PMID- 11884248 TI - Reporting of measures of accuracy in systematic reviews of diagnostic literature AB - Abstract | Background | There are a variety of ways in which accuracy of clinical tests can be summarised in systematic reviews. Variation in reporting of summary measures has only been assessed in a small survey restricted to meta-analyses of screening studies found in a single database. Therefore, we performed this study to assess the measures of accuracy used for reporting results of primary studies as well as their meta-analysis in systematic reviews of test accuracy studies. Methods | Relevant reviews on test accuracy were selected from the Database of Abstracts of Reviews of Effectiveness (1994 --2000), which electronically searches seven bibliographic databases and manually searches key resources. The structured abstracts of these reviews were screened and information on accuracy measures was extracted from the full texts of 90 relevant reviews, 60 of which used meta-analysis. Results | Sensitivity or specificity was used for reporting the results of primary studies in 65/90 (72%) reviews, predictive values in 26/90 (28%), and likelihood ratios in 20/90 (22%). For meta-analysis, pooled sensitivity or specificity was used in 35/60 (58%) reviews, pooled predictive values in 11/60 (18%), pooled likelihood ratios in 13/60 (22%), and pooled diagnostic odds ratio in 5/60 (8%). Summary ROC was used in 44/60 (73%) of the meta-analyses. There were no significant differences in measures of test accuracy among reviews published earlier (1994 --97) and those published later (1998 --2000). Conclusions | There is considerable variation in ways of reporting and summarising results of test accuracy studies in systematic reviews. There is a need for consensus about the best ways of reporting results of test accuracy studies in reviews. Keywords: Background : The manner in which accuracy of clinical tests is mathematically summarised in the biomedical literature has important implications for clinicians. Appropriate accuracy measures would be expected to sensibly convey the meaning of the study results with scientifically robust statistics without exaggerating or underestimating the clinical significance of the findings. Lack of use of appropriate measures may lead authors of primary accuracy studies to draw biased conclusions. In systematic reviews of test accuracy literature, there are many ways of synthesising results from several studies, not all of which are considered to be scientifically robust. For example, measures such as sensitivity and specificity commonly used in primary studies are not considered suitable for pooling separately in meta-analysis. Variations in reporting of summary accuracy and use of inappropriate summary statistics may increase the risk of misinterpretation of clinical value of tests. A recent study evaluated a small sample of meta-analytical reviews of screening tests to demonstrate the variety of approaches used to quantitatively summarise accuracy results. This study confined itself to a limited Medline search. It exclusively examined meta-analytical studies so reviews not using quantitative synthesis were excluded. It did not look at accuracy measures used to report results of primary studies separately from those used for meta-analyses. In order to address these issues, we undertook a comprehensive search to survey systematic reviews (with and without meta-analysis) of test accuracy literature to assess the measures used for reporting results of included primary studies as well as their quantitative synthesis. Methods : We manually searched for relevant reviews in the Database of Abstracts of Reviews of Effectiveness (DARE). In order to limit the impact of human error inherent in manual searching, we complemented it with electronic searching. DARE was searched electronically with word variants of relevant terms (diagnostic, screening, test, likelihood ratio, sensitivity, specificity, positive and negative predictive value) combined using OR. From 1994 to 2000 DARE has identified 1897 reviews of different types by regular electronic searching of several bibliographic databases, hand searching of key major medical journals, and by scanning grey literature (search strategy and selection criteria can be found at ). The structured abstracts of these reviews were screened independently by the authors to identity systematic reviews of test accuracy. The full texts were obtained of those abstracts judged to be potentially relevant. Reviews addressing test development and diagnostic effectiveness or cost effectiveness were excluded. Any disagreements about review selection were resolved by consensus. Information from each of the selected reviews was extracted for the measures of test accuracy used to report the results of the primary studies included in the review. If a meta-analysis was conducted, information was also extracted for the summary accuracy measures. The various accuracy measures are shown in Table . We sought the following in the primary studies: sensitivity or specificity, predictive values, likelihood ratios and diagnostic odds ratio. For meta-analysis, we sought the summary measures pooling the above results and summary receiver operating characteristics (ROC) plot or values. All extracted data were double-checked. We divided the reviews into two groups arbitrarily according to time of publication; one group covering the period 1994 --97 (50 reviews) and another covering 1998 --2000 (40 reviews). This allowed us to assess whether there were any significant differences in measures being used to report test accuracy results among reviews published earlier and those published later. As the approaches to summarising results are not mutually exclusive, we evaluated and reported the most commonly used measures and their most common combinations. We used chi-squared statistical test for comparison of differences between proportions. Table 1 | Measures of accuracy of dichotomous test results Results : Of the abstracts available in DARE, 150 were considered to be potentially relevant. Excluding reviews that addressed test development and diagnostic effectiveness or cost, 90 reviews of test accuracy were left for inclusion in our survey. There were 45 reviews of dichotomous test results, 42 reviews of continuous results dichotomised by the original authors, and 3 reviews that contained both result types. Meta-analysis was used in 60/90 (67 %) reviews, 50 in 1994 --97 and 40 in 1998 --2000. (See : BMC_IncludedRefList_04032002 for a complete listing of the 90 reviews included in our study). As shown in Table , sensitivity or specificity was used for reporting the results of primary studies in 65/90 (72%) reviews, predictive values in 26/90 (28%), and likelihood ratios in 20/90 (22%). For meta-analysis, independently pooled sensitivity or specificity was used in 35/60 (58%) reviews, pooled predictive values in 11/60 (18%), pooled likelihood ratios in 13/60 (22%), and pooled diagnostic odds ratio in 5/60 (8%). Summary ROC was used in 44/60 (73%) of the meta-analyses. There were no significant differences between reviews published earlier and those published later as shown in Table . Table 2 | Measures of test accuracy reported in review of diagnostic literature (1994 --2000) Discussion : Our study showed that sensitivity and specificity remain in frequent use, both for primary studies and for meta-analyses over the time period surveyed. Sensitivity and specificity are considered inappropriate for meta-analyses, as they do not behave independently when they are pooled from various primary studies to generate separate averages. In our survey, separate pooling of sensitivities or specificity was used frequently in meta-analyses where summary ROC would have been more appropriate. . Our findings about reporting of summary accuracy measures in meta-analyses are different to those reported previously. We found a higher rate of use of summary ROC, though use of independent summaries of sensitivity, specificity and predictive values were similar. These differences may be due to differences in searching strategies (databases and time frames) and selection criteria. Our search was more recent and comprehensive, using DARE, which has covered seven different databases (Medline, CINAHL, BIOSIS, Allied and Alternative Medicine, ERIC, Current Contents clinical medicine and PsycLIT), and hand-searched 68 peer-reviewed journals and publications from 33 health technology assessment centres around the world since February 1994. Moreover, as we did not restrict our selection to meta-analytical reviews only, we were able to examine reviews summarising accuracy results of primary studies without quantitative synthesis, which constituted 33% (30/90) of our sample. Therefore, compared to the previous publication on this topic, our survey provided a broader and more up-to-date overview of the state of reporting of accuracy measure in test accuracy reviews. Conclusions : The use of inappropriate accuracy measures has the potential to bias judgement about the value of tests. Of the various approaches to reporting accuracy of dichotomous test results, likelihood ratios are considered to be more clinically powerful than sensitivities or specificities. Crucially, it has been empirically shown that authors of primary studies may overstate the value of tests in the absence of likelihood ratios. There is also evidence that readers themselves may misinterpret test accuracy measures following publication. It is conceivable that the problem of inconsistent usage of test accuracy measures in published reviews, as found in our survey, may contribute to misinterpretation by clinical readership. The reason for variation in reported accuracy measures may, in part, be attributed to a lack of consensus regarding the best ways to summarise test results. It is worth noting that despite authoritative publications about appropriate summary accuracy measures in the past, (we have only quoted a few references) inconsistent and inappropriate use of summary measures has remained prevalent in the period 1994 --2000. Our paper highlights the need for consensus to support change in this field of research. Competing interest : None declared Backmatter: PMID- 11914161 TI - A randomised controlled trial of a patient based Diabetes recall and Management system: the DREAM trial: A study protocol [ISRCTN32042030] AB - Abstract | Background | Whilst there is broad agreement on what constitutes high quality health care for people with diabetes, there is little consensus on the most efficient way of delivering it. Structured recall systems can improve the quality of care but the systems evaluated to date have been of limited sophistication and the evaluations have been carried out in small numbers of relatively unrepresentative settings. Hartlepool, Easington and Stockton currently operate a computerised diabetes register which has to date produced improvements in the quality of care but performance has now plateaued leaving substantial scope for further improvement. This study will evaluate the effectiveness and efficiency of an area wide 'extended' system incorporating a full structured recall and management system, actively involving patients and including clinical management prompts to primary care clinicians based on locally-adapted evidence based guidelines. Methods | The study design is a two-armed cluster randomised controlled trial of 61 practices incorporating evaluations of the effectiveness of the system, its economic impact and its impact on patient wellbeing and functioning. Keywords: Background : Delivering care to people with diabetes | There is broad, international agreement over what constitutes high quality health care for people with diabetes . This will be enshrined in a National Service Framework for people with diabetes, due in summer 2002. However, in the face of poor current performance the most efficient method of delivering care remains unclear . Following a 1994 systematic literature review suggesting structured care improved patient care, an editorial in the British Medical Journal concluded that more evaluative research was needed before widespread adoption of any of the models could be recommended . A subsequent systematic review of routine surveillance of patients with diabetes by Griffin and Kinmonth concluded "Computerised central recall, with prompting for patients and their family doctors, can achieve standards of care as good or better than hospital outpatient care, at least in the short term. The evidence supports provision of regular prompted recall and review of people with diabetes by willing general practitioners and demonstrates that this can be achieved, if suitable organisation is in place'. However, the evidence base on which these conclusions are based is limited in several ways. Firstly there are only five randomised controlled trials (RCTs) involving 1058 patients. All of these studies are 'patient randomised" trials, thus potentially under-estimating the effectiveness of the intervention (see Study Design). They were all evaluating more or less selected patients and general practices and none of them were explicitly evaluating a UK National Health Service (NHS) service area wide intervention. Only one of the four UK based studies evaluated patient based outcomes and included an economic assessment and this study only involved patients from three general practices . Thus, the effectiveness of an area wide, patient focussed, structured recall and management system (in terms of process of care, patient outcome and economic impact) remains unknown. The current system | The current computerised diabetes management system runs in Hartlepool, Easington and Stockton, three Primary Care Group (PCG) areas, in the Northern and Yorkshire Region. It was introduced to all 36 general practices in Hartlepool and Easington Districts in mid-1995. Stockton (25 practices) agreed to join the system in 1999 and it was operational there by October 2000. There are three key components to the current system: 1. A central register of patients with diabetes. 2. A structured minimum dataset to be completed and returned to the central register. 3. The provision of both patient specific and aggregated data to both patients and clinicians. The system (developed by Westman Medical Software) allows three methods of collection of data at each contact with a patient with diabetes who is registered on the database. Two methods use a standard form completed by clinicians to collect data concordant with the UK minimum data set . Within secondary care, forms are completed at every new patient or annual review. In primary care, forms are completed by the practice nurse (usually) or general practitioner, either opportunistically or at practice diabetic clinics. In both cases, the completed data forms are sent to the Diabetes Register Facilitator for data entry. Thirdly, the hospital laboratory provides a monthly download of laboratory test details (e.g. HbA1c). A patient can be identified as having diabetes and added to the register by any permutation of one or more of these three routes. Feedback of individual patients' data, including review status, is provided to general practices quarterly. This feedback is 'passive' in that it does not explicitly prompt either patients or doctors as to required actions. Audit packages within the software can audit on every variable collected. District wide audit is provided on anonymised aggregated data; individual practice audits (with comparisons to other practices) are provided to participating practices at least annually. Feedback of the data to the patient (for hospital patients only) is by a patient information sheet and to the GP as a standardised letter. A Diabetes Register Facilitator co-ordinates and updates the register. A steering group composed of GH, the Diabetes Register Facilitator and representatives of the PCGs and patients, oversees the register and deals with issues such as confidentiality. Impact of the system to date | Measures of the impact of the system to date relate only to Hartlepool, Easington and Stockton. The main impact on patient registration was in its first 12 to 18 months of operation: during 1995, 747 patients were registered on the system (0.4% prevalence) which had increased to 3867 (1.8% prevalence) by the end of 1996. The increase in registration has stabilised since then, reaching 4324 (2% prevalence) by 1999. During 1999, 70% of registered patients attended a clinic; 52% had their feet examined and 51% had their eyes examined. Seventy three per cent had an HbA1c result recorded and 69% a blood pressure measurement. These figures are similar to those reported by other centres using the same system . The need for an extended system | Recording of clinical measures increased during the first few years of operation of the system but began to plateau more recently (for example, 50% of patients had an HbA1c recorded during 1996, compared to 60% in 1997 and 63% in 1998). This plateauing of performance has been reported by others . We believe that this is due to a lack of coordination (patients being lost to follow up) and lack of prompting of clinicians to deliver appropriate clinical interventions. Furthermore, given that most patients with diabetes are primarily seen in primary care the greatest potential impact is from optimising and extending the system in primary care. In order to address these shortcomings the additional key components, over and above those already in the system, will be: 1. Locally adapted evidence based guidelines for the management and follow up of patients with diabetes. 2. Automated prompts to patients and primary care clinicians that a review consultation is necessary. 3. A structured management sheet (including patient specific management suggestions based on (1)). 4. An enhanced monitoring system to follow up reasons for non-attendance from both patients and clinicians and to re-schedule appointments, based on nonreturn of a completed management sheet. 5. Patient feedback for patients in primary care. There is some limited supportive trial evidence for these developments, although the existing studies involved small sample sizes and may not be generalisable to the NHS . In evaluating the system with these extended features this study will also address the design shortcomings of previous studies of shared care in diabetes . It will be tailored to each practice, PCG defined areas will be studied, rather than an unrepresentative sample of general practices; and the system will be transparent and replicable in other areas. Methods : Design of the study | The study design is a pragmatic two-arm cluster randomised controlled trial. The unit of randomisation will be the general practice. Simple patient randomised trials are rightly considered the most robust method of assessing most health care innovations . This design, however, cannot be regarded as the gold standard for evaluating systematic approaches to chronic disease management, an essentially behavioural field of research . If both intervention and control patients were to be cared for within the same practice there is the risk that the management of control patients would be influenced by the practitioners knowledge of the care of intervention patients. This would result in an underestimation of the effect of the intervention . Therefore, practices rather than patients are the appropriate unit of randomisation and analysis. As the current system has been in place for different lengths of time within the three participating PCGs, we will stratify the randomisation by PCG. Randomisation will be performed by a statistician independent of the research team using computer generated numbers to avoid allocation bias . Study setting and recruitment of practices | The study will be based in the general practices of the three PCGs of Easington, Hartlepool and Stockton. Since the recent merger of Hartlepool and North Tees Acute Trusts all three PCGs are now exclusively served by one secondary care diabetes service (and thus the one diabetes register). GH is the lead clinician for diabetes services in the new Trust. The 61 general practices in the three PCGs constitute the target practices for the study and we will attempt to recruit all practices. The PCG diabetes leads or the PCG clinical governance leads in all three PCGs have provided letters confirming their support for the project. We do not envisage major difficulties with recruitment, given the need to agree local guidelines as part of the process involved in the Trust merger, the likely requirements in the forthcoming National Service Framework for diabetes, and the 100% practice coverage with the current diabetes system. We will (through the PCGs) write to all practices, giving information about the project to the senior partner or diabetes lead and practice manager of practices. Practices will be invited to opt out if they do not wish to be included in the study -- this is an approach we have used successfully before. The PCG diabetes lead, clinical governance lead and GH will be co-signatories of this letter. If practices do decline we will collect data on characteristics of non-participating practices to assess the impact on the generalisability of the trial's findings. Finally, if there are significant problems with recruitment, there are other practices which could be approached in a nearby PCG (South Tyneside) which uses the same software for its diabetes register. Details of the intervention | Local guidelines and management prompts | A guideline development group will be established to develop local guidelines for the management of diabetes, based upon available evidence based guidelines (Scottish Intercollegiate Guidelines Network (SIGN, 1996, 1997a, 1997b, 1997c), and Effective Care Bulletins . They will also use the forthcoming national diabetes guidelines as these become available. The group will be multidisciplinary and contain primary and secondary care doctors and nurses, patients and the Diabetes Register Facilitator . The group will define review periods for specified patient groups (e.g. patients with diabetes satisfactorily controlled on diet alone should be reviewed every 12 months), referral criteria for patients moving from primary to secondary care and back and simple decision rules for the management prompts. These would be of two types. The first would prompt for actions to be performed and only require their performance to be documented (e.g. asking for a foot examination to be performed in a patient who does not have a recorded foot examination). The second would be more complex and suggest alterations to clinical management on the basis of data in the database (e.g. patients with persistently raised blood pressure should have their anti-hypertensive medication increased). These decision rules will be integrated into the recall and management system. Running the system | The proposed enhancements to the system are designed to require the primary care team to perform no additional work over and above the current configuration. The current database has a patient identifier, a minimum dataset and retrieval systems to support the structured recall of patients. Westman Medical Software has agreed to amend the system as required. A 'circle of information exchange' will be established between the participating general practices and the database. The local guidelines will be used to adapt the current centralised database, along with the practices' preferred method of following up patients (for example, within consultations in routine surgeries or within special clinics). The central database system will identify when patients are due for review (based upon the local guidelines) and will generate a letter to the patient asking them to make an appointment for a review consultation. Patient information or educational materials could be included with the letter. At the same time, the central database will generate a letter to the practice stating that the patient should be making a review appointment in the near future. The letter to the practice will include a management sheet (to be held in the patient's record) to capture an agreed minimum data set to be collected during the consultation. This management sheet will also contain the relevant prompts (as described above). When the patient is seen in the practice, the primary care professional (currently this is usually done by the practice nurse) will complete the management sheet and return a copy for entry onto the central register within a designated period of time. This circle of information is broken if the patient does not visit the general practice as planned or the general practice does not return the management sheet to the central register. If this happens, the central register would alert the Diabetes Register Facilitator who will ascertain the reason for failure and take appropriate action, (e.g. send a reminder to the patient, prompt the practice to return the management sheet). A range of educational activities will be provided for intervention practices, as part of the usual local structures for contact with practices, with some additions, These will include: distribution of information about the trial in local newsletters; meetings with practice clinical governance leads; evening meetings for practice nurses (with small group discussion of the practical implications for intervention practices); and a telephone meeting with the practice diabetes lead (usually the practice nurse) in each intervention practice. Practices in the control arm will continue to receive the recall system as currently configured. Logistical considerations | From the prevalence of patients with diabetes on the current register, there will be about 7500 patients on the system if 61 practices are recruited. Half of these will be in intervention practices. On current patterns of usage, we anticipate there being the need for 1.5 recalls per annum per patient on the register, resulting in about 6000 recalls per year for the intervention group. Assuming a 40 week working year, the system will need to dispatch, receive and process about 150 forms per group per week. Identification of patients | Patients for the structured recall and management system are already identified on the Hartlepool and North Tees database. As some practices have children registered on the system, who are under the care of an exclusively secondary care adolescent service, an age limit of 18 years or over will be set for inclusion. Practices will be asked to check lists of their patients on the database regularly throughout the study. The central database will remove patients from the recall system who are known to have died or moved away. Patient consent | Patients have already consented, or are being consented, to their data being held within the current diabetes register. The study will involve no extra 'routine' data being collected, and this data will be anonymised before being sent for analysis; all data held for analysis will be held in accordance with the Data Protection Act. For the patient-based questionnaire study, we will seek additional patient consent to complete one survey. The three relevant Local Research Ethics Committees have approved the trial. Data collection | The main study outcome measures will be rates of performance of process of care and the patient based measures of functional and psychosocial wellbeing. Data will be collected for 15 months after the start of the intervention. Fifteen months was chosen to allow for patients who are reviewed every 12 months but fail to attend on initial invitation. Process of care variables | Process of care variables will be collected via the computerised database. The exact data to be collected will be determined by both the current content of the database and the guidelines but will include such data items as rates of attendance at clinics and annual reviews, conduct of eye and feet examinations, performance of investigations and prescribing. We will also collect data on clinical measures (e.g. HbA1c, and blood pressure levels). Outcome of care measures | Outcome of care data will be collected, by postal questionnaire, 15 months after commencement of the study. A portfolio of validated and responsive generic and disease specific instruments will be used to measure functional and psychosocial variables that will be potentially influenced by the intervention. These will include: i) The SF36 health status profile which we will use to generate Mental (MCS) and Physical Component Summary Scales (PCS) . ii) The Newcastle Diabetes Symptoms Questionnaire . iii) The Bradley Treatment Satisfaction Questionnaire . Patient costs questions will be developed by the study health economist. We have successfully used such packages of questionnaires within trials before and have achieved response rates in excess of 70% in similar surveys in this region. . Sample size considerations | On the basis of previous work we have made the following assumptions. The mean number of patients per practice for whom we will be able to collect process data will be 30 and the ICC (a measure of the lack of independence of responses from patients from the same practice) calculated from our local data is 0.14 for measures of process (whether a blood pressure measurement and whether an HbA1c measurement has been recorded in a 12 month period). Standard methods for determining the sample size requirements for a cluster randomised trial indicate that we need 60 practices to detect a difference of 15% (42.5% v 57.5%) with 80% power assuming a significance level of 5%. Assessment of outcome of care will be based on health status scales such as the SF-36. Previous work has shown that this type of intervention is likely to produce an effect size of approximately 0.25 in such measures and that the ICCs for such measures will be approximately 0.07. The most efficient study design (that minimises the number of patients required) is one that makes use of all the available practices. A sample of 27 patients from each of 61 practices will give us 85% power to detect an effect size of 0.25 assuming a significance level of 5%. With a predicted response rate of approximately 70% (based on our experience in the COGENT study ) after two reminders, our starting sample size will need to be 2379 patients (approximately 39 patients per practice). Principles of data analysis | Analysis will be by intention to treat. Multilevel modelling (using the MlwiN package ) will be used to take into account the clustering of patients within practices . Both binary variables (when a process was undertaken or not) and continuous variables (such as the physical health component of the SF-36) can be analysed using these techniques. For both types of variable, variation between practices will be fitted as a random effect and the difference between intervention and control practices will be fitted as a fixed effect. In the case of binary variables, a logit link function will be used. Economic evaluation | The economic impact of implementing the new structured recall and management system will be evaluated in terms of the marginal costs of adapting and running the system; the costs of developing and disseminating the guidelines; the educational activities for intervention practices; the implications for the use of health care services; and the costs to the patients and their carers. The benefits will be measured as described earlier on in the clinical study. The estimation of health service resource use will relate to diabetes-specific clinical visits, tests, investigations, and procedures. This data will be routinely collected as part of the management system implementation and subsequent costing, using health service pay and price data, will be undertaken using a mixed approach based on micro-costing and gross-costing methods . Use of drugs, referrals to secondary care and the impact of the intervention on the change of use of patients' and their carers' time will also be monitored through postal questionnaires at the end of the follow-up period. A sensitivity analysis will be undertaken to test the robustness of the results to the uncertainty not related to sampling variations and to enhance the generalisability of the results . We are aware that the costs of the system might be balanced only in the longer term against the cost savings related to averted complications . However, the assessment of the benefits in terms of final outcomes (e.g lives saved, or QALYs) and long term costs is beyond the objective of the present study. Competing interests : None declared Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 11914164 TI - Inter-rater agreement in the scoring of abstracts submitted to a primary care research conference AB - Abstract | Background | Checklists for peer review aim to guide referees when assessing the quality of papers, but little evidence exists on the extent to which referees agree when evaluating the same paper. The aim of this study was to investigate agreement on dimensions of a checklist between two referees when evaluating abstracts submitted for a primary care conference. Methods | Anonymised abstracts were scored using a structured assessment comprising seven categories. Between one (poor) and four (excellent) marks were awarded for each category, giving a maximum possible score of 28 marks. Every abstract was assessed independently by two referees and agreement measured using intraclass correlation coefficients. Mean total scores of abstracts accepted and rejected for the meeting were compared using an unpaired t test. Results | Of 52 abstracts, agreement between reviewers was greater for three components relating to study design (adjusted intraclass correlation coefficients 0.40 to 0.45) compared to four components relating to more subjective elements such as the importance of the study and likelihood of provoking discussion (0.01 to 0.25). Mean score for accepted abstracts was significantly greater than those that were rejected (17.4 versus 14.6, 95% CI for difference 1.3 to 4.1, p = 0.0003). Conclusions | The findings suggest that inclusion of subjective components in a review checklist may result in greater disagreement between reviewers. However in terms of overall quality scores, abstracts accepted for the meeting were rated significantly higher than those that were rejected. Keywords: Background : Interest in the peer review process and research aimed at determining the method of obtaining the best quality reviews has grown in recent years. Checklists have been developed that aim to guide reviewers when assessing the quality of papers, but little evidence exists concerning the extent of agreement between two referees when evaluating the same paper. In addition, little is known about which dimensions of a checklist are likely to result in greater agreement between referees. There were two aims of this study: (1) to examine inter-rater agreement of the quality of abstracts submitted to a primary care research conference (Annual Meeting of the South West Association of University Departments of General Practice, Exeter 2000, UK), and (2) to compare the scores of abstracts accepted and rejected for the meeting. Materials and Methods : Abstracts were anonymised and scored using a structured assessment comprising seven categories: (1) importance of the topic (2) originality (3) overall quality of the study design (4) appropriateness of the design used (5) achievement of aim (6) contribution to academic primary care (7) likelihood of provoking discussion. For comparison purposes, we have classified the assessment of categories 1, 2, 6 and 7 as more 'subjective' in nature, and categories 3, 4 and 5 as more 'objective'. Between one (poor) and four (excellent) marks were awarded for each category, giving a maximum possible score of 28 marks. Every abstract was assessed independently by two referees (AM and AG). Agreement between referees was assessed using intraclass correlation coefficients (ICC), a chance corrected measure of agreement. The ICC indicates perfect agreement only if the two assessments are numerically equal and is preferable to the more usual (Pearson) correlation coefficient. The crude ICC is lowered by any systematic differences between referees' scores. In terms of a plot of the two referees' scores, a line with a non-zero intercept will further lower the ICC irrespective of any disagreement, represented by deviation of the slope of the line away from unity and scatter around the line. In a further analysis, this effect was investigated by subtracting the mean difference for each component from the higher of the two referees' scores. The ICCs were then recalculated, giving estimates of agreement corrected for both systematic differences and chance. There are no universally applicable standard values for the ICC that represent adequate agreement, but the following convention is used here to aid interpretation: ICC <0.20 'slight agreement'; 0.21 --0.40 'fair agreement'; 0.41 --0.60 'moderate agreement'; 0.61 --0.80 'substantial agreement'; >0.80 'almost perfect agreement'. Scores from referees from three different institutions were summed to give each abstract an overall score. Abstracts were ranked by this overall score and the top 45 were accepted for oral presentation at the meeting. Of the 52 abstracts refereed by AM and AG, mean total scores of those accepted and rejected for the meeting were compared using an unpaired t test. Results : Chance corrected agreement between the two referees' scores measured using crude ICCs was greater for the three components relating to design and execution of the study (Table : items 3 to 5) compared to those relating to more subjective elements of the abstract (Table : items 1, 2, 6, 7). After adjustment for systematic differences in referees' scores, ICCs for items 3 to 5 remained highest, demonstrating fair to moderate agreement. Table 1 | Inter rater agreement between two referees for 52 abstracts submitted for a primary care research conference A total of 76 abstracts were submitted for the meeting. Of 52 received by the authors for assessment, 26 were accepted for oral presentation . Abstracts accepted for the meeting had a significantly higher mean score than those that were rejected (95% CI for difference 1.3 to 4.1, p = 0.0003) . Table 2 | Summary statistics of abstracts accepted and rejected for oral presentation at a primary care research conference Discussion : This study has shown that when using a structured assessment form, two independent reviewers were more likely to agree on design or methodological components of a checklist than on subjective components of abstracts submitted for an annual research meeting. Abstracts accepted for the meeting had significantly higher total scores, but overlapped considerably with rejected abstracts. This was due to acceptance for the meeting being determined by an overall aggregate of scores awarded by referees from three institutions. While the subject of inter-reviewer agreement on different components of a checklist is relatively under-researched, some previous studies offer support for our finding that agreement is better when reviewers can be more objective in their assessments. Among a group of reviewers asked to rate a series of review articles, agreement on scientific quality of the papers was very high (60% of ICCs > 0.7) both within and between groups with varying levels of research training and expertise. All 10 dimensions of the checklist that reviewers rated could be regarded as objective. Divergent reviewers have been identified in a study comparing an overall rating score that indicated a recommendation to publish rather than individual dimensions of a review checklist. This study does have limitations. Importantly, we assessed agreement between only two reviewers on a relatively small number of abstracts. This could be addressed by having more abstracts assessed by a greater number of reviewers. However the study was conducted pragmatically within the time and administrative constraints of a small annual scientific meeting rather than submissions to a journal over an extended period. Another limitation is that the reviewer checklist was constructed prior to conceiving the study. If future meetings are to be used to investigate the content of structured reviewer assessments, such checklists should be constructed with specific hypotheses in mind. Characteristics associated with good peer review are age under 40 years and training in epidemiology or statistics, characteristics that applied to both reviewers in the present study. Structured assessment forms that ask the reviewer for their opinion of a paper's interest, originality or likelihood of provoking discussion may be more likely to result in scores that reflect the reviewer's own research interests. This is not necessarily a criticism -- it is perhaps only natural that individuals will differ in their opinions of how interesting they find, and think others will find, a particular paper. It is interesting that the two components with the lowest agreement, importance of the topic and originality of the study, both require more knowledge about a specific subject area than either of the other two subjective questions. Journal editors and meeting organisers should be aware that including subjective components in review checklists may result in greater disagreement between reviews. Conclusions : This study provides some evidence that inclusion of subjective components in a review checklist may result in greater disagreement between reviewers. An interesting area for further research would be to investigate the effects of attaching different weights to subjective and objective components of a checklist, or to exclude subjective components altogether from overall quality scores and simply use them a guide to acceptance or rejection. Competing interests : None declared. Figure 1 | Difference between referees' scores versus mean score Difference between referees' scores versus mean score Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 11943069 TI - The Caenorhabditis elegans Y87G2A.14 Nudix hydrolase is a peroxisomal coenzyme A diphosphatase AB - Abstract | Background | The number of Nudix hydrolase family members varies widely among different organisms. In order to understand the reasons for the particular spectrum possessed by a given organism, the substrate specificity and function of different family members must be established. Results | The Y87G2A.14 Nudix hydrolase gene product of Caenorhabditis elegans has been expressed as a thioredoxin fusion protein in Escherichia coli and shown to be a CoA diphosphatase with catalytic activity towards CoA and its derivatives. The products of CoA hydrolysis were 3',5'-ADP and 4'-phosphopantetheine with Km and kcat values of 220 muM and 13.8 s-1 respectively. CoA esters yielded 3',5'-ADP and the corresponding acyl-phosphopantetheine. Activity was optimal at pH 9.5 with 5 mM Mg2+ and fluoride was inhibitory with a Ki of 3 muM. The Y87G2A.14 gene product has a potential C-terminal tripeptide PTS1 peroxisomal targeting signal -- SKI. By fusing a Y87G2A.14 cDNA to the C-terminus of yeast-enhanced green fluorescent protein, the enzyme appeared to be targeted to peroxisomes by the SKI signal when transfected into yeast cells. Deletion of SKI abolished specific targeting. Conclusions | The presence of related sequences with potential PTS1 or PTS2 peroxisomal targeting signals in other organisms suggests a conserved peroxisomal function for the CoA diphosphatase members of this group of Nudix hydrolases. Keywords: Background : The Nudix hydrolase family comprises enzymes that hydrolyse predominantly the diphosphate (pyrophosphate) linkage in a variety of nucleoside triphosphates, dinucleoside polyphosphates, nucleotide sugars and nucleotide cofactors having the general structure of a nucleoside diphosphate linked to another moiety, X . They are found in archaea, eubacteria, animal, plants, and fungi and all possess the Nudix box sequence signature motif Gx5Ex5 [UA]xREx2EExGU (where U is an aliphatic hydrophobic amino acid) . The proposed functions of this family are to eliminate potentially toxic nucleotide metabolites from the cell and to regulate the concentrations of nucleotide cofactors and signalling molecules for optimal cell growth and survival. The number of genes encoding Nudix hydrolases varies widely, from zero in Mycoplasma genitalium to 22 in Deinococcus radiodurans. This variation presumably reflects the growth or environmental adaptability, stress tolerance and metabolic capacity of the different organisms. The Nudix hydrolases thus offer an ideal system with which to study the evolution of a largely inessential protein family and its contribution to the individual biology of an organism. Understanding such variation requires a combination of detailed biochemical, genetic and cellular studies to reveal the individual functions of family members within the set in any given system. In the case of multicellular eukaryotes, the nematode Caenorhabditis elegans offers a genetically amenable model system with which to carry out such studies. There are 11 members of the Nudix hydrolase family in C. elegans. So far only two of these have been characterized -- a diadenosine tetraphosphate pyrophosphohydrolase (the orthologue of human NUDT2) and an NADH diphosphatase . Sequence comparisons would predict the existence of an ADP-sugar diphosphatase (NUDT5 orthologue) , an ADP-ribose diphosphatase (NUDT9 orthologue) , a diphosphoinositol polyphosphate pyrophosphohydrolase (NUDT3/4 orthologue) , two probable coenzyme A diphosphatases, one of which is highly similar to the mouse Nudt7 CoA diphosphatase , and 4 proteins of unknown function, including one with a strong similarity to the Saccharomyces cerevisiae PSU1/DCP2 protein and another similar to the developmentally-regulated mouse RP2 protein . Recent characterization of the S. cerevisiae NADH and CoA diphosphatases and the mouse Nudt7 CoA diphosphatase has revealed that they are located in peroxisomes. The function of these peroxisomal enzymes may be to regulate the concentration of these essential nucleotide cofactors for peroxisomal metabolism or, by analogy with the E. coli MutT 8-oxo-dGTPase, to eliminate toxic modified cofactor metabolites from the highly oxidizing peroxisomal environment. In order to investigate these possibilities in the C. elegans model system, we have cloned and characterised the putative C. elegans Y87G2A.14 CoA diphosphatase and shown that it displays the expected enzymatic activities and that it appears to be targetted to peroxisomes by a C-terminal PTS1 targeting signal. Results and discussion : Cloning, expression and purification of Y87G2A.14 | The C. elegans Y87G2A.14 gene encodes a 234 amino acid protein with an expected molecular weight of 26,601 Da. It was amplified by PCR from a C. elegans cDNA library. The PCR fragment was inserted into the pET-32b(+) expression vector and the nucleotide sequence of the insert was determined to be exactly the same as that submitted to GenBank under accession no. CAB54476. The recombinant plasmid pETY87G2A.14 was then used to transform E. coli BL21 (DE3) cells to generate a His-tagged thioredoxin fusion protein with an expected molecular mass of 43,731 Da. When the Trx-Y87G2A.14 fusion protein was expressed at 37C, it was confined to inclusion bodies, so the induction temperature was decreased to 25C to enhance protein solubility. As the expression level was low at this temperature, the induction time was increased to 8 h. These conditions markedly increased the solubility of Trx-Y87G2A.14 which was then purified from the soluble fraction (Fig , lane 2) to apparent homogeneity on NiCAMTM-HC resin (Fig , lane 3). To determine the molecular weight of the Y87G2A.14 protein itself, the Trx-Y87G2A.14 fusion was cleaved with thrombin, which generated Y87G2A.14 with an apparent molecular weight of 27 kDa (expected molecular weight, 29,807 Da) and thioredoxin (15 kDa, Fig , lane 4). Figure 1 | Purification and cleavage of Trx-Y87G2A.14 fusion protein. Purification and cleavage of Trx-Y87G2A.14 fusion protein. Samples were analysed by SDS-PAGE (15% polyacrylamide) and stained with Coomassie Brilliant blue R 250. Lane 1, 2 mug protein standards: bovine serum albumin (66 kDa), ovalbumin (45 kDa), glyceraldehyde 3-phosphate dehydrogenase (36 kDa), carbonic anhydrase (29 kDa), trypsinogen (24 kDa), soybean trypsin inhibitor (20 kDa) and alpha-lactalbumin (14.2 kDa); lane 2, soluble cell extract of BL21 (DE3) cells transformed with recombinant plasmid pETY87G2A.14 and induced with 1 mM IPTG for 8 hours at 25C before applying to a column of NiCAMTM-HC resin ; lane 3, 3 mug purified Trx-Y87G2A.14 fusion protein; lane 4, 3 mug purified Trx-Y87G2A.14 fusion protein after cleavage with thrombin. Substrate specificity and product analysis | Purified Trx-Y87G2A.14 was inactive towards the following nucleotides when assayed at a fixed concentration of 0.5 mM: NADH, NAD+, NDP-sugars, 5'-(d)NTPs, 5'-NDPs, 5'-NMPs and diadenosine polyphosphates. High activity was found with CoA and its derivatives. HPLC analysis of CoA hydrolysis by Trx-Y87G2A.14 showed that the enzyme was a CoA diphosphatase, cleaving the diphosphate linkage in CoA to yield adenosine 3',5'-bisphosphate (3',5'-ADP) and 4'-phosphopantetheine . Figure 2 | Identification of reaction products of CoA hydrolysis. Identification of reaction products of CoA hydrolysis. Reaction mixtures containing 0.5 mM CoA were incubated at 37C for 20 min with or without 0.1 mug Trx-Y87G2A.14 fusion protein and the products separated by HPLC as described in Materials and methods. Without enzyme (------), with enzyme , gradient ( --- --- --- --- ---). Positions of authentic standards are indicated. Reaction requirements and kinetic parameters | Trx-Y87G2A.14 displayed optimal activity with 0.5 mM CoA as a substrate at pH 9.5. A divalent metal ion was absolutely required for activity, with optimal activity at 5 mM MgCl2. In common with all other Nudix hydrolases tested, fluoride was a strong inhibitor with a Ki value of approximately 3 muM (results not shown). Km, and kcat values for CoA, CoA esters and oxidized CoA were calculated by non-linear regression from data obtained by HPLC analysis . A graphical example of the data for CoA in the form of a hyperbolic plot and double reciprocal plot show that the enzyme obeys simple Michaelis-Menten kinetics. The kcat / Km ratios show that the enzyme prefers reduced forms of CoA to oxidized CoA with CoA itself the best substrate of those tested. Figure 3 | Lineweaver-Burk and Michaelis-Menten (inset) plots for the hydrolysis of CoA. Lineweaver-Burk and Michaelis-Menten (inset) plots for the hydrolysis of CoA. Reaction mixtures containing various concentrations of CoA (0.05 --0.7 mM) were incubated at 37C for up to 20 min with 0.1 mug Trx-Y87G2A.14 fusion protein. Initial rates of hydrolysis were determined after separation of the products by HPLC as described in Materials and methods. Table 1 | Kinetic parameters for the hydrolysis of CoA and CoA derivatives by Trx-Y87G2A.14 fusion protein Subcellular localization | Y87G2A.14 has the C-terminal tripeptide sequence SKI. This conforms to the pattern typical of PTS1 peroxisomal targeting signals found in many peroxisomal matrix proteins, suggesting that Y87G2A.14 may be targeted to these organelles . However, possession of a potential PTS1 sequence is not always sufficient on its own to result in peroxisomal targeting and other elements of the protein sequence may also be involved. Since targeting of animal peroxisomal proteins expressed in yeast has often been observed, yeast cells were transformed with expression plasmids encoding C-terminal or N-terminal fusions of Y87G2A.14 to yeast-enhanced green fluorescent protein (yEGFP) in order to determine the subcellular location of Y87G2A.14. The cells were then examined by confocal microscopy. Cells transformed with pY87G2A.14-yEGFP, in which the C-terminus of Y87G2A.14 is fused to the N-terminus of yEGFP, showed a diffuse, cytoplasmic fluorescence with no clear subcellular localization . In contrast, cells transformed with pyEGFP-Y87G2A.14, in which the C-terminal tripeptide SKI is free to act as a targeting signal showed the clear punctate fluorescence that is indicative of yeast peroxisomes . The identity of SKI as the targeting signal was confirmed by transformation of cells with pyEGFP-Y87G2A.14Delta SKI, in which the C-terminal tripeptide was deleted during construction. This again showed a diffuse, cytoplasmic, fluorescence . Together, these results strongly suggest that Y87G2A.14 is targeted to peroxisomes by its C-terminal tripeptide, SKI. Figure 4 | Subcellular localization of Y87G2A.14 by fluorescence confocal microscopy. Subcellular localization of Y87G2A.14 by fluorescence confocal microscopy. yEGFP fluorescence of yeast cells transformed with (a) pY87G2A.14-yEGFP; (b) pyEGFP-Y87G2A.14 and (c) pyEGF-Y87G2A.14DeltaSKI Conclusions : On the basis of its sequence, the C. elegans Y87G2A.14 gene product was predicted to be a peroxisomal coenzyme A diphosphatase. In addition to the Nudix motif, Y87G2A.14 possesses the PROSITE UPF0035 motif , which we have previously suggested confers a specificity for coenzyme A and its derivatives , and a C-terminal tripeptide, SKI, that conforms to the pattern typical of PTS1 peroxisomal targeting signals. The experiments described here confirm these predictions. Fig shows a multiple sequence alignment of the motif-containing region of Y87G2A.14 with related sequences from other organisms. Those marked with a tick have been experimentally shown to be coenzyme A diphosphatases . In most cases, higher organisms possess two related sequences, e.g. mouse Nudt7 and Nudt8, one of which encodes a peroxisomal enzyme (e.g. Nudt7). However, S. cerevisiae has only one sequence containing the UPF0035 motif while Arabidopsis thaliana has three, and the second of the two Drosophila melanogaster sequences, RH61317, is currently only represented in GenBank by a single expressed sequence tag, so its status is still questionable. For the peroxisomal enzymes, either a putative C-terminal PTS1 or an N-terminal PTS2 targeting signal is present. Interestingly, in each case, the putative PTS2 signal is contained within or near a predicted mitochondrial targeting or chloroplast transit peptide sequence , suggesting a possible dual location for these proteins. Such a possibility has not yet been experimentally observed; however, mutation of a glutamate five residues to the C-terminal side of the PTS2 of rat peroxisomal 3-ketoacyl-CoA thiolase to a neutral or basic amino acid has been shown to result in partial mitochondrial targeting, suggesting that the negative charge on glutamate may normally block translocation to the mitochondria . Whether or not a system exists in vivo to regulate dual targeting is clearly a topic requiring further investigation. The non-peroxisomal sequences provide no clear indication of possible subcellular location, hence they are likely to be cytoplasmic. Given the existence of mitochondrial, peroxisomal and cytoplasmic pools of CoA and CoA esters , it would not be surprising to find CoA diphosphatase activity in all these locations. However, the precise substrate specificities of the "cytoplasmic" activities remain to be determined. Figure 5 | Partial sequence alignment of Y87G2A.14 and related sequences. Partial sequence alignment of Y87G2A.14 and related sequences. The partial sequence of Y87G2A.14 containing the UPF0035 and Nudix motifs (arrowed) was aligned using the Clustal W program with related sequences from other organisms retrieved from a BLAST search. Organisms and database accession numbers are: Caenorhabditis elegans Y38A8.1, Q23236; Homo sapiens NUDT7, XP_058753; H. sapiens NUDT8, AI743601; Mus musculus Nudt7, Q99P30; M. musculus Nudt8, AK009700; Drosophila melanogaster CG11095, Q9VY79; D. melanogaster RH61317, BI631687; Schizosaccharomyces pombe YDH5, Q92350; S. pombe YDZA, 013717; Ambidopsis thaliana At2g33980, 022951; A. thaliana At1g28960, Q9SHQ7; A. thaliana At5g45940, BAB09322; Saccharomyces cerevisiae PCD1, Q 12524; Escherichia coli YeaB, P43337; Deinococcus radiodurans DR1184, Q9RV46. Sequences encoding experimentally confirmed CoA diphosphatases are marked with a tick. Columns on the right indicate whether the full sequence contains a putative peroxisomal targeting signal (PTS1 or PTS2) and/or a putative mitochondrial targeting peptide (mTP) or chloroplast transit peptide (cTP). Regarding the possible function of these enzymes in general, and the C. elegans peroxisomal enzyme in particular, a recent functional genomic analysis by RNA-mediated interference of C. elegans chromosome I, on which the Y87G2A.14 gene is located, revealed no phenotype in relation to growth, survival, fecundity or morphology when the expression of Y87G2A.14 was ablated . This would indicate that, within the limitations of RNAi, the CoA diphosphatase activity of Y87G2A.14 is not essential. However, now that the biochemical properties of this protein have been established, a more detailed and targeted biochemical analysis can be undertaken that should reveal its cellular function and benefit to the organism. Nudix hydrolases are believed to regulate the concentrations of nucleotides for optimal cell performance and also to eliminate potentially toxic nucleotide metabolites from the cell. With regard to regulation, CoA diphosphatase activity is associated with the 400 kDa CoA synthesizing protein complex from S. cerevisiae, in which it forms part of an alternative pathway for CoA biosynthesis that differs from the principal route of 3'-dephospho-CoA and CoA synthesis by this complex . This CoA/4'-phosphopantetheine cycle involves hydrolysis of CoA to 3',5'-ADP and 4'-phosphopantetheine, which then reacts with ATP to give 3'-dephospho-CoA then CoA. Whether such a pathway operates in peroxisomes and whether the C. elegans Y87G2A.14 protein is involved remain to be established. With regard to the elimination of toxic nucleotide metabolites, the 13-fold higher kcat / Km ratio for oxidized CoA (CoASSCoA) compared to CoA for the S. cerevisiae PCD1 CoA diphosphatase previously suggested to us that this enzyme might preferentially remove non-functional and potentially toxic oxidized CoA and CoA esters from within the oxidizing environment of the peroxisomes . However, neither the mouse Nudt7 nor the C. elegans Y87G2A.14 proteins show this preference. Nevertheless, the potential production of adenine ring-oxidized derivatives of CoA by reactive oxygen species generated in the peroxisomes analogous to the 2-oxo-dATP and 8-oxo-dATP substrates of the mammalian MTH1 Nudix hydrolase suggests that such species could be more relevant substrates for peroxisomal CoA diphosphatases in vivo. The amenability of C. elegans to studies of cellular and molecular stress will allow the question of the biological function of these enzymes to be addressed. Materials and methods : S. cerevisiae strain BY4741 (MAT a; his3D1; leu2D0; met15D0; ura3D0) was from Research Genetics. Calf intestinal alkaline phosphatase, yeast inorganic pyrophosphatase, EcoR1 and BamH1 were from Roche while BspH1 (Pag1) was from Helena Biosciences. Pfu DNA polymerase was from Stratagene. All other chemicals and nucleotides were from BDH or Sigma. The E. coli expression vector pET-32b(+) was from Novagen and the yeast-enhanced green fluorescent protein (yEGFP) fusion vectors pUG35 and pUG36 were a gift from J.H. Hegemann, Institute of Microbiology, University of Dusseldorf, Germany. The C. elegans cDNA library was prepared from adult nematodes by H. M. Abdelghany, School of Biological Sciences, University of Liverpool, U.K. Cloning of Y87G2A.14 from C.elegans | A cDNA corresponding to the C. elegans Y87G2A.14 gene on chromosome 1 (GenBank accession no. CAB54476) was amplified from a cDNA library by PCR using as forward and reverse primers 5' GCAAATCATGAAGTGTGTGGTTAGCCGAGCTG 3' and 5' TAAATGAATTCACTAAATTTTGGATTTCGGTTC 3' respectively. These primers provided a BspH1 restriction site at the start of the amplified gene and an EcoR1 site at the end. After amplification with Pfu DNA polymerase, the DNA was recovered by phenol/chloroform extraction and digested with BspH1 and EcoR1. The digest was gel-purified and the restriction fragment ligated into the Nco1 and EcoR1 restriction sites of pET-32b(+) as both BspH1 and Nco1 form compatible ends with each other. The resulting construct, pETY87G2A.14, yielded Y87G2A.14 downstream of the 109-amino acid thioredoxin (Trx) fusion and His-tag and S-tag sequences under the control of an IPTG-inducible promoter. The structure of the insert was confirmed by sequencing. The construct was propagated by transformation of E. coli XL1-Blue cells. Expression of Y87G2A.14 in E. coli and protein purification | E. coli strain BL21(DE3) was transformed with pETY87G2A.14. A single colony was picked from an LB agar plate containing 50 mug/ml ampicillin and inoculated into 10 ml LB medium containing 50 mug/ml ampicillin and incubated at 37C. When the cells reached an A600 of 0.5, they were transferred to 1 litre of fresh LB medium containing 50 mug/ml ampicillin and grown to an A600 of 0.3 at 37C, then transferred to an incubator at 25C. Isopropyl-1-thio-beta-D-galactopyranoside (IPTG) was added to 1 mM at an A600 of 0.8, and the cells induced for 8 h. The induced cells (4 g) were harvested, washed and resuspended in 20 ml breakage buffer (50 mM Tris-HCl, pH 8.0, 50 mM NaCl, 1 mM DTT). The cell suspension was sonicated and the cell lysate was then cleared by centrifugation at 10,000 x g at 4C for 15 min. The supernatant was applied to a 15 x 50 mm column of NiCAMTM-HC resin (Sigma) equilibrated with 50 mM Tris-HCl, pH 8.0, 0.5 M NaCl, 1 mM 2-mercaptoethanol at a flow rate of 0.5 ml/min. After eluting the unbound proteins, a linear gradient of 0 --40 mM histidine in the same equilibration buffer was applied at flow rate of 1 ml/min and 1 ml fractions collected and analysed by SDS-PAGE. Those containing pure Trx-Y87G2A.14 fusion protein were collected, dialysed overnight at 4C against 1 litre of 20 mM Tris-HCl, pH 8.0, 50 mM NaCl, ImM DTT and then concentrated by ultraflltration (Amicon) and stored at -20C in 50% glycerol. Enzyme assays | Potential substrates were screened by measuring the Pi released after nucleotide hydrolysis in presence of inorganic pyrophosphatase or alkaline phosphatase . The standard assay (200 mul) for phosphodiester substrates was incubated at 37C for 30 min and contained 50 mM l,3-bis [tris(hydroxymethyl)-methylamino]propane-HCl (BisTrisPropane-HCl), pH 8.0, 5 mM MgCl2, 1 mM DTT, 0.5 mM substrate, 0.1 mug of Trx-Y87G2A.14 fusion protein and 0.5 mug (1 unit) alkaline phosphatase. Assays with phosphomonoester substrates were as above, except 0.5 mug (100 mU) inorganic pyrophosphatase was used instead of alkaline phosphatase. The Pi released in each case was measured colorimetrically. Chromatographic analysis | Kinetic parameters and reaction products generated from hydrolysis of CoA and its derivatives were measured by high performance anion-exchange chromatography. The reaction mixtures (100 mul) contained 50 mM BisTrisPropane-HCl, pH 9.5, 5 mM MgCl2, 1 mM DTT (in cases of substrates requiring reducing conditions), substrate in the range of 0.05 --0.7 mM, (0.1 --1 mM in the case of oxidized CoA) and were incubated at 37C for up to 20 min (during which time the reaction rates remained linear) with 0.1 mug Trx-Y87G2A.14 fusion protein. A 90 mul sample of each reaction mixture was applied to a 1 ml Resource-Q column (Amersham Pharmacia Biotech) equilibrated with 0.045 M CH3COONH4 (pH 4.6, adjusted with H3PO4), and eluted with a linear gradient from 0 to 0.45 M NaH2PO4 (pH 2.7 adjusted with CH3COOH) for 10 min at a flow rate of 2 ml/min . Elution was monitored at 259 nm and peaks identified with the aid of standards and quantified by area integration. GFP fusion constructs and subcellular localization | Expression plasmids encoding C-terminal and N-terminal fusions of Y87G2A.14 to yeast-enhanced green fluorescent protein (yEGFP) were constructed by amplification of the coding region of Y87G2A.14 from C. elegans cDNA by PCR using the same forward primer 5' CGACGGATCCATGAAGTGTGT 3' and one of the reverse primers 5' TAAATGAATTCACTAAATTTTGGATTTCGGTTC 3', 5' CACTAAGAATTCTATTTCGGTTCAAATTTCCTACTTGC 3', or 5' GCTCGAATGAATTCAATTTTGGATTTCGGTTC 3' to give PCR products "C", "CDeltaSKI" or "N" respectively. These primers provided a BamH1 restriction site at the start of the amplified gene and EcoR1 sites at the end. PCR products "C" and "CDeltaSKI" were cloned as C-terminal fusion proteins to yEGFP, while PCR product "N" with a deletion of the Y87G2A.14 termination codon was cloned as an N-terminal fusion to yEGFP. After amplification with Pfu DNA polymerase, the DNA products were recovered by phenol/chloroform extraction and digested with BamH1 and EcoR1. The digested PCR products "C", and "CDeltaSKI" were gel purified and the restriction fragments ligated between the BamH1 and EcoR1 restriction sites of pUG36 (yEGFP-C-fusion) to give pyEGFP-Y87G2A.14 and pyEGFP-Y87G2A.14DeltaSKI respectively. The digested PCR product "N" was gel purified and the restriction fragment ligated between the BamH1 and EcoR1 restriction sites of pUG35 (yEGFP-N-fusion) to give pY87G2A.14-yEGFP. The structures of the inserts were confirmed by sequencing. The plasmids were propagated by transformation of E. coli XL 1-Blue cells. For microscopy, S. cerevisiae strain BY4741 was transformed with pyEGFP-Y87G2A.14, pyEGFP-Y87G2A.14DeltaSKI or pY87G2A.14-yEGFP and grown on solid SC-Ura medium containing 2% glucose. Cells were viewed by conventional and confocal fluorescent microscopy on a Zeiss LSM510 confocal microscope with a 100 x 1.4 NA objective. Other methods | Protein concentrations were estimated by the Coomassie blue binding dye-based colorimetric method using equal weights of bovine serum albumin, conalbumin, cytochrome c and myoglobin as standards . Backmatter: PMID- 11914163 TI - Outcomes research in the development and evaluation of practice guidelines AB - Abstract | Background | Practice guidelines have been developed in response to the observation that variations exist in clinical medicine that are not related to variations in the clinical presentation and severity of the disease. Despite their widespread use, however, practice guideline evaluation lacks a rigorous scientific methodology to support its development and application. Discussion | Firstly, we review the major epidemiological foundations of practice guideline development. Secondly, we propose a chronic disease epidemiological model in which practice patterns are viewed as the exposure and outcomes of interest such as quality or cost are viewed as the disease. Sources of selection, information, confounding and temporal trend bias are identified and discussed. Summary | The proposed methodological framework for outcomes research to evaluate practice guidelines reflects the selection, information and confounding biases inherent in its observational nature which must be accounted for in both the design and the analysis phases of any outcomes research study. Keywords: Background : The development of practice guidelines | In clinical medicine, variations exist that do not appear to be related to variations in the clinical presentation and severity of disease . In response, practice guidelines have been developed in an attempt to reduce the wide practice variations and, through this process, to increase the appropriateness and quality of medical care and to reduce health care costs . Despite the publication and dissemination of practice guidelines , there has been relatively little evaluation of the application and impact of clinical practice guidelines . Some of the difficulty in the evaluation of these guidelines relates to the methods that were used to develop them . Guidelines have often have been developed before adequate data have been available to assess the relationship between clinical practice patterns and desired clinical outcomes. Nevertheless, there have been some reviews of practice guideline evaluation . While epidemiological designs are commonly used to evaluate the effectiveness of health care interventions, never has this been discussed in the context of outcomes research. We propose the use of a methodological framework for outcomes research to evaluate practice guidelines. Methodological issues with the measurement of practice variations | In the debate about reasons to promote the development of practice guidelines, few have questioned whether the variations are real, or alternatively, whether they are simply a function of methodological flaws in the measurement of medical practices themselves, the result of variations in practice patterns across groups of patients with a similar diagnosis, or both. Furthermore, few studies have addressed whether practice variations, in fact, lead to outcome variations. Finally, little attention has been paid to the identification and measurement of initial conditions, that is, the potentially confounding factors and effect modifiers of the practice patterns outcomes relationship. Measurement of practice pattern variation | The measurement of medical practice patterns is susceptible to error. Measurement error may affect the validity of medical practice measurement in three major ways. First, it may lead to selection bias, in that subjects are selected to belong to a certain group based on an erroneous diagnosis. Secondly, it may lead to misclassification of exposure (information bias), in that patients treated with a specific practice pattern are classified in the wrong diagnostic group. Thirdly, it may lead to misclassification of outcomes, in that patients with a given outcome are classified in the wrong diagnostic group. Potential problems with the measurement of practice variations relate to the mechanisms that underlie the choice of groups that are compared in studies of practice variations. These mechanisms must be defined clearly to minimize selection bias. In many studies of practice variations, populations are arbitrarily divided according to hospitals, regions, counties, or countries. Little information is available about the factors that lead these groups to go to a particular hospital, live in a particular region, go to a particular doctor, etc. The population base from which each comparison group is derived should, in principle, be quite similar for all groups. Basically, if the groups are drawn from a similar population, unmeasurable and potentially confounding variables are more likely to be equally distributed between groups. In addition, the measurement of practice variations cannot be valid without information on relevant "initial conditions". Initial conditions are all confounding factors and effect modifiers, other than the treatment/practice patterns, that may cause or influence the clinical outcomes of interest. These factors may explain practice variations among groups that do not share similar initial conditions. To evaluate practice patterns-outcomes associations, potential confounders must be identified and controlled for in the analysis. Aside from clinical presentation and severity of illness; the initial conditions to be identified and characterized as completely as possible include physician, patient, and practice environment factors . Measurement of such factors is essential to minimize the chance of a systematic error following confounding biases and effect modification . Figure 1 | Table 1 | Initial conditions to be taken into account when making inferences about practice patterns-outcomes associations Identification and measurement of outcomes of interest | Limitations to the development and evaluation of practice guidelines also include the absence of a clear concept of the targeted outcomes and the paucity of outcomes data to support these guidelines . There appears to be only a weak relationship between the purpose of guidelines and many of the outcomes usually measured in clinical research, that is, the source of evidence for guideline development (evidence-based). The initial goals of establishing practice guidelines -- to reduce costs and enhance the quality and appropriateness of treatment -- are, in fact, rarely the basis for guideline development, since little data is available for these outcomes. To some degree, the development of guidelines has been driven by the availability of data on clinical outcomes, such as morbidity and mortality, rather than those outcomes related to the primary goals of the guidelines. The evaluation of practice guidelines | Throughout the development of practice guidelines, the major deficiency has been the lack of an evaluative method . Thus, we suggest a methodological framework for outcomes research to be applied to evaluate practice guidelines. Outcomes research evaluates practice patterns as they occur in actual clinical settings. This type of research can describe practice patterns, evaluate their divergence from practice guidelines and determine the effect of practice variations on outcomes. Outcomes research is necessarily observational in nature and, although observational studies have been used to evaluate health care interventions, the proposed methodological framework has yet to be applied to outcomes research. Why should outcomes research be used to evaluate and validate practice guidelines? The primary goal of practice guidelines is the consistent adherence by physicians to practice patterns that achieve the "best" outcomes at the lowest cost. Outcomes research evaluates practice patterns as they occur in actual clinical settings, and is thus the logical method to evaluate practice guidelines. In fact, outcomes research and practice guidelines are connected through concepts that relate to efficacy and effectiveness research . Efficacy studies, which normally complement practice guideline development, are those performed in highly selected groups of patients to investigate if a particular intervention works under controlled conditions set by the study investigators. In contrast, outcomes research evaluates practice as it occurs in actual clinical settings . Research in these settings is called effectiveness research because the investigators have limited control over the conditions that qualify the practice settings. The difference between efficacy and effectiveness research can be summarized as follows: does it work at all (efficacy) or does it work in the real world (effectiveness)? Thus, there exists a dynamic process in which evidence from both effectiveness and efficacy studies feeds into the development and evaluation of practice guidelines, as depicted in Figure . Figure 2 | Relationship between outcomes research and practice guidelines Relationship between outcomes research and practice guidelines Most practice guidelines are derived from efficacy studies rather than effectiveness studies. Therefore, it is not surprising that practice guidelines are not fully applicable in actual clinical practice. We suggest that effectiveness studies be used not only as a method to evaluate practice guidelines but also as a basis for their development. These could include both observational studies and effectiveness trials. Outcomes research better reflects practice in the real world and may make guidelines more likely to be applied. However, to date, little attention has been paid to the epidemiological underpinnings of the methods used to conduct outcomes research. Discussion : We will first propose a methodological framework for outcomes research. Then, we will show how it can be used to evaluate practice guidelines. Finally, we will address the limitations of the proposed methodological framework. Generic epidemiological issues in outcomes research | In the proposed methodological framework, the generic issues related to outcomes research will be discussed in sequential order. In outcomes research, the first step is to identify the study population and the groups (hospitals, providers, regions, etc.) that will be compared. The next step is the measurement of practice patterns and outcomes. After groups are compared on the basis of the treatment they receive and outcomes of interest, associations are sought between practice patterns and the various measures of outcome. This step of the methodological framework raises issues of confounding bias because not all factors that can confound these associations are measured and controlled or even known. The presence or absence of confounding bias can be affected by the other sources of bias namely selection and information biases. Lastly, we discuss the issue of temporal trends. In the evaluation of practice guidelines, the measurement of practice patterns may not be contemporaneous with the publication of practice guidelines. This may explain and even lead to the frequently observed discrepancy between the actual practice and what the guidelines state that it should be. Finally, two particularities of outcomes research 1) the presence of ecological exposures in individual level studies and 2) the common use of large administrative databases are discussed. Specification of the model | Definition of the elements of the proposed epidemiological model for outcomes research | In the proposed model for outcomes research designed to evaluate practice guidelines, the outcome of interest can be a disease . For example, if the practice patterns that are being studied pertain to coronary revascularization, complications such as mortality and reinfarction after acute myocardial infarction may constitute the outcome of interest. Finally, the consequences of different practice patterns on medical resources (cost, quality and appropriateness) may be another possible outcome of interest. Table 2 | Epidemiological model for outcomes research to evaluate practice guidelines In the studies of outcome research, practice patterns, (which constitute the exposure in the proposed model), range from the use of medication, diagnostic tests and therapeutic procedures to the length of hospital stay, transfer to other facilities and/or scheduled physicians visits. The primary goal of outcomes research is the evaluation of the effects of the selected practice patterns on the outcomes of interest. Consequently, any inference made about this association must be evaluated as a function of the potential selection, information (measurement error) and confounding biases. A limitation of outcomes research as it is most often performed is the lack of attention given to the measurement of each of the elements of the epidemiological model shown in Table 3. The basis of the proposed methodological framework will be the identification of generic sources of potential bias that relate to each element of the proposed model. Selection bias | Since outcomes research is observational in nature, the choice of the study population and of the compared groups is highly susceptible to selection bias. As applied to outcomes research, selection bias is defined as a distortion in the estimate of the practice patterns outcomes association due to the way that subjects are selected for inclusion in the study population and in the different groups to be compared . A major consequence of selection bias is the potential confounding of inferences made about practice patterns-outcomes associations. This occurs when some characteristics of the subjects related to practice patterns or clinical outcomes influence the selection or exclusion of individual subjects, groups of subjects or practice environments. The selection process should be such that patients included in the study population come from the same target population . Furthermore, patients or study members should have a similar probability of being selected and included in the actual population. Inclusion and exclusion criteria must be clearly defined in order to characterize the actual population as precisely as possible. Judging the internal validity of a study is more feasible when there is a detailed account of how the individuals were selected to become members of the actual population. Finally, the study population, also needs to be carefully characterized so that the inferences derived from the analysis of the study population can be evaluated for both internal validity (based on the data analyzed in the study) and external validity (the extent to which results obtained from the data analyzed in a particular study can be generalized to populations outside of the study). Any systematic differences between those actually studied and the source (target) population could result in biased estimates of the impact of a practice pattern on a clinical outcome. In many studies of outcomes research, groups exposed to different practice patterns are compared. The identification of such groups of patients is sought to assess the impact of different practice patterns on various outcomes in actual clinical settings and, as previously mentioned, can be used to assess practice guidelines. Because of such study design, it becomes unclear as to what the target population precisely is. Is it the group (the set of patients in a given environment) or is it the individuals receiving the various practice patterns within each group? For example, in a study of regional variations in the treatment of acute myocardial infarction in the U.S., the treatment of patients (practice patterns) was compared across different regions of the U.S. In this study, one wishes to generalize the findings about practice patterns-outcomes associations to all individuals with acute myocardial infarction (individual level). One also wishes to generalize the effect of the exposure, which is in this case practice patterns, to those prevalent in a given region (ecological level). The presence of these two levels, the individual and the ecological levels, introduces an added level of complexity in terms of the assessment of the effect of the exposure on outcome. When comparing practice patterns across regions using individual data, there is a certain degree of correlation brought about by the clustering of practice patterns that needs to be taken into account. Such a correlation is very difficult to quantify. In contrast, when assessing the effect of the exposure at the individual level, there are ecological factors (initial conditions particular to a given region) that need to be taken into account. The data originating from studies with mixed design, which are often the design of outcomes research studies, need to be analyzed with special attention to the degree of correlation between the individual covariates and to the presence of ecological exposure variables. Another potential source of selection bias is the choice of the groups to be compared, which depends on the criteria used to divide the groups. Individuals included in the groups to be compared should have the same probability of being included in these groups. Not infrequently in outcomes research, geographic criteria (such as country, regions, hospitals) are used because such criteria allow the identification of clinically comparable groups that receive very different treatments, whose resulting outcomes can then be assessed. However, such a process must be scrutinized for the possibility of selection bias other than the treatments that are being evaluated. Such selection bias would make groups not comparable as to clinical and other factors that could affect outcomes. The presence of a biased selection process could lead to confounding bias when practice patterns-outcomes associations are assessed. Such a situation may occur when the study groups are not comparable with regard to some characteristics of the subjects related to practice patterns or clinical outcomes that influenced the selection or exclusion of individual subjects, groups of subjects or practice environments. For example, in the same study of regional variations in the treatment of acute myocardial infarction, census regions of the U.S. were arbitrarily chosen as a basis for comparison. In this example, patients with similar risk of developing the outcome of interest, which is defined here as a complication after acute myocardial infarction, may not have had the same probability of being included in the different groups to be compared. Confounders may then bias the practice patterns/outcomes association if the selection of different risk groups is related to practice patterns. Selection bias can also affect the assessment of outcomes. Potential sources of this bias include loss to follow-up or missing data. Follow-up data is difficult to obtain in outcomes research studies, which often rely on administrative databases for data acquisition. Linkage, either of different databases or of the same database over time, is often performed . A failure to link the databases for a number of individuals presents a problem equivalent to having data missing for these individuals. Information bias | The second step in outcomes research studies is the measurement of practice patterns and of the outcomes of interest. Here, issues of information bias must be considered. Information bias can be defined as a distortion of the potential practice patterns outcomes association due to misclassification of subjects with regard to practice patterns, outcome measures or both, or due to measurement error . There are two major ways in which practice patterns can be misclassified. They relate to the sensitivity and specificity of the tests that are used for the diagnosis for which practice patterns are being evaluated and for the classification of the outcomes of interest. The measurement of the different practice patterns and their related outcomes largely depend on the identification of a group of patients who have a given diagnosis and require a given treatment. The characteristics that make a diagnosis more amenable to outcomes research are the following: 1) a precise diagnostic definition, 2) a diagnostic test with high sensitivity and specificity, 3) reproducibility among different individuals and locations, 4) easily coded, 5) related to a procedure, and 6) common and costly, so that it is likely to be collected in large, administrative databases frequently used in outcomes research. Because of such requirements, only a limited number of clinical conditions are amenable to outcomes research. Acute myocardial infarction is an example of a diagnosis that can be made with a high level of certainty because it has a precise diagnostic definition and well-defined diagnostic criteria, which, when taken together, have high sensitivity and specificity for the correct classification of patients. Therefore, it is easy to identify a study population that, in fact, has this disease and to describe their treatment. Thus, in order to minimize the misclassification of relevant practice patterns, the methods used to classify the disease and the outcomes that relate to the practice patterns under investigation must have high sensitivity and specificity . Given the principles underlying the measurement of practice patterns and outcomes, how are the measurements generally made in outcomes research studies? The measurement of the exposure (practice patterns) in outcomes research is valid only if it corresponds to the "true" practice as performed in the clinical setting. Again, practice can only be "true" if the diagnosis is correct. The identification of both patients with the disease of interest and their treatment requires a source of information that has the features of a diagnostic test. In outcomes research, administrative databases are often used as an information source to identify a study population and to obtain data on exposure. The database coding of diagnoses and procedures can be used as a "diagnostic test" to identify the clinical condition for which practice patterns will be described and to classify the practice patterns themselves and the outcomes of interest. Such a "diagnostic test" will have higher sensitivity and specificity values for some diagnoses than for others. For example, administrative database coding will have higher sensitivity and specificity for procedure-related diagnoses (such as hip fracture) because the diagnostic code is related to a major operation and is likely to be recorded for administrative purposes. In contrast, a diagnostic criterion for osteoarthritis can be quite vague and administrative coding is likely to have very low sensitivity and specificity for this diagnosis. The use of databases as a diagnostic test must be validated in all outcomes research studies, especially those using administrative databases. Methods to validate these databases include chart reviews, a priori coding systems or both. These validation methods ensure that coding is as accurate and reproducible as possible, thus allowing the database to be used as a diagnostic test to identify the study population and the practice patterns and the outcomes in outcomes research. However, these validation methods are rarely used. Finally, appropriate measures of outcomes that will serve to evaluate practice guidelines must be identified. This presents a problem because most practice guidelines aim to reduce practice variations, which will, in turn, lead to improved appropriateness and quality of care. However, how appropriateness and quality of care are measured is controversial and will not be discussed here . Nevertheless, defining the outcomes that will be used to evaluate practice guidelines is a crucial step in this process. Quality of life and functional status measures constitute another group of outcome measures that should be included for the evaluation of practice guidelines. These dimensions of outcomes have received more attention from health providers, while consumers have become more concerned about outcomes of care. However, these outcomes also are difficult to measure, because they rely heavily on patient interviews and questionnaires. They are likely to vary with patient expectations, culture, and climate and are thus potentially to be measured with error and be misclassified. A few reliable, valid instruments have been developed to assess health-related quality of life , but such instruments are not easily used to collect this information from large databases. There is a need to develop instruments to measure these types of outcomes, whether they are conversion factors for existing databases (such using length of stay as a proxy for cost) or new measures that could easily be integrated in administrative databases. Such measures could include estimates of functional class or severity of illness. At present, many outcomes research studies measure mortality and disease-specific morbidity. The validity of the measurement of these outcomes is limited by the type of database that is used. For example, using death registries to obtain causes for death is a notoriously invalid source for this type of information. There are many examples of poor correlation between cause of death as established by death registries versus disease registries. Death certificates in New York City during 1992 were assessed to determine the accuracy and frequency of reporting tuberculosis as a cause of death. Of 310 persons who died with active tuberculosis in 1992 (based on a disease-specific registry), only 34% had tuberculosis listed on their death certificate. Thus, in this example, as in many others like it, using death certificates led to an inaccurate measure of disease burden . Confounding bias | In outcomes research terms, confounding bias is present when the effect of the practice variations on the outcomes of interest is distorted because of the effects of extraneous variables (variables that are causally associated with the practice variations and the outcomes of interest) . This issue is crucial in outcomes research because, while outcomes research shares the purpose of a clinical trial (to evaluate different treatments), it primarily uses observational methods -- investigators conducting outcomes research have limited control over potentially confounding factors (the initial conditions of individual groups of patients). Because outcomes research builds on existing practice variations and analyses the natural ongoing experiment, there is ample opportunity for confounding bias to invalidate any inference made about practice patterns outcome associations . For example, variations in practice patterns could reflect variation not only in the use of a given procedure but also in the severity of disease. Assignment of patients to certain procedures on the basis of the severity of illness makes sense clinically, but in outcomes research, it is a common and important source of confounding if the procedure is either efficacious or particularly harmful in high-risk patients. Many indices have been developed to measure the severity of illness when using existing databases to correct for such confounding, but one can never be sure that this type of confounding has been entirely controlled . This presents an intrinsic limitation of outcomes research. Avoidance of confounding bias is limited by the source of data used to describe practice patterns, particularly when observational data, such as the large Medicare administrative databases, are used to compare outcomes among patients who receive different treatments. The potential for confounding bias arises because many factors other than the treatment under evaluation may affect patient outcomes. These factors include comorbid diseases, severity of illness, and patient, physician and environmental factors. Such factors are likely to influence treatment decisions but are difficult to capture fully in recorded data. Researchers cannot adjust for imbalances in prognostic factors that are unmeasured or poorly categorized and administrative data, in particular, may lack the precise and accurate coverage of clinical details needed to permit full and fair adjustments. Further data collection might solve this issue, but it is not always possible to collect additional information. Standard statistical modeling can attempt to adjust for the known differences between the groups, but this might not be sufficient for unmeasured differences. Several alternative methods have been suggested. One method is subgroup analysis to adjust for unmeasured differences between groups of individuals who differ on known risk factors. Another method consists of the use of instrumental variables . Instrumental variables are observable factors that influence treatments but do not directly affect patient outcomes. This approach uses the so-called instrumental variables to mimic a randomization of patients to different likelihoods of receiving alternative treatments. McClellan et al. applied this methodology to assess whether more aggressive use of invasive cardiac procedures improved outcomes in the elderly. In this study, the instrumental variable was the distance of the patient's residence from the nearest hospital with on-site angiography. The authors noted lower mortality among elderly individuals who received more aggressive treatment than among those treated more conservatively. Temporal trend bias | We propose a bias called a "temporal trend bias" that is particular to the use of outcomes research to evaluate practice guidelines. This bias results from the inability to control for secular trends. It reflects the fact that by the time practice guidelines are published and disseminated, new treatments and technology are being incorporated into clinical practice. Thus, it is difficult to identify a pure application of a practice guideline whose application is not undermined by recent advances in medicine and technology. For example, we evaluated the effect of a specific set of guidelines on return to work after acute myocardial infarction. The use of these guidelines had been successful in a university setting; this study assessed their use in a community setting. During the 5 years that elapsed between these two studies, practices changed. The use of guidelines was less successful in the community not only because they did not influence practice but also because usual care had grown closer to the proposed guidelines . Ecological exposure in individual level studies | A frequently encountered particularity of outcomes research study design is the presence of both ecological exposure and individual level covariates in the same analysis. Because the unit of analysis is a group, but inferences are made about the impact of a given practice pattern on individual outcomes, many outcomes research analyses have elements of both individual and ecological analyses . In our study of regional variations in the treatment of acute myocardial infarction, measures describing practice patterns at the regional level, ecological exposure, (proportion of patients receiving angiography, angioplasty, and coronary artery bypass surgery) were linked to the outcome measures of mortality adjusting for individual level variables that measured severity of disease. Then, inferences were made about the use of these procedures at the patient level. Although the unit of analysis is the region, which would demand an ecological analysis, there are individual level covariates, which are likely to be correlated within each region, that need to be taken into account. When group measures are used that contain individual-level variability with some degree of correlatedness (within region) and aggregate-level variability (between regions), specific analytic tools must be used. It has been suggested that hierarchical logistic regression modeling be used to examine the interplay between sources of variation in the use of health-care services, that is, between ecological-level and individual-level sources. This type of modeling is designed to separate true variability across areas from observed variability. An application of this method is the work by Gatsonis et al. who found that practice variations across regions of the U.S. in the use of angiography after acute myocardial infarction were largely explained by differences in patient characteristics and geographic region. However, states that had more on-site availability of angiography still tended to have higher angiography rates after accounting for between-region and within-region variability. After analysis for sources of variability, more reliable inferences about the associations between practice patterns and outcomes can be made. Sources of data | The application of the proposed methodological framework for outcomes research largely depends on the sources of data that are used to evaluate the effect of the practice variations on outcomes . Most commonly, the study design is a retrospective cohort analysis and the dataset that is used has been obtained either for administrative purposes (discharge databases) or for a randomized clinical trial that addressed a different question . Less often, a prospective cohort study is designed to evaluate a particular set of practice guidelines . Although a prospective design provides more control in data collection than a retrospective analysis, both designs are subject to selection, information and confounding biases. The ideal database to use for the evaluation of practice guidelines is one that allows the precise measurement of the practice patterns (exposure) and outcomes (disease) as well as the measurement of potential confounders (severity of illness, precision of diagnosis, socioeconomic characteristics). Unfortunately, such a database probably does not exist. The strength of administrative databases, such as that of Medicare is that they allow the observation of large numbers of patients for which practice patterns can be evaluated as they occur in actual clinical practice. Furthermore, administrative databases allow the observation of practice patterns outcomes associations in large numbers of unselected patients. However, the limitations of such databases include the missing information about potential confounding factors, such as severity of illness, and the limited ability to measure exposure and outcome accurately. Many databases that are not designed for clinical research either mismeasure patient outcomes or fail to capture outcomes that are important to both physicians and patients (such as quality of life and functional status). The control of these biases was the basis of the methodological framework for outcomes research proposed in this chapter. The application of outcomes research methods to practice guideline evaluation | The application of outcomes research methods to practice guideline evaluation can accomplish several goals. One important goal is the evaluation of practice guidelines, that is, to determine to what extent the guidelines accomplished their primary goals after their dissemination. We have suggested the model of chronic disease epidemiology as the methodological framework for outcomes research to evaluate practice guidelines. The steps to evaluate practice guidelines using outcomes research when the basic design is a retrospective cohort study are summarized in Figure Some limitations to the application of this model exist. The reasons for the inability of the proposed methodological framework to deal completely with the intrinsic biases in outcomes research are listed in Figure . They relate mostly to the databases usually used in studies of outcomes research. Figure 3 | Steps to evaluate practice guidelines using outcomes research Steps to evaluate practice guidelines using outcomes research Figure 4 | Reasons for the inability of the proposed methodological framework to deal with biases in outcomes research Reasons for the inability of the proposed methodological framework to deal with biases in outcomes research Summary : The proposed methodological framework for outcomes research to evaluate practice guidelines reflects the selection, information and confounding biases inherent in its observational nature which must be accounted for in both the design and the analysis phases of any outcomes research study. Indeed, a major limitation of outcomes research is the inability to account for unobserved heterogeneity that directly correlates with practice patterns and/or health outcomes. This may lend bias to any inferences made about practice variations and outcomes. "Researchers cannot correct for the subtle reason doctors choose one treatment over another for a particular patient. That bias, in turn, can undermine the entire premise of outcomes research" . These are intrinsic properties of outcomes research that can be dealt with only in part, by applying the principles of chronic disease epidemiology. Thus, this proposed methodology can serve as a framework for the conduct of outcomes research in the evaluation of practice guidelines but its application will be limited. Competing interests : none declared Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 11914162 TI - Organization specific predictors of job satisfaction: findings from a Canadian multi-site quality of work life cross-sectional survey AB - Abstract | Background | Organizational features can affect how staff view their quality of work life. Determining staff perceptions about quality of work life is an important consideration for employers interested in improving employee job satisfaction. The purpose of this study was to identify organization specific predictors of job satisfaction within a health care system that consisted of six independent health care organizations. Methods | 5,486 full, part and causal time (non-physician) staff on active payroll within six organizations (2 community hospitals, 1 community hospital/long-term care facility, 1 long-term care facility, 1 tertiary care/community health centre, and 1 visiting nursing agency) located in five communities in Central West Ontario, Canada were asked to complete a 65-item quality of work life survey. The self-administered questionnaires collected staff perceptions of: co-worker and supervisor support; teamwork and communication; job demands and decision authority; organization characteristics; patient/resident care; compensation and benefits; staff training and development; and impressions of the organization. Socio-demographic data were also collected. Results | Depending on the organization, between 15 and 30 (of the 40 potential predictor) variables were found to be statistically associated with job satisfaction (univariate analyses). Logistic regression analyses identified the best predictors of job satisfaction and these are presented for each of the six organizations and for all organizations combined. Conclusions | The findings indicate that job satisfaction is a multidimensional construct and although there appear to be some commonalities across organizations, some predictors of job satisfaction appear to be organization and context specific. Keywords: Background : There appears to be no one commonly accepted definition for quality of work life. In healthcare organizations, quality of work life (QWL) has been described as referring to the strengths and weaknesses in the total work environment . Characteristics that describe the overall organization are viewed as part of the behaviour and reward system of the staff working in that setting. Organizational features such as policies and procedures, leadership style, operations, and general contextual factors of the setting, all have a profound effect on how staff view the quality of their work life. QWL is an umbrella term which includes many concepts. Therefore, concentrating on only one job characteristic, whether it is wages or management style, is an inadequate approach to assessing QWL. Because the perceptions held by employees play an important role in their decisions to enter, stay with or leave an organization, it is important that staff perceptions be included when assessing QWL. And although job satisfaction is not QWL, perception of QWL is often assessed using job satisfaction surveys. Previous studies have shown that low job satisfaction is a major cause of turnover among health care providers . In addition, job satisfaction may affect the quality of service and organizational commitment and may be a contributing factor associated with shortages of health care providers . Such findings have recently increased interest in studying job satisfaction among health care providers . The results of a 1993 meta-analysis of 48 studies looking at work satisfaction in over 15,000 nurses revealed that job satisfaction was associated strongly with reduced work stress, organizational commitment, communication with supervisors, autonomy, employee recognition, fairness, locus of control, years of experience, education, and professionalism. This study also found a strong relationship between job satisfaction and QWL for nurses . After reviewing the literature on QWL and job satisfaction, and considering the wide variety of health care settings, situational contexts, and organizational structures (including management styles, reporting structures, staffing complements, and levels of training and experience) in which employees work, we hypothesized that the predictors of job satisfaction would vary depending on the organization. The purpose of this study was to identify organization specific predictors of job satisfaction within a health care system that consisted of six independent and distinct organizations located in five communities in Central West Ontario, Canada. Methods : Setting | The settings for this study included six independent and distinct health care organizations providing varying levels and types of care. All six organizations were affiliated with the St. Joseph's Health System (SJHS) located in five Central West Ontario communities. Collectively, the SJHS is one of the largest corporations in Canada devoted to health care. At the time of the study (2000), the SJHS employed 5,486 full, part and casual time (non physician) staff. Additional information about of each of the six organizations and their respective communities is provided in Table . Table 1 | Characteristics of the Organizations within the St. Joseph's Health System. Questionnaire development | Items included in the "Quality of Work Life Survey 2000" were selected after a review of the literature and extensive consultation between research team members and the QWL Task Force (a management group consisting of representatives from each of the six SJHS organizations). The initial selection of items was influenced by a recently published Canadian study and reports from two meta-analyses . The QWL Task Force then refined these items to consider, among other things, issues of accuracy, relevance, readability, grammar, potential for offensiveness, and appearance of cultural or gender bias. After several months of development, the instrument was pretested on a small group of staff at two of the participating organizations (Site 2 and Site 4 -- see Table ). This pretesting was done to ensure that individuals could follow the instructions associated with the format, to obtain estimates of the time required to complete the survey instrument, to identify items that were poorly written or ambiguous, and to identify an appropriate implementation strategy. The questionnaire and implementation strategies were revised accordingly. The final 65-item survey contained nine sections representing topic areas considered relevant to assessing QWL in the SJHS. Eight scale scores were developed from the individual items (see below and : Statistically Significant Organization Specific (Univariate) Predictors of Job Satisfaction). The Co-worker and supervisor support section included 10 closed-ended and 1 open-ended questions. A 3-item supervisor social support scale included questions about supervisor helpfulness, concern about the welfare of employees, and ability to facilitate effective interaction among employees. Co-worker support was measured by a 7-item scale reflecting the extent to which co-workers were seen as competent, understanding, and supportive of employees. Both scales where adapted from Woodward et al. (1999) . The Teamwork and Communication section included 9 closed-ended and 1 open-ended questions. For determining teamwork, a 7-item scale was adapted from Taylor and Bowers (1972) to measure the extent to which one's work unit coordinates efforts, solves problems and works together effectively . A 2-item scale developed for this project measured how communication was practiced within the organization. The Job Demands and Decision Authority section included 15 closed-ended and 1 open-ended questions. It included a 4-item scale adapted from Brosnan and Johnson (1980) to measure clarity regarding responsibilities, workloads and conflicting demands . There was also a 9-item scale adapted from Karasek et al. (1998) to measure the extent to which respondents' jobs gave them autonomy or decision-making latitude , and 2 questions which reflected the demands of one's work . The Characteristics of Your Organization section included 6 closed-ended and 1 open-ended questions. This section was adapted from Woodward et al. (1999) and included a 4-item scale that inquired about the extent to which the organization encouraged the best efforts from staff, and how employees were treated . Two additional questions examined the extent to which staff were kept informed, and organizational recognition of employee contributions. The Patient/Resident Care section included 5 closed-ended and 1 open-ended questions. The questions (developed for this project) were used to measure employees' perceptions of the quality and timeliness of care provided for patients and residents at their respective organizations. The Compensation and Benefits section included 10 closed-ended and 1 open-ended questions. These questions were developed for this project to determine employee satisfaction concerning a number of employee benefits and level of pay. The Staff Training and Development section included 6 closed-ended and 1 open-ended questions. These questions (developed for this project) measured the extent to which each organization supports its staff in training, educational development and opportunities for advancement. The Overall Impressions of Your Organization section included 4 closed-ended and 4 open-ended questions. All of the questions (developed for this project) assessed staffs' impressions of and overall satisfaction with their organization. The question "Overall, how satisfied are you with your job?" was used as the outcome variable in this study. The Staff Socio-Demographic Information section included 10 closed-ended questions (developed for this project) to collect information on gender, age, marital status, education, length of employment, supervisory status, time spent on job activities, job status and job classification. Within each of the first 8 sections, employees were asked to circle the response that best described their feelings using 5-point Likert scales. Employees were also asked for written comments pertaining to each of the sections and were provided space to comment on other issues they felt were important. Survey Procedure | Because of the diversity of organizations and staff within the SJHS, it was decided by the QWL Task Force, organization administrators and researchers that the implementation of the survey would be customized to best fit each of the organizations. It was felt that a varied approach would be more feasible for the organizations and that this would help maximize response rates. Although the procedures were not identical, all of the organizations provided as a minimum: advance notification (written or voice mail) of the survey to all staff (eligibility was based on whether the worker was active on the organization's pay roll at the time of the study and was not a physician); access to questionnaires for all staff (the QWL Task Force felt that each staff member in the SJHS should have the opportunity to complete a questionnaire); one or more reminder notices (e.g., letters, newsletters, voice mail, personal communication); and sealed drop off boxes for completed questionnaires. Pilot testing of the questionnaire revealed that employees felt that tracking individual employees for the purpose of follow-up (i.e., to increase response rate), violated the perception of anonymity and confidentiality. Therefore, to help ensure anonymity and confidentiality, follow-up attempts were limited to general reminder notices to all staff. Analysis | All closed-ended (or quantitative) responses were entered directly from the questionnaires into SPSS (version 10.0.5 for Windows, SPSS, Inc., Chicago, 1999). Prior to data analysis, most of the survey questions were re-coded. Questions which asked participants to select one response within a five point scale (never to always; very dissatisfied to very satisfied; very poor to very good; no, definitely not to yes, definitely) were collapsed into two categories. For example, for the response scale (1=very dissatisfied, 2=dissatisfied, 3=not sure, 4=satisfied, 5=very satisfied) those who indicated they were either satisfied or very satisfied were re-coded as "satisfied" while all others were re-coded "not satisfied" by default. In several instances, it was appropriate to combine two or more of the questions into a composite scale score. See "Questionnaire Development" section and : Statistically Significant Organization Specific (Univariate) Predictors of Job Satisfaction for additional details on how the composite scale scores were calculated. In total, there were eight scale scores (supervisor social support; co-worker support; teamwork; communication; role clarity; decision latitude; organization/staff relations; patient/resident care). Scale scores were generated by summing the participant responses (i.e. one to five) for all questions that made up the scale. In the rare situation where a participant failed to answer one or more of the questions that made up a scale score, missing values were replaced with mean values for that organization. Scale scores were categorized into meaningful dichotomous categories prior to analysis (e.g., satisfied or not satisfied). For the purpose of this study, QWL was operationally defined using the global question "Overall, how satisfied are you with your job?". Employees rated job satisfaction from very dissatisfied to very satisfied using a five point scale (very dissatisfied, dissatisfied, not sure, satisfied, very satisfied). For the analysis, however, those indicating they were either satisfied or very satisfied were considered to be "satisfied" with their jobs. All others were considered "not satisfied" with their jobs. Prior to analysis, study researchers reached a consensus on which survey questions to include as potential predictors of job satisfaction. In total, there were eight scale scores and 32 questions that were rationalized a priori as potential predictors of job satisfaction. Data from each of the organizations, as well as all of the organizations combined (representing the SJHS), were analyzed separately to identify predictors of job satisfaction. T-test, chi-square analyses and, when appropriate, Fisher exact tests were used to determine which of the variables were statistically associated with job satisfaction i.e., were potential predictors of job satisfaction. Descriptive information (numbers and percentages) for each of the variables was calculated by whether or not staff were satisfied with their jobs. In addition, p-values, odds ratios, and 95% confidence intervals for the odds ratios were calculated for each potential predictor of job satisfaction. Separate logistic regression analyses were used to identify the best predictors of job satisfaction for each organization and for all organizations combined (SJHS). Only variables which had a statistically significant association with job satisfaction were included in these analyses. Adjusted odds ratios and corresponding 95% confidence intervals are reported for each organization and the SJHS. The logistic regression analyses produces odds ratios which have been simultaneously adjusted for all other variables in their respective final models. The goodness of fit of the logistic regression models were assessed using the rho-squared statistic . A rho-square value between 0.20 and 0.40 suggests a very good fit of the model. A probability level of <0.05 was used to determine statistical significance. SPSS and Epi-Info (version 6.04a, Centers for Disease Control and Prevention, Atlanta, 1995) were used for statistical computations. Results : Table provides additional information about each of the six health care organizations, including the type of organization, number of staff, number of beds or visits/year, and the size of the community where the organization was located. Respondent participation rate | Response rates are often used as an indicator of the representativeness of a sample of respondents. Of the combined 5,486 staff, 1,819 (33.2%) returned a completed questionnaire. Organization specific response rates varied from 25.3% to 55.3% . In an attempt to assess the representativeness of respondents, a comparison was made of available socio-demographic information between respondents and all staff within each of the organizations. Overall, female employees were more likely to respond than male employees (it should be noted, however, that the vast majority of staff (82% to 98%), were females within each of these organizations), as were full-time employees compared to part-time, casual or temporary employees. There were also some differences in respondents, across organizations, based on job classification. All organizations, however, had respondents within each job classification. A statistical estimating procedure was also used to assess how accurately respondents represent staff at each of the organizations . This calculation suggests that the organization specific findings were accurate plus or minus 3.6% to 8.8%, 19 times out of 20 . Table 2 | Response rates and accuracy of responses by organization. Potential predictors of job satisfaction | Organization specific and combined SJHS (univariate) analyses (t-test, chi-square analyses and, when appropriate, Fisher exact tests) were used to determine which of the potential predictor variables were statistically associated with job satisfaction. Included in these analyses were the 40 potential predictor variables (8 scale scores and 32 individual questions). See : Statistically Significant Organization Specific (Univariate) Predictors of Job Satisfaction for a list of all variables. The number of statistically significant variables ranged from 15 to 30 depending on the organization and 32 for all organizations (SJHS) combined (see : Statistically Significant Organization Specific (Univariate) Predictors of Job Satisfaction). Best predictors of job satisfaction | Separate logistic regression analyses were then used to identify the best predictors of job satisfaction for each organization and for all organizations combined (SJHS). All variables found to be statistically associated with job satisfaction from the univariate analyses were entered into these logistic regressions analyses. The best predictors of job satisfaction are presented in Table . The ranking assigned to these variables relates to the order in which variables were added to the logistic regression models. For example, the rank "1" refers to the first variable that was added to the model i.e., the variable which best improved the fit of the model (or the most important variable). A more detailed description of the magnitude (as represented by the size of the odds ratios) and statistical significance (as represented by the 95% confidence intervals of the odds ratios) of the association between each of these predictors and job satisfaction is presented below for each organization and all organizations combined (SJHS). The best predictors of job satisfaction are again ranked according to their importance. All of the odds ratios presented below have been simultaneously adjusted for all other variables in their respective final logistic regression models. All logistic regression models achieved a rho-square between 0.20 and 0.40 suggesting they were very good (fitting) models for predicting job satisfaction. Table 3 | Best Predictors of Job Satisfaction1 Ranked by Organization2. Site 1 (community hospital) | The most important predictors of job satisfaction were: 1) being satisfied with the organization's recognition of employee contributions (OR 5.01, 95% CI 1.59 to 15.81), 2) good decision authority (OR 7.91, 95% CI 1.46 to 42.92), 3) being satisfied with patient resident care (OR 4.66, 95% CI 1.36 to 15.97), and 4) good role clarity (OR 4.24, 95% CI 1.16 to 15.49). The final model achieved a rho-square of 0.30. Site 2 (community hospital/long-term care facility) | The most important predictors of job satisfaction were: 1) good open communication between staff (OR 2.55, 95% CI 1.03 to 6.35), 2) good supervisor social support (OR 6.27, 95% CI 1.36 to 29.00), 3) organization keeps staff informed (OR 3.73, 95% CI 1.51 to 9.20), 4) good decision authority (OR 3.49, 95% CI 1.25 to 9.73), and 5) being satisfied with pay level (OR 2.47, 95% CI 1.14 to 5.34). The final model achieved a rho-square of 0.24. Site 3 (visiting nurse organization) | The most important predictors of job satisfaction were: 1) less frequently (never/seldom/sometimes) asked to do an excessive amount of work (OR 7.22, 95% CI 2.22 to 23.46), 2) being satisfied or very satisfied that the organization keeps employees informed (OR 4.52, 95% CI 1.43 to 14.32), 3) belief the organization carries out its Mission statement (OR 11.17, 95% CI 2.04 to 61.14, and 4) good decision authority (OR 5.29, 95% CI 1.32, to 21.22). The final model achieved a rho-square of 0.34. Site 4 (long-term care facility) | The most important predictors of job satisfaction were: 1) belief the organization carries out its Mission statement (OR 4.63, 95% CI 1.77 to 12.51), 2) good supervisor social support (OR 3.32, 95% CI 1.22 to 9.04), 3) good decision latitude (OR 11.61, 95% CI 1.33 to 101.8), 4) often or always given enough time to get the job done (OR 3.05, 95% CI 1.00 to 9.35), and 5) spending 38 hours or more on the job or job related activities (OR 3.55, 95% CI 1.32 to 9.59). The final model achieved a rho-square of 0.34. Site 5 (community hospital) | The most important predictors of job satisfaction were: 1) belief the organization carries out its Mission statement (OR 3.42, 95% CI 1.82 to 6.43), 2) satisfied that the organization keeps staff informed (OR 2.62, 95% CI 1.48 to 4.65), 3) not being asked frequently to do an excessive amount of work (OR 2.41, 95% CI 1.36 to 4.27), 4) good decision latitude (OR 5.65, 95% CI 2.09 to 15.25), 5) being satisfied with pay level (OR 2.41, 95% CI 1.37 to 4.23), 6) being female (OR 2.99, 95% CI 1.29 to 6.90), and 7) good role clarity (OR 2.45, 95% CI 1.02 to 5.86). The final model achieved a rho-square of 0.25. Site 6 (tertiary care hospital/community health centre) | The most important predictors of job satisfaction were: 1) belief the organization carries out its Mission statement (OR 3.99, 95% CI 2.52 to 6.31), 2) good communication (OR 3.00, 95% CI 1.85 to 4.88), 3) being given enough time to get the job done (OR 2.63, 95% CI 1.58 to 4.40), 4) being a member of the nursing staff (OR 2.73, 95% CI 1.75 to 4.26), 5) good organization support for training and development (OR 3.51, 95% CI 1.59 to 7.76), 6) good decision latitude (OR 2.57, 95% CI 1.30 to 5.09) and 7) being satisfied with the organization's recognition of employee contributions (OR 2.05, 95% CI 1.07 to 3.91). The final model achieved a rho-square of 0.25. All sites combined (SJHS) | The most important predictors of job satisfaction after adjusting for site were: 1) belief the organization carries out its Mission statement (OR 2.79, 95% CI 2.07 to 3.77), 2) good communication (OR 1.87, 95% CI 1.33 to 2.62), 3) less frequently being asked to do an excessive amount of work (OR, 1.80, 95% CI 1.33 to 2.43), 4) good decision latitude (OR 3.28, 95% CI 2.09 to 5.17), 5) being satisfied with pay level (OR 1.61, 95% CI 1.21 to 2.15), 6) being satisfied with the organization's recognition of employee contributions (OR 1.57, 95% CI 1.07 to 2.29), 7) being female (OR 2.83, 95% CI 1.81 to 4.42), 8) good role clarity (OR 1.73, 95% CI 1.17 to 2.56), 9) being satisfied that the organization keeps employees informed (OR 1.35, 95% CI 1.00 to 1.85), 10) good teamwork (OR 1.45, 95% CI 1.01 to 2.09), 11) being given enough time to get the job done (OR 1.57, 95% CI 1.10 to 2.23), and 12) good organization/staff relations (OR 2.02, 95% CI 1.13 to 3.62). The final model achieved a rho-square of 0.26. Discussion : The results of this survey were intended to assist decision-makers in identifying key workplace issues, as perceived by employees, in order to develop strategies to address and improve the quality of working conditions for staff within each of the individual health care organizations and the SJHS as a whole. This research represents the first step of an ongoing process to ensure better QWL for employees. In addition to the findings presented here, information from the survey's open-ended written comments have also been summarized for each of the six organizations (L Lohfeld, K Brazil, P Krueger, G Edward, D Lewis, E Tjam, E., personal communication, 2001) and the SJHS as a whole (St. Joseph's Health System Quality of Work Life Technical Reports 2000). This open-ended information provides additional and complementary information to that which is provided in this report. Together, these findings are currently being used by decision-makers at each of the organizations, and the SJHS, in an effort to improve employee QWL. It should be noted that at the time of this survey, all of the hospitals included in this study (as well all other hospitals within the Province of Ontario) were operating in an environment of restructuring and change. This was a time of anxiety for many health care professionals, hospital staff and the general public. In 1996, the Ontario government created a Health Services Restructuring Commission (HSRC) with a four year mandate to restructure Ontario's hospitals and health services system. The HSRC was given authority under the Public Hospitals Act and The Ministry of Health Act to direct public hospitals to change their roles, transfer services and programs, amalgamate or close. The HSRC completed its mandate, announced its decisions and was terminated in March 2000. The timing for this study was after the decisions of the HSRC were announced. All of the organizations included in this study were impacted to varying degrees either directly or indirectly the HSRC decisions. The most notable impacts occurred at Site 1 and Site 2. Site 1 (a community hospital) was ordered closed effective March 2001 with programs and services to be transferred to the other local community hospital while site 2 (a community hospital/long-term care facility) was ordered to transfer its acute care services to the other local hospital in its community thereby becoming a long-term care facility. During the time of the survey, a new building (adjacent to the current facility) for the new long-term care facility was under construction and was scheduled to open in 2002. These contextual issues could have influenced employee responses and therefore the predictors of job satisfaction for all of these organizations, particularly for site 1 and site 2. There are several positive attributes of this study. First, to our knowledge, it is the largest QWL investigation of health care workers in Canada with 1,819 completed interviews. Second, it is also unique in that we collected information from staff at six distinct and functionally diverse health care organizations. Third, because we could not find an "off-the-shelf" QWL instrument that suited our needs and collected all the information desired by key stakeholders, we developed (through a combination of modifying existing instruments and creating our own questions and scales) our own questionnaire. Finally, although the response rates were not as high as we would have hoped, the findings: appear to be consistent with what we expected a priori (the study's investigators had offices within 5 of the 6 organizations thus having inside knowledge about these organizations); appear consistent with the published literature; and were judged credible by management and staff at each of the sites. The statistical estimating procedure to assess how accurately respondents represent staff at each of the organizations also suggest that our findings were fairly representative of staff within these organizations, particularly the larger organizations. Conclusions : The results of this research show that job satisfaction is a multidimensional construct and is a product of the global evaluation of one's work place and context. This report provides valuable information about how employees in specific health care settings view their work environment. A number of organization specific predictors of job satisfaction were identified as a result of this study. The implications of these findings are currently being deliberated as they relate to improving QWL within each of the six health care organizations that make up the SJHS. These findings, may also be of relevance and value to employees, researchers, evaluators, human resource planners and administrators of similar health care organizations. The results of this survey can also be used as baseline measures against which the findings of future job satisfaction surveys can be compared. Such comparisons place this type of research within a continuous quality improvement framework. Competing interests : None declared. Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12022922 TI - Identification of protein tyrosine phosphatase 1B and casein as substrates for 124-v-Mos AB - Abstract | Background | The mos proto-oncogene encodes a cytoplasmic serine/threonine-specific protein kinase with crucial function during meiotic cell division in vertebrates. Based on oncogenic amino acid substitutions the viral derivative, 124-v-Mos, displays constitutive protein kinase activity and functions independent of unknown upstream effectors of mos protein kinase. We have utilized this property of 124-v-Mos and screened for novel mos substrates in immunocomplex kinase assays in vitro. Results | We generated recombinant 124-v-Mos using the baculovirus expression system in Spodoptera frugiperda cells and demonstrated constitutive kinase activity by the ability of 124-v-Mos to auto-phosphorylate and to phosphorylate vimentin, a known substrate of c-Mos. Using this approach we analyzed a panel of acidic and basic substrates in immunocomplex protein kinase assays and identified novel in vitro substrates for 124-v-Mos, the protein tyrosine phosphatase 1B (PTP1B), alpha-casein and beta-casein. We controlled mos-specific phosphorylation of PTP1B and casein in comparative assays using a synthetic kinase-inactive 124-v-Mos mutant and further, tryptic digests of mos-phosphorylated beta-casein identified a phosphopeptide specifically targeted by wild-type 124-v-Mos. Two-dimensional phosphoamino acid analyses showed that 124-v-mos targets serine and threonine residues for phosphorylation in casein at a 1:1 ratio but auto-phosphorylation occurs predominantly on serine residues. Conclusion | The mos substrates identified in this study represent a basis to approach the identification of the mos-consensus phosphorylation motif, important for the development of specific inhibitors of the Mos protein kinase. Keywords: Background : Mos belongs to a small family of cytoplasmic protein serine/threonine kinases having oncogenic activity . It is highly expressed in germ cells but barely detectable in a variety of somatic tissues . Studies in Xenopus oocytes have established a role for c-mos in a) initiation of the maturation process and the meiosis I / meiosis II transition and b) in metaphase II arrest in mature oocytes . In mouse c-Mos is apparently not required for initiation of maturation, however, like in Xenopus it is absolutely essential for the metaphase II arrest . The 124-v-mos oncogene represents one of several transforming gene isolates of the moloney murine sarcoma virus and shows unique constitutive protein kinase activity and enhanced transforming activity when compared to other v-Mos proteins or to c-Mos [,-]. The transforming mechanism of Mos involves signalling through the MAP kinase pathway as phosphorylation of MEK by c-Mos has been demonstrated and mapping analyses have shown that Mos and Raf phosphorylate identical sites on MEK . The upstream events of the Mos/MEK/MAPK signalling cascade have not as yet been identified. In earlier studies we have shown that an activating mechanism of c-Mos is likely to involve a conformational change which is mimicked when a single amino acid is exchanged in the alpha-helix C loop of the kinase domain (Arg145-Gly) resulting in constitutive active c-Mos . Recently Fisher and co-workers proposed an activating mechanism of c-Mos by sequential association with Hsp70 and Hsp90, in addition to phosphorylation . Presence of the activating Arg145-Gly amino acid substitution in 124-v-Mos does not change kinase specificity but is sufficient for constitutive kinase activity . Hence the kinase activity of 124-v-Mos is independent of upstream effectors and we have used this oncogenic Mos derivative to identify substrates for the Mos protein kinase in vitro. Using the baculo virus expression system we have expressed active 124-v-Mos protein kinase, as demonstrated by its ability to auto-phosphorylate, predominantly on serine residues, and to phosphorylate vimentin in vitro. We have analysed a panel of acidic and basic substrates in immunocomplex protein kinase assays and identified two novel in vitro substrates for 124-v-Mos, the protein tyrosine phosphatase 1B and alpha/beta-casein. Results : Three tryptic 124-v-Mos peptides include target sites for auto-phosphorylation | We have expressed 124-v-Mos with the baculovirus system in Sf9 insect cells and immunopurified 124-v-Mos using the anti-Mos N13 antiserum . As a control, a Mos-unrelated protein, a synthetic kinase-inactive construct of PKC, PKCgammaK380R, was expressed in Sf9 cells. Mos kinase assays, completed in the presence of [gamma-32P]ATP, were resolved using SDS-PAGE and the Coomassie blue staining of the protein gel showed visible amounts of immunopurified 124-v-Mos (fig. , arrowhead). The corresponding autoradiograph in figure demonstrates that 124-v-Mos is expressed as a constitutive active protein kinase indicated by its ability to auto-phosphorylate in vitro. Further, a parallel kinase reaction was used for phosphoamino acid analyses which confirmed that 124-v-Mos auto-phosphorylation occurred predominantly on serine residues (fig. ) and a two-dimensional resolution of a tryptic digest of auto-phosphorylated 124-v-Mos showed that three tryptic peptides include auto-phosphorylation target sites (fig. ), demonstrating that auto-phosphorylation occurs on multiple sites of the Mos protein . Figure 1 | Constitutive kinase activity of immunopurified 124-v-Mos from baculovirus expressing Sf9 insect cells. Constitutive kinase activity of immunopurified 124-v-Mos from baculovirus expressing Sf9 insect cells. Auto-phosphorylation of immunopurified 124-v-Mos expressed in Sf9 cells is shown in B (Coomassie stained 10% SDS-PAGE) and A (corresponding autoradiograph). Parallel 124-v-Mos kinase assays were subjected to a two-dimensional phosphoamino acid analysis (C) or a tryptic digestion followed by a two-dimensional resolution (D). Arrowheads indicate the origin of sample application in (C,D) and the position of 124-v-Mos (A,B). 124-v-Mos phosphorylates vimentin but not tubulin in vitro | Initially, we tested the kinase activity of 124-v-Mos using previously identified Mos substrates. It has been shown that 124-v-Mos, derived from mos-transformed fibroblasts, phosphorylates vimentin in vitro and as presented here in figure , in vitro kinase assays using immunopurified 124-v-mos from Sf9 insect cells showed strong vimentin phosphorylation. In contrast, tubulin which has been shown to be phosphorylated in vivo and in vitro by Xenopus c-Mos was not a substrate for 124-v-Mos in vitro (fig. ). We have tested tubulin purified from various organs (mouse brain, testis and spleen) either polymerised, unpolymerised or pretreated with phosphatases but in none of these states found tubulin to be phosphorylated by 124-v-Mos (data not shown). Figure 2 | 124-v-Mos phosphorylates vimentin but not tubulin. 124-v-Mos phosphorylates vimentin but not tubulin. In vitro 124-v-Mos kinase assays with either vimentin (C,D) or purified tubulin from brain (A,B) as substrates were electrophoresed using 10% SDS-PAGE and Coomassie stained (B,D), the corresponding autoradiographs are shown in (A,C). Immunoprecipitates of Sf9 cells expressing the kinase-inactive PKCgammaK380R were indicated as controls. Demonstration of alpha and beta-casein phosphorylation by 124-v-Mos | In search of further substrates for the 124-v-Mos protein kinase we tested MBP; histone HI, H2AS, H3; protamine; protaminsulphate; purified PKC-alpha/-beta II/gamma and alpha- and beta-casein. With the exception of alpha- and beta-casein (fig. ) none of these substrates were phosphorylated by 124-v-Mos (data not shown). The possibility that factors other than 124-v-Mos in the immunoprecipitate might be responsible for the observed casein phosphorylation was eliminated by including a synthetic kinase-inactive construct of 124-v-Mos, 124-v-MosK121R, as a control in addition to the Mos-unreleated protein, PKC_K380R. A comparison of background phosphorylation on beta-casein in the immunoprecipitates of both controls and 124-v-Mos specific phosphorylation showed that 124-v-Mos phosphorylates beta-casein 7fold relative to background (fig. ). Critically, a tryptic digest of phosphorylated beta-casein revealed that 124-v-Mos phosphorylates a specific tryptic peptide in beta-casein which shows no background phosphorylation in either controls (fig. , arrowhead) strongly supporting that 124-v-Mos is able to phosphorylate beta-casein. Further, a two-dimensional phosphoamino acid analysis (fig. ) showed that 124-v-Mos phosphorylates alpha- and beta-casein on serine and threonine residues at a ratio of 1:1. Figure 3 | 124-v-Mos phosphorylates alpha- and beta-casein in vitro. 124-v-Mos phosphorylates alpha- and beta-casein in vitro. Mos kinase assays, in the presence of alpha- and beta-casein, were resolved using 10% SDS-PAGE; the Coomassie stained protein gel shown in 3A, right panel and the corresponding autoradiograph on the left panel. Arrowheads indicate the position of 124-v-Mos, alpha- and beta-casein and the antibody. Using two control immunoprecipitates of Sf9 cells expressing the synthetic kinase-inactive constructs, 124-v-MosK121R or PKCgammaK380R, Mos-specific beta-casein phosphorylation was demonstrated in 3B and 3C: Mos kinase assays were blotted on nylon-membrane, the phospho-beta-casein bands (B, arrowhead) excised and 32P-Cerenkov counts recorded (B). Alternatively, the excised phospho-beta-casein bands were digested with trypsin and electrophoresed using 16% SDS-PAGE (C). The arrowhead in 3C indicates the tryptic beta-casein peptide phosphorylated by wild-type 124-v-Mos only. Further, two-dimensional phosphoamino acid analyses of 124-v-Mos phosphorylated alpha-casein (D, left panel) and beta-casein (D, right panel) were completed, the arrowheads indicating the origins of sample application. The protein tyrosine phosphatase 1B is a novel in vitro substrate for 124-v-Mos | Protein tyrosine phosphatases constitute a diverse family of enzymes that can be divided into several subgroups, including receptor and non-receptor PTPs . The non-transmembrane protein tyrosine phosphatase PTP-1B, a major intracellular PTP is widely expressed. PTP-1B has been demonstrated to be phosphorylated on multiple sites in a cell cycle specific manner whereby mitotic hyper-phosphorylation occurs, reflected by a protein mobility shift in SDS-PAGE analyses . Using purified PTP-1B as a substrate, we show here that 124-v-Mos can phosphorylate PTP-1B in vitro (fig. ). We controlled this result by using immunoprecipitates from Sf9 cells expressing the synthetic kinase-inactive 124-v-Mos construct or purified PTP-1B alone in parallel kinase assays (fig. ). Other kinases such as PKC and CKII that phosphorylate PTP-1B in vitro are unable to induce a mobility shift of PTP-1B as observed in mitotic cells . Likewise, as shown in figure , a Mos-dependent phosphorylation did not result in a mobility shift of PTP-1B. Figure 4 | PTP-1B is a substrate for 124-v-Mos in vitro. PTP-1B is a substrate for 124-v-Mos in vitro. In vitro Mos kinase assays, using purified PTP-1B as a substrate, were resolved using 10% SDS-PAGE and the autoradiograph is shown in 4A. Immunoprecipitates of Sf9 cells expressing the kinase-inactive 124-v-MosK121R variant or PTP-1B alone were included as controls (A,B). A parallel kinase assay was blotted on nylon-membrane and PTP-1B was detected (B) using the PTP-1B-specific antiserum FG6 , arrowheads indicate the position of 124-v-Mos and PTP-1B. Discussion : In this study we have expressed constitutive active 124-v-Mos using the baculovirus expression system and identified novel in vitro substrates for Mos by immunocomplex kinase assays. It has been shown that 124-v-Mos from mos-transformed mouse fibroblasts phosphorylates vimentin in vitro and that v-Mos is physically associated with vimentin in transformed cells . We have used vimentin as a positive control for 124-v-Mos kinase assays in vitro to demonstrate protein kinase activity of baculovirus expressed 124-v-Mos (fig. ). It is known that the kinase activity of c-Mos is regulated by cellular factors and therefore we have chosen the oncogenic variant of c-Mos, 124-v-Mos, in our study since it is independent of activating mechanisms. Recently it has been shown that Hsp70 and Hsp90 physically interact with c-Mos in Xenopus oocytes and are required for c-Mos activation . Another factor controlling c-Mos kinase activity in Xenopus oocytes was identified by Chen and colleagues to be CKII, a tetrameric holoenzyme composed of two catalytic alpha-subunits and two regulatory beta-subunits . In Xenopus oocytes c-Mos kinase activity is inhibited by binding to the C-terminus of CKII beta-subunit and by over-expression of the alpha-subunit of CKII this effect can be neutralized suggesting a binding competition between c-Mos and the alpha-subunit of CKII . Another protein that interacts with c-Mos in Xenopus oocytes is tubulin. Tubulin not only co-precipitates with c-Mos but also serves as an in vivo and in vitro substrate . In contrast, tubulin was not a substrate for 124-v-Mos in our immunocomplex kinase studies (fig. ). Possibly, this indicates that a cellular factor present in Xenopus oocytes and co-precipitating with c-Mos might be necessary for tubulin phosphorylation by the Mos protein kinase. This factor might not interact with the v-Mos protein, be absent in Sf9 insect cells or unable to interact with v-Mos. Interestingly, we have not detected any co-precipitation of the _-subunit of CKII from Sf9 cells with 124-v-Mos in our immunoprecipitates (data not shown). However, as previously mentioned, Hsp70 is known to interact also with 124-v-Mos . Having established that our recombinant 124-v-Mos protein is active in vitro, we tested a variety of molecules in immunocomplex kinase assays and identified alpha- and beta-casein as very good substrates in vitro (fig. ). This phosphorylation was specific to active 124-v-Mos as the overall phosphorylation on casein was significantly reduced using the synthetic kinase-inactive construct 124-v-MosK121R and more importantly, a tryptic peptide of casein was identified to be phosphorylated by 124-v-Mos only and not by either of the controls used in this study. As expected, casein phosphorylation occured on serine and threonine residues. The Mos-specific consensus phosphorylation site has not as yet been identified and only the mos-phosphorylation sites on MAP kinase kinase have been mapped revealing them to be identical to raf-phosphorylation sites . Using the mos substrates identified in this study, it may be possible to determine the specific consensus phosphorylation site for the mos protein kinase as a basis for developing Mos-specific inhibitors. We have also identified protein tyrosine phosphatase 1B (PTP-1B) as a substrate for 124-v-mos in vitro (fig. ). PTP-1B is phosphorylated on multiple sites in vivo and during mitosis becomes hyper-phosphorylated resulting in a mobility shift in SDS-PAGE . Protein kinase C and CKII phosphorylate PTP-1B in vitro but neither is responsible for the observed mitotic hyper-phosphorylation in vivo . We show here that likewise PTP-1B phosphorylation by 124-v-mos is insufficient to effect a mobility shift (fig. ). PTP-1B phosphorylation occurs on serine 386, a phosphoacceptor site for Cdc2/cyclin B in vitro and serine 352, phosphorylated by an unknown kinase. The serine 352 phosphorylation site either might not be a target for Mos in vitro or PTP-1B may be sequentially phosphorylated by multiple kinases in vivo. Interestingly, it has been shown that PTP-1B hyper-phosphorylation does not occur uniquely in mitosis but also during osmotic shock and is induced by several other stress stimuli . Given that activation of c-Mos is dependent on its interaction with the heatshock proteins, Hsp70 and Hsp90, it is tempting to speculate that the Mos kinase may phosphorylate PTP-1B also in vivo. Conclusions : The crucial biological functions of c-mos during meiosis have been analysed by antisense experiments in Xenopus lavis and by generating mos-deficient mice establishing mos as the main player in metaphase II arrest. In contrast, not much is known about activating mechanisms of mos and biochemical properties such as the mos-specific consensus phosphorylation site. In this study we immunopurified an oncogenic and constitutive active variant of mos, 124-v-Mos, and identified novel phosphorylation substrates, PTP1B and alpha- and beta-casein. Our substrates represent a basis to determine the consensus mos-specific phosphorylation site and further, to analyze this phosphorylation ability functionally in vivo. Materials and Methods : Protein expression and in vitro immunocomplex protein kinase assays | The construction and isolation of recombinant baculoviruses expressing active 124-v-Mos and the synthetic kinase-inactive variant of 124-v-Mos, 124-v-MosK121R, is described in detail elsewhere . According to the standard procedure published by Summers & Smith , recombinant proteins were expressed at 27C in Sf9 cells for 48 hrs. and mos was immunopurified using the anti-Mos N13 antiserum as stated in . Mos kinase assays were carried out in 50 _l kinase reaction buffer (10 mM HEPES pH 7.3, 150 mM NaCl, 0.1% Triton X-100, 2 mM DTT, 15 mM MnCl2, 5 mM MgCl2, 2.5 mM beta-glycerophosphate, 2.5 mM NaF, 20 muM ATP/ 10 muCi [_gamma32P]ATP), incubated for 20 min. at 25C and stopped by the addition of Laemmli buffer. For in vitro substrate kinase assays, 2 mug of substrate was added to each kinase reaction. Phosphoproteins were resolved using 10% SDS-PAGE, Coomassie stained, dried and compared with the corresponding autoradiograph. Immunodetection of western blots were performed using the ECL system and protocol (Amersham). Substrates for in vitro immunocomplex kinase assays | alpha- and beta-casein (dephosphorylated, bovine origin) were purchased from Sigma and vimentin from Roche. Purified PTP-1B and the PTP-lB-specific antiserum FG6 were provided by N. Tonks, Cold Spring Harbor . Tubulin was purified from either mouse brain, testis or spleen by F. Propst, Vienna. Two-dimensional phosphoamino acid analyses | Two-dimensional phosphoamino acid analyses were completed according to Boyle and colleagues . Briefly, phosphoproteins were separated using SDS-PAGE, blotted on nylon-membrane and the desired protein bands were excised. The membrane strips were washed sequentially with 100% methanol and water and the phosphoproteins hydrolysed for 60 min. at 110C in 5.7 N HCl. The hydrolysed samples were lyophilised, resuspended in 2.5% formic acid, 7.8% acetic acid and mixed at 15:1 with a non-radioactive amino acid standard (1 mg/ml of each phospho-serine, -threonine, -tyrosine; Sigma). Finally, samples were spotted on thin-layer chromatography plates and separated in two dimensions using the HTLE-7000 apparatus and manufacture's procedure (Two-Dimensional Peptide Mapping And Phosphoamino Acid Analysis, Featuring The Hunter Thin Layer Plate Electrophoresis System. B. Boyle & T. Hunter, C.B.S. Scientific Company, Del Mar, USA). First dimension: 20 min. electrophoresis at 0.8 bar, 1 kV in 2.5% formic acid, 7.8% acidic acid. Second dimension: 16 min. at 0.8 bar, 1.3 kV in 5% acidic acid, 0.5 % pyridine. The phosphoamino acids were fixed for 10 min. at 65C and the standard non-radioactive amino acids visualised by spraying the chromatography plates with 0.25% ninhydrin followed by incubation for 15 min. at 65C. The phosphoamino acids were located by comparing the autoradiograph with the stained standard amino acids. Tryptic digests and one- or two-dimensional separation of tryptic phosphopeptides | According to Boyle and colleagues phosphorylated proteins were proteolytically digested with trypsin by incubating twice for 2 hrs. at 37C, on each occasion with 10 mug trypsin (Promega, modified trypsin, sequencing grade) in 200 mul 50 mM NH4HCO3 and a two-dimensional separation of tryptic phosphopeptides was completed using the HTLE-7000 apparatus and manufacture's protocol: electrophoretic separation was performed on thin layer chromatography plates for 25 min. at 0.8 bar and 1 kV, followed by conventional chromatography in 39.25% n-butanol, 30.25% pyridine, 6.1% acetic acid. One-dimensional separation of tryptic phosphopeptides was achieved using 16% SDS-PAGE according to Schagger and von Jagow . List of Abbreviations used : Sf9, Spodoptera frugiperda cell line; MAPK, mitogen-activated protein kinase; MEK, MAP and erk kinase; Hsp, heat-shock protein; PTP, protein tyrosine phosphatase; MBP, myelin basic protein; PKC, protein kinase C; CKII, casein kinase II. Backmatter: PMID- 12014993 TI - Mitochondria from cultured cells derived from normal and thiamine-responsive megaloblastic anemia individuals efficiently import thiamine diphosphate AB - Abstract | Background | Thiamine diphosphate (ThDP) is the active form of thiamine, and it serves as a cofactor for several enzymes, both cytosolic and mitochondrial. Isolated mitochondria have been shown to take up thiamine yet thiamine diphosphokinase is cytosolic and not present in mitochondria. Previous reports indicate that ThDP can also be taken up by rat mitochondria, but the kinetic constants associated with such uptake seemed not to be physiologically relevant. Results | Here we examine ThDP uptake by mitochondria from several human cell types, including cells from patients with thiamine-responsive megaloblastic anemia (TRMA) that lack a functional thiamine transporter of the plasma membrane. Although mitochondria from normal lymphoblasts took up thiamine in the low micromolar range, surprisingly mitochondria from TRMA lymphoblasts lacked this uptake component. ThDP was taken up efficiently by mitochondria isolated from either normal or TRMA lymphoblasts. Uptake was saturable and biphasic with a high affinity component characterized by a Km of 0.4 to 0.6 muM. Mitochondria from other cell types possessed a similar high affinity uptake component with variation seen in uptake capacity as revealed by differences in Vmax values. Conclusions | The results suggest a shared thiamine transporter for mitochondria and the plasma membrane. Additionally, a high affinity component of ThDP uptake by mitochondria was identified with the apparent affinity constant less than the estimates of the cytosolic concentration of free ThDP. This finding indicates that the high affinity uptake is physiologically significant and may represent the main mechanism for supplying phosphorylated thiamine for mitochondrial enzymes. Keywords: Background : Thiamine is a water-soluble, B-complex vitamin that cannot be synthesized by mammals, and thus thiamine can be obtained only from dietary intake. This can lead to severe consequences in humans when thiamine is limiting; thiamine deficiency may result in beriberi and the Wernike-Korsakoff syndrome . Being positively charged and present in relatively low plasma concentrations, thiamine movement across cellular membranes requires transporters. Upon being taken up by a cell, thiamine is rapidly diphosphorylated by thiamine diphosphokinase to give thiamine diphosphate (ThDP) . Thus, thiamine represents only a few percent of the total cellular thiamine/thiamine phosphate derivatives. ThDP serves as a cofactor for several enzymes that are found both in the cytosol (transketolase) and mitochondria (alpha-ketoglutarate dehydrogenase complex being the most studied example). The intracellular concentration of ThDP has been estimated at 30 muM, with only about 7 percent being free cytosolic and the remainder being enzyme-bound with much of this within mitochondria . Previous findings indicate a complex, cell-type dependent regulation of compartmentalization and intracellular pools of thiamine and its phosphorylated derivatives in response to fluctuating extracellular thiamine levels . Hence, thiamine transport, including that by mitochondria, is of interest. Thiamine entry into mammalian cells occurs by a saturable, high affinity transporter that is deficient in humans with thiamine-responsive megaloblastic anemia (TRMA) . Thiamine uptake by mitochondria has been demonstrated , yet thiamine diphosphokinase is cytosolic and mitochondria cannot convert thiamine to ThDP . Barile and coworkers demonstrated saturable uptake of ThDP by rat liver mitochondria characterized by a Km of around 20 muM. Although the estimated concentration, in mice and human cells, of intracellular ThDP is about 30 muM, much of this is found in a low turnover pool representing enzyme-bound ThDP . The estimated concentration of free cytosolic ThDP available for intracellular transport is about 10% of the total concentration and thus 2 to 3 muM . Hence the physiological significance of the mitochondrial ThDP uptake just decribed is uncertain. Within mitochondria, ThDP can be converted to thiamine monophosphate . Thiamine or ThDP entry into mitochondria from TRMA cells has not been studied. Interestingly, ThDP-utilizing enzymes in mitochondria are much less affected (as revealed by loss of activity) upon progressive depletion of thiamine available to TRMA cells than are ThDP-utilizing enzymes of the cytosol . For these reasons, we have examined the uptake of thiamine and especially ThDP by mitochondria from several human cell types, including cells from TRMA patients. Results : Uptake of thiamine by cells and mitochondria | Although our interests primarily were in mitochondrial uptake of thiamine and its derivatives, we first examined cellular uptake of thiamine by the lymphoblast cell lines and found thiamine uptake properties typical of other mammalian cells. Figure indicates thiamine uptake by normal lymphoblasts and lymphoblasts derived from a TRMA patient. The high affinity transport of thiamine by normal lymphoblasts is abolished in the presence of a 100 fold excess of unlabeled thiamine. Under such conditions, some uptake continues from a low affinity (Km in the mM range) transport mechanism and/or from diffusion that characterizes thiamine uptake in all mammalian cells examined to date. Using an expanded range of thiamine concentrations from that shown in fig. in multiple experiments resulted in a Km of 1.0 +- 0.9 muM for the high affinity transport by normal lymphoblasts. As expected, lymphoblasts derived from the TRMA patient showed no high affinity thiamine transport as revealed by thiamine uptake being the same in the absence and presence of excess unlabelled thiamine. Figure 1 | Uptake of radioactive thiamine by normal and TRMA lymphoblasts and mitochondria isolated from the lymphoblasts. Uptake of radioactive thiamine by normal and TRMA lymphoblasts and mitochondria isolated from the lymphoblasts. A. Late log phase lymphoblasts from normal (squares) or TRMA individuals (circles) were incubated for 30 minutes with various concentrations of radioactive thiamine. Incubations were carried out in the absence (unfilled symbols) or presence (filled symbols) of a 100 fold excess of non-radioactive thiamine (at each concentration). Cell-associated counts per minute were determined, and the velocity (V) (pmol thiamine per 2 x 106 cells per min.) is plotted versus the concentration (in micromolar) of radioactive thiamine.). Error bars represent SEM for two independent experiments. B. Mitochondria were isolated from lymphoblasts derived from normal (squares) or TRMA individuals (circles) were incubated for 15 minutes with various concentrations of radioactive thiamine. Incubations were carried out in the absence (unfilled symbols) or presence (filled symbols) of a 100 fold excess of non-radioactive thiamine (at each concentration). Mitochondrial-associated counts per minute were determined, and the velocity (V) (pmol thiamine per mg mitochondrial protein per min.) is plotted versus the concentration (in micromolar) of thiamine.). Error bars represent +- SEM for two independent experiments. C. Western anaylsis indicating the presence of the thiamine transporter in plasma membrane fractions and in mitochondrial fractions. Equivalent volumes of subcellular fractions were electrophoretically separated, blotted to a filter, and probed using antisera specific for the human thiamine transporter that is mutated in TRMA individuals. Lane 1, plasma membrane fraction; 2, initial mitochondrial fraction; 3 and 4, successive washes of the mitochondrial fraction; 5, final mitochondrial fraction. 75 micrograms of protein were loaded into each lane with the exception of the lanes containing the washes (3 and 4) which were not quantitated. A faint but reproducible (using different preparations) band was found in the final mitochondrial fraction. Mitochondria isolated from normal lymphoblasts also were found to take up thiamine (fig. ) in a manner similar to that of cellular uptake, with both a high and low affinity component. Using an expanded range of thiamine concentrations in multiple experiments resulted in a Km of 2.1 +- 0.4 muM for high affinity thiamine uptake by normal lymphoblast mitochondria. Interestingly, mitochondria from TRMA lymphoblasts did not possess a "high affinity" thiamine transport capacity as did mitochondria form normal lymphoblasts. No difference in uptake of thiamine by TRMA mitochondria was found in the presence and absence of excess unlabelled thiamine (fig. ). This finding suggests that cellular and mitochondrial uptake of thiamine may be mediated by the same transporter since TRMA is defined by mutation within the thiamine transporter located on the plasma membrane . Using antiserum specific for the human thiamine transporter that is mutated in TRMA individuals, western analysis consistently resulted in a faint but detectable band within the isolated mitochondrial suspension (fig. ), even after extensive and multiple washing. Uptake of ThDP by mitochondria | Although mitochondria from lymphoblasts (above) and other mammalian cells were found to take up thiamine, the physiological significance of the uptake is unknown given that thiamine diphosphokinase is cytosolic and mitochondria cannot convert thiamine to ThDP . We thus were interested in uptake of ThDP by mitochondria, a possibility demonstrated previously with rat liver . Mitochondria from normal human lymphoblasts were able to take up ThDP in a saturable, biphasic manner (fig. ) with a first saturation in the submicromolar range (described below) and a second at much higher concentrations of ThDP. Time course experiments indicated ThDP uptake was linear for at least 20 minutes (data not shown), and uptake was linear with the amount of mitochondrial protein added (inset of fig. ). Figure 2 | The rate of uptake of ThDP by mitochondria isolated from normal lymphoblasts. The rate of uptake of ThDP by mitochondria isolated from normal lymphoblasts. Mitochondria were isolated from normal lymphoblasts and were incubated for 15 minutes with various concentrations of radioactive ThDP. Mitochondrial-associated counts were determined, and the velocity (V) (pmol ThDP per mg mitochondrial protein per min.) is plotted versus the concentration in micromolar of ThDP ([S]). Error bars represent SEM for four independent experiments. The inset shows uptake (V, pmol ThDP per mg mitochondrial protein per min.) versus varying amounts of resuspended mitochondria (mug of protein) in the presence of 2 M radioactive ThDP. Although the biphasic nature of the uptake is readily seen in various plots of the data, the existence of the high affinity component perhaps is best illustrated in fig. in which the uptake at submicromolar concentrations of ThDP is illustrated (open squares). Uptake of ThDP was repeated in the presence of 30 muM non-radioactive ThDP, a concentration that is 100 to about 40 fold excess over the concentration of radioactive ThDP that was used. Given the Km values for the high and low affinity components (see below), this excess should abolish most of the high affinity uptake but have little to no effect on the low affinity uptake. The uptake due solely to the low affinity component was calculated using the kinetic parameters determined for this component and was plotted as open triangles. As seen in fig. , uptake in the presence of 30 muM non-radioactive ThDP (open circles) was essentially identical to that calculated for the low affinity component and abolition of the high affinity uptake indeed was observed. Figure 3 | The rate of uptake of ThDP at submicromolar concentrations by mitochondria from normal lymphoblasts. The rate of uptake of ThDP at submicromolar concentrations by mitochondria from normal lymphoblasts. Mitochondria were isolated from normal lymphoblasts and were incubated for 15 minutes with various concentrations of radioactive ThDP in the absence (open squares) or presence (open circles) of 30 muM nonradioactive ThDP. Mitochondrial-associated counts were determined, and the velocity (V) (pmol ThDP per mg mitochondrial protein 15 min.) is plotted versus the concentration in micromolar of radioactive ThDP ([S]). The calculated V versus [S] for the low affinity component only, using the kinetic parameters for that component, also is plotted (open triangles). Error bars represent +- SEM for four independent experiments in the absence and three experiments in the presence of non-radioactive ThDP. Using the data from 4 independent experiments resulted in the determination of a Km of 0.38 muM for the high affinity (table ) and 115 muM for the low affinity uptake components. The high affinity Km value compares favorably with the estimated 2 to 3 muM intracellular concentration of free ThDP. Table 1 | Kinetic constants for the high affinity component of ThDP uptake by mitochondria isolated from various cell types. Mitochondria isolated from TRMA lymphoblasts took up ThDP in an essentially identical saturable, biphasic fashion as uptake by normal mitochondria (fig. ). The inset compares the Lineweaver-Burk plot of the high affinity uptake component for mitochondria of both cell types. A high affinity Km of 0.60 (table ) was calculated for mitochondrial uptake of ThDP for TRMA lymphoblasts, a value that is essentially the same as that for mitochondria isolated from normal lymphoblasts. The results indicate that although mitochondria from TRMA lymphoblasts cannot take up thiamine with high affinity, they can efficiently import ThDP, the active form of thiamine. Figure 4 | The rate of uptake of ThDP by mitochondria isolated from normal and TRMA lymphoblasts. The rate of uptake of ThDP by mitochondria isolated from normal and TRMA lymphoblasts. Mitochondria were isolated from normal (open squares) or TRMA (open circles) lymphoblasts and were incubated for 15 minutes with various concentrations of radioactive ThDP. Mitochondrial-associated counts were determined, and the velocity (V) (pmol ThDP per mg mitochondrial protein per min.) is plotted versus the concentration in micromolar of ThDP ([S]). The inset shows the Lineweaver-Burk plot (1/V vs. 1/S) of the high affinity component of ThDP uptake for normal (open squares) and TRMA (open circles) derived mitochondria. Previous work has indicated that different cell types may differentially regulate intracellular pools of thiamine and/or its phosphorylated derivatives . Thus, we examined ThDP uptake by several other cell types. As shown in table , high affinity uptake of ThDP by mitochondria from fibroblasts and neuroblastoma cells was essentially the same as that for normal and TRMA lymphoblasts as revealed by the similar Km values. The apparent affinity for ThDP characteristic of the high affinity uptake component was essentially identical for all cell types examined, however variation was seen in transport capacity, as revealed by substantially different values for Vmax (table ). Mitochondria from all of the cell types examined also possessed the low affinity uptake characterized by Km's similar to that of normal lymphoblasts but with a greater range of values being found (20 to 115 muM). The final entry in table is for ThDP uptake by mitochondria isolated from glyB cells. GlyB cells are a Chinese hamster ovary cell line that is deficient in the transport of folate into mitochondria. For reasons discussed below, ThDP uptake was examined in these cells. As seen in table , high affinity uptake was found with kinetic constants similar to those of the human cell types. The low affinity uptake component was also similar to that by human mitochondria (data not shown). Discussion + Conclusion : As has been reported for mitochondria of rat liver , mitochondria from human lymphoblasts were found herein to take up thiamine in a saturable manner characterized by a Km of 2.1 muM. Upon entry into a cell, thiamine is rapidly diphosphorylated to ThDP, resulting in a low intracellular thiamine concentration . The Km determined here is about an order of magnitude greater than the estimated intracellular thiamine concentration , raising questions about the efficiency of such uptake. Surprisingly, mitochondria derived from TRMA lymphoblasts lacked the high affinity uptake of thiamine. TRMA is caused by mutations which destroy the high affinity thiamine transporter of the plasma membrane . The similar Km values found for cellular and mitochondrial uptake of thiamine for normal lymphoblasts and the lack of such uptake by TRMA mitochondria and TRMA cells suggests that high affinity thiamine import into mitochondria may be carried out by the same transporter or a variant form, perhaps generated by differential splicing, of that serving on the plasma membrane. Although western analysis using anti-human transporter (that is mutated in TRMA individuals) antiserum supports this interpretation, further experiments need to be carried out to substantiate the suggestion of a shared thiamine transporter between these two membrane systems. Even if true, the physiological significance of thiamine uptake by mitochondria is unknown since mitochondria cannot form ThDP from thiamine . We find that mitochondria from a variety of human cell types efficiently take up ThDP. Uptake is biphasic with a high and a low affinity component. The Km values characteristics of the high affinity uptake component (all around 0.4 muM) are comparable to the estimated intracellular concentration of free (non-enzyme bound) cytosolic ThDP of around 3 muM . This suggests that the high affinity uptake system is the physiologically relevant mechanism responsible for ThDP entry into mitochondria. Earlier work with rat liver mitochondria identified a ThDP uptake system with an estimated Km of around 20 muM . This is of the same order of magnitude that we find for the low affinity component in the human cells examined herein. The previous work used a less sensitive procedure of examining ThDP uptake and did not examine uptake below 10 muM. This would explain the lack of identification in the previous work of the high affinity uptake component. There is a high degree of amino acid similarity among folate transporters and the thiamine transporter of the plasma membrane . Recently, it was found that in murine cells there can be a substantial efflux of ThDP mediated by the reduced folate carrier protein . GlyB cells are a Chinese hamster ovary cell line derivative that are deficient in the transport of folates into mitochondria, and the responsible mitochondrial transporter has recently been identified and its gene cloned . We wondered if the mitochondrial folate transporter was responsible for uptake of ThDP into mitochondria, making analogies to the ability of the plasma membrane folate transporter being able to transport ThDP. However, this is not the case as glyB cells that lack the mitochondrial folate transporter were found to take up ThDP with high affinity kinetics similar to that of mitochondria of human cells. Mitochondria from three human cell types -- lymphoblasts, fibroblasts, and neruoblastoma cells -- all possessed a high affinity ThDP uptake component characterized by equivalent apparent affinity for ThDP as revealed by essentially identical Km values. Previous studies indicate the existence of a complex, cell-type dependent regulation of compartmentalization and intracellular pools of thiamine and/or its phosphorylated derivatives in response to fluctuating extracellular thiamine levels . The cell lines used here were also used in the studies leading to this conclusion. Clearly, differences in mitochondrial transporter affinities do not contribute to the cell-type dependent regulation of ThDP compartmentalization. However, we did find significant differences in ThDP uptake capacity with respect to cell type as revealed by variation in the Vmax values. Neuroblastoma mitochondria possessed the largest uptake capacity, having a Vmax 4 to 30 fold higher than that of the other cell types examined. Interestingly, of the three cell types neuroblastoma cells also are the most resistant to changes in mitochondrial ThDP-utilizing enzyme activity upon progressive depletion of the thiamine made available to the cells . This suggests that cell-dependent variation in ThDP uptake capacity by mitochondria may contribute to the cell-dependent regulation of ThDP compartmentalization. As such regulation was most clearly revealed upon progressively depleting thiamine from cells , it will be of interest to examine possible changes in mitochondrial transport capacity in response to thiamine depletion. Studies on the sensitivity of ThDP-utilizing enzymes to progressive depletion of thiamine that is available to the cell indicate that such enzymes in mitochondria are significantly less sensitive than cytosolic enzymes in TRMA cells . This could be interpreted as efficient import of thiamine/ThDP into mitochondria in TRMA cells even though thiamine inefficiently enters these cells due to the lack of the high affinity thiamine transporter. Although we found a lack of mitochondrial uptake of thiamine in TRMA cells, our finding of an intact, high affinity ThDP transport mechanism for TRMA mitochondria is consistent with and offers an explanation for such an interpretation. Materials and Methods : Radiochemicals | [3H]thiamine (1 Ci/mmol, radiochemical purity greater than 97%) and [3H]thiamine diphosphate (1.4 Ci/mmol, radiochemical purity greater than 98 %) were purchased from Moravek Biochemical Inc (Brea, CA). Cell culture | Normal lymphoblasts, TRMA lymphoblasts and fibroblasts cell lines were obtained and have been characterized as described . The human neuroblastoma cells were an SY-SY5Y cell line, a thrice-cloned subline of SK-N-SH . GlyB cells, a Chinese hamster ovary K1 subline, are deficient in the transport of folate into mitochondria and were a kind gift from L. Chasin (Columbia University). All cell types were growth at 37C in the presence of 10 muM thiamine in RPMI 1640 medium supplemented with 10% heat-inactive fetal calf serum, 2 mM L-glutamine and 1 g/L penicillin/streptomycin, except glyB cells which were grown in MEM medium. Cells were grown and used at late log phase or at 80 --90% confluency. Cellular thiamine transport | Cells were harvested, washed four times with 40 ml of ice-cold transport buffer (145 mM NaCl, 1 mM MgCl2, 1 mM CaCl2, 10 mM glucose, 10 mM HEPES, pH 7.4), titered, and preincubated for 30 min. at 37C after resuspending 3 x 107 cells in 1 ml of transport buffer. Various amounts of [3H]thiamine (to give submicromolar and micromolar final concentrations) were added and the reactions were incubated for 30 min. The specific activity of the radioactive thiamine was the same for each concentration used. The cells were collected by rapid filtration onto glass fiber filters (type A/E, Gelman Sciences, Ann Arbor, MI) and washed via filtration with 10 ml of cold transport buffer. After thorough drying (overnight at 60C), the amount of labeled thiamine taken up by the cells was determined by scintillation counting . For each concentration, the uptake in the presence of a 100-fold excess of unlabelled thiamine was performed to assess the contribution to uptake from a low affinity (Km in the mM range) component and/or from diffusion . Isolation of mitochondria | Mitochondria were isolated from about 3 x 108 cells according to published procedures . The final mitochondrial pellet was suspended in suspension buffer (140 mM KCl, 0.3 mM EDTA, 5 mM MgCl2, 10 mM HEPES, pH 7.4) to give a protein concentration of 3 --4 mg/ml. Protein concentration was determined using the Bio-Rad DC Protein Assay Kit (Bio-Rad Laboratories, Hercules, CA). The isolation and purity of the mitochondrial preparations were monitored by western analysis using anti-cytochrome C antiserum and by microscopy. Western analysis also was performed on the subcellular fractions using antiserum raised against a human thiamine transporter-specific peptide. The antiserum detects a protein of 55 KD (predicted size of the plasma membrane thiamine transporter) that is not detected in cells from TRMA individuals that possess a premature stop codon within the transporter gene (unpublished results). Uptake of thiamine and ThDP by mitochondria | Uptake of thiamine and ThDP by mitochondria was determined by a rapid filtration procedure . Incubations were performed at 37C by rapidly mixing 30 mul of mitochondrial suspension (ca. 100 micrograms of protein) with 220 mul of incubation buffer (140 mM KCl, 0.3 mM EDTA, 5 mM MgCl2, 10 mM Mes, pH 6.5) containing labeled thiamine or ThDP at various concentrations. The uptake was stopped at 15 min. by the addition of 2 ml of ice-cold stop buffer (100 mM KCl, 100 mM mannitol, 10 mM potassium phosphate, pH 7.4) and the mitochondria were collected by rapid filtration on 0.45 muM Millipore membrane filters. The filters were immediately washed with 5 ml of stop buffer via filtration, and they were subsequently dried at 60C overnight. The amount of labeled thiamine or ThDP taken up by the mitochondria was determined by scintillation counting. Background binding was determined by using a 100 fold excess of unlabelled thiamine or ThDP in parallel reactions. Authors' Contributions : QS carried out most of the experiments and participated in writing the manuscript. CKS conceived of the study, participated in its design and coordination, performed a few of the thiamine uptake by cells experiments, and participated in writing the manuscript. Abbreviations : thiamine diphosphate (ThDP); thiamine-responsive megaloblastic anemia (TRMA) Backmatter: PMID- 11972899 TI - Interconversion of Anthozoa GFP-like fluorescent and non-fluorescent proteins by mutagenesis AB - Abstract | Background | Within the family of green fluorescent protein (GFP) homologs, one can mark two main groups, specifically, fluorescent proteins (FPs) and non-fluorescent or chromoproteins (CPs). Structural background of differences between FPs and CPs are poorly understood to date. Results | Here, we applied site-directed and random mutagenesis in order to to transform CP into FP and vice versa. A purple chromoprotein asCP (asFP595) from Anemonia sulcata and a red fluorescent protein DsRed from Discosoma sp. were selected as representatives of CPs and FPs, respectively. For asCP, some substitutions at positions 148 and 165 (numbering in accordance to GFP) were found to dramatically increase quantum yield of red fluorescence. For DsRed, substitutions at positions 148, 165, 167, and 203 significantly decreased fluorescence intensity, so that the spectral characteristics of these mutants became more close to those of CPs. Finally, a practically non-fluorescent mutant DsRed-NF was generated. This mutant carried four amino acid substitutions, specifically, S148C, I165N, K167M, and S203A. DsRed-NF possessed a high extinction coefficient and an extremely low quantum yield (< 0.001). These spectral characteristics allow one to regard DsRed-NF as a true chromoprotein. Conclusions | We located a novel point in asCP sequence (position 165) mutations at which can result in red fluorescence appearance. Probably, this finding could be applied onto other CPs to generate red and far-red fluorescent mutants. A possibility to transform an FP into CP was demonstrated. Key role of residues adjacent to chromophore's phenolic ring in fluorescent/non-fluorescent states determination was revealed. Keywords: Background : Recently, homologs of the well-known green fluorescent protein (GFP) from jellyfish Aequorea victoria were discovered in Anthozoa species . These proteins can be subdivided into two main types. First type, fluorescent proteins (FPs), emit a significant portion (25 --80%) of the absorbed photons. Second type, chromoproteins (CPs), effectively absorb but practically do not emit light. Peculiarities of structure that make each GFP-like protein fluorescent or non-fluorescent are poorly understood to date. Only the importance of position 148 (we will use numbering in accordance to GFP, see Fig. ) was demonstrated in experiments on appearance of fluorescence in CPs . Introduction of Ser148 into several CPs made them clearly fluorescent, although the emission brightness of these mutants was significantly lower in comparison with wild type FPs. Figure 1 | Sequence alignment of asCP, GFP, and DsRed proteins. Sequence alignment of asCP, GFP, and DsRed proteins. The numbering is based on GFP. Introduced gaps are represented by dashes. The residues whose side chains form the interior of the beta-can are shaded. Mutations introduced in asCP and DsRed are designated under and below their sequences, respectively. Due to a great and still growing popularity of GFP and novel FPs in biotechnology, a comprehension of structure-function correlations in GFP-like proteins has both a scientific and a practical significance, showing novel possibilities to achieve desirable protein properties artificially. Here, we applied mutagenesis to a chromoprotein asFP595 (asCP) and a red fluorescent protein drFP583 (DsRed) to study transformation of a chromoprotein into a fluorescent protein and vise verse. Results : Although sequence comparison of known GFP-like proteins does not reveal absolutely invariable differences between FPs and CPs, one can draw attention to three positions, specifically, 148, 165, and 203, which are occupied by noticeably different residues in the two types of proteins (Fig. , Table ). Since residues at these positions are in a close proximity to chromophore , it is reasonable to presume that they can participate in the determination of the state (fluorescent or non-fluorescent) of a particular protein. Figure 2 | Schematic outline of the chromophores and selected neighboring residues in GFP (A), DsRed (B, D), and DsRed-NF (C, E, F) in "sticks" and "spacefill" representation. Schematic outline of the chromophores and selected neighboring residues in GFP (A), DsRed (B, D), and DsRed-NF (C, E, F) in "sticks" and "spacefill" representation. Carbon atoms are gray, nitrogen atoms are blue, and oxygen atoms are red. Images were generated by RasMol 2.6 software. Computer modeling for DsRed-NF was performed using Swiss-PdbViewer and HyperChem 5.01 software. Table 1 | Amino acids occupying positions 148, 165, and 203 (GFP numbering) in known GFP-like proteins. Random mutagenesis of asCP at position 148 | Earlier, we demonstrated for several CPs that Ser-148 containing mutants possess red fluorescence . To check other residue we fulfilled mutagenesis using degenerated primers encoding any amino acid at position 148. Visual inspection of about 50 recombinant clones and sequence analysis of the selected clones showed the following. Only Ser148 ensured clear fluorescence. Several intensively colored non-fluorescent clones contained Ala, Cys, Asn, or Gly at position 148 (remarkably, known wild type CPs carry the very Ala, Cys, or Asn at this position). All other substitutions of Ala148 appeared to be intolerable for proper protein folding and chromophore maturation. Mutagenesis of asCP at position 165 | First of all, we tested a substitution S165V because several FPs carry Val at this position. This mutation resulted in the appearance of a clearly visible red fluorescence with a maximum at 620 nm (Fig. , Table ). Interestingly, in comparison with the wild type asCP, the mutant asCP-S165V showed a strongly modified absorption spectrum which included an additional peak at 390 nm. Absorption at this wavelength produced a very weak (about 10-fold less than the red fluorescence) blue fluorescence at 465 nm. Figure 3 | Normalized spectra for selected mutants of asCP and DsRed. Normalized spectra for selected mutants of asCP and DsRed. Absorption (black solid lines), excitation (colored dashed lines), and emission (colored solid lines) spectra are shown for each mutant. Blue, green, or red excitation-emission lines correspond to color of fluorescence. (A) asCP-S165V. Blue fluorescence is about tenfold weaker than red. (B) asCP-S165A. (C) asCP-S165C. (D) asCP-S165T. Green emision is about twofold stronger than red. (E) DsRed-NF. Green emission peak is about threefold lower than red one. Table 2 | Spectral characteristics for some mutants of asCP and DsRed. To reveal other substitutions at position 165 that could lead to fluorescence appearance we exploited randomization at this position. As a result, several red fluorescent clones of different brightness were selected. The most bright clones carried the already known substitution S165V. All other fluorescent mutants were considerably (5 --10 fold) dimmer and contained Ala, Cys, or Thr165 (in decreasing brightness order). Absorption spectra for these mutants have a characteristic peak at about 390 nm, but it produces no detectable blue fluorescence . An interesting feature of these low fluorescent mutants is that their excitation spectra for red emission do not coincide with the absorption spectra. This phenomenon implies the existence of different spectral forms within a spectrally heterogeneous population of the mutant protein molecules. Red emitting spectral forms are underrepresented or they possess a very low extinction coefficient. At the same time, the major red light-absorbing spectral forms are non-fluorescent. Random mutagenesis of asCP | To extend the search of amino acid substitutions that are able to convert asCP into a fluorescent protein, we used random mutagenesis of the whole asCP gene. Visual screening of about 5000 recombinant clones revealed only one brightly fluorescent colony. Sequence analysis showed that this fluorescent mutant contained the already known substitution A148S. After a more thorough visual inspection we found several very weakly fluorescent clones containing the following substitutions: S68G; I72N; H176R/K219I; H203R; H203Q; Q220L . Importantly, two independent clones carrying different substitutions at position 203 were collected. Summing up, this experiment has not highlighted novel important sites, because all random mutants were considerably (3 --5 fold) dimmer than the mutants at positions 148 and 165 mentioned above. One can conclude that positions 148, 165 are probably the most important sites that influence the state of asCP. Mutagenesis of DsRed | Finally, we attempted to transform the fluorescent DsRed into a chromoprotein. First of all, mutation S148A was tested. Unexpectedly, this substitution did not exert a strong influence on the fluorescence -- quantum yield for DsRed-S148A mutant decreased by a factor of 1.5 only in comparison to the wild type protein . Then, on the base of this mutant, a series of mutants carrying substitutions I165S, K167M, and S203A,L in different combinations was generated. Position 167 was added to mutagenesis considering the crystallographic studies that revealed a direct interaction between Lys167 and chromophore's Tyr66 . This bond appeared to stabilize the ionized form of the DsRed fluorophore. Mutant proteins containing Leu203 were colorless because of unsatisfactory protein folding in E. coli. Following the spectral properties of other mutants, one can notice a gradient of emission intensity and conclude that all positions mentioned above are important for DsRed fluorescence . However, even a quadruple mutant S148A/I165S/K167M/S203A displayed a clearly visible fluorescence comparable to that of some asCP fluorescent mutants (e.g., asCP-S165V). Thus, this DsRed mutant can not be regarded as a true chromoprotein, although it is very close to the CP state because it possesses hundredfold decreased fluorescence in comparison to DsRed. Then, we tested Cys and Asn that are characteristic for some other known CPs at positions 148 and 165, respectively. Triple mutant DsRed-S148C/I165N/S203A possessed a low quantum yield similarly to the mutant S148A/I165S/K167M/S203A mentioned above. When a substitution K167M was added, the final quadruple mutant S148C/I165N/K167M/S203A became practically non-fluorescent . At the same time, this mutant named DsRed-NF intensively absorbed light. Altogether, these properties make DsRed-NF practically indistinguishable from wild type CPs. Spectra for DsRed-NF are shown in Fig. . An extremely weak dual-color fluorescence can be detected at high protein concentration only. Similarly to the low fluorescent mutants of asCP mentioned (see Fig. ), absorption and excitation spectra for DsRed-NF strongly differ from each other. Interestingly, excitation spectrum for green emission displays 2 peaks: a major peak at 410 nm and a minor peak at 490 nm. Such a shape of the excitation curve is similar to that of wild type GFP and has never been detected for DsRed mutants (to date, only EGFP-like single-peak excitation spectra were described for green-emitting mutants of DsRed ). Probably, the short-wave excitation peak corresponds to a neutral (protonated) form of GFP-like chromophore within DsRed-NF. Discussion : Great diversity of fluorescent and non-fluorescent colors in the family of GFP-like proteins poses an challenging problem of understanding its structural background. Mutagenetic studies lately demonstrated various transitions of fluorescence color in Anthozoa proteins: from red to green , from yellow to green, from green to yellow, and from green to red . Also, red and far-red fluorescent mutants of non-fluorescent CPs were generated . Basic investigation of relationship between fluorescent and non-fluorescent GFP-like proteins was the main goal of the present work. However, some practical applications of the results obtained can be considered. The first part of our work, attempts to convert asCP into FP, revealed importance of position 165 for fluorescence appearance. This finding can be applied on other CPs. To date, mutagenesis of natural CPs is the only way to generate a far-red FPs that are in high demand for various applications. Additional far-red fluorescence color broadens abilities of multicolor labeling and assays based on fluorescence resonance energy transfer (FRET). Knowledge about the ways of transforming CPs into FPs could help to generate novel far-red FPs when novel CPs with red-shifted absorption spectra are found. The second part of our work was to transform DsRed into CP. At first glance, such fluorescence quenching can not be used in practice. However, we found that DsRed-NF mutant can be used to resolve a problem of DsRed tetramerization that is the main disadvantage of this tag . When DsRed is fused with a target protein, especially with oligomeric protein, it often results in improper folding and functioning of the tagged partners as well as intensive aggregation of the fusion protein. To neutralize injurious consequences of DsRed tetramerization we suggest to use a simultaneous co-expression of DsRed-tagged proteins with excess free DsRed-NF. In this case mixed heterotetramers are formed so that DsRed becomes a "monomeric" tag (this approach will be published elsewhere). It was recently demonstrated that DsRed and asCP carry chemically distinct chromophores . Theoretically, spectral differences between DsRed and asCP and generally between FPs and CPs may be explained by the diversity of their chromophores. If so, the appearance of fluorescence in asCP mutants and the disappearance of fluorescence in DsRed mutants should resulted from formation of altered chromophores within these mutants. Alternatively, one may suggest that each chromophore type in GFP-like proteins can be fluorescent or non-fluorescent depending on the protein environment. Some observations speak in favor of this hypothesis. First, all key residues mentioned above (positions 148, 165, 167, and 203) are grouped in a close proximity to the phenolic ring of Tyr66 . Thus, they can more likely participate in stabilization and positioning of the chromophore but not in chromophore cyclization events that result in the diversity of chromophores. Second, asCP demonstrates a striking phenomenon of light-induced reversible increasing of fluorescence . This photoconversion clearly shows that an initially non-fluorescent protein molecule can be switched into a fluorescent state due to some conformation changes. It is well-known that GFP-like chromophores and other chromophores that are capable of cis-trans isomerization are practically non-fluorescent in solution because of fast relaxation of the excited state through chromophore isomerization . Probably, chromophore in FPs must be strongly stabilized by the amino acid environment to ensure high quantum yield, while chromophore surrounding within CPs should be more relaxed to allow energy of absorbed light to dissipate into heat. From this point of view, we can draw the following scheme of DsRed chromophore stabilization. According to the crystal structure of DsRed Ser148 and Lys167 hold the chromophore by a direct interaction with phenolate oxygen . Bulky Ile165 supports the ring of Tyr66 and prevents its movement required for the chromophore isomerization . Although Ser203 has no direct H-bonds with the chromophore in the wild type DsRed, such bonds could be formed in mutants with altered 148, 165 and 167 positions. Possibly, Ser203 in DsRed mutants can turn similarly to GFP Thr203 that forms an H-bond with chromophore's phenolate oxygen . Quantitative data on the influence of each substitution on fluorescence intensity speak in favor of this scheme. Comparing in pairs quantum yields for the available DsRed mutants that differ from each other by one substitution (see Table ), one can note the following. The contribution of each substitution strongly depends on mutation order: the later the substitution is introduced the stronger the impact is. For instance, the mutant S203A demonstrates the same quantum yield as the wild type protein. At the same time, an addition of S203A to the mutant S148A leads to a 1.5-fold decrease in quantum yield. Then, introducing Ala-203 into a double mutant S148A/K167M results in a 2.4-fold decreased fluorescence. Analogously, mutation K167M leads to 2-, 3.2-, or 8.6-fold decrease of quantum yield when Met167 is introduced as second, third or fourth substitution, respectively. Also, 4.8- or 12.9-fold decrease of fluorescence intensity is associated with substitution I165S added to S148A/S203A or S148A/K167M/S203A mutants, respectively. The model of several chromophore-stabilizing interactions mentioned above implies such tendency because the importance of each interaction must progressively increase in absence of one, two or more other bonds. Computer modeling of the chromophore environment within DsRed-NF showed the following . In contrast to Ser148 and Lys167 in DsRed, Cys148 and Met167 in DsRed-NF are incapable of stabilizing the chromophore by H-bonds with phenolate oxygen. Moreover, substitution I165N generates a vacant space near the chromophore (compare Fig. and ). We believe that this space is sufficient to ensure the chromophore cis-trans isomerization after light absorption . Thus, absence of phenolate-stabilizing interactions together with free space around the chromophore can explain an extremely low fluorescence quantum yield of DsRed-NF. Unfortunately, no protein structures for CPs were published to date. Obviously, further crystallographic studies of FPs, CPs, and their mutants are required to make valid conclusions about the structural background of differences between FPs and CPs. Conclusions : The ability for fluorescence of GFP-like proteins depends to a great extent on the surrounding of the phenolic ring of the chromophore. For asCP chromoprotein, mutations at positions 148 and 165 can lead to red fluorescence appearance. For DsRed red fluorescent protein, fluorescence can be quenched by mutagenesis at positions 148, 165, 167, and 203. This knowledge can be applied to other GFP-like proteins in effort of customizing spectral characteristics of FPs and CPs. Materials and Methods : Mutagenesis and protein expression | Site-directed mutagenesis was performed by PCR with primers containing target substitution using the overlap extension method . The Diversity PCR Random Mutagenesis kit (Clontech) was used for random mutagenesis of asCP, in conditions optimal for 4 --5 mutations per 1000 bp. All mutants were cloned into pQE30 vector (Qiagen), so that recombinant proteins contained 6-histidine tag at their N-termini. To express mutant proteins E. coli XL1 Blue cells were transformed with the plasmids according to standard protocols and spread onto 3 --4 Petri dishes with LB agar media supplemented with ampicillin for selection. After overnight growth at 37C the plates were stored for 2 --5 days at room temperature or 4C to allow proteins to mature completely. Then, the plates were washed with PBS. Cells were disrupted by sonication, and soluble recombinant proteins were purified on the TALON metal-affinity resin (Clontech). Spectroscopy | Absorption spectra were recorded on a Beckman DU520 UV/VIS Spectrophotometer. A Cary Eclipse Fluorescence Spectrophotometer (Varian) was used for measuring excitation-emission spectra. For molar extinction coefficient determination, we relied on measuring mature chromophore concentration rather than total protein concentration. DsRed and its mutants were alkali-denatured with equal volume of 2 M NaOH. asCP and its mutants were acid-denatured with equal volume of 2 M HCl. Under these conditions, DsRed and asCP chromophores absorb at 452 and 430 nm, respectively . The amounts of chromophore (that correspond to amounts of matured protein) were equalized among samples, absorption spectra for the native proteins were collected. Absorbance intensities were compared to that of DsRed (extinction coefficient is 75,000 M-1cm-1) or asCP (extinction coefficient is 56,000 M-1cm-1), and molar extinction coefficient for each mutant was estimated. For quantum yield determination, the fluorescence of the mutants was compared to equally absorbing DsRed (quantum yield for DsRed was measured to be 0.70 ). Backmatter: PMID- 12019031 TI - Expression of monolysocardiolipin acyltransferase activity is regulated in concert with the level of cardiolipin and cardiolipin biosynthesis in the mammalian heart AB - Abstract | Background | Monolysocardiolipin acyltransferase (MLCL AT) catalyzes the acylation of monolysocardiolipin to cardiolipin in mammalian tissues. We previously reported that cardiac cardiolipin levels, MLCL AT and cardiolipin synthase activities were all elevated in rats made hyperthyroid by thyroxine treatment. In this study, we examined if cardiac mitochondrial MLCL AT activity was dependent upon the biosynthesis and level of cardiolipin in the heart. Rat heart mitochondrial MLCL AT activity was determined under conditions in which the levels of cardiac cardiolipin and cardiolipin synthase activity were either reduced or unaltered using four different disease models in the rat. In addition, these parameters were examined in a murine model of cardiac cell differentiation. Results | In rats made hypothyroid by treatment with 6-n-propyl-2-thiouracil in the drinking water for 34 days, cardiac cardiolipin content was decreased 29% (p < 0.025) and this was associated with a 32% decrease (p < 0.025) in cardiolipin synthase and a 35% reduction (p < 0.025) in MLCL AT activities. Streptozotocin-induced diabetes or hyperinsulinemia in rats did not affect cardiac cardiolipin content nor MLCL AT and cardiolipin synthase activities. Finally, cardiolipin content, MLCL AT and cardiolipin synthase activities were unaltered during murine P19 teratocarcinoma cell differentiation into cardiac myocytes. In all models, phospholipase A2 activities were unaltered compared with controls. Conclusion | We propose a general model in which the expression of MLCL AT activity is regulated in concert with the biosynthesis and level of cardiolipin in the heart. Keywords: Background : Phospholipids are important structural and functional components of the biological membrane . Structurally, as major components of the biological membrane, they define compartmentalization of organelles and the protective barrier, the cell membrane, which surrounds cells. An important class of phospholipids are the polyglycerophospholipids. Cardiolipin (CL), the first polyglycerophospholipid discovered, was isolated from beef heart by Pangborn in 1942 . In the heart, CL represents approximately 12 --16% of the entire cardiac phospholipid mass and is found exclusively in mitochondria . In mammalian tissues CL is required for the reconstituted activity of a number of key mitochondrial enzymes involved in cellular energy metabolism including for example cytochrome c oxidase, carnitine palmitoyltransferase, creatine phosphokinase, pyruvate translocator, mono-, di- and tricarboxylate carriers, glycerol-3-phosphate dehydrogenase, phosphate transporter, ATP/ADP translocase and ATP synthase . Under experimental conditions in which CL was removed or digested away from these proteins with phospholipases, denaturation and complete loss in activity of many of these proteins was observed. CL interaction with these proteins was specific since substitution with other phospholipids did not fully reconstitute activity. The fatty acyl composition of CL also appeared to be important for this functional reconstitution. For example, the activity of delipidated rat liver cytochrome c oxidase was reconstituted by the addition of CL . The specific activity of reconstituted cytochrome c oxidase varied significantly with different fatty acyl compositions of CL. A strong positive correlation has been established between fatty acid unsaturation of CL and antioxidant production in cells . In staurosporine-treated granulosa cells undergoing apoptosis CL levels were reduced . Peroxidation of CL induced release of cytochrome c from mitochondria into the cytosol and this was associated with the induction of apoptosis . Suppression of CL peroxidation inhibited release of cytochrome c from mitochondria . Thus, the activities of the enzymes that synthesize and remodel CL play a pivotal role in maintaining the content and molecular composition of CL and hence may regulate a plethora of cellular processes from energy metabolism to apoptosis. In mammalian tissues CL is synthesized by condensation of phosphatidylglycerol with cytidine-5'-diphosphate-1,2-diacyl-sn-glycerol catalyzed by CL synthase [for review see ]. Thyroxine treatment of rats stimulated the expression of rat liver and heart mitochondrial CL synthase activities . The increase in CL synthase activity accounted for the elevated levels of CL observed in these organs. We recently identified and characterized the activity of monolysocardiolipin acyltransferase (MLCL AT), the enzyme responsible for monolysocardiolipin (MLCL) acylation to CL in mammalian tissues . In another study, we showed that thyroxine treatment of rats elevated cardiac MLCL AT activity and postulated that MLCL AT may be a rate-limiting enzyme for the molecular remodeling of CL in the heart . The above studies prompted us to examine if cardiac MLCL AT activity was linked to CL biosynthesis and content in the heart. Our results, using four different disease models in the rat in which the level of cardiac CL is either reduced, elevated or unaltered, demonstrate that this is the case. We also observe this relationship in a model of murine cardiac cell differentiation. Results : Cardiac CL content, CL synthase and MLCL AT activities are reduced in hypothyroid rats | In previous studies we observed that cardiac MLCL AT activity was elevated when the cardiac CL content and CL synthase activity were elevated in hyperthyroid rats . We examined if cardiac CL content, CL synthase and MLCL AT activities were reduced in hypothyroid rats. Rats were made hypothyroid by the addition of 0.05% PTU to their drinking water for 34 days. This protocol was shown to produce decreased serum thyroid hormone levels and result in cardiac atrophy in the rat . As seen in Table , in rats that received PTU there was a 48% decrease (p < 0.025) in heart weight compared to controls. In addition, the heart to body weight ratio decreased, indicative of cardiac atrophy. Growth failure was demonstrated by the decreased body weights of the hypothyroid animals compared to controls. As a further control, the activity of an inner mitochondrial membrane marker, succinate dehydrogenase, was determined. Cardiac mitochondrial succinate dehydrogenase activity was reduced 23% (p < 0.025) from 30 +- 3 mumol/minmg to 23 +- 3 mumol/minmg protein in PTU-treated rats. These are documented characteristics of hypothyroidism . Table 1 | Body weight, heart weight, heart CL content, CL synthase and MLCL AT activities in normal and hypothyroid rats. Heart mitochondrial fractions were prepared from rats made hypothyroid by the addition of 0.05% PTU to their drinking water for 34 days and CL content, CL synthase and MLCL AT activities determined. We initially determined the activity of cardiac mitochondrial PA:CTP cytidylyltransferase, a rate-limiting enzyme of CL biosynthesis . PA:CTP cytidylyltransferase activity was 15.1 +- 1.2 pmol/min/mg protein and unaltered (14.7 +- 1.1 pmol/min/mg protein) in cardiac mitochondrial fractions prepared from hypothyroid rats. Hence, PA:CTP cytidylyltransferase served as a control for a mitochondrial enzyme not affected by hypothyroidism. When compared to controls, heart mitochondria prepared from hypothyroid rats exhibited a 29% decrease (p < 0.025) in CL content, a 32% decrease (p < 0.025) in CL synthase activity and a 35% decrease (p < 0.025) in MLCL AT activity . PLA2 activity was 4.2 +- 0.7 nmol/minmg protein and unaltered (4.0 +- 0.5 nmol/minmg protein) in cardiac mitochondria prepared from hypothyroid rats. Thus, cardiac mitochondrial CL content, CL synthase and MLCL AT activities were all reduced in hypothyroid rats. Cardiac CL content, CL synthase and MLCL AT activities are unaltered in streptozotocin-induced diabetic rats and in hyperinsulinemic rats | Previously we showed that cardiac phosphatidylglycerol levels were reduced in streptozotocin-induced diabetic rats but CL synthase activity and CL content were unaltered . We examined if streptozotocin-induced diabetes in rats altered MLCL AT activity or if hyperinsulinemia in rats altered CL synthase and MLCL AT activities in cardiac mitochondria. Rats were made diabetic by injection of steptozotocin or hyperinsulinemic by intraperitoneal addition of insulin. Subsequently, the hearts were removed and mitochondrial fractions prepared. Cardiac CL synthase activities were 3.0 +- 0.5 pmol/minmg protein in hyperinsulinemic rats and did not differ from control (3.1 +- 0.6 pmol/minmg protein) non-insulin injected animals. Cardiac CL content was 5.9 +- 0.5 nmol/mg heart and unaltered compared to controls (6.1 +- 0.1 nmol/mg heart). Cardiac MLCL AT activities were 38 +- 6 pmol/minmg protein in diabetic rats and 43 +- 9 pmol/minmg protein in hyperinsulinemic rats and did not differ from controls (40 +- 6 pmol/minmg protein saline injected and 41 +- 9 pmol/minmg protein non-insulin injected animals, respectively). Thus, in streptozotocin-induced diabetes and hyperinsulinemia, conditions in which the CL content and CL synthase activities were unaltered, MLCL AT activity was unaltered. CL content, CL synthase and MLCL AT activities are unaltered during cardiac cell differentiation | As a distinct model, we examined if CL synthase activity was altered in murine P19 teratocarcinoma cells induced to undergo differentiation into cardiac myocytes. We chose this model since differentiation of murine P19 cells into cardiac myocytes results in an increase in phosphatidylethanolamine biosynthesis, phosphatidylethanolamine mass and lysophosphatidylethanolamine acyltransferase activities . The cells were harvested at various times, 0 --8 days post DMSO addition, and MRNA analysis of markers of cardiac cell differentiation performed on cell lysates. GATA-4 is a member of the GATA family of zinc finger transcription factors and is an early marker of cardiac cell differentiation. As seen in Figure , GATA-4 was expressed at 4 days post DMSO addition relative to the constitutive expression of tubulin. As expected GATA-4 expression preceded the expression of other cardiac genes including B-natriuretic peptide (BNP), alpha myosin heavy chain (alphaMHC), beta myosin heavy chain (betaMHC) and Troponin C relative to the constitutive expression of tubulin. Thus, the P19 cells used in this study differentiated into cardiac myocytes. As previously shown CL content, MLCL AT and PLA2 activities were unaltered during P19 cell differentiation into cardiac myocytes . CL synthase activity was 2.7 +- 0.5 pmol/minmg protein in undifferentiated and unaltered (2.8 +- 0.3 pmol/minmg protein) in differentiated P19 cells. Together, the above five models using hyper- and hypothyroid, diabetic and hyperinsulinemic rats and murine P19 cell differentiation into cardiac myocytes all indicate that expression of mammalian cardiac mitochondrial MLCL AT activity appears to be regulated in concert with the biosynthesis and content of CL in the heart. Figure 1 | Expression of genes during differentiation of P19 cells into cardiac myocytes. Expression of genes during differentiation of P19 cells into cardiac myocytes. P19 cells were incubated with 1% DMSO for up to 8 days. At various times, 0 --8 days post DMSO addition, cells were harvested and mRNA levels of GATA-4, BNP, alpha MHC, beta MHC, troponin C and tubulin were determined by quantitative RT-PCR analysis. Discussion : Previous and current studies in the mammalian heart and liver support the hypothesis that CL content is regulated in concert with the level of CL synthase activity | In the CL biosynthetic pathway, PG is converted to CL by condensation with CDP-DG catalyzed by CL synthase . In vitro studies have indicated that alteration in cellular CL levels appears to have functional consequences. For example, reduction in the content of CL was shown to reduce oxygen consumption in mitochondria prepared from rat liver . Thus, maintenance of the appropriate content of CL in mammalian mitochondria is essential for proper mitochondrial function. Thyroxine treatment of rats was shown to stimulate the activity of rat liver mitochondrial CL synthase 2.5-fold . This elevation in rat liver mitochondrial CL synthase activity was suggested to account for the elevated levels of CL observed in livers prepared from hyperthyroid rats. In addition, CL synthase was shown to be elevated in heart mitochondria prepared from hyperthyroid rats and this was correlated with an increase in cardiac CL content . Prior to the current study, CL synthase activity had not been determined in any model of hypothyroidism. Hypothyroidism in the rat resulted in a 25% reduction in cardiac CL synthase activity. This reduction in CL synthase activity likely accounted for the reduced levels of cardiac CL observed in hearts prepared from hypothyroid animals. Previous studies in the rat have indicated that hypothyroidism also results in reduced CL levels in the liver [for review see ]. Thus, it is reasonable to assume that CL synthase activity would also be reduced in the liver of hypothyroid animals. In the current study, CL synthase activity was unaltered in diabetic and hyperinsulinemic rats and in a model of murine cardiac cell differentiation. In these models, the content of CL was unaltered. These data suggest that the level of CL produced in the mammalian heart is regulated in concert with the level of CL synthase activity. Previous and current studies in the heart support the hypothesis that cardiac MLCL AT activity may be regulated in concert with CL content and CL synthase activity | The data presented in this paper are entirely consistent with the conclusion that the expression of MLCL AT activity in the heart is regulated in concert with the biosynthesis and content of cardiac CL. Previously, we demonstrated that thyroxine-treatment of rats resulted in an increase in cardiac CL content, CL synthase and MLCL AT activities . In the present study rats made hypothyroid with PTU in the drinking water had reduced cardiac CL content, CL synthase and MLCL AT activities. In contrast, in streptozotocin-induced diabetes and hyperinsulinemia, pathological conditions in which cardiac mitochondial CL content and CL synthase were unaltered , MLCL AT activities were unaltered. In addition, CL content, CL synthase and MLCL AT activities were unaltered during cardiac cell differentiation. It is reasonable to propose that when the rate of synthesis and level of CL is either reduced or elevated expression of the activities of the enzymes that remodel CL should be correspondingly reduced or elevated. The activity of cardiac mitochondrial PLA2 was high (100-fold) relative to cardiac mitochondrial MLCL AT activity and was unaltered in all models examined . Since MLCL AT activity was either increased or decreased under conditions in which elevated or reduced CL remodelling was required, i.e. elevated or reduced CL synthesis, it is possible that MLCL AT may be rate-limiting for MLCL acylation to CL in the mammalian heart. However, it should be considered that other factors such as the intra-mitochondrial level of MLCL may be limiting for the acylation of MLCL to CL. Conclusions : A summary of our findings is presented in Table . In hyperthyroidism, when cardiac CL synthase activity and CL content are elevated an increase in MLCL AT activity is observed. In hypothyroidism, when cardiac CL synthase activity and CL content are reduced a decrease in MLCL AT activity is observed. Finally, when cardiac CL synthase activity and CL content are unaltered in streptozotocin-induced diabetes, hyperinsulinemia and murine P19 cell differentiation into cardiac myocytes, MLCL AT activity is unaltered. Thus, expression of MLCL AT activity is regulated in concert with the biosynthesis and content of cardiac CL. Table 2 | Summary of mammalian cardiac mitochondrial MLCL AT activities, CL synthase activities and CL content in various cardiac models. Materials and Methods : Male Sprague Dawley rats (125 --175 g) were used throughout the study and were housed in a temperature and light controlled room. They were maintained on Purina rat chow and tap water ad libitium. Treatment of animals conformed to the Guidelines of the Canadian Council on Animal Care. Rats were made hypothyroid by administration of (0.5% w/v) 6-n-propyl-2-thiouracil (PTU) in their drinking water for 34 days. Rats were made diabetic by injection of 60 mg/Kg steptozotocin. Hyperglycemia was confirmed 24 h later by urine and blood glucose analysis. Rats were made hyperinsulinemic by intraperitoneal addition of 3 units/day of insulin for 28 days. Murine P19 teratocarcinoma cells were kindly provided by Dr. Mona Nemer, Institute of Cardiovascular Research, University of Montreal, Montreal, Quebec, Canada. [1-14C]Linoleoyl-Coenzyme A was obtained from American Radiochemical Co., St. Louis MO. All other radiochemicals were obtained from Dupont, Winnipeg, Canada. MLCL was obtained from Avanti Polar Lipids, Alabaster, AL. Thin-layer chromatography (Silica gel 60, 0.25 mm thickness) plates were obtained from BDH, Toronto, Canada. Cell culture and reagents were products of Canadian Life Technologies (GIBCO) Burlington, Ontario, Canada. Ecolite scintillant was obtained from ICN Biochemicals, Costa Mesa, CA.. Lipids standards were obtained from Serdary Research Laboratories, Englewood Cliffs, NJ., USA. All other biochemicals were of analytical grade and obtained from either Fisher Scientific, Edmonton, Canada, Sigma Chemical Co., St. Louis, MO. or CanLab Division of Baxter Co. Winnipeg, Canada. The protocol for differentiation and culturing of murine P19 teratocarcinoma cells into cardiac myocytes was performed as described . P19 cells (5 x 105 cells/ml) were placed into 60 mm bacterial dishes, 1% dimethylsulfoxide (DMSO) was added and incubation continued for 48 h. The cells began to aggregate at this point. Cells were then transferred to a 100 mm bacterial dish and 1% DMSO was added for another 48 h. The cells in these 100 mm bacterial dishes differentiated into the cardiac cell lineage within eight days. At various days (0 --8) post DMSO addition cells were harvested and mRNA expression of GATA-4, BNP, alphaMHC, betaMHC, troponin C and tubulin were determined using quantitative RT-PCR analysis as described . A 10% homogenate from rat hearts or P19 cells was prepared in buffer (0.25 M sucrose, 0.145 M NaCl, 10 mM Tris-HCl, pH 7.4) and centrifuged for 10 min at 600 x g (Beckman J2-H with JA-20 rotor). The resulting pellet was washed once, resuspended in 5 ml buffer by 15 strokes of a hand-held Dounce (tight fitting) tissue grinder and designated the crude nuclear fraction. The supernatant from the first centrifugation was centrifuged at 10,000 x g for 10 min. The resulting pellet was resuspended in 1.5 ml buffer as described above and used as the source of mitochondrial fraction for enzyme assays. Protein in this fraction was determined by the method of Bradford . Phospholipase A2 (PLA2) was determined as described using phosphatidyl [14C]glycerol as substrate . CL synthase and phosphatidic acid (PA):CTP cytidylyltransferase activities were determined as described . MLCL AT activities were determined as described . Mitochondrial fractions (50 mug) were incubated for 30 min at 25C in 50 mM Tris-HCL, pH 8.0, 33 muM [1-14C]linoleoyl-Coenzyme A (68,700 dpm/nmol), 0.3 mM MLCL in a final volume of 0.35 ml. The reaction was initiated by the addition of [1-14C]linoleoyl-Coenzyme A substrate and terminated by addition of 3 ml of chloroform:methanol (2:1, by vol). 0.8 ml of KCL was added to facilitate phase separation. The aqueous phase was removed and the organic phase dried under nitrogen and resuspended in 25 mul of chloroform:methanol (2:1, by vol). A 20 mul aliquot was placed on a thin layer plate and CL was separated from other phospholipids in a solvent system containing chloroform:hexane:methanol:acetic acid (50:30:10:5, by vol). The silica gel corresponding to CL was removed and placed in plastic scintillation vials with 5 ml of aqueous counting scintillant. Radioactivity incorporated into CL was determined approximately 24 h later using a liquid scintillation counter. CL content was determined as described . Mitochondrial succinate dehydrogenase activity was determined as described . Students t-test was used for the determination of statistical significance. The level of significance was defined as p < 0.025. List of abbreviations : CL, Cardiolipin; MLCL AT, monolysocardiolipin acyltransferase; MLCL, monolysocardiolipin; PTU, 6-n-propyl-2-thiouracil; PLA2, phospholipase A2; PA, phosphatidic acid; DMSO, dimethylsulfoxide; CTP, cytidine-5'-triphosphate; ATP, adenosine-5'-triphosphate; ADP, adenosine-5'-diphosphate; BNP, B-natriuretic peptide; alphaMHC, alpha myosin heavy chain; betaMHC, beta myosin heavy chain; mRNA, messenger ribonucleic acid; STZ, Stre ptozotocin Authors contribution : Mr. William A. Taylor intitiated the experimental and edited the manuscript. Dr. Fred Y. Xu performed experimental studies related to the P19 cells and edited the manuscript. Mr. Brian J. Ma performed experimental studies related to the hyperinsulinemic rat model and edited the manuscript. Dr. Thomas C. Mutter performed experimental studies related to the hypothyroid rat model. Mr. Vernon W. Dolinsky initiated experimental studies related to the hypothyroid rat model and edited the manuscript. Prof. Grant M. Hatch conceived of the study, participated in its design and coordination, wrote and edited the manuscript. All authors have read and approved the final manuscript. Backmatter: PMID- 12052259 TI - The role of the Zn(II) binding domain in the mechanism of E. coli DNA topoisomerase I AB - Abstract | Background | Escherichia coli DNA topoisomerase I binds three Zn(II) with three tetracysteine motifs which, together with the 14 kDa C-terminal region, form a 30 kDa DNA binding domain (ZD domain). The 67 kDa N-terminal domain (Top67) has the active site tyrosine for DNA cleavage but cannot relax negatively supercoiled DNA. We analyzed the role of the ZD domain in the enzyme mechanism. Results | Addition of purified ZD domain to Top67 partially restored the relaxation activity, demonstrating that covalent linkage between the two domains is not necessary for removal of negative supercoils from DNA. The two domains had similar affinities to ssDNA. However, only Top67 could bind dsDNA with high affinity. DNA cleavage assays showed that the Top67 had the same sequence and structure selectivity for DNA cleavage as the intact enzyme. DNA rejoining also did not require the presence of the ZD domain. Conclusions | We propose that during relaxation of negatively supercoiled DNA, Top67 by itself can position the active site tyrosine near the junction of double-stranded and single-stranded DNA for cleavage. However, the interaction of the ZD domain with the passing single-strand of DNA, coupled with enzyme conformational change, is needed for removal of negative supercoils. Keywords: Background : Escherichia coli DNA topoisomerase I is a representative example of type IA DNA topoisomerase (for reviews, see refs ). Its major biological role in the bacterial cell is the removal of excessive negative supercoils from DNA to maintain the DNA at optimal superhelical density along with DNA gyrase . The enzyme has a molecular weight of 97 kDa and the active site tyrosine responsible for DNA cleavage is found in the 67 kDa N-terminal transesterification domain. The structure of this 67 kDa domain has been determined by X-ray crystallography to be torus-like, indicating the need for protein conformational change for strand passage to take place after DNA cleavage . Relaxation activity requires the presence of the Zn(II) binding tetracysteine motifs found between the 67 kDa N-terminal domain (Top67) and the 14 kDa C-terminal single-stranded DNA binding domain . The three tetracysteine motifs do not form a stably folded structure on its own, but when combined with the 14 kDa C-terminal domain, forms a stably folded 268 amino acid DNA binding domain (ZD domain) that has higher affinity for single-stranded DNA than the 121 amino acid 14 kDa C-terminal region by itself . Recent sequence and structural analysis suggests that the 14 kDa domain is evolutionarily related to the three tetracysteine motifs and belongs to the zinc ribbon family . The ZD domain in E. coli topoisomerase I probably evolved from a domain that binds five Zn(II) originally. Figure 1 | Domain organization of E. coli DNA topoisomerase I. Domain organization of E. coli DNA topoisomerase I. Removal of negative supercoils from DNA by bacterial type IA topoisomerase involves the following steps: (1) binding of the enzyme to the junction of double-stranded and single-stranded DNA ; (2) cleavage of a single-strand of DNA near the junction with cleavage sequence preference of a cytosine in the -4 position to form the covalent intermediate ; (3) conformational change of the covalent enzyme-DNA complex to result in physical separation of the 5' phosphate covalently linked to the active tyrosine, and the 3' hydroxyl of the cleaved DNA; (4) passage of the complementary single strand through the break; (5) enzyme conformational change to bring the 5' phosphoryl end back into the proximity of the 3' hydroxyl group of the cleaved DNA; (6) religation of the phosphodiester bond. Although it is known that the ZD domain can function as a DNA binding domain, its exact role in these individual steps of removal of a negative superhelical turn from DNA by E. coli topoisomerase I remains to be defined. Using purified 67 kDa transesterification domain and 30 kDa ZD domain, results from experiments described here provide new insight into the action of these two individual domains in the enzyme mechanism. Results : Partial restoration of relaxation activity from mixing of Top67 and ZD domains | As reported previously , the N-terminal transesterification domain Top67 by itself did not exhibit any relaxation activity when assayed with negatively supercoiled plasmid DNA . The 30 kD C-terminal ZD domain also had no relaxation activity by itself, as expected. Partial relaxation of the input supercoiled DNA was detected when Top67 was mixed with the ZD domain prior to addition of DNA. A ratio of 2 ZD molecules added for each Top67 was found to be sufficient for maximum relaxation activity, with no increase in activity when higher ratio of ZD/Top67 was used (data not shown). The specific activity observed under this optimized condition was still about 10 fold lower than that of the intact enzyme. Analysis of the time course of relaxation with 6 pmoles of topoisomerase I or top67 reconstituted with ZD showed that negative supercoils were removed at a much slower rate by the reconstituted activity. Figure 2 | Partial restoration of relaxation activity by complementation of Top67 and ZD domains. Partial restoration of relaxation activity by complementation of Top67 and ZD domains. (a). Agarose gel electrophoresis was carried out to analyze the relaxation reaction products after 1 h of incubation. Lane 1: supercoiled plasmid DNA with no protein added; lane 2: Top67 alone; lane 3: ZD alone. Lanes 4 --7 (Top67 reconstituted with ZD) and lanes 8 --11 (topoisomerase I) have 6, 1.2, 0.24 and 0.05 pmoles of proteins added. (b). Time course of relaxation reaction catalyzed by 6 pmoles of topoisomerase I, or Top67 reconstituted with ZD Top67 and ZD domains have comparable binding affinities to single-stranded DNA but significantly different affinities for double-stranded DNA | The gel mobility shift assay was used to compare the binding affinities of Top67 and the ZD domain to a 5' end-labeled single-stranded oligonucleotide 35 base in length. As shown in Figure , these two domains had similar affinities for binding to the single-stranded substrate. The half maximal binding values based on the average of results from three different experiments were 0.02 muM for Top67 and 0.04 muM for the ZD domain. However, with the same oligonucleotide in a duplex form , Top67 exhibited much higher affinity (half maximal binding value = 0.07 muM) than the ZD domain (half maximal binding value > 5 muM). Figure 3 | Binding of Top67 and ZD to single-stranded and double-stranded DNA. Binding of Top67 and ZD to single-stranded and double-stranded DNA. The gel mobility shift assay was used to compare the binding affinities. The substrates used are (a): single-stranded 5'GAAAACTCACAGGAAGCGGCCGAAGCGATTCGTCC 3'; (b): the same labeled strand of hybridized to its complementary strand. Open circles: Top67; solid circles: ZD. Top67 can recognize cleavage sites preferred by E. coli DNA topoisomerase I | Previous studies have shown that E. coli DNA topoisomerase I cleavage of single-stranded DNA occurs with selectivity for sites with the C nucleotide base at the -- 4 position and that the enzyme preferentially cleaves at junctions of double-stranded and single-stranded DNA . Several different 5'-end labeled substrates were prepared and used in cleavage assays to compare the cleavage sites selected by Top67 versus topoisomerase I. The results showed that with single-stranded substrates, Top67 also preferred cleavage sites with a C nucleotide base at the -4 position as reported for most of the type IA topoisomerases . There were some differences from topoisomerase I in the relative distribution of cleavage products among the potential cleavage sites . Top67 appeared to be more non-discriminatory in selection of the possible cleavage sites with the C nucleotide in the -4 position. Addition of the ZD domain had no effect on the cleavage selectivity of Top67. A substrate with both single-stranded and double-stranded regions was constructed to mimic such junction in negatively supercoiled DNA. Top67 and topoisomerase I recognised the same cleavage site on this substrate . Maximal yield of cleavage products was obtained for both Top67 and topoisomerase I within seconds after mixing of the enzyme and DNA so any potential difference in cleavage rates between the Top67 and topoisomerase I is unlikely to account for the difference in relaxation efficiency. Figure 4 | Cleavage selectivity of topoisomerase I and Top67. Cleavage selectivity of topoisomerase I and Top67. This was analyzed using single-stranded 32mer (a), 31mer (b), or substrate with both single- and double-stranded regions (c). For (a) and (b), lane 1: no protein added, lane 2: topoisomerase I, lane 3: Top67, lane 4: Top67 mixed with ZD. For (a), lane 5: ZD alone. For (c), lane 1: topoiosmerase I, lane 2: Top67, lane 3: no protein added, lane 4: DNase I digestion pattern. Top67 cleavage sites are religated upon addition of high salt and Mg2+ | To test the religation capability of Top67, a 5'-end labeled oligonucleotide 61 base in length was first incubated with the enzyme in low ionic strength buffer to allow formation of the cleaved complex. Sodium chloride concentration was then increased to 1 M to induce reversal of cleavage and dissociation of the enzyme from the DNA. We observed that more complete and consistent reversal of cleavage was obtained with both topoisomerase I and Top67 if a low concentration of Mg2+ (4 mM) was also added with the NaCl. This is consistent with an early observation of dissociation of the enzyme-DNA complex in high salt upon addition of Mg2+. It has also been reported that addition of Mg2+ was apparently not required for observation of this reversal of cleavage. However, it is possible that some enzyme preparations may contain bound Mg2+ and the low concentration of bound Mg2+ might have been sufficient for reversal of cleavage, as postulated previously to explain the data . The results of this cleavage reversal experiment indicated that the ZD domain was not required for efficient reversal of cleavage and Top67 could carry out religation of cleaved DNA. Again the reversal of cleavage was complete for both Top67 and topoisomerase I within seconds after the addition of high salt and Mg2+ even when the reactions were carried out on ice (data not shown) so the lack of relaxation activity by Top67 is unlikely to be due to deficiency in religation. Figure 5 | Reversal of DNA cleavage by topoisomerase I and Top67. Reversal of DNA cleavage by topoisomerase I and Top67. A 5'-end labeled 61mer was used as substrate. C: no enzyme added. Lane 1: enzyme cleavage reaction stopped with SDS; Lane 2: enzyme cleavage reaction incubated with 1 M NaCl before SDS treatment; Lane 3: enzyme cleavage reaction incubated with 1 M NaCl and 4 mM MgCl2 before SDS treatment. The ZD domain is not required for catenation of double-stranded DNA circles | E. coli topoisomerase I can catalyze catenation of double-stranded DNA circles if the molecules contain single-strand scissions . To test if the Top67 can carry out double-stranded DNA passage at enzyme cleavage sites across from the DNA nicks, the yield of DNA catenanes were compared with that obtained with topoisomerase I. In contrast to the relaxation activity, the catenating activity of Top67 shown in figure was as efficient as that of full-length topoisomerase I, and the addition of the ZD domain had no effect . The rate of catenane formation for Top67 alone was similar to that of topoisomerase I . This catenation activity observed with topoisomerase I and Top67 was unlikely to be due to contaminating topoisomerase III activity since it was not observed with the ZD domain purified under almost identical procedures and a site-directed mutant with substitution of the active site Tyr319 by phenylalanine also did not exhibit this activity . Figure 6 | Catenation of nicked double-stranded DNA circles by topoisomerase I and Top67. Catenation of nicked double-stranded DNA circles by topoisomerase I and Top67. (a). Phage PM2 DNA circles with one or more single-strand scissions were incubated with 5 pmoles of proteins for 1 h. Lane 1: no enzyme added; Lane 2: topoisomerase I; Lane3: Top67; Lane 4: Top67 and ZD domain; Lane 5: ZD domain; Lane 6: mutant with Y319F substitution. (b). Aliquots of the reaction for topoisomerase I and Top67 were removed at different time points to analyse the time course of catenation. Discussion : There are two homologous type IA topoisomerases present in E. coli. Topoisomerase III has potent DNA decatenating activity for resolution of plasmid DNA replication intermediates, but much weaker relaxation activity than topoisomerase I . To exhibit maximal relaxation activity, topoisomerase III requires high temperature (52C) along with low magnesium and monovalent ion . In contrast, E. coli topoisomerase I was not active in the in vitro assay for resolution of plasmid DNA replication intermediates . Removal of the C-terminal 49 amino acids from the 653 amino acid topoisomerase III protein resulted in drastic reduction of catalytic activity . Fusion of the carboxyl-terminal 312 amino acid residues of E. coli topoisomerase I, which includes the entire ZD domain, onto the 605 N-terminal amino acids of topoisomerase III generated a hybrid topoisomerase that has relaxation activity resembling topoisomerase III along with weak decatenating activity . Although preferring single-stranded DNA as binding substrate, topoisomerase I had been shown to also bind double-stranded DNA , but there is no data available to indicate which domain in the enzyme is responsible for this interaction. The experiments described here measured directly the interaction of the ZD domain with both double-stranded and single-stranded DNA substrates. ZD domain was found to bind to single-stranded DNA, but not double-stranded DNA, with high affinity. This result indicated that with regard to the mechanism of E. coli topoisomerase I, the ZD domain was likely to function as a single-stranded DNA binding domain instead of having double-stranded DNA binding function as previously suggested . Even though Zn(II) binding transcription factors that recognise specific double-stranded DNA are well represented in eukaryotes , there are also numerous examples of Zn(II) coordination being required for interaction with single-stranded nucleic acid or damaged DNA with single-strand characteristics . The effect of removal of the ZD domain on the individual step of enzyme action was also investigated using Top67. The results indicated that Top67 was effective in binding to both double-stranded and single-stranded DNA. As a result, Top67 could position itself in the absence of ZD domain at the junction of double- and single-stranded DNA for subsequent DNA cleavage, as previously observed for intact topoisomerase I . Reversal of DNA cleavage also took place readily with Top67 upon addition of 1 M NaCl and 4 mM MgCl2. The ZD domain also was not required for selectivity of a cytosine in the -4 position relative to the cleavage sites. Despite its ability to recognise the DNA substrate and carry out DNA cleavage-religation, Top67 by itself cannot catalyze change of linking number in the relaxation of supercoiled DNA. The single-strand DNA substrate designated for the ZD domain in the catalytic mechanism of the enzyme may be the strand of DNA complementary to the strand first cleaved by the enzyme to form the covalent complex. This interaction with the passing strand of DNA would not be needed for the first two steps of enzyme mechanism up to the formation of the covalent complex. Our results showed that adding the purified ZD domain partially restored the relaxation activity. Therefore the ZD domain can supply the function that is missing in Top67 even when the two domains are not covalently linked. However, the resulting relaxation activity is much less efficient than that of the intact enzyme, suggesting that coordinated actions of the two domains are required for efficient removal of negative supercoils from DNA. The requirement of specific protein-protein interactions between the two domains could also account for the weak relaxation activity observed for the hybrid topoisomerase with ZD linked to topoisomerase III sequence . This proposed role for the ZD domain in interacting with the passing single-strand of DNA is also supported by the observation that there is no difference between Top67 and intact topoisomerase I in the formation of catenanes. This reaction involves passage of another double-stranded DNA circle, instead of the complementary DNA strand through the break generated by DNA cleavage so the ZD domain would not be expected to play any significant role. High concentration of DNA substrate is required to favor formation of catenanes catalyzed by topoisomerase I, and the enzyme also has to be present in higher concentration compared to the relaxation reaction. The double-stranded DNA-binding activity in E. coli topoisomerase III required for highly efficient decatenation activity is attributed to a 17-amino-acid residue with no counterpart in E. coli topoisomerase I . It may be required for interaction with the passing double-strand of DNA in the decatenation mechanism. The presence of this decatenation loop instead of the Zn(II) binding ZD domain in topoisomerase III may account for the dominance of the decatenation activity over the relaxation activity. Based on these results, we propose a model for the relaxation of supercoiled DNA by E. coli topoisomerase I modified from previous versions that have a number of common features but differ most significantly in the role of the Zn(II) binding domain . In this model, the subdomains in Top67 is responsible for interacting with the G-strand of DNA both upstream and downstream of the cleavage site. The ZD domain interacts with the passing single-strand DNA to be transported (T-strand). After cleavage of the DNA gate strand which becomes covalently linked to Tyr319 on Top67 (step 2), protein conformational change involving both Top67 and the ZD domain increases the distance between the covalently bound 5' phosphate and non-covalently bound 3' hydroxyl of the cleaved DNA gate strand while the passing DNA strand (T-strand) is guided through the "gate" via interaction with the ZD domain (step 3) to lead to change in linking number. A second enzyme conformational change positions the cleaved DNA ends for religation (step 4). The ZD domain can still interact with the T-strand of DNA even when not linked to Top67 in the same polypeptide, but efficiency of catalysis is reduced as a result, probably due to loss of coordinated action by the two domains. The presence of the ZD domain may enhance the transition of Top67 from a closed conformation to a more open conformation so that strand passage can take place through the "DNA gate". Previous data showed that although Zn(II) binding is not absolutely required for formation of the cleaved complex, it increased the amount of cleaved complex that can be isolated . When linked to Top67, the ZD domain also has some influence on the cleavage site selections. It has previously been observed that a mutation in the Zn(II) binding motif can affect the cleavage site selectivity of topoisomerase I even though Top67 by itself can recognize both the cytosine in the -4 position and the junction of single- and double-stranded DNA. To gain further details for this model of enzyme action, we are characterizing the protein-protein interactions between the Top67 transesterification domain and the ZD domain, as well as the protein conformational changes that can take place when the enzyme interacts with DNA substrate. Figure 7 | Model for removal of a negative supercoil by E. coli DNA topoisomerase I. Model for removal of a negative supercoil by E. coli DNA topoisomerase I. Subdomains I, II, III, IV found in the crystal structure of Top67 (4) are illustrated schematically along with the potential site for ZD domain (Z). G-strand: DNA strand cleaved to provide "DNA Gate". T-strand: DNA strand to be transported through the "DNA Gate". The hyperthermophilic topoisomerase I from Thermotoga maritima has been shown to coordinate one Zn(II) with a unique tetracysteine motif Cys-X-Cys-X16-Cys-X-Cys but Zn(II) binding is not required for relaxation activity . The sequence of this unique tetracysteine motifs is somewhat different from those present in other type IA topoisomerases in that the other tetracysteine motifs always had at least two amino acids separating the pairs of cysteines (Cys-X2-11-Cys), instead of just one amino acid (Cys-X-Cys) in T. maritima topoisomerase I . Therefore the structure and function of the single Zn(II) binding motif in T. maritima may differ from the multiple Zn(II) binding motifs in E. coli topoisomerase I. Direct interaction between DNA and the T. maritima Zn(II) binding motif has not been demonstrated. It has been suggested that the mechanisms of these two enzymes may be different . Direct interaction between the enzyme and the passing strand may not be necessary for the T. maritima topoisomerase I activity. The relaxation and decatenation activities of T. maritima topoisomerase I appear to be significantly more efficient than those of the E. coli topoisomerase I . Based on their primary sequences, a number of bacterial topoisomerase I enzymes do not appear to coordinate any Zn(II) with tetracysteines motifs while other type IA topoisomerase has up to 4 tetracysteine motifs . The topoisomerase I from Mycobacterium smegmatis has been demonstrated biochemically not to bind Zn(II) . In contrast, mutation disrupting the fourth Zn(II) motif of Helicobacter pylori topoisomerase I abolished enzyme function in vivo. Therefore there may be significant differences in the mechanisms of type IA topoisomerases from different organisms with respect to requirement of Zn(II) binding for relaxation activity. There is also another possible explanation for the varied number of tetracysteine motifs and requirement of Zn(II) for relaxation activity found in different type IA topoisomerases. The 14 kDa C-terminal region of E. coli topoisomerase I has been classified based on its structure to be in the Zn-ribbon superfamily [SCOP release 1.50, 7] even though it does not bind Zn(II). It also has high affinity for binding to single-stranded DNA on its own when separated from the three tetracysteine motifs . Based on the structural and DNA-binding properties of the E. coli topoisomerase I 14 kDa domain, one can conclude that it is possible for a subdomain in topoisomerase I to lose the Zn(II) binding cysteines during evolution and still maintains the Zn-ribbon structure and single-strand DNA binding properties . Finally, the in vivo catalytic activities of eukarytotic type IA topoisomerases, the topoisomerase III from various higher organisms may be related to their sequences. The transesterification domains of these eukaryotic enzymes have high degrees of identity to E. coli DNA topoisomerase III . However, the decatenation loop is not present in the eukaryotic topoisomerase III sequences and to date the decatenation activity has not been demonstrated for these enzymes. The number of potential Zn(II) binding cysteine motifs range from none in S. cerevisiae DNA topoisomerase III to four highly conserved tetracysteine motifs in the beta family of the topoisomerase III enzymes . The Zn(II) domain formed by these tetracysteine motifs may be required for interaction with single-strand DNA in removal of hypernegative supercoils or disruption of early recombination intermediates between inappropriately paired DNA molecules . Conclusions : We have shown that the ZD domain of E. coli DNA topoisomerase I is not required for the substrate recognition and DNA cleavage-religation action of the enzyme. We propose that the ZD domain interacts with the passing single-strand of DNA in the relaxation of negatively supercoiled DNA by this enzyme. Materials and methods : Enzyme and DNA | E. coli DNA topoisomerase I and the ZD domain were expressed and purified as described . To express the 67 kDa N-terminal transesterification domain (Top67), a stop codon at amino acid 598 was introduced into plasmid pJW312 used for topoisomerase I expression by site-directed mutagenesis employing the Chameleon-Mutagenesis kit from Stratagene. Top67 was expressed and purified with the same procedures as topoisomerase I. The oligonucleotides were custom synthesized by Genosys. The single-strand substrates and the top strand of the duplex substrates were labeled at the 5' termini with T4 polynucleotide kinase and gamma32P-ATP. The labeled oligonucleotides were purified by electrophoresis in a 12 or 15% sequencing gel. After elution from the gel slice, the labeled single-stranded oligonucleotides were desalted by centrifugation through a Sephadex G10 spin column. The duplex or heteroduplex substrates were prepared by mixing the labeled top strand with 4 fold excess of the unlabeled bottom strand, heating at 80C for three minutes, cooling to room temperature and purified by electrophoresis in a 20% non-denaturing polyacrylamide gel with TBE buffer. Plasmid pJW312 DNA used in relaxation assay was purified by CsCl centrifugation. Phage PM2 DNA was extracted from infected Pseudoalteromonas espejiana cells and PM2 DNA with one or more single-chain scissions used in the catenation assay was prepared as described . DNA relaxation assay | Top67 and the ZD domains at different concentrations were mixed and incubated at 37C for 10 min before addition to the 0.3 mug of supercoiled plasmid DNA in 20 mul of 10 mM Tris-HCl pH 8.0, 2 mM MgCl2, 0.1 mg/ml gelatin. After incubation at 37C for up to 1 h, the reaction was stopped by addition of 50 mM EDTA and electrophoresed in a 0.7% agarose gel and visualized by ethidium bromide staining as described . Gel mobility shift assay | The proteins were mixed with the 1 pmole of the labeled DNA substrates in 10 mul of 20 mM Tris-HCl pH 8.0, 100 mug/ml BSA, 12% glycerol and 0.5 mM EDTA. The samples were incubated at 37C for 5 min and then loaded onto a 6% polyacrylamide gel and electrophoresed with buffer of 45 mM Tris-borate pH 8.3, 1 mM EDTA. Electrophoresis was carried out at room temperature at 2 V/cm for 2 h. After drying of the gel, bands corresponding to the protein-bound oligonucleotides and unbound oligonucleotides were visualized by autoradiography, excised and counted in a Scintillation counter for quantitation. DNA cleavage assay | The cleavage assays were carried out with 1 pmole of 5' 32P-end labeled DNA substrate and 5 --10 pmoles of topoisomerase I or Top67 in 10 mul of the buffer used for the gel mobility shift assay. After incubation at 37C for up to 20 min, an equal volume of 90% formamide, 10 mM KOH, 0.25% bromophenol blue and 0.25% xylene cyanol was added to stop the reactions. The samples were analyzed by electrophoresis in a 12% sequencing gel followed by autoradiography. Salt and Mg2+ induced reversal of cleavage | The conditions were modified from those described previously . The cleavage reactions were incubated at 37C for 5 min and then divided into three aliquots. The cleavage products were trapped in one aliquot by the addition of SDS to 1%. NaCl (1 M) alone or NaCl with MgCl2 (4 mM) were added to the other aliquots followed by further incubation at 37C for up to 30 min before the addition of SDS. The products were analyzed as described for the cleavage reactions. Catenation of nicked DNA circles | The catenation reaction was carried out with 1.4 mug of nicked PM2 phage DNA in 20 mul of 10 mM Tris-HCl, pH 8.0, 0.1 mM EDTA, 10 mM KCl, 10 mM MgCl2. After incubation at 37C for up to1 h, the reactions were stopped with the addition of 1% SDS and 50 mM EDTA. The products were analyzed as described for the relaxation assay. Authors' contributions : Author 1 (A.A.) carried out all the experiments except the catenation assay. Author 2 (Y.T.) conceived of the study, participated in its design and coordination and carried out the catenation assay. All authors read and approved the final manuscript. Backmatter: PMID- 11996675 TI - Hisactophilin is involved in osmoprotection in Dictyostelium AB - Abstract | Background | Dictyostelium cells exhibit an unusual stress response as they protect themselves against hyperosmotic stress. Cytoskeletal proteins are recruited from the cytosolic pool to the cell cortex, thereby reinforcing it. In order to gain more insight into the osmoprotective mechanisms of this amoeba, we used 1-D and 2-D gel electrophoresis to identify new proteins that are translocated during osmotic shock. Results | We identified hisactophilin as one of the proteins that are enriched in the cytoskeletal fraction during osmotic shock. In mutants lacking hisactophilin, viability is reduced under hyperosmotic stress conditions. In wild type cells, serine phosphorylation of hisactophilin was specifically induced by hypertonicity, but not when other stress conditions were imposed on cells. The phosphorylation kinetics reveals a slow accumulation of phosphorylated hisactophilin from 20 --60 min after onset of the hyperosmotic shock condition. Conclusion | In the present study, we identified hisactophilin as an essential protein for the osmoprotection of Dictyostelium cells. The observed phosphorylation kinetics suggest that hisactophilin regulation is involved in long-term osmoprotection and that phosphorylation occurs in parallel with inactivation of the dynamic actin cytoskeleton. Keywords: Background : Cells steadily face changes of the external osmolarity, to which they have to adapt. To withstand a steep increase in osmolarity, eukaryotic cells activate responses like "regulatory volume increase", accumulation of compatible osmolytes and stimulated expression of stress proteins . Recently, an exception from this scheme has been identified: Dictyostelium cells protect themselves against hyperosmolarity by largely rearranging cellular proteins, whereas no "regulatory volume increase", no accumulation of compatible osmolytes and no change of the expression pattern of the most abundant proteins were observed . Among the translocated proteins identified, cytoskeletal proteins appear to be predominant. In particular, the rearrangement of actin and myosin II to the cell cortex beneath the plasma membrane was shown to constitute a pivotal element of osmoprotection in Dictyostelium. These two proteins form the core of a rigid network resembling a shell-like structure . Conversely, DdLIM, a cytoskeletal protein involved in the formation of protrusions , is depleted from the cytoskeletal fraction under hypertonic conditions, which is in consistence with the rounding up and the retraction of protrusions of cells . 30 --60 min after onset of the shock condition, progressive actin phosphorylation is observed , which presumably leads to the inactivation of the dynamic actin system . The pivotal role of the cytoskeleton in hyperosmotic stress response is also reflected by the fact, that the cytoskeletal proteins beta-actinin and "120 kDa gelation factor" are essential to ensure survival under hyperosmotic conditions . However, the signalling pathways mediating protein translocation and actin phosphorylation largely remain to be elucidated. The best understood signalling pathway involves myosin II translocation: upon hyperosmotic stress an increase in cGMP concentration was observed, which is due to guanylyl cyclase activation. This in turn results in the activation of the kinase, which eventually phosphorylates myosin II heavy chain . Thereby, disassembly of myosin II filaments is induced as a prerequisite for their reassembly at the cell cortex , which is accomplished within 10 min after onset of hyperosmotic shock . In addition, DokA, a homologue of bacterial histidine kinases and cAMP are involved in hyperosmolar signal transduction , however, molecular targets of these signalling pathways during osmoregulation have not been identified yet. Hence, the translocation and phosphorylation of cytoskeletal proteins constitute pivotal osmo-responses in Dictyostelium. To further elucidate this unusual hyperosmolar stress response we attempted to identify additional proteins, which are translocated under hypertonic conditions and investigated, whether protein phosphorylation of a candidate protein is regulated in response to hyperosmolarity. We could identify hisactophilin as such a protein, which is translocated and phosphorylated during hyperosmotic shock. Hisactophilin is a 15 kDa protein in Dictyostelium, that consists of two highly identical isoforms Hisactophilin I and II, which are myristoylated at the N-terminus . The two isoforms exhibit 84% sequence identity and are both myristoylated and distributed between plasma membrane and cytoplasm . Due to this high degree similarity, both isoforms are in this manuscript referred to as hisactophilin, without distinguishing between them. The biochemical properties of hisactophilin, namely actin- and membrane binding have recently shown to be strongly pH-dependent, which is due to the high content of 26 --30% histidine residues . Therefore, a role also as pH-sensor was postulated for this protein . In addition to myristoylation, phosphorylation was shown to be a posttranslational modification of hisactophilin ; the physiological role of hisactophilin phosphorylation however is unknown. Hisactophilin appears to play an essential role in osmoprotection, as hisactophilin null cells are osmosensitive. Results : Hisactophilin is enriched in the cytoskeletal fraction of wild type cells exposed to hyperosmotic shock | To identify proteins, which are translocated to or depleted from the cytoskeleton upon hyperosmotic stress, we used 2-D electrophoresis as a differential method. Wild type cells were subject to hypertonic shock in liquid culture prior to cell lysis. The Triton X-100-insoluble cytoskeletal fraction was isolated and the proteins were separated by 2-D electrophoresis. Samples from three independently grown cell cultures were analyzed in parallel. As control, the same procedure was performed with wild type cells shaken in SPB buffer. Computer analysis of the silver-stained gels with the software package Melani II (Biorad) revealed, that a 15 kDa protein, consisting of several isoforms with a pI of 7.3 --7.5, is enriched in the cytoskeletal fraction of hyperosmotically shocked cells compared to control cells (a typical result is shown in Fig. , upper panels). Correspondingly, reduction of the protein amount in the cytosolic pool was observed under hypertonic conditions compared to the control (Fig. , lower panels), suggesting, that the protein is translocated from the cytosol to the cytoskeleton under hypertonic conditions. Search of protein databases for Dictyostelium proteins exhibiting a molecular weight of 15 kDa and a pI of 7.3 --7.5 resulted in hisactophilin as a candidate cytoskeletal protein, which consists of two isoforms, hisactophilin I and II . The identity of the characterized protein as hisactophilin was demonstrated by immunostaining with a monoclonal alpha-hisactophilin antibody . In addition, analysis of digested peptides of the protein eluted from 2-D gels by mass spectrometry demonstrated the identity of hisactophilin (data not shown). The enrichment of hisactophilin in the cytoskeletal fraction of hyperosmotically shocked cells was also confirmed by analyzing cytoskeletal fractions by 1-D SDS-PAGE followed by immunostaining with an alpha-hisactophilin antibody . As the actin-rich cytoskeleton is primarily found in the cell cortex of hypertonically shocked cells , we investigated, whether a hisactophilin II-GFP fusion protein expressed in wild type cells is translocated to the cell cortex after onset of hyperosmotic shock. In fact, the hisactophilin II-GFP fusion protein was found to form a thick layer beneath the plasma membrane of cells exposed to hypertonicity for 10 min, whereas only a weak staining of the plasma membrane was observed in control cells (data not shown). Figure 1 | Hisactophilin is enriched in the cytoskeletal fraction of hyperosmotically shocked cells. Hisactophilin is enriched in the cytoskeletal fraction of hyperosmotically shocked cells. (A) Hisactophilin is enriched in the cytoskeletal fraction of cells hyperosmotically shocked for 2 h with respect to control cells suspended in SPB buffer for 2 h. (upper frames). Correspondingly, less hisactophilin was found in the cytosol of cells exposed to hypertonicity when compared to cells suspended in SPB buffer (lower frames). Proteins from cytosolic and cytoskeletal fractions were separated by 2-D electrophoresis. The frames show a silver-stained region of the gel corresponding to a molecular weight of 15 kDa and to an IEP of 7.3 --7.5. The predominant spot of each panel is indicated with an arrow. (B) Identification of the protein in (A) as hisactophilin by Western blotting of a typical gel of the cytoskeletal fraction and immunostaining of with a alpha-hisactophilin antibody. (C) The enrichment of hisactophilin in the cytoskeletal fraction of cells exposed to hypertonicity was confirmed by 1-D SDS-PAGE. The cytoskeletal fraction was isolated from cells suspended in SPB buffer (-) or in SPB buffer/400 mM sorbitol (+) for 2 h. The proteins were separated by SDS-PAGE and blotted. Immunostaining was performed with an alpha-hisactophilin antibody (right panel). The membrane was subsequently coomassie-stained (left panel). Abb.: M= marker. Hisactophilin is phosphorylated upon hyperosmotic stress | As phosphorylation is an important regulatory modification of actin and myosin II under hypertonic conditions, we investigated, whether this also applies for hisactophilin. Wild type cells deprived of phosphate were suspended in Mes buffer and were metabolically labeled with 32P-orthophosphate. After 1 h incubation, hyperosmotic shock was initiated and samples were withdrawn at intervals. Hisactophilin was isolated after cell lysis by immunoprecipitation. The samples were analyzed by autoradiography and subsequently by immunostaining with an alpha-hisactophilin antibody. Hisactophilin, that was isolated from cells shocked for 2 h, was found to be phosphorylated, whereas no detectable hisactophilin phosphorylation was observed in control samples isolated from cells suspended in SPB buffer for 1 h or 3 h . Analysis of the kinetics of the phosphorylation reaction revealed, that the phosphorylation does not constitute an acute stress response . Phospho-hisactophilin progressively accumulated 20 --60 min after onset of the shock condition. The result of the densitometric analysis revealed that the kinetics of phospho-hisactophilin can be fitted by a quadratic equation . Figure 2 | Hisactophilin is phosphorylated in wild type cells exposed to hyperosmotic conditions Hisactophilin is phosphorylated in wild type cells exposed to hyperosmotic conditions (A). Cells were metabolically labeled with 32P-orthophosphate prior to hypertonic shock. Hisactophilin was immunoprecipitated from samples withdrawn at T = 0 and T = 2 h. The probe was separated by SDS-PAGE and blotted. Autoradiography (left panels A,B) was performed for 48 h and hisactophilin was subsequently identified by immunostaining (right panels A, B). (+) denotes samples from shocked cells, (-) denotes samples from control cells suspended in SPB buffer. The arrow indicates the position of 15 kDa proteins. (B) The amount of phospho-hisactophilin progressively increased in hyperosmotically shocked cells. The experiment was performed as described in (A). (C) Densitometric analysis of the autoradiography in (B) reveals an exponential increase in phospho-hisactophilin upon hyperosmotic stress: the kinetics can be fitted by a quadratic function displayed in the figure. The densities are given in relative arbitrary units, setting the background as 0. To test the specificity of hisactophilin phosphorylation as a hyperosmotic stress response, wild type cells were labeled with 32P-orthophosphate and were then exposed to various other stress conditions. Subsequently, samples were withdrawn and hisactophilin phosphorylation was detected as described above. The phosphorylation reaction was found to be specific for hypertonic conditions, as other stress conditions as heat stress, acid stress, oxidative stress and energy depletion did not result in a detectable hisactophilin phosphorylation . Figure 3 | Hisactophilin phosphorylation is not a general stress response. Hisactophilin phosphorylation is not a general stress response. Autoradiographies are shown on the left panels, the corresponding immunostainings are displayed on the right panels. (A) Wild type cells were metabolically labeled with 32P-orthophosphate and were then exposed to various stress conditions. Hisactophilin phosphorylation was detected as described in Fig. . : Control (cells shaken in SPB buffer). 2: Cellular Acidification. 3: Heat stress. 4: Energy depletion. 5: Oxidative stress. 6: Hyperosmotic stress. (B) Hisactophilin phosphorylation is dependent on osmolarity, but not on the osmolyte. Wild type cells were metabolically labeled with 32P-orthophosphate and were then exposed to either 200 mM NaCl, 350 mM glucose or 400 mM sorbitol for 1 h. As control, cells were shaken in SPB buffer (1). Hisactophilin phosphorylation was detected as described in Fig. 2. To prove that the phosphorylation of hisactophilin is dependent on osmolarity, but not on the osmolyte used, wild type cells were exposed to hypertonic stress conditions, using either NaCl, glucose or sorbitol as osmolyte. The concentrations of the osmolytes were adjusted to 400 mOsm in the case of sorbitol (400 mM) and NaCl (200 mM) and a slightly reduced osmolarity (350 mOsm) was chosen in the case of glucose. Analysis of hisactophilin phosphorylation in response to the hypertonic stress conditions revealed that phosphorylation occurred to the same extent in the presence of all three osmolytes, indicating that hisactophilin phosphorylation is a hyperosmolar stress response . Serine phosphorylation of Hisactophilin | Recently, it has been demonstrated in vitro using a crude kinase fraction, that hisactophilin is phosphorylated on threonine (95%) and on serine residue(s) (5%) . We addressed the question, whether the phosphorylated residues are identical under hyperosmolar conditions in vivo. We therefore labeled wild type cells with 32P-orthophosphate, exposed them to hypertonic stress and immunoprecipitated hisactophilin. The protein probe was hydrolyzed in the presence of constant boiling HCl and was then separated together with phosphoamino acid standards by 2-D thin layer electrophoresis. 32P-labeled phosphoamino acids originating from phospho-hisactophilin were detected by autoradiography. Identification of the phosphoamino acid was performed by staining the standards with Ninhydrin and subsequent comparison of the staining pattern with the autoradiography. As shown in Fig. , only phospho-serine could be detected under in vivo conditions, whereas no phospho-threonine or phospho-tyrosine was observed, indicating that serine phosphorylation is favored over threonine phosphorylation under hyperosmotic stress conditions in vivo. Figure 4 | Hisactophilin is phosphorylated on serine residue(s) in vivo. Hisactophilin is phosphorylated on serine residue(s) in vivo. Wild type cells were metabolically labeled with 32P-orthophosphate and were then exposed to 400 mM sorbitol for 2 h. Hisactophilin was isolated by immunoprecipitation and was analyzed by 2-D thin layer electrophoresis. Phosphoamino acid standards were detected by Ninhydrin staining (positions indicated by lines) and were compared to the autoradiography. Abbreviation: part. hydrol.=partially hydrolyzed Hisactophilin-deficient cells exhibit reduced viability under hyperosmotic stress conditions | To investigate whether hisactophilin plays a role in osmoregulation, we determined whether survival of his--cells was affected under hypertonic conditions. The viability of his--cells was found to be reduced by 75% 2 h after onset of the stress condition, whereas viability of the wild type cells was not significantly affected . Hence, the presence of hisactophilin is essential to ensure viability under hypertonic conditions. Figure 5 | Viability of his--cells is reduced under hypertonic conditions. Viability of his--cells is reduced under hypertonic conditions. his--cells and wild type cells were suspended in either 400 mM sorbitol or SPB buffer for 2 h. Samples were plated together with bacteria and plaques were counted after 2 --4 days incubation (n = 5). Discussion : Dictyostelium was shown to employ an unusual mechanism to cope with the effects of osmotic stress by rearranging its cellular cortex . As the cells round up, actin and myosin moieties are being redistributed underneath the plasma membrane. In this process the distribution of the actin-associated protein Hisactophilin was investigated in order to gain a better insight into osmo-protective mechanism of the cell. Hisactophilin was found to be enriched in the cytoskeletal fraction of wild type cells exposed to hyperosmotic stress . In addition, a thick layer of Hisactophilin II-GFP was formed at the cell cortex of cells, that were shocked for 10 minutes (data not shown), which correlates with the observation that actin and myosin II are translocated to the cell cortex within 10 minutes after onset of the stress condition . As hisactophilin was shown to enhance actin polymerization in vitro and as cells overexpressing hisactophilin II exhibit an increased F-actin content , this protein could concur in the formation of the rigid filamentous network of actin and myosin II filaments. The essential role of hisactophilin becomes evident when hisactophilin negative cells are investigated under osmotic stress conditions. They show a markedly reduced survival rate in a test were they are first exposed to high osmolarity for 2 hours and are then plated onto agar plates containing bacteria. This procedure involves a volume reduction and a subsequent volume increase caused by the re-dilution in the plating process. It therefore remains to be investigated whether hisactophilin deploys its function in the process of rearranging the cell's cortex upon volume reduction or is essential for reinforcing the cortex in its rounded up state. A role comparably to the family of MARCKS (myristoylated alanine-rich protein kinase C substrate) proteins in mammalian cells was postulated recently . These proteins are also myristoylated at the N-terminus and crosslink actin into a rigid meshwork at the membrane. Upon phosphorylation, MARCKS proteins are rejected from the membrane by electrostatic interactions, which results in the spatial separation of the cytoskeleton from the membrane . In fact, hisactophilin phosphorylation was found to be specifically induced under hypertonic conditions (Fig. , ). However, the phosphorylation reaction is too slow to account for the translocation of the protein within 10 minutes : phospho-hisactophilin progressively accumulated within 60 min with a kinetics strikingly similar to the kinetics of actin phosphorylation under hyperosmotic conditions . This points out the possibility, that actin and hisactophilin cooperate in two distinct phases of osmoregulation. During the first phase, the proteins are translocated within the first 10 minutes of stress which is accompanied by cortical reinforcement. The second phase is characterized by actin and hisactophilin phosphorylation reinforced after about 30 min of hypertonic shock. Actin phosphorylation was shown to correlate with an inactivation of the dynamic actin system . The identical kinetics suggests a similar role for hisactophilin during this process. In addition, the coinciding phosphorylation kinetics raises the possibility of a common regulation of hisactophilin and actin phosphorylation. The signalling pathways regulating actin and hisactophilin phosphorylation are unknown, however, the second messengers involved in osmoregulation identified so far, namely cGMP and cAMP could be excluded as modulators of hisactophilin phosphorylation, as an increase or decrease in these messengers did not affect hisactophilin phosphorylation (data not shown). This suggests the presence of an additional signalling pathway concomitantly regulating a serine kinase specific for hisactophilin and, due to the identical phosphorylation kinetics, a tyrosine kinase specific for actin . A conceivable signalling pathway regulating the hisactophilin translocation is a change in pH, as the biochemical properties of Hisactophilin are strongly pH-dependent: actin-binding is strongly increased in vitro when pH is lowered to = 7.0 . In fact, progressive cytosolic acification was observed during hyperosmotic stress, resulting in a pH drop from 7.5 to 6.8 15 min after onset of the stress condition . Hence, the translocation of hisactophilin to the actin-rich cytoskeleton can be attributed to the cytosolic acidification. Also, the enrichment of the elongation factor 1alpha in the cytoskeletal fraction of hyperosmotically shocked cells can be explained by cytosolic acidification, as enhanced binding of this protein to F-actin at pH 6.5 in vitro has been described recently . These examples are in accordance with the notion of pH changes as "synergistic messenger" . Changes in proton concentration affect the biochemical properties of proteins and can thereby act as signals. However, other signals have to concur to elicit the osmoprotective responses: the translocation of hisactophilin potentially caused by the drop in cytosolic pH is complemented by the relocation of myosin II mediated by cGMP and of actin by a yet unknown mechanism. Therefore, the role of hisactophilin in osmoprotection also exemplifies the numerous interactions of the signalling pathways that are necessary for osmoprotection in Dictyostelium. Conclusions : We could demonstrate that Hisactophilin is both translocated to the cytoskeleton and phosphorylated during hyperosmotic stress in Dicytostelium. The phosphorylation kinetics suggests a slow accumulation of phosphorylated hisactophilin resembling the phosphorylation kinetics for actin, which is known to eventual lead to cytoskeletal inactivation. Second messengers known to play an essential role during osmoprotection are not involved in hisactophilin phosphorylation suggesting a further signalling pathway being involved in the unusual osmoprotection mechanisms in Dictyostelium. Materials and Methods : Growth of Dictyostelium cells | D. discoideum strain AX2-214 (AX-2), referred to as wild type, was cultivated axenically in shaken suspension at 21C and harvested during exponential growth. Clones of HsII-GFP were cultivated as described for AX-2, but under the selection of 20 mug/ml Geneticin (G418). Clones of his--cells were cultivated as described for AX-2, but under the selection of 7.5 mug/ml Geneticin. Hyperosmotic shock in liquid culture | Hyperosmotic shock experiments were performed according to Schuster et al. (1996). Axenically grown Dictyostelium cells were harvested, washed twice in Soerensen Phosphate Buffer (SPB: 2.0 mM Na2HPO4, 14.6 mM KH2PO4, pH 6.0) and were then adjusted to 2*107 cells/ml with SPB buffer (osmolarity: 34 mOsm). Cell density was 1,6*107 cells/ml during the osmoshock, adjusted by the addition of SPB buffer/2 M sorbitol. To the control cells, SPB buffer was added until the same cell density was reached. The cell suspension was shaken for 1 h. SPB buffer/2 M sorbitol was added to a final concentration of 400 mM sorbitol (osmolarity of the buffer: 434 mOsm). The cell suspension was incubated up to 2 hours according to the experiment. Viability of Dictyostelium cells after exposure of the cells to hyperosmotic stress was determined as described . In short, his--cells were shaken in SPB buffer for 1 h prior to the hyperosmotic shock. Samples were withdrawn at intervals and cell suspension corresponding to 200 cells was plated on agar plates together with Klebsiella aerogenes bacteria. Plaques in the bacterial lawn were counted after 2 --4 days. Exposure to stress conditions | To acidify Dictyostelium cells, axenically grown wild type cells were washed twice in SPB buffer and were resuspended in SPB buffer/5 mM propionic acid, pH 6.0. To impose oxidative stress, the cells were incubated in SPB buffer/3 mM H2O2 for 60 min. To study heat stress, the cells were exposed to 30C for 30 min. Energy depletion was achieved by incubating the cells in SPB buffer/50 muM 2,4-dinitrophenol for 60 min. 2-D Gel electrophoresis | Preparation of the cytoskeletal fraction from hyperosmotically shocked cells and separation of the proteins by 2-D gel electrophoresis was performed as described recently . The cytoskeletal fraction was obtained by isolating the Triton X-100-insoluble fraction. The corresponding soluble fraction was defined as cytosolic fraction. 100 mug total protein was applied per gel. Metabolic labeling with 32P-orthophosphate | 50 ml cell suspension of axenically grown wild type cells were transferred to 300 ml phosphate-free medium buffered with 20 mM Mes, pH 6.0. The cells were shaken for 16 h and were subsequently washed twice in ice-cold 20 mM Mes, pH 6.0 (Mes buffer). The cells were adjusted to 2*107 cells/ml with Mes buffer. After incubating the cells for 30 min, 32P-orthophosphate (Amersham Pharmacia) was added to a final activity of 0.25 mCi/ml in the cell suspension. The cell suspension was shaken for 60 min, before the cells were exposed to hyperosmotic shock. Samples of 1 ml were withdrawn at intervals, the cells were washed twice with Mes buffer and were resuspended in 800 mul RIPA buffer (50 mM Tris/HCl, pH 7.5; 150 mM NaCl, 1% TritonX-100; 1% Natriumdeoxycholate; 0.1% SDS; 1 mM DTT; 1 mM EDTA; 100 muM Na-orthovanadate; Complete Protease Inhibitor Cocktail (Boehringer Mannheim)). The cell lysate was immediately frozen in liquid nitrogen. As control, samples were prepared from a cell suspension to which Mes buffer was added instead of Mes buffer/2 M sorbitol. Immunoprecipitation of hisactophilin | Thawed cell lysate from 32P-labeled cells was centrifuged for 20 min at 10000 g (4C). alpha-hisactophilin antibody 54-11-10 was added to the cleared lysate and was incubated for 3 h. Immunoprecipitation was performed by adding fixed Staphylococcus aureus cells (Pansorbin, Calbiochem) preincubated with rabbit-alpha-mouse IgG (Sigma). After 45 min incubation, the immunoprecipitate was washed twice with TBS and was analyzed by SDS-PAGE, Western blotting and autoradiography. Fluorescence microscopy | Axenically grown HisII-GFP cells were shaken in SPB buffer for 1 h at a cell density of 2*107 cells/ml. SPB buffer/2 M sorbitol was added to a final concentration of 400 mM sorbitol (cell density: 1,6*106 cells/ml). An aliquot of the living cells was allowed to adhere to glass cover slips for 10 min. Subsequently, fluorescence microscopic observations were carried out. As control, cells suspended in SPB buffer were analyzed. The equipment consisted in a Leitz Aristoplan fluorescence microscope, set at 450 --490 nm for excitation. Phosphoamino acid analysis | Phosphoamino acid analysis was performed as described . In short, hisactophilin was immunoprecipitated from 32P-labeled cells exposed to hyperosmotic stress for 60 min. Acid hydrolysis of the immunoprecipitate was performed, followed by the separation of the phosphoamino acids by electrophoresis in two dimensions on TLC plates together with phosphoamino acid standards. The plates were analyzed by autoradiography and by the detection of amino acids with Ninhydrin. Miscellaneous | Protein quantification, SDS-PAGE and Western blotting onto PVDF membrane were performed according to the published methods of Bradford , Laemmli and Towbin et al.. Western blots were treated with the mouse monoclonal alpha-hisactophilin antibody 54-11-10 . Antibodies were detected using peroxidase-coupled rabbit-alpha-mouse IgG and the Renaissance system (Du Pont). Prestained Seeblue marker (Novex) was used as a protein molecular weight standard. Backmatter: PMID- 12057022 TI - Training practitioners in preparing systematic reviews: a cross-sectional survey of participants in the Australasian Cochrane Centre training program AB - Abstract | Background | Although systematic reviews of health care interventions are an invaluable tool for health care providers and researchers, many potential authors never publish reviews. This study attempts to determine why some people with interest in performing systematic reviews do not subsequently publish a review; and what steps could possibly increase review completion. Methods | Cross-sectional survey by email and facsimile of the 179 participants in Australasian Cochrane Centre training events between 1998 and 2000. Results | Ninety-two participants responded to the survey (51 percent). Response rate of deliverable surveys was 82 percent (92/112). The remainder of the participants had invalid or no contact information on file. More than 75 percent of respondents felt that the current workshops met their needs for training. The most critical barriers to completion of a Cochrane review were: lack of time (80 percent), lack of financial support (36 percent), methodological problems (23 percent) and problems with group dynamics (10 percent). Conclusions | Strategies to protect reviewer time and increase the efficiency of the review process may increase the numbers of trained reviewers completing a systematic review. Keywords: Background : Increasing emphasis is being placed on evidence-based medicine. The best evidence for treatment interventions comes from systematic reviews of randomised controlled trials, however, systematic reviews with meta-analysis of published and unpublished data are time- and labour-intensive. A major player in the evidence-based medicine movement has been the Cochrane Collaboration, an international organisation committed to 'preparing, maintaining and promoting the accessibility of systematic reviews of the effects of health care interventions.' The Australasian Cochrane Centre is one of the component centres of the Cochrane Collaboration, providing ongoing training and support to people in the Australasian region who prepare systematic reviews that are subsequently published on the Cochrane Library. The ultimate goal of the training program is to increase the number and quality of completed Cochrane reviews, and ensure that they are routinely updated. Currently, training occurs over two days, with the first day focusing on developing a protocol for a systematic review and the second, statistical analysis and interpretation of the review. Training sessions occur several times per year in locations throughout Australasia. In May 2001, the Australasian Cochrane Centre surveyed all of the participants of its training workshops during the years 1998 to 2000. We undertook this survey to: 1. determine why some people who attend Cochrane training workshop do not go on to publish a systematic review; and 2. what steps could increase systematic review completion. Methods : Survey Population and Contact Information | Our survey population included the 179 participants of the Australasian Cochrane Centre's training workshops from 1998 to 2000. Contact information for the year 2000 participants was drawn from the Australasian Cochrane Centre database (initiated in January 2000) of all workshop participants and current Cochrane reviewers in Australasia. Several partial data systems provided contact information for 1998 --1999 participants. Questionnaire | Participants were asked to respond to a cross-sectional survey of 21 'yes' or 'no' questions with a provision for open-ended comments. [See for the questionnaire used in this study.] The format of our internally developed questionnaire did not lend itself to statistical testing for internal consistency. No demographic details were collected, so that responses could be deidentified after receipt of the questionnaire. The initial questionnaire was sent via email or fax to all participants. A second email or fax was sent four weeks later, with a third survey sent via post to those participants who had not responded to the first two requests and had a postal address on file. Analysis | Both quantitative and qualitative methods were used in data analysis. The structured questions were analysed by SPSS for Windows software, release 10.0.5, and results were presented in the form of frequencies and percentages. Content analysis was used to analyse the qualitative responses. Publication status of workshop participants was determined by searching the Cochrane Library by each workshop participants' last name and first initial. Results : Response rate | Thirty-five questionnaires were returned as undeliverable (20 percent), with an additional 32 participants having no contact information on file with the Australasian Cochrane Centre (18 percent). [see Figure ] Of the 112 deliverable questionnaires (63 percent of participants), 54 participants (47 percent) responded after the first two mailings. With the third mailing, an additional 38 participants responded, creating an overall response rate for deliverable surveys of 82 percent. Eighteen people returned the survey without response, which provided us with 74 valid surveys (41 percent of participants) for analysis. Figure 1 | Survey Response of Workshop Participants Survey Response of Workshop Participants Publication status | Approximately 40 percent of these 1998 --2000 workshop participants have published a protocol or review on the Cochrane Library, as of 2001 issue 4. However, due to issues of anonymity of questionnaire response, we are unable to stratify the results based upon publication status. Findings from Questionnaire: Factors that Interfere with Systematic Review Completion | Lack of time emerged as the most critical barrier to completion of Cochrane reviews, with 80 percent of respondents citing this factor. Twenty respondents (36 percent) acknowledged lack of financial support, and twelve (21 percent), lack of institutional encouragement and support, as barriers to completion of their systematic review. Of the 31 people who offered qualitative comments, 11 suggested a need for dedicated time and support, with two needing better time management, and one wanting a grant to supply a research assistant. In a similar tone, seven people stated that other work or personal commitments take priority over Cochrane activities. [see Table ] Table 1 | Barriers for Training in, and Completion of, Systematic Review of the Literature Methodological barriers affected 23 percent of respondents. Three respondents specifically mentioned problems with translation of non-English papers, and two had no access to EMBASE (Excerpta Medica). In addition, one person was frustrated that a similar title for a Cochrane review had been registered, but not completed in a timely manner. Ten percent of respondents had problems with group dynamics, with two stating this factor as a significant delay to publication, and one complaining of contradictory editorial commentary. Thirteen respondents (22 percent) had yet to start a review. Nine respondents (16 percent) were working on a systematic review outside of the Cochrane Collaboration, with one qualitative commentator stating that the employer insisted on publication of the review prior to submission to Cochrane. A substantial minority (38 percent) of respondents felt that follow-up communication, especially via email, would help them complete their review. From the qualitative comments, this email could occur every one to six months, with 11 of the 12 respondents suggesting a time period of every three months or less. Findings from Questionnaire: Training Issues | Scheduling and content of the training workshops appeared to meet the needs of participants. Over 75 percent of respondents felt that the workshops were not too infrequent, too inconvenient, too far away, assumed too much knowledge, or were too long. Qualitative comments stressed the need for ongoing support after training. Three respondents requested a specific contact person, whereas another four asked for support to discuss general problems, statistical analysis issues, help selecting trials, and the creation of a bulletin board-style mentoring system. Three more respondents requested advanced training. In January 2000, the Australasian Cochrane Centre began collecting formal immediate evaluations from workshop participants. Analysis of these data indicates that the average participant self-ranks only slightly above an absolute beginner in systematic review methodology. Mean satisfaction with content, presentations, opportunity to ask questions, and overall satisfaction range between 4.5 to 4.7, on a scale from 1 = highly dissatisfied to 5 = highly satisfied. Finally, 83 percent of respondents stated that their Cochrane training had made them more likely to use the Cochrane Library to answer clinical questions. Only 45 percent, however, felt that their practice patterns had changed. Discussion : Lack of time is by far the most commonly reported barrier to completion of a systematic review. Financial, institutional, personal, and methodological issues are not problematic for the majority of respondents. These findings have important implications for both reviewers and entities commissioning systematic reviews. In the absence of funding for all or, at least, part of a systematic review, it will continue to be difficult for reviewers to do this research by taking more time out of their busy lives. The challenge becomes the ability to support researchers in minimising the time they spend on the process of the review, without affecting quality. Limitations in this study are several. First of all, the subjective nature of this study may have lead reviewers to state that lack of time was a barrier to completion of the project, when another reason, such as change in professional priorities was the 'true' barrier. Second, searching publications by last name and first initial may lead to either overstatement or understatement of publication status, as reviewers may have common last names, have changed their last names, or use a different first initial to the first name they use routinely. Finally, we had a large number of invalid addresses on file at the Australasian Cochrane Centre. The disruption of a move of the Australasian Cochrane Centre from Adelaide to Melbourne in 1999, with resultant staff turnover and loss of records, is responsible for much of our difficulty. However, we still had invalid contact information for almost one-quarter of the 2000 participants, whose training occurred less than 18 months prior to this survey. This lack of contact information is problematic as this survey suggests email communication could facilitate review completion. Strategies we plan to implement to improve the currency of our database include: 1. discussion at training events of the importance of informing the Australasian Cochrane Centre of change of address, 2. reminder notices on newsletters, and 3. liaison with other Cochrane entities. The year 2000 also saw a change in enrolment policy for the training workshops, allowing people to attend the course only if they had a title or protocol registered with the relevant Cochrane entity. Therefore, our results are likely to be more representative of recent participants of our program, committed to the Cochrane system. Conclusions : Lack of time, not skill, is the most common barrier to completion of a systematic review for authors who have attended such training given by the Australasian Cochrane Centre. One solution lies in advancements in methodology to make systematic reviews more efficient, such as automating aspects of the review or providing authors with specialised, centralised services in literature searches, project management, and the like. We expect that these advances in systematic review methodology, increased capacity for completing systematic reviews in Australasia, and realistic expectations regarding required investment of time for valid evidence will lead to an increased number and quality of systematic reviews published on the Cochrane Library. Ultimately, consumers and practitioners of health care around the globe should benefit from this increased knowledge. Competing interests : All authors were under the employ of the Australasian Cochrane Centre at the time of this survey. Authors' contributions : The authors contributed equally to the design and analysis of this project prior to Dr. Silagy's death in December 2001. Doctors Piehl and Green are responsible for all subsequent editorial changes. Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12028592 TI - The regulation of protein synthesis and translation factors by CD3 and CD28 in human primary T lymphocytes AB - Abstract | Background | Activation of human resting T lymphocytes results in an immediate increase in protein synthesis. The increase in protein synthesis after 16 --24 h has been linked to the increased protein levels of translation initiation factors. However, the regulation of protein synthesis during the early onset of T cell activation has not been studied in great detail. We studied the regulation of protein synthesis after 1 h of activation using alphaCD3 antibody to stimulate the T cell receptor and alphaCD28 antibody to provide the co-stimulus. Results | Activation of the T cells with both antibodies led to a sustained increase in the rate of protein synthesis. The activities and/or phosphorylation states of several translation factors were studied during the first hour of stimulation with alphaCD3 and alphaCD28 to explore the mechanism underlying the activation of protein synthesis. The initial increase in protein synthesis was accompanied by activation of the guanine nucleotide exchange factor, eukaryotic initiation factor (eIF) 2B, and of p70 S6 kinase and by dephosphorylation of eukaryotic elongation factor (eEF) 2. Similar signal transduction pathways, as assessed using signal transduction inhibitors, are involved in the regulation of protein synthesis, eIF2B activity and p70 S6 kinase activity. A new finding was that the p38 MAPK alpha/beta pathway was involved in the regulation of overall protein synthesis in primary T cells. Unexpectedly, no changes were detected in the phosphorylation state of the cap-binding protein eIF4E and the eIF4E-binding protein 4E-BP1, or the formation of the cap-binding complex eIF4F. Conclusions | Both eIF2B and p70 S6 kinase play important roles in the regulation of protein synthesis during the early onset of T cell activation. Keywords: Background : The initiation of translation of mRNAs is an important control point in protein synthesis in eukaryotes and requires a set of initiation factors (eIFs). The cap-binding protein eIF4E recognises the 5'cap-structure of the mRNA, and is a component of the eIF4F complex consisting of eIF4E, eIF4G, a scaffolding protein , and eIF4A, an RNA helicase . Any secondary structure in the 5'untranslated region of the mRNA is thought to be unwound by eIF4A together with eIF4B or eIF4H . The 40S subunit of the ribosome binds to the eIF4F complex through an association between eIF4G and eIF3, which interacts directly with the 40S ribosomal subunit. The preinitiation complex, containing the 40S ribosomal subunit, eIF4F, eIF4B, and Met-tRNAi*eIF2*GTP, scans the 5'UTR until the AUG start codon is located. The subsequent hydrolysis of the GTP bound to eIF2 is promoted by eIF5, after which eIF2*GDP leaves the ribosome. The 60S ribosomal subunit can then join and the 80S complex is formed. eIF2 in the GDP-bound state is inactive and, in order to return to the active form again, the GDP is exchanged for GTP in a step promoted by the guanine nucleotide exchange factor eIF2B. The next stage in the translation process, the elongation step, can be regulated via changes in the activity of eEF2 . Phosphorylation of eEF2 at Thr56 results in its complete inactivation . Human primary T-cells are metabolically quiescent, with little ongoing DNA, RNA or protein synthesis . The low protein synthesis rate in quiescent T cells is associated with low levels of initiation factors in these cells. The rate of protein synthesis increase 2 --4 fold after 4 h of mitogenic stimulation , and it has been reported that the mRNA and protein levels for several translation initiation factors increased during T cell activation. The mRNA levels of eIF4A, eIF2alpha, and eIF4E increased rapidly after stimulation . However, the increase in the levels of the corresponding protein lagged significantly behind. It is therefore likely that increased levels of translation factors contribute to the pronounced stimulation of protein synthesis that occurs during T cell activation at later times, while modulation of the activity of several translation initiation factors e.g. by phosphorylation or association with binding proteins is important in the early phase of T cell activation . Increased phosphorylation of eIF4E in T lymphocytes has been reported under several conditions. Activation of quiescent mature porcine peripheral blood mononuclear cells with phorbol 12-myristate 13-acetate (PMA) or concanavalin A or stimulation of human primary T cells with PHA , PMA, or PMA plus ionomycin caused a rapid increase in the phosphorylation of eIF4E. Similarly, stimulation of the T cell receptor in the human leukaemic T cell line Jurkat with OKT-3, or treatment with PMA, increased eIF4E phosphorylation , and a significant increase in the amount of eIF4F complexes was also detected. The activity of eIF4E can also be modulated by its association with eIF4E-binding proteins, of which 4E-BP1 is the best-studied. Phosphorylation of 4E-BP1 leads to its dissociation from eIF4E, leaving eIF4E free to bind eIF4G and form eIF4F complexes . In a murine cytotoxic T cell line, interleukin-2 induced the phosphorylation of 4E-BP1 . 4E-BP1 is present in human primary T lymphocytes and becomes phosphorylated in response to PMA or PMA plus ionomycin . In several cell lines, an increase in eIF2B activity coincides with an increase in protein synthesis . One mechanism to regulate the activity of eIF2B is via phosphorylation of its epsilon-subunit (eIF2Bepsilon) by GSK-3, which causes a decrease in eIF2B activity . Stimulation of T cells with PMA plus ionomycin caused a rapid rise in eIF2B activity, which coincided with inactivation of GSK-3 , suggesting a role for dephosphorylation of eIF2Bepsilon. The activity of eIF2B can also be modulated by phosphorylation of the alpha-subunit of eIF2. eIF2 phosphorylated in its alpha-subunit acts as a competitive inhibitor of eIF2B . Stimulation of T cells with PHA did not cause significant changes in the phosphorylation state of eIF2alpha , excluding this mechanism of regulation under this condition. In this study we used the antibodies, alphaCD3 and alphaCD28, to activate resting human primary T lymphocytes. Engagement of alphaCD3 activates the T cell receptor, while cross-linking of alphaCD28 with the B7 receptor will supply a co-stimulatory signal, which is required for full activation of a resting T cell . We have studied the effects of T cell activation on protein synthesis and on the activities and/or phosphorylation states of several translation initiation factors. Furthermore, the signalling pathways involved in these changes have been investigated. Results : Activation of primary T cells with a physiological stimulus increased protein synthesis | We activated T cells with the antibodies, alphaCD3 and alphaCD28, for up to 24h and measured protein synthesis and the activation of several signalling pathways that are important for the regulation of translation factors and that have been shown to increase after activation of T cells . Figure 1 | Activation of T cells with alphaCD3 and alphaCD28. Activation of T cells with alphaCD3 and alphaCD28.A. Primary T lymphocytes were activated with both alphaCD3 and alphaCD28 for 4, 8, 16 or 24 h and for the last 45 min 10 muCi/ml of [35S]-labelled methionine was present. The experiment was performed in duplicate. Incorporation of [35S]-labelled methionine into equal amounts of protein was measured. Protein synthesis in control cells was set at 100%. Methionine incorporation ranged between 1000 and 2000 cpm per 50 mug of protein. (control cells at t = 0 h and t = 24 h are not significantly different, n = 3). B. T cells were activated by alphaCD3 and alphaCD28 or with PMA. Cells were activated for 30 min, harvested and 50 --80 mug of lysate was analyzed by SDS-PAGE and Western blotting. Antibodies that recognize the phosphorylated form of either ERK (pp42) or p38 MAPK (pp38) were used. Even loading of the gel was verified using anti-ERK2 (p42). Similar results were obtained in three sets of experiments. C. PKB activity was measured as described in Materials and Methods (p<0.05, n=4). The rate of protein synthesis in resting T cells is low, and activation of the cells with alphaCD3 and alphaCD28 led to a substantial increase in the incorporation of [35S]-methionine into protein. Within 24 h of activation, the rate of protein synthesis was increased 6-fold , depending on the blood donor. After 30 min of activation, phosphorylation of ERK and p38 MAPK, and the activity of PKB were measured. Phosphorylation of ERK2 and p38 MAPK increased already after 5 min and reached a maximum after 30 --60 min of treatment (data not shown). After 30 min of treatment, a clear phosphorylation of ERK and p38MAPK was detected . Treatment with PMA, a potent activator of PKC, was used as a positive control. PMA induced, in particular, a greater extent of phosphorylation of ERK than alphaCD3 plus alphaCD28. For p38 MAPK phosphorylation, the difference between these stimuli was less pronounced . An 1.5 fold increase in PKB activity was detected within 30 min of activation of T cells by alphaCD3 and alphaCD28. The increase in protein synthesis and the stimulation of various signalling pathways indicated that treatment of T cells with alphaCD3 and alphaCD28 led to activation of the cells. The cellular protein levels of eIF4E and eIF2Bepsilon do not change during the early phase of T cell activation | Previous studies using mitogenic stimuli showed that the levels of several initiation factor proteins increase later (>16 h) following T cell activation and probably contribute to the increase in protein synthesis . To examine whether activation of T cells with alphaCD3 and alphaCD28 also affected initiation factor levels, the amounts of the cap-binding protein, eIF4E, and of the catalytic subunit of the eIF2B complex, eIF2Bepsilon, were assessed at different time points . The amounts of eIF4E and eIF2Bepsilon protein each remained constant during the first 6 h. Figure 2 | The amount of eIF4E and eIF2Bepsilon protein did not change in the early phase of T cell activation. The amount of eIF4E and eIF2Bepsilon protein did not change in the early phase of T cell activation. T cells were activated with alphaCD3 and alphaCD28 for the indicated times. In each case, 600 mug of protein was used for an m7GTP Sepharose pull down to detect the amount of eIF4E or used in an immunoprecipitation reaction with alphaeIF2Bepsilon coupled to protein G to detect eIF2Bepsilon. The pull downs were analyzed by SDS-PAGE and Western blotting. The experiment was performed in duplicate. In this study, we have focused on the mechanisms underlying the initial response after activation of primary T cells by alphaCD3 and alphaCD28 and the concomitant increase in protein synthesis. Since the levels of initiation factor proteins did not change in this early phase of T cell activation, we considered the possibility that changes in the phosphorylation state and/or activities of several translation factors were involved in the initial activation of protein synthesis in T cells. Protein synthesis is regulated via multiple signalling pathways | Primary T lymphocytes were activated with alphaCD3 and alphaCD28, and after 1 h of activation, protein synthesis was increased 1.2 fold . To study the signalling events involved in this increase in protein synthesis, we performed the experiment in the presence of different specific signal transduction pathway inhibitors . The increase in overall protein synthesis was consistently blocked by each of the signal transduction inhibitors used, i.e. the PI 3-kinase inhibitor wortmannin, the mTOR inhibitor rapamycin, the p38 MAPKalpha/beta inhibitors SB203580 and SB202190, and the MEK inhibitor PD98059. It appears that the immediate activation of protein synthesis in T cells involves interplay between several signalling pathways. Figure 3 | The increase in protein synthesis requires signalling via several pathways. The increase in protein synthesis requires signalling via several pathways. T cells were preincubated with wortmannin (W, 100 nM), rapamycin (R, 100 nM), SB203580 (10 muM), SB202190 (10 muM) or PD98059 (PD, 50 muM) for 30 min before activation with both alphaCD3 and alphaCD28 for 1 h or the cells were treated with PMA (1 muM) for 1 h. For the last 30 min of activation 10 muCi/ml [35S]-labelled methionine was present. The experiment was performed in duplicate. Protein synthesis in control cells was set at 100%. Methionine incorporation ranged between 600 and 1500 cpm per 50 mug of protein. Incorporation of [35S]-labelled methionine into equal amounts of protein was measured using hot TCA precipitation (p<0.05 for the alphaCD3/28 and PMA treated samples; all other samples are not significantly different from the control, n=6). Stimulation of T cells with the more potent stimulus PMA for 1 h led to a substantially larger increase in protein synthesis (1.8 fold) compared to activation with alphaCD3 and alphaCD28 . Phosphorylation of eIF4E, eIF4F complex formation and 4E-BP1 phosphorylation remain unchanged after T cell activation | Phosphorylation of the cap-binding protein eIF4E can be regulated via the ERK and p38 MAPKalpha/betapathways , two pathways that appear to be important for the regulation of protein synthesis in T cells . Furthermore, the phosphorylation of eIF4E has been reported to be increased in response to several different treatments of primary T lymphocytes or the Jurkat T cell line . However, activation of T cells with alphaCD3 and alphaCD28 for up to 60 min did not cause a significant change in the phosphorylation state of eIF4E . Stimulation of T cells with either alphaCD3 or alphaCD28 alone was also insufficient to change the phosphorylation state of eIF4E (data not shown). We did detect a marked change in phosphorylation of eIF4E after 30 min of PMA treatment, indicating that the cells respond to this stimulus . In addition, treatment of the Jurkat T cell line with alphaCD3 and alphaCD28 caused phosphorylation of eIF4E that was already detectable after 30 min, demonstrating the effectiveness of the antibodies (alphaCD3 and alphaCD28) used . Figure 4 | Regulation of eIF4E phosphorylation and eIF4F formation. Regulation of eIF4E phosphorylation and eIF4F formation.A. Jurkat T cells were activated for 30 or 60 min by both alphaCD3 and alphaCD28. The primary T lymphocytes were activated for the indicated times with both alphaCD3 and alphaCD28 or with PMA. eIF4E was purified using m7GTP Sepharose, analyzed on a one-dimensional iso-electric focusing gel and detected by Western blotting. 4E and 4E-P indicate unphosphorylated and phosphorylated eIF4E respectively. B. 100 mug of total cell lysate from primary T cells treated for 1 h with alphaCD3 and alphaCD28 or PMA was analyzed by SDS-PAGE and Western blotting to detect 4E-BP1. (-) indicates untreated cells. The lane with 4E-BP1 from HeLa cell extract was obtained from a shorter exposure from the same blot. C. Formation of eIF4F was analyzed after 60 min activation of primary T cells with either alphaCD3, alphaCD28 or both. eIF4E was purified as described above and its association with eIF4G was analyzed by SDS-PAGE and Western blotting. An eIF4E blot was used to verify equal loading of all lanes. Similar results for eIF4E, eIF4G and 4E-BP1 were obtained in three independent experiments. An important way of regulating eIF4F assembly is through eIF4E-binding proteins such as 4E-BP1. Phosphorylation of 4E-BP1 leads to its release from eIF4E, allowing the latter protein to bind eIF4G . The phosphorylation of 4E-BP1 can be detected by virtue of a reduction in its mobility upon SDS-PAGE . As a control to demonstrate that different forms of human 4E-BP1 can be resolved on our gel system, we used HeLa cell extract and in this case three separate bands (alpha, beta and gamma) were indeed detected , indicative of differently phosphorylated forms. In resting T-cells, 4E-BP1 was mainly present in the unphosphorylated form (alpha-form) as reported before . We were unable to detect any changes in mobility of 4E-BP1, and therefore its phosphorylation after stimulation of the cells with alphaCD3 plus alphaCD28 or PMA . Formation of eIF4F complexes was studied by purification of eIF4E on m7GTP-Sepharose followed by a Western blot to detect associated eIF4G. In resting T cells, eIF4F complexes are already present, and after 1h of activation with alphaCD3, alphaCD28, or both antibodies, the amount of eIF4G bound to eIF4E remained unchanged . Similar results were obtained after 30 min of activation (data not shown). Surprisingly, no 4E-BP1 associated with eIF4E was detected, even though up to 2 mg of T cell extract was used in a m7GTP Sepharose pull down (data not shown). This could be due to low amounts of 4E-BP1 protein present in resting T cells. These data indicate that increased formation of eIF4F complexes is not required for the activation of protein synthesis in the early phase of T cell activation. Regulation of eIF2B activity after activation of T cells | In several cell types, an increase in overall protein synthesis coincides with an increase in eIF2B activity . We therefore examined the activity of eIF2B after activation of primary T cells . After 1 h of activation with alphaCD3 and alphaCD28, the activity of eIF2B increased 2.2 fold. The increase in eIF2B activity was not caused by a change in the phosphorylation state of the alpha-subunit of eIF2 or by an increase in the amount of eIF2Bepsilon protein (Figs. and and ). Immunoprecipitation of different amounts of T cell extracts showed that the eIF2Bepsilon antibody was able to detect different levels of protein in the immunoprecipitations within the same range used in Fig. (bottom panel). Taken these results together it suggested that eIF2B was regulated directly, e.g. via phosphorylation. To study which signal transduction pathways are involved in the regulation of the activity of eIF2B, the cells were activated in the presence of specific signal transduction pathway inhibitors and eIF2B activity was measured. The basal activity of eIF2B was slightly affected in the presence of the PI 3-kinase inhibitor wortmannin, the mTOR inhibitor rapamycin, and the p38 MAPKalpha/beta inhibitor SB203580, however the alphaCD3 plus alphaCD28-induced increase in eIF2B activity was completely blocked in the presence of each inhibitor. The MEK inhibitor PD98059 did not affect the basal eIF2B activity and was also able to inhibit the alphaCD3 and alphaCD28-induced increase in eIF2B activity, showing that all these signalling pathways are required to mediate the activation of eIF2B . Figure 5 | Regulation of eIF2B activity. Regulation of eIF2B activity.A. T cells were preincubated with wortmannin (W, 100 nM), rapamycin (R, 100 nM), SB203580 (SB, 10 muM), or PD98059 (PD, 50muM) for 30 min and left untreated (white bar) or the cells were activated with both alphaCD3 and alphaCD28 (black bar) for 1 h. Simultaneously cells were activated with PMA (hatched bar) for 1 h. An eIF2B assay was performed as described in Materials and Methods. Bars marked with * are significantly different from the untreated cells (p<0.05, n=5). B. Cell lysates from resting and stimulated cells (1 h alphaCD3 and alphaCD28) were analyzed by SDS-PAGE and Western blotting to detect phosphorylated eIF2alpha. eIF2B was immunoprecipitated from 400 mug of lysate using alphaeIF2Bepsilon and the amount of protein was analyzed by SDS-PAGE and Western blotting. Similar results were obtained in three experiments C. To test the sensitivity of the eIF2Bepsilon antibody different amounts of T cell extracts (as indicated) were immunoprecipitated with alphaeIF2Bepsilon and analyzed by SDS-PAGE and Western blotting. Similar results were obtained in two experiments. D. T cells were left untreated (white bar) or the cells were activated with both alphaCD3 and alphaCD28 (black bar) or PMA (grey bar) for 1 h. GSK-3alpha and beta were immunoprecipitated together from 400 mug of lysate and a kinase assay was performed as described in Materials and Methods (p<0.05, n=4). E. eIF2Bepsilon was immunoprecipitated from 400 mug of lysate from resting and stimulated cells (1 h alphaCD3 and alphaCD28) and the total amount of eIF2Bepsilon and the phosphorylation state of Ser540 were analyzed by SDS-PAGE and Western blotting. Similar results were obtained in two experiments. Activation of eIF2B after stimulation of the cells with PMA was about 2 fold higher than after stimulation with the antibodies. An increase in eIF2B activity after stimulation of primary T-cells with PMA/ionomycin has been reported before . It has been suggested that GSK-3 may be an important regulator of eIF2B activity, i.e. in response to insulin and during cell survival . Phosphorylation of eIF2Bepsilon by GSK-3 inhibits the activity of the eIF2B complex . GSK-3 activity is decreased only by a small extent (15%) after T cell activation with alphaCD3 and alphaCD28 . In contrast, PMA treatment reduced GSK-3 activity by about 50%, which is similar to previously reported data. GSK-3 phosphorylates eIF2Bepsilon on Ser540. Therefore, we analyzed the phosphorylation state of this site using a phospho-specific antibody. We were unable to detect any change in the phosphorylation of this site in response to alphaCD3 and alphaCD28, excluding a role for GSK-3 in the regulation of eIF2B activity in T cells under these conditions. Dephosphorylation of eEF2 | Elongation factor 2 (eEF2) plays an important role in the regulation of the rate of elongation, and therefore in the regulation of the rate of overall protein synthesis. Phosphorylation of eEF2 causes its inactivation . Phosphorylation of eEF2 was rapidly but only transiently decreased after activation of the primary T lymphocytes with alphaCD3 and alphaCD28 . Within 3 min, maximum dephosphorylation was reached and the phosphorylation level returned to a level similar to that of resting T cells by 10 min. Given the transient nature of these changes, it is unlikely that regulation of eEF2 plays a role in the sustained increase in the rate of protein synthesis after activation of T cells. However, dephosphorylation of eEF2 could play a role very early in T cell activation. Figure 6 | Dephosphorylation of eEF2 after T cell activation. Dephosphorylation of eEF2 after T cell activation. T cells were activated with alphaCD3 and alphaCD28 for the indicated times. 80 mug of protein was analyzed by SDS-PAGE and Western blotting. Phosphorylation of eEF2 was detected using a phospho-specific antibody (eEF2-P). ERK2 was detected as a loading control. Similar results were obtained in three experiments. Regulation of p70 S6 kinase upon T cell activation | Activation of p70 S6 kinase and phosphorylation of the ribosomal protein S6, an in vivo substrate of p70 S6 kinase, coincide with increased translation of specific mRNAs, namely the 5'TOP mRNAs . However, a recent report has questioned the role of p70 S6 kinase in 5'TOP messenger translation . We studied the effect of activation of T cells with alphaCD3 plus alphaCD28 on these proteins. The activity of p70 S6 kinase was increased about 1.5 fold after 60 min, and the increase in p70 S6 kinase activity was blocked by each of the signal transduction inhibitors used, i.e. wortmannin, PD98059, SB203580, and rapamycin . Similar results were obtained when the phosphorylation of the S6 protein was examined as a cellular read-out of p70 S6 kinase activity . Phosphorylation of S6 in response to PMA was greater than in response to alphaCD3 and alphaCD28, similarly to the situation for several translation factors, as described above. Figure 7 | Regulation of p70 S6 kinase. Regulation of p70 S6 kinase.A. Primary T lymphocytes were preincubated with wortmannin (W, 100 nM), rapamycin (R, 100 nM), SB203580 (SB, 10 muM), or PD98059 (PD, 50 muM) for 30 min before activation with both alphaCD3 and alphaCD28 (black bars) for 60 min. Bars marked with * are significantly different (p<0.01) from the activity in untreated cells (n=4). B. T cells were preincubated with wortmannin (W, 100 nM), rapamycin (R, 100 nM), SB203580 (SB, 10 muM), or PD98059 (PD, 50 muM) for 30 min before activation with both alphaCD3 and alphaCD28 for 60 min or the cells were stimulated with PMA for 60 min. Phosphorylation of the S6 protein was analyzed by SDS-PAGE and Western blotting using a phospho-specific S6 (Ser235) antibody. Similar results were obtained in four experiments. Discussion : The mechanisms underlying the regulation of protein synthesis following activation of resting primary T cells has not been widely studied. Early reports showed that stimulation of primary T cells with pharmacological stimuli, e.g. PHA or PMA, led to an increase in protein synthesis and protein levels of certain translation initiation factors within 16 h . More recently the regulation of protein synthesis and translation factors after a 6 h stimulation of primary T cells with PMA or PMA/ionomycin was described in detail . We investigated the regulation of protein synthesis and translation factors during the early phase of activation of resting T cells by alphaCD3 and alphaCD28, i.e. 1 h of activation. The protein synthesis rate in T cells rapidly increased after treatment , and several signalling pathways, i.e. ERK and p38 MAPK phosphorylation and PKB activation, were stimulated, showing the efficacy of the alphaCD3 and alphaCD28 antibodies in activating the cells. The increase in protein synthesis after 1 h was mediated via multiple signalling pathways, e.g. the MEK, p38 MAPKalpha/beta, PI 3-kinase and mTOR pathway, as indicated by the use of signal transduction inhibitors . In several cell types, the involvement of either MEK , PI 3-kinase or mTOR [,,-] in the activation of protein synthesis has been described. However, this is the first time that a role for the p38 MAPKalpha/betapathway in the regulation of overall protein synthesis has been described. This role is supported by the fact that two structurally unrelated p38 MAPKalpha/beta inhibitors, i.e., SB202190 and SB203580, were each able to block the increase in protein synthesis. The increase in protein synthesis after activation of T cells with alphaCD3 and alphaCD28 coincided with an increase in the activities of p70 S6 kinase and eIF2B and dephosphorylation of eEF2, which is indicative of an increase in its activity . The dephosphorylation of eEF2 was very transient, and therefore it is unlikely that eEF2 plays an important role in the sustained increase in protein synthesis after T cell activation. However, activation of eEF2 could be important for the initial increase in protein synthesis. The stimulation of p70 S6 kinase was mediated via similar signalling pathways (Fig. 7) to those underlying the activation of overall protein synthesis. Inhibition of PI 3-kinase, mTOR, p38 MAPKalpha/beta, or MEK during T cell activation by alphaCD3 and alphaCD28 prevented the activation of p70 S6 kinase, indicating that multiple signalling pathways are required for regulation of p70 S6 kinase activity. The effect of rapamycin on the activation of p70 S6 kinase in response to alphaCD3 and alphaCD28 has been reported previously . Inhibition of the activation of p70 S6 kinase by SB203580 has been described before in insulin-stimulated rat vascular smooth muscle cells [,-]. Furthermore, it has been reported that SB203580 (at the concentration used, 10 muM) can inhibit phosphorylation of PKB at Threonine308 and thus its activation . Since PKB is an upstream component of the signalling pathway towards p70 S6 kinase, this could provide a mechanism by which SB203580 blocks activation of p70 S6 kinase. Each of the other signalling pathways studied here has also been implicated in the regulation of p70 S6 kinase activity in a variety of cell types under a range of conditions [,-]. However, in human primary T cells all of them appear to be important for the regulation of p70 S6 kinase activity after stimulation with alphaCD3 and alphaCD28. The activity of the guanine nucleotide exchange factor, eIF2B, also was mediated via similar signalling pathways as the increase in protein synthesis . The increase in eIF2B activity after activation of T cells with alphaCD3 and alphaCD28, at the early times we examined, was not due to an increase in eIF2B protein level or to changes in eIF2alpha phosphorylation. Therefore, modification of the eIF2B protein complex probably caused the increase in eIF2B activity. The modulation of the activity of the eIF2B complex after activation of T cells with alphaCD3 and alphaCD28 required several different signalling pathways, e.g. MEK, p38 MAPKalpha/beta, mTOR and the PI 3-kinase pathway. These signalling pathways have been reported separately to be involved in the regulation of eIF2B activity . However, this is the first report where all these pathways are involved in the regulation of eIF2B in a single cell type. A small inactivation of GSK-3 was detected after activation of primary T cells with alphaCD3 and alphaCD28. However, no dephosphorylation of Ser540 in eIF2Bepsilon was detected , excluding a role for GSK-3 in regulating the activity of eIF2B under these conditions. In contrast, studies employing PMA/ionomycin-activated T cells , insulin treatment of various cell types , and cell survival have implied a role for GSK-3 in regulating eIF2B activity. In contrast to previously reported data using mitogenic stimuli to activate primary T cells or Jurkat T cells , eIF4E phosphorylation, association of eIF4G with eIF4E and 4E-BP1 phosphorylation remained unchanged after T cell activation using alphaCD3 and alphaCD28 as a stimulus . Signal transduction inhibitor studies showed that the MEK and p38 MAPKalpha/beta pathways are important for eIF4E phosphorylation in Jurkat T cells , and a role for MEK was demonstrated previously in primary T cells . The weaker activation of the ERK pathway in particular by alphaCD3 and alphaCD28 in primary T cells may well account for the absence of increased phosphorylation of eIF4E under these conditions. We did not observe increased phosphorylation of 4E-BP1 in response to alphaCD3 and alphaCD28, even though it has been reported to occur after cytokine stimulation of a murine cytotoxic T cell line or after 6 h of mitogenic stimulation of human primary T cells . However, Grolleau et al. showed that 4E-BP1 was present in primary T cells mainly in its dephosphorylated form, and no significant change was detected after PMA treatment. Our results are consistent with this last finding; 4E-BP1 is mainly present as one band, and no change in mobility is observed upon cell treatment. This is consistent with the observation that eIF4F complex formation did not alter. However, it remains surprising that eIF4F complexes are present when 4E-BP1 is completely dephosphorylated, and therefore presumably associated with eIF4E. We were unable to detect any 4E-BP1 associated with eIF4E, which is probably due to the low 4E-BP1 protein levels in resting T lymphocytes, thus explaining basal eIF4F formation. Conclusions : The treatment of primary T lymphocytes with alphaCD3 and alphaCD28 activates two key components of the translational machinery, p70 S6 kinase and eIF2B. The activities of these translation factors were regulated similarly to the activation of protein synthesis, consistent with an important role for the components in the activation of protein synthesis by alphaCD3 and alphaCD28. Interestingly, activation of protein synthesis, p70 S6 kinase, and eIF2B is inhibited by rapamycin, a compound that was first discovered as an immunosuppressant, suggesting that mTOR regulated translation is involved in the process of T cell proliferation. The activation of p70 S6 kinase is related to the regulation of translation of specific mRNAs, while the activation of eIF2B is likely to be required for stimulation of general protein synthesis . This suggests that increases in both specific and general protein synthesis are important in the early phase of T cell activation. Materials and methods : Primary T cell isolation and cell treatment | Buffy coats used for the isolation of T cells were prepared from freshly drawn blood from healthy human donors and were obtained from the Scottish National Blood Transfusion Service (Edinburgh, UK). Mononuclear leukocytes were isolated by Ficoll-Hypaque (Amersham-Pharmacia) gradient centrifugation. T cells were further enriched using nylon-wool columns. T cells were suspended in RPMI 1640 medium supplemented with 10% (v/v) heat-inactivated foetal calf serum, 1 mM glutamine and antibiotics/antimycotics (100 units/ml penicillin G sodium, 100 mug/ml streptomycin sulphate and 0.25 mug/ml amphotericin B). The cells were kept in 75 cm2 tissue culture flasks at a density of 4 x 106 cells/ml at 37C and 5% CO2. All tissue culture reagents were obtained from Gibco BRL. Measurement of protein synthesis rate | Cells were treated with alphaCD3 mouse IgG2a mAb (33/2A3) (1:1000 dilution from a hybridoma supernatant) and alphaCD28 mouse IgM mAb (CK243) (1:12 dilution from a hybridoma supernatant) for 1 h in the absence or presence of signal transduction inhibitors. For the last 30 min, 10 muCi/ml [35S]methionine was added to the cells. To harvest the cells, the cells were transferred to a microfuge tube and centrifuged at 6000 x g for 20 s. The cell pellet was lysed in 20 mM Hepes pH7.4, 50 mM beta-glycerophosphate, 0.2 mM EDTA, 1% Triton X-100, 10% (v/v) glycerol, 1 mug/ml leupeptin, 1 mug/ml pepstatin, 1 mug/ml antipain, 1 mM benzamidine, and 1 mM DTT. Part of the sample was used to measure the protein content with Protein Assay Reagent (Bio-Rad) and the rest was spotted in duplicate on to Whatman 3 MM paper and subjected to 'hot TCA precipitation'. Gel electrophoresis and Western blotting | T cells were activated with alphaCD3 and alphaCD28 for times indicated in the figure legends, harvested in Laemmli sample buffer and analyzed by SDS-PAGE and Western blotting. Phospho-p42/44(ERK) and phospho-p38 MAPK antibodies were obtained from New England Biolabs, the phospho-eIF2alphaantibody was a kind gift from Dr. Gary Krause (Detroit, USA), the phospho-S6 (Ser235) antibody was a kind gift from Dr. Dario Alessi (University of Dundee), the 4E-BP1 antibody was obtained from Santa Cruz (SC-6025), the eIF2Bepsilon antibody was raised in rabbit against the whole protein expressed in the baculovirus system , the phospho-specific antibody for Ser540 in eIF2Bepsilon was raised against the peptide SEEPDS(P)RGGC (S(P) indicates the phosphoserine) in sheep, and the phospho-eEF2 (Thr56) antibody was raised against the peptide GETRFT(P)DTRK (T(P) indicates phosphothreonine) . Kinase assays | For PKB assays, the cells were pelleted at 6000 xg for 20 s and harvested in 50 mM Tris-HCl pH 7.5, 1 mM sodium orthovanadate, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 50 mM NaF, 5 mM sodium pyrophosphate, 0.27 M sucrose, 1 muM microcystin LR, 1 mug/ml leupeptin, 1 mug/ml pepstatin, 1 mug/ml antipain, and 1 mM benzamidine-HCl. Antibodies directed against the three PKB isoforms (alpha, beta, and gamma) were simultaneously bound to protein G-Sepharose, and about 100 mug of protein was used in immunoprecipitation reactions. The immunoprecipitation and PKB assays were performed as described before . For p70 S6 kinase and GSK-3 assays, the cells were harvested in a buffer containing 50 mM Tris-HCl pH 7.5, 50 mM beta-glycerophosphate, 0.5 mM sodium vanadate, 1.5 mM EDTA, 1.5 mM EGTA, 0.5% Triton X-100, 1 mug/ml leupeptin, 1 mug/ml pepstatin, 1 mug/ml antipain, 1 mM benzamidine, and 1 mM DTT. A polyclonal antibody raised against a peptide sequence from p70 S6 kinase-1 was bound to protein G-Sepharose, and about 100 mug of extract was used in the immunoprecipitation reaction. Immunoprecipitation and the p70 S6 kinase assays (using a peptide substrate) were performed as described before . GSK-3alpha and beta were immunoprecipitated together from 150 mug of cell lysate and kinase assays were performed as described . Phosphorylation of eIF4E and eIF4F complex formation | Cells were pelleted at 6000 x g for 20 s and harvested in a buffer containing 20 mM Hepes pH7.4, 50 mM beta-glycerophosphate, 0.2 mM EDTA, 1% Triton X-100, 10% (v/v) glycerol, 1 mug/ml leupeptin, 1 mug/ml pepstatin, 1 mug/ml antipain, 1 mM benzamidine, and 1 mM DTT. Using m7 GTP Sepharose 4B (Amersham-Pharmacia; 15 mul of slurry diluted with 15 mul of Sepharose CL-4B), eIF4E was purified from approximately 2 mg of extract. For SDS-PAGE, Laemmli sample buffer was added and the samples were heated at 95C for 10 min. For one-dimensional iso-electric focusing analysis, the appropriate sample buffer was added . The samples were run on a 12.5% SDS-PA gel or on a one-dimensional iso-electric focusing gel, transferred to PVDF, and detected by Western analysis. eIF4E was detected with a polyclonal antibody raised in rabbit , and eIF4GI with a polyclonal antibody raised in sheep against the peptide CKKEAVGDLLDAFKEAN. Measurement of eIF2B activity | The cells were pelleted at 6000 x g for 20 s and lysed in a buffer containing 20 mM Tris-HCl pH 7.5, 50 mM beta-glycerophosphate, 100 mM KCl, 0.2 mM sodium orthovanadate, 0.2 mM EDTA, 0.2 mM EGTA, 1% Triton X-100, 10% glycerol, 1 mug/ml leupeptin, 1 mug/ml pepstatin, 1 mug/ml antipain, 1 mM benzamidine, and 1 mM DTT. About 50 mug of cell lysate was used for the eIF2B assay, which was performed as described previously . Authors' contributions : Author 1 (MK) carried out all the experiments. Author 2 (CGP) participated in the design and coordination of this study. All authors read and approved the final manuscript. Abbreviations : eIF, eukaryotic initiation factor; eEF, eukaryotic elongation factor; ERK, extra-cellular regulated kinase; GSK-3, glycogen synthase kinase-3; MAPK, mitogen-activated protein kinase; MEK, mitogen-activated protein kinase kinase; m7GTP, 7-methyl guanosine triphosphate; mTOR, mammalian target of rapamycin; PI 3-kinase, phosphoinositide 3-kinase; PHA, phytohemagglutinin; PKB, protein kinase B; PKC, protein kinase C; PMA, phorbol 12-myristate 13-acetate; 4E-BP1, eIF4E-binding protein 1. Backmatter: PMID- 12006105 TI - Management of obstetric anal sphincter injury: a systematic review & national practice survey AB - Abstract | Background | We aim to establish the evidence base for the recognition and management of obstetric anal sphincter injury (OASI) and to compare this with current practice amongst UK obstetricians and coloproctologists. Methods | A systematic review of the literature and a postal questionnaire survey of consultant obstetricians, trainee obstetricians and consultant coloproctologists was carried out. Results | We found a wide variation in experience of repairing acute anal sphincter injury. The group with largest experience were consultant obstetricians (46.5% undertaking >= 5 repairs/year), whilst only 10% of responding colorectal surgeons had similar levels of experience (p < 0.001). There was extensive misunderstanding in terms of the definition of obstetric anal sphincter injuries. Overall, trainees had a greater knowledge of the correct classification (p < 0.01). Observational studies suggest that a new 'overlap' repair using PDS sutures with antibiotic cover gives better functional results. However, our literature search found only one randomised controlled trial (RCT) on the technique of repair of OASI, which showed no difference in incidence of anal incontinence at three months. Despite this, there was a wide variation in practice, with 337(50%) consultants, 82 (55%) trainees and 80 (89%) coloproctologists already using the 'overlap' method for repair of a torn EAS (p < 0.001). Although over 50% of colorectal surgeons would undertake long-term follow-up of their patients, this was the practice of less than 10% of obstetricians (p < 0.001). Whilst over 70% of coloproctologists would recommend an elective caesarean section in a subsequent pregnancy, only 22% of obstetric consultants and 14% of trainees (p < 0.001). Conclusion | An agreed classification of OASI, development of national guidelines, formalised training, multidisciplinary management and further definitive research is strongly recommended. Keywords: Background : The importance of highlighting the problem of incontinence to professionals, and the need to focus on reducing underlying causes is emphasised in recent Department of Health documents . Anal incontinence may be defined as 'faecal or flatus incontinence which is a social or hygienic problem' . There is little doubt that vaginal delivery in general, and obstetric anal sphincter injury (OASI) in particular, are significant contributory factors in the development of anal incontinence . In the UK, anal incontinence in the year after birth is thought to affect nearly 40,000 mothers (1 in 20) annually . Post-partum anal incontinence may affect mothers psychologically as well as physically but many do not seek medical attention because of embarrassment or because they are easily discouraged from discussing it . In one study only one third of individuals with faecal incontinence had ever discussed the problem with a physician . In recent correspondence to the Continence Foundation, a woman describes 'the eternal shame of being with another person when the worst occurs' . The impact of this complication on the vulnerable postnatal mother and her baby is potentially catastrophic. Aside from the potential clinical and social implications there are important medico-legal issues . Furthermore, the treatment of postpartum anal incontinence itself is associated with very high cumulative costs . Perineal trauma after childbirth is of further importance because it lies behind the growing clamour for the right of a woman to choose whether to deliver by elective caesarean section . Obstetric anal sphincter injuries may be seen at the time of birth ('overt') or may be detected only after additional ultrasound investigation, after birth ('occult'). The incidence of 'overt' anal sphincter injury has previously been reported as being between 0.5 --3% of vaginal deliveries . Until recently, anal incontinence not due to 'overt' anal sphincter injury was attributed to pelvic neuropathy . The advent of anal endosonography altered this view by identifying further 'occult' obstetric trauma to the anal sphincter. This has been reported in 35% of primiparous women and a significant association has been demonstrated between these sonographic defects and anal incontinence. However, it has not been established whether these injuries were genuinely 'occult' or whether they had been missed by the doctor or midwife at delivery. There is evidence from one study that perineal anatomy is poorly understood by midwives and trainee doctors, who perform the bulk of deliveries in the UK. In this study, 41% of trainees and 16% of midwives incorrectly classified a partial or complete tear of the EAS as 'second degree'. Inconsistency in classification of tears would allow many injuries to pass, unrecognised. Clearly where an injury occurs, but is not detected, the incidence of anal incontinence may approach 100%. These women will frequently be referred to colorectal surgeons for further evaluation and possible 'secondary' repair. Even with recognition and 'primary' repair, the incidence of anal incontinence has been reported as over 50% and the actual incidence may be even higher . The reasons for the apparently poor outcome after primary repair are not clear, particularly as there is considerable controversy in the literature regarding the optimal obstetric management of OASI. Variation in outcome may be due to different methods and materials being used or to deficiencies in skill and training . Given the sub-optimal outcomes achieved when these injuries are repaired by obstetricians, it has been suggested that better results may be obtained if colorectal surgeons perform the primary surgery . The aims of this study were firstly, to establish the best available evidence for the management of OASI by conducting a systematic review of the literature. Secondly, we aimed to audit current practice amongst trainee and consultant obstetricians. Finally, we wished to explore the views of consultant colorectal surgeons with respect to optimal management of OASI. Methods : Literature review | The Cochrane Library and Cochrane Register of Controlled Trials were searched for relevant Randomised Controlled Trials (RCT), systematic reviews and meta-analysis. A search of MEDLINE and PUBMED (electronic databases) from 1966 up to April 2001 was also carried out. The databases were searched using the relevant MeSH terms, including all sub-headings and this strategy was combined with a Key-word search using -- Human; Female; Childbirth; Obstetric; Perineum; Anal sphincter; Tear; Injury; Rupture; Damage; Incontinence; Faecal; Anal; Repair; Surgery; Sutures; Randomised controlled trials; Meta-analysis. The bibliographies of retrieved articles were searched manually as well as conference proceedings and abstracts from obstetrics and gynaecology and coloproctology meetings. Survey participants | The study sample included all consultant obstetricians & gynaecologists in the Royal College of Obstetricians & Gynaecologists (RCOG) UK database. As a comprehensive national register of obstetric trainees is unavailable, we surveyed all trainee obstetricians in the two regions in which the authors are based (West Midlands and South West Thames regions). We also included all consultant members of the Association of Coloproctology of Great Britain and Ireland. Development of a questionnaire | The postal questionnaire that was sent to the consultant and trainee obstetricians enquired about those aspects of knowledge and practice that our preliminary literature review suggested would be linked to outcome. Questions relating to current experience, classification of perineal injury, acute management, follow-up, management of subsequent deliveries and training received were all included. A similar questionnaire inquiring about the management of primary OASI repair was sent to the consultant coloproctologists. Data processing and analysis | A Freepost envelope was included with the questionnaire. Non-responders received a second questionnaire. Data was entered onto an Access database. The software used for analysis was StatXact Turbo (CYTEL), Cambridge, Massachusetts. The Kruskal-Wallis test was used to compare ordinal data between the 3 groups and chi square tests were applied when the data were nominal. In order to highlight the differences between current evidence and current practice we have presented the results of our literature search in combination with the results of the survey. Results : Respondents | At the time of the survey, 1441 names appeared in the RCOG consultant database (UK) of whom 152 were non-practising obstetricians and 96 had retired. Of the 1193 consultants in active obstetric practice, 672 (56%) completed and returned the questionnaire. Of the 235 trainee obstetricians in the West Midlands and South West Thames regions, 148 (63%) completed and returned the questionnaire. Only 90 (23%) of the 385 members of the Association of Coloproctology replied, despite 2 mailings. The overall response rate was 50.2% (910/1813). The majority of responding consultant obstetricians, 438 (65%) and coloproctologists, 53 (59%) had been in post for more than 5 years. Of the trainee obstetricians-108 (73%) were post-MRCOG. The average numbers of acute OASI repairs (primary) performed annually by consultant obstetricians, obstetric trainees and colorectal surgeons are shown in Table . There were significant differences between the three groups. Table 1 | Number of acute OASI repairs performed per year Definition of OASI | The evidence | The literature review revealed a lack of consistency in the classification of OASI. One study, which examined all the obstetric texts (n= 65) in the RCOG library, found that 22% of authors classified anal sphincter injury as 'second degree' and a further 17% did not mention any classification . Obstetric anal sphincter injury (OASI) is classified as a '3rd degree' tear when there is any involvement of the external anal sphincter (EAS) but when the anal epithelium is involved it is '4th degree' and this was incorporated in the RCOG guidelines on the management of perineal trauma . The RCOG definitions of perineal injury, which are included in the Green Top Guidelines for the Management of Third and Fourth-Degree Perineal Tears Following Vaginal Delivery are shown in Table . A more descriptive classification suggested by Sultan was agreed at a recent consensus meeting on OASI where third degree tears were further classified into three subgroups according to the extent of damage to the external anal sphincter (EAS) and internal anal sphincter (IAS) . Table 2 | Classification of Injury The survey | Two hundred and twenty (33%) consultant obstetricians and 30 (22%) trainees considered a complete or partial external sphincter tear to be 'second degree' . There was widespread regional variation in the 'misclassification' of OASI as 'second degree'. This is emphasised by a ten-fold difference between some regions (higher in the northern regions) in rates of respondents considering a complete EAS tear to be 'second degree' . Table 3 | Classification of perineal tears Table 4 | Geographical variation in the definition of OASI Technique of repair following OASI | The evidence | The most common type of repair is an end-to-end repair, where either interrupted or figure-of-eight sutures are inserted into the sphincter muscle. There is little variation on the standard technique reported. One study described the end-to-end approximation of the anal sphincter, by suturing the outer fascial layer without inserting sutures in the muscle. However, this study has not been reproduced. When employed for secondary repair, mobilisation, scar excision and direct apposition lead to an incidence of sphincter repair disruption of 40% . A modification of an overlapping technique for sphincter repair, described by Parks for the secondary repair of OASI , was first described for acute OASI by Sultan in 1999 . The technique includes identifying the internal anal sphincter which, if torn, is repaired as a separate layer. Using this technique the authors found a significant reduction in anal incontinence (to 8%), which can be compared with 41%, seen in a previous study where the end-to-end technique was employed . There is only one published, prospective randomised study, comparing end-to-end and the overlap techniques . In this series of 112 primiparous women, no significant difference in continence was identified at 3 months' follow-up. The techniques used in this study were different to those described by Sultan . In particular, internal sphincter injury was not identified and repaired separately. There are two ongoing randomised controlled trials comparing the two methods (in Stoke and Liverpool) registered in the Cochrane clinical trials register . The survey | Coloproctologists favour the overlap technique for primary repair of OASI . This technique is reportedly being used by large proportions of obstetricians particularly trainees. Over 55% of responding consultant obstetricians (375/672) and coloproctologists (60/90) said that they would be interested in participating in a trial to compare the two methods of repair. Table 5 | Techniques of external anal sphincter repair Suture material | The evidence | No randomised controlled studies to assess the best suture material for sphincter repair were identified. Most texts that describe the repair still mention the use of chromic catgut . However, monofilament suture materials, such as Polydioxanone (PDS) or Polypropelene (Prolene) are thought to be better than catgut or Polyglactin (Vicryl) because of their longer half life. There is good evidence from randomised trials that synthetic materials such as Vicryl or Polyglycolic acid (Dexon) are preferable to catgut for repair of the perineum . Catgut sutures, made from bovine intestinal material, have recently been withdrawn from UK and European countries. The survey | Vicryl was the material most frequently used for sphincter repair (by 505 (75%) consultant obstetricians and 97 (66%) trainees). By contrast, only 22 (24%) coloproctologists used Vicryl, with greater numbers preferring PDS (41 (46%)). Obstetric experience with PDS was minimal (used by only 51 (8%) consultant obstetricians and 20 (14%) trainees). Antibiotic usage | The evidence | Intra-operative and post-operative broad spectrum antibiotics are recommended because the development of infection may be linked to a breakdown of the anal sphincter repair . Although data exists relating to antibiotic prophylaxis in colorectal surgery there are no randomised controlled studies which examine antibiotic use following OASI repair. The survey | Peri-operative antibiotics were recommended by 82 (91%) coloproctologists and by 528 (79%) consultant obstetricians and 130 (88%) obstetric trainees. Stool softeners | The evidence | A passage of a hard bolus of stool may disrupt repair and therefore most surgical textbooks and experts recommend use of laxatives. However the use of laxatives and stool softeners after OASI repair has not been evaluated in a randomised controlled trial. Indeed medical 'bowel confinement' practised by some colorectal surgeons following secondary anal sphincter repair, has been shown in a randomised trial to confer no benefit in terms of septic complications or functional outcomes . The survey | Stools softeners or laxatives were advised by 70 (78%) coloproctologists. Six hundred and eighteen (94%) consultants and 121 (82%) trainee obstetricians routinely prescribed laxatives following repair. Colostomy | The evidence | In a recent report comprising of 4 patients, temporary defunctioning colostomy has been described when failure to recognise and repair rectal mucosal injury lead to significant early perineal contamination . We could find no studies to support colostomy in the management of acute OASI. A small randomised trial showed no conclusive evidence that a defunctioning stoma confers any benefit for those patients undergoing secondary repair in terms of functional outcome and may be associated with higher morbidity and longer hospital stays related to the stoma closure . The survey | Twenty-seven (30%) coloproctologists recommend a covering colostomy for third and fourth degree tears (of whom 25 said 'only for 4th degree tears'). None of the obstetricians either performed or requested a colostomy for acute OASI. Follow-up | The evidence | There are no controlled trials comparing different protocols. We found two publications suggesting that all women with OASI should be followed-up by coloproctologists . At a recent consensus meeting on OASI follow up in a multidisciplinary clinic was recommended for women with persistent bowel symptoms after delivery. The survey | Coloproctologists suggest a follow-up period of more than 12 months, compared to the majority of obstetricians who follow-up patients for only six weeks . Table 6 | Duration of follow-up after OASI repair Management of subsequent delivery | The evidence | There were no data in the literature from controlled studies regarding the best mode of subsequent delivery following OASI. Evidence that women who experience even transient anal incontinence after vaginal delivery may be at increased risk of faecal incontinence after a further vaginal delivery has lead to calls that such women should be offered caesarean section. However, compared with vaginal delivery caesarean section carries a higher mortality and other forms of morbidity and should therefore not be offered routinely to all women . In the prospective study from Fynes and colleagues , women with transient anal incontinence or occult sphincter injury after first delivery were at high risk of faecal incontinence after a second vaginal delivery. Furthermore it is not clear whether or not pregnancy per se influences postpartum anal incontinence. Symptomatic anal incontinence has been reported after both elective and emergency caesarean deliveries. An alternative approach may be to routinely assess function by means of a continence questionnaire, anorectal physiology tests and endosonography and selectively offer caesarean section to those women with some degree of compromised function. Clearly those who have had successful continence surgery should be delivered by caesarean section . However it remains to be established whether caesarean section would be beneficial to pregnant women with severe incontinence who are due to have continence surgery some time after delivery. The survey | In the present survey most coloproctologists recommended caesarean section. By contrast most obstetricians allowed vaginal delivery after a previous OASI. Table 7 | Mode of subsequent delivery: Role of the coloproctologist in the management of OASI | The evidence | Although some authors suggest that coloproctologists are best trained to repair OASI, there are no data to support this statement and the debate continues . The survey | In the present survey only small numbers of consultants (103 (15%)), and fewer trainees (5 (3%)), called a coloproctologist or a general surgeon to assist during acute OASI repair, the indication in these cases being 'severe anatomical disruption'. Few of the respondents believed that a coloproctologist should be routinely involved in acute management of OASI; 17 (19%) coloproctologists, 169 (25%) consultants and 16 (11%) trainees). Training received in the management of OASI | The evidence | There are no research studies comparing different methods of training. There is only one study that has evaluated training in repair of OASI . Ninety four percent reported unsatisfactory training at the time of performing their first unsupervised OASI repair. The survey | 445 (64%) consultants and 184 (64%) trainees reported either 'a lack of' or 'unsatisfactory' training in the management of OASI. Discussion : Our literature review and national survey have shown that obstetric anal sphincter injury is an area of childbirth that has been largely ignored, in both research and educational terms. There is a poor evidence base for practice and there has been only one published randomised controlled trial in this area. Our survey had a good obstetric response rate but a poor response from the colorectal surgeons (this may be attributed to the fact that, only about one third of coloproctologists have a specialised interest in anorectal incontinence). Our survey shows wide variation within specialities, levels of seniority and different regions of the country. These findings were echoed at a national professional consensus which highlighted the importance of research into anal incontinence, including the problems of failed recognition and uncertainty about the best method of repair . There are two probable reasons for the underestimation of OASI. Firstly, as a result of lack of a consistent classification, OASI can be wrongly classified as a 2nd degree tear and therefore managed inappropriately. We found that 33% of consultants and 22% of trainees classified a partial or complete tear of the EAS as 'second degree' which echoes the previous smaller study by Sultan . One reason for misclassification may be a degree of confusion created by popular obstetric texts . Our analysis of the regional trends in classification indicates that more consultants in the Northern regions of the country prefer to classify OASI as 2nd degree. This may reflect the teaching in a popular obstetric textbook . To avoid confusion we support the unified and descriptive classification given by Sultan in which third degree tears are further sub-classified into 3a (less than 50% of thickness of external sphincter torn), 3b (more than 50% of thickness of external sphincter torn) and 3c (internal sphincter torn). Figure 1 | Regional variation of the definition of anal sphincter injury Regional variation of the definition of anal sphincter injury Secondly, underestimation may be due to lack of recognition of OASI because of lack of training. A recent study has shown that immediate assessment of all perineal tears following childbirth by a trained and experienced obstetrician can significantly increase the detection rate. In this study, rates of up to 15% were found, which might provide an explanation for the high incidence of 'occult' sphincter defects reported in previous prospective studies . We have recently presented national data showing nearly a hundred fold variation in detection rates of OASI . The reason for the poor outcome following repair can be attributed to inappropriate technique or to lack of operator expertise or to a combination of these factors. Despite a primary repair of acute obstetric anal sphincter injury, up to 59% of women suffer from persistent anal incontinence and persistent sonographic sphincter defects have been identified in over 80% . Anal incontinence may present for many years following OASI and can also deteriorate with time. Schofield and Grace have looked at the differential rates of the components of anal incontinence (faecal incontinence and flatus and urgency) in five studies which examined these outcomes after primary repair of third degree tears. The intervals of follow-up ranged from 6 weeks to 10 years. The overall mean rates were 25% (flatus +/- urgency) and 9% (faecal incontinence). The longer the follow-up, the higher the rate of incontinence. We have identified another 11 studies with long term follow-up (a mean of 41 months) after 3rd degree tears and 20 to 59% (mean 40%) reported anal incontinence symptoms . Responses to this national survey showed the principle area of disagreement to be 'the optimal method of repair'. However, the fact that half of all obstetricians now claim to be using the 'overlapping' method suggests that there is a change in practice taking place, despite the absence of good quality evidence to support this. Given that untutored use of the overlap method could potentially increase morbidity (as it requires more dissection and mobilisation prior to repair and could also result in a sphincter which is too tight), this may be considered an area of concern. In a recent 5 year follow-up study of incontinent women who had a secondary overlapping repair for obstetric trauma , although 50% improved only 4 of 38 patients were totally continent. The overall success of the overlap method 'seemed to deteriorate with time'. The authors wondered whether the technique itself contributed to this deterioration. Clearly there is an urgent need for further properly controlled trials of method of repair, with adequate long-term follow-up. As shown in our survey, there is a widespread interest in participating in a controlled trial in this area. If advantage is taken of this 'window of opportunity', considerable benefits for women should arise. The finding that only one third of UK consultant obstetricians reported that they were adequately trained to perform anal sphincter repairs requires further attention as this may well have serious clinical and medico-legal implications . We believe that obstetricians need more intensive and focused training in OASI and repair. A series of hands-on-workshops in the management of OASI, utilising a specifically designed model and animal models, has been initiated at St George's Hospital, London. Although it has been suggested that as colorectal surgeons are trained to perform a secondary sphincter repair they should be performing the primary repairs, there was little evidence that current systems could support such a significant change in organisation. Indeed, our survey showed that most coloproctologists have little or no experience of managing acute OASI. A lack of understanding of the circumstance of childbirth by coloproctologists may explain why 30% believed that a colostomy is appropriate management. Further interdisciplinary co-operation is clearly required. It is apparent from our survey that most obstetricians use Vicryl sutures and antibiotics, and prescribe stool softeners after repair. The ideal suture material for primary sphincter repair is not known although good results have been described using a delayed absorbable monofilament material such as PDS and some surgeons prefer prolene. Further research in this area is required. Given the large proportions of women who may suffer impaired anal continence even following repaired OASI, it is imperative to establish a system for follow-up. Further research into the most efficient and effective systems is required. Increasing awareness amongst women and community health professionals about the possible sequelae of OASI is important and easy access for appropriate follow-up and further investigation is essential. There is no evidence to indicate the ideal and safest mode of subsequent delivery. Until further research has been undertaken individual cases will need to be managed empirically. Where there has been secondary surgery or where symptoms have taken some time to improve the threshold for elective caesarean section will be lower. Conclusions : To encourage women to consider vaginal delivery positively, adverse outcomes need to be minimised. The results of our literature review and professional survey are informing the development of national guidelines based on current available best evidence . Support for research in this area has a broad mandate and reflects a need noted by both the research community and the research consumer . Randomised controlled trials of overlap versus end-to-end repair are currently underway . It is clearly of great importance that this research is supported and that further studies addressing the detail on training, prevention, recognition and management of OASI are commenced. Competing interests : None declared. Appendix : What is known about the subject / What the paper adds | Appendix see : Appendix Anal Sphincter Paper Authors' contributions : This study was conceived by RF, RJ, AS and SR. RF, RJ, AS, SR and PJ contributed to the design of the study. RF carried out the survey and data collection. RF and PJ analysed the data. The paper was written jointly by all authors. RF and RJ act as guarantors for this paper. Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12052260 TI - Mutational analysis of human profilin I reveals a second PI(4,5)-P2 binding site neighbouring the poly(L-proline) binding site AB - Abstract | Background | Profilin is a small cytoskeletal protein which interacts with actin, proline-rich proteins and phosphatidylinositol 4,5-bisphosphate (PI(4,5)-P2). Crystallography, NMR and mutagenesis of vertebrate profilins have revealed the amino acid residues that are responsible for the interactions with actin and poly(L-proline) peptides. Although Arg88 of human profilin I was shown to be involved in PI(4,5)-P2-binding, it was suggested that carboxy terminal basic residues may be involved as well. Results | Using site directed mutagenesis we have refined the PI(4,5)-P2 binding site of human profilin I. For each mutant we assessed the stability and studied the interactions with actin, a proline-rich peptide and PI(4,5)-P2 micelles. We identified at least two PI(4,5)-P2-binding regions in human profilin I. As expected, one region comprises Arg88 and overlaps with the actin binding site. The second region involves Arg136 in the carboxy terminal helix and neighbours the poly(L-proline) binding site. In addition, we show that adding a small protein tag to the carboxy terminus of profilin strongly reduces binding to poly(L-proline), suggesting local conformational changes of the carboxy terminal alpha-helix may have dramatic effects on ligand binding. Conclusions | The involvement of the two terminal alpha-helices of profilin in ligand binding imposes important structural constraints upon the functions of this region. Our data suggest a model in which the competitive interactions between PI(4,5)-P2 and actin and PI(4,5)-P2 and poly(L-proline) regulate profilin functions. Keywords: Background : The small actin binding protein profilin has multiple binding partners and is thought to play a key-role in the regulation of actin dynamics . Originally, profilin was identified as an actin sequestering protein but recently more complex effects on actin polymerization have been proposed because actin-profilin complexes can add to free barbed ends thereby stimulating actin polymerization . Profilins bind poly(L-proline) sequences and many proteins containing proline-rich stretches have been identified as profilin ligands. Of these the interaction with the enabled/vasodilator stimulated phosphoprotein (Ena/VASP) family is best documented . For several proline-rich proteins a direct link with signal transduction pathways has been described , thus positioning profilins at crossroads of multiple pathways that lead to actin remodeling . With the elucidation of the profilin-beta-actin crystal structure, the residues at the interface of both proteins were identified . Additionally, crystalographic, mutagenesis and spectroscopic studies have addressed the poly(L-proline) binding site and showed that a hydrophobic pocket between the amino and carboxy terminal alpha-helices forms the binding site for poly(L-proline) sequences [,-]. The interaction of profilin with phosphatidylinositol lipids has been functionally studied. In vitro, PI(4,5)-P2 dissociates actin:profilin complexes and these and other authors also demonstrated the specificity of the interaction between profilin I and PI(4,5)-P2 in both micellar form as well as in lipid vesicles . More recently it was shown that phosphatidylinositol (3,4)-bisphosphate and phosphatidylinositol (3,4,5)-triphosphate bind to profilin with even higher affinity than PI(4,5)P2 and that phosphatidylinositol (3,4,5)-triphosphate inhibits profilin sequestering activity much better than PI(4,5)P2. In addition, PI(4,5)-P2, bound to profilin, can only be hydrolyzed by phospholipase Cgamma1 (PLCgamma1), when this lipase is phosphorylated and activated, which occurs in response to transmembrane signaling . This leads to two, not mutually exclusive scenarios that profilins are involved in phosphoinositide metabolism or that PI(4,5)-P2 hydrolysis causes translocation of profilin from the membrane to the cytosol where it can interact with actin or other ligands. This suggests an important role for profilin-phosphoinositide interaction in vivo. The structural basis for this interaction is, however, only partly resolved (see below). The interaction of actin binding proteins with PI(4,5)-P2 is usually assigned to the binding of the negatively charged headgroup of the phoshoinositide to basic amino acids. In agreement with this is that the more positivily charged Acanthamoeba profilin II isoform has highest affinity for PI(4,5)-P2. Similarly, the more basic human profilin I isoform interacts better with PI(4,5)-P2 than does profilin IIa . The identity of the amino acids responsible for binding of profilins to PI(4,5)-P2 is a matter of debate, because there are discrepancies between studies on profilins from lower eukaryotes and from vertebrates . Based on comparison of the crystal structure of the two Acanthamoeba profilin isoforms, Fedorov and co-workers proposed that a surface with positive electrostatic potential, formed by residues 71, 80, 81 and 115 (corresponding to residues 74, 88, 90 and 125 in human profilin), was the main PI(4,5)-P2 binding site in Acanthamoeba profilin. This surface largely overlaps with the actin binding surface and hence this model explained the observed competition between actin and PI(4,5)-P2 for binding to profilin . Mutagenesis of the yeast homologue partially confirmed this model as residue 71, but not residue 80, is implicated in phosphoinositide binding . Based on the structural model, we previously suggested that Glu56 in mammalian profilin IIa would be responsible for the weaker interaction of this isoform because the negative charge of this residue reduces the large, positively charged surface around the hypothetical PI(4,5)-P2-binding site . In profilin I, which has a serine at position 56 this is less the case. In human profilin, however, only Arg88 and not Arg74, was argued to be involved in PI(4,5)-P2-binding since only the mutant in Arg88 showed decreased inhibition of PI(4,5)-P2 hydrolysis by PLCgamma . We and others have speculated that basic residues in the carboxy terminal alpha-helix of vertebrate profilins may be involved in PI(4,5)-P2-binding. First, Yu and coworkers postulated that the residues 126 to 136 (KCYEMSHLRR) of human profilin I are a modified version of the PI(4,5)-P2-binding motif in gelsolin (KSGLKYKK). Second, using photoactivatable homologues of PI(4,5)-P2, it was hypothesized that carboxy terminal basic residues in human profilin I are involved in contacting the negative headgroups of PI(4,5)-P2. Third, the observed competition between poly(L-proline) and PI(4,5)-P2 for binding to profilin is consistent with the proposal that the carboxy terminus of profilin is involved in PI(4,5)-P2-binding . Fourth, we have shown that mammalian profilins I and IIa have clearly different affinities for PI(4,5)-P2, even though their actin binding surface including Arg74 and Arg88, are well conserved. This suggests that still other residues must be involved in PI(4,5)-P2-binding. In this study we experimentally investigated this hypothesis using site directed mutagenesis of human profilin I. Our data clearly show that, in addition to Arg88, also Arg136 in the carboxy terminal helix has a major contribution to PI(4,5)-P2-binding. Given that mutant R136D, but not R88A, displays wild type actin binding activity, we propose that the PI(4,5)-P2 and actin binding sites only partly overlap. Our data also suggest a connection between PI(4,5)-P2-binding and the interaction with proline-rich ligands, since the profilin IIa mutant W3A, defective in poly(L-proline) binding shows increased PI(4,5)-P2-binding. Given the observed conformational changes upon poly(L-proline) and PI(4,5)-P2-binding we propose that correct orientation of the terminal alpha-helices is important for ligand binding. This is strengthened by the fact that the addition of a myc tag to the carboxy terminal helix of profilin IIa abolishes poly(L-proline) binding completely. Results and discussion : Mutational analysis of human profilin I | The goal of this study was to get a better insight into the structural basis of the interaction of vertebrate profilins with PI(4,5)-P2. To investigate the possible role of the above mentioned residues (see Background) in PI(4,5)-P2-binding and to obtain profilins that have reduced PI(4,5)-P2-binding capacity, we created a set of single and double mutants in the residues Ser56, Arg74, Arg88, Arg135 and Arg136 of human profilin I (Figure and Table ) and a mutant W3A defective in poly(L-proline) binding. Figure 1 | Three dimensional structure of human platelet profilin I (PDB entry, 1 fik). Three dimensional structure of human platelet profilin I (PDB entry, 1 fik). Helices are shown in red, beta-strands in blue, beta-turns in green and loops in grey. Residues mutated in this study are indicated with space filling: Trp3 in yellow (poly(L-proline) binding), Ser56 and Arg135 in pink, Arg 74 in green (actin binding), Arg88 and Arg136 in blue (PI(4,5)-P2 binding). Table 1 | Interaction of wild type and mutant profilins with a proline-rich peptide Wild type human profilin I as well as the mutants listed in Table were expressed in E. coli and all could be purified by poly(L-proline) affinity chromatography, except for the W3A mutant which does not bind poly(L-proline) (see below). We initially included the R88E mutant, but due to its instability, we were unable to purify this protein in sufficient amounts for biochemical analysis. Mutants have a similar fold and stability as wild type profilin I | We first probed whether the introduced mutations did not affect the conformation and stability by analyzing the conformational integrity of the mutants using circular dichroism (CD) spectra. We measured and compared spectra for wild type and mutant profilins between 184 and 260 nm . All mutants adopt a very similar fold as wild type profilin I. The wavelengths at which maximal and minimal peak values are observed do not or only slightly change. The small shoulders at lower wavelength, observed for the double mutants with R136D, suggests that mutation of this residue to aspartic acid affects in some way the stability or the position of the carboxy terminal alpha-helix. The differences are however too small to be interpreted quantitatively. Figure 2 | Circular dichroism spectra show that the mutants have a similar fold as wild type profilin I. Circular dichroism spectra show that the mutants have a similar fold as wild type profilin I. The molar ellipticity per residue weight is shown. The spectra of several single mutants (A) and of double mutants with altered PI(4,5)-P2 binding (B) are compared with that of wild type profilin I. To further test the stability of the mutants, especially the ones that show greatly altered binding to PI(4,5)-P2 (see below), we measured urea denaturation curves . For R136D and R88A/R136D we observed a very small shift of the transition to lower urea concentration when compared to wild type profilin I. On the contrary, R88E/R136D, which has the most pronounced phenotype (see below) displays a denaturation curve very similar to that of wild type profilin I. Together, these data show that the mutants are stable and correctly folded under the conditions used in the assays described below. Figure 3 | Urea denaturation curves for human profilin I and the three mutants that have strongly reduced PI(4,5)-P2 binding. Urea denaturation curves for human profilin I and the three mutants that have strongly reduced PI(4,5)-P2 binding. For each profilin the ratio of the intrinsic fluorescence (F) at two different wavelengths F(352 nm)/F(332 nm) is plotted versus the urea concentration. Wild type profilin I (closed squares), R136D (open circles), R88A/R136D (open triangles), R88E/R136D (closed circles). The inserted table lists the urea concentration at the midpoint of the fluorescence transition. These values are a measure for the stability of the proteins. Poly(L-proline) binding | To sensor more subtle effects on the poly(L-proline) binding of the mutants, we used surface plasmon resonance technology to monitor the binding of the mutants to the (GP5)3 peptide derived from VASP. The measured reasonance units (RU) for each mutant at three different concentrations are given in Table . Although it is not possible to calculate a Kd for profilin I by this method , from the obtained RU-values we can deduce relative affinities for the mutants as compared to wild type profilin I . The most severe effects are observed for R135D, R136D and double mutants containing one of these mutations. This is logical because Arg135 and Arg136 are located in the carboxy terminal alpha-helix, which is involved in poly(L-proline) binding. These residues do, however, not directly contact the proline-rich peptide nor do they stabilize any of the crucial poly(L-proline) binding residues . Instead they are oriented outward, away from the poly(L-proline) moiety in the co-crystal. Therefore, the mutations may induce a conformational change in the carboxy terminal helix, which distorts correct orientation of the poly(L-proline) binding residues. But as judged from the CD-spectra and modeling experiments, this structural change is probably very subtle . In addition, mutations may inhibit or facilitate the previously observed conformational changes that occur in profilin upon binding of poly(L-proline) . Even though mutations at positions 56, 74 or 88 and combinations thereof are distant from the poly(L-proline) binding site, they also result in lowered poly(L-proline) binding. Remarkably, mutations in this region in yeast profilin caused a similar phenotype . Apparently these mutations cause allosteric conformational changes, resulting in less efficient binding of the proline-rich peptide. Interaction of mutants with actin | We determined the dissociation constants of our mutants for alpha-skeletal muscle actin using capped filaments . Under these conditions, profilin displays only G-actin sequestering activity. In addition, we studied the effect of each mutant on non-steady state actin polymerization . To analyze the obtained curves we determined the amount of F-actin formed at a time point (indicated in Figure as T1/2) where the amount of F-actin in the absence of profilin is 50% of the amount formed after 1500 sec. In the presence of WT profilin I, only 12% of F-actin is formed at this time point. The values for the mutant profilins are given in Table . Figure 4 | Time course of alpha-actin polymerization in the absence or presence of several mutant profilins. Time course of alpha-actin polymerization in the absence or presence of several mutant profilins. 10 muM actin and 5 muM profilin are pre-incubated prior to addition of KCl and MgCl2 to a final concentration of 100 mM and 2 mM, respectively. Curves for actin alone (closed triangles), or in the presence of either wild type profilin I (closed circles), R74E (open squares), R136D (open circles) are shown. T1/2 is the time point where the F-actin amount in the actin alone sample reaches 50% of the total F-actin formed after 1500 sec. For each profilin I mutant the percentage of F-actin at T1/2 is determined and given in Table . Table 2 | The interaction of wild type and mutant profilins with alpha-actin. We could not calculate a Kd value for R74A, R74E, S56E/R74E, S56E/R74A and R74E/R88E because the concentration of the actin-profilin complex was nearly zero, leading to very high Kd estimates. This is consistent with the observation that these mutants have no activity in the time course polymerization assay. As determined from the crystal structure of the actin-profilin complex and mutagenesis studies , Arg74 is a crucial residue for actin binding, since it forms a salt bridge with the carboxyl group of Phe375. Consequently, changing the arginine to an alanine or glutamic acid abolishes this interaction completely. Arg88 is also part of the actin-profilin interface, but changing it to alanine decreases the affinity only three-fold, indicating that the binding is less stringent than for Arg74. Mutating Arg88 to leucine or to glutamic acid in combination with S56E (which on its own has no effect), however, abolished actin binding completely. Arg135 and Arg136 locate in the carboxy terminal helix on the opposite side of the molecule and do not participate in actin binding. As a consequence, mutations in these residues do not affect the affinity for actin to a significant extent (Figure and Table ). PI(4,5)-P2 binds to two distinct regions in human profilin I | We used microfiltration and gel filtration to assay the ability of the mutants to bind PI(4,5)-P2 (Figure , Table ). The results of both assays were comparable. Based on analogy with invertebrate profilins (see background) and combined with sequence comparison of profilin I and IIa, we expected S56E to contribute negatively to PI(4,5)-P2-binding. This is, however, not the case and thus this amino acid difference between profilin I and IIa cannot explain the different affinities of the two profilin isoforms for PI(4,5)-P2. A further difference with invertebrate profilins is the observation that mutating Arg74 to leucine, glutamic acid or alanine (this study and ) does not significantly affect PI(4,5)-P2-binding. Since substitution to an acidic residue at this position results in only a slight effect, we consider the contribution of Arg74 in PI(4,5)-P2-binding to be of minor importance. Consequently, also the double mutants S56E/R74A and S56E/R74E show nearly wild type PI(4,5)-P2-binding. Previously, it was shown that Arg88 is involved in PI(4,5)-P2-binding of human profilin I , in agreement with several crystal structures showing a phosphate or sulfate anion associated with Arg88 and surrounding residues . In our assays, R88A has a small effect on PI(4,5)-P2-binding . Unfortunately we were unable to purify mutant R88E for which we expected a more pronounced phenotype. The effect of the latter mutation can, however, be inferred from the double mutants R74E/R88E and S56E/R88E. Both mutants show reduced PI(4,5)-P2-binding, compared to S56E, R74E and S56E/R74E which display nearly wild type binding capacity . Figure 5 | PI(4,5)-P2-binding of profilin mutants. PI(4,5)-P2-binding of profilin mutants. A. Microfiltration of profilin-PI(4,5)-P2 complexes. 4 muM profilin is incubated with increasing concentrations of PI(4,5)-P2 as indicated and applied to a filter with MWCO of 30.000. Non-bound profilin passes through the filter upon centrifugation. The flowthrough is analyzed by SDS-PAGE and is shown here for wild type profilin, R135D, R136D and R135A/R136A. B. Examples of gel filtration experiments. Profilin (10 muM) was pre-incubated with increasing concentrations of PI(4,5)-P2 and run over a SMART Superdex75 gel filtration column. Free profilin elutes at 1.62 ml, while the profilin-PI(4,5)-P2 complex elutes in the void (0.96 ml). The profilin peak shifts to the void fraction upon binding to PI(4,5)-P2. Elution pattern of wild type profilin alone (black line), profilin with 40 muM PI(4,5)-P2 (dark grey line) and profilin with 150 muM PI(4,5)-P2 (light grey line) are shown. We calculated the peak surface of free profilin to determine the percentage of bound profilin for different PI(4,5)-P2 concentrations. These data were then plotted in curves as shown in C. C. Percentage of bound profilin in function of PI(4,5)-P2 concentration as determined from the gel filtration curves. Wild type profilin (closed circle), R136D (open circle), R88A (closed triangle), R88A/R136D (open triangle) and R88E/R136D (closed square) in the gel filtration experiment. The concentration of PI(4,5)-P2 where 50% of profilin is bound to the micelles was derived from these curves and is given in Table for the different mutants. Table 3 | PI(4,5)-P2-binding of mutants assayed by gel filtration Interestingly, mutant R136D has a more pronounced effect than R88A (Figure and Table ). In contrast, mutating the neighboring residue Arg135 has only a small effect on PI(4,5)-P2-binding. Combining mutations in Arg88 and Arg136 has an additive effect : R88A/R136D and R88E/R136D show a much larger reduction in PI(4,5)-P2-binding than the single mutants . This suggests that the reduced PI(4,5)-P2-binding seen for R136D is due to a direct loss of an interaction. Although we cannot exclude contribution from allosteric effects, modeling experiments substituting R136 with an aspartic acid (data not shown) show no significant change in position of the side-chain or of the carboxy terminal alpha-helix. We conducted gel filtration experiments at high profilin to PI(4,5)-P2 ratio's for wild type profilin I and the R136D mutant to assess if the mutation affects overall saturable binding ability. This seems, however, not to be the case (data not shown), since we found for both wild type and mutant a ratio of ten profilin molecules per PI(4,5)-P2 micelle, suggesting a stoichiometry of 1:8 profilin : PI(4,5)-P2 molecules, consistent with a previous report . Depending on the assay conditions used, variable values for the stoichiometry of the profilin : PI(4,5)-P2 complex were found, varying between 1:4 and 1:10 . Given this 1:8 stoichiometry, it is difficult to observe the loss of one interaction using PI(4,5)-P2 micelles. We note, however, that in case of the mutant higher concentrations of profilin and PI(4,5)-P2 than for wild type profilin were required to obtain saturation, in agreement with the lower affinity of this R136D mutant Lassing and Lindberg showed that the inhibition on actin polymerization of wild type human profilin I decreases in the presence of PI(4,5)-P2. If Arg136 is involved in PI(4,5)-P2-binding, then this mutant should be less affected in its inhibitory activity in the presence of PI(4,5)-P2. This is indeed what we observe . R136D behaves similar to wild type profilin I in the absence of PI(4,5)-P2 (Figure and ). In the presence of a 9-fold molar excess of PI(4,5)-P2 we observe, however, a significant difference. For R136D we measure only a small reduction in sequestering activity compared to an almost complete inhibition of the sequestering activity of wild type profilin I. In the presence of a 25-fold molar excess of PI(4,5)-P2, however, R136D loses its sequestering activity completely (data not shown), indicating that the mutation did not entirely abolish PI(4,5)-P2-binding. This is consistent with the results from the gel filtration experiment and implicates a role for other residues such as Arg88. Figure 6 | PI(4,5)-P2 inefficiently competes with actin for binding to R136D profilin I. PI(4,5)-P2 inefficiently competes with actin for binding to R136D profilin I. The curves shown are : 8 muM Mg2+-ATP-G-alpha-actin (5% pyrene labeled) alone (closed triangles) or with 4 muM wild type profilin I (closed circles), 4 muM R136D (open circles), 4 muM wild type profilin I and 36 muM PI(4,5)-P2 (closed squares), 4 muM R136D and 36 muM PI(4,5)-P2 (open squares). Recently we demonstrated that profilin IIa has a lower affinity for PI(4,5)-P2 than profilin I . This can be explained with the data presented in this paper. In profilin I, Arg136 is important for PI(4,5)-P2-binding. In profilin IIa, there is an aspartic acid at this position (Asp136) and the profilin I R136D mutant thus mimics the profilin IIa isoform with respect to PI(4,5)-P2-binding. An indirect role of tryptophan 3 in PI(4,5)-P2-binding | Based on experiments with photoactivatable PI(4,5)-P2 analogues, Chaudhary and coworkers (1998) suggested that hydrophobic residues in the amino terminal helix are involved in the interaction with PI(4,5)-P2. Trp3, the fluorescence of which is quenched in the presence of PI(4,5)-P2, is spatially close to Arg136 (see Figure ). Therefore we mutated the former residue to alanine, thereby reducing the hydrophobic moiety. Trp3 is a crucial residue for the interaction of profilin with poly(L-proline) [,-,] and as expected the W3A mutants of profilin I and IIa lack poly(L-proline) binding and were thus purified using alternative methods (see Materials and Methods). The dissociation constant for the actin-profilin I W3A-complex was similar to that of wild type profilin I . The profilin I W3A mutant did not show a significant decrease in PI(4,5)-P2-binding, suggesting this residue does not directly contribute to the interaction. Interestingly, the profilin IIa W3A mutant shows increased affinity for PI(4,5)-P2 and the affinity is comparable with that of wild type profilin I . Given the profilin I W3A data presented here and in view of the conformational changes observed upon ligand binding , we propose that mutating Trp3 in profilin IIa promotes/induces a conformation which is more competent for PI(4,5)-P2-binding (see below). Figure 7 | Profilin IIa W3A mutant has increased affinity for PI(4,5)-P2. Profilin IIa W3A mutant has increased affinity for PI(4,5)-P2. Percentage of bound profilin in function of PI(4,5)-P2 concentration as determined from gel filtration experiments described in Figure . The concentration of PI(4,5)-P2 where 50% of profilin is bound to the micelles was derived from these curves and is given in Table . Model for regulation of profilin-ligand interactions | The data presented here show that in addition to Arg88, Arg136 is involved in PI(4,5)-P2-binding of mammalian profilin I. Based on our quantitative gel filtration assay, the contribution of Arg136 is in fact more important than that of Arg88 and the double mutant hardly binds PI(4,5)-P2 micelles. We conclude that the PI(4,5)-P2 binding sites of profilin are located in two distinct regions of the molecule that are approximately 31 A apart (see Figure ). It is remarkable that there are no corresponding positively charged residue(s) in the carboxy terminus of yeast and Acanthamoeba profilins that could account for a similar interaction as found here for human profilin I. This may indicate that the structural basis for the interaction of PI(4,5)-P2 with profilins from lower and higher eukaryotes is partially different. We also note that Acanthamoeba profilin II has a ten fold lower affinity for PI(4,5)-P2 than human profilin I . Both PI(4,5)-P2-binding regions in vertebrate profilins are implicated in the interaction with another profilin ligand. Arg136 is close to several poly(L-proline) binding residues. Not surprisingly, mutations in Arg136 have also strongly decreased poly(L-proline) affinity, although Arg136 itself is not directly contacting proline-rich ligands. On the other hand, Arg88, involved in PI(4,5)-P2-binding is also part of the actin binding site . A partial overlap of actin- and PI(4,5)-P2-binding sites was also observed for actophorin and gelsolin , suggesting this is the basis for a general regulatory mechanism for several actin binding proteins, whose function is inhibited by PI(4,5)-P2. Our data thus offer an explanation for the previously observed competition between PI(4,5)-P2 and the two other profilin ligands : actin and poly(L-proline) . This offers a nice model for the regulation of profilin with its different ligands. Since PI(4,5)-P2 inhibits both actin and poly(L-proline) binding , it is conceivable that PI(4,5)-P2 may have a master regulatory function in the cell. When PI(4,5)-P2 is hydrolyzed after cell stimulation, profilin may be set free to interact both with proteins containing proline-rich regions and with actin to regulate actin dynamics. The concerted action in vivo of profilin-actin complexes with several proline-rich proteins such as Ena/VASP proteins, N-WASP and formins for the promotion of actin polymerization was suggested previously . Several of our mutants suggest allosteric communication within vertebrate profilins. Arg88 mutants have reduced poly(L-proline) binding, although this residue is not part of the poly(L-proline) binding pocket. Conversely, W3A (in profilin IIa) influences PI(4,5)-P2-binding, but appears not to be directly involved in PI(4,5)-P2-binding as suggested by the data on profilin I W3A, although a W3N mutation in profilin I results in a higher affinity for PI(4,5)-P2. These results suggest that the interaction of profilin with PI(4,5)-P2 and poly(L-proline) involve conformational changes, which have been experimentally observed before . The interaction of profilin with PI(4,5)-P2 induces an increase in alpha-helical content . We propose that the local structure of the neighboring binding sites may change upon binding of PI(4,5)-P2 and poly(L-proline). The fact that W3A of profilin I binds PI(4,5)-P2 similar to wild type, suggests that profilin I already has the correct conformation for optimal binding of PI(4,5)-P2 and that mutating Trp3 to alanine does not ameliorate this conformation further (see also below), while an asparagine at position 3 does have a positive effect . In contrast, the W3A mutation in profilin IIa increases the affinity for PI(4,5)-P2, suggesting that this mutation induces a conformational change which optimizes the interaction with PI(4,5)-P2 despite the presence of an aspartic acid at the nearby position 136. The profilin IIa structure is, however, optimal for strong poly(L-proline) binding. In modeled and energy minimized profilin IIa structures we observed that the terminal alpha-helices are further apart from each other suggesting better access to the poly(L-proline) binding cleft . From this point of view, it is logical to assume that changing the position of these terminal alpha-helices has dramatic effects on ligand binding. This idea is consistent with our observation that the addition of a myc-tag to the carboxy terminal end of profilin IIa results in the dramatic loss of poly(L-proline) binding despite the fact that all known proline interacting residues are present . The suggested conformational change in profilin IIa-myc does, however, not significantly influence the affinity for actin . Similarly, mouse profilin IIb, which has six additional amino acids at its carboxy terminus, does not bind poly(L-proline) . In addition, it has been reported that both amino- and carboxy terminal GFP fusion proteins of mammalian profilins display a dramatic loss in poly(Lproline) binding . Some fusion proteins even lack complete poly(L-proline) binding. Therefore we believe that the correct positioning of the terminal alpha-helices of profilin is a primary requirement for ligand interaction. It is clear that any distortion of the alpha-helices will reduce the interaction with poly(L-proline). Figure 8 | Addition of carboxy terminal myc-tag to profilin IIa dramatically reduces poly(L-proline) binding. Addition of carboxy terminal myc-tag to profilin IIa dramatically reduces poly(L-proline) binding. A. Biacore binding curves for 100 muM wild type profilin IIa (blue), 1 muM wild type profilin IIa (green), or 100 muM profilin IIa-myc (red) to the (GP5)3 peptide derived from VASP. Resonance units (R.U.) are a measure for the number of profilin molecules retained by the peptide on the sensor chip and this is also concentration dependent (see B.). Even at a 100 times higher concentration, profilin IIa-myc (100 muM) binds less efficient to the peptide than wild type profilin IIa (1 muM). B. R.U. values obtained with different concentrations of wild type profilin IIa and profilin IIa-myc. Note that the value for 100 muM wild type profilin IIa is different from the one in Table , due to a different amount of peptide coupled to the sensor chip. Conclusions : We have identified Arg136, besides the previously identified Arg88, of human profilin I as an important residue for the interaction with PI(4,5)-P2. Since Arg136 is part of the poly(L-proline) binding helix and Arg88 is located in the actin binding surface, we suggest that the interaction of profilin with its different ligands is regulated by competitive interactions, which may be partly allosteric. Our results also indicate that the position of the two large terminal alpha-helices is crucial for optimal ligand binding. The addition of (protein or peptide) tags to the carboxy terminus results in dramatic decreased affinity for poly(L-proline) ligands. Conceivably, this will result in altered interactions in cells and in vivo data obtained with tagged profilin isoforms should be carefully (re)interpreted. Materials and methods : Profilin mutagenesis and purification | The profilin I cDNA amplified by polymerase chain reaction from a human cDNA library was subcloned into pET11d . Site directed mutagenesis was performed by polymerase chain reaction with mutated oligonucleotide primers and pfu polymerase. Mutations were verified by sequencing. MC1061 E. coli harboring the pT7POL26 plasmid were used for expression of wild type and mutant profilin I, Proteins were subsequently purified by poly(L-proline) affinity chromatography . W3A mutants do not bind poly(L-proline), thus the flow-through of the poly(L-proline) column was loaded onto a DEAE column equilibrated in buffer A (20 mM Tris-HCl, pH 8.1, 1 mM EDTA, 1 mM DTT). The column was eluted with a 0 to 500 mM NaCl gradient in buffer A. Profilin eluted with 60 to 130 mM NaCl. The profilin containing fractions were pooled and loaded on a MonoQ column. The flowthrough of this column contained profilin and only very few other proteins. These contaminating proteins were then removed by gel filtration in buffer A. Other Protein preparations | We purified actin from rabbit skeletal muscle and isolated it as calcium G-actin by Sephadex G200 chromatography in G-buffer (5 mM Tris-HCl, pH 7.7, 0.1 mM CaCl2, 0.2 mM ATP, 0.2 mM dithiothreitol, 0.01% sodium azide) . Actin was pyrene labeled on cysteine 375 . Gelsolin was purified from human plasma . Circular dichroism | We performed CD measurements in the far UV region (184 --260 nm) for WT and mutant profilins at a concentration of 15 muM in 7 mM TRIS/HCl, pH 8 in a JASCO J-170 spectropolarimeter using a 1 cm pathway cell. The step resolution was 0.5 nm and the scan speed 20 nm/min. For each sample the average of 9 scans was obtained and spectra were normalized for concentrations. Denaturation curves | Profilin was diluted to 2 muM in increasing concentrations of urea (0 to 8 M) in 20 mM Tris-HCl pH 8.1, 1 mM EDTA, 1 mM DTT. The samples were incubated for 15 min. at room temperature and the intrinsic fluorescence change during a wavelength scan between 300 and 400 nm was measured in a Hitachi F4500 spectrophotometer with the excitation wavelength set at 295 nm. We recorded a shift of the emission peak from 332 nm to 352 nm upon denaturation with urea. For each sample we plotted the ratio F(352 nm)/F(332 nm) versus the concentration of urea in that sample (see Figure ) . Polyproline binding | A (GP5)3 peptide, derived from VASP, was used to compare the affinities of the profilin I mutants on a BiacoreX (Pharmacia). The amino terminally biotinylated peptide was coupled to a streptavidin coated Biacore biosensor chip (Pharmacia). The experiments were carried out and analyzed as described in . Actin binding assays | The affinity of the profilin mutants for alpha-actin was determined using gelsolin capped filaments as described in . To determine the effect on non-steady state actin polymerization we pre-incubated 10 muM actin (5 % pyrene labeled) with or without 5 muM profilin for 15 minutes at room temperature prior to the addition of a final concentration of 2 mM MgCl2 and 100 mM KCl. The fluorescence change was recorded using a Hitachi F4500 spectrophotometer. PI(4,5)-P2-binding | Microfiltration was performed as described using 4 muM profilin and different concentrations of PI(4,5)-P2 (Sigma) as indicated in Figure . For gel filtration experiments, 10 muM profilin was pre-incubated with PI(4,5)-P2 micelles for 30 min on ice prior to loading on a Superdex75 gel filtration column (SMART, Pharmacia). The peak surface of free profilin was determined and used to calculate the percentage of bound and free profilin in each sample. The competition experiment between actin and PI(4,5)-P2 was performed with 8 muM Mg2+-ATP-G-alpha-actin (5% pyrene labeled), 4 muM profilin and 36 muM PI(4,5)-P2 in 5 mM Tris-HCl, 0.2 mM ATP, 0.2 mM dithiothreitol, pH 7 in the absence of Ca2+ and Mg2+ to avoid precipitation of PI(4,5)-P2. Profilin and PI(4,5)-P2-micelles were incubated for 10 minutes on ice prior to addition of actin and subsequent incubation for 10 minutes at room temperature. Polymerization was started by adding KCl to a final concentration of 50 mM. Abbreviations : Circular dichroism : CD; enabled : Ena; phosphatidylinositol 4,5-bisphosphate : PI(4,5)-P2; phospholipase Cgamma1 : PLCgamma1; vasodilator stimulated phosphoprotein : VASP; reasonance units : RU. Authors' contributions : A.L. participated in design of the study, carried out the mutagenesis, the stability and CD experiments and the actin and PI(4,5)P2 binding studies and drafted the manuscript. V.J. purified the proteins and carried out the Biacore and microfiltration experiments. D.D. helped with the mutagenesis. J.V. participated in the design of the study. C.A. conceived the study, participated in the coordination and in the design of the study. Backmatter: PMID- 12069692 TI - Gene expression profile of HIV-1 Tat expressing cells: a close interplay between proliferative and differentiation signals AB - Abstract | Background | Expression profiling holds great promise for rapid host genome functional analysis. It is plausible that host expression profiling in an infection could serve as a universal phenotype in virally infected cells. Here, we describe the effect of one of the most critical viral activators, Tat, in HIV-1 infected and Tat expressing cells. We utilized microarray analysis from uninfected, latently HIV-1 infected cells, as well as cells that express Tat, to decipher some of the cellular changes associated with this viral activator. Results | Utilizing uninfected, HIV-1 latently infected cells, and Tat expressing cells, we observed that most of the cellular host genes in Tat expressing cells were down-regulated. The down-regulation in Tat expressing cells is most apparent on cellular receptors that have intrinsic receptor tyrosine kinase (RTK) activity and signal transduction members that mediate RTK function, including Ras-Raf-MEK pathway. Co-activators of transcription, such as p300/CBP and SRC-1, which mediate gene expression related to hormone receptor genes, were also found to be down-regulated. Down-regulation of receptors may allow latent HIV-1 infected cells to either hide from the immune system or avoid extracellular differentiation signals. Some of the genes that were up-regulated included co-receptors for HIV-1 entry, translation machinery, and cell cycle regulatory proteins. Conclusions | We have demonstrated, through a microarray approach, that HIV-1 Tat is able to regulate many cellular genes that are involved in cell signaling, translation and ultimately control the host proliferative and differentiation signals. Keywords: Background : Whole-genome expression profiling exemplified by the development of DNA microarrays represents a major advance in genome-wide functional analysis . In a single assay, the transcriptional response of each gene to a change in cellular state can be measured, whether it is a viral infection, host cell cycle changes, chemical treatment, or genetic perturbation. Specifically, systematic approaches for identifying the biological functions of cellular genes altered during these changes, such as HIV-1 infection, are needed to ensure rapid progress in defining significant host and viral genome sequences in directed experimentation and applications. Therefore, host cellular states can be inferred from the expression profiles, and the notion that the global transcriptional response constitutes a detailed molecular phenotype, such as class discovery, class prediction, drug target validation, and the classification of tumors by expression profiling has begun to receive considerable attention . Since its discovery, much of the mainstream human immunodeficiency virus type 1 (HIV-1) Tat research has focused on its ability to activate the HIV-1 LTR. However, to date, besides the transactivation activity on the HIV-1 promoter, few other effects exerted by HIV-1 Tat on cellular and viral genes has also been observed. The Tat protein has been shown to transcriptionally repress host cellular genes and be involved in the immunosuppression associated with viral infection. For instance, HIV-1 infection is able to down-regulate major histocompatibility complex type I (MHC-I) by various different viral proteins, including Tat which represses the transcription of MHC-I, Vpu which retains nascent MHC-I chains in the endoplasmic reticulum, and Nef which can mediate selective internalization of MHC-I molecules from the plasma membrane. MHC class I gene expression has also been shown to be reduced upon infection with the wild-type LAI virus or a Tat exon one recombinant virus . Tat has been shown to down-regulate mannose receptor, EDF-1, CD3-gamma, and TCR/CD3 surface receptor . Tat reduces mannose receptor levels and promoter activity in mature macrophages and dendritic cells by interfering with the host transcriptional machinery; resulting in decreased levels of surface mannose receptor needed for Ag (mannosylated albumin uptake) or pathogen capture (Pneumocystis carinii phagocytosis), and eventual delivery to MHC class II-containing intracellular compartments . EDF-1, a gene down-regulated when endothelial cells are induced to differentiate in vitro, was shown to be down-regulated by Tat at the transcriptional level, resulting in the inhibition of endothelial cell growth and in the transition from a nonpolar cobblestone phenotype to a polar fibroblast-like phenotype . When examining the in vivo effects of HIV-1 Tat protein in the Xenopus embryo, it was found that upon injection of synthetic Tat mRNA into zygotes, a marked delay in gastrulation occurred. This led to the altered specification of the anterior-posterior axis and partial loss of the anterior embryo structures. Mechanistically, HIV-1 Tat elicited a general suppression of gene expression, including that of Xbra and gsc, two early genes whose expression are required for proper gastrulation . In relation to the cell cycle, Tat has also been shown to bind to p53 and inhibit the transcription of p53 responsive elements, such as the p21/Waf1 gene promoter. Consequently, upon introduction of stress signals (e.g., gamma irradiation), HIV-1-infected cells lose their G1/S checkpoint, enter the S-phase inappropriately, and apoptose . Finally, the inhibition of Tat on translational machinery has also been noted. The potential translational inhibitory effects of the TAR RNA region is mediated by the activation of p68 (the interferon-induced 68-kilodalton protein kinase) kinase, which was down-regulated by Tat during productive HIV-1 infection . Although the mechanism of the host cellular down-regulation remains largely unknown, few reports have attempted to decipher the mechanism of the observed inhibition. For instance, the addition of Tat to PC12 cells up-regulated the expression of the inducible cAMP early repressor (ICER), a specific member of the cAMP-responsive element modulator transcription factor family, in a cAMP-dependent manner. In turn, ICER overexpression abrogated the transcriptional activity of the TH promoter, strongly suggesting ICER's involvement in Tat-mediated inhibition of gene expression . Aside from induction of ICER, Tat is capable of forming complex (es) with a component of TFIID, TAFII250 and Tip60 both of which contain histone acetyltransferase (HAT) activity. In these cases, Tat-TAFII250 and Tat-Tip60 do not affect the transcription from the HIV-1 LTR, but interfere with the transcription activity of cellular genes. It is postulated that different targets of HATs by Tat have different consequences. The interaction of Tat with p300/CBP and P/CAF stimulates its ability to transactivate LTR-dependent transcription, while Tat-TAFII250 or Tat-Tip60 interactions control the transcription of cellular genes. Here to better understand the host response to Tat, we have performed microarray experiments on HIV-1 infected cells expressing the Tat protein. To our surprise many host cellular genes were down-regulated when comparing HIV-1 infected latent cells to uninfected parental cells. Because most, if not all, latent infected cells available to date (e.g., ACH2, U1, J1.1, OM.10) have various expression levels of doubly spliced viral mRNAs, including Tat, Rev, Nef, Vpr, and other accessory proteins, we decided to perform the microarray in a system where Tat was constitutively expressed; asking whether Tat by itself, or in the absence of other accessory proteins, could still down-regulate host cellular genes. Consistent with latently infected cells, we found many cellular genes to be down-regulated in Tat expressing lymphocytes. The down-regulation is most apparent on cellular receptors that have intrinsic receptor tyrosine kinase (RTK) activity and signal transduction members that mediate RTK function; including the Ras-Raf-MEK pathway, and co-activators such as p300/CBP and SRC-1, which mediate gene expression related to hormone receptor genes. Interestingly, we also observed up-regulation of S-phase genes, as well as ribosomal genes involved in translation. Functionally, down-regulation of receptors may allow latent HIV-1 infected cells to either hide from the immune system or avoid extracellular differentiation signals normally regulated by receptors. Up-regulation of S-phase and translation genes may allow speeding of cells through the S-phase and subsequent accumulation at the G2 phase, where most of the cellular and viral translation may take place. Therefore, the presence of Tat may not only control activated transcription on HIV-1 LTR, but also aid in the subsequent translation of viral mRNA in the cytoplasm. Results and discussion : Host expression profiling in a sufficiently large and diverse set of profiles could allow additional hypotheses to be drawn regarding the function of genes based on the regulatory characteristics of their own transcripts . Here, we describe the effect of one of the most critical viral activators, Tat, involved in HIV-1 infection and pathogenesis. The rationale for these experiments came from the fact that many AIDS-infected patients who are either at stage III (non-progressors) or under highly active antiretroviral therapy (HAART) treatment show some level of doubly spliced viral messages in their infected cells. One of these messages, Tat, has been well studied and characterized, both from a viral activator standpoint, its effect on few host cellular genes, and its effect as an extracellular cytokine. However, to date there are no reported Tat gene expression analyses that detect more than a few cellular genes. We performed our microarray analysis first with a cDNA blot of 588 genes, which is known to contain various receptors, cytokines, transcription factors, DNA replication genes, and other additional well characterized genes. Figure shows the results of a microarray experiment from uninfected (CEM) and latently HIV-1 infected (ACH2) cells. By definition, latently HIV-1 infected cells contain integrated HIV-1 sequences in the host genome. To our surprise, we detected many cellular genes that were down-regulated in ACH2 cells as compared to CEM uninfected cells. ACH2 cells, similar to many other latent HIV-1 infected cells including U1, OM10.1, 8E5, and J1-1, express multiple doubly spliced messages including Tat, Rev, Nef, and Vpr . Therefore, it would be difficult to determine which one of these viral open reading frames was in fact controlling the observed changes in host gene expression. Figure 1 | Gene expression analysis of uninfected and HIV-1 infected cells. Gene expression analysis of uninfected and HIV-1 infected cells. A) Both CEM (uninfected) and ACH2 (latently HIV-1 infected) cells were grown to mid-log phase of growth and processed for RNA isolation. Total RNA was labeled with 32P-ATP and hybridized to human cDNA filters (Clontech, 588 genes). Blots were hybridized overnight, washed the next day, and exposed to a PhosphorImager cassette. B) Same as in panel A, except all the 588 genes were plotted as fold change vs. gene index (individual genes). Examples of three genes such as prothymosin-alpha, C-myc, and p21/Waf1, is shown on the diagram. C) Northern blot analysis of prothymosin-alpha, C-myc, p21/Waf1 and ubiquitin using 10 mug of total RNA, separated on 0.8% formaldehyde gel, and probed with 40 mer anti-sense oligos against respective genes. Bottom of panel C, last insert shows RNA ethidium bromide stain from CEM and ACH2 cells. Nonetheless, when mapping all the 588 genes, we found that 139 genes were activated above 1 fold and 449 genes were expressed below 1 fold . This is in sharp contrast to latent HTLV-1 infected Tax expressing cells, where more than two-thirds of the same set of genes were activated and scored above one . Some of the genes from HIV-1 latent cells were further processed as control experiments using Northern blot analysis. As can be seen in Figure , and consistent with previously published reports, p21/Waf1 was down-regulated and C-myc and pro-thymosin-alpha were up-regulated in HIV-1 latent cells. Collectively, these experiments point toward host cellular changes in the presence of doubly spliced HIV-1 RNA; however they do not explain whether Tat or other viral genes are responsible for the observed cellular changes. Therefore, we focused our attention on HIV-1 Tat protein, since Tat could readily be detected in immunoprecipitations from 35S labeled latent ACH2, U1, OM10.1, 8E5, and J1-1 cells (20, data not shown). We utilized a well-characterized system of H9 and H9/Tat cell lines for our next set of microarray experiments and increased our repertoire of the known cDNA genes from 588 to 2400. This was accomplished by using glass slides which were printed with cDNAs of 500 bases or higher, and could be used in hybridization with two different sets of RNAs labeled with either Tyramide linked Cy-3 or Cy-5. Results of such an experiment is shown in Figure , where H9 cytoplasmic Poly A+ selected RNA was labeled with Cy-5 and the H9/Tat RNA with Cy-3 prior to hybridization. When all the 2400 genes were plotted, some 695 genes were shown to be up-regulated above one fold and 1705 genes were down-regulated below one fold (right hand graphs). This was consistent with the results obtained from the ACH2 microarray experiment, where more than 2/3 of the cellular genes were down-regulated by one or more of the doubly spliced genes. We then arbitrarily chose a cut-off of three fold change for our next set of analyses. This was mainly because many of our in house microarray experiments with HIV-1, HTLV-1, and HHV-8 infection has shown a reproducible correlation between protein and mRNA levels when gene expression levels were up- or down-regulated by more than three fold (data not shown). A collection of all the genes above and below three fold are shown in Tables , , and . Based on existing literature, we categorized all of these genes into known pathways. For instance, genes in Table belong to receptors, co-receptors, and co-activators, genes in Table are all the translation related factors, and those in Table indicate genes that are involved in cytoskeleton, signal transduction, cell cycle, DNA repair, transcription, and chromatin remodeling processes. All genes have a number ratio of Cy3 to Cy5 (C3/C5) indicating a ratio of mRNA from H9/Tat over H9 cells. A brief name description and gene ID accession number is given to the right hand side of each ratio. All genes are divided into up (3 fold and higher) and down (3 fold and lower) regulated in Tat expressing cells. Below is a brief description of genes that we, along with the existing literature, were able to correlate with proliferative and/or differentiation signals. Figure 2 | Gene expression analysis from Tat expressing cells. Gene expression analysis from Tat expressing cells. Both H9 and H9/Tat cells were grown to mid-log phase of growth, processed for RNA preparation, and labeled with Tyramide linked Cy-5 (H9) or Cy-3 (H9/TAT). Labeled RNAs were hybridized simultaneously to a glass slide containing 2400 known cDNA genes (NEN Inc.). All genes were plotted similar to Figure and genes above & below 1 fold change were plotted (on the right hand side) to show activation and suppression of all genes. Table 1 | Receptors Table 2 | Translation Factors Table 3 | Receptor family members | It has long been known that infection by HIV-1 commonly leads to the down-regulation and the disappearance of CD4 receptors from the plasma membrane, a phenomenon referred to as receptor down-modulation. This, in turn, renders cells refractory to subsequent infection by the same or other viruses that use the CD4 receptor for entry; thus creating a state of super-infection immunity. Results in Table indicate that although few receptor genes were up-regulated, most of the cellular receptors in general, were down-regulated in the presence of the Tat protein. Most of these receptors or membranous proteins were initially discovered from immune or neuronal cells, hence they were given names related to the immune or nervous system. For instance, mRNA for the neuropeptide Y-like receptor (Acc# X71635), which was up-regulated in Tat expressing cells, was initially discovered as a G-protein coupled neuropeptide Y receptor, and later found to be homologous to the co-receptor CCR5 needed for HIV-1 infection of monocyte/macrophage cells. Therefore, most of the receptors listed in Table may in fact be expressed in various tissues and have multiple functions. Consistent with our microarray results on CCR5 up-regulation, experiments performed in peripheral blood mononuclear cells (PBMCs) with soluble Tat has shown selective entry and replication of CCR5 virus into cells . Up-regulation of HIV-1 coreceptor by Tat has also been reported, where a synthetic Tat protein that was immobilized on a solid substrate, up-regulated the surface expression of the chemokine receptors in purified populations of primary resting CD4+ T cells. Also, a similar result was seen from Tat protein actively released by HIV-1 infected cells, implying a potentially important role for extracellular Tat in rendering the bystander CD4+ T cells more susceptible to infection . We therefore tested whether H9/Tat cells, which showed an increase in CCR5 expression, could in fact allow better entry and infection of the CCR5 (R5) virus into cells. Figure shows the result of such an experiment, where H9/Tat cells allowed a better replication profile of the R5 than the CXCR4 (X4) virus. The increase in viral titer peaked after some 18 days of infection with the R5 virus, further implying that the CCR5 co-receptor allowed a better selection of R5 virus in Tat expressing cells. Figure 3 | Functional and physical confirmation of few genes from Tat expressing cells. Functional and physical confirmation of few genes from Tat expressing cells. A) Infection of mono- and T-tropic viruses into Tat expressing cells. Both HXB-2 and BaL strains of HIV-1 were infected into H9 and H9/TAT cells. Supernatants were collected every 3 days and further processed for p24 gag ELISA assays. B) Western blot analysis from H9 and H9/TAT expressing cells using co-activators (SRC-1), DNA damage (DNA-PK), activator (p300), and signal transduction (Ras, RAF, and MAPK) antibodies. TBP stands for TATA binding protein, which served as positive control in western blots. C) Western blot analysis from CEM (uninfected T-cell), ACH2 (infected T-cell), U937 (uninfected promonocytic), U1 (infected promonocytic), and PBMCs treated with purified Tat wild type or K41A mutant (100 ng/ml) proteins. Fifty microgram of whole cells lysates were processed for western blots with anti-DNA-PK, p300, RAF, and TBP antibodies. Another example of co-receptors with multiple functions is the leukotriene family member B4, which was down-regulated in Tat expressing cells (Acc# D89078, Table ). The cysteinyl leukotrienes (CysLT), LTC, LTD, and LTE, were first shown to be essential mediators in asthma . However, when the mouse leukotriene B4 receptor (m-BLTR) gene, was cloned it was shown to have significant sequence homology with chemokine receptors (CCR5 and CXCR4), co-receptors for many different HIV-1 clades . Along the same lines, when cells were infected with 10 primary clinical isolates of HIV-1, leukotriene B4 receptor was primarily utilized for efficient entry into cells which were mainly of the syncytium-inducing phenotype . Therefore, up-regulation of neuropeptide Y-like receptor and down-regulation of leukotriene B4 receptor in Tat expressing cells indicates a selective advantage of one class of virus (CCR5) over another (CXCR4). Other examples of consistency between our microarray results on receptors and the HIV-1 Tat literature, include the down-regulation of gene expression in uPAR (Acc# X74039), IP3 (Acc# D26070, D26351), Glu R flop (Acc# U10302), PPAR (Acc# L07592), alpha-2 macroglobulin receptor protein (Acc# M63959), and receptor tyrosine kinase (Acc# L36645, U66406) genes. The transmembranous urokinase-type plasminogen activator receptor (uPAR; CD87) focuses the formation of active plasmin at the cell surface, thus enhancing directional extracellular proteolysis. Interestingly, the promoter activity of the CD87 gene was shown to decline after infection , implying that post integration of HIV-1 may in fact down-regulate CD87 gene expression. Similarly, inositol 1,4,5-trisphosphate receptors (IP3R) are intracellular calcium release channels involved in diverse signaling pathways and are required for the activation of T lymphocytes . Tat (also implicated as a neurotoxin) has been shown to release calcium from inositol 1,4, 5-trisphosphate (IP3) receptor-regulated stores in neurons and astrocytes causing premature apoptosis . Down-regulation of IP3 may therefore contribute to viral latency and maintenance of an anti-apoptotic state in cells. HIV-1 infection can cause extensive neuronal loss and clinically, a severe dementia. The cause of the neurotoxicity remains unclear as neurons are not infected, but the disturbance of glutamate-linked calcium entry has been implicated. It has been shown that HIV-infected brain has a decrease of mRNA and protein of the GluR-A flop subtype of alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) glutamate receptor in cerebellar Purkinje cells. The observed disturbance of AMPA receptors may contribute to the neurotoxic process in other vulnerable brain regions and clinically to the development of dementia . Interestingly, in a mouse model AMPA receptors in the cortex, striatum, hippocampus, and cerebellum declined by 29 --50% as early as 8 weeks post-retroviral inoculation. Thus, the reduction in AMPA receptor density may contribute to the development of the cognitive abnormalities associated with HIV-1 infection . Finally, patients with AIDS who are receiving therapy with HIV-1 protease inhibitors have been reported to be afflicted with a syndrome characterized by lipodystrophy (fat redistribution favoring the accumulation of abdominal and cervical adipose tissue), hyperlipidemia, and insulin resistance. Potential mechanisms for altered adipocyte function include, direct binding to PPARgamma or inhibition of transcription of PPARgamma promoter . The lipodystrophy syndrome may be a result of the inhibition of 2 proteins involved in lipid metabolism that have significant homology to the catalytic site of HIV proteases; namely cytoplasmic retinoic acid binding protein type 1 and low density lipoprotein-receptor-related protein . An additional mechanism of PPAR down-regulation may be related to Tat expression in latent cells. Translation associated factors | Viruses have evolved a remarkable variety of strategies to modulate the host cell translation apparatus with the aim of optimizing viral mRNA translation and replication. For instance, viruses including Herpes simplex virus type 1 (HSV-1) have been known to induce severe alterations of the host translational apparatus, including the up-regulation of ribosomal proteins and the progressive association of several nonribosomal proteins, such as VP19C, VP26, and the poly(A)-binding protein 1 (PAB1P) to ribosomes . In the case of HIV-1, approximately one infectious HIV-1 genome in an infected cell could be transcribed and translated into 50,000 to 100,000 physical particles . This poses an immense challenge for the virus to be able to transcribe, splice, transport, and translate its RNA into fully packaged virions in a timely fashion. Therefore, it would be advantageous for the virus to set the stage for each successive step necessary for viral progeny formation. One such event is Tat's ability to control genes that aid in translational machinery. As seen in Table , many of the critical components of a functional ribosome, including large subunits L 3, 6, 26, 31, 38, and 41, as well as S 6, 12, 20, and 24, and many of the translation initiation factors are up-regulated by Tat. This would imply that Tat up-regulates many ribosomal genes that may be necessary to produce functional ribosomes needed for viral mRNA translation. Therefore, interfering with translation could provide a new strategy for anti-HIV treatment. Along these lines, when the aminogylcosides (kanamycin, hygromycin B, paromycin and neomycin) due to their ability to inhibit protein synthesis by affecting ribosomal fidelity, or puromycin because of its competing ability with tRNAs for binding on the large ribosomal subunit, or cycloheximide which inhibit the large ribosomal subunit by preventing ribosomal movement along the mRNA, were used in active HIV-1 infection, it was found that both cycloheximide and puromycin produced the greatest decrease in HIV-1 inhibition, presumably by inhibiting the large subunit of the ribosome . Translation of HIV-1 RNAs pose a challenge since they all contain a TAR sequence at their 5' end. The Tat-responsive region (TAR) of HIV-1 exhibits a trans-inhibitory effect on translation by activating the interferon-induced 68-kilodalton protein kinase. Productive infection by HIV-1 has been shown to result in a significant decrease in the amount of cellular p68 kinase. The steady-state amount of p68 kinase was found to be reduced in cells stably expressing Tat. Thus, the potential translational inhibitory effects of the TAR RNA region, mediated by activation of p68 kinase, may be down-regulated by Tat during activation of the latent virus . Along these lines, a Tat peptide antagonist, which bound specifically to TAR RNA and competed with Tat for binding, reduced Tat-dependent translation . Finally, upregulation of translation genes in Tat expressing cells is specially intriguing in light of the recent discovery of internal ribosome entry sites (IRESs) in HIV-1 gag ORF . IRESs are thought to promote initiation of translation by directly binding to ribosomes, in a manner independent of the mRNA cap or of scanning through upstream sequences. Since, the TAR is located at the 5' end of all HIV-1 RNA transcripts and the presence of secondary structure at or near the 5' end of RNAs reduces the accessibility of the 5' cap to eIF4F, it is thought that this feature of HIV-1 mRNAs can inhibit their cap-dependent translation . Therefore, a possible function of the HIV-1 gag IRES might be to serve as a mechanism to bypass the structural barriers to cap-dependent translation by recruiting ribosomes easily and directly to the gag ORFs. IRES entirely contained within a translated ORF has been shown in the MMLV gag , and host mRNA encoding p110PITSLRE and p58PITSLRE. Along these lines, cap-dependent translation may be cell cycle regulated, especially when cells are arrested at the G2 phase of the cell cycle, where the cap-dependent translation of most cellular host cell mRNAs is inhibited . Modulation of signal transduction pathway | Results in Table indicate that many seemingly different pathways are being regulated by Tat. However, the signal transduction pathway, MAPK, has been shown to control and be upstream of DNA-replication, transcription, and cell cycle pathways . The mitogen-activated protein kinase (MAPK) pathway, consisting of the MAP kinase kinases (MKKs) 1 and 2, and extracellular signal-regulated kinases (ERKs) 1 and 2, which have been implicated in diverse cellular processes including proliferation, transformation, and cell differentiation . The MAP kinase (MAPK) pathway has emerged as a crucial route between membrane-bound Ras and the nucleus. This MAPK pathway encompasses a cascade of phosphorylation events involving three key kinases, namely Raf, MEK (MAP kinase kinase) and ERK (MAP kinase). The MAPK pathway controls ERKs 1 and 2, c-Jun N-terminal kinase (JNK), and p38. These signaling pathways in turn, activate a variety of transcription factors including NF-kappaB (p50/p65), AP-1 (c-Fos/c-Jun), and CREB phosphorylation, which in turn coordinate the induction of many genes encoding inflammatory mediators. Cytokine receptors such as IL-3, GM-CSF, and the interferons transmit their regulatory signals primarily by the receptor-associated Jak family of tyrosine kinases, and activate STAT transcription factors. Activated STAT5 proteins are detected in reduced levels in lymphocytes recovered from HIV-infected patients and immunocompromised mice. Both of these types of receptor signaling pathways have recently been shown to interact with serine/threonine kinases such as MAP kinases. A common intermediate pathway initiating from receptors to the nucleus is the Ras/Raf/MEK/ERK (MAPK) cascade, which can result in the phosphorylation and activation of additional downstream kinases and transcription factors such as p90Rsk, CREB, Elk, and Egr-1 . Therefore, it is intriguing that Tat expressing cells show down-regulation of MAPK components (Table , Figure and ), essential mediators between receptors and nuclear transcription factors. This would imply that latently infected cells that express Tat (doubly spliced RNA) and not the whole virus (all three classes of the RNA), can control signal transduction related to membrane and transcriptional signaling . Interestingly, Tat, through the RGD motif, which controls integrin-based cell signaling, has been reported to mediate the activity of phosphotyrosine phosphatase(s). This in turn which would lead to a decrease in the levels of phosphotyrosine-containing proteins such as ERK-2/p42MAPK kinases . Cysteine-rich and basic Tat peptides have been shown to inhibit VEGF-induced ERK activation and mitogenesis. These peptides also inhibited proliferation, angiogenesis, and ERK activation induced by basic fibroblast growth factor with similar potency and efficacy . Consistent with this model, it has been shown that treatment of neural cells with culture supernatants from HAART-treated subjects, which presumably contain extracellular Tat, resulted in down-regulation of the JNK, AKT, and ERK kinases . Finally, activation of MAPKs has been shown to activate the singly spliced and unspliced (genomic) latent HIV-1 virus. For instance, the signal transduction pathways that regulate the switch from latent to productive infection have been linked to MAPK. The induction of latent HIV-1 expression has been shown to be inhibited by PD98059 and U0126, specific inhibitors of MAPK activation. The MAPK acts by stimulating AP-1 and a subsequent physical and functional interaction of AP-1 with NF-kappaB, resulting in a complex that synergistically transactivates the HIV-1 . At the level of infection and entry, the activation of MAPK through the Ras/Raf/MEK (MAPK kinase) signaling pathway enhances the infectivity of HIV-1 virions. Virus infectivity can be enhanced by treatment of cells with MAPK stimulators, such as serum and phorbol myristate acetate, as well as by coexpression of constitutively activated Ras, Raf, or MEK in the absence of extracellular stimulation . Also, following infection, efficient disengagement of the reverse transcription complex from the cell membrane and subsequent nuclear translocation, requires phosphorylation of the reverse transcription complex components by ERK/MAPK; demonstrating a critical regulation of an early step in HIV-1 infection by the host cell MAPK signal transduction pathway . Therefore, Tat down-regulation of the MAPK pathway in latent cells implies that much of the host signal transductions connected to activation are down-regulated, and at the same time, these cells may be refractory to subsequent infection by other viruses. Thymosin family members, and cell cycle | Prothymosin alpha (ProTalpha) belongs to the alpha-Thymosin family which comprises different polypeptides widely distributed within animal tissues. Although its role has remained controversial, it is involved in the increase of immediate early genes such as c-myc , which is upstream of cyclin D synthesis and necessary for cell division . In humans, ProTalpha is coded by a gene family of six members. One of them contains introns, exons and classic regulatory signals, while the remaining five are intronless genes located on chromosome 2 . There are two mRNA transcripts, which arise in a ratio of 9:1 (shorter/longer form), where only the long transcript is regulated by extracellular signals. It has been demonstrated that malignant tissues with accelerated cell cycle show higher levels of ProTalpha expression than normal or surrounding healthy tissues . ProTalpha was shown as a marker for breast cancer , hepatocarcinoma , and plasma levels of its derivative Talpha1 been proposed as a marker for the prognosis of lung cancer . In ligand blotting assays, ProTalpha bound only to chromatin pools and nuclear fractions where histone H1 was present . The analysis of the interaction of ProTalpha with H1-containing chromatin suggests a putative role for ProTalpha in the fine-tuning of the stoichiometry and/or mode of interaction of H1 with chromatin . Interestingly, HL-60 cells overexpressing ProTalpha show an enhancement of accessibility of micrococcal nuclease to chromatin, implying relaxed chromatin structure for enhanced cell cycle gene expression . A broad study using several mononuclear and fibroblastic cell lines has shown that ProTalpha mRNA accumulation is cell cycle phase-dependent. In the U937 monocytic cell line, ProTalpha mRNA peaked at the end of S/G2 phase and fell towards the entry into the new G1 phase. More prominent mRNA regulation was found in the fibroblastic cell lines CV1 and NIH3T3, with peak mRNA levels at the end of S-phase. In all cases the expression pattern coincided with that of cyclin B and Cdc2/cyclin B activation . It is interesting to note that Cdc2 (Acc# X05360), Cdc10 homolog (Acc# S72008), and Cdc37 (Acc# U43077) were all up-regulated in Tat expressing cells. Cdc2, a catalytic subunit of cyclin-dependent kinases, is required for both the G1-to-S and G2-to-M transitions. In the fission yeast Schizosaccharomyces pombe, the execution of Start requires the activity of the Cdc2 protein kinase and the Cdc10/Sct1 transcription complex. The loss of any of these genes leads to G1 arrest . Cdc37 encodes a 50-kDa protein that targets intrinsically unstable oncoprotein kinases including Cdk4, Raf-1, and v-src to the molecular chaperone Hsp90, an interaction that is thought to be important for the establishment of signaling pathways. Cdc37 expression may not only be required to support proliferation in cells that are developmentally programmed to proliferate, but may also be required in cells that are inappropriately induced to initiate proliferation by oncogenes. For instance, MMTV-Cdc37 transgenic mice develop mammary gland tumors at a rate comparable to that observed previously in MMTV-cyclin D1 mice, indicating that Cdc37 can function as an oncogene in mice and suggests that the establishment of protein kinase pathways mediated by Cdc37-Hsp90 can be a rate-limiting event in transformation . Also, analysis of proteins that co-immunoprecipitated with Cdk6 and Cdk4 has shown complexes containing both Hsp90 and Cdc37 . Cdc37 also promotes the production of Cak1. Cak1 in yeast is the human homolog of CAK trimeric enzyme containing CDK7, cyclin H, and MAT1. Both human and yeast Caks function as RNA polymerase II CTD kinase, Cdk activating kinase, and DNA damage/repair enzymes. Cdc37, like its higher eukaryotic homologs, promotes the physical integrity of multiple protein kinases, perhaps by virtue of a cotranslational role in protein folding . Finally, Hsp90/Cdc37 has recently been shown in the stabilization/folding of Cdk9 as well as the assembly of an active Cdk9/cyclin T1 complex responsible for P-TEFb-mediated Tat transactivation . Transcription and chromatin remodeling factors | A highly ordered chromatin structure presents a physical obstacle for gene transcription; presumably by limiting the access of transcription factors and RNA polymerase II core machinery to target DNA . In concert with the observation that corepressors are associated with HDAC activities , it appears that the transcriptional outcome of nuclear receptors is determined by the balance of histone acetylation and deacetylation activities, and that ligands serve as a switch to recruit HATs with the concomitant dismissal of HDACs. Signal transduction pathways add another layer of regulation to the functions of CBP/p300. In the case of the POU homeodomain factor Pit-1, transcriptional activity is potentiated by MAPK pathways . Therefore, down-regulation of MAPK pathway members in Tat expressing cells, as seen in Table , is consistent with decreased phosphorylation of DNA binding factors such as Pit-1, and overall lower DNA binding activity. Here, we describe the effect of coactivator proteins SRC-1 (Acc# AJ000882, U90661, Table ) and p300 (Acc# U01877, Table ), and their relation to differentiation genes such as retinoic acid receptor (RAR/PML, Acc#: X06614, Table ), and Leptin receptor variant (Acc#: U66496, Table ); all of which are down-regulated in Tat expressing cells . Figure 4 | Synthesis of IL-8 in Tat expressing cells. Synthesis of IL-8 in Tat expressing cells. Hela cells (pCEP4, and eTat) were either unblocked (unt), or blocked with hydroxyurea (Hu) (2 mM) for 14 h, released, washed twice with phosphate-buffered saline (PBS) and subsequent addition of complete medium . Supernatants were collected at 9 hrs after release for ELISA. All remaining suspension cells were treated with 1% serum for 48 hrs prior to addition of Hu. PHA-activated PBMCs were kept in culture for 2 days prior to addition of Tat protein. Approximately 5 x 106 PBMCs were used for treatment with either wild type or K41A Tat mutant (100 ng/ml) proteins. After an initial incubation for one hr with Tat proteins, cells were washed and cultured in complete media for 24 hrs, prior to IL-8 ELISA. Figure 5 | Predictive model for control of gene expression and signal transduction by constitutive Tat expressing cells. Predictive model for control of gene expression and signal transduction by constitutive Tat expressing cells. Down-regulation of SWI/SNF components such as BAF 170 and 60 along with coactivators CBP/p300 and SRC-1 may down-regulate a subset of cellular genes that depend on chromatin remodeling and/or co-activator function for their gene expression. Such genes depend on the presence of ligand receptors that require either both SRC-1 and p300 or individual co-activator for their activity. Over the past three decades a great deal of evidence has accumulated in favor of the hypothesis that steroid receptor hormones act via regulation of gene expression. The action is mediated by specific nuclear receptor proteins, which belong to a superfamily of ligand-modulated transcription factors that regulate homeostasis, reproduction, development, and differentiation . This family includes receptors for steroid hormones, thyroid hormones, hormonal forms of vitamin A and D, peroxisomal activators, and ecdysone . Nuclear hormone receptors are ligand-dependent transcription factors that regulate genes critical to such biological processes as development, reproduction, and homeostasis. Interestingly, these receptors can function as molecular switches, alternating between states of transcriptional repression and activation, depending on the absence or presence of a cognate hormone, respectively. In the absence of cognate hormone, several nuclear receptors actively repress transcription of target genes via interactions with the nuclear receptor corepressors SMRT and NCoR. Upon binding of the hormone, these corepressors dissociate from the DNA-bound receptor, which subsequently recruits a nuclear receptor coactivator (NCoA) complex. Prominent among these coactivators is the SRC (steroid receptor coactivator) family, which consists of SRC-1, TIF2/GRIP1, and RAC3/ACTR/pCIP/AIB-1. These cofactors interact with nuclear receptors in a ligand-dependent manner and enhance transcriptional activation via histone acetylation/methylation and recruitment of additional cofactors such as CBP/p300 . CBP/p300 has been implicated in the functions of a large number of regulated transcription factors based primarily on physical interaction and the ability to potentiate transcription when overexpressed . In the case of nuclear receptors, the interaction with CBP/p300 is ligand-dependent and relies on the conserved nuclear receptor functional domain, AF-2 (activation function 2). In vivo studies have supported the conclusion that CBP/p300 are components of the hormonal-regulation of transcription in fibroblasts isolated from a p300-/- mouse; and loss of the p300 gene severely affects retinoic acid (RA)-dependent transcription . In a separate study using hammerhead ribozymes that specifically cleave CBP or p300 mRNA, Kawasaki et al reported that reduced cellular CBP or p300 levels resulted in compromised expression of endogenous RA-inducible genes such as p21/Waf1 and p27 cdk inhibitors. Along this line, Tat expressing cells have lower levels of p21/Waf1 presumably due to inactivation of p53 and a lack of p300/RA- induced gene expression. Consistent with this interpretation, CBP and p300 harbor transcriptional activation of ligand-induced RA or ER function on a chromatinized template . The NcoA family members constitute SRC-1/NcoA-1 , TIF2/GRIP1/NcoA-2, and pCIP/ACTR/AIB1 proteins, which interact with liganded RA receptor (RAR), and CBP/p300. Overexpression of these NCoA factors enhances ligand-induced transactivation of several nuclear receptors . A weak intrinsic HAT activity has been reported in SRC-1/NCoA-1 and pCIP/ACTR/AIB1, suggesting that chromatin remodeling may also be a function of these NCoA factors ; although they do not appear to contain regions homologous to the HAT domains of CBP/p300 or p/CAF. Structure-function analysis of the NCoAs have revealed multiple copies of a signature motif, LXXLL, with conserved spacing that is required for interaction with nuclear receptors and CBP/p300 . Intriguingly, different LXXLL motifs are required for PPARgamma (Peroxisome Proliferator activated receptor gamma, a gene down-regulated in Tat expressing cells; Acc# L07592, Table ) function in response to different classes of ligands, suggesting distinct configuration of assembled complexes. Taken together, through the use of microarray technology, we have described one of the first observations about how Tat is able to control various host cellular machineries. Although our data is consistent with most of the cited literature on the effects of Tat in infected host and uninfected bystander cells, we caution that the transcriptional profiling in chronically infected cells such as ACH2 or H9/Tat cells may not necessarily be representative of the pattern of expression observed in most cells infected by other group M, N, or O HIV-1 isolates. We recently extended our observations by utilizing other HIV-1 infected cells which normally express Tat (U1), and addition of exogenous purified Tat to uninfected PBMCs. Preliminary results using western blots supports the idea that genes which were altered in H9/Tat system also showed a similar level of change in few of the tested genes . This notion of consistency was further confirmed using the IL-8 activation by Tat. Interleukin-8 (IL-8) belongs to the CXC chemokine family and is secreted by several different cell types, including monocytes, neutrophils, endothelial cells, fibroblasts, and T lymphocytes. IL-8 production (induced by several stimuli, including IL-1, TNF-, and phorbol myristate acetate) is primarily regulated at the transcriptional level. IL-8 is a potent chemotactic factor for granulocytes and T lymphocytes, and is found in HIV-infected individuals. The CXC chemokine IL-8 does not bind to CCR5. It has previously been shown that IL-8 mRNA induction was seen less then 1 h after Tat (72aa) stimulation, and levels remained elevated for up to 24 h, leading to IL-8 protein production . Along these lines, we have previously shown that the IL-8 gene is expressed in a cell cycle-dependent manner in cells that express the Tat protein, and the induction is during the S phase of the cell cycle and regulated by stable NF-kB binding to the IL-8 promoter . When looking for IL-8 at the G1/S border, we found that all Tat containing cells, including PBMCs that were treated with exogenous Tat showed an up-regulation of IL-8 in the supernatant , further implying that results obtained from the H9/Tat system may infact be of general physiological relevance in vivo. Figure 6 | Proposed model for changes in signal transduction. Proposed model for changes in signal transduction. A) Down-regulation of receptor tyrosine kinases (RTK) by Tat which modulates the phosphorylation and transcription of downstream effectors such as Ras, Raf, MEK, MAPK, and control transcription factor phosphorylation. B) Role of Tat in the increase of genes necessary for proliferation, such as Cdc2, Cdc37, and Prothymosin alpha, and down-regulation of differentiation genes, such as receptors, co-receptors, and signal transduction genes. Finally, throughout the current study we came across some technical findings that were critical in the confirmation of most of our results. For instance, few genes did not correlate in their activation or suppression levels when comparing fold changes between microarrays and protein levels using western blot analysis. We suspect this is because many genes that are transcribed may not necessarily be translated, due to their cell cycle stage, 5' stem and loop RNA structures, varying half-lives of proteins and mRNAs, and a host of other unknown variables. Also, specific changes that occur in a cell may not be required in redundant pathways that score for a specific function. This is commonly seen in the differences between HIV-1 infected or Tat expressing in vitro cell lines and AIDS patients PBMC samples. Therefore, other microarrays would have to be performed on purified infected PBMCs to confirm most of the changes observed in Tables , , and . Unfortunately, to date this particular issue is not feasibly addressable, since it is not possible to isolate a homogenous population of infected T- or Monocytic cells from AIDS patients. Also, confirmatory tests for protein expression would have to be done with both hydrophilic and hydrophobic extraction buffers. For instance, we have observed that PCNA protein, which is up-regulated in Tat expressing cells, extract best with hydrophobic buffers from the nucleus, presumably due to its binding to DNA replication machinery (data not shown). Future experiments will address issues related to differences between various HIV-1 Tat clades, host expression levels between T- and Monocytic cells, and its effect at various stages of the cell cycle. Conclusions : Expression profiling from HIV-1 or Tat expressing cells holds great promise for rapid functional analysis. Here, we have described the effect of Tat and its alterations with the host cellular gene expression. We observed that more than 2/3 of the cellular genes tested were down-regulated by Tat. These genes belong to receptor, co-receptor, and co-activator pathways that are part of serine/threonine receptor tyrosine kinase, Ras/Raf/MEK/ERK (MAPK) cascade, which control proliferative and/or differentiation signals. We also observed a great deal of increase in the host cell translation apparatus with the possible aim of optimizing viral mRNA translation prior to viral maturation and release. Therefore, HIV-1 accessory doubly spliced messages such as Tat, may control the host gene expression in latently infected cells, and determine not only viral transcription, but also the fate of post-transcriptional events. Materials and method : Cell culture | ACH2 cells are HIV-1 infected CD4 lymphocytic cells, with an integrated wild-type single-copy chromatinized DNA. The CEM T cell (12D7) is the parental cell for ACH2 cells. ACH2 cell lines has a single copy of LAI strain proviral sequence. The TAR has a point mutation at (C37 -> T), which no longer responds (efficiently) to Tat. However, the cell line is fully capable of making infectious virus in presence of TNF, PHA, PMA, and a host of other stimuli. H9 and H9/Tat cells are both CD4+ Lymphocytic cells, where H9 cells carry a control integrated vector without the Tat open reading frame, and H9/Tat cells carry integrated Tat expression vector. Both cell lines were a generous gift of George Pavlakis (NCI, NIH). U1 is a monocytic clone harboring two copies of the viral genome from parental U973 cells. All cells were cultured at 37C up to 105 cells per ml in RPMI-1640 media, containing 10% Fetal Bovine Serum (FBS) treated with a mixture of 1% streptomycin and penicillin antibiotics, and 1% L-glutamine (Gibco/BRL). Phytohemagglutinin-activated PBMC were kept in culture for 2 days prior to addition of Tat protein. Isolation and treatment of PBMC were performed by following the guidelines of the Centers for Disease Control. Approximately 5 x 106 PBMC were used for treatment of wild type and K41A Tat mutant (100 ng/ml) proteins. After an initial incubation for one hr with Tat proteins, cells were washed and cultured in complete media for 24 hrs, prior to western blots. pCEP4, eTat cells were HeLa cells stably transfected with either a backbone control plasmid (pCEP4; Invitrogen) or a plasmid expressing Tat (1 --86) with a C-terminal epitope tag (eTat) . HeLa cell lines containing either the control or eTat plasmid were selected by single-cell dilution. Both cell types were selected and maintained under 200 mug of hygromycin per ml. Verification of Tat transcriptional activity was achieved by electroporation of reporter plasmids as previously described . Cell cycle analysis | Hela cells were blocked with hydroxyurea (Hu) (2 mM) for 14 h. Following the block, cells were released by being washed twice with phosphate-buffered saline (PBS) and by the addition of complete medium. All suspension cells were treated with 1% serum for 48 hrs prior to addition of Hu. Supernatants were collected and analyzed by an IL-8 ELISA according to the manufacturer's instructions (Biosource International). For controls, each sample, approximately 1 x 106 cells was processed for cell sorting. Cells were washed with PBS and fixed by addition of 500 mul of 70% ethanol. For fluorescence-activated cell sorting (FACS) analysis, cells were stained with a cocktail of propidium iodide (PI) buffer (PBS with Ca2+ and Mg2+, RNase A [10 mug/ml], NP-40 [0.1%], and PI [50 mug/ml]) followed by cell-sorting analysis. FACS data acquired were analyzed by ModFit LT software (Verity Software House, Inc.). Cell extract preparation and immunoblotting | All cells were cultured to mid-log phase of growth, washed with PBS without Ca2+ and Mg2+, and lysed in a buffer containing 50 mM Tris-HCl (pH 7.5), 120 mM NaCl, 5 mM EDTA, 50 mM NaF, 0.2 mM Na3VO4, 1 mM DTT, 0.5% NP-40 and protease inhibitors (Protease inhibitor cocktail tablets, Boehringer Mannheim, one tablet per 50 ml). The lysate was incubated on ice for 15 min, and microcentrifuged at 4C for 10 min. Total cellular protein was separated on 4 --20% Tris-glycine gels (Novex, Inc.) and transferred to a polvinylidene difluoride (PVDF) membranes (Immobilon-P transfer membranes; Millipore Corp.) overnight at 0.08 A. Following the transfer, blots were blocked with 5% non-fat dry milk in 50 ml of TNE 50 (100 mM Tris-Cl [pH 8.0], 50 mM NaCl, 1 mM EDTA) plus 0.1% NP-40. Membranes were probed with a 1:200 --1:1000 dilution of antibodies at 4C overnight, followed by three washes with TNE 50 plus 0.1% NP-40. All antibodies used in this study were purchased from Santa Cruz Biotechnology. The next day, blots were incubated with 10 ml of 125I-protein G (Amersham, 50 mul/10 ml solution) in TNE 50 plus 0.1% NP-40 for 2 hrs at 4C. Finally, blots were washed twice in TNE 50 plus 0.1% NP-40 and placed on a PhosphorImager cassette for further analysis. Total RNA purification | Cells were grown to mid-log phase of growth (5.0 x 106), pelleted, and washed twice with cold D-PBS without Ca2+/Mg2+. Total RNA was extracted on ice using Trizol Reagent (Life Technologies, Inc.). Purified RNA was then analyzed on a 1% agarose gel for quality and quantity prior to each experiment. Glass slide microarray | Gene expression analysis was performed using MicromaxTM: Human cDNA Microarray System I (cat# MPS101, NEN Life Science Products). On a glass microarray slide, 2400 know human genes were arrayed into 4 separate grids (A, B, C, D), containing 600 genes each (gene description and location on microarrays available at NEN website: ). All human genes were ~2200 bp cDNAs, and were characterized from 50+ human cDNA libraries (AlphaGene, Inc., Woburn, MA). In addition to the human genes, three plant control genes were spotted on each grid and were utilized to balance the Cyanine-3 (Cy-3) and Cyanine-5 (Cy-5) fluorescence signals. A total of 8 mug each of H9 (control sample) and H9/Tat (test sample) mRNAs were reverse transcribed into Biotin and Dinitrophenyl (DNP) labeled cDNA, respectively. After cDNA quality and quantities were analyzed, both cDNAs were then pooled and simultaneously hybridized overnight at 65C onto the glass microarray. The next day, the microarray slide was serially washed in 0.5x SSC (Sodium Citrate-Sodium Chloride) + 0.01% SDS (Sodium Dodecyl Sulfate), 0.06x SSC + 0.01% SDS, and 0.06x SSC. Next, the Tyramide Signal Amplification (TSATM) was then used to amplify the Cy-3 and Cy-5 signals using antibody-enzyme conjugates, alpha-DNP-Horseradish peroxidase (HRP) and alpha-Streptavidin-HRP with Tyramide linked Cy-3 and Cy-5. Screening and data analysis was performed by NEN. cDNA filter hybridization | Gene expression of CEM and ACH2 were performed using Atlas Human cDNA Expression Array (Clontech Laboratories Inc., Palo Alto, CA) according to the manufacturer's directions. One mug of poly A+ RNA each was DNase I treated, purified using a CHROMA SPIN-200 column, and reverse transcribed into 32P-labeled cDNA. The CHROMA SPIN-200 column was used to purify the 32P-labeled cDNA from unincorporated 32P-labeled dNTPs and small (<0.1 kb) cDNA fragments. Each sample was then hybridized to a human cDNA expression array overnight with continuous agitation at 68C. The next day, the array was washed three times with gentle agitation, first wash with 2x SSC + 1% SDS and the last two washes with 0.1x SSC + 0.5% SDS at 37C. Array was exposed to a PhosphorImager Cassette and analyzed using ImageQuant software. Northern blots | Total cellular RNA was extracted using the RNAzol reagent (Gibco/BRL). Total RNA (20 mug) was isolated from various cells and ran on a 1% formaldehyde-agarose gel overnight at 75 V, transferred onto a 0.2 mum nitrocellulose membrane (Millipore Inc.), UV cross-linked, and hybridized overnight at 42C with 32P-end-labeled 40 mer oligo probes including p21/Waf1, C-myc, Pro-thymosin, Actin, Tat, and Ubiquitin (Loftstrand, Gaithersburg, Md.). Next day, membranes were washed two times for 15 min each, with 10 ml of 0.2% SDS-2XSSC at 37C, exposed, and counted on PhosphorImager Cassette. Viral infection and ELISA assay | Both H9 and H9/Tat cells were infected in the presence of 10 ug of polybrene. For PBMC infections, PHA activated PBMCs were kept in culture for 2 days prior to each infection. Isolation and treatment of PBMCs were performed by following guidelines from the CDC (Isolation, culture, and identification of HIV, Procedural Guide, July 1991, Atlanta, GA). Approximately 2 x 10 6 of H9 or H9/Tat cells, and 5 x 10 6 PBMC cells were infected with either an HXB-2 (CXCR4, T-tropic), or BaL (CCR5, Macrophage-tropic) at 5 ng of p24 gag antigen/ HIV-1 strain. Both viral isolates were obtained from the NIH AIDS research and reference reagent program. After 8 hrs of infection, cells were washed and fresh media was added. Samples were collected every 3rd day and stored at -20 C for p24 gag ELISA. Media from HIV-1 infected cells were centrifuged to pellet the cells and supernatants were collected, and diluted to 1:100 to 1:1000 in RPMI 1640 prior to ELISA. The p24 gag antigen level was analyzed by HIVAGTM-1 Monoclonal antibody Kit (Abbott Laboratories, Diagnostics Division). Authors' Contributions : CF, and FS carried out the ACH2 and H9/Tat microarrays. LD, CE, IZ, CGL, and KW aided in westerns, northerns, p24 and ELISA assays. AM, KK, SB, AP, and FK aided in data interpretation, Bioinformatics, literature searches and writing the manuscript. Backmatter: PMID- 12052258 TI - Hospital competition, resource allocation and quality of care AB - Abstract | Background | A variety of approaches have been used to contain escalating hospital costs. One approach is intensifying price competition. The increase in price based competition, which changes the incentives hospitals face, coupled with the fact that consumers can more easily evaluate the quality of hotel services compared with the quality of clinical care, may lead hospitals to allocate more resources into hotel rather than clinical services. Methods | To test this hypothesis we studied hospitals in California in 1982 and 1989, comparing resource allocations prior to and following selective contracting, a period during which the focus of competition changed from quality to price. We estimated the relationship between clinical outcomes, measured as risk-adjusted-mortality rates, and resources. Results | In 1989, higher competition was associated with lower clinical expenditures levels compared with 1982. The trend was stronger for non-profit hospitals. Lower clinical resource use was associated with worse risk adjusted mortality outcomes. Conclusions | This study raises concerns that cost reductions may be associated with increased mortality. Keywords: Introduction : The last two decades brought about fundamental changes in the organization and delivery of medical services in the United States as payers seek to control the escalation in health care expenditures. Policies addressing these issues have been of two types. The first relies on containing costs through control of the prices paid to providers, beginning with the Prospective Payment System (PPS) for hospitals in 1983, Resource Based Relative Value Units (RBRVUs) payment for physicians in 1992 and the most recently implemented prospective payment for nursing homes. Such prospective payment systems provide the same incentives to cut costs to all providers, irrespective of the markets in which they are located and the competitiveness of their markets. The other relies on changing the focus of competition among health care providers, from quality based competition to price based competition . These policy changes were successful in bringing about a deceleration in hospital revenues and expenditures growth . Little is known, however, about what specific strategies hospitals adopted and the impact these strategies may have had on the quality of care patients receive. Previous studies found that hospitals increased efficiency in all clinical services following selective contracting in California. California hospitals also tended to specialize and differentiate themselves from similar hospitals in response to competitive pressures . A similar response, of increased specialization, was observed for a national sample of hospitals following implementation of PPS . In this paper we study another potential strategy that hospitals may adopt and which has not been addressed in the literature to-date. We investigate the hypothesis that hospitals in an increasingly price competitive environment, shift resources from activities related to clinical services, which are not easily observed and evaluated by patients, into hotel services which are easily observed. We study hospitals in California, comparing resource allocation during a regime dominated by quality competition and a regime dominated by price competition. We then examine the association between risk adjusted excess hospital mortality and resource use in clinical services, to investigate the potential impact on quality of medical care and health outcomes. Competition in California in the 1980s: a case study | The implementation of selective contracting in California in 1982 offers a unique natural experiment to study the response of hospitals to changes in the nature of competition. Unlike other health care markets, in which the change from quality to price based competition was gradual, driven by continuously increasing penetration of managed care, and often time confounded by other secular trends, the California legislation changed market conditions very rapidly for all hospitals in the state by permitting all health plans for the first time to contract with only a subset of hospitals. It thus allows a pre/post study design: hospitals resource allocation decisions during the quality competition regime (pre period) can be compared to decisions made during the price competition regime (post period). This natural experiment allows us to test the hypothesis that changes in the nature of competition are more likely to be associated with a shift of resources from clinical to hotel activities (and a concomitant deterioration in mortality outcomes) in more competitive hospital markets. As the level of hospital competition has not changed during the period (the Herfindahl-Hirschman Index (HHI) remained stable in all markets), this test is limited to the change in the nature of competition and is not confounded by the impact of changes in the level of competition on resource allocation decisions. During the same time period, California hospitals were also subject for the first time to price regulation, due to implementation of the Medicare PPS. While PPS also provided hospitals with incentives to lower their costs, it did so in a distinctively different manner than price competition. The PPS set the price per discharge hospitals were paid, thus creating incentives to lower costs, irrespective of market structure (7). The intensity of price based competition, on the other hand, is highly sensitive to the competitiveness of the hospital market. The analytical strategy of this paper is based on this distinction. Hospital competition, quality, resource allocation and health outcomes | Competition focused on prices, as is often the case in markets dominated by managed care, creates incentives to increase efficiency and possibly curtail resource use. With the exception of possible increases in administrative activities designed to contain costs in other areas (e.g. billing and utilization review) or to increase marketing efforts, such incentives to cut costs are likely to affect all aspects of hospital activities. Hospitals may also compete on quality, both quality of medical services and quality of hotel services and amenities. The importance of competition for quality is likely to be greater in markets in which hospitals compete for patients directly, as they do for all fee-for-service patients and for those enrolled in HMOs that offer a choice of hospitals within their market. Furthermore, to the extent that HMOs make their contracting decisions based on beneficiary hospital preferences, perceptions of quality are important competitive tools. Competition for quality , unlike competition for price , may lead to increased costs. Furthermore, it may affect clinical and hotel services differently. In markets where patients' choice of hospitals are increasingly important, hospitals are likely to compete more on quality attributes that patients observe and value. Given the difficulty that patients have in directly determining the quality of medical care they receive, and the relative ease with which they can evaluate the quality of hotel services (e.g. condition of the facility, quality of food) hospitals face incentives to shift resources from clinical activities to amenities. On the other hand, if patients rely on their physician's recommendations in choosing hospitals , and to the degree that physicians can assess clinical quality, albeit imperfectly, hospitals are faced with counter incentives, incentives that would promote resource use in clinical activities rather than hotel services. As a result, hospitals may face conflicting incentives: incentives to maintain or enhance the quality of hotel services on the one hand, and incentives to maintain activities that contribute to the quality of clinical care and health outcomes on the other. The actual choices that hospitals make about resource allocation depend on the relative strength of these opposing incentives. A model of changes in resource allocation | As the main hypothesis of interest is that the change in the nature of competition was associated with changes in hospital resource allocations, the model we hypothesize allows the marginal effects of market and hospital characteristics (Xt) on resource allocation (Yt), as measured by the coefficients in a regression model (betat), to vary over time: (1) Yt = alphat + betatXt + epsilont The models we estimated were difference models of the form (2) DeltaY = Deltaalpha + Deltabeta * X0 + betat * DeltaX + Deltaepsilon where Delta is the difference between year t and the base year, indicated by t = 0. From (2) it follows that betat, the vector of coefficients multiplying the change variables, measures the marginal effect of the variable in the end year, while Deltabeta, the vector of coefficients multiplying the level variables, measures the change in the marginal effect. The marginal relationship in year 0 is given by beta0 = betat - Deltabeta.. Methods : Sample | The initial sample included all 338 acute care hospitals in California that were in operation during both 1982 and 1989. Of those, 18 (5.3%) were excluded from the resource allocation analyses and 8 (2.4%) were excluded from the mortality analyses, because of incomplete data. Data sources | Financial, ownership and utilization data were obtained from the Hospital Annual Financial Disclosure Reports, filed annually by all California hospitals with the Office of Statewide Health Planning and Development (OSHPD). Risk adjusted mortality data were obtained from the Medicare Hospital Information Report published by the Health Care Financing Administration . Variable definitions | I. Resource allocation variables | Resource allocation was measured by expenditures per adjusted discharge. Adjusted discharges are a composite measure of input designed to account for both inpatient discharges and outpatient visits, using the methodology developed by the American Hospital Association. Expenditures per adjusted discharge were calculated separately for three categories: clinical, hotel and administrative services. Expenditures were aggregated by cost center, with each cost center assigned to one of the three services. The table in the lists all hospital cost centers and their assignment to hotel, clinical and administrative categories. The dependent variables in the resource allocation analyses were defined as the differences in expenditures per adjusted discharge, between 1989 and 1982, for each of the three categories. II. Quality variables | The dependent variables for the analyses of quality of clinical care were excess death rates from all causes and from 4 specific medical conditions that have relatively high death rates: acute myocardial infarction, congestive heart failure, pneumonia and stroke. We included in the analyses measures based on cause specific mortality in addition to overall mortality because prior studies have shown that these measures tend to be uncorrelated, and that hospitals performing well in one clinical area do not necessarily perform well in others. Measures based on overall mortality may therefore be biased towards zero, showing less variation compared with cause specific measures. Excess mortality was defined as the difference between the observed mortality rate for the hospital and a predicted, risk adjusted mortality rate. Observed and predicted mortality rates were obtained from the Medicare Hospital Reports . They are based on Medicare discharges and include all deaths within 30 days of admission, irrespective of the location of death. The risk adjustment methodology used by the Health Care Financing Administration, incorporates individual patients' age, gender, specific diagnoses and comorbidities, admission source, emergency or elective admission and the patient's risk group based on hospitalizations during the preceding 6 months . III. Independent variables | Competition was measured by the Herfindahl-Hirschman Index (HHI), defined as the sum of squared market shares of all hospitals competing in the same area. Hospital market areas and the HHI were calculated based on all payer zip code level patient flows, as described in Zwanziger et al. . To control for financial pressures hospitals may have been experiencing in addition to competition, we included variables measuring bad debt and charity as percent of total revenues and percent occupancy. To control for potential economies of scale the estimated models included total clinical standard units of measures reported in the California Financial Disclosure Reports. Ownership indicator variables included for-profit, not-for profit public, and not-for profit district ownership. The omitted category was private not-for profit hospitals. Teaching status was defined as hospitals with some residents. All payer DRG-based case mix index was included to account for differences in patients' severity. Median family income measured for the hospital's zip code area was included to capture demand effects and as a proxy for cross sectional wage variations. Analyses | We estimated regression models in which change in expenditures per adjusted discharge and excess mortality in 1989 were the dependent variables. Because all models were heteroskedastic, all reported tests of significance are based on White's robust standard errors . The Ramsey RESET test for specification errors was applied to all models to rule out the need for non-linear and interaction terms. The mortality models were weighted by the inverse of the standard error for the predicted mortality rate, to account for differences across hospitals in the accuracy of the excess mortality measures, which are due to differences in sample sizes . Since initial analyses indicated different associations (different betas) for for-profit and non-profit hospitals, we estimated fully interacted models, in which all variables were interacted with for-profit status. The hypotheses of significant marginal effect were therefore tested for the non-profit hospitals by a t test of the main effect and for the for-profit hospital by an F test of the linear restriction that the sum of the coefficients of the main and interaction effect are zero. Results : Description of sample hospitals | Table presents descriptive statistics for the hospitals included in the study. The majority of hospitals (52.6%) were private non-profit with the second largest group (26.0%) being for-profit institutions. Fifteen point four percent were teaching hospitals. The average hospital size did not change significantly over the 1982 --1989 period, remaining at 190 --200 beds. Occupancies declined significantly, from an average of 62.3% to 55.2%, and inpatient case mix increased significantly from 1.17 to 1.27, indicating that hospitals were treating sicker and more expensive patients at the end of the period. Both total expenditures and expenditures per adjusted discharge increased significantly. Table 1 | Descriptive Statistics The degree of competition among hospitals has not changed between 1982 and 1989. The HHI of around 0.3 suggests that competition was limited. (Markets with HHI values below 0.18 are considered moderately or very competitive ). The large variation in the HHI, however, indicates that many hospitals were located in competitive markets, with 25% of hospitals in markets with HHI below 0.17. Overall mortality rates averaged 10%. Average observed and predicted rates were very similar, but the variation in rates was higher for the observed rates compared with the predicted rates, suggesting substantial variations in excess mortality and quality across the sample. Resource allocation changes | Table reports the mean values for the dependent and independent variables included in the multivariate regressions. Table reports results by ownership -- for-profit and non-profit. These results are based on a fully interacted model estimated over pooled data by ownership. All models were highly significant (p < 0.01). The clinical services model explained 51% of the variation in expenditures per adjusted discharge, while the hotel and administrative services models explained 24% and 26% respectively. Table 2 | Means and standard deviations of variables included in the multivariate analyses Table 3 | Expenditures per adjusted Discharge -- Multivariate regression results Effect of competition | Table presents the marginal effect of competition, calculated from the regression results and using equations 2 and 3, for 1982 and 1989, as well as the change in these coefficients between the two years. (Note that the regression coefficients for the HHI variable were multiplied by -1 in table , such that a positive association means that expenditures per discharge increase with increased competition.) Table 5 | Marginal effect of competition on Expenditures per adjusted discharge (Positive values indicate increase with increased competition) Non-profit hospitals in more competitive areas had higher expenditures per adjusted discharge in all three categories in all years. The marginal effect was highest in clinical areas and lowest in administrative services. It declined significantly over time in both clinical and hotel services, but not in administrative services. The decline was almost three times as large in the clinical services compared with hotel services. By 1989, while the marginal effect of competition on expenditures in these services was still positive, it was no longer significantly different from zero. These findings are consistent with the hypothesis that the focus of competition on quality in 1982 has diminished significantly over the seven-year period we studied. The results for for-profit hospitals present a different picture. First, the association between expenditures per adjusted discharge and competition was statically significant only for administrative services. This association was also by far the strongest. It was negative, indicating that hospitals in more competitive markets spent less per discharge on administrative activities. The marginal effect was slightly smaller in 1989, suggesting that hospitals in more competitive areas may have reallocated resources into administrative services. They may have, for example, invested in better information and management systems that would allow them to better control costs. The association between competition and resource use was negative in clinical services and positive in hotel services. This is consistent with the hypothesis that for-profit hospitals compete on quality in those areas that can be easily observed by patients, namely hotel services, and cut back on resources in clinical services, where quality is more difficult for patients to evaluate directly. The change over the 1982 through 1989 period is also consistent with this hypothesis: the negative association between competition and clinical resources increased in 1989 as did the positive association between hotel resources and competition. The lack of significance of the associations may reflect the smaller number of for-profit hospitals in the sample and the resulting lower statistical power. (There were 83 for-profit hospitals and 13 independent variables, compared with 237 non-profit hospitals.) Other hospital and market characteristics | The strongest and most consistent relationship was between all payer case mix and expenditures per adjusted discharge in all categories (see table ). The association was substantially stronger for the clinical category and in non-profit compared with for-profit hospitals. Most other variables either exhibited no significant associations or no clear patterns. There were no significant differences in resource allocations by ownership. Teaching status was positively associated with clinical and hotel expenditures among the non-profit hospitals but exhibited a negative association among the for-profit hospitals. Percent bad debt and charity and percent occupancy had no significant relationship with expenditures among the non-profit but were associated with lower clinical and administrative expenditures among the for-profit hospitals. Median family income was associated with higher expenditure levels for all services among the non-profit hospital, but only with the clinical services among the for-profits. Risk adjusted excess mortality | Table reports the results of the regressions modeling the association between excess mortality (defined as the difference between observed and predicted mortality rate) and clinical expenditures per adjusted discharge, competition, ownership, and teaching status. The models explained between 1% and 5% of the variation in excess mortality. In all cases there was a negative association between clinical expenditure levels and excess mortality, implying that increased resources were associated with better mortality outcomes. This relationship was present for mortality from all causes as well as from the four specific causes, and was statistically significant at the 0.10 level or better. Table 4 | Risk adjusted excess mortality -- Multivariate Regression Results Table shows the increase in excess mortality that is associated with a decrease of 1 standard deviation (SD) in clinical expenditures per adjusted discharge, based on the estimated regression coefficients. For comparison, the table also provides the magnitude of 1 SD in excess mortality among the study hospitals. In all cases, the effect of a 1 SD in resources was less than a 1 SD in excess mortality. As expected due to the potential bias towards zero in the measures based on all causes, the associations were larger for the cause specific measures. Table 6 | Increase in excess mortality rates associated with 1 standard deviation decrease in clinical expenditures per adjusted discharge As discussed earlier, we hypothesized that competition may affect quality, including clinical quality, not only through its impact on resource use but also due to incentives to compete on quality. If indeed hospitals were competing on clinical quality, the association between the HHI and excess mortality, controlling for resource use, should have been positive and significant. In all cases, except for pneumonia, we did not find a significant relationship. In most cases there was also no significant relationship between ownership or teaching status and excess mortality. Discussion : In this paper we present a test of the hypothesis that changes in the nature of competition among California hospitals, resulting from selective contracting, were associated with changes in hospitals' resource allocation decisions. We find empirical evidence to suggest that resources have been shifted from clinical activities (which are not observed by patients) and into hotel services (which are more readily observable). These changes in resource allocation tended to be larger in hospitals located in more competitive areas. As the level of competition has not changed during the study period, the change in hospital behavior is likely to be a response to the change in the nature of competition. The relationship between resource allocation and competition differed by ownership. For-profit hospitals in more competitive areas had lower expenditures levels compared with those in less competitive areas. Among non-profit hospitals, we found the opposite -- clinical expenditures per adjusted discharge increased with competition. The trend over time, however, even though it was much stronger among the non-profit hospitals, was the same for both types. For both, the change in the marginal effect of competition (Deltabeta) was negative. As a result, the positive association between competition and clinical resource use among non-profit hospitals diminished and the negative association among the for-profit hospitals increased. The analyses of excess mortality demonstrate that clinical quality, at least as measured here, is positively associated with the amount of resources used in producing clinical services. Therefore, policies that create incentives for hospitals to limit resource use are likely to have an impact on health outcomes. Furthermore, as clinical quality is not easily observable by consumers, leading to disparity in incentives to provide hotel and clinical quality, more competitive areas are likely to experience a larger relative decline in resources allocated to clinical activities and hence worse mortality outcomes. The impact on health outcomes in non-profit hospitals may not be as large as it might have been because expenditure levels were curtailed not only in clinical areas, but in hotel services as well, although to a much lesser degree. This strategy spread the burden of cost containment efforts beyond clinical activities. If non-profit hospitals would have concentrated all their cost cutting efforts in clinical services, as did the for-profit hospitals, the impact on costs, and potentially on mortality outcomes, would have been 35% higher (see table ). The generalizability of the findings presented here is limited in several ways. First, quality was measured only in terms of excess mortality. While this is an important aspect of quality, it is likely to be an insensitive measure. Because measures based on mortality do not tend to be correlated with measures based on other outcomes , such as complications, one cannot deduce from this study that other aspects of clinical quality have been affected similarly by the changes in competition during the period. Furthermore, the mortality models we estimated, unlike the expenditures models, were cross sectional and are therefore subject to the usual concerns about potential bias due to omitted hospital specific effects. It should be noted, however, that the models did include the variables most likely to be associated with excess mortality -- patient level risks, expenditures, competition, ownership and teaching status. It should also be noted that while our findings with respect to changes in resource allocation are based on total expenditures, thus reflecting care for all patients, the mortality outcomes are based on the experience of Medicare patients only. Prior studies, however, suggest that patient care given by same provider does not vary by payer status . A more important generalizability question arises due to the sample selected for this study -- namely hospitals located only in California during the 1980s. Can the behavior of hospitals observed in this local and during this time period be assumed to generalize to other markets and other times? While the magnitude of the effects we measure in this study are clearly not generalizable, the directions of the effects are likely to transcend time and place, as they reflect basic behavioral responses to market incentives in accordance with theory. While this sample selection may be viewed as a limitation of this study, it is also its strength. By focusing on a period in which levels of competition were stable and where the change in the nature of competition can be traced to a specific legislative act, this study is able to test for the differential impact of price vs. quality competition, without confounding by other factors. Conclusion : In summary, the results of this study should be viewed as raising a cautionary question: are the hospital cost reductions that have been observed in California and nationally associated with increased mortality? Authors' Contributions : DM designed the study, performed the analyses, and wrote the manuscript. AB prepared the data set, and JZ participated in the design and analyses. Competing interests : None declared. Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12079499 TI - Functional group interactions of a 5-HT3R antagonist AB - Abstract | Background | Lerisetron, a competitive serotonin type 3 receptor (5-HT3R) antagonist, contains five functional groups capable of interacting with amino acids in the 5-HT3R binding site. Site directed mutagenesis studies of the 5-HT3AR have revealed several amino acids that are thought to form part of the binding domain of this receptor. The specific functional groups on the ligand that interact with these amino acids are, however, unknown. Using synthetic analogs of lerisetron as molecular probes in combination with site directed mutagenesis, we have identified some of these interactions and have proposed a model of the lerisetron binding site. Results | Two analogs of lerisetron were synthesized to probe 5-HT3R functional group interactions with this compound. Analog 1 lacks the N1 benzyl group of lerisetron and analog 2 contains oxygen in place of the distal piperazine nitrogen. Both analogs show significantly decreased binding affinity to wildtype 5-HT3ASRs. Mutations at W89, R91, Y142 and Y152 produced significant decreases in binding compared to wildtype receptors. Binding affinities of analogs 1 and 2 were altered only by mutations at W89, and Y152. Conclusions | Based on the data obtained for lerisetron and analogs 1 and 2, we have proposed a tentative model of the lerisetron binding pocket of the 5-HT3ASR. According to this model, The N-benzyl group interacts in a weak interaction with R91 while the benzimidazole group interacts with W89. Our data support an interaction of the distal amino nitrogen with Y142 and Y152. Keywords: Background : The cysteine-loop family of ligand gated ion channels (LGIC) is comprised of receptors with pentameric quaternary structure and at least two ligand binding sites present at the subunit interfaces . This receptor family is characterized by the presence of a critical disulfide loop structure within the binding site and an integral ion selective channel. LGIC receptors are found in both the peripheral and central nervous systems. Members of this family include the acetylcholine receptors , the gamma-amino butyric acid type A receptor (GABAAR) , and the glycine receptor (GlyR) . The first subunit of the 5-HT3R was cloned in 1991 . The sequence of this subunit was shown to be highly homologous to LGIC receptors and thus identified the 5-HT3R receptor as another member of this superfamily . Similar to other LGIC receptors, more than one subtype has been identified. Two splice variants of an A subunit (long and short forms), and a single B subunit have been cloned . Both the long and short forms of the A subunit are capable of forming functional homomeric receptors [5-HT3ALR and 5-HT3ASR] although some differences between an agonist and partial agonist activity have been observed . A third subtype is formed by a combination of the A and B subunits to produce a heteromeric receptor of unknown stoichiometry . Heteromeric receptors are pharmacologically and functionally distinct from the homomeric 5-HT3AL and 5-HT3AS receptors . 5-HT3Rs are distributed throughout the central and peripheral nervous system, playing a significant role in phenomenon such as anxiety, emesis and alcoholism. Antagonists to 5-HT3Rs are clinically efficacious in the treatment of chemotherapy-induced emesis and recent studies on human subjects have suggested their potential application in the treatment of early onset alcoholism . Hibert et al proposed an early model for the antagonist pharmacophore of the 5-HT3R . According to this model, all 5-HT3R antagonists contain an aromatic ring, a carbonyl oxygen or bioisosteric equivalent, and a basic nitrogen. According to Hibert's model, the basic nitrogen is located 5.2A from the centre of the aromatic ring and approximately 1.7A above plane of the ring. The carbonyl oxygen and the aromatic ring are coplanar and separated by a distance of 3.3A. Recent studies have expanded on this model to include another lipophilic region and a second hydrogen bonding interaction two atoms away from the first . A compound that contains all five pharmacophoric regions was synthesized by Orjales et al. This compound (1-(phenylmethyl)-2-piperizinyl benzimidazole or lerisetron) is shown in figure and is a potent 5-HT3R antagonist. Functional groups on this compound capable of forming interactions with the receptor are the distal amino group, a benzimidazole and a benzyl group in the N1 position of the benzimidazole. While Lerisetron contains no carbonyl group, the second nitrogen contained in the benzimidazole heterocycle could act as bioisostere of this functional group . Orjales demonstrated the importance of the N-benzyl group by synthesizing several N1 substituted analogs of Lerisetron. Removal of the N-benzyl group produced a 80-fold decrease in affinity, indicating a role for this group in interacting with the 5-HT3R. Other studies have supported this observation and suggest a more specific electrostatic interaction . Figure 1 | 5-HT3R antagonists. 5-HT3R antagonists. Granisetron and Lerisetron are shown along with the two analogs utilized to probe the functional group interactions of binding site amino acids. Analog 1 lacks the N-benzyl functional group. Analog 2 contains oxygen in place of the distal piperazine nitrogen. While structure-activity relationship studies and molecular modeling have led to the development of a detailed pharmacophore model, determining specific point interactions between 5-HT3 antagonists and binding site amino acids has proven difficult. Mutagenesis studies have identified the interaction of amino acids W89 and R91 in the binding of 5-HT3R ligands . Studies conducted in our laboratory have identified three additional putative binding site residues (Y140, Y142, and Y152) . W89 and R91 are present in a conserved region of LGIC receptors often referred to as loop D . Similarly, Y140, Y142 and Y152 are located in the region homologus to E loop region of nicotinic AchR. In this study, we have endeavoured to identify the amino acids interacting with the different functional groups present on the lerisetron molecule in order to develop a model for interaction of this compound with the 5-HT3R. Using site directed mutagenesis in combination with analogs of lerisetron, we have identified amino acids that appear to interact selectively with the terminal amino group, the N-benzyl group and the aromatic benzimidazole. Results : Functional activity of lerisetron | Whole cell patch-clamp experiments were performed to test the functional activity of lerisetron. No response was observed when lerisetron was applied alone (data not shown). When co-applied with 5-HT, lerisetron inhibited the absolute magnitude of the response with no apparent alteration of the response profile . The combination of several identical inhibition experiments produced a Ki value of 0.2 +- 0.03 nM for lerisetron inhibition of the 5-HT induced response. These data correspond well with previously reported data for this compound and verify the competitive antagonist action of lerisetron. Analogs of lerisetron have been shown to inhibit 5-HT3Rs in a similar manner . Figure 2 | Inhibition of 5-HT induced responses by Lerisetron. Inhibition of 5-HT induced responses by Lerisetron. tsA 201 cells transfected with 5-HT3ASR cDNA were clamped at a holding potential of -60 mV and perfused with 4 muM 5-HT (lower trace) or 4 muM 5-HT and 0.3 nM lerisetron (upper trace). Lerisetron reduced the peak 5-HT induced currents from 2.0 nA to 1.0 nA with no change in the response characteristics. An IC50 value was determined by a combination of similar experiments and determined to be 0.4 +- 0.03 nM. Importance of the N-benzyl and distal piperazine nitrogen to binding of lerisetron | The Ki value for lerisetron inhibition of [3H]-granisetron binding to wildtype receptors was determined to be 0.8 +- 0.19 nM (Figure and Table ). This value agrees with previously published data for this compound. The Ki values for analogs 1 and 2 under identical conditions are 25 +- 3.2 nM and 320 +- 82 nM respectively (Figure , Table ). The observed change in Ki represents the decreases in binding energy resulting from removal of the N-benzyl group (analog 1) and the distal piperazine nitrogen (analog 2). Figure 3 | Inhibition of [3H]-granisetron binding to mutant and wildtype receptors by lerisetron. Inhibition of [3H]-granisetron binding to mutant and wildtype receptors by lerisetron. Competition binding assays were performed as described in Materials and Methods. Each point represents a mean +- SE of n>4 experiments. Ki values calculated from this data and the relative magnitude of the shift in Ki compared to wildtype receptors are shown in Table . Y140A produced no significant change in Ki values. W89F, W89Y and R91A mutations produced moderate increases in Ki. Y142A and Y152A mutations resulted in increases in Ki of 160 and 190 fold respectively. Figure 4 | Inhibition of [3H]-granisetron binding to wildtype and mutant receptors by analog 1 and analog 2. Inhibition of [3H]-granisetron binding to wildtype and mutant receptors by analog 1 and analog 2. Competition binding assays were performed as described in Materials and Methods. Each point represents a mean +- SE of n> 4 experiments. IC50 values were determined from the data and a Ki value calculated using the Cheng-Prusoff equation. Ki values for inhibition on each mutant are shown in Table for all three compounds. A. Wildtype B. W89F C. W89Y D. R91A E. Y142A F. Y152A Table 1 | Summary of inhibition data obtained for lerisetron, analog 1 and analog 2 on mutant and wildtype receptors. Ki values were calculated from IC50 values as previously described. All data is the result of n > 4 experiments and is expressed as the mean +- SE. The relative change in Ki was calculated as the Ki (mutant) / Ki (wildtype) and indicates the increase in Ki relative to wildtype +- SE. * indicates significant difference in fold change value. Kd values for [3H]granisetron binding to wt and mutant receptors were reproduced from reference 25. Identification of amino acids interacting with Lerisetron | In order to determine the nature of the amino acids interacting with the distal amino and N-benzyl groups of lerisetron, we constructed 5-HT3ASRs containing mutations at W89, R91, Y140, Y142 and Y152. Figure shows inhibition of [3H]-granisetron binding by lerisetron on wildtype and mutant receptors. For most amino acids, an alanine substitution was constructed in order to effectively remove any amino acid interaction with the ligand. For W89, an alanine substitution has been shown to prevent binding of [3H]granisetron; therefore a less severe mutation was constructed. The W89F mutation produces a 18-fold change in Kd for [3H]-granisetron binding (18 +- 2 nM) and the W89Y mutation produces a 5.8-fold change in Kd (5.7 +- 0.7 nM). Mutation of amino acid R91 to alanine produced a 5-fold change in Kd for [3H]granisetron binding (4.9 +- 0.7 nM) . These data agree well with previously reported values . Alanine mutations at the tyrosine positions Y140, Y142 and Y152 also produced minor increases in Kd for [3H]granisetron binding (2.7 +- 0.19 nM, 4.5 +- 0.5 nM and 7.8 +- 1.1 nM respectively) . Only small changes in Ki for lerisetron were observed for the Y140A mutation while the Y142A and Y152A mutations produced large increases in the Ki (Table and Figure ). For W89F and W89Y, the changes in Ki observed for lerisetron were much smaller than for the alanine mutations at Y142A and Y152A, as would be expected for the less severe nature of these mutations. The changes were, however significant (p < 0.001 in both cases) and are similar to the changes in Kd reported for [3H]granisetron. The increase in Ki on the W89F mutant receptor was 4.8 +- 0.56 fold and the increase in Ki on the W89Y receptor was 3.6 +- 0.4 fold. The R91A mutant produced an increase in Ki of 7.6 +- 1.5 fold as compared to the wildtype receptor. These data indicated potential interactions of lerisetron with W89, R91, Y142 and Y152. Mutation of W89 | As mentioned above, the lack of [3H]-granisetron binding to W89A mutant receptors necessitated the use of W89F and W89Y mutations to analyze functional group interactions. The effects of these mutations on the Ki for analogs 1 and 2 are shown in figure , and Table . Analog 1 inhibited [3H]-granisetron binding to W89F receptors with a Ki of 170 +- 54 nM (7 +- 3.2 fold increase, p < 0.001) and W89Y receptors with a Ki of 81 +- 14 nM (3.2 +- 0.6 fold increase, p < 0.001). This reflects a significant increase in Ki and reflects a potential interaction of analog 1 with W89. The strength of this interaction is apparently similar to the strength of the interaction with [3H]-granisetron and lerisetron since the magnitude of the change is similar in both cases. Analog 2 also showed a significant increase in Ki as a result of the W89F and W89Y mutations. The magnitude of the change for W89F (5.1 +- 1.3 fold, p < 0.05) was similar to that observed for lerisetron and analog 1. The W89Y mutation produced a 6.8 +- 1.6 fold change in Ki (p < 0.05). Thus, all three compounds appear to form similar interactions with W89. Mutation of R91 | Mutation of R91 to alanine (R91A) resulted in a significant, but small increase in Ki for lerisetron of 7.6 +- 1.5 fold (p < 0.01). Figure shows the inhibition of [3H]-granisetron binding by analogs 1 and 2 at R91A mutant receptors. No significant change in Ki was observed on these receptors for either analog 1 (0.9 +- 0.28 fold) or analog 2 (0.56 +- 0.14 fold) as compared to the wildtype receptor. Mutation of Y142 | Mutation of Y142 to alanine produced one of the largest observed changes in Ki for lerisetron (Figure and Table ). The Ki obtained for lerisetron was 130 +- 28 nM, reflecting a change of 160 +- 37 fold compared to wildtype receptors. The Ki value for analog 1, in contrast, increased only 6.8 +- 2.3 fold (p < 0.01) as a result of this mutation (Figure and Table ). The Ki for analog 2 showed a similar change of 17 +- 0.77 fold (p < 0.01). While these Ki values are significantly different from wildtype values for each analog, the lack of larger effects suggests that neither analog 1 nor analog 2 bind as strongly as lerisetron to Y142. Mutation of Y152 | The Y152A mutation showed the most variability in its effects on Ki values for lerisetron, analog 1 and analog 2 (Figure , Figure and Table ). Lerisetron inhibited [3H]-granisetron binding with a Ki value of 150 +- 36; an increase of 190 +- 43 fold compared to wildtype values. The Ki value for analog 1 increased from 25 +- 3.2 nM (wildtype) to 2.5 +- 0.40 muM. This change of 100 +- 16 fold is slightly smaller, but not significantly different from the relative change observed for lerisetron. The Ki for analog 2 increased from 0.32 +- 0.08 muM on wildtype to 13 +- 4.2 muM on Y152A mutant receptors (40 +- 12 fold increase). The increase observed for analog 2 was significantly less than that observed for both lerisetron and analog 1. The smaller change in Ki for analog 2 suggests that analog 2 binds weakly to Y152 while lerisetron and analog 1 bind more tightly. Discussion : The purpose of this study was to ascertain the functional group interactions of lerisetron with specific amino acids present in the 5-HT3R binding site. To this end, we examined the effects of specific amino acid mutations on the binding of lerisetron and two analogs. These analogs, shown in Figure , are identical to lerisetron in structure; however, each analog lacks a single functional group found in lerisetron. By comparing the effects of a specific amino acid mutation on the Ki for inhibition of [3H]-granisetron binding, we can begin to identify the functional groups likely to participate in that interaction. Five different amino acids were tested: W89, R91, Y140, Y142 and Y152. Each of these amino acids has been shown in previous studies to alter the binding of 5-HT3R ligands . A summary of the changes in Ki is presented as a bar chart in Figure and . Figure 5 | Relative increases in Ki for lerisetron, analog 1 and analog 2 on mutant and wildtype receptors. Relative increases in Ki for lerisetron, analog 1 and analog 2 on mutant and wildtype receptors. The relative change in Ki was determined as Ki (mutant)/ Ki (wildtype) for each compound. A. W89Y, W89F, R91A B. Y140A, Y142A, Y152A. Lerisetron showed a significant increase in Ki for all mutants except Y140A. Increases in Ki were observed for analog 1 and 2 on W89 and Y152A. While lerisetron binds to wildtype receptors with high affinity, analog 1 binds with lower affinity (31 fold increase in Ki compared to lerisetron). Analog 2, which lacks the terminal nitrogen of lerisetron, binds with even lower affinity (a 400-fold fold increase in Ki compared to lerisetron -- Table ). These data demonstrate the relative importance of the N-benzyl and the terminal nitrogen moieties for lerisetron binding. Cation-pi interactions involving aromatic residues and a quaternary ammonium have been demonstrated to play a major role in binding of nicotinic ligands to nAchRs . Our data support a similar role for Y142 and Y152 in binding of lerisetron to the 5-HT3A receptor. Functional group interactions of W89 | The W89F mutation produced a significant increase in Ki for all three compounds. The magnitude of the change was similar in all cases. In addition, the increase in Ki was identical to the increase in Kd that has been observed for [3H]granisetron binding on this mutant . Alterations in Ki resulting from the W89Y mutation were slightly less, however the change was again the same for lerisetron, analog 1 and analog 2. These data suggest that all three compounds form binding site interactions with W89. The interaction between lerisetron and W89 is unlikely to be via the N-benzyl functional group since the Ki for analog 1 was also altered by this mutation. The same argument can be made for the distal piperazine nitrogen since the Ki for analog 2 also increased. The portion of the molecule common to all three compounds, the aromatic benzimidazole, is thus the most likely point of interaction for W89. Functional group interactions of R91 | The R91A mutation increased the Ki value for lerisetron inhibition of [3H]granisetron binding by 7.6 fold. This is a moderately small change for an alanine mutation, particularly considering that the smallest change in Ki for removal of a functional group on lerisetron (the N-benzyl group) was 31 fold. It is therefore likely that this interaction is either extremely weak or the change in Ki is the result of a structural change in the binding site. Previous studies concluded that R91 was an important interaction for the 5-HT3R agonist 5-hydroxytryptamine (5-HT), since the Ki for 5-HT inhibition increased over 3000 fold as a result of the R91A mutation . A change in Kd for [3H]granisetron binding to R91A was also observed. In order to determine whether the N-benzyl or distal piperazine nitrogen of lerisetron was involved in an interaction with R91, we tested both analog 1 and 2 on R91A mutant receptors. No change in Ki was observed for either compound. This result makes it much more difficult to assign the correct functional group to this amino acid since it suggests that one or both of the compounds is no longer binding the receptor in precisely the same manner as lerisetron. Considering the small change observed for lerisetron binding as a result of this mutation, even a slight reorientation of the molecule in the binding site could result in the loss of this interaction. Functional group interactions of Y142 | The Ki values for inhibition of [3H]granisetron binding by analogs 1 and 2 were altered only slightly by the Y142A mutation. The magnitude of the increase in Ki for lerisetron, however, was considerably larger (160 fold) and is indicative of an important interaction of the compound with Y142. The lack of a large change in Ki for both analogs makes it difficult to interpret this data since one or both of the compounds appears to be interacting differently with the binding site than lerisetron. Analogs 1 and 2 differ from each other both in the functional groups contained in the molecule and their structural similarity to lerisetron. Analog 2 is most similar in overall structure. The substitution of oxygen for the distal amino nitrogen alters the potential interactions formed at this position, but is likely to have a small effect on the overall size and shape of the molecule. Analog 1 is far less similar to lerisetron and more similar to the 5-HT3R antagonist granisetron. Previous studies have shown that the binding of granisetron is not affected by the Y142A mutation . Analog 1 may bind more similar to granisetron than lerisetron and thus would be unaffected by mutations at Y142. This is less likely to be the case with analog 2. The strength of the putative interaction at Y142 can be identified by examining the change in binding of lerisetron as a result of the Y142A mutation. The Y142A mutation produced a 160 fold change in Ki. This change reflects the binding energy lost as a result of the alanine substitution. The observed change in Ki on wt receptors is much larger than that observed for removal of the N-benzyl group (31 fold), but is similar to that observed for substitution of the distal amino nitrogen in analog 2 (400 fold). Taken together with the close structural similarity of analog 2 to lerisetron, it can be concluded that comparison of analog 2 and lerisetron should provide the best means of identifying the interaction at Y142. No change in Ki was observed for analog 2 as a result of the Y142A mutation indicating a lack of any significant interaction of this compound with Y142. These data support our hypothesis that Y142 interacts with the distal piperazine nitrogen of lerisetron. A second amino acid may also be involved since the change in Ki for lerisetron binding as a result of the Y142A mutation was smaller than the change produced by substitution of the piperazine nitrogen. As described below, one candidate for this second amino acid is Y152. Functional group interactions of Y152 | The Y152A mutation produced increases in Ki for all three compounds although the magnitude of the change differed. The increases in the Ki values were 190 fold for lerisetron, 98 fold for analog 1 and only 40 fold for analog 2. Thus, analog 1 retains much of its ability to interact with Y152 despite the absence of the N-benzyl group, while analog 2 interacts more weakly with this amino acid. Since the Ki for analog 1 is increased by the Y152A mutation, it is unlikely that the N-benzyl group interacts with Y152. The small change in Ki for analog 2 supports a partial interaction of Y152 with the distal piperazine nitrogen although some interaction with another group is also apparent. This other group would be expected to be in close proximity to the distal nitrogen. The most likely candidate is the other nitrogen of the piperazine ring. Thus Y152A may form a partial interaction with both piperazine nitrogens. Conclusions : Figure shows a hypothetical model of the lerisetron-binding site supported by our observations. The model illustrates the secondary structure of the region of the receptor from Y140 -- Y152 in a loop configuration. This structure is supported by site-directed mutagenesis data as well as structural predictions obtained from other LGIC receptors . The recent determination of the structure of a nicotinic acetylcholine binding protein that shares significant homology with the LGIC family also supports a loop structure in this part of the protein. The region from W89 through Y93 is shown as a beta-sheet as has been hypothesized based on site-directed mutagenesis studies of this strand of the 5-HT3R . Our data indicate the functional groups of lerisetron that may interact with W89, R91, Y142 and Y152. Figure 6 | Model of the 5-HT3R binding site. Model of the 5-HT3R binding site. The model illustrates the major points discussed in the text. Lerisetron is shown spanning the region between the D and E loops of the receptor. Both loops are illustrated as beta-sheets as described. The side chains are presumed to point in or out of the page depending on their position. W89, R91, Y140, Y142 and Y152 all presumably extend into the binding site. The terminal piperazine nitrogen is shown interacting primarily with Y142, but is also capable of forming an interaction with Y152. Y152 may also form an interaction with the second piperazine nitrogen. W89 is illustrated as forming an aromatic interaction with the benzimidazole. While R91 is shown in close contact to the N-benzyl group, the data support only a weak interaction at this point. W89 is shown interacting with the aromatic benzimidazole group of lerisetron although the precise position of W89 relative to this group is not known. The W89 interaction with this group is supported by the observed increase in Ki for lerisetron, analog 1 and analog 2. Since the benzimidazole group is common to all three compounds it is the most likely point of interaction. W89 also represents a common interaction in the binding site for both lerisetron and [3H]granisetron. Y142 is shown interacting with the distal piperazine nitrogen possibly through a cation-pi interaction. This orientation of an amino group interacting with an aromatic amino acid in a cation-pi interaction has been shown for the nicotinic acetylcholine receptor and has been hypothesized for many LGIC receptors [-,]. This conclusion is based on both the magnitude of the change observed on the wild type receptor for removal of the amino group (400 fold) compared to the effect of the Y142A mutation on lerisetron binding (160 fold) and the lack of any major change in Ki for analog 2 as a result of this mutation. Our data does not support an interaction of this amino acid with either the N-benzyl or benzimidazole portions of lerisetron. Y152 is shown positioned between the two piperazine nitrogens. This conclusion is supported by the smaller increase in Ki for analog 2 (40 fold) compared to that observed for lerisetron (190 fold). These results suggest a partial interaction of Y152 with the distal piperazine nitrogen. Since some change was observed, a second interaction is also likely. The functional group in closest proximity to the distal piperazine nitrogen is the other nitrogen on the piperazine ring. Another possibility would be the N-benzyl interaction, however, since the Y152A mutation also produced a large increase in Ki for analog 2, this conclusion is not supported by our data. R91 is shown as interacting with the N-benzyl group. This is a difficult conclusion to make considering the small effect of the R91A mutation on lerisetron binding. The interaction is included in the model based on structural information obtained from the crystal structure of AChBP . The region of this protein homologous to loop E and loop D of the 5-HT3A receptor suggests a loop structure from Y140 to Y152 and a 3-residue turn containing a glycine at position 147 and the beta-strand from W89 through Y93 oriented as shown in Figure . The orientation of lerisetron between W89 and Y142A as shown would enable the N-benzyl group to be positioned in close proximity to R91. If this is the case, then a small alteration in position of analog 1 or 2 in the binding site could result in the loss of this presumably weak interaction. The apparent alterations in the binding site location of analog 1 would be consistent with this hypothesis. An alternate hypothesis would place the N-benzyl group in a different position, interacting with another amino acid; either solely or in concert with R91. Our data support a binding site for lerisetron on the 5-HT3R that spans the D and E loop regions. Table shows the sequence alignment for the 5-HT3R, the alpha7 receptor and the AChBP for these loops. Sequence alignment of mouse 5-HT3 AR, alpha7 nAchR and AChBP result in alignment of the proposed D and E loop of the 5-HT3 AR with corresponding regions of the alpha7 nAchR and AChBP. The amino acids W89, R91 Y140, Y142 and Y152 of the 5HT3 A R can be aligned with W53, Q55, L102, R104, and M114 of the AChBP . These amino acids form a cluster in the proposed acetylcholine binding domain of AChBP similar to that proposed in our model. Both loops have been identified on the complementary face of the binding site of the nAChR. It is unknown if lerisetron utilizes amino acids on the principal face although none have been identified. The model for lerisetron binding will be further refined as its interactions with other binding site amino acids are investigated . Of particular interest would be potential interactions of the N-benzyl group that would account for the decrease in binding affinity of analog 2. Additional information gained from comparison of our model with the recent crystal structure of the AChBP demonstrates that lerisetron can be roughly 'fit' into the binding site such that all the residues line up as shown in our model. While this is not direct evidence that the model is correct, subsequent molecular modeling of the data presented in this paper may provide further support for our hypothesis. Our current model provides an initial working hypothesis that can form the basis of further investigation. Also, while it is unclear whether the information obtained in this study can be extended to other 5-HT3R ligands, a similar approach would be useful in identifying functional group interactions for mCPBG, 5-HT, dtC and granisetron. Table 2 | Sequence alignment of the mouse 5-HT3R with the mouse nicotinic alpha7 receptor and the AChBP. Two sequences are shown. The top sequence is the putative D loop containing W89 and R91 (homologous to W53 and Q55 of the AChBP). The lower sequence shows the putative E loop and contains Y140, Y142 and Y152 (homologous to L102, R104 and M114 of the AChBP). Materials and Methods : Mutagenesis | Wild type 5-HT3AS mouse receptor cDNA was derived from N1E-115 neuroblastoma cells as previously described . Mutant receptors were constructed using polymerase chain reaction (Quick change mutagenesis kit, Promega). All mutations were confirmed by DNA sequencing. Cell culture methods and transfections | tsA201 cells (a derivative of the HEK293 cell line) were grown in Dulbecco's modified Eagles medium (D-MEM) containing 10% FBS and 100-units/ml penicillin/streptomycin. Cultures were maintained in humidified atmosphere of 5% CO2 at 37C. For binding studies, tsA201 cells were plated at a density of 5 x 106 cells/75 cm2 and grown for 9 hours prior to transfection. Cells were transfected with 10 mug murine 5-HT3ASR cDNA using calcium phosphate co-precipitation (New Life Technologies, NY), then incubated 36 hours prior to harvesting. For whole cell patch clamp experiments, tsA201 cells were plated at a density of 0.25 x 106 cells/27 cm2 dish and grown 12 hours prior to transfection. Cells were washed with fresh culture medium then transfected with 10 mug 5-HT3ASR cDNA using Qiagen Superfect transfection reagent (Qiagen, CA). Transfected cells were incubated with this mixture for 2.5 hours, then divided into 35 mm culture dishes at a density of approximately 5 x 104 cells/dish and incubated for 24 hours at 37C before recording. Radioligand Binding Assay | Transfected cells were scraped from the dishes, washed twice with Dulbecco's PBS (New Life Technologies, NY), then resuspended in 1.0 ml PBS/100 mm dish. Cells were either used fresh or frozen at this step until needed. Immediately prior to use, cells were homogenized in PBS using a glass tissue homogenizer then centrifuged at 35 000 x g for 30 minutes in a Beckman JA20 rotor. Membranes were washed once more with PBS then resuspended in 1 ml PBS/100 mm dish. Protein content was determined using a Lowry assay (Sigma. Diagnostics, St. Louis, MO). Binding assays were performed in PBS. For Kd determinations, 100 mul of homogenate was incubated at 37C for 1 hour with varying concentrations of [3H] granisetron (NEN, MA). Specific binding of [3H] granisetron was determined as the bound [3H] granisetron not displaced by a saturating concentration of a competing ligand (100 muM mCPBG or 10 muM MDL-72222). Kd values were determined by fitting the binding data to the following equation using GraphPad PRISM (San Diego CA): B = Bmax [L] n / ([L] n + Kn), where theta is bound ligand, Bmax is the maximum binding at equilibrium L is the free ligand concentration and n is the Hill coefficient. For Ki determinations, 100 mul of homogenate was incubated at 37C for 2 hours with varying concentrations of inhibitor and [3H] granisetron (NEN, MA). Binding was terminated by rapid filtration onto a GF/B filters. The IC50 values were calculated by fitting the data to the following equation using GraphPad PRISM (San Diego CA): theta = 1/ (1+(L/IC50)), where theta is the fractional amount of [3H] granisetron bound in the presence of inhibitor at concentration L as compared to the amount of [3H] granisetron bound in the absence of inhibitor. IC50 is the concentration at which theta = 0.5. The Ki is calculated from the IC50 value using the Cheng-Prusoff equation. Electrophysiological Methods | Transfected tsA201 cells were transferred to a recording chamber and submerged in extracellular recording buffer containing 25 mM HEPES pH 7.4, 140 mM NaCl, 1.7 mM MgCl2, 5 mM KCl, 1.8 mM CaCl2. Patch electrodes (2 --2.5 MOmega) were filled with intracellular recording buffer containing 25 mM HEPES pH 7.4, 145 mM KCL, 2 mM MgCl2 and 1 mM EGTA. Cells were clamped in whole cell configuration at a holding potential of -60 mV. Currents elicited by agonist application were measured using an Axopatch 200B amplifier (Foster City, CA) under computer control (DataPac 2000, RUN Technologies). Agonists and antagonists were dissolved in extracellular solution and delivered to cells using a rapid perfusion system (Warner Instruments, Hamden, CT). For EC50 determinations, responses were normalized to the maximum response obtained from the full agonist 5-HT and fitted to the equation Psi= 1/1+(EC50/ [C] n), where Psi is the normalized current at 5-HT concentration [C], EC50 is the concentration of 5-HT needed to obtain half maximal activation and n is the apparent Hill coefficient. For inhibition experiments, cells were exposed to inhibitor alone for 30 s prior to co-exposure with 5-HT. Inhibited responses were calculated as a fraction of the response to 5-HT alone. Data were plotted as the fractional response versus the concentration of inhibitor and analysed using GraphPad software. The IC50 value was calculated as the concentration of antagonist inhibiting the 5-HT evoked response by 50%. A Ki value was calculated from the IC50 using the Cheng-Prusoff equation. Synthesis of Lerisetron and its analogs | All target molecules were prepared according to a general 2-step synthesis reported previously by Orjales et al. with only slight modification . Commercially available 2-chlorobenzimidazole, in dry DMF was treated with a slight excess of NaH, (1.1eq). After stirring for 1 hour at room temperature, one equivalent of the appropriate alkyl bromide was added slowly and the reaction mixture heated under reflux for > 5 hours (the reaction was monitored by TLC). Reaction product was partitioned between water and methylene chloride; organic layer was dried (Na2SO4) and concentrated in vacuum. The solid residue was purified by Flash chromatography, which afforded the corresponding N-substituted 2-Chlorobenzimidazole intermediates in good yield. The final step involved a nucleophilic substitution of the 2-chloro group by piperazine at high temperatures. The reaction was performed neat using 4 --10 fold excess piperazine and typically heated for a short period only, (30 --45 min). Similar work-up afforded a residue that was purified by either crystallization or chromatography. The yields ranged from 40 --95%. All compounds were characterized by NMR, MS, HRMS, and /or elemental analysis or were identical to literature reports. Materials | D-MEM, Penicillin-Streptomycin, fetal bovine serum, and Trypsin were obtained from New Life Technologies. 5-HT and MDL-72222 were obtained from RBI. [3H]-granisetron (84 Ci/mmol) was purchased from New England Nuclear. Backmatter: PMID- 12079498 TI - Low solubility of unconjugated bilirubin in dimethylsulfoxide -- water systems: implications for pKa determinations AB - Abstract | Background | Aqueous pKa values of unconjugated bilirubin are important determinants of its solubility and transport. Published pKa data on an analog, mesobilirubin-XIIIalpha, studied by 13C-NMR in buffered solutions containing 27 and 64 vol% (C2H3)2SO because of low aqueous solubility of mesobilirubin, were extrapolated to obtain pKa values in water of 4.2 and 4.9. Previous chloroform-water partition data on bilirubin diacid led to higher estimates of its pKa, 8.12 and 8.44, and its aqueous solubility. A thermodynamic analysis, using this solubility and a published solubility in DMSO, suggested that the systems used to measure 13C-NMR shifts were highly supersaturated. This expectation was assessed by measuring the residual concentrations of bilirubin in the supernatants of comparable DMSO-buffer systems, after mild centrifugation to remove microprecipitates. Results | Extensive sedimentation was observed from numerous systems, many of which appeared optically clear. The very low supernatant concentrations at the lowest pH values (4.1-5.9) were compatible with the above thermodynamic analysis. Extensive sedimentation and low supernatant concentrations occurred also at pH as high as 7.2. Conclusions | The present study strongly supports the validity of the aqueous solubility of bilirubin diacid derived from partition data, and, therefore, the corresponding high pKa values. Many of the mesobilirubin systems in the 13C-NMR studies were probably supersaturated, contained microsuspensions, and were not true solutions. This, and previously documented errors in pH determinations that caused serious errors in pKa values of the many soluble reference acids and mesobilirubin, raise doubts regarding the low pKa estimates for mesobilirubin from the 13C-NMR studies. Keywords: Background : The saturation status of unconjugated bilirubin (UCB) is relevant to understanding the pathophysiology of jaundice and to interpreting experiments with UCB . UCB, in its diacid form (H2B), has a low solubility (So) of 51 nM in water . The total solubility (S) at any pH is determined by So, pKa values, and pH : S = [H2B] + [HB-] + [B=] = So(1 + Ka1/ [H+] + Ka1Ka2/ [H+]2) (Eq. 1) Self-association of UCB dianion, B=, can also increase S . Beginning in 1995, a series of papers reported the use of 13C-NMR to study the ionization of a close analogue of bilirubin-IXalpha, mesobilirubin-XIIIalpha (MBR), along with several soluble reference acids, in buffered mixtures of (C2H3)2SO and water . Due to the very low aqueous solubility of MBR, data were obtained only at two high concentrations of (C2H3)2SO (64 and 27 vol%). Its pKa values in "water" (actually in 1 vol% of ((C2H3)2SO), obtained by an extrapolation procedure based on the behavior of soluble reference acids, were 4.2 and 4.9, far lower than values of 8.12 and 8.44 obtained by solvent partition between chloroform and buffered aqueous solutions . These serious discrepancies led us to reexamine several experimental aspects of the 13C-NMR studies and the implications of their purported low pKa values for the solubility of UCB. Inaccuracies in the measurements of buffer pH and of the pKa values of the reference 13C-carboxylic acids, related to failure to correct for the strong effects of DMSO on the pKa values of weak acids, have been previously noted and acknowledged . In addition, as discussed later, a thermodynamic theory about solubility in mixed solvents, using the So values of 51 nM in water and 10 mM in pure DMSO , suggested that So of UCB would be only about 0.15 muM in 27 vol% DMSO and 2.2 muM in 64 vol% DMSO, which are well below concentrations used in the MBR studies. This implies serious supersaturation effects. In fact, turbidity was reported for some of the systems used in the 13C-NMR papers , indicating the formation of coarse suspensions. By definition, as stressed in the Conclusions section, pKa determinations are valid only for monomeric species and require that solutions are below saturation at all pH values studied. Even reversible aggregation of monomers in undersaturated solutions is known to affect pKa estimates . If supersaturation leads to the formation of fine sols or coarse suspensions, the data are unacceptable for determination of the pKa values of monomers. In the present paper, we assess whether some of the systems used in the 13C-NMR studies were supersaturated with MBR. Our experimental work on sedimentation was done with unconjugated bilirubin (UCB) and we used DMSO instead of its deuterated analogue, (C2H3)2SO. It is known that, when alkaline aqueous UCB is acidified to neutral or low pH's, "Usually a colloid suspension of bilirubin is formed and the solution remains clear, as observed by the naked eye, thus inviting an erroneous interpretation" . Such systems, which simulate true solutions, often show sedimentation on centrifugation . We here report that sedimentation of UCB from apparently clear solutions can likewise be extensive when UCB in DMSO is diluted with aqueous buffers to final mole fractions (N) of DMSO = 0.025, 0.086 and 0.31, corresponding to 9, 27, and 64 vol% of DMSO. A thermodynamic theory is used to examine the effect of added DMSO on So. Our findings, consistent with our partition-derived So, S and pKa values of UCB in aqueous buffers , indicate that the recently reported low pKa values of MBR in comparable (C2H3)2SO -water systems , were determined above saturation. Results : Residual UCB in the supernatants after centrifugation | Many systems initially appeared optically clear, but well over 90% of initial UCB sedimented on centrifugation of most samples below pH 7.2. Overall recoveries (UCB in supernatant + UCB in redissolved precipitate) were between 90 and 100%. Only residual [UCB] in supernatants are reported. Figs. plot the measured residual [UCB] in the supernatant vs. the initial [UCB] at NDMSO = 0.31, 0.086 and 0.025. Deviations of their ratio below unity (dotted lines) represent a decrease in [UCB], mainly from precipitation, but possibly also from limited degradation. At N = 0.31 , [UCB] after sedimentation varied from 2.8 muM at pH 5.86 to 166 muM at pH 8.38. At N = 0.086 , [UCB] ranged from 0.3 to 2.4 muM at pH 4.50 and from 1.4 to 6.4 muM at pH 7.05 and most of the UCB sedimented. By contrast, at pH 7.56 and 7.70 (phosphate), only minor precipitation was observed. At N = 0.025 ; [UCB] ranged from 0.1 to 2.7 muM at all three pH values (4.15 to 7.18). Figure 1 | A-C. Residual vs. initial UCB concentrations in buffered DMSO/water systems. A-C. Residual vs. initial UCB concentrations in buffered DMSO/water systems. Measured residual [UCB] in supernatants of buffered DMSO/water systems, after centrifugation for 5 min at 14,000 g, are plotted against the pipetted, initial UCB concentrations before centrifugation. Mole fractions (N) of DMSO were 0.310 (A), 0.086 (B) and 0.025 (C). Panels B and C are log-log plots. The dashed unity line indicates complete retention of UCB in supernatant. The buffers used and the corrected pH values of the DMSO-buffer systems were: Panel A -- (circle) Phosphate pH 8.4, (square) Tris pH 7.1, (diamond) Acetate pH 6.9, (triangle) Acetate pH 5.9; Panel B -- (circle) Phosphate pH 7.6 --7.7, (square) Tris pH 7.1, (triangle) Acetate pH 4.5; Panel C -- (circle) Phosphate pH 7.2, (square) Tris pH 7.0, (triangle) Acetate pH 4.15. [UCB] in each supernatant was determined from A at 458 nm, using the extinction coefficient of 0.0634 muM-1cm-1. Sedimentation of bilirubin-albumin complexes | The supernatant from the original supersaturated UCB-HSA system, when diluted with 1/8th vol. of buffer and centrifuged again, showed more sedimentation and a progressive rise in [UCB] as one moved down the column of fluid. By contrast, after dilution with 1/8th volume of DMSO to decrease UCB saturation, the supernatants produced no further sedimentation and there were no significant differences in [UCB] or protein concentrations along the axis of the fluid column. Thus, the UCB-HSA complex, which constitutes over 99.9% of the UCB in this system , did not sediment. Discussion : Findings and their relation to thermodynamic theory | Supersaturated aqueous systems of UCB, that are optically clear before centrifugation, may exhibit considerable variation in the extent of sedimentation [-,]. Although sedimentation is often extensive, it is generally incomplete and may not be observed at all. Our data in DMSO-water show the same features, which are expected from the complex kinetics of nucleation and growth of insoluble aggregates of UCB diacid (H2B), leading to the formation of a new solid phase . Our centrifugation, 5 min at 14,000 g, was quite mild, and the short 20-minute period between preparation of UCB-DMSO-water systems and centrifugation severely limited the time-dependent growth to large aggregates. Lack of sedimentation of the UCB-HSA complex (mol. wt. 68,000) indicates that fine colloids composed of 100 UCB molecules would be too small to sediment. Thus, supersaturated systems lacking coarse, insoluble aggregates may not show sedimentation, but any sedimentation observed indicates their presence. To evaluate the important effect of pH on sedimentation efficiency, we calculated S in water using chloroform-water partition data on UCB and the best measure of So in chloroform, 0.88 mM . S at any pH, e.g. 62 nM at pH 7.4 and 0.32 muM at pH 8.5, can be calculated from the fitted partition data, or, equivalently, from Eq. 1, using the partition-derived So in water of 51 nM and pKa values of 8.12 and 8.44 . In aqueous systems , the lowest [UCB] in water, below which no sedimentation was observed at 100,000 x g for a few hours, was 100 nM at pH 7.4, modestly higher than our partition-derived S of 62 nM . Even under such vigorous centrifugation, the lowest [UCB] increased rapidly with increasing pH, to 17 muM (150 times S) at pH 8.05 and 34 muM (230 times S) at pH 8.2 . This indicates increasing charge-stabilization of fine, non-sedimenting colloids of H2B by adsorbed UCB anions . In contrast, below pH 6.7, sedimentation of 10 muM UCB was nearly complete . This is compatible with a dearth of stabilizing UCB anions at this pH, as expected from the high pKa values of 8.12 and 8.44 . Our present data on residual [UCB] in DMSO-water systems likewise show decreased sedimentation with increasing pH . At each NDMSO, the lowest [UCB] were at the lowest pH values: 0.1 muM (N = 0.025, pH 4.15); 0.3 muM (N = 0.086, pH 4.5); and 2.8 μM (N = 0.31, pH 5.9). As in water, these are likely to be closest to the So values at each N. Indeed, they are only moderately higher than the corresponding So values of 0.07 muM, 0.15 muM and 2.2 muM, respectively, calculated from Equation 2 using So values of 51 nM in water and 10 mM in DMSO . log So,mixed = log So,water + (log So,DMSO - log So,water) x N (Eq. 2) Equation 2 is a thermodynamic relationship based on assumptions of complete ideality of mixing . In general, a roughly linear variation of log So with N at low N is expected. For example, data from 1-naphthoic acid in DMSO-water show that log So is a linear function of N up to N = 0.35. Such a relationship leads to a relatively small effect of low N values on So. Thus, according to Equation 2, So increases by a factor of only 1.4 at N = 0.025 and 2.9 at N = 0.086, but by a relatively larger factor of 44 at N = 0.31. This would markedly reduce the supersaturation factor ([UCB]/ So), which is a measure of the tendency of UCB to come out of solution at N = 0.31. This explains in part the relatively high [UCB] at high pH at N = 0.31 . The pH effects on [UCB] at each NDMSO are of interest also. The lowest [UCB] at each pH registered relatively small increases with significant increases in pH: for example from 0.1 muM (pH 4.14) to 0.2 muM (pH 7.0) at N = 0.025; from 0.3 muM (pH 4.5) to 1.4 muM (pH 7.1) at N = 0.086; and from 111 muM (pH 7.1) to 166 muM (pH 8.4) at N = 0.31. These increases are probably caused mainly by increasing charge-stabilization of colloidal aggregates, as in aqueous media . If, instead, the relatively small increases are ascribed entirely to increases in true solubility (S) at the high pH (Eq. 1), the required pKa values are about 7 at N = 0.025 and 0.086, and 8.5 at N = 0.31. The true pKas of UCB in DMSO-water are thus probably significantly higher. We note that some variability in sedimentation results from our short-term experiments, most evident at the low residual [UCB] in Figs. and , in part magnified by the log-log scale used. Some variability is expected, however, because of the complexity of the kinetic processes of nucleation, growth and flocculation that precede sedimentation. In Fig. , the difference between acetate and Tris buffers is quite small (note the linear scale), compatible with the 58% higher [H+] in the acetate buffer. In Fig. , the markedly lower sedimentation from phosphate buffers at pH 7.6 --7.7, as compared to Tris buffer at pH 7.1, can be ascribed mainly to the much higher pH values and ionic strength of the phosphate systems. Another significant factor may be the difference in charge between the buffer salts; phosphate is anionic whereas Tris is cationic and zwitterionic. The cationic species of Tris can, in principle, reduce the negative charges on the surface of the colloidal H2B sufficiently to facilitate the formation of coarser particles and, thus, increase sedimentation. Implications for pKa values of mesobilirubin-XIIIalpha (MBR) | In the recent 13C-NMR studies of the ionization of the 13C-COOH groups of MBR , it was assumed that the relevant physical properties of UCB and MBR, and of (CH3)2SO (DMSO) and (C2H3)2SO, are similar. Actually, as expected from the replacement of two vinyl groups in UCB with two ethyl groups in MBR, MBR is slightly more soluble in organic solvents and has a higher Rf on silica gel t.l.c. ; MBR is thus more hydrophobic and should be less soluble in water than is UCB. Our low [UCB] in DMSO-water systems at comparable N, therefore, indicate that many of the (C2H3)2SO/buffer systems used in the 13C-NMR studies were likely supersaturated with MBR. In those studies, the MBR concentrations used were stated to be 1 to 100 muM at N = 0.086 , compared to our lowest [UCB] of 0.3 muM at pH 4.5 and 1.4 muM at pH 7.05. At this N, 9 of 11 MBR data points were obtained at pH below 7.05 and 5 below pH 4.5 , so that even 1 muM MBR was likely to be supersaturated. At N = 0.31, our lowest [UCB], 2.8 muM at pH 5.9, was close to the lower limit of the 2 to 800 muM range of [MBR] used . Thus, many data points, obtained at pH values down to 2 , were probably from supersaturated systems, despite being optically clear. As noted here and elsewhere [-,], optical clarity gives no assurance of the absence of supersaturation. Actually, turbidity was reported in some of the 13C-NMR samples , indicating that coarse, insoluble aggregates of MBR were present. The claim that such turbidity did not affect 13C-NMR measurements contrasts with evidence that even small multimers can change NMR chemical shifts . It should be noted also that, at high concentrations of B=, extensive, reversible self-association of B= can lead to apparently stable supersaturation with no separation of an insoluble phase . For example, at pH 8.5 and a UCB concentration of 20 muM (63 times S), the weight-average aggregation number of UCB has been found to be 7.17 , corresponding to a molecular weight of 4,195. The aggregation number remained fairly high, 4.2, in 60% (w/v) ethanol . The successful application of equilibrium ultracentrifugation for that study suggests a complete absence of even small colloidal species of UCB. Self-association of MBR dianions in (C2H3)2SO-water mixtures cannot be ruled out on a priori grounds. It has been shown that neglect of self-association of B= leads to an artefactually low estimate of pKa values for UCB . In addition to the problems of insolubility, supersaturation and self-aggregation of the MBR systems in (C2H3)2SO-water , we had shown previously that inaccuracies in the pH measurements affected both the magnitude of DeltapKa (the change in pKa on adding (C2H3)2SO to water), as well as the degree of the variation of DeltapKa with N . This is important for extrapolating pKa values in (C2H3)2SO-water to pure water (N = 0). Indeed, remeasurement of one soluble acid raised its pKa by as much as 3 units at N = 0.31 . Thus, the inaccuracies in pH measurement produced serious errors in reported pKa values of more than fifteen soluble acids used as models for MBR, as well as for MBR itself . Many methods, using appropriate pH measurements, have been applied in the past to determine thermodynamic pKa values of soluble acids in non-aqueous or partially aqueous media, including DMSO-water systems . Many other relevant references were given in our prior paper . In that paper, our pKa measurements on acetic acid in DMSO-water systems were based on the potentiometric method, using properly calibrated glass electrodes, which determine the activity of H+, and on estimates of the activity coefficients of the acetate ion. This method, which is well established for aqueous solutions, yielded results in good agreement with data from the literature that was based on a very different method, measurements of electrical conductivity . In the 13C-NMR papers, therefore, it was not justified, to assume that pH values do not change on adding DMSO , or to use uncalibrated pH measurements for determination of the pKa values of soluble acids . Our sedimentation data and their interpretation indicate that significant additional uncertainties, not important for the soluble acids investigated, exist for the reported pKa values of the relatively insoluble MBR in (CD3)2SO-water (4.2 and 4.9 at N = 0.086 and 4.3 and 5.0 at N = 0.31), as well as their extrapolation to obtain pKa values of 4.2 and 4.9 in water . Indeed, if these low aqueous pKa values, along with the experimental S values at pH 8.5 of 0.32 muM , or 0.6 muM , are applied to Eq. 1, the calculated extremely low So values of UCB diacid of 4 or 8 x 10-15 M are seven orders of magnitude lower than the experimental So, 5.1 x 10 --8 M . Applying the So of 4 or 8 x 10-15 M to Eq. 2, moreover, would indicate massive supersaturation (up to 8 to 10 orders of magnitude) of MBR at the concentrations (1 --800 muM) used in the 13C-NMR studies . Conclusions : The present sedimentation data for UCB in DMSO-water demonstrate that the true solubilities of UCB, even at fairly high pH values, are low at DMSO mole fractions up to 0.31. The results and related considerations are compatible with similar results in purely aqueous solutions , and support both the estimated solubility (So) of 5.1 x 10 --8 M for uncharged UCB (H2B) in water, and the corresponding high aqueous pKa values of 8.12 and 8.44, derived from our partition studies . These were performed in undersaturated systems and took into account the self-association of B=. Our experimental data indicate problems of insolubility, supersaturation and self-aggregation of UCB in DMSO-water mixtures with compositions similar to the MBR systems in (C2H3)2SO -water . In (C2H3)2SO-water, DMSO-water or any other medium , properly determined pKa values for the dissociation equilibria of a diacid H2A (H2A <--> HA- + H+ and HA- <--> A= + H+) must pertain to monomeric H2A, HA- and A=, the solute species involved in the stated equilibria, and require unambiguous determination of [H+] or pH. Unless pKa values are determined for monomeric systems, relative concentrations of H2A, HA- and A= cannot be determined from the pKa values and the pH. The 13C-NMR data, suggesting low pKa values for MBR in (C2H3)2SO -water and water [-,], did not meet these essential requirements of proper pH measurements nor provide assurance that the MBR in every system was below saturation and not self-associated . The issues raised are not trivial, since the pKa and So values of UCB are clinically relevant to the effects of pH on the precipitation of calcium bilirubinates in pigment gallstones and the neurotoxicity caused by UCB diacid in severely jaundiced neonates . Materials and Methods : Materials | UCB (Calbiochem) was purified by alkaline extraction of a chloroform solution, recrystallized twice from chloroform-methanol , dried under Argon, stored invacuo in the dark and used within 6 weeks. DMSO was spectroscopic grade, 99.8% pure (UVASol, Merck). Human serum albumin (HSA, lot 903635) was from Calbiochem-Boehringer. All other chemicals were reagent grade (Merck). Water used was deionized and distilled. All flasks and tubes were Kimax glass, washed with 0.1 N HCl and rinsed 4X with water and then dried before use. Stock buffers, 1.0 M, were: Tris-HCl, pH 7.01; Na- phosphate, pH 6.85, or 6.99; and Na-acetate, pH 4.01. Stock UCB in DMSO (4 to 6 mM) and stock HSA, 613 muM in 0.1 M Tris-HCl buffer, pH 7.01, were prepared freshly for each experiment. Preparation of UCB-DMSO-buffer systems | Test systems (4.0 mL) of UCB were prepared in duplicate: to 0.4 mL of the stock buffer were added successively the appropriate volumes of water, DMSO and, finally, up to 150 muL of stock UCB/DMSO solution. To minimize UCB oxidation, all tubes and solutions were deoxygenated with Argon and kept in the dark . To determine the [UCB] in the UCB/DMSO stock, 6.0 muL, was added to 3.0 mL of the HSA stock. The absorbance, A, at 460 nm was read against a blank containing 2.0 mL of HSA stock plus 4.0 muL DMSO, and the [UCB] calculated using the extinction coefficient, epsilon, 47,000 M-1.cm-1. The pH measurement of each final DMSO/aqueous buffer system included an electrode calibration using the strong acid, HClO4. The 0.1 M phosphate buffer in N = 0.31 DMSO was centrifuged because of partial insolubility and only the supernatant was used. Centrifugation and analysis of residual UCB in supernatants | Fifteen min after mixing, samples were assessed visually for turbidity or precipitation. After Vortex-mixing, duplicate 1.8 mL aliquots were transferred to polypropylene tubes and centrifuged for 5 min at 25C and 14,000 g (Mikroliter centrifuge, Hettich, Tuttlingen, Germany). The supernatants were assayed spectrophotometrically within 20 min. The precipitates were washed once with 1.8 mL of water, again centrifuged, the water aspirated, and the packed precipitate dissolved in DMSO for spectrophotometry. Absorbance (A) was measured in triplicate at 458 nm on each sample, diluted, when necessary, with DMSO or DMSO/buffer mixture to A of 0.2 to 0.8; a comparable medium without UCB was used as a blank. In all systems, including the redissolved UCB sediments, we applied the extinction coefficient of 0.0634 muM-1cm-1 at the peak wavelength of 458 nm for UCB in pure DMSO . Preliminary calibration studies. of the effects of DMSO concentration and buffer composition on A had confirmed this value for pure UCB in pure DMSO and in DMSO/buffer systems containing 64 vol% DMSO. In the systems containing 27 and 9 vol% DMSO, the spectrum developed a plateau between 458 and 450 nm, but A at 458 nm remained within +- 10% of the value expected from applying the extinction coefficient of 0.0634 muM-1cm-1 to the measured quantity of UCB dissolved in each system. The variability is in part due to degradation, discussed above, and in part due to the low absorbances at [UCB] below the saturation limit in some buffer/DMSO systems. For these reasons, no corrections were made for these minor differences in A. Sedimentation of bilirubin-albumin complexes | To determine if UCB bound to HSA would sediment, we prepared a system containing HSA, 300 muM, and UCB 170 muM, in 0.1 M Tris-HCl buffer, pH 7.01. Microcentrifugation for 5 min yielded a small amount of precipitated UCB. Three aliquots of the clear supernatant were diluted with 1/8th volume of DMSO and a fourth aliquot diluted with 1/8th volume of buffer. The diluted samples (in duplicate) were then microcentrifuged for another 10 min and 25 muL samples taken from the top, middle and bottom of the fluid column in each tube, using a Hamilton syringe. Protein concentrations were determined with the Bio-Rad bicinchonic acid method, which is unaffected by bilirubin. After dilution with 1.8 mL DMSO, triplicate A readings were taken at 458 nm. Abbreviations : UCB, unconjugated bilirubin; H2B, UCB diacid; B=, UCB dianion; DMSO, dimethylsulfoxide; MBR, mesobilirubin XIIIalpha; N = mole fraction of DMSO in DMSO-aqueous buffer systems; NMR, nuclear magnetic resonance; S, solubility of UCB or MBR at a given pH; So, solubility of UCB diacid. Authors' note : An abstract of this work has been published (Gastroenterology 2000; 118:A1477) Authors' contributions : All three authors collaborated in the conception, design and writing of this study. The work was performed by JDO while he was a visiting professor at the Academic Medical Center in Amsterdam, the Netherlands. All authors read and approved the final manuscript. Backmatter: PMID- 12079500 TI - Identification of critical residues in loop E in the 5-HT3ASR binding site AB - Abstract | Background | The serotonin type 3 receptor (5-HT3R) is a member of a superfamily of ligand gated ion channels. All members of this family share a large degree of sequence homology and presumably significant structural similarity. A large number of studies have explored the structure-function relationships of members of this family, particularly the nicotinic and GABA receptors. This information can be utilized to gain additional insights into specific structural and functional features of other receptors in this family. Results | Thirteen amino acids in the mouse 5-HT3ASR that correspond to the putative E binding loop of the nicotinic alpha7 receptor were chosen for mutagenesis. Due to the presence of a highly conserved glycine in this region, it has been suggested that this binding loop is comprised of a hairpin turn and may form a portion of the ligand-binding site in this ion channel family. Mutation of the conserved glycine (G147) to alanine eliminated binding of the 5-HT3R antagonist [3H]granisetron. Three tyrosine residues (Y140, Y142 and Y152) also significantly altered the binding of 5-HT3R ligands. Mutations in neighboring residues had little or no effect on binding of these ligands to the 5-HT3ASR. Conclusion | Our data supports a role for the putative E-loop region of the 5-HT3R in the binding of 5-HT, mCPBG, d-tc and lerisetron. 5-HT and mCPBG interact with Y142, d-tc with Y140 and lerisetron with both Y142 and Y152. Our data also provides support for the hypothesis that this region of the receptor is present in a loop structure. Keywords: Background : Sequence homology between the serotonin type 3 receptor (5-HT3R), the nicotinic acetylcholine receptor (nAChR), the GABAA receptor and the glycine receptor suggests a large amount of structural similarity within this superfamily of ligand gated ion channels. This presumed structural homology can be used to guide site directed mutagenesis studies of particular receptor subtypes. On a gross level, all members of this superfamily assemble as pentameric receptors . In some cases, receptors can be assembled from a single subunit (5-HT3AR and nicotinic alpha7 receptors) [,-]. In other cases, at least two different subunits are required . The ligand binding site is thought to be formed at the junction between two subunits . Based on data obtained from affinity labeling and site directed mutagenesis studies, several regions of the nicotinic acetylcholine receptor have been implicated in ligand binding . Six essential loops appear to contribute to the binding site with the position of each loop differing slightly depending on whether the receptor is heteromeric or homomeric. For homomeric receptors such as the alpha7 nAChR, the loops are identified as A, B, C, D, E and F . The amino acids that interact with ligands presumably extend into the binding site from these loops. Subunits are composed of 2 faces (+ and -) with the + face of one subunit forming one side of the binding site and the -- face of another subunit forming the complimentary side. Some binding loops (A, B and C) are present on the -- face while the remaining loops (D and E) are present on the + face. . The individual amino acids that form binding interactions with the functional groups present on a ligand are likely to differ for each receptor subtype, reflecting the specificity of a particular binding site, however, the overall structure of the binding domain may be similar even for binding sites with different ligand specificities. Identification of ligand specificity requires identification of both the location or structure of individual binding site loops and the amino acids present in a particular receptor subtype. The purpose of this study is to extend the information available from other members of this receptor family to the 5-HT3R. Sequence homology and a presumed structural similarity to other ligand gated ion channels suggests that the E loop region of the 5-HT3AR forms part of the ligand binding domain for 5-HT3R ligands. This region extends from Y140 to K153 and is shown in Table . The homologous sequences of other representative members of this family are also shown. In the center of this region is a critical glycine residue that is thought to play a role in establishing a hairpin loop . Recent x-ray crystallographic data obtained from an ACh binding protein (AChBP) shows a loop structure in this region resulting from a 3 residue turn containing a glycine homologous to G147 of the 5-HT3R. On either side of this putative turn region are residues that have been identified as important to receptor binding . The formation of this loop structure brings amino acids on either side of glycine into close proximity and may form a binding pocket that will accommodate one or more functional groups. In order to identify the interaction of amino acids in this binding loop with 5-HT3R ligands, we have constructed alanine mutations of residues throughout this region and evaluated the alteration in binding affinity of 5 different classes of 5-HT3R ligands . Our data identifies 3 tyrosine residues that appear to interact selectively with each structural class and supports the existence of a loop structure in this region of the receptor. Figure 1 | 5-HT3R ligands. 5-HT3R ligands. Table 1 | Sequence Comparison of Putative Binding Loop Region. Results : All mutant receptors were tested for their ability to bind the 5-HT3R antagonist [3H]granisetron. Table shows the Kd values for wildtype mouse 5-HT3ASRs and the 13 alanine mutations we evaluated. [3H]granisetron is a potent antagonist of the wt 5-HT3R (Kd = 0.98 +- 0.12 nM). This value agrees with published data for this compound . Bmax values range from the 5.5 pmoles/mg protein observed for E148A to 0.30 pmoles/mg protein for the K153A mutation, indicating some variability in expression of the different receptors. In general, however, receptor expression was similar to that reported by other laboratories [,-]. No detectable binding was observed for G147A and V149A mutant receptors. For all other mutants, decreases in binding affinity (increased Kd) were observed although the magnitude of the change was less than 10 fold in all cases. A bar graph showing the change in Kd value resulting from each alanine mutation is shown in Figure . The largest decreases in binding affinity were observed for Y142A (4.6 fold, Figure ), E148A (5.3 fold) and Q150A -- K153A (6 -- 8 fold). Figure 2 | Changes in Ki or Kd (for binding of [3H]granisetron) as a result of alanine mutations. Changes in Ki or Kd (for binding of [3H]granisetron) as a result of alanine mutations. The values shown are the ratio of the Kd or Ki for wildtype and mutant receptors. The length of each bar illustrates the magnitude of change in either the Kd or Ki relative to wildtype while the direction indicates an increase (up) or decrease (down) in the Kd or Ki. A. Most alanine mutations produced little change in Kd. No binding was detected for the G147A and V149A receptors therefore no Kd could be determined. B. Changes in Ki were small for most receptors although large changes were observed for select ligands at Y140A, Y142A and Y152A receptors. The relative increases in Ki for these amino acids are as follows: 5-HT: Y140A 1.3 +- 0.30 fold, Y142A 110 +- 25 fold, Y152A 24 +- 5.6 fold mCPBG: Y140A 7.3 +- 1.7 fold, Y142A 160 +- 36 fold, Y152A 24 +- 5.6 fold d-tc: Y140A 50 +- 12 fold, Y142A 6.5 +- 1.5, Y152A 10.0 +- 2.3 fold lerisetron: Y140A 4.6 +- 1.1 fold, Y142A 160 +- 37 fold, Y152A 190 +- 43 fold Figure 3 | Saturation binding isotherms for binding of [3H]granisetron to wildtype, Y140A, Y142A and Y152A receptors. Saturation binding isotherms for binding of [3H]granisetron to wildtype, Y140A, Y142A and Y152A receptors. The data shown represent the specific binding of [3H]granisetron as determined from at least four identical experiments. Specific binding was determined as the fraction of total binding not inhibited by a saturating concentration of mCPBG or MDL-72222. Data were combined and fitted using non-linear curve fitting. Only small shifts in binding affinity to [3H]granisetron were observed on all three mutants. Kd values, Bmax and Hill coefficients are shown in Table . Table 2 | [3H]Granisetron binding to wildtype and mutant 5-HT3ASRs. Inhibition binding assays were also conducted. Four test compounds with structures representative of the major classes of 5-HT3R ligands were chosen: serotonin (5-HT, the endogenous agonist), m-chlorophenylbiguanide (mCPBG, agonist), d-tubocurarine (d-tc, antagonist) and lerisetron (antagonist). The Ki values for inhibition of [3H]granisetron binding by all four compounds are shown in Table . Little change in Ki value was observed for the majority of mutations. The values highlighted in bold in Table represent the Ki values for inhibition of [3H]granisetron binding on mutant receptors that increased over 10 fold compared to the Ki obtained for wildtype receptors. The bar chart in Figure illustrates the changes in Ki resulting from each mutation on the test compounds. Ratios of Ki are shown as positive for increases in Ki on mutant receptors versus wildtype and negative for decreases. A positive change thus corresponds to a decrease in binding affinity for the compound as a result of the mutation. Large decreases in binding affinity were observed for select compounds only on the Y140A, Y142A and Y152A mutations. Table 3 | Ki values for Inhibition of [3 H]granisetron binding to wildtype and mutant 5-HT3ASRs. A more detailed analysis of the competition binding data obtained for the Y140A, Y142A and Y152A mutations is shown in Figure . For 5-HT, the Y142A mutation produced a 110 fold increase in Ki and Y152A produced a 24 fold increase. No change in Ki was observed for the Y140A mutation . The 5-HT3R agonistmCPBG showed a similar profile for the changes in Ki values resulting from mutations of the three tyrosines . As was observed for 5-HT, the Y142A mutation produced a large increase in Ki (160 fold) while the Y14 0A and Y152A mutations produced only 7 and 24 fold changes respectively. Figure 4 | Inhibition of [3H]granisetron binding by 5-HT3R ligands. Inhibition of [3H]granisetron binding by 5-HT3R ligands. The data represent the combined results of at least four experiments. Assays were conducted as described in Materials and Methods. Fractional response is the fraction of [3H]granisetron binding obtained in the presence and absence of inhibitor. An IC50 value was determined from the data using non-linear curve fitting as described in the methods and the Ki value calculated using the Cheng-Prusoff equation, the Kd as determined in Figure and the concentration of ligand used to obtain the inhibition data. Kd values are reported in Table . A. 5-HT B. mCPBG C. dtC D. Lerisetron The Ki value for d-tc inhibition of [3H]granisetron binding was altered only slightly by the Y142A or Y152A mutations (6.5 fold and 10 fold changes respectively). The Y140A mutation, however, produced a 50 fold increase in the Ki compared to wildtype receptors . The Ki value for lerisetron inhibition of [3H]granisetron binding was increased 160 fold by the Y142A mutation and 190 fold by the Y152A mutation. Only a 4.6 fold change in Ki resulted from the Y140A mutation. Lerisetron was the only compound for which a large increase in the Ki was observed on the Y152A mutation . This mutation produced smaller changes in Ki for 5-HT and mCPBG (24 fold for both) and only a 10 fold change for d-tc. In whole cell patch clamp studies, 5-HT perfusion of cells transfected with Y140A and Y142A cDNA produced no responses at 5-HT concentrations of up to 1 mM although specific binding to these receptors was identified in receptor binding studies. Unlike Y140A and Y142A, Y152A receptors responded to application of 5-HT. Due to the low potency of 5-HT on these receptors, only a portion of the concentration response curve could be determined (up to 1 mM). The EC50 value was estimated as greater than 370 muM . This value shows a greater than 140 fold increase in EC50 compared to wildtype receptors; a larger change than was observed for the Ki (24 fold). The most dramatic change observed for whole cell currents was an alteration in the kinetics of the response elicited by application of 5-HT. Y152A mutant receptors displayed much slower rise times compared to wildtype receptors at all concentrations tested. Peak wt responses were typically obtained in less than 80 ms while Y152A responses required several seconds to plateau. Desensitization kinetics were also altered. While wt receptors desensitized rapidly, mutant receptors showed no desensitization during the 8 s perfusion time. Figure 5 | Whole cell patch clamp recording from tsA 201 cells transfected with wildtype and Y152A cDNA. Whole cell patch clamp recording from tsA 201 cells transfected with wildtype and Y152A cDNA. Data were obtained using whole cell patch clamp on tsA201 cells transfected with wt or Y152A mutant 5-HT3ASR cDNA. Cells were clamped at a holding potential of -60 mV and 5-HT applied with a rapid perfusion system. The application time for 5-HT is indicated by the bar above the curves. The responses indicated in A and B are representative examples of at least 4 sets of experiments. A. Wildtype 5-HT3ASR. 5-HT was applied as indicated for a 4 s duration. The concentrations of 5-HT used and the response amplitudes are: a. 0.3 muM (0.03 nA) b. 1.0 muM (0.96 nA) c. 3.0 muM (2.5 nA) d. 30.0 muM (4.4 nA), e. 100.0 muM (3.9 nA). B. Y152A mutant receptors. 5-HT was applied as indicated for an 8 s duration. The curves in B were offset vertically for clarity. The dashed line in each curve indicates the baseline level for that response. The concentrations of 5-HT used and the response amplitudes are: f. 3.0 muM (0.10 nA) g. 100 muM (0.14 nA) h. 300.0 muM (0.57 nA) i. 1.0 mM (1.0 nA). C. Concentration response curve for 5-HT activation of wt 5-HT3ASRs. Each data point represents the combined results of 4 experiments. Responses were normalized to the response obtained at 30 muM and the curves fit as described in the methods section. EC50 = 2.7 +- 0.25 muM, n = 1.6 +- 0.22. D. Concentration response curve for 5-HT activation of Y152A 5-HT3ASRs. Each data point represents the combined results of 4 experiments. The response did not plateau in the range of concentrations used and hence responses were normalized to the maximum concentration of 5-HT (1 mM). The EC50 was estimated by fitting the data to the following equation: I = Imax/(1 +(EC50/ [C])n), where is the normalized current at 5-HT concentration [C], EC50 is the concentration of 5-HT needed to obtain half maximal activation and n is the apparent Hill coefficient. Imax in this case represents the estimated maximum current relative to the response obtained at 1 mM. To obtain the curve shown in the figure, the Hill coefficient was fixed at the wt value of 1.6. The EC50 value obtained from this fit is 370 +- 27 muM. Discussion : The putative E-loop region of the LGIC family of receptors is homologous to residues Y140 through K153 in the 5-HT3R [,-]. Structure-function studies of this region have been conducted in several other members of this family of receptors including GABAA and nAChR subtypes. In each case, residues have been identified that alter either the binding of selective ligands or receptor function [,-]. In order to determine if this loop also contains residues critical to the structure or function of the 5-HT3R, we have constructed alanine mutations of amino acids throughout the homologous region of the mouse 5-HT3ASR and investigated the affects on binding of 5 different structural classes of 5-HT3R ligands. We have identified three tyrosine residues that appear to play a role in binding of selective ligands to this receptor. In addition, our data support the existence of a loop structure in this region as has been hypothesized for the nAChR and identified in a homologous AChBP . Representative members of 5 major structural classes of 5-HT3R ligands were tested on all mutants . These ligands include the antagonists [3H]granisetron, d-tc and lerisetron, and the agonists 5-HT and mCPBG. Most of the mutations tested produced only minor changes in binding affinity for these ligands. Large changes in binding are only apparent on Y140A, Y142A and Y152A receptors for select ligands. The resulting effects of individual mutations are specific to particular structural classes of ligands. The Y140A mutation altered the Ki for d-tc inhibition, but did not alter the Ki obtained for any other compound tested, while Y142A altered the Ki for mCPBG, 5-HT and lerisetron but had little if any effect on inhibition by d-tc. These data indicate the highly specific nature of the effects introduced by the alanine mutations and appear to reflect specific changes in ligand/receptor interaction. Binding of [3H]granisetron is altered only slightly by the alanine mutations introduced in this study (<10 fold). This result indicates the lack of involvement of amino acids in this binding loop in the binding of [3H]granisetron. It is apparent, however, that [3H]granisetron does occupy the same binding cleft as other 5-HT3R ligands as evidenced by the ability of 5-HT, mCPBG, d-tc and lerisetron to displace it from the binding site. The lack of any large change in binding of granisetron supports our contention that there is little global structural perturbation of the binding site resulting from the introduction of each individual alanine mutation. The Ki for inhibition of [3H]granisetron binding by 5-HT was increased 110 fold by the Y142A mutation. A similar increase was also observed for mCPBG (160 fold). In contrast, the Y140A and Y152A mutations produced relatively small changes in the Ki. Since both compounds are 5-HT3R agonists, they are likely to share the same binding interactions. One of these interactions appears to be with Y142. Other studies have also identified binding site interactions for agonists. R91, E106, F107, W183 and several residues adjacent to the M1 region have all been demonstrated to alter the action of 5-HT and/or mCPBG . Within the crystal structure of the AChBP, homologous residues are located in the apparent binding site . The effects of agonist on the 5-HT3R are mediated by their interaction with these binding site residues, resulting in stable receptor conformations, including the channel open state. Identification of interacting amino acids and their location in the tertiary structure of the receptor may provide clues to the mechanism of channel opening. For example, amino acids homologous to Y142 and W183 (R104 and W143 respectively) are in close proximity in the AChBP, although on complementary faces of the receptor subunits . Spier et. al. have suggested that W183 may be involved in a cation-pi interaction with the amino group of 5-HT and mCPBG . If W183 and Y142 are located near each other but on opposite faces of the binding site in the 5-HT3R, then they could potentially act in concert to help stabilize a conformation of the receptor leading to channel opening. d-Tubocurarine inhibition was uniquely altered by the Y140A mutation. An increase in Ki of 50 fold was observed for d-tc on this mutant. Little if any change resulted for any other mutation. Thus, while d-tc may form an interaction with the receptor at this binding loop, it appears to interact with Y140 rather than Y142 or Y152. This difference in binding site interactions of the antagonist d-tc and the agonists 5-HT and mCPBG could be the result of a slightly different positioning of d-tc in the binding site. The antagonists d-tc and [3H]granisetron also interact differently with this region since [3H]granisetron binding was not altered by mutations at any of the amino acids tested. While the binding of d-tc appears to involve some interaction with this binding loop, the binding of [3H]granisetron does not. Previous studies have indicated a point of overlap between [3H]granisetron and d-tc at W89 of the 5-HT3ASR. . These studies indicate that the binding regions for these antagonists are partially overlapping at W89 but not at Y140. In addition to Y140 and W89, d-tc has also been observed to interact with D206 and several other residues in loop C . The d-tc binding site appears to involve interaction with at least three different binding loops (A, E and C) although additional interacting amino acids may be found. Data obtained for d-tc is particularly valuable due to the rigid nature of this molecule. Determination of interacting functional groups and their relationship to individual amino acids could enable d-tc to be used as a molecular ruler to determine relative positions of these amino acids. Lerisetron inhibition of [3H]granisetron binding to mutant receptors was also investigated. Lerisetron is a potent 5-HT3R antagonist first synthesized by Orales et. al.. As was observed for 5-HT3R agonists, lerisetron binding was altered by the Y142A mutation. This mutation produced an increase in the Ki of 160 fold compared to wildtype receptors. This increase in Ki indicates an important interaction of lerisetron with Y142 and a similarity between the binding location of lerisetron, 5-HT and mCPBG. Lerisetron is the only antagonist tested that shares a binding site interaction in this region with agonists. In contrast to 5-HT and mCPBG, however, lerisetron also interacts with Y152 as indicated by the 190 fold increase in Ki on Y152A mutant receptors. Since neither d-tc nor [3H]granisetron interacts with Y142 or Y152, there appears to be a difference between the interactions formed by these antagonists compared to lerisetron. The binding sites of the three antagonists tested differ with respect to this binding loop. Since the role of a competitive antagonist is simply to block the binding of agonists and prevent channel opening, the specific amino acids that interact with the ligand can vary for different antagonists. This is less likely to be the case with agonists since they must produce a conformational change in the protein to exert their effects. Only two mutant receptors failed to bind [3H]granisetron; G147A and V149A. G147 is the conserved glycine in this putative binding loop. Chirara et. al. have suggested that the highly conserved nature of the glycine in this region may indicate the existence of a loop structure consisting of either gamma or a loose three residue-turn in the nAChR . Either of these turns would bring the two putative beta-strands together such that gammaL109, gammaY111 and gammaS115 and gammaY117 are all on the same side of an antiparallel beta-sheet. These residues have been identified by affinity labeling, site-directed mutagenesis or cysteine substitution to lie on the same surface. A classic 2-residue beta-turn would place these residues on opposite surfaces . Substitution of the conserved glycine by alanine may disrupt the structure of this region and prevent assembly or expression of the receptor. The recent determination of the crystal structure of an AChBP supports this hypothesis. The AChBP displays a large amount of homology to the amino terminal of LGIC receptors and thus may be similar in structure . The crystal structure of this protein reveals a loose 3 residue turn incorporating the conserved glycine residue . Homologous residues in other LGIC subunits have also been identified and are shown in Figure . The residues identified in this study as altering binding affinity of 5-HT3R ligands would also be present on the same surface if this structure is present in the 5-HT3R. While Y140 lies somewhat outside the region identified by Chiara in the nAChR (homologous to gammaN107), gammaL109 and gammaL119 are homologous to Y142 and Y152 of the 5-HT3ASR. The ability of lerisetron to interact with both Y142 and Y152 also supports the hypothesis that these two amino acids are present in a loop structure since the eight intervening residues would position Y142 and Y152 too far apart to permit them both to interact with a single ligand even if they were interacting with functional groups on opposite ends of the molecule. A loop structure would bring them into closer proximity and permit interaction with the small molecule lerisetron. All three tyrosine mutations were investigated using a whole cell patch clamp assay to determine if functional changes could be observed. Whole cell responses could not be obtained for Y140A or Y142A, although specific binding of [3H]granisetron was observed. These data suggest that, while the receptors do assemble and are capable of binding [3H]granisetron, they are either not transported to the cell surface or are non-responsive to 5-HT at concentrations of 1 mM or less. Y152A does produce functional channels however they display distinctly altered response kinetics when compared to wildtype receptors. Y152A responses do not show the rapid rise times observed in wt receptors. The extremely slow rise times observed for Y152A receptors may indicate a change in rate constants preceding channel opening. These changes in the rate constants for either agonist binding or channel opening also produce a 140 fold decrease in the observed EC50 for 5-HT activation. The slow rise is followed by a non-desensitizing phase of the response that is dramatically different from the fast desensitization observed for wt 5-HT3ASRs. Lack of desensitization could result from either a stabilization of the open state of the channel or a destabilization of the desensitized state. Mutations of homologous or nearby residues in both the nAChR and GABAA receptors have also been demonstrated to alter the agonist response. Mutation of the homologous residue in the GABAA receptor gamma-subunit (T142) to serine altered the efficacy of the agonist Flumazenil, converting it to a partial agonist . In the nAChR, mutation of mouse and rat epsilonP121 to leucine altered both the binding of acetylcholine and the stability of the open state of the channel. epsilonP121 is homologous to P154 in the 5-HT3R and is only two residues away from Y152. The authors of this study concluded that this portion of the acetylcholine binding site was closely linked to the channel opening region of the receptor . It is reasonable to conclude that the homologous region in the 5-HT3R may perform a similar function. The link between an agonist binding domain and a conformational change leading to channel opening is not unexpected since the two must obviously be linked. If binding to this region of the receptor is shown to be a critical step between the binding of agonists and the opening of the channel, further investigation of the amino acids in this loop may provide valuable clues to molecular basis of this process. Conclusions : Our data indicate an important role for this putative binding site loop in the interaction of the 5-HT3R with different ligands and illustrate the difference in binding of different structural classes of ligands. Each structural class shows different patterns of interaction with amino acids in this region of the receptor. [3H]granisetron does not appear to interact with any of the amino acids tested while d-tc interacts with only Y140, mCPBG and 5-HT with Y142 and lerisetron with both Y142 and Y152. Similar selective effects have been observed on other residues including W183, W89, F107 and E106 among others. Our data in conjunction with those of other laboratories indicates the differences in orientation of different ligands within the same binding cleft. These differences in orientation result in different amino acid/functional group interactions. As the structural detail of these interactions emerges, these differences could potentially be exploited to produce more potent and specific ligands. For example construction of a "hybrid" ligand that combines the interactions of [3H]granisetron with those of d-tc, 5-HT or mCPBG could produce an antagonist with increased affinity due to the additional binding energy of these interactions and a greater specificity since it would utilize more structural features of the binding site. The requirement of a glycine at position 147 and the ability of lerisetron to interact with both Y142 and Y152 also support the hypothesis that the secondary structure in this region of the receptor is formed by a loop structure. If the loose 3 residue turn proposed by Chiara et. al. and shown for the AChBP is present in the 5-HT3R, then Y140, Y142 and Y152 would be present on the same side of the sheet and all three would be capable of extending into the binding site. A similar observation has been made for the nAChR where gammaS111, gammaY117, gammaL119, deltaR113 and deltaT119 of the mouse nAChR receptor and gammaL109 and gammaY111 of the torpedo nAChR are all thought to be present in the binding site . Materials and Methods : Materials | [3H]granisetron was purchased from New England Nuclear, 5-HT from Spectrum, and mCPBG and d-tc from Research Biochemical International. Lerisetron was provided by Dr. Karen Kirschbaum at The University of Louisiana at Monroe, Monroe, LA. All other reagents were obtained from commercial sources. Site directed mutagenesis | Wild type 5-HT3AS mouse receptor cDNA was obtained from Dr. Michael White . Mutant receptors were constructed using either the Quick Change Mutagenesis kit (Stratagene) or the Altered Sites Mutagenesis kit (Promega). All mutations were confirmed by commercial DNA sequencing. Cell culture and transfection | tsA201 cells, a derivative of HEK293 cells, were seeded at a density of 5 x 106 cells/100 mm dish. Cells were grown in DMEM medium containing 10% FBS, 100 units/ml penicillin/streptomycin for nine hours in 5% CO2 and transfected with 10 mug mouse 5-HT3ASR cDNA per 100 mm dish using the calcium phosphate technique (New Life Technologies, NY). Media was changed 12 --14 hrs after transfection. The cells were allowed to grow for another 24 hours and then harvested. For whole cell patch clamp experiments, tsA201 cells were seeded to a density of 0.25 x 106 cells/60 mm dish. Cells were grown in DMEM culture medium containing 10% FBS and 100 units/ml penicillin/streptomycin for 12 hours prior to transfection. Transfections were performed using Superfect Transfection Reagent (Qiagen, CA). Ten mug of cDNA were mixed with DMEM medium containing no serum or antibiotics in a volume of 150 mul. Twenty mul of Superfect reagent were then added and the mixture incubated at room temperature for 15 min. The reaction was terminated by adding 1 ml of DMEM medium containing 10% FBS and 100 units/ml penicillin/streptomycin and the entire mixture added to cells in the 60 mm dish. Cells were exposed to Superfect Reagent for 3 hours. At that time, the reagent was replaced with DMEM medium containing 10% FBS and 100 units/ml penicillin/streptomycin and incubated for an additional 24 hours prior to use. Binding assays | Transfected cells were scraped from the dishes, washed twice with Dulbecco's PBS (New Life Technologies, NY), then resuspended in 1.0 ml ice cold PBS/100 mm dish. Cells were either used fresh or frozen at this step until needed. Immediately prior to use, cells were homogenized on ice in PBS using a glass tissue homogenizer then centrifuged at 35 000 x g for 30 minutes in a Beckman JA20 rotor (4C). Membranes were washed once more with PBS at 4C then resuspended in 1 ml PBS/100 mm dish. Protein content was determined using a Lowry assay (Sigma. Diagnostics, St. Louis, MO). Membranes were initially prepared and Bmax and Kd values determined in the presence and absence of a cocktail of protease inhibitors (Complete Protease Inhibitor Cocktail, Roche Diagnostics, Mannheim Germany). No change in Bmax was observed as a result of omitting the protease inhibitor (Bmax (+ protease inhibitor)= 3.1 +- 0.11 pmoles/mg protein vs Bmax (- protease inhibitor) = 3.2 +- 0.13 pmoles/mg protein.) hence all assays were performed in the absence of the cocktail. Binding assays were performed in PBS. For Kd determinations, 100 mul of homogenate was incubated at 37C for 1 hour with varying concentrations of [3H]granisetron (NEN, MA). Specific binding of [3H]granisetron was determined as the bound [3H]granisetron not displaced by a saturating concentration of a competing ligand. Kd values were determined by fitting the binding data to the following equation using Graphpad PRISM (San Diego CA): B = Bmax [L]n / ([L]n + Kn), where B is bound ligand, Bmax is the maximum binding at equilibrium L is the free ligand concentration and n is the Hill coefficient. Ki determinations, 100 mul of homogenate was incubated at 37C for 2 hours with varying concentrations of inhibitor and [3H]granisetron (NEN, MA). Binding was terminated by rapid filtration onto a GF/B filters. The IC50 values were calculated by fitting the data to the following equation using Graphpad PRISM (San Diego CA): theta = 1/ (1+(L/IC50)), where theta is the fractional amount of [3H]granisetron bound in the presence of inhibitor at concentration L as compared to the amount of [3H]granisetron bound in the absence of inhibitor. IC50 is the concentration at which theta = 0.5. The Ki is calculated from the IC50 value using the Cheng-Prusoff equation. Electrophysiological Recordings | Transfected cells were transferred to a recording chamber containing extracellular solution (140 mM NaCl, 1.7 mM MgCl2, 5 mM KCl, 1.8 mM CaCl2, 25 mM HEPES pH 7.4). Patch electrodes of resistance 2.5 --3.0 MOmega were filled with filtered intracellular solution containing 145 mM KCl, 2 mM MgCl2, 1 mM EGTA, 25 mM HEPES (pH 7.4). Cells were clamped in whole cell configuration at a holding potential of -60 mV. A continuous extracellular solution flow (0.8 ml/min) was maintained throughout the recording procedure. 5-HT was dissolved in extracellular solution and delivered to cells using a rapid perfusion system (Warner Instruments, Hamden, CT) at a rate matching the extracellular solution flow rate. The drug perfusions lasted for a period varying from 4 to 8 seconds. Currents elicited by agonist application were measured using an Axopatch 200 B amplifier (Foster City, CA). The data were plotted and analyzed by non-linear curve fitting (Graphpad PRISM, San Diego CA) according to the following equation: I = 1/(1 +(EC50/ [C])n), where is the normalized current at 5-HT concentration [C], EC50 is the concentration of 5-HT needed to obtain half maximal activation and n is the apparent Hill coefficient. Backmatter: PMID- 12086585 TI - Insertion of a small peptide of six amino acids into the beta7 --beta8 loop of the p51 subunit of HIV-1 reverse transcriptase perturbs the heterodimer and affects its activities AB - Abstract | Background | HIV-1 RT is a heterodimeric enzyme, comprising of the p66 and p51 subunits. Earlier, we have shown that the beta7-beta8 loop of p51 is a key structural element for RT dimerization (Pandey et al., Biochemistry 40: 9505, 2001). Deletion or alanine substitution of four amino acid residues of this loop in the p51 subunit severely impaired DNA binding and catalytic activities of the enzyme. To further examine the role of this loop in HIV-1 RT, we have increased its size such that the six amino acids loop sequences are repeated in tandem and examined its impact on the dimerization process and catalytic function of the enzyme. Results | The polymerase and the RNase H activities of HIV-1 RT carrying insertion in the beta7-beta8 loop of both the subunits (p66INS/p51INS) were severely impaired with substantial loss of DNA binding ability. These enzymatic activities were restored when the mutant p66INS subunit was dimerized with the wild type p51. Glycerol gradient sedimentation analysis revealed that the mutant p51INS subunit was unable to form stable dimer either with the wild type p66 or mutant p66INS. Furthermore, the p66INS/p66INS mutant sedimented as a monomeric species, suggesting its inability to form stable homodimer. Conclusion | The data presented herein indicates that any perturbation in the beta7-beta8 loop of the p51 subunit of HIV-1 RT affects the dimerization process resulting in substantial loss of DNA binding ability and catalytic function of the enzyme. Keywords: Background : Human immunodeficiency virus type-1 reverse transcriptase (HIV-1 RT) is a product of the gag-pol polyprotein precursor, which is subsequently cleaved by the pol-encoded protease to yield the active form of the enzyme . This multifunctional enzyme is responsible for copying the single stranded viral RNA genome into double stranded proviral DNA . HIV-1 RT is a heterodimer consisting of a 66 and 51 kDa polypeptide chain designated as p66 and p51, respectively. The p51 subunit is generated via endoproteolytic cleavage of the p66 subunit between Phe 440 and Tyr 441 . The larger subunit (p66) contains both polymerase and RNase H activities, while the smaller subunit (p51) lacks these functions, in context of the heterodimer . However, both the p66 and p51 monomers are functionally inactive when dissociated from each other . Several years have passed since it was first suggested that agents that could specifically disrupt the dimerization of HIV-1 RT might prove a worthwhile antiretroviral strategy , though such agents have yet to be developed. Despite the fact that p51 shares an identical amino acid sequence with the N-terminal portion of p66, the two subunits assume different global folding patterns in the formation of the asymmetric heterodimer . Structural determination through X-ray crystallography has revealed that the p66 subunit of HIV-1RT has its polymerase domain in an "open" conformation, with its subdomains forming a large cleft which accommodates DNA. In contrast, the p51 subunit assumes a compact folded conformation that causes the active site residues in this subunit to be buried and therefore, nonfunctional . It has been proposed that the open conformation of p66 is supported by interactions with a closed and compact p51 molecule . The two subunits interact mainly via their connection subdomains. Additional contacts, between the thumb subdomain of p51 and RNase H subdomain of p66 are also substantial . Although there have been conflicting reports regarding the DNA polymerase activity of recombinant preparations of the p51 homodimer , it has become clear that p51 mainly plays a supportive role in context of the p66/p51 heterodimer. Assembly of chimeric heterodimers formed by mixing subunits of HIV-1 RT and FIV-1 RT, has demonstrated that the p51 subunit of HIV-1 RT helps to preserve the functional integrity of the HIV-1 RT heterodimer . Despite the fact that several functions have been proposed for the p51 subunit, the mechanism whereby p51 performs these functions has remained largely undefined. Some of the proposed functions for p51 include: (i) stabilizing the t-RNA primer binding for the initiation of reverse transcription , (ii) enhancement of strand displacement DNA synthesis , and (iii) as a processivity factor in DNA synthesis . Cys->Ser mutation at position 280 in the p51 subunit has been shown to alter the RNase H activity of the heterodimeric enzyme, indicating that this residue in the thumb subdomain of p51 plays an important role in support of the RNase H activity of p66 . The emergence of a strain of HIV-1 resistant to the non-nucleoside RT inhibitor TSAO (Tertbutyldimethyl silyspiro amino oxathioledioside) displaying Glu->Lys mutation at position 138 in the p51 subunit of HIV-1 RT has also been reported , thus implicating p51 to play a more direct role in drug binding and/or the enzymatic activities of HIV-1RT. This report was initially surprising, since Glu138 of p51 was thought to be quite distant from the purported dNTP-binding pocket of HIV-1RT, as well as the NNRTI binding pocket. However, in light of our recent findings implicating this loop region of p51 as a critical structural element supporting the catalytic functions of p66, it seems feasible that mutation at position 138 in p51 effectively altered the binding of TSAO through its influence on the p66 catalytic subunit . Examination of the crystal structure of HIV-1 RT reveals the presence of a small groove like region on the floor of the polymerase cleft of p66 . The beta7-beta8 loop of p51, comprising of six amino acids denoted as SINNET appears to fit into this groove-like region and likely stabilizes the polymerase domain of p66. In an earlier communication, we have shown that the p51 subunit of HIV-1 RT is required to load the p66 subunit on to the template primer for DNA synthesis . Our recent studies indicate that the beta7-beta8 loop of the p51 subunit is essential for the catalytic function of the p66 subunit. Deletion of this loop or substitution of four amino acid residues with alanine within the beta7-beta8 loop of p51 severely impaired the DNA polymerase activity of the enzyme as a consequence of the inability of the enzyme to form stable dimers . These findings clearly establish the absolute requirement of the beta7-beta8 loop of p51 for RT dimerization. Nonetheless, the question regarding the optimal size and composition requirement of this loop for efficient dimerization remains unanswered. In the present article, we have addressed the impact of increasing the size of the beta7-beta8 loop on the dimerization process. As a preamble to these studies, we have increased the size of this loop by repeating its six amino acid sequence in tandem. The rationale for duplicating the loop sequence was to increase the size of this loop without significantly disrupting the interactions seen with the wild type beta7-beta8 loop. The resulting mutant derivatives of HIV-1 RT containing insertion of six amino acids in the beta7-beta8 loop in either or both the subunits were analyzed for their ability to form stable dimers and other biochemical characteristics. In this article, we present evidence that HIV-1 RT mutants, carrying insertion of six amino acids in the beta7-beta8 loop specifically in the second subunit, do not form stable dimers. This inability to dimerize substantially decreases the enzymes affinity for DNA consequently impairing its polymerase and RNase H activities. Results : Glycerol gradient ultra-centrifugation analysis | The ability of the HIV-1 RT mutants, carrying insertion of six residues in the beta7-beta8 loop in either one or both the subunits was analyzed by glycerol gradient sedimentation analysis. Fractions of 200 muL were collected from the bottom of the tube and aliquots of every third fraction were subjected to SDS PAGE. The results shown in Figure , indicate that the sedimentation peak for the wild type p66/p66 homodimer was between fractions 16 --19 of the gradient (Panel A), whereas the wild type p51 species sedimented between fractions 22 --28, as a monomeric protein (Panel C). The sedimentation pattern for the p66WT/p51INS mutant indicated two distinct peaks. While the p66WT subunit predominantly sedimented in between fractions 16 --19, at the predicted position of the homodimer, the p51INS subunit sedimented between fractions 22 --28, indicating a monomeric conformation (Panel D). This sedimentation profile indicates that the p66WT/p51INS mutant carrying the 6 amino acid insertion in the beta7-beta8 loop of its p51 subunit is unable to form stable heterodimer. Under these conditions, the p66WT subunit of the p66WT/p51INS mutant tends to homodimerize. The mutants, p66INS/p66INS (Panel B) and p66INS/p51INS (Panel F), exhibited a distinct sedimentation peak between fractions 22 --28 of the gradient, indicating that these enzymes are unable to form stable dimers. In contrast, the p66INS/p51WT mutant (Panel E), in which only the p66 subunit carried an insertion in the beta7-beta8 loop sedimented between fractions 16 --19 of the gradient, indicating that these two subunits can form stable heterodimers. These analyses imply that the proper size of the beta7-beta8 loop in the second subunit of HIV-1 RT is important for the formation of a stable dimeric enzyme. Figure 1 | (A) Sedimentation profile of the wild type and mutant HIV-1 RT and (B) Polymerase activity profile of the glycerol gradient fractions. (A) Sedimentation profile of the wild type and mutant HIV-1 RT. Mutant enzymes carrying 6 amino acids insertion in the beta7-beta8 loop of either or both the subunits were applied on 10 --30% linear glycerol gradient (5 mL) and centrifuged at 48,000 rpm in SW 50.1 rotor for 22 hours. Gradients were fractionated from the bottom and subjected to SDS polyacrylamide gel electrophoresis and Coomassie Blue staining. (B) Polymerase Activity profile of the glycerol gradient fractions. Every third fraction between 7 and 33 of the glycerol gradient was diluted 10-fold and analyzed for its polymerase activity on poly (rA)(dT)18 as described under Materials and Methods. The reaction products were resolved on an eight percent denaturing poly-acrylamide urea gel and subjected to PhosphorImaging. In order to correlate the sedimentation profile of these insertion mutants with their functional activity, we analyzed the polymerase activity in the various gradient fractions. These results are presented in Fig. . The polymerase activity profile of the gradient fractions of the wild-type p66/p66 and the p66INS/p51WT mutant revealed major polymerase activity peaks corresponding to fractions 16 --19 . This activity peak correlates with the protein band intensity seen in Fig. (panels A and E) and is also in agreement with the sedimentation pattern of these two enzymes. Interestingly, the activity profile of the p66WT/p51INS mutant also yielded a peak corresponding to gradient fractions 16 --19 , thus substantiating our contention that the p66WT subunit of the p66WT/p51INS mutant tends to self-dimerize and form the catalytically active p66 homodimer. The wild type p51 and the two mutants, p66INS/p66INS and p66INS/p51INS, the sedimentation profile of which indicated a monomeric conformation were conspicuously devoid of any polymerase activity . These results imply that the beta7-beta8 loop of the second subunit of HIV-1 RT is critical in forming functionally active dimeric enzyme. DNA polymerase activities of wild type HIV-1 RT and its insertion mutants | DNA polymerase activity of the wild type HIV-1 RT and its mutant derivatives were quantitatively determined on both heteropolymeric RNA (U5-PBS RNA) and DNA (49-mer) templates, primed with 32P labeled 17-mer PBS primer. As documented in Table , insertion of the six amino acid segment in both the subunits (p66INS/p66INS or p66INS/p51INS), resulted in significant loss of polymerase activity corresponding to greater than 90% reduction. Interestingly, when the p66INS mutant was dimerized with the wild type p51 (p66INS/p51WT), the polymerase activity was restored to wild type levels. The same result was obtained when p66INS was dimerized with p51 having Asp->Ala mutation at amino acid position 186 (one of the catalytically crucial carboxylate triad) but having an intact beta7-beta8 loop. The rationale for using p51D186A mutant was to ascertain that the wild type polymerase activity observed with p66INS/p51WT was not due to residual polymerase activity of p51. These results suggest that insertion of six residues in the beta7-beta8 loop of p51 but not in p66 is detrimental to the function of the heterodimeric enzyme. Table 1 | Effect of duplication of the beta7 --beta8 loop in either or both the subunits of HIV-1 RT on the polymerase activity of the enzyme As shown in figure , evaluation of the polymerase activity of the wild type HIV-1RT and its insertion mutants by primer extension assay generally mirrored the results observed in the TCA precipitation assay, summarized in Table . Insertion of six amino acid residues in the beta7-beta8 loops of both the subunits (p66INS/p66INS or p66INS/p51INS) resulted in significant impairment of polymerase activity of the mutant enzymes. On an RNA template, both these enzymes exhibited total lack of polymerase activity , though the p66INS/p51INS mutant exhibited residual primer extension capability on a DNA template , while the p66INSp/66INS mutant was inactive on this template. Consistent with the results of the TCA precipitation assay, both the mutants, p66INS/p51WT and p66INS/p51D186A, exhibited near wild type (p66WT/p51WT) polymerase activity on both RNA and DNA templates. Thus, the impairment of the polymerase activity seen with these mutants carrying insertion in either both the subunits or specifically in the second subunit may be related to the inability of these mutants to form stable dimers resulting in a closed polymerase cleft. Figure 2 | Analysis of polymerase products catalyzed by insertion mutants of HIV-1 RT. Analysis of polymerase products catalyzed by insertion mutants of HIV-1 RT. Primer extension reactions catalyzed by the wild type and mutant enzymes were on DNA (panel A) and RNA (panel B) templates, primed with 5'-32P labeled 17-mer primer. Each set of reactions was carried out for 30 seconds (lane 1) and 60 seconds (lane 2) at 37C and quenched by the addition of equal volume of Sanger's gel loading dye. The reaction products were resolved on an 8% polyacrylamide-7M urea gel and subjected to PhosphorImager analysis. Effect of insertion in either or both the subunits of HIV-1 RT on the DNA binding function of the enzyme | Earlier we have shown that substitution of four amino acids on the beta7-beta8 loop with alanine reduced the DNA binding ability of the enzyme. Similar results were obtained when four amino acids were deleted from the loop. We therefore, concluded that the effect was exerted via p51 since alanine substitution or deletion specifically in the p66 subunit had no effect on DNA binding and polymerase activity of the enzyme. We speculated that alanine substitution or deletion in the loop of the p51 subunit may shorten the interacting sphere of the loop which may not be able to induce opening of the polymerase cleft in p66, which is essential for DNA binding. Given the fact that shortening of this loop in p51 prevented DNA binding, we were interested in examining the effect of six amino acids insertion on DNA binding. Therefore, we determined the equilibrium dissociation constants (Kd) of E-TP binary complexes for the wild type enzyme and its mutant derivatives by gel mobility shift assay. For this purpose, we used a 33-mer heteropolymeric DNA template primed with 5' 32P-labeled 21-mer DNA. Results shown in figure and table indicate a 24-fold reduction in DNA binding affinity when both the subunits carried insertion (p66INS/p51INS) in their beta7-beta8 loops. Similar results were obtained with the p66INS/p66INS mutant. Interestingly, the DNA binding affinity was restored to the wild type levels, when the mutant p66 subunit was dimerized with the wild type p51 (p66INS/p51WT). We expected that the p66WT/p51INSmutant would also exhibit reduction in its affinity for DNA, since insertion of six residues in the p51 subunit was speculated to be detrimental for the dimerization process. However, the DNA binding affinity of the p66WT/p51INS mutant was similar to the wild type enzyme. Our glycerol gradient sedimentation analysis had revealed that the p66WT/p51INS mutant does not form stable dimers, rather the p66WT subunit tends to self dimerize. Thus, the wild type DNA binding affinity seen in case of the p66WT/p51INS mutant may be attributed to the presence of these p66WT/p66WT homodimeric species. Based on our observations from the sedimentation analysis, we propose that the loss of DNA binding function in case of the HIV-1RT insertion mutants results from a failure to form stable dimers. The wild type p51 alone exhibited very low DNA binding affinity with a Kd[DNA]1300-fold higher than the wild type heterodimeric enzyme. These results suggest that the optimal size of the beta7-beta8 loop in the second subunit of the HIV-1 RT dimer is crucial for opening the polymerase cleft of the p66 subunit. Figure 3 | DNA binding affinity of mutant HIV-1 RT carrying insertion in the beta7 --beta8 loop of either or both the subunits. DNA binding affinity of mutant HIV-1 RT carrying insertion in the beta7 --beta8 loop of either or both the subunits. The 49-mer DNA primed with 5'-32P labeled 21-mer DNA primer was incubated with varying concentrations of the individual enzyme at 4C for 10 min. The mixture was electrophoresed under non-denaturing conditions on a 6% (w/v) polyacrylamide gel and analyzed on phosphorImager (Left panel). The positions of the free template primer (TP) and enzyme bound template-primer (E-TP binary complex) are indicated for the wild type enzyme. The percent of E-TP complex formed as a function of enzyme concentration was plotted for determining the Kd values (Right panel). Table 2 | DNA and dNTP binding affinities of the mutant HIV-1 RT carrying insertion in the beta7 --beta8 loop of either or both the subunits Ternary complex formation by the wild type and mutant enzymes | In the crystal structures of the ternary complex of HIV-1 RT (E-DNA-dNTP) the finger subdomain moves by 20A towards the palm subdomain . In this finger closing conformation, the DNA is locked in a stable ternary complex poised for catalysis. An in vitro assay using dideoxy terminated primer annealed with the template which allows the next correct dNTP to bind in the ternary complex without actual DNA synthesis has recently been reported . Using this assay system, we have evaluated the ability of the insertion mutants to form the ternary complexes and the effect of DNA trap on such complexes. Since binding of dNTP to the enzyme is an ordered mechanism which occurs only after DNA binding, the extent of labeled TP remaining bound to the enzyme in the presence of dNTP and DNA trap represents the extent of ternary complex formed. The E-TP binary complex was formed at enzyme concentrations which binds 100% of the labeled template primer. The preformed E-TP complex was then incubated in the presence of next correct dNTP followed by addition of 300-fold molar excess of unlabeled TP as the DNA trap. We found that E-TP binary complex was completely competed out by the DNA trap (data not shown) while a significant amount of the E-TP binary complex converted to E-TP-dNTP ternary complex was resistant to competition with DNA trap suggesting the stability of the ternary complex. Table lists the apparent dNTP binding affinity for the WT enzyme and its insertion mutants determined from data shown in Fig. . It was observed that although the DNA binding affinity was severely affected in case of the p66INS/p51INS mutant, its apparent dNTP binding affinity in the ternary complex did not change with respect to the wild type enzyme. However, the p66INS/p66INS mutant was unable to form a ternary complex. These data suggest that these two mutants may have a different conformation and mode of interaction in the ternary complex. It is apparent that the p66INS/p66INS binds to TP in a nonproductive manner which may have a direct impact on dNTP binding in the ternary complex. Figure 4 | Apparent dNTP binding affinity of mutant HIV-1 RT carrying insertion in the beta7 --beta8 loop of either or both the subunits. Apparent dNTP binding affinity of mutant HIV-1 RT carrying insertion in the beta7 --beta8 loop of either or both the subunits. The 33-mer DNA primed with 5'-32P labeled dideoxy (ddC) terminated 21-mer DNA primer was incubated with the individual enzyme at 4C for 10 min. The binding affinity [Kd (dNTP)] of the wild type enzyme and its mutant derivatives in the ternary complex was determined by incubating the preformed E-TP binary complex of the individual enzyme species at different concentrations (0.26 --800 muM) of the next correct dNTP (dGTP). Following incubation with dGTP, 300 fold molar excess of the unlabeled TP was used as trap to remove the readily dissociable binary complexes from the ternary complex population. The extent of stable ternary complexes formed were resolved on 6% native polyacrylamide gel and analyzed by phosphorImaging. The [Kd [dNTP]], in the ternary complex for each enzyme was determined by quantifying the E-TP-dNTP ternary complexes as a function of increasing concentrations of dGTP and fitting the data to the equation for single-site ligand binding using Sigma-Plot. Lanes 1 --6 represent the ternary complex formed in the presence of dGTP at 0.2, 1.28, 6.4, 32, 160 and 800 muM, concentrations. Steady state kinetic analysis of HIV-1 RT and its insertion mutants | In order to determine whether alteration in DNA binding without any change in the apparent dNTP binding affinity of the insertion mutants is consistent with their kinetic parameters, we analyzed the steady-state kinetic parameters of these mutants. The results of this investigation are summarized in Table . On poly (rA).(dT)18, only the p66INS/p66INS mutant showed a significant increase in Km [dNTP]. This observation is in agreement with the apparent dNTP binding affinity data in the ternary complex, where p66INS/p66INS mutant was found to be defective in forming a productive ternary complex. This observation is also consistent with our suggestion that p66INS/p66INS binds nonproductively to TP that may influence the formation of ternary complexes. Interestingly, the p66INS/p51INSmutant carrying insertion in both the subunits did not display the same reduction in dTTP binding affinity. However, the p66INS/p66INS and p66INS/p51INS mutants displayed nearly 6,000-fold and 400-fold reduction in catalytic efficiency (kcat/Km) compared to their wild type counterparts, respectively, on this template primer. A 10-fold reduction in catalytic efficiency in case of the p66INS/p51WT was noted only on poly (rA). (dT)18 and may be template-primer specific. None of the enzymes displayed a significant reduction in Km [dNTP] when the heteropolymeric DNA\DNA template primer was used, although the p66INS/p66INS and p66INS/p51INS mutants exhibited drastic reduction in catalytic efficiency. The p66WT/p51INSdisplayed no change in either dNTP binding or catalytic efficiency. These results are in keeping with our analysis of the polymerase and DNA binding assay for this mutant. Once again, we believe that the wild type p66 subunit in this enzyme preparation tends to homodimerize since the p51INS fails to participate in stable dimer formation. This phenomenon masks the deleterious effect of the insertion mutation. Table 3 | Steady-State kinetic parameters of mutant HIV-1 RT carrying insertion in the beta7 --beta8 loop of either or both the subunits RNase H activity of the insertion mutants | Since the polymerase activity of the homo- and hetero-dimeric enzymes carrying insertion in the beta7-beta8 loop in both the subunits (p66INS/p66INS and p66INS/p51INS) was drastically impaired, it was of interest to examine how this insertion affects their RNase H activity. To evaluate this, we employed a 30-mer RNA-DNA hybrid, and examined the cleavage of the 5'-32P-RNA strand of the duplex by the wild type enzyme and its mutant derivatives. The result of this analysis is presented in Fig. . Similar to their polymerase activities, the RNase H activities of the p66INS/p66INS and p66INS/p51INS mutants were severely impaired. This is not surprising, since our analysis of the DNA binding function of these two mutants had indicated a substantial loss of DNA binding affinity, which in turn is expected to affect both the polymerase and RNase H functions. Dimerization of the p66INS subunit with the wild type p51 (p66INS/p51WT) carrying an intact beta7-beta8 loop resulted in substantial recovery of the RNase H activity. The RNase H activity seen in case of the p66WT/p51INS mutant was not surprising since the p66WT subunit of the p66WT/p51INSmutant tends to self-dimerize and form the catalytically active p66 homodimer. Figure 5 | RNase H activity of the wild-type HIV-1 RT and its mutant derivatives. RNase H activity of the wild-type HIV-1 RT and its mutant derivatives. The individual enzymes were incubated with 5'-32P-labeled 30-mer RNA annealed with 30-mer complementary DNA strand, at 37C for 30 sec and 1 min as described in Materials and Methods. The cleavage products were resolved on an 8% denaturing polyacrylamide-urea gel and analyzed on a phosphorImager. Lanes 1 and 2 represent reactions carried out for 30 sec and 1 min, respectively. Discussion : In an earlier investigation on the role of the p51 subunit of HIV-1 RT, we demonstrated that decrease in size of its beta7-beta8 loop impairs the catalytic function of the heterodimer . In the present studies, we demonstrate that maintaining the wild type size of this loop in the p51 subunit is critical for dimerization of the enzyme and its catalytic activity. Duplication of the beta7-beta8 loop sequence selectively in the p66 subunit did not affect the dimer formation, DNA binding or polymerase activity of the p66INS/p51WT mutant. However, insertion of the same amino acid residues in the beta7-beta8 loop of p51 prevented stable dimerization of the p51INS subunit with either p66INS or p66WT and adversely impacted the DNA binding, polymerase and RNase H activities. Earlier, we have shown that p51 facilitates the loading of the p66 subunit on to the template primer . Therefore, the impaired polymerase activity and template-primer binding affinity of HIV-1 RT mutants carrying insertion in p51 may be due to their inability to load the catalytic p66 (p66INS) on the template primer. These altered biophysical/enzymatic properties of these insertion mutants may be attributed to the reduced dimer stability. Crystal structures of HIV-1 RT show that p66 and p51 assume different folding patterns and tertiary structures . It has been proposed that p66 in a monomeric form exists in a closed conformation similar to p51 . Following dimerization with another molecule of p51 (or p66), it assumes an open conformation . The polymerase domain of p51 is buried within its core. This difference in tertiary structures between the two subunits makes the dimer asymmetric. The amino acid residues at the contact interface differ with respect to their position and location in the 3-D structure. The counterpart of amino acid residues of p66 located at the contact interface are buried in the p51 folded conformation, whereas those of p51 are scattered in the p66 'open' conformation . Figure 6 | Amino acid residues in the p51 subunit in contact with the p66 subunit. Amino acid residues in the p51 subunit in contact with the p66 subunit. The Calpha backbone of p51 and p66 is shown in orange and green, respectively. Amino acids residues of p51 making contact with the p66 subunit are represented as 'sticks' and are shown encircled. The location of analogous residues in p66 are also encircled. It may be noted that these residues in p66 are scattered. The residues circled within 'blue' are from fingers; those in light blue' are from 'connection' and those in pink are from 'thumb' subdomains. The template (blue) and primer (gray) are also displayed in this figure. The residues in blue circle are V21, K22, P25, P52, E53, N57, T131, N136, N137, E138, T139, in light blue circle are P392, I393, Q394, E396, T397, T400, W401, N418, P420, L422, and those in pink circle are N255, Q258, V261, N265, V276, L283, T286, L289. Of the several domain interactions between p66 and p51, the beta7-beta8 loop of p51 is strategically positioned to interact with the residues on the floor of the palm subdomain of p66. It has been suggested that the stability of the dimer is related to the buried surface area between the two subunits . In the nevirapine-bound HIV-1 RT crystal structure, the total contact surface area between the subunits is approximately ~4600 A2. The two major contact regions between the subunits which provide it stability are their connection subdomains and the thumb of p51 and RNase H domain of p66. These contacts account for approximately two third of the total buried surface area. Interestingly, the marginal decrease in the total surface area due to deletion of four residues in the beta7-beta8 loop does not account for the dimer instability, thus suggesting that polar interactions of residues in the beta7-beta8 loop of p51 with the palm subdomain of p66 may play a role in conferring stability to the heterodimer. The observation that a single point mutation at L289 of p66, a residue not in direct contact with p51, also destabilizes the dimer indicates that other factors may also contribute towards dimer stability. In order to analyze the impact of inserting the six amino acid peptide in the beta7-beta8 loop of p51, we used the molecular modeling approach. A search in the database of known protein structures employing the 'loop-search' algorithm of SYBYL yielded 100 loops, of which only five were sterically permissible. The loop exhibiting the best homology was incorporated in the modeled structure . This loop of 9 amino acids (RFNAHGDVN) from the protein S. lectin formed a short anti-parallel two strand beta-sheet. This inserted loop lies in the vicinity of the palm subdomain of p66 and exhibits additional hydrophobic and polar interactions with residues in the palm subdomain of p66, not seen in the wild type structure. These additional interactions are expected to enhance the stability of the dimer. However, sedimentation analysis indicates that the insertion mutants form unstable dimer. This implies that the insertion may have altered the relative position of the subdomains in the two subunits thereby perturbing the dimer stability. In summary, the beta7-beta8 loop of p51 is an important structural element involved in imparting stability to the heterodimer and in opening the polymerase cleft of p66 for catalysis. Figure 7 | Molecular model of the inserted peptide in the beta7 --beta8 loop of the p51 subunit. Molecular model of the inserted peptide in the beta7 --beta8 loop of the p51 subunit. Molecular modeling of the extended loop was performed by the 'Loop Search' option of SYBYL version 6.5 (Tripos Associates, St. Louis, MO). A total of 100 loops, matching the Calpha distances constrain between the end points of the loop residues, were obtained. The loops were examined for 'bumps' with the neighboring protein structure. Selection of the appropriate loop was based on the least root mean square deviation and maximum homology with the inserted loop sequence. A loop satisfying these criteria containing two small anti-parallel beta-strands is shown in cyan. The side chains of amino acid residues of p66 are shown in green and those of p51 (wild type) are shown in orange. The side chains of interacting residues in the inserted peptide are colored violet and are indicated by their position number in parentheses. Interactions seen in the wild type loop are shown in blue dotted line while additional interactions are shown in black dotted lines. Materials and methods : PfuTurbo DNA polymerase and PCR reagents were obtained from Stratagene, Inc. Restriction endonucleases, DNA modifying enzymes and HPLC-purified dNTPs were purchased from Roche Molecular Biochemicals. Fast flow chelating sepharose (iminodiacetic-Sepharose) for immobilized metal affinity chromatography (IMAC), Phosphocellulose and Q-sepharose was purchased from Amersham Pharmacia Biotech. The alpha-32P-dNTPs and gamma-32P-ATP were purchased from Perkin Elmer life sciences. The DNA oligomers were synthesized at the Molecular Resource Facility at UMDNJ. All other reagents and chemicals were of the highest available purity grade and purchased from Fisher, Millipore Corp., Roche Molecular and Bio-Rad. Plasmid and clones | The expression vector pET-28a and E. coli expression strain BL21 (DE3) were obtained from Novagen. The HIV-1RT expression clones (pKK223-3 RT66 and pET-28a-RT51) constructed in this laboratory were used for PCR amplification and construction of the insertion mutants in the p66 and p51 subunits of HIV-1 RT. An HIV-RNA expression clone pHIV-PBS was a generous gift from Dr. M. A. Wainberg . Insertion of 6 amino acid residues in the beta7 --beta8 loop | The pKK-RT66 clone containing two unique restriction sites, Hpa1 and Stu1, at codons 136 and 140 in the RT coding region was used for insertion of 6 amino-acid residues in the beta7-beta8 loop of the p66 and p51 subunit. The pKK-RT66 clone was digested with HpaI restriction enzyme to generate a blunt end at codon 136. For insertion, two complementary pre-kinased 18-mer synthetic DNA oligos having the following sequences: 5'-ATA AAC AAT GAG ACA ATA-3 (sense strand) and 3'-TAT TTG TTA CTC TGT TAT-5' (antisense strand) were hybridized. The 18-mer duplex DNA encoding the insertion peptide (Ile-Asn-Asn-Glu-Thr-Ile) was ligated with Hpa1 digested pKK-RT66 in between codon 135 and 136. The positive clones were screened in E. coli HB101 by the absence of an Hpa1 site and the correct orientation of the insertion was confirmed by DNA sequencing. This construct expresses the p66+6aa subunit without His tag sequences. A His-tag at the N-terminal of the p66+6aa subunit was introduced by sub cloning the Bal-I and Hind III fragment of pKKRT66+6aa into pET-28a-RT66 expression cassette. A unique Sac I site was also introduced in pKK-RT66 template at codon 440. The construction of P51+6aa was carried out by removal of the 360 bp fragment from pKK-RT66+6aa by restriction digestion with SacI followed by re-ligation of the vector ends. The insertion mutant in pET28a and pKK223-3 vectors were introduced into E. coli BL-21 (DE3) pLys S and E. coli JM109, respectively, for expression. Induction of the enzyme protein was carried out as described before for the wild type HIV-1RT . The enzyme with the hexahistidine-tag was purified from bacterial lysates by immobilized metal affinity chromatography , while non-hexahistidine-tagged enzyme was purified using the phosphocellulose and Q-Sepharose columns as described previously . Preparation of the heterodimeric enzyme with subunit specific insertion | The p51 subunit with a hexahistidine-tag and a non-tagged p66 were used to generate the heterodimers containing insertion in either or both of the subunits. For each set of heterodimers, 260 mug of p51 was mixed with 660 mug of p66 in the buffer containing 50 mM Tris HCl, pH 7.8, 60 mM KCl and 5 mM MgCl2. The rationale for using a 1:3 ratio of p51 to p66 was to saturate the His-tagged p51 with the non-tagged p66, ensuring heterodimer formation and eliminating excess p66 during IMAC purification. The mixture was incubated for 16 hours at 4C and applied to (0.5 mL) Ni2+ iminodiacetic-Sepharose (IDA-Sepharose) column, which was pre-equilibrated with the binding buffer (20 mM Tris HCl pH 7.8, 500 mM NaCl and 5 mM Imidazole). The column was washed with 15 mL of the same buffer to remove the excess of p66 that was not dimerized with p51 bound to the IDA-sepharose column. The heterodimeric RT was then eluted from the column with elution buffer (20 mM Tris HCl pH 7.8, 500 mM NaCl and 250 mM imidazole). Fractions of 0.5 mL were collected and an aliquot of each fraction was analyzed by SDS-PAGE using Coomassie Blue stain. The fractions containing approximately equal band intensity of p66 and p51 were dialyzed against a storage buffer (50 mM Tris HCl pH 7.0, 200 mM NaCl and 50% Glycerol) and this enzyme preparation was used in all experiments. Glycerol gradient ultra centrifugation | Fifty micrograms of the enzyme protein in 100 muL of buffer (50 mM Tris HCl, pH 7.8, 1 mM DTT and 400 mM NaCl) was carefully loaded onto 5 mL of 10 --30% glycerol gradients prepared in the same buffer. The gradients were centrifuged at 48,000 rpm in an SW48 rotor for 22 h at 4C. Fractions (200 muL) were collected from the bottom of the tube and aliquots of these fractions were electrophoresed using SDS PAGE and Coomassie Blue stain to identify the protein peak. The polymerase activity in the gradient fractions were analyzed by extension of the labeled (dT)18 annealed to poly (rA) template. Every third fraction between 7 and 33 of the glycerol gradient was diluted 10-fold and analyzed for its polymerase activity. Reactions were carried out at 37C for 2 min at 20 muM dTTP concentration and quenched with Sanger's gel loading dye . The reaction products were resolved by denaturing polyacrylamide-urea gel electrophoresis and analyzed on a PhosphorImager (Molecular Dynamics, Inc.). DNA polymerase assay | Polymerase activity of the HIV-1RT WT and insertion mutant enzymes was determined using two different template-primers: U-5PBS HIV-1 RNA and synthetic 49-mer U5-PBS DNA templates primed with the 17-mer PBS primer . Assays were carried out in a 50 muL volume containing 50 mM Tris HCl, pH 7.8, 100 mug/mL bovine serum albumin, 5 mM MgCl2, 1 mM dithiothreitol, 60 mM KCl, 100 nM template-primer, 50 muM of each of the four dNTPs with one of them being 32P-labeled (0.1 muCi/nmol dNTP) and 21 nM enzyme. Reactions were incubated at 37C for 3 min and terminated by the addition of ice-cold 5% trichloroacetic acid containing 5 mM inorganic pyrophosphate. Following termination, the reaction mixtures were filtered on Whatman GF/B filters. The filters were then dried, immersed in scintillation fluid and counted in a liquid scintillation counter. Gel analysis of RNA and DNA dependent polymerase activities | The U5-PBS HIV-1 RNA and heteropolymeric synthetic U5-PBS HIV-1 DNA templates primed with the 17-mer PBS DNA primer were used to assess the polymerase activities of the wild type and mutant heterodimeric enzymes. The primers were 5'-labeled using gamma-32P-ATP and T4 polynucleotide kinase according to the standard protocol . Polymerase reactions were carried out by incubating 2.5 nM template primer with 50 nM of the wild type HIV-1RT or its mutant derivative in a total reaction volume of 6 muL containing 25 mM Tris-HCl, pH 7.5, 10 mM dithiothreitol, 100 mug/mL bovine serum albumin, 5 mM MgCl2 and 50 muM of each dNTP. Reactions were initiated by the addition of enzyme and terminated by the addition of an equal volume (6 muL) of Sanger's gel loading dye . The reaction products were resolved by denaturing poly acrylamide-urea gel electrophoresis and analyzed on a PhosphorImager (Molecular Dynamics, Inc.). Template-Primer (TP) binding affinity of the wild type enzyme and its mutant derivatives | The dissociation constants (Kd) of the E-TP binary complexes of the wild type HIV-1 RT and its mutant derivatives were determined as described by Tong et al. . The heteropolymeric 33-mer DNA (0.4 nM) annealed to 5'-32P-labeled 21-mer primer (0.3 nM) was incubated with varying concentrations of the wild type enzyme and its mutant derivatives in a total volume of 10 muL containing 50 mM Tris-HCl, pH 7.8, 5 mM MgCl2 and 0.01 % BSA. Following incubation of the mixture for 10 min at 4C, equal volume of 2x gel-loading dye containing 0.25% bromophenol blue and 20% glycerol was added. The E-TP binary complexes formed were resolved at 4C on 6% native polyacrylamide gel using Tris-Borate buffer (85 mM Tris, 85 mM Boric acid, pH 8.0). The amounts of the TP in the binary complex (E-TP) and in free form with respect to the varying concentrations of the enzyme protein were determined by PhosphorImager (Molecular Dynamics, Pharmacia) analysis of the gel. The fraction of the bound DNA was plotted against enzyme concentration and the Kd [DNA] value was determined as the RT concentration at which 50% of DNA is bound. Ternary complex formation assay | The ternary complex (E-DNA-dNTP) formation was assessed by incubating the binary complexes of enzyme and dideoxy terminated template primer in the presence of next correct dNTP . The binary complexes were formed by incubating 10 --50 nM of the wild type enzyme or its mutant derivatives with 0.3 nM of 5'-32P-labeled dideoxy terminated 33-mer/21-mer template-primer as described above. The chosen concentration of enzyme was such that resulted in almost complete shift during E-TP complex formation. The E-TP-dNTP ternary complex formation was assessed by the addition of dNTP complementary to the next template base (in this case dGTP, 200 muM). Following incubation with dNTP at 4C for 10 min, 300 fold molar excess of a DNA trap was added to the incubation mixture to assess the stability of the binary and ternary complexes formed by the enzyme. The complexes were resolved on a 6% native polyacrylamide followed by phosphorImaging. The extent of E-TP-dNTP ternary complexes formed was quantified using ImageQuant software. RNase H activity assay | We used a 5'-32P labeled 30-mer synthetic U5-PBS RNA template annealed with a complementary 30-mer DNA to determine the RNase H activity of the enzymes . The reaction mixture contained labeled RNA-DNA hybrid (10 K Cerenkov cpm), 50 mM Tris-HCl pH 8.0, 60 mM KCl, 10 mM dithiothreitol, 0.1 mg/ml bovine serum albumin, 5 mM MgCl2, and 20 ng of enzyme in a final volume of 5 mul. Reactions were carried out at 37C for 30 sec and 1 min and terminated by the addition of equal volume of Sanger's gel loading dye . The cleavage products were analyzed on an 8% denaturing polyacrylamide-urea gel and scanned on a phosphorImager (Molecular Dynamics Inc.). Backmatter: PMID- 12097150 TI - Molecular cloning and tissue distribution of mammalian L-threonine 3-dehydrogenases AB - Abstract | Background | In mammals, L-threonine is an indispensable amino acid. The conversion of L-threonine to glycine occurs through a two-step biochemical pathway involving the enzymes L-threonine 3-dehydrogenase and 2-amino-3-ketobutyrate coenzyme A ligase. The L-threonine 3-dehydrogenase enzyme has been purified and characterised, but the L-threonine 3-dehydrogenase gene has not previously been identified in mammals. Results | Transcripts for L-threonine 3-dehydrogenase from both the mouse and pig are reported. The ORFs of both L-threonine dehydrogenase cDNAs encode proteins of 373 residues (41.5 kDa) and they share 80% identity. The mouse gene is located on chromosome 14, band C. The amino-terminal regions of these proteins have characteristics of a mitochondrial targeting sequence and are related to the UDP-galactose 4-epimerases, with both enzyme families having an amino-terminal NAD+ binding domain. That these cDNAs encode threonine dehydrogenases was shown, previously, by tiling 13 tryptic peptide sequences, obtained from purified L-threonine dehydrogenase isolated from porcine liver mitochondria, on to the pig ORF. These eukaryotic L-threonine dehydrogenases also have significant similarity with the prokaryote L-threonine dehydrogenase amino-terminus peptide sequence of the bacterium, Clostridium sticklandii. In murine tissues, the expression of both L-threonine dehydrogenase and 2-amino-3-ketobutyrate coenzyme A ligase mRNAs were highest in the liver and were also present in brain, heart, kidney, liver, lung, skeletal muscle, spleen and testis. Conclusions | The first cloning of transcripts for L-threonine dehydrogenase from eukaryotic organisms are reported. However, they do not have any significant sequence homology to the well-characterised Escherichia coli L-threonine dehydrogenase. Keywords: Background : In husbanded animals, dietary supplements of mixtures of indispensable amino acids regulate their growth and the activity of amino acid metabolising enzymes . The liver plays a critical role in regulating amino acid metabolism. Porcine-derived hepatocytes are being used in clinical studies of bioartificial liver organs . The regulation of amino acid supply to bioartificial organs and maintaining the activity of the amino acid-metabolising enzymes will be important in their development. From a metabolic perspective, threonine is one of three indispensable amino acids. In biological tissues, there are two major and one minor L-threonine degradation pathways. L-threonine is either catabolised by L-threonine 3-dehydrogenase (EC 1.1.1.103; gene abbreviation TDH) to 2-amino-3-ketobutyrate or by L-serine/threonine dehydratase (EC 4.2.1.16; L-threonine deaminase, gene abbreviation SDH) to NH4+ and 2-ketobutyrate in the major pathways and by threonine aldolase (EC 4.1.2.5; gene abbreviation GLY1) to yield glycine and acetaldehyde in the minor pathway . In both prokaryotic and eukaryotic cells, the conversion of L-threonine via 2-amino-3-ketobutyrate to glycine takes place in a two-step process . L-threonine dehydrogenase catalyses the reaction: L-threonine + NAD+ = 2-amino-3-ketobutyrate + NADH. The subsequent reaction between 2-amino-3-ketobutyrate and coenzyme A to form glycine and acetyl-CoA is catalysed by 2-amino-3-ketobutyrate coenzyme A ligase (EC 2.3.1.29; gene name KBL). If not processed by 2-amino-3-ketobutyrate coenzyme A ligase, the highly reactive intermediate, 2-amino-3-ketobutyrate, rapidly undergoes decarboxylation to form aminoacetone and CO2. L-threonine dehydrogenase and 2-amino-3-ketobutyrate coenzyme A ligase have been found to be associated physically on the inner membrane-matrix of mitochondria where the two enzymes form a complex with a stoichiometry of one threonine dehydrogenase tetramer to two 2-amino-3-ketobutyrate coenzyme A ligase dimers . The three-dimensional structure of Escherichia coli 2-amino-3-ketobutyrate coenzyme A ligase has been determined recently . Recently, we have cloned the murine and human 2-amino-3-ketobutyrate coenzyme A ligase cDNAs and wished to clone the first enzyme in this L-threonine catabolic pathway. The E. coli L-threonine dehydrogenase gene and its activity have been well characterised . I describe the first isolation of eukaryotic cDNA sequences encoding L-threonine dehydrogenases that have significant identity to 13 peptide sequences from the purified porcine L-threonine dehydrogenase enzyme , and the amino-terminus peptide of L-threonine dehydrogenase from the Gram+ Firmicutes bacteria, Clostridium sticklandii . However, they have no significant homology to the E. coli enzyme at the primary amino acid sequence level. Results : Previously we identified, by homology to the E. coli gene, both the mouse and human cDNAs for 2-amino-3-ketobutyrate coenzyme A ligase, the second enzyme in the biochemical pathway that converts L-threonine to glycine . In a search for the mouse cDNA of L-threonine dehydrogenase, which is the first enzyme in this pathway, I initially used the same approach. However, only expressed sequence tags belonging to sorbital dehydrogenase (and, with a much lower degree of homology, numerous isoforms of alcohol dehydrogenase) were identified. Nor were other candidate genes found in the human genomic sequence. Fortunately, Kao and Davis (1994) had previously purified and characterised the porcine L-threonine dehydrogenase protein, that they had isolated from liver mitochondria and partially peptide, sequenced. This peptide sequence was used to identify mouse ESTs with significant homology by a back translation to nucleotides search. The program ESTblast was used to construct a tentative mouse contiguous sequence from EST sequences. PCR primers were designed to match the 5' and 3' ends of the electronic contiguous sequence and used to amplify the gene from mouse liver and lung cDNA. After agarose gel electrophoresis, each primer set produced a single amplicon indicating that the gene is not alternatively spliced. The PCR products were cloned and sequenced. A blast search with the murine threonine dehydrogenase cDNA sequence of the pig EST database identified similar 5' and 3' ESTs (accession Nos. BE233801 and BI400146 respectively) and the sequence of these ESTs was utilised to design primers to amplify the pig L-threonine dehydrogenase from hepatocytes by RT-PCR. Analysis of murine and porcine L-threonine dehydrogenase cDNAs | The 1508 bp mouse sequence has an ORF which encodes a 373 residue protein and has a ATTAAA polyadenylation signal at 1460 --1465 (GenBank accession No. AY116662) . A second clone (accession No. AF134346) includes 63 bp of 5'UTR and utilises a more 5' ATTAAA polyadenylation signal at 1350 --1355. The predicted protein has a 41,461 Da molecular mass and an isoelectric point 8.45. The mouse genomic sequence for this cDNA is located on chromosome 14, band C (accession No. NW_000100, The Sanger Institute, UK). The gene spans 16.4 kb and consists of 9 exons. There is a 329 bp CpG island (64% CG) spanning the 5' untranslated exon . Figure 1 | The cDNA sequence and translation of murine L-threonine dehydrogenase. The cDNA sequence and translation of murine L-threonine dehydrogenase. There are two potential polyadenylation signals (attaaa at 1350 --1355 and 1460 --1465) shown in bold and underlined with the polyadenylation sites indicated by a. An * indicates the taa stop codon. The bold type and underlined nucleotide pairs indicate the positions of the exon/exon/ boundaries. Figure 2 | The gene structure of the murine TDH gene. The gene structure of the murine TDH gene. The gene spans 16.4 kb and consists of 8 translated exons and a 5' untranslated exon on chromosome 14, band C (supercontig accession No. NW_000100, The Sanger Institute, UK). There is a CpG island spanning the 5' untranslated exon. The ORF is indicated by closed boxes and the sizes, in bp, of the exons and introns are indicated. The pig sequence (GenBank accession No. AY095535) also has an ORF that encodes a 373-residue protein with a 41,432 Da molecular mass and an isoelectric point 7.67 . At the nucleotide level, the porcine and mouse ORFs have 78% identity and at the protein level have 81% identity and 94% similarity. The potential polyadenylation signal on the pig sequence is homologous to the most 5' signal on the mouse sequence. Figure 3 | The porcine L-threonine dehydrogenase cDNA sequence and translation. The porcine L-threonine dehydrogenase cDNA sequence and translation. The ORF (accession No. AY095535) is shown in bold and the 5' EST (accession No. BE233801) and the 3' EST (accession No. BI400146) are shown in lower case. The potential polyadenylation signal (attaaa) is shown in bold and underlined. Comparison of the porcine L-threonine dehydrogenase ORF with sequenced peptides from the porcine L-threonine dehydrogenase enzyme | Evidence that the porcine cDNA encodes for L-threonine dehydrogenase comes from the high degree of similarity to sequenced peptides from the purified and structurally characterised porcine L-threonine dehydrogenase protein isolated from liver mitochondria . The sequences of 13 porcine peptides have been aligned with the porcine ORF protein and have 98% identity over 212 residues . The 5 mismatched residues are probably due to errors in peptide sequencing since they are located towards the end of the sequences. Figure 4 | Comparison of the translation of the porcine L-threonine dehydrogenase cDNA and sequenced peptides from porcine L-threonine dehydrogenase. Comparison of the translation of the porcine L-threonine dehydrogenase cDNA and sequenced peptides from porcine L-threonine dehydrogenase. Porcine cDNA sequence, Ss cDNA; porcine sequenced peptides (Pig PEP). Individual pig peptides are alternatively, not underlined, then underlined and shown in bold. Conserved residues are shown by an (*) and strongly conserved residues by (:). Import into mitochondria | Mammalian L-threonine dehydrogenase is a nuclear encoded gene; the protein is synthesised in the cytoplasm and imported into mitochondria. The amino-terminal of the mature porcine L-threonine dehydrogenase protein isolated from mitochondria corresponds to amino acid residue 51 on the porcine L-threonine dehydrogenase ORF , which suggests that the pro-protein is cleaved to produce a 36 kDa mature enzyme. This value is close to that which would be expected since the mature porcine enzyme has a subunit molecular mass of 37 kDa on SDS-PAGE . The amino-terminal region of the mouse, fly and nematode L-threonine dehydrogenase proteins all have characteristics of mitochondrial targeting sequences , despite being the region of lowest similarity within the protein, having a high content of basic amino acids and few acidic amino acids . Figure 5 | Comparison of eukaryotic L-threonine dehydrogenase protein sequences. Comparison of eukaryotic L-threonine dehydrogenase protein sequences. The L-threonine dehydrogenase sequences are: mouse, MmTDH; pig, SsTDH; Drosophila melanogaster, DmCG5955; Caenorhabditis elegans, CeF3.4 (corrected as described in results). NAD, indicates those conserved residues likely to contact the nicotinamide cofactor by homology with the crystal structure of the UDP-galactose 4-epimerases. The locations of the exon/exon boundaries are shown on the translated protein as underlined residues. Sequence homology in other species | A database search revealed the presence of L-threonine dehydrogenase genes in the genome of other organisms. The fly, Drosophila melanogaster has a 6 exon gene located on chromosome 3L which translates into a cDNA of 1288 bp, encoding the 367 residue CG5955 protein (accession No. AAF51607) . The nematode, Caenorhabditis elegans has a 5-exon gene located on chromosome V (encompassing only the first 5 exons of the predicted 10 exons of the hypothetical gene product F08F3.4, accession No. AAB04871). By extending the fifth exon to the next polyadenylation site a 1217 bp cDNA is formed, encoding a 359-residue protein . The cDNA sequences of both these genes are supported by EST data. The fly and nematode proteins have over 52% identity and 88% similarity to the mammalian proteins in the 306-residue, central core of the enzyme. Four exon/exon boundaries are conserved in two or more of the genes . The search also revealed similar L-threonine dehydrogenase ESTs in amphibians, bony fishes, tunicates, flies, moths, mites, nematodes and trypanosomes, but not in higher plants and yeasts. Similarly, the gene for the second enzyme in this pathway, KBL, is also absent from the yeast, Saccharomyces cerevisiae , and no L-threonine dehydrogenase activity has been found in S. cerevisiae . That L-threonine dehydrogenase sequences have been evolutionarily conserved between the Gram+ bacteria and mammals is shown by the homology between mouse and the amino-terminus peptide sequence from the threonine dehydrogenase of the Gram+ Firmicutes bacteria, C. sticklandii which has 54% identity and 93% similarity over 28 residues . C. sticklandii is an amino acid fermenting anaerobic bacterium that can grow on threonine as a sole substrate. Together, the mouse and C. sticklandii sequences enabled the identification of putative L-threonine dehydrogenase genes in a number of bacterial species such as Thermoplasma acidophilum, T. volcanium and Staphylococcus epidermidis. An alignment with the putative L-threonine dehydrogenase sequence, the SAV0542 gene, from S. aureus that has 41% identity and 75% similarity to the mouse protein is shown in Fig. . Figure 6 | Comparison of murine and prokaryotic L-threonine dehydrogenase protein sequences. Comparison of murine and prokaryotic L-threonine dehydrogenase protein sequences. The L-threonine dehydrogenase sequences are: mouse, MmTDH; Staphylococcus aureus hypothetical protein from gene SAV0542 (accession No. BAB46113) , S. aureus; Clostridium sticklandii threonine dehydrogenase amino-terminal peptide, C.sticklandii NT . Mammalian threonine dehydrogenases have an NAD+ binding domain | A search of the protein structural database revealed that the closest matches with 19% identity were UDP-galactose 4-epimerases (GALE) from E. coli and Homo sapiens . GALE is a mixed alpha-helices/beta-sheet protein with a N-terminal NAD+ binding Rossmann-fold and belongs to the tyrosine-dependent oxidoreductase protein family (also known as short-chain dehydrogenases). The characteristic Tyr-X-X-X-Lys couple (residues 195 and 199) found in all family members are important for catalysis with the conserved tyrosine serving as the active-site base . By comparison with the crystal structures of the GALE proteins, two domains were identified and it is likely that the substrate, L-threonine, is located in the cleft between the two domains. The larger amino-terminus domain (residues 58 --231 on the mouse sequence) has a NAD+ binding motif. There are 12 conserved residues in the murine L-threonine dehydrogenase protein that are likely to contact the nicotinamide cofactor (Gly-62, Gly-65, Gly-68, Asp-88, Ile-107, His-127, Leu-131, Asn-147, Ser-169, Tyr-195, Lys-199 and Tyr-222) . The smaller carboxy-terminus domain (residues 232 --335) has little similarity to GALE and is likely to be involved in substrate binding. Expression of L-threonine dehydrogenase mRNA in mouse tissues | To identify which tissues are likely to contribute to L-threonine dehydrogenase activity in the mouse, reverse-transcriptase real time PCR was used to examine the tissue distribution of L-threonine dehydrogenase mRNA. By reverse-transcriptase real time PCR L-threonine dehydrogenase expression was found in all tissues examined, being highest in liver, high in testis and spleen and lowest in skeletal muscle, relative to the expression of beta-actin . Similar results were also obtained with another set of L-threonine dehydrogenase primers (located on exons 6 and 7) (data not shown). The expression of 2-amino-3-ketobutyrate coenzyme A ligase was also found in all tissues examined, being highest in liver and high in kidney. The expression level of the housekeeping gene beta-actin was similar in all tissues examined. Another housekeeping gene, glyceraldehyde-3-phosphate dehydrogenase (G3PDH) was also used to standardise expression levels, but G3PDH expression showed greater variation between tissues, having higher expression in heart and skeletal muscle and lower expression in testis relative to beta-actin (data not shown). After 40 cycles of PCR amplification, the amplicons were specific as verified by melting curve analysis (data not shown) and agarose gel electrophoresis . Figure 7 | Expression of L-threonine dehydrogenase mRNA in tissues by semiquantitative PCR. Expression of L-threonine dehydrogenase mRNA in tissues by semiquantitative PCR. (A) The expression level in each tissue cDNA was normalised to the expression level of the housekeeping gene beta-actin. The ratio of L-threonine dehydrogenase mRNA to beta-actin mRNA (Y axis, arbitrary units) from each tissue was standardised to that of liver, which was taken as 100. (B) Ethidium bromide stained agarose gel of PCR products after 40 cycles of amplification. Threonine dehydrogenase, TDH; 2-amino-3-ketobutyrate coenzyme A ligase, KBL. The tissues examined were: brain, Br; Heart, He; kidney, Ki; liver, Li; lung, Lu; skeletal muscle, SM; spleen, Sp; testis, Te; and no cDNA control, -c; 100 bp ladder, m. Discussion : L-threonine dehydrogenase is the first enzyme in the biochemical pathway converting L-threonine, via 2-amino-3-ketobutyrate, to glycine. In a search for mammalian L-threonine dehydrogenase genes, clones were isolated from mouse and pig with ORFs very similar to sequenced peptides from the pig L-threonine dehydrogenase protein . The mouse and pig L-threonine dehydrogenase proteins also have significant homology to amino-terminus peptide sequences from chicken (Gallus gallus) and the bacteria, C. sticklandii . I conclude that the cloned cDNAs encode L-threonine dehydrogenase genes. The mammalian proteins belong to the tyrosine-dependent oxidoreductase protein family (also known as short-chain dehydrogenases). They do not have any significant homology to the well-characterised bacterial L-threonine dehydrogenase enzymes from E. coli and the Gram-negative plant pathogenic bacteria Xanthomonas campestris , which belong to the medium-chain, zinc-binding alcohol/polyol dehydrogenase family. However, they are distantly related with both short and long-chain dehydrogenases having characteristics of the NAD(P)-binding Rossmann-fold. Both short- and long-chain L-threonine dehydrogenases have no significant homology to the D-threonine dehydrogenase gene from Pseudomonas cruciviae that is a member of the 6-phosphogluconate dehydrogenase family. However, both types of L-threonine dehydrogenase utilise NAD+ as a co-factor, but the C. sticklandii enzyme only utilises NAD+ as an electron acceptor whereas the E. coli enzyme also catalyses the reduction of 3-acetylpyridine adenine dinucleotide and thionicotinamide adenine dinucleotide . In L-threonine dehydrogenases purified from vertebrate livers NAD+ is the lead substrate followed by L-threonine and the products are released in the order CO2, aminoacetone and NADH . All these reaction products inhibit threonine dehydrogenase activity. The short-chain and medium-chain threonine dehydrogenase also differ in their requirement for zinc ions with the E. coli enzyme requiring one ion per subunit and the C. sticklandii enzyme being inhibited by zinc ions . Their substrate specificity differs since the E. coli enzyme will also oxidizes L-serine in addition to L-threonine . In the mouse, expression of both threonine dehydrogenase and 2-amino-3-ketobutyrate coenzyme A ligase mRNAs were found in all tissues examined, both being highest in liver. In vertebrates, threonine dehydrogenase enzymatic activity is thought to be mostly confined to the liver when the mass of the organ is taken into consideration. In the rat, the activity in kidney and heart was 30% and in brain 10% of the hepatic activity: no activity was detected in other tissues . In the pig, activity in the pancreas was similar to that of liver, but no activity was found in other tissues . The mRNA data supports the suggestion that this pathway has a wider tissue distribution than had been thought previously. In vertebrates, L-threonine is degraded by two major enzymatic pathways. In normally-fed pigs and rats 80 to 87% of L-threonine catabolism occurs via threonine dehydrogenase . Probably, the 2 major pathways have different physiological roles with the cytosolic L-serine/threonine dehydratase enzyme being highly inducible by a high protein diet, starvation and cortisone, and thought to aid the homoeostatis of blood glucose by mobilising threonine and serine for hepatic gluconeogenesis. Whereas the mitochondrial threonine dehydrogenase enzyme is thought to act in the maintenance of free somatic threonine concentration derived from dietary threonine . With the identification of threonine dehydrogenase and L-serine/threonine dehydratase genes from rodents, the way is open to study their relative contribution to threonine catabolism under different physiological states. Conclusions : The mouse and pig L-threonine 3-dehydrogenase cDNAs have been cloned. They encode 373 residue proteins. The mouse gene is located on chromosome 14, band C. Mammalian L-threonine 3-dehydrogenases are mitochondrial enzymes that utilise NAD+. In agreement with this, the predicted proteins have mitochondrial targeting sequences in their amino-termini followed by an NAD+ binding domain. In murine tissues, the expression of L-threonine dehydrogenase mRNA was highest in the liver. Materials and Methods : Molecular cloning of murine and porcine L-threonine dehydrogenase cDNAs | Clones encoding the mouse L-threonine dehydrogenase cDNA sequence were obtained by touchdown PCR amplification from mouse liver and lung cDNAs using the Advantage cDNA polymerase mix (Clontech, UK) on a Perkin-Elmer 2400 thermocycler. The cycle conditions for the first 10 cycles were 94C for 5 sec, 72C less 0.4C per cycle for 3 min and for the next 20 cycles 94C for 5 sec, 68C for 10 sec, 72C for 3 min per cycle using primers (100 nM) derived from the sequence of the mouse ESTs D21787, forward 5'-CCGGCTCCCGCGTGGCGTTCTCAGCATCCA-3' and AV100443, first reverse 5' TTTTTTTTTTTTTTTTTGATACTTAAATTG-3' and second reverse 5'-TTTTTTTTTTTGCAAGCGATCGTT-3' (Amersham-Pharmacia Biotech, UK). The porcine (Sus scrofa) L-threonine dehydrogenase cDNA was cloned from primary hepatocytes (generously provided by Dr Leonard J. Nelson, Liver Cell Biology Laboratory, University of Edinburgh, UK). Total RNA was extracted using guanidine thiocyanate and treated with DNase-I to remove any contaminating genomic DNA (SV total RNA isolation system, Promega, UK). A comparison of the murine threonine dehydrogenase cDNA sequence with the pig EST database identified similar 5' and 3' ESTs (accession Nos. BE233801 and BI400146 respectively). From these EST sequences a gene-specific reverse transcriptase primer and PCR primers were designed to amplify a cDNA region encompassing the porcine L-threonine dehydrogenase ORF. Total RNA was reversed transcribed with AMV RNase H- reverse transcriptase (ThermoScript, Life Technologies, UK) at 50C using primer 5'-GTTTAATACAAATGTCACATACTTCCACAT-3'. The cDNA was amplified by touchdown PCR, as described above, with a final annealing temperature of 58C using the PCR primers 5'-GGAAGGCTGTTGGCAGAGACC-3' and 5'-CGCTTCCTCCGGCACTTCA-3'. PCR products were examined by agarose gel electrophoresis and stained with ethidium bromide. For cloning, PCR products were excised from low-melting point agarose gels and the agarose digested with agarase (Promega, U.K.). The PCR products were cloned into the T-A vector pCR-II-TOPO (Invitrogen, The Netherlands) and sequenced in both directions using the big dye terminator cycle sequencing ready reaction kit and AmpliTaq DNA polymerase FS and run on an ABI 373XL Stretch Sequencer (both from PE Applied Biosystems, UK). Tissue distribution of L-threonine dehydrogenase mRNA by semiquantitative PCR | Murine cDNA from 8 tissues (BD Clontech, UK) was analysed for the relative expression of L-threonine dehydrogenase and the housekeeping gene, beta-actin, by real time PCR. This was carried out on a GeneAmp 5700 Sequence Detection System using a SYBR Green I double-stranded DNA binding dye assay (both from AB Applied Biosystems). Approximately 0.4 mug of cDNA from each tissue was amplified by PCR using Taq Gold polymerase. Tissue master mixes were divided into gene specific mixes with the addition of PCR primers to a final concentration of 300 muM. The L-threonine dehydrogenase primers were; sense 5'-AAGCACGCGCCTGACTTC-3' and antisense 5'-CCGAGCATTGCTGTCATCTAGA-3' and produced a 99 bp amplicon. The 2-amino-3-ketobutyrate coenzyme A ligase primers were; sense 5'-TGCACGGTTGTCTTCTCAAATG-3' and antisense 5'-ACACAGCGGTCAATGTCTTCCT-3' and produced a 150 bp amplicon. The beta-actin primers were; sense 5'-CAGAAGGAGATTACTGCTCTGGCT-3' and antisense 5'-GGAGCCACCGATCCACACA-3' and produced a 93 bp amplicon. The PCR primers were designed (with the aid of the Primer Express program, AB Applied Biosystems) to amplify cDNA with both sense primers located on the penultimate exon and the antisense primers located on the last exon. The beta-actin primers were designed to amplify beta-actin specifically and not other actin isoforms or pseudogenes. The amplification conditions were; a 10 min hot start to activate the polymerase followed by 40 cycles of 95C for 15 sec and 60C for 1 min. The number of cycles required for the SYBR Green I dye fluorescence to become significantly higher than background fluorescence (termed cycle threshold [Ct]) was used as a measure of abundance. A comparative Ct method was used to determine L-threonine dehydrogenase gene expression. Expression levels of L-threonine dehydrogenase in each tissue cDNA sample was normalised to the expression levels of the housekeeping gene beta-actin (DeltaCt). The ratios of L-threonine dehydrogenase mRNA/beta-actin mRNA from each tissue were standardised to that of liver which was taken as 100% (DeltaDeltaCt). The formula 2-DeltaDeltaCt was used to calculate relative expression levels assuming a doubling of the DNA template per PCR cycle. To confirm amplification specificity, the PCR products from each sample were examined by melting curve analysis and subsequent agarose gel electrophoresis. Backmatter: PMID- 12084180 TI - A comparison of hospital readmission rates between two general physicians with different outpatient review practices AB - Abstract | Background | There has been a relentless increase in emergency medical admissions in the UK over recent years. Many of these patients suffer with chronic conditions requiring continuing medical attention. We wished to determine whether conventional outpatient clinic follow up after discharge has any impact on the rate of readmission to hospital. Methods | Two consultant general physicians with the same patient case-mix but markedly different outpatient follow-up practice were chosen. Of 1203 patients discharged, one consultant saw twice as many patients in the follow-up clinic than the other (Dr A 9.8% v Dr B 19.6%). The readmission rate in the twelve months following discharge was compared in a retrospective analysis of hospital activity data. Due to the specialisation of the admitting system, patients mainly had cardiovascular or cerebrovascular disease or had taken an overdose. Few had respiratory or infectious diseases. Outpatient follow-up was focussed on patients with cardiac disease. Results | Risk of readmission increased significantly with age and length of stay of the original episode and was less for digestive system and musculo-skeletal disorders. 28.7% of patients discharged by Dr A and 31.5 % of those discharged by Dr B were readmitted at least once. Relative readmission risk was not significantly different between the consultants and there was no difference in the length of stay of readmissions. Conclusions | Increasing the proportion of patients with this age- and case-mix who are followed up in a hospital general medical outpatient clinic is unlikely to reduce the demand for acute hospital beds. Keywords: Background : The number of emergency general medical admissions continues to grow relentlessly, many patients having repeated admissions due to exacerbations of chronic disease. The rate of readmission is sometimes used as an indicator of discharge practice performance and targets to reduce this rate have been set in the UK National Health Service . In the United Kingdom National Health Service, there are three common patterns of follow up care following an emergency general medical admission: 1) in the community, by the patient's general practitioner (primary care physician), 2) at a hospital outpatient clinic provided by the general physician who supervised the inpatient episode and 3) at a hospital specialist clinic appropriate to the patient's condition. One potential method for reducing the rate of readmissions is more frequent outpatient clinic review of patients after discharge. However, there is limited evidence regarding the effectiveness of the current practice of outpatient review following discharge from hospital in reducing readmissions. We took the opportunity of studying two physician colleagues with similar clinical practice who differed in their outpatient follow up practices. Following discharge, one reviewed twice as many patients in his personal general medical outpatient clinic than the other. The aim of this study was to identify any difference in the rate of readmission and the length of stay of readmissions associated with these different review practices. Methods : The emergency take at Birmingham Heartlands Hospital is sub-specialised, patients with acute myocardial infarction, respiratory conditions and infectious diseases being admitted directly under the appropriate specialist team. All other patients come under the care of the general physician of the day who practices general medicine separately from his or her speciality interest. The two physicians were general physicians at Birmingham Heartlands Hospital and had been appointed in 1993 and 1995. Both had an equal share of the general medical admitting rota and so received an equivalent random sample of the total emergency medicine admissions over the course of a 12 month period. Both had a special interest in renal medicine but this study only relates to their general medical practice. Inpatient data was taken retrospectively by DC from the routinely collected hospital activity computer database, which records all inpatient episodes. Coding was performed by professional coders using diagnoses provided by clinicians at the time of discharge. All patients who had been admitted as an emergency to General Medicine, managed by Dr Rayner or Dr Temple and subsequently discharged by them between April 1997 and March 1998 were included in the analysis. Factors associated with readmission were explored using multiple logistic regression analysis with independent variables of age, sex, length of stay of original admission episode, HRG coding and discharging consultant. Length of stay was divided into three categories, <3 days, 3 --7 days and >7 days. Since the decision to review the patient in the clinic would be affected by the HRG code and length of stay of the original admission, as well as the consultant discharging the patient, whether the patient was reviewed in the clinic was not included as an independent variable. Results : A total of 1366 inpatient episodes were collected from 1289 patients, 70 patients having more than one episode. In 1280 discharge episodes (1203 patients) the patient left the hospital alive and so outpatient follow up was possible. This cohort was used for subsequent analysis. Because of the specialisation in the admitting arrangements relatively few patients were coded to respiratory system (HRG3 D) (table ). There was no significant interaction between consultant and HRG code in the logistic regression analysis, confirming that the case-mix was not significant different between the two consultants. There was similarly no difference in sex (Dr A = 51.7% male, Dr B = 51.2% male) or the age distribution . Figure 1 | Age distribution for patients discharged by each consultant Age distribution for patients discharged by each consultant Table 1 | HRG3 Chapter Codes for patients discharged by the two consultants. The number of patients offered and either attending or not attending a first outpatient appointment within 90 days of discharge is shown in Table . Dr B saw twice as many patients as Dr A, 124 (19.6% of discharges) versus 63 (9.8%). The timing of outpatient clinics of those patients reviewed following discharge is shown in Figure . Dr A reviewed patients slightly earlier than Dr B. The age and HRG distributions of patients seen were very similar, 66.3% being for the cardiac HRG chapter. Hence, Dr B saw twice as many patients coded in the cardiac group (HRG3 Chapter E) as Dr A, 84 patients (30% of possible follow ups) versus 40 (15.8%). Dr B also had a higher subsequent appointment rate (58.6% versus 32.1%), giving an average number of appointments per patient of 1.59 and 1.32 respectively. Figure 2 | Cumulative percentage of patients attending first general medical out-patient review clinic with time after discharge for each consultant. Cumulative percentage of patients attending first general medical out-patient review clinic with time after discharge for each consultant. Table 2 | Number patients attending and not attending (in brackets) the first follow-up appointment within 90 days of discharge. Following discharge, a number of patients were referred to other specialist clinics rather than being followed up in the general medical clinic of the admitting consultant. Similar numbers were referred to other clinics by each consultant, 177 (25%) by Dr A and 156 (23%) by Dr B, with a similar spread of specialities. There was no difference in the referral rate to cardiology clinics (33 patients from each consultant). Re-admission rates | From the cohort of 1203 patients discharged, 411 patients were readmitted to the Trust as an emergency under any speciality in the twelve months after discharge. 28.7% of patients discharged by Dr A and 31.5 % of those discharged by Dr B were readmitted at least once. Many patients had more than one re-admission and the total number of episodes generated by these 411 patients was 800 , Dr A's cohort generating 388 and Dr B's 412 episodes. Table 3 | Readmission rates following discharge episode. The age distribution for re-admissions was more skewed to the elderly than the initial cohort . The greatest rate of re-admission occurred in the first 6 weeks after discharge and continued steadily over the subsequent 11 months . Figure 3 | Age distribution for patients readmitted following discharge by each consultant. Age distribution for patients readmitted following discharge by each consultant. Figure 4 | Cumulative growth of readmissions during 12 month period after discharge by each consultant. Cumulative growth of readmissions during 12 month period after discharge by each consultant. The HRG coding for the re-admission episode closely matched that for the initial admission (see Additional file: ). Cardiac disorders (HRG3 Chapter E) was the dominant group. 76 (28.9%) of the 263 patients in this group discharged by Dr A (38% of the discharge cohort), were re-admitted at least once, generating 143 episodes (36.9% of readmission episodes). 64 (22.1%) of the 289 patients in this group discharged by Dr B (43% of the discharge cohort), were re-admitted at least once, generating 160 episodes (38.8% of re-admission episodes). Cardiac disorders (HRG3 Chapter E) made up 66% of patients followed up in outpatients A similar proportion of the patients who attended the follow-up general medical clinics for each consultant were readmitted within the 12 months following discharge, 6 of 63 for Dr A (10%) and 11 of 124 for Dr B (9%). Variables independently associated with readmission were increasing age (expressed in units of 10 years), particular HRG codes (both Digestive system, and Musculo-skeletal system, in comparison with all other HRG's -- see table ), and increasing length of stay (LoS) of the original inpatient episode. The discharging consultant was not a "risk factor" for readmission, and there was no interaction between consultant and HRG code. Table 5 | Variables independently associated with readmission from multiple logistic regression analysis. Length of stay for categories 3 --7 days and >7 days was compared to <3 days. Length of stay | Patients who were subsequently readmitted had a longer original length of stay than those not readmitted (mean 4.9 v 3.6 days, median 3 v 2 days). The length of the readmission episodes was much longer and did not differ between the consultants (mean 8.1 v 8.1 days, median 4 v 4 days). Discussion : This observational study compares two consultant physicians whose general medical practice, due to the specialist admissions arrangements for respiratory or infectious diseases, focussed on cardiovascular and cerebrovascular disease and deliberate self-harm. About one-third of patients were readmitted at least once over the twelve month period following discharge and the total number of readmission episodes equalled about half of the number of discharges. The risk of readmission was greater in older patients and in those with a longer initial inpatient stay. A longer length of stay may indicate more severe disease or less adequate social circumstances. The risk was less in those with digestive and musculo-skeletal conditions (HRG3 Chapters F and H), which would have included conditions that are less likely to recur, such as gastro-intestinal bleeding and musculoskeletal pain. For the middle length of stay group (3 --7 days) and digestive conditions, p-values are marginally non-significant; but the former is part of a clear trend with increasing length of stay, and the latter is so sharply distinguished from any other risks in specific HRG's that it was included. The rate of readmission for both physicians was high. However, similar rates of readmission have been reported from the USA and Australia in similar patient groups. The risk of readmission in these studies was also associated with patient characteristics such as age, co morbidity, depression and previous readmission events rather than the quality of patient care . As both consultants reviewed a proportion of patients in the outpatient clinic, it is not possible to exclude totally an effect of general medical clinic review on the rate of readmissions. Furthermore, only one outcome of treatment, readmission, has been looked at in this study and we have no way of comparing the quality of patient care or patient satisfaction. The rate of readmission was highest in the first six weeks after discharge. Approximately 50% of follow-up appointments were made for more than six weeks after discharge and hence these could not be expected to affect the early readmission rate. It would be interesting to study the impact of earlier outpatient review on readmission rates. This study compared two physicians with different thresholds for offering general medical outpatient clinic review following discharge. Although carried out by different consultants, the clinics were otherwise run in a similar and conventional way. It did not compare different types of outpatient clinic, which may be more effective than the conventional follow up clinic. Others have found similar difficulty in reducing the rate of readmission, either by involving a general practitioner in the discharge process or running a nurse-led heart failure clinic . More promising results have been reported in two studies of intensive intervention in the patients' home following admission for stroke , in which physician or physiotherapist visits reduced readmission rates, and for heart failure , in which visits by a cardiac nurse had a similar effect. Conclusions : One can conclude that a practice of following up a larger proportion of patients with this age- and case-mix in the general medical outpatient clinic does not reduce the requirements for inpatient acute medical beds. These results would not support an increase in the use of general medical review clinics, of the type described in this study, in an attempt to control the growth in emergency admissions. Competing interests : None declared. Authors' contributions : HCR conceived the study and wrote the manuscript. RMT participated in the design of the study and writing of the manuscript. TM performed the statistical analysis. DC performed the data extraction and analysis. All authors read and approved the final manuscript. Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12110157 TI - Application of the development stages of a cluster randomized trial to a framework for evaluating complex health interventions AB - Abstract | Introduction | Trials of complex health interventions often pose difficult methodologic challenges. The objective of this paper is to assess the extent to which the various development steps of a cluster randomized trial to optimize antibiotic use in nursing homes are represented in a recently published framework for the design and evaluation of complex health interventions. In so doing, the utility of the framework for health services researchers is evaluated. Methods | Using the five phases of the framework (theoretical, identification of components of the intervention, definition of trial and intervention design, methodological issues for main trial, promoting effective implementation), corresponding stages in the development of the cluster randomized trial using diagnostic and treatment algorithms to optimize the use of antibiotics in nursing homes are identified and described. Results | Synthesis of evidence needed to construct the algorithms, survey and qualitative research used to define components of the algorithms, a pilot study to assess the feasibility of delivering the algorithms, methodological issues in the main trial including choice of design, allocation concealment, outcomes, sample size calculation, and analysis are adequately represented using the stages of the framework. Conclusions | The framework is a useful resource for researchers planning a randomized clinical trial of a complex intervention. Keywords: Background : Antibiotic use in nursing home residents is an important public health issue given that 50% to 75% of residents are exposed to at least one course of antibiotics annually . Such overuse can potentially lead to the development of multi-drug antibiotic resistant bacteria, drug-related adverse effects, and harmful drug interactions . The appropriate use of antibiotics in this setting is challenging. The diagnosis of infection in residents of nursing homes is often difficult, since symptoms may be vague and signs difficult to elicit . The evaluation of residents is often made by nursing staff who have varying levels of expertise and knowledge about infectious diseases. Frequently, treatment decisions are made by busy clinicians who rely on nursing assessment as their primary source of information. Interventions to optimize antibiotic use in this setting therefore need to directly involve both nurses and physicians and have the potential for strong "buy-in" from these healthcare professionals. Such interventions are aimed at changing behavior and by necessity will involve multiple components. Recently, Campbell and colleagues published a framework for the design and evaluation of complex health interventions . Derived from a document drafted by members of the MRC Health Services and Public Health Research Board, the paper is based on the premise that a different construct is required for trials that evaluate complex interventions. Examples include interventions to improve delivery of health services, to change health professionals' behavior, or to promote health through strategies aimed at individuals, schools, healthcare facilities, or communities. It is apparent that such trials pose very different challenges when compared to clinical trials designed, for example, to assess the efficacy of pharmacologic therapy . Amongst these are the development, identification, documentation, and reproduction of the interventions. Although such challenges are well known , few published strategies exist for researchers about to embark on a clinical trial of these interventions. Campbell et al. propose that the development and evaluation of such complex interventions be considered in a series of phases, analogous to the sequential phases of drug development . The design and evaluation framework proposed by Campbell et al. appears to be well suited to interventions for optimizing antibiotic use in long-term care facilities. The objective of this paper is to evaluate the extent to which the various steps in the development of a cluster randomized trial to optimize antibiotic use in nursing homes are represented in the framework for the design and evaluation of complex health interventions . Often it is not possible to evaluate complex health interventions, such as strategies to improve antibiotic use in long-term care facilities, by randomizing individuals. Cluster randomized trials randomize groups of individuals, healthcare professionals, or institutions. Applying the development steps of this trial to the framework proposed by Campbell and colleagues allows for a formal evaluation of the framework's utility. Methods : The five phases of the framework of Campbell et al include 1) a pre-clinical or theoretical phase, 2) Phase I, identification of the components of the intervention, 3) Phase II, definition of the trial and intervention design, 4) Phase III, methodological issues for the main trial, and 5) Phase IV, promoting effective implementation of the intervention. Using the framework, the corresponding stages of the nursing home trial development are identified and described. Results : Preclinical or theoretical phase | Overview | The purpose of this phase is to determine if the intervention has the potential to have the desired effect. This step can take into account previous studies demonstrating empirical evidence for the intervention as well as the theoretical basis for the intervention. Review of the theoretical basis for an intervention may lead to an altered hypothesis and improved specification of "potentially active ingredients" . The first step in the development of the algorithms was to systematically review the literature for data that could lead to strategies for reducing antibiotic use. Lower respiratory infections, urinary infections, and skin and soft tissue infections account for the majority of bacterial infections in residents of long-term care facilities . No data to support reducing antibiotics for lower respiratory or skin infections was located. In contrast, data to support reducing antibiotic use for urinary indications was found. Asymptomatic bacteriuria, the presence of bacteria in the urine in the absence of urinary symptoms, occurs in up to 50% of older institutionalized women and 35% of institutionalized older men . There are four randomized controlled trials that demonstrate no benefit in treating asymptomatic bacteriuria in residents of long-term care facilities . Despite the evidence, one third of all prescriptions for urinary indications in residents of nursing homes are for asymptomatic bacteriuria . Accordingly, the algorithms indicate that urine should not be cultured and antibiotics should not be prescribed in the absence of clinical features of urinary infection (Figures and ). The algorithms also consider the increased risk of urinary infection with indwelling urinary catheters and with specific urinary symptoms . Figure 1 | Diagnostic algorithm. Diagnostic algorithm. This algorithm guides physicians and nurses in the ordering of urine cultures for nursing home residents with suspected infections. Figure 2 | Treatment algorithm. Treatment algorithm. This algorithm allows physicians and nurses to optimize antibiotic use in residents with suspected infections. Phase I: defining components of the intervention | Overview | This phase, often inter-related to the previous phase, emphasizes modeling or simulation studies to improve understanding of the components of an intervention. Use of qualitative research and descriptive studies may help define components of the intervention. Furthermore, in this phase potential barriers in trials that seek to change behavior may be elicited . In order to assess the potential effectiveness of the two algorithms, we applied them to data from a study assessing antibiotic use in Ontario chronic care facilities . In this modeling exercise, we found that the algorithms could reduce antibiotic use for urinary indications by 80%. Previously, we had conducted focus groups among physicians and nurses who provide care to nursing home residents. Our objective was to explore the perceptions, attitudes, and opinions of those involved in the process of prescribing antibiotics for asymptomatic bacteriuria . Twenty-two physicians and 16 nurses, each in two focus groups, participated. Both physicians and nurses emphasized that there was a wide range of indicators, more subtle than the classic symptoms of urinary tract infection, that influence the ordering of cultures and the prescribing of antibiotics. Because residents in long-term care facilities frequently have cognitive impairment and are unable to articulate their symptoms, health care providers rely instead on signs of more subtle functional and behavioural changes. Misconceptions about the definition of symptoms of urinary tract infection were also discussed. Although the nurses participating in the focus groups had learned that foul-smelling urine alone was not sufficient reason to order a culture, a number of physicians still felt that this was one of the main reasons why cultures are ordered. Other key findings included the influence nurses have over physician prescribing through their description of the clinical situation, often conveyed over the telephone. Physicians attributed their use of antibiotics to treat bacteriuria in residents with no urinary symptoms to relatively vague non-urinary symptoms (e.g. lethargy, change in functional status). To these physicians, this meant that such residents were not in fact "asymptomatic", despite the fact that absence of urinary symptoms alone was used to define asymptomatic bacteriuria in the clinical trials . This was an important barrier to evidence-based care that would not have been discovered without using a qualitative approach. To assess the factors leading to the ordering of urine cultures, we conducted a prospective survey in four nursing homes. Nursing staff were asked to complete a brief questionnaire every time a request for a urine culture was made. They were asked to specify the signs and symptoms present at the time the culture was ordered and to specify who initiated the request for the culture. Physicians, after they received the results of the urine culture, were asked to indicate the clinical reasons for their therapeutic decision. Forty-eight percent of the 545 urine cultures obtained over six months were ordered by nursing staff and 52% by physicians. The urine culture was suggested by nursing staff 56% of the time. For 67% of urine culture orders, there were no urinary symptoms present at the time the culture was requested. One in every three urine cultures ordered resulted in an antibiotic prescription. Of these prescriptions, one third were for asymptomatic bacteriuria. Phase II: defining trial and intervention design | Overview | In this phase, the information gathered in phase I is used to develop the optimum intervention and study design. The feasibility of delivering the intervention is tested and the acceptability to healthcare providers and patients is assessed . The diagnostic and treatment algorithms were tested and further refined in a three-month pilot study conducted in four nursing homes. Administrators and directors of nursing of the homes felt that the algorithms should be introduced to the nursing staff by each facility's infection control practitioner. To ensure that the infection control practitioners themselves understood the algorithms, they were asked to decide whether urine cultures should be obtained and antibiotics ordered for six case-scenarios using the algorithms. Using this "train the trainer" approach, the practitioners then conducted brief seminars with registered nursing staff to introduce the algorithms in their facilities. This approach was unsuccessful however. Semi-structured interviews with staff revealed that in two of the homes there were nurses who were completely unfamiliar with the algorithms two-thirds of the way into the pilot study. This led to the development of a standardized 30-minute presentation of six case-scenarios facilitated by one of two study investigators. Participation was active, staff were asked to decide whether to order antibiotics and urine cultures and to justify their answers using the algorithms. In addition, a video-tape replicating the seminar material was made and distributed to the nursing homes. The use of logs to document appropriate adherence to the algorithms was also piloted during the three-month pilot. Office visits to introduce the algorithms to nursing home physicians were conducted using the six case-scenarios presented to nursing staff. Feedback about the algorithms from both physicians and nurses was positive, both groups felt that the algorithms were user friendly and feasible in the long-term care setting. Additionally, the initial poor adherence to the algorithms indicated that regularly scheduled on-site visits should be another component of the intervention, serving as an important reminder or cue to the participants. Meetings with the local champion for the study (usually the director of nursing) to discuss barriers or obstacles to completing the study were therefore held during the pilot. Feedback about adherence to the protocol was given. Such follow up visits were acceptable to staff and administrators as they were not overly time consuming. Phase III: methodological issues for the main trial | Overview | Campbell and colleagues state that this step addresses issues normally posed by randomized trials, such as sample size, inclusion and exclusion criteria, methods of randomization, as well as the challenges of complex interventions . The main antibiotic trial uses a randomized matched pair design where one of 11 pairs of nursing homes (matched by size and case-mix), has been randomized to the clinical algorithms and the other to usual management of presumptive urinary tract infections. Only free-standing, community-based nursing homes are eligible i.e. those representative of the majority of nursing homes in Canada and the United States. To be eligible, nursing homes must also: 1) have 100 or more residents (since the required sample size of antibiotics courses will be achieved sooner); 2) have no stated policy for diagnosis or treatment of urinary tract infections; 3) agree to refrain from introducing new management strategies for antibiotic utilization or urinary tract infection during the study. The primary outcome of this trial is the total number of antibiotic courses prescribed. Other outcomes include 1) the proportion of antibiotic courses prescribed for urinary indications, 2) rates of urine cultures ordered, 3) hospitalization rates for urinary tract infections, and 4) mortality rates. The last two outcomes are being measured to ensure that the algorithms can safely be applied. Within a nursing home, randomization of individual healthcare providers to the algorithm likely would introduce bias due to contamination. Therefore, for the quantitative component of this study, the nursing home will serve as the unit of allocation and analysis. To conceal allocation, one nursing home in each pair was arbitrarily assigned an even number (and the other odd) by a hospital administrative assistant not associated with the study or any of the investigators and who was blinded to the nursing homes' identity. A statistician who is not associated with the study and who was unaware of the nursing home pairs, generated a list of 11 random numbers using a random numbers table, one for each pair of nursing homes. The corresponding odd or even numbered nursing home in each pair was assigned the intervention. An absolute reduction in 20% of all antibiotic use, which corresponds to a reduction in the proportion of antibiotic prescriptions for urinary indications from 30% to 10%, was considered the minimal clinically important effect worth detecting. To detect this difference, for an alpha of 0.05 and 80% power, adjusting for the effect of within cluster dependency, 4513 prescriptions need to be collected in total. This means that 20 or 10 pairs of nursing homes will need to be followed for 10 months. Given the potential for withdrawals of nursing homes, 22 homes will be recruited. A paired t-test will be used to analyse the within-pair differences between the proportions of antibiotics prescribed for urinary indications in matched pairs of nursing homes. This way, the fact that the denominator of the proportions is actually an outcome is taken into consideration. Differences in rates of overall antibiotic use (antibiotic courses per 1000 resident days) will be compared using a paired t-test. Rates of antibiotic use for urinary indications (antibiotic courses per 1000 resident days and defined daily dosages/1000 resident days), rates of urine cultures obtained (urine cultures per 1000 resident days), rates of hospitalization (per 1000 resident days), and overall mortality rates will be compared using paired t-tests and Wilcoxon signed rank test. Logistic regression analysis that adjusts for the effect of clustering will be performed to adjust for potentially important co-variates such as nursing home size, proportion of residents bed/wheelchair bound, and pharmacy automatic stop dates. Phase IV: promoting effective stability of the intervention, assessing for the existence of adverse effects | Qualitative studies to assess the attitudes and perceptions of both nurses and physicians will help us understand the key elements to maintaining sustainability should the intervention prove to be successful in reducing antibiotic use. We plan to conduct focus group studies for nursing staff and semi-structured interviews amongst physicians and directors of nursing care. If the intervention is not successful, this qualitative approach may help explain the barriers leading to failure of the intervention. Although the trial itself is meant to examine the effectiveness of the intervention, including adherence and adverse effects, if the results demonstrate that the intervention can reduce overall antibiotic use and is safe, one possibility is to conduct a cohort study of the algorithms in nursing homes over a large geographic area will be made. Discussion : The application of the trial to improve antibiotic use in nursing home residents to the framework demonstrates that the framework is a useful resource for health services researchers. Not surprisingly, the stages of the framework have been used in previous studies of complex interventions, even before the framework was published. One recent example is the Rapid Early Action for Coronary Treatment (REACT), a multi-component, community-based educational intervention of community development, public education, provider education, and patient education (31). The description of the trial described in this paper however allows for a focused assessment of the components of this framework. The pre-clinical phase of the antibiotic trial involved a summary of the evidence needed to construct the algorithms. This phase should include a systematic review of the literature with respect to the intervention , a point omitted in the pre-clinical phase description of the design paper but discussed in the Medical Research Council paper described in the text . It was also evident that the preclinical and first phase cannot be clearly demarcated, given that it was the modeling study in phase I which actually determined if the intervention had the potential to have the desired effect, the stated aim of the preclinical phase . The use of qualitative research to help define the intervention was extremely valuable. Focus group studies helped to accurately identify and anticipate barriers to the intervention, even prior to the pilot study. Determining physicians and nurses attitudes, beliefs, and behaviors toward antibiotic use for urinary infections highlighted the importance of nonspecific, non-urinary symptoms and the key role of nurses in the process of antibiotic treatment of asymptomatic bacteriuria in institutionalized older adults. As specified in the framework, these findings illustrate the utility of using qualitative methods early in the development of complex interventions. Focus groups and semi-structured interviews will also be conducted at the end of the main trial. We anticipate that such qualitative data will help us identify strategies to sustain the algorithms should the algorithms prove to be effective in reducing antibiotic use . Results of the questionnaires administered to nurses and physicians used in developing the algorithms provided convincing evidence about the need to target both ordering of urine cultures and antibiotic prescribing in order to reduce overall antibiotic use. The particular importance of introducing the algorithms to nurses in order to culturing of urine was emphasized. This confirmed the qualitative findings. It is notable that although the phases of the framework are described as being sequential , components of the frameworks might vary temporally, as exemplified by the qualitative and questionnaire data described in phase I being collected prior to the actual development of the algorithms. The pilot study was instrumental in refining and introducing the algorithms. One of the most valuable aspects of the pilot study was finding the optimal way to introduce the algorithms to nurses and physicians. Perhaps more emphasis could have been placed on strategies to introduce the complex health interventions in the design framework . An important methodological issue for randomized trials of complex health interventions is allocation concealment. This was a key consideration in the design and implementation of the antibiotic trial. Empiric evidence exists to suggest that failure to conceal allocation can result in larger biased estimates of effect than failure to blind . When a limited number of clusters are randomized, such as in this trial, there is extra concern . Although it is often impossible to blind patients, practitioners, and researchers to the intervention, it is always possible to conceal allocation . Such issues are well described in the Medical Research Council paper described in the text . Sample size calculation is an important challenge in cluster-randomized trials. Cluster randomization reduces power and the sample size generally needs to be substantially increased. Further elaboration on this topic in the Phase III of the framework would be of benefit to researchers. Conclusion : The application of a randomized trial to optimize antibiotic use in nursing homes to the framework reported by Campbell and colleagues demonstrated that the framework is a useful resource for researchers embarking on a clinical trial of a complex health intervention. Competing interests : None declared. Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12121577 TI - Cloning and characterisation of hAps1 and hAps2, human diadenosine polyphosphate-metabolising Nudix hydrolases AB - Abstract | Background | The human genome contains at least 18 genes for Nudix hydrolase enzymes. Many have similar functions to one another. In order to understand their roles in cell physiology, these proteins must be characterised. Results | We have characterised two novel human gene products, hAps1, encoded by the NUDT11 gene, and hAps2, encoded by the NUDT10 gene. These cytoplasmic proteins are members of the DIPP subfamily of Nudix hydrolases, and differ from each other by a single amino acid. Both metabolise diadenosine-polyphosphates and, weakly, diphosphoinositol polyphosphates. An apparent polymorphism of hAps1 has also been identified, which leads to the point mutation S39N. This has also been characterised. The favoured nucleotides were diadenosine 5',5"'-pentaphosphate (kcat/Km = 11, 8 and 16 x 103M-1s-1 respectively for hAps1, hAps1-39N and hAps2) and diadenosine 5',5"'-hexaphosphate (kcat/Km = 13, 14 and 11 x 103M-1s-1 respectively for hAps1, hAps1-39N and hAps2). Both hAps1 and hAps2 had pH optima of 8.5 and an absolute requirement for divalent cations, with manganese (II) being favoured. Magnesium was not able to activate the enzymes. Therefore, these enzymes could be acutely regulated by manganese fluxes within the cell. Conclusions | Recent gene duplication has generated the two Nudix genes, NUDT11 and NUDT10. We have characterised their gene products as the closely related Nudix hydrolases, hAps1 and hAps2. These two gene products complement the activity of previously described members of the DIPP family, and reinforce the concept that Ap5A and Ap6A act as intracellular messengers. Keywords: Background : In addition to the canonical ribonucleoside and deoxyribonucleoside phosphates and cofactors, cells contain a large number of minor nucleotides. Among these are the diadenosine polyphosphates (ApnA, where n = 2 --7 ). Ap3A and Ap4A are the most intensively studied of these and are generally present in the soluble fraction of eukaryotic and prokaryotic cells at concentrations between 10 nM and 5 muM . Platelet dense granules, adrenal chromaffin granules and certain synaptic vesicles have been reported to contain high concentrations of Ap5A and Ap6A in addition to Ap3A and Ap4A, all of which can be exocytosed following appropriate stimuli and bind to target cell purinoceptors causing a variety of physiological responses in the cardiovascular and central and peripheral nervous systems [,-]. However, although Ap6A has been detected in erythrocytes , there are no substantiated measurements of Ap5A and Ap6A in the soluble fraction of nucleated cells, and it is likely that they are typically present at concentrations much lower than those of Ap3A and Ap4A. The estimated 7 muM Ap5A reported in cardiac muscle may also be confined to granules. Whilst the extracellular diadenosine polyphosphates have partially characterised signalling properties, the possible functions of the soluble, intracellular compounds remain unclear. On the one hand, they may simply be by-products of several enzymic reactions (e.g. those catalysed by aminoacyl-tRNA synthetases and other ligases ). On the other, they may have one or more of a number of important regulatory roles, including involvements in DNA replication/repair, metabolic stress responses, the determination of cellular fate, and the regulation of enzyme activities and ion channels (see ). For example, Ap5A is a potent inhibitor of adenylate kinase . It is clear that cells have a variety of relatively specific enzymes able to degrade these compounds. These include symmetrically-cleaving Ap4A hydrolases (in prokaryotes), members of the histidine triad protein family such as the FHIT tumour suppresser protein and (in lower eukaryotes) the related Ap4A phosphorylases, and most widespread of all, members of the Nudix hydrolase family (see for reviews). Nudix hydrolases cleave predominantly the diphosphate linkage in compounds of general structure: nucleoside diphosphate linked to another moiety, X. Some family members have a relatively broad substrate specificity, while others have a much more restricted range. They all possess the Nudix sequence signature motif Gx5Ex5 [UA]xREx2EexGU (where U is an aliphatic hydrophobic amino acid), which represents the catalytic site of the enzyme. Animal and plant Ap4A hydrolases (EC 3.1.6.17) degrade Ap4A, Ap5A and Ap6A, always producing ATP as one of the products. They are most active towards Ap4A; however, the related enzymes from Escherichia coli and Rickettsia prowazekii show a marked preference for Ap5A. More recently, distinct members of the Nudix family that prefer Ap6A have been purified and characterised. The three human enzymes termed diphosphoinositol polyphosphate phosphohydrolases 1 , 2alpha and 2beta (hDIPP-1, -2alpha and -2beta) as well as Schizosaccharomyces pombe Aps1 and Saccharomyces cerevisiae Ddp1p all favour Ap6A over Ap5A. Activity with Ap4A is lower, and is in fact lacking in the case of Ddp1p. Remarkably these Ap6A hydrolases are also able to utilise the structurally unrelated non-nucleotide substrates, diphosphoinositol pentakisphosphate (PP-InsP5) and bis-diphosphoinositol tetrakisphosphate ([PP]2-InsP4), hence they are commonly referred to as the DIPP subfamily. These substrates are generally favoured in vitro, with kcat/Km ratios being 50 --500-fold higher for PP-InsP5 compared to Ap6A, except for DIPP-2beta, which has a significantly reduced relative activity with PP-InsP5 compared to the others. The intracellular levels of diphosphoinositol polyphosphates appear to be lower than those of the nucleotide substrates. PP-InsP5 levels are often between 1 and 3% of the levels of InsP6, which has been estimated as being between 15 and 100 muM (see ) and [PP]2-InsP4 levels are an order of magnitude lower , and as such are virtually undetectable in many cell types. PP-InsP5 turnover is regulated by the tumour promoter thapsigargin , while [PP]2-InsP4 turnover is regulated by cyclic AMP- and cyclic GMP-dependent processes via an undefined mechanism . However, as for the ApnAs, no clear functions have yet been found for the diphosphoinositol polyphosphates, although they have been implicated in vesicle trafficking , apoptosis and DNA repair . We describe here two further members of the human DIPP subfamily with novel properties. These proteins, hAps1 and hAps2, products of the NUDT11 and NUDT10 genes respectively, closely resemble the other DIPPs in primary structure, but show a selectivity towards Ap5A and Ap6A and reduced activity towards PP-InsP5 and [PP]2-InsP4. They also display a novel pattern of tissue-specific expression. Results : Sequence alignment | An initial BLAST search of the GenBank expressed sequence tag (EST) database with the hDIPP-1 sequence identified two closely related but previously uncharacterised predicted proteins, which we have called hAps1 and hAps2 (human ApsixA hydrolases 1 and 2). Alignment of the cDNA sequences with the human genome indicated that the genomic sequences encoding these proteins, FLJ10628 and LOC139770, lie about 150 kb apart on the X chromosome at Xp11.23 and are transcribed in opposite directions . In accordance with the guidelines for the Nudix protein family, the genes for hAps1 and hAps2 have been designated NUDT11 and NUDT10 respectively by the HUGO Gene Nomenclature Committee (see ). These proteins are each 164 amino acids long and are most similar to hDIPP-2beta, hAps1 showing 76, 90 and 91% identity at the amino acid level to hDIPP-1, 2alpha and 2beta, respectively. In particular, they both possess the additional Gln residue (Q86 in DIPP-2beta) that distinguishes hDIPP-2beta from hDIPP-2alpha and hDIPP-1 and which is responsible for the reduced activity of hDIPP-2beta towards diphosphoinositol polyphosphates compared to the other DIPPs. hAps1 and hAps2 are also identical to the hDIPP-2 enzymes throughout the Nudix motif . These observations led us to anticipate that hAps1 and hAps2 would be potent dinucleoside polyphosphate hydrolases with reduced activity towards diphosphoinositol polyphosphates. The existence of two such similar expressed gene products was surprising. Comparison of the open reading frames (ORFs) of hAps1 and hAps2 revealed 6 differences at the nucleotide level, leading to five silent changes (T8, E28, E69, D88 and stop), and an Arg-Pro substitution at position 89. A comparison of the cDNAs for the two gene products showed further divergence in the 3' untranslated region. It appears that the evolution of these closely related genes occurred recently, as hAps1 and hAps2 show greater similarity with each other than with the Aps-like sequences from mouse (of which two genes, XM_135784 and XM_135786 appear to yield the same protein product) or cow . Further examination of all available hAps1 ESTs revealed a sequence divergence, leading to a point mutation S39N, in the coding region. This is present in two ESTs and was detected in 10% (3 from 30) of polymerase chain reaction (PCR) products generated from a range of normal human cDNA samples from diverse sources. In all other respects investigated, including the 3' untranslated region, it appeared to be identical to hAps1. This would suggest that this minor variant, hAps1-39N, represents a polymorphism present in the human population. Its properties were also examined in this study. Figure 1 | The NUDT11/NUDT10 region of chromosome Xp11. The NUDT11/NUDT10 region of chromosome Xp11.23. The two known exons of NUDT11, encoding hAps1, and NUDT10, encoding hAps2 are shown represented as boxes. Introns are represented by angled connecting lines. The coding region of each gene is shaded. Solid arrows indicate the direction of transcription of each gene. Figure 2 | Sequence alignment of hAps1, hAps2, hDIPP-1 and hDIPP-2 (alpha and beta). Sequence alignment of hAps1, hAps2, hDIPP-1 and hDIPP-2 (alpha and beta). The figure shows an alignment of the amino acid sequences of hAps1, hAps2, hDIPP-1 and hDIPP-2 (alpha and beta), generated by the MAP algorithm. Black shading denotes amino acids in at least two of the proteins were identical. Physicochemical similarity is denoted by grey shading. Figure 3 | Comparison of hAps1 with translated EST sequences from mouse and cow. Comparison of hAps1 with translated EST sequences from mouse and cow. No full-length bovine EST for the hAps1 homologue is available, hence the composite of two separate sequences is shown. A predicted rat protein from genomic sequencing is also shown. Table 1 | Tissue sources of known ESTs for hAps1 and hAps2 Purification and properties of hAps1 and hAps2 | The ORFs for hAps1, hAps1-39N and hAps2 were cloned into the pGEX6P-1 expression vector and expressed in Escherichia coli (E. coli). After incubation with isopropyl beta-D-thiogalactoside (IPTG) (100 muM) at 26C for 8 hours, a major protein of the expected size was induced and expressed in a soluble form in each case. The glutathione S-transferase (GST)-tagged recombinant proteins were purified by chromatography on Glutathione Sepharose 4 Fast Flow resin . Figure 4 | SDS-PAGE of purified hAps1, hAps1-39N and hAps2. SDS-PAGE of purified hAps1, hAps1-39N and hAps2. GST-tagged hAps proteins were prepared as described under "Materials and methods". Approximately 10 mug of (a) hAps1, (b) hAps1-39N and (c) hAps2 were analysed by SDS-PAGE. Purity was determined by Coomassie-blue staining a 4 --12% Bis-Tris NuPage gel (Novex). Molecular weight standards (MultiMark multi-colored standards) were from Novex. Initially, the ability of hAps1 and hAps2 to metabolise Ap5A was determined at 37C in the presence of 50 mM HEPES, pH 7.6 and 100 muM substrate. A divalent cation was essential for activity, with Mn2+ by far the most effective between 2 and 6 mM . Cu2+ supported less than 30% and Zn2+ and Co2+ each less than 3% of the maximum activity. Ni2+, Ca2+ and, surprisingly, Mg2+ were unable to activate hAps1 or hAps2. When assayed in the presence of 1 mM MnCl2, hAps1 and hAps2 showed alkaline pH optima of pH 8.5 , a feature common among Nudix hydrolases. Activity increased from undetectable at pH 6 to over 70% maximum between 7.5 and 9. Previous characterisation of hDIPP-1, 2alpha and 2beta, as well as S. pombe Aps1 was performed in a buffer containing 1 mM MnCl2 at pH 7.6. In order to compare results for hAps1 and hAps2 directly with hDIPP-1, hDIPP-2 and S. pombe Aps1, we have performed assays using identical conditions. Figure 5 | Properties of hAps1 and hAps2. Properties of hAps1 and hAps2. (a) Mn2+-dependency of hAps1 (filled circles) and hAps2 (open circles) activity towards Ap5A (100 muM). (b) pH dependency of hAps1 (filled circles) and hAps2 (open circles) activity towards Ap5A (100 muM). When hAps1, hAps1-39N and hAps2 were assayed with a wide range of potential substrates all three proteins were found, as expected, to metabolise Ap5A and Ap6A. Of the other substrates tested, only p4A and p4G showed significant rates of hydrolysis by all three, while Ap4A and Gp4G were also effective substrates for hAps2. All three enzymes showed a marked preference for adenine over guanine nucleotides. Table shows rates of metabolism and products formed where identified. Km and kcat values for each enzyme with Ap5A and Ap6A are shown in Table . The most rapidly metabolised substrate appeared to be Ap5A, although Ap6A was bound with higher affinity. Table 2 | Nucleotide substrate utilisation by hAps1, hAps1-39N and hAps2 Table 3 | Kinetic parameters of hAps1, hAps1-39N and hAps2 Prolonged treatment of Ap6A and Ap5A identified the major products as adenosine 5'-monophosphate (AMP) plus adenosine 5'-pentaphosphate (p5A) and AMP plus adenosine 5'-tetraphosphate (p4A) respectively. In neither case did p5A or p4A accumulate as these are metabolised almost as rapidly as they are formed, p5A being hydrolysed sequentially to p4A then ATP. In this respect, hAps1 and hAps2 show similar modes of action to hDIPP-1 and hDIPP-2 (S.T. Safrany, data not shown). Nudix hydrolases have been found to be sensitive to inhibition by fluoride. Hydrolysis of Ap5A by hAps1 and hAps2 was found to be inhibited non-competitively with Ki values of 13.0 +- 0.5 and 30 +- 1 muM respectively (data not shown). Despite the overall high degree of sequence similarity between hAps1, hAps2 and the DIPPs, hAps1 and hAps2 were found to have little activity towards PP-InsP5 or [PP]2-InsP4. hAps1 and hAps1-39N showed first order rate constants (k-1) with PP-InsP5 of 19 +- 4 and 7 +- 5 mug-1min-1 respectively. Activity of hAps2 was similar to that of hAps1, with a k-1 = 13 +- 6 mug-1min-1. These compare with k-1 values of 2200, 2000 and 220 mug-1min-1 for hDIPP-1, -2alpha and -2beta respectively . Km and kcat values were determined for each of the proteins and these suggested that the reduction in activity towards PP-InsP5 compared with hDIPP-1 and hDIPP-2alpha was primarily due to a reduced affinity for this substrate. Activity towards [PP]2-InsP4 was even weaker. hAps1 and hAps2 gave k-1 = 1.0 +- 0.2 and 0.9 +- 0.2 mug-1min-1 respectively. hAps1-39N showed no activity towards [PP]2-InsP4 under the conditions tested. In contrast, hDIPP-1, -2alpha and -2beta gave k-1 values of 320, 90 and 32 mug-1min-1 respectively, whereas S. pombe Aps1 and S. cerevisiae Ddp1p showed similar activity with k-1 = 8.3 and 1.7 mug-1min-1 respectively . Tissue distribution | Analysis of the tissue distribution of hAps1 and hAps2 gene expression was performed using PCR and human tissue cDNA samples. Using primers selective for hAps1, PCR products were readily visible from brain, pancreas and testis and less so from lung, thymus, prostate, ovary, small intestine and heart (the least) . Selective primers for hAps2 produced strong signals from brain and liver, but also from heart, placenta, lung, kidney, pancreas, spleen, prostate, testis and ovary . No signal for hAps1 was detected in placenta, liver, skeletal muscle, kidney, spleen, colon or peripheral blood leukocytes. Likewise, no signal for hAps2 was found in skeletal muscle, thymus, small intestine, colon or peripheral blood leukocytes. A number of tumour cell line cDNA samples were also tested; hAps1 was present in prostate PC3, and ovary GI-102 cells. The prostate PC3 cell line was also found to express hAps2. Tumour cell lines from breast GI-101, lung LX-1, colon CX-1, lung GI-117, colon GI-112 and pancreas GI-103 were found to express neither hAps1 nor hAps2. These results are in broad agreement with the sources of ESTs for these two protein families . This list is for comparison and is not exhaustive. In contrast, hDIPP-1 was found in all the tissues and cell types tested above (S.T. Safrany, unpublished data). Figure 6 | Tissue distribution of hAps1, hAps2 and beta-actin. Tissue distribution of hAps1, hAps2 and beta-actin. PCR using human tissue first-strand cDNA as template was performed with primers specific for beta-actin and for hAps1 or hAps2. 35 cycles of amplification were used for beta-actin, 40 cycles for hAsp1 and hAps2. Whilst an exhaustive analysis has not been performed for hDIPP-2, it was found to be highly expressed in heart, skeletal muscle, kidney and pancreas, and weakly expressed in brain, placenta lung and liver . Subcellular localisation | The subcellular localisation of hAps1, hAps1-39N and hAps2 was investigated by individual expression of the three proteins tagged at the N-terminus with EGFP in HEK293 and PC12 cells, and visualisation of the expressed fusion proteins by deconvolved fluorescence microscopy. These results show that in both cell types EGFP-tagged proteins showed a cytosolic location indistinguishable from EGFP-hDIPP-2alpha, EGFP-hDIPP-2beta and EGFP alone (Fig. and ). Similar data were obtained with proteins tagged at the C-terminus with EGFP (data not shown). This is in agreement with localisation predicted using PSORTII. Figure 7 | Subcellular localisation of EGFP, EGFP-hDIPP-2 (alpha and beta) and hAps1 and hAps2. Subcellular localisation of EGFP, EGFP-hDIPP-2 (alpha and beta) and hAps1 and hAps2. Proteins fused to the C-terminus of EGFP were expressed in (a) HEK293 or (b) PC12 cells for 24 h and visualised by fluorescence microscopy. Discussion : The human genome contains at least 18 Nudix genes, several of which encode multiple products, and 9 Nudix hydrolases have been characterised so far to varying degrees (see ). These include hDIPP-1, hDIPP-2alpha and hDIPP-2beta, encoded by the NUDT3 (hDIPP-1) and NUDT4 (hDIPP-2alpha and hDIPP-2beta) genes, all of which hydrolyse both the higher order diadenosine polyphosphates and diphosphoinositol polyphosphates. In this report, we describe the cloning and functional characterisation of hAps1 and hAps2, two new members of the DIPP subfamily with a novel pattern of expression and substrate specificity. A polymorphism of hAps1 (hAps1-39N) has also been found, its activity and localisation appearing broadly similar to those of hAps1. Unlike the other subfamily members, hAps1 and hAps2 show little activity towards diphosphoinositol polyphosphates, both gene products favouring diadenosine polyphosphates, ApnA, where n = 5 or 6. In each case, Ap6A binds with higher affinity than Ap5A, whereas the kcat for the latter is greater. Given that no other PP-InsP5 hydrolase activity was observed during the purification of rat hepatic DIPP (most similar to hDIPP-1) and assuming that tissue expression and physiology are broadly conserved between human and rat, these results would suggest that Aps-like proteins do not contribute significantly to inositol phosphate metabolism. Both hAps1 and hAps2 show typical Nudix hydrolase requirements, namely an alkaline pH optimum in vitro and a requirement for divalent cations . An unusual feature of hAps1 and hAps2 is that they both require Mn2+ for activation, and that Mg2+ is without effect. Optimal [Mn2+] is approximately 4 mM, which is in contrast to that estimated for free intracellular [Mn2+] of below 0.5 muM . Free [Mn2+] in our in vitro assays was not determined. It is well acknowledged that Mn2+ binds serum albumin very tightly, but despite this we would anticipate free [Mn2+] in the assays to be far in excess of intracellular levels. It is known that the activity of metal ion-requiring enzymes in different cellular compartments can be regulated through the controlled trafficking of these ions and therefore through the controlled access of these enzymes to their required ions. This principle has been most thoroughly studied in the case of Ca2+, although analogous observations have been made with Cu2+ and Mn2+. Since the [Mn2+] throughout the cell is known to be non-uniform, the unusual requirement of the hAps enzymes suggests a possible mechanism of regulation through access to this ion. Several biochemically distinct Ap5/6A hydrolases have been described previously. S. pombe Aps1, S. cerevisiae Ddp1p and hDIPP-1 and hDIPP-2 have all been shown to metabolise Ap5A and Ap6A. The activities of hAps1 and hAps2 towards Ap5A and Ap6A are comparable to hDIPP-1, but their activities towards the diphosphoinositol polyphosphates PP-InsP5 and [PP]2-InsP4 are greatly reduced, thus making them some 100 --300-fold more selective towards Ap5A and Ap6A than hDIPP-1 or hDIPP-2. S. pombe Aps1 and S. cerevisiae Ddp1p, the sole representatives of the DIPP subfamily in the two yeast genomes, appear to favour the diphosphoinositol polyphosphates. Their modes of action have been shown to differ from their mammalian counterparts. The major products of Ap6A hydrolysis by S. pombe Aps1 and S. cerevisiae Ddp1p are ADP and p4A . As with hDIPP-1 and hDIPP-2 (Safrany, data not shown), hAps1 and hAps2 produced AMP and p5A as major products. With Ap5A as substrate, S. pombe Aps1 favoured the production of ADP plus ATP, whereas S. cerevisiae Ddp1p favoured production of AMP and p4A. In this respect hAps1 and hAps2 as well as hDIPP-1 and hDIPP-2 resemble Ddp1p. Due to the chromosomal proximity of Aps1 and Aps2, and the greater degree of sequence identity with each other than with the Aps-like genes from mouse, rat or cow, it seems likely that this gene pair has arisen as the result of a recent gene duplication event on the short arm of the X chromosome . A popular view of the evolution of catalytic motifs envisages the duplication of relatively non-specific progenitors followed by adaptation to perform special tasks, and the Nudix hydrolase motif appears to be a good illustration of this process. The existence of multiple forms of these enzymes would suggest that the diadenosine polyphosphates have significant biological functions, with the specialised task of hAps1 and/or hAps2 being the regulation of Ap5A and Ap6A levels. Conclusions : Previously, the human DIPP family comprised hDIPP-1, a ubiquitous enzyme able to hydrolyse both inositol phosphates and diadenosine polyphosphates, and hDIPP-2alpha and -2beta, which are predominantly diphosphoinositol polyphosphate phosphohydrolases highly expressed in heart, skeletal muscle, kidney and pancreas. We have now identified two further members, hAps1 and hAps2, which have high selectivity towards the diadenosine polyphosphates. The distinction between DIPP and ApnA hydrolase activities reinforces the importance of Q86 in DIPP-2beta and the role of Q85 in hAps1 and hAps2. The addition of this amino acid in hDIPP-2 has been previously shown to have a substantial negative effect upon its activity towards diphosphoinositol polyphosphates, whereas Ap5A hydrolase activity was increased. The expression of hAps1 and hAps2 is also restricted. Of the tissues tested, only pancreas showed high levels of hDIPP-2 and hAps1, and only heart showed high levels of hDIPP-2 and hAps2, although ESTs were found suggesting that hDIPP-2 is expressed in testis, ovary and prostate. The presence of hAps1 and hAps2 in such a variety of tissues suggests that Ap5A and Ap6A may have roles other than as secreted molecules. We find that hAps1 is present in the pancreas, but there is no suggestion that it is secreted. For this reason, hAps1 and hAps2 are expected to control intracellular Ap5A and Ap6A levels. In conclusion, we have identified two human Nudix hydrolases that share a novel substrate specificity, metabolising predominantly Ap5A and Ap6A. These two gene products complement the activity of previously described members of the DIPP family, and reinforce the concept of Ap5A and Ap6A acting as intracellular messengers. Authors' contributions : NRL performed the cloning and subcloning, determined subcellular localisation in mammalian cells, participated in sequence analysis and obtained the genomic alignment. AGMcL assisted in the writing of the manuscript. STS conceived of this study, prepared recombinant enzymes, performed enzyme assays, determined the tissue distribution, participated in sequence analysis and drafted the manuscript. All authors read and approved the final manuscript. Materials and Methods : Materials | Ap4G, Ap5G and p4A were purchased from Jena Bioscience, Jena, Germany. Other nucleotides were purchased from Sigma. Non-radioactive PP-InsP5 was a kind gift from J.R. Falck (University of Texas Southwestern Medical Center, Dallas, TX, USA). [3H]-PP-InsP5 and [3H]-[PP]2-InsP4 were prepared as previously described . Cloning and plasmid construction | PCR products corresponding to the hAps1 sequence were amplified from a human testis library (Clontech), using the primers SSD019 (5'-ggcaggatccaagtgcaaacccaaccagacg-3') and SSD020 (5'-ggcaggatccttagggatcgctatctggcg-3') (BamHI sites are underlined). PCR was carried out using a HiFidelity Expand Kit (Roche), and the products cloned into the vector pCR2.1TOPO (Invitrogen) and sequenced. Surprisingly, three variant sequences were found (termed hAps1, hAps1-39N and hAps2). These were subcloned as BamHI fragments into the BamHI site of pGEX-6P1 (Amersham Pharmacia Biotech) (starting at K2, to compare with K3 of hDIPP-1 and hDIPP-2) and fully re-sequenced. The 3' nucleotide of SSD019 introduced a silent mutation into hAps2, and was allowed to remain. Likewise, SSD020 also introduced a silent mutation into the stop codon of hAps2. This was also allowed to remain. Vectors for the expression of Nudix hydrolase proteins fused to the C-terminus of Enhanced Green Fluorescent Protein (EGFP) were produced by ligating a cDNA encoding each protein as a BamHI digestion fragment from the GST-fusion vectors described above, into pEGFP-C1 (Clontech) previously digested with BglII and BamHI. Bioinformatics | Multiple sequence alignments were performed using MAP and BoxShade, both run by EMBnet, Lausanne, Switzerland. Blast searches of EST and genomic databases were performed using the facilities provided by NCBI, Bethesda, MD, USA. Expression and purification of hAps1, hAps1-39N and hAps2 | Full-length hAps1, hAps1-39N and hAps2 were expressed and purified from E. coli as GST-fusion proteins. Expression plasmids based on pGEX-6P1 were transformed into E. coli strain BL21, and induced at 26C overnight with 100 muM IPTG. Cells were harvested in buffer A (20 mM Tris, 150 mM NaCl, 2 mM DTT, 0.1 mM EGTA, pH 7.5), supplemented with 5 mug/ml leupeptin and 1 mug/ml aprotinin, and disrupted by sonication (3 x 15s). Particulate matter was removed by centrifugation. The supernatant was applied to (and subsequently eluted from) a 5 mm x 5 cm Glutathione Sepharose 4 Fast Flow column (Amersham Pharmacia Biotech) at a flow rate of 1 ml/min. The column was washed for 5 min with buffer A. Bound protein was eluted with a gradient generated by mixing buffer A with buffer B (buffer A plus 25 mM glutathione, pH 7.5) as follows: 0-10 min, 0-50% buffer B; 10-13 min, 100% buffer B. Fractions containing pure hAps1, hAps1-39N or hAps2 were supplemented with glycerol (10% v/v final), and stored at -80C. Tissue distribution | The tissue distribution of hAps1 and hAps2 expression was determined by PCR from human tissue cDNA samples (human cDNA panels, Clontech) and human cDNA libraries (testis from Clontech, brain from Dr Peter Cheung of the MRC Protein Phosphorylation Unit, Wellcome Trust Biocentre, Dundee, Scotland, UK). beta-Actin-specific and hAps1 and hAps2-specific primers were used. The hAps1 primers were SSD049 (5'-cgtcttcgaacagaaccaggatcg-3'), and SSD045 (5'-caaaagccacacacatggtgcc-3') and the hAps2 primers were SSD050 (5'-cgtcttcgaacagaaccaggaccc-3') and SSD046 (5'-gtgcaacaacctggagaatagtcattgta-3'). Primers SSD045 and SSD046 were selective for the 3' untranslated regions of hAsp1 and hAsp2 respectively. beta-Actin cDNA was amplified using beta AS (5'-acactgtgcccatctacgaggg-3') and beta AA (5'ccttctgcatcctgtcagcaatg-3'). PCR was carried out using a HiFidelity Expand Kit (Roche), and cDNA samples from heart, brain, placenta, lung, liver, skeletal muscle, kidney, pancreas (human MTC panel I #K1420-1, Clontech), spleen, thymus, prostate, testis, ovary, small intestine, colon, peripheral blood leukocytes (human MTC Panel II #K1421-1, Clontech). The following tumour cell lines were also tested: breast GI-101, lung LX-1, colon CX-1, lung GI-117, prostate PC3, colon GI-112, ovary GI-102 and pancreas GI-103 (human MTC Tumor Panel #K1422-1, Clontech). Subcellular localisation | The subcellular localisation of hAps1, hAps1-39N and hAps2 proteins fused to EGFP was investigated using transfected HEK 293 epithelial cells and PC12 pheochromocytoma cells. PC12 cells were grown in DMEM/5% foetal bovine serum/5% horse serum. HEK293 were grown in DMEM/10% foetal bovine serum. Both cell types were grown on coverslips and transfected using Fugene-6 (Roche) with expression vectors for each Nudix hydrolase fused to the C-terminus of EGFP. 24 hours after transfection, cells were observed using a Leica inverted stage fluorescence microscope and a Hamamatsu Orca charge-coupled-device camera. Images were analysed using Improvision OpenLab deconvolution software. Enzyme assays | For the analysis of nucleotide metabolism, assays were performed as previously described , unless otherwise stated. The UV-absorbing (260 nm) nucleotide reaction products were analysed using a 4.6 x 125 mm Partisphere 5 mum SAX HPLC column. Substrates and products were eluted at 1 ml/min by the following gradient generated by mixing water with buffer C (1.3 M (NH4)2HPO4, pH 4.8 with H3PO4): 0-5 min, 0% buffer C; 5-55 min, 0-50% buffer C; 56-70 min, 0% buffer C. pH optima of recombinant enzymes were determined in a series of buffers comprising 1 mM MnCl2, 100 mug/ml BSA and 15 mM MES, 15 mM HEPES or 15 mM Tris (pH 6-9 with HCl or NaOH). Divalent ion-dependency was determined in a buffer comprising 50 mM HEPES, 100 mug/ml BSA, pH 7.6, and 50 muM-10mM (where possible) divalent metal as the chloride salt. Inositol phosphate metabolism was determined exactly as described previously . Km and kcat values were determined by hyperbolic regression analysis. Authors' note : During the reviewing of this manuscript a paper was submitted and accepted in the Journal of Biological Chemistry (K Hidaka, JJ Caffrey, L Hua, T Zhang, JR Falck, GC Nickel, L Carrel, LD Barnes, SB Shears: An adjacent pair of human NudT genes on chromosome X are preferentially expressed in testis and encode two new isoforms of diphosphoinositol polyphosphate phosphohydrolase. J Biol Chem in press). DIPP-3beta corresponds to hAps1, whereas DIPP-3alpha corresponds to hAps2. Backmatter: PMID- 12149129 TI - Improvement of Drosophila acetylcholinesterase stability by elimination of a free cysteine AB - Abstract | Background | Acetylcholinesterase is irreversibly inhibited by organophosphate and carbamate insecticides allowing its use for residue detection with biosensors. Drosophila acetylcholinesterase is the most sensitive enzyme known and has been improved by in vitro mutagenesis. However, it is not sufficiently stable for extensive utilization. It is a homodimer in which both subunits contain 8 cysteine residues. Six are involved in conserved intramolecular disulfide bridges and one is involved in an interchain disulfide bridge. The 8th cysteine is not conserved and is present at position 290 as a free thiol pointing toward the center of the protein. Results | The free cysteine has been mutated to valine and the resulting protein has been assayed for stability using various denaturing agents: temperature, urea, acetonitrile, freezing, proteases and spontaneous-denaturation at room temperature. It was found that the C290V mutation rendered the protein 1.1 to 2.7 fold more stable depending on the denaturing agent. Conclusion | It seems that stabilization resulting from the cysteine to valine mutation originates from a decrease of thiol-disulfide interchanges and from an increase in the hydrophobicity of the buried side chain. Keywords: Background : Acetylcholinesterase (AChE, EC 3.1.1.7) is a serine hydrolase, which catalyzes the hydrolysis of acetylcholine. This enzyme is the target of organophosphate and carbamate insecticides which phosphorylate or carbamoylate the serine of the active site blocking the hydrolysis of the neurotransmitter acetylcholine. The post-synaptic membrane then remains depolarised and synaptic transmission cannot take place so the insect dies. These compounds are used to control various agricultural pests: insects, acari and nematodes. One of the consequences is that pesticide residues remain in the environment and are potentially toxic for all animals including humans since cholinergic transmission is well conserved. Insecticide residues can be detected with biosensors using AChE as biological element. Such a method has been described to detect low levels of contaminants in crops, soil, water or food samples using various kinds of biosensors . Drosophila AChE was found to be the most sensitive enzyme when compared to enzymes of non-insect origin and in-vitro-mutagenesis was used to select enzymes up to 300-fold more sensitive . But like most enzymes from mesophilic organisms, Drosophila AChE is not stable, and this instability precludes its utilization in biosensors. It can be stabilized by additives: protein such as bovine serum albumin, reversible inhibitor or polyethylene glycol. Alternatively, stabilization can also be achieved by encapsulation in liposomes . Another way to stabilize the enzyme is to use in vitro mutagenesis to modify the primary structure of the protein. This method has the advantage of stabilizing the enzyme during its synthesis leading to higher production and higher purification yields. Several kinds of mutation may stabilize a protein, among them there is elimination of free sulfhydryl groups. Wang et al. suggested that removal of nonessential free cysteines by serine could be a general method for preparing muteins with higher stability, specific activity and homogeneity. Subsequent assays have shown that mutation of free cysteines has been found to increase the half-life of proteins . However, some exemples have been reported where elimination of a free cysteine did not improve the stability of protein and even decreased it . There are eight cysteines in the wild type cholinesterase , six are involved in intrachain disulfide bonds, one is involved in an interchain disulfide bond and one, at position 290 (328 with the precursor numbering) remains free . This free cysteine is half buried in the protein, located in a loop near the disulfide bond formed by cysteines 292 and 307 (Fig. and ) . When comparing the primary sequence of cholinesterases, it appears that this cysteine is not conserved and thus is not essential for the function of the enzyme. In the present study, we used mutagenesis to change this free cysteine for a hydrophobic residue with a similar size, a valine to enhance the stability of the enzyme. Figure 1 | Position of cystein 290 near the disulfide bond formed by cysteines 292 and 307. Position of cystein 290 near the disulfide bond formed by cysteines 292 and 307. Figure 2 | Orientation of cysteine 290 towards the center of the protein. Orientation of cysteine 290 towards the center of the protein. Results and discussion : Thermodynamic and thermal (also called long-term) stabilities are usually closely related since thermal denaturation is considered to be a two-step process involving a reversibly unfolded intermediate . But the two stabilities do not necessarily correlate due to independent unfolding of microdomains . Therefore, several denaturation methods were applied. In all situations analyzed, denaturation of AChE was irreversible and followed apparent first order kinetics. Stability was then characterized by the half-life (t50), the time at which 50% of an initial enzymatic activity is preserved. The effect of freezing was measured from the recovery after one freeze-thaw cycle . Table 1 | Stability of wild type and mutated AChE (mean +- SEM) and an index of the increase in stability i.e. the ratio between the measurement made for the mutant and the wild type enzymes. We first analyzed denaturation with the most common methods used to study protein denaturation: incubation at high temperature or in the presence of urea. Incubation of the wild type enzyme at 45C resulted in the progressive irreversible denaturation of the enzyme with a 17 min half-life. Mutation of cysteine 290 to valine increased this half-life to 24 min. It also provided resistance to urea denaturation: the mutated enzyme was 2.7-fold more stable than the wild type if we consider the ratio of half-life time in 4 M urea . Lyophilisation is the most efficient method to store Drosophila AChE, however freezing denatures the enzyme at dilutions below the concentration of 1 mmole per liter . We thus checked if the C290V mutation protects the enzyme at the concentration of 10 nM. Following one freezing to -20C, 75% of the wild type enzyme was denatured while only 50% of the mutated enzyme activity was lost. The utilization of AChE in a biosensor can last several days in some continuous flow devices. In fact, as the inhibition is irreversible, the longer AChE is incubated with the sample, the lower the detection limit of the biosensor. However, the enzyme activity has to decrease from inhibition and not from spontaneous denaturation of the protein. We thus incubated the wild type and the mutated enzyme at 20C. Half the enzyme was denatured in 1.5 days; the mutation doubled this half life. Detection of insecticides in food requires their extraction with organic solvent. Although the solvent should be eliminated before the assay, low amounts may remain in solution and inactivate the enzyme. We used acetonitrile as model because it is soluble in water. This organic solvent strongly denatured the protein which had a half-life in 20% acetonitrile of 1.3 minutes. This half-life was slightly increased to 1.4 minutes by the mutation. As some proteases may be present in the sample to be assayed for pesticide residues, we tested the effect of the mutation on the proteolytic sensitivity of the enzyme. Actually, it has been reported that a protein's resistance to proteolysis increases with its conformational stability due to the fact that the susceptibily to proteolysis reflects the rate of local unfolding . Our data support this relation between stability and proteolytic susceptibility since the mutated enzyme was 1.3 more resistant to proteolytic attack than the wild type enzyme in the experimental conditions used. Several hypotheses can be put forward to explain this increased stability. A first hypothesis involves the deletion of the highly reactive sulfhydryl group which could interact with the neighboring disulfide bond. Thiol-disulfide interchanges have been reported in proteins including acetylcholinesterase and changes involving a free sulfhydryl group and a disulfide bond have been found to result in the irreversible denaturation of proteins . As the disulfide exchange reaction originates from nucleophilic attack on a sulfur atom of the disulfide, the proximity of the disulfide bond formed by cysteines 292 and 307 enforces this hypothesis. The inter Cbeta distance between Cys 290 and Cys 292 or Cys 307 are 5 A or 7.7 A respectively . Actually, mutation of cysteine 292 to valine resulted in an inactive protein suggesting that disruption of the bond between cysteines 292 and 307 results in misfolding since the loop is not involved in the structure of the active site. A free sulfhydryl group may also mediate the formation of dimer and oligomer. Although oligomerization usually has a stabilizing effect , it can sometimes lead to loss of biological activity . As the side chain of cysteine 290 points towards the interior of the protein and is inaccessible to the solvent, it is unlikely that this cysteine is responsible for the formation of tetramers. Air oxidation of thiols is an important source of thermal inactivation of lysozyme . To test the relevance of this phenomenon for stability, we used the "foot printing system" of Tullius and Dombrowski using the EDTA complex of iron(II) to generate hydroxyl radical form hydrogen peroxide as previously performed with AChE . The C290V mutation did not increase the half life of the protein. Thus, stabilization does not seem to originate from the elimination of an oxidation site. The contribution of individual amino acids to the stability of a specific protein has often been found to be difficult to assess since the elimination of one type of interaction often affects other types of interactions. This can happen in our mutation: by replacing a cysteine by a valine, we not only eliminated a reactive thiol, we also increased the hydrophobicity of the side chain which is oriented towards the heart of the protein . As increasing the hydrophobicity of buried residues usually increases the stabilityof proteins , we may hypothesize that this rule applied here. Thus, removal of the free cysteine 290 most probably stabilized Drosophila AChE by two mechanisms: inhibition of disulfide exchange and increase of the hydrophobicity of the side chain. Methods : Protein production and purification | cDNA encoding Drosophila melanogaster AChE and mutant were expressed with the baculovirus system . We expressed a soluble dimeric form deleted from a hydrophobic peptide at the C-terminal end with a 3 x histidine tag replacing the loop from amino-acids 103 to 136. This external loop is at the other end of the molecule with respect to the active and its deletion does not affect either the activity or the stability of the enzyme. Secreted AChE was purified to homogeneity using the following steps, ammonium sulfate precipitation, ultrafiltration with a 50 kDa cut off membrane, affinity chromatography with procainamide as ligand, NTA-nickel chromatography and anion exchange chromatography . Activity was recorded at 25C in 25 mM phosphate buffer pH 7, with 1 mM acetylthiocholine iodide as substrate using the method of Ellman et al. . Residue numbering followed that of the mature protein. Denaturation | All denaturation experiments were performed with 10 picomoles enzyme in 1 ml 25 mM phosphate buffer pH7 at 25C. AChE was incubated in denaturing conditions and the remaining activity was measured by taking aliquots at regular times since residual enzymatic activity is dependent on the time of incubation. To analyze thermosensitivity, enzymes were incubated at 45C instead of 25C and 1 mg/ml Bovine Serum Albumin was added to the buffer. Aliquots were mixed with cold buffer chilled on ice and the solution was incubated at 25C for ten minutes before recording the remaining activity. For urea denaturation, unfolding of AChE was induced by adding 4 M urea to the incubation buffer. Aliquots were diluted 10-fold in 25 mM phosphate buffer pH7 before measuring the remaining activity. The effect of freezing was estimated by measuring the remaining activity after one day at -20C. Stability at room temperature was measured by recording the activity after incubation at 20C. The effect of organic solvent was followed by incubation of the enzyme in 20% acetonitrile. The effect of protease sensitivity was determined by incubation of AChE with 0.1 mg/ml pronase. To follow the effect of oxygen radical on denaturation 10 mM H2O2, 1 mM FeCl3, 2 mM EDTA and 40 mM ascorbic acid were added to the incubation medium following the procedure of Weiner et al. . Authors' contributions : A.B-L performed the in vitro mutagenesis experiment, I.F. and M.A. produced and purified the proteins. S.M. and C.L. carried out the stability measurements, D.F. designed the study. All authors read and approved the final manuscript. Backmatter: PMID- 12126482 TI - Priority setting for new technologies in medicine: A transdisciplinary study AB - Abstract | Background | Decision makers in health care organizations struggle with how to set priorities for new technologies in medicine. Traditional approaches to priority setting for new technologies in medicine are insufficient and there is no widely accepted model that can guide decision makers. Discussion | Daniels and Sabin have developed an ethically based account about how priority setting decisions should be made. We have developed an empirically based account of how priority setting decisions are made. In this paper, we integrate these two accounts into a transdisciplinary model of priority setting for new technologies in medicine that is both ethically and empirically based. Summary | We have developed a transdisciplinary model of priority setting that provides guidance to decision makers that they can operationalize to help address priority setting problems in their institution. Keywords: Background : Priority setting is a challenge for every health care system in the world because demand for health care outweighs the supply of resources allocated to finance it. Which programs should a regional health authority fund? Which drugs should be placed on the drug benefit formulary? Which patients should be admitted to a hospital's critical care unit? Which priorities should a research granting body support? There are no widely accepted models for legitimate and fair priority setting in health care to help address these questions. Traditionally, approaches from health economics are promoted as the solution to the problem of priority setting. Health economics provides necessary but insufficient tools (e.g. cost-effectiveness analysis, program budgeting and marginal analysis ) to aid priority setting decision makers. Studies of actual priority setting show that these tools have only limited influence on decision making and the analyses are often unavailable when needed . Priority setting for new technologies, for example, is frequently conducted under conditions of varying degrees of evidence about the safety, effectiveness, and appropriateness of particular interventions . Moreover, there is no consensus regarding the values these approaches emphasize (i.e. efficiency) . The Institute of Medicine Panel on Cost-Effectiveness in Health and Medicine argued that, "... CEA [should] be used as an aid to decision makers who must weigh the information it provides in the context of ... other values ." Other values important to priority setting include equity, the health of individuals as against communities, the "rule of rescue," and democratic decision making . Unpacking these values helps to clarify choices, but reasonable people, having diverse moral views, disagree about what constitutes a fair allocation of resources to meet competing health care needs. In the absence of consensus on guiding principles, the problem of priority setting becomes one of procedural justice -- legitimate institutions using fair processes . Discipline-specific ethical approaches to priority setting (e.g. from philosophy, law, political science, medicine, and health economics) are insufficient because they are not grounded in actual experiences of priority setting in health care institutions and because the values that they contribute to priority setting conflict. Empirical descriptions of priority setting that focus, for example, on considerations of safety, effectiveness, and appropriateness of particular preventive, diagnostic or therapeutic interventions are also insufficient. This is because, though grounded in actual experiences of priority setting, they cannot alone provide normative guidance about what should be done. What is needed, therefore, is a synthesis that incorporates both ethical and empirical considerations. A transdisciplinary model, based in empirical realities of decision making and justified by ethical values, can provide such a synthesis. In this paper, we develop a transdisciplinary model of health care priority setting. Transdisciplinary research -- "researchers work [ing] jointly using shared conceptual framework drawing together disciplinary specific theories, concepts and approaches to address common problems" -- can be contrasted with interdisciplinary research -- "researchers work [ing] jointly but still from a disciplinary-specific basis on [a] common problem" -- and multidisciplinary research -- "researchers work [ing] in parallel from their respective disciplinary bases to address a common problem" . By integrating an ethically based account of health care priority setting (accountability for reasonableness) with an empirically based one (diamond model) as described below, we have engaged in a transdisciplinary effort. The result is a model because it consists of plausible relationships proposed among concepts and sets of concepts . However, the innovation here is not the resulting model, though the transdisciplinary model is an advance on its parent-models, but rather the transdisciplinary approach to the problem of priority setting. Discussion : Accountability for reasonableness: an ethically-based account of priority setting | Daniels and Sabin identify two key problems at the heart of health care priority setting: legitimacy and fairness . The legitimacy problem poses the question: under what conditions should authority over priority setting decisions be placed in the hands of a particular organization, group or person? The fairness problem poses the question: when does a patient or clinician have sufficient reason to accept priority setting decisions as fair? An institution's priority setting decisions may be considered legitimate and fair if they satisfy four conditions of accountability for reasonableness: relevance, publicity, appeals, and enforcement . Table 1 | The four conditions of accountability for reasonableness Although developed in the empirical context of US managed care organizations, the strength of accountability for reasonableness lies in its ethical roots found in theories of democratic deliberation, grounded in the disciplines of philosophical ethics and political philosophy. These roots connect the priority setting process in accountability for reasonableness with broader democratic processes. The limitation of accountability for reasonableness is that it does not sufficiently explain how an institution might go about operationalizing the model. For example, how should institutional administrators go about putting the relevance condition into practice? Diamond model of health care priority setting: An empirically-based account | Our case study of priority setting for new technologies in medicine in two health care institutions provides a descriptive account of how priority setting decisions are actually made . In the case study, we analyzed documents, interviewed people and observed meetings of groups making priority setting decisions. Based on these data, we developed the diamond model of health care priority setting . Figure 1 | The diamond model of priority setting The diamond model of priority setting The six elements are described briefly below: * Institutions | Priority setting occurs in an institution with a mandate to set priorities; variations between institutional cultures affect priority setting. * People | Groups of people contributing a spectrum of factors and arguments are involved in priority setting. * Factors | Factors considered important by priority setting decision makers include benefit, evidence, cost, cost-effectiveness, and equity. * Reasons | Priority-setting decisions involve clusters of factors that vary according to the decision. The previous conception of priority setting as trade-off (eg, equity vs efficiency) was too simplistic and abstract to describe actual priority setting reasoning. Reasons are assembled by combining these factor-clusters in support of a particular decision. Finally, each decision and rationale is compared with previous decisions and rationales in a casuistry or "case law" that helps to ensure consistency. * Process | Priority setting includes certain procedural safeguards such as transparency, disclosure of conflicts of interest, fair access to decision makers, and fair chairing and leadership of the priority setting group. * Appeals | Decisions are open to review based on new information or arguments. The diamond model reflects the fact that legitimacy and fairness are not all or none phenomena. Instead, they lie along a spectrum and priority setting can be more or less legitimate and fair. Much as a gem's flawlessness depends on the flawlessness of its facets, legitimate and fair priority setting depends on the legitimacy and fairness of each of six elements identified in our study. The strength of the diamond model lies in its empirical roots, grounded in the discipline(s) of the social sciences. Because it reflects the realities of actual decision making, it can more easily be put into practice than an ethically based account. The major limitation of the model is that simply because a group makes priority setting decisions in a particular way does not make these decisions "right". An ideal model of health care priority setting will need to specify what should be done (i.e. justified ethically) and how it can be done (i.e. based in empirical reality). A transdisciplinary model of priority setting in health care institutions | By combining accountability for reasonableness with the diamond model, we developed a transdisciplinary model of priority setting that has a solid ethical justification and deep empirical roots so it is more easily put into practice in a particular context (see Figure ). Figure 2 | A model of accountability for priority setting in health care institutions A model of accountability for priority setting in health care institutions We integrated these two accounts as follows . First, because reasons are central to priority setting, we collapsed together the elements of "reasons" and "factors" in our diamond model into the concept of "rationales" and positioned this as the central concept of the transdisciplinary model. Second, we arranged three elements of the diamond model -- processes, people, and appeals -- around the central concept of rationales (see Table , left hand column), because we discovered that conditions of accountability for reasonableness -- relevance, publicity, and appeals (Table , middle column) -- bound these elements to rationales. Third, we replaced these conditions with operational goals: reasonableness, transparency, responsiveness, and accountability (Table , right hand column). Finally, we situated the model in the "institutions" element of the diamond model, which accords with the enforcement condition of accountability for reasonableness, and supports an operational goal of accountability. Table 2 | Relationships among diamond model elements, accountability for reasonableness conditions, and transdisciplinary model operational goals This new model is a more practical and user-friendly version of accountability for reasonableness that can guide health care decision makers in actual priority setting. There would be constraints in each domain from accountability for reasonableness. Institutions would have to be duly delegated to make these decisions; people would have to represent an appropriate array of perspectives; the people in the context would have to determine what factors should be included in the rationales; the process would have to make reasons transparent not only to participants but to people affected by the decisions; and appeals would have to be responsive to the rationales. Implications for decision makers | Imagine that a health care institution such as a hospital, managed care organization, or regional health authority is charged with priority setting for new technologies in medicine. The following steps, based on the transdisciplinary model, offer guidance to the institution to ensure that their priority setting is legitimate and fair: Step 1 | Bring together a committee of fair-minded people to make priority setting decisions. The group should include a broad spectrum of individuals from within the organization, as well as patients and members of the public. They should focus on the reasonableness of the rationales for each decision. Step 2 | Put into place processes to ensure the transparency of the decision making activities (e.g. public notices and town hall meetings) and to ensure that the rationales are widely available. Step 3 | Design methods to hear appeals regarding the rationales for decisions and to respond to further evidence or arguments. Step 4 | Develop mechanisms of institutional accountability to ensure that the first three steps are followed. Comparison with current practice | Although studies of actual priority setting practices are rare, they often resemble aspects of the transdisciplinary model. In the US, accountability for reasonableness was itself developed in the context of private sector managed care institutions and is being applied to pharmaceutical benefit management . In Canada, the diamond model was developed in the context of public sector institutions making decisions about new technologies in cancer and cardiac care. In the UK, Oxfordshire Health Authority has developed a "Priorities Forum" that illustrates many aspects of the transdisciplinary model . There may be differences in how some public or private institutions might go about implementing the transdisciplinary model. Many public institutions are already obligated by law to disclose their decision making practices and outcomes and to include a broad representation of public stakeholders. Adoption of the transdisciplinary model would therefore entail a less radical transformation of their institutional culture and practices than will be the case for many private institutions that have, up until this point, been under no such obligations . Both accountability for reasonableness and the diamond model -- and therefore the transdisciplinary model -- were developed in the context of priority setting for new technologies. It would be reasonable to expect that the transdisciplinary approach might be more broadly applicable, for example, to decision making about bed closures or space allocation or staffing, or around strategic, capital, and operational planning both at the hospital level and at the clinical level. If this exercise were repeated in different contexts, the resulting transdisciplinary models may very well look quite different. This transdisciplinary model may not be fully generalizable to all contexts, though we expect that it will offer some guidance to decision makers across many priority setting contexts. Summary : Priority setting decision makers have had little guidance to help them develop their processes. Traditional discipline-specific approaches are insufficient because they cannot tell decision makers what to do in specific contexts, and empirically derived models of actual decision making are insufficient because they cannot provide normative guidance. We have developed a transdisciplinary model of priority setting that provides guidance to decision makers that they can operationalize to help address priority setting problems in their institution. Funding : The project was funded by a grant from the Medical Research Council of Canada/Canadian Institutes of Health Research and the Physicians Services Incorporated Foundation. Dr. Gibson is supported by a Canadian Health Services Research Foundation post-doctoral award. Dr. Singer is supported by a Canadian Institutes of Health Research Investigator award. Competing interests : None declared Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12204944 TI - Peripheral Blood Lymphocyte Subsets in Adolescents: a Longitudinal Analysis from the REACH Project AB - Flow cytometry analysis of lymphocyte subset markers was performed for a group of sexually active, human immunodeficiency virus (HIV)-negative adolescents over a 2-year period to establish normative data. Data were collected in the REACH Project (Reaching for Excellence in Adolescent Care and Health), a multicenter, longitudinal study of HIV-positive and high-risk HIV-negative adolescents. Two- and three-color flow cytometry data were collected every 6 months for these subjects. We determined the effects of gender, race, and age on the following lymphocyte subset markers: total CD4+ cells, CD4+ naive cells, CD4+ memory cells, all CD8+ cells, CD8+ naive cells, CD8+ memory cells, CD16+ natural killer cells, and CD19+ B cells. Gender was the demographic characteristic most frequently associated with differences in lymphocyte subset measures. Females had higher total CD4+ cell and CD4+ memory cells counts and lower CD16+ cell counts than males. Age was associated with higher CD4+ memory cell counts as well as higher CD8+ memory cell counts. For CD19+ cells, there was an interaction between age and gender, with males having significantly lower CD19+ cell counts with increasing age, whereas there was no age effect for females. Race and/or ethnicity was associated with differences in total CD8+ cell counts and CD8+ memory cell counts, although both of these associations involved an interaction with gender. Keywords: Introduction : The interaction between infection with human immunodeficiency virus (HIV) and immunologic responses to infection has led to the development of new and more detailed laboratory methods. There are limited normative data available for immunologic assays and flow cytometry measures, especially for specific age populations, such as adolescents. Adolescence is a unique period of development, characterized by profound physiologic and psychosocial changes . Gender and age differences in immunologic cell numbers and function have been observed for both healthy and diseased subjects. Studies of adolescents and adults have demonstrated gender differences as well as some age differences in immune cell counts as characterized by flow cytometry studies and limited number of functional assays . Age may be an important factor influencing immunologic responses to infection, especially in younger children .Other factors, including race and genetic characteristics, may also influence immunologic cell numbers. Understanding how these factors influence the immune system in healthy individuals is key to beginning to understand age, gender, and race differences in immune system-based diseases and the adolescent's immunologic response to infection with HIV and other infectious agents. HIV infection is a major issue confronting adolescents . Recent data show marked increases in the number of HIV infections in adolescents and young adults, especially in the number of infections due to heterosexual transmission in young women and in the number due to male-to-male transmission in young men . The REACH (Reaching for Excellence in Adolescent Care and Health) Project of AMHARN (The Adolescent Medicine HIV/AIDS Research Network) recruited and longitudinally followed a cohort of high-risk youths not infected with HIV to establish normative data for this population . The study set out to establish gender, age, and racial differences in a set of immunologic markers in a longitudinal analysis. Before comparisons can be made for HIV-infected adolescents, normative data must be established for groups related to age, race and/or ethnicity, and gender. We present data for a group of phenotypic markers in an HIV-negative adolescent cohort. MATERIALS AND METHODS : The REACH Project recruited HIV-positive and high-risk HIV-negative adolescents (age range, >=13 to <19 years) into a study of biomedical and behavioral features of HIV infection as observed while under medical care for HIV infection and adolescent health. The HIV-negative subjects served to establish adolescent normative data for a number of biological measures. HIV-negative subjects were recruited from adolescent clinics serving high-risk adolescents based on the high seroprevalence rates in the geographic areas and on sexual risk or needle-using risk of the adolescents. The characteristics of the cohort, recruitment and eligibility criteria, and study design have been reported elsewhere . The HIV-negative subjects were determined to be so on the basis of results from an enzyme-linked immunosorbent assay (ELISA) performed within 30 days of enrollment into the REACH study. The HIV ELISA was repeated annually to confirm HIV-negative status. The HIV-negative youth, to qualify for enrollment into the study, had a history of either sexual intercourse or injection drug use. Blood samples for HIV-negative subjects were collected at 15 clinical sites every 6 months (see Appendix). The following flow markers were analyzed, along with an automated differential count, at a local AIDS Clinical Trials Group-certified laboratory: CD3+/CD4+ (helper T cell), CD3+/CD8+ (suppressor and/or cytotoxic T cell), CD3-/CD56+/CD16+ (natural killer cell), and CD19+ (total B cell). Additional flow markers were analyzed centrally at the Immunology Core Laboratory at The Children's Hospital of Philadelphia, as previously reported . These markers included the following cell subsets: CD4+/CD45RO-/CD45RA+ (naive helper cell), CD4+/CD45RO+/CD45RA- (memory helper cell), CD8+/CD45RO-/CD45RA+ (naive suppressor cell), and CD8+/CD45RO+/CD45RA- (memory suppressor cell). A cross-sectional analysis of 192 subjects by two-sided tests with a type I error of 5% and a power of 80% can identify a standardized effect of 0.20, considered small, or a correlation of 0.20. Repeated analyses, with subject's age at the data collection visit as the time variable, were performed using Generalized Estimating Equations (GEE) software (SAS software, version 8.1; [SAS Institute]). P values of 0.05 or less were considered significant for two-sided tests. Multiple comparison corrections were not used. Summary statistics for cell counts are presented as means and standard deviations, which include only one observation per subject in each age category, although individual subjects may have contributed to more than one age category. For analysis, counts were log-transformed to make skewed distributions more nearly normal, as assessed by the Shapiro-Wilks test. Separately for each lymphocyte subset, GEE models were used to regress log counts on gender, subject's present age, and race and/or ethnicity of these HIV-negative subjects. The three two-way interactions of the demographic variables were also examined; models with interactions also retained main effects. Because multiple visits in the same age category were included, these GEE models used more data than appear in the summary of means and standard deviations. RESULTS : At the time of this analysis, 198 HIV-negative subjects were enrolled in the REACH cohort. Two HIV-negative transgender males were excluded and 4 subjects did not have a complete lymphocyte subset panel evaluation performed, leaving 192 subjects eligible for analysis. The demographic characteristics of the eligible subjects are shown in Table . For each eligible subject, all study visits at which a complete lymphocyte subset panel evaluation had been performed were included in the analysis. The total number of study visits included was 779, with one to nine visits per subject. The distribution of ages at the times the visits were made is also shown in Table . The means and standard deviations by gender and age for each of the eight subsets are provided in Table . Figures to are a graphical display of means by gender and age. The significant statistical associations, by cell type, are shown in Table . There was a significant association with one or more of the demographic characteristics for 6 of the 8 lymphocyte subsets that were examined. No association with the three demographic characteristics was found for CD4+ naive cells or for CD8+ naive cells. Gender was the demographic characteristic most frequently associated with differences in lymphocyte subset counts. Gender had a significant main effect or interaction for 6 of the 8 subsets: total CD4+ cells, CD4+ memory cells, total CD8+ cells, CD8+ memory cells, CD16+ cells, and CD19+ cells. Females had higher total CD4+ cell counts, higher CD4+ memory cell counts, and lower CD16+ cell counts. For total CD8+ cells and CD8+ memory cells, there was an interaction between gender and race. African-American males had significantly lower total CD8+ counts than did other males; there was not a significant difference in total CD8+ counts by race for females, although African-American females had nearly significantly higher total CD8+ counts. For CD8+ memory cells, African-American and Hispanic males had higher mean counts than males classified as Caucasian or other, while African-American and Hispanic females had lower mean counts than females classified as Caucasian or other. None of these differences were statistically significant when data for males and females were modeled separately. For CD19+ cells, there was an interaction between gender and age. Males had significantly lower CD19+ cells counts with increasing age, whereas there was no age effect for females. Age was associated with differences in three of the eight subsets: CD4+ memory cells, CD8+ memory cells, and CD19+ cells. Older subjects had higher CD4+ memory cell counts and higher CD8+ memory cell counts. The gender by age interaction for CD19+ cells is described in the preceding paragraph. Race and/or ethnicity was associated with differences in only two of the eight subsets: total CD8+ cells and CD8+ memory cells. Both of these associations involved an interaction with gender, as was discussed above. FIG.1. | Graphical representation of peripheral blood lymphocyte subsets (CD4 cells) observed in adolescents. Graphical representation of peripheral blood lymphocyte subsets (CD4 cells) observed in adolescents. FIG. 3. | Graphical representation of peripheral blood lymphocyte subsets (CD16 and CD19 cells) observed in adolescents. Graphical representation of peripheral blood lymphocyte subsets (CD16 and CD19 cells) observed in adolescents. TABLE 1 | Characteristics of the cohort TABLE 2 | Mean counts (and standard deviations) of lymphocyte subsets by gender and age TABLE 3 | Significant associations between demographic characteristics and lymphocyte subsets, calculated by multivariate linear regression DISCUSSION : We have investigated the impact of gender, race, and age on T-lymphocyte subset populations as measured by flow cytometry in adolescents identified as at high risk for HIV infection in a national, longitudinal study. Gender was the demographic factor most frequently associated with differences in lymphocyte subsets. In females, we found higher CD 4+ T-lymphocyte counts and higher CD4+ memory T-lymphocyte counts, and lower CD 16+ T-lymphocyte counts than in males. Our study demonstrated some limited gender differences in CD8+ cells and CD19+ cells counts, but only with the influence of either race or age. Since this was a longitudinal study in a predominantly African-American and Hispanic group of adolescents, it serves as an important reference when studying disease processes that affect the immune system in the adolescent population. Our data for CD4+ T lymphocyte differences by gender are consistent with those previously reported for minority and Caucasian adolescents. Bartlett and colleagues studied a group of minority youth and found higher levels (both totals and percentages) of CD4+ T lymphocytes in females than in males and lower levels of B cells in females than in males . In another study of primarily Caucasian adolescents from 12 to 19 years, Tollerud and colleagues found higher levels of CD4+ T lymphocytes in females than in males among older adolescents . Our studies demonstrate that much of this gender difference is mainly in the memory CD4+ T-cell subset. We did not find significantly higher B-cell levels in males than in females, as was found by Bartlett. There is no direct explanation for this; however, we had fewer adolescents in the younger age ranges, which may explain some of this difference. A number of studies have reported increases in CD4+ T cells in females . Studies of adults have identified some similar gender differences. Higher mean total CD3+ T-cell and mean total CD4+ T-cell counts have been observed in females than in males; in addition, lower CD16+ cells have also been noted in adult females than in males . Differences in immune cell numbers may be secondary to hormonal differences (androgens or estrogen or both). Gender may affect the immune system at many levels. Differences in sex steroid regulation and production are present in males and females early in development, with the most profound differences noted during adolescence with the onset of puberty. Pubertal hormonal changes occur under the control of the hypothalamic-pituitary-gonadal axis, which leads to increases in sex hormone levels . Gender has an effect on both B- and T-cell responses and the secretion of immune modulators, including cytokines and chemokines . Cell proliferation occurs under the influence of cytokines, and some differences in T-cell numbers between males and females may be related to differences in cytokine regulation . Whether the increased levels of circulating CD4+ T cells found in females in this study is secondary to the influence of specific cytokines requires further investigation. Sex hormones may influence the immune cell populations through several indirect pathways. Sex hormones may impact immune cell numbers through thymic pathways. Studies of castrated male mice demonstrate increased number of both CD4+ T cells and CD8+ T cells . In addition, evidence suggests that thymic involution is modulated by androgens . Sex hormones may also influence T cells directly through cell receptors for sex steroids, as has been found for CD8+ T cells . During adolescence, the hormonal influence on immune cell numbers may be most apparent; however, the impact of sex hormones on immunity may begin much earlier. There are detectable estrogen receptors in fetal thymus as well as fetal lymphoblast cells, suggesting that sex hormones may influence lymphocyte maturation in the thymus . Whether the interactions between lymphocyte receptors and sex hormones lead to differences in immune cells numbers later in life is not clear. Gender differences in circulating T-cell subsets most likely reflect a complex regulatory mechanism, some of which may be under the influence of gender differences in sex steroids. The impact of gender on the immune system is evident from gender differences in immune system diseases particularly autoimmune disorders. Systemic lupus erythematosis, rheumatoid arthritis, myasthenia gravis, and Hashimoto's disease all have a higher prevalence among women . We found no independent influences of race and/or ethnicity on any of the immunologic measures in our study. We are limited, however, by the relatively small numbers of Caucasion subjects in our cohort. Rates of new HIV infections in adolescents appear to be greatest in minority groups . Thus, racial or ethnic differences in immune responses would be important to determine in a more heterogeneous population. We also found limited influence of age on the T-cells subsets studied with the only independent impact of age on memory CD4+ T cells and CD8+ T cells. Again, our study is limited by small numbers of younger adolescent subjects. Studies in younger children show a significant impact of age on T-cell populations, with signficantly higher numbers of CD4+ T lymphocytes in young infants . Differences between adolescents and adults have been reported: adolescent females begin to demonstrate CD4+ T-cell levels that are increased compared to those observed in males, and females also begin to demonstrate CD8+ T cell levels that are lower (such levels in adults have also been reported) . Our numbers may be too small to demonstrate subtle differences in the transition from adolescence to adulthood. Our study is unique as a longitudinal analysis of gender, age, and race and/or ethnicity differences in specific immunologic cells in high-risk youth. Although we found some significant differences, our study has several limitations. First, as most of our youth were sexually mature at the time of entry into the study, we were unable to analyze the data by Tanner-stage subgroups. Second, the mechanisms leading to gender differences were beyond the scope of this study. Finally, our population was primarily minority; thus, the findings may not apply to all populations. However, understanding the impact of gender, race, and age on immune cell numbers in healthy individuals is essential for subsequent evaluation in disease states. APPENDIX : The following investigators, listed in order by the numbers of subjects enrolled at their respective clinics, are participating in this study: L. Friedman, L. Pall, D. Maturo, A. Pasquale (University of Miami); D. Futterman, D. Monte, M. Alovera-DeBellis, N. Hoffman, and S. Jackson (Montefiore Medical Center); D. Schwarz and coauthor B. Rudy (University of Pennsylvania and the Children's Hospital of Philadelphia); M. Tanney and A. Feldman (Children's Hospital of Philadelphia); M. Belzer, D. Tucker, C. Hosmer, K. Chung (Children's Hospital of Los Angeles); S. E. Abdalian, L. Green, C. McKendall, and L. Wenthold (Tulane Medical Center); L. J. D'Angelo, C. Trexler, C. Townsend-Akpan, R. Hagler, and J. A. Morrissy (Children's National Medical Center); L. Peralta, C. Ryder, S. Miller, and S. Calianno (University of Maryland); L. Henry-Reid and R. Camacho (Cook County Hospital/University of Chicago); M. Sturdevant, A. Howell, and J. E. Johnson (Children's Hospital, Birmingham); A. Puga, D. Cruz, and P. McLendon (Children's Diagnostic and Treatment Center); M. Sawyer, J. Tigner, and A. Simmonds (Emory University); P. Flynn, K. Lett, J. Dewey, S. Discenza (St. Jude Children's Research Hospital); L. Levin and M. Geiger (Mt. Sinai Medical Center); P. Stanford and F. Briggs (University of Medicine and Dentistry of New Jersey); and J. Birnbaum, M. Ramnarine, and V. Guarino (SUNY Health Science Center at Brooklyn). The following investigators have been responsible for the basic science agenda: C. Holland (Center for Virology, Immunology, and Infectious Disease, Children's Research Institute, Children's National Medical Center), A. B. Moscicki (University of California at San Francisco), D. A. Murphy (University of California at Los Angeles); S. H. Vermund (University of Alabama at Birmingham); P. Crowley-Nowick (The Fearing Laboratory, Brigham and Women's Hospital, and Harvard Medical School, Boston, Mass.), coauthor S. D. Douglas (University of Pennsylvania and the Children's Hospital of Philadelphia). Staff from sponsoring agencies include A. Rogers and A. Willoughby, (NICHD), K. Davenny and V. Smeriglio (NIDA), E. Matzen, (NIAID), B. Vitiello (NIMH). Backmatter: PMID- 12204952 TI - Comparison between the Skin Snip Test and Simple Dot Blot Assay as Potential Rapid Assessment Tools for Onchocerciasis in the Postcontrol Era in Ghana AB - Successful control of onchocerciasis through mass distribution of ivermectin needs to be coupled with reliable, sensitive, specific, yet affordable diagnostic methods to monitor and ensure the efficacy of such measures. The effort put into the development of diagnostic methods for onchocerciasis that can substitute for or work in combination with the present "gold standard," the skin snip test, has resulted in the discovery of a number of immunogenic proteins with potential use as diagnostic tools in the postcontrol era. Most of these proteins have now been produced through recombinant DNA techniques. However, when costs are not a trivial issue, none of them have yet found their way into the areas where the disease still exists. In the present study, we have evaluated the performance of a simple dot blot assay which uses a mixture of native proteins designated PakF as a serious contender in the quest for a less invasive and more sensitive method to detect Onchocerca volvulus infection in areas with diverse endemicities. Our results indicate that the assay we propose is more sensitive than the skin snip test and shows high specificity, both characteristics required for a suitable tool for the monitoring of onchocerciasis in the postcontrol era. Keywords: Introduction : Onchocerciasis (river blindness) occurs in 34 countries in Africa, Latin America, and the Arabian Peninsula. The vast majority of the estimated 18 million persons infected with the causative agent, the parasite Onchocerca volvulus, live in Africa . One of the goals of the World Health Organization is to eliminate onchocerciasis as a public health problem. The classical method to detect O. volvulus infection consists of examining small skin biopsy specimens (skin snips) by microscopy for the presence of O. volvulus microfilariae (mf). However, this very specific method is invasive. The process may be uncomfortable, which may result in poor compliance from individuals, especially in communities subjected to repeated tests . Much effort has been put into the development of less invasive serodiagnostic methods to support the initiatives aimed at elimination of the disease. The program in West Africa, known as the Onchocerciasis Control Program (OCP), has been successful in meeting its goal of eliminating onchocerciasis as a public health problem in the 11 countries where the program has been active . As a result of the OCP's success, the efforts of the program are now turning to surveillance, with the goal of rapidly detecting and controlling outbreaks of infection in the onchocerciasis-free zones . In the postivermectin era, however, there is still no diagnostic test available for large-scale use in onchocerciasis-free zones or areas where the disease is still endemic, although with very low mf density. We have previously described a simple dot blot assay (DBA) using a native low-molecular-weight antigen fraction of female O. volvulus parasites, designated PakF , which is a useful diagnostic test in settings where the disease is nonendemic , and with its high specificity and sensitivity, it could have potential as a screening tool for onchocerciasis under certain circumstances. The DBA using this antigenic fraction (PakF-DBA) showed very high sensitivity and specificity when tested in Guatemala (9a), a country where onchocerciasis is endemic but where no other human filariasis has been reported . In this study, we describe the evaluation of the PakF-DBA for sera from individuals from localities with diverse endemicities in the Volta region of Ghana and compare the results with those obtained using the skin snip test. This region of Ghana has pockets of high and low transmission of onchocerciasis, and mass treatment has been practiced in some of these areas. Screening for onchocerciasis in these areas has always been done using the skin snip test; however, repeated ivermectin treatment results in decreased mf load, making the skin snip test less sensitive and reliable. We also evaluate the potential cross-reactivity of the assay with other filarial infections and show that sera from patients with circulating Wuchereria bancrofti do not cross-react in this assay. We discuss the potential use of the PakF-DBA as a less expensive and therefore affordable method for rapid screening and monitoring of areas where control measures have been established. (G. Guzman conducted this research in partial fulfillment of the requirements for his Ph.D. thesis at Karolinska Institute, Stockholm, Sweden, 2002.) MATERIALS AND METHODS : Study population. | The study was carried out in five localities of the district of Ho in the Volta region of Ghana with diverse endemicities (hypo- and hyperendemicity). Thirty individuals, who volunteered to participate, were selected from each locality, half of whom were within the age range of 5 to 17 years. Informed consent was obtained from all the participants in the study or their parents or guardians. The localities were as follows: (i) Aflapke (LOC1; male-to-female ratio, 0.9; child-to-adult ratio, 1), (ii) Klave (LOC2; male-to-female ratio, 0.9; child-to-adult ratio, 1), (iii) Aboete, Anorme, Asifiafe, and Luvudo (LOC3; male-to-female ratio, 2.3; child-to-adult ratio, 1), (iv) Honuta Gborgame (LOC4; male-to-female ratio, 0.8; child-to-adult ratio, 1), and Kpedze Anoe (LOC5; male-to-female ratio, 1.0; child-to-adult ratio, 1). Study samples. | Capillary blood was obtained by pricking the middle finger with a disposable hemolancet. Five circles of blood were blotted onto previously labeled 90-mm-diameter round, fast-drying filter paper (no. 597; Schleicher & Schuell, Dassel, Germany). When the blood spots were dry, filters were placed inside labeled plastic zip-lock bags and stored in the freezer compartment of a refrigerator (approximate temperature, -20C). Blood collection was done according to the ethical and safety policies in Ghana. Pools of sera from mf+ and nonexposed Guatemalan individuals from a previous study were used as positive and negative controls, respectively. A panel of Indian sera from areas where W. bancrofti is endemic, including 9 microfilaremic patients (W. bancrofti mf+; age range, 10 to 40 years) and 16 patients with chronic infection but no circulating mf in the blood (chronic patients [CP]; age range, 20 to 50 years), were used as other filaria controls. In addition, 15 individuals living in the same area with no evidence of filariasis (endemic controls [EN]) were included. These sera were from an area where onchocerciasis is not reported. Diagnosis of onchocerciasis. | Parasitological examination for O. volvulus was done by four-site skin snips and microscopic examination for the presence of mf following 24-h incubation in saline at room temperature (approximately 20C). The presence of mf of O. volvulus in any of the skin snips confirmed the diagnosis. Skilled technicians could distinguish O. volvulus from other filaria, such as Streptocerca, which is prevalent in the study area. The microfilarial density (MFD) was determined by dividing the average number of mf by the average weight of the skin snips in milligrams. Thus, an mf+ diagnosis corresponded to an individual with an MFD of >0. Antigen preparation (TSF). | Intact noncalcified nodules obtained in a previous study were digested by the collagenase method of Schulz-Key et al. . Female worms were washed and frozen at -70C in RPMI 1640 medium containing gentamicin (0.2 mg/ml). Tris-soluble antigen fractions (TSF) were prepared as previously described and stored at -20C . The total protein concentration was estimated by the Bradford assay . IEC. | PakF fractions were prepared separately from TSF by ion-exchange chromatography (IEC) as described previously, except that neither a fast protein liquid chromatography system nor a peristaltic pump was used . We simplified the IEC preparation by using a handheld ion-exchange column as described elsewhere (unpublished data). One hundred and twelve micrograms of TSF (1 to 1.3 mug/mul) was diluted to 0.5 ml in equilibration buffer (50 mM Tris-HCl, pH 9.0, containing 0.05% Tween 20). This was passed manually through a 1-ml HiTrap-Q ionic column (Pharmacia, Uppsala, Sweden) at a flow rate of 0.5 ml/min using a 1-ml syringe as a reservoir. The sample was eluted with equilibration buffer at 0.5 ml/min, and the eluate was collected in eight 350-mul aliquots before the buffer was changed to equilibration buffer containing 0.5 M NaCl. Two final 350-mul aliquots were collected using this buffer. Each fraction was tested for antigenicity and polypeptide composition using a DBA (see below) and one-dimensional gel electrophoresis as previously described . Serum elution from blood spots. | Blood was extracted as previously described . Briefly, one disk, 6 mm in diameter, was cut out from a blood blot on each filter and extracted for 2 h in 2 ml of phosphate-buffered saline (PBS; pH 7.4) containing 0.05% Tween 20 (PBST). After slow vortexing, the disk was removed and the eluted serum was collected. In some instances, eluted sera were tested before and after heat treatment at 56C for 30 min. DBA. | The DBA was performed as described previously with some modifications . Briefly, nitrocellulose strips, 4 mm wide and 3 cm long, were placed in mini-incubation trays (Bio-Rad Laboratories, Richmond, Calif.), and droplets (0.5 mul) of unconcentrated and undiluted PakF (approximately 5 ng) and TSF (64 ng; positive control) were adsorbed on the strips. The strips were allowed to dry at room temperature and were then blocked with 1% nonfat milk powder in PBST for 1 h under gentle agitation on a rocking shaker. The strips were incubated for 1 h with 1 ml of eluted serum. The strips were washed three times for 5 min each time with PBST and then incubated for 1 h with alkaline phosphatase-labeled goat anti-human immunoglobulin G (Calbiochem, La Jolla, Calif.) diluted 1:2,000 in PBST. The strips were washed three times as before with PBST and twice with PBS only before development of the immunoreactive spots with BCIP (5-bromo-4-chloro-3-indolylphosphate)-nitroblue tetrazolium (Sigma Chemical Co., St. Louis, Mo.). The color reaction was allowed to proceed for 10 min and then was stopped with several washes with distilled water. The strips were air dried in the dark. The DBA was performed blind, and thus the results from the skin snips were not made available to the person performing the assay until after all the analyses had been done. Control sera were tested first, diluted 1:4,000 in PBST. Evaluation of the dot blots. | Nitrocellulose strips were evaluated by visual examination and, for purposes of quantification, also by densitometry. Densitometric values were obtained as the integrated intensity of all the pixels in a spot excluding the background and were expressed as arbitrary units (Bio Image, Ann Arbor, Mich.). A visually positive sample was determined by identifying two spots of reactivity where PakF was applied to the nitrocellulose. The absence of reactivity in those spots was assessed as a negative result . Analysis of results. | Prevalence was defined as the number of skin snip-positive (MFD > 0) and DBA-positive (by visual examination) individuals in each locality studied. The sensitivity and positive predictive values of the PakF-DBA were calculated by comparison to the skin snip as a "gold standard." Sensitivity was calculated as the proportion of positive individuals detected by both methods to those detected by the skin snip test. Similarly, the positive and negative predictive values were calculated as the proportions of positive and negative individuals by both methods to those positive or negative by DBA . Statistics. | Pearson's correlation coefficients (Pcc) were used to compare the prevalences of the skin snip test and PakF-DBA, and the significance was tested using the Fisher z transformation. FIG. 1. | (A) Reproducibility in the production of PakF. (A) Reproducibility in the production of PakF. Lanes 1 to 4 show the electrophoretic patterns of PakF fractions obtained at different times, showing the mixture of at least five distinct bands that can be detected by silver staining. (B) Clear-cut difference between positive (+) and negative (-) samples in the PakF-DBA test. In a typical onchocerciasis-positive sample, the two spots where PakF had been adsorbed, as well as the procedural control (TSF), are recognized by the serum sample. Negative samples do not show reactivity against PakF, although they may still react against some proteins present in the whole soluble extract, TSF. (C) Graphic representation of the densitometric data obtained by measuring the densities of the spots in all the individual DBA strips representing the five localities under study. A line at 0.1 U of integrated intensity indicates the cutoff value for visually negative samples. RESULTS : Skin snip test results. | Of the 150 individuals tested, 99 were skin snip positive and 51 were skin snip negative. A cluster of 40 negative individuals was found in the age range 5 to 20 years, and the remaining 11 were distributed between 21- and 60-year-olds . No Streptocerca mf were detected in the skin snips of any of the participants in the study. The prevalence of individuals with mf+ skin snips varied from 43.3 to 76.7% in the five localities . A closer evaluation of the <18-year-old age group shows that only one of the 15 individuals in LOC2 had a positive skin test, resulting in a prevalence of 6.7% . In this age group, the highest prevalence was in locality LOC3. DBA results. | Figure shows the reproducibilities in the production of PakF at different times. Figure shows the result of a positive DBA test. For a sample to be DBA positive, the two dots where PakF has been adsorbed need to show reactivity. TSF, included as a procedural control, may or not be reactive depending on the degree of cross-reactivity of the proteins in the whole soluble extract. In the case of onchocerciasis, infected individuals normally recognize TSF even if they do not recognize PakF, as they may react to other proteins present in the whole extract. In order to represent these results in a more graphic manner, densitometric values were generated for each sample, obtaining the mean of the integrated intensity of each pair of spots on each DBA strip. Using this system, we were able to assign a numeric value of intensity to spots which were visually weak or very weak. Those visually negative fell in the intensity range of 0 to 0.1, and those visually positive had an integrated intensity above 0.1. This allowed us to generate a graph of integrated intensities for the entire study population . Figures and show the results obtained by DBA in all the areas under study. While the individual intensities of the spots varied, there was a clear-cut difference between positive and negative controls. In total, 125 samples were positive by DBA and 25 were negative, with a cluster of 22 DBA-negative samples within the age range 5 to 17 years. Individual analysis of each locality showed that the majority of positive individuals in the age range 5 to 17 years were inhabitants of the villages grouped as LOC3 and those from LOC5 . The MFD in relation to age followed a distribution pattern very similar to that of the DBA versus age (not shown). Comparison of DBA and skin snip results. | Tables and show the comparisons of the DBA and the skin snip test results, using visual assessment of the strips, for the whole study population and those aged 5 to 17 years, respectively. Stratified by age , the results clearly show that the DBA results closely followed the skin snip results in all age groups. However, the DBA consistently showed more positive individuals in all groups except the older age groups, with the biggest discrepancy between the DBA and skin snip test seen in the younger age groups. A closer look at the 5-to-17-year-old age group in the individual localities showed a similar trend between the DBA and skin snip test, with LOC2 having the lowest DBA positivity but LOC5 showing DBA prevalence akin to that of LOC3. The PakF-DBA detected more positive individuals than the skin snip test. Moreover, 98 out of the 99 skin snip-positive individuals were detected by DBA, for a sensitivity of 99% and a predictive value of 78%. A positive correlation (Pcc = 0.899; P < 0.05) between the prevalences estimated by the skin snip test and PakF-DBA was found when the whole study population was analyzed, which did not vary appreciably when the 5-to-17-year-old population was analyzed separately (Pcc = 0.893; P > 0.05). There was no significant correlation (r = 0.102; P > 0.05) between the prevalences estimated by the two methods among those older than 18 years. Cross-reactivity of PakF-DBA with other filarial sera. | The specificity of the PakF-DBA was assessed using sera from Indian individuals with bancroftian infection with apparently no exposure to O. volvulus. None of the W. bancrofti mf+ sera or the control sera were positive by DBA. However, two CP and two EN sera reacted positively on the DBA . Pooled Guatemalan mf+ and nonexposed sera, used as positive and negative controls, were positive and negative, respectively, when tested at the same dilution. In total, 4 out of 40 Indian sera were positive by DBA, resulting in a 90% specificity of the assay. FIG. 2. | DBA results obtained from serum eluted from finger prick blood, using PakF as an antigen. DBA results obtained from serum eluted from finger prick blood, using PakF as an antigen. Each strip represents an individual serum. The picture shows the immune reactivity of each individual to duplicate spots of PakF (top) and the unfractionated TSF (bottom). The samples are from LOC1, but the last two strips correspond to a pool of mf+ (GUmf+) and nonexposed (N-GU) controls, respectively. The strips were visually scored as positive (+) or negative (-) as indicated in the legend to Fig. . FIG. 3. | DBA results from LOC2, LOC3, LOC4, and LOC5. DBA results from LOC2, LOC3, LOC4, and LOC5. Visual assessment of the DBA strips was performed as indicated in the legend to Fig. . FIG. 4. | Specificity of PakF-DBA. Specificity of PakF-DBA. The graphic representation of densitometric data shows that the PakF-DBA is highly specific when sera from individuals with other filarial infections, other than onchocerciasis, are analyzed. Two W. bancrofti CP and two EN were positive by PakF-DBA, whereas none of the microfilaremic (MF) individuals reacted to PakF. Pools of Guatemalan O. volvulus-infected (GUmf+) and nonexposed (N-GU) samples were used as positive and negative controls, respectively. Visual assessment: samples classed as negative had an integrated intensity of <0.1 U, whereas those classed as positive had an integrated intensity of >0.1 U. TABLE 1 | Comparison of PakF-DBA and skin snip test in diagnosis of onchocerciasis in different age ranges TABLE 2 | Performance of PakF-DBA compared to skin snip test for diagnosis of onchocerciasis in the whole study population (5 to 78 years old) TABLE 3 | Performance of PakF-DBA compared to skin snip test for diagnosis of onchocerciasis in the age range 5 to 17 years DISCUSSION : Since the advent of successful control of onchocerciasis through measures including mass distribution of ivermectin, the need for developing surveillance programs to detect recrudescence of infection in the OCP area, as well as safe, inexpensive, and rapid diagnostic tests for O. volvulus infection, has been highlighted. In the present study, we have evaluated the performance of a simple DBA (PakF-DBA) as a serious contender in the quest for a less invasive and more sensitive method to detect O. volvulus infection in areas with diverse endemicities. With the PakF-DBA, we were able to detect a trend similar to that of the skin snip test, but the PakF-DBA detected a higher proportion of infected individuals than the skin snip test. These included 98 out of 99 detected by the more invasive gold standard test. This could indicate that the sensitivity of the DBA described here is higher than that achieved with the skin snip test. Although we have used analytical densitometry to generate comprehensive graphs, we have shown that the results obtained by visual examination of the DBA strips correspond well with densitometric data (Guzman et al., submitted). The clear-cut difference between positive and negative samples allows for visual assessment instead of the use of a sophisticated apparatus such as a densitometer. This is of particular interest in areas where the use of delicate or sophisticated readout devices is not practical. Therefore, the visual results of the DBA strips were enough to establish a correlation between the prevalences estimated by the skin snip test and those estimated by DBA in all age groups. The fact that the DBA results were comparable to those obtained by skin snip tests indicates the possibility that the overall estimates of prevalence generated by DBA from a sentinel population could replace those generated by skin snips. The major difference in the number of positive individuals detected by DBA and skin snip test was observed in the groups aged 5 to 17 years . When analyzed separately, a positive and significant correlation between the prevalences estimated with the PakF-DBA and the skin snip test was found for the population aged 5 to 17 years. However, no such correlation was observed in the group of adults studied (>17 years old). Due to the long life span of the adult worms, the continuous exposure in areas with ongoing transmission of onchocerciasis, and the unknown duration of circulation of antibodies against the proteins contained in PakF, it is not likely that an antibody-based assay like the PakF-DBA would show a better performance than the skin snip test when an age group of >20 years is selected as a sentinel population to assess the efficacy of control measures. Our results are in accordance with those of other studies that have shown that the most useful indicator group for detecting recrudescence, as well as for determining the intensity of infection in a community, is the population 5 to 15 years of age . The PakF-DBA may be positive in more specimens due to a number of reasons, including higher sensitivity, lower specificity, or the capacity to detect even early (prepatent) infection. The possibility of detecting prepatent disease can only be evaluated in long-term longitudinal studies. Our results suggest that the PakF-DBA has a higher sensitivity coupled with high specificity. There is an emphasis in the World Health Organization strategy on the need for surveillance methods to be highly specific even at the cost of low sensitivity. The problem of specificity is that of the quality of the gold standard. If the gold standard has low sensitivity, then assays with higher sensitivities can be viewed as having low specificities. This is the problem with the highly specific but low-sensitivity skin snip test. However, we have previously established the specificity of the PakF-DBA assay to be 100% using a panel of sera from noninfected individuals from areas where no other filaria is present (Guatemala, in Central America) (Guzman et al., submitted). In the later studies, we used sera from individuals infected with other coendemic parasites in the area, mainly Ascaris lumbricoides, in order to rule out any cross-reactivity due to the carbohydrate-containing antigens widely shared among nematodes, primarily phosphocholine-containing antigens . In addition, we ruled out cross-reactivity with a series of sera from filaria-infected patients. In the present study, we have also approached this question by extending the pool of test samples with well-characterized sera from individuals with other filarial infections from areas where onchocerciasis is not endemic. The sensitivity of the PakF-DBA using these sera was 90%. Of importance, however, is the fact that those positive in the PakF-DBA were not the individuals with circulating bancroftian mf, for whom nine of nine sera were negative by the PakF-DBA test. Combined analysis using the present data and the additional panels of sera from areas where other filarias, but not O. volvulus, are endemic resulted in 96% specificity (Guzman et al., submitted). These results also suggest that the PakF-DBA could be useful in screening children and adolescents born after control measures have been established, since they form a desirable sentinel population to detect recrudescence of infection in controlled areas . In the areas of endemicity, the need for assays that provide rapid results is advocated. The time needed to obtain qualitative results from the PakF-DBA was comparable to that for the rapid version of the skin snip test, i.e., 3 h. Thus, although a quantification of the level of infection is beyond the capabilities of the DBA, in terms of a rapid assessment for surveillance purposes, it represents an affordable alternative that offers more sensitivity than the skin snip test. Moreover, the test can be performed using only a few drops of finger prick blood, and this represents an advantage in terms of a higher compliance from the individuals in areas under constant screening to evaluate the impact of control measures. Several promising diagnostic methods, based on detection of specific antibodies, parasite DNA, or parasite antigens in clinical specimens, have been developed for onchocerciasis in recent years . However, none of these have been implemented under field conditions in the areas of endemicity. Antibody tests using recombinant antigens have been based on standard, indirect enzyme-linked immunosorbent assays, which can be performed in a simple laboratory. The antigen and DNA detection methods are performed over 2 or 3 days, and they require more, in terms of laboratory infrastructure, equipment, and expensive reagents, than the enzyme-linked immunosorbent assay . In either case, the costs involved in both the acquisition of instrumentation and the transfer of the technology are not trivial, thus making the development of less expensive, stable, and reliable methods a necessity in low-income countries where onchocerciasis is being controlled or still exists. Simple yet reliable approaches like the use of a less expensive diagnostic technique that could be implemented under field conditions are of particular importance, especially when the successor to OCP, the African Program of Onchocerciasis Control, is facing funding difficulties which could threaten its future success . The PakF-DBA described here, unlike antigen and DNA detection methods, does not require pretreatment of samples or expensive supplies and equipment (e.g., PCR). We have evaluated the PakF-DBA in heat-inactivated eluted sera (56C; 30 min) and found no difference in intensity between heated and unheated sera (data not shown). In addition, we have observed that PakF is highly stable at working temperatures normally found in the areas of endemicity, and it does not require the use of anti-immunoglobulin G4 conjugates to reduce the background, since the molecules comprising PakF do not seem to be recognized by such an isotype (Guzman et al., submitted). A possible disadvantage, compared to antigen or DNA detection methods, is that the PakF-DBA does not distinguish between past and present infections, but this is a problem of all antibody-based assays. Nevertheless, the PakF-DBA may still be a suitable tool for following changes in transmission when used in younger age groups. Furthermore, antibody testing still holds great promise as a means for monitoring changes in the transmission of O. volvulus after mass treatment of populations . Criticisms of the use of native human-derived material in the PakF-DBA have previously been made in terms of reproducibility of material and safety. We have made many preparations of PakF, and the reproducibility in performance in the DBA has been high. A reliable, sensitive, and specific recombinant protein would be valuable, but in its absence, native proteins still have a role. Where the question of possible exposure to contaminant viruses is concerned, PakF, which is a protein fraction stable at high temperatures, may be heat treated before use (Guzman et al., submitted). In addition, the PakF-DBA can be used with heat-inactivated sera. Concern about the availability of parasite material to prepare PakF is justified. However, in areas where no control measures have been established, the relative abundance of nodules from which the worms are freed makes it possible to obtain large amounts of PakF to be used in rapid screening of sentinel populations. In our hands, the amount of TSF obtained from a batch of one to three worms is enough to run a HiTrap-Q column up to 10 times. Each purification round produces 350 mul of PakF that can be applied to nitrocellulose strips without further treatment. The yield after 10 purification rounds (3.5 ml) allows the analysis of at least 3,000 samples, using the format of duplicate PakF spots per strip. Even if the format is changed to use 1 mul per spot, the number of samples that could be tested is large enough, considering that the material comes from only one to three worms. In the areas of endemicity, where nodulectomy is a common practice not only for control but also for aesthetic purposes, the availability of worms for PakF production would not be an insurmountable difficulty. We and others have shown variations in the protein compositions of O. volvulus parasites from different geographic areas . However, we have prepared PakF from worms isolated in both Ghana and Guatemala, used it in crossed analysis with sera from the two countries, and obtained similar results regardless of the origin of the parasite material (Guzman et al., submitted). Therefore, geographical variations would not be of concern in screening a particular area with parasite material obtained from worms from another region. In conclusion, the present study provides evidence that it is still possible to implement simple techniques for the monitoring of onchocerciasis which offer high sensitivity and specificity while being affordable in financially challenged areas. Where costs are an issue, and in the absence of more elaborate diagnostic tests in the areas of endemicity, simple tests like the PakF-DBA could still deliver reliable information that can be used to monitor changes in the transmission of onchocerciasis. Backmatter: PMID- 12204955 TI - Specificities and Opsonophagocytic Activities of Antibodies to Pneumococcal Capsular Polysaccharides in Sera of Unimmunized Young Children AB - An enzyme immunoassay (EIA) for antibodies to pneumococcal capsular polysaccharides (Pnc PSs) detects in some cases antibodies that are cross-reactive within different Pnc PSs. Recently, it has been suggested that for detection of only serotype-specific antibodies, EIA can be modified by removing cross-reactive antibodies by absorption with an irrelevant PS, e.g., the type 22F PS. The opsonophagocytosis assay measures the functional activities of antibodies in vitro, and the results of that assay correlate with in vivo protection better than measurement of the antibody concentration by EIA. We compared these different methods for measuring antibodies to type 1, 6B, 11A, 14, 19F, and 23F Pnc PSs in the sera of unimmunized young children who had been monitored for pneumococcal carriage, acute otitis media, and acquisition of antibodies to Pnc PSs from 2 to 24 months of age. Serum samples with antibody increases after contact with a pneumococcus of a homologous serotype contained specific antibodies and often had opsonophagocytic activity (OPA) (20 of 46). In samples with antibody increases from children who had not had contact with a pneumococcus of a homologous serotype, the antibodies found to be type specific by conventional EIA were usually cross-reactive and infrequently had OPA (10 of 68). When type 22F PS absorption was used in the EIA, most of the false antibody increases were eliminated, but most of the true antibody increases were still detected and the association between the antibody concentration detected by EIA and OPA was improved. However, there were serotype-dependent differences in the frequency of OPA. Use of absorption with a heterologous PS in EIA should be encouraged, and both the specificity of EIA and the sensitivity of opsonophagocytic assays should be further evaluated and improved. Keywords: Introduction : Immunity against Streptococcus pneumoniae (pneumococcus) is mediated by phagocytosis in the presence of complement and antibodies to pneumococcal capsular polysaccharides (Pnc PSs) . The in vitro opsonophagocytic activities (OPAs) of serum antibodies are believed to represent the functional activities of the antibodies in vivo and thus to correlate with protective immunity . Enzyme immunoassay (EIA) for the measurement of the concentrations of antibodies to Pnc PSs has been widely used to measure immunity to pneumococci and the immunogenicities of pneumococcal vaccines. However, for the estimation of immunity, a good correlation between the concentration of immunoglobulin G (IgG) measured by EIA and the OPAs of antibodies is needed. The correlation between the two methods has been reasonably good with postimmunization serum samples from infants and adults . However, sera from unimmunized individuals may have lower OPAs than expected on the basis of the antibody concentration obtained by EIA . The Pnc PS preparations used in the present EIAs are contaminated with a common cell wall PS (CPS) , and antibodies to CPS should be absorbed to improve the specificity of the EIA . Recently, several investigators have reported that despite absorption with CPS, antibodies cross-reactive with several types of Pnc PSs are still measured by EIA . The reason for this cross-reactivity has not been confirmed. It has been suggested that the Pnc PS preparations used as EIA antigens contain impurities or cross-reactive epitopes common to many serotypes . Removal of the cross-reactive antibodies by absorption with an irrelevant heterologous PS, e.g., the type 22F PS, improves the correlation between the antibody concentration obtained by EIA and the OPA . Thus, type 22F PS absorption has been suggested as an additional step in EIAs for antibodies to Pnc PSs. Cross-reactive antibodies are found more often in the sera of unimmunized infants and adults than in the sera of infants and adults immunized with pneumococcal vaccines, suggesting that the majority of the antibodies induced by vaccination are Pnc PS type or group specific . The origin and development of the cross-reactive antibodies by age has not been studied. We have previously described the natural development of antibodies to Pnc PSs, as detected by EIA, during the first 2 years of life , and antibody responses in children with pneumococcal acute otitis media (AOM) in a Finnish Otitis Media (FinOM) Cohort Study . Pneumococcal carriage and AOM induced antibodies to the homologous Pnc PS, but there were serotype-specific differences . Low concentrations of antibodies were also produced after contact with pneumococci with heterologous serotypes and even with no detectable contact with pneumococci . In this study, we evaluated the specificities and OPAs of antibodies in selected serum samples from our previous study . We used a set of samples that had twofold or greater increases in antibody concentrations compared to the concentration in a sample taken 6 months earlier and with approximately 1 mug or more of anti-Pnc PS antibodies per ml, as measured by the conventional EIA. Such samples were selected from children with and without a previous contact with pneumococci of the homologous serotype. The effect of the type 22F PS absorption step in the EIA was also evaluated. MATERIALS AND METHODS : Study population and sera. | Serum samples for this study were selected from among the samples from participants in the FinOM Cohort Study used in our previous study . The study population consisted of 329 healthy Finnish children who were monitored prospectively in a special study clinic from 2 to 24 months of age . During scheduled visits at 2, 3, 4, 5, 6, 9, 12, 15, 18, and 24 months of age, interview data and nasopharyngeal (NP) swab specimens for detection of pneumococcal carriage were obtained. In addition, NP aspirates (NPAs) and middle ear fluid (MEF) samples were obtained from patients with respiratory infection and from patients with a diagnosis of AOM, respectively. Serum samples (5 ml of venous blood) were obtained from the children at scheduled visits at 6 (+-14 days), 12 (+-14 days), 18 (+-28 days), and 24 (+-28 days) months of age. The sera were stored at -20C. The NP swabs, NPAs, and MEF samples were collected and cultured as described previously for detection of pneumococcal carriage or the etiology of AOM. Pneumococci were identified and serotyped by standard methods . The development of antibodies by age has been reported earlier . Serum samples used in this study. | Selected serum samples taken at the scheduled visits at 18 or 24 months of age were used in this study. The following criteria were used for selection: a twofold or greater increase in the concentration of antibody to PS of type 1, 6B, 11A, 14, 19F, or 23F compared to the concentration in the sample taken during the previous scheduled visit (i.e., at 12 or 18 months, respectively) and the presence of an antibody concentration of approximately 1 mug/ml or greater. An antibody concentration of at least 1 mug/ml is needed to achieve detectable OPA in infant sera taken after vaccination . The serum samples were further grouped according to pneumococcal culture findings . Sera in the Pnc PS contact-positive (Pnc 6B+, Pnc 11A+, Pnc 14+, Pnc 19F+, and Pnc 23F+) groups were from children who had had at least one culture-confirmed contact with a pneumococcus of the indicated serotype (pneumococci of that serotype were cultured from NP swabs, NPAs, and/or MEF samples) between the two scheduled visits. Accordingly, sera in the Pnc PS contact-negative (Pnc 1-, Pnc 6B-, Pnc 11A-, Pnc 14-, Pnc 19F-, and Pnc 23F-) groups were from children who had not had any detectable contact with the respective serotype by the indicated age, despite the increase in antibody concentrations. These children may have had contacts with other pneumococcal serotypes or no contacts with pneumococci at all. Data were derived from a total of 82 serum samples from 75 children. Some samples were included more than once. PSs. | Capsular PSs of S. pneumoniae serotypes 1 (lot 1211368), 11A (lot 963596), 14 (lot 2020510), 19F (lot 2033178), 22F (lots 1450702 and 2045906), and 23F (lot 1417200) were obtained from American Type Culture Collection (ATCC; Manassas, Va.). Capsular PS of serotype 6B was received via collaboration with the National Institute of Public Health and the Environment (Bilthoven, The Netherlands). The CPS used for absorption of anti-CPS antibodies from the sera was from the Statens Serum Institut (Copenhagen, Denmark). EIA. | The conventional EIA was performed as described earlier by Kayhty et al. . The results are expressed as micrograms per milliliter, calculated on the basis of the officially assigned IgG concentrations in reference serum sample 89-SF . In addition to the conventional EIA, antibody concentrations were determined by EIA with an absorption step with type 22F PS; the lowest serum dilution was incubated with 10 mug of CPS per ml and 30 mug of type 22F PS per ml for 1 h at room temperature before the EIA. The interassay coefficient of variance for both EIA methods was <15%. Inhibition EIA. | Serum samples were diluted 1:100 in 10% fetal bovine serum (Life Technologies, Ltd., Paisley, Scotland) in phosphate-buffered saline containing 10 mug of CPS per ml and aliquoted and placed into three separate tubes. Either homologous (type 1, 6B, 11A, 14, 19F, or 23F) or heterologous (type 22F) Pnc PS (30 mug/ml) was added, and the tubes were incubated for 1 h at room temperature. A control tube containing only CPS was included. The concentration of 30 mug/ml was found to be optimal for absorption for the PS types studied . After absorption, EIA was performed as described above. The percent inhibition was calculated from the optical density values. Inhibition of antibody binding by <=20% was considered low, and inhibition of antibody binding by >=80% was considered effective. Opsonophagocytosis assay. | The OPAs of the antibodies were analyzed by measuring the killing of live pneumococci by differentiated HL-60 cells in the presence of serum antibody and complement. A modification of the method described by Romero-Steiner et al. was used. S. pneumoniae serotypes 1, 6B, 14, 19F, and 23F (reference strains received from the Centers for Disease Control and Prevention, Atlanta, Ga.) and serotype 11A (reference strain obtained from the World Health Organization Collaborating Center for Reference and Research on Streptococci, Prague, Czech Republic) were grown in Todd-Hewitt broth supplemented with 0.5% yeast extract and kept frozen (-70C) in aliquots in Todd-Hewitt broth with 15% glycerol. OPA is expressed as a titer that is the reciprocal of the serum dilution with 50% killing compared with the bacterial growth in the controls without serum. A titer of 4 was given to sera with undetectable OPAs. To demonstrate the specificities of opsonic antibodies, serum samples were preincubated with either CPS or homologous or heterologous Pnc PS (30 min at room temperature with 1 mg of PS or CPS in 1 ml of undiluted sera) prior to the assay for OPA. OPA was totally removed only from samples preincubated with the homologous Pnc PS. Statistical analyses. | The average antibody concentrations were expressed as geometric mean concentrations (GMCs) and the average OPAs were expressed as geometric mean OPAs (GMOPAs). TABLE 1 | Selection of serum samples from among the FinOM Cohort Study samples for the different groups RESULTS : In the whole FinOM Cohort Study population, the number of serum samples with twofold or greater increases in antibody concentrations that yielded an approximately 1 mug/ml or higher concentration in the sample obtained at 18 or 24 months of age depended on the serotype . Among the serum samples from children without a previous contact with the homologous serotype, increases in the concentrations of antibodies to type 1, 6B, 11A, 14, and 23F PSs were detected in 4 to 9% of serum samples and increases in the concentrations of antibodies to type 19F PS were detected in 20% of the serum samples . Among the children with antibody increases but without contact with homologous type 1, 6B, 11A, 14, 19F, or 23F pneumococci, 8 of 13, 7 of 8, 16 of 20, 7 of 10, 20 of 34, and 5 of 12, respectively, had had contact with other serotypes of pneumococci. Antibody increases occurred more frequently (11 to 70%) in children with contact with a homologous serotype . Specificities of antibodies and effect of type 22F PS absorption on the antibody concentration measured by EIA. | The specificities of antibodies to serotypes 1, 6B, 11A, 14, 19F, and 23F measured by EIA were analyzed by inhibition of antibody binding with the homologous PS (type 1, 6B, 11A, 14, 19F, or 23F) or a heterologous PS (type 22F). Serotype-specific antibodies were considered those that were not inhibited (<20%) by the heterologous PS but that were effectively inhibited (>80%) by the homologous PS. Cross-reactive, nonspecific antibodies were inhibited (>20%) by heterologous PS. The antibodies to serotypes 6B and 14 in the serum samples from the Pnc 6B+ and Pnc 14+ groups were highly specific. In most cases the level of inhibition by heterologous PSs was <20%, and it was never >80% . Furthermore, they were mainly inhibited only by the homologous PSs: >80% inhibition by homologous PSs was seen in three of four and five of six serum samples, respectively . However, poorer inhibition by homologous PSs was found in the serum samples from the Pnc 6B- and Pnc 14- groups. It should be noted that three of the eight serum samples in the Pnc 6B- group were from children who had had contact with serotype 6A, and in two of the three serum samples the antibodies to PS type 6B were specific. Consequently, in all groups the concentrations of antibodies to type 6B and type 14 PSs obtained either by the conventional EIA or by EIA with the type 22F PS absorption step were equal; in addition, neither the GMCs nor the antibody concentrations in the individual serum samples (data not shown) were affected by the use of type 22F PS absorption. Antibodies to serotype 11A were highly specific in the Pnc 11A+ group: heterologous inhibition of <20% was detected in all 16 serum samples and homologous inhibition of >80% was detected in 15 of the 16 serum samples . In the Pnc 11A- group, however, heterologous inhibition of >20% was detected in 9 of 17 serum samples. Consequently, the EIA with type 22F PS absorption gave lower antibody concentrations than the conventional EIA for the Pnc 11A- group but not the Pnc 11A+ group . Very similar observations were made for serotype 23F: antibodies were specific in the Pnc 23F+ group but mostly cross-reactive in the Pnc 23F- group. The type 22F PS absorption step in EIA markedly reduced the concentrations of antibodies to type 23F PS in the Pnc 23F- group but not the Pnc 23F+ group . Antibodies to type 19F PS had a complex pattern. The specificities of the antibodies were somewhat better in the Pnc 19F+ group than in the Pnc 19F- group; 7 of 13 and 4 of 13 serum samples, respectively, were specific and were not inhibited (>20%) by the heterologous PS . However, many serum samples showed antibody binding that could not be effectively (>80%) inhibited by either type 19F or type 22F PS. All the serum samples from the Pnc 19F- group had antibodies with poor specificities; 5 were effectively (>80%) inhibited by heterologous PS, in addition to the type 19F PS, and in 5 additional serum samples the level of inhibition by either the homologous PS or the heterologous PS was low. The EIA with the type 22F PS absorption step gave lower concentrations than the conventional EIA for both the Pnc 19F- and Pnc 19F+ groups. However, the difference was more remarkable in the Pnc 19F- group . Serotype 1 was not cultured from any of the samples during the study. In spite of this, 13 children had increased levels of antibodies to the type 1 Pnc PS, as measured by conventional EIA . The antibodies to serotype 1 were not specific: inhibition by the homologous serotype as well as inhibition by the heterologous serotype was >80% for all serum samples . The EIA with type 22F PS absorption detected practically no antibodies to the type 1 Pnc PS . OPAs of antibodies. | Higher OPAs against pneumococci of serotypes 6B, 11A, 14, and 23F were measured in the sera of children taken after contact with a homologous serotype than in the sera of children with no contact with a homologous serotype . Even though the concentrations of specific antibodies (determined by EIA with type 22F PS absorption) to type 6B and type 14 PSs were equal, the OPAs were higher and were found more frequently in the Pnc 6B+ and Pnc 14+ groups than in the Pnc 6B- and Pnc 14- groups . Although most (16 of 26) of the serum samples in the Pnc 19F+ and 19F- groups had concentrations of antibodies to type 19F PS of >=1 mug/ml, as measured by EIA with type 22F absorption, none of the samples had OPAs against the type 19F strain . Antibodies to the type 1 PS were removed from all serum samples by type 22F PS absorption. Accordingly, none of the serum samples had OPAs against the type 1 strain . Overall, the sera of children with contact with pneumococci of a homologous serotype had OPA (titers, >=8) more often than the sera of children without homologous contact (20 of 46 versus 10 of 68 for the pneumococcal serotype contact-positive versus the pneumococcal serotype contact-negative groups combined; Table ). Furthermore, the antibodies were more often specific in the samples of children with contact with pneumococci of the homologous serotype than in the samples of children without contact with pneumococci of the homologous serotype . Only 3 of the 45 samples with cross-reactive antibodies had OPAs, whereas 27 of the 69 samples with specific antibodies had OPAs . Association between antibody concentration and OPA. | The concentrations of antibodies to type 6B, 11A, 14, and 23F PSs in serum measured by the conventional EIA and the EIA with type 22F PS absorption were compared to the OPAs . For these analyses, serum samples in the pneumococcal serotype contact-positive group and the pneumococcal serotype contact-negative group were combined. Serotypes 1 and 19F were not included because no OPAs against these serotypes were detected . For serotypes 6B and 14, the two EIAs gave very similar antibody concentrations , and thus, the type of EIA did not influence the association of concentration and OPA . For serotypes 11A and 23F the cross-reactive nonfunctional antibodies were markedly inhibited by type 22F PS and the association between the concentrations and the OPAs of antibodies was improved when EIA with the type 22F PS absorption step was used to determine the antibody concentration . FIG. 1. | Association between IgG concentration (in micrograms per milliliter), as measured by conventional EIA or by EIA with type 22F PS absorption, and OPAs (as titers) of antibodies to type 6B, 11A, 14, and 23F Pnc PSs. Association between IgG concentration (in micrograms per milliliter), as measured by conventional EIA or by EIA with type 22F PS absorption, and OPAs (as titers) of antibodies to type 6B, 11A, 14, and 23F Pnc PSs. Serum samples were taken from unimmunized children at 18 or 24 months of age. Horizontal lines, a titer of 4 was given to sera with undetectable OPA; vertical lines, a concentration of 1.0 mug/ml was used as a cutoff for the EIAs. TABLE 2 | Number of serum samples with levels of inhibition of <20% and >80% with the indicated homologous and heterologous (type 22F) Pnc PSs, GMCs of antibody measured by conventional EIA or EIA with type 22F PS absorption, and GMOPAs of antibodies to Pnc PSs 1, 6B, 11A,14, 19F, and 23F in sera of unimmunized children in the different groups TABLE 3 | Numbers of serum samples taken at 18 and 24 months of age from unimmunized children in the different groups containing either specific or cross-reactive antibodies to type 1, 6B, 11A, 14, 19F, and 23F Pnc PSs and OPAs of the samples DISCUSSION : We have previously studied the natural development of antibodies to Pnc PSs and its association with previous contacts with pneumococci, nasopharyngeal carriage, or AOM in the FinOM Cohort Study population . Previous contacts were associated with increased concentrations of antibodies to the homologous serotype, but there were serotype-specific differences. However, the children developed low concentrations of antibodies to all the Pnc PS types studied even without contact with the homologous serotype . The reason seems to be the fact that the detection of antibodies to Pnc PSs is affected by the cross-reactivity of the Pnc PS antigens used in the present EIA . We studied here the specificities and functional activities, as measured by opsonophagocytosis assay, of these antibodies and correlated these characteristics with culture-confirmed contacts with pneumococci. We also evaluated the possibility of using the type 22F PS absorption step in EIA to improve the specificity. For the whole study population, antibody increases yielding >=1 mug of antibodies per ml in the children without contact with the homologous serotype were rare compared to the antibody increases in children with contact with the homologous serotype: 20 versus 29% for serotype 19F and 4 to 9% versus 11 to 70% for the other five serotypes studied. We included in this study serum samples taken at either 18 or 24 months of age; these ages of sample collection were chosen to exclude maternal antibodies. Sera were grouped according to the pneumococcal contacts that the children had had by the indicated age. In general, serum samples from children with contact with a homologous serotype contained serotype-specific antibodies. Moreover, these samples often had OPAs, although there were clear serotype-specific differences. In contrast, antibodies in serum samples from children without contacts with a homologous serotype were mostly cross-reactive and infrequently had OPAs. Previous studies by us and other investigators with sera from adults have shown that detection of antibody to serotype 14 by EIA is serotype specific, whereas detection of antibody to the other serotypes is hampered by cross-reactivity . The present data with sera from a pediatric population confirm that antibodies to type 14 PS are highly specific. All serum samples in the Pnc 14+ group also had OPA against type 14. In contrast, only a few serum samples in the Pnc 14- group had OPA against type 14. This is interesting since this difference in OPA between the Pnc 14+ and the Pnc 14- groups cannot be explained by differences in antibody concentrations or antibody specificity. The origins of the antibodies in samples in the Pnc 14- group are not known; the antibodies were not absorbed by type 22F PS, and in a few cases they also had OPA. It is possible that some contacts with pneumococci remained undetected or that the children had had contacts with bacteria with PSs similar to type 14, e.g., the type III PS from group B streptococci . In the case of serotype 14, the increase in antibody levels or the concentration alone was not as good a marker of previous contact with type 14 as OPA was. The results of the present study confirm that the Pnc PS type 6B preparation used here and in our previous studies (-) is specific. However, although the specificities of antibodies to PS 6B were good in both the Pnc 6B+ and the Pnc 6B- groups, only a few serum samples had OPA. This may be due to the overall low antibody concentrations that were insufficient for detection of OPA by the present opsonophagocytosis assay. A previous contact with serotype 6A, which is cross-reactive with 6B, explained the increases in the levels of antibodies to type 6B PS in three of the eight serum samples in the Pnc 6B- group, and in two of the three samples the antibodies specificities were high. However, the antibodies in these three samples had no OPA against type 6B, despite the presence of anti-type 6B antibody concentrations equal to those in the samples from the Pnc 6B+ group showing OPA. This is in accordance with data showing that antibodies to type 6A elicited by vaccines containing type 6B may not be functional against type 6A strains or that more antibodies to type 6B PS are needed for killing of a type 6A strain than for killing of a type 6B strain . Altogether 26 serum samples with >=1 mug of IgG against type 19F PS per ml, as detected by the conventional EIA, were analyzed, and none of these serum samples had OPA against type 19F. The specificities of the antibodies in samples from children without previous contact with type 19F were poor. However, despite the higher concentrations and better specificities of antibodies in sera from children with cultures positive for type 19F, the sera showed no OPA. The specificities of antibodies to type 19F, as determined by EIA, in general, were confusing; in some cases antibody binding could not be inhibited by either the homologous PS or the heterologous PS. Furthermore, high concentrations of antibodies to type 19F PS even in the sera of vaccinated children do not seem to offer protection against AOM . Also, higher anti-type 19F antibody concentrations than anti-type 6B antibody concentrations are needed in the sera of vaccinated children for OPA . We have previously described that antibodies to type 1 PS can be detected by conventional EIA in serum samples of children participating in the FinOM Cohort Study, even though type 1 pneumococcus was not detected in any of the NP, NPA, or MEF samples during the study . We have now shown that these antibodies are not specific for type 1 PS and therefore are probably produced by a non-type-specific antigenic stimulus. These antibodies possessed no OPA against type 1 pneumococci. Use of the type 22F PS absorption step in the EIA removed all antibodies to type 1 PS detected by the conventional EIA in all samples. The data obtained in this study highlight the need for heterologous type PS absorption in EIA, especially when samples from an unimmunized population are studied. Type 22F PS absorption did not have a notable effect on the concentrations of antibodies to type 6B and type 14 Pnc PSs. However, the type 6B antigen was especially chosen for this study , and the commercially available preparations seem to be more cross-reactive , indicating that, in general, the type 22F PS absorption step is also needed for the anti-type 6B EIA. Because the nature of the cross-reactivity is unknown , the specificity and amount of antibodies inhibited by type 22F PS cannot be determined. It is, however, probable that the antibodies removed are not serotype specific and, in principle, should not be detected. Data from us and others show that the type 22F PS absorption step increases the association between the antibody concentration obtained by EIA and OPA. Furthermore, the results of the present study suggest that if antibody detection by EIA is used to study the seroepidemiology of pneumococcal infections, an EIA with type 22F PS absorption instead of the conventional EIA is highly recommended. The present opsonophagocytosis assay requires approximately 1 mug of antibody per ml for detection of OPA in the sera of immunized infants . As shown here, the concentration of specific antibodies needed may be even higher in sera of unimmunized children and may be different for the different serotypes. In the present study, high concentrations of specific antibodies to type 11A and 19F PSs were insufficient for OPA, while lower concentrations were associated with OPA against type 14 and 23F strains. It is possible that in addition to the antibody concentration and specificity, other qualitative characteristics of the antibodies, e.g., antibody avidity, affect OPA . Contacts with pneumococci seem to elicit the development of specific and functional antibodies, and thus, detection of OPA in sera could be an indicator of a serotype-specific pneumococcal contact. However, the sensitivity of the assay for OPA described here should be improved to detect activity in the sera of unimmunized individuals. Backmatter: PMID- 12204964 TI - Affordable CD4+-T-Cell Counting by Flow Cytometry: CD45 Gating for Volumetric Analysis AB - The flow cytometers that are currently supported by industry provide accurate CD4+-T-cell counts for monitoring human immunodeficiency virus disease but remain unaffordable for routine service work under resource-poor conditions. We therefore combined volumetric flow cytometry (measuring absolute lymphocyte counts in unit volumes of blood) and simpler protocols with generic monoclonal antibodies (MAbs) to increase cost efficiency. Volumetric absolute counts were generated using CD45/CD4 and CD45/CD8 MAb combinations in two parallel tubes. The percentage values for the various subsets were also determined within the leukocyte and lymphocyte populations utilizing a fully automated protocol. The levels of agreement between the newly developed method and the present industry standards, including both volumetric and bead-based systems using a full MAb panel for subset analysis, were tested by Bland-Altman analyses. The limits of agreement for CD4 counts generated by the volumetric methods using either CD45/CD4 (in a single tube) or the full Trio MAb panel (in three tubes) on the CytoronAbsolute flow cytometer were between -29 and +46 cells/mm3 with very little bias for CD4 counts (in favor of the Trio method: +8 CD4+ lymphocytes/mm3; 0.38% of lymphocytes). The limits of agreement for absolute CD4 counts yielded by the volumetric CD45/CD4 method and the bead-based method were between -118 and +98 cells/mm3, again with a negligible bias (-10 CD4+ lymphocytes/mm3). In the volumetric method using CD45/CD8, the strongly CD8+ cells were gated and the levels of agreement with the full Trio showed a minor bias (in favor of the Trio; +40 CD8+ cells/mm3; 5.2% of lymphocytes) without a significant influence on CD4/CD8 ratios. One trained flow cytometrist was able to process 300 to 400 stained tubes per day. This workload extrapolates to a throughput of >30,000 samples per year if both CD45/CD4 and CD45/CD8 stainings are performed for each patient or a throughput of >60,000 samples if only CD45/CD4 counts are tested in a single tube. Thus, on the basis of the high efficiency and excellent agreement with the present industry standards, volumetric flow cytometers with automated gating protocols and autobiosamplers, complemented by generic CD45, CD4, and CD8 MAbs used in two-color immunofluorescence, represent the most suitable arrangements for large regional laboratories in resource-poor settings. Keywords: Introduction : Dedicated flow cytometers are designed to enumerate the absolute numbers and percentages of lymphocyte populations, such as subsets of T cells, B cells, and NK cells. In the clinical service for monitoring human immunodeficiency virus (HIV) disease, the primary aim is to deliver absolute CD4+-T-cell counts, and this is achieved with a remarkably high level of precision . Nevertheless, the various cytometric systems differ in complexity . It has recently been documented that routine CD4-T-cell enumeration can be simplified without compromising quality , leading to cost-effective services for patients who receive generic antiretroviral drugs in resource-poor settings . Among the flow cytometers, dedicated instruments operating as "single platforms" are preferred due to their convenience and accuracy . These single platforms are based either on a volumetric principle by counting CD4+ T cells in a unit volume of blood or on the concept of adding known numbers of fluorospheres, or "microbeads," to each sample . These beads are, however, precision products that can increase running costs. Consequently, services handling large numbers of samples had to revert to "double platforms" operating a panleucogating strategy with CD45 monoclonal antibody (MAb) to secure a much less expensive but still accurate mode of operation . Indeed, the CD45-based gating, an example of the "heterogeneous" gating strategy, is a more reliable protocol when used with autogating in aging samples than the conventional gating strategies that utilize morphological scatter gates . Despite the present interest in improving the efficacy of routine flow cytometry , the performance of volumetric flow cytometric systems operating with CD45-based gating and generic MAbs has not yet been assessed. We have therefore investigated the following topics: (i) the agreement between the results of CD45/CD4 staining using simple panleucogating on volumetric single platforms and those obtained on the full volumetric and bead-based systems during CD4+-T-cell enumeration, including both absolute counts and CD4 percentage values (among leukocytes and lymphocytes); (ii) the increased sample throughput using CD45/CD4 staining; (iii) the extension of this protocol to include a second tube for CD45/CD8 staining in order to obtain CD4-plus-CD8 counts and CD4/CD8 ratios; and finally, (iv) the use of volumetric CD45 staining for generating absolute and differential counts for leukocyte subsets . Our study reveals the practical advantages of volumetric two-color flow cytometry with CD45/CD4 and CD45/CD8 staining using generic MAbs. Volumetric cytometers, equipped with biosamplers of high capacity, Microsoft Windows-based autogating software, and reporting systems, efficiently handle 300 to 400 samples during a working day. As many as 15 parameters, including CD4 and CD8 analysis together with hematological leukocyte differentials, can be generated for cost-efficient monitoring of HIV-infected patients in large regional laboratories. (Part of this research was presented at Monitoring and Diagnostic Tools for the Management of Antiretroviral Therapy in Resource-Poor Settings, a workshop held in Bethesda, Md., 11 to 13 November 2001, and arranged by Virology Education, B.V., Utrecht, The Netherlands.) FIG. 1. | Recent events leading to affordable CD4-T-cell enumeration by flow cytometry. Recent events leading to affordable CD4-T-cell enumeration by flow cytometry. NIBSC, National Institute for Biological Standards and Control; NEQAS, UK National External Quality Assessment Service, Sheffield, United Kingdom; QASI, Quality Assessment & Standardization for Immunology, Ottawa, Canada. MATERIALS AND METHODS : Clinical samples. | Samples (n = 93) were received for routine immunological diagnosis at an HIV-immunology laboratory and included patients at various stages of HIV infection as part of the routine diagnostic and quality assurance activity at the Royal Free Hospital. No extra specimens from HIV-seropositive patients were required. Twelve additional samples were taken from healthy volunteers 21 to 59 years of age as approved by the Institutional Ethics Committee . These whole-blood samples were collected in EDTA and analyzed within 24 h using a "lyse-no-wash" procedure . Briefly, in each tube, 10 or 20 mul of a diluted mixture of antibodies was admixed with 50 or 100 mul of whole blood, respectively. After 15 min of incubation at room temperature, 2.0 ml of lysing solution (0.17 M NH4Cl) was added. The samples were counted after a final 15-min incubation . Instrumentation. | Absolute lymphocyte subset counting was performed on three systems: (i) a CytoronAbsolute (Ortho Diagnostic Systems Inc., Raritan, N.J.) operating with Ortho Trio reagents and Immunocount II software to provide absolute counts by a volumetric method ; (ii) a FACSCalibur (Becton Dickinson Immunocytometry Systems, Oxford, United Kingdom) operating with TruCOUNT tubes preloaded with a known number of beads (46,295 beads per tube); cell concentrations were calculated with the formula (number of events in the region containing the cell population/number of events in the region containing beads) x (46,295/test volume [i.e., 50 mul]); and (iii) a CytoronAbsolute system operating with Immunocount II software using panleucogating in order to record three parameters: side scatter (SSc), green fluorescence (CD45-fluorescein isothiocyanate [FITC]), and orange fluorescence (CD4-phycoerythrin [PE] or CD8-PE). The volumetric absolute counting properties of the Cytoron were utilized with both systems i and iii. . The capacities of the automatic biosampler devices used were 100 (systems i and iii) and 40 (system ii) tubes. Total and differential white blood cell (WBC) counting was performed on a Bayer 120 hematology analyzer in the hospital's hematology laboratory as part of the routine service. Reagents. | During the volumetric-control procedure on the Ortho CytoronAbsolute, Ortho Trio MAbs were used . These included three tubes comprising in tube 1 isotype controls (immunoglobulin G1 [IgG1] plus IgG2a-FITC-IgG1 plus IgG2a-PE-IgG2a-PECy5), in tube 2 CD4(OKT4)-FITC-CD8(OKT8)-PE-CD3(OKT3)-PECy5, and in tube 3 CD16(3G8)-FITC-CD19(OKB9)-PE-CD3(OKT3)-PECy5 (original clone designations are shown in italics). During the bead-based control procedure on the FACSCalibur, TruCOUNT tubes were used in combination with MultiTEST reagents (Becton Dickinson Immunocytometry Systems) to obtain absolute CD4 counts . These included CD3(SK7)-FITC-CD8(SK1)-PE-CD45(2D1)-PerCP-CD4(SK3)-APC. The new CD45-based protocol was also based on the volumetric procedure performed on the Ortho Cytoron. Two tubes containing two-color immunofluorescence (IF) reagents were each tested. Tube 1 contained CD45(2D1)-FITC-CD4(RFT4)-PE, and tube 2 comprised CD45(2D1)-FITC-CD8(RFT8)-PE. These generic reagents are available in unconjugated form from the National Institute for Biological Standards and Control (Potters Bar, United Kingdom). Gating strategies for CD4 and CD8 enumeration. | On the Ortho Cytoron, we employed the Trio reagents and obtained absolute counts for the following cell types: T cells (CD3+; low side scatter), CD4+ T lymphocytes (CD3+ CD4+ CD8-), CD8+ T lymphocytes (CD3+ CD8+ CD4-), B cells (CD19+; low side scatter), NK cells (CD3- CD16+), and total lymphocytes (CD3+ T plus CD19+ B plus CD16+ NK cells referred to as Immunosum) . Percentage values for CD4+ T lymphocytes (CD4%) were derived as the number of CD4+ T cells divided by the total number of lymphocytes based on the criteria of CD4+, CD3+, and CD8- cells/Immunosum . The CD4/CD8 ratios were calculated as (CD3+ CD4+)/(CD3+ CD8+) values. The internal quality control for pipetting errors was based on CD3 replicates using Immunocount II software : samples for which the CD3 replicates differed from the average absolute CD3 count by >5% were automatically flagged for further inspection. The event threshold was set to operate on forward scatter. On the FACSCalibur, the gating strategy recommended by the manufacturer was used, with the threshold set for red fluorescence (CD45) in a single tube . For the new protocol, the gating strategy was based on CD45 panleucogating . A threshold was first set for green (CD45) fluorescence, and all WBCs were identified using a heterogeneous CD45/SSc dual-parameter histogram (CD45+ to CD45+++ in gate A). All WBC events in gate A were then sent to a CD4/SSc histogram, where CD4 T cells were counted (CD4++/SSc+ in gate E ). The same gating strategy was applied for CD8 counting in a second tube. Here, only lymphoid cells with bright CD8 expression were counted as CD8 T cells (CD8++/SSc+ in gate F . The CD4/CD8 ratios were calculated as (CD4++ SSc+)/(CD8++ SSc+) values. All these gating strategies were set to operate automatically and printed with all details . The internal quality control for pipetting errors was based on CD45 WBC replicates using the Immunocount II program. If the CD45 total WBC replicates differed from the average absolute CD45 count by >5%, the samples were flagged. All flagged samples or those where the operator had detected gating irregularities were subsequently reanalyzed. In the second stage of the analysis, the different CD45 staining intensities among the leukocyte populations were used to identify lymphocytes (CD45+++/SSc+ in gate B), monocytes (CD45++/SSc++ in gate C), and granulocytes (CD45+/SSc+++ in gate D). Using absolute counting and a two-color IF panel in two parallel tubes, the following 15 parameters were distinguished and stored: (i) total WBC counts, (ii) absolute CD4-T-cell counts, (iii) CD4-T-cell percentage among WBC, (iv) CD4-T-cell percentage among lymphocytes, (v) absolute CD8-T-cell counts, (vi) CD8-T-cell percentage among WBC, (vii) CD8-T-cell percentage among lymphocytes, (viii) absolute CD4- plus CD8-T-cell counts, (ix) CD4/CD8 ratio, (x) absolute lymphocyte counts, (xi) absolute monocyte counts, (xii) absolute granulocyte counts, (xiii) lymphocyte percentage among WBC, (xiv) monocyte percentage among WBC, and (xv) granulocyte percentage among WBC. In samples where a single tube was analyzed with CD45/CD4, 10 parameters (i to iv and x to xv) were recorded. Data handling and statistical analysis. | All results have been recorded in Microsoft Access-based spreadsheets. Following consultations with clinicians, forms were created for reports. Depending on the requests, these could include the single parameter of absolute CD4 count or all 10 to 15 parameters recorded above. After we tested whether the differences between the methods were normally distributed , Bland-Altman plots were used to investigate the agreement between the results obtained in two different systems, such as the panleucogating analysis on a volumetric flow cytometer versus a conventional "industry-standard" method. The standard techniques included the volumetric flow cytometer, CytoronAbsolute, using the full Trio panel, as well as the Becton Dickinson FACSCalibur running the TruCOUNT bead-based system. The absolute CD4+- and CD8+-lymphocyte counts and the percentages of these subsets among leukocytes and lymphocytes were studied. Bland-Altman (or bias) plots examine whether two methods have sufficient agreement to be used interchangeably. The average of values obtained by the two methods is plotted on the x axis, and the difference between the methods is plotted on the y axis. The average difference between the methods (bias), its 95% confidence intervals, and the limits of agreement (bias +- 2 standard deviations) are shown on the plots. The Pollock modification is identical to the Bland-Altman analysis except that the percentage difference is expressed between the compared methods, which is best suited to illustrate a systematic bias across a wide range of absolute counts. FIG. 2. | CD45/CD4 double staining on blood using panleucogating with volumetric absolute counting. CD45/CD4 double staining on blood using panleucogating with volumetric absolute counting. The graph report form is printed to document the autogating procedure. First, the CD45 side scatter histogram and gate A (all leukocytes) are established (left). All events of gate A are sent to the second display of CD4 side scatter (right). The CD4+ and lymphoid cells (in E) are automatically gated to provide the absolute CD4-T-cell count. E/A x 100 is the value of the CD4% among all leukocytes. The number of events in gate B (absolute lymphocyte count), gate C (absolute monocyte count), and gate D (absolute granulocyte count) are also defined. E/B x 100 is the CD4% value among lymphocytes. A parallel tube for CD45/CD8 double staining can also be run to provide the total of 15 parameters listed in Materials and Methods. QA, quality assurance; NEQAS, UK National External Quality Assessment Service. TABLE 1 | Age, sex, and HIV status of patient population studied RESULTS : CD4+-T-cell enumeration using primary CD45 and CD4 gating. | Total lymphocytes were identified by volumetric counting (i) as the sum of T cells, B cells, and NK cells (Immunosum) using Trio MAbs (referred to as a full Trio panel) and (ii) as cells with a bright CD45 expression and lymphoid side scatter in the CD45-based protocol. No significant systematic bias was observed between the two methods (bias = -8 lymphocytes/mm3 ). Next, CD4+-T-lymphocyte counts were determined by the full Trio panel and the CD45/CD4-based protocol, i.e., in the presence and absence of a CD3 reagent, respectively. The CD4% values among lymphocytes generated by the two methods showed a minimal bias of +0.38% (in favor of the Trio MAbs ). The absolute CD4-T-cell counts yielded by the two methods also showed excellent agreement (bias = +8 CD4+ cells/mm3; limits of agreement, between -29 and +46 CD4 cells/mm3 ). The agreement between the volumetric absolute CD4-T-cell counts using CD45/CD4 gating and the bead-based TruCOUNT tube was also determined. An average bias of -10 CD4 cells/mm3 was observed with widened limits of agreement (-118 and +98 CD4 cells/mm3), similar to the values previously observed between the standard volumetric and bead-based single-platform technologies . Efficiency of the CD45/CD4 gating protocol on a volumetric system. | After having documented CD4 enumeration using a CD45/CD4-based protocol by volumetric flow cytometer, we assessed the sample throughput of the system. A technician with a month of experience in flow cytometry processed 100 clinical samples using the autobiosampler. The steps of the procedure were monitored for time. A batch of 100 samples was processed in 95 min. The automated acquisition on the flow cytometer lasted for 120 min , allowing the operator a 25-min break before starting to prepare the following batch. Three batches of 100 samples could be prepared within the normal 8-h working day. An additional batch was also prepared at the end of the day for unattended acquisition during the late hours, to be ready for inspection by the next morning. In total, 300 to 400 clinical samples could be processed each day. CD8+-T-cell enumeration using primary CD45 and CD8 gating. | We next evaluated the agreement for CD8 enumeration between the CD45-based protocol and the full Trio panel. With the latter protocol, similar to the other currently recommended protocols, CD8+ T cells are counted when they coexpress both CD3 and CD8 molecules. Using the CD45-based protocol in the absence of a CD3 reagent, a tight gate was placed around lymphoid cells with bright CD8 expression (median, 130 x 103 antibody binding capacity per cell), excluding most NK cells (range, 10 x 103 to 110 x 103 antibody binding capacity per cell; median, 24 x 103 ABC per cell) from the CD8 gate. The correlation was excellent throughout the whole CD8 range , but a systematic bias was observed (+40 CD8 T cells/mm3; 95% confidence interval, +32 to +48) , representing +5.2% of CD8+ lymphocytes . The CD4/CD8 ratios were also evaluated in samples by the volumetric method, using the CD45/CD4 and CD45/CD8 protocols in two parallel tubes, and compared to the CD4/CD8 ratios observed with the full Trio panel . The agreements were good with virtually no bias (-0.05 CD4/CD8; limits of agreement, between -0.21 and +0.12 CD4/CD8) . We extended our study to investigate the agreement between the sum of the CD4+ and CD8+ T cells generated with the volumetric CD45 protocol and the T cells observed using CD3 staining in the full Trio protocol. The CD45-based protocol failed to identify CD3+ lymphoid cells doubly negative for CD4 and CD8 antigens (CD3+ CD4- CD8-). There was a bias of +125 T cells/mm3 in favor of the CD3+-T-cell counts recorded on the full Trio panel. These CD3+ CD4- CD8- T cells represented a 10.1% bias throughout the whole range of the T-cell counts . WBC subset enumeration using CD45-based protocols on a volumetric flow cytometer. | The expression of CD45 antigen is the common feature of all WBCs, and the CD45 staining intensity plus SSc distinguishes lymphocytes, monocytes, and granulocytes . We investigated the agreement between counts generated by this method and those yielded by a hematology analyzer in the routine hematology laboratory of our institution. The agreements for total WBCs, lymphocytes, and granulocytes were good, with a minimal bias of -136 WBCs/mm3, -70 lymphocytes/mm3, and +78 granulocytes/mm3 . However, the agreement for monocyte enumeration was poor. The hematology analyzer underestimated the monocyte counts with a bias of -179 monocytes/mm3. This is a large value, representing 35 to 40% of the total monocyte counts, as already reported for several hematology analyzers . FIG. 3. | Bland-Altman plots to establish the agreements between the volumetric CD45/CD4 protocol, single tube, and the "state-of-the-art" single-platform technology. Bland-Altman plots to establish the agreements between the volumetric CD45/CD4 protocol, single tube, and the "state-of-the-art" single-platform technology. The parameters studied were the total absolute lymphocyte counts (a), CD4-T-cell percentage values among lymphocytes (b), and absolute CD4-T-cell counts (c and d). The standard technologies used were the full lymphocyte subset panel (three tubes) tested with the Ortho Trio panel on the CytoronAbsolute (a, b, and c) and the TruCOUNT bead-based method (one tube) performed on a FACSCalibur (d). FIG. 4. | Evaluation of the agreement between the volumetric absolute CD8-T-cell counts generated with the CD45/CD8 protocol (CD8++ SSc+) and the full Trio panel (CD3+ CD8+). Evaluation of the agreement between the volumetric absolute CD8-T-cell counts generated with the CD45/CD8 protocol (CD8++ SSc+) and the full Trio panel (CD3+ CD8+). The results using linear regression (a) and the Bland-Altman plot (b) and its Pollock modification (c) are shown. In the Pollock modification, the differences between the two methods of counting CD8 T cells were expressed as a percentage of the total CD8 counts to illustrate the systematic bias at a 5% level. FIG. 5. | Bland-Altman plots (a and b) and the Pollock modification (c) to establish agreements on the Cytoron between the volumetric CD45/CD4-plus-CD45/CD8 two-tube protocol and the standard volumetric method using Trio reagents. Bland-Altman plots (a and b) and the Pollock modification (c) to establish agreements on the Cytoron between the volumetric CD45/CD4-plus-CD45/CD8 two-tube protocol and the standard volumetric method using Trio reagents. The parameters studied were the CD4/CD8 ratios (a) and the sum of the absolute CD4- plus CD8-T-cell counts versus the CD3+-T-cell counts (b and c, respectively). In the Pollock modification (c), the differences in total T-cell counts were expressed as percentages of T-cell counts to illustrate the regular underestimation of total CD3+-T-cell counts, at a 10% level, by the CD45 protocol. This bias is due to the existence of CD3+ CD4- CD8- T lymphoid cells. TABLE 2 | Timetable of routine operation using a biosampler with a 100-tube capacity DISCUSSION : The need to improve laboratory services for regions of the world where the HIV epidemic threatens to destroy the fabric of life has revitalized efforts to identify the most efficient techniques for monitoring HIV disease. The present changes relate to the common areas of routine immunology and hematology, such as quality assurance, sample processing, and transportation, as well as to the challenges of how to optimally count blood cells . Immunological methods, based on the specificities and discriminating capacities of MAbs, have recently made an impact by recognizing even minor subsets of functionally divergent blood cells . By using flow cytometry, the true power of directly identifying cells by antibodies, as opposed to first investigating merely their morphological features, is now documented, and the strategy of primary immunological gating is widely used . The relevant examples include CD45 for leukocytes and their subpopulations , CD3 for T cells , CD4 for the major T-cell subset and monocytes , and CD8 for the minor T-cell subset and some NK cells . Reliable total lymphocyte counts have been achieved by the Immunosum technique , providing the sum of the immunogated CD3+ (T), CD19+ (B), and CD16+ (NK) cells instead of using only the lymphocytic scatter appearance. The commonly used display on the cytometers is referred to as a heterogeneous, or morphospectral, protocol to show the IF of cells stained with MAb (on one axis) and the side scatter profile of cells (on the other axis) . In our study, we combined immunological CD45 gating with volumetric absolute counting on single platforms in order to introduce a robust method for WBC counting and for enumerating CD4+ and CD8+ T cells. Our four main findings are as follows. First, this study confirms our previous work, also performed on volumetric flow cytometers , as to the good agreement between the absolute CD4+-T-cells counts obtained by direct CD4 gating and by CD4+ CD3+ coexpression . Importantly, however, when we previously used CD4 MAb on its own without CD45, reliable CD4%-per-lymphocyte values could be obtained only with the constant vigilance of an experienced operator, who frequently had to manually modify the lymphocyte gates, a time-wasting procedure . We have now added CD45 gating to the protocol and report the excellent agreement between lymphocyte counts determined by CD45-side scatter and by the Immunosum method using the full Trio panel . Thus, CD45 staining improves the efficiency of the new autogated protocol , saving effort and technicians' time. Similarly, this gating strategy shows no bias compared to the bead-based CD4 counts but reveals occasional differences leading to a wider spread . This discrepancy might be a bead-related phenomenon, because similar results are seen when CD4 counts obtained by panleucogating on a double platform are compared to counts obtained by the bead-based method . The second conclusion is that the CD45/CD4 protocol on a volumetric cytometer provides an efficient system in which a trained flow cytometrist can run large numbers of tubes per day (>300 samples using CD45/CD4 alone ). If two parallel tubes are used with CD45/CD4- and CD45/CD8-double-stained cells, >150 blood samples can be studied. Obviously, such intensive diagnostic activity needs to be supplemented by clerical help and supervisory capacity. Nevertheless, this capacity illustrates the high efficiency of flow cytometry compared to that of manual methods such as the Dynabeads system , where a single assistant can manually handle only 15 to 20 samples per day. A hugely increased workload for CD4-T-cell enumeration is in line with the expected demand generated by the arrival of generic drugs for antiretroviral therapy. The larger regional centers dedicated to nationwide support with organized sample transportation using TransFix blood stabilizers will require this increased service capacity. The technical efficiency of this technology is directly related to three factors: (i) the fluent operation with a robust autogating process, where only <2 to 4% of samples need attention for regating (see above) , (ii) the use of an efficient autobiosampler , and (iii) a convenient system using a Windows environment and a Microsoft Access database for feedback to the clinicians. Flexible reporting, based on consultation with clinicians, may include only CD4 counts or any of the 15 parameters listed in Materials and Methods. The third finding of our study is related to the use of CD3, the specific T-cell marker. Arguably, CD3 is not required to identify CD4+ T cells . However, the CD8+-lymphocyte populations are more complex and display CD8 antigen over a wide range (15 x 103 to 140 x 103 CD8 molecules/cell ). The CD8+ cells include 80 to 92% proper CD8+ CD3+ T cells that display CD8 at a high level (CD8++; 80 x 103 to 140 x 103/cell) and 8 to 20% CD8+ CD3- NK cells that express CD8 at a lower level (CD8+; <80 x 103/cell). It is therefore logical to place a tight gate around the CD8++ population and compare these results with those obtained by counting CD3+-gated CD8 T cells . The results described above show that the CD8++ gate underestimates CD3+ CD8+ counts by 5.2% . This bias is apparently too modest to influence the CD4/CD8 ratios (bias, -0.05 ). An extra advantage of running both CD45/CD4 and CD45/CD8 tubes is the availability of CD4- plus CD8-T-cell counts that disregard the CD3+ CD4- CD8- T cells. We have argued elsewhere that these double-negative T cells represent a functionally different, mostly T-cell receptor alphabeta-negative subset that should not be included in the total T-cell counts . Finally, Loken et al. have documented the differential expression of CD45 antigen on lymphocytes, granulocytes, and CD14+ monocytes. In our study, the CD45 analysis is combined with volumetric counting in order to generate absolute leukocyte differential counts. These parameters, when defined on hematological counters, can be error prone , and the monocyte counts are frequently underestimated . On the other hand, the monocyte counts obtained by CD45 gating and carefully confirmed by the CD14 monocytic marker expression are more accurate. Consequently, the methods described above, in combination with the use of stabilized blood preparations with long shelf lives , will assist the establishment of long-awaited quality assurance schemes for leukocyte differentials and absolute counts, which are required to coordinate the performance of the wide variety of different hematology analyzers. In conclusion, the present volumetric CD45/CD4 flow cytometry, assisted by more affordable sources of MAbs, has wide applicability in the routine laboratories operating in economy-conscious environments. The specification required for the two-color IF plus side scatter used in this study is within the reach of the newly designed, battery-operated, smaller-volumetric-flow cytometers that carry red diodes or other small light sources as the sole source of light excitation and are also capable of performing bead-based enzyme-linked immunosorbent assays with the multiplexing technology in the area of the differential diagnosis of infectious diseases . TABLE 3 | Comparative performances of a hematology analyzer (Bayer 120) and the simplified CD45-based protocol on the CytoronAbsolute for absolute WBC enumeration Backmatter: PMID- 12204943 TI - Laboratory Diagnosis of Visceral Leishmaniasis AB - Keywords: null: null EPIDEMIOLOGY : Leishmania infections are worldwide in distribution: they are found in five continents. The disease is endemic in the tropical and subtropical regions of 88 countries. There are an estimated 12 million cases worldwide; 1.5 to 2 million new cases occur every year. Cutaneous forms are most common (1 to 1.5 million cases per year), representing 50 to 75% of all new cases, and 500,000 cases of VL occur every year . The geographical distribution of leishmaniasis is limited to the areas of natural distribution of the sandfly, the vector for the disease. Economic development, including widespread urbanization, deforestation, and development of newer settlements, besides migration from rural to urban areas, is responsible for the spread of the sandfly as well the reservoir system of leishmania . Moreover, the number of new host populations, i.e., populations of immunodeficient HIV-infected patients, is increasing, especially in southern Europe and Africa . Leishmania-HIV coinfection is regarded as an emerging disease especially in southern Europe, where 25 to 70% of adults with VL have AIDS as well; leishmaniasis behaves as an opportunistic infection, and it has been proposed that it be included as an AIDS-defining illness. Moreover, the presence of the leishmania parasite outside the reticuloendothelial system, e.g., in the peripheral blood, in HIV-infected patients makes these patients a reservoir and source of infection for the vectors. The parasite load in peripheral blood is generally so high that transmission among intravenous drug users by use of shared syringes has also been demonstrated . The resurgence of leishmaniasis, its emergence in newer geographical areas and in newer hosts, besides changing the clinical profile of infected patients, has put forward newer challenges in the areas of diagnosis, treatment, and disease control. PRINCIPLES FOR DIAGNOSIS OF LEISHMANIASIS : The diagnosis of VL is complex because its clinical features are shared by a host of other commonly occurring diseases, such as malaria, typhoid, and tuberculosis; many of these diseases can be present along with VL (in cases of coinfection); sequestration of the parasite in the spleen, bone marrow, or lymph nodes further complicates this issue. Laboratory diagnosis of leishmaniasis can be made by the following: (i) demonstration of parasite in tissues of relevance by light microscopic examination of the stained specimen, in vitro culture, or animal inoculation; (ii) detection of parasite DNA in tissue samples; or (iii) immunodiagnosis by detection of parasite antigen in tissue, blood, or urine samples, by detection of nonspecific or specific antileishmanial antibodies (immunoglobulin), or by assay for leishmania-specific cell-mediated immunity. Demonstration and isolation of parasite. | The commonly used method for diagnosing VL has been the demonstration of parasites in splenic or bone marrow aspirate. The presence of the parasite in lymph nodes, liver biopsy, or aspirate specimens or the buffy coat of peripheral blood can also be demonstrated. Amastigotes appear as round or oval bodies measuring 2 to 3 mum in length and are found intracellularly in monocytes and macrophages. In preparations stained with Giemsa or Leishman stain, the cytoplasm appears pale blue, with a relatively large nucleus that stains red. In the same plane as the nucleus, but at a right angle to it, is a deep red or violet rod-like body called a kinetoplast . After identification, parasite density can be scored microscopically by means of a logarithmic scale ranging from 0 (no parasite per 1,000 oil immersion fields) to +6 (>100 parasites per field) . The sensitivity of the bone marrow smear is about 60 to 85%. Splenic aspirate, though associated with risk of fatal hemorrhage in inexperienced hands, is one of the most valuable methods for diagnosis of kala-azar, with a sensitivity exceeding 95%. It requires no special equipment, from the patient's standpoint is generally preferable to the more painful bone marrow aspirate, and has proven to be safe and relatively easy to perform in experienced hands. For patients suspected to have VL, splenic aspirate can be performed even when spleen is not palpable, after demarcating the area of splenic dullness by percussion. The only risk of splenic puncture is bleeding from a soft and enlarged spleen. At our treatment center, fatal bleeding has occurred only twice in 9,612 splenic aspirate procedures performed over the last 10 years. To avoid the risk of excessive blood loss, splenic puncture should be avoided in patients with a platelet count of less than 40,000 platelets/mul and a prothrombin time of more than 5 s over the control. A tissue specimen, e.g., a spleen, liver, or lymph node tissue specimen, may be subjected to imprint cytology by the repeated pressing of its cut flat surface on microscopic slides. The smear is fixed with absolute alcohol and stained with Giemsa stain. In imprint cytology, a monolayer of cells is formed and amastigotes are easily identifiable. The results are expressed as the number of leishmania per 100 host cell nuclei. Tissue specimens can also be subjected to histology, and the presence of parasites can be demonstrated by standard hematoxylin and eosin stain. Tissue specimens are usually uneven in thickness; consequently the amastigotes are unevenly distributed. Long searches may be required to demonstrate the parasite. The sensitivity of the test can be increased by staining the specimen with fluorescent dye-tagged antibodies to the surface receptors of the parasite. Fluorescein isothiocyanate isomer- or rhodamide B isothiocyanate-conjugated antiserum is usually used for this purpose. Fluorescent dye-conjugated monoclonal antibodies are also used for speciation of the parasite. Culture of parasite can improve the sensitivity of detection of parasite, but leishmania culture is rarely needed in routine clinical practice. However, cultures are required for (i) obtaining a sufficient number of organisms to use an antigen for immunologic diagnosis and speciation, (ii) obtaining parasites to be used in inoculating susceptible experimental animals, (iii) in vitro screening of drugs, and (iv) accurate diagnosis of the infection with the organism (as a supplement to other methods or to provide a diagnosis when routine methods have failed). Leishmania strains can be maintained as promastigotes in artificial culture medium. The culture media used may be monophasic (Schneider's insect medium, M199, or Grace's medium) or diphasic (Novy-McNeal Nicolle medium and Tobies medium). We prefer diphasic medium containing modified diphasic rabbit blood agar overlaid with RPMI 1640 (Gibco BRL, Grand Island, N.Y.) for primary isolation, and we prefer M199 medium containing 20% fetal calf serum to amplify parasite numbers . Hockmeyer's medium, which is Schneider's commercially prepared culture medium supplemented with 30% heat-inactivated fetal calf serum with 100 IU of penicillin and 100 mug of streptomycin, is simple to use and satisfactory for diagnosis of VL, but it is expensive . Culture tubes are inoculated with 1 to 2 drops of bone marrow or splenic aspirate and incubated at a temperature between 22 and 28C. The tubes are examined weekly for the presence of promastigotes by phase-contrast microscopy or by wet mount of culture fluid for 4 weeks before being discarded as negative. If promastigotes are present, they are maintained by weekly passage to fresh medium. Blood can also be used to isolate the parasite, but the method is slow and takes longer. Aseptically collected blood (1 to 2 ml) is diluted with 10 ml of citrated saline, and the cellular deposit obtained after centrifugation is inoculated in culture media. Contamination of the culture media by bacteria or yeast species or other fungi usually complicates the culture but can be avoided by use of good sterile techniques and by the addition of penicillin (200 IU/ml) and streptomycin (200 mug/ml) to the medium (for bacteria), as well as 5-flucytosine (500 mug/ml) (as an antimycotic agent) . In vitro culture of the amastigotes is done for chemotherapeutic studies and to study the interrelationship of the amastigotes and macrophages. The amastigotes are grown in tissue or macrophage culture. These cell lines are produced from (i) human peripheral blood monocytes, after these are set apart by density sedimentation with lymphocyte separation medium (LSM; Organon-Teknika, Durham, N.C.), in which case a new batch of macrophages must be produced anew ; (ii) macrophage cell lines, e.g., P388D and J774G8 lines from mice; and (iii) dog sarcoma and hamster peritoneal exudates of cell lines, in which case continuous culture can be achieved . The parasite can also be demonstrated after inoculation of laboratory animals (such as hamsters, mice or guinea pigs) with infected specimen . Animal inoculation is not usually employed as a diagnostic test, since several months may be required to obtain a positive result. Golden hamster is the animal of choice for maintaining L. donovani complex . It can be infected via many routes, including across mucous membranes, but intraperitoneal and intrasplenic routes are preferred. Both amastigotes and promastigotes can infect the animal. After inoculation, the animal is examined weekly for signs of infection, such as cutaneous lesions, hepatosplenomegaly, or metastatic lesions. Amastigotes can be harvested by biopsy from the spleen and the liver of an animal that is under anesthesia and that is allowed to survive following the procedure as a source of infective parasite. In the absence of signs of obvious infection, the animal is generally sacrificed after 4 months, at which point liver and spleen samples are examined for the presence of the parasite. In areas of endemicity, recognition of species of leishmania is rarely required. However, identification of an organism to the species level is helpful epidemiologically and is also important for the treatment of and prognosis determination for global travelers who are not immune to the parasite and tend to develop unusual manifestations of the disease . Identification of species of the L. donovani complex is particularly difficult, because morphologically the species are almost indistinguishable from each other. For species-level identification, a large amount of promastigotes is obtained by culture of the organism and the species-specific isoenzyme pattern is analyzed by cellulose acetate electrophoresis . Typing of washed live promastigotes by direct agglutination test with species-specific monoclonal antibodies is another highly sensitive taxonomic tool frequently utilized for this purpose . Species-level identification can also be done by analysis of amplified minicircle kinetoplast DNA (KDNA), by choosing primers from conserved regions of different leishmania species KDNA minicircles . Yet another method used for identification of species of leishmania is the analysis of the in vitro promastigotes' released antigenic factors, which are different for different leishmanial species . Although demonstration of even a single amastigote upon microscopic examination of tissue smears or multiple promastigotes in cultures is considered sufficient for positive diagnosis of the disease, the sensitivity of the tissue examination, except in the case of splenic aspirate, is low. Moreover, the procedure(s) for obtaining tissue specimen(s) is traumatic and associated with considerable risk. Identification of amastigotes requires considerable expertise and training and is subject to the ability of the observer. Besides, culturing parasites is expensive and time consuming and requires expertise and costly equipment, severely restricting its use in routine clinical practice. DNA detection method. | Due to the limitations inherent in techniques used for detection of parasites, new approaches to the detection of parasites, such as DNA hybridization, have been attempted since the early 1980s. Although these methods had considerable sensitivity (detecting as few as 50 to 100 parasites) , their potential use in routine diagnosis is hampered by the complex procedure of hybridization. The development of PCR has provided a powerful approach to the application of molecular biology techniques to the diagnosis of leishmaniasis. Primers designed to amplify conserved sequences found in minicircles of KDNA of leishmanias of different species were tested in various tissues of relevance. Such a target was eminently suitable because the kinetoplast is known to possess thousands of copies of minicircle DNA. In recent years, PCR-based diagnostic methods with a wide range of sensitivities and specificities have been described . In a study reported from Sudan, PCR was found to be more sensitive than microscopy for the detection of Leishmania parasites in lymph node and bone marrow aspirations. However, its sensitivity for the detection of Leishmania DNA in the blood of parasitologically proven VL cases was only 70% . In another study reported from India, in which a species-specific primer for L. donovani (LDI primer) was used, the sensitivity of PCR with whole blood from VL patients was 96% and Leishmania DNA was detected in skin specimens from 45 of 48 patients with PKDL (sensitivity, 93.8%) . A PCR-enzyme-linked immunosorbent assay (ELISA) technique using a primer that was able to identify 33 L. infantum strains from 19 different zymodemes has been developed. It has a sensitivity higher than that of other diagnostic techniques, e.g., indirect fluorescent-antibody (IFA) test, parasite culture, or microscopy, and was able to detect a minimum of 0.1 promastigote or 1 fg of genomic material. This PCR-ELISA technique can potentially be used for diagnosis of VL from peripheral blood samples . PCR done from blood spots on filter paper can also be used as a screening test to identify Leishmania infection in immunocompromised patients with high parasite loads in peripheral blood. The sensitivity of this technique for detecting leishmania (75%) was considerably higher than the respective sensitivities of microscopy (26.3%) and blood culture (42.3%) (17). However, PCR assay with buffy coat preparations to detect Leishmania was 10 times more sensitive than that with whole-blood preparations, and particularly good results were obtained when proteinase K-based methods were used. Proteinase K-based PCR was able to detect 10 parasites/ml . A fluorescent DNA probe specific for a conserved region of the small subunit rRNA gene of Leishmania and a pair of flanking primers, when used for DNA amplification in one assay, proved to be a highly specific and rapid diagnostic modality to detect infection with Leishmania . Using this rapid fluorogenic PCR technique, DNA could be amplified from 27 strains of cultured Leishmania, and the turnaround time from fresh human tissue biopsy to test result was found to be less than 24 h . Besides being a highly sensitive and specific tool for diagnosis of both VL and PKDL and a useful method for species identification , PCR can also be used to distinguish between relapse and reinfection in treated VL patients. Restriction fragment length polymorphism analysis of the PCR-amplified minicircle of leishmanial DNA can be utilized for this purpose . PCR could also prove to be an important tool in assessing the success of VL treatment: of patients treated for VL who tested negative by PCR with lymph node tissue, none relapsed or developed PKDL, while more than half of patients who tested positive by PCR with lymph node tissue either relapsed or developed PKDL after apparent cure of disease following supervised treatment . On the other hand, a substantial number of the patients who tested positive by PCR, after apparent cure, did not relapse or develop PKDL, a result that suggests the limitation of PCR in deciding the end point of treatment. The PCR positivity observed in these patients may be due to nonviable parasite. Similarly, PCR results for healthy endemic controls may be positive , which may lead to the erroneous conclusion that they suffer from VL. In these healthy endemic controls, a combination of direct agglutination test (DAT) (which shows low titers in healthy endemic controls) and PCR may be helpful in defining the status of these patients. Immunodiagnosis. (i) Antigen detection. | Antigen detection is more specific than antibody-based immunodiagnostic tests . This method is also useful in the diagnosis of disease in cases where there is deficient antibody production (as in AIDS patients). De Colmenares et al. (20) from Spain have reported two polypeptide fractions of 72-75 kDa and 123 kDa in the urine of kala-azar patients. The sensitivities of the 72-75-kDa fractions were 96%, and the specificities were 100%. Besides, these antigens were not detectable within 3 weeks of anti-kala-azar treatment, suggesting that the test has a very good prognostic value . A new latex agglutination test (KATEX) for detecting leishmanial antigen in urine of patients with VL has showed sensitivities between 68 and 100% and a specificity of 100% in preliminary trials. The antigen is detected quite early during the infection and the results of animal experiments suggest that the amount of detectable antigen tends to decline rapidly following chemotherapy. The test performed better than any of the serological tests when compared to microscopy. Large field trials are under way to evaluate its utility for the diagnosis and prognosis of VL . (ii) Antibody detection. | For several decades, nonspecific methods, which depend upon raised globulin levels, have been used in the diagnosis of VL. Some of the tests used for detecting these nonspecific immunoglobulins are Napier's formol gel or aldehyde test and the Chopra antimony test. Since these tests depend upon raised globulin levels, results can be positive in a host of conditions . Lack of specificity, as well as varying sensitivities, renders them highly unreliable. Several immunodiagnostic methods which are more sensitive and specific have been developed. They are useful in identifying specific cases and can be used for community surveillance. The human body makes an attempt to fight against VL by producing some of the highest levels of antibodies found in response to any disease, all to no avail. This is due to polyclonal activation of the B cells, resulting in marked elevation of levels (in serum) of immunoglobulin G (IgG) and IgM against various nonspecific proteins and haptens . The consistent presence of high levels of antibodies against parasite antigens can simplify diagnosis of VL. Several serological techniques are based on detection of these antibodies. The specificity of the antibody depends upon the antigen or epitope used in the test, as the parasite stimulates production of a wide array of antibodies, including group-, genus-, and species-specific antibodies. Therefore, the sensitivity may depend upon the test and its methodology, but the specificity will depend on the antigen rather than the serological procedure used. In most serological tests, the sensitivity and specificity data are compared against demonstration of parasites in various tissues. Conventional methods for antibody detection included gel diffusion, complement fixation test, indirect hemagglutination test, IFA test, and countercurrent immunoelectrophoresis . However, aside from practical difficulties at peripheral laboratories, the sensitivities and specificities of most of the above tests have been the limiting factors. Except for the IFA test, which is used on a limited scale, these tests are rarely used for routine diagnosis of VL. In 1988, a modified DAT was reported to be useful in kala-azar and is being used in several countries of endemicity . In this test, the trypsinized whole promastigotes are formalin fixed and stained with Coomasie brilliant blue; serum from the patient is then incubated with the antigen, and agglutination is observed the next day. Use of an 0.8% concentration of 0.1 M 2-mercaptoethanol in the sample diluent further improves its performance . DAT in various studies has shown to be 91 to 100% sensitive and 72 to 100% specific . In Sudan, in specially set up field laboratories, Medecins Sans Frontieres uses DAT for diagnosis of VL; patients with high titers receive treatment, and a confirmatory parasitic diagnosis is done in those with low titers . From India, several laboratories reported satisfactory sensitivity and specificity levels for this test . Although DAT showed a high degree of repeatability within the centers, its reproducibility across the centers was quite weak . Moreover, difficult field conditions, the fragility of aqueous antigen, the lack of cold chain, and batch-to-batch variations in the antigen, along with the nonstandardization of test readings, have severely limited its widespread applicability in regions of endemicity. Freeze-dried antigens developed in Belgian and Dutch laboratories are likely to overcome some of these handicaps . Unless this improved antigen is produced indigenously to make it affordable and DAT is made user friendly with one-step dilution and reduced incubation time, its field use is unlikely in countries of endemicity like India. Like most antibody-based tests, DAT may yield positive results for a long time after complete cure and thus has not proved to be of much prognostic value . ELISA has been used as a potential serodiagnostic tool for almost all infectious diseases, including leishmaniasis. The technique is highly sensitive, but its specificity depends upon the antigen used. Several antigens have been tried. The commonly used antigen is a crude soluble antigen (CSA). It is prepared by repeated freezing and thawing (four to six cycles) of a suspension of promastigotes in phosphate-buffered saline, followed by cold centrifugation at 10,000 to 20,000 x g. The supernatant is used as soluble antigen and is used to coat ELISA plates after estimation of protein content (100 to 5,000 ng/ml). The sensitivity of ELISA using these concentrations of CSA is reported to range from 80 to 100%, but cross-reactions with sera from patients with trypanosomiasis, tuberculosis, and toxoplasmosis have been recorded . On the other hand, when various selective antigenic masses (116 kDa, 72 kDa, and 66 kDa) were used, a specificity of 100% could be achieved, but only at the cost of sensitivity, which went down to as low as 37.5% . Palatnik-de-Souza et al. described the use of fucose-mannose ligand as the antigenic molecule. It is a 36-kDa glycoprotein present throughout the life cycle of leishmania (amastigote and promastigote stages). Its use in ELISA has been found to result in 100% sensitivity and 96% specificity . In a recent study, it was found that the sensitivity and specificity of ELISA in diagnosing VL could also be increased by the use of soluble antigens derived from promastigotes cultivated in a protein-free medium. One study, done with 129 VL and 143 cutaneous leishmaniasis patients, showed a sensitivity of 95% . A recombinant antigen, rK39, has been shown to be specific for antibodies in patients with VL caused by members of the L. donovani complex . This antigen, which is conserved in the kinesin region, is highly sensitive and predictive of the onset of acute disease. The antigen is derived from L. chagasi, which in the United States is used for veterinary purposes, though it is not approved for human use. High antibody titers in immunocompetent patients with VL have been demonstrated. This antigen has been reported to be 100% sensitive and 100% specific in the diagnosis of VL and PKDL by ELISA . Another important facet of anti-rK39 antibody is that the titer correlates directly with the disease activity, indicating its potential for use in predicting response to chemotherapy. It was previously shown that anti-rK39 antibody titers were 59-fold higher than those of antibody against CSA at the time of diagnosis, and with successful therapy, it fell sharply at the end of treatment and fell further during follow-up monitoring. In patients who experience disease relapse, the titer rose steeply again . The diagnostic and prognostic utility of rK39 for HIV-infected patients has also been demonstrated . Because of the conditions prevailing in areas of endemicity, any sophisticated method cannot be employed on a wider scale. There is a need for a simple rapid and accurate test with good sensitivity and specificity, which can be used without any specific expertise. A promising ready-to-use immunochromatographic strip test based on rK39 antigen has been developed as a rapid test for use in difficult field conditions. The recombinant antigen is immobilized on a small rectangular piece of nitrocellulose membrane in a band form, and goat anti-protein A is attached to the membrane above the antigen band. After the finger is pricked, half a drop of blood is smeared at the tip of the strip, and the lower end of the strip is allowed to soak in 4 to 5 drops of phosphate-buffered saline, placed on a clean glass slide or tube. If the antibody is present, it will react with the conjugate (protein A colloidal gold) that is predried on the assay strip. The mixture moves along the strip by capillary action and reacts with rK39 antigen on the strip, yielding a pink band. In the strip of patients who are infected, two pinkish lines appear in the middle of the nitrocellulose membrane (the upper pinkish band serves as a procedural control). In the first extensive field trial in 323 patients, we found the strip test to be 100% sensitive (confidence interval, 98 to 100%) and 98% specific (confidence interval, 95 to 100%) . Several studies from the Indian subcontinent reported the test to be 100% sensitive . However, when evaluated in Sudan, the sensitivity of the test was only 67%. In the Sudan study, all the parasitologically confirmed VL patients who tested negative by the rK39 strip test showed IgG against rK39 by micro-ELISA (though at lower titers) . In a study done in southern Europe, the rK39 strip test results were positive in only 71.4% of the cases of VL . These differences in sensitivity may be due to differences in the antibody responses observed in different ethnic groups . When tested for PKDL, the test had a 91% sensitivity . High levels of specificity (97 to 100%) have been reported uniformly for this test; however, with a later version of the rK39-treated strips, some (12.5%) healthy endemic control subjects also tested positive . While such reactions might be considered to be false positive, these probably represent subclinical infections: PCR assay for L. donovani was positive in a few of these cases . Anti-rK39 IgG may be present in serum for an extended period after successful treatment for VL; thus, patients with suspected relapse of VL with a past history of infection would not be candidates for diagnosis by strip testing. Another drawback of this format is that an individual with a positive rK39 strip test result may suffer from an illness(es) (malaria, typhoid fever, or tuberculosis) with clinical features similar to those of VL yet be misdiagnosed as suffering from VL. Notwithstanding these limitations, the rK39 immunochromatographic strip test has proved to be versatile in predicting acute infection, and it is the only available format for diagnosis of VL with acceptable sensitivity and specificity levels which is also inexpensive (similar1 to 1.5 U.S. dollars) and simple and can be performed even by paramedics in prevailing difficult field conditions. Specific antibodies can also be detected by Western blotting. For this type of testing, promastigotes of L. donovani are grown to log phase and lysed and the soluble protein is run on sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels. The separated proteins are electroblotted onto a nitrocellulose membrane and probed with serum from the patient. The sensitivity of this technique can be enhanced using the chemiluminescent antibody probes. Using Western blotting, one can find even minor antigenic differences among various organisms and thus detect cross-reactive antigens. However, the process is time consuming, technically cumbersome, and expensive . (iii) Skin testing. | Delayed type hypersensitivity (DTH) or T-cell-mediated immunity is a group-specific immune response. The Montenegro skin test (leishmanin skin test) is a test for DTH specific to leishmaniasis, but its role is limited . In this method, 0.5 ml of phenol-killed whole parasites (5 x 107 promastigotes) is injected on the volar aspect of the forearm of the patient. After 48 to 72 h, the size of induration is measured and compared with the size of induration produced by injection of a phenol-saline control in the other forearm. Presently, there is no available standardized leishmanin reagent. All leishmanins are said to be alike and nonspecific. The test is negative in acute cases of VL due to the absence of DTH and is positive only in cases where kala-azar has been cured . FIG. 1. | Microphotograph showing intracellular and extracellular L. donovani Microphotograph showing intracellular and extracellular L. donovani bodies in splenic aspirate from a patient with visceral leishmaniasis. HIV-LEISHMANIA COINFECTION : Atypical clinical presentations of VL in HIV-infected patients pose a considerable diagnostic challenge. In fact, the clinical triad of fever, splenomegaly, and hepatomegaly is found in less than half of such patients, though more so in patients with low CD4 counts (<50 CD4 cells/mm3) . In these patients, leishmaniasis can present with gastrointestinal involvement (stomach, duodenum, or colon); ascites; pleural or pericardial effusion; involvement of lungs, tonsils, and skin; and even as widely disseminated disease . The diagnostic principles remain essentially the same as those for non-HIV-infected patients. The presence of amastigotes may be demonstrated in buffy coat preparation. Sometimes the presence of amastigotes in unusual sites may be demonstrated (e.g., amastigotes may be present in specimens from bronchoalveolar lavage, pleural fluid, or biopsy specimens from the gastrointestinal tract). For HIV patients, the sensitivity of antibody-based immunologic tests like the IFA test and ELISA is low . Since the parasite load is quite heavy in these patients, the presence of leishmania amastigotes in the bone marrow can often be demonstrated, but there are well-described instances in the literature where amastigotes were not demonstrable on bone marrow, though they were found at unexpected locations like the stomach, the colon, or the lungs. PCR analysis of the whole blood or its buffy coat preparation may prove a useful screening test for these patients, obviating the need for traumatic procedures. CONCLUSIONS : Various noninvasive tests, with various specificities and sensitivities, are available for the diagnosis of leishmaniasis ; however, none have become popular in areas of endemicity. Very few are commercially available; generally speaking, they also are expensive, require skilled personnel, expensive equipment, and electricity, and are technically demanding. Parasite diagnosis by splenic, marrow, or skin lesion remains the "gold standard," with its usual limitations. DAT can be performed only in a few centralized laboratories that are equipped for the purpose (and have trained personnel); cost, multiple steps, incubation, and antigenic variations are limiting factors. The rK39 strip test has the potential to be used for diagnosis of VL under field conditions. Other tests, which are likely candidates for diagnosis and prognosis of leishmaniasis in the future, are KATEX and a field-adaptable version of PCR, which would be simple, inexpensive, and easily available. TABLE 1 | Sensitivities and specificities of various methods used for diagnosis of visceral leishmaniasis Backmatter: PMID- 12204970 TI - Performance of Two Commercial Glycoprotein G-Based Enzyme Immunoassays for Detecting Antibodies to Herpes Simplex Viruses 1 and 2 in Children and Young Adolescents AB - In 61 patients 1 to 14 years of age, the Gull/Meridian enzyme-linked immunosorbent assay (ELISA) had a sensitivity of 100% for herpes simplex virus type 1 (HSV-1) and specificities of 74% for HSV-1 and 48% for HSV-2. In 128 similarly aged patients, the HerpeSelect ELISA (Focus Technologies) showed sensitivities of 80% for HSV-1 and 88% for HSV-2, and specificities of 97% for HSV-1 and 100% for HSV-2. Keywords: Introduction : Herpes simplex virus type 1 (HSV-1) infection is common in childhood and may be indistinguishable from viral or bacterial stomatitis, otitis media, and upper respiratory tract infections (, , 14). HSV-2 infections are unusual after the neonatal period and before sexual debut; seroprevalence begins to rise in late adolescence . Virologic methods for diagnosing HSV infection in children are limited by the need to collect samples early in the clinical course and by the need to perform the vigorous swabbing that is necessary to obtain infected cells from mucosal surfaces or from lesions. Serologic tests to detect HSV antibodies are available commercially. Some tests can distinguish HSV-1 from HSV-2 antibodies on the basis of type-specific antigens of glycoprotein G-1 (gG-1) and gG-2, respectively . The first such test to be approved by the Food and Drug Administration was an enzyme-linked immunosorbent assay (ELISA) from Gull Laboratories, Salt Lake City, Utah that was sold under the Premier brand by Meridian Diagnostics (Cincinnati, Ohio) (termed the Gull/Meridian ELISA). This ELISA was both sensitive and specific in a premarket evaluation of adult sera . To assess the accuracy of the Gull/Meridian ELISAs for children and adolescents, we tested blood samples from healthy children from southern Texas (n = 61; mean age, 7.4 years; range, 1 to 13 years) with kits purchased from Gull Laboratories and compared these results to those obtained by Western blotting (WB), a well-validated "gold standard" . Later, when the Gull/Meridian ELISAs were withdrawn from the market, we extended this study to evaluate the HerpeSelect HSV-1 and HSV-2 ELISAs from Focus Technologies (formerly MRL) on pediatric sera (n = 128; mean age, 5.7 years; range, 1 to 13 years) that had been sent to the University of Washington Virology Laboratory for HSV type-specific serology. Our comparison studies revealed substantial differences among the performances of these ELISAs. The seroprevalence of HSV-1 determined by WB in Texas-based patients was 49% (30 of 61 positive); no patient was positive for HSV-2. Of the sera from Seattle, 46 of 128 (36%) were seropositive for HSV-1 by WB and 8 of 125 (6%; ages 11 days to 14 years) were positive for HSV-2. All eight samples with discordant results for HSV-1 by the Gull/Meridian ELISA and WB were false positive by the ELISA. The sensitivity for the HSV-1 Gull/Meridian ELISA was 100%, with a negative predictive value (NPV) of 100% . The specificity for HSV-1 was 74%, with a positive predictive value (PPV) of 79%. The sensitivity of the Gull/Meridian HSV-2 ELISA could not be evaluated (none were HSV-2 WB positive). Thirty-two of 61 samples (52%) were positive by the Gull/Meridian HSV-2 ELISA, giving a specificity of 48% and a PPV of 0%. TABLE 1 | Results of WB and two ELISA for HSV-1 and HSV-2 antibodies in patients aged 1 to 14 years from Texas and Washington The results for 5 of 128 pediatric sera tested by the HerpeSelect HSV-1 ELISA were equivocal, and the sera could not be classified as negative or positive for comparison. For the remaining 123 sera, the HerpeSelect HSV-1 ELISA had a sensitivity of 80% and a specificity of 97%, with a PPV and an NPV of 95 and 89%, respectively . Two sera had equivocal results with the HerpeSelect HSV-2 ELISA. Neither was positive by WB for HSV-2; one was positive for HSV-1 antibody by both WB and the HerpeSelect HSV-1 ELISA. In addition, three sera had atypical HSV-2 results by WB and could not be scored by that test as either negative or positive. Of the 123 evaluable result sets for HSV-2 antibody, HerpeSelect HSV-2 ELISA had a sensitivity of 88%, a specificity of 100%, a PPV of 100%, and an NPV of 99% . Thus, two commercial ELISAs had very different performance characteristics with pediatric sera. The Gull/Meridian ELISA (based on immunoaffinity-purified gG-1 and gG-2) suffered from very low specificity and unacceptably low PPVs, especially for HSV-2. At least one widely used reference test, the gG-1 and gG-2 immunodot enzyme assay, is based on antigens similar to those in the Gull/Meridian ELISAs . The HerpeSelect ELISAs had high specificity for both HSV-1 and HSV-2 but surprisingly low sensitivity for HSV-1 (80%). Our study raises important issues about the use of these gG-based HSV type-specific serologic tests in children. The very low specificities of the Gull/Meridian ELISAs were surprising in light of previous data for adults . Although these tests could be more sensitive than WB, this higher sensitivity (if true) appears to apply only to children . Alternatively, some young people may have circulating factors that nonspecifically bind glycoprotein G in this particular ELISA. A previous report may have given an early, unrecognized warning of a unique problem with Gull/Meridian ELISAs for pediatric sera . It is unclear whether the specificity problem with Gull/Meridian ELISAs for pediatric sera is restricted to the Gull antigen. Studies using immunoaffinity-purified HSV-2 antigens in gG-2-based tests other than the Gull/Meridian ELISA suggest a reasonably low HSV-2 prevalence rate in children ; these studies give no insight into ELISA performance for antibodies to HSV-1 in pediatric sera. The HerpeSelect HSV-1 and HSV-2 ELISAs are based on baculovirus recombinant gG-1 and gG-2 . Unlike the Gull/Meridian ELISAs, these tests with pediatric sera gave no indication of excessive false-positive HSV-1 results (two of 77 sera) or HSV-2 results (0 of 116 sera) . However, the sensitivity for HSV-1 was only 80%. Two of nine false-negative sera were from infants under 1 year of age and could have represented low-titer maternal antibodies. The other seven sera were from children with a median age of 4.5 years (range, 3 to 12) who might have been in the process of seroconverting, based on the appearance of the WB profiles (data not shown). The accuracy of HSV serologic testing for children appears questionable with current ELISAs. False-positive HSV-1 results may lead to inappropriate treatment or to unnecessary antiviral prophylaxis in immunosuppressed patients. Positive HSV-2 tests for children suggest the occurrence of sexual abuse. Our limited testing of the HerpeSelect ELISAs provides cautious optimism that these tests are reasonably accurate for children. However, prospective studies using virologic diagnosis of infection as the gold standard are needed. A serologic diagnosis of HSV infection in children should be made with caution. A negative test should be followed by testing in 6 to 8 weeks to detect seroconversion. Positive results for HSV-2 antibodies should be confirmed by a second type-specific test, such as WB or the Focus HerpeSelect immunoblot . Backmatter: PMID- 12204953 TI - Calcium Phosphate Nanoparticles Induce Mucosal Immunity and Protection against Herpes Simplex Virus Type 2 AB - Previously we reported that calcium phosphate nanoparticles (CAP) represented a superior alternative to alum adjuvants in mice immunized with viral protein. Additionally, we showed that CAP was safe and elicited no detectable immunoglobulin E (IgE) response. In this study, we demonstrated that following mucosal delivery of herpes simplex virus type 2 (HSV-2) antigen with CAP, CAP adjuvant enhanced protective systemic and mucosal immunity versus live virus. Mice were immunized intravaginally and intranasally with HSV-2 protein plus CAP adjuvant (HSV-2+CAP), CAP alone, phosphate-buffered saline, or HSV-2 alone. HSV-2+CAP induced HSV-specific mucosal IgA and IgG and concurrently enhanced systemic IgG responses. Our results demonstrate the potency of CAP as a mucosal adjuvant. Furthermore, we show that systemic immunity could be induced via the mucosal route following inoculation with CAP-based vaccine. Moreover, neutralizing antibodies were found in the sera of mice immunized intranasally or intravaginally with HSV-2+CAP. Also, the results of our in vivo experiments indicated that mice vaccinated with HSV-2+CAP were protected against live HSV-2 infection. In conclusion, these preclinical data support the hypothesis that CAP may be an effective mucosal adjuvant that protects against viral infection. Keywords: Introduction : Since mucosal surfaces act as the primary point of entry for most pathogens and the first line of defense against them, vaccines inducing effective mucosal immunity may reduce rates of infection and decrease the morbidity and mortality of infectious diseases. Currently, no safe and effective mucosal vaccine adjuvants are approved for human use. Mucosal vaccine delivery is a promising strategy. Mucosal vaccines administered in one part of the body can elicit an antibody response in mucosal tissues remote from the site of initial antigen exposure. This effect occurs because of the common mucosal immune system . A major obstacle to developing a mucosal vaccine in humans is finding a safe and effective adjuvant. Experimental mucosal adjuvants include cholera toxin, heat-labile enterotoxin, mutant toxins (LTK63 and LTR72), CpG oligodeoxynucleotide, polymerized liposomes, microparticles, and interleukins or immune modulators. None of these adjuvants is approved for use in humans . Biodegradable calcium phosphate particles have been investigated as an alternative to aluminum adjuvants for parenteral vaccines. Clinical studies conducted in France described the use of a calcium phosphate adjuvant for secondary or booster immunizations against diphtheria and tetanus . Calcium phosphate has also been used for allergen desensitization . Early studies indicated that calcium phosphate particles produce strong adjuvant effects, induced less immunoglobulin E (IgE) than aluminum adjuvants, and elicited only minimal local irritation in animal experiments and human clinical trials . Here, we describe a unique formulation of calcium phosphate nanoparticles (CAP) which is distinct from the formulations of calcium phosphate described by European scientists and demonstrate its use as an effective mucosal adjuvant. Our results indicate that following viral challenge, mice immunized with CAP-based formulations of herpes simplex virus type 2 (HSV-2) glycoprotein exhibited significantly increased survival rates and less severe clinical infection than controls. These findings demonstrate that CAP delivered as a mucosal adjuvant confers protective antiviral immunity. MATERIALS AND METHODS : Formulation of subunit vaccine. | The preparation of partially purified HSV-2 glycoproteins has been described previously . Briefly, infected cells were collected and sonicated. The viral suspension was centrifuged at 5,500 x g for 15 min. Supernatant was collected and treated with 1% IGEPAL (Sigma Chemical Co., St. Louis, Mo.) lysis buffer for 30 min on ice. The solution was centrifuged at 18,500 x g for 2 h. The supernatant was dialyzed against phosphate-buffered saline (PBS) at 4C and stored at -80C. Then 1 mg of HSV-2 protein was added to 7.5 ml of 12.5 mM calcium chloride, followed by the addition of 7.5 ml of 12.5 mM dibasic sodium phosphate and 1.5 ml of 15.6 mM sodium citrate. The solution was stirred until the final average particle size was less than 1.2 mum, as determined with a Coulter N4Plus Submicron particle sizer, and was treated with 129 mM cellobiose overnight. The total protein inside CAP was 123 mug. The particle containing HSV-2 protein was coated again with 3.877 mg of HSV-2 proteins by coincubation for 1 h at 4C. The final concentration of CAP plus HSV solution was 2 mg of HSV/ml and 10 mg of CAP/ml. The control vaccines were PBS, CAP alone, and HSV-2 protein alone. Animals. | Female BALB/c mice, 6 to 8 weeks old and weighing 25 g, were obtained from Charles River Laboratories. The mice were maintained in standard housing with a normal diet of Purina rodent chow 5001. Immunization and sample collection. | Eight groups of five female BALB/c mice were inoculated intravaginally or intranasally with HSV-2+CAP (20 mug of viral protein plus 100 mug of CAP per dose per mouse), HSV-2 alone (20 mug per dose per mouse), or CAP alone (100 mug per dose per mouse) in a total volume of 50 mul (intravaginally) or 10 mul (intranasally). The mice received two inoculations, on days 0 and 7. Samples were collected 7, 14, and 38 days after the last immunization. Blood was obtained from the orbital sinus, and the serum samples were stored at -20C. Mucosal samples were collected 14 days after the last immunization by vaginal lavage with 100 mul of PBS. The sediments were removed by centrifugation, and samples were pooled and stored at -20C. ELISA. | HSV-specific antibodies were detected by an end-point dilution enzyme-linked immunosorbent assay (ELISA) as described previously . Titers for IgG in plasma samples were expressed as group mean +- standard error of the mean of values for individual animals. Titers for IgA and IgG in mucosal samples were expressed as the mean of triplicate assays from pooled mucosal samples. HSV-2 challenge experiment. | Using methods reported previously , mice were injected subcutaneously with DepoProvera (Upjohn, Kalamazoo, Mich.) at a concentration of 2 mg/mouse in 50 mul of distilled water on the 45th day following primary immunization. Five days later, the mice were challenged intravaginally with 106 PFU of HSV-2. Mice were examined daily for genital pathology, and the clinical scoring was performed by an investigator blinded to the animal's immunization status. Clinical pathology was scored on a 5-point scale: 0, no apparent infection; 1, slight redness of external vagina; 2, severe redness and swelling of external vagina; 3, genital ulceration with severe redness, swelling, and hair loss of genital and surrounding tissue; 4, severe ulceration of genital and surrounding tissue and paralysis; and 5, death. Neutralization assay. | As reported previously , Vero cells were propagated in culture plates. Pooled mouse serum samples from day 38 after the last immunization were incubated with HSV-2 and assessed for the presence of HSV-2-specific neutralizing antibodies by plaque assay. The titer is the reciprocal of the serum dilution required to inhibit the cytolysis of a confluent monolayer of Vero cells by 50%. Statistical analysis. | Pathological data were analyzed by analysis of variance to determine the difference between groups. RESULTS : As indicated in Fig. , both the intranasal and intravaginal HSV-2+CAP-vaccinated mice showed a high titer of HSV-specific mucosal IgA and IgG in vaginal lavage fluid at 14 days after the last immunization. Serological IgG and IgG2a titers determined on day 38 after the last immunization showed a systemic response in the mice after intranasal or intravaginal immunization with HSV-2+CAP compared to PBS, CAP alone, or HSV-2 alone . The neutralization assay was performed at day 38 following secondary immunization. Neutralizing antibodies were found in both the intranasally and intravaginally HSV-2+CAP-immunized mice at titers of 1:40 and 1:80, respectively. Notably, neutralizing antibodies were absent in the mice inoculated with PBS alone, CAP alone, or HSV-2 alone. Resistance to HSV-2 infection was evaluated by monitoring clinical pathology. On days 6, 8, and 10, the reduced clinical severity in mice intravaginally immunized with HSV-2+CAP achieved statistical significance (P < 0.05) compared to mice immunized with PBS, CAP alone, or HSV-2 alone (Fig. , right panel). One of five mice intravaginally inoculated with HSV-2+CAP died from HSV-2 infection, whereas all of the mice intravaginally vaccinated with PBS, HSV-2 alone, and CAP alone developed severe disease and died by day 8 or 10. Similarly, the mice vaccinated intranasally with HSV-2+CAP showed reduced clinical severity compared with mice immunized with PBS, CAP alone, or HSV-2 alone at days 8 and 10 (Fig. , left panel). Two of five mice intranasally vaccinated with CAP+HSV-2 died, compared with the controls (i.e., recipients of PBS, CAP alone, and HSV-2 only), all of which died eventually. All surviving mice were kept for 2 more weeks and recovered gradually. FIG. 1. | Groups of five female BALB/c mice were immunized on days 0 and 7 by intranasal or intravaginal delivery of PBS (vertically striped bars), CAP alone (open bars), HSV-2 alone (horizontally striped bars), or HSV-2+CAP (solid bars). Groups of five female BALB/c mice were immunized on days 0 and 7 by intranasal or intravaginal delivery of PBS (vertically striped bars), CAP alone (open bars), HSV-2 alone (horizontally striped bars), or HSV-2+CAP (solid bars). The antigen concentration and vaginal lavage fluid dilution used in the ELISA were 100 mug/ml and 1:1, respectively. FIG. 2. | The antigen concentration and antibody dilution used in the IgG ELISA were 6 mug/ml and 1:200, respectively. The antigen concentration and antibody dilution used in the IgG ELISA were 6 mug/ml and 1:200, respectively. The antigen concentration and antibody dilution used in the IgG2a ELISA were 100 mug/ml and 1:50, respectively. Each bar represents the group mean antibody level for mice immunized intranasally or intravaginally with PBS (vertically striped bars), CAP alone (open bars), HSV-2 alone (horizontally striped bars), or HSV-2+CAP (solid bars). FIG. 3. | Five BALB/c mice per group were immunized intranasally or intravaginally with PBS (vertically striped bars), CAP alone (open bars), HSV-2 alone (horizontally striped bars), or HSV-2+CAP (solid bars) and challenged intravaginally with 106 PFU of HSV-2 at 43 days after the last immunization. Five BALB/c mice per group were immunized intranasally or intravaginally with PBS (vertically striped bars), CAP alone (open bars), HSV-2 alone (horizontally striped bars), or HSV-2+CAP (solid bars) and challenged intravaginally with 106 PFU of HSV-2 at 43 days after the last immunization. Clinical pathology was scored as described in the text. The stars indicate P values of <0.05 for HSV-2+CAP versus PBS, CAP alone, and HSV-2 alone. DISCUSSION : The mucosal tissues are the primary routes of entry into the body for microbial pathogens. Vaccines inducing mucosal immunity prevent the transmission of infection via mucosal surfaces. However, no mucosal vaccine adjuvant is currently approved for human use. Because of the weak inherent immunogenicity of some antigens targeted for vaccine development, such as epitope subunits and recombinant peptides, there is a great need for safe and efficient mucosal adjuvants. The only adjuvants used in licensed vaccines in the United States are aluminum compounds, which effectively enhance immune responses . However, human studies have shown them to be weak adjuvants for inducing cell and humoral immunity to some virus protein subunits (S. J. D. Bell, personal observation). Additionally, alum can elicit an IgE antibody response that increases the risk of allergic reactions. We have reported previously that CAP delivered intraperitoneally with HSV-2 and Epstein-Barr virus proteins induced high titers of IgG2a antibody and neutralizing antibody and facilitated a high degree of protection against viral infection in a murine model . In this study, using HSV-2 protein as a model antigen, we evaluated the immunity and efficacy of an HSV-2+CAP experimental vaccine. Our results indicated that mice vaccinated either intravaginally or intranasally with HSV-2+CAP had high antibody levels at mucosal surfaces and effective neutralizing antibody titers and were protected against virus infection. We assumed that the neutralizing antibody prevented the attachment of pathogens to the epithelial surfaces and conferred protection against subsequent viral infection. Our findings also confirmed the previous studies showing that antibodies can efficiently neutralize virus in mucosal areas. The immune system within the female lower genital tract is the initial defense against sexually transmitted diseases. Our study suggested that intravaginal immunization induced relatively higher levels of mucosal IgG and IgA than intranasal immunization, providing optimal protection against HSV-2 infection. This observation supports the findings of others and suggests that genital local immunity and Th1 response in association with other protecting factors, such as local production of antibodies and viral clearance from the vaginal mucosa, play a major role in HSV-2 infection in mice. Our next step is to prove that CD4+ T cells secreting gamma interferon and B cells or natural antibodies are critical for immune protection against lethal genital HSV-2 reinfection. The exact mechanism of the adjuvant action of CAP is not fully understood. M cells in the mucosal tissues are known to reside exclusively in the epithelium and deliver foreign material by transepithelial transport from the lumen to the underlying mucosa-associated lymphoid tissue. Particulate antigens are desirable because they permit M cells to translocate across the tight epithelial barrier to mucosal dendritic cells. Therefore, the particulate mucosal vaccine created from the combination of soluble antigens formulated within CAP provides the desirable size and functional attributes to induce effective mucosal immunity. Recent comparative studies have indicated that microparticles are potent adjuvants for mucosal delivery . However, microparticles are not an ideal size for inducing cellular immunity because they tend to be too large, and it is believed that M cells, dendritic cells, macrophages, and local lymph nodes are more effective at taking up smaller particles. Advantageously, CAP are generally in the preferred size range (i.e., less than 1.2 mum, versus 1 mum-sized polymers) and also stimulate cellular immunity and cytotoxic T lymphocyte responses (unpublished data). Based on these results, we conclude that (i) the CAP-based HSV-2 subunit vaccine appears to concurrently induce both systemic and mucosal immunity and (ii) CAP shows great potential as a safe and effective mucosal vaccine adjuvant for humans, given its relative absence of side effects and lack of IgE antibody induction. Backmatter: PMID- 12204973 TI - Determination of the Nucleotide Sequences of Heat Shock Operon groESL and the Citrate Synthase Gene (gltA) of Anaplasma (Ehrlichia) platys for Phylogenetic and Diagnostic Studies AB - The 1,670-bp nucleotide sequence of the heat shock operon groESL and the 1,236-bp sequence of the citrate synthase gene (gltA) of Anaplasma (Ehrlichia) platys were determined. The topology of the groEL- and gltA-based phylogenetic tree was similar to that derived from 16S rRNA gene analyses with distances. Both groESL- and gltA-based PCRs specific to A. platys were also developed based upon the alignment data. Keywords: Introduction : Anaplasma (Ehrlichia) platys is a bacterial parasite of dog platelets that causes infectious cyclic thrombocytopenia . A. platys has been shown to be closely related to Anaplasma marginale, Anaplasma centrale, and Anaplasma phagocytophila, including the former human granulocytic ehrlichia (HGE) agents Ehrlichia equi and Ehrlichia phagocytophila, based on 16S rRNA gene sequences ; however, little information is available regarding the natural history of the pathogen. The 16S rRNA gene had been the only known gene sequence of A. platys before the heat shock protein gene (groEL) was sequenced recently . A groEL sequence analysis supported the phylogenetic relationship between A. platys and related species. The groESL operon contains a spacer region between groES and groEL which is thought to be more divergent than the coding regions . However, the nucleotide sequence of the spacer region of A. platys has not been studied yet. Thus, the nucleotide sequences of groES and the spacer region between groES and groEL were analyzed for additional phylogenetic characterization of A. platys. More recently, we sequenced the citrate synthase gene (gltA) of 13 species, including Ehrlichia, Anaplasma, and Neorickettsia, for phylogenetic analyses and found higher variation than for the 16S rRNA gene . The topology of the gltA-based phylogenetic tree confirmed the reorganization of genera in the families Rickettsiaceae and Anaplsmataceae reported recently . However, the gltA sequence of A. platys has yet to be determined. Thus, the nucleotide sequences of A. platys gltA were also analyzed to support the phylogenetic relationship of A. platys among related species. We also propose to use sequence data from groESL and gltA with greater differences among species to develop an A. platys-specific PCR method. New PCR primers to specifically detect A. platys fragments were developed based on the alignment data of these two genes for a diagnostic assay. The A. platys DNA analyzed in this study was supplied from a dog infected in Somieres, France . The dog had a history of a clotting disorder. At the time of bleeding, the platelet count was 256,000/mul, which was in the normal range, and A. platys was observed within 58% of platelets on a Giemsa-stained peripheral blood smear. To evaluate the species-specific PCR designed in this study, DNA from other strains of A. platys were used. DNA from a dog infected with A. platys in Venezuela was kindly provided by E. B. Breitschwerdt, North Carolina State University . DNA from a dog infected with A. platys in Okinawa, Japan, was also used . For amplification of the groESL operon of A. platys, EEgro1F and EEgro2R were used to amplify an approximately 1,700-bp fragment with an annealing temperature of 55C . The amplification products were purified using the QIAquick PCR purification kit (Qiagen GmbH, Hilden, Germany) and sequenced. Five primers shown in Table were used to complete the sequence of groESL. Fluorescence-labeled dideoxynucleotide technology was used for DNA sequencing reactions (Perkin-Elmer, Applied Biosystems Division, Foster City, Calif.). Samples were then sequenced using a Perkin-Elmer ABI Prism 377 automated DNA sequencer at the DNA Core Facility of the Center for Gene Research, Yamaguchi University. TABLE 1 | Oligonucleotide sequences of primers used in this study The strategy for determining the gltA sequence was similar to that used in our previous report . A partial sequence of A. platys gltA was first determined using two sets of degenerate primers, F1b and EHR-778R, and F1b and HG1085R . These primers were designed based upon the sequence of gltA of A. phagocytophila, A. marginale, and A. centrale. The amplification conditions were the same as in the previous study , with an annealing temperature of 53C. The amplification products were purified and sequenced as described above. After a partial determination of the sequence, the unknown areas of the 3' and 5' ends of the gene were determined using the Universal Genome Walker kit (Clontech Laboratories, Palo Alto, Calif.). Briefly, genomic DNA was digested with EcoRV, DraI, PvuII, StuI, and ScaI. DNA fragments were ligated with a Genome Walker adaptor, which had one blunt end and one end with a 5' overhang. A ligation mixture of the adaptor and ehrlichial genomic DNA fragments was used as a template for PCR. This PCR was performed using an adaptor primer supplied by the manufacturer and A. platys gltA-specific primers to walk downstream on the DNA sequence . For the amplification, the conditions were as in our previous report . The sequences of A. platys and the registered sequences of other related species deposited in GenBank were analyzed for phylogenetic relationships. Multiple alignment analysis, the calculation of distance matrices, and the construction of phylogenetic trees were performed with the ClustalW program version 1.8 in the DNA Data Bank of Japan (Mishima, Japan; ). The distance matrices for the aligned sequences with all gaps ignored were calculated using the Kimura two-parameter method , and the neighbor-joining method was used for constructing a phylogenetic tree . The stability of the tree obtained was estimated by bootstrap analysis for 100 replications using the same program. Tree figures were generated using the TreeView program, version 1.61 . A primer set, forward primer PLA-HS475F and reverse primer PLA-HS1198R, was designed based upon the alignment data to specifically amplify an A. platys groESL fragment. Another set of primers, PLA-CSM136F and PLA-CS1359R, was also designed based upon the alignment data of gltA. PCR conditions were the same as described above but with an annealing temperature of 58C and the use of 40 cycles. The specificity of the reaction was tested with DNA extracted from the three strains of A. platys and related species, including A. phagocytophila (formerly the HGE strain Webster) (J. S. Dumler), E. equi strain California (J. E. Madigan), E. phagocytophila strain 1602 (A. Garcia-Perez), A. marginale strain Florida (G. H. Palmer), E. canis strain Oklahoma (J. Dawson), Wolbachia pipientis (M. Taylor), and Neorickettsia helminthoeca (Y. Rikihisa). The sensitivities of both PCR systems were also examined using DNA from A. platys strain France. The DNA was diluted 10-fold from 1:1 to 1:10,000 with distilled water. A 1,670-bp groESL fragment of A. platys was determined to contain 41 bp of the partial groES, 51 bp of the spacer region, and 1,577 bp of the groEL coding region. The groESL operon was compared with that of other Anaplasma and Ehrlichia bacteria reported previously and was found to be closely related to the operon of A. phagocytophila and A. marginale, with 81.4 and 78.8% identity, respectively. The level of similarity among groESL sequences was much lower than that for the 16S rRNA gene sequence in the same species (98.6% with A. phagocytophila and 96.1% with A. marginale). The spacer length of 51 nucleotides was similar to that for related Anaplasma species: 52 bp for A. phagocytophila and 47 bp for A. marginale. The percent identities of the nucleotide sequence in the spacer region of A. platys compared to that of A. phagocytophila and to that of A. marginale were 74.5 and 72.3%, respectively, revealing a greater degree of divergence than for the entire groEL coding region. After the initial identification of the 955-bp partial sequence of A. platys gltA using the degenerate PCR strategy, the full-length open reading frame extending from the ATG start codon to the TAA stop codon was determined using the Genome Walker PCR method. The length of the gltA open reading frame was 1,236 bp and encoded a protein of 411 amino acids. The complete gltA sequence was compared with that of other Anaplasma and Ehrlichia bacteria reported previously and found to be closely related to those of A. phagocytophila and A. marginale, with 62.7 and 63.2% identity, respectively. The level of similarity among ehrlichial gltA was much lower than that for the 16S rRNA gene sequence in the same species. Thus, the sequence of gltA is much more variable among these species than is groESL. The length of the gltA sequence of A. platys (1,236 bp) is the same as that of A. phagocytophila but slightly shorter than that of A. marginale and A. centrale (1,254 bp). In topology, the gltA-based phylogenetic tree was very similar to the tree derived from analysis of the 16S rRNA gene analyses and the groEL-based tree reported previously . However, the trees constructed from gltA and groESL nucleotide sequences showed more distance than the 16S rRNA-based trees. These findings also support the use of gltA-based and groESL-based comparisons in determining the phylogeny of Anaplasma, Ehrlichia, and Neorickettsia agents and strengthen the 16S rRNA- and groESL-based phylogeny reported recently . There have been reports of A. platys-like organisms identified based on 16S rRNA analysis, including bacteria from white-tailed deer in North America and from a cow in South Africa . It would be interest to determine the phylogenetic position of these agents by analyzing the gltA and groESL gene sequences. FIG. 1. | Phylogenetic relationship of various Anaplasma, Ehrlichia, and Neorickettsia spp. Phylogenetic relationship of various Anaplasma, Ehrlichia, and Neorickettsia spp. based on the nucleotide sequences of gltA (a) and groESL (b) genes. The neighbor-joining method was used to construct the phylogenetic tree with the ClustalW program. The scale bar represents 10% divergence. The numbers at nodes are the proportions of 100 bootstrap resamplings that support the topology shown. The GenBank accession numbers of the groESL sequences used to construct the phylogenetic tree and aligned data are as follows: A. phagocytophila (strain HGE agent), ; E. equi, ; E. phagocytophila, ; A. marginale, ; E. chaffeensis, ; E. canis, ; E. muris, ; Ehrlichia sp. detected from I. ovatus, ; Ehrlichia (Cowdria) ruminantium, ; Neorickettsia (Ehrlichia) risticii, ; Neorickettsia (Ehrlichia) sennetsu, ; Rickettsia prowazekii, ; Bartonella henselae, . The GenBank accession numbers of the gltA sequences used for comparative analysis are as follows: A. phagocytophila (strain HGE agent), ; E. equi, ; E. phagocytophila, ; A. marginale, ; A. centrale, ; E. chaffeensis, ; E. canis, ; E. muris, ; Ehrlichia sp. detected from I. ovatus, ; E. (C.) ruminantium, ; N. (E.) risticii, ; N. (E.) sennetsu, : N. helminthoeca, ; R. prowazekii, ; B. henselae, . PCR is a powerful tool for epidemiological or diagnostic purposes because of its high sensitivity and specificity; however, there are few molecular tools available for A. platys. All PCR assays reported previously were developed based upon 16S rRNA gene sequences . In the present study, new PCR primers were designed to amplify A. platys DNA specifically and were based upon the alignment data of groESL and gltA of A. platys and closely related Anaplasma and Ehrlichia species. The PCR produced a fragment of 724 bp from groESL and 1,459 bp from gltA with the DNA from A. platys strain France . The sensitivities of the PCR systems were examined by using diluted DNA. Both groESL- and gltA-based PCRs detected DNA diluted 1:100 , similar to the 16S rRNA-based method . As 5 mul of the original DNA solution contained genomic DNA from approximately 375 platelets infected with A. platys, both PCR systems can detect DNA from 3.75 infected platelets in a reaction mixture. The specificity of both systems was also examined using DNA of three A. platys strains from different geographic locations, France, Japan, and Venezuela, and using DNA from related species, including A. phagocytophila and A. marginale. Figure shows that both PCR systems were specific for A. platys. Furthermore, the gltA-based PCR amplifies the whole gltA sequence of A. platys, which contains both start and stop codons. As the sequences of gltA and groESL have greater variation than the sequence of 16S rRNA, the sequence analysis of PCR products may supply useful information for phylogenic studies of the agents. Our findings suggest that both groESL- and gltA-based PCRs are useful for the specific detection of A. platys DNA, and they would be additional molecular tools for both phylogenetic study and diagnosis in veterinary medicine. FIG. 2. | The sensitivities of the A. platys The sensitivities of the A. platys-specific PCR based upon the groESL (lanes 1 to 6) and gltA (lanes 7 to 12) genes were evaluated. DNA equivalent to that from 375, 37.5, 3.75, 0.375, and 0.0375 infected platelets was used as a template for the amplicons demonstrated in lanes 1 and 7, 2 and 8, 3 and 9, 4 and 10, and 5 and 11, respectively. Distilled water was used in lanes 6 and 12 as a negative control. Lane M, 100-bp DNA ladder. FIG. 3. | A. platys A. platys-specific PCRs based upon the groESL (A) and gltA (B) genes were evaluated for their specificity with DNA from A. platys strains from Somieres, France (lane 1), Okinawa, Japan (lane 2), and Venezuela (lane 3), and from A. marginale (lane 4), A. centrale (lane 5), A. phagocytophila strain HGE agent (lane 6), E. equi (lane 7), E. phagocytophila (lane 8), W. pipientis (lane 9), E. canis (lane 10), and N. helminthoeca (lane 11), with distilled water as a negative control (lane N). Positive bands (arrows) were observed only with DNA of the three strains of A. platys in both groESL- and gltA-based PCRs. Nucleotide sequence accession numbers. : The nucleotide sequences of groESL and gltA of A. platys strain France determined herein have been deposited in the GenBank database under the accession numbers and , respectively. Backmatter: PMID- 12204966 TI - Suppression of Growth by All-trans Retinoic Acid Requires Prolonged Induction of Interferon Regulatory Factor 1 in Cervical Squamous Carcinoma (SiHa) Cells AB - All-trans retinoic acid (ATRA) suppresses growth of cervical dysplasias in vivo, although the sensitivity to retinoids is frequently lost during cervical carcinogenesis. It has been suggested that prolonged treatment or use of higher doses of retinoids might offer favorable response rates. We found SiHa cervical squamous carcinoma cells that were virtually resistant to ATRA-induced growth-inhibitory effects at physiological doses (10-7 to 10-6 M) to be more responsive at pharmacological doses (10-5 to 10-4 M). The growth inhibition by high-dose (10-4 M) ATRA was associated with a sustained activation of interferon regulatory factor 1 (IRF-1), while a low dose (10-6 M) of ATRA activated IRF-1 only transiently. Antisense IRF-1 inhibited the high-dose (10-4 M), ATRA-mediated growth arrest; forced expression of IRF-1 caused a significant reduction in cell growth. High-dose (10-4 M) ATRA increased binding of NF-kappaB and STAT1 proteins to sequences that originated from the IRF-1 promoter region, while low-dose (10-6 M) ATRA induced only NF-kappaB binding. A delayed tyrosine phosphorylation of the signal transducer and activator of transcription-1 (STAT1) was observed after high-dose (10-4 M) but not low-dose (10-6 M) ATRA treatment. In agreement with this, induction of IRF-1 mRNA by ATRA was only modest and transient in a STAT1 knockout cell line, suggesting the importance of STAT1 in sustained IRF-1 expression. Our data showed that ATRA is capable of inducing dose-dependent cellular changes, which might be appropriate to overcome resistance to retinoids that frequently develops during cervical carcinogenesis. Keywords: Introduction : Vitamin A and its natural or synthetic derivatives (collectively known as retinoids) are potent regulators of growth of various malignancies, including cervical cancer . In clinical trials, ATRA could reverse or suppress low-grade or moderate- to high-grade cervical dysplasias . However, retinoids were not effective in patients with more advanced dysplasias ; this finding is similar to the resistance to ATRA observed with HPV16-transformed cervical keratinocytes in vitro . It has been suggested that prolonged treatment or use of higher doses of retinoids might offer a favorable response rate . Retinoids are potent modulators of cellular proliferation and differentiation. In cervical carcinoma cells, retinoic acid induces interferon regulatory factor (IRF-1) , which is responsible for growth arrest . In growth-arrested cells, IRF-1 mRNA expression is markedly elevated, but its expression declines prior to and during DNA synthesis: in this context, IRF-1 is a tumor suppressor . IRF-1 expression can be stimulated by ATRA at the level of transcription through a gamma interferon-activated site (GAS) , via an NF-kappaB site , or directly via a retinoid-responsive element found in the promoter of the IRF-1 gene. Accordingly, our aim was to determine the effects of different doses of ATRA on the expression and regulation of IRF-1 and on the subsequent inhibition of growth in cervical squamous carcinoma (SiHa) cells. MATERIALS AND METHODS : Cell lines. | The HPV16-positive cervical squamous carcinoma cell line (SiHa) was purchased from ATCC and maintained at 37C in Dulbecco's modified Eagle medium supplemented with 10% fetal calf serum in a 5% CO2 atmosphere. 2fTGH and U3A fibroblast cell lines were gifts from George Stark (The Cleveland Clinic Foundation Research Center, Cleveland, Ohio), and they were maintained at 37C in Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum plus 250 mug of hygromycin/ml in a 5% CO2 atmosphere. Plasmids. | The expression vector that contains IRF-1 cDNA (pHuIRF3-1) was kindly provided by Tadatsugu Taniguchi (Department of Immunology, Graduate School of Medicine and Faculty of Medicine, University of Tokyo, Tokyo, Japan). The IRF-1 cDNA was PCR amplified and subcloned into the pCR3.1 bidirectional eukaryotic TA cloning vector (Invitrogen, Carlsbad, Calif.) in both sense and antisense orientations, according to the manufacturer's instructions. Antibodies. | Anti-STAT1 and anti-phospho-STAT1 antibodies were purchased from Upstate Biochemical, Inc. (Lake Placid, N.Y.). The anti-IRF-1 antibody was purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, Calif.). MTT assay. | The growth rates of cells were measured with the 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay. Briefly, 2 x 103 cells in 0.2 ml of culture medium were plated in each well of a 96-well culture plate. A total of 10 wells per time point were used for treatment. Cells were analyzed at regular intervals of 2, 4, 6, 8, and 10 days by the addition of 40 mul of MTT, 1.25 mug of MTT/ml of phosphate-buffered saline (PBS), to each well of the plate. The cells were incubated at 37C for 2.5 h, the medium was aspirated, and the cells were then lysed in 100 mul of dimethyl sulfoxide. Conversion of MTT to formazan by metabolically viable cells was monitored at 570 nm in an enzyme-linked immunosorbent assay reader, and the results were analyzed by regression analysis from triplicate experiments. Western blot analysis. | Cells were washed, scraped in PBS, and then centrifuged. The pellet was resuspended in a RIPA buffer containing protease inhibitors and kept on ice for 60 min. After centrifugation the protein content of the supernatant (whole-cell extract) was determined by a Bio-Rad DC (Hercules, Calif.) method. Fifty micrograms of whole-cell extract was analyzed by standard methods on a sodium dodecyl sulfate-polyacrylamide gel as described earlier . Proteins were transferred to a polyvinylidene difluoride membrane (Bio-Rad) and incubated with the primary antibody at 4C overnight. Protein bands were detected by an enhanced chemiluminescent method (ECL; Amersham Pharmacia, Piscataway, N.J.). Bands were analyzed on an AlphaImager system (Alpha Innotech, San Leandro, Calif.). Blots were stripped and reprobed several times. Equal protein loads were checked by rehybridizing the blots with a glyceraldehyde-3-phosphate dehydrogenase (G3PDH) antibody (Advanced Immunochemical, Long Beach, Calif.) that served as a constitutively expressed internal control. RNA isolation and semiquantitative reverse transcription-PCR. | Cells were washed with PBS and directly lysed in TriReagent-LS (Molecular Research Center, Inc., Cincinnati, Ohio) and precipitated according to the manufacturer's recommendations. One microgram of RNA was reverse transcribed (SuperScript II; Gibco/BRL, Grand Island, N.Y.) and subjected to PCR amplification as described earlier . Primers for IRF-1 were custom designed and synthesized (Genosys, The Woodlands, Tex.). Primers for G3PDH were purchased from Clontech (Palo Alto, Calif.). PCR fragments were resolved by agarose gel electrophoresis and transferred to a nylon membrane (Amersham Pharmacia, Piscataway, N.J.) and hybridized with end-labeled oligonucleotide probes. Oligonucleotide probes were custom designed and synthesized. Autoradiograms were analyzed by densitometry (AlphaImager; Alpha Innotech, San Leandro, Calif.). Electrophoretic mobility shift assay (EMSA). | Nuclear extracts were prepared by lysing cells in a buffer containing 10 mM HEPES (pH 7.9), 60 mM KCl, 1 mM dithiothreitol, 1 mM EDTA, 0.5% NP-40, and 1 mM phenylmethylsulfonyl difluoride. Five to ten micrograms of nuclear extracts were incubated in a binding buffer (10 mM Tris-HCl [pH 7.6], 50 mM KCl, 5 mM MgCl2, 0.1 mM EDTA, 6% glycerol, 1 mM dithiothreitol, 1 mug of polydIC) in the presence or absence of a 100-fold molar excess of unlabeled competitor DNA and/or appropriate antibodies on ice for 10 min. Following incubation, 10,000 cpm of 32P-end-labeled oligonucleotide probe was added and the reaction was incubated at room temperature for an additional 30 min. The DNA-protein complexes were separated from free probe by electrophoresis on a 4 to 6% polyacrylamide gel. The gel was dried and subjected to autoradiography. Oligonucleotides were custom designed and synthesized (Genosys, Inc.). Annealing of individual oligonucleotides was done according to standard protocols. EMSA oligonucleotides. | For gel shift competitions and experiments, the following double-stranded oligonucleotides were used. (The sense strand is shown. The oligonucleotides were synthesized by Genosys, Inc., and annealed according to a standard protocol.) IRF1-NF-kappaB , TTAGCGGGATTCCCCAGCCCT; IRF1-GAS , AGCCTGATTTCCCCGAAATGAC. Plasmid transfection. Cells were transfected with the recombinant IRF-1 plasmid using a GenePORTER 2 transfection reagent (GTS Inc., San Diego, Calif.). Cells were plated at 5 x 105 cells/60-mm dish density in 1 ml of culture medium 24 h before the transfection so that they will be 60 to 90% confluent on the day of transfection. Then, 4 mug of plasmid DNA was diluted in 80 mul of serum-free medium plus 20 mul of GenePorter 2, which gives a final volume of 100 mul. This mixture was incubated at room temperature for 10 min and then was added to the cells. After 24 h, fresh medium was added. Transfections were made in triplicate. Controls such as mock-transfected and vector-transfected cells were included. RESULTS : ATRA inhibits growth of SiHa cells in a dose-dependent fashion. | SiHa cells were grown in the presence or absence of ATRA for various time points . Cell growth was determined by an MTT assay. Apparently, physiological (low) doses (10-7 and 10-6 M) of ATRA did not inhibit cell growth significantly, while administration with pharmacological (high) doses (10-5 and 10-4 M) of ATRA caused significant growth inhibition. IRF-1 participates in ATRA-induced growth inhibition. | IRF-1 is an important regulator of cell proliferation , so its role in ATRA-induced growth inhibition was investigated. Accordingly, protein lysates from SiHa cells treated with low-dose (10-6 M) or high-dose (10-4 M) ATRA were evaluated by Western immunoblotting. ATRA at a concentration of 10-6 M increased protein levels of IRF-1 at a peak of fourfold 6 h after treatment , while induction of IRF-1 protein by 10-4 M ATRA was sustained. We also determined IRF-1 protein levels 24 h after treatment with various concentrations of ATRA . Apparently, induction of IRF-1 at 24 h was dose dependent; significant induction occurred at high doses. In another set of experiments, SiHa cells were transiently transfected with an IRF-1 expression vector. Using a trypan blue exclusion test, we determined the cell number 48 h posttransfection. The results (representing the mean of three independent measurements in which the standard deviation did not exceed 4%) showed that IRF-1 transfection significantly reduced cell numbers, to 65% of the level of untreated control cells. Also, transient transfection of an antisense IRF-1 construct abrogated ATRA-induced growth arrest. Prolonged induction of IRF-1 requires STAT1. | IRF-1 mRNA levels were increased only moderately and transiently after treatment with ATRA in a STAT1 knockout cell line (U3A), while treatment of the parental (STAT1 wild-type) line (2fTGH) with ATRA increased it significantly and for a longer time . Only high-dose ATRA induces phosphorylation of STAT1. | Considering the role of STAT1 in regulation of IRF-1 expression, we determined the phosphorylation status of STAT1 after ATRA treatment. Treatment with 10-6 M ATRA did not affect STAT1 tyrosine phosphorylation . In contrast, 10-4 M ATRA treatment significantly elevated STAT1 tyrosine phosphorylation between 3 to 12 h posttreatment. STAT1 levels, however, remained unchanged during ATRA treatment. NF-kappaB binding to an oligonucleotide from the IRF-1 promoter is independent of ATRA concentrations. | IRF-1 transcription is regulated via NF-kappaB sites in the IRF-1 gene promoter . Accordingly, we tested NF-kappaB binding to oligonucleotides corresponding to NF-kappaB-binding sites in the IRF-1 promoter by a gel-shift assay. As shown, NF-kappaB (mostly the p65-p50 complex) binding to the IRF-1 promoter was increased after 30 min by both low- and high-dose ATRA treatment (Fig. and B). STAT1 binding to an oligonucleotide from the IRF-1 promoter depends on ATRA concentration. | IRF-1 transcription could be also regulated via GAS/SIE sites in the IRF-1 gene promoter . Accordingly, we tested STAT1-binding oligonucleotides corresponding to a GAS site in the IRF-1 promoter by a gel-shift assay. In the extracts of ATRA-treated cells, binding of STAT1 to a GAS site was observed only after high-dose, but not low-dose, ATRA treatment . FIG. 1. | Effects of ATRA on cell growth. Effects of ATRA on cell growth. SiHa cells were seeded in a multiwell plate and treated with either low-dose (10-7 and 10-6 M) or high-dose (10-5 and 10-4 M) ATRA for the times indicated. An MTT assay was employed to determine the amounts of viable cells. Results are shown as means +- standard deviations (n = 3). FIG. 2. | Effects of ATRA on IRF-1 protein levels in SiHa cells. Effects of ATRA on IRF-1 protein levels in SiHa cells. Confluent cultures of SiHa cells were treated with either low-dose (10-6 M) or high-dose (10-4 M) ATRA for the times indicated. Top, Western immunoblotting was performed on the total cell lysate as described in Materials and Methods. Bottom, Densitometric results are expressed as fold change to the untreated controls. Values are shown as means +- standard deviations (n = 3). FIG. 3. | Dose-dependent induction of IRF-1 by ATRA. Dose-dependent induction of IRF-1 by ATRA. Confluent cultures of SiHa cells were treated for 24 h with various doses of ATRA as indicated. Western immunoblotting was performed on the total cell lysate as described in Materials and Methods. Densitometric results are expressed as fold change to the untreated controls. Values are shown as means +- standard deviations (n = 3). FIG. 4. | Prolonged induction of IRF-1 requires STAT1. Prolonged induction of IRF-1 requires STAT1. STAT1 knockout (U3A) or STAT1 wild-type (2fTGH) cells were treated with ATRA (10-6 M) for different time points. IRF-1 mRNA levels were determined by reverse transcription-PCR together with the constitutively expressed G3PDH. IRF-1 levels are given as ratios of IRF-1/G3PDH. Values are shown as means +- standard deviations (n = 3). FIG. 5. | Effects of different doses of ATRA on tyrosine phosphorylation of STAT1 in SiHa cells. Effects of different doses of ATRA on tyrosine phosphorylation of STAT1 in SiHa cells. Confluent cultures of SiHa cells were treated with low-dose (10-6 M) or high-dose (10-4 M) ATRA for the times indicated. Western immunoblotting was performed using anti-phosphoSTAT1 (STAT1p) as well as anti-STAT1 antibodies. Experimental data shown are representative of three independent experiments. FIG. 6. | Effects of different doses of ATRA treatment on binding of cellular proteins to oligonucleotides representing NF-kappaB or GAS sites of the IRF-1 promoter. Effects of different doses of ATRA treatment on binding of cellular proteins to oligonucleotides representing NF-kappaB or GAS sites of the IRF-1 promoter. Nuclear extracts from SiHa cells treated with low-dose (A and C) or high-dose (B and D) ATRA were incubated with labeled oligonucleotides representing the NF-kappaB (A and B) or GAS (C and D) binding sites of the IRF-1 promoter, and an EMSA was performed as described in Materials and Methods. In panels A and B we showed only the 30-min time point that represented the first appearance of NF-kappaB binding. STAT1 binding to the GAS oligomer (C and D) appeared after 6 h of treatment. Next, 100x cold ligand was used to demonstrate specificity of both probes (data not shown). Tetradecanoyl phorbol acetate (TPA)- or gamma interferon (IFNgamma)- treated cell extracts were used as positive controls. Preincubations with specific anti-NF-kappaB p65 and p50 or STAT1 antibodies were performed to demonstrate the presence of these proteins in the observed complexes (data not shown). DISCUSSION : Induction of IRF-1 mRNA and the consequent growth inhibition in cervical squamous carcinoma cells by ATRA have been demonstrated . Clinically, retinoids reversed or suppressed low-grade or moderate- to high-grade cervical dysplasias but were not effective in patients with more advanced dysplasias . Although immortalized keratinocytes are more sensitive to growth control by retinoids than their normal counterparts , at later stages transformed keratinocytes frequently become resistant to retinoids . SiHa cervical squamous carcinoma cells have been reported to be unresponsive to retinoids . In our experiments, SiHa cells exhibited resistance to 10-7 to 10-6 M ATRA-induced growth inhibition, but 10-5 M and especially 10-4 M ATRA significantly inhibited the growth of SiHa cells . A hallmark of growth inhibition is the high expression of the IRF-1 gene . Ultimately, this induction of IRF-1 is responsible for the observed growth inhibition of target cells. Introduction of an IRF-1 expression vector into SiHa cells caused growth arrest(see above). Similarly, in the presence of an antisense IRF-1 expression plasmid, high-dose (10-4 M) ATRA did not inhibit cell growth (viable cells were at 95% of control level, compared to 60% with ATRA but without the IRF-1 expression plasmid). IRF-1 expression was elevated in ATRA-treated cells in a dose-dependent fashion: low-dose (10-6 M) ATRA increased it temporarily, while high-dose (10-4M) ATRA caused a sustained expression of IRF-1 (Fig. and ). However, only high-dose (10-5 to 10-4 M) but not low-dose (10-7 to 10-6 M) ATRA could significantly inhibit cell growth in vitro. These data strongly suggested that duration of IRF-1 expression is critical for cell growth inhibition in this system. Our next set of experiments aimed to determine the molecular events behind this dose-dependent induction of IRF-1 by ATRA. In STAT1 knockout cells , ATRA moderately and temporarily induced IRF-1 mRNA , while IRF-1 induction in the parental cell line (2fTGH) was long lasting under the same conditions. Activation (tyrosine phosphorylation) of STAT1 occurred only after high-dose (10-4 M), but not low-dose (10-6 M), ATRA treatment in SiHa , suggesting that the capability of high-dose ATRA to activate STAT1 was responsible for the long-lasting activation of IRF-1 expression and consequent inhibition of cell growth. Interestingly, the kinetics of this tyrosine phosphorylation was delayed (3 to 13 h) compared to the immediate-early (5 to 15 min) phosphorylation by cytokines . Similarly, binding of the activated STAT1 to the GAS site was a late event (6 h). We observed that ATRA treatment ---at both 10-6 and 10-4 M concentrations ---induced binding of NF-kappaB to a double-stranded oligonucleotide originating from the IRF-1 promoter , similar to that described by others . However, binding of STAT1 to its cognate GAS elements in the IRF-1 promoter was observed only after high-dose (10-4 M) ATRA treatment , suggesting that STAT1 activation was responsible for the sustained IRF-1 expression. Phosphorylation of STAT1 and its binding to the IRF-1 promoter occurred with delayed kinetics compared to IRF-1 activation via NF-kappaB binding. The effects of GAS and NF-kappaB sites on the activity of the IRF-1 promoter might be synergistic and could be accountable for the long-lasting activation of IRF-1 expression. It has been postulated that retinoids directly increase the expression of transcription factors (STAT1 or IRF-1) that play key roles in JAK-STAT signaling, thereby restoring interferon (IFN) sensitivity . Our data suggest that ATRA could restore sensitivity to itself by a dose-dependent induction of STAT1, resulting in a sustained activation of the IRF-1 signal that might be critical in determining downstream cellular responses. The downstream targets might include the activation of apoptotic events associated with the IRF-1/CAS pathway or NF-kB induction or both. The role of these pathways in high-dose ATRA-induced inhibition of cell growth, however, needs further evaluation. Backmatter: PMID- 12204956 TI - In Vitro Whole-Blood Analysis of Cellular Immunity in Patients with Active Coccidioidomycosis by Using the Antigen Preparation T27K AB - Measurement of cellular immunity in human coccidioidomycosis has important diagnostic and prognostic implications. The coccidioidin skin test has been the standard for the measurement of this, but it is not available in the United States. We examined the utility of measuring surface expression of CD69 on T lymphocytes in whole blood incubated with the coccidioidal antigen preparation T27K as an alternative to the skin test. Seventy donors with active coccidioidomycosis were studied. The mean fluorescent intensity (MFI) of CD69 expression on CD3 lymphocytes in response to T27K was 28.61 +- 1.77, significantly greater than the control response of 11.45 +- 0.78 (P < 0.001). The MFI CD69 response to T27K above that for the control (MFI CD69 above control) was 6.35 +- 2.18 for seven subjects with disseminated coccidioidomycosis who were studied within 5 months of diagnosis. This was significantly below the value of 20.17 +- 3.17 for 18 subjects with pulmonary coccidioidomycosis studied within 5 months of diagnosis and the value of 19.58 +- 2.91 for 27 subjects with disseminated coccidioidomycosis studied after 5 months of diagnosis (for both, P < 0.05). There was an inverse correlation between coccidioidal clinical score and MFI CD69 above control for all 34 subjects with disseminated coccidioidomycosis (r = 0.362; P = 0.036) but not for the 36 subjects with pulmonary disease (r < 0.001; P = 0.993). Among 30 subjects for whom data were available, there was a highly significant association between the MFI CD69 above control and the supernatant concentrations of gamma interferon, interleukin-2 (IL-2), and tumor necrosis factor alpha (for all, P < 0.001), but not for IL-4, IL-5, or IL-10. These data indicate that in vitro assessment of CD69 expression on T lymphocytes by using T27K may be a useful measure of cellular immune response among subjects with active coccidioidomycosis. Keywords: Introduction : The association between expression of delayed-type hypersensitivity and the severity of illness in human coccidioidomycosis has been recognized for more than 50 years . Specifically, expression of delayed-type hypersensitivity is predominantly absent often among individuals who have severe, extrathoracic dissemination. On the other hand, delayed-type hypersensitivity is usually present among those with mild, self-limited illness . Moreover, the failure to express coccidioidal delayed-type hypersensitivity appears to predict subsequent relapse after therapy for patients with active coccidioidomycosis is halted . Immunologically, delayed-type hypersensitivity is an expression of cell-mediated immunity. While measurement of induration at the site of intradermal injection of antigen has been the traditional way to measure cell-mediated immunity, in vitro laboratory techniques also are capable of assessing cellular immunity in coccidioidomycosis . In vitro methods are advantageous in that they do not require exposure of the subject to foreign antigen or its preservative and results are not dependent on injection technique, time of reading, and measurement of induration size. In coccidioidomycosis, the use of in vitro methods to measure cell-mediated immunity has taken on further importance because no skin-test reagent has been available in the United States for the past several years. In the present study, we assessed in vitro expression of cellular immunity in coccidioidomycosis by measuring the T-lymphocyte surface expression of CD69, a glycoprotein that is expressed by activated lymphocytes and NK cells , after incubation of whole blood with the antigen preparation T27K. In the present work, we compared CD69 expression with simultaneous cytokine production among donors with various forms of active coccidioidomycosis. We have previously shown that in vitro assessment of cellular immunity by flow cytometry with the antigen T27K distinguishes immune and nonimmune healthy donors on the basis of coccidioidal skin testing when the antigen was still available . MATERIALS AND METHODS : Human subjects. | Seventy patients with active coccidioidomycosis were sequentially recruited and tested as they entered the Valley Fever Clinic of the Southern Arizona Veterans Affairs Health Care System. In this medical center, all those with a diagnosis of coccidioidomycosis are referred to the Valley Fever Clinic. Donors known to be infected with human immunodeficiency virus or known to have undergone allogeneic transplantation were excluded from the study. A variety of clinical data were collected at the time of study, and a clinical score of severity of coccidioidomycosis was determined. This score is based on clinical symptoms, the severity and number of clinical sites of disease, and the immune diffusion complement-fixing coccidioidal antibody titer. In this system, a higher score is associated with more severe coccidioidal illness . The study was approved by the Human Subjects Committee of the University of Arizona. In addition to these patients, 20 healthy donors with known reactivity to coccidioidin and without active coccidioidomycosis and 11 healthy donors without known coccidioidal immunity were also studied. Whole-blood preparation and incubation. | Approximately 5 ml of blood was drawn by venipuncture from each donor and placed into tubes containing lithium heparin. Aliquots of 0.5 ml were added to 15-ml polypropylene conical centrifuge tubes (Corning, Corning, N.Y.) to which 10 or 20 mug of T27K per ml was added. Previous studies did not reveal differences between the two concentrations. Nothing was added to control tubes. T27K was prepared as a solution as described previously . Samples were incubated for 18 to 24 h at 37C in 95% air-5% CO2, with the caps of the tubes kept loose and without rocking. At the end of this time, 50 mul of 20 mM EDTA (GIBCO, Grand Island, N.Y.) was added to each sample for 15 min at room temperature. Subsequently, 5.0 ml of FACSlyse (Becton Dickinson Immuno-cytometry Systems, [BDIS], San Jose, Calif.) was added to each tube to lyse red blood cells. The tubes were incubated for 10 min at room temperature, and then the contents were mixed. Following this, the samples were centrifuged, resuspended in 5.0 ml of wash buffer (phosphate-buffered saline with 1% bovine serum albumin and 0.1% NaN3), and recentrifuged; and the pellet was resuspended in 1.0 ml of wash buffer. To 200 mul of each sample was added 20 mul of phycoerythrin (PE)-labeled anti-CD69 and 20 mul of peridinin chlorophyll protein-labeled anti-CD3 (BDIS), and the mixture was incubated for 30 min at room temperature in the dark. Then, 2.0 ml of wash buffer was added to each sample, the samples were centrifuged again, and the pellet was resuspended in 500 mul of phosphate-buffered saline. After this step, the samples were immediately used for flow cytometry. Flow cytometry. | Cells were analyzed with a single argon-ion laser flow cytometer (488 nm; FACScan; Becton Dickinson, Mountain View, Calif.) with Cell Quest software (Becton Dickinson). Prior to each assay, general calibration and compensation were established by using two-color CaliBRITE beads (Becton Dickinson), prepared according to the directions of the manufacturer, with the results analyzed by using FACSComp software (version 4.1; Becton Dickinson). For every use, a printout was generated detailing the photomultiplier tube (PMT) voltages, laser current, and power. Over the 6-month period encompassing this study, the FLI PMT voltage varied by 14 V (620 to 634 V), the FL2 PMT voltage varied by 26 V (664 to 680 V), and the FL3 PMT voltage varied by 32 V (704 to 736 V). In addition, the laser current varied between 5.18 and 5.22 A, and the laser power varied between 15.00 and 15.05 mW. In addition to these general steps, a sample was prepared from T27K-stimulated cells for each experiment containing fluorescein isothiocyanate-labeled immunoglobulin G2a, PE-labeled immunoglobulin G1, and PerCP-labeled anti-CD3. From this, a CD3+-lymphocyte gate was established and the PMT voltage for each channel was set. Compensation was then optimized by analyzing samples containing fluorescein isothiocyanate-, PE-, and PerCP-labeled anti-CD8 within the CD3+-lymphocyte gate. Compensation was considered optimized when the positive populations were aligned over the negative populations. The surface expression of CD69 was then assessed among CD3 lymphocytes after counting of 10,000 of these cells for each sample and was recorded as the mean fluorescent intensity (MFI). Cytokine bead assay. | The plasma supernatant was drawn off some of the samples prior to preparation for flow cytometry and was frozen at -70C. Subsequently, these samples were thawed and the concentrations of the cytokines gamma interferon (IFN-gamma), interleukin-2 (IL-2), tumor necrosis factor alpha (TNF-alpha), IL-4, IL-5, and IL-10 were measured by a flow cytometric bead assay (BDIS). Samples with levels >5,000 pg/ml, the upper limit of the assay, were not rerun at dilution but were instead listed as 5,000 pg/ml. Statistics. | Statistical analysis of continuous variables between two groups was performed by the Student t test, and the paired t test was used for comparison of data within subjects. Linear regression was used to determine the association of continuous variables. A contingency table was used to assess differences in frequencies between groups by the Fisher exact test being used for two-by-two tables. In all cases, a P value of <0.050 was considered statistically significant. RESULTS : Description of patients. | The 70 patients consisted of 34 individuals with disseminated coccidioidomycosis and 36 patients with pulmonary disease. Reflecting the population of veterans from which the study population was derived, 67 of the subjects in the study were male, with a median age of 66.9 years (age range, 26.8 to 85.4 years). In addition, 21 were African American, 67 were non-Hispanic, and 5 were active cigarette smokers. The median time since the diagnosis of coccidioidomycosis upon entry into the study was 8.9 months (range, 0.2 to 289.4 months). Forty-seven patients were receiving some type of antifungal therapy at the time of the study, while 23 were not. Thirty-four of the patients had underlying diseases. Of these, diabetes was the most common and occurred in 16 patients. Expression of CD69 in response to T27K. | Among all 70 patients, the mean +- standard error of the mean MFI CD69 in response to incubation with T27K was 28.61 +- 1.77, significantly greater than the response of the control, whose samples were not incubated with T27K, of 11.45 +- 0.78 (P < 0.001). In order to account for the low but persistent expression of CD69 by CD3 lymphocytes without antigen stimulation, subsequent data are expressed as the difference, for each donor, in MFI CD69 between the sample incubated with T27K and the control sample not incubated with T27K (MFI CD69 above control). Hence, the results for each donor were normalized by subtracting the results for the sample containing T27K from the results for the control sample that did not contain T27K. The mean +- standard error of the mean MFI CD69 above control in response to T27K for all 70 patients was 17.16 +- 1.66 and ranged from -1.45 to 53.7. The MFI CD69 above control for the 34 patients with disseminated coccidioidomycosis was 16.86 +- 2.52 and was not significantly different from value for the 36 patients with pulmonary disease, 17.44 +- 2.22 (P = 0.861). The results for the patients with active coccidioidomycosis were not significantly different from the value of 15.21 +- 1.58 obtained for the 20 healthy immune donors with known coccidioidal immunity (P = 0.549). On the other hand, the MFI CD69 above control for 11 healthy nonimmune donors was only 1.58 +- 0.69, significantly less than the values for the patients with active coccidioidomycosis and the healthy, immune donors (for all, P < 0.001). There were no differences in the MFI CD69 above control response to T27K of CD3 cells on the basis of underlying disease, including diabetes, age of 60 years or more, or race (for all, P > 0.050) (data not shown). Because antifungal therapy may have changed the initial coccidioidal cellular immune response over time, particularly among those with disseminated disease, data were next analyzed by comparing subjects who were studied within 5 months of diagnosis to those who were studied after 5 months of diagnosis. These data are displayed in Fig. . Among the seven subjects with disseminated coccidioidomycosis studied within 5 months of diagnosis, the MFI CD69 above control was 6.35 +- 2.18, significantly less than the value of 20.17 +- 3.17 for the 18 subjects with pulmonary coccidioidomycosis studied within 5 months of diagnosis (P = 0.016) and the value of 19.58 +- 2.91 for the 27 patients with disseminated coccidioidomycosis studied more than 5 months after diagnosis (P = 0.031). This difference was not due to a difference in background expression of CD69. The expression of CD69 in control, unstimulated samples from those with a diagnosis of pulmonary coccidioidomycosis of less than 5 months' duration was 8.45 +- 1.38, not different from the value of 10.83 +- 3.30 for those with disseminated disease of less than 5 months' duration (P = 0.434). Association of CD69 expression in response to T27K and clinical coccidioidomycosis score. | As shown in Fig. , among the 34 subjects with disseminated coccidioidomycosis, the MFI CD69 above control in response to T27K was found to be significantly and inversely associated with the clinical coccidioidomycosis score (r = 0.362; P = 0.036). On the other hand, among the 36 donors with pulmonary coccidioidomycosis, there was no correlation between MFI CD69 above control in response to T27K and the clinical coccidioidomycosis score (r < 0.001; P = 0.993). Association of CD69 response with cytokine production. | For 30 subjects, the plasma supernatant was saved and subsequently tested to determine the relationship between CD69 expression and cytokine production after incubation with 10 mug of T27K per ml. Fifteen of these donors had disseminated coccidioidomycosis and 15 had pulmonary disease. As depicted in Fig. , there was a highly significant and positive association between MFI CD69 above control and the concentrations of IFN-gamma, IL-2, and TNF-alpha in the plasma supernatant (for all three, P < 0.001). The levels of IL-4, IL-5, and IL-10 were low and not significantly associated with MFI CD69 above control. The relation between CD69 expression and cytokine production was not different among the donors with disseminated or pulmonary coccidioidomycosis (data not shown). FIG. 1. | MFI CD69 above control for CD3 lymphocytes in response to T27K among 34 donors with disseminated coccidioidomycosis (black boxes) and 36 donors with pulmonary coccidioidomycosis (open boxes). MFI CD69 above control for CD3 lymphocytes in response to T27K among 34 donors with disseminated coccidioidomycosis (black boxes) and 36 donors with pulmonary coccidioidomycosis (open boxes). Patients studied within 5 months of diagnosis were compared to those studied more than 5 months after diagnosis. *, P < 0.05 when donors with disseminated coccidioidomycosis studied within 5 months of diagnosis are compared to donors with pulmonary disease studied within 5 months of diagnosis and donors with disseminated disease studied after 5 months of diagnosis. Values in parentheses indicate the number of subjects in each group. FIG. 2. | Correlation between clinical coccidioidomycosis score and MFI CD69 above control in response to T27K among 34 subjects with disseminated coccidioidomycosis and 36 subjects with pulmonary coccidioidomycosis. Correlation between clinical coccidioidomycosis score and MFI CD69 above control in response to T27K among 34 subjects with disseminated coccidioidomycosis and 36 subjects with pulmonary coccidioidomycosis. FIG. 3. | Comparison of MFI CD69 above control for CD3 lymphocytes after incubation with 10 mug of T27K per ml and the concentrations of IFN-gamma, IL-2, TNF-alpha, IL-4, IL-5, and IL-10 in plasma supernatants among 30 subjects with coccidioidomycosis. Comparison of MFI CD69 above control for CD3 lymphocytes after incubation with 10 mug of T27K per ml and the concentrations of IFN-gamma, IL-2, TNF-alpha, IL-4, IL-5, and IL-10 in plasma supernatants among 30 subjects with coccidioidomycosis. DISCUSSION : Measurement of delayed-type hypersensitivity by skin testing has been the standard technique by which cell-mediated immunity has been determined among individuals with coccidioidomycosis . This test is not available in the United States, and alternatives are required to assess coccidioidal cellular immunity. Assessment of cellular immunity in coccidioidomycosis serves several purposes. First, it may be used to establish the prevalence of coccidioidal infection in a population within a geographic region . Second, it may be used to determine new infection, if prior assessment of immunity has been performed and has been found to be lacking. Finally, expression of cell-mediated immunity appears to reflect a protective host response and may auger an improved clinical outcome among individuals with symptomatic coccidioidomycosis . In the present study, the surface expression of CD69 on CD3 lymphocytes after incubation with the coccidioidal antigen preparation T27K was used to ascertain the cellular immune status of donors with active coccidioidomycosis. The response to T27K of samples from all donors was significantly above that for unstimulated control samples and above the values obtained for a group of healthy, nonimmune donors. When the MFI CD69 above control in response to T27K was examined, there was no overall difference in response between donors with disseminated coccidioidomycosis and those with pulmonary disease, nor were there differences in response on the basis of underlying disease, age, or race. The average age of the patients with active coccidioidomycosis was nearly 67 years. While this appears old, recent data indicate that symptomatic coccidioidomycosis is far more prevalent among those aged 60 years and over than among younger subjects . Because all patients with active coccidioidomycosis at the medical center were referred to the coccidioidomycosis clinic and were eligible for study, the older age most likely reflects the predisposition for symptomatic illness among an older age group rather than any other factor. However, there were significant differences on the basis of the length of time between diagnosis and study. When studied early in the course of their illness, those with disseminated disease had a decreased response to T27K compared to the response of either those studied later or those with pulmonary disease. These data suggest that antifungal therapy may ameliorate the depression in cellular immune response that occurs in disseminated coccidioidomycosis that has been observed previously . Barbee and Hicks have demonstrated such a return in cellular immune response with therapy in a small cohort of patients using an in vitro lymphocyte transformation assay. We are prospectively examining the cellular immune response over time in patients with various forms of coccidioidomycosis on antifungal therapy to more fully explore this question. The coccidioidal antigen preparation T27K used in these experiments is the soluble, aqueous supernatant obtained after mechanically disrupted thimerosal-preserved spherules are centrifuged at 27,000 x g. Previous work with formalin as the preservative has shown that it protects mice from experimental coccidioidal infection when combined with alum . To date, the components of T27K have not been fully defined, but several coccidioidal antigens have been identified in the preparation, including chitinase, chitobiase, aspartyl protease, Ag2/proline-rich antigen, alkaline phosphatase, serine protease, and tube precipitin. Some of these antigens have been identified in both coccidioidin and spherulin, but neither of the latter two preparations has ever been fully characterized. The use of T27K in an in vitro assay is in contrast to earlier studies, in which either coccidioidin or spherulin was used as the skin-test reagent. While it would be of great interest to compare these older antigen preparations to T27K, they are, unfortunately, no longer available. Because of the laboratory and bioterrorism risk, preparation of either requires a licensed biosafety level 3 facility. Moreover, significant lot-to-lot variability occurs with these preparations, making any comparisons difficult. In previous work, we have found that the in vitro response to T27K among healthy donors corresponds closely with skin-test positivity to spherulin . The clinical score is a measure of coccidioidal severity and is increased in patients with severe disease and declines as disease improves with antifungal therapy. There was a significant inverse association between the score of coccidioidal illness among donors with disseminated coccidioidomycosis and the expression of CD69. However, no such association was seen among those with active pulmonary disease. This observation is consistent with observations first made by Smith and Beard with the use of coccidioidin skin testing and with the findings of other investigators who used both skin testing and in vitro assays of cell-mediated immunity . CD69 expression on CD3 lymphocytes in response to T27K among donors with coccidioidomycosis was directly and positively associated with production of the cytokines IFN-gamma and IL-2 but not with the production of the T-helper type 2 cytokines IL-4, IL-5, and IL-10. This suggests that CD69 expression is associated with a T-helper type 1 cellular immune response in patients with coccidioidomycosis. These data and our previous work demonstrating clear distinctions between healthy immune and nonimmune donors suggest that T27K is a biologically useful antigen in determining the cellular immune response to coccidioidomycosis. The close association between the production of T-helper type 1 cytokines, particularly IFN-gamma, and CD69 expression in response to T27K suggests that assessment of these cytokines alone or in association with CD69 expression might be useful in measuring the coccidioidal cellular immune response. At this time, we do not have enough data to determine whether this is the case. Studies to examine this are planned for the future. Expression of CD69 on lymphocytes has been shown to be associated with cellular immune function in patients with a variety of other conditions . In addition, other rapid assays have been found to be useful in determining the cellular immune response in patients with other granulomatous conditions. In particular, the measurement of IFN-gamma in whole blood incubated with tuberculin antigen has been found to compare favorably with purified protein derivative skin testing in patients with a variety of forms of tuberculosis . In the present study, expression of CD69 on CD3 lymphocytes in response to the coccidioidal antigen preparation T27K was a rapid, easily performed assay that required less blood than many standard clinical laboratory tests and provided useful information regarding coccidioidomycosis-specific cellular immunity. Therefore, even if the reagents for performing coccidioidal skin testing were still available, there would be several advantages to using an in vitro system, such as measurement of CD69 expression. The present study, while demonstrating clinical associations with the in vitro assay, did not clearly show a direct relationship between the assay result and clinical disease or outcome. It is imperative that further studies examining the utility of in vitro measurement of coccidioidal cellular immunity be performed to further assess this possibility. Backmatter: PMID- 12204957 TI - Immunoglobulin G Antibody against Helicobacter pylori: Clinical Implications of Levels Found in Serum AB - The clinical significance of high levels of antibody against Helicobacter pylori is still unclear. We sought to evaluate whether the serum antibody levels could predict the presence of macroscopic gastroduodenal disease, to identify factors that correlate with antibody levels in a multivariate context, and to determine the predictive value of antibody levels for diagnosing H. pylori infection. The grades of gastritis and density of H. pylori colonization were scored separately using the updated Sydney system for antral and body mucosa. An enzyme-linked immunosorbent assay (ELISA) for the quantitative detection in serum of IgG antibodies to H. pylori was performed. Of the 170 dyspeptic patients, 105 (62%) had H. pylori infection. There was no difference in antibody levels among endoscopic findings of normal mucosa, chronic gastritis, and duodenal ulcer. On multivariate linear regression analysis, the status of H. pylori infection, mononuclear cell infiltration of body mucosa, and age correlated with antibody levels. The negative predictive value for antibody levels of <30 U/ml is 94%, and the positive predictive value of antibody levels of >70 U/ml is 98%. We conclude that serum antibody levels do not predict the severity of gastroduodenal diseases or the density of H. pylori colonization in H. pylori-infected dyspeptic patients. Higher levels are associated with the presence of H. pylori infection, the chronic gastritis score of the corpus, and older age. Setting a gray zone is necessary for ELISA, since the accuracy in this zone does not allow a precise determination of H. pylori status. Keywords: Introduction : Helicobacter pylori has been established as an important etiological factor for chronic gastritis and duodenal ulcer . It is also associated with gastric ulcer and gastric cancer . As the test-and-treat policy for H. pylori infection is gradually being accepted by general practitioners, a simple, reliable, and noninvasive diagnostic test for H. pylori has become essential in clinical practice . An easier and cheaper way to diagnose H. pylori infection noninvasively is to test for antibodies to the infection. Enzyme-linked immunosorbent assay (ELISA) has been the most commonly used serological test because it is suited for screening large populations . Clinically, some patients are very concerned about contracting H. pylori infection when they are told they have high antibody levels. However, the clinical significance of the levels is still unclear. The correlation between anti-H. pylori antibody levels and the severity of histological gastritis or H. pylori density has been studied with conflicting results . This discrepancy in results may arise from differences in classification and grading of gastritis, the numbers of subjects, consideration of confounding factors, and statistical methods. One study reported that immunoglobulin G (IgG) antibody levels correlated with the grades of antral polymorphonuclear leukocyte infiltration and antral bacterial density , while another study demonstrated that having higher levels implies advanced antral gastritis with either acute or chronic inflammation . Both studies concluded that levels in themselves do not provide a method for predicting the presence of an ulcer. However, the sample size was small in one study, and no body mucosa was evaluated in either study. Therefore, we conducted this study to identify which factors correlate with antibody levels in a multivariate context and to determine if antibody levels could predict the presence of macroscopic gastroduodenal disease. The predictive value of antibody levels for diagnosing H. pylori infection was also evaluated. MATERIALS AND METHODS : Dyspeptic patients who were scheduled for upper gastrointestinal endoscopy were recruited into the study. Patients with any of the following conditions were excluded: (i) ulcer complications, such as bleeding, stenosis, or perforation; (ii) previous stomach surgery; (iii) intake of any substitute for benzimidazoles or preparations containing bismuth within 1 month prior to administration of the test; or (iv) having been treated with or currently on anti-H. pylori therapy. For serology studies, blood was drawn immediately after endoscopy, and sera were collected and stored at -70C until they were assayed. Histology. | During endoscopy, two sets of biopsy specimens from the antrum and the greater curvature of the midbody were obtained for rapid urease testing (CLO Test; Delta West, Bentley, Australia) and histology. Hematoxylin and eosin stains were used to grade the gastritis. The grades of gastritis were assessed for infiltration of mononuclear cells or neutrophils, mucosal atrophy, and intestinal metaplasia using visual analogue scales described in the updated Sydney system on a four-point scale (0, normal; 1, mild; 2, moderate; and 3, marked). A modified Giemsa stain was used to assess the density of H. pylori . The number of lymphoid follicles was recorded. All histological sections were evaluated by the same pathologist, who was blinded to the patients' clinical conditions. Quantitative ELISA. | Serum specimens were tested for the presence of IgG antibodies against H. pylori using a quantitative ELISA (HEL-pTEST II; AMRAD, Kew, Australia). Reference standards were used to produce a standard curve to quantitate H. pylori antibody levels in patient samples. The results were expressed in arbitrary units per milliliter. The antigen was an inactivated native antigen of H. pylori. On the day of testing, we added 100 mul of diluted specimens, diluted positive and negative controls, and duplicates of reference standards 1 to 4 to the appropriate wells of the microtiter plate. The plate was incubated for 15 min at room temperature and then washed six times with a wash buffer. After the washing, 100 mul of sheep anti-human IgG conjugated to horseradish peroxidase was added to each well. After a further 15 min of incubation, 100 mul of substrate reagent was added to each well and incubated in the dark for 15 min at room temperature. Then, stopping solution (100 mul of H2SO4) was added to each well to terminate the enzymatic reaction. The absorbance was read within 30 min using a 450-nm-pore-size filter with a 620-nm-pore-size filter as a reference. H. pylori status. | A patient was classified as H. pylori infected if both the CLO and histological tests were positive. A patient was classified as non-H. pylori-infected if both methods were negative. Patients who had only one positive CLO or histology test were considered to be of indeterminate status. Statistical analysis. | Continuous data were analyzed by the two-tailed Student's t test or analysis of variance (ANOVA). When ANOVA of multiple groups revealed a difference at the 5% level, the post hoc multiple-comparison technique was used to determine which pairs of groups were responsible for the overall difference. Spearman's rank correlation was used to examine the relationship between antibody levels and grades of gastritis. Multiple stepwise linear regression was also used to identify which factors were related to the antibody levels. The data were processed using SPSS software. Positive and negative predictive values were calculated for the ELISA (HEL-pTEST II), using gastric biopsy results as the "gold standard." RESULTS : One hundred and seventy patients (88 men and 82 women; mean age +- standard deviation [SD], 43.7 +- 13.5) were included in the study. There were 34 patients with normal endoscopic findings, 62 patients with gastritis, 57 patients with duodenal ulcers, 5 patients with gastric ulcers, 2 patients with combined gastric and duodenal ulcers, and 10 patients with other findings. Based on the proposed gold standard, 105 patients were infected, 64 were not infected, and 1 was of indeterminate status. Anti-H. pylori antibody levels did not show a normal distribution (skewness, 1.888; kurtosis, 2.839), but its logarithm did (skewness, 0.033; kurtosis, -0.967), so logarithmic transformation was performed in the assessment of anti-H. pylori antibody levels during the two-tailed Student's t-test or ANOVA by Scheffe's method or a linear regression. There was no difference in antibody levels between men and women or smokers and nonsmokers or among different blood groups. The influence of drinking habits on antibody levels was not evaluated because only six patients were regular drinkers. For dyspeptic patients with H. pylori infection, there was no difference in antibody levels among the endoscopic findings for the normal mucosa (n = 12), chronic gastritis (n = 36), and duodenal ulcers (n = 46) (223 +- 220, 227 +- 223, and 228 +- 196 U/ml [mean +- SD], respectively; P = 0.99). The mean serum anti-H. pylori antibody levels in relation to the grading of the four histological features of the antral mucosa and body mucosa are illustrated in Fig. and , respectively. We see a trend for the mean antibody levels to rise as the grades of gastritis and H. pylori density increase. Significant correlations were found between the antibody levels and the grades of gastritis in the antral mucosa (Spearman's rank correlation: r = 0.586 and P < 0.001 for neutrophil infiltration, r = 0.620 and P < 0.001 for mononuclear cell infiltration, r = 0.429 and P < 0.001 for atrophy, r = 0.630 and P < 0.001 for H. pylori density, and r = 0.438 and P < 0.001 for lymphoid follicles) and in the body mucosa (Spearman's rank correlation: r = 0.598 and P < 0.001 for neutrophil infiltration, r = 0.618 and P < 0.001 for mononuclear cell infiltration, r = 0.481 and P < 0.001 for atrophy, r = 0.680 and P < 0.001 for H. pylori density, and r = 0.190 and P < 0.05 for lymphoid follicles). However, if the calculation was restricted to H. pylori-infected patients, only the grades of neutrophil infiltration and mononuclear cell infiltration, atrophy, and the number of lymphoid follicles in the body mucosa retained significant correlation with antibody levels (r = 0.248 and P < 0.05, r = 0.292 and P < 0.005, r = 0.218 and P < 0.05, and r = 0.230 and P < 0.05, respectively). This is also noted in estimating correlation between antibody levels and age. Significant correlations were found when all subjects were considered (Spearman's rank correlation: r = 0.200; P < 0.01), but there was no significant correlation in H. pylori-infected patients (Spearman's rank correlation: r = 0.154; P = 0.117). However, if age is divided into older (>=45 years) and younger (<45 years) groups, the difference in antibody levels between the two groups in H. pylori-positive patients approached significance (median, 164, and range, 38 to 823 versus 124 and 21 to 804; P = 0.09). Intestinal metaplasia in antral specimens was observed in 18 patients (15 with score 2, 2 with score 3, and 1 with score 4), so scores 2, 3, and 4 were grouped together. There was no difference in the mean antibody levels between patients with and without intestinal metaplasia (P = 0.984). No intestinal metaplasia was observed in the body specimens. A multiple stepwise linear regression was performed to analyze the relationship between antibody levels and the following predictors: sex, age, smoking, blood groups, grades of antral and body gastritis, number of lymphoid follicles, H. pylori density, and status of H. pylori infection. Table shows that the status of H. pylori infection, grade of mononuclear cell infiltration of body mucosa, and age were three independent factors correlated with antibody levels (R = 0.780; R2 = 0.608). Figure shows the distribution of anti-H. pylori antibody (IgG) levels in H. pylori-infected and uninfected patients (n = 169). The negative predictive value of antibody levels of <20 U/ml is 100%, and that for levels of <30 is 94%. The positive predictive value of antibody levels of >70 U/ml is 98%. There were 27 patients (16%) whose antibody levels fell between 30 and 70 U/ml. Among them, 14 patients were H. pylori infected and the rest were non-H. pylori infected, judged by the reference standards. This suggests that little more than a guess is involved when making a positive-negative decision in this gray zone. FIG. 1. | Comparison of serum anti-H. pylori Comparison of serum anti-H. pylori antibody (IgG) levels among groups based on grading scores of four histological features of antral mucosa. Scores: 0, normal; 1, mild; 2, moderate; and 3, marked. Each column represents the mean +- standard error of the mean. **, P < 0.01 compared to a score of 0; ++, P < 0.01 compared to a score of 1. FIG. 2. | Comparison of serum anti-H. pylori Comparison of serum anti-H. pylori antibody (IgG) levels among groups based on grading scores of four histological features of body mucosa. Scores: 0, normal; 1, mild; 2, moderate; and 3, marked. Each column represents the mean +- standard error of the mean. **, P < 0.01 compared to a score of 0; ++, P < 0.01 compared to a score of 1. FIG. 3. | Distribution of anti-H. pylori Distribution of anti-H. pylori antibody (IgG) levels in H. pylori-infected and uninfected patients (n = 169). TABLE 1 | Multiple stepwise linear regression analysis of factors correlated with antibody levels DISCUSSION : In our present study, there was no difference in antibody levels of the normal mucosa, gastritis, and duodenal ulcers in H. pylori-infected patients. This is in accordance with the study by Sharma et al., who reported that H. pylori serology cannot be used to predict the presence or absence of gastroduodenal ulcer disease . Similarly, Sheu et al. reported no difference in antibody levels between nonulcer and ulcer subgroups , and Hsu et al. reported that levels of anti-H. pylori in serum are not predictive of the severity of macroscopic gastrointestinal damage. The correlation between anti-H. pylori antibody levels and the severity of histological gastritis or H. pylori density has been studied with conflicting results . Significant correlation was reported in studies in which non-H. pylori-infected patients were included in the analysis . On the other hand, no such correlation was found in one report in which only H. pylori-infected patients were analyzed . More than 90% of H. pylori-infected patients have detectable serum IgG antibodies, and the vast majority of H. pylori-infected patients have chronic active gastritis in the antral biopsy specimens. Therefore, the status of H. pylori infection is a crucial confounding factor in analyzing the relationship between antibody levels and histological gastritis. If non-H. pylori-infected patients are included in the analysis, the correlation between antibody levels and gastritis scores or H. pylori density may be misleading due to their common association with the status of H. pylori. Our study also supports these observations. In this study, grades of active and chronic inflammation, atrophy, lymphoid follicles, and H. pylori density were correlated with antibody levels in the antral mucosa but were not correlated if only H. pylori-infected patients were analyzed. Our results had some differences from the reports of Hsu et al. and Sheu et al. . In the study by Hsu et al., antibody levels did not correlate with grades of antral mononuclear cell infiltration, mucosal atrophy, or intestinal metaplasia but did correlate with grades of antral neutrophil infiltration and antral bacterial density in H. pylori-infected patients. The ELISA kit they used was the same as ours. Since anti-H. pylori antibody levels do not show a normal distribution, logarithmic transformation is more appropriate when performing a parametric test. Therefore, the variance may result from the small number of cases in their study (n = 36) and different statistical methods. In the study by Sheu et al., antibody levels correlated with the severity of acute or chronic inflammation. However, the rate of absence of neutrophil infiltration (score, 0) in H. pylori-infected patients in their study was 63%. This is contrary to the well-established concept that the majority of H. pylori-infected patients have chronic active gastritis in the antral biopsy specimens. In addition, the ANOVA model with Duncan's test is not an appropriate statistical method for data with unequal numbers in the groups. In this study, grades of active and chronic inflammation, atrophy, and lymphoid follicles in the body mucosa remained significantly correlated with antibody levels in H. pylori-infected patients. The correlation coefficients, however, were reduced to half of those when non-H. pylori-infected patients were included. This indicates that the status of H. pylori infection is an important contributing factor, which has been demonstrated in the multivariate linear regression model. The data in this study were obtained from dyspeptic patients, so they may not generalize to asymptomatic subjects. There has been a report showing that diffuse antral gastritis with prominent lymphoid follicles has the highest antibody levels . However, in that study, no statistical data were presented, so we calculated the difference between diffuse antral gastritis and chronic atrophic gastritis from the data presented in the table and found no statistical difference (0.274 +- 0.025 versus 0.266 +- 0.013 [mean +- SD]; P = 0.757). In our study, antibody levels of <30 or >70 U/ml had good predictive value for the absence or presence, respectively, of H. pylori infection. For individuals having a value between 30 and 70 U/ml, the so-called gray zone, a different diagnostic method is recommended to assess H. pylori status more accurately. When antibody levels are >70 U/ml, higher levels do not improve the positive predictive value. However, our previous study of the rapid serological test has demonstrated that antibody levels in ELISA are correlated with the density of the reaction band in the rapid serological test . The higher the antibody levels, the more clearly the band is noted. The lower the antibody levels, the more likely that it will be regarded as negative in the rapid serological test due to a very weak color change of the reaction band. Thus, higher antibody levels have a favorable effect on the diagnosis of H. pylori infection in rapid serological tests. There were several studies demonstrating that the sensitivity of serological diagnosis of H. pylori infection was higher in the older group (>=45 years) than in the younger group (<45 years) . Our study showing higher antibody levels in the older group in H. pylori-infected patients could explain this phenomenon, because the higher the antibody levels, the fewer false negatives there will be. In conclusion, levels of IgG anti-H. pylori antibody in the serum do not predict the presence of macroscopic gastroduodenal diseases or the density of H. pylori colonization in H. pylori-infected dyspeptic patients. The presence of H. pylori infection, higher updated Sydney system chronic gastritis scores of the corpus, and older age are associated with higher antibody levels. Setting a gray zone is necessary for ELISA, since the accuracy in this zone does not allow a precise determination of H. pylori status. Backmatter: PMID- 12204962 TI - Clinical and Epidemiological Relevance of Quantitating Hepatitis E Virus-Specific Immunoglobulin M AB - Diagnosis of acute hepatitis E by detection of hepatitis E virus (HEV)-specific immunoglobulin M (IgM) is an established procedure. We investigated whether quantitation of HEV IgM and its ratio to HEV total Ig furnished more information than conventional IgM tests that are interpreted as positive or negative. A previously described indirect immunoassay for total Ig against a baculovirus-expressed HEV capsid protein was modified to quantitate HEV-specific IgM in Walter Reed (WR) antibody units by using a reference antiserum and the four-parameter logistic model. A receiver-operating characteristics curve derived from 197 true-positive specimens and 449 true-negative specimens identified 30 WR units/ml as an optimum cut point. The median HEV IgM level in 36 patients with acute hepatitis E fell from 3,000 to 100 WR units/ml over 6 months, suggesting that 100 WR units/ml would be a more appropriate cut point for distinguishing recent from remote IgM responses. Among three hepatitis E case series, determination of the HEV IgM-to-total-Ig ratio in acute-phase serum revealed that most patients had high ratios consistent with primary infections whereas a few had low ratios, suggesting that they had sustained reinfections that elicited anamnestic antibody responses. The diagnostic utility of the new IgM test was similar to that of a commercially available test that uses different HEV antigens. In conclusion, we found that HEV IgM can be detected specifically in >95% of acute hepatitis E cases defined by detection of the virus genome in serum and that quantitation of HEV IgM and its ratio to total Ig provides insight into infection timing and prior immunity. Keywords: Introduction : Hepatitis E is acute, self-limited hepatitis caused by a virus of the same name (hepatitis E virus [HEV]) that is excreted in feces and transmitted orally. In large parts of Asia and Africa, this disease is common, causing sporadic and epidemic illness . Diagnosis of acute hepatitis E is based on detection of the HEV genome in serum or feces by reverse transcription-PCR (RT-PCR) or detection of newly elicited antibodies to HEV, in particular HEV-specific immunoglobulin M (IgM) . An IgM test is marketed in Asia ; this test uses recombinant HEV antigens derived from the carboxyl terminus of the capsid protein (ORF-2) and ORF-3. The good diagnostic utility of the marketed test has been characterized . Moreover, several research laboratories have developed IgM tests based on alternative recombinant HEV (rHEV) antigens expressed in bacteria or by use of the baculovirus system . Recently, we reported an indirect enzyme immunoassay (EIA) for total Ig against a baculovirus-expressed HEV capsid protein that quantitated antibodies to HEV in Walter Reed (WR) antibody units by using a reference antiserum and the four-parameter logistic model . We modified this test to detect HEV-specific IgM and employed the IgM and total-Ig tests together to characterize serum specimens from patients with suspected acute hepatitis E. We investigated whether quantitation of HEV IgM and its ratio to HEV total Ig furnished more diagnostic or epidemiological information than conventional IgM tests that are interpreted as positive or negative. Here we report the development of an HEV IgM quantitation standard, the protocol for the IgM test, the kinetics of HEV IgM and total-Ig responses over 6 months in a case series of patients with hepatitis E, an extensive characterization of the test's sensitivity and specificity, the use of the IgM-to-total-Ig ratio to identify rare cases of clinically overt reinfection, and our test's good concordance with the marketed IgM test. We found that quantitation of IgM and total Ig together furnished novel insight into infection timing and prior immunity. MATERIALS AND METHODS : RT-PCR. | Serum specimens were tested for the HEV genome, indicating viremia during acute infection, by use of previously published protocols that detect either a conserved region of ORF1 or ORF2 . The previously unpublished HEV ORF2 nested PCR primers, designated set 3, are listed in Table . Reference human antibodies. | Equal aliquots of acute-phase serum from 20 hepatitis E patients from Nepal were pooled; each case was diagnosed by detection of HEV viremia by RT-PCR. Pool 6, created by diluting the acute-phase serum pool with approximately 3 volumes of serum with HEV-specific total-Ig levels of <0.1 WR unit/ml, was designated the HEV IgM quantitation standard. Pool 7, created by diluting pool 6 with approximately 3 more volumes of the same negative serum, was designated the IgM positive control. Relative potency. | The relative potencies of reference antisera and working antigen lots were determined by parallel line assay and calculation of a common slope, as previously described . rHEV antigens. | The antigen for all assays was a 56-kDa recombinant capsid protein truncated at the amino and carboxyl ends to comprise amino acids 112 to 607 of the 660-amino-acid protein. The protein, made in Spodoptera frugiperda cells by using a baculovirus expression vector, was prepared by Novavax as previously described . All tests used 33 WR antigen units/ml; antigens were from one of the lots previously characterized . EIA protocols. | The IgM assay protocol was identical to the total-Ig protocol except that the goat anti-human Ig-horseradish peroxidase (HRP) conjugate was replaced with goat anti-human IgM-HRP (Kirkegaard and Perry). The optimal 1:4,000 dilution of anti-IgM conjugate was determined by testing twofold dilutions to find the highest signal-to-noise ratio. Serum specimens. | Serum specimens stripped of personal identifiers were from archives at the Walter Reed Army Institute of Research (WRAIR, Silver Spring, Md.) and the Armed Forces Research Institute of Medical Sciences (Bangkok, Thailand). All were from volunteers enrolled in research protocols approved by local institutional review boards and the Human Subjects Research Review Board of the U.S. Army Surgeon General. The majority of hepatitis E serum specimens came from three consecutive case series in Nepal: pregnant women enrolled in an observational cohort study, a cross-sectional study of intrafamilial HEV transmission that identified hepatitis E patients and subclinical infections in persons domiciled with them, and surveillance of hepatitis E among soldiers. IgM EIA control parameters. | Control parameters were developed to ensure accuracy and consistency. Eighteen wells on each 96-well plate were utilized for the following duplicate controls: six half-log dilutions of the HEV IgM quantitation standard (pool 6), the positive control (pool 7), the same negative control as for the total-Ig test, and a no-serum control. Thirty consecutive technically adequate runs were used to calculate limits for control parameters as the mean +- 1.96 (standard deviation) of log-transformed values (expressed as optical density [OD] or WR units per milliliter, as appropriate). Thereafter, assays were accepted according to these limits. Quantitation. | In using the four-parameter logistic model for quantitation, accuracy is greatest at the midpoint of the standard curve and least at the lower and upper limits. We used the procedures for quantitation developed for the total-Ig test (testing in duplicate, OD limits prompting sample dilution and retesting) to ensure consistency . Comparison of antibody potency by WRAIR and Genelabs Technologies IgM EIAs. | To characterize the relationship between antibody potency determined by the WRAIR IgM EIA and that determined by a widely used commercial test (HEV IgM enzyme-linked immunosorbent assay; Genelabs Technologies, Singapore), we tested dilutions of three specimens in both tests. The commercial test employs a mixture of recombinant HEV ORF2 and ORF3 polypeptides expressed in Escherichia coli which are absent from the WRAIR test's 56-kDa rHEV capsid antigen and the ORF2 polypeptide SG3 (amino acids 334 to 660), also expressed in E. coli, which overlaps 53% of the 56-kDa rHEV capsid antigen. The commercial-test results were expressed in OD units, whereas WRAIR EIA results were expressed in WR units per milliliter. Three regression lines were derived, and the portions of each titration curve above each assay's cut point were compared. Additionally, serial serum specimens from six patients with hepatitis E confirmed by detection of HEV viremia were tested in both assays, and their results were compared. TABLE 1 | HEV ORF2 "set 3" nested PCR primers RESULTS : Potency of reference antibodies. | As previously described , we quantified HEV-specific total Ig by using arbitrary WR units per milliliter defined by a convalescent-phase reference serum (pool 4). We chose to quantify HEV IgM also by using WR units per milliliter, in this case by an IgM quantitation standard called pool 6, arbitrarily defined to contain 860 WR units of HEV IgM/ml. The relationship between a unit of HEV IgM and a unit of HEV total Ig is undefined, but these units are roughly equivalent, as pool 6, defined to contain 860 WR units of HEV IgM/ml, was determined to contain 820 WR units of HEV total Ig/ml by parallel line assay against pool 4. Dilutions of pool 6 were used on each assay plate to establish an IgM standard curve. Pool 7 also was used on every assay plate as an IgM positive control. EIA control parameters and assay stability. | To ensure EIA accuracy and consistency, we empirically set control parameters and used individual plate standards and controls . These control parameters were derived from 30 consecutive technically adequate assays. The limited variation of these parameters over those 30 assays is evidence that a skilled serologist can achieve consistent assay performance . Kinetics of IgM and total-Ig responses in patients with hepatitis E. | Consecutive patients with suspected acute viral hepatitis hospitalized at one center in Nepal were evaluated for hepatitis E by an RT-PCR test for viremia. Thirty-six cases of hepatitis E were identified, and HEV IgM and total-Ig levels were determined for four serial serum specimens from each of these patients collected over approximately 6 months . The initial geometric mean IgM level was 3,000 WR units/ml at a median of 8 days after illness onset; the level declined slightly over the next 2 weeks and then exponentially over the next 5 months, nevertheless remaining easily detectable at 100 WR units/ml at a median of 190 days after illness onset. Geometric mean total Ig against HEV was greater than IgM at every time point analyzed, as was expected, but the decline in total Ig mirrored that in IgM. IgM test sensitivity, specificity, and cut point. | After determining the range of IgM levels likely to occur among patients, we analyzed the performance of the IgM test in order to identify a cut point. Specificity was assessed with serum specimens from 449 persons at low risk for acute HEV infection ; these were considered true-negative specimens. Approximately two-thirds were from healthy persons of all ages, including some residing in Nepal, a country where hepatitis E is endemic. Specimens from Nepal residents were collected several months before the annual epidemic, when the monthly infection rate is <2 per 10,000 (M. P. Shrestha and R. M. Scott, unpublished data). Other healthy donors were from the United States or Thailand, where hepatitis E is not endemic. Additionally, about one-third of the true-negative specimens were from patients in Thailand recently infected with bacteria or viruses other than HEV that cause hepatitis. The distribution of HEV IgM levels detected in this negative specimen set is shown in the upper histogram of Fig. . Most values were between 1 and 16 WR units/ml. Sensitivity was assessed with serum specimens from 197 persons with acute hepatitis E infection diagnosed by detection of HEV viremia. Of these, 94% had clinically overt hepatitis E, while 6% had had subclinical infections. These patients were predominantly from Nepal; a few were from Indonesia or Bangladesh; all had their specimens submitted to our laboratories for serologic diagnosis. The distribution of HEV IgM levels detected in this positive specimen set is shown in the lower histogram of Fig. . Most values were between 200 and 20,000 WR units/ml. A receiver-operating characteristics curve was constructed from the test results described above by using cut points of 10, 20, 30, 40, 60, and 100 WR units/ml . A cut point of 30 WR units/ml combines maximal sensitivity and specificity. Nevertheless, because elevated HEV IgM levels appear to persist for 6 months after illness onset, a cut point of 100 WR units/ml seems more appropriate for distinguishing recent from remote infection. Among the 36 cases for which data are summarized in Fig. , the proportion with HEV IgM levels above the cut point of 100 WR units/ml fell from 92% within 3 weeks of illness onset to 83% at 8 weeks after onset and 53% at 6 months after illness onset. Ratio of HEV IgM to total Ig. | We determined the ratio of HEV IgM to total Ig for all 197 serum specimens in the true-positive specimen set. The histogram of those results shows that most values were between 0.1 and 10 and were distributed symmetrically around a median value of 1.0. Nevertheless, <5% of specimens (n = 8) had distinctly low ratios, such that these specimens appeared to constitute a separate population. Serology data for acute-phase and early-convalescent-phase serum specimens from cases representing the 25th, 50th, and 75th percentiles were compared to data from the eight outlying cases with markedly low HEV IgM-to-total-Ig ratios . All patients had HEV viremia demonstrated in the first specimens. Of the eight outlying cases, seven represented patients with clinically overt hepatitis E whereas one case was an inapparent infection detected in a family member of a patient with hepatitis E. The timing of the first specimen collection after illness onset for these eight cases was similar to, or even earlier than, that for the 25th-, 50th-, and 75th-percentile cases, excluding late collection as an explanation for low IgM levels. Nevertheless, there are striking contrasts: among the eight cases with low IgM-to-Ig ratios, the levels of IgM are low and the levels of total Ig are extremely high. These findings suggest that typical cases of hepatitis E represent primary infections with a typical evolution of Ig isotypes from IgM to IgG, whereas a minority of cases represent reinfection (secondary infections), with a typical anamnestic Ig response characterized by low levels of IgM and extremely high levels of IgG. Comparison of WRAIR and Genelabs Technologies IgM EIAs. | We found a consistent and highly correlated relationship between antibody binding assessment by the WRAIR IgM EIA and that by a commercial test widely available in Asia, for pool 6 and for acute-phase serum specimens from two typical hepatitis E cases in Nepal . Nevertheless, a proportion of each titration curve was to the right of the WRAIR cut point (30 WR units/ml) (i.e., positive) but below the Genelabs cut point (i.e., negative). The greater sensitivity of the WRAIR test for low levels of antibody may confer some advantage on that test. The relative utility of the two tests was compared directly in a sample of 19 specimens collected at varying intervals after disease onset (median of 40 days of follow-up after disease onset) from six hepatitis E patients diagnosed by detection of viremia. All specimens were positive in both tests. A line plot of results for each case, determined by the WRAIR test (Fig. , upper graph) and the Genelabs test (Fig. , lower graph), shows that the kinetic responses were similar in four cases and different in two (cases 7 and 10). Overall, the performance of the two tests appeared similar. FIG. 1. | Plots of the HRP conjugate and of positive (pool 7), negative, and no-serum (antigen) controls over 30 technically adequate assays. Plots of the HRP conjugate and of positive (pool 7), negative, and no-serum (antigen) controls over 30 technically adequate assays. FIG. 2. | Levels of HEV IgM (solid line) and total Ig (dashed line) among 36 patients with hepatitis E proven by detection of HEV viremia. Levels of HEV IgM (solid line) and total Ig (dashed line) among 36 patients with hepatitis E proven by detection of HEV viremia. Each patient had four consecutive serum specimens collected. Data are grouped by median day of specimen collection; these values plus the associated interquartile ranges in parentheses are given below the plot of HEV IgM. Error bars represent the 95% confidence interval for the geometric mean. Horizontal dotted reference lines are drawn at 30 and 100 WR units/ml. The table below the line plot gives the proportions of patients with HEV IgM detected at four median time points for assay cut points of 100 and 30 WR units/ml. FIG. 3. | (Top) Histogram of HEV IgM levels determined in a true-negative specimen set (n = 449). (Top) Histogram of HEV IgM levels determined in a true-negative specimen set (n = 449). The interval scale is logarithmic; the mode is 6 WR units/ml. Vertical dotted reference lines are drawn at 30 and 100 WR units/ml. (Bottom) Histogram of HEV IgM levels determined in a true-positive specimen set (n = 197). The interval scale is logarithmic; the mode is 2,512 WR units/ml. Vertical dotted reference lines are drawn at 30 and 100 WR units/ml. FIG. 4. | Receiver-operating characteristic plot for IgM EIA based on true-negative and true-positive specimen sets. Receiver-operating characteristic plot for IgM EIA based on true-negative and true-positive specimen sets. FIG. 5. | Histogram of HEV IgM-to-total-Ig ratios for the true-positive specimen set (n = 197). Histogram of HEV IgM-to-total-Ig ratios for the true-positive specimen set (n = 197). The interval scale is logarithmic; the mode is 1.6. Vertical reference lines mark cumulative distribution percentiles, as follows: dotted lines, 5 and 95%; solid lines, 25 and 75%; dashed-and-dotted line, 50%. Cases to the right of the 5% reference line (n = 189) appear to represent primary infections, whereas cases to the left of the 5% reference line (n = 8) appear to represent secondary infections with anamnestic antibody responses. FIG. 6. | Comparison of IgM quantitation by WRAIR and Genelabs Technologies EIAs. Comparison of IgM quantitation by WRAIR and Genelabs Technologies EIAs. Titration of three acute-phase serum specimens from hepatitis E patients is represented. Horizontal dotted reference line, cut point for the Genelabs test; vertical dotted reference line, cut point (30 WR units/ml) for the WRAIR test. FIG. 7. | Serial determinations of HEV IgM levels for six patients with acute hepatitis E confirmed by RT-PCR detection of HEV viremia in the initial specimen. Serial determinations of HEV IgM levels for six patients with acute hepatitis E confirmed by RT-PCR detection of HEV viremia in the initial specimen. The upper graph shows results obtained by using the WRAIR IgM test; the bottom graph shows results obtained by using the Genelabs Diagnostics test. All 19 specimens were positive in each test. The cut point for each test is shown as a dashed line. TABLE 2 | IgM EIA control parameters defined by their variation over 30 consecutive assays TABLE 3 | Composition of the true-negative specimen set (n = 449) TABLE 4 | Serology data for all secondary cases and representative primary cases of hepatitis E in a sample of 197 patients from areas where hepatitis E is endemic DISCUSSION : We have adapted our previously reported indirect EIA, which is able to quantitate HEV total Ig accurately and reproducibly, to quantitate HEV IgM. This is the third report describing an IgM EIA that uses a recombinant HEV capsid protein expressed in the baculovirus system, but the first HEV IgM test to offer quantitation or to be so exhaustively characterized. The performance of the IgM test was extensively evaluated, and assay specificity, sensitivity, and consistency were carefully documented. We used the IgM and previously described total-Ig tests to determine HEV IgM levels, total-Ig levels, and their ratio in acute-phase serum specimens from almost 200 cases of hepatitis E occurring in countries where hepatitis E is endemic. This dual testing allowed us to make the novel observation that a small proportion of disease is characterized by an anamnestic antibody response, suggesting that these cases result from secondary infection of a person who had previously recovered from a primary infection. We chose an indirect EIA format for detecting and quantitating HEV IgM despite its inferior specificity compared to an IgM isotype-capture EIA. We rejected developing an IgM isotype-capture EIA after initial experiments demonstrated poor sensitivity despite use of substantially greater amounts of the rHEV antigen. We inferred from this observation that the rHEV antigen, in the physical form used in this assay (thawed from -70C storage one to three times before use) and at concentrations that were economically feasible, was inefficient at bridging layers of specific Ig. This behavior of our rHEV antigen limited the assay format to an indirect test. One of the weaknesses of an indirect test is that IgM-rheumatoid factor in a test serum, which has activity against the Fc portion of IgG directed against HEV antigen, may elicit a false-positive result . We substantially reduced the risk of such nonspecific reactions by testing serum initially diluted 1:1,000. The other potential weakness of an indirect EIA format is reduced sensitivity due to competition between virus-specific IgM and IgG for antigen binding sites. The test sensitivity of 92 to 97%, depending on the cut point, suggests that IgG competition is not a limitation of this particular assay. A necessary step in developing the quantitative IgM EIA was the creation of reference antibody pools of HEV IgM by using human serum from Nepal, where hepatitis E is endemic. By trial and error, we set the potency of the IgM quantitation standard so that a WR unit of IgM and a WR unit of total Ig (M plus G plus A isotypes) were comparable. There is no international HEV IgM reference standard, as the available World Health Organization HEV antibody standard contains only low levels of HEV-specific IgM (105.6 WR units/ml). We can provide samples of our IgM and total-Ig reference standards (available from the Department of Virus Diseases, WRAIR, upon request) to interested laboratories who wish to prepare their own in-house standards. Of the several approaches for EIA quantitation of an unknown by using a standard curve, we chose the four-parameter logistic model, which is generally considered the most accurate and reproducible . We retained all procedures from our total-Ig test, previously shown to be reproducibly accurate to below the cut point (7% median error in quantitation of the mid-range standard). To achieve accurate and reproducible quantitation, an operator must perform this test with great care, using well-controlled reagents. This is more likely to be possible when the test is performed routinely, as might be expected in a research serology laboratory or a regional or national public health laboratory. We had access to serial serum specimens collected over 6 months from 36 hepatitis E patients. These specimens demonstrated that IgM antibody levels were very high soon after illness onset, declined little over several weeks, and then declined rapidly to low levels over the next 4 to 6 months. This is typical of IgM responses to other acute, self-limited, systemic viral infections . The weeks-long duration of markedly elevated IgM levels after disease onset means that diagnosis using even relatively insensitive IgM detection methods should be successful, even if patients come to medical attention late. Moreover, the months-long duration of IgM responses to HEV may be a boon to hepatitis E outbreak investigations, since these typically commence months after the index case occurs. A sensitive IgM test should be able to identify most disease cases from late-convalescent-phase serum specimens; the WRAIR test meets this criterion by detecting HEV IgM above a cut point of 30 WR units/ml in 92% of specimens collected a median of 2 months after disease onset and in 83% of specimens collected 6 months after disease onset. Additionally, since many HEV infections are known to be subclinical , sensitive tests for IgM may enable identification of all infected persons rather than those with disease only. The preceding paragraph illustrates some of the ways in which an HEV IgM test might be used. For different uses, different assay cut points may be appropriate. The receiver-operating characteristics curve identified 30 WR units/ml as the cut point suitable for outbreak investigations, in which it is necessary to find remotely infected persons, whereas it identified 100 WR units/ml as a marginally less sensitive cut point that could distinguish a recent infection from a remote one. We began this study anticipating that an HEV IgM test would improve serological diagnosis of hepatitis E, then based on detection of HEV-specific total Ig or IgG. We confirmed that detection of HEV IgM is the best serological test for diagnosis of hepatitis E. Yet the most interesting aspect of our work was the observation that in some cases of hepatitis E, there was a weak or absent IgM response. By combining HEV IgM and total-Ig tests, we identified primary and anamnestic HEV immune responses among adolescent and adult hepatitis E patients in areas of HEV endemicity, distinguished by widely divergent HEV IgM-to-total-Ig ratios. We inferred that anamnestic responses resulted from reinfection (secondary infection) of a person having an immunologic memory of prior HEV infection. In our experience, secondary infections are uncommon, comprising <5% of overt hepatitis E cases. This is an important observation because previously, some authorities have speculated that waning immunity explained why most cases of hepatitis E occurred among adults. The fact that more than 90% of hepatitis E cases in areas of HEV endemicity occur in patients who have a primary antibody response refutes this speculation of waning immunity, since previously exposed persons should mount an anamnestic response upon reexposure. On the other hand, we do speculate that asymptomatic secondary infections may be more common than symptomatic ones. Ultimately, the true prevalence of secondary infection and disease must be assessed prospectively. Such studies also may identify the risk factors that are associated with partial failure of immunity and determine whether clinical outcomes differ between primary and secondary disease. If secondary disease is associated with high-dose HEV exposure or waning immunity, hepatitis E vaccines now under development may have public health utility even among previously exposed persons, if they can be demonstrated to boost protective immunity. Finally, we wanted to evaluate the WRAIR IgM test against the test marketed widely in Asia by Genelabs Diagnostics. The comparison was initiated by testing in parallel serial dilutions of three acute-phase specimens using the WRAIR and Genelabs Diagnostics tests. We found that the tests performed similarly across a range of dilutions, suggesting that these tests were comparable for serological diagnosis of acute disease. On the other hand, the results suggested that the WRAIR test would be more versatile in outbreak investigations based on its apparent superiority at detecting low levels of IgM reflecting remote infection. The comparison was completed by testing a separate panel of 19 specimens from 6 hepatitis E patients with viremia by both tests. In this limited assessment, both tests sensitively detected true acute HEV disease over a median of 40 days of follow-up. This result is consistent with an earlier report by Ghabrah and others , who found 96% concordance (kappa, 0.87) between an IgM test using an antigen similar to that used in the WRAIR test and the Genelabs IgM test. In conclusion, the test method described here enables accurate quantitation of HEV IgM, including low levels associated with remote infection. Detection of HEV IgM is the method of choice for laboratory diagnosis of hepatitis E. By combining quantitation of HEV IgM and total-Ig levels, a new class of secondary infections can be detected. The epidemiological implications of secondary infections are clear, but their clinical implications must be assessed. Backmatter: PMID- 12204960 TI - Use of Egg Yolk-Derived Immunoglobulin as an Alternative to Antibiotic Treatment for Control of Helicobacter pylori Infection AB - The present study evaluated the potential use of immunoglobulin prepared from the egg yolk of hens immunized with Helicobacter pylori (immunoglobulin Y [IgY]-Hp) in the treatment of H. pylori infections. The purity of our purified IgY-Hp was 91.3%, with a yield of 9.4 mg of IgY per ml of egg yolk. The titer for IgY-Hp was 16 times higher than that for IgY in egg yolk from nonimmunized hens, and IgY-Hp significantly inhibited the growth and urease activity of H. pylori in vitro. Bacterial adhesion on AGS cells was definitely reduced by preincubation of both H. pylori (108 CFU/ml) and 10 mg of IgY-Hp/ml. In Mongolian gerbil models, IgY-Hp decreased H. pylori-induced gastric mucosal injury as determined by the degree of lymphocyte and neutrophil infiltration. Therefore, in this experimental model, H. pylori-associated gastritis could be successfully treated by orally administered IgY-Hp. The immunological activity of IgY-Hp stayed active at 60C for 10 min, suggesting that pasteurization can be applied to sterilize the product. Fortification of food products with this immunoglobulin would significantly decrease the H. pylori infection. In conclusion, the IgY-Hp obtained from hens immunized by H. pylori could provide a novel alternative approach to treatment of H. pylori infection. Keywords: Introduction : Helicobacter pylori is the most common cause of gastritis and gastric ulcers and plays a pivotal role in the development of gastric carcinomas . Often a significant portion of those infected do not show symptoms, although chronic infection increases the risk of the development of H. pylori-associated gastric disease. There have been tremendous efforts to evaluate numerous therapeutic regimens for eradication of H. pylori infections. Successful treatment of H. pylori infections most often employs antibiotic therapy, consisting of some combination of metronidazole, amoxicillin, clarithromycin, and either bismuth or a proton pump inhibitor . However, antibiotic therapy fails in 10 to 15% of cases due to the development of antibiotic resistance . Increasing occurrence of antibiotic resistance would further complicate the treatment of H. pylori infections. Consequently, it is important to seek new therapies for a wider means of treating, suppressing, or preventing H. pylori infection without drug resistance problems. The concept of protecting a host with passively derived antibodies is not new. It has been shown that oral administration of antibacterial or antiviral immunoglobulins, through infant formulae or other diet, is effective in preventing intestinal infection . However, oral administration of antibodies is prohibitively expensive when large amounts of antibodies are required . Recently, chicken egg yolk was recognized as an inexpensive, alternative antibody source, and the usefulness of egg yolk immunoglobulin Y (IgY) has been assessed for therapeutic application by passive immunization therapy through oral ingestion of IgY, as in fortified food products for prevention or control of intestinal infections, such as those caused by enterotoxigenic Escherichia coli , Salmonella enterica serovar Typhimurium , and rotavirus . These studies, taken together, provide the potential advantage of using IgY with specificity to H. pylori (IgY-Hp) for controlling H. pylori-associated gastric disease and subsequently prevent disease resulting from chronic infection. Nevertheless, there has been no report so far on the use of IgY in the prevention and treatment of H. pylori infections. Furthermore, for the practical application of IgY-Hp, together with food or pharmaceutical materials to prevent H. pylori-associated disease, the stability of IgY toward heat, acid, and pepsin should be ensured. The aim of this study was to evaluate the potential use of IgY-Hp in the prevention and treatment of H. pylori infections. To achieve this objective, we studied, in vitro and in vivo, the activity of IgY-Hp against H. pylori. The effect of IgY-Hp on gastric mucosal injury induced by H. pylori was evaluated in vivo using Mongolian gerbils. The stability of IgY-Hp was also investigated. MATERIALS AND METHODS : H. pylori preparation and immunization. | H. pylori (ATCC 43504) was cultured in brucella broth (Difco Laboratories, Detroit, Mich.), supplemented with 5% (vol/vol) bovine calf serum (PAA Laboratories Inc., Parker Ford, Pa.) and antibiotics (amphotericin B, 2.5 mug/ml; vancomycin, 10 mug/ml; trimethoprim, 5 mug/ml; and polymyxin B, 2.5 IU/ml; all were from Sigma Chemical Co. [St. Louis, Mo.]) at 37C under 10% CO2 at 200 rpm. The H. pylori was harvested by centrifugation at 12,000 x g for 10 min and disrupted by sonication. Cellular material was removed by centrifugation, and the supernatant was collected (H. pylori whole-cell lysate). The protein concentrations were determined by bicinchoninic acid methods (Pierce, Rockford, Ill.). Brown Leghorn hens (25 weeks old; n = 15) were immunized intramuscularly with H. pylori whole-cell lysate (200 mug/ml, protein) using an equal volume of Freund's complete adjuvant (Difco Laboratories). Each hen was injected at four different sites (250 mul per site) of the leg muscle. Three booster injections, with Freund's incomplete adjuvant, were given at 2-week intervals following the first injection. One month after immunization, the eggs laid were collected daily for 1 month and stored at 4C. The egg yolk was separated, pooled, and frozen prior to purification of IgY. Isolation and purification of IgY-Hp. | Isolation of IgY-Hp was carried out by the method described by Akita and Nakai with modification. Egg yolk was separated from the white, and the yolk preparation was mixed with an equal volume of distilled water for 30 min, followed by the addition of 0.15% (wt/vol) lambda-carrageenan (Wako Pure Chemical Laboratory, Osaka, Japan). After centrifugation at 10,000 x g at 20C for 30 min, the water-soluble fraction (WSF) was collected and filtered through a Whatman filter paper (no. 1) to remove solid lipid materials. The resulting IgY-containing filtrate was further purified by salt precipitation (19% [wt/vol] sodium sulfate) and ultrafiltration (UF) using a UF membrane (Millipore Corp., Bedford, Mass.) cartridge with a molecular weight cutoff of 100 kDa. Purity and yield of IgY were monitored at various stages by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The IgY content was measured by its absorbance at 280 nm. SDS-PAGE. | According to the method of Laemmli , 10% PAGE was done with a Mini-PROTEAN II Cell (Bio-Rad Laboratories, Hercules, Calif.). Under nonreducing conditions, samples were diluted 1:4 with sample buffer (62.6 mM Tris-HCl, pH 6.8, 25% [vol/vol] glycerol, and 2% SDS [vol/vol]). Under reducing conditions, samples were diluted with sample buffer containing 5% (vol/vol) beta-mercaptoethanol and heated for 5 min at 100C. Fifteen microliters of the samples was loaded into each well (3 mug of protein per well). Prestained SDS-PAGE standards (Bio-Rad Laboratories) and standard chicken IgY (Promega, Madison, Wis.) were used as molecular weight markers. ELISA. | To assess the antibody activity of IgY-Hp to H. pylori, we performed the enzyme-linked immunosorbent assay (ELISA) by Akita et al. with modification. Ninety-six-well plates were coated with H. pylori whole-cell lysate (500 ng/well). After blocking with 1% (wt/vol) bovine serum albumin, 100 mul of IgY-Hp (1 mg/ml) was added using a twofold serial dilution. The plates were then washed with phosphate-buffered saline (PBS)-Tween (0.05% Tween 20 in PBS [pH 7.2]) and incubated for 1 h after the addition of alkaline phosphatase-conjugated goat anti-chicken IgY (Promega). The plates were washed with PBS-Tween, and disodium p-nitrophenyl phosphate (Sigma) was added as substrate to each well. After incubation for 10 min, the reaction was stopped by addition of 3 M NaOH. The absorbance was measured at 405 nm using a microplate reader (Multiskan MS; Thermo Labsystems, Helsinki, Finland). Heat, acid, and pepsin stability of IgY-Hp. | IgY-Hp solutions were incubated at 0, 4, 10, 25, 35, 60, 70, 80, and 90C for 10 min. The heat-treated IgY-Hp was cooled in an ice bath. For the pH stability test, the pHs of IgY-Hp solutions were adjusted to the desired pH (pH 2 to 8) with NaOH or HCl, the solutions were incubated at 37C for 4 h, and then each IgY-Hp solution was neutralized. For the pepsin stability tests, the pHs of the IgY-Hp solutions were adjusted to pH 2, pH 4, and pH 6, respectively. IgY-Hp solution of each pH were incubated with 15 mug of pepsin (Sigma)/ml at 37C for 0.5, 1, 2, and 4 h. After incubation, each IgY-Hp solution was neutralized to inactivate the pepsin. The remaining antibody activity was measured by ELISA, following the heat, pH, and pepsin treatments. Antibody activity was represented as a percentage of control. Colony counting. | H. pylori (108 CFU/ml) was incubated with IgY-Hp (1 and 10 mg/ml) for 6 h at 37oC under 10% CO2 at 50 rpm. After incubation, H. pylori was diluted with brucella broth via a 10-fold series dilution. Each 100 mul was inoculated onto brucella agar containing 5% (vol/vol) bovine calf serum, and the plate contents were incubated at 37C under 10% CO2 for 10 days. The colonies were identified as H. pylori by Gram staining and urease, oxidase, and catalase activities. The growth rate (percentage of control) was calculated by colony counting compared to results for the control. Urease activities. | Both IgY-Hp (1 and 10 mg/ml) and H. pylori (108 CFU/ml) were incubated at 37C under 10% CO2 for 6 h at 50 rpm, and then 50 mul of urea base (2% urea and 0.03% phenol red) was added and allowed to react for 30 min. Urease activity was quantified by measuring the optical density at 550 nm, using a modification to the method of Fauchere and Blaser , and was represented as % of control. H. pylori adhesion on AGS cells. | The human gastric carcinoma cell line AGS was obtained from the Korean Cell Line Bank (Seoul, Korea). Cells (105 cells/ml) were cultured in RPMI 1640 medium (Gibco BRL, Grand Island, N.Y.), supplemented with 10% (vol/vol) fetal bovine serum (HyClone Laboratories, Logan, Utah) and antibiotics (100 U of penicillin/ml and 100 mug of streptomycin/ml; HyClone) at 37C under 5% CO2 for 24 h. AGS cells were checked for their H. pylori colonization level in the presence or absence of IgY-Hp using scanning electron microscopy (SEM). SEM. | For electron microscopy investigation of bacterial adhesion, 300 mul of preincubated IgY-Hp (10 mg/ml) and H. pylori (108 CFU/ml) was added to the AGS cells on the chamber slide (Nalge Nunc International, Naperville, Ill.). After incubation for 4 h, the chamber slide was washed with PBS and fixed with 2.5% (vol/vol) glutaraldehyde for 24 h. After washing with PBS, the material was postfixed with 1% OSO4 for 60 min and then washed twice with PBS. Fixed material was dehydrated through a graded ethanol series from 50 to 95%, followed by two washes with absolute ethanol for 15 min. Dehydrated material, in absolute ethanol, was critical point dried in a critical point drier (HCP-2; Hitachi, Tokyo, Japan). Critical point dried material was mounted on an aluminum stub and coated with gold. Adherence of H. pylori to AGS cells was observed by SEM (S2500; Hitachi). Passive immunization with IgY-Hp against H. pylori infection in Mongolian gerbils. | Mongolian gerbils (6-week-old males) were purchased from Bio Animal 21 (Sung-Nam, Korea). The gerbils were infected with H. pylori (0.5 ml, 108 CFU/ml, orally) three times at 12-h intervals. Two weeks after inoculation, immune response to infection was monitored by ELISA. H. pylori-infected gerbils were randomly divided into three groups as follows: (i) H. pylori infection only (n = 10), (ii) treatment with 1 mg of IgY-Hp treatment/ml (n = 10), and (iii) treatment with 10 mg of IgY-Hp/ml (n = 10). Negative control groups were administered IgY-Hp (1 and 10 mg/ml) alone (n = 10). IgY-Hp (1 ml) was administered daily, using an oral feeding needle, for 4 weeks. After the gerbils were fasted for 24 h, the antral portion of stomach was quickly removed and used for histological examinations. Formalin-fixed tissue was processed routinely in paraffin and stained with hematoxylin and eosin for examination by light microscopy. Gastric mucosal injury was classified and scored on a scale of 0, 1, 2, or 3, according to the updated Sydney system . Statistical analysis. | All data were expressed as the means +- standard deviations (SD). The statistical significance was evaluated by Student's t tests. P values of <0.05 were considered statistically significant. RESULTS : Purification and characterization of IgY-Hp from egg yolk. | Isolation of egg yolk IgY, using the lambda-carrageenan method followed by a UF system, was effective. The purity of the IgY obtained was 91.3%, with a yield of 9.4 mg of IgY per ml of egg yolk . The egg yolk proteins obtained during purification of IgY were analyzed using SDS-PAGE. As shown in Fig. , the IgY finally purified, using the UF system, was pure and dissociated into heavy and light chains of 64 and 25 kDa, respectively. The electrophoretic patterns were in perfect accordance with commercially available IgY (Promega) (Fig. , lane 2). After treating the egg yolk with lambda-carrageenan, the number of proteins in the WSF were significantly decreased. These contaminating proteins were removed by salting out with 19% sodium sulfate. The IgY preparation was concentrated and desalted using the UF system. The IgY and IgY-Hp obtained from hens without and with H. pylori immunization, respectively, were examined for immunological properties by ELISA. The ELISA titers were 640 and 10,240 for IgY and IgY-Hp, respectively. These results indicate that IgY-Hp is highly specific to H. pylori . The heat, pH, and pepsin stabilities of IgY-Hp were evaluated by ELISA . The IgY-Hp was stable at 40C. However, at 60C for 10 min, IgY-Hp lost approximately 20% of its antibody activity. The antibody activity of IgY-Hp significantly decreased at 80C and lost up to 90% of its activity. IgY-Hp showed broad stability between the pHs 4 and 8. However, IgY-Hp lost 80 and 70% of its initial activity at pH 2 and 3, respectively. The antibody activity of IgY-Hp, as determined by ELISA, was almost lost when incubated with pepsin at pH 2, while about 92 and 99% of the activity remained at pH 4 and 6, respectively. Effects of IgY-Hp on growth and urease activity of H. pylori in vitro. | To determine the effect of IgY-Hp on the growth and urease activity of H. pylori, we compared growth rates and urease activity with those of the control. The growth rates for H. pylori were 60.0% +- 9.5% and 10.0% +- 10.8% compared with those found for the control, after incubation with 1 mg and 10 mg of IgY-Hp/ml, respectively . When H. pylori (108 CFU/ml) was incubated with IgY-Hp (1 and 10 mg/ml) for 6 h, the urease activity of H. pylori was 75.4% +- 8.5% and 15.5% +- 7.4% compared with that of the control . Effect of IgY-Hp on H. pylori-induced gastritis in the Mongolian gerbils model. | Mongolian gerbils were inoculated with H. pylori. Two weeks after the inoculation, the status of infection was confirmed by measuring the serum antibody (IgG) against H. pylori. The H. pylori-infected gerbils were determined by high IgG titer (data not shown). Oral administration of 10 mg of IgY-Hp/ml for 4 weeks to the H. pylori-infected gerbils improved lymphocyte infiltration compared to that found for the H. pylori infection group with no IgY-Hp (1.2 +- 0.8 versus 2.5 +- 0.5; P < 0.01) . However, the group that was administered 1 mg of IgY-Hp/ml showed no statistically significant difference with the H. pylori infection group (2.4 +- 1.1 versus 2.5 +- 0.5) . The group treated with 10 mg of IgY/ml significantly improved neutrophil infiltration (0.9 +- 0.7 versus 2.3 +- 0.8; P < 0.01) . There was no statistically significant difference between the group treated with 1 mg of IgY-Hp/ml and the H. pylori infection group (2.2 +- 0.5 versus 2.3 +- 0.8) . A high dose of IgY-Hp treatment decreased H. pylori-induced lymphocyte and neutrophil infiltration in gastric mucosa. However, a low dose of IgY failed to protect gerbils from H. pylori-induced gastric mucosal injury. Effect of IgY-Hp on H. pylori attachment on AGS cells. | Figure shows SEM of AGS cells and those infected with H. pylori (108 CFU/ml) in the presence and absence of IgY-Hp (10 mg/ml). Numerous rod-shaped H. pylori organisms, attached to the AGS cells' surfaces, were observed when cells were cultured with H. pylori . On the other hand, the number of adhering H. pylori organisms dramatically decreased when cells were pretreated with IgY-Hp . FIG. 1. | SDS-PAGE patterns of IgY purified from egg yolk. SDS-PAGE patterns of IgY purified from egg yolk. (A) Coomassie-stained SDS-10% PAGE under nonreducing conditions. (B) Reducing conditions of SDS-10% PAGE. Lanes: 1, molecular size marker, 2, chicken IgY, standard immunoglobulin (Promega), 3, egg yolk, 4, WSF after lambda-carrageenan treatment, 5, IgY obtained after salting out, and 6, final IgY obtained by UF. FIG. 2. | The reaction between IgY-Hp and H. pylori The reaction between IgY-Hp and H. pylori. One hundred microliters of IgY-Hp (1 mg/ml) was added with a twofold serial dilution in 96-well plates coated with H. pylori whole-cell lysate (500 ng/well), and the titers were measured using ELISA. IgY was isolated from the egg yolk of nonimmunized hens, and IgY-Hp was obtained from H. pylori-immunized hens. FIG. 3. | Heat, pH, and pepsin stability of IgY-Hp. Heat, pH, and pepsin stability of IgY-Hp. IgY-Hp was treated at various temperatures for 10 min (A), at various pHs for 4 h (B), and with pepsin (15 mul/ml) (C) at pH 2, 4, and 6 for 0.5, 1, 2, and 4 h. Remaining activities after the treatments were measured using ELISA and are expressed as a percentage of the initial activity. FIG. 4. | Effect of IgY-Hp on the growth and urease activity of H. pylori. H. pylori Effect of IgY-Hp on the growth and urease activity of H. pylori. H. pylori (108 CFU/ml) was incubated with of IgY-Hp (1 and 10 mg/ml) for 6 h at 37C under 10% CO2 at 50 rpm. After incubation, bacteria were inoculated on agar plates. The growth rate (percentage of control) was calculated by colony counting compared to the control. Both IgY-Hp (1 and 10 mg/ml) and H. pylori (108 CFU/ml) were incubated at 37oC under 10% CO2 for 6 h at 50 rpm and then 50 mul of urea base (2% urea-0.03% phenol red) was added and reacted for 30 min. Urease activity was measured by a spectrophotometer at 550 nm and is represented as a percentage of the control. The results are shown as the mean +- SD. * and **, statistically significant differences from the values for the control (P < 0.05 and P < 0.01, respectively). FIG. 5. | Effect of IgY-Hp on H. pylori Effect of IgY-Hp on H. pylori-induced gastric mucosal injury. IgY-Hp (1 and 10 mg/ml) was orally administered to Mongolian gerbils for 4 weeks. The lymphocytes and neutrophils were measured by the degree of lymphocyte and neutrophil infiltration, respectively. The results are shown as the mean +- SD. *, statistically significant differences from the values for the H. pylori infection group (P < 0.01). FIG. 6. | SEM findings for H. pylori SEM findings for H. pylori-infected AGS cells in the presence or absence of IgY-Hp. (A) AGS cells. (B) AGS cells infected with H. pylori (108 CFU/ml). (C) AGS cells infected with H. pylori (108 CFU/ml) pretreated with IgY-Hp (10 mg/ml). TABLE 1 | Summary of yield and purity of IgY from egg yolk DISCUSSION : The present study demonstrates that IgY-Hp prepared from the egg yolk of hens immunized with H. pylori is effective in the treatment of H. pylori infections. H. pylori infections are widespread in humans, and, although they can be cured using antimicrobial therapy, this large-scale use of antibiotics leads to the emergence of antibiotic-resistant strains . Moreover, antimicrobial therapeutic cures of H. pylori infections do not lead to immunity from reinfection , and the emergence of antibiotic-resistant strains can increase failure-of-therapy and relapse rates . Consequently, there have been tremendous efforts to seek alternatives to antibiotic-based therapies for a more widely available means of treating, suppressing, or preventing H. pylori infections without drug resistance problems. Recent work in several laboratories, using several animal models, has shown that immunization with defined native and recombinant antigens of H. pylori can protect against H. pylori infections . The mechanisms of protection are still poorly understood. Although prophylactic and therapeutic immunization has been successful in animal models, efficacy data for humans are still lacking. On the other hand, treatment of H. pylori infection by oral administration of active antibodies specific to H. pylori may have merit, due to the antibody being recognized by H. pylori, thus inhibiting adhesion of the bacterium to human epithelial cells more efficiently. Many studies have shown that egg yolk from an immunized hen has an antibody capable of specific recognition in an abundant quantity and is therefore economical . We estimated that 1 ml of the egg yolk (15 ml per egg) contains about 9.4 mg of IgY. A hen lays about 250 eggs (about 4,000 ml of egg yolk) in a year; thus, the eggs laid by an immunized hen in a year would yield 40 g of IgY. Oral administration of IgY from chicken egg yolk has been used successfully by many researchers in preventing many intestinal diseases, such as those caused by enterotoxigenic E. coli and human rotavirus . Sugita-Konishi et al. reported that IgY, obtained from hens immunized with a mixture of formalin-treated pathogenic bacteria, inhibited the growth of Pseudomonas aeruginosa. The production of Staphylococcus aureus enterotoxin A and adhesion of Salmonella typhimurium serovar Enteritidis to cultured human intestinal cells were also inhibited. IgY antibodies inhibited the colonization of teeth by Streptococcus mutans, thus preventing plaque formation in humans . However, there has been no report so far on the use of IgY in the prevention and treatment of H. pylori infections. In this study, IgY-Hp, obtained from hens immunized with H. pylori whole-cell lysate, dramatically inhibited the growth of H. pylori in vitro. If the antibody IgY had no specific effect, no inhibition of bacterial growth would occur. It seems clear from our observations, using gerbil models, that IgY-Hp decreased H. pylori-induced gastric mucosal injury, as determined by the degree of lymphocyte and neutrophil infiltration. Therefore, the therapeutic value of orally administered IgY-Hp, against the experimental model in gerbils, lies in its ability to inhibit the bacterial organism. More convincing evidence in support of the specificity of these antibodies comes from inhibition of H. pylori attachment to AGS cells, as confirmed by SEM. Although the mechanisms by which IgY-Hp prevented H. pylori colonization are yet not elucidated, our results suggest that IgY-Hp would inhibit H. pylori adherence properties. Of interest, the IgY-Hp used in this study significantly inhibited urease activity. The relationship between inhibitions of urease activity and adhesion properties needs to be clarified. In this study, IgY-Hp seems to have two properties: inhibiting adhesion of the bacterium to gastric epithelial cells and demonstrating powerful urease-inhibiting activity. For practical application of IgY-Hp together with food or pharmaceutical materials to treat H. pylori infection, the stability of IgY-Hp toward heat, acid, or pepsin was studied by measuring the remaining activity by ELISA. Immunologically, IgY-Hp stayed active at 60C for 10 min, suggesting that pasteurization can be applied to sterilize the product. The fortification of food products with this immunoglobulin, together with its higher productivity and effectiveness, would significantly decrease the activities of H. pylori infections. In conclusion, the encouraging results of this study indicate that the IgY-Hp obtained from hens immunized by H. pylori may provide a novel approach to the management of H. pylori infections in humans. However, many problems remain unsolved for the clinical application, such as the effect of IgY-Hp in humans, the persistence of the effect of IgY-Hp after cessation of the application, and the potential for eradicating an established infection. Backmatter: PMID- 12204959 TI - Bacillus Species Are Present in Chewing Tobacco Sold in the United States and Evoke Plasma Exudation from the Oral Mucosa AB - Five Bacillus species, predominantly Bacillus megaterium and Bacillus pumilus, were isolated from two popular brands of commercially available chewing tobacco [(5.0 +- 1) x 106 CFU/ml of supernatant; results for four experiments]. Moreover, the supernatant of the Bacillus culture evoked plasma exudation from postcapillary venules in the intact hamster cheek pouch, exudation that was mediated by the kallikrein/kinin metabolic pathway. Taken together, these data indicate that Bacillus species contaminate chewing tobacco commercially available in the United States and elaborate a potent exogenous virulence factor(s) that injures the oral mucosa. Keywords: Introduction : Habitual use of chewing tobacco is on the rise in the United States . This practice may be associated with oral mucosa inflammation and injury in susceptible individuals and may predispose chewing tobacco users to oral epithelial cell dysplasia and cancer . Previous studies from our laboratory showed that chewing tobacco elicits plasma exudation from the oral mucosa by activating oral keratinocytes and the kallikrein/kinin metabolic pathway . However, the toxic constituent(s) of chewing tobacco that accounts for these responses has not been characterized . A physiological disorder of the tobacco plant, termed "frenching," has been known to the tobacco industry since colonial times . Although the etiology of this condition is uncertain, so-called organic toxins produced by Bacillus species and found in the soil where tobacco plants are grown have been implicated . It is conceivable that Bacillus spores could contaminate chewing tobacco processed for human use and germinate when placed onto the oral mucosa. These bacteria could then elaborate potent virulence factors, such as proteinases, that activate oral keratinocytes and the kallikrein/kinin metabolic pathway in the oral mucosa, leading to plasma exudation and tissue injury . The purpose of this study was to begin to address this issue by determining whether Bacillus species contaminate commercially available chewing tobacco and, if so, whether they evoke plasma exudation from the intact oral mucosa. MATERIALS AND METHODS : Culture of chewing tobacco. | One box each of two popular chewing tobacco brands (moist snuff; Skoal Cherry and Skoal Spearmint [U.S. Tobacco Co., Richmond, Va.]) were purchased at a local grocery store in Chicago, Ill. One-half gram of chewing tobacco from each box was mixed with 50 ml of sterile, double-distilled water for 5 min. Then samples of the mixture were transferred onto blood, mannitol salt, and MacConkey agar plates by sterile cotton swabs and streaked for isolated colonies . The blood and mannitol salt agar plates were incubated under aerobic, anaerobic, and 10% CO2 conditions. MacConkey agar plates were incubated aerobically at 37C. After 48 h, the cultures were examined and a Gram stain of the isolated colonies was performed . After an additional 48 h of incubation 13 biochemical tests were conducted to identify the isolated gram-positive, spore-forming Bacillus species . A viable count was done to determine the number of organisms present in each sample. The most frequently isolated Bacillus species was inoculated into 7 ml of Trypticase soy broth and incubated for 72 h . Thereafter, the culture was centrifuged (15,000 x g), and the supernatant was filtered through a 0.22-mum-pore-size filter, snap-frozen in liquid nitrogen, and stored at -70C until used (see below). Determination of macromolecular efflux from the oral mucosa. | To determine the effects of the Bacillus supernatant on plasma exudation from the intact oral mucosa, we used the in situ microcirculation of the hamster cheek pouch as previously described in our laboratory and by other investigators (, , , -, ). Briefly, we used adult, male golden Syrian hamsters (mean body weight, 130 +- 1 g) that had been anesthetized with pentobarbital sodium and on which tracheostomics had been performed. The cheek pouch microcirculation was visualized with a fluorescence microscope (magnification, x40). Macromolecular leakage was determined by extravasation of fluorescein isothiocyanate-labeled dextran (FITC-dextran; molecular mass, 70 kDa), the intravascular tracer, which appeared as fluorescent spots or leaky sites around postcapillary venules. The number of leaky sites in three random microscopic fields was counted, averaged, and expressed as the number of leaky sites per 0.11 cm2 of cheek pouch, which corresponds to the area of one microscopic field (, , , -, ). With a spectrophotofluorometer, the concentration of FITC-dextran in the plasma and suffusate was determined based on a standard curve of FITC-dextran concentration versus percent emission. The clearance of FITC-dextran was determined by calculating the ratio of the concentration of FITC-dextran in suffusate (nanograms per milliliter) to that in plasma (milligrams per milliliter) and multiplying this ratio by the suffusate flow rate (2 ml/min). The experimental design has been used previously in our laboratory (, , -, ). After the buffer was suffused for 30 min (the equilibration period), FITC-dextran was injected intravenously and the number of leaky sites and the clearance of FITC-dextran were determined for 30 min. Next, the Bacillus supernatant (1:1,000 dilution in bicarbonate buffer, pH 7.4) was suffused over the cheek pouch for 20 min. The number of leaky sites was determined every minute for 20 min and at 5-min intervals for 60 min thereafter. The clearance of FITC-dextran was determined before the application of the supernatant and every 5 min after for 60 min. In another group of animals, 1 muM Hoe140 or 1 muM NPC 17647, both structurally distinct selective bradykinin B2 receptor antagonists (, , -), was suffused over the check pouch for 30 min before the check pouch was suffused with the Bacillus supernatant (1:1,000 dilution in bicarbonate buffer, pH 7.4) for 20 min. The observer of the cheek pouch microcirculation was unaware which hamster belonged in which treatment group. In preliminary studies, we determined that suffusion of bicarbonate buffer (vehicle) or of a 1:1,000 dilution of the bacteriological medium in bicarbonate buffer (pH 7.4) for the entire duration of the experiment or for 20 min, respectively, was associated with no visible leaky-site formation and no significant increase in the clearance of FITC-dextran (four animals; P > 0.5). The concentration of the Bacillus supernatant used in these experiments was based on preliminary experiments. The concentrations of Hoe140 and NPC 17647 used in these experiments were based on previous studies in our laboratory . Chemicals and drugs. | FITC-dextran was obtained from Sigma Chemical Co. Hoe140 and NPC 17647 were gifts from Aventis Pharmaceuticals and Nova Pharmaceutical Corporation, respectively. All drugs were dissolved in saline. Drugs were prepared fresh before each experiment and were diluted in saline to the desired concentrations. Data and statistical analyses. | When a test compound was suffused over the cheek pouch, we determined the maximal change in the number of leaky sites and the clearance of FITC-dextran and noted it as the response to that compound. Data are expressed as means +- standard errors of the means, except for body weight, which is expressed as mean +- standard deviation. Statistical analysis was performed using two-way analysis of variance and the Newman-Keuls test for multiple comparisons (StatView; SAS Institute Inc., Cary, N.C.) on a personal computer. A P of <0.05 was considered statistically significant. RESULTS : Culture of chewing tobacco. | Gram stain of cultures of both brands of chewing tobacco revealed large gram-positive, spore-forming bacilli in all samples (four separate experiments were conducted in duplicate). Biochemical analysis of these isolates revealed the presence of five distinct species of Bacillus, i.e., eight isolates of Bacillus megaterium, six of Bacillus pumilus, five of Bacillus brevis, two of Bacillus licheniformis, and one of Bacillus subtilis in the four experiments. The mean colony count was (5.0 +- 1.0) x 106 CFU/ml of supernatant (7.0 x 106 CFU/ml for experiments 1 and 2, 3.0 x 106 CFU/ml for experiment 3, and 2.1 x 106 CFU/ml for experiment 4). No other bacteria were isolated from either brand of chewing tobacco. Determination of macromolecular efflux from the oral mucosa. | Suffusion of saline (vehicle) alone for the entire duration of the experiment evoked no visible leaky-site formation and no significant increase from baseline in the clearance of FITC-dextran (four animals; P > 0.5). Repeated suffusions of B. megaterium supernatant (1:1,000 dilution in bicarbonate buffer, pH 7.4) for 20 min, each with 30-min suffusions of saline in between, were associated with significant and reproducible leaky-site formation [(5.4 +- 0.9)/0.11 cm2 and (6.0 +- 1.1)/0.11 cm2; each group contained four animals; P < 0.05 in comparison to results with saline, and P > 0.5 for within-group comparisons] and an increase in the clearance of FITC-dextran [(25.2 +- 7.5 ml)/min x 10-6 and (26.9 +- 5.9) ml/min x 10-6; each group contained four animals, P < 0.05 in comparison to results with saline, and P > 0.5 for within-group comparisons). Leaky-site formation was visible within 7 to 8 min after the start of suffusion and was maximal 4 to 5 min after the suffusion of B. megaterium supernatant was stopped. The number of leaky sites and the clearance of FITC-dextran returned to baseline 20 min after the suffusion of B. megaterium supernatant was stopped. Repeated suffusions of 1 muM Hoe140 or 1 muM NPC 17647 alone for 30 min, like suffusion of saline alone, evoked no visible leaky-site formation and no significant increase from baseline in the clearance of FITC-dextran (each group contained four animals; P > 0.5). Suffusion of Hoe140 and NPC 17647 (1 muM each) for 30 min significantly attenuated B. megaterium supernatant-induced leaky-site formation and the increase in clearance of FITC-dextran from the cheek pouch (each group contained four animals; P < 0.05 in comparison to results for B. megaterium supernatant alone). FIG. 1. | Effects of B. megaterium Effects of B. megaterium supernatant (1:1,000 dilution in bicarbonate buffer, pH 7.4; CT Bacillus), suffused for 20 min onto the intact hamster cheek pouch, on leaky-site formation (A) and the clearance of FITC-dextran (molecular mass, 70 kDa) (B) in the absence and presence of Hoe140 and NPC 17647 (1 muM each), two structurally distinct selective bradykinin B2 receptor antagonists. Data are given as means +- standard errors of the means and represent maximal values observed 5 min after the suffusion of the supernatant was stopped. Each group contained four animals. *, P of <0.05 in comparison to results with saline; +, P of <0.05 in comparison to results with B. megaterium supernatant alone. DISCUSSION : This study produced two new findings. First, we found that five distinct Bacillus species, predominantly B. megaterium and B. pumilus, contaminate, in relatively large numbers, two popular brands of chewing tobacco commercially available in the United States. Second, suffusion of a diluted B. megaterium supernatant onto an intact hamster cheek pouch, an established animal model for studying the mechanisms underlying the deleterious effects of smokeless tobacco in the oral mucosa , significantly increases macromolecular efflux from the cheek pouch. This response was mediated by local elaboration of bradykinin, because Hoe140 and NPC 17647, two structurally distinct selective bradykinin B2 receptor antagonists (, , -), significantly attenuated B. megaterium supernatant-induced responses. Taken together, these data indicate that Bacillus species contaminate commercially available chewing tobacco and elaborate a potent exogenous virulence factor(s) that activates the kallikrein/kinin metabolic pathway in the intact oral mucosa. This, in turn, evokes plasma exudation from postcapillary venules, which leads to interstitial edema and tissue dysfunction. Whether Bacillus species contaminate other brands of commercially available chewing tobacco remains to be determined. The mechanisms underlying bradykinin production by B. megaterium supernatant in the intact hamster cheek pouch were not elucidated. Nonetheless, the results of this study indicate that a potent bacterium-derived exogenous virulence factor(s) activates the kallikrein/kinin metabolic pathway in the oral mucosa to elaborate bradykinin, a potent edemagenic mediator . Previous studies have implicated proteinases released by oral bacteria, such as subtilisin and gingipain RgpA, in bradykinin-induced oral mucosa injury and inflammation in vivo (, , -). It remains to be determined whether Bacillus species contaminating commercially available chewing tobacco elaborate proteinases that directly activate the kallikrein/kinin metabolic pathway in the oral mucosa or whether a virulence factor(s) released by these bacteria stimulates oral keratinocytes, the first cells in the oral mucosa exposed to chewing tobacco and Bacillus species, to elaborate proteinases that activate this metabolic pathway (, , -). Likewise, the role of various chemical constituents of chewing tobacco in modulating the germination and expression of virulence factors by Bacillus species in the oral mucosa should be addressed . Notwithstanding the findings described above, the relevance of the present acute experimental study of the intact hamster cheek pouch to oral mucosa inflammation and injury observed in susceptible habitual chewing tobacco users is uncertain; the results of this study do not include the number of bacteria required to elaborate sufficient quantities of soluble virulence factors to injure the oral mucosa or whether Bacillus species spores and bacteria adhere to oral keratinocytes (, , , -, , ). However, habitual users consume relatively large quantities of chewing tobacco continually and spit it out rather than rinsing their mouths thoroughly after each application . This practice creates a local environment in the oral mucosa that is conducive to the germination of Bacillus species spores and to the elaboration of soluble virulence factors . Certain Bacillus species, including B. licheniformis and B. pumilus, isolated from chewing tobacco in the present study, have been shown to cause opportunistic infections and pulmonary inflammation in humans . The results of the present study extend these observations by showing that Bacillus species contaminating commercially available chewing tobacco elicit oral mucosa inflammation in experimental animals. Clearly, additional experimental and clinical studies are warranted to address these issues. In summary, we found that five distinct Bacillus species, predominantly B. megaterium and B. pumilus, contaminate in relatively large numbers chewing tobacco commercially available in the United States. Once germinating, these bacteria elaborate a potent exogenous virulence factor(s) that activates the kallikrein/kinin metabolic pathway in the intact oral mucosa, leading to plasma exudation and tissue dysfunction. Backmatter: PMID- 12204954 TI - Serodiagnosis of Mice Minute Virus and Mouse Parvovirus Infections in Mice by Enzyme-Linked Immunosorbent Assay with Baculovirus-Expressed Recombinant VP2 Proteins AB - Mice minute virus (MMV) and mouse parvovirus (MPV) type 1 are the two parvoviruses known to naturally infect laboratory mice and are among the most prevalent infectious agents found in contemporary laboratory mouse colonies. Serologic assays are commonly used to diagnose MMV and MPV infections in laboratory mice; however, highly accurate, high-throughput serologic assays for the detection of MMV- and MPV-infected mice are needed. To this end, the major capsid viral protein (VP2) genes of MMV and MPV were cloned and MMV recombinant VP2 (rVP2) and MPV rVP2 proteins were expressed by using a baculovirus system. MMV rVP2 and MPV rVP2 spontaneously formed virus-like particles that were morphologically similar to empty parvovirus capsids. These proteins were used as antigens in enzyme-linked immunosorbent assays (ELISAs) to detect anti-MMV or anti-MPV antibodies in the sera of infected mice. Sera from mice experimentally infected with MMV (n = 43) or MPV (n = 35) and sera from uninfected mice (n = 30) were used to evaluate the ELISAs. The MMV ELISA was 100% sensitive and 100% specific in detecting MMV-infected mice, and the MPV ELISA was 100% sensitive and 98.6% specific in detecting MPV-infected mice. Both assays outperformed a parvovirus ELISA that uses a recombinant nonstructural protein (NS1) of MMV as antigen. The MMV rVP2 and MPV rVP2 proteins provide a ready source of easily produced antigen, and the ELISAs developed provide highly accurate, high-throughput assays for the serodiagnosis of MMV and MPV infections in laboratory mice. Keywords: Introduction : Mice minute virus (MMV) (formerly minute virus of mice) and mouse parvovirus (MPV) type 1 are among the most prevalent infectious agents found in contemporary laboratory mouse colonies . Although the immunosuppressive strain of MMV (MMVi) can induce a potentially lethal renal hemorrhagic disease when experimentally inoculated into neonatal mice , clinical disease and histologic lesions have not been reported for mice naturally infected with MMV. Similarly, clinical disease and histologic lesions have not been observed in mice naturally or experimentally infected with MPV . Despite the absence of clinical disease and histopathology, murine parvoviruses can have significant deleterious effects on research due to their immunomodulatory effects both in vivo and in vitro (, , -). In addition, there is significant potential for MMV and MPV to be transmitted among animals within a facility due to the high degree of environmental stability of these agents . Therefore, identification of infected laboratory mice is critical to minimize the impact of murine parvovirus infections on research. Serologic evaluation for the presence of antiparvovirus antibodies has typically been used to diagnose MMV and MPV infections in mice , although the type of immunoassay and the source of diagnostic antigen vary significantly among rodent diagnostic laboratories. The most common methods used for the diagnosis parvovirus infections in mice include the enzyme-linked immunosorbent assay (ELISA), the indirect fluorescent-antibody assay (IFA), and the hemagglutination inhibition (HAI) assay, with the ELISA being preferred due to its high-throughput capability. More important than methodology is the choice of parvovirus antigen. MMV antigens have been generated from cell lines experimentally infected with the prototype strain of MMV (MMVp), which grows very well in vitro, yielding highly concentrated preparations of MMV antigen. This antigen works well for the detection of MMV infections in mice by the ELISA, IFA, and HAI assay formats , but MMVp antigen also displays some cross-reactivity with antibodies directed against MPV . This cross-reactivity is primarily the result of antibodies generated to the nonstructural (NS) proteins of MPV binding to the NS proteins of MMVp in the antigen preparation, reflecting the high degree of homology among the NS proteins of the rodent parvoviruses . Cell culture-propagated MPV can be used as antigen to detect anti-MPV antibodies . However, high-titer stocks of MPV are extremely difficult and expensive to obtain via cell culture propagation, and this has largely precluded the use of MPV antigens in ELISAs. For the past few years, a recombinant nonstructural protein of MMV (rNS1), which is highly conserved among rodent parvoviruses, has been used as antigen in serologic ELISAs to detect MMV and MPV infections . However, recent findings indicate that this assay lacks sensitivity in detecting infections, particularly MPV infections . Thus, serodiagnostic assays for MMV and MPV that use capsid proteins as antigens are needed. The goals of this study were to develop sensitive and specific recombinant antigen-based ELISAs to detect parvovirus-infected mice. Here we report on the cloning of the genes that encode the major capsid proteins (VP2) of MMV and MPV, the expression of recombinant VP2 (rVP2) of MMV and MPV with baculovirus vectors, and the use of these recombinant proteins as antigens in ELISAs to detect anti-MMV or anti-MPV antibodies in sera from experimentally infected mice. Our results show that the newly developed ELISAs are highly sensitive and specific for detecting parvovirus infections in mice. MATERIALS AND METHODS : Viruses. | MPV-1b was propagated in CTLL-2 murine cytotoxic T cells as described previously . MMVi was a kind gift from Dave Pintel of the University of Missouri. Construction of recombinant baculoviruses. | Standard molecular biology techniques were performed as described previously . The MMV VP2 and MPV VP2 genes were amplified by PCR. The MMV VP2 gene was amplified with platinum Pfx DNA polymerase (Invitrogen, San Diego, Calif.) according to the protocol of the manufacturer. Each 50-mul reaction mixture contained 1.25 U of polymerase, 290 ng of MMVp pML-n plasmid (an MMVp clone that was a kind gift from Dave Pintel, University of Missouri) as template, and primers MMV VP2 2664-2682 forward (5'-gcggatccgtcgacATGAGTGATGGCACCAGCC-3') and MMV VP2 4436-4413 reverse (5'-gcggtaccgcggccgcTTAGTAAGTATTTCTAGCAACAGG-3'). The uppercase letters of the primers correspond to MMV genome sequences, while the lowercase letters of the forward primer represent engineered restriction endonuclease sites for BamHI and SalI and the lowercase letters of the reverse primer represent engineered restriction endonuclease sites for KpnI and NotI. Thermocycling parameters consisted of an initial denaturation step (94C, 2 min), followed by 35 cycles of denaturation (94C, 15 s), annealing (55C, 30 s), and extension (68C, 2 min). The MPV VP2 gene was amplified by a nested PCR strategy. A region of the MPV genome was amplified with platinum Pfx DNA polymerase (Invitrogen) according to the protocol of the manufacturer with the primers MPV 2634-2653 forward (5'-GCACAGCAAAGAACTCAGAC-3') and MPV 4440-4419 reverse (5'-CAGAAAGAAAGAACATGGTTGG-3') and an annealing temperature of 55C. Each 50-mul reaction mixture contained 1.25 U of polymerase and 370 ng of template DNA from MPV-infected CTLL-2 murine cytotoxic T cells that was purified by the DNeasy (Qiagen Inc., Valencia, Calif.) protocol. An aliquot (0.5 mul) of the amplification product from the first PCR was used as template to amplify the MPV VP2 gene by using the Expand High Fidelity PCR system (Roche, Indianapolis, Ind.) and internal primers MPV VP2 2664-2682 forward (5'-gcgaattcggatccgcATGAGTGATGGCGCCGAGC-3') and MMV VP2 4436-4413 reverse (see above) with an annealing temperature of 55C. The uppercase letters of the primers correspond to regions of the MPV genome, and the lowercase letters of the forward primer represent engineered restriction endonuclease sites for EcoRI and BamHI. The amplified MMV VP2 and MPV VP2 genes were electrophoresed for 1 h at 80 V on a 1% SeaPlaque GTG agarose gel (FMC BioProducts, Rockland, Maine) that contained 0.5% ethidium bromide. Amplicons were visualized under UV light, excised from the gel, and purified with the Zymoclean Gel DNA recovery kit (Zymo Research, Orange, Calif.); and DNA concentrations were determined by measuring the A260 with a spectrophotometer (Gene Quant; Amersham Biosciences Corp., Piscataway, N.J.). The purified MMV and MPV VP2 PCR products were digested with BamHI and NotI. The MMV VP2 gene was directionally cloned, in frame, into the His-tagged fusion vector pFastBac HTb (Invitrogen) to create the plasmid MMV-VP2-pFastBac HTb. The MPV VP2 PCR product was directionally cloned, in frame, into the His-tagged fusion vector pFastBac HTc (Invitrogen) to create the plasmid MPV-VP2-pFastBac HTc. Plasmid MMV-VP2-pFastBac HTb or MPV-VP2-pFastBac HTc was used to construct independent recombinant baculovirus expression vectors MMV-VP2 Bac and MPV-VP2 Bac, respectively, by using the Bac-to-Bac Baculovirus expression system (Invitrogen). High-titer stocks of recombinant baculoviruses were generated in Spodoptera frugiperda (Sf-9) insect cells (Invitrogen), and viral titers were determined by plaque assay . Production of recombinant proteins and purification of ELISA antigens. | Standard insect cell and baculovirus propagation protocols were followed . Pilot experiments were performed to determine the kinetics for maximal production of each recombinant protein . To generate large quantities of MMV rVP2 or MPV rVP2 protein for use as antigens, 1-liter pendulum spinner flasks (Wheaton Science Products, Millville, N.J.) were seeded with 600 ml of High Five insect cells (Invitrogen) at a density of 9 x 105 cells per ml in complete insect cell medium (Insecta-FIVE medium; Mediatech, Inc., Herndon, Va.) supplemented with 1.36 mg of l-glutamine per ml and 5% fetal bovine serum. Cells were infected with recombinant baculovirus at a multiplicity of infection of 1.0 for MMV-VP2 Bac or 0.1 for MPV-VP2 Bac. At 72 h postinoculation the cells were pelleted by centrifugation at 500 x g for 15 min at 4C, resuspended in 50 ml of phosphate-buffered saline (PBS), and disrupted by six 30-s cycles of sonication (Sonifer Cell Disrupter 185; Branson Ultrasonics Corp., Danbury, Conn.) with a cup sonicator (Heat Systems Ultrasonics, Farmingdale, N.Y.) on ice. Insoluble material was pelleted by centrifugation at 20,000 x g for 1 h at 4C. The clarified supernatant of the MMV rVP2 preparation (semipurified protein) was harvested and used as antigen in the MMV rVP2 ELISA. Uninfected High Five insect cells cultivated and processed in the same manner were used as controls in the MMV ELISA. To further purify the MPV rVP2 for use as antigen in the MPV ELISA and to purify the MMV rVP2 and MPV rVP2 proteins for electron microscopy, clarified supernatants of vector-infected insect cells were loaded onto 10 to 40% cesium chloride gradients and centrifuged at 26,500 rpm in an SW48 rotor (Beckman Coulter, Inc., Fullerton, Calif.) for 16 h at 4C in an ultracentrifuge (Beckman Coulter, Inc.). Following centrifugation, the area of the gradient containing the rVP2 proteins was visualized as a blue band at a density of 1.31 g/cm3. This fraction was harvested, and the recovered proteins were dialyzed overnight at 4C against PBS through a 10,000-Da-cutoff membrane (Pierce Endogen, Rockford, Ill.). The concentrations of the recovered proteins were determined by a bicinchoninic acid procedure , and the proteins were stored at -80C until use. Western blot analysis. | Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to Immobilon-P transfer membranes (Millipore, Bedford, Mass.) . Molecular weight markers were included on each gel to facilitate the calculation of the sizes of the protein bands. Membranes were blocked with 5% nonfat dry milk in PBS (1.4 mM NaH2PO4, 10.8 mM Na2HPO4, 137.9 mM NaCl [pH 7.4]) and probed with diluted sera (1:200) from mice experimentally infected with MMV, mice naturally infected with MPV, or parvovirus-free mice. The membranes were incubated sequentially in solutions of biotinylated goat anti-mouse antibody diluted 1:1,000 (Jackson Immunoresearch Laboratories, Inc., West Grove, Pa.) and avidin DH (Vector Laboratories, Inc.) and were developed with peroxidase substrate solution (Kirkegaard & Perry, Gaithersburg, Md.). Parvovirus serology. | For the MMV ELISA, semipurified MMV rVP2 protein was diluted to 1.0 mug/ml in coating buffer (0.1 M NaHCO3 [pH 9.6]) and insect cell control protein was diluted to 0.5 mug/ml in coating buffer; 200 mul of each was coated onto separate wells of an Immulon 2 96-well plate (Dynex Technologies, Inc., Chantilly, Va). For the MPV ELISA, CsCl-purified MPV rVP2 protein was diluted to 0.1 mug/ml in coating buffer, and 200 mul was loaded into test wells of an Immulon 2 96-well plate. Control wells were uncoated. Proteins were allowed to bind to the plates for 48 h at 4C. The wells were emptied and washed three times in PBS with 0.05% Tween 20 (PBS-Tween). Blocking buffer (PBS with 0.5% nonfat dry milk and 0.05% Tween 20) (300 mul) was added to each well, the mixture was incubated for 30 min at room temperature, and the buffer was discarded; 200 mul of serum diluted 1:100 in blocking buffer was added to experimental and control wells. After incubation at 37C for 2 h, the plates were washed three times in PBS-Tween. Two hundred microliters of alkaline phosphatase-labeled secondary antibody [F(ab')2 fragment goat anti-mouse immunoglobulin G (heavy and light chains); Jackson Immunoresearch Laboratories, Inc.]) diluted 1:7,000 in PBS-Tween was added to each well, and the plate was incubated at 37C for 2 h. The plates were washed five times in PBS-Tween. One hundred microliters of phosphatase substrate solution (1 mg of Sigma 104 phosphatase substrate [Sigma Chemical Company, St. Louis, Mo.] per 1 ml of substrate buffer [2 mM MgCl2, 27.5 mM NaCO3, 22.5 mM NaHCO3 {pH 9.8}]) was added to each well, and the plate was incubated in the dark at room temperature for 45 min. The optical density (OD) of each well was measured (Kinetic Reader EL312E; Bio-Tek Instruments, Inc., Winooski, Vt.) and automatically calculated as the OD at 405 nm (OD405) minus the OD490 to correct for optical interference. For each serum sample, the final OD was calculated as the OD of the test well minus the OD of the control well. The cutoff OD value for each ELISA was set at 3 standard deviations above the mean for 30 serum samples from mice negative for parvovirus infections. The MMV rVP2 ELISA and the MPV rVP2 ELISA were compared to a parvovirus ELISA that uses a recombinant nonstructural protein (rNS1) of MMV as antigen. The rNS1 ELISA was performed as described previously . Performance of ELISAs. | The following were calculated for each ELISA: sensitivity = TP/(TP + FN) x 100, specificity = TN/(TN + FP) x 100, positive predictive value = TP/(TP + FP) x 100, negative predictive value = TN/(TN + FN) x 100, and accuracy = (TP + TN)/(TP + FP + TN + FN), where TP is the number of samples with true-positive results, TN is the number of samples with true-negative results, FP is the number of samples with false-positive results, and FN is the number of samples with false-negative results. For each ELISA, a Kruskal-Wallis one-way analysis of variance on ranks was performed to test for differences between groups of mice. Statistical significance was set at a P value of <=0.01. Mice, experimental infections, and sera. | Four-, 8-, and 12-week-old male Hsd:ICR(CD-1) mice and 12-week-old male BALB/cAnNHsd, C3H/HeNHsd, C57BL/6NHsd, and DBA/2NHsd mice were obtained (Harlan Sprague-Dawley, Inc., Indianapolis, Ind.). The mice were separated into experimental and control groups of three to five animals each according to strain, age, and dose of viral inoculum. Mice were inoculated with 5 x 104 to 1 x 106 50% tissue culture infective doses of MPV-1b diluted in TE (50 mM Tris base, 10 mM EDTA [pH 8.7]) or with TE by intragastric gavage under isofluorane anesthesia. At 4 weeks postinoculation, the mice were euthanized with an inhaled overdose of carbon dioxide, blood was collected by cardiocentesis, and the sera were collected. For the MMV experimental infections, pregnant female C3H/HeNHsd mice were obtained (Harlan Sprague-Dawley), and neonatal male and female pups were inoculated oronasally with 5 x 104 50% tissue culture infective doses of MMVi. Mice were euthanized with an inhaled overdose of carbon dioxide at 8 weeks postinoculation, blood was collected by cardiocentesis, and the sera were collected. Sera from all mice were diluted 1:5 in PBS, heat inactivated at 56C for 30 min, and stored at -20C until use. The source colonies for all mice were documented to be free of known mouse pathogens, including MPV and MMV. Each experimental and control group was housed separately in microisolation cages, and all animal manipulations were performed in a class II biological safety cabinet by standard microisolation technique. All animal studies were approved by the University of Missouri or the University of Arizona Institutional Animal Care and Use Committee. Sera from MPV-inoculated mice that were positive for anti-MPV antibodies by an MPV IFA, sera from MMV-inoculated mice that were positive for anti-MMV antibodies by an MMV IFA, and sera from uninfected mice were used in the ELISAs. The MMV and MPV IFAs were performed as described previously . In Western blot experiments, sera from MMV-inoculated mice that were positive for anti-MMV antibodies by IFA were used as MMV-positive sera, sera from FVB/NCr mice that were naturally infected with MPV were used as MPV-positive sera, and sera from uninfected mice were used as negative sera. The mice naturally infected with MPV were positive for anti-MPV antibodies by MPV IFA, positive for MPV DNA by MPV PCR of mesenteric lymph nodes, negative for anti-MMV antibodies by MMV IFA, and negative for MMV DNA by PCR of mesenteric lymph nodes. Electron microscopy. | The MPV rVP2 and MMV rVP2 proteins were expressed and purified by CsCl gradient centrifugation as described above, with the exception that the recovered fractions were diluted in PBS and centrifuged at 41,000 rpm for 5 h in an SW 41 rotor (Beckman) at 4C to pellet the proteins. The pellets of each protein were resuspended in 1% ammonium acetate. A small droplet of each sample in 4% phosphotungstate was deposited on a carbon-coated grid and examined with a Hitachi H-6700 transmission electron microscope. RESULTS : MMV rVP2 and MPV rVP2 self-assemble into VLPs. | The VP2 genes of MMV and MPV were cloned to create the baculovirus vectors MMV-rVP2 Bac and MPV-rVP2 Bac. These vectors were used to express the recombinant capsid proteins MMV rVP2 and MPV rVP2 in insect cells. It is known that many parvovirus recombinant VP2 proteins, including MMV rVP2, assemble into virus-like particles (VLPs) when expressed in baculovirus expression systems . To determine if baculovirus-expressed MMV rVP2 or MPV rVP2 proteins self-assembled into VLPs, we subjected clarified supernatants of insect cells infected with the MMV-rVP2 Bac vector or the MPV-rVP2 Bac vector to CsCl gradient ultracentrifugation. Each gradient yielded a visible blue band at a density of 1.31 g/cm3 of CsCl, which is slightly less dense than previously reported densities of MMV VLPs and empty parvovirus capsids (1.32 g/cm3 of CsCl) . Transmission electron microscopic examination of negative stained preparations of the collected bands revealed the presence of VLPs morphologically similar to empty parvovirus capsids . The MMV VLPs and MPV VLPs were isometric and approximately 20 nm in diameter. Western blots of MMV rVP2 and MPV rVP2. | To determine if MMV rVP2 and MPV rVP2 reacted with natural antibodies to MMV or MPV and if the recombinant proteins shared cross-reactive epitopes, CsCl-purified MMV rVP2 and MPV rVP2 were subjected to SDS-PAGE and Western blot analysis. The Western blots were probed with MMV-positive sera, MPV-positive sera, or parvovirus-negative sera. MMV rVP2 was visualized as a 68-kDa immunoreactive protein band, and MPV rVP2 was visualized as a 69-kDa immunoreactive protein band . The molecular masses of the recombinant proteins represent the molecular mass of the VP2 protein plus the molecular mass of the leader peptide of the fusion protein. The MMV rVP2 protein reacted strongly with four of five MMV-positive serum samples, reacted weakly with one of five MMV-positive serum samples, and displayed cross-reactivity with two of five MPV-positive serum samples. The MPV rVP2 protein reacted strongly with all five MPV-positive serum samples and cross-reacted weakly with two of five MMV-positive serum samples. The MMV rVP2 and MPV rVP2 proteins did not react with sera from uninfected mice. In an ELISA format, MMV-positive sera reacted only with MMV rVP2, MPV-positive sera reacted only with MPV rVP2, and parvovirus-negative sera did not react with either antigen. MMV- and MPV-specific ELISAs. | The baculovirus-expressed MMV rVP2 and MPV rVP2 proteins yielded large amounts of readily produced antigens to develop the MMV- and MPV-specific ELISAs. From 600-ml suspension cultures of vector-infected insect cells, 250 mg of semipurified MMV rVP2 antigen or 1.2 mg of CsCl gradient-purified MPV rVP2 was produced. The semipurified preparation of MMV rVP2 was used as antigen for the MMV ELISA, and the CsCl-purified MPV rVP2 was used as antigen for the MPV ELISA. These antigen preparations were chosen on the basis of preliminary ELISA experiments that showed the marked immunoreactivity of semipurified MMV antigen with MMV-positive sera and the poor immunoreactivity of semipurified MPV antigen with MPV-positive sera. Strong immunoreactivity of MPV rVP2 in an ELISA was obtained only when CsCl-purified MPV rVP2 was used as antigen (data not shown). The baseline cutoff OD405 values used to designate positive sera were calculated to be 0.160 for the MMV ELISA and 0.093 for the MPV ELISA. To compare the performances of the MMV ELISA and the MPV ELISA to that of a previously described generic parvovirus ELISA that detects anti-NS1 antibodies, sera from mice experimentally infected with MMV or MPV and sera from uninfected mice were tested by the MMV ELISA, the MPV ELISA, and the rNS1 ELISA . The MMV ELISA was 100% sensitive and 100% specific in detecting MMV-infected mice and had a 100% positive predictive value, a 100% negative predictive value, and an accuracy of 100%. The MPV ELISA was 100% sensitive and 98.6% specific in detecting MPV-infected mice and had a 97.2% positive predictive value, a 100% negative predictive value, and an accuracy of 99%. The rNS1 ELISA was 21% sensitive and 100% specific in detecting MPV-infected mice, was 0% sensitive and 100% specific in detecting MMV-infected mice, and had a 100% positive predictive value, a 28% negative predictive value, and an accuracy of 76% for detecting mice infected with MMV or MPV. These results demonstrate that the MMV ELISA and the MPV ELISA are sensitive and specific and markedly outperform the rNS1 ELISA for detecting antibodies to MMV or MPV. The performances of the MMV ELISA and the MPV ELISA were further characterized by examining the intensities of the OD values generated when sera from MMV-infected mice, MPV-infected mice, and parvovirus-negative mice were evaluated . For the MMV ELISA, the median absorbance value for sera from MMV-infected mice was statistically higher than the median absorbance values for sera from mice infected with MPV or sera from parvovirus-free mice (P < 0.01). For the MPV ELISA, the median absorbance value for sera from MPV-infected mice was significantly higher than the median absorbance values for sera from mice infected with MMV or sera from parvovirus-free mice (P < 0.01). Prevalence of MMV and MPV infections in research mice. | In addition to evaluating sera from experimentally infected mice, we also evaluated all mouse serum samples that were submitted for routine serologic evaluation over a 4-week-period to the University of Missouri Research Animal Diagnostic Laboratory (Columbia, Mo.). These sera were from 83 research institutions throughout the United States and Canada and were obtained from a variety of immunocompetent mouse strains and stocks. Of 2,473 mouse serum samples tested, 205 serum samples (8.3%) were positive for MPV by ELISA, 38 serum samples (1.5%) were positive for MMV by ELISA, and 32 serum samples (1.3%) were positive for both MMV and MPV by ELISA. Seventeen institutions had sera positive for MPV, three institutions had sera positive for MMV, and five institutions had sera positive for both agents. A group of 22 mice that were serologically positive for MMV and MPV by ELISA were also evaluated for the presence of MMV and MPV DNA by PCR analysis of the mesenteric lymph nodes. Of these 22 mice, 19 were positive for MMV and MPV by PCR. Collectively, these data suggest that MPV and MMV infections are present in research mouse colonies, MPV infections are more prevalent than MMV infections, and dual infections with MPV and MMV occur in some mice. FIG. 1. | Transmission electron micrographs of negative stained preparations of MPV VLPs produced from baculovirus-expressed MPV rVP2 protein (A) and MMV VLPs produced from baculovirus-expressed MMV rVP2 protein (B). Transmission electron micrographs of negative stained preparations of MPV VLPs produced from baculovirus-expressed MPV rVP2 protein (A) and MMV VLPs produced from baculovirus-expressed MMV rVP2 protein (B). The VLPs were purified by CsCl gradient ultracentrifugation. FIG. 2. | Western blot analysis of MPV rVP2 VLPs or MMV rVP2 VLPs separated by SDS-PAGE. Western blot analysis of MPV rVP2 VLPs or MMV rVP2 VLPs separated by SDS-PAGE. The blot was probed with sera from MPV-infected mice (lanes 1 to 5), MMV-infected mice (lanes 6 to 10), and uninfected mice (lanes 11 and 12). Each lane contains 5 mug of protein; test sera were run at a 1:200 dilution. The positions of the MPV rVP2 and the MMV rVP2 are indicated by arrows on the left and right, respectively. The positions of the molecular mass standards are indicated on the left. FIG. 3. | Serum antibody responses of mice experimentally infected with MMV or MPV or of uninfected mice tested by MMV ELISA (A) or MPV ELISA (B). Serum antibody responses of mice experimentally infected with MMV or MPV or of uninfected mice tested by MMV ELISA (A) or MPV ELISA (B). The datum points are presented as box plots; for each experimental group the lower boundary of the box indicates the 25th percentile, the upper boundary of the box indicates the 75th percentile, and the line within the box marks the median value. The line below the box marks the 10th percentile, the line above the box marks the 90th percentile, and datum points outside this range are indicated (*). *, P < 0.01 compared to each other experimental group. TABLE 1 | Comparison of serologic test results for uninfected mice and mice experimentally infected with MPV or MMV DISCUSSION : In this report, we describe the development of ELISAs that detect anti-MMV or anti-MPV antibodies in the sera of parvovirus-infected mice. The MMV ELISA and the MPV ELISA displayed high sensitivities and high specificities for the detection of species-specific parvovirus antibodies in MMV- and MPV-infected mice, and each ELISA had high positive and negative predictive values. These assays are ideally suited for use as primary serologic tests to screen mouse colonies for MMV and MPV infections because the assays were highly accurate and because ELISAs are readily automated, making it possible to test large numbers of samples quickly and at relatively little expense. In addition, large quantities of the MMV rVP2 and the MPV rVP2 antigens used in these assays were readily produced with the baculovirus expression system, providing ample antigens for high-volume testing. This is particularly significant for MPV because it is difficult and expensive to produce large quantities of MPV-specific ELISA antigens in vitro from MPV-infected cells. The MMV and MPV ELISAs performed with high sensitivities in detecting MMV and MPV infections in mice, respectively, and these two assays were much more sensitive in detecting parvovirus infections in mice than a generic parvovirus ELISA that uses a recombinant nonstructural protein (rNS1) of MMV as antigen. This is consistent with previous work that shows that mice experimentally inoculated with MPV, especially mice infected after 12 weeks of age, develop anticapsid antibodies more so than anti-NS1 antibodies . The VP2 protein is the major capsid protein of the parvovirus virion and can be presented to the immune system upon exposure to the virus or during viral replication. However, nonstructural proteins, including NS1, are not present in the parvovirus virion and are produced more transiently upon infection. Thus, an undetectable anti-NS1 humoral immune response could result from a nonproductive parvovirus infection or from a productive infection that produces an insufficient amount of NS1 to induce an immune response . The MMV and MPV ELISAs also performed with high specificities in detecting MMV and MPV infections in mice, respectively. Previously described ELISAs that use whole-virus preparations of MMV or rNS1 proteins as antigens are not specific for the identification of MMV- and MPV-infected mice and require the use of secondary testing with MMV- or MPV-specific HAI assays to elucidate the true infection status of the animal. Thus, the high specificities of the newly developed MMV and MPV ELISAs eliminate the need for secondary testing by the HAI assay. An interesting feature of the baculovirus-expressed MMV and MPV rVP2 proteins was that they self-assembled into VLPs that were morphologically similar to empty parvovirus capsids. Self-assembly of baculovirus-expressed recombinant parvovirus VP2 proteins has been reported for many parvoviruses including MMV ; however, this is the first report of self-assembly of MPV rVP2 into VLPs. MMV rVP2 and MPV rVP2 were expressed as fusion proteins that contained an N-terminal polyhistidine tag with the intent of purifying these proteins by metal-affinity chromatography. In preliminary experiments, MMV rVP2 and MPV rVP2 could not be successfully purified by affinity chromatography (data not shown). However, this proved to be unnecessary because the semipurified preparation of MMV rVP2 was highly immunoreactive and therefore required no further purification. While the semipurified preparation of MPV was weakly immunoreactive, highly immunoreactive MPV rVP2 in the form of MPV VLPs was easily purified by CsCl ultracentrifugation. The N-terminal polyhistidine tag and leader sequences added an estimated 3.6 kDa to each rVP2. While the leader peptides did not alter the ability of the rVP2 proteins to assemble into VLPs, they may have caused a slight decrease in the densities of the VLPs (1.31 g/cm3 of CsCl) compared to previously reported densities of MMV VLPs and empty parvovirus capsids (1.32 g/cm3 of CsCl) . The formation of VLPs by the MMV and MPV rVP2 proteins may have contributed to the high sensitivities of the MMV and MPV ELISAs. Interestingly, the MMV rVP2 protein that was subjected to Western blotting reacted strongly with only four of five MMV-positive serum samples. The one MMV-positive serum sample that reacted weakly with MMV rVP2 by Western blotting was strongly reactive with MMV rVP2 by ELISA. It is possible that conformational epitopes present on the exterior of the MMV VLPs were not present when MMV rVP2 was disrupted or denatured for Western blot analysis. Our data also suggest that formation of MMV and MPV VLPs likely enhanced the specificities of the MMV and MPV ELISAs over those which would be attained with unassembled MMV rVP2 or MPV rVP2 proteins as antigens. Sera from two MPV-infected mice strongly cross-reacted with MMV rVP2 proteins on Western blots. These sera did not cross-react with MMV rVP2 in an ELISA format. In addition, two serum samples from MMV-infected mice displayed weak cross-reactivity with MPV rVP2. These two serum samples were not cross-reactive with the MPV rVP2 protein in an ELISA format. A possible explanation for these observations is that in the ELISA format, the MMV rVP2 and MPV rVP2 VLPs behave antigenically like whole virions and anticapsid antibodies are able to associate only with antigenic epitopes present on the exterior of the VLPs. However, when the VLPs are subjected to SDS-PAGE and Western blotting, the VLPs are disrupted and the proteins are denatured, possibly exposing cross-reactive epitopes. The MMV and MPV ELISAs were used to survey sera obtained from numerous mouse colonies from geographically disparate locations in North America. The prevalence of sera positive for a parvovirus was nearly 10%, with the majority of these sera being positive for MPV. This is consistent with a recent survey that showed that in the United States the prevalence of parvovirus infections among laboratory mouse colonies is high . Given the potential for parvovirus infections to alter or invalidate studies that use infected mice (, , -), there is a need for highly accurate, high-throughput assays to detect parvovirus infections in mice. Baculovirus-expressed MMV rVP2 and MPV rVP2 proved to be excellent antigens for the development of such assays. The MMV and MPV ELISAs performed with high sensitivities and specificities and markedly outperformed another widely used parvovirus ELISA that detects antibodies to NS1 proteins in infected mice. The two newly developed serologic assays offer rapid, inexpensive, and accurate methods for the screening of research mouse colonies for MMV and MPV infections. Backmatter: PMID- 12204972 TI - CD4+ CD25+ T-Cell Production in Healthy Humans and in Patients with Thymic Hypoplasia AB - Regulatory T cells are found primarily in the CD4+ CD25+ fraction of T cells and play an important role in the prevention of autoimmunity. We examined CD4+ CD25+ T cells in 33 healthy children and adults and compared them to a population with an inherited form of thymic hypoplasia and a predisposition to autoimmune disease. Absolute numbers of CD4+ CD25+ T cells were markedly higher in healthy infants than in infants with chromosome 22q11.2 deletion syndrome. Keywords: Introduction : Regulatory T cells constitute an important contributor to the state of tolerance . Little is known about their function and production in humans, and nothing is known about their development in humans . Murine studies first defined these cells as a component of CD4+ CD25+ (CD4 CD25) T cells. Several important features of these murine cells have been described. They exit the thymus from day 4 of life onwards, and they appear to require a T-cell receptor with high affinity . They are antigen specific in their activation but are non-antigen specific as effectors. As effectors, they are naturally nonproliferative, and their function is generally detected by their ability to inhibit proliferation of other cells or other effector functions, such as cytolytic activity. Although their exact mechanism of action is controversial, there is agreement that cell contact is required. These cells have been demonstrated to be responsible for the prevention of organ-specific autoimmunity in several murine models. Mice which have undergone a thymectomy between days 1 and 3 of life have a delay in the appearance of CD25+ T cells. These mice develop organ-specific autoimmune disease; the specific disease is dependent on the strain of mice used . Infusion of CD25+ T cells can abolish the organ-specific autoimmune disease. In thymectomy models, the ratio of CD25+ to CD25- T cells is normal. The CD25+ T cells appear late and in lower numbers but retain their normal proportion in the T-cell compartment. This has suggested that adequate quantitative production of regulatory T cells early in life may be a critical part of the development of tolerance. Human cells with regulatory properties express very high levels of CD25 and have suppressive effects on effector functions similar to that seen in mice . They are present in adults at levels similar to those seen in mice, i.e., approximately 6 to 10% of CD4 T cells. The studies described above documented the important role of the thymus in the production of CD4 CD25 T cells and the role of CD4 CD25 T cells in the prevention of autoimmunity. In this study, we examined a population of children who have thymic hypoplasia as a result of a heterozygous deletion of chromosome 22q11.2. This deletion results in a phenotype which has also been called DiGeorge syndrome, velocardiofacial syndrome, or conotruncal anomaly face syndrome. Approximately 80% of patients with the deletion will have an immunodeficiency as a consequence of the thymic hypoplasia . The immunodeficiency is typically a pure defect in T-cell production, and patients have on average 50 to 70% of the normal numbers of T cells . Autoimmune disease develops in approximately 10% of patients, although the pathophysiologic mechanism for this is not understood. This study examined the developmental appearance of CD4 CD25 T cells in the peripheral blood of 33 children and adults with chromosome 22q11.2 deletion and in 33 healthy age-matched controls in an effort to determine whether thymic hypoplasia could be associated with diminished numbers of this important T-cell subset, as is seen in mice with neonatal thymectomies. Unselected patients with hemizygous deletions of chromosome 22q11.2 as detected by the Oncor/Vysis N25 probe were included as "patients" in this analysis. Control samples were selected from clinics performing well-child care or allergy evaluations (children with negative RAST tests). Institutional review board approval was obtained for this study, and consent was obtained from each participating patient or parent of the patient. Three-color flow cytometry was performed to define CD3, CD4, and CD25. Only CD25hi CD4 T cells were included in the CD25-positive gate, in accordance with what is known regarding the expression of CD25 on human regulatory T cells . The fraction of CD25 T cells within the CD3 population and the frequency of CD25 T cells within the CD4 population did not vary with age and showed no difference between controls and patients. In both patients and controls, the CD25 fraction of CD4 T cells was typically 6 to 15%. In both patients and controls, the CD25 fraction of CD3 T cells was typically 2 to 10%. Differences were readily apparent between patients and controls when the CD4 CD25 proportion of total lymphocytes was examined and the CD4 CD25 absolute number was determined . There is a distinct age-related decline in the CD4 CD25 fraction and number of cells from birth through approximately 36 months of age in the controls. The differences between the CD4 CD25 T-cell numbers in children less than 3 years of age and those over 9 years of age are significant, with P being 0.001. In contrast, the fraction and number of CD4 CD25 T cells were similar throughout all ages examined in the chromosome 22q11.2 deletion population. As a result, the fraction and the absolute number are markedly lower in the patients than controls in infancy. The mean CD4 CD25 T-cell count in patients under 3 years of age was 7 cells/mm3, while it was 75 cells/mm3 in the controls (P = 0.001). The mean CD4 CD25 T-cell count in patients 3 to 9 years of age was 3 cells/mm3, while it was 14 cells/mm3 in controls (P = 0.005). The mean CD4 CD25 T-cell count in patients over the age of 9 years was 3 cells/mm3, while it was 10 cells/mm3 in controls (P = 0.002). Six patients had autoimmune disease. The mean CD4 CD25 T-cell count in these six patients was no different than the mean in the non-autoimmune disease patients. FIG. 1. | (Top) Fractions of CD4 CD25 T cells within the total lymphocyte gate; (bottom) absolute numbers of CD4 CD25 T cells. (Top) Fractions of CD4 CD25 T cells within the total lymphocyte gate; (bottom) absolute numbers of CD4 CD25 T cells. Gray diamonds represent patients with autoimmune disease. m, months. This is the first study to examine peripheral blood CD4 CD25 T cells in human children. We have shown that newborns generally have high numbers that decline rapidly over the first 2 to 3 years of life. This is dependent on T-cell production, because the fraction of CD4 CD25 cells within the T-cell subset remains relatively constant throughout childhood. Patients with chromosome 22q11.2 deletion with developmental thymic hypoplasia have markedly fewer CD4 CD25 T cells in infancy. The significance of these findings is twofold. Patients with chromosome 22q11.2 deletion syndrome have a relatively pure quantitative defect in T-cell production. Thus, this study suggests that regulation of CD4 CD25 T-cell production early in life in humans is directly related to thymic capacity. CD4 CD25 T cells are produced in similar proportions with other T cells postnatally. Secondly, this study suggests that one contributor to the increased prevalence of autoimmune disease could be diminished CD4 CD25 T cells. This is comparable to what was seen in the murine neonatal thymectomy model. The CD4 CD25 T cells appear late, and the absolute numbers remain diminished. In this murine model, approximately 20% of the mice developed autoimmune gastritis by 1.5 to 3 months of age after neonatal thymectomy . Extrapolating from the murine model, our finding of markedly diminished CD4 CD25 T-cell numbers early in life suggests that this phenomenon could play a role in the predisposition to autoimmune disease in patients with chromosome 22q11.2 deletion syndrome. Other factors are required to define the type of autoimmune disease and to further determine susceptibility. Backmatter: PMID- 12204961 TI - Immunoglobulin G Antibody Response to Infection with Coccoid Forms of Helicobacter pylori AB - An increasing number of studies support a potential role for coccoid forms in Helicobacter pylori infection. Evidence for this was obtained through scanning microscopy, genetic analysis for virulence traits, examination of the presence and activity of key enzymes, and other methods. We studied the serum immunoglobulin G responses to coccoid H. pylori forms by enzyme-linked immunosorbent assay (ELISA) and immunoblotting and compared them with those of bacillary cells. Sera from a total of 295 infected individuals were studied; these included sera from 100 patients with duodenal ulcers, 98 patients with nonulcer dyspepsia, 11 patients with gastroduodenal cancer, and 86 asymptomatic individuals. Initially, we characterized and selected coccoid and bacillary antigenic preparations by one-dimensional (1-D) and 2-D gel electrophoresis and immunoblotting. Data showed that coccoid and bacillary preparations with comparable protein contents have similar patterns in 1-D and 2-D electrophoresis gels and antigenic recognition at blotting. These results revealed that coccoid and spiral antigens in ELISA can equally recognize specific antibodies to H. pylori in sera from infected individuals. The analysis of the spiral and coccoid preparations by Western blotting showed no major differences in antigen recognition. No specific bands or profiles associated with a single gastric condition were identified. Keywords: Introduction : Helicobacter pylori spiral-coccoid dimorphism is observed both in vivo and in vitro. It is generally accepted that the coccoid forms arise as a response to stress conditions, e.g., in vitro aerobiosis , temperature changes , extended incubation , and in vivo antibiotic treatment . Since they were first described, coccoid H. pylori forms were considered an irreversible phase that leads to cell death . Present knowledge suggests that coccoid cells are not dead but actually dormant . Coccoid forms may therefore play a role in the survival, and eventually in the transmission, of this microorganism. A number of reports suggest that coccoid forms maintain cell structures, metabolism, DNA indemnity , and gene expression . There are also reports indicating that H. pylori cells are able to survive for prolonged periods in the environment, especially in water and under conditions of starvation . This would not be surprising if we took into account that coccoid forms are biologically important for other pathogenic bacteria, such as Vibrio vulnificus or Campylobacter jejuni . This study of coccoid forms may help us to better understand the natural history of H. pylori infection. H. pylori infection induces a strong local inflammatory response which often is insufficient to eradicate the pathogen, and this failure may be responsible for the chronicity that these gastric diseases often demonstrate. It is not fully understood how the immune system is involved in clinical outcomes. One point upon which investigators agree is that the presence of specific antibodies can be used as an epidemiological indicator of infection . Some studies suggest that noninvasive serologic tests may be of value to confirm treatment success . Although many studies have focused on the impact of bacillary H. pylori cells on immune status , there is no information on the potential role of coccoid forms. The aim of this work was to study the immunoglobulin G (IgG) immune response of colonized individuals against coccoid H. pylori forms and compare it with that elicited by its spiral counterpart. MATERIALS AND METHODS : Strains. | We studied 21 strains of H. pylori isolated in our laboratory from gastric biopsy samples of Chilean adults. The isolates were confirmed through microscopy, culture, and rapid urease testing. Antigen preparation. | All strains were grown under microaerophilic conditions at 37C on chocolate agar and a Skirrow antibiotic pool. Spiral cells were collected after 3 days in phosphate-buffered saline (PBS). The coccoid cells were harvested after 30 days at room temperature under aerobic conditions. Coccoid morphology was confirmed by Gram stain (100 fields) and by the strains' inability to grow in appropriate conditions. The coccoid and bacillary antigens were prepared by the acid glycine extraction method , standardized in their protein content (Bio-Rad Labs, Hercules, Calif.), and maintained frozen (-20C) until analysis. SDS-PAGE antigen characterization. | H. pylori proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), with 4 and 7% stacking and running gels, respectively. The bands were visualized with silver stain, and the gels were analyzed by Quantity One software (Bio-Rad). 2-D electrophoretic antigen characterization. | The H. pylori preparations were first separated by isoelectric focusing according to a procedure described by Celis et al. . Antigens (200 mug/capillary) were incubated at room temperature with 40 mul of lysis solution (9.8 M urea, 2% NP-40, 2% ampholyte 3/10, 100 mM dithiothreitol) for 15 min. Preparations were loaded into the capillaries and covered with 20 mul of overlay solution (8 M urea, 1% ampholyte 7/9, 5% NP-40, 100 mM dithiothreitol). Gels were run at 200 V for 2 h, 500 V for 2 h, and 800 V for 16 h in two-dimensional (2-D) electrophoresis equipment (Protean II; Bio-Rad). After an electrophoretic run under similar conditions, the protein spots were visualized by silver staining and analyzed by 2-D Bio-Rad software. Western blot antigen analysis. | The H. pylori coccoid and bacillary antigens were evaluated by Western blot analysis . In brief, strips were blocked with skimmed milk, confronted with 1:150 serum dilutions, and maintained overnight at room temperature. Membranes were then incubated with an anti-human IgG alkaline phosphatase conjugate (Sigma). Reaction was revealed with 5-bromo-4-chloroindolylphosphate, nitroblue tetrazolium (GIBCO), and MgCl2 (Merck). The molecular masses of the proteins were calculated by interpolating the proteins in a curve constructed with reference markers (Bio-Rad). Serum panel. | The sera obtained previously were from 295 colonized individuals: 100 duodenal ulcer (DU) patients, 98 nonulcer dyspepsia (NUD) patients, 11 gastric cancer patients, and 86 asymptomatic individuals. Colonization of asymptomatic subjects was confirmed by the presence of IgG antibodies to H. pylori. ELISA for IgG antibodies to spiral and coccoid preparations. | Enzyme-linked immunosorbent assay (ELISA) was previously standardized in our laboratory ; the conditions were optimized by check board analysis. In brief, flat-bottom plates (Maxisorp; Nunc, Roskilde, Denmark) were covered with either bacillary or coccoid H. pylori surface antigen preparation (0.125 mug/well). After blocking with PBS containing 10% skim milk, serum diluted 1:600 was added, the plates were incubated for 2 h at 37C, and their reactivity was revealed with anti-human IgG alkaline phosphatase conjugate diluted 1:1,000 (Sigma). Wells were washed, and the substrate was added (p-nitrophenyl phosphate; Sigma) and maintained for 30 min at 37C. The reaction was stopped with 3 M NaOH. Results were expressed as optical density (OD) at 405 nm. All serum samples were evaluated in duplicate. Adsorption assay. | Four sera with high ELISA levels were tested to evaluate the presence of common antigens between both preparations. Serum samples were assayed before and after adsorption with bacillary and coccoid antigens separately. Briefly, a total of 25 mug of antigen was incubated with 400 mul of diluted sera (in PBS diluted 1/20) under gentle agitation for 1 h at 37C and overnight at 4C. Adsorbed sera were centrifuged at 6,000 x g, and the supernatant was saved and kept frozen at -20C until ELISA was performed. Statistics. | Comparisons between values obtained by ELISA based on coccoid and bacillary forms were performed through the Student's t test (STATISTICA software). RESULTS : H. pylori antigens. | The coccoid and spiral preparations revealed numerous electrophoretic bands ranging from 10 to 29 and 11 to 26 bands, respectively . The molecular weights were calculated to be between 6.5 and 169 for the bacillary form and 6.5 and 200 for the coccoid form. Based on the frequency of band occurrence (50% or higher), a common profile of the following approximate molecular weights of 18 bands was drawn: 6.5, 7.5, 9, 12, 14.5, 21, 27, 30, 35, 53.5, 57, 61.5, 68, 72, 78, 84, 94, and 104. Occasionally, the coccoid or spiral preparation showed a single band (200 or 66) not visible in the other form. 2-D electrophoretic analysis. | 2-D electrophoretic analysis revealed that coccoid preparations produced fewer spots than their spiral counterparts (135 and 163 spots, respectively). The spiral preparations showed spots in a molecular weight range of 22 to 219, and the coccoid forms showed spots in a range of 25 to 245. The pI range of the protein spots was similar for both preparations, ranging from 4.5 to 8.5. Only a few strains (n = 3) produced some spots observed exclusively in the coccoid forms (molecular weights of 245, 149, 136, 71, and 42 with pIs of 6.5, 7.1, 7, 4.4, and 4.5, respectively). In another three strains, the intensity of the spots was higher in the coccoid form . Western blot analysis. | All 21 H. pylori strains were challenged with homologous sera. We found variability in the number and intensity of bands recognized by the homologous sera. The number of reactive bands for the different strains ranged from 9 to 20 for the coccoid antigens and from 9 to 19 for the bacillary form. ELISA antigen selection. | Twelve out of 21 strains were selected due to the higher number of bands detected on the blot , and a higher protein content was revealed by 1- and 2-D electrophoresis. Strains with exclusive spots or higher intensity in the coccoid forms were also included . ELISA for IgG antibodies to spiral and coccoid preparations. | The IgG responses to H. pylori against coccoid and bacillary preparations are shown in Table . As observed, all the serum panels tested were reactive with both the coccoid and bacillary H. pylori preparations. The OD reading values showed high variability, which was reflected in elevated standard deviation values. The average ELISA values for patients with different gastric conditions were similar whether coccoid or spiral capturer antigens were used. The only exception was the cancer group, which showed lower values against both coccoid- and spiral-based antigens by ELISA. When the values were compared within each patient group, results showed significantly higher OD readings for the coccoid capturer antigen (P < 0.05). The only exception was the NUD patients, who had similar values for both antigens (P = 0.6551) . Adsorption assay. | The results of this assay showed that all the OD values decreased similarly (more than 90%) when sera were adsorbed with either coccoid or bacillary antigens. Western blot analysis with the antigen preparation used in ELISA. | A representative subset of serum samples tested (22 DU patients, 14 NUD patients, 11 cancer patients, and 21 asymptomatic individuals) was blotted against the coccoid and spiral capturer antigens used by ELISA. The results showed that sera from infected individuals recognized H. pylori epitopes with a wide individual variability. Accordingly, we could not find a common recognizable profile from patients with similar gastric conditions. However, some bands appeared with a higher frequency in most sera (molecular weights of ca. 104, 87, 66, 63, 61, 35, and 28) . Some bands were recognized only among the antigens presented by the bacillary form (molecular weights of 49, 32, 30, and 19) in serum samples from ulcer patients. Other bands were recognized only in the coccoid antigen by serum samples from asymptomatic individuals (molecular weights of 171, 156, 135, 116, 104, and 32), and yet other bands were recognized by sera from the ulcer patients (molecular weights of 144, 135, and 98). Serum samples from many asymptomatic subjects had antibodies to a 77-molecular-weight antigen; this epitope was not recognized by serum samples from symptomatic patients . To get a general picture of the epitopes recognized by blotting, we tested the antigens applied during ELISA against four pools of six sera representing each patient group. The results showed that the four groups of sera recognized the same bands in coccoid and bacillary antigens, with some minor differences in band intensity. The common profile recognized by the four sera groups is shown in Fig. . FIG. 1. | SDS-PAGE of bacillary (lanes 1 through 6) and coccoid (lanes 7 through 12) antigens of six Helicobacter pylori strains included for antigen characterization. SDS-PAGE of bacillary (lanes 1 through 6) and coccoid (lanes 7 through 12) antigens of six Helicobacter pylori strains included for antigen characterization. MW, molecular weight standard. FIG. 2. | Comparative 2-D gel electrophoresis of bacillary (a) and coccoid (b) protein preparations from a single strain. Comparative 2-D gel electrophoresis of bacillary (a) and coccoid (b) protein preparations from a single strain. Arrows show two spots found only in the coccoid form. The ovals show the spots that became more intense in the coccoid preparation. MW, molecular weight standard. FIG. 3. | Immunoblot of bacillary (even lanes) and coccoid (pair lanes) preparations from the 12 selected H. pylori Immunoblot of bacillary (even lanes) and coccoid (pair lanes) preparations from the 12 selected H. pylori strains for ELISA. Each antigenic preparation was tested against homologous serum. Lanes: MW, molecular weight standard; 1 and 2, strain 1; 3 and 4, strain 2; 5 and 6, strain 3; 7 and 8, strain 4; 9 and 10, strain 5; 11 and 12, strain 6; 13 and 14, strain 7; 15 and 16, strain 8; 17 and 18, strain 9; 19 and 20, strain 10, 21 and 22, strain 11; 23 and 24, strain 12. FIG. 4. | Antigenic protein bands of bacillary (lanes 1, 3, 5, and 7) and coccoid (lanes 2, 4, 6, and 8) pools used in ELISA recognized by pooling of sera from patients with different gastroduodenal conditions. Antigenic protein bands of bacillary (lanes 1, 3, 5, and 7) and coccoid (lanes 2, 4, 6, and 8) pools used in ELISA recognized by pooling of sera from patients with different gastroduodenal conditions. Lanes: 1 and 2, sera from gastroduodenal ulcer patients; 3 and 4, sera from NUD patients; 5 and 6, sera from cancer patients; 7 and 8, sera from asymptomatic individuals. TABLE 1 | IgG responses in OD units to bacillary and coccoid antigens in different groups of patients TABLE 2 | Frequency of band recognition of bacillary and coccoid antigen pools by sera from patients with different gastroduodenal conditions DISCUSSION : The amount of scientific information gained in the last 2 decades on H. pylori is impressive. Genetic and biochemical characteristics, and many aspects of virulence, have already been well described . The mechanism of transmission of H. pylori, however, is less well understood. Much research confirms that the oral-oral route is how one becomes infected with H. pylori . This does not, however, fully explain how H. pylori-related infections are so widespread the world over and why the prevalence is significantly higher in less-developed areas. One can postulate that the fecal-oral route, via contaminated water and/or foodstuffs, could be the missing link that explains the high prevalence of H. pylori infection in areas with poor sanitation . In order to satisfy this theory, one must explain how spiral cells are able to survive the hostile environmental factors that do not support them. Because coccoid cells are normally present both in vivo and in vitro, these forms could explain the persistence of H. pylori under hostile environmental conditions. To explore this issue, we attempted to determine whether the coccoid forms have any impact on the immune response to H. pylori in colonized individuals. We established a specific ELISA aimed at evaluating and comparing the immune response to coccoid and spiral forms against a panel of sera from symptomatic and asymptomatic infected individuals. The results revealed that all infected individuals either with or without symptoms were highly reactive to both forms. However, the coccoid antigen preparations usually gave significantly higher OD readings than their bacillary counterparts. The biologic significance of these differences is probably not relevant. It is tempting to speculate that the higher values found with coccoid antigen in ELISA may be due to the overexpression of one or more major epitopes in the coccoid morphology. However, the results obtained in the blotting assays make this explanation less probable. This is because we did not find any major differences in band recognition when sera were assayed against coccoid or bacillary antigens. The data from 68 serum samples analyzed indicated that the IgG response to the coccoid antigen was as strong as that to the spiral one . Most of the bands recognized by blotting were similar to those described by Nilsson et al. . We found no specific bands or profiles that could be associated with a particular morphological state or specific gastric condition. Kimmel et al. obtained similar results: when comparing antigenic patterns recognized in sera from different patients, they found no association of specific H. pylori antigens with antibodies in patients with specific gastroduodenal pathologies. The results presented here indicate that the coccoid forms of H. pylori are able to induce a humoral immune response. They also show that their immune response is similar to that induced by the bacillary forms. The analysis of the antigenic preparations applied in this study by 1-D and 2-D electrophoresis and immunoblotting showed that the total protein content of the coccoid forms is very similar to that of the spiral ones. Furthermore, it established that the cell proteins remain intact, as demonstrated by their immunogenic potential. The results of ELISA with adsorbed sera support this idea. These assays using cross-adsorbed sera confirm the presence of common antigens in coccoid and bacillary preparations. An important subject that remains to be elucidated is whether coccoid forms also elicit a cellular immune response or, in particular, whether there are differences in Th1 or Th2 involvement when coccoid antigens are presented. Future studies in our laboratory will focus on some of these issues. Backmatter: PMID- 12204951 TI - Antibody Responses of Cattle with Respiratory Coronavirus Infections during Pathogenesis of Shipping Fever Pneumonia Are Lower with Antigens of Enteric Strains than with Those of a Respiratory Strain AB - The serum antibody responses of cattle with respiratory coronavirus infections during the pathogenesis of shipping fever pneumonia were analyzed with different bovine coronavirus antigens, including those from a wild-type respiratory bovine coronavirus (RBCV) strain (97TXSF-Lu 15-2) directly isolated from lung tissue from a fatally infected bovine, a wild-type enteropathogenic bovine coronavirus (EBCV) strain (Ly 138-3), and the highly cell culture-adapted, enteric prototype strain (EBCV L9-81). Infectivity-neutralizing (IN) and hemagglutinin-inhibiting (HAI) activities were tested. Sequential serum samples, collected during the onset of the respiratory coronavirus infection and at weekly intervals for 5 weeks thereafter, had significantly higher IN and HAI titers for antigens of RBCV strain 97TXSF-Lu15-2 than for the wild-type and the highly cell culture-adapted EBCV strains, with P values ranging from <0.0001 to 0.0483. The IN and HAI antibody responses against the two EBCV strains did not differ significantly, but the lowest titers were detected with EBCV strain L9-81. Keywords: Introduction : The role of coronaviruses as respiratory pathogens in cattle was recently recognized when numerous coronavirus strains were isolated from the nasal secretions and lung tissues from cattle with fatal cases of shipping fever pneumonia (SFP). These isolates are referred to as respiratory bovine coronaviruses (RBCV) . The enteropathogenic potential of coronaviruses was identified previously when these viruses were isolated from diarrhea fluid or intestinal samples from neonatal calves with severe diarrhea . In addition, coronaviruses were implicated in winter dysentery of adult cattle . These coronaviruses are referred to as enteropathogenic bovine coronaviruses (EBCV). Significant phenotypic, antigenic, and genetic differences separate the newly recognized RBCV from EBCV . Furthermore, infectivity-neutralizing (IN) monoclonal antibodies (MAb) distinguished between the EBCV wild-type strain LY138-3 and the highly cell culture-adapted EBCV strain L9-81, as well as the Norden vaccine strain . Bovine coronaviruses are enveloped, have a positive single-stranded RNA genome of about 31 kb, and belong to the Coronaviridae family of the order Nidovirales . The spike (S) and the hemagglutinin esterase (HE) glycoproteins of the five structural proteins function in the infectious process through adsorption and coadsorption to the receptors on susceptible host cells . The S and HE glycoproteins are both capable of hemagglutinating rodent and chicken red blood cells (RBC). The HE but not S glycoprotein facilitates the release of viruses adsorbed to RBC in the receptor-destroying enzyme (RDE) test mediated by an acetylesterase (AE) component of the HE (-, , ). These two glycoproteins induce IN and hemagglutinin-inhibiting (HAI) antibodies. The objectives of the present investigations were to assess the antibody responses of cattle during a naturally occurring and experimentally monitored infection that had induced severe pneumonic signs, including SFP. The kinetics of IN and HAI antibody responses of these cattle were analyzed with antigens of an endogenous RBCV strain, which were isolated from lung tissue from a fatally infected bovine from this SFP epizootic, and with antigens of the wild-type and the highly cell culture-adapted EBCV strains. The analysis was conducted with sera sequentially collected from cattle which had primary and secondary immune responses to the RBCV infection. MATERIALS AND METHODS : Experimental design. | One hundred five 6- to 8-month-old cattle were involved in this specific epizootic and were subjected to nasal swab and blood sampling and testing at the time of assembly at the order-buyer's barn (day 0), after transport on day 7, and thereafter on days 14, 21, 28, and 35, during the pathogenesis of and recovery from SFP . Nasal swab samples were taken for virological and bacteriological studies, while blood samples for serum harvest were collected for immunological investigations. The cattle were separated into five response groups based on clinical signs of respiratory tract diseases and results of RBCV isolations . Response groups 1 and 5 were utilized in the present analysis. Response group 1 included 72 cattle that exhibited clinical signs of respiratory tract disease and that nasally shed RBCV on day 0, day 7, or both. Seven randomly selected cattle of the group shedding RBCV on day 7 were investigated in detail. Seven calves in response group 5 were also included in these serological investigations because they remained clinically healthy and RBCV isolation negative during the epizootic and the 5-week follow-up period. To analyze sequential serum samples from the 14 cattle selected for their IN and HAI antibodies, we used antigens of three specific strains of bovine coronaviruses. Cell line, virus isolates, and virus purification. | Clone G of human rectal tumor (HRT-18) cells was used at the 24th passage level for RBCV and EBCV propagation . The wild-type RBCV strain 97TXSF-Lu15-2 was used at its second passage in clone G cells after initial isolation from lung tissue from a calf that died on day 8 of the epizootic . The highly cell culture-adapted EBCV strain L9-81 was originally isolated from bovine fetal kidney cell cultures as the EBCV Mebus strain . The EBCV L9-81 strain was propagated in 75 sequential passages in different bovine cell cultures, and the final six passages were done in clone G cells. The virulent wild-type EBCV strain LY138-3 from intestinal samples of a diarrheic calf was initially maintained by oral inoculation of conventional calves with induction of enteritis . It was used as an antigen in its third clone G passage. Partially purified preparations of RBCV strain 97TXSF-Lu15-2 and EBCV strains L9-81 and LY138-3 were employed for serological tests . IN assay. | Serum samples were prepared in quadruplicate and tested for IN activities according to a previously reported method . Serum 1745 with an IN titer of 128 and BCV antibody-free serum from a normal calf with an IN titer of <8 were included as positive and negative controls, respectively . The serum IN titers were expressed as the reciprocal of the highest serum dilution that completely inhibited cytopathic changes in 50% of the wells. Assays for HA and RDE. | The assays for hemagglutinin (HA) and RDE were performed as reported previously with washed rat RBC prepared as 0.5% suspensions with phosphate-buffered saline (pH 7.4) containing 0.05% bovine serum albumin. HAI test. | The HAI titers of serum samples were determined with partially purified virus stocks, which were diluted to contain 8 to 16 U of both HA and RDE as antigens . Again, serum 1745 with an HAI titer of 128 was included as a positive control, while BCV antibody-free serum from a normal calf with an HAI titer of <8 was used as a negative control . The serum HAI titers were expressed as the reciprocal of the highest dilution of serum sample that completely inhibited the aggregation of rat RBC. Data analysis and statistical methods. | The serum IN and HAI titers were transformed to base 2 logarithms for statistical analysis. All data were presented as means +- standard errors of the means. The IN and HAI activities of the two response groups were compared by an analysis of variance of repeated measures designed with a split-plot arrangement of treatments. Pairwise comparisons of treatment and day differences were assessed with Scheffe's test. Interaction effects were examined with pairwise t tests of least-squares means for preplanned comparisons of treatments at specific day levels. All tests were considered significant at a probability of <0.05. RESULTS : Comparison of IN activities of serial serum samples with antigens of RBCV strain 97TXSF-Lu15-2 and EBCV strains L9-81 and LY138-3. | The antibody responses with serum samples collected from cattle in response groups 1 and 5 had significantly higher overall kinetics of IN activities against RBCV strain 97TXSF-Lu15-2 than against EBCV strain L9-81, with P values of 0.0001 and 0.0001, and EBCV strain LY138-3, with P values of 0.0001 and 0.0025, respectively. Significant differences in IN activities against RBCV strain 97TXSF-Lu15-2 and EBCV strain L9-81 were observed in serum samples collected from cattle in response group 1, with P values of 0.0439, 0.001, <0.0001, <0.0001, and <0.0001 for results on days 7, 14, 21, 28, and 35, respectively. Similarly, the cattle in response group 5 had significantly lower levels on days 7, 14, and 28, with P values of 0.0095, 0.0164, and 0.0164, respectively. Significant differences between serum IN activities against RBCV strain 97TXSF-Lu15-2 and these against the EBCV strain LY138-3 antigen were also detected for cattle in response group 1, with P values of 0.0029, <0.0001, 0.001, and <0.0001 for results for days 14, 21, 28, and 35, respectively, and for cattle of response group 5 on days 7, 14, and 28 (P values of 0.0446, 0.0399, and 0.0335, respectively). Interestingly, IN antibody titers against EBCV strains L9-81 and LY138-3 in cattle of response groups 1 and 5 did not differ significantly during this 5-week investigation. Comparison of HAI activities of serial serum samples with antigens of RBCV strain 97TXSF-Lu15-2 and EBCV strains L9-81 and LY138-3. | The kinetics of the responses in HAI antibodies against RBCV strain 97TXSF-Lu15-2 were significantly greater than those against EBCV strain L9-81, with P values of 0.0001 and 0.0001 for the results for cattle of response groups 1 and 5, respectively. The differences between RBCV strain 97TXSF-Lu15-2 and EBCV strain LY138 were significant at a P value of 0.0001 for both groups 1 and 5 . The HAI activities against RBCV strain 97TXSF-Lu15-2 and EBCV strain L9-81 were remarkably different in serum samples collected from cattle in response group 1 for days 7, 14, 21, 28, and 35, respectively, with P values of 0.0483, 0.0004, 0.0035, <0.0001, <0.0001, and <0.0001. The HAI activities showed similar differences among cattle in response group 5 from day 0 onward with a P value of <0.0001. Levels of HAI antibodies against RBCV strain 97TXSF-Lu15-2 and EBCV strain LY138-3 were also substantially different for the cattle in response group 1 for days 7, 14, 21, 28, and 35 (P values of 0.0035, 0.0483, <0.0001, 0.0001, and <0.0001, respectively), and for the cattle in response group 5 for days 0, 7, 14, 21, 28, and 35 (P values of 0.0024, <0.0001, <0.0001, <0.0001, <0.0001, and <0.0001, respectively). Significant differences between serum HAI antibody kinetics against EBCV strain L9-81 and EBCV strain LY138-3 were not observed in cattle of either response group. FIG. 1. | IN (A and B) and HAI (C and D) antibody responses against RBCV strain 97TXSF-Lu15-2 (circles), EBCV strain L9-81 (triangles), and EBCV strain LY138-3 (squares) in serum samples from cattle. IN (A and B) and HAI (C and D) antibody responses against RBCV strain 97TXSF-Lu15-2 (circles), EBCV strain L9-81 (triangles), and EBCV strain LY138-3 (squares) in serum samples from cattle. Cattle in response group 1 (A and C) shed RBCV on day 7 and had clinical signs of respiratory tract disease, while cattle in response group 5 (B and D) did not develop adverse clinical signs and remained virus isolation negative, despite exposure to the animals that were shedding virus (response group 1). Data are means +- standard errors of the means (error bars). Results are shown for seven cattle (A and C) and for seven cattle (B and D). DISCUSSION : Prospectively designed and experimentally controlled sampling and testing of cattle during a naturally occurring SFP epizootic created fortuitous circumstances for analyzing the kinetics of antibody responses of cattle to a native, wild-type RBCV infection . This experiment also provided an excellent opportunity to compare antibody reactivities with both the wild-type RBCV and the highly cell culture-adapted and wild-type EBCV strains in serum samples collected from cattle involved in a naturally occurring RBCV infection. These conditions differ from experimentally induced infections that routinely involve giving virus inocula adapted to, and propagated in, cell cultures. Importantly, the IN and HAI antibody responses against RBCV were significantly greater than those detected with the wild-type EBCV strain LY138-3 and the highly cell culture-adapted EBCV. The EBCV strain L9-81 is virtually identical to the Norden vaccine strain that is used in attempts to protect against intestinal coronavirus infections of neonatal calves. Antigens of the highly cell culture-adapted EBCV strains reacted minimally with antibodies induced by naturally occurring RBCV infections. These findings call for analysis and a search for appropriate antigens in efforts to immunize cattle against these currently prevailing RBCV infections. Bovine coronaviruses contain a large, single-stranded RNA genome and may mutate in the natural host to generate quasi-species differences . Phenotypic changes may result from adaptation to, and propagation, of the virus in cell cultures, effecting the selection of host cell virus mutants. Differences in infectivity between the highly cell culture-adapted EBCV strain L9-81 and the wild-type strains of RBCV and EBCV consisted of a greatly expanded range of permissive host cells. This EBCV strain replicates readily, particularly in the presence of trypsin and in cultured bovine fetal kidney, spleen, thyroid, or adrenal cells, as well as in Vero cells . In contrast, the wild-type RBCV strains had restricted host cell ranges, because only clone G cells were permissive for initial isolation from clinical samples at low passage levels . Differences between the RBCV and EBCV strains also were observed in the HA patterns, cell fusion, and AE activities . Antigenic differences between RBCV and EBCV could not be detected by tests such as immunodiffusion immunofluorescence, immunoblotting, or enzyme-linked immunosorbent assays with antigens that include all the structural proteins of these coronaviruses . MAb specific for S glycoprotein with IN activity distinguished between EBCV strain L9-81 and several wild-type strains of EBCV . Differences between EBCV strain L9-81 and the wild type strain LY138-3 of EBCV in inhibition of AE activity with HE-specific MAb and in HA (4) reactivities were also detected. Genotypic differences consisted of changes in the regions of the polymerase and associated genes, in the genes of S and HE glycoproteins, and in other nonstructural genes . Nucleotide and deduced amino acid mutations were within the A and B immunoreactive domains of the S-I subunit of the spike protein of the RBCV strains but not the EBCV strains, which could be the basis for the observed diversity in antigenic stimulation during these naturally occurring infections . Comparatively, the differences in the kinetics of the primary bovine antibody responses to naturally occurring and experimentally monitored RBCV infections with the antigens of the wild-type RBCV and EBCV strains as well as the highly cell culture-adapted EBCV appear to be similar to responses of chickens to infections with the coronavirus of infectious bronchitis. Early antibodies generated during a primary immune response distinguished several serotypes of infectious bronchitis in both IN and HAI tests . Backmatter: PMID- 12204950 TI - Comparison of In-House and Commercial Slides for Detection by Immunofluorescence of Immunoglobulins G and M against Bartonella henselae and Bartonella quintana AB - We compared the sensitivities and specificities of indirect fluorescent antibody tests developed in our laboratory and commercially available from Focus Technologies (FT; formerly MRL Diagnostic) for detection of serum antibodies to Bartonella spp. Serum samples tested were from patients with culture- or PCR-confirmed Bartonella quintana or B. henselae infections causing cat scratch disease (CSD), chronic bacteremia, or endocarditis. At a cutoff titer of 64, the FT test had higher sensitivity than our in-house test in detecting anti-B. henselae immunoglobulin G (IgG) antibodies in CSD patients (91.2 versus 52.9%; P < 0.001). The specificity in serum samples from 85 control patients was, however, lower with the FT test (87%) than with the in-house test (98.8%) (P = 0.002). A cutoff titer of 128 improves the specificity for the FT test but lowers the sensitivity to 85%. For patients infected with B. henselae, our in-house test, but not the FT test, enabled endocarditis to be detected more reliably. With the in-house test, titers of IgG against B. henselae of >=1,024 were found only in endocarditis patients and not in CSD patients. With the FT test, 19.1% of CSD patients had titers of IgG against B. henselae of >=1,024 (P < 0.001). Our in-house technique also improved detection of anti-B. quintana antibodies in homeless patients with endocarditis. IgG titers of >=1,024 were present in 75% of serum samples, but only in 16.7% of serum samples with the FT test (P = 0.004). Since each test has advantages over the other, the serological diagnosis of Bartonella infections would benefit if both tests were used concurrently. Keywords: Introduction : There are now 18 species within the genus Bartonella. Three are well-known human pathogens: Bartonella bacilliformis , B. quintana , and B. henselae . Others have been isolated from patients on single occasions only: B. elizabethae (one endocarditis case) , B. vinsonii subsp. berkhoffii (one endocarditis case) , and B. vinsonii subsp. arupensis (one case of bacteremia) . B. clarridgeiae , and B. washoensis have been tentatively associated with cat scratch disease (CSD) and myocarditis, respectively. B. quintana and B. henselae are species of worldwide distribution . B. quintana was first described as the agent of trench fever in 1918 and is now known to be responsible for louse-borne bacteremia and endocarditis in homeless people and bacillary angiomatosis in AIDS patients . Humans are the only known reservoir of B. quintana, and transmission among people occurs via the body louse . B. henselae, a species first recognized in 1990 , is the main etiological agent of CSD and is also responsible for bacillary angiomatosis and peliosis hepatitis in immunocompromised (mainly AIDS) patients , as well as bacteremia and endocarditis . Cats are the main reservoir of B. henselae, and people become infected following cat scratches or bites. A role for the cat flea (Ctenocephalides felis) as a vector for human transmission has been proposed . B. henselae comprises two distinct genotypes, B. henselae Houston and B. henselae Marseille . Diagnostic techniques for infections with Bartonella spp. include culture of the pathogen , molecular biology techniques, especially PCR, to amplify Bartonella sp. genes , and serology . The usefulness of these various techniques may vary with the disease present. Culture of Bartonella spp. may be successful with blood from bacteremic patients , from cardiac valve specimens from endocarditis patients , and from skin, lymph node, or other organ biopsy samples from patients with bacillary angiomatosis . By contrast, B. henselae is only rarely isolated from patients with CSD . Amplification of Bartonella sp. DNA in tissue biopsy samples is an invasive technique and is mainly useful for patients with CSD (lymph node specimen) , bacillary angiomatosis (angiomatous lesions of skin or other organs) , or endocarditis (cardiac valve when removed) . Serology is the only noninvasive diagnostic technique, and it has been evaluated in the diagnosis of CSD , bacteremia , and endocarditis . The indirect fluorescent antibody (IFA) test was first described for CSD by Regnery et al. , who used B. henselae grown in Vero cells as antigens. It remains the most frequently used technique, and many laboratories have performed Bartonella serology using tests developed in-house, with reported sensitivities varying considerably, from nearly 100% to less than 30% . Commercially prepared antigen slides are now available for B. henselae and B. quintana serology , and in this report we compare the sensitivity and specificity of one of these tests with our in-house IFA test, which has been used by our laboratory for 10 years. We compared the abilities of the two tests to detect immunoglobulin G (IgG) and IgM antibodies in serum samples from patients known to have CSD (B. henselae), chronic bacteremia (B. quintana), or endocarditis (B. henselae and B. quintana). MATERIALS AND METHODS : Patients and serum samples. | Serum samples from 104 patients infected with Bartonella spp., comprising CSD patients (n = 68), patients with endocarditis (n =17; 5 patients with B. henselae endocarditis and 12 with B. quintana endocarditis), and homeless people (n = 19) with chronic B. quintana bacteremia but without endocarditis, were used in the study. Diagnoses of CSD were based on PCR amplification of B. henselae DNA from lymph node samples of patients with regional lymphadenopathy and close contact with cats . Bartonella endocarditis was diagnosed by positive blood cultures or PCR amplification of DNA of the organism from valve samples . Diagnoses of B. quintana bacteremia for the 19 homeless patients were based on blood culture results . Controls included 40 apparently healthy blood donors (group A), 10 patients initially thought to have CSD because of lymphadenopathy but later, by culture, found to have tuberculosis (group B), and 35 patients who were seropositive for bacterial or viral diseases not related to Bartonella species and for whom stored serum samples were available (group C). Group C included patients seropositive for cytomegalovirus , Epstein-Barr virus , hepatitis B virus , human immunodeficiency virus , acute Q fever , chronic Q fever , or Mediterranean spotted fever caused by Rickettsia conorii . Immunofluorescence assays. | All serum samples were examined by IFA for the presence of antibodies to B. quintana and B. henselae. For antigens we used either our slides prepared in-house or commercial slides from Focus Technologies (FT), Cypress, Calif. (distributed in France by Eurobio, Paris, France). We prepared in-house slides by using B. henselae strain Houston-1 or B. quintana strain Oklahoma as the antigen, as previously described . Briefly, bacteria were grown in the human endothelial cell line ECV 304 for 2 weeks before infected cells were recovered and lysed by sonication. Cell suspensions were centrifuged at 1,500 rpm (700 x g) for 10 min to remove cell debris, and bacteria were purified by using a sucrose gradient (phosphate-buffered saline [PBS] with 25% sucrose; 6,000 x g for 30 min). The resulting pellet was washed in PBS twice (6,000 x g for 10 min), and purified bacteria were pelleted by centrifugation (10,000 x g for 10 min), resuspended in 1 ml of 0.1% paraformaldehyde, and incubated at 4C for 24 h for fixation. The bacterial suspension was repelleted by centrifugation (10,000 x g for 10 min), resuspended in sterile distilled water, and stored at -80C before use. This whole-cell antigenic preparation was plated onto Dynatech slides and used for detection of either IgM or IgG antibodies by IFA. FT uses the same B. henselae Houston-1 and B. quintana Oklahoma strains for preparation of its commercial slides. Bacteria are grown in Vero cells for preparation of slides used to detect IgG and on blood agar before being harvested and diluted in yolk sac fluid for slides used to detect IgM. For both the FT and in-house tests, IgG titers of >=64 and IgM titers of >=20 were used as cutoff titers with B. henselae or B. quintana antigens. Statistical analysis. | Differences in antibody titers between groups and/or between IFA techniques were analyzed by using the chi2 test for categorical variables or Student's t test for continuous variables. A two-sided P value of <0.05 was considered significant. RESULTS : With serum samples from CSD patients, the sensitivities of the in-house and FT tests using a cutoff titer of 64 for detection of anti-B. henselae IgG antibodies were 52.9 and 91.2%, respectively (P < 0.001). Median IgG titers were 64 with the in-house test (range, <64 to 256) and 256 with the FT test (range, <64 to 1,024) . IgM antibodies were detected at a low titer (i.e., 20) for only two patients with the in-house test and one patient with the FT test. IgG antibodies cross-reacting with B. quintana were detected in most of the serum samples (35 of 36 [97.2%]; median titer, 64) from CSD patients with the in-house test, but in only one serum sample with the FT test (P < 0.001). All endocarditis patients had IgG titers of >=64 against both B. quintana and B. henselae antigens. However, significant differences in antibody titers were found between the in-house and FT tests. With the in-house test, 9 of 12 (75%) serum samples from patients with B. quintana endocarditis had homologous IgG titers of >=1,024, whereas such high titers were present in only 2 of 12 (16.7%) patients with the FT test (P = 0.004). Also, median homologous IgG titers in serum samples from B. quintana endocarditis patients were 1,024 with the in-house test but only 256 with the FT test (P < 0.01) . Anti-B. henselae cross-reacting antibodies were found in all B. quintana endocarditis patients (median IgG titers of 1,024 and 256 with the in-house and FT tests, respectively). All five B. henselae endocarditis patients had homologous IgG titers of >=1,024 with both tests (range, 1,024 to 4,096; median titers of 2,048 and 1,024 with the in-house and FT tests, respectively). Most of them had IgG antibodies which cross-reacted with the B. quintana antigen, albeit at lower titers (median titers of 512 and 256 with the in-house and FT tests, respectively). No antibodies to B. quintana were found in 2 of 19 (10.5%) and 3 of 19 (15.8%) serum samples from homeless patients with B. quintana bacteremia with the in-house and FT tests, respectively. For the remaining patients, anti-B. quintana IgG antibodies were found at low titers (range, 64 to 512; median, 128 with both tests) . Cross-reacting anti-B. henselae IgG antibodies were found for most of these patients, at similar titers. The two bacteremic patients with a homologous IgG titer of 512 had no echocardiographic evidence of endocarditis. None of the 40 serum samples from blood donors (group A) were found to contain antibodies to B. quintana or B. henselae by either of the two tests. Among the group B controls, only one had an IgG titer (of 64) detectable by the in-house test, while with the FT test, serum samples from this and two additional controls had IgG titers of 64. No reactive antibodies were found for the 35 group C controls when the in-house test was used. In contrast, serum samples from 8 of 35 (22.8%) group C controls had IgG titers (of 64 to 256) to Bartonella spp. in the FT test. Cross-reacting IgG antibodies were found in serum samples from three cytomegalovirus patients, two chronic Q fever patients, and three patients with Mediterranean spotted fever. Thus, by using a cutoff titer of >=64, the specificities of the in-house and FT tests for reactive IgG were both 100% for blood donors (group A); 90 and 70%, respectively, for group B; and 100 and 90%, respectively, for group C (P = 0.002). For all 85 controls taken together, the specificities of the FT and in-house tests were 87 and 98.8%, respectively (P = 0.002). To ensure a specificity of >=95% with the FT test, a titer of >=128 rather than >=64 should have been considered the cutoff. With this cutoff titer, the sensitivity would be lower (85.3%) for CSD patients, but this would still be superior to that of the in-house test. FIG. 1. | Anti-B. henselae Anti-B. henselae IgG titers in CSD patients. FIG. 2. | Homologous IgG titers in patients with B. quintana Homologous IgG titers in patients with B. quintana endocarditis. FIG. 3. | Anti-B. quintana Anti-B. quintana IgG titers in homeless patients with chronic B. quintana bacteremia. DISCUSSION : Bartonella serology has been performed in our laboratory for several years now, by using B. quintana or B. henselae grown in cell culture as the antigen. The aim of our study was to compare the sensitivity and specificity of our in-house test to that of a recently available commercial test from FT. With CSD patients, our results confirm the previously reported high sensitivity of the FT serological test , whereas a lower sensitivity and significantly lower median IgG titers were found with our in-house test. Reactive IgM antibodies were rarely detected for CSD patients by either test, and these antibodies would thus appear to be of little use diagnostically, as previously mentioned . The sensitivity of the B. henselae IFA test first described by Regnery et al. for detection of CSD has been reported from different laboratories as 32% to 100% . Enzyme immunoassay (, ,, , , ) and Western blotting techniques have not been consistently superior to IFA for the diagnosis of CSD. Several factors may influence the apparent sensitivity of a serological test. First, sensitivity may change according to disease case definition. Historically, a CSD case has been defined as the combination of regional lymphadenopathy with characteristic histopathological features (granuloma), the presence of an inoculation skin lesion (e.g., cat scratch), a history of cat contact, a positive result on a skin test using an antigen prepared from a CSD lymph node exudate, and negative laboratory results for other causes of lymphadenopathy . More recently, less restrictive case definitions have been used (e.g., regional lymphadenopathy and a recent contact with a cat ), since the skin test antigen is no longer available and lymph node biopsies are performed infrequently. However, a number of patients (e.g., 14% in the study of Flexman et al. ) with elevated anti-B. henselae antibody levels report no previous contact with a cat, and many patients present with clinically atypical cases of CSD . Use of a restrictive case definition for CSD may erroneously increase the sensitivity of a serological test because only patients with evident epidemiological and clinical findings suggestive of CSD will be considered. Thus, in a series of 354 suspected cases of CSD where serum samples were examined for anti-B. henselae antibodies, Flexman et al. reported a sensitivity of 18% (with a cutoff titer of >=1:64). However, when only the 98 patients who met the clinical case definition of CSD (lymphadenopathy and recent contact with a cat) were considered, the sensitivity was 54%. Accurate determination of the sensitivity and specificity of a serological test requires well-defined patient groups (infected and uninfected), identified by objective diagnostic procedures rather than clinical and epidemiological data. In a few recent studies, B. henselae DNA amplification from lymph node biopsy samples was used for confirmation of the diagnosis of CSD in a portion of cases . Giladi et al. recently evaluated an enzyme immunoassay for detection of anti-B. henselae antibodies in PCR-confirmed cases of CSD. We report the first evaluation of IFA tests for Bartonella-infected patients for whom diagnoses were unambiguously established by either culture or DNA amplification using PCR. Diagnoses of diseases not related to Bartonella spp. in control patients were also carefully documented The apparent sensitivity of a serological test may also be greatly influenced by the cutoff values used. Usually, cutoff titers are chosen to obtain a specificity of at least 95% in control patients . The control groups are usually composed of a mixture of apparently healthy blood donors and patients with a disease distinct from the disease being tested. Using an IFA test for detection of anti-B. henselae IgG antibodies at a cutoff titer of 64 in CSD patients, Regnery et al. , and Zangwill et al. in the United States, reported sensitivities of 88 and 84%, respectively, but the specificity was only similar90%. In each study, specificity was high at similar99% when a titer of 512 was considered the cutoff, but sensitivities decreased to only 67 and 64%, respectively. Bergmans et al. in The Netherlands reported a sensitivity as low as 31.8% by using a cutoff titer of 512, which was needed to obtain a specificity of >=95% in blood donors. The cutoff titer may depend on the seroprevalence of antibodies to Bartonella spp. in the population, and this may vary with geographical region . We found a specificity of 98.8% with the in-house test but only 87% with the FT test for the 85 controls we studied. With the FT test, the specificity was only 70% for the 10 patients with tuberculous lymphadenopathy. In our laboratory, cultures of about 5% of lymph node biopsy samples submitted yield Mycobacterium tuberculosis, while Bartonella spp. are demonstrated in approximately 30% of samples by culture or PCR. A highly specific test for CSD is critical in order to accurately rule out other, more serious diseases which may present as chronic lymphadenopathy, including tuberculosis or lymphoma. To ensure a specificity of >=95% with the FT test, a cutoff titer of >=128 should be used instead of >=64. It is becoming clear that some patients with typical CSD have no detectable antibodies to B. henselae, as shown in our study. Demers et al. reported a seronegative patient diagnosed by culture of B. henselae from a lymph node. Serology may be negative early in the course of the disease . Also, the antigenic preparation used for serology may greatly influence the results. Low antibody titers are usually found when B. henselae grown on agar is used as the antigen rather than cell culture-grown organisms . Poor antibody responses to B. henselae antigen have also been ascribed to the antigenic variability between the two B. henselae serotypes that have been described . Both the in-house and FT tests use B. henselae strain Houston-1 grown in cell culture as the antigen. Discrepancies between the two techniques with CSD and endocarditis patients may be explained by the mode of antigen preparation and conservation, as this may greatly influence antigenic-epitope recognition by serum antibodies . Finally, other Bartonella species (including B. clarridgeiae) may cause CSD , and antibodies to these organisms do not react with B. henselae. This was not the case in our study, since our PCR results showed that all CSD patients were infected with B. henselae, and the lack of detectable antibodies to B. henselae in some patients may then truly represent a lack of sensitivity in the serological tests. For homeless patients with chronic B. quintana bacteremia, both the FT and in-house tests detected only low levels of homologous IgG antibodies, and both failed to detect any reactive antibodies in many patients with culture-positive B. quintana bacteremia. There are few serological data on B. quintana infections. Although two large epidemics of trench fever occurred during the first and second World Wars , B. quintana had not yet been cultured and serology was unavailable . More recently, data from homeless people with B. quintana infections indicate that most suffer from chronic, intermittent bacteremia and that many do not have detectable antibodies to B. quintana when infections are diagnosed . The in-house and FT tests both detected homologous antibodies in patients with B. henselae endocarditis. However, only IgG titers obtained with the in-house technique allowed efficient discrimination between endocarditis patients and those with CSD. All B. henselae endocarditis patients presented with homologous IgG titers of >=1,024, whatever the technique used. In contrast, none of the CSD patients presented with IgG titers of >=1,024 with the in-house technique, whereas 13 of 68 (19.1%) had IgG titers of >=1,024 with the FT technique (P < 0.001). Thus, our in-house technique, but not the FT technique, may help define patients with a high likelihood of Bartonella-related endocarditis. In our experience, this may be critical, because Bartonella sp. endocarditis is a chronic, clinically unspecific disease. In this study, an anti-B. henselae IgG titer of >=1,024 with the in-house technique was 100% predictive of endocarditis , which is in accordance with previous reports from our laboratory . The in-house technique was more effective than the FT technique in detecting homologous IgG antibodies in serum samples from homeless patients with B. quintana endocarditis. The strain of B. quintana and the number of passages in culture have previously been reported to influence seroreactivity . The FT and in-house techniques use the same B. quintana Oklahoma strain, grown in cell cultures. The use of a cell culture- rather than an agar-grown antigen may favor the detection of specific antibodies, as previously described for B. henselae . With the in-house test, a homologous IgG titer of >=1,024 was 100% predictive of B. quintana endocarditis. Many B. quintana endocarditis patients, however, presented with lower IgG titers, and titers of 256 or 512 were found in 5 of 12 (41.6%) endocarditis patients. Such titers were also found in 7 of 19 (36.8%) homeless people with chronic bacteremia but with no endocarditis. These two groups of patients, then, could not be reliably differentiated by using IgG titers to B. quintana. Cross-reactions have been described between various Bartonella spp. and with Chlamydia trachomatis , Coxiella burnetii , and many other species, including Rickettsia rickettsii, Ehrlichia chaffeensis, Treponema pallidum, Francisella tularensis, and Mycoplasma pneumoniae . Cross-reactivity of serum samples containing IgM antibodies directed at Epstein-Barr virus capsid antigen has also been reported . The cross-reactivity between Bartonella sp. and Chlamydia sp. antigens has been particularly troublesome, and endocarditis supposedly due to a Chlamydia sp. on the basis of serology has later been shown to be due to a Bartonella sp. . The possibility of cross-reactions interfering with the serological diagnosis of Bartonella endocarditis is removed if all suspect serum samples are also tested for antibodies to C. burnetii and Chlamydia spp. In conclusion, our study has shown the high sensitivity of the FT test in detecting IgG against B. henselae in CSD patients. It should, however, be used with a cutoff titer of 128 to raise the specificity of the test to >=95%. Highly specific tests are essential for CSD, because a misdiagnosis may delay the detection of more serious conditions (e.g., lymphoma). Our in-house test had an advantage over the FT test in that it allowed CSD patients to be reliably differentiated from those with B. henselae endocarditis. An IgG titer of >=1,024 was 100% predictive of B. henselae endocarditis. Our in-house test was also more effective than the FT test at detecting antibodies to B. quintana in homeless patients, especially those with endocarditis. An anti-B. quintana IgG titer of >=1,024 was 100% predictive of endocarditis, although many patients with endocarditis presented with lower antibody levels. Since both tests have advantages, concurrent use of both tests should be considered in order to improve the diagnosis of B. henselae and B. quintana infections. Backmatter: PMID- 12204967 TI - Double-Blind Study To Evaluate Flow Cytometry Analysis of Anti-Live Trypomastigote Antibodies for Monitoring Treatment Efficacy in Cases of Human Chagas' Disease AB - The validation of flow cytometry analysis of anti-live trypomastigote antibodies (FC-ALTA) to monitor cure after treatment of Chagas' disease was evaluated with serum samples from treated and nontreated chagasic patients. After optimization of the original technique, toward better sensitivity and applicability to field surveys, we design a double blind study of 94 coded samples classified into the following categories: patients not treated (NT) and patients treated but not cured (TNC), both presenting positive conventional serology and xenodiagnosis; patients treated and cured (TC), showing negative serology and xenodiagnosis; and patients treated under evaluation (TUE), who presented positive or oscillating conventional serology (CSA) but negative xenodiagnosis. Coded samples, diluted 1:256, were assayed by incubation with live cell culture trypomastigotes, which were subsequently stained with fluorescein isothiocyanate-conjugated anti-human immunoglobulin G, with prior fixation and analysis by flow cytometry. The results were expressed as the percentages of positive fluorescent parasites (PPFP) for each individual sample, establishing 20% PPFP as the cutoff between negative and positive results. Our data demonstrated that all NT and TNC presented positive results while all but one TC had a PPFP lower than 20%. Analysis of TUE demonstrated a wide degree of reactivity, with PPFP values that were negative (PPFP <= 20%), low positive (20% < PPFP <= 50%), and high positive (PPFP > 50%). As TUE with negative PPFP presented negative xenodiagnosis and positive or oscillating CSA, they were classified as dissociated according to the criteria of Krettli and Brener (J. Immunol. 128:2009-2012, 1982) and could indeed be considered cured after chemotherapy. This study demonstrates and validates the use of FC-ALTA to easily identify anti-live trypomastigote membrane-bound antibodies, offering another approach for investigating and monitoring the efficacy of specific chemotherapy in cases of human Chagas' disease. Keywords: Introduction : Chagas' disease is a parasitic infection caused by the protozoan Trypanosoma cruzi, which is widespread in South and Central America, and transmitted mainly by an invertebrate vector, triatomine . Although autochthonous cases are rare in other localities, transmission of Chagas' disease by blood transfusion may represent an increasing public health problem in countries in which Chagas' disease is not endemic . Chemotherapy with nitroheterocyclic compounds has been indicated for the treatment of acute cases . In patients with chronic disease, however, the value of chemotherapy is still controversial. So far, it has been recommended for the treatment of patients with recent chronic disease who have not yet developed clinical symptoms . One of the major challenges regarding the evaluation of treatment effectiveness is the lack of truthful laboratorial approaches for use as tools for cure criteria. Two categories of tests are available, including serological and parasitological methods. Parasitological tests are based on parasite demonstration by hemoculture, xenodiagnosis, or parasitological molecular test (PCR), whereas serological methods evaluate the presence of specific antibodies by immunological methods, such as indirect hemagglutination (IHA), indirect immunofluorescence assay (IFA), and enzyme-linked immunosorbent assay (ELISA). Using serological approaches, Krettli and Brener proposed that sera from chronic chagasic patients present two types of anti-parasite antibodies with different functional activities named lytic antibodies (LA) and conventional serology antibodies (CSA). LA are associated with resistance in active ongoing infection and can be detected by complement-mediated lysis (CoML) and indirect immunofluorescence, all with live trypomastigotes . On the other hand, CSA are neither associated with resistance nor able to bind to live parasites, but they do react to soluble or fixed epimastigote antigens and are detected by different immunological methods . Considering these findings, clinical trials have demonstrated that LA gradually become negative after treatment, whereas CSA can remain positive over decades. So it has been proposed that the cure criterion for Chagas' disease should be based on either negative parasitological tests or the absence of anti-live trypomastigote antibodies (ALTA) . A flow cytometric method to detect ALTA (FC-ALTA) has been previously described as a new approach for monitoring the efficacy of treatment in cases of human Chagas' disease. The performance of FC-ALTA has been demonstrated to be comparable to that of CoML, offering an alternative method for easily identifying anti-live T. cruzi membrane-bound antibodies . In this study, we describe an optimization of FC-ALTA analysis and evaluate its performance for clinical studies. A double-blind study, with serum samples from treated and nontreated chagasic patients, was conducted to validate the method, reemphasizing its applicability for monitoring cure after treatment of Chagas' disease. Moreover, the data presented here offer an optimization of the original technique, making our method more sensitive and applicable to field studies. MATERIALS AND METHODS : Patients. | The inclusion of all subjects in our investigation had the approval of the FIOCRUZ Ethical Committee (Brazilian Health Ministry). In this study, we analyzed 94 patients (53 females and 41 males) ranging in age from 6 months to 68 years assisted by one of us at the Faculdade de Medicina, Universidade Federal de Goias, GoiAnia, Goias, Brazil. Chagas' disease diagnosis was established in all patients by positive xenodiagnosis and three positive serological tests, including IHA, IFA, and ELISA. All patients in the acute (n = 13) and subacute (n = 3) phases were treated with benznidazole (BZ) (60 mg/kg of body weight/day for 60 days). Patients in the chronic phase (n = 78) were treated with BZ (n = 44), allopurinol (n = 3), nifurtimox (n = 3), or BW349C59 (n = 6) or not treated (n = 22). In this study, we did not focus on the efficacy of the different therapeutic schemes used since it was not the major goal of our present investigation. After clinical, parasitological, and serological follow-up studies, ranging from 3 to 26 years, patients were classified into three different categories: not treated (NT) (n = 22), all patients with chronic disease; treated but not cured (TNC) (n = 19), 4 patients with acute- and 15 patients with chronic-phase disease; treated and cured (TC) (n = 25), 8 patients with acute-, 3 patients with subacute-, and 14 patients with chronic-phase disease; and treated under evaluation (TUE) (n = 28), 1 patient with acute- and 27 patients with chronic-phase disease . NT and TNC patients had positive results on both serological and parasitological tests. Patients were considered TC only when both tests, serological and parasitological, were consistently and repeatedly (at least eight times) negative. Patients presenting with positive (n = 17) or oscillating (n = 11) serology with negative parasitological tests were considered TUE. Samples were coded and stored pure at -20C or in 50% glycerol at room temperature. Conventional serological tests. | Three kinds of serological tests were performed as recommended by the World Health Organization : IHA (Imunoserum, Sao Paulo, Brazil), IFA (Biolab, Rio de Janeiro, Brazil), and ELISA (Abbott). At the beginning of the study, we used the complement fixation reaction, which was available at that time, to screen some patients. Xenodiagnosis. | Xenodiagnosis was performed on all patients before and after treatment as described by Cerisola et al. by using triatomine from Dipetalogaster maximus and Triatoma infestans according to availability. Forty triatomines, after a 15-day fast, were allowed to feed on each patient for 30 min. Microscopic examination of intestinal contents was carried out 30 and 60 days after feeding. The number of xenodiagnoses per patient ranged from 8 (320 triatomines) to 114 (4,500 triatomines). Cell line and parasites. | NTCT clone 929 cells (L929, ATCC CCL 1) were maintained in our laboratory by serial passages and frozen in liquid nitrogen. For the assays, 5 x 105 L929 cells were seeded in tissue culture flasks (25 or 75 cm2, Falcon; Becton Dickinson, San Jose, Calif.) with 10 ml of RPMI medium (GIBCO, Grand Island, N.Y.) containing 10% fetal bovine serum (FBS) and incubated at 37C in humidified air containing 5% CO2. After 2 or 3 days, the monolayer was infected with 3 x 106 to 7 x 106 trypomastigotes of T. cruzi strain CL . Trypomastigotes were obtained initially from infected mice and subsequently from the supernatant of another infected monolayer. Infected cells were washed daily with RPMI-5% FBS. The cultures were maintained in RPMI and 10% FBS at 33C, 5% CO2, and 95% humidity . After 5 to 7 days of incubation, the trypomastigotes started being released into the supernatant. The cell culture was transferred to a 50-ml tube and centrifuged at low speed (100 x g for 10 min at room temperature) to remove cell debris and some contaminating amastigotes. Trypomastigotes recovered from the supernatant were washed three times (1,000 x g for 15 min at 4C) and resuspended in phosphate-buffered saline (PBS) and 10% FBS at a concentration of 1 x 106 parasites/ml. A 50-mul sample was used for FC-ALTA. FC-ALTA. | The immunofluorescence reaction was performed as described by Martins-Filho et al. , modified as follows. Live trypomastigotes (500,000) were incubated at 37C for 30 min in the presence of a 1:256 final dilution of serum (whole or diluted 1:1 in glycerol), using a coded serum sample. After incubation with sera, parasites were washed once with PBS containing 10% FBS. Parasite suspensions were reincubated at 4C for 1 h or 37C for 30 min in the dark in the presence of fluorescein isothiocyanate (FITC)-conjugated anti-human immunoglobulin G (IgG) antibody (anti-whole molecule) (Sigma, St. Louis, Mo.) diluted 1:400 in PBS-10% FBS. The use of different temperatures and times of incubation with FITC-conjugated anti-human IgG antibody represented the major change from the original method described by Martins-Filho et al. , where incubation was performed only at 4C for 60 min. Each assay included an internal control of nonspecific binding as well as positive and negative controls. To monitor unspecific binding, parasites not exposed to human serum were incubated with FITC-conjugated anti-human IgG. Positive controls included sera from NT chagasic patients, whereas samples from uninfected individuals were included as negative controls. After being stained, labeled parasites were washed with PBS-10% FBS and fixed on ice for 30 min with a fix solution (10 g of paraformaldehyde per liter, 1% sodium cacodylate, 6.65 g of sodium chloride per liter, and 0.01% sodium azide, pH 7.2). Flow cytometry analysis was performed up to 24 h after parasite fixation. FACScan data storage and analysis. | Flow-cytometric measurements were performed on a Becton Dickinson FACScan interfaced to an Apple Quadra FACStation. CellQuest software was used for both data storage and analysis. Trypomastigotes were identified based on their specific forward (FSC) and side (SSC) laser-scattering properties. Following FSC and SSC adjustments, parasites were localized on FSC (size) by SSC (internal complexity-granularity) dot plot distribution . Parasites were then selected by gating on the FSC by SSC dot plot distribution. An average of 8,000 to 9,000 gated trypomastigotes was analyzed for relative fluorescence intensity by using single FITC histograms for each individual sample. A control marker of up to 2% of the parasites that were fluorescence positive was set up on the FITC-conjugated internal control histogram . This marker was used to determine the percentage of positive fluorescent parasites (PPFP) for each sample. Figures show representative histograms for the negative and positive controls, respectively. Data analysis was performed by establishing 20% PPFP as the cutoff between negative and positive results as described by Martins-Filho et al. . Thus, samples were considered positive when the PPFP was >20% and negative when the PPFP was <=20%. Positive results were additionally classified according to the method of Cordeiro et al. as low positive (20% < PPFP <= 50%) and high positive (PPFP > 50%). Statistics. | Statistical analysis was performed using one-way analysis of variance followed by Student's t test. Differences were considered significant when the P value was <0.05. FIG. 1. | FC-ALTA. FC-ALTA. (a) Dot plot analysis of a representative trypomastigote distribution (R1) based on size and granularity. Single histograms represent PPFP values for FITC-conjugated anti-human IgG internal control (b), negative control (c), and positive control (d). TABLE 1 | Patients, conventional serology, and xenodiagnosis RESULTS : Optimization of original method for FC-ALTA. (i) Effect of incubation temperature on PPFP values. | In order to evaluate the effect of incubation conditions on the performance of FITC-conjugated anti-human IgG antibody binding, we performed a parallel study in which a 1-h incubation at 4C was compared to a 30-min incubation at 37C. Figure shows representative histogram analyses demonstrating that the latter condition led to higher PPFP values for samples from one NT patient, whereas no change was detected for a sample from a TC patient . Using a range of samples, our data demonstrated that incubation of parasites for 30 min at 37C increases the sensitivity of FC-ALTA (P = 0.0007 for NT patients and P = 0.0001 for TNC patients) with no effect on its ability to discriminate samples with PPFP values of <=20% (P = 0.68) . Therefore, we have adjusted this condition in the original method in order to improve the sensitivity of FC-ALTA. (ii) Comparison of PPFP values between pure and GPS. | With the purpose of fitting the method for field studies, in which storage of frozen samples (FS) may not be possible, we tested the applicability of our method to glycerol-preserved samples (GPS) that can easily be maintained at room temperature. Parallel studies which used 70 pairs of FS and GPS samples were carried out, and the PPFP values for each individual sample are presented in Fig. . Data analysis was performed by grouping the samples based on the PPFP values obtained with FS as negative (PPFP <= 20%), low positive (20% < PPFP <= 50%), or high positive (PPFP > 50%) followed by comparison with the PPFP values obtained with GPS. One out of 29 samples with negative results (PPFP = 20.0%) became low positive (PPFP = 25.0%) when tested as GPS. Two out of 11 samples with low-positive results showed a distinct result when tested as GPS. One became negative (PPFP = 26.0 to 17.0%) and another one became high positive (PPFP = 50.0 to 66.0%). No changes were observed in the performance of FC-ALTA for all 40 samples that showed high-positive results. Statistical analysis did not show any differences between the mean PPFP values obtained from GPS and FS (P = 0.89 for results with PPFP <= 20%, P = 0.66 for results with 20% < PPFP <= 50%, and P = 0.68 for results with PPFP > 50%). Performance of FC-ALTA with treated and nontreated patient samples. (i) PPFP values from NT and TNC patients. | PPFP values from chagasic patients with positive serology and xenodiagnosis are presented in Fig. . Our data demonstrated that all NT patients showed positive PPFP values. Moreover, it was interesting that NT samples were always high positive (PPFP > 50%) . The TNC patients were subdivided into two groups as they received treatment during the acute or chronic phase of the disease . In the same manner, all TNC patients presented PPFP values greater than the 20% cutoff, and then they did not differ from NT patients. However, in this group we found two low-positive samples with PPFP values of >20% and <=50% (samples 38 and 76), suggesting that the therapy was able to decrease parasitemia and also lower antibody titers but not enough to make them free of infection . (ii) PPFP values from TC individuals. | PPFP results for TC individuals are presented in Fig. . The patients were subdivided into three groups as they received treatment during the acute, subacute, or chronic phase of the disease. All but one (sample 17) TC individual had PPFP values lower than 20%. With the aim of further investigating this unexpected result, after decoding the patient identification number, we retested this sample from 1997 along with two other previous samples from 1981 and 1989, taken before treatment and early after treatment, respectively. The results of this analysis confirm the high reactivity of this patient before and at different times after treatment (PPFP = 95.5, 94.3, and 95.1%, respectively). (iii) PPFP values from TUE patients. | PPFP results for TUE individuals are presented in Fig. . The data demonstrated a wide degree of reactivity, including 12 samples classified as high positive (PPFP > 50%), 9 samples classified as low positive (20% < PPFP <= 50%), and 7 samples presenting negative results (PPFP <= 20%) . These seven samples with negative PPFP results are classified as dissociated according to the Krettli and Brener criteria . As previously described , these individuals should be considered cured since they presented negative results for xenodiagnosis and PPFP. Additional analysis demonstrated that all patients with less than 10 years of follow-up after treatment (9 of 9) showed positive PPFP values and 78.0% of them (7 of 9) had high-positive results (PFPF > 50%) . Patients with more than 10 years of follow-up after treatment presented a wide range of results. The frequency of samples with high positive results dropped to 31.5% (6 of 19), and the frequency of low-positive and negative results increased to 31.5% (6 of 19) and 37.0% (7 of 19), respectively. Statistical analysis demonstrated that results for patients in the <=10 years after treatment group differed significantly from those for patients in the >10 years after treatment group (P < 0.05). To further address the differential reactivity of the TUE patient group, we subdivided the patients based on their results from CSA analysis . Then the PPFP values from patients with positive CSA and those with oscillating CSA were compared. The data demonstrated that both subgroups had negative, low-positive, and high-positive results. The percentages of samples with negative, low-positive, and high-positive PPFP values were 36.5% (4 of 11), 36.5% (4 of 11), and 27.0% (3 of 11) for patients with oscillating CSA and 18.0% (3 of 17), 29.0% (5 of 17), and 53.0% (9 of 17) for those with positive CSA, respectively. Interestingly, the frequency of high-positive results was higher within the group with positive CSA (53.0 versus 27.0%). Together, our data support the hypothesis that high-positive results with FC-ALTA are in general associated with positive CSA, regardless of the clinical group analyzed (NT, TNC, or TUE). FIG. 2. | Representative histograms showing effect of incubation temperature on PPFP values. Representative histograms showing effect of incubation temperature on PPFP values. Parasites were incubated in parallel study with a 1:256 dilution of PBS-10% FBS-diluted human sera (from NT and TC patients) followed by a 1-h incubation at 4C (a) or a 30-min incubation at 37C (b) with FITC-conjugated anti-human IgG (whole-molecule) antibody. (c) A range of samples were tested for both protocols (-4C for 60 min and 37C for 30 min [*]), including samples from the NT (n = 12), TNC (n = 9), and TC (n = 4) patient groups. Statistical analysis demonstrated significant differences between PPFP values obtained at 37C for 60 min in comparison to those obtained at 4C for 60 min for the NT and TNC patient groups but not for the TC patient group. FIG. 3. | Comparison of PPFP values between pure and GPS. Comparison of PPFP values between pure and GPS. Parallel studies with pairs of FS (*) and GPS samples were carried out, and the results were presented as PPFP values. Samples were grouped based on the PPFP values obtained with FS followed by comparison with the PPFP values obtained with GPS. No significant differences were observed between mean PPFP values obtained from FS and GPS. FIG. 4. | Analysis of PPFP values from patients with positive xenodiagnosis and positive conventional serology. Analysis of PPFP values from patients with positive xenodiagnosis and positive conventional serology. The performance of FC-ALTA with samples from the NT (a) and TNC (b) patient groups was investigated by using FS, and data analysis was performed by establishing a 20% PPFP value as the cutoff between negative and positive results, as described by Martins-Filho et al. . Positive results were additionally classified according to the method of Cordeiro et al. as low positive (20% < PPFP <= 50%) and high positive (PPFP > 50%). The TNC patients were subdivided into two groups, as they received treatment during the acute (A) or chronic (C) phase of the disease. FIG. 5. | Analysis of PPFP values from patients with negative xenodiagnosis and negative conventional serology. Analysis of PPFP values from patients with negative xenodiagnosis and negative conventional serology. The performance of FC-ALTA for samples from the TC patient group was investigated by using FS, and data analysis was performed by establishing a 20% PPFP value as the cutoff between negative and positive results, as described by Martins-Filho et al. . Positive results were additionally classified according to the method of Cordeiro et al. as low positive (20% < PPFP <= 50%) and high positive (PPFP > 50%). The TC patients were subdivided into three groups, as they received treatment during the acute (A), subacute (SA), or chronic (C) phase of the disease. FIG. 6. | Analysis of PPFP values from patients with negative xenodiagnosis but positive or oscillating conventional serology. Analysis of PPFP values from patients with negative xenodiagnosis but positive or oscillating conventional serology. The performance of FC-ALTA for samples from the TUE patient group was investigated by using FS, and data analysis was performed by establishing a 20% PPFP value as the cutoff between negative and positive results, as described by Martins-Filho et al. . Positive results were additionally classified according to the method of Cordeiro et al. as low positive (20% < PPFP <= 50%) and high positive (PPFP > 50%). The TC patients were subdivided into groups, as they received treatment during the acute (A) or chronic (C) phase of the disease (b), had positive or oscillating serology (a), and had less or more than 10 years of follow-up after treatment (c). DISCUSSION : Different studies performed with either human or murine models have focused on strategies to evaluate the efficacy of specific chemotherapy against Chagas' disease . Conflicting reports have arisen because of differences in the approaches applied as cure criteria . Positive parasitological tests indicate the persistence of circulating parasites and represent a definitive marker of ongoing infection and treatment failure . However, lack of sensitivity is the major feature that leads to a low-negative predictive value of these methods . On the other hand, most types of CSA tests, evaluating the presence of specific antibodies, remain positive for long periods after treatment. Therefore, despite their high sensitivity, CSA tests are not recommended as cure criteria . For this reason, there is an urgent need for the development of serological tests, which avoid the occurrence of positive serology after parasite elimination. One of the strategies is to search for new immunological techniques, which are more reliable, to be used in follow-up studies after treatment. The finding of two types of anti-T. cruzi antibodies (LA and CSA) enabled Krettli and Brener to describe a phenomenon of serology dissociation and postulate that the absence of LA should be considered a cure marker, despite residual positive CSA tests. Considering the technical laboriousness of CoML, we have used an alternative tool (FC-ALTA) in our laboratory that was shown to be comparable to CoML for monitoring the treatment effectiveness in cases of human Chagas' disease . In fact, the sensitivity of FC-ALTA was much higher than that of CoML, as shown by the optimal dilutions required, 1:256 and 1:4, respectively. Furthermore, the capacity of flow cytometry to count thousands of trypomastigotes per assay improved the data confidence. Like CoML, the original FC-ALTA method allows for distinguishing sera from NT and TNC patients from both TC and noninfected individuals. However, the original FC-ALTA let two out of nine TNC patients with positive CoML results fall below the 20% cutoff level. For this reason, in the present investigation we designed an optimization of the original method in order to improve its performance and sensitivity. Our data demonstrated that incubation with FITC-conjugated anti-human IgG antibody for 30 min at 37C led to higher PPFP values for samples from chagasic patients, with no changes for samples from cured individuals. Therefore, we have incorporated this adjustment in the original method for further studies. Another great advance achieved by this study was the perspective of fitting the method for field surveys since the PPFP values were not adversely affected when testing GPS. Parallel analysis of FS and GPS demonstrated that only 3 out of 70 samples showed discrepancies in PPFP values. From those, only two samples interchanged between negative and low-positive results within PPFP intervals of 20% +- 10%. Then we recommended that when testing GPS, the interpretation of PPFP values should take this variation into account. In the study described here, the major goal of the research was to conduct a double-blind investigation to reinforce the use of FC-ALTA for clinical survey. For this purpose, we tested coded serum samples from four different categories of chagasic patients, including TC, TNC, TUE, and NT patients. Our data confirmed, in another study population, our previous statement that patients with ongoing infection (NT and TNC), in this case identified by positive xenodiagnosis, had positive results with FC-ALTA. It was interesting that the optimization of the original method, as described above, held all TNC patient samples with PPFP values higher than 20%, improving the performance of the original technique, in which 2 out of 9 TNC patient samples fell bellow the cutoff level. In our pioneer study, we suggested that those low PPFP values should be the result of antibody shedding from the parasite membrane and that shorter incubation times at a lower temperature should resolve this question. Here, we confirm that lower incubation time but higher temperature was the optimal condition to improve the sensitivity of FA-ALTA with no effect on its ability to discriminate negative samples. Together, the data from NT and TNC patient samples reemphasizes that the persistence of circulating parasites is associated with the presence of ALTA, which can be used as a valid immunological marker for active infection. Furthermore, it should be stressed that, in general, patients with negative CSA and negative xenodiagnosis persistently had PPFP values under 20%. Unexpectedly, we found that one cured patient had PPFP values over 50% in sequential samples and repeated tests. Interestingly, we found that this patient had been treated for tuberculosis 10 years before inclusion in this study. This patient had more than 140 negative xenodiagnoses and persistent negative CSA results over the past 10 years (data not shown). Several studies have reported the possibility of cross-reactivity in the serology for Chagas' disease in patients with tuberculosis . We are now testing to see if samples from patients with tuberculosis may present cross-reactivity in FC-ALTA assay. Considering these observations, we suggest the association of FC-ALTA with other tests already indicated as cure criteria to evaluate therapeutic efficacy in Chagas' disease. Parasitological and CSA tests should be performed repeatedly before a single FC-ALTA. Individuals presenting negative results on conventional tests should be considered cured, without requirement of confirmation by FC-ALTA. However, the FC-ALTA should be indicated, to confirm cure, in cases of individuals with several negative parasitological tests and positive or oscillating CSA tests (TUE patients), since this approach could save several years of inconclusive diagnosis. In fact, we have found that seven samples within the TUE patient population had negative PPFP results and could be classified as dissociated according to the Krettli and Brener criteria . As previously described , these individuals should be considered cured since they had negative results for xenodiagnosis and PPFP. We therefore postulated that patients who have been treated and show negative parasitological tests and the absence of ALTA should be considered cured despite residual positive or oscillating CSA. Here, using a 1:256 serum dilution from a TUE patient, we have observed that a period of 10 years after treatment was required to identify negative results with FC-ALTA for patients treated during the chronic phase of the disease. Then, we hypothesized that FC-ALTA follow-up with serial dilution could identify the antibody clearance sooner by the shift of positive PPFP values to lower titers. This hypothesis is currently under investigation in our laboratory. Finally, our study optimized and validated our method, offering another tool, less laborious than CoML and parasitological tests, for investigating and monitoring results of specific chemotherapy. Furthermore, these data reinforce the relevance of studies regarding etiological treatment of chronic patients since a considerable number of successful treatments were achieved during the chronic phase of Chagas' disease. Backmatter: PMID- 12204965 TI - Gamma Interferon Inhibits Production of Anti-OspA Borreliacidal Antibody In Vitro AB - The ability of a Lyme borreliosis vaccine to induce and maintain sustained levels of borreliacidal antibody is necessary for prolonged protection against infection with Borrelia burgdorferi. Vaccination against infection with B. burgdorferi could be improved by determining the mechanism(s) that influences the production of protective borreliacidal antibody. Borreliacidal antibody was inhibited in cultures of lymph node cells obtained from C3H/HeJ mice vaccinated with formalin-inactivated B. burgdorferi and cultured with macrophages and B. burgdorferi and treated with recombinant gamma interferon (rIFN-gamma). The suppression of production of outer surface protein A (OspA) borreliacidal antibody by rIFN-gamma was not affected by the time of treatment. In addition, treatment with rIFN-gamma inhibited the production of other anti-B. burgdorferi antibodies. By contrast, treatment of cultures of immune lymph node cells with anti-IFN-gamma marginally increased the production of borreliacidal antibody and enhanced the production of other antibodies directed against B. burgdorferi. These results show that IFN-gamma does not play a major role in the production of anti-OspA borreliacidal antibody. Additional studies are needed to determine which cytokine(s) will enhance production of borreliacidal antibody. Keywords: Introduction : Public health concerns about the morbidity associated with Lyme borreliosis have prompted the development of several vaccines to prevent infection with Borrelia burgdorferi . The efficacies of the vaccines are based on the ability of a major outer surface protein (Osp) of B. burgdorferi, specifically OspA, to induce antibody that can kill the Lyme spirochete in vaccinees or ticks . In an extensive field trial involving 10,936 participants, the vaccine was 76% effective in preventing infection with B. burgdorferi after three inoculations . However, the duration of protection had not been determined . In addition, we showed previously that vaccination of humans with recombinant OspA (rOspA) induced only low levels of anti-OspA borreliacidal antibody and that the borreliacidal response waned rapidly . Only one individual had detectable anti-OspA borreliacidal antibody after 180 days. A similar anti-OspA borreliacidal antibody response was detected in hamsters vaccinated with rOspA . The poor antibody response induced by vaccination may have contributed to the withdrawal of the vaccine. The ability of rOspA or other protective immunogens to induce high and sustained levels of borreliacidal antibody is necessary to ensure prolonged protection against infection with B. burgdorferi. Recently, we developed an in vitro system to determine the effects that immunologic mediators have on the production and regulation of borreliacidal antibody. Anti-OspA borreliacidal antibody was readily produced when lymph node cells obtained from vaccinated mice were cultured with macrophages and B. burgdorferi . When OspA borreliacidal antibody-producing cells were exposed to a known B-lymphocyte-stimulating factor , interleukin 4 (IL-4), borreliacidal-antibody production was inhibited. Furthermore, treatment of the immune lymph cell cultures with anti-murine IL-4 did not alter the production of anti-OspA borreliacidal antibody. These results suggested that IL-4 plays a minor role in the production and up-regulation of borreliacidal antibody. The inability of IL-4-stimulated immune lymph node cells to increase production of borreliacidal antibody may be due to down-regulation of gamma interferon (IFN-gamma). It is known that IL-4 strongly down-regulates functions promoted by IFN-gamma , especially class switching to immunoglobulin G2a (IgG2a) by B lymphocytes . Since B. burgdorferi organisms are killed by IgG2a and complement , we sought evidence of whether IFN-gamma augments anti-OspA borreliacidal-antibody production. Such information could provide insight into the mechanism of borreliacidal-antibody production and contribute to the development of a more efficacious Lyme borreliosis vaccine. MATERIALS AND METHODS : Mice. | Eight- to 12-week-old inbred C3H/HeJ mice were obtained from our breeding colony located at the Wisconsin State Laboratory of Hygiene. Mice weighing 20 to 40 g were housed four per cage at an ambient temperature of 21C. Food and acidified water were provided ad libitum. Organism. | B. burgdorferi sensu stricto isolate 297 was originally isolated from human spinal fluid . Low-passage (<6) organisms were cultured once in modified Barbour-Stoenner-Kelly (BSK) medium containing screened lots of bovine serum albumin to a concentration of 5 x 107 spirochetes per ml. Five-hundred-microliter samples were then dispensed into 1.5-ml screw-cap tubes (Sarstedt, Newton, N.C.) containing 500 mul of BSK supplemented with 10% glycerol (Sigma Chemical Co., St. Louis, Mo.), sealed, and stored at -70C. When necessary, a frozen suspension of spirochetes was thawed and used to inoculate fresh BSK medium. The spirochetes were viewed by dark-field microscopy and enumerated using a Petroff-Hausser counting chamber. Preparation of vaccine. | B. burgdorferi organisms were grown in 1 liter of BSK medium for 6 days, pelleted by centrifugation (10,000 x g; 15C; 10 min), and washed three times with phosphate-buffered saline (PBS; pH 7.4). The washed pellet was resuspended in 1% formalin, incubated at 32C for 30 min with periodic mixing, washed three times by centrifugation with PBS (12,000 x g; 10C; 15 min), and resuspended in PBS. Subsequently, the formalin-inactivated spirochetes were mixed in a volume of a 1% suspension of aluminum hydroxide (alum; Reheis, Berkeley Heights, N.J.) to yield 4 x 106 spirochetes/ml. Vaccination of mice. | We showed previously that mice vaccinated with whole cells of B. burgdorferi with or without alum yielded higher levels of anti-OspA borreliacidal antibody than those vaccinated with rOspA in the presence or absence of alum. Therefore, whole cells of B. burgdorferi were used in this investigation. Whole cells are not recommended as a vaccine for human usage. The ability of whole cells to consistently induce anti-OspA borreliacidal antibody in mice permits evaluation of cytokine mechanisms responsible for control of anti-OspA borreliacidal ability. Sixty mice were mildly anesthetized with methoxyflurane contained in a mouth-and-nose cup and vaccinated subcutaneously in the inguinal region with 0.25 ml (similar106B. burgdorferi organisms) of the formalin-inactivated vaccine preparation. The suspension contained approximately 100 mug of borrelial protein. Sham-vaccinated mice were injected with BSK medium or alum alone. Recovery of macrophages. | Five to 10 mice per experimental protocol were mildly anesthetized with methoxyflurane contained in a mouth-and-nose cup and injected intraperitoneally with 2 ml of 3% 3-week-old thioglycolate in PBS. Four days after injection, the mice were euthanized by CO2 asphyxiation, and 8 ml of cold Hanks' balanced salt solution (Sigma) was injected intraperitoneally. The peritoneal cavity was massaged for similar1 min, and the exudate cells were recovered by aspiration with a syringe. The suspension of peritoneal exudate cells was centrifuged at 1,500 rpm (IEC Centra-7) for 10 min at 4C. The supernatant was decanted, and the cells were resuspended in Dulbecco's modified Eagle's medium (DMEM; Sigma) that was free of antimicrobial agents but supplemented with 10% heat-inactivated (56C; 45 min) fetal bovine serum (HyClone Laboratories, Logan, Utah), 5 x 10-5 M 2-mercaptoethanol (Sigma), and l-glutamine (2.92 mg/ml; Sigma). Aliquots of the cell suspension were then poured over polystyrene tissue culture dishes (100 by 20 mm; Corning Glass Works, Corning, N.Y.) and incubated at 37C in a humidified atmosphere of 5.0% CO2 for 4 to 6 h. After incubation, the nonadherent cells were aspirated from the tissue culture dishes. The dishes were gently rinsed twice with 8-ml portions of warm Hanks' balanced salt solution to further eliminate nonadherent cells. Five milliliters of cold, nonenzymatic cell lifter (Sigma) was then added to each tissue culture dish and incubated at 4C for 30 min. The macrophages were detached from the surfaces of the dishes by vigorously tapping and gently scraping the inside of the tissue culture dishes with a sterile rubber policeman. Suspensions of macrophages from several tissue culture dishes were aspirated, pooled, and centrifuged at 1,500 rpm for 10 min at 4C. The supernatant was decanted, and the pellet was resuspended in 1 ml of DMEM. Cell viability was determined by trypan blue exclusion. The preparations of macrophages obtained by this method were 98% free of lymphocyte contamination. Isolation of lymph node cells. | Mice were euthanized by CO2 inhalation 17 days after vaccination with formalin-inactivated B. burgdorferi. The inguinal lymph nodes were removed from vaccinated and nonvaccinated mice and placed in cold DMEM. Single-cell suspensions of lymph node cells were prepared by teasing apart the lymph nodes with forceps and gently pressing them through a sterile stainless steel 60-mesh screen into antimicrobial-free cold DMEM supplemented with 10% heat-inactivated fetal bovine serum, l-glutamine, and 2-mercaptoethanol. The lymph node cells were washed twice by centrifugation (1,500 rpm; 4C; 10 min) with DMEM. The supernatants were decanted, and the pellets were resuspended in 1 ml of cold DMEM. Cell viability was assessed by trypan blue exclusion. Production of antibody in vitro. | Sterile six-well flat-bottom tissue culture dishes (Becton Dickinson, Lincoln Park, N.J.) were inoculated with lymph node cells (5 x 106) obtained from vaccinated or nonvaccinated mice, macrophages (105), and 106 live B. burgdorferi organisms. DMEM was added to the suspensions of cells to bring the final volume to 3 ml. On days 3, 6, 9, 12, and 15 after cultivation at 37C in the presence of 5.0% CO2, 1.0-ml samples of the supernatants were removed after gentle agitation and replaced with equal volumes of warm DMEM. In some experiments, rIFN-gamma at quantities ranging from 0.1 to 10 mug or rat anti-murine IFN-gamma at quantities ranging from 25 to 75 mug (R&D Systems, Minneapolis, Minn.) was added to cultures of immune lymph node cells, macrophages, and B. burgdorferi at 10 min and 2 and 4 days of incubation. In other experiments, rat anti-murine CD119 at quantities ranging from 1 to 100 mug (PharMingen, San Diego, Calif.) was added to cultures of immune lymph node cells, macrophages, and B. burgdorferi at 10 min of incubation. In similar fashion, control cultures were treated with a rat isotype-nonspecific antibody. Supernatants were collected after centrifugation at 13,000 rpm for 8 min to remove spirochetes and other cellular debris. Supernatants were stored at -70C until they were used. Detection of borreliacidal antibody by membrane filtration. | The frozen supernatants were thawed, heat inactivated (56C, 30 min), sterilized with a 0.22-mum-pore-size filter (Acrodisk; Gelman Sciences, Ann Arbor, Mich.), and serially twofold diluted (from undiluted to 1:8,192) with fresh BSK medium. One hundred-microliter aliquots of each dilution were transferred to 1.5-ml screw-cap tubes (Sarstedt), and 100 mul of BSK containing 104B. burgdorferi organisms per ml was added along with 20 mul of sterile guinea pig complement (Sigma). The tubes were then gently shaken and incubated for 3 days at 32C. Controls included filter-sterilized supernatants obtained from suspensions of nonimmune lymph node cells with macrophages and B. burgdorferi. Other controls included supernatants from nonimmune lymph node cells, macrophages alone, and DMEM. After incubation, 100 mul of each suspension was removed and placed in individual 1.5-ml screw-cap tubes (Sarstedt). Subsequently, 100 mul of a solution of propidium iodide (1.0 mg/ml; Molecular Probes, Eugene, Oreg.) diluted 1:20 in sterile PBS was added. The suspensions were briefly mixed before being incubated at 56C for 30 min to permit intercalation of propidium iodide into the spirochetes. One hundred microliters of each sample was then filtered through 0.22-mum-pore-size Nuclepore polycarbonate membrane filters (47-mm diameter; Whatman Nuclepore, Clifton, N.J.) under negative pressure with a single-place sterility test manifold (Millipore Corporation, Bedford, Mass.) attached to a vacuum pump. The membrane filters were washed with similar8 ml of sterile double-distilled H2O (ddH2O), removed from the vacuum apparatus, allowed to dry, and placed onto glass microscope slides. Coverslips were placed on the filters before they were viewed with a Laborlux S fluorescence microscope (Leitz, Wetzlar, Germany) using an x50 oil immersion objective. The number of spirochetes on each filter was determined by viewing similar30 fields. The borreliacidal-antibody titer was defined as the reciprocal of the dilution preceding the dilution at which the number of spirochetes or clumping was equal to that of the control. Generally, individual spirochetes with a few clumps were uniformly distributed throughout the fields on filters of the control supernatants. Western immunoblotting. | B. burgdorferi 297 organisms were grown in 1 liter of BSK medium for 6 days, pelleted by centrifugation (10,000 x g; 15C; 10 min), and washed three times with PBS at pH 7.4. The washed pellet was resuspended in 1% formalin and incubated at 32C for 30 min with periodic mixing and then washed three times by centrifugation with PBS (12,000 x g; 10C; 15 min) and resuspended in PBS. The spirochetes were suspended in sodium dodecyl sulfate-polyacrylamide gel electrophoresis sample buffer and boiled for similar3 min. One hundred twenty micrograms of B. burgdorferi lysate was loaded onto a preparative 12% acrylamide gel, and the proteins were resolved by overnight electrophoresis at similar7 mA constant current with the buffer system of Laemmli . The proteins were transferred onto a nitrocellulose membrane for 1 h at 20 V, using a semidry blotting apparatus (Bio-Rad Laboratories, Hercules, Calif.). The nitrocellulose membrane was incubated overnight at 4C in 5% milk dissolved in Tris-buffered saline with 0.05% Tween 20 (TBS-T; pH 7.4) to block nonspecific reactivity, washed two times each with TBS-T and ddH2O, allowed to dry, and finally cut into strips. The strips were washed three times with TBS and subsequently incubated for 1 h with a 1:1,000 dilution of an alkaline phosphatase-conjugated goat anti-murine IgG (heavy and light chain specific; Kirkegaard & Perry Laboratories, Gaithersburg, Md.) in 5% milk in TBS-T. This was followed by four washes with TBS. The strips were developed by the addition of 5-bromo-4-chloro-3-indolylphosphate-nitroblue tetrazolium substrate (Kirkegaard & Perry). The reactions were stopped after 2 min with several large volumes of chilled ddH2O. Flow cytometric analysis of immune supernatants. | Suspensions of immune lymph node cells containing macrophages and B. burgdorferi in the presence or absence of IFN-gamma, anti-IFN-gamma, or isotype-nonspecific antibody were analyzed for the numbers of cells, B lymphocytes, and T lymphocytes on day 4 of incubation by using flow cytometry. Briefly, suspensions (1 ml) of immune lymph node cells were placed in chilled centrifuge tubes, and the total number of lymphocytes was determined. The suspensions of cells were then mixed with both phycoerythrin-conjugated anti-murine CD3 (5 mul of a 1:5 dilution; PharMingen) and fluorescein isothiocyanate-conjugated anti-murine CD45R/B220 (10 mul of a 1:50 dilution; PharMingen) and incubated at 4C for 15 min. The cells were washed twice by centrifugation with PBS containing 0.1% bovine serum albumin (1,500 rpm; 4C; 10 min). The pellets of cells were resuspended in 250 mul of cold DMEM and kept in the dark at 4C until they were analyzed by flow cytometry. One hundred microliters of 50-mug/ml propidium iodide (Sigma) was then added to each tube just prior to acquisition by the flow cytometer to discriminate viable and nonviable cells. Data were acquired on a FACSCalibur flow cytometer (Becton Dickinson, San Jose, Calif.) using CellQuest acquisition and analysis software (Becton Dickinson). Twenty thousand events were detected by forward and side angle light scatter and by propidium iodide, phycoerythrin, and fluorescein isothiocyanate fluorescence. A dot blot profile of forward angle light scatter and propidium iodide fluorescence enabled identification and gating of live lymphocyte populations. The gated events were subsequently analyzed by quadrant dot blots of phycoerythrin and fluorescein isothiocyanate fluorescence for enumeration of CD3+ and B220+ lymphocytes in a given sample. Statistical analysis. | A t test was used to determine significant differences in the titers of borreliacidal antibody and in lymphocyte counts among supernatants. In addition, titers that were determined during kinetic studies of in vitro borreliacidal antibody production were tested by analysis of variance, utilizing the Minitab statistical analysis program. The Fisher least-significant-difference test was used to examine pairs of means when a significant F ratio indicated reliable mean differences. The alpha level was set at 0.05 before the experiments were started. RESULTS : Modulation of in vitro production of anti-OspA borreliacidal antibody by IFN-gamma. | We showed previously that anti-OspA borreliacidal antibody is readily detected in supernatants of lymph node cells obtained from mice vaccinated with formalin-treated B. burgdorferi in adjuvant and cultured with macrophages and B. burgdorferi for several days. The data presented in Fig. confirms and extends these findings. High levels of anti-OspA borreliacidal antibody (titer, 256) were detected in supernatants on day 6 of culture of immune lymph node cells with macrophages and B. burgdorferi. The peak anti-OspA borreliacidal activity (titer, 1,024) was detected with supernatants obtained on days 9 to 15 of culture. When cultures of immune lymph node cells with macrophages and B. burgdorferi were exposed to 0.1, 1.0, or 10 mug of rIFN-gamma, the levels of anti-OspA borreliacidal antibody were significantly reduced 16-fold or more. Maximum suppression of anti-OspA borreliacidal antibody occurred in cultures of lymph node cells exposed to 0.1 mug of rIFN-gamma followed by 1.0 and 10 mug of rIFN-gamma. The borreliacidal activity was due to the production of anti-OspA antibody. Adsorption of the supernatant from immune lymph node cells with rOspA reduced the borreliacidal-antibody titer. No borreliacidal antibody was detected in supernatants obtained from nonimmune lymph node cells cultured with macrophages and B. burgdorferi. When these studies were repeated three times, similar results were obtained. In other studies, immune lymph node cells were treated with 25, 50, or 75 mug of anti-murine IFN-gamma. Anti-OspA borreliacidal-antibody production was increased two- to fourfold in supernatants obtained from immune lymph node cell cultures treated with 50 or 75 mug of anti-IFN-gamma for 9 days or more compared to the levels of anti-OspA borreliacidal antibody detected in cultures of untreated immune lymph node cells . However, treatment of immune lymph node cells with 25 mug of anti-IFN-gamma did not affect the production of borreliacidal antibody. No borreliacidal antibody was detected in supernatants obtained from nonimmune lymph node cells cultured with macrophages and B. burgdorferi. Similar results were obtained when these studies were repeated. Temporal effects of rIFN-gamma and anti-IFN-gamma on the production of borreliacidal antibody. | Immune lymph node cell cultures were treated with rIFN-gamma or anti-IFN-gamma at 10 min or 2 or 4 days after cultivation. Production of anti-OspA borreliacidal antibody was inhibited by rIFN-gamma at all treatment intervals, although less effect was detected with rIFN-gamma treatment on day 4 of cultivation . Treatment of immune lymph node cell cultures with anti-IFN-gamma slightly enhanced production of anti-OspA borreliacidal antibody at all treatment intervals . When these studies were repeated, similar results were obtained. Effects of rIFN-gamma and anti-IFN-gamma on production of anti-B. burgdorferi antibody. | Figure shows that supernatant obtained from immune lymph node cells (day 17 after vaccination) cultured with macrophages and B. burgdorferi for 9 days produced predominately anti-OspA (31-kDa) antibody. Treatment of the lymph node cell cultures with rIFN-gamma abrogated the antibody response . By contrast, treatment of immune lymph node cell cultures with anti-IFN-gamma induced production of antibodies to multiple B. burgdorferi antigens . Furthermore, supernatant from anti-IFN-gamma-treated lymph node cell cultures had an anti-OspA Western immunoblot titer of 40,960 compared to a titer of 2,560 for supernatant obtained from untreated immune lymph node cell cultures. Flow cytometric analysis of lymph node cell cultures. | Immune lymph node cells cultured with macrophages and B. burgdorferi in the presence or absence of rIFN-gamma or anti-IFN-gamma for 4 days were analyzed for viability and numbers of T and B lymphocytes by using flow cytometry. Figure shows that the number of viable cells in lymph node cell cultures treated with rIFN-gamma decreased approximately 25% from the number of viable cells detected in cultures of lymph node cells treated with an isotype-nonspecific antibody or anti-IFN-gamma . In addition, the number of B lymphocytes (B220 marker) decreased approximately 50% in lymph node cell cultures treated with rIFN-gamma compared to the control . By contrast, a 10% increase in B lymphocytes was detected in lymph node cell cultures treated with anti-IFN-gamma . Similarly, the number of T lymphocytes (CD3 marker) decreased approximately 25% in cultures of lymph node cells treated with rIFN-gamma , while a 6% decrease was detected in cultures treated with anti-IFN-gamma compared to control cultures . When these studies were repeated, similar results were obtained. Effect of rIFN-gamma on macrophages. | In other studies, the lymph node cell cultures treated with an isotype-nonspecific antibody (control), rIFN-gamma, or anti-IFN-gamma were gently washed to remove nonadherent cells and viewed by light microscopy. The macrophages cultured with rIFN-gamma were rounded and lacked pseudopodia , whereas macrophages treated with the isotype-nonspecific antibody or anti-IFN-gamma were spindle shaped with long pseudopodia . FIG. 1. | Effect of 0.1 , 1 , or 10 mug (*) of rIFN-gamma on production of borreliacidal antibody by lymph node cells obtained from 17-day-vaccinated mice cultured with macrophages and B. burgdorferi Effect of 0.1 , 1 , or 10 mug (*) of rIFN-gamma on production of borreliacidal antibody by lymph node cells obtained from 17-day-vaccinated mice cultured with macrophages and B. burgdorferi for 3, 6, 9, 12, and 15 days. Controls included lymph node cells obtained from vaccinated and nonvaccinated mice cultured with macrophages and B. burgdorferi. The error bars indicate standard deviations. FIG. 2. | Effect of 25 , 50 , or 75 mug (*) of anti-IFN-gamma on production of borreliacidal antibody by lymph node cells obtained from 17-day-vaccinated mice cultured with macrophages and B. burgdorferi Effect of 25 , 50 , or 75 mug (*) of anti-IFN-gamma on production of borreliacidal antibody by lymph node cells obtained from 17-day-vaccinated mice cultured with macrophages and B. burgdorferi for 3, 6, 9, 12, and 15 days. Controls included lymph node cells obtained from vaccinated and nonvaccinated mice cultured with macrophages and B. burgdorferi. The error bars indicate standard deviations. FIG. 3. | Borreliacidal-antibody response of lymph node cells obtained from 17-day-vaccinated mice cultured with B. burgdorferi Borreliacidal-antibody response of lymph node cells obtained from 17-day-vaccinated mice cultured with B. burgdorferi and macrophages and treated with rIFN-gamma (A) or anti-IFN-gamma (B) 10 min (large-grid bars) or 2 (small-grid bars) or 4 (cross-hatched bars) days after cultivation. Control cultures (solid bars) were treated with PBS (A) or an isotype-nonspecific antibody (B). The error bars indicate standard deviations. FIG. 4. | Western immunoblots obtained with supernatants from cultures of immune lymph node cells (day 17 after vaccination) cultured with macrophages and B. burgdorferi Western immunoblots obtained with supernatants from cultures of immune lymph node cells (day 17 after vaccination) cultured with macrophages and B. burgdorferi without treatment (A) or treated with IFN-gamma (B) or anti-IFN-gamma (C). The supernatants were diluted 1:160 (lanes I) or 1:640 (lanes II). FIG. 5. | Flow cytometric analysis of lymph node cells (day 17 after vaccination) cultured with macrophages and B. burgdorferi Flow cytometric analysis of lymph node cells (day 17 after vaccination) cultured with macrophages and B. burgdorferi in the presence of isotype-nonspecific antibody (A), rIFN-gamma (B), or anti-IFN-gamma (C) for 4 days. The cells were labeled with both phycoerythrin-conjugated anti-murine CD3 (T cells) and fluorescein isothiocyanate-conjugated anti-murine CD45R/B220 (B cells). Viability was determined by exclusion of propidium iodide. FIG. 6. | Macrophages (adherent cells) detected in cultures of lymph node cells cultured with macrophages and B. burgdorferi Macrophages (adherent cells) detected in cultures of lymph node cells cultured with macrophages and B. burgdorferi and treated with isotype-nonspecific antibody (A) or rIFN-gamma (B) for 4 days after removal of nonadherent cells. Macrophages detected in cultures of lymph node cells treated with anti-IFN-gamma were similar to those pictured in panel A. DISCUSSION : B. burgdorferi organisms are easily killed when they bind specific antibody that activates complement . Several outer surface proteins of B. burgdorferi, including OspA, OspB, OspC, and the 39-kDa protein (, , -, ), induce complement-dependent borreliacidal antibody. This feature makes these outer surface proteins ideal candidates for development of a vaccine to prevent infection with B. burgdorferi. Of these proteins, OspA has been used to vaccinate humans and animals . Although rOspA is immunogenic, the antibody response is dominated by production of antibody that is not protective . Most importantly, vaccinated animals challenged with B. burgdorferi during periods with concomitant high levels of nonbactericidal anti-OspA and low levels of anti-OspA borreliacidal antibodies develop arthritis . Recently, we showed that hamsters vaccinated with 30, 60, or 120 mug of rOspA with or without alum developed severe destructive arthritis when challenged with B. burgdorferi . Similarly, hamsters vaccinated with a commercially available canine rOspA vaccine developed severe destructive arthritis after challenge with the Lyme spirochete . Only the presence of high levels of borreliacidal antibody prevented infection with B. burgdorferi from inducing arthritis in vaccinated animals . Maximum protection due to borreliacidal antibody, however, is restricted to 7 to 9 weeks after vaccination. One approach to increase the duration of protection and lessen the potential for adverse effects is to determine the immune mediators responsible for the production and maintenance of borreliacidal antibody. We showed previously that IL-4 does not augment the production of anti-OspA borreliacidal antibody, even though IL-4 is known to up-regulate B-lymphocyte growth and differentiation . In fact, IL-4 inhibited the production of anti-OspA borreliacidal antibody, including IgG2a, by lymph node cells obtained from vaccinated mice . It is known that IgG2a is negatively regulated by IL-4 but up-regulated by IFN-gamma . Our data and those of others suggested that IL-4 counteracted the effects of IFN-gamma on the production of borreliacidal antibody. When lymph node cells producing borreliacidal antibody were exposed to rIFN-gamma, borreliacidal antibody production was inhibited. The time of exposure (10 min to 4 days) of immune lymph node cells to rIFN-gamma did not affect the inhibition of production of borreliacidal antibody. In addition, treatment with rIFN-gamma inhibited the production of other anti-B. burgdorferi antibodies. The suppression of borreliacidal and other anti-B. burgdorferi antibodies by rIFN-gamma was unexpected. Flow cytometric analysis of rIFN-gamma-treated and untreated immune lymph node cells showed that treatment with rIFN-gamma reduced the number of viable lymphocytes, especially B lymphocytes. In addition, macrophages cultured in the presence of exogenous rIFN-gamma were rounded and exhibited rare pseudopodia. By contrast, macrophages obtained from cultures of untreated lymph node cells producing borreliacidal antibody were spindle shaped with many pseudopodia. IFN-gamma has been shown to induce apoptosis in T and B lymphocytes. These events could have affected the activation of macrophages and their ability to process borrelial antigen. It is possible that inhibition of borreliacidal antibody was due to the use of toxic concentrations of rIFN-gamma. However, when more physiologically relevant concentrations (1.0 or 0.1 mug) of rIFN-gamma were added to immune lymph node cells, borreliacidal activity also failed to increase. In addition, immune lymph node cells were treated with anti-CD119, which blocks the binding of murine IFN-gamma to cellular receptors . No significant decrease in borreliacidal-antibody production was detected in these cultures compared to immune lymph node cells treated with an isotype-nonspecific antibody (data not shown). These results suggest that IFN-gamma is not a major force in driving the production of borreliacidal antibody. Perhaps the most compelling evidence that IFN-gamma is not responsible for production of borreliacidal antibody came from experiments utilizing neutralizing antibody to IFN-gamma. Treatment of immune lymph node cells with various concentrations of anti-IFN-gamma failed to suppress borreliacidal activity. In fact, borreliacidal antibody production was marginally enhanced, especially when treatment with anti-IFN-gamma occurred early in cultivation of the immune lymph node cells. Furthermore, treatment with anti-IFN-gamma resulted in polyclonal expansion of the anti-B. burgdorferi antibody response. In support of this idea, by using flow cytometric analysis, we also detected a significant increase in the number of B lymphocytes in cultures of immune lymph node cells treated with anti-IFN-gamma. IFN-gamma performs numerous immunologic functions, including T helper lymphocyte differentiation and stabilization , enhancement of major histocompatibility complex expression on both B lymphocytes and macrophages , antiviral effects , and amelioration of production of IgG2a . Our results, however, show that IFN-gamma suppresses antibody production, including IgG2a borreliacidal antibody . An explanation may be that IgG2a antibody expression is not completely dependent upon IFN-gamma . In addition, IFN-gamma may prevent the expression of IL-4 functions . However, we showed previously that IL-4 suppressed the production of borreliacidal antibody . This suggests that other cytokines, but not IL-4 or IFN-gamma, may be responsible for the induction, production, and maintenance of borreliacidal antibody. In support of this idea, treatment with anti-IFN-gamma augmented not only borreliacidal antibody but other antibodies directed against B. burgdorferi. Additional studies are needed to determine which cytokines are responsible for the production of borreliacidal antibody. In conclusion, we showed that IFN-gamma plays a major role in suppression of the production of borreliacidal antibody. Effective neutralization of endogenous IFN-gamma slightly augmented the production of borreliacidal antibody and expanded the anti-B. burgdorferi antibody responses. Determination of the mechanism that inhibits the production of borreliacidal antibody by rIFN-gamma may lead to the development of a safe and effective Lyme borreliosis vaccine. Backmatter: PMID- 12204945 TI - Active Surveillance for Scrapie by Third Eyelid Biopsy and Genetic Susceptibility Testing of Flocks of Sheep in Wyoming AB - Control of scrapie, an ovine transmissible spongiform encephalopathy or prion disorder, has been hampered by the lack of conventional antemortem diagnostic tests. Currently, scrapie is diagnosed by postmortem examination of the brain and lymphoid tissues for PrPSc, the protein marker for this group of disorders. For live, asymptomatic sheep, diagnosis using tonsil or third-eyelid lymphoid tissue biopsy and PrPSc assay has been described. To evaluate the feasibility and efficacy of third-eyelid testing for identification of infected flocks and individual infected sheep, 690 sheep from 22 flocks were sampled by third-eyelid lymphoid tissue biopsy and immunohistochemistry. Sheep were further evaluated for relative genetic susceptibility and potential contact exposure to scrapie. Third-eyelid testing yielded suitable samples for 80% of the sheep tested, with a mean of 18.1 lymphoid follicles (germinal centers) per histologic section. Three hundred eleven of the sheep were sampled through passive surveillance programs, in which only sheep with potential contact with an infected sheep at a lambing event were tested, regardless of their scrapie susceptibility genotype. In addition, 141 genetically susceptible sheep with no record of contact with an infected animal at a lambing event were sampled through a targeted active surveillance program. Ten PrPSc-positive sheep were identified through the passive surveillance program, and an additional three PrPSc-positive sheep, including two from flocks with no history of scrapie, were identified through the active surveillance program. All PrPSc-positive sheep had the highly susceptible PrP genotype. Third-eyelid testing is a useful adjunct to flock monitoring programs, slaughter surveillance, and mandatory disease reporting in a comprehensive scrapie eradication and research program. Keywords: Introduction : Scrapie is a fatal neurologic disease of sheep and goats, introduced into the United States in 1947 and now endemic in many states . Scrapie is a member of the heterogeneous group of prion diseases, or transmissible spongiform encephalopathies (TSE), which includes bovine spongiform encephalopathy (BSE), cervid chronic wasting disease, and transmissible mink encephalopathy. Although scrapie is not a zoonotic disease, the apparent transmission of BSE to humans in the United Kingdom has resulted in a call for eradication of all TSE in food-producing animals. The transmissible agent in scrapie remains controversial because no conventional bacterial, viral, fungal, or toxic source has been identified. The major component in infectious tissue homogenates is a relatively denaturation-resistant conformational isoform (PrPSc) of the highly conserved mammalian PrP protein (PrPC) . The prion model proposes that PrPSc and other less well characterized cofactors induce conversion of PrPC to the pathogenic isoform PrPSc through aggregation and posttranslational structural changes. Regardless of the etiology of scrapie, PrPSc is a reliable marker for infection. Scrapie transmission is thought to occur by oral exposure to the causative agent, uptake through the intestinal wall, amplification of the agent in the gut-associated lymphoid and nervous tissue , and transport to the brain through the autonomic fibers of the vagus nerve . In U.S. sheep, PrPSc detected by immunoassay or a transmissible agent detected by rodent bioassay is found in some lymphoid tissue by the age of 14 months and in the brain by 2.5 years, approximately 6 months before onset of clinical signs. Early accumulation of PrPSc in lymphoid tissue is the basis for preclinical scrapie diagnosis by immunohistochemistry (IHC) assay of postmortem tonsil and antemortem third-eyelid lymphoid tissue or tonsil biopsy samples . Disease transmission depends on both exposure to the infectious agent and susceptible genetics in the host. Sheep housed with an infected postparturient ewe (within 60 days of lambing or abortion) are considered to be at increased risk of scrapie, presumably because they are exposed to the infectious agent in placental and fetal tissues or fluids . Following contact with an infected postparturient ewe, disease occurs almost exclusively in sheep with particular polymorphisms in the host PrP gene . In the U.S. sheep population, scrapie (whether associated with clinical disease or with PrPSc accumulation in the absence of clinical signs) has been confirmed only in sheep homozygous for the PrP allele encoding glutamine at codon 171 (QQ171) , regardless of breed. In Europe and the United Kingdom, relative susceptibility of sheep with several other PrP genotypes is reported in various breeds, a finding useful in national scrapie control programs . Scrapie is typically diagnosed by IHC assay of brain tissue from sheep with clinical signs of scrapie, notably weight loss, wool loss, and ataxia. Following diagnosis of an index animal, control measures in the flock include culling or quarantine of sheep at risk due to susceptible genetics and/or possible contact with the infected postparturient ewe . However, many infected sheep die from intercurrent diseases or routine culling without diagnostic testing, and identification of infected flocks by surveillance for clinical scrapie alone may not be sufficient for disease eradication. Cooperative state-federal programs that meet or exceed the current control measures and that are designed to test or improve program procedures or facilitate research are allowed as pilot programs . The state of Wyoming has proposed a pilot program in which passive surveillance (quarantine, testing, or removal of sheep having potential contact with an infected ewe) is supplemented with active surveillance, including live-animal testing of genetically susceptible, clinically normal sheep with no reported contact with an infected sheep during a lambing event. The objective of this study was to investigate the feasibility and efficacy of including third-eyelid testing and susceptibility genetics in active and passive surveillance programs. MATERIALS AND METHODS : Study flocks. | Twenty-two flocks in Wyoming were included in this study. Ninety sheep from the University of Wyoming Animal Science Department were sampled by third-eyelid biopsy to estimate the efficiency of eyelid biopsy sampling (percentage of sheep with suitable biopsy samples) for sheep of various ages and breeds. The sheep were aged 14 months to 7 years and represented four breeds (Suffolk, Columbia, Hampshire, and Rambouillet) . In addition, 600 sheep from privately owned flocks in Wyoming were tested . These sheep were of Suffolk and Hampshire origin. All sheep over the age of 14 months at the time of testing were sampled in flocks 1 to 5 and 7 to 13. Only sheep over the age of 14 months and considered at risk due to potential contact with a postparturient ewe subsequently diagnosed with scrapie (see "Relative risk designation" below) were sampled in flocks 6 and 14 to 16. The group designated flock 16 represents one sheep with potential contact risk from each of five flocks and two sheep with potential contact risk from a sixth flock. Flocks were included in the study following identification of an infected animal or an animal with potential contact exposure, either born into the flock (flocks 3 to 5) or purchased (flocks 6 to 16). Flocks 1 and 2 were volunteered by the producers for educational purposes or due to industry concerns about scrapie. Restraint. | One eye on each sheep, usually the left eye, was pretreated with topical 0.5% sterile ophthalmic proparacaine HCl (Bausch and Lomb, Tampa, Fla.) several minutes before the procedure was performed. The sheep was haltered and restrained manually. All sheep were sampled on the farm of origin, and equipment for restraint of the sheep varied among the facilities. Many farms were equipped with a fitting stand, in which the head was immobilized by resting the chin on a horizontal support and securing a restraint strap behind the head. Alternative restraint devices included a tilt table in which the sheep was restrained in a squeeze chute, or a head catch in which partial restraint was afforded by a stanchion. In all cases, immobilization of the head was achieved through traction on the halter strap. Optimally, one operator restrained the sheep's body, one operator maintained control of the halter, one person retracted the third eyelid, and a fourth person collected the biopsy specimen. If only two operators were available, equipment suitable for restraint of the sheep was needed; one operator maintained control of the halter and retracted the third eyelid, and one operator collected the sample. Biopsy sample collection. | After adequate immobilization of the sheep, the third eyelid was retracted by using 1 by 2 toothed forceps (Econo Thumb Tissue forceps 21-762; Sklar Instruments, West Chester, Pa.) and treated again with proparacaine. The lymphoid tissue was visualized on the bulbar surface as a slightly raised pink tissue on either side of the midline (Fig. , lower arrow). The color, location, and size of the tissue varied slightly among individual sheep. Sheep were sampled even if there was no gross evidence of lymphoid tissue. The tissue was grasped with a second set of forceps, raised slightly from the underlying connective tissue, and excised with single-use curved Metzenbaum scissors (Econo Metzenbaum scissors 21-336; Sklar Instruments) by using a medial to lateral cut; typically, two or three sequential cuts with the scissors were required to acquire a single biopsy specimen of sufficient size containing the entire area of lymphoid tissue on one side of the midline (Fig. , upper arrow). The biopsy sample was fixed in 10% neutral buffered formalin for at least 24 h. Tissue processing. | Typically, fixed biopsy samples were observed to be folded in half with the lymphoid tissue facing outward, secured by the indentation formed by the forceps teeth. Biopsy samples were opened using iris scissors and placed flat between biopsy sponges in plastic histopathology cassettes. The lymphoid tissue was placed down in the cassettes. Cassettes with properly positioned biopsy samples were immersed in 95 to 98% formic acid for 1 h to decontaminate the tissue . Samples were rinsed in three changes of deionized water and for at least 10 min in running deionized water, then reequilibrated in 10% neutral buffered formalin. Tissues were processed conventionally and embedded in paraffin; care was taken to position the biopsy samples in the paraffin blocks so that the lymphoid tissue was facing and on the same plane as the surface of the block, to provide the largest possible section when cut. Adhesion of formic acid-decontaminated tissue to glass slides required additional attention during preparation. Three- to five-micrometer-thick sections were cut with a microtome and applied to adhesive-coated aminosilane-treated slides (Newcomer Supply, Middleton, Wis.). Residual moisture under the section was removed with absorbent laboratory tissue paper, and slides were positioned vertically to dry for at least 1 h. Slides were baked at 57C for approximately 15 h (overnight) to remove excess paraffin. Immunostaining. | Sections were rehydrated through xylene substitutes and graded alcohols, then treated with 95% formic acid for 5 min and neutralized by three changes of 0.1 M Tris-HCl, pH 7.6. Antigen retrieval was performed by incubation at 120C in a medical pressure cooker (BioCare Medical, Walnut Creek, Calif.) for 20 min in a modified citrate buffer (Target Retrieval solution, pH 6.1; Dako Corporation, Carpinteria, Calif.). IHC was performed on an automated immunostainer (NexES; Ventana Medical Systems, Tucson, Ariz.) by using the manufacturer's detection reagents (antibody diluent, AEC detection kit, hematoxylin, bluing, and mouse immunoglobulin G1 [IgG1] negative control antibody). The primary antibody was a cocktail of two monoclonal antibodies (MAbs) to separate epitopes on the PrP molecule (MAb F99/97.6.1 and MAb F89/160.1.5 ; Veterinary Medical Research Development, Inc., Pullman, Wash.) at 5 mug/ml each in antibody diluent (Ventana Medical Systems) filtered through a 0.22-mum-pore-size syringe top filter. The primary antibody was applied for the maximum time available with the automated system (32 min), followed by biotinylated goat anti-mouse IgG, streptavidin-conjugated horseradish peroxidase, and 3-amino-9-ethylcarbazole as a chromogen. The default settings for incubation time and temperature provided by the manufacturer were used for all detection steps. The slides were counterstained with hematoxylin, rinsed, and covered with coverslips by use of a compatible mounting medium (Faramount aqueous mounting medium; Dako). Lymphoid tissue (tonsil or retropharyngeal lymph node) from an infected sheep was used as a positive control. Negative controls included replacement of the primary anti-PrP antibodies with an isotype control antibody (Ventana) and assay of tissues from uninfected sheep. Lymphoid follicles appearing as discrete oval or round areas with a darker border were counted on each section. Sections were considered "PrPSc detected" if a granular red precipitate was noted in lymphoid follicles (germinal centers) , regardless of the number of follicles in the section. Sections were considered "no PrPSc detected" if no precipitate was noted in a section. A minimum of six lymphoid follicles per section were needed for this designation. Sections with no immunostaining and fewer than six lymphoid follicles were scored "insufficient lymphoid tissue for determination." All IHC assays were scored by trained observers with no knowledge of the potential contact or genetic susceptibility status of the sheep. Relative risk designation. | Sheep were scored for potential contact risk by using producer records and the definitions established in the Code of Federal Regulations 79.4 . Briefly, sheep were defined as high risk by potential contact if they were born in a flock during the same lambing season as a lambing event by a sheep subsequently diagnosed with scrapie. Lambing events include the birth of an infected sheep as well as parturition or abortion by an infected ewe in any year of her life. Sheep were considered to be at lower risk by contact if they were born on the farm at other times, in different lambing facilities on the same farm, or on farms without reported cases of scrapie. If an infected ewe had produced at least one lamb at the facility and no birth records were available for the flock, all sheep born in the flock during the years that the infected ewe lambed or aborted were designated high risk. Relative contact risk status was based on producer records, and no outside measure of the accuracy of those records is available. No producer could rule out a previously undetected scrapie case, because some of the clinical signs (particularly weight loss) occur frequently in these flocks and result in culling of the animal without diagnostic testing. Sheep were scored for relative genetic susceptibility based on the deduced amino acid sequence at residue 171. Blood was collected by jugular venipuncture into EDTA-treated vacuum tubes and shipped to a commercial laboratory for scrapie susceptibility testing (GeneCheck, Fort Collins, Colo.). DNA was extracted from the buffy coat, amplified by PCR, and analyzed by a DNA mismatch binding assay. The diploid genotype was reported as QQ, QR, or RR. However, this assay does not distinguish between sequences encoding glutamine (Q) and sequences encoding histidine (H) (both are reported as 171Q), and samples scored QQ may actually be QQ, QH, or HH. Samples scored QR may be QR or HR. Samples scored RR are unequivocally homozygous for arginine (R) at position 171. Sheep were considered to have high genetic susceptibility if their genotype was 171QQ, 171QH, or 171HH and to have lower genetic susceptibility if their genotype was 171QR, 171HR, or 171RR. Surveillance programs. | Under the current program, passive surveillance includes testing of all sheep in flocks in which an infected animal was born or produced a lamb and all sheep sold from those flocks if they were born in a year in which the infected ewe produced a lamb. The passive surveillance program therefore includes sheep with potential contact with the transmissible agent shed at parturition, even if those sheep are at low genetic risk. Active surveillance includes only genetically susceptible sheep with no reported contact risk. In some flocks, all sheep over the age of 14 months were sampled, even if they were not eligible for either program. The number of sheep sampled in each surveillance program (or sampled as whole-flock testing but enrolled in neither program) is shown in Table . FIG. 1. | Bulbar surface, nictitating membrane of a Suffolk sheep. Bulbar surface, nictitating membrane of a Suffolk sheep. The lymphoid tissue is typically visualized as a slightly raised, red to pink area (lower arrow). Complete excision of the lymphoid tissue (upper arrow) is visible approximately 2 months after biopsy. The lymphoid tissue at this site does not regenerate. FIG. 2. | IHC detection of PrPSc in the nictitating membrane lymphoid tissue. IHC detection of PrPSc in the nictitating membrane lymphoid tissue. A third-eyelid biopsy specimen from a live sheep with preclinical scrapie is shown. Lymphoid follicles (discrete round to oval accumulations of lymphocytes, macrophages, and dendritic cells with a dark border) show granular PrPSc immunoreactivity (red) when immunostained with a cocktail of MAbs F89/160.1. and F99/97.6.1. Bar, 200 mum. TABLE 1 | Suitability of eyelid biopsy samples from sheep of four breeds, aged 1 to 7 years TABLE 2 | Flocks and sheep sampled by third-eyelid biopsy and PrP genotyping RESULTS : Third-eyelid biopsy samples were collected from a university-owned flock to determine if particular breeds of sheep or age groups were unsuitable for this procedure. Ninety sheep of four breeds were available for sampling . Suitable samples were those with at least six lymphoid follicles per section. Nine sheep were sampled twice because the left eye did not appear to have adequate lymphoid tissue when examined grossly. Five of these nine animals had suitable biopsy specimens from both eyes. One animal, a 5-year-old Columbia, had unsuitable samples from both eyelids. Three animals had one suitable sample and one unsuitable sample; this group included a 3-year-old Hampshire, a 4-year-old Columbia, and a 2-year-old Suffolk. The overall efficiency of testing was 76% (68 of 90) when the results from either eyelid were used to score an individual sheep. There was no breed for which all samples were unsuitable. There was a trend toward a lower percentage of suitable samples from sheep older than 3 years, but a larger study population is needed before conclusions on the feasibility of testing older sheep can be drawn. A total of 600 sheep from 21 producer-owned flocks were assessed for evidence of infection with the scrapie agent by IHC of third-eyelid biopsy samples and for relative genetic susceptibility to scrapie by DNA genotyping. Sampled sheep were over the age of 14 months and were primarily of the Suffolk or Hampshire breed. Sheep were housed in flocks with no reported cases of scrapie, flocks in which an infected ewe was born or had produced at least one lamb, and flocks with at least one purchased animal with potential contact exposure in the flock of origin . The PrP genotypic ratio in this sample was 0.48:0.52 (290 sheep with the highly susceptible 171QQ, -QH, or -HH genotype and 310 sheep with the lower-susceptibility 171QR, -RH, or -RR genotype). This ratio is not significantly different from that reported for an independent sample of 1,000 U.S. Suffolk sheep in 1996 . Three hundred eleven of the 600 sheep in the study were tested under passive surveillance programs , representing sheep with potential contact risk from exposure to an infected ewe at parturition. Eighty-five of 311 had the highly susceptible genotype. One hundred forty-one of the 600 sheep were tested under the Wyoming active surveillance program and represent genetically susceptible sheep with no record of contact with an infected sheep at a lambing event, whether they resided in flocks with a purchased animal with potential contact risk at a different facility (n = 96) or in flocks with no history of scrapie contact in either native or purchased animals (n = 45). One hundred forty-eight sheep in the lower-genetic-susceptibility group and with no reported contact with an infected postparturient ewe were tested because they were housed with sheep in the active surveillance trial. Sheep were tested for PrPSc, a marker for scrapie, by third-eyelid lymphoid tissue biopsy and IHC analysis . Biopsy samples of the third eyelid from 481 (80%) of the sheep tested contained at least six lymphoid follicles and were considered suitable for diagnosis, although all biopsy samples were analyzed by IHC. The mean follicle count in suitable biopsy samples was 18.1 (standard deviation, 13.3). PrPSc was detected in 13 of the biopsy samples, collected from sheep on seven farms . All 13 sheep with PrPSc detected in lymphoid tissue had the highly susceptible PrP 171 genotype QQ, QH, or HH. Of the 13 sheep with positive eyelid biopsy samples, 4 were removed for immediate necropsy, and scrapie was confirmed by IHC of brain and/or lymphoid tissues. Nine sheep were moved to quarantine. Scrapie has been confirmed in seven of these nine sheep by postmortem analysis following development of clinical signs and in one of these nine sheep following acute loss (found dead with no clinical signs). The last of the nine sheep brought to quarantine is asymptomatic at the age of approximately 36 months. Ten of the 13 PrPSc-positive sheep were identified by the passive surveillance program, and 3 were identified by the active surveillance program. The positive samples were collected from sheep in flocks in which an infected sheep had been born (n = 7), flocks with no reported cases of scrapie (n = 2), and flocks with at least one sheep purchased from an infected flock (n = 4), although in one case, the PrPSc-positive sheep was not the purchased sheep with contact risk but an unrelated animal purchased from a different flock. The efficiency of the surveillance programs was estimated as the number of PrPSc-positive sheep per number of sheep in each program (10 of 311, or 3.2%, in the passive surveillance program and 3 of 141, or 2.1%, in the active surveillance program). TABLE 3 | Flock status at the time of testing and risk status of PrPSc-positive sheep DISCUSSION : Scrapie was reported as a clinical entity nearly 300 years ago and was introduced into the United States in 1947. In the United States, control programs have relied on total or partial flock depopulation following diagnosis of the disease in a sheep with clinical signs of scrapie. The insidious nature of the disease, the long preclinical incubation period, the lack of a preclinical test, loss of infected sheep from intercurrent disease before development of clinical signs, and underreporting of scrapie due to the negative economic effect of a diagnosis have contributed to the failure of these programs to control the prevalence of scrapie, which is estimated at 0.07% in the United States (based on a mailed producer survey conducted in 1996 by the National Animal Health Monitoring System of the U.S. Department of Agriculture (USDA) Animal and Plant Health Inspection Service). State and federal flock certification programs, in which producers of monitored flocks agree to purchase breeding stock from similarly monitored flocks, permanent identification of individual sheep and/or flock of origin, and restrictions on interstate movement of suspect and high-risk animals have been in place since 1992. The USDA Animal and Plant Health Inspection Service recently expanded the program to include identification of breeding and mature sheep and breeding goats in interstate commerce, uniform standards for state scrapie control programs, indemnity payments, slaughter surveillance, and third-eyelid testing of exposed animals and/or flocks. These programs are being implemented to eradicate scrapie from the United States. The development of the third-eyelid test for antemortem diagnosis of scrapie and the identification of PrP genotypes associated with clinical and subclinical PrPSc accumulation have provided additional means for controlling scrapie in the U.S. sheep population. The estimated specificity of the third-eyelid test approaches 100% . The estimated sensitivity of the test is 85 to 90%. The false-negative results are due to a number of factors. Some infected sheep have PrPSc detectable in the brain but not in lymphoid tissues; in other cases, PrPSc is detectable in tonsil but not in antemortem eyelid biopsy specimens, due in part to inadequate biopsy sample size. Optimal sampling conditions include a restraint device such as a tilt table, fitting stand, or head catch and adequately trained personnel for immobilization of the head, retraction of the eyelid, and biopsy sample collection. Laboratory handling of the tissue to maximize the number of follicles available for inspection in a single histologic section requires laboratory personnel trained in recognizing the lymphoid side of the tissue and embedding a flat sample in the plane of the microtome cut. In other trials, techniques such as embedding strips of lymphoid tissue at right angles to the plane of the microtome cut or embedding folded or rolled tissues typically resulted in small numbers of follicles available for inspection in each section. In the present study, sampling was performed by one of three veterinarians (J.V.D., J.R.L., and E.S.W.), and IHC of samples from producer flocks was performed by one person (A.K.A.). When sampling, tissue handling, and assay were performed under these circumstances, approximately 80% of the sheep in each flock could be assessed. In this study, the third-eyelid test identified PrPSc-positive sheep in flocks with previous scrapie exposure, although in two cases the individual sheep identified by eyelid biopsy had no reported contact with an infected ewe at parturition. The efficiency of the programs (2 to 3% in this relatively small sample) exceeds the estimated scrapie prevalence level of 0.07% nationally and illustrates the potential value of preclinical surveillance. However, there are some limitations to the use of the eyelid test in a scrapie eradication program. One hundred to 150 sheep can be sampled per day in a single location with proficient personnel using restraint equipment available on most facilities; testing small numbers of sheep at multiple locations decreases the efficiency of the testing personnel. The costs of personnel, equipment, topical anesthetics, and sample testing are borne by the state and federal governments through a joint scrapie eradication program. Although the procedure is performed under local anesthesia, third-eyelid sampling may be perceived as objectionable by some producers. The third-eyelid test is most useful in animals between the ages of 14 and 36 months, because older animals tend to have smaller areas of lymphoid tissue. Further, the estimated test sensitivity is approximately 85 to 90%. Therefore, the third-eyelid test is not suitable as a stand-alone eradication tool. The test is more appropriately considered an adjunct surveillance tool to identify infected flocks which could then be managed through an integrated program of testing, necropsy or quarantine, genetic selection, and husbandry modifications. The test is also useful for identification of naturally infected sheep for transfer to research facilities. Although active surveillance is expensive, cost-effectiveness is increased by targeting the testing program to 14- to 24-month-old sheep of heavily impacted breeds in flocks with a history of purchasing breeding ewes from multiple sources. Active surveillance is a justifiable addition to live-animal passive surveillance, in which flocks are identified only after submission of a clinical suspect. One of the sheep quarantined after selection by this program was found dead with no history of clinical signs, and another remains clinically normal at the age of 36 months. Inapparently infected sheep such as these produce lambs annually and serve as a continuing source of the transmissible agent. Slaughter surveillance and active live-animal surveillance programs such as the Wyoming pilot program will be critical for identification of infected flocks and eventual disease eradication. This study demonstrated that identification of scrapie-infected flocks by third-eyelid testing is technically feasible with proficient personnel and adequate field and laboratory resources. Third-eyelid testing and genotype analysis are useful in identifying infected flocks for regulatory intervention and infected animals suitable for research purposes. Active surveillance for scrapie is a useful component of an integrated scrapie control program based on permanent identification, flock monitoring, slaughter surveillance, and passive surveillance by diagnosis of clinical suspects with testing of their scrapie-exposed flockmates. Backmatter: PMID- 12204969 TI - Immunoblot Analysis of the Humoral Immune Response to Leishmania donovani Polypeptides in Cases of Human Visceral Leishmaniasis: Its Usefulness in Prognosis AB - Sera from Indian patients with parasitologically confirmed visceral leishmaniasis were studied by immunoblot analysis in order to identify a specific pattern for Leishmania infection. A soluble extract of Leishmania donovani was used as antigen. At diagnosis the sera from patients with visceral leishmaniasis specifically recognized fractions represented by bands of 201 kDa (50% of serum samples), 193 kDa (60%), 147 kDa (50%), 120 kDa (60%), 100 kDa (50%), 80 kDa (80%), 70 kDa (70%), 65 kDa (100%), 50 kDa (50%), 36 kDa (50%), 20 kDa (70%), and 18 kDa (50%). The 65-kDa band, common to all patients infected with Leishmania parasites, was found at the time of diagnosis. However, the immunoblot pattern changed after patients were treated and cured with sodium antimony gluconate (SAG; n =10) or miltefosine (n =10), as was evident from blots of sera obtained pretreatment and at 1, 3, and 6 months posttreatment. At 6 months posttreatment, immunoblots of sera from patients on the SAG regimen showed the disappearance of all bands except the 70-kDa band. Similarly, sera from those on the miltefosine regimen showed the disappearance of all bands except the 65- and 70-kDa bands. This study shows that Western blot analysis is a sensitive test for detection of anti-Leishmania antibodies. Moreover, the persistence of reactivity with the 65- and 70-kDa bands in the sera of all groups shows its promise as a diagnostic and prognostic tool. Keywords: Introduction : Human visceral leishmaniasis is caused by a protozoan parasite of the Leishmania donovani complex, namely, Leishmania donovani. This infection is characterized clinically by fever, hepatosplenomegaly, anemia, and weight loss. A marked hypergammaglobulinemia and absence of detectable cell-mediated immunity are the principal immunological features of the disease. Although patients generate high levels of nonspecific immunoglobulin as well as specific antileishmanial antibodies, they fail to respond to parasite antigen in the delayed-type hypersensitivity skin test . A kinesin-related protein-encoding gene has been discovered in Leishmania chagasi and contains a repetitive 117-bp sequence encoding 39 amino acid residues (K39) conserved at the C-terminal end in all of the visceral leishmaniasis-causing isolates examined so far . The recombinant product of K39 (rK39) has proved to be a very sensitive and specific antigen in an enzyme-linked immunosorbent assay (ELISA) for the serodiagnosis of visceral leishmaniasis from the foci of endemicity in Brazil, China, Pakistan, and Sudan . Kumar et al. reported extremely high levels of anti-rK39 antibodies in patients with visceral leishmaniasis, suggesting the application of rK39 for a sensitive and specific means of serodiagnosis and the potential of rK39 ELISA in the monitoring of drug therapy and the detection of disease relapses. To date, several L. donovani-specific proteins, including species-specific and subspecies-specific molecules, have been identified with monoclonal antibodies or polyvalent antisera. Several of these proteins elicit a strong humoral immune response during visceral leishmaniasis, allowing the specific diagnosis of visceral leishmaniasis by a competitive ELISA or a direct dot blot assay . Immunoblots have been useful in studies of host serological responses during infection, and a few antigens with potential diagnostic value have been selected . The humoral immune response occurs during the active phase of infection, with the appearance of low titers of antibodies that disappear some months after the end of treatment representing a temporary response . This fact has made the assessment of antibody titers a useful tool for patient follow-up after therapy and for defining the period during which clinicians should observe patients in order to make a prognosis. In the present study, we have examined patients' antibody responses at the time of diagnosis and various stages posttreatment using the immunoblot technique with soluble L. donovani antigen. We had the double objective of identifying a particular band pattern present in the patients affected by visceral leishmaniasis (confirmed by the presence of parasites in a sample taken from the spleen) and describing any possible variation in this band pattern following antileishmanial therapy with sodium antimony gluconate (SAG) or miltefosine. MATERIALS AND METHODS : Antigen. | L. donovani (MHOM/IN/96/B.H.U.70) promastigotes were cultivated in tissue culture flasks with RPMI 1640 medium (Hi-Media, Mumbai, India) supplemented with 10% fetal calf serum (Gibco, Grand Island, N.Y.) and antibiotics (gentamicin) . Parasites were taken at the late-logarithmic phase of growth, washed five times at 4C with sterile phosphate-buffered saline (PBS), and centrifuged at 1,400 x g for 15 min. The parasite pellet was resuspended in 1 ml of PBS, and the mixture was immediately frozen at -70C. In order to make up only one batch of antigen for Western blot analysis, all parasites were kept frozen at this temperature until there were sufficient parasites from which soluble antigen could be obtained. To prepare the soluble antigen, a method described by Isaza et al. was used, with slight modifications. Briefly, the parasites were defrosted and resuspended in 2 ml of lysis buffer (20 mM Tris HCl [pH 7.4] containing 40 mM NaCl, 10 mM EDTA, 2 mM phenylmethylsufonyl fluoride [BDH, Mumbai, India], and 0.4% sodium dodecyl sulfate [BDH]) . The mixture was left on ice for 30 min, with vortex agitation every 10 min. It was then centrifuged at 6,000 x g for 20 min at 4C. The supernatant was removed and kept at -70C until use. A small sample was used for protein determination by a method modified from that of Lowry et al. . By this method the final antigen protein concentration was found to be 9.4 mg/ml. Human sera. | Serum samples were collected from patients with parasitologically confirmed visceral leishmaniasis (kala azar) (L. donovani body score in splenic aspirate of 2+ to 4+, i.e., between >1 to 10 parasites/100 field and >1 to 10 parasites/field) at the time of diagnosis. Sera were obtained by venipuncture from patients and controls registered at Kala-azar Medical Research Centre, Muzaffarpur, India, and Sir Sundar Lal Hospital, Banaras Hindu University, Varanasi, India. Blood was allowed to coagulate at room temperature and was then centrifuged at 1,400 x g for 5 min. All sera were stored at -70C until required. Western blot analysis. | SDS-polyacrylamide gel electrophoresis was done with a vertical (Bangalore Genei, Peenya Bangalore, India) gel apparatus. The antigen was boiled for 5 min in sample buffer (two times) and was immediately subjected to electrophoresis in an SDS-10% polyacrylamide gel containing 0.1% SDS as described by Laemmli . The slab gel was run with two lanes per comb: a 100-mm lane for the parasite antigen sample and a 7-mm lane for a wide-range molecular mass marker (kind gift of David Sacks, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Md.). Three hundred micrograms of protein was used for gel electrophoresis. The gels were run at 15 mA in the stacking gel and 30 mA in the resolving gel until the bromophenol blue dye migrated 1 cm from the bottom of the gel in Tris-glycine-SDS buffer (pH 8.3) . Immunoblotting. | Transfer of polypeptides from SDS-polyacrylamide gels onto nitrocellulose membrane (pore size, 0.45 mum; kind gift of R. T. Kenney, Center for Biologics Evaluation and Research, Food and Drug Administration, Rockville, Md.) was done in a transblotting chamber at 4 V/cm and 4C for overnight in 25 mM Tris-192 mM glycine-20% (vol/vol) methanol . Following the blotting, the membranes were stained with 0.5% ponceau red for 5 min and were destained with distilled water. The blotted gels were stained with Coomassie brilliant blue in order to make sure that all polypeptides had been completely transferred. Detection of antibodies from patients' sera bound to antigens was done with a 5-bromo-4-chloro-3-indolylphosphate-Nitro Blue Tetrazolium system (BCIP-NBT; Kirkegaard & Perry Laboratories, Gaithersburg, Md.). Four-millimeter-wide strips were cut from the previously blotted membrane and blocked in 5% bovine serum albumin-0.3% Tween 20 in 100 mM Tris-buffered saline (TBS; pH 7.4) for an hour at room temperature. After six washes of 15 min each in TBS-0.1% Tween 20, the strips were incubated with serum (diluted 1:100 in 1% bovine serum albumin-TBS) for 2 h at 37C with constant stirring. After incubation with the primary antibody, the strips were washed four times for 15 min each time with Tween 20-TBS. After the last wash, immune complexes were detected with a 1:500 dilution of alkaline phosphatase (Bangalore Genei) conjugated with goat anti-human immunoglobulin G (IgG), and the mixture was incubated with constant shaking for 2 h at room temperature. After incubation the membrane was washed with Tween 20-TBS three or four times and then washed with TBS only to remove the Tween 20. The membranes were developed with the BCIP-NBT system (Kirkegaard & Perry Laboratories) for 5 min. The reaction was stopped with PBS-EDTA (20 mM EDTA). Statistical analysis. | Analysis of the corresponding polypeptides in the blotted membranes was done by using an exponential regression curve with the molecular mass markers of each membrane, calculating the relative mobility (Rf) of each protein , and looking for its corresponding molecular mass. RESULTS : Study population. | The clinical and laboratory features of the visceral leishmaniasis patients on admission are summarized in Table . The response to SAG or miltefosine therapy was prompt, and the splenic aspirate smears of all patients showed no parasites (L. donovani bodies) at the end of treatment (the patients were parasitologically cured). These patients were subsequently monitored. Pretreatment sera. | In general, the sera from patients (n = 10) with visceral leishmaniasis specifically recognized polypeptides ranging from 18 to more than 201 kDa, with the frequently occurring bands being of 201 kDa (50% of serum samples), 193 kDa (60%), 147 kDa (50%), 120 kDa (60%), 100 kDa (50%), 80 kDa (80%), 70 kDa (70%), 65 kDa (100%), 50 kDa (50%), 36 kDa (50%), 20 kDa (70%), and 18 kDa (50%). We found one band of 65 kDa in sera from 100% of patients with active disease . Immunoblot evaluation. (i) SAG treatment group. | We studied the sera of patients (n = 10) who were being treated with SAG. We did antibody tests at regular intervals for up to 6 months posttreatment. All patients were cured after the end of treatment, as confirmed by laboratory and parasitological tests . At 1 month after treatment, the bands of 201, 193, 100, and 36 kDa disappeared. At 3 and 6 months posttreatment all bands except the 70-kDa band disappeared . (ii) Miltefosine treatment group. | We also studied patients (n = 10) who were being treated with the oral drug miltefosine. We compared pretreatment sera with sera obtained at 1, 3, and 6 months posttreatment. At 1 month posttreatment only the 20- and 80-kDa bands had disappeared. At 3 months posttreatment all bands except the bands at 36, 65, and 70 kDa had disappeared; and at 6 months posttreatment all bands except the 65- and 70-kDa bands had disappeared. The 65- and 70-kDa bands remained at all time points. Sera from all patients had antibodies to the 65- and 70-kDa bands in common at pretreatment and 1, 3, and 6 months (follow up) posttreatment . Interestingly, all patients in both the SAG and miltefosine treatment groups showed the exact same response pattern after treatment and follow-up. FIG. 1. | Immunoblotting with L. donovani Immunoblotting with L. donovani soluble antigen with the alkaline phosphatase system. VL lanes 1 to 10, sera from patients with visceral leishmaniasis at the time of diagnosis; EC lanes 1 to 3, sera from controls from an area of endemicity; and NEC lanes 1 to 3, sera from controls from an area of nonendemicity. FIG. 2. | Immunoblotting with soluble L. donovani Immunoblotting with soluble L. donovani antigen with the alkaline phosphatase system. Lane 1, serum from patients at the time of diagnosis; lanes 2, 3, and 4, patterns obtained for all patients (n = 10) with serum taken on days 30, 90, and 180, respectively, after treatment with SAG. FIG. 3. | Immunoblotting with soluble L. donovani Immunoblotting with soluble L. donovani antigen with the alkaline phosphatase system from a patient with visceral leishmaniasis. Lane 1, serum from patients at the time of diagnosis; lanes 2, 3, and 4, the patterns obtained for all patients (n = 10) with serum taken on days 30, 90, and 180, respectively, after treatment with miltefosine. TABLE 1 | Clinical and laboratory features of patients with visceral leishmaniasis DISCUSSION : Parasitological methods have always been considered first-choice procedures for the diagnosis and prognosis of leishmaniasis due to their 100% specificities, although their sensitivities may vary depending upon the experience of the diagnostic team. However, in some instances it is very difficult to demonstrate the presence of the parasite, such as in patients with mucocutaneous leishmaniasis and visceral leishmaniasis. In these patients, serodiagnosis becomes an important alternative for demonstrating the presence of the parasite . Specific serological techniques are based on the determination of antibodies produced against the circulating Leishmania-specific antigens. Both clinical and epidemiological data provide presumptive evidence in the diagnosis of visceral leishmaniasis. However, on the basis of these same criteria, other diseases must be considered in the differential diagnosis; among these are malaria, typhoid fever, tuberculosis, and leukemias. Definitive diagnosis is based on demonstration of the parasites, which may be found in aspirates of the spleen (98% of the aspirates are positive), (sternal) bone marrow (85% of the aspirates are positive), and liver (60% of the aspirates are positive). Splenic aspiration is the most reliable procedure, but it is a high-risk procedure. Because in many cases demonstration of parasites is difficult, serological tests have provided a useful alternative. Among the most recently described methods, the Western blot technique is highly sensitive and specific and provides broader information about the parasite's antigenic profile . In the present study, a 65-kDa antigenic component was recognized by 100% of serum specimens from patients with clinically and pasitologically confirmed visceral leishmaniasis. It was never identified in the control sera tested (100% specificity). There was little variation in the patterns of the bands recognized by each of the serum specimens . Comparing our results with those of other groups, we found differences in the band patterns recognized by sera from patients with visceral leishmaniasis and cutaneous leishmaniasis using several Leishmania isolates as antigen. The studies of Dos Santos et al. showed reactivities with the 119- and 123-kDa bands when an extract of L. infantum was used as the antigen with sera from patients with visceral leishmaniasis. In additon, Rolland-Burger et al. found that the sera of patients with visceral leishmaniasis reacted to a 119-kDa band when L. infantum was used as the antigen. If antibodies remain in the patient for years and the patient's serum has a Western blot pattern similar to the one obtained at the time of diagnosis, then serious doubts arise as to whether a patient is ever completely cured (i.e., the patient has no remaining parasites) or whether treatment has induced a chronic subpatent infection, leaving the patient susceptible to relapse later on; for example, immunosuppression can lead to the reemergence of active infection, as in the case of visceral leishmaniasis in patients with AIDS . A test as sensitive as Western blotting may be able to identify immunosuppressed patients who could be at risk of a relapse of leishmaniasis. The only difference between the pre- and posttreatment sera of the groups treated with SAG and miltefosine was the disappearance of all protein bands except those of 65 and 70 kDa, respectively. We could detect a 65-kDa polypeptide in 100% of the pretreatment sera from kala azar patients . Vinayak et al. reported on a 63- to 68-kDa complex which is an immunodominent surface molecule of L. donovani that can be used as an immunoprophylactic agent. Investigations have indicated that major surface glycoproteins with molecular masses ranging from 63 to 68 kDa (known as the glycopeptide 63 [gp63] complex) are important ligands in cell-to-cell interactions and cell infectivity . Vinayak et al. have shown that promastigotes of L. donovani use the gp66 molecule not only for identification but also for attachment to macrophages. Sera from patients with confirmed leishmaniasis have been shown to identify the gp63 complex molecule in Western immunoblots of Leishmania crude antigen . Further work is needed to define the possible value of the 65-kDa band as a prognostic indicator. We suggest that the band of 65 kDa could be useful for diagnosis of Indian visceral leishmaniasis. Furthermore, the results of our studies suggest that Western blotting could be a sensitive technique for differentiation between the acute and chronic stages of disease. The presence of bands of 201, 193, 147, 100, 80, 70, 65, 50, 36, 20, and 18 kDa (Fig. and ) and the subsequent disappearance of all bands except those of 65 and 70 kDa during the course of treatment suggest the usefulness of Western blotting in making a clear-cut prognosis. The most interesting observation was the disappearance of protein bands unique to both the SAG and the miltefosine treatment groups, as the sera of all patients in the trial groups showed the same band patterns at the respective time points. This suggests that the type of drug used for treatment probably has a bearing on the immune response. Our data also suggest that Western blotting could successfully be used when there is no splenomegaly. Western blotting could also be a prognostic tool, as the disappearance or reappearance of specific bands could predict cure or relapse. Moreover, Western blotting analysis showed a large diversity in the antibody response to visceral leishmaniasis and permitted detection of antibodies to Leishmania in serum. Backmatter: PMID- 12204946 TI - Augmentation of the Lipopolysaccharide-Neutralizing Activities of Human Cathelicidin CAP18/LL-37-Derived Antimicrobial Peptides by Replacement with Hydrophobic and Cationic Amino Acid Residues AB - Mammalian myeloid and epithelial cells express various peptide antibiotics (such as defensins and cathelicidins) that contribute to the innate host defense against invading microorganisms. Among these peptides, human cathelicidin CAP18/LL-37 (L1 to S37) possesses not only potent antibacterial activity against gram-positive and gram-negative bacteria but also the ability to bind to gram-negative lipopolysaccharide (LPS) and neutralize its biological activities. In this study, to develop peptide derivatives with improved LPS-neutralizing activities, we utilized an 18-mer peptide (K15 to V32) of LL-37 as a template and evaluated the activities of modified peptides by using the CD14+ murine macrophage cell line RAW 264.7 and the murine endotoxin shock model. By replacement of E16 and K25 with two L residues, the hydrophobicity of the peptide (18-mer LL) was increased, and by further replacement of Q22, D26, and N30 with three K residues, the cationicity of the peptide (18-mer LLKKK) was enhanced. Among peptide derivatives, 18-mer LLKKK displayed the most powerful LPS-neutralizing activity: it was most potent at binding to LPS, inhibiting the interaction between LPS and LPS-binding protein, and attaching to the CD14 molecule, thereby suppressing the binding of LPS to CD14+ cells and attenuating production of tumor necrosis factor alpha (TNF-alpha) by these cells. Furthermore, in the murine endotoxin shock model, 18-mer LLKKK most effectively suppressed LPS-induced TNF-alpha production and protected mice from lethal endotoxin shock. Together, these observations indicate that the LPS-neutralizing activities of the amphipathic human CAP18/LL-37-derived 18-mer peptide can be augmented by modifying its hydrophobicity and cationicity, and that 18-mer LLKKK is the most potent of the peptide derivatives, with therapeutic potential for gram-negative bacterial endotoxin shock. Keywords: Introduction : Peptide antibiotics exhibit potent antimicrobial activities against both gram-positive and gram-negative bacteria, fungi, and viruses, and they form one group of effector components in the innate host defense system . The peptide-based defense in mammals against invading microbes relies on two evolutionarily distinct groups of antimicrobial peptides, defensins and cathelicidins, which have been identified in several epithelial tissues and in the granules of phagocytes (, , -, , , ). Defensins contain six conserved cysteine residues in their sequences and exhibit characteristic beta-sheet structures stabilized by three intramolecular disulfide bonds . In contrast, cathelicidins are characterized by highly conserved cathelin-like prosequences and variable carboxyl-terminal sequences that correspond to the mature antibacterial peptides . About 30 cathelicidin members from various mammalian species have been identified; however, only one cathelicidin, hCAP18 (human cationic antibacterial protein of 18 kDa), has been found in humans, and its carboxyl-terminal antibacterial peptide, called LL-37, which comprises 37 amino acid residues (L1LGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES37), has recently been identified . In addition to exhibiting potent antibacterial activities against gram-positive and gram-negative bacteria, peptides derived from LL-37 can bind to lipopolysaccharide (LPS) and neutralize its biological activities . LPS is a major constituent of the outer membranes of gram-negative bacteria and is recognized as a key molecule in the pathogenesis of endotoxin shock associated with gram-negative bacterial infections . Recently, we have shown that LL-37 exerts protective action against endotoxin shock by blocking the binding of LPS to CD14+ cells, thereby suppressing the production of cytokines by these cells . Thus, LL-37 and its related derivatives could be attractive candidates for therapeutic agents that can be used for endotoxin shock and sepsis caused by gram-negative bacterial infections . Although antibacterial peptides are diverse in their sizes, structures, and activities, they are mostly amphipathic, retaining both cationic (positively charged) and hydrophobic faces . These features facilitate interactions with negatively charged microbial surface membranes, followed by insertion into the microbial lipid membrane, which alters membrane permeability and impairs internal homeostasis . Moreover, the positively charged amphipathic structures are assumed to be important for interaction with a negatively charged amphipathic LPS . Interestingly, secondary-structure predictions indicate that antibacterial peptides of some cathelicidin members, such as hCAP18/LL-37, rabbit CAP18-derived peptide, and guinea pig CAP11 (cationic antibacterial polypeptide of 11 kDa), adopt an alpha-helical amphipathic conformation ; the helical wheel regions are clearly amphipathic and subtended by the hydrophilic (positively charged) and hydrophobic sectors . Interestingly, structure-activity relationship (SAR) studies using different kinds of natural and synthetic model peptides have revealed that the potencies and spectra of the antibacterial activities of amphipathic alpha-helical antimicrobial peptides can be influenced by interrelated structural and physicochemical parameters such as charge (cationicity), hydrophobicity, and amphipathicity . Thus, it could be anticipated that by changing these parameters one could design novel antimicrobial peptides with increased potency and directed activity which would be effective as therapeutic agents for bacterial infections and their related symptoms. FIG. 1. | Helical wheel projections for LL-37 and its 18-mer peptide derivatives. Helical wheel projections for LL-37 and its 18-mer peptide derivatives. The sequences of alpha-helical peptides LL-37, 18-mer K15-V32, 18-mer LL, and 18-mer LLKKK are presented according to the Shiffer-Edmundson wheel projection analyzed with a Genetyx-Mac computer system (Software Development, Tokyo, Japan). To increase hydrophobicity, E16 and K25 in 18-mer K15-V32 were replaced by L16 and L25 in 18-mer LL, respectively. Furthermore, to increase cationicity, Q22, D26, and N30 in 18-mer LL were replaced by K22, K26, and K30 in 18-mer LLKKK, respectively. Positively charged residues are circled, whereas negatively charged residues are boxed. Hydrophobic residues are outlined, while neutral hydrophilic residues are not. The hydrophilic and hydrophobic sectors are divided by dashed lines. During analysis of the biological activities of hCAP18/LL-37-derived antimicrobial peptides of different sizes , it was found that a short fragment peptide (18-mer; K15 to V32 [K15-V32]) of LL-37 displayed an amphipathic alpha-helical structure and possessed LPS-neutralizing activity almost equal to that of the parent peptide, LL-37; like LL-37, the 18-mer peptide inhibited the binding of LPS to CD14+ cells and suppressed LPS-induced cytokine production by these cells. Thus, the 18-mer peptide may be a good template for development of therapeutic agents that can be used for prevention of gram-negative bacterial sepsis and endotoxin shock. In this study, therefore, to develop 18-mer peptides with improved LPS-neutralizing activities, we modified the hydrophobicity and cationicity of the peptide by substitution of leucine and lysine residues, respectively, and evaluated the activities of those peptide derivatives by using the CD14+ murine macrophage cell line RAW 264.7 and the murine endotoxin shock model. MATERIALS AND METHODS : Reagents. | Fluorescein isothiocyanate (FITC)-conjugated or unconjugated LPS (from Escherichia coli O111:B4), the 3,3',5,5'-tetramethylbenzidine (TMB) liquid substrate system, and d-galactosamine were purchased from Sigma (St. Louis, Mo.). In some experiments, LPS was biotinylated with biotin-LC-hydrazide, based on the manufacturer's protocol (Pierce, Rockford, Ill.). A 37-mer peptide of hCAP18 (LL-37; L1LGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES37) and its 18-mer derivatives (18-mer K15-V32, 18-mer LL, and 18-mer LLKKK) were synthesized by the solid-phase method on a peptide synthesizer (model PSSM-8; Shimadzu, Kyoto, Japan) by fluorenylmethoxycarbonyl (Fmoc) chemistry. The peptides were eluted from the resin and purified to homogeneity by reversed-phase high-performance liquid chromatography on a Cosmosil 5C18 column (Nacalai Tesque, Kyoto, Japan) by use of a 0 to 70% acetonitrile gradient in 0.1% trifluoroacetic acid. The molecular masses of the peptides synthesized were confirmed on a mass spectrometer (model TSQ 700; Thermo Quest Finnigan, Manchester, United Kingdom). The sequences of the peptides were as follows: 18-mer K15-V32, K15EFKRIVQRIKDFLRNLV32; 18-mer LL, K15LFKRIVQRILDFLRNLV32; 18-mer LLKKK, K15LFKRIVKRILKFLRKLV32 (underlining indicates amino acid substitutions introduced into the original 18-mer peptide to increase hydrophobicity or cationicity). Tissue culture supplies were obtained from Iwaki Glass (Tokyo, Japan). Antibodies. | As anti-LPS-binding protein (LBP) antibodies, mouse anti-human LBP monoclonal antibody (MAb) 6G3 (HyCult Biotechnology, Uden, The Netherlands), which can cross-react with bovine LBP, and rat anti-mouse LBP MAb clone 39 (class 2) were used. These anti-LBP MAbs can recognize both free LBP and LBP-LPS complexes, and they inhibit the transfer of LPS to CD14. As anti-CD14 antibodies, FITC-conjugated rat anti-mouse CD14 MAb rmC5-3 (BD PharMingen, San Diego, Calif.) and FITC-conjugated or unconjugated rat anti-mouse CD14 MAb 4C1 were utilized. MAb 4C1 can block the binding of LPS to CD14+ cells, whereas rmC5-3 has little effect on LPS binding . Cells. | The murine macrophage cell line RAW 264.7 was obtained from the American Type Culture Collection (Manassas, Va.) and cultured in RPMI 1640 medium (Nissui Pharmaceutical, Tokyo, Japan) supplemented with 10% fetal bovine serum (FBS; Sanko Junyaku, Tokyo, Japan) at 37C under 5% CO2. Confluent RAW 264.7 cells were detached by washing with 0.05% EDTA in phosphate-buffered saline (PBS) (137 mM NaCl, 2.7 mM KCl, 8.1 mM Na2HPO4, 1.5 mM KH2PO4 [pH 7.4]) and suspended in RPMI 1640 containing 10% FBS. Assay for binding of FITC-conjugated LPS to RAW 264.7 cells. | RAW 264.7 cells (5 x 105/ml) were incubated with FITC-conjugated LPS (100 ng/ml) in the absence or presence of 18-mer peptides or LL-37 (0.01 to 10 mug/ml) in RPMI 1640 containing 10% FBS for 15 min at 37C. After cells were washed with PBS, the binding of FITC-conjugated LPS was analyzed by flow cytometry (FACScan; Becton Dickinson, Rutherford, N.J.), and median fluorescence intensity was determined. Alternatively, RAW 264.7 cells were incubated with FITC-conjugated LPS in the presence of anti-human LBP MAb 6G3 or anti-mouse CD14 MAb 4C1 at 5 mug/ml in RPMI 1640 containing 10% FBS, and the binding of FITC-conjugated LPS was analyzed as described above. In some experiments, RAW 264.7 cells were preincubated with 18-mer peptides or LL-37 at 1 mug/ml in RPMI 1640 containing 10% FBS for 10 min at 37C. After a wash, cells were incubated with FITC-conjugated LPS (100 ng/ml) in RPMI 1640 containing 10% FBS for 15 min at 37C, and LPS binding was evaluated by flow cytometry. In separate experiments, using RPMI 1640 containing 10% mouse serum, we confirmed that FITC-conjugated LPS could bind to RAW 264.7 cells, and the binding was completely suppressed by anti-mouse LBP MAb clone 39 and anti-mouse CD14 MAb 4C1. Moreover, we observed that 18-mer peptides and LL-37 inhibited the binding of FITC-conjugated LPS to RAW 264.7 cells and LPS-induced tumor necrosis factor alpha (TNF-alpha) expression by these cells (see below) in the medium containing 10% mouse serum, as in the medium containing 10% FBS (data not shown). These findings suggest the possibility that the peptides could function in mice. Evaluation of TNF-alpha expression. | RAW 264.7 cells (106/well in a 24-well microplate) were incubated with LPS (100 ng/ml) in the absence or presence of 18-mer peptides or LL-37 at 1 mug/ml in 500 mul of RPMI 1640 containing 10% FBS for 4 h at 37C. After incubation, cells were detached by a wash with 0.05% EDTA-PBS, and expression of TNF-alpha mRNA and protein was analyzed by Northern and Western blotting, respectively . In brief, total cellular RNA (2.5 mug) was separated by electrophoresis on a 1% agarose-formaldehyde gel, and Northern blot hybridization was performed by using cDNA probes labeled with a digoxigenin-High Prime DNA labeling kit (Roche Diagnostics, Mannheim, Germany); probes used were the 0.39-kb mouse TNF-alpha cDNA (encompassing nucleotides 427 to 819) and the 2.1-kb beta-actin cDNA (pHFbetaA-1) . In Western blot analysis, cell sonicates in PBS containing 1 mM diisopropyl fluorophosphate (3 x 105 cells for TNF-alpha and 3 x 104 cells for beta-actin) were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis on a 7.5 to 20% linear gradient of polyacrylamide under reducing conditions. Resolved proteins were electrotransferred to an Immobilon-P membrane (Millipore, Bedford, Mass.), and blots were probed with a rabbit anti-mouse TNF-alpha antibody (Ab) (Genzyme Diagnostics, Cambridge, Mass.) or an anti-beta-actin MAb (Sigma). Blots were further probed with horseradish peroxidase (HRP)-conjugated goat anti-rabbit immunoglobulin G (IgG) (Organon Teknika, Durham, N.C.) or goat anti-mouse IgG/IgM (Chemicon International, Temecula, Calif.), and proteins were finally detected with SuperSignal West Pico chemiluminescent substrate (Pierce). In some experiments, RAW 264.7 cells were incubated with LPS in the medium containing 10% mouse serum, and the effects of 18-mer peptides and LL-37 on LPS-induced TNF-alpha expression were investigated as described above. Measurement of LPS-binding activities of 18-mer peptides. | Ninety-six-well microtiter plates (Immulon 2H; Dynex Technologies, Ashford, United Kingdom) were coated with LPS (100 ng/well) as described previously . Excess binding sites were blocked with 100 mul of PBS containing 1% bovine serum albumin (BSA)/well, and 18-mer peptides and LL-37 (0.02 to 0.5 mug/well) were incubated in the plates for 1 h at 37C in 50 mul of RPMI 1640 without phenol red (Life Technologies, Grand Island, N.Y). Alternatively, 18-mer peptides and LL-37 (0.1 mug/well) were incubated in the presence of LPS (0.05 to 2.5 mug/well) in 50 mul of RPMI 1640. After a wash, an affinity-purified rabbit anti-CAP18 Ab (50 mul/well; 2 mug/ml in PBS containing 0.1% BSA) was added and incubated for 1 h at 37C in the plates. The Ab solution was then rinsed out, and HRP-conjugated goat anti-rabbit IgG (50 mul/well; diluted 2,000-fold in PBS containing 0.1% BSA) was incubated for 1 h at room temperature. Finally, a TMB liquid substrate (100 mul/well) was incubated for 5 to 15 min. The reaction was stopped by addition of 100 mul of 0.18 M sulfuric acid/well, and absorbances at 450 and 560 nm were quantitated in a microtiter plate reader. An anti-CAP18 serum was raised in rabbits by use of the LL-37 peptide covalently coupled to keyhole limpet hemocyanin, and the Ab was purified by affinity chromatography using LL-37 peptide-conjugated, epoxy-activated Sepharose (Amersham Pharmacia Biotech). In preliminary experiments, we confirmed that the anti-CAP18 Ab could recognize not only LL-37 but also 18-mer K15-V32, 18-mer LL, and 18-mer LLKKK (data not shown). Assay for interaction of LPS with LBP. | LPS (100 ng/well) was used to coat the 96-well-microtiter plates as described above. After blocking, RPMI 1640 containing 0.1, 1, or 10% FBS (50 mul/well) was added and incubated for 1 h at 37C. Plates were then washed, and 50 mul of anti-LBP MAb 6G3 (25 nM in PBS containing 0.1% BSA)/well was incubated in the plates for 1 h at 37C. The MAb solution was rinsed out, and 50 mul of HRP-conjugated rabbit anti-mouse IgG (Dako, Glostrup, Denmark; diluted 1,000-fold in PBS containing 0.1% BSA)/well was incubated for 1 h at room temperature. Finally, binding of LBP to the immobilized LPS was detected by incubation with TMB liquid substrate (100 mul/well). Alternatively, the LPS microtiter plates were preincubated with 18-mer peptides or LL-37 (0.025 to 0.2 mug/well) in 50 mul of RPMI 1640/well for 1 h at 37C. After a wash, RPMI 1640 containing 10% FBS (50 mul/well) was added, and binding of LBP was determined as described above. Assay for binding of LBP to 18-mer peptides. | Microtiter plates were coated with 18-mer peptides or LL-37 (2.5 mug/well) by incubating 50 mul of 50-mug/ml peptides in PBS/well overnight at room temperature. After blocking, RPMI 1640 containing 10% FBS (50 mul/well) was added and incubated for 1 h at 37C. Plates were then washed, and 50 mul of anti-LBP MAb 6G3 (25 nM in PBS containing 0.1% BSA)/well was incubated in the plates for 1 h at 37C. The MAb solution was rinsed out and replaced with 50 mul of HRP-conjugated rabbit anti-mouse IgG (diluted 1,000-fold in PBS containing 0.1% BSA)/well for 1 h at room temperature. Finally, binding of LBP to the immobilized peptides was detected by TMB reaction. As a positive control, biotinylated LPS (100 ng/well) was added to plates coated with 18-mer peptides or LL-37, and the plates were incubated for 1 h at 37C in 50 mul of RPMI 1640/well. The LPS solution was then rinsed out, and 50 mul of HRP-conjugated streptavidin (Dako; diluted 5,000-fold in PBS containing 0.1% BSA)/well was incubated for 1 h at 37C. Binding of biotinylated LPS to the peptides that had been used to coat the wells was finally detected by TMB reaction. Flow cytometric assay for expression of CD14 and binding of 18-mer peptides to CD14. | To analyze the effects of 18-mer peptides and LL-37 on CD14 expression, RAW 264.7 cells (5 x 105/ml) were incubated without or with the peptides (1 mug/ml) or LPS (100 ng/ml) in RPMI 1640 containing 10% FBS for 15 min at 37C and were further incubated with FITC-conjugated rat anti-mouse CD14 MAb rmC5-3 (2.5 mug/ml), or with FITC-conjugated rabbit anti-mouse IgG (Dako) as a negative control, for 15 min at 37C. After a wash, the binding of the anti-CD14 MAb was measured by flow cytometry. Furthermore, the binding of 18-mer peptides and LL-37 to CD14 was examined by using the neutralizing anti-mouse CD14 MAb 4C1, which recognizes the murine CD14 epitope located near the LPS-binding site . RAW 264.7 cells (5 x 105/ml) were preincubated without or with the peptides (1 mug/ml), or with LPS (100 ng/ml) as a positive control, for 15 min at 37C in RPMI 1640 containing 10% FBS. Cells were then added with 50 ng of FITC-conjugated anti-CD14 MAb 4C1/ml, or with 50 ng of FITC-conjugated rabbit anti-mouse IgG/ml as a negative control, and were further incubated for 15 min at 37C. After a wash, the binding of FITC-conjugated anti-CD14 MAb 4C1 was analyzed by flow cytometry. Evaluation of effects of 18-mer peptides on the murine endotoxin shock model. | To determine the protective activities of 18-mer peptides against lethal LPS activity, we utilized d-galactosamine-sensitized mice, which are highly susceptible to LPS . d-Galactosamine (18 mg/0.3 ml of saline), LPS (200 ng/0.2 ml of saline), and 18-mer peptides or LL-37 (1 mug/0.2 ml of saline) were sequentially injected intraperitoneally (i.p.) into male C57BL/6 mice aged 10 weeks (Charles River Japan, Kanagawa, Japan), and deaths were recorded every 24 h until day 6 after injection. Furthermore, 75 min after LPS challenge, mice were sacrificed by drawing blood from the heart, and serum samples were prepared. Serum TNF-alpha levels were determined by using a commercially available mouse TNF-alpha enzyme-linked immunosorbent assay kit (Endogen, Woburn, Mass.) that can detect <9 pg of TNF-alpha/ml. In some experiments, 75 min after challenge with d-galactosamine (18 mg), FITC-conjugated LPS (200 ng), and 18-mer peptides or LL-37 (1 mug), peritoneal fluids were harvested by washing the peritoneal cavities with PBS, and peritoneal macrophages were recovered. Binding of FITC-conjugated LPS to peritoneal macrophages was analyzed by flow cytometry, and expression of TNF-alpha in peritoneal macrophages was investigated by Northern and Western blotting, as described above. The experiments were carried out in accordance with institutional guidelines, and mice received proper care and maintenance. Statistical analysis. | Data are shown as means +- standard deviations (SD). Statistical significance was determined by one-way analysis of variance with a multiple-comparison test (StatView; Abacus Concepts, Berkeley, Calif.) unless otherwise noted. Survival in mice after LPS administration was assessed by a chi2 test (StatView). A P value of <0.05 was considered to be significant. RESULTS : Effects of 18-mer peptides on binding of FITC-conjugated LPS to RAW 264.7 cells. | We first examined the effects of 18-mer peptides on the binding of FITC-conjugated LPS to CD14+ cells by flow cytometry using the murine macrophage cell line RAW 264.7. When FITC-conjugated LPS was incubated with RAW 264.7 cells in the presence of serum, it bound to the cells . However, when FITC-conjugated LPS was incubated with RAW 264.7 cells in the presence of a neutralizing anti-LBP MAb or anti-CD14 MAb, LPS binding was completely inhibited to the background level (the level without FITC-conjugated LPS), indicating that LPS binds to the cellular CD14 molecule via the action of LBP in serum. Interestingly, 18-mer peptides as well as LL-37 (1 mug/ml each) markedly suppressed the binding of FITC-conjugated LPS to RAW 264.7 cells . Their effects were dose dependent , and 18-mer LLKKK (50% inhibitory concentration [IC50], 0.09 mug/ml, or 0.043 muM) was the most potent of the peptides tested at inhibiting the binding of FITC-conjugated LPS to RAW 264.7 cells (for 18-mer LL, the IC50 was 0.23 mug/ml, or 0.102 muM; for 18-mer K15-V32, the IC50 was 0.31 mug/ml, or 0.136 muM; for LL-37, the IC50 was 0.31 mug/ml, or 0.070 muM). Furthermore, the effects of 18-mer peptides on TNF-alpha expression were examined by Northern and Western blot analyses . The 18-mer peptides and LL-37 (1 mug/ml) apparently suppressed LPS-induced TNF-alpha expression by RAW 264.7 cells at both the mRNA and protein levels, and the effect of 18-mer LLKKK was the most prominent. Eighteen-mer LLKKK also exerted the most suppressive action on LPS-induced TNF-alpha expression by RAW 264.7 cells at 0.1 mug/ml among the peptide derivatives investigated (data not shown). LPS-binding activities of 18-mer peptides and their effects on the binding of LPS to LBP. | In order to clarify the mechanisms by which 18-mer peptides inhibit the binding of LPS to CD14+ cells, we investigated the LPS-binding activities of these peptides by using LPS-coated microtiter plates. The 18-mer peptides and LL-37 bound to the LPS-coated plates in a dose-dependent fashion , and the binding was dose-dependently inhibited by addition of LPS to the plates . Notably, the binding of 18-mer LLKKK was more potently inhibited by LPS addition (IC50, 0.08 mug of LPS/well) than that of the other peptides; IC50s of LPS, in micrograms per well, were 0.16 for18-mer LL, 0.25 for 18-mer K15-V32, and 0.34 for LL-37. Using the LPS-coated plates, we next examined the effects of 18-mer peptides on the interaction between LPS and LBP, which catalyzes the transfer of LPS to CD14 . When LPS-coated plates were incubated with FBS, the LBP in the serum bound to the LPS-coated plates proportionately to the concentrations of serum used . When the LPS-coated plates were pretreated with the peptide derivatives, binding of LBP to the LPS-coated plates was inhibited in a dose-dependent manner by addition of 18-mer peptides or LL-37 to the plates . Importantly, 18-mer LLKKK (IC50, 0.018 mug/well, or 7.9 pmol/well) was the most potent of the 18-mer peptides at inhibiting LPS-LBP binding; other IC50s were 0.024 mug/well, or 10.9 pmol/well, for 18-mer LL; 0.035 mug/well, or 15.2 pmol/well, for 18-mer K15-V32; and 0.025 mug/well, or 5.4 pmol/well, for LL-37. Furthermore, we evaluated the interaction of LBP with the peptides by using 18-mer peptide- or LL-37-coated microtiter plates. Apparently, substantial amounts of biotinylated LPS, a positive control, bound to the 18-mer peptide- or LL-37-coated plates. In contrast, LBP hardly bound to the plates (data not shown). The observations described above indicate that the 18-mer K15-V32, 18-mer LL, and 18-mer LLKKK peptides have potential to bind to LPS, but not LBP, thereby inhibiting the interaction of LPS with LBP. Moreover, 18-mer LLKKK appears to be the most potent at binding to LPS and inhibiting LPS-LBP interaction. Effects of 18-mer peptides on CD14 expression and binding of 18-mer peptides to CD14+ cells. | It is possible that antimicrobial peptides may alter CD14 expression, thereby affecting the binding of LPS to CD14+ cells. To check this, we investigated the expression of CD14 after treatment of RAW 264.7 cells with 18-mer peptides by flow cytometry using FITC-conjugated anti-mouse CD14 MAb rmC5-3. Neither 18-mer peptides LL-37 (1 mug/ml each), nor LPS (100 ng/ml) altered CD14 expression on RAW 264.7 cells . Likewise, it is conceivable that antimicrobial peptides bind to the CD14 molecule on the cells and influence the binding of LPS to CD14+ cells without affecting CD14 expression. To confirm this, RAW 264.7 cells were preincubated with 18-mer peptides, and after a wash, the binding of FITC-conjugated LPS was analyzed by flow cytometry. Interestingly, preincubation with 18-mer LLKKK or 18-mer LL (1 mug/ml) inhibited the binding of FITC-conjugated LPS to RAW 264.7 cells by 33.9% +- 0.1% or 22.5% +- 2.6%, respectively (n = 3; P < 0.001 for comparison with binding of FITC-conjugated LPS in the absence of peptides), while preincubation with 18-mer K15-V32 or LL-37 inhibited LPS binding by only 5 to 7% . To further clarify the binding of peptides to the CD14 molecule, we utilized anti-CD14 MAb 4C1, which can recognize the murine CD14 epitope and inhibit the binding of LPS to CD14+ cells . As expected, when RAW 264.7 cells were preincubated with LPS (100 ng/ml), binding of anti-CD14 MAb 4C1 (50 ng/ml) to the cells was inhibited by 40.2% +- 11.9% (n = 3) , suggesting that the epitope, which can be recognized by MAb 4C1, is located on murine CD14 next to the LPS-binding site. Importantly, preincubation with 18-mer LLKKK or 18-mer LL (1 mug/ml each) inhibited the binding of FITC-conjugated anti-CD14 MAb 4C1 to RAW 264.7 cells by 33.5% +- 5.2% or 26.2% +- 9.8%, respectively (n = 3 or n = 5; P < 0.05 for comparison with the binding of FITC-conjugated anti-CD14 MAb 4C1 in the absence of peptides), whereas preincubation with 18-mer K15-V32 or LL-37 inhibited the binding by only 10 to 14% . These observations likely suggest that among 18-mer peptides, 18-mer LLKKK has the most potent activity to bind to CD14 near the LPS-binding site, thereby influencing the binding of LPS to CD14. Effects of 18-mer peptides on the murine endotoxin shock model. | Using d-galactosamine-sensitized mice, we assessed the effects of 18-mer peptides on lethal LPS activity in vivo . d-Galactosamine administration sensitized mice to the lethal effect of LPS, and 94% of the mice died within 24 h after i.p. injection of 200 ng of LPS . Notably, administration of 18-mer K15-V32, 18-mer LL, 18-mer LLKKK, or LL-37 (1 mug/mouse) increased the survival rate to 33, 50, 67, or 47%, respectively, all significantly higher than that of the group with d-galactosamine and LPS administration alone (P < 0.05); 18-mer LLKKK administration was the most protective. In agreement with this, administration of an 18-mer peptide or LL-37 markedly lowered the LPS-induced increase in serum TNF-alpha levels (P < 0.001) , and 18-mer LLKKK had the greatest effect. In addition, the effects of 18-mer peptides on the binding of FITC-conjugated LPS to peritoneal macrophages (CD14+ cells) were analyzed by flow cytometry. Administration of the peptides, especially 18-mer LLKKK, markedly suppressed the binding of FITC-conjugated LPS to peritoneal macrophages (P < 0.001) . Concurrently, TNF-alpha expression in peritoneal macrophages was investigated. Northern and Western blot analyses indicated that administration of the peptides, particularly 18-mer LLKKK, strongly suppressed LPS-induced TNF-alpha mRNA and protein expression by peritoneal macrophages (data not shown), just as was observed in LPS-stimulated RAW 264.7 cells . Together, these observations likely suggest that among 18-mer peptides, 18-mer LLKKK exerts the most protective action against murine endotoxin shock by blocking the binding of LPS to CD14+ cells, thereby suppressing the production of cytokines (such as TNF-alpha) by these cells, possibly via its potent binding to LPS and CD14. FIG. 2. | Effects of 18-mer peptides on binding of FITC-conjugated LPS to RAW 264. Effects of 18-mer peptides on binding of FITC-conjugated LPS to RAW 264.7 cells. (A) RAW 264.7 cells (5 x 105/ml) were incubated with 100 ng of FITC-conjugated LPS/ml in the absence or presence of 5 mug of anti-LBP MAb 6G3 or anti-CD14 MAb 4C1/ml in RPMI 1640 containing 10% FBS for 15 min at 37C. (B) Alternatively, RAW 264.7 cells were incubated with FITC-conjugated LPS in the presence of 1 mug of peptide (18-mer K15-V32, 18-mer LL, 18-mer LLKKK, or LL-37)/ml in RPMI 1640 containing 10% FBS. After a wash, the binding of FITC-LPS was analyzed by flow cytometry. Background (BG) was assessed by using RAW 264.7 cells incubated without FITC-LPS. Data are from one of five separate experiments. FIG. 3. | Dose-dependent inhibition of the binding of FITC-conjugated LPS to RAW 264. Dose-dependent inhibition of the binding of FITC-conjugated LPS to RAW 264.7 cells by 18-mer peptides. RAW 264.7 cells (5 x 105/ml) were incubated with 100 ng of FITC-conjugated LPS/ml in the absence or presence of 0.01 to 10 mug of 18-mer K15-V32, 18-mer LL, 18-mer LLKKK, or LL-37/ml in RPMI 1640 containing 10% FBS for 15 min at 37C. After a wash, the binding of FITC-LPS was analyzed by flow cytometry, and median fluorescence intensity was determined. Binding of LPS was expressed as a percentage of that obtained by using RAW 264.7 cells incubated with FITC-conjugated LPS alone. Data are means +- SD from three to five separate experiments. Values for 18-mer K15-V32 are compared with those for 18-mer LL, 18-mer LLKKK, or LL-37. *, P < 0.05; **, P < 0.01; ***, P < 0.001. FIG. 4. | Effects of 18-mer peptides on LPS-induced TNF-alpha expression by RAW 264. Effects of 18-mer peptides on LPS-induced TNF-alpha expression by RAW 264.7 cells. RAW 264.7 cells (106/well in a 24-well microplate) were incubated without (Control) or with 100 ng of LPS/ml in the absence (LPS) or presence of 1 mug of peptide (18-mer K15-V32, 18-mer LL, 18-mer LLKKK, or LL-37)/ml in 500 mul of RPMI 1640 containing 10% FBS for 4 h at 37C. After incubation, cells were recovered, and expression of TNF-alpha mRNA and protein was analyzed by Northern (A) and Western (B) blotting, respectively. As a control, expression of beta-actin mRNA and protein was also analyzed. Data are from one of three separate experiments. FIG. 5. | Evaluation of LPS-binding activities of 18-mer peptides. Evaluation of LPS-binding activities of 18-mer peptides. (A) LPS-binding activities of the peptides were investigated by incubating 0.02 to 0.5 mug of 18-mer K15-V32, 18-mer LL, 18-mer LLKKK, or LL-37 in LPS-coated 96-well microtiter plates (100 ng of LPS/well) for 1 h at 37C in 50 mul of RPMI 1640. Bound peptides were detected by TMB reaction by using a rabbit anti-CAP18 Ab and HRP-conjugated goat anti-rabbit IgG. (B) The peptide (0.1 mug of either 18-mer K15-V32, 18-mer LL, 18-mer LLKKK, or LL-37) was incubated in LPS-coated 96-well microtiter plates for 1 h at 37C in the absence or presence of added LPS (0.05 to 2.5 mug/well) in 50 mul of RPMI 1640, and bound peptides were detected as described above. Binding of peptides to LPS-coated plates was expressed as a percentage of that observed when 0.5 mug (for panel A) or 0.1 mug (for panel B) of each peptide was incubated in the absence of added LPS. Data are means +- SD from three to five separate experiments. Values for 18-mer K15-V32 are compared with those for 18-mer LL, 18-mer LLKKK, or LL-37. *, P < 0.05; ***, P < 0.001. FIG. 6. | Effects of 18-mer peptides on the interaction of LPS with LBP. Effects of 18-mer peptides on the interaction of LPS with LBP. (A) LPS-LBP binding was examined by incubating 50 mul of RPMI 1640 containing 0.1, 1, or 10% FBS in LPS-coated 96-well microtiter plates (100 ng of LPS/well) for 1 h at 37C. After incubation, bound LBP was detected by TMB reaction using anti-LBP MAb 6G3 and HRP-conjugated rabbit anti-mouse IgG. (B) LPS-coated microtiter plates were preincubated with either 18-mer K15-V32, 18-mer LL, 18-mer LLKKK, or LL-37 at 0.025 to 0.2 mug/well in 50 mul of RPMI 1640 for 1 h at 37C. Thereafter, LPS-LBP binding was determined by incubating 50 mul of RPMI 1640 containing 10% FBS in the microtiter plates as described above. LPS-LBP binding was expressed as a percentage of that obtained by incubation with RPMI 1640 containing 10% FBS in the absence of added peptide. Data are means +- SD from three to eight separate experiments. Values for 18-mer K15-V32 are compared with those for 18-mer LL, 18-mer LLKKK, or LL-37. *, P < 0.05; **, P < 0.01; ***, P < 0.001. FIG. 7. | Effects of 18-mer peptides on expression of CD14 by RAW 264. Effects of 18-mer peptides on expression of CD14 by RAW 264.7 cells. RAW 264.7 cells (5 x 105/ml) were incubated without or with 100 ng of LPS/ml (A) or 1 mug of peptide (18-mer K15-V32, 18-mer LL, 18-mer LLKKK, or LL-37)/ml (B) in RPMI 1640 containing 10% FBS for 15 min at 37C, after which cells were further incubated with 2.5 mug of FITC-conjugated rat anti-mouse CD14 MAb rmC5-3/ml for 15 min at 37C. After a wash, the binding of FITC-anti-CD14 MAb rmC5-3 was measured by flow cytometry. Background (BG) was assessed by using RAW 264.7 cells incubated with FITC-conjugated rabbit anti-mouse IgG (a negative control for nonspecific binding). Data are from one of three separate experiments. FIG. 8. | Evaluation of the binding of 18-mer peptides to CD14+ cells. Evaluation of the binding of 18-mer peptides to CD14+ cells. (A) RAW 264.7 cells (5 x 105/ml) were preincubated with 1 mug of peptide (18-mer K15-V32, 18-mer LL, 18-mer LLKKK, or LL-37)/ml in RPMI 1640 containing 10% FBS for 10 min at 37C. After a wash, cells were incubated with 100 ng of FITC-conjugated LPS/ml in RPMI 1640 containing 10% FBS for 15 min at 37C, and the binding of FITC-conjugated LPS was analyzed by flow cytometry. Background (BG) was assessed by using RAW 264.7 cells incubated without FITC-conjugated LPS. (B and C) RAW 264.7 cells (5 x 105/ml) were preincubated without or with 100 ng of LPS/ml or 1 mug of peptide (18-mer K15-V32, 18-mer LL, 18-mer LLKKK, or LL-37)/ml for 15 min at 37C in RPMI 1640 containing 10% FBS. Cells were further incubated with 50 ng of FITC-conjugated neutralizing anti-CD14 MAb 4C1/ml for 15 min at 37C, and after a wash, the binding of FITC-conjugated anti-CD14 MAb 4C1 was analyzed by flow cytometry. Background (BG) was assessed by using RAW 264.7 cells incubated with FITC-conjugated rabbit anti-mouse IgG (a negative control for nonspecific binding). Data are from one of three to six separate experiments. FIG. 9. | Protective effects of 18-mer peptides on survival and serum TNF-alpha levels in LPS-challenged mice. Protective effects of 18-mer peptides on survival and serum TNF-alpha levels in LPS-challenged mice. (A) Mice were i.p. injected with d-galactosamine alone at 18 mg/mouse (Control) or with LPS at 200 ng/mouse with or without 1 mug of peptide (18-mer K15-V32, 18-mer LL, 18-mer LLKKK, or LL-37)/mouse. After injection, deaths were recorded every 24 h until day 6 (16 mice in each group). (B) After LPS challenge (75 min), serum samples were prepared, and serum TNF-alpha levels were determined by using a commercially available mouse TNF-alpha enzyme-linked immunosorbent assay kit. Data are means +- SD for 16 mice in each group. Values without and with peptide administration are compared, as are values with and without LPS administration. *, P < 0.05; **, P < 0.01; ***, P < 0.001. FIG. 10. | Effects of 18-mer peptide administration on the binding of FITC-conjugated LPS to mouse peritoneal macrophages. Effects of 18-mer peptide administration on the binding of FITC-conjugated LPS to mouse peritoneal macrophages. Mice were i.p. injected with d-galactosamine alone at 18 mg/mouse (Control) or with FITC-conjugated LPS at 200 ng/mouse with or without 1 mug of peptide (18-mer K15-V32, 18-mer LL, 18-mer LLKKK, or LL-37)/mouse. After LPS challenge (75 min), peritoneal macrophages were recovered, and binding of FITC-conjugated LPS to the cells was analyzed by flow cytometry. (A) Data are from 1 of 16 mice in each group. (B) Data are means +- SD for 16 mice in each group. Values without and with peptide administration are compared, as are values with and without LPS administration. ***, P < 0.001. DISCUSSION : For prevention of bacterial infections and their related symptoms (e.g., gram-negative bacterial septic shock), much attention has been focused on the low-molecular-weight cationic antimicrobial peptides that possess both antibacterial and LPS-neutralizing activities . Previously, it was demonstrated that in addition to exhibiting potent antibacterial activity against gram-positive and gram-negative bacteria , hCAP18/LL-37 could bind to the lipid-A moiety of LPS and inhibit the interaction of LPS with LBP, which transports LPS to CD14+ cells, thereby suppressing the binding of LPS to CD14+ cells and possibly attenuating Toll-like receptor-mediated CD14+ cell activation . In this study, we have revealed that the LPS-neutralizing activity of the amphipathic hCAP18/LL-37-derived 18-mer peptide can be augmented by modifying its hydrophobicity and cationicity, and that 18-mer LLKKK, with increased hydrophobicity and cationicity, is the most potent of the peptide derivatives. Defensins and cathelicidins are the two major families of mammalian antimicrobial peptides, which contribute to the innate host defense against invading microorganisms by virtue of their broad spectra and potent antimicrobial activities (, , -, , ). Defensins are characterized by the presence of six invariant cysteine residues in their sequences and exhibit beta-sheet structure . In contrast, antimicrobial peptides of cathelicidins display remarkable structural variety; some are alpha-helical (e.g., hCAP18/LL-37, rabbit CAP18-derived peptide, and guinea pig CAP11), and others are proline and arginine rich, showing a polyproline type structure (e.g., porcine PR-39 and bactenecins), whereas porcine protegrins exhibit beta-sheet structure . Notably, alpha-helical cathelicidin peptides such as hCAP18/LL-37- and rabbit CAP18-derived peptides and guinea pig CAP11 have been shown to exert antimicrobial activities even under physiological conditions in which salt (150 mM NaCl) and serum are present, where the antimicrobial activities of defensins are substantially diminished . In addition, among cathelicidin peptides, only CAP18-derived peptides (human and rabbit) and CAP11 have been revealed to possess LPS-binding and -neutralizing activities . In contrast, defensins exhibit little, if any, LPS-binding or -neutralizing activity . Thus, CAP18- and CAP11-derived peptides could have therapeutic potential for bacterial infections and gram-negative bacterial endotoxin shock in vivo. In this study, to develop cathelicidin-derived antimicrobial peptides with improved LPS-neutralizing activities, we utilized the 18-mer peptide (K15-V32) of human cathelicidin hCAP18/LL-37 as a template and evaluated the activities of its peptide derivatives. By replacement of E16 and K25 with two L residues, the hydrophobicity of the peptide was increased and the hydrophobic sector in the helix was extended (Fig. and ; compare 18-mer K15-V32 with 18-mer LL). Concomitantly, the pI was increased from 11.50 to 12.22. Furthermore, by replacement of Q22, D26, and N30 with three K residues, the hydrophilicity of the peptide was enhanced and the positively charged hydrophilic sector in the helix was expanded (18-mer LL versus 18-mer LLKKK), accompanied by a further increase in pI from 12.22 to 12.82. Among 18-mer peptides, 18-mer LLKKK displayed the most powerful LPS-neutralizing activity: it was most potent at binding to LPS, inhibiting LPS-LBP interaction, and attaching to the CD14 molecule, thereby suppressing the binding of LPS to CD14+ cells and attenuating the production of TNF-alpha, an important inflammatory mediator for the development of endotoxin shock , by these cells. Furthermore, 18-mer LLKKK most effectively protected mice from lethal endotoxin shock. Thus, the enhanced hydrophobicity and cationicity (positive charge) of 18-mer LLKKK are likely important for the expression of its augmented activity in binding with and neutralizing a negatively charged amphipathic LPS. Furthermore, these features may facilitate the interaction of the peptide with the CD14 molecule. Taken together, these observations likely indicate that the alpha-helical amphipathic antimicrobial peptides with increased hydrophobicity and cationicity, such as 18-mer LLKKK, can interact more potently with negatively charged amphipathic LPS than the parent 18-mer peptide, thereby exerting augmented LPS-neutralizing activities. In agreement with this, SAR studies have indicated that the antibacterial activities of alpha-helical antimicrobial peptides could be modulated by interrelated structural and physicochemical parameters (such as charge, hydrophobicity, and amphipathicity) and that the maximum antibacterial potency of alpha-helical antimicrobial peptides could be obtained when a high charge (cationicity) and amphipathicity were achieved . Thus, our present findings suggest that the results of SAR studies on the antibacterial activities of alpha-helical antimicrobial peptides could be applied to the estimation of the LPS-neutralizing activities of these peptides and that novel antimicrobial peptides with enhanced LPS-neutralizing activity and therapeutic potential may be evolved by changing those parameters. Cationic antimicrobial peptides target cell surface anionic lipids such as phosphatidyl glycerol and cardiolipin that are abundant in microorganisms; the action is not receptor based but involves a less specific interaction with microbial membrane components . In contrast, the mammalian cell membrane is mainly composed of electrically neutral phospholipids such as phosphatidylcholine and sphingomyelin, for which the affinity of the antimicrobial peptides is generally low . The simple electrostatic interaction between cationic antimicrobial peptides and microbial membrane lipids provides selective toxicity (bacteria versus mammalian cells) as well as a broad spectrum of antimicrobial activities. Moreover, development of microbial resistance is assumed to be low, because the target molecules (anionic lipids) are important components conserved among microorganisms, and the molecular recognition between cationic peptides and target molecules is rather lenient . In addition, the peptides are small and relatively easy to synthesize. From these points of view, cationic antimicrobial peptides could be promising candidates for new antibiotics with therapeutic value. In this study, we have revealed that by modifying the hydrophobicity and cationicity of the amphipathic alpha-helical antimicrobial peptide (hCAP18/LL-37-derived 18-mer peptide), the in vitro and in vivo biological activities of the peptide could be enhanced. Among the peptide derivatives, 18-mer LLKKK obviously possessed more potent LPS-neutralizing activities than the parent peptide and appeared to have therapeutic potential for gram-negative bacterial endotoxin shock. However, it should be noted that as cationic antimicrobial peptides act principally via electrostatic attraction with, and hydrophobic partitioning into, the membrane targets, they could also bind to various host components such as plasma lipoproteins and anionic constituents of host cell membranes, leading to potentially harmful side effects on the host . In this context, it has been demonstrated that high concentrations of cationic antimicrobial peptides are occasionally toxic to host cells and that cytotoxicity is correlated with the extent of the hydrophobic regions in the peptides . Actually, we observed in the present study that 18-mer LL and 18-mer LLKKK with increased hydrophobicity were toxic to mammalian cells such as peritoneal macrophages and RAW 264.7 cells at >10 mug/ml; however, they exerted strong anti-LPS activities without exhibiting any cytotoxic effect on these cells at <10 mug/ml (data not shown). Thus, cationic antimicrobial peptides should be cautiously administered in vivo, considering their toxic effects on host cells. Although many problems still need to be solved , antimicrobial peptides could become one of the new classes of antibiotics that can be used for treatment of bacterial infections and their related symptoms in the future. FIG. 11. | Hydrophilicity/hydrophobicity plots of LL-37 and its 18-mer peptide derivatives. Hydrophilicity/hydrophobicity plots of LL-37 and its 18-mer peptide derivatives. Hydropathy indices (+, hydrophilicity; -, hydrophobicity) of LL-37 and its derivatives (18-mer K15 to V32, 18-mer LL, and 18-mer LLKKK) were calculated by the algorithm of Hopp and Woods by use of a Genetyx-Mac computer system, and the pIs of the peptides were also determined by the same system. Horizontal axes display the amino acid position number. Backmatter: PMID- 12204942 TI - Biological and Clinical Significance of Lipids as Modulators of Immune System Functions AB - Keywords: null: null TYPES OF FATTY ACIDS THAT ALTER IMMUNE SYSTEM FUNCTIONS : Overall, the integrated action of several polyunsaturated or monounsaturated fatty acids on the immune system includes the alteration of different immune functions in both humans and animals. Therefore, lymphocyte proliferation , production of cytokines , modification of cell surface molecules , changes in phagocytic activity , and alteration of natural killer (NK) cell activity are susceptible to be modulated by dietary lipid administration or by culture of cells in the presence of several free fatty acids. However, it is important to note that not all of the fatty acids have the same immunomodulatory properties. Thus, long-chain n-3 polyunsaturated fatty acids, such as EPA (20:5n-3) or DHA (22:6n-3) (which belong to n-3 fatty acid series derived from fish oil) are particularly known to exert a profound influence on the immune system, reducing many of these functions previously described for both humans and animals . Other fatty acids belonging to n-6 fatty acid series such as AA (20:4n-6) have also been reported to affect the immune system; although AA has not been shown to produce an immunosuppression as relevant as long-chain n-3 fatty acids, it is also considered an immunosuppressive agent. Nevertheless, both n-6 polyunsaturated fatty acids and saturated fatty acids have been associated to a chemopromotive role in cancer, probably due in part to a suppression of apoptosis . However, oleic acid (a monounsaturated fatty acid [18:1n-9] that is the main component of olive oil belonging to n-9 fatty acid series) is also capable of acting as a modulator of immune functions . Accordingly, olive oil (a traditional and essential component of the Mediterranean diet) also plays an important role in the modulation of immune functions, and therefore it has been efficiently applied in the alleviation of the symptomatology caused by rheumatoid arthritis . BIOLOGICAL CONSEQUENCES ON IMMUNE FUNCTIONS ATTRIBUTED TO FATTY ACID ADMINISTRATION : Differential properties of fatty acids and their role in the modulation of immune system functions. | The changes attributed to fatty acid administration include the reduction of lymphocyte proliferation, which is modified by polyunsaturated fatty acids (n-3 or n-6) or monounsaturated fatty acids (n-9). Studies carried out in both humans and animals have revealed that the administration of high levels of dietary n-3 polyunsaturated fatty acids or the inclusion in parenteral regimens of lipid emulsions rich in n-3 or n-9 fatty acids are related to the reduction of lymphocyte proliferation during the supplementation . Thus, concanavalin-A or lipopolysaccharide-stimulated lymphocytes have reduced the cellular proliferation in assays carried out in vitro or ex vivo in the presence of free fatty acids or in cell cultures from both animals and humans fed dietary lipids, respectively . Cytokine production is reduced by the action of n-3 polyunsaturated fatty acids or n-9 monounsaturated fatty acids , cytokine receptor expression is also affected , natural killer (NK) activity is significantly suppressed , phagocytic activity of macrophages is modified , and the antigen-presenting function of human monocytes is inhibited . Based on these experimental observations, we can affirm that n-3 polyunsaturated fatty acids rather than n-6 polyunsaturated or n-9 monounsaturated fatty acids are directly associated to the alteration of immune and inflammatory response. This modulation depends on different factors, such as the nature of fatty acids added to diets, the concentration of fatty acids, the duration of supplementation with dietary lipids, or differences among animal species fed dietary lipids . Mechanisms of action. | As mentioned previously, for several years numerous studies have attempted to elucidate the mechanisms by which some dietary lipids produce a potential effect on immune system functions . Several hypotheses have been suggested as possible mechanisms. Experimental investigations have confirmed that several fatty acids exert changes in the phospholipids of plasma membrane which affect the membrane fluidity ; they also alter eicosanoid production , produce lipid peroxides , or regulate the transcription factors . Overall, long-chain n-3 polyunsaturated fatty acids, such as EPA, are incorporated into cell membranes, replacing AA (the most important of the eicosanoid precursors). This process reduces the production of a biological mediator as prostaglandin E2 responsible for inhibiting IL-1 and tumor necrosis factor (TNF) production . Similarly, EPA inhibits gene transcription because it reduces the translocation of NF-kappaB . In addition to these important considerations, a crucial role of lymphocyte subsets has been suggested to explain the effects of different dietary polyunsaturated fatty acids on the immune system, because they alter the number of lymphocyte subsets as well as the proliferation of these cells . Thus, antigen presentation and the proportion of T-cell subsets are modified after dietary lipid manipulation. In fact, recent studies have determined that polyunsaturated fatty acids inhibit the surface expression of major histocompatibility complex (MHC) class II molecules, as well as some adhesion molecules on human monocytes . Similarly, consumption of a monounsaturated fatty acid-rich diet by humans is also involved in the reduction of adhesion molecules from peripheral blood mononuclear cells . Overall, these actions may explain the different expression of CD4 or CD8 on the T-cell surfaces of peripheral blood mononuclear cells of mice that were fed diets supplemented with DHA . The biological consequences of these changes are still unclear. It is probable that the alteration experimented by the lymphocyte population enhances the host susceptibility against opportunistic infections. In other words, the modulation of lymphocyte subsets may lead in part to a modification of cytokine production, NK cell activity, antibody production, etc. Accordingly, Th1-type cytokines such as IL-2 are more susceptible to polyunsaturated fatty acid effects than are Th2-type cytokines, such as IL-4, which indicates that these fatty acids are capable of altering the balance of Th1- and Th2-type cytokines in mice . As a direct consequence of a reduction of IL-2 production, Th1 cell activation is suppressed, whereas an increase of Th2 cell activation due to an enhancement of IL-4 production suppresses Th1 cell proliferation in animals fed a diet containing fish oil . Therefore, the anti-inflammatory effects derived from long-chain n-3 polyunsaturated fatty acid administration may be the result of these two interconnected mechanisms. Hence, it is obvious that these alterations may affect significantly the interaction between host natural resistance and infectious agents due to the crucial importance of Th1-mediated response in the infectious processes promoted by viruses or bacteria . Although results from epidemiological, experimental, or clinical studies have demonstrated an important association between n-3 polyunsaturated fatty acids and immunosuppressive properties in both humans and animals, their effects are complex and still not clearly known. In fact, some discrepant results have suggested that n-3 polyunsaturated fatty acids have stimulatory effects on TNF and IL-1 production . Evidence for the involvement of fatty acids in apoptosis modulation. | Programmed cell death or apoptosis has been defined as an essential mechanism responsible for the regulation of homeostasis, tissue development, or immune functions. Thus, damaged, aberrant, or unnecessary cells must be eliminated to ensure the correct development of multicellular organisms. Irrespective of its crucial role in normal cell control, apoptosis regulates pathological processes including human clinical disorders such as cancer, autoimmune diseases, viral or bacterial infections, and neurodegenerative disorders. Several biological factors (many viruses, bacteria, or parasites), chemical factors (glucocorticoids, ceramides, etc.), and physical factors (irradiation) are responsible for apoptosis induction . Moreover, different studies with both animals and humans, involving in vitro assays or administration of dietary lipids, have recently described an important role of several fatty acids in the induction or inhibition of apoptosis. As a result, different mechanisms of action have been proposed in order to explain the action of fatty acids on apoptosis modulation. Thus, polyunsaturated fatty acids, such as EPA and DHA, saturated fatty acids, such as PA, or fats such as fish oil administered in the diets have been defined as substances capable of inducing cell death via a mitochondrial process or by downregulation of Bcl-2 . Different mechanisms have been proposed in order to explain the induction of apoptosis by certain fatty acids. Thus, PA (a saturated fatty acid) added to cellular cultures in in vitro assays induces apoptosis via a direct effect on mitochondria, because it causes a dissipation of the mitochondrial transmembrane potential (DeltaPsim), an event that precedes nuclear apoptosis . Another possible partial explanation for the modulation of apoptosis by long-chain n-3 polyunsaturated fatty acids is that it is caused by direct action of these substances on the cells and by activation of the caspase cascade through cytochrome c release coupled with a modulation of mitochondrial membrane depolarization . Recent studies have also determined the crucial importance of dietary fatty acids in the reduction of Bcl-2 expression as well as an increase of Fas ligand (Fas-L) expression. Therefore, when the concentration of polyunsaturated fatty acids augments, Bcl-2 expression is reduced and cell death occurs . In other words, long-chain fatty acids contained in fish oil trigger apoptosis by suppressing Bcl-2 expression and increasing Fas-L expression. On the other hand, it has been recently determined that Ras membrane localization (a protein that plays a critical role in the cell growth and apoptosis) is reduced after administration of fish oil, which indicates that dietary fish oil may play a protective role against colon cancer development . Finally, it is important to note that lipid peroxidation is of crucial importance in apoptosis induction. This argument suggests that the effects of polyunsaturated fatty acids can be directly associated to specific alterations on the expression of Bcl-2, Fas, and Ras . CLINICAL CONSEQUENCES OF FATTY ACID ADMINISTRATION : Early studies revealed the importance of n-3 polyunsaturated fatty acids as possible therapeutic substances capable of reducing the incidence of inflammatory disorders in humans . Autoimmune diseases such as rheumatoid arthritis, psoriasis, or systemic lupus erythematosus are characterized by the production of an inflammatory response due to a marked increase of proinflammatory cytokines. In recent years, numerous epidemiological studies have described widely the reduction in the incidence of rheumatoid arthritis in populations who consume fish oil or olive oil in their diets, which are related to the prevention of inflammatory disorder incidence, although these events occur after a prolonged ingestion of these fats (, -, , ). Nevertheless, in spite of the importance of these substances as biological mediators in the inflammation and their application in the resolution of inflammatory diseases, in the present review we will summarize other important features of fatty acids from a clinical point of view, focusing on the potential effect of fatty acids in cancer incidence and the involvement of these substances in the alteration of host natural resistance against infectious agents. Crucial role of fatty acids in the incidence of cancer. | Many lines of evidence have indicated the essential role of fish oil in the incidence of cancer . Numerous reports have suggested that high intakes of long-chain n-3 fatty acids or n-9 fatty acids (the most important fatty acid contained in olive oil) may reduce the risk of breast cancer in both humans and animals. Thus, n-3 polyunsaturated fatty acids may inhibit colon cancer in rats as well as reduce the risk of colorectal cancer development . In contrast, n-6 fatty acids or saturated fatty acids may be involved in the increase of mammary or colon tumorigenesis by altering membrane phospholipid turnover. Accordingly, AA is released from plasma membrane that produces an alteration in the synthesis of prostaglandins via cyclooxygenase enzyme . Epidemiological studies have determined the protective role of n-3 polyunsaturated fatty acids against colon cancer in populations of Alaskan and Greenland Eskimos . Subsequently, experimental studies demonstrated that a fish oil diet appears to exert a protective effect against experimentally induced colon cancer . Similarly, a recent study has determined that olive oil also exerts a protective role against the development of colorectal cancer . Although the mechanisms that contribute to this protective effect have not been elucidated, a recent investigation has reported that fish oil diets are able to increase apoptosis and cell differentiation in induced rat colon tumorigenesis . Based on current knowledge, an interesting study has recently determined that the Western diet (which consists predominantly of a mixture of saturated, monounsaturated, and polyunsaturated fats) produces dysplastic lesions in the colon, which is indicative of a tumorigenic process. As a result, the administration of a diet containing mixed lipids and high levels of saturated fatty acids (similar to the lipids contained in the Western diet) induces colon carcinogenesis in mice by an unknown mechanism . Nevertheless, the most interesting results from this investigation reveal that an alteration of cyclooxygenase activity as well as apoptosis suppression appears to play a crucial role in the colon tumorigenesis induced by this diet in animal models . In addition, our experimental observations have indicated that the survival of mice fed a diet containing fish oil and transplanted with a T-lymphoma cell line is reduced in comparison to that of mice fed a diet containing low fat. We speculate that this event may be attributed to the impairment in the function of T cells or NK cells due to polyunsaturated fatty acid administration, which reduces the cytotoxic activity of NK cells . Recently, an interesting investigation has shown that n-3 fatty acids alter cell transformation in a mouse epidermal cell line because DHA or EPA inhibited transcription activator protein AP-1 whereas AA did not produce any effect on AP-1 activity. Therefore, this finding confirms previous reports of the chemoprotective role of diets containing n-3 fatty acids . In contrast, the most significant relationship between fatty acids and reduction of cancer incidence is found in the Mediterranean diet, which could decrease the development of cancer by up to 10 to 15% in the populations of highly developed Western countries . In spite of the wide information currently available, the potential role of dietary lipids in cancer etiology remains controversial. In fact, although most of the studies have shown that diets containing n-3 polyunsaturated fatty acids might be beneficial during inflammatory conditions, their effects are not elucidated because they promote an impairment of the host defense to infection. Therefore, future studies are needed in order to confirm or refute the different hypotheses and arguments suggested hitherto. Relevance of dietary lipids in the reduction of host natural resistance and susceptibility to infection. | As mentioned above, nutrient intake is considered a critical determinant of immunocompetence. Many investigations have reported the modulatory functions exerted by fatty acids and the clinical benefits of dietary lipid supplementation with fish oil or olive oil in both humans and animals. As a consequence, this dietary supplementation has been applied in the treatment of patients suffering from inflammatory diseases, because unsaturated fatty acids (mainly n-3 or n-9 fatty acids) reduce the levels of many biological mediators associated to the promotion of the inflammatory events that participate in an inappropriate immune response. However, the reported reduction of immune response caused by the administration in the diet of some fatty acids may exert a detrimental effect, and therefore, they may impair the cellular immunity to pathogenic agents . For obvious reasons, the altered resistance to bacterial infection has been analyzed in animal models in which the administration of diets containing fish oil reduces the clearance of bacteria from liver or spleen and significantly reduces the survival during the course of Listeria monocytogenes infection. As a result, the elimination of pathogenic agents (bacteria, viruses, or parasites) is more difficult. Different reports have described the clinical consequences of dietary supplementation rich in n-3 fatty acids that suppress the immune system functions. Thus, a significant reduction of survival percentage after feeding experimental mice with a diet containing fish oil has been determined whereas an increase of survival percentage was observed after challenge with L. monocytogenes in animals fed diets containing either lard oil, hydrogenated coconut oil, or palm oil . These mice were inoculated with a lethal dose of a virulent L. monocytogenes strain, a facultative intracellularly growing bacterium that has been used as a model of infectious and pathogenic processes. After the administration of this diet different results were observed. Bacterial clearance from liver or spleen was increased in these animals , bactericidal activity of peritoneal cells was significantly reduced , and cytotoxic effects due to bacterial infection were increased , whereas the susceptibility of cells to adhesion or invasion by L. monocytogenes infection was substantially modified . These observations indicate the loss of capacity of the immune system from animals fed a diet containing fish oil to destroy and eliminate the infectious agents . A recent investigation has contributed to explaining in part the reasons for which n-3 polyunsaturated fatty acids reduce host defense against L. monocytogenes: consumption of EPA or DHA (both contained in fish oil) impairs the production of IL-12 and IFN-gamma, cytokines that play an essential role in the innate and adaptive responses of host immune system . Hence, the reduction of IL-12 levels may explain the impaired bacterial clearance and the reduction of survival to L. monocytogenes infection . Another possible explanation for the reduction of host resistance is based on the inhibition of the expression of MHC class II (called Ia in mice) that is reduced in mice fed a fish oil diet and infected with L. monocytogenes . Similarly, it is important to note that these effects may be associated with the reported increase of reactive oxygen species, whose numbers were significantly increased in mice after the animals were experimentally infected with this bacterium and fed a diet containing hydrogenated coconut oil . In fact, the role of cellular oxidative processes requires considerable attention because they play an important role in the elimination of infectious agents. However, not all polyunsaturated fatty acids are capable of exerting the same effects, since many studies are not in accordance with the previously described arguments. Thus, a recent investigation has determined that conjugated linoleic acid does not alter the resistance of mice after L. monocytogenes infection . Similarly, a previous study indicated the irrelevant effect of dietary lipid manipulation on the survival of mice in two models of peritonitis, one with Pseudomonas aeruginosa and the other with Salmonella enterica serovar Typhimurium . Irrespective of these findings, other experimental studies have demonstrated the adverse effects of dietary lipids containing n-3 or n-6 polyunsaturated fatty acids in the impairment of immune functions in animals experimentally infected with different pathogenic agents such as S. enterica serovar Typhimurium , P. aeruginosa , Staphylocococcus aureus , or Mycobacterium tuberculosis . By contrast, it has been reported that the administration of diets containing fish oil does not reduce the survival percentage after experimental infection with Klebsiella pneumoniae . In fact, dietary fish oil supplementation increased resistance to infection due to an enhancement of IL-1 and TNF production by peritoneal cells in mice . It is probable that the discrepancy in these findings is due to different factors that include the duration of the supplementation period, the concentration of fatty acids used, or the cell population affected by fatty acid administration. Finally, experimental observations with viruses have also demonstrated that influenza virus infection in animals fed fish oil diet delays the clearance due to an impairment of primary virus-specific T-cell cytotoxicity but has no effect on NK cytotoxicity . CONCLUDING REMARKS : Nutritional status is generally recognized as an essential factor involved in the modulation of immune response, which may be determinant in the development of the clinical effects derived from a malnutrition process . Thus, in recent years the effects that different dietary lipids exert upon immune functions have received considerable attention, because different functions of immune system are altered after dietary lipid administration. Hence, lymphocyte proliferation, cytokine production, phagocytic activity, adhesion molecule expression, and NK cell activity are susceptible to modification by the action of certain lipids in both animals and humans. Different mechanisms of action have been proposed to be involved in these processes, such as lipid peroxidation, changes in the plasma membrane, eicosanoid production, or alteration of gene expression. Both epidemiological and experimental studies have applied the beneficial properties associated to a modulation of the immune functions. Thus, different fats such as fish oil or olive oil have been applied in the amelioration of symptoms related to rheumatoid arthritis and other inflammatory disorders due to their anti-inflammatory properties. Similarly, they have also been applied as biological mediators capable of exerting a protective effect against colon or breast cancer, whereas other fatty acids such as n-6 polyunsaturated fatty acids exert a chemopromotive role in tumorigenic processes. In addition, these effects may be related to the ability of several fatty acids to modulate apoptosis by different pathways. Nevertheless, the immunosuppressant role associated with dietary polyunsaturated fatty acids containing fish oil may produce adverse effects because they reduce the host's natural resistance against pathogenic agents. Therefore, in the light of current experimental observations, the benefits of administering these dietary lipids should be weighed before they are applied as immunosuppression factors in order to prevent detrimental or adverse effects caused by an excessive suppression of immune functions. FIG. 1. : Schematic representation of potential role of dietary lipids and biological and clinical consequences of the administration of several dietary lipids. Schematic representation of potential role of dietary lipids and biological and clinical consequences of the administration of several dietary lipids. PUFAs, polyunsaturated fatty acids; MUFAs, monounsaturated fatty acids; SFAs, saturated fatty acids. Backmatter: PMID- 12204975 TI - Centrifugation of Human Lung Epithelial Carcinoma A549 Cells Up-Regulates Interleukin-1beta Gene Expression AB - Keywords: Cytokines are important mediators for host defense and inflammation to bacterial infections. Chlamydia pneumoniae, an obligate intracellular bacterium that causes acute respiratory infections in human hosts, has been shown to induce cytokine expression in various human and murine model systems . In most in vitro studies of C. pneumoniae, a centrifugation step is employed, and it has been demonstrated that the addition of this centrifugation step can increase the formation of inclusion bodies . However, the stress of centrifugation may also affect some cellular activities, such as proliferation and gene expression; certain cytokines, such as interleukin-1beta (IL-1beta), can also be induced under stress conditions . Therefore, the effects of centrifugation of cells should be considered when performing in vitro studies of C. pneumoniae, particularly studies using lung epithelial cells, which serve as a major target for C. pneumoniae infection. We intended to determine if centrifugation stress can affect the expression of proinflammatory cytokine IL-1beta in a human lung epithelial carcinoma A549 cell line. Centrifugation of A549 cells for 1 h at 2,000 rpm (similar760 x g) at room temperature did not affect cell proliferation or cell viability. Furthermore, C. pneumoniae infection with or without centrifugation (direct inoculation) also did not alter cell proliferation or viability. The effect of centrifugation stress on IL-1beta expression was then examined by reverse transcription-PCR (RT-PCR). The primers used were 5'-AAA CAG ATG AAG TGC TCC TTC CAG G-3' and 5'-TGG AGA ACA CCA CTT GTT GCT CCA-3' . beta-Actin was used as an amplification control, detected by primer sequences 5'-CGG GAC CTG ACT GAC TAC-3' and 5'-GAA GGA AGG CTG GAA GAG-3' . As shown in Fig. , the stress of centrifugation clearly increased the expression of the IL-1beta gene. This increase in expression of IL-1beta was present even at 24 h after centrifugation. FIG. 1. : Centrifugation stress up-regulates IL-1beta gene expression. Centrifugation stress up-regulates IL-1beta gene expression. A549 cells were centrifuged at 427 x g, 760 x g, or 1,187 x g for 1 h at room temperature. Four hours later, total RNA was extracted, and cellular levels of IL-1beta were measured by RT-PCR. beta-Actin was used as internal control (Con). (A) Electrophoretogram of RT-PCR. The RT-PCR products were run on agarose gels, stained with ethidium bromide, and visualized using UV. (B) Densitometric quantification of the electrophoretogram in panel A after normalization to beta-actin. The inclusion bodies of C. pneumoniae in both centrifuged and uncentrifuged A549 cells were also counted, and the inclusion body formation units (IFU) were calculated as (3.58 +- 0.61) x 108 and (1.03 +- 0.49) x 108 IFU/ml, respectively. There was a significant difference between the two values as determined by the Student t test (P < 0.01), suggesting that centrifugation does increase the formation of inclusion bodies. However, even without the centrifugation, C. pneumoniae could still efficiently infect A549 cells. Although centrifugation has been shown to improve inclusion body formation, it is always better to eliminate as many confounding factors as possible that may interfere with the host-pathogen interactions. In this case, we have shown that the stress of centrifugation alone can up-regulate the expression of the IL-1beta gene without affecting cell proliferation and viability. Thus, the IL-1beta induction data from some earlier studies where centrifugation was used to infect cells may be compromised. Therefore, reevaluation of earlier data to identify the true effects of C. pneumoniae infection on cytokine expression may be warranted. Although we did not test the effects of centrifugation on cytokines other than IL-1beta, they may also be affected by centrifugation. Cellular responses to stress are very complex, and different cell types have differing responses to the same stressor. Thus, it is noted that the results we observed using A549 cells may or may not be applicable to other cell types. Proper controls should always be run to clarify the role that the stress of centrifugation plays in cellular cytokine gene expression. Backmatter: PMID- 12204947 TI - Impaired Macrophage Phagocytosis of Apoptotic Neutrophils in Patients with Human Immunodeficiency Virus Type 1 Infection AB - Dysfunction of neutrophils (polymorphonuclear leukocytes [PMNL]) and macrophagic cells occurs as a consequence of human immunodeficiency virus type 1 (HIV-1) infection. Macrophages contribute to the resolution of early inflammation ingesting PMNL apoptotic bodies. This study investigated macrophage ability to phagocytose PMNL apoptotic bodies in patients with HIV-1 infection in comparison with uninfected individuals and the effect of HIV Nef protein on apoptotic body phagocytosis to determine if phagocytic activity is impaired by HIV infection. Monocytes/macrophages were isolated from 10 HIV-1-infected patients and from five healthy volunteers, whereas PMNL were isolated from healthy volunteers. Macrophage phagocytosis of apoptotic PMNL was determined by staining of apoptotic bodies with fluorescein-conjugated concanavalin A or with fluorescein-labeled phalloidin. Our data show significant impairment of PMNL apoptotic body macrophage phagocytosis in subjects with HIV-1 infection presenting a concentration of CD4+ T lymphocytes of >200/mm3 and in particular in those with <200 CD4+ T lymphocyte cells/mm3. In addition, HIV-1 recombinant Nef protein is able to decrease phagocytosis of apoptotic PMNL from normal human macrophages in a dose-dependent manner. The results of our study suggest that impaired macrophage phagocytosis of PMNL apoptotic bodies may contribute to the persistence of the inflammatory state in HIV-infected subjects, especially during opportunistic infections that are often favored by defective phagocytic activity. Keywords: Introduction : Neutrophil (polymorphonuclear leukocytes [PMNL]) function, including chemotaxis, phagocytosis, oxidative burst capacity, and bacterial killing, is impaired in the course of human immunodeficiency virus type 1 (HIV-1) infection, particularly in the later stages of the disease, and this abnormal function may predispose to some secondary bacterial infections and/or to opportunistic infections . PMNL have the shortest half-life of all circulating leukocytes and are programmed to die within 1 day. These aging leukocytes spontaneously undergo apoptosis and are recognized and phagocytosed by macrophages . Pitrak et al. have demonstrated that the rate of PMNL apoptosis is accelerated in AIDS patients, and this defect is intrinsic and not an effect of endogenous serum factors. Moreover, it has been proposed that the ingestion of apoptotic PMNL triggers production of anti-inflammatory mediators from macrophages , whereas persistent PMNL-rich inflammatory infiltrates have been associated with unresolved inflammatory reactions, including adult respiratory distress syndrome and rheumatoid arthritis . Thus, the removal of apoptotic cells appears to be critical in the resolution of inflammation. We have previously demonstrated that macrophages from HIV-positive subjects have a reduced ability to phagocytose Candida albicans cells, and there is a significant decrease in oxidative processes for the intracellular killing. These phenomena seem to be induced, at least in part, by HIV Nef protein . Since the effects of macrophage phagocytosis of apoptotic PMNL have not been completely investigated, especially in HIV-positive subjects, the purpose of this study was to evaluate phagocytosis of PMNL apoptotic bodies performed by macrophagic cells obtained from HIV-1-positive subjects and in parallel by the macrophages obtained from healthy individuals. Furthermore, we studied the effect of Nef protein on PMNL apoptotic body macrophagic phagocytosis, since this viral protein is able to depress both specific and nonspecific immune responses in HIV-infected patients, particularly microbial phagocytosis . MATERIALS AND METHODS : Subjects. | Ten HIV-1-infected subjects (mean age, 35.3 +- 5.8 years) were enrolled and five healthy volunteers (mean age, 37.1 +- 4.4 years), without HIV-1 risk factors, served as controls. Five of the HIV-1-infected subjects presented more than 200 CD4+ T lymphocytes/mm3 (mean = 517 +- 225), had a CD4/CD8 T-cell ratio of 0.5 +- 0.1, and had mean HIV-1 RNA levels in plasma of 32,628 +- 42,188 copies/ml. Five patients had less than 200 CD4+ T lymphocytes/mm3 (mean = 101 +- 70), a ratio of 0.2 +- 0.2, and mean HIV-1 RNA levels of 375,000 +- 246,815/ml. All these patients had a moderate anemic state and were generally studied before receiving antiretroviral therapy. In fact, a significant part of our study population included individuals who had ignored their seropositive condition for a long time and in consequence came to medical evaluation late. Monocyte and PMNL isolation and apoptotic body preparation. | Monocytes obtained from peripheral blood of healthy subjects and HIV-1-infected patients were selected as adhering cells after separation of peripheral blood mononuclear cells with a Ficoll Paque gradient (Pharmacia, Uppsala, Sweden). After repeated washes, the adherent cells were harvested by trypsinization and resuspended at a concentration of 106/ml in RPMI-1640 medium (Gibco, Paisley, Scotland), supplemented with 10% fetal calf serum (Celbio, Milan, Italy). Then, macrophage colony-stimulating factor (10 ng/ml; Genzyme, Milan, Italy) was added to the medium in order to differentiate monocytic cells into activated macrophages, and the cultures were maintained at 37C with 5% CO2. Cellular cultures were enriched for macrophages by adherence and repeated trypsinization and, to obtain a pure population, by phenotypic evaluation (more than 98% of cells). After 7 days of incubation in flasks (25 cm2; Nunc Kamstrup, Baltimore, Md.), the cells were harvested by trypsin treatment and transferred, at a final concentration of 5 x 105/ml, to four-well chamber slides (Lab-Tek; Nunc). Cultures of PMNL were obtained from peripheral blood of healthy subjects by separation with a Polymorphoprep gradient (Pharmacia). PMNL cells spontaneously go to apoptosis after 24 h of culture, and this method was adopted to induce apoptosis. In order to separate the apoptotic bodies of PMNL, after centrifugation at 600 rpm for 10 min (to remove nonapoptotic cells), the supernatants were centrifuged at 3,000 rpm, and the pellets containing PMNL apoptotic bodies were resuspended at the indicated concentrations. Fluorescent microscopy for phagocytosis detection. | Two staining methods were used to detect apoptotic bodies of PMNL in macrophagic cells. In the first method, 5 x 106 apoptotic bodies/ml were incubated for 20 min at room temperature with fluorescein-conjugated concanavalin A (ConA; Calbiochem Corp., La Jolla, Calif.) at a final concentration of 150 mug/ml. Then the apoptotic bodies were washed twice by centrifugation for 10 min at 3,000 rpm and resuspended in RPMI-1640 with 10% fetal calf serum at a concentration of 5 x 106/ml to obtain a ratio of 5:1 per macrophagic cell (this preparation was diluted 1:2 in the wells containing macrophagic cells). This method allows uniform staining of the apoptotic body surface, making the apoptotic bodies easily detectable in macrophagic cells without having any significant influence on phagocytosing activity. In the second method, fluorescein-labeled phalloidin (Sigma-Aldrich Chemicals, Milan, Italy) was used at a final concentration of 5 mug/ml. This product is able to stain the interior of the apoptotic bodies, binding firmly to the microtubular structures. This staining method was used like the first one. In a subsequent set of experiments, macrophage phagocytosis of PMNL apoptotic bodies was also investigated in the presence of various concentrations of recombinant Nef protein (American Biotechnologies, Inc., Cambridge, Mass.), using macrophagic cells obtained from four healthy individuals and performing the experiments in triplicate. The challenge with stained apoptotic bodies, using both the first and second methods, was made at a ratio of 5:1. After 2 h of incubation in the presence of apoptotic bodies, the supernatants were removed, and slides were detached from their supports and washed twice with phosphate-buffered saline (PBS). Then the slides were fixed with 10% ethyl alcohol solution, mounted in phosphate-buffered glycerol (30% PBS and 70% glycerol, vol/vol), and examined under a fluorescence microscope at x400 by four blinded microscopists. At least five microscopic fields were observed for each sample. Macrophage oxidative phenomenon evaluation. | We indirectly studied the oxidative phenomena of macrophages in the absence or presence of Nef protein, by evaluating the antioxidant power of the supernatants, as previously described . For this purpose we used the antioxidant power (PAO) kit furnished by Med. Dia S.r.l. (San Germano, Vercelli, Italy). The test is based on the detection of Cu+ ions produced by the reduction of a known amount of Cu2+; this reduction is induced by the antioxidant factors present in the culture, the activity of which decreases in an inversely proportional manner to the activity of macrophage oxidative phenomena. The concentration of Cu2+ ion was detected through the formation of complexes consisting of Cu+ and the chelating agent betacuproine disulfonate (2,9-dimethyl-4,7-diphenyl-1,10-phenanthrolinedisulfonate), and it was measured spectrophotometrically at 490 nm (Metertech spectrophotometer; Medical System, Genoa, Italy). The data are expressed as microequivalents per liter of reducing equivalents +- standard deviation, corresponding to PAO units. Statistical analysis. | The results were expressed as mean and standard deviation. The differences among the experimental and control groups were statistically evaluated using Student's t test. Statistical significance was defined as P <= 0.05. Moreover, a linear correlation test was employed for the correlation study reported in the results. RESULTS : In the first set of experiments, we studied the phagocytosis of PMNL apoptotic bodies stained with fluoresceinated ConA from macrophages of patients with HIV-1 infection and from those of uninfected healthy individuals. A significant decrease in the percentage of phagocytosing cells was observed in HIV-1-positive subjects presenting a CD4+ T-cell concentration of >200 cells/mm3 (24.2% +- 12.3%; P = 0.002) and in those with <200 CD4+ T cells/mm3 (23.7% +- 6.9%; P < 0.001), compared to healthy control subjects (50.1% +- 3.7%) . We point out that by analyzing all the HIV-positive subjects, the mean number of phagocytosing cells was 23.9% +- 10% in 52% of the controls (P < 0.001). Since fluoresceinated ConA only stains the surface of apoptotic bodies, we performed further experiments with fluoresceinated phalloidin, which is endowed with a specific ability to bind to microtubular structures. We also observed a reduction of about 50% in phagocytosing cells in HIV-positive subjects compared to the controls (unreported data). As reported in Table , a significant decrease in the number of PMNL apoptotic bodies phagocytosed per macrophagic cell was observed in HIV-1 subjects with >200 CD4 T cells/mm3 (0.6 +- 0.4; P = 0.001) and in those with <200 CD4 T cells/mm3 (0.3 +- 0.1; P < 0.001) compared to the controls (2.4 +- 0.7). Evaluating all the HIV-positive subjects, the mean number of apoptotic bodies phagocytosed per cell was 0.45 +- 0.3 (P < 0.001). As can be seen in Table , the reduction of the mean number of PMNL apoptotic bodies phagocytosed per cell in HIV-positive subjects compared to the controls is fivefold, and the percent reduction in phagocytosing cells in infected patients is twofold. This observation suggests a decreased ability of single phagocytosing cells to swallow apoptotic bodies. In a further set of experiments, we studied the phagocytic activity of PMNL apoptotic bodies in normal human macrophages preincubated with recombinant Nef protein. The results obtained demonstrate that this protein is able to inhibit the phagocytosis of PMNL apoptotic bodies. In particular, preincubation with Nef (2 h before the challenge with apoptotic bodies) inhibited macrophage phagocytosis of PMNL apoptotic bodies by about 50% (P = 0.01) at the concentration of 1 mug/ml and 65% at the concentration of 2 mug/ml (P < 0.001). When the preincubation time was prolonged to overnight, inhibition of macrophage phagocytosis was 80% at a Nef protein concentration of 2 mug/ml. Table reports the mean number of PMNL apoptotic bodies stained with fluoresceinated phalloidin phagocytosed per cell in the absence or presence of Nef protein. As shown, the reduction was about 34%, 90%, and 92% with Nef at 0.5, 1, and 2 mug/ml, respectively. Consequently, in these experiments the reductions induced by Nef were also more evident when evaluating the number of PMNL apoptotic bodies phagocytosed per cell than the percentage of phagocytosing cells. Moreover, we observed a negative and significant correlation between the Nef protein concentration and the number of PMNL apoptotic bodies phagocytosed per cell (r = -0.65; P = 0.0221). The mean number of apoptotic bodies phagocytosed related to 106 macrophages was 1.18 x 106 in the absence of HIV infection (23% of the total) and 48,000 with a viral load in plasma of 8 x 105 HIV-1 RNA copies/ml (about 1%). A negative correlation was found between the HIV RNA load and the number of apoptotic bodies phagocytosed (r = -0.623; P < 0.05). Finally, we evaluated the antioxidant power (PAO) of the macrophage supernatants in the absence or presence of various concentrations of Nef, and we found a significant increase in PAO that meant a reduction of oxidative phenomena in the presence of the viral protein compared to the control values, especially at a concentration of 2 mug/ml (PAO units = 164.1 +- 8.0 with Nef at 2 mug/ml versus 105.6 +- 12.9 without Nef; P < 0.001). TABLE 1 | Percentage of macrophages phagocytosing PMNL apoptotic bodies stained with fluoresceinated ConA: ex vivo test of macrophages from uninfected and HIV-1-infected subjects TABLE 2 | Number of PMNL apoptotic bodies phagocytosed per macrophage TABLE 3 | Effect of Nef viral protein on phagocytosis of PMNL apoptotic bodies DISCUSSION : The results of this study show that phagocytosis of PMNL apoptotic bodies by macrophages of HIV-1-infected patients is impaired and that Nef, a regulatory viral protein of HIV, is able to decrease phagocytosis of PMNL apoptotic bodies by human normal macrophages. In addition, macrophagic oxidative phenomena are depressed by Nef protein. This is relevant because these effects are involved in intracellular killing processes and consequently in the destruction of phagocytosed particles . The decrease in oxidative processes and the inhibition of macrophagic function seem to occur in parallel. Aging PMNL spontaneously undergo apoptosis and are recognized and phagocytosed by monocytes and macrophages . PMNL function is impaired in all stages of HIV-1 infection and especially in the terminal stage of the disease . Pitrak et al. have shown that abnormalities of PMNL function observed in HIV-positive subjects might partly depend on the accelerated apoptosis induced by HIV infection. In fact, during all stages of HIV-1 infection, there is an increased number of apoptotic PMNL which are unable to function as host defenders . The removal of apoptotic cells appears to be central to the resolution of inflammation. In fact, the clearance of apoptotic PMNL not only prevents the release of toxic and immunogenic intracellular contents, but also stimulates the macrophages to produce inflammatory mediators, including transforming growth factor beta1, prostaglandin E2, and platelet-activating factor, and inhibits the production of tumor necrosis factor, interleukin-1beta, and interleukin-8 . Dysregulation of PMNL function, along with that of monocytic macrophagic cells, in HIV-1-infected patients is reflected in the increased incidence of some microbial infections among these patients . During several microbial infections, PMNL migrate and accumulate at the inflammatory sites, followed by removal of inflammatory cells. This occurs mainly by apoptosis and by phagocytosis of apoptotic bodies. In parallel, some antimicrobial pathogens can be phagocytosed and killed by macrophages. These phenomena appears to be critical to the resolution of inflammation and infection. However, the decreased phagocytosis of apoptotic PMNL by macrophages in HIV-1-infected patients and the accelerated apoptosis of PMNL lead to accumulation of apoptotic inflammatory PMNL and a decrease in their clearance. The persistence of apoptotic PMNL and their apoptotic bodies at the inflammatory site may maintain the inflammatory state through persistent stimulation of proinflammatory cytokines . This can explain some of the pathological conditions in the gastrointestinal tract or bronchoalveolar tract that have been reported . Backmatter: PMID- 12204948 TI - Evaluation of an In-House-Developed Radioassay Kit for Antibody Detection in Cases of Pulmonary Tuberculosis and Tuberculous Meningitis AB - A radioassay for the detection of antitubercular antibody has been developed. The technique involves the addition of 125I-labeled Mycobacterium tuberculosis antigen as a tracer, diluted clinical sample (serum or cerebrospinal fluid [CSF]), and heat-inactivated Staphylococcus aureus to capture the antibody, incubation for 4 h, and quantitation of the amount of antibody present in the sample. A total of 330 serum samples from patients with pulmonary tuberculosis and 138 control serum samples from individuals who were vaccinated with M. bovis BCG and from patients with pulmonary disorders of nontubercular origin were analyzed. Also, 26 CSF samples from patients with tuberculous meningitis and 24 CSF samples as controls from patients with central nervous system disorders of nontuberculous origin were analyzed. Sensitivities of 80 and 73% were observed for patients with pulmonary tuberculosis and tuberculous meningitis, respectively, and specificities of 90 and 88% were seen for the two groups of patients, respectively. The sensitivity was lower, however, for human immunodeficiency virus-infected patients coinfected with M. tuberculosis. The control population could be differentiated from the patient population. This assay is rapid and user friendly and, with its good sensitivity and specificity, should benefit the population by providing diagnoses early in the course of disease and, hence, permit the early administration of appropriate chemotherapy. Keywords: Introduction : Tuberculosis, caused by Mycobacterium tuberculosis, represents a major health problem worldwide. It is estimated that almost one-third of the entire world's population is exposed to this disease . In 1993, the World Health Organization declared tuberculosis a global emergency and estimated that by the start of the new millennium there would be more than 1 billion newly infected individuals and more than 17 million deaths due to this disease if control programs were not implemented properly. These control programs are dependent on good tests for early diagnosis, constant monitoring of disease status, and response to and completion of treatment. The diagnosis of tuberculosis has never been simple, especially in children and in patients with extrapulmonary tuberculosis, from whom clinical specimens for microscopy and culture are not easily obtained. Even in patients with pulmonary tuberculosis, diagnosis by microscopy is rather insensitive and culture is time-consuming. Since the mid-1970s radioimmunoassay and enzyme-linked immunosorbent assay (ELISA) techniques have been developed for the detection of M. tuberculosis antigen and anti-M. tuberculosis antibody from clinical specimens, and these assays have various degrees of sensitivity and specificity (, , , 7, , , -, , -, , ). Almost all the assays developed at the Laboratory Nuclear Medicine Section of the Bhabha Atomic Research Centre require the detection of both antigen and antibody from clinical samples (, , -, -). With the incidence of tuberculosis being on the rise and with the added risk of coinfection with human immunodeficiency virus (HIV), the need at this time is for a test with good sensitivity and specificity for rapid and early diagnosis. With this as the aim, we have modified the antibody test and describe here the evaluation of a single test kit for the diagnosis of tuberculosis. MATERIALS AND METHODS : Antigens. | M. tuberculosis H37Rv was grown in Youman's synthetic medium for 4 to 6 weeks. Cultures were heat inactivated by autoclaving at 120C for 20 min, followed by sonication for extraction of antigen from the cells. The suspension was ultracentrifuged at 105,000 x g for 90 min, and the supernatant was used as the source of antigen. This sonicate antigen was iodinated with 125I by the iodogen method . The iodinated antigen was purified from free iodide by chromatography on a Sepharose 6B column. The immunoreactive peak was used as the radioactive antigen or tracer. Antibody. | Rabbit anti-M. bovis BCG antibody (M/S Dakopatts, Copenhagen, Denmark) was commercially obtained and was used for the preparation of quality control (QC) samples. Solid phase. | Staphylococcus aureus (strain Cowan I) was used as the solid phase. S. aureus has protein A on its surface and can bind to immunoglobulins (immunoglobulin G [IgG]) efficiently through the Fc region. The S. aureus cultures were grown in nutrient medium for 3 to 4 days. After the cells were harvested, they were heat inactivated at 80C for 40 min. A 10% suspension was made and was used as the solid phase. Assay procedure. | The antibody assay developed is a single-step assay. A total of 0.1 ml of clinical sample (a serum sample diluted 1:25 or a neat cerebrospinal fluid [CSF] sample) whose antibody levels are to be determined was incubated with 0.1 ml of an S. aureus suspension diluted 1:3 and 0.3 ml of 125I-labeled M. tuberculosis antigen. Assay control samples and QC samples were also treated in the same way as the patient samples. The reaction mixture was incubated for 4 h at room temperature with constant shaking. Subsequently, 2 ml of 0.02 M barbitone buffer (pH 7.6) was added, and the entire mixture was centrifuged at 2,000 x g for 40 min. The supernatant was aspirated, the pellet was counted in a gamma counter, and the antibody concentration in the samples was determined by calculating the ratio as described in the Results. Determination of assay cutoff and assay control samples. | In order to determine a cutoff limit for negativity above which the clinical samples could be classified as positive, a standard curve was constructed by using anti-BCG IgG as the standard, with concentrations ranging from 5 to 200 mug/dl, as shown in Fig. . The assay was performed as described above with 77 control serum samples. These samples were taken from healthy individuals and patients with pulmonary infections of nontubercular nature. Antibody levels were quantitated by referring to the standard curve and are expressed as micrograms per deciliter. The values obtained were calculated for the upper limit of negativity as the mean + 3 standard deviations (SDs), which was found to be 25 mug of IgG per dl, as shown in Fig. . With this as the basis, two assay control samples were included in the kit and for calculation purposes. Control C1, which represented nonspecific binding, was the assay buffer with no antibody, and control C2 represented a concentration of 25 mug of anti-BCG IgG per dl in the same assay buffer. Stability of the kit. | The stability of the kit was determined for a period of 7 weeks both at room temperature (RT) and at 4C. In this case, the kits were kept at RT and at 4C and were evaluated at 0, 1, 2, 3, 4, and 7 weeks. QC samples Q1 and Q2 were also included to determine the reproducibility of the assay. Sample selection. (i) Pulmonary tuberculosis. | Three hundred and thirty blood samples were collected from patients attending a tuberculosis clinic. The patients had pulmonary tuberculosis at various stages of the disease and were at various stages of treatment. For all patients, the diagnosis of tuberculosis was made on the basis of the following criteria, determined by taking a detailed history: (i) presenting clinical signs and symptoms such as cough and fever; (ii) BCG vaccination status, past history of tuberculosis, or presence of a close contact with tuberculosis; (iii) routine laboratory tests which included assay of sputum for acid-fast bacilli (AFB), culture of sputum, skin test, and chest X ray; and (iv) response to antituberculosis treatment, as against specific treatment for chronic obstructive pulmonary disease (COPD) and asthma. On the basis of the criteria described above, the patients were diagnosed as having tuberculosis and classified into the groups described below. Group I comprised 86 patients who were positive by both smear for AFB and culture. Of these, 20 patients had fresh cases of tuberculosis, 14 had been treated for tuberculosis with antituberculosis drugs for less than 6 months, 11 had been treated for 6 months to 1 year, and 33 had been treated for more than 1 year; data on the treatment given were not available for 8 patients. Group II comprised 126 patients who were negative by smear for AFB but positive by culture. In this group, 23 patients had fresh cases of tuberculosis, 15 had been treated for less than 6 months, 16 had been treated for 6 months to 1 year, and 57 had been treated for more than a year; data on the type of treatment given were not available for 15 patients. Group III comprised 118 patients who were negative by both smear for AFB and culture. However, 99 of these patients had previously received a diagnosis of tuberculosis on the basis of a positive smear for AFB and culture and were receiving antituberculosis treatment. Nine patients had been treated with antituberculosis drugs for less than 6 months, 14 had been treated for 6 months to 1 year, and 45 had been treated for more than 1 year; data on the type of treatment given were not available for 31 patients. Only 19 patients in this group had fresh cases of tuberculosis, and diagnosis was made on the basis of a combination of tests and criteria such as X ray, skin test, and presence of a contact. Further analysis indicated that 33 of the 330 tuberculosis patients were HIV positive. Of these, 7 patients were in group I, 16 were in group II, and 10 were in group III. The control group comprised 138 individuals between the ages of 21 and 50 years who had no prior history of tuberculosis but who had all been vaccinated with BCG. Thirty-one of these were healthy volunteers, 75 had bronchitis leading to COPD, and 28 were asthma patients. Diagnosis of COPD and asthma was done on the basis of the criteria given earlier, such as X ray, sputum cultures, clinical presentation, and response to antibacterial treatment. COPD patients were treated with aminophylline and theophylline, and asthma patients were treated with bronchodilators. (ii) Extrapulmonary tuberculosis. | Fifty CSF samples were collected from individuals with signs and symptoms of meningitis and other neurological disorders. Of these, 26 were diagnosed as having tuberculous meningitis on the basis of the results of clinical and biochemical investigations. Twenty-four of the patients had central nervous system (CNS) disorders of nontuberculous origin. FIG. 1. | (a) Standard curve for the antibody assay; (b) determination of the cutoff value. (a) Standard curve for the antibody assay; (b) determination of the cutoff value. RESULTS : Calculation. | The amount of antibody in the clinical samples was determined by using the two assay controls (C1 and C2) included in the kit. Control C1 represented nonspecific binding with no antibody, and control C2 represented an anti-BCG IgG concentration of 25 mug/dl. The ratio X (as a percentage) was calculated by using the counts in C1, C2, and the unknown, as follows: [(counts in unknown - counts in control C1)/(counts in control C2 - counts in control C1)] x 100. Samples that gave values of more than 100 were considered positive. Stability of the kit. | In order to ascertain the stability of the kit, the reproducibilities of values for the QC samples in the kit were determined. The expected value of the ratio for QC Q1 was 200, and the mean observed values over a period of 7 weeks were 198 and 205 at RT and 4C, respectively. The SD and percent coefficient of variation (CV) were 22.8 and 11.5%, respectively, at RT and 18.7 and 9.1%, respectively, at 4C. The expected value of the ratio for the second QC, Q2, was 400, with mean values of 344 and 355 at RT and 4C, respectively. The SD and CV were 50.8 and 14.7%, respectively, at RT and 41.9 and 11.8%, respectively, at 4C. These CVs were within the statistically acceptable range for interassay variability. These results are represented graphically in Fig. Pulmonary tuberculosis. | Table presents the antibody responses in patients with pulmonary tuberculosis with respect to the period of treatment given, and Table presents the antibody responses in patients with HIV and M. tuberculosis coinfection. It was observed that 14 of the 138 control samples had antibody levels above the ratio cutoff value of 100, contributing to a rate of nonspecific detection of 10% and a specificity of 90%. Group I comprised 86 patients who were positive by both smear and culture and had 69 patients who had high antibody ratios, giving the test a sensitivity of 80.2% and a positive predictive value of 83.1%. Of the 7 HIV-positive patients in this group, 4 patients had high titers of antitubercular antibody. Group II had 126 patients who were negative by smear but positive by culture. Of these, 92 were positive for antibody, giving the test a sensitivity of 73% and a positive predictive value of 86.8%. Six of the 16 patients in this group were positive for both HIV and M. tuberculosis and had high titers of antibodies against M. tuberculosis. Group III comprised 118 patients who were negative by both smear and culture and had 71 patients with high antibody titers, giving the test a sensitivity of 60.2% and positive predictive value of 83.5%. Three of the 10 HIV-positive patients in this group also had high antibody levels. The detailed analysis of each subgroup is presented in Tables and 2. Figure provides a graphical representation of the response ratios for the different groups of patients with pulmonary tuberculosis. Compared to the antitubercular antibody levels in the control group, the patient groups had high levels of antitubercular antibody. The patient population was further classified as those who had received antituberculous treatment at the time of sample collection and those who had not received any treatment at the time of sample collection (fresh cases). The test showed a sensitivity of 62.9% and a positive predictive value of 73.6% for the patients with fresh cases. For the treated patients, the sensitivity was 72% and the positive predictive value was 93.2%, as depicted in Table . Extrapulmonary tuberculosis. | Table represents the percent positivity of antibody detection in patients with tuberculous meningitis. Of the 26 patients with clinically proven tuberculous meningitis, 19 had high antibody levels by this test, giving a sensitivity of 73%. A false-positivity rate of 12% was observed for the control group, giving a specificity of 88% and a positive predictive value of 91%. Figure is a graphical representation of the antibody response ratios for patients with CNS disorders of tuberculous and nontuberculous origin. FIG. 2. | Stability of the kit at RT (a) and at 4C (b). Stability of the kit at RT (a) and at 4C (b). *, C1; , C2; , QC1; , QC2. FIG. 3. | Antibody levels in patients with pulmonary tuberculosis. Antibody levels in patients with pulmonary tuberculosis. , group I; , group II; , group III; *, group IV (controls). FIG. 4. | Antibody levels in patients with tuberculous meningitis. Antibody levels in patients with tuberculous meningitis. , patients with tuberculous meningitis; , controls (patients with CNS disorders of a nontubercular nature). TABLE 1 | Antibody response in pulmonary tuberculosis TABLE 2 | Antibody responses in patients with HIV and M. tuberculosis coinfection TABLE 3 | Predictive values for treated and nontreated patients TABLE 4 | Antibody response in patients with tuberculous meningitis DISCUSSION : The need for a serological test for the diagnosis of tuberculosis is emphasized on the grounds of potential benefits for rapid and differential diagnosis, particularly in culture- and/or smear-negative patients. During bacterial diseases like tuberculosis, the immune response leads to a rise in the titers of antibodies against different antigenic determinants of the organisms, and hence, the specific antibodies present may range from those that are species specific to those that share certain common antigenic determinants with the causative organisms. The antibody response in tuberculosis has been studied for a long time (, , , , 13, , -, , -, , ). Although one of the major drawbacks, especially in serological tests, is the persistence of antibodies even after effective treatment, measurement of the antibody response to 38-kDa antigen has been found to be useful in assessing the effectiveness of treatment in patients with tuberculous meningitis . A radioimmunoassay for the detection of M. tuberculosis antigen and anti-M. tuberculosis antibodies developed at the Laboratory Nuclear Medicine Section of the Bhabha Atomic Research Centre has been tested with samples from patients with various clinical manifestations of tuberculosis (, , , , -). It was demonstrated that detection of antigen and antibody gave a good sensitivity and specificity for diagnosis. It was also shown that antibody detection alone could be useful for the diagnosis of tuberculous meningitis. Antigen and antibody detection in patients with pulmonary tuberculosis was done from immune complexes, not serum. In this particular study, antibody levels in serum samples from patients with pulmonary tuberculosis were measured. The aim of the present work was to make the test user friendly and evaluate detection of antibody from serum rather than immune complexes in the case of pulmonary tuberculosis. The modified antibody assay described here is simple and rapid and involves the addition and incubation of all reagents together for 4 h at room temperature with a final centrifugation step. This not only reduces the pipetting errors and the number of centrifugation steps required but also reduces the incubation time, facilitating early diagnosis and hence the early institution of treatment. The most significant aspect of this test was that it was useful for the diagnosis of tuberculosis in those patients for whom the rapid test for AFB was negative, which included patients from group II and group III. Although a false-positivity rate of 10% was observed for the control population, this was acceptable, as India is an area with a high prevalence of tuberculosis and the entire population is vaccinated with BCG. Early diagnosis and early treatment of tuberculosis in patients coinfected with HIV are mandatory, as delayed therapy results in increased rates of mortality . In the present study, in the case of patients coinfected with M. tuberculosis and HIV, antibodies against M. tuberculosis persisted even after antituberculosis treatment. However, the percent positivity of antibody detection for the HIV-infected tuberculous patient population is lower than that for the entire tuberculous patient population (Table and Table ). These findings were similar to those presented in other reports, in which the sensitivity dropped from 62% for non-HIV-infected tuberculous patients to 28% for HIV-infected tuberculous patients . It has been reported that no antibody is detected by an immunochromatographic test (ICT) for tuberculosis in any of the serum samples of HIV-infected patients coinfected with M. tuberculosis or mycobacteria other than M. tuberculosis . Evaluation of the Mycodot immunodiagnostic test for the diagnosis of tuberculosis, which detects antibodies to lipoarabinomannan antigen, showed that it has very good specificity. The sensitivity of detection for HIV-negative tuberculous patients was 56% and much lower, as low as 25%, for HIV-positive patients . In general, serological assays have high negative values for HIV-positive individuals, low sensitivities, and low negative predictive values, which markedly decreases their utility in populations in which HIV infection is prevalent . Diagnosis of tuberculous meningitis has always been a challenge, especially in children. Data from a previous study in which the 38-kDa antigen in the ELISA system was used for the detection of antibodies in CSF samples showed sensitivities of 60% for patients proven to have tuberculosis postmortem, 80% for patients with culture-proven tuberculous meningitis, and 62.5% for patients with clinically proven tuberculous meningitis . The sensitivity in that study was superior to those reported elsewhere . Use of the 19-kDa antigen for the detection of antibodies in CSF of patients with tuberculous meningitis could clearly differentiate the patient group from the control group in studies with a European population, but no such differentiation was observed in the study with an Indian population . The modified antibody assay mentioned here has a fairly good sensitivity of 73% for patients with tuberculous meningitis and a specificity of 88%, and above all, it could differentiate between the patient group and control group in the Indian population. In Western countries, the prevalence of tuberculosis is low and a clear distinction between patient and control populations is possible. An ELISA with the 38-kDa antigen specific for the M. tuberculosis complex showed a sensitivity of 83% and a specificity of 99% . In a study with a Chinese population, however, the sensitivity of detection of IgG antibodies to the 38-kDa antigen was 64% and the specificity was 81% . Similarly, an ELISA performed with the Ag5 antigen to study the antibody response in patients with pulmonary tuberculosis showed a sensitivity of 84% and a specificity of 96% for populations in low-prevalence areas, while the sensitivity was 94% and the specificity was 80% for populations in high-prevalence areas . The same ELISA for the detection of antibodies used elsewhere showed sensitivities that varied from 49 to 89% and specificities that ranged from 88 to 98% . The A60 antigen, which is a thermostable component of the purified protein derivative, has been used in the serodiagnosis of tuberculosis. In patients with smear-negative but culture-positive pulmonary tuberculosis, measurement of both IgM and IgG by ELISA revealed a sensitivity of 68% and a specificity of 100% . In the pediatric group, at a chosen specificity of 98%, anti-A60 IgG antibodies were observed in 68% of children with clinically active tuberculosis . In the Indian population, an ELISA with the A60 antigen for the detection of IgM antibodies in patients with pulmonary tuberculosis had a sensitivity of 77.5% and a specificity of 87.5% , while quantification of antibodies in patients with neurotuberculosis showed a sensitivity of 63% and a specificity of 100% . When titers of antibodies to the 19-kDa antigen in the Indian population were measured by ELISA, the test was found to have a poor sensitivity (8%), but the same test used for a population in the United Kingdom was found to have a sensitivity of 58% and a specificity of 97% . An ELISA for the detection of antilipoarabinomannan antibodies in sera from patients with pulmonary, miliary, and pleural tuberculosis and pleural lymphadenitis and in CSF from patients with tuberculous meningitis had an overall sensitivity of 61 to 72% and a specificity of 72% . Among the HIV-negative patients with active tuberculosis, the sensitivity of antilipoarabinomannan antibody detection ranged from 21.5 to 89%, but the sensitivity of detection was generally poor, about 7 to 40%, for HIV-positive patients . In developing countries like India, where the prevalence of tuberculosis is high and where individuals are by and large vaccinated with BCG or exposed to environmental mycobacteria, antibodies are present even in the healthy control population, so that the cutoff for any diagnostic test is high. The antibody test with partially purified antigen developed in the study described here shows a sensitivity of 80% and a specificity of 90% for patients with pulmonary tuberculosis. These results are quite acceptable and compare well with those of tests from Western countries. Previous reports have indicated that ELISA kits for the detection of antibody for the diagnosis of tuberculosis have poor sensitivities and specificities . Clinical evaluation of the Andelisa IgG kit, which uses the A60 antigen derived from M. bovis BCG, gave a sensitivity of 48.3% and a specificity of 92%, while the Kreatech IgA kit, which uses the KP90 antigen derived from M. tuberculosis, showed a sensitivity of 49.7% and a specificity of 84% . Simultaneous evaluation of seven serological tests, i.e., two ICTs and five ELISAs, with the same clinical samples also showed poor to modest sensitivities that ranged from 16 to 57%. The specificities of these tests ranged from 80 to 97% with sera from the Mantoux test controls and 62 to 100% with sera from the anonymous controls . These results show lower sensitivities and specificities, indicating considerable overlap in antibody levels in the healthy control group and the patient group and indicating that none of the tests has utility for the diagnosis of tuberculosis. Evaluation of the validity of the ICT kit in a study with a Korean population showed a sensitivity of 73% for patients with fresh cases of tuberculosis and 87% for patients with reactivated tuberculosis, while the specificity was 88% for the healthy group and 94% for the nontuberculous patients and hospital workers . A similar evaluation of the ICT kit in Madagascar showed sensitivities of 68.2% for patients with pulmonary tuberculosis and 65.2% for patients with extrapulmonary tuberculosis and a specificity of 83.3% . However, evaluation of the ICT kit with the Indian population showed an overall sensitivity of 20% for patients in the first month of the disease (fresh cases) and a very low sensitivity of 4%. The test was also positive for 30% of cases of disease caused by mycobacteria other than M. tuberculosis, indicating cross-reactivity. The overall specificity of the test was 89% . The ELISA kit for the detection of IgG antibodies against tuberculous glycolipid antigen containing cord factor (TBGL Ag) had a sensitivity of 87% and a specificity of 100%. The other findings included in the report were an increase in the antibody titers after the start of antituberculosis chemotherapy and the persistence of antibodies in the circulation even after the completion of treatment . The ICT kit developed for detection of antibodies against the 38-kDa antigen of M. tuberculosis had an overall sensitivity of 70 to 92% and an overall specificity of 92 to 93% for patients with pulmonary tuberculosis. For patients with extrapulmonary tuberculosis, the overall sensitivity was 76% and the overall specificity was 92% . Results from a study conducted in the Europe demonstrated a low sensitivity of 50% but a high specificity of 100% . This low sensitivity will result in many false-negative results, whereby the negative predictive value will be unacceptably low even in countries with a high prevalence of tuberculosis . A specificity of 100% is not possible in developing countries, particularly in tropical areas where people are permanently in contact with various pathogens and develop cross-reacting antibodies, resulting in poor test specificities. Hence, every new serodiagnostic test should be validated with controls from countries where the test would be applied . In this context a cutoff of 100 was determined on the basis of the values obtained for the control population, which included contacts and patients with nontubercular disease. It was observed, however, that if the cutoff were raised to 125, the specificity would go up to 97% and the sensitivities would be 69.7, 66.6, and 51.6%, for groups I to III, respectively, as shown in Table , with a positive predictive value of 94%. Thus, it is possible to alter the cutoff and obtain appropriate sensitivity, specificity, and predictive values according to the prevalence of the disease in different areas. The antibody test kit described here involves very few steps in terms of the addition of reagents and a shorter incubation time, which make it user friendly. Above all it has a good sensitivity and a good specificity. This kit should benefit the population by providing diagnoses early in the course of disease and, hence, permitting the early administration of appropriate chemotherapy. Backmatter: PMID- 12204974 TI - Diagnostic Techniques To Detect Cryptic Leishmaniasis in Dogs AB - This study of several techniques for detecting cryptic leishmaniasis in dogs from areas in Spain where Leishmania infantum is highly endemic concludes that immunological techniques (enzyme-linked immunosorbent assay, immunofluorescence antibody test, Western blotting, delayed-type hypersensitivity reaction, and in vitro lymphocyte proliferation assay) do not clearly differentiate between noninfected and infected asymptomatic dogs and that culture and PCR are more reliable diagnostic tools. Keywords: Introduction : Seroepidemiological studies of canine leishmaniasis have revealed a large number of asymptomatic seropositive animals . Moreover, in areas where leishmaniasis is highly endemic, a high proportion of apparently healthy animals show low levels of anti-Leishmania antibodies. Serological follow-up of these animals has revealed that some of them are in the prepatent infection period, which will lead to increased antibody titers in subsequent blood extractions. Others have regressive forms of the disease, and their antibody levels will decrease in the following months or years; still others maintain low levels of antibodies without developing the disease for many years . However, the total number of infected animals is unknown. The detection of the extent of the infection, particularly among asymptomatic dogs, is of great importance for the control of leishmaniasis. Most epidemiological and control studies of canine leishmaniasis are performed by serological methods. Although such methods are traditionally considered to be more sensitive than parasitological techniques for the diagnosis of the disease, they underestimate the prevalence and incidence of the infection relative to those estimated by culture and PCR . Indeed, experimentally infected dogs that develop the disease have an anti-Leishmania humoral immunoresponse while those that remain asymptomatic present a cellular response . The application of highly sensitive techniques, such as PCR and Western blotting (WB) , as well as the optimization of culture , have improved the rate of detection of leishmaniasis. Here, we compare the results obtained by various diagnostic methods (direct examination, culture, PCR, enzyme-linked immunosorbent assay [ELISA], WB, delayed-type hypersensitivity reaction [DTH], and a lymphocyte proliferation assay [LPA]) for dogs from two areas in Spain where canine leishmaniasis is endemic, and we consider the results in the context of the clinical status of the animals. The study was carried out with 72 animals, 38 of which were from the Priorat, an area in northeast Spain where leishmaniasis is highly endemic. The remaining 34 dogs were from an animal protection society in Madrid, Spain. Animals were selected for the study after extensive serological screening for anti-Leishmania-specific antibodies, and subjects were sometimes chosen for screening because of the presence of sick animals in the kennel. In other cases, we picked asymptomatic animals with specific antibodies or under other irregular circumstances. Thus, the dog samples were biased and cannot be considered representative of the whole dog population in the areas studied. Clinical symptoms and lesions consistent with canine leishmaniasis (skin abnormalities, onychogryposis, weight loss, epistaxis, apathy, ocular and other lesions, lymph node and spleen enlargement, etc.) were recorded for all dogs. Blood was collected by cephalic or jugular venipuncture for complete blood count and biochemical analysis (serum proteins, the renal markers urea and creatinine, and the hepatic markers aspartate aminotransferase and alanine aminotransferase), for detection of anti-Leishmania antibodies, and for PCR and LPA. Popliteal lymph node aspirates were obtained for direct examination, culture, and PCR. Needle aspiration skin microbiopsy was also performed to obtain samples for PCR. Anti-Leishmania antibodies were quantitatively detected by an immunofluorescence antibody test (IFAT) and an ELISA for immunoglobulin G2 (ELISA-IgG2) for dogs from Madrid and by ELISAs for protein A (ELISA-protein A) and for IgG2 for dogs from the Priorat. The IFAT was performed by standard methods using a fluorescent conjugated anti-dog IgG (heavy plus light chains; Jackson ImmunoResearch, West Grove, Pa.) at a 1:50 dilution. ELISAs were performed as described elsewhere . Horseradish peroxidase-conjugated protein A (dilution, 1:30,000; Sigma, St. Louis, Mo.) and anti-dog IgG2 (dilutions, 1:2,000 to 1:5,000; Bethyl Laboratories, Montgomery, Tex.) were used as second antibodies. The reaction results were quantified in units relative to those of a positive serum sample that was used as a calibrator and arbitrarily set at 100 U. WB analyses of sera from the Priorat area were performed at the Faculty of Pharmacy, Barcelona, Spain, as described elsewhere . Those sera that developed any of the polypeptide fractions of 14 or 16 kDa, previously reported as 12 and 14 kDa , were considered positive. WB analyses of sera from the Madrid area were performed at the Faculty of Veterinary Science, Madrid, as described previously . The sera that revealed the polypeptide fractions of 30, 42, 50, and 57 kDa were considered positive, as reported previously . The correspondence between positive results detected by each method has been established (S. Mendez, M. J. Aisa, F. J. Fernandez-Perez, L. Iniesta, M. Portus, J. M. Alunda, and M. T. Gomez-Munoz, submitted for publication). DNA for PCR analysis was extracted with a PCR template preparation kit (Boehringer Mannheim, Mannheim, Germany). Leishmania-specific oligonucleotide primers A2 (5'-GGGAGAAGCTCTATTGTG-3') and B1 (5'-ACACTCAGGTCTGTAAAC-3') were used to amplify a 650-bp fragment in the internal transcribed spacer region of genomic DNA . The DTH was performed by inoculation into the skin of the groin of 0.1 ml of Leishmania infantum antigen (3 x 108 promastigotes/ml), which was kindly supplied by the Instituto de Salud Carlos III, Majadahonda, Spain. Reactions were read after 48 h and measured by the ballpoint pen method. The LPA was performed with peripheral blood mononuclear cells as described elsewhere . Crude results are shown in Table . Each laboratory (Madrid and Barcelona) applied its current serological tests in addition to the ELISA-IgG2, which was performed in both laboratories with the same conjugates, the same quantification system, and the same calibrator serum. The IgG2 results correlated highly with the ELISA-protein A results (Spearman's rank correlation coefficient, 0.950; P = 0.000) and the IFAT results (Spearman's rank correlation coefficient, 0.627; P = 0.000). Therefore, to simplify this discussion, only results from the ELISA-IgG2 were included in the statistical analysis and comparison of techniques. In contrast, no correlation was observed between the techniques for detecting the cellular immunoresponse, DTH, and LPA (Spearman's rank correlation coefficient, 0.052; P = 0.75), and so the results obtained in each determination were analyzed independently. Attempts to link DTH and LPA findings for use in detecting human leishmaniasis have led to contradictory results. DTH and LPA results correlated in L. major-infected children of Jericho and Tunisia . Nevertheless, the lymphoproliferative response to the L. major antigen by peripheral blood mononuclear cells from Sudanese individuals with a positive leishmania skin test and no history of cutaneous leishmaniasis was similar to the response observed in Danish people with no reported exposure to Leishmania parasites . TABLE 1 | Clinical and laboratory findings and results of diagnostic methods for visceral leishmaniasis in dogs The dogs that were positive by direct examination, culture, PCR, or at least two immunological methods were considered probably infected. Among them, those with external signs of leishmanial disease were considered to be symptomatic and those without external signs of disease, despite the detection of biochemical and hematological (analytical) disorders, were considered to be asymptomatic or to have cryptic infections. The specific IgG2 levels detected by ELISA were directly related to the pathophysiological disorders detected during the physical and analytical examinations . All symptomatic dogs (10 of 10) had high antibody titers as determined by the ELISA-IgG2, while only 35 of 51 animals with cryptic infections were positive by this technique. FIG. 1. | IgG2 expression in L. infantum IgG2 expression in L. infantum-infected dogs is related to the pathophysiology. Columns indicate mean values, and error bars indicate standard deviations. Dogs are divided into cohorts as follows: (column 1) symptomatic dogs, i.e., infected animals with external signs of leishmanial disease; (column 2) infected dogs, i.e., animals found positive by direct examination, culture, PCR, or at least two immunological techniques; (column 3) infected asymptomatic dogs with biochemical alterations; (column 4) infected asymptomatic dogs without biochemical alterations; (column 5) noninfected dogs. n, number of dogs in each cohort. The dotted line indicates the cutoff separating positive and negative results. The positive and negative results obtained with the techniques were compared two by two by the McNemar and Pearson tests in the whole dog sample studied, in animals considered to be infected (symptomatic and asymptomatic animals) and in those found to have cryptic infections . Culture and PCR provided very similar results; they gave the same proportion of positives (McNemar test, P = 0.629) and a high degree of association (Pearson test, P < 0.001). The culture and PCR results differed for 17 of the 72 dogs studied (10 animals positive by PCR and negative by culture and 7 animals positive by culture and negative by PCR). In 11 of these animals, immunological techniques confirmed the presence of parasites. In three cases PCR was the only positive test, and in two cases only the culture was positive, as confirmed by new samples and repeated analyses. TABLE 2 | Statistical analysis of results (positive/negative) obtained by various diagnostic techniques in three dog cohorts: (i) the whole dog population studied, (ii) dogs considered to be infected with L. infantum, and (iii) asymptomatic dogs with a cryptic Leishmania infection The humoral immunoresponse, detected as L. infantum-specific IgG2 expression, was also associated with PCR and culture results when the whole dog population was considered (Pearson test, P <= 0.001) but was independent of those results when only the animals with cryptic infections were considered. The results obtained by immunological techniques were independent when compared two by two, even when these techniques were considered to have the same target as those of antibody detection (ELISA-IgG2 and WB) or cellular immunoresponse (DTH and LPA) techniques. The lack of a "gold standard" for diagnosis of asymptomatic infections caused by L. infantum is a drawback to epidemiological studies of the disease. Parasitological techniques like direct microscopic examination and culture offer the only reliable evidence of the presence of parasites in a sample. However, direct examination lacks sensitivity when the parasite number is small, and the growth capacity in vitro varies from one Leishmania strain to another. Other analytical techniques, such as PCR and immunological methods, lack some specificity and sensitivity. There is evidence of Leishmania persistence inside the host after recovery from the disease , and a positive response to Leishmania antigen detected by DTH and LPA has been associated with asymptomatic infection . However, cellular immunoresponse or low antibody levels may be due to immunological memory rather than to the presence of the parasite in the host, especially in areas of endemicity. Moreover, positive lymphoproliferative responses to Leishmania antigen have been observed in nonexposed humans . Whether a dog can be considered infected or uninfected is the key determinant of parameters such as sensitivity, specificity, and predictive values for diagnostic techniques. If we consider that only those animals that are positive by direct examination or culture are parasitized, the sensitivities and specificities of other methods such as PCR and immunological techniques are affected. Therefore, these parameters were not determined in the present study. The arbitrary definition of a cutoff in the ELISA and the lack of a clear separation between positive and negative results entail a lack of specificity when this cutoff is established to improve the sensitivity of the technique. The low humoral immunoresponse characteristic of asymptomatic animals frequently places their antibody levels at the borderline of detectability, and positive results may be caused by artifacts and cross-reactions. This may account for the association between the PCR and ELISA-IgG2 results for the total dog sample (Pearson test, P = 0.001) and for their independence for the sample comprising only animals with cryptic infections. WB analysis is regarded as a highly specific technique, when the specific bands are considered, and is very useful with samples with low antibody concentrations . However, bands are difficult to identify, since several antigen fractions have similar molecular weights and subjectivity in reading the results cannot be avoided when very weak bands are detected. This may explain the independent results (Pearson test, P > 0.05) provided by the ELISA-IgG2 and WB. Moreover, since the antigen treatments in the ELISA and WB differed strongly, each technique identified a different set of epitopes, and this explains the lack of association in animal cohorts with a predominance of individuals with low humoral immunoresponses. In conclusion, visceral leishmaniasis in sick dogs from areas of endemicity can be easily diagnosed. It is characterized by a high humoral immunoresponse that is measurable by conventional serological techniques (e.g., IFAT and ELISA) and a high parasite burden that is detectable by parasitological and PCR methods. However, when we deal with asymptomatic animals, particularly in epidemiological studies, immunological techniques do not discriminate between infected and noninfected animals. Antibody detection alone and DTH or LPA, which are used to estimate the infected dog population , are not suitable tools,and parasitological methods and PCR offer more accurate results. Backmatter: PMID- 12204949 TI - Functional and Phenotypic Changes in Circulating Lymphocytes from Hospitalized Zambian Children with Measles AB - Measles is associated with immunosuppression and increased susceptibility to secondary infections and is a particular problem in developing countries. Lymphocyte changes accompanying immune activation and regulation of the immune response may contribute to immunosuppression. To evaluate lymphocyte changes during measles, children (n = 274) hospitalized with measles in Lusaka, Zambia, were evaluated at entry, discharge, and 1-month follow-up and compared to healthy Zambian children (n = 98). Lymphopenia was present on hospital admission and reflected decreased CD4 and CD8 T cells but resolved quickly. Lymphopenia was most marked in girls, in those with temperatures of >38.5C, and in malnourished children. CD4/CD8 ratios were decreased at all time points and were lower in boys than in girls at discharge and follow-up. Spontaneous death occurred in cultured lymphocytes, and the proportions of freshly isolated cells undergoing apoptosis, based on annexin V and propidium iodide staining, were increased. Surface Fas was increased on both CD4 and CD8 T cells compared to controls, and expression was greater on CD4 T cells and was inversely correlated with lymphocyte viability in culture at study entry. Mitogen stimulation of lymphocytes improved viability, but inhibitors of Fas, tumor necrosis factor (TNF)-related apoptosis-inducing ligand, and TNF did not. Plasma levels of beta2 microglobulin and soluble Fas, Fas ligand, CD8, CD4, and TNF receptor were increased, and soluble CD8 was higher in boys than in girls. The multiple effects of measles on lymphocytes from Zambian children include decreased numbers in circulation, increased activation, and increased susceptibility to cell death, with substantive differences in the magnitude of these changes between boys and girls. Keywords: Introduction : Measles causes nearly 1 million deaths per year worldwide, with the heaviest burden occurring in sub-Saharan Africa . The high morbidity and mortality associated with measles virus (MV) infection is due primarily to secondary infections, particularly of the respiratory and gastrointestinal tracts . Increased susceptibility to other infectious diseases has been linked to the immune suppression associated with this infection. Cell-mediated immunity, manifested by decreased delayed-type hypersensitivity responses to skin test antigens and decreased proliferation of lymphocytes after stimulation in vitro, is suppressed, and this suppression may persist for several months after recovery . The mechanism of MV-induced immune suppression is incompletely understood, and few studies of children in developing countries, where mortality rates are highest, have been done. Lymphocyte activation is a consistent feature of acute measles in North and South American children, and we have proposed that the observed cytokine "shift" from type 1 cytokines early during infection to type 2 cytokines during and after recovery inhibits development of effective type 1 cytokine-mediated cell-mediated immunity responses upon subsequent exposure to new pathogens . In addition, lymphocyte apoptosis, often associated with immune activation, is a feature of acute infections with other immunosuppressive murine and human viruses , and it has been hypothesized that increased apoptosis of lymphocytes in individuals with acute measles may contribute to immune suppression . To better understand these processes in African children, we have evaluated the functional and phenotypic changes that occur during acute MV infection and recovery from measles in children hospitalized with measles in Lusaka, Zambia. MATERIALS AND METHODS : Study population. | The children studied (n = 274; mean age, 3.05 +- 0.21 years; median age, 1.4 years; range, 2 months to 14.3 years) were a subgroup of children enrolled in a study of the clinical manifestations and immune responses of human immunodeficiency virus (HIV)-infected and non-HIV-infected Zambian children hospitalized with measles between January 1998 and August 2000 . Children admitted to the infectious-disease isolation ward at the University Teaching Hospital (UTH) in Lusaka, Zambia, with the clinical diagnosis of measles were prospectively enrolled. Measles was confirmed by the presence of immunoglobulin M (IgM) antibody against MV (Wampole, Cranbury, N.J.), and children who were HIV infected, as determined by the presence of antibody to HIV (Organon Technika, Boxtel, The Netherlands) and reverse transcription-PCR detection of HIV type 1 RNA (Amplicor version 1.5; Roche Pharmaceuticals Inc., Branchburg, N.J.) were excluded. Children who were moribund at the time of hospitalization were not enrolled in the study, although many of those enrolled had severe or complicated infections and there were four deaths among the study children. Samples from all children were not included in all assays or at all time points due to the limited amounts of blood available, difficulty with blood drawing, or parental withdrawal from the study. Not all children returned for the 1-month follow-up visit after discharge, and active tracing was not performed. Healthy control children (n = 98; mean age, 3.47 +- 0.31 years; median age, 1.75 years; range, 6 months to 9.4 years; P = 0.03) were enrolled from two sources located within 2.5 miles of the UTH: the clinic for well-child care at Chilenge Health Center (<5 year olds) and the Regiment Primary School (6 to 7 year olds). Control children were documented to be negative for MV IgM and for HIV antibody and RNA. The study populations were equally matched for sex, with 49% boys in the measles group and 50% boys in the control group. Written informed consent was obtained from the parents or guardians of the children studied. The study was reviewed and approved by the Committee on Human Research of the Johns Hopkins Bloomberg School of Public Health and the Ethics Committee of the UTH. Nutritional status. | Weight-for-height and height-for-age z scores were determined with Epi-Info software (Centers for Disease Control and Prevention). Wasting was defined as a weight-for-height >2 standard deviations below the norm, and stunting was defined as a height-for-age >2 standard deviations below the norm (defined by National Center for Health Statistics reference values). Nine percent of the children with measles and 2.2% of the control children were wasted (P < 0.05), and 48.6% of the children with measles and 33.7% of the control children were stunted (P < 0.02; chi2 test) Blood collection and processing. | Peripheral blood, collected in a sterile tube containing EDTA, was obtained at study entry (mean, 3.6 +- 0.1 days after onset of rash; range, 1 to 10 days after onset of rash; n = 252), at hospital discharge (mean, 6.5 +- 0.18 days after onset of rash; range, 2 to 17 days after onset of rash; n = 185), and at a 1-month follow-up visit (mean, 37.6 +- 0.5 days after onset of rash; range, 25 to 61 days after onset of rash; n = 78). White blood cell counts were determined manually. Differential counts were performed on Wright-Giemsa-stained smears by the hematology laboratory at the UTH. Mononuclear cells were separated by Ficoll-Hypaque (Pharmacia, Piscataway, N.J.) density gradient centrifugation. Plasma was frozen in aliquots at -80C. Lymphocyte viability studies. | Washed fresh peripheral blood mononuclear cells (PBMCs) were suspended at a concentration of 106/ml in RPMI 1640 medium supplemented with 10% fetal bovine serum, 10 mM HEPES, and 10 mug of gentamicin/ml. The cells were cultured, with or without stimulation with phytohemagglutinin (PHA; 5 mug/ml; Sigma Chemical Co., St. Louis, Mo.), in 96-well round-bottom microtiter plates at a volume of 200 mul/well. For some studies, cells were cultured with the following inhibitors of apoptosis: tumor necrosis factor (TNF)-related apoptosis-inducing ligand-Fc IgG chimera (20 ng/ml), Fas-Fc IgG chimera (10 ng/ml), TNF receptor II (TNFRII)-Fc IgG chimera (20 ng/ml) (R&D Systems, Minneapolis, Minn.), and mouse anti-human Fas ligand (FasL) monoclonal antibody (1 mug/ml; PharMingen, San Diego, Calif.). Cell viability was determined by trypan blue exclusion. All studies were performed on fresh samples at the Virology Laboratory in Lusaka. Flow cytometry. | All analyses were performed with fresh samples at the Virology Laboratory in Lusaka on a FACScan flow cytometer using Cell Quest software (Becton Dickinson, San Jose, Calif.). Directly conjugated mouse monoclonal antibodies to the following human antigens were used: CD3-fluorescein isothiocyanate (FITC), CD4-phycoerythrin (PE), CD8-PE (Becton Dickinson), and Fas-FITC (Antigenix America, Huntington Station, N.Y.). In each case, staining was compared with that of the appropriately labeled isotype control antibody. The percentages of CD4 and CD8 T cells were determined after staining whole blood. CD3+/CD4+ and CD3+/CD8+ cells were enumerated with lymphocyte gates set on the basis of forward and side scatter information adjusted to exclude monocytes based on percent CD3-/CD4+ cells. Fas expression was measured on CD4+ and CD8+ subsets using the same gates. The percentage of freshly isolated mononuclear cells staining with annexin V only was determined after simultaneous staining with annexin V-FITC and propidium iodide (PI) (R&D Systems) using a wide gate based on forward and side scatter information which included dead and dying cells. The percentage of annexin V-positive cells at entry among subsets was assessed after 18 h of culture by dual staining for annexin V and CD4 or CD8. Soluble factors in plasma. | The levels of gamma interferon (IFN-gamma; BD PharMingen), soluble CD8 (sCD8), sCD4 (Endogen, Woburn, Mass.), sTNFRII (R&D Systems), sFas, and sFasL (Medical Biological Laboratories, Nagoya, Japan) in plasma were measured by enzyme immunoassay according to the manufacturer's instructions. beta2 microglobulin was measured by radioimmunoassay (Pharmacia & Upjohn, Uppsala, Sweden). Values below the limit of detection were assigned values midway between zero and the lower limit of detection for the purposes of analysis. Statistical analysis. | Data are presented as means +- standard errors of the mean unless otherwise indicated. All analyses were performed using STATVIEW or STATA software (SAS, Cary, N.C.). Mann-Whitney U, Wilcoxon signed-rank, and Spearman rank correlation coefficient tests were used as appropriate, with the level of significance set at a P value of 0.05. No corrections were made for multiple comparisons. RESULTS : Lymphopenia. | The numbers and types of leukocytes in the peripheral blood were measured at study entry (usually the day after admission to the hospital), at discharge from the hospital, and approximately 1 month after discharge, and the values were compared to those of healthy children . On entry, total leukocyte counts were not different between measles patients (8,400 +- 287 cells/mul) and controls (8,300 +- 345 cells/mul), but they were increased in measles patients by the time of discharge (10,579 +- 425 cells/mul; P < 0.0001) and at follow-up (10,610 +- 498 cells/mul; P = 0.0005) . Monocyte counts were increased at all times in measles patients compared to controls . Lymphocyte counts were decreased for measles patients at study entry (3,681 +- 164 cells/mul) compared to controls (5,650 +- 240 cells/mul; P < 0.0001) and were lower in girls than in boys . The lymphocyte counts rose to 5,835 +- 298 cells/mul at the time of discharge and to 6,362 +- 359 cells/mul (P = 0.27) at follow-up . The lymphocyte counts in the four children who died (girls aged 6, 8, and 34 months; boy aged 8 months) were 1,200, 2,196, 2,280, and 2,500 cells/mul, all in the lowest quartile at study entry (306 to 2,873 cells/mul) for children with measles. However, the lymphocyte count on entry did not predict the disease severity as measured by days of hospitalization (rs = 0.013; P = 0.85; Spearman rank correlation test). As in control children (rs = -0.549; P < 0.0001), younger children with measles tended to have higher lymphocyte counts than older children at study entry (rs = -0.456; P < 0.0001) and higher numbers of CD4 T cells (rs = -0.418; P < 0.0001; Spearman rank correlation test for all comparisons). Lymphopenia was not prolonged in children <1 year of age (4,652 +- 316 cells/mul on entry [n = 85; P < 0.0001], 7,118 +- 489 cells/mul on discharge [n = 50], and 7,417 +- 580 cells/mul on follow-up [n = 40]) compared to age-matched controls at 7,323 +- 530 cells/mul [n = 24]; Mann-Whitney U test). Lymphocyte counts on entry were lower in children with evidence of malnutrition (wasted, 2,821 +- 317 cells/mul [n = 20]; not wasted, 3,869 +- 191 cells/mul [n = 181; P = 0.0668]) and were higher in children with stunted growth (stunted, 3,911 +- 233 cells/mul [n = 124]; not stunted, 3,184 +- 225 cells/mul [n = 99; P = 0.0209]; Mann-Whitney U test). Fever and the release of proinflammatory cytokines induced by infection can lead to rapid depletion of lymphocytes from the blood by increasing the rate of lymphocyte entry into secondary lymphoid tissue and reducing the rate of return to the circulation . To determine if fever played a role in lymphopenia during measles, patients from this and the larger cohort were stratified according to temperature at the time of study entry . The lymphocyte count was significantly lower in children with fever of >38.5C (2,782 +- 261 cells/mul) than in children with temperatures of <38.5C (3,766 +- 131 cells/mul; P = 0.007; Student's t test). The lymphocyte count did not correlate with plasma IFN-gamma levels on entry (rs = -0.16; P = 0.549; n = 15; Spearman rank correlation test). The absolute numbers of CD4 and CD8 T lymphocytes were determined by flow cytometry . The numbers of both CD4 T cells (1,013 +- 67 cells/mul) and CD8 T cells (782 +- 56 cells/mul) were depressed at study entry compared to controls (CD4, 2,007 +- 93 cells/mul [P < 0.0001]; CD8, 1,347 +- 73 cells/mul [P < 0.0001]). CD4 T-cell counts remained depressed at discharge (1,813 +- 155 cells/mul [P = 0.022]) but had returned to normal (2,117 +- 157 cells/mul) at the 1-month follow-up visit. These changes in CD4 and CD8 T-cell numbers led to lower CD4/CD8 ratios at study entry (1.492 +- 0.1; P = 0.0024), hospital discharge (1.243 +- 0.07; P < 0.0001), and follow-up (1.527 +- 0.11; P = 0.041) compared to controls (1.679 +- 0.072). Because CD8 T cells tended to increase more rapidly in boys than in girls, the CD4/CD8 ratios were significantly different by sex at discharge and at follow-up . Lymphocyte viability. | To determine whether lymphocytes from measles patients were more likely to undergo spontaneous cell death in culture, the viability of fresh PBMCs cultured without stimulation for 3 days was assessed . Fewer cells from measles patients at study entry than from control children were viable when assessed 1, 2, and 3 days after being placed in culture (P < 0.0001 for all comparisons). On day 3, viability was 74.7% +- 1.3% in cultures of cells from measles patients compared to 85.2% +- 0.9% in cultures of cells from control children. The tendency to undergo spontaneous cell death improved with time after acute measles, but differences from controls persisted at hospital discharge (day 3 viability, 78.7% +- 1.5%; P = 0.0007) and at the follow-up evaluation 1 month later (day 3 viability, 81.5% +- 1%; P = 0.0122). To determine whether stimulation of lymphocytes in vitro would alter this pattern of cell death, the viabilities of cells cultured in the presence of PHA were assessed . Cells from children with measles had lower day 3 viability (79.1% +- 1.2%) at study entry than cells obtained from control children (85% +- 0.7%; P = 0.0001). However, at later times during recovery there were no differences in the viabilities of PHA-stimulated cells obtained from children with measles and those obtained from control children. Surface binding of annexin V. | To determine whether lymphocyte death was apoptotic and whether cells were primed for apoptosis in vivo, PBMCs from patients with measles were stained simultaneously with annexin V and PI . Annexin V binds phosphatidyl serine residues exposed on the surfaces of apoptotic and necrotic cells, while PI enters only necrotic cells. The percentage of annexin V+/PI- cells was increased in cells isolated from measles patients at the time of study entry (28.7% +- 2.2%) and at hospital discharge (30.8% +- 3.1%) compared to cells obtained from control children (22.2% +- 2.2%; P = 0.0289 and P = 0.0134, respectively). Both CD4 and CD8 T cells from measles patients at the time of study entry bound annexin V , with a larger proportion of annexin V-positive CD8 T cells (17.7% +- 1.6%) than CD4 T cells (8.7% +- 1.4%; P = 0.0007) after overnight culture. Surface expression of Fas. | Surface expression of Fas (CD95), a member of the TNFR family, is often upregulated during T-cell activation and can render lymphocytes susceptible to apoptosis upon interaction with cells expressing FasL . There was a marked increase in the percentage of CD4 T cells expressing Fas at the time of study entry (25.9% +- 3%) in measles patients compared to control children (1.1% +- 0.16%; P < 0.0001), and this remained elevated at discharge (22.5% +- 3.4%; P < 0.0001) and 1 month later (8.2% +- 2.3%; P = 0.0012) . The percentage of CD4 T cells expressing Fas at study entry was inversely correlated with the viabilities of cultured PBMCs on days 1 (rs = -0.752; P = 0.0067) and (rs = -0.598; P = 0.0384; Spearman rank correlation test). There was no significant correlation of viability with Fas expression on CD4 T cells at discharge or follow-up (data not shown). A smaller percentage of CD8 T cells expressed Fas . This percentage was elevated in measles patients at the time of study entry (11.8% +- 1.9%) compared to control children (0.8% +- 0.3%; P < 0.0001) and at the time of discharge (11.5% +- 4.4%; P = 0.0004), but it had returned nearly to normal (1.95% +- 0.6%; P = 0.0358) by the 1-month follow-up visit. There was no correlation of lymphocyte viability in culture with Fas expression on CD8 T cells (data not shown). Levels of soluble cell surface proteins in plasma. | Soluble forms of Fas are produced by activated PBMCs and inhibit Fas-induced cell death . Plasma sFas levels were elevated (1,910 +- 132 ng/ml) in measles patients at the time of hospital discharge compared to control levels (1,342 +- 481; P = 0.0006) but did not differ significantly from that of controls at entry (1,463 +- 124) or at 1-month follow-up (1,491 +- 115) . sFasL is produced by proteolytic cleavage of FasL from the surfaces of activated FasL-expressing T cells . sFasL is postulated to suppress Fas-mediated death signals by competitive binding to Fas without inducing apoptosis . Children with measles had normal levels of sFasL in their plasma at study entry (379 +- 36 ng/ml). These levels rose by hospital discharge (432 +- 42; P = 0.0776) and were significantly increased at the 1-month follow-up visit (518 +- 0.047 ng/ml; P = 0.0048) compared to the levels in the plasma of control children (345 +- 36) . Soluble forms of CD4, CD8, and TNFRII are shed from the surfaces of activated lymphocytes, and levels of beta2 microglobulin are increased as a consequence of immune activation . Plasma levels of sCD8, sTNFRII, and beta2 microglobulin were elevated at entry and discharge compared to those in controls , and levels of sTNFRII and beta2 microglobulin remained elevated at follow-up. At entry, sCD8 levels were higher in boys (1,056 +- 599 U/ml; n = 45) than in girls (761 +- 468; n = 31; P = 0.0169; Mann-Whitney U test), while none of the other molecules measured had levels that were significantly different in boys and girls (data not shown). Effects of inhibitors of receptor-induced apoptosis on viability of cultured cells. | Several cell surface receptors in addition to Fas can induce cell death after ligand binding. To determine if blocking any of these receptor-ligand interactions would protect cultured lymphocytes from measles patients from spontaneous cell death, Fc IgG chimeras of TNF-related apoptosis-inducing ligand, Fas, and TNFRII and a FasL-blocking antibody were added to cultured PBMCs. Viability was determined after 3 days in culture. None of these inhibitors prevented the spontaneous death of PBMCs obtained from children with measles (data not shown). FIG. 1. | Effect of measles on numbers of leukocytes in the peripheral blood. Effect of measles on numbers of leukocytes in the peripheral blood. The mean absolute numbers of leukocytes (A), monocytes (B), lymphocytes (C), CD4 T cells (D), and CD8 T cells (E) and CD4/CD8 T-cell ratios (F) in children with measles at study entry, hospital discharge, and 1-month follow-up compared to controls are shown. The numbers of CD4 and CD8 T cells were calculated using the absolute lymphocyte count and percentages of CD3+ T cells coexpressing CD4 or CD8, as measured by flow cytometry. The number of children in each group is indicated at the base of each bar. The error bars indicate standard errors of the mean. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05 (Mann-Whitney U test). FIG. 2. | Association of fever with numbers of lymphocytes in circulation; mean lymphocyte counts on entry for children with measles stratified by presence or absence of fever of >38. Association of fever with numbers of lymphocytes in circulation; mean lymphocyte counts on entry for children with measles stratified by presence or absence of fever of >38.5C. The number of children in each group is indicated at the base of each bar. The error bars indicate standard errors of the mean. **, P = 0.007 (Student's t test). FIG. 3. | Viability of mononuclear cells in culture; mean percent viability of mononuclear cells isolated from patients with measles compared to controls after culture without stimulation (A) and with PHA stimulation (B). Viability of mononuclear cells in culture; mean percent viability of mononuclear cells isolated from patients with measles compared to controls after culture without stimulation (A) and with PHA stimulation (B). Mononuclear cells were cultured with or without 10 mug of PHA/ml. Viability was determined by trypan blue exclusion on days 1, 2, and 3 after initiation of cultures. The numbers of unstimulated samples assessed were 68 to 73 at entry, 35 to 47 at discharge, 42 to 46 at follow-up, and 54 to 71 for controls. The numbers of PHA-stimulated samples assessed were 53 to 57 at entry, 20 to 29 at discharge, 32 to 36 at follow-up, and 54 to 70 for controls. The error bars indicate the standard error of the mean. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05 (Mann-Whitney U test). FIG. 4. | Annexin V binding to lymphocytes. Annexin V binding to lymphocytes. Annexin V binding to the cell surfaces of PI-negative PBMCs (A) and of CD4-PE- or CD8-PE-stained T lymphocytes obtained at entry and cultured overnight (B) was measured by flow cytometry using a wide forward and side scatter gate. The number of samples in each group is indicated at the base of each bar. The error bars indicate the standard error of the mean. *, P < 0.05 (Mann-Whitney U test); ***, P < 0.001 (Wilcoxon rank test). FIG. 5. | Fas (CD95) expression on the surfaces of lymphocytes during measles. Fas (CD95) expression on the surfaces of lymphocytes during measles. (A and B) Mean percentages of CD4 (A) and CD8 (B) T cells from measles patients and control children with surface expression of Fas. Fas expression was determined by flow cytometry of freshly isolated CD4+ and CD8+ lymphocytes gated by forward and side scatter. The number of samples in each group is indicated at the base of each bar. The error bars indicate the standard error of the mean. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01; *, P < 0.05 (Mann-Whitney U test). (C) Correlation of percent Fas-positive CD4 T cells at entry with lymphocyte death in culture. P = 0.0067 (Spearman rank correlation). FIG. 6. | Levels of soluble forms of cell surface proteins, sFas(A), sFasL (B), sTNFRII (C), sCD4 (D), and sCD8 (E), and beta2 microglobulin (F) in plasma of measles patients and controls as determined by enzyme or radio immunoassay. Levels of soluble forms of cell surface proteins, sFas(A), sFasL (B), sTNFRII (C), sCD4 (D), and sCD8 (E), and beta2 microglobulin (F) in plasma of measles patients and controls as determined by enzyme or radio immunoassay. The number of samples in each group is indicated at the base of each bar. The error bars indicate the standard error of the mean. ****, P < 0.0001; ***, P < 0.001; **, P < 0.01 (Mann-Whitney U test). TABLE 1 | Changes in numbers of lymphocytes, CD4 T cells, and CD8 T cells and in CD4/CD8 ratios in boys and girls with measles and controls DISCUSSION : During measles, the rash marks the onset of the virus-specific immune response, the initiation of virus clearance, and the appearance of a variety of immunologic abnormalities that often persist for weeks to months after recovery . In this study, we have evaluated the functional and phenotypic changes in lymphocytes from Zambian children with measles during acute infection and after recovery from infection. Lymphopenia occurred early and affected both CD4 and CD8 T cells. Cells in circulation were activated and susceptible to apoptosis as indicated by increased expression of Fas, binding of annexin-V, and spontaneous death in culture. During the recovery phase, lymphocyte counts normalized rapidly, expression of Fas decreased, sFas and sFasL levels in plasma became elevated, and mitogen stimulation protected cells from death in culture. Lymphocyte responses differed between boys and girls. In boys, lymphopenia was less marked, CD8 T cells in circulation increased more rapidly, and the sCD8 level was higher than in girls. Therefore, in children with measles, T lymphocytes undergo profound changes associated with altered peripheral circulation, activation, apoptosis, and susceptibility to death in culture. The differences in responses between boys and girls may be linked to sex-based differences in measles mortality and complications associated with measles immunization. Lymphopenia is a characteristic of many acute viral infections, including measles . Profound lymphopenia has previously been associated with severe measles and high mortality in African children , and the four children in our study who died during hospitalization had lymphocyte counts in the lowest quartile at study entry. More profound decreases in T-cell counts during acute measles have also been associated with malnutrition , and the acutely malnourished (wasted) children enrolled in our study had lymphocyte counts lower than those of children who were not wasted. Interestingly, stunted children had higher lymphocyte counts, but the reason for and importance of this observation are unclear. In our study, profound lymphopenia was also associated with female sex. Lymphocyte counts did not predict the severity of disease as measured by days of hospitalization; however, this parameter does not accurately measure the disease severity of those who died during hospitalization. Furthermore, some of the most severely ill children were not enrolled in the study, limiting our ability to detect these relationships. Lymphopenia was transient in this study, and counts normalized quickly as the fever and rash resolved during hospitalization. Although persistent lymphopenia has been associated with ages of <1 year in Japanese children , this was not seen in Zambian infants. Genetic factors may contribute to these differences. In addition, environmental factors, such as exposure to other infections, nutritional status, and prior immunization, may also play as yet undefined roles. The mechanism by which the numbers of lymphocytes in circulation are decreased during the acute phases of viral diseases is not clear. Although lymphopenia has been postulated to be due to MV-induced death of T cells , MV replicates preferentially in macrophages , and circulating lymphocytes represent only a small proportion of the total lymphocyte population. Therefore, alterations in lymphocyte trafficking are more likely to be responsible for profound, transient changes in the numbers of circulating lymphocytes. Increased exit of naive T cells from the circulation into secondary lymphoid tissue will improve the chances that lymphocytes capable of recognizing viral antigens will contact appropriate antigen-presenting cells . In macaques with measles, lymphocyte counts decrease during viremia and increase after the onset of the rash . One potential common mechanism for induction of altered trafficking is the effect of increased temperature. Fever can be associated with induction of inflammatory cytokines, increased adhesion of lymphocytes to lymph node vessels, and depletion of lymphocytes from the circulation . Fever is a universal feature of measles, typically beginning 2 to 3 days before the onset of the rash as a part of the prodrome , a time when lymphocyte counts are decreased . In our study, fever at the time of study entry was specifically associated with low lymphocyte counts. Lymphopenia during acute simian immunodeficiency virus infection of macaques has been attributed to alterations in lymphocyte trafficking associated with production of IFN-gamma . Although we did not find a correlation between plasma IFN-gamma and lymphocyte counts in measles, IFN-gamma levels are highest early during the rash and could contribute to altered trafficking and a reduction in the numbers of circulating lymphocytes. Thus, lymphopenia during measles is likely to involve increased trafficking of peripheral blood lymphocytes into secondary lymphoid tissue during the acute phase of the immune response to infection. The increasing numbers of lymphocytes in circulation during recovery probably correspond with the release of activated T cells from secondary lymphoid tissue for dissemination to sites of virus infection Our previous studies have shown that activated, proliferating CD4 and CD8 T cells are found in the circulation of measles patients after the onset of the rash , and the present studies show that Fas, another marker of lymphocyte activation , is also highest during the rash. Decreases in the numbers of both CD4 and CD8 T cells contributed to the lymphopenia, but CD4 T cells were more depressed for a longer time than CD8 T cells, resulting in decreased CD4/CD8 T-cell ratios at all time points. Late in infection, boys were the major contributors to the overall CD4/CD8 T-cell ratio suppression. A previous study of measles patients in Senegal also noted the tendency for girls to have higher percentages of CD4 T cells and higher CD4/CD8 T-cell ratios during convalescence . In our study, boys also had higher levels of sCD8 in plasma than girls, further suggesting greater stimulation and proliferation of CD8 T cells in response to infection. There is intense interest in understanding how girls and boys differ in their responses to measles and measles immunization. Studies from Guinea-Bissau and South Africa have observed increased mortality due to measles in girls , and three of the four deaths in our study were those of girls, reflecting the higher mortality in girls in the larger study . Studies of a high-titer measles vaccine showed long-term increased mortality in girls , and girls are more likely to develop fever and rash after routine measles-mumps-rubella immunization . Our data showing more profound lymphopenia and less evidence of CD8 T-cell activation in girls add to the accumulating evidence for an important effect of sex on responses to MV and suggest that the CD8 T-cell response may be an important aspect of the differences. The vigor and timeliness of the CD8 T-cell response is likely to be important for recovery from measles. Other studies have shown that circulating CD8 T cells have cytotoxic activity against MV-infected cells , and CD8 T cells are present at sites of MV replication and are presumed to play an important role in the clearance of MV from infected tissues . Lymphocytes were susceptible to cell death when placed in culture long after lymphocyte numbers in the blood had returned to normal, and in other studies this abnormality was present for up to 6 months after recovery . The propensity for lymphocytes to die in culture is assumed to reflect the in vivo process whereby virus-specific cells are eliminated after virus clearance . The death of activated T cells can be induced by at least two general pathways: an active antigen-dependent pathway that is mediated through death receptors, such as Fas and TNFR, and a passive, antigen-independent lymphokine withdrawal pathway regulated by members of the Bcl-2 family . During acute measles in our study, large percentages of lymphocytes were annexin V positive and PI negative, indicating apoptotic cell death. The process of deletion of virus-specific CD8 T cells is the best studied, involves multiple cell death-inducing mechanisms, and may be preceded by functional inactivation . Over 17% of CD8 T cells bound annexin, suggesting that these cells were apoptotic. Expansion of CD4 T cells is generally slower and downregulation is a much more prolonged process than for CD8 T cells. Therefore, during convalescence, virus-specific CD4 T cells may be more numerous than CD8 T cells. This pattern is consistent with our observations that indicators of CD8 T-cell activation, such as levels of sCD8 and IFN-gamma in plasma, are increased during the first 7 to 10 days after the onset of rash and then decrease, while evidence of CD4 T-cell activation persists (, , -, ). During measles, CD4 T cells were more likely to be Fas-positive at all time points than CD8 T cells, and 8% of circulating CD4 T cells were still expressing Fas a month after discharge. The activation-induced death of CD4 T cells is often mediated by FasL , but expression of Fas does not necessarily correlate with susceptibility to FasL-induced death. Lymphocyte death in culture was associated with Fas expression on CD4 T cells at study entry, suggesting that Fas may be involved in CD4 T-cell death early in infection. Although FasL antagonists did not block death, it is possible that apoptosis initiated in vivo is not influenced by attempts at blocking death stimuli in vitro. Increased circulating levels of sFas during hospitalization and of sFasL at follow-up suggest that this pathway is active in vivo. However, the mechanism for death in culture during recovery from measles may be different, since there was no association with Fas expression at later times. In previous studies of PBMCs from measles patients, stimulation with anti-CD3 enhanced or had no effect on death in culture, while stimulation with phorbol ester and ionomycin reduced lymphocyte death . In the present study, PHA stimulation partially protected cells from measles patients against death in culture. Lymphocyte viability was improved at the earliest time points and was completely restored at later times, again suggesting that more than one mechanism is involved in spontaneous death and that the mechanisms may change during the acute and recovery phases of measles. As in other viral infections (, , -, , , , , ), lymphokine deprivation may play a role in spontaneous death, since we have shown that lymphocytes from measles patients have increased expression of CD25, the alpha chain of the interleukin 2 (IL-2) receptor , and levels of IL-2 are increased in plasma for the first 2 weeks after the onset of the rash and then rapidly decline . IL-2 is particularly important for maintaining CD4 T cells and partially reverses the in vitro defect in mitogen-induced proliferation seen in measles . PHA-induced protection may occur by inducing production of growth factors or transcription factors that promote viability without inducing proliferation . In summary, we have shown that MV infection of hospitalized African children causes transient lymphopenia that involves CD4 and CD8 T cells and is associated with fever, malnutrition, and the female sex. During hospitalization and at the time of discharge, subpopulations of both CD4 and CD8 T cells are apoptotic, but other types of cell death, as well as alterations in lymphocyte trafficking, may contribute to lymphopenia. The magnitude of these changes differs between boys and girls, providing new insights into potential mechanisms of sex-based differences in the outcome of MV infection and immunization. Backmatter: PMID- 12204958 TI - Spontaneous Cytokine Production and Its Effect on Induced Production AB - Cytokines regulate cellular immune activity and are produced by a variety of cells, especially lymphocytes, monocytes, and macrophages. Multiparameter flow cytometry is often used to examine cell-specific cytokine production after in vitro phorbol 12-myristate 13-acetate and ionomycin induction, with brefeldin A or other agents added to inhibit protein secretion. Spontaneous ex vivo production reportedly rarely occurs. We examined the spontaneous production of interleukin 2 (IL-2), IL-4, IL-6, IL-8, IL-10, tumor necrosis factor alpha (TNF-alpha), and gamma interferon (IFN-gamma) by peripheral-blood B lymphocytes, T cells, CD8- T cells, CD8+ T cells, CD3- CD16/56+ lymphocytes (natural killer [NK] cells), CD3+ CD16/56+ lymphocytes (natural T [NT] cells), and/or monocytes of 316 acutely ill hospitalized persons and 62 healthy adults in Malawi, Africa. We also evaluated the relationship between spontaneous and induced cytokine production. In patients, spontaneous TNF-alpha production occurred most frequently, followed in descending order by IFN-gamma, IL-8, IL-4, IL-10, IL-6, and IL-2. Various cells of 60 patients spontaneously produced TNF-alpha; for 12 of these patients, TNF-alpha was the only cytokine produced spontaneously. Spontaneous cytokine production was most frequent in the immunoregulatory cells, NK and NT. For IL-2, IL-4, IL-6, IL-8, and IL-10, spontaneous cytokine production was associated with greater induced production. For TNF-alpha and IFN-gamma, the relationships varied by cell type. For healthy adults, IL-6 was the cytokine most often produced spontaneously. Spontaneous cytokine production was not unusual in these acutely ill and healthy persons living in an area where human immunodeficiency virus, mycobacterial, malaria, and assorted parasitic infections are endemic. In such populations, spontaneous, as well as induced, cell-specific cytokine production should be measured and evaluated in relation to various disease states. Keywords: Introduction : Cytokines regulate cellular immune interactions and are produced by lymphocytes, monocytes, macrophages, and, for some cytokines, also fibroblasts, neutrophils, endothelial cells, or mast cells (for a review, see reference ). Cytokines function on a microenvironmental level, but human cytokines are most commonly assessed at the macro level by measuring their levels in serum or plasma or in the supernatant of in vitro-stimulated blood cells. In recent years, improved reagents have permitted flow cytometric cell-specific cytokine assessments in which peripheral blood mononuclear cells (PBMCs) are stimulated ex vivo in the presence of an agent to inhibit protein secretion. This stimulation is followed by cell permeabilization, fixation, fluorescent staining, and cytokine detection . With multiparameter flow cytometry, specific cell populations can be identified by surface antigen staining and examined for cytokine production without cell separation or cloning. Unstimulated cells are usually run in parallel, as a negative control and to assist in setting quadrants for assessing positivity, much as is done with an isotype control. Spontaneous ex vivo cytokine production by PBMCs has been reported not to occur in healthy individuals or in persons with various disorders, including allergic asthma , active and inactive systemic lupus erythematosus , and sarcoidosis , and those recovering from vascular surgery . We ourselves have rarely, if ever, noted spontaneous cytokine production in PBMCs of healthy U.S. blood donors (reference and unpublished data). However, in a multicenter, multiphase study of the immune correlates of bloodstream infections in developing countries , we noticed individuals whose PBMCs produced cytokines without stimulation. Other researchers, measuring either protein or mRNA, have detected spontaneous cytokine production by bronchoalveolar lavage fluid cells of patients with asthma , interacting umbilical vein endothelial cells-monocyte cells , human milk mononuclear cells , peripheral blood lymphocytes of persons with atopic dermatitis , peripheral blood monocytes of patients with multiple sclerosis , or CD4+ T cells of persons with active Mycobacterium tuberculosis pulmonary infection . The presence of spontaneous cytokine production ex vivo implies prior in vivo stimulation. Since the process of cell stimulation often leads to programmed cell death, in vivo stimulation, preceding in vitro stimulation, is not only of potential physiologic importance but may also be of practical importance in terms of interpreting the results of in vitro stimulation of cytokine production. At our study site in Malawi, Africa, we acquired a sufficient number of participants to examine spontaneous cytokine production in some depth. For this study, we defined spontaneous cytokine production in a simple, quantitative manner and then assessed the cell types spontaneously producing cytokines, the cytokines and cytokine arrays being produced, the relationships between spontaneous production and subsequent induced cytokine production, and the relationship between spontaneous production and serum cytokine levels. MATERIALS AND METHODS : Participants. | During three periods in 1997 and 1998, we enrolled all 561 febrile (oral temperature, >=38C) adults (>=13 years old) and all 244 acutely ill children (<13 years old) admitted to Lilongwe Central Hospital, Malawi, Africa, into a study of bloodstream infections . All hospitalized children were included because infected children often do not present with fever. A random subset including 166 adult and 150 pediatric patients had immune studies done at admission; this subset was comparable to the general study population and was the study group analyzed here. Demographic, epidemiological, and clinical data will be presented elsewhere. We also enrolled 62 asymptomatic, healthy, employed adults; children were not included in this group for ethical reasons. Epidemiological data were not obtained for these participants. The study protocol was approved by the institutional review boards of the Centers for Disease Control and Prevention (CDC) and the Malawian Health Sciences Research Committee; informed consent was obtained from all participants and/or their guardians. As in most developing nations, human immunodeficiency virus (HIV)-infected persons in this study were neither receiving antiviral therapies nor being monitored for changes in CD4+-cell counts or HIV RNA levels in plasma. Laboratory methods. (i) Microbiological. | HIV antibody testing was done at study enrollment, using enzyme-linked immunosorbent assay kits (Murex Diagnostics Inc., Norcross, Ga.). HIV type 2 has not been reported in Malawi. Whole-blood samples were cultured as described previously . (ii) Cytokine stimulation. | Heparinized blood was either stimulated for 5 h at 37C with phorbol 12-myristate 13-acetate (200 ng/ml) (Sigma Chemical Co., St. Louis, Mo.) and ionomycin (4 mug/ml) (Sigma) in the presence of brefeldin A (40 mug/ml) (Sigma) and RPMI 1640 with l-glutamine (induced, or stimulated, cytokine expression) or retained in identical medium without phorbol 12-myristate 13-acetate and ionomycin but with brefeldin A (spontaneous, or unstimulated, cytokine expression) . No serum was added to the cultures. After being washed, the red blood cells were lysed with ammonium chloride solution, and the lymphocytes were permeabilized and fixed using Permeafix (Ortho Diagnostic Systems, Inc., Raritan, N.J.). Cell counts before and after stimulation could not be done, and markers of apoptosis were not assessed; therefore, the issue of programmed cell death could not be addressed in our analyses. After being processed, samples were shipped at 4 to 8C to CDC for further analysis. (iii) Flow cytometric reagents. | The surface antigens assessed in this study were ones previously shown in our laboratory to be stable with this permeabilization-fixation protocol, i.e., using these techniques, we had comparable results for the surface-related antigens when staining was done either pre- or postpermeabilization. Fluorescein isothiocyanate (FITC)-conjugated, phycoerythrin (PE)-conjugated, peridinin chlorophyll protein-conjugated, or allophycocyanin (APC)-conjugated murine monoclonal antibodies were obtained from the following sources: (i) Becton Dickinson Immunocytometry Systems/PharMingen (BD/PMG; San Jose, Calif.) (CD8-FITC and -PE [clone SK1], CD3-peridinin chlorophyll protein and -APC [clone SK7], CD4-APC [clone SK3], CD45-FITC [clone 2D1], CD19-APC [clone SJ25C1], CD14-PE [clone MoP9], CD16-PE [clone B73.1], CD56 [clone MY31], interleukin 4 [IL-4]-PE [clone 8D4-8], IL-8-PE [clone G265-8], and IL-10-PE [clone JES3-9D7]), (ii) Research and Diagnostics (Minneapolis, Minn.) (IL-6-PE [clone 1927.311]), and (iii) Immune Source (Reno, Nev.) (CD8-APC [clone KL.12], IL-2-APC [clone R-56.2], tumor necrosis factor alpha [TNF-alpha]-FITC [clone DTX.34], and gamma interferon [IFN-gamma]-APC [clone 13.TR]). Isotype controls were obtained from BD/PMG. Stimulation was confirmed using anti-CD69-FITC [clone L78] (BD), and permeabilization was confirmed using anti-microtubulin [clone DM1A] (Sigma) custom conjugated to FITC by CalTag, South San Francisco, Calif. (iv) Flow cytometry. | All staining was done at room temperature for 30 min in the dark after permeabilization, fixation, and shipment to CDC. Staining was followed by a buffered saline wash. Four-color cytofluorometry was done using a single FACSort (1997 and 1998) or FACSCalibur (1998) cytometer and CellQuest software (BD/PMG). Between 50,000 and 80,000 ungated events were collected from each tube in the panel. Instrument settings were standardized each day using CaliBRITE Beads (BD) and chicken red blood cells (BioSure Controls; Riese Enterprises, Inc., Grass Valley, Calif.); these settings varied little within each phase of the study. (v) Serum cytokines. | Serum samples were analyzed for IL-2, IL-4, IL-6, IL-8, and IL-10 (all provided by BD/PMG) and for IFN-gamma and TNF-alpha (both provided by Genzyme Diagnostics, Cambridge, Mass.) by enzyme-linked immunosorbent assays using pairs of cytokine-specific monoclonal antibodies according to the manufacturers' instructions. Details are provided elsewhere . No healthy participants and few children had serum cytokines assessed (see Results). Analytic techniques. (i) Definition. | For the sake of simplicity and consistency, we defined spontaneous cytokine production as occurring if the percentage of a particular cell type producing a certain cytokine was above the 95th percentile for the entire pediatric or adult patient group . For example, 7 of 158 (<5%) of the adult patients had >1.7% of unstimulated CD3+ lymphocytes read as being positive for intracellular IL-2, and 6 of the 150 (<5%) pediatric patients had >2.2% of unstimulated CD3+ lymphocytes read as being positive for intracellular IL-2. These 13 individuals were classified as having CD3+ lymphocytes spontaneously making IL-2; the remaining patients were classified as having CD3+ lymphocytes that were not spontaneously making IL-2. The number of healthy participants producing cytokines spontaneously was fairly small; the results for the healthy participants will be provided separately from those of patients. (ii) Cells assessed. | For each participant, analyses were done for all lymphocytes and various combinations of CD3+ lymphocytes (T cells), CD3+ CD8+ lymphocytes, CD3+ CD8- lymphocytes, CD3+ CD16/56+ lymphocytes (natural T [NT] cells), CD3- CD16/56+ lymphocytes (natural killer [NK] cells), CD19+ (B) lymphocytes, and monocytes, depending upon the tube configuration . The tube configurations varied slightly among the three phases of the study . (iii) Flow analyses. | For consistency, all data in this study concerning spontaneous cytokine production were based on analyses of flow data by a single individual. Quadrants were based on both isotype controls and parallel unstimulated samples; these quadrants did not tend to vary between patients. An example of an instance of spontaneous cytokine production is provided in Fig. . Lymphocytes were defined on the basis of forward and side scatter; monocytes were defined on the basis of a wide gate based on forward and side scatter of stimulated and unstimulated CD14+ cells. (iv) Statistical techniques. | Comparisons of continuous variables between those with and without spontaneous cytokine production were made using Wilcoxon rank sum tests. Correlations were assessed using Pearson and Spearman correlation coefficients (rp and rs, respectively). The significance level was set at 0.05. FIG. 1. | Dot plot of unstimulated CD3+ lymphocytes of an individual spontaneously producing IL-6. Dot plot of unstimulated CD3+ lymphocytes of an individual spontaneously producing IL-6. FL1, anti-CD8-FITC; FL2, anti-IL-6-PE. TABLE 1 | Percentages of cells staining with monoclonal antibodies to various cytokines by cell and cytokine types and spontaneous cytokine production category RESULTS : Individual cytokines and cell types. | Spontaneous cytokine production was defined for seven cytokines and assessed in up to seven cell types . The numbers of patients with cell-specific spontaneous cytokine production according to our definition are provided in Table , as are the median and range of the percentage of each cell type producing each cytokine. The overlap in the ranges between those defined as being with and without spontaneous production is due to differences in the ranges for adult and pediatric patients and the differences in the age distributions of those with and without spontaneous cytokine production. Except in the case of IL-6 (see below), these age-associated differences tended to be slight and were not in any consistent direction (data not shown). Our definition of spontaneous cytokine production was chosen to be quite conservative, in that the lower limit for spontaneous production was in all cases well over 1% . The highest percentages of any cells spontaneously producing a cytokine were for IL-6; the majority of these instances were produced by the PBMCs of children, and the percentages themselves were higher in children (e.g., for CD3+ lymphocytes, the figures were as follows: median, 69%; range, 68 to 71%; n = 6 [children] versus 21 and 61%, n = 2 [adults]). The percentages of cells producing IL-6 with or without stimulation were strongly associated with serum transferrin receptor levels, an indicator of chronic iron deficiency . Using the definition of spontaneous cytokine production established for this study, a slightly higher proportion of B lymphocytes than T (CD3+) lymphocytes spontaneously produced IL-4. The percentages of CD8- and CD8+ T cells producing IL-6 were comparable to one another and higher than the percentages of monocytes producing IL-6 spontaneously. Immunoregulatory NT cells were proportionately highest in producing IFN-gamma and TNF-alpha spontaneously, followed distantly by NK cells and CD8+ T cells. Using the definition provided above, 33 of the 62 healthy persons had peripheral blood cells that spontaneously produced one (n = 27 persons), two (n = 5 persons), or three (n = 1 person) cytokines. The T cells of three healthy participants spontaneously produced IL-2, and those of three spontaneously produced IL-10. The B cells of one healthy person made IL-4. The CD8- T cells of two healthy persons made IL-8, and those of nine made IL-6. The CD8+ T cells of three healthy persons made IL-8, those of three made IFN-gamma, and those of nine made IL-6. The NK cells of three healthy persons made TNF-alpha, as did the NT cells of two healthy persons. Monocytes spontaneously produced IL-6 (one healthy person), IL-8 (six persons), IL-10 (two persons), or TNF-alpha (two persons). Thus, for these healthy persons, few made more than one cytokine spontaneously and IL-6, not the proinflammatory cytokine TNF-alpha, was the cytokine most commonly made. There was no predominant cell type making cytokines spontaneously, but immunoregulatory cells predominated in spontaneous TNF-alpha production. Cytokine and cell type constellations. | Unlike the healthy participants, most patients producing cytokines spontaneously produced more than one cytokine type . For both those spontaneously producing only one and those spontaneously producing more than one cytokine, the cytokine most frequently made was the proinflammatory cytokine TNF-alpha, followed by IFN-gamma. Production of a given cytokine by multiple cell types was frequent but did not occur in the majority of cases. For example, for IL-4, 4 patients had both T and B cells spontaneously producing the cytokine while a total of 18 patients had T cells and 17 had B cells spontaneously producing it. Similarly, three patients had both T cells and monocytes producing IL-6 and five had both CD8- and CD8+ T cells making IL-6. For IL-8, these values were four and six patients, respectively. Four patients had both CD8- and CD8+ T cells spontaneously producing IFN-gamma. The overlap was greatest between NT and NK cells, with fully half (nine patients) of those producing IFN-gamma spontaneously in NT or NK cells producing it in both these immunoregulatory cell populations. This was not the case for TNF-alpha, of which only two patients had spontaneous production in both NT and NK cells. Thirteen adult patients and nine children had two to eight combinations of cells spontaneously producing type 1 cytokines without any cell type spontaneously producing any type 2 cytokine. Conversely, five adult patients and seven children had two to five combinations of cells spontaneously producing type 2 cytokines without any cell type spontaneously producing any type 1 cytokine. Nineteen adult patients and 17 children had both one to six combinations of cells spontaneously producing type 1 cytokines and one to five combinations of cells spontaneously producing type 2 cytokines. Four adult patients and six children had two to six cell types spontaneously producing TNF-alpha and IL-8; two adult patients and four children had two or three cell types spontaneously producing IL-4 and/or IL-10. For healthy participants, the mix of cell types spontaneously producing a given cytokine was similar to that of the patients. For six persons, both the CD8+ and CD8- T cells spontaneously produced IL-6; for three persons each, only CD8- or CD8+ T cells spontaneously produced IL-6; and for one person, only monocytes spontaneously produced IL-6. For two persons, both CD8+ and CD8- T cells spontaneously produced IL-8; for one person, only CD8+ T cells spontaneously produced IL-8; and for six persons, only monocytes spontaneously produced IL-8. For three healthy persons, only T cells spontaneously made IL-10, and for two persons, only monocytes spontaneously made IL-10. For three healthy persons, only NK cells spontaneously made TNF-alpha; for one person, only NT cells spontaneously made TNF-alpha; for one person, only monocytes spontaneously made TNF-alpha; and for one person, monocytes and NT cells spontaneously made TNF-alpha. Of the six healthy persons producing more than one cytokine spontaneously, two made both IL-6 and TNF-alpha, one made IL-6 and IL-10, one made IL-8 and IL-10, one made IL-8 and TNF-alpha, and one made IL-4, IL-8, and IFN-gamma. Effects of spontaneous cytokine production on induced (stimulated) production by the same cell type. | For IL-2, IL-4, IL-6, IL-8, and IL-10, spontaneous cytokine production by a given cell type tended to be or was significantly associated with a higher percentage of the same cell type producing the same cytokine with stimulation, with the single exception of children's T-cell production of IL-4, for which there was a nonsignificant trend in the opposite direction (data not shown). Findings for IFN-gamma varied by cell type, with NK and NT responding in a pattern similar to those of other cytokines, CD3+ CD8- cells without an obvious effect, and the percentage of CD3+ CD8+ cells producing IFN-gamma with stimulation being lower in those for whom this cell type was spontaneously producing the cytokine . For TNF-alpha, the only significant effect was found in NK cells, again in a positive direction . These trends were present in both adult and pediatric patients (data not shown). The numbers of healthy participants making cytokines spontaneously were too small for statistical analyses, except in the cases of IL-6 production by T cells and IL-8 production by monocytes (see below). The trends were in the same direction found with the patients, except that spontaneous production of IL-10 by T cells did not affect induced production (data not shown). In healthy persons, as in the patients, the percentage of each cell type producing IL-6 with induction was significantly higher in those with spontaneous IL-6 production than in those without spontaneous production (medians for CD8- T cells, 9.3 versus 0.3% [P < 0.001], and for CD8+ T cells, 7.6 versus 1.2% [P < 0.001]). As with patients, the spontaneous production of IL-8 by monocytes did not have a significant effect on the percentages of monocytes producing IL-8 with induction (74.2 versus 58.3%; P = 0.398). Effects of spontaneous cytokine production on induced (stimulated) production by other cell types. | We next sought evidence of intercellular regulation of cytokine production by examining the relationships between spontaneous production of a cytokine by one cell type and the induced production of that same cytokine by another cell type. Few relationships were found; those for IL-4, IL-6, and IFN-gamma were in a positive direction . For two proinflammatory cytokines, IL-8 and TNF-alpha, in adults but not in children, the spontaneous production of a cytokine was associated with a lower percentage of another cell type producing that cytokine with stimulation. This was especially the case for TNF-alpha . Findings for adult and pediatric patients were similar to one another in regard to the other cytokines (data not shown). For healthy participants, the effects of spontaneous production of IL-6 on induced production by other cell types were similar to those for the patients. For example, for healthy persons, the median percentages of CD8- T cells and of monocytes making IL-6 with induction were higher for those whose CD8+ T cells made IL-6 spontaneously than for those whose CD8+ T cells did not make IL-6 spontaneously (7.4 versus 0.3% [P < 0.001] and 12.2 versus 1.0% [P < 0.001], respectively). Unlike in the patients, in these healthy persons, spontaneous production of IL-8 by monocytes did not affect the percentage of CD8+ T cells making induced IL-8 (medians, 1.7 versus 2.8%; P = 0.365). Relationships between the percentages of cells spontaneously producing cytokines and serum cytokine levels. | Only one patient (each) with spontaneous cell-specific IL-2 or IL-6 production had the matching serum cytokines assessed; each had detectable levels of the matching serum cytokine: IL-2, 17 pg/ml, and IL-6, 445 pg/ml. For the other cytokines, the numbers of patients with spontaneous cell-specific production and serum cytokine levels assessed were 9 for IL-4, 18 for IL-8, 11 for IL-10, 32 for IFN-gamma, and 34 for TNF-alpha. Of these, the numbers with detectable serum cytokines were 0 for IL-4, 13 for IL-8, 8 for IL-10, 19 for IFN-gamma, and 6 for TNF-alpha. No healthy participants and few children had serum cytokines assessed. No child with spontaneous cellular production had detectable levels of IL-4, IL-10, or TNF-alpha. All three children with spontaneous cellular IL-8 production also had detectable serum IL-8, and both children with spontaneous cellular IFN-gamma production had detectable serum IFN-gamma. Similarly, of 5 adult patients with spontaneous cellular IL-4 production, none had detectable serum IL-4; of 15 adult patients with cellular IL-8, 10 had detectable serum IL-8; of 30 adult patients with cellular IFN-gamma, 17 had detectable serum IFN-gamma; and of 29 adult patients with cellular TNF-alpha, only 6 had detectable serum TNF-alpha. We did correlation analyses between the percentages of various cell types spontaneously producing a given cytokine and the levels of that cytokine in the serum, with the caveat that the numbers (see above) were quite small. In analyses where undetectable serum cytokine levels were treated as being half the detection limit, correlations were significant between the percentage of monocytes spontaneously producing IL-8 and serum IL-8 levels (rs = +0.55; P = 0.018; n = 18), between the percentages of lymphocytes and of T cells spontaneously producing IFN-gamma and serum IFN-gamma levels (rs = +0.44, P = 0.014, n = 31 and rp = +0.36, P = 0.043, n = 32, respectively), and between the percentages of lymphocytes and of NT cells spontaneously producing TNF-alpha and serum TNF-alpha levels (rs = +0.37, P = 0.030, n = 34 and rs = +0.48, P = 0.004, n = 34, respectively). When those with undetectable levels were excluded from the analyses, correlations were significant between the percentage of lymphocytes spontaneously producing IFN-gamma and serum IFN-gamma levels (rs = +0.57; P = 0.014; n = 18) and between the percentages of lymphocytes spontaneously producing TNF-alpha and serum TNF-alpha levels (rs = +0.90; P = 0.015; n = 6). TABLE 2 | Numbers of patients spontaneously making >=1 cytokine TABLE 3 | Median percentages of cells expressing cytokine with stimulation, by presence or absence of spontaneous production of that cytokine TABLE 4 | Median percentages of cells producing cytokine with stimulation, by presence or absence of spontaneous production of that cytokine by a different cell type DISCUSSION : We report here that the peripheral blood cells of some hospitalized and some apparently healthy persons in Malawi produced cytokines ex vivo without additional stimulation . Spontaneous ex vivo cytokine production by peripheral blood cells rarely occurs in healthy U.S. or European individuals or even in those with a wide variety of chronic and/or systemic diseases, including allergic asthma, collagen vascular diseases, and sarcoidosis . The clinical and demographic characteristic of the Malawi patients spontaneously producing cytokines will be discussed elsewhere, but suffice it to say, these participants were not analogous to healthy U.S. blood donors. HIV, malaria, numerous intestinal parasites, and mycobacterial infections are all endemic in Malawi. In addition, by the enrollment criteria, the patients in this study were acutely ill. Since there is great interest in the roles of cytokines in acute and chronic infections, spontaneous cytokine production by the peripheral blood cells of persons within our study population warranted further investigation. We therefore examined data from these Malawi patients and the ostensibly healthy individuals to characterize this previously unstudied phenomenon and to assess its relationship to induced cell-specific cytokine production. By the criteria used, spontaneous cytokine production was not rare in this study group and often involved NK and NT cells, two immunoregulatory cell populations important in mycobacterial and HIV infections but only infrequently examined in clinical studies. TNF-alpha, a proinflammatory cytokine, was the cytokine most often produced spontaneously by patients, consistent with these patients having acute inflammation. The type 1 cytokine IFN-gamma was next in frequency, often being made by NK and NT cells, followed by the chemokine IL-8, which activates neutrophils and neutrophil adhesion to vascular endothelium. This order appears reasonable given the participants' acutely ill state and the frequency of HIV and mycobacterial infections in this population . Spontaneous production of type 2 cytokines was less frequent than that of the proinflammatory and type 1 cytokines, and a higher proportion of patients spontaneously produced the type 2 cytokine IL-4 than the type 2 cytokine IL-10. Production of a given cytokine by one cell type was not necessarily associated with production by other cell types; however, most patients who produced cytokines spontaneously produced more than one type of cytokine. For the healthy individuals, IL-6 was the cytokine most frequently produced spontaneously, and spontaneous production of a single cytokine was the rule, not the exception. Spontaneous cytokine production has both physiologic and practical implications. Cytokines affect infectious-disease morbidity and mortality and the interaction between various opportunistic infections and HIV. Presumably, spontaneous ex vivo cytokine production reflects in vivo stimulation and thus represents cell-specific activity in the context of a current infection(s) and/or physical condition and in the actual natural milieu of that cell. Consistent with this assumption, in previous assessments of this Malawian study group we found spontaneous IFN-gamma production by NT cells to be associated with mortality and the spontaneous production of a number of other cytokines by various cell types to be associated with malaria parasitemia , iron deficiency , or vitamin A deficiency . In nonhospitalized HIV-infected persons in the United States, we found significant declines in spontaneous TNF-alpha and IL-4 production by T cells to be associated with the acute response to highly active antiretroviral therapy . These previously published data support the potential physiological importance of measurable spontaneous cytokine production. However, we and many others have also reported significant relationships between induced cytokine production and a wide variety of illnesses and physiologic states. Thus, we were concerned about whether spontaneous cytokine production might obscure or alter the more commonly measured induced cytokine production. Ex vivo induction theoretically expands a stimulated population to measurable levels; however, human data to support the idea that this expanded population is representative of in vivo cell-specific cytokine profiles are sparse at best. As with measurements of serum cytokine levels, the underlying assumption is that peripheral-blood findings are of importance in systemic, or even localized, diseases. This is somewhat of an assumption in regard to both measures. We have previously shown that, in general, serum and cell-specific cytokines are at best only weakly associated with one another . Correlations between cell and serum cytokine parameters were not infrequently even negative, consistent with the hypothesis that serum and cell-specific cytokines measure different things . Interestingly, for the patients with spontaneous cytokine production discussed here, the correlations between serum and peripheral-blood cell-specific cytokines were all in a positive direction but, as with the more general population, generally with only weak correlation coefficients (this study). Theoretically, if cells have been stimulated by in vivo antigen exposure, secondary ex vivo induction with phorbol esters could lead to in vitro apoptosis and obscure the role of that cell-specific cytokine in the physiologic process. This would be analogous to measuring expression of variable beta-chain families of the T-cell receptor too long after superantigen stimulation . In this Malawi study, we found that in many instances, spontaneous cytokine production was not associated with a significant difference in induced production. When a relationship was found, it was usually in a positive direction, not a negative one, as would have been expected with apoptosis. IL-6 findings were especially interesting in this regard. They strongly followed the described pattern; however, for this cytokine, the cell percentages were actually higher for unstimulated than for induced cells (e.g., for patients' CD3+ CD8+ cells, the medians were 85.9 versus 53.7%), suggesting that some degree of activation-associated cell death might have been taking place but that cell division and cytokine induction were generally keeping in balance with the negative process. Further, this IL-6 finding strongly suggests that the cytokine was far more strongly induced by events occurring in vivo, e.g., iron deficiency , than by the stimulation provided in vitro. Finally, evidence of down modulation of induced cytokine production by one cell type in the presence of spontaneous production by a different cell type was found only for TNF-alpha (among T, NT, and NK cells) for adult patients . In conclusion, spontaneous cytokine production, albeit unusual in industrialized societies, can occur in some individuals and may not be that unusual in persons in developing countries. These findings strongly suggest that in certain study populations ---e.g., those living in developing countries, with acute illness, and/or with HIV infection ---spontaneous cell-specific cytokine measurements should be recorded in addition to the measurement of induced cell-specific cytokine production. One parameter does not negate the potential importance of the other; both may provide useful information concerning the roles of cytokines in various disease states. In light of the variable effects of spontaneous cytokine production on induced production, it may also be useful to calculate and assess the subtractive or proportionate difference between the percentage of cells producing a given cytokine without stimulation and the percentage producing that cytokine with stimulation. Backmatter: PMID- 12204968 TI - Modulation of Human Immunodeficiency Virus (HIV)-Specific Immune Response by Using Efavirenz, Nelfinavir, and Stavudine in a Rescue Therapy Regimen for HIV-Infected, Drug-Experienced Patients AB - Analysis of the virologic and immunomodulatory effects of an association of efavirenz (EFV), nelfinavir (NFV), and stavudine (d4T) was performed in 18 human immunodeficiency virus (HIV)-infected and highly active antiretroviral therapy (HAART)-experienced patients who failed multiple therapeutic protocols. Patients (<500 CD4+ cells/mul; >10,000 HIV copies/ml) were nonnucleoside reverse transcriptase inhibitor (NNRTI)-naive and were treated for 10 months with EFV (600 mg/day) in association with NFV (750 mg three times daily) and d4T (30 or 40 mg twice daily). Measurement of HIV peptide- and mitogen-stimulated production of interleukin-2 (IL-2), gamma interferon (IFN-gamma), IL-4, and IL-10 as well as quantitation of mRNA for the same cytokines in unstimulated peripheral blood mononuclear cells were performed at baseline and 2 weeks (t1), 2 months (t2), and 10 months (t3) into therapy. The results showed that HIV-specific (but not mitogen-stimulated) IL-2 and IFN-gamma production was augmented and IL-10 production was reduced in patients who received EFV, NFV, and d4T. Therapy was also associated with a reduction in HIV RNA in plasma and an increase in CD4+ cell count. These changes occurred in the first year of therapy (t2 and t3) and were confirmed by quantitation of cytokine-specific mRNA. Therapy with EFV, NFV, and d4T increases HIV-specific type 1 cytokine production as well as CD4 counts and reduces plasma viremia. This therapeutic regimen may be considered for use in cases of advanced HIV infection. Keywords: Introduction : The progression of human immunodeficiency virus (HIV) infection to AIDS is associated with increasing HIV viral load and qualitative and quantitative defects affecting CD4 T lymphocytes and cell-mediated immunity . Drugs designed as therapeutic tools against HIV infection should thus ideally be capable of stimulating cell-mediated immunity and reducing HIV viremia. Nevertheless, whereas highly active antiretroviral therapy (HAART) can achieve suppression of HIV viremia in HIV-infected individuals, the efficacy of these antiretroviral therapies in reconstituting immune function has been less dramatic . Nonnucleoside reverse transcriptase inhibitors (NNRTI), nucleoside reverse transcriptase inhibitors (NRTI), and protease inhibitors (PI) are currently used in the treatment of HIV-1 infection (reviewed in references , , , and ). Efavirenz (EFV) is an NNRTI which in combination with other antiretroviral drugs induces viral suppression as well as increases in CD4 counts. Treatment with EFV may be provided on a once-a-day, no-food-interaction schedule that makes its compliance higher than that of other drugs. Furthermore, although cross-resistance within the NNRTI class is extensive and the presence of point mutations (K103N, Y181C) may exclude the efficacy of the entire class, there is some evidence that EFV may retain full activity against more than 25% of Y181C-mutated viruses , even though an NNRTI sequential treatment is still not recommended by international guidelines . Because of these interesting pharmacokinetic properties we decided to evaluate the potential usefulness of this compound in advanced HIV infection. Therefore, we investigated immune and virologic parameters in patients undergoing therapy with an EFV-containing regimen. In particular, because progression to AIDS and loss of CD4+ T cells is associated with the impairment of type 1 cytokine production, we verified that EFV-containing antiviral therapies stimulated cell-mediated immunity and a type 1 cytokine profile. MATERIALS AND METHODS : Patients. | Eighteen HIV-infected patients were randomly selected among those enrolled in a study to evaluate the efficacy of an EFV-, nelfinavir (NFV)-, and stavudine (d4T)-based rescue therapy . The only selection bias was the patient's willingness to be included in the study. These patients had failed previous combination therapies (including multiple nucleoside analogues and at least one PI) and they (i) were failing the current PI-containing therapy, (ii) were naive for NNRTIs, (iii) had never received NFV, (iv) had no evidence of active HIV-related infections or neoplasms; and (v) had never been treated with immunomodulants (e.g., interleukin-2 [IL-2]). EFV was administered at a dose of 600 mg daily at bedtime, NFV was given at 750 mg every 8 h, and d4T was given at 30 or 40 mg, depending on body weight, twice daily. Blood sample collection. | Whole blood was collected by venipuncture immediately before treatment (t0) and 2 weeks (t1) as well as 2 months (t2) and 10 months (t3) after the start of therapy in EDTA Vacutainer tubes (Becton Dickinson & Co., Rutherford, N.J.). Peripheral blood mononuclear cells (PBMC) were separated on lymphocyte separation medium (Organon Teknika Corp., Durham, N.C.) and washed twice in phosphate-buffered saline, and the number of viable leukocytes was determined by trypan blue exclusion. PBMC were resuspended in RPMI 1640 (Sigma, St. Louis, Mo.) plus 2% AB serum (Sigma). All experiments were performed using cryopreserved PBMC. Antigen- and mitogen-stimulated cytokine production. | Gamma interferon (IFN-gamma), IL-2, IL-4, and IL-10 production by PBMC was determined by culturing 3 x 106 PBMC per well in 24-well LINBRO plates (Flow Laboratories, Inc.) (37C; 7% CO2). PBMC were unstimulated or stimulated with phytohemagglutinin (PHA) (M form; Sigma, St. Louis, Mo.) (final dilution, 1:100) for 48 h or a pool of five synthetic peptides from the gp160 envelope (Env) of HIV type 1 (HIV-1) (2.5 mM final concentration) for 5 days. The following peptides were utilized: T1 (KQIINMWQEVGKAMYA), HIV-1/IIIb Env amino acid residues 428 to 443; T2 (HEDIISLWDQSLK), HIV-1/IIIb Env amino acid residues 112 to 124; TH4.1 (DRVIEVVQGAYRAIR), HIV-1/IIIb Env amino acid residues 834 to 848; p18-IIIb (RIQRGPGRAFVTIGK), HIV-1/III IIIB Env amino acid residues 315 to 329; and p18-MN (RIHIGPGRAFYTTKN), HIV-1 (MN) Env region homologous to p18-IIIB. Cytokine production was evaluated in the supernatants with commercially available enzyme-linked immunosorbent assays (Endogen, Woburn, Mass.). RNA extraction. | RNA was extracted from unstimulated lymphocytes with the acid guanidium thiocyanate-phenol-chloroform method and dissolved in RNase-free water; purity was determined by spectrophotometry. Genomic DNA contamination was removed by RNase-free DNase (RQ1 DNase; Promega, Madison, Wis.). Reverse transcription (RT). | One microgram of RNA was reverse transcribed into first-strand cDNA in a 20-mul final volume containing 1 mM random hexanucleotide primers, 1 mM oligo(dT), and 200 U of Moloney murine leukemia virus reverse transcriptase (Clontech, Palo Alto, Calif.); cytokine mRNA expression was normalized for beta-actin cDNA content by semiquantitative PCR (Clontech). Quantification of cytokine cDNA by PCR. | PCR was performed in 50 mul of reaction mixture containing 10 mul of RT reaction mixture, 1x PCR buffer (20 mM Tris-HCl, 100 mM KCl, 0.1 mM EDTA, 1 mM dithiothreitol, 0.5% Tween 20, 0.5% Nonidet P-40, 50% glycerol), 200 mM dNTPs, 1.25 U of Taq polymerase (Takara, Otsu, Japan), 0.4 mM beta-actin primers, and 0.4 mM (each) cytokines. The following primers were used: IL-4 5' primer, 5'CGGCAATTTGACCACGGACACCCGTGCATA3'; IL-4 3' primer, 5'ACGTACTCTGGTTGGCTTCCTTCACAGGACAG3'; IL-10 5' primer, 5'AAGCTGAGAACCAAGACCCAGACATCAAGGCG3'; IL-10 3' primer, 5'AGCTAT CCCAGAGCCCCAGATCCGATTTTGG3'; IFN-gamma 5' primer, 5'GCATCGTTTTGGGTTCTCTTGGCTGTTACTGC3'; IFN-gamma 3' primer, 5'CTCCTTTTTCGCTTCCCTGTTTTAGCTGCTGG3'; tumor necrosis factor alpha (TNF-alpha) 5' primer, 5'GAGTGACAAGCCTGTAGCCCATGTTGTAGCA3'; TNF-alpha 3' primer, 5'GCAATGATCCCAAAGTAGACCTGCCCAGACT3'; TNF-beta 5' primer, 5'ATGACACCACCTGAACGTCTCTTC3'; TNF-beta 3' primer, 5'GAAGGCTCCAAAGAAGACAGTACT3'. Thermal cycling was performed in a Touchdown Hybaid apparatus (Celbio, Milan, Italy) as follows: initial denaturation at 95C for 10 min; 35 cycles of denaturation at 95C for 30 s, annealing at 60C for 30 s, and extension at 72C for 30 s; and a final amplification step at 72C for 10 min. PCR products were electrophoresed in a 10% acrylamide gel and stained with 0.5 mg of ethidium bromide per ml; the size of cDNA products was determined by comparison to a DNA size marker, pbR322 (Sigma). The bands on the gels were scanned by transmission densitometry to quantify relative levels of gene expression, and the areas of the peaks were calculated in arbitrary units. To evaluate the relative levels of expression of the target genes in RT-PCR, the value of the internal standard (beta-actin) was used as the baseline gene expression of that sample, and the relative value was calculated for the target genes amplified in that reaction. Statistical analysis. | Statistical analysis was based on a nonparametric Jonckheere-Terpstra test for trends. Data were also analyzed by a nonparametric Kruskal-Wallis test; P values were two sided. RESULTS : Demographic profile of the study cohort. | The patients enrolled in the protocol (7 females and 11 males) had a mean age of 33.8 years. Four had a diagnosis of AIDS in their past medical history and eight had a CD4+ cell count below 200 cells/mul at the baseline. None of the patients had been immunized with candidate HIV vaccines; patients had previously received a mean of 3.6 NRTI and 2 PI. All of the patients were tested at the time points shown in the text. CD4+ cell counts progressively increased in all individuals after the initiation of therapy. Thus, mean (range) CD4 cell counts were as follows: t0, 212 (115-328) CD4 cells/mul; t1, 227 (131- 350) CD4 cells/mul; t2, 291 (174-444) CD4 cells/mul; t3, 399 (312-497) CD4 cells/mul. The change was significant (P = 0.011) when the baseline value was compared with the value observed after 10 months of therapy (t3). HIV genomic RNA was detectable in all patients at the baseline (mean, 109,217; range, 62,488-198,714 copies/mul) and decreased during therapy in all cases (t1 mean, 104,100; range, 58,227-196,515 copies/mul; t2 mean, 39,744; range, 21,057-74,480 copies/mul; t3 mean, 11,518; range, 499 to 15,511 copies/mul). HIV RNA at t3 was undetectable for 12 of 18 subjects. A trend toward a greater increase in CD4 counts was seen in the 12 patients who achieved suppression of plasma viremia below the limit of detection. Modulation of immune functions. | Env-stimulated production of IL-2 and IFN-gamma was augmented whereas IL-10 production was reduced in the patients. These results (means +- standard errors) are shown in Fig. . Thus, Env-stimulated IL-2 and IFN-gamma production was significantly increased 2 months after initiation of therapy (t2); the increase was still observed at t3 (P < 0.01 in all cases). IL-10 production was significantly diminished at the last time point (t3) (P < 0.01) . Mitogen-stimulated IL-2 production was marginally improved at t3; the production of both IFN-gamma and IL-10 upon mitogen stimulation was augmented, albeit not significantly, at t1 and t2 and returned to baseline by the end of the study period . Modulation of cytokine mRNA. | Cytokine-specific mRNA was quantified in resting PBMC of all patients at the beginning and after 10 months of therapy (cytokine mRNA could not be examined at t1 and t2 because of limitations in the number of available PBMC). Because TNF-alpha exerts a direct effect on viral replication and TNF-beta is suggested to be a mediator of apoptotic cell death in HIV infection, the expression of these two cytokines was measured as well. Results showed that after 10 months of therapy, (i) the expression of IFN-gamma was higher and (ii) the expression of IL-10, IL-4, TNF-alpha, and TNF-beta was lower. IL-2 was not detectable in resting PBMC (Fig. and Table ). FIG. 1. | (A) Env-stimulated cytokine production (IL-2, IFN-gamma, IL-4, and IL-10) by PBMC of HIV-infected individuals at different time points (before treatment [t0] and 2 weeks [t1], 2 months [t2], and 10 months [t3] into therapy). (A) Env-stimulated cytokine production (IL-2, IFN-gamma, IL-4, and IL-10) by PBMC of HIV-infected individuals at different time points (before treatment [t0] and 2 weeks [t1], 2 months [t2], and 10 months [t3] into therapy). The results are shown as means +- standard errors. (B) PHA-stimulated production of the same cytokines. *, values are significantly different from those at t0. FIG. 2. | Quantification of mRNA for IL-4, IL-10, IFN-gamma, TNF-alpha, and TNF-beta in resting PBMC of one representative HIV-infected individual before (t0) and after 10 months of (t3) therapy. Quantification of mRNA for IL-4, IL-10, IFN-gamma, TNF-alpha, and TNF-beta in resting PBMC of one representative HIV-infected individual before (t0) and after 10 months of (t3) therapy. The upper bands show cDNA retrotranscribed from cytokine mRNA extracted from patient PBMC; the lower bands show cDNA retrotranscribed from beta-actin mRNA extracted from patient PBMC and used to normalize the PCRs. TABLE 1 | Cytokine-specific mRNA in HIV-infected patients undergoing a rescue therapy DISCUSSION : The activities of an EFV-, NFV-, and d4T-based therapy on immunologic and virologic parameters were evaluated by analyzing the effects of this association in 18 randomly selected, HIV-infected, heavily pretreated individuals who failed previous therapeutic regimens . The combination was well tolerated, resulted in increased CD4 cell counts and reduced HIV plasma viremia, and at the same time induced a positive trend in HIV-specific type 1 cytokine production as well as a significant reduction in the generation of IL-10. To verify if the variations in cytokine production were secondary to modifications in their patterns of expression, the amount of cytokine-specific mRNA was analyzed in resting PBMC. The results showed that IFN-gamma mRNA was more abundant while IL-10 and IL-4 mRNAs were expressed less after 10 months of therapy. Because the generation of type 1 and type 2 cytokines is cross-regulatory, it is not surprising that during follow-up the decreased expression and production of IL-10 coincided with an augmented generation of IL-2 and IFN-gamma. Env-stimulated IL-2 and IFN-gamma production increased before down modulation of IL-10 production could be observed. A speculative hypothesis could be that the drug regimen studied here preferentially stimulates type 1 cytokine production and that the increase of these cytokines results in the subsequent cross-regulation of IL-10 production. TNF-alpha expression was reduced in this study. Because TNF-alpha directly stimulates viral replication , the modulation of this cytokine could contribute to the reduction of HIV viral load observed in our patients. CD4 cell counts progressively improved concomitantly with a reduction of TNF-beta expression in the study presented here. Because TNF-beta is known to be a mediator of apoptotic cell death in HIV infection it is tempting to speculate that increases in CD4 counts are at least partially secondary to a reduced susceptibility of CD4 cells to apoptosis. The analysis of the efficacy data relative to the entire cohort of patients showed that plasma viral load was reduced by 1.7 log at 12 months with a sustained increase in CD4+-T-cell count . HIV plasma viremia was undetectable in 12 of the 18 patients after 1 year of therapy. In light of recent data suggesting that HIV-specific T-cell function would not be recovered in patients with fully suppressed viral replication (reviewed in reference ), we compared Env-specific cytokine production in patients in whom viremia was or was not still detectable after therapy. We did not observe any correlation between residual viremia and improvement in immune function. This result can probably be explained considering the small number of individuals studied and the fact that our patients had advanced illness and had failed previous combination therapies. It is interesting to observe that modulation of HIV-specific immune responses was detected in our patients within a few months. These results contrast with recent data showing that improvement in HIV-specific immunity may not be seen in the first year of antiviral therapy . The discrepancy could be secondary to the fact that, whereas we measured Env-specific responses, Angel et al. and Blankson et al. stimulated PBMC with Gag or Pol. Numerous studies showed that HAART positively influences quantitative and qualitative immune parameters in both acute and chronic HIV infection, but the immune reconstitution that ensues nevertheless is never complete . Thus, triple therapy is associated with improvement in CD4 counts and functions that are more easily seen in early- than in late-stage patients (, , , , , , , , , , , ; M. Albrecht, D. Katzenstein, R. Bosch, and S. Liou, Abstr. 7th Conf. Retrovir. Opportunistic Infect., abstr. 531, 2000). Our data show that partial modulation of immune functions in chronic HIV infection can be seen even in patients who had previously failed multiple regimens. NNRTIs, in addition to NRTIs and PIs, have gained a definite place in the treatment of HIV-1 infection (reviewed in references , , , and ). EFV is an NNRTI which is used in combination with other antiviral compounds in the treatment of HIV infection and is shown to knock out virus replication and to delay virus resistance from arising (; Albrecht et al., 7th Conf. Retrovir. Opportunistic Infect.). Treatment guidelines were recently modified to include NNRTI as an acceptable substitute for a PI in primary therapy . Results stemming from a number of protocols in which this therapeutic approach was used in HIV-infected individuals confirmed that NNRTI-including regimens are capable of achieving a very consistent suppression of HIV plasma viremia. For the 006 study participants were randomized to receive EFV in association either with nucleoside analogues or with PIs. The results showed that suppression of plasma HIV-1 RNA to undetectable levels was achieved in a higher percentage of patients in the NNRTI plus nucleoside analogues group than in the PI plus nucleoside analogues group . In the ACTG 302 study NNRTI- and PI-naive patients who had previously failed an NRTI therapy were randomized to receive EFV, NFV, or both drugs in combination to 1 or 2 NRTIs. By week 40 the percentage of individuals with a viral load below the detection limit was significantly higher in the EFV- than in the NFV-based arms (Albrecht et al., 7th Conf. Retrovir. Opportunistic Infect.). Finally, recent data comparing the virologic effect of NNRTI- and PI-based antiretroviral combinations suggest that the use of NNRTI is associated with a faster and more frequent initial virologic response . In conclusion, our results suggest that NNRTI-employing antiviral therapies could be useful as a rescue regimen in advanced cases of HIV infection. Because these data are based on a small group of patients and because the design of this preliminary study did not involve a control arm, further studies will be necessary to clarify the best way to utilize this strategy in advanced HIV disease. Backmatter: PMID- 12204971 TI - Helicobacter pylori-Specific Immune Responses of Children: Implications for Future Vaccination Strategy AB - We analyzed the specific anti-Helicobacter pylori immunoglobulin G (IgG) antibody profile for a sample of 824 asymptomatic schoolchildren in southern Germany (mean age, 10.7 +- 0.65 years) with an H. pylori-specific IgG enzyme-linked immunosorbent assay and Western blot analysis. The prevalence of infection was 19.8% (95% confidence interval, 17.1 to 22.7%). The immunoresponses were characterized predominantly by antibodies against low-molecular-mass antigens of 14 and 29 kDa, with a significant difference between children of German and Turkish nationalities (P = 0.0012 and P < 0.0001, respectively). Keywords: Introduction : Helicobacter pylori plays an etiologic role in gastritis, peptic ulcer disease, gastric cancer, and mucosa-associated lymphoid tissue lymphoma , and acquisition of the infection seems to occur mainly in the first years of life . Genetic typing has shown that H. pylori is extremely diverse as a species and possesses a number of antigens that induce a specific immune response in infected individuals . The specific immune response is part of the natural history of the infection in children, which remains poorly characterized, and may mirror the specific interaction of the host with the bacterium (, , , 21). This immune response may be of special importance with regard to vaccination, as it may determine the antigens considered for a vaccine. However, the few existing studies of the infection in children are based mainly on symptomatic and small clinical samples . The aim of this study was to analyze the specific anti-H. pylori immunoglobulin G (IgG) antibody profile in a sample of asymptomatic schoolchildren. This study is part of a population-based investigation performed by the Baden-Wurttemberg State Health Office between October 1999 and April 2000 with a sample of schoolchildren (fourth graders; mean age +- standard deviation, 10.7 +- 0.65 years; range, 9 to 13 years; median, 11 years) living in southern Germany in four different areas of Baden-Wurttemberg (Stuttgart, Mannheim, Kehl, and Ravensburg). Participation in the study was voluntary. The study was approved by the Ethics Board of the University of Ulm. The parents of 1,118 children filled out a questionnaire which requested information regarding sociodemographic factors, family characteristics, and medical history. Serum samples from 824 (73.7%) of those 1,118 children were available. H. pylori IgG antibodies were determined by enzyme-linked immunosorbent assay (GAP; Bio-Rad Laboratories Diagnostics Group, Munich, Germany). According to the manufacturer's instructions, levels of >20 U/ml were considered positive. This test has previously been evaluated with 310 children (compared with the histology, culture, and urease test, the sensitivity of our test was 94.9% and the specificity was 92.4%) . The serum samples of the IgG-positive children were further analyzed for the presence of antibodies against H. pylori-specific antigens with a commercial Western blot method (Helicobacter pylori Western blot; AID, Strasberg, Germany) as described previously . Of the 824 children included in the study, 53.3% were boys. About two-thirds, or 63.5%, of the schoolchildren were of German nationality, 12.5% were Turkish, and 12.7% were of other nationalities (for 89 of the children (10.8%), there were no data available concerning nationality). Information gathered on family characteristics showed that 10.2% of the children had no siblings, 43.3% had one, and 38.1% had two siblings; 45.5% of the fathers and 37.5% of the mothers had 9 or fewer years of school education. One hundred sixty-three children (19.8% of the total number, 95% confidence interval [CI], 17.1 to 22.7%) had positive serology results for H. pylori . The seroprevalence in children of German nationality was 13.9% (95% CI, 11.0 to 17.1%) and varied from 12.4% (95% CI, 9.6 to 15.6%) for the German children born in Germany to 42.3% (95% CI, 23.4 to 63.1%) for the German children born elsewhere (mostly children of families who had lived in countries of Eastern Europe for generations and who had returned to Germany quite recently). The prevalence of H. pylori in children of Turkish nationality was 37.9% (95% CI, 28.5 to 48.0%) (in Turkish children, place of birth was not associated with seroprevalence), and for children of other nationalities, it was 30.5% (95% CI, 21.9 to 40.2%). These strong differences among ethnic groups were independent of socioeconomic factors, such as family size and school education of parents, and persisted after these factors had been controlled for. On the other hand, duration of school education of parents and number of siblings were not factors associated with seroprevalence if ethnic origin was considered in the analysis (data not shown). TABLE 1 | Prevalence of H. pylori infection in children according to nationality Western blots were used to detect and analyze eight different H. pylori-specific immunoreactive bands . TABLE 2 | Prevalence of specific H. pylori IgG antibodies in H. pylori-infected children (n = 163) Table shows that antibodies to the 25-kDa antigen and the 14-kDa antigen were detected much more frequently in infected children of Turkish or other nationality than in infected children of German nationality (P values of 0.0012 and <0.0001, respectively). TABLE 3 | Prevalence of H. pylori-specific antibodies identified by immunoblotting among infected children according to nationality of the children (n = 144) This study, conducted with a population-based sample of asymptomatic schoolchildren in Germany, demonstrated that the specific systemic immune response in children to H. pylori infection is particularly pronounced against low-molecular-weight proteins of H. pylori. Furthermore, there was a large variation in both the prevalence of infection and the specific antibody patterns with respect to the nationalities of the children, although the children were living in the same geographic region. As shown in a previous study, differences between ethnic groups are not explained by differences in socioeconomic factors and most likely mirror strain differences among these groups. A recent French study of 80 pediatric patients found the highest prevalence (88.7%) for antibodies to a 26-kDa antigen . Little is known about the function of the 26-kDa antigen, which may be identical to the 25-kDa antigen in our study. However, recently it has been shown that a 25-kDa outer membrane protein acts in a lectin-like manner with lipopolysaccharide to mediate attachment of H. pylori to laminin and may therefore be important for the colonization process in the early phase of the infection . Much interest exists in the seroprevalence in children of CagA antibodies, which, according to a study by Mitchell et al. , correlated with age and increased up to the age of 15 years. There are few data available for children, and the results of recent studies have been inconsistent concerning the prevalence and association with disease of CagA-positive strains in both adults and children (, 6, , , , , ). This may be due partly to the considerable variation in the frequencies of vacAsl subtypes and iceA alleles in children originating from different geographic regions . As for adults, a high prevalence of CagA-positive H. pylori strains was found in Japanese children; however, there was no association of this prevalence with nodular gastritis or peptic ulcer disease in Japanese children (mean age, 12 +- 3 years). This result challenges the role of CagA as a marker for more serious gastroduodenal disease in children as well as in adults in Japan . The highest prevalence of CagA antibodies in Europe (80%), has been found in a study of infected children in Poland ; the prevalence in infected schoolchildren in Estonia (9 to 15 years old) was 46% . Raymond et al. described the prevalence of CagA antibodies in children in France who had abdominal pain to be 43% . In conclusion, the differences in the antibody patterns between children of various ethnic groups or found in our study may indicate strain differences and should be considered if antigens are chosen as targets for a vaccination approach. Backmatter: PMID- 12204963 TI - Cytokine Gene Expression by Peripheral Blood Leukocytes in Horses Experimentally Infected with Anaplasma phagocytophila AB - Human granulocytic ehrlichiosis (HGE), a tick-borne zoonosis, is caused by an obligatory intragranulocytic bacterium, the HGE agent, a strain of Anaplasma phagocytophila. The equine model of HGE is considered valuable in understanding pathogenic and immune mechanisms of HGE. In the present study, cytokine mRNA expression by peripheral blood leukocytes (PBLs) in horses was examined during the course of infection by intravenous inoculation of A. phagocytophila or by allowing feeding by infected ticks. The p44 genes encoding the major outer membrane protein P44s of A. phagocytophila were detected by PCR in PBLs of all four horses from 4 to 20 days postexposure. During the 20-day infection period, interleukin-1beta (IL-1beta) and tumor necrosis factor alpha (TNF-alpha) mRNA expression was upregulated in PBLs of all four horses, and IL-8 mRNA expression was upregulated in three horses. Gamma interferon, IL-10, and IL-12 p35 mRNAs were weakly expressed in only one horse each. IL-2, IL-4, IL-6, and IL-12 p40 mRNA expression , however, could not be detected in the PBLs of any of the four horses. These results suggest that IL-1beta, TNF-alpha, and IL-8 generation during A. phagocytophila infection has a primary role in HGE pathogenesis and immunomodulation. Keywords: Introduction : Human granulocytic ehrlichiosis (HGE) is characterized by fever, chills, headache, myalgia, and laboratory findings including leukopenia, anemia, and thrombocytopenia, as well as elevated liver enzyme activities . Delayed treatment, misdiagnosis, and/or the presence of immunosuppression may lead to a severe or fatal outcome . HGE has been increasingly recognized in the United States and various parts of Europe . HGE is caused by a strain of Anaplasma phagocytophila (the HGE agent), a gram-negative obligatory intragranulocytic bacterium. The white-footed mouse (Peromyscus leucopus) is considered to be a major wild-animal reservoir of A. phagocytophila in the northeastern United States, and Ixodes spp. ticks, the vector of Borrelia burgdorferi, the agent of Lyme disease, are considered to be the vector . Strains of A. phagocytophila have also been known to cause tick-borne fever in ruminants in Europe and to cause equine ehrlichiosis in the United States (it has hence formerly been called Ehrlichia phagocytophila and Ehrlichia equi). Pathogenesis and cellular immune responses of HGE are not well defined. The presence of low levels of A. phagocytophila in the blood of patients in the acute stage and autopsied patients' tissues and the nature of clinical signs and laboratory findings suggest the involvement of proinflammatory cytokines and chemokines in the pathogenesis. In vitro, we found that A. phagocytophila strain HZ, isolated from an acute-stage HGE patient, induced rapid and strong proinflammatory cytokine (interleukin-1beta [IL-1beta], tumor necrosis factor alpha [TNF-alpha], and IL-6) mRNA expression by human peripheral blood leukocytes (PBLs) and monocytes within 2 h and protein secretion within 24 h . However, only IL-1beta is upregulated in neutrophils, and the mRNA of these three cytokines is not upregulated in HL-60 cells. Within 2 h of incubation with A. phagocytophila, IL-8, IL-10, gamma interferon (IFN-gamma), IL-2, and transforming growth factor beta mRNAs are not consistently upregulated in human PBLs, suggesting that for these cytokines, gene expression either is not induced or is induced at later time points in vitro. Akkoyunlu et al. reported IL-8 production by retinoic acid-treated HL-60 cells (human promyelocytic leukemia cell line) after 24 h (negative at 12 h for both mRNA and protein) and by human neutrophils 7 h postincubation with A. phagocytophila in vitro. That report also indicated that no IL-1alpha, IL-1beta, or TNF-alpha is detected in the culture supernatant of retinoic acid-treated and nontreated HL-60 cells at 6 days after the addition of A. phagocytophila. Klein et al. reported that in infected dimethyl sulfoxide-treated HL-60 cells or human bone marrow cells, IL-8 and other chemokine levels, but not that of IL-1, IL-6, or TNF-alpha in the culture medium, were significantly elevated at 48 h postinfection . IL-8 levels are significantly increased in the blood of HGE patients . Dumler et al. reported that IFN-gamma and IL-10 levels are elevated in acute-phase (median, 4 days after onset) sera compared with convalescent (median, 31 days after onset) sera from HGE patients or normal controls whereas serum IL-1beta, TNF-alpha, and IL-4 levels are not elevated . Using C3H mice, which do not show clinical signs and spontaneously clear the infection by 15 days after intraperitoneal inoculation of A. phagocytophila, splenic IL-12 and IFN-gamma mRNA and serum IFN-gamma levels are significantly elevated starting at 2 days postinfection, although the clearance of A. phagocytophila is not delayed in IFN-gamma-/- mice . Considering apparently conflicting data in vitro using cell lines, bone marrow cells, or human PBLs which were assayed at varied time points of incubation or in vivo using a transient mouse infection model or patients' sera, an equine model of HGE may be useful for clarifying the importance of these cytokines in HGE. Several reports have suggested that the horse serves as a more useful infection and disease model of HGE than the mouse, since A. phagocytophila isolated from patients induces clinical signs in horses similar to those of HGE and equine ehrlichiosis . However, cytokine responses in horses infected with A. phagocytophila have not been described so far. In the present study, we investigated cytokine mRNA expression by PBLs in horses following transmission of A. phagocytophila by intravenous inoculation or tick bite. MATERIALS AND METHODS : A. phagocytophila. | HZ strain A. phagocytophila, isolated from an HGE patient , was propagated in HL-60 cells as previously described . The percentage of infected cells was determined by Diff-Quik (Baxter Scientific Products, Obetz, Ohio) staining as previously described . Horse infection. | Four horses purchased from Ohio horse farms were kept in vector-proof stalls in the Equine Isolation Unit (College of Veterinary Medicine, The Ohio State University). All horses were mixed breed. Horse 1 was a 2-year-old male (380-kg body weight [BW]), horse 2 was a 5-year-old male (200-kg BW), horse 3 was a 5-year-old male (180-kg BW), and horse 4 was a 3-year-old female (170-kg BW). Prior to A. phagocytophila infection, these animals were confirmed as seronegative and PCR negative for A. phagocytophila. Horses 1 and 2 were intravenously (i.v.) inoculated with A. phagocytophila of high passage (HP; i.e., more than 50 passages in HL-60 cells) and low passage (LP; i.e., fewer than 10 passages), respectively, at 107 infected HL-60 cells (approximately 80% infected cells) in 5 ml of RPMI 1640 medium. A total of 100 and a total of 47 infected Ixodes scapularis adults were allowed to feed on horses 3 and 4, respectively. As nymphs, these ticks were acquisition fed on experimentally infected DBA/2 mice . Uninfected nymphs were kindly provided by S. R. Telford (Harvard School of Public Health, Boston, Mass.). All horses were euthanized at 24 days postinoculation (PI) or postattachment (PA). The use of horses for this study has been approved by the Ohio State University Institutional Animal Care and Use Committee. Clinical evaluation, hematology, PBL preparation, morula examination, and indirect fluorescent antibody (IFA) testing. | Fever, depression, anorexia, diarrhea, leukopenia, laminitis, and other clinical signs were monitored daily for 20 days. For complete blood cell counts, blood samples were taken from the jugular veins of horses prior to injection or tick attachment (day 0) and every 4 days until 20 days PI or PA. To prepare the PBL fraction, the blood samples, which were collected in acid citrate dextrose anticoagulant tubes, were centrifuged at 500 x g for 5 min. Erythrocytes in the pellet were lysed in sterile 0.83% NH4Cl solution for 3 min at room temperature. Cells were washed twice by centrifugation (500 x g for 5 min) in phosphate-buffered saline (137 mM NaCl, 10 mM Na2HPO4, 2.7 mM KCl, 1.8 mM KH2PO4 [pH 7.2]). Viabilities of PBLs were >98%, as assessed by the trypan blue dye exclusion test. Intracytoplasmic microcolonies (morulae) of A. phagocytophila were examined in PBLs by using Diff-Quik staining and IFA using monoclonal antibody (MAb) 5C11 . The IFA test was performed as previously described . The serum antibody titer was expressed as the reciprocal of the highest dilution of serum that showed a positive reaction. Reverse transcription (RT)-PCR for cytokine genes. | Total RNA was extracted from equine PBLs (107) by TRIzol reagent (GIBCO-BRL) and resuspended in 90 mul of diethyl pyrocarbonate-treated sterile water. As a positive control, PBLs (107) of uninfected healthy horses were stimulated with either concanavalin A (ConA; Sigma, St. Louis, Mo.) (10 mug/ml) or Escherichia coli lipopolysaccharide (LPS) (10 mug/ml) in RPMI 1640 medium at 37C for 4 h. Total RNA (2 mug) was reverse transcribed in a 30-mul reaction mixture containing 1x reaction buffer (50 mM Tris-HCl [pH 8.3], 75 mM KCl, 3 mM MgCl2), 0.5 mM each deoxynucleoside triphosphate (dNTP), 1 U of RNase inhibitor (GIBCO-BRL), 1.5 muM oligo(dT) primer, and 10 U of Moloney murine leukemia virus reverse transcriptase (GIBCO-BRL) at 42C for 1 h. The reaction was terminated by heating at 94C for 5 min, and the cDNA was used in the PCR. The cDNA (2 mul) was amplified in a 50-mul reaction mixture containing 1x PCR buffer (10 mM Tris-HCl [pH 8.3] and 50 mM KCl), 1.5 mM MgCl2, 0.2 mM each dNTP, and 0.4 muM (each) 3' and 5' primers (0.5 muM each primer for equine IL-8 mRNA). The primers used for equine IL-1beta, IL-2, IL-4, IL-6, IL-10, IL-12 p35/40, TNF-alpha, IFN-gamma, and beta-actin mRNA detection were as previously described . Primer sequences for equine IL-8 mRNA expression, designed based on the equine cDNA sequence (GenBank accession number ), were kindly provided by S. Giguere (Department of Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida). The sequences were as follows: sense, 5'-GACTTCCAAGCTGGCTGTTGC-3', and antisense, 5'-GTCCTCTTTAGAAACGCCTGC-3'. To reduce nonspecific priming, all PCRs were performed by the hot-start method: Taq DNA polymerase (2 U/reaction; GIBCO-BRL) was added after incubation of the mixture at 94C for 5 min. PCR was 25 cycles (27 cycles for equine IL-8), consisting of denaturation at 94C for 45 s, annealing at 60C (62C for IL-8) for 45 s, and extension at 72C for 2 min. The final extension was for 7 min. These conditions were within the linear range for PCR, as determined previously . PCR products (10 mul each) were electrophoresed in 1.5% agarose gel containing ethidium bromide (final concentration, 0.5 mug/ml) at 95 V for 1 h and photographed under UV illumination with a gel video system (Gel Print 2000i; Biophotonics Corporation, Ann Arbor, Mich.). DNA size markers (HaeIII fragments of phiX174 replicative-form [RF] DNA; GIBCO-BRL) providing bands from 1,353 to 72 bp were run in parallel. PCR of p44 genes of A. phagocytophila. | To detect A. phagocytophila DNA in equine PBLs, nested PCR, based on the multigene family p44 genes encoding immunodominant outer membrane protein P44s , was performed. The two primer pairs p3726 (5'-GCTAAGGAATTAGCTTATGA-3')-p4257 (5'-AGAAGATCATAACAAGCATTG-3') and p3761 (5'-CTGCTCKGCCAARACCTC-3')-p4183 (5'-CAATAGTYTTAGCTAGTAACC-3') (K = T+G; R = A+G; Y = C+T [mixed bases]) that are located in conserved regions of p44 genes (Q. Lin, N. Zhi, H.-Y. Kim, H. Horowitz, G. Wormser, and Y. Rikihisa, 101st Gen. Meet. Am. Soc. Microbiol., abstr. D-204, p. 319, 2001) were used. The first and second PCR product sizes were 531 and 422 bp, respectively. The first PCR was performed with 3 min of denaturation at 94C followed by 27 cycles consisting of 1 min each of denaturation at 94C, annealing at 55C, and extension at 72C. In the first cycle, each 50-mul PCR mixture contained 2 mug of RNA, 5 mul of 10x reaction buffer, 0.2 mM each dNTP, 1.5 mM MgCl2, 1.25 U of Taq polymerase, and 25 pmol of each primer of the primer pair p3726-p4257. The second PCR was performed the same as the first PCR, except that the annealing temperature was 52C, 1 mul of the first PCR product was used as template, and 10 pmol of each primer of the primer pair p3761-p4183 was used. RESULTS : Clinical signs and hematology. | Horse 1, which had been inoculated i.v. with HP A. phagocytophila, developed fever (body temperature > 38.9C; normal range, 37.5 to 38.6C ) during days 5 to 8 and 15 to 18 PI and thrombocytopenia (<1.1 x 1011 platelets/liter; normal value for total equine platelets, [2.2 +- 0.93] x 1011/liter ) during days 4 to 8 PI. Horse 2, inoculated i.v. with LP A. phagocytophila, developed fever only at day 8 and thrombocytopenia (1.1 x 1011 platelets/liter) and slight neutropenia (1.5 x 109 leukocytes/liter; normal value, [4.4 +- 2.0] x 109/liter ) on day 16. Horses 3 and 4 did not show any significant clinical signs. Total red blood cell counts (5.9 x 1012 to 13.9 x 1012 cells/liter; normal value, [7.5 +- 2.5] x 1012/liter ) of all four horses were within the normal range during the 20 days PI. However, at 8 days PI, variable sizes (anisocytosis) and abnormal shapes (poikilocytosis) were detected in less than 1% of the red blood cells in all 4 horses. Leukocyte levels (5.1 x 109 to 16.7 x 109/liter; normal value, [7.6 +- 3.0] x 109/liter ) and differential cell count levels (1.0 x 109 to 7.9 x 109 lymphocytes/liter [normal value, (3.0 +- 1.9) x 109/liter] and 0.1 x 109 to 0.8 x 109 monocytes/liter [normal value, (0.25 +- 0.25) x 109/liter] ) were within normal ranges in all four horses. The reactive lymphocytes (less than 1%) were observed for horse 1 (days 8, 12, 16, and 20), for horse 2 (day 16), for horse 3 (days 16 and 20), and for horse 4 (every day examined). Slight eosinophilia was observed during days 12 to 16 in horse 3 (0.4 x 109 to 0.5 x 109 eosinophils/liter; normal value, [0.15 +- 0.15] x 109/liter ). In horses 1 and 2, which were inoculated i.v., examination of PBLs by using Diff-Quik staining or IFA staining with MAb 5C11 revealed the presence of A. phagocytophila in neutrophils during days 4 to 12 PI (horse 1, 1.0% of neutrophils infected on day 4, 4.9% infected on day 8, and 2.2% infected on day 12; horse 2, 1% of neutrophils infected on day 4 and 3% infected during days 8 to 16) . Morulae were not detected in the granulocytes of horses 3 and 4 throughout 20 days. Detection of p44 genes of A. phagocytophila. | The presence of immunodominant major outer membrane protein P44s has been demonstrated in all strains of A. phagocytophila so far examined . Since P44 proteins are encoded by a multigene family, we have developed a highly sensitive p44-based nested-PCR method for HGE diagnosis (Lin et al., abstr. 101st Gen. Meet. Am. Soc. Microbiol., 2001). beta-Actin DNA served as a control for the amount of horse PBL DNA across the samples. In all four horses, p44 genes were detected during days 4 to 20 PI , indicating that ehrlichemia was established in all four horses. Although the nested PCR used was not quantitative, lack of progressive increase in band intensities following the day of infection in all four horses suggests an immune mechanism to contain the infection. IFA titers. | In the two horses inoculated i.v. with HP and LP A. phagocytophila, respectively, similar IFA immunoglobulin G (IgG) titers were observed . In horses 1 and 2, seroconversions (>1:20) were detected from day 8 and day 4 PI, respectively. The maximum IgG titers of the 2 horses were 1:320 on days 12 and 16 PI, respectively. The IgG titers in horse 3, which had been infected by attaching infected ticks, were less than 1:20 throughout 20 days, and horse 4 showed a peak IgG titer of 1:160 on day 12 PA . Equine cytokine mRNA expression. | Expression of several equine mRNAs (IL-1beta, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12 p35/40, TNF-alpha, IFN-gamma, and beta-actin) by PBLs in the four horses infected with A. phagocytophila by i.v. inoculation or by attaching infected ticks was measured following the time course of infection by RT-PCR at a linear range of 25 or 27 PCR cycles. Contamination of RNA preparations with the genomic DNA was negligible, because PCR products from "minus-reverse transcriptase" negative controls were not detected in all specimens. Constitutively expressed beta-actin mRNA served as a control for the amount of input RNA across the samples. As positive controls, ConA- and/or E. coli LPS-induced IL-1beta, TNF-alpha, IL-6, IFN-gamma, IL-8, IL-2, IL-4, IL-12 p35, IL-12 p40, and IL-10 mRNAs were clearly detectable in uninfected equine PBLs, using the primers and our RT-PCR conditions . For horse 1 only, the results for all 10 cytokine and beta-actin mRNAs examined are shown, and for the remaining horses, the results for only the cytokines whose mRNAs were expressed and IL-8 and beta-actin mRNAs are shown . Regardless of method of inoculation (i.e., intravenous inoculation or tick transmission), expression of IL-1beta and TNF-alpha mRNAs was induced in all four horses by A. phagocytophila infection . IL-8 mRNA expression was detected in horses 1, 2, and 4 but not in horse 3. IFN-gamma mRNA expression was weakly induced only in horse 2. A weak IL-10 mRNA expression level was also detected in horse 1. In horse 4, IL-12 p35 mRNA was weakly detected . Expression of IL-2, IL-4, IL-6, and IL-12 p40 mRNAs was not detectable in the PBLs of any infected horses. FIG. 1. | Microcolonies (morulae) of A. phagocytophila Microcolonies (morulae) of A. phagocytophila (black arrowheads) found in peripheral blood neutrophils from HP A. phagocytophila-infected horse 1 by using Diff-Quik staining and IFA with MAb 5C11 (insert, white arrowhead). Magnification, x625. FIG. 2. | p44 gene detection of A. phagocytophila p44 gene detection of A. phagocytophila in the PBLs. The levels of A. phagocytophila in PBLs in horses infected by i.v. inoculation and/or by attachment of infected ticks were examined by PCR using p44 gene-specific primers. Input DNA was normalized by beta-actin PCR. The PCR products were resolved on agarose gels containing ethidium bromide. M, DNA size markers (HaeIII fragments of phiX174 RF DNA); N, no template; P, positive-control DNA from cell-cultured A. phagocytophila HZ. FIG. 3. | IFA titers, showing development of IgG antibodies against A. phagocytophila IFA titers, showing development of IgG antibodies against A. phagocytophila in the four horses infected by intravenous inoculation of HP or LP A. phagocytophila or by attachment of HP A. phagocytophila-infected ticks. FIG. 4. | Induction of cytokine mRNAs in PBLs from four horses infected with A. phagocytophila Induction of cytokine mRNAs in PBLs from four horses infected with A. phagocytophila. Total RNA was extracted from equine PBLs and subjected to RT-PCR. The amount of cDNAs was normalized against beta-actin mRNA in corresponding samples. M, DNA size markers (HaeIII fragments of phiX174 RF DNA); N, no template. Uninfected horse PBLs were stimulated with ConA (C) or E. coli LPS (L) for 4 h in vitro. The results shown are representative of more than three assays. DISCUSSION : In our equine model of HGE, only IL-1beta and TNF-alpha mRNAs were consistently upregulated in all four horses, regardless of A. phagocytophila cell culture passage number or route of infection. This suggests that A. phagocytophila-induced mild fever, neutropenia, and thrombocytopenia are mediated in horses through the induction of these two major proinflammatory cytokines, which are known to cause these changes . In our previous study, A. phagocytophila or rP44 induced IL-1beta, TNF-alpha, and IL-6 mRNAs within 2 h PI and induced protein within 24 h PI in human PBLs in vitro in a dose-dependent manner . However, IL-6 mRNA expression was not detectable in our equine model. The lack of IL-6 mRNA induction by PBLs in infected horses may be due to low numbers of A. phagocytophila in the blood of horses, since IL-6 mRNA induction in vitro requires 10-fold higher levels of A. phagocytophila than TNF-alpha or IL-1beta mRNA induction does . The lack of IL-6 mRNA induction may be also related to the relatively mild nature of clinical signs of our horses. Coinfection of C3H mice with B. burgdorferi and A. phagocytophila was reported to increase the levels of A. phagocytophila and elevate IL-6 levels in the sera at 2 weeks postinfection beyond those of infection with B. burgdorferi alone . One report observed that in sera of acute-phase patients, the levels of IL-1beta, TNF-alpha, and IL-6 are not elevated . Since, unlike those of experimentally infected animals, it is difficult to preserve these rapid-turnover cytokines in the specimens of human patients , it is premature to dismiss the role these cytokines play in HGE pathogenesis. Acute-phase HGE patient sera contain significantly higher levels of IFN-gamma than convalescent-phase sera . A. phagocytophila can transiently establish infection in C3H mice without any clinical signs, and these mice develop significant IFN-gamma levels in serum during days 2 to 8 postintraperitoneal inoculation with A. phagocytophila-containing blood from infected SCID mice . Although in IFN-gamma-/- mice, initial infection levels with A. phagocytophila are greater than in control mice, the speed of clearance of A. phagocytophila does not differ . In our study, IFN-gamma mRNA expression was seen only in one horse. This finding may be related to the lack of spontaneous clearance of A. phagocytophila from the blood of all horses at even 20 days PI or PA. The role of IFN-gamma in clearance of A. phagocytophila remains to be studied. In the present study, IL-8 mRNA expression was upregulated in 3 horses; however, horse 3 had the weakest cytokine response overall and did not show IL-8 mRNA expression. Recently, Akkoyunlu et al. proposed that IL-8 generated by neutrophils in response to A. phagocytophila facilitates the infection. Klein et al. reported that in dimethyl sulfoxide-treated infected HL-60 cells or human bone marrow cells, IL-8 and other chemokine levels, but not that of IL-1, IL-6, or TNF-alpha in the culture medium, were significantly elevated at 48 h postinfection . Since findings, including those of our present horse study, indicate that IL-8 mRNA is not upregulated at early stages of infection in vitro and in vivo, IL-8 does not seem to be critical for the initial establishment of infection. Overall clinical signs for our horses, which were inoculated with a strain from an HGE patient in the state of New York, were milder than those reported with studies using the BDS or Webster strain from Wisconsin but rather similar to those using inoculation with a New York strain as reported by Chang et al. . In the study of Pusterla et al. , at day 16 after i.v. inoculation and at day 18 after inoculation by tick attachment, A. phagocytophila Webster and BDS strains became undetectable by real-time quantitative PCR assays, based on the presence of the 16S rRNA gene in the blood of horses. However, A. phagocytophila HZ strain p44 genes were detected at day 20 of infection in four horses in the present study, suggesting that A. phagocytophila establishes subclinical persistent infection in horses, as suggested by Chang et al. . A pioneering phylogenic study, comparing different A. phagocytophila strains from patients and animals in the United States and Europe by using the ank gene as a marker, showed that Wisconsin strains are closer to Ehrlichia equi (A. phagocytophila, horse isolate, 99.18 to 99.24% amino acid identity) than New York strains (97.90 to 97.92% amino acid identity) . Thus, it is possible that Wisconsin strains are more pathogenic to horses but cleared more rapidly than New York strains. Disease severity in horses inoculated i.v. with A. phagocytophila strain Webster was reported to be dependent on A. phagocytophila passage numbers in culture . This finding appears to be among those of studies that are not evident unless larger number of animals are tested, since we did not observe clear differences in clinical manifestations between horses infected with HP and LP A. phagocytophila HZ strains. Using a real-time quantitative PCR assay, Pusterla et al. reported 55-fold-lower initial ehrlichial load in PBLs from tick-infected horses than in those from horses inoculated i.v. . In the same study, however, higher ehrlichial levels in the blood of horses infected by tick attachment were observed at 7 days PI. According to the findings based on the use of PCR with p44 genes, our study appears to support their observation with respect to the lower rate of increase of ehrlichial growth with tick transmission than with intravenous inoculation. Pusterla et al. did not see significant difference in antibody titer at day 30 PI between intravenous and tick transmission. In the present study, antibody responses were lower by tick attachment than by intravenous inoculation. More studies are needed to learn whether tick inoculation modulates host immune response to facilitate the infection. Backmatter: PMID- 12186657 TI - A ~35 kDa polypeptide from insect cells binds to yeast ACS like elements in the presence of ATP AB - Abstract | Background | The S. cerevisiae origin recognition complex binds to the ARS consensus sequence in an ATP dependent fashion. Recently, the yeast Cdc6 has been reported to have DNA binding activity. Conservation of replication proteins among different species strongly supports their functional similarity. Here we report the results of an investigation into the DNA binding activity of human Cdc6 protein. Cdc6 was expressed and purified from baculovirus infected Sf9 (Spodoptera frugiperda) insect cells as GST fusion protein (GST-Cdc6) and its DNA binding activity was tested. Results | Partially purified fractions containing GSTCdc6 or GST showed an ACS binding activity in an ATP dependent manner. However, further purification revealed the presence of a putative 35 kDa insect cell protein (p35) which was found responsible for the DNA binding activity. A close match to the 9/11 bases of the ARS consensus sequence was sufficient for p35 binding activity. A DNA fragment from the human c-myc origin region containing yeast ACS like elements also showed p35 binding activity. Conclusions | We have identified a Spodoptera frugiperda protein with ATP dependent DNA binding activity to ACS like elements. ACS like elements have been reported to be essential for ORC binding and replication initiation in yeast but their role in higher eukaryotes still remains elusive. Like the ARS consensus sequence elements of yeast, ACS like elements found in c-myc and lamin beta 2 origin regions may play similar roles in replication and indicate a conserved role for this DNA motif among eukaryotes. Keywords: Background : Proteins recognizing specific DNA sequences play an important role in the regulation of gene expression and in DNA replication. Nearly all eukaryotic genes transcribed by RNA Polymerase II for instance, contain the conserved TATA box which is present upstream of the transcription start site. The TATA-box binding protein, a ~30 kDa component of the TFIID complex binds specifically to the heptanucleotide A and T residues and forms the core of the transcription initiation complex. Additionally, many specific transcription factors bind to the upstream promoter in a sequence specific manner and regulate gene expression. For example in Drosophila, a heat-shock transcription factor (HSTF) can bind to consensus heat-shock response elements and regulates expression of heat/stress inducible genes. Promoter activity in later branching eukaryotes is greatly modulated by enhancer and repressor sequences which have no activity of their own but are the targets of DNA biding proteins or protein complexes which can remodel the promoter chromatin to make it more or less accessible to RNA polymerase. Biochemical assays have shown that the action of ATP-dependent chromatin remodelling activities increase the accessibility of DNA within chromatin templates. S. cerevisiae SWI/SNF , Ino80 complex , Drosophila NURF are examples of some high molecular weight chromatin remodelling factors which can facilitate transcription by binding to chromatinized DNA templates. However, none of the above chromatin remodelling factors binds to specific DNA sequences. Unlike transcription, the role of sequence specific DNA binding proteins in eukaryotic DNA replication is not well characterized. In higher eukaryotes finding of specific DNA sequences essential for DNA replication has been elusive so far. In yeast Saccharomyces cerevisiae, a six-protein origin recognition complex binds to ARS consensus sequence (ACS) in a sequence specific manner . Individual ORC subunits have not been demonstrated to show DNA binding activity in vitro. Recently, in an in vitro study ScCdc6 has been shown to bind double stranded DNA . The minimal requirement for the binding of Cdc6 to DNA has been mapped within its N-terminal 47-amino acid sequence. Saccharomyces pombe ORC4 subunit has been reported to contain DNA binding activity by using its N-terminal AT hook region . Neither ScCdc6 nor SpORC4 showed any sequence specific DNA binding activity. Recombinant six protein Drosophila ORC (DmORC) binds to ACE region of the Drosophila chorion gene . In vivo, DmORC co-localized with the amplified chorion gene locus. In Xenopus, biochemical analysis of replication and cell cycle events using egg extracts has helped to understand the mechanism of eukaryotic DNA replication . However, two dimensional gel electrophoresis analysis of the rDNA locus showed that replication initiated at all sites tested . All six human homologs of yeast and Drosophila ORC subunits have been cloned and characterized . Other replication proteins like Cdc6, Cdt1, MCMs, Cdc45 that are essential for initiation of DNA replication have also been reported . Conservation of replication factors among higher eukaryotes suggests that functionally they may play similar roles. In an attempt to identify DNA binding activity of human Cdc6, it was expressed and purified as a GST-Cdc6 fusion protein from baculovirus infected Sf9 insect cells. Partially purified fractions (reduced glutathione eluate) containing GSTCdc6 or GST showed an ACS binding activity in an ATP dependent manner. The GSTCdc6 protein fraction contained both the GSTCdc6 and a 35 KDa S. frugiperda protein. The DNA binding activity was confined to a 35 kDa polypeptide. It was latter found that the p35 has an intrinsic affinity to GST. This polypeptide bound to yeast ACS like elements in the presence of ATP. 9/11 matches to ARS consensus sequence were found to be essential for this DNA binding activity both by gel shift assay as well as by in vitro foot printing assay. A DNA fragment containing 9/11 matches from human c-myc replication origin region also showed p35 binding activity suggesting that this polypeptide has intrinsic DNA binding activity. The implications of this DNA binding activity are discussed here. Results : Partially purified protein fractions containing GSTCdc6 or GST contain an ACS binding activity | We infected Sf9 insect cells with the baculovirus expressing GSTCdc6. Cells were harvested 48 hours post infection and the proteins were extracted according to the procedures described in materials and methods. The GSTCdc6 protein was partially purified by pull down on glutathione beads . The partially purified protein was used in DNA binding assays with a 240 bp DNA fragment containing all three conserved boxes (A, B1 and B2) of the ARS consensus sequences (Fig. &). As a control, we used GST alone, which was purified using the same strategy used for GSTCdc6 purification. A DNA protein complex was formed in both the cases as evidenced by the retarded mobility of the free 32P phosphate labelled probe . The specificity of the DNA binding was examined in a competition reaction by increasing the amount of unlabeled DNA fragment containing ACS like elements. It was determined that the DNA-Protein complex could be competed efficiently by increasing amount of unlabelled ACS like DNA (20x, 50x, 100x and 200x respectively) . A ~350 bp DNA fragment from pBlueScript KS+ (HinfI digested and subsequently gel purified) of similar base composition did not compete with the complex formation in the Electrophoretic Mobility Shift Assay when added at a similar concentration indicating a degree of specificity in the DNA-Protein complex formation . Figure 1 | Purification of GST-Cdc6 and its DNA binding activity. Purification of GST-Cdc6 and its DNA binding activity. (A) SDS-PAGE of GST-Cdc6 purified by glutathione agarose beads purification technique from Sf9 insect cells infected with baculovirus expressing GST-Cdc6. The polypeptide was visualised by commassie stain. Molecular mass markers are indicated on the right. (B) Gel retardation assays were performed in the presence of reduced glutathione eluted GST or GST-Cdc6 fusion protein. S. cerevisiae ARS1 fragment (240 bp) was end labelled using gamma-32P ATP and incubated with GST or GST-Cdc6 as described in the materials and methods. Both GST and GST-Cdc6 showed strong band shifts. Figure 2 | Specificity of the ARS1 binding Specificity of the ARS1 binding. Gel retardation assays were performed in the absence (-) or presence (+) of GST-Cdc6 using radiolabeled ARS1 probe as described in Fig. . The binding reaction was competed either with cold ARS1 fragment (lanes 3 --6) or a 350 bp fragment from the plasmid BlueScript KS(+) digested with HinfI followed by gel purification (lanes 7 --10). Free probe and the band shifts are indicated by arrow. An unidentified ~35 kDa protein from baculovirus infected insect cells is responsible for DNA binding activity | Partially purified fractions containing GSTCdc6 or GST showed DNA binding activity. To further fractionate the proteins present in the partially purified GSTCdc6 fraction the glutathione column was washed with 200 mM and 300 mM sodium chloride prior to GSTCdc6 elution with reduced glutathione. Western blot analysis using anti GST antibodies revealed that neither 200 mM nor 300 mM fractions contained GSTCdc6 . GSTCdc6 protein was present only in the proteins eluted by reduced glutathione. Proteins released by different salt fractionation were separated on an SDS-PAGE and visualized by silver stain . A prominent band of molecular mass ~35 kDa was visualised both in the 200 mM and 300 mM salt eluate. Figure 3 | Identification of p35 Identification of p35. (A) Western blot analysis of salt eluate and reduced GSH eluate using anti GST antibody. GST-Cdc6 beads were washed using wash buffer containing NaCl (200 mM and 300 mM respectively). Beads following salt wash were treated with reduced glutathione and released proteins were run in SDS-PAGE along with salt eluate followed by western blot analysis. GST-Cdc6 was found only in GSH eluate. (B) Silver staining of salt eluate: Salt eluates were run in SDS-PAGE and proteins were visualised by silver stain. A distinct band of p35 was found both in 200 mM and 300 mM salt eluate. Molecular mass markers are shown on the right. (C) DNA binding activity of p35. Gel shift assays were performed using salt wash (300 mM NaCl eluate) and GSH eluate following 300 mM salt wash of GST-Cdc6 beads. ARS 240 bp DNA fragment was used as probe. 300 mM salt eluate gave a strong band shift whereas GSH eluate failed to do so though it contained GST-Cdc6 as shown in figure . To check whether the DNA binding activity was due to the presence of ~35 kDa band or GSTCdc6 itself, gel shift assays were performed either using 50 ng ~35 kDa protein obtained from the salt wash or GSTCdc6 by itself . Salt eluate gave a strong band shift which was identical with the band shift found with GSTCdc6 found in Fig. . GSTCdc6 eluted from the beads following high salt wash failed to give any band shift suggesting that the ~35 kDa polypeptide was responsible for the DNA binding activity. To further test whether the presence of p35 is absolutely required for DNA binding activity, the 300 mM salt eluate was dialysed against low salt buffer H/0.15 and then passed through Superose 12 gel filtration column. Each fraction was checked for DNA binding activity by gel retardation assay using a 32P labelled DNA fragment containing ACS elements. DNA binding activity was found only in high molecular weight fraction (~670 kDa) . Proteins present in the gel filtration fractions in the high molecular weight range (fractions 13 --17) were separated by SDS-PAGE followed by silver stain. Surprisingly, p35 was found to be present only in the fraction 15 which contains the DNA binding activity . The presence of p35 in the same fraction containing the DNA binding activity strongly suggests that p35 is responsible for the binding activity. Figure 4 | Purification of p35 by gel filtration chromatography Purification of p35 by gel filtration chromatography. (A) Gel retardation assay. 300 mM salt eluate (as shown in fig. ) was dialysed against Buffer H/0.15 and passed through Superose-12 Gel filtration column. Fractions 13 --25 were tested for ARS 240bp DNA binding activity. Only fraction 15 showed a band shift compared to the free probe. Molecular mass markers are shown on the top. (B) Silver stain of gel filtration fractions. 0.25 ml of each gel filtration fractions (fraction 13 --17) were precipitated using trichloro acetic acid (TCA). The samples were run in SDS-PAGE and proteins were visualised by silver stain. p35 was found only in the fraction 15. DNA binding activity is ATP dependent | One of the hallmarks of yeast ORC binding to yeast ARS consensus sequences is its ATP dependence . We were interested to see whether the DNA binding activity of the ~35 kDa protein is ATP dependent or not. Gel shift assays were performed either in the absence or in the presence of increasing amount of ATP in the reaction mixture . In the absence of ATP, a very weak binding was observed whereas with increasing amount of ATP strong binding was detected. There is a threshold of ATP concentration (6 mM and onwards) which stimulated the binding remarkably. A nonhydrolysable ATP analog, ATPgammaS was used in the binding reaction to see whether ATP hydrolysis is required for this binding. With increasing amount of ATPgammaS, the band shift was completely inhibited suggesting that ATP hydrolysis is required for this DNA binding activity. This was confirmed by adding ATP back in the reaction mixture when ATPgammaS was already present in the reaction. Under these reaction conditions, increasing amount of ATP again stimulated the DNA binding activity even in the presence of ATPgammaS suggesting that ATPgammaS can be competed with ATP and it is the ATP hydrolysis which is essential for this DNA binding activity (Fig. , lanes 13 and 14). Figure 5 | ATP hydrolysis is required for the DNA binding activity of p35 ATP hydrolysis is required for the DNA binding activity of p35. Gel shift assays were performed using 32P labelled ARS fragment either in the presence of ATP (lanes 3 --7) or ATPgammaS (lanes 8 --12) or both (lanes 13 --14). A strong band shift was observed in the presence of ATP whereas ATPgammaS completely inhibited the binding. The inhibition can be reversed by adding back ATP in the reaction mixture. p35 binds to A, B1 and B2 boxes of ARS consensus sequences as revealed by footprinting assay | After establishing the fact that a protein of approximately 35 kDa binds to DNA fragment containing ARS consensus sequences, the exact site of binding of the protein on the DNA was mapped by copper-phenanthroline footprinting assay. Unlike DNaseI, 1,10-phenanthroline-cuprous complex is a small chemical probe which can demark the boundaries of the protected region clearly. The protein was bound to a 5' 32P labelled 240 bp DNA fragment containing yeast ARS consensus sequence and separated from the free DNA by gel shift assay. The gel was then treated with copper-phenanthroline reagent as described in materials and methods. The bound and unbound DNA was purified and allowed to run in a sequencing gel . On the T-rich strand (bottom strand) two protected regions were observed. The first region entirely covered the A box of the ARS consensus sequence and the second region covered the overlapping regions of B1 and B2. Therefore the 35 kDa protein has a strong affinity to bind to A, B1 and B2 boxes of the ARS consensus sequences. The long stretch of protection could be due to multimeric form of p35 or could be due to the formation of a higher order nucleoprotein complex. Figure 6 | Copper-Phenanthroline footprint analysis of ARS DNA fragment bound to p35 Copper-Phenanthroline footprint analysis of ARS DNA fragment bound to p35. ARS DNA fragment was 5' end labelled at the T rich strand and incubated with p35 (300 mM salt eluate) and finally subjected to gel retardation assay. The gel was then treated with copper-phenanthroline mixture and DNA from bound (B) and unbound (U) fractions were extracted from gel, denatured and loaded in a sequencing gel. A+G ladder was also loaded to figure out the position of footprint. The relative positions of A, B1 and B2 boxes were indicated. The protected regions are shown by solid lines and asteric (*). The hypersensitive site is denoted by arrowhead. Figure 7 | At least one ACS like element is required for p35 binding At least one ACS like element is required for p35 binding. (A) ARS consensus sequence. (B) 240 bp yeast S. cerevisiae ARS1 fragment containing several functional elements required for ARS activity (Boxes A, B1, B2 and B3 respectively). Several oligos spanning the entire ARS1 as indicated in the figure were designed and subsequently cloned into p21N fragment. Double stranded oligo 'a' (ARSa) does not contain any ARS like element. Oligo 'b' (ARSb) contains one 11/11 match and one 9/11 match to ACS element. Oligo 'c' (ARSc) and oligo 'd' (ARSd) contain one 9/11 match to ACS element each. In the fragment dmut, 'A's and 'T's within the core ARS were mutated to 'G's and 'C's. (C) Nucleotide sequences of ARSd and ARSdmut. (D) Gel shift assay was performed either using ARSwt (lanes 1,8), or p21N (lanes 2,3 and 9) or the constructs derived from p21N by cloning fragments a-d and dmut (lanes 4 --7, 10,11) as described in (B). p35 does not bind to p21N or p21N + ARSa or p21N + ARSdmut. All the other DNA fragments showed DNA binding activity. p35 binds to A, B1 and B2 boxes and mutation in these boxes abolish DNA binding activity | Previous studies of the structure of ARS1 in both plasmid and chromosome contexts have shown that it contains one essential DNA element, A, that includes a perfect (11/11) match to the ARS consensus sequence (ACS, Fig. ), and three additional elements, B1, B2, and B3, with 9 out of 11 bases match to the ACS that are also important for ARS function . We were interested to see whether A, B1 and B2 boxes were sufficient to allow p35 binding activity. We used the p21N protein (N terminal 150 bases of coding region of p21) as a control. This DNA when incubated with ~35 kDa protein does not form a DNA protein complex . The DNA fragment was divided into four subfragments (a, b, c and d; Fig. ) and subcloned in the middle of p21N fragment. Fragment 'a' does not contain any ACS sequence whereas fragments b, c and d contain at least one ACS like element. All the ACS containing sub fragments (b, c, d) showed a mobility shift , which suggests that at least one ACS like element (either 11/11 match or 9/11 match to ARS consensus sequence) is essential and sufficient for p35 binding activity. This was further confirmed by using a subfragment d mutated at the ACS motif . Subfragment d as shown if Fig lane 10 can bind strongly to the 35 kDa protein, however mutation altering the A and Ts of the core ACS sequence to G and Cs to generate the dmut oligonucleotide results in the abolition of its ability to bind p35. Therefore the ACS sequence is essential for the p35-DNA protein complex formation. 35 kDa polypeptide binds to ARS consensus sequence found in c-myc origin of replication | In S. cerevisiae, ARS elements have been implicated to be important both for ORC binding and origin function . We looked for the availability of such sequences in known human origins of replication like c-myc, lamin beta 2 and Dnmt1. An origin of replication was mapped previously by nascent strand abundance analysis within 2.0 kb zone immediately upstream of c-myc gene . Detailed analysis of the 2.0 kb upstream sequences revealed the presence of two ACS like elements separated by 148 nucleotides (AAAAGATAAAG and AAAAGAAAAAA). A 300 bp DNA fragment containing both the ACS elements was amplified by polymerase chain reaction and the product was used subsequently for p35 binding studies. A strong band shift was observed which could be competed out using a 50 bp long double stranded oligo containing two ACS like elements (oligo 'b', Fig. ). Increasing amount of ~65 bp long unrelated (non specific) double stranded oligo did not have any effect on this binding activity suggesting that binding of p35 to c-myc origin region is specific. Interestingly, analysis of Lamin beta 2 origin of replication region also revealed the presence of two ACS (9/11 match) like elements . Further studies are required to find out whether p35 also binds to lamin beta 2 origin region. Figure 8 | p35 binds to DNA fragment containing ACS like elements from c-myc origin region p35 binds to DNA fragment containing ACS like elements from c-myc origin region. A 300 bp DNA fragment containing two ACS like elements (10/11 matches) was PCR amplified using primers specific for the zone of replication from c-myc locus. PCR product was end labelled using gamma-32P ATP and used for gel shift assays. Competition reactions were performed by using different quantities (50x, 100x and 200x) of 50 bp long double stranded oligos (oligo 'b', Fig. ) in the reaction mixtures corresponding to lanes 3 --5. Lanes 6 --8 contained 65 bp double stranded DNA fragment (plasmid BlueScript KS(+) digested with HinfI followed by gel purification). p35-ACS interaction is sensitive to high salt, temperature and EDTA | The stability of p35-ACS DNA complex was further tested either by changing NaCl concentration in the reaction mixture, or by shifting reaction temperature or by adding EDTA. The DNA binding activity was found to be sensitive to NaCl concentration . Strong band shift was obtained up to 0.2 M NaCl. NaCl concentration at 0.5 M and above completely inhibited the binding activity. Higher temperature also showed a drastic effect on the binding activity. Normal DNA binding activity was observed up to 42C. Temperature higher than 42C completely abolished the binding activity suggesting that the off rate of p35 from DNA is much faster at higher temperature. Finally inclusion of EDTA in the reaction mixture inhibited the binding reaction suggesting that the divalent cations are essential for this binding activity. Figure 9 | Effect of salt concentration, temperature and EDTA on p35 gel shift assay Effect of salt concentration, temperature and EDTA on p35 gel shift assay. Gel retardation assay was performed either using no extra salt or different concentration (lanes 1 --5) of NaCl. Band shift was completely inhibited at 0.5 M NaCl concentration and above. Effect of temperature on binding reaction was tested by incubating the reaction mixture (lanes 6 --9) at different temperature. Binding reaction was completely inhibited at 55C and above. All the other reactions were carried out at room temperature. Effect of bivalent cations were tested by incubating the reaction mixture in the presence of 50 mM and 100 mM EDTA (lanes 10 --11). Presence of EDTA completely inhibited the binding reaction. Discussion : Few proteins have been reported in the literature, which are capable of binding to DNA in a sequence specific ATP dependent manner. Although transcription factors bind to specific DNA sequences, the binding activity is not dependent on ATP hydrolysis. In contrast, chromatin remodelling factors like SWI/SNF, ISW1, BRG1 facilitate transcription from chromatinised templates in the presence of ATP . However, these factors do not bind to specific DNA sequences. In eukaryotic DNA replication, sequence specific ATP dependent DNA binding activity has been demonstrated in yeast S. cerevisiae where ORC, a six polypeptide complex binds to yeast ARS consensus sequence in an ATP dependent manner . The binding sites for other ORCs are not very clear at present. DmORC binds the critical elements of well-characterized, chromosome III amplification domain (ACE3 and ori-beta, though the precise sequence recognized by DmORC within ACE3 and ori-beta have not been identified . Studies of both ScORC and DmORC indicate the ATP binding by Orc1p is required for DNA binding. However ATP hydrolysis is not required for DNA binding for both the cases suggesting that ATP hydrolysis may be required for further downstream processes. Chromatin immunoprecipitation (ChIP) studies have demonstrated the association of SpORC with S. pombe origins and human ORC with the EBV OriP . Whether ORC binds to these sequences directly or indirectly with the help of other proteins are subject to in vitro DNA binding assays using purified ORC proteins. In this study, we report a ~35 kDa protein from the baculovirus infected Sf9 insect cells that binds to yeast ACS sequences in an ATP dependent fashion. p35 was purified as high salt (300 mM NaCl) eluate from the GST-Cdc6 beads. GST-Cdc6 eluted from the beads following high salt wash failed to show any DNA binding activity whereas high salt eluate containing only p35 showed strong DNA binding activity suggesting that p35 not Cdc6 is responsible for the binding activity. This experiment was repeated several times and always the protein preparations containing p35 showed DNA binding activity. p35 has an intrinsic affinity to GST moieties. Sf9 insect cells were infected with baculovirus expressing GST alone. The cell lysate was allowed to bind to GST beads. High salt eluate (300 mM NaCl) from GST beads was tested for DNA binding activity. Surprisingly, we observed a very similar band shift as obtained previously using high salt eluate from GSTCdc6 (data not shown). Further, high salt eluate from GSTORC2 and GSTORC4 (GST fusion protein containing human origin recognition complex subunit 1 and 2 respectively) also showed DNA binding activity (data not shown). Therefore, it can be concluded that p35 interacts with GST and high salt concentration is required to disrupt this interaction. The nature and the specificity of the interaction between GST and p35 are not clear at this moment. It is important to note that majority of the p35 bound to GST or GST fusion proteins are released mostly at high salt concentration (300 mM) allowing us to get rid of most of the impurities by stringent washing of the GST beads with buffer containing 250 mM NaCl. A weak DNA binding activity was found till 4 mM ATP concentration . A strong stimulation was obtained at 6 mM ATP concentration. It is possible that p35 is purified as ATP bound form but an associated weak ATPase activity does not allow it to give a strong binding activity. It is suggested that 6 mM ATP concentration may be sufficient to overcome this inhibitory effect. ATP hydrolysis is essential for DNA binding activity of p35 since ATPgamma S, a nonhydrolysable analog of ATP completely inhibited the binding activity . It is possible that ATP is required for strand opening which needs to be further explored. An insect cell factor, polyhedrin promoter binding protein has been reported previously, capable of binding to AT rich DNA sequence . However, the reported DNA binding activity was unusual in a sense that the activity was heat and salt concentration resistant. 100 mM EDTA concentration did not affect the DNA binding activity. The DNA binding activity reported in this study was found to be temperature, EDTA and salt concentration sensitive suggesting that this polypeptide is completely different from the polyhedrin promoter binding protein. p35 showed a strong DNA binding affinity towards ACS like elements. This was confirmed first by using unlabeled specific competitor DNA which completely abolished the binding of 32P labelled probe. Competition using unrelated DNA did not affect the DNA binding activity. We took p21N, which normally does not bind to p35, to further test the binding specificity. Introduction of a single ACS like element (9/11 match) in p21N (ARSc and ARSd) showed a strong band shift suggesting that only one ACS like element is sufficient for p35 binding. This was further confirmed by making mutations in the ARSd fragment. The resulting ARSdmut did not show any p35 binding activity suggesting that the ACS like element itself but not the adjacent sequences are essential for the p35 binding activity. However, a systematic mutational analysis of ACS like elements will be required to explore the exact binding specificity of p35 towards ACS like elements. ACS elements are normally AT rich. However, p35 did not bind to random AT rich sequences. p35 did not bind to p21N + ARSa which contains ARSa oligo (78% AT rich) with no ACS like element. Surprisingly, p35 showed strong binding activity in the presence of the oligo ARSb (64% AT rich), ARSc (66% AT rich) and ARSd (74% AT rich) respectively . ARSb, c and d contain at least one ACS like element . Finally the copper phenanthroline footprint analysis confirmed that p35 binds to A and B1-B2 boxes of the ARS1 DNA fragment. At this moment, the function of p35 is not very clear. It may play major role(s) in the transcription of certain insect cell genes. It may as well be responsible for DNA replication. The fact that it binds to yeast ACS and to a DNA fragment from c-myc origin of replication region containing yeast ACS like elements in an ATP dependent manner further strengthen the hypothesis. S. cerevisiae origin recognition complex (ORC) binds to ARS consensus sequences in an ATP dependent fashion and this binding is essential for both origin function and activity. It is interesting to note that a huge six protein origin recognition complex binds to yeast ARS1 whereas p35, a small protein is showing same kind of protection as evidenced by foot print analysis . We believe that p35 forms an oligomeric structure or it maintains a multimeric form which may explain the wide footprint over the ARS1 fragment. The presence of p35 in high molecular weight fraction (~670 kDa) following superose 12 gel filtration chromatography strengthen this hypothesis and clearly suggests that p35 forms an oligomeric structure. ARS consensus sequence has been found near the vicinity of c-myc, lamin beta 2 and Dnmt1 replication origin . Therefore, Identification and characterization of this protein from insect cells and finding its human counterpart will greatly help in elucidating its possible function in DNA replication. Conclusions : The data presented here leads to the identification and characterization of a polypeptide from insect cells with ATP dependent DNA binding activity. This is an important and unique observation. In S. cerevisiae, ACS elements have been reported to be essential for ORC binding and replication initiation. Yeast ACS like elements found in c-myc and lamin beta 2 origin region may play similar roles in replication initiation. However, it is also possible that p35 is a transcription factor which may facilitate transcription of some insect cell genes. Further characterization of p35 from insect cells and finding its human homolog will be very helpful to dissect its functional role in replication and/or transcription. Materials and Methods : Plasmid construction | Cloning of human Cdc6 cDNA is described elsewhere. Coding sequence of human Cdc6 was cloned in pFastBac-GST vector (Life Technologies, Inc.) to express GST fusion protein. A 240 bp DNA fragment from S. cerevisiae ARS1 chromosomal DNA replication origin containing all the key elements including boxes A, B1, B2 and B3 was subcloned in pBlueScript KS(+) between EcoRI and HindIII. Subsequently EcoRI-HindIII fragment was end labelled using gamma32P ATP and used either for gel shift assay or copper-phenanthroline foot print assay. p21N, (N terminal ~150 bp of p21) was previously cloned in pBlueScript KS(+) between EcoRI and HindIII sites. Complementary oligos corresponding to ARSa, b, c, d and dmut were synthesized (~50 bases in length) and subsequently annealed to get double stranded oligos. p21N/KS(+) construct contains only one StuI site which is present within the p21N insert. All the annealed double stranded oligos were cloned into the StuI site using blunt end ligation. The sequences of ARSa, b, c and d are followed: ARSa: ttagtttttcggtttactaaatcgtaatagaaatgtagaacaataaaatgt ARSb: tctaaaatacaaatctagaaaatacgaacgaaaagttttccggacgtccgt ARSc: cggacgtccgttcacgtgtttgttatgaatttatttatgatgagtcattat ARSd: tgagtcattattggataaagaatcgtaaaaactgctttaaacgataaaa Plasmid containing 2.5 kb DNA fragment from c-myc origin region was a kind gift from Michael Leffak, Wright State University Ohio. Forward and reverse PCR primers (5'-gaagaaaaactctcttttc-3' and 5'-atttgctgggttgaaaaatg-3' respectively) were used to amplify 300 bp region containing two ACS like elements. Expression of GSTCdc6 and GST in insect cells and purification | Baculoviruses were produced from the recombinant pFB-GST plasmid using Bac-to-Bac expression system (Life Technologies Inc.). Sf9 cells (Invitrogen) were infected with the pFB-GSTCdc6 or pFBGSTbaculovirus according to the manufacturers' recommendations. Cells were harvested 48 hours post-infection. The cell pellet was washed once in cold phosphate-buffered saline and subsequently resuspended in hypotonic lysis buffer (10 mM Tris.Cl, pH 7.9, 10 mM KCl, 1.5 mM MgCl2, 1 mM phenylmethylsulfonyl fluoride, 2 mug/ml pepstatin, 2 mug/ml leupeptin, 5 mug/ml aprotinin, 1 mM dithiothreitol). The cell suspension was homogenized in a Dounce homogenizer using a B-type pestle followed by centrifugation at 3000 rpm for 7 min. The pellet containing the nuclei was lysed in buffer H/0.15 (50 mM HEPES/KOH, pH 7.5, 150 mM KCl, 0.02% Nonidet P-40, 5 mM magnesium acetate, 1 mM EDTA, 1 mM EGTA, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, 2 mug/ml pepstatin, 2 mug/ml leupeptin, 5 mug/ml aprotinin, 1 mM dithiothreitol). The resulting suspension was subjected to ammonium sulphate precipitation (starting with 10% followed by 30% and finally 50%). The pellet after the 50% ammonium sulphate cut was resuspended in buffer H/0.0 (no salt) and then dialyzed overnight against buffer H/0.15. The dialyzed sample was then bound to GST beads (Sigma) and washed three times with buffer H*/0.15 (containing 150 mM NaCl instead of 150 mM KCl). Proteins were eluted using reduced glutathione elution buffer (50 mM Tris.Cl, pH 8.0, 20 mM reduced glutathione, 0.01% Nonidet P-40, 100 mM NaCl). Immunoblotting and Silver stain | Anti-GST polyclonal antibodies were purchased from Santa Cruz Biotechnologies. Western blotting technique was carried out using standard protocol. The silver stain protocol is described elsewhere . Gel retardation assay | Gel retardation assay was performed with slight modification of the protocol used by Mukhopadhyay et al . The DNA fragments to be used for gel retardation assay were endlabeled with gamma32P ATP. The binding reactions were performed in 20 mul of T buffer (50 mM Tris-HCl, pH 7.4, 50 mM KCl, 50 mM NaCl, 10 mM MgCl2, 0.1 mM EDTA, 0.5 mM DTT, 30 mug/ml BSA) supplemented with 5 mM ATP and 6% (v/v) glycerol for most of the reactions. The mixture was incubated at 37C for 10 min and loaded directly on a 5% polyacrylamide gel in TBE buffer (89 mM Tris base, 89 mM boric acid, 2.5 mM EDTA, pH 8.3). The gel was run at 150 V for 2 hours, dried and autoradiographed. Copper-phenanthroline footprint assay | The Copper-phenanthroline footprint assay was performed essentially by using the protocol described by Kuwabara et al . The DNA-protein gel is run in the absence of free radical scavengers as described in gel retardation assay. The gel is placed in 200 ml of 50 mM Tris-HCl, pH 8.0. The gel is further incubated for 10 minutes in a solution containing equal volume of solution A (40 mM 1,10 Phenanthroline monohydrate in 100% EtOH and 9 mM Cupric sulphate mixed with equal volume followed by 1:10 dilution with water) and solution B (0.5% 3-Merceptopropionic acid in water). Finally the gel is soaked in solution C (28 mM 2,9 Dimethyl-1,10 Phenanthroline in 100% EtOH) for 2 min. The gel is washed twice in deionised water. After the pre-treatment of the gel, it is autoradiographed and the retarded band is cut from the gel and placed in an eppendorf tube. The DNA is eluted from the gel slice, denatured and loaded in a sequencing gel. The sequencing gel is fixed, dried and the bands were visualised by autoradiography. Authors' contributions : SKD designed, performed and co-ordinated the whole study. NM participated in insect cell culture and helped in making and amplification of baculovirus expressing GST-Cdc6 and GST. RKS helped in making the figures and drafting the manuscript. GM helped in analysing the data and critically reviewed the manuscript. All authors read and approved the final manuscript. Backmatter: PMID- 12221110 TI - Role of Fission Yeast Tup1-like Repressors and Prr1 Transcription Factor in Response to Salt Stress AB - In Schizosaccharomyces pombe, the Sty1 mitogen-activated protein kinase and the Atf1 transcription factor control transcriptional induction in response to elevated salt concentrations. Herein, we demonstrate that two repressors, Tup11 and Tup12, and the Prr1 transcription factor also function in the response to salt shock. We find that deletion of both tup genes together results in hypersensitivity to elevated cation concentrations (K+ and Ca2+) and we identify cta3+, which encodes an intracellular cation transporter, as a novel stress gene whose expression is positively controlled by the Sty1 pathway and negatively regulated by Tup repressors. The expression of cta3+ is maintained at low levels by the Tup repressors, and relief from repression requires the Sty1, Atf1, and Prr1. Prr1 is also required for KCl-mediated induction of several other Sty1-dependent genes such as gpx1+ and ctt1+. Surprisingly, the KCl-mediated induction of cta3+ expression occurs independently of Sty1 in a tup11Delta tup12Delta mutant and so the Tup repressors link induction to the Sty1 pathway. We also report that in contrast to a number of other Sty1- and Atf1-dependent genes, the expression of cta3+ is induced only by high salt concentrations. However, in the absence of the Tup repressors this specificity is lost and a range of stresses induces cta3+ expression. Keywords: INTRODUCTION : Exposure of cells to environmental stress triggers a rapid increase in the transcription of genes whose products have protective functions . Key to this response are stress-activated protein kinase (SAPK) pathways that transmit the signal from stress sensors to the transcription factors that regulate gene expression. These pathways are evolutionarily conserved, and homologs of the mammalian SAP kinases, p38/RK/CSBP , are present in both Saccharomyces cerevisiae (Hog1) and Schizosaccharomyces pombe (Sty1/Spc1) . The Hog1 pathway in S. cerevisiae is activated essentially by hyperosmolarity , whereas the S. pombe Sty1 pathway, like mammalian p38, is activated by a range of adverse conditions . Models of SAPK-dependent regulation of transcription have been almost exclusively based upon the positive control of activators. However, recent analysis of S. cerevisiae has demonstrated that the Sko1 repressor regulates the expression of Hog1-dependent osmostress genes, such as ENA1 and GRE2, via recruitment of the Ssn6(Cyc8)-Tup1 global corepressor complex . Ssn6-Tup1 mediates its function via the organization of repressive chromatin structures and by inhibition of the basal transcription machinery . This global repressor controls the expression of numerous genes through interaction with a variety of site-specific DNA binding proteins . Relief from this repression is achieved by control of the proteins that serve to tether the complex to DNA; for example, Sko1 is phosphorylated by Hog1 at three sites in its N-terminal region, disrupting the interaction with Ssn6-Tup1 . Therefore, a component of the osmotic induction of some genes occurs via derepression rather than by activation. In fission yeast, Sty1 operates via the transcriptional activators Atf1/Gad7 and Pap1 . Atf1 is phosphorylated in a Sty1-dependent manner and loss of Atf1 results in hypersensitivity to osmotic stress, high levels of calcium, and an inability to respond to deteriorating nutritional conditions . In addition, Atf1 forms a heterodimeric complex with Pcr1, a related ATF/CREB factor, which is also required for transcriptional induction of some stress genes . Pap1 activates transcription in response to oxidative stress, and its subcellular localization is regulated in a Sty1-dependent manner . Recently, Prr1, a homolog of Skn7 in S. cerevisiae , has also been implicated in the transcriptional response to oxidative stress . Skn7 and Prr1 have heat-shock factor-like DNA binding domains and also share homology with bacterial "two-component" response regulators that are controlled by histidine-to-aspartate phosphorelay systems . Herein, we have addressed the roles of the Tup-like repressors Tup11 and Tup12 in the response to stress in S. pombe. We find that deletion of both tup genes in combination results in hypersensitivity to KCl and CaCl2, and we also identify cta3+ as a novel stress gene that is negatively regulated by Tup11-Tup12. The expression of cta3+ is rapidly and specifically induced in response to salt shock in a Sty1- and Atf1-dependent manner, but the dependence on the Sty1 pathway for induction is lost in a tup11Delta tup12Delta mutant. Furthermore, Tup11 and Tup12 proteins function as specificity factors by preventing induction of cta3+ in response to inappropriate stresses such as heat and oxidative stress. We also reveal a new role for the "response regulator" protein Prr1 and demonstrate that it is required for proper KCl-mediated transcriptional induction of Sty1-dependent genes such as cta3+, ctt1+, and gpx1+. MATERIALS AND METHODS : Strains | Routine culture of S. pombe and general genetic methods were performed as described in . The strains used in this study are described in Table . The cta3+ gene was disrupted using a polymerase chain reaction (PCR)-based approach as described by . Oligonucleotides 5' KO (5'-TTTGATTTTACTTATATTTCTCCCCTTCTACTCATCCCGATATATTCTTACTTCCTTGATTCAATCTCAAATATTGTTCAGCTTAGCTACAAATCCCACT-3') and KO 3' (5'-ATAAATCCTTTACGATTTGTCGGTTCTGTGAAAACGATACACTCACGCATATTCATATACATATTCATGGCAAGAAAACATCTGACATAAAACGCCTAGG-3') were used to amplify a 1.6-kb ura4+-containing fragment from pRep42. The amplified fragment was used to transform strain NT5 strain to Ura+, creating strain SW95. Integration at the correct locus was confirmed by PCR analysis. Table 1 | Strains used in this study Plasmids | The tup11+ coding sequence was amplified by PCR from a cDNA library by using the following primers: 5'-GCACGGATCCCATGGCGTCAGTGGAGGATGC-3' and 5'-CTAGGGATCCAATTCAA-GGAGATGCAGGGTC-3'. The tup12+ coding sequence was amplified using primers 5'-GCACGGATCCCATGATTACTGTCCGCCAATC-3' and 5'-CTGCTAGGCATATGGCGCTCATGAAACAAACCG-3'. Fragments were cleaved with BamHI and cloned into the BamHI site of derivatives of pRep41 and pRep42 vectors that allow the expression of proteins as N-terminal HA or 6HisMyc fusions . RNA Analysis | RNA samples were prepared from 0.25 to 0.5 x 109 cells. Pellets were washed in H2O and resuspended in 200 mul of RNA buffer (50 mM Tris-HCl pH 8.0, 100 mM NaCl, 50 mM EDTA pH 8.0, and 0.25% SDS) with 200 mul of phenol/chloroform. Cells were disrupted with 0.75 ml of glass beads (0.5 mm; Biospec Products, Bartlesville, OK) in a Ribolyser (Hybaid, Middlesex, United Kingdom). A further 0.75 ml of RNA buffer was added followed by spinning in a microfuge for 10 min. The aqueous layer was subjected to two further phenol/chloroform extractions before the RNA was precipitated with 0.1 volume of sodium acetate, pH 5.2, and 0.6 volume of isopropanol. RNA pellets were washed in 70% ethanol and resuspended in H2O. RNA analysis was as described by . Briefly, a 10 --15-mug sample of total RNA was denatured with glyoxal, separated on a 1.2% agarose gel prepared in 15 mM sodium phosphate, pH 6.5, and transferred to a GeneScreen hybridization membrane (PerkinElmer Life Sciences, Boston, MA). The his3+ probe has been described previously . Other gene-specific probes were produced by PCR amplification from genomic DNA by using the appropriate primers. All probes were labeled with [alpha-32P]dCTP by using a Prime-a-Gene labeling kit (Promega, Madison, WI). Transcript levels were quantified relative to the loading control using a PhosphorImager BAS-1500 (Fuji Photo Film, Tokyo, Japan). beta-Galactosidase Assays | Assays were performed as described previously . Coprecipitations | Whole cell extracts were prepared as described by with some modification. Cultures were grown to mid-log phase (OD595 = 0.25 --0.5) in EMM medium. Cells were harvested washed once and snap frozen. Pellets were washed in 1 ml of lysis buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 0.5% NP-40, 10 mM imidazole, 2 mug/ml pepstatin, 2 mug/ml leupeptin, 2 mug/ml aprotinin, and 100 mug/ml phenylmethylsulfonyl fluoride). Cells were disrupted with 2 ml of glass beads by vortexing twice for 45 s with 1-min incubation on ice in between. Protein extracts were recovered and centrifuged at 13,000 rpm for 10 min at 4C. Protein precipitations were performed by adding 25 mul of nickel-agarose (50% slurry in lysis buffer) to 1 mg of whole protein extract and incubating at 4C for 1 h with gentle agitation. Precipitates were recovered by centrifugation and washed four times with lysis buffer containing 200 mM NaCl and 20 mM imidazole. Samples were analyzed by SDS-PAGE and proteins were transferred to nitrocellulose membrane and subjected to Western blotting by using monoclonal hemagglutinin (HA) (12CA5) antibody (Babco, Berkeley, CA). Electrophoretic Mobility Shift Assays (EMSAs) | Whole cell extracts were prepared as described above except that cells were grown in YE5S medium and extracts were prepared in buffer containing 25 mM HEPES pH 7.6, 0.1 mM EDTA, 150 mM KCl, 0.1% Triton X-100, 25% glycerol, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride 2 mug/ml pepstatin, 2 mug/ml leupeptin, and 2 mug/ml aprotinin. Radiolabeled DNA fragments were prepared using PCR amplification as described in . The oligonucleotides used for amplification of probe 1 were 5'-TAAAACACCGACATGTAGCC-3' and 5'-TTGAGAGAAACTAACCAAGG-3'. The oligonucleotides for probe 2 were 5'-CTCTGTCATGGAAATCCACAC-3' and 5'-ATAAGCAGCAAAGCTTGCCTG-3'. Binding reactions were performed by adding 15 mug of whole cell extract to 20-mul reactions containing 25 mM HEPES pH 7.6, 34 mM KCl, 5 mM MgCl2, and 2 mug of poly[d(I-C)]. Reactions were incubated for 10 min at room temperature before the addition of similar0.5 ng of radiolabeled probe DNA followed by a further 20-min incubation. Samples were analyzed by electrophoresis through 4% polyacrylamide gels run in 0.5x Tris-borate-EDTA buffer. Antibody supershift was performed by adding 0.2 mug of monoclonal HA antibody (12CA5) (Babco) 10 min after the addition of probe DNA. RESULTS : To address the role of repressors in the transcriptional response to stress in fission yeast we examined whether cells lacking the tup genes tup11+ and tup12+ exhibited any stress-related phenotypes. We found that single and double tup mutant strains exhibited an increased tolerance to cadmium but that the tup11Delta tup12Delta mutant strain had decreased tolerance to elevated levels of Ca2+ and K+ ions. The degree of sensitivity to these salt stresses was similar to that associated with loss of either the Atf1 transcription factor or the Sty1 MAP kinase that are known to control the induction of genes in response to elevated cation concentrations . The tup11Delta tup12Delta double mutant strain was only slightly less sensitive to KCl than the sty1-1 and atf1Delta strains and the tup11Delta tup12Delta strain was actually more sensitive to CaCl2 than strain lacking Sty1 (Figure A). In contrast, the double tup mutant strain had wild-type levels of tolerance to high sorbitol concentrations, indicating that although they are K+- and Ca2+-intolerant they are not osmosensitive (our unpublished data). Figure 1 | Stress-related phenotypes of cells lacking Tup repressors. Stress-related phenotypes of cells lacking Tup repressors. (A) Exponentially growing cultures (similar0.4 x 107 cells/ml) were diluted serially, spotted onto YE5S agar, and incubated for 2 --3 d at 30C or spotted onto YE5S agar supplemented with CaCl2, KCl, or CdSO4 at the indicated concentration and incubated for 3 --4 d at 30C. (B) Exponentially growing cultures of wild type (w.t.) (NT4), atf1Delta (NT147), and tup11Delta tup12Delta (SW76) strains were treated with H2O2 (to a final concentration of 50 mM), and viable cell numbers were determined by plating onto YE5S agar. The similarity in the sensitivities of tup11Delta tup12Delta and atf1Delta strains to elevated K+/Ca2+ ions was unexpected because Tup11 and Tup12 have been previously demonstrated to be repressors , whereas Atf1 is primarily a transcriptional activator. We therefore investigated whether tup11Delta tup12Delta cells shared any other phenotypes with atf1Delta cells. It has recently been demonstrated that atf1Delta cells are sensitive to an acute oxidative stress . When challenged with a high dose of H2O2 (50 mM) atf1Delta cells rapidly lose viability (Figure B). In contrast, tup11Delta tup12Delta cells were only slightly more sensitive than wild-type cells in this assay. Furthermore, although atf1Delta cells conjugate poorly , a tup11Delta tup12Delta strain conjugates in nutrient-rich media . Hence, tup11Delta tup12Delta cells and atf1Delta cells share only a subset of phenotypes. Taken together, these findings are consistent with Tup11 and Tup12 having overlapping functions and indicate that Tup11 and Tup12 play roles in the cellular response to stress. Tup11 and Tup12 Negatively Regulate Expression of Salt-Stress gene cta3+ | Cells lacking atf1+ and both tup genes have similar sensitivities to salt stress and so we examined whether the expression of genes known to be induced by exposure to salt stress, via Atf1, were also regulated by the Tup repressors (Figure , A and B). We found that expression of cta3+, which encodes a cation-transporting P-type ATPase , was markedly influenced by loss of the Tup proteins; deletion of both tup genes together resulted in a large increase in the basal level of expression (Figure , A and B). Loss of both Tup proteins also resulted in a large increase in expression after exposure to CaCl2 (Figure , A and B) and KCl (Figures , , and ). Thus, Tup11 and Tup12 function in a partially redundant manner to repress cta3+ expression and limit the level of induction. As previously observed , the induction of cta3+ expression is completely dependent upon both the Sty1 MAP kinase and the Atf1 transcription factor, and thus cta3+ displays a novel pattern of stress regulation that is positively controlled by Sty1 and Atf1 but negatively regulated by the Tup repressors. Indeed, deletion of the tup genes either singly or in combination had only minor effects on the expression levels of other Sty1-dependent genes such as gpd1+, ctt1+, and gpx1+ in unstressed cells and in cells subjected to a CaCl2 shock (Figure , A and B). Figure 2 | Tup11 and Tup12 repress the transcription of the salt stress gene cta3+. (A) Strains used are indicated above the lane and were wild type (w.t.) (NT4), tup12Delta (BSP03), tup11Delta (SW53), tup11Delta tup12Delta (SW76), atf1Delta (78 Tau147), and sty1-1 (JM1144). Tup11 and Tup12 repress the transcription of the salt stress gene cta3+. (A) Strains used are indicated above the lane and were wild type (w.t.) (NT4), tup12Delta (BSP03), tup11Delta (SW53), tup11Delta tup12Delta (SW76), atf1Delta (78 Tau147), and sty1-1 (JM1144). Log phase cultures growing at 30C in YE5S (lanes 1, 4, 7, 10, 13, and 16) were incubated with CaCl2 (to final concentration of 0.3 M) for 15 min (lanes 2, 5, 8, 11, 14, and 17) and 30 min (lanes 3, 6, 9, 12, 15, and 18). Total RNA was extracted, separated by electrophoresis, and Northern blots were analyzed with the indicated probes. The level of his3+ mRNA was used as a loading control. (B) Quantification of cta3+ mRNA levels in A. (C) Influence of Tup proteins on the activity of a cta3-lacZ reporter. beta-Galactosidase assays were performed on extracts derived from exponentially growing cells (open bars) and cells treated with KCl to 0.6 M for 60 min (black bars). The strains used were wild type (w.t.) (HAI003) and tup11Delta tup12Delta (SW107). Data is the mean of three independent cultures. (D) Deletion of cta3+ does not rescue the salt sensitivity of tup11Delta tup12Delta cells. Exponentially growing cultures (similar0.4 x 107 cells/ml) were diluted serially, spotted onto YE5S agar and incubated for 2 --3 d at 30C or spotted onto YE5S agar supplemented with CaCl2 or KCl (at the indicated concentration) and incubated for 3 --4 d at 30C. Strains used were wild type (w.t.) (NT4), tup11Delta tup12Delta (SW76), cta3Delta (SW95), and cta3Delta tup11Delta tup12Delta (SW93). To confirm that increased level of cta3+ transcripts associated with deletion of tup genes was due to an effect on transcription and not mRNA stability, we measured the expression of an integrated cta3+ promoter ---lacZ reporter . It is highly unlikely that Tup repressors would specifically influence the stability of lacZ transcripts. Consistent with the Northern analysis, deletion of both the tup genes resulted in 14-fold increase in expression of the lacZ reporter relative to the wild-type control (Figure C). Furthermore, exposure of cells to high KCl concentrations (0.6 M for 1 h) increased the level of expression sevenfold in wild-type cells and threefold in the cells lacking the Tup repressors (Figure C). These results suggest that S. pombe Tup proteins exert their effects at the level of transcription. To determine whether the high level of cta3+ expression observed in the tup11Delta tup12Delta double mutant confers the increased sensitivity of this strain to elevated K+ and Ca2+ concentrations we examined the effect of deleting the cta3+ gene in the presence or absence of the tup genes. Loss of Cta3 function has previously been reported to result in increased sensitivity to elevated Ca2+ concentrations . In contrast, found that cta3 null cells did not exhibit any detectable change in resistance to K+, Ca2+, or Na+ ions. In agreement with the latter study our cta3Delta mutant exhibited wild-type levels of resistance to both Ca2+ and K+ (Figure D). Moreover, deletion of the cta3+ gene in a tup11Delta tup12Delta strain did not rescue the salt-sensitive phenotype associated with the loss of the tup genes (Figure D), and plasmid-mediated overexpression of cta3+ in wild-type cells did not result in any increased sensitivity to KCl or CaCl2 (our unpublished data). These results indicate that the salt sensitivity of tup- cells is not simply due to the elevated expression of the cta3+ gene. To date, the fbp1+ gene encoding fructose 1,6-bisphosphatase is the only gene that has been identified as a target gene for Tup11-Tup12 --mediated repression . The expression of fbp1+ is also positively regulated by the Sty1 pathway , but its expression is induced by carbon limitation and not by other acute stresses that activate Sty1, such as heat shock, oxidative stress, and osmotic shock. Furthermore, the cAMP pathway negatively regulates the expression of fbp1+, and mutations that disrupt this pathway result in increased expression under repressing (glucose-rich) conditions . In contrast, growing cells under carbon-limiting conditions did not induce the expression of cta3+ nor were mRNA levels influenced by a deletion of git2+ that encodes adenylate cyclase (our unpublished data). Formation of Protein Complexes on the cta3+ Promoter | Studies have indicated that Atf1 binds constitutively to a CRE binding site in the gpd1+ promoter . In contrast, EMSAs of the fbp1+ promoter have demonstrated that Atf1 associates with a CRE-like element in UAS1 only under activating (glucose-limiting) conditions . We therefore examined the ability of Aft1 to bind to the cta3+ promoter. Inspection of the DNA sequence revealed the presence of a number of potential CRE-like Atf1 binding sites located between -1111 and -1401 relative to the initiation codon (Figure A). We performed EMSAs by using whole cell extracts and a DNA fragment corresponding to the -1477 to -1297 region of the promoter. This region includes a near consensus CRE element and two CRE-like sequences containing the highly conserved ACGT core sequence. A major slow-migrating complex was formed on this probe (Figure B). This binding activity was not changed by subjecting cells to stress (KCl 0.6 M for 15 min) before extract preparation. The complex was also present in extracts derived from atf1Delta cells, indicating that it does not require Atf1. Furthermore, the mobility of the complex was unchanged when HA antibody was included in reactions containing HA epitope-tagged Atf1 (our unpublished data). Next, we examined the ability of complexes to form on a probe corresponding to the -1249 and -1058 region of the promoter that contains a single CRE element. In this case, we also observed a binding activity that was Atf1-independent (Figure C). However, we also detected a slow-migrating complex that was absent in reactions lacking Atf1. Also, the mobility of this complex was reduced by the addition of the HA antibody to reactions containing HA epitope-tagged Atf1. This Atf1-dependent complex was present in reactions using extracts derived from unstressed and stressed cells, indicating that at least under these experimental conditions Atf1 binds constitutively to this region of the cta3+ promoter. We were unable to properly assess the role of Tup proteins on DNA binding activity; when tup- extracts were used a marked reduction in the level of complex formation on both probes was observed. However, this seemed to be due to difficulties in preparing extracts from these cells rather than a specific effect because we found that extracts lacking Tup proteins also showed a reduced ability to form complexes on a DNA probe unrelated to the cta3+ promoter (our unpublished data). Figure 3 | Complexes formed on the cta3+ promoter. Complexes formed on the cta3+ promoter. (A) Schematic of the CRE-like elements in the cta3+ promoter and the probes used for EMSAs. Consensus or near consensus CRE elements are shaded black, and the CRE-like sequences containing an ACGT core are shaded gray. Their locations are given relative to the initiation codon. (B) EMSAs were performed with probe 1 by using whole cell extracts derived from exponentially growing atf1HA (KS14709), w.t. (NT4), or atf1Delta (NT147) cells as indicated above the lanes. The extract used in lane 3 was derived from atf1HA6H cells treated with KCl (0.6 M for 15 min) immediately before extract preparation. Lane 1 is a probe alone control. (C) EMSAs were performed with probe 2 by using extracts derived from exponentially growing cells (lanes 2, 4, and 5) or cells treated with KCl (0.6 M for 15 min) immediately before extract preparation (lane 3). The identity of the extract is indicated above the lane. The reaction in lane 4 contained 0.2 mug of HA antibody. Lane 1 is a probe alone control. Arrows mark the positions of the Atf1-dependent shifted (S) and supershifted (SS) complexes. Tup11 and Tup12 Interact | Our data indicate that Tup11 is capable of repressing cta3+ expression in the absence of Tup12 and vice versa. It is also known that S. cerevisiae Tup1 tetramerizes through its N-terminal domain and based on homology it is very likely that the S. pombe Tup proteins form homotetramers. However, it is possible that in addition to functioning in homomeric complexes Tup11 and Tup12 may also function in a heteromeric complex. Therefore, we investigated the ability of Tup11 and Tup12 to interact using a coprecipitation assay. Whole cell extracts were prepared from wild-type cells that expressed 6His-tagged Tup11 (or Tup12) and coexpressed HA-tagged Tup11 (or Tup12). Ni2+-agarose was then used to precipitate His-tagged Tup proteins, and the presence of HA-tagged Tup proteins was examined by Western blotting . In these experiments, Tup11 copurified with Tup12 and vice versa, indicating that Tup11 and Tup12 physically interact. The specificity of this interaction was demonstrated by the absence of HA-tagged Tup proteins in control precipitates derived from cells extracts expressing the empty 6His vector (Figure , lanes 7 and 8). Thus, Tup11 and Tup12 have the potential to regulate gene expression in the same protein complex. Figure 4 | Tup11 and Tup12 coprecipitate. Tup11 and Tup12 coprecipitate. Whole cell extracts were prepared from wild-type cells containing plasmids expressing epitope-tagged Tup proteins: pRep42-HisMycTup12 (lanes 1 --3), pRep42-HisMycTup11 (lanes 4 --6), pRep42-HisMyc empty vector (lanes 7 and 8), pRep41-HATup12 (lanes 1, 4, and 7), and pRep41-HATup11 (lanes 2, 5, and 8). Extracts were precipitated with Ni2+-agarose analyzed on 8% SDS polyacrylamide gels and subjected to Western blotting by using HA monoclonal antibody. Tup Repressors Link Transcriptional Induction to the Sty1 Pathway | To test whether the high level of cta3+ expression observed in the absence of the Tup repressors was dependent upon the Sty1 MAP kinase cta3+ mRNA levels were examined in a strain that lacks both Sty1 and Tup function (sty1-1 tup11Delta tup12Delta). In this mutant the level of cta3+ transcripts was similar to that observed in tup11Delta tup12Delta cells, indicating that Sty1 is not required for basal levels of expression (Figure , A and B). Surprisingly, exposure of this strain to a KCl-mediated shock resulted in induction of cta3+ expression, indicating that in the absence of the Tup proteins the Sty1 MAP kinase is not required for the stress-mediated induction of cta3+. Consistent with these observations the expression of cta3+ was also induced by salt shock (0.6 M KCl) in a sty1Delta tup11Delta tup12Delta strain (our unpublished data). The expression of other genes such as pyp2+ and gpd1+ was not induced in the sty1-1 tup11Delta tup12Delta triple mutant, although deletion of the tup genes did restore the basal level of expression in sty1- cells (Figure A). The kinetics of KCl-mediated induction of cta3+ were similar in wild-type and tup11Delta tup12Delta cells, with mRNA levels peaking at 20 min but elevated mRNA levels persisted for a greater length of time in cells lacking the Tup proteins (Figure , C and D). In the sty1-1 tup11Delta tup12Delta triple mutant strain induction was delayed and peak mRNA levels were not observed until 30 min after the addition of KCl. Figure 5 | Deletion of tup11+ and tup12+ allows Sty1-independent transcriptional induction of cta3+. (A) Strains used are indicated above the lanes and were wild type (w.t.) (NT4), sty1-1 (JM1144), sty1-1 tup12Delta (SW88), sty1-1 tup11Delta (SW89), tup11Delta tup12Delta (SW76), and sty1-1 tup11Delta tup12Delta (SW90). Deletion of tup11+ and tup12+ allows Sty1-independent transcriptional induction of cta3+. (A) Strains used are indicated above the lanes and were wild type (w.t.) (NT4), sty1-1 (JM1144), sty1-1 tup12Delta (SW88), sty1-1 tup11Delta (SW89), tup11Delta tup12Delta (SW76), and sty1-1 tup11Delta tup12Delta (SW90). Log phase cultures growing at 30C in YE5S (lanes 1, 3, 5, 7, 9, and 11) were incubated with KCl (to a final concentration of 0.6 M) for 15 min (lanes 2, 4, 6, 8, 10, and 12). Total RNA was extracted, separated by electrophoresis, and analyzed by Northern blotting with the indicated probes. The level of his3+ mRNA was used as a loading control. (B) Quantification of the cta3+ mRNA levels in A. (C) Kinetics of the induction of cta3+ mRNA in response to KCl shock. Log phase cultures growing at 30C in YE5S (lanes 1, 7, and 13) were incubated with KCl (to a final concentration of 0.6 M) for 10 min (lanes 2, 8, and 14), 20 min (lanes 3, 9, and 15), 30 min (lanes 4, 10, and 16), 40 min (lanes 5, 11, and 17), or 50 min (lanes 6, 12 and 18). The strains used were wild type (w.t.) (NT4), tup11Delta tup12Delta (SW76), and sty1-1 tup11Delta tup12Delta (SW90). Total RNA was extracted, separated by electrophoresis, and analyzed by Northern blotting with the indicated probes. (D) Quantification of the cta3+ mRNA levels in C. We next determined whether removal of Tup11 and Tup12 rescued any of the other phenotypes associated with loss of Sty1. We examined the ability of cells to grow on medium supplemented with cadmium. Deletion of the tup genes in a sty1+ background increases resistance to cadmium (Figure A) but unexpectedly deletion of tup11+ and tup12+ in a sty1-1 background reduced cadmium tolerance (Figure A). Thus, the resistance of tup- cells to cadmium depends on Sty1 function and in its absence they become hypersensitive. The elongated cell morphology of sty1-1 cells that is indicative of a G2 cell cycle delay was slightly exacerbated by deletion of the tup genes (Figure B). Furthermore, deletion of the tup genes in an aft1Delta or a sty1-1 background resulted in a small increase in sensitivity to KCl and the tup11Delta tup12Delta atf1Delta triple mutant strain was slightly less tolerant to CaCl2 than the parental strains (Figure C). Figure 6 | Genetic interactions. Genetic interactions. (A) Deletion of the tup genes does not rescue phenotypes associated with the sty1-1 mutation. The indicated strains were subcultured onto YE5S agar (control) or subcultured onto YE5S agar supplemented with CdSO4 (to the indicated concentration) and incubated at 30C for 3 --4 d. (B) Comparison of the morphology of wild type (w.t.) (NT4), sty1-1 (JM1144), tup11Delta tup12Delta (SW76), and sty1-1 tup11Delta tup12Delta (SW90) cells. (C) Exponentially growing cultures were diluted serially, spotted onto YE5S agar or YE5S agar supplemented with CaCl2 or KCl (at the indicated concentration), and incubated for 2 d at 30C. The strains used were wild type (w.t.) (NT4), tup11Delta tup12Delta (SW76), sty1-1 (JM1144), sty1-1 tup11Delta tup12Delta (SW90) atf1Delta (NT147), and atf1Delta tup11Delta tup12Delta (SW92). We next addressed whether removal of Tup11-Tup12 repression rendered the induction of cta3+ independent of Atf1. Deletion of atf1+ in a tup11Delta tup12Delta mutant strain resulted in a further increase in cta3+ transcript levels in unstressed cells, suggesting that nonactivated Atf1 may have a repressive effect on transcription that is independent of Tup11 and Tup12 (Figure , A and B). A similar effect has been observed previously; the decrease in the basal level of ctt1+ mRNA associated with loss of Sty1 function is suppressed by deletion of atf1+ . In the atf1Delta tup11Delta tup12Delta background, cta3+ mRNA levels did not increase after exposure to KCl (0.6 M), indicating that Atf1 is absolutely required for induction of cta3+ in response to a salt shock. Figure 7 | Transcription induction of cta3+ in cells lacking Tup11 and Tup12 is Atf1 dependent. Transcription induction of cta3+ in cells lacking Tup11 and Tup12 is Atf1 dependent. (A) Strains used are indicated above the lanes and were wild type (w.t.) (NT4), atf1Delta (NT147), tup11Delta tup12Delta (SW76), and atf1Delta tup11Delta tup12Delta (SW92). Log phase cultures growing at 30C in YE5S (lanes 1, 3, 5, and 7) were incubated with KCl (to a final concentration of 0.6 M) for 15 min (lanes 2, 4, 6, and 8). Total RNA was extracted, separated by electrophoresis and subjected to Northern analysis with the indicated probes. The level of cdc2+ mRNA was used as a loading control. (B) Quantification of cta3+ mRNA levels in A. (C) As for A, except strains used were wild type (w.t.) (NT4), pcr1Delta (JX25), tup11Delta tup12Delta (SW76), and pcr1Delta tup11Delta tup12Delta (RJP59). (D) Quantification of cta3+ mRNA levels in C. The bZIP transcription factor Pcr1 that can heterodimerize with Atf1 is also required for stress-mediated induction of cta3+ expression (Figure , C and D). Examination of cta3+ mRNA levels in a pcr1Delta tup11Delta tup12Delta strain revealed that Pcr1 is not required for the high level of basal expression, and furthermore in this strain expression of cta3+ was partially induced in response to a salt shock. Thus, the Tup repressors ensure that induction of cta3+ remains dependent upon the Sty1 MAP kinase and to a lesser extent, the activator Pcr1. Prr1 Is Involved in Regulation of Gene Expression in Response to Elevated K+ Ions | Our analysis suggests that another factor may regulate transcription of cta3+ independently of Sty1. The Prr1 transcription factor is known to regulate oxidative stress responsive genes , but there is no evidence that Sty1 regulates its activity directly. Therefore, we analyzed mRNA levels in a prr1Delta strain and found that the level of cta3+ transcripts after exposure to KCl was significantly reduced in comparison with the wild-type strain (Figure , A and B). Furthermore, the influence of Prr1 was not confined to the cta3+ gene because KCl-mediated induction of both ctt1+ and gpx1+ expression was also significantly reduced in the prr1Delta mutant strain. This was surprising because Prr1 has previously been reported not to be involved in the transcriptional response to high salt and indeed KCl-mediated induction of some genes such as gpd1+ occurs independently of Prr1 (; Figure , A and B). To determine the role that Prr1 plays in control of cta3+ expression, we measured cta3+ mRNA levels in a tup11Delta tup12Delta prr1Delta triple mutant strain. In this background the expression of cta3+ was induced by exposure to high concentrations of KCl (0.6 M) (Figure , C and D). However, the deletion of prr1+ in a tup11Delta tup12Delta background resulted in a decrease in the basal level of cta3+ mRNA (Figure , C and D). Thus, Prr1 activity contributes to the high basal level of expression that is associated with loss of the Tup repressors. Figure 8 | Prr1 is involved in the regulation of gene expression in response to high salt. Prr1 is involved in the regulation of gene expression in response to high salt. (A) Strains used are indicated above the lanes and were wild type (w.t.) (NT4) and prr1Delta (SW97). Log phase cultures growing at 30C in YE5S (lanes 1 and 3) were incubated with KCl (to a final concentration of 0.6 M) for 15 min (lanes 2 and 4). Total RNA was prepared and subjected to Northern analysis with the indicated probes. The level of his3+ mRNA was used as a loading control. (B) Quantification of the mRNA levels in A. (C) Strains used are indicated above the lanes and were tup11Delta tup12Delta (SW76) and prr1Delta tup11Delta tup12Delta (SW96), and the treatment was as described in A. (D) Quantification of the mRNA levels in C. In vitro experiments have demonstrated that recombinant Prr1 binds to a heat shock-like element in the ste11+ promoter . Analysis of the cta3+ promoter revealed the presence of such an element (GGAAAATTC) located at -2068 relative to the initiation codon. However, in assays using this region of the promoter and whole cell extracts we were unable to detect a Prr1-dependent binding activity (our unpublished data). Therefore, we cannot exclude the possibility that the role of Prr1 in regulation of cta3+ expression is indirect. Tup11 and Tup12 Prevent Induction in Response to Inappropriate Stresses | Sty1, and thus in turn Atf1-Pcr1, is activated in response to a number of environmental insults. Accordingly, the expression of Atf1- and Pcr1-dependent genes such as ctt1+, pyp2+, and gpx1+ are induced in response to a variety of stresses such as UV irradiation, heat shock, and hyperosmolarity and an oxidative stress elicited by exposure to high concentrations of H2O2 . However, some Atf1-Pcr1 target genes are induced only by a subset of these stresses. For example, we found that cta3+ expression was induced specifically in response to salt shock but not by oxidative stress (6 mM H2O2) or by heat shock (15 min at 42C) (Figure , A and D). In contrast, ctt1+ and gpx1+ mRNA levels were both induced by these treatments, indicating that Atf1 (and Pcr1) were active under these conditions. This indicates that activation of the Sty1 pathway per se is not sufficient to induce the expression of cta3+ and mechanisms must exist to prevent induction of gene expression in response to such "inappropriate stresses." We wanted to examine the possibility that Tup11 and Tup12 play a role in this process. Therefore, we measured the levels of cta3+ mRNA after exposing a tup11Delta tup12Delta strain to an oxidative stress (6 mM H2O2) and a heat shock (15 min at 42C). In contrast to the wild-type strain, the expression of cta3+ was significantly induced by heat stress and by exposure to high levels of H2O2. In addition, the level of cta3+ transcripts was induced by hypotonic conditions in a tup11Delta tup12Delta strain but not in a wild-type strain (our unpublished data). We also examined cta3+ transcript levels in a sty1Delta tup11Delta tup12Delta triple mutant strain (Figure , B and D). This revealed that the induction in expression in response to heat shock was partly independent of the MAP kinase. In contrast, the induction of cta3+ expression in response to oxidative stress mediated by H2O2 was completely dependent upon Sty1, suggesting a difference in the mechanism of induction. Further analysis indicated that the response to heat shock was independent of Prr1 (Figure , C and D) but dependent upon Atf1 (our unpublished data). Taken together, these findings indicate that the Tup repressors function as part of the mechanism that ensures the specificity of stress-mediated transcriptional induction at the cta3+ promoter. Figure 9 | Tup11 and Tup12 prevent induction in response to inappropriate stresses. Tup11 and Tup12 prevent induction in response to inappropriate stresses. (A) Strains used were wild type (w.t.) (NT4) and tup11Delta tup12Delta (SW76). Mid log cultures growing at 30C in YE5S (lanes 1 and 4) were incubated with H2O2 (final concentration 6 mM) for 15 min (lanes 2 and 5) or shifted to 42C for 15 min (lanes 3 and 6). Total RNA was prepared and subjected to Northern analysis with the indicated probes. The level of cdc2+ mRNA was used as a loading control. (B) Strain used was sty1Delta tup11Delta tup12Delta (SW91) and the treatments were as described in A. (C) Strain used was prr1Delta tup11Delta tup12Delta (SW91) and the treatments were as described in A. (D) Quantification of the cta3+ mRNA levels in A, B, and C. The strains are indicated above the graphs and the treatments are indicated below. DISCUSSION : In this study, we reveal roles for S. pombe Tup11 and Tup12 in the cellular response to elevated K+ and Ca2+ levels. We identify cta3+ as a novel stress-induced gene whose transcription is coregulated by the Sty1 MAP kinase pathway and the Tup repressors. Our results indicate that the Tup repressors fulfill a number of functions in the control of cta3+ expression. First, they maintain low levels of basal expression and limit the level of induction. Second, they ensure that induction of expression is linked to the Sty1 pathway. And third, they maintain the specificity of induction. We also reveal a new role for the response regulator Prr1 and demonstrate that it functions to regulate gene expression in response to elevated salt concentrations. Prr1 is known to contribute to the regulation of several genes whose expression is induced by oxidative stress via the Pap1 transcription factor , and so Prr1 regulates gene expression in response to a number of stresses. Tup11 --Tup12 Interaction | Our data and that of others suggest that Tup11 and Tup12 can function in homomeric complexes. In addition, we demonstrate that Tup11 and Tup12 have the potential to form a heteromeric complex. This is significant because full function requires both repressors; some derepression of a fbp1::lacZ reporter is observed upon deletion of a single tup gene . Furthermore, single tup mutants have demonstrable phenotypes such as increased resistance to cadmium. Thus, regulation of some genes may depend upon both repressors and the formation of heteromeric Tup complexes. Relief of Tup11-Tup12 --mediated Repression | The Hog1 MAP kinase in S. cerevisiae plays a direct role in relieving Ssn6-Tup1 --mediated repression at osmostress genes; Hog1 phosphorylates Sko1, reducing its affinity for the corepressor complex . It is possible that the Sty1 MAP kinase may similarly antagonize the action of Tup11-Tup12; however, our results demonstrate that the Atf1, Pcr1, and Prr1 transcription factors are required for relief from Tup-mediated repression at the cta3+ promoter. It is probable that the S. pombe Tup proteins function, at least in part, through the organization of repressive chromatin structures , and therefore it is possible that Atf1-Pcr1 and Prr1 overcome this repression by recruiting positive-acting chromatin remodeling complexes such as Swi-Snf or histone acetylase complexes (HATs). In support of this, DNA binding by the Atf1-Pcr1 heterodimer is known to alter local nucleosome positioning at the ade6-M26 hotspot and thereby promote meiotic recombination . Moreover, genes such as SUC2 in S. cerevisiae are regulated by the interplay between Ssn6-Tup1 repression and Swi-Snf --mediated activation . In S. cerevisiae, Hog1-dependent transcriptional induction of HAL1 requires the Gcn4 activator that relieves Tup1-Ssn6 --mediated repression by competing with Sko1 for the occupancy of a single CRE binding site . This CRE element functions as a dual control element and integrates both positive and negative regulatory signals. Furthermore, analysis of the S. pombe fbp1+ promoter, which is regulated by both Atf1-Pcr1 and Tup11-Tup12, has demonstrated the presence of a control element (called UAS2) that contains a CRE-like sequence and is bound by multiple activators and repressors . Interestingly, the Atf1-Pcr1 transcription factor does not bind to UAS2 directly, but it does influence the protein complexes that assemble on it . The cta3+ promoter contains a number of CRE-like elements at least one of which mediates Atf1 binding (Figure C). It will be interesting to determine the contributions of these elements to activation and repression of cta3+ transcription. Tup Repressors Maintain Specificity of Induction | The advent of stressful conditions results in the rapid and Sty1-dependent phosphorylation of Atf1 . Although the precise role of Atf1 phosphorylation remains obscure, it is evident that transcriptional activation by Atf1 is dependent upon Sty1. However, deletion of the tup genes allows transcriptional induction of cta3+ to occur in sty1- cells. Furthermore, in a tup11Delta tup12Delta mutant induction of cta3+ expression does not require Pcr1 or Prr1. Thus, the Tup repressors function to "wire" induction to the Sty1 pathway, insulating it from interfering signals. These results also suggest that Atf1 activity can be "uncoupled" from Sty1 in this specific case and that an additional mechanism for activating transcription that requires Atf1 exists. The finding that Prr1 also controls expression of cta3+ suggests that it may function as part of this mechanism. The Sty1 pathway in S. pombe is fundamentally different to the Hog1 pathway in S. cerevisiae because it is triggered by exposure to a wide range of adverse environmental conditions. As a consequence, a large number of Sty1 target genes are up-regulated by multiple stresses. The products of such genes may comprise a set of "general stress response proteins" that are necessary because a single environmental insult may result in multiple classes of intracellular stress . Nonetheless, discrete stimuli also produce distinct transcriptional outputs, because there are subsets of Sty1-dependent genes, such as cta3+, that are induced only by specific stresses. A major question to be addressed is the mechanism by which Sty1 signaling is integrated into the regulation of such genes. Expression of cta3+ is not induced by oxidative stress, heat shock, carbon limitation, or sexual differentiation (Figure ; our unpublished data), and furthermore cta3+ is only poorly induced by an osmotic shock mediated by high sorbitol concentrations . Thus, the transcriptional response is triggered essentially by elevated intracellular cation concentrations rather than by an osmotic effect (i.e., decrease in turgor pressure across the plasma membrane). The cta3+ gene encodes a putative intracellular P-type ATPase transporter that is involved in cation extrusion or sequestration into intracellular compartments. Loss of function leads to an accumulation of cytoplasmic Ca2+ levels , although recent evidence suggests that Cta3 is primarily a K+ ion pump . It is thus consistent that it is salt stress that specifically that triggers its transcriptional induction. However, removal of the constraints imposed by Tup repressors allows cta3+ to be induced in response to other stresses such as elevated temperature and oxidative stress. Thus, the Tup repressors function as a part of a mechanism that adds specificity to Sty1-dependent transcriptional induction. Our results also indicate that activation of the Sty1 pathway alone is insufficient to induce cta3+ expression and implies that an elevated cation concentration triggers an additional pathway that is required to circumvent repression . In this respect it may be significant that Prr1 is involved in the regulation of cta3+ expression because its structure suggests that it may be one part the target of a histidine-aspartate phosphorelay pathway. Recent work has identified several of these pathways in fission yeast , and current experiments are addressing its contribution to the regulation of Prr1 in the response to stress. Figure 10 | Model for the regulation of cta3+ expression. Model for the regulation of cta3+ expression. Under nonstress conditions, cta3+ expression is repressed by Tup11 and/or Tup12 that are tethered to the promoter through interaction with a site-specific DNA binding protein "X." Activation of the Sty1 pathway alone is insufficient to induce expression, and the Tup repressors prevent activation by Atf1-Pcr1 and Prr1. Elevated Ca2+ or K+ concentrations trigger the activity of other pathways (indicated by dashed lines) that interfere with Tup repression and/or facilitate activation via the response regulator Prr1 and Atf1-Pcr1. In cells lacking the Tup repressors, specificity is lost and expression is induced in response to a range of stresses. Backmatter: PMID- 12221111 TI - The Roles of Bud-Site-Selection Proteins during Haploid Invasive Growth in Yeast AB - In haploid strains of Saccharomyces cerevisiae, glucose depletion causes invasive growth, a foraging response that requires a change in budding pattern from axial to unipolar-distal. To begin to address how glucose influences budding pattern in the haploid cell, we examined the roles of bud-site-selection proteins in invasive growth. We found that proteins required for bipolar budding in diploid cells were required for haploid invasive growth. In particular, the Bud8p protein, which marks and directs bud emergence to the distal pole of diploid cells, was localized to the distal pole of haploid cells. In response to glucose limitation, Bud8p was required for the localization of the incipient bud site marker Bud2p to the distal pole. Three of the four known proteins required for axial budding, Bud3p, Bud4p, and Axl2p, were expressed and localized appropriately in glucose-limiting conditions. However, a fourth axial budding determinant, Axl1p, was absent in filamentous cells, and its abundance was controlled by glucose availability and the protein kinase Snf1p. In the bud8 mutant in glucose-limiting conditions, apical growth and bud site selection were uncoupled processes. Finally, we report that diploid cells starved for glucose also initiate the filamentous growth response. Keywords: INTRODUCTION : Cells of the yeast S. cerevisiae can undergo a developmental switch from a yeast form of growth to a filamentous form of growth (; for reviews, see ; ). In haploid cells, one of the triggers for the switch to the filamentous form is glucose starvation . The developmental switch to filamentation has at least three components. First, cells change their budding pattern. For example, haploid cells switch from an axial budding pattern, in which new buds emerge at sites adjacent to the birth scar or the site of the preceding bud, to a unipolar-distal budding pattern, in which new buds emerge at the pole distal to the birth scar . Second, the cells become elongated. Third, the cell surface changes, enabling the cells to adhere to each other and to invade the agar substratum. In this article, we investigate the requirements for the change in budding pattern associated with filamentation in haploid cells. Are proteins required for bud site selection in yeast-form cells necessary for the budding pattern observed during filamentation? If so, which ones, and how are their activities modulated to affect the unipolar pattern during filamentation? In vegetative cells, the budding pattern is controlled by cell type (; ; ; for reviews, see ; ; ; ; ; ; ,). As described above, haploid cells bud in an axial pattern. Diploid cells, on the other hand, bud in a bipolar pattern. A new bud can emerge from either the birth scar pole or the distal pole, although there is a bias for distal pole budding in the first bud formed . A GTPase module is required for cells to display either of these budding patterns; in its absence, cells bud in a random pattern . The module is composed of a RAS-like GTPase, Rsr1p/Bud1p; its GTPase-activating protein, Bud2p; and its guanine nucleotide exchange factor, Bud5p . Bud2p and Bud5p are localized to axial positions in haploid cells and bipolar positions in diploid cells , where they direct bud emergence, in part through interaction with polarity establishment proteins . The recruitment of the GTPase module to the appropriate site is controlled by other bud-site-selection proteins. In haploid cells, axial budding requires Bud3p, Bud4p, and Axl2p/Bud10p/Sro4p . These proteins are localized to the mother-bud neck and together recruit Bud5p to the axial position . In addition, Axl1p is a haploid-specific protein required for axial budding . Loss of Bud3p, Bud4p, Axl2p, or Axl1p causes bipolar budding in haploid cells, but does not affect budding pattern in diploid cells. A different set of factors is required to orchestrate bipolar, rather than axial, budding in diploid cells. Genetic studies suggest that Bud8p and Bud9p mark the poles distal and proximal to the birth scar, respectively . For example, mutants deleted for BUD8 bud exclusively from the proximal pole. Moreover, green fluorescent protein (GFP)-tagging and immunofluorescence studies reveal that Bud8p is located at the distal pole and Bud9p at the proximal pole, implying that they may comprise part of the marks that identify these poles to the GTPase module . In addition, Bud6p and Bni1p, which form a protein complex , are also required for bipolar budding . Loss of any of these four proteins disrupts bipolar budding in diploid cells, but does not affect axial budding in haploid cells. Pea2p and Spa2p are also components of the Bud6p/Bni1p protein complex and are important determinants of bipolar budding and of polarized growth . The switch in budding pattern during filamentous growth is particularly striking in the case of the axial-to-unipolar transition of haploid cells deprived of glucose . We show herein that Bud8p is localized to the distal tip of haploid cells, and under glucose-limiting conditions it directs bud emergence to the distal pole. Glucose depletion results in the Snf1p-dependent disappearance of Axl1p, providing one mechanism by which glucose modulates budding pattern in haploid cells. MATERIALS AND METHODS : Strains, Plasmids, and Microbiological Techniques | The yeast strains used in this study are listed in Table . All of the strains were derived from HYL333 and HYL334 of the filamentous Sigma1978b background (provided by G. Fink, Whitehead Institute for Biomedical Research, Cambridge, MA); these strains exhibit particularly robust filamentous growth compared with other strains from the Sigma1978b background (H. Madhani, UCSF, San Francisco, CA; personal communication). To construct SY3687 and SY3688, HYL334 was made his3::URA3 or leu2::URA3, by using a polymerase chain reaction (PCR)-based method (, and references therein) and plasmid pRS306 as a template . The resulting strains were then made Ura3- by selection on 5-fluoroorotic acid (Biovectra, Oxford, CT). Disruption of BNI1 was performed using plasmid p321, provided by C. Boone . Disruptions of PEA2 and SPA2 were performed using plasmids pNV44 and p210, provided by I. Herskowitz . The plasmid used to disrupt GRR1, pBM2101, was provided by M. Johnston . Other gene disruptions were performed by PCR-based methods (, and references therein) to remove the entire open reading frame with plasmids described by , or other plasmids containing auxotrophic markers from Candida glabrata (for LEU2 and HIS3) and Kluyveromyces lactis (for URA3) and that were provided by I. Herskowitz. Integrated GFP fusions and GAL1 promoter fusions were made by PCR-based methods with plasmids provided by J. Pringle . Gene disruptions and integrated promoter and protein fusions were confirmed by PCR analysis and by phenotype. All of the GFP- and hemagglutinin (HA)-tagged fusion proteins used in this study were functional with respect to bud site selection and invasive growth phenotypes. Table 1 | Yeast strains Yeast and bacterial strains were propagated using standard methods . Yeast peptone dextrose (YPD) and synthetic complete dextrose (SCD) media have been described previously . Yeast transformations were performed as described previously . Bacterial transformations, bacterial DNA preparations, and plasmid constructions were performed by standard methods . Genes controlled by a galactose-inducible promoter were induced in SC or YP medium containing 2% galactose (Gal) as indicated. Geneticin (Biovectra) selection was performed as described previously . Protein Localization | The localization of Bud8p was determined using plasmid YEpGFP*-BUD8 (provided by J. Pringle; ), in which the GFP-Bud8p fusion was expressed from its own promoter. Wild-type cells containing YEpGFP*-BUD8 were grown in synthetic medium lacking leucine (SCD-LEU) to stationary phase and spread onto SCD-LEU or SC-LEU medium for 16 h at 25C. A coverslip was placed directly onto the plates, and GFP-Bud8p was visualized by fluorescence microscopy with a fluorescein isothiocyanate (FITC) filter at 100x. The localization of GFP-Bud2p was determined using plasmid pHP726 (provided by H.-O. Park; ) carried in bud2 and bud2 bud8 strains. The localizations of Bud3p, Bud4p, and Axl2p were determined using C-terminal GFP fusions that were integrated into the genome. Actin staining was performed as described previously . Cells were incubated in SCD or SC medium and fixed in 3.7% formaldehyde for 1 h. Fixed cells were incubated with rhodamine (Rh)-phalloidin (Molecular Probes, Eugene, OR). Cells were washed twice and visualized by florescence microscopy at 100x by using an Rh filter. Invasive Growth Assays | The single cell invasive growth assay was performed as described previously . For some experiments, cells were scraped from plates by using 4 ml of distilled water, concentrated by centrifugation, resuspended in 20 mul of water, and visualized by microscopy. For other experiments, a coverslip was placed directly on the agar medium and cells were visualized directly by microscopy. The plate-washing assay was performed essentially as described previously . Equal concentrations of cells were spotted onto YPD or YPGal medium as specified, invasion was allowed to proceed for 2 d at 30C, and then plates were washed vigorously with water and rubbed with a wet finger to remove cells that did not invade the agar. In some cases, invasion was allowed to proceed for 5 d, during which time the cells became more than twice as elongated as observed in the single cell assay. Cell-cell adhesion was assessed by a standard flocculation assay, as described previously . Microscopy | Differential interference contrast (DIC) and fluorescence microscopy with Rh and FITC filter sets were performed using an Axioplan 2 microscope (Zeiss, Jena, Germany), a black-and-white Orca II digital camera (Hamamatsu, San Jose, CA), and the Openlab software program (Improvision, Coventry, UK). Only brightness and contrast digital adjustments were performed on photographs. Budding Pattern Analysis | Budding pattern determination was performed as described previously , with the following modifications. Equal concentrations of cells were spotted onto YPD medium and incubated for 2 d at 30C. Plates were washed, and invaded cells were excised from the agar by using a toothpick. Cells were resuspended in water containing 1 mug/ml calcofluor (Sigma-Aldrich, St. Louis, MO), and after a 10-min incubation, bud scars were visualized directly by fluorescence microscopy. The enhanced cohesion of cells in the filamentous background facilitated the distinction between proximal and distal bud scars by their position relative to the cell-cell orientation. A bud scar was scored as distal if it was at the pole opposite to the birth scar, or if it was present at the distal pole of a cell that comprised a filament whose growth direction was obvious. A bud scar was scored as proximal if it was at the same pole as the birth scar or at the same pole as the attached parent cell. Bud scars in the middle third of the cell were scored as equatorial. At least 200 bud scars were scored for each experiment. Previously, wild-type cells in glucose-limiting conditions were shown to bud at the distal pole for 95% of all first buds . Subsequent buds were more frequently observed at the proximal pole. In the bud scar counts in the present work, all budding events were considered, resulting in the lower percentage of bud scars observed at the distal pole (similar70%). Budding patterns were corroborated by using the single cell invasive growth assay . Equal concentrations of cells were spread onto SCD or SC medium, and budding pattern was assessed directly by microscopic examination. Microcolonies at the 10-cell stage or less were chosen for analysis. For some experiments cells were placed onto SCD or SC medium by micromanipulation and allowed to grow to the 10-cell stage, which showed the exact lineage of cells within the microcolony. The precise position of bud placement was determined for a subset of experiments by photographing cells and aligning the photographs to an arbitrary model cell. Western Blot Analysis and Determination of Axl1p Abundance | Western blots were performed as described previously . Proteins were separated by 10% SDS-PAGE, transferred to nitrocellulose, and visualized by probing with antibodies specific to GFP (Roche Applied Science, Indianapolis, IN), HA, or Dpm1p (provided by Tom Stevens, Institute of Molecular Biology, University of Oregon, Eugene, OR), which served as a loading control. Band intensity was determined using ImageQuant software (Amersham, Piscataway, NJ), and, where indicated, the values reported were normalized to Dpm1p levels. The abundance of Axl1p was measured using plasmid p151 (provided by C. Boone; ), which expresses a functional Axl1p-HA fusion protein expressed from the AXL1 promoter. In some experiments, Axl1p-HA abundance in yeast-form and filamentous cells was determined by incubating wild-type cells containing p151 (SY3718) on SCD-URA or SC-URA solid agar medium. Cells were harvested from plates, resuspended in water, and adjusted to equal density by measuring optical density. Proteins were then extracted and subjected to Western analysis. In other experiments, Axl1p-HA abundance was measured through the course of a growth cycle in cells incubated in SD-URA liquid medium for various times. In addition, the glucose-limited disappearance of Axl1p-HA was measured in p151-containing wild-type (SY3718) and snf1 mutant (SY3720) cells. Cells were grown to early log phase in liquid SCD-URA medium at 30C, and each culture was split, washed twice with water, and incubated in liquid SC-URA or SCD-URA medium prewarmed to 30C for various times. RESULTS : Bud8p Is Required for Haploid Invasive Growth, whereas Bud9p Impedes Invasion | Disruption of BUD8 in a haploid strain of the filamentous background caused an invasive growth defect in the plate-washing assay (Figure A). In addition, both the single cell invasive growth assay and bud scar staining demonstrated that the bud8 mutant was defective in budding at the distal pole (Figure B; Table ). The mutant cells budded at the distal pole at a frequency of only 11%, in contrast to 70% for wild-type cells. The bud8 mutant cells were as elongated as wild-type cells (Figure C), suggesting that in the bud8 mutant, apical growth and the selection of budding sites were independent. The morphology of the bud8 mutant microcolony was a rosette, a morphology that contrasted strikingly with the linear form of wild-type filamentous cells (Figure C). Figure 1 | Bud8p is required for agar invasion and distal pole budding in haploid cells; Bud9p impedes invasion. Bud8p is required for agar invasion and distal pole budding in haploid cells; Bud9p impedes invasion. (A) Plate washing assay. Equal concentrations of wild-type (SY3687), bud8 (SY3689), and bud9 (SY3692) cells were spotted onto YPD medium and incubated for 2 d at 30C. The plate was photographed (left), washed, and photographed again (right). (B) Single cell invasive growth assay. Equal concentrations of wild-type (top) or bud8 mutant (bottom) cells were spread onto SC medium, incubated for 16 h at 25C, scraped from the plates, and photographed. Arrows indicate the first bud produced by the first daughter cell. (C) Prolonged incubation illustrates the difference between wild-type (left) and the bud8 mutant (right). Equal concentrations of cells were spotted onto YPD medium and grown for 5 d at 30C. The lower right panel shows bud scar (calcofluor) staining for the bud8 mutant . Bars, 5 mum. Table 2 | Budding patterns of mutants known to have defects in bipolar bud site selection during haploid invasive growth Disruption of BUD7 caused an invasive growth defect similar to that of the bud8 mutant in the plate-washing assay, although the bud7 mutant was slightly more invasive (our unpublished data). The bud7 mutant also exhibited a distal pole bud site selection defect . Disruption of bud7 and bud8 together had an invasive growth defect equivalent to either single mutant, and the double mutant had a similar (although slightly more severe) budding-pattern defect than either single mutant (Table ; our unpublished data). Taken together, these data suggest that Bud7p and Bud8p may be components of the same genetic pathway. In contrast to the noninvasive phenotype of bud7 and bud8 mutants, the bud9 mutant exhibited hyperinvasive growth (Figure A). The bud9 mutant also had a higher percentage of distal pole buds compared with wild-type cells , which may account for its hyperinvasive growth phenotype. Disruption of BUD8 in the bud9 mutant caused invasive-growth and distal-pole budding defects equivalent to those of the bud8 single mutant , consistent with the budding pattern observed in bud8 bud9 homozygous diploid cells during vegetative growth . Disruption of BUD7 also suppressed the hyperinvasive growth of the bud9 mutant but to a lesser extent than did disruption of BUD8, consistent with the phenotypes of the bud7 and bud8 single mutants . In glucose-rich conditions, haploid bud7, bud8, and bud9 mutants did not show a budding pattern defect (; ; our unpublished data). In summary, genes identified by virtue of their role in diploid cell budding pattern determination also have a role in haploid cells, specifically during invasive growth that occurs under glucose limitation. Bud8p Is Localized to Distal Pole in Haploid Cells | Bud8p is localized to the distal pole of diploid cells . We examined the localization of Bud8p in haploid cells by using a plasmid containing a functional GFP-Bud8p fusion under the control of the BUD8 promoter . GFP-Bud8p was observed at the distal pole of haploid cells grown in glucose-limiting conditions . GFP-Bud8p was also observed at the distal pole in glucose-rich conditions , a situation in which distal pole budding does not occur. Western blot analysis confirmed that the level of Bud8p was equivalent in glucose-rich and limiting conditions (our unpublished data). Thus, the Bud8p at the distal pole of haploid cells is apparently recognized only under glucose-limiting conditions. Figure 2 | Bud8p is localized to the distal tip of haploid cells, and its localization is dependent upon Bud6p and Bni1p. Bud8p is localized to the distal tip of haploid cells, and its localization is dependent upon Bud6p and Bni1p. Strain backgrounds (wild-type, SY3695; pea2, SY3698; bud6, SY3696; and bni1, SY3697) are as shown. Growth on SCD (+Glu) or SC medium (-Glu) is as indicated. Left, DIC. Right, FITC filter. All pictures were taken at the same scale; bar, 5 mum. In diploid cells, Bud8p's localization to the distal pole is dependent upon Bni1p . We examined the localization of Bud8p in bni1 and related mutants in haploid cells. Distal-pole localization of GFP-Bud8p was not observed in the bni1 mutant (<0.2% of cells had GFP-Bud8p at the distal pole compared with >50% for wild-type cells), but GFP-Bud8p was observed throughout the cell periphery . In the bud6 mutant, distal-pole localization of GFP-Bud8p was observed in a lower percentage of cells than wild type (5% of cells had GFP-Bud8p at the distal pole), and in bud6 cells in which GFP-Bud8p was at the distal pole, the fluorescence intensity was reduced . The abundance of GFP-Bud8p was equivalent in bud6, bni1, and wild-type cells (by Western blot). Consistent with the peripheral localization pattern of Bud8p, bud scar staining of invaded cells showed that bud6 and bni1 mutants had random budding patterns . No budding pattern defects were detected in the mutants in glucose-rich conditions, under which Bud8p was also mislocalized; the cells showed normal axial budding. The Pea2p and Spa2p proteins were not required for distal-pole localization of Bud8p (Figure ; our unpublished data), and pea2 and spa2 mutants maintained the unipolar budding pattern . However, the pea2, spa2, bud6, and bni1 mutants were all defective in the extended apical growth that results in elongated cells during haploid invasion . Consequently, the four mutants all exhibited an invasive growth defect. Bud8p Is Required for Distal Pole Localization of Bud2p and Actin in Glucose-limiting Conditions | Bud2p, the GTPase-activating protein for Rsr1p, has been shown to localize to the incipient bud site, where it presumably recruits Rsr1p to the bud site . In diploid cells, this localization is dependent upon Bud8p . We examined the localization of a functional GFP-Bud2p fusion (provided by Hay-Oak Park, Ohio State University, Columbus, OH) expressed from a high-copy plasmid in bud2 and bud2 bud8 strains. In glucose-limiting conditions, Bud2p was observed at the distal pole of the cell, directly underneath the emerging bud, and at the mother-bud neck of small distal buds . In contrast, GFP-Bud2p localization in a bud2 bud8 mutant was mostly at the proximal pole , a result consistent with the proximal budding pattern observed in the bud8 mutant. In glucose-rich conditions, Bud2p was observed adjacent to the previous bud site (our unpublished observations; ). Figure 3 | Bud8p is required for localization of Bud2p and actin to the distal pole in glucose-limiting conditions. Bud8p is required for localization of Bud2p and actin to the distal pole in glucose-limiting conditions. Left four panels, GFP-Bud2p localization. bud2 (top, SY3707) and bud2 bud8 (bottom, SY3708) strains containing a GFP-BUD2 fusion plasmid were grown in glucose-limiting conditions. Cells were photographed using DIC (left) or FITC filters (right). Right four panels, actin localization. Wild-type (SY3687) and bud8 mutant (SY3689) cells grown in glucose-limiting conditions were stained with rhodamine-phalloidin and photographed using DIC (left) or Rh filters (right). Arrows indicate regions of highly localized actin patches. All pictures were taken at the same scale; bar, 5 mum. Bud site selection components (e.g., Bud8p and Bud2p) are known to recruit actin to the incipient bud site, an event required for bud emergence . We examined actin localization in filamentous cells and found that in glucose-limiting conditions, actin localized to the distal tip of daughter cells . In contrast, actin accumulation was observed at the proximal pole in the bud8 mutant and in wild-type cells grown in glucose-rich conditions (our unpublished data). Thus, glucose limitation caused the localization of Bud2p and bud emergence machinery (e.g., actin) to the distal pole of the cell, an event that was dependent upon Bud8p. Proteins Required for Axial Budding Are Localized Appropriately in Filamentous Cells | We hypothesized that the disappearance of an axial cue in glucose-limiting conditions might result in Bud8p-dependent distal pole budding. In particular, the axial cues Bud3p and Bud4p are transient and are reported to disappear in nutrient-limiting conditions, because they are cell cycle regulated and absent in the Go phase of the cell cycle . The localization of the known axial cues (Bud3p, Bud4p, and Axl2p) was examined in cells grown in both glucose-rich and glucose-limiting conditions. Functional Bud3p-GFP, Bud4p-GFP, and Axl2p-GFP fusions were expressed from chromosomal loci and were found to be localized to the mother-bud neck in cells grown under both conditions (Figure A). Thus, an explanation other than axial cue disappearance must be invoked to explain the change in budding pattern to Bud8p-dependent bud site selection. Figure 4 | Localization and role of axial cues in haploid filamentous cells. Localization and role of axial cues in haploid filamentous cells. (A) Localization of axial cues in cells grown in glucose-rich and glucose-limiting conditions. Equal concentrations of cells (Bud3p-GFP, SY3709; Bud4p-GFP, SY3710; and Axl2p-GFP, SY3711) were spread onto either SCD (+Glu) or SC (-Glu) medium, incubated for 16 h at 25C, scraped from plates, and photographed using DIC (left) or FITC filters (right). Bar (all panels), 5 mum. (B) Loss of axial cues causes hyperinvasive growth. Strains were spotted onto YPD medium and incubated for 2 d at 30C. The plate was photographed (left), washed, and photographed again (right). Strains: wt (SY3687), bud3 (SY3712), bud4 (SY3713), and axl2 (SY3714). (C) Suppression of the bud8 invasive growth defect by disruption of axial cues. Strains were spotted onto YPD medium and incubated for 2 d at 30C. The plate was photographed (left), washed, and photographed again (right). Strains: wt (SY3687), bud8 (SY3689), bud8 bud3 (SY3715), bud8 bud4 (SY3716), and bud8 axl2 (SY3717). Although Bud8p is the primary bud site cue in haploid cells undergoing filamentous growth, genetic evidence suggests that axial cues are used to some degree. First, disruption of BUD3, BUD4, or AXL2 caused hyperinvasive growth (Figure B), due to a significant decrease in bud site selection at the proximal pole . Second, disruption of axial cues in the bud8 mutant ablated the proximal budding , correlating with partial suppression of the invasive growth defect (Figure C). The increased percentage of distal pole budding in the axl2 bud8 mutant, compared with the bud3 bud8 and bud4 bud8 mutants, may be due to the enhanced apical growth that was observed in axl2 mutant cells (our unpublished data). Table 3 | Budding patterns of mutants lacking axial-specific cues or both axial cues and Bud8p Control of Axl1p Abundance by Glucose and Snf1p | Because AXL1 is transcriptionally repressed in diploid cells, an event sufficient to prevent axial budding in wild-type cells , we considered the possibility that Axl1p protein abundance is regulated by glucose in haploid cells. Axl1p protein levels were measured in cells expressing an Axl1p-HA fusion from the AXL1 promoter (provided by C. Boone). We initially observed that Axl1p-HA was not present in filamentous cells (Figure A). Axl1p-HA abundance was examined in cells grown throughout a culture growth cycle. In reference to a control protein, the amount of Axl1p-HA increased steadily during early log phase and was highest in mid-log phase (Figure B). Axl1p-HA abundance declined as growth rate slowed and was significantly reduced in stationary phase (Figure B). To confirm that glucose influenced Axl1p abundance, cells containing the Axl1p-HA fusion were grown to early log phase and shifted to medium lacking glucose. As expected, the level of Axl1p-HA declined markedly upon the shift to glucose-limited medium (Figure C). Figure 5 | Axl1p abundance is controlled by glucose and Snf1p, and Axl1p is an inhibitor of invasive growth. Axl1p abundance is controlled by glucose and Snf1p, and Axl1p is an inhibitor of invasive growth. (A) Axl1p is absent in filamentous cells. Western blot of protein extracts from wild-type cells containing p151 (SY3718) incubated on SCD-URA (+Glu) or SC-URA (-Glu) solid agar medium for 16 h at 25C and probed with antibodies against HA (to detect Axl1p-HA) or Dpm1p (see MATERIALS AND METHODS). (B) Abundance of Axl1p correlates with the culture growth cycle. Wild-type cells containing p151 (SY3718) were grown through a growth cycle in SCD-URA liquid medium at 30C, and protein extracts were harvested from cells at the times indicated. Quantitation of Western blots of Axl1p levels (adjusted to Dpm1p levels) is shown (filled circles). Optical density at 600 nm is also shown (open squares). (C) Decrease in Axl1p abundance upon a shift to glucose-limited medium is dependent upon Snf1p. Wild-type cells containing p151 (SY3718) were grown to early log phase and shifted to SCD-URA (open squares, +Glu) or SC-URA (open triangles, -Glu) medium. A snf1 mutant containing p151 (SY3720) was grown in the same way (closed circles, -Glu). Quantitation of Western blots of Axl1p levels (adjusted to Dpm1p levels) is shown. (D) Hyperinvasive growth in axl1 mutants. Equal concentrations of wild-type (SY3687), axl1 (SY3721), axl1 bud8 (SY3722), and bud8 (SY3689) cells were spotted onto YPD medium and incubated for 2 d at 30C. The plate was photographed (left), washed, and photographed again (right). The colonies shown are all from the same plate. (E) Overproduction of Axl1p prevents invasive growth. Wild-type (SY3687) and GAL1-AXL1 ( AXL1, SY3723) cells were grown to saturation in YPGal medium, and equal concentrations of cells were spotted onto YPGal medium for 2 d at 30C. The plates were photographed (left), washed, and photographed again (right). The Snf1p protein kinase is a global regulator of glucose response , and we showed previously that Snf1p is required for unipolar budding during haploid invasive growth . We investigated the role of Snf1p in the glucose-dependent regulation of Axl1p. In contrast to the observations reported above for wild-type cells, Axl1p-HA protein abundance remained high in a snf1 mutant after a shift to glucose-limited medium (Figure C). Thus, Snf1p is required for the disappearance of Axl1p in glucose-limiting conditions. Genetic analysis also supports a role for Axl1p in the transition to filamentous growth. Disruption of AXL1 caused hyperinvasive growth (Figure D) due to distal pole budding in both glucose-rich and glucose-limiting conditions . These phenotypes were largely suppressed by disruption of BUD8 (Figure D and Table ). The axl1 bud8 double mutant invades better than the bud8 single mutant because it has constitutive nonaxial budding, whereas the bud8 single mutant buds almost exclusively from the proximal pole under both glucose-rich and glucose-limiting conditions. In contrast, overexpression of AXL1 suppressed agar invasion by wild-type cells (Figure E) due to an increase in proximal budding . Thus, the disappearance of the Axl1p protein in glucose-limiting conditions is sufficient to explain the Bud8p-dependent budding during haploid invasive growth. Table 4 | Budding patterns of strains deleted for AXL1 and/or BUD8 or overexpressing AXL1 Unipolar Budding in rsr1 Mutant Due to Increased Apical Growth during Haploid Invasion | It has been reported that ablation of the general bud-site-selection machinery does not disrupt the ability of haploid cells to undergo agar invasion . This result is seemingly at odds with our finding that mutants that are defective for distal-pole budding cannot invade agar. Therefore, we characterized the rsr1 mutant, which is lacking the core bud-site-selection GTPase, in detail. In the single cell invasive growth assay, the rsr1 mutant formed filaments composed of elongated cells emanating away from the mother cell, suggesting a unipolar-distal pattern (Figure A); however, buds emerging from the equatorial regions of cells were also observed (Figure A, black arrows). To determine more precisely the budding pattern of first and second buds for the rsr1 mutant, cells were placed onto glucose-limited medium by micromanipulation and budding pattern was assessed by microscopic examination. Strikingly, the first bud produced was at the distal pole >90% of the time . The second buds emerged uniformly around the entire surface of the cell . In glucose-rich conditions, a less dramatic distal-pole bias was observed in the rsr1 mutant , consistent with previous reports . Figure 6 | Distal pole budding in the rsr1 mutant during haploid invasive growth. Distal pole budding in the rsr1 mutant during haploid invasive growth. (A) Wild-type (SY3687) and rsr1 mutant (SY3724) cells were spread onto SC medium, incubated for 16 h at 25C, and photographed. Bar, 5 mum. (B) Diagram of precise position of buds as determined by microscopic examination. Microcolonies at the 10-cell stage (or less) were visualized by light microscopy and photographed. Photographs were analyzed for bud position; only buds whose precise locations were unambiguous were chosen for analysis. The position of 50 buds is shown for each depiction. (C) Bud8p is not required for agar-invasion in the rsr1 mutant. Wild-type (SY3687), bud8 (SY3689), rsr1 (SY3724), and rsr1 bud8 mutant (SY3725) cells were spotted onto YPD medium and grown for 2 d at 30C. Plates were photographed (top), washed, and photographed again (bottom). Table 5 | Bud position of the rsr1 mutant grown on glucose-rich or glucose-limited medium We hypothesized that the distal-pole budding observed in the rsr1 mutant was due to the lengthened period of apical growth that leads to cell elongation during haploid invasion . Precise mapping of bud placement showed that in the rsr1 mutant, buds emerged within an arc that included the distal pole, whereas in wild-type cells, bud emergence was confined to the extreme tip of the cell (Figure B), suggesting that the mechanism of distal budding in the rsr1 mutant was different than in wild-type cells. Indeed, disruption of BUD8 did not affect distal-pole budding or agar invasion of the rsr1 mutant (Figure C), demonstrating that the distal-pole budding in the rsr1 mutant was not due to Bud8p-dependent bud site selection. To gain further support for the idea that enhanced apical growth can influence budding pattern in the rsr1 mutant, we used the grr1 mutation , which causes hyperpolarized growth . grr1 mutant cells were elongated in glucose-rich conditions, but their budding pattern was axial . The rsr1 grr1 double mutant, however, had a clear distal pole bias compared with the rsr1 single mutant in glucose-rich medium , supporting the idea that in the rsr1 mutant, hyperpolarized growth leads to distal-pole budding. Contribution of Different Aspects of Filamentous Growth to Agar Invasion | Filamentous growth is characterized by several physiological events: a change in budding pattern, an increase in cell length, and enhanced cell-cell adhesion. We found that disruption of FLO11, which is required for cell-cell adhesion and haploid invasive growth , did not affect cell elongation or unipolar-distal budding (Figure A), implicating cell-cell adhesion as the primary defect in the flo11 mutant. We directly compared mutants defective primarily in a single aspect of filamentous growth to a ste20 mutant, which is defective in all three aspects. In particular, we compared a bud8 mutant (defective for distal-pole budding but not elongation or adhesion), a pea2 mutant (defective for elongation but not distal-pole budding or adhesion), and a flo11 mutant (defective for adhesion but not distal-pole budding or elongation). A defect in any single aspect of filamentous growth caused a partial invasive growth defect, although the flo11 mutant was less invasive than the pea2 and bud8 mutants (Figure B). These results were extended by examining the phenotypes of double and triple mutants. The bud8 pea2, bud8 flo11, and pea2 flo11 double mutants were less invasive than any single mutant, and in the pea2 bud8 flo11 triple mutant, no agar invasion was observed (Figure B), implying that the three physiological events affected by these mutations are the major contributors to agar invasion. Figure 7 | Genes involved primarily in a single aspect of filamentous growth contribute independently to agar invasion. Genes involved primarily in a single aspect of filamentous growth contribute independently to agar invasion. (A) Flo11p is not required for unipolar-distal budding or cell elongation. Wild-type (SY3687) and flo11 (SY3729) cells were spread onto SC medium, incubated for 16 h at 25C, and photographed. Bar, 20 mum. (B) Contribution of individual filamentation functions to agar-invasion. Equal concentrations of cells were spotted onto YPD medium and grown for 2 d or 4 d at 30C, as indicated. Plates were photographed, washed, and photographed again. Strains are as indicated: wild-type (SY3687), bud8 (SY3689), pea2 (SY3698), flo11 (SY3729), ste20 (SY3728), and triple, the pea2 bud8 flo11 triple mutant (SY3733). Diploid Cells Exhibit Filamentous Growth in Response to Glucose Depletion | Filamentous growth has been thought to be induced by different cues in haploid cells (glucose limitation; ) and diploid cells (nitrogen limitation; ). We investigated the effect of glucose depletion on diploid cells and found that on glucose-limited medium, diploid cells were more elongated than on glucose-rich medium (Figure A). The elongated phenotype was apparent within the first cell division and was observed for >80% of cells. In addition to the increase in cell length, diploid cells budded in a unipolar-distal pattern upon glucose limitation. Bud scar staining of invaded diploid cells showed unipolar-distal bud scars 85% of the time (15% had scars at both poles); in contrast, diploid cells in glucose-rich medium had unipolar-distal bud scars 51% of the time (45% had scars at both poles and 4% had at least one scar in an equatorial site). Wild-type diploid cells also exhibited robust agar invasion by the plate-washing assay (Figure B). Thus, diploid cells can undergo filamentous growth in response to limiting glucose. Figure 8 | Diploid cells exhibit filamentous growth in response to limiting glucose. Diploid cells exhibit filamentous growth in response to limiting glucose. (A) Single cell assay. Diploid cells (SY3734) were spread onto SC (-Glu) and SCD (+Glu) medium, incubated for 16 h at 25C, and photographed. Bars, 10 mum. (B) Equal concentrations of diploid cells (SY3734) and haploid cells (SY3687) were spotted onto YPD medium and grown for 2 d at 30C. Plates were photographed (left), washed, and photographed again (right). DISCUSSION : Bipolar Bud-Site-Selection Components Are Required for Haploid Invasive Growth | We investigated the role of bud-site-selection components in haploid invasive growth and found that proteins required for bipolar budding in diploid cells were required in haploid cells for distal pole budding during invasive growth. In particular, our data indicate that Bud8p marks the distal pole of haploid cells and is recognized upon glucose depletion to direct budding to the distal pole. Bud8p had not previously been implicated in haploid invasion , perhaps because bud8 mutants exhibit only a partial invasive growth defect. We have shown, however, that Bud8p has a profound role in bud site selection under conditions of glucose limitation, even though it has no known role when glucose is abundant. This contrasts with Bud8p's role in diploid cells, where it controls budding pattern during both yeast form and filamentous growth . Thus, an additional level of regulation is required in haploid cells to prevent Bud8p from directing budding pattern in glucose-rich conditions (see below). Our results also indicate that Bud7p, Bud9p, Bni1p, and Bud6p are required in glucose-limiting conditions for normal haploid invasive growth. As in diploid cells, Bni1p is required in haploid cells to localize Bud8p to the distal pole. We have also found that Bud6p is important, although not essential, for Bud8p localization in haploid cells, an apparent difference from previously reported results with diploid cells ; this may either represent a genuine difference between haploid and diploid cells or just be a function of different strain backgrounds. (Note that previous studies have also shown less precise use of the distal pole in bud6 mutant diploids; ; ; ). Two pieces of evidence in our study support the conclusion that the bipolar budding machinery is in place, albeit dormant, in haploid cells. First, Bud8p is present and localized appropriately in haploid cells in glucose-rich conditions. Second, Bni1p is required for localization of Bud8p to the distal pole in glucose-rich conditions. Because these proteins are not required for bud site selection during yeast-form growth in haploid cells, we suggest that the bipolar budding machinery constitutes a default program in haploid cells that becomes active in glucose-limiting conditions. Indirect evidence from a number of studies has indicated that haploid cells have properly located proximal and distal cues , and we have provided direct evidence for this conclusion. In glucose-rich environments, the axial Bud3p/Bud4p/Axl2p cues are chosen, whereas glucose limitation causes the distal cue Bud8p to be chosen in preference to Bud3p/Bud4p/Axl2p. The presence of both sets of cues enables an individual cell to reorient bud growth rapidly in response to changing nutrient availability. Glucose Controls Bud Site Selection by the Snf1p-dependent Regulation of Axl1p | How is unipolar-distal budding prevented in haploid cells growing in glucose-rich conditions? We show here that the abundance of the axial-promoting factor Axl1p is controlled by glucose. Axl1p levels decline sharply upon glucose limitation , an event that is concomitant with the appearance of unipolar buds (our unpublished data). Thus, it is reasonable to speculate that the disappearance of Axl1p is a trigger for distal-pole budding. We found that the Snf1p protein kinase, which plays a role in the derepression of glucose-repressed genes , is required for the disappearance of Axl1p in glucose-limiting conditions. Snf1p may exert its effect on Axl1p indirectly, by allowing derepression of a gene whose product regulates Axl1p. Alternatively, Snf1p may directly phosphorylate Axl1p and target it for degradation. Irrespective of the mechanism by which Snf1p regulates Axl1p abundance, the genetic evidence that we have presented identifies Axl1p as an important regulator of haploid invasive growth. Loss of Axl1p permits Bud8p-dependent unipolar-distal budding, whereas overexpression of Axl1p suppresses unipolar-distal budding and agar invasion in glucose-limiting conditions. How Axl1p functions to direct budding to sites marked by Bud3p/Bud4p/Axl2p is not known but is a question that is crucial to the ultimate understanding of bud site determination in yeast. Coordination of Unipolar Budding and Polarized Growth | A lengthened period of polarized, apical growth promotes bipolar budding in diploid cells and also promotes unipolar-distal budding in haploid cells during invasive growth . In fact, in this latter case, we showed that an extended period of apical growth could confer distal-pole budding to a mutant lacking a functional bud-site-selection system (rsr1 mutant). In wild-type cells, polarized growth and localization of the bud site machinery are coordinated. For example, Bni1p, which is involved in polarized growth, is also required for localization of Bud8p to the distal tip of the daughter cell in both haploids and diploids . These two processes are not inextricably linked, however. Pea2p (and Spa2p) were shown to be required for cell elongation during invasive growth but not for Bud8p localization or distal-pole budding. Conversely, loss of Bud8p disrupted the budding pattern but did not affect cell elongation. Thus, two distinct cues mark the distal pole of daughter cells: a bud site cue (Bud8p) to direct bud site selection and a second cue to direct polarized growth. The notion that the components of invasive growth could be genetically isolated was extended to cell-cell adhesion. Disruption of FLO11 prevented adhesion but had no effect on cell elongation or the budding pattern. Thus, haploid invasive growth can be divided into three separate processes: cell elongation, unipolar-distal budding, and cell adhesion. Loss of any one of these processes partially compromises invasive growth; loss of all three prevents it entirely. Diploid Cells Starved for Glucose Initiate the Filamentous Growth Response | Filamentous growth was first characterized in diploid cells and was described as a response to nitrogen limitation . Mutations in nitrogen-sensing and nitrogen utilization pathways confirmed that pseudohyphal growth was a response to low levels of environmental fixed nitrogen . Subsequently, haploid cells were also shown to undergo a filamentation-like process in which they invaded the agar substratum . Diploid pseudohyphal growth and haploid invasive growth were presumed to be similar processes because they required some of the same signal transduction pathways, but differences are apparent, in particular in the degree of agar invasiveness. Recently, we showed that glucose depletion was a trigger for haploid invasive growth . Hence, it was proposed that diploid cells initiate filamentous growth in response to limiting nitrogen, whereas glucose depletion triggers haploid invasive growth . However, we have shown here that diploid cells manifest all of the characteristics of filamentous growth in response to glucose limitation. Perhaps variation within the Sigma1278b background has caused confusion as to the cues that underlie filamentous growth. For example, strains of yeast capable of starch degradation have been reported to initiate filamentous growth upon either carbon or nitrogen source depletion . Diploid cells sporulate upon limitation of both carbon and nitrogen cues ; depletion of either single cue, however, triggers filamentous growth. Backmatter: Abbreviations used: : DIC = differential interference contrast FITC = fluorescein GFP = green fluorescent protein HA = hemagglutinin Rh = rhodamine SC = synthetic complete URA = uracil YPD = yeast peptone dextrose wt = wild-type PMID- 12221112 TI - Pkh1 and Pkh2 Differentially Phosphorylate and Activate Ypk1 and Ykr2 and Define Protein Kinase Modules Required for Maintenance of Cell Wall Integrity AB - Saccharomyces cerevisiae Pkh1 and Pkh2 are functionally redundant homologs of mammalian protein kinase, phosphoinositide-dependent protein kinase-1. They activate two closely related, functionally redundant enzymes, Ypk1 and Ykr2 (homologs of mammalian protein kinase, serum- and glucocorticoid-inducible protein kinase). We found that Ypk1 has a more prominent role than Ykr2 in mediating their shared essential function. Considerable evidence demonstrated that Pkh1 preferentially activates Ypk1, whereas Pkh2 preferentially activates Ykr2. Loss of Pkh1 (but not Pkh2) reduced Ypk1 activity; conversely, Pkh1 overexpression increased Ypk1 activity more than Pkh2 overexpression. Loss of Pkh2 reduced Ykr2 activity; correspondingly, Pkh2 overexpression increased Ykr2 activity more than Pkh1 overexpression. When overexpressed, a catalytically active C-terminal fragment (kinase domain) of Ypk1 was growth inhibitory; loss of Pkh1 (but not Pkh2) alleviated toxicity. Loss of Pkh2 (but not Pkh1) exacerbated the slow growth phenotype of a ypk1Delta strain. This Pkh1-Ypk1 and Pkh2-Ykr2 dichotomy is not absolute because all double mutants (pkh1Delta ypk1Delta, pkh2Delta ypk1Delta, pkh1Delta ykr2Delta, and pkh2Delta ykr2Delta) were viable. Compartmentation contributes to selectivity because Pkh1 and Ypk1 were located exclusively in the cytosol, whereas Pkh2 and Ykr2 entered the nucleus. At restrictive temperature, ypk1-1tsykr2Delta cells lysed rapidly, but not in medium containing osmotic support. Dosage and extragenic suppressors were selected. Overexpression of Exg1 (major exoglucanase), or loss of Kex2 (endoprotease involved in Exg1 processing), rescued growth at high temperature. Viability was also maintained by PKC1 overexpression or an activated allele of the downstream protein kinase (BCK1-20). Conversely, absence of Mpk1 (distal mitogen-activated protein kinase of the PKC1 pathway) was lethal in ypk1-1tsykr2Delta cells. Thus, Pkh1-Ypk1 and Pkh2-Ykr2 function in a novel pathway for cell wall integrity that acts in parallel with the Pkc1-dependent pathway. Keywords: INTRODUCTION : A cascade of protein kinases is a commonly used mechanism for amplifying and disseminating signals that control metabolism, growth, survival, and differentiation in eukaryotic cells. In animal cells, recruitment of phosphatidylinositol 3-kinase by growth factor receptors generates 3-phosphoinositides, which stimulate 3-phosphoinositide-dependent protein kinase-1 (PDK1) (for review, see ; ). Activated PDK1 phosphorylates and activates multiple downstream targets, including protein kinase B/c-Akt , p70 S6 kinase , protein kinase C (PKC) isoforms , and serum- and glucocorticoid-inducible protein kinase (SGK) isoforms , thereby eliciting physiological responses. We have demonstrated previously that, in budding yeast (Saccharomyces cerevisiae), Pkh1 and Pkh2 are the homologs and functional equivalents of mammalian PDK1. Pkh1 and Pkh2 share an essential function because pkh1Delta and pkh2Delta single mutants are viable, whereas a pkh1Delta pkh2Delta double mutant is inviable. Expression of human PDK1 rescues the lethality of a pkh1Delta pkh2Delta strain . The PDK1 enzymes from Caenorhabditis elegans, Drosophila melanogaster, and Homo sapiens all possess a C-terminal pleckstrin homology (PH) domain that binds phosphatidylinositol (PtdIns)(3,4,5)P3 and PtdIns(3,4)P2 . However, S. cerevisiae does not produce PtdIns(3,4,5)P3 or PtdIns(3,4)P2 . Moreover, Pkh1 and Pkh2 lack discernible PH domains, and PDK1 lacking its PH domain was sufficient to rescue the growth of pkh1Delta pkh2Delta cells , suggesting that the activity of Pkh1 and Pkh2 in yeast does not depend on phosphoinositides. It was shown subsequently that sphingosine (4-dehydro-sphinganine) can also stimulate mammalian PDK1 autophosphorylation and increase its ability to phosphorylate in vitro known PDK1 substrates, such as c-Akt and PKCbeta . Correspondingly, it has been reported recently that Pkh1 and Pkh2 can be activated in vitro by nanomolar concentrations of the major sphingoid base in yeast, phytosphingosine (4-hydroxy-sphinganine) . Moreover, endocytosis in yeast seems to require sphingoid base synthesis and overexpression of Pkh1 or Pkh2 can suppress this requirement , suggesting that sphingoid bases activate a signaling pathway involving Pkh1 and Pkh2. Mammalian PDK1 activates its downstream targets by phosphorylating a Thr residue (starred) in a sequence motif, Thr*-Phe-Cys-Gly-Thr-X-Glu-Tyr (where X represents any amino acid), that lies within the "activation loop" of their catalytic domains and is unique to and conserved in all known PDK1 substrates. Full activation of c-Akt/PKB and other PDK1 targets also seems to require phosphorylation at a second site (starred) situated in a hydrophobic motif, Phe-X-X-Ar-Ser*/Thr*-Ar (where Ar represents an aromatic residue), that is located near the C terminus of each of these enzymes . In S. cerevisiae, four previously characterized protein kinases possess both of these motifs, suggesting that they are physiological substrates of Pkh1 and/or Pkh2. These four protein kinases are the products of the following genes: YPK1 , YKR2/YPK2 , PKC1 , and SCH9 . Studies from this laboratory have demonstrated that Ypk1 is a direct substrate of Pkh1 and that Ykr2 is phosphorylated by Pkh2 . Similarly, it has been shown that Pkc1 can also be phosphorylated by Pkh1 and Pkh2 . Reduced Pkc1 activity was observed in a pkh1-1ts pkh2Delta strain, and the temperature sensitivity of this strain was partially suppressed by a dominant PKC1(R398P) allele, suggesting that Pkh1 and Pkh2 are required for Pkc1 function in vivo . The catalytic domains of Ypk1 and Ykr2 are 88% identical and these proteins also share extensive homology across their N- and C-terminal extensions. Moreover, the catalytic domains of Ypk1 and Ykr2 closely resemble (55% identity) that of mammalian SGK. Indeed, cells lacking Ypk1 or Ykr2 are viable, whereas cells lacking both Ypk1 and Ykr2 are inviable , and expression of mammalian SGK rescues this inviability . Furthermore, both purified PDK1 and purified Pkh1 phosphorylate the same residue (Thr504) in the consensus motif in purified Ypk1, and Ypk1 phosphorylation is significantly diminished in vivo in cells lacking Pkh1 . Thus, just as SGK is a downstream target of PDK1 in animal cells, Ypk1 and Ykr2 seem to act downstream of Pkh1 and Pkh2 in yeast. Moreover, lipid-derived signals are required as upstream activators in both pathways, 3-phosphoinositides and sphingosine in the case of PDK1 and closely related sphingoid bases in the case of Pkh1 and Pkh2. Consistent with this view, overexpression of Ypk1 confers resistance to myriocin (ISP-1), an antibiotic that specifically inhibits serine C-palmitoyltransferase (product of the LCB1 gene), which is the enzyme responsible for sphinganine biosynthesis . Herein, we describe experiments that address the genetic and biochemical interrelationships between Pkh1 and Pkh2 and Ypk1 and Ykr2, which we undertook to try to understand the reason for the redundancies within these protein kinase cascades. To provide further insight, we also investigated the subcellular localization of all four proteins. Finally, as two independent approaches for discerning the physiological function of the Ypk1 and Ykr2 enzymes, we selected for dosage suppressors and also for chromosomal mutations that suppress the lysis phenotype of ypk1-1ts ykr2Delta cells. MATERIALS AND METHODS : Strains and Growth Conditions | Yeast strains used in this study are listed in Table . Standard rich (YP) and defined minimal (SC) media , containing either 2% glucose (Glc), 2% raffinose (Raf), or 2% galactose (Gal) as the carbon source and supplemented with appropriate nutrients to maintain selection for plasmids, were used for yeast cultivation. For gene expression from the galactose-inducible GAL1 promoter in liquid media, cells were pregrown to mid-exponential phase in SC containing 2% raffinose-0.2% sucrose (Raf/Suc) and then Gal was added to a final concentration of 2% and incubation continued for 2 h. In experiments involving growth on solid medium containing 5-fluoroorotic acid, 5-fluoroorotic acid was used at a concentration of 0.5 mg/ml . Cells were grown routinely at 30C, except for strains carrying temperature-sensitive mutations, which were propagated at their permissive temperature (26C). Table 1 | S. cerevisiae strains used in this study Recombinant DNA Methods | Escherichia coli strain DH5alpha was used for the construction and propagation of plasmids. Conventional recombinant DNA methods were used for the construction of plasmids . The sequences of constructs that contained DNA fragments amplified by polymerase chain reaction (PCR) were verified by the dideoxy chain termination-sequencing method . Native and Turbo Pfu polymerases (Stratagene, La Jolla, CA) were used for PCR, unless noted otherwise. Plasmids | Plasmids pYPK1, pYKR2, pGAL-YPK1, pGAL-YKR2 (pAM1), pGAL-Ypk1-Myc (pAM54), pADH-YPK1, pRS316-YKR2 (pAM12), pGAL-PKH1 (pAM73), and pGAL-PKH2 (pAM79) have been described previously . To create plasmid pADH-YKR2 (pAM4), which constitutively overexpresses YKR2 from the ADH1 promoter, a 2.4-kb XhoI (blunt)-SalI fragment containing the entire YKR2 gene was excised from pYKR2, gel purified, and inserted into vector pAD4 M that had been linearized with SmaI/SalI. To generate a version of Ykr2 tagged at its C-terminal end with the c-Myc epitope , a PCR-based method for precise gene fusion was performed using the YKR2 sequence cloned in pUC18 as one template (pYKR2), and as the other template, pOGFP (E. Swartzman, this laboratory), which contains a sequence encoding the 16-residue version of the c-Myc epitope followed by a (His)6 tag cloned in pBluescript (Stratagene); with three appropriate synthetic oligonucleotide primers: T3 (Stratagene); 5'-GGA CAT ATT GCA CTG TGT G-3' (RMN5), corresponding to sequences in YKR2 overlapping a DraIII site near the C terminus; and a "joiner" primer, 5'-TTC AGA AAT CAA CTT TTG TTC ACT AAT GCT TCT CCC CTG-3' (RMC), corresponding to the 3' end of the YKR2 coding sequence and the first several residues of the c-Myc epitope. An similar1.6-kb DraIII/KpnI fragment of the resulting PCR product was used to replace the corresponding segment in pYKR2, yielding pYkr2-Myc (pAM24). An similar3-kb NcoI/HindIII fragment from pYkr2-Myc was gel purified and used to replace the corresponding similar2.2-kb NcoI/HindIII segment in pGAL-YKR2 to create a 2-mum DNA-containing, LEU2-marked plasmid, pGAL-Ykr2-Myc (pAM59), that overexpresses Ykr2-Myc upon galactose induction. To generate a catalytically inactive ("kinase-dead") version of Ypk1, a PCR-based method for site-directed mutagenesis was performed using pYPK1 as the template and three appropriate synthetic oligonucleotide primers: 5'-CTT GAA CAC AGT AAG TAA CGG-3' (PKC2), corresponding to the flanking genomic sequence commencing 68-base pairs downstream of the stop codon; 5'-CAC AAA AAG TAT ACG CCT TGG CGG CAA TCA G-3' (PKD), where the underlined nucleotide is a silent mutation to introduce a BglI site, and the bold nucleotides correspond to an introduced alanine codon (GCG) in place of the native lysine codon (AAG); and 5'-GTC CAT CGA TGA TTT CGA TC-3' (Pseq2), corresponding to the coding strand of YPK1 starting at nucleotide position 1024. The resulting similar1.1-kb PCR product was digested with ClaI and NcoI, and the resulting similar850-base pair fragment was used to replace the corresponding segment in pYPK1, yielding pYPK1(K376A-KD) (pAM46). Conversion of the Lys residue at the equivalent position in all other protein kinases examined to date eliminates their catalytic activity . To generate a catalytically inactive (kinase-dead) version of Ykr2, a similar PCR-based approach for site-directed mutagenesis was performed using pYKR2 as the template, and three appropriate synthetic oligonucleotide primers: 5'-AGT ATA GCC CTG CCC CAA C-3' (Rseq2), corresponding to the noncoding strand of YKR2 commencing at nucleotide position 1544; 5'-CCC AAA AGA TTT ACG CCT TGG CGG CTC TGA G-3' (RKD), where the underlined nucleotide is a silent mutation to introduce a BglI site, and the bold nucleotides correspond to an introduced alanine codon (GCG) in place of the native lysine codon (AAG); and 5'-CGT GGG GTA ATG GCC TG-3' (Rseq3), corresponding to the coding strand of YKR2 starting at nucleotide position 66. The resulting similar1.4-kb PCR product was digested with NcoI and DraIII and used to replace the corresponding segment in pYKR2, yielding pYKR2(K373A-KD) (pAM47). Plasmid pYPK1(K376A-KD) was digested with AlwnI, converted to flush ends by treatment with T4 polymerase (NEB) and all four dNTPs then digested with SalI. The resulting 3.3-kb YPK1(K376A-KD) --containing fragment was gel purified and ligated into YEp351GAL that had been linearized by digestion with XbaI, converted to flush ends by incubation with T4 polymerase and all four dNTPs, and then digested with SalI. The resulting plasmid, pGAL-YPK1(K376A-KD) (pAM48), expresses a catalytically inactive allele [Ypk1-(K376A-KD)] from a 2-mum DNA-containing, LEU2-marked plasmid under control of the GAL1 promoter. An similar1.2-kb NcoI/SalI fragment from pYpk1-Myc was gel purified and used to replace the corresponding NcoI/SalI segment in pGAL-YPK1(K376A-KD) to create pGAL-YPK1(K376A-KD)-Myc (pAM49). An similar2.4-kb XhoI-HindIII fragment containing the entire YKR2(K373A-KD) allele was excised from pYKR2(K373A-KD), gel purified, and inserted into YEp351GAL that had been linearized with SalI and HindIII, to create a 2-mum DNA-containing, LEU2-marked plasmid, pGAL-YKR2(K373A-KD) (pAM50), that overexpresses catalytically inactive Ykr2 upon galactose induction. Galactose-inducible expression vectors that are URA3 based were constructed as follows. An similar3.3-kb BamHI/HindIII fragment carrying YPK1 was excised from pGAL-YPK1, gel purified, and ligated into YEp352GAL , which had been linearized with BamHI/HindIII, yielding YEp352GAL-YPK1 (pAM75). An similar3.8-kb BamHI/HindIII fragment from pGAL-Ypk1-Myc was gel purified and ligated into YEp352GAL that had been linearized with BamHI/HindIII, yielding YEp352GAL-Ypk1-Myc (pAM76). An similar2.2-kb BamHI/HindIII fragment from p2GAL-YKR2 was gel purified and ligated into YEp352GAL, which had been linearized with BamHI/HindIII, yielding YEp352GAL-YKR2 (pAM77). An similar3.0-kb BamHI/HindIII fragment from p2GAL-Ykr2-Myc was gel purified and ligated into YEp352GAL, which had been linearized with BamHI/HindIII, yielding YEp352GAL-Ykr2-Myc (pAM78). To generate an amino-terminal truncation of Ypk1, the following two-step approach was taken. First, an similar1.1-kb fragment corresponding to the last 344 amino acids of Ypk1 was amplified by PCR from pYPK1 with the following oligonucleotides: 5'-GGC GGA TCCATG TCC AGA AAT AAA CCT TTG TCC-3' (PCT), corresponding to sequences in the middle of the YPK1 coding sequence, just upstream of the beginning of the catalytic domain, where the underlined nucleotides correspond to an introduced BamHI restriction site, and the bold nucleotides represent an introduced start codon (ATG); and 5'-CTT GAA CAC AGT AAG TAA CGG-3' (PKC2), corresponding to the flanking genomic sequence commencing 68-base pairs downstream of the stop codon. The resulting PCR product was digested with BamHI and NcoI, gel purified, and used to replace an similar2.6-kb BamHI/NcoI fragment in pRS315-YPK1(B/H). The resulting CEN-containing, LEU2-marked plasmid encodes an amino-terminal truncation of Ypk1, which contains only the catalytic domain, but essentially no promoter sequence, and is called pRS315-YPK1-DeltaN (pAM55). An similar1.2-kb NcoI/SalI fragment from pRS315-Ypk1-myc was gel purified and used to replace the corresponding NcoI/SalI segment in pRS315-YPK1-DeltaN to create a CEN-containing, LEU2-marked plasmid, pRS315-Ypk1-DeltaN-myc (pAM56), that encodes a myc-tagged version of the Ypk1 catalytic domain. To insert a promoter, an similar2.3-kb BamHI/SalI fragment from pRS315-YPK1-DeltaN was gel purified and inserted into YEp351GAL that had been linearized with BamHI/SalI to create a 2-mum DeltaNA-containing, LEU2-marked plasmid, pGAL-YPK1-DeltaN (pAM99), that overexpresses the amino-terminal truncation of Ypk1 upon galactose induction. Likewise, an similar2.0-kb BamHI/SalI fragment from pRS315-Ypk1-DeltaN-Myc was gel purified and inserted into YEp351GAL that had been linearized with BamHI/SalI to create a 2-mum DeltaNA-containing, LEU2-marked plasmid, pGAL-Ypk1-DeltaN-Myc (pAM100), that overexpresses a myc-tagged version of the amino-terminal truncation of Ypk1 upon galactose induction. To move these truncated Ypk1 derivatives into URA3-marked plasmids, an similar2.3-kb BamHI/SalI fragment from pGAL-YPK1-DeltaN was gel purified and inserted into YEp352GAL that had been linearized with BamHI/SalI to create a 2-mum DNA-containing, URA3-marked plasmid, YEp352GAL-YPK1-DeltaN (pAM101), that overexpresses the amino-terminal truncation of Ypk1 upon galactose induction. Similarly, an similar2.0-kb BamHI/SalI fragment from pGAL-Ypk1-DeltaN-Myc was gel purified and inserted into YEp352GAL that had been linearized with BamHI/SalI to create a 2-mum DNA-containing, URA3-marked plasmid, YEp352GAL-Ypk1-DeltaN-myc (pAM102), that overexpresses a myc-tagged version of the amino-terminal truncation of Ypk1 upon galactose induction. To generate a catalytically inactive derivative of the Ypk1-DeltaN allele, a 2.3-kb ClaI/HindIII fragment from pGAL-YPK1-KD, encoding the carboxy terminus (containing the K376A kinase-dead mutation of Ypk1) was gel purified and used to replace the corresponding segment in YEp352GAL-Ypk1-DeltaN-Myc. The resulting plasmid, YEp352GAL-Ypk1-DeltaN-KD (pFR30), overexpresses a catalytically inactive derivative of the amino-terminal truncation of Ypk1 upon galactose induction. Protein Localization by Using Chimeras Containing Green Fluorescent Protein (GFP) | To create vectors for galactose-inducible expression of YPK1, YKR2, and PKH2, each fused to the carboxy terminus of a protein comprising three tandem repeats of an enhanced (S65T V163A) mutant of GFP, the following approach was taken. Two primers, 5'-GCG AGC GGG ATC CAT G, the first 18 bases of the gene-3' (primer A), where underlined bases correspond to an introduced BamHI site and start codon in bold; and 5'-GGC ACG CGT CGA CTT A, the last 18 bases of the gene-3' (primer B), where underlined bases correspond to an introduced SalI site and stop codon in bold, were used to amplify the entire open reading frames of the corresponding genes from genomic DNA. The PCR products were digested with BamHI and SalI and ligated into vector pGS836 (YCpGAL-3GFP) that had been digested with BamHI and SalI, yielding plasmids pGAL-3GFP-YPK1 (pFR33), pGAL-3GFP-YKR2 (pER2), and pGAL-3GFP-PKH2 (pER3). To create pGAL-3GFP-PKH1 (pFR37), the same approach was used, but due to the presence of BamHI and SalI restriction sites in the gene, two PCR products were made, one with primer A and a primer corresponding to the sequence 3' to the ClaI site present in PKH1, and the other with a primer 3' of the ClaI site and primer B. A three-way ligation was then used to ligate the two PCR products digested with BamHI and ClaI, or ClaI and SalI, into pGS836 that had been digested with BamHI and SalI. Cells expressing the GFP constructs were grown to mid-exponential phase at 30C in SC-Leu containing Raf/Suc and then induced with 2% galactose for 3 h. Nuclear DNA was stained with 4,6-diamidino-2-phenylindole (DAPI) by adding the dye directly in the medium at a concentration of 1 mug/ml for the last hour of growth. Rhodamine-labeled phalloidin was purchased from Molecular Probes (Eugene, OR). Samples of each culture were viewed directly with a TE300 fluorescence microscope (Nikon, Melville, NY) equipped with a 100x/1.4 Plan-Apo objective and a 1.4 numerical aperture condenser. Digital images were acquired with a bottom-ported Orca 100 charge-coupled device camera (Hamamatsu, Bridgewater, NJ) and Phase 3 Imaging Systems software (Northern Exposure, Inc., Glen Mills, PA). The fraction of lysed cells in cultures was assessed by direct counting of at least 200 cells after staining with a vital dye (LIVE/DEAD Yeast Viability kit; catalog no. 7009; Molecular Probes). Antibody Production | To express a glutathione S-transferase (GST)-Ypk1 fusion protein in E. coli, an similar370-base pair fragment corresponding to the first 114 amino acids of Ypk1 was amplified by PCR from pYPK1 by using the following oligonucleotides: 5'-GGG GGG GGA TCCATG TAT TCT TGG AAG TCA AAG TTT-3', where the underlined nucleotides correspond to an introduced BamHI site, and the bold nucleotides correspond to the start codon; and 5'-GGG GGG AAT TCT CAG GTG GCA TCA TTG GGT GTC CC-3', where the underlined nucleotides correspond to an introduced EcoRI site, and the bold nucleotides correspond to the reverse complement of an introduced stop codon. This PCR fragment was then digested with BamHI and EcoRI, gel purified, and ligated into the pGEX2T vector (Pharmacia, Peapack, NJ), which had been linearized with BamHI and EcoRI to create plasmid pGEX-Ypk1 (pAM5). To express a GST-Ykr2 fusion protein in E. coli, an similar360-base pair fragment corresponding to the first 111 amino acids of Ykr2 was amplified by PCR from pYKR2 by using the following olignucleotides: 5'-GGG GGG GGA TCCATG CAT TCC TGG CGA ATA TCC AAG-3', where the underlined nucleotides correspond to an introduced BamHI site, and the bold nucleotides correspond to the start codon; and 5'-GGG GGG AAT TCT CAA CTC GGT CCC TGC GTC TCA GT-3', where the underlined nucleotides correspond to an introduced EcoRI site, and the bold nucleotides correspond to the reverse complement of an introduced stop codon. This PCR fragment was then digested with BamHI and EcoRI, gel purified, and ligated into the pGEX2T vector, which had been linearized with BamHI and EcoRI to create plasmid pGEX-Ykr2 (pAM6). To prepare antigen, expression of GST-Ypk1(1-114) and GST-Ykr2(1-111) fusions from plasmids, pGEX-Ypk1 and pGEX-Ykr2, respectively, were induced in a protease-deficient E. coli strain BL21 (DE3)[pLys] by addition of isopropyl-beta-d-thiogalacto-pyranoside to a final concentration of 0.2 mM followed by incubation with aeration for 2 h at 30C. Cells were harvested, washed once with ice-cold wash buffer (50 mM Tris-HCl pH 8, 0.5 mM dithiothreitol [DTT], 100 mM KCl, 1 mM phenylmethylsulfonyl fluoride, 0.05% NP-40, 1 mM EGTA, and 1 mM EDTA), and resuspended in 1/20 volume of wash buffer containing 1 M NaCl. Cells were disrupted by digestion with lysozyme (final concentration, 2 mg/ml) followed by sonication. Insoluble material was removed by centrifugation at 12,000 x g, and the soluble GST-Ypk1(1-114) or GST-Ykr2(1-111) proteins were purified by adsorption to, and elution from, glutathione-agarose beads (Pharmacia), essentially as directed by the manufacturer, except that elution was performed in the presence of 1 M NaCl and 20 mM glutathione. The purified proteins were used as immunogens to raise polyclonal antisera in adult female New Zealand White rabbits following standard immunization protocols . The resulting anti-Ypk1 antibodies (serum #1446) and anti-Ykr2 antibodies (serum #1732) are specific to Ypk1 and Ykr2 and do not display any cross-reaction against the incorrect antigen. Anti-GFP antibodies were the generous gift of Roger Tsien and Charles Zuker (Department of Cellular and Molecular Medicine, University of California, San Diego, CA). Preparation of Cell Extracts and Immunoblot Analysis | Yeast cells were grown at 30C to mid-exponential phase (A600 nm = 0.5 --1), either in SC medium supplemented in a manner appropriate for maintenance of plasmids or in rich medium (YPGlc). If cells required galactose induction for expression from the GAL1 promoter, galactose was added to a final concentration of 2% and the cultures were incubated at 30C for an additional 2 h. Cells were harvested by brief centrifugation, washed twice by resuspension and resedimentation in ice-cold lysis buffer (50 mM Tris-HCl pH 7.5, 5 mM EDTA, 3 mM DTT and 1 mM phenylmethylsulfonyl fluoride), and resuspended in 200 mul of the same buffer. Prechilled glass beads (0.45 --0.6 mm in diameter) were added to the meniscus of the cell suspension, and lysis was achieved by vigorous vortex mixing for six 1-min intervals, with intermittent cooling on ice. To remove the glass beads, the bottom of the Eppendorf tube was punctured with a syringe needle (<0.5 mm in diameter) and inserted into another tube; the lysate was collected into the fresh tube by brief centrifugation in a clinical centrifuge. The crude extract was subjected to centrifugation at 30,000 x g for 15 min to remove unbroken cells and large debris. The protein concentration of the crude extract was measured using a dye-binding method with a protein assay kit as instructed by the manufacturer (, Hercules, CA), by using bovine serum albumin (, Beverly, MA) as the standard. For immunoblot analysis, samples (50 mug of total protein) were diluted into SDS-PAGE sample buffer , subjected to electrophoresis in an 8 --12% gel, and then transferred to nitrocellulose . To detect Ypk1, rabbit polyclonal anti-Ypk1 antiserum #1446 was used at a dilution of 1:3000. To detect Ykr2, rabbit polyclonal anti-Ykr2 antiserum #1732 was used at a dilution of 1:3000. To detect proteins using the anti-c-Myc monoclonal antibody (mAb) 9E10, ascites fluid containing this mAb was used at a dilution of 1:10,000 . Immobilized immune complexes were detected using a commercial chemiluminescence detection system (Renaissance; PerkinElmer Life Sciences, Boston, MA) and x-ray film (Biomax MR; Eastman Kodak, Rochester, NY). Immunoprecipitations | Yeast cultures to be used for immunoprecipitation analysis were grown as described above, and then rinsed in ice-cold IP buffer (20 mM Tris-HCl pH 7.5, 125 mM potassium acetate, 0.5 mM EDTA, 0.5 mM EGTA, 1 mM DTT, 1 mug/ml leupeptin, 1 mug/ml pepstatin A, 0.1% Triton X-100, and 12.5% glycerol). Glass beads were added to the meniscus of the cell suspension, and lysis was achieved by vigorous vortex mixing for eight 30-s intervals with intermittent cooling on ice. The lysate was clarified by centrifugation at 14,000 x g at 4C for 30 min. The clarified extract was assayed for protein concentration, and a sample (1 mg of total protein) was diluted to a final volume of 200 mul in IP buffer. An aliquot (20 mul) of protein G/protein A-agarose beads (30% slurry) (Oncogene Science, Cambridge, MA) and a sample of an appropriate control antibody, either 2 mul of preimmune rabbit serum or 1 mug of purified mouse anti-T-cell receptor antibody (gift of James Allison, Department of Molecular and Cell Biology, University of California, Berkeley, CA), were added. The samples were then incubated on a roller drum for 1 h at 4C to adsorb proteins that bound nonspecifically to the solid support and to rabbit or mouse IgG (preclearing). The beads were removed by centrifugation for 10 min in a microfuge, and the supernatant fraction was transferred to a fresh tube containing another aliquot (15 mul) of protein G/protein A-agarose beads and either 2 mul of anti-Ypk1 (or anti-Ykr2) polyclonal antiserum or 1 mul of anti-c-Myc (mAb 9E10) ascites, and incubated on a roller drum for 1 to 3 h at 4C. The beads were sedimented by brief centrifugation in a microfuge and washed three times (1 ml each) with ice-cold IP buffer and collected by centrifugation for 1 min in a microfuge on maximum speed. Bead-bound immune complexes were solubilized in SDS-PAGE sample buffer and immediately boiled for 5 min in a water bath and then clarified by brief centrifugation in a Microfuge before resolution by SDS-PAGE. The proteins of interest were visualized as described above. Immune-Complex Protein Kinase Assays | Cells expressing either wild-type or kinase-dead Ypk1-myc (or Ykr2-myc) under control of the GAL1 promoter were grown in SC containing Raf/Suc to an A600 nm = 0.6, induced by addition of galactose (2% final concentration), incubated with shaking at 30C for 2 h, collected by centrifugation, washed with ice-cold 1x phosphate-buffered saline, resuspended in 0.2 ml of ice-cold IP buffer, and lysed as described above. The resulting lysates were clarified by centrifugation at 4C for 30 min at 30,000 x g. Protein concentration in the resulting crude extracts was determined by the method. A volume of extract containing 1 mg of total protein was immunoprecipitated with mAb 9E10 as described above. The immunoprecipitates were washed once with ice-cold IP buffer, once with ice-cold IP buffer containing 0.5 M NaCl, and twice with ice-cold buffer A (50 mM Tris-HCl pH 7.5, 0.1 mM EGTA, and 0.1% [by vol] 2-mercaptoethanol). As part of the final wash, the slurry of beads was split into two equal portions. For immunoblot analysis, SDS-PAGE sample buffer (similar15 mul) was added directly to one sample of each bead suspension. For protein kinase assays, the activity of the Ypk1-myc or Ykr2-myc immune complex was assayed by adding 30 mul of a mixture containing 1 muM microcystin-LR, 10 mM Mg-acetate, 100 muM [gamma-32P]ATP (200 --400 cpm/pmol), and 100 muM Cross-tide (GRPRTSSFAEG) , which we have documented previously is an excellent peptide phospho-acceptor substrate for Ypk1 and Ykr2 . After incubation for 15 min at 30C, each reaction was terminated by spotting a portion (45 mul) of the reaction mixture onto small squares of phosphocellulose paper (P81; Whatman, Maidstone, United Kingdom), which were washed and analyzed as described in detail previously . In some experiments, samples of the immunoprecipitates were resuspended in an appropriate buffer (20 mM Tris-HCl pH 8.8 and 10 mM MgCl2) and treated with shrimp alkaline phosphatase (0.25 U; US Biochemical, Cleveland, OH) in either the absence or presence of a mixture of inhibitors of this phosphatase (25 muM Na-orthovanadate and 100 muM beta-glycerol-phosphate, adjusted to pH 8). Bioassays for Drug Sensitivity | An agar diffusion (halo) assay was performed to test the relative sensitivity of various strains to rapamycin, valinomycin, hygromycin B, cycloheximide, and polyoxin D. Nascent lawns of the strains to be tested were prepared by mixing similar2 x 106 cells from a saturated culture with 2 ml of molten (55C) 1% agar. The cell-containing agar was rapidly mixed and immediately poured evenly onto plates containing an appropriate medium. Various concentrations of rapamycin (50 and 500 ng/mul), hygromycin B (5 and 50 mug/mul), or the other drugs indicated, were spotted in the same volume (10 mul) onto sterile cellulose filter discs (0.6 cm), which were placed on the nascent lawn. The plates were incubated at 30C, and photographed after 2 d. Selection and Analysis of Dosage Suppressors | A library of restriction fragments of yeast genomic DNA cloned into a URA3-marked, 2-mum DNA-based vector, YEp352 , was introduced into strain YPT40 (ypk1-1ts ykr2Delta) by selecting transformants on SCGlc-Ura medium at 26C. Temperature-resistant clones were then selected by their ability to grow at 35C. One suppressor plasmid obtained carried the EXG1 locus as the sole open reading frame. To create a plasmid that expressed EXG1 from a high-level constitutive promoter, two primers, 5'-GCG TCT CGA GAT GCT TTC GCT TAA AA-3', where the underlined bases correspond to an introduced XhoI site, and the start codon is in bold; and 5'-CGC CGG AGC TC T TAG TTA GAA ATT GTG CC-3', where the underlined bases correspond to an introduced SacI site, and the stop codon is in bold, were used to amplify the entire open reading frame of EXG1 from the library plasmid that was originally isolated as a dosage suppressor of the ypk1-1ts ykr2Delta strain (see RESULTS). This 1.4-kb PCR product was digested with XhoI and SacI and ligated into vector pAD4 M that had been digested with SalI and SacI, yielding pADH-EXG1 (pAM88). To express PKC1 from its own promoter on a high copy number (2-mum DNA) plasmid, a 4.2-kb NsiI fragment containing the entire open reading frame of PKC1 as well as 560 base pairs upstream of the ATG and 175 base pairs downstream of the stop codon, was cut out of a genomic clone obtained by F. Owen Fields (this laboratory; ; ) and ligated into the URA3-based plasmid YEp352 that had been linearized with PstI, yielding plasmid pPKC-PN, or into the LEU2-based plasmid YEp351 that had been linearized with PstI, yielding plasmid pFR32. To express BCK1-20 in a LEU2-based plasmid, the PvuI-PvuI fragment of plasmid pRS314-BCK1 --20 was replaced by the equivalent LEU2-containing fragment of plasmid pRS315, yielding plasmid pRS315-BCK1-20. Selection and Analysis of Chromosomal Suppressor Mutations | A genomic DNA library containing Tn3::LacZ::LEU2 insertions (generous gift of Michael Snyder, Department of Biology, Yale University) was introduced into strain YPT40 (ypk1-1ts ykr2Delta) by selecting transformants on SCGlc-Leu medium at 26C. Temperature-resistant clones were then selected by their ability to grow at 35C. Plasmids carrying genomic DNA corresponding to the sites of insertion were recovered as described in detail previously and characterized by direct nucleotide sequence analysis. RESULTS : Ypk1 Has a More Prominent Role in Maintaining Viability Than Ykr2 | The two protein kinases encoded by the YPK1 and YKR2/YPK2 genes are very similar to each other and functionally redundant at the genetic level . Yeast cells missing either Ypk1 or Ykr2 are viable, but cells lacking both proteins are inviable, indicating that these enzymes share an essential function. However, several observations suggest that Ypk1 plays the predominant role in executing this essential function. First, ypk1Delta mutants are slow growing at 30C , whereas ykr2Delta cells do not display any obvious growth phenotype, compared with otherwise isogenic YPK1+ YKR2+ control cells . Moreover, we found that the slow-growth phenotype of ypk1Delta cells is strongly exacerbated at lower temperatures, even at 26C . In essence, ypk1Delta mutants are cold sensitive and ykr2Delta mutants are not. Figure 1 | Phenotypes of ypk1Delta mutants. Phenotypes of ypk1Delta mutants. ypk1Delta cells (but not ykr2Delta cells) grow slowly at 30C and are cold sensitive. Serial dilutions of exponentially growing wild-type (WT; YPH499) and derived ypk1Delta (YES3) and ykr2Delta (YES1) mutants were spotted on YPGlc plates and grown for 3 d at the indicated temperatures (26, 30, and 35C). Among the close mammalian relatives of Ypk1 and Ykr2 is p70 S6 kinase (50% identity in the catalytic domain). Activation of mammalian p70 S6 kinase by mitogens is blocked by rapamycin, an immunosuppressive drug . Therefore, we tested whether loss of either Ypk1 or Ykr2 conferred on yeast cells elevated sensitivity to this agent. Compared with wild-type cells, we found that ypk1Delta cells, but not ykr2Delta cells, were hypersensitive (similar10-fold) to the growth inhibitory effect of rapamycin , providing a second distinction between ypk1Delta and ykr2Delta mutants. We found that ypk1Delta cells (but not ykr2Delta cells) were also hypersensitive to hygromycin B, valinomycin, polyoxin D, cycloheximide , and caffeine (Bezman, unpublished observations). These results suggested that ypk1Delta cells are generally more permeable to drugs and provided indirect evidence for a defect or perturbation in the cell envelope in ypk1Delta cells (see below). When overexpressed, either YPK1 or YKR2 was able to restore the normal level of drug sensitivity (Roelants, unpublished observations). Given the fact that Pkh1 and Pkh2 are upstream activators of Ypk1 and Ykr2 , it was of interest to test whether loss of either Pkh1 or Pkh2 might also confer a drug-sensitive phenotype. However, neither a pkh1Delta mutant nor a pkh2Delta mutant showed any degree of hypersensitivity to the compounds mentioned above, compared with the parental strain . Catalytically inactive (kinase-dead) alleles of protein kinases frequently act as dominant-negatives . Hence, we tested whether overexpression of catalytically inactive versions of Ypk1 and Ykr2 would be toxic to cells. We constructed two such derivatives altered in the invariant Lys found in conserved kinase motif II . We showed that Ypk1(K376A) and Ykr2(K373A) are indeed catalytically nonfunctional in vitro and unable to complement the lethality of ypk1Delta ykr2Delta cells in vivo . To test whether these alleles behave in a dominant-negative manner when overexpressed from an inducible promoter, plasmids pGAL-YPK1-(K376A-KD) and pGAL-YKR2-(K373A-KD) were introduced into wild-type cells (YPH499) and into ypk1Delta (YES3) and ykr2Delta (YES1) mutants, selecting for transformants on SC-Leu medium containing Glc as the carbon source. The resulting transformants then were streaked onto SC-Leu medium containing Gal/Suc to select for maintenance of the plasmid and to induce expression of either kinase-dead Ypk1 or kinase-dead Ykr2. High-level expression of catalytically inactive Ypk1 was growth inhibitory to all three cell types; however, the strongest growth inhibition was observed in ypk1Delta cells and the mildest was observed in wild-type cells (Figure A). In contrast, high-level expression of catalytically inactive Ykr2 had no detectably detrimental effect on growth in any of the strains (Figure A). To rule out the possibility that this differential effect was due to a difference in the level of expression of these proteins, Ypk1, Ypk1(K376A), Ykr2, and Ykr2(K373A) were tagged at their C termini with a c-Myc epitope and introduced into ypk1Delta or ykr2Delta strains. Identical amounts of protein from extracts of the resulting transformants were immunoprecipitated with anti-Myc mAb 9E10 antibodies, resolved by SDS-PAGE, and visualized by immunoblotting with polyclonal anti-Ypk1 or anti-Ykr2 antibodies. This analysis verified that the proteins were expressed at equivalent levels (Figure B). The fact that kinase-dead Ypk1 was able to effectively impede its own function and that of Ypk2, and the fact the converse was not true, provided a third independent indication that Ypk1 plays the more predominant role. Figure 2 | Overexpression of catalytically inactive Ypk1 inhibits growth. Overexpression of catalytically inactive Ypk1 inhibits growth. (A) Dominant-negative effect of inactive Ypk1. Wild-type cells (WT; YPH499), and ypk1Delta (YES3) and ykr2Delta (YES1) mutants were transformed, as indicated, with either an empty vector (YEp352GAL) or the same vector expressing from the GAL1 promoter either full-length Ypk1 (pGAL-YPK1), a catalytically inactive (kinase-dead) ypk1 allele (pGAL-YPK1-KD; pAM48), full-length Ykr2 (pGAL-YKR2; pAM59), or a catalytically inactive (kinase-dead) ykr2 allele (pGAL-YKR2-KD; pAM61). Transformants were selected on glucose-containing medium and then representative isolates were streaked to single colonies on selective medium containing either Glc (left) or Gal (right) as the carbon source and grown for 3 d at 30C. (B) Equivalent overexpression confirmed by immunoblotting. Cultures of the resulting transformants were grown at 30C in SCGal/Suc-Leu medium, induced by addition of galactose (2% final concentration) for 3 h. After harvesting the cells by centrifugation, extracts were prepared, and an identical amount of total protein (1 mg) from each extract was subjected to immunoprecipitation with anti-Myc mAb 9E10, as described in MATERIALS AND METHODS. After washing, the immunoprecipitates were solubilized and resolved by electrophoresis on an SDS-slab gel. The species corresponding to Ypk1 and Ypk2 were visualized by immunoblotting with rabbit polyclonal anti-Ypk1 (left) or anti-Ykr2 (right) antibodies, respectively. Pkh1 Preferentially Activates Ypk1 and Pkh2 Preferentially Activates Ykr2 | Purified Pkh1 phosphorylates and activates purified Ypk1 in vitro . To examine the state of activation of Ypk1 and Ykr2 in cell extracts and its dependence on the function of Pkh1 and Pkh2, we developed an immune-complex kinase assay. Cell extracts were prepared from strains expressing c-Myc epitope-tagged derivatives of Ypk1 or Ykr2, immunoprecipitated with anti-Myc mAb 9E10, and samples of the resulting immunoprecipitates were examined for protein content by SDS-PAGE and for catalytic activity using a specific peptide substrate (Cross-tide) and [gamma-32P]ATP in a filter binding assay . As independent negative controls to assess the nonspecific background, extracts were prepared from cells expressing untagged Ypk1 and Ykr2 and from cells expressing kinase-dead derivatives of Ypk1 and Ykr2. Immune complexes from wild-type cells expressing wild-type Ypk1-myc showed similar10-fold increase in phosphotransferase activity compared with both negative controls: immune complexes from wild-type cells expressing untagged Ypk1 (Figure A, top) and immune complexes from wild-type cells expressing catalytically inactive Ypk1-myc . When Ypk1-myc was isolated from pkh1Delta cells, however, the increase in activity was reproducibly 40 --60% lower than that observed in wild-type cells, whereas within experimental error, the recovery of Ypk1-myc activity was unaffected in pkh2Delta cells (Figure A, top). Most tellingly, activity was greatly increased (>=4-fold) when Ypk1-myc was isolated from cells cooverexpressing Pkh1 but only modestly elevated (similar1.5-fold) when Pkh2 was cooverexpressed (Figure A, top). Because Ypk1-myc and Pkh1 (or Pkh2) were both expressed from multicopy plasmids carrying the GAL1 promoter, which compete for a limiting pool of the Gal4 transactivator, the total amount of Ypk1 produced was reduced (Figure A, bottom); thus, the increase in specific activity when Ypk1-myc and Pkh1 were co-overexpressed is even more dramatic than the observed increase in total activity. Moreover, immunoprecipitated Ypk1-myc ran as a set of multiple bands that were collapsed into a single band of faster mobility upon treatment with a phosphatase (Figure A, bottom). Phosphatase-treated samples were no longer catalytically active (Torrance, unpublished observations). These findings suggest that activation is due to phosphorylation and that activation of Ypk1 is more dependent on Pkh1 than on Pkh2. Figure 3 | Preferential activation of Ypk1 and Ykr2 by Pkh1 and Pkh2. Preferential activation of Ypk1 and Ykr2 by Pkh1 and Pkh2. (A) Pkh1 differentially activates Ypk1. Top, wild-type (WT; W303-1B), pkh1Delta (AC301), and pkh2Delta (YPT67) cells overexpressing from the GAL1 promoter on plasmids either Ypk1 (pAM75) or Ypk1-myc (pAM76), in the absence or presence of co-overexpression from the GAL1 promoter on plasmids of either Pkh1 (pAM73) or Pkh2 (pAM79), as indicated, were grown to mid-exponential phase and induced with galactose for 3 h. Extracts were prepared and identical amounts of total protein (1 mg) were immunoprecipitated with anti-Myc mAb 9E10. The immune complexes were washed extensively with protein kinase assay buffer and then incubated with a specific peptide substrate (Cross-tide) and [gamma-32P]ATP, and the resulting product measured, as described in MATERIALS AND METHODS. Bottom, samples of each immunoprecipitate were treated with phosphatase in the absence and presence of phosphatase inhibitors, and examined by SDS-PAGE, as described in MATERIALS AND METHODS. (B) Pkh2 differentially activates Ykr2. Wild-type (WT; W303-1B), pkh1Delta (AC301), and pkh2Delta (YPT67) cells overexpressing from the GAL1 promoter on plasmids either Ykr2 (pAM1) or Ykr2-Myc (pAM78), in the absence or presence of co-overexpression from the GAL1 promoter on plasmids of either Pkh1 (pAM73) or Pkh2 (pAM79), as indicated, were grown to mid-exponential phase and induced with galactose for 3 h. Extracts were prepared and activity was measured, as in A. Values shown in A and B represent the average of three independent experiments, each performed in duplicate, and the error bars represent the range of values observed. Immune complexes from wild-type cells expressing wild-type Ykr2-myc showed approximately a 10-fold increase in phosphotransferase activity compared with both negative controls: immune complexes from wild-type cells expressing untagged Ykr2 (Figure B) and immune complexes from wild-type cells expressing catalytically inactive Ykr2-myc . When Ykr2-myc was isolated from either pkh1Delta or pkh2Delta cells, however, the increase in activity was reproducibly lower than that observed in wild-type cells, suggesting that both Pkh1 and Pkh2 contribute to phosphorylation and activation of Ykr2. Total activity was stimulated approximately threefold when Ykr2-myc was recovered from cells co-overexpressing Pkh2, but only similar1.5-fold when Pkh1 was co-overexpressed (Figure B), indicating that phosphorylation and activation of Ykr2 are more responsive to Pkh2 than to Pkh1. Loss of PKH2 (but Not PKH1) Exacerbates Slow Growth of ypk1Delta Cells | The biochemical assays discussed above indicated that Pkh1 preferentially activates Ypk1 and Pkh2 preferentially activates Ykr2. Given that both phk1Delta pkh2Delta and ypk1Delta ykr2Delta double mutants are inviable , if the discrimination observed in vitro with overexpressed proteins is even more stringent in vivo when these enzymes are expressed at their normal levels then it might be expected that certain mutant combinations might display genetic interaction. To construct all possible double mutant combinations between either ypk1Delta or ykr2Delta and either pkh1Delta or pkh2Delta, a ypk1Delta haploid was crossed to a pkh1Delta strain and to a pkh2Delta strain to create two diploid strains: PKH1/pkh1Delta::TRP1 YPK1/ypk1Delta::HIS3 and PKH2/pkh2Delta::HIS3 YPK1/ypk1Delta::TRP1. Likewise, a ykr2Delta haploid was crossed to a pkh1Delta strain and to a pkh2Delta strain to create two additional diploid strains: PKH1/pkh1Delta::TRP1 YKR2/ykr2Delta::HIS3 and PKH2/pkh2Delta::HIS3 YKR2/ykr2Delta::TRP1. The four doubly heterozygous diploid strains were sporulated. After tetrad dissection, viable Trp+ His+ haploid spores were readily recovered from all four diploids, indicating that all four double mutants (pkh1Delta ypk1Delta, pkh2Delta ypk1Delta, pkh1Delta ykr2Delta, and pkh2Delta ykr2Delta) are viable (Figure A). However, all of the pkh2Delta ypk1Delta spore clones grew significantly more slowly than any of the ypk1Delta spore clones or any of the pkh2Delta spore clones, when either streaked to single colonies on plates (Figure A) or examined by more definitive spot tests (Figure B). This finding provides genetic evidence in support of the biochemical results that the primary and physiologically relevant upstream activator of Ykr2 in vivo is Pkh2 (and not Pkh1). Nonetheless, a yeast cell can survive with either member of these two tiers of protein kinases. Hence, Pkh1 and Pkh2 must each be able to phosphorylate and activate either Ypk1 or Ykr2 to at least some significant degree. Figure 4 | Slow growth of ypk1Delta cells is exacerbated by absence of Pkh2 (but not Pkh1). Slow growth of ypk1Delta cells is exacerbated by absence of Pkh2 (but not Pkh1). (A) Four spores of tetratype asci produced by doubly heterozyous diploids derived from the following four crosses were streaked to single colonies and tested for growth on YPGlc plates at 30C: top left, a ypk1Delta strain (YFR107) against a pkh1Delta strain (YFR105); top right, a ypk1Delta strain (YFR107) against a pkh2Delta strain (YFR106); bottom left, a ykr2Delta strain (YFR119) against a pkh1Delta (YFR105) strain; and bottom right, a ykr2Delta strain (YFR64) against a pkh2Delta (YFR106) strain. (B) Serial dilutions of exponentially growing cultures of the indicated genotypes derived, respectively, from the crosses ypk1Delta (YFR107) against pkh1Delta (YFR105), left side, and ypk1Delta (YFR107) against pkh2Delta (YFR106), right side, were spotted on YPGlc plates and grown for 2 d at 30C. Loss of PKH1 (but Not PKH2) Alleviates Toxicity of Hyperactive Ypk1 | Ypk1 and Ykr2 share 88% identity within their 252-residue kinase domains and 75% identity within their downstream 75-residue C-terminal extensions. Ypk1 and Ykr2 also share considerable similarity within their 350 --353-residue amino-terminal extensions: 22% identity within the first similar100 residues and strikingly, 65% identity within the next 250 residues. To investigate what the large amino-terminal domain might contribute to the function of Ypk1, a truncation allele of Ypk1 was constructed in which the entire amino terminus (residues 2 --336) was deleted. The YPK1(Delta2--336) allele, encoding Ypk1-DeltaN, commences with Ser337; in Ypk1, the first Gly of the GxGxxG motif conserved in all protein kinases lies at residue 354. To permit its conditional expression, YPK1(Delta2-336) was inserted in a multicopy vector under control of the GAL1 promoter. Induction of the resulting plasmid, pGAL-YPK1-DeltaN (pAM101), in a wild-type strain (W303-1B) on galactose-containing medium was toxic as judged by the exceedingly slow growth of single colonies . Despite its toxicity, overexpression of Ypk1-DeltaN was capable of restoring growth (albeit very slowly) to an otherwise inviable ypk1Delta ykr2Delta double mutant . The observed toxicity required the catalytic activity of Ypk1-DeltaN because a kinase-dead derivative, Ypk1(K376A)-DeltaN, was not detectably growth inhibitory , even though immunoblotting indicated that, after induction, Ypk1-DeltaN and Ypk1(K376A)-DeltaN were expressed at equivalently high levels (Torrance, unpublished observations). Revealingly, the toxicity of Ypk1-DeltaN was also alleviated in cells lacking Pkh1, but not in cells lacking Pkh2 . The fact that Pkh1 was required for the dominant toxicity of Ypk1-DeltaN provides genetic evidence in support of the biochemical results that the primary and physiologically relevant upstream activator of Ypk1 in vivo is Pkh1 (and not Pkh2). Figure 5 | Absence of Pkh1 (but not Pkh2) alleviates toxicity of constitutively active Ypk1. Absence of Pkh1 (but not Pkh2) alleviates toxicity of constitutively active Ypk1. Wild-type (WT; W303-1B), pkh1Delta (AC301), or pkh2Delta (YPT67) strains were transformed with either an empty vector (YEp352GAL) or the same vector expressing from the GAL1 promoter either normal Ypk1 (pAM75), the C-terminal catalytic domain of Ypk1 (Ypk1-DeltaN; pAM101), or a catalytically inactive derivative of the C-terminal catalytic domain of Ypk1 (Ypk1-KD-DeltaN; pFR30). The resulting transformants were selected on glucose-containing medium and then representative isolates were streaked onto selective medium containing either Glc (left) or Gal (right) as the carbon source. Growth was assessed after 3 d at 30C. Differential Subcellular Localization of Pkh1, Pkh2, Ypk1, and Ykr2 | One explanation for the observed preferential phosphorylation of Ypk1 by Pkh1, and of Ykr2 by Pkh2, is that the activating enzyme and its downstream target are confined to the same subcellular compartment. As an initial approach to examine localization, each of these four proteins was tagged at its N terminus with three tandem in-frame repeats of GFP and expressed from the GAL1 promoter in a multicopy plasmid. Both 3GFP-Pkh1 and 3GFP-Pkh2 were able to complement the temperature sensitivity of a pkh1ts pkh2Delta strain at restrictive temperature (37C) on galactose-containing medium and even on glucose-containing medium (Roelants, unpublished observations), indicating that each construct was functional. Similarly, 3GFP-Ypk1 and 3GFP-Ykr2 retained their biological function (and transcriptional control was tighter) because each construct was able to complement the temperature sensitivity of ypk1-1ts ykr2Delta cells at the nonpermissive temperature (37C) on galactose-containing medium (but not on glucose-containing medium) (Roelants, unpublished observations). In addition, as judged by immunoblotting with anti-GFP antibodies, each of the four tagged proteins was expressed intact and had the molecular weight expected for the full-length chimeric protein (Roelants, unpublished observations). Live wild-type cells expressing each of the four fusions to 3GFP were examined under the fluorescence microscope . The 3GFP-Ypk1 chimera was found exclusively in the cytosol and was excluded from both the vacuole (whose position was observed by phase contrast microscopy of the same field) and the nucleus (whose position was revealed by growing the cells in the DNA-specific dye DAPI). In contrast, the 3GFP-Ykr2 chimera accumulated in the nucleus, congruent with the DAPI-stained DNA, although it was also readily detectable in the cytoplasm. Interestingly, when fused to 3GFP, the catalytic domain of Ypk1 (Ypk1DeltaN), which by itself is toxic when overexpressed (see above), was located predominantly in the nucleus, unlike full-length 3GFP-Ypk1, suggesting that its toxicity may arise largely from its mislocalization. The same patterns of distribution for Ypk1 and Ykr2 were also observed if the cells were fixed, permeabilized, and stained, respectively, with polyclonal anti-Ypk1 and anti-Ykr2 antibodies (Torrance, unpublished observations). Likewise, identical patterns of distribution were observed when cells expressing Ypk1-myc or Ykr2-myc were examined by indirect immunofluorescence using anti-c-Myc mAb 9E10 (Roelants, unpublished observations). Figure 6 | Subcellular localization of GFP-tagged Ypk1, Ykr2, Pkh1, and Pkh2. Subcellular localization of GFP-tagged Ypk1, Ykr2, Pkh1, and Pkh2. Wild-type (YPH499) cells were transformed with low copy number (CEN DNA-based) plasmids expressing from the GAL1 promoter either 3GFP-Ypk1 (pFR33), 3GFP-Ypk1-DeltaN (pFR34), 3GFP-Ykr2 (pER2), 3GFP-Pkh1 (pFR37), or 3GFP-Pkh2 (pER3), as indicated. The transformants were grown to mid-exponential phase at 30C in SCRaf/Suc-Leu, induced with galactose (2%) for 3 h, and samples of each culture were viewed directly under a fluorescence microscope. To permit visualization of the position of the nucleus, DAPI was added to the medium (1 mug/ml final concentration) during the last hour of induction. As observed for 3GFP-Ypk1, 3GFP-Pkh1 was localized exclusively to the cytosol, and clearly excluded from both the vacuole and the nucleus . The most prominent feature of the 3GFP-Pkh1 staining was, however, bright puncta or larger patches situated at the cell cortex. These dots are not congruent with actin patches, as was revealed by costaining with rhodamine-labeled phalloidin (Roelants, unpublished observations). The same distribution pattern was observed if cells expressing Pkh1-(HA)3 were fixed, permeabilized, and examined by indirect immunofluorescence by using an anti-HA mAb and an appropriate fluorescently tagged secondary antibody (Roelants, unpublished observations). Unlike 3GFP-Pkh1, 3GFP-Pkh2 was not excluded from the nucleus, but like 3GFP-Pkh1, the most prominent feature of the staining was a large number of punctate bodies immediately subtending the plasma membrane , which were also distinct from actin patches (Roelants, unpublished observations). Thus, taken together, these observations indicate that Pkh2 and Ykr2 are able to enter a compartment (the nucleus) from which Pkh1 and Ypk1 are normally excluded. Thus, these findings help to explain, at least in part, the greater dependence of Ypk1 activation on Pkh1 and the greater dependence of Ykr2 activation on Pkh2. Genetic Analysis of Ypk1 and Ykr2 Function by Selection of Dosage Suppressors | To identify gene products that may be involved in processes both upstream and downstream of Ypk1 and Ykr2, we selected, first, for genes that when overexpressed from a URA3-marked multicopy vector, were able to restore growth to ypk1-1ts ykr2Delta cells at an otherwise nonpermissive temperature (35C), as described in MATERIALS AND METHODS. From 20,000 Ura+ transformants, we recovered 18 plasmids that were able to support growth reproducibly at the restrictive temperature . As expected, seven independent isolates of YKR2 and one isolate of YPK1 were obtained. Two of the other suppressor genes obtained, one encoding a putative chaperone (HLJ1) and the other encoding a component of RNA polymerase II holoenzyme (SRB4), may rescue because they stabilize or elevate expression of the temperature-sensitive Ypk1-1 enzyme, although this hypothesis was not tested directly. Another suppressor plasmid carried the YPC1 gene, which encodes an enzyme that can generate phytosphingosine from the corresponding phytoceramide and hence presumably rescues by hyperstimulating Pkh1 and Pkh2. Indeed, we have shown previously that elevated Pkh1 can restore growth to ypk1-1ts ykr2Delta cells at otherwise restrictive temperature . Indeed, when excised from the original isolate and overexpressed from a completely different vector, YPC1 rescues the temperature sensitivity of ypk1-1ts ykr2Delta cells (Roelants, unpublished observations). In contrast, at least one other of the genes from the same insert, RPS6B, was not a suppressor on its own (Roelants, unpublished observations). Table 2 | Dosage suppressors of the temperature-sensitive lethality of ypk1-1ts ykr2Deltacells Revealingly, among the seven remaining dosage suppressors, two plasmids carried a single intact open reading frame corresponding to the EXG1 gene, which encodes the major exo-beta(1,3)-glucanase involved in cell wall remodeling . Indeed, when excised from the original isolate and expressed from a constitutive promoter (ADH1) in a completely different multicopy vector (see MATERIALS AND METHODS), elevated expression of EXG1 reproducibly suppressed, albeit weakly, the temperature-sensitive growth defect of ypk1-1ts ykr2Delta cells, even at 37C (Figure A). Also obtained were two isolates of a locus (YBL104c) of unknown function, but which seems from the phenotype of a null allele to also have effects on cell wall structure . Another suppressor plasmid isolated carries multiple open reading frames, one of which is a candidate chitinase , which may also influence cell wall structure. At least one of the other genes carried on this same plasmid, FRQ1 , is not responsible for the suppression and does not contribute to the suppression (Roelants, unpublished observations). The final two dosage suppressors encoded proteins that might act by enhancing the efficiency with which enzymes involved in cell wall biosynthesis or remodeling are delivered to their final destination and/or are activated there. One plasmid carried GOT1, which specifies a membrane protein thought to enhance the function of a t-SNARE heavy chain, Sed5, involved in vesicle-mediated protein transport from the endoplasmic reticulum (ER) to the Golgi . The other plasmid carried only PLB1, which specifies the phospholipase B that is primarily responsible for the conversion of phosphatidylcholine and phosphatidylethanolamine in the exocellular leaflet of the plasma membrane to lysophosphatidylcholine (and glycerophosphocholine) and lyso-phosphatidylethanolamine (and glycerophosphoethanolamine), respectively . It is well documented that the activity of many classes of membrane-associated enzymes can be influenced dramatically (either stimulated or inhibited), depending on the nature of the phospholipids (or their derivatives) with which those enzymes associate . Figure 7 | Relationship between dosage suppressor (EXG1) and extragenic suppressor (kex2Delta). Relationship between dosage suppressor (EXG1) and extragenic suppressor (kex2Delta). (A) Overexpression of EXG1 suppresses the temperature sensitivity of ypk1-1ts ykr2Delta cells. A temperature-sensitive ypk1-1tsykr2Delta strain (YPT40) was transformed with either an empty vector (pAD4M) or the same vector expressing from the ADH1 promoter either YPK1 (pADH-YPK1) or EXG1 (pAM88). Growth of the resulting transformants was assessed after 2 d at 30 and 37C, as indicated. (B) Deletion of KEX2 suppresses the temperature sensitivity of ypk1-1tsykr2Delta cells. A ypk1-1tsykr2Delta strain (YAN2) was crossed to a kex2Delta strain (KRY24), and the resulting diploid cells (YFR66) were sporulated and dissected. The four spores of a tetratype ascus derived from this diploid were recovered and tested for growth on YPGlc at 30 and 37C, as indicated. (C) Suppression by kex2Delta requires EXG1. A MATaypk1-1tsykr2Delta kex2Delta strain, derived as described in B, was crossed to an exg1Delta strain (YFR84) and the resulting diploid cells were sporulated and dissected. Serial dilutions of cultures of the indicated genotypes were spotted on YPGlc plates and growth was assessed after 3 d at 30 and 37C, respectively. (D) Either kex2Delta cells or rot2Delta cells lacking Mpk1 are inviable. A kex2Delta strain (KRY24) was crossed to a mpk1Delta strain (YFR128), and the resulting diploid cells were sporulated and dissected on plates containing 1.2 M sorbitol. The four spores of a tetratype ascus derived from this diploid were recovered and tested for growth on YPGlc and YPGlc containing 1 M sorbitol at 30C. The same procedure was applied to diploid cells resulting from crossing a rot2Delta strain (YFR129) and an mpk1Delta strain (YFR127). Taken together, the nature of the dosage suppressors obtained, along with our observation that ypk1Delta cells showed a general increase in permeability to inhibitory drugs of several different chemical classes (see above), strongly suggested that the primary defect in Ypk1- and Ykr2-deficient cells involved some aspect of cell wall biosynthesis and/or structure. Genetic Analysis of Ypk1 and Ykr2 Function by Selection of Suppressor Mutations | To gain further insight, and to corroborate the conclusion that absence of Ypk1 and Ykr2 compromises some aspect of cell wall structure, we also performed a selection for genes that, when interrupted by insertion of a transposon (Tn3::LacZ::LEU2), were able to restore the ability of ypk1-1ts ykr2Delta cells (strain YPT40) to grow at high temperature (35C), as described in MATERIALS AND METHODS. From 80,000 Leu+ transformants, 16 haploid isolates were obtained that contained a transposon insertion and for which Leu+ segregated with the ability to grow at high temperature when the isolate was backcrossed to a ypk1-1ts ykr2Delta cell of opposite mating type (strain YAN2) . One extragenic suppressor obtained inactivated a gene (SRN2) that interacts genetically with the machinery involved in nucleocytoplasmic transport. This mutation may simply enhance export of YPK1 mRNA and thus expression of the temperature-sensitive Ypk1-1 enzyme, although this hypothesis was not tested directly. Two other insertions disrupted uncharacterized loci of unknown function. Reassuringly, however, the 13 remaining extragenic suppressors fell in genes of known function, which were all involved in processes required for biosynthesis of normal cell wall glycoproteins. Table 3 | Transposon insertions that suppress the temperature-sensitive lethality of ypk1-1ts ykr2 cells Four independent isolates represented transposon insertions in the ALG5 gene. Alg5 is an integral membrane enzyme that transfers glucose from UDP-Glc to the dolichol carrier that is used to attach the Glc residues to the immature Asn-linked (GlcNAc)2(Man)9(Glc)3 core oligosaccharide, which is added en bloc to cell wall mannoproteins and other secreted glycoproteins in the lumen of the ER . Three independent suppressors were insertions in a gene (ROT2) encoding the integral membrane enzyme that trims two alpha(1,3)-linked glucose residues from the (GlcNAc)2(Man)9(Glc)3 core during subsequent maturation of secreted glycoproteins . Two additional extragenic suppressors inactivated two other genes, ALG3/RHK1 and ALG9 , that encode mannosyltransferases involved in adding the sixth and seventh mannose residues, respectively, to the (GlcNAc)2(Man)9(Glc)3 core during its biosynthesis. Two other suppressor mutations fell in the gene (DNF3), which encodes an apparent transport ATPase for aminophospholipids (phosphatidylethanolamine and phosphatidylserine) , which may affect the composition of the ER membrane and thereby influence the activity of one or more of the enzymes involved in Asn-linked oligosaccharide biosynthesis mentioned above. Strikingly, two more independent isolates corresponded to transposon insertions in the KEX2 gene, which encodes a Golgi-localized endoprotease that participates in maturation of secreted precursor glycoproteins by cleaving on the C-terminal side of pairs of basic residues . To verify that suppression was due to Kex2 loss of function, we crossed a ypk1-1ts ykr2Delta strain against a kex2-Delta2::LEU2 strain, in which the entire KEX2 open reading frame was deleted and replaced by the LEU2 gene, and examined the phenotype of the spores derived from resulting tetratype asci. Just like the original transposon insertions, a standard kex2 null allele also rescued the growth of ypk1-1ts ykr2Delta cells at the restrictive temperature (Figure B). One possible explanation for this suppression is that, normally, Ypk1 and/or Ykr2 are negative regulators of Kex2 synthesis, function, or intracellular trafficking. In this regard, it was noteworthy that Kex2 is reportedly involved in processing of the precursor of the Exg1 exoglucanase . Thus, the fact that overexpression of Exg1 also rescued the temperature sensitivity of a ypk1-1ts ykr2Delta strain (Figure A) suggested that, perhaps, it is the unprocessed form of Exg1 that is responsible for the suppression, because this precursor form presumably accumulates when Kex2 is absent (due to mutation) or if Kex2 is limiting (when Exg1 is overproduced). To determine whether amelioration of the temperature sensitivity of ypk1-1ts ykr2Delta cells by loss of Kex2 involved Exg1, we deleted EXG1 in the ypk1-1ts ykr2Delta kex2Delta strain. Indeed, the absence of Exg1 greatly reduced the ability of the ypk1-1ts ykr2Delta kex2Delta cells to grow at nonpermissive temperature (Figure C), suggesting that an intact EXG1 gene is required for mediating, at least in part, the suppressive effect of loss of Kex2. The residual growth observed could be explained by the fact that the S. cerevisiae genome encodes 12 other demonstrated and presumptive glucanases whose precursors may also require Kex2-mediated processing. However, there was another equally plausible explanation for the ability of kex2 mutations to suppress the temperature sensitivity of ypk1-1ts ykr2Delta cells that was consistent with all of the above-mentioned observations. Specifically, absence of Kex2 prevents processing of certain secreted cell wall mannoproteins, causing defects in the cell wall . Moreover, various defects in the cell wall trigger activation of the Pkc1-dependent mitogen-activated protein (MAP) kinase Mpk1/Slt2 and induction of genes under its control , including EXG1 . Indeed, in agreement with the hypothesis that the Pkc1-Mpk1 pathway is induced when cell wall structure is perturbed by a kex2Delta mutation and by at least one other of the extragenic suppressors (rot2Delta) we isolated, we found that these mutants are inviable when Mpk1 is absent but rescued on medium containing an osmotic support (Figure D). In fact, this synthetic lethality suggests that the only reason that kex2Delta and rot2Delta mutants are able to survive is that they induce the Pkc1 --Mpk1 pathway, which up-regulates glucan synthases and many other enzymes necessary to repair, modify, and maintain the otherwise abnormal cell wall . Thus, collectively, the above-mentioned findings suggested that Ypk1 and Ypk2 participate in a signaling pathway required for optimal cell wall integrity and that all of the dosage suppressors and extragenic suppressors rescue the lethality of ypk1-1ts ykr2Delta cells because they cause additional cell wall perturbations that induce the alternative Pkc1 --Mpk1 cell wall integrity signaling pathway, and therefore bypass the need for efficient Ypk1- and Ykr2-dependent signaling. Ypk1 and Ykr2 Are Involved in a Novel Cell Wall Integrity Signaling Pathway | One diagnostic property of many of the conditional mutations whose primary defect is perturbation of cell wall structure is that the cells lose viability rapidly at the restrictive temperature because they undergo lysis (for review, see ). Indeed, we found that >50% of the population of ypk1-1ts ykr2Delta cells underwent lysis by 2 h after shift to nonpermissive temperature (37C), and >90% of the cells were lysed by 4 h after the shift, as judged by staining with a commercial vital dye and by plating the cells for viable titer, whereas <2% of control cells were lysed under the same conditions . A second hallmark of mutations that lead directly or indirectly to defects in cell wall structure is that inviability can be rescued in medium containing an osmotic support . Again, consistent with a primary defect in cell wall integrity, the lysis phenotype of ypk1-1ts ykr2Delta cell could be completely prevented by the presence of an osmotic support (1.2 M sorbitol) in the growth medium (Figure A). Third, there is coupling between normal cell wall assembly and proper organization of the actin cytoskeleton . In normal cells, actin patches are confined to the bud and bundles of actin cables are found only in the mother cell . We found, first, that compared with isogenic wild-type cells, ypk1Delta cells showed a marked depolarization of the actin cytoskeleton even at 30C, with pronounced actin patches in the mother cell and no detectable bundles of actin cables (Figure B). Likewise, at 30C, ypk1-1ts ykr2Delta cells displayed normal actin polarization, whereas after shift to restrictive temperature (37C) and before lysis, ypk1-1ts ykr2Delta displayed a pronounced defect in actin polarization, with numerous actin patches present in the mother cell at all stages of the cell cycle and no detectable bundles of actin cables (Figure B). Figure 8 | Ypk1 and Ykr2 are required for maintenance of cell wall integrity. Ypk1 and Ykr2 are required for maintenance of cell wall integrity. (A) Osmotic support permits growth of ypk1-1tsykr2Delta cells at restrictive temperature. Either ypk1-1tsykr2Delta cells (YPT40) or wild-type cells (WT; YPH499) were streaked onto YPGlc plates or YPGlc plates containing 1.2 M sorbitol and incubated for 2 d at 30 and 37C, respectively. (B) Loss of Ypk1 causes defects in actin organization. Wild-type cells (WT; YPH499), top line, and ypk1Delta cells (YES3), second line, were grown to mid exponential phase at 30C. Also, ypk1-1tsykr2Delta cells (YPT40) were grown in YPGlc at 26C for 18 h to an A600 nm = 0.5, third line, and then a portion of the same culture was shifted to 37C for 2 h, bottom line. Samples of all four cultures were then fixed, stained with a reagent specific for visualizing F-actin (rhodamine-labeled phalloidin), and examined under the fluorescence microscope. (C) Hyperactivation of the Pkc1-dependent Mpk1 MAP kinase pathway rescues the temperature sensitivity of a ypk1-1tsykr2Delta cells. Strain YPT40 (ypk1-1ts ykr2Delta) was transformed with an empty 2-mum DNA-based vector (YEp351) or with the same vector expressing from their native promoters either YPK1 (pAM21), PKC1 (pFR32), or BCK1-20 (pRS315-BCK1-20), or a different 2-mum DNA-based vector expressing EXG1 from the ADH1 promoter (pAM88). The resulting transformants were streaked onto SCGlc-Leu and growth was assessed after 3 d at the indicated temperatures (26, 35, and 36C). (D) ypk1-1tsykr2Delta cells lacking Mpk1 are inviable. A heterozygous ypk1-1ts/YPK1 ykr2Delta/YKR2 kex2Delta/KEX2 diploid strain (YFR66) was deleted for MPK1, and the resulting diploid strain was sporulated and dissected on YPGlc containing 1.2 M sorbitol. The indicated strains were streaked onto YPGlc plates and growth was assessed after 3 d at 26 and 30C. All of the above-mentioned phenotypes are also displayed by mutations that compromise the Pkc1-dependent cell wall integrity signaling pathway (for review, see ). In this pathway, Pkc1 (a protein kinase C-like enzyme) activates a MAP kinase cascade composed of a mitogen-activated protein kinase kinase kinase (Bck1), two redundant mitogen-activated protein kinase kinases (Mkk1 and Mkk2), and the Mpk1/Slt2 MAP kinase. Therefore, we tested directly the hypothesis that induction of the Pkc1-Mpk1 signaling pathway can bypass the defect in cells deficient in Ypk1- and Ykr2-mediated signaling. Consistent with this idea, we found that overexpression of PKC1 from its own promoter on a multicopy (2 mum DNA) plasmid was sufficient to rescue the temperature sensitivity of ypk1-1ts ykr2Delta cells (strain YPT40) and did so more efficiently than any of the dosage suppressors we selected directly, including EXG1 (Figure C). Similarly, PKC1 overexpression, like expression of YKR2 (as a control), was able to rescue the actin polarization defects of ypk1-1ts ykr2Delta cells at nonpermissive temperature (Roelants, unpublished observations). Likewise, expression of a constitutively active allele of BCK1 (BCK1-20) was also able to restore growth at high temperature, although somewhat less efficaciously than PKC1 (Figure C). Thus, in some senses, Pkc1 and the Mpk1 MAP kinase pathway act downstream of Ypk1 and Ykr2. On the other hand, several observations indicated that Ypk1 and Ykr2 do not act upstream of Pkc1, but rather in a parallel pathway. First, none of the dosage suppressors, none of the extragenic suppressors, and neither osmotic support nor overexpression of PKC1 or BCK1 --20 was able to rescue the inviability of ypk1Delta ykr2Delta cells (Roelants and Torrance, unpublished observations). Thus, it seems that ypk1-1ts ykr2Delta cells retain some low level of throughput at the restrictive temperature and that elevation of Pkc1-induced processes can act in conjunction with that small signal but cannot substitute completely for it. Second, just as overexpression of PKC1 was unable to rescue ypk1Delta ykr2Delta cells, overexpression of YPK1 or YKR2 could not suppress the temperature sensitivity of the pkc1-2ts mutation (Roelants, unpublished observations). Third, and most revealingly, we found that a ypk1-1ts ykr2Delta mpk1Delta triple mutant was inviable on YPGlc medium at 30C, conditions under which otherwise congenic ypk1-1ts ykr2Delta mutants, ykr2Delta mutants, mpk1Delta mutants, and ykr2Delta mpk1Delta mutants all grow well (Figure D). This synthetic lethality indicates that efficient Ypk1- and Ykr2-dependent signaling and signaling via the MAP kinase pathway downstream of Pkc1 are both required for viability. Hence, Ypk1 and Ykr2 most likely function in a pathway that acts in parallel to the Pkc1 pathway (or perhaps through Pkc1 effectors distinct from Mpk1). DISCUSSION : Relative Roles of Ypk1 and Ykr2 Protein Kinases | Although either Ypk1 or Ykr2 is able to perform their shared, essential function at normal growth temperature (30C), several lines of evidence support the conclusion that Ypk1 is the primary enzyme. First, although both ypk1Delta and ykr2Delta single mutants are alive , we found that ypk1Delta cells grow slowly at 30C, are cold sensitive, and are hypersensitive to antibiotics and caffeine, whereas ykr2Delta cells display none of these phenotypes. The dramatic cold sensitivity of ypk1Delta cells is probably explained by the fact that recent analysis of global transcription patterns by using DNA microarrays has clearly shown that YKR2 (but not YPK1) is a gene strongly induced by heat stress . Inspection of the 5'-flanking region of the YKR2 locus yields three matches to the consensus nucleotide sequence TTC(N)2 --3GAA for the binding of the heat shock transcription factor (Hsf1) at -497, -404, and -104 from the ATG initiator codon. Thus, at lower temperatures, ypk1Delta cells presumably express YKR2 poorly and behave, therefore, like cells deficient in both enzymes, which are inviable. Other data also support the conclusion that, under normal growth conditions, Ypk1 is the enzyme most important for carrying out the function(s) essential for cell viability. Overexpression of catalytically inactive Ypk1 (Ypk1-KD) had detrimental effects on cell growth, whereas an equivalent level of overexpression of the analogous kinase-dead allele of Ykr2 did not. These detrimental effects were significantly exacerbated when there was no wild-type Ypk1 present to compete with the mutant protein. In other words, the presence of wild-type Ykr2 (in a ypk1Delta cell) was insufficient to overcome the toxicity imposed by Ypk1-KD, whereas the presence of wild-type Ypk1 (in a ykr2Delta cell) was able to ameliorate this toxicity to some extent. Furthermore, at steady state, Ypk1 was located exclusively in the cytosol, whereas Ykr2 was largely sequestered in the nucleus, suggesting that cytosolic targets of these enzymes are more important for normal growth than nuclear targets. Indeed, overexpression of the catalytic domain of Ypk1 was toxic, and this toxicity may be due to mislocalization because, when tagged with 3GFP, this constitutively active fragment accumulated in the nucleus. Nonetheless, overexpression of Ykr2 can overcome all of the phenotypes of ypk1Delta cells, providing evidence that both enzymes are able to perform the same essential function(s). Thus, the phenotypic differences between ypk1Delta and ykr2Delta mutants presumably arise primarily from differential expression of the corresponding genes, as discussed above, and from the differential subcellular localization of these proteins, rather than from differences in the intrinsic specificity of these enzymes for their substrates. Division of Labor between Upstream Protein Kinases Pkh1 and Pkh2 | Pkh1 and Pkh2 are responsible for the activation of several downstream protein kinases, and we have shown that either enzyme can fulfill this function . This role is sufficient to explain why Pkh1 and Pkh2 are essential gene products because the targets of Pkh1 and Pkh2 include protein kinases that are themselves known to be required for cell viability, such as Pkc1 and, together, Ypk1 and Ykr2 . Our findings suggest, however, that Pkh1 and Pkh2 have differential roles with regard to phosphorylation and activation of Ypk1 and Ykr2. First, as measured by immune-complex kinase assays, absence of Pkh1 substantially reduced Ypk1 activity in cell extracts, whereas absence of Pkh2 did not; in contrast, absence of Pkh2 did reduce Ykr2 activity. Conversely, overproduction of Pkh1 increased Ypk1 activity in cell extracts much more than overproduction of Pkh2, whereas overproduction of Pkh2 increased Ykr2 activity more than overproduction of Pkh1. Second, as expected if Pkh1 is the primary activator of Ypk1, absence of Pkh1 (but not Pkh2) suppressed the toxicity resulting from overexpression of a constitutively active, carboxy-terminal fragment containing the catalytic domain of Ypk1. This result is in accord with our previous observations that overexpression of Pkh1 (but not Pkh2) was able to suppress the temperature sensitivity of ypk1-1ts ykr2Delta cells and that absence of Pkh1 (but not Pkh2) resulted in a detectable decrease in incorporation of [32PO43-] into Ypk1 that was immunoprecipitated from extracts of metabolically labeled cells . Third, as expected if Pkh2 is the primary activator of Ykr2, the slow growth and drug sensitivity of ypk1Delta cells were exacerbated by absence of Pkh2 (but not Pkh1). Fourth, at steady state, Pkh1 and Ypk1 are confined to the cytosol and excluded from the nucleus, whereas Pkh2 and Ykr2 are not excluded from the nucleus. Taken together, these findings support the conclusion that, in the cell, Pkh1 preferentially activates Ypk1 and Pkh2 preferentially activates Ykr2. However, several other findings indicate that this separation is by no means complete. First, when overexpressed, either Pkh1 or Pkh2 was able to stimulate either Ypk1 or Ykr2 activity over that observed in the controls. Second, in the case of Ykr2, absence of either Pkh1 or Pkh2 caused a similar reduction in Ykr2 activity. Finally, and most convincingly, all four double mutant combinations (pkh1Delta ypk1Delta, pkh1Delta ykr2Delta, pkh2Delta ypk1Delta, and pkh2Delta ykr2Delta) were recovered as viable haploids at the expected frequency after tetrad dissection of the appropriate doubly heterozygous diploid strains. Therefore, despite the apparent Pkh1-Ypk1 and Pkh2-Ykr2 dichotomy, the capacity exists for significant cross talk between these enzyme-substrate pairs. Physiological Function of Pkh1-Ypk1 and Pkh2-Ykr2 Cascades | There is compelling evidence in both yeast and mammalian cells that membrane microdomains enriched in sphingolipids and sterols, referred to as rafts, are involved in the biosynthetic delivery of certain proteins to the cell surface. Likewise, there is substantial evidence both in yeast and in animal cells that sphingolipid rafts also play a role in a clathrin-independent route of endocytosis. It has been reported recently that the sphingoid base requirement for the internalization step of endocytosis may be to activate Pkh1 and Pkh2, and that Pkc1, a known Pkh1 and Pkh2 target , acts as a downstream effector in this signaling pathway. In this regard, it is noteworthy that we found GFP-tagged Pkh1 and Pkh2 both localized primarily to prominent cortical puncta, distinct from actin patches, that seem to be membrane-associated, as judged by three-dimensional reconstruction of images taken using deconvolution fluorescence microscopy (Roelants, unpublished observations). We are currently exploring whether these structures represent microdomains enriched in sphingolipids, although there are, to our knowledge, no reliable cytological markers for such structures currently available. While our studies were in progress, it was also reported that the growth inhibitory effect of an antibiotic myriocyn (also known as ISP-1) that causes sphingolipid depletion can be overcome by overexpression of Ypk1 , suggesting that Ypk1 may also be a downstream target in response to phytosphingosine-dependent activation of Pkh1 and Pkh2, in agreement with our previous findings demonstrating that Pkh1 and Pkh2 act upstream of phosphorylate and activate Ypk1 and Ykr2 . Even more recently, it has been claimed that Ypk1 (but not Ykr2) is involved directly in endocytosis . However, proper actin assembly is critical for endocytosis in yeast , and we found that loss of Ypk1 alone had profound effects on actin organization. Thus, the apparent role of Ypk1 in endocytosis may have been inferred from a rather indirect effect that is secondary to its primary function. Several properties of ypk1-1ts ykr2Delta cells, including rapid lysis at restrictive temperature, rescue of the lysis phenotype by osmotic support, altered actin organization, and increased sensitivity to many different toxic agents, were all consistent with a cell wall defect. To gain further insight about the physiological role of Ypk1 and Ykr2, we selected both dosage suppressors and transposon insertions that restored viability to ypk1-1ts ykr2Delta cells at an otherwise nonpermissive temperature. The majority of both classes of suppressors provided additional evidence for a direct connection between Ypk1 and Ykr2 and cell wall biosynthesis. In most cases, suppression could be attributed to imposition of further cell wall damage sufficient to trigger the Pkc1-mediated activation of the Mpk1 MAP kinase-dependent cell wall integrity pathway, suggesting that Ypk1 and Ykr2 themselves are components of a novel pathway also responsible for activation of the transcription of genes involved in cell wall maintenance and remodeling. Indeed, while our article was in preparation, a study was published that also linked Ypk1 and Ykr2 to cell integrity signaling on the basis of a completely independent approach . Because the rate of phytosphingosine generation will depend on the phytoceramide concentration in the plasma membrane, the Pkh1-Ypk1 and Pkh2-Ykr2 cascades could represent a feedback control mechanism whereby membrane growth via insertion of sphingolipid-enriched vesicles is monitored and coordinately coupled to appropriate expansion of the cell wall . This signaling route seems to be distinct from, but work in parallel with, the Pkc1 pathway for cell wall maintenance, in which plasma membrane proteins (such as Wsc1/Slg1 and Mid2) serve as sensors of cell turgor pressure and function by stimulating guanine nucleotide exchange factors for the small GTPase Rho1, a known activator of Pkc1 (for review, see ). Figure 9 | Pkh1-Ypk1 and Pkh2-Ykr2 cascades and cell wall integrity signaling. Pkh1-Ypk1 and Pkh2-Ykr2 cascades and cell wall integrity signaling. For proper cell enlargement, plasma membrane growth needs to be accommodated by expansion of the cell wall. This coordinated coupling is achieved through at least two pathways. As shown previously, one route for monitoring plasma membrane growth is via the insertion of transmembrane proteins that affect the state of activation of the small GTPase, Rho1. Rho1, in turn, stimulates the protein kinase, Pkc1, and the downstream MAP kinase kinase cascade (Bck1 -> Mkk1 and Mkk2 -> Mpk1). The Mpk1 MAP kinase kinase activates transcription factors, including Rlm1, that regulate genes for cell wall enzymes. As shown in this study, a two-tiered cascade of functionally redundant protein kinases (Pkh1-Ypk1 and Pkh2-Ypk2) appears to monitor growth of the plasma membrane via a distinct mechanism. In this novel pathway, a product (phytosphingosine) derived from the vesicle-mediated insertion of sphinolipids into the plasma membrane stimulates Pkh1 and Pkh2. These enzymes, in turn, phosphorylate and activate Ypk1 and Ykr2, primarily in the cytosol and the nucleus, respectively. Genetic evidence indicates that Ypk1 and Ykr2 may act through the transcription factor, Smp1, which shares DNA-binding specificity with Rlm1. Nuclear entry is only one potential level at which the activity of these transcription factors might be regulated by these signaling pathways. Cross talk between these pathways probably occurs at two levels: full activation of Pkc1 requires phosphorylation by Pkh1 and/or Pkh2 and Ypk1 (and perhaps Ykr2) may contribute (directly or indirectly) to Mpk1 activation . See DISCUSSION for additional details. One possibility is that Ypk1 and Ykr2 feed into the known cell integrity signaling pathway by providing a Pkc1-independent route to activate Mpk1 itself. Consistent with this idea is our observation that ypk1-1ts ykr2Delta cells are inviable when Mpk1 is absent and with the fact that Mpk1 phosphorylation induced by heat stress is reduced in a ypk1Delta mutant . Moreover, the strongest suppressors we obtained were loss-of-function mutations in the Kex2-processing enzyme, which is known to cause abnormalities in the cell wall , and such abnormalities evoke Pkc1- and Mpk1-dependent signaling . Not all Pkc1-dependent responses are achieved via the Mpk1 MAP kinase pathway . However, we found that kex2Delta mutants are inviable when Mpk1 is absent, suggesting that Kex2-deficient cells can only survive because genes under Mpk1 control are activated. Likewise, it has been observed by others that kex2Delta mutants seem to have cell wall defects (Fuller, personal communication) and are inviable when they are defective in components of a signaling pathway that monitors extracellular Ca2+, including calmodulin (Cmd1) and calmodulin-activated phosphoprotein phosphatase 2B/calcineurin (Cna1, Cna2, and Cnb1) (Davis, personal communication), or the calcineurin-activated C2H2-type zinc finger transcription factor (Crz1/Tcn1) (Cunningham, personal communication), or are exposed to known calcineurin inhibitors, cyclosporin A, or FK506 (Fuller, personal communication). Strikingly, it has been shown previously that Crz1 provides an independent means to regulate many of the same genes that are under the control of the Pkc1- and Mpk1-dependent pathway . Additional evidence that absence of Kex2 leads to up-regulation of the transcriptional initiation of genes essential for viability (including, presumably, cell wall synthesis) is provided by two previously obscure findings. First, it was observed that kex2Delta mutations bypass temperature-sensitive mutations in the largest subunit of RNA polymerase II (Rpo21) and also temperature-sensitive mutations in other RNA polymerase II subunits . Second, and conversely, kex2Delta mutations are synthetically lethal when combined with otherwise viable null mutations (ppr2Delta) in the gene encoding a factor (TFIIS) needed for efficient transcriptional elongation by RNA polymerase II . However, for several reasons, we currently favor the idea that Ypk1 and Ykr2 activate genes for cell wall remodeling independently of the Pkc1-activated Mpk1 MAP kinase kinase pathway. One of the transcription factors under Mpk1 control is Rlm1 (676 residues), a member of the Mcm1, Agamous, Deficiens, Serum Response Factor-box family of transactivators . In this regard, it seems more than a coincidence that one of the dosage suppressors of the temperature-sensitive lethality of ypk1-1ts ykr2Delta cells that we isolated carries the SMP1 gene, which encodes another Mcm1, Agamous, Deficiens, Serum Response Factor-box transcription factor related to Rlm1. Smp1 (452 residues) shares near identity to Rlm1 in its similar60 residue, N-terminal, DNA-binding domain (but bears little similarity to Rlm1 beyond that), recognizes the same (but a somewhat more extended) sequence motif as Rlm1, and is even able to form heterodimers with Rlm1 . Indeed, when excised from the original isolate and expressed from a completely different multi-copy vector, elevated SMP1 expression reproducibly suppressed the temperature-sensitive growth defect of ypk1-1ts ykr2Delta cells and did so better than YPC1 and as well as EXG1 (Roelants, unpublished observations). Moreover, Smp1 contains three consensus Ypk1 phosphorylation sites (-R-x-R-x-x-S/T-Hyd-, where Hyd indicates any bulky hydrophobic residue; ), and in preliminary experiments, Smp1 can be phosphorylated by Ypk1 in vitro (Roelants, unpublished observations). Hence, currently we favor the idea that Ypk1 and Ykr2 do not act by leading to Mpk1 activation per se, but rather by providing an independent input through Smp1 that acts in parallel to or in conjunction with Rlm1 to activate genes involved in cell wall metabolism, including EXG1 . Experiments to test the above-mentioned ideas, including analysis of global transcription profiles with DNA microarrays, are underway. Backmatter: PMID- 12221113 TI - Distinct Chromosome Segregation Roles for Spindle Checkpoint Proteins AB - The spindle checkpoint plays a central role in the fidelity of chromosome transmission by ensuring that anaphase is initiated only after kinetochore-microtubule associations of all sister chromatid pairs are complete. In this study, we find that known spindle checkpoint proteins do not contribute equally to chromosome segregation fidelity in Saccharomyces cerevisiae. Loss of Bub1 or Bub3 protein elicits the largest effect. Analysis of Bub1p reveals the presence of two molecular functions. An N-terminal 608-amino acid (nonkinase) portion of the protein supports robust checkpoint activity, and, as expected, contributes to chromosome segregation. A C-terminal kinase-encoding segment independently contributes to chromosome segregation through an unknown mechanism. Both molecular functions depend on association with Bub3p. A 156-amino acid fragment of Bub1p functions in Bub3p binding and in kinetochore localization by one-hybrid assay. An adjacent segment is required for Mad1p binding, detected by deletion analysis and coimmunoprecipitation. Finally, overexpression of wild-type BUB1 or MAD3 genes leads to chromosome instability. Analysis of this activity indicates that the Bub3p-binding domain of Bub1p contributes to this phenotype through disruption of checkpoint activity as well as through introduction of kinetochore or spindle damage. Keywords: INTRODUCTION : Protein components of the spindle checkpoint were first defined genetically through studies in the budding yeast Saccharomyces cerevisiae by analysis of mutants that lack the ability to arrest in the presence of spindle damage introduced by antimicrotubule drug exposure or by manipulation of temperature conditional spindle proteins . The spindle checkpoint thus defined has been shown to control at least two functionally distinct steps within mitosis. First, at metaphase, the checkpoint acts to detect a lack of bipolar attachment or tension for any sister chromatid pair. This condition delays anaphase in the presence of even a single unattached kinetochore or a lack of tension on a single chromatid pair . Second, entry into G1 (mitotic exit) is prevented in cells that have suffered spindle damage sufficient to preclude the delivery of a daughter nucleus into the bud (for review, see ; ). Metaphase arrest due to activation of the spindle checkpoint depends upon a well-conserved pathway that regulates the degradation of the anaphase inhibitor protein Securin (budding yeast Pds1p; for review, see ; ). Anaphase is normally initiated as Separin (Esp1p) is liberated from its binding partner Securin (Pds1p) after Securin is targeted for degradation by the Cdc20-associated form of the anaphase promoting complex. Cdc20p is a target of the metaphase checkpoint arrest pathway and physically interacts with other checkpoint proteins during metaphase arrest (for review, see ; ; ). Maintenance of the arrest induced by kinetochore damage also requires arrest of the mitotic exit pathway , indicating a functional connection between the distinct control pathways that operate at anaphase initiation and mitotic exit. In budding yeast, both of these steps are inhibited by the presence of Pds1 protein , and thus anaphase and exit control may be related to one another by a key role played by Pds1p or Esp1p at both cell cycle positions . Initiation of metaphase arrest in response to a lack of bipolar attachment requires at least six proteins in S. cerevisiae: Mad1p, Mad2p, Mad3p, Bub1p, Bub3p, and Mps1p . These proteins function together in kinetochore surveillance, activating a checkpoint-governed arrest in response to microtubule defects, kinetochore protein defects, or centromere DNA mutations . The proteins involved in kinetochore surveillance are remarkably conserved in eukaryotes, and homologs have been found in fission yeast, flies, maize, frog, mouse, and human (for review, see ). In systems with robust cytology, homologus of Mad1p, Mad2p, Bub1p, Bub3p, and Mps1p have been observed to concentrate at unattached kinetochores in prometaphase. Thus, these proteins behave as expected components of a molecular structure that broadcasts an inhibitory signal that will be extinguished upon achievement of bipolar attachment and/or associated tension from spindle forces exerted in opposite directions (for review, see ; ; ). Physical association studies have shown that the metaphase arrest proteins reside in several complexes that contain overlapping components, and that these complexes exhibit alterations in a cell cycle-regulated manner. In budding yeast, Mad1p/Mad2p, Bub1p/Bub3p, and Mad3p/Bub3p complexes are detected in interphase, whereas cells in damage-induced metaphase arrest contain a Bub1p/Bub3p/Mad1p complex, as well as a Cdc20p/Mad2p/Mad3p/Bub3p complex . Moreover, at metaphase arrest, both Bub1p and Mad1p exhibit shifts in gel migration consistent with hyperphosphorylated states . Movement of constituents among protein complexes may represent the spatial communication from an activated (unattached) kinetochore to site(s) where the Cdc20-associated form of the anaphase promoting complex is poised to initiate the degradation of Pds1p. Although biochemical characterization of protein complexes has provided insight into features of checkpoint activation, the nature of the spatial regulation imposed at metaphase by the presence of unattached kinetochores has not been precisely elucidated. Indeed, it is possible that different kinetochore states, such as kinetochore-microtubule attachment or the presence of tension, may be handled at metaphase by either overlapping or distinct signaling pathways . Chromosome missegregation associated with loss of kinetochore surveillance by the spindle checkpoint has been observed. In budding yeast, observed an increase in chromosome missegregation in mad1, mad2, and mad3 mutants upon recovery from aberrant mitoses induced by exposure to the antimicrotubule drug nocodazole. In the absence of intentional spindle damage, chromosome missegregation has been detected in budding yeast bub1 and mad2 mutants as well as in Drosophila melanogaster bub1 , Schizosaccharomyces pombe Deltabub1 , and Caenorhabditis elegans mdf-1 and mdf-2 mutants . The chromosome missegregation phenotypes observed suggest that the spindle checkpoint plays a role in many cell cycles (even in the absence of induced damage), or that Bub1 and Mad2 checkpoint proteins have additional roles in kinetochore function. In this report, we present a quantitative survey of the segregation roles of five nonessential metaphase checkpoint proteins that govern kinetochore surveillance (Mad1, Mad2, Mad3, Bub1, and Bub3) in cells without additional spindle damage. We find that these spindle checkpoint proteins differ in their contributions, and that the absence of Bub1p or Bub3p has the greatest impact on segregation. Further analysis of the role of Bub1p leads to a model in which Bub1 protein provides chromosome stability through two separate mechanisms. MATERIALS AND METHODS : Yeast Media | All yeast media are as described in . Yeast Strains | Strains used in this study are listed in Table . Except where noted, experiments were conducted in an S288c laboratory background and are related by DNA-mediated transformation or isogenic mating and sporulation. Figure B, C, and E show data for strains that are derivatives of W303-1a. The one-hybrid assay was carried out in YJL128 and transformants derived from it. Table 1 | Yeast strains used in this study Figure 1 | Spindle checkpoint mutants exhibit different rates of chromosome loss. Spindle checkpoint mutants exhibit different rates of chromosome loss. (A) Null mutant sectoring phenotypes. The strains shown are wild type (YPH278), bub1Delta (YFP2), bub3Delta (YFS1100), mad1Delta (YFS1120), mad2Delta (YCD173), and mad3Delta (YFS1205). (B) Chromosome loss rates in null mutants determined by half-sector analysis. Wild type: 49 half-sectored colonies/61,276 total colonies (YPH278); 8/9,305 (YKH231). bub1Delta: 192/4,784 (YFP2); 89/2,121 (YRJ112). bub2Delta: 17/22,101 (YCD165); 3/3,440 (YRJ113). bub3Delta: 137/3,362 (YFS1100); 63/2,237 (YRJ114). mad1Delta: 139/12,394 (YFS1120); 32/8,552 (YMB111). mad2Delta: 87/10,153 (YCD173); 9/3,106 (YMB113). mad3Delta: 57/29,364 (YFS1205); 29/13,186 (YRJ111). (C) Immunoblots showing overexpression from a MET25 promoter. Extracts were taken after 2 h of inductionin media lacking methionine and were analyzed by Western blot using antibody specific for each protein. The left lane (vector, p423MET) shows the endogenous Bub1p expression level where detected; the right lane ( --MET) shows protein expressed from the MET25 promoter. All strains were generated from YKH231 by introduction of p423MET-derived plasmids containing full-length open reading frames cloned adjacent to the MET25 promoter. (D) Chromosome loss associated with overexpression of checkpoint genes. Half-sector analysis was performed after plating the strains in C on plates lacking methionine. Vector: 14/19,030. METpBUB1: 195/17,640. METpBUB3: 17/17,875. METpMAD1: 28/11,920. METpMAD2: 92/17,065. METpMAD3: 137/12,115. Two or more additional independent transformants tested for each construct showed the same chromosome instability phenotype by colony sectoring assay. (E) Benomyl sensitivity of checkpoint null mutants. Log phase cultures were spotted in a 10-fold dilution series on rich medium (YPD) or rich medium plus Benomyl. Strains were mad1Delta (YMB111), mad2Delta (YMB113), mad3Delta (YRJ111), bub1Delta (YRJ112), and bub3Delta (YRJ114). Chromosome Loss Rate | This assay was performed as previously described . Strains containing a nonessential SUP11-marked test chromosome and plasmids were grown in selective media and were plated at a density of similar200 colonies per plate on minimal (SD) medium, including 20 mug/ml uracil, 40 mug/ml l-lysine, 6 mug/ml adenine sulfate, 20 mug/ml l-histidine, 30 mug/ml l-tryptophan, and 220 mug/ml l-leucine when required to cover auxotrophies. The limiting adenine supplementation was used to facilitate red pigment development in ade2-101 cells. Chromosome loss events during the first cell division were visualized as colonies that were at least one-half red. The loss rate for the SUP11-marked chromosome is expressed as loss per chromosome per cell division, and is calculated by dividing the number of half-sectored colonies by the total number of colonies scored. Determination of the bub1-1 Mutation | The bub1-1 allele was captured on a yeast-bacterial shuttle vector by gap repair from MAY1726 , and the entire open reading frame was sequenced from two independent transformants. The sole change observed was G to A at position 997, which substitutes a conserved glutamic acid with lysine in the putative Bub3p binding region. Therefore, bub1-1 is referred to as bub1-E333K. Introduction of bub1 Mutations at the Genomic Locus | Plasmid-borne mutant alleles were created adjacent to a HIS3 marker, amplified in a single DNA fragment with the selectable marker by high-fidelity polymerase chain reaction (PCR), and integrated by homologous recombination into the native BUB1 locus. The resulting genomic structure contained the native BUB1 promoter, a mutant bub1 allele, an HIS3 downstream marker gene, and finally, natural BUB1 3'-flanking sequence. Details of the constructions are available upon request. One-Hybrid Assay | The one-hybrid assay was performed essentially as described . YJL128 was transformed with activation domain fusion constructs, and multiple independent transformants were plated on SD-LEU supplemented with 5 mM 3-amino-triazole (3-AT). Plates were incubated at 30C for up to 2 wk. MPS1 Overexpression | A GAL-MPS1 allele was created by integration of pAFS120 at the MPS1 locus of YFS589 yielding yeast strain YML101. BUB1 overexpression plasmids were introduced into YML101, and two independent transformants were picked. Cultures were grown overnight in selective media lacking histidine and uracil supplemented with 2% raffinose, diluted into selective media lacking histidine, uracil, and methionine (to derepress the MET25 promoter) supplemented with 2% raffinose and were grown to early log phase. To induce MPS1 overexpression, galactose was added to a final concentration of 3%. Samples taken at t = 0 and t = 4 h were fixed in1 M sorbitol, 50 mM KPO4, pH 7.5, and 3.7% formaldehyde, 4,6-diamidino-2-phenylindole (DAPI) stained, and scored for bud and nuclear morphology. A minimum of 200 cells was scored for each sample. Plasmids | All overexpression constructs were made in either p423MET (2mu HIS3) or p415MET (CEN/ARS/LEU2) vectors containing the methionine-repressible MET25 promoter and the CYC1 terminator sequence flanking the multiple cloning site . For one-hybrid analysis, GAL4-AD fusions were constructed by cloning each PCR-generated open reading frame into pGADT7 (, Palo Alto, CA). All plasmids generated by PCR were verified by sequence analysis. Details are available upon request. Immunoblotting and Coimmunoprecipitation | Immunoblotting and coimmunoprecipitation were carried out as described previously . The lysis buffer for coimmunoprecipitation was 50 mM HEPES, pH 7.6, 75 mM KCl, 50 mM NaF, 1 mM Na vanadate, 1 mM MgCl2, 1 mM EGTA, 0.1% Na Deoxycholate, 1 mM phenyl methyl sulfoxide, "complete EDTA-free protease inhibitor cocktail" (Roche, Indianapolis, IN), and 1 mM dithiothreitol. Rabbit alpha-Mad1, Mad2, Mad3, Bub1, and Bub3 antibodies have been previously described . RESULTS : Spindle Checkpoint Mutants Exhibit Different Rates of Chromosome Loss | In previous work, it was apparent that chromosome loss of bub1 and mad2 mutants differed from one another . To determine the requirement for spindle checkpoint proteins in accurate chromosome segregation during normal unperturbed mitosis, null mutants of six checkpoint genes (BUB1, BUB2, BUB3, MAD1, MAD2, and MAD3) were generated in an otherwise isogenic background. To follow chromosome segregation fidelity, the loss of a nonessential test chromosome was ascertained by a colony color assay . In this assay, haploid colonies containing the test chromosome bearing a SUP11 (ochre-suppressing tRNA) gene are white, whereas cells that have lost the test chromosome accumulate a red pigment due to the host ade2-101 (ochre) mutation. Thus, loss events give rise to red sectors during colony growth. The checkpoint mutant strains were plated on color indicator plates and chromosome loss rates were evaluated visually by colony sectoring morphology and by half-sector analysis . In half-sector analysis, the rate of first division missegregation events is directly measured by observing the number of colonies that are at least one-half red, and dividing by the total number of colonies that were established by cells with a test chromosome. In an S288c background, bub1Delta and bub3Delta cells exhibited the highest rates of chromosome loss, 50-fold higher than the wild-type rate of 0.8 loss events per 1000 divisions (Figure , A and B). mad1Delta and mad2Delta strains also showed an increased chromosome loss rate, but at a level two- to threefold lower than bub1Delta and bub3Delta. mad3Delta exhibited a slight increase above wild type, whereas bub2Delta was indistinguishable from control. To test the generality of this result, the null mutants were characterized in a different laboratory strain background, W303-1a. The strong phenotypes for bub1Delta and bub3Delta were again observed, but the smaller differences among the mad null mutants were less apparent in W303-1a strains. At a minimum, the chromosome loss phenotypes indicate that Bub1 and Bub3 proteins have an additional role that is important to chromosome segregation during culture in the absence of intentional spindle damage. Kinetochore surveillance checkpoint proteins perform their functions in the context of multiprotein complexes. To test whether cells are sensitive to protein dosage, each full-length open reading frame was placed under the control of the MET25 promoter (MET25p), whose transcriptional strength is controlled by altering the environmental methionine concentration . MET25p-controlled expression of the five checkpoint proteins led to steady-state protein levels in excess of wild type (Figure C). The MET25p-controlled alleles were introduced into a wild-type yeast strain containing the test chromosome for monitoring chromosome segregation. Cultures grown in the presence of methionine were diluted in water and plated at similar200 cells/plate on media lacking methionine. Half-sector analysis indicated that overexpression of Bub1p and Mad3p led to a 15-fold increase in test chromosome missegregation over the wild-type rate (0.7 loss events per 1000 divisions; Figure D). High-level expression of Mad1p and Mad2p caused a smaller increase in chromosome missegregation (three- and sevenfold), whereas high level expression of Bub3p had no effect. Commonly used assays for the presence of checkpoint deficiency measure cell survival in the presence of antimicrotubule drugs such as Benomyl. The Benomyl sensitivity elicited by the absence of each kinetochore surveillance checkpoint protein was determined using the panel of null alleles. Strain viability was tested in the presence of a concentration of drug that delays but does not arrest wild-type cell growth. Cells containing bub1 and bub3 mutations were more Benomyl sensitive than mad1, mad2, or mad3 mutants by an order of magnitude (Figure E). Thus, checkpoint proteins differ in their contribution to the maintenance of cell viability in response to mild spindle damage. Note that the order of Benomyl sensitivity correlates with the relative intrinsic chromosome loss rates observed (Figure A). In principle, Benomyl sensitivity of mutants in this assay may reflect a sum of defective mechanisms contributing to cell death, including drug-induced hindrance of microtubule dynamics, null mutant kinetochore structural defects, and inappropriate cell cycle progression. BUB1 and BUB3 Cooperate in a Chromosome Segregation Role | To determine whether the overexpression phenotype of Bub1 was due to discrete domain(s), a series of BUB1 truncation alleles was constructed capable of expressing the N-terminal 210, 367, or 608 amino acids as well as amino acid segments 211-1021 and 211 --367 (Figure A). Western analysis using a Bub1p-specific antibody raised to the N-terminal 216 amino acids indicated that MET25-promoted expression led to significant protein accumulation in cells grown in the absence of methionine (Figure B). A serial dilution analysis indicated that the full-length overexpression product reached similar50-fold that of wild type. Figure 2 | The N terminus of Bub1p contributes to the overexpression phenotype and is counterbalanced by additional BUB3. The N terminus of Bub1p contributes to the overexpression phenotype and is counterbalanced by additional BUB3. (A) Diagram of BUB1 protein and protein fragments. The boxes indicate positions of conserved regions of BUB1p. Black: Mad3 like. White: Bub3 binding. Hatched: kinase domain. Star: E333K mutation. (B) Western blot detection of BUB1 overexpression alleles. Left: Wild-type cells (YPH278) containing MET25-promoted Bub1 alleles in p423MET were grown in the absence of methionine. Western blot analysis using an antibody raised to the N-terminal 216 amino acids of Bub1p detects protein bands with migrations consistent with eachconstruct, in addition to faster migrating degradation products. Endogenous Bub1p is not detected at this exposure (vector lane). Right: A dilution series Western blot indicates that the overexpression level for full-length BUB1 is similar50-fold. (C) Chromosome missegregation induced by BUB1 overexpression alleles. Left: Full-length and partial BUB1 alleles expressed from the MET25 promoter of p423MET were introduced into a wild-type strain (YPH278). Chromosome loss was determined by half-sector analysis after plating to methionine-free medium. Vector (no insert) and pBUB1 data are from D. p[1-210]: 25/14,087. p[1-367]: 127/5,741. p[1-608]: 96/6,592. p[211-1021]: 157/10,697. p[211-367]: 247/13,781. p[211-367*]: 9/5,117. Right: Chromosome loss was determined in YPH278 containing plasmid pairs as shown. p423MET + p415MET: 1/2,616. p423MET + pBUB3: 7/4,998. p[211-367] + p415MET: 37/3,015. p[211-367] + pBUB3: 6/4,510. At least two additional independent transformants of each construct were tested by visual sectoring assay and showed the same chromosome instability phenotype. (D) Chromosome loss in bub1 and bub3 null mutants. Left: Half sector analysis was used to compare chromosome loss rates of bub1Delta, bub3Delta, and bub1Delta bub3Delta mutant strains derived from sporulation of a wild-type diploid (YCD251) into which heterozygous bub1Delta::natMX and bub3Delta::kanMX alleles were introduced by transformation. bub1Delta: 186/5019 (one spore); bub3Delta: 262/5406 (one spore); bub1Delta bub3Delta: 601/12566 (four spores). Right: Chromosome missegration in a bub3Delta strain (YFS1100) containing the vector p423MET (284/5,675) or overexpression plasmid p[211-367] (289/6,069). Each of the truncation constructs was introduced into wild-type cells on high-methionine medium, where expression is suppressed. Chromosome loss was quantitated for several independent transformants by half-sector analysis after plating on low-methionine medium (Figure C, left). Under these conditions, the full-length construct exhibits a 15-fold increase in loss (from Figure D). Overexpression of the N-terminal 367 or 608 amino acids of Bub1p from plasmids (p[1 --367] and p[1 --608], respectively) caused a 30- and 20-fold increase in chromosome loss. Overexpression of the N-terminal 210 amino acids had little effect (2.4-fold, Figure C). The segment common to p[1 --367] and p[1 --608], but absent from p[1 --210], contains a well-conserved homology box predicted to mediate association between Bub1p and Bub3p . This result suggested that overexpression of a Bub3-binding region of Bub1p might cause the chromosome loss. To test this hypothesis, a construct expressing only amino acids 211 --367 under the control of the MET25 promoter was created. It too was found to induce chromosome loss at high expression levels (26-fold greater than the vector control). Several additional lines of in vivo evidence now strongly support the interpretation that the overexpression phenotype is mediated through disruption of a Bub1p/Bub3p interaction. First, the bub1-1 point mutation , which is suppressed by a low-level increase in BUB3 gene dosage, was cloned and identified as E333K (see "Materials and Methods"). This mutation is located within the 211 --367 segment. Second, when the E333K mutation was introduced into the p[211 --367] plasmid, this allele failed to induce chromosome instability (Figure C, left panel). Third, expression of additional BUB3 from a MET25-controlled allele (from plasmid pBUB3) on a centromere vector (p415MET) reversed the chromosome instability phenotype of p[211 --367] (Figure C, right). Fourth, a BUB1/BUB3 cooperative role in chromosome segregation implied by this interpretation was tested by analyzing the chromosome loss rate of a bub1Delta bub3Delta mutant (Figure D, left). The rate observed in the double mutant (48 events in 1000 divisions) is consistent with a shared role for Bub1p and Bub3p. Finally, if the presence of excessive bub1[211 --367]p interferes with a Bub1p-Bub3p association, then this protein fragment should not elicit additional missegregation in the absence of the complex. Indeed, its overexpression does not augment chromosome loss in a bub3Delta null mutant (Figure D, right). We conclude that an interaction between Bub1 and Bub3 proteins is likely to be mediated by amino acids 211 --367 of Bub1p in vivo, and disruption of this interaction contributes significantly to the overexpression phenotype associated with excessive Bub1p. Yeast Bub1p Can Associate with Kinetochores in a One-Hybrid Assay | Previous experiments have demonstrated kinetochore localization of Bub1 protein in experimental systems where these structures are cytologically visible . To date, localization of checkpoint proteins to budding yeast kinetochore structures has not been achieved. In a one-hybrid assay , kinetochore protein components can activate a HIS3 reporter gene located immediately adjacent to the centromere of chromosome III. This reporter system depends on the presence of an active centromere, reveals association of known kinetochore components, and has been successfully used to identify new kinetochore proteins . We used this assay to ask whether kinetochore association of full-length Bub1p or truncation alleles can be detected in budding yeast. Independent transformants containing Gal4-activation domain fusions of Bub1 and Bub1 fragments were spotted onto minimal medium lacking leucine and histidine supplemented with 5 mM 3-AT. Figure shows that fusions of BUB1[1 --367], BUB1[1 --608], and BUB1[211 --367] can activate transcription of the centromere reporter, indicating that these proteins can localize to kinetochores. Reporter activation with the full-length BUB1 fusion protein is not observed, likely due to a lower level of fusion protein accumulation or the presence of a nonfunctional conformation. Figure 3 | Localization of activation domain fusions to kinetochores in a one-hybrid assay. Localization of activation domain fusions to kinetochores in a one-hybrid assay. In the one-hybrid assay, fusion of the GAL4-activation domain to a kinetochore-binding protein induces transcription of a centromere-adjacent HIS3 reporter allele . GAL4AD fusion constructs were introduced into strain YJL128. The fusion moiety is indicated to the right; GAL4AD-CTF13 (top) served as a positive control. Four independent transformants were grown to saturation in SD-LEU, diluted to 1.5 x 107 cells/ml, and spotted (3 mul) on SD-HIS, LEU + 5 mM 3-AT. The spots shown were incubated at 30C for 14 d. The large papillae that appear occasionally in the GAL4AD-BUB1 transformants (observed in similar25% of transformants) may reflect the occurrence of truncating mutations. All constructs were similarly tested in YJL148, a strain containing a mutant centromere sequence adjacent to the HIS3 reporter . No growth above vector background was observed in these controls. All fusion constructs shown were functional in a two-hybrid assay. Activation domain fusions of other kinetochore surveillance proteins were also tested. MAD1 and MAD3 fusions activated the reporter , whereas BUB3 and MAD2 fusions did not. CTF13-AD is shown as a control: it activates the HIS3 reporter and supports growth on 3-AT within 3 --4 d at 30C, whereas the checkpoint fusion proteins tested require up to 14 d to show evidence of activation over vector background. This weak signal is consistent with a transient association, as would be expected for proteins that associate with a subset of kinetochores for a subset of the cell cycle. An unequal transcriptional activation efficiency for different fusion proteins may also be a contributing factor. We conclude from these experiments that the budding yeast kinetochore surveillance proteins Bub1, Mad1, and Mad3 can associate with yeast kinetochores, as is predicted from localizations of their studied orthologs. The BUB1 Overexpression Phenotype Includes Disruption of Both Checkpoint and Segregation Functions | High-level expression of Bub1p (and fragments of this protein) may disrupt kinetochore checkpoint signaling, a segregation function, or both. To address whether checkpoint signaling was disrupted, strains overexpressing full-length Bub1p or protein fragments were tested for checkpoint competence in two different assays. The first took advantage of the spindle checkpoint-dependent delay exhibited by ctf18Delta cells, which is associated with a partial defect in sister chromatid cohesion . This delay is detected as an accumulation of G2/M phase cells during early log phase using flow cytometry (Figure A, left column). MET25-controlled alleles were introduced into ctf18Delta cells and transformants were selected on high-methionine medium. Four independent transformants were then grown in medium without methionine for 18 --24 h (O.D. similar0.4) and were analyzed for DNA content using flow cytometry (Figure A). Diminution of the G2/M phase peak indicated that the delay can be disrupted by overexpression of full-length Bub1p, Bub1[1 --367]p, Bub1[1 --608]p, and Bub1p[211 --367]p. The G2/M reduction is consistent with an observed decrease in the proportion of budded cells (Figure A), as well as a reduction in viability determined by growth on solid medium with or without methionine (Figure B). The degree of delay diminution and reduced viability correlates with the amount of chromosome loss induced by the overexpression alleles (see Figure B). We observed similar loss of delay and viability in cells lacking CTF19 (C.D. Warren, unpublished data), a gene that encodes a nonessential kinetochore protein . We conclude from these experiments that overexpression of Bub1p, and Bub1p fragments that cause chromosome loss, does have the capacity to disrupt a spindle checkpoint-dependent delay. Figure 4 | Overexpression of Bub1p or Bub1p fragments both disrupts the spindle checkpoint and causes damage. Overexpression of Bub1p or Bub1p fragments both disrupts the spindle checkpoint and causes damage. (A) Disruption of a ctf18Delta-induced checkpoint delay. ctf18Delta strains containing p423MET plasmids expressing no (vector), full-length (BUB1), or partial (1 --210, 1 --367, 211 --367, 1 --608) alleles of BUB1 were created by transformation of YFS377. Cells were grown to early log phase in the absence of methionine for 18 --24 h, and were prepared for flow cytometry (as in ). A representative histogram is given for each genotype; four independent transformants were analyzed. The fractions of budded and unbudded cells were determined and are shown as mean +- SD (3 d.f.). (B) Disruption of the ctf18Delta delay results in decreased viability. Two independent isolates from each of the ctf18Delta strains described in A were grown to early log phase in media containing methionine. Cells were spotted onto solid medium without methionine in a 10-fold serial dilution series. Overexpression of bub1-[211-367]p in wild-type or bub3Delta cells did not result in a significant reduction in mortality (bottom). (C) Competence to arrest in response to MPS1 overexpression. Log phase cells grown in raffinose media lacking methionine (to induce expression of the BUB1 alleles) were treated with 3% galactose to induce MPS1 expression. Samples were taken at t = 0 and t = 4 h, formaldehyde fixed, DAPI stained, and scored for bud and nuclear morphology. The graph shows the percentage arrested (large-budded uninucleate) cells at each time point. Two independent transformants were analyzed (average +- range indicated). All strains were derived from YML101 (GAL-MPS1). All plasmids overexpressing BUB1 alleles in this strain were derived from p423MET (vector). YCD362 (bub1Delta) was included as a control for the assay. (D) Chromosome missegregation in a mad3Delta host. Chromosome loss rates were determined by half-sector analysis on plates lacking methionine. Vector: 5/3,327. pBUB1: 18/2,295. p[1 --210]: 4/2,408. p[1 --367]: 51/2,500. p[1 --608]: 42/2,766. p[211 --367]: 37/2,349. Data for a representative transformant are shown; at least two independent transformants were analyzed for each construct. The strains were YFS1205 derivatives created by introduction of p423MET (vector) and related BUB1 allele overexpression plasmids. In a second system, we obtained evidence that the checkpoint arrest pathway is not completely dysfunctional. In this experiment, checkpoint activation by overexpression of the MPS1 protein kinase was used to cause cell cycle arrest . MPS1-induced arrest is dependent on each of the known BUB and MAD checkpoint genes . If overexpression of Bub1p or Bub1p fragments completely disrupts the checkpoint pathway, similar to a null mutant, then MPS1 overexpression will not cause cell cycle arrest. Wild-type strains containing integrated GAL1-MPS1 as well as MET-controlled BUB1 overexpression plasmids were grown in medium lacking methionine to induce high-level expression of the BUB1 truncation alleles. Samples were taken before and 4 h after addition of galactose (for overexpression of MPS1). Formaldehyde-fixed cells were stained with DAPI and were scored for morphological evidence of metaphase arrest. Strains overexpressing the full-length Bub1p, Bub1[1 --367]p, or Bub1[1 --608]p arrested in response to MPS1 overexpression, whereas control bub1Delta cells did not (Figure C). The presence of an MPS1-induced arrest indicates that the overexpression of Bub1 full-length protein or the truncation alleles does not fully abrogate spindle checkpoint function. We speculate that even a single remaining active kinetochore may be sufficient to arrest cells in response to MPS1 overexpression. Both of the experiments above address the checkpoint competence of strains containing extra Bub1p or Bub1p fragments. Neither addresses whether disruption of checkpoint control is responsible for the chromosome loss introduced by overexpression of Bub1p, or whether a separate mechanism causes missegregation (e.g., competition for a kinetochore structural component). Note that mad3Delta null cells have a quite modest chromosome instability phenotype (Figure , A and B), although they are markedly defective in preanaphase arrest (; ; our unpublished data). To explore the cause of chromosome loss, overexpression interference of BUB1 alleles was tested in a mad3Delta yeast host (Figure D). The chromosome loss rates observed closely parallel those induced by the Bub1 overexpression alleles in wild-type cells. This result indicates that the mechanism responsible for chromosome loss incorporates a defect distinct from loss of a functional checkpoint pathway. Two Domains of Bub1p Play Distinct Roles in Chromosome Segregation | Because the molecular defects engendered by overexpression may be complex, genomic loss-of-function alleles have also been characterized. Chromosome loss rates of two bub1 missense mutants were measured by half-sector analysis (Figure , A and B). The bub1-1 allele (i.e., bub1-E333K) has apparent partial function because its checkpoint defect is suppressible by additional copies of the BUB3 gene on a centromere plasmid, whereas that of a bub1 null mutation is not . However, the chromosome missegregation rate measured for bub1-1 (Figure , A and B) is similar to that of bub1Delta (Figure , A and B). Like the bub1-1 checkpoint defect, the bub1-1 chromosome missegregation phenotype is suppressible by extra copies of the BUB3 gene introduced on a centromere plasmid (Figure B). Figure 5 | Chromosome missegregation associated with genomic alleles of BUB1. Chromosome missegregation associated with genomic alleles of BUB1. (A) Colony sectoring phenotypes of bub1-E333K (bub1-1) and bub1-K733R. Strains were YCD280 and YCD281. (B) Chromosome loss caused by genomic alleles. Left: Half sector analysis was used to analyze the mutants as shown. WT: 8/8,503. bub1-E333K: 248/5,964. bub1-K733R: 83/6,576. bub1[1-210]: 354/6,769. bub1[1-367]: 477/9,019. bub1[1-608]: 222/14,105. bub1[211-1021]: 355/14,837. Strains were YCD279, YCD280, YCD281, YCD371, YCD358, and YCD407. Right: A centromere plasmid containing a MET25-inducible BUB3 gene (pBUB3) or vector alone (p415MET) was introduced into wild-type (WT, YPH278), bub1-E333K (YCD280), or bub1-K733R (YCD281) strains. Half sector analysis was performed after plating on methionine-free media. WT + p415MET: 1/3,728. WT + pBUB3: 3/4,925. bub1-E333K + p415MET: 104/1,567. bub1-E333K + pBUB3: 48/3,941. bub1-K733R + p415MET: 53/3,654. bub1-K733R + pBUB3: 36/4,102. The bub1-K733R allele, which alters a conserved lysine residue in the Bub1 protein kinase domain, has been previously characterized as deficient in checkpoint competence and protein kinase activity . The chromosome stability defect of bub1-K733R is less severe than that of bub1-E333K (Figure , A and B). The addition of extra BUB3 gene copies to bub1-K733R does not markedly alter its chromosome segregation phenotype. BUB1 truncation alleles were tested at single copy in the native genomic locus under the control of the BUB1 promoter in haploid cells. Although the genomic bub1[1-367] allele exhibits a phenotype similar to the null mutant, the bub1[1-608] allele supports a chromosome loss rate that is intermediate (16 events per 1000 divisions, Figure B). This rate is similar to that observed for the kinase region missense mutant bub1-K733R (13 events per 1000). Thus, an intermediate level of chromosome stability is observed for two alleles of BUB1, with defective or absent protein kinase activity, indicating a role for the N-terminal portion of Bub1p in segregation. Note that the chromosome segregation competence conferred by the N-terminus of Bub1p does not account for the very high fidelity of segregation in wild-type cells. To test if chromosome stability can be provided by a Bub1p C-terminal protein fragment, a deletion allele expressing amino acids 211-1021 was constructed at the genomic locus. This mutant has a chromosome stability phenotype that is between wild-type and bub1Delta, at 24 events per 1000 (Figure B). We conclude that the C-terminal portion of Bub1p also contributes to chromosome stability. In summary, the bub1-E333K allele appears to be virtually null for both checkpoint and chromosome segregation activities, although it encodes a protein whose functions are rescued by additional expression of its binding partner Bub3p. This argues that an association with Bub3p is required for both checkpoint and segregation activities. Separate N-terminal (bub1[1-608]) and C-terminal (bub1[211-1021]) protein fragments contribute to chromosome stability, each providing an intermediate level of segregation fidelity. We note that the loss rates of these two partial protein alleles sum to a value that is the same as the loss rate observed in the null mutant (40 per 1000, Figure B). An N-Terminal Segment of Bub1p Is Necessary and Sufficient for Its Checkpoint Function | Genomic loss-of-function alleles were tested for checkpoint competence by evaluating arrest after spindle damage. An arresting concentration of the antimicrotubule drug nocodazole (15 mug/ml) was added to asynchronous cultures grown in rich medium at t = 0. Samples at 4, 6, and 8 h were fixed in formaldehyde, DAPI stained, and scored for the frequency of uninuclear large-budded (arrested) or multibudded (inappropriately progressing) cellular phenotypes (Figure , A and B). The bub1[1-367] and bub1-E333K mutants behaved like bub1Delta, consistent with their null chromosome instability phenotypes. However, bub1[1-608] gave results similar to wild type for both arrest and inappropriate progression tests. This indicates that the kinase domain can be deleted without loss of the checkpoint arrest function of Bub1p. Figure 6 | Checkpoint competence of BUB1 genomic alleles. Checkpoint competence of BUB1 genomic alleles. (A) Cell cycle arrest. Logarithmically growing cultures of strains containing integrated alleles were transferred to YPD + 15 mug/ml nocodazole. At t = 0, 4, 6, and 8 h after shift into nocodazole, aliquots were formaldehyde fixed and stained with DAPI. Two hundred cells from each were scored for the arrested fraction (large-budded uninucleate cells), and the mean +- standard deviation for three independent integrants is shown. The strains were YPH278, YFP2, YCD371, YCD358, YCD281, and YCD280. (B) Failure of cell cycle arrest. The same samples were scored for the fraction of cells that exhibited multibudded uninucleate cells, an indication mitotic exit in the absence of nuclear division. (C) Anti-Bub1p immunoprecipitates from the genotypes indicated were analyzed by Western blot for the presence and abundance of Bub1 protein (as described in ). Immunoprecipitation product from the wild-type cells was loaded in a dilution series (1, 0.5, and 0.25) for comparison with lanes containing immunoprecipitations from mutant extracts. The strains were YFP2, YPH278, YCD280, and YCD281. The previous report that bub1-K733R is checkpoint deficient led to the hypothesis that the kinase-encoding portion of BUB1 is the checkpoint-functional moiety of the protein. We observe that although bub1-K733R cells are indeed checkpoint deficient, the timing of arrest failure indicates the presence of partial function (Figure A). Moreover, the bub1-[211-1021] allele failed to exhibit a checkpoint arrest in nocodazole. These results, taken together with the arrest competence of bub1-[1-608], indicate that the protein kinase activity of Bub1p is not responsible for nocodazole-induced arrest. The proteins encoded by bub1-E333K and bub1-K733R alleles were further investigated. Western blot analysis of anti-Bub1p immunoprecipitates reveals the presence of a stable protein pool in bub1-E333K mutant cells (Figure C). The bub1-E333K protein is significantly underphosphorylated, strongly suggesting that function of the wild-type Bub1 protein depends upon its phosphorylation. In contrast, bub1-K733Rp appears to be less abundant, and modified forms are readily detected (Figure C). The low steady-state abundance of bub1-K733Rp may reflect a high protein turnover rate. This prediction suggests an hypothesis in which the inability of bub1-K733Rp to maintain a checkpoint arrest is in part due to a gradual loss of the mutant protein in arrested cells. Formation of a Mad1p-Bub1p-Bub3p complex is crucial for spindle checkpoint function . Therefore, we tested whether the Bub1 protein fragment alleles could form such a complex by assaying for coimmunoprecipitation with Mad1p or myc-tagged Bub3p . First, immunoprecipitates prepared with an alpha-Bub1p antibody were characterized for the presence of Bub1p and Mad1p (Figure A). Full-length Bub1p and bub1[1-608]p expressed from the genomic locus were found to coprecipitate Mad1p in nocodazole-arrested cells, whereas bub1[1-367]p did not. Second, immunoprecipitation was carried out to test for association between a genomic myc-tagged BUB3 allele and Bub1 truncation proteins expressed from the MET25 promoter on a 2-mum plasmid (Figure B). Anti-myc precipitates containing equivalent amounts of Bub3-myc protein (Figure B, bottom) also contained appreciable amounts of full-length Bub1p, bub1[1-367]p, and bub1[1-608]p, but not bub1[1-210]p. Figure 7 | Bub1-[1-608]p associates with Mad1p and Bub3p. Bub1-[1-608]p associates with Mad1p and Bub3p. (A) Coimmunoprecipitation of full-length Bub1p and bub1-[1-608]p with Mad1p. The strains shown, containing integrated Bub1 alleles expressed from the wild-type BUB1 promoter, were grown to log phase and were incubated with +- 15 mug/ml nocodazole for 2 h at 24C. Immunoprecipitates were prepared using an alpha-Bub1p antibody, separated by SDS-PAGE, and transferred to nitrocellulose. The immunoblots were then probed with alpha-Bub1p and alpha-Mad1p rabbit antibodies as indicated. The strong band labeled (*) in the Bub1 blot is IgG heavy chain from the immunoprecipitation. Strains shown are YPH278, YFP2, YCD358, and YCD371. (B) Coimmunoprecipitation of full-length Bub1p, bub1-[1-367]p, and bub1-[1-608]p with Bub3p. All strains contained a BUB3-myc allele in the genome. The experimental strains contained a wild-type BUB1 gene in addition to episomal MET25-promoted alleles as indicated. A bub1Delta strain served as control. Left: Bub1p Western blot using a rabbit alpha-Bub1p antibody. Right: Immunoprecipitation with an alpha-myc antibody recovered an equivalent amount of myc-tagged Bub3 protein (bottom). The immunoprecipitates were probed with rabbit alpha-Bub1p antibody (top). The strains were YKH300 (bub1Delta) or YKH238 with pBUB1, p[1-210], p[1-367], or p[1-608]. In summary, bub1[1-608]p exhibits biochemical characteristics of a functional Bub1 protein capable of coprecipitation with both Mad1p and Bub3p. Moreover, bub1[1-608]p is heavily phosphorylated in all of our Western blots (Figures B and ; confirmed by lambda protein phosphatase treatment; K.G. Hardwick, unpublished data), whereas bub1[1-367]p and bub1[1-210]p are not. In a functional assay, a genomic allele of bub1[1-608] supports a robust checkpoint arrest in the presence of nocodazole. Thus, we conclude that the bub1[1-608] protein is sufficient for BUB1 checkpoint arrest function, and exhibits biochemical properties expected for this activity. DISCUSSION : Nonessential spindle checkpoint proteins from budding yeast differ in their importance to chromosome stability in cells where spindle assembly dynamics are not challenged by intentional introduction of damage. The disparity in chromosome loss rates observed among the checkpoint null mutants indicates the presence of functional differentiation. bub2Delta cells exhibit a wild-type chromosome loss rate, in agreement with BUB2's primary role in mitotic exit rather than in kinetochore surveillance at metaphase. Each of the other mutants conferred a chromosome loss rate higher than wild type, indicating one or more roles important for high-fidelity chromosome transmission. BUB1 and BUB3 genes in particular appear to influence chromosome segregation more strongly than MAD1, MAD2, and MAD3 genes. We speculate that differential roles among these genes may include distinct kinetochore structural contributions that influence segregation, detection of different types of kinetochore status in the context of checkpoint signaling (e.g., tension vs. attachment), or communication of checkpoint signaling to diverse target molecules that mediate different aspects of checkpoint delay or recovery. It was recently argued that although mammalian Mad2p responds to the lack of microtubule attachment, the Bub proteins respond to both microtubule attachment and a lack of tension . However, evidence from budding yeast suggests that the spindle checkpoint in this organism responds to the lack of tension in a mitotic spindle, and that this checkpoint-associated delay is Mad2 dependent . Further work is needed to clarify roles of the checkpoint proteins. In this work, we have endeavored to explain the relatively high rate of loss exhibited by bub1Delta cells and to find evidence for the presence of two distinct contributions to chromosome segregation. One is encoded within the first 608 amino acids in a protein segment that is both necessary and sufficient for a nocodazole-induced checkpoint arrest. The other is encoded in the kinase domain, which is not required for checkpoint arrest and whose function is unknown. Previous work in budding yeast has indicated that a missense allele predicted to disrupt kinase activity (bub1-K733R) was also defective in checkpoint arrest . In apparent contradiction, an in vitro experiment using a Xenopus extract system has provided evidence that a kinase-defective missense allele can support an active checkpoint . Here, we find that the genomic bub1-[1-608] allele, entirely lacking the conserved kinase domain, exhibits checkpoint competence after spindle disruption, whereas bub1-K733R exhibits a transient arrest that decays rapidly. Examination of the steady-state abundance of bub1-K733R encoded protein indicates a decreased accumulation. Taken together, these studies indicate that the checkpoint defect associated with bub1-K733R is more likely due to insufficient gene product than to a dysfunctional kinase activity. The protein encoded by bub1-[1-608] exhibits several interesting properties relevant to its checkpoint function. The immunoprecipitation experiments reveal association of this truncation product with both Bub3p and Mad1p. The BUB1 partial protein allele series indicates the involvement of specific amino acid segments of Bub1p in complex formation. The segment from amino acid 211 to 367 is required for complex formation with Bub3p. Similarly, the segment from 367 to 608 is required for Mad1p association. In our analysis of the partial protein alleles, the presence of both Bub1p and Mad1p binding correlates with the accumulation of phosphorylated forms of Bub1p, as well as the presence of checkpoint arrest competence. These observations strongly support the current model that a Bub1-Bub3-Mad1 protein complex is required for checkpoint arrest, and they suggest that the phosphorylation in the N-terminal one-half of Bub1p may also be a requirement. The 211 --367-amino acid segment can localize a GAL4 transcriptional activation domain to the yeast kinetochore in the one-hybrid assay. This activity, as well as Bub3p binding, is consistent with previous work on murine Bub1p, which defined a conserved homology with similar functions in an overexpression assay. In general, our overexpression results in budding yeast parallel studies in mammalian cells where overexpression of Bub1p mutant alleles from an ectopic promoter leads to disruption of checkpoint function . However, the mammalian studies have been controversial (see ) due to differing outcomes from similar experiments. In budding yeast, under partial induction of the checkpoint (e.g., in ctf18Delta or ctf19Delta cycling populations), overexpression of Bub1p or fragments was sufficient to "silence" checkpoint signaling. We assume that in ctf18Delta or ctf19Delta mutants, many cells experience a delay due to the failure of one kinetochore (or a few) to achieve stable bipolar attachment with normal timing. In contrast, under the same Bub1-overexpression conditions, checkpoint activation by extra Mps1p was sufficient for cell cycle arrest. We speculate that because the checkpoint is strongly induced with overexpression of Mps1p, even a single remaining active checkpoint-signaling complex may cause cell cycle arrest. Comparison of the results from these two tests for checkpoint function in yeast highlights a cautionary note where partial induction or disruption of checkpoint activity is involved. For example, in vertebrate cell culture systems, seemingly subtle variation (e.g., in genotype or culture conditions) may contribute to quantitative aspects of checkpoint competence and may affect the outcome. In the overexpression survey of checkpoint proteins, extra Mad3p caused a chromosome loss rate similar to that conferred by extra Bub1p. Although Mad3p exhibits similarity in protein alignment to the Bub1p N-terminal segment, each gene is independently required for checkpoint activity and, therefore, they are not functionally equivalent. The mad3 null chromosome loss rate is notably subtle in comparison with the MAD3 overexpression phenotype, indicating that compromise of Mad3p's overexpression binding partners is more important to segregation than Mad3 protein itself. Because the interaction between Bub1p and Bub3p contributes to the BUB1 overexpression phenotype, and because Mad3p associates with Bub3p , it is likely that interference with Bub3p function is causal for the chromosome missegregation induced by Mad3p overexpression. Interestingly, the amounts of Bub3p at a single human kinetochore have been estimated to be around 1000 copies , an abundance that is suggestive of its having a structural role as well as a signaling one. In conclusion, a quantitative study of the roles played by spindle checkpoint genes in chromosome segregation indicates the presence of functional differentiation beyond their essential contributions to the spindle checkpoint. Further studies of loss-of-function alleles that define distinct functional contributions, and overexpression alleles that disrupt in vivo relationships, hold promise for elucidating the in vivo importance of biochemical properties of checkpoint components. Backmatter: PMID- 12221114 TI - Activation of Mitogen-activated Protein Kinase (Mitogen-activated Protein Kinase/Extracellular Signal-regulated Kinase) Cascade by Aldosterone AB - Aldosterone in some tissues increases expression of the mRNA encoding the small monomeric G protein Ki-RasA. Renal A6 epithelial cells were used to determine whether induction of Ki-ras leads to concomitant increases in the total as well as active levels of Ki-RasA and whether this then leads to subsequent activation of its effector mitogen-activated protein kinase (MAPK/extracellular signal-regulated kinase) cascade. The molecular basis and cellular consequences of this action were specifically investigated. We identified the intron 1-exon 1 region (rasI/E1) of the mouse Ki-ras gene as sufficient to reconstitute aldosterone responsiveness to a heterologous promotor. Aldosterone increased reporter gene activity containing rasI/E1 threefold. Aldosterone increased the absolute and GTP-bound levels of Ki-RasA by a similar extent, suggesting that activation resulted from mass action and not effects on GTP binding/hydrolysis rates. Aldosterone significantly increased Ki-RasA and MAPK activity as early as 15 min with activation peaking by 2 h and waning after 4 h. Inhibitors of transcription, translation, and a glucocorticoid receptor antagonist attenuated MAPK signaling. Similarly, rasI/E1-driven luciferase expression was sensitive to glucocorticoid receptor blockade. Overexpression of dominant-negative RasN17, addition of antisense Ki-rasA and inhibition of mitogen-activated protein kinase kinase also attenuated steroid-dependent increases in MAPK signaling. Thus, activation of MAPK by aldosterone is dependent, in part, on a genomic mechanism involving induction of Ki-ras transcription and subsequent activation of its downstream effectors. This genomic mechanism has a distinct time course from activation by traditional mitogens, such as serum, which affect the GTP-binding state and not absolute levels of Ras. The result of such a genomic mechanism is that peak activation of the MAPK cascade by adrenal corticosteroids is delayed but prolonged. Keywords: INTRODUCTION : Small, monomeric Ras GTP-binding proteins initiate pleiotropic signaling cascades to affect many aspects of cellular physiology. Ras signaling through the extracellular signal-regulated kinase (ERK) cascade mediated by mitogen-activated protein kinases (MAPKs) 1/2, for instance, is well documented to play a pivotal role in cellular growth and differentiation. Protein hormones, which target GTP exchange factors and GTPase-activating proteins via plasma membrane receptors, activate the MAPK cascade by increasing the GTP-bound state of Ras proteins and not the absolute levels of these proteins. Thus, the "classic" paradigm of Ras -> MAPK signaling involves posttranslational control of Ras. Emerging evidence suggests that numerous steroids, including aldosterone, also affect Ras signaling . The molecular basis and end effect of this steroid action remain, for the most part, not well described. Because steroids control cell activity through receptors that function as trans-acting factors to modulate gene expression, it is possible that steroids act on the Ras signaling cascade via a "genomic" mechanism that is dependent on transcription and subsequent translation to increase Ras protein levels and thus, distinct from the classic mechanism. The current study, which investigated this possibility, identifies a novel paradigm by which corticosteroid activate the Ras -> MAPK signaling cascade. The adrenal cortical steroid hormone aldosterone is the major endocrine factor regulating Na+ and K+ homeostasis. Aldosterone, consequently, plays a central role in maintaining electrolyte and water balance . Aldosterone also plays a direct role in pathological remodeling of the heart, possibly by promoting fibrosis and cellular proliferation both of which are generally known to be impacted by Ras signaling via ERK cascades . Although the systemic effects and target tissues of aldosterone are well known, little is actually known about its cellular mechanisms of action. Many integral membrane proteins involved in epithelial cell transport, such as the epithelial Na+ channel (ENaC), apical membrane potassium channel, Na+/Cl- cotransporter, H+/K+-ATPase, and Na+/K+-ATPase are end effectors of aldosterone signaling . Although aldosterone affects cell activity by modulating gene expression, the expression levels of these proteins involved in transport, however, are not themselves initially controlled by the steroid. This has led to the proposal that aldosterone must control expression of factors that initiate or impinge upon signal transduction. Adrenal corticosteroids, including aldosterone, increase the levels of the small, monomeric GTP-binding protein Kirsten Ras (Ki-Ras; ; ; ; ). Aldosterone preferentially increases expression in epithelia of the A splice variant of Ki-Ras via control of transcription with induction of Ki-ras mRNA being a primary response to steroid that is independent of de novo protein synthesis and begins within 30 min after steroid addition. Induction of Ki-RasA is necessary and sufficient for aldosterone action, in part, on Na+ transport . In addition, Ki-RasA activates ENaC when both proteins are overexpressed in a heterologous system and increases the open probability of this channel in native epithelia . Consequently, Ki-RasA is a likely candidate in some instances to transduce information form the nucleus to final effectors in response to aldosterone. Induction of Ras expression by steroids may impact more than just ENaC and epithelial transport for glucocorticoids and estrogen increase Ras expression in mammary epithelia with enhanced expression possibly being associated with tumor formation and metastasis . Indeed, it has long been recognized that in cells that lack mutant Ras, elevated levels of normal Ras can lead to cell growth and/or transformation, presumably through inappropriate stimulation of Ras effector cascades . This suggests that through mass action, induction of Ras leads to activation of this protein and subsequent signaling. The general consequences and in particular those associated with cell signaling of steroid-dependent induction of Ki-ras are not well understood. It also is not clear whether increases in Ki-RasA levels in response to aldosterone result from actions mediated by nuclear receptors and whether steroid-sensitive increases in Ki-RasA result in concomitant increases in functional GTP-bound Ki-RasA. Similar to the other Ras proteins (Ha-Ras, N-Ras, and Ki-RasB), active Ki-RasA initiates many different intracellular signaling cascades, including the MAPK cascade. This cascade is known to affect several aldosterone-target proteins, such as ENaC, Na+/K+-ATPase, Na+/H+ exchanger, Na+/Cl-, and Na+/bicarbonate cotransport; and Na+/Ca2+ exchange proteins . Thus, the MAPK cascade may play a pivotal role in signaling aldosterone action secondarily to stimulation of Ki-RasA or may ultimately be involved in a negative feedback pathway initiated by this steroid. The current work tested the hypothesis that aldosterone-stimulated Ki-RasA activates the MAPK cascade in renal epithelia. In addition, we asked whether Ki-RasA and the MAPK cascade are activated in response to aldosterone via nuclear steroid receptors, and whether increases in Ki-RasA expression in response to steroid result in increases in functional Ki-RasA:GTP levels. Through the course of this work, we also investigated possible molecular mechanisms by which aldosterone induces Ki-Ras expression and compared aldosterone effects on Ki-Ras -> MAPK signaling with that of a traditional mitogen, such as serum. EXPERIMENTAL PROCEDURES : Cell Culture | All experiments were performed with renal A6 epithelial cells (passages 75 --81; American Type Culture Collection, Manassas, VA). Cells were cultured on polycarbonate supports (Transwell-Clear Inserts, pore size 0.4 muM, growth area 4.7 cm2; , Cambridge, MA) and allowed to form polar monolayers by using standard methods described previously (,, , ). In brief, cells were maintained at 26C in 4% CO2 with complete amphibian medium (3/10 Coon's F-12, 7/10 Leibovitz's L-15) supplemented with fetal bovine serum (10%). Basic medium was devoid of serum and aldosterone. High-resistance (>2 KOmega), polarized A6 cell monolayers were used for all experiments. To observe the full action of aldosterone, confluent cells were treated with basic media for 48 to 72 h before experimentation. Molecular Biology | Plasmid Preparation and Isolation of Ki-ras Intron 1-Exon 1. | The pMMrasDN plasmid was a kind gift form Dr. G. Firestone (University of California at Berkeley, Berkeley, CA). In brief, this construct allows glucocorticoid-inducible expression of dominant-negative Ha-RasN17. Similar to that described previously by the Firestone laboratory for Con8 rat mammary epithelial cells , this construct in conjunction with G418 selection was used to create clonal A6 cell lines stably expressing inducible dominant-negative RasN17. The firefly luciferase reporter plasmid pGL2-TK was generated by subcloning the minimal herpes simplex virus thymidine kinase promotor from pRL-TK (Promega, Madison, WI) into pGL2 Basic Vector (Promega) with HindIII and BglII. (pGL2-TK was a kind gift from Dr. A. Firulli, University of Texas, San Antonio, TX.) The control pRL-CMV plasmid contains the cytomegalovirus promotor upstream of Renilla luciferase (Promega). Mouse c-Ki-ras2 exon 1 plus its 5'-flanking region (intron 1-exon 1 region; nucleic acids -165 --153 as labeled from the adenosine of the translation start codon ATG in exon 1; see GenBank accession numbers , 52798, , and ; ; ) were amplified with a standard polymerase chain reaction by using mouse whole tail genomic DNA and the 5'-ATGCGGTACCGACTTACAGGTTACTC (incorporating KpnI site, underlined) and 5'-GCATCTCGAGCTGCCGTCCTTTACAAGCG (incorporating XhoI site, underlined) upstream and downstream primers, respectively. The 321-base pair product from this polymerase chain reaction was subcloned into pGL2-TK with KpnI and XhoI to produce pGL2-TK-rasI/E1. Luciferase Reporter Gene Assay. | A quantitative assay with a Renilla luciferase internal control was used to measure the firefly luminescent signal in A6 cells overexpressing reporter genes. In brief, A6 cells plated at 80% confluence on 100- x 20-mm2 culture dishes were transfected with 100 ng of pRL-CMV in addition to 3 mug of the firefly luciferase reporter plasmid (either pGL2-TK or pGL2-TK-rasI/E1) by using the LipofectAMINE Plus (Invitrogen, Carlsbad, CA) system per the manufacturer's instructions with the exception that cells were exposed to transfection reagents for similar8 h. Twenty-four hours after transfection and 24 h before performing assay, cells were replated in a 96-well culture plate. Luciferase activity then was measured with the Dual-Luciferase Reporter assay system (Promega) per the manufacturer's instructions directly following the experimental treatment period (i.e., exposure to 1.5 muM aldosterone for 4 h.) and 1-h extract preparation period required with passive lysis buffer (see Dual-Luciferase Reporter instructions). An MLX microtiter plate luminometer (Dynatech Labs, Chantilly, VA) was used to record luminescent signal. For these experiments, all firefly luciferase activity data are normalized to the internal Renilla luciferase control. Biochemistry | Western Blot Analysis. | Whole A6 cell lysate was extracted after three washes with Tris-buffered saline by using standard procedures . Cells were scraped and then maintained for 1 --2 h at 4C in gentle lysis buffer (GLB) (76 mM NaCl, 50 mM HCl-Tris, 2 mM EGTA plus 1% Nonidet P-40, and 10% glycerol, pH 7.4) and protease inhibitors (phenylmethylsulfonyl fluoride, leupeptin, tosylphenylalanyl chloromethyl ketone, and 1-chloro-3-tosylamido-7-amino-2-heptanone). For Western blot analysis of phosphorylated proteins, GLB was supplemented with 0.1 mM NaPPi, 0.5 mM NaF, 0.1 mM Na2MoO4, 0.1 mM ZnCl2, and 0.04 mM Na3VO4 prepared fresh from 1000x stocks. After clearing cellular debris, standardizing total protein concentration, and addition of Laemmli sample buffer (0.005% bromphenol blue, 10% glycerol, 3% SDS, 1 mM EDTA, 77 mM HCl-Tris, and 20 mM dithiothreitol), lysates were heated to 85C for 10 min. Proteins were then separated by standard SDS-PAGE and subsequently electrophoretically transferred to nitrocellulose (0.2 muM). Western blot analysis was performed using standard techniques and appropriate antibodies (,,, ; see below; primary and secondary antibodies were used at 1/1000 and 1/20000, respectively). Tween 20 (0.1%) and 5% dried milk (Carnation) were used as blocking reagents. Band intensity was quantified with densitometric scanning using Sigmagel (Jandel Scientific, Costa Madre, CA). When possible, the flood configuration with the highest practical threshold was used to measure band density. Western blots were often stripped of primary and secondary antibody to subsequently reprobe with a control antibody. All Western blots were stripped in 100 mM 2-mercaptoethanol, 62.5 mM Tris-HCl, pH 6.7, and 2% SDS for 30 min at 55C with constant agitation. After removal of antibodies, nonspecific interactions were reblocked by incubating in TBS-Tween, 5% milk for 2 h before reprobing with primary antibody. Ras:GTP Assay. | Raf-1 RBD agarose was from Upstate Biotechnology (Lake placid, NY). This immobilized fusion protein corresponds to the human Raf-1 Ras binding domain (RBD) (residues 1 --149). Raf-1 RBD binds Ras complexed with GTP. Pull-down experiments were performed in 400 mul (0.4 mg of total protein) of whole A6 cell lysate isolated with GLB. Lysates were incubated with 30 mul of Raf-1 RBD agarose overnight (at 4C and with constant agitation); pellets were washed five times with 2 volumes of fresh GLB each time for a total wash time of 2 h; and after resuspending in sample buffer and heating, Raf-1 RBD agarose precipitated proteins were separated by SDS-PAGE and Ki-RasA:GTP identified by immunoblotting. MAPK Assay. | MAPK activity in lysates prepared from cells treated with and without aldosterone was assayed by quantifying phosphorylation of exogenous myelin basic protein (MBP). MAPK activity was measured in whole A6 cell lysate (2 mg/ml) extracted in the presence of phosphatase inhibitors as described above. MAPK activity was measured for 30 min at 30C in the following assay dilution buffer (ADB; Upstate Biotechnology): 20 mM MOPS pH 7.2, 25 mM beta-glycerophosphate, 5 mM EGTA, 0.4 mM MnCl2, 0.4 mM CaCl2, 1 mM sodium orthovanadate, and 1 mM dithiothreitol. The final reaction contained 10 mul of substrate cocktail (from stock of 2 mg/ml dephosphorylated MBP in ADB), 10 mul of inhibitor cocktail (from stock of 20 muM PKC inhibitor peptide, 2 muM PKA inhibitor peptide, and 20 muM compound R24571 in ABD), 10 mul of A6 cell extract, and 10 mul of Mg2+/ATP cocktail (from a stock of 75 mM MgCl2, 500 muM ATP in ADB). Reactions were initiated with Mg2+/ATP and terminated with Laemmli sample buffer (described above). Phosphorylation of MBP (in 10 mul of final reaction) was assessed after SDS-PAGE by immunoblot analysis with a specific anti-phospho-MBP antibody. Electrophysiology | Transepithelial Na+ current was calculated as described previously (,,, ), from Ohm's law as the ratio of transepithelial voltage to transepithelial resistance under open circuit conditions by using a Millicel Electrical Resistance System with dual Ag/AgCl pellet electrodes (Millipore, Bedford, MA) to measure voltage and resistance. Materials | All reagents unless indicated otherwise were from either BIOMOL Research Laboratories (Plymouth Meeting, PA), Calbiochem (San Diego, CA), Invitrogen, or Sigma-Aldrich (St. Louis, MO). Phosphorothiate oligonucleotides were synthesized by the Emory University Microchemical Facility and stored frozen as 10 mM (in water) stocks. Aldosterone, dexamethasone, and mifepristone (RU486) were stored frozen as 1.5, 0.1, and 1.0 mM (in dimethyl sulfoxide [DMSO]) stocks. Cycloheximide (in MeOH) and emetine (in H2O) were stored frozen as 1.0-mg/ml stocks. Actinomycin D was stored at 4C as a 1.0-ng/ml (in MeOH) stock. PD-98059 and U-0126 were prepared fresh (in DMSO) before each experiment at stock concentrations of 10 and 5 mM, respectively. All reagents used for Western blot analysis unless noted otherwise were from (Hercules, CA) and (Rockford, IL). For each lysate, protein concentration was determined with the bicinchoninic acid protein assay. Kodak BioMax Light-1 film and Chemiluminescence Reagents Plus (PerkinElmer Life Sciences, Boston, MA) were used to develop Western blots. Antibodies | The rabbit polyclonal anti-MAPK 1/2 (Erk 1/2-CT) antibody was from Upstate Biotechnology. The mouse monoclonal anti-c-Raf-1 antibody was from Transduction Laboratories (Lexington, KY). The rabbit polyclonal anti-MKP-1 (V-15; MAPK phosphatase), anti-Fra-2 (L-15), and anti-K-Ras2A antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). This latter antibody recognizes only the Ki-RasA isoform of Ras proteins. The mouse monoclonal anti-v-Ha-Ras antibody was from Oncogene Science (Cambridge, MA). This antibody recognizes all isoforms of Ras protein, including Ha-Ras, Ki-RasA, Ki-RasB, and N-Ras. All phospho-specific antibodies were from Cell Signaling Technologies (Beverly, MA). All secondary horseradish peroxidase-conjugated antibodies were from Kirkegaard and Perry Laboratories (Gaithersburg, MD). Statistics | All values reported as mean +- SEM. Statistical significance (p <= 0.05) was determined using the t test for differences in mean values, and a one-way analysis of variance in conjunction with the Student-Newman-Keuls test for multiple comparisons. RESULTS : Aldosterone Increases Absolute and Active GTP-bound Levels of Ki-RasA | Aldosterone via transcriptional control increases Ki-rasA mRNA and Ki-RasA protein levels in renal A6 epithelial cells. Experiments in Figure tested the hypothesis that aldosterone in these cells also increases the amount of active Ki-RasA. Activated Ras bound by GTP interacts with the RBD of Raf . The representative Western blots of Figure A show that the addition of aldosterone for 3 h to A6 cell monolayers markedly increased Ki-RasA (middle) and Ki-RasA:GTP levels (bottom), but had little effect on total Ras levels (top). It is known that Ki-RasA is expressed at levels much lower than other Ras isoforms . Thus, the lack of a marked change in total Ras was not unexpected. For these experiments, Ras:GTP was isolated using GST-RBD agarose from whole cell lysates from cells treated with vehicle (CON; 0.1% DMSO) and aldosterone (ALDO; 1.5 muM) for 3 h. After isolating total cellular Ras:GTP, Ki-RasA:GTP was identified with anti-K-Ras2A antibody, which reacts only with Ki-RasA . Total Ras protein was identified with the anti-v-Ha-Ras antibody, which is reactive with Ki-RasA and B, Ha-Ras, and N-Ras isoforms. The summary graph in Figure B shows the relative change in response to aldosterone for the levels of Ras (1.6 +- 0.2, n = 11), Ki-RasA (3.1 +- 0.4, n = 8), and Ki-RasA:GTP (2.8 +- 0.4, n = 8). Aldosterone, compared with vehicle, significantly increased Ki-RasA, and Ki-RasA:GTP levels (p < 0.005 for both). Although aldosterone also significantly increased total Ras levels (p = 0.03) at a similar time point, as reported previously by our laboratory , the relative increase in total Ras was significantly less than that of Ki-RasA (p = 0.002) and Ki-RasA:GTP (p = 0.009). Importantly, the relative changes in Ki-RasA vs. Ki-RasA:GTP levels were not different (p = 0.60). Figure 1 | Aldosterone increases active, GTP-bound Ki-RasA levels. Aldosterone increases active, GTP-bound Ki-RasA levels. (A) Western blot analysis of total Ras (top), Ki-RasA (middle), and Ki-RasA: GTP (bottom) in A6 cells treated without (CON) and with ALDO for 3 h. The representative blot probed for Ki-RasA: GTP (bottom) contained the precipitant from whole cell lysate by using Raf-RBD agarose, which specifically binds GTP-bound Ras. Blot probed with anti-Ki-RasA antibody to detect Ki-RasA: GTP. (B) Summary graph showing the relative changes in the levels of Ras, Ki-RasA and Ki-RasA: GTP in response to aldosterone treatment for 3 --4 h. * indicates significantly greater relative increases for Ki-RasA and Ki-RasA: GTP vs. Ras. Aldosterone Increases MAPK Activity | One potential signaling pathway activated by Ki-RasA is the MAPK cascade. Aldosterone stimulation of MAPK activity in A6 cells was measured in an in vitro reaction by following phosphorylation of exogenous MBP. The typical Western blots in Figure , A and B, showing the time course of aldosterone stimulation of MAPK activity, were probed with anti-phospho-MBP antibody. For these blots, each lane contained 3.5 mug of MBP processed for 30 min at 30C by A6 whole cell lysate (equal concentrations of total cellular protein; supplemented with MBP, Mg2+/ATP, and protein kinase C, protein kinase A, and calcium-calmodulin-dependent protein kinase II, and phosphatase inhibitors) from cells treated with aldosterone for the indicated times (in minutes for Figure A and hours for Figure B). The bar graph in Figure C summarizes such experiments. Aldosterone significantly increased MAPK activity 6.6 +- 1.2-, 10.5 +- 1.8-, 11.5 +- 2.5-, 12.0 +- 4.2-, 7.1 +- 2.8-, and 3.1 +- 0.9-fold at the 15 and 30 min, and 1-, 2-, 4-, and 6-h time points (n >= 4). At 1 and 5 min, MAPK activity was 1.4 +- 0.2- and 3.3 +- 0.8-fold higher (n = 2), respectively. MAPK activity in response to aldosterone increased steadily peaking between 0.5 and 2 h and waning after 4 h. Figure 2 | Aldosterone increases MAPK activity. Aldosterone increases MAPK activity. (A and B) MAPK activity was assessed by measuring phosphorylation of exogenous MBP. Exogenous MBP was added to equal amounts of A6 cell lysate from cells treated with aldosterone for the indicated times (in minutes, A; and hours, B). These typical Western blots were probed with anti-phospho-MBP antibody. (C) Summary graph of such experiments. *p <= 0.05 vs. time 0. Time Course of Aldosterone Activation of MAPK Is Distinct from That of Classic Mitogens | Results in Figure suggested that compared with classic mitogens, aldosterone activates MAPK signaling in a distinct manner, taking longer (up to 1 --2 h) to reach maximal activity and having a persistent signal (up to 4 h). Such a time course would be consistent with aldosterone-dependent activation of the MAPK requiring a latent period necessary for transcription and translation of Ki-RasA. Experiments in Figure were performed to determine whether the time course of aldosterone-dependent activation (phosphorylation) of MAPK in A6 cells was indeed distinct from activation by the traditional mitogen, serum. For these experiments, determining phospho-MAPK levels assessed activation of MAPK. The Western blots in Figure A were probed with anti-phospho-MAPK antibody. These blots contained equal amounts of whole cell lysate from A6 cells treated with serum (10% FBS; top) and aldosterone (1.5 muM) in the absence (middle) and presence (bottom) of the corticosteroid receptor antagonist RU486 (mifepristone; 0.1 muM) for the indicated times in minutes. Western blots in Figure A were stripped and subsequently reprobed with anti-MAPK antibody (Figure B, same order). Serum and aldosterone clearly have temporally distinct effects on activation of MAPK with the effects of serum peaking in the first 15 --30 min and waning thereafter, whereas those of aldosterone rising from 30 min onwards. The actions of aldosterone on MAPK were completely reversed by RU486, suggesting that this steroid stimulates MAPK signaling via its genomic actions. As shown in Figure C, similar results were observed when MAPK activity was assessed in an in vitro reaction by following phosphorylation of exogenous MBP. This blot was probed with anti-phospho-MBP antibody and each lane contained 3.5 mug of MBP processed by A6 whole cell lysate (equal concentrations of total cellular protein) from cells treated with serum (10% FBS) for the indicated times (in minutes). Figure 3 | Aldosterone compared with serum activates MAPK signaling in a delayed manner. Aldosterone compared with serum activates MAPK signaling in a delayed manner. (A) Western blots of A6 whole cell lysate from cells treated with serum (10% FBS, top), aldosterone (1.5 muM), and aldosterone plus RU486 (0.1 muM) for the indicated time in minutes. All blots probed with anti-phospho-MAPK antibody. (B) Blots in A stripped and reprobed with anti-MAPK antibody. (C) Western blot of MBP processed by A6 whole cell lysates from cells treated with serum (10% FBS) for the indicated times in minutes. Blot probed with anti-phospho-MBP antibody. Persistent Stimulation of MAPK by Aldosterone Is Mediated by Nuclear Steroid Receptors | Experiments in Figure were performed to further characterize the relation of aldosterone-stimulated MAPK at the 2-h time point with the genomic effects of this steroid. Both the effects of aldosterone on activation (phosphorylation) of MAPK in the presence of inhibitors of transcription and translation (Figure A), and the effects of aldosterone on MAPK activity in the presence of inhibitors of nuclear corticosteroid receptors, translation and mitogen-activated protein kinase kinase (MEK) (Figures , B and C) were determined. Figure 4 | Aldosterone stimulates MAPK signaling in A6 cells through induction of gene expression mediated by the glucocorticoid receptor. Aldosterone stimulates MAPK signaling in A6 cells through induction of gene expression mediated by the glucocorticoid receptor. (A) Western blot containing A6 whole cell lysate from cells treated with aldosterone (1.5 muM), CON, actinomycin D (1 mug/ml), cycloheximide (1 mug/ml), and aldosterone plus actinomycin D or cycloheximide for 2 h. This blot was probed with anti-phospho-MAPK antibody (top), stripped, and then reprobed with anti-MAPK antibody (bottom). (B) This typical Western blot shows phosphorylation of exogenous MBP added to A6 cell lysate from cells treated with aldosterone, and steroid plus RU486 (0.1 muM), emetine (1 mug/ml), cycloheximide (1 mug/ml), PD-98059 (10 muM), and U-0126 (0.5 muM) for 2 h. This typical blot was probed with anti-phospho-MBP antibody. (C) Below is a summary graph of five such experiments showing relative MAPK activity. The typical Western blot of Figure A was probed with anti-phospho-MAPK antibody (top), stripped, and then reprobed with anti-MAPK antibody (bottom). This blot contained equal amounts of A6 whole cell lysate from cells treated with ALDO (1.5 muM), CON (vehicle), actinomycin D (1.0 mug/ml), cycloheximide (1.0 mug/ml), and aldosterone plus actinomycin D or cycloheximide for 2 h. Although actinomycin D and cycloheximide had no overt effect on activation of MAPK when added alone, when added simultaneously with aldosterone, they abolished steroid-dependent activation of MAPK, demonstrating that transcription and translation are necessary for aldosterone to activate MAPK signaling. The typical Western blot of Figure B shows phosphorylation of exogenous MBP added to A6 cell lysate from cells treated with aldosterone for 2 h in the absence and presence of the corticosteroid receptor inhibitor RU486 (mifepristone; 0.1 muM) and inhibitors of translation (cycloheximide and emetine 1.0 mug/ml) and MEK (PD-98059 and U-0126 at 10 and 0.5 muM, respectively). For these experiments, inhibitors were added simultaneously with aldosterone. This Western blot was probed with anti-phospho-MBP antibody and contained equal amounts of exogenous MBP processed by the respective A6 cell lysate (equal concentration of total protein). As shown in the summary graph of Figure C, RU486 significantly decreased relative aldosterone-stimulated MAPK activity to 0.2 +- 0.04 (n = 5). Similarly, relative aldosterone-induced MAPK activity was decreased to 0.2 +- 0.1 (n = 5) and 0.3 +- 0.2 (n = 4) by emetine and cycloheximide, respectively. Moreover, PD-98059 and U-0126 decreased relative aldosterone-sensitive MAPK activity to 0.2 +- 0.1 (n = 5) and 0.2 +- 0.1 (n = 5), respectively. At this time point (2 h), the negative control for U-0126, U-0124, did not affect MAPK signaling and Ki-RasA levels were 53, 61, 62, 96, and 108% of (aldosterone-treated) control in the RU486, emetine, cycloheximide, PD-98059, and U-0126 groups, respectively (our unpublished data). Ki-ras Gene Contains a Functional Steroid Response Element That Confers Aldosterone Responsiveness to a Heterologous Promotor | The results described above and those reported previously by us and others suggest that aldosterone via steroid receptors directly affects Ki-ras expression and that this then impinges upon MAPK signaling. However, the molecular basis of this regulation has not been studied. Glucocorticoids and aldosterone ultimately target similar cis-acting elements through either the glucocorticoid or mineralocorticoid receptor (reviewed in ). The human and rat Ki-ras genes contain partially characterized cis-acting elements within the intron 1-exon 1 region that are trans-activated by the glucocorticoid --steroid receptor complex . Conserved elements have similarly been identified in the Ha-ras gene as responsive to glucocorticoids . Using this paradigm, we prepared a reporter plasmid containing the mouse c-Ki-ras2 intron 1-exon 1 region (-165 --153 from adenosine of the translation start codon within exon 1; pGL2-TK-rasI/E1), which contains several putative steroid response elements, and as shown in Figure , tested whether this region conferred functional aldosterone responsiveness to a heterologous promotor in A6 epithelial cells. Luciferase activity in cells transfected with pGL2-TK-rasI/E1 in the presence of vehicle and aldosterone (1.5 muM, 4 h) was 1.5 +- 0.3 and 4.6 +- 0.9 (n = 6), respectively. Thus, aldosterone significantly increased (threefold) luciferase expression driven by the pGL2-TK-rasI/E1 chimeric reporter plasmid. Simultaneous addition of RU486 (0.1 muM, n = 3) with aldosterone markedly decreased luciferase activity 50% to 2.3 +- 0.9 in cells transfected with pGL2-TK-rasI/E1. Dexamethasone (DEX) (0.1 muM, 4 h) had a similar effect as aldosterone increasing luciferase activity vs. vehicle 2.5-fold to 3.7 +- 0.9 (n = 3; p = 0.06; our unpublished data). Luciferase activity in the presence of either steroid in cells transfected with pGL2-TK-rasI/E1 was significantly greater (p > 0.05) than that in cells transfected with pGL2-TK and treated similarly (ALDO = 1.0 +- 0.2, n = 6; DEX = 0.7 +- 0.2, n = 3). In contrast, luciferase activity in the presence of vehicle was not different (p = 0.3) between pGL2-TK (0.8 +- 0.3, n = 4) and pGL2-TK-rasI/E1 --transfected cells. Reporter gene activity in cells transfected with the minimal promoter thymidine kinase luciferase plasmid (pGL2-TK) or pRL-CMV control plasmid alone was unaffected by steroid treatment (our unpublished data). Figure 5 | rasI/E1 region of c-Ki-ras2 is sufficient for steroid responsiveness. rasI/E1 region of c-Ki-ras2 is sufficient for steroid responsiveness. A6 cells were transiently transfected with pRL-CMV plus either control reporter plasmid (pGL2-TK), which contained firefly luciferase expression driven by the minimal thymidine kinase promotor, or reporter plasmid that contained the rasI/E1 region (pGL2-TK-rasI/E1). Luciferase activity in cells treated with vehicle, aldosterone, and aldosterone plus RU486 was quantified 48 h after transfection. *p < 0.05 vs. pGL2-TK treated with aldosterone and vehicle, and pGL2-TK-rasI/E1 treated with vehicle. Activation of MAPK Cascade by Aldosterone Is Dependent on Ki-RasA Expression | Although all of the above-reported results support the idea that aldosterone activates Ki-RasA via transcriptional control mediated by nuclear steroid receptors, it is unclear whether the aldosterone-sensitive MAPK signaling reported in Figures -- is in fact dependent on induction of Ki-RasA. Experiments in Figure directly test the link between aldosterone-dependent induction of Ki-RasA and steroid-dependent activation of MAPK signaling. The typical Western blots in Figure A are of lysates extracted from cells treated with aldosterone (1.5 mum; control) and aldosterone after pretreatment (24 h, 10 muM) with sense and antisense Ki-rasA oligonucleotides. Use of these oligonucleotides in A6 cells has been described previously . The top blot in Figure A was probed with anti-K-Ras2A antibody and demonstrates the efficacy of the antisense oligonucleotide to decrease Ki-RasA levels. Ki-RasA levels in the antisense group were similar40% of those in the control and sense groups (n = 6). In response to Ras signaling, Raf and MAPK become phosphorylated. The top middle blot in Figure A was probed with anti-phospho-Raf antibody and demonstrates that the Ki-ras antisense oligonucleotide attenuates aldosterone-sensitive phosphorylation of Raf. Phospho-Raf levels in the antisense group were <20% of those in the control and sense groups (n = 3). Similarly, as shown by the bottom middle blot, antisense inhibited aldosterone-dependent activation of MAPK with phospho-MAPK levels in the antisense group being <25% of those in the control and sense groups (n = 3). In contrast, as shown by the bottom blot, antisense had no effect on total MAPK levels with all three groups having similar levels of MAPK. Figure 6 | Aldosterone activates the MAPK cascade through induction of Ki-RasA. Aldosterone activates the MAPK cascade through induction of Ki-RasA. (A) These typical Western blots show the effects of aldosterone on Ki-RasA (top), phospho-Raf (top middle), active (phospho)-MAPK (bottom middle), and absolute MAPK (bottom) levels in A6 cells treated with steroid alone (vehicle) and in the presence of sense (SENSE) and antisense (ANTI) Ki-ras oligonucleotide. (B) Typical Western blot probed with anti-Ras antibody containing equal amounts of whole cell lysate extracted from cells stably expressing inducible DNRasN17 treated with vehicle (CON) and aldosterone for 2 h. (C) Typical Western blot probed with anti-phospho-MAPK (top) and then stripped and reprobed with anti-MAPK (bottom) antibodies containing equal amounts of whole cell lysate extracted from control cells and cells stably expressing inducible DNRasN17 treated with vehicle (-) and aldosterone (+) for 2 h. The representative Western blot in Figure B shows the effects of aldosterone on two distinct clonal A6 cell lines (DNRas1 and DNRas2) stably expressing corticosteroid-inducible dominant-negative RasN17. For such experiments (n = 2), confluent cells were treated with vehicle (CON) or aldosterone (ALDO) for 2 h. This blot, which contains equal amounts of total protein for each lysate, was probed with anti-Ras antibody and demonstrates that aldosterone increases total Ras levels in these clonal lines. Because aldosterone has little effect on total Ras expression in untransfected A6 cells (Figure A), these results demonstrate that these clonal lines stably express corticosteroid-inducible DNRasN17. The effects of aldosterone on activation of MAPK in these two clonal lines were determined next (Figure C). This typical Western blot (n = 2), which contained lysates with equal amounts of total protein from control cells and cells stably expressing inducible DNRasN17 treated with vehicle (-) and aldosterone (+) for 2 h, was probed with anti-phospho-MAPK antibody (top). This blot was subsequently stripped and reprobed with anti-MAPK antibody (bottom). In A6 cells stably expressing inducible DNRasN17, aldosterone had less of an effect on activation of MAPK compared with untransfected cells. Interestingly, stable cells had elevated levels of phospho-MAPK in the absence of aldosterone with steroid actually decreasing these levels. This may reflect a feedback response to chronically depressed MAPK signaling due to DNRasN17 leak. Nonetheless, these results show that for aldosterone to stimulate MAPK signaling, Raf kinase must be available to Ki-RasA, which it is not in the presence of DNRasN17. These results also are consistent with those in Figure A and together suggest that Ki-RasA transduces the aldosterone signal onto the MAPK cascade. Peak Activation of MAP Kinase Cascade by Aldosterone-stimulated Ki-RasA Is Delayed and Prolonged | The experiments in Figure temporally map the actions of aldosterone on MAPK signaling at several discrete levels within the transduction cascade. The representative Western blot of Figure A shows that whereas aldosterone increases the amount of active (phosphorylated) MAPK, it does not affect the total cellular pool of MAPK. For these blots, each lane contained 50 mug of total cellular protein, and the aldosterone (1.5 muM) treatment time is indicated in hours. For this set of experiments, cells representing time 0 were washed with vehicle 2 h before extraction. No difference was observed between time zero and wash for these and all other experiments. The top blot in this experiment was stripped after being probed with anti-phospho-MAPK (ERK 1/2) antibody and subsequently reprobed with anti-MAPK 1/2 antibody (bottom blot). Figure 7 | Aldosterone activates the MAPK cascade in a delayed but prolonged manner. Aldosterone activates the MAPK cascade in a delayed but prolonged manner. (A) Western blot analysis of aldosterone effects on MAPK and active (phosphorylated) MAPK levels in A6 cells. Top blot probed with anti-phospho-MAP kinase antibody. This blot was subsequently stripped and reprobed with anti-MAPK antibody (bottom). (B) Western blot analysis testing the effects of aldosterone on c-Raf and phosphorylated c-Raf. Top and bottom blots probed with anti-phospho-Raf and anti-Raf antibody, respectively. (C) Western blot analysis testing the effects of aldosterone on the MAPK cascade effectors, RSK (top), Fra-2 (middle), and MKP-1 (bottom). Top blot probed with anti-phoshpo-p90RSK antibody; middle blot probed with anti-Fra-2 antibody; and bottom blot probed with anti-MKP-1 antibody. Each lane for blots in A --C contained equal amounts of whole A6 cell lysate extracted from cells treated with aldosterone for the indicated time in hours. (D and E) Summary graphs showing the time course of the relative (vs. time 0) actions of aldosterone on Ki-RasA, phospho-MAPK, MAPK, phospho-Raf, Raf, MKP-1, phospho-RSK, Fra-2, and Na+ transport. The blots in Figure B show that similar to MAPK, aldosterone increases the levels of phospho-Raf compared with total cellular Raf. These blots contain lysate from cells treated with aldosterone for the indicate times (in hours). Top and bottom blots were probed with anti-phospho-(Ser259)-Raf and anti-Raf antibody, respectively. Each lane contained similar50 mug of total protein. The actions of aldosterone on downstream effectors of the MAPK cascade are shown in Figure C. These Western blots are of the same lysates (similar to that for Raf and phospho-Raf in 7B from the indicated time points after aldosterone treatment. Each lane contained 50 mug of total protein. Top, middle, and bottom blots were probed with anti-phospho-RSK-1 (p90 ribosomal S6 kinase; also referred to as MAPKAP kinase-1), Fra-2, and MKP-1 antibody, respectively. Clearly, all three of these MAPK effector proteins are either activated/phosphorylated, as for RSK-1, or induced, for Fra-2 and MKP-1, in response to aldosterone. The summary plots of Figure D show relative changes in Ki-RasA (inverted triangles), phospho-MAPK (diamonds), and MAPK (squares) levels in response to aldosterone at 0-, 0.5-, 1-, 2-, 4-, and 6-h time points. Also shown in this graph are the effects of aldosterone on Na+ transport (current, circles) across A6 cell monolayers at each time point. Figure E shows diary plots of relative changes in phospho-Raf (circles) and total Raf (gray triangles) in response to aldosterone. Also shown in this graph are the temporal actions of aldosterone on expression of the MAPK effectors Fra-2 (inverted triangles) and MKP-1 (diamonds). Included, in addition, are the effects of aldosterone on phosphorylation of the MAPK effector RSK-1 (squares). An expanded time course for aldosterone actions of affecters and effectors of MAPK is included in Table . Table 1 | Temporal effects of aldosterone on MAPK signaling and current DISCUSSION : The present results support a novel mechanism whereby aldosterone induces Ki-RasA expression at the level of transcription and then through stoichiometric increases in the levels of active, GTP-complexed Ki-RasA stimulates the MAPK cascade. Figure compares this genomic mechanism with the classic mechanism initiated by traditional mitogens. Traditional mitogens, such as serum, in contrast, stimulate MAPK signaling via a mechanism involving posttranslational control of active, GTP-complexed Ras levels without effects on absolute Ras levels. Compared with the classic mechanism, the distinct but possibly complimentary genomic mechanism activated by aldosterone leads to delayed but prolonged MAPK signaling. It is speculated that the delayed but prolonged MAPK signaling in response steroids will differentially affect cellular activity compared with traditional mitogens. Figure 8 | Schematic diagram showing distinct but complementary mechanisms of activation of the MAPK cascade in response to classic mitogens and corticosteroids. Schematic diagram showing distinct but complementary mechanisms of activation of the MAPK cascade in response to classic mitogens and corticosteroids. Classic mitogens using a posttranslational mechanism affect the GTP-binding state of Ras compared with steroids, which through genomic actions lead to stoichiometric increases in total and active Ki-RasA levels without changes in GTP-binding kinetics. Once stimulated by either mechanism, Ki-RasA then activates effector kinases through a common pathway with the distinction being that the genomic response compared with the posttranslational mechanism has a delayed onset and prolonged signal. Aldosterone to Ki-RasA | Aldosterone increases both Ki-RasA mRNA and protein levels in amphibian renal epithelial cells . The current results in Figure are consistent with this finding. The analogous findings in mammals, however, have been more controversial (reviewed in ; ) with corticosteroids increasing Ki-ras and Ki-RasA expression in mammalian colonic and mammary epithelial cells (; ; ; ; Fuller, personal communication) and cardiac fibroblasts , but not kidney epithelia (; Verrey, personal communication). The underlying molecular and cellular basis for these apparent discrepancies are currently unclear, but may be related to similarly undetermined mechanisms resulting in tissue-selective aldosterone induction of other proteins, such as ENaC and serum and glucocorticoid-inducible kinase (reviewed in ). Nevertheless, in A6 cells and in the heterologous Xenopus laevis oocyte expression system, Ki-RasA activates ENaC, which is one final effector of aldosterone signaling in epithelial cells . The molecular basis whereby aldosterone induces Ki-RasA has not been described. Similarly, it also is unclear whether merely increasing absolute Ki-RasA levels in response to aldosterone is sufficient to activate Ki-RasA leading to dependent stimulation of its effector MAPK cascades. Results in Figure demonstrate that aldosterone increases active, GTP-bound Ki-RasA levels proportionately with absolute Ki-RasA levels. This is consistent with a mechanism where through mass action, aldosterone-induced Ki-RasA leads to concomitant increases in the active pool of Ki-RasA. Results in Figures , , , and are consistent with this aldosterone-increased active pool of Ki-RasA then subsequently stimulating effector MAPK signaling. The results in Figure identify a putative cis-acting element/region (-165 --153 for mouse c-Ki-ras2) within the Ki-ras gene that possibly bestows aldosterone-responsiveness at the level of transcription to the Ras -> MAPK signaling cascade. Sequence analysis of this region of mouse c-Ki-ras2 reveals the presence of several potential sites responsive to corticosteroids with one hexanucleotide (TGTTCT; -50 to -45) half-site identical to those modulating corticosteroid-responsiveness in other genes . This hexanucleotide half-site is the most highly conserved portion of the palindromic GRE (5'-GGTACAnnnTGTTCT-3') and has been shown to bind GR and trans-activate in response to activated receptor . It is provocative that this half-site contained in the noncoding region is absolutely conserved in sequence identity and relative position in the mouse, rat, and human Ki-ras genes (see accession numbers , , and , respectively). Sequence data for the corresponding region in the X. laevis Ki-ras gene has not been published (see accession number ); therefore, it currently is unclear whether Ki-ras in this species also contains a similar element. The Ha-ras gene also contains conserved regions similar to those in the -165 --153 region of mouse Ki-ras. In addition, Ha-ras is induced by corticosteroids in some tissues, but pointedly not induced in renal and colonic epithelia (; ; Fuller, personal communication). It currently is unclear how common aldosterone effects on Ha-Ras expression are and what are the underlying molecular bases, if any, allowing for discretionary induction of Ki-RasA vs. Ha-Ras in response to aldosterone and other corticosteroids in a tissue- and species-specific manner. The effects of Ha-Ras on aldosterone effectors, moreover, have not been investigated. Although we observe a small but significant increase in total Ras levels in response to aldosterone (Figure ; ), it is not clear what fraction of this increase results from induction of Ki-RasA vs. Ha-Ras. However, the effects of aldosterone on MAPK signaling in A6 cells reported in the current study are abolished by inhibiting Ki-RasA expression with an antisense oligonucleotide, suggesting that this species of Ras and not Ha-Ras is the primary mediator of aldosterone actions in these cells. These findings are consistent with previous findings showing that Ki-RasA is necessary and sufficient to reconstitute, in part, aldosterone actions on ENaC . Ki-RasA to MAPK | The current study is the first to directly link the effects of aldosterone on Ki-RasA expression with activation of the MAPK cascade. Figure shows, using an in vitro assay, that aldosterone increases MAPK activity in A6 cells. Activity peaked between 1 and 2 h and began to wane by 4 h. Similar results were observed when activation of MAPK signaling was assessed using an in vivo assay of MAPK phosphorylation (Figures and ). In contrast to the time course of aldosterone effects on MAPK signaling, the classic mitogen, serum, stimulated this cascade much quicker, reaching peak activation within the first 30 min and waning thereafter . An alternative to aldosterone affecting the MAPK cascade through genomic actions dependent on induction of Ki-RasA expression is that this steroid activates MAPK signaling independent of its nuclear effects. Indeed, report that aldosterone modulates Na+/H+ exchange in Madin-Darby canine kidney cells through MAPK signaling, and that due to the rapidity of this action it is likely independent of steroid effects on gene expression. Similarly, report that aldosterone, independent of modulating gene expression, increases phospho-MAPK levels within 3 to 5 min with levels waning soon thereafter. We believe that the current results are more consistent with aldosterone activating the MAPK cascade in A6 cells via a genomic mechanism for several reasons. This is particularly true when the maximal effects of steroid are considered. First, as mentioned above, absolute and active Ki-RasA levels increased by the same amount in response to aldosterone , and aldosterone's effects on MAPK signaling at peak activation are entirely dependent on Ki-RasA expression and activation of its downstream effectors, such as Raf and MEK . Second, activation (phosphorylation) of MAPK in response to aldosterone was not apparent until 30 min after treatment . Third, at all time points assayed the stimulatory effects of aldosterone on phosphorylation of MAPK were attenuated by treatment with the nuclear corticosteroid receptor antagonist RU486 . Similarly, the maximum stimulatory effects of aldosterone (at 2 h) on phosphorylation of MAPK and MAPK activity (Figures and ) were sensitive to blockade of nuclear corticosteroid receptor, transcription, and translation. These results strongly support the contention that for aldosterone to affect MAPK signaling in A6 cells, its genomic actions are absolutely required. However, it was not our intention to investigate nongenomic regulation of MAPK signaling by aldosterone and thus, our experimental design cannot definitively exclude this possibility for the earlier time points (<30 min; see below). It is possible that in the current study there was a nongenomic response superimposed on a slower developing genomic response. Further support for a genomic mechanism driving the activation of MAPK in response to aldosterone observed in the current study is provided by comparison of the aldosterone-dependent time course of MAPK activity and phosphorylation vs. that of serum (Figures and ). Presumably, as shown previously for aldosterone , the nongenomic effects of this steroid would have a rapid time course more similar to traditional mitogens. This is not the case in the current study. We find that aldosterone activates the MAPK cascade via a similar genomic mechanism dependent on induction of Ki-RasA in rat cardiac fibroblasts . It is not clear why some cells respond to aldosterone via nongenomic activation of the MAPK cascade and others with a genomic mechanism. It is likely that these complimentary mechanisms are manifested in a cell-specific manner. Importantly, the systemic and cellular effects of aldosterone on classic target tissues, such as the distal nephron and colon, are mediated primarily through the genomic actions of this steroid (reviewed in ). Further comparison of the maximal effect of aldosterone on MAPK activity and activation of MAPK (Figures and and Table ) shows that MAPK activity and phosphorylation increase similar12 and similar2-fold, respectively. This difference may reflect signal amplification where a unit increase in MAPK phosphorylation yields a higher increase in activity. Comparison also of the effects of aldosterone on MAPK activity and phosphorylation at the 5-min time point shows that although MAPK activity already has increased by 5 min, there is no apparent increase in phosphorylation of MAPK. At first glance, the rapid effects of aldosterone on MAPK activity seem to be consistent with an early nongenomic response superimposed on a genomic response. We do not believe this to be the case, as argued above. Moreover, it is not commonly accepted that A6 cells have a nongenomic response to aldosterone. It is likely that, although performed in the presence of several different kinase inhibitors, the in vitro assay used to quantify aldosterone actions on MAPK activity was somewhat biased by the presence of uninhibited kinases or other cellular factors that either directly impinge upon phosphorylation of MBP or MAPK activity when taken out of cellular context. However, it also is possible that the assay of MAPK activity was more sensitive than that assessing activation of MAPK. The results showing that MAPK activity is increased 12-fold, whereas MAPK phosphorylation increased only 2-fold by 2 h would also be consistent with this possibility. If this is the case then nongenomic actions of aldosterone on MAPK signaling in A6 cells during the first 30 min or so cannot be wholly excluded by the current results. However, if a nongenomic response was superimposed on a slower developing genomic response we would have expected aldosterone-stimulated MAPK activity to peak at two distinct time points, which was clearly not the case. Nevertheless, the current results definitively demonstrate that peak activation of MAPK signaling at the 2-h time point in response to aldosterone is absolutely dependent on a genomic event and induction of Ki-RasA. Time Course and Effects of Aldosterone-stimulated MAPK Signaling | The current results show for the first time the temporal effects of aldosterone on the active levels of several different affecters and effectors of MAPK (Figure and Table ). Similar to MAPK, phosphorylation of Raf in response to aldosterone was delayed and prolonged. Phosphorylation of Raf was dominant between 2 and 6 h and thus Raf was phosphorylated (on Ser259) at a time point later than phosphorylation of MAPK. This finding is consistent with Raf being phosphorylated in response to aldosterone in a negative feedback manner. In fact, the Western blots in Figure actually assessed negative regulation of Raf that had previously been active. The serine/threonine kinase Akt/PKB an effector of phosphatidylinositol 3-kinase (PI3-K) is responsible for phosphorylation of active Raf at Ser259, which then leads to inactivation . PI3-K is also a first effector of Ki-Ras and activated by aldosterone . Thus, this feedback regulation of Raf may reflect parallel activation by aldosterone-induced Ki-RasA of both the MAPK and PI3-K signaling cascades. Consistent with these findings are those showing that the MEK inhibitors PD-98059 and U-0126 attenuate aldosterone-sensitive phosphorylation of MAPK but not Raf; and that the PI3-K inhibitor LY 294002 decreases aldosterone-sensitive Raf phosphorylation but not phosphorylation of MAPK (our unpublished observations). In addition to MAPK, MEK, and Raf, aldosterone increased the levels of phospho-RSK1 beginning at 30 min, reaching a peak by 4 h. RSK-1 is well known to be a target regulated at the posttranslational level in response to MAPK signaling. Interestingly, aldosterone induced expression of Fra-2 and MKP-1, which are known to be regulated at the level of transcription in response to MAPK signaling. Fra-2 is a transcription factor related to Fos. Thus, this action may enable aldosterone to secondarily impinge upon subsequent rounds of transcription. An alternative mechanism is that aldosterone directly affects transcription/translation of these proteins independent of MAPK signaling. Indeed, recent results from support such a mechanism of aldosterone action on Fra-2. To date, a direct effect of aldosterone on MKP-1 transcription/translation has not been demonstrated. Interestingly, induction of MKP-1, which is a phosphatase that dephosphorylates MAPK in a negative feedback manner to dampen MAPK signaling, in conjunction with feedback phosphorylation of Raf on Ser259 by Akt may result in the deactivation of MAPK signaling after 4 h observed in the current study. All of the current results are consistent with the idea that activation of the MAPK cascade in response to aldosterone is mediated by induction of Ki-RasA and activation of signaling constituents, such as Raf and MEK, that lie between this Small, monomeric G protein and MAPK. Consequently, after increases in total Ki-RasA and active Ki-RasA:GTP levels, aldosterone-dependent MAPK signaling follows a normal progression with the major exception being that signaling is prolonged. The explanation for prolonged MAPK signaling in response to aldosterone clearly then must come from the mechanism of initiation: transcriptional control of Ki-RasA. Such a mechanism is distinct from that used by classic mitogens, and leads to a delayed but prolonged signaling event. The delay results from the latent period required for increased transcription/translation of Ki-RasA. Prolongation results from stoichiometric increases in total and active Ki-RasA with absolute but not relative levels of Ki-RasA:GTP increasing. MAPK activity in response to aldosterone would then primarily be dependent on Ki-RasA protein turnover and feedback regulation. It is predicted that prolonged steroid-dependent MAPK signaling produces unique changes in cellular activity compared with a classic response. Although it is known that activation of Ki-RasA by aldosterone is necessary and sufficient for induced Na+ transport and ENaC activity in A6 cells and for ENaC activation when this G protein is overexpressed along with the channel in X. laevis oocytes , the actions of MAPK signaling on transport and ENaC seem to be inhibitory . Thus, it can be speculated that the prolonged activation of the MAPK cascade described in the current study is either a component of a feedback system or impacts Na+-transporting epithelia independently of directly affecting Na+ transport. The possible physiological and pathophysiological roles for genomic activation of MAPK signaling by corticosteroids at the tissue and systemic levels remain to be elucidated. Backmatter: PMID- 12221115 TI - Functional Heterogeneity of Bone Morphogenetic Protein Receptor-II Mutants Found in Patients with Primary Pulmonary Hypertension AB - Germline mutations in the BMPR2 gene encoding bone morphogenetic protein (BMP) type II receptor (BMPR-II) have been reported in patients with primary pulmonary hypertension (PPH), but the contribution of various types of mutations found in PPH to the pathogenesis of clinical phenotypes has not been elucidated. To determine the biological activities of these mutants, we performed functional assays testing their abilities to transduce BMP signals. We found that the reported missense mutations within the extracellular and kinase domains of BMPR-II abrogated their signal-transducing abilities. BMPR-II proteins containing mutations at the conserved cysteine residues in the extracellular and kinase domains were detected in the cytoplasm, suggesting that the loss of signaling ability of certain BMPR-II mutants is due at least in part to their altered subcellular localization. In contrast, BMPR-II mutants with truncation of the cytoplasmic tail retained the ability to transduce BMP signals. The differences in biological activities among the BMPR-II mutants observed thus suggest that additional genetic and/or environmental factors may play critical roles in the pathogenesis of PPH. Keywords: INTRODUCTION : Vascular development and homeostasis are regulated by a number of cytokines, including the members of the transforming growth factor-beta (TGF-beta) superfamily. The TGF-beta superfamily includes various proteins with similar dimeric structures, e.g., activins, nodal, bone morphogenetic proteins (BMPs), and growth/differentiation factors . BMPs were originally identified as osteoinductive cytokines at extraskeletal sites in vivo . Subsequently, BMPs have been shown to exhibit multifunctional activities in various types of cells. They regulate cell growth, apoptosis, and differentiation, and participate in patterning and specification of various tissues and organs . BMPs transduce their signals via two types of serine/threonine kinase receptors, type I and type II receptors, both of which are required for their signal transduction . BMPs bind to three different type II receptors, i.e., activin type II receptors (ActR-IIA and ActR-IIB) and BMPR-II , and three different type I receptors, i.e., activin receptor-like kinase (ALK)-3/BMPR-IA, ALK-6/BMPR-IB, and ALK-2 (,; ; ; ; ). On binding of BMPs, type II receptors phosphorylate type I receptors, which in turn phosphorylate intracellular signal-transducing molecules Smad1, 5, and 8 . ALK-3 and ALK-6 activate these three Smads, whereas ALK-2 activates only Smad1 and Smad5 but not Smad8 . Recently, heterozygous germline mutations of the BMPR2 gene encoding BMPR-II were found in patients with primary pulmonary hypertension (PPH) , suggesting that BMPs may play important roles in homeostasis of the pulmonary vascular system. PPH is a disorder of the pulmonary arteries characterized by formation of plexiform lesions and obliteration of small pulmonary arteries . Subsequently, sporadic form of PPH was also shown to be associated with germline mutations of BMPR2 in at least 26% of cases . Although BMP signals are involved in the regulation of proliferation of human pulmonary smooth muscle cells , it has not been determined whether all cases of PPH carrying mutations within the BMPR2 gene are caused by perturbation of BMP signals. Mutations are distributed throughout the coding region of the BMPR2 gene, suggesting heterogeneity of their contribution to the pathogenesis of PPH. Furthermore, many PPH kindreds carrying mutations of the BMPR2 gene do not develop any signs or symptoms, suggesting that additional environmental and/or genetic factors may be necessary for development of symptoms . These findings raised the following questions: 1) whether the signaling components of BMP/Smad pathways are present in human pulmonary endothelial and smooth muscle cells, 2) whether BMP signals are impaired by all types of mutations found in PPH patients, and 3) how signal-transducing capabilities are disrupted in the BMPR-II mutant proteins. In this study, we used various types of BMPR-II mutants found in patients with PPH to investigate their ability to transduce BMP signals and the biochemical mechanisms by which BMPR-II mutants interfere with BMP signaling. First, we showed that human pulmonary artery endothelial cells (HPAECs) and smooth muscle cells (PASMCs) expressed BMP/TGF-beta signaling components, suggesting that these cells may potentially transduce their signals. Next, we showed that some BMPR-II mutants lost most signal-transducing abilities, such as transcriptional activity and phosphorylation of Smad proteins, whereas others retained most of them. Some of the mutants with defects in signaling activities were predominantly located in cytoplasm and may bind a cytoplasmic pool of type I receptors. Taken together, the findings of the present study suggest that perturbation of BMP signaling in the pulmonary vascular system by some types of mutations may be involved in the pathogenesis of PPH, whereas with other types of mutations signals can still be transduced, suggesting that additional factors may be required for the development of PPH. MATERIALS AND METHODS : Cell Culture | HPAECs and PASMCs were obtained from Clonetics (San Diego, CA) and were maintained in EGM-2 and SmGM-2 (Clonetics), respectively. COS-7 and R-mutant mink lung epithelial cells were maintained in DMEM (Sigma-Aldrich, St. Louis, MO) containing 10% fetal bovine serum, 100 U/ml penicillin, and 100 mug/ml streptomycin. Reverse Transcription-Polymerase Chain Reaction (RT-PCR) Analysis | Total RNA was isolated from HPAECs and PASMCs with ISOGEN (NipponGene, Tokyo, Japan), and first-strand cDNA was synthesized using the Superscript First-Strand Synthesis System (Invitrogen, Carlsbad, CA) with random hexamer primers. Expression of various signaling components was compared by semiquantitative RT-PCR analysis. A human beta-actin primer set was used to normalize the amount of total cDNA in each sample. PCR products were separated by electrophoresis in agarose gel (1%) and visualized with ethidium bromide. The primer sequences, PCR programs, and expected sizes of PCR products are available online as indicated in Table 1. As controls, RNAs from HPAECs and PASMCs were analyzed for beta-actin expression without the prior generation of cDNA, and a PCR reaction for each set of primers was run against H2O. Plasmid Construction | Plasmids of the BMPR-II, ALKs, and Smads were described previously . Various mutant forms of BMPR-II were constructed by a PCR-based approach. An EcoRI and an XhoI site were added to the N terminus and C terminus of the BMPR-II cDNA, respectively, and the resulting fragments were subcloned into pcDNA3-FLAG and pcDNA3-HA, which add a FLAG-tag and hemagglutinin (HA)-tag, respectively, C-terminally to the insert . To increase levels of expression, inserts were subcloned into another expression vector, pcDEF3 . All of the PCR products were sequenced. The sequences of the mutagenesis primers are available upon request. Transfection, Immunoprecipitation, and Immunoblotting | COS-7 cells were transiently transfected using FuGENE6 (Roche Applied Science, Mannheim, Germany). The amounts of plasmids transfected are available online in Table 2. Immunoprecipitation and immunoblotting were performed as described previously using anti-HA 12CA5 (for immunoprecipitation; Roche Applied Science), anti-HA 3F10 (for immunoblotting; Roche Applied Science), anti-FLAG M2 (Sigma-Aldrich), and anti-phosphoserine antibodies (Zymed Laboratories, South San Francisco, CA). Luciferase Assay | R-mutant mink lung epithelial cells were transiently transfected with an appropriate combination of reporter constructs, expression plasmids, and pcDNA3. Total amounts of transfected DNAs were the same in each experiment. Luciferase activities were normalized using cotransfected sea pansy luciferase activity under the control of thymidine kinase promoter. Affinity Cross-Linking and Immunoprecipitation | Iodination of BMP-6, affinity cross-linking, and subsequent immunoprecipitation were performed as described previously . Briefly, recombinant BMP-6 was iodinated using the chloramine T method, and cross-linking was performed with disuccinimidyl suberate (, Rockford, IL). Cells were lysed and subjected to immunoprecipitation with anti-FLAG antibody followed by SDS-PAGE. Cross-linked receptor complexes were visualized by using a BAS 1800 Bio-Image Analyzer (Fuji Photo Film, Tokyo, Japan). Immunofluorescence Labeling | Immunohistochemical staining of FLAG-tagged BMPR-II in transiently transfected COS-7 cells was performed using anti-FLAG M2 antibody (Sigma-Aldrich), followed by incubation with fluorescein isothiocyanate-labeled goat anti-mouse IgG as described previously . Nuclei of the cells were stained by 4,6-diamidino-2-phenylindole. Subcellular localization was determined by confocal laser scanning microscopy (, Hercules, CA). RESULTS : Profiles of Expression of TGF-beta Superfamily Signaling Components in Pulmonary Vascular Cells | Recently, showed that PASMCs express type I (ALK-1, 4, 5, and 6) and type II (TGF-beta type II receptor [TbetaR-II], ActR-II, and BMPR-II) receptors for the TGF-beta superfamily. To further evaluate the expression of TGF-beta superfamily signaling components in HPAECs and PASMCs, we performed RT-PCR analysis to detect mRNA transcripts for ligands (BMP-2 and TGF-beta1), type I (ALK-1, 2, 3, 4, 5, and 6), type II receptors (BMPR-II, ActR-IIA, ActR-IIB, and TbetaR-II), endoglin, betaglycan, and Smads (Smad1, 2, 3, 4, and 5) . Figure 1 | Expression of TGF-beta superfamily signaling components in HPAECs and PASMCs. Expression of TGF-beta superfamily signaling components in HPAECs and PASMCs. RNA samples from HPAECs and PASMCs were analyzed by RT-PCR for expression of TGF-beta superfamily-signaling components and the housekeeping gene beta-actin. Two alternatively spliced forms, WT and SH, of BMPR-II mRNA transcripts were detected. As controls, RNAs from HPAECs and PASMCs were analyzed for beta-actin expression without the prior generation of cDNA, and a PCR reaction for each set of primers was run against H2O. Transcripts for both BMP-2 and TGF-beta1 were present in HPAECs and PASMCs. Among BMP type I receptors, ALK-2 and ALK-6 were expressed in both types of cells, whereas ALK-3 was expressed only in PASMCs. ALK-1 is a TGF-beta type I receptor that has been reported to be predominantly expressed in endothelial cells . We detected mRNA transcripts for ALK-1 in HPAECs but only very weakly in PASMCs, whereas we detected those for ALK-5 in both types of cells. Two alternatively spliced forms of BMPR-II mRNA transcripts have been reported . To examine which forms of BMPR-II are expressed in pulmonary vascular cells, we designed PCR primers that are able to generate distinct PCR products from the two spliced variants. As shown in Figure , transcripts for both the wild-type (WT) and short (SH) form of BMPR-II were detected in both types of cells, although intensities of the bands of BMPR-II (SH) were much weaker than those of BMPR-II (WT) for both types of cells. We also detected transcripts for other type II receptors, i.e., ActR-IIA, ActR-IIB, and TbetaR-II, and endoglin and betaglycan in both types of cells. Finally, the expression of Smads was examined in HPAECs and PASMCs. We detected mRNA transcripts for receptor-regulated Smads specific for BMPs (Smads 1, 5, amd 8), and those for TGF-betas and activins (Smads 2 and 3), and common-partner Smad (Smad4), in both of the cell types. Thus, both HPAECs and PASMCs express transcripts for most components of BMP and TGF-beta --signaling pathways, suggesting that pulmonary vascular cells are capable of responding to BMPs and TGF-betas. However, responses to these ligands may differ between HPAECs and PASMCs, because of their differences in expression profiles of type I receptors ALK-1 and ALK-3. Construction of BMPR-II Mutants Found in Patients with PPH | Because it seemed that BMP signals are intact in pulmonary vascular cells, we attempted to characterize the biological activities of the mutant forms of BMPR-II found in patients with PPH . BMPR-II has a structure essentially similar to those of other type II receptors for members of the TGF-beta superfamily. However, BMPR-II (WT) has a long cytoplasmic tail, the roles of which are not well understood (Figure A). In addition, an alternatively spliced form (SH) lacking the cytoplasmic tail exhibited no functional differences from BMPR-II (WT) when assayed using Xenopus embryos . Figure 2 | Biological activities of wild-type and mutant BMPR-II. Biological activities of wild-type and mutant BMPR-II. (A) Structure and location of mutations of WT, SH, and mutant BMPR-II used in the following experiments. Numbers indicate amino acid positions. Mutations are denoted by asterisks. Missense mutations in extracellular (E1) and kinase (K1 and K2) domain mutants and substituted amino acid residues are shown. Cytoplasmic tail mutants (T1 and T2) have frameshift or nonsense mutations resulting in truncated tails. (B and C) Transcriptional activation by wild-type and mutant BMPR-II. p3TP-Lux reporter gene was cotransfected into R-mutant mink lung epithelial cells with ALK-3 and wild-type and/or mutant forms of BMPR-II as indicated, and cells were stimulated with or without BMP-2 (100 ng/ml for B and 50 ng/ml for C). Luciferase activity was normalized against cotransfected sea pansy luciferase activity. Expression of cotransfected BMPR-II mutants was confirmed by immunoblotting of cell lysates with anti-FLAG antibodies (C, right). At least four types of germline mutations of the BMPR2 gene have been reported . The first type (type X) has nonsense or frameshift mutations in the extracellular domain, which lead to premature truncation of the transcripts and absence of the production of transmembrane BMPR-II proteins. The second type (type E) has missense mutations in the extracellular domain, most of which involve highly conserved cysteine residues. The third type (type K) has either missense or frameshift mutations in the kinase domain. The fourth type (type T) has frameshift or nonsense mutations within the cytoplasmic tail, resulting in cytoplasmic truncation of the receptor protein. To investigate the biological activities of the BMPR-II mutants, we constructed one or two of each of the three types of BMPR-II mutant (E1, K1, K2, T1, and T2) reported by the International PPH Consortium (Figure A). BMPR-II Mutants Found in PPH Patients Exhibited Differences in Transcriptional Activities | We first examined the transcriptional activities mediated by wild-type or mutant forms of BMPR-II by using p3TP-Lux, a TGF-beta --responsive promoter-reporter construct, which weakly responds to BMP signals . Coexpression of a BMP type I receptor (ALK-3) and WT or SH of BMPR-II induced transcriptional activation of p3TP-Lux, which was further enhanced in the presence of BMP-2 (Figure B). None of the E1, K1, or K2 mutants induced transcriptional activation of the reporter gene. In contrast, the T1 and T2 mutants maintained the ability to induce transcription from p3TP-Lux, suggesting that truncation of the cytoplasmic tail does not efficiently disrupt the transcriptional activity of BMPR-II. Essentially similar results were obtained using 3GC2-lux , a BMP-specific promoter-reporter construct (our unpublished data), suggesting that the transcriptional activities induced by BMPR-II mutants found in patients with PPH differ between the type E and K mutants and type T mutants. Because heterozygous mutations of the BMPR2 gene were reported to cause PPH, we examined the effects of the BMPR-II mutants on the p3TP-Lux transcriptional activity induced by BMPR-II (WT) (Figure C, left). When the E1 or K1 mutants were cotransfected with BMPR-II (WT), they repressed the transcriptional activity induced by BMPR-II (WT) in a dose-dependent manner, suggesting that the E1 and K1 mutants behave as dominant negative mutants. In contrast, the T1 or T2 mutant that retained transcriptional activities exhibited less dominant negative effect than the E1 and K1 mutants. In addition, the K2 mutant also showed less dominant negative effect, suggesting the functional heterogeneity within the type K mutants. BMPR-II Mutants Differentially Induce Phosphorylation of Smad5 | BMP receptor complexes propagate signals mainly through phosphorylation of Smads 1, 5, and 8, although there is evidence for involvement of Smad-independent pathways in this propagation . To elucidate whether the differences in transcriptional activities induced by BMPR-II mutants involve the activation of Smads, we analyzed the phosphorylation of Smad5 cotransfected with wild-type or mutant forms of BMPR-II into COS-7 cells . The WT and SH forms of BMPR-II phosphorylated Smad5, whereas the E1 and K1 mutants failed to do so. Phosphorylation of Smad5 by the K2 mutant was also significantly reduced (our unpublished data). In agreement with the transcriptional activities, the T1 mutant phosphorylated Smad5, although less efficiently than BMPR-II (WT). These findings suggest that the differences in transcriptional activities mediated by BMPR-II mutants found in PPH patients are due to their abilities to activate BMP-specific Smads. Figure 3 | Phosphorylation of FLAG-tagged Smad5 mediated by HA-tagged wild-type or mutant BMPR-IIs in transfected COS-7 cells. Phosphorylation of FLAG-tagged Smad5 mediated by HA-tagged wild-type or mutant BMPR-IIs in transfected COS-7 cells. Top, cell lysates were immunoprecipitated (IP) with anti-FLAG antibody followed by immunoblotting with anti-phosphoserine (P-Ser) antibody. Expression of Smad5 (middle) and BMPR-II (bottom) was confirmed by immunoblotting of cell lysates with anti-FLAG and anti-HA antibodies, respectively. Ligand-binding Abilities of E1 and K1 Mutants Are Decreased | To investigate the biochemical mechanisms by which the E1 and K1 mutants lost signal-transducing abilities, we examined the ligand-binding abilities of the wild-type and mutant forms of BMPR-II. COS-7 cells were cotransfected with ALK-3 and wild-type or mutant forms of BMPR-II, affinity cross-linked using 125I-BMP-6, and subjected to immunoprecipitation by using anti-FLAG antibody for BMPR-II. As shown in Figure , WT, SH, K2, T1, and T2 mutant receptors bound BMP-6 efficiently in the presence of ALK-3. In contrast, the E1 mutant carrying a mutation in the extracellular ligand-binding domain did not bind BMP-6, suggesting that its loss of ligand-binding ability resulted in the loss of Smad-phosphorylating ability. Intriguingly, we also found significant reduction of the ligand-binding ability of the K1 mutant carrying a mutation in the kinase domain, which may have, at least in part, caused its loss of Smad-phosphorylating ability. Figure 4 | Ligand-binding abilities of BMPR-II mutants. Ligand-binding abilities of BMPR-II mutants. COS-7 cells were transfected with FLAG-tagged BMPR-II (BMPR-II-FLAG) and HA-tagged ALK-3 (ALK-3-HA), followed by affinity cross-linking with 125I-BMP-6, and lysates were immunoprecipitated (IP) with anti-FLAG M2 antibody. Immuno-complexes were subjected to SDS-PAGE and visualized by Fuji BAS bio-image analyzer (top). Expression of BMPR-II (middle) and ALK-3 (bottom) was confirmed by immunoblotting of cell lysates with anti-FLAG and anti-HA antibodies, respectively. E1 and K1 Mutants Exhibit Altered Subcellular Localization | To determine how the ligand-binding abilities of the E1 and K1 mutants were reduced, we examined the subcellular localization of wild-type and mutant forms of BMPR-II. COS-7 cells transfected with the wild-type or mutant forms of BMPR-II were subjected to immunofluorescence staining. WT (Figure A) and the T1 mutant (Figure E) exhibited intense staining of the plasma membrane as well as the cytoplasm. In contrast, the E1 and K1 mutants carrying missense mutations of cysteine residues within the extracellular and kinase domains, respectively, were observed mostly in the cytoplasm (Figure , B and C), suggesting that reduction of the ligand binding abilities of the E1 and K1 mutants was due to their altered subcellular localization. The K2 mutant, carrying a missense mutation of aspartic acid within the kinase domain, was mainly located on the plasma membrane, suggesting that the mechanism of its loss of signal-transducing ability may be due to perturbation of kinase activity. Figure 5 | Differential subcellular localization of wild-type and mutant BMPR-II. Differential subcellular localization of wild-type and mutant BMPR-II. Subcellular distribution of FLAG-tagged wild-type (A), E1 (B), K1 (C), K2 (D), or T1 (E) mutant BMPR-II in transfected COS-7 cells. Permeabilized cells were subjected to immunofluorescence (fluorescein isothiocyanate; green) staining and observation by confocal laser scanning microscopy after nuclear staining with 4,6-diamidino-2-phenylindole (red). E1 and K1 Mutants Are Retained in the Intracellular Compartments with Type I Receptors | Many membrane and secreted proteins are posttranslationally modified by the addition of N-linked oligosaccharides. We expected that the altered subcellular localization of E1 and K1 mutants would be confirmed by their posttranslational modification. The E1 mutant protein was observed as a fast-migrating band compared with WT (Figure , bottom), suggesting that the E1 protein is retained in the intracellular compartments as a glycoprotein containing high-mannose-type oligosaccharides. The K1 mutant was observed as two bands, i.e., a fast-migrating band similar to the E1 mutant and a slowly migrating band similar to the BMPR-II (WT) protein. This finding suggests that a considerable portion of the K1 mutant is also retained in the intracellular compartments. Figure 6 | Hetero-oligomerization of FLAG-tagged BMPR-II with HA-tagged ALK-3 in transfected COS-7 cells. Hetero-oligomerization of FLAG-tagged BMPR-II with HA-tagged ALK-3 in transfected COS-7 cells. Top, cell lysates were immunoprecipitated with anti-FLAG antibody followed by immunoblotting with anti-HA antibody. Expression of ALK-3 (middle) and BMPR-II (bottom) was confirmed by immunoblotting of cell lysates with anti-FLAG and anti-HA antibodies, respectively. Fast- and slowly migrating bands of BMPR-II and ALK-3 are indicated by open and closed triangles, respectively. To determine how the altered subcellular localization of the BMPR-II mutants affects complex formation with type I receptors, we examined the hetero-oligomerization of BMPR-II mutants with ALK-3. COS-7 cells cotransfected with ALK-3 and wild-type or mutant forms of BMPR-II were subjected to FLAG-immunoprecipitation for BMPR-II, followed by HA-immunoblotting for ALK-3. BMPR-II (WT), BMPR-II (SH), and the T1 mutant formed complexes with slowly migrating forms of ALK-3, whereas the E1 and K1 mutants formed complexes predominantly with fast-migrating forms of ALK-3, which may contain high-mannose-type oligosaccharides (Figure , top). These results suggest that the E1 and K1 mutants are located in the intracellular compartments and that they may preferentially form complexes with the type I receptors located in the same compartments. DISCUSSION : Roles of BMP and TGF-beta Signaling in Maintenance of Vascular Systems | TGF-beta plays important roles during yolk sac vasculogenesis as well as late stages of angiogenesis by growth inhibition and production of extracellular matrix of endothelial cells . In endothelial cells, two types of TGF-beta type I receptors, ALK-1 and ALK-5, mediate TGF-beta signaling. ALK-5 is ubiquitously expressed in TGF-beta --responsive cells and activates Smad2 and Smad3. In contrast, ALK-1 is predominantly expressed in endothelial cells and activates BMP-specific Smad1 and Smad5. These observations suggest that balance between Smad1/5/8 and Smad2/3 pathways is important in determining vascular endothelial properties during angiogenesis . Endoglin is a dimeric glycoprotein with a short intracellular region that is structurally similar to betaglycan (also known as TGF-beta type III receptor). Endoglin binds TGF-beta as well as BMP-2 and BMP-7, suggesting that it may regulate both TGF-beta and BMP signaling pathways . Interestingly, mutations of ALK-1 and endoglin have been found in patients with hereditary hemorrhagic telangiectasia . Taken together with the findings that the BMPR2 gene is mutated in PPH patients, our findings suggest that TGF-beta/BMP signals mediated by Smad1, 5, and 8 may play important roles in maintenance of vascular homeostasis. Recently, showed that PASMCs express receptors for TGF-beta and BMPs, and that BMP suppressed the DNA synthesis and proliferation of PASMCs from patients with secondary pulmonary hypertensions, but did not suppress those from patients with PPH . The present study showed that both HPAECs and PASMCs express most of the signaling components required for TGF-beta/BMP signal transduction, including ligands, receptors, and Smads . However, response to TGF-beta and BMPs may differ between HPAECs and PASMCs. Because HPAECs express both ALK-5 and ALK-1, TGF-beta may activate Smad2/3 and Smad1/5 pathways, similar to other endothelial cells. Because PASMCs do not express ALK-1, the Smad1/5 pathways may not be activated by TGF-beta. Intriguingly, HPAECs express ALK-2 and ALK-6, but not ALK-3, suggesting that they respond to BMP-6 and BMP-7 through ALK-2 and ALK-6, but not to BMP-4, which binds to ALK-3 . In contrast, PASMCs express ALK-2, 3, and 6, suggesting that they respond to BMP-6 and -7 as well as to BMP-4. One of the features of PPH is overproliferation of endothelial cells and smooth muscle cells. Taken together with results of previous studies showing that BMPs have growth inhibitory effects on smooth muscle cells , these findings suggest that it is likely that BMP signals maintain pulmonary vascular integrity by suppressing the overproliferation of cells and that reduction of BMP signals caused by mutations of the BMPR2 gene eventually results in symptoms of PPH. How Did Type E and K Mutants Lose Their Signal-transducing Abilities? | In the present study, we generated five BMPR-II mutants, i.e., those mutated in the extracellular domain (E1), kinase domain (K1 and K2), or cytoplasmic tail (T1 and T2) (Figure A), to examine the biological activities of the BMPR-II mutants found in PPH patients. We found that the type E and K mutants lost their transcriptional activities, whereas the type T mutants maintained transcriptional activities although they were less potent than those of BMPR-II (WT). This suggests that these BMPR-II mutants have different biological activities. To date, all missense mutations within the extracellular domain of BMPR-II have been found at cysteine residues in PPH patients . Interestingly, extracellular cysteine residues have been shown to be essential for formation of proper three-dimensional structure and to be required for membrane targeting of some receptors . Consistent with this, we found that most of the E1 mutant proteins mutated at cysteine-118 were present in the cytoplasm (Figure B). These results suggest that loss of signal-transducing abilities due to missense mutations in the extracellular ligand-binding region is due not only to loss of ligand-binding ability of the extracellular domain but also to altered subcellular localization. Notably, the E1 mutant protein migrated faster than the BMPR-II (WT) protein (Figure , bottom), implying differential posttranslational modification due to abnormal subcellular localization of the E1 protein. When ALK-3 was coexpressed with the E1 mutant, only fast-migrating protein bands of ALK-3 formed complexes with the E1 mutant proteins (Figure , top). Many membrane-targeted proteins are posttranslationally modified by addition of N-linked oligosaccharides during transport through the Golgi apparatus. Treatment of ALK-3 and BMPR-II with N-glycosidase F resulted in shift of slowly migrating bands of ALK-3 and BMPR-II to fast-migrating bands (our unpublished data), suggesting that the fast-migrating proteins of the E1 mutant and ALK-3 may contain high-mannose-type oligosaccharides and that they are retained in the cytoplasm as a complex. These results raised the possibility that dominant negative effects of the E1 mutant against BMPR-II (WT) may be due to sequestration of BMP type I receptors in the intracellular compartments. On the other hand, missense mutations within the kinase region were identified at various amino acid residues, including cysteine, aspartic acid, and arginine residues . The K1 mutant, with substitution of cysteine-347 by tyrosine, exhibited a reduced ligand-binding ability than BMPR-II (WT). This can be explained by the distribution of mutant proteins partially in cytoplasm, as demonstrated by immunohistochemical analysis and by the presence of fast-migrating bands on immunoblot analysis (Figures C and , bottom). However, this distribution profile of the K1 mutant proteins cannot fully explain the loss of signal-transducing ability and gain of dominant negative activity by them, which were equivalent to those of the E1 mutant. Kinase activity was probably lost in the K1 mutant, resulting in the potent dominant negative effects of this mutant. In agreement with this, a BMPR-II kinase negative mutant exhibited a dominant negative effect against ActR-II in transcriptional activation activity . The K2 mutant, with substitution of aspartic acid-485 by glycine, exhibited normal ligand-binding ability and subcellular localization (Figure D), but lost signal-transducing ability (Figure B). Kinase activity was probably lost in the K2 mutant, which may have caused the loss of transcriptional activity; however, how the K2 mutant has less dominant negative effect remains unknown. Role of BMPR-II Mutants with Truncation of Cytoplasmic Tail in Pathogenesis of PPH | BMPR-II is structurally similar to other type II receptors of the TGF-beta superfamily, e.g., TbetaR-II, ActR-IIA, and ActR-IIB. However, BMPR-II has a long cytoplasmic tail that is not found in other type II receptors in mammals. The functions of the cytoplasmic tail of BMPR-II are not yet clear. The fact that truncation of the cytoplasmic tail of BMPR-II was found in type T mutants from patients with PPH suggests novel functions for this region. Compared with the E1 and K1 mutants, however, the T1 mutant retained most of its biological activity with the exception that it phosphorylated Smad5 less efficiently than WT or SH forms of BMPR-II. analyzed the transcriptional activities of BMPR-II mutants K2 and T1 according to our nomenclature, in NMuMG cells in which endogenous BMP signaling pathways are intact. Although they concluded that both of mutants lost their signaling capabilities, their results showed that only the K2 mutant, but not the T1 mutant, inhibited endogenous BMP signals. Thus, there may be significant differences in biological activities between the K2 and T1 mutants, consistent with our results. Recently, showed that BMPR-II mutants completely lacking the cytoplasmic tail were capable of transducing BMP-2 signals similar to BMPR-II (SH). Taken together with the present findings, these results suggest that the cytoplasmic tail of BMPR-II may not be essential for transduction of BMP signals through Smads, although it is possible that it has yet unidentified functions in BMP signaling. It will be important to determine whether other factors, such as additional genetic mutations and/or environmental factors, play important roles in the pathogenesis of PPH. Backmatter: PMID- 12221116 TI - Aurora B Kinase Exists in a Complex with Survivin and INCENP and Its Kinase Activity Is Stimulated by Survivin Binding and Phosphorylation AB - Aurora B regulates chromosome segregation and cytokinesis and is the first protein to be implicated as a regulator of bipolar attachment of spindle microtubules to kinetochores. Evidence from several systems suggests that Aurora B is physically associated with inner centromere protein (INCENP) in mitosis and has genetic interactions with Survivin. It is unclear whether the Aurora B and INCENP interaction is cell cycle regulated and if Survivin physically interacts in this complex. In this study, we cloned the Xenopus Survivin gene, examined its association with Aurora B and INCENP, and determined the effect of its binding on Aurora B kinase activity. We demonstrate that in the Xenopus early embryo, all of the detectable Survivin is in a complex with both Aurora B and INCENP throughout the cell cycle. Survivin and Aurora B bind different domains on INCENP. Aurora B activity is stimulated >10-fold in mitotic extracts; this activation is phosphatase sensitive, and the binding of Survivin is required for full Aurora B activity. We also find the hydrodynamic properties of the Aurora B/Survivin/INCENP complex are cell cycle regulated. Our data indicate that Aurora B kinase activity is regulated by both Survivin binding and cell cycle-dependent phosphorylation. Keywords: INTRODUCTION : Defects in chromosome segregation can generate aneuploidy, a condition that is found in almost all human tumors and is a major cause of miscarriages and birth defects. The complex process of chromosome segregation must be highly regulated to ensure fidelity and to prevent aneuploidy. Many of the mitotic events are regulated by the kinetochore, a proteinaceous structure assembled on centromeric DNA that coordinates at least three mitotic functions (for review, see ). First, the kinetochore is the chromosomal site of microtubule attachment and movement. Second, the kinetochore is the major site of cohesion between sister chromatids. This cohesion must be maintained through metaphase and its dissolution is the critical event that triggers anaphase. Third, kinetochores that are not attached to microtubules send signals to the cell cycle machinery to prevent this dissolution of cohesion, a process referred to as the spindle assembly checkpoint. This checkpoint ensures that all chromatids are attached before the onset of anaphase. How the kinetochore coordinates these various functions is a critical unanswered question. A group of mitotic regulators that includes Aurora B kinase and the inner centromere protein (INCENP) has been given the name chromosomal passengers . The passenger proteins are defined by a dynamic pattern of localization . In G2 and prophase, the passenger proteins have a general chromosomal localization. During prometaphase and metaphase, they move from chromosomes to inner centromeres. When chromosomes move to the poles during anaphase, passenger proteins remain at the spindle midzone where they eventually become part of the midbody . The phenotypes of the chromosomal passenger proteins suggest that they are critical regulators of various steps of chromosome segregation. RNAi experiments in Caenorhabditis elegans embryos and Drosophila cell lines suggest that cells lacking Aurora or INCENP have similar mitotic defects. First, the passenger proteins are necessary for the proper segregation of DNA. During anaphase, the chromosome masses do not properly segregate, leaving a chromatin bridge between the major DNA masses . Second, cytokinetic furrows begin but fail to fully progress in cells lacking either Aurora or INCENP . This suggests a second role for Aurora and INCENP in cytokinesis. Third, studies from budding yeast suggest that Aurora is involved in bipolar attachment of microtubules to kinetochores. The Aurora homolog Ipl1 is required during chromosome segregation, as cells without Ipl1 activity have massive nondisjunction and often segregate both sisters' chromosomes to the same pole. Ipl1 is required for kinetochores to release spindle microtubules both in vitro and in vivo . Current models propose that Ipl1 ensures that each chromosome obtains bipolar attachment of microtubules by coordinating the release of kinetochores attached from a single pole, thereby allowing kinetochores to rebind microtubules until proper bipolar attachment is achieved. This model predicts that Aurora B kinase activity is inactivated by bipolar attachment, but we presently know little about how Aurora B activity is regulated. Some overlapping phenotypes are seen in embryos lacking the Survivin/Bir1 protein (Survivin). Survivin is required in both budding and fission yeast for proper chromosome segregation . RNAi experiments in C. elegans have shown that embryos lacking Survivin display abnormal chromosome condensation, disrupted mitotic spindles, and were ultimately unable to complete cytokinesis, resulting in multinucleate embryos . Survivin-null mouse embryos displayed polyploidy, abnormal mitotic spindles, and failed cytokinesis . The similarities in phenotypes suggest that Aurora, INCENP, and Survivin could function together. INCENP and Survivin have both been shown to genetically interact with Aurora B kinases. Chan and colleagues first identified mutants of the budding yeast INCENP homolog (Sli15) that are synthetically lethal with temperature-sensitive mutants of the Aurora homolog Ipl1. Moreover, they showed that sli15 cells had phenotypes identical to those of ipl1 yeast. As discussed earlier, similar phenotypes are also seen in fission yeast, C. elegans, and Drosophila cells lacking either Survivin, Aurora, or INCENP (for review, see ). Aurora B kinase is not localized to the kinetochores in fission yeast or C. elegans embryos lacking Survivin . Similarly, in C. elegans embryos and Drosophila cells, loss of INCENP by RNAi also leads to the mislocalization of Aurora B kinase . Biochemical evidence has shown that Aurora B physically interacts with INCENP. Sli15p and Ipl1p form a complex in budding yeast , and a complex containing both INCENP and Aurora B kinase has been purified from Xenopus laevis mitotic extracts . There is no conclusive biochemical evidence that Aurora kinases or INCENP are physically associated with Survivin, although a recent report has shown that human Survivin can interact with either Xenopus INCENP (xINCENP) or Aurora B kinase in both two-hybrid and in vitro pull-down assays . Therefore, some essential questions are whether Aurora, INCENP, and Survivin physically interact in vivo, whether complex formation is cell cycle regulated, and how each subunit interacts in the complex. Moreover, it is critical to identify the molecular function(s) of each protein in the complex. To understand the interrelationship of the passenger proteins and to further understand how Aurora B kinase is regulated, we have cloned the Xenopus Survivin (xSurvivin) gene. xSurvivin is shown to exist in a complex with both xINCENP and Xenopus Aurora B kinase (xAurora B) in S-phase (interphase) and mitotic Xenopus extracts. Moreover, immunodepletion of xAurora B kinase can completely remove xSurvivin and xINCENP from Xenopus extracts, suggesting that all of the xSurvivin and xINCENP is physically associated with xAurora B kinase. We show that the N terminus of xAurora B kinase interacts with the conserved C terminus of xINCENP, whereas xSurvivin interacts with the N terminus of xINCENP. Furthermore, xAurora B activity is stimulated at least 10-fold in mitotic extracts, and this stimulation is shown to be phosphatase sensitive. Adding recombinant xSurvivin protein to xAurora B immunoprecipitations (IPs) stimulates the mitotic kinase activity an additional 10-fold, suggesting that xSurvivin binding to Aurora B plays a regulatory role similar to cyclin binding of CDKs. Therefore, our data suggests that xAurora B kinase is regulated by both complex formation and phosphorylation. MATERIALS AND METHODS : Materials | All chemicals were purchased from Sigma (St. Louis, MO) unless stated otherwise. All DNA restriction enzymes were purchased from (Beverly, MA). Adult wild-type Xenopus laevis were purchased from Nasco (Fort Atkinson, WI). Xenopus Interphase and Mitotic Extracts | Interphase extracts were prepared as previously described . Interphase extracts were driven into mitosis by the addition of 66 nM nondegradable glutathione S-transferase (GST)-cyclin B for 20 min at room temperature. Fusion Constructs and Protein Purification | Based on the sequence of xAurora B , primers 1292098 (5'-GCGGATCCATGGAGTACAAAGAGAATCTCAACCC) and 1292100 (5'-CGGCGGCCGCTTTTGATTGGGTGGACTGGTAGAC) were used to amplify xAurora B from a Xenopus stage 11.5 --14 cDNA library. This polymerase chain reaction (PCR) fragment was subcloned into the NcoI-NotI sites of pET 28 (Novagen, Madison, WI) to create pET 28-xAurB that expresses C-terminal 6His-tagged xAurora B. xAurora B was also amplified using primers 1292098 and 1292099 (5'-GGCTCGAGAAGCTTCATTTTGATTGGGTGGACTGG). The resulting PCR fragment was subcloned into the NcoI-HindIII sites of pET 41 (Novagen) to yield pET 41-xAurB that expresses N-terminal GST-tagged xAurora B. The gene was sequenced by the University of Virginia Biomolecular Research Facility (Charlottesville, VA), and its sequence was deposited into GenBank (accession no. ). Our isolated gene has only two amino acid changes from the published gene. As both of these changes are also found on each of the GenBank deposited expressed sequence tags (ESTs), they are unlikely to be mutations caused by the PCR isolation. Primers 1292099 and 1984819 (5'-GCCCATGGAATTCCCATTGGGGAAGGGG) were used to amplify the xAurora B kinase domain. The resulting PCR fragment was subcloned into the NcoI-HindIII sites of pFastBac and was subsequently subcloned into the NcoI-HindIII sites of pET 41B to create pET 41-xAurB (99 --384) that expresses N-terminal GST-tagged xAurora B kinase domain. To clone xSurvivin, a BLAST search of the EST database was conducted to find a Xenopus EST with high homology to human and mouse Survivin. xSurvivin was then amplified from a Xenopus stage 11.5 --14 cDNA library using primers 1242204 (5'-CTGGCCGGCCCCATATGTATTCTGCCAAGAACAGG) and 1242206 (5'-CGCTCGGGTGGTCGAGATCTATGGAGCACTG). This PCR fragment was subcloned into the NdeI-XhoI sites of pET 41 (Novagen) to yield pET 41-xSurvivin that expresses C-terminal 6His-tagged xSurvivin. xSurvivin was also amplified using primers 1242204 and 1242205 (5'-CCGGCGCGCCTCAGTGGTCAAGATCTATGGAGCAA). The resulting PCR fragment was subcloned into the NdeI-Asc1 sites of pGEXcsFA (a generous gift from Ethan Lee, Harvard Medical School, Boston, MA) to yield pGEXcsFA-xSurvivin that expresses N-terminal GST-tagged xSurvivin. xSurvivin sequence was deposited into GenBank (accession no. ). The modeled crystal structure coordinates were generated by SWISS-MODEL . The figures were constructed with RasMol. A XhoI site was engineered into the N terminus of our pCS2+xINCENP clone using PCR with the oligonucleotides (5'-CCGCGCTCGAGAACGATGCAGA-GTGCCGTGCCC) and (5'CCGGCGGGGCCCTCTAGAGGATCCTCGTATTTGAGGCCATAACC). The resulting product was cloned into the XhoI-Xba sites of Super GFP-Wee1 to form a GFP-xINCENP fusion protein. A HindIII to ApaI fragment containing GFP and the N terminus of xINCENP was cloned back into pCS2+xINCENP, and the remaining PCR product was confirmed by sequence to generate pCS2+GFP-xINCENP. The XhoI to XbaI fragment was cloned into pET 28B in the XhoI and Bpu1103 sites, and finally a NheI to PstI fragment of pET 28B-xINCENP was removed to generate pET 28-xINCENP (677 --874) that expresses N-terminal 6His-tagged xINCENP (677 --874). All proteins were expressed in the Escherichia coli strain BL21 (DE3 pLysS; Novagen). 6His-tagged proteins were purified on Ni2+-NTA agarose (Qiagen, Valencia, CA) as instructed by the manufacturer. GST-tagged proteins were purified on glutathione agarose . Antibody Production, IP, and Immunoblotting | All polyclonal antibodies were made by Covance Research Products (Denver, PA). To make anti-xAurora B antibodies, rabbits 315 and 316 were immunized with purified C-terminal 6His-tagged xAurora B. To make anti-xSurvivin antibodies, rabbits 342 and 343 were immunized with purified C-terminal 6His-tagged xSurvivin. Anti-xINCENP antibodies were produced in rabbits 354 and 355 immunized with N-terminal 6His-tagged xINCENP (677 --874) encoding the C-terminal fragment of xINCENP. All antibodies were affinity purified on the corresponding immunizing protein coupled to a cyanogen bromide-activated Sepharose column (Amersham Biosciences, Piscataway, NJ) as described . After affinity purification, the antibodies were dialyzed into XB no Ca2+ (10 mM HEPES, pH 7.7, 1 mM MgCl2, 100 mM KCl, and 50 mM sucrose), aliquoted, and stored at -80C. Immunoblotting was performed on Immobilon-P (Millipore, Bedford, MA) and was detected using horseradish peroxidase-conjugated-anti-rabbit secondary antibodies (Jackson ImmunoResearch Laboratories, West Grove, PA) by chemiluminescence (ECL; Amersham Biosciences). To immunoprecipitate complexes from Xenopus extracts, 20 mug of preimmune sera (Pre-I) 1 mug of affinity-purified anti-xAurora B, anti-xSurvivin, or anti-xINCENP antibodies were bound to 50 mul of protein A Sepharose beads (Amersham Biosciences) in phosphate-buffered saline (PBS) for 1 h at room temperature. The beads were washed five times in XB no Ca2+, and one-half was added to 40 mul of interphase extract and the other one-half was added to 40 mul of interphase extract driven into mitosis by the addition of GST-cyclin B. Beads were incubated with extracts for 1 h at 4C and were subsequently washed five times in XB no Ca2+ containing an additional 200 mM NaCl and 0.1% Brij 35. The concentration of xAurora B, xSurvivin, and xINCENP in Xenopus extracts was determined by quantitatively comparing interphase extracts with recombinant GST-xSurvivin, GST-xAurora B, and 6His-xINCENP (677 --874) levels on immunoblots. For quantification, the amount of INCENP in Xenopus extracts was compared with titrations of 6His-xINCENP (677 --874) transferred to Immobilon by slot blots rather than by SDS-PAGE followed by transfer. To quantify immunoblots, films were scanned with densitometer hardware (Molecular Dynamics, Sunnyvale, CA) into ImageQuant 5.0. Data were then exported into Microsoft Excel and graphed for analysis. Cell Culture and Immunofluorescence | XTC cells were obtained from Dr. Douglas DeSimone (University of Virginia, Charlottesville, VA) and were grown at room temperature in 70% L-15 media supplemented with 10% fetal bovine serum and 1 mM sodium pyruvate. XTC cells were grown on coverslips and fixed with 4% paraformaldehyde in PHEM buffer (60 mM PIPES, pH 6.9, 25 mM HEPES, 10 mM EGTA, and 4 mM MgCl2) containing 0.75% Triton-X-100. Coverslips were subsequently washed in PBS containing 0.1% Tween 20 (PBS-T). Fixed cells were blocked in 20% heat-inactivated goat serum (Invitrogen, Carlsbad, CA) and were then incubated for 1 h at room temperature with monoclonal anti-tubulin antibody at 1:500 and either affinity-purified anti-xAurora B (1 mug/ml) or anti-xINCENP (1 mug/ml) antibodies diluted in PBS-T. Coverslips were then washed three times with PBS-T and then incubated with fluorescein isothiocyanate-conjugated anti-mouse immunoglobulin G at 1:1000 and Cy3-conjugated anti-rabbit immunoglobulin G (Jackson ImmunoResearch Laboratories) at 1:1000, washed three times in PBS-T, stained with 1 mug/ml Hoechst No. 33342 for 30 s, and mounted with Vectashield (Vector Laboratories, Burlingame, CA). Images were collected at x100 using an E600 Eclipse microscope (Nikon, Tokyo, Japan) equipped with a CH350 cooled charged-coupled device camera (Photometrics, Tucson, AZ). Images were acquired and analyzed using version 4.5 Meta View software. Gel Filtration Chromatography and Sucrose Gradient Sedimentation | For gel filtration analysis, 100 mul of interphase extract was diluted 1:5 in XB no Ca2+. In parallel, microcystin (5 muM) was added to 100 mul of mitotic extracts and the mitotic extract was diluted 1:5 in 80 mM beta-glycerophosphate, 20 mM EGTA, 10 mM MgCl2, and 5 mM NaF. Clarified interphase and mitotic extracts were prepared by ultracentrifugation at 40,000 rpm for 1 h at 4C in an S120-AT2 rotor (Sorvall, Kendro Laboratory Products, Newtown, CT). The extracts were then concentrated in a Microcon (Millipore) and were subsequently loaded onto a Superose 6 column (Amersham Biosciences). Thirty-two fractions of 500 mul were collected and trichloroacetic acid precipitated as described . Alternating fractions were loaded onto 8 and 15% SDS-PAGE gels and immunoblotted for xINCENP, xAurora B, and xSurvivin. Similarly clarified extracts were loaded onto 5 --30% sucrose gradients and centrifuged at 30,000 rpm for 30 h at 4C in a SW-40 rotor. Sixteen fractions were collected, trichloroacetic acid precipitated, and equal volumes of sample were separated on 8 and 15% gels for immunoblot analysis. Molecular weight markers were run in a parallel gradient. Mapping of Passenger Protein-Binding Domains | To identify binding domains among the passenger proteins, the indicated truncated or wild-type proteins were translated in vitro in the presence of [35S]methionine using the TNT Coupled Reticulocyte Lysate System (Promega, Madison, WI). These proteins were allowed to integrate into endogenous complexes by mixing 10 mul of the in vitro translation reaction with 20 mul of Xenopus extract for 30 min, and then the endogenous protein was immunoprecipitated. Labeled proteins were detected by phosphorimager analysis of dried SDS-PAGE gels. Proteins were scored as interacting if the amount of labeled protein immunoprecipitated was fivefold higher than the amount precipitated by Pre-I; typically >80% of a labeled protein was precipitated in these experiments. The deletion set of xINCENP was a generous gift of Susannah Rankin and Marc Kirschner (Harvard Medical School). Kinase Assays | For each kinase assay, xAurora B kinase was immunoprecipitated from 50 mul of Xenopus extract on similar10 mul of protein A Sepharose beads (Amersham Biosciences). The beads were washed four times in XB no Ca2+ containing an additional 300 mM NaCl and 0.1% Triton-X-100, followed by an additional wash in kinase buffer (20 mM Tris, pH 7.5, 1 mM MgCl2, 25 mM KCl, 1 mM dithiothreitol, and 40 mug/ml bovine serum albumin). Beads were aliquoted into tubes containing the indicated recombinant protein in 25 mul of kinase buffer supplemented with 100 muM gamma[32P]ATP (2 muCi; Perkin Elmer-Cetus Life Sciences, Norwalk, CT) and 1 mug of myelin basic protein (MBP; Invitrogen). Reactions were incubated in a Mixer 5432 (Eppendorf; Brinkmann Instruments, Westbury, NY) at room temperature, and after 10 min, the reactions were stopped by the addition of sample buffer. One-half of the reaction was loaded onto a gel that was Coomassie stained, dried on 3MM paper (Whatman, Clifton, NJ), and analyzed for 32PO4 incorporation with phosphorimager hardware (Molecular Dynamics) and ImageQuant 5.0 software. The other one-half of the reaction was loaded onto a second gel and the amount of xAurora kinase was quantified by immunoblot. In kinase reactions involving lambda phosphatase, beads were washed into lambda phosphatase buffer, treated at room temperature for 30 min with either 200 U of lambda phosphatase or buffer, subsequently washed four times in kinase buffer, and assayed for kinase activity. RESULTS : Cloning the Xenopus Survivin Gene | To biochemically dissect the regulation of passenger proteins, we used embryonic extracts of the frog Xenopus laevis. These extracts are excellent tools for studying mitotic regulators for three reasons. First, cell cycle regulators are stockpiled in Xenopus eggs and are more abundant than in somatic cells. Second, the cell cycle state of these extracts can be easily controlled and, third, the extracts can recapitulate mitotic events in vitro. We have previously cloned the xINCENP gene and demonstrated that the protein is specifically phosphorylated in mitosis . A gene encoding the xAurora B protein was recently identified , and a highly related gene was isolated from a Xenopus stage 11.5 --14 plasmid cDNA library by PCR. The xSurvivin cDNA was cloned by identifying a Xenopus EST with high homology to human and mouse Survivin. The Survivin cDNA was then isolated from a Xenopus stage 11.5 --14 cDNA library by PCR. The Xenopus Survivin cDNA encodes a protein that is 46% identical to human Survivin (Figure A). The predominant motif on the Survivin protein is an inhibitor of apoptosis domain, a motif that has been shown in a related baculoviral protein to inhibit Caspase-3 activity (for review, see ; ), although the Survivin family does not inhibit Caspase-3 . The xSurvivin protein has the recognizable motifs of Survivin including the inhibitor of apoptosis repeat and a Cdc2 kinase phosphorylation site. Two independent groups have recently published the crystal structure of human Survivin . The Xenopus sequence was modeled into these crystal structures by the program SWISS-MODEL. The program predicts that the backbone structures of the Xenopus and the human proteins are highly similar (Figure , B and C). A prominent motif that is not included on the model is a C-terminal extended alpha helix (Figure , B and C). This is not included because this region is divergent in the primary sequence (Figure A); however, a SOPM secondary structure of this region predicted a long C-terminal alpha helix. Therefore, there is structural similarity between the Survivin homologs throughout the protein. Noel and colleagues identified two potential interaction regions (an acidic patch and a basic patch) on the surface of human Survivin. To determine if these regions are conserved in xSurvivin, we compared the charged amino acid residues of the xSurvivin model (Figure D) with the placement of charged residues in the solved crystal structure of human Survivin (Figure E). The placement of surface charge on xSurvivin is predicted to be almost identical to human Survivin. Figure 1 | Structural and sequence homology of Xenopus Survivin. Structural and sequence homology of Xenopus Survivin. (A) Clustal-W alignment of the amino acids encoding sequenced vertebrate Survivins. Significant residues from the crystal structure of human Survivin are conserved and are indicated by letters or symbols above the alignment: P-cdc2 phosphorylation site; star-residues that coordinate the Zn2+ in the Bir domain. (B-E) Comparison of solved human Survivin structure and a structural prediction of Xenopus Survivin protein produced by the SWISS-MODEL server . Ribbon diagram of Xenopus Survivin (B) and human Survivin (C). Charge distribution in Xenopus (D) and human (E) space-filled Survivin models where acidic charged residues are shown in red and basic residues are shown in blue. Note that the long C-terminal alpha helix is present in the Xenopus sequence; however, the homology is too low to fit in the modeled structure. Characterization of Antibodies | After affinity purification, anti-xAurora B and anti-xINCENP antibodies are specific as they predominantly recognize a band of 41 and 130 kDa, respectively, in a Xenopus interphase extract and a Xenopus tissue culture cell line (XTC) lysate (Figure , B-C). Anti-xSurvivin antibody recognizes a predominant band at 18 kDa (Figure A). Even after affinity purification, the anti-xSurvivin antibody recognizes six additional bands; therefore, it was not used for immunofluorescence. The antibodies are highly sensitive as they all recognize <1 ng of recombinant protein (unpublished data, M. Bolton, P.T. Stukenberg). By comparing immunoblot signals of Xenopus interphase extracts with known amounts of recombinant protein, we estimate that the concentration of xAurora B protein in these extracts is 500 nM, the concentration of xSurvivin is similar250 nM, and the concentration of xINCENP is 100 nM (unpublished data, M. Bolton, P.T. Stukenberg). Figure 2 | Specificity of affinity-purified anti-xSurvivin, anti-xAurora B, and anti-xINCENP antibodies in Xenopus interphase extracts and XTC cells. Specificity of affinity-purified anti-xSurvivin, anti-xAurora B, and anti-xINCENP antibodies in Xenopus interphase extracts and XTC cells. (A) Affinity-purified anti-xSurvivin antibody recognizes a band of 18 kDa in an interphase extract. (B) Affinity-purified anti-xAurora B antibody recognizes a band of 41 kDa in an interphase extract and in XTC cells. (C) Affinity-purified anti-xINCENP antibody recognizes a band of 130 kDa in an interphase extract and in XTC cells. (D-AA) Immunofluorescence localization of xAurora B (D-O) and xINCENP (P-AA) in XTC cells. (D-I) Immunolocalization of xAurora B during interphase (D) and throughout mitosis (E-I). (P-U) Immunolocalization of xINCENP during interphase (P) and throughout mitosis (Q-U). Colocalization of either xAurora B (J-O) or xINCENP (V-AA) with DNA and microtubules. Blue, DAPI; green, microtubules; red, xAurora B in (K-M), xINCENP in (W-Y); yellow and orange, overlap of xAurora B or xINCENP with microtubules. XTC cells stained with the anti-xAurora B antibody (Figure , D-O) and anti-xINCENP antibody (Figure , P-AA) display typical passenger protein immunolocalization. The concentration of both proteins is low in interphase (Figure , D, J, P, and V), but by prophase, the antibodies recognize both chromatin and kinetochores (Figure , E, K, Q, and W). Both proteins move to kinetochores at prometaphase-metaphase (Figure , F-G, L-M, R-S, and X-Y), and finally to bundles of overlapping microtubules in anaphase (Figure , H, N, T, and Z) where they remain throughout cytokinesis until they are discarded with the midbody (Figure , I, O, U, and AA). These data suggest that the roles of passenger proteins are conserved in Xenopus and demonstrate that the anti-xAurora B and anti-xINCENP antibodies are highly specific. xAurora B, xINCENP, and xSurvivin Are Physically Associated In Vivo | To determine the native molecular weight of the passenger proteins, Xenopus interphase and mitotic extracts were size separated by both Superose 6 gel filtration chromatography and 5 --30% sucrose gradients, and the location of each of the proteins was detected by immunoblot. In interphase egg extracts, xAurora B, xINCENP, and xSurvivin comigrate in a broad peak from 900 to 600 kDa, whereas all three proteins migrate in a sharp peak at 900 kDa in mitotic extracts (Figure A). A slower migrating form of xSurvivin is often seen in SDS-PAGE gels. Detection of this form depends on phosphatase inhibitors and its appearance is not cell cycle dependent (unpublished data, M. Bolton, W. Lan, and P.T. Stuken-berg). xAurora B, xINCENP, and xSurvivin also comigrate in a sucrose gradient, sedimenting at 9.5S in interphase and at 4.5S in mitosis (Figure B). A number of control proteins, including Eg2, the Xenopus Aurora A homolog, migrated similarly in interphase and mitosis, demonstrating that the differences in migration of the passenger protein complex were caused by cell cycle changes and not due to differences in the manipulations of the sucrose gradients (Figure B). Because of the contributions of shape, neither gel filtration nor sedimentation can accurately predict native molecular weight. Gel filtration measures the Stokes radius, which is determined by the overall length of the molecule, whereas sucrose gradients measure sedimentation, which is retarded by a long thin shape. Therefore, highly elongated molecules such as myosin migrate faster than spherical proteins of similar molecular weight in gel filtration, whereas they sediment more slowly than spheres in sucrose gradients . have derived equations using the hydrodynamic characteristics of molecules to overcome the contributions of shape and more accurately measure molecular weight. Such calculations suggest that the mitotic complex is one-half the size (similar250 kDa) of the interphase complex (similar490 kDa). Moreover, in both complexes, the overall length is comparable because they migrate similarly by gel filtration. The predicted axial ratio is >20, indicating a size and shape reminiscent of fibrinogen, but not as elongated as myosin. INCENP has a long coiled-coil domain, which is a motif that often contributes to such elongated shapes in proteins. Thus, the passenger proteins complex has a highly elongated shape in both interphase and mitosis, and the mass of the complex doubles in interphase. Figure 3 | xSurvivin, xAurora B, and xINCENP cofractionate on gel filtration columns and sucrose density gradients during interphase and mitosis. xSurvivin, xAurora B, and xINCENP cofractionate on gel filtration columns and sucrose density gradients during interphase and mitosis. (A) Interphase and mitotic extracts were separated by Superose 6 gel filtration chromatography and equal volumes of alternate fractions were run on SDS-PAGE gels. The fractions were then immunoblotted for xINCENP, xAurora B, and xSurvivin to determine the molecular weight of the passenger protein complex. The molecular weights of the void (2000 kDa), thyroglobulin (670 kDa), bovine gamma globulin (158 kDa), chicken ovalbumin (44 kDa), and equine myoglobin (17 kDa) are shown. (B) xINCENP, xAurora B, and xSurvivin cofractionate on a sucrose density gradient, but the interphase complex migrates faster than the mitotic complex. Interphase and mitotic extracts were sedimented on 5 --30% sucrose density gradients. Equal volumes of each fraction were run on SDS-PAGE gels and immunoblotted for xINCENP, xAurora B, xAurora A (Eg2), and xSurvivin. Thyroglobulin (19S), bovine gamma globulin (7S), chicken ovalbumin (3.5S), and equine myoglobin (2S) were sedimented in a parallel gradient as markers. To directly test if the passenger proteins are in the same complex in vivo, we immunoprecipitated each of the passenger proteins from Xenopus extracts and immunoblotted the precipitates to determine if other passenger proteins were bound. After IP with xAurora B antibodies, both xINCENP and xSurvivin are also precipitated (Figure A). Because equivalent amounts of these three passenger proteins are detected in both interphase and mitotic extracts, it does not appear that their association is cell cycle regulated in the early embryo. Pre-I controls do not precipitate any of the passenger proteins. Moreover, neither xINCENP nor xSurvivin is immunoprecipitated by the xAurora B antibodies if recombinant xAurora B protein is added to the extract before the IP, demonstrating that the interaction is specific to xAurora B antibodies (unpublished data, M. Bolton and P.T. Stukenberg). In xINCENP IPs, both xSurvivin and xAurora B (Figure A) are detected by immunoblot. Finally, in xSurvivin IPs, both xAurora B and xINCENP are detected (Figure A). Therefore, in Xenopus embryos during both interphase and mitosis, there is a physical interaction between xSurvivin, xINCENP, and xAurora B kinase. Figure 4 | xSurvivin, xAurora B, and xINCENP are physically associated in vivo during both interphase and mitosis. xSurvivin, xAurora B, and xINCENP are physically associated in vivo during both interphase and mitosis. (A) xAurora B (xAurB), xINCENP, and xSurvivin were immunoprecipitated from interphase and mitotic Xenopus extracts. To determine if the chromosomal passenger proteins are physically associated, the immunoprecipitated samples were loaded onto 8 and 15% gels and were immunoblotted for xINCENP, xAurora B, and xSurvivin. Note that xINCENP is phosphorylated during mitosis and therefore exhibits retarded mobility. (B) In Xenopus extracts, the majority of xINCENP, xAurora B, and xSurvivin is physically associated. Interphase and mitotic extracts were immunodepleted with anti-xAurora B antibodies. Samples of the immunodepleted extract (xAurB Depleted Extract), beads, and interphase and mitotic whole cell extracts (WCE) were run on 8 and 15% gels and were immunoblotted for xINCENP, xAurora B, and xSurvivin. xINCENP and xSurvivin were depleted below detection levels (<1 ng) and were detected on the beads. Because the three proteins comigrate on a gel filtration column and sucrose gradient , it is possible that the majority of each passenger protein exists in this complex. This hypothesis was directly tested by immunodepleting xAurora B complex from a Xenopus low-speed extracts and immunoblotting the depleted extracts for xINCENP and xSurvivin. When 95% of the xAurora B is removed from extracts, xINCENP and xSurvivin are depleted to similar levels (Figure B). This experiment demonstrates that in the Xenopus embryo, most, if not all, of xINCENP and xSurvivin is physically associated with xAurora B kinase. Mapping the Chromosomal Passenger Complex Interactions | INCENP Is a Scaffold Protein with xSurvivin Binding Its N Terminus and xAurora B Binding Its C Terminus. An assay was developed to identify the domains on xINCENP that interact with xAurora B and xSurvivin (Figure A). 35S-labeled full-length xINCENP protein was translated in vitro and incubated in a Xenopus extract. This protein incorporates into the endogenous complex because it can be immunoprecipitated with our anti-xAurora B antibody (Figure B) and anti-xSurvivin antibody (unpublished data, M. Bolton, W. Lan, and P.T. Stukenberg). These interactions are specific, as they are not detected after IP with Pre-I controls (Figure B). The interaction between xAurora B and in vitro translated xINCENP is quite robust as the xINCENP is quantitatively immunoprecipitated from the extract (unpublished data, M. Bolton and P.T. Stukenberg), and, like the endogenous complex, the interaction is detected in both interphase and mitotic extracts (Figure , B and C). We determined whether xAurora B interacted with a series of deletion and truncated xINCENP proteins using this assay (Figure C). xAurora B interacts with all of the constructs that contain the C terminus of xINCENP except for the construct xINCENP Delta(119 --242). We do not understand why the xINCENP Delta(119 --242) construct does not interact with Aurora B, but it is likely that either the C terminus is misfolded or a regulatory region is missing. The simplest interpretation of the data is that there is an xAurora B interaction domain in the C-terminal 200 amino acids of xINCENP that is both necessary and sufficient for xAurora B binding. This region contains the IN-box that was previously shown to be an xAurora B binding domain in C. elegans and mouse . More recently, it was shown that yeast Aurora (Ipl1p) binds directly to the C terminus of the INCENP homolog Sli15 . Figure 5 | xAurora B, xINCENP and xSurvivin complex interactions. xAurora B, xINCENP and xSurvivin complex interactions. (A) Schematic of an assay used to study xAurora B, xINCENP, and xSurvivin interaction. Plasmids encoding full-length xINCENP as well as xINCENP deletion constructs (pictured in C) were in vitro translated in the presence of [35S]methionine and were subsequently incubated with Xenopus interphase and mitotic extracts. Anti-xAurora B antibodies, anti-xSurvivin antibodies, and Pre-I were used to immunoprecipitate the immunogenic protein and interacting xINCENP fragments from the extract. (B) An example of the assay where full-length xINCENP was immunoprecipitated with anti-xAurora B antibodies, but not with Pre-I. (C) The xINCENP constructs pictured were tested for their ability to coimmunoprecipitate with xAurora B and xSurvivin. Precipitation greater than fivefold above the amount detected by Pre-I controls is indicated by a + whereas less than fivefold was indicated by a -. Full-length xINCENP includes the following domains (N-terminal to C-terminal): the centromere-interacting domain, the chromosome-binding domain, the coiled-coil/microtubule-binding domain, and the IN-box (see Figure ). (D) The interaction of xSurvivin with the complex is NaCl sensitive. xAurora B was immunoprecipitated from a mitotic extract and was washed in XB buffer containing the indicated amount of additional NaCl. After washing, the concentration of xAurora B and xSurvivin in the IPs was determined by quantitative immunoblot. Densitometric analysis of this experiment indicates that after washing with 300 mM NaCl, there is 10-fold more xAurora B kinase than xSurvivin. (E) The N terminus of Aurora B is required to interact with xINCENP. Full-length xAurora B (circles) and xAurora B (99 --384) (triangles) were in vitro translated in the presence of [35S]methionine and were subsequently incubated with Xenopus interphase extracts. Anti-xINCENP antibodies, anti-xSurvivin antibodies, and Pre-I were used to immunoprecipitate the resulting complex from the extract, and the IPs were washed with XB + 0.1% Triton-X 100 and the indicated extra NaCl (mM). The presence of either full-length or kinase domain xAurora was detected by SDS-PAGE and subsequent phosphorimager analysis. Using this assay, we also mapped the xSurvivin interaction domain to the N-terminal 119 amino acids of xINCENP (Figure C). The three constructs that contained the first 119 amino acids of INCENP could be immunoprecipitated by xSurvivin antibodies, whereas the two constructs that lacked this region do not. Again, we see no cell cycle differences in the xINCENP/xSurvivin interaction. The N terminus of chicken INCENP has been implicated in centromere targeting as well as chromosome and midzone binding . We are currently testing if these biochemical activities are mediated by xSurvivin. xSurvivin Binding to the Complex Is NaCl Sensitive. The salt sensitivity of the interaction between the endogenous xSurvivin and xAurora B complex was examined. xAurora B IPs were washed with buffers containing 0.1% Triton-X 100 and varying NaCl concentrations and were subjected to quantitative Western blot (Figure D). The endogenous complex of xAurora B and xINCENP is stable in high salt. However, the interaction between xAurora B and xSurvivin is salt sensitive. There is twofold more xAurora B kinase in Xenopus extracts than xSurvivin, but after washing with only 100 mM NaCl, the ratio of the two proteins is 6:1. This ratio increases to 10:1 after a 300 mM NaCl wash, and no xSurvivin is detected after washing the complex with 1 M NaCl (Figure D). These data indicate that the interaction between xINCENP and xAurora B does not require xSurvivin because xAurora B and xINCENP remain tightly associated after removal of xSurvivin. Reciprocal IPs were performed by in vitro translating xAurora B, incubating the translation mix with Xenopus interphase extracts to allow labeled xAurora B to incorporate into the endogenous complex, and immunoprecipitating with either anti-xSurvivin or anti-xINCENP antibodies. The resulting IPs were washed with buffers containing 0.1% Triton-X 100 and varying NaCl concentrations. The amount of xAurora B bound was quantified by phosphorimager and plotted (Figure E). As seen in Figure D, the xAurora-xINCENP interaction is highly stable; however, the xAurora-xSurvivin interaction is NaCl sensitive. These data confirm two conclusions made from Figure D: the Survivin interaction is salt sensitive, suggesting that it is not tightly associated with the complex; and the interaction between xINCENP and xAurora B does not require xSurvivin. The N Terminus and Kinase Domain of xAurora B Both Interact with INCENP and the N Terminus May Weakly Bind xSurvivin. The first 98 amino acids of xAurora B define an N-terminal region of unknown function and the rest of the protein is mostly kinase domain. We demonstrate that the N terminus of xAurora B is required for the highly stable interaction with xINCENP (Figure E). A construct encoding the C-terminal amino acids 99 --384 of xAurora B was radioactively translated in vitro and was mixed with interphase Xenopus extracts. Unlike full-length xAurora B, which efficiently immunoprecipitates with xINCENP, the xAurora B kinase domain construct is not quantitatively precipitated. Whereas 70% of the full-length xAurora B coimmunoprecipitates with xINCENP, only 30% of the xAurora kinase domain coimmunoprecipitates. After washes with increasing concentrations of NaCl, the amount of xAurora kinase domain bound to xINCENP falls from 30 to 5%. The weak, salt-sensitive interaction between xAurora B kinase domain and xINCENP suggests that the N terminus of xAurora B provides the majority of the binding to xINCENP, whereas the kinase domain only weakly interacts. It has been reported that xAurora B kinase interacts with both xINCENP and human Survivin in both two-hybrid and in vitro pull-down assays . Therefore, it is surprising that we have identified xINCENP constructs that robustly bind either xSurvivin or xAurora B, suggesting that xAurora and xSurvivin do not interact directly but rather bind opposite ends of xINCENP. If this were true, we would expect that anti-xSurvivin antibodies would pull down xAurora B kinase domain through its interaction with xINCENP (Figure E). However, we find that xSurvivin antibodies were not able to precipitate any xAurora B kinase domain, even though at low NaCl concentrations, there was some interaction between the kinase domain and xINCENP. The simplest interpretation of this data is that there may be a weak direct interaction between the N terminus of xAurora B and xSurvivin that stabilizes the complex. xSurvivin Protein Can Stimulate Mitotic xAurora B Kinase Activity | Because xSurvivin could be washed off xAurora B kinase IPs, the dependence of xAurora B kinase activity on complex formation was examined. xAurora B kinase was immunoprecipitated from interphase and mitotic extracts, washed with a buffer containing 300 mM NaCl so that the ratio of xAurora B to xSurvivin was 10:1, and the kinase activity was measured by the incorporation of 32PO4 from gamma[32P]ATP onto a MBP substrate (Figure A). Kinase activity is readily detected when the IP is performed with Aurora B antibodies, but not if Pre-I is used. In correlation with previous studies, we find that this activity is cell cycle regulated, as at least 10-fold greater activity is detected if the xAurora B is precipitated from mitotic extracts as compared with interphase extracts . The kinase stimulation is due to an increase in specific activity of the enzyme as similar amounts of protein are detected in the IP by Coomassie stain of the gel (Figure , A and C) or xAurora B immunoblot (Figure B). Figure 6 | The activity of xAurora B is stimulated by stoichiometric xSurvivin binding and mitotic phosphorylation. The activity of xAurora B is stimulated by stoichiometric xSurvivin binding and mitotic phosphorylation. (A) Recombinant GST-Survivin can stimulate an xAurora B IP kinase assay. xAurora B IP kinase assay from either interphase or mitotic extracts was conducted in the presence or absence of 10 ng of recombinant GST-Survivin. The top panel is an autoradiogram of [32P]O4 incorporation into a MBP substrate; the bottom panel is a Coomassie-stained gel showing similar amounts of xAurora B in each reaction. (B) Titration of the amount of xSurvivin needed to activate the xAurora B kinase immunoprecipitated from interphase and mitotic extracts. The top panel is a graph of the kinase activity of the immunoprecipitated xAurora B in the presence of the indicated amount of GST-Survivin. The bottom panel is an xAurora B immunoblot analysis demonstrating similar amounts of xAurora B kinase in the xAurora B IP kinase assay. (C) xAurora B kinase was immunopurified from either interphase or mitotic Xenopus egg extracts; Pre-I was used as a negative control. The IPs were washed extensively and either incubated in phosphatase buffer (-) or phosphatase buffer and lambda phosphatase (+), and were then washed again extensively and the kinase activity was assayed on MBP. A Coomassie-stained gel of the IP kinase assay demonstrates that there are similar amounts of xAurora B kinase in each reaction (bottom panel). (D) The phosphorylation state of xAurora B directly regulates its kinase activity. The xAurora B IPs were washed extensively, incubated in either phosphatase buffer (-) or phosphatase buffer and lambda phosphatase (+), washed again, and assayed for kinase activity on MBP substrate. Samples were also immunoblotted for xINCENP, xAurora B, and xSurvivin to determine relative amounts of the proteins. We find that adding 10 ng of recombinant GST-Survivin stimulates the mitotic kinase activity similareightfold, but has limited affect on xAurora B isolated from interphase extracts (Figure A). This stimulation of xAurora B kinase activity was titrated in Figure B and was saturated at 10 ng of recombinant GST-Survivin. As a negative control, we added 10 ng of GST-cyclin B, which did not stimulate kinase activity (unpublished data, S. Powers, P.T. Stukenberg). We estimate that there is similar20 --30 ng of xAurora B in these IPs (Figure D), suggesting that xSurvivin is stoichiometrically required for xAurora B kinase activity. xAurora B Kinase Activity Is also Regulated by Phosphorylation | The kinase activity of Aurora A is regulated by phosphorylation . To determine if Aurora B kinases are also regulated by phosphorylation, xAurora B IPs from interphase and mitotic extracts were washed with low-salt buffer, which allows most of the endogenous xSurvivin to remain bound. This reaction was split, and one-half of the reaction was treated with lambda phosphatase, subsequently washed to remove the phosphatase, and assayed for kinase activity (Figure C). The mitotically stimulated xAurora B kinase activity is sensitive to treatment with lambda phosphatase, demonstrating that cell cycle-specific phosphorylation regulates the specific activity of xAurora B (Figure C; ). The simplest interpretation of the phosphatase sensitivity of Aurora B kinase activity is that, as in many other kinases, the phosphorylation state directly regulates kinase activity. However, dephosphorylation may also disturb the xSurvivin interaction. To distinguish between these two possibilities, we phosphatase treated an IP of xAurora B and assayed for both kinase activity and for the relative amount of the three subunits by immunoblot (Figure D). Phosphatase treatment eliminated both the kinase activity and greatly increased the gel mobility of xINCENP, demonstrating that the phosphatase treatment had worked efficiently. However, the concentration of each subunit in the complex did not change. Therefore, maximum kinase activity of xAurora B requires both interaction with xSurvivin and the mitotic phosphorylation state of the complex. DISCUSSION : During both interphase and mitosis, xAurora B kinase is in a complex with the majority of xINCENP and xSurvivin protein in the Xenopus embryo. The complex is cell cycle regulated as the kinase is activated >10-fold in mitosis, and hydrodynamic characterization suggests that there is a dramatic change in size that corresponds to changes in the cell cycle. We have determined two independent steps for mitotic xAurora B kinase activation: stoichiometric xSurvivin binding and a phosphorylation requirement. Therefore, like the CDK1 mitotic kinase, xAurora B activation can be regulated by both its phosphorylation state and the binding of a regulatory subunit. Our demonstration that xSurvivin, xINCENP, and xAurora B kinase are in a complex can simply explain the molecular details underpinning a number of in vivo observations. First, in C. elegans double-stranded RNA-mediated interference (RNAi) of the INCENP homolog can mislocalize the Aurora B kinase and loss of either protein causes both chromosome segregation and cytokinetic defects . Second, after RNAi of the Survivin homolog in C. elegans, the Aurora B homolog is mislocalized and embryos lacking either protein have similar phenotypes . Our data suggest that these phenotypes could either be caused by mislocalization of the proteins when the complex is not properly formed or by inactive Aurora B kinase in the absence of the Survivin protein. The N terminus of human INCENP (1 --405) has dominant-negative activity and can inhibit the final stages of cytokinesis when it is overexpressed in tissue culture cells . Our mapping of xSurvivin binding to the N terminus of INCENP and xAurora B binding to the C terminus suggests that overexpression of INCENP (1 --405) would sequester Survivin away from Aurora B and inhibit its kinase activity. A Complex of Chromosomal Passenger Proteins in the Xenopus Early Embryo | An in vivo passenger protein complex of xSurvivin, xINCENP, and xAurora B kinase has been identified in Xenopus early embryos. Our mapping data are consistent with a scaffold protein role for xINCENP, which binds xSurvivin on its N terminus and xAurora B on its C terminus (Figure A). The N terminus of INCENP has been implicated in centromere targeting in metaphase and midzone localization in anaphase . Our localization of the Survivin interaction to this region suggests that these targeting events could be mediated through the Survivin protein. The C terminus of INCENP contains the IN-box, which appears to be a highly conserved Aurora B interaction motif as it also interacts with Aurora B homologs in C. elegans, mouse, and budding yeast . We have shown that both the kinase domain of xAurora B and the N terminus provide interactions with the C terminus of xINCENP. We have also shown that the N terminus of Aurora B stabilizes xSurvivin binding in high-salt conditions. The simplest interpretation of this finding is that the major interaction between xSurvivin and xAurora is mediated through xINCENP, although there is a weak direct interaction between xSurvivin and the N terminus of xAurora B (Figure B). Figure 7 | Model of xAurora B, xINCENP, and xSurvivin interactions and regulation. Model of xAurora B, xINCENP, and xSurvivin interactions and regulation. (A) Interactions within the passenger protein complex. xSurvivin binds the N terminus of xINCENP, whereas the N terminus of xAurora B binds the C terminus of xINCENP. (B) Cell cycle regulation of the passenger protein complex. The interphase complex is hypophosphorylated and its migration in a sucrose gradient is consistent with either additional subunits (X) or a dimerization of the complex. A weak interaction between xSurvivin and xAurora may also exist. In mitosis, all three proteins are both present and phosphorylated, but the complex is smaller. Both the mitotic phosphorylation of the complex and the presence of the xSurvivin subunit are required for high kinase activity (On). (C) We present a highly speculative model in which the interaction between xSurvivin and xAurora is regulated within mitosis to control kinase activity. We propose that during mitosis, xAurora B kinase activity is inactivated by disrupting the interaction of xAurora B with xSurvivin (Off), perhaps through a mechanical force. It is also possible that Survivin binding to the complex could be spatially controlled. Such a mechanism could permit spatial regulation within parts of the same mitotic cell, allowing for pools of active Aurora kinase with Survivin-bound (On) versus inactive Aurora kinase with no Survivin interaction (Off). It has been reported that Survivin and Aurora B interact directly in a two-hybrid assay and that in an in vitro pull-down assay, the complex is stable in treatments as harsh as 3 M NaCl . Perhaps high concentrations of interacting proteins in the two-hybrid and pull-down systems uncovered the weak interaction between N terminus of Aurora B and Survivin. Alternatively, the difference between the two results may be a problem with Aurora B folding in overexpression systems. We have found that soluble recombinant Aurora B can be purified from E. coli; however, protein from these preparations migrates in the void fractions of subsequent gel filtration columns (unpublished data, W. Lan, P.T. Stukenberg). We feel that it is likely that recombinant Aurora B purified from E. coli is at least partially denatured and therefore the high NaCl sensitivity in the pull-down assay may be an artifact of an in vitro experiment. Because of this problem with recombinant protein, we have been forced to design all activity and interaction experiments with endogenous xAurora B. For further biochemical examination of the passenger protein complex, it will be necessary to develop overexpression systems that can generate the complex in its native folded state. Although a physical interaction between xAurora B, xINCENP, and xSurvivin is detected during both interphase and mitosis, the hydrodynamic properties of the complex change dramatically during the cell cycle. Specifically, the peak off a gel filtration column is much broader in interphase than in mitosis, but peaks in the same fractions (migrating at 107 A). Also, our sucrose density gradient data indicate that the interphase complex sediments significantly faster than the mitotic complex (9.5S and 4.5S, respectively). The simplest model to explain such hydrodynamic characteristics is that the active complex in mitosis is a highly elongated molecule that contains at least one molecule of xINCENP, xSurvivin, and xAurora B; in interphase, two of these complexes dimerize to double the molecular weight (Figure B). It is equally possible that there are unknown subunits that are bound in interphase. There is a 2:1 M ratio of xAurora B:xSurvivin in extracts, and all of the xSurvivin protein is complexed with xAurora B. This finding was surprising because Survivin has been shown to have two functions: one as a mitotic spindle regulator and another as a chromosome passenger protein required for chromosome segregation and cytokinesis. Because Aurora B kinase has not been localized to the metaphase mitotic spindle, we expected xSurvivin to exist in at least two complexes. It is likely that the mitotic spindle requirement of Survivin is not required in the early embryo, but becomes essential during the mitosis of somatic cells. Interestingly, the metaphase spindle/centrosome localization of Survivin has been implicated to regulate apoptosis, a process that is not present in the Xenopus early embryo until after the midblastula transition (12th division). xAurora B Kinase Activity Is Regulated by xSurvivin Binding and Phosphorylation | Mitotic kinases are highly regulated to ensure that their activity is restricted to the proper time and location of the cell cycle. For example, both the activity and localization of CDK1 kinase is regulated by binding of a cyclin subunit . CDK1 kinase activity is also both positively and negatively regulated by phosphorylation . Our goal in beginning this study was to describe in molecular detail the mechanisms of Aurora B kinase activation in mitosis. To this end, we have developed a robust IP kinase assay in Xenopus extracts to measure xAurora B kinase activity and found that the kinase is stimulated >10-fold in mitosis. Using this assay, we have made two significant contributions to our understanding of xAurora B kinase regulation. First, we have demonstrated that xAurora B kinase activity is stimulated by xSurvivin binding. Second, we have confirmed that the phosphorylation state of the mitotic complex is critical for the stimulation of kinase activity (Figure B; ). Survivin binding could increase Aurora kinase activity by two mechanisms. It could either change the substrate recognition of xAurora kinase or it could increase the rate of catalysis. Our data cannot distinguish between these two models. During the preparation of this manuscript, Chan and colleagues published that the in vitro activity of the budding yeast Aurora (Ipl1p) is stimulated 10- to 20-fold by direct binding of the INCENP homolog (Sli15p). Thus, it is either possible that Aurora B activity requires binding by both INCENP and Survivin or that Sli15p has acquired Survivin's activation role. We have not detected stimulation of Aurora kinase activity by the addition of the C-terminal fragment of xINCENP to our Aurora B IPs. However, this is not surprising as xINCENP, unlike Survivin, is not washed off in the salt washes. Unfortunately, recombinant xINCENP and xAurora B from both E. coli and baculovirus appear highly misfolded after purification, making direct activation experiments in the Xenopus system impossible at this time. We have also shown that activation of the xAurora B kinase depends upon mitotic phosphorylation. This corroborates a recent study of Xenopus Aurora B kinase and is consistent with previous work that suggests that Aurora A kinase is regulated by phosphorylation . Currently, it is unknown if the activating mitotic phosphorylation is on the xAurora B kinase or an interacting protein. Recent data from our laboratory indicate that all three proteins are phosphorylated in mitosis, and we have previously demonstrated that the xINCENP protein is specifically hyperphosphorylated in mitosis . We have extended this knowledge by mapping 26 cell cycle-specific phosphorylation sites on Xenopus INCENP (F. White, D. Hunt, W. Lan, and P.T. Stukenberg, unpublished data). Survivin has been shown previously to be a substrate of CDK1 in mitosis , and if phosphatase inhibitors are used, we often see multiple bands in Survivin blots consistent with heterogeneous phosphorylation . We have mapped seven cell cycle-specific phosphorylation sites on the xAurora B protein that are currently being characterized (F. White, D. Hunt, W. Lan, and P.T. Stukenberg, unpublished data). Therefore, the phosphorylation-dependent stimulation of Aurora B activity is likely to be very complex. Why does Survivin binding regulate kinase activity? The simplest model is that the kinase is not active until the entire complex is properly assembled. The most provocative interpretation of our data is a highly speculative model that we propose in Figure C. In this model, the interaction between Survivin and Aurora is regulated within mitosis to control kinase activity. One way to regulate the interaction of Survivin with Aurora could be through a mechanical force, which could separate Aurora and Survivin in the same complex, or Survivin could simply dissociate from INCENP. Such a mechanism could permit spatial regulation within parts of the same mitotic cell, allowing for pools of active Aurora kinase with Survivin bound versus inactive Aurora kinase with no Survivin interaction. However, we have yet to find conditions where xSurvivin binding to xAurora B is used as a regulatory mechanism in the Xenopus early embryo. Similar amounts of xSurvivin are bound to xAurora B in both interphase and mitotic extracts . We have also immunoblotted Xenopus embryos traversing the early embryonic cell cycles and we do not detect a gross difference of xSurvivin levels (M. Bolton, P.T. Stukenberg, unpublished data). However, the fact that xSurvivin is stable in interphase is probably a unique feature of embryonic systems, as the protein levels peak during mitosis in somatic cells (; Figure ). Backmatter: Abbreviations used: : EST = expressed sequence tag GST = glutathione S-transferase I = interphase INCENP = inner centromere protein IP = immunoprecipitation MBP = myelin basic protein PBS = phosphate-buffered saline PBS-T = PBS containing 0.1% Tween 20 PCR = polymerase chain reaction Pre-I = preimmune sera xAurora B and xAurB = Xenopus Aurora B xINCENP = Xenopus inner centromere protein xSurvivin = Xenopus Survivin PMID- 12221117 TI - ADP-Ribosylation Factor (ARF) Interaction Is Not Sufficient for Yeast GGA Protein Function or Localization AB - Golgi-localized gamma-ear homology domain, ADP-ribosylation factor (ARF)-binding proteins (GGAs) facilitate distinct steps of post-Golgi traffic. Human and yeast GGA proteins are only similar25% identical, but all GGA proteins have four similar domains based on function and sequence homology. GGA proteins are most conserved in the region that interacts with ARF proteins. To analyze the role of ARF in GGA protein localization and function, we performed mutational analyses of both human and yeast GGAs. To our surprise, yeast and human GGAs differ in their requirement for ARF interaction. We describe a point mutation in both yeast and mammalian GGA proteins that eliminates binding to ARFs. In mammalian cells, this mutation disrupts the localization of human GGA proteins. Yeast Gga function was studied using an assay for carboxypeptidase Y missorting and synthetic temperature-sensitive lethality between GGAs and VPS27. Based on these assays, we conclude that non-Arf-binding yeast Gga mutants can function normally in membrane trafficking. Using green fluorescent protein-tagged Gga1p, we show that Arf interaction is not required for Gga localization to the Golgi. Truncation analysis of Gga1p and Gga2p suggests that the N-terminal VHS domain and C-terminal hinge and ear domains play significant roles in yeast Gga protein localization and function. Together, our data suggest that yeast Gga proteins function to assemble a protein complex at the late Golgi to initiate proper sorting and transport of specific cargo. Whereas mammalian GGAs must interact with ARF to localize to and function at the Golgi, interaction between yeast Ggas and Arf plays a minor role in Gga localization and function. Keywords: INTRODUCTION : The Golgi-localized, gamma-ear homology domain, ARF-binding protein (GGA) family of proteins facilitates the sorting and transport of proteins out of the trans-Golgi network (for review, see ). GGA proteins interact directly with ADP-ribosylation factor (ARF) proteins, a family of 21-kDa GTP-binding proteins that have been implicated as regulators of membrane traffic at many steps in the secretory, endocytic, and recycling pathways (,; ; ; ). GGA proteins interact with the activated, GTP-bound form of ARF, establishing them as effectors of ARF signaling . Three GGA genes have been described in mammalian cells: GGA1, GGA2, and GGA3. Alternative splicing of GGA3 results in a long form and a short form of the GGA3 protein. The yeast Saccharomyces cerevisiae has two GGA genes, GGA1 and GGA2. The expressed yeast proteins, Gga1p and Gga2p, share 50% amino acid identity. The human and yeast GGA proteins are similar25% identical. Although the sequence identity is low, the domain structure of GGA proteins is conserved between species. Four domains are apparent when the amino acid sequences of yeast and human GGA proteins are aligned (see Figure A). An N-terminal VHS domain (150 residues) resembles the VHS domain present in Vps27p, HRS, and STAM proteins . The 170-residue GAT domain is the most highly conserved (65% identity between human GGA proteins) and contains two predicted coiled-coil regions. The GAT domain was so named for its sequence homology to a protein named Tom1 (GGA and Tom1). A "hinge" region of variable length contains one or more clathrin-binding domains but no other significant homology to each other or other known proteins. A C-terminal ear domain (120 amino acids; 40% identity between human GGA proteins) is homologous to the ear domain of gamma-adaptin. Figure 1 | ARF-binding region of GGA proteins and effects of mutations. ARF-binding region of GGA proteins and effects of mutations. (A) Comparison of ARF-binding domains from yeast and human GGA proteins. GGA proteins have four domains, with the highest identity between yeast and human GGA proteins in the ARF-binding domain. Residues identical in at least three proteins are highlighted in red. Dashes indicate gaps in the alignment. The numbers indicate the first and last residues shown for each GGA protein. Asterisks indicate mutated residues. (B) Loss of ARF binding by two-hybrid analysis. Gal4AD fusions of wild-type and mutant GGA proteins were assayed for binding to Gal4BD fusions of activated ARF in the two-hybrid assay. Serial 10-fold dilutions of each strain were replica plated on SD plates lacking tryptophan and leucine (-trp-leu) and SD plates lacking histidine, tryptophan, and leucine with 25 mM 3AT. Growth on -trp-leu shows that all strains are viable. Interaction is positive when colonies grow on 3AT and turn blue in X-gal assays. Scores in the right-hand column are defined in MATERIALS AND METHODS. (C) Loss of Arf binding by using affinity chromatography. Increasing concentrations of Arf1pGDP or Arf1pGTPgammaS (as indicated) were incubated with GST fusion proteins immobilized on glutathione agarose beads: GST-Gga2p (lanes 1 --6), GST-Gga2pI207N (lanes 7 --12), and GST only (lanes 13 --14). Retained proteins were eluted and analyzed by Coomassie staining to visualize the GST fusion proteins (bottom) or by immunoblotting to detect Arf1p (top). Arf1p was retained by GST-Gga2p, but not by mutant GST-Gga2pI207N or GST only. Work in both mammalian cells and S. cerevisiae suggests that GGA proteins function in trafficking at the TGN. In yeast, deletion of either gene alone causes no (GGA1 deletion) or minor (GGA2 deletion) defects, whereas deletion of both genes disrupts distinct post-Golgi trafficking events . Several proteins that are transported between the TGN and early or late endosomes, including carboxypeptidase Y (CPY), carboxy peptidase S, Pep12p, and Kex2p, are at least partially missorted in cells lacking both GGA1 and GGA2. For example, similar40% of newly synthesized CPY is secreted in gga1gga2 cells . These data support a role for Gga proteins in a TGN-to-early endosome pathway, in a TGN-to-late endosome pathway, or both. The function of mammalian GGA proteins has been deduced based on the protein interactions of each domain. Mammalian GGA proteins interact directly with a defined subset of cargo that trafficks between the TGN and lysosomes. Sortilin, LRP3, and both the cation-independent and cation-dependent forms of mannose 6-phosphate receptor (M6PR) contain an acidic cluster-dileucine motif that interacts with the VHS domain of human GGA proteins . Many other transmembrane proteins that lack this motif do not interact with GGA proteins, suggesting that GGAs are specific for a subset of transported proteins. It is not known whether a similar motif (acidic cluster-dileucine) is recognized by yeast Gga proteins, or even whether cargo is sorted via direct interaction with Gga proteins in yeast. The hinge domain of mammalian and yeast GGA proteins interacts directly with clathrin in vitro. The ear domain strengthens this interaction, and the ear domains of certain GGA proteins also interact with clathrin on their own . These findings suggest that GGA proteins can recruit clathrin to sites of vesicle budding on the TGN. Overexpression in cultured cells of a GGA3 construct lacking the hinge and ear domains causes M6PR to accumulate in the TGN, supporting a role for GGA-dependent clathrin recruitment in the packaging of these cargo into vesicles . Thus, it is likely that both yeast and mammalian GGA proteins are monomeric clathrin adaptors that function at the TGN. The ear domain of human GGA proteins interacts with gamma-synergin , a Golgi-localized protein partner of gamma-adaptin that has an unknown function . Thus, the function of the GGA ear was accurately predicted by its homology to gamma-adaptin. However, gamma-synergin has no obvious homolog in yeast, and the ear domains of yeast and human GGA proteins are not well conserved; hence, the role of the ear domain in yeast remains unclear. In mammalian cells, endogenous GGA1, GGA2, and GGA3 localize predominantly to the trans-Golgi region . Three lines of evidence show that TGN localization of GGAs requires interaction between the GAT domain and activated ARF proteins. First, treatment with brefeldin A causes rapid translocation of GGA proteins to the cytosol in a time frame indistinguishable from that of ARF itself . Second, the interaction between the GAT domain and ARF is strong enough to direct a green fluorescent protein (GFP) fusion of GAT onto the Golgi, even in the absence of the VHS, hinge, and ear domains . Third, and most convincingly, point mutations within the ARF-binding domain that abolish ARF interaction also cause loss of Golgi localization . Together, these data suggest that GTP-bound ARF recruits mammalian GGA proteins from the cytosol onto the late Golgi membrane by interacting with the GAT domain. The function of GGA proteins in mammalian cells is also intimately tied to activated ARF. First, expression of the constitutively active ARF1Q71L in mammalian cells causes dramatic expansion of the Golgi apparatus . Overexpression of GGA1 or GGA3 in these same cells prevents Golgi expansion . This alteration of ARF1 function indicates a functional interaction between ARF1 and GGAs in vivo. Second, as described above, mutations within the GAT domain of GGA3 disrupt its Golgi localization. These same mutations prevent GGA3 overexpression from displacing adaptor complexes or fragmenting the Golgi apparatus , which usually results from GGA overexpression in mammalian cells. Although these mutations in GGA3 might disrupt other interactions, current evidence supports the model that loss of ARF interaction disrupts GGA function in mammalian cells. Together, these data suggest that GGA proteins are ARF-dependent, monomeric clathrin adaptors that facilitate the sorting of specific cargo at the TGN into vesicles destined for endosomes. Much of the data comparing yeast and mammalian GGA function assumes that the mechanisms of GGA function are well conserved between these organisms. However, the mechanism of yeast GGA function has not been analyzed. In particular, it is not known whether Gga function or localization in yeast depends on its binding to Arf (via the GAT domain) or other proteins (via the VHS and/or hinge and ear domains). In this article, we show that yeast Gga1p and Gga2p do not require Arf for their localization or function, although in vivo binding to Arf is detectable. We propose that yeast Ggas are targeted to the Golgi membrane through interactions with Golgi-localized proteins other than Arf, which then stabilize the Gga --Arf interaction. MATERIALS AND METHODS : Materials | Unless otherwise specified, chemicals and reagents were purchased from Sigma-Aldrich (St. Louis, MO) or Fisher Scientific (Fair Lawn, NJ). Yeast and bacterial media reagents were purchased from Difco (Detroit, MI). Restriction and other enzymes were purchased from Promega (Madison, WI) or Invitrogen (Carlsbad, CA). TA cloning vector pGEMT-easy was purchased from Promega. Unique Site Elimination Mutagenesis kit was purchased from (Palo Alto, CA). Yeast Methods | Yeast were grown under standard conditions at 30C unless otherwise noted. Transformations into yeast were performed using the LiOAc method, with herring sperm DNA as a carrier . Yeast Strains | Yeast strains used in this study are listed in Table . Diploid strains YAB538 and YAB539 were generated by mating YAB531 (gga1::TRP1) and YAB532 (gga2::HIS3) . Tetrad dissections generated several gga1gga2 strains used for trafficking assays and generation of triple knockout strains (YAB538 T6a, YAB538 T2d, YAB539 T3d; ). Strain YAB650 (gga1gga2apm1) was generated by mating and tetrad dissections of YAB539 T3d and Piper1878. Strain YAB658 (gga1gga2apm3) was generated by mating and tetrad dissections of YAB538 T2d and Piper1022. Strain YAB667 (gga1gga2vps1) was generated by transforming plasmid pCKR3A, digested with SacI and XbaI, into strain YAB538 T6a and selecting for LEU auxotrophy. Strains YAB679 and YAB680 (gga1gga2vps27 in SEY6210 background) were generated by mating and tetrad dissections of ABC114 T6b and vps27 strain JUY68. Strain YAB677 (gga1gga2vps27 in S288C background) was generated by transforming plasmid pKJH2, digested with BamHI and PstI, into strain YAB538 T6a and selecting for LEU auxotrophy. Gene deletions were confirmed by Western blotting for Gga1p and Gga2p, and by phenotype for VPS27 (class E compartment), APM3 (ALP processing), and VPS1 (CPY secretion). Multiple isolates of gga1gga2apm1 were analyzed. All other strains were generated by transformation. Plasmid Construction | Plasmids used in this study are listed in Table . Oligonucleotide primers used for polymerase chain reaction (PCR) amplification are listed in Table . All constructs generated by PCR amplification were verified by automated sequencing (Advanced Genetic Analysis Center, University of Minnesota). Table 1 | Plasmids used in this study Two-Hybrid Plasmids. | Gal4 binding domain (Gal4BD) fusions of yeast and human ARFs were generated from plasmid pBG4D (a kind gift from Rob Brazas, University of California, San Francisco, CA), as described previously . Gal4 activation domain (Gal4AD) fusions of yeast and human GGA proteins were generated from plasmid pACT2 (a kind gift from Steve Elledge, Baylor University, Waco, TX), as described previously . A single point mutation within the Arf-binding domain of yeast Gga1p was fortuitously identified after PCR amplification of the GGA1 VHS/GAT domain and subcloning into the two-hybrid vector pACT2, resulting in plasmid pAB362. This single-base substitution changed leucine at position 203 to a glutamine (L203Q). Homologous mutations in yeast Gga2p (I207N), human GGA1 (L182Q), and human GGA3 (L149Q) were generated by two-step PCR mutagenesis with oligos specified in Table . The PCR products were generated with BamHI and XhoI sites at the 5' and 3' ends of the open reading frames, respectively, and subcloned into vector pACT2 to generate Gal4AD fusion proteins of mutant Gga2p (pAB462), GGA1 (pAB418), and GGA3 (pAB503). Alanine substitutions in Gga2p (PEDL->AAAL and ANKL->AAAL) were generated using Unique Site Elimination mutagenesis in plasmid pAB473 (see below). The open reading frames were then subcloned into pACT2 by using NcoI and XhoI, resulting in plasmids pAB580 and pAB581. Yeast Expression Vectors. | Plasmids for expression of Gga1p (pAB470) and Gga2p pAB441) were described in . An NcoI site was generated at the start ATG of GGA2 in pAB441 by using two-step PCR to generate plasmid pAB471. The GGA2(NcoI) gene was then subcloned into vector pRS315 by using BamHI and XhoI to generate pAB473. Unique site elimination mutagenesis of pAB473 was used to generate PEDL->AAAL (pAB493) and ANKL->AAAL (pAB509). The mutated BamHI-XhoI fragments of pAB493 and pAB509 were then subcloned into pRS316 to generate pAB529 and pAB530, respectively. The mutated NcoI-XhoI fragments of pAB493 and pAB509 were also subcloned into pACT2, as described above. For expression of Gga1pL203Q, the mutated region from pAB362 was subcloned into pAB470, resulting in pAB469. For expression of Gga2pI207N, two-step PCR mutagenesis was used to generate the I207N mutation from plasmid pAB441. The PCR product was then subcloned into pRS316, resulting in pAB456. The amino terminal truncation of Gga2p(110 --585) was generated by PCR (oligos A053 and A038) followed by subcloning the NcoI-XhoI --digested products into pAB473. Then the NotI-XhoI cassette was subcloned into pRS316 to generate pAB555. The carboxy-terminal truncation of Gga2p(1 --326) was generated by PCR (oligos A037 and A049) and subcloned with BamHI and XhoI into pRS316, resulting in pAB468. Hemagglutinin (HA)-tagged GGA Proteins. | Unique site elimination was used to generate an NheI site immediately preceding the stop codon of GGA1 and GGA2, resulting in plasmids pAB487 and pAB483. The XbaI fragment from pMR2654 (a kind gift from Mark Rose, Princeton University, Princeton, NJ) containing a triple HA repeat was inserted into the NheI-digested plasmids, resulting in plasmids pAB492 and pAB491. Orientation of the HA tag was determined by sequencing. Mutant alleles were generated by subcloning the mutated portion of the open reading frames into the HA-tagged plasmids, resulting in plasmids pAB579 (GGA1) and pAB578 (GGA2). Glutathione S-Transferase (GST) Fusion Proteins. | All constructs were subcloned from pACT2 (described above) into pGEX5X-2 (Stratagene, La Jolla, CA) by using BamHI and XhoI sites. The following constructs were used for these studies: GST-Gga1p1 --331 (pAB391), GST-Gga1p1 --331, L203Q (pAB476), GST-Gga2p1 --326 (pAB382), and GST-Gga2p1 --326, I207N (pAB472). GFP Fusion Proteins. | Plasmids for expression of GFP-Gga1p (pCS135) and GFP-Gga2p (pCS136) were obtained from Chris Stefan and Scott Emr (University of California, San Diego, CA). Both are pRS426-based plasmids with the CPY promoter driving expression of the GFP open reading frame. A multiple cloning site follows GFP. Plasmid pAB505 for expression of GFP-Gga1pL203Q was generated by subcloning the KpnI-KpnI fragment of pAB469 into pCS135. Plasmid pAB504 for expression of GFP-Gga2pI207N was generated by PCR amplification of the open reading frame from pAB456 to incorporate an SalI site at the 5' end then subcloning the SalI-KpnI fragment into pCS136. The truncated forms of GFP-Gga1p were all generated by PCR amplification from either pCS135 or pAB505 then subcloned into pCS135 by using SalI-KpnI or KpnI-KpnI sites. Mammalian Expression Vectors. | Plasmids for expression of human GGA proteins in normal rat kidney (NRK) cells were generated from pcDNA3 as described in . GGA1L182Q was generated by two-step PCR mutagenesis with mutant oligos A025 and A026 and flanking oligos 797 and A028 then subcloned into pcDNA3-HA with EcoRI and XbaI, resulting in pAB443. GGA3L149Q was generated by subcloning the mutated region of GGA3 (HindIII-XhoI) from pAB503 into vector pAB366 (pcDNA3-HA-GGA3; ), resulting in pAB506. Truncated GGA1DeltaVHS (pAB318) begins at residue 145, as described previously . GGA1DeltaVHS,L182Q (pAB582) was generated by subcloning the BamHI-XhoI fragment of pAB418 into pcDNA3. An ATG immediately following the BamHI site acts as the start ATG. Two-Hybrid Assays | Filter assays for beta-galactosidase activity were performed as described previously with 1 mg/ml X-gal as the substrate. Colony color was scored every hour for 3 h then allowed to develop overnight. Strong positives (+++) turned blue within 1 h, moderate positives (++) turned blue within 3 h, weak positives (+) turned blue overnight, and negatives (-) remained white throughout the development period. Assays for histidine auxotrophy were performed by replica plating yeast strains on SD-trp-leu plates and 3AT plates (SD plates lacking tryptophan, leucine, and histidine, and containing 25 mM 3-amino triazole) and scoring growth after 3 --4 d. GST-Affinity Chromatography | Bacterially expressed Arf1pQ71L was purified as described previously . Bacterially expressed GST fusion proteins were purified on glutathione agarose beads as described previously . The beads were washed three times with TND (25 mM Tris pH 7.5, 100 mM NaCl, and 1 mM dithiothreitol) and once with binding buffer (20 mM HEPES pH 7.5, 1 mM EDTA, 100 mM NaCl, 0.5 mM MgCl2, 1 mM dithiothreitol, 50 mug/ml bovine serum albumin, and 1% Triton); 30 mul of bead volume was used for each binding reaction. Bacterially expressed yeast Arf1pQ71L (6 muM) was incubated for 30 min at 30C in binding buffer with 10 muM guanosine-5'-O-(3-thio)triphosphate (GTPgammaS) or GDP. GST fusions were mixed with 100 mul of 0.6, 1.8, or 6 muM Arf1p for 30 min. The beads were then washed three times with 1 ml of binding buffer containing 10 muM nucleotide. Bound proteins were eluted from the beads with an equal volume of 2x Laemmli sample buffer and run on a 12% SDS-PAGE gel. The final wash was also analyzed to ensure complete washing. Duplicate gels were either stained with Coomassie brilliant blue or developed by immunoblotting, by using a polyclonal rabbit antiserum (R23; ) to detect yeast Arf1p. Antibodies and Immunoblotting | Immunoblotting was performed as described previously . Antibodies were used at the following dilutions, as indicated for each experiment: 12CA5 (to HA epitope; BabCo, Berkeley, CA), 1:10,000; R89675 (to Gga1p), 1:10,000; R89013 (to Gga2p), 1:10,000; R23 (to Arf1p, a gift from R. Kahn, Emory University, Atlanta, GA), 1:2000; 10A5-B5 (to yeast CPY; Molecular Probes, Eugene, OR), 1:5000; and anti-ALP (a gift from S. Nothwehr, University of Missouri, Columbia, MO), 1:5000. Bound antibodies were detected with horseradish peroxidase-linked secondary antibodies and enhanced chemiluminescence detection reagents (Amersham Biosciences, Piscataway, NJ). Films were scanned using a flatbed scanner and figures were prepared using Adobe PhotoShop 5.0 (Adobe Systems, Mountain View, CA). Affinity Purification of Antibodies to Gga1p and Gga2p. | For Gga1p, antibodies were adsorbed onto a nitrocellulose strip to which GST-Gga1p was transferred. After washing, bound antibodies were stripped with 0.2 M glycine, pH 2.8. For Gga2p, antibodies were mixed with GST-Gga2p bound to glutathione agarose, washed, and eluted with 0.2 M glycine. The efficacy of the affinity purification was analyzed by immunoblotting (our unpublished data). Yeast Fractionation | Yeast cells (0.15 g) were suspended in 100 mul of fractionation buffer (0.4 M sucrose, 25 mM KPO4 pH 7.0, and 2 mM EDTA) and 5 mul of protease inhibitor cocktail for yeast (Sigma-Aldrich). Cells were lysed by vortexing with glass beads for 2x 3 min, on ice between. Mixture was centrifuged for 5 min at 3000 x g to remove nuclei and unbroken cells; the supernatant from this step is the lysate. The lysate was centrifuged at 100,000 x g for 30 min at 4C to generate supernatant (S100) and pellet (P100) fractions. Equal fractions of lysate, S100, and P100 were loaded on 12% acrylamide gels and immunoblotted as described above. Secretion Assays | CPY Pulse-Chase Assay. | Processing and sorting of CPY was assayed as described previously by using anti-CPY antisera (provided by Tom Stevens, University of Oregon, Eugene, OR; or Elizabeth Jones, Carnegie Mellon University, Pittsburgh, PA). Films were scanned on a flatbed scanner and processed using Adobe Photoshop 5.0 (Adobe Systems, Mountain View, CA). CPY Colony Immunoblotting. | Serial dilutions of yeast cultures were replica plated on YPD plates and incubated at 30C for 12 --24 h. Nitrocellulose filters (0.45-mum BA85; Schleicher & Schuell, Keene, NH) were overlaid for an additional 12 h at 30C. Cells were thoroughly rinsed from the filters with distilled H2O. Immunoblotting with CPY monoclonal antibodies (1:1500; Molecular Probes) or Arf1p and alkaline phosphatase polyclonal antibodies (to control for cell lysis) was performed as described above. Staining with FM 4-64. | Cultures were grown in YPD to an optical density of 0.5 --0.8. FM 4-64 (10 mM stock in dimethyl sulfoxide; Molecular Probes) was added to 500 nM; cells were incubated at 30C for 30 min. Cells were washed with water, resuspended in YPD lacking FM 4-64, and incubated an additional 30 min at 30C. Cells were concentrated 10-fold by brief centrifugation, mounted with AntiFade (Molecular Probes), and observed immediately. To assay for temperature sensitivity, strain YAB679 was shifted to 37C for 2 h before labeling. Labeling with FM 4-64 and chase incubations were performed at 37C. Cells were observed and photographed on a E600 microscope (Nikon, Melville, NY) equipped with a Spot II digital camera and MetaMorph 3.6 software (Universal Imaging, Downingtown, PA). Images were processed using Adobe Photoshop 5.0. Indirect Immunofluorescence Microscopy | pcDNA3-based plasmids for expression of GGA1, GGA1L182Q, GGA3, and GGA3L149Q with an HA epitope tag at the amino terminus were transiently transfected into NRK cells by using FuGENE 6 (Roche Applied Science, Indianapolis, IN). The cells were analyzed by indirect immunofluorescence, as described previously by using antibodies to GGA1 (R79709), the HA epitope (12CA5; BabCo), or mannosidase II (53FC3; Covance Research Products, Richmond, CA). GFP Fluorescence Microscopy | For analysis of GFP-tagged proteins by fluorescence microscopy, mid-log cultures were diluted with 0.5 volumes of GFP KILL buffer (1 M Tris pH 8.0 and 5% sodium azide; ). Samples were mixed with an equal volume of AntiFade (Molecular Probes) and observed within 1 h. Cells were photographed as described for FM 4-64 staining. RESULTS : Mutations in GAT Domain Eliminate Interaction with ARF | The Arf-binding domain of human and yeast GGA proteins was previously mapped by truncation analysis to a highly conserved region within the GAT domain . Several residues within this domain are conserved in all known GGA proteins (Figure A, highlighted residues). To disrupt binding to ARF without disrupting overall protein folding, selected point mutations within this region were generated and tested for binding to ARF by using the two-hybrid assay. To compare between species, homologous mutations were generated in both human and yeast GGA proteins. We changed leucine at position 203 of Gga1p to glutamine (Gga1pL203Q), isoleucine 207 of Gga2p to asparagine (Gga2pI207N), leucine 182 of GGA1 to glutamine (GGA1L182Q), and leucine 149 of GGA3 short form to glutamine (GGA3L149Q). Additional mutations in the GAT domain of Gga2p were also generated: residues 212 --214 were changed to alanines (referred to as Gga2pPEDL -> AAAL) and residues 219 and 220 were changed to alanines (referred to as Gga2pANKL -> AAAL). None of these mutations affect protein expression or stability, because similar levels of protein were detected by Western blotting in yeast strains that expressed wild-type or mutant forms of each protein as either Gal4AD fusions or untagged constructs (our unpublished data). These mutations are not predicted to disrupt the coiled-coils in the GAT domain, determined using the COILS program . Wild-type and mutant human GGA1 and GGA3 as Gal4AD fusions were tested for binding to ARF3Q71L as a C-terminal Gal4BD fusion. As described previously , wild-type GGA1 and GGA3 interact strongly with ARF3Q71L, shown herein by growth on medium lacking histidine and supplemented with 3-amino triazole (3AT; Figure B) and by blue color development in the beta-galactosidase assay (X-gal; Figure B). In contrast, GGA1L182Q and GGA3L149Q show no interaction by growth on 3AT plates or in the X-gal assay (Figure B). These results showed that our mutations abolished detectable binding of mammalian GGAs to ARF. We then tested wild-type and mutant yeast Gga1p and Gga2p as Gal4AD fusions for binding to Arf2pQ71L as a C-terminal Gal4BD fusion. Arf2pQ71L was used for these experiments because Arf1pQ71L is lethal in this construct . As expected , wild-type yeast Gga1p and Gga2p interact with Arf2p, shown herein by growth on 3AT plates and blue color development in the X-gal assay (Figure B). Our mutations effectively abolished this interaction. Gga1pL203Q and Gga2pI207N showed no detectable binding to Arf2p; strains grew much slower on 3AT plates and remained white in the X-gal assay, even after extended times. The overall folding of these mutant proteins was unperturbed, because these mutants remained active for binding to a novel Gga-binding protein (our unpublished data). To confirm that the loss of signal in our two-hybrid assays was due to loss of affinity for Arf, we analyzed the interaction between Arf1p and Gga2p or Gga2pI207N by affinity chromatography (Figure C). GST alone was used as a negative control. Bacterially expressed Arf1p was incubated with GTPgammaS or GDP then mixed with purified GST fusion proteins on glutathione agarose beads. Retained Arf1p was detected by immunoblotting (top). Equal amounts of GST-Gga2p, GST-Gga2pI207N, or GST were present in each reaction, as shown by Coomassie staining (bottom). Arf1pGTPgammaS, but not Arf1pGDP, bound to GST-Gga2p; the amount of Arf1p retained by Gga2p increased linearly with the concentration of Arf1p in the binding reaction. Identical experiments were performed with GST-Gga1p, but no binding was detected at these concentrations (our unpublished data). This suggests a lower affinity of Arf1 for Gga1p than for Gga2p. Because of the unique nucleotide-binding properties of Arf proteins , we do not know the actual concentration of GTP-bound Arf1p in these samples and we cannot determine a Kd value for the binding of Arf1p. However, we can detect as little as 3 ng of Arf1p by immunoblotting, which is 10-fold less than present in lane 4 (GST-Gga2p, 0.6 muM Arf1pGTPgammaS) and 100-fold less than lane 6 (GST-Gga2p, 6 muM Arf1pGTPgammaS) as determined by densitometry. Strikingly, we could not detect any binding of Arf1pGTPgammaS to GST-Gga2pI207N or GST alone, even with 6 muM Arf1p. This suggests that the I207N mutation reduced the binding to Arf1p by at least 99% in vitro. Based on the two-hybrid and in vitro binding experiments, we concluded that these conserved residues in the GAT domain are critical for interaction with GTP-bound ARF proteins. ARF Interaction Is Required for Localization of Mammalian GGA Proteins | We next tested the mammalian GGA mutants for localization in vivo. Plasmids encoding HA-tagged wild-type or mutant GGA1 or GGA3 were transfected into normal NRK fibroblasts and localized by indirect immunofluorescence (IIF) by using antibodies to GGA1 or the HA epitope (12CA5). The Golgi apparatus was double labeled using antibodies to mannosidase II (Figure , b, d, and f) or beta-COP (Figure h). Transfected wild-type GGA1 localized predominantly to the trans-Golgi region (Figure a), as expected. In contrast, GGA1L182Q was distributed throughout the cytosol, with no detectable Golgi staining (Figure c). The staining shown in Figure , a and c, used antibodies against GGA1 at a dilution that fails to detect endogenous protein, thus this staining detected only the transfected proteins. Similar results were seen with antibodies against the HA epitope (our unpublished data). At higher concentrations of GGA1 antibody, the endogenous GGA1 was detected at the Golgi region (our unpublished data), suggesting that the Golgi localization of endogenous GGA1 was not altered by overexpression of GGA1L182Q. Similarly, transfected wild-type GGA3 localized to the trans-Golgi region (Figure e), whereas GGA3L149Q was distributed throughout the cytosol (Figure g). Double labeling with antibodies to ARF (1D9; our unpublished data), mannosidase II (Figure d), or beta-COP (Figure h) showed that ARF localization and Golgi structure were unaffected by these mutant GGA proteins. These results suggest that GGA1L182Q and GGA3L149Q fail to interact with ARF in vivo as well as in vitro, and that interaction with ARF is required for normal GGA localization to the Golgi membrane in mammalian cells. Figure 2 | Localization of wild-type and mutant human GGA proteins in NRK fibroblasts. Localization of wild-type and mutant human GGA proteins in NRK fibroblasts. Cells were double labeled by indirect immunofluorescence to visualize transfected GGA proteins (left) and Golgi markers mannosidase II or beta-COP (right). Wild-type GGA1 and GGA3 are localized to the Golgi region, whereas mutants GGA1L182Q and GGA3L149Q localized throughout the cytosol. These data are consistent with a recent report that different mutations in the GAT domain of GGA3 disrupt ARF interaction and Golgi localization. Also consistent with previous work, we found that removal of the VHS or ear domains does not alter GGA localization in NRK cells (Figure ; and our unpublished data), suggesting an exclusive role for the GAT/Arf-binding domain in Golgi localization of mammalian GGA proteins. Our new results identify a distinct residue (L182 in GGA1, L149 in GGA3) in the GAT domain that is essential for interaction with ARF in vivo. Mutant Yeast GGA Proteins Retain Golgi Localization | Based on the above-mentioned results for mutant human GGAs and the high conservation of the ARF-binding domain between human and yeast orthologs, we expected the homologous mutations in yeast Gga1p and Gga2p to disrupt their localization and function. Surprisingly, this was not the case. To investigate whether Ggas and Arf interact in vivo, we analyzed the localization of wild-type and mutant Gga proteins by fluorescence microscopy and subcellular fractionation. Antibodies specific to each Gga protein were initially used for indirect immunofluorescence of yeast cells. Affinity-purified antibodies to Gga1p and Gga2p showed faint but distinct puncta throughout the yeast cytoplasm, consistent with Golgi staining (Figure , a and c). Yeast strains deleted of both GGA genes failed to stain with either antibody, indicating specificity (Figure , b and d). Endogenous Gga1p and Gga2p did not colocalize with an HA-tagged form of the early Golgi protein Och1p (; our unpublished data). Technical difficulties have hampered our attempts to colocalize Ggas with late Golgi markers. Because our detection of endogenous proteins was weak, we next localized GFP-tagged Gga proteins. GFP-Gga1p and GFP-Gga2p fully complement the trafficking defects of gga1gga2 cells (Stefan, personal communication; Figure ). These GFP-constructs were each expressed in wild-type strain BY4735. Western blots with antibodies to Gga1p or Gga2p indicated that the GFP-Ggas are expressed at significantly higher levels than endogenous Gga1p and Gga2p (our unpublished data). When analyzed by fluorescence microscopy, we detected a distinct punctate fluorescence pattern for GFP-Gga1p (Figure e) that was very similar to, but significantly brighter than, the endogenous Gga1p signal. GFP-Gga2p gave similar staining, plus high levels of cytosolic staining not seen with endogenous Gga2p (our unpublished data). We therefore used GFP-Gga1p for the studies described below. GFP-Gga1p did not colocalize with a red fluorescent protein (RFP)-tagged form of the early Golgi protein Gyp1p (; our unpublished data), consistent with our IIF results. In vps27Delta cells, which accumulate TGN, endosomal, and vacuolar components in the class E compartment, GFP-Gga1p but not RFP-Gyp1p localizes to this compartment (Figure , m and p). These data suggest that Gga proteins are localized to the late Golgi, consistent with their function and with a recent report using HA-tagged Gga1 in a class E mutant strain . Surprisingly, the L203Q mutation in GFP-Gga1p did not alter the staining pattern (Figure f), suggesting that the Golgi localization of Gga1p did not require Arf interaction. Identical results were observed with GFP-Gga2pI207N (our unpublished data). These data were in stark contrast to their mammalian counterparts. Figure 3 | Localization of Gga1p and Gga2p in yeast. Localization of Gga1p and Gga2p in yeast. Top and middle, indirect immunofluorescence of Gga1p or Gga2p in wild-type strain BY4704 (a and c) or gga1gga2 strain YAB538 T6a (b and d). Bottom, GFP fluorescence of GFP-Gga1p or mutant GFP-Gga1pL203Q in wild-type strain BY4735. ARF proteins (except ARF6) cycle on and off their target membranes in concert with GTP binding and hydrolysis; GTP-bound ARF is membrane associated, and GDP-bound ARF is cytosolic. However, after cell lysis and centrifugation, the majority of yeast and mammalian ARF is found in a soluble fraction. Mammalian GGA proteins behave similarly , consistent with ARF-dependent membrane association. We analyzed yeast Ggas by subcellular fractionation to determine whether yeast Gga proteins fractionate with membranes or cytosol, and to determine whether mutations in the Arf-binding domain altered the fractionation profile. Lysates from stains expressing endogenous or HA-tagged Gga1p or Gga2p were centrifuged at 100,000 x g to generate soluble (S100) and membrane (P100) fractions, which were analyzed on immunoblots with antibodies against endogenous Gga1p or Gga2p (our unpublished data) or the HA tag (Figure A), with identical results. Duplicate gels were probed with R23 antibodies against Arf1p. As expected, Arf1p was found almost exclusively in the S100 fraction (Figure A). In contrast, yeast Gga1p and Gga2p partitioned equally between the soluble and membrane fractions (Figure A, left), strongly suggesting that yeast Gga proteins are bound to membranes by interactions with proteins other than, or in addition to, Arf. In addition, we found that the expression levels of HA-Gga1p and HA-Gga2p are nearly identical. Each is expressed from its own promoter, suggesting that endogenous Gga1p and Gga2p expression levels are also similar. This differs from a previous report stating that Gga1p is expressed at significantly lower levels than Gga2p. Figure 4 | Immunoblots of yeast lysate fractions. Immunoblots of yeast lysate fractions. (A) Fractionation of wild-type and mutant Gga proteins. Total lysates (T) from strains expressing HA-tagged Gga2p, Gga1p, Gga2pI207N, or Gga1pL203Q were centrifuged at 100,000 x g to generate supernatant (S) and pellet (P) fractions. Duplicate gels were probed with antibodies to the HA tag (top) or Arf1p (bottom). Arf1p was predominantly in the soluble fraction, whereas all Gga proteins fractionated equally between the soluble and pellet fractions. (B) Fractionation of Gga2p in the absence of Arf1p. Fractions from wild-type or arf1Delta strain TT104 were prepared as in A, and probed with antibodies to Gga2p. Gga2p was predominantly in the pellet fraction in both strains. In the same experiments, untagged (our unpublished data) or HA-tagged mutants Gga1pL203Q and Gga2pI207N also partitioned equally between the soluble and membrane fractions (Figure A, right), confirming that Arf is not the primary mediator of Golgi localization for yeast Gga proteins. In some of our fractionation experiments, the wild-type and mutant Gga proteins partitioned exclusively into the membrane fraction. We cannot yet explain why this was so, except to note that the wild-type and mutant Ggas behaved identically in any given experiment. To independently test a role for Arf in localizing Ggas to membranes, Gga2p was analyzed in yeast cells devoid of Arf1p (arf1Delta strain TT104; ). Consistent with our mutant analysis, Gga2p fractionated with membranes in the absence of Arf1p (Figure B). This strain expresses low levels of Arf2p (10% that of Arf1p; ), unlikely to account for all membrane localization of Gga2p. We conclude that Arf is not required for recruiting Gga proteins to membranes in yeast. GGA Mutants Complement CPY Sorting Defects | Yeast strains deleted of GGA1 and GGA2 have been analyzed for defects in trafficking of several proteins, including CPY . To determine whether defects in binding to Arf disrupt the trafficking functions of Gga proteins, we tested whether Gga1pL203Q or Gga2pI207N could complement the CPY sorting defect of gga1gga2 strains. The wild-type and mutant alleles of GGA1 and GGA2 were subcloned into low copy vectors under control of their endogenous promoters and transformed into gga1gga2 yeast strain YAB538 T6a . All constructs were expressed at levels similar to the endogenous Gga proteins, as determined by Western blotting (our unpublished data). Table 2 | Yeast strains used in this study Table 3 | Oligonucleotide primers used in this study As expected, the gga1gga2 strain was defective in sorting of CPY, as shown by pulse-chase analysis of newly synthesized CPY (Figure A) and colony immunoblot (Figure B). Also as expected, wild-type Gga2p fully complemented the defects in both assays (Figure , A and B). Surprisingly, the mutant Gga2pI207N also fully complemented these defects, as shown by pulse-chase analysis (Figure A) and colony immunoblotting (Figure B). We concluded that either Arf interaction is not required for Gga2p function, or these mutants can still bind Arf in vivo. Expression of Gga1p complemented the CPY-processing and -sorting defects only partially (Figure , A and B). Our first indication of a functional defect in our mutant Gga proteins was seen for Gga1pL203Q, which partially complemented the CPY processing defect in the pulse-chase assays (Figure A) but failed to complement the CPY sorting defect (Figure B). There was minor variability in the extent of partial complementation by Gga1pL203Q (compare Figure , A and B), but it is always less than wild-type Gga1p. This suggests that the sorting function of Gga1p is partially disrupted by the mutation that blocks Arf binding. Figure 5 | CPY sorting is complemented by Arf-binding domain mutants in yeast. CPY sorting is complemented by Arf-binding domain mutants in yeast. (A) Pulse-chase analysis of newly synthesized CPY in wild-type, gga1gga2, and gga1gga2 harboring low copy plasmids of Ggas. Time course is indicated in minutes. Expression of Gga2p or Gga2pI207N (right) fully complemented the CPY processing defect in gga1gga2 strains. Expression of Gga1p or Gga1pL203Q (middle) partially complemented this defect. P1, unglycosylated endoplasmic reticulum form of CPY; P2, glycosylated Golgi form; M, fully processed mature (vacuolar) form. (B) Detection of secreted CPY by colony immunoblotting. Serial dilutions of yeast cultures were replica plated onto YPD plates. A signal on colony immunoblots indicates that CPY is secreted from the cell, suggesting that CPY is missorted. The CPY sorting defect of gga1gga2 strains (DeltaDelta) was complemented by wild-type Gga2p and mutant Gga2pI207N. Wild-type Gga1p partially complemented the CPY sorting defect, whereas mutant Gga1pL203Q failed to complement. Although the homologous mutations in human GGA1 and GGA3 disrupt ARF interaction in vivo, this single mutation might not fully eliminate Arf binding in the yeast Gga2p mutants. To address this possibility, we made two clusters of alanine substitution mutations in the Arf-binding domain of Gga2p. We replaced three residues of a conserved PEDL sequence (P212A, E213A, and D214A), designated Gga2pPEDL -> AAAL, or two residues in a conserved ANKL sequence (N219A and K220A), designated Gga2pANKL -> AAAL. The latter includes the asparagine residue mutated by to disrupt ARF --GGA3 interaction. Neither mutant Gga2pPEDL -> AAAL nor Gga2pANKL -> AAAL interacted with Arf2p in the two-hybrid assay (Figure B). However, both mutants still complemented the CPY defect when assayed by either pulse chase or colony immunoblot (our unpublished data), showing that they were functional in yeast. These data suggest that reduction of Arf interaction does not affect the ability of Gga2p to sort CPY in yeast. Genetic Interaction between GGAs and VPS27 | As a complementary approach to understand the function of Gga proteins in yeast, we took advantage of the nonlethal phenotype of GGA gene deletions. Specifically, we tested for synthetic interactions between GGA deletions and genes that function in distinct post-Golgi trafficking pathways: APM1, APM3, VPS1, and VPS27. APM1 encodes the medium (mu) subunit of the adaptor protein (AP)-1 clathrin adaptor complex, which is thought to function in anterograde or retrograde traffic between the TGN and early endosome . However, we note that deletion of APM1 does not eliminate AP-1 function, because a related protein, Apm2p, can substitute in the AP-1 complex . APM3 encodes the medium (mu) subunit of the AP-3 clathrin adaptor complex, which mediates a direct TGN-to-vacuole pathway . Deletion of APM3 eliminates AP-3 function. VPS1 encodes yeast dynamin, a protein involved in the scission of clathrin-coated vesicles from donor membranes . Vps1p mediates both TGN-to-endosome and TGN-to-vacuole traffic, and deletion causes missorting of vacuolar proteins via the plasma membrane . VPS27 is a class E gene that regulates exit from the prevacuolar compartment (PVC) and entry into intralumenal vesicles . Strains were constructed with triple gene deletions and analyzed for growth defects at 30 or 37C. Deletion of either apm3 or vps1 in the gga1gga2 knockout strain showed no synthetic growth defects (Figure A, strains YAB658 and YAB667). Deletion of apm1 in the gga1gga2 knockout produced a very slight growth defect at 37C (Figure A, strain YAB650). Furthermore, these three triple-knockout strains showed no synthetic defects in trafficking of alkaline phosphatase, CPY, or Vps10p (our unpublished data). Figure 6 | (A) Synthetic ts growth phenotype of gga1gga2vps27 strains. (A) Synthetic ts growth phenotype of gga1gga2vps27 strains. Strains were tested for synthetic growth defects at 30 or 37C. Strains were streaked onto YPD plates and incubated for 2 d. Deletion of the AP-1 medium subunit (apm1), the AP-3 medium subunit (apm3), or the dynamin homolog VPS1 (vps1) showed no synthetic growth phenotypes with gga1gga2. In contrast, deletion of VPS27 in two different gga1gga2 strain backgrounds was (ts for growth. Strains: wild-type (BY4704), gga1gga2 (YAB538 T6a), gga1gga2apm1 (YAB650), gga1gga2apm3 (YAB658), gga1gga2vps1 (YAB667), gga1gga2vps27 in S288C background (YAB677), gga1gga2vps27 in SEY6210 background (YAB679), and vps27 (JUY68). (B) Synthetic ts growth phenotype of gga1gga2vps28 strain. As in Figure A, strains were streaked onto YPD plates and incubated for 2 d at 30 or 37C. Strains: wild-type (SEY6210), gga1gga2 (GPY2385), vps27 (JUY68), vps28 (YCS41), gga1gga2vps27 (YAB680), and gga1gga2vps28 (YCS195). All strains grew at 30C. Deletion of GGA1, GGA2, and VPS28 is ts for growth, similar to deletion of GGA1, GGA2, and VPS27. (C) Class E compartment forms at 30 and 37C in gga1gga2vps27 strain. Cells grown at 30 or 37C were stained with FM 4-64 to visualize the vacuole and prevacuolar compartment. FM4-64 fluorescence (left) and differential interference contrast images (right) are shown. Wild-type cells show vacuolar staining with no enlarged PVC (our unpublished data). In contrast, the gga1gga2vps27 cells had enlarged PVC, indicated by bright staining adjacent to the vacuole, at both 30 and 37C. In contrast, cells deleted of GGA1, GGA2, and VPS27 were temperature sensitive (ts) for growth in two different strain backgrounds (Figure A, strains YAB677 and YAB679). The slight differences in strain background are currently unexplained. This ts growth defect was reversible after at least 3 d at the restrictive temperature, indicating growth inhibition rather than lethality (our unpublished data). This tight and reversible ts growth phenotype in the SEY6210 background was found in multiple isolates of the genotype, including strain YAB680 shown in Figure B. To determine whether the ts phenotype was related to the VHS domain of Vps27p, we tested another class E gene for genetic interaction with GGA deletion. Deletion of GGA1, GGA2, and VPS28 in the SEY6210 background was synthetically ts for growth (Figure B). This suggests that the synthetic phenotypes are due to the function of class E genes rather than the VHS domain of Vps27p. A notable defect of class E mutants (such as vps27) is the enlargement of the PVC, or the class E compartment. Using FM 4-64 as a marker for the vacuole and PVC, we tested for PVC enlargement in the triple knockout strain to determine whether GGA deletion prevented the formation of this compartment. Both vps27 (our unpublished data) and gga1gga2vps27 strains (Figure C) contained characteristic class E compartments at 30 and 37C, suggesting that the PVC can form in the absence of GGA gene function. We also noted that the vacuolar morphology was altered in these strains, with a bubble-like appearance. This is consistent with a vacuolar phenotype of gga1gga2 strains reported by others and not observed in our previous experiments with a different strain background. GGA Mutants Complement gga1gga2vps27 Phenotype | To determine whether binding to Arf was required for Gga1p or Gga2p to complement the synthetic ts phenotype of the gga1gga2vps27 strain, the wild-type and mutant alleles of GGA1 and GGA2 were transformed into yeast strain YAB680. Transformants were streaked onto YPD and incubated at 30 or 37C for 2 d. The ts phenotype was fully complemented by expression of plasmid-borne wild-type GGA2 . The three mutant alleles Gga2pI207N, Gga2pPEDL->AAAL, and Gga2pANKL->AAAL also fully complemented the growth defect at 37C when expressed in this strain . Thus, these mutants seemed to complement all Gga trafficking functions. We concluded that Gga2p function is not dependent on Arf, or that these mutants do not disrupt Gga2p-Arf binding in vivo. As in the CPY assay, expression of Gga1p partially complemented the ts growth defect . High copy expression of GFP-Gga1p fully complemented the ts growth defect , showing that the partial complementation by Gga1p could be overcome by higher expression (see DISCUSSION). In contrast, expression of mutant Gga1pL203Q failed to complement the ts growth defect; only very slow growth was restored . This suggests that Arf binding is important for Gga1p function. At high expression levels, GFP-Gga1pL203Q partially complemented the ts growth defect . This suggests that the mutation reduced, but did not eliminate, Gga1p function. Figure 7 | Synthetic ts phenotype of gga1gga2vps27 is fully complemented by mutant Gga2p. Synthetic ts phenotype of gga1gga2vps27 is fully complemented by mutant Gga2p. Plasmids encoding wild-type or mutant Gga1p or Gga2p were transformed into gga1gga2vps27 strain YAB680. GFP fusions of Gga1p are expressed at high copy, all others at low copy. Serial dilutions of wild-type, YAB680, and each transformed strain were replica plated on YPD and incubated at 30 or 37C. Complementation of the ts growth defect is indicated by restored growth at 37C. N- and C-Terminal Domains Confer Golgi Localization of Yeast GGA Proteins | Because interaction with Arf was not required for Golgi localization of Gga proteins, we made a series of truncation constructs of GFP-Gga1p to test the VHS, hinge, and ear domains for roles in Gga localization. All constructs that included the GAT domain were made as wild-type and L203Q mutants to analyze the role of Arf interaction in the truncated proteins. We used Gga1p for these experiments because the L203Q mutation reduces function of the full-length protein. All proteins were expressed at levels similar to full-length GFP-Gga1p, as determined by Western blotting (our unpublished data). Each GFP fusion protein was localized in wild-type strain BY4735 by fluorescence microscopy. Full-length wild-type GFP-Gga1p and mutant GFP-Gga1pL203Q are shown for comparison (Figure , a and b). Figure 8 | Schematic of truncation constructs of GFP-Gga1p and Gga2p. Schematic of truncation constructs of GFP-Gga1p and Gga2p. Figure 9 | Localization of truncated GFP-Gga1p. Localization of truncated GFP-Gga1p. (a --h) Strains expressing wild-type or mutant versions of the indicated constructs were analyzed by GFP fluorescence. Punctate staining, characteristic of Golgi localization, was observed in a, b, e, g, and h. Cytosolic staining was observed in c, d, and f. (i --l) Localization of wild-type and mutant human GGA1DeltaVHS. NRK cells were transiently transfected with pcDNA3-based vectors for expressing GGA1DeltaVHS (residues 145 --639, i --j) or mutant GGA1DeltaVHS, L182Q (k and l). Cells were double-labeled for IIF with antibodies to GGA1 (R79709, i and k) and mannosidase II (j and l). GGA1DeltaVHS localizes to the Golgi region, whereas GGA1DeltaVHS, L182Q is cytosolic. (m --p) Localization of yeast Gga1p constructs in a class E mutant strain. The indicated constructs were expressed in a strain deleted of VPS27. Full-length GFP-Gga1p (m) and GFP-Gga1p1 --331 (n) localize to the class E compartment. In contrast, GFP-Gga1p141 --557 (o) localizes to distinct puncta throughout the cytoplasm in addition to limited staining near the vacuole. The early Golgi protein Gyp1p, expressed as an RFP fusion, localizes exclusively to Golgi puncta (p). A construct expressing the yeast GAT domain alone (GFP-Gga1p141 --331) was localized in the cytosol (Figure c), in stark contrast to the GAT domain of mammalian GGAs. A similar construct as a Gal4AD fusion protein interacts with Arf2pQ71L in the two-hybrid assay , yet Arf interaction is not sufficient to localize the yeast GAT domain to membranes in vivo. The mutant GFP-Gga1p141 --331,L203Q also localized to the cytosol (Figure d). Similarly, the VHS domain alone was cytosolic (our unpublished data). Importantly, a construct expressing the VHS/GAT domains (GFP-Gga1p1 --331) localized to the Golgi and was indistinguishable from full-length GFP-Gga1p (Figure e). When this same construct carried the L203Q mutation, GFP-Gga1p1 --331,L203Q was found exclusively in the cytosol (Figure f). We made two conclusions from these results. First, we concluded that binding to Arf contributes to stable association of Gga1p with membranes and that the L203Q mutation indeed disrupts interaction with Arf in vivo. Second, we concluded that the VHS domain confers weak Golgi localization, which is stabilized by GAT binding to Arf. A construct expressing the GAT/hinge/ear domains (GFP-Gga1p141 --557) localized to Golgi-like puncta (Figure g). This suggests that the hinge and/or ear domains also contribute to stable association with membranes. This construct carrying the L203Q mutation (GFP-Gga1p141 --557,L203Q) also localized to Golgi-like puncta (Figure h), suggesting that the hinge and/or ear domains interact with some Golgi component with high enough affinity to drive the GFP reporter onto the Golgi, regardless of Arf interaction. Consistent with this observation, a hinge/ear construct of Gga1p also seems membrane associated (our unpublished data). In contrast, comparable constructs of human GGA1 confirm that only the GAT domain confers Golgi localization in mammalian cells. IIF of GGA1DeltaVHS in NRK cells shows Golgi localization indistinguishable from full-length GGA1, whereas GGA1DeltaVHS,L182Q is entirely cytosolic (Figure , i --l). Interestingly, the GFP-Gga1p141 --557 construct lacking the VHS domain seemed to stain more puncta than did full-length Gga1p, suggesting that the VHS domain may restrict Gga1p to distinct Golgi cisternae (likely the TGN). To test for altered localization of the truncated constructs, each was expressed in a vps27Delta strain (Figure , m --p). Full-length GFP-Gga1 and GFP-Gga1p1 --331 localized exclusively to the class E compartment, suggesting that removal of the hinge/ear domain did not alter the localization of Gga1p. In contrast, GFP-Gga1p141 --557 localized to spots throughout the cytoplasm in addition to some accumulation near the vacuole (Figure o). The class E compartment formed in these cells, as determined by FM 4-64 staining (our unpublished data). Early Golgi cisternae do not collapse into the class E compartment; RFP-Gyp1p localized to puncta in vps27Delta cells (Figure p). These data suggest that removal of the VHS domain caused mislocalization of Gga1p to other organelles such as early Golgi. We concluded that at least three domains on Gga proteins interact with Golgi-associated proteins to recruit and stabilize Gga proteins at the TGN and that Arf interaction occurs in vivo but is neither sufficient nor required for Gga localization. Both VHS and Hinge/Ear Domains Are Required for GGA Function | To determine whether the VHS and hinge/ear domains are required for Gga function, we tested truncations of Gga2p and GFP-Gga1p (diagrammed in Figure ) for complementation of the CPY sorting defect of gga1gga2 or ts growth defect of gga1gga2vps27 strains. Each construct was expressed and stable, as determined by Western blotting (our unpublished data). In contrast to the GAT domain mutants, deletion of the VHS domain or the hinge and ear domains eliminated the function of Gga2p and GFP-Gga1p. A construct expressing the VHS and GAT domains of Gga2p (Gga2p1 --326) failed to complement the gga1gga2vps27 ts growth defect or the CPY sorting defect when assayed by pulse-chase or colony immunoblot (our unpublished data). In all three assays, the strains expressing Gga2p1 --326 were indistinguishable from strains lacking Gga2p entirely. This indicates that the hinge and/or ear domains are essential for Gga2p function. Similarly, a construct expressing the GAT, hinge, and ear domains of Gga2p (Gga2p110 --585) failed to complement the gga1gga2vps27 ts growth defect or the CPY sorting defect (our unpublished data). These results indicate that the VHS domain is essential for Gga2p function. Figure 10 | VHS and ear domains are required for Gga function in yeast. VHS and ear domains are required for Gga function in yeast. Indicated constructs of Gga2p or GFP-Gga1p were transformed into gga1gga2vps27 strain YAB680. Serial dilutions were replica plated on YPD and grown at 30 or 37C for 2 d. None of the truncated proteins complemented the ts growth defect, indicating that the VHS and ear domains are required for Gga function in yeast. Identical results were obtained using truncations of GFP-Gga1p to complement the gga1gga2vps27 growth defect . Neither GFP-Gga1p1 --331 nor GFP-Gga1p141 --557 complemented the defect. Not surprisingly, a construct expressing only the GAT domain (GFP-Gga1p141 --331) also failed to complement the gga1gga2vps27 growth defect. Together, these data show that the VHS, hinge, and ear domains are important not only for localization of Ggas to the Golgi but also for the function of the Ggas at the Golgi. DISCUSSION : Our main focus of this work was to analyze the functional significance of Gga --Arf interaction in yeast. Because of the high homology between human and yeast GAT domains, particularly in the ARF-binding domain, we expected that the function of this domain would be conserved between human and yeast GGA proteins. Likewise, we expected that mutations within this domain would have similar effects on the localization and function of yeast and human GGA proteins. Surprisingly, our data refute this hypothesis and show strong differences between human and yeast GGA proteins. In mammalian cells, ARF interaction is necessary and sufficient for recruitment of GGAs to the Golgi. In yeast, the VHS, GAT, and hinge/ear domains all contribute to Golgi localization; when expressed individually, only the hinge/ear domain is sufficient for Golgi localization. Although we can detect interaction with Arf in vivo, this interaction is not required for localization or function of yeast Gga proteins. Mutant GGA Proteins | We are confident that the mutations we generated in the GAT domain eliminate binding to activated ARF. First, our two-hybrid assays show loss of interaction with all mutations, and affinity chromatography confirms the loss of affinity between Gga2p and Arf1p. Second, the mutations in human GGA1 and GGA3 cause a complete shift from the Golgi to the cytosol, consistent with a recent report describing a mutation in the GAT domain of GGA3 . Third, in the context of a truncated GFP-Gga1p (but not in full-length GFP-Gga1p), the L203Q mutation causes a complete shift from the Golgi to the cytosol. Our results extend those in a recent report characterizing a mutation in human GGA3 that eliminates ARF --GGA interaction. Herein, we identified at least two additional amino acids in the ARF-binding domain that are also essential for the ARF --GGA interaction. Notably, showed that mutation of the aspartate in the GGA3 PEDL sequence (D189) had no effect on ARF interaction. Because our PEDL->AAAL mutation eliminates Arf2p binding, we conclude that the proline or glutamate residues must be important for ARF binding. The growing number of residues within the GAT domain now reported as essential for interaction with ARF proteins is interesting, particularly that alteration of any one residue eliminates interaction with ARF. Perhaps the affinity of the GGA --ARF interaction is low, and elimination of any contributing element reduces the affinity below detection by two-hybrid analysis. Alternatively, the mutations may alter the structure of the GAT domain sufficiently to disrupt interaction. Structures of this domain and cocrystal structures of ARF and GGAs will surely reveal much about the interaction. Differences between Yeast and Human GGA Proteins | In mammalian cells, ARF interaction is necessary for recruitment of GGAs to the Golgi. In yeast, our mutants have minimal effects on both localization and function, suggesting that Arf interaction is not required. The two most surprising results are that Gga1pL203Q remained Golgi localized and that all Gga2p mutants are fully functional. One possible explanation for our data is that yeast Ggas interact with Arf in vivo, and the mutations described herein reduce the affinity for Arf but do not fully disrupt the interaction. We note that the ARF-GGA interaction is significantly weaker in yeast than in mammalian cells, because the GAT domain of yeast Gga1p is not targeted to the Golgi. However, the GAT domain does contribute to Golgi localization: VHS/GAT is Golgi localized, whereas VHS alone or VHS/GATL203Q is cytosolic. The strength of the interactions between Ggas and other proteins (via the VHS and hinge/ear domains) could stabilize the Gga --Arf interaction at the Golgi and allow signaling from activated Arf. We favor a model in which interaction between Gga and Arf is stabilized by other protein interactions, hence reducing the effect of the mutations in vivo. We conclude that the role of Arf --Gga interaction in yeast is not to recruit Ggas to the Golgi, but rather to alter the structure of Ggas to allow sequential interactions with clathrin or other components. Unfortunately, we cannot detect expression of mammalian GGAs in yeast. We have tried multiple promoters, fusion constructs, and truncation constructs; only the two-hybrid constructs are expressed at a detectable level. Furthermore, we cannot detect yeast Ggas in mammalian cells after either plasmid transfection or protein microinjection. Hence, we cannot test for complementation of the yeast phenotypes by wild-type or mutant human GGAs. Differences between Gga1p and Gga2p | Previous reports concluded that Gga1p and Gga2p are functionally redundant, because expression of either Gga1p or Gga2p can complement the defects of gga1gga2 strains . However, our data reveal differences between Gga1p and Gga2p. Three mutant alleles of Gga2p function fully in all trafficking-related assays, whereas the function of mutant Gga1p is reduced. Also, full complementation of gga1gga2 defects by Gga1p requires high expression levels. There are several possible explanations for these differences. First, Gga1p and Gga2p may perform distinct functions in yeast and only Gga1p function is dependent upon interaction with Arf. This scenario is unlikely, because all gga1gga2 defects described to date can be rescued by expression of either Gga1p or Gga2p. Second, the different effect of the mutations may reflect differences in affinities for ARF. We feel this is the best explanation of our data. Our two-hybrid and affinity chromatography analyses indicate that the interaction between Gga1p and Arfs is weaker than that between Gga2p and Arfs. Whether these differences in affinity are a true reflection of affinity in vivo is not known. If the mutations reduce affinity for Arf, this reduction could ablate Gga1p --Arf interaction in vivo, but not fully ablate Gga2p --Arf interaction. Perhaps we would see an effect of the Gga2p mutants if we lowered the level of Gga2p expression. Third, Gga1p may interact with different proteins than Gga2p. Future screens to identify binding partners will address this possibility. Fourth, Gga1p may interact with the same proteins as Gga2p, but with different affinities. Perhaps these interactions are stronger with Gga2p than with Gga1p, thus stabilizing the Gga2p --Arf interaction and masking any effects of the mutants. High expression of Gga1p could force interaction with these proteins, thus complementing the double deletion. Fifth, the level of expression of Gga2p may be higher than that of Gga1p, as noted previously . However, our results with HA-tagged Gga1p and Gga2p argue that the expression levels are in fact very similar. Finally, the L203Q mutation may reduce interaction with another, unknown partner of Gga proteins that is involved in Gga function. Although we cannot rule out this possibility, the current evidence strongly favors Gga --Arf interaction via the GAT domain. Roles of VHS Domain | The VHS domain is required for function; a construct lacking the VHS domain fails to rescue the ts phenotype of gga1gga2vps27. This conclusion is consistent with two recent reports showing that the VHS domain is required for CPY sorting. If the yeast and mammalian VHS domains of GGA proteins have similar functions then the VHS domain of Gga1p and Gga2p should interact with cargo that is sorted into Gga-dependent pathways. The mammalian GGA proteins interact with an acidic cluster-dileucine motif in transmembrane cargo receptors that traffic from the TGN to endosomes . The closest yeast homolog to these cargoes is the CPY receptor Vps10p. We cannot detect direct interaction between the Vps10p cytoplasmic tail and Gga proteins by two-hybrid analysis (our unpublished data). It will be important to identify the VHS binding partners in yeast and to test whether the yeast VHS domain functions to recognize and sort cargo. Because the VHS domain of Vps27p cannot substitute for the Gga VHS domain , it will also be interesting to identify essential residues within the GGA VHS domain required for function. We also show that the VHS domain contributes to the Golgi localization of Gga proteins, likely through interaction with other Golgi-associated proteins. This contribution is seen most definitively with the VHS/GAT construct of GFP-Gga1p. This construct (GFP-Gga1p1 --331) is localized to Golgi-like puncta, whereas addition of the L203Q mutation causes exclusively cytosolic staining. It is possible that the L203Q mutation alters folding of the VHS/GAT construct such that interactions with the VHS domain are disrupted. We do not favor this explanation, because the L203Q mutation has only minor effects on Gga1p function. If L203Q altered VHS folding, we would expect the total loss of function phenotype of the VHS deletion. Hence, we conclude that both the VHS domain and the GAT domain interact weakly with Golgi-localized proteins, and that together these interactions confer Golgi localization. It seems that the GFP-Gga1p construct lacking the VHS domain localizes to more puncta than does full-length GFP-Gga1p. There are two possibilities for this staining pattern. First, this construct may fragment the late Golgi. Second, the truncated Gga1p may mislocalize to other membranes such as early Golgi or endosomes. Our data using vps27Delta cells show that the class E compartment forms but the construct still localizes to distinct puncta, supporting the second possibility. These data suggest that the VHS domain confers specificity to the late Golgi, perhaps by binding to cargo proteins that are present at the TGN. Because the VHS/GAT construct localizes to Golgi but GAT alone does not, a genetic screen for mutants that disrupt Golgi localization of VHS/GAT may identify the VHS-binding partners. Role of Hinge/Ear Domains | The hinge and/or ear domains are required for function; when both domains are deleted, Gga1p and Gga2p are nonfunctional. reported that deletion of the ear domain partially reduced Gga function. Hence, our results suggest that the hinge is required for Gga function. Because clathrin interacts with the hinge domain and facilitates vesicle formation , this loss of function is likely due to loss of clathrin interaction, leading to loss of vesicle formation. However, the hinge and ear domains also confer Golgi localization to yeast Gga proteins. This contrasts with the human GGA proteins, for which the hinge/ear domain is soluble. Two models could explain this result. First, clathrin is present at the TGN membrane in yeast and helps to recruit Gga proteins. Second, a trans-Golgi component interacts with the ear domain to recruit Ggas to the TGN, which in turn recruit clathrin. Many have shown that clathrin is recruited from the cytosol to the membrane by adaptor proteins (for review, see ), supporting the second model. However, the identity of these potential ear binders is unknown, and the role of the ear domain is still unclear. Perhaps the ear domain interacts with Golgi components to increase cargo specificity, stabilize the interactions during vesicle formation, or disassemble the complex after vesicle formation. As discussed for the VHS domain, it will be very informative to identify Gga ear partners. Mammalian GGA Function | We show that ARF interaction is required for the localization of human GGA1 and GGA3 to the TGN. This is consistent with a recent report showing that a mutant GGA3 is cytosolic with dramatically reduced function. We found that high levels of expression of mutant GGA3 shift M6PR to the plasma membrane (our unpublished data). This was similar to the shift we observed with wild-type GGA3 overexpression , suggesting that mutant GGA3 in fact retains some function. There are two possible explanations for this observation. First, the mutant GGA3 may interact weakly with ARF, and high levels of expression allow a functional interaction at the TGN. However, we do not see any accumulation of mutant GGA3 at the TGN in our immunofluorescence analyses. Second, mutant GGA3 may interact with binding partners in the cytosol rather than at the TGN, sequestering these components away from their functional sites on the TGN or plasma membrane. The most likely candidate is clathrin, shown to be recruited by GGAs to the TGN . This would be interesting, because it would indicate that GGA --clathrin interaction is not dependent upon an established ARF --GGA interaction, but that GGA and clathrin are recruited together to the TGN through interaction between GGA and ARF. If this sequestering occurs, then we would expect that clathrin-dependent endocytosis would be reduced. We have preliminary data showing that transferrin receptor accumulates on the plasma membrane in cells overexpressing GGA3, supporting this possibility. Further studies must be done to show interaction of the mutant GGA3 with clathrin in cytosol. Genetic Interactions | Our characterization of the synthetic growth defect in gga1gga2vps27 and gga1gga2vps28 strains supports the hypothesis that Gga proteins are involved in delivery of select cargo proteins to the endosomes. In vps27 and vps28 strains, vacuolar function is maintained due to mixing of vacuolar components into the enlarged PVC, or class E compartment, allowing cell growth . We show that the enlarged PVC is still able to form in the gga1gga2vps27 strain, even at the restrictive temperature, indicating that not all traffic to the PVC is blocked. If some essential proteins cannot reach the PVC due to loss of Gga function then vacuolar function would be further reduced. When cells are stressed by high temperature, this reduced function may be exacerbated, leading to growth inhibition. Because both Vps27p and Gga proteins contain a VHS domain, we tested whether the synthetic interaction is related to VHS function or common to all class E VPS genes. The ts phenotype of gga1gga2vps28 strains supports the latter. In our genetic analyses, we also tested for synthetic interactions with VPS1 and with the medium subunits of adaptor complexes AP-1 and AP-3. Others have recently reported genetic interactions between GGAs and the large subunits of AP-1 . We show that deletion of the medium subunit of AP-1 does not exacerbate the gga1gga2 phenotype, except for very minor ts growth inhibition. Our result is likely due to incomplete loss of AP-1 function when the medium subunit is deleted, because Apm2p can function in AP-1 complexes . We also show that VPS1 and GGAs do not display genetic interactions. It is possible that Vps1p is involved with Ggas in vesicle formation, but deletion of GGAs does not accentuate the more severe vps1 phenotype. Because a number of lipid modifying enzymes act upstream or downstream of ARF1, it will be interesting to test for genetic interactions between GGAs and these same genes. Together, our data suggest a model in which yeast Gga proteins assemble a protein complex at the TGN that is required for proper sorting and transport of proteins to endosomes. The overall function of GGA proteins is thus conserved between yeast and humans. However, the mechanics of the interactions seem to be very different. In mammalian cells, interaction between activated ARF and GGA proteins recruits GGAs to the Golgi membrane, where interactions with cargo and clathrin can occur. In yeast, it seems that interactions between Gga proteins and cargo and/or other Golgi components drive the recruitment of Ggas to the Golgi membrane, where Arf interaction can occur. We speculate that a conformational change in Gga proteins upon Arf activation then drives the formation of a functional complex. The mutant Gga proteins with reduced Arf-binding affinity still function in yeast, because Arf and Ggas are juxtaposed at the Golgi membrane by other interactions, thus stabilizing the Arf --Gga interaction. Backmatter: Abbreviations used: : ARF = ADP-ribosylation factor Gal4AD = Gal4 activation domain Gal4BD = Gal4 binding domain GAT = GGA and Tom1 GFP = green fluorescent protein Golgi-localized = GGA, Golgi-localized, gamma-ear homology domain, ARF-binding protein GST = glutathione S-transferase M6PR = mannose 6-phosphate receptor PVC = prevacuolar compartment RFP = red fluorescent protein TGN = trans-Golgi network VHS = domain present in Vps27p, HRS, and STAM proteins VPS = vacuolar protein sorting PMID- 12221118 TI - Inversin Forms a Complex with Catenins and N-Cadherin in Polarized Epithelial Cells AB - Nephrogenesis starts with the reciprocal induction of two embryonically distinct analages, metanephric mesenchyme and ureteric bud. This complex process requires the refined and coordinated expression of numerous developmental genes, such as inv. Mice that are homozygous for a mutation in the inv gene (inv/inv) develop renal cysts resembling autosomal-recessive polycystic kidney disease. The gene locus containing inv has been proposed to serve as a common modifier for some human and rodent polycystic kidney disease phenotypes. We generated polyclonal antibodies to inversin to study its subcellular distribution, potential binding partners, and functional aspects in cultured murine proximal tubule cells. A 125-kDa inversin protein isoform was found at cell-cell junctions. Two inversin isoforms, 140- and 90-kDa, were identified in the nuclear and perinuclear compartments. Plasma membrane allocation of inversin is dependent upon cell-cell contacts and was redistributed when cell adhesion was disrupted after incubation of the cell monolayer with low-calcium/EGTA medium. We further show that the membrane-associated 125-kDa inversin forms a complex with N-cadherin and the catenins. The 90-kDa nuclear inversin complexes with beta-catenin. These findings indicate that the inv gene product functions in several cellular compartments, including the nucleus and cell-cell adhesion sites. Keywords: INTRODUCTION : Organogenesis of the mammalian kidney involves the coordinated regulation of gene expression that occurs during the reciprocal induction of two embryonically distinct analages, mesenchymal metanephric blastema and epithelial ureteric bud. Successful nephrogenesis and maturation of renal tubules requires a combination of growth pattern control, cell fate determination, and cell cycle control. A relatively common derangement of tubule maturation is the growth of fluid-filled epithelial cysts. Mutations in one of several genetic loci can lead to a cystic phenotype characterized by epithelial cell proliferation, reversal of cell polarity, and alterations in apoptosis, extracellular matrix, and transepithelial transport of fluids and electrolytes (for review, see ). Rodent models have proved valuable in elucidating the mechanisms underlying renal epithelial cyst formation. One such model, the inversion of embryonic turning (inv/inv) mutant mouse was created by insertional mutagenesis and exhibits an autosomal recessive form of situs inversus and renal cysts , the latter resembling autosomal recessive polycystic kidney disease (PKD). The predicted sequence of inversin, the inv gene product, has 15 Ank/Swi6 motifs arranged in a tandem array located toward the amino-terminal side of the protein . The C-terminal one-half of inversin has no similarity to other proteins . Subcellular distribution and potential binding partners of inversin have yet to be discovered. Although the function of the inv gene product is unknown, inversin's potential contribution to renal cystogenesis is supported by quantitative trait localization data. Quantitative trait localization indicates that inv is one of many genes within a locus that contains one or more modifying genes in the pcy, cpk, jck, and bpk mouse models of human PKD. (; ; ; D.D. Woo, personal communication). Modifying genes have been proposed to explain the diversity of PKD phenotypes observed in humans and animal models . The wide range of PKD phenotypes in humans and animal models, combined with the numerous gene defects underlying cyst formation, suggests there is an abundance of proteins that participate in a common pathway of cystogenesis. Inversin appears to be one of these proteins. In this paper, the subcellular distribution of inversin is characterized. In a murine S1 proximal tubule cell line, affinity-purified antibodies raised against inversin's c-terminal domain show nuclear, perinuclear, and plasma membrane labeling by confocal microscopy and the detection of 140-, 125-, and 90-kDa isoforms on immunoblots. The 140- and 90-kDa isoforms were enriched in nuclear and cytoplasmic extracts, whereas the 125-kDa isoform was most abundant in membrane fractions. Inversin antibodies immunoprecipitated N-cadherin and alpha, beta, and gamma-catenins, but not E-cadherin, beta1-integrin, or vinculin. Reverse immunoprecipitations with antibodies to N-cadherin and beta-catenin confirmed that inversin forms a stable complex with these integral membrane proteins. Like inversin, N-cadherin has been implicated in defective left-right asymmetry , and transgenic mice overexpressing beta-catenin develop cystic kidneys . Inversin's localization at the lateral membrane is dependent upon cell-cell contacts and can be disrupted by incubating the monolayer with low-calcium/EGTA medium. These findings suggest that isoforms of inversin may participate in intercellular junction biogenesis and gene transcription similar to beta-catenin's cellular function. MATERIALS AND METHODS : Cell Culture | An immortalized cell line derived from the early segment (S1) of the proximal tubule (S1 cells) was a kind gift from Dr. G.T. Nagami (UCLA School of Medicine, Los Angeles, CA; ). Cells were maintained in a 50:50 mixture of Hams-F12:DMEM media supplemented with 2 mM l-glutamine, 10 mM Na-HEPES, 2 mM sodium pyruvate, insulin, sodium-selenite, and sodium-bicarbonate, and 7% fetal calf serum, penicillin, and streptomycin (, 2000). Cells were grown on dishes, coverslips, and polyester membrane filters (0.4-mum pore size; Transwell, , Cambridge, MA) coated with type I rat-tail collagen (Invitrogen, Carlsbad, CA; ). Reagents and Antibodies | Cell media, fetal calf serum, and reagent grade chemicals were supplied by Sigma (St. Louis, MO). Mouse monoclonal antibody to histone H1 was obtained from StressGen Biotechnologies (Victoria, British Columbia, Canada). Mouse-monoclonal antibodies to alpha-, beta-, and gamma-catenin, rat monoclonal antibody to E-cadherin, and polyclonal rabbit antibody to beta-catenin were obtained from Zymed Laboratories (South San Francisco, CA). Pan-cadherin antibody was purchased from Sigma. Polyclonal rabbit antibody to N-cadherin was obtained from Calbiochem (La Jolla, CA). Mouse monoclonal antibodies to vinculin, beta1-integrin, and E- and N-cadherin were purchased from BD Transduction Laboratories (Franklin Lakes, NJ). All secondary antibodies were obtained from Jackson ImmunoResearch Laboratories (West Grove, PA). 4',6-Diamidino-2-phenyindole (DAPI) nucleic acid stain, rhodamine-phalloidin, and SYPRO protein gel stain were purchased from Molecular Probes (Eugene, OR). All other chemicals and reagents were supplied by Fisher Scientific (Pittsburgh, PA). Inversin Polyclonal Antibody | An expressed sequence tag (EST; NCBI accession no. ) was obtained from Incyte Genomics (St. Louis, MO; ) and was sequenced to confirm homology to the 3' end of the inv gene (Lark Technologies, Houston, TX). A 459-base pair (bp) segment of the EST clone was ligated into pRSETB (Invitrogen) and was sequenced to confirm the clone was in-frame with a polyhistidine tag. The recombinant plasmid was transfected into Escherichia coli strain BL21(DE3)pLysS (Invitrogen) to synthesize a bacterial fusion-protein containing the C-terminal 153 amino acids of the EST clone. After induction with 1 mM isopropyl-beta-d-thiogalactopyranose, the recombinant protein was expressed and isolated from bacterial lysate by Ni2+ affinity chromatography (ProBond Resin; Invitrogen). Bound protein was eluted with 500 mM NaCl and 20 mM NaPO4, pH 4.0, and was then dialyzed against 50 mM Tris-Cl, pH 8.0, 1 mM CaCl2, and 0.1% Tween-20, after which the histidine tag was cleaved by enterokinase digestion (EKMax; Invitrogen). Digest products were separated on a 15% SDS-PAGE gel, and protein was recovered by gel excision under visualization using SYPRO protein gel stain. Recombinant protein (17 kDa) was electroeluted from gel slices using an electroeluter (, Richmond, CA). Successful cleavage of the histidine tag was confirmed by immunoblotting with a mouse monoclonal antibody specific for histidine epitope tags (Anti-HisG; Invitrogen). Rabbits were immunized with the purified 17-kDa inversin miniprotein by Covance Research Products (Denver, PA). Antibodies were purified by immunoaffinity chromatography using the inversin mini protein coupled to activated beads (Affigel 10; ) according to the manufacturer's directions. Antibody was eluted in 100 mM glycine, pH 2.4, and the eluate was neutralized with Tris base. Protein Extraction | For total protein extracts, confluent S1 cells were washed with cold phosphate-buffered saline (PBS), scraped from the dish with a rubber-policeman, and pelleted by centrifugation at 850 x g for 5 min at 4C. Cell pellets were resuspended and homogenized using a ball-bearing homogenizer in extraction buffer (150 mM NaCl, 50 mM Tris-Cl, pH 8.0, 4 mM EDTA, 1 mM phenyl methyl sulfonyl fluoride [PMSF], and protease inhibitor cocktail [Sigma] with or without Triton X-100 at vol/vol of 1.0%). Cell lysates were centrifuged at 15,000 x g for 10 min at 4C. Supernatants were mixed with Laemmli buffer (2% SDS, 100 mM Tris-Cl, pH 6.8, 25% [vol/vol] glycerol, 10 mM dithiothreitol, 0.001% [wt/vol], and bromphenol blue) and boiled for 10 min . SDS-PAGE and Immunoblot Analysis | Proteins were separated on 7.5% SDS-PAGE gels and transferred to nitrocellulose membranes (Amersham Pharmacia Biotech, Piscataway, NJ; ). Membranes were blocked with 3% newborn calf serum (NCS) dissolved in Tris-buffered saline (TBS) containing 0.1% (vol/vol) Tween-20 and incubated for 45 min with primary antibodies diluted in 3% NCS in TBS. Membranes were washed in TBS-Tween 20, incubated with horseradish peroxidase-conjugated secondary antibodies in 3% NCS in TBS for 45 min and washed as above. Chemiluminescence was used for detection (, Rockford, IL). Mass Spectrometry Analysis of Inversin Immunoprecipitate | Total protein extracts from cultured S1 cells were immunoprecipitated with the inversin antibody in 1% (vol/vol) Triton X-100, 150 mM NaCl, 10 mM Tris-Cl, pH 8.0, 1 mM EDTA, 1 mM EGTA, 0.2 mM Na3VO4, 2 mM PMSF, and protease inhibitor cocktail, 1:100. Precipitates were recovered by protein A conjugated to magnetic beads (Dynal, Lake Success, NY). Beads were collected in a magnetic particle concentrator and were washed with three buffers featuring different ionic strengths (buffer I: 150 mM NaCl, 50 mM Tris-Cl, pH 8.0, and 1% [vol/vol] Triton X-100; buffer II: 500 mM NaCl and 50 mM Tris-Cl, pH 8.0; and buffer III: 50 mM Tris-Cl, pH 8.0). Precipitates were resuspended in Laemmli buffer and boiled for 10 min to elute complexes. Proteins were separated by 7.5% SDS-PAGE and visualized with Coomassie stain (Pierce). The 140-kDa band was excised, trypsin digested, and analyzed with a Finnigan (Thermoquest) LCQ mass spectrometer (Biochemistry Biotechnology Facility, Indiana University School of Medicine, Indianapolis, IN). A free-ware program was used to compute and map the cleavage fragments obtained from a virtual trypsin digest of the inv sequence (Client Paws; ProteoMetrics, New York, NY). Immunohistochemistry and Fluorescence Microscopy | Cells grown on filters were fixed in 4% paraformaldehyde in PBS for 10 min. The fixation reaction was quenched in 100 mM NH4Cl dissolved in PBS. Samples were incubated in blocking buffer (1% bovine serum albumin and 0.1% Triton X-100 in PBS) for 10 min before labeling. Cells were incubated in primary antibodies and rhodamine-phalloidin, followed by incubation with fluorescein- or Cy5-conjugated secondary antibodies (Jackson ImmunoResearch Laboratories) and washing in PBS. Nuclei were labeled with DAPI diluted in blocking buffer for 10 min. Filters were placed in 2% PFA in PBS, mounted with Mowoil (Calbiochem), and examined with a LSM 510 laser scanning microscope (Carl Zeiss North America, Thornwood, NY) equipped with a UV argon laser, a visible argon laser, and two helium-neon lasers. Images were collected sequentially and processed by Adobe Photoshop 5.5 (Adobe Systems, Mountain View, CA). Iodixanol Gradient Fractionation | Iodixanol subcellular fractionation was performed as previously described . Briefly, confluent S1 cells were homogenized in isotonic sucrose buffer [0.25 M sucrose, 90 mM KOAc, 2 mM Mg(OAc)2, 20 mM HEPES-KOH, pH 8.0, 2 mM PMSF, 0.5 mM 4-(2-aminoethyl)benzensulfonyl-fluoride, and protease inhibitor cocktail 1:100] by repeated passage through a ball-bearing homogenizer. Eight volumes of postnuclear supernatant were combined with 10 volumes of a 54% iodixinol solution that was prepared from a 60% stock solution [diluted with 1/10 vol of 900 mM KOAc, 20 mM Mg(OAc)2, and 200 mM HEPES-KOH, pH 8.0]. Samples were centrifuged at 354,000 x g for 1 h at 4C in a TLN-100 rotor (Beckman Instrument, Fullerton, CA). Fractions were collected and proteins were separated by 7.5% SDS-PAGE, transferred to a nitrocellulose membrane, and immunoblotted with antibodies specific for beta-catenin, pan-cadherin, and inversin. Extraction of Nuclear Protein Complexes | Nuclear proteins were extracted by stepwise lysis of cells to ensure minimal cross-contamination, as described earlier . Briefly, confluent S1 cells were washed with cold PBS, scraped in PBS from the dish with a rubber-policeman, and pelleted by centrifugation at 500 x g for 3 min at 4C. Using NE-PER nuclear extraction reagents (Pierce), cells were chemically lysed and nuclei were isolated after centrifugation (16,000 x g for 5 min) and resuspension per the manufacturer's instructions. Nuclear cell debris and DNA were removed by centrifugation (16,000 x g for 5 min) and nuclear proteins were mixed with Laemmli buffer, boiled for 10 min, and loaded on 7.5% SDS-PAGE gels. Extraction of Membrane Protein Complexes | Plasma membranes were purified by the technique of aqueous two-phase partition . Briefly, confluent S1 cells were washed with cold PBS, scraped from the dish with a rubber-policeman, and pelleted by centrifugation at 500 x g for 3 min at 4C. Using Mem-PER extraction reagents (Pierce), cells were chemically lysed and membrane proteins were solubilized. The mixture was incubated at 37C and centrifuged at 10,000 x g for 2 min to separate hydrophobic from hydrophilic proteins. The viscous phase containing the membrane protein fraction was collected and proteins were purified by precipitation with 10% trichloroacetic acid. Recovered proteins were mixed with Laemmli buffer, boiled for 10 min, and loaded on 7.5% SDS-PAGE gels. Ca2+ Switch | S1 cells were grown in normal-calcium (1.8 mM Ca2+) medium (NCM, see cell culture for formula). Low-calcium medium (LCM) was prepared by supplementing NCM with EGTA to a final concentration of 4 mM, as previously described . The pH was adjusted to its initial value with NaOH, and the medium was sterile filtered. S1 cells were washed with PBS, and NCM was replaced with LCM. At 45 and 90 min of incubation in LCM, cells were fixed and processed for immunofluorescence microscopy. Immunoprecipitation | Confluent S1 cells grown on collagen I-coated tissue culture dishes were washed with PBS and incubated on ice for 30 min in CSK buffer [50 mM NaCl, 300 mM sucrose, 10 mM PIPES, pH 6.8, 3 mM MgCl2, 0.5% (vol/vol) Triton X-100, 2 mM PMSF, 0.5 mM 4-(2-aminoethyl)benzensulfonylfluoride, and protease inhibitor cocktail 1:100; ]. Cells were scraped from dishes and insoluble material was removed by centrifugation at 10,000 x g for 10 min at 4C. Cell extracts were incubated with primary rabbit polyclonal antibodies for 1 h. Immune complexes were recovered by incubation with protein A-Sepharose (Amersham) for 1 h. Protein A-Sepharose beads were washed three times in extraction buffer before protein complexes were released by boiling in Laemmli buffer for 10 min. Precipitated proteins were separated by SDS-PAGE followed by immunoblotting. RESULTS : Characterization of the Polyclonal Inversin Antibody | A 17-kDa recombinant protein fragment was expressed as a fusion protein containing 153 amino acids from the C-terminal portion of the EST that lacks significant homology with any other protein in the available databases . Affinity-purified antibody detected three bands estimated at 140, 125, and 90 kDa on immunoblots containing proteins from confluent S1 cells extracted in 1% Triton X-100 (Figure A, left lane). Antibody recognition of both bands was competitively blocked when the affinity-purified antibody was preincubated with the immunizing inversin miniprotein (Figure A, right lane). Figure 1 | Characterization of inversin antibody. Characterization of inversin antibody. (A) Total protein extracted from confluent S1 cells (1% Triton X-100 buffer) was separated by 7.5% SDS-PAGE, transferred to membranes, and incubated with affinity-purified inversin antibody. Three bands were detected at 140, 125, and 90 kDa with anti-inversin alone, but no bands were detected when inversin antibody was preincubated with the immunizing recombinant protein. (B) Postnuclear homogenate from confluent S1 cells was immunoprecipitated with affinity-purified inversin antibody and was resolved by 7.5% SDS-PAGE. The 140-kDa band was excised, trypsin digested in situ, and analyzed by mass spectrometry. Table shows peak values measured from mass spectrum as compared with calculated fragment masses after trypsin digestion of inversin. Differences between these values ranged in the expected variability. Mass Spectrometry Confirmation of Antibody Specificity | Mass spectrometry analysis of the 140-kDa protein detected by the affinity-purified antibody confirmed that the antibody specifically bound to inversin. Proteins from S1 cell extracts were immunoprecipitated with inversion antibody and separated by SDS-PAGE. The 140-kDa band was digested with trypsin and the resultant fragments were analyzed by mass spectrometry (Figure C). Measured mass-to-charge ratios (m/z) of tryptic peptides significantly correlated with calculated fragment m/z ratios obtained from computed analysis of the inversin protein sequence. Masses of fragments were also compared with the masses of a peptide database calculated from NCBI's nr protein database . The inversin protein was identified as the top ranking of the candidate proteins based on its calculated posterior probability . The next top nine candidate proteins did not produce significant alignments to the inversin protein sequence . Similarly, the antibody to inversin precipitated two proteins of 165- and 90-kDa that matched inversin when analyzed by mass spectrometry (unpublished data; ). Subcellular Localization of Inversin: Laser Confocal Microscopy | Subconfluent (Figure , A, C, and E) and confluent (Figures , B, D, and F, and 3) S1 cells were grown on filters and stained with inversin antibody (Figures , A and B and 3B) and DAPI, a nucleic acid stain (Figures , C and D and 3A). To show the cellular architecture, the actin cytoskeleton was labeled with rhodamine phalloidin (Figures , E and F and 3C; ). In subconfluent cells, inversin was predominantly distributed in nuclei (compare Figure , A with C). Weak perinuclear and membrane staining was also observed (Figure A). No signal was detected when cells were incubated with the inversin antibody in the presence of the immunizing peptide or the secondary antibody alone (Figure A, insets). In separate experiments, perinuclear inversin staining colocalized with anti-KDEL antibody staining, suggesting that inversin is in the rough endoplasmic reticulum (unpublished data). When incubated with anti-inversin, confluent cells showed distinct labeling of cell membranes (Figures B and B) that was only weakly apparent (Figure A, arrowheads) in subconfluent cells with adjacent cell borders (Figure E). A vertical image section (X-Z axis) of confluent S1 cells shows a lateral distribution of inversin (Figure B). Similar labeling by inversin antibody was also observed in other cultured renal epithelial cell lines such as MDCK cells or cells derived from the distal convoluted tubule (unpublished data; ), indicating that various cell lines carry this protein in a membrane-associated fashion. Figure 2 | Confocal microscopy of subconfluent (A, C, and E) and confluent (B, D, and F) S1 cells fixed in paraformaldehyde and triple-labeled with inversin antibody (A and B), DAPI (C and D), and phalloidin-rhodamine (E and F). Confocal microscopy of subconfluent (A, C, and E) and confluent (B, D, and F) S1 cells fixed in paraformaldehyde and triple-labeled with inversin antibody (A and B), DAPI (C and D), and phalloidin-rhodamine (E and F). In both subconfluent (A) and confluent (B) S1 cells, the inversin antibody stained nuclei and less intensely the perinuclear compartment. Cell membranes appeared uniformly stained in confluent cells (B), but were focally stained in subconfluent cells at early cell-cell contacts (A, arrowheads). Staining seen with inversin antibody (A and B) was absent in cells stained with inversin antibody preincubated with immunizing protein, and secondary antibody alone (A, insets). Figure 3 | Vertical sections (X-Z axis) of inversin staining. Vertical sections (X-Z axis) of inversin staining. Confluent S1 cells were grown on filters, fixed in paraformaldehyde, and triple-labeled with DAPI (A), inversin antibody (B), and phalloidin-rhodamine (C). The vertical sections (X-Z axis) demonstrate distribution of inversin to membranes of cell-cell contacts, but not to apical or basal plasma membranes. (D) overlay of all three channels. Subcellular Distribution of Inversin: Iodixanol Gradient Fractionation | To identify which inversin isoforms (i.e., 140-, 125-, or 90-kDa bands) are responsible for the membrane staining patterns, S1 cell extracts were fractionated by detergent-free iodixanol density gradient centrifugation . In iodixanol gradients, low-density plasma membrane and cytoplasmic particles (<1.1 g/ml) are recovered in the top layer of the gradient (fraction 1), while cytosolic proteins exhibiting a higher density (1.26 g/ml) migrate to the bottom of the gradient during centrifugation (fraction 12). Postnuclear supernatants of confluent S1 cells were fractionated by iodixanol gradient, and fractions were subsequently analyzed by immunoblotting with antibodies to inversin, pan-cadherin, and beta-catenin. Figure A shows the immunoblot analysis of fractions collected from the iodixanol gradient. The inversin antibody detected a band at 140 kDa in all fractions, with the greatest intensity in fractions rich in cytosolic proteins. The weakest 140-kDa bands were in the fractions rich in plasma membrane particles. Additionally, in fraction 1, the inversin antibody detected a 125-kDa band that was not detected in any other fraction. The 90-kDa inversin signal was detected in all fractions with progressively increasing signal intensity from fraction 1 to fraction 12. When the immunoblots were subjected to extended exposure times, a 165-kDa band was detected in the cytoplasmic-enriched fractions 11 and 12 (unpublished data). Pan-cadherin and beta-catenin antibodies detected bands in fractions 1 and 2 containing low-density plasma membrane particles, in agreement with previous findings . Taken together, these data suggest that the 125-kDa band is only found in the fraction that is enriched for plasma membranes, whereas the 90-kDa isoform is found in all fractions. Figure 4 | Subcellular distribution of inversin. Subcellular distribution of inversin. (A) Iodixinol subcellular fractionations were collected from the lightest (top fraction 1) to the heaviest (bottom fraction 12) of the gradient. Equivalent fraction volumes were resolved by 7.5% SDS-PAGE and immunoblotted with antibodies to inversin, pan-cadherin, or beta-catenin. Bands (140- and 90-kDa) were detected with anti-inversin in all fractions, but a 125-kDa band was detected only in membrane fraction 1. The bands detected by anti-pan-cadherin and anti-beta-catenin were restricted to the lightest fractions. The accompanying graph expresses phosphorimager values for each band as a percentage of the total volume of all fractions measured for each antibody. (B) Nuclear and membrane protein extracts from confluent S1 cells were separated by 7.5% SDS-PAGE and were immunoblotted with anti-inversin. The inversin antibody detected only one band of 125 kDa in the membrane protein extract and two bands of 140- and 90- kDa in the nuclear protein extracts. (C) Total protein was extracted from confluent S1 cells with and without 1% Triton X-100 followed by inversin immunoblot analysis. Anti-inversin detected bands at 140 and 90 kDa in both extracts, but only in the presence of Triton X-100 did the inversin antibody detect a band at 125 kDa. Immunoblot Analysis | To confirm the enrichment of the 125-kDa inversin protein in cell membranes, immunoblots were performed on membrane protein complexes extracted from confluent S1 cells using aqueous two-phase partitioning. The inversin antibody detected a single band at 125 kDa (Figure B, right lane) migrating at the same molecular weight as the 125-kDa band in fraction 1 of the iodixanol gradient (Figure A). Nuclear protein complexes were isolated from confluent S1 cells, separated by SDS-PAGE. and transferred to membranes for immunoblot analysis. Successful extraction of nuclear proteins was confirmed on immunoblots by monoclonal antibody detection of histone H1, an exclusive nuclear protein (unpublished data). Inversin antibody detected bands at 90 and 140 kDa in immunoblots of nuclear extracts (Figure B, left lane), but the 125-kDa band was not detectable. To confirm that the 125-kDa inversin protein is restricted to detergent-soluble membrane fractions, protein was extracted from confluent S1 cells in a solution with or without 1.0% Triton X-100. The 140- and 90-kDa bands were detected on immunoblots containing proteins extracted without detergent (Figure C, left lane). When S1 cell proteins were extracted with 1.0% Triton X-100, bands were detected at 140, 125, and 90 kDa (Figure C, right lane). Inversin Forms a Complex with Catenins and N-Cadherin | As described above, inversin localizes to the lateral plasma membrane at regions of cell-cell contacts in confluent S1 cells. To identify proteins that potentially interact with inversin, we immunoprecipitated S1 cell homogenates with the inversin antibody and probed the precipitates with mouse monoclonal antibodies to the cell adhesion proteins alpha-, beta-, and gamma-catenins, pan-, E-, and N-cadherins, vinculin, and beta1-integrin. Bands of the expected molecular weights were detected with antibodies to alpha-, beta-, and gamma-catenins, and pan- and N-cadherin, but not with antibodies to E-cadherin, vinculin, or beta1-integrin despite extended exposure times (Figure A). Figure 5 | (A) Homogenates from confluent S1 cells were immunoprecipitated with inversin antibody and precipitates were immunoblotted with a panel of antibodies. (A) Homogenates from confluent S1 cells were immunoprecipitated with inversin antibody and precipitates were immunoblotted with a panel of antibodies. Arrowheads indicate expected molecular weights. Bands were detected for alpha-, beta-, and gamma-catenin, and pan-, and N-cadherin, but no bands were detected for E-cadherin, B1-integrin, and vinculin despite long film exposure. (B) Confluent S1 cell homogenates were immunoprecipitated with anti-beta-catenin (lane 1) and anti-N-cadherin (lanes 2 and 3), and precipitates were immunoblotted with anti-inversin or anti-beta-catenin. The inversin antibody detected 125- and 90-kDa bands in both the beta-catenin precipitate (left panel) and the N-cadherin precipitate (middle panel). The N-cadherin precipitate that was immunoblotted for beta-catenin detected a 92-kDa band (right panel). (C) Iodixanol fractions 1 and 12 were collected as for Figure A, immunoprecipitated with inversin antibody, and immunoblotted with antibodies to beta-catenin and N-cadherin. Inversin precipitates from fraction 1 contained beta-catenin and N-cadherin (lanes 1 and 2), but no bands were detected in fraction 12 (lanes 3 and 4). Protein extracts in A, B, and C were separated by 7.5% SDS-PAGE and transferred to nitrocellulose membranes. Inverse immunoprecipitations were performed on S1 cell homogenates. Complexes that were precipitated using polyclonal rabbit antibodies to beta-catenin or N-cadherin contained inversin as detected in immunoblots using the inversin antibody (Figure B). Both beta-catenin and N-cadherin precipitates contained the 125- and 90-kDa inversin isoforms. Protein complexes precipitated by anti-N-cadherin contained beta-catenin, as shown in Figure B. To confirm that protein complexes containing inversin, beta-catenin, and N-cadherin are membrane associated in S1 cells, we analyzed membrane and cytosolic fractions obtained from the iodixanol density gradient as described for Figure A. Protein complexes from fractions enriched in membranes (fraction 1) or cytosol (fraction 12) were immunoprecipitated with anti-inversin followed by SDS-PAGE and immunoblot analysis with monoclonal antibodies to beta-catenin and N-cadherin. Coprecipitation of inversin with beta-catenin and N-cadherin was found only in the membrane fraction 1, but not in the cytosolic fraction 12 (Figure C). Codistribution of Inversin and the Cadherin/Catenin Complex | Confluent S1 cells were double-labeled with antibodies to inversin and beta-catenin or inversin and N-cadherin. Both inversin and alpha-catenin were located at the lateral cell membranes and in the nucleus (Figures , A and B). Most inversin staining colocalized with beta-catenin at the plasma membrane, consistent with our findings that these proteins are part of a complex (Figure C). In the nucleus, the antibody labeling of beta-catenin and inversin showed partial colocalization in S1 cells (Figure C) and was weak to absent in MDCK cells (unpublished data). Anti-N-cadherin strongly stained lateral cell membranes (Figure E). Figure F shows that membrane-associated inversin and N-cadherin colocalized exclusively at the plasma membrane (Figure F). No staining was observed when cells were labeled with secondary antibodies alone, confirming the specificity of the primary antibodies used (insets in Figure , A, B, D, and E). Figure 6 | Colabeling of inversin with beta-catenin or N-cadherin. Colabeling of inversin with beta-catenin or N-cadherin. Confluent S1 cells were double-labeled with antibodies to inversin (A) and beta-catenin (B) or inversin (D) and N-cadherin (E) and were analyzed by confocal microscopy. Inversin colocalized with beta-catenin at cell membranes, but there was partial overlap in nuclei (C, yellow overlap). Inversin colocalized with N-cadherin only at cell membranes. Inversin Membrane Assembly Is Calcium Dependent | Inversin's membrane distribution may be dependent on cell-cell adhesions. To study whether this distribution is calcium dependent, confluent S1 cells were switched from medium containing 1.8 mM calcium to media containing 1.8 mM calcium plus EGTA. Cells fixed at various time points (0, 45, and 90 min) were labeled with anti-inversin, anti-beta-catenin, and fluorescein isothiocyanate-conjugated phalloidin (Figure , A-I) or anti-inversin, anti-N-cadherin, and fluorescein isothiocyanate-conjugated phalloidin (Figure , J-R). In 1.8 mM calcium medium (0 min), cells were confluent, and staining for F-actin was continuous at regions of cell-cell contacts (Figure , G and P). Similar to the results shown in Figure , inversin and beta-catenin localized to the cell membrane and nuclei (Figure , A, D, and J). Under these conditions, anti-N-cadherin also labeled cell membranes (Figure M). Incubating the cells in calcium-chelated media induced changes in cell shape, cell adhesion, and redistribution of analyzed proteins. After 45 min in calcium-chelated media, cell contacts were discontinuous, and phalloidin staining illustrated a retraction of the F-actin belt (Figure , H and Q). Membrane labeling of inversin, beta-catenin, and N-cadherin was dramatically weakened, whereas nuclear staining of inversin and beta-catenin appeared unchanged (Figure , B, E, K, and N). After 90 min in LCM, cells were spherical and contacts with neighboring cells were minimal or absent (Figure , I and R). Staining of inversin and beta-catenin was almost exclusively confined to the nuclear compartment (Figure , C, F, and L). Only a few membrane segments showed residual staining of beta-catenin and N-cadherin (Figure , F and O). These results demonstrate that membrane assembly of inversin is calcium dependent. Changes in membrane assembly of inversin occurred over the same time course as the changes in N-cadherin and beta-catenin membrane assembly (unpublished data). Low calcium-induced redistribution of inversin was almost complete after 45 min. Similarly, complete redistribution of N-cadherin and beta-catenin occurred at the 45-min time point. Notably, intranuclear inversin staining was unaffected by changes in extracellular calcium. Figure 7 | Calcium depletion in confluent S1 cells by confocal microscopy. Calcium depletion in confluent S1 cells by confocal microscopy. Confluent S1 cells were calcium depleted by incubating with medium containing 4 mM EGTA. At time 0, 45, and 90 min, cells were triple labeled with anti-inversin (A-C), anti-beta-catenin (D-F), and phalloidin (G-I) or anti-inversin (J-L), anti-N-cadherin (M-O), and phalloidin (P-R). At 45 and 90 min of calcium depletion, cells progressively lost cell-cell contacts, as displayed by staining of the F-actin cytoskeleton (H, I, Q, and R). Calcium depletion also led to diminished staining of inversin (B, C, K and L), beta-catenin (E and F), and N-cadherin (N and O) from cell membranes. Nuclear staining of inversin and beta-catenin remained unchanged under low calcium conditions (C, F, and L). DISCUSSION : The inv gene was originally identified in the OVE210 model of reversal of embryonic turning and PKD . The activity of the inv gene product, inversin, is not known, but the defects seen in the inv/inv mouse suggest that inv plays a crucial role in the establishment of the left-right axis and maturation of epithelial structures such as renal tubules. Using an inversin-specific polyclonal antibody, this study identifies at least three inversin proteins of 140-, 125-, and 90- kDa. The existence of multiple isoforms is supported by the identification of alternatively spliced inv transcripts. described alternative splicing in exon 13 of the full-length inv sequence predicting isoforms of 99, 104, and 118 kDa. Using reverse transcriptase-polymerase chain reaction, we found splice variants lacking exon 4 (accounting for 6.2 kDa) or exon 10 (4 kDa) in mouse kidney and exon 11 (7.9 kDa) in MDCK cells (canine; unpublished data). Using the NCBI database, we identified 16 ESTs with 90% homology to inv, but when completely sequenced, none of these encoded the full-length (1062 amino acids) sequence. Fifteen of the ESTs had limited homology to exon 16 (unpublished data), but one EST was a splice variant that encoded a 76-kDa in-frame inversin protein that we used to generate antigen for our inversin antibody. We used web-based software to predict posttranslational modifications of inversin and we found potential phosphorylation sites (50 serine, 11 threonine, and 6 tyrosine; ), as well as several type O-glycosylation sites (11 serine and 2 threonine; ). Some of these posttranslational modifications may account for the difference between the predicted size of the 118- and 140-kDa bands detected by the inversin antibody. Potential Cellular Function of Inversin | Of the two lower molecular-weight inversin isoforms that precipitate with beta-catenin and N-cadherin, the 125-kDa band was most enriched in the cell membrane fractions. Like the cadherin/catenin complex , membrane-associated inversin is calcium dependent. Analysis of the inversin protein sequence did not reveal any known transmembrane sequence motifs; therefore, the effect of calcium depletion on inversin redistribution may be mediated via N-cadherin or beta-catenin. These findings suggest that the 125-kDa inversin may play a role in regulating the molecular architecture of cell-cell junctions. The inversin proteins responsible for the nuclear and perinuclear immunofluorescence appear to be the 140- and 90-kDa isoforms. The import of inversin into the nucleus likely involves the recognition of a nuclear localization signal sequence (NLS; ). The full-length inversin sequence exhibits one classical type of NLS (KHRR at aa 735; ) and two bipartite NLSs (RKDAAAKKREEENKRKE at aa 589 and KRQDRAARPRGASQKRR at aa 782; ). Rheinhardt's method for cytoplasmic/nuclear discrimination also predicted a nuclear localization of inversin , providing further evidence for a nuclear function of inversin. Based on PROSITE database analysis, the inversin sequence does not exhibit sequence motifs that are involved in DNA binding . Of the two inversin proteins that were enriched in nuclear fractions, only the 90-kDa protein complexed with beta-catenin, the latter a protein involved in transcription regulation . Therefore, a nuclear inversin protein may modulate gene transcription via interaction with transcription factors such as beta-catenin. Potential Role of Inversin in Determination of the Left-Right Axis | Mice lacking a functional inv gene exhibit defects in left-right patterning , suggesting a role of inversin in the determination of the left-right axis. In the earliest stages of asymmetrical development, a nodal flow generated by motile cilia is postulated to initiate expression of genes, such as Nodal, that regulate embryonic turning . Although little is known about the function of inversin, a reduced nodal flow was observed in the inv/inv mouse and has been proposed to be associated with the reversal of embryonic turning in this model . We found inversin localization at nuclei and at basolateral membranes. When MDCK cells were costained for inversin and tubulin, we did not observe inversin staining in monocilia (unpublished data). However, inversin precipitated with proteins that are involved in axis development, i.e., N-cadherin and beta-catenin. N-cadherin (neural cadherin), a classical or type I cadherin, is expressed in a developmental manner , regulates migration of cortical and neural crest cells , and plays a role in the embryonic development of the kidney . However, N-cadherin is also involved in establishment of embryonic left-right asymmetry. found that chicken embryos treated with anti-N-cadherin exhibit a randomization of left-right asymmetry. Mechanistically, N-cadherin, which is localized to the right side of the node, may restrict activation of Nodal , which in turn controls embryonic turning . A defective inv gene product may adversely impact N-cadherin function, resulting in defects of establishing the left-right axis. Inversin also interacts with beta-catenin, a protein that plays a crucial role in the Wnt/beta-catenin signaling pathway . Nuclear beta-catenin complexes with LEF/TCF transcription factors and subsequently activates LEF/TCF target genes , including the promoter of Nodal . Hence, the defect in the left-right axis in the inv/inv mouse could also result from deregulated beta-catenin signaling. However, altered left-right asymmetry has not been described in transgenic mice expressing an activated mutant of the beta-catenin gene . Potential Role of Inversin in Kidney Development | Renal cyst formation has been proposed to result from the disruption of a multicomponent membrane-spanning polycystin complex . Several proteins have been identified that participate in this complex, including polycystin-1, polycystin-2, the catenins, and E-cadherin . Defects in these proteins and deregulation of the linked Wnt/beta-catenin signaling cascade may also play a key role in the cystogenic pathway . Recently, have shown that transgenic mice expressing a mutant form of beta-catenin develop polycystic kidneys . Similarly, overexpression of c-myc, which is a target gene of beta-catenin signaling , results in a cystic phenotype in transgenic mice . In this study, we found inversin coimmunoprecipitating with molecules that complex with polycystin. As with other junctional proteins , the 125-kDa membrane-enriched inversin may modulate beta-catenin function. The finding of inversin and beta-catenin colocalization in nuclei provides further evidence of a partnership of these two proteins in pathways that may include cystogenesis. Backmatter: PMID- 12221119 TI - Modulation of Cellular Cholesterol Transport and Homeostasis by Rab11 AB - To analyze the contribution of vesicular trafficking pathways in cellular cholesterol transport we examined the effects of selected endosomal Rab proteins on cholesterol distribution by filipin staining. Transient overexpression of Rab11 resulted in prominent accumulation of free cholesterol in Rab11-positive organelles that sequestered transferrin receptors and internalized transferrin. Sphingolipids were selectively redistributed as pyrene-sphingomyelin and sulfatide cosequestered with Rab11-positive endosomes, whereas globotriaosyl ceramide and GM2 ganglioside did not. Rab11 overexpression did not perturb the transport of 1,1'-dioctadecyl-3,3,3',3'-tetramethyl-indocarbocyanine-perchlorate --labeled low-density lipoprotein (LDL) to late endosomes or the Niemann-Pick type C1 (NPC1)-induced late endosomal cholesterol clearance in NPC patient cells. However, Rab11 overexpression inhibited cellular cholesterol esterification in an LDL-independent manner. This effect could be overcome by introducing cholesterol to the plasma membrane by using cyclodextrin as a carrier. These results suggest that in Rab11-overexpressing cells, deposition of cholesterol in recycling endosomes results in its impaired esterification, presumably due to defective recycling of cholesterol to the plasma membrane. The findings point to the importance of the recycling endosomes in regulating cholesterol and sphingolipid trafficking and cellular cholesterol homeostasis. Keywords: INTRODUCTION : Cholesterol is an essential constituent of membranes in mammalian cells and a precursor for steroid hormone and bile acid synthesis. Cellular cholesterol levels are tightly regulated at the level of synthesis, esterification, and exchange with plasma lipoproteins . The route of low-density lipoprotein (LDL)-cholesterol uptake is hitherto the best characterized cellular cholesterol-trafficking pathway. The role of the LDL receptor in LDL internalization, the breakdown of the lipoprotein particle in acidic organelles, and the homeostatic mechanisms regulating the LDL-receptor levels have been unraveled . However, the contribution of other endocytic routes on cholesterol transport and balance and their interplay with the LDL-receptor route are so far poorly understood at the molecular level. The endocytic organelles have been mainly defined based on the flow of different cargo molecules to early, recycling, and late compartments. Internalized molecules are initially transported to early endosomes (also termed sorting endosomes) from where they can be delivered to late endosomes and lysosomes for degradation or become recycled to the plasma membrane either directly or via a recycling endosomal membrane system . Recycling endosomes are considered to be cholesterol enriched . The cholesterol content of the early or late endocytic membranes has not been determined, but late endocytic circuits are considered to be important for the regulation of the cellular free cholesterol content. This is highlighted in the late endosomal/lysosomal cholesterol storage disorder Niemann-Pick type C (NPC) disease. In this disease, cholesterol as well as other lipids and proteins accumulate in late endocytic organelles due to mutations in either of two recently cloned gene products, NPC1 or NPC2/HE1 . Consequently, cholesterol homeostatic responses in the endoplasmic reticulum fail, as manifested by defective cholesterol esterification and inappropriately high cholesterol synthesis . NPC1 is a polytopic membrane protein of late endocytic membranes, whereas NPC2 is a cholesterol-binding soluble protein that is also targeted to the late endocytic organelles. The precise functions and trafficking itineraries of both NPC1 and NPC2 remain to be elucidated. We have recently reported that the clearance of lysosomal cholesterol deposits can be inhibited by Rab-GDP dissociation inhibitor . This protein controls multiple vesicular transport pathways by sequestering GDP-bound (inactive) forms of the small GTPases of the Rab family in the cytoplasm. Rab proteins and their effectors coordinate consecutive stages of membrane transport, such as vesicle formation, movement, and tethering of vesicles to their target compartment . To gain further insight into the Rab-dependent endosomal cholesterol-trafficking mechanisms, we screened the potential contribution of selected endosomal Rab proteins 1) morphologically by using filipin staining and 2) biochemically by measuring cholesterol esterification. Based on the results obtained, our further analyses focused on the role of Rab11 in controlling endocytic cholesterol routing. We provide evidence that Rab11-dependent membrane trafficking modulates endosomal cholesterol levels independent of LDL uptake and serves as an important regulator of cellular cholesterol balance. MATERIALS AND METHODS : Antibodies and Reagents | Mouse monoclonal anti-transferrin receptor (TfR) and rabbit polyclonal anti-Rab11 antibodies were from Zymed Laboratories (South San Francisco, CA), mouse monoclonal anti-lysosome-associated membrane protein (lamp) 1 antibody was from Developmental Studies Hybridoma Bank (University of Iowa, Iowa City, IA), and mouse monoclonal anti-LDL receptor antibodies (C7) were from American Type Culture Collection (Manassas, VA). IgG antibodies against sulfatide and IgM antibodies against globotriaosyl ceramide and GM2 were generous gifts from Jan-Eric Mansson (Sahlgrenska University Hospital, Molndal, Sweden); anti-lysobisphosphatidic acid (LBPA) antibody was from Jean Gruenberg (University of Geneva, Geneva, Switzerland), and anti-HE1/NPC2 antibody was from Naomichi Okamura (University of Tsukuba, Tusbuba, Japan). Anti-NPC1 antibody has been described previously . Fluorescein isothiocyanate (FITC)- and tetramethylrhodamine B isothiocyanate-conjugated anti-IgG secondary antibodies were from Immunotech (Marseille, France), and Cy3-conjugated streptavidin and tetramethylrhodamine B isothiocyanate-conjugated secondary antibodies against mouse IgM were from Jackson Immunoresearch Laboratories (West Grove, PA). FuGENE6 transfection reagent was from Roche Applied Science (Indianapolis, IN). Filipin, FITC lentil lectin, methyl-beta-cyclodextrin (mbeta-CD), fatty-acid free bovine serum albumin (BSA), chymostatin, leupeptin, antipain, and pepstatin A, cell culture media, cholesterol, and other unlabeled lipids were from Sigma-Aldrich (St. Louis, MO). [9,10(n)-3H]Oleic acid (specific activity 7.5 Ci/mmol), cholesteryl[1-14C]oleate (specific activity 57 mCi/mmol), [4-14C]cholesterol (specific activity 55 mCi/mmol), and [3H]acetic acid (specific activity 9 Ci/mmol) were from Amersham Biosciences (Piscataway, NJ). Alexa 568-conjugated anti-IgG secondary antibodies, Texas Red transferrin, 1,1'-dioctadecyl-3,3,3',3'-tetramethyl-indocarbocyanine-perchlorate --labeled low-density lipoproteins (DiI-LDLs), and Alexa 594-conjugated cholera toxin subunit B (CTxB) were from Molecular Probes (Eugene, OR). Pyrenyldecanoylsphingomyelin (Pyr10SM) was prepared as described previously . gamma-Cyclodextrin (gamma-CD) was from Cyclodextrin Technologies Development (High Springs, FL), and 3-beta-[2-(diethylamino)ethoxy]-androst-5-en-17-one (U18666A) was from Upjohn (Puurs, Belgium). Biotin-2xFYVE was a generous gift from Harald Stenmark. Cell Culture and Transfections | COS-1 cells were cultured in DMEM containing 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 mug/ml streptomycin. F92-99 control fibroblasts and 93.41 NPC fibroblasts were obtained and cultured as described previously . Cells were transfected using FuGENE6 according to the manufacturer's instructions and used for experiments at 40 --48 h (COS-1 cells) or 65 --72 h (human primary fibroblasts) posttransfection. DNA Constructs | pEGFP-C3 was from (Palo Alto, CA). Green fluorescent protein (GFP)-wtRab5, GFP-wtRab6, GFP-wtRab7, and GFP-wtRab11 were as described previously . GFP-Rab5Q79L, GFP-Rab11Q70L, and GFP-Rab11S25N were generous gifts from Marino Zerial (Max Planck Institute for Molecular Cell Biology and Genetics, Dresden, Germany). The human TfR cDNA was in pCDNA3.1. Human LDL-receptor cDNA in pCB6 and human NPC1 in pCR3.1 have been described previously . Cholesterol Esterification Assays | Cells on 12-well plates were transfected, and the following day the medium was replaced with fresh culture medium. Alternatively, to deplete cholesterol the cells were incubated with medium containing 5% lipoprotein-deficient serum (LPDS), prepared as in for 24 h before labeling. To analyze esterification in the presence of LDL, cells grown in culture medium were washed with phosphate-buffered saline (PBS) and labeled with [3H]oleic acid (5 muCi/ml) in serum-free, 2% defatted BSA medium supplemented with 50 mug/ml LDL for 4 h. After labeling, the cells were washed with ice-cold PBS on ice and scraped into PBS, harvested by centrifugation, and resuspended in 2% NaCl. Aliquots were removed for determining the protein concentration. A chromatography recovery standard was added (2.5 --5 nCi of [14C]cholesteryl oleate) and the lipids extracted with 2 ml of methanol and 1 ml of chloroform as described previously . After subsequent centrifugation, 1/10 of the supernatant was removed for liquid scintillation counting to determine the [14C]cholesteryl oleate radioactivity. The extracted lipids were separated by thin layer chromatography on silica gel plates by using hexane/diethyl ether/acetic acid (80:20:1) as the solvent. The cholesteryl ester band was determined based on the comigration of a cholesteryl ester standard, scraped, and 3H and 14C radioactivity measured by liquid scintillation counting. The results were corrected for the volume and procedural losses based on the recovery of 14C radioactivity and plotted against the total amount of protein in the sample. The protein concentration was determined according to . To analyze esterification in delipidated cells, cells grown in 5% LPDS medium for 24 h were washed with PBS and labeled with [3H]oleic acid (5 muCi/ml) in serum-free, 2% defatted BSA medium for 4 h. Lipids were analyzed as described above. To analyze esterification in cells loaded with cholesterol/mbeta-CD-complex the cells were initially delipidated as described above and labeled with [3H]oleic acid (5 muCi/ml) in serum-free, 2% defatted BSA medium for 4 h. During the labeling, cholesterol/mbeta-CD-complex prepared as described previously was added at 50 mug/ml concentration of cholesterol at staggered time points to yield the final loading times indicated. The basal rate of esterification as determined by samples labeled without cholesterol/mbeta-CD-complex was subtracted from the values at all time points. Western Blot Analysis | Cells were harvested in 1% Nonidet-P40 in PBS supplemented with protease inhibitors (chymostatin, leupeptin, antipain, and pepstatin, at 25 mug/ml each). Aliquots of the cell lysate (20 mug of protein) were separated by SDS-PAGE, and the proteins were transferred to Hybond-C Extra membrane (Amersham Biosciences). After blocking with 5% nonfat milk in Tris-buffered saline containing 0.2% Tween 20 for 1 h at 37C, the membrane was incubated overnight at 4C with rabbit polyclonal anti-GFP antibodies . The membrane was then washed and incubated with horseradish peroxidase-conjugated anti-IgG secondary antibodies (, Hercules, CA). The staining was visualized using enhanced chemiluminescence Western blotting detection reagent (Amersham Biosciences). Immunocytochemistry | The cells were fixed with 4% paraformaldehyde for 20 min and quenched with 50 mM NH4Cl for 10 min. Cells were permeabilized either with 0.1% Triton X-100 for 4 min and blocked with 10% FBS for 30 min at 37C, or alternatively the blocking solution was supplemented with 0.05% filipin to permeabilize the cells. The primary antibodies were diluted in 5% FBS and incubated for 1 h at 37C or overnight at 4C and the secondary antibodies for 30 min at 37C. For filipin staining only, the fixed and quenched cells were incubated with 0.05% filipin in PBS for 15 min and washed with PBS. The coverslips were mounted with Mowiol and the antifading reagent 1,4 diazobicyclo-(2.2.2) octane and viewed with TCS SP confocal microscope (, Deerfield, IL), Axiophot photomicroscope (Carl Zeiss, Thornwood, NY), or IX70 inverted microscope (, Tokyo, Japan) equipped with a Polychrome IV monochromator (TILL Photonics, Eugene, OR) with appropriate filters. Labeling with Texas Red Transferrin | Cells double transfected with the indicated cDNAs and TfR were starved in serum-free culture medium for 1 h at 37C. Cells were then incubated with 50 mug/ml Texas Red transferrin in Eagle's minimum essential medium supplemented with 0.2% BSA, 0.35 g/l NaHCO3, 100 U/ml penicillin, 100 mug/ml streptomycin, and 10 mM HEPES, pH 7.4, for 30 min on ice at 4C. After labeling the cells were incubated in serum-free culture medium supplemented with 0.2% BSA for 30 min at 37C, fixed, and processed for immunofluorescence microscopy as described above. Labeling with Biotin-2xFYVE | Cells fixed and quenched as described above were blocked and permeabilized with 10% FBS supplemented with 0.05% filipin for 30 min. Cells were then incubated with 50 mug/ml biotin-2xFYVE in 10% FBS for 30 min at room temperature, washed 3 x 5 min with PBS, and further incubated with 1 mug/ml Cy3-conjugated streptavidin in 10% FBS for 30 min, washed 3 x 5 min with PBS, and mounted. Labeling with Alexa 594-conjugated CTxB | Cells were incubated with 2 mug/ml Alexa 594-conjugated CTxB in Eagle's minimum essential medium supplemented with 0.01% BSA, 0.35 g/l NaHCO3, 100 U/ml penicillin, 100 mug/ml streptomycin, and 10 mM HEPES, pH 7.4, for 1 h on ice at 4C. After labeling the cells were incubated in serum-free culture medium supplemented with 0.01% BSA for 2 h at 37C and fixed. Labeling with Pyr10SM | To prepare the Pyr10SM/gamma-CD-complex the lipid was dried under argon and desiccated in the vacuum for 30 min. gamma-CD (100 mM in PBS) was added on the lipid film in a molar ratio of 1000:1, and the suspension was sonicated 3 x 2 min . Cells were labeled with Pyr10SM/gamma-CD-complex at 10 nmol/ml concentration of the lipid for 10 min at 37C and incubated in culture medium for 2 h at 37C before fixation. The fluorescence was excited at 345 nm and visualized at 480/80 nm. The degradation rate of Pyr10SM in COS-1 cells was determined by high-performance liquid chromatography using on-line fluorescence detection . Electron Microscopy | Cells were fixed with 4% paraformaldehyde in 0.25 M HEPES pH 7.4, scraped, and infiltrated in 1.75 M sucrose in 0.25 M HEPES containing 4% paraformaldehyde for 48 h at 4C. Droplets of cells in sucrose were mounted on pins and frozen in liquid nitrogen. Ultrathin cryosections were labeled with polyclonal rabbit anti-GFP antibody (a generous gift from Graham Warren (Yale University School of Medicine, New Haven, CT) and David Shima (Imperial Cancer Research Fund, London, UK), followed by protein A coupled to 10-nm gold particles. Sections were examined and photographed at 80 kV with a 1200 EX electron microscope (, Tokyo, Japan). Labeling with DiI-LDL | Cells were incubated in medium containing 5% LPDS for 24 h then labeled with 10 mug/ml DiI-LDL in serum-free medium for 15 min at 37C. After washing with PBS, the cells were either fixed or further incubated in serum-free medium for 2 h at 37C before fixation. Cholesterol Biosynthesis | Cells on six-well plates were transfected and at 6 h posttranfection the culture medium was changed to medium supplemented with 5% LPDS and [14C]cholesterol (100 nCi/ml) for 41 h. The cells were washed with PBS, pulse labeled with [3H]acetic acid (250 muCi/ml) in serum-free medium for 15 min at 37C, and chased in serum-free medium supplemented with 10 muM lovastatin and 25 mM mevalonate for 90 min at 37C. The cells were washed with ice-cold PBS on ice, scraped into PBS, harvested by centrifugation, and resuspended in 2% NaCl. Aliquots were removed for determining the protein concentration. Lipids were extracted as described above, separated by thin layer chromatography, and analyzed by high-performance liquid chromatography as described previously . Nascent cholesterol was quantified as 3H radioactivity in the cholesterol peak, corrected for the volume and procedural losses based on the recovery of 14C radioactivity, and plotted against the amount of protein in the sample. RESULTS : Redistribution of Free Cholesterol and Inhibition of Cholesterol Esterification upon Overexpression of Endosomal Rab Proteins | COS-1 cells were transiently transfected for 40 --48 h with GFP-fusions of Rab proteins reported to regulate early, late, or recycling endocytic transport events, represented by Rab5, 7, and 11, respectively. Rab5 promotes homotypic fusion of early endosomes . Overexpression of the late endosomal Rab7, on the other hand, has been shown to affect early-to-late endosomal transport and lysosome biogenesis , whereas Rab11 regulates the function of the recycling endosomes . Rab6 that is involved in retrograde trafficking in the Golgi , and soluble GFP were used as controls. To visualize the distribution of free cholesterol, the cells were fixed and stained with the fluorescent sterol-binding antibiotic filipin. In COS cells, the perinuclear area of the cell is strongly filipin positive. In addition, the plasma membrane and punctate peripheral structures are stained, albeit at lower intensity . The prominent perinuclear filipin staining colocalizes with a Golgi marker lentil lectin but several filipin-positive punctae also colocalize with lysosomal or early endosomal markers as visualized by antibodies against lysosomal membrane protein lamp1, or labeling with peptide 2xFYVE that binds the early endosomal phosphatidylinositol-(3)-phosphate (PI-3-P) . The perinuclear aspect of endogenous Rab11 staining also partially overlaps with that of filipin. However, the small peripheral Rab11-positive dots are not resolved by filipin staining. The filipin staining pattern characteristic of untransfected cells was also seen in cells expressing soluble GFP, Rab6, or Rab7 . In Rab5-overexpressing cells, numerous brightly filipin-positive peripheral dots were observed. These structures were also positive for Rab5, indicating that they represent early endosomes . The most pronounced redistribution of filipin staining was seen in Rab11-overexpressing cells. In these cells, intensely filipin stained and Rab11-positive tubular elements extending to the cell periphery were observed . For Rab5, the effects of the GTPase-deficient mutant (Rab5Q79L) are significantly more pronounced than that of the wild-type protein, resulting in massive enlargement of early endosomes . Considering the moderate effect of wtRab5 on filipin staining pattern we also analyzed the effect of Rab5Q79L. In cells expressing this protein, early endosomes became heavily enlarged and their membranes were intensely filipin positive . Figure 1 | Distribution of free cholesterol in COS-1 cells. Distribution of free cholesterol in COS-1 cells. Left, cells were stained with FITC-conjugated lentil lectin, anti-lamp1 antibodies, biotin-2xFYVE, and Cy3-conjugated streptavidin or anti-Rab11 antibodies. Right, filipin stainings of the respective cells. The arrowheads indicate colocalization of endosomal markers with filipin staining. Images were obtained using wide-field microscope. Bar, 8 mum. Figure 2 | Distribution of free cholesterol in COS-1 cells overexpressing soluble GFP, GFP-fused wtRab5, wtRab6, wtRab7, wtRab11, and Rab5Q79L. Distribution of free cholesterol in COS-1 cells overexpressing soluble GFP, GFP-fused wtRab5, wtRab6, wtRab7, wtRab11, and Rab5Q79L. Left, cells overexpressing GFP or GFP-fused Rab proteins. Right, filipin stainings of the respective cells. The arrowheads indicate free cholesterol redistributed in GFP-Rab --positive organelles. Images were obtained using wide-field microscope. Bar 8, mum. To test whether overexpression of the Rab proteins was accompanied by biochemical effects on cholesterol homeostasis, we analyzed cholesterol esterification by measuring the incorporation of [3H]oleic acid into cholesteryl esters at 40 --48 h posttransfection. The values obtained with Rab overexpressions were compared with those obtained with overexpressed GFP alone. Strikingly, overexpression of the Rabs with the most pronounced effects on filipin distribution, Rab5Q79L and Rab11, also caused the strongest inhibition in cholesterol esterification (similar50% inhibition with both; Figure A). Overexpression of Rab7 was slightly inhibitory (25 --30% inhibition), whereas Rab6 overexpression was without effect. The expression levels of the individual Rabs were closely similar with the 50 --70% transfection frequencies obtained, as assessed by Western blotting with anti-GFP antibodies (Figure B). The effects of the individual Rabs on cholesterol esterification could be observed already at 24 h posttransfection (our unpublished data). Furthermore, the Rab11-induced redistribution of cholesterol was morphologically apparent already at 8 h posttransfection, at a stage when the GFP-Rab11 decorated small punctate structures throughout the cytoplasm (Figure A). By 16 h of transfection, the Rab11- and filipin-positive organelles had attained a more tubular appearance and by 24 h, larger vesicular and tubular profiles containing both Rab11 and cholesterol were generated (Figure A). Figure 3 | Rate of cholesterol esterification in COS-1 cells overexpressing soluble GFP, GFP-fused Rab5Q79L, wtRab6, wtRab7, and wtRab11. Rate of cholesterol esterification in COS-1 cells overexpressing soluble GFP, GFP-fused Rab5Q79L, wtRab6, wtRab7, and wtRab11. (A) Transfected cells grown in culture medium were pulsed for 4 h with [3H]oleic acid in the presence of 50 mug/ml LDL. The rate of esterification is shown as percentage of esterification in cells transfected with GFP alone. Each bar represents three to nine samples from one to four individual experiments; the SEs are indicated. The mean rate of esterification in the GFP control sample was 34 dpm/mug protein/h. (B) Western blot analysis of GFP and different GFP-fused Rab proteins. The cells were transfected for 40 h, and the cell lysate was separated by SDS-PAGE and analyzed by immunoblotting with anti-GFP antibodies. Figure 4 | (A) Formation of GFP-Rab11 --positive organelles with increasing expression times. (A) Formation of GFP-Rab11 --positive organelles with increasing expression times. COS-1 cells overexpressing GFP-wtRab11 were fixed 8, 16, or 24 h posttransfection and stained with filipin. The areas indicated in the top panels are shown in the bottom panels. The arrowheads indicate GFP-wtRab11 -- and filipin-positive organelles. (B) Distribution of TfR and Texas Red transferrin in COS-1 cells coexpressing GFP-wtRab11 and TfR. Cells overexpressing GFP or GFP-wtRab11 (left) and TfR as visualized by anti-TfR antibody. (C) Localization of transferrin in the cells overexpressing GFP-wtRab11 and TfR (unpublished data). The cells were labeled with Texas Red transferrin for 30 min on ice and transferrin internalized for 30 min at 37C. Images in A were obtained using wide-field microscope. Images in B and C are confocal and represent a single focal plane. Bar, 8 mum. Rab11-positive Organelles Accumulate Transferrin Receptor and Internalized Transferrin | In baby hamster kidney and Chinese hamster ovary cells, Rab11 localizes with internalized transferrin in the pericentriolar recycling compartment and in HeLa cells, Rab11 overexpression leads to morphological alterations of the TfR-containing compartments . We therefore analyzed the effect of Rab11 overexpression on the distribution of TfR and its ligand in COS cells. Cells were cotransfected with wtRab11 and TfR cDNAs, and at 40 h posttransfection Texas Red transferrin was bound to the cells for 30 min on ice, followed by 30-min internalization at 37C. In control cells expressing soluble GFP and TfR, the receptor was localized in the perinuclear region and in small punctate and tubular structures throughout the cell (Figure B). In Rab11-expressing cells, the TfR staining was concentrated in larger tubular structures that also contained Rab11 (Figure B). Moreover, labeled transferrin accumulated readily in these structures (Figure C). These results suggest that also in COS cells, Rab11 regulates the dynamics of the endosomal recycling compartment as probed by using TfR and its ligand as markers. Both GTPase-deficient and Dominant Negative Mutants of Rab11 Alter TfR Distribution, Cholesterol Distribution, and Cholesterol Esterification | The GTPase-deficient Rab11 mutant (Rab11Q70L) and the dominant negative mutant (Rab11S25N) differentially affect TfR distribution in HeLa cells . To further characterize the effect of Rab11 on endocytic recycling in COS cells, we studied the distribution of Texas Red transferrin in cells coexpressing TfR and mutant Rab11 proteins. To analyze whether cholesterol distribution was affected, the transfected cells were also stained with filipin. In Rab11Q70L cells, Texas Red transferrin and filipin colocalized in extended tubulovesicular structures that were also strongly positive for the mutant Rab11 . These structures were reminiscent of those observed upon overexpression of wtRab11. Rab11S25N gave a predominantly cytosolic staining pattern in accordance with previous results . In these cells, thinner transferrin-positive tubular elements forming a perinuclearly concentrated network were visualized. This meshwork was also visualized with filipin . We then measured [3H]oleic acid incorporation into cholesteryl esters upon expression of the GTPase-deficient or dominant negative Rab11 mutants. This revealed that both Rab11Q70L and Rab11S25N markedly inhibited cholesterol esterification. The extent of inhibition by either mutant did not differ significantly from that observed with wtRab11 (our unpublished data). Figure 5 | Effect of GFP-Rab11Q70L and GFP-Rab11S25N on the distribution of Texas Red transferrin and free cholesterol in COS-1 cells. Effect of GFP-Rab11Q70L and GFP-Rab11S25N on the distribution of Texas Red transferrin and free cholesterol in COS-1 cells. The cells coexpressing GFP-Rab11Q70L or GFP-Rab11S25N (left) and TfR (unpublished data) were labeled with Texas Red transferrin for 30 min on ice and transferrin internalized for 30 min at 37C (middle). Right, filipin stainings of the respective cells. The arrowheads indicate redistribution of free cholesterol in tubular organelles. Images were obtained using wide-field microscope. Bar, 8 mum. Localization of Other Lipids Enriched in Plasma Membrane and Endosomal Compartments in Rab11-overexpressing Cells | Given the Rab11-induced redistribution of cholesterol, we next analyzed whether Rab11 would also affect the subcellular localization of other endosomal lipids. Rab11-expressing cells were labeled with the PI-3-P binding peptide 2xFYVE . We found that the distribution of PI-3-P, typically in small vesicular structures characteristic of early endosomes, was not altered in Rab11-expressing cells and did not overlap with that of Rab11 . We then analyzed the distribution of the late endosomal acidic phospholipid LBPA by antibody staining. Also this staining was similar in Rab11-expressing cells compared with nonexpressing cells and did not colocalize with Rab11 . Figure 6 | Localization of PI-3-P, LBPA, globotriaosyl ceramide, and GM2 ganglioside is not altered in GFP-wtRab11 --expressing COS-1 cells. Localization of PI-3-P, LBPA, globotriaosyl ceramide, and GM2 ganglioside is not altered in GFP-wtRab11 --expressing COS-1 cells. Cells overexpressing GFP-wtRab11 (left) were permeabilized with filipin and stained with biotin-2xFYVE and Cy3-conjugated streptavidin, anti-LBPA, anti-globotriaosyl ceramide, or anti-GM2 antibodies. Images were obtained using confocal microscope and represent a single focal plane. Bar, 8 mum. Because cholesterol is thought to associate with sphingolipids in membranes, we analyzed the localization of select glycolipids as well as sphingomyelin in Rab11-expressing cells. Monoclonal antibodies against globotriaosyl ceramide gave prominent plasma membrane staining, and this pattern was not altered upon Rab11 expression . GM2 ganglioside is found in late endosomes , and the punctate anti-GM2 staining did not overlap with that of Rab11 . In contrast, the antibody against the glycosphingolipid sulfatide revealed that in some of the Rab11-overexpressing cells, this lipid was redistributed to Rab11-positive organelles as shown in Figure A. This labeling pattern was observed in 57% of the Rab11-overexpressing cells (n = 200). The glycolipid redistribution upon Rab11 overexpression was not as marked as that of cholesterol that consistently colocalized with GFP-Rab11 in the same cells (our unpublished data). In control cells, the anti-sulfatide antibodies visualized the plasma membrane and perinuclear organelles. Some of this labeling may correspond to recycling endosomes as the anti-sulfatide staining partially colocalized with transferrin in control cells (Figure A). Figure 7 | Effect of GFP-wtRab11 on the distribution of sulfatide, CTxB, and Pyr10SM in COS-1 cells. Effect of GFP-wtRab11 on the distribution of sulfatide, CTxB, and Pyr10SM in COS-1 cells. (A) Cells overexpressing TfR (unpublished data) were labeled with Texas Red transferrin for 30 min on ice and transferrin internalized for 30 min at 37C. The cells were then permeabilized with filipin and stained with anti-sulfatide antibodies (top). GFP-wtRab11 and nonexpressing cells were permeabilized with filipin and stained with anti-sulfatide antibodies (bottom). (B) Cells overexpressing GFP-wtRab11 and nonexpressing cells were labeled with Alexa-conjugated CTxB for 1 h on ice and CTxB internalized for 2 h at 37C. (C) Cells overexpressing GFP or GFP-wtRab11 were labeled with Pyr10SM for 10 min at 37C and incubated for further 2 h at 37C in the absence of the lipid. The arrowheads indicate GFP-wtRab11 --positive organelles containing Pyr10SM. (D) Electron micrograph of GFP-Rab11 --overexpressing COS-1 cell. The anti-GFP antibodies decorate clustered tubulovesicular membranes. Images in A and B are confocal and represent a single focal plane. Images in C were obtained using wide-field microscope. Bars, 8 mum (A --C) and 200 nm (D). To visualize the distribution of GM1 ganglioside, we labeled the cells with Alexa 594-conjugated CTxB for 1 h on ice followed by 2 h internalization at 37C. This resulted in prominent perinuclear staining in control cells (Figure B), in accordance with the transport of the toxin to the Golgi . On Rab11 overexpression CTxB was largely redistributed to GFP-Rab11 --positive organelles (Figure B). To visualize sphingomyelin, cells were labeled with the fluorescent Pyr10SM for 10 min and chased for 2 h before fixation. At this time point Pyr10SM was not significantly degraded because >95% of the label was still associated with sphingomyelin. In GFP-expressing cells, the staining was visualized in the plasma membrane and in punctate perinuclear structures (Figure C), probably representing late endocytic organelles and the Golgi apparatus, analogously to the distribution of BODIPY-Sphingomyelin . In Rab11 cells, Pyr10SM was additionally distributed to more peripheral structures throughout the cell that colocalized to a large extent with Rab11 (Figure C). These results indicate that the Rab11-induced lipid redistribution is highly selective. The GFP-Rab11 --containing compartments exclude select early and late endosomal lipids while including specific sphingolipids and even distinguishing between different glycosphingolipid classes. When the ultrastructure of the Rab11-containing organelles was analyzed by electron microscopy the GFP-Rab11 --positive structures were resolved as tubulovesicular clusters of membranes (Figure D). Late Endosomal Cholesterol Transport in Rab11-overexpressing Cells | One possible explanation for the accumulation of cholesterol in Rab11-containing compartments and the inhibition in cholesterol esterification could be that Rab11 interferes with LDL-cholesterol internalization. To test this possibility, we added DiI-LDL to living cells for 15 min and either fixed the cells directly or after a 2-h chase. The distribution of DiI was visualized in cells costained with antibodies against the lysosomal membrane protein lamp1. In GFP- and in Rab11-expressing cells, the markers were segregated at 15 min, whereas at 2 h extensive colocalization was observed, indicating transport of DiI to lysosomes (Figure , A and B; our unpublished data). In contrast, in Rab5Q79L-expressing cells, DiI-LDL labeling was observed in the cores of enlarged early endosomes that were delineated by Rab5Q79L, at both time points (Figure , A and B). These data indicate that the transport of DiI-LDL to late endosomes and lysosomes was blocked in Rab5Q79L-expressing cells. However, this was not the case in Rab11-overexpressing cells. Figure 8 | Distribution of internalized DiI-LDL in COS-1 cells overexpressing GFP-Rab5Q79L or GFP-wtRab11. Distribution of internalized DiI-LDL in COS-1 cells overexpressing GFP-Rab5Q79L or GFP-wtRab11. Cells transfected with indicated constructs were cultured for 24 h in 5% LPDS medium, labeled with DiI-LDL for 15 min, and fixed (A) or labeled with DiI-LDL for 15 min and incubated for 2 h in serum-free medium (B). The cells were then stained with anti-lamp1 antibodies. The colocalization of DiI and lamp1 is indicated by arrowheads in respective panels and as purple color in the overlay. (C) Localization of LDL receptor in COS-1 cells double-overexpressing soluble GFP or GFP-wtRab11 and the LDL-receptor. LDL receptor was visualized by anti-LDL receptor antibodies. Images were obtained using confocal microscope and represent a single focal plane. Bars, 8 mum. We also examined the distribution of the LDL-receptor in Rab11-overexpressing cells. The localization of the receptor was visualized in cells coexpressing the LDL receptor and Rab11 and compared with cells overexpressing the LDL receptor and soluble GFP. In both cases, anti-LDL receptor antibodies visualized surface staining and small punctate structures (Figure C). In the Rab11-overexpressing cells, these structures partially colocalized with Rab11. However, the prominent plasma membrane staining of the receptor was not appreciably altered by Rab11 overexpression. Together, these results suggest that the cholesterol accumulation in Rab11-containing organelles is not likely to be explained by sequestration of the LDL receptor and its ligand. Rab11 and the NPC Phenotype | The phenotype of the Niemann-Pick type C disease cells indicates that the NPC1 and NPC2 proteins have important functions in endocytic cholesterol trafficking. When the distribution of endogenous NPC1 or NPC2 was analyzed in wtRab11 or Rab11Q70L-overexpressing cells, we found no colocalization of the proteins with the overexpressed Rab . Instead, NPC1 and NPC2 colocalized with late endocytic markers as reported previously (our unpublished data; ; ). Interestingly, in Rab5Q79L-expressing cells, both NPC1 and NPC2 accumulated in the enlarged early endosomes. The proteins were often visualized inside the organelles surrounded by Rab5Q79L . This suggests that NPC1 and NPC2 may be more closely connected by membrane trafficking with early than with recycling endocytic compartments. Figure 9 | Localization of NPC1 and NPC2 in COS-1 cells overexpressing GFP-wtRab11, GFP-Rab11Q70L, or GFP-Rab5Q79L. Localization of NPC1 and NPC2 in COS-1 cells overexpressing GFP-wtRab11, GFP-Rab11Q70L, or GFP-Rab5Q79L. Endogenous NPC1 and NPC2 were visualized by anti-NPC1 or anti-NPC2 antibodies. Images were obtained using confocal microscope and represent a single focal plane. Bar, 8 mum. To rule out that the cholesterol deposition in Rab11-positive organelles was a phenomenon limited to COS cells, we analyzed the cholesterol distribution upon Rab11 overexpression in primary fibroblasts cooverexpressing TfR. Also in these cells, the Rab11-positive organelles accumulated TfR as well as free cholesterol (Figure A). Cholesterol accumulation was not observed in cells overexpressing soluble GFP and TfR (Figure A). We then tested whether the Rab11-induced cholesterol accumulation could be observed in cells exhibiting a lysosomal cholesterol transport block. This was achieved by transfecting NPC patient fibroblasts with the Rab11 construct. Also in these cells, the Rab11-induced cholesterol deposits were observed (Figure B). These structures were typically more peripherally localized and less intensively stained with filipin compared with the lysosomal deposits. Similar accumulation of cholesterol in Rab11-containing organelles was observed in COS cells when a lysosomal cholesterol transport block was introduced pharmacologically using U18666A (our unpublished data). Figure 10 | Effect of GFP-wtRab11 on the distribution of free cholesterol in normal and NPC fibroblasts and the complementation of NPC fibroblasts by the NPC1 protein. Effect of GFP-wtRab11 on the distribution of free cholesterol in normal and NPC fibroblasts and the complementation of NPC fibroblasts by the NPC1 protein. (A) Human control fibroblasts coexpressing TfR and soluble GFP (left) or TfR and GFP-wtRab11 (right) were stained with anti-TfR antibodies and filipin. (B) NPC fibroblasts transfected with GFP-wtRab11 and stained with filipin. (C) Human NPC fibroblast cooverexpressing GFP-wtRab11 and NPC1 (as visualized by anti-NPC1 antibodies) and filipin staining of the respective cell. The arrowheads indicate GFP-wtRab11 --positive organelles containing free cholesterol. The panels in A and B and filipin panel in C were imaged using wide-field microscope. Other panels in C are confocal and represent a single focal plane. Bar, 8 mum. We next investigated whether Rab11 overexpression would interfere with the clearance of the late endocytic cholesterol storage by the NPC1 protein. NPC fibroblasts were cotransfected with NPC1 and Rab11 and imaged 3 d posttransfection. We found that NPC1 did not colocalize with Rab11 and was capable of complementing the NPC cells as shown by the disappearance of the filipin-positive lysosomal cholesterol stores (Figure C). Moreover, the weaker peripheral filipin-positive staining typical to Rab11-overexpressing fibroblasts was observed in these cells. Together, these data indicate that both the buildup and the disappearance of late endocytic cholesterol accumulation can take place irrespective of Rab11 overexpression and that the Rab11 induced cholesterol deposition can be observed in cells with a late endocytic cholesterol transport block. Rab11-induced Decrease in Cholesterol Esterification Is Not Dependent on LDL-Cholesterol and Can Be Bypassed by Adding Cholesterol in a Cyclodextrin Complex | Our morphological data suggested that the cholesterol accumulation upon Rab11 overexpression may not depend on LDL-cholesterol internalization. We therefore incubated COS cells in lipoprotein-deficient medium after Rab11 transfection. Filipin staining of the cells revealed that the Rab11-containing organelles were indeed cholesterol loaded also under these conditions (Figure A). Moreover, the Rab11-induced decrease in cholesterol esterification was also independent of LDL-cholesterol as shown by the decrease in [3H]oleic acid incorporation into cholesteryl esters in the presence of lipoprotein-deficient medium (Figure B). Under delipidating conditions, the overall rate of esterification was decreased expectedly (as observed by comparing the absolute dpms to those in Figure ), but a Rab11-induced inhibition of esterification was nevertheless clearly observed. Interestingly, under these conditions Rab6 overexpression enhanced esterification slightly compared with the GFP control (Figure B). Figure 11 | (A) Distribution of free cholesterol in delipidated COS-1 cells expressing GFP-wtRab11. (A) Distribution of free cholesterol in delipidated COS-1 cells expressing GFP-wtRab11. Cells transfected with GFP-wtRab11 were grown in 5% LPDS medium for 24 h before fixation and stained with filipin. Images were obtained using wide-field microscope. Bar, 8 mum. (B) Rate of cholesterol esterification in delipidated COS-1 cells overexpressing soluble GFP, GFP-wtRab6, or GFP-wtRab11. Cells transfected with GFP, GFP-wtRab6, or GFP-wtRab11 were cultured in 5% LPDS medium for 24 h before labeling with [3H]oleic acid for 4 h. The rate of esterification is shown as percentage of esterification in cells transfected with GFP alone. Each bar represents 5 --10 samples from two to four individual experiments; the SEs are indicated. The mean rate of esterification in the GFP control sample was 5 dpm/mug protein/h. (C) Rate of cholesterol esterification in GFP-wtRab11 --overexpressing COS-1 cells upon loading with cholesterol/mbeta-CD-complex. Cells transfected with GFP or GFP-wtRab11 were cultured in 5% LPDS medium for 24 h before labeling with [3H]oleic acid for 4 h. Cholesterol/mbeta-CD-complex was added during the labeling to yield the loading time points indicated. Values represent the average of duplicate samples from a representative experiment; the variation between samples is indicated. To analyze the possibility that the cholesterol load in Rab11-positive organelles was due to elevated de novo cholesterol synthesis, we labeled GFP and GFP-Rab11 transfected COS cells with [3H]acetate and analyzed the amount of [3H]cholesterol formed. There was no increase in the rate of cholesterol biosynthesis in Rab11-transfected cells (207 +- 18 dpm/mug protein; SEM, n = 3) compared with GFP control (194 +- 8 dpm/mug protein; SEM, n = 3). Considering the recycling characteristics of the Rab11-harboring organelles, we reasoned that the block in cholesterol esterification could be due to defective recycling of cholesterol from endocytic organelles to the plasma membrane. We therefore tested whether the inhibition of cholesterol esterification by Rab11 overexpression could be overcome by providing exogenous cholesterol to the plasma membrane. This was achieved by incubating cells with a cholesterol/mbeta-CD-complex, which results in efficient cholesterol loading of cells and induces a rapid and massive compensatory increase in cholesterol esterification. When the increase in [3H]oleic acid incorporation upon cholesterol addition was compared between GFP- and GFP-Rab11 --expressing cells no significant differences were detected at any time point analyzed (Figure C). This result shows that the Rab11-induced block in cholesterol esterification could be bypassed by adding cholesterol to the plasma membrane. DISCUSSION : The role of Rab proteins as specific regulatory switches of protein transport is well appreciated. In this work, we provide the first evidence for selective regulation of cholesterol trafficking and homeostasis by endocytic Rab proteins. The well-characterized Rabs 5, 7, and 11 were chosen to encompass aspects of early, late, and recycling endocytic membrane trafficking, respectively. Yet, the precise boundaries between endocytic compartments cannot be determined by Rab proteins and to a large extent, these boundaries still remain to be established . Enlargement of the early endosomal compartment by the dominant active mutant of Rab5 was accompanied by sequestration of cholesterol in these organelles and a concomitant decrease in cholesterol esterification. This may be explained, at least partially by inhibition of LDL-cholesterol transport, as observed by the accumulation of DiI-LDL and the LDL-receptor in the Rab5Q79L-positive organelles (Figure ; our unpublished data). In contrast, Rab7 overexpression did not appreciably alter cholesterol distribution as assessed by filipin staining, and its effect on cholesterol esterification was also more moderate than that of Rab5. This was somewhat surprising considering that NPC1, a key regulator of endocytic cholesterol flow, is localized in Rab7-positive late endosomes and that the cholesterol accumulation in NPC disease is most pronounced in late endocytic organelles. Because Rab5 (and Rab11; see below) overexpression instead had marked effects on cholesterol balance, one possible explanation is that the bulk of endocytic cholesterol flow normally occupies earlier endocytic compartments than the Rab7-regulated organelles. However, the transient transfection approach used and differences between individual Rabs (e.g., with respect to their GTP-GDP cycle or potential effects on LDL uptake or degradation), preclude comparisons regarding the quantitative contribution of each Rab in regulating cholesterol flow. The most pronounced effects on cholesterol balance were observed upon Rab11 overexpression. The wild-type, GTPase-deficient, and dominant inhibitory Rab11 were all effective, analogously to the effects of Rab11 and its mutants on transferrin recycling and the transport of shiga toxin B subunit . The Rab11-regulated, cholesterol-sequestering organelles have characteristics of recycling endosomes based on the accumulation of TfR and internalized transferrin. We observed that two of the glycosphingolipids studied, sulfatide and GM1, cosequestered with cholesterol in the Rab11-positive organelles, as judged by antibody staining and cholera toxin labeling, respectively. In addition, pyrene-labeled sphingomyelin was found to colocalize with Rab11. This is in accordance with previous reports showing that recycling endosomes are enriched with sphingomyelin and cholesterol , and that these two lipids have high affinities toward each other . On the other hand, the early endosomal PI-3-P and the late endosomal LBPA were not redistributed to Rab11 organelles. Moreover, the distributions of two other glycolipids, globotriaosyl ceramide and GM2 ganglioside, were not altered. The former localized mostly on the plasma membrane and the latter in late endocytic organelles in both Rab11-overexpressing and control cells. Our results provide evidence that the specificity of lipid and protein transport along the endocytic pathway is maintained in Rab11-overexpressing cells. Furthermore, they reinforce the emerging concept of differential sorting of glycolipids along the endocytic pathways . The mechanisms by which the selective accumulation of lipids upon Rab11 overexpression is generated, remain to be elucidated. It could potentially derive from selective retention in recycling endosomes or from altered sorting at an endocytic step before recycling endosomes (or a combination of both). Because the LDL-receptor route of cholesterol internalization is well characterized the potential contribution of this route to the Rab11-induced cholesterol sequestration was studied. The following data suggest that the Rab11-regulated cholesterol transport route is separate from the LDL-cholesterol route and that the majority of the cholesterol trapped in Rab11-containing organelles is not directly derived from LDL. First, DiI-LDL was transported to lysosomes in Rab11-overexpressing cells and the bulk of the LDL-receptor was not sequestered in Rab11 organelles. Second, both the morphologically detected accumulation of free cholesterol and the inhibition of cholesterol esterification were LDL independent. Third, cholesterol deposition in Rab11-positive organelles was seen also in cells exhibiting lysosomal cholesterol accumulation. Finally, introduction of the NPC1 protein into NPC patient fibroblasts allowed correct localization of the protein and this lead to clearance of the late endocytic cholesterol deposits irrespective of Rab11 overexpression. Entrapment of the recycling marker TfR in the Rab11-containing organelles implies that in COS cells Rab11 regulates recycling of select cargo to the plasma membrane as has been observed for other cell types . We therefore hypothesized that the cholesterol entrapment in Rab11 organelles may result from reduced recycling of cholesterol to the plasma membrane. This would be in accordance with recent data demonstrating that expression of a dominant negative Rme-1 retards the return of dehydroergosterol to the cell surface . Although recycling takes place from several stations along the endocytic route, recycling endosomes are thought to be more plastic than e.g., sorting or late endosomes and could store accumulating cargo. Moreover, their membranes could have particularly high affinity for cholesterol . Even moderate stagnation in recycling may be sufficient to eventually manifest as massive deposition of cholesterol. If the decrease in cholesterol esterification upon Rab11 overexpression were due to its endosomal entrapment and inaccessibility to the plasma membrane, one should be able to bypass the effect by adding excess cholesterol on the plasma membrane. To test this, we loaded the plasma membrane with cholesterol using a cyclodextrin carrier. Indeed, the Rab11-induced inhibition of cholesterol esterification was now bypassed. However, the intensity of the plasma membrane filipin staining was not appreciably reduced upon Rab11 overexpression. It is therefore plausible that the entrapment of cholesterol in the Rab11 endosomes is by itself sufficient to explain the reduced accessibility of cholesterol for esterification. An important concept emphasized by the present study is the modulation of cholesterol homeostasis by perturbation of recycling membrane trafficking. This effect is partially but not fully analogous to the effects seen in cholesterol balance in the NPC-related cholesterol transport block. Rab11 overexpression causes the accumulation of free cholesterol in organelles with recycling characteristics, whereas loss of NPC1 function results in the accumulation of cholesterol in late endocytic compartments. Both result in a defect in cholesterol esterification. However, in NPC cells cholesterol biosynthesis is increased, but in the Rab11 cells, cholesterol synthesis remains unaltered. Neither in NPC nor in Rab11 cells is the acyl coenzyme A-cholesterol acyltransferase (ACAT) activity reduced (as suggested by the normal esterification upon cholesterol/mbeta-CD addition in Rab11 cells). This suggests that in both cases, defective esterification is due to reduced substrate availability to ACAT. This again is likely to be associated with the endocytic cholesterol transport problems. Our data support the idea that several membrane-trafficking pathways can feed ACAT with cholesterol. The role of NPC1 in regulating membrane trafficking has been reinforced in several studies . In this work, we demonstrate that cholesterol esterification can be modulated by a number of Rab proteins associated with distinct membrane-trafficking pathways. Interestingly, although the focus of the present study was on selected endosomal Rabs (all of which inhibited esterification to a various extent) we also noted that the Golgi-associated Rab6 enhanced cholesterol esterification when the cells were cultured in lipoprotein-deficient serum and the basal esterification rate was slow. This points to the intriguing possibility that the Rab6-regulated retrograde Golgi transport pathway may carry cholesterol back to the endoplasmic reticulum where ACAT esterifies it. This route may be physiologically relevant as the dominant negative mutant of Rab6 inhibited esterification when the cells were cultured in the presence of serum lipoproteins (our unpublished data). Interestingly, another retrograde Golgi transport route via COPI-coated vesicles is depleted of cholesterol , supporting the idea that there may be preferential membrane carriers for cholesterol both along the exocytic and endocytic pathways. In addition, nonvesicular cholesterol-trafficking itineraries operate in parallel. For instance, digestion of plasma membrane sphingomyelin by neutral sphingomyelinase has been shown to lead to plasma membrane vesiculation and stimulation of cholesterol esterification in an ATP-independent manner . It is conceivable that cholesterol transferred to cells from the cyclodextrin complex may use such a mechanism because Rab11 does not inhibit this pathway. In conclusion, our data reveal a novel role for the recycling endosomal circuits regulated by Rab11, in endocytic cholesterol trafficking. The effects of Rab11 are not explained by perturbations on LDL-cholesterol transport but rather point to the role of the endocytic recycling compartment in membrane cholesterol cycling and its significance in maintaining cellular cholesterol homeostasis. Backmatter: Abbreviations used: : ACAT = acyl coenzyme A-cholesterol acyltransferase CD = cyclodextrin CTxB = cholera toxin subunit B DiI-LDL = 1,1'-dioctadecyl-3,3,3',3'-tetramethyl-indocarbocyanine-perchlorate --labeled low-density lipoprotein LBPA = lysobisphosphatidic acid LDL = low-density lipoprotein LPDS = lipoprotein-deficient serum mbeta-CD = methyl-beta-cyclodextrin NPC = Niemann-Pick type C PI-3-P = phosphatidylinositol 3-phosphate Pyr10SM = pyrenyldecanoylsphingomyelin TfR = transferrin receptor PMID- 12221120 TI - All Small Nuclear RNAs (snRNAs) of the [U4/U6.U5] Tri-snRNP Localize to Nucleoli; Identification of the Nucleolar Localization Element of U6 snRNA AB - Previously, we showed that spliceosomal U6 small nuclear RNA (snRNA) transiently passes through the nucleolus. Herein, we report that all individual snRNAs of the [U4/U6.U5] tri-snRNP localize to nucleoli, demonstrated by fluorescence microscopy of nucleolar preparations after injection of fluorescein-labeled snRNA into Xenopus oocyte nuclei. Nucleolar localization of U6 is independent from [U4/U6] snRNP formation since sites of direct interaction of U6 snRNA with U4 snRNA are not nucleolar localization elements. Among all regions in U6, the only one required for nucleolar localization is its 3' end, which associates with the La protein and subsequently during maturation of U6 is bound by Lsm proteins. This 3'-nucleolar localization element of U6 is both essential and sufficient for nucleolar localization and also required for localization to Cajal bodies. Conversion of the 3' hydroxyl of U6 snRNA to a 3' phosphate prevents association with the La protein but does not affect U6 localization to nucleoli or Cajal bodies. Keywords: INTRODUCTION : The nucleolus is the site of ribosome biogenesis (for review, see ). However, the nucleolus seems to be plurifunctional and contains RNA used for other events, such as the RNA component of RNase P, which catalyzes the 5' processing of pre-tRNA , signal recognition particle RNA that assembles with proteins in the nucleolus and telomerase RNA . Recently, it has been reported that several of the small nuclear RNAs (snRNAs) pass through the nucleolus before their nucleoplasmic destination where splicing occurs. U6 snRNA transiently localizes to nucleoli where it seems to undergo 2'-O-methylation and pseudouridylation of defined nucleotides, guided by small nucleolar RNAs (snoRNAs) . Similarly, U2 snRNA is found in nucleoli where it seems to be modified by guide snoRNAs, probably after reimport to the nucleus from the cytoplasm . In addition, guide snoRNAs for modification of several spliceosomal snRNAs have also been identified . In this report, we present direct evidence that two of the targets of modification, U4 and U5 snRNAs, localize to nucleoli. Therefore, the list of snRNAs associated with nucleoli is expanded to U6 , U2 , and U4 and U5 snRNAs (this report). The observation of nucleolar localization of U6 snRNA raises the question at which point of its life cycle U6 enters the nucleolus. Upon transcription by RNA polymerase III , the first protein to associate with U6 snRNA is La , which generally binds to the 3' termini of nascent RNA polymerase III transcripts and a number of viral RNAs . Subsequently, the 3'-poly U end of U6 elongates during maturation and then is trimmed to approximately five uridines with conversion of the 3' end of U6 snRNA from a hydroxyl group to a 2',3'-cyclic phosphate . At this point, La is replaced by the evolutionarily conserved Sm-like (Lsm) proteins Lsm 2 --8 , which facilitate the formation of the [U4/U6] di-snRNP . The Lsm proteins are specific in their binding to U6 snRNA and do not associate with U1, U2, U4, or U5 snRNAs that instead are bound by Sm proteins . Then, the [U4/U6.U5] tri-snRNP forms to which Lsm2-Lsm8 and other proteins are associated , and the use of U6 snRNA in the nucleoplasm for splicing ensues. The experiments presented herein demonstrate that the 3' end of U6, which in vivo associates with the La protein and subsequently with the Lsm protein complex to form the [U4/U6.U5] tri-snRNP, is essential for nucleolar localization as well as localization to Cajal (coiled) bodies but not for assembly with U4 snRNA. Furthermore, by alteration of the 3' end of U6 snRNA, we can exclude the possibility that La plays a role in localization of U6 to the nucleolus or to Cajal bodies. In addition, we demonstrate that all snRNAs of the [U4/U6.U5] tri-snRNP localize to nucleoli. MATERIALS AND METHODS : In Vitro Transcription and Labeling of RNA | In vitro transcription reactions using polymerase chain reaction (PCR)-generated DNA templates produced the labeled RNAs used in the present study. The templates and primers used for PCR are given below. Templates. | The starting material for the template for in vitro transcription of U6 snRNA was the human U6 clone pT7U6 , which carries a U6 gene that is identical in sequence to Xenopus tropicalis except for a one base difference at nucleotide (nt) 6. An appropriate 5' primer was used to give a PCR product identical to the Xenopus U6 snRNA gene sequence, as described previously ; this was subcloned into pCR3.1 (Invitrogen, Carlsbad, CA) and its sequence was confirmed. The template and primers for wild-type (WT) U6 snRNA have been described in . The primers to generate templates by PCR for in vitro transcription of mutant U6 snRNA and of wild-type U4 or U5 snRNA are listed below. Clones containing the genes for Xenopus laevis U5 snRNA and chicken U4B snRNA were kindly provided by I.W. Mattaj (European Molecular Biology Laboratory, Heidelberg, Germany) in the pUC9 plasmid; the corresponding snRNAs were used in this study because their structure --function relationships were previously extensively characterized in Xenopus oocytes . Transcripts of U3 snoRNA were prepared as described previously . A triple repeat of the 3' end of U6 snRNA (nt 87 --107, called 3'-end nucleolar localization element [NoLE]) was generated by PCR and its sequence confirmed. This NoLE construct was generated using 5' and 3' primers listed below. The minus strand template was 5'-TTG CCG AGG AGC TT(A AAA ATA TGG AAC GCT TCA CG)(A AAA ATA TGG AAC GCT TCA CG)(A AAA ATA TGG AAC GCT TCA CG) C CCT ATA GTG AGT CGT ATT A-3'. The first 14 nucleotides of this oligonucleotide are derived from the genomic sequence that follows the 3' end of the U6 snRNA coding region of X. tropicalis. The 21 nt shown in parentheses and repeated three times are complementary to nt 87 --107 at the 3' end of U6 snRNA. The nt in italics are the first 18 nucleotides of the T7 promotor. Restriction of the 3'-end PCR template with MseI (site is underlined above) removed the genomic sequence flanking the gene; subsequent in vitro transcription produced the 65-nt 3'-end RNA construct (triple repeat preceded by GG from the T7 promoter). The 3'-end NoLE construct just described was also tested for nucleolar localization in a heterologous context by coupling it to the 3' end of a synthetic control RNA . The template for this construct (control RNA/3' end) was created by annealing (complementary sequences are underlined) and PCR extension of the following two oligonucleotides: 5'-TAA TAC GAC TCA CTA TAG GGT CCT GTC GAC TCC TCC TCC TCC TCC TCC GCG GAT TTA CCT CGG CAA CGT-3' and 5'-TAA ATG AGG AGC TT(A AAA ATA TGG AAC GCT TCA CG)(A AAA ATA TGG AAC GCT TCA CG)(A AAA ATA TGG AAC GCT TCA CG)T TGC CGA GGT AAA TCC GCG GA-3'; the T7 promotor is shown in italics and the 3'-end NoLE (nt 87 --107) repeated three times are enclosed by parentheses. The synthetic control RNA by itself was created by PCR by using the template and primers described by . U2 snRNA was used as a control for immunoprecipitation experiments and stability assays and was derived as described previously . 5' End Primers (T7 Promotor Shown in Italics). | U4 WT 5'-TAA TAC GAC TCA CTA TAG GGA GCT TTG CGC AGT GGC AGT ATC-3'; U5 WT 5'-TAA TAC GAC TCA CTA TAG GGA TAC TCT GGT TTC TCT TCA AAT TCG AAT AAA TC-3'; U6 Delta1 --19 5'-TAA TAC GAC TCA CTA TAG GGA TAT ACT AAA ATT GGA-3'; U6 Delta20 --25 5'-TAA TAC GAC TCA CTA TAG GGT GCT TGC TTC GGC AGC ACT AAA ATT GG-3'; U6 sub20 --25 (substitution is underlined) 5'-TAA TAC GAC TCA CTA TAG GGT GCT TGC TTC GGC AGC ACG AGC GGT AAA ATT GG-3'; U6 Delta26 --42 5'-TAA TAC GAC TCA CTA TAG GGT GCT TGC TTC GGC AGC ACA TAT ACA GAG AAG AT-3'; and T7 promoter for 3'-end construct 5'-TAA TAC GAC TCA CTA TAG GG-3'. 3' End Primers. | U4 WT 5'-CAG TCT CCG TAG AGA CTG TCA-3'; U5 WT 5'-TAC CTG GTG TGA ACC AGG CTT C-3'; U6 Delta43 --56 5'-AAA AAT ATG GAA CGC TTC ACG AAT TTG CGT GTC ATC CTT GCG CAG GGG CCA GTA TCG TTC C-3'; U6 Delta57 --81 5'-AAA AAT ATG GAA CGC TTC ACG AAT TTT GCT AAT CTT-3'; U6 Delta43 --81 5'-AAA AAT ATG GAA CGC TTC ACG AAT TTG TAT CGT TCC-3'; U6 Delta82 --95 5'-AAA AAT ATG GAA GCG TGT CAT CCT TGC-3'; U6 Delta96 --99 5'-AAA AAT ATC GCT TCA CGA ATT T-3'; U6 Delta100 --102 5'-AAA AAG GAA CGC TTC ACG AAT TTG CGT GTC ATC CTT G-3'; U6 Delta103 --107 5'-TAT GGA ACG CTT CAC GAA TTT GCG TGT CAT CCT TG-3'; U6 Delta100 --107 5'-GGA ACG CTT CAC GAA TTT GCG TGT CAT CCT TG-3'; U6 Delta82 --107 5'-GCG TGT CAT CCT TGC GCA GGG GCC-3'; and U6 3'-end 5'-TTG CCG AGG AGC TTA AA-3' (genomic sequence downstream of U6). In vitro transcripts of RNA were made with either fluorescein-12-UTP (PerkinElmer Life Sciences, Boston, MA) and/or alpha-[32P]UTP (PerkinElmer Life Sciences) label that was added to a T7 megascript in vitro transcription kit (Ambion, Austin, TX). The T7 transcripts were purified according to . They all contained GG at their 5' ends from the T7 promoter and were capped with m7G(5')ppp(5')G (Ambion) to improve stability. Oocyte Microinjection | Stage V-VI oocytes from X. laevis were obtained as described previously . For fluorescence analysis of nucleolar localization as well as for stability assays, oocyte nuclei were injected with 0.8 ng of wild-type U4, U5, and U6 snRNA or U3 snoRNA in 9.2 nl of H2O. For the U6 snRNA mutants that were shorter than the wild-type and for the 3'-end construct, amounts equimolar to the wild-type U6 RNA (0.8 ng = similar23 fmol of in vitro-transcribed WT U6 snRNA) were injected. The concentration of U6 used for injection was determined by titration to be the lowest possible that still gave specific labeling of nucleoli (our unpublished data) and is in the range of the concentration of endogenous U6 (similar7.5 fmol/stage V-VI oocyte as determined by Northern blot analysis), whereas there is nuclear retention of up to similar500 --600 fmol of U6 per Xenopus oocyte . The concentration used for injection of our transcripts is also in the range of those used by for oocyte injection of U1, U2, and U5 snRNA. For the 40-nt negative control RNA, 0.8 ng/oocyte was injected, which is equivalent to similar62 fmol/oocyte. A further control was the injection of an excess of fluorescein-labeled UTP at 5 pmol/oocyte. We confirmed that microinjected U4 and U6 snRNA transcripts could participate in their normal functional pathway and form a di-snRNP. This was achieved by coimmunoprecipitation (see section below) with an anti-Sm antibody as described previously . Endogenous U6 and U4 snRNA were disrupted through RNase H-mediated destruction by two nuclear injections spaced 4 h apart of 9.2 nl of each of the following antisense oligonucleotides at a concentration of 3 mug/mul (28 ng/oocyte): a combination of two oligonucleotides complementary to nt 20 --53 (5'-TAA TCT TCT CTG TAT CGT TCC AAT TTT AGT ATA T-3') and nt 75 --102 (5'-TAT GGA ACG CTT CAC GAA TTT GCG TGT C-3') was used for U6 depletion. U4 depletion was carried out with an oligonucleotide complementary to nt 51 --83 (5'-GGG TAT TGG GAA AAG TTT TCA ATT AGC AAT A-3'). Nucleolar Localization Assay | After incubation of the oocytes for a stipulated time (1.0 --1.5 h unless specified otherwise), nuclear spreads were made as described previously using a method for preparation of lampbrush chromosomes . For Cajal body immunostaining, the slides were fixed for 1 h in 2% paraformaldehyde in phosphate-buffered saline (PBS) (137 mM NaCl, 3 mM KCl, 6.4 mM Na2HPO4, and 1.5 mM KH2PO4, pH 7.0), washed in PBS, and unspecific binding sites saturated with 10% bovine serum albumin in PBS for 20 min. Subsequently, rabbit polyclonal serum against a synthetic 21 amino-acid fragment of Xenopus coilin (kindly provided by J.G. Gall, Carnegie Institution, Baltimore, MD) was applied as a primary antibody for immunostaining of Cajal bodies at a dilution of 1:1000 in PBS for 20 min at 4C. The preparations were rinsed 3x for 5 min each in PBS and incubated for 20 min at 4C with the secondary antibody (goat anti-rabbit; Molecular Probes, Eugene, OR) coupled to the dye Alexa 594. Subsequently, slides were washed 3x with PBS for 5 min each and the DNA stained with 200 ng/ml 4'-6-diamidino-2-Phenylindole (DAPI) in PBS for 5 min. Fluorescence microscopy was performed as described previously with the exception that ProLong mounting medium (Molecular Probes) was used. Nucleolar preparations were analyzed with an Axiophot epifluorescence microscope (Carl Zeiss, Thornwood, NY) equipped with a 100x Neofluar Ph 3 objective and a 100-W mercury lamp. Pictures were taken with constant exposures for each filter set (for DAPI, Alexa 594, or fluorescein) by using Ektachrome 400x professional film (Eastman Kodak, Rochester, NY). Therefore, any difference in signal strength between various samples is directly visualized without the interference of software or automatic camera settings. snoRNA Stability Assay | To determine the stability of the various in vitro transcripts after injection into oocyte nuclei, U2 snRNA was coinjected and served as an internal control to normalize for any differences in injection or recovery of the samples. At defined time points after injection of the oocytes with [alpha-32P]UTP --labeled RNAs, the RNA of four nuclei per sample was recovered and analyzed as described previously . Immunoprecipitation of U6 snRNA from Oocytes | For immunoprecipitation experiments, 0.8 ng/oocyte of the purified U6 snRNA (colabeled with alpha-[32P]UTP and with fluorescein-12-UTP) and either 0.8 ng/oocyte fluorescein-12-UTP --labeled U4 snRNA or alpha-[32P]UTP --labeled U3 snoRNA were coinjected into Xenopus oocytes and incubated for 1 or 4 h (depending on the experiment) at 20C. For each sample, 10 nuclei were homogenized on ice in 50 mul of isolation buffer (50 mM NaCl, 10 mM Tris pH 8.0, 1 mM dithiothreitol, 100 U/ml RNase inhibitor [Roche Applied Science, Mannheim, Germany], one tablet of protease inhibitor cocktail [Roche Applied Science] per 10 ml of isolation buffer), and spun for 1 min in a microcentrifuge at 10,000 rpm. The supernatant was removed, spun again twice, and then added to 240 mul of IP 150 (150 mM NaCl, 10 mM Tris pH 8.0, 0.1% NP-40, 1 mM dithiothreitol, 10 U/ml RNase inhibitor [Roche Applied Science], and one tablet of protease inhibitor cocktail [Roche Applied Science] per 10 ml of isolation buffer), and 20 mul of protein A-Sepharose beads. The beads had been coupled either to rabbit-anti-Xenopus La antibody (immune serum 79, provided by S. Clarkson, University Medical Center-CMU, Geneva, Switzerland; ) or to preimmune serum as a control, or in other cases to monoclonal Y12 mouse anti-Sm antibody or mouse IgG as a control by incubation of 150 mul of preswollen beads with 150 mul of IP 500 (500 mM NaCl, 10 mM Tris pH 8.0, 0.1% NP-40, and 0.1% sodium azide) and 200 mul of antibody for 4 h at 4C with end over end rotation before they were spun and washed 3x in IP 150 in a microcentrifuge at 1000 rpm. The mixture of nuclear extract and antibody-coupled beads was rotated end over end for 90 min (anti-La antibody) or 8 h (anti-Sm antibody) at 4C before the beads were spun and washed 5x in IP 150 in a microcentrifuge at 1000 rpm. Then the RNA was isolated and purified. Precipitated RNA and supernatant were analyzed on a denaturing 7 M urea, 8% polyacrylamide gel (1 mm in thickness, 35 cm in length). It is important to note that although U4 transcript association with Sm proteins and U6 is initiated upon injection into the oocyte, further association can also occur during coincubation of nuclear lysate with antibody-coupled beads. Therefore, accurate kinetics of in vivo association of U4/U6 cannot be carried out. For alteration of the 3'-hydroxyl end of U6 to a 3'-phosphate group, 10 mul with the in vitro transcript (similar1 mug) was incubated in Whitfield's reagent (25 mM NaIO4 and 1 M lysine, pH 8.5) for 2 h at 45C . This treatment of oxidation-beta elimination also shortens the input RNA by 1 nt . RESULTS : Nucleolar Localization of U4 and U5 snRNA | U4 snRNA and subsequently U5 snRNA and U2 snRNA associate with U6 to form the [U4/U6] di-snRNP, [U4/U6.U5] tri-snRNP, and [U2.U4/U6.U5] tetra-snRNP . To complete the picture, we analyzed whether U4 and U5 snRNAs associate with nucleoli, like U6 and U2 . This was monitored by a technique used previously to define the nucleolar localization elements (NoLEs) of snoRNAs from various families (,,, ; ,) as well as nucleolar localization of U6 snRNA . Fluorescein-labeled in vitro transcripts were injected into Xenopus oocyte nuclei to allow direct visualization of the labeled RNA in nucleolar preparations made subsequently. In the present study, fluorescein-labeled in vitro transcripts of U4 snRNA and U5 snRNA were injected into Xenopus oocyte nuclei. Controls included injection of nucleolar U3 snoRNA as a positive control, and injection of a 40-nt synthetic RNA or fluorescein-UTP as negative controls to rule out nonspecific nucleolar staining (e.g., by diffusion of excess material). After 8 min, 1 h, and 24 h (the longest time point feasible to be studied), the oocytes were manually dissected and the nuclear contents, including nucleoli, were centrifuged onto a microscope slide. As shown in Figure a, strong fluorescent signals depicting nucleolar localization of U4 and U5 snRNA were detected 8 min and 1 h after injection of 0.8 ng of transcript per oocyte nucleus. Likewise, nucleoli were stained by a positive control (U3 snoRNA), injected at the same concentration that had been previously optimized for the nucleolar localization assay . In contrast, negative controls (40-nt synthetic RNA = control RNA, or unincorporated fluorescein-UTP), both injected in excess (see MATERIALS AND METHODS) did not stain nucleoli. Twenty-four hours after incubation, only U3 snoRNA retained strong signals in nucleoli, whereas U4 snRNA labeling was reduced and U5 snRNA signals were close to background (Figure a). This observation is consistent with the nucleolus being the functional compartment for U3 snoRNA and with the predicted transient nature of snRNA localization to nucleoli, as has been already shown for U6 snRNA in the same system . Figure 1 | Nucleolar localization of U4 and U5 snRNA. Nucleolar localization of U4 and U5 snRNA. (a) Fluorescein-labeled U4 snRNA or U5 snRNA were injected into the nuclei of X. laevis oocytes to analyze nucleolar localization. Controls included injection of nucleolar U3 snoRNA as well as synthetic RNA or fluorescein-UTP to rule out nonspecific nucleolar staining. After 8 min, 1 h, or 24 h, nuclear spreads were prepared and analyzed by phase contrast (PC) or fluorescence microscopy. Nucleoli can be distinguished from other nonchromosomal nuclear bodies by staining of the rDNA (DAPI, blue). Strong nucleolar labeling (FL, green) is seen as early as 8 min after nuclear injection of U4 or U5 snRNA and continues at 1 h after injection. Likewise, nucleoli were stained by a positive control (U3 snoRNA) at these time points, whereas negative controls (control RNA or unincorporated fluorescein-UTP), both injected in excess (see MATERIALS AND METHODS), did not stain nucleoli. Twenty-four hours after incubation only U3 snoRNA retains strong signals in nucleoli, whereas U4 snRNA labeling is reduced and U5 snRNA signals are close to background. Stronger nucleolar staining by U4 snRNA compared with U5 snRNA or U6 snRNA was generally observed at all time points. Bar, 10 mm. (b) 32P-Labeled transcripts of U4 and U5 snRNA were injected into oocyte nuclei to determine their stability; the RNAs were isolated and analyzed by gel electrophoresis as described previously . U2 snRNA transcripts were coinjected and served as an internal control to normalize for any differences in injection or recovery of the samples. U4 and U5 snRNA are stable 24 h after oocyte injection, compared to the 0-h control. Similarly, as published previously U3 snoRNA, U6 snRNA, or the control RNA were stable through 24 h. The nucleolar localization of fluorescent U4 and U5 snRNA was specific, because injection of an unrelated control RNA, even at similar3 times the molar amount of U4 or U5, did not stain nucleoli. Additional controls demonstrated that the observed fluorescent signals were not due to degradation of fluorescent snRNA and subsequent reutilization of the label by other nuclear components. First, injection of a similar75-fold molar excess of fluorescein-UTP alone did not label the nucleoli (Figure a). Second, stability assays using 32P-labeled transcripts demonstrated that U4 and U5 snRNA transcripts were stable at the times the nucleolar localization assay was performed (Figure b). Similarly, as previously published for U3 snoRNA, U6 snRNA, or the control RNA , U4 and U5 snRNAs were stable over 24 h. Sequences of U6 snRNA Essential for Nucleolar Localization | To define cis-acting elements of U6 snRNA necessary for nucleolar localization, the localization of mutant transcripts was compared with that of wild type. The scheme in Figure (modified from and ) points out areas of functional interest in wild-type U6 snRNA and also gives an overview of the various U6 mutants designed for the present study. Sequences of mature wild-type U6 snRNA and of the mutants are listed in detail below the scheme. Figure 2 | Sequence and mutations of U6 snRNA. Sequence and mutations of U6 snRNA. The sequence of Xenopus U6 snRNA , which is highly conserved across the eukaryotic kingdom , is shown. The figure has been modified from and . Shaded areas are sites of base pairing with U2 snRNA in the spliceosome and nt 49 --74 base pairs with U4 snRNA before association with the spliceosome. Several of the modifications (psi, pseudouridine; m, methylated nucleotides) occur in areas of U6 snRNA involved in intermolecular base pairing during splicing . Most mutations designed for this study were deletions covering the nucleotides indicated by lines. Delta1 --19 was designed to remove the 5'-terminal stem. Nucleotides used for substitution of nt 20 --25, which is a sequence needed for nuclear import after injection of U6 into the cytoplasm , are shown in a box. In various mutations of the middle part of the molecule, sites of base pairing with U2 and U4 snRNA were removed. In addition various deletions were introduced between nt 82 and 107, which includes regions that base pair in the spliceosome with U2 snRNA as well as the 3' terminus of U6 that binds the La protein and subsequently the Lsm proteins (see text). Furthermore, a triple repeat of nt 87 --107 of the U6 3' end (3' end) was designed (open box at top of diagram). The lower portion of this figure lists the sequences of mutant U6 snRNAs. Nucleotides that are the same as in wild-type U6 are shown by dots in the sequence alignment, and deletions are indicated by dashes. There is no previous information at which stage of its life cycle U6 enters nucleoli. It might enter nucleoli as the individual snRNP or in a complex as a di-snRNP or tri-snRNP directly bound to U4 and in conjunction with U5 snRNA, both of which are also found in nucleoli (see above; Figure a). Therefore, it was of particular interest to analyze whether the sites in U6 (nt 49 --74) that base pair with U4 before association with the spliceosome and that are important for snRNP assembly might play a role in U6 nucleolar localization. However, deletions within that sequence (Delta43 --56, Delta57 --81; Figure ) or total deletion of the entire middle part of the molecule, including the entire U4 binding site (Delta43 --81; Figure ), do not appreciably influence nucleolar localization of U6 snRNA compared with the wild type. This area also contains all sites for nucleolar posttranscriptional methylation of U6 as well as a region that base pairs with U2 snRNA in the spliceosome . Similarly, U6 snRNA carrying a deletion of the 5' stem (Delta1 --19) or a deletion of nt 26 --42 containing a U2-binding site retained the ability to localize to nucleoli . Therefore, areas of U6 snRNA that will base pair with U4 snRNA or later with U2 snRNA during splicing are not required for nucleolar localization. Interestingly, a deletion or substitution of nt 20 --25 does not affect U6 nucleolar localization . This sequence element is responsible for nuclear import of U6 snRNA after injection into the cytoplasm . Although U6 during maturation normally does not travel to the cytoplasm, a mutation of this region was used previously to study the effect of additional mutations on the natural nuclear retention of U6 by preventing nuclear reimport of molecules leaking into the cytoplasm . Figure 3 | Nucleolar localization of U6 snRNA mutated in various positions throughout the molecule. Nucleolar localization of U6 snRNA mutated in various positions throughout the molecule. Fluorescein-labeled U6 snRNA mutants were injected into the nuclei of X. laevis oocytes and 1.5 h later the localization assay performed. U6 snRNA carrying a deletion of the 5' stem (Delta1 --19) or a deletion or substitution of nt 20 --25 localized as well to nucleoli as the wild-type (WT) molecule (FL, green). Similarly, various mutants with deletions throughout the middle part of the molecule (Delta26 --42, Delta43 --56, Delta57 --81, and Delta43 --81) retained full ability to localize to nucleoli. In contrast, a deletion of nt 82 --107 completely abolished nucleolar localization. Other details as in Figure a. In contrast to the mutations described above, nucleolar localization of U6 could be completely abolished by a deletion of nt 82 --107 of U6 snRNA . This area contains the 3' terminus of U6, which binds the La protein and subsequently the Lsm proteins (see INTRODUCTION). The 3'-terminal U residues also may somewhat assist nuclear retention of U6, although after deletion of the 3' end the majority of U6 injected into Xenopus oocytes still remained in the nucleus even after long incubation times . The present study now demonstrates that the 3' end of U6 snRNA is essential for nucleolar localization, whereas the remainder of the molecule representing more than three-quarters of the U6 snRNA sequence, including sites absolutely required for [U4/U6] di-snRNP formation, splicing complex assembly, and splicing activity , lacks elements important for nucleolar localization. In additional control experiments , we confirmed that nucleolar localization is indeed mediated independently from [U4/U6] snRNP assembly because 1) the synthetic U6 snRNA and U4 snRNA transcripts generated here by in vitro transcription retain their functional activity to form a [U4/U6] snRNP, 2) the mutant U6 Delta43 --81 still localizes to nucleoli even though it lacks the U4 base-pairing sites and cannot interact with U4, and 3) the 3'-NoLE U6 mutant that does not localize to nucleoli (Delta82 --107) is still able to assemble in a [U4/U6] snRNP. The experiments leading to these observations involved coinjection of U6 snRNA (colabeled with [32P]UTP and fluorescein-UTP) with U4 snRNA transcripts (labeled with fluorescein-UTP) into U6- and U4-depleted Xenopus oocytes and subsequent coimmunoprecipitation of these snRNAs with an anti-Sm antibody. Because the Sm proteins are bound to U4 and not to U6 snRNA, precipitation of U6 occurs only if it is associated with U4 snRNA; this technique was used by to identify domains of U4 and U6 required for snRNP assembly. Figure shows that U6 WT and the NoLE 3'-mutant (Delta82 --107) but not U6 mutated in nt 43 --81 were coprecipitated with U4 when using an anti-Sm antibody. Moreover, in the converse experiment, U4 carrying a deletion of the base-pairing sites for U6 (Delta1 --18/56 --63) failed to precipitate wild-type U6 snRNA. Figure 4 | [U4/U6] snRNP assembly after nuclear injection. [U4/U6] snRNP assembly after nuclear injection. U6 snRNA (colabeled with [32P]UTP and fluorescein-UTP) and U4 snRNA transcripts (labeled with fluorescein-UTP) were coinjected into Xenopus oocytes depleted of endogenous U4 and U6 snRNA. After 4 h of incubation the functional ability of the in vitro transcripts to form a [U4/U6] snRNP was analyzed by immunoprecipitation from the nuclear lysate with an anti-Sm protein antibody. The equivalents of 10 nuclei/sample of the precipitated RNA (pellet) and 0.1 nuclei/sample of the supernatant (control for equal amounts injected) were analyzed on a denaturing 8% polyacrylamide, 8 M urea gel and by autoradiography. [U4/U6] snRNP assembly occurs between wild-type U6 snRNA (U6 WT) and wild-type U4 snRNA (U4 WT), also with the U6 NoLE mutant (Delta82 --107), and even with the truncated 3'end mutant (Delta82-107), consistent with . Ability of the transcripts to form a snRNP was disrupted by using either mutant U6 (Delta43 --81) or mutant U4 (Delta1 --18/Delta56 --63) that lack the sites to base pair with each other and that are important for snRNP assembly before association with the spliceosome . It is notable that this U6 mutant (Delta43 --81) is still capable of localizing to nucleoli (see Figure ). None of the samples were precipitated by beads coupled to control antibody. Although not the primary focus of the present study, we also analyzed association of U6 with Cajal bodies, identified by their immunostaining for coilin . One hour after nuclear injection of fluorescein-labeled wild-type U6 snRNA, the Cajal bodies are stained, although not as strongly as nucleoli . Four hours after injection, consistent with the previously reported kinetics of nucleolar localization of U6 snRNA , the nucleoli are weakly stained. At this time point, Cajal body staining by U6 snRNA remains at low levels (our unpublished data). It has been reported that snoRNA lacking the box C/D NoLE accumulates in Cajal bodies , and so we investigated whether a similar phenomenon occurred for U6 snRNA. However, U6 snRNA lacking its NoLE no longer localized to Cajal bodies or nucleoli . Therefore, the 3'-end of U6 snRNA is essential not only for localization to nucleoli but also to Cajal bodies. Figure 5 | The 3' end of U6 is essential for localization to Cajal bodies. The 3' end of U6 is essential for localization to Cajal bodies. Fluorescein-labeled wild-type U6 snRNA as well as the 3' mutant (Delta82 --107) were injected into the nuclei of Xenopus oocytes and the localization assay performed. Immunostaining of Cajal bodies (CB, red) was carried out in addition to DNA staining of nucleoli (DAPI, blue) to distinguish between these two sites of U6 snRNA localization (FL, green). Injection of wild-type U6 snRNA results in strong labeling of nucleoli and weaker labeling of Cajal bodies 1 h after injection. In contrast, deletion of the U6 3' end (Delta82 --107) completely abolished localization to nucleoli as well as to Cajal bodies, comparable with background levels seen in uninjected control oocytes (no injection). Other details as in Figure a. Further mutational analysis was carried out to reveal whether specific sequences of a few nucleotides within the 3' part of U6 are essential for nucleolar localization, similar to the discrete NoLEs of snoRNAs. Four mutations were designed that spanned the region of interest in U6 snRNA (Delta82 --95, Delta96 --99, Delta100 --102, and Delta103 --107) . In addition, a mutation spanning the last 8 nt (Delta100 --107) was used, which drastically decreased the interaction of U6 with the La protein in Xenopus oocytes . All five mutations of the 3' end of U6 (Delta82 --95, Delta96 --99, Delta100 --102, Delta103 --107, and Delta100 --107) significantly impaired localization . However, weak fluorescent signals were still obtained by all five U6 mutants and localization was not entirely abolished. This is in contrast to the U6 mutant, which lacks the entire 3' end (Delta82 --107) and which was entirely incapable of localization to nucleoli . Therefore, nucleolar localization of U6 snRNA relies on many nucleotides in the 3'-end region, rather than on just a few nucleotides. Figure 6 | Sequences within the 3' end of U6 snRNA that mediate nucleolar localization. Sequences within the 3' end of U6 snRNA that mediate nucleolar localization. Mutational analysis of the 3' portion of U6 snRNA was carried out to analyze whether specific sequences within this region are essential for nucleolar localization. Various deletion mutants of the 3' end of U6 (Delta82 --95, Delta96 --99, Delta100 --102, Delta103 --107, and Delta100 --107) were impaired in their nucleolar localization 1.5 h after injection. However, they still labeled nucleoli weakly, unlike the larger deletion of the 3' end of U6 (Delta82 --107; Figure ) that had no nucleolar signal. Therefore, nucleolar localization relies on many nucleotides in the 3' region. In addition, a triple repeat of the NoLE (3'-end) was tested in the nucleolar localization assay and revealed that nt 87 --107 are not only important but also sufficient for nucleolar localization because the construct was able to localize to nucleoli. Similarly, this 3'-end NoLE also localized to nucleoli when it was coupled to a synthetic control RNA (control RNA/3'-end), whereas the control RNA by itself did not. Other details as in Figure a. It was important to ascertain the stability of each mutant U6 snRNA transcript, to guard against the possibility that failure of some mutants to localize to nucleoli was simply due to their degradation. Stability assays using 32P-labeled transcripts demonstrated that all transcripts were sufficiently stable 1.5 h after injection into oocyte nuclei (the time when localization assays were carried out) . This included U6 mutant Delta82 --107 that failed to localize to nucleoli, as well as the control RNA. In addition, the use of U6 wild-type as well as U6 mutants (Delta43 --81 and Delta82 --107) in immunoprecipitation experiments confirmed their stability when labeled with fluorescein-12-UTP and [alpha-32P]UTP. Although well beyond the time frame of the localization assay, we were also interested to see whether any of the U6 mutants would show significant instability after longer incubation periods in oocytes. By assaying the long-term stability of transcripts 24 h after injection, we found that several U6 mutants of the 3' end as well as the 3'-end NoLE constructs (3' end, control RNA/3' end) were significantly less stable than wild-type U6 snRNA (our unpublished data). This precluded us from carrying out long-term localization studies. Figure 7 | Stability of wild-type and mutated U6 snRNA. Stability of wild-type and mutated U6 snRNA. 32P-Labeled U6 snRNAs (mutants or wild type or the 3'-end NoLE constructs) were injected into oocyte nuclei, and the RNAs were isolated and analyzed by 8% polyacrylamide, 8 M urea gel electrophoresis. The top panel shows the 0-h controls and the bottom panel shows the short-term stability at 1.5 h (the time when localization assays were carried out) after injection into oocyte nuclei. To determine the stability of the various RNAs after nuclear injection, U2 snRNA transcripts were coinjected and served as an internal control to normalize for any differences in injection or recovery of the samples. All the U6 mutants as well as the control RNA were stable at the time point used for analysis of nucleolar localization. For some snoRNAs it has been shown that the NoLEs not only are essential but also are sufficient for nucleolar localization because just the box C/D core structure of C/D snoRNAs can target synthetic RNA sequences to the nucleolus . Therefore, we analyzed if the 3' sequence of U6 snRNA by itself is not only essential but also sufficient for nucleolar targeting. As shown in Figure , a transcript of just the 3' end of U6 is sufficient for nucleolar localization. Similarly, this 3'-end NoLE construct also localized to nucleoli when it was coupled to a synthetic control RNA (control RNA/3' end), whereas the 40-nt control RNA by itself did not . Thus, the nucleotides that are both essential and sufficient for nucleolar localization are nt 87 --107, which represent the NoLE for U6 snRNA. Does the 3' End of U6 snRNA Function as a NoLE by Binding to the La Protein? | It has been hypothesized that NoLEs of snoRNAs act by binding protein(s) that either transport the snoRNA from the nucleoplasm to the nucleolus and/or anchor it within the nucleolus. Similarly, binding of specific protein(s) to the NoLE of U6 might initiate nucleolar localization. Deletion of the 8 nt of the U6 NoLE at the 3' end drastically decreases the interaction of U6 with the La protein in Xenopus oocytes . La is the first protein to associate with U6 snRNA, and later it is replaced by the evolutionarily conserved Lsm proteins, which leads to formation of the di- and tri-snRNPs (see INTRODUCTION). It has been shown that a 3'-end alteration modulates the binding of La protein to U6 snRNA . This alteration allowed us to find out whether interaction of U6 with La might be involved in U6 nucleolar localization. The La protein binds the stretch of uridylates at the 3'-hydroxyl end of newly synthesized U6 snRNA both in vivo and in vitro but is unable to bind to U6 snRNA that lacks a 3' hydroxyl, and conversion of the U6 3' end in vivo to a 2',3'-cyclic phosphate leads to a replacement of La by Lsm proteins . Consequently, we generated an in vitro transcript of U6 with an altered 3' end that was unable to bind La but would allow the nucleolar localization assay to be carried out to determine whether La plays a role in nucleolar localization. For alteration of the 3'-hydroxyl end, U6 snRNA was incubated with Whitfield's reagent; this treatment via oxidation-beta elimination converts the 3' hydroxyl to a 3'-phosphate group, resulting in a loss of La binding . Figure a shows migration on a polyacrylamide gel of U6 transcripts (with or without a 3'-hydroxyl group) that were colabeled with 32P for immunoprecipitation and with fluorescein for further studies in a nucleolar localization assay. The transcripts were coinjected into Xenopus oocytes, and binding to the La protein was analyzed by immunoprecipitation (Figure b) to confirm that only U6 snRNA with a 3'-hydroxyl group but not a 3'-phosphate group effectively binds to La and can be precipitated by anti-La antibodies. The same antibodies precipitated only traces of U3 snoRNA transcripts that were coinjected as a control, and the control serum did not precipitate any of the samples. The gel also shows that the transcripts remain stable over the time of incubation in the oocyte, which is equivalent to the time of incubation for the localization assay (1.5 h). Regardless of the alteration, both the 3'-OH and 3'-phosphate forms of U6 localized strongly to nucleoli and, although weaker, to Cajal bodies as well (Figure c), indicating that the mechanism of U6 localization to either nuclear subcompartment does not require a 3'-hydroxyl terminus nor binding of the La protein to the 3' NoLE. Figure 8 | U6 localization to nucleoli and Cajal bodies does not require a 3'-hydroxyl terminus nor binding of the La protein. U6 localization to nucleoli and Cajal bodies does not require a 3'-hydroxyl terminus nor binding of the La protein. (a) U6 snRNA in vitro transcripts with either a 3'-hydroxyl (3'-OH) or a 3'-phosphate (3'-P) group (colabeled with [32P]UTP and fluorescein-UTP [FL]) were run on a denaturing 8% polyacrylamide, 8 M urea gel and analyzed by autoradiography (left two lanes) and under UV light (right two lanes). Because the 3' modification via oxidation-beta elimination also results in shortening of the RNA by one nucleotide, the transcripts can be distinguished by their different migration. (b) Colabeled U6 snRNA (32P; FL) with either a 3'-OH or 3'-P group and a control (32P-labeled U3 snoRNA) were coinjected into Xenopus oocytes. One hour after injection, binding of the injected RNA to the La protein was analyzed by immunoprecipitation with anti-Xenopus La (La) or preimmune serum as a control (Ctr). The equivalents of five nuclei per sample of the precipitated RNA (pellet) and 0.2 nuclei per sample of the supernatant were analyzed by PAGE as in (a). Only U6 snRNA with a 3'-OH group but not a 3'-P group effectively binds to La and can be found in the precipitate. The control preimmune serum did not precipitate any of the samples. (c) Colabeled U6 snRNA transcripts with either a 3'-OH or 3'-P group were injected into the nuclei of X. laevis oocytes and the localization assay carried out. Both transcripts (FL, green) localized strongly to nucleoli (DAPI, blue) and with weaker signals to Cajal bodies (CB, red), indicating that the conversion of the 3'-OH group to a 3'-P group, which blocks La-binding does not affect localization to nucleoli or to Cajal bodies. Other details as in Figure a. DISCUSSION : All snRNA Components of [U4/U6.U5] Tri-snRNP Localize to Nucleoli | There are significant differences in the maturation and trafficking of U6 snRNA compared with other spliceosomal snRNAs, including U4 and U5 of the [U4/U6.U5] tri-snRNP, and it was unknown whether these snRNAs localize to nucleoli, similar to U6 snRNA . U6 is transcribed by RNA polymerase III and retained in the nucleus . In contrast, the other spliceosomal snRNAs are transcribed by RNA polymerase II and are exported to the cytoplasm where the 5' cap is converted from a monomethyl G (7mGpppG) to a trimethyl G (2,2,7mGpppG) and Sm proteins are bound; after these events these snRNAs are reimported back into the nucleus to function in splicing (for review, see ). The data presented herein demonstrate that U4 and U5 snRNAs, like U2 snRNA and U6 snRNA , associate with nucleoli. These conclusions are based on the nucleolar localization of injected synthetic T7 polymerase U4 and U5 snRNA transcripts with a monomethyl G cap like their newly synthesized in vivo counterparts. Previous studies showed that such transcripts microinjected into Xenopus oocytes exhibit normal nucleo/cytoplasmic traffic as their endogenous counterparts . In addition, functionality of the synthetic RNAs was shown by the specificity of their nucleolar localization and also confirmed by the ability of U5 (our unpublished data) and U4 snRNA to associate with Sm proteins and, moreover, by the capacity of U4 to form the [U4/U6] snRNP via base pairing. Injection of synthetic transcripts allows the visualization of RNA that transiently passes through the nucleolus. In contrast, detection by in situ hybridization of endogenous RNA can only detect steady state levels, which for U4 and U5 are weak at background levels for nucleoli in contrast to snurposomes . Injection of synthetic transcripts also allows a kinetic analysis of nucleolar localization, which showed that U4 snRNA labeling was reduced and U5 snRNA signals were close to background 24 h after incubation. This observation is consistent with the predicted transient nature of snRNA localization to nucleoli, as has been already shown for U6 snRNA in the same system . Furthermore, the kinetic analysis in the present study suggests that nucleolar localization of U4 snRNA and U5 snRNA can occur early during their maturation and before export to the cytoplasm. It seems that nuclear import after microinjection of various snRNAs, including U4 and U5 snRNA, into Xenopus oocytes is a time-limiting factor and consumes several hours , whereas efficient nuclear export occurs within 2 h (, and references therein). Even when U5 snRNA was equipped with a mature trimethyl G cap instead of a monomethyl G cap, efficient nuclear import still required 4 h of incubation . In the present study, nucleolar localization of injected U4 and U5 snRNA with a monomethyl G cap (same as after synthesis in vivo), however, took place within 8 min after nuclear injection. Thus, it is unlikely that the snRNAs traveled to the cytoplasm and were reimported into the nucleus within that short time frame. It is probable that the injected snRNAs localized to nucleoli before export to the cytoplasm. Because U4 and U5 snRNAs localize to nucleoli similar to U6 snRNA , it was possible that U6 snRNA might passively be carried to the nucleolus as part of the [U4/U6] di-snRNP or [U4/U6.U5] tri-snRNP. The present study rules out this possibility. First, we have shown herein that deletions of the proposed [U4/U6] interaction domain as well as the sequences flanking the interaction domain, which are essential for [U4/U6] snRNP assembly , do not affect nucleolar localization. Second, the 3' end of U6 is identified herein as the NoLE; it is not only essential but also sufficient for nucleolar localization and can be targeted to nucleoli by itself. The U6 3' end alone, however, is unable to assemble into a [U4/U6.U5] tri-snRNP . In contrast, the 3'-NoLE U6 mutant that does not localize to nucleoli is still able to assemble into a [U4/U6] snRNP ; similarly, 3'-truncated U6 can form a tri-snRNP . Although [U4/U6] snRNP formation is one prerequisite for splicing, recent studies have shown that regions of U6 essential for splicing exceed those essential for [U4/U6] snRNP assembly and cover most of the molecule, including its middle part and 3' terminus. An important conclusion of the present study is that the 3' sequence of U6, while equally important for further function in splicing, enjoys a role as a NoLE fundamentally distinct from internal sequences in U6 that are used for [U4/U6] snRNP formation. Similarly, functionally important regions of box C/D snoRNAs involved in pre-rRNA processing are distinct from the box C/D NoLEs needed for nucleolar targeting of these RNAs (,,). Thus, formation of a [U4/U6] snRNP is not a qualifying event for nucleolar localization because U6 snRNA seems fully capable of localizing to nucleoli independent of an association with U4 snRNA. Nevertheless, this capacity might be retained when U6 is part of the di-snRNP or tri-snRNP. Alternatively, it could be hypothesized that nucleolar localization of U6 snRNA requires association with the snoRNAs guiding its modification; they might anchor U6 snRNA in the nucleolus. Just as in the scenario discussed above, U6 snRNA would passively localize in nucleoli due to interactions with other molecules that themselves contained the NoLEs. We have ruled out this possibility because the U6 3' construct of nt 87 --107 (3' end) by itself can localize to nucleoli, although it does not include any of the sites to be methylated or pseudouridylated. Thus, the interaction of guide snoRNAs with regions to be modified in U6 snRNA cannot be essential for nucleolar localization of U6. In line with previous reports , we have also shown here that U6 snRNA localizes to Cajal bodies. The relationship between Cajal bodies and nucleoli is not yet fully understood. However, data from yeast suggest that box C/D snoRNAs pass through Cajal bodies when trafficking to nucleoli . Moreover, it has been reported that disruption of the snoRNA C/D motif seemed to block transfer from Cajal bodies to nucleoli . Those results suggest that the sequence essential for nucleolar localization of box C/D snoRNAs is not involved in localization to Cajal bodies. To date, only one sequence essential for RNA localization to Cajal bodies has been identified, namely, the Sm site in U7 , an snRNA that is not found in nucleoli. The present study shows that U6 snRNA after mutation of the 3' NoLE loses its ability to localize to both nucleoli and Cajal bodies, suggesting that this sequence may be both a Cajal body localization element (CaBLE) and a NoLE. It seems that the nucleolus is an organelle where specific internal posttranscriptional modifications of snRNAs (2'-O-methylation and psi) may be carried out by snoRNAs. Three box C/D snoRNAs have already been identified as guide RNAs for the 2'-O-methylation of U6, and other results suggest that factors needed to form 2'-O-methylation and psi of U6 snRNA function in the nucleolus . In addition, correlative evidence suggests that U2 snRNA modifications occur in the nucleolus rather than the Cajal bodies . Because the 3' end of U6 snRNA is required for localization to both nucleoli and Cajal bodies, it cannot be discerned whether snRNA modification occurs in just one or both of these organelles. Candidate Proteins That May Interact with the 3' NoLE of U6 snRNA | NoLEs of snoRNAs are recognized by proteins that might transport the snoRNA from the nucleoplasm to the nucleolus and/or anchor it within the nucleolus. Recent evidence supports this idea, because all four proteins associated with the box C/D motif are needed for efficient nucleolar localization of U14 snoRNA . Yeast box C/D snoRNAs interact with Srp40p in the Cajal body and then seem to be delivered by Nsr1p to the nucleolus . Similarly, the nucleolar localization of U6 might be mediated by proteins that assemble on its NoLE before engagement in a di- or tri-snRNP. Although, several proteins make direct contact with U6 snRNA during conversion from free U6 to the tri-snRNP , only the La protein and the Lsm protein complex bind to the 3' end of U6 snRNA (see INTRODUCTION), which is the NoLE essential for U6 nucleolar localization as shown in the present report. La is the first protein to associate with U6 snRNA upon transcription , and it also has been shown that deletion of the 8 nt at the U6 3' end decreases the interaction of U6 with the La protein and significantly reduces nucleolar localization (Delta100 --107; this study). Although La is found predominantly in nuclear speckles , it might transiently and/or in low amounts appear in nucleoli where it has been observed . Various RNA polymerase III transcripts, which require binding of La to their precursors to guarantee a normal pathway of maturation , are found in nucleoli , including pre-tRNAs and U6 . All these observations suggest that La could be involved in nucleolar localization of RNAs. It has even been proposed that La is needed to stabilize some RNA polymerase II transcripts like U4 , shown here to localize to nucleoli and also binds U3 snoRNA . Intriguing although the idea may be, in the present report we have ruled out a role of La for U6 snRNA nucleolar localization. We show herein that U6 snRNA that lacks a 3'-hydroxyl group and therefore cannot associate with the La protein localizes efficiently to nucleoli nonetheless. Therefore, we conclude that it is not La but instead is another factor that mediates nucleolar localization by binding to the 3' NoLE of U6. Similarly, it seems unlikely that La mediates nucleolar localization of certain other RNA polymerase III transcripts (e.g., SRP RNA, RNase P RNA; for review, see ). In the in vivo situation, during the maturation of U6 snRNA, La is replaced with the Sm-like protein complex Lsm2 --8, which is a likely candidate to play an important role in nucleolar localization of U6 snRNA. Interestingly, Lsm proteins also associate with pre-RNase P RNA , which is found in nucleoli . The Lsm complex binds to the 3' end of U6 snRNA with a 3'-hydroxyl end and to mature U6 snRNA whose 3' end is a 2',3'-cyclic phosphate . Similarly, we found that U6 snRNA with either a 3' hydroxyl or a 3' phosphate end localized to nucleoli. Moreover, no other part than the U6 3' end associates with the Lsm proteins stably enough to be coprecipitated with an anti-Lsm4 antibodies . Our finding that the 3' end of U6 is the NoLE is consistent with the hypothesis that the Lsm complex might mediate U6 nucleolar localization. Although the five uridine residues at the extreme 3' end are required for Lsm binding , they are not sufficient. Similarly, we found that U6 mutants with deletion of areas within nt 82 --107 showed weak although greatly diminished nucleolar localization, suggesting that the NoLE cannot be subdivided. Another protein, Prp24, interacts with Lsm proteins and binds to U6 snRNA, promoting formation of the [U4/U6] di-snRNP , but it cannot play an independent role in nucleolar localization of U6 because it binds to U6 at positions 39 --56 and 64 --76 that are not NoLEs as we have demonstrated here. Nucleolar localization of U6 snRNA via the 3' NoLE might occur after U6 engagement in a di- or tri-snRNP. However, in the more likely scenario, nucleolar localization is initiated for noncomplexed U6 before formation of the di- or tri-snRNP, because we have shown herein that association with U4 snRNA is not a prerequisite for localization. Therefore, the present report suggests the nucleolar entry of U6 snRNA at a point during its life cycle when Lsm proteins replace La in binding to the NoLE of U6 and before association of U6 with U4 in the di-snRNP. Backmatter: PMID- 12221121 TI - A Role for Caenorhabditis elegans Importin IMA-2 in Germ Line and Embryonic Mitosis AB - The importin alpha family of nuclear-cytoplasmic transport factors mediates the nuclear localization of proteins containing classical nuclear localization signals. Metazoan animals express multiple importin alpha proteins, suggesting their possible roles in cell differentiation and development. Adult Caenorhabditis elegans hermaphrodites express three importin alpha proteins, IMA-1, IMA-2, and IMA-3, each with a distinct expression and localization pattern. IMA-2 was expressed exclusively in germ line cells from the early embryonic through adult stages. The protein has a dynamic pattern of localization dependent on the stage of the cell cycle. In interphase germ cells and embryonic cells, IMA-2 is cytoplasmic and nuclear envelope associated, whereas in developing oocytes, the protein is cytoplasmic and intranuclear. During mitosis in germ line cells and embryos, IMA-2 surrounded the condensed chromosomes but was not directly associated with the mitotic spindle. The timing of IMA-2 nuclear localization suggested that the protein surrounded the chromosomes after fenestration of the nuclear envelope in prometaphase. Depletion of IMA-2 by RNA-mediated gene interference (RNAi) resulted in embryonic lethality and a terminal aneuploid phenotype. ima-2(RNAi) embryos have severe defects in nuclear envelope formation, accumulating nucleoporins and lamin in the cytoplasm. We conclude that IMA-2 is required for proper chromosome dynamics in germ line and early embryonic mitosis and is involved in nuclear envelope assembly at the conclusion of mitosis. Keywords: INTRODUCTION : The regulated distribution of proteins between the nucleus and the cytoplasm is critically important for maintenance of the cell cycle, differentiation of cells and tissues, and the development of a complete organism . The bidirectional movement of proteins between the cytoplasm and nucleus is due to intrinsic peptide sequences in each protein. The importin beta/karyopherin beta family of proteins specifically recognizes many of these sequences and chaperones the proteins between the two compartments through the nuclear pore complex . A subset of nuclear import factors known as the importin alphas recognizes proteins containing classical nuclear localization sequences (cNLSs) . This family of proteins shares two common structural features: a central armadillo repeat-containing domain that recognizes the cNLS and an amino terminal importin beta binding (IBB) domain that binds to the cargo carrier importin beta1 (for review, see ; ). The interaction of importin alpha with importin beta allows cNLS-containing proteins to be translocated across the nuclear pore complex, thus the importin alpha family can be thought of as a set of adapter proteins for transport. Various studies suggest that there is both redundancy and specificity in the recognition of cNLSs by the importin alpha family . The number of importin alpha genes increases with the complexity of the organism; budding yeast have a single gene, whereas humans have at least eight. Phylogenetic analyses group the importin alphas into three conserved clades, although several proteins cannot be assigned to any clades . The restriction of the alpha2 and alpha3 clades to animals suggests that these importins have specific roles in animal development . Recently, we described the identification of three importin alpha proteins in Caenorhabditis elegans: IMA-1, IMA-2, and IMA-3 (an alpha3) . IMA-1 and IMA-2 are among the small group of unusual importin alphas that cannot yet be classified into distinct clades. The key to regulation of the nuclear-cytoplasmic transport system is the small GTPase Ran and its associated factors that modulate nucleotide binding and hydrolysis . The Ran-GTPase network also regulates DNA replication, the exit from mitosis, microtubule polymerization, and accurate chromosome segregation in mitosis and meiosis . Although some aspects of this regulation may be simply a requirement to transport the necessary factors into the nucleus, other nontransport roles for Ran are now evident. Recent studies in Xenopus laevis egg extracts have shown that Ran-GTP modulates the release of factors that control mitotic spindle formation from importin alpha and importin beta and can direct nuclear envelope (NE) formation . Conditional mutations or depletion of the Saccharomyces cerevisiae importin alpha Srp1p result in a mitotic cell cycle arrest at G2/M accompanied by chromosome condensation and segregation defects . Mutations in one of the two Schizosaccharomyces pombe importin alphas, cut15, lead to mitotic progression without chromosome condensation, resulting in the septum bisecting the nuclear material, a characteristic "cut" phenotype . cut15-85 mutants do not have gross defects in nuclear protein import, suggesting that the cut phenotype is not due to failure in protein import. The Drosophila melanogaster importin alpha2 pendulin may also have a direct role in mitosis because it accumulates in embryonic nuclei at the onset of mitosis . Although these results have been suggestive of a nontransport role in mitosis for members of the importin alpha family, only the sequestration and release of TPX2 from importin alpha to regulate mitotic spindle assembly has been directly implicated in a mitotic function . MATERIALS AND METHODS : General Procedures and Nematode Strains | The wild-type N2 Bristol strain was maintained at 20C on NGM plates seeded with Escherichia coli OP50. Antibody Production | A polymerase chain reaction product encoding amino acids 512 --531 of IMA-2 was inserted in frame with the glutathione S-transferase (GST) protein of the pGEX-4T-1 vector (Amersham Biosciences, Piscataway, NJ). Recombinant GST-IMA-2 protein induction and purification on glutathione-Sepharose were performed as described in the manufacturer's instructions (Amersham Biosciences). HTI Bio-Products (Ramona, CA) prepared rabbit antiserum for the GST fusion protein. Anti-IMA-2 antibodies were affinity purified on full-length IMA-2 S peptide-tagged fusion proteins immobilized on polyvinylidene difluoride membranes. The specificity of the affinity-purified antibodies was determined by immunoblotting whole worm lysates and bacterial lysates containing expressed recombinant IMA proteins . Indirect Immunofluorescence Microscopy | Detection of IMA proteins in wild-type and RNA-mediated gene interference (RNAi) germ lines was performed on extruded gonads fixed in 1% paraformaldehyde for 4 min . All subsequent staining and washing steps were performed in Tris-buffered saline containing 0.1% Triton X-100 (TBST). Nucleoporins were detected with MAb414 diluted 1:4000 in TBST (Covance Research Products, Richmond CA). Affinity-purified anti-IMA-2 antibodies were diluted 1:200 in TBST. No staining was observed with the anti-IMA-2 antibodies in the presence of the immunogen. beta-Tubulin was localized with the monoclonal antibody (mAb) E7 as described previously . The E7 mAb developed by Michael Klymkowsky was obtained from the Developmental Studies Hybridoma Bank (Iowa City, IA) developed under the auspices of the National Institutes of Child Health and Human Development and maintained by The University of Iowa, Department of Biological Sciences, Iowa City, IA). Embryos extended from gravid hermaphrodites were fixed and stained with mAb414 and rat anti-LMN-1 (a generous gift of Drs. Kathy Wilson, Johns Hopkins University School of Medicine, Baltimore, MD, and Yosef Gruenbaum, The Hebrew University of Jerusalem, Jerusalem, Israel) as described previously . DNA was stained by inclusion of 0.1 mug/ml 4',6-diamidino-2-phenylindole (DAPI) or 1 mug/ml TOTO-3 iodide (Molecular Probes, Eugene, OR) in the final wash buffer. Immunofluorescence images were obtained with an LSM510 laser scanning confocal microscope (Carl Zeiss, Thornwood, NY). Images of DAPI-stained nuclei were obtained with either Eclipse E800 (Nikon, Melville, NY) or Axioskop (Carl Zeiss) microscopes equipped with digital cameras. The digital images were processed in MetaMorph, version 4.0 (Universal Imaging, Downington, PA) and Photoshop, version 5.0 (Adobe Systems, Mountain View, CA). RNA Interference Assays | Double stranded RNA (dsRNA) was generated from linearized plasmids of full-length EST yk96a12 in pBluescript II SK(-) by in vitro transcription with T3 and T7 RNA polymerases. After annealing, the double-stranded RNAs were microinjected into the intestines of L4 larvae at a concentration of either 0.5 or l mg/ml with equivalent results . As a control, distilled H2O was injected into the intestines of L4 larvae. Between 24 and 72 h postinjection, gonads were extruded and processed as described previously . For RNAi soaking experiments, RNA was prepared as described and L3 and L4 worms were soaked for 24 h at 18C . After soaking, the worms were washed in water and transferred to OP50-seeded plates. The P0s were transferred to new plates every 24 h. Embryos were obtained from the adults 24 h after soaking. RESULTS : Localization of IMA-2 | The three C. elegans importin alpha genes (ima-1, ima-2, and ima-3) are differentially expressed during development and each protein has a unique germ line localization. IMA-3 is expressed in both somatic and germ cells and reduced expression of IMA-3 stops the progression of germ cells through pachytene of meiotic prophase I . To understand the role of IMA-2 in the germ line, we first examined the localization of the protein in the adult hermaphrodite. Previously, we have demonstrated that ima-2 mRNA was weakly expressed in embryonic and larval stages but expression increased in L4 and adult animals. A mutant strain of C. elegans [glp-1(q224ts)] that do not develop a germ line as adults and did not express ima-2 mRNA when grown at the nonpermissive temperature indicated that ima-2 is a germ line intrinsic gene . Two independent genome-wide analyses of germ-line gene expression subsequently confirmed our identification of ima-2 as a germ line intrinsic gene . Affinity-purified antibodies to IMA-2 detected the protein only in germ line cells . In embryos the maternal IMA-2 was diluted during early cell divisions and was expressed at detectable levels only in the germ line precursor cells Z2 and Z3, not in the somatic cells (Figure , A and B). We have not determined when in the germ cell lineage ima-2 expression was activated. In the adult hermaphrodite germ line, IMA-2 was present within all germ cells from the distal end of the germ line to the proximal oocyte (Figure C). Note that IMA-2 was not detected in sperm (Figure C), consistent with the absence of an NE in these cells. In distal germ cells, IMA-2 was predominantly cytoplasmic and NE associated. However, in the developing oocytes IMA-2 was predominantly cytoplasmic and intranuclear with no apparent enrichment at the NE. When distal germ cells in prometaphase and metaphase were evident by DAPI staining of the DNA, IMA-2 was enriched at the region immediately surrounding the condensed chromosomes (Figure , D and E). IMA-3 was dispersed throughout the mitotic cells in the distal germ line with no obvious enrichment near the chromosomes (our unpublished data). Figure 1 | Immunofluorescence localization of IMA-2 in embryos and adults. Immunofluorescence localization of IMA-2 in embryos and adults. IMA-2 was localized by indirect immunofluorescence with affinity-purified antibodies to IMA-2. (A) IMA-2 localization in red in a tadpole stage embryo. (B) Overlay of IMA-2 staining in red and nucleoporin staining with MAb414 in green. (C) The two cells positive for IMA-2 are the germ line precursor cells Z2 and Z3. An extruded hermaphrodite germ line with IMA-2 is shown in red. The distal end is to the left and the spermatheca is to the right (indicated by the bracket). IMA-2 is cytoplasmic and nuclear in all germ cells but is not detected in sperm. IMA-2 (D) and DNA (E) show a higher magnification of a distal germ line with a germ cell in metaphase (arrow) to highlight the nuclear localization of IMA-2 at mitosis. IMA-2 Expression in Eggs and Early Embryos | Because early embryonic cells are larger than germ cells, we localized IMA-2 in early embryos by indirect immunofluorescence to better define the timing of nuclear association for IMA-2. The oocyte nucleus is in diakinesis of meiotic prophase I, completing meiosis only upon fertilization. After fertilization, the oocyte nucleus is positioned in the anterior end of the embryo and completes its two meiotic divisions, producing two polar bodies. The female and male pronuclei then form and move toward each other, meeting in the posterior half of the embryo. After the two pronuclei make contact and move back toward the center of the egg, the mitotic spindle rotates onto the A-P axis, the NEs break down, and the chromosomes move to the metaphase plate. Mitosis progresses through anaphase and is completed with NEs reforming around each set of chromosomes at telophase . In the fertilized egg before pronuclear migration, IMA-2 was predominantly cytoplasmic and NE associated in both pronuclei . The intranuclear IMA-2 in the proximal oocyte had dispersed into the oocyte cytoplasm at NE breakdown (our unpublished data), but returned to the NE upon NE reformation around the pronuclei. Because sperm do not contain IMA-2 , the male pronuclear IMA-2 must have originated in the egg cytoplasm. As the pronuclei became associated, the intensity of IMA-2 staining at the NE decreased. After pronuclear fusion, IMA-2 completely filled the space surrounding the chromosomes but did not seem to be enriched on the surface of the condensed chromosomes (Figure C). This perichromosomal localization persisted during congression of the chromosomes toward the metaphase plate but had decreased by the time of metaphase plate formation. Early in anaphase as chromosome separation first became apparent, IMA-2 staining could still be seen in the region surrounding the chromosome, but was significantly decreased compared with earlier stages. Later in telophase, IMA-2 again became associated with the NE as soon as the structure was detectable around the daughter nuclei. IMA-2 staining persisted at the NE through cytokinesis and into interphase of the next cell cycle. The mitotic IMA-2 staining was not coincident with beta-tubulin staining in the spindle microtubules or the spindle poles. Figure 2 | Immunofluorescence localization of IMA-2 in fertilized zygotes. Immunofluorescence localization of IMA-2 in fertilized zygotes. IMA-2 localization is shown in red, tubulin in green, and DNA in blue. Each panel is a different embryo with the anterior end of the embryo oriented to the left. The series shows the progression through pronuclear formation after fertilization (A), pronuclear meeting and spindle rotation (B), pronuclear fusion/prometaphase (C), metaphase (D), early anaphase (E), and telophase and NE reformation (F). The embryos shown are representative of typical localization patterns. We further investigated the localization of IMA-2 in two-cell embryos to see whether the mitotic localization of IMA-2 was preserved in later cell divisions. Two mechanisms could explain the accumulation of IMA-2 around the chromosomes at mitosis. IMA-2 could enter the nucleus before NE breakdown by entry through the nuclear pores or it could surround the chromosomes by interaction with another component after NE breakdown. Mitosis in C. elegans is unique in that the NE becomes permeable to proteins only in late prometaphase and only fully disassembles during anaphase in early embryos . To define the timing of NE breakdown in our experiments, we immunolocalized IMA-2 along with nucleoporins or beta-tubulin in two-cell embryos. As seen in Figure , IMA-2 surrounded the chromosomes only in nuclei that were in late prometaphase or later (Figure C). Nuclei containing fully condensed chromosomes, but in which the microtubules had not penetrated, excluded IMA-2 (Figure A). Note, however, that by this point IMA-2 was only weakly detected at the NE. Nuclei in which peripheral nucleoporin staining could be detected did not accumulate IMA-2 around the chromosomes (Figure B). The localization patterns for IMA-2 presented in Figures and suggest that IMA-2 entered the nucleus only after the point when the NE was permeable to proteins. Figure 3 | Immunofluorescence localization of IMA-2 in two-cell embryos. Immunofluorescence localization of IMA-2 in two-cell embryos. (A) Fixed and permeabilized embryos stained for IMA-2 in red, tubulin in green, and TOTO-3 --stained DNA in white. (B) IMA-2 is red, nucleoporins are in green, and DNA is in white. In A, the cell to the left is in metaphase and the cell to the right is in late prophase (arrow), and the microtubules have not penetrated the NE. In B, the cell to the left is in early anaphase and the cell to the right (arrow) is in late prophase. IMA-2 RNA Interference | Gene expression can be specifically down-regulated in C. elegans by dsRNA interference . We reduced the expression of ima-2 by injection of the dsRNA into the intestines of L4 or young adult worms. Between 24 and 72 h postinjection, fixed and extruded germ lines were stained with DAPI to visualize the chromosomes . The hermaphrodite germ line is organized with the cells most distal to the uterus forming a mitotic stem cell population. Between 10 and 20 cell diameters away from the distal tip cell, the germ cells initiate meiosis I in a region called the transition zone. Proximally to the transition zone, germ cells progress from pachytene through diakinesis of meiotic prophase with some of the cells eventually developing into oocytes . ima-2(RNAi) had a dramatic effect on germ line morphology in 25% of the injected animals. In these affected animals, the distal germ line was highly disorganized with DAPI-staining material occupying the central core. The germ line contained fewer germ cells than control animals and the germ cell chromatin in the distal arm had an abnormal appearance. The nuclei in the distal end seemed to have become aneuploid with both larger and smaller than normal DAPI-staining bodies present. In spite of this dramatic effect on the mitotic germ line, some germ cells seemed to enter meiosis normally. These germ cells had likely entered the meiotic cell cycle before the full effect of the RNAi. Proximal to the transition zone, most of the germ cells had normal pachytene morphology with condensed cable-like chromosomes at the nuclear periphery, although several small highly condensed DAPI-staining masses were also interspersed throughout this region. The developing oocytes had the normal complement of six bivalent chromosomes and both the nuclear and cytoplasmic oocyte volumes increased normally. These observations and the fewer number of germ cells in the ima-2(RNAi) worms compared with control animals suggested that germ cells that initiated meiosis before full penetrance of the ima-2(RNAi) were fertilized but not fully replenished. The majority of injected germ lines had no obvious morphological nuclear defects but produced nonviable embryos with aneuploid nuclei (see below). Figure 4 | Phenotype of ima-2(RNAi). Phenotype of ima-2(RNAi). Extruded germ lines from control injected (A) or ima-2(RNAi) (B --E) germ lines were stained with DAPI to visualize the DNA. (C --E) Higher magnifications of the mitotic region, pachytene region, and oocytes, respectively. The nuclei in the ima-2(RNAi) germ line contain unequal amounts of DNA. Note also that the regular organization of the mitotic germ cells is disrupted. The F1 progeny from the injected hermaphrodites exhibited 97% embryonic lethality before the 200-cell stage . A small number of progeny produced shortly after RNA injection died as larvae (0.1%) or survived to adulthood (3%). All of the embryos that developed into adults were produced in the first 18 h post dsRNA injection, suggesting that the RNAi phenotype was fully penetrant by 18 h. Of the worms that survived to adulthood, 27% (55/204) were phenotypically wild type. The remainder of the F1 adults had germ line mitotic chromatin defects similar to the injected animals and produced embryos with aneuploid nuclei. Table 1 | Phenotypes of ima-2(RNAi) F1 Progeny Embryos obtained from adults treated for RNAi by soaking exhibited greater than 99% lethality either before or early in gastrulation. The terminal phenotype of the F1 embryos was characterized by severe aneuploidy (Figure , A and B) making it difficult to determine the precise point of arrest. The nuclei displayed unequal amounts of DNA and some areas of the embryos seemed to lack DNA entirely, indicative of a chromosome segregation defect. We localized microtubules and DNA in very early embryos with anti-beta tubulin antibodies and with DAPI to determine the state of the chromatin and organization of the mitotic spindles. Most embryos contained disorganized chromosome masses between the spindle poles (Figure , C --F). The DNA in these embryos was not uniformly stained with DAPI, suggesting different states of chromosome condensation and we rarely observed clearly individualized chromosomes. At telophase, some chromosome masses failed to separate to daughter nuclei and the chromosomes remained as single masses of chromatin. In some embryos, two unequal chromosome masses separated, frequently connected by a chromatin bridge or trailing a strand of chromatin. We observed a small number of embryos with the chromosomes either peripherally associated with the spindle or completely separated from the spindle (Figure , G --J). In these embryos, the DNA mass was frequently bisected by the plasma membrane between two daughter cells. In most ima-2(RNAi) embryos examined, the mitotic spindles seemed structurally normal and were properly oriented, although an occasional embryo was seen with a chromosome mass associated with multiple spindles (our unpublished data). Some asynchrony in the well-defined embryonic cell division pattern may have also occurred resulting in early embryos with an abnormal appearance (Figure , E and F and I and J). Figure 5 | Terminal phenotype of ima-2(RNAi) embryos. Terminal phenotype of ima-2(RNAi) embryos. Fixed and permeabilized embryos were stained with DAPI to visualize the DNA. (A) Embryos from hermaphrodites soaked in RNAi buffer alone. (B) ima-2(RNAi) embryos. Embryos in A and B were obtained from age-matched hermaphrodites and were extruded 24 h postsoaking. The embryos in C --J were obtained from ima-2(RNAi)-injected hermaphrodites 48 h postinjection. C and D, E and F, G and H, and I and J are four different embryos stained for DNA with DAPI (C, E, G, and I) or tubulin (D, F, H, and J). Note that in G and I, the DNA is not associated with the spindles (arrows). In I, the arrow indicates a chromatin bridge, and the arrowheads indicate what seem to be individual chromosomes. The disorganized state of the chromatin in the ima-2(RNAi) embryos led us to examine the NE by indirect immunofluorescence to determine whether each chromatin mass was surrounded by an NE. ima-2(RNA)i embryos obtained by soaking were collected 24 h after treatment. Immunostaining with antibodies to nucleoporins (Figure , A --D) or the C. elegans lamin LMN-1 (Figure , E --F) revealed incompletely formed NEs and mislocalized nucleoporins and lamin in the embryos. Some chromatin masses seemed to be completely surrounded by nucleoporins, whereas others had very weak perinuclear nucleoporin staining or were associated with large immunoreactive spots and patchy nuclear staining. In control embryos, the NE as defined by nucleoporin or lamin localization is closely apposed to the surface of the chromatin (Figure , A and E). Frequently when nuclei in ima-2(RNAi) embryos had what seemed to be a complete NE, the NE was detached from the surface of the chromatin mass (Figure B). Immunostaining with rat anti-LMN-1 antibodies showed a similar lack of a complete NE around individual chromatin masses. These results indicate that IMA-2 is involved in NE assembly after mitosis. Figure 6 | Nuclear envelope assembly in ima-2(RNAi) embryos. Nuclear envelope assembly in ima-2(RNAi) embryos. Fixed and permeabilized embryos were stained for nucleoporins with mAb414 (A --D) or lamin with rat anti-LMN-1 (E and F). Nucleoporins and lamin are in green. DNA was stained with TOTO-3 in red. A and E are representative wild-type embryos. B --D and F --H are representative ima-2(RNAi) embryos. The arrow in B points to a separation of the nuclear envelope from the chromatin. Both nucleoporins and lamin accumulate in the cytoplasm of the treated embryos. Note also the more intense DNA stain in nuclei of ima-2(RNAi) embryos. DISCUSSION : IMA-2 Has Multiple Roles in Mitotic Cell Cycle | Adult C. elegans express three importin alpha proteins: IMA-1, IMA-2, and IMA-3. IMA-3 is expressed in somatic and germ line cells, whereas IMA-1 and IMA-2 expression is confined to the germ line . The expression of multiple members of the importin alpha family of nuclear transporters within the same cells and tissues suggests a redundancy of function as has been described for the importin beta family in yeast . This redundancy could mask certain phenotypes in RNAi or other loss of function experiments, but also may reveal unique roles for each protein. Within the germ line, IMA-2 is present in mitotic and meiotic germ cells, suggesting that it has a role in transporting proteins involved in both types of cell cycle. We did not observe any defects in meiosis in our experiments, suggesting that either IMA-2 does not play a critical role in meiosis or that IMA-1 or IMA-3 can compensate for the reduction in IMA-2 levels. In contrast, ima-3(RNAi) leads to a block early in meiosis but does not have a discernable effect on mitosis . The dramatic defects in the mitosis resulting from ima-2(RNAi) were most apparent in embryos. The terminal embryonic phenotype was arrest before or early in gastrulation characterized by severe aneuploidy of the embryonic cells. The unequal amounts of DNA between cells as well as the appearance of chromatin bridges and lagging chromosomes are consistent with a chromosome segregation defect. In the most extremely affected embryos, the partially condensed DNA mass could be seen in a single cell, completely separated from any spindles. At this time, we cannot determine whether the chromosome segregation defect is due to a failure in the nuclear transport function of IMA-2 or is related to its localization at mitosis, or a combination of the two. ima-2 has also been a target in two large-scale analyses of gene function by RNA interference. These studies have also found that ima-2(RNAi) results in embryonic lethality before the 200-cell stage . Conservation of Importin alpha Functions | Phylogenetic analyses of importin alpha protein sequences from plants, fungi, and animals group the proteins into three clades. IMA-2 cannot be placed into any of the three clades of importin alpha genes, indicating the protein sequence has likely diverged from the importin alphas of other metazoans . Despite this apparent divergence in sequence, some functional characteristics of IMA-2 may have been conserved with importin alpha proteins in other organisms. Drosophila melanogaster pendulin (an alpha2) and S. pombe Cut15p (an alpha1) exhibit a similar mitotic nuclear localization. Pendulin accumulates rapidly in the nucleus at the onset of prophase of early embryos . The intensity of nuclear staining decreases through metaphase and anaphase and increases again at telophase. During decondensation of the chromatin as the cells enter interphase, pendulin redistributes into the cytoplasm. The S. pombe Cut15 polypeptide has a similar transient nuclear localization . Cut15-GFP has strong NE localization throughout the cell cycle with intranuclear localization increasing through mitosis, reaching a maximum at prometaphase/metaphase. The intranuclear Cut15p then decreases through the remainder of mitosis. IMA-2 is distributed between the cytoplasm and the NE during interphase in the C. elegans embryo. The NE-associated protein disappears early in prophase followed by a rapid accumulation of IMA-2 in the region surrounding the mitotic chromosomes during prometaphase. Through metaphase and anaphase the localization of IMA-2 around the chromosomes decreases until telophase when IMA-2 again localizes strongly to the NE. A mutant allele of cut15, cut15-85, and the reduction of IMA-2 expression both result in chromosome segregation defects . The phenotype of cut15-85 cells is similar to the phenotypes of top2 and cut14 mutant cells, suggesting that Cut15p might interact with topoisomerase II or condensin inside the nucleus during mitosis. The cut15-85 allele is synthetically lethal with three alleles of cut3, a condensin subunit, supporting a direct role for Cut15p in chromosome condensation. Cut3p is localized to the nucleus in cut15-85 cells at the nonpermissive temperature, supporting the interpretation that Cut15p may interact with the condensin complex inside the nucleus. RNAi of the C. elegans Cut3 homolog F35G12.8 results in a "cross-eyed" phenotype in which the daughter nuclei stay close to the central cortex and multiple nuclei form in each daughter blastomere . This phenotype is associated with lagging chromosomal material that retains daughter nuclei together at the central cortex. A similar chromatin morphology is seen in some ima-2(RNAi) embryos. The single S. cerevisiae homolog Srp1p has been reported to have a role in regulation of the ubiquitin-proteasome system distinct from its role in nuclear protein import . The srp1-31 allele arrests at the G2/M boundary and arrested cells are unable to degrade the cell cycle regulator Clb2p . Depletion of Srp1p results in a terminal phenotype very similar to the terminal phenotype of IMA-2 depletion; abnormal appearance of mitotic chromatin, and dissociation of the chromosomes from the mitotic spindle . Together, the mitotic localization of the yeast, nematode, and fly importin alphas and the phenotypes associated with depletion or mutation of these proteins argue for a direct role for the importin alphas in mitosis distinct from their protein transport functions. Our results presented herein show that a role for importin alpha proteins in mitotic chromosome behavior is conserved in animals and yeast. Association of IMA-2 with Nuclear Mitotic Structures | The localization of IMA-2 around mitotic chromosomes in prometaphase and later suggests that IMA-2 may have a mitotic role distinct from its role as a nuclear transporter. Permeabilization of the NE in C. elegans embryos occurs late in prophase without significant loss of nucleoporins or lamin from the NE . Because IMA-2 is not detected around the chromosomes until prometaphase, it is likely that the protein is entering the nucleus by diffusion after early fenestration of the NE. The accumulation of IMA-2 around the chromosomes at metaphase cannot simply be due to constraining the protein within the partially disassembled NE. have reported that nucleoporins and lamins are present in the NE of early embryos until anaphase; however, in our hands, nucleoporins and lamin are undetectable in the NE by metaphase (Figures and ; our unpublished data). Alternatively, the rapid appearance of IMA-2 in the prometaphase nucleus could be the result of rapid accumulation of the protein during a transport event immediately before fenestration of the envelope, similar to the accumulation of cdc2/cyclin B1 . The sudden nuclear accumulation and gradual dissipation of the accumulated IMA-2 to the cytoplasm during metaphase and anaphase suggest that IMA-2 may associate with other factors important for mitosis that are "anchored" near the mitotic spindle. Recently, roles for importin beta and importin alpha in regulating promoting spindle formation have been described previously . A local high concentration of RanGTP near the mitotic chromosomes is believed to release the spindle-promoting factors from the importins . However, we do not observe any defects in mitotic spindle formation when IMA-2 levels are reduced. If IMA-2 has a direct role in sequestering factors that regulate mitosis, this role must be distinct from a role in spindle formation. The localization of IMA-2 near the condensing chromosomes at the onset of mitosis suggests that the protein is important for an early mitotic event, possibly interacting with another factor or factors involved in chromosome condensation or kinetochore formation. The localization and timing of IMA-2 appearance around the mitotic chromosomes is in some ways similar to Skeletor, a protein that has been suggested to be a component of a spindle matrix . Whether IMA-2 is associated with a spindle matrix-type structure will be an area for active future investigation. IMA-2 in Nuclear Envelope Formation | One of the most dramatic effects of the ima-2(RNAi) was the inability to reform correct NE structures after mitosis. Immunolocalization of nucleoporins and lamins demonstrated that both NE components are mislocalized in the ima-2(RNAi) embryos. This could be the result of the failure to transport the sole C. elegans lamin LMN-1 into the forming nucleus at telophase and into interphase. Because nuclear envelope assembly is a coordinated process to assemble a lamina, membrane, and pore complexes, failure to assemble an adequate lamina would also result in a failure to assemble pore complexes, resulting in their accumulation in the cytoplasm. A failure to import an adequate amount of LMN-1 to maintain a functional lamina could also result in a failure to complete replication or organize the nucleus correctly, leading to defects in chromosome condensation and segregation . Videomicroscopic analysis of ima-2(RNAi) embryo phenotypes identified a possible defect in NE formation in early embryos . The NE defects observed could also be an indirect effect of the failure to transport other factors that are required upstream of nuclear envelope assembly. In support of a direct role for IMA-2 in LMN-1 transport, we identified a fragment of LMN-1 containing the putative NLS in a two-hybrid screen with IMA-2. In vitro binding assays indicate that both IMA-2 and IMA-3 are capable of directly binding the NLS of LMN-1 in solution (our unpublished data). lmn-1(RNAi) results in similar chromosomal and pore complex organizational defects as seen for ima-2(RNAi) . The Ran GTPase network is required for NE assembly at the conclusion of mitosis . It has recently been demonstrated that importin beta is required for NE assembly induced by Ran in Xenopus egg nuclear assembly assays . Additionally, the Drosophila Ketel dominant negative mutations in the importin beta gene block the formation of the NE in cleavage nuclei . Importin beta provides a link between nuclear transport, mitotic spindle formation, and NE assembly all tied to regulation by the RanGTPase network. The importins are believed to act as chaperones to sequester spindle assembly factors or NE components until they can be released at the proper time by Ran-GTP (for review, see ). Because the importin alphas are complexed to importin beta through their IBB domain, they are a component of this general mechanism. A role for an importin alpha in sequestering the spindle assembly factor TPX2 in Xenopus has been described recently . It is likely that IMA-2 is operating in mitosis through a similar mechanism. Because IMA-2 cannot be grouped into a phylogenetic clade with any of the other identified importin alphas, it will be interesting to determine whether any of the importin alphas from higher animals have retained the localization pattern and functions of IMA-2. Backmatter: PMID- 12221122 TI - The Major Sites of Cellular Phospholipid Synthesis and Molecular Determinants of Fatty Acid and Lipid Head Group Specificity AB - Phosphatidylcholine and phosphatidylethanolamine are the two main phospholipids in eukaryotic cells comprising similar50 and 25% of phospholipid mass, respectively. Phosphatidylcholine is synthesized almost exclusively through the CDP-choline pathway in essentially all mammalian cells. Phosphatidylethanolamine is synthesized through either the CDP-ethanolamine pathway or by the decarboxylation of phosphatidylserine, with the contribution of each pathway being cell type dependent. Two human genes, CEPT1 and CPT1, code for the total compliment of activities that directly synthesize phosphatidylcholine and phosphatidylethanolamine through the CDP-alcohol pathways. CEPT1 transfers a phosphobase from either CDP-choline or CDP-ethanolamine to diacylglycerol to synthesize both phosphatidylcholine and phosphatidylethanolamine, whereas CPT1 synthesizes phosphatidylcholine exclusively. We show through immunofluorescence that brefeldin A treatment relocalizes CPT1, but not CEPT1, implying CPT1 is found in the Golgi. A combination of coimmunofluorescence and subcellular fractionation experiments with various endoplasmic reticulum, Golgi, and nuclear markers confirmed that CPT1 was found in the Golgi and CEPT1 was found in both the endoplasmic reticulum and nuclear membranes. The rate-limiting step for phosphatidylcholine synthesis is catalyzed by the amphitropic CTP:phosphocholine cytidylyltransferase alpha, which is found in the nucleus in most cell types. CTP:phosphocholine cytidylyltransferase alpha is found immediately upstream cholinephosphotransferase, and it translocates from a soluble nuclear location to the nuclear membrane in response to activators of the CDP-choline pathway. Thus, substrate channeling of the CDP-choline produced by CTP:phosphocholine cytidylyltransferase alpha to nuclear located CEPT1 is the mechanism by which upregulation of the CDP-choline pathway increases de novo phosphatidylcholine biosynthesis. In addition, a series of CEPT1 site-directed mutants was generated that allowed for the assignment of specific amino acid residues as structural requirements that directly alter either phospholipid head group or fatty acyl composition. This pinpointed glycine 156 within the catalytic motif as being responsible for the dual CDP-alcohol specificity of CEPT1, whereas mutations within helix 214 --228 allowed for the orientation of transmembrane helices surrounding the catalytic site to be definitively positioned. Keywords: INTRODUCTION : Phospholipids are the major components of cellular membranes. Phosphatidylcholine (PtdCho) and phosphatidylethanolamine (PtdEtn) are the two most abundant phospholipids present in eukaryotic cell membranes, comprising similar50 and 25% of phospholipid mass, respectively . Both PtdCho and PtdEtn are actively metabolized through both agonist stimulated and constitutive processes to release a plethora of biologically active molecules, including arachidonic acid for the synthesis of the inflammatory mediators prostaglandins and leukotrienes, and diacylglycerol for the activation of signaling molecules, which include members of the protein kinase C family . In the face of this complex metabolic regulation the levels of PtdCho and PtdEtn remain essentially unchanged, because the cell is capable of responding to alterations in lipid catabolism through increases in synthesis and subsequent transport of lipids from their site of synthesis to the intracellular destination at which the lipid has been catabolized. Indeed, in any model of intracellular lipid homeostasis one must couple lipid metabolism at a particular site within the cell to increased de novo lipid synthesis and targeted transport of the lipid to the site of its catabolism. The regulation of lipid transport may overlap with vesicle trafficking processes; indeed, the major cellular phospholipid PtdCho is believed to be inhibitory to fission of vesicles from the Golgi , whereas its metabolites phosphatidic acid and diacylglycerol appear to be positive regulators of vesicle transport . Research into the transport of the major membrane lipids PtdCho and PtdEtn has been hampered by a lack of knowledge with regard to precise sites of synthesis of these lipids within the cell. PtdCho is synthesized almost exclusively through the CDP-choline pathway in all mammalian cell types (except the liver; Figure ; ; ; ). The first step in the synthesis of PtdCho through the CDP-choline pathway is the phosphorylation of choline by a cytoplasmic choline kinase to form phosphocholine. In the second step, CMP is transferred from CTP to phosphocholine to form CDP-choline by the rate-limiting CTP:phosphocholine cytidylyltransferase (CT). There are two isoforms of CT found in mammalian cells, with CTalpha being ubiquitous and likely the major contributor to PtdCho synthesis in most cell types, whereas the second CT isoform, CTbeta, is found in a much more restricted tissue distribution. . CTalpha is directed to the nucleus in most mammalian cell types including the Chinese hamster ovary cells (CHO) cells used in this study by a basic motif found in its N-terminal region, whereas CTbeta is extranuclear in all cell types thus far examined . The final step in the synthesis of PtdCho is catalyzed by a cholinephosphotransferase activity that transfers phosphocholine from CDP-choline to diacylglycerol to form PtdCho . The intracellular location of the cholinephosphotransferase reaction defines the site of PtdCho synthesis. Figure 1 | Synthesis of PtdCho and PtdEtn by the CDP-alcohol pathways. Synthesis of PtdCho and PtdEtn by the CDP-alcohol pathways. The gene names of the enzymes that catalyze each step are indicated. EKI, ethanolamine kinase; ET CTP:phosphocholine cytidylyltransferase; CEPT1, choline/ethanolaminephosphotransferase; CKI, choline kinase; CT, CTP:phosphocholine cytidylyltransferase; CPT1, cholinephosphotransferase. PtdEtn is synthesized through an analogous pathway using a soluble ethanolamine kinase to produce phosphoethanolamine, which is converted by an endoplasmic reticulum bound CTP:phosphoethanolamine cytidylyltransferase to CDP-ethanolamine. Phosphoethanolamine is transferred from CDP-ethanolamine to diacylglycerol by an ethanolaminephosphotransferase activity of unknown intracellular location to produce PtdEtn. . PtdEtn can also be synthesized by the decarboxylation of phosphatidylserine with the contribution of the CDP-ethanolamine versus phosphatidylserine decarboxylation pathways being cell type dependent . Our laboratory recently identified the first mammalian cholinephosphotransferase and ethanolaminephosphotransferase encoding cDNAs . CPT1 codes for a CDP-choline specific cholinephosphotransferase for the exclusive synthesis of PtdCho, whereas CEPT1 codes for a dual specificity choline/ethanolaminephosphotransferase that can use both CDP-choline and CDP-ethanolamine as substrates to synthesize both PtdCho and PtdEtn. An analysis of the human genome and those of other eukaryotic cells indicates that these two genes likely code for the entire set of cholinephosphotransferases and ethanolaminephosphotransferases in mammalian cells. Genetic inactivation of the analogous genes in yeast resulted in complete loss of both enzyme activities in vitro and an inability to metabolically reconstitute lipid synthesis by either route in vivo . Human CEPT1 and CPT1, by virtue of their choice in diacylglycerol fatty acyl and CDP-alcohol species, affect the form and function of PtdCho and PtdEtn . In this study we have pinpointed the sites of PtdCho and PtdEtn synthesis in mammalian cells. We have also taken advantage of the dual CDP-alcohol specificity of CEPT1 to identify amino acid residues that determine its ability to synthesize PtdCho versus PtdEtn and have identified amino acid residues that alter its diacylglycerol fatty acyl species preference. MATERIALS AND METHODS : Materials | [Methyl-14C]Cytidine 5'-diphosphocholine and [ethanolamine 1,2-14C]cytidine 5'-diphosphoethanolamine were purchased from American Radiolabeled Chemicals (St. Louis, MO). [Methyl-14C]-choline chloride was purchased from NEN Life Science Products (Boston, MA). [Ethanolamine 1,2-14C]-ethanolamine hydrochloride was purchased from ICN (Costa Mesa, CA). All materials used in the preparation of bacterial and yeast media were purchased from Difco Laboratories (Detroit, MI). Lipids were purchased from Avanti Polar Lipids (Birmingham, AL). The T7 mouse mAb and T7 mouse mAb conjugated to horseradish peroxidase were purchased from Novagen (Madison, WI). The anticalnexin rabbit polyclonal antibody was a product of StressGen Biotechnologies Corp (Victoria, British Columbia, Canada). The Lap-2 mouse mAb was purchased from Transduction Laboratories (Lexington, KY). The CTalpha rabbit polyclonal antibody was a gift from Dr. Martin Post (Hospital for Sick Children, Toronto, Canada). FITC-labeled Lens culinaris (LcH) lectin was purchased from Sigma (St. Louis, MO). Goat anti-mouse Texas Red, goat anti-rabbit Texas Red, goat anti-mouse FITC, and goat anti-rabbit FITC secondary antibodies were products of Molecular Probes (Eugene, OR). Generation of Constructs for Expression in Mammalian Cells | Human CEPT1 was amplified by PCR from our original cDNA using the oligonucleotide primers, 5'-GCGGGATCCATGAGTGGGCATCGATCAACA-3' (forward) and 5'-GCGGTCGACTTAATGATTAGAATGAGCTGT-3' (reverse), which contain BamHI and SalI restriction sites, respectively. The PCR product was subcloned into the pCR2.1-Topo vector (Invitrogen, Carlsbad, CA), excised with BamHI and SalI and subcloned into the Escherichia coli expression vector pET23a (Novagen), resulting in the addition of an 11-residue, T7 epitope tag to the N-terminus of the protein. The T7-tagged version of CEPT1 was excised using BglII and SalI and subcloned into the BamHI and SalI sites of the mammalian expression vector pcDNA3 (Invitrogen). GFP was added to the C terminus of T7-tagged CEPT1 by amplification of CEPT1 by PCR using the oligonucleotide primers, 5'-GCAAGATCTATGGCTAGCATGACTGGTGGA-3' (forward) and 5'-GCGCAGAATTCGATGATGATTAGAATGAGC-3' (reverse), which have BglII and EcoRI restriction sites, respectively, built into the primers, and subcloned into the BglII and EcoRI sites of the mammalian expression vector pEGFP-N1 (, Palo Alto, CA). Expression of this construct in mammalian cells results in the production of a CEPT1 protein with an N-terminal T7 epitope tag and a C-terminal green fluorescent protein (GFP). The open reading frame of human CPT1 was amplified by PCR using the oligonucleotide primers 5'-GCCAGATCTATGGCGGCGGCGCCGGGGCC-3' (forward) and 5'-GCCGTCGACTCAATCCATGTTATTCTGATG-3' (reverse), which have BglII and SalI restriction sites, respectively, and subcloning into the TA cloning into the pCR2.1-TOPO (Invitrogen). The BglII/SalI fragment was excised and subcloned into the BamHI and SalI sites of the pET23a vector, resulting in the addition of a T7 epitope tag to the N-terminus of CPT1. The T7-tagged version of CPT1 was digested with BglII and NotI and subcloned into the BamHI and NotI restriction sites of pcDNA3 (Invitrogen). The DNA sequence of all PCR-amplified constructs was confirmed by sequencing both DNA strands. Site-directed mutants were made using the MORPH site-directed mutagenesis kit (5 Prime 3 Prime, Inc., West Chester, PA). Mutagenic oligonucleotides were 5'-phosphorylated and purchased from Life Technologies (Rockville, MD). Mutations were confirmed by manual dideoxy sequencing using the T7sequencing kit (Amersham Pharmacia Biotech, Piscataway, NJ). Cell Culture | CHO-K1 cells were transiently transfected using LIPOFECTAMINE Reagent (Life Technologies) in DMEM containing 5% (vol/vol) fetal calf serum and 33 mug/ml proline and maintained at 37C in an atmosphere of 5% CO2. The calcium chloride method of cell transfection was used for the preparation of stable cell lines in 100-mm dishes. Transfected cells were incubated overnight at 37C in DMEM containing 5% (vol/vol) fetal calf serum and 33 mug/ml proline. Medium was replaced with 8 ml of CHO-K1 media containing 500 mug/ml G418 to start selection of stably transfected clones. Subsequently, the medium was replaced every 2 --3 d with 8 ml of fresh CHO-K1 medium containing 500 mug/ml G418 to select for clones that were stably transfected with the desired plasmid. Single colonies of stably transfected CHO-K1 cells were selected by dilution cloning. Positive colonies were initially identified by Western blotting with the T7-horseradish peroxidase --conjugated antibody and their homogeneity confirmed by immunofluorescence. Immunofluorescence | All cells used for immunofluorescence were grown on glass coverslips in 60-mm dishes at a density of 2 x 105 cells per dish. Cells were fixed in 3% (vol/vol) formaldehyde in 10 mM sodium phosphate (pH 7.4), 225 mM NaCl, 2 mM MgCl2 (PBS-B) for 15 min at room temperature. The coverslips were washed twice at room temperature with PBS-B containing 5 mM ammonium chloride for 5 min each wash, with PBS-B for 5 min, and then permeabilized by adding PBS-B containing 0.05% Triton X-100 and incubating at 40C for 15 min and then room temperature for 15 min. The coverslips were washed twice with PBS-B containing 1% fatty acid --free BSA (PBS/BSA) for 5 min per wash, once for 15 min, and treated with primary antibody in PBS/BSA for 1 h at room temperature. The coverslips were washed twice with PBS/BSA for 5 min each. LcH-lectin (which was conjugated directly to FITC) and secondary antibody was added in PBS/BSA, and cells were incubated for 1 h at room temperature and washed twice with PBS/BSA for 5 min each, and the coverslips were mounted on slides with 2.5% (wt/vol) 1,4-diazadicyclo-[2,2,2]-octane in 50 mM Tris-HCl (pH 9.0) and 90% (vol/vol) glycerol. T7 antibodies were used at 1:1000 dilution, CTalpha antibodies at 1:4000, and calnexin antibodies at 1:200. Secondary antibodies and FITC-coupled LcH-lectin were used at a 1:4000 dilution. Very faint intranuclear staining was observed with the T7 mAb in mock-transfected cells. To stain the mitochondria, cells were grown on glass coverslips as described above. Before fixing, the medium was aspirated from the cells and replaced with fresh CHO-K1 medium containing 200 nM MitoTracker Red CMXRos (Molecular Probes). The cells were incubated with the dye for 45 min at 37C and washed twice with warm PBS. Cells were fixed and mounted as described above. Cells transiently transfected with GFP-fusion and YFP-fusion proteins were grown on glass coverslips, washed three times with 58 mM Na2HPO4, 17 mM NaH2PO4, 68 mM NaCl (PBS-C, pH 7.4), and either viewed directly or fixed by incubating with PBS-C/4% formaldehyde for 30 min at room temperature, washed twice with PBS-C, and mounted as described above. The YFP-Golgi marker is a fusion protein consisting of enhanced yellow fluorescent protein and the sequence encoding the N-terminal 81 amino acids of human beta-1,4-galactosyltransferase . Fluorescence microscopy was performed using a Zeiss axiophot microscope (Thornwood, NY). Green and yellow fluorescent proteins and FITC-coupled antibodies were visualized with Zeiss filter number 10, which excites at 450/490 and emits at 550/565. Texas Red and MitoTracker were visualized with Zeiss filter number 15, which excites at 546/560 and emits at 590. Nuclear Fraction Isolation | NE-PER Nuclear and Cytoplasmic Extraction Reagents Kit (, Rockford, IL) was used to isolate nuclei. We seeded 5 x 105 CHO cells in a 100-mm plate containing DMEM with 5% (vol/vol) fetal calf serum and 33 mug/ml proline and maintained at 37C. The cells were maintained in an atmosphere of 5% CO2 for 4 d. The plates were placed on ice, and the medium was aspirated. The cells were scraped into 1 ml ice-cold PBS and spun at 500 x g at 4C for 3 min. The PBS was aspirated, and the cell pellet was resuspended in 200 mul cytoplasmic extraction reagent I. The sample was vortexed at the highest setting for 15 s to fully resuspend the cell pellet. The sample was incubated on ice for 10 min, and 11 mul cytoplasmic extraction reagent II was added. The sample was vortexed 5 s and incubated on ice for 1 min. The sample was vortexed 5 s on the highest setting and then spun at 16,000 x g and 4C for 5 min. The supernatant was transferred to a new, precooled tube and mixed with an equal volume of SDS-loading buffer. The pellet was resuspended in 100 mul nuclear extraction reagent. The sample was vortexed for 15 s and incubated on ice for 10 min. This was repeated for a total of 40 min. The sample was spun at 16,000 x g and 4C for 10 min. The supernatant was transferred to a new, precooled tube and then mixed with an equal volume of SDS-loading buffer. Samples were stored at -20C. A 20-mul sample of nuclear and extranuclear extracts was loaded and proteins separated by SDS-PAGE and identified by Western blot. Western Blot Analysis | HJ091 Saccharomyces cerevisiae cells (ahis3-Delta1 leu2-3 leu2-112 ura3-52 trp1 --289 cpt1::LEU2 ept1-) were grown to midlog phase in synthetic dextrose medium containing the appropriate nutrients to ensure plasmid maintenance , and microsomal membranes were prepared as described . Protein extracts were incubated with SDS-PAGE sample buffer at 37C for 30 min, separated on a 12% SDS-polyacrylamide gel, and transferred to polyvinylidene difluoride membranes. Blots were probed with a T7 epitope tag-specific mAb coupled directly to horseradish peroxidase (1:5000, Novagen) for subsequent detection using the ECL system (Amersham Pharmacia Biotech). Enzyme Assays | To determine in vitro choline- and ethanolaminephosphotransferase enzyme activities of wild-type and site-directed CEPT1 mutants the microsomal membranes were isolated from HJ091 yeast (cpt1::LEU2 ept1-) grown to midlog phase in synthetic dextrose media containing the appropriate nutrients to ensure plasmid maintenance of wild-type or mutant versions of CEPT1 from the constitutive glyceraldehyde-3-phosphate dehydrogenase promoter in the expression vector p416GPD . A mixed micelle assay was used as previously described . All assays were performed at least three times in duplicate and the indicated values represent their mean. SEs were <15% of the mean for each experiment. Cholinephosphotransferase and ethanolaminephosphotransferase activities in CHO-K1 cell lines and CHO-K1 cells transiently transfected with CEPT1-GFP were determined after placing the cells on ice and washing them twice with ice-cold PBS. The cells were scraped into a microfuge and centrifuged at 15,000 x g for 5 min at 4C. The cell pellet was resuspended in 0.5 ml of 10 mM HEPES-HCl (pH 7.4), 50 mM KCl, 1 mM EDTA, and Complete protease inhibitor cocktail (Roche Molecular Biochemicals, Indianapolis, IN)) and passed through a 23-gauge needle 20 times to lyse the cells. The mixture was centrifuged at 15,000 x g for 30 s at 4C, and the supernatant was centrifuged at 450,000 x g for 15 min at 4C to pellet cellular membranes. Membranes were resuspended in 10 mM HEPES-HCl (pH 7.4), 50 mM KCl, 1 mM EDTA, and Complete protease inhibitor cocktail (Roche Molecular Biochemicals) by using a Teflon pestle and stored at -70C. The mixed micelle assay for cholinephosphotransferase activity was used as previously described . Metabolic Labeling | S. cerevisiae HJ091 cells (cpt1::LEU2 ept1-) transformed with either wild-type or mutant versions of CEPT1 in the constitutive expression vector p416GPD were grown to midlog phase in synthetic dextrose media containing the appropriate nutrients to ensure plasmid maintenance . [14C]Choline (10 muM, 1 x 105 dpm/nmol) or [14C]ethanolamine (6.7 muM, 2.2 x 105 dpm/nmol) was added to the cultures for 1 h at 30C. Cells were then concentrated by centrifugation and incorporation of radiolabel into lipids was performed as described . All labeling experiments were performed at least three times in duplicate and the values indicated represent their mean. SEs were <15% of the mean for each experiment. Protein and Lipid Determination | Protein was determined by the method of using bovine serum albumin as standard. Phospholipid phosphorus was determined by the method of . RESULTS : Sites of PtdCho and PtdEtn Synthesis | The CEPT1 open reading frame was fused with that of the GFP, and this construct was transiently transfected into Chinese hamster ovary (CHO-K1) cells. The CEPT1-GFP chimeric protein was functional as cells transiently transfected with CEPT1-GFP at 20% efficiency had a fourfold increase in cholinephosphotransferase activity. Indeed, transfection efficiency of CEPT1-GFP essentially mirrored the increase in cholinephosphotransferase activity, implying that all of the CEPT1-GFP fusion protein was enzymatically active (our unpublished results). As an initial step in determining the intracellular site of CEPT1 we treated the cells with brefeldin A, which essentially collapses the Golgi apparatus into the endoplasmic reticulum, resulting in a dramatic relocalization of most Golgi resident proteins. Brefeldin A treatment did not result in the relocalization of CEPT1, whereas the Golgi marker was effectively relocalized (Figure A). Similar results were observed whether live cells were viewed or if cells were fixed before microscopy (our unpublished results). Figure 2 | Effect of brefeldin A on CEPT1 and CPT1 subcellular localization. Effect of brefeldin A on CEPT1 and CPT1 subcellular localization. (A) CHO-K1 cells transiently transfected with T7-CEPT1-GFP or the first 81 amino of the Golgi resident beta-1,4-galactosyltransferase fused to yellow fluorescent protein were treated with 2 mug/ml brefeldin A for 30 min. Live and fixed cells resulted in identical images. (B) CHO-K1 cells stably expressing T7-CPT1 were treated with brefeldin A and T7 monoclonal antibodies followed by Texas Red --conjugated secondary antibodies were used to determine the location of CPT1. The location of the Golgi was determined by the addition of FITC-coupled L. culinaris lectin. To further define the intracellular site of CEPT1 and CPT1, stable CHO-K1 cell lines expressing T7-epitope --tagged CEPT1 and CPT1 were constructed. These same T7-tagged CEPT1 and CPT1 proteins expressed in yeast devoid of their endogenous cholinephosphotransferase and ethanolaminephosphotransferase activities were previously used to characterize the substrate specificity of the CEPT1 and CPT1 enzymes in vitro by enzyme assay and in vivo through metabolic labeling experiments . In addition, increased expression of T7-tagged CEPT1 was demonstrated to prevent farnesol induced inhibition of cholinephosphotransferase activity in CHO-K1 cells, once again demonstrating that the tagged versions of these enzymes are active and can reconstitute the CDP-alcohol pathways in a variety of eukaryotic cell types, indicating they are properly localized . We isolated five CHO-K1 cell lines expressing T7-tagged CEPT1 or CPT1 that possessed barely perceptible increases in CEPT1 and CPT1 enzyme activity compared with those with up to fivefold increased expression of CPT1 or CEPT1. We found that the level of expression of CPT1 and CEPT1 did not alter their intracellular locations (our unpublished results), and representative images of cell lines with an estimated twofold increase in CEPT1 and CPT1 expression are presented (Figures and ). Figure 3 | Intracellular localization of CPT1. Intracellular localization of CPT1. The intracellular location of CPT1 was determined in CHO-K1 cells stably expressing T7-CPT1 using T7 monoclonal primary antibodies. The location of the endoplasmic reticulum was assessed using primary antibodies to calnexin. Golgi location was determined by the addition of FITC-coupled L. culinaris lectin. MitoTracker dye was used to determine the location of the mitochondria. Secondary antibodies were coupled to either FITC or Texas Red. The yellow color in merged images indicates overlap between CPT1 and the organelle marker. Brefeldin A treatment of CHO cells expressing T7-CPT1 resulted in a redistribution of CPT1 from a large punctate region to a more diffuse region (Figure B). This implies that CPT1 synthesizes PtdCho in the Golgi apparatus. Coimmunofluorescence of CPT1 and CEPT1 with organelle-specific markers was performed. CPT1 colocalized with the Golgi-specific L. culinaris lectin and was independent of the endoplasmic reticulum and mitochondrial markers . This was consistent with the redistribution of CPT1 in response to brefeldin A treatment. The CEPT1 protein colocalized with the endoplasmic reticulum marker calnexin and was not found colocalized with the Golgi or mitochondrial markers . The CEPT1-GFP chimera was also found to colocalize with the endoplasmic reticulum resident protein disulfide isomerase (our unpublished results). Figure 4 | Intracellular localization of CEPT1. Intracellular localization of CEPT1. The intracellular location of CEPT1 was determined in CHO-K1 cells stably expressing T7-CEPT1 using T7 monoclonal primary antibodies. The location of the endoplasmic reticulum was assessed using primary antibodies to calnexin. Golgi location was determined by the addition of FITC coupled L. culinaris lectin. MitoTracker dye was used to determine the location of the mitochondria. Secondary antibodies were coupled to either FITC or Texas Red. The yellow color in merged images indicates overlap between CEPT1 and the organelle marker. Reconstitution of the PtdCho Biosynthetic Pathway | Closer observation of the merged CEPT1/calnexin image resulted in our observance of a small region around the nuclear membrane that did not appear to completely merge with the endoplasmic reticulum marker. The synthesis of CDP-choline for consumption by cholinephosphotransferase is the rate-limiting step in the synthesis of PtdCho and is catalyzed by a pair of CTP:phosphocholine cytidylyltransferase (CT) enzymes. CTalpha is the main isoform present in most cell types including CHO cells , and upon activation of PtdCho synthesis (e.g., through fatty acid supplementation) CTalpha translocates to the nuclear membrane . In our hands, activation of CTalpha also resulted in its colocalization to the nuclear membrane, and this overlapped with the portion of CEPT1 that did not colocalize with the endoplasmic reticulum (Figure A). To ensure that a portion of CEPT1 was indeed associated with the nuclear membrane, we separated CHO-K1 cellular nuclei from endoplasmic reticulum by subcellular fractionation. The nuclei are clearly separated from the endoplasmic reticulum because the endoplasmic reticulum marker calnexin is found exclusively in the extranuclear fraction, whereas the nuclear membrane marker Lap-2 is found exclusively in the nuclear fraction. Similar to our coimmunofluorescence experiments, the bulk of CEPT1 was associated with the extranuclear fraction, whereas a smaller proportion was found in the nucleus. These experiments demonstrate that CEPT1 resides in both the endoplasmic reticulum and nuclear membranes, and increasing PtdCho synthesis via CTalpha translocation is by redistribution of this rate-limiting enzyme to nuclear located CEPT1. However, whether CEPT1 and CTalpha reside within the same nuclear membrane and/or directly interact will require further experimentation. Figure 5 | PtdCho synthesis pathway reconstitution at the nuclear membrane. PtdCho synthesis pathway reconstitution at the nuclear membrane. (A) CHO-K1 stably expressing T7-CEPT1 were treated with oleic acid (500 muM in 0.5% bovine serum albumin) for 24 h to translocate CTalpha from its inactive soluble intranuclear location to its active nuclear membrane location before immunofluorescence. (B) Nuclei (N) were separated from extranuclear structures including the endoplasmic reticulum (E) by subcellular fractionation as described in MATERIALS AND METHODS, and the distribution of nuclear and endoplasmic reticulum markers are compared with CEPT1. Structure/Function Analysis of CEPT1 | The first genes identified to code for cholinephosphotransferase and ethanolaminephosphotransferase activities were isolated from the yeast S. cerevisiae . The yeast CPT1 gene product utilizes CDP-choline in vitro and in vivo, whereas the EPT1 gene product can use both CDP-choline and CDP-ethanolamine . Genetic inactivation of the CPT1 and EPT1 genes resulted in the complete loss of measurable cholinephosphotransferase and ethanolaminephosphotransferase in vitro enzyme activity and an inability to incorporate specific radiolabeled precursors for each pathway into either PtdCho or PtdEtn . Inactivation of the both CDP-alcohol pathways in most cell types would be lethal; however, yeast cells are still viable because they can de novo synthesize phosphatidylserine, which can be decarboxylated to PtdEtn and methylated to PtdCho . The ability of yeast to survive in the absence of functional CDP-choline and CDP-ethanolamine pathways for PtdCho and PtdEtn synthesis allowed for their use as a null expression system for structure/function analysis of human CEPT1. Extrapolation of previous chimeric enzyme analysis of the yeast Cpt1p and Ept1p enzymes to human CEPT1 positions the outside limit of the CDP-alcohol --binding region to amino acid residues 88 --213, whereas the diacylglycerol binding site is located to the transmembrane helices spanning residues 88 --258 (Figure ; ; ). Within this region is a CDP-alcohol phosphotransferase motif, DG(x)2AR(x)8G(x)3D(x)3D, that runs from residues 136 --158 in human CEPT1. The final two aspartates within this motif are required for catalysis with the remainder of the conserved residues responsible for substrate affinity or steric stability . Secondary structure predictions and hydropathy plots weave either three of four membrane-spanning domains through the substrate binding region of the yeast and human cholinephosphotransferases and choline/ethanolaminephosphotransferases . Amino acids residues within the CDP-alcohol phosphotransferase motif were changed from those found in the dual specificity human CEPT1 and yeast Ept1p to those found in CDP-choline --restricted human CPT1 and yeast Cpt1p (Figure A) to determine if the divergent residues within this region impart CDP-alcohol specificity. Figure 6 | Predicted secondary structures for human CEPT1. Predicted secondary structures for human CEPT1. Seven or eight membrane spans are strongly predicted with helix 181 --199 positioned either in (A) or outside (B) of the membrane using the SMART or TmPred algorithms. The numbers denote amino acid residues with the black filled circles representing those mutated in the current study. Based on comparisons to studies on the yeast Cpt1p and Ept1p enzymes, the dark gray circles represent the maximum region required for diacylglycerol binding, and the light gray the maximum region required for CDP-alcohol binding. The conserved residues within the CDP-alcohol phosphotransferase motif, DG(x)2AR(x)8G(x)3D(x)3D, are also indicated. Figure 7 | CEPT1 CDP-alcohol phosphotransferase motif mutants. CEPT1 CDP-alcohol phosphotransferase motif mutants. (A) The site-directed mutations made in CEPT1 are indicated. The numbers denote CEPT1 amino acid residues. (B) Western blot of wild-type and mutant versions of CEPT1. CEPT1 CDP-Alcohol Site-directed Mutants | Plasmids carrying wild-type and site-directed mutants of CEPT1 were transformed into S. cerevisiae HJ091, a yeast strain that is devoid of cholinephosphotransferase and ethanolaminephosphotransferase activities due to inactivating mutations within its endogenous CPT1 and EPT1 genes (cpt1::LEU2 ept1-; ; ; ). This expression system ensures that any cholinephosphotransferase or ethanolaminephosphotransferase activity detected would be plasmid encoded. Western blot analysis of HJ091 membrane fractions expressing the wild-type and mutant versions of CEPT1 established that similar amounts of full-length enzyme were being expressed (Figure B). The CDP-alcohol specificities of CEPT1 and the CDP-alcohol phosphotransferase motif mutants were determined by in vitro enzyme assay. There were very small differences in enzyme activity or CDP-alcohol specificity with the CEPT1 N144G, S146Q, or S146C mutations compared with the wild-type enzyme . A decrease in both cholinephosphotransferase and ethanolaminephosphotransferase activity to similar50% wild-type activity was seen in the K138 M mutant; however, this demonstrates an overall effect on activity but not substrate specificity because use of both CDP-choline and CDP-ethanolamine as substrates was equally affected. Mutation of glycine 156 to either alanine, serine, or cysteine also decreased cholinephosphotransferase activity to 50% wild-type, but more importantly abolished the ability of CEPT1 to utilize CDP-ethanolamine as a substrate . Increasing the specific radioactivity of the CDP-ethanolamine substrate to values 10-fold normally used still did not allow for detectable enzyme activity. This indicates a specific role for glycine 156 in CDP-ethanolamine binding by CEPT1. Figure 8 | Enzyme activity of CEPT1 CDP-alcohol phosphotransferase motif mutants. Enzyme activity of CEPT1 CDP-alcohol phosphotransferase motif mutants. Enzyme activities were determined from microsomal membrane preparations of S. cerevisiae cells (cpt1:: LEU2 ept1-) constitutively expressing CEPT1 or the indicated site-directed mutants. The ability to use either CDP-choline (black) or CDP-ethanolamine (white) as a substrate is indicated. To confirm the in vitro CDP-alcohol substrate specificity results, the ability of wild-type and mutant versions of CEPT1 to reconstitute the synthesis of PtdCho or PtdEtn from radiolabeled choline or ethanolamine, respectively, was tested in S. cerevisiae HJ091 cells (cpt1::LEU2 ept1-). All of the yeast expressing mutant CEPT1 proteins were able to take up choline and synthesize PtdCho at levels similar to yeast expressing parental CEPT1 (Figure A). All of the yeast expressing mutant CEPT1 proteins were also capable of taking up ethanolamine at levels near to those expressing parental CEPT1; however, none of the CEPT1 proteins containing a substitution for glycine 156 were able to reconstitute de novo PtdEtn synthesis (Figure B). Hence, the in vivo metabolic pathway reconstitution results for the CEPT1 mutants were consistent with the in vitro assessment of their CDP-alcohol specificities and further support a role for glycine 156 in CEPT1 CDP-alcohol specificity for human CEPT1. Figure 9 | Metabolic reconstitution of PtdCho and PtdEtn synthesis by CEPT1 CDP-alcohol phosphotransferase mutants. Metabolic reconstitution of PtdCho and PtdEtn synthesis by CEPT1 CDP-alcohol phosphotransferase mutants. Exponentially growing S. cerevisiae cells (cpt1:: LEU2 ept1-) constitutively expressing CEPT1 or the indicated site-directed mutants were radiolabeled with (A) [14C]choline to label PtdCho or (B) [14C]ethanolamine to label PtdEtn, for 1 h. Radiolabel incorporated into each phospholipid was determined by scintillation counting. Identification of Diacylglycerol Fatty Acyl Specificity Residues | The 214 --234 membrane spanning helix of CEPT1 lies within the predicted diacylglycerol binding region of CEPT1 . N-terminal to residues 214 --234 is a helix that spans residues 181 --199 that is weakly predictive to span the membrane. If helix 181 --199 does indeed span the membrane, then the orientation of helix 214 --234 within the membrane would be flipped . Because the incoming phosphobase is transferred directly from the CDP-alcohol onto diacylglycerol without going through a membrane-bound intermediate , then the relevant amino acids within helix 214 --234 need to be in close proximity to the soluble CDP-alcohol phosphotransferase motif. Because diacylglycerol only spans half of the membrane bilayer, amino acid residues that encompass the half of the membrane that interact with diacylglycerol would have an effect on enzymatic activity and/or diacylglycerol specificity, whereas residues mutated on the opposite side of the helix would have little effect. A number of residues spanning the entire membrane spanning 214 --234 helix were targeted for site-directed mutagenesis (Figure A) to determine if residues within this helix interact with diacylglycerol and to allow positioning of this helix in the active site of CEPT1 relative to the CDP-alcohol phosphotransferase motif. Figure 10 | CEPT1 diacylglycerol specificity mutants. CEPT1 diacylglycerol specificity mutants. (A) The site-directed mutations made in CEPT1 are indicated. The same region in human CPT1 is provided for comparison. The numbers denote CEPT1 amino acid residues. (B) Western blot of wild-type and mutant versions of CEPT1. Plasmids carrying wild-type CEPT1 and various helix 214 --234 mutations were transformed into S. cerevisiae strain HJ091 (cpt1::LEU2 ept1-). Western blot analysis of HJ091 membrane fractions established that similar levels of full-length CEPT1 protein were expressed (Figure B). The cholinephosphotransferase activities of CEPT1 mutants were determined by in vitro enzyme assay. The wild-type CEPT1 enzyme had a broad diacylglycerol substrate specificity with di16:1 >= 16:0/22:6 >= 16:0/18:1 >= di18:1 > di10:0 >= 16:0(O):20:4 (platelet activating factor [PAF] precursor). There was very little detectable activity toward di16:0, di14:0, or di12:0 diacylglycerols , and no activity was detected using 16:0(O):2:0 as substrate for the direct synthesis of PAF (our unpublished results). Each of the CEPT1 mutants was also analyzed for cholinephosphotransferase activity using a wide variety of defined diacylglycerols . A number of the mutations, namely T214A, V216A, and I221A, altered the profile of diacylglycerol utilization when compared with that of wild-type CEPT1 and also resulted in modest reductions in enzyme activity. The E215A, E215D, and E215Q mutations resulted in a much more dramatic reduction in CEPT1 enzyme activity. E215A and E215D did not alter diacylglycerol specificity, whereas the E215Q demonstrated altered diacylglycerol specificity. In contrast, the mutations L226A and V228A did not alter cholinephosphotransferase activity or diacylglycerol specificity of CEPT1. These results indicate that residues 226 and 228 are located on the opposite side of the helix from the catalytic site, whereas residues 214, 215, 216, and 221 are in close proximity to the catalytic site. Based on these results, it is predicted that the model proposed in Figure B is correct with regards to the orientation of membrane spanning helix 214 --234 to the CDP-alcohol phosphotransferase motif. Figure 11 | Enzyme activity of CEPT1 diacylglycerol specificity motif mutants. Enzyme activity of CEPT1 diacylglycerol specificity motif mutants. Enzyme activities were determined from microsomal membrane preparations of S. cerevisiae cells (cpt1:: LEU2 ept1-) constitutively expressing CEPT1 or the indicated site-directed mutants. DISCUSSION : PtdCho and PtdEtn comprise 50 --75% of cellular phospholipid mass. In this study we have identified the sites at which CEPT1 and CPT1, the enzymes that directly synthesize these two lipids, reside and provided molecular insights into specific amino acid determinants for both the fatty acid and lipid head group specificity of CEPT1. CEPT1 is a dual specificity enzyme that can synthesizes both PtdCho and PtdEtn, and glycine 156 within the catalytic CDP-alcohol phosphotransferase motif of CEPT1 was required for dual CDP-alcohol specificity. Any other amino acid at this position resulted in a loss of ethanolaminephosphotransferase activity by CEPT1. Mutagenesis of amino acid residues within the predicted transmembrane helix 214 --234 of CEPT1 altered diacylglycerol substrate specificity, and enzyme activity confirming this helix spans the membrane. Only mutations in residues 214 --221 of this helix affected diacylglycerol substrate specificity and enzyme activity, and because the phosphobase is transferred directly from the CDP-alcohol onto diacylglycerol without passing through an enzyme-bound intermediate , the first half of helix 214 --234 must juxtapose the CDP-alcohol --binding site. This enabled the orientation of this diacylglycerol binding helix within the membrane span to be accurately positioned. Our study also provides formal proof that PtdCho and PtdEtn are de novo synthesized at specific sites within the cell. CPT1 synthesizes PtdCho exclusively and was found in the Golgi, whereas the dual specificity CEPT1, which synthesizes both PtdCho and PtdEtn, was found in both endoplasmic reticulum and nuclear membranes. The rate-limiting step in the synthesis of PtdCho is catalyzed by CTalpha in most cell types and produces the CDP-choline that is used by cholinephosphotransferase activities to produce PtdCho. In most cell types, CTalpha is found in the nucleus as an amphitropic protein that is stored as an inactive soluble protein that translocates to the nuclear membrane to upregulate PtdCho synthesis . We demonstrated that activation of CTalpha through translocation to the nuclear membrane brings it into proximity with the nuclear membrane portion of CEPT1. This demonstrates that the activation of the CDP-choline pathway is likely through redistribution of the rate-limiting penultimate enzyme to the site of the ultimate step within the pathway. Whether CTalpha and CEPT1 reside in the same nuclear bilayer and/or physically interact remains to be determined, as does their colocalization with other upstream enzymes including CTbeta and CTP:phosphoethanolamine cytidylyltransferase. The role of diacylglycerol in the regulation of vesicle transport from the Golgi is well established in yeast and mammalian cells. In mammalian cells diacylglycerol production is required for recruitment of the Golgi vesicle biogenesis factor protein kinase D to the trans-Golgi network . The regulation of Golgi diacylglycerol levels thus appears to be an important regulatory event in Golgi derived vesicle transport in mammalian cells. Naturally, the production of diacylglycerol in the Golgi for Golgi-derived vesicle transport must be balanced by diacylglycerol clearance. Our identification of a Golgi specific localization for mammalian CPT1 implies its consumption of diacylglycerol could play a role in the regulation of Golgi derived vesicle transport. A role for diacylglycerol consumption by the CDP-choline pathway during de novo PtdCho biosynthesis is supported by data obtained in yeast. The yeast PtdCho/phosphatidylinositol transfer protein Sec14p is a soluble protein that translocates to Golgi membranes, and ablation of Sec14p function results in decreased Golgi-derived vesicle transport that eventually results in cell death . Genetic inactivation of the CPT1-catalyzed step in the synthesis of PtdCho results in bypass of the essential function of Sec14p and allows cells to now live because of their ability to regain Golgi-derived vesicle transport. Decreasing PtdCho synthesis and thus increasing diacylglycerol levels by preventing its consumption through the CDP-choline pathway suppresses the need for Sec14p in Golgi transport , implying increased Golgi diacylglycerol promotes Golgi-derived vesicle transport . However, inactivation of the EPT1 pathway for PtdCho synthesis does not bypass the ability of cells to live in the absence of Sec14p, even if Ept1p is contributing to 50% of net CDP-choline --derived PtdCho synthesis . Our determination that human CEPT1 resides in the endoplasmic reticulum and nuclear membrane, whereas human CPT1 is in the Golgi, may provide an explanation for the ability of loss of function of yeast CPT1, but not EPT1, to bypass the essential function of Sec14p. Although speculative, if yeast Cpt1p and Ept1p localize to similar intracellular locations as their mammalian counterparts, then inactivation of yeast CPT1 would increase diacylglycerol and decrease PtdCho levels in the Golgi and thus provide a favorable shift in Golgi lipid levels for bypass of Sec14p function. Inactivation of EPT1 would primarily alter endoplasmic reticulum and nuclear lipid levels and would not have dramatic affects on Golgi diacylglycerol or PtdCho levels. Increased expression of EPT1 (and human CEPT1) in yeast was able to restore PtdCho synthesis to cells lacking their CPT1 gene, and this also prevented bypass of loss of Sec14p function in yeast carrying an inactivated CPT1 gene. We predict that increased expression of Ept1p and human CEPT1 results in either the mislocalization of a portion of these enzymes to the Golgi or drives increased synthesis of endoplasmic reticulum PtdCho, which is then transported to the Golgi. Our attempts to express human CPT1 in yeast were at levels too low to allow for complete restoration of PtdCho synthesis , and this is likely why human CPT1 was unable to bypass of the essential function of Sec14 due to inactivation of yeast CPT1. The precise site of yeast Cpt1p, Ept1p, and their human counterparts expressed in yeast is currently under investigation. In this study we have addressed a fundamental issue of cell biology, where are phospholipids made? We have also demarcated molecular determinants that define product formation. The sites of synthesis of PtdCho and PtdEtn is requisite knowledge for any model attempting to describe how these lipids are transported to other organelles for restoration of PtdCho and PtdEtn levels subsequent to their signal transduction --mediated catabolism or for providing new membrane for cell growth. The data have also added credence to the model that interprets how Golgi-derived vesicle transport may be regulated by the consumption of diacylglycerol during the formation of PtdCho. Backmatter: Abbreviations used: : PtdCho = phosphatidylcholine PtdEtn = phosphatidylethanolamine CPT1 = cholinephosphotransferase of S. cerevisiae EPT1 = choline/ethanolaminephosphotransferase of S. cerevisiae CEPT1 = human choline/ethanolaminephosphotransferase CPT1 = human cholinephosphotransferase CT = CTP:phosphocholine cytidylyltransferase PMID- 12221123 TI - An Essential Subfamily of Drs2p-related P-Type ATPases Is Required for Protein Trafficking between Golgi Complex and Endosomal/Vacuolar System AB - The Saccharomyces cerevisiae genome contains five genes encoding P-type ATPases that are potential aminophospholipid translocases (APTs): DRS2, NEO1, and three uncharacterized open reading frames that we have named DNF1, DNF2, and DNF3 for DRS2/NEO1 family. NEO1 is the only essential gene in APT family and seems to be functionally distinct from the DRS2/DNF genes. The drs2Delta dnf1Delta dnf2Delta dnf3Delta quadruple mutant is inviable, although any one member of this group can maintain viability, indicating that there is a substantial functional overlap between the encoded proteins. We have previously implicated Drs2p in clathrin function at the trans-Golgi network. In this study, we constructed strains carrying all possible viable combinations of null alleles from this group and analyzed them for defects in protein transport. The drs2Delta dnf1Delta mutant grows slowly, massively accumulates intracellular membranes, and exhibits a substantial defect in the transport of alkaline phosphatase to the vacuole. Transport of carboxypeptidase Y to the vacuole is also perturbed, but to a lesser extent. In addition, the dnf1Delta dnf2Delta dnf3Delta mutant exhibits a defect in recycling of GFP-Snc1p in the early endocytic-late secretory pathways. Drs2p and Dnf3p colocalize with the trans-Golgi network marker Kex2p, whereas Dnf1p and Dnf2p seem to localize to the plasma membrane and late exocytic or early endocytic membranes. We propose that eukaryotes express multiple APT subfamily members to facilitate protein transport in multiple pathways. Keywords: INTRODUCTION : P-type ATPases are a large family of multitransmembrane domain, ATP-dependent transporters. The P-type designation derives from a covalent aspartyl-phosphate catalytic intermediate formed during ATP hydrolysis, which is coupled to conformational changes that can drive ion transport across membranes against an electrochemical gradient. Based on the substrate transported, P-type ATPases are classified as heavy metal ion ATPases, such as the copper transporters that are mutated in patients with Menkes or Wilson disease; and nonheavy metal ion ATPases, such as Ca2+-, H+-, and Na+,K+-ATPases . From sequence analyses, members of a third subfamily of P-type ATPases have been distinguished from the above-mentioned two types and are proposed to be aminophospholipid translocases (APTs) or flippases . APTs translocate phosphatidylserine (PS) and/or phosphatidylethanolamine (PE) from one leaflet of the membrane bilayer to the other, and seem to be responsible for concentrating PS and PE on the cytosolic leaflet of biological membranes. Genome sequencing projects have identified a large number of potential APT subfamily genes in eukaryotic organisms, although no prokaryotic member has been identified yet. There are five members of this subfamily in both Saccharomyces cerevisiae and Schizosaccharomyces pombe . The Caenorhabditis elegans genome contains six members, Arabidopsis has 11 , and there seem to be at least 21 members in humans. Little is known about the cellular function of the APT subfamily of proteins or why eukaryotes maintain so many of these genes. However, a few biological disorders have been linked or attributed to genes from this subfamily. FIC1 mutations cause familial intrahepatic cholestasis in humans, which is a defect in bile secretion . The ATP10C gene has been linked to Angelman syndrome and autism in some patients , and the mouse ATP10C homolog pfatp is a candidate gene for increased body fat . In addition, the PDE1 gene is a pathogenic factor for the rice blast fungus, Magnaporthe grisea , and the ALA1 gene in Arabidopsis is required for growth at low temperature . The first member of the APT subfamily to be identified genetically and cloned was DRS2 from S. cerevisiae (yeast) . Drs2p is 47% identical in amino acid sequence to the mammalian chromaffin granule ATPase II , which is proposed to be the PS translocase characterized with intact chromaffin granules . DRS2 was initially identified in a genetic screen for mutants that exhibit a cold-sensitive defect in ribosome synthesis and was shown to be a nonessential gene, although disruption of DRS2 causes a strong cold-sensitive growth defect . Subsequently, we identified DRS2 in another genetic screen for mutations that are synthetically lethal with arf1 . ADP-ribosylation factor (ARF) is a small GTP-binding protein whose functions in initiating COPI and clathrin-coated vesicle (CCV) budding have been well studied . The drs2Delta null allele is also synthetically lethal with clathrin heavy chain temperature-sensitive (ts) alleles but not with ts alleles of COPI subunits . These genetic interactions between DRS2, ARF, and clathrin suggested a protein transport function for Drs2p at the trans-Golgi network (TGN). In fact, Drs2p localizes to the TGN and the drs2Delta mutant exhibits TGN defects comparable with those exhibited by clathrin mutants. These include 1) the mislocalization of TGN proteins, resulting in a defect in the proteolytic processing of pro-alpha-factor; 2) the accumulation of aberrant Golgi cisternae; and 3) a deficiency of CCVs in subcellular fractions . These phenotypes are much more severe at the nonpermissive growth temperature. Recently, however, we have found that the drs2Delta mutant exhibits a defect at permissive growth temperatures in generating a specific class of CCVs carrying invertase and acid phosphatase to the plasma membrane . The ribosome synthesis defect observed originally in drs2Delta is likely a secondary effect of its protein transport defect. Many secretory mutants globally down-regulate ribosomal protein synthesis through a protein kinase C signal transduction pathway . In addition, the severe protein transport defects are observed immediately after shifting cells to the nonpermissive temperature, whereas several hours are required before a protein synthesis defect is observed . The drs2Delta mutant has also been reported to exhibit a defect in APT activity at the plasma membrane , although others report no defect in this activity . Moreover, the localization of Drs2p suggests a primary function for this protein in the TGN rather than the plasma membrane . Although PS and/or PE translocation is the most likely biochemical activity for Drs2p, direct evidence for this remains elusive. In fact, only an erythrocyte plasma membrane APT activity has been reconstituted in vitro with purified enzyme . The drs2Delta strain grows as well as wild type at 24C or above, but fails to grow at 23C or below. One possible explanation for the cold sensitivity of drs2Delta null strain is that other proteins could replace the essential function of Drs2p at higher temperatures, with the best candidates being the Drs2p homologs encoded within the yeast genome. A phylogenetic sequence comparison of the 16 P-type ATPases encoded in the yeast genome suggested that five of these proteins belong to the APT subfamily . Two of these genes (DRS2 and NEO1) have been identified genetically, whereas the remaining three genes (YER166W, YDR093W, and YMR162C) have not been characterized. We have designated the three uncharacterized genes DNF1 (YER166W), DNF2 (YDR093W), and DNF3 (YMR162C) for DRS2/NEO1 family. The similarity of these proteins range from 39% for DNF3 and NEO1 to 83% for DNF1 and DNF2 . We have constructed strains carrying all possible combinations of drs2Delta and dnfDelta null alleles and have found that DRS2 and the DNF genes constitute an essential gene family that exert overlapping functions in protein transport. Drs2p and Dnf1p have redundant functions in protein transport from the TGN to the vacuole, and Dnf1p, Dnf2p, and Dnf3p have redundant functions in recycling the v-SNARE Snc1p from endosomes to the TGN. The potential role of these proteins in generating membrane asymmetry will be the focus of another report. Table 1 | Similarity and identity of DRS2/NEO1 genes MATERIALS AND METHODS : Media, Strains, and Plasmids | Yeast were grown in standard rich medium (YPD) or SD minimal media containing required supplements . The calcofluor white (CW) sensitivity test was done on YPD 2% agar plates containing 10, 30, or 50 mug/ml calcofluor white (F6259; Sigma-Aldrich, St. Louis, MO). Yeast strains used in this study are listed in Table . The original DRS2/NEO1 family gene deletion strains were made by Saccharomyces genome deletion project and were purchased from American Type Culture Collection (Manassas, VA). The drs2/dnf combination deletion strains were generated by standard genetic crosses or gene disruptions. The genotype of each spore was determined by a polymerase chain reaction (PCR) method as described by the Saccharomyces genome deletion project . ZHY3271G was generated by mating two dnf1,2,3Delta strains generated from tetrad dissection. The other dnf deletion diploid strains were generated by transforming cells with pHO , encoding the HO endonuclease. The strains that harbor Myc-tagged proteins were generated by PCR-based gene targeting into the BY4742 strain and selected on G418 plates. The pFA6a-13Myc-kanMX6 was used as PCR template , and the Myc integrated strains were confirmed by PCR. The epitope-tagged Drs2 and Neo1 proteins are functional because they support cell growth at all temperatures tested. We assume that the epitope-tagged Dnf proteins are functional although this has not been tested directly. Table 2 | Yeast strains used in this study pRS416-DRS2 was produced by subcloning of a SpeI-SalI fragment from pRS315-DRS2 into SpeI-SalI digested pRS416. pZH523 (DRS2 disruption plasmid) was constructed by replacing a BamHI-SnaBI fragment from pPR10 with a similar2.1-kb BamHI-PvuII fragment containing LEU2 from pJJ283 . The plasmid pZH523 was linearized with SacI and HpaI and transformed into different strains to delete DRS2. The full-length NEO1 gene was PCR amplified using primers (forward 5'-AAGGATATGTCTTTTCGTGTACGTG and reverse 5'-GCCGCTTGATATGTAAGTTCGTCA) and cloned into pYGW1 by using the GeneWeaver I kit from Research Genetics (Huntsville, AL). The EcoRI-SalI fragment of the NEO1 gene from pYGW1-NEO1 was subcloned into EcoRI-SalI site of pBM743 so that NEO1 is under the control of GAL1 promoter and can be suppressed in glucose media . GFP-ALP , pRS416 SNC1-GFP , pRS426 STE2-GFP , and p426 KH (2 mu Kex2-HA in pRS426) were transformed into strains as indicated. Sporulation Test | To test for sporulation efficiency, cells were grown overnight in YPD medium, diluted into PSP2 (1% KOAc, 0.1% yeast extract, and 0.67% yeast nitrogen base) presporulation medium, and cultured for 18 --24 h to late log phase. Cells were then washed once in distilled H2O and resuspended in 1% KOAc sporulation medium at 1 --2 x 107 cells/ml. After shaking for 3 --7 d at 30C, the sporulating cells were examined under the microscope for visible tetrads. In addition, cells that were in sporulation medium for 5 --7 d were treated with glusulase and ether to kill the nonsporulated cells and spread on YPD plates to germinate. Strains scored as "+" in Table showed no visible or viable spores by these tests. Immunological Methods | Cell labeling, immunoprecipitation , and immunoblotting were performed as described previously. Anti-alpha-factor , anti-carboxypeptidase Y (CPY) , and anti-ALP serum were used (1 mul/OD of cells) for immunoprecipitations. The 9E10 mouse monoclonal c-Myc antibody (Oncogene Research Products, Darmstadt, Germany) was used at 1:2000 for Western blot and 1:100 for immunofluorescence. Polyclonal affinity-purified rabbit anti-hemagglutinin (HA) antibody (Zymed Laboratories, South San Francisco, CA) was used at 1:100 dilution to detect Kex2-HA by immunofluorescence. Alexa-488 goat anti-rabbit IgG and Alexa-594 goat anti-mouse IgG (Molecular Probes, Eugene, OR) were used at 1:200 as secondary antibodies for immunofluorescence. Monoclonal CPY and ALP antibody (Molecular Probes) were used at 1:2000 and 1:125 in Western blot experiments. NIH Image 1.62 was used to quantify the band intensities from scanned images of Western blots. Immunofluorescence and green fluorescent protein (GFP) fluorescence were observed using an Axioplan microscope (Carl Zeiss, Thornwood, NY), and fluorescent images were quantified and processed using MetaMorph 4.5 software (Universal Imaging, Downingtown, PA). Samples for electron microscopy were prepared as described previously . Sections (50 --60 nm) were viewed on a CM12 electron microscope (Philips, Eindhoven, Netherlands). Invertase Secretion Assay | Cells were grown overnight in YP medium with 5% glucose at 30C. Cells were diluted to 0.1 --0.2 OD600/ml in YP medium with 5% glucose and were grown at 30C to log phase (0.6 --1.0 OD600/ml). Cells were pelleted and resuspended in 0.1% glucose and incubated at 30C. Aliquots of cells were collected at 0, 15, 30, and 45 min after induction and processed for invertase activity assay . External invertase was assayed at 30C as described by . Total invertase activity was determined by assaying whole cell lysate prepared by freeze-thaw method. Invertase secretion ratio is calculated by dividing total invertase activity by external activity. The small amount of invertase activity at time 0 was subtracted from all time points. RESULTS : NEO1 Is an Essential Gene at All Temperatures | NEO1 was identified in a screen for genes that confer resistance to the aminoglycoside neomycin upon overexpression and was reported to be an essential gene at 30C . Because disruption of DRS2 causes a temperature-conditional growth phenotype, we tested the neo1Delta strain for viability at several different temperatures. NEO1 was placed under transcriptional regulation of the GAL1 promoter, which is tightly repressed when cells are grown on glucose and expressed strongly on galactose . The neo1Delta GAL:NEO1 strain grew well on galactose but failed to grow when plated on media containing glucose at any temperature tested (Table , neo1Delta). Thus, NEO1 provides an essential function that cannot be performed by DRS2, DNF1, 2, or 3. We also found that overexpression of NEO1 could not suppress the cold-sensitive growth defect of drs2Delta strains, nor could overexpression of DRS2 suppress the lethality of neo1Delta, suggesting that there is no significant genetic overlap between DRS2 and NEO1. Table 3 | Growth profile of DRS2/NEO1 family deletion mutants DRS2 and DNF Genes Constitute an Essential Family | Except for NEO1, none of the other members of the DRS2/NEO1 family are essential . To test whether there is any functional redundancy between DRS2 and the DNF genes, we generated strains carrying all possible combinations of the drs2 and dnf null alleles (Tables and and Figure A). All possible double and triple mutants were recovered among the progeny of standard crosses, although this technique failed to yield a viable quadruple mutant. To further test whether the quadruple mutant is viable, the dnf1,2,3Delta strain was transformed with a second copy of the DRS2 gene carried on URA3 marked plasmid (pRS416-DRS2), and the chromosomal copy of DRS2 was disrupted. Serial dilutions of these strains (dnf1,2,3Delta pRS416-DRS2 and drs2Delta dnf1,2,3Delta pRS416-DRS2) were replica-plated to minimal medium containing 5-fluoroorotic acid (5-FOA) to select against cells that retained the pRS416-DRS2 plasmid. As shown in Figure B, the dnf1,2,3Delta strain could lose the DRS2 plasmid, whereas the drs2Delta dnf1,2,3Delta strain could not. In fact, the drs2Delta dnf1,2,3Delta pRS416-DRS2 strain failed to yield colonies on 5-FOA plates at any temperature tested (15, 20, 24, 30, and 37C), indicating that the drs2Delta dnf1,2,3Delta strain is not viable. The fact that any one member of the four genes can maintain the viability of the cells (Table and Figure ) and that the quadruple mutant is not viable indicates that DRS2 and the DNF genes constitute an essential gene family with substantial functional overlap. Figure 1 | DRS2 and DNFs constitute an essential gene family. DRS2 and DNFs constitute an essential gene family. (A) BY4742 (WT), ZHY615D1C (drs2Delta), BY4742 YER166W (dnf1Delta), ZHY2143A (drs2Delta dnf1Delta), ZHY709 (drs2Delta dnf1Delta dnf2Delta), ZHY708 (drs2Delta dnf1Delta dnf3Delta), ZHY7282C (drs2Delta dnf2Delta dnf3Delta), and PFY3273A (dnf1Delta dnf2Delta dnf3Delta) strains were grown on YPD plates at 30, 37, or 20C as indicated in the lower right panel. (B) The dnf1,2,3Delta drs2Delta quadruple mutant is inviable. Serial dilutions of strains PFY3273A pRS416-DRS2 (dnf1,2,3Delta pRS416-DRS2) and ZHY704 (dnf1,2,3Delta drs2Delta pRS416-DRS2) were spotted on YPD or minimal 5-FOA plates, and incubated at 30C (shown), and 37, 24, 20, or 15C (not shown). Growth Profiles of Double and Triple Mutants | Most of the double mutants grow as well as their parental single mutants except for drs2Delta dnf1Delta, which exhibits a strong growth defect at 30C (Figure A). Interestingly, this double mutant grows much better at 37C (Figure A) and can be propagated at this temperature. The remaining double mutants carrying drs2Delta exhibit the same cold-sensitive growth defect as the drs2Delta single mutant, whereas the dnf double mutants grow as well as the wild-type strain . These data indicate that DRS2 and DNF1 show the greatest overlap in function, whereas DNF2 or DNF3 contribute little to viability in the absence of DRS2, which was surprising considering the sequence similarity between DNF1 and DNF2. The drs2Delta dnf1,2Delta and drs2Delta dnf1,3Delta triple deletion strains exhibit a severe growth defect as predicted, because drs2Delta dnf1Delta exhibits a strong growth defect. Interestingly, the drs2Delta dnf1,2Delta and drs2Delta dnf1,3Delta strains grew slightly better than drs2Delta dnf1Delta at 30C but fared worse at 37C (Figure A). The drs2Delta dnf2,3Delta strain showed a slight temperature-sensitive growth phenotype compared with drs2Delta single mutant and the dnf1,2,3Delta strain grew nearly as well as the wild-type strain at all temperatures (Figure A), indicating that DRS2 by itself can fully maintain cell growth. Sporulation Defect and Calcofluor White Hypersensitivity of drs2/dnf Mutants | During the course of generating the strains shown in Table , we noticed that some diploid deletion mutants could not sporulate. To further define this defect, we examined several diploid strains homozygous for the null alleles indicated in Table for their ability to sporulate. Unexpectedly, the homozygous dnf1,2,3Delta diploid exhibited a severe sporulation defect even though it does not show an obvious growth defect. The drs2Delta, dnf1,2Delta, and dnf1,3Delta homozygous strains also exhibited a sporulation defect. However, dnf2,3Delta sporulated normally, as did all the single deletion dnf strains . Table 4 | Sporulation defect and CW sensitivity of drs2/dnf deletion mutants The sporulation defect of homozygous diploids correlated to an increased sensitivity of the haploid mutants to CW . Strains deficient for cell wall chitin are resistant to CW, whereas other mutants that increase chitin deposition or perturb cell wall structure are hypersensitive to this agent. The dnf1,2,3Delta strain exhibited the greatest sensitivity to CW, whereas the single mutants were similar to the wild-type strain. The drs2Delta and dnf1,2Delta strains showed an intermediate sensitivity . We also noticed that the dnf1,2,3Delta cells are elongated when grown at lower temperature or in minimal media (our unpublished observation), which may be caused by a weakened cell wall at the bud tip. Although it is not clear whether the sporulation defect and CW hypersensitivity of these strains are caused by a common defect, these data do indicate that Dnf1p, Dnf2p, and Dnf3p overlap to exert an important cellular function that cannot be replaced by Drs2p or Neo1p. dnf1Delta Exacerbates Membrane Accumulation Defect of drs2Delta | Mutants that perturb protein transport in the secretory pathway typically accumulate abnormal membrane structures. The drs2Delta mutant accumulates abnormal membrane-bound structures containing Golgi enzymes that are similar to Berkeley bodies in morphology . Examination of drs2Delta dnf1Delta cells by electron microscopy (EM) revealed a massive accumulation of membrane-bound structures resembling Berkeley bodies , a more severe phenotype than observed in the drs2Delta single deletion cells. This is consistent with the severe growth defect of drs2Delta dnf1Delta cells and suggested that the drs2Delta dnf1Delta might exhibit a stronger defect in protein transport from the Golgi complex than either single mutant. In contrast, the dnf1,2,3Delta cells did not accumulate abnormal membrane structures and was similar in appearance to wild-type cells . Strains constructed in the BY4741/BY4742 background exhibit more internal membranes than other wild-type strains (e.g., SEY6210), so the appearance of the dnf1,2,3Delta cells shown is not significantly different from wild-type. Figure 2 | drs2Delta dnf1Delta mutant massively accumulates abnormal membrane bound structures resembling Berkeley bodies. drs2Delta dnf1Delta mutant massively accumulates abnormal membrane bound structures resembling Berkeley bodies. BY4742 (WT), ZHY2149D (drs2Delta dnf1Delta), and PFY3273A (dnf1,2,3Delta) cells were prepared for electron microscopy as described previously . Numerous double-membrane ring structures resembling Berkeley bodies (indicated with arrows) accumulate in drs2Delta dnf1Delta cells, whereas the dnf1,2,3Delta cells were similar to wild-type cells. drs2 dnf1 Exhibits a Defect in Protein Transport to Vacuole | The functional redundancy between DRS2 and the DNF genes suggested that the Dnf proteins might be involved in protein transport in the secretory and/or endocytic pathways. To test this, we examined the transport of ALP and CPY to the vacuole. ALP is transported directly from the TGN to the vacuole in adaptor protein (AP)-3 coated vesicles and is processed from a high molecular weight precursor form to the mature form in the vacuole . To determine whether the Dnf proteins are involved in the ALP pathway, we first examined the localization of a GFP-ALP fusion protein in the different mutant strains. In wild-type cells, GFP-ALP was primarily localized to the vacuole membrane, and <4% of the cells contained one to three punctate structures outside of the vacuole (Figure , A and B). In drs2Delta cells, most of the GFP-ALP localized to the vacuole membrane, but an increase in the percentage of cells exhibiting extravacuolar puncta was observed. This defect in GFP-ALP localization was strongly exacerbated in the drs2Delta dnf1Delta cells, approaching the defect observed in the apl5Delta (AP-3) mutant (Figure , A and B). Close to 50% of the drs2Delta dnf1Delta cells contained more than four puncta, with many of the cells also containing a normal-appearing vacuole(s). This suggests that the primary defect in these cells is GFP-ALP mislocalization rather than vacuole fragmentation. This defect is specific to the drs2Delta dnf1Delta mutant cells, because neither dnf1Delta nor the dnf1,2,3Delta mutant exhibited a defect in ALP localization (Figure A). Figure 3 | Drs2p and Dnf1p are required for ALP transport to the vacuole. Drs2p and Dnf1p are required for ALP transport to the vacuole. (A) Localization of GFP-ALP. BY4742 (WT), ZHY615D1C (drs2Delta), ZHY2149D (drs2Delta dnf1Delta), BY4742 YER166W (dnf1Delta), PFY3273A (dnf1,2,3Delta), and 6210 apl5Delta harboring pGO41 (GFP-ALP) were grown at 30C to log phase and examined by fluorescence microscope. (B) Percentage of cells exhibiting abnormal GFP-ALP localization. Cells containing 0, 1 --3, 4 --6, 7 --9, or >=10 punctate structures bearing GFP-ALP were counted and presented as the percentage of the total number of cells counted. Data shown is the average from counting >100 cells from three independent transformants of each strain. (C) Steady-state distribution of ALP forms. Cells were grown in YPD media at 30C to log phase. Whole cell lysates were prepared and immunoblotted to detect precursor (pro) and mature (m) forms of ALP. (D) Kinetics of ALP transport to the vacuole. Cells were grown to log phase at 30C, labeled with [35S]methione/cysteine for 5 min and chased at 0, 5, 10, and 20 min at 30C. ALP was recovered from each sample by immunoprecipitation and subjected to SDS-PAGE. (E) CPY and ALP processing rate assessed from pulse-chase experiments were plotted relative to WT cells. Data shown is the average of at least three experiments. Western blots indicated that the drs2Delta dnf1Delta cells accumulated more ALP precursor (49 +- 29% of total ALP, n = 7) than the wild-type cells (4 +- 0%). The drs2Delta dnf1Delta strains tended to pick up faster growing suppressors, which probably contributed to the wide range of error for this mutant. There was also an intermediate level of proALP accumulation in the drs2Delta cells (18 +- 8%), whereas dnf1Delta was indistinguishable from wild type (Figure C). To follow the transport kinetics of ALP in these mutant strains, cells were pulse-labeled and chased for the times indicated in Figure D. Consistent with the previous results, ALP processing is significantly delayed (similar5-fold) in the drs2Delta dnf1Delta mutant compared with the wild-type cells (Figure E). The drs2Delta mutant again showed an intermediate phenotype, whereas the dnf1,2,3Delta cells exhibited normal kinetics of ALP processing (Figure D). We conclude from these data that Drs2p and Dnf1p are required for the ALP transport pathway, whereas dnf1,2,3Delta does not affect this pathway. We next examined the CPY transport pathway in these mutants. CPY is synthesized in the endoplasmic reticulum (ER) as the p1 precursor form and is modified on N-linked oligosaccharides by Golgi mannosyltransferases to form the p2 precursor. p2 CPY is sorted from secreted proteins in the TGN and is transported through the late endosome to the vacuole where it is processed to the mature form (mCPY) . The CPY pathway is independent of AP-3 but is perturbed in vps mutants, which secrete the p2 form of CPY . For example, vps35 secretes p2 CPY and produces a dark signal on a colony blot probed for CPY (Figure A). The drs2Delta strain exhibits a very modest CPY secretion phenotype, which does not seem to be exacerbated in the drs2Delta dnf1Delta double mutant. Neither dnf1Delta nor the dnf1,2,3Delta triple mutant secreted CPY (Figure A). None of the drs2/dnf deletion strains tested accumulated a significant amount of p2 CPY as detected by Western Blot (Figure B), although a modest increase was observed in the drs2Delta dnf1Delta mutant (Figure B). To further explore the kinetics of CPY processing, these strains were subjected to a pulse-chase analysis at 30C. The drs2Delta mutant exhibited a 1.7 fold kinetic delay in CPY transport, whereas the drs2Delta dnf1Delta mutant exhibited a 3.2-fold kinetic delay (Figures E and C). We had previously noted a partial, Golgi-specific glycosylation defect in drs2Delta that prevents a clear separation of p1 and p2 CPY by SDS-PAGE . This is also evident in the drs2Delta dnf1Delta cells (Figure C), making it difficult to assess the kinetics of ER-to-Golgi transport for CPY. However, analysis of alpha-factor transport (described below) indicated that protein transport from the ER to the TGN is unaffected in the drs2Delta dnf1Delta mutant. Thus, we see exacerbation of the CPY transport defect in the drs2Delta dnf1Delta double mutant, although not as substantial as observed for the ALP pathway. In addition, the Western blot seemed to show that there was less CPY precursor in the dnf1,2,3Delta mutant. However, the pulse-chase result indicated that the kinetics of CPY transport was normal in this mutant (Figure C). Similarly, the other mutants displayed wild-type kinetics of CPY transport (Figure C). Figure 4 | CPY pathway is partially affected in drs2/dnf deletion strains at 30C. CPY pathway is partially affected in drs2/dnf deletion strains at 30C. (A) Colony-blot analysis of CPY secretion. Freshly streaked BY4742 (WT), BY4742 YER166W (dnf1Delta), PFY3273A (dnf1,2,3Delta), ZHY615D1C (drs2Delta), ZHY2149D (drs2Delta dnf1Delta) and 6210 vps35Delta cells were grown on YPD plates at 30C (shown), and 37 or 20C (data not shown) for 24 h. Nitrocellulose membranes were overlaid onto the colonies and incubated for another 24 h. Membranes were washed and probed with a monoclonal CPY antibody. (B) Steady-state distribution of CPY precursor forms. Cells were grown in YPD medium at 30C to log phase. Whole cell lysates were prepared and subjected to SDS-PAGE and Western blotting to detect p1, p2 and mature forms of CPY. (C) Kinetics of CPY transport. Cells were grown to log phase at 30C, labeled with [35S]methionine/cysteine for 5 min, and chased at 0, 5, and 10 min at 30C. CPY was recovered from each sample by immunoprecipitation and subjected to SDS-PAGE. dnf1Delta Does Not Exacerbate Pro-alpha-Factor Processing and Invertase Secretion Defects of drs2Delta | The yeast alpha-factor mating pheromone is synthesized in the ER as a high molecular weight precursor and further modified in the Golgi complex to produce the heterogeneously glycosylated pro-alpha-factor form, which migrates as a high molecular mass smear by SDS-PAGE. This precursor is further processed in the TGN through a series of proteolytic events, initiated by Kex2p, to produce the mature alpha-factor peptide, which is secreted . Kex2p normally cycles between the TGN and the endosomal system . Mutants that mislocalize Kex2p and other TGN proteases required for pro-alpha-factor processing secrete the high molecular weight Golgi form of pro-alpha-factor . The drs2Delta mutant exhibited a defect in pro-alpha-factor processing at 30C (Figure A), although this phenotype is more severe at lower temperatures . The drs2Delta dnf1Delta cells also exhibited an incomplete processing of pro-alpha-factor, but this defect was actually less severe than that observed in the drs2Delta mutant (Figure A). The rapid disappearance of the ER form of pro-alpha-factor in the mutant cells indicates that ER-to-Golgi transport was not perturbed. As shown before, the clathrin light chain mutant (clc1Delta) exhibited a substantial defect with very little mature alpha-factor being produced. Furthermore, the dnf1Delta, dnf1,2,3Delta, and dnf2,3Delta strains processed pro-alpha-factor normally (our unpublished observation). Thus, dnf1Delta does not exacerbate the Kex2p mislocalization phenotype of drs2Delta. Figure 5 | dnf1Delta does not exacerbate the pro-alpha-factor processing and invertase secretion defects of drs2Delta. dnf1Delta does not exacerbate the pro-alpha-factor processing and invertase secretion defects of drs2Delta. (A) Kinetics of pro-alpha-factor processing. Cells were grown to log phase at 30C, labeled with [35]methione/cysteine for 5 min, and chased at 0, 5, and 10 min at 30C. Proalpha-factor was recovered from each sample by immunoprecipitation and subjected to SDS-PAGE. (B) Invertase secretion assay. Cells were grown in 5% YP glucose media at 30C to log phase and shifted to 0.1% YP glucose media to start the induction of invertase. Aliquots of cells were collected at 0, 15, 30, and 45 min after invertase is induced. Secreted and total invertase activities were measured and expressed as the percentage of invertase secreted. There is evidence for two classes of late secretory vesicles in yeast: a dense class carrying invertase and a lighter class carrying Pma1p . In addition, clathrin is required to sort invertase into the dense vesicles, and in the absence of clathrin, invertase is shunted to the Pma1p vesicle pathway . Thus, secretion of invertase is not blocked, although a kinetic defect is observed. To test whether the Dnf proteins are involved in exocytosis, we measured the kinetics of invertase secretion in the different dnf deletion strains. Cells were grown in 5% YP glucose media to log phase and then shifted to 0.1% YP glucose media to induce invertase synthesis. The rate of induction was equivalent for all strains tested (our unpublished observation), so the initial rate of invertase appearance outside of the cell reflected the rate of secretion. In wild-type cells, 40% of newly synthesized invertase was secreted within 15 min after induction and started to plateau (70%) at similar30 min (Figure B). The dnf1,2,3Delta cells exhibited wild-type kinetics of invertase secretion while the drs2Delta and clathrin mutants exhibited a significant delay in this process. Although the drs2Delta dnf1Delta mutant exhibits severe growth and membrane accumulation defects, the invertase secretion defect is no more severe than that of the drs2Delta cells (Figure B). These results suggest that Drs2p is the only family member that contributes significantly to the invertase secretion pathway, and that the severe growth and membrane accumulation defects of the drs2Delta dnf1Delta cells, compared with drs2Delta cells, is not due to a more severe defect in secretion. dnf1,2,3Delta Exhibits a Defect in Recycling and Polarized Localization of GFP-Snc1p | Other than the involvement of Dnf1p in the vacuolar protein transport pathways with Drs2p, we have not found a requirement for the other Dnf proteins in the secretory pathway. We examined the endocytic pathway in these mutants by labeling and chasing the cells with FM4-64, a fluorescent endocytic marker . All the dnf mutants, including the double and triple deletion mutants, were able to endocytose the FM4-64 and deliver the dye to the vacuole at 30C as well as the wild-type strain (our unpublished observations). Interestingly, at 15C the dnf1,2,3Delta cells seemed to internalize FM4-64 from the plasma membrane more rapidly than wild-type cells, but exhibited a delay in transport from endosomes to the vacuole . Surprisingly, the dnf1,2Delta and dnf1,3Delta exhibited a delay in FM4-64 internalization from plasma membrane at 15C, but without accumulation of internal punctate structures. The reason for these differences between the dnf1,2,3Delta triple mutant and the double mutants is unclear. However, all these defects are cold sensitive and reversible, because when cells were shifted up to 22C, the FM4-64 moved quickly to the vacuole (our unpublished observations). Figure 6 | dnf1,2,3Delta mutant exhibits a cold-sensitive defect in FM4-64 transport to the vacuole. dnf1,2,3Delta mutant exhibits a cold-sensitive defect in FM4-64 transport to the vacuole. Cells were stained with FM4-64 on ice and shifted to 15C for 0, 1, or 4 h and then viewed by fluorescence microscopy. During the FM4-64 labeling experiments, we found that the drs2Delta and dnf1,2,3Delta cells always seemed to internalize FM4-64 faster than wild-type cells at either 15 or 30C. Others have proposed that after internalization of FM4-64 to an early endosome, a portion of the dye follows a recycling pathway to the TGN and then back to the plasma membrane . In this case, a defect in the recycling pathway will cause an apparent increase in the rate of FM4-64 internalization. To directly test whether the recycling pathway is perturbed, we examined the localization of a GFP-Snc1p fusion protein as a marker to follow this pathway. Snc1p is an exocytic v-SNARE protein that has been shown to recycle by endocytosis from the plasma membrane to an early endosome, and from there it is sorted to the TGN. From the TGN, Snc1p can be packaged into exocytic vesicles to target them to the plasma membrane . In wild-type cells, GFP-Snc1p was primarily localized to the plasma membrane, concentrating in the bud or the regions of polarized growth. Only a small fraction of the GFP-Snc1p was observed in punctate structures within the cell (Figure A). In contrast, there was a substantial decrease of GFP-Snc1p localized to the plasma membrane in dnf1,2,3Delta cells, with a concomitant increase to internal punctate structures (Figure A), which could be early endosomes or TGN compartments. The GFP-Snc1p that remains on the plasma membrane was evenly distributed between the mother and daughter cells, suggesting a defect in the polarized delivery of Snc1p to the bud. The mislocalization of Snc1p in the dnf1,2,3Delta cells supports the hypothesis that recycling from the early endosome back to the plasma membrane is disturbed in this mutant. Because the kinetics of secretion seems normal in the dnf1,2,3Delta cells (Figure B), we propose that the dnf1,2,3Delta mutant is defective in the endosome to TGN branch of this pathway. The dnf single and dnf2,3Delta mutants traffic GFP-Snc1p normally; however, the dnf1,2Delta and dnf1,3Delta mutants do exhibit a noticeable defect, although less severe than in the dnf1,2,3Delta triple mutant (our unpublished observation). The drs2Delta cells exhibit a similar GFP-Snc1p mislocalization defect as shown in dnf1,2,3Delta cells (Figure A), which is also consistent with the more efficient internalization of FM4-64 observed in these cells. However, this could be explained by the exocytic defect of drs2Delta, and so it is unclear whether drs2Delta also has an endosome to TGN transport defect. Figure 7 | dnf1,2,3Delta mutant exhibits a defect in Snc1p-GFP recycling. dnf1,2,3Delta mutant exhibits a defect in Snc1p-GFP recycling. (A) pRS416 SNC1-GFP was introduced into BY4742 (WT), PFY3273A (dnf1,2,3Delta) and ZHY615D1C (drs2Delta) strains. Cells were grown at 30C to log phase and examined by fluorescence microscopy. (B) Cells harboring pRS426 STE2-GFP were grown at 30C to log phase and examined by fluorescence microscopy. MetaMorph 4.5 was used to process the image and measure the fluorescence intensity on cell surface and in the vacuoles. To rule out the possibility that Snc1p is actually internalized faster in the dnf1,2,3Delta cells rather than a slowed recycling back to the plasma membrane, we examined the localization of GFP-Ste2p in these cells (Figure B). Ste2p, the alpha-factor receptor, is constitutively internalized from the plasma membrane and delivered through the endosomal system to the vacuole for degradation . If endocytosis is enhanced in the dnf1,2,3Delta cells, we would expect to see a decrease in plasma membrane fluorescence relative to internal signal as shown for Snc1p. However, in both wild-type and the dnf1,2,3Delta cells, the ratio of Ste2p-GFP fluorescence intensity on the plasma membrane to that in the vacuole was 0.2 (plasma membrane pixel value / vacuolar pixel value). This result suggests that the mislocalization of Snc1p in the dnf1,2,3Delta cells is not caused by enhanced endocytosis, but is due to inefficient transport in the endosome -> TGN -> plasma membrane loop. Dnf Protein Expression and Localization | To study the expression pattern of the Dnf proteins, we epitope-tagged all five DRS2/NEO1 family members with c-Myc at the carboxyl-terminal end of these proteins. Immunoblotting of whole cell extracts indicated that all three previously uncharacterized genes (DNF1, 2, and 3) were expressed under normal growth conditions (Figure A). Neo1p and Drs2p are expressed at the highest level, whereas Dnf3p is expressed at a very low level compared with the other proteins (Figure B). This is consistent with the observation that Dnf3p seems to contribute the least to both cell viability and CW sensitivity. The protein level for each family member was greatly decreased when the cells were grown at 37C and were moderately higher at 20C as revealed by both Western blot and immunofluorescence analyses (our unpublished observations). Consistent with the growth profiles, this suggests that cells have a greater requirement for Drs2p and Dnf proteins at lower temperatures. The marked susceptibility to proteolysis, noted previously for Drs2p, is also the case for Neo1p and the Dnf proteins. We cannot distinguish at this time whether the proteolysis occurs in vivo or during preparation of the cell lysates. Figure 8 | Expression levels of Drs2/Neo1 family proteins. Expression levels of Drs2/Neo1 family proteins. (A) Strains BY4742 (control), ZHYNEO1-MYC, ZHYDRS2-MYC, ZHYDNF1-MYC, ZHYDNF2-MYC, and ZHYDNF3-MYC were grown in YPD media at 30C. Equal amounts of whole cell lysates were subjected to SDS-PAGE and immunoblotting with a monoclonal c-Myc antibody to detect the fusion proteins. The migration of standard protein markers is labeled at the left. Predicted molecular mass of each full-length protein fused with 13XMyc tag is Neo1p, 146 kDa; Drs2p, 170 kDa; Dnf1p, 194 kDa; Dnf2p, 198 kDa; and Dnf3p, 204 kDa. Each lane contained an equivalent amount of CPY (not shown). (B) Full length and near full-length bands from A were analyzed by NIH Image 1.62. The signal intensity of each fusion protein was plotted as the percentage of Drs2p signal strength. Data shown in this figure is representative of three experiments. We then examined the localization of the epitope-tagged proteins by immunofluorescence microscopy and found that all the proteins exhibited a punctate staining pattern. However, the three Dnf proteins show different localization patterns (Figure A). Dnf1p is found in internal membranes and the plasma membrane, but it was more concentrated at the emerging bud site, small buds, and the mother-daughter neck of dividing cells (Figure , a and e), which is very similar to the localization of cortical actin patches. This suggests that Dnf1p may be concentrated in exocytic vesicles targeted to polarized sites of growth. Dnf2p is found in unique, smaller punctate structures that lie directly underneath the plasma membrane (Figure , b and f), but it does not exhibit a polarized distribution as Dnf1p does. The punctate structures might be endocytic vesicles or plasma membrane invaginations. In some cells, small regions of continuous plasma membrane staining could be seen for Dnf2p. Dnf3p staining reveals punctate structures evenly spread throughout the cell, which resembles the staining pattern for Drs2p (Figure , c, d, g, and h). Figure 9 | Localization of Dnf proteins. Localization of Dnf proteins. Cells were grown in YPD media at 30C and prepared for immunofluorescence. (A) Cells were visualized with mouse monoclonal c-Myc antibody (a --d). (B) Cells transformed with a 2 mu KEX2HA plasmid were stained with a mouse monoclonal c-Myc antibody and a rabbit polyclonal HA antibody to visualize Myc-tagged Drs2/Dnf proteins (red) and HA-tagged Kex2p (green) (e --h). The yellow punctate structures indicate the colocalization of Drs2/Dnf proteins with Kex2p. Even though Dnf1p and Dnf2p exhibit unique localization patterns clustered on or near the plasma membrane, both were also found distributed in internal structures as well. Drs2p colocalizes substantially with the TGN marker Kex2p. We examined the colocalization of the Dnf proteins with Kex2p and found that all of them colocalized with Kex2p to some extent (Figure B). Dnf3p showed the greatest overlap with Kex2p (Figure B, g) comparable with the staining pattern observed for Drs2p (Figure B, h). Dnf1p and Dnf2p showed the least overlap with Kex2p, and further experiments will be required to determine whether a small percentage of Dnf1p and Dnf2p is localized to the TGN. The localization of the Dnf1 and Dnf2 proteins to plasma membrane and what seem to be transient exocytic and/or endocytic membranes suggests that these proteins might cycle in a TGN-plasma membrane --endosome --TGN pathway. This is consistent with the defect in GFP-Snc1 recycling in the dnf1,2,3Delta mutant. DISCUSSION : We have previously implicated Drs2p in clathrin function at the TGN , and this report focuses on the requirements for Drs2p and Dnfs in protein transport. We find that 1) the DRS2 and DNF genes constitute an essential gene family with overlapping functions. 2) Specific protein transport steps in the late secretory, endocytic, and vacuolar pathways have specific requirements for Drs2/Dnf proteins . 3) Drs2p, Dnf1p, and Dnf2p exhibit unique localization patterns, suggesting that each may have a primary function at different transport steps. Figure 10 | Model for the roles of Drs2p and Dnf proteins in late secretory, vacuolar, and endocytic pathways. Model for the roles of Drs2p and Dnf proteins in late secretory, vacuolar, and endocytic pathways. PM, plasma membrane; EE, early endosome; LE, late endosome. Of the five DRS2/NEO1 family genes, NEO1 is the only essential gene and it does not seem to overlap genetically with the other four family members. Yeast can survive with multiple combinations of null alleles for DRS2, DNF1, 2, and 3 but the quadruple mutant is inviable. The fact that any single member of the Drs2/Dnf group can support cell growth demonstrates a functional interchangeability for these proteins, and suggests that each protein shares a common biochemical activity. A substantial body of work suggests that these proteins are APTs, although this activity has not yet been reconstituted with purified enzyme. Although it is possible that the Drs2 and Dnf proteins transport different substrates, it seems more likely that the existence of multiple members of this family reflects a requirement for these proteins at different subcellular organelles. This view is supported by the unique localization patterns observed for Drs2p, Dnf1p, and Dnf2p. Role of Drs2p and Dnf1p in Protein Transport from TGN | Drs2p and Dnf1p seem to contribute most of the essential function of the Drs2/Dnf subgroup because the drs2Delta dnf1Delta double mutant grows very slowly. In addition, dnf1Delta strongly exacerbates the membrane accumulation phenotype of drs2Delta, suggesting an overlap in function for protein and membrane exit from the Golgi complex. Multiple exit pathways were examined and we found that the ALP pathway was most significantly perturbed in the drs2Delta dnf1Delta double mutant. A less severe synthetic defect in the CPY pathway was also observed in this double mutant, although the defects in the invertase secretion kinetics or pro-alpha-factor processing observed in the drs2Delta mutant were not exacerbated in the double mutant. The redundant roles for Drs2p and Dnf1p in the ALP pathway are somewhat surprising because only a fraction of Dnf1p seems to localize to the TGN, where Drs2p is localized. Rather, Dnf1p is found in membranes concentrated near sites of polarized growth in wild-type cells. These include the incipient bud site, tips of small buds, and the mother/bud neck during cytokinesis, a distribution very similar to that of actin cortical patches in yeast. This dynamic localization pattern suggests that Dnf1p is cycling between the exocytic and endocytic pathways and could therefore be a transient occupant of the TGN. Because drs2Delta partially perturbs the exocytosis of invertase, it is possible that Dnf1p exocytosis is also inefficient, resulting in an increased residence in the TGN of drs2Delta cells. This could allow Dnf1p to partially compensate for the loss of Drs2p in the ALP pathway. We have observed modest defects in ALP transport in the drs2Delta single mutant but not in the dnf1Delta mutant. Therefore, we suggest that Drs2p play a primary role, and Dnf1p a compensatory role, in budding AP-3 --coated vesicles carrying ALP to the vacuole . Clathrin has recently been implicated in the invertase secretion pathway with the observation that clathrin mutants missort invertase into the less dense vesicle fractions containing Pma1p . Mutations that perturb the actin cytoskeleton also affect invertase secretion and cause the accumulation of the invertase class of exocytic vesicles . This is presumably because the targeted delivery of these vesicles to polarized sites of growth requires actin. We have found that Drs2p and clathrin are required to form vesicles that accumulate upon actin disruption and that these vesicles contain acid phosphatase, another cargo of the dense exocytic vesicles (Gall et al., submitted). Because Drs2p localizes to the TGN and this seems to be site where invertase and Pma1p are segregated into separate exocytic pathways, we prefer a model in which the invertase vesicles bud directly from the TGN for targeting to polarized sites of growth [Figure , Invertase (A)]. Surprisingly, the sorting of invertase and Pma1p into two separate exocytic pathways at the TGN is also perturbed by several mutations that disrupt protein trafficking through the late endosome. These observations have led others to propose that invertase is sorted first to a late endosome and is then packaged into clathrin-coated vesicles for exocytosis [Figure , Invertase (B)]. However, other interpretations are possible. For example, late Golgi proteins continuously cycle between the TGN and endosomes and so a strong block in endosome trafficking could rapidly deplete the TGN resident proteins required for sorting invertase into CCVs. In addition, there is no direct evidence that invertase passes through the late endosome (or prevacuolar compartment) of wild-type cells before it is secreted. Further work will be required to clearly define the donor compartment that buds the invertase vesicles. A function for Drs2p in vesicle budding was first suggested by the discovery of a synthetic lethal interaction between drs2Delta and arf1 mutations . ARF has been implicated in recruiting tetrameric adaptins (AP-1, AP-3, and AP-4) and clathrin to the TGN and endosomes . The CPY pathway may use clathrin and the GGA adaptins, also ARF-dependent for membrane association , for exit from the TGN. The requirements for Drs2p and Dnf proteins in these different ARF-dependent pathways are very interesting. Drs2p seems to be the only family member capable of supporting the clathrin-dependent invertase pathway because the drs2Delta single mutant exhibits a defect in this pathway (Figure B; ). However, both Drs2p and Dnf1p seem to be capable of contributing to the AP-3/ALP pathway. The CPY transport defect observed in the drs2Delta dnf1Delta mutant suggests that the Drs2/Dnf proteins may also contribute to a GGA-dependent pathway as well . Unfortunately, we have not yet successfully isolated conditional alleles of the DRS2/DNF genes that would allow us to score the immediate consequences of a complete loss of function for all four proteins. Requirement for Dnf1, 2, and 3p in Snc1p Recycling Pathway | The dnf1,2,3Delta mutant grows well and does not exhibit a defect in the ALP, CPY, or invertase pathways. However, this mutant does exhibit a striking defect in GFP-Snc1p recycling. Snc1p is a v-SNARE that is thought to function in exocytic and endocytic vesicles, but at steady-state most of the GFP-Snc1p fluorescence is concentrated on the plasma membrane of wild-type cells . This localization pattern suggests that internalization of GFP-Snc1p from the plasma membrane is rate limiting in the recycling pathway shown in Figure . This localization pattern is altered in the dnf1,2,3Delta mutant with most of the GFP-Snc1p concentrated in internal structures. Because exocytosis does not seem to be perturbed, we suggest that endosome to TGN transport is partially defective in this mutant. The recycling defect could also explain the CW hypersensitivity and sporulation defects of the dnf1,2,3Delta mutant. Chitin synthase III (Chs3p) is known to be dynamically recycled through endocytic intermediates for targeting to polarized sites of growth . Recent work strongly suggests that ARF, AP-1, and clathrin are required for transport of Chs3p from an early endosome to the TGN. Mutations in these clathrin coat components results in an unregulated transport of Chs3p to the cell surface, and increased sensitivity of cells to CW . Clathrin and AP-1 have also been implicated in endosome -> TGN as well as TGN -> plasma membrane routes in mammalian cells . The formation of spore membranes and cell walls may also require a dynamic recycling of materials. In fact, Chs3p is delivered to the prospore membranes for synthesis of the spore cell wall , and homozygous clathrin mutants also fail to sporulate . The localization of Dnf2p to both internal structures as well as the plasma membrane suggests that this protein may also be cycling between the endocytic and exocytic pathways. Thus, Dnf1p and Dnf2p would be positioned to directly contribute to the clathrin/AP-1 --dependent endosome to TGN route. Remarkably, our studies position a Drs2-related P-type ATPase in each pathway where an ARF/AP/clathrin-dependent vesicle budding event has been implicated. In fact, we have observed a dramatic defect in the ability of drs2Delta mutants to produce CCVs at 15C, whereas normal-appearing CCV can be isolated at 30C. We postulate that Dnf1, 2, and 3 proteins facilitate CCV budding in the drs2Delta mutant at 30C but fail to do so at lower temperatures. This would explain why the clathrin deficiency phenotypes of drs2Delta are much more severe at 20C or below. APT Genes in Other Organisms | The DRS2/NEO1 family of potential APT genes exists ubiquitously in eukaryotes, and a few other members of this large family have been identified genetically. The PDE1 gene from rice blast fungus was identified in a screen for pathogenicity factors and is most similar to DNF3. Mutation in PDE1 perturbs formation of penetration hyphae, a narrow outgrowth of the cell surface used to puncture the plant cuticle. The Talbot group has suggested that Pde1p could play a role in the massive recycling of membrane and cell wall biosynthetic enzymes required for the formation of penetration hyphae . The defect in GFP-Snc1p recycling and polarized delivery we have observed in the dnf1,2,3Delta mutant is consistent with this interpretation. The ALA1 gene from Arabidopsis is most similar to DRS2 and remarkably, decreased expression of ALA1 causes cold-sensitive growth of Arabidopsis. In addition, the ALA1 gene can complement the cold-sensitive growth defect of the drs2Delta mutant underscoring the conservation of Drs2p/Ala1 function across phylogenetic boundaries. The ALA1 gene also complements a cold-sensitive defect in APT activity at the plasma membrane of drs2Delta cells . Because Drs2p is not localized to the plasma membrane but is required for protein transport in the late secretory pathway, this defect may be caused by mislocalization of other APTs. Our localization data suggest that Dnf1 and Dnf2 are the best candidates for the plasma membrane APT in yeast, and we are currently testing this possibility. The roles of the Drs2/Dnf family in protein traffic in yeast could also shed light on the function of their homologs implicated in human disease. Both FIC1 and BSEP mutations cause familial intrahepatic cholestasis in humans. FIC1 is most homologous to the DNF1 and DNF2 genes in yeast and BSEP is a bile salts export pump (in the ABC ATPase family) localized to the bile canalicular plasma membrane of hepatocytes . How FIC1 is involved in bile production is not clear, but based on the phenotype of the drs2/dnf yeast mutants, it is possible that FIC1 might be required for the polarized delivery of BSEP to the canalicular membrane. ATP10C, another member of the APT family in humans, has been linked to the neurological disorders Angelman syndrome and autism . Recycling of membrane from synapses plays a critical role in synaptic vesicle production, a process where ARF and AP-3 have been implicated . In addition, the recycling or polarized delivery of AMPA receptors to specific sites on the plasma membrane is essential for proper function of neurons and may contribute to long-term potentiation . Interestingly, mutations in the mouse ATP10C homolog pfatp are linked to increased body fat . This phenotype could have a neurological basis or could reflect perturbations in the trafficking of glucose transporters or insulin receptors in adipocytes. The precise mechanism by which Drs2p and the Dnf proteins contribute to protein trafficking remains to be elucidated. In the case of Drs2p, mutation of aspartic acid critical for ATPase activity strongly perturbs the formation of clathrin-coated vesicles targeted to the plasma membrane . This suggests that lipid translocation is required to either generate a membrane environment permissive for vesicle budding or perhaps to help deform the membrane during vesicle budding (see , for further discussion on the potential roles of lipid translocases in vesicle budding). Further biochemical analysis of these proteins will be required to better assess their role in protein trafficking. Backmatter: PMID- 12221124 TI - Regulation of Focal Adhesion Kinase by a Novel Protein Inhibitor FIP200 AB - Focal adhesion kinase (FAK) is a major mediator of integrin signaling pathways. The mechanisms of regulation of FAK activity and its associated cellular functions are not very well understood. Here, we present data suggesting that a novel protein FIP200 functions as an inhibitor for FAK. We show the association of endogenous FIP200 with FAK, which is decreased upon integrin-mediated cell adhesion concomitant with FAK activation. In vitro- and in vivo-binding studies indicate that FIP200 interacts with FAK through multiple domains directly. FIP200 bound to the kinase domain of FAK inhibited its kinase activity in vitro and its autophosphorylation in vivo. Overexpression of FIP200 or its segments inhibited cell spreading, cell migration, and cell cycle progression, which correlated with their inhibition of FAK activity in vivo. The inhibition of these cellular functions by FIP200 could be rescued by coexpression of FAK. Last, we show that disruption of the functional interaction between endogenous FIP200 with FAK leads to increased FAK phosphorylation and partial restoration of cell cycle progression in cells plated on poly-l-lysine, providing further support for FIP200 as a negative regulator of FAK. Together, these results identify FIP200 as a novel protein inhibitor for FAK. Keywords: INTRODUCTION : Focal adhesion kinase (FAK) is a major mediator of signal transduction by integrins, which has been implicated in the regulation of cell spreading, migration, survival, and proliferation . FAK activation and tyrosine phosphorylation have been shown in a variety of cell types to be dependent on integrins binding to their extracellular ligands . On its activation, FAK is autophosphorylated at Y397, which mediates FAK association with a number of Src homology 2 (SH2) domain-containing signaling molecules, including Src family kinases , p85 subunit of PI3K , phospholipase C-gamma , and Grb7 . FAK binding to Src family kinases has been proposed to allow phosphorylation of Y925 of FAK by Src, which binds to the SH2 domain of Grb2 . The FAK/Src complex formation also leads to tyrosine phosphorylation of a number of other substrates, including paxillin , p130cas , and Shc . Recent studies have shown that Grb7 is phosphorylated by FAK in a Src-independent manner . FAK and its downstream signaling pathways have been shown to play important roles in the regulation of cell spreading and migration . FAK-/- fibroblasts derived from FAK-knockout mouse embryo showed a significant decrease in cell migration compared with the cells from wild-type mice . Similarly, inhibition of FAK by the FAK C-terminal recombinant protein (i.e., FRNK) caused decreased motility of both fibroblasts and endothelial cells , as well as a reduced rate of fibroblast spreading . Last, overexpression of FAK in a number of cell lines, including the FAK-/- cells, promoted their migration on fibronectins (FN) . FAK signaling pathways have also been shown to regulate cell survival and cell cycle progression in integrin-mediated cell adhesion. Overexpression of FAK protected cells from apoptosis induced by cell detachment, serum withdraw, or other treatments in MDCK cells or primary fibroblasts . Conversely, inhibition of FAK by treatment of tumor cell lines with FAK antisense oligonucleotides or by microinjection of CEF cells with an anti-FAK monoclonal antibody- (mAb; ) induced apoptosis. Microinjection of the C-terminal fragment of FAK into either fibroblasts or endothelial cells inhibited cell cycle progression as measured by bromodeoxyuridine (BrdU) incorporation . Inhibition of FAK tyrosine phosphorylation by disruption of FN matrix assembly also resulted in the delay of the G1 to S transition, suggesting a role for FAK in cell cycle progression . Finally, using a tetracycline-regulated expression system, we have shown recently that expression of wild-type FAK accelerated G1 to S transition, whereas expression of a dominant negative FAK mutant inhibited cell cycle progression at G1 phase . In contrast to rapid progress in elucidating the FAK downstream signaling pathways, relatively little is known about the mechanisms of regulation of FAK activity and its associated cellular functions. Using the yeast two-hybrid screen, we have recently identified a novel protein, FAK-family interacting protein of 200 kDa (FIP200), that is associated with the FAK-related tyrosine kinase Pyk2 . Our initial analysis indicated that FIP200 could inhibit the kinase and cellular activity of Pyk2 by binding to its kinase domain directly. Furthermore, FIP200 could also bind to FAK. Interestingly, both FIP200 and FAK are widely expressed in a variety of tissues and cell lines in contrast to the limited expression pattern of Pyk2 . This suggests a potentially important role for FIP200's interaction with FAK in some fundamental cellular functions. In this report, we show that FIP200 could also bind to the kinase domain of FAK and function as a protein inhibitor for FAK kinase activity and its associated cellular functions. MATERIALS AND METHODS : Antibodies | Polyclonal antibodies against the C-terminal FIP200 (residues 1374 --1591; anti-FIP200C; ), rabbit antiserum against FAK , mouse mAb KT3 , and mouse mAb 12CA5 that recognize the hemagglutinin (HA) epitope tag have been described previously. Antiserum against the N-terminal segment of FIP200 was prepared in rabbits using a glutathione S-transferase (GST)-fusion protein containing residues 1 --373 within N terminus of FIP200. Anti-FIP200N antibodies were then affinity purified from the antiserum using the same fusion protein immobilized on glutathione-Sepharose as an affinity matrix. Mouse mAbs against FAK, Pyk2, and paxillin, and antiphosphotyrosine antibody, PY20, were purchased from Transduction Laboratories (Lexington, KY). Rabbit antibody against phosphorylated Y397 of FAK (anti-pFAKY397) was purchased from Biosource (Camarillo, CA). Rabbit anti-HA (HA probe), mouse mAb against Myc epitope tag (9E10), and rabbit polyclonal anti-green fluorescent protein (GFP) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit polyclonal anti-beta-Gal was from 5 prime- 3 prime, Inc. (Boulder, CO). Mouse monoclonal anti-Flag, anti-BrdU, fluorescein-conjugated goat anti-rabbit immunoglobulin (Ig) G, and rhodamine-conjugated goat anti-mouse IgG were purchased from Sigma (St. Louis, MO). Construction of Expression Vectors | The expression vectors pSG5-FIP200, pSG5-N-terminal-FIP (NT-FIP), and pSG5-C-terminal-FIP (CT-FIP) encoding Flag-tagged full-length NT-FIP and CT-FIP have been described previously . pSG5-middle domain-FIP (MD-FIP) encoding Flag-tagged middle domain of FIP200 was generated by amplifying residues 639-1373 of FIP200 using primers with EcoRV site at the 5' end and BglIsite at the 3' site. The region was subsequently cloned into the corresponding cloning sites in pSG5 vector. Similarly, expression vectors pKH3-FIP200, pKH3-NT-FIP, pKH3-MD-FIP, and pKH3-CT-FIP encoding HA-tagged FIP200 or fragments were generated by amplifying residues 1 --1591, 1 --638, 639-1373, and 1374 --1591 with the addition of SmaI site at 5' end and EcoRV site at 3' end. These fragments were subsequently digested and cloned into corresponding cloning sites in pKH3 vector. pGEX-CT-FIP has been described previously . pGEX-NT-FIP was constructed by performing polymerase chain reaction (PCR) to generate a 1.1-kb N-terminal fragment corresponding to residues 1 --373 within NT-FIP with the addition of SmaI site at the 5' end and EcoRV site at the 3' end. This fragment was digested with SmaI and EcoRV and was inserted into the corresponding cloning site of pGEX-2T vector. pGEX-MD-FIP was generated by amplifying region corresponding to residue 639-1373 of plasmid encoding full-length-FIP200. The primers included a SmaI site at 5' end and EcoRV site at 3' end. The fragment was digested with SmaI and EcoRV and was inserted into corresponding sites into the pGEX-2T vector. FAK segment containing N-terminal domain (NT-FAK) was generated by PCR amplification using the forward (5'-CTGGATCCATGGCAGCTTACCTTG-3') and reverse (5'-ATGATATCTTAAGTATCTTC TTCATC-3') primers. The PCR product was digested with BamHI and EcoRV and was cloned into pKH3 at BamHI and SmaI site to generate pKH3-NT-FAK. FAK segment containing the kinase domain (KD-FAK) was generated by PCR amplification using the forward (5'-ATGATATCAACCAGAGATTATGAAATTC-3') and reverse (5'-GCTTTAAATTAAGTAAACCTGGGTCGTCTAC-3') primers. The PCR product was digested with EcoRV and DraI and was cloned into pKH3 at SmaI site to generate pKH3-KD-FAK. The same primers were used to amplify the kinase domain with K454 to R mutation using a FAK cDNA with this mutation as the template . This fragment was then cloned into pKH3 vector to make the HA-tagged KDKR construct. The expression vectors encoding full-length HA-tagged FAK and the C-terminal FAK have been described previously . The kinase domain of Pyk2 was generated by PCR amplification using the sense (5'-CCAGGATCCGGCATTGCCCGTGAAGATG-3') and antisense (5'-ATGAATTCGCTTCACACCAGCTCGGTG-3') oligonucleotides. The product was then inserted into pKH3 to generate pKH3-KD-Pyk2. The vector encoding full-length Pyk2 has been described previously . The expression vectors encoding HA-tagged Grb7 and the control protein (Grb7-SH2 domain) have been described previously . Expression vectors encoding GFP-paxillin, HA-Shc, and Myc-p130cas were kind gifts from Drs. C. Turner (Upstate Medical Center, Syracuse, NY), D. Schlaepfer (Scripps Research Institute, La Jolla, CA) and S. Hanks (Vanderbilt University, TN), respectively. In Vitro Binding | GST fusion proteins were produced and purified using a protease-defective Escherichia coli strain BL21-Dex, as described previously . GST fusion proteins (3 mug) were immobilized on glutathione-agarose beads and were then incubated for 4 h at 4C with lysates (200 mug) prepared from 293 cells that had been transfected with expression vectors encoding kinase domain of Pyk2, HA-FAK, or its fragments. After washing, the bound proteins were analyzed by Western blotting with anti-HA (1:2000) as described below. For binding to the recombinant FAK, His-tagged recombinant FAK was purified from baculovirus-infected sf21 cells as described previously . GST-fusion proteins (5 mug) were equalized for amount of glutathione agarose beads and were incubated with 1 mug of purified His-tagged FAK in binding buffer (20 mM Tris, pH 8.0, 137 mM NaCl, 1 mM MgCl2, and 1% Triton) overnight at 4C with rotation. The samples were then washed five times with binding buffer, boiled in SDS buffer, resolved by SDS-PAGE, and western blotted with alpha-FAK antibody. Immunoprecipitation and Western Blot | For most experiments, cells were lysed with 1% NP-40 lysis buffer (20 mM Tris, pH 8.0, 137 mM NaCl, 1% Nonidet P40, 10% glycerol, 1 mM Na3VO4, 1 mM phenyl methyl sulfoxide, 10 mug/ml aprotinin, and 20 mug/ml leupeptin). For experiments to detect phosphorylation of HA-Shc, cells were lysed in the modified RIPA lysis buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 0.3% sodium deoxycholate, 0.1% Nonidet P-40, 10% glycerol, 1.5 mM MgCl2, 1 mM EDTA, 0.2 mM EGTA, 20 mM NaF, 25 muM ZnCl2, 1 mM NaVO4, 1 mM phenyl methyl sulfoxide, 10 mug/ml aprotinin, and 2 mug/ml leupeptin) as described previously . Immunoprecipitation was carried out at 4C by incubating cell lysates for 2 --4 h with indicated antibodies followed by incubation for 1 h with Protein A-Sepharose or Protein G-Plus. Immunoprecipitates were washed three times in lysis buffer without protease inhibitors. The beads were then resuspended in SDS-PAGE sample buffer, boiled for 5 min, and resolved by SDS-PAGE. Western blotting was performed with appropriate antibodies as indicated, using the Amersham enhanced chemiluminescent system (Arlington Heights, IL), as described previously . In some experiments, whole cell lysates were analyzed directly by Western blotting. FAK In Vitro Kinase Assay | FAK was immunoprecipitated from Chinese hamster ovary cells overexpressing FAK . Aliquots of the immune complex were assayed for kinase activity as described previously in the presence of various amounts of GST fusion proteins containing FIP200 segments or GST alone. Measurement of Cell Spreading | NIH3T3 cells were transfected using the LipofectAmine and PLUS transfection reagents (Life Technologies, Grand Island, NY) according to the manufacturer's instructions. One day after transfection, the cells were replated on FN (10 mug/ml), fixed in formaldehyde, and processed for immunofluorescence staining (see below). Alternatively, cells were cotransfected with a plasmid encoding beta-gal along with the indicated vectors. One day after transfection, cells were replated on FN (10 mug/ml) for 30 min, fixed, and assayed for beta-gal activity as described previously . At least 60 positively transfected cells (blue) were counted for their spreading phenotype in each transfection in three independent experiments. Cell Migration Assays | NIH 3T3 cells were cotransfected with various vectors along with a plasmid encoding GFP in 7:1 ratio using the LipofectAmine and PLUS transfection reagents (Life Technologies) according to the manufacturer's instructions. One or 2 d after transfection, the cell monolayer was wounded with a p10 tip. The plates were then washed and incubated at 37C in growth medium for 8 h. Phase contrast and fluorescence images were taken every 2 h until the wound closed (similar10 h). The rate of migration was calculated by measuring the distance moved toward the center of the wound in 8 h. Motility assays using OMAware were as described previously . Measurement of Cell Cycle Progression by BrdU Incorporation | BrdU incorporation assays were performed as described previously . Briefly, NIH 3T3 cells were transfected using the LipofectAmine and PLUS transfection reagents (Life Technologies) according to the manufacturer's instructions. The subconfluent transfected cells were serum starved for 24 h in DME with 0.5% CS. They were then replated on FN (10 mug/ml) and incubated for 16 h with 100 muM BrdU (Sigma) in DME plus 10% CS. For experiments with FAK-KDKR mutant, cells were serum starved for 30 h in 0.5% serum. They were then replated on FN (10 mug/ml) or poly-l-lysine (PLL; 0.1 mg/ml) and incubated for 20 h with 100 muM BrdU in 1% serum. Cellular DNA was digested with 0.5 U/mul DNaseI (, Beverly, MA) for 30 min at 37C. Cells were then processed for double immunofluorescence staining with polyclonal anti-HA (HA probe; 1:300) and monoclonal anti-BrdU (1:300) as described below. At least 80 positively transfected cells (as recognized by anti-HA) in multiple fields were scored for BrdU staining in each independent experiment. For FAK rescue experiments, an expression plasmid encoding beta-Gal was also included in transfections. Cells were then analyzed for BrdU incorporation as described above, except that the positively transfected cells were identified by immunostaining with polyclonal anti-beta-Gal. The percentage of BrdU+/beta-Gal+ cells was determined by analyzing 40 --50 beta-Gal+ cells for each transfection in multiple fields. Immunofluorescence Staining | Cells were processed for immunofluorescence staining as described previously . The primary antibodies used were polyclonal anti-FIP200N (1:200), monoclonal anti-FAK (1:100), polyclonal anti-HA (1:200), polyclonal anti-beta-Gal (1:300), monoclonal anti-BrdU (1:200), and monoclonal antivinculin (1:50). The secondary antibodies used were fluorescein-conjugated goat anti-rabbit IgG (1:300) and rhodamine-conjugated goat anti-mouse IgG (1:200). The cells were mounted on Slowfade (Molecular Probes, Eugene, OR) and examined. The image of stained cells was captured using an immunofluorescence microscope (, Tokyo, Japan) and a charged-coupled device camera. RESULTS : Association of Endogenous FIP200 with FAK | To explore the mechanism and potential function of FIP200 interaction with FAK, we first analyzed interaction of endogenous FIP200 and FAK. Lysates were prepared from cells that had been suspended or replated on FN, type IV collagen, or type I collagen. They were immunoprecipitated by an antibody against FIP200 and then subjected to western blotting with anti-FAK to detect associated FAK in the immune complexes. Figure A shows association of endogenous FAK with FIP200 and that the association was decreased upon cell adhesion to FN, and to a less extent, type IV collagens or type I collagen. Western blotting of the immunoprecipitates with another antibody against FIP200 showed similar amounts of FIP200 precipitated from cells lysates under these different conditions (Figure B). Consistent with previous studies , cell adhesion to FN, and to a less extent, type IV collagens or type I collagen, activated FAK that lead to increased FAK autophosphorylation at Y397 (Figure , C and D). These results suggest that FIP200 dissociation from FAK is correlated with FAK activation during cell adhesion, which is consistent with our previous finding that FIP200 may also function as a protein inhibitor for FAK . These coimmunoprecipitation analyses also detected association of endogenous FIP200 and FAK in several other cell lines, including rat aortic smooth muscle cells, 293 cells, and NIH 3T3 cells (S. Abbi and J. Guan, unpublished data). Figure 1 | Association and localization of FIP200 and FAK. Association and localization of FIP200 and FAK. (A-D) Lysates were prepared from MDA-MB231 breast carcinoma cells that had been suspended, or replated on FN, type IV collagen, or type I collagen, as indicated. They were immunoprecipitated by anti-FIP200N and then analyzed by Western blotting with anti-FAK (A) or anti-FIP200C (B). The whole cell lysates (WCL) were also analyzed directly by Western blotting with anti-pFAKY397 (C) or anti-FAK (D). (E) NIH 3T3 cells (top panels) or ECV304 cells transfected with FIP200 (bottom panels) were processed for immunofluorescence as described in "Materials and Methods" using anti-FIP200N or anti-FAK, as indicated. Examples of colocalization of FIP200 and FAK in peripheral focal contacts are marked by arrows. Our previous studies suggested that FIP200 was predominantly localized in the cytoplasm . Using the new polyclonal antibody against the N-terminal domain of FIP200, we detected presence of endogenous FIP200 in the focal contacts in the cell periphery in addition to the cytoplasmic staining in a fraction of the cells (Figure E). Costaining with anti-FAK (Figure E, top panels) or anti-vinculin (S. Abbi and J. Guan, unpublished data) showed partial colocalization of FIP200 with FAK and vinculin in the focal contacts. This partial colocalization of FIP200 with FAK in focal contacts in the periphery of the cells was also seen more clearly in cells transfected with the full-length FIP200 (Figure E, bottom panels). These results suggested that at least part of FIP200 was partially colocalized with FAK. FIP200 Association with FAK through Multiple Interaction Domains | To define the FAK-binding domains within FIP200, we coexpressed HA-tagged FAK with Flag-tagged FIP200 and several FIP200 segments (see Figure A) in 293 cells. Immunoprecipitations were performed with anti-Flag antibody and were followed by western blotting with anti-FAK antibody. As shown in Figure B, FAK is coprecipitated with the full-length FIP200 and the CT-FIP, which is consistent with our previous result . Surprisingly, however, both the FIP200 NT-FIP and MD-FIP segments also associated with FAK in these experiments. We then performed in vitro-binding assays to determine whether all three FIP200 segments bound to the same region on FAK. Figure C shows that GST fusion proteins containing any of the three FIP200 segments bound to the full-length FAK, whereas GST alone did not. Interestingly, GST fusion proteins containing NT-FIP or MD-FIP bound to the kinase domain of FAK, whereas GST fusion protein containing CT-FIP bound to the N-terminal region of FAK. None of the GST fusion proteins bound to the C-terminal region of FAK. The interaction of NT-FIP and MD-FIP with the kinase domain of FAK was specific because they did not interact with the kinase domain of Pyk2, a homolog of FAK, in the same experiment. Likewise, GST alone did not bind to any of the FAK domains as expected. To examine whether all three FIP200 fragments bound to FAK directly or indirectly through other proteins in the 293 cell lysates, we used purified recombinant FAK from insect cells in the same in vitro-binding assays. Figure D shows that GST fusion proteins containing NT-FIP, MD-FIP, or CT-FIP, but not GST alone, bound to the recombinant FAK. Taken together, these results demonstrate that FIP200 could associate directly and specifically with FAK through multiple interaction domains. Figure 2 | Analysis of FIP200 association with FAK. Analysis of FIP200 association with FAK. (A) A schematic of full-length FIP200 is shown on top. NT-FIP and CT-FIP segments, and MD-FIP of FIP200 are shown below. (B) 293T cells were transfected with an expression vector encoding HA-FAK and vectors encoding Flag-FIP200, its segments or empty vector control (V), as indicated. Lysates were immunoprecipitated with anti-Flag followed by Western blotting with anti-FAK to detect associated HA-FAK (top panel) or with anti-Flag to verify similar amounts of samples in the immunoprecipitates (bottom panel; Flag-FIP200 and segments are marked by arrowheads). (C) 293T cells were transfected with vectors encoding HA-FAK (WT) or its fragments (N-terminal, kinase domain and C-terminal) or with kinase domain of Pyk2, as indicated. Lysates from the transfected cells were then incubated with immobilized GST fusion proteins containing FIP200 segments or GST alone, as indicated. The bound proteins were resolved on SDS-PAGE and were analyzed by Western blotting with mAb 12CA5 (anti-HA). (D) Equal amounts of immobilized GST-fusion proteins containing FIP200 fragments, or GST alone, were incubated with 1 mug of recombinant FAK. The bound proteins were resolved on SDS-PAGE and were analyzed by Western blotting with anti-FAK (top panel). The membrane was also stained with Ponceau S stain to detect GST-fusion proteins (marked with arrowheads; bottom panel). FIP200 Inhibition of FAK Kinase Activity and Autophosphorylation | The binding of FIP200 to FAK kinase domain raised the possibility that FIP200 may have an effect on FAK kinase activity. To test this directly, we performed FAK in vitro kinase assays using E4Y1 as an exogenous substrate in the presence of different amounts of purified GST fusion protein containing the FIP200 segments or GST alone as a control. Figure A shows that the GST fusion proteins containing the NT-FIP and MD-FIP inhibited FAK kinase activity, whereas GST alone did not have any effect. GST fusion protein containing NT-FIP showed a significantly greater inhibitory effect than GST fusion protein containing CT-FIP. GST fusion protein containing MD-FIP showed an intermediary activity, which was also significantly higher than GST fusion protein containing CT-FIP. In particular, at lower concentrations (e.g., <5 pmol/reaction), GST fusion proteins containing NT-FIP or MD-FIP reduced FAK kinase activity, whereas CT-FIP did not, suggesting that NT-FIP and MD-FIP are more effective than CT-FIP in the inhibition of FAK kinase activity in vitro. These FIP200 segments also inhibited FAK from SYF cells (deficient in Src, Yes, and Fyn expression) to the same extent as FAK from wild-type control cells (S. Abbi and J. Guan, unpublished data), suggesting that FIP200 inhibited the kinase activity of FAK directly, but not through its potential effects on the associated Src family kinases. Figure 3 | Inhibition of FAK activity by FIP200. Inhibition of FAK activity by FIP200. (A) The kinase activity of FAK was assayed in the presence of various amounts of GST fusion proteins containing FIP200 segments, or GST alone, as indicated. Relative kinase activities were normalized to FAK activity in the absence of GST fusion protein. The mean +- SE of relative kinase activities from three independent experiments are shown. The inset shows Coomassie blue staining of a representation preparation of the fusion proteins (1, GST-NT-FIP; 2, GST-MD-FIP; and 3, GST-CT-FIP). (B) 293T cells were cotransfected with plasmid encoding HA-FAK and HA-FIP200, its segments, or vector control, as indicated. One day after transfection, cells were trypsinized and either kept in suspension (S) or replated on FN (10 mug/ml) for 30 min. They were then lysed and immunoprecipitated with anti-HA followed by Western blotting with PY20 to detect FAK phosphorylation (top panel) or anti-HA to verify similar FAK expression levels (middle panel). The corresponding whole cell lysates (WCL) were resolved on a SDS-PAGE gel and were western blotted with anti-HA to detect similar amounts of FIP200 and its fragments (marked by arrowheads, bottom panel). We next examined the effect of FIP200 and its segments on cell adhesion-induced FAK phosphorylation in intact cells. As shown in Figure B (Top and middle panels), expression of FIP200 suppressed tyrosine phosphorylation of FAK after adhesion to FN (Figure B, compare lanes FIP200 and V). Expression of NT-FIP or MD-FIP also inhibited FAK phosphorylation (Figure B, compare lanes NT-FIP and MD-FIP with lane V), whereas CT-FIP did not have any effect (Figure B, compare lanes CT-FIP and V). Similar expression levels of FIP200 fragments were verified by western blotting with anti-HA (Figure B, lower panel). Together, these results indicate that binding of FIP200 to FAK through interactions at different domains could inhibit FAK kinase activity in vitro. However, they also suggest that the quantitative difference of the in vitro inhibitory activity of FIP200 segments could lead to a differential inhibition of FAK activity in intact cells by NT-FIP and MD-FIP, but not by CT-FIP. Effects of FIP200 on FAK Downstream Signaling | Activation and autophosphorylation of FAK have been suggested to lead to tyrosine phosphorylation of several other cellular proteins, including paxillin, p130cas, Grb7, and Shc . Therefore, we examined the effects of FIP200 on the FAK-promoted activation of these downstream targets. Figure shows that overexpression of FAK induced tyrosine phosphorylation of all four potential substrates, paxillin, p130cas, Grb7, and Shc, as observed previously . Interestingly, overexpression of NT-FIP, which had maximum inhibition of FAK activation and phosphorylation among the segments (see Figure ), reduced cell adhesion-dependent paxillin and Shc phosphorylation by FAK (Figure , A and C), but had little effect on p130cas and Grb7 phosphorylation (Figure , B and D). The mechanism of FIP200's selective inhibition of FAK downstream targets is unknown. It is possible that the threshold activity of FAK required for its phosphorylation of these substrates is different. Thus, inhibition of FAK by FIP200 under these experimental conditions could be sufficient to inhibit paxillin and Shc phosphorylation, but not for p130cas and Grb7 phosphorylation. Figure 4 | Effects of FIP200 on FAK downstream signaling. Effects of FIP200 on FAK downstream signaling. 293 cells were cotransfected with plasmids encoding FAK, NT-FIP, or empty vector as controls, along with vectors encoding GFP-paxillin (A), Myc-p130cas (B), HA-Shc (C), or HA-Grb7 (D), as indicated. One day after transfection, cells were trypsinized and replated on FN (10 mug/ml) for 30 min. Whole cell lysates were then immunoprecipitated with anti-GFP, anti-Myc, or anti-HA to pull-down epitope-tagged paxillin, p130cas, Shc, and Grb7, respectively. The immune complexes were then analyzed by Western blotting with PY20 to detect their phosphorylation (top panels), or with antipaxillin, anti-Myc, or anti-HA to show their respective expression levels (bottom panels). The position of HA-Shc is indicated by arrows in C. Effects of FIP200 on FAK-Dependent Cell Spreading and Migration | We next examined the effects of FIP200 on FAK-regulated cellular functions, including cell spreading and migration, and cell cycle progression. To study cell spreading, we transiently transfected NIH3T3 cells with the expression vectors encoding FIP200 or its fragments (see Figure A). The effects on cell spreading on FN were assessed initially by immunofluorescent staining with anti-HA antibody to mark the positively transfected cells, and with anti-vinculin antibodies to mark the background untransfected cells in the same field. Figure A shows that transfection of the cells with full-length FIP200 prevented cell spreading on FN (Figure , top panels), whereas expression of the CT-FIP did not affect cell spreading (Figure , bottom panels). Similar studies showed that expression of NT-FIP or MD-FIP also inhibited cell spreading (S. Abbi and J. Guan, unpublished data). The differences in cell spreading were still apparent 4 h after replating on FN (S. Abbi and J. Guan, unpublished data), although all cells were completely spread after overnight incubation (see Figure A), suggesting that inhibition of cell spreading by FIP200 was transient. We also used cotransfection of an expression vector encoding beta-gal to identify the positively transfected cells in the cell spreading assays. Figure B shows similar results using this method. Expression of FIP-200, NT-FIP, or MD-FIP inhibited cell spreading by similar50% compared with control untransfected cells or cells expressing CT-FIP. The correlation of cell spreading inhibition by NT-FIP and MD-FIP, but not CT-FIP, with their inhibition of FAK activity (see Figure ) suggested that FIP200 might inhibit cell spreading by its inhibition of endogenous FAK functions. Consistent with this possibility, coexpression of FAK with FIP200 rescued inhibition of cell spreading by FIP200 (compare the first and the last lane in Figure B), although overexpression of FAK alone had no effect on cell spreading under these conditions. Western blotting of aliquots of lysates from the transfected cells showed similar expression levels of FIP200 and its fragments and a lack of effects of FAK coexpression on the levels of FIP200 (Figure , C and D). Figure 5 | Inhibition of cell spreading by FIP200. Inhibition of cell spreading by FIP200. (A) NIH3T3 cells were transfected with expression vectors encoding FIP200 or CT-FIP, as indicated. One day after transfection, cells were trypsinized and replated on FN (10 mug/ml) for 45 min. Cells were fixed and processed for immunostaining with anti-HA to detect the positively transfected cells (green) and with antivinculin to visualize all cells (red). (B-D) NIH3T3 cells were transfected with vectors encoding FIP200, its segments, or empty vector control, and plasmid encoding a HA-tagged FAK (in some experiments), along with a plasmid encoding beta-Gal, as indicated. One day after transfection, cells were trypsinized and replated on FN (10 mug/ml) for 45 min. The cells were then fixed, and beta-Gal assays were performed to identify the positively transfected cells. The mean +- SE of percentage of spread cells from three independent experiments are shown (B). Aliquots of whole cell lysates (WCL) were also analyzed directly by Western blotting using anti-Flag to detect FIP200 and its fragments (C) or anti-HA to detect FAK (D). The effect of FIP200 and its fragments on cell migration was assessed by using monolayer-wounding assays after transient transfection of NIH3T3 cells with expression vectors encoding FIP200 or its fragments along with a plasmid encoding GFP. Phase contrast and fluorescence images were captured at regular intervals after wounding to monitor the movement of cells from the wound edge to the center of the wound. The rate of migration was then calculated for transfected cells at the edge of the wound by measuring the distance that the GFP-positive cells moved toward the center of the wound in 8 h. As shown in Figure A, cells transfected with the control vector (Figure V) moved toward the center of the wound at the same rate as the surrounding untransfected cells. In contrast, the FIP200-transfected cells moved much less than the surrounding untransfected cells. Quantification of the rate of migration showed that FIP200, NT-FIP, and MD-FIP inhibited cell migration by similar60 --80%, whereas CT-FIP had no effect (Figure B). Furthermore, coexpression of FAK or paxillin with FIP200 rescued inhibition of cell migration by FIP200 (Figure B). Similar results were also obtained using an alternative cell migration assays employing a time-lapse imaging-based computerized motility analysis method OMAware, as described previously . Although FIP200 inhibited cell migration, coexpression with FAK reversed this inhibition to control levels (Figure C). Taken together, these results demonstrate that FIP200 inhibition of FAK leads to inhibition of FAK-dependent cell spreading and migration and suggest that inhibition of paxillin phosphorylation downstream of FAK might be responsible for these effects. Figure 6 | Inhibition of cell migration by FIP200. Inhibition of cell migration by FIP200. (A and B) NIH3T3 cells grown on FN (10 mug/ml) were transfected with FIP200, its segments, or empty vector control, with vectors encoding FAK or paxillin in some experiments, along with a plasmid encoding GFP in 7:1 ratio, as indicated. One day after transfection, the cell monolayer was wounded with a p10 tip, incubated at 37C, and images were captured at 2-h intervals until 8 h. Images from representative experiments are shown in A. The rate of migration was measured by quantifying the total distance that the positively transfected cells (GFP+) moved from the edge of the wound toward the center of the wound in 8 h. (B) mean +- SE of the rate of migration from three independent experiments. *p = 0.48, 0,76, 0.32, and 0.89 for samples from left to right, in comparison with vector alone transfected cells. (C) Motility of cells on FN (5 mug/ml) were also assayed using OMAware based on time-lapse video microscopy as described in "Materials and Methods." Representative field of cell tracks of control untransfected cells (a), cells transfected with expression vector encoding FIP200 (b), and cells transfected with both vectors encoding FIP200 and FAK (c) are shown. The arrows denote positively transfected cells. The untransfected cells in the same field serve as internal controls. FIP200 Inhibition of Cell Proliferation and Its Rescue by FAK | To explore a potential role for FIP200 in cell cycle progression, we transiently transfected NIH3T3 cells with the expression vectors encoding FIP200 or its fragments (see Figure A), and then measured the extent of BrdU incorporation. Figure A shows that overexpression of FIP200 inhibited cell cycle progression as measured by BrdU incorporation (Figure A, top panels). Expression of a control vector encoding an irrelevant protein did not affect BrdU incorporation under the same conditions (Figure A, bottom panels). Quantitative analysis indicated that FIP200 inhibited cell cycle progression by similar90% compared with cells transfected with the control plasmid or mock-transfected cells (Figure B). Similar analysis showed that NT-FIP and MD-FIP also inhibited BrdU incorporation to a similar extent as the full-length FIP200, whereas CT-FIP did not have any effect. There was no evidence of apoptosis in any of the transfected cells (S. Abbi, H. Ueda, and J. Guan, unpublished data), suggesting that the cell cycle effects are not due to possible role of FIP200 or its fragments in cell survival or apoptosis. Figure 7 | Regulation of cell cycle progression by FIP200. Regulation of cell cycle progression by FIP200. (A and B) NIH3T3 cells were transfected with expression vectors encoding HA-FIP200 or its segments, or an irrelevant control protein (C) or were mock transfected, as indicated. They were then analyzed for BrdU incorporation as described in "Materials and Methods." (A) Representative fields for cells transfected with HA-FIP200 or the control. Immunostaining with anti-HA identifies positively transfected cells (green), and staining with anti-BrdU shows cells with new DNA synthesis (red). (B) The mean +- SE of three independent experiments of the percentage of BrdU+/positively transfected cells as determined by analyzing at least 80 positively transfected cells for each transfection in multiple fields. (C) NIH3T3 cells were cotransfected with an expression vector encoding FAK and that encoding HA-FIP200 or empty vector control (V), as indicated. A plasmid encoding beta-Gal was also included. They were then analyzed for BrdU incorporation as described in "Materials and Methods." The positively transfected cells were identified by immunostaining with anti-beta-Gal. The percentage of BrdU+/beta-Gal+ cells was determined by analyzing 40 --50 beta-Gal+ cells for each transfection in multiple fields. The percentage of BrdU-positive cells was normalized to the vector control of 100%. The results show mean +- SE of three independent experiments. *p = 0.18 and 0.59 are values for empty vector plus FAK transfection and HA-FIP200 plus FAK transfection, respectively, in comparison with value from empty vector alone transfection. Inset shows similar expression levels of FIP200 (alpha-HA blot) with or without cotransfection of FAK (KT3 blot). FAK has been shown to play a role in cell cycle progression , and we have shown that FIP200 can inhibit FAK activity. Therefore, we examined if overexpression of FAK along with FIP200 could rescue this inhibition of cell cycle progression. FAK alone did not promote cell proliferation under these conditions, but it rescued the inhibition of BrdU incorporation by FIP200 to the control levels (Figure C). Western blotting of aliquots of cell lysates showed that coexpression of FAK did not affect the expression levels of FIP200 (Figure C, inset). Together, these data indicate that FIP200 inhibition of FAK also leads to inhibition of FAK-dependent cell cycle progression. Effects of Disruption of Functional Interaction between Endogenous FIP200 and FAK | We also investigated the role of FIP200 as a protein inhibitor for FAK by disrupting the functional interaction of these two proteins. Although FIP200 can associate with FAK through more than one domains (see Figure ), FIP200 binding to FAK kinase domain is responsible for its inhibition of FAK kinase and cellular activities in vivo (see Figures --). Therefore, we designed a FAK construct, designated KDKR, which contains only the kinase domain of FAK (residues 403 --672) with the kinase-defective mutation (K454 to R). Overexpression of KDKR should titrate out the FIP200 functional binding sites for the FAK kinase domain, thus relieving its inhibition of FAK. The kinase-defective mutant was used instead of the wild-type kinase domain construct to minimize potential effects of expressing this domain (as a kinase) other than its competing with endogenous FAK kinase domain binding to FIP200. This mutation did not affect its binding to FIP200 as KDKR bound to FIP200 as efficiently as its wild-type kinase domain counterpart (S. Abbi and J. Guan, unpublished data). We first examined the effects of KDKR on FAK phosphorylation during cell adhesion. Consistent with FIP200 being an inhibitor for FAK, overexpression of KDKR led to an increased tyrosine phosphorylation of FAK in cells plated on PLL in comparison with cells transfected with a control plasmid (Figure A). The specificity of KDKR to affect FIP200/FAK interaction was supported by its lack of an effect on tyrosine phosphorylation of Pyk2 with or without stimulation by sorbitol (Figure B). We then examined the effects of KDKR on FAK-dependent cell cycle progression by measuring BrdU incorporation of cells plated on FN or PLL. Figure C shows that a significant fraction (similar50%) of cells plated on FN progressed to the S-phase of cell cycle under the experimental conditions, whereas only a small portion (similar10%) of cells plated on PLL entered the S-phase. In contrast, overexpression of KDKR led to a partial rescue of the reduced cell cycle progression on PLL (similar30%). These results suggest that disruption of FAK inhibition by FIP200 could lead to an increased FAK phosphorylation as well as a partial restoration of cell cycle progression in the absence of cell adhesion to FN. It is likely that additional signals from FN other than FAK phosphorylation are necessary for a full restoration of cell cycle progression. Nevertheless, these data provide further support for FIP200 as a protein inhibitor for FAK. Figure 8 | Disruption of endogenous FIP200 interaction with FAK. Disruption of endogenous FIP200 interaction with FAK. (A and B) NIH3T3 cells were cotransfected with plasmid encoding HA-tagged KDKR or an irrelevant control protein (Grb7 SH2 domain, designated as C) and plasmids encoding HA-FAK or HA-Pyk2, as indicated. One day after transfection, cells were trypsinized and replated on PLL (0.1 mg/ml) or FN (10 mug/ml; A) or serum starved and treated with or without sorbitol (400 mM, 10 min; B), as indicated. Cell lysates were then immunoprecipitated with anti-HA or anti-Pyk2 and were western blotted with PY20 (top panel) or anti-HA (middle panel). Whole cell lysates (WCL) were also blotted with anti-HA (bottom panel) to show levels of transfected KDKR and control protein (C). (C) NIH3T3 cells were transfected with plasmid encoding HA-tagged KDKR or an irrelevant control protein, or mock transfected, as indicated. They were then analyzed for BrdU incorporation on both FN (10 mug/ml) and PLL (0.1 mg/ml) as described in "Materials and Methods." The percentage of BrdU+/positively transfected cells was calculated and normalized to that of untransfected cells in each experiment. The mean +- SE are shown for data from three independent experiments. DISCUSSION : FIP200 is a novel cellular protein that was recently found to interact with the FAK-related kinase, Pyk2, using the yeast two-hybrid screen . Furthermore, FIP200 could inhibit the kinase and cellular activities of Pyk2 by binding to its kinase domain directly. In contrast to the restricted expression pattern of Pyk2 , FIP200 is widely expressed in many tissues and cell lines , suggesting that it may play important functions in some fundamental cellular processes involving FAK. In this report, we provide evidence demonstrating that FIP200 is a novel protein inhibitor for FAK. FAK has been well documented to play an important role in signal transduction by integrins. Recent studies have identified multiple signaling molecules that interact with FAK and mediate its downstream pathways in the regulation of cellular functions . To our knowledge, however, FIP200 is the first reported protein inhibitor for FAK that functions by directly bind to its kinase domain thus inhibiting its kinase and cellular activities. Together with our previous report , these results suggest that FIP200 could function as a protein inhibitor for both members of FAK family tyrosine kinases, FAK and Pyk2. Furthermore, inhibition of FAK and Pyk2 by FIP200 may be mediated by similar mechanisms because both involve binding of FIP200 to the catalytic domains of the kinases. In contrast, we could not detect binding of FIP200 to another tyrosine kinase Src (S. Abbi, H. Ueda, and J. Guan, unpublished data), suggesting specificity of FIP200 toward FAK family kinases. Both FAK and Pyk2 have been shown to associate with Src family kinases upon their activation . However, we showed previously that FIP200 inhibited the kinase activity of Pyk2 from SYF cells (deficient in Src, Yes, and Fyn expression) to the same extent as Pyk2 from wild-type control cells . Similar results were obtained for FAK isolated from SYF or control cells (S. Abbi and J. Guan, unpublished data). These results supported that FIP200 inhibited the kinase activity of FAK and Pyk2 directly, with little effect on the associated Src family kinases. Despite these similarities of FIP200 inhibition of FAK and Pyk2, different segments of FIP200 are involved in its interaction with the catalytic domains of these two kinases. Although CT-FIP, but not NT-FIP or MD-FIP, bound to Pyk2 kinase domain (; see Figure C), we found here that both NT-FIP and MD-FIP associated with FAK kinase domain and inhibited its kinase and cellular activities. Furthermore, CT-FIP bound to the N-terminal domain of FAK, reduced FAK kinase activity in vitro only when used at high concentrations (but no effect at lower concentrations), and did not inhibit FAK functions in vivo. These results suggest that different residues of FIP200 are involved in its binding to FAK and Pyk2 despite their homologous kinase domains. Indeed, in spite of the homology between FAK and Pyk2, they are activated by different signals within the cell and play different functional roles in vivo. In addition to common binding partners, there are also proteins that bind to one but not the other. For example, NIRS (mammalian homolog of Drosophila retinal degradation B; ) binds to Pyk2 but not FAK. Similarly, talin can differentiate between the C-terminal domains of FAK and Pyk2, which are 39% identical, and binds only to FAK . Recently, a novel protein, PSGAP, has been discovered that can bind both Pyk2 and FAK in vitro, but plays a role only in Pyk2-mediated signaling . Thus, it seems that these proteins are able to utilize the subtle differences in Pyk2 and FAK to mediate differential interactions and functions. The differential binding of FIP200 to FAK and Pyk2 also raised the interesting possibility that FIP200 may interact with both kinases and coordinate their signaling functions under certain conditions. Experiments are in progress to define the sequence motifs involved in FIP200 interaction with FAK as well as Pyk2 that should provide further insights into the molecular mechanisms and possible relationship of the interactions. Integrin signaling through FAK has been shown to regulate a variety of cellular functions, including cell spreading, migration, and cell cycle progression . Consistent with its being a protein inhibitor for FAK, overexpression of FIP200 or its fragments in fibroblasts inhibited these cellular functions. Several lines of evidence suggest that FIP200 affects these cellular functions through its inhibition of FAK, although we cannot completely exclude the possible involvement of other mechanisms. First, inhibition of cell spreading, migration, and cell cycle progression by FIP200 was completely rescued by coexpression of FAK. Second, inhibition of these cellular activities by FIP200 segments correlated with their abilities to bind FAK kinase domain and inhibit its biochemical activities (as measured by autophosphorylation) in vivo (Figure B). It is also interesting to note that expression of either the full-length FIP200 (Figure E, bottom panels) or its fragments (S. Abbi and J. Guan, unpublished data) did not affect FAK localization in focal contacts, suggesting that FIP200 did not inhibit these cellular activities by altering FAK localization. Third, expression of NT-FIP reduced tyrosine phosphorylation of several FAK downstream targets, including paxillin and Shc , which have been shown to play a role in these cellular functions. Fourth, FIP200 did not inhibit migration of FAK-/- cells on FN (S. Abbi, H. Ueda, and J. Guan, unpublished data), suggesting that its effect on cellular functions is specifically through its interaction with FAK. Also, disruption of functional interaction of endogenous FIP200/FAK with a FAK kinase domain (with kinase-defective mutation) construct increased FAK phosphorylation and partially restored cell cycle progression for cells plated on PLL . The specificity of this construct is supported by a lack of effect on stimulation of Pyk2 activation (Figure B). It is also supported by the fact that no other proteins are known to interact with this region of FAK (residues 403 --672, which exclude FAK motifs such as Y397 or P712/715), thus potentially be affected nonspecifically. It is a kinase-dead version, therefore the potential nonspecific effect is minimized here also. Finally, it only affected FAK phosphorylation and BrdU incorporation for cells plated on PLL (when there is FIP200 association with FAK; see Figure ) but not for cells plated on FN (when there is minimal FIP200/FAK complex; see Figure ). If it enhanced BrdU incorporation by other mechanisms, one would expect it to have an effect under both conditions (e.g., cotransfection of v-Src would lead to enhanced FAK phosphorylation for cells on PLL and FN; ). Taken together, these data indicate that FIP200 also functions as an inhibitor for FAK in appropriate cellular contexts. Based on these results, we propose the following working hypothesis for the role of FIP200 interaction with FAK in integrin-mediated cell adhesion and signaling . In untransfected control cells (Figure , A, left), some FIP200 and FAK is complexed under suspended conditions. On cell adhesion and integrin binding to ligands, FIP200 is dissociated from FAK. This release from a negative inhibitor may contribute to FAK activation and phosphorylation in cell adhesion, which trigger downstream signaling pathways in various cellular functions such as cell migration and proliferation. Overexpression of FIP200 in these cells (Figure , B, Right) drives the equilibrium toward more association of FIP200 with FAK (even in adherent cells), thus leading to inhibition of FAK signaling and function. These hypotheses are consistent with results in Figure and other observations . Although this model implies some role for FIP200 in the regulation of FAK activation by integrins, it is important to note that other factors are also likely to be critical in the activation of FAK by integrins or other receptors. Figure 9 | A working hypothesis of FIP200 interaction with FAK during cell adhesion. A working hypothesis of FIP200 interaction with FAK during cell adhesion. One potential concern for our proposed role of FIP200 as a protein inhibitor for FAK is that the data are largely based on the overexpression of FIP200 or its fragments. It is possible that proteins of components of positive active complexes might act as dominant inhibitors when overexpressed (e.g., overexpression of the p85 subunit inhibits the PI3K function of the p85/p110 complex). In this study, however, the overexpression studies are supported by data from other and complementary approaches. These include the association and regulation of endogenous proteins , in vitro studies using purified proteins (Figures and A), and expression of an FAK segment that disrupts the functional interaction of FIP200 with FAK . Given the consistent results from these other approaches, it is very unlikely that the endogenous FIP200 functions as a part of positive FAK complex. It was interesting that FIP200 inhibited FAK-mediated activation of paxillin and Shc, whereas it had no effect on p130cas and Grb7 phosphorylation. It is possible that there is difference in the threshold activity of FAK required to activate its various substrate, and although the inhibition of FAK activity by FIP200 was sufficient to block its activation of paxillin and Shc, it did not effect the activation of other downstream targets. It is also possible that there are separate complexes of FAK with its various substrates, and their interaction with FIP200 is differentially regulated within the cell. In any case, these data suggest that inhibition of FAK-mediated tyrosine phosphorylation of paxillin and/or Shc by FIP200 is at least partially responsible for the inhibition of various cellular activities by FIP200. Interestingly, inhibition of cell spreading by FRNK correlated with a decreased tyrosine phosphorylation of paxillin . Furthermore, it was reported recently that tyrosine phosphorylation of paxillin and its association with Crk stimulated migration of a tumor cell line NBT-II on collagen . Also, the phosphatase PP2A that dephosphorylates paxillin negatively regulates cell cycle progression and cell motility . Consistent with a role for paxillin in cell motility, we also observed that overexpression of paxillin rescued FIP200 inhibition of cell migration (Figure B). Further studies will be necessary to clarify the roles of various FAK-downstream targets in the regulation of cellular activities by FIP200. Previous studies have shown a number of protein tyrosine phosphatases that inhibit FAK signaling by dephosphorylation of FAK (; ; ; ; ; ;). However, all these inhibitory events required the enzymatic activities of the phosphatases. In contrast, FIP200 inhibited FAK by binding to its kinase domain, which offers the potential opportunity to derive small peptide inhibitors for FAK. It is interesting that two FIP200 segments (NT-FIP and MD-FIP) could both inhibit FAK by apparently similar mechanisms. There are several regions of high homology (similar30% identity) between NT-FIP and MD-FIP. Future studies will be necessary to determine whether these common regions play a role in FIP200 interaction with FAK. The possible generation of small peptides or their derivatives as inhibitors for FAK is also an exciting future avenue of research, especially because activation of FAK has been implicated in diseases such as cancer metastasis . Backmatter: Abbreviations used: : BrdU = bromodeoxyuridine CT-FIP = C-terminal FIP FAK = focal adhesion kinase FIP200 = FAK-family interacting protein of 200 kDa FN = XXXX GFP = green fluorescent protein GST = glutathione S-transferase HA = hemagglutinin Ig = immunoglobulin mAb = monoclonal antibody MD-FIP = middle domain FIP NT-FIP = N-terminal FIP PCR = polymerase chain reaction PLL = poly-l-lysine SH2 = Src homology 2 PMID- 12221125 TI - The Product of the Survival of Motor Neuron (SMN) Gene is a Human Telomerase-associated Protein AB - Telomerase is a ribonucleoprotein (RNP) complex that is minimally composed of a protein catalytic subunit, the telomerase reverse transcriptase (TERT), and an RNA component, the telomerase RNA. The survival of motor neuron (SMN) gene codes for a protein involved in the biogenesis of certain RNPs. Here, we report that SMN is a telomerase-associated protein. Using in vitro binding assays and immunoprecipitation experiments, we demonstrate an association between SMN and the telomerase RNP in vitro and in human cells. The specific immunopurification of SMN from human 293 cells copurified telomerase activity, suggesting that SMN associates with a subset of the functional telomerase holoenzyme. Our results also indicate that the human telomerase RNA and the human (h) TERT are not associated with Sm proteins, in contrast to Saccharomyces cerevisiae telomerase. Immunofluorescence analysis showed that hTERT does not specifically colocalize with wild-type SMN in gems or Cajal bodies. However, a dominant-negative mutant of SMN (SMNDeltaN27) previously characterized to elicit the cellular reorganization of small nuclear RNPs caused the accumulation of hTERT in specific SMNDeltaN27-induced cellular bodies. Furthermore, coexpression of SMNDeltaN27 and hTERT in rabbit reticulocyte lysates decreased the efficiency of human telomerase reconstitution in vitro. Our results establish SMN as a novel telomerase-associated protein that is likely to function in human telomerase biogenesis. Keywords: INTRODUCTION : Telomere maintenance in most eukaryotic cells is established by the ribonucleoprotein (RNP) enzyme telomerase. The telomerase RNP minimally consists of an RNA molecule and a protein catalytic subunit, the telomerase reverse transcriptase (TERT). Using an internal template sequence in the telomerase RNA subunit, this specialized reverse transcriptase synthesizes simple guanine-rich sequences at the 3'-end of chromosomal DNA. The telomeric DNA repeats and associated telomere-binding proteins protect chromosomes from nuclease digestion, end-to-end fusions, and other DNA rearrangement events . The lengths and nucleotide sequences of the telomerase RNA subunits are highly divergent . Secondary structure prediction suggests the presence of a small nucleolar (sno) RNA H/ACA box motif in the 3'-end of vertebrate telomerase RNAs . Mutations that perturb the folding of the H/ACA box of human and mouse telomerase RNAs prevent both their cellular accumulation and their ability to properly localize to the nucleolus of microinjected Xenopus oocytes . Notably, the human telomerase RNA (hTR) associates with the H/ACA box snoRNA-binding proteins dyskerin, hGAR1, NH2P, and NOP10 . The role of these proteins in vertebrate telomerase function is unclear. The X-linked form of the disease dyskeratosis congenita is caused by mutations in the gene that encodes the protein dyskerin . Cells from individuals affected by this disease have low levels of hTR and telomerase activity as well as short telomeres , supporting an important role for snoRNP proteins in telomerase biogenesis. Human telomerase activity is detected in >85% of cancers and transformed cell lines, whereas it is absent from most normal human cells . Inhibition of human telomerase from immortal and cancer cell lines results in telomere shortening and, in certain cell types, cell death or senescence . Consequently, a better understanding of the mechanisms involved in the assembly and regulation of the human telomerase RNP will be important for the rational design of telomerase inhibitors. In Saccharomyces cerevisiae, the telomerase RNA (TLC1) associates with the heptameric Sm protein complex and acquires a 5'-2,2,7-trimethylguanosine (TMG) cap structure ; both of these events are hallmarks of small nuclear (sn) RNP assembly. Yet, little is known about the molecular machinery involved in the localization and assembly of vertebrate telomerase. The product encoded by the survival of motor neuron (SMN) gene is present both in the cytoplasm and in the nucleus, where it localizes in different nuclear structures: gems, Cajal bodies (CBs), and the nucleolus . SMN and a set of associated proteins (Gemins) form a complex involved in the biogenesis of at least four uridine (U)-rich snRNPs, U1, U2, U4, and U5, all major constituents of the splicing machinery . Biochemical data demonstrate that SMN interacts directly with the arginine-glycine --rich domain of a subset of Sm proteins . The Sm core complex contains seven proteins (B/B', D1 --D3, E, F, and G) predicted to form a closed ring structure around a conserved sequence motif within some of the UsnRNAs . The precise function of the SMN complex is not completely characterized; however, evidence strongly suggests that the SMN complex facilitates or stabilizes the association of Sm proteins with U1, U2, U4, and U5 snRNAs and the functional maturation of UsnRNPs . Furthermore, using antibody addition experiments and a dominant-negative version of SMN that lacks the first 27 amino acids (SMNDeltaN27), demonstrated that SMN might also have a more direct role in pre-mRNA splicing. As expected from the wide range of cellular pathways in which SMN is implicated , disruption of the gene encoding SMN in different organisms is lethal . The role of the Sm protein complex in S. cerevisiae telomerase biogenesis and the recent observation that SMN associates with snoRNP proteins , prompted us to examine whether SMN and/or Sm proteins are involved in human telomerase function. We report that SMN is a novel telomerase-associated protein. The human Sm protein complex does not interact with human (h) TERT, hTR, or catalytically active telomerase, suggesting that the association of SMN with human telomerase is independent of Sm proteins. A previously characterized dominant-negative SMN protein (SMNDeltaN27) has the ability to perturb the normal subcellular localization of hTERT and decrease the efficiency of in vitro reconstitution of telomerase in rabbit reticulocyte lysates (RRLs). On the basis of these and other recent results , we suggest that SMN is involved in human telomerase biogenesis as an H/ACA snoRNP. MATERIALS AND METHODS : Experimental Procedures | Construction of Plasmids. The cDNAs encoding human SMN and SMNDeltaN27 were amplified by RT-PCR using total cellular RNA extracted from HeLa cells. SmB, SmD1, and SmD3 cDNAs were amplified by PCR from IMAGE clones. Expression of the Myc-tagged version of these proteins in cultured human cells or rabbit reticulocyte lysates was performed by cloning the DNA fragments corresponding to the above-mentioned cDNAs into a modified pcDNA3.1 vector (InVitrogen, San Diego, CA) containing the sequence for the Myc-tag epitope . The expression construct for FLAG-hTERT was a gift from Dr. Lea Harrington (Amgen, University of Toronto). For subcellular localization experiments, hTERT cDNA was subcloned into the pEGFP-C1 vector (Clonetech, Cambridge, UK). Antibodies. The antibodies used were as follows: mouse monoclonal FLAG antibody (Sigma, St. Louis, MO); affinity-purified goat anti-GST serum (Amersham Biosciences, Arlington Heights, IL); affinity-purified rabbit anti-hTERT serum ; mouse anti-Sm ; affinity-purified rabbit anti-TEP1 serum ; mouse monoclonal anti-SMN (clone 8; Transduction Laboratories, Lexington, KY); mouse monoclonal anti-Myc (9E10; ATCC hybridoma; American Type Culture Collection, Manassas, VA); and mouse monoclonal 2,2,7-TMG-specific antibody (Oncogene Research Products, San Diego, CA). Cell Culture and Manipulations. Human embryonic kidney cells (293) and HeLa cells were grown in DMEM with 10% fetal bovine serum and antibiotics. Transient transfections of 293 and HeLa cells were performed using Lipofectamine 2000 (Invitrogen) with 1 --2 mug expression constructs combined per 35-mm dish. In Vitro Binding Assays. Reconstitution of active human telomerase by coexpression of GST-hTERT and the hTR in yeast was described previously . To investigate SMN and telomerase interaction in vitro, we generated [35S]methionine-labeled Myc-SMN and luciferase using an RRL kit as described per the manufacturer's instructions (Promega, Madison, WI). Equal amounts of labeled proteins were first precleared overnight in 1 ml of in vitro binding buffer (50 mM Tris, pH 7.5, 200 mM NaCl, 2 mM EDTA, 0.1% NP-40, and protease inhibitors). Immunopurified human telomerase RNP was prepared by incubating protein extracts from 50 ml yeast pellets with GST-specific antibody (Amersham Biosciences) and protein-A --Sepharose (Sigma) in yeast lysis buffer (10 mM Tris, pH 7.5, 2 mM MgCl2, 5.0 mM beta-mercaptoethanol, 20% glycerol, 1% NP-40, 0.25 mM sodium deoxycholate, 1.0 mM EGTA, and 150 mM NaCl, plus protease and RNase inhibitors). After a 1-hour incubation at 4C, beads were washed 5 times in yeast lysis buffer supplemented with 500 mM NaCl. The precleared RRL-synthesized labeled proteins were then incubated with the immunopurified human telomerase RNP for an additional 2 h at 4C. After washing 5 times with 1 ml of in vitro binding buffer, bound proteins were eluted by boiling in SDS-PAGE loading dye and subjected to electrophoresis. Immunoprecipitation, Telomerase Activity, and Northern Blotting. Twenty to 24 h after transfection, cells were washed two times with PBS and resuspended in 500 mul of lysis buffer (20 mM HEPES, pH 7.9, 2 mM MgCl2, 0.2 mM EGTA, 10% glycerol, 1 mM DTT, 150 mM NaCl, and 1.0% NP-40, plus protease and RNase inhibitors). After homogenization by forcing the cells five times through a 25-gauge needle, the cell suspension was left rotating at 4C for 30 min before the lysate was cleared in a microcentrifuge for 15 --20 min. For protein coimmunoprecipitation experiments, cell lysates were incubated with antibodies for 30 min before the addition of protein-A --Sepharose for an additional 1-hour incubation at 4C. After the beads had been washed four times with 1 ml of lysis buffer, the bound proteins were eluted and subjected to SDS-PAGE and immunoblotting. Five to 10% of immunoprecipitates were assayed for telomerase activity by Telomeric repeat amplification protocol (TRAP) as previously described . The preparation of HeLa nucleolar-enriched nuclear extracts was based on a previously described protocol . To analyze immunoprecipitated RNAs, HeLa total cell lysates were prepared from 70 --80% confluent cells in a 10-cm dish in lysis buffer (20 mM HEPES, pH 7.9, 300 mM KCl, 10% glycerol, 0.5 mM DTT, 1 mM EDTA, 2 mM MgCl2, and 1% NP-40, plus protease and RNase inhibitors) and subjected to immunoprecipitation as previously described . Probes used for Northern blotting were DNA oligonucleotides complementary to human U1, U2, and U6 snRNAs and hTR. Indirect Immunofluorescence. HeLa cells were cultured on coverslips in 6-well dishes. Twenty hours after transfection, HeLa cells were fixed for 5 min in 1.0% paraformaldehyde in PBS, pH 7.5, and then permeabilized for 5 min in 0.5% Triton X-100 in PBS. Myc-tagged SMN and SMNDeltaN27 proteins were labeled with anti-Myc antibody (9E10; 1:400). Cells were then washed with 0.1% Triton X-100 in PBS, followed by PBS, and were then incubated with secondary antibody (goat anti-mouse Cy3 from Chemicon, Temecula, CA) for 30 min. Cells were rinsed with 0.1% Triton X-100 in PBS or in PBS alone and then mounted in 1 mg/ml para-phenylenediamine in PBS/90% glycerol that also contained DAPI at 1 mug/ml. Digital imaging was performed with a SPOT cooled CCD camera (Diagnostic Instruments, Inc., Burroughs, MI) mounted on a Zeiss Axioplan immunofluorescence microscope. RESULTS : SMN Associates with the Human Telomerase RNP In Vitro and In Vivo | We used functional recombinant human telomerase expressed in S. cerevisiae to investigate whether the SMN protein can form a complex with telomerase in vitro. First, the GST-hTERT/hTR telomerase complex was immunopurified from yeast extracts using an affinity-purified GST-specific antibody as described in Experimental Procedures. [35S]methionine-labeled luciferase and SMN proteins were generated in RRLs. The labeled proteins were incubated with the recombinant telomerase RNP previously immobilized on antibody-coated Sepharose beads. After a 2-h incubation, the complexes were washed extensively, eluted, and analyzed by SDS-PAGE. Figure A shows that SMN specifically bound to the GST-hTERT/hTR complex (lane 6), whereas SMN did not bind to GST alone (lane 5). Treatment of the GST-hTERT/hTR complex with a cocktail of RNases before addition of [35S]-labeled SMN did not affect or disrupt the SMN-telomerase interaction (data not shown). Although incomplete RNA digestion cannot be ruled out, the results suggest that the association of in vitro --synthesized SMN with recombinant hTERT is mediated via direct contact with hTERT or a yeast TERT-associated protein. Figure 1 | SMN associates with hTERT in vitro and in vivo. SMN associates with hTERT in vitro and in vivo. (A) Human telomerase was reconstituted by coexpression of GST-hTERT and hTR in S. cerevisiae as previously described . GST (lanes 2 and 5) and GST-hTERT/hTR (lanes 3 and 6) were affinity-purified from equal volumes of yeast extracts and incubated with in vitro --translated [35S]methionine-labeled luciferase (lanes 1 --3) and SMN (lanes 4 --6). After extensive washing, bound proteins were analyzed by SDS-PAGE and autoradiography. The input lanes (1 and 4) show 5% of the RRL lysate used in the binding reaction. Molecular mass markers are indicated on the left (in kilodaltons, kDa). (B) 293 cells were transiently transfected with a DNA construct expressing FLAG-tagged hTERT. At 20 h after transfection, a total cell lysate was prepared and subjected to immunoprecipitation (IP) without antibody or using anti-GST, anti-FLAG, anti-Sm (Y12), or anti-TEP1. Immunoprecipitates were analyzed by SDS-PAGE and Western blotting for endogenous SMN. The lysate lane corresponds to 5% of the total cell lysate used for the immunoprecipitation. (C) Nucleolar-enriched nuclear extracts were prepared from HeLa cells and subjected to immunoprecipitation (IP) using anti-GST, anti-Sm (Y12), and two different affinity-purified hTERT antibodies. Immunoprecipitates were analyzed by SDS-PAGE and Western blotting for endogenous hTERT (top) and SMN (bottom). Five percent of the cytosolic (C) and nucleolar-enriched nuclear (N) extracts were also loaded. We also used transient expression of a FLAG-tagged hTERT protein in telomerase-positive 293 cells to demonstrate the association between SMN and telomerase. Total cell extracts from 293 cells transiently transfected with a FLAG-hTERT construct were subjected to immunoprecipitation using different antibodies and analyzed by immunoblotting with a mouse monoclonal anti-SMN antibody. As previously demonstrated , an antibody specific to the Sm protein complex efficiently coimmunoprecipitated SMN from 293 cellular extracts (Figure B, lane 5). The precipitation of FLAG-tagged hTERT using anti-FLAG also coimmunoprecipitated SMN from 293 cell extracts (Figure B, lane 4). An antibody specific to a telomerase-associated protein (TEP1) , GST antibody, or protein-A --Sepharose alone did not precipitate the SMN protein (Figure B, lanes 6, 3, and 2, respectively). Nucleolar-enriched nuclear extracts were prepared from HeLa cells to determine whether an endogenous SMN-hTERT complex exists in cells. The SMN protein was found in both the cytosolic and nuclear extracts (Figure C, lanes 1 and 2), as expected from its previously determined subcellular localization . Confirming the efficiency of our nuclear extract preparation, the hTERT protein was primarily nuclear (Figure C, lanes 1 and 2), in agreement with the previous immunofluorescence analysis of hTERT and mTERT . Proteins from the nuclear extracts were subjected to immunoprecipitation using the Sm protein --specific antibody (Y12), an affinity-purified hTERT antibody , and a GST-specific antibody as a negative control. After extensive washing of the antibody-coated beads, the immunoprecipitated proteins were analyzed for recovery of SMN and hTERT as determined by Western blotting with antibodies specific for the respective proteins. As demonstrated in Figure B, immunoprecipitation performed with the Y12 antibody coprecipitated SMN from HeLa cell nuclear extracts, but not the hTERT protein (Figure C, lane 4). hTERT-specific immunoprecipitation also recovered SMN from the HeLa nuclear extracts (lane 5). The coimmunopurification of SMN and hTERT was also confirmed using a different affinity-purified hTERT antibody (Figure C, lane 6). As a control, neither SMN nor hTERT was present in anti-GST immunoprecipitates (lane 3). We conclude that hTERT and SMN can associate in vitro and in human cells. SMN Is Associated with Catalytically Active Human Telomerase | We used transient transfection experiments to investigate whether SMN is associated with active human telomerase. A Myc-tagged version of SMN and FLAG-hTERT were transiently expressed in human 293 cells, and total cell lysates were prepared for immunoprecipitations using either Myc or FLAG antibodies. As a control, a lysate from mock-transfected cells was used. TRAP assays demonstrated that equal levels of telomerase activity were present in the different total cell lysates before immunopurification (Figure , lanes 1 --3). As previously demonstrated , FLAG antibody resin precipitated the FLAG-hTERT protein (data not shown) and human telomerase activity from the FLAG-hTERT --containing extracts (Figure , lane 8). However, no telomerase activity was present on the FLAG-antibody resin incubated with Myc-SMN --containing extracts (lane 7). Telomerase activity (lane 5) was also recovered from Myc-antibody resin prepared from lysates containing the Myc-SMN protein, but not from anti-Myc immunoprecipitates prepared from lysates of mock-transfected and FLAG-hTERT --transfected cells (lanes 4 and 6, respectively). Similar results were observed using HeLa cells (data not shown). These results indicate that SMN associates with a fully assembled and catalytically active telomerase RNP. Figure 2 | SMN associates with catalytically active telomerase. SMN associates with catalytically active telomerase. 293 cells were transiently transfected with DNA constructs expressing Myc-tagged SMN (lanes 2, 5, and 7), FLAG-tagged hTERT (lanes 3, 6, and 8), or with vector alone (mock; lanes 1 and 4). Total cell lysates were prepared and subjected to immunoprecipitation using either anti-Myc (lanes 4 --6) or anti-FLAG (lanes 7 and 8). Immunoprecipitates were analyzed for telomerase activity by the TRAP assay. Of the total cell lysates (lanes 1 --3), 0.5% were also assayed for telomerase activity. The Sm Protein Complex Is Not Associated with Active Human Telomerase and hTR | The Sm proteins form an RNA-binding complex that interacts with a specific region of UsnRNAs . Direct interactions between SMN and specific members of the Sm proteins are thought to recruit the SMN complex to the Sm-snRNA complex . Interestingly, the S. cerevisiae telomerase RNA subunit, TLC1, contains an Sm protein --binding site, as determined by the coimmunoprecipitation of the TLC1 RNA and yeast telomerase activity with epitope-tagged versions of the SmD1 and SmD3 proteins . Myc-tagged versions of the human SmB, SmD1, and SmD3 proteins were used to examine whether the association of SMN with telomerase could be mediated by the involvement of the Sm protein complex in human telomerase biogenesis. Addition of the Myc epitope to the N-terminus of the SmB, D1, and D3 proteins did not affect their function, as determined by their ability to coimmunoprecipitate SMN and by immunofluorescence analysis (data not shown). Constructs expressing the Myc-SmB, Myc-SmD1, and Myc-SmD3 proteins were transfected into human 293 cells, and total cell extracts were prepared for immunopurification using the Myc antibody. As was previously shown , Myc-antibody resin incubated with Myc-SMN --containing extracts recovers telomerase activity (Figure , lane 9). However, immunoprecipitations performed using the Myc antibody and prepared from either the Myc-SmB --, Myc-SmD1 --, or Myc-SmD3 --containing extracts did not recover levels of human telomerase activity (Figure , lanes 10 --12, respectively) significantly higher than background (lane 8). Western blot analysis of the immunoprecipitated proteins (Figure A, bottom) revealed that the myc-tagged Sm proteins were expressed and immunoprecipitated to considerably higher levels than myc-tagged SMN, yet only myc-SMN copurified human telomerase activity. As a control, anti-FLAG antibody efficiently precipitated telomerase activity from lysates prepared from FLAG-hTERT --transfected cells (lane 13). These results, in addition to the absence of hTERT in Y12 immunoprecipitates (Figure C), indicate that the Sm protein complex is not associated with the human telomerase RNP and suggest that Sm proteins do not mediate the SMN-telomerase association. Figure 3 | The hTR and telomerase activity are not associated with the Sm protein complex. The hTR and telomerase activity are not associated with the Sm protein complex. (A) 293 cells were transiently transfected with DNA constructs expressing Myc-SMN (lanes 3 and 9), Myc-SmB (lanes 4 and 10), Myc-SmD1 (lanes 5 and 11), Myc-SmD3 (lanes 6 and 12), FLAG-hTERT (lanes 7 and 13), or with vector alone (mock; lanes 2 and 8). At 20 h after transfection, total cell lysates were prepared and subjected to immunoprecipitation using either anti-Myc (lanes 8 --12) or anti-FLAG (lane 13). Immunoprecipitates were analyzed for telomerase activity by the TRAP assay (top) and for protein content by Western blotting (WB) using anti-myc (bottom). Of the total cell lysates (lanes 2 --7) or lysis buffer (lane 1), 1% were analyzed for telomerase activity. (B) A total cell lysate from HeLa cells was prepared and subjected to immunoprecipitation using anti-Sm (Y12-lane 2), affinity-purified anti-hTERT (lane 3), anti-TMG (lane 4), or anti-GST (lane 5). Immunoprecipitates were analyzed by denaturing PAGE and Northern blotting for endogenous hTR (top), U1 snRNA (middle), and U6 snRNA (bottom). The input lane (lane 1) was loaded with 2.5% of the total RNA extracted from the HeLa total cell lysate. For hTR, exposure time was three times longer than for U1 and U6. To further investigate whether the Sm protein complex is involved in human telomerase biogenesis, we determined whether endogenous hTR from HeLa total cell extracts copurifies with the Sm complex using the Y12 antibody. As previously demonstrated , endogenous U1 and U6 snRNAs are coprecipitated with Sm proteins using the Y12 antibody (Figure B, lane 1). However, the Y12 antibody did not coimmunoprecipitate hTR (Figure B, lane 2). hTERT-associated hTR was coprecipitated using an affinity-purified hTERT antibody (Figure B, lane 3), whereas an anti-GST immunoprecipitation did not recover U1, U6, or hTR (Figure B, lane 5). We also determined whether a subpopulation of the hTR contains a TMG cap structure as was previously shown for the yeast telomerase RNA . This type of hypermethylated 5'-cap structure is a well-known characteristic of some spliceosomal snRNAs, such as U1, U2, U4, and U5 . HeLa total cell lysates were subjected to immunoprecipitation using a TMG-specific monoclonal antibody, and the copurified RNAs were analyzed by Northern blotting. Both U1 and U2 snRNAs were recovered from the anti-TMG immunoprecipitate (Figure B, lane 4 and data not shown), whereas the presence of hTR in TMG-specific immunoprecipitates was undetectable (Figure B, lane 4). Similarly, U6 snRNA was not coimmunoprecipitated by the TMG-specific antibody (lane 4) and was used as a negative control because it does not have a 5'-TMG cap structure . In conclusion, these results indicate that the Sm proteins, whether endogenous or transiently overexpressed, associate neither with hTR nor with the catalytically active human telomerase RNP. Our results further suggest that hTR does not acquire a 5'-TMG cap. Expression of a Dominant-Negative Version of SMN (SMNDeltaN27) Perturbs the Subcellular Localization of hTERT | The SMN protein is present in the cytoplasm as well as in the nucleus of cells, where it is known to concentrate in nuclear structures such as gems, CBs, and nucleoli . A previous study characterizing a dominant-negative mutant of SMN lacking its first N-terminal 27 amino acids (SMNDeltaN27) revealed that this mutant protein causes the reorganization of snRNPs in the nucleus and in the cytoplasm and also negatively affects pre-mRNA splicing in vitro . This SMN mutant protein was also used to demonstrate the functional interaction between SMN and the RNA polymerase II complex as well as with the snoRNP proteins fibrillarin and hGAR1 . We generated SMNDeltaN27 to investigate the effect of this dominant-negative mutant SMN on hTERT cellular localization. Indirect immunofluorescence with the Myc antibody was used to detect the Myc-SMN and Myc-SMNDeltaN27 proteins in transfected HeLa cells. The Myc-SMN protein localized in the cytoplasm and in the nucleus, where it accumulated in gems and CBs (Figure , A and B, c). When the Myc-SMN construct was cotransfected with a plasmid expressing a yellow fluorescent protein (YFP)-SmB fusion, Myc-SMN and YFP-SmB colocalized in gems and CBs (Figure A, a --d), as previously reported . In cells transfected with Myc-SMNDeltaN27, the mutant protein accumulated in large cytoplasmic bodies that partially redistributed the YFP-SmB fusion (Figure A, e --h). In contrast, the YFP-SmB fusion protein was barely detectable in the cytoplasm of Myc-SMN transfected cells (Figure A, b) and untransfected cells (data not shown), consistent with the localization of endogenous Sm proteins as detected using the Sm-specific Y12 antibody . Similar results were obtained when a YFP-SmD1 fusion was used (data not shown), in agreement with the observations that a mutant of SMN lacking the first 27 amino acids causes a reorganization of Sm snRNPs proteins . Figure 4 | The expression of a dominant-negative mutant of SMN (SMNDeltaN27) perturbs the nuclear localization of hTERT. The expression of a dominant-negative mutant of SMN (SMNDeltaN27) perturbs the nuclear localization of hTERT. (A) HeLa cells were transiently cotransfected with DNA constructs expressing either Myc-SMN and YFP-SmB (a --d) or Myc-SMNDeltaN27 and YFP-SmB (e --h). The fixed and permeabilized cells were stained for Myc-SMN (c) and Myc-SMNDeltaN27 (g) using anti-Myc. DNA stained with DAPI shows the nucleus of each cell (a and e). Images b and c, and f and g, are merged to form d and h, respectively. (B) HeLa cells were transiently cotransfected with DNA constructs expressing either Myc-SMN and GFP-hTERT (a --d) or Myc-SMNDeltaN27 and GFP-hTERT (e --l). The fixed and permeabilized cells were stained for Myc-SMN (c) and Myc-SMNDeltaN27 (g and k) using anti-Myc. DNA stained with DAPI shows the nucleus of each cell (a, e, and i). Images b and c, f and g, and j and k are merged to form d, h, and l, respectively. The arrows point to the nuclear gems and the arrowheads to the SMNDeltaN27-induced cytoplasmic accumulations. Bar, 12 mum. We used GFP-tagged hTERT to monitor the steady-state subcellular localization of hTERT. Addition of GFP to the N-terminus of hTERT did not alter its catalytic function, because the GFP-hTERT fusion reconstituted human telomerase activity when expressed in telomerase-negative human fibroblasts (data not shown). The GFP-hTERT fusion protein showed a diffuse nucleoplasmic distribution both in untransfected (data not shown) and in Myc-SMN --transfected HeLa cells (Figure B, b), consistent with previous reports of TERT localization . GFP-tagged hTERT did not specifically colocalize with Myc-SMN in gems and CBs (Figure B, a --d) but did frequently localize to the nucleolus of transfected cells (Figure B and data not shown). GFP-hTERT expressed in Myc-SMNDeltaN27 --transfected cells accumulated prominently in the cytoplasm in structures that colocalized with the Myc-SMNDeltaN27 protein (Figure B, e --l), in striking contrast to the restricted and diffuse nucleoplasmic localization of hTERT in Myc-SMN --transfected and untransfected cells. The colocalization of hTERT and Myc-SMNDeltaN27 was also observed in the nucleus, where GFP-hTERT accumulated in gems and CBs (Figure B, f --h). The observation that hTERT subcellular localization is affected by expression of SMNDeltaN27 suggests a functional relationship between the SMN complex and human telomerase. SMNDeltaN27 Affects Human Telomerase Reconstitution In Vitro | Coexpression of hTERT and hTR reconstitutes human telomerase activity in RRLs . Studies of human and Tetrahymena telomerase suggest that proteins present in reticulocyte extracts may be involved in reconstitution of telomerase assembly and/or activity . Western blot analysis of crude RRL extracts using a monoclonal SMN antibody revealed the presence of a single 38-kDa protein (data not shown), suggesting the presence of rabbit SMN in reticulocyte lysates. This latter observation and the profound effects of SMNDeltaN27 on hTERT subcellular localization led us to examine whether expression of SMNDeltaN27 in RRL would affect the reconstitution of human telomerase activity. In vitro --transcribed hTR was added to RRL programmed to express hTERT alone or to coexpress hTERT with SmB, wild-type human SMN, or the SMNDeltaN27 mutant. Protein synthesis was allowed to proceed for different lengths of time, followed by analysis of telomerase activity by TRAP. Human telomerase activity was undetectable after 30 min, whether hTERT was expressed alone or with another protein (Figure A, lanes 1, 4, 7, and 10). At 60 min, robust telomerase activity was observed in the control RRL reaction, in which hTERT was expressed alone (lane 2). Telomerase activity was also detected when SmB or SMN was coexpressed with hTERT (lanes 5 and 8). Lower levels of telomerase activity were observed in RRL reactions coexpressing hTERT and SmB/SMN than in the control reaction (compare lanes 5 and 8 with lane 2) because of the lower levels of hTERT protein synthesized when additional DNA is present in the RRL reaction (Figure B and data not shown). Human telomerase activity reconstituted after 60 min was barely detectable from the reaction that coexpressed hTERT and SMNDeltaN27 (lane 11). Similarly, after 90 min, the amount of telomerase activity reconstituted in the SMNDeltaN27-programmed RRL was considerably lower than in SmB- and SMN-containing RRLs (compare lane 12 with lanes 6 and 9). As can be seen in Figure B, these differences were not attributed to drastically different levels of hTERT protein (lanes 2 --4). Telomerase activity was similar to the results seen in Figure A when luciferase was expressed with hTERT rather than SmB (data not shown). The decreased efficiency of human telomerase reconstitution when SMNDeltaN27 is expressed in RRL was reproduced in three independent experiments. Telomerase activity levels were quantified from the three independent experiments and averaged, and a percentage was calculated relative to the levels of activity reconstituted in the control RRL reaction, in which hTERT was expressed alone (Figure C). These results indicate that a previously characterized dominant-negative SMN mutant, SMNDeltaN27, significantly decreases the efficiency of human telomerase reconstitution in vitro. Figure 5 | Expression of SMNDeltaN27 decreases human telomerase activity reconstitution in RRLs. Expression of SMNDeltaN27 decreases human telomerase activity reconstitution in RRLs. (A) hTERT was synthesized in RRL in the presence of hTR, [35S]methionine, and equal amounts of DNA constructs expressing Myc-SmB (lanes 4 --6), Myc-SMN (lanes 7 --9), Myc-SMNDeltaN27 (lanes 10 --12), or no additional DNA (lanes 1 --3). At 30, 60, and 90 min after the start of the RRL reactions, telomerase activity was assayed by the TRAP assay. Each TRAP reaction included an internal control (IC) to normalize for variation in PCR efficiency. (B) At 90 min after the start of the RRL reactions, equal amounts were analyzed by SDS-PAGE and autoradiography. (C) The telomerase activity was calculated as the ratio between the intensity of the telomerase ladder products and the intensity of the internal PCR control. A ratio of this telomerase activity value to the amount of in vitro --translated hTERT measured by the intensity of the S35-labeled hTERT was calculated to generate the relative telomerase activity. The activities from three independent experiments performed in the presence of SmB (diamonds), wild-type SMN (squares), and SMNDeltaN27 (triangles) were averaged and compared relative to RRL reactions in which no additional DNA construct was included. DISCUSSION : Recent results suggest that spliceosomes and transcriptosomes are preassembled in a substrate-independent manner . Similarly, telomerase is most likely preassembled into a functional RNP before recruitment to its site of action, the telomere. The assembly and maturation of many RNP particles is believed to occur in nuclear structures such as the CBs and nucleoli . Vertebrate telomerase RNA is a member of the H/ACA box family of sno RNA , and a fraction of hTR localizes to the nucleolus , suggesting that vertebrate telomerase assembly and/or maturation transits through the nucleolus. Furthermore, recently reported that in vitro --transcribed human and Xenopus telomerase RNAs microinjected into Xenopus oocyte nuclei localize not only to nucleoli, but also to CBs. Our results support a model in which the human telomerase RNP is assembled and/or matured into a functional enzyme by transit through the nucleoli and/or the CBs. The physical association between endogenous hTERT and SMN, a protein involved in RNP assembly that localizes in both nucleoli and CBs, strongly suggests that SMN plays a role in human telomerase biogenesis. Two experimental observations are in support of the association between SMN and telomerase does not merely reflect the fact they both colocalize to similar nuclear structures: nucleoli and/or CBs. First, recombinant telomerase can specifically bind in vitro --translated SMN (Figure A). Second, hTERT protein and telomerase activity is undetectable in immunoprecipitates performed using antibodies specific to the box C/D snoRNP nucleolar protein fibrillarin (data not shown). The effect of SMNDeltaN27 on hTERT subcellular localization and telomerase reconstitution in vitro further supports a functional role for SMN in telomerase assembly. The mechanism by which SMNDeltaN27 disturbs the cellular organization of snRNPs and snoRNPs has not yet been defined. GFP-hTERT did not specifically colocalize with wild-type SMN in gems and CBs, yet it accumulated and colocalized with SMNDeltaN27 . These results suggest dynamic and transient interactions between the SMN complex and components of the telomerase RNP. Similar results are observed in immunofluorescence analyses of several components of the RNA polymerase II transcription machinery upon overexpression of SMN and SMNDeltaN27 . The effect of SMNDeltaN27 on hTERT cellular localization is consistent with the proposed view that this SMN mutant sequesters associated proteins in cellular bodies by blocking or retarding their release and/or their transport between different nuclear bodies . SMNDeltaN27 expression also resulted in the accumulation and the detection of GFP-hTERT in the cytoplasm of cotransfected cells (Figure B). The accumulation of GFP-hTERT in the cytoplasm was never observed in cells either cotransfected with wild-type SMN (Figure B) or transfected with the GFP-hTERT construct alone (data not shown). Sm proteins also colocalize with the SMNDeltaN27-induced cytoplasmic accumulations (Figure A) , possibly as a result of a perturbed interaction between endogenous SMN and Sm proteins in the cytoplasm . Yet, SMNDeltaN27 does not elicit the cytoplasmic accumulation of other SMN-associated proteins such as p80coilin , components of the RNA pol II complex , and snoRNP proteins . Our results demonstrating the accumulation of GFP-hTERT in cytoplasmic SMNDeltaN27-containing aggregates suggest that the interaction between the SMN complex and hTERT could be initiated in the cytosol. The SMN-hTERT complex could then relocalize to the nucleus and encounter a fully processed hTR-snoRNP protein complex in subnuclear domains such as the nucleolus and/or the CBs . How might SMN be involved in human telomerase biogenesis? The best-characterized function of SMN is its role in snRNP assembly. Experiments in Xenopus oocytes and those using a cell-free system for in vitro reconstitution of UsnRNP assembly suggest that the SMN complex is involved in facilitating the association of distinct snRNAs with Sm proteins . hTERT, hTR, and human telomerase activity are not detected in anti-Sm immunoprecipitates (Figures and ), suggesting that the Sm protein complex is not involved in human telomerase biogenesis. The lack of association between hTR and Sm proteins was previously noted . Thus, on the basis of our results and recent studies that report interactions between SMN and snoRNP proteins , we propose that SMN may function in human telomerase assembly through the association of SMN with snoRNP proteins such as hGAR1. The human GAR1 snoRNP protein is an attractive candidate, because it associates with hTR . Using antibodies specific for H/ACA snoRNP proteins, we were unable to immunodeplete the SMN-associated telomerase activity (data not shown). However, hGAR1 and dyskerin, two hTR-associated H/ACA snoRNP proteins , and SMN are present in partially purified human telomerase fractions generated using anion and size exclusion chromatography followed by differential-density ultracentrifugation through a cesium sulfate gradient (data not shown). These observations, coupled with the nucleolar localization of hTR , hTERT (Figure and data not shown), SMN , and hTR-associated snoRNP proteins , suggest that SMN is involved in human telomerase biogenesis as an H/ACA snoRNP. Our results also establish major differences between S. cerevisiae and vertebrate telomerase RNPs. The budding yeast telomerase RNA associates with Sm proteins and gains a 5'-TMG cap structure ; both events are characteristic of snRNP assembly. However, hTR was not recovered from anti-Sm and anti-TMG immunoprecipitates , suggesting that human telomerase is not processed as an snRNP. This conclusion is supported by recent experiments in which in vitro --synthesized hTR was microinjected into Xenopus oocyte nuclei . SnoRNAs are generated from two different genomic contexts: pre-mRNA introns or their own independent transcription unit . The U3, U8, and U13 box C/D snoRNAs are transcribed from their own promoters and receive a TMG cap, whereas most intron-generated snoRNAs do not undergo 5' hypermethylation . The lack of a TMG cap at the 5' end of hTR is thus surprising, because it is expressed from its own promoter as an RNA polymerase II transcript . However, to the best of our knowledge, none of the metazoan H/ACA box snoRNAs have been shown to receive a TMG cap. Further studies will be necessary to better understand the processing and maturation events required for functional hTR formation. Previous data support the view that telomerase reconstitution in RRLs is facilitated by the action of proteins present in the extracts . The negative effect of SMNDeltaN27 on in vitro telomerase reconstitution in RRL is consistent with this view. However, when SMNDeltaN27-containing extracts from RRL or 293 cells were added to previously reconstituted human telomerase, the activity of telomerase was not affected (data not shown). Thus, the mutant form of SMN may affect telomerase not by inhibiting its catalytic activity but rather by affecting assembly in vitro. The incomplete inhibition of human telomerase reconstitution by SMNDeltaN27 in vitro could reflect a partial decrease in the efficiency of reconstitution or the inability to obtain concentrations of SMNDeltaN27 sufficient to completely sequester the endogenous RRL proteins involved in telomerase assembly. The detailed characterization of the cellular components and progressive steps involved in human telomerase assembly will be critical for the rational design of new telomerase inhibitors. The identification of SMN as a telomerase-associated protein suggests that it will be an important player in the functional assembly and activation of human telomerase. Future studies will focus on understanding the specific role performed by SMN and its associated proteins in human telomerase biogenesis. Backmatter: PMID- 12221126 TI - alpha4beta1 Integrin Regulates Lamellipodia Protrusion via a Focal Complex/Focal Adhesion-independent Mechanism AB - alpha4beta1 integrin plays an important role in cell migration. We show that when ectopically expressed in Chinese hamster ovary cells, alpha4beta1 is sufficient and required for promoting protrusion of broad lamellipodia in response to scratch-wounding, whereas alpha5beta1 does not have this effect. By time-lapse microscopy of cells expressing an alpha4/green fluorescent protein fusion protein, we show that alpha4beta1 forms transient puncta at the leading edge of cells that begin to protrude lamellipodia in response to scratch-wounding. The cells expressing a mutant alpha4/green fluorescent protein that binds paxillin at a reduced level had a faster response to scratch-wounding, forming alpha4-positive puncta and protruding lamellipodia much earlier. While enhancing lamellipodia protrusion, this mutation reduces random motility of the cells in Transwell assays, indicating that lamellipodia protrusion and random motility are distinct types of motile activities that are differentially regulated by interactions between alpha4beta1 and paxillin. Finally, we show that, at the leading edge, alpha4-positive puncta and paxillin-positive focal complexes/adhesions do not colocalize, but alpha4beta1 and paxillin colocalize partially in ruffles. These findings provide evidence for a specific role of alpha4beta1 in lamellipodia protrusion that is distinct from the motility-promoting functions of alpha5beta1 and other integrins that mediate cell adhesion and signaling events through focal complexes and focal adhesions. Keywords: INTRODUCTION : Cell migration is essential for a variety of biological events, including embryonic development, wound healing, inflammation, and metastasis of malignant cells. Cell migration along a substratum is regulated by extracellular signals transduced into cells partly through adhesive interactions between the cell and its surrounding extracellular matrix (ECM). Integrins, the major receptors that mediate cell --ECM interactions , play important roles in regulating cell motility. Integrins are a large family of heterodimeric cell adhesion receptors. Many integrins, including alpha5beta1 and alphaVbeta3, mediate cell-ECM adhesion by forming junctional complexes called focal adhesions, which bind extracellularly to specific ECM components and intracellularly to cytoskeletal proteins and signaling molecules. In cultured adherent cells, such as fibroblasts, focal adhesions play key roles in regulating motility . When fibroblasts begin to migrate on an ECM substratum, small nascent focal complexes assemble in plasma membrane protrusions at the leading edge of the cell. These complexes grow larger and subsequently recruit alpha5beta1 and other integrins as they evolve into highly organized focal adhesions . As a cell moves forward, focal adhesions not only act as anchors but also function as nucleation and activation sites for signaling proteins, which in turn activate an intracellular signaling network, leading to actin cytoskeletal reorganization and generation of cell motility . alpha4beta1, a member of the integrin family, is not localized in focal adhesions in most cell types, yet this integrin also plays important roles in cell migration. alpha4beta1 binds to an alternatively spliced V25 (also called CS-1) region of fibronectin (FN) instead of the RGD sequence that is recognized by alpha5beta1 and other integrins localized to focal adhesions . alpha4beta1 also binds to vascular cell adhesion molecule-1 (VCAM-1), a member of the immunoglobulin superfamily . alpha4beta1 is expressed in many migratory cell types in vivo, including neural crest cells and their derivatives , smooth muscle cells of newly formed blood vessels , hematopoietic cell lineages , and epicardial progenitor cells . Furthermore, the migration of neural crest cells and hematopoietic cells on FN can be disrupted in culture by antibodies that specifically inhibit binding between alpha4beta1 and FN , and progenitor cells fail to migrate on the heart to form the epicardium in mouse embryos deficient in alpha4beta1 . Although an important role for alpha4beta1 in cell migration has been well documented, questions remain as to how alpha4beta1 promotes cell migration. Because alpha4beta1 is not localized in focal adhesions in most cell types and has a ligand-binding specificity different from integrins in focal adhesions, this integrin may promote cell migration by a mechanism distinct from that of alpha5beta1 and other integrins in focal adhesions. This idea is also supported by an observation that the cytoplasmic tails of alpha4 and alpha5 subunits confer different cellular activities with the alpha4 tail conferring migratory activities and the alpha5 tail conferring adhesive activities . In this article, we examined the migratory behaviors of Chinese hamster ovary (CHO) cells ectopically expressing alpha4beta1, by using a scratch-wound assay. Our data show that alpha4beta1 plays a unique role in promoting lamellipodia protrusion through a focal complex/focal adhesion-independent mechanism. MATERIALS AND METHODS : Construction of Plasmids | For expressing green fluorescent protein (GFP)-tagged alpha4 integrin in CHO cells, we constructed a plasmid pQN4G. To construct this plasmid, upstream human alpha4 cDNA (; obtained from American Type Culture Collection, Rockville, MD) and downstream mouse alpha4 cDNA (; a generous gift from Dr. Martin Hemler, Dana-Farber Cancer Institute, Boston, MA) were joined at a unique and conserved KpnI site, and the 3' end of this chimeric alpha4 cDNA was ligated to the 5' end of GFP cDNA by insertion into PGBI25-fN1 GFP plasmid vector (Quantum Biotechnologies, Montreal, Quebec, Canada). The fusion protein's expression was driven by a cytomegalovirus promoter. The pQN4Y991AG plasmid was the same as pQN4G except that the tyrosine at position 1093, equivalent to position 991 in human alpha4 cDNA product, of the alpha4 tail region was mutated to alanine by polymerase chain reaction. Purified Ligands | Mouse plasma FN was purchased from Invitrogen (Carlsbad, CA). A recombinant FN fragment containing FN type III repeats 12 --15 and the CS-1 region and recombinant soluble VCAM-1 were provided by Richard Hynes (Massachusetts Institute of Technology, Cambridge, MA) and Roy Lobb (Biogene, Cambridge, MA), respectively. Cells, Transfections, and Cell Culture | CHO cells were maintained in DMEM containing 10% fetal bovine serum (FBS) (Hyclone Laboratories, Logan, UT), l-glutamine, and antibiotics. CHO-alpha4 cells , provided by Martin Hemler (Dana-Farber Cancer Institute), were maintained in F-12 (Invitrogen) containing 10% FBS, l-glutamine, antibiotics, and 0.4 mg/ml G418 (Life Invitrogen). CHOB2, CHOB2-alpha4, and CHOB2-alpha5 cell lines, provided by Cary Wu (University of Pittsburgh, Pittsburgh, PA) and Michael DiPersio (Albany Medical College, Albany, NY), were maintained in minimal essential medium-alpha (Invitrogen) containing 10% FBS, l-glutamine, antibiotics, and 0.4 mg/ml G418. CHO cells were transfected with pQN4G and pQN4Y991AG by using Lipofectin/Optimem (Invitrogen) following manufacturer's instructions. CHO-alpha4/GFP and CHO-alpha4Y991A/GFP clones were selected in F-12 containing 0.8 mg/ml G418 and screened for alpha4/GFP and alpha4Y991A/GFP expression by using fluorescence-activated cell sorting. Stably transfected cell lines were maintained using the initial concentration of G418. Analysis of Lamellipodia Protrusion and alpha4-Positive Puncta Formation at Edges of Scratch-Wounds | For the studies using time-lapse microscopy, cells were plated onto glass bottom Microwell dishes (MatTek, Ashland, MA) coated with 10 mug/ml FN or VCAM-1 for 2 h at 37C. At confluence, the cell monolayer was scraped with a Pipetman tip to generate scratch-wounds. The wounded surface was washed with phosphate-buffered saline (PBS) and then returned to serum-containing medium. After 2-h incubation, media were replaced with Leibovitz's L-15 medium (Invitrogen) containing 10% FBS. Migration at the wound edge was monitored by phase or fluorescence microscopy by using an Axiovert 135 TV microscope (Carl Zeiss, Thornwood, NY) equipped with a temperature controller (Harvard Apparatus, Holliston, MA). Cell movement was recorded with a charge-coupled device camera (Roper Photometrics, Trenton, NJ) by using IPLab-Spectrum software (Scanalytics, Fairfax, VA). The last frame of each time-lapse movie was analyzed for the percentage of cells at wound edges that protruded broad lamellipodia. For studies using regular microscopy, cells were plated on tissue culture plates or coverslips coated with 10 mg/ml FN and scratch-wounded as described above. Nonoverlapping fields were photographed at designated time points by phase (250x) or fluorescence microscopy (630x). The percentage of cells at wound edges that protruded broad lamellipodia or formed alpha4-positive puncta was scored using the phase or fluorescence micrographs, respectively. This method was also used in an antibody perturbation experiment on the fanning activity of CHOB2-alpha4 cells. In this experiment, an anti-alpha4 antibody P1H4 (Chemicon International, Temecula, CA), which is identical to a functional blocking antibody, P4C2, was added to the cells at 25 mg/ml. The cells were preincubated with the antibody for 2 h at room temperature, plated on FN and cultured in the presence of the antibody before and after scratch-wounding. At the 2-h time point, the cells at wound edges were photographed and scored for the percentage of cells at wound edges that protruded broad lamellipodia. Flow Cytometry | Flow cytometry analysis was performed as described by with some modifications. Washed cells were resuspended at 2 x 106 cells/ml in PBS, containing 5% normal goat serum (Vector Laboratories, Burlingame, CA) and 1% bovine serum albumin (BSA) (PBS/NGS/BSA), and blocked on ice for 20 min. Cells (100 mul) were mixed with 100 mul of one of the following primary antibodies at 20 mug/ml: mouse anti-alpha4 (alpha4-PUJ1; Upstate Biotechnology, Lake Placid, NY), mouse anti-hamster alpha5, PB1 , or mouse anti-hamster beta1, 7E2 . PB1 and 7E2 were provided by Rudy Juliano (Department of Pharmacology, University of North Carolina, Chapel Hill, NC). After 45 min on ice, cells were washed with PBS and resuspended in 100 mul of PBS/normal goat serum/BSA containing 20 mug/ml of either fluorescein- or R-phycoerythrin --conjugated secondary antibodies (BioSource International, Camarillo, CA). After 45 min and a final wash with PBS, cells were resuspended in 0.5 ml of 2% paraformaldehyde in PBS and analyzed on a FACStar Plus with an Innova-90 laser (Coherrent, Santa Clara, CA) exciting at 488 nm wavelengths and running at 100 mW. Adhesion Assay | The adhesion assay was performed as in with the following modifications. Triplicate wells of 96-well plates were coated with 10 mug/ml FN, CS-1, or VCAM-1 at 37C for 2 h. Then 5 x 104 cells were plated per well and allowed to adhere in a tissue culture incubator. After 15 min, nonadherent cells were removed by submerging the plate in PBS and shaking off the cells. Seven nonoverlapping high-power fields (200x) along the diameter of each well were photographed, and the number of adherent cells per field was counted. Migration Assays | For the scratch-wound cell migration assay, the cells were plated on wells of 24-well tissue culture plates coated with 10 mg/ml FN and scratch-wounded as described above to generate scratch-wounds 0.28 --0.56 mm in width. Scratch-wounds were allowed to heal in medium containing 10% FBS in a tissue culture incubator. Photographs were taken at designated time points with a phase microscope (Nikon, Melville, NY). By using the photographs, the distance cells migrated was calculated as a percentage of wound closure. For each data point, 10 --30 nonoverlapping measurements were taken from multiple wells; mean and SDs were calculated from three independent experiments. Rates of cell proliferation were measured by immunohistochemical detection of 5-bromo-2-deoxyuridine (BrdU; Sigma-Aldrich, St. Louis, MO) incorporation . Cells were plated on FN-coated coverslips and wounded as described above. Fifteen minutes before the time point, the culture medium was replaced by medium containing 50 muM BrdU and 10% FBS. After 15-min culturing, cells were washed in PBS, fixed with cold methanol, washed with PBS, and treated with 1.5 M HCl in the tissue culture incubator for 40 min. Cells were then washed with PBS and stained with an anti-BrdU antibody (Sigma-Aldrich). For each cell line, microscopic fields (630x) from three coverslips were photographed. Percentage of cells with BrdU incorporation in each microscopic field was determined and mean and SDs were calculated. The Transwell cell migration assay was performed as described in with the following exceptions. Transwell inserts were coated with 10 mug/ml FN in serum-free F-12 for 2 h at 37C. Media from the top chamber were replaced with 200 mul of cell suspension (1.5 x 105 cells/ml in F-12), and chambers were incubated for 4 h at 37C. Cell Surface Biotinylation and Immunoprecipitation | CHO-alpha4 cells were surface biotinylated by resuspending at 5 x 106 cells/ml in cell wash buffer (50 mM Tris pH 7.5, 0.15 M NaCl, 1 mM CaCl2, and 5 mM MgCl2) and incubating with EZ-Link NHS-LC-biotin (, Rockford, IL) for 60 min at room temperature. Cells were lysed for 15 min at 4C in ice-cold extraction buffer (0.5% NP-40, 2 mM phenylmethylsulfonyl fluoride, and 0.02 mg/ml aprotonin in cell wash buffer). For coimmunoprecipitation studies, CHO-alpha4/GFP and CHO-alpha4Y991A/GFP cells were washed three times with PBS and lysed in ice-cold lysis buffer (20 mM Tris-HCl pH 7.4, 150 mM NaCl, 10 mM EDTA, 1% Triton X-100, 0.05% Tween 20, 2 mug/ml aprotinin, and 0.5 mug/ml leupeptin) for 30 min at 4C. The cell lysates were cleared with protein-G agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA) and subjected to immunoprecipitation by using an anti-alpha4 antibody (5B10G; Upstate Biotechnology) and protein-G agarose beads. Immunoprecipetates were analyzed on Western blots for the presence of alpha4/GFP, alpha4Y991A/GFP, and paxillin, by using an anti-GFP antibody (Molecular Probes, Eugene, OR) and an anti-paxillin antibody (Transduction Laboratories, Lexington, KY), respectively. Fluorescence and Confocal Microscopy | Cells were plated on glass coverslips coated with 10 mg/ml FN and scratch-wounded. At 3 h or otherwise designated time points after scratch-wounding, the coverslips were collected. For GFP fluorescence, cells were washed three times in PBS, fixed for 15 min in 4% paraformaldehyde (Fluka Chemical, Ronkonkoma, NY) in PBS, and mounted. For immunofluorescence staining, cells were washed three times in PBS, fixed for 15 min in 4% paraformaldehyde in PBS, permeabilized for 15 min in 0.5% NP-40 (Sigma-Aldrich) in PBS, and incubated with antibodies against paxillin (349 from Transduction Laboratories; and 165, a gift from Christopher Turner, SUNY Upstate Medical University, Syracuse, NY) at 37C. After 30 min, coverslips were washed three times in PBS, incubated with a secondary antibody (BioSource International) at 37C for 30 min, and washed three times in PBS. Both primary and secondary antibodies were diluted in 10% normal goat serum in PBS. Fluorescent images were obtained using an Axioskop 2 microscope (Carl Zeiss) in conjuction with a Coolsnap fx charge-coupled device camera (Photometrics, Tuscson, AZ) controlled by IPLab-Spectrum software. Confocal images were obtained using the Oz confocal laser scanning microscope system (Noran, Middleton, WI), with Intervision software, version 6.5, on a Silicon Graphics O2 platform. Online Supplemental Material | The online version of this article contains QuickTime movies that accompany Figures , , , and . The speed of the movies is 60x faster than real time. Videos 1 --5 accompany Figure , videos 6 --8 accompany Figure , video 9 accompanies Figure , and video 10 accompanies Figure . Online supplemental material is available at Figure 1 | Time-lapse microscopy of CHO, CHO-alpha4, and CHO-alpha4/GFP cells at wound edges. Time-lapse microscopy of CHO, CHO-alpha4, and CHO-alpha4/GFP cells at wound edges. CHO cells on FN (row A), CHO-alpha4 cells on FN (row B), CHO-alpha4/GFP cells on FN (row C), CHO-alpha4 cells on VCAM-1 (row D), and CHO-alpha4 cells on RGD (row E) were plated and wounded. Two hours after wounding, the cells were photographed every 2 min for 2 h. Four frames of each cell type at 0-, 40-, 80-, and 120-min time points are shown. Bar, 50 mum. Figure 2 | Time-lapse microscopy of CHOB2, CHOB2-alpha5 and CHOB2-alpha4 cells at wound edges. Time-lapse microscopy of CHOB2, CHOB2-alpha5 and CHOB2-alpha4 cells at wound edges. CHOB2 cells (row A), CHOB2-alpha5 cells (row B), and CHOB2-alpha4 cells (row C) were plated and wounded. Two hours after wounding, the cells were photographed every 2 min for 2 h. Three frames of each cell type at 0-, 40-, 80-, and 120-min time points are shown. Bar, 50 mum. RESULTS : alpha4beta1 Integrin Promotes Lamellipodia Protrusion When Ectopically Expressed in CHO Cells | It was previously shown that CHO cells express alpha5beta1 but not alpha4beta1 . Using a Transwell assay, Martin Hemler and colleagues have shown that when alpha4beta1 is ectopically expressed in CHO cells, this integrin enhances cell motility . To determine how alpha4beta1 promotes cell motility, we used a scratch-wound assay and time-lapse microscopy to examine the migratory behaviors of a CHO cell line that stably expresses alpha4 (CHO-alpha4 cells) . In the scratch-wound assay, the CHO-alpha4 cells and the parental CHO cells were plated on FN-coated dishes; as the cells formed a confluent monolayer, a scratch-wound was made in the monolayer to induce cell migration into the wound. Two hours later (2-h time point), the cells at wound edges were imaged by time-lapse microscopy. As shown in Figure , A and B (Videos 1 and 2), CHO-alpha4 cells and the parental CHO cells displayed very different migratory behaviors. CHO cells at wound edges migrated as a mass. While the monolayer of CHO cells pushed toward the scratch-wound, individual cells protruded short-lived membrane extensions in random directions with little persistent polarity toward the wound. No prominent lamellipodia were observed (Figure A). In contrast, some CHO-alpha4 cells at the wound edge migrated into the wound as individual cells by forming fan-shaped broad lamellipodia with a persistent polarity toward the wound (Figure B). At the 4-h time point (2 h after starting to take the movies) 21% of all the cells at the wound edges exhibited this "fanning" behavior . We also performed the scratch-wound assay and photographed the cells at the 12- and 18-h time points. At these later time points, the majority of the CHO-alpha4 cells at wound edges had the fan shape, whereas none of the CHO cells at wound edges did (our unpublished data). We generated several stable cell lines expressing alpha4, which was tagged with GFP (CHO-alpha4/GFP, see below for characterization of these cell lines), and found that all of the alpha4/GFP-expressing cell lines also had the fanning phenotype (Figure C and Video 3) to a similar degree . Therefore, the fanning phenotype was not due to a cloning artifact. Figure 3 | Percentage of fanning cells at wound edges. Percentage of fanning cells at wound edges. The last frame of each movie in Figures and were analyzed. Cells at the edge of the scratch-wound (n = 30 --50) were scored for fanning, and the percentage of these cells was calculated. This was done for each of the following cell lines: CHO, CHO-alpha4, CHO-alpha4/GFP, CHO-alpha4Y991A/GFP, CHOB2, CHOB2-alpha5, and CHOB2-alpha4 on a FN substrate as well as for CHO-alpha4 on both VCAM-1 and RGD. alpha4beta1 Promotes Lamellipodia Protrusion Independent of alpha5beta1 | The CHO-alpha4 and CHO-alpha4/GFP cells also express alpha5beta1. Our flow cytometry analyses showed that the expression level of alpha5 at the cell surface was slightly reduced in these cells (Figure A). Thus, the fanning behavior may be promoted directly by alpha4beta1, or indirectly due to decreased expression of alpha5beta1 at the cell surface. To distinguish between these possibilities and determine whether the fanning phenotype depends on alpha4beta1, we tested the CHO-alpha4 cells in the scratch-wound assay by using an alpha4beta1-specific ligand, VCAM-1, or an alpha5beta1-specific ligand, the RGD peptide, as the substrate. The cells fanned on VCAM-1 (Figure D and Video 4) but not on the RGD peptide (Figure E and Video 5). At the 4-h time point, CHO-alpha4 cells plated on RGD alone showed no evidence of fanning . However, 27% of the CHO-alpha4 cells on VCAM-1 at wound edges exhibited the fanning behavior. Because alpha4beta1 is the only receptor for VCAM-1 in CHO-alpha4 cells, this result indicates that the binding between alpha4beta1 and VCAM-1 is sufficient for the fanning phenotype, and that the RGD-recognizing integrins, including alpha5beta1, are not sufficient. Figure 4 | GFP-tagged alpha4 integrin is functionally normal. GFP-tagged alpha4 integrin is functionally normal. (A) Flow cytometry analysis. Unlabeled CHO cells served as negative controls (column 1). Cells were labeled with either anti-human alpha4 (column 2), anti-hamster beta1 (column 3), or anti-hamster alpha5 (column 4). Fluorescence intensity is shown in logarithmic scale. (B) alpha4/GFP, alpha4, or GFP alone was each transfected into CHO cells. The CHO-alpha4 cells (a) and the parental CHO cells (c) were stained with an anti-alpha4 antibody by immunofluorescence. alpha4/GFP (b) localized at the surface of transfected cell in the same pattern as alpha4 in the CHO-alpha4 cells (a). The parental CHO cells had only background staining (c). GFP alone localized diffusely in the nucleus and cytoplasm and did not reach the cell surface (d). Bar, 10 mum. (C) Number of adherent cells per high-power field (n = 7) was determined for each of the triplicate wells for CHO, CHO-alpha4, and CHO-alpha4/GFP cells adhering to FN, CS-1, or VCAM-1. The mean values and standard derivations for the triplicates were graphed. Note that the CHO-alpha4 and CHO-alpha4/GFP cell lines but not the CHO cells adhered to CS-1 and VCAM-1. (D) Cells were photographed at 0-, 12-, and 18-h time points. The percentage of wound closure was determined (n = 20 --30) from at least two (3 for all but alpha4/GFP) independent experiments. The mean values and SDs were graphed. Note that CHO-alpha4 and CHO-alpha4/GFP cells both had a much faster wound closure rate than that of CHO cells, with no significant difference between the wound closure rates of CHO-alpha4 and CHO-alpha4/GFP. To test directly the roles of alpha4beta1 and alpha5beta1 in the fanning phenotype, we took advantage of a CHO-derived cell line, named CHOB2, which expressed a negligible level of alpha5beta1 . It has previously been shown that alpha5beta1, when stably transfected into CHOB2 cells, can rescue the ability of CHOB2 cells to adhere to and migrate randomly on FN . When alpha4beta1 is stably expressed in CHOB2 cells, it can also rescue the cells for adhesion and migration . We compared the alpha5- and alpha4-expressing CHOB2 cell lines for their migratory behaviors at edges of scratch-wounds, by using FN as the substrate. The results are shown in Figure . CHOB2 cells adhere poorly to FN. To examine their migratory behaviors at wound edges, the cells were plated and scratch-wounded, but the washes were omitted to allow the cells to remain on the dish. These cells (Figure A and Video 6) at wound edges did not show any polarization, although the cells were able to protrude membrane extensions in a random manner. CHOB2-alpha5 cells (Figure B and Video 7) adhered and migrated much in the same manner as the CHO cells. At the 4-h time point, CHOB2 and CHOB2-alpha5 both showed no evidence of fanning at wound edges . In contrast, some CHOB2-alpha4 cells at wound edges (Figure C and Video 8) migrated by forming lamellipodia with persistent polarity toward the wound, although to a lesser extent compared with CHO-alpha4 cells. At the 4-h time point, 11% of CHOB2-alpha4 cells at wound edges exhibited the fanning behavior . These results show that 1) alpha5beta1 alone does not promote fanning on FN and alpha4beta1 is required; and 2) alpha4beta1 is sufficient to promote fanning on FN, but optimal fanning on FN also requires alpha5beta1. To test further whether alpha4beta1 is required for the fanning phenotype, we performed an antibody perturbation experiment, by using an anti-alpha4 antibody that specifically disrupted binding between alpha4beta1 and FN . When this antibody was added to scratch-wounded CHOB2-alpha4 cells, considerably fewer cells exhibited the fanning activity than did the CHOB2-alpha4 cells in the absence of the antibody . This result indicates that the ligand-binding activity of alpha4beta1 is required for promoting the fanning behavior. Table 1 | Percentage of cells fanning at the wound edge at the 2-h time point Total number of cells was 340 and 654 for no antibody and with antibody, respectively. We conclude that alpha4beta1 and alpha5beta1 play different roles in the migration of CHO cells at wound edges, and alpha4beta1 plays a unique role in promoting lamellipodia protrusion. alpha4 Integrin Is Functional When Tagged with GFP | To study the mechanisms by which alpha4beta1 promotes lamellipodia protrusion, we attached GFP to the cytoplasmic tail of the alpha4 subunit, producing the alpha4/GFP fusion protein. Two independent cell lines stably expressing alpha4/GFP were generated and analyzed. Cells from both cell lines migrated in the same manner as the CHO-alpha4 cells (Figure C and Figure ). To determine more closely whether the GFP tag interfered with the function of alpha4beta1, we used the CHO-alpha4 cell line as a control to characterize the GFP-tagged alpha4 for its surface expression, adhesive activity, and migration-promoting activity. Flow cytometry analysis showed that the surface expression levels of alpha4/GFP in the transfected cell lines were similar to that of alpha4 in the control CHO-alpha4 line (Figure A). Both the CHO-alpha4 and CHO-alpha4/GFP cell lines also had surface expression levels of beta1 similar to CHO cells, yet they had slightly decreased levels of alpha5 compared with the parental CHO cells. We also compared the surface distribution of alpha4 with and without the GFP tag (Figure B). alpha4/GFP and alpha4 were both expressed over the entire cell surface when the cells were plated sparsely, as assayed by GFP fluorescence and anti-alpha4 immunofluorescence, respectively. alpha4 and alpha4/GFP were also both detected within the cells, possibly in the ER and Golgi complexes. In control cells transfected with GFP cDNA alone, GFP was expressed in the cytoplasm and nucleus but not on the cell surface. These data showed that alpha4/GFP was distributed normally at the cell surface. Because alpha5beta1 binds to the RGD region of FN , and alpha4beta1 binds to the CS-1 region of FN, we predicted that the parental CHO cells, which express alpha5beta1 but not alpha4beta1, should adhere to full-length FN (containing both regions) but not to a fragment of FN that contains only the CS-1 region. In contrast, CHO-alpha4 and CHO-alpha4/GFP should adhere to both. Because alpha4beta1 also binds to VCAM-1, we predicted that, if the (alpha4/GFP)beta1 protein was functional, the transfected cells would adhere to VCAM-1, whereas the parental CHO cells would not. Cell adhesion assays on these cells fulfilled our predictions (Figure C), demonstrating that (alpha4/GFP)beta1 has normal adhesive activities. As discussed above, we have shown that CHO-alpha4 and CHO-alpha4/GFP cells both display a fanning behavior at edges of scratch-wounds. We also compared their rates of wound closure and found that both CHO-alpha4 and CHO-alpha4/GFP cells closed the scratch-wounds much faster than the CHO cells, with the migration rates of CHO-alpha4 and CHO-alpha4/GFP cells not significantly different from each other (Figure D). In parallel with the wound assay, cell proliferation rates were measured for these cell lines under the same plating and wounding conditions as the cells in the wound assays. Our data showed that CHO, CHO-alpha4,and CHO-alpha4/GFP cells proliferated at the same rate . Thus, the faster wound-closure rates of CHO-alpha4 and CHO-alpha4/GFP cells were not due to a faster proliferation rate but due to a faster migration rate of these cells. Table 2 | Proliferation rates of CHO, CHO-alpha4 and CHO-alpha4/GFP cells in wound assays In summary, we show that (alpha4/GFP)beta1 has the same localization, adhesive activities, and migration-promoting activities as untagged alpha4beta1. We conclude that (alpha4/GFP)beta1 is functionally normal. alpha4beta1 Forms Transient Puncta at Leading Edge of Cells That Begin to Protrude Lamellipodia in Response to Scratch-Wounding | Using a fluorescence microscope attached to a time-lapse imaging system, we examined the surface dynamics of (alpha4/GFP)beta1 on the CHO-alpha4/GFP cells that displayed fanning activity . At low magnification, (alpha4GFP)beta1 was seen over the entire surface of the cells. The alpha4/GFP fluorescence was particularly strong in membrane ruffles at the leading edge (arrowheads in Figure , a --c). At higher magnification, transient alpha4/GFP-positive puncta were found in cells that had just begun to fan into the scratch-wounds. The alpha4/GFP-positive puncta were seen in some areas at the leading edge where the ruffles flattened out and the membrane extended into smooth edged but small lamellipodia (arrows in Figure , e and k). The puncta were located right along the leading edge of these small lamellipodia. As the lamellipodia continued to extend and a new cell front formed, the puncta stayed at their original positions (arrow in Figure f). The extended membrane then began to ruffle again, whereas the alpha4/GFP puncta gradually disappeared (Figure g, h, and l). As the new cell front was ruffling, alpha4/GFP fluorescence was again seen in the ruffles (arrowheads in Figure , j and l). This sequence of events was repeated continuously as the cell formed a broad lamellipodia and moved forward toward the wound (Video 9). This pattern of surface dynamics may be related to the unique function of alpha4beta1 in promoting broad lamellipodia protrusion. Figure 5 | Time-lapse microscopy of alpha4/GFP on the surface of CHO-alpha4/GFP cells migrating into a scratch-wound. Time-lapse microscopy of alpha4/GFP on the surface of CHO-alpha4/GFP cells migrating into a scratch-wound. CHO-alpha4/GFP cells were plated and wounded. After 3 h, the cells with fanning activity were photographed under a fluorescence microscope at 2-min intervals for 26 min. Three frames at low magnification (a --c) and nine frames at high magnification (d --l) are shown representing this time period. Asterisks are fixed reference marks to highlight the movement of the cell. Note that alpha4/GFP fluorescence was very strong in ruffles (arrowheads). As the ruffles flattened out, alpha4/GFP fluorescence was seen as puncta along the leading edge of small lamellipodia protrusions (e and k, arrows). As the leading edge continued to extend forward, the puncta of fluorescence stayed at a fixed position behind of the new cell front (f, arrow), and they disappeared when the cell front formed new ruffles (j and l, arrowheads). Bars, 10 mum. Disrupting alpha4/Paxillin Binding Allows a Faster Response of Cells to Scratch-Wounding That Correlates with Formation of alpha4-Positive Puncta at the Leading Edge | Paxillin is a signaling adaptor protein , which binds to the cytoplasmic tail of alpha4 . A point mutation at the alpha4 tail, Y991A, disrupts this binding . It has been shown that this mutation reduces random cell motility . We reasoned that, if this mutation affects the alpha4beta1-dependent fanning phenotype as well as the formation of the alpha4-positive puncta at the leading edge, we would be able to establish a mechanistic relationship between fanning and puncta formation. Therefore, we generated and analyzed two independent CHO cell lines that stably express alpha4 cDNA carrying this mutation; the mutant alpha4 was tagged with GFP (the cell lines are referred to as CHO-alpha4Y991A/GFP). To confirm that the Y991A mutation disrupted the paxillin-binding in the CHO-alpha4Y991A/GFP cells, we performed a coimmunoprecipition experiment. We showed that the amount of paxillin that coimmunoprecipitated with the mutant alpha4 was much less than that coimmunoprecipitated with the wild-type alpha4 (Figure B). We also showed that both alpha4/GFP and alpha4Y991A/GFP remained intact when expressed in CHO cells (Figure A). To confirm that the Y991A mutation reduces random motility as reported by Ginsberg and colleagues , we compared the abilities of the CHO-alpha4/GFP and CHO-alpha4Y991A/GFP cells to migrate in the Transwell assay. We found that the motility of CHO-alpha4Y991A/GFP cells on FN was reduced by similar55% (Figure C). Figure 6 | Characterization of CHO-alpha4Y991A/GFP cells. Characterization of CHO-alpha4Y991A/GFP cells. (A) Tagged GFP was not cleaved from the alpha4/GFP fusion protein when the protein is expressed in CHO cells. CHO-alpha4 cells were surface biotinylated, lysed, and immunoprecipitated with an anti-alpha4 antibody, by which alpha4 integrin was detected as cleavage fragments of 70 and 80 kDa. CHO-alpha4/GFP and CHO-alpha4Y991A/GFP cells were lysed and subjected to immunoblot analysis by using an anti-GFP antibody, which revealed a 100-kDa cleavage fragment of alpha4/GFP fusion protein (the 70-kDa cleavage fragment of alpha4 plus 30 kDa of GFP), showing that the tagged GFP was not cleaved from the fusion protein. (B) Y991A mutation reduced binding between alpha4 and paxillin. CHO-alpha4/GFP and CHO-alpha4Y991A/GFP cells were lysed, immunoprecipitated with an anti-alpha4 antibody, and analyzed by immunoblotting by using anti-paxillin or anti-GFP antibody. Total cell lysate for each cell type was analyzed alongside. Note that the amount of paxillin that coimmunoprecipitated with alpha4Y991A/GFP was significantly less than that with the wild-type alpha4/GFP. (C) Transwell cell migration assay with the membrane coated on both sides with FN. The number of cells migrated per field (n = 3) was determined for each of the triplicate wells for CHO-alpha4/GFP and CHO-alpha4Y991A/GFP cells. Data were means and standard derivations of triplicate experiments (p < 0.001). Note that the Y991A mutation reduced random cell motility. (D) Cells were photographed at 0-, 6-, and 12-h time points. The percentage of wound closure was determined (n = 10), and the mean values and SDs were graphed (p < 0.01). Note that the Y991A mutation enhanced the wound closure rate. We then compared the CHO-alpha4Y991A/GFP and CHO-alpha4/GFP cells by using the scratch-wound assay. To our surprise, all CHO-alpha4Y991A/GFP cell lines, when plated on FN and tested in the scratch-wound assay, had a faster wound closure rate than CHO-alpha4/GFP cells (Figure D). The faster wound closure rate was not due to faster cell proliferation, because CHO-alpha4Y991A/GFP and CHO-alpha4/GFP cells had the same proliferation rates as assayed under the same plating and wounding conditions as in the wound assays . The same results were obtained from two independent cell lines, indicating that the faster wound closure rate was not due to a cloning artifact. This result suggests that the scratch-wound assay and the Transwell assay measure different types of motile activities. Table 3 | Proliferation rates of CHO-alpha4/GFP and CHO-alpha4Y991A/GFP cells in wound assay The numbers represent the mean and standard deviation of the percentages of cells with BrdU incorporation (n = ~500). There are no statistically significant differences between the two cell lines (p > 0.1). We suspected that the faster wound closure rate of the mutant cells was likely due to enhanced fanning activity of the CHO-alpha4Y991A/GFP cells at wound edges. To test this idea, we compared the migratory behaviors of CHO-alpha4/GFP and CHO-alpha4Y991A/GFP cells at wound edges by time-lapse microscopy and found that the CHO-alpha4Y991A/GFP cells indeed displayed a much higher degree of fanning at wound edges (Figure A and Video 10). At the 4-h time point (2 h after starting the time-lapse movie), the percentage of the CHO-alpha4Y991A/GFP cells that exhibited fanning activity exceeded that of CHO-alpha4 and CHO-alpha4/GFP cells by at least 45% . To evaluate the fanning activities more closely, the CHO-alpha4Y991A/GFP and CHO-alpha4/GFP cells at wound edges (n = similar300) were photographed at 0.5-, 1-, 2-, and 3-h time points and scored for the percentage of cells that displayed the fanning behavior (Figure B). We found that at the 0.5-h time point, the CHO-alpha4/GFP cells at wound edges had little fanning activity (<5% fanning cells), whereas at this time point, similar30% fanning CHO-alpha4Y991A/GFP cells was seen at wound edges (Figure B). There was a steady increase of the percentage of fanning cells for both cell types over time. At the 3-h time point, while the percentage of fanning CHO-alpha4/GFP cells remained low (18.6%), that of fanning CHO-alpha4 Y991A/GFP cells reached 53.8%. This result showed that the CHO-alpha4Y991A/GFP cells responded to scratch-wounding faster and fanned earlier than the CHO-alpha4/GFP cells. Figure 7 | Y991A mutation enhances cell fanning and migration into the scratch-wound. Y991A mutation enhances cell fanning and migration into the scratch-wound. (A) CHO-alpha4/GFP and CHO-alpha4Y991A/GFP cells were plated on FN and wounded. Two hours after wounding, the cells were photographed every 2 min for 2 h. Four frames of both cell types at 0-, 40-, 80-, and 120-min time points are shown. Bar, 50 mum. (B) CHO-alpha4/GFP and CHO-alpha4Y991A/GFP cells were plated on FN-coated coverslips and wounded. Percentages of cells at the wound edge (n = similar300) with fanning activity were scored at 0.5-, 1-, 2-, and 3-h time points. To relate the formation of alpha4-positive puncta at the leading edge to the fanning activity, we examined a large number of CHO-alpha4/GFP and CHO-alpha4Y991A/GFP cells at wound edges and scored the percentage of cells with the alpha4-positive puncta (n = similar110) (Figure K). The alpha4-positive puncta were found at the leading edge of both cell types (Figure , E and F), and the cells that displayed the alpha4-positive puncta were largely those that had just begun to fan and migrate into the scratch-wounds, which were frequently found among CHO-a4/GFP and CHO-a4Y991A/GFP cells at the 3-h (Figure C) and 0.5-h (Figure B) time points, respectively. In these cells, the puncta were again located along the leading edge of small, newly formed lamellipodia protrusions (arrows in Figure , E and F). However, the cells that had already formed broad lamellipodia and migrated into the wounds did not display the alpha4-postive puncta at the leading edge (Figure D). Therefore, while the CHO-alpha4Y991A/GFP cells had fanning activity at earlier time points after scratch-wounding than the CHO-alpha4/GFP cells, the onset of puncta formation was also earlier in these cells. To compare the fanning activity and puncta formation more closely, we scored CHO-alpha4/GFP and CHO-alpha4Y991A/GFP cells at wound edges for the percentage of cells with fanning activity (n = similar300) at the 0.5- and 3-h time points after scratch-wounding, by using phase micrographs at a lower magnification, which provided a better view of cell morphology at wound edges (Figure , G --J). The cells with fanning activity were placed in one of two categories: 1) initiating fanning or 2) having formed broad lamellipodia and migrated into the scratch-wounds. We found that the degree to which cells had initiated fanning at the wound edge correlated with the percentage of cells exhibiting alpha4-positive puncta. CHO-alpha4/GFP cells at the 0.5-h time point (Figure G) showed little fanning activity at the wound edges and little evidence of puncta formation (Figure K). However, by the 3-h time point, when some cells began to fan and move into the wound (Figure I), the appearance of puncta increased correspondingly (Figure K). On the other hand, CHO-alpha4Y991A/GFP cells had considerable number of cells initiating fanning at the 0.5-h time point (Figure H), and at this time point a corresponding percentage of the cells had the alpha4-positive puncta (Figure K). By the 3-h time point (Figure J) when the majority of the CHO-alpha4Y991A/GFP cells have formed broad lamellipodia and migrated into the scratch-wounds, the percentage of the cells with alpha4-positive puncta had drastically decreased (Figure K). These results show that there is a close correlation between the formation of alpha4-positive puncta and the initial protrusion of broad lamellipodia into the scratch-wound. Figure 8 | alpha4-Positive puncta are present in cells that begin to protrude broad lamellipodia. alpha4-Positive puncta are present in cells that begin to protrude broad lamellipodia. CHO-alpha4/GFP (A, C, E, G, and I) and CHO-alpha4Y991A/GFP (B, D, F, H, and J) cells were plated on FN-coated coverslips and scratch-wounded. At 0.5- (A, B, F, G, and H) and 3-h (C, D, E, I, and J) time points, cells were fixed and analyzed by fluorescence (A --F) and phase microscopy (G --J). E and F were higher magnification of C and B, respectively. The histogram (K) shows the percentage of cells (n = similar110) with the alpha4-positive puncta, and the percentages of cells (n = similar300) that begin to fan and of those that have migrated into the scratch-wounds. Note that at the 0.5-h time point, CHO-alpha4/GFP cells (A and G) showed little fanning activity, whereas CHO-alpha4Y991A/GFP cells (B and H) had already initiated fanning. By the 3-h time point, although CHO-alpha4/GFP cells (C and I) had just begun to fan, many CHO-alpha4Y991A/GFP cells (D and J) had already formed broad lamellipodia and migrated into the scratch-wounds. The percentage of cells with the presence of alpha4-positive puncta closely correlated with the percentage of cells initiating fanning (K), in which the puncta were located along the leading edge of small lamellipodia protrusions (E and F, arrows). Asterisks (*) in C and B and arrowheads in E and F indicate locations of alpha4-positive puncta. Bar, 10 mum. alpha4beta1 Colocalizes with Paxillin Partially in Leading Edge Ruffles But Is Not Localized in Focal Adhesions and Focal Complexes | To understand how the paxillin-binding regulates the lamellipodia-promoting activity of alpha4beta1, we performed immunofluorescence studies to determine whether alpha4beta1 and paxillin colocalize at the leading edge. It has been shown by others that in the CHO cells that do not express alpha4beta1, paxillin is localized in focal adhesions, focal complexes, and ruffles . In migrating CHO cells, paxillin is recruited into newly formed focal complexes near the leading edge . We found that paxillin was also localized in these areas in CHO-alpha4/GFP cells. (alpha4/GFP)beta1 partially colocalized with paxillin in ruffles (Figure J, arrow). At the leading edge of cells migrating into scratch-wounds, (alpha4/GFP)beta1 was localized in puncta as seen in the time-lapse studies (Figure F, arrows), but these alpha4-positive puncta clearly did not colocalize with the paxillin-positive focal complexes, which are also seen at the leading edge. (alpha4/GFP)beta1 was also absent from most of the paxillin-positive focal adhesions (Figure , C and F). A very small number of paxillin-positive focal adhesions overlapped with some alpha4-positive spots (Figure F, arrowhead), but given that the majority of focal adhesions had no alpha4 staining, we conclude that alpha4beta1 is not localized in focal adhesions. We also examined the CHO-alpha4Y991A/GFP cells by paxillin staining and GFP fluorescence and found that the localization patterns of alpha4 and paxillin were not altered by the Y991A mutation (our unpublished data). Paxillin and the mutant alpha4beta1 were both found in ruffles, although our biochemical data clearly showed that the binding between alpha4beta1 and paxillin was drastically reduced by the Y991A mutation (Figure B). Figure 9 | alpha4beta1 is localized in ruffles but not in paxillin-positive focal complexes/focal adhesions. alpha4beta1 is localized in ruffles but not in paxillin-positive focal complexes/focal adhesions. CHO-alpha4/GFP cells were wounded and allowed to heal for 3 h, after which the cells were fixed, stained with an anti-paxillin antibody, and analyzed by fluorescence (A --F) or confocal (G --J) microscopy. Paxillin (A, D, and G, single exposures) was localized in focal adhesions and ruffles. alpha4/GFP was localized in ruffles and puncta at the leading edge (B, E, and H, single exposures). Paxillin and alpha4/GFP were colocalized in ruffles, giving yellow color in double exposures (pointed by arrowheads in C, I, and J). An edge view of ruffles show that paxillin and alpha4/GFP colocalize partially in the ruffles (J, oblique edge view of I). However, alpha4/GFP did not localize with paxillin-positive focal complexes/focal adhesions (C and F, double exposures), and paxillin did not localize in alpha4-positive puncta (pointed by arrows in F). The area marked between two asterisks (*) in J shows artifact colors of reflection from glass coverslips beneath the cell. Bar, 5 mum. DISCUSSION : Cell migration is an integrated process involving multiple steps, including membrane protrusion, formation of stable attachments near the leading edge of the protrusion, forward locomotion of the cell body, release of adhesions, and cell rear retraction . The assembly and disassembly of focal adhesions play critical roles in the formation and release of stable attachments of the cell to its substratum . The assembly of focal adhesions involves sequential recruitment of adhesion components, including integrins such as alpha5beta1, into nascent focal complexes at the leading edge of membrane protrusions . The integrins in focal complexes and focal adhesions not only provide anchors for the cells to generate motile force but also their adhesive activities, when modulated, can regulate migration speed . Furthermore, when the cell forms membrane protrusions, these protrusions are stabilized by focal complexes that mediate stable cell-substratum adhesion. This relatively stable adhesion allows persistent membrane protrusions but is not required for the initial formation of the protrusions . In this article, we provided evidence for a role of alpha4beta1 integrin in the formation of membrane protrusions that is independent of focal complexes and focal adhesions. We show that alpha4beta1 promotes broad lamellipodia protrusion when ectopically expressed in CHO cells that do not express this integrin endogenously, whereas alpha5beta1 does not have this effect. This protrusion-promoting activity of alpha4beta1 is consistent with an observation that the protrusive activity of T lymphocytes on FN can be inhibited by an anti-alpha4 antibody but not an anti-alpha5 antibody . In migrating cells, while alpha5beta1 is recruited into focal complexes , we show that alpha4beta1 forms transient puncta at the leading edge, which do not colocalize with focal complexes and focal adhesions. It is likely that the alpha4-positive puncta contribute to the lamellipodia protrusion activity of the cells, because the Y991A mutation in the cytoplasmic tail of alpha4 results in an earlier onset of alpha4-positive puncta formation as well as earlier initiation of lamellipodia protrusion in response to scratch-wounding. We found that the alpha4-positive puncta formed along the leading edge of small protrusions before they developed into broad lamellipodia, supporting a role of these puncta in the initiation stage of lamellipodia formation. The dynamic nature of the alpha4-positive puncta strongly suggests that the puncta result from transient clustering of the alpha4beta1 molecules. While alpha4beta1 is sufficient and required for CHO cells to protrude broad lamellipodia in response to scratch-wounding, under the same conditions CHO cells expressing endogenous or exogenous alpha5beta1 only randomly protrude short-lived membrane extensions. This result indicates that alpha5beta1 is not sufficient to promote broad lamellipodia protrusion, but optimal lamellipodia protrusion of CHO-alpha4 cells requires alpha5beta1. This result is consistent with alpha5beta1 playing a role in stabilizing broad lamellipodia after they are formed. We propose that transient clustering of alpha4beta1 molecules at the leading edge mediates strong but transient adhesion of the leading edge membrane to the ECM substrate, whereas the focal complexes evolve into focal adhesions that mediate stable adhesion. Both transient and stable adhesions are required for optimal protrusive activity, but the transient clustering of alpha4beta1 may be the primary adhesive event that initiates broad lamellipodia protrusion in response to scratch-wounding. Lamellipodia are broad membrane extensions of cells comprised of a planar meshwork of actin filaments . The formation of lamellipodia involves proteins that regulate actin dynamics, such as vasodilator-stimulated phosphoprotein , Wiskott-Aldrich syndrome protein , and the Arp2/3 complex , and their upstream regulators, including Rac and its effector p21-activated kinase . Broad lamellipodia protrusion does not necessarily require cell-substratum contacts. When stimulated by soluble chemoattractants, some cell types can protrude broad lamellipodia rapidly in the absence of any contact with the substratum . But lamellipodia protrusion stimulated by mechanical cues does require cell-substratum contact . We speculate that scratch-wounding somehow generated a mechanical cue at the wound edge, which induced clustering of alpha4beta1 molecules at the leading edge of the cells. The clustering event not only provides transient contact of the leading edge to the substratum but also may activate a signaling cascade leading to actin cytoskeletal reorganization and the formation of broad lamellipodia. The cytoplasmic domain of alpha4 can be phosphorylated and may play an active role in these signaling events. Therefore, we propose that while the integrins in focal complexes/focal adhesions mediate cell-substratum adhesion to stabilize protrusions, alpha4beta1 may act at the leading edge as a mediator for sensing migratory cues to initiate polarized protrusions. Besides alpha4beta1, other integrins may play a similar role at the leading edge. For example, activated alphaVbeta3 can be recruited to the leading edge of cells when stimulated with fibrinogen or basic fibroblast growth factor, and this recruitment is required for directional motility . It is conceivable that integrin clustering and activation may play a similar role at the leading edge in regulating polarized membrane protrusions. We show that the lamellipodia-promoting activity of alpha4beta1 is negatively regulated by paxillin. This negative regulation is likely to occur in the ruffles at the leading edge where paxillin and alpha4beta1 colocalize. We propose that the binding between paxillin and the alpha4 tail prevents alpha4beta1 in the ruffles from clustering at the leading edge. This negative regulation may be relieved by an inside-out signaling pathway, which is activated in response to scratch-wounding. We found that, after scratch-wounding, the cells expressing alpha4Y991A began to fan sooner than the cells expressing wild-type alpha4, suggesting that (alpha4Y991A)beta1 may be constitutively active in its lamellipodia-promoting activity due to reduced paxillin binding. Interestingly, the Y991A mutation reduces random motility of the cells in Transwell assays while enhancing lamellipodia protrusion in response to scratch-wounding. Therefore, lamellipodia protrusion and random motility are distinct types of motile activities, which are differentially regulated by interactions between alpha4beta1 and paxillin. Paxillin is a multidomain adaptor protein that is recruited into focal complexes and focal adhesions, where it provides docking sites for cytoskeletal and signaling proteins . Furthermore, paxillin has been shown to play essential roles in cell motility, possibly by recruiting signaling components into focal complexes via the PKL-PIX-PAK complex and by regulating the turnover of focal adhesions via a FAK-mediated pathway when paxillin and FAK interact in focal adhesions . Ginsberg and colleagues show that disrupting the binding between paxillin and alpha4beta1 affects the kinetics of FAK phosphorylation. They propose that the alpha4 tail changes the kinetics of FAK phosphorylation through alpha4/paxillin interactions and promotes random cell motility by allowing a more efficient activation of FAK signaling that is primarily mediated by the integrins in focal adhesions, including alpha5beta1 . Thus, in addition to the cooperative role of alpha4beta1 and alpha5beta1 in promoting lamellipodia protrusion as discussed above, these two integrins may also play a synergistic role in regulating random cell motility, where alpha4beta1 somehow amplifies the alpha5beta1-mediated signaling events although not localized in focal adhesions. In fact, alpha4beta1 can rescue the ability of CHOB2 cells, which lack alpha5beta1, to adhere and migrate on FN . Alternatively, paxillin may play a more active role. It has been reported that overexpression of a paxillin LD4 domain deletion mutant, which disrupted binding of paxillin to PKL and localization of PKL to focal adhesions, caused an increase in cell protrusiveness and random motility but an inhibition of the cells to migrate into scratch-wounds . Therefore, paxillin/PKL binding and paxillin/alpha4beta1 binding seem to have opposite effects on cell motility, suggesting that paxillin may play an active role in achieving a balance between random and polarized lamellipodia protrusive activities. In summary, we show that alpha4beta1 can mediate specific events at the leading edge of migrating cells to initiate the formation of broad lamellipodia and that this lamellipodia-promoting activity of alpha4beta1 is independent of focal complexes/focal adhesions. In addition, we demonstrated that lamellipodia protrusion and random motility can be differentially regulated by interactions between alpha4beta1 and paxillin. These findings provide new insight into how cell migration can be regulated by integrin-mediated cell --ECM interactions. Backmatter: Abbreviations used: : CHO = Chinese hamster ovary ECM = extracellular matrix FN = fibronectin GFP = green fluorescent protein VCAM-1 = vascular cell adhesion molecule-1 PMID- 12221127 TI - Regulation of Airway Tight Junctions by Proinflammatory Cytokines AB - Epithelial tight junctions (TJs) provide an important route for passive electrolyte transport across airway epithelium and provide a barrier to the migration of toxic materials from the lumen to the interstitium. The possibility that TJ function may be perturbed by airway inflammation originated from studies reporting (1) increased levels of the proinflammatory cytokines interleukin-8 (IL-8), tumor necrosis factor alpha (TNF-alpha), interferon gamma (IFN-gamma), and IL-1beta in airway epithelia and secretions from cystic fibrosis (CF) patients and (2) abnormal TJ strands of CF airways as revealed by freeze-fracture electron microscopy. We measured the effects of cytokine exposure of CF and non-CF well-differentiated primary human airway epithelial cells on TJ properties, including transepithelial resistance, paracellular permeability to hydrophilic solutes, and the TJ proteins occludin, claudin-1, claudin-4, junctional adhesion molecule, and ZO-1. We found that whereas IL-1beta treatment led to alterations in TJ ion selectivity, combined treatment of TNF-alpha and IFN-gamma induced profound effects on TJ barrier function, which could be blocked by inhibitors of protein kinase C. CF bronchi in vivo exhibited the same pattern of expression of TJ-associated proteins as cultures exposed in vitro to prolonged exposure to TNF-alpha and IFN-gamma. These data indicate that the TJ of airway epithelia exposed to chronic inflammation may exhibit parallel changes in the barrier function to both solutes and ions. Keywords: INTRODUCTION : Tight junctions (TJs) are characteristically located at the apicolateral borders of adjacent epithelial cells and are responsible for the selective regulation of the passage of ions and neutral molecules through the paracellular space. In addition to their key role in maintaining barrier function of the epithelium, TJs have been implicated in the pathogenesis of several diseases. One group of disorders, inflammatory bowel disease, displays a highly deregulated barrier function that appears to be a fundamental event in disease pathogenesis . Whether the alteration in TJ barrier function is a primary phenomenon associated with the disease or an acquired response caused by the high degree of inflammation associated with the disease is unclear. If, indeed, the TJ response is acquired, this would suggest that other diseases known to be associated with a high level of inflammation may display a characteristic change in TJ barrier function. Cystic fibrosis (CF) serves as an excellent model for the study of the effects of inflammation on TJ barrier function. The airway disease of CF is characterized by an inflammatory response with a marked influx of neutrophils and chronic bacterial infection with Pseudomonas aeruginosa. Proinflammatory cytokines detected in CF airways serve in part to perpetuate this inflammatory response. High levels of the proinflammatory cytokines tumor necrosis factor alpha (TNF-alpha), interleukin (IL)-8, and IL-1beta as well as the soluble intercellular adhesion molecule (ICAM)-1 have been measured in the airways of patients with CF . The airway epithelium of patients with CF also expresses an increased level of signal transducer and activator of transcription-1 (STAT-1), a component of the interferon gamma (IFN-gamma) signaling cascade, indicating that IFN-gamma may also be increased in CF airways . In addition, increased levels of IFN-gamma mRNA have also been detected in the airway epithelium of CF patients . Freeze-fracture electron microscopy (EM) has shown previously that the TJ strands of inflamed CF airways appear to be altered compared with those from non-CF patients . The cytokines IL-1, IL-4, IL-10, IL-13, TNF-alpha, and IFN-gamma have all been shown to regulate the TJ of both epithelia and endothelia . In addition, IL-1beta has been shown to alter TJ permeability through an effect on the claudin family of transmembrane proteins thought to be important in maintaining junctional integrity in astrocytes . However, these studies have not focused on the effects of chronic cytokine exposure on airway epithelial TJ function. The TJ is a complex structure composed of several protein components, some of whose function(s) remain largely unclear. The claudin family of transmembrane proteins are thought to be key components of the TJ, but the growing number of members of this family impedes a complete understanding of their function. Several other components, such as occludin, junctional adhesion molecule (JAM), and more recently the coxsackievirus and adenovirus 2/5 receptor, have been localized to the TJ . Occludin, the first transmembrane TJ protein identified, has been localized to the TJ strand with claudins. JAM has been identified as a component of the TJ necessary to allow the transmigration of monocytes through the paracellular space, which makes it a likely target for inflammation-induced alterations in TJ function. Which of these components is significant in maintaining the integrity of the airway TJ during inflammation is unknown. The regulation of TJ function is incompletely understood and may vary among different cell types. Several protein kinase C (PKC) isoforms have been associated with changes in paracellular permeability. Previous studies have shown a correlation between the level of membrane-associated PKCalpha and the extent of paracellular permeability, and PKCbeta has been shown to play a role in the regulation of endothelial TJs . In addition, the atypical PKC isoforms PKCzeta have been localized to the TJ of MDCK cells . Atypical PKC isotype-specific interacting protein may also localize to the TJ via interactions with PKCiota/lambda and PKCzeta, forming a complex that is tethered to the junction by a direct interaction with JAM . Although PKC isoforms appear to play a functional role in the TJ, factors regulating their expression remain unclear, in particular, how these isoforms are affected by biological stimuli, including inflammation. The difference in the morphology of CF airways as determined by freeze-fracture EM suggests that the environment of the CF airway might lead to disruption of the barrier function of the TJ. To determine whether the chronic inflammation of CF airways leads to modulation of airway TJs, we exposed CF and non-CF well-differentiated (WD) primary human airway epithelial (HAE) cells to cytokines that are upregulated in CF, including IL-1beta, TNF-alpha, and IFN-gamma. We then assessed the effects of this treatment on TJ barrier function and on components of the TJ. To correlate the effects of cytokines on primary HAE cells in vitro with the effects seen in vivo, we performed immunofluorescence localization of TJ components in freshly excised human CF and non-CF large airways. MATERIALS AND METHODS : Chemicals and Antibodies | TNF-alpha and IFN-gamma were purchased from Sigma Chemical Inc. (St. Louis, MO). Rabbit polyclonal antibodies to ZO-1, claudin-1, and occludin and mouse monoclonal antibodies (mAbs) to occludin, ZO-1, and ICAM-1 were purchased from Zymed Laboratories (San Francisco, CA). Rabbit polyclonal and goat polyclonal antibodies to ICAM-1 were purchased from Santa Cruz Biochemical (Santa Cruz, CA). The 3D8 mouse mAb to JAM was kindly provided by Dr. Kenji Ishii (Kyoto, Japan). Rabbit polyclonal antibodies to both PKCiota/lambda and PKCzeta were purchased from Santa Cruz. The mouse anti-claudin-4 mAb was generously provided by James M. Anderson (Yale University, New Haven, CT). The FITC-conjugated dextrans were purchased from Sigma. The PKC inhibitor H7 and the tyrosine kinase inhibitor genistein were purchased from Sigma. Chelerythrine chloride was purchased from Alexis Corp. (San Diego, CA). Cell Culture | Primary airway cells from human subjects were obtained in accordance with guidelines approved by the Committee on the Protection of the Rights of Human Subjects. Bronchial cells of normal (non-CF) and CF type were isolated from surgical specimens, plated at a density of 2 x 105 cells/12-mm Transwell-Col (0.4-mum pore size) insert, and maintained in a 50:50 mixture of LHC Basal Medium (Biofluids, Rockville, MD) and DMEM-H medium supplemented with growth factors, retinoic acid, and BSA as previously described . After cultures reached confluence, medium was aspirated from the apical surface, and cells were maintained at an air-liquid interface for 4 wk. Cultures with 10% cilia as determined by microscopy and a transepithelial resistance (RT) of 600 Omega-cm2 measured with an ohmmeter (EVOM; World Precision Instruments, Sarasota, FL) were selected for experiments. Cultures were exposed to cytokines on the basolateral surface for 24, 48, or 72 h. Transepithelial Cell Permeability | Permeation of FITC-dextrans of 10 and 2000 kDa across primary HAE cells was measured after 24, 48, or 72 h exposure to cytokines. Compounds were added to the apical surface (200 muL of a 5-mg/mL solution in HEPES-buffered Ringer's solution containing 1.3 mM CaCl2), and samples (10 muL) were removed from the apical and basolateral compartments at 10, 20, 30, 40, and 60 min. The rate of permeation was determined by measuring the sample fluorescence at 496 nm in a 96-well fluorescent plate reader. The paracellular permeability to hydrophilic solutes (Papp) coefficients were calculated as previously described . Electrophysiological Measurements of Dilution Potential | Polarized CF primary HAE cultures were treated with IL-1beta for 72 h and mounted in modified Ussing chambers interfaced with an electrometer, in which transepithelial potential difference (VT) and RT were measured continuously under open-circuit conditions. The temperature of all solutions was maintained at 37C, and pH was regulated by bubbling with 95% O2 --5% CO2. Basal VT, RT, and current were recorded across cultures in a buffer containing high NaCl (in mM: 120 NaCl, 10 HEPES at pH 7.4, 10 NaHCO3, 1.2 CaCl2, 5 KCl, and 1 MgSO4) and 10-4 M amiloride. The change in VT (DeltaVT) in response to luminal substitution with a buffer containing low NaCl (in mM: 60 NaCl, 120 mannitol, 10 HEPES at pH 7.4, 10 NaHCO3, 1.2 CaCl2, 5 KCl, and 1 MgSO4) and 10-4 M amiloride was subsequently recorded . Blank filters were used to determine background and subtracted from subsequent measurements. Dilution potentials (DeltaVT) and PCl-/PNa+ were calculated as previously described . All measurements were performed in a minimum of six total cultures isolated from two patients. Immunofluorescence Labeling and Confocal Microscopy | Cells were permeabilized with methanol at -20C for 30 min. Antibodies to ZO-1, occludin, claudin-1, claudin-4, PKCiota/lambda, and JAM diluted to 1:1000 were added to the luminal surface for 1 h. Cells were washed with PBS, and Texas Red --labeled secondary antibodies (Amersham), diluted 1:600 in 10% goat serum/PBS, were added to the luminal surface. For occludin and claudin-1 double labeling, a mouse mAb to occludin conjugated to FITC was incubated with rabbit polyclonal antibody to claudin-1 and incubated for 1 h at room temperature. For JAM and ZO-1 double labeling, 3D8 mouse anti-JAM and rabbit anti-ZO-1 were added to the culture. After washing, anti-rabbit Texas Red and antimouse FITC were incubated in 10% goat serum/PBS for 1 h at room temperature. Cells were postfixed with 4% paraformaldehyde. Transwell-Col inserts were excised and mounted on slides with 100 muL Vectashield (Vector Laboratories, Burlingame, CA) containing 4',6-diamidino-2-phenylindole (DAPI). Images were captured with a confocal laser-scanning microscope (, Exton, PA). Freshly excised human bronchi from non-CF and CF (DeltaF/DeltaF) subjects were embedded in OCT, and frozen sections (10 mum) were cut. Sections were incubated in 95% ethanol followed by a brief wash in PBS. After incubation in acetone, sections were washed in 0.1% Triton X-100 in PBS, and then primary and secondary antibodies added as described above. Western Blotting | Lysates of primary HAE cells were prepared with 0.1% Triton X-100 extraction buffer containing phenylmethanesulfonyl fluoride (PMSF) and dithiothreitol (DTT). Equal amounts of protein (30 mug) were loaded onto Tris-glycine gels (Novex, San Diego, CA). After electrophoresis for 1 h at 200 V, protein was transferred to nitrocellulose or polyvinylidine difluoride (PVDF) membrane at 33 V and blocked in 5% fat-free milk. Membranes were probed at antibody dilutions of 1:1000 in Tris-buffered saline --Tween-20 (TBS-T). Protein was visualized with a peroxidase-conjugated secondary antibody (1:20,000) by enhanced chemiluminescence (Amersham Biosciences, Piscataway, NJ) or Supersignal west femto maximum sensitivity substrate for JAM blotting (, Rockford, IL). For immunoprecipitation of PKCiota/lambda and zeta and JAM, lysates were incubated with rabbit polyclonal antibodies to PKCiota/lambda or zeta or mouse mAb to JAM for 2 h (PKC) or overnight (JAM) at 4C, and Sepharose G beads were added for an additional 1 h. After centrifugation, beads were washed in cell lysis buffer and then heated at 95C for 15 min. Loading buffer was added and Western blotting performed as described above. Blots were stripped with Restore Western blot stripping buffer according to the manufacturer's protocol (Pierce). Human bronchi from non-CF and CF (DeltaF/DeltaF) lungs were dissected immediately after removal from patients undergoing clinical lung transplantation (freshly excised) and placed in ice-cold PBS, and epithelium was removed with a scalpel. Protein was isolated in 7 M urea buffer containing SDS and beta-mercaptoethanol and sonicated for 30 s. Protein concentration was determined with the RC/DC Protein Assay (, Hercules, CA), 50 mug total protein run on Tris-glycine gels and transferred to PVDF membrane at 33 V for 2 h. Western blotting was performed as described above. For Western blotting of JAM, membranes were incubated with 3D8 antibody overnight at 4C. Occludin Western blotting was performed with the rabbit polyclonal antibody as described above. Densitometry was performed with Scion Image for Windows (Scion, Frederick, MD) to determine average band intensity and percent difference between non-CF and CF. The intensity of lanes containing no bands was set to the background level. RT-PCR and Quantitative RT-PCR | Total RNA was isolated with Qiagen RNeasy Protect according to the manufacturer's protocol (Qiagen, Valencia, CA). RNA was then treated with DNase (Ambion, Austin, TX). For complementary DNA synthesis, 1 mug total RNA was used in a 20-muL reaction containing 1 mM deoxynucleotide triphosphates (dNTPs), 2.5 mM random hexamers or oligo dT, 1000 U/ml RNase inhibitor, 0.1 volume 10X buffer (supplied by manufacturer), and 2500 U/ml murine leukemia virus reverse transcriptase (Invitrogen, Carlsbad, CA). The reverse transcription (RT) reaction was carried out at 1 cycle in a thermal cycler at 42C for 50 min, followed by 15 min incubation at 70C. For competitive quantitative RT-PCR (cQRT-PCR), an internal standard was generated containing a 60-nucleotide deletion that would be recognized by the same primers as the target sequence . RNA for this internal standard was generated by ligation of a T7 promoter followed by in vitro transcription. Before RT, the standard was added at a concentration range of 0.0001 to 1 ng, and cDNA was generated. For semiquantitative RT-PCR (QRT-PCR), primers to the gene of interest and those to GAPDH were added simultaneously to the PCR reaction. PCR was carried out with Taq DNA polymerase for 25 cycles. PCR products were separated on a 1.5% agarose gel containing ethidium bromide. The relative band intensities were then determined with Scion Image, and the amount of JAM mRNA expression was calculated as previously described . The amount of standard added to the reaction (in ng) versus the ratio of standard (DeltaJAM) to transcript was plotted on a double logarithmic plot. The amount of transcript was then determined by measuring the point on the plot at which the y-intercept equals 1, indicating an equal ratio of standard to transcript, and then calculating the x-intercept. Enzyme-linked Immunosorbent Assay | Primary HAE cells were treated with cytokines for 24, 48, or 72 h, and total protein was isolated as described above. For ICAM-1, a 96-well assay plate was coated overnight with mouse monoclonal ICAM-1, then lysate was added at 1, 3, and 10 mug and incubated for 2 h. After washing, rabbit polyclonal anti-ICAM-1 was added for 2 h, followed by horseradish peroxidase (HRP)-labeled secondary antibody. For JAM, plates were coated overnight with 1, 3, 10, or 30 mug lysate and incubated with mouse monoclonal anti-JAM (3D8) for 2 h and visualized with antimouse HRP. Protein was detected by addition of a 1:1 ratio of H2O2:tetramethylbenzidine. Optical density was determined at 450 nm. Data are expressed as percent change in JAM or ICAM-1 relative to vehicle controls. Freeze-Fracture EM | Cultures treated with vehicle or cytokines were fixed in 2% glutaraldehyde/2% paraformaldehyde in 0.1 M phosphate buffer (pH 7.2) at 4C overnight. The epithelium was gently removed from the Transwell-Col with a scalpel and rinsed in phosphate buffer containing 0.2 M sucrose at room temperature, followed by a 25% glycerol cryoprotectant solution. The epithelium was sandwiched between gold double-replica mounts and frozen in liquid nitrogen --cooled Freon. Specimens were fractured in a Balzers BAF 400T freeze-fracture plant at a stage temperature of -100C, and replicas were made at a 30o angle with platinum/carbon shadowing. The replicas and adherent tissue were removed by placing in distilled water, followed by transferring to a solution of 5% sodium dichromate in 50% sulfuric acid for cleaning. Replicas were then moved again to distilled water, where they were retrieved and placed onto standard copper microscopy grids. The replicas were examined, and fields exhibiting TJs at a plate magnification of 20,000x were photographed with a Zeiss EM-10A at an accelerating voltage of 60 kV. Morphometric analysis of TJ strands was performed as previously described . Photomicrographs were enlarged to a final magnification of 60,000x, and TJs were transected by lines perpendicular to the luminal border with adjacent transects no closer than 1 mum apart. Strand depth was calculated at the transected lines by measuring the distance from the most luminal strand to the most distal strand. Statistics | Data are presented as mean +- SEM. A one-way analysis of variance (ANOVA) and Bonferroni's correction for multiple comparisons were used to determine statistical significance (p < 0.05). RESULTS : Effect of Cytokine Treatment on RT and Papp | WD primary HAE cells of non-CF and CF types were exposed to cytokines involved in CF inflammation, IL-1beta, IFN-gamma, and TNF-alpha. To determine whether treatment with either cytokine could elicit effects on airway TJ permeability, we exposed the basolateral surface of HAE cells to IL-1beta (100 ng/ml), TNF-alpha (10 ng/ml), or IFN-gamma (100 ng/ml) alone or in combination and determined the effects on RT and Papp at 24, 48, and 72 h. These concentrations were chosen because they are similar to levels detected in the airways of CF patients . After 24 h exposure of non-CF WD HAE to IL-1beta, TNF-alpha, or IFN-gamma alone, there were no significant effects on Rt. However, in cultures exposed to combined treatment of TNF-alpha and IFN-gamma, RT at 24 h decreased to 65% of control (vehicle-treated) cultures (Figure A). After 48 h exposure, cultures exposed to IFN-gamma alone exhibited RT values that were 80% of vehicle cultures, whereas RT in those exposed to TNF-alpha or IL-1beta alone decreased to 60%. The RT of HAE cells treated with a combination of TNF-alpha and IFN-gamma decreased to 30% of control cultures (Figure A). The effect of cytokines on RT was even more dramatic after 72 h exposure, with RT of cultures treated with TNF-alpha and IFN-gamma in combination falling to 8% of control cultures (Figure A), whereas exposure to IL-1beta, TNF-alpha, or IFN-gamma alone decreased RT to similar60% of vehicle controls. Exposure of cultures with TNF-alpha, IFN-gamma, and IL-1beta simultaneously did not alter the kinetics or magnitude of either RT or Papp (data not shown). Upon removal of cytokines at 72 h, RT returned to control levels by 24 h posttreatment (Figure B). Figure 1 | Effect of cytokines on RT. Effect of cytokines on RT. (A) RT of non-CF WD primary HAE cells exposed to TNF-alpha, IFN-gamma, or IL-1beta alone, or TNF-alpha and IFN-gamma in combination. (B) Recovery of RT after removal of TNF-alpha and IFN-gamma at 4, 12, and 24 h after washing. (C) Effect of combined TNF-alpha and IFN-gamma treatment on RT in non-CF and CF primary HAE cells. (D) Effect of IL-1beta on RT in non-CF and CF primary HAE cells. *Significantly different from (A) vehicle and TNF-alpha, IFN-gamma, or IL-1beta alone or from (C) non-CF controls, at p < 0.001. Data presented are a minimum of n = 12 cultures from at least three patients. CF cultures displayed a different kinetic profile of cytokine-induced changes in RT than did non-CF cultures. Whereas the RT of non-CF cultures was only modestly affected by cytokine treatment at 24 h (61% of vehicle), the RT of CF cultures exposed to combined TNF-alpha and IFN-gamma treatment was reduced to 18% of vehicle-treated controls (Figure C). This trend continued after 48 h exposure, with RT in non-CF cultures reduced to 32%, compared with 10% of vehicle controls in CF cultures. However, by 72 h exposure, the RT in both non-CF and CF cultures was reduced to 5 --10% of vehicle controls (Figure C). In contrast, there was no significant difference between non-CF and CF cultures after exposure to IL-1beta at any time point (Figure D). To correlate the changes in RT induced after cytokine exposure to alterations in the barrier function of the TJ, the permeability coefficients of cultures to both a small solute, a 10-kDa FITC-labeled dextran, and a larger solute, 2000-kDa FITC-labeled dextran, were measured after cytokine exposure. In non-CF cultures, there was no difference in the Papp to the 10- and 2000-kDa dextrans by 24 h (Figure , A and B) in cultures treated with TNF-alpha and IFN-gamma in combination. However, pronounced increases in Papp after 48 and 72 h exposure were detected. The Papp to the 10-kDa dextran increased by 8-fold and that to the 2000-kDa 19-fold when treated with TNF-alpha and IFN-gamma in combination for 48 h. By 72 h after treatment, Papp to the 10-kDa dextran was increased 17-fold and that to the 2000-kDa dextran 25-fold (Figure , A and B, respectively). There were no differences in the Papp between cultures treated with vehicle and those treated with IL-1beta alone at any time point (Figure , A and B). There was also no difference in cultures exposed to TNF-alpha or IFN-gamma alone (data not shown). To provide evidence that cytokines were increasing Papp via passive mechanisms typical of the paracellular route, the permeability to solutes from the basolateral to apical compartments were measured and were equal to those from the apical to basolateral compartments (data not shown). Figure 2 | Effect of cytokines on Papp. Effect of cytokines on Papp. Papp to (A) 10-kDa or (B) 2000-kDa FITC-labeled dextran in WD primary HAE exposed to IL-1beta alone or TNF-alpha and IFN-gamma in combination for 24, 48, or 72 h. (C, D) Comparison of the effects of TNF-alpha and IFN-gamma combined treatment on Papp in non-CF vs. CF WD primary HAE cells to (C) 10-kDa or (D) 2000-kDa FITC-labeled dextran. *Significantly different from vehicle (A, B) or from non-CF (C, D) controls at p < 0.05. Data presented are a minimum of n = 12 cultures from at least three patients. The kinetics of the changes in Papp of CF cultures treated with TNF-alpha and IFN-gamma in combination were more significant than those of non-CF cultures. After 24 h exposure, Papp of CF cultures was 10-fold greater than both vehicle-control and non-CF cultures. At 48 h, the Papps of CF cultures were similartwofold greater than those of non-CF cultures. However, like RT, Papp after 72 h exposure was equal between non-CF and CF cultures (Figure C). In addition, Papp to the 2000-kDa dextran was also significantly increased in CF versus non-CF cultures (Figure D). Although there was no increase in Papp in non-CF cultures after 24 h, there was a fivefold increase in CF cultures. In addition, at 48 and 72 h, there is an similar1.5-fold greater increase in Papp of CF cultures than non-CF (Figure D). IL-1beta Effects on Na+ and Cl- Permeability | Because treatment of primary HAE cells with IL-1beta did not induce alterations in Papp to dextrans but did lead to a modest decrease in RT, we evaluated more subtle effects on TJ function by measuring the relative ion selectivity of the paracellular path in CF cultures at 72 h after IL-1beta exposure. The effect of NaCl ion substitution on the transepithelial dilution potentials were determined and used to calculate the ratio of the relative permeabilities of Cl- to Na+ (PCl-/PNa+). To exclude ion permeation via the transcellular pathway, experiments were performed in CF cultures that do not express the CF transmembrane conductance regulator (CFTR) channel on the apical membrane, thereby excluding Cl- permeation via the transcellular route. To block Na+ transcellular permeation, all luminal solutions contained amiloride (10-4 M). Cultures treated with IL-1beta relative to vehicle exhibited a decrease in DeltaVT in response to dilute NaCl solutions from 9.00 +- 0.28 mV (vehicle) to 3.50 +- 0.30 mV (IL-1beta), indicating an effect on ion permeation . This reduction in DeltaVT correlated with an increase in PCl-/PNa+ in cultures exposed to IL-1beta for 72 h, from 0.29 +- 0.03 to 0.73 +- 0.30 . The twofold increase in PCl-/PNa+ in IL-1beta --treated cultures also correlated with a similartwofold increase in conductance (GT), an increase from 1.20 +- 0.20 mS/cm2 (vehicle) to 2.43 +- 0.13 mS/cm2 (IL-1beta). Because IL-1beta led to increased GT and a hyperpolarization of VT while exhibiting a net increase in PCl-/PNa+, it can be concluded that IL-1beta leads to a selective increase in paracellular permeability to Cl-. Figure 3 | Effect of IL-1beta on relative ion selectivity. Effect of IL-1beta on relative ion selectivity. Polarized CF WD primary HAE cells were exposed to IL-1beta for 72 h. Electrophysiological measurements and dilution potential (DeltaVT) were measured and permeability ratio of Cl-/Na+ (PCl-/PNa+) calculated as described in MATERIALS AND METHODS. Shown are the transepithelial voltage recordings with intermittent current pulses in amiloride (Amil)-treated cultures before and after luminal substitution with a low-NaCl solution. Current pulses and voltage scales are the same. Summary data are presented in the table. Data presented are representative of recordings from 12 cultures from 2 patients. *Significantly different from vehicle-treated controls. Cytokines and TJ Morphology | To determine whether the effects of TNF-alpha and IFN-gamma cotreatment on the distribution of TJ proteins also affected the morphology of TJ strands, we performed freeze-fracture EM and quantified the effect of cytokine treatment on junctional depth and strand number. Freeze fracture splits the cell membrane along a hydrophobic core, thereby allowing an en face view of the strands of the TJ. Electron micrographs of freeze-fracture replicas from non-CF cultures treated with TNF-alpha and IFN-gamma for 72 h revealed that treatment decreased overall junctional depth and number of strands (Figure , A and C). In cultures exposed to cytokines, junctional depth was measured at 0.295 +- 0.020 mum, whereas that of vehicle controls was 0.469 +- 0.103 mum as determined by morphometric analysis (Figure C). Strand number was decreased from 6.90 +- 0.372 in vehicle controls to 4.85 +- 0.214 in cultures exposed to TNF-alpha and IFN-gamma. Figure 4 | Morphology of TJs of primary HAE cells exposed to TNF-alpha and IFN-gamma. Morphology of TJs of primary HAE cells exposed to TNF-alpha and IFN-gamma. Freeze-fracture analysis of TJ strands of WD non-CF (A) or CF (B) HAE cells exposed to vehicle or cytokines for the indicated times. (C) Quantification of strand number and depth. Micrographs are representative of 40 mum (vehicle non-CF), 20 mum (vehicle CF), 70 mum (non-CF 72 h), 60 mum (CF 24 h), and 50 mum (CF 72 h) of TJ analyzed. Significantly different from vehicle controls (*), vehicle controls and 24 h exposure (**), and from non-CF 72 h at p < 0.05. Unlike non-CF cultures, which were unaffected by cytokines at 24 h, the effect of TNF-alpha and IFN-gamma on RT of CF cultures was pronounced by 24 h after treatment (Figure C). For this reason, freeze-fracture analysis was also performed on CF cultures at 24 h. After exposure for 24 h, CF cultures displayed a significant change in both depth and strand number, with changes similar to those seen in non-CF cultures after 72 h cytokine exposure (Figure B). By 72 h, the number of strands had decreased to 1.97 +- 0.214 and junctional depth to 0.068 +- 0.007 mum, compared with a strand number of 6.7 +- 0.227 and a depth of 0.501 +- 0.170 mum in vehicle-controls. Although the RT and Papp of CF cultures at 72 h after cytokine exposure do not differ significantly from non-CF, TJ morphology differs significantly, suggesting an increased sensitivity of CF cultures to TNF-alpha and IFN-gamma. Cytokine Effects on the Integrity of TJ-Associated Proteins | To determine whether exposure to TNF-alpha and IFN-gamma disrupted the organization of TJ-associated proteins, we performed immunofluorescence localization of ZO-1, JAM, claudin-1, claudin-4, and occludin. When double-immunofluorescence localization was performed for the cytoplasmic protein ZO-1 and the integral membrane protein JAM, we saw redistribution of both proteins at 48 h and 72 h (Figure A), with less obvious alterations at 24 h (data not shown). The apparent disruption was detected only in cultures treated with the combination of TNF-alpha and IFN-gamma and not in cultures treated with IL-1beta, TNF-alpha, or IFN-gamma alone (data not shown). CF cultures at 24 h exposure to TNF-alpha and IFN-gamma exhibited a similar pattern of reorganization of ZO-1 and JAM as non-CF cultures at 72 h (Figure A). CF cultures exposed to cytokines for 72 h displayed a similar relocalization of ZO-1 and JAM as non-CF at 72 h (data not shown). The apparent change in localization of ZO-1 and JAM shown by immunofluorescence was also associated with a decrease in ZO-1 and JAM expression by Western blot analysis of non-CF cultures exposed to TNF-alpha and IFN-gamma for 72 h (Figure B). Figure 5 | Effect of TNF-alpha and IFN-gamma treatment on expression and localization of JAM and ZO-1. Effect of TNF-alpha and IFN-gamma treatment on expression and localization of JAM and ZO-1. (A) Immunofluorescent staining for JAM and ZO-1 in non-CF and CF primary HAE cells exposed to TNF-alpha and IFN-gamma for 72 or 24 h. Red staining (left) represents ZO-1, green staining (middle) is JAM, and (right) merged image of both ZO-1 and JAM. Areas of colocalization appear as yellow. Images are representative of a minimum of n = 6 cultures from two patients. (B) Western blotting of ZO-1 (top) and immunoprecipitation of JAM (middle) in non-CF cultures. Western blot for ZO-1 was stripped and reprobed with an antibody specific for claudin-1 to control for loading (bottom). However, immunofluorescence localization of claudin-1 showed no apparent change in either the intensity or distribution of fluorescent staining for either protein (Figure A). Occludin exists in both a high-molecular-weight (HMW) and low-molecular-weight (LMW) form. Although immunofluorescence did not reveal a change in the expression or localization of occludin, the antibody used recognized only the LMW form of the protein (Figure A). To reconcile whether there were TNF-alpha and IFN-gamma --induced effects on either occludin expression or phosphorylation, we performed Western blots with two antibodies, one recognizing only the LMW and one recognizing both forms. Samples isolated from cultures exposed to TNF-alpha and IFN-gamma for 72 h and probed with an antibody recognizing both the LMW and HMW forms appeared to express more of the 73-kDa occludin than vehicle-controls (Figure B, lanes 4 --6). However, when the membrane was reprobed with an antibody that recognizes only the LMW form of the protein, there was no change in the 65-kDa occludin band after treatment (Figure C). This finding would indicate that the relative increase seen correlates with a shift from the LMW form to the HMW, indicating an increase in the phosphorylation state of the protein. Although there appeared to be effects on occludin levels, claudin-1 expression remained unchanged (Figure C). The distribution and expression of claudin-4, the other claudin species identified in our culture system, was also unaffected by TNF-alpha and IFN-gamma treatment (Figure , B and C). The finding with regard to claudin-4 was confirmed by comparison of relative fluorescence intensity of immunofluorescence images and by densitometry of Western blots (data not shown). Figure 6 | TNF-alpha and IFN-gamma effect on occludin, claudin-1, and claudin-4 distribution and expression. TNF-alpha and IFN-gamma effect on occludin, claudin-1, and claudin-4 distribution and expression. (A) Immunofluorescent staining for occludin and claudin-1 in primary HAE cells exposed to TNF-alpha and IFN-gamma for 72 h. Occludin staining in green (middle), claudin-1 in red (left), and merged image (right). Colocalization appears as yellow. (B) Immunofluorescence staining of claudin-4 in vehicle-treated cultures and in cultures exposed to TNF-alpha and IFN-gamma for 72 h. (C) Western blot analysis of occludin, claudin-1, and claudin-4. Top, Probed with an antibody recognizing both the HMW and LMW forms of occludin. This blot was then stripped and reprobed with an antibody recognizing only the LMW form of the protein. Claudin-1 and claudin-4 expression are below. Lanes 1 --3, vehicle; 4 --6, exposed to TNF-alpha and IFN-gamma for 72 h. Images are representative of a minimum of n = 6 cultures from two patients. Cytokine Exposure and JAM and ICAM-1 Expression | Redistribution of JAM by cytokines (Figure A) would be expected to decrease monocyte transmigration across the airway, whereas increased ICAM-1 expression would be required for neutrophil transmigration . Because neutrophils are the predominant inflammatory cell in the CF airway, we investigated whether treatment of primary non-CF HAE cells with TNF-alpha and IFN-gamma altered the levels of expression of JAM and ICAM-1. After exposure of cultures for 24, 48, or 72 h with TNF-alpha and IFN-gamma, the level of JAM and ICAM-1 mRNA was determined by cQRT-PCR or semiquantitative RT-PCR (QRT-PCR), respectively, and compared with the expression in untreated cultures. Two methods of mRNA quantification were chosen because more subtle changes were expected in JAM expression level. The level of JAM mRNA decreased to 58 +- 12% of vehicle controls after 72 h cytokine exposure as determined by cQRT-PCR (Figure A). Similar results were found after QRT-PCR (data not shown). In contrast, ICAM-1 mRNA expression increased with increasing duration of exposure to cytokines (Figure B). By 48 and 72 h exposure, the level of ICAM-1 mRNA was significantly greater in cytokine-treated than in control cultures. No ICAM-1 was detectable in vehicle-treated HAE cultures, consistent with previously published data . After 24 h exposure to cytokines, there was a significant increase in ICAM-1 mRNA expression, which increased further to 214 +- 14.5% of the 24-h level by 48 h and 242 +- 8.5% at 72 h. Figure 7 | Coordinate regulation of JAM and ICAM-1 in response to TNF-alpha and IFN-gamma exposure. Coordinate regulation of JAM and ICAM-1 in response to TNF-alpha and IFN-gamma exposure. (A) JAM mRNA expression as determined by cQRT-PCR with an internal standard (DeltaJAM) at the indicated concentration (1 to 0.0001 ng). Left, Vehicle, and right, cultures exposed to TNF-alpha and IFN-gamma for 72 h. (B) Semiquantitative RT-PCR (QRT-PCR) analysis of ICAM-1 mRNA expression normalized to GAPDH. Top, ICAM-1 expression in vehicle or at 24, 48, or 72 h after exposure to TNF-alpha and IFN-gamma. Bottom, GAPDH expression. (C, D) JAM (C) and ICAM-1 (D) protein expression determined by ELISA at 24, 48, and 72 h after exposure to TNF-alpha and IFN-gamma. Data are from a minimum of n = 6 cultures from at least three patients. To correlate the changes in mRNA expression after cytokine exposure with protein levels, we performed an enzyme-linked immunosorbent assay (ELISA) on vehicle and cytokine-treated cultures for 24, 48, or 72 h . The level of JAM expression decreased to 80 +- 10%, 70 +- 10%, and 40 +- 4% of control cultures at 24, 48, and 72 h, respectively. In contrast, ICAM-1 expression increased to 141 +- 9%, 227 +- 7%, and 356 +- 22% of controls at 24, 48, and 72 h, respectively. Thus, TNF-alpha and IFN-gamma downregulate JAM expression while upregulating ICAM-1 expression. Mechanism of Cytokine-induced Alterations in Permeability | Cytokine-induced changes in TJ permeability have previously been linked to alterations in tyrosine kinases and PKC . We examined the ability of a tyrosine kinase inhibitor, genistein, and the nonspecific PKC inhibitor H7 to block the effect of TNF-alpha and IFN-gamma on RT and Papp to small and large solutes, respectively. Cultures were pretreated bilaterally with the inhibitor at concentrations ranging from 1 to 300 muM for 1 h. After this initial incubation, TNF-alpha and IFN-gamma in the presence of the inhibitor were incubated for 72 h, the time point at which the greatest effect of cytokine was observed . Genistein did not inhibit the effect of cytokines on RT or Papp at any concentration (data not shown), whereas H7 (10 and 30 muM) significantly inhibited the effect on both RT and Papp. At 72 h, H7 (10 muM) displayed partial inhibition of the TNF-alpha and IFN-gamma --induced changes in RT (Figure A). Figure 8 | Effect of selective and nonselective inhibitors of PKC on cytokine-induced changes in RT, Papp, and JAM and ICAM-1 expression. Effect of selective and nonselective inhibitors of PKC on cytokine-induced changes in RT, Papp, and JAM and ICAM-1 expression. Cultures were exposed to TNF-alpha and IFN-gamma in the presence or absence of H7 (nonselective) or chelerythrine chloride (selective) inhibitors of PKC. (A) RT and (B) Papp to 10-kDa FITC-dextran. JAM (C) and ICAM-1 (D) protein expression by ELISA in cultures treated with TNF-alpha and IFN-gamma for 72 h in the absence or presence of 10 muM or 30 muM H7. *Significantly different from vehicle at p < 0.05. Data presented are a minimum of n = 12 cultures from at least three patients. We next determined whether H7 would also induce a partial inhibition of the cytokine-mediated increase in Papp by measuring Papp of the 10-kDa FITC-dextran at 72 h after cytokine exposure. Papp in cultures treated in the absence of H7 was increased 17-fold over vehicle controls. In contrast, cultures treated with TNF-alpha and IFN-gamma in the presence of H7 (10 muM) displayed a negligible increase in Papp, with Papp to the FITC-dextran only twofold greater than that of vehicle controls (Figure B). Because the nonspecific inhibitor H7 may also inhibit protein kinase A (PKA), we also evaluated the effect of the specific PKC inhibitor chelerythrine, which inhibits the catalytic domain of all PKC isoforms on RT and Papp. Like H7, chelerythrine (1 muM) inhibited the cytokine-induced decrease in RT and the associated increase in Papp to a similar degree as H7 (Figure , A and B), suggesting that the effects on RT and Papp were mediated through PKC signaling pathways. Effect of H7 on Cytokine-induced Alterations JAM and ICAM-1 Expression | In addition to eliciting a profound effect on RT and Papp, exposure of WD primary HAE cells to combined treatment of TNF-alpha and IFN-gamma caused changes in the expression of JAM and ICAM-1 . To determine whether cotreatment of H7 with cytokines could inhibit the effect on JAM and ICAM-1 protein expression, we performed ELISA analysis after 72 h exposure to TNF-alpha and IFN-gamma in the absence or presence of H7. JAM and ICAM-1 expression were 44 +- 6% and 378 +- 7% of vehicle controls after treatment in the absence of H7 (Figure , C and D). However, JAM expression in cultures treated in the presence of H7, 10 and 30 muM, was 65 +- 20% and 90 +- 16% of control, respectively (Figure C). ICAM-1 expression in the presence of H7 also remained at or near control levels, 189 +- 15% and 102 +- 15% in cultures treated with 10 and 30 muM H7, respectively (Figure D). Effect of PKC Inhibitors on Distribution of TJ-associated Components | The nonspecific (H7) and specific (chelerythrine) PKC inhibitors prevented the cytokine-induced changes in junctional barrier function (as assessed by RT and Papp measurements) and the changes in the reciprocal regulation of JAM and ICAM-1 expression . To determine whether these inhibitors could also inhibit the changes in the distribution of ZO-1 and JAM, which occurred after 72 h TNF-alpha and IFN-gamma treatment, we performed immunofluorescence localization after treatment of cultures in the presence of H7 and chelerythrine. Although a subtle redistribution of JAM and ZO-1 was detected after cytokine exposure in the presence of H7 or chelerythrine, the extent of this relocalization was less pronounced than with cytokine exposure alone . Claudin-1 and occludin distribution was not effected either in the presence or absence of the inhibitors (data not shown). Figure 9 | Effect of H7 and chelerythrine on the reorganization of ZO-1 and JAM. Effect of H7 and chelerythrine on the reorganization of ZO-1 and JAM. Immunofluorescent staining for JAM and ZO-1 in primary HAE cells exposed to TNF-alpha and IFN-gamma for 72 h in the presence or absence of H7 (10 muM) or chelerythrine (10 muM). Red staining (left) represents ZO-1, green staining (middle) is JAM, and right is a merged image of both ZO-1 and JAM. Areas of colocalization appear yellow. Characterization of PKC Isomers Mediating Cytokine-induced Changes in TJ Permeability | Because H7 and chelerythrine inhibited the effects of TNF-alpha and IFN-gamma on RT and Papp, we determined which PKC isoform was involved in the cytokine-mediated alteration of the TJ. We focused on the atypical PKC isoforms iota/lambda and zeta because of their link to the TJ . In particular, because JAM expression is highly regulated by cytokine exposure, we examined whether expression of PKCiota/lambda, an isoform thought to bind directly to JAM, would be altered in the same manner. After 72 h exposure, PKCiota/lambda protein expression was greatly increased in cultures exposed to TNF-alpha and IFN-gamma . Immunofluorescence revealed an increase in PKCiota/lambda expression, correlating with enhanced staining of the nuclei, plasma membrane, and the apical portion of the cytoplasm (Figure A). In addition, immunoprecipitation of PKCiota/lambda and PKCzeta followed by Western blot analysis demonstrated an increase in the PKCiota/lambda isoform after cytokine exposure, but not PKCzeta (Figure B). Treatment of cells with cytokines in the presence of H7, which prevented the changes in barrier function and JAM and ICAM-1 expression (Figures and ), also partially inhibited the increase in PKCiota/lambda at 72 h (Figure B). In contrast, the specific PKC inhibitor chelerythrine, which prevented the cytokine-induced alteration in RT and Papp , did not inhibit the increase in PKCiota/lambda expression (Figure B), suggesting that the reduction in PKC iota/lambda expression by H7 was mediated through PKA. PKA is known to regulate expression of aPKCs (see below) in cells that express adenosine receptors . Figure 10 | Cytokine effects on PKC isomer expression. Cytokine effects on PKC isomer expression. WD primary HAEs were exposed to TNF-alpha and IFN-gamma in the absence or presence of H7, and expression of PKCiota/lambda and PKCzeta was determined. (A) Immunofluorescence localization of PKCiota/lambda after 72 h exposure to TNF-alpha and IFN-gamma, with or without H7. Increased staining in cultures exposed to TNF-alpha and IFN-gamma is evident. (B) Immunoprecipitation followed by Western blot analysis of PKCiota/lambda (top) and PKCzeta (bottom) after 72 h exposure to TNF-alpha and IFN-gamma, with or without H7 (left) or chelerythrine (right). Images and gel are representative of a minimum of n = 6 cultures from at least 3 patients. Expression of Junctional Components and ICAM-1 In Vivo | We have demonstrated that prolonged cytokine treatment induces alteration in the barrier function of the TJ, presumably via PKC-induced changes in protein components of the TJ. Because the airways of CF patients are chronically exposed to an inflammatory milieu, we examined the expression of these same TJ components in large airways excised from non-CF and CF patients after lung transplantation. We found that the changes in ICAM-1, JAM, and ZO-1 induced by cytokine treatment in our in vitro system were also present in vivo. Bronchi from CF patients expressed increased ICAM-1, particularly at the basolateral surface, and decreased ZO-1 and JAM staining as assessed by immunofluorescence . However, no apparent change in the fluorescence intensity of occludin, claudin-1, or claudin-4 was detected. Figure 11 | In vivo localization of ICAM-1 and junctional components in freshly excised non-CF and CF (DeltaF/DeltaF) bronchus. In vivo localization of ICAM-1 and junctional components in freshly excised non-CF and CF (DeltaF/DeltaF) bronchus. Bronchi from non-CF and CF patients were isolated and prepared for staining as described in MATERIALS AND METHODS. Frozen sections were stained for ICAM-1, ZO-1, JAM, claudin-1, occludin, and claudin-4. Blue, DAPI-stained nuclei and red, positive protein staining. Images are representative of sections from a total of four patients of non-CF and CF (DeltaF/DeltaF) type each. Top, Hematoxylin and eosin (H+E) staining of a representative section of non-CF or CF type to illustrate section morphology. To further quantify these changes, we performed Western blot analysis of protein isolated from bronchi of non-CF donors and CF patients after transplant. We found that, as occurs with cytokine exposure in our in vitro system, bronchial epithelia from CF patients in vivo exhibit a decrease in ZO-1 and JAM and a marked increase in ICAM-1 and occludin . CF bronchial epithelia in vivo expressed 46 +- 14% less ZO-1 and 54 +- 19% less JAM than non-CF donors. In contrast, CF bronchial epithelia in vivo demonstrated a 196 +- 18% increase in ICAM-1 expression over non-CF. These data are similar to those of previously published reports measuring ICAM-1 expression in CF and non-CF epithelium . Similarly, occludin Western blotting showed an increase of 234 +- 36% over non-CF donor epithelium. When the polyclonal antibody to occludin was used in Western blotting, only a single band was detected in the range of 65 --73 kDa. There was no significant change in the expression of either claudin-1 or -4 in CF epithelium (97 +- 14% and 96 +- 6% of non-CF, respectively), consistent with our in vitro data. Figure 12 | In vivo expression of ICAM-1 and junctional components in non-CF and CF (DeltaF/DeltaF) freshly excised bronchus. In vivo expression of ICAM-1 and junctional components in non-CF and CF (DeltaF/DeltaF) freshly excised bronchus. Total protein isolated from the bronchi of non-CF and CF patients after transplant was subjected to Western blot analysis of ICAM-1 and TJ components as described in MATERIALS AND METHODS. Band intensity was analyzed and expression in CF patients presented as a percentage of non-CF (bottom). *Significantly different from non-CF (p < 0.05). DISCUSSION : We hypothesized that the chronic inflammation in the CF lung may lead to changes in the integrity of the TJ in CF patients, leading to decreased barrier function and alterations in ion selectivity. To address this hypothesis, we exposed non-CF and CF WD primary HAE cultures to cytokines known to be upregulated in CF. We found that exposure of WD primary HAE to TNF-alpha and IFN-gamma combined, but not separately, induced significant changes in the barrier function and rearrangement of structural components of the TJ. In addition, it appears that these cytokine-induced changes may be regulated via a signaling cascade involving upregulation of PKCiota/lambda, an atypical PKC isoform. Prolonged exposure of the airway to a combination of TNF-alpha and IFN-gamma led to significant changes in TJ barrier properties, with significant alterations in both RT and Papp (Figures and ). However, there was a less pronounced effect when cultures were treated with either of these cytokines alone or with IL-1beta. In addition, the kinetics of the effects of TNF-alpha and IFN-gamma on RT and Papp were different in CF WD primary HAE cells from those in non-CF controls. Although cultures of the non-CF type were resistant to the effects of cytokine treatment until 48 h of exposure, CF cultures showed a significant decrease in RT and an increase in Papp by 24 h (Figure , C and D), indicating some properties of the CF epithelium make it respond more rapidly to the effects of cytokines on the TJ. Perhaps primary cells isolated from CF airways are already sensitized to the effects of cytokines because of retention of changes in gene expression induced by the inflammatory environment present in vivo. Our data measuring the effects of cytokines on RT and Papp for noncharged solutes and dilution potentials allowed us to identify differences in TJ responses to individual cytokines. For example, IL-1beta treatment modestly decreased RT, while exhibiting no effect on Papp to small or large noncharged solutes, suggesting that the barrier function of the TJ to noncharged solutes was unchanged, whereas the ion permeability of the junction was altered. Measurements of the effects of 72 h IL-1beta treatment on GT and transepithelial dilution potentials suggested that the increase in GT (or reduction in RT) could be accounted for solely by the increase in relative permeability for Cl- across the paracellular pathway . Previously published data have established a pattern of ion transport defects in CF epithelium both in vitro and in vivo (for review, see , ). Although CF epithelia are impermeable to Cl- ions via the transcellular pathway, because of the CFTR defect, they exhibit an increased rate of Cl- absorption in vivo, which would reflect increased permeability of the paracellular pathway to Cl- ions in CF . Because IL-1beta levels have been shown to be markedly increased in CF sputum, our data showing increased PCl-/PNa+ ratio induced by IL-1beta may in part explain some of the enhanced absorption of Cl- ions in CF epithelia. Isotonic transcellular hyperabsorption of sodium (Na+), which occurs because of absent or defective apical membrane CFTR Cl- channels that are responsible for inhibiting epithelial sodium channels, is accompanied by hyperabsorption of Cl- via the paracellular path in CF epithelia and has been proposed as a mechanism by which defective cellular Cl- transport leads to decreased airway surface liquid height and volume, decreased mucociliary clearance, bacterial infection, and bronchiectasis in CF . This finding might also suggest the induction of a claudin with selective Cl- permeability, which is in contrast to the recently reported decreased Na+ permeability induced by expression of claudin-4 . In contrast, the large changes in RT and in large solute permeability (Papp) induced by TNF-alpha and IFN-gamma suggest that these cytokines are acting via a distinct pathway that either degrades the TJ barrier function greatly or alters the properties of a claudin member with very large equivalent pore radii. The claudin family of transmembrane proteins play a critical role in maintaining TJ integrity and may be responsible for the selective passage of ions and molecules through the paracellular space (for review, see ; ). However, claudin-1 and claudin-4 do not appear to be affected by the exposure of primary HAE cells to TNF-alpha and IFN-gamma , suggesting that claudin-1 and -4 are not involved in the regulation of the TJ mediated by cytokines in the airway epithelium. Freeze-fracture EM revealed that as a correlate to altered barrier properties of the TJ to ions and solutes, there was also an effect on the morphology of the TJ strands. Non-CF cultures exposed to cytokines had a decrease in the extent of their junctional depth and a modest decrease in strand number (Figure A). These data are consistent with previously published data suggesting that cytokine exposure altered TJ ultrastructure in intestinal cell lines . In contrast to non-CF cultures, those of the CF type appear to exhibit a unique kinetic profile, as evidenced by a significant decrease in RT and Papp after 24 h TNF-alpha and IFN-gamma treatment (Figure , C and D). The more rapid effect of cytokines on CF cultures was also seen in freeze-fracture EM at 24 and 72 h after treatment. At 24 h after treatment, the depth and strand number of CF strands resembled those of the non-CF type exposed to cytokines for 72 h. Similarly, JAM and ZO-1 redistribution in CF cultures at 24 h was similar to that of non-CF cultures at 72 h . To determine whether TNF-alpha and IFN-gamma exposure of primary HAE cells altered the organization of TJ components, we performed QRT-PCR, Western blotting, ELISA, and immunofluorescence studies. We found that several components of the TJ are affected by cytokine exposure. In particular, ZO-1 and JAM were downregulated and redistributed after cytokine exposure . Previous studies of the effects of cytokines on proteins involved in the maintenance of TJ integrity have focused primarily on cell lines of intestinal origin. Previous work in HT-29/B6 cells, a human intestinal cell line, suggested that the effects of TNF-alpha on the TJ are mediated via downregulation of occludin . Occludin exists in both an LMW and an HMW form, with the HMW form corresponding to an increased state of phosphorylation. In our study, the expression of the LMW form of occludin remained unchanged, whereas the expression of the HMW form increased, indicating a role for occludin phosphorylation in the cytokine-mediated effects on the TJ, perhaps as an attempt to reseal leaky junctions . JAM is a member of the immunoglobulin superfamily localized to the TJ and may play a key role in monocyte transmigration across the paracellular pathway . Therefore, redistribution from the TJ and decreased expression of JAM after TNF-alpha and IFN-gamma exposure may suggest a decreased ability of monocytes to penetrate to the lumen and perpetuate the inflammatory response. Because CF airways are filled with neutrophils, we postulated that reciprocal regulation of JAM and ICAM-1 might occur with cytokine treatment. Chronic TNF-alpha and IFN-gamma treatment increased ICAM-1 expression at both the mRNA and protein levels, with a relative decrease in JAM . These data indicated a shift in the epithelium chronically exposed to TNF-alpha and IFN-gamma, from that which allows the migration of monocytes to one that is much more prone to the passage of neutrophils via the paracellular pathway. This observation was supported by published data showing increased levels of neutrophils and soluble ICAM-1 in CF sputum, consistent with a predominant influx of neutrophils through the paracellular pathway into the lumen . The in vivo results presented here in CF freshly excised airways are consistent with this hypothesis (Figures and ). The effect of TNF-alpha and IFN-gamma on the airway TJ appears to reflect widespread effects on several protein components. To investigate which signaling pathways activated by the combined treatment of TNF-alpha and IFN-gamma may be relevant to TJ function, we tested agents reported to inhibit two pathways known to regulate TJ permeability, tyrosine kinases and PKC. The tyrosine kinase inhibitor genistein was ineffective in blocking the cytokine-mediated effect on TJ barrier function. In contrast, the PKC inhibitors H7 and chelerythrine significantly inhibited the increase in Papp to both intermediate- and large-sized molecules induced by cytokines . However, the inhibition of the decrease in RT was less complete, indicating that whereas PKC is involved in the pathway regulating the cytokine-induced effects on permeability to solutes, it may have limited effects on ionic permeability. These data support a model of cytokine regulation of TJ barrier function in which there are distinct properties for the regulation of permeability to solutes and ions, one involving PKC. The family of aPKC kinases consists of two distinct members, PKCiota/lambda and PKCzeta, that are insensitive to Ca2+, diacylglycerol, and phorbol esters, which are known to be potent activators of the calcium-dependent (cPKCs) and novel (nPKCs) subfamilies but are instead regulated by lipid cofactors . We focused on the effects of TNF-alpha and IFN-gamma on the expression of both PKCiota/lambda and PKCzeta, because of its reported interaction with JAM and our data indicating that cytokine exposure affected JAM localization and expression. After cytokine treatment, there was an increase in the expression of PKCiota/lambda, as revealed by Western blot analysis and immunofluorescence, while PKCzeta remained constant . Of note, the diffuse cytoplasmic staining for PKCiota/lambda in the apical region of the cells is consistent with studies of PKCiota/lambda in MDCK cells . In addition, this increase was blocked when cells were cotreated with H7, presumably because of the chronic exposure of the epithelium to H7 for 72 h. Because H7 also inhibits PKA, we also examined the effect of chelerythrine on the increase in PKCiota/lambda expression. Whereas H7 partially inhibited the increase in expression of PKCiota/lambda, chelerythrine did not, presumably indicating the inhibition of PKA by H7 upstream of PKCiota/lambda. This conclusion is supported by published data implicating a role for PKA-mediated regulation of aPKCs by adenosine receptors , which are known to be expressed in HAE cells. The aPKC isoform PKCiota/lambda has also been identified as a potential regulator of apoptosis in several cell types, and inhibition of this isoform may be necessary for the induction of apoptosis . We did detect a modest increase in lactate dehydrogenase (LDH) levels after treatment with TNF-alpha and IFN-gamma in combination (data not shown), consistent with the possible apoptotic events. However, the reversible effect of TNF-alpha and IFN-gamma treatment, with RT levels returning to normal levels within 24 h (Figure B) and specific effects on TJ proteins, would suggest a role for PKCiota/lambda independent of apoptosis in the cytokine-induced alteration at the TJ, perhaps via its interactions with JAM. In addition, LDH levels were similar in cultures treated with TNF-alpha and IFN-gamma alone to those treated with TNF-alpha and IFN-gamma in combination, although only combined treatment elicits effects on junctional components (data not shown). In summary, we examined the effects on TJ structure and function of exposure of primary HAE cultures to cytokines known to be present in CF. Chronic exposure of TNF-alpha and IFN-gamma caused significant effects on both RT and Papp. The mechanism by which TNF-alpha and IFN-gamma altered TJ permeability may involve signaling via PKCiota/lambda, because inhibitors of PKC, H7 and chelerythrine, inhibited the effects of cytokines on RT and Papp and prevented the increase in expression of ICAM-1 and decrease in JAM. Although the majority of our studies were conducted in a model system of WD polarized airway epithelium, changes in expression of JAM, ZO-1, and ICAM-1 induced by cytokines correlated with the expression of these proteins in chronically inflamed excised CF airways. In addition, the distinct effects induced by TNF-alpha and IFN-gamma synergistically compared with IL-1beta suggest that epithelium exposed to a mixture of cytokines, as is the CF airway, may exhibit extreme alterations in the barrier of the TJ to solutes and ion species. In our study, the cytokine-induced alteration in the TJ may help explain several associated findings in CF, namely, increased Cl- absorption through the paracellular pathway in response to accelerated Na+ absorption and the ability to increase neutrophil transmigration caused by increased ICAM-1 levels. With regard to other inflammatory diseases, the pattern of cytokine-induced alteration of TJs may similarly contribute to the severity and type of inflammation observed. Given the ability of H7 and chelerythrine to inhibit TJ effects on permeability and ICAM-1 expression, a clearer understanding of cytokine effects on TJs may therefore permit the ability to pharmacologically modulate the severity of chronic inflammation. Backmatter: PMID- 12221128 TI - Centrosomal Proteins CG-NAP and Kendrin Provide Microtubule Nucleation Sites by Anchoring gamma-Tubulin Ring Complex AB - Microtubule assembly is initiated by the gamma-tubulin ring complex (gamma-TuRC). In yeast, the microtubule is nucleated from gamma-TuRC anchored to the amino-terminus of the spindle pole body component Spc110p, which interacts with calmodulin (Cmd1p) at the carboxy-terminus. However, mammalian protein that anchors gamma-TuRC remains to be elucidated. A giant coiled-coil protein, CG-NAP (centrosome and Golgi localized PKN-associated protein), was localized to the centrosome via the carboxyl-terminal region. This region was found to interact with calmodulin by yeast two-hybrid screening, and it shares high homology with the carboxyl-terminal region of another centrosomal coiled-coil protein, kendrin. The amino-terminal region of either CG-NAP or kendrin indirectly associated with gamma-tubulin through binding with gamma-tubulin complex protein 2 (GCP2) and/or GCP3. Furthermore, endogenous CG-NAP and kendrin were coimmunoprecipitated with each other and with endogenous GCP2 and gamma-tubulin, suggesting that CG-NAP and kendrin form complexes and interact with gamma-TuRC in vivo. These proteins were localized to the center of microtubule asters nucleated from isolated centrosomes. Pretreatment of the centrosomes by antibody to CG-NAP or kendrin moderately inhibited the microtubule nucleation; moreover, the combination of these antibodies resulted in stronger inhibition. These results imply that CG-NAP and kendrin provide sites for microtubule nucleation in the mammalian centrosome by anchoring gamma-TuRC. Keywords: INTRODUCTION : The microtubule fulfills essential functions in chromosome segregation in mitosis and in determining organelle positioning and cell polarity. Nucleation of microtubules is initiated by the gamma-tubulin ring complex (gamma-TuRC) at microtubule organizing centers, such as the centrosome of mammalian cells or the spindle pole body (SPB) of yeast. gamma-TuRC is a ring-shaped multiprotein complex containing gamma-tubulin, which is found in the cytoplasm as well as in the centrosome . The cytoplasmic pool of gamma-TuRC may be a source of nucleating complexes that are recruited to the centrosome when increased microtubule nucleation is required . Budding yeasts have a simple gamma-TuRC consisting of three proteins, Spc97p, Spc98p, and Tub4p , which are conserved among other organisms . In mammalian cells, at least 6 proteins have been identified in gamma-TuRC, including GCP2, GCP3 (also named HsSpc98), and gamma-tubulin, which are the orthologues of Spc97p, Spc98p, and Tub4p, respectively . Little is known about how gamma-TuRCs are anchored at the centrosome in mammalian cells. Mammalian centrosome is composed of a pair of centrioles surrounded by an electron-dense cloud of pericentriolar material (PCM) . PCM seems to be an interconnected meshwork of proteins that forms a matrix or lattice composed of a high proportion of coiled-coil proteins. The microtubule nucleating capacity of the centrosome is localized to the PCM . In budding yeast, Spc72p in outer plaque and Spc110p in inner plaque have been identified as mediating microtubule nucleation by anchoring gamma-TuRC . Spc110p is a coiled-coil protein and anchors gamma-TuRC through interaction with Spc97p and Spc98p at the amino-terminal region. The carboxyl-terminal region associates with Cmd1p (calmodulin) that is integrated into the SPB near the nuclear envelope . Several groups have attempted to identify anchoring proteins for gamma-TuRC in other organisms. Human and Xenopus laevis Spc110p-related proteins were found by cross-reactivity of monoclonal antibodies to Spc110p . Mouse pericentrin and gamma-tubulin are found in a protein complex and organized into a lattice at the centrosome . Recently, a larger isoform of pericentrin, human kendrin (or pericentrin B) was found to share homology with the Cmd1p-binding domain of the yeast Spc110p and thus seems to be the mammalian orthologue of Spc110p . However, it is not known whether these candidates can mediate microtubule nucleation or bind to the gamma-TuRC. We have reported that a giant coiled-coil protein, CG-NAP (centrosome and Golgi localized PKN-associated protein), is localized to the centrosome throughout the cell cycle and the Golgi apparatus at interphase . CG-NAP shares partial homology with pericentrin . Its localization to the centrosome is independent of microtubules, whereas that to the Golgi apparatus is disrupted by microtubule depolymerization. Furthermore, CG-NAP associates with protein kinases (PKN, PKA, and PKCepsilon) and protein phosphatases (PP1 and PP2A) . CG-NAP may thus coordinate the location and activity of these enzymes to regulate the phosphorylation states of specific substrates at the centrosome and the Golgi apparatus. In this study, we demonstrate that CG-NAP (also named AKAP350 and AKAP450 ) and kendrin anchor gamma-TuRC through binding with GCP2 and/or GCP3 at their amino-terminal regions. CG-NAP and kendrin are localized to the centrosome via their carboxyl-terminal regions, which interact with calmodulin. Furthermore, endogenous CG-NAP and kendrin form complexes together with GCP2 and gamma-tubulin in vivo. Microtubule aster formation from isolated centrosomes is suppressed by pretreatment with antibodies to CG-NAP and/or kendrin. These observations imply that CG-NAP and kendrin provide microtubule nucleation sites by anchoring gamma-TuRC in mammalian centrosome. MATERIALS AND METHODS : Cell Culture and Transfection | COS7, HeLa S3, and HEK293T cells were grown in DMEM supplemented with 10% heat-inactivated fetal bovine serum, 50 U/ml penicillin, and 50 mug/ml streptomycin at 37C in a humidified 5% CO2 atmosphere. For Chinese hamster ovary (CHO)-K1 cells, Ham's F12 medium was used as a basal medium. COS7 cells were transfected with expression plasmid(s) by electroporation using GenePulser II (, Hercules, CA). HEK293T cells were transfected by use of LipofectAMINE PLUS Reagent (Invitrogen, Carlsbad, CA). Preparation of Full-Length cDNA of Human Kendrin, GCP2, GCP3, and gamma-Tubulin | The full-length cDNA sequence of kendrin was found in GenBank with accession number . KIAA0402 encoding the carboxyl-terminal half of kendrin (amino acids 1511 to the stop codon) was obtained from Kazusa DNA Research Institute Foundation (Kazusa, Japan). The amino acid sequence encoded by KIAA0402 was slightly different from that of kendrin : it lacks amino acids 2828 --2906, and the carboxyl-terminal 13 amino acids were replaced with 16 amino acids. The rest of the kendrin cDNA was obtained by PCR using the HeLa Marathon-ready cDNA library (, Palo Alto, CA) with appropriate oligonucleotide primer sets for the cDNA fragments encoding amino acids 1 --744 and 745-1845 (see Figure A). These fragments were assembled into pBluescript to create the full-length cDNA of kendrin encoding 3246 amino acids. Expression plasmids for the full-length and deletion mutants of kendrin were constructed by subcloning the corresponding fragments into pTB701-HA (hemagglutinin) or pTB701-FLAG . Full-length cDNAs for human GCP2 and GCP3 were obtained by PCR using the HeLa Marathon-ready cDNA library with appropriate oligonucleotide primer sets. Then the cDNA fragments were cloned into pTB701-HA or pTB701-FLAG. gamma-Tubulin cDNA was obtained by PCR using the same library with a primer set designed to create the XhoI site in place of the stop codon, and then cloned into pcDNA3.1-MycHis (Invitrogen) to express gamma-tubulin carboxyl-terminally tagged with Myc and His6. Antibodies | Polyclonal antibodies to human CG-NAP (alphaEE and alphaBH) were described previously . Immunoprecipitation of endogenous CG-NAP was performed by combination of alphaEE and alphaBH. Immunostaining and immunoblotting were done with alphaEE. Polyclonal antibodies to rat CG-NAP (alpharXN), human kendrin (alphaKen), and human GCP2 (alphaGCP2) were generated by immunizing rabbits with bacterially expressed glutathione S-transferase (GST)-tagged antigens prepared as follows. A partial rat CG-NAP cDNA fragment (1.34 kb) was cloned by hybridization screening of rat brain cDNA library by using the human CG-NAP cDNA fragment #2 --43 as a probe. The cDNA sequence was submitted to DDBJ/GenBank/EMBL Data Bank with accession number . The nucleotide number 757-1341 was subcloned into pGEX4T (Amersham Biosciences, Piscataway, NJ) to express GST-tagged antigen. The cDNA fragments encoding amino acids 744 --909 of kendrin and 762 --902 of GCP2 were cloned into pGEX4T. GST-tagged antigens were expressed in Escherichia coli and purified by use of glutathione-Sepharose beads (Amersham Biosciences) as described previously . Antibodies were affinity-purified by use of antigen-coupled Sepharose beads according to the manufacturer's instruction (Amersham Biosciences). The following antibodies were purchased: anti-gamma-tubulin GTU88, anti-alpha-tubulin DM1A, and anti-FLAG M2 (Sigma, St. Louis, MO); rat anti-HA 3F10 (Roche Diagnostics, Basel, Switzerland); anti-His6 antibody RGS-His (Qiagen, Hilden, Germany); rhodamine-conjugated anti-rabbit IgG and FITC-conjugated anti-mouse IgG (Chemicon International, Temecula, CA); and horseradish peroxidase (HRP)-conjugated anti-Myc and HRP-conjugated secondary antibodies (Santa Cruz Biotechnology, Santa Cruz, CA). Immunofluorescence Microscopy | Cells grown on cover glasses were fixed with cold methanol for 5 min with or without prior extraction by 0.1% Triton X-100 in 80 mM 1,4-piperazinediethanesulfonic acid at pH 6.9, 5 mM EDTA, and 1 mM MgCl2 at room temperature for 2 min. The fixed cells were blocked with 5% normal donkey serum in phosphate-buffered saline-containing Triton X-100 (PBST) (20 mM phosphate buffer at pH 7.5, 150 mM NaCl, and 0.03% Triton X-100) and then incubated with the relevant antibody for 1 h at room temperature. After the cells had been washed with PBST, the primary antibody was visualized by subsequent incubation with the appropriate secondary antibody conjugated with either rhodamine or FITC. The fluorescence of rhodamine and FITC was observed under a fluorescence microscope (Zeiss, Oberkochen, Germany) equipped with a CCD camera (Hamamatsu Photonics, Hamamatsu, Japan). Yeast Two-Hybrid Screening and Interaction Analysis | The cDNA fragment encoding CG-NAP3510 --3828 was fused to the Gal4 DNA binding domain (Gal4bd) by subcloning into pGCKT7 . The resultant plasmid was transformed to the yeast reporter strain AH109 together with a HeLa cDNA library fused to the Gal4 activation domain (Gal4ad) in pGAD GH . Screening was done as described previously . GST Pull-down Assay | Bacterial expression plasmids for GST-tagged CG-NAP3510 --3829 and His6-tagged calmodulin 2 were constructed by inserting the corresponding cDNA fragments into pGEX4T and pRSET (Invitrogen), respectively. Tagged proteins were purified as described previously . The His6-tagged calmodulin 2 was incubated with the GST-tagged CG-NAP3510 --3829 in a buffer containing 50 mM Tris-HCl at pH 7.5, 1% NP-40, 150 mM NaCl, 1 mM EDTA, 1 mM dithiothreitol (DTT), 2 mM MgCl2, 10 mug/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride (PMSF) at 4C for 1 h. Then, glutathione-Sepharose 4B beads (Amersham Biosciences) were added and incubated for 30 min. After the resin had been washed with the same buffer, the bound proteins were analyzed by immunoblotting. Immunoprecipitation and Immunoblotting | Cells were lysed with buffer containing 50 mM Tris-HCl at pH 7.5, 1% NP-40, 150 mM NaCl, 0.5 mM EDTA, 0.5 mM EGTA, 1 mM DTT, 1.5 mM MgCl2, 1 mM PMSF, 20 mug/ml aprotinin, and 10 mug/ml leupeptin. After centrifugation, cell extracts were saved and incubated with an appropriate antibody at 4C for 2 h. Then protein A --Sepharose beads (Amersham Biosciences) were added and incubated for 30 min. The resin was washed with the same buffer, and the bound proteins were processed for immunoblotting. Isolation of Centrosomes and In Vitro Microtubule Nucleation | Centrosomes were isolated from CHO cells according to the method of with minor modifications. In brief, CHO cells were incubated with 10 mug/ml nocodazole and 5 mug/ml cytochalacin B for 2 h, rinsed with isolation buffer (1 mM Tris-HCl at pH 8.0, 0.5 mM EDTA, and 0.1% beta-mercaptoethanol), and then lysed by swaying the dishes in the isolation buffer containing 0.5% NP-40 at 4C for 10 min. The lysates were spun at 2500 x g for 10 min to remove cell debris and nuclei. The resultant supernatant was subjected to the discontinuous sucrose density gradient with 60 and 40% (wt/wt) solutions from the bottom and spun at 120,000 x g for 1 h. Fractions were collected from the bottom and stored at -80C. For immunoblotting, fractions were eightfold diluted with 10 mM PIPES at pH 6.9 followed by centrifugation at 18,500 x g for 10 min to precipitate centrosomes, and then the precipitates were dissolved in SDS sample buffer. The microtubule nucleating activity of the isolated centrosomes was tested according to . The centrosome suspension was incubated with 2.5 mg/ml bovine tubulin (Cytoskeleton) and 1 mM GTP at 37C for 4 --8 min. After glutaraldehyde fixation and sedimentation on glass coverslips, the microtubules were visualized by use of an anti-alpha-tubulin antibody. To test the effect of antibodies on microtubule nucleation, the centrosome suspension was preincubated with affinity-purified antibodies for 30 min at 4C. Then, a nucleation reaction was done for 4 min. RESULTS : CG-NAP Is Localized to the Centrosome via the Carboxyl-Terminal Region | A schematic representation of the centrosomal proteins CG-NAP, kendrin, and their deletion mutants used in this study is shown in Figure . In the text, deletion mutants are designated as CG-NAPX-X or kendrinX-X, where X-X represents amino acid residues. We first analyzed subcellular localization of various deletion mutants of CG-NAP expressed in COS7 cells. Most of the deletions were distributed in the cytosol (for instance, Figure Ab). On the other hand, the carboxyl-terminal fragment CG-NAP2875 --3899 and a further deletion CG-NAP3510 --3828 were well colocalized with gamma-tubulin at the centrosomes (Figure A, c, e, and f, respectively). Moreover, CG-NAP1 --2876 lacking the carboxyl-terminal region distributed diffusely at perinuclear area that was not colocalized with gamma-tubulin (Figure 2Ad). These results indicate that the carboxyl-terminal region containing the amino acid residues 3510 --3810 is responsible for the centrosomal localization of CG-NAP. BLAST search of CG-NAP3510 --3828 revealed that this region shares high homology with the carboxyl-terminal region of another centrosomal coiled-coil protein kendrin (Figure B). CG-NAP and kendrin have three coiled-coil regions flanked by noncoiled regions . BLAST search using full-length CG-NAP yielded kendrin at two regions with relatively high homology (Figure , shaded areas), and the carboxyl-terminal part contained the sequence shown in Figure B. Figure 1 | Schematic representation of CG-NAP, kendrin, and their deletion mutants. Schematic representation of CG-NAP, kendrin, and their deletion mutants. Schematic structure of CG-NAP and kendrin are shown with predicted coiled-coil regions in shaded boxes. Positions of the deletion mutants of CG-NAP and kendrin are shown with amino acid residues on the upper and lower sides, respectively. Shaded areas between CG-NAP and kendrin represent the regions sharing homology found by BLAST search. Total amino acid residue of kendrin used in this study was 3246, as described in MATERIALS AND METHODS, which is shorter than that (3321) deposited to GenBank with accession number . Figure 2 | Centrosomal localization of the carboxyl-terminal region of CG-NAP. Centrosomal localization of the carboxyl-terminal region of CG-NAP. (A) Subcellular localization of deletion mutants of CG-NAP. HA-tagged full-length and deletion mutants of CG-NAP were transiently expressed in COS7 cells, and then the cells were fixed with methanol directly (a --d) or after brief extraction with detergent to visualize proteins associated with intracellular structures (e, f). Then, the cells were double-stained with anti-HA and anti-gamma-tubulin (alpha-gammaTub). Bar, 10 mum. (B) Sequence homology of the centrosomal-localization region of CG-NAP with kendrin. Aligned sequences are the result of BLAST search with CG-NAP3510 --3828. The amino acid residues of kendrin shown are of and are different from those of our kendrin construct. Possible calmodulin binding sequence of kendrin homologous to that of yeast Spc110p is underlined. The Carboxyl-Terminal Region of CG-NAP Associates with Calmodulin | To search for the proteins interacting with the centrosomal-localization region of CG-NAP, we used yeast two-hybrid screening of a HeLa cDNA library using CG-NAP3510 --3828 as bait. Among similar1500 clones obtained, most of the clones carried calmodulin 2 and calmodulin 3 cDNAs. Calmodulin 2 and 3 have identical amino acid sequences, although they are coded by distinct genes . These clones contained varying lengths of 5'- and 3'-noncoding sequences, indicating that they were independent clones. We confirmed the interaction between CG-NAP3510 --3828 and calmodulin by different combinations in a yeast two-hybrid system (Figure A). Furthermore, GST pull-down assay revealed that CG-NAP3510 --3828 interacts directly with calmodulin in a Ca2+-dependent manner (Figure B). However, exogenously expressed CG-NAP3510 --3828 in COS7 cells was coimmunoprecipitated with calmodulin in either the presence or absence of Ca2+ (Figure C). A mobility shift of calmodulin on electrophoresis confirmed that the Ca2+-dependent conformational change occurred under these experimental conditions. Figure 3 | Association of the centrosomal-localization region CG-NAP3510 --3828 with calmodulin. Association of the centrosomal-localization region CG-NAP3510 --3828 with calmodulin. (A) Yeast two-hybrid analysis of interaction between CG-NAP3510 --3828 and calmodulin 2. Interaction of the proteins fused to the Gal4 DNA binding domain (BD) and activation domain (AD) was assessed by growth and development of blue color of the transfected yeasts. Combinations of BD and AD constructs are shown on the right. p53 and SV40 large T antigen were used as controls. (B) Direct and Ca2+-dependent binding of CG-NAP3510 --3828 with calmodulin. Bacterially expressed His6-tagged calmodulin 2 and GST-tagged CG-NAP3510 --3828 were mixed and incubated in the presence of 2 mM CaCl2 or EGTA. After removal of aliquots (Input), glutathione-Sepharose beads were added to the mixture and incubated further, and then the proteins bound to the beads were collected (Output) and immunoblotted with anti-His. Black and white arrowheads indicate the positions of Ca2+-unbound and -bound forms of calmodulin, respectively. (C) Ca2+-independent coimmunoprecipitation of calmodulin with CG-NAP3510 --3828. HA-tagged calmodulin 2 and FLAG-tagged CG-NAP3510 --3828 were coexpressed in COS7 cells, and then cell extracts (Extr) were prepared in the presence of 2 mM CaCl2 or EGTA. The extracts were immunoprecipitated with anti-FLAG (alphaFL) or control (Ctr) mouse IgG followed by immunoblot with anti-HA (top) or anti-FLAG (bottom). Black and white arrowheads indicate the positions of calcium-unbound and -bound forms of calmodulin, respectively. The Amino-Terminal Region of CG-NAP Associates with gamma-TuRC through Interaction with GCP2 and/or GCP3 | During the course of this study, kendrin was proposed to be a human orthologue of yeast Spc110p, because kendrin has a calmodulin-binding sequence homologous to that of Spc110p . A similar sequence was also found in CG-NAP3510 --3828 (Figure B), which indeed interacted with calmodulin . Therefore, CG-NAP may function as a mammalian orthologue of Spc110p and anchor gamma-TuRC. To test this possibility, we first carried out an immunoprecipitation study of endogenous proteins. Endogenous gamma-tubulin was coimmunoprecipitated with CG-NAP by anti-CG-NAP antibody from HeLa cell extracts. However, gamma-tubulin was not coimmunoprecipitated with any of the deletion mutants of CG-NAP expressed in COS7 cells (our unpublished results). Spc110p indirectly associates with gamma-tubulin through binding with Spc97p and Spc98p, the components of gamma-TuRC . Mammalian orthologues of Spc97p and Spc98p were identified as GCP2 and GCP3 (also named HsSpc98p), respectively . We thus examined whether CG-NAP binds with these proteins. GCP2 was efficiently coimmunoprecipitated with the amino-terminal region of CG-NAP, CG-NAP16 --1229, and weakly with CG-NAP1229 --1917 (Figure A). The binding property of GCP3 to CG-NAP was similar to that of GCP2 (our unpublished results). We next examined whether CG-NAP16 --1229 interacts with gamma-tubulin through binding with GCP2 or GCP3 (Figure B). As expected, gamma-tubulin was coimmunoprecipitated with CG-NAP16 --1229 only when GCP2 or GCP3 was coexpressed (Figure B, lanes 3, 6, and 9). The binding affinity of GCP3 to CG-NAP16 --1229 seemed to be lower than that of GCP2. GCP3 might indirectly associate with CG-NAP16 --1229 through interaction with GCP2, because GCP3 was coimmunoprecipitated with GCP2 (our unpublished results). These results indicate that CG-NAP anchors gamma-TuRC through binding with GCP2 and/or GCP3. Figure 4 | Association of CG-NAP16 --1229 with gamma-tubulin through binding with GCP2/GCP3. Association of CG-NAP16 --1229 with gamma-tubulin through binding with GCP2/GCP3. (A) Association of GCP2 with deletion mutants of CG-NAP. FLAG-tagged (FL) deletion mutants of CG-NAP were expressed in COS7 cells together with HA-tagged GCP2. Then the cell extracts were immunoprecipitated (IP) with anti-FLAG followed by immunoblot (IB) with anti-HA. (B) Indirect association of gamma-tubulin (gammaTub) with CG-NAP16 --1229 through binding with GCP2 or GCP3. COS7 cells were transfected with various combinations of expression plasmids as shown at the top. Then the cell extracts were immunoprecipitated with anti-FLAG or control mouse IgG followed by immunoblot with anti-Myc for gamma-tubulin (top), anti-HA for GCP2 or GCP3 (middle), or anti-FLAG for CG-NAP16 --1229 (bottom). The Amino-Terminal Region of Kendrin Also Associates with gamma-TuRC through Interaction with GCP2 | Although kendrin was proposed to be a mammalian orthologue of Spc110p, its interaction with gamma-TuRC has not been presented. We thus asked whether kendrin also interacts with gamma-TuRC in a manner similar to CG-NAP. To do this, we prepared full-length cDNA of kendrin and specific antibody to kendrin (alphaKen) as described in MATERIALS AND METHODS (Figure A). HA-tagged kendrin expressed in COS7 cells and endogenous kendrin in HeLa or CHO cells were well recognized by alphaKen in immunoblotting and immunoprecipitation (Figure B). The mobility of these bands may agree with the deduced molecular mass of similar370 kDa. We next examined the subcellular localization of kendrin. alphaKen gave centrosomal staining in HeLa cells (Figure C, a and b) and other cells such as CHO (our unpublished results). HA-tagged full-length kendrin was also localized to the centrosome (Figure D). The carboxyl-terminal fragment of kendrin, kendrin2136 --3246, was localized to the centrosome and associated with calmodulin in a manner similar to CG-NAP2875 --3899 (our unpublished results). Figure 5 | Centrosomal localization of endogenous and recombinant full-length kendrin. Centrosomal localization of endogenous and recombinant full-length kendrin. (A) Preparation of full-length cDNA of kendrin. Schematic representation of kendrin is shown with the fragments used to construct full-length cDNA as described in MATERIALS AND METHODS. Positions are shown with the corresponding amino acid residues. KIAA0402 was obtained from Kazusa DNA Research Institute. Position of the fragment to generate specific antibody is shown as Antigen. (B) Specificity of anti-kendrin (alphaKen) antibody. Cell extracts of COS7 cells expressing HA-tagged kendrin, HeLa, or CHO cells were immunoprecipitated with alphaKen followed by immunoblotting with anti-HA (lanes 1 --3) or alphaKen (lanes 4 --10). (C) Centrosomal localization of endogenous kendrin. HeLa cells were fixed with MeOH, and then double-stained with alphaKen and anti-gamma-tubulin (a, b) or anti-CG-NAP (alphaEE) and anti-gamma-tubulin (c, d). (D) Centrosomal localization of recombinant kendrin. COS7 cells expressing HA-tagged kendrin were fixed with MeOH, and then double-stained with anti-HA and anti-gamma-tubulin. Bars, 10 mum. The amino-terminal fragment of kendrin, kendrin1 --1189, associated with GCP2, and with gamma-tubulin when GCP2 was coexpressed (Figure , lanes 3 and 6). However, association of kendrin1 --1189 with GCP3 was very weak, and coexpression of GCP3 had little effect on the association with gamma-tubulin (Figure , lane 9). These results indicate that kendrin also associates with gamma-TuRC through binding with GCP2, although it remains unclear whether GCP3 is involved in this association. Figure 6 | Association of kendrin with gamma-tubulin (gammaTub) through binding with GCP2. Association of kendrin with gamma-tubulin (gammaTub) through binding with GCP2. COS7 cells were transfected with various combinations of expression plasmids as shown at the top. Then the cell extracts (Extr) were immunoprecipitated with anti-FLAG (alphaFL) or control (Ctr) mouse IgG followed by immunoblot with anti-Myc for gamma-tubulin (top), anti-HA for GCP2 or GCP3 (middle), or anti-FLAG for kendrin1 --1189 (bottom). Endogenous CG-NAP and Kendrin Form Complexes and Associate with GCP2 and gamma-Tubulin | CG-NAP and kendrin were both localized to the centrosome (Figure C) and have coiled-coil regions that are thought to mediate protein-protein interaction. We thus examined whether CG-NAP interacts with kendrin. Endogenous kendrin was coimmunoprecipitated with endogenous CG-NAP from HeLa cell extracts by anti-CG-NAP antibody (Figure A, lane 3), and vice versa (Figure A, lane 4). The interaction was also observed between exogenously expressed CG-NAP and kendrin (Figure B). These results indicate that CG-NAP and kendrin are associated in vivo at the centrosome. Figure 7 | Association of endogenous kendrin, CG-NAP, gamma-tubulin, and GCP2. Association of endogenous kendrin, CG-NAP, gamma-tubulin, and GCP2. (A) Association of endogenous CG-NAP and kendrin. HeLa cell extracts (Extr) were immunoprecipitated with anti-CG-NAP (alphaEE+alphaBH) or anti-kendrin (alphaKen) or control (Ctr) rabbit IgG followed by immunoblot with anti-CG-NAP (alphaEE) (top) or alphaKen (bottom). (B) Association of recombinant kendrin with CG-NAP. 293T cells were cotransfected with FLAG-tagged (FL) CG-NAP and HA-tagged kendrin. Then the cell extracts were immunoprecipitated with anti-FLAG or control mouse IgG followed by immunoblot with anti-FLAG (lanes 1 --3) or anti-HA (lanes 4 --6). (C) Specificity of anti-GCP2 antibody. The extracts of COS7 cells transfected with HA-tagged GCP2 or HeLa cells were immunoprecipitated with alphaGCP2 or control rabbit IgG followed by immunoblot with alphaGCP2 (lanes 1 --3) or anti-HA (lanes 4 --6). (D) Association of GCP2 with CG-NAP and kendrin as well as with gamma-tubulin. HeLa cell extracts were immunoprecipitated with alphaGCP2 or control rabbit IgG followed by immunoblot with anti-CG-NAP (alphaEE), alphaKen, alphaGCP2, or anti-gamma-tubulin. To examine whether CG-NAP and kendrin form complexes with gamma-TuRC in vivo, we performed an immunoprecipitation study using anti-GCP2 antibody (alphaGCP2). alphaGCP2 was prepared as described in MATERIALS AND METHODS and was confirmed to recognize and immunoprecipitate endogenous GCP2 (Figure C). Endogenous CG-NAP and kendrin were significantly coimmunoprecipitated with GCP2 together with gamma-tubulin (Figure D). These data suggest that CG-NAP and kendrin form complexes and associate with gamma-TuRC in vivo. Antibodies to CG-NAP and Kendrin Inhibit Microtubule Nucleation from Isolated Centrosomes | To examine whether CG-NAP and kendrin were present in isolated centrosomes, we fractionated the lysates of nocodazole/cytochalacin B --treated CHO cells by sucrose density gradient centrifugation. Fractions were subjected to immunoblotting and assayed for microtubule nucleation in vitro. gamma-Tubulin was enriched in the fractions 10 --12 corresponding to the sucrose densities of between 40 and 60% (Figure A). CG-NAP and kendrin were cosedimented with gamma-tubulin in these fractions (Figure A). Microtubule asters were efficiently assembled in vitro by using the fractions 10 and 11, and CG-NAP and kendrin were detected in the center of microtubule asters (Figure B). Figure 8 | Inhibition of microtubule nucleation from isolated centrosomes by antibodies to CG-NAP and kendrin. Inhibition of microtubule nucleation from isolated centrosomes by antibodies to CG-NAP and kendrin. (A) Presence of CG-NAP and kendrin in the fractions enriched with centrosomes. CHO cell lysates were fractionated by sucrose density gradient as described in MATERIALS AND METHODS. Then the fractions were examined for the presence of gamma-tubulin (left), CG-NAP, and kendrin (right) by immunoblotting. Fractions 10 --12 corresponding to the interphase between 40 and 60% sucrose were enriched with centrosomes. (B) Localization of CG-NAP and kendrin at the center of microtubule asters formed in vitro. Centrosome fractions were incubated with bovine tubulin in the presence of 1 mM GTP for 8 min at 37C as described in MATERIALS AND METHODS. The resultant microtubule asters were fixed with glutaraldehyde and spun down on coverslips. Then the coverslips were processed for double-staining with anti-CG-NAP (alphaEE) and anti-alpha-tubulin (a --c), or alphaKen and anti-alpha-tubulin (d --f). Bar, 10 mum. (C) Inhibition of microtubule nucleation by pretreatment of the centrosomes with antibodies to CG-NAP and kendrin. Centrosome fractions were pretreated with antibodies to CG-NAP (alpharXN) or kendrin (alphaKen) at a concentration of 1.2 mg/ml or with the combination of alpharXN and alphaKen (1.2 mg/ml each), or control rabbit IgG (2.4 mg/ml) for 30 min on ice. Then microtubule nucleation was performed by incubation with bovine tubulin for 4 min. Microtubule asters were visualized by immunostaining with anti-alpha-tubulin. Bars, 10 mum. The results shown are representative of five different experiments. Next, we examined the effect of antibodies to CG-NAP and kendrin on the microtubule nucleation. The centrosome fractions were preincubated with alphaCG-NAP and/or alphaKen or control IgG and then assayed for the microtubule nucleation in vitro. Normal rabbit IgG had no effect on asters formation (Figure 8Ca), whereas alphaCG-NAP and alphaKen moderately reduced the number of microtubules nucleated from the centrosomes (Figure C, b and c, respectively). Furthermore, combination of the two antibodies inhibited the microtubule nucleation more efficiently (Figure 8Cd). It is of interest that microtubules escaping from this inhibition had almost the same length as those in the control. These results suggest that antibodies to CG-NAP and kendrin inhibit the initiation of nucleation but not the elongation process of microtubules. DISCUSSION : Several candidates have been proposed as mammalian anchoring proteins for gamma-TuRC at the centrosome, such as Ctr100, pericentrin, and kendrin . However, none of them have been demonstrated to interact with components of gamma-TuRC or to mediate microtubule nucleation, although pericentrin was found in the complex containing gamma-tubulin . In this study, we have shown that the amino-terminal regions of CG-NAP and kendrin indirectly associated with gamma-tubulin through interaction with other components of gamma-TuRC, GCP2/GCP3 and GCP2, respectively. Interaction among endogenous proteins was also detected by immunoprecipitation study , although it remains to be established that CG-NAP and kendrin constitute primary anchoring sites for gamma-TuRC in vivo. Furthermore, pretreatment of isolated centrosomes by antibodies to CG-NAP and/or kendrin suppressed the initiation of microtubule nucleation (Figure C), suggesting the involvement of CG-NAP and kendrin in microtubule nucleation from the centrosome. The antigens used to generate these antibodies are located in the amino-terminal region that interacted with gamma-TuRC; thus, the epitopes might be close to or overlap with the binding sites for gamma-TuRC. It was revealed that the amount of gamma-tubulin was decreased in the centrosomes treated with antibodies to CG-NAP and/or kendrin by immunoblotting analysis (our unpublished results). Therefore, these antibodies suppressed microtubule nucleation, probably by displacement of gamma-TuRC from CG-NAP and/or kendrin in the centrosome. The majority of microtubule nucleation appears to be limited to the centrosome despite the presence of substantial amounts of cytoplasmic gamma-TuRC. Therefore, on recruitment to CG-NAP and/or kendrin in the centrosome, gamma-TuRC might represent the active form, or activation of gamma-TuRC might be prerequisite for the recruitment. Additional studies will be necessary to address each of these possibilities. CG-NAP and kendrin were localized to the centrosome via their carboxyl-terminal regions, which were found to associate with calmodulin. The role of calmodulin in the targeting of CG-NAP and kendrin to the centrosome remains unclear. It was reported that green fluorescent protein --tagged calmodulin is localized to the centrosome only at mitotic phase in HeLa cells . We could also detect only a small amount of calmodulin in the centrosomal fractions and almost no specific staining of endogenous calmodulin at the centrosomes (our unpublished results). Centrosomal targeting of CG-NAP and kendrin might be mediated by calmodulin at mitotic phase and by some undefined protein(s) at interphase. Another possibility is that calmodulin may serve to chaperone the carboxyl-terminal region of CG-NAP and kendrin, similar to the role of Cmd1p in the proper assembly of Spc110p at SPB , and be released after centrosomal targeting, as discussed by . It is also unclear whether calmodulin binding to these proteins is regulated by Ca2+: it was Ca2+-dependent in vitro (Figure B) but Ca2+-independent in the immunoprecipitation study (Figure C). The interaction between these proteins might be modified by the presence of some other protein(s) under cellular conditions. In addition to the calmodulin-binding region, the centrosomal-localization regions contain several conserved sequences between CG-NAP and kendrin (Figure B), which may provide interaction sites for other proteins. Our screening using CG-NAP3510 --3828 as bait thus far has yielded only calmodulin clones, which might be attributed to the abundance of calmodulin mRNA in mammalian cells. Further screening by using different bait constructs may be one approach to assess these possibilities. CG-NAP and kendrin have coiled-coil regions that may form a filamentous complex. Indeed, endogenously and exogenously expressed CG-NAP and kendrin formed complexes (Figure , A and B). Moreover, certain combinations of CG-NAP and kendrin deletions were coimmunoprecipitated (our unpublished results), suggesting that the interaction between CG-NAP and kendrin is direct rather than indirect as components of a large protein complex. Heterodimers (or oligomers) of CG-NAP and kendrin may serve as components of PCM and provide anchoring sites for gamma-TuRC. CG-NAP also forms a homodimer (or oligomer) ; thus, CG-NAP homodimers (or oligomers) may also provide the anchoring sites. Do CG-NAP and kendrin play distinct roles in the centrosome? In yeast, gamma-TuRC anchoring proteins Spc110p and Spc72p play independent roles by their different subcellular localization, inner plaque and outer plaque, respectively. Although CG-NAP and kendrin represent differences in terms of the additional localization of CG-NAP in the Golgi apparatus, they are localized to the centrosome in a very similar manner (Figure C). Some difference was observed in the efficiency of binding with gamma-TuRC. CG-NAP16 --1229 associated with GCP2 and gamma-tubulin more efficiently than kendrin1 --1189 (compare Figures B and B), which might be attributed to the relatively low homology between these regions . We may not be able to conclude that CG-NAP has a higher affinity to gamma-TuRC, because these results were obtained with deletion constructs of CG-NAP and kendrin and from cells at random stages of the cell cycle. Cell cycle --dependent phosphorylation and regulation of yeast SPB proteins, such as Spc110p and Spc98p , have been demonstrated. Similarly, centrosomal proteins may be regulated by phosphorylation, for instance, at mitotic phase. From interphase to metaphase, the amount of gamma-tubulin at the centrosome increases at least threefold and decreases rapidly by late anaphase . The recruitment of gamma-TuRC to the centrosome may contribute to the increase in microtubule number associated with mitotic versus interphase centrosomes . It is attractive to postulate that mitotic phosphorylation of CG-NAP or kendrin or GCP2/GCP3 alters affinity among these proteins or to the component(s) of PCM, which may result in increased recruitment of gamma-TuRC to the centrosomes. We have found that CG-NAP interacts with Rho-activated protein kinase PKN , PKA, PKCepsilon, and protein phosphatases PP1 and PP2A, and thus, CG-NAP may target them to the proximity of specific substrates at the centrosome and the Golgi apparatus . It is possible that the complex containing CG-NAP, kendrin, and gamma-TuRC serves as a substrate for these enzymes and is regulated downstream of various signals such as Rho and cAMP. Kendrin also anchors PKA . CG-NAP and kendrin may form matrix as integral components of PCM to provide anchoring sites for gamma-TuRC as well as serving as targeting machinery for various signaling enzymes to the centrosome. Further studies will be necessary to elucidate the role of CG-NAP and kendrin in the regulation of recruitment and activity of gamma-TuRC at the centrosome. Backmatter: Abbreviations used: : SPB = spindle pole body gamma-TuRC = gamma-tubulin ring complex GCP = gamma-tubulin complex protein PCM = pericentriolar material PKA = protein kinase A HA = hemagglutinin PMID- 12221129 TI - DNA Repair and Transcriptional Effects of Mutations in TFIIH in Drosophila Development AB - Mutations in XPB and XPD TFIIH helicases have been related with three hereditary human disorders: xeroderma pigmentosum, Cockayne syndrome, and trichothiodystrophy. The dual role of TFIIH in DNA repair and transcription makes it difficult to discern which of the mutant TFIIH phenotypes is due to defects in any of these different processes. We used haywire (hay), the Drosophila XPB homolog, to dissect this problem. Our results show that when hay dosage is affected, the fly shows defects in structures that require high levels of transcription. We found a genetic interaction between hay and cdk7, and we propose that some of these phenotypes are due to transcriptional deficiencies. We also found more apoptotic cells in imaginal discs and in the CNS of hay mutant flies than in wild-type flies. Because this abnormal level of apoptosis was not detected in cdk7 flies, this phenotype could be related to defects in DNA repair. In addition the apoptosis induced by p53 Drosophila homolog (Dmp53) is suppressed in heterozygous hay flies. Keywords: INTRODUCTION : The TFIIH DNA repair/transcription factor provides an outstanding example of the complexity encountered in genotype --phenotype relationships. Mutations in some of the TFIIH components in humans may produce three hereditary disorders: xeroderma pigmentosum (XP), Cockayne syndrome (CS), and trichothiodystrophy (TTD) . XP patients present sunlight hypersensitivity, abnormal skin pigmentation, and a high skin cancer predisposition . CS individuals have slow postnatal growth and defects in nervous system development . On the other hand, TTD patients share some of the neurological problems present in CS and have the particular phenotypes of brittle hair, fragile nails and ichthyosis . In addition defects in TFIIH may have a role in the generation of cancer (; Liu et al., 2000). TFIIH takes part in nucleotide excision repair (NER) (; ; ; Drapkin and Reinberg, 1999; ; ). In eukaryotes, NER repairs many types of lesions that cause a distortion of the DNA helical structure, including pyrimidine dimers . It has also been reported that TFIIH may participate in base excision repair (BER) when DNA suffers oxidative damage , increasing the number of roles that TFIIH plays in the eukaryotic genome maintenance. TFIIH is formed by the DNA helicases XPB and XPD; the p62, p52, p44, and p34 polypeptides; and the complex known as cyclin-dependent kinase (Cdk)-activating kinase or CAK, which is formed by three proteins: Cdk7, CycH, and Mat1. Cdk7, CycH, and Mat1 are not involved in DNA repair, and there are no reported syndromes related to defects in these genes . The Cdk7 kinase phosphorylates the C-terminal domain of RNA polymerase II. This phosphorylation is necessary for RNA polymerase II elongation . It has been suggested that the Drosophila Cdk7 homolog may also have a role in cell cycle control . Therefore, the central role of TFIIH factor in transcription, DNA repair, and probably in cell cycle, explains the extremely pleiotropic phenotypes observed in XP, TTD, and CS disorders. Functional analysis of TFIIH has been done using yeast, human cells, and in vitro transcription/DNA repair systems. More recently, the use of transgenic mice has allowed the generation of a TTD mouse model by introducing a mutation found in a human TTD patient into the mouse XPD gene . The mouse carrying the TTD allele has shown clearly TTD phenotypes. Unfortunately, this is the only example of a mammalian model that reproduces some manifestations found in humans affected in TFIIH . On the other hand, experiments in Drosophila have demonstrated that the fly is an excellent model for understanding and assaying the developmental function of genes that encode for the TFIIH complex components . In Drosophila the XPB homolog was identified as the haywire (hay) gene . Alleles of hay mimic in the fly some of the defects found in XP and CS . In this work we identified some phenotypes caused by defects either in transcription or by deficient DNA repair in hay flies, showing that the complex phenotype --genotype relationship of TFIIH genes in Drosophila can be genetically analyzed in detail. Our results show that when TFIIH is not functional during development, a high degree of apoptosis is induced. We also found that there is a genetic interaction between Dmp53 and hay similar to the one found in human cells affected in XPB. MATERIALS AND METHODS : Drosophila Strains | Wild-type strain in all experiments was OreR. hay alleles used in this work were haync2, haync2rv1 --4, and haync2rv8 reported by . Two new alleles, hayXPCS and hayTTD, were constructed. Deficiency Df(3L)lxd6 uncovers the hay gene . cdk7P140S is a transgenic conditional negative dominant allele. Therefore, all the crosses with this allele were performed at 29C. The genotype of the cdk7 stock is as follows: w Df(1)JB254 Pw+ [snf+, dhd+]/w Df(1)JB254 Pw+[snf+, dhd+]; +/+; Pw+[Dmcdk7P140S] Sb/TM3 Ser; Df(1)JB254 uncovers the 4F1-2 region of the X chromosome . For p53 experiments the stocks used were w118; Dmp53/CyO , w1118; Dmp53R155H; Dmp53H159N; Dmp53K259H; and Dmp53C+ and MS1096 as a GAL4 wing imaginal disk driver located in the X chromosome . Phenotypic Analyses of Wings, Cuticles, and Bristles | Wings were dehydrated and dissected in ethanol before being mounted in Permount (Fisher Scientific, Pittsburgh, PA) and then visualized with an optic microscope. The cuticle phenotypes were observed with a stereoscopic microscope; subsequently, abdominal regions of adult flies were dissected and prepared for electron microscopy as described previously and were examined in an EM900 transmission electron microscope (Carl Zeiss, Thornwood, NY). Bristle defects were visualized in a stereoscopic microscope. Flies were fixed in glutaraldehyde, postfixed in osmium tetroxide, dehydrated through a graded series of alcohol, and critical point dried before mounting on stubs and coated with carbon and gold. Samples were analyzed with a 5410 LV scanning electron microscope (, Tokyo, Japan). Transgenic Flies | Mutations in the hay cDNA were introduced using a polymerase chain reaction-based method . For the generation of the hayTTD allele a base substitution (A-C) in position 385 on the hay cDNA was made using the oligonucleotide 5'-AGTACAAACTCCCCGCATACAGTTTATATG-3'. To generate a change in the reading frame hayXPCS was made by introducing a guanine in position 2293. The oligonucleotide used was 5'-CCGACACGTCTCACCGATGCCGCCCG-3'. Mutant cDNAs were cloned by using the appropriate restriction enzymes in pCaSper hsp83 vector, and the whole genes were sequenced to corroborate the mutations. Transgenic flies were constructed following a standard protocol . Two hayTTD and four hayXPCS independent lines in the second chromosome were isolated. Staining of Imaginal Discs and CNS | Third instar imaginal discs and larval CNSs were dissected in 1x phosphate-buffered saline and stained with acridine orange vital dye (Sigma-Aldrich, St. Louis, MO) or Nile blue (Sigma-Aldrich) at a concentration of 5 and 100 mug/ml, respectively . Samples were analyzed with a conventional fluorescence microscope. For irradiated larvae, the imaginal discs were dissected and stained 24 h after irradiation. Terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay was performed using the In Situ Cell Death Detection kit from Roche Applied Science (Indianapolis, IN) following the recommended protocol. The tissue was visualized on an MRC-600 confocal microscope (, Hercules, CA). Slot Blot Hybridization | Total RNA from adult flies was purified using the TRIzol kit (Invitrogen, Carlsbad, CA) following the manufacturer's protocol. The integrity of the RNA was verified in a typical formamide-agarose gel. The RNA was loaded in Hybond-N+ membranes (Amersham Biosciences, Piscataway, NJ) by using a slot blot manifold (Hoeffer, San Francisco, CA). The blot was hybridized with labeled Pcp-1 and Actin probes in a 50% formamide hybridization solution. After hybridization, the membrane was washed at 65C several times in 0.2x SSC, 0.1% SDS and exposed to XAR-5 film (Eastman Kodak, Rochester, NY). As control the membrane was hybridized with total rRNA labeled with [alpha-32P]dCTP in a reverse transcriptase reaction. Signal quantification was performed by using the Scan Image system . UV Irradiation | Third instar wild-type and haync2/haync2 larvae were irradiated with the wild-type half LD by using 254-nm UV light with a germicide lamp (UVP), and the irradiation was measured using a UVX radiometer (UVP). Antibody Generation and Western Blot | A Hay-Glutathione S-transferase recombinant protein was constructed using the first 70 residues from the Hay protein. The protein was purified as described previously and used to elicit polyclonal antibodies in Wistar rats. The antibody obtained was used in Western blot experiments. In general, total protein soluble extracts were prepared from adult flies and standardized. Samples were loaded in 12% SDS-PAGE gels (Laemmli, 1970), blotted in nitrocellulose, and immunostained . Hay protein was visualized using a 1:1000 dilution of the anti-Hay antibody. Peroxidase-conjugated goat anti-rat was used as secondary antibody (Zymed Laboratories, South San Francisco, CA). RESULTS : Multiple Cuticular Defects Are Present in haywire Mutant Flies | Several hypomorphic and antimorphic hay alleles have been analyzed at the genetic and molecular levels . It has been reported that mutations in the hay gene produce an increase to UV light sensitivity, male sterility in combination with mutations in a particular tubulin gene, and some defects in the development of the nervous system . These phenotypes resemble some of the human manifestations produced in humans affected with XPB. However, except for the sensitivity phenotype caused by UV irradiation, which is attributed to defects in DNA repair, other phenotypes can be the consequence of DNA repair, transcriptional problems, or both. To understand the phenotypes produced by hay mutations in more detail, we reduced hay activity to its lowest viable level by making heteroallelic combinations of different antimorphic or hypomorphic alleles and two human-like alleles with the conditional hypomorphic haync2 mutation (Figure A). Progeny from these crosses had abdominal and wing defects, as well as deformations in the bristles (Table and Figure ). The abdominal abnormalities appeared as the loss of some cuticle portions. Electron microscopy showed that this phenotype is a consequence of a reduction of the deepest layers of the lamellate procuticle, whereas the superficial layers were not affected (Figure B). Both cuticular layers are derived from the same cell type , which suggest that this phenotype is due to a reduction of protein synthesis, rather than to the absence of these cells. Figure 1 | hay gene structure, and heteroallelic and transheterozygous phenotypes of hay and cdk7 mutants. hay gene structure, and heteroallelic and transheterozygous phenotypes of hay and cdk7 mutants. (A) Gene structure and localization of the mutations of the different hay alleles. The red boxes are the helicase domains, the black boxes are nuclear localization signals, the green box is the ATP-binding region, and the pink box is the COOH-terminal region affected in the XPCS human allele. The nucleotide sequence of the fly alleles has been reported previously . The hayTTD human-like allele has a change of Ser 119 for Pro , and the hayXPCS human-like allele is an insertion of one base at the intron acceptor site that produces a phase change in the open reading frame at the COOH terminus, which is described in the human allele . Hypomorphic alleles are indicated with a blue arrow, antimorphic alleles with a red arrow, and the human alleles introduced in transgenic flies with a black arrow. The original haync2 allele was originally described as antimorphic in terms of its interactions with tubulin mutants ; however, in our genetic analysis it behaves as hypomorphic. (B) Dorsal view of hay heteroallelic flies with cuticle deformations in abdominal tergites (top panels) and electron microscopy pictures of cuticle preparations of wild-type and mutant flies (bottom panels). The arrows indicate the exocuticular (Ec) and endocuticular (En) layers, ms is muscle, and pc is parenchyma. (C) Brittle bristle phenotype. Scanning micrographs of wild-type and brittle bristles from the scutellum of hay heteroallelic flies. (D) Wing phenotypes A and B of haync2/haync2rv1 and cdk7P140S/haync2rv8 flies. Note that both phenotypes are very different. The relevant genotypes are indicated. (E) Cuticle preparations of cdk7P140S/haync2rv8 flies showing the brittle bristle and abdominal cuticular phenotypes. Complete genotypes are given in MATERIALS AND METHODS. Table 1 | Viability and frequency of defects in heteroallelic hay flies The data in table are percentages obtained from at least 150 individuals of each genotype. An example of each phenotype is shown in Figure . Compared with wild-type flies, bristles present in hay heteroallelic flies were very fragile and deformed in shape and texture (Figure C). This phenotype affects both the machrochaete and the microchaete. Some of the bristles have a fork-like structure at the tip and in general the macro- and microchaete in the thorax do not show the organization found in wild-type flies. Because of their similarities with the brittle hair phenotype present in TTD patients, we named this phenotype "brittle" bristles. Although the origin, function, and structure of the hair in mammals are basically different to the fly bristles, they do have molecular similarities in their development. The integrity of both hair and bristles requires high transcription levels of genes' structural coding for proteins that assemble these complex structures . Genetic evidence that the brittle bristle phenotype is due to transcriptional defects is presented below. In addition to the cuticular and brittle bristle defects, wing defects were also observed in some of the hay heteroallelic combinations (Table and Figure D). This phenotype, which we named wing phenotype A, is not as penetrant as are the brittle bristle and abdominal cuticular phenotypes. The haync2rv8 and haync2rv4 alleles as well as the Df(3L)lxd6 with the haync2 allele do not show wing defects . These results suggest that these wing defects were not due to the gene product dosage, but to allele-specific interactions. Transgenic Flies Carrying Human-like Alleles with Mutations Reported in Human Patients Reproduce Defects Observed in hay Heteroallelic Flies | In addition to known EMS alleles , we constructed two new hay alleles containing two mutations found in XPB patients. One of these mutations causes TTD, whereas the other produces both XP and CS manifestations . These transgenes were named hayTTD and hayXPCS, respectively (Figure A). Although wild-type transgenes were able to rescue lethality of hay homozygous flies as well at the abdominal, bristle, and wing defects, mutant transgenes (hayTTD and hayXPCS) were not (our unpublished data). The presence of one copy of the hayTTD or hayXPCS transgenes in a haync2/haync2 background decreases fly viability dramatically . Transgenic flies carrying one copy of hayXPCS, one copy of haync2 allele, and one copy of the wild-type gene reproduced defects observed in heteroallelic combinations of hay, but in a more severe manner. For instance, bristles were much more fragile and thinner and wings were practically amorphous (Figure A). In addition, locomotion impairments were observed. None of these defects were observed in transgenic flies carrying two wild-type hay alleles, demonstrating that, in individuals with this genotype, hayXPCS is not dominant. Both transgenes overexpress the corresponding RNA (our unpublished data). A polyclonal rat antibody against the N-terminal domain of the wild-type Hay protein was produced (see MATERIALS AND METHODS). By using this antibody, we analyzed the Hay protein produced by the transgenic flies harboring the hayTTD and hayXPCS constructs. We found that both transgenic lines have a band corresponding to the expected size of wild-type and mutant proteins (similar89 kDa) and a truncated product 30 kDa smaller than the wild type (Figure B). This result suggests that these mutant forms produce a nonstable protein that is processed. Table 2 | Fly viability in heteroallelic and homozygous haync2flies in the presence of the hayTTD and hayXPCS alleles Data are presented as the percentage of expected individuals of each genotype and are an average of at least 500 embryos of all genotypes laid in each cross. Figure 2 | hayXPCS allele enhances the hay phenotypes. hayXPCS allele enhances the hay phenotypes. (A) Phenotype of a haync2 heterozygous fly carrying one copy of the hayXPCS allele expressed from the fly hsp83 promoter. Similar phenotypes are presented with the hayTTD allele. haync2/+ flies have a wild-type phenotype. (B) Western analysis of soluble proteins from wild-type and transgenic third instar larvae expressing either the HayTTD and HayXPCS proteins. After Western transfer Hay proteins were visualized using a polyclonal antibody against a glutathione S-transferase fusion protein harboring the N terminus of Hay. The genotypes of the larvae are wild-type: OreR; XPCS: w118, hayXPCS/CyO; TTD: w118, hayTTD/CyO. The wild-type protein is indicated by a red arrow and the blue arrow indicates a truncated form which is observed in both mutants. cdk7 Genetically Interacts with Hay | The observed hay phenotypes could be due to problems either in DNA repair or in transcription. It has been demonstrated that Cdk7 participates in transcription as a component of TFIIH and is also involved in the control of the cell cycle, but it does not participate in DNA repair . Thus, it is possible that cdk7 mutants would present a subset of defects observed in hay flies affected in these processes. In an attempt to dissect the causes of different phenotypes observed in hay mutants, we performed crosses of hay flies with flies harboring different doses of cdk7 wild-type gene, with or without the conditional dominant negative cdk7P140S (; see MATERIALS AND METHODS). If at least some of the defects observed in single cdk7 or hay mutant flies are caused by defective transcription due to different TFIIH abnormal subunits, we should observe a genetic interaction in transheterozygous flies. The results are shown in Table . We found that when the only source of Cdk7 comes from the mutant allele cdk7P140S, flies incubated at the restrictive temperature presented wing, cuticular, and bristle phenotypes. The last two were similar in appearance to the ones presented by single hay homozygous or heteroallelic mutants (Table , and Figure , D and E). Most of the hay alleles show a genetic interaction with cdk7P140S increasing the penetrance of bristle and cuticular phenotypes . A stronger interaction was found between cdk7P140S and haync2rv8 alleles. Based on the appearance of the common bristle and cuticular phenotypes by single hay and cdk7 mutants, and the enhancement of these phenotypes in transheterozygous cdk7/hay flies, we suggest that these two phenotypes were caused by transcriptional defects. These phenotypes were only observed at the restrictive temperature for cdk7P140S. In addition to this result, no interaction was observed in Df(1)JB254/+; hay/+ flies , indicating that the bristle and cuticle defects are only due to the presence of the cdk7P140S mutant. Thus, the possibility of a second mutation that interacts with hay producing these phenotypes could be discarded. On the other hand, bristle and cuticle phenotypes emerge in homozygous Df(1)JB254 flies in the presence of one copy of the cdk7P140S allele , reflecting that these phenotypes were due to the loss of function of cdk7. Table 3 | Wing and bristles phenotypes in cdk7 flies and genetic interaction with hay In contrast, wing defects observed in cdk7P140S flies were clearly different from the ones presented by hay mutant flies, thus we named it wing phenotype B (Figure D). This phenotype was present even when there is a wild-type cdk7 copy, which is not the case for bristle and cuticular phenotypes. cdk7P140S flies present small and deformed wings (phenotype B), whereas wings of heteroallelic hay flies show loss of marginal regions and the presence of blisters (Figure D, phenotype A). All transheterozygous hay/cdk7 combinations showed the cdk7P140S wing phenotype B. Unexpectedly, the wing phenotype of cdk7P140S is not enhanced by the addition of a defective hay allele with the exception of the haync2rv8 allele, showing that the interaction between cdk7P140S and haync2rv8 is allele specific (Table , in bold). Lack of enhancement of cdk7P140S wing phenotype with different hay alleles, with the exception of haync2rv8 (see DISCUSSION), supports that the different wing defects in hay and cdk7 mutants (phenotype A and B) are caused by failure in different mechanisms. Thus, because Cdk7 is only required for transcription and for the progression of cell cycle, the hay wing phenotype is probably caused by a deficiency in DNA repair rather than by transcription. To complement the genetic interaction data between cdk7 and hay, mRNA levels of cuticular protein Pcp-1 and Actin , which should be related to the observed phenotypes, were analyzed in haync2rv8/cdk7P140S flies and compared with heterozygous and wild-type flies. haync2rv8 flies were crossed with cdk7P140S organisms and the obtained transheterozygous adults were incubated at the restrictive temperature (29C) at different times (0, 4, 8, and 12 h). Then total RNA from three different experiments was purified, and similar amounts of for each sample were loaded in a slot blot and hybridized against a Pcp-1 and Actin cDNA probes. Because mutations in the TFIIH components should affect a large number of genes transcribed by RNA polymerase II we used as control the levels of rRNA. Figure shows that RNA levels of Pcp-1 and Actin were reduced in the haync2rv8/cdk7P140S flies incubated at 29C compared with hay and cdk7 heterozygous flies incubated at a similar temperature. Quantification of the Pcp-1 and actin mRNA levels in haync2rv8/cdk7P140S flies, of three independent blots, indicates that the reduction was of similar35 and 40%, respectively, if compared with the wild-type at 12 h of incubation at the restrictive temperature (Figure , A and B). The reduction of both mRNA levels occurs at 29C, confirming that this is due to the cdk7P140S conditional mutant. These results support that the interaction between cdk7 and hay mutants has an effect on transcription that may result in the observed phenotypes. Figure 3 | RNA levels of the Pcp-1 cuticular and Actin proteins in cdk7P140S and hay double mutant flies. RNA levels of the Pcp-1 cuticular and Actin proteins in cdk7P140S and hay double mutant flies. Slot blot hybridization of the Pcp-1 (A) and Actin probes (B) against total RNA purified at different times from adult flies incubated at 29C. The quality of the RNA was confirmed in typical formamide-agarose gels. Independent membranes were hybridized for each probe. The different genotypes are indicated in the figure. The same blots were washed and hybridized against rRNA as loading control. Note that in the cdk7P140S/hayrv8 flies (indicated with an arrow) the Pcp-1 and Actin transcript levels are reduced with the incubation time. The graphs show for each genotype, the average (n = 3) percentage of reduction in specific transcript levels after 12 h of incubation at the restrictive temperature (29C) compared with rRNA levels. Flies Affected in Hay Have a High Rate of Apoptosis in Imaginal Discs and in CNS | It has recently been shown that human cells with defects in XPB are not able to repair oxidative DNA damage by BER during transcription (transcription-coupled repair or TCR) . These defects in TCR by XPB have been related to neurodegenerative problems found in CS patients, suggesting that deficient TCR may induce apoptosis . To know whether apoptosis was increased in hay flies, we analyzed the presence of apoptotic bodies in larval CNS and in the imaginal discs. Interestingly, an increase of apoptotic bodies in both kinds of tissues in hay larvae was found (Figure , A and B). The amount of apoptotic cells is increased in hay discs after UV irradiation compared with irradiated wild-type discs (Figure C). In contrast, we could not detect abnormal apoptotic bodies in cdk7P140S imaginal discs and in the CNS, even in cdk7P140S/haync2rv8 transheterozygous flies (Figure D). Because cdk7 only participates in transcription and in cell cycle control, but not in DNA repair, we suggest that apoptosis in haync2 mutants is due to DNA repair defects. These results also show that DNA repair by TFIIH is required during development and constitutes the first in vivo evidence that defects in the hay gene may induce apoptosis. cdk7P140S and hay flies present different wing phenotypes. For cdk7, the allele-specific interaction with haync2rv8 suggests that wing phenotype is related to deficient transcription, although a deficiency in cell cycle cannot be ruled out. With the evidence presented herein, we propose that hay wing phenotype is probably a consequence of an abnormal high apoptosis found in wing imaginal discs of mutant hay flies, caused by a deficient DNA repair during development. Figure 4 | Apoptosis in hay imaginal discs and brains. Apoptosis in hay imaginal discs and brains. (A) Wing and leg disc preparations from wild-type and haync2/haync2 flies stained with acridine orange (red) and TUNEL assay (green) for the presence of apoptotic cells. (B) CNS preparations from wild-type and haync2/haync2 flies stained with acridine orange (red) and TUNEL (green). Note the high number of apoptotic bodies in the hay larval tissues. (C) Wild-type and haync2/haync2 wing discs stained for apoptosis 24 h after UV irradiation (half LD for the wild-type, 164 J/m2). (D) cdk7P140Shay+/hay+ and cdk7P140Shay+/haync2rv8 discs. Note that only background is detected as in the wild-type discs. In all cases cell death is also detected with Nile blue (our unpublished data). Relevant genotypes of these flies are as in Figure and MATERIALS AND METHODS. Acridine orange-treated samples were observed under a fluorescence microscope. The TUNEL assay was visualized using confocal microscopy. For each tissue sample maximum projections of the whole set of images are shown. Dmp53/hay Genetic Interaction | In mammalian cells, apoptosis induced by DNA damage is mediated by p53 . A physical interaction between p53 and TFIIH has been demonstrated , and p53-mediated apoptosis seems to be defective in XPB and XPD mutant cells . The Drosophila p53 homolog gene (Dmp53) can activate apoptosis in response to DNA damage by gamma-irradiation . Wild-type Dmp53 overexpression in the eye and wing discs induces apoptosis and severe deformations in both adult organs (; ; Figure B). Wing deformations are characterized by abnormal shape and a dramatic reduction of the wing blade as a consequence of massive cell death (Figure B). To know whether there is an interaction between Dmp53 and TFIIH in Drosophila, we used transgenic flies with either wild-type p53 (Dmp53) gene or with several dominant negative alleles affected in the DNA binding domain (Dmp53R155H, Dmp53H159N, Dmp53K259H, and Dmp53C) . Both wild-type and mutant p53 alleles were overexpressed under GAL4-UAS system in the third instar larval wing discs of hay mutant individuals, by using the GAL4 wing driver MS1069 (see MATERIALS AND METHODS). Figure 5 | hay and Dmp53 genetic interaction. hay and Dmp53 genetic interaction. Wild-type Dmp53 and a dominant negative Dmp53R155H, Dmp53H159N, Dmp53K259H, and Dmp53C+ alleles were overexpressed in the wing disc by using the MS1096 GAL4 driver in the presence or the absence of hay alleles. (A) Wild-type OreR wing. (B) Overexpression of the wild-type Dmp53 produces aberrant wings. (C and D) Wing overexpression of the wild-type Dmp53 in haync2 and haync2rv1 heterozygous backgrounds. Note that the apoptotic phenotype is reduced. (E) Overexpression of the dominant negative Dmp53R155H allele does not produces defects in the wing. (F) Example that shows that no genetic interaction occurs between Dmp53 negative dominant mutants and hay alleles. We found that wing defects produced by the Dmp53 overexpression were suppressed in the presence of a hay mutant allele. Although this suppression was partial, the penetrance was 100%, even in the presence of a single hay mutant allele (haync2/+ and haync2rv1/+ flies; Table and Figure , C and D). These results showed that Dmp53 needs an intact TFIIH to induce apoptosis because it occurs in human cells derived from patients affected in XPB and XPD . Table 4 | Genetic interaction between Dmp53 and hay Data represent percentages from at least 50 individuals of each genotype. MS1096 is a wing imaginal disc GAL4-driver. Dmp53R155H is a p53 negative dominant mutation that is not able to induce apoptosis . DISCUSSION : Phenotypic defects produced by some of the components of TFIIH have been difficult to characterize in mammalian systems. Analyzing combinations of mutant alleles of genes encoding two TFIIH subunits (hay and cdk7), we have been able to dissect phenotypes associated with transcription and with DNA repair. Brittle Bristles and Cuticular Phenotypes Are Associated with TFIIH Transcriptional Deficiencies | Reduction of the Hay activity in different heteroallelic combinations produces analogous defects to the ones observed in humans affected in the XPB gene. Besides the increase in sensitivity to UV irradiation and sterility that were described by , the most obvious defects in adult hay heteroallelic flies are severe cuticular deformations in the abdomen, brittle bristles, and aberrant wings. These defects can be rescued by overexpression of the Hay wild-type form in all the heteroallelic combinations. From the data reported herein, we propose that both the cuticular abdominal deformations and the brittle bristles are associated with deficiencies in transcription caused by defective TFIIH. For the adult cuticles, electron microscopy studies showed that this phenotype is due to a reduction of the deeper lamellate procuticle layers. As in the cuticle, construction of hair-like structures of bristle requires the abundant transcription of genes that encode bristle structural components . Bristles in hay flies are very fragile and they have severe deformations in their structure and texture. These defects resemble defects observed in TTD patients' hair caused by reduced transcription of genes encoding structural components such as the sulfur-rich proteins of the hair . This phenotype is also caused by a mutation in cdk7 producing a defective TFIIH. We found a clear genetic interaction between hay and cdk7 for the brittle bristle and the abdominal cuticular phenotypes. Because these two TFIIH subunits are involved in transcription, but Cdk7 is not involved in DNA repair, we concluded that lower levels of transcription cause brittle bristles and the abdominal cuticular phenotypes. This conclusion was supported by the fact the double hay/cdk7 mutants can affect transcript levels of genes related to the cuticular and bristle defects . In addition, we have observed similar defects by the genetic interaction of hay and mutant alleles of other basal transcription factors genes (our unpublished data). The general transcription machinery is affected in hay mutants used in this work. Nevertheless, some hay individuals, although defective in many senses, are able to have a proper cell differentiation and to develop until adults. However, we believe that in hay cells that require the overexpression of specific genes, the TFIIH machinery is probably exhausted before the transcriptional program is completed, resulting in defects such as cuticle and bristles phenotypes. These defects could be the result of a partial reduction in the transcriptional rate of highly expressed genes. This could be similar to the proposed explanations for the brittle hair and ichthyosis manifestations observed in TTD patients . The penetrance of the defects associated with hay is enhanced in the presence of hayXPCS and the haync2 alleles . The XPCS patient carrying the mutation similar to the hayXPCS allele is heterozygous, and it has been speculated that this mutation could be a dominant allele . However, the presence of a second mutation in the control region of the other XPB allele (paternal) or in other gene has not yet been ruled out, in particular, because it was observed that the transcript of the paternal allele is not detectable in the patient . Our results show that in Drosophila, hayXPCS allele is not dominant, and it only produces a hay phenotype when it is in combination with other hay mutant alleles. It is important to note that, although the nature of the two mutations (TTD and XP/CS) is different, in Drosophila they both produce a truncated Hay form that could affect the function of TFIIH. Interestingly, reported that haync2 mutant also accumulates a truncated product of similar size to the one found in this work, suggesting that in the fly some hay mutations produce a nonstable product. It has not been tested whether this truncated protein retains some function or whether it could be assembled into TFIIH complexes. Wing Defects in hay Flies Can Be Associated with Defects in DNA Repair during Development | Both cdk7P140S and hay flies have aberrant wings, but these defects are different in each case. Results presented herein show that the wing phenotype present in cdk7P140S (phenotype B) is not enhanced by most of the hay alleles, with the exception of haync2rv8 allele (see below). This suggests that wing defects observed in both mutants are due to deficiencies in different processes. Wing defects in cdk7 mutant can be due to a deficiency in the cell cycle control and/or in transcription . However, an allele-specific interaction with haync2rv8 is observed also for wing defects. haync2rv8 allele has a modest effect in the bristles in combination with haync2, but it does not cause any wing defects in this situation . In contrast, a single copy of this allele in the presence of a wild-type hay gene strongly enhances cdk7P140S wing phenotype (Table , in bold). We propose that the nature of these two mutations produces an interaction that has a strong effect on TFIIH activity, most likely affecting transcription. Conversely, aberrant wings observed in hay heteroallelic combinations (phenotype A) seem not to be related to transcription. This hypothesis is supported by the fact that most of the hay alleles do not increase wing phenotype B observed in cdk7P140S flies. Taking these results plus the fact that apoptosis is not detected in cdk7P140S/haync2rv8 flies as it occurs in the haync2/haync2 background, we suggest that aberrant wings observed in hay mutant flies are not due to transcription but to DNA repair. Defects in hay Increase Apoptosis during Fly Development: Interaction with Dmp53 | A higher rate of apoptotic bodies is detected in hay heteroallelic combinations and in haync2 homozygous flies. Apoptosis is also dramatically increased in hay imaginal discs after UV irradiation but not in wild-type discs. As stated above, apoptosis is not observed in cdk7 mutant flies, suggesting that the presence of a larger number of apoptotic cells in hay flies is due to abnormal DNA repair. These results suggest that DNA repair function of TFIIH is required during development even without the challenge of external physical or chemical agents that damage DNA. This is the first in vivo evidence that defective hay, an XPB homolog, may induce cell death during development. The TFIIH function in DNA repair could be either in NER or in BER. TFIIH is required for TCR of 8-oxo-guanine and thymine glycol caused by oxidative damage of DNA . It has been proposed that in CS patients the accumulation of 8-oxo-guanine and/or thymine glycol caused by defective TCR may induce apoptosis during development of the nervous system . We also propose that apoptosis observed in hay imaginal discs and in CNS is mostly due to an accumulation of oxidative damage that normally occurs during development. This accumulation of oxidative damage may be caused by defective TFIIH. Interestingly, the presence of a hay mutant allele suppresses the penetrance of the wing defects due to the overexpression of Dmp53 in wing disc. It is remarkable that this suppression is fully penetrant , showing a clear genetic interaction between Dmp53 and hay. This information is of particular relevance because it has been reported that primary cultured fibroblast derived from individuals with xeroderma pigmentosum, which are deficient in DNA repair by mutations in XPD or XPB, are not able to undergo p53-induced apoptosis . This deficiency can be rescued by transferring the wild-type XPD or XPB genes in the corresponding mutant cells, suggesting that XPB and XPD are components of the p53-mediated apoptosis pathway . The results presented in this work are in agreement with this information and confirm that the fly is an excellent animal model to study the role of TFIIH during development. However, a paradox emerges from these results. Mutations in hay induce apoptosis during development, but the same mutations suppress the apoptotic effect of the Dmp53 overexpression. Dmp53 apparently needs an intact TFIIH to induce apoptosis as in cultured human cells from XP patients. Then how is the apoptosis produced by hay mutations activated? It is possible that the apoptosis seen in hay mutants could be triggered by a Dmp53-independent mechanism. Alternatively, it cannot be ruled out that Dmp53 may still respond to other molecules that may activate it as a response to DNA damage produced by a defect in TFIIH during development. TFIIH is able to phosphorylate p53 in vitro, and the phosphorylated p53 is able to bind more efficiently to its target DNA sequences . There is evidence that p53 can modulate the TFIIH-associated nucleotide excision repair activity . Thus, there is an intricate cross talk between p53 and TFIIH. In conclusion, mutations in TFIIH components produce as diverse phenotypes in Drosophila as they do in humans. Some of these defects are caused by faulty transcription, such as the bristle and cuticular phenotypes. Other defects, such as aberrant wings in hay flies are related to problems in DNA repair, which could increase cell death during development. A genetic interaction between hay and the fly homolog of p53 seems to be similar in some aspects to what has been found with human alleles in the same genes, confirming the value of the fly as a model for human diseases produced by mutations in TFIIH. Studies of humans affected in TFIIH have proposed that some mutant alleles of XPB and XPD genes may impair only DNA repair or transcription and others may affect both. Most of the hay (XPB homolog) alleles analyzed in this work seem to affect both mechanisms. This work shows that by using Drosophila genetics, the developmental role of the different components of the TFIIH complex can be analyzed in detail and that this information is relevant for the analysis of the syndrome manifestations in humans. Backmatter: PMID- 12221130 TI - Agonist-induced PIP2 Hydrolysis Inhibits Cortical Actin Dynamics: Regulation at a Global but not at a Micrometer Scale AB - Phosphatidylinositol 4, 5-bisphosphate (PIP2) at the inner leaflet of the plasma membrane has been proposed to locally regulate the actin cytoskeleton. Indeed, recent studies that use GFP-tagged pleckstrin homology domains (GFP-PH) as fluorescent PIP2 sensors suggest that this lipid is enriched in membrane microdomains. Here we report that this concept needs revision. Using three distinct fluorescent GFP-tagged pleckstrin homology domains, we show that highly mobile GFP-PH patches colocalize perfectly with various lipophilic membrane dyes and, hence, represent increased lipid content rather than PIP2-enriched microdomains. We show that bright patches are caused by submicroscopical folds and ruffles in the membrane that can be directly visualized at similar15 nm axial resolution with a novel numerically enhanced imaging method. F-actin motility is inhibited significantly by agonist-induced PIP2 breakdown, and it resumes as soon as PIP2 levels are back to normal. Thus, our data support a role for PIP2 in the regulation of cortical actin, but they challenge a model in which spatial differences in PIP2 regulation of the cytoskeleton exist at a micrometer scale. Keywords: INTRODUCTION : Phosphatidylinositolbisphosphate (PIP2) controls many cell processes, ranging from channel gating to vesicle trafficking and the actin cytoskeleton. In this latter case, PIP2 is thought to directly interact with actin binding proteins, influencing the equilibrium of monomeric (G-) to filamentous (F-) actin at the level of G-actin availability (e.g., plectin; , profilin; , ) and at the level of polymerization into actin fibers. Examples are members of the F-actin --severing and --capping protein family including gelsolin and CapG , and protein beta2 . Cytoskeletal interactions with integral membrane proteins can also be regulated by PIP2, as is the case with ezrin binding to ICAM and CD44 . Many of these actin binding proteins interact with PIP2 via pleckstrin homology domains . The majority of data on the interaction of PIP2 with the cytoskeleton stem from in vitro binding studies and lipid biochemistry. If membrane lipids are to exert a local signaling function, local differences in content or availability must exist. However, although the phosphoinositide levels at the membrane appear to be tightly regulated by a multitude of specific reactions (; for review see ), little is known about their distribution along the membrane. Biochemical studies have reported the existence of separate PIP2 pools within cells , and PIP2 was reportedly enriched in detergent-insoluble membrane fractions (rafts, caveolae; ; ), although a very recent electron microscopy study challenges these results . In addition to these biochemical reports, a limited number of in vivo studies have implicated PIP2 levels in the regulation of the cytoskeleton. For example, increasing [PIP2] by overexpression of the type I PIP-kinase alpha lead to an increase of stress fibers in CV1 cells , and sequestering of this lipid using a membrane-permeable PIP2-binding peptide blocked motility . A detailed understanding of a role for PIP2 as a local signal requires techniques to study cellular processes with spatial and temporal resolution in single living cells. Recently, an approach to image PIP2 in living cells was pioneered in the labs of Meyer and Balla . Both groups used the PIP2-binding pleckstrin homology domain of PLCdelta1, fused to GFP (GFP-PH), to study PIP2 in the membrane in vivo. In resting cells, GFP-PH is bound to the membrane, and it translocates to the cytosol after agonist-induced PIP2 hydrolysis. Translocation can be detected at very high sample rate using fluorescence resonance energy transfer . Imaging of GFP-PH bound to the membrane also provides spatial resolution. Interestingly, initial studies reported a rather uniform distribution of GFP-PH along the membrane in unstimulated cells , whereas a recent study reported that GFP-PH displays distinct bright patches on a uniformly labeled background. These bright patches are highly dynamic and rich in F-actin content, and they often colocalize with membrane ruffles and microvilli-like structures. Development of new membrane ruffles was reported to start with local concentration of GFP-PH. Based on these observations, and in line with the hypothesized function of PIP2 as a local regulatory factor in cytoskeletal dynamics, GFP-PH patches were interpreted as local PIP2 enrichments . This study was undertaken to determine to what extent local differences in membrane PIP2 content influence the cytoskeleton in vivo. To this goal, we investigated 1) whether changes in PIP2 level induced by physiological PLC-activating agonists modulate cortical actin dynamics and 2) whether physiologically relevant differences in membrane [PIP2] (which, be it relative enrichments or depletions, will here be collectively termed PIP2 patches) exist locally at the plasma membrane of cultured cells. MATERIALS AND METHODS : Materials | 1-Oleoyl-LPA, endothelin, bradykinin, histamine, phenylarsine oxide, and quercetin were from Sigma Chemical Co. (St. Louis, MO), and ionomycin and neurokinin A were from Calbiochem-Novabiochem Corp. (La Jolla, CA). H2O2 was from Riedel-deHaen (Germany), and thrombin receptor activating peptide (SFLLRN) was synthesized in house. Membrane dyes were from Molecular Probes Inc. (Eugene, OR); listed are the dye name, with the structures in parentheses: diphenyl DiI (1,1'-dioctadecyl-5,5'-diphenyl-3,3,3',3'-tetramethylindocarbocyanine chloride); SP-DiIC18(3) (1,1'-dioctadecyl-6,6'-di(4-sulfophenyl)-3,3,3',3'-tetramethylindocarbocyanine); SP-DiOC18(3) (3,3'-dioctadecyl-5,5'-di(4- sulfophenyl) oxacarbocyanine, sodium salt); DiIC18(5) oil (1,1'-dioctadecyl-3,3,3',3'-tetramethylindodicarbocyanine perchlorate); TRITC DHPE (N-(6-tetramethylrhodaminethiocarbamoyl) -1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine, triethylammonium salt); bis-BODIPY FL C11-PC (1,2-bis-(4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-undecanoyl)-sn-glycero-3-phosphocholine); BODIPY 564/570 C11 (4,4-difluoro-5-styryl-4-bora-3a,4a-diaza-s-indacene-3-undecanoic acid); NBD-PA (7-nitrobenz-2-oxa-1,3-diazol-PA). Constructs and Transfection | The pcDNA3 expression vectors with inserts eGFP-PH(PLCdelta1), eCFP-PH(PLCdelta1), eYFP-PH(PLCdelta1), and eGFP-CAAX were described elsewhere . pEYFP-Mem was from (Palo Alto, CA), pcDNA3 with insert GFP-Actin (from the N-terminus: GFP, flexible linker [GGGLDPRVR] and actin) was obtained from Dr. J. Neefjes, Division of Tumor Biology, and vectors containing the endothelin B receptor, and a human NK2 receptor with C-terminal truncation at position 328 were obtained from Dr. W. Moolenaar, Division of Cellular Biochemistry at our institute. Constructs were transfected using calcium phosphate precipitate, at similar0.8 mug DNA/well. After transfection for 12 h, cells were washed with fresh medium and incubated for 4 --24 h until usage. Cell Culture and Stimulation | N1E-115 neuroblastoma cells and NIH-3T3 fibroblast cells were seeded in six-well plates at similar25.000 cells per well on 25-mm glass coverslips and cultured in 3 ml DMEM supplemented with 10% serum and antibiotics. Agonists and inhibitors were added from concentrated stock solutions. It was verified that the PIP kinase inhibitors PAO, quercetin and H2O2 did not noticeably affect cell viability over the time course of the experiments. Determination of PLC-mediated PIP2 Breakdown by Fluorescence Resonance Energy Transfer | Monitoring of dynamics of PLC activation with FRET was described in detail elsewhere . In brief, cells were transiently transfected with YFP-PH and CFP-PH, at 1:1 ratio. When bound to PIP2 at the membrane, these constructs are in close proximity and show FRET; upon PIP2 hydrolysis, CFP-PH and YFP-PH dilute out into the cytosol and FRET ceases. Excitation of CFP-PH was at 425 +- 5 nm, and emission was collected simultaneously at 475 +- 15 (CFP) and 540 +- 20 nm (YFP). FRET was expressed as ratio of CFP to YFP signals, and changes were expressed as percent deviation from the initial value. Confocal Microscopy | For imaging, coverslips with cells were transferred to a culture chamber and mounted on an inverted microscope. All experiments were performed in bicarbonate-buffered saline (containing in mM: 140 NaCl, 5 KCl, 1 MgCl2, 10 glucose, and 10 HEPES Ca2+), pH 7.2, kept under 5% CO2, at 37C. Confocal imaging was with a DM-IRBE inverted microscope fitted with TCS-SP scanhead (, Mannheim, Germany). Excitation of eGFP, eYFP, DiO, NBD-PA, and BODIPY-FL was with the 488-nm laserline, and emission was collected at 500 --560 nm. For DiI, Bodipy 564/570, and TRITC, excitation was with 568 nm, and emission was collected at 590 --650 nm. DiD was excited at 633 nm, whereas emission was collected at 645 --700 nm. Cross-talk between channels was checked and where necessary corrected using Confocal Software. Image Analysis | Timelapse Analysis. | For time lapse studies, series of confocal images were taken at 5 --30-s time intervals and stored on harddisc. Visualization and analysis was performed off-line using TCS and Qwin software and a suite of analysis routines that were written by one of the authors using the APL+Win development platform (APL2000 Inc., Bethesda, MD), as detailed below. Motility Assay. | To detect motility of actin and GFP-PH patches, pairs of images from a stored timelapse series (8-bit grayscale values) were analyzed essentially for correlation of pixel intensities, using the following steps: (i) Within a series of N images (1...N), pairewise comparison was carried out for images (1 with J + 1), (2 with J + 2),... ([N - J] with N), with J selected to obtain a lapse of 30 --120 s between the image pair, which appeared optimal for the detection of changes. (ii) Analysis of movement is best understood by referring to Figure A, lower scatterplot. Pixels that are intensely fluorescent in the first but not in the second image appear in the scatterplot below the blue diagonal line. Slope and abscissa of this line can be set for optimal rejection of unaltered pixels; in addition, a threshold can be included for rejection of background noise. If If is the intensity of a given pixel in the first image, and Is that in the second image, then for these pixels, where A is the parameter that determines detection threshold (background; usually <30), B the X-abscissa (range, 5 --40) and C the slope (range, 1 --1.5). Similarly, for pixels above the red diagonal where B is the Y-abscissa. The degree of dissimilarity DD is then calculated as (iii) Finally, motility traces are constructed by plotting corresponding DD versus image index for (1...[N - J]). The reproducibility and sensitivity of this algorithm were checked using simulated and real data. Further details are available upon request from K.J. Figure 1 | Influence of agonist-induced PLC activation on cortical actin dynamics. Influence of agonist-induced PLC activation on cortical actin dynamics. (A) Detection of actin motility using intensity scatter plot. Time-lapse series of confocal images of the basal membrane of GFP-actin expressing N1E-115 cells were collected. For each pair of subsequent images X and Y, intensities of each pixel were plotted as dots in a (X, Y) scatterplot (see MATERIALS AND METHODS for further details). Multiple occurrence of identical coordinates is color-coded. For a pair of images collected 2 s apart, the scatterplot shows a distribution along the diagonal (left scatterplot), with some divergence due to the inherent photon noise. For pairs of images taken at longer time intervals (top panel, t = 0 and 2 min), bright actin structures that have moved in between collection of the images will be apparent as off-diagonal clusters of dots (right scatterplot). Dots within the blue-dashed area in this plot represent pixels of which the intensity has significantly decreased in 2 min; the red-dashed area contains all pixels whose intensity has increased after 2 min. Taking a fixed time interval (usually 30 --120 s), actin motility as a function of time can be quantified extremely sensitively by calculating the fraction of pixels that are off-diagonal for subsequent images in a time-lapse series. In the composite (rightmost) photomicrograph, pixels from the dashed areas are superimposed on the image. Scale bar, 0.5 mum. (B --F) Comparison of agonist-induced PIP2 hydrolysis (left panels) and actin motility (right panels) in separate N1E-115 cells. Shown are responses from single cells (B) expressing the endothelin B receptor and stimulated with 20 nM endothelin (ET); (C) stimulated with 10 muM histamine (HIS); (D) stimulated with 1 muM bradykinin (BK); and (E) expressing a desensitization-defective neurokinin A receptor and stimulated with 1 muM neurokinin A (NKA) and (F) pretreated with 1 muM phenylarsine oxide (PAO) for 5 min and stimulated with 1 muM bradykinin (BK). Red marks indicate the moment of agonist addition. Imaging of Reconstructed Axial PSF (RAP Imaging). | Stacks of X/Y images of the basal membrane were captured at 40- or 80-nm axial distance, using a 63x, 1.32 NA oil immersion Planapochromatic objective and a pinhole setting of 1 airy disk. Excitation was at 488 nm, and emission was collected at 525 nm. Because the axial resolution at these conditions is similar1.05 mum, the objective point spread function (PSF; i.e., the gaussian intensity profile that is detected when a true point source is imaged with an objective) is thus oversampled up to 25 times (see Figure ). Before processing, individual images were smoothed once. For each pixel (x,y), plotting its intensity in image 1 to n of the stack of n images (I1...In) versus axial position reconstructs the axial PSF, because the thickness of the membrane (similar5 nm) can be ignored. The "center of intensity" of the axial fluorescence intensity profiles was then determined by a calculation-efficient, simplified fitting algorithm as follows. First, image numbers were sorted and arranged in order of decreasing intensity. Then, these numbers were multiplied by an array of n weighting values (W1...Wn). Values for W were chosen to progressively reduce the influence of out-of-focus (dim) images (see below). The sum P of these products is proportional to the axial position of the center of intensity. For example, be the order of intensity for a given (x,y) pixel: then the expression for P is: Three-dimensional surface profiles (see Figure ) were then constructed by plotting P as a function of x and y, using the Surface Plotter plugin of the public domain software ImageJ, version 1.24t (Wayne Rasband, National Institutes of Health, Bethesda, MD). The array of weighting values W was chosen so that A trivial example is the case where W1 = 1, and W2...Wn = 0. In this case, P will simply be the image number of the highest pixel intensity. Because of the large Poisson-distributed photon noise in the confocal images, this results in extremely noisy surface plots. Good results were obtained with linear weighting arrays that emphasize the images of highest intensity, such as the array Wi = 0.14, 0.13, 0.12, 0.11, 0.10, 0.09, 0.08, 0.07, 0.06, 0.05, 0.03, 0.02, 0, 0, 0, ..., 0. It should be noted that the described calculation-effective algorithm performed equally well as pixel-by-pixel curve-fitting algorithms that took dramatically longer computer processing. P was calibrated numerically as well as experimentally, using a slightly tilted mirror as the object. It is estimated that a Z-axis resolution up to similar15 nm is obtained. Further details are available on request from K.J. Membrane Staining | Membrane dye stocks were mixed by vigorous pipetting with bicarbonate-buffered saline to a final concentration of 1 muM. Cells were incubated with the mix for 5 --15 min at 37C. Before imaging, cells were washed three times with bicarbonate-buffered saline. Fluorescence Recovery after Photobleaching | For fluorescence recovery after photobleaching (FRAP) experiments, cells were imaged using a TCS-SP confocal microscope equipped with 63x (NA 1.3) oil immersion objective. Spots were bleached with the 488-nm argon laser line (Bis-Bodipy FL C11-PC; 0.2s) or 568-nm krypton laser line (DiI, 0.2 s), and recovery was sampled at 10 Hz. Data were corrected for slight (<5%) background bleaching and fitted with single exponents using Clampfit software (Axon Instruments, Union City, CA). RESULTS : Cortical Actin Motility Correlates with Agonist-induced PIP2 Breakdown | We recently reported the use of a FRET-based assay to monitor the kinetics of receptor-mediated PIP2 breakdown in single cells . We showed that distinct G protein --coupled receptors (GPCRs) induce PIP2 hydrolysis with their own characteristic kinetic profiles. Representative examples are depicted in Figure , B --F (left traces). To establish the relationship between membrane [PIP2] and actin dynamics, we set out to study the actin cytoskeleton under identical conditions. N1E-115 cells do not possess actin stress fibers, but the cortical actin cytoskeleton is well developed, and it mediates agonist-induced cell shape changes . Cortical actin dynamics were studied by in vivo time-lapse imaging of cells that express GFP-tagged actin. GFP-actin at the cell cortex displays an inhomogeneous, patchy distribution (Figure A, top panel), and it is highly dynamic, with individual structures showing seemingly random as well as directed movements. Strikingly, after addition of PIP2-hydrolyzing agonists, such as endothelin and neurokinin A, actin movements are inhibited within a minute. To study these effects in more detail, we set up an assay for the quantification of GFP-actin dynamics. Essentially, in this assay motility is expressed as change (lack of correlation) between successive images in a time-lapse series (see the legend to Figure and MATERIALS AND METHODS). We compared the kinetics of actin dynamics to those of the concomitant PIP2 decreases after agonist addition. The strong PLC activator endothelin (ET, Figure B) causes transient retardation of actin dynamics that correlates well with the decrease in membrane PIP2 content (representative result of 5 experiments). Agonists that induce weaker PIP2 hydrolysis, such as histamine, LPA, and thrombin , caused less pronounced or undetectable inhibition of actin motility (n = 9; Figure C). Bradykinin, which induces a short-lived drop in [PIP2] caused a minor and transient drop in actin motility (n = 6; Figure D), whereas the sustained PIP2 hydrolysis evoked by a desensitization-defective mutant of the NKA receptor correlates with prolonged immobilization of cortical actin (Figure E, n = 6). These results show that cortical actin dynamics correlate well with membrane PIP2 content, providing evidence for the causal relationship that was hypothesized in the recent literature. To further address that actin motility changes are secondary to PIP2 hydrolysis, we inhibited PIP2 resynthesis by blocking PIP kinases with a low dose (1 --4 muM) of phenylarsine oxide (PAO). As shown in Figure F, pretreatment with this drug did not influence basal [PIP2] or actin motility, but it completely blocked recovery of PIP2 to basal levels after BK-induced PLC activation . This was paralleled by persistent reduction of actin dynamics. Similar observations were made with the PIP kinase inhibitor quercetin, which blocks ATP binding to the kinase domain, and with H2O2 (; J. Halstead and N. Divecha, personal communication), which is thought to disrupt PIPkinase function by modifying critical thiol groups outside the ATP binding domain. Figure 2 | Colocalization of GFP-PH and DiI. Colocalization of GFP-PH and DiI. Living N1E-115 cells expressing GFP-PH were stained with the lipophilic membrane dye DiI. (A) Images of the fluorescent protein and DiI were collected from both medial (top panels) and basal (bottom panels) sections, using a confocal microscope. Scale bar, 5 mum. (B) Pixel intensity for DiI was plotted against GFP-PH intensity in a scatter plot. Note the difference from the scatter plots in Figure , where one fluorophore is imaged at two points in time. GFP-PH Distribution at the Plasma Membrane Is Not Homogenous | It has been hypothesized that local differences in membrane PIP2 content may transmit extracellular signals into local cytoskeletal changes. Do such local enrichments or depletions in PIP2 content (PIP2 patches; note that although for brevity we will often mention "PIP2 enrichments" our analyses were equally focused on PIP2 enrichments and decreases) at the plasma membrane exist? To study the spatial distribution of PIP2, cells expressing a GFP-tagged pleckstrin homology domain derived from PLCdelta1 were imaged on the confocal microscope. We noted that GFP fluorescence along the plasma membranes does not appear to be homogenous. Rather, in several cell types, including N1E-115 mouse neuroblastoma cells (Figure A), NIH-3T3 mouse fibroblasts and HEK293 human embryonic kidney cells, GFP fluorescence shows distinct bright patches that are two- to threefold more intense than the rest of the membrane. Bright patches can be observed in medial sections through the cells (top left panel), where they often colocalize with membrane ruffles and lamellae. Patches are also particularly apparent in the basal membranes of cells grown on coverslips, usually displaying a slender, elongated shape (bottom left panel). Similar observations were recently reported by . When imaged in living cells these patches appear highly dynamic: over time, individual patches may disappear, show directed movements, and occasionally branch. These structures further colocalize with F-actin but not with vinculin or other components of focal adhesions. Because many actin-binding proteins can interact with PIP2 in vitro, GFP-PH patches were interpreted to represent local concentrations of PIP2 . However, alternative explanations for the local concentration of GFP-PH have not been addressed. GFP-PH Staining Pattern Reflects Membrane Content | We set out to address the possibility that bright GFP-PH patches reflect local increases in membrane area, due to local membrane folding. Cells expressing GFP-PH were stained with the lipophilic membrane dye diphenyl-DiI and simultaneously imaged for GFP and dye fluorescence on a confocal microscope. Images were collected at the basal membrane and at medial sections through the cell. Strikingly, we observed strong colocalization of GFP and DiI fluorescence in all sections in N1E-115 cells (Figure A), NIH-3T3 cells, and several other cell types. A colocalization analysis was carried out by constructing scatter plots to compare pixel intensities of GFP-PH and DiI (Figure B). In these plots, any structures present in the GFP-PH, but not in the DiI image, will be apparent as off-diagonal clusters of dots. No evidence was found for GFP-PH enrichment beyond the level predicted by lipid mass as detected by DiI fluorescence (>35 images analyzed). Conceivably, however, the DiI dye might localize preferentially to places enriched in PIP2 or F-actin. We therefore used a panel of different membrane markers with widely different physicochemical properties, including lipophilic dyes that intercalate in the lipid doublelayer, phospholipids with fluorescently labeled acyl chain or headgroup, and fluorescent proteins, targeted to the membrane with lipid anchors . All membrane markers colocalized with GFP-PH, strongly suggesting that patches represent sites with increased membrane content. Table 1 | Membrane markers used in colocalization studies Strong colocalization of GFP-PH with membrane markers was not restricted to the PH domain from PLCdelta1, because similar observations were obtained using the PIP2-specific PH domain derived from PLCdelta4, and a PIP2-specific mutant (E41K) derived from the PH domain of Bruton's tyrosine kinase (Btk; ). Furthermore, bradykinin-induced PLC activation caused GFP-PH patches to disappear, whereas the DiI staining pattern remained unaltered. As shown for medial and basal confocal sections in Figure , GFP-PH returned to the exact same sites to colocalize again with the DiI patches after resynthesis of PIP2. Taken together, these data demonstrate that patches enriched in GFP-PH are the consequence of locally increased membrane area (i.e., folds and ruffles), rather than of local PIP2 enrichment. Figure 3 | Localization of GFP-PH and DiI during PIP2 hydrolysis. Localization of GFP-PH and DiI during PIP2 hydrolysis. N1E-115 cells expressing GFP-PH were stained with DiI and imaged on the confocal microscope in medial (A) and basal (B) sections. Bradykinin (BK, 1 muM) induced PIP2 breakdown, resulting in translocation of GFP-PH. After resynthesis of PIP2 (at 120 s), GFP-PH returned to the membrane. In contrast, DiI patches (see arrow) remain at the membrane during PIP2 hydrolysis. Scale bar: (A) 5 mum, (B) 1 mum. Membrane Folds at GFP-PH Patches Can Be Directly Visualized | Although the above results indicate that GFP-PH patches are sites of membrane folding, detailed confocal imaging studies failed to directly visualize folds in a considerable subset of the patches. This is perhaps not surprising, because the resolution of the confocal microscope is limited by the objective point spread function (PSF), this is the Gaussian intensity profile that is detected when a true point source is imaged with an objective. The PSF of the best objectives are close to two orders of magnitude larger than the thickness (5 nm) of the lipid bilayer. Two types of experiments were performed to investigate whether subresolution membrane folds cause the bright fluorescence in all membrane patches. First, cells expressing GFP-PH were swollen by a hypotonic shock. The medium was diluted from 350 mOsmol to a final value of 120 mOsmol, while confocal images were continuously collected (Figure A). This caused an increase in cell volume of 40 --50% that led to straitening out of the membrane with consequent disappearance of bright GFP-PH patches. Disappearance of bright patches was not due to swelling-induced PIP2 hydrolysis, because [PIP2] in these cells remained constant (Figure B). These experiments strongly argue that bright fluorescent patches are in fact folds in the lipid bilayer. Figure 4 | GFP-PH patches disappear after osmotic swelling. GFP-PH patches disappear after osmotic swelling. (A) Confocal images of NIH-3T3 cells expressing GFP-PH, subjected to hypotonic swelling. Osmolarity of the medium was adjusted from a basal value of 350 mOsmol (1) to 235 (2) and 120 mOsmol (3). Ionomycin, 5 muM, was added (4) to cause complete translocation of the fluorescent proteins. (B) PIP2 degradation was essayed by FRET in a single cell, subjected to the same protocol. Note that although swelling eliminates the patches, total PIP2 in the cell is virtually unaltered. Scale bar, 5 mum. Second, we set out to directly visualize subresolution membrane folding, using a numerical approach to increase the axial resolution. In these studies, we recorded stacks of images at 40- or 80-nm axial distance, thus oversampling the axial resolution (PSF) of the objective up to 25 times (see Figure ). The intensity of a small region of interest (ROI) was plotted against the axial position for the images in this stack. Because the thickness of the membrane (similar5 nm) can be ignored, the resulting curve basically reconstructs the axial PSF (Figure C). When such reconstructed axial PSFs are compared for ROIs inside (blue mask) and just outside (red mask) of the GFP-CAAX patches, the normalized curves consistently show small offsets, indicating differences in Z-position of the fluorescent membrane. It should be noted that by fitting the PSFs, the Z-axis offset can in fact be estimated with a precision considerably higher that the axial step size (see MATERIALS AND METHODS). By applying this technique on a pixel-by-pixel basis to the image stack, the three-dimensional surface profile of the basal membrane could be visualized with similar15-nm axial resolution (Figure D). In these images, upward and downward protrusions measuring between 15 and 150 nm in the basal membrane are observed (Figure D) that correspond to the bright GFP-CAAX patches. Therefore, these data directly demonstrate that patches in effect represent submicrometer peaks and valleys in the landscape of the basal membrane. Figure 5 | Direct visualization of subresolution membrane folds at the patches. Direct visualization of subresolution membrane folds at the patches. (A) Stack of confocal XY-images, taken 80 nm apart in Z direction, showing the basal membrane of GFP-CAAX expressing N1E-115 cell. (B) Using image analysis software, an area was manually assigned within (blue) and just outside (red) a patch, and applied to the entire image stack. Scale bar, 2 mum. (C) Plotting the measured mean intensity within the blue and red area versus the image number in the stack reconstructs the axial pointspread function of the objective. The PSF (full-width at half-maximum) of the used 63x oil 1.32 NA planapochromatic objective is 1.05 mum, at a confocal pinhole setting of 0.58 mum and at emission lambda = 525 nm. By analysis of the offset of the normalized curves, the Z-position of the membrane can be estimated with high precision. (D) The three-dimensional surface profile of a patch is visualized with similar15 nm resolution by applying this analysis on a point-by-point basis to the image stack. Further analysis of these structures is described in the Supplementary Data section, online. It is demonstrated that GFP-PH patches do not represent sites of cell adhesion, such as focal adhesions. In addition, we show that motility of GFP-PH patches is potently inhibited by cytochalasin D and myosin light chain kinase blockers, indicating that patches are the result, rather than the cause, of local actin-dependent forces. Diffusion Limits the Establishment and Maintenance of PIP2 Gradients | Summarizing the above results, GFP-PH labeling indicates that spatial differences in the concentration of unbound PIP2, at least at a micrometer scale, do not exist in our cells. However, there might be [PIP2] differences on a more global scale, e.g., between the leading and trailing edge of polarized cells or in organelles such as lamellipodia. For 3'-phosphorylated phosphatidylinositols, it has been shown that such gradients can exist in chemotactic and phagocytic cells . At what scale can spatial differences in [PIP2] be induced in the plasma membrane of living cells? Gradients in [PIP2] are the result of local synthesis and breakdown, combined with lateral diffusion of the lipid in the membrane. We studied the diffusion kinetics of bodipy-labeled PIP2 in N1E-115 cells using laser-induced photobleaching (see MATERIALS AND METHODS). As can be seen in Figure A, the fluorescence in micrometer-sized photobleached spots in the plasma membrane completely recovered within seconds. Thus, it is likely that at this scale, induced gradients in [PIP2] rapidly dissipate by diffusion. The observed recovery rate is similar to those of other freely diffusible membrane labels, including DiI and bodipy-labeled phosphatidylcholine (Figure A). Published diffusion coefficients (D) for these labels average similar1 --2 mum2/s , in reasonable agreement with the value reported recently for PIP3 (0.5 mum2/s; ). These D values therefore predict that gradients in PIP2 at a larger scale may occur. To test whether PIP2 concentration differences can be induced and detected on this scale, we used focal stimulation with Neurokinin A from a micropipette . This caused a rapid initial translocation of GFP-PH at the stimulus site that subsequently spread to the neighboring membrane. In particular in cases where diffusion is restricted, such as in neurites (Figure A), sustained local decreases in [PIP2] could reliably be evoked (Figure B). Taken together, these observations show that agonists can induce PIP2 differences along the membrane, whereas at a micrometer-scale lateral diffusion limits maintenance of such gradients. Figure 6 | Local PIP2 gradients are limited by diffusion. Local PIP2 gradients are limited by diffusion. (A) Cells were labeled with DiI, bodipy-TR PtdIns(4,5)P2, or bis-bodipy-FL C11-phosphatidylcholine. Spots were bleached at the plasma membrane in neurites (red trace) or the soma (black trace) with a 0.2-s pulse of 488- or 568-nm laser light, and recovery was quantified from confocal images. Note that recovery upon spot photobleaching in neurites (red lines) was reduced by orders of magnitude. (B) N1E-115 cell expressing GFP-PH and a desensitization-defective NK2 receptor was imaged at the indicated time points using a confocal microscope. PLC was stimulated locally in a developing neurite by a precisely confined stream of neurokinin A (NKA) generated by means of an application (A) and a suction pipette (S). The resulting PIP2 degradation was monitored as GFP-PH translocation (see arrow). Similar results were obtained with bradykinin. The intensity of the images is color-coded with the "glow" color look-up table of TCS software. Scale bar, 5 mum. DISCUSSION : In this study, we used GFP-PH as in vivo PIP2 tag to investigate the hypothesis that local pools of free PIP2 at the plasma membrane (patches) spatially regulate cytoskeletal remodeling. The existence of PIP2 patches and their involvement in the local regulation of cellular physiology (e.g., actomyosin remodeling, vesicle budding, etc.) has been hypothesized by several groups (; ; ; ; for review see ; ). In addition, some data have been presented that seems to support this notion . At odds with these reports, the detailed spatial analysis presented here indicates, within the limits of optical microscopy (similar200 nm), that in all cell lines checked the GFP-PH labeling pattern simply reflects the amount of membrane rather than local PIP2 enrichment. We have previously shown that the PH domain from PLCdelta1 detects PIP2 (rather than IP3) in vivo and that membrane-bound GFP-PH is in rapid (similar1 s) equilibrium with a cytosolic pool of approximately equal size. Because in the present article it was shown that gradients in [PIP2] could be induced and detected, it is fair to argue that GFP-PH is capable of reporting, at least qualitatively, the local free PIP2 concentration in single, living cells. From our failure to observe micrometer-sized PIP2 patches, we conclude that spatial differences in PIP2 regulation of the cytoskeleton do not exist at this scale. The above-mentioned contrast between our findings and literature reports deserves further attention. Because this study addressed the hypothesized spatial regulation of the cytoskeleton by PIP2, we used detection by decoration with GFP-PH to visualize the distribution of the free (unbound) pool of this lipid along the membrane. PIP2 that is bound to proteins at the plasma membrane constitutes a second pool that, likely, may display local molar enrichment or gradients due to, respectively, clustering at specific sites or gradients of the PIP2-interacting proteins. In this case, PIP2 enrichments are the result, rather than the cause of protein clustering, and because such enrichments are not available for decoration with GFP-PH, our studies will fail to detect them. However, lipid-biochemical approaches, such as those used to study rafts, obviously would. These rafts, which are detergent-insoluble microdomains in the membrane, contain signal-transducing proteins and may have roles in membrane trafficking and signaling (reviewed by ). The resolution of the here used imaging techniques is not sufficient to investigate rafts (<70 nm; ; ). Nevertheless, the observed highly homogeneous distribution of GFP-PH along the membrane suggests that hypothetical rafts are either very abundant and evenly distributed along the membrane or they do not contain excess free PIP2. Reports that rely on labeling of PIP2 with specific antibodies require fixation and permeabilization of the cells . However, this procedure may well have caused artifacts . For example, it was observed by Laux and colleagues that the size of PIP2 clusters observed in monkey kidney epithelial cells depends on the fixation method. Furthermore, in a recent electron microscopy study that used immunogold labeling of PLCdelta1-PH-GST to visualize PIP2 at the membrane, it was shown that lipids are not efficiently fixed with aldehyde at room temperature. Interestingly, these authors showed good labeling results at 0C but did not observe PIP2 patches at the membrane. The published literature generally lacks rigorous colocalization analysis to exclude the possibility of "apparent PIP2 enrichment" by membrane folding. In our studies, we used a panel of 10 different membrane stains with widely different physicochemical properties, and we have used three different GFP-tagged PIP2-specific pleckstrin homology domains. Strong colocalization of GFP-PH with lipid dyes was observed in all cell types tested, irrespective of the phase of the cell cycle and of growth conditions (i.e., both when cultured in serum-free and in serum-containing medium). Furthermore, because our data directly contradict the interpretation put forward by , we have addressed this issue by additional, independent approaches. These included experiments that involve hypotonic cell swelling, interference reflection microscopy, and the here-introduced technique of reconstructed axial PSF imaging. Together, the experiments show conclusively that GFP-PH patches represent submicroscopic folds and protrusions in the membrane that are caused by actomyosin-based forces, rather than local PIP2 enrichments. In addition, in vivo FRAP experiments were carried out using fluorescently labeled lipids that incorporated in the plasma membrane after addition to the culture medium. These experiments should be interpreted with some caution, because membrane insertion and possible binding to proteins of the exogenously added PIP2 has not been characterized extensively. With this caveat, the analysis revealed that diffusion of PIP2 occurs freely at a rate comparable to that of phosphatidylcholine, showing that micrometer-sized patches will dissipate within seconds. The recently reported gradients in 3' phosphorylated phosphoinositides (3' PI) in chemotactic and phagocytic cells were analyzed by Haugh and coworkers. In a model of local synthesis and breakdown of 3' PI, the experimentally revealed unhindered diffusion along the membrane (D = similar0.5 mum2/s) and the lifetime of similar40 s of 3' PI were consistent with the gradients observed in vivo. Our FRAP experiments indicate very similar diffusion rates for PIP2. Furthermore, from the observation that agonist-induced PIP2 breakdown and subsequent resynthesis can take place well within 2 min, we conclude that under these conditions the PIP2 lifetime may be short enough to enable buildup of gradients. Taken together, our results challenge the view that PIP2 regulation of the actin cytoskeleton is localized at a micrometer scale. However, on a larger scale and in cases where diffusion is limited, sustained PIP2 gradients may exist for prolonged periods of time. We also observed that agonist-induced PIP2 changes as detected using the FRET assay correlate well with the dynamics of cortical actin as quantified using the assay that was introduced in this article, both with respect to the magnitude and the time course of the response. A possible delay of the actin response was not detected, although it should be noted that the temporal resolution of the assay is limited to similar30 s (see MATERIALS AND METHODS). These observations provide experimental evidence for the actin-modulatory role of PIP2 that has been widely hypothesized (e.g., ; ; ; ), based on the PIP2-binding properties displayed by many actin-binding proteins. Furthermore, this correlation held true when agents were used that circumvent receptor activation to cause PIP2 decrease (the PIPkinase inhibitors PAO and quercetin, and H2O2; ; J. Halstead and N. Divecha, personal communication), suggesting that the observed actin changes are not due to signaling events independent from PIP2 metabolism. In support of this viewpoint, a recent report from Cunningham and coworkers showed that incubation of cells with a membrane permeable peptide with potent PIP2-binding activity inhibited cell migration . Indeed, we observed that the transient drop in F-actin dynamics during agonist-induced PIP2 breakdown was mirrored in loss of motility of the leading edge in fibroblasts. Thus, our data show, for the first time, that the PIP2 decreases triggered by agonist-induced PLC activation suffice to influence cortical actin motility in vivo. Finally, a notable observation is the existence of highly dynamic, actin-rich structures in the basal membrane of nonmigratory cells. These structures represent places of close membrane-substrate proximity but do not appear to relate to sites of cell adhesion. Whereas it is striking that their dynamic behavior appears PIP2 dependent, a possible function awaits future investigations. Backmatter: Abbreviations used: : CFP = cyan fluorescent protein FRAP = fluorescence recovery after photobleaching FRET = fluorescence resonance energy transfer GFP = green fluorescent protein GPCR = G protein-coupled receptor PH = pleckstrin homology PLC = phospholipase C PIP2 = phosphatidylinositol(4,5) bisphosphate PSF = pointspread function YFP = yellow fluorescent protein PMID- 12221132 TI - Calcineurin, a Calcium/Calmodulin-dependent Protein Phosphatase, Is Involved in Movement, Fertility, Egg Laying, and Growth in Caenorhabditis elegans AB - Calcineurin is a Ca2+-calmodulin --dependent serine/threonine protein phosphatase that has been implicated in various signaling pathways. Here we report the identification and characterization of calcineurin genes in Caenorhabditis elegans (cna-1 and cnb-1), which share high homology with Drosophila and mammalian calcineurin genes. C. elegans calcineurin binds calcium and functions as a heterodimeric protein phosphatase establishing its biochemical conservation in the nematode. Calcineurin is expressed in hypodermal seam cells, body-wall muscle, vulva muscle, neuronal cells, and in sperm and the spermatheca. cnb-1 mutants showed pleiotropic defects including lethargic movement and delayed egg-laying. Interestingly, these characteristic defects resembled phenotypes observed in gain-of-function mutants of unc-43/Ca2+-calmodulin --dependent protein kinase II (CaMKII) and goa-1/Go-protein alpha-subunit. Double mutants of cnb-1 and unc-43(gf) displayed an apparent synergistic severity of movement and egg-laying defects, suggesting that calcineurin may have an antagonistic role in CaMKII-regulated phosphorylation signaling pathways in C. elegans. Keywords: INTRODUCTION : Calcineurin (CaN), a protein phosphatase 2B (PP2B), is a serine/threonine phosphatase under the control of Ca2+/calmodulin . Although CaN is a member of a family of protein phosphatases, it is structurally and functionally distinct from alkaline and acid phosphatases . Calcineurin is a heterodimer of an similar60-kDa catalytic subunit, calcineurin A (CNA), and a 19-kDa regulatory subunit, calcineurin B (CNB; ), and the two-subunit structure is well conserved from yeast to human. CaN is abundantly expressed in the brain and broadly distributed in nonneural tissues as well . Among its several functions in controlling intracellular Ca2+ signaling, CaN participates in gene regulation and external signal-mediated biological responses in many organisms and in many cell types . CaN functions have been extensively studied in the yeast, Saccharomyces cerevisiae. To investigate the biological role of CaN, immunosuppressant drugs Cyclosporin A (CsA) and FK506 were used to inhibit CaN function . Calcineurin has been shown to regulate Ca2+ pumps and exchangers to maintain Ca2+ homeostasis . Calcineurin is also known to regulate adaptation to high salt stress . However, in higher animals, it is better known to regulate the transcription of the T-cell growth factor, interleukin-2 . Dephosphorylation of the transcription factor NF-ATp (nuclear factor --activated T cells) by CaN is required for the translocation of NF-AT from the cytoplasm to the nucleus, in response to an increased intracellular Ca2+ level. Calcineurin also plays a role in programmed cell death and in hippocampal long-term depression . Furthermore, studies reveal that CaN plays a critical role in the pathogenesis of hypertrophic cardiomyopathy . Thus, calcineurin as a key signaling molecule has been shown to be involved in diverse types of physiological processes. Caenorhabditis elegans has been an ideal model to study gene functions especially at the organism level using a genetic approach. Moreover, C. elegans has been useful for identifying interactions between molecules in biochemical signaling pathways that are associated with a certain behavior or phenotype. In this study, we have identified and characterized the C. elegans homologue of calcineurin B, which binds calcium and functions together with calcineurin A as a heterodimeric protein phosphatase. Null mutants of calcineurin B showed multiple adverse phenotypes including defects in locomotion and egg laying. Interestingly, these phenotypes are quite similar to those observed in gain-of-function mutants of unc-43, which encodes the Ca2+-calmodulin --dependent protein kinase CaMKII. Recently, a G-protein signaling pathway regulated by unc-43 has been found to be involved in locomotory and egg-laying functions . Our results describing the relationship between the cnb-1 null mutant and mutants of unc-43 indicate that calcineurin and CaMKII may have opposing and complementary functions in this G-protein signaling pathway in C. elegans. MATERIALS AND METHODS : C. elegans Strains | The following strains were obtained from the Caenorhabditis Genetics Center (CGC) at the University of Minnesota, Duluth, MS: Bristol N2, CB224 dpy-11(e224)V, CB1482 sma-6(e1482)II, MT1092 unc-43(n498)IV, unc-43(n1186)IV, and CB1282 dpy-20(e1282) IV. KJ300 cnb-1(jh103) was isolated by reverse genetics method . cnb-1(ok276) was isolated by the C. elegans Knockout Consortium (G. Molder, Oklahoma). Worm breeding and handling were conducted as described . Cloning of C. elegans Calcineurin A and Calcineurin B cDNAs and Northern Analysis | To obtain full-length cDNA clones of C. elegans calcineurin A (cna-1) or calcineurin B (cnb-1), cDNA library screening was conducted following plaque hybridization procedure . A mixed-stage worm cDNA library (kindly provided by P. Okkema and A. Fire) was probed with the partial cDNA clones, yk375h10 or yk496e12 (obtained from Y. Kohara). The largest inserts of 2869 base pairs and 943 base pairs for cna-1 and cnb-1, respectively, were sequenced and confirmed by Northern analysis. For Northern blotting, total RNA was prepared from staged animals and transferred to a Zeta probe membrane (, Hercules, CA) as described . The membrane was hybridized in hybridization buffer (0.25 M Na2HPO4, pH 7.2), 0.25 M NaCl, 1 mM EDTA, 7% SDS, 50% Formamide, 5% dextran sulfate, and 100 mug/ml denatured salmon sperm DNA at 42C with random-primed 32P-labeled probes. Exposure of the blots after high stringent washing in 0.5x SSC and 0.1% SDS was performed on x-ray films (Fuji) overnight at -80C. Expression and Purification of GST-CNA-1 and GST-CNB-1, Ca2+-binding, and Yeast Two-hybrid Assays | Complementary DNAs encoding the entire open reading frames of cna-1 or cnb-1 were subcloned in pGEX4T-3 or pGEX4T-1 (Pharmacia, Piscataway, NJ), encoding GST-CNA-1 or GST-CNB-1, respectively. Syntheses of GST-fused proteins were carried in Escherichia coli strains (BL21) at 30C for CNA-1 and at 37C for CNB-1 in the presence of 1 mM IPTG. Cells were harvested and sonicated according to the methods described . Ca2+-binding assay for CNB-1 was performed as described earlier . Purified GST-CNB-1 (see above) was resolved by 12% SDS-PAGE, transferred to PVDF membrane, and then probed with 45Ca2+ . Purified GST, bovine calmodulin (CaM; Sigma), and C. elegans calsequestrin (CSQ-1) were used as controls. Wild-type cna-1 and cnb-1 cDNAs, covering the entire open reading frames, were fused in-frame to the GAL4 DNA-binding domain and GAL4 activation domain of the yeast vectors pAS2 --1 and pACT2 (, Palo Alto, CA) to produce the plasmids CJ1 and CJ2, respectively. Plasmids were transformed into the yeast strain AH109 according to the manufacturer's protocol . Transformants were plated on synthetic dropout (SD) media lacking Trp and Leu. Interaction assays were conducted on plates containing 5 mM 3-aminotriazole (3-AT) in SD without Trp, Leu, His, and Ade in absence or presence of 2 mM CaCl2. In Vitro Phosphatase Assay | Equimolar concentrations (0.016 nmoles each) of purified GST-CNA-1 and GST-CNB-1 (see above) were used to conduct phosphatase assay (Promega) using a phosphopeptide (100 muM) as a substrate. To test the Ca2+ dependency of calcineurin, 0.2 mM EGTA (Ca2+ chelator) was used in the reactions. Inhibition of phosphatase activity by the immunosuppressant drug, cyclosporin A, CsA (Calbiochem, La Jolla, CA), was tested by preincubating GST-CNA-1 and GST-CNB-1 together with 1 muM each of CsA and bovine cyclophilin (Sigma, St. Louis, MO) at 4C for 1 h. The dephosphorylation of the phosphopeptide was determined spectrophotometrically at 595 nm. The optical density was further converted to pmol phosphate release/min/mug protein using appropriate standards supplied with the kit. Construction and Expression of gfp Fusion Constructs | Two cosmid clones C02F4 and F55C10 were obtained from A. Coulson (The Sanger Center, UK). To check the reporter gene (green fluorescent protein, gfp) expression, promoterless GFP vectors, pPD95.70 containing the nuclear localization signal (NLS), pPD95.79 and pPD95.75 (vectors provided by A. Fire) were used in the present study. All constructs (pJJ001 --pJJ004) were generated as translational fusions with gfp. Microinjection using pRF4 (dominant rol-6) as a transformation marker was performed as described by . Figure 1 | Genomic organization of calcineurin genes in C. elegans. Genomic organization of calcineurin genes in C. elegans. (A) Genetic and physical maps of the cnb-1 region. The cnb-1 locus relative to the nearby genetic markers on the gene cluster region of LG V is shown. cnb-1, which is mapped to the cosmid F55C10 (Accession No. ), and the neighboring cosmids are indicated. The predicted F55C10.1 gene (cnb-1) structure has been revised and is presently shown to contain six exons (black boxes). (B) Northern blot analysis detects a 1-kb band in RNA samples of wild-type N2 worms. (C) Amino acid sequence alignment for calcineurin B of C. elegans (CECNB-1), Drosophila (DMCNB-1, Accession No. ), human (HSCNB-1, Accession No. NP000936), and yeast (SCCNB-1, Accession No. NP_012731). Regions of identity among calcineurin B homologues are boxed. Four Ca2+-binding sites are shown (I --IV). The myristoylation signature is indicated by a black bar. Immunofluorescence and Immunogold Microscopy | Wild-type C. elegans was immunostained as described . For gonad immunostaining, gonads were extruded by decapitating adult C. elegans and fixed with 3% formaldehyde, 0.1 M K2HPO4 (pH 7.2), for 1 h and postfixed with cold (-20C) 100% methanol for 5 min. Antibody incubations and washes were performed as described . Anti --CNA-1 mouse monoclonal and anti --CNB-1 rabbit polyclonal antibodies were used as primary antibodies. TRITC-conjugated anti-mouse (Sigma) and FITC-conjugated anti-rabbit IgG (Santa Cruz Biotechnology, Santa Cruz, CA) were used as secondary antibodies. Sperm immunostaining was conducted according to the methods described previously . Immunogold staining with anti --CNB-1 antibody for N2 worms was carried out as described previously . Samples were examined under a transmission electron microscope (Jeol 1200 EXII, Peabody, MA). Isolation of a cnb-1 Deletion Mutant from a Mutagenized DNA Library | TMP (Trimethylpsoralen)/UV method was used to generate C. elegans deletion mutant. Screening of mutants from the mutagenized DNA library was carried out by a nested PCR-based method and subsequent sib selections as described . Primers were designed based on the predicted sequences spanning the full genomic DNA of cnb-1: outer upstream primer (5'-ACA TTC TAC TAC ATT CTG GCT GTG TGA TCC-3') and downstream primer (5'-ATG AGC ATC ATT TAT TTG GCG GAC C-3'), inner upstream primer (5'-AAG CCC TCT GCT GGA CTG CTG TCC ACC-3'), and downstream primer (5'-AAT GCG AGG AAA CGC TTC CCA ATT GGC-3'). A homozygous line of animals with a 950-base pair deletion relative to the wild-type was isolated. This animal was outcrossed six times to wild-type animals to establish the strain KJ300 cnb-1(jh103) and was used in subsequent analysis. Deletion region for the cnb-1 hermaphrodites was determined by nested PCR followed by sequencing the PCR products. Construction of Double Mutant Strain and Phenotype Analysis | Double mutants of unc-43(n498);cnb-1(jh103), unc-43(n1186);cnb-1(jh103), and goa-1(n1134);cnb-1(jh103) were constructed by standard genetic methods. dpy-20(e1282) was used as a genetic marker. PCR was used to detect the cnb-1 mutant. The brood size of N2, cnb-1, unc-43, and unc-43(n498);cnb-1(jh103) hermaphrodites was determined by placing individual worms on seeded plates and allowing self-fertilization at the indicated temperature. The P0 mother was then transferred to a fresh plate at 24-h intervals for each of the next 4 days. Total F1 progeny on the plates were counted. The brood size of crossed progeny was determined by placing a single N2 or cnb-1 hermaphrodite with three wild-type males on seeded plates and allowing crossing at 20C for 2 d. Total F1 progeny and the number of male progeny on the plates were counted. Body length, body width, and uterine embryos were carefully examined under a dissecting microscope. L4-stage larvae were transferred to a new plate and allowed to grow for 36 h. The resulting 1-d adult worms were then examined for phenotypes. Movement of animals was examined by placing five adult hermaphrodites on a bacterial lawn, and after 5 min the tracks made by the worms were photographed. Serotonin-mediated egg-laying phenotypes were examined as previously described . Levamisole-mediated egg-laying phenotypes were examined in the same way at concentrations ranging from 25 to 100 mum. Sperm morphology was analyzed in hermaphrodites as previously described by hand-dissecting hermaphrodite spermatheca in sperm medium (50 mM NaCl, 25 mM KCl, 5 mM CaCl2, 1 mM MgSO4, 5 mM HEPES, 10 mM dextrose, 4 mM levamisole, pH 7.8). The spermatheca was observed with a Zeiss Universal microscope equipped with Nomarski optics. Transformation Rescue | To test for rescue of cnb-1(jh103) homozygotes, cnb-1 genomic DNA cloned in pGEM-T Easy vector (Promega) and csq-1::gfp fusion construct as a transformation marker at final concentrations of 50 and 50 mug/ml, respectively, were coinjected into cnb-1 mutant animals using standard methods . The injected parents were allowed to self-fertilize. Individual GFP-expressed progeny were then picked and examined for phenotypes. RESULTS : Characterization of the C. elegans Calcineurin B (cnb-1) Gene | C. elegans calcineurin B-like genes have been physically mapped to the cosmids F55C10 and F30A10, respectively. However, alignment studies and molecular cloning (this study) have revealed that F55C10.1 is the putative calcineurin B (cnb-1), the regulatory subunit of the protein phosphatase 2B with four EF-hand motifs for Ca2+ binding (see below). The cnb-1 locus is physically mapped to the gene cluster region of chromosome V (LGV) on the cosmid F55C10 that corresponds to the region between unc-42 and sma-1 loci on the genetic map (Figure A). Northern blot analysis confirmed a single 1.0-kb mRNA transcript of cnb-1 (Figure B). C. elegans calcineurin B, as in other species, is also a small protein of 171 amino acids exhibiting high sequence homology with other organisms (80% identity with human and Drosophila calcineurin B and 58% with the yeast; Figure C). A putative initiation Met (ATG) codon is located at nucleotide position 51, which is surrounded by a relatively good consensus sequence for translation also found in other C. elegans genes . Additionally, the cDNA appears to contain a partial SL1 leader sequence at the 5' end. The C. elegans calcineurin A (cna-1) gene was recently identified by . The tax-6 locus was shown to encode calcineurin A, and our results confirmed the cna-1/tax-6 sequence and gene structure from previously obtained results . CNB-1 Is a Ca2+-binding Protein and Enhances CNA-1 Phosphatase Activity | It is well known that CaN binds Ca2+ and calmodulin (CaM), as originally shown by . Calcineurin B, the regulatory subunit of CaN binds 4 molecules of Ca2+ with high affinity (Kd <= 10-6 M), and has sequence homology with CaM and troponin C, two other calcineurin binding proteins . Recombinant GST-CNB-1 had strong Ca2+-binding activity (Figure A, lane 4) whereas GST alone showed no Ca2+-binding activity (Figure A, lane 1). Thus, Ca2+ overlay experiments confirmed that CNB-1 has high affinity for Ca2+ as in other Ca2+-binding proteins such as C. elegans calsequestrin CSQ-1 and bovine CaM (Figure A, lane 2 and lane 3, respectively). Figure 2 | Biochemical characterization of C. elegans Biochemical characterization of C. elegans calcineurin. (A) cnb-1 is a Ca2+-binding protein. Ca2+-binding assay was performed with GST (glutathione-S-transferase) (lane 1), C. elegans calsequestrin, GST-CSQ-1 (lane 2), bovine CaM (lane 3), and GST-CNB-1 (lane 4). 45Ca2+-labeled proteins are shown by the dark signals on the x-ray film. GST (lane 1) is used as a negative control. A corresponding gel shown in the right panel is stained with Coomassie Blue. Molecular weight markers (in kDa) are denoted by arrows. (B) Interaction between CNA-1 and CNB-1 in the yeast two-hybrid system. Growth of yeast transformants on SD -Trp -His -Leu -Ade with 5 mM 3-AT (panel 2) and 5 mM 3-AT + 2 mM CaCl2 (panel 3) are shown. Plates supplemented with 5 mM 3-AT + 2 mM CaCl2 select for a positive interaction between the fusion proteins. Positive interaction between AD/T-antigen and DNA-BD/p53 fusion proteins and the negative or no interaction between AD/T-antigen and DNA-BD/lamin C were used as positive (+ control), and negative controls (- control), respectively. The array of the yeasts containing the different constructs is indicated in the schemes shown in panel 1. (C) Serine/threonine phosphatase activity of CNA-1. Dephosphorylation of the phosphopeptide by GST, GST-CNA-1 (CnA), and GST-CNA-1 + GST-CNB-1 (CnAB) are shown. Ca2+-dependence of the phosphatase activity (-Ca2+) was determined by adding 0.2 mM EGTA (Ca2+-chelator) during the CnAB reaction. Inhibition of phosphatase activity was determined by adding 1 muM each of CsA and bovine cyclophilin (+CsA) in the CnAB reaction mixture. Assays were performed in triplicate with equimolar concentrations of CNA-1 and CNB-1. (D) Western blot analysis with anti --CNA-1 and anti --CNB-1 antibodies. Worm protein extracts prepared from N2 wild-type worms or cnb-1(jh103) mutants (CNB, lane 2). The band is absent in the cnb-1(jh103) deletion worms indicating they are functionally null mutants. Calcineurin is a tight heterodimer composed of the catalytic A subunit and the regulatory B subunit . This interaction is necessary for the function and stabilization of the enzyme. We utilized the yeast two-hybrid system to check if CNA-1 interacted with CNB-1. As shown in Figure B, the subunits interacted with each other in the presence of Ca2+ as indicated by the growth of colonies in the selective media (compare panels 2 and 3). To further test if C. elegans calcineurin exhibits phosphatase activity in vitro, a serine/threonine phosphatase assay was conducted with purified full-length GST-CNA-1 and GST-CNB-1 expressed in E. coli. It was previously demonstrated that GST does not interfere with or possess phosphatase activity . We carried out experiments with or without GST-CNB-1 to check whether the rate of dephosphorylation can be regulated in the absence or presence of the regulatory subunit, and at the same time tested whether there was basal activity of GST-CNA-1. As shown in Figure C, GST-CNA-1 alone showed a dephosphorylation activity (similar10 pmol phosphate/min/mug protein), whereas addition of GST-CNB-1 elevated the activity by similar2-fold indicating that the full activity of the enzyme requires both subunits of calcineurin. Addition of EGTA (a Ca2+-chelator) abolished the phosphatase activity to control levels indicating that the phosphatase required Ca2+ for its activity (Figure C). Furthermore, we tested if the immunosuppressant, cyclosporin A (CsA), affects the phosphatase activity. Cyclophilin-CsA complex has long been known to block the phosphatase activity of calcineurin . Previously, several cyclophilin (cyp) genes had also been cloned and characterized in C. elegans indicating once again a well-conserved role(s) for cyclophilin in signal transduction and protein folding . In the present study, we used bovine cyclophilin with CsA to test if C. elegans calcineurin activity is affected. As expected, the phosphatase activity of CNA-1 was completely inhibited (Figure C) in the presence of cyclophilin-CsA implicating a conserved phenomenon of phosphatase inhibition by the immunosuppressant drug. Expression and Localization of Calcineurin in C. elegans | We examined the temporal and spatial pattern of cna-1 and cnb-1 expression using the gfp reporter driven by cna-1 or cnb-1 5'-upstream regulatory sequences. Both cna-1 and cnb-1 reporter transgenes expressed in diverse tissues. Expression was detected at all stages of development starting from early comma stage embryos to adult stages. Calcineurin is expressed in vulval muscle, body-wall muscle, spermatheca (Figure , A --C), and in a majority of neuronal cell bodies in the head and tail similar to previously obtained results Figure 3 | Expression and localization of calcineurin. Expression and localization of calcineurin. Images show only representative expression. CNA-1::GFP expressed in (A) ventral nerve cord (arrow) and vulva muscles (arrowhead) and (B) body-wall muscles of the midbody region. CNB-1::GFP expresses in spermatheca (C; arrowhead) and intestine (C; arrow). Nerve cord expression can also be observed in this image. (D) Immunostaining of wild-type worms with anti --CNB-1 antibodies detected CNB-1 in hypodermal seam cells (arrow). Bar, 20 mum. (E --G) Immunogold electron micrograph (EM) showing the subcellular localization of CNB-1. (E) Transverse section through the midbody of a hermaphrodite showing CNB-1 localization as indicated by the signals of gold particles in the seam cells of the lateral hypodermis (LH) closely apposed to the cuticle (arrowhead). (F) Another region of seam cell showing signals. Lateral hypodermis (LH), cuticle (arrowhead), and excretory canal (arrow) are labeled. (G) Localization of CNB-1 in a wild-type male gonad. Distinct and scattered signals of CNB-1 are observed mainly in the cytoplasmic regions of the cellularized spermatids. Bar, 1 mum. Polyclonal antibodies against CNA-1 and CNB-1 were generated, and western blots with total protein extracts from wild-type worms detected a single band around 60 kDa for CNA-1 and 16 kDa for CNB-1, respectively (Figure D). Immunostaining performed with these antibodies for both subunits of calcineurin showed similar localization to GFP expression patterns and additionally showed localization in hypodermal seam cells (Figure D). We localized CNB-1 at the subcellular level in specific tissues of wild-type animals by immunogold electron microscopy (EM). Signals of CNB-1 were observed in the seam cells of the lateral hypodermis of wild-type hermaphrodites consistent with our immunostaining data (Figure , D --F). Additionally, the male gonad also expressed CNB-1. This was evident from the scattered and distinct cytoplasmic signals of CNB-1 surrounding the cellularized spermatids (Figure G). Based on these observations we further examined wild-type male sperm and immunostained with anti-CNA-1 or anti --CNB-1 antibodies. As expected, we observed robust staining in the wild-type sperm for both proteins and the staining was distinctly cytoplasmic (Figure A). We also confirmed CNA-1 and CNB-1 localization in the spermatheca by immunostaining isolated gonads (Figure B). Hence, electron microscopy and immunostaining data reveal that calcineurin is expressed in the C. elegans male germline, and therefore may have possible roles in germline development. Figure 4 | Calcineurin in sperm and the spermatheca. Calcineurin in sperm and the spermatheca. (A) Immunostaining of wild-type sperm with anti-CNA-1 (top left panel) and anti --CNB-1 (bottom left panel) antibodies and the respective nuclei are shown by DAPI staining (right panels). Bar, 2 mum. (B) Immunostaining of wild-type and cnb-1 mutant spermatheca with calcineurin antibodies. Anti-CNA-1 (top left panel) and anti --CNB-1 (top middle panel) antibodies stain wild-type spermatheca. Spermatheca of cnb-1 mutant worms do not show any anti --CNB-1 staining (top right panel). DAPI staining of spermatheca is shown in the bottom panels. Bar, 20 mum. (C) Sperm and spermatheca defects. Compared with wild-type (left panel) cnb-1 mutant worms display defects in sperm and the spermatheca. cnb-1 sperm (middle and right panel). Arrows point to cnb-1 sperm, which are fewer in number and show a smoother appearance compared with N2 sperm (left panel). The surrounding cnb-1 spermatheca appears to contain debris. (D) Mating of N2 and cnb-1 hermaphrodites with N2 males. Three N2 males were mated with one N2 or one cnb-1 hermaphrodite for 2 days. Total progeny and outcrossed progeny are indicated by white and black bars, respectively. cnb-1 Is Involved in Normal Cuticle Formation, Sperm Morphology, and Brood Size | We have isolated cnb-1(jh103) deletion mutants by PCR-based TMP/UV mutagenesis method. The cnb-1(jh103) null mutants are viable but exhibit multiple phenotypes. A second deletion mutant strain, cnb-1(ok276), was also isolated (kindly donated by the C. elegans Knockout Consortium, Oklahoma) and exhibited identical phenotypes to cnb-1(jh103). We further checked the protein profile of the cnb-1(jh103) deletion mutants and found no protein band on western blots confirming that the deletion mutants are functionally null (Figure D, lane 2). The cnb-1(jh103) mutants were examined for phenotypic defects compared with the wild-type worms. At the surface of the worm, the cuticle of cnb-1 mutants appears to thin out, resulting in the animals having a more transparent appearance. This phenotype seems consistent with a loss of calcineurin in hypodermal tissue in cnb-1 mutant worms. cnb-1 mutants also have a significantly decreased brood size compared with wild-type . This observation taken together with the immunostaining data led us to believe that there may be defects in cnb-1 sperm. Indeed, under close inspection, cnb-1 sperm morphology was different from wild-type sperm. The defective sperm were smaller and smoother than wild-type sperm (Figure C) and displayed smaller pseudopods. Moreover, sperm, which normally gather in high numbers in the spermatheca, were scarcely found in mutant worms. In addition, these cnb-1 --deficient spermatheca were filled with oocyte debris indicative of a possible endomitotic oocyte (emo)-like defect caused by spermatheca defects. To distinguish whether the small brood size phenotype observed in the mutant was a direct cause of defective sperm or spermatheca defects, we mated N2 male worms with cnb-1 hermaphrodites. As Figure D shows, wild-type sperm only partially rescues the low brood size, and the percentage of outcrossed progeny of cnb-1 hermaphrodites is almost the same as outcrossed of N2 hermaphrodites. This suggests that altered sperm alone cannot account for the decreased fertility in cnb-1 mutants, and that defective spermatheca and/or oocytes may also play a role in this phenotype. Table 1 | Brood size, body size, and egg-laying phenotypes Defects in cnb-1 Mutants Resemble Defects Observed in unc-43 Mutants | Along with its small brood size, cnb-1 shows several other characteristic phenotypes. Firstly, mutant worms have small and slender bodies compared with wild-type animals. cnb-1 worms also retained embryos in the uterus varying from early to very late in development, such as threefold stage embryos, which was not seen in the wild-type worms. cnb-1 mutations also cause the worms to be uncoordinated (unc), or lethargic in their movement. Generally wild-type worms move rapidly in a sinusoidal pattern as evidenced from the tracks made by the worms (Figure A), but cnb-1 mutants moved slowly and with decreased amplitude of tracks. (Figure B). Each of these phenotypes was not constant over the lifetime of the worm, but rather all became more severe as the worm grew older. Brood size, body size, and embryo retention phenotypes are quantified and summarized in Table . We then sought to determine whether cnb-1 genomic DNA could rescue the cnb-1(jh103) phenotypes by microinjection. Movement of these transgenic animals appeared to be fully rescued (Figure C) and there was a significant recovery in brood size (146 +- 27 progeny; Table ). The small body size and retention of late-stage embryos in the uteri were almost completely restored to normal state indicating that these pleiotropic phenotypes observed in the mutant were specifically caused by the loss of cnb-1 gene. Figure 5 | Locomotion defects in mutants of cnb-1 and unc-43. Locomotion defects in mutants of cnb-1 and unc-43. Five animals were placed in the center of a bacterial lawn and photographed 5 min later. (A) Tracks of movement by wild-type animals; (B) cnb-1 mutants; (C) rescued cnb-1 mutants; (D) unc-43(lf) mutants; (E) unc-43(gf) mutants; (F) cnb-1; unc-43(gf) double mutants. Bar, 1 mm. Although the phenotypes of cnb-1 mutants appeared to be pleiotropic, it is interesting to note that similar phenotypes were observed in other particular mutants. Mutants of the unc-43 gene, which encodes the Ca2+/calmodulin-dependent protein kinase CaMKII also show defects in brood size and progressive defects in body size, egg laying, and movement . We compared phenotypes of gain-of-function and loss-of-function mutants of unc-43 with those of the cnb-1 null mutant and observed some interesting relationships. cnb-1 mutants and unc-43(n498)[gain-of-function (gf)] mutants were quite similar in phenotype, with both mutants showing much lower brood sizes, delayed egg laying, and smaller body size . In addition, both worms displayed severe uncoordinated movement (Unc) phenotypes (Figure , B and E). Conversely, unc-43(n1186) [loss-of-function (lf)] mutants showed phenotypes mostly opposite to that of unc-43(gf) and cnb-1 mutants, displaying hyperactive movement (Figure D), a significantly higher brood size, and earlier egg-laying compared with that of cnb-1. On the other hand, sma-6 mutants, which also have a small body size phenotype, do not show other defects present in cnb-1 mutants suggesting that small body size alone does not directly affect these other phenotypes and that sma-6 and cnb-1 likely function in different pathways. Phenotypes Related to G-protein --coupled Phosphorylation Signaling Pathways in C. elegans | It was shown that UNC-43/CaMKII regulates a G-protein pathway involving the Go-protein alpha-subunit, goa-1, in locomotory and egg-laying behavior . A transgenic gain-of-function mutant of goa-1, syIs9[goa-1(gf)], displayed lethargic movement and egg retention behavior similar to phenotypes seen in unc-43(gf) mutants . Conversely, loss-of-function mutants of goa-1 showed hyperactive movement and premature egg laying . Thus, mutations in goa-1 result in similar phenotypes to those of unc-43 mutants, and, consequently, opposite phenotypes to that of cnb-1 mutants. We decided to further test whether cnb-1 could be operating in similar G-protein pathways. Along with the phenotypes tested in Table , goa-1 also had defects in serotonin-regulated egg laying . Thus, we tested whether cnb-1 and unc-43 mutants had defects in this egg-laying behavior as well. Exogenous serotonin and imipramine were shown to stimulate egg laying in wild-type worms . However, both cnb-1 and unc-43(gf) mutants failed to respond to exogenous serotonin, whereas unc-43(lf) mutants were hypersensitive to serotonin treatment by laying even more eggs than wild-type . cnb-1 and unc-43(gf) mutants also show a decreased sensitivity to imipramine, an agent that induces endogenous release of serotonin from stores in the presynaptic HSN neuron, compared with wild-type animals. When compared with the egg-laying defects observed in serotonin-treated syIs9[goa-1(gf)] mutants reported elsewhere cnb-1 and unc-43(gf) mutants show slightly more severe phenotypes in response to the treatment. Table 2 | Characterization of serotonin-mediated egg-laying phenotypes We next tested whether the serotonin-mediated egg-laying phenotypes observed in cnb-1 and unc-43 mutants were concentration-dependent. Unlike wild-type animals, which show increased egg laying in response to elevated concentrations of serotonin (Figure A), all three mutants displayed resistance to serotonin even at high concentrations (Figure , B and C). We also tested the egg-laying response to another exogenous agent, levamisole. Levamisole is an agonist of the UNC-29 nicotinic acetylcholine receptor localized in postsynaptic muscle, and levamisole treatment to wild-type C. elegans results in muscle hyper-contraction and subsequent egg laying (Figure A). In contrast, cnb-1 mutants and both unc-43(gf) and unc-43(lf) mutants are resistant to levamisole at all concentrations (Figure , B, C, and D). Finally, goa-1(n1134) loss-of-function mutants also show concentration-dependent resistance to levamisole (Figure E). This further suggests that cnb-1, like unc-43 and goa-1, are involved in similar aspects in the regulation of egg laying. Figure 6 | Dose-dependent serotonin-mediated egg laying. Dose-dependent serotonin-mediated egg laying. Color-coded area in bars represent percentage of worms from the total number tested. Respective number of eggs were laid after serotonin treatment at the given concentrations. In each case N = 24. (A) N2 worms show serotonin-mediated egg laying in a dose-dependent manner, whereas (B) cnb-1 and (C) unc-43(gf) mutants and (D) unc-43(gf);cnb-1 double mutants were mostly resistant even at high concentration of serotonin. Figure 7 | Dose-dependent levamisole-mediated egg laying. Dose-dependent levamisole-mediated egg laying. Color-coded area in bars represent percentage of worms from the total number tested. Respective number of eggs were laid after levamisole treatment at the given concentrations. In each case N = 24. (A) N2 worms show levamisole-mediated egg laying in a dose-dependent manner, whereas (B) cnb-1, (C) unc-43(gf), (D) unc-43(lf), (E) goa-1(n1134) mutants, and (F) unc-43; cnb-1 double mutants were mostly resistant even at high concentrations of levamisole. To further validate the possibility that calcineurin, UNC-43, and GOA-1 may be functioning together to regulate these various functions, we attempted to generate double mutants between cnb-1 and unc-43 and goa-1 mutants and assess possible genetic relationships. Although we generated both unc-43(lf);cnb-1 and goa-1(n1134);cnb-1 double mutants, the resulting animals were arrested in early larval stages with severe morphological defects. Thus, epistatic phenotypes could not be assessed in these worms. On the other hand, unc-43(gf);cnb-1 double mutants were viable and developed to adulthood. Since unc-43(gf) encodes a protein kinase, mutants should have hyperactive phosphorylation activity in pathways where unc-43 may be involved in. Likewise, cnb-1, which encodes a protein phosphatase that was shown to have enzyme activity, should produce null mutants that have a loss of dephosphorylation activity in pathways which calcineurin may function in. If these two proteins function within a G-protein-coupled phosphorylation pathway in an opposing manner, then a double mutant of cnb-1 and unc-43(gf) would have phosphorylation pathways that are hyperpolarized without any opposing dephosphorylation activity, and, thus, should display more severe phenotypes than animals with single mutations alone. As we had predicted, this double mutant had even more severe defects than the individual mutants by themselves. The worms displayed extremely lethargic movement (Figure F), an average body size of less than half that of normal worms, synergistically small brood sizes with most worms being sterile, and severe egg retention leading to internally hatched young and eventual death of the parent worm . unc-43(gf);cnb-1 double mutants also result in increased resistance to serotonin, imipramine, and levamisole treatment compared with the single mutants alone (Table ; Figures D and F). In comparison to single mutants, synergistic effects in the double mutants can be clearly seen in fertility and movement phenotypes, and possibly egg-laying phenotypes as well. Therefore, our results suggest that calcineurin can regulate multiple functions in C. elegans that both unc-43 and goa-1 are known to be involved in, and a role for calcineurin phosphatase activity in a CaMKII-dependent G-protein coupled phosphorylation signaling pathway in C. elegans is a distinct possibility. DISCUSSION : Calcineurin, a serine/threonine protein phosphatase, plays important roles in the transduction of Ca2+ signals to regulate various cellular processes . In this study, we identified and characterized calcineurin from C. elegans, both at its molecular and cellular levels. In contrast to higher animals where calcineurin subunits are encoded by more than one gene, C. elegans calcineurin subunits are each encoded by a single gene. Results reported here also show that the homologues of the mammalian and Drosophila calcineurin subunits exist in C. elegans, and thus, they represent a conserved branch of the PP2B family of protein phosphatases having important roles in normal physiology . The subunits of C. elegans calcineurin were shown to interact with one another, and this interaction conferred strong phosphatase activity to the heterodimer, which is consistent with studies on other conserved forms of calcineurin . In addition, Ca2+-dependent phosphatase activity of CNA-1 was potently inhibited by the immunosuppressant CsA (Figure C) in agreement with previous reports . Sequence homology and Ca2+ overlay experiments verified that CNB-1 has strong calcium binding affinity. Thus, we believe that the biochemical function of calcineurin as a calcium-binding heterodimeric protein phosphatase is conserved in C. elegans. A deletion null mutant of cnb-1 was isolated by target-selected mutagenesis. cnb-1(jh103) mutants are viable although loss of calcineurin activity leads to pleiotropic defects for the worms. It was shown that transgenic mice expressing a mutated form of CnB that cannot bind the CnA subunit will die early in embryo development . These mice exhibit vascular developmental abnormalities that stem from its inability to dephosphorylate and subsequently translocate the transcription factor NFAT from the cytoplasm into the nucleus. Thus, the dephosphorylation activity of calcineurin is dependent on both CnA and CnB subunits in vivo. This is also observed in S. cerevisiae as mutations in either CnA or CnB lead to a complete loss of calcineurin phosphatase activity and function in yeast . Finally, a recent study showed that a chemotaxis-defective mutant tax-6(p675), which displays thermotaxis, body size, and growth defects as well, carries a mutation in the cna-1 locus . Thus, a viable mutant of cnb-1 in a multicellular organism like C. elegans, although quite surprising, is important for further studies in the delineation of calcineurin's diverse functions in vivo. Mutants of cnb-1 showed pleiotropic phenotypes including cuticle defects, small body size, decreased brood size, and locomotory and egg-laying defects. These defects were consistent with the loss of calcineurin function in tissues that normally express the protein phosphatase; these include hypodermal tissue, spermatheca, sperm, body-wall muscle, and vulva muscle. Most of the phenotypes that we observed in cnb-1 mutants appeared to be very similar to and characteristic of those observed in unc-43(gf) yet opposite to those in unc-43(lf) mutants. A double mutant between this cnb-1 null mutant and an unc-43(gf) mutant causes a synergistic effect of movement, fertility, and possibly egg-laying phenotypes as well, which may suggest a complementary relationship between the Ca2+-calmodulin --dependent protein kinase CaMKII and the Ca2+-calmodulin --dependent protein phosphatase calcineurin. Opposing functions of calcineurin and CaMKII have also been observed biochemically in integrin signaling, in the receptor associated protein RAP, in skeletal muscle dystrophin, and in T-cell signaling . Thus, this kind of biochemical role for calcineurin in C. elegans is a reasonable and plausible function. It was previously shown that mutations in goa-1, the homologue of the Goalpha-subunit of the Go protein, can suppress locomotory defects associated with unc-43(n498) gain-of-function mutants . Besides an uncoordinated movement phenotype, loss-of-function mutants of goa-1 also exhibited defects in egg laying and reduced brood size similar to those seen in unc-43(lf) mutants and opposite to those observed in our cnb-1 mutant. We tested to see if cnb-1 was also defective in another goa-1 related phenotype, serotonin-mediated egg laying . cnb-1 mutants exhibited defects in serotonin-induced egg-laying similar to those seen in unc-43(gf) mutant and a transgenic goa-1 gain-of-function mutant syIs9[goa-1(gf)]. In addition, the dose-dependent serotonin-mediated and levamisole-mediated egg-laying curves confirmed that calcineurin is involved in similar aspects of C. elegans egg laying as unc-43 and goa-1 are involved in. From these data we suggest that calcineurin may be involved in G-protein --coupled phosphorylation pathways in locomotion, egg laying, and brood size in C. elegans. The involvement of calcineurin in G-protein-mediated signaling has been observed in many different pathways including regulation of NFAT in cardiac myocytes, T helper cell immunity, and cardiac hypertrophy . In addition, phosphatase activity as an inhibitor of G-protein phosphorylation signaling is a common method of pathway regulation so it is likely that calcineurin may be fulfilling this role in C. elegans. Finally, calcium oscillations evoked by G-protein coupled receptors and stimulated by regulators of G-protein signaling (RGS) proteins could initiate Ca2+/calmodulin-dependent calcineurin activity in C. elegans which has two homologous RGS proteins, EGL-10 and EAT-16, directly involved in G-protein pathways . Although movement and egg-laying phenotypes are known to be involved in a G-protein-mediated pathway, other phenotypes observed in cnb-1 mutants such as body size and fertility were not yet determined to be involved in this specific pathway. Body-size phenotypes are not likely regulated by this specific pathway since small body size in unc-43 mutants is caused by tonically contracted body-wall muscle , which we did not observe in cnb-1 mutants. On the other hand, the relationship of the brood size phenotype to the unc-43-regulated G-protein signaling pathway may be more difficult to determine. We showed that calcineurin expresses in the spermatheca and sperm, and loss of calcineurin function results in burst oocytes derived from defects in the spermatheca and sperm morphology defects. Attempted rescue of this phenotype by mating cnb-1 hermaphrodites with wild-type males showed that the cause of this defect may be from multiple factors. To assess whether this phenotype is involved in this same pathway similar experiments need to be performed on mutant goa-1 and unc-43 worms. Just as calcineurin is involved in many types of signaling, CaMKII and Goalpha are also general signaling molecules and have diverse functions. unc-43 mutants are not only involved in defective locomotion, egg laying, and brood size, but also show abnormal defecation behavior . We also assessed the defecation cycle in cnb-1 mutants but worms showed normal defecation behavior. In the same way, goa-1 mutants also show other defective behaviors. Among these behaviors, goa-1 males have problems executing the "turning" behavior of male mating, which may be a result of defective diagonal muscles . We also tested whether calcineurin RNAi-affected male worms show defects in turning behavior, but no significant defects nor any expression of calcineurin in diagonal muscle tissue could be detected. The common defects observed in all cnb-1, unc-43, and goa-1 mutants suggest that locomotion and egg-laying defects may be specific to a CNB-1/UNC-43 G-protein --coupled phosphorylation pathway in C. elegans. Calcineurin may function upstream or downstream of unc-43, which is known specifically to regulate locomotion via the Go/Gq signaling network . If a double mutant of cnb-1 and either unc-43(lf) or goa-1(lf) was generated, epistatic phenotypes between these functionally antagonistic proteins would distinguish where calcineurin may function in relation to UNC-43 or GOA-1. Unfortunately, both unc-43(lf);cnb-1 and goa-1(n1134);cnb-1 resulted in worms that were arrested in the L1 larval stage with severe morphology defects, indicating that both UNC-43 and GOA-1 in association with calcineurin are essential for developmental signaling pathways distinct from the G-protein pathway described above. Although egg-laying phenotypes could not be assessed in these double mutants, we attempted to observe locomotory behavior. Although goa-1(n1134);cnb-1 movement phenotypes were not assessed, cnb-1;unc-43(lf) double mutants exhibited hyperactive movement similar to unc-43(lf) mutants, although it was difficult to quantify this behavior due to the small size of the worm and its severe defects. Nevertheless, our observations indicate that unc-43 might be epistatic to cnb-1, which could further verify the role of calcineurin in a G-protein-mediated signaling pathway. Backmatter: PMID- 12221133 TI - Membrane Trafficking of Heterotrimeric G Proteins via the Endoplasmic Reticulum and Golgi AB - Membrane targeting of G-protein alphabetagamma heterotrimers was investigated in live cells by use of Galpha and Ggamma subunits tagged with spectral mutants of green fluorescent protein. Unlike Ras proteins, Gbetagamma contains a single targeting signal, the CAAX motif, which directed the dimer to the endoplasmic reticulum. Endomembrane localization of farnesylated Ggamma1, but not geranylgeranylated Ggamma2, required carboxyl methylation. Targeting of the heterotrimer to the plasma membrane (PM) required coexpression of all three subunits, combining the CAAX motif of Ggamma with the fatty acyl modifications of Galpha. Galpha associated with Gbetagamma on the Golgi and palmitoylation of Galpha was required for translocation of the heterotrimer to the PM. Thus, two separate signals, analogous to the dual-signal targeting mechanism of Ras proteins, cooperate to target heterotrimeric G proteins to the PM via the endomembrane. Keywords: INTRODUCTION : Heterotrimeric G proteins transduce signals from cell-surface receptors to intracellular effectors. To do this, G proteins must associate with the cytoplasmic face of the plasma membrane (PM). The Galpha, Gbeta, and Ggamma subunits of G proteins are synthesized in the cytosol on free polysomes and must be posttranslationally modified to traffic to the PM. Three types of posttranslational modifications are known to occur on subunits of G proteins (for review, see ). alpha-Subunits can be myristoylated and/or palmitoylated, whereas the Ggamma subunits contain a CAAX motif similar to those of the Ras and Rho families of monomeric GTPases. The CAAX motif is modified by a well-characterized, three-step process yielding a prenylated and carboxyl methylated C-terminus . The Gbeta subunit is unmodified but remains tightly associated with a Ggamma subunit. The various contributions of myristoylation, palmitoylation, and CAAX-box processing of individual G-protein subunits to PM association of the trimer have not been thoroughly investigated. For Ras and Rho family small GTPases, two signals cooperate to target the monomeric GTPase to the PM . Processing of the CAAX box is necessary and sufficient to target newly synthesized small GTPases to the cytoplasmic leaflet of endoplasmic reticulum (ER), where AAX proteolysis and carboxyl methylation take place. Final PM targeting, however, requires a second signal within the same polypeptide. This second signal consists of either a cluster of basic residues (polybasic region) or one or more palmitoylated cysteine residues immediately upstream of the CAAX box. Mutation of the second signal results in retention of the GTPase on endomembrane. The targeting of mammalian G-protein alpha subunits has been extensively studied. Binding of the Gbetagamma dimer promotes stable membrane association of Galphas and Galphaq subunits . Mutations that disrupt the binding of Galpha to Gbetagamma also disrupt membrane association of these Galpha subunits, suggesting that the palmitoylation of Galpha alone is insufficient for stable membrane association. Palmitoylation of Galphas and Galphaz and their association with Gbetagamma act cooperatively . This suggests a model for Galpha localization that involves a dual signal, analogous to that defined for Ras. Previous analyses of Galpha targeting leave unanswered the question of the targeting of Gbetagamma to the PM. The observation that heterotrimeric G proteins can be mislocalized by ectopic targeting of Gbetagamma suggests that final localization is dictated by Gbetagamma. If Galpha follows Gbetagamma, then understanding the intrinsic targeting of the latter is critical. The CAAX motifs of Ggamma can be either farnesylated (Ggamma1) or geranylgeranylated (most other Ggamma subunits) . Analysis of the sequences of the mammalian gamma subunits reveals no obvious PM-targeting second signal similar to those found in Ras proteins. If the model established for Ras and Rho proteins also applies to the Ggamma subunits, then it would be expected that the Ggamma subunit would localize on endomembrane. Among the possible explanations for PM localization of Gbetagamma are the existence of a previously uncharacterized second signal in the Ggamma polypeptide and the contribution of such a signal by Galpha subunits after assembly of the trimeric complex on endomembrane. To distinguish between these possibilities, we expressed Ggamma and/or Galpha subunits tagged with green fluorescent protein (GFP) or spectral mutants of GFP with and without coexpression of Gbeta and analyzed the localization of the fusion proteins in living cells. Our results demonstrate that the CAAX processing of Ggamma targets this subunit, in complex with Gbeta, to the ER and that translocation from endomembrane to the PM requires both expression and acylation of Galpha. Association of Galpha with Gbetagamma does not serve merely to stabilize PM association but rather occurs initially on Golgi and is critical for translocation to PM. Thus, PM targeting of G proteins, like Ras proteins, requires two signals, but whereas the Ras signals are on the same polypeptide, they are on different subunits for heterotrimeric G proteins. In addition, we found that carboxyl methylation was necessary for stable membrane association of farnesylated Ggamma1 but not geranylgeranylated Ggamma2. MATERIALS AND METHODS : Cell Culture and Transfection | COS-1 and MDCK cells were obtained from American Type Tissue Collection, Manassas, VA. These cells were grown in DMEM containing 10% fetal bovine serum (Cellgro, Herndon, VA) at 5% CO2 and 37C. Spontaneously immortalized murine embryonic fibroblasts (MEFs) from both prenylcysteine carboxyl methyltransferase (pcCMT) --null (CMT-/-) mice and CMT+/+ littermates were generated as we have described and cultured in DMEM with 15% fetal bovine serum (Colorado Serum Company, Denver, CO), nonessential amino acids, beta-mercaptoethanol, and l-glutamine at 5% CO2 and 37C. For all microscopy, cells were plated, transfected, and imaged in the same 35-mm culture dish that incorporated a No. 1.5 glass coverslip --sealed 15-mm cutout on the bottom (MatTek, Ashland, MA). Transfections of COS-1 and MDCK cells were performed 1 day after plating at 50% confluence using SuperFect according to the manufacturer's instructions (Qiagen, Hilden, Germany). MEFs were transfected using Lipofectamine Plus according to the manufacturer's instructions (Invitrogen, San Diego, CA). In some experiments, 50 muM 2-bromopalmitate (2-BP) (Sigma-Aldrich, St. Louis, MO) was added at the time of transfection. Unless otherwise noted, for coexpression, a 1:2:2-mug plasmid DNA ratio of gamma:beta:alpha was used. Control transfections omitting beta and gamma contained an equivalent amount of vector DNA. Transiently transfected cells were analyzed 1 day after transfection. Plasmids | The plasmids pCMV-Galphai, pCMV-Galphai1Q204L, pCMV-Galphai2, pCMV-Galphai2Q205L, pCMV-Galphaq, pCMV-Galphasshort (pCMV-Galphass), pCMV-GalphassQ213L, and pCMV-Galphaslong (pCMV-Galphasl) were generous gifts of Dr. Susanne Mumby, University of Texas (Dallas, TX). Plasmid pcDNA-Galphai2 was obtained from the Guthrie Institute (Sayre, PA). Galphai2 was subcloned into pCFP-N1 (Clonetech, Cambridge, UK) for production of Galphai2-cyan fluorescent protein (CFP). The plasmids pEV-Ggamma1, pEV-Ggamma2, and pEV-Gbeta1 were gifts of Dr. N. Gautam, Washington University, St. Louis, MO. Ggamma subunits were subcloned into pEGFP-C3 (Clonetech) for production of GFP-Ggamma fusion proteins and into pYFP-C1 (Clonetech) for production of yellow fluorescent protein (YFP)-Ggamma. The beta-subunit was subcloned into pcDNA3.1+. The 11-amino-acid tails of the Ggamma subunits were produced by PCR amplification using primers bracketing the C-terminal 11 amino acids of each subunit and were then cloned into pEGFP-C3. The C3S mutation of pcDNA-Galphai2 was produced by PCR amplification using primers that included the appropriate Cys-to-Ser mutation at position 3 (counting from the initial methionine that is cleaved off in myristoylated proteins), followed by cloning into pcDNA3.1+. This mutant was also subcloned into pCFP-N1 (Clonetech) for production of Galphai2C3S-CFP. Galpha expression levels were similar for both pCMV and pcDNA vectors, as was the effect on GFP-Ggamma localization. Fluorescence Microscopy | Live cells were examined 24 h after transfection with a Zeiss Axioscope epifluorescence microscope (63x PlanApo 1.4 NA objective) (Zeiss, Oberkochen, Germany) equipped with a Princeton Instruments cooled CCD camera and MetaMorph digital imaging software (Universal Imaging, West Chester, PA) or a Zeiss 510 laser scanning confocal microscope (100x PlanApo 1.4 NA objective). Digital images were processed with Adobe Photoshop 6.0 (Adobe, San Jose, CA). RESULTS : Ggamma Subunits Are Targeted to the ER | GFP is a hydrophilic protein that localizes homogeneously throughout the cytosol and nucleoplasm (Figure a). Addition of a four-amino-acid CAAX motif, such as the CAIL sequence of Ggamma2, to the C terminus of GFP (GFP-CAAX) changed its localization dramatically to an endomembrane pattern that included ER, nuclear envelope, and Golgi but excluded PM (Figure b). Thus, as we have previously demonstrated, the CAAX motif alone is an efficient endomembrane targeting signal . For the Ras and Rho families of GTPases, sequences within the C-terminal 10 --20 amino acids (the hypervariable region) are sufficient to give GFP a pattern of membrane expression identical to that of the full-length GFP-tagged protein . GFP-tagged H-Ras (GFP-H-Ras) and GFP extended with the last 10 amino acids of H-Ras (GFP-H-Ras tail) both localize to PM and Golgi (Figure , c and d). Similarly, GFP extended with the last 11 amino acids of Rac1 (GFP-Rac tail) gave a pattern of membrane localization indistinguishable from that of GFP-tagged full-length Rac1 (Figure , e and f). Thus, for Ras and Rho family proteins, the final 10 --20 amino acids contain all the necessary information for membrane localization. Figure 1 | Ggamma subunits are targeted by their CAAX sequence to endomembrane but lack intrinsic secondary PM targeting signals. Ggamma subunits are targeted by their CAAX sequence to endomembrane but lack intrinsic secondary PM targeting signals. COS-1 cells were transfected with plasmids directing expression of GFP alone (a), GFP fused to the Ggamma2 CAAX sequence, CAIL (b), GFP-H-Ras (c), GFP-H-Ras tail (d), GFP-Rac1 (e), GFP-Rac1 tail (f), GFP-Ggamma1 (g), GFP-Ggamma1 tail (h), GFP-Ggamma2 (i), or GFP-Ggamma2 tail (j) and imaged 24 h later alive by digital epifluorescence microscopy using a high-resolution cooled CCD camera. Bars, 10 mum. Amino acid sequence comparison of the C-terminal hypervariable regions of H-Ras, Rac1, Ggamma1, and Ggamma2 (k). The CAAX motif is underlined, the palmitoylation sites of H-Ras are shown in outline font, and the polybasic region of Rac1 is shaded. In contrast, GFP fused to full-length Ggamma1 or to the C-terminal 11 amino acids of Ggamma1 (GFP-Ggamma1, GFP-Ggamma1-tail) and expressed in COS-1 cells (Figure , g and h) or MDCK cells (not shown) localized to the ER and Golgi in a pattern identical to that observed for GFP-CAAX (Figure a). Similar results were obtained (Figure , i and j) using GFP fused to full-length Ggamma2 or to the C-terminal 11 amino acids of Ggamma2 (GFP-Ggamma2, GFP-Ggamma2-tail). Thus, unlike the Ras and Rho family proteins, Ggamma polypeptides lack a second signal for PM targeting. Analysis of the amino acid sequences of the C-termini of these molecules (Figure k) revealed that whereas H-Ras has sites for palmitoylation near the CAAX motif and Rac1 has a polybasic region adjacent to the CAAX motif, neither Ggamma1 nor Ggamma2 has analogous sequences. As with the Ras and Rho family proteins , neither farnesylation alone (Ggamma1) nor geranylgeranylation alone (Ggamma2) is sufficient to target GFP to the PM. We conclude that the intrinsic membrane targeting of G-protein gamma subunits is for ER and Golgi and that an extrinsic factor(s) must therefore be required for translocation of these proteins from endomembrane to PM. Coexpression of Gbeta and Galpha Targets GFP-Ggamma to the PM | We next tested the effect of coexpression of Gbeta and Galpha subunits on GFP-Ggamma localization. GFP-Ggamma1 was coexpressed in COS-1 cells with either Gbeta1 alone or with Gbeta1 and a variety of Galpha subunits . Gbeta subunits form tight complexes with Ggamma subunits. GFP-Ggamma1 co-overexpressed with Gbeta1 (Figure a) showed the same ER pattern as seen with GFP-Ggamma1 expressed alone . Thus, Gbeta subunits do not alter the intrinsic targeting of Ggamma1 to the endomembrane. In contrast, when GFP-Ggamma1 was co-overexpressed with Gbeta1 and with the Galphai2 (Figure b), Galphaq (Figure c), or Galphas (Figure d), a predominantly PM and Golgi pattern was observed that was identical to that observed for GFP-H-Ras at steady state (Figure c). The same results were obtained with MDCK cells (not shown). As with farnesylated GFP-Ggamma1, geranylgeranylated GFP-Ggamma2 coexpressed with Gbeta1 alone (Figure e) showed the same ER/Golgi pattern as seen with the Ggamma2 subunit expressed alone . Coexpression of Gbeta and each Galpha subunit with GFP-Ggamma2 (Figure , f --h) resulted in PM and Golgi localization. Thus, neither the targeting of GFP-Ggamma alone to the ER nor the heterodimer to the Golgi and PM was affected by the length of the polyisoprene lipid that modified Ggamma. Coexpression of GFP-Ggamma2 with Gbeta1 and with constitutively active mutants of Galpha, which are unable to bind to Gbetagamma, did not promote PM localization of GFP-Ggamma (not shown). We conclude that heterotrimer formation is required for PM targeting and that sequences within the Galpha subunit act in trans with the Ggamma CAAX motif to deliver the trimer as a complex to the PM. Moreover, the appearance at steady state of GFP-Ggamma on the Golgi as well as the PM, a pattern identical to that of GFP-H-Ras that transits the Golgi en route to the PM , suggests that heterotrimer formation occurs on the Golgi. These results are in agreement with a previous study that showed that PM localization of Ggamma3 was facilitated by interaction with the Galpha and Gbeta subunits. However, our results suggest that the role of heterotrimer formation is not simply to add affinity for the PM but rather to permit protein trafficking from endomembrane to PM. Figure 2 | Galpha subunits provide a PM-targeting second signal for Gbetagamma. Galpha subunits provide a PM-targeting second signal for Gbetagamma. COS-1 cells were transfected with GFP-Ggamma1 (a --d) or GFP-Ggamma2 (e --h) and cotransfected with Gbeta1 alone (a, e), Gbeta1 and Galphai2 (b, f), Gbeta1 and Galphaq (c, g), or Gbeta1 and Galphas (d, h) and imaged as in Figure . Arrows indicate PM, arrowheads indicate Golgi, and the positions of nuclei are marked (N). a and e, The nuclear envelope and Golgi are purposely overexposed to reveal the peripheral reticulum of the ER. Bars, 10 mum. Palmitoylation of Galpha Is Necessary for PM Localization of the Trimer | The PM and Golgi localization of GFP-Ggamma coexpressed with Gbeta1 and Galpha is very similar to the localization pattern seen with GFP-H-Ras (Figure c). H-Ras is palmitoylated on Golgi membranes , and palmitoylation is required for trafficking of H-Ras from the endomembrane to the PM, as demonstrated by inhibition of palmitoylation with 2-BP or expression of GFP-H-RasC181,184S, which lacks palmitoylation sites (Figure , a --c). All three of the Galpha subunits tested, Galphas, Galphaq, and Galphai2, are palmitoylated: Galphas is singly palmitoylated, Galphaq is doubly palmitoylated, and Galphai2 is myristoylated and palmitoylated . To test whether the palmitate modification of the Galpha subunit functions like that of H-Ras in providing the second signal required for PM targeting, we tested the ability of unpalmitoylated Galpha subunits to promote PM trafficking of Gbetagamma. GFP-Ggamma2 was coexpressed with Gbeta1 and Galphai2 in the presence or absence of 2-BP. Whereas coexpression of GFP-Ggamma2 with Gbeta1 and Galphai2 resulted in PM localization (Figure d), GFP-Ggamma2 remained endomembrane-associated in the presence of 2-BP (Figure e). Similar results were obtained using Galphas and Galphaq in the presence and absence of 2-BP (not shown). To distinguish an effect on Galpha binding of Gbetagamma from an effect on heterotrimer trafficking, we determined whether unpalmitolyated Galpha could bind Gbetagamma. A palmitoylation-deficient mutant of the Galphai1 subunit has previously been shown to interact normally with Gbetagamma subunits . We confirmed that palmitoylation-deficient Galphai2C3S can interact with Gbetagamma by demonstrating that this Galpha, when coexpressed with GFP-Ggamma2 and Gbeta1, was efficiently ADP-ribosylated by pertussis toxin (not shown), a modification that requires heterotrimer formation. Coexpression of GFP-Ggamma2 and Gbeta1 with a palmitoylation-deficient Galphai2C3S resulted in retention of GFP-Ggamma2 on the endomembrane (Figure f). Similar results were obtained with GFP-Ggamma1 (Figure , g --i). Thus, palmitoylation of the Galpha subunit functions like acylation of H-Ras to provide a second signal for engagement of a transport pathway from the endomembrane to the PM. Interestingly, whereas the dual signals of CAAX processing and acylation occur in cis on H-Ras, they are in trans on heterotrimeric G proteins. More important, these data show that interaction between Galpha and Gbetagamma does not stabilize independent binding to PM of palmitoylated Galpha and prenylated Gbetagamma, as previously thought , but rather that association occurs even in the absence of palmitoylation and that PM targeting occurs after trimer formation on endomembrane. Figure 3 | Galpha palmitoylation acts as a second signal for PM targeting of GFP-Ggamma. Galpha palmitoylation acts as a second signal for PM targeting of GFP-Ggamma. COS-1 cells were transfected with GFP-H-Ras (a, b), GFP-H-Ras with mutated palmitoylation sites (c), GFP-Ggamma2 (d --f), or GFP-Ggamma1 (g --i). The Gbeta1 subunit was coexpressed with GFP-Ggamma2 or GFP-Ggamma1 as indicated (d --i) along with either Galphai2 (d, e, g, h) or palmitoylation-deficient Galphai2C3S (f, i). An inhibitor of palmitoylation, 2-BP, was added in b, e, and h. Bars, 10 mum. Galpha Interacts with Gbetagamma on Golgi | To confirm directly that Galpha interacts with Gbetagamma on endomembrane, we tagged with CFP the C-termini of Galphai2 and Galphai2C3S and coexpressed these fusion proteins with or without Gbeta1 and Ggamma2 tagged at the N-terminus with YFP. Galphai2-CFP expressed with Gbeta1 and YFP localized on both the PM and Golgi (Figure a). The PM localization is most likely a consequence of association with endogenous Ggamma. When Galphai2-CFP was coexpressed with Gbeta1 and YFP-Ggamma2, the two tagged subunits colocalized on PM and Golgi, but only YFP-Ggamma2 was observed on ER (Figure b). This observation suggests that whereas CAAX-processed Gbetagamma traffics from cytosol to ER and then onto Golgi and PM, association with Galpha takes place on the Golgi, a compartment on which palmitoyltransferase activity resides . Galphai2C3S-CFP expressed with Gbeta1 and YFP was largely cytosolic, although some of the fusion protein was enriched in the paranuclear region around the Golgi area (Figure c). When Galphai2C3S-CFP was coexpressed with Gbeta1 and YFP-Ggamma2, the palmitoylation-deficient Galpha was recruited to the Golgi region in association with YFP-Ggamma2, but neither of the fusion proteins was observed on the PM (Figure d). Thus, unpalmitoylated Galpha associated with Gbetagamma on the Golgi, but in the absence of palmitoylation, neither subunit was translocated to the PM. Figure 4 | Galpha and Gbetagamma colocalize on Golgi. Galpha and Gbetagamma colocalize on Golgi. COS-1 cells were cotransfected with Galphai2-CFP (a and b) or Galphai2C3S-CFP (c and d) and either YFP plus Gbeta1 (a and c) or YFP-Ggamma2 plus Gbeta1 (b and d). Dual color images of living cells were imaged 24 h after transfection with a Zeiss 510 LSM. The CFP channel is assigned red, the YFP channel is assigned green, and colocalization is indicated by yellow pseudocolor. Arrows indicate PM ruffles, and the arrowhead indicates Golgi. Bars, 10 mum. Carboxyl Methylation Is Necessary for Endomembrane Targeting of Farnesylated but Not Geranylgeranylated Ggamma Subunits | Membrane localization of farnesylated Ras proteins is dependent not only on prenylation but also on carboxyl methylation of the CAAX motif . However, in vitro analysis of the association of prenylated peptides with liposomes suggested that the added hydrophobicity of the 20-carbon geranylgeranyl modification found on most Ggamma subunits may be sufficient for membrane association in the absence of carboxyl methylation . To test the role of carboxyl methylation in the localization of farnesylated Ggamma1 and geranylgeranylated Ggamma2 on endomembrane, we expressed GFP-tagged Ggamma1 and Ggamma2 in spontaneously immortalized MEFs derived from mouse embryos null for pcCMT (CMT-/-) or their wild-type littermates (CMT+/+). Laser scanning confocal microscopy was used to analyze MEFs. The morphology of the endomembrane system of MEFs (revealed by observing live cells expressing GFP-CAAX; data not shown) differed from that observed in established cell lines such as COS-1 or MDCK in that, whereas the ER of COS-1 cells consisted of a well-defined nuclear envelope and peripheral reticulum, the endomembrane system of MEFs appeared polymorphic, with reticulum interspersed with numerous cytoplasmic vesicles. This pattern was also observed when GFP-tagged Ggamma1 or Ggamma2 was expressed in CMT+/+ cells (Figure , a and c). An identical pattern was observed when GFP-Ggamma2 was expressed in CMT-/- cells (Figure d), indicating that geranylgeranylated Ggamma2 did not require carboxyl methylation for stable association with the endomembrane. In contrast, GFP-Ggamma1 was observed in the cytosol and nucleoplasm of CMT-/- cells (Figure b), indicating that carboxyl methylation was necessary for stable membrane association of this farnesylated molecule. Thus, although geranylgeranylated Ggamma subunits are substrates for carboxyl methylation , this modification is not required for stable association with endomembrane. Figure 5 | Endomembrane targeting of farnesylated GFP-Ggamma1, but not geranylgeranylated GFP-Ggamma2, requires pcCMT. Endomembrane targeting of farnesylated GFP-Ggamma1, but not geranylgeranylated GFP-Ggamma2, requires pcCMT. GFP-Ggamma1 (a, b, e --j) or GFP-Ggamma2 (c and d) were transfected into CMT+/+ (a, c, e, g, and i) or CMT-/- (b, d, f, h, and j) MEFs alone (a --d) or with either Gbeta1 only (e and f), or Gbeta1 and Galphai2 (g and h), or Gbeta1 and Galphaq (i and j) and imaged alive 24 h later by LSM. GFP-Ggamma1 remained in the cytosol and nucleoplasm in CMT-/- cells when expressed alone or coexpressed only with Gbeta1 but was localized to PM when coexpressed with Gbeta1 and either Galpha subunit. Bars, 10 mum. Having determined that carboxyl methylation is required for stable association of Ggamma1 with endomembrane, we next tested whether carboxyl methylation of farnesylated Ggamma1 is required for the Galpha-mediated delivery of the trimer to the PM. GFP-Ggamma1 was coexpressed with Gbeta1 alone or with Gbeta1 and Galphai2 in CMT+/+ and CMT-/- cells. GFP-Ggamma1 coexpressed with Gbeta1 alone was localized to internal membranes in CMT+/+ cells (Figure e) but was cytosolic in CMT-/- cells (Figure f), similar to results obtained with GFP-Ggamma1 expressed alone (Figure , a and b). Nevertheless, GFP-Ggamma1 coexpressed with Gbeta1 and Galphai2 was localized to the PM in both CMT+/+ and CMT-/- cells (Figure , g and h). Similar results were obtained with Galphaq (Figure , i and j). This indicates that the more stable endomembrane association of Gbetagamma1 dimers mediated by carboxyl methylation is not required for heterotrimer formation and trafficking to the PM. Thus, co-overexpression of Galpha rescues the trafficking defect of farnesylated but unmethylated Ggamma1. Whether unmethylated Ggamma1 becomes associated with Galpha subunits that have reached the PM by virtue of association with endogenous Ggamma or whether unmethylated Ggamma1, despite markedly diminished affinity for endomembrane, can associate with Galpha on the Golgi before transport to the PM remains unresolved. DISCUSSION : The CAAX motif, shared by Ras and Rho family proteins and the Ggamma subunits of heterotrimeric G proteins, signals for prenylation that targets the protein to the ER, where it encounters the Rce1 protease and pcCMT . Whereas N-Ras and H-Ras then traffic by vesicular transport to the PM via the Golgi, K-Ras4B takes an alternative, as yet uncharacterized path . The signal in Ras for engagement of each of these pathways is contained in the so-called "second signal" that lies adjacent to the CAAX motif and consists of either cysteines that are sites of palmitoylation (N-Ras and H-Ras) or a polybasic domain (K-Ras4B). The trafficking of Rho family GTPases is more complex, because several members of this family bind to RhoGDIalpha, a ubiquitously expressed chaperone that has the capacity to retain C-terminally processed Rho proteins in the cytosol. Thus, for the Rho proteins, a combination of CAAX motif processing, a second signal, and affinity for RhoGDIalpha determines their final membrane localization . Heterotrimeric G proteins reside in the PM in an inactive, GDP-bound, trimeric form until association with an activated receptor triggers nucleotide exchange and dissociation of Galpha from Gbetagamma. The mechanisms that target newly synthesized Galpha subunits to the PM have been explored in some depth. A combination of palmitoylation and association with Gbetagamma is necessary for stable PM association of Galpha . However, if Gbetagamma is necessary for proper targeting of Galpha to the PM, what targets Gbetagamma? Recent evidence suggests that the Gbetagamma3 dimer is found predominantly on internal membranes in the absence of Galpha . The conclusion that these authors drew from this observation was that Gbetagamma interaction with Galpha serves to stabilize the otherwise transient PM association of Galpha. Our study presents evidence that the role of Gbetagamma is not to stabilize independent PM association of Galpha but rather to act cooperatively with Galpha to target the entire trimer from the Golgi to PM. We demonstrate that the intrinsic targeting of Gbetagamma is to the ER and Golgi, and only when complexed with Galpha is there further trafficking to the PM. GFP-tagged Ggamma expressed alone or coexpressed with Gbeta appeared predominantly on the ER, whereas GFP-tagged Ggamma coexpressed with Gbeta and Galpha appeared on the PM and Golgi. When palmitoylation of Galpha is prevented, either by mutation of the palmitoylated cysteine residue to serine or by treatment with 2-BP, Gbetagamma accumulates on ER, and the heterotrimer remains on the Golgi. This result would not be expected if association of Galpha and Gbetagamma occurred initially at the PM. However, it is the result that would be expected if G protein heterotrimers behave like H-Ras, which at steady state appears in the PM and Golgi but is retained in the ER if palmitoylation is blocked . We conclude that a combination of targeting elements within Gbetagamma (the CAAX motif) and Galpha (palmitoylation and/or myristoylation) acts cooperatively in trans to target the entire trimer to the PM. Accordingly, nascent trimer formation must occur on endomembrane before translocation of the complex to the PM. Coexpression of Galpha and Gbetagamma tagged with resolvable spectral mutants of GFP revealed colocalization on Golgi and PM but not ER, suggesting that heterotrimer formation occurs on Golgi. This division of the two targeting signals into different subunits may have evolved to ensure that only a complete trimer (which represents an inactive signaling unit) can reach the PM. Because Gbetagamma signaling requires only release from Galpha on receptor-mediated nucleotide exchange, it is imperative that nascent Gbetagamma is able to reach the PM only in association with GDP-bound Galpha. If Gbetagamma alone, like Ras, could reach the PM in the absence of GDP-bound Galpha, there would be nothing to stop the Gbetagamma subunits from prematurely engaging their downstream effectors even in the absence of receptor activation. Thus, it is possible that the two signals in the cis mechanism that targets monomeric GTPases to PM have been modified for the heterotrimeric G proteins into a two signal in trans mechanism to avoid premature Gbetagamma signaling. Although no protein palmitoyltransferase has been characterized at the molecular level, an important conclusion that can be deduced from the data presented here is that at least one enzyme that palmitoylates Galpha is localized in an endomembrane compartment, most likely Golgi. Although such a conclusion is contrary to the prevailing view that places Galpha palmitoyltransferase in the PM , Galpha palmitoyltransferase activity has, in fact, been detected in Golgi fractions . Moreover, the Golgi has been implicated as the compartment in which the enzyme that palmitoylates H-Ras resides . Similarly, in vitro palmitoyltransferase activity for the neuronal plasticity protein GAP-43 was found in Golgi . Prenylcysteine carboxyl methylation is a modification of the CAAX motif that has been well conserved from yeast to humans, although its precise role in protein targeting and GTPase signaling remains uncertain. Whereas carboxyl methylation of yeast GTPases is not required for growth , disruption of the CMT gene by homologous recombination has revealed that the gene is required for mouse development (embryonic lethal day 10.5). Ras proteins are mislocalized in CMT-/- cells . It has been suggested that the elimination of the negative charge of the carboxy terminus of prenylated proteins accomplished by carboxyl methylation adds to the hydrophobicity of the C terminus and that the additional hydrophobicity is of much greater consequence to proteins modified by the 15-carbon farnesyl polyisoprene than to those modified by the 20-carbon geranylgeranyl lipid . Our study directly tests this hypothesis in live cells by examining the localization of GFP-Ggamma1 and GFP-Ggamma2 in CMT-/- MEFs . In these cells, farnesylated GFP-Ggamma1 was unable to associate stably with endomembranes, even though geranylgeranylated GFP-Ggamma2 localized normally on ER. The conservation through evolution of two different CAAX prenyl transferases suggests distinct biological roles for the farnesyl and geranylgeranyl modifications. Our data suggest that whereas geranylgeranylation imparts a relatively high affinity for membranes independent of carboxyl methylation, farnesylation affords only a weak affinity that is then modulated by carboxyl methylation. Because Galphai2 is myristoylated even in the absence of palmitoylation and palmitoylation-deficient Galphai2C3S failed to cooperate with processed Ggamma to target heterotrimers to the PM, we conclude that myristoylation alone is not able to act as a second signal for PM targeting. This observation has implications for transducin, because Galphat is modified only with a myristoyl group. The combination of the myristoylated Galphat and the farnesylated Ggamma1 would be predicted to yield a heterotrimer whose PM targeting may be inefficient and whose endomembrane association is dependent on carboxyl methylation. This is in agreement both with the relatively high amount of transducin found in the soluble fraction of retinal preparations and with the observation that unmethylated Gbetagamma1 associates with membranes only poorly . This weak association of the transducin trimer and its subunits with cellular membranes is in sharp contrast to the relatively stable membrane interactions of most other fully processed heterotrimers studied. It is interesting to speculate that the relatively weak membrane targeting signals and unique requirement for carboxyl methylation, a modification that is reversible under physiological conditions, of transducin play a role in the biology of the visual signaling pathway. Together, our data support a model for G protein trafficking analogous to that described for Ras . Prenylation of the CAAX motif of Ggamma directs Gbetagamma to the ER, where the prenyl-CAAX sequence is cleaved and carboxyl methylated, the latter modification contributing significantly to the affinity of farnesylated Ggamma1 for the endomembrane. Nascent Galpha then associates with Gbetagamma on the Golgi, where it is palmitoylated, a modification that serves as a signal for further transport to the PM. This model is consistent with previous data on Galpha membrane association but adds a trafficking dimension largely overlooked in previous studies of G proteins. Figure 6 | Model of G-protein trafficking to the PM. Model of G-protein trafficking to the PM. All three G protein subunits are synthesized in the cytosol on free polysomes. Gbeta and Ggamma immediately dimerize on the basis of their high affinity for each other (1). Whether this occurs before or after Ggamma prenylation is uncertain. Prenylation of Ggamma (2) drives Gbetagamma to the cytosolic face of the ER (3), where it encounters the CAAX protease and carboxyl methyltransferase (4). Fully processed Gbetagamma is then delivered to the cytosolic face of the Golgi, where it recruits Galpha (5), which is then acylated by a Golgi resident acyl transferase (6). Acylation then allows the G protein heterotrimer to be transported as a holoenzyme to the PM via a route (7) that may be the classic secretory pathway. Backmatter: PMID- 12221134 TI - Regulation of Flagellar Dynein by Calcium and a Role for an Axonemal Calmodulin and Calmodulin-dependent Kinase AB - Ciliary and flagellar motility is regulated by changes in intraflagellar calcium. However, the molecular mechanism by which calcium controls motility is unknown. We tested the hypothesis that calcium regulates motility by controlling dynein-driven microtubule sliding and that the central pair and radial spokes are involved in this regulation. We isolated axonemes from Chlamydomonas mutants and measured microtubule sliding velocity in buffers containing 1 mM ATP and various concentrations of calcium. In buffers with pCa > 8, microtubule sliding velocity in axonemes lacking the central apparatus (pf18 and pf15) was reduced compared with that of wild-type axonemes. In contrast, at pCa4, dynein activity in pf18 and pf15 axonemes was restored to wild-type level. The calcium-induced increase in dynein activity in pf18 axonemes was inhibited by antagonists of calmodulin and calmodulin-dependent kinase II. Axonemes lacking the C1 central tubule (pf16) or lacking radial spoke components (pf14 and pf17) do not exhibit calcium-induced increase in dynein activity in pCa4 buffer. We conclude that calcium regulation of flagellar motility involves regulation of dynein-driven microtubule sliding, that calmodulin and calmodulin-dependent kinase II may mediate the calcium signal, and that the central apparatus and radial spokes are key components of the calcium signaling pathway. Keywords: INTRODUCTION : Our goal is to understand the molecular mechanism by which calcium regulates the size and shape of ciliary and flagellar bends to modulate motility. For many organisms or cell types, external stimuli trigger changes in cytosolic free calcium concentration, which in turn produce altered ciliary and flagellar motility. Much of what we know about calcium modulation of ciliary and flagellar motility comes from studies of isolated axonemes or demembranated cell models reactivated to beat in vitro. For example, in the biflagellate green alga Chlamydomonas reinhardtii, calcium is required for phototaxis as well as the photophobic response (reviewed in ). In vitro studies using reactivated cell models indicate that small increases in calcium (pCa9 - pCa7) differentially activate one flagellum or the other . A larger increase in calcium (pCa5 - pCa4) causes a momentary cessation of motility followed by a complete switch from an asymmetric to a symmetric waveform (Figure ; ; ; ). Axonemes isolated from sea urchin sperm and reactivated in vitro under low calcium conditions beat with a symmetric waveform. Upon increasing calcium in the buffer, the axonemes beat with increasing asymmetry ; at extremely high calcium concentrations, quiescence is induced . For reactivated cell models of Paramecium and Tetrahymena, an increase in calcium induces reversal of swimming direction by changing the direction of the ciliary effective stroke . Although the axonemal response to changes in calcium concentration has been well described, we still do not know the precise molecular mechanism by which motility is modulated by fluctuations in the concentration of cytosolic free calcium. Figure 1 | (a) In vivo, Chlamydomonas cells normally swim forward, toward the light, with an asymmetric, ciliary waveform. (a) In vivo, Chlamydomonas cells normally swim forward, toward the light, with an asymmetric, ciliary waveform. During the photophobic response, bright light induces a shift from an asymmetric waveform to a symmetric, flagellar waveform, and the cells swim in reverse. The arrows indicate swimming direction (for example, see and ). (b) This change in waveform can be induced in vitro. Isolated axonemes lacking membranes and soluble flagellar matrix components beat with an asymmetric waveform in buffers of pCa < 8 and beat with a symmetric waveform in buffers of pCa4. (Waveform traces adapted from .) The in vitro reactivation experiments described above clearly demonstrate that all of the regulatory proteins required for modulating motility, including key calcium sensors, are structural components of the axoneme. Several highly conserved calcium-binding proteins are associated with the axoneme. Calmodulin has been identified as a component of ciliary and flagellar axonemes of Chlamydomonas, Paramecium, Tetrahymena, Elliptio, and sperm cells from mammals and echinoderms (reviewed in ; ; ). In Chlamydomonas, a subset of axonemal calmodulin is associated with the radial spokes . In addition, cilia and flagella also contain the calcium-binding protein centrin/caltractin , which is a component of a subset of inner dynein arms . And, have determined that the 18-kDa light chain of the outer dynein arm in Chlamydomonas is a calcium-binding protein with homology to both calmodulin as well as centrin. Therefore, cilia and flagella contain at least three different classes of calcium-binding proteins that predictably mediate calcium control of motility. In addition to sensing changes in calcium, the axoneme must also possess a mechanism for converting the calcium signal into altered axonemal bends, presumably resulting from localized modulation of dynein-driven microtubule sliding (reviewed in ). The relationship between changes in intraflagellar free calcium concentration and predicted changes in dynein activity has not yet been determined. To test the hypothesis that calcium regulates axonemal dynein, our strategy was to assess dynein activity in axonemes isolated from mutant and wild-type cells using an in vitro assay to measure dynein-driven microtubule sliding velocity . This assay has two key advantages. First, measurement of microtubule sliding in isolated axonemes assesses dynein activity in situ with most or all of the endogenous regulatory components intact. Second, the availability of Chlamydomonas mutants with axonemes lacking particular structures provides an opportunity to detect regulatory mechanisms not easily revealed in wild-type axonemes. For example, although axonemes isolated from radial spoke and central apparatus defective mutants cannot be reactivated in vitro in buffers containing 1 mM ATP, dynein activity in these mutants can still be assessed using the microtubule sliding assay . Studies using this assay have provided crucial information toward the development of a model in which axonemal dynein is regulated by the coordinate action of several kinases and phosphatases anchored to the axoneme (, ; reviewed in ). To define the role of calcium in regulating dynein, and hence flagellar waveform, we used the microtubule sliding assay to measure dynein activity in axonemes isolated from wild-type and mutant Chlamydomonas strains in response to calcium. In low calcium conditions, dynein activity is reduced in axonemes lacking the radial spokes and central apparatus. However, in high calcium conditions, dynein activity is restored to nearly wild-type levels in mutant axonemes lacking the entire central apparatus. Furthermore, the increase in dynein activity is inhibited by the addition of either calmodulin or calmodulin-dependent kinase II antagonists. These studies provide evidence that dynein activity is regulated by calcium, that this regulation involves a signaling pathway that includes an axonemal calmodulin and calmodulin-dependent kinase, and that the calcium control system includes the radial spokes and central apparatus. MATERIALS AND METHODS : Cell Strains and Growth Conditions | Strain A54-e18 (nit1-1, ac17, sr1, mt+) is the "wild-type" strain used in motility assays and is the strain used in transformation experiments to obtain the insertional pf16 allele, pf16C . The central pair --defective strains, pf18, and pf15, and the radial spoke --defective strains pf14 and pf17 were obtained from the Chlamydomonas Genetics Center (Duke University). All cells were grown in constant light in TAP media . Isolation of Axonemes and the Microtubule Sliding Assay | Flagella were severed from cell bodies by the dibucaine method and isolated by differential centrifugation in buffer A (10 mM HEPES, pH 7.4, 5 mM MgSO4, 1 mM DTT, 0.5 mM EDTA, and 50 mM potassium acetate). Axonemes were isolated by adding NP-40 (Calbiochem, La Jolla, CA) to flagella for a final concentration of 0.5% (wt/vol) to remove flagellar membranes. Measurement of sliding velocity between doublet microtubules was based on the methods of . Approximately 8 mul of axonemes were applied to a perfusion chamber ; the chamber was perfused with wash buffer (buffer A containing 1 mM ATP) to remove nonadherent axonemes. To initiate microtubule sliding, the chamber was perfused with motility buffer (buffer A containing 1 mM ATP (Roche Molecular Biochemicals, Indianapolis, IN) and 2 mg/ml Nagarse (Type XXVII Protease; Sigma Chemical Co., St. Louis, MO). Although all of the experiments in this report were performed using Nagarse, it should be noted that this protease is no longer available. The supplier recommended replacement is Type VIII protease (catalogue number P-5380; Sigma). We have recently used Type VIII protease in microtubule sliding assays and detected no qualitative or quantitative differences in microtubule sliding. For experiments involving buffers with different concentrations of free calcium, all buffers were made as described in minus polyethylene glycol, creatine kinase, and phosphocreatine. For pharmacological treatments, inhibitors were added to isolated axonemes followed by a 20-min incubation at room temperature. Inhibitors were maintained in both the wash and motility buffers when appropriate. Microtubule sliding was observed using an Axioskope 2 microscope (Zeiss Inc., Thornwood, NY) equipped for dark-field optics including a Plan-Apochromate 40x oil immersion objective with iris and ultra dark-field oil immersion condenser. Images were recorded by a silicon-intensified target camera (VE-1000 SIT; Dage-MTI, Inc., Michigan City, IN) through a time-date generator, on videotape by a videocassette recorder (AG-1980; Panasonic, Secaucus, NJ). Microtubule sliding velocity was measured manually from calibrated video screens using the jog/shuttle device to measure displacement versus time. All data are presented as mean +- SD. The Student's t test was used to determine the significance of differences between means. Kinase, Phosphatase, and Calmodulin Inhibitors | All inhibitors were purchased from Calbiochem. Microcycstin-L-R was prepared as a 500 muM stock in methanol. DRB (5, 6-dichloro-1-b-d-ribofuranosylbenzimidazole) was prepared as a 10 mM stock solution in ethanol. For microtubule sliding assays, the final concentrations of these inhibitors were: microcystin, 1 muM; DRB, 100 muM. KN-92 and KN-93 were prepared as 1.0 mM stocks in water and were used at a final concentration of 1.0 muM . The calmodulin binding domain peptide (Leu-Lys-Lys-Phe-Asn-Ala-Arg-Arg-Lys-Leu-Lys-Gly-Ala-Ile-Leu-Thr-Thr-Met-Leu-Ala), calmodulin inhibitory peptide (Arg-Arg-Lys-Trp-Gln-Lys-Thr-Gly-His-Ala-Val-Arg-Ala-Ile-Gly-Arg-Leu), and calmodulin inhibitory peptide control peptide (Arg-Arg-Lys-Glu-Gln-Lys-Thr-Gly-His-Ala-Val-Arg-Ala-Ile-Gly-Arg-Glu) were prepared as 6.0 mM stocks in water and used at a final concentration of 60 muM or as indicated in figures . The autocamtide-2 --related inhibitory peptide (Lys-Lys-Ala-leu-Arg-Arg-Gln-Glu-Ala-Val-Asp-Ala-Leu) was stored as a 300 muM stock and used at a concentration of 3 muM or as indicated . All stock solutions were stored at -20C. RESULTS : Calcium Regulation of Microtubule Sliding | Many studies have indicated that the central apparatus and radial spokes are involved in a signal transduction pathway that includes several axoneme-associated enzymes that regulate dynein activity to produce the bending motion characteristic of eukaryotic ciliary and flagellar motility. It is also well established that changes in intraflagellar calcium modulate the size and shape of ciliary and flagellar bends (see Figure ). Based on these observations, we postulated that changes in free calcium concentration would differentially affect dynein activity in Chlamydomonas mutants lacking the radial spokes or central apparatus. To test this, we used a microtubule-sliding assay to measure dynein-driven microtubule sliding velocity in axonemes isolated from radial spoke and central apparatus defective Chlamydomonas mutants and as a function of calcium. Table 1 | Chlamydomonas mutants for this study We first compared microtubule sliding velocity in motility buffer containing 10-4 M calcium (pCa4) with that in buffer containing <10-8 M calcium (Figure a). Microtubule sliding velocity in wild-type axonemes was not affected by changes in the concentration of free calcium (17.0 --19.0 mum/s, Figure a). Similarly, microtubule sliding velocity in axonemes isolated from the radial spoke defective mutants pf14 and pf17 and the C1 central microtubule defective strain pf16 was not affected by changes in calcium concentration. Axonemes isolated from these mutants have the same slow microtubule sliding velocity in buffers of either low or high calcium (between 7.5 and 8.5 mum/s, Figure a). In striking contrast, the velocity of microtubule sliding in mutant axonemes lacking the entire central apparatus, pf18 and pf15, was calcium sensitive. The velocity of microtubule sliding was slow in low calcium buffer (between 7.5 and 8.5 mum/s, Figure a), yet increased to nearly wild-type velocity in high calcium buffer (16.0 --17.0 mum/s, Figure a). Evidently, the inhibition of dynein activity caused by the lack of the central apparatus is bypassed by the presence of high calcium. However, the calcium-induced rescue of dynein activity fails if the C2 microtubule of the central apparatus is present (pf16) or if radial spoke components are lacking (pf14 and pf17). Figure 2 | (a) Microtubule sliding velocities in wild-type and mutant axonemes in low calcium (pCa > 8, black bars) and high calcium (pCa4, gray bars) buffers. (a) Microtubule sliding velocities in wild-type and mutant axonemes in low calcium (pCa > 8, black bars) and high calcium (pCa4, gray bars) buffers. Sliding velocities in wild-type, radial spoke --defective (pf14 and pf17), and C1 central tubule --defective (pf16) axonemes in low calcium buffer are not significantly different from those in high calcium buffer. In contrast, sliding velocities in axonemes completely lacking the central apparatus (pf15 and pf18) in high calcium buffer are significantly increased from those in low calcium buffer (p < 0.001; Student's t test). All bars represent the mean of >60 measurements +- SD from a minimum of three experiments. (b) Microtubule sliding velocity in pf18 axonemes in microtubule sliding buffer with varying concentrations of free calcium. As the concentration of free calcium increases, microtubule sliding velocity in pf18 axonemes increases to nearly wild-type level. Each point represents the mean of >40 measurements +- SD from two experiments. The sliding velocity of axonemes isolated from pf18 increased in a linear manner with increasing free calcium concentration (Figure b). The half-maximal velocity occurs between pCa5 and pCa6, the same concentration of free calcium that induces the switch from ciliary to flagellar waveforms . These results provide evidence that dynein activity is modulated by calcium and indicate that dynein activity is regulated by the response of a particular enzyme to increasing concentrations of free calcium. The Calcium Signaling Pathway Acts Independently of Casein Kinase 1 and Protein Phosphatase-2A Activity | Previous studies have indicated that in conditions of low calcium, the central apparatus and radial spokes form a signaling pathway that regulates dynein activity, at least in part, through the action of casein kinase 1 (CK1) and protein phosphatase-2A (PP2A) located in the axoneme . This conclusion is based on the observation that upon the addition of DRB, a CK1 inhibitor, dynein activity is restored in mutant axonemes lacking the radial spokes or central apparatus . Moreover, in low calcium buffer, the DRB induced increase in sliding velocity observed for both radial spoke and central apparatus defective axonemes requires the presence of an active phosphatase, most likely PP2A . The addition of microcystin, an inhibitor of PP1 and PP2A, blocks the DRB-induced increase in dynein activity in these mutants . To investigate the relationship between the calcium-mediated increase in dynein activity and the CK1/PP2A-mediated regulation of dynein activity, we compared the effect of CK1 and PP2A inhibitors on microtubule sliding velocity in low versus high calcium buffer conditions. First, we determined whether the DRB-induced increase in dynein activity for pf16 and pf14 axonemes is calcium sensitive by testing whether DRB restores dynein activity to axonemes isolated from these mutants in pCa4 buffer. As previously reported, we found that DRB increases the microtubule sliding velocity of axonemes isolated from pf14 in low calcium buffer (Figure a; ). However, the sliding velocity in pf14 axonemes in pCa4 buffer with DRB is significantly different from sliding velocity in pf14 axonemes in both pCa4 buffer alone (p < 0.001) as well as low calcium buffer with DRB (p < 0.001; Figure a). Therefore, the DRB-induced increase in dynein activity for pf14 axonemes is to some degree calcium sensitive. Even more pronounced effects of calcium were observed for pf16 axonemes. The addition of DRB to pf16 axonemes in low calcium buffer increased dynein activity to nearly wild-type levels . However, the sliding velocity in pf16 axonemes in pCa4 buffer with DRB is not significantly different from that observed in either high or low calcium buffer alone (Figure a). Evidently, in the presence of high calcium, the inhibition of CK1 fails to restore dynein activity to pf16 axonemes. Figure 3 | (a) Microtubule sliding velocities in pf14 and pf16 axonemes. (a) Microtubule sliding velocities in pf14 and pf16 axonemes. Dynein-driven microtubule sliding velocities in axonemes isolated from the radial spoke defective mutant pf14 and central apparatus defective mutant pf16 significantly increase (p < 0.001; Student's t test) upon the addition of DRB compared with control, untreated axonemes in low calcium buffer (C = pCa > 8). In high calcium buffer, (Ca = pCa4) DRB is less effective at increasing microtubule sliding velocity in pf14 axonemes and is completely ineffective at increasing sliding velocity in pf16 axonemes. (b) The increase in sliding velocity observed for pf18 axonemes in high calcium buffer is not inhibited by the addition of microcystin (MC). All bars represent the mean +- SD. N >= 60 from a minimum of three experiments. Second, we used specific inhibitors to investigate the involvement of axonemal phosphatases in regulating dynein activity. As noted, the addition of microcystin, a potent inhibitor of phosphatases PP1 and PP2A, blocks the DRB induced increase in sliding velocity in both radial spoke and central apparatus defective mutants . To determine whether the calcium-mediated increase in dynein activity for pf18 axonemes also requires the presence of PP1 or PP2A, axonemes were incubated with microcystin before and during induction of microtubule sliding in pCa4 buffer. The addition of microcystin does not inhibit the calcium-induced increase in sliding velocity in pf18 axonemes (Figure b). Therefore, neither PP1 nor PP2A activity is required for the calcium-induced increase in dynein activity. These combined results indicate that the calcium-induced increase in dynein activity in pf18 axonemes occurs by a mechanism not directly dependent on the activity of CK1 or PP2A. Calmodulin Plays a Role in the Calcium-dependent Regulatory Pathway | As demonstrated above, high calcium does not increase sliding velocities in axonemes isolated from either pf14 or pf17 mutants. Therefore, we predicted that an important component of the calcium-mediated signaling pathway must be either missing or inactivated in pf14 and pf17 axonemes. Several calcium-binding proteins reside within the axoneme, including calmodulin . have recently shown that a fraction of axonemal calmodulin is associated with the radial spokes; axonemes isolated from pf14 lack this fraction of calmodulin. One possibility is that calcium-induced increase in dynein activity in pf18 axonemes involves a calmodulin-mediated regulatory pathway. If this were the case, we predicted that the addition of calmodulin antagonists would inhibit dynein activity in pf18 axonemes in high calcium buffers. A significant advantage of using the microtubule sliding assay to measure dynein activity is that we have complete experimental access to the regulatory machinery associated with isolated axonemes. Therefore, the use of pharmacological agents, such as peptide inhibitors, has been particularly effective in elucidating signal transduction components important for dynein regulation. To test whether calmodulin is involved in calcium-induced dynein regulation, we tested two peptide inhibitors of calmodulin for their ability to reduce dynein activity in pf18 axonemes in pCa4 buffer. These peptide inhibitors represent the calmodulin binding sites for two different enzymes, myosin light chain kinase, and CaM-kinase II; when bound to calmodulin, they prevent interaction with calmodulin-binding proteins . Both the calmodulin-binding domain (CBD) peptide of CaM-kinase II as well as the calmodulin inhibitory peptide (CIP, calmodulin-binding domain of myosin light-chain kinase) inhibit the high calcium-induced increase in dynein activity in pf18 axonemes (Figure a). The velocity of microtubule sliding in pf18 axonemes incubated in pCa4 buffer with either of these calmodulin inhibitors is not significantly different from that in pf18 axonemes incubated in low calcium buffer. Importantly, sliding velocity in pf18 axonemes incubated in pCa4 buffer with the control peptide for the calmodulin inhibitory peptide (CIPc) was not significantly different from that in pf18 axonemes in pCa4 buffer alone. The velocity of microtubule sliding in wild-type axonemes in pCa4 buffer was unaffected by the addition of the CBD peptide. Figure 4 | (a) The addition of calmodulin inhibitors blocks the calcium-induced increase in sliding velocity observed for pf18 axonemes in pCa4 buffer (Ca). (a) The addition of calmodulin inhibitors blocks the calcium-induced increase in sliding velocity observed for pf18 axonemes in pCa4 buffer (Ca). The sliding velocity in pf18 axonemes incubated in pCa4 buffer with either the calmodulin-binding domain peptide (CBD, 60.0 muM) or the calmodulin inhibitory peptide (CIP, 60.0 muM) is not significantly different from that in pf18 axonemes incubated in low calcium buffer. Sliding velocity in pf18 axonemes in pCa4 buffer incubated with the control peptide (CIPc, 60.0 muM) for the calmodulin inhibitory peptide is not significantly different from that in pf18 axonemes in pCa4 buffer alone. All bars represent the mean of >60 measurements +- SD from a minimum of three experiments. (b) Microtubule sliding velocity in pf18 axonemes in microtubule sliding buffer with varying concentrations of the calmodulin binding domain peptide (CBD) inhibitor. Increasing concentration of CBD decreases sliding velocity in pf18 axonemes in high calcium buffer. Each point represents the mean of >40 measurements +- SD from two experiments. To investigate the effective concentration range of calmodulin inhibitor necessary for reducing dynein activity, pf18 axonemes were incubated in various concentrations of the CBD peptide, and the microtubule sliding assay was performed. The CBD peptide decreased microtubule sliding velocity with half-maximal inhibition at similar6.0 x 10-7 M (Figure b). These results provide evidence that the increase in sliding velocity of pf18 axonemes in high calcium buffers is mediated by an axonemal calmodulin. The Calmodulin-dependent Increase in Sliding Velocity is Mediated by a Calmodulin-dependent Kinase | Our results indicate that the increase in microtubule sliding velocity in pf18 axonemes in high calcium buffer is mediated by calmodulin. Calmodulin is known to bind to a variety of enzymes including calmodulin-dependent kinases (CaM-kinase), calcineurin or protein phosphatase 2B (PP2B), and cyclic nucleotide phosphodiesterases (reviewed in ). To investigate whether any of these enzymes are possible targets of the calmodulin-mediated increase in microtubule sliding velocity, we tested available inhibitors in our sliding assay for their ability to block the calcium-induced increase in dynein activity in pf18 axonemes in high calcium buffer. An inhibitor (KN-93) of calmodulin-dependent kinase II (CaM-KII) significantly reduced dynein activity in pf18 axonemes in high calcium buffer (Figure a). The microtubule sliding velocity in pf18 axonemes incubated with KN-93 in the presence of high calcium is not significantly different from that in pf18 axonemes incubated in low calcium buffer alone. Importantly, the control compound KN-92 did not reduce microtubule sliding velocity in pf18 axonemes in pCa4 buffer. The sliding velocity in pf18 axonemes incubated with KN-92 in the presence of high calcium is not significantly different from that in pf18 axonemes incubated in high calcium buffer alone. Figure 5 | (a) The addition of CaM-KII inhibitors blocks the calcium-induced increase in sliding velocity observed for pf18 axonemes in pCa4 buffer (Ca). (a) The addition of CaM-KII inhibitors blocks the calcium-induced increase in sliding velocity observed for pf18 axonemes in pCa4 buffer (Ca). The sliding velocity in pf18 axonemes incubated in pCa4 buffer with either the compound KN-93 (1.0 muM) or the autocamtide-2 related inhibitory peptide (AIP, 3.0 muM) is not significantly different from that in pf18 axonemes incubated in low calcium buffer. Sliding velocity in pf18 axonemes in pCa4 buffer incubated with the control compound KN-92 (1.0 muM) is not significantly different from that in pf18 axonemes in pCa4 buffer alone. All bars represent the mean of >60 measurements +- SD from a minimum of three experiments. (b) Microtubule sliding velocity in pf18 axonemes in pCa4 microtubule sliding buffer with varying concentrations of the autocamtide-2 related inhibitory peptide (AIP). Increasing concentration of AIP decreases sliding velocity in pf18 axonemes in high calcium buffer. Each point represents the mean of >40 measurements +- SD from two experiments. We also investigated whether a specific peptide inhibitor of CaM-KII reduced dynein activity in pf18 axonemes in high calcium buffer. The autocamtide-2 --related inhibitory peptide (AIP) is a potent and specific inhibitor of CaM-KII with a reported Ki of 2 --8 x 10-9 M . On the addition of AIP, dynein activity in pf18 axonemes in pCa4 buffer is significantly reduced (Figure a); the velocity of microtubule sliding in pf18 axonemes incubated with AIP in pCa4 buffer was not significantly different from that in low calcium buffer alone. In experiments using varying concentrations of AIP, half-maximal inhibition was achieved at AIP concentrations of 1.6 x 10-7 M (Figure b). These results indicate that a signaling pathway that includes calmodulin and a calmodulin-dependent kinase controls dynein-driven microtubule sliding in response to calcium. DISCUSSION : In the work described here, we present data demonstrating that calcium regulation of flagellar motility involves regulation of dynein-driven microtubule sliding. In addition, our results suggest that calmodulin is a key axonemal calcium sensor and that a calmodulin-dependent kinase may mediate the calcium signal. Finally, these studies reveal that the calcium control system is regulated by the central apparatus and radial spokes. We propose that in wild-type axonemes, the central apparatus locally controls the calcium sensor to locally regulate microtubule sliding and to modulate the size and shape of flagellar bends. These conclusions are consistent with genetic analyses implicating the radial spokes and central apparatus in control of flagellar waveform . Calcium Modulates Dynein Activity | Our analysis of dynein-driven microtubule sliding velocity in central apparatus --defective mutants has revealed a role for calcium in modulating dynein activity. This conclusion is founded on the observation that buffers with high calcium concentration restored dynein activity to nearly wild-type levels in mutant axonemes completely lacking the central apparatus. In addition, dynein activity increased in a linear manner with increasing concentrations of calcium. This result was surprising because the calcium-induced switch from asymmetric to symmetric waveform occurs somewhat abruptly with increasing calcium. One explanation for this result is that increasing calcium concentration affects the concentration of MgATP in the buffer, which in turn, directly affects dynein activity. However, this explanation is unlikely because increased dynein activity is not observed in all mutants and is abolished upon the addition of specific inhibitors. report the concentration of CaATP2- at pCa4 to be similar10-5 M in buffer virtually identical to that used in our assay. A second explanation is that the effect of calcium on dynein activity is mediated by the response of a particular enzyme to increasing concentrations of free calcium. Based on their studies of isolated axonemes reactivated in vitro, concluded that the flagellar response to calcium is a multicomponent process involving subtle quantitative changes in flagellar movement that ultimately affect beat frequency and waveform. The microtubule sliding assay used here to assess dynein activity in particular mutants has provided us with a unique opportunity to resolve subtle details of dynein regulation that would not easily be detected in wild-type axonemes. Our results using mutant axonemes and inhibitors of specific enzymes are most consistent with a model in which calcium acts directly through an enzyme-driven mechanism to ultimately affect dynein activity. The question is, what is the molecular mechanism by which calcium modulates dynein activity? One possibility is that the direct binding of calcium to dynein arm components modulates the activity of the dynein heavy chains. For example, centrin is component of the inner dynein arms , and a calcium-binding light chain is a component of the outer dynein arms . A second possibility is that calcium activates a signal transduction cascade that ultimately regulates dynein activity by posttranslational modification of axonemal components, possibly including the dynein arms. Although these two possibilities are not mutually exclusive, our data provides greater support for the hypothesis that calcium modulates dynein activity by activating a signal transduction pathway that includes calmodulin and a calmodulin-dependent kinase. Several studies have demonstrated that changes in calcium concentration result in the altered phosphorylation of particular axonemal components . More recent studies have confirmed that axonemal dynein is regulated in part by a network of kinases and phosphatases that are structural components of the axoneme (reviewed in ). In particular, the central apparatus --radial spoke system has been implicated in a signal transduction cascade that controls the activity of an axonemal CK1 and PP2A to regulate the I1 inner dynein arm subform . Also, analyses of phototaxis-defective mutants have suggested that changes in the phosphorylation state of inner dynein arm I1 may play a role in regulating motility during phototaxis, which is a calcium-dependent response . Therefore, one hypothesis is that calcium regulates dynein through a pathway that directly alters the activity of axonemal CK1 and PP2A. For example, high calcium may inhibit CK1 either directly or indirectly, to increase dynein activity in axonemes lacking the central apparatus. Alternatively, calcium alters dynein-driven motility through a separate signaling pathway. Our data supports the hypothesis that calcium does not modulate dynein activity in pf18 and pf15 axonemes by inhibiting CK1. The sliding velocity in pf15 axonemes did not increase upon the addition of DRB but did increase in high calcium. In contrast, dynein activity in both pf14 and pf16 axonemes was restored after the addition of DRB in low calcium buffer but not in high calcium buffer. That sliding velocities do not increase in pf16 axonemes in high calcium in either the presence or absence of DRB suggests that the C2 central microtubule may be sufficient to maintain dynein inhibition in the presence of high calcium regardless of whether CK1 is inhibited. Our data also support the hypothesis that neither PP1 nor PP2A is required for the calcium-mediated regulation of dynein. First, the presence of high calcium restores wild-type sliding velocity to pf18 axonemes in the presence of microcystin, an inhibitor of PP1 and PP2A. Second, pf15 axonemes lack PP2A and yet, dynein activity in pf15 axonemes increases in the presence of high calcium. Therefore, the calcium-mediated increase in dynein activity for these central apparatus defective axonemes does not require either PP1 or PP2A. These results do not, however, rule out the possibility that additional phosphatases such as calcineurin are involved in the calcium-mediated signaling pathway. The Calcium-signaling Pathway Includes Calmodulin and a Calmodulin-dependent Kinase | Our pharmacological data suggest that the calcium-induced increase in dynein activity in central pairless axonemes is mediated by calmodulin and CaM-KII. The addition of either of two specific peptide inhibitors of calmodulin blocks the calcium-mediated increase in dynein activity in central pairless axonemes. It is possible that the calmodulin peptide inhibitors bind to centrin and/or the 18-kDa calcium-binding light chain. However, these proteins share only 45% amino acid identity with calmodulin, whereas Chlamydomonas calmodulin shares 85% amino acid identity with vertebrate calmodulin . Based on the high degree of specificity of these inhibitors, the simplest interpretation is that they bind to and block the function of an axonemal calmodulin. This interpretation is also supported by our results using inhibitors to CaM-KII. KN-93, a specific inhibitor of CaM-KII , and AIP, a highly specific peptide inhibitor of CaM-KII also inhibit dynein activity in central pairless axonemes in high calcium buffer. Importantly, both the control calmodulin inhibitory peptide as well as the control compound for the CaM-KII inhibitor (KN-92) fail to inhibit the calcium-induced increase in dynein activity. Based on these results, and given the concentration at which half-maximal inhibition is achieved, we believe that the affect of these inhibitors is specific. The Radial Spokes and Central Apparatus Are Key Components of the Calcium-signaling Pathway | The results using axonemes from pf18 imply that the calmodulin-dependent mechanism that mediates calcium-induced change in dynein activity is not located in the central apparatus. Our observation that high calcium does not increase sliding velocity in pf14 axonemes suggests that an important component of the calcium-signaling pathway is either missing or inactivated in radial spoke defective axonemes. One prediction is that the calmodulin and calmodulin-dependent kinase that bind to the CBD and AIP peptides, respectively, are components of the radial spokes. As noted, one component of the radial spokes is calmodulin . If the radial spoke associated calmodulin is located in the spoke stalk and is necessary and sufficient for calcium-induced rescue of wild-type sliding velocity, we predicted that the pf17 mutant (lacking the radial spoke heads; , ) would also have restored dynein activity in high calcium. However, the velocity of microtubule sliding in pf17 axonemes was not significantly different from that in pf14 in high calcium buffer. In addition, microtubule sliding in axonemes isolated from pf16 was not restored in high calcium buffer, even although pf16 axonemes contain wild-type radial spokes. Evidently, the assembly of wild-type radial spokes alone is not sufficient for restoring dynein activity in high calcium buffer. Therefore, either the radial spoke --associated calmodulin is not the target of the inhibitors used in these studies, or, in the absence of the C1 central microtubule or radial spoke heads, the radial spoke calmodulin is unable to respond to increases in intraflagellar calcium. The latter hypothesis is intriguing given that the central apparatus rotates during flagellar beating (reviewed in ) and the central apparatus projections make transient contact with the radial spoke heads in active regions of microtubule sliding . Moreover, in recent functional analyses of reactivated axonemes isolated from sea urchin sperm, demonstrate that calcium-induced changes in microtubule sliding are mediated by a rotatable component and suggest that this component is most likely the central apparatus. Based on our observation that changes in calcium concentration differentially affect dynein activity in radial spoke mutants compared with central apparatus defective mutants, we propose that these structures are part of a control system that modulates dynein-driven microtubule sliding to regulate the size and shape of flagellar bends in response to calcium. In contrast, , , and have proposed that the radial spokes and central apparatus are not essential for calcium-induced waveform conversion. Their conclusion is based on the observation that isolated axonemes from central pairless and radial spokeless mutants reactivated at low ATP concentration or in the presence of certain organic compounds undergo waveform conversion in response to changes in calcium concentration. Apparently, at low ATP concentration or in organic compounds dynein is activated and this activation bypasses the requirement for intact radial spokes and central apparatus complex. However, for wild-type axonemes in 1 mM ATP, calcium-mediated change in waveform requires the presence of the central apparatus . propose that key calcium sensors may be localized to the axoneme in positions other than the radial spokes or central apparatus. Our results support the hypothesis that the calcium sensor is not exclusively localized to the central apparatus. However, our results do not rule out the possible involvement of the central apparatus in regulating the calcium sensor to modulate dynein-driven microtubule sliding during calcium-dependent waveform conversion. What Are the Targets of Calcium-mediated Regulation? | The biggest challenge to understanding the mechanism of flagellar motility is determining how asymmetric regulation of dynein activity is achieved. At any particular moment during flagellar beating, all of the dynein arms are not simultaneously active. The experiments described here do not address the possibility that under specific conditions, the dynein arms on different subsets of doublet microtubules are differentially affected by calcium. The only structure within the flagellum that is asymmetric in both structure and composition is the central apparatus. There are two possibilities to explain how calcium may produce asymmetric regulation of dynein activity to modulate waveform. The key components may be uniformly distributed within the axoneme, and the central apparatus modulates their activity in the presence of calcium. Alternatively, the key components may be asymmetrically organized onto axoneme structures such as the radial spokes or specific doublet microtubules. In either case, it is crucial to determine the location of these components and to identify the dynein arm subforms involved. To determine which dynein arms are the targets of the calmodulin-mediated increase in sliding velocity observed for pf18 axonemes in pCa4 buffer, we are currently constructing double mutants that lack the central apparatus as well as specific dynein arm subforms. Mitchell and Rosenbaum have reported that although the switch from asymmetric to symmetric waveform can be ellicited in outer dyenin armless mutants, the response of outer dyenin armless mutants is abnormal . Unfortunately, double mutants lacking the central apparatus and outer dynein arms have very short flagella. Because of these technical limitations, we have been unable to obtain interpretable results using axonemes isolated from these mutants in our microtubule sliding assay. Therefore, we must develop alternative methods to investigate the involvement of the outer dynein arms in regulating sliding velocity in high calcium conditions. The construction of central pairless mutants lacking subforms of inner dynein arms is underway. We are also investigating whether CaM-KII is a structural component of the axoneme. The Chlamydomonas EST database contains several cDNAs that show a high degree of amino acid identity with subunits of human CaM-KII. However, positive identification of an axonemal CaM-KII will require further biochemical and functional analyses as well as definitive localization. Experiments are underway to identify an axonemal calmodulin-dependent kinase as well as additional calmodulin-binding proteins within the axoneme. Backmatter: PMID- 12221135 TI - Analysis of Sec22p in Endoplasmic Reticulum/Golgi Transport Reveals Cellular Redundancy in SNARE Protein Function AB - Membrane-bound soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins form heteromeric complexes that are required for intracellular membrane fusion and are proposed to encode compartmental specificity. In yeast, the R-SNARE protein Sec22p acts in transport between the endoplasmic reticulum (ER) and Golgi compartments but is not essential for cell growth. Other SNARE proteins that function in association with Sec22p (i.e., Sed5p, Bos1p, and Bet1p) are essential, leading us to question how transport through the early secretory pathway is sustained in the absence of Sec22p. In wild-type strains, we show that Sec22p is directly required for fusion of ER-derived vesicles with Golgi acceptor membranes. In sec22Delta strains, Ykt6p, a related R-SNARE protein that operates in later stages of the secretory pathway, is up-regulated and functionally substitutes for Sec22p. In vivo combination of the sec22Delta mutation with a conditional ykt6-1 allele results in lethality, consistent with a redundant mechanism. Our data indicate that the requirements for specific SNARE proteins in intracellular membrane fusion are less stringent than appreciated and suggest that combinatorial mechanisms using both upstream-targeting elements and SNARE proteins are required to maintain an essential level of compartmental organization. Keywords: INTRODUCTION : In the eukaryotic secretory pathway, a multiplicity of proteins, lipids, and cofactors is required for organized transport. Proper organization is due in part to highly specific homotypic and heterotypic membrane fusion events that depend on a family of proteins termed soluble N-ethylmaleimide-sensitive factor attachment protein receptors, or SNAREs . This family is typified by a conserved heptad repeat sequence or "SNARE motif" adjacent to a membrane-bound segment. Certain sets of SNARE proteins form stable complexes through assembly of their heptad repeat regions into a parallel four-helix coiled-coil structure . A crystal structure of the neuronal SNARE complex consisting of synaptobrevin-II, syntaxin 1A, and SNAP-25B revealed a four-helix bundle held together by 16 layers of largely hydrophobic residues but with an ionic "zero layer" near the center of this bundle . The ionic layer seems to be a conserved feature of many different SNARE complexes and in most instances consists of one arginine residue contributed by a synaptobrevin-like protein or R-SNARE and three glutamines residues contributed from three Q-SNARE helices . The assembly of these four helix bundles with cognate sets of SNARE proteins contributed from opposing membranes has been proposed to catalyze bilayer fusion and to encode compartmental specificity . However, other studies have suggested that SNARE proteins are not the sole determinants of intracellular fusion reactions and upstream targeting or tethering machines may work in concert with SNARE complexes to impart specificity (reviewed by ). For example, a given SNARE protein can assemble into complexes with multiple SNARE partners and function in multiple membrane fusion reactions . Moreover, studies with purified SNARE proteins demonstrate that stable complexes between some noncognate SNARE proteins form promiscuously , although these associations may not reflect a capacity to fuse lipid bilayers. To test the role of SNARE proteins in specifying membrane fusion, a comprehensive study of SNARE proteins from Saccharomyces cerevisiae was undertaken to identify combinations that catalyze bilayer fusion when reconstituted with proteoliposomes bearing purified SNARE proteins. In large part, the compartmental specificity of intracellular membrane fusion was recapitulated with cognate SNARE proteins . In S. cerevisiae, genetic, biochemical, and morphological evidence indicates that the SNARE proteins Sed5p, Bet1p, Bos1p, and Sec22p mediate fusion of endoplasmic reticulum (ER)-derived transport vesicles with an early Golgi compartment . Indeed, membrane fusion reactions reconstituted with purified SNARE proteins in proteoliposomes demonstrated that of 11 SNARE proteins tested, only Bet1p sustained fusion activity when a ternary complex of Sed5p, Bos1p, and Sec22p was present on the opposing membrane . Moreover, this set of SNARE proteins efficiently forms a stable quaternary complex upon mixing of purified components . However, the action of this SNARE complex in cellular membrane fusion has been enigmatic because the SEC22 gene is dispensable for growth . In contrast, the genes encoding other members of this complex (SED5, BET1, and BOS1) are essential. These observations raised the possibility of redundancy in Sec22p function or that certain SNARE activities can somehow be bypassed. In this report, we investigate the role of Sec22p in transport between the ER and Golgi by using a cell-free assay that reconstitutes this stage of transport in yeast . We show that in wild-type cells, Sec22p is required for transport to the Golgi but in its absence, Ykt6p, a protein with high sequence identity to Sec22p, is up-regulated and functionally substitutes for Sec22p. Given this apparent redundancy in SNARE protein function, our results suggest that additional specificity factors operate in concert with SNARE proteins to achieve an essential level of membrane organization. MATERIALS AND METHODS : Strains and Plasmids | Yeast strains CBY740 (MATalpha his3 leu2 lys2 ura3) and CBY773 (MATalpha his3 leu2 lys2 ura3 sec22Delta::KAN) were purchased from Research Genetics (Huntsville, AL) and are isogenic to BY4742 . Strain CBY1108 (MATalpha his3 leu2 lys2 ura3 with pYKT6-2 mum-URA3) contains plasmid pSK60 in CBY740. Wild-type strain FY834 (MATalpha his3 leu2 lys2 ura3 trp1) has been described previously , and CBY1236 (MATa his3 leu2 lys2 ura3 trp1 sec22Delta::KAN) was constructed by backcrossing CBY773 with FY834 multiple times. The ykt6-1 temperature-sensitive strain SARY166 (MATalpha his3 leu2 ura3 trp1 ykt6::LEU (CEN6, TRP1, ykt6-1) and isogenic wild-type SARY189 (MATalpha his3 leu2 ura3 trp1 ykt6Delta::LEU (2 mu, TRP1, YKT6) were from D. Banfield . The plasmid pGEX-2T-SEC22 was constructed by subcloning a 556-base pair polymerase chain reaction fragment carrying SEC22 (1 --180 aa) into the BamHI-EcoRI sites of the pGEX-2T vector (Pharmacia, Peapack, NJ). Strain CBB1136 contains pGEX-2T-SEC22 in XL-1 Blue cells (Stratagene, La Jolla, CA). Yeast strains were grown in either rich medium (1% Bacto-yeast extract, 2% Bacto-peptone, and 2% dextrose) or selective medium (0.67% nitrogen base without amino acids, 2% dextrose) and required supplements. Bacterial strains were grown in LB medium (1% NaCl, 1% peptone, and 0.5% yeast extract) containing 100 mug/ml ampicillin. Antibodies and Immunoblotting | Antibodies directed against alpha-1,6-mannose linkages Ypt1p, Sec61p, Bos1p, Bet1p, Sed5p, Sec23p, Ykt6p, and Sly1p have been described previously . Antibodies against Erv25p and Erv41p were also used. Polyclonal antibodies were raised against a GST-Sec22p (NH2-terminal 1 --180 aa) fusion protein expressed from plasmid pGEX-2T-SEC22. The fusion protein was purified according to the manufacturer's specifications (Pharmacia) and used to immunize rabbits by standard procedures. For affinity purification of anti-Sec22p antibodies, purified GST-Sec22p protein was coupled to Affi-Gel 10 as recommended by the manufacturer (, Hercules, CA). Anti-Sec22p antibodies were bound and eluted from this matrix and then concentrated by centrifugation in a Centricon 30 microconcentrator (Amicon, Beverly, MA). Affinity-purified anti-Ykt6p antibodies were prepared as described previously and preimmune IgGs isolated on protein A-Sepharose . Immunoblots were developed using the enhanced chemiluminescence method (Pharmacia). For densitometric analyses, films were scanned and plotted using NIH Image 1.52. Immunoprecipitations | Native immunoprecipitation of Bos1p from detergent-solubilized membranes was performed as follows. Wild-type and sec22Delta strains were grown at a permissive temperature of 25C. Semi-intact yeast cells were prepared and a 60-mul aliquot was incubated in 180 mul of lysis buffer (25 mM HEPES pH 7.0, 150 mM KOAc, 10 mM EDTA, and 0.5 mM phenylmethylsulfonyl fluoride) with 8 U of apyrase at 25 or 35C for 10 min. Lysed cells were washed with 1 ml of lysis buffer and sedimented by centrifugation at 18,000 x g (14,000 rpm) for 3 min at 4C. Pellets were resuspended with 200 mul of lysis buffer (containing 13 U of apyrase), and an equal volume of lysis buffer containing 2% Triton X-100 was added to solubilize membranes on ice for 10 min. The detergent extract was centrifuged at 100,000 x g (50,000 rpm) at 4C for 10 min in a TL-100 ultracentrifuge (Beckman Coulter, Inc., Fullerton, CA). The supernatant fraction (380 mul) was mixed with 100 mul of buffer A (25 mM HEPES pH 7.0, 100 mM KOAc, and 0.1% Triton X-100) and 30 mul of anti-Bos1p antibodies linked to protein A beads (50% solution) or 75 mul of protein A beads (20%) was added. After a 2-h incubation at 4C with gentle mixing, the beads were washed five times with cold buffer B (25 mM HEPES pH 7.0, 150 mM KOAc, and 0.1% Triton X-100) and bound proteins eluted by heating in 30 mul of 2% SDS at 95C for 1 min. The eluted proteins were diluted in SDS-PAGE sample buffer, resolved on 12.5% polyacrylamide gels, and transferred to nitrocellulose for immunoblot analysis. In Vitro Vesicle Budding, Tethering, and Transport Assays | Yeast semi-intact cells from either wild-type (CBY740) or sec22Delta (CBY773) strains were prepared for in vitro tethering and transport assays as described previously . For Sec22p immunodepletion experiments from COPII vesicles, budded vesicles were prepared and isolated from microsomes with the following modification. Before the addition of COPII proteins, microsomes (0.1 ml of 4 mg/ml) containing [35S]gp-alpha-factor were incubated in the presence or absence of affinity-purified anti-Sec22p antibodies (15 mug/ml) on ice for 15 min. Depleted and wild-type budded vesicles were then isolated from density gradients and added to acceptor membranes . For in vitro assays, data points are the average of duplicate determinations and the error bars represent the range. RESULTS : Sec22p Is Required for Anterograde Transport to Golgi Complex In Vitro | Genetic and biochemical experiments suggested a requirement for Sec22p in anterograde transport from the ER to the Golgi by forming a SNARE complex with Bos1p, Bet1p, and Sed5p . Further studies showed that Sec22p also functions in retrograde transport from the Golgi to the ER and acts with the ER-localized SNARE protein Ufe1p . When examined in cell-free assays that reproduce anterograde and retrograde transport , the thermosensitive sec22-3 allele inhibited retrograde and not anterograde transport at restrictive temperatures. These findings suggest that Sec22p does not act directly in anterograde transport although other explanations, such as allele specificity, are possible. Therefore, as an independent test of Sec22p function in anterograde transport, we prepared affinity-purified anti-Sec22p antibodies to neutralize Sec22p function in a reconstituted cell-free assay that measures transport to the Golgi. In this assay, washed semi-intact cell membranes containing [35S]glycopro-alpha-factor (gpalphaf) are incubated with purified transport factors (COPII, Uso1p, and LMA1) to drive transport of [35S]gpalphaf to the Golgi . On delivery to the Golgi complex, gpalphaf receives outer-chain alpha1,6-mannose residues that can be immunoprecipitated with alpha1,6-mannose --specific antiserum to quantify [35S]gpalphaf transport . As seen in Figure A, reconstituted transport was sensitive to anti-Sec22p antibody, whereas preimmune IgGs at comparable concentrations do not inhibit transport. Cell-free transport can be divided into subreactions, each following movement of [35S]gpalphaf . Incubation of washed semi-intact cells with the purified COPII proteins generates freely diffusible vesicles containing gpalphaf that can be separated from larger membranes by differential centrifugation. Using these assays, we found that the inhibitory Sec22p antibodies specifically blocked the vesicle fusion stage of the reaction because COPII-dependent budding and Uso1p-dependent tethering were unaffected (Figure , B and C). The fact that anti-Sec22p antibodies had no effect on vesicle budding and tethering excluded the possibility that inhibition was due to aggregation of membranes. Taken together, these results indicated that Sec22p was directly required for anterograde transport of [35S]gpalphaf to the Golgi complex. Figure 1 | Anti-Sec22p antibodies inhibit transport to the Golgi complex in vitro. Anti-Sec22p antibodies inhibit transport to the Golgi complex in vitro. (A) Wild-type (CBY740) semi-intact cells containing [35S]gp-alpha-factor in the ER were incubated with COPII proteins, Uso1p, LMA1, and an ATP regeneration system. After 80 min at 23C, the amount of Golgi-modified [35S]gp-alpha-factor was measured to determine transport efficiency. Varying amounts of affinity-purified anti-Sec22p antibodies or preimmune IgGs were added into cell-free transport reactions. In this experiment, background transport (semi-intact cells with an ATP regeneration system) was 1.3% and reconstituted transport in the absence of antibodies was 12.1%. (B) Semi-intact cells prepared as in A were incubated with COPII or COPII plus Uso1p to measure budding and tethering in the presence or absence of anti-Sec22p (15 mug/ml). After 30 min at 23C, freely diffusible vesicles containing [35S]gp-alpha-factor were separated from semi-intact membranes by centrifugation at 18,000 x g. (C) A parallel transport reaction for the experiment shown in B to demonstrate effective inhibition of fusion with anti-Sec22p. Unlike the other ER/Golgi SNAREs required for anterograde transport (i.e., BET1, BOS1, and SED5), the SEC22 gene is not essential although the sec22Delta allele results in slowed growth and temperature sensitivity . Given our in vitro findings, these observations raise an interesting paradox. If Sec22p is required for an essential transport step, how do cells survive in its absence? We speculated that the Sec22p-dependent step was somehow bypassed or that a redundant activity substituted for Sec22p function. To address these possibilities, we first investigated anterograde transport in sec22Delta cells by using the reconstituted cell-free assay. As seen in Figure A, similar40% of the [35S]gpalphaf was budded into diffusible vesicles in a wild-type strain when COPII proteins were added. Addition of Uso1p-tethered COPII vesicles to acceptor membranes resulted in an similar50% reduction of diffusible vesicles (Figure A, open bars). Efficient fusion of tethered vesicles required the addition of LMA1 and yielded similar18% transport of [35S]gpalphaf to the Golgi complex in wild-type membranes (Figure B). In sec22Delta semi-intact cell membranes, reconstituted budding and transport efficiencies were significantly reduced, whereas Uso1p-dependent tethering remained similar50% efficient . It is not entirely clear why budding and fusion are compromised in sec22Delta cells, although it is known that this deletion causes activation of the unfolded protein response . Regardless, these results indicated that a COPII- and Uso1p-dependent transport pathway was operational in the complete absence of Sec22p. Therefore, we consider it unlikely that the normal anterograde transport pathway was bypassed in sec22Delta cells. Figure 2 | Influence of sec22Delta on steps in ER-Golgi transport. Influence of sec22Delta on steps in ER-Golgi transport. Washed semi-intact cells containing [35S]gp-alpha-factor were prepared from wild-type (CBY740) and sec22Delta (CBY773) strains. (A) Vesicle budding and tethering in reactions that contained an ATP regeneration system alone (no addition, hatched bars), plus COPII (solid bars) or plus COPII and Uso1p (open bars). (B) Transport assays contained an ATP regeneration system alone (no addition, hatched bars) or were supplemented with COPII, Uso1p, and LMA1 (reconstituted, solid bars). Ykt6p Assembles into Specific ER/Golgi SNARE Complex in sec22Delta Strains | We next investigated whether a redundant activity was substituting for Sec22p function. Of the 21 predicted SNARE proteins in S. cerevisiae, Ykt6p shares the highest degree of amino acid identity (28%) with Sec22p. Furthermore, the core sequences of Ykt6p and Sec22p share an even greater degree of amino acid identity (40%) with the zero layer arginine residue present in both proteins. Thus, substitution of Sec22p with Ykt6p in a tetrameric SNARE complex consisting of Sed5p, Bet1p, Bos1p, and Ykt6p would preserve a 3Q:1R ratio. Ykt6p has been reported to act in multiple trafficking steps in yeast, including retrograde transport to the cis-Golgi , homotypic vacuole fusion , and anterograde transport from the Golgi complex to the vacuole . Although Ykt6p has not been directly implicated in anterograde transport from the ER to the Golgi complex in yeast, the temperature-sensitive sec22-1 allele is suppressed by multicopy YKT6 . To investigate the possibility that Ykt6p functionally replaced Sec22p, we first examined the expression level of the ER/Golgi SNAREs in whole cell membranes . Sed5p, Sly1p, and Bos1p (our unpublished data) expression levels were unchanged in the sec22Delta strain, whereas Ykt6p expression was increased 3.4-fold. Furthermore, Ykt6p remained membrane bound in the sec22Delta strain, indicating efficient posttranslational prenylation of the overexpressed protein. Other proteins involved in budding (Sec23p) and tethering (Ypt1p) of ER-derived transport vesicles were not elevated in the sec22Delta strain. Interestingly, a 1.6-fold increase in the ER-translocon protein Sec61p was observed in the sec22Delta strain and was probably due to activation of the unfolded protein response (UPR) caused by this deletion . However, it should be noted that the expression level of Ykt6p was not induced simply by activating the UPR with dithiothreitol (our unpublished data), a result that is in accord with microarray analysis of UPR-induced messages . In summary, deletion of SEC22 increases the expression level of Ykt6p, the yeast SNARE protein that shares highest identity to Sec22p. Figure 3 | Deletion of SEC22 increases Ykt6p expression level. Deletion of SEC22 increases Ykt6p expression level. Immunoblot to monitor expression levels and subcellular distributions of proteins in WT (CBY740) and sec22Delta (CBY773) cells. Washed semi-intact cells were incubated in buffer (20 mM HEPES pH 7.0, 150 mM KOAc, and 5 mM MgOAc) and then fractionated by centrifugation at 100,000 x g for 10 min to generate supernatant (Sup) and pellet fractions. The total lysate and aliquots of each fraction were resolved by SDS-PAGE, followed by immunoblot for Sec23p, Sec61p, Sly1p, Ypt1p, Sed5p, Ykt6p, and Sec22p. Previous reports suggested that Ykt6p functions on Golgi membranes and later compartments of the secretory pathway, including the vacuole . If Ykt6p was substituting for Sec22p function in transport between the ER and Golgi, we expected that a fraction of overexpressed Ykt6p would be found on ER-derived transport vesicles. To explore this possibility, we compared COPII vesicles isolated from wild-type and sec22Delta strains. Vesicle budding was reconstituted from ER membranes by incubating purified COPII proteins with washed membranes . Membranes lacking Sec22p produced COPII-coated vesicles, albeit less efficiently than wild-type, as evidenced by budding of the vesicle marker protein Erv25p . This result was also in accord with decreased [35S]gpalphaf budding observed in Figure . Although budding was less efficient in sec22Delta membranes, COPII vesicles from these membranes contained an elevated level of Ykt6p compared with other vesicle marker proteins. We conclude that overexpressed Ykt6p was contained on COPII vesicles and therefore in a location to participate in this stage of transport. Figure 4 | Packaging of Ykt6p into COPII-coated vesicles. Packaging of Ykt6p into COPII-coated vesicles. COPII budding reactions with semi-intact cells from WT (CBY740) and sec22Delta (CBY773) strains. One-tenth of a total reaction (T), budded vesicles isolated after incubation with COPII proteins (+), or a mock reaction without COPII proteins (-). Samples were resolved by SDS-PAGE, followed by immunoblot for Sec61p (ER resident protein), Erv25p (vesicle protein), and Ykt6p. We next tested whether the overexpressed Ykt6p detected in sec22Delta strains was associated with other SNARE proteins that operate in transport between the ER and Golgi complex. Previous reports indicated that some Ykt6p coimmunoprecipitated with Sed5p from detergent-solubilized membranes . We immunoprecipitated Bos1p from wild-type and sec22Delta-solubilized membranes and monitored the amount of Ykt6p and other proteins that coprecipitated (Figure A). Membranes were preincubated for a brief period at 25 or 35C to monitor the relative stability of these associations. After preincubation at 25C, equivalent amounts of Sed5p, Sly1p, and Bet1p (our unpublished data) coimmunoprecipitated with Bos1p in wild-type and sec22Delta strains; however, the amount of Ykt6p associated with Bos1p immmunoprecipitates was increased 2.7-fold in sec22Delta membranes. Erv41p, an integral membrane protein that localizes to ER/Golgi membranes , was not efficiently immunoprecipitated and served as a negative control for these experiments. If membranes were preincubated at 35C before Bos1p immunoprecipitation, the level of Sed5p and Sly1p was unchanged; however, the amount of bound Ykt6p was decreased 1.5-fold in sec22Delta membranes compared with wild type. Comparable levels of Bos1p were recovered from both membrane preparations at 25 or 35C. To summarize, more Ykt6p was associated with Bos1p in strains lacking Sec22p and this association was thermosensitive. The observed instability of this Bos1p complex in sec22Delta membranes may underlie the temperature sensitivity exhibited by sec22Delta strains. Figure 5 | Incorporation of Ykt6p into ER-Golgi SNARE complexes. Incorporation of Ykt6p into ER-Golgi SNARE complexes. Solubilized proteins were bound to anti-Bos1p coupled to protein A beads (+) or beads alone (-) as described in MATERIAL AND METHODS. (A) Semi-intact cells from WT (CBY740) and sec22Delta (CBY773) strains were preincubated at 25 or 35C and then placed on ice before solubilization with Triton X-100. Total solubilized extracts (T) and immunoprecipitates (IP) were resolved on SDS-PAGE followed by immunoblot for indicated proteins. (B) Semi-intact cells from WT (CBY740), WT+pYKT6-2 mum (CBY1108), and sec22Delta (CBY773) strains were preincubated at 25C, solubilized, and immunoprecipitated with anti-Bos1p --coupled protein A beads as described above. The association of Ykt6p with Bos1p was also examined in a wild-type strain that overproduced Ykt6p (Figure B). We were concerned that association of Ykt6p with Bos1p was a nonspecific consequence of Ykt6p overexpression in the sec22Delta strain. However, if Sec22p and Ykt6p competed for a specific association with a Sed5p-Bet1p-Bos1p SNARE complex, one would expect that YKT6 overexpression in a wild-type strain would yield less Ykt6p in association with Bos1p than in a sec22Delta strain. Indeed, a threefold overproduction of Ykt6p in the presence of normal levels of Sec22p resulted in a modest increase in Ykt6p that was coimmunoprecipitated with Bos1p. This level was significantly less than the amount of Ykt6p associated with Bos1p in sec22Delta membranes (Figure B, compare Ykt6p in lanes 3, 6 and 9). These observations indicate that in the absence of Sec22p, Ykt6p protein levels are increased and Ykt6p assembles into a specific SNARE complex with Bos1p. Ykt6p Functionally Substitutes for Sec22p | If Ykt6p functionally substitutes for Sec22p in sec22Delta strains, we hypothesized that strains lacking Sec22p would be sensitive to inhibitors of Ykt6p activity in the ER/Golgi cell-free transport assay. To test this idea, we used the reconstituted transport assay described in Figure and selectively neutralized Sec22p or Ykt6p activity with affinity-purified antibodies directed against these proteins. Addition of anti-Sec22p antibodies to wild-type reactions inhibited transport by 90%, whereas addition of an identical dose of antibodies to sec22Delta reactions reduced transport by 10% (Figure A). In contrast, wild-type transport reactions were insensitive to anti-Ykt6p antibodies but sec22Delta membranes were strongly inhibited by this addition (Figure B). Moreover, anti-Ykt6p antibodies did not inhibit COPII vesicle budding in sec22Delta membranes (our unpublished observation), indicating a specific block in the fusion stage. These findings demonstrate that Ykt6p is functionally required for anterograde transport to the Golgi complex in sec22Delta strains. Figure 6 | Selective inhibition of ER-Golgi transport by anti-Ykt6p antibodies. Selective inhibition of ER-Golgi transport by anti-Ykt6p antibodies. Cell free transport assays in wild-type (CBY740) and sec22Delta (CBY773) semi-intact cells. (A) Assays contained an ATP regeneration system alone (no addition, hatched bars), plus reconstitution proteins (solid bars) or reconstitution proteins and anti-Sec22p. (B) As in A except anti-Ykt6p (75 mug/ml) antibodies were used instead of anti-Sec22p antibodies. In these experiments, maximal transport of [35S]gp-alpha-factor was similar15% for the wild-type and similar5% for the sec22Delta semi-intact cells. Efficient Transport Requires Sec22p Activity on ER-derived Vesicles or Acceptor Membranes | Having established that Sec22p was directly required for anterograde transport, we next investigated whether this activity localized to vesicle or acceptor membranes. Our previous experiments indicated that Bet1p and Bos1p were functionally required on vesicles, whereas Sed5p acted on the acceptor membrane fraction . To determine sites of action, we isolated COPII vesicles containing [35S]gpalphaf from wild-type or sec22Delta membranes and added equal amounts of each to wild-type or sec22Delta acceptor membranes. Vesicles or acceptor membranes lacking Sec22p fused efficiently when mixed with their wild-type counterpart (Figure A). Only when Sec22p was absent from both vesicles and acceptor membranes was transport significantly reduced. This level corresponded to that observed when overall transport was reconstituted in sec22Delta membranes (Figure B). Presumably, this transport level is sustained by Ykt6p substitution. These results indicate that either the vesicle or the acceptor membrane fraction can provide Sec22p activity. Figure 7 | Sec22p acts on vesicle or acceptor membranes. Sec22p acts on vesicle or acceptor membranes. (A) COPII vesicles containing [35S]gp-alpha-factor were synthesized from wild-type (WT) or sec22Delta (22Delta) membranes and mixed with wild-type or sec22Delta semi-intact cell acceptor membranes in transport assays. No addition indicates ATP regeneration system alone, and reconstituted indicates addition of Uso1 and LMA1. (B) Immunoblot of floated ER-derived vesicles that had been depleted of Sec22p by addition of anti-Sec22p during COPII budding. (C) Incubation of wild-type or Sec22p-depleted vesicles with wild-type acceptor membranes in transport assays. Addition of anti-Sec22p antibody inhibited fusion efficiency in wild-type and depleted vesicles. Because the sec22Delta mutation reduced the efficiency of vesicle budding and fusion assays, we sought a second line of experimentation to confirm the localized requirements for Sec22p by using wild-type membranes. Previous reports have shown that antibodies directed against specific ER-vesicle proteins can inhibit their incorporation into these vesicles when added during vesicle-budding reactions . Presumably, antibody-bound proteins are not recognized by the COPII-budding machinery and therefore are not efficiently packaged into transport vesicles. Therefore, we sought to deplete Sec22p from ER-derived vesicles by adding affinity-purified anti-Sec22p antibodies to a COPII-budding reaction. As seen in Figure B, inclusion of anti-Sec22p inhibited Sec22p packaging into COPII-synthesized vesicles but did not inhibit overall vesicle budding because other vesicle proteins, including Bos1p, Sed5p (our unpublished data), Erv25p, and Erv46p, were packaged efficiently in the presence of this antibody. We then purified wild-type and Sec22p-depleted vesicles on density gradients and measured their capacity to fuse with wild-type acceptor membranes (Figure C). No anti-Sec22p antibodies were detected on vesicles after gradient-purification (our unpublished observation); therefore, any influence on fusion efficiency can be attributed to depletion and not carryover of antibody. In these experiments, we observed that fusion of purified wild-type vesicles remained sensitive to anti-Sec22p antibodies, indicating a direct role for Sec22p in this fusion stage of anterograde transport. We also found that Sec22p depletion from vesicles reduced their fusion efficiency (similar2-fold) but some fusion activity remained. This residual vesicle fusion activity relied on Sec22p because addition of anti-Sec22p antibody inhibited this fusion signal to near background levels. Therefore, a >90% reduction in Sec22p from transport vesicles caused only a 50% reduction in fusion efficiency. Together with the sec22Delta experiments, these results suggest that optimal fusion efficiency requires Sec22p on vesicles and acceptor membranes, but fusion can proceed if activity is present on either membrane. These findings are similar to those reported for Nyv1p-dependent fusion of vacuoles . In this situation, deletion of Nyv1p (an R-SNARE) from one vacuole reduced but did not block fusion, whereas deletion from both vacuoles blocked membrane fusion. Genetic Experiments Reveal a Synthetic Lethal Relationship between sec22Delta and ykt6-1 | If Ykt6p substitutes for Sec22p in vivo, we hypothesized that a crippled version of Ykt6p may not fulfill this requirement. YKT6 is an essential gene ; however, a previous report described a temperature-sensitive ykt6-1 allele that inhibits intra-Golgi and/or post-Golgi transport when incubated at restrictive temperatures . We tested the genetic relationship between the sec22Delta and ykt6-1 alleles by crossing strain CBY1236 (sec22Delta) with SARY166 (ykt6Delta, pykt6-1) and SARY189 (ykt6Delta, pYKT6). After sporulation and tetrad dissection of the sec22Delta X ykt6Delta, pykt6-1 cross, no haploid strains containing the sec22Delta, ykt6Delta and pykt6-1 alleles were recovered (Figure A). In similar analyses of sec22Delta X ykt6Delta, pYKT6 tetrads, several spores containing the sec22Delta, ykt6Delta and pYKT6 alleles were recovered (Figure B). Based on these results, we conclude that sec22Delta strains cannot survive if YKT6 function is compromised. These in vivo results corroborate our in vitro findings indicating that wild-type Ykt6p can substitute for Sec22p in fusion of ER-derived transport vesicles. Figure 8 | Synthetic lethal interaction between ykt6-1 and sec22Delta.. Synthetic lethal interaction between ykt6-1 and sec22Delta.. (A) A diploid strain generated from the cross of CBY1236 (sec22Delta) and SARY166 (ykt6Delta, pykt6-1) was sporulated, dissected, and incubated on YPD plates at room temperature for 5 d. Seven (from a total of 14) representative tetrads are shown. No strains containing sec22Delta ykt6Delta and pykt6-1 were recovered. (B) As in A, except CBY1236 (sec22Delta) was crossed with SARY189 (ykt6Delta, pYKT6) and sporulated. Seven (from a total of 14) representative tetrads are shown. Circled colonies indicate strains in which the sec22Delta ykt6Delta and pYKT6 alleles cosegregated. Some inviability results from a failure to segregate the pYKT6 plasmid into ykt6Delta spores. DISCUSSION : SNAREs in ER/Golgi Transport | In this report, we investigated the mechanisms by which yeast cells lacking the Sec22p R-SNARE protein can maintain anterograde transport between the ER and Golgi complex. Our experiments demonstrated that Ykt6p, the R-SNARE most related to Sec22p, was up-regulated and formed a specific SNARE complex with Bos1p and Sed5p when Sec22p was absent. Under this condition, Ykt6p was also efficiently packaged into ER-derived transport vesicles and was required for fusion of these vesicles with acceptor Golgi membranes. Although Ykt6p can substitute for Sec22p activity, replacement was not optimal because cell growth rates are reduced and in vitro transport efficiency was decreased . When Ykt6p function was further compromised in a sec22Delta background, cell viability was lost. Based on these findings, we conclude that Ykt6p provides a redundant activity for Sec22p. Our findings answer a long-standing question concerning the viability of sec22Delta strains and their resulting phenotypes . A role for Sec22p in retrograde transport from the Golgi complex to the ER had been suggested and demonstrated , but a direct requirement for Sec22p activity in anterograde transport to the Golgi has not been reported . Perhaps the conditional sec22-3 allele used in these experiments selectively inhibits retrograde and not anterograde transport. Regardless, studies now indicate direct requirements for Sed5p, Bet1p, Bos1p, and Sec22p in anterograde traffic to the Golgi complex . We propose that a SNARE complex formed from Sed5p, Bet1p, Bos1p, and Sec22p catalyzes membrane fusion in accord with studies showing formation of a stable quaternary complex between these proteins and a capacity for this subset of SNAREs to fuse proteoliposomes . In vitro data suggest that Bos1p and Bet1p act on ER-derived vesicles, whereas Sed5p acts on acceptor membranes . In this report, we show that Sec22p acts on either vesicles or acceptor membranes. These findings contrast the minimal fusion assay where only a combination of Sed5p, Bos1p, and Sec22p in one bilayer fused with partner liposomes containing Bet1p . It remains to be determined how SNARE regulatory proteins such as Sly1p may influence these topological requirements. Last, we hypothesize that the Sed5p-Bet1p-Bos1p-Sec22p complex acts in fusion of ER-derived membranes with Golgi acceptor membranes that house outer-chain oligosaccharide modification activities; however, these SNAREs may also act in homotypic fusion of ER-derived vesicles in a step that precedes heterotypic fusion . If Sec22p is required for retrograde transport from the Golgi to ER , does Ykt6p also substitute in the retrograde pathway when Sec22p is absent? We speculate that Ykt6p satisfies this requirement as well. Other characterized proteins that operate in retrograde traffic to the ER are essential ; therefore, it seems unlikely that a parallel pathway operates in the absence of Sec22p. Rather, Ykt6p may substitute for Sec22p yielding a SNARE complex consisting of Ufe1p, Bos1p, Bet1p, and Ykt6p that catalyzes retrograde fusion in sec22Delta strains. Given that Ufe1p is largely ER localized and Sed5p is Golgi localized, it is not at all clear how anterograde and retrograde vesicles are distinct with respect to their SNARE machinery. Perhaps upstream-tethering components that probably include Uso1p, Ypt1p, and TRAPP for anterograde movement, and Sec20p, Tip20, and Dsl1p for retrograde transport, could decipher features on these distinct carrier vesicles. Substantial progress has been made in characterizing SNARE proteins that mediate transport through the early secretory pathway in mammalian cells. The mammalian homologs of Sed5p (syntaxin 5), Bet1p (rbet), Bos1p (membrin), Sec22p (Sec22b), and Ykt6p (Ykt6) have been functionally implicated in transport between the ER and Golgi complex . Interestingly, antibodies against mammalian Ykt6p inhibited a late stage of ER-Golgi transport of vesicular stomatitis virus-G protein protein in vitro , in contrast to our observation in yeast. However, it may be difficult to draw direct parallels between yeast and mammals because there seem to be multiple isoforms of ER/Golgi SNARE proteins that localize to distinct compartments in mammalian cells, and it seems that the organization of the early secretory pathway across species is distinct . Specificity of SNAREs | Other studies in yeast have suggested cellular redundancy in SNARE protein functions through genetic experiments although in these instances the data could be explained by substitution or by activation of parallel processes. Indeed, the situation is complicated because a single SNARE protein can operate in multiple trafficking pathways , and transport between some membranes can use multiple routes . Importantly, the findings in this report demonstrate that a single SNARE protein that normally operates in other trafficking steps can be conscripted to act in another. This apparent flexibility in SNARE protein requirements seems inconsistent with a role in specifying fusion partners . Biochemical studies indicate significant promiscuity in SNARE complex assembly when purified cognate and noncognate SNARE proteins are mixed in solution . In contrast, reconstituted liposome fusion assays suggest that cognate SNARE complexes are largely required to drive bilayer fusion. For example, proteoliposomes containing a Sed5p-Bos1p-Sec22p complex fused specifically with partner liposomes containing Bet1p but not with 10 other SNARE proteins tested . Interestingly, Ykt6p was not able to substitute in this assay when modified with a lipid anchor. When a transmembrane domain was fused to Ykt6p, this integral membrane species promoted fusion with a plasma membrane SNARE complex consisting of Sso1p-Sec9p. The transmembrane-anchored form of Ykt6p was apparently not tested in combinations with Sed5p, Bos1p, and Bet1p. It seems probable that Ykt6p would at least partially substitute for Sec22p in the reconstituted fusion assay although additional SNARE regulatory factors may be required to recapitulate this reaction. Ykt6p may be well suited for promiscuous behavior because it is a lipid-anchored protein that is partially soluble and displays a broad intracellular distribution . In fact, Ykt6p substitution for other related R-SNAREs could explain the nonlethal phenotypes associated with strains lacking Snc1p/Snc2p or Nyv1p . The Snc1p/Snc2p R-SNAREs operate in fusion at the cell surface, and when both are deleted, fusion of exocytic vesicles is reduced but cells remain viable. The other Q-SNAREs that operate in this stage of transport, Sso1p/Sso2p and Sec9p, are essential. Therefore these properties are reminiscent of the ER/Golgi situation because the R-SNARE (Sec22p) is nonessential and the other Q-SNAREs (Sed5p, Bos1p, and Bet1p) are essential. In homotypic vacuole fusion, the R-SNARE proteins Nyv1p and/or Ykt6p are thought to act with the Q-SNAREs Vam3p, Vam7p, and Vti1p . The phenotypes associated with nyv1Delta are mild compared with the vacuolar fragmentation patterns displayed by deletion of the associated Q-SNAREs . Again, Ykt6p may substitute for Nyv1p in this fusion pathway, a proposal that is supported by in vitro studies showing Ykt6p competes with Nyv1p for binding to a ternary SNARE complex consisting of Vam3p, Vam7p, and Vti1p . Is the situation with Sec22p in yeast an isolated case or could SNARE substitution be more widespread in nature? Given that deletion of the R-SNARE synaptobrevin in flies , worms , and mice does not block fusion, redundancy seems a probable explanation. Closer examination of the synaptrobrevin knockout mice by electrophysiology reveals that spontaneous synaptic vesicle fusion was decreased 10-fold in the neural synapse . Indeed, one explanation given for the reduced fusion efficiency was that a noncognate SNARE that did not normally function in synaptic vesicle fusion could partially substitute for synaptobrevin in the mutant neurons . Given our current findings, it may be informative to test whether the mammalian version of Ykt6p is expressed and can substitute for synaptobrevin at the neural synapse. If there are SNARE proteins that can operate in many steps, and substitute for one another, how is compartmental organization maintained? Perhaps some inappropriate fusion can be tolerated although within the limits of detection these events seem minor. Alternatively, membrane fusion reactions could be highly selective. The collective data on SNARE proteins now suggest they provide some selectivity but are unlikely to be the sole determinants of specificity. Previous studies with Rab GTPase chimeras indicated that they too are unlikely to provide a needed level of specificity . Therefore, a more likely explanation is that combinatorial mechanisms that use upstream targeting elements and SNARE proteins are required to maintain compartmental identity. Backmatter: PMID- 12221136 TI - Proteasome Regulates the Delivery of LDL Receptor-related Protein into the Degradation Pathway AB - The low-density lipoprotein receptor (LDLR)-related protein (LRP) is a multiligand endocytic receptor that has broad cellular and physiological functions. Previous studies have shown that both tyrosine-based and di-leucine motifs within the LRP cytoplasmic tail are responsible for mediating its rapid endocytosis. Little is known, however, about the mechanism by which LRP is targeted for degradation. By examining both endogenous full-length and a minireceptor form of LRP, we found that proteasomal inhibitors, MG132 and lactacystin, prolong the cellular half-life of LRP. The presence of proteasomal inhibitors also significantly increased the level of LRP at the cell surface, suggesting that the delivery of LRP to the degradation pathway was blocked at a compartment from which recycling of the receptor to the cell surface still occurred. Immunoelectron microscopy analyses demonstrated a proteasomal inhibitor-dependent reduction in LRP minireceptor within both limiting membrane and internal vesicles of the multivesicular bodies, which are compartments that lead to receptor degradation. In contrast to the growth hormone receptor, we found that the initial endocytosis of LRP minireceptor does not require a functional ubiquitin --proteasome system. Finally, using truncated cytoplasmic mutants of LRP minireceptors, we found that a region of 19 amino acids within the LRP tail is required for proteasomal regulation. Taken together our results provide strong evidence that the cellular turnover of a cargo receptor, i.e., LRP, is regulated by the proteasomal system, suggesting a broader function of the proteasome in regulating the trafficking of receptors into the degradation pathway. Keywords: INTRODUCTION : The low-density lipoprotein receptor (LDLR)-related protein (LRP) is a large endocytic receptor that belongs to the emerging LDLR family . LRP is a unique receptor in the family because of its ability to bind and endocytose a variety of structurally and functionally distinct ligands and its important role during embryonic development and the pathogenesis of various diseases (reviewed in ; ; ; ). Ligands of LRP include proteins that are involved in lipid metabolism, proteinase regulation, blood coagulation/fibrinolysis cascades, and several membrane proteins, including urokinase plasminogen activator receptor (uPAR) and beta-amyloid precursor protein (APP). LRP's extracellular domain contains 31 ligand-binding repeats grouped into four clusters of 2, 8, 10, and 11 repeats, respectively . However, only the second and the fourth clusters have been shown to mediate ligand binding . The extracellular domain of LRP precedes a single membrane-spanning segment, which is followed by a 100-amino acid cytoplasmic tail. Biochemical studies have shown that LRP is synthesized as a single polypeptide chain of similar600 kDa that is cleaved by furin in the trans-Golgi compartments into two subunits of 515 and 85 kDa, which remain associated noncovalently with one another as they mature to the cell surface . Another significant feature of LRP is its rapid endocytosis when compared with that of other members of the LDLR family. Although two copies of the NPxY motif, which has been shown to mediate LDLR endocytosis , are present within the LRP tail, recent studies from our laboratory have shown that the primary endocytosis signals for LRP are a YxxL motif and a di-leucine motif . Our studies have also shown that the initial endocytosis of LRP is further regulated by cyclic AMP --dependent protein kinase A (PKA) phosphorylation on a serine residue within LRP's cytoplasmic tail . These studies together indicate that the endocytic trafficking of LRP is unique compared with other LDLR family members. Previous studies on LRP trafficking have focused on the early secretory pathway and the early events of endocytosis . However, little is known regarding the trafficking of LRP to the degradation pathway. The ubiquitin --proteasome system plays an important role in mediating both receptor endocytosis and sorting to the degradation pathway for several cell surface receptors . Although polyubiquitination of proteins leads to their degradation via the 26S proteasome , monoubiquitination has recently been shown to modulate the function and trafficking of various cellular proteins . To date the most definitively studied examples of the involvement of the ubiquitin --proteasome regulation of cell surface receptor include the yeast G protein-coupled receptors , the growth hormone receptor (GHR; ; ), the human epidermal growth factor receptor (EGFR; ; ; ) and the mammalian beta2-adrenergic receptor . In case of the GHR, it was found that although ubiquitination of GHR itself is not required, functional ubiquitination and proteasomal systems are both essential for its endocytosis and trafficking to the degradation pathway . Despite recent evidence for the engagement of LRP in several signal transduction pathways , the multiple ligands for and rapid endocytosis rate of LRP suggest that a major function of this receptor is the cellular transport of ligands via receptor-mediated endocytosis. To examine whether the transport of such a signal transducing/cargo receptor is regulated by the ubiquitin --proteasome system, we analyzed the effects of proteasomal inhibitors on the endocytic trafficking and cellular turnover of LRP. We found that the delivery of LRP to the degradation pathway is blocked within a compartment from which recycling of the receptor still occurs. MATERIALS AND METHODS : Materials | Recombinant human RAP was produced as a glutathione S-transferase fusion protein and purified as described before . All tissue culture media, sera, and plastic supplies were from Life Technologies (Grand Island, NY). Na125I was purchased from NEN Life Sciences (Boston, MA). [35S]Cysteine was from ICN (Costa Mesa, CA). IODO-GEN reagent was from Pierce (Rockford, IL). MG132 (Z-Leu-leu-leu-H aldehyde) proteasomal inhibitor was from Peptide Institute, Inc (Minosh-shi Osaka, Japan). Clasto-lactacystin-beta-lactone (referred to as lactacystin in this manuscript) proteasomal inhibitor was from Calbiochem-Novabiochem (La Jolla, CA). Complete protease inhibitor cocktail (Complete) was from Roche (Indianapolis, IN). Mouse anti-HA antibody was from Babco (12CA5; Richmond, CA). FITC-conjugated goat anti-mouse antibody was from BD Biosciences-PharMingen (San Diego, CA). Horseradish peroxidase --linked anti-mouse antibody, ECL detection system, and Rainbow molecular size markers were from Amersham Pharmacia Biotech (Piscataway, NJ). All other chemicals were reagent grade and from Sigma (St. Louis, MO). Cell Culture | Human hepatoma HepG2 cells were cultured in minimum essential medium (MEM) supplemented with 10% fetal bovine serum, 100 U/ml penicillin, 100 mug/ml streptomycin, and 300 mug/ml l-glutamine. Chinese hamster ovary (CHO)-K1 and CHO LRP-null cells were cultured in F-12 Ham's media with 10% fetal bovine serum, 100 U/ml penicillin, 100 mug/ml streptomycin, and 300 mug/ml l-glutamine. Ts20 cells were grown in MEMalpha supplemented with 10% fetal calf serum, 4.5 g/l glucose, 100 U/ml penicillin, and 100 mug/ml streptomycin. CHO-mLRP4, ts20-mLRP4, and ts20-GHR cells were cultured under the same conditions as their parent CHO cells, except with the addition of 350 mug/ml (for CHO-mLRP4 cells) or 450 mug/ml (for ts20-mLRP4 and ts20-GHR cells) geneticin. Ts20 cells were maintained at 30C; all others were maintained at 37C in a humidified incubator with 5% CO2. Metabolic Pulse-chase Labeling and Immunoprecipitation | HepG2, CHO-K1, or CHO-mLRP4 cells were seeded at 5 x 105 cells/well density in six-well plates and cultured overnight before experiments. HepG2 cells transiently transfected with mLRP4 by calcium phosphate method were replated after overnight transfection and cultured for an additional 24 h before experiments. Metabolic pulse-chase labeling of LRP or mLRP4 with [35S]cysteine was performed essentially as described before . Briefly, after depletion of endogenous cysteine, cells were pulse-labeled with [35S]cysteine (200 muCi/ml) in cysteine-free MEM and chased with serum-containing medium in the presence or absence of 20 muM MG132 for various times as indicated in each experiment. Cells were then lysed at 4C in 0.5 ml PBSc (phosphate-buffered saline supplemented with 1 mM CaCl2 and 0.5 mM MgCl2) containing 1% Triton X-100, 1 mM PMSF, 1x Complete, and 10 mM N-ethymaleimide (NEM; lysis buffer). Cell lysates were incubated with an excess of either anti-LRP antibody (for endogenous LRP with HepG2 cells) or monoclonal anti-HA antibody (for mLRP4 with CHO-mLRP4 cells), followed with protein-A agarose beads. Immunoprecipitated protein was released from the beads by boiling in Laemmli sample buffer under reducing conditions and analyzed via SDS-PAGE. Band intensities were quantitated using a phosphoimager (Storm 840; Molecular Dynamics, Sunnyvale, CA). Western Blotting of LRP Minireceptor | CHO-mLRP4 cells or CHO LRP-null cells stably transfected with mLRP4 tail truncation mutants were seeded at 1.5 x 106 cells per 60-mm dish and cultured overnight before experiments. Cells were incubated in the presence or absence of proteasome inhibitors (20 muM MG132 or lactacystin) at 37C as indicated in each experiment. Cell monolayers were then washed twice in prechilled PBSc and lysed with 0.5 ml of lysis buffer. Equal quantities of protein were subjected to SDS-PAGE (6%) under reducing conditions. After transfer to PVDF membrane, successive incubations with anti-HA antibody and horseradish peroxidase --conjugated goat anti-mouse IgG were carried out for 60 min at room temperature. The immunoreactive proteins were then detected using the ECL system. Films showing immunoreactive bands were scanned by Kodak Digital Science DC120 Zoom Digital Camera and analyzed with Kodak Digital Science1D Image Analysis Software (Eastman-Kodak, Rochester, NY). Flow Cytometric Analysis of Cell Surface LRP Minireceptors | CHO-mLRP4 cells were seeded at 1.2 x 106 cells per T25 flask and cultured overnight before experiments. Cells were washed in serum-free medium and incubated with or without proteasomal inhibitors (20 muM MG132 or lactacystin) for varying lengths of time. Cells were then detached by incubation with nonenzymatic cell dissociation solution (Sigma). Successive incubations with affinity-purified anti-HA IgG (25 mug/ml) and goat anti-mouse Ig-FITC were carried out at 4C for 45 min each. Background fluorescence intensity was assessed in the absence of primary mAb and subtracted from all samples. Mean fluorescence values were obtained in at least triplicate on a FACScalibur (BD Biosciences-PharMingen), and data were analyzed with Cell Quest software. Kinetic Analysis of mLRP4 Endocytosis | Fifty micrograms of RAP were iodinated using the Iodogen method as described . Kinetic analysis of endocytosis was performed essentially as described . Briefly, CHO-mLRP4 cells were plated in 12-well plates at a density of 2 x 105 cells/well and used after overnight culture. Cells were pretreated in the presence or absence of MG132 (20 muM) for either 30 min or 2 h. 125I-RAP was added at 5 nM final concentrations in cold ligand binding buffer (0.5 ml/well). The binding of 125I-RAP was carried out at 4C for 30 min with gentle rocking. Unbound ligand was removed by washing cell monolayers three times with cold binding buffer. Ice-cold stop/strip solution (0.2 M sodium acetate/acetic acid, pH 2.6, 0.1 M NaCl) was added to one set of plates without warming up. The remaining plates were then placed in a 37C water bath, and 0.5 ml of ligand binding buffer with or without MG132 and prewarmed to 37C was quickly added to the cell monolayers to initiate internalization. After each time point, the plates were quickly placed on ice and the ligand binding buffer was replaced with cold stop/strip solution. Ligand that remained on the cell surface was stripped by incubation of cell monolayers with cold stop/strip solution for a total of 20 min (0.75 ml for 10 min, twice) and counted. Cell monolayers were then solubilized with low-SDS lysis buffer (62.5 mM Tris-HCl, pH 6.8, 0.2%SDS, 10% vol/vol glycerol, and 0.02% bromophenol blue) and counted in a gamma counter. The sum of ligand that was internalized plus that remained on the cell surface after each assay was used as the maximum potential internalization. The fraction of internalized ligand after each time point was calculated and plotted. Uptake of Cy3-labeled Ligands and Confocal Immunofluorescence Microscopy | Ts20-mLRP4 or ts20-GHR cells were grown on glass coverslips overnight before experiments. Cells were then incubated at 30, 41.5, or at 30C with either DMSO or DMSO containing 20 muM MG132 for 1 h. Incubation was continued for 10 min with Cy3-RAP (20 nM) or 30 min with Cy3-GH (8 nM) at 30 or 41.5C. After medium aspiration, cells were washed once with PBS and once with fixative (3% paraformaldehyde in 0.1 M sodium phosphate, pH 7.2), fixed for 3 h in fixative, washed with PBS and water, and embedded in Mowiol. The laser scanning microscopy was performed using a TCS 4D system and 63x oil immersion lens (Bensheim, Germany). Uptake and Degradation of Radiolabeled Ligands | Iodinated uPA-PAI-1 complexes were added to vehicle or drug-pretreated cells as described above. Radiolabeled ligand with or without excess unlabeled RAP (0.5 muM) was incubated for 4 h at 37C. The overlying media were collected from cells and precipitated with 20% trichloroacetic acid (TCA). TCA soluble counts, after subtraction of TCA-soluble counts in well contained no cells, were used to represent cell-mediated degradation of radiolabeled ligand. Unbound ligand was removed by washing cell monolayers three times with cold binding buffer. Ligand that remained on the cell surface was stripped by incubation of cell monolayers with cold stop/strip solution for a total of 20 min and counted. Cell monolayers were then washed twice in PBSc and lysed in low-SDS buffer to quantitate cell-associated ligand. Immunoelectron Microscopy | CHO-mLRP4 cells were incubated in the absence or presence of 20 muM MG132 for 2 h and thereafter fixed in a mixture of 2% paraformaldehyde and 0.5% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4, for 1 h and then stored in the same buffer containing 1% paraformaldehyde until use. Cryosections of 50 --100 nm were picked up from the diamond knife in a sucrose/methylcellulose mixture and sequentially incubated with mouse monoclonal HA antibody, polyclonal rabbit anti-mouse IgG (Dako, Copenhagen, Denmark), and 10-nm protein A --conjugated gold particles (for technical details see ). mLRP4 labeling at the cell surface was quantitatively evaluated in a total of 50 electron micrographs at x15,000 taken randomly from sections of two blocks with control and two blocks of MG132-treated cells. Totals of gold particles and measured plasma membrane lengths in control cells and MG132 cells were 680/969 and 1733/949 mum, respectively. mLRP4 labeling densities represented by the ratios of gold particles over plasma membrane length were 0.76 +- 1.12 and 1.78 +- 2.08 (means +- SD) in control and MG132 cells, respectively. RESULTS : Inhibition of Proteasome Activity Increases the Half-Life and the Steady State Levels of LRP | To determine whether proteasomal inhibitors influence the cellular turnover of LRP, we analyzed the half-life of LRP in the absence or presence of the proteasome inhibitor MG132 in HepG2 cells. These cells were metabolically pulse labeled with [35S]cysteine for 30 min and chased for 0.5, 1, 2, 4, or 8 h, in the absence or presence of MG132 (20 muM). After each chase time, cell lysates were quantitatively immunoprecipitated with anti-LRP antibody and analyzed by SDS-PAGE under reducing conditions (Figure A). When the band intensities were quantitated with a phosphorimager, we found that the half-life of LRP is prolonged from similar3 h in the absence of MG132 to >8 h in the presence of MG132 (Figure B). Similar results were obtained when lactacystein, was used as the proteasomal inhibitor. Figure 1 | The proteasomal inhibitor MG132 prolongs the cellular half-life of LRP. The proteasomal inhibitor MG132 prolongs the cellular half-life of LRP. (A) HepG2 cells were metabolically labeled with [35S]cysteine for 30 min and chased for the indicated times, in the absence or presence of MG132 (20 muM). After each chase, cell lysates were immunoprecipitated with anti-LRP antibody and analyzed with SDS gels (5%) under reducing conditions. The 200-kDa molecular size marker is indicated. (B) The band intensities in A were quantitated and plotted against chase times. Many of our recent studies on LRP endocytosis and trafficking have been carried out mostly using a minireceptor form of LRP, which includes the fourth ligand-binding domain and the entire carboxyl-terminus of the receptor . This receptor, referred to as mLRP4, exhibits similar biogenesis and trafficking with the endogenous full-length receptor, including the furin cleavage to generate the two subunits . One advantage of this minireceptor is the ability to readily distinguish the mature form from the precursor form of the protein, which is extremely difficult for the endogenous full-length LRP because of its large size . Additionally, mLRP4 is tagged with an HA epitope near its amino-terminus, which facilitates various immunodetection strategies including immunoelectron microscopy (see below). To examine whether the cellular turnover of mLRP4 is also regulated by the proteasome, we performed metabolic pulse-chase labeling with mLRP4 stably expressed in CHO LRP-null cells (; stable cell line is termed CHO-mLRP4). Seen in Figure are results from a representative experiment. As seen in the figure, the presence of MG132 significantly increased the half-life of the mature forms of mLRP while exhibiting little effect on the precursor form; the half-life of mLRP4 was increased from similar2.5 to similar5.5 h. To exclude the possibility that the effect of proteasome inhibitor on the half-life mLRP4 was due to the different cell lines used in our studies, we also performed pulse chase experiments to examine mLRP4 half-life in mLRP4-transiently transfected HepG2 cells. We found that MG132 increased the half-life of mLRP4 in HepG2 cells from similar4.2 to 8.1 h. These results demonstrate that the proteasome inhibitor MG132 increases the half-life of mLRP4 in both CHO LRP-null cells and HepG2 cells. Figure 2 | The proteasomal inhibitor MG132 increases the half-life of a LRP minireceptor, mLRP4. The proteasomal inhibitor MG132 increases the half-life of a LRP minireceptor, mLRP4. CHO-mLRP4 cells were metabolically labeled with [35S]cysteine for 30 min and chased for the indicated times, in the absence or presence of MG132 (20 muM). After each chase, cell lysates were immunoprecipitated with anti-HA antibody and analyzed with SDS gels (6%) under reducing conditions. The molecular size markers in this and subsequent figures are given in kilodaltons on the left. The positions of the precursor and the mature forms of mLRP4 are marked. (B) The band intensities in A (includes the precursor and the mature forms) were quantitated and plotted against chase time. To examine whether an increase of mLRP4 half-life results in an increase in the steady state level of mLRP4, we treated CHO-mLRP4 cells for various periods of time and analyzed the total cellular mLRP4 via Western blotting with anti-HA antibody. Only the precursor form and the ligand-binding subunit of the mature form are shown in the blot because these forms, but not the 85-kDa subunit mature form, contain the HA epitope. As seen in Figure A, a significant increase of the mature form, which is present at the cell surface and later compartments of the endocytic pathway, is seen in the presence of MG132. An average of similar2.5-fold increase of mature mLRP4 was consistently observed using this method. The increase in steady state level of mLRP4 was also seen upon incubation with lactacystin but not with the vehicle DMSO alone (Figure B). These results suggest that the cellular turnover of LRP is mediated via or regulated by proteasomal activity. Figure 3 | Treatment of CHO-mLRP4 cells with proteasomal inhibitors increases the steady state levels of mLRP4. Treatment of CHO-mLRP4 cells with proteasomal inhibitors increases the steady state levels of mLRP4. (A) CHO-mLRP4 cells were treated with either vehicle DMSO or DMSO containing MG132 (20 muM) for various periods of time. Equal quantities of cell lysates were separated via SDS gel (6%) and blotted with anti-HA antibody. Only the precursor form and the ligand-binding subunit of the mature form are detected in the blot because these forms, but not the 85-kDa subunit mature form, contain the HA epitope. Note the increase of the mature form, but not the precursor form of mLRP4 after MG132 treatment. (B) The effects of proteasomal inhibition on steady state levels of mLRP4 were also seen with lactacystin, but not with vehicle DMSO alone. Treatment of Cells with Proteasomal Inhibitors also Increased Cell Surface mLRP4 | To analyze whether upon treatment of cells with proteasome inhibitors there was also a change of cell surface mLRP4, we performed FACS analysis using the anti-HA antibody. As seen in Figure A, we consistently observed an approximately threefold increase (average of 5 experiments) of cell surface mLRP4 after a 2-h treatment of CHO-mLRP4 cells with MG132. To examine the kinetics of the cell surface increase of mLRP4, we treated CHO-mLRP4 cells with MG132 for increasing periods of time and measured cell surface mLRP4 via FACS analyses. Shown in Figure B are results from a representative experiment. As seen in the figure, there is a gradual increase of cell surface mLRP4 upon MG132 treatment, suggesting that the cell surface increase of mLRP4 is cumulative. A similar increase in cell surface mLRP4 was also observed, at the selected time points, with lactacystin, but not with vehicle DMSO alone (Figure B). These results suggest that proteasomal inhibitors likely block LRP trafficking at an endosomal compartment (e.g., sorting endosome) from which the receptor is still able to recycle to the cell surface. Figure 4 | Proteasomal inhibitors increase cell surface mLRP4. Proteasomal inhibitors increase cell surface mLRP4. (A) CHO-mLRP4 cells were treated with DMSO or DMSO containing MG132 (20 muM) for 2 h. Cells were then dissociated and immunostained with anti-HA antibody. Cell surface mLRP4 was detected with goat anti-mouse Ig-FITC and FACS analysis. Background fluorescence intensity was assessed in the absence of primary antibody and subtracted from each assay. The mean values from representative triplicate determinations were averaged with the SE value given as error bars. (B) Kinetic analyses of cell surface mLRP4 after treatment with proteasomal inhibitors. CHO-mLRP4 cells were treated with MG132 (20 muM) for various periods of time as indicated. Treatment with the alternative proteasome inhibitor lactacystin (20 muM) or vehicle DMSO alone were also performed at selected time points. Cell surface mLRP4 was determined via FACS analysis as indicated above. Shown in the figure are results of a representative experiment. Note the gradual increase of cell surface mLRP4 upon treatment with proteasomal inhibitors, but not DMSO. Proteasomal Inhibitors Slow, but Do Not Block the Initial Endocytosis of mLRP4 | Recent studies have shown that the proteasome regulates not only the degradation but also the initial endocytosis of several cell surface receptors including the GHR . To determine whether the endocytosis of mLRP4 is regulated by proteasomal activity, we quantitated the initial endocytosis rate of mLRP4 after pretreatment of cells with MG132 for either 30 min (Figure A) or 2 h (Figure B). As seen in Figure A, there was no significant change of mLRP4 endocytosis rate when cells were pretreated with MG132 for only 30 min. However, a significant (similar40%) decrease in mLRP4 endocytosis rate was seen when cells were pretreated for 2 h with MG132 (Figure B). It is important to point out that although the initial (<=30 s) endocytosis rate was decreased in the presence of MG132, the overall ability of mLRP4 to endocytose ligands was essentially indistinguishable in the absence or presence of MG132 when endocytosis evaluated >120 s. Figure 5 | The initial endocytosis of mLRP4 is slowed, but not blocked by proteasomal inhibitors. The initial endocytosis of mLRP4 is slowed, but not blocked by proteasomal inhibitors. CHO-mLRP4 cells were pretreated with DMSO or DMSO containing MG132 (20 muM) for either 30 min (A) or 2 h (B). The initial endocytosis rates of mLRP4 were measured using 125I-RAP as described in MATERIALS AND METHODS. Data represent triplicate determination with SE values given as error bars. To determine whether a functional ubiquitin system is required for LRP-mediated endocytosis, we compared the endocytosis of ligands for mLRP4 (RAP) and GHR (GH) in ts20 cells stably expressing mLRP4 (ts20-mLRP4) or GHR (ts20-GHR). The ts20 cells express a thermolabile, ubiquitin-activating enzyme, E1 . At the permissive temperature (30C), both Cy3-RAP (Figure A) and Cy3-GH (Figure B) were endocytosed. However, at the nonpermissive temperature (41.5C), wherein the ubiquitin activating enzyme is inactivated, Cy3-RAP (Figure C), but not Cy3-GH (Figure D), were internalized. This indicates that the initial endocytosis of GHR, but not mLRP4, requires a functional ubiquitin-activating/conjugating system . Using similar assays at the permissive temperature, we also found that the proteasomal inhibitor, MG132, inhibits the endocytosis of GH (Figure F), but not RAP (Figure E), consistent with our previous observation that the initial endocytosis of the GHR, but not mLRP4, requires proteasomal activity (; also see Figure ). We confirmed the clathrin-dependence of mLRP4 and GHR endocytosis because potassium-depletion inhibited Cy3-RAP and Cy3-GH endocytosis. Figure 6 | A functional ubiquitin conjugation system is not required for mLRP4-mediated endocytosis. A functional ubiquitin conjugation system is not required for mLRP4-mediated endocytosis. CHO-ts20-mLRP4 or CHO-ts20-GHR cells were incubated with Cy3-RAP (20 nM) for 10 min or Cy3-GH for 30 min, respectively, under the conditions indicated. Cells were washed, fixed, and examined with a confocal microscope. At the permissive temperature (30C), endocytosis of both Cy3-RAP (A) and Cy3-GH (B) is observed by the appearance of punctuate fluorescence representing ligand distribution in endocytic compartments. The endocytosis of Cy3-GH, but not Cy3-RAP, was inhibited at the nonpermissive temperature (41.5C), and in the presence of the proteasomal inhibitor MG132 (C-F). Our previous studies have shown that mLRP4 is capable of binding and endocytosing several LRP ligands, including uPA-PAI-1 complexes . To examine the fate of mLRP4 ligands in the absence or presence of proteasomal inhibitors, we performed ligand uptake and degradation assays using CHO-mLRP4 cells. In this experiment, 125I-uPA-PAI-1 was incubated with CHO-mLRP4 cells at 37C for 4 h. Cell-mediated degradation and intracellular and cell surface bound ligands were determined at the end of assay. As seen in Figure A, MG132 significantly inhibited cellular degradation of 125I-uPA-PAI-1. A corresponding increase of intracellular 125I-uPA-PAI-1 was seen (Figure B), suggesting that the block in 125I-uPA-PAI-1 degradation likely involves inhibition of ligand delivery to the lysosomal compartment. A small increase in cell surface 125I-uPA-PAI-1 was also observed (Figure C), consistent with the fact that MG132 slows mLRP4-mediated endocytosis (see Figure ). Taken together, these results indicate that proteasomal activity is required for intracellular trafficking of mLRP4 and its ligand to the degradation pathway. Figure 7 | Treatment of CHO-mLRP4 cells with proteasomal inhibitors results in intracellular accumulation and inhibition of degradation of mLRP4 ligand. Treatment of CHO-mLRP4 cells with proteasomal inhibitors results in intracellular accumulation and inhibition of degradation of mLRP4 ligand. 125I-uPA-PAI-1 (2 nM) was incubated with CHO-mLRP4 cells at 37C for 4 h. Cell-mediated degradation and intracellular and cell surface bound ligands were determined as described in MATERIALS AND METHODS. Data represent triplicate determinations with SE values given as error bars. Trafficking of mLRP4 to the Multivesicular Bodies Is Blocked on Proteasome Inhibition | To examine morphologically the effects of proteasome inhibitors on LRP endocytic trafficking, we performed immunogold EM labeling of mLRP4 in MG132-treated and control cells. Differences in mLRP4 labeling between control and MG132 cells mainly concerned the plasma membrane and late endocytic compartments. First, in control cells most of the surface mLRP4 was localized to clathrin-coated pits (Figure A), similar to what we have observed previously for the endogenous LRP . However, in MG132-treated cells, mLRP4 is observed in both clathrin-coated pits, and outside the pits (Figure , B and E). A quantitative evaluation of the labeling at the plasma membrane revealed a statistically significant 2.5-fold greater labeling density of mLRP4 in MG132 cells than in control cells (see MATERIALS AND METHODS for quantitation results). Second, mLRP4 in control cells was found throughout the endocytic pathway, i.e., primary endocytic vesicles, multivesicular bodies (MVBs; Figure , C and D), and lysosomes. Within the MVBs, mLRP4 labels were associated with both the limiting membrane and the internal vesicles, thought to be destined for degradation . The limiting membrane showed triple-layered, clathrin-coated lattices, likely involved in the recruitment of membrane proteins for targeting to the internal vesicles . In contrast, MG132 cell profiles showed no or few MVBs with scarce mLRP4 labeling. Instead, mLRP4 was found intensely in long peripheral tubules, probably representing early endosomes/recycling tubules (Figure E). Figure 8 | Proteasome inhibitors increase surface localization of mLRP4 and block the delivery of mLRP4 into MVBs. Proteasome inhibitors increase surface localization of mLRP4 and block the delivery of mLRP4 into MVBs. Cryosections of CHO-mLRP4 cells were immunogold labeled with anti-HA antibodies and 10 nm gold particles. (A, C, and D) control cells; (B and E) MG132-treated cells. In control cells (A) by far the majority of mLRP4 at the plasma membrane is present in clathrin-coated pits, whereas in cells treated with MG132 (B and E) label is also present outside the pits. (C and D) MVBs in control cells with label present at the limiting membrane and associated with the internal vesicles. Note the thick clathrin lattices at the limiting membrane (arrows). (E) Peripheral cytoplasm of a MG132-treated cell with labeled clathrin-coated vesicles and long recycling tubules (arrowheads). All bars represent 200 nm. Residues 60 --78 of the LRP Tail Are Required for Proteasomal Regulation | To begin to identify cis-elements within the LRP tail that are required for proteasomal regulation, we examined the effects of MG132 treatment on CHO LRP-null cells stably expressing various tail-truncation mutants of mLRP4 . As seen in Figure A, upon MG132 treatment increase of cell surface mLRP4 was nearly abolished when the tail is truncated to 59 amino acid residues but not when truncated to 78 amino acids. To examine whether these differences in response to MG132 also apply to the steady state levels of these truncated minireceptors, we compared their steady state levels in the absence or presence of MG132 treatment. As seen in Figure B, each of these mLRP4 minireceptors is expressed at a similar basal level when the precursor forms are compared. The increased ratios of mature to precursor forms for mLRP4Tailess and mLRP4T59 when compared with that of mLRP4T100 are due to the slower endocytosis rates mediated by these two truncated minireceptors . More importantly, we demonstrated that the proteasome inhibitor MG132 increased the steady state levels of the mature forms of mLRP4T78 and mLRP4T100 but not those of mLRP4Tailess and mLRP4T59. These results together suggest that the region of the tail between amino acids 60 and 78 may contain an important sequence or binding site that when present contributes to mLRPs regulation by the proteasome. Additionally, because mLRP4T59 still exhibits significant endocytosis , but is minimally regulated by the proteasome, the results further suggest that the endocytosis and proteasomal regulation of LRP are separate events. Figure 9 | The increase in cell surface mLRP4 upon MG132 treatment requires residues 60 --78 of the LRP tail. The increase in cell surface mLRP4 upon MG132 treatment requires residues 60 --78 of the LRP tail. CHO LRP-null cells stably transfected with cDNAs of mLRP4(T100), mLRP4T78, mLRP4T59, or mLRP4T tailess were treated with DMSO or DMSO containing MG132 (20 muM) for 2 h at 37C. (A) Cell surface mLRP4 minireceptor levels were determined as described in Figure . Values are the average of triple determinations with the SE value indicated by error bars. (B) The steady state levels of mLRP4 minireceptors were analyzed as described in Figure . DISCUSSION : LRP has emerged as a unique receptor in the LDLR family in several aspects. First, LRP is a multifunctional receptor with the ability to bind >30 distinct ligands . Second, its endocytosis is significantly faster than other members of the LDLR family . Third, its physiological functions are carried out by both receptor-mediated endocytosis and signal transduction . The high level expression of this LRP in the brain and liver as well as its multiligand nature are consistent with its requirement for normal embryonic growth and development . Additionally, important roles for LRP in the pathogenesis of disease (e.g., Alzheimer's disease) have recently been recognized . However, despite advances in understanding the biogenesis of this giant receptor, including the role of its intracellular chaperone RAP , little is known about the mechanisms governing its cellular turnover. In this study, we demonstrate that inhibition of proteasomal activity results in an increase of cellular half-life of LRP and its minireceptor mLRP4. Additionally, we showed that proteasomal inhibitors block the trafficking of mLRP4 into MVBs. The block in mLRP4 trafficking to the MVBs was also indicated as the mLRP4, which accumulated in the presence of proteasome inhibitors, recycled back to the cell surface. Thus, we conclude that the trafficking of LRP to the degradation pathway is regulated by proteasomal activity. The importance of the proteasome activity in receptor trafficking to the degradation pathway has also been shown recently for the EGFR . In that case, inhibition of proteasome also blocked receptor degradation and promoted receptor recycling . Cell surface receptors entering sorting endosomes can be either recycled to the plasma membrane or degraded via delivery to the lysosome . Increasing evidence has shown that the recycling of receptors is the default pathway, whereas delivery to the degradation pathway is signal mediated. Critical step in the latter process occurs in the MVBs where the limiting membrane invaginates and buds into the lumen of the MVBs/late endosome . One of the best-studied examples of this sorting process is the downregulation of the EGFR. On EGF activation, the delivery of the EGFR to the MVBs/late endosome requires sorting signals within the EGFR tail and the ubiquitin ligase c-Cbl . Recently, it has been shown that the ubiquitin-dependent sorting into the MVBs pathway requires the function of a conserved endosomal protein sorting complex, ESCRT-1 . This complex, composed of three products of the class E vacuolar protein sorting (VPS) genes (Vps23, Vps 28, and Vps 37), recognizes ubiquitinated cargo molecules and initiates their sorting into the lumen of the MVBs. Thus, one approach to define the mechanism of LRP downregulation by the ubiquitin --proteasome system is to determine whether the cytoplasmic tail of this receptor can be ubiquitinated. Despite a recent report that suggests that the extracellular subunit of LRP can be ubiquitinated , we have failed to detect any ubiquitination of LRP after an extensive series of approaches, including both in vivo immunodetection of ubiquitinated LRP species in the presence of proteasomal inhibitors and in vitro ubiquitination assays of GST-LRP tail fusion proteins . These results suggest that an ancillary protein, which itself may be ubiquitinated and thus regulated by the ubiquitin --proteasome system, may function as a regulatory protein for LRP turnover. In support of this hypothesis, we show herein that a region of the LRP tail (residue 60 --78) is required for its proteasomal regulation. It is possible that this region of the LRP tail contains a sequence element important for interaction with the regulatory protein, perhaps similar to that in the growth hormone receptor or beta-arrestin, associated with the beta2-adrenergic receptor . An alternative mechanism underlying proteasomal regulation of LRP may involve a short-lived protein that functions in endosomal sorting and is regulated by the proteasome. Our previous studies have shown that a functional ubiquitin conjugation system is required for the initial endocytosis of the GHR . The likely mechanism appears to be that the ubiquitination of an ancillary protein that interacts with GHR and serves as the endocytosis signal for the receptor. Our current study shows that such a functional ubiquitination system is not required for LRP endocytosis. This is not surprising because recent studies have shown that the rapid endocytosis of LRP is mediated by both a tyrosine-based signal as well as a di-leucine motif . The slower endocytosis of LRP seen upon prolonged treatment of cells with proteasomal inhibitors may be due to an indirect effect of the inhibitors on the turnover of some component(s) of the endocytic machinery. However, the initial endocytosis of the transferrin receptor is not affected by these same proteasomal inhibitors. This suggests that the mechanism(s) underlying LRP endocytosis may utilize partially a distinct mechanism for its rapid endocytosis. The need for downregulation of signal transducing receptors is for signal desensitization. However, the significance of downregulation of receptors whose primary function is cargo transport is less clear. Receptors such as the LDLR and the transferrin receptor, whose sole recognized function is cargo transport, typically exhibit long half-lives . While examining the turnover of LRP in various cell types, we noted half-lives ranging from 3.5 h in HepG2 cells to >8 h in U87 cells . Thus, the function of LRP in cargo transport and/or signal transduction may vary among cell types and/or in the presence of different ligands. LRP may represent a distinct class of receptor with both cargo transport as well as signal transduction activity) whose cellular turnover is regulated by more than one mechanism. In this regard it is tempting to speculate that the regulation of LRP endocytic trafficking is also unique and may involve both a specific cis-element within its cytoplasmic tail, as well as an unidentified cytosolic protein that recognizes this tail element. Backmatter: Abbreviations used: : LDLR = low-density lipoprotein receptor LRP = LDLR-related protein GH = growth hormone GHR = growth hormone receptor EGF = epidermal growth factor EGFR = epidermal growth factor receptor CHO = Chinese hamster ovary RAP = receptor-associated protein EM = electron microscopy MVBs = multivesicular bodies PMID- 12221137 TI - A Ypt32p Exchange Factor Is a Putative Effector of Ypt1p AB - Ypt1p regulates vesicle tethering and fusion events from the ER to the Golgi and through the early Golgi. Genetic studies have suggested a functional relationship between Ypt1p and Ypt31p/Ypt32p. Ypt31p and Ypt32p are a pair of functionally redundant GTPases that act after Ypt1p to mediate intra-Golgi traffic or the budding of post-Golgi vesicles from the trans-Golgi. Here we report that a novel Ypt32p exchange factor is a putative effector of Ypt1p. These findings implicate small GTP-binding proteins of the Ypt/Rab family in a signal cascade that directs membrane traffic through the secretory pathway. Keywords: INTRODUCTION : Newly synthesized secretory proteins are translocated into the ER and then transported to the plasma membrane via the Golgi apparatus by carrier vesicles. Intracellular vesicle traffic requires an efficient mechanism to direct vesicles to their appropriate target compartment. The Ypt/Rab family of Ras-related small GTP-binding proteins are involved in the regulation of protein transport through the different steps of the exocytic pathway . The Saccharomyces cerevisiae genome encodes 11 Ypt/Rab proteins . Ypt1p, the ortholog of the mammalian small GTP-binding protein Rab1, acts in both ER-to-Golgi and intra-Golgi traffic , whereas Ypt31p and its functional homologue Ypt32p have been implicated in traffic through and from the Golgi . Ypt/Rab proteins function as molecular switches by cycling between an inactive GDP-bound and active GTP-bound conformation . This cycle of activation and inactivation is regulated by guanine nucleotide exchange factors (GEFs) and GTPase-activating proteins (GAPs). GEFs promote GDP dissociation and GTP uptake, which converts Ypt/Rab proteins to their active form. A nucleotide exchange activity for Ypt1p located on Golgi membranes is essential for Ypt1p mediated fusion events . Recently, we have demonstrated that this Ypt1p exchange factor is TRAPP, a highly conserved multiprotein complex that peripherally associates with the Golgi . There are two forms of the TRAPP complex, TRAPP I and TRAPP II. The two TRAPP complexes share seven subunits (Bet5p, Trs20p, Bet3p, Trs23p, Trs31p, Trs33p, and Trs85p), whereas three subunits (Trs65p, Trs120p, and Trs130p) are unique to TRAPP II. Although both complexes localize to the same early Golgi compartment, mutational analysis and in vitro transport studies have revealed they mediate different transport steps. TRAPP I is required for the tethering of ER-derived COP II vesicles to the Golgi, whereas TRAPP II has been implicated in Golgi traffic . Both forms of TRAPP, TRAPP I and TRAPP II, can exchange nucleotide on Ypt1p. The finding that Ypt1p is activated by two distinct but related exchange factors explains how this small GTP-binding protein acts in two different transport events . In addition to Ypt1p, TRAPP was also reported to act as an exchange factor for Ypt31p/32p . In their active GTP-bound state, Ypt/Rab proteins interact with downstream effectors that control the targeting, docking and fusion of transport intermediates with their appropriate acceptor compartments . Here we report that a novel Ypt32p exchange activity is a putative effector of Ypt1p. Furthermore, we show that TRAPP is the major exchange factor for Ypt1p, but not Ypt32p. Our findings imply that small GTP-binding proteins of the Ypt/Rab family may act in a signal cascade to direct membrane traffic. MATERIALS AND METHODS : Purification of TRAPP | TRAPP was purified from a strain in which the sole copy of Trs33p was TAP tagged . Approximately 10,000 OD599 units of cells were washed and lysed with a bead beater in 70 ml of buffer I (20 mM HEPES, pH 7.2, 150 mM NaCl). The salt concentration was adjusted to 300 mM before centrifugation at 25,000 x g for 20 min, and the supernatant was incubated with 0.4 ml of IgG-Sepharose beads (Amersham Biosciences, Piscataway, NJ) for 2 h. The beads were first washed with IPP150 buffer (10 mM Tris, pH 8.0, 150 mM NaCl) and then TEV buffer (10 mM Tris, pH 8.0, 150 mM NaCl, 0.5 mM EDTA, 1 mM DTT) and incubated with 700 U of the TEV protease (GIBCO BRL Product, Rockville MD) at room temperature for 2 h. The released TRAPP was bound to 0.3 ml of calmodulin affinity resin (Stratagene, La Jolla, CA) in binding buffer (10 mM Tris, pH 8.0, 150 mM NaCl, 1 mM Mg(OAc)2, 1 mM imidazole, 2 mM CaCl2, 10 mM 2-mercaptoethanol). The beads were washed with binding buffer and buffer II (50 mM Tris, pH 8.0, 5 mM MgCl2, 2 mM CaCl2, 1 mM ATP, 1 mM DTT, 1 mg/ml BSA) and then resuspended in uptake buffer (50 mM Tris, pH 8.0, 5 mM MgCl2, 1 mM EDTA, 1 mM ATP, 1 mM DTT, 1 mg/ml BSA) and aliquoted into six equal portions for the nucleotide exchange assay. Protein A (PrA)-tagged TRAPP was purified as described before . Preparation of TRAPP-depleted Cytosol | TRAPP was depleted from cytosol prepared from a strain in which the DSS4 gene was disrupted. DSS4 was disrupted by replacing the ORF with the HIS3 gene. Briefly, a hybrid sequence containing the HIS3 gene, flanked by part of DSS4, was amplified by PCR from plasmid pFA6a-His3MX6 . This product was then transformed into SFNY1086 (MATalpha ura3-52 BET3Delta::URA3 leu2-3,112, BET3-protein A::LEU2, his3-Delta200) and Ura+ Leu+ His+ colonies were selected and purified. The disruption of DSS4 was confirmed by PCR, and one of the transformants was named SFNY1088 (MATalpha ura3-52 BET3Delta::URA3 leu2-3,112 BET3-protein A::LEU2, his3-Delta200, DSS4Delta::HIS3). To deplete TRAPP from cytosol prepared from SFNY1088, 1000 OD599 units of cells were converted to spheroplasts as described before and lysed with a Wheaton dounce homogenizer in 5 ml of buffer containing 20 mM HEPES, pH 7.2, 150 mM NaCl, 1 mM DTT, and 1x protease inhibitor cocktail (PIC; ). The lysate was centrifuged at 60,000 x g for 30 min, and 15.5 mg of the supernatant was incubated overnight with 0.1 ml of packed IgG-Sepharose beads. The beads were spun, and the supernatant was incubated two more times with 0.1 ml of fresh beads for 4 h. The same amount of supernatant was mock treated with Sepharose 4B (Sigma, St. Louis, MO) in the same way. The presence of TRAPP subunits in cytosol was detected by Western blot analysis using the ECL method. Nucleotide Exchange Assays | The GTP uptake and GDP dissociation assays were performed as described in . In the GTP uptake assay, 5 pmol of recombinant (His6)-Ypt1p or (His6)-Ypt32p, produced in Escherichia coli, was incubated at room temperature in uptake buffer with an aliquot (50 mul) of calmodulin-coated agarose beads +- TRAPP. Assays were performed in the presence of 5 pmol of [35S]GTPgammaS (NEN Life Science Products, Boston, MA) for the indicated times (see Figure ). The exchange reaction was stopped by adding 1 ml of ice cold stop buffer (20 mM Tris, pH 8.0, 25 mM MgCl2) and filtered through a nitrocellulose filter (Millipore, Bedford, MA). The filter was washed three times with 3 ml of the stop buffer and then dried. The radioactivity bound to the filter was counted in a scintillation counter. Figure 1 | Chemically pure TRAPP stimulates GTPgammaS uptake onto Ypt1p but not Ypt32p. Chemically pure TRAPP stimulates GTPgammaS uptake onto Ypt1p but not Ypt32p. TRAPP was purified by affinity purification from a strain in which Trs33p was TAP tagged. Calmodulin agarose beads with or without TRAPP were incubated with 5 pmol of Ypt1p or Ypt32p at room temperature in the presence of [35S]GTPgammaS. At the time intervals indicated, radioactivity that bound to protein was measured by the filter-binding assay described in MATERIALS AND METHODS. The data are expressed as picomoles of GTPgammaS retained on the filter. The GDP dissociation assay was performed in the same way with the following modifications. Ypt1p (2 mM) or Ypt32p (2 mM) was preloaded with 12 mM of [8'-3H]GDP (Amersham Biosciences) at 30C for 15 min in preloading buffer (20 mM HEPES, pH 7.2, 5 mM EDTA, 1 mM DTT). At the end of the incubation, 10 mM MgCl2 was added to each assay. [3H]GDP-Ypt1p (5 pmol) or [3H]GDP-Ypt32p (5 pmol) was incubated at 30C in release buffer (20 mM HEPES, pH 7.2, 5 mM MgCl2, 1 mM DTT, 0.75 mM GTP, 0.75 mM GDP, 1 mg/ml BSA) with 0.61 pmol of PrA TRAPP, or 14 mg/ml lysate, or an aliquot of glutathione agarose (80 mul) coated with GST-Ypt1p or GST for varying periods of time. In Vitro Binding to Ypt1p-GTPgammaS | Recombinant GST-Ypt1p and GST-Ypt51p were purified from 1 liter of E. coli (BL21). Expression was induced during a 15-h incubation at 20C by the addition of IPTG (1 mM). The fusion protein was bound to 0.6 ml of glutathione Sepharose (Amersham Biosciences) according to the manufacturer's protocol and stored in phosphate-buffered saline (PBS) with 5 mM MgCl2. Beads bound with 4 mg of GST-Ypt1p or GST-Ypt51p were washed with buffer A (PBS plus 0.5 mM MgCl2, 1 mM DTT, 1 mg/ml BSA, and 1x PIC) containing 10 muM GTPgammaS or with buffer B (PBS plus 5 mM MgCl2, 10 mM EDTA, 1 mM DTT, 1 mg/ml BSA, and 1x PIC) containing 10 muM GDP. The beads were then incubated with buffer A in the presence of 1.5 mM GTPgammaS or buffer B in the presence of 1.5 mM GDP for 30 min at room temperature. The washes and incubations were repeated once more before the beads were washed with buffer C (PBS plus 10 mM MgCl2, 1 mM DTT, 1 mg/ml BSA) containing 0.2 mM GTPgammaS or GDP. A wild-type yeast lysate (SFNY26-3a, MATa ura3-52) was prepared by converting 15,000 OD599 units of cells to spheroplasts and lysing the spheroplasts in 100 ml of lysis buffer (20 mM HEPES, pH 7.2, 100 mM NaCl, 10 mM MgCl2, 1% Triton X-100, 1 mM DTT, and 1x PIC) with a Wheaton dounce homogenizer. The lysate was centrifuged at 18,000 x g for 15 min, and 1 g of the supernatant was incubated with GST-Ypt1p or GST-Ypt51p, prepared as described above, for 2 h at 4C in the presence of 0.2 mM GTPgammaS or GDP. The beads were then washed sequentially with buffer C containing 10 muM GTPgammaS or 10 muM GDP and buffer D (20 mM HEPES, pH 7.2, 5 mM MgCl2, 1 mM DTT, 1 mg/ml BSA) containing 0.75 mM GTP or GDP, resuspended in release buffer and aliquoted into five equal portions for nucleotide exchange assays. Cell Fractionation and Elution of Ypt32p Exchange Activity | Approximately 1000 OD599 units of SFNY1088 or SFNY26-3a cells were converted to spheroplasts as described by and lysed in 10 ml of lysis buffer (20 mM HEPES, pH 7.2, 5 mM MgCl2, 1 mM DTT, and 1x PIC). The unbroken cells were removed during a 5-min spin at 500 x g, and the supernatant (lysate) was centrifuged at 12,000 x g for 10 min. The pellet (P12) was resuspended in lysis buffer ( of the initial volume), and the supernatant (S12) was centrifuged at 100,000 x g for 1 h to obtain S100 and P100 fractions. The P100 fraction was resuspended in lysis buffer ( of the initial volume) and incubated for 1 h on ice with an equal volume of lysis buffer with or without 1 M NaCl. The mixture was centrifuged at 100,000 x g for 1 h, and the supernatant was assayed for exchange activity. Purification of Uso1p | Uso1p was purified from SFNY779 (MAT a ura3-52 pUSO1-myc [2 mum, URA3]) as described before with the following minor modifications. Cytosol was loaded onto a Mono Q HR10/10 column (Amersham Biosciences), and the column was eluted with a 20-ml linear salt gradient (0.5 --2 M KOAc). Fractions of 0.5 ml were collected, and the presence of Uso1p in each fraction was detected by Western blot analysis with 9E10 mAb. The peak of Uso1p from the Mono Q column was applied to a Superdex-200 gel filtration column, and the fractions containing Uso1p were pooled and concentrated. RESULTS : Chemically Pure TRAPP Stimulates Guanine Nucleotide Exchange on Ypt1p but not Ypt32p | We previously reported that TRAPP is an exchange factor for Ypt1p . Subsequently, TRAPP was also reported to be an exchange factor for Ypt32p . In these earlier studies, partially purified preparations of TRAPP were assayed for exchange activity. Thus, it is formally possible that one or both of these activities is the consequence of a contaminant in these preparations. To directly compare the exchange activity of TRAPP on Ypt1p and Ypt32p, we purified TRAPP by tandem affinity purification from a yeast strain containing TAP-tagged Trs33p. Briefly, the two forms of TRAPP were recovered from extracts by affinity absorption onto IgG-Sepharose beads. TEV protease was then added to release the bound material, and the eluate was incubated with calmodulin-coated agarose beads. This second affinity step removed the TEV protease as well as other contaminating proteins. TAP-tagged Trs33p yielded chemically pure TRAPP that was previously shown on a silver-stained gel to only contain TRAPP subunits . To determine the nucleotide exchange activity of this highly purified preparation of TRAPP, a GTPgammaS uptake assay was performed. TRAPP, immobilized on calmodulin agarose beads, was incubated with GTPgammaS in the presence of recombinant Ypt1p or Ypt32p, and the amount of GTPgammaS that bound to protein was measured. Although TRAPP stimulated the uptake of GTPgammaS onto Ypt1p, it did not appreciably stimulate the uptake of GTPgammaS onto Ypt32p . These results, together with previously published results showing that TRAPP does not stimulate nucleotide exchange on Sec4p , imply that TRAPP is a specific guanine nucleotide exchange factor for Ypt1p. The Depletion of TRAPP from Cytosol Abolishes the Ypt1p but not Ypt32p GDP Release Activity | To determine if TRAPP is the only exchange factor for Ypt1p, we depleted TRAPP from cytosol and then assayed for GDP release activity. For these studies, cytosol was prepared from a strain (SFNY 1088) in which the DSS4 gene was disrupted, and the sole copy of Bet3p was tagged with Protein A (PrA). Dss4p, a putative chaperonin for the nucleotide-free form of Rabs, stimulates the dissociation of GDP from both Ypt1p and Sec4p . TRAPP was depleted from cytosol by affinity absorption onto IgG-Sepharose beads. As Bet3p is found in TRAPP I and TRAPP II , both forms of the complex were precipitated onto the beads. To completely deplete cytosol of TRAPP, we found it necessary to treat cytosol three times with IgG-Sepharose beads. Each incubation was done with a fresh batch of beads. The amount of TRAPP remaining in the cytosol was detected by Western blot analysis using anti-Trs33p antibody (Figure C). Quantification of the data indicated that similar99% of the TRAPP was removed from the cytosol. Because Trs33p is present in both forms of TRAPP, this result confirms that TRAPP I and TRAPP II were depleted from the cytosol that we used to assay exchange activity. Figure 2 | Depletion of TRAPP abolishes the Ypt1p exchange activity from cytosol. Depletion of TRAPP abolishes the Ypt1p exchange activity from cytosol. TRAPP was depleted by incubating IgG-Sepharose beads with cytosol prepared from a strain in which Bet3p is Protein A tagged and the DSS4 gene is deleted. Samples containing 0.2 mg of mock-treated or 0.5 mg of IgG-Sepharose --treated cytosol were resolved by SDS-PAGE and analyzed by Western blot analysis with alpha-Trs33p serum (C). TRAPP-depleted or mock-treated cytosol was incubated with 5 pmol of Ypt1p (A) or Ypt32p (B) preloaded with [3H]GDP. At the time intervals indicated, radioactivity that bound to protein was determined by filter-binding, and the data are expressed as the percentage of label bound to Ypt1p or Ypt32p. The intrinsic rate of [3H]GDP release from Ypt1p or Ypt32p was measured in the presence of BSA. TRAPP-depleted or mock-treated cytosol (treated with Sepharose beads) was then assayed for its ability to displace GDP from Ypt1p and Ypt32p. Recombinant Ypt1p, or Ypt32p, preloaded with [3H]GDP was incubated with the depleted cytosol, and the radioactivity that remained bound to protein was counted. The intrinsic rate of [3H]GDP release was measured in the presence of BSA. Although TRAPP stimulated dissociation on Ypt1p is concentration dependent, the affinity depletion of TRAPP abolished this activity (Figure A). In contrast, the GDP release activity on Ypt32p remained unchanged (Figure B). We conclude that TRAPP is the major exchange factor for Ypt1p, but not Ypt32p. The Ypt32p Exchange Factor Is Associated with the P100 Fraction | To begin to characterize the Ypt32p exchange factor, differential fractionation experiments were performed. Cell lysates were centrifuged at 12,000 x g to generate supernatant (S12) and pellet (P12) fractions. The S12 was centrifuged at 100,000 x g, and the GDP release activity of the S100 (supernatant) and P100 (pellet) fractions were compared with that of the lysate, S12 and P12 fractions. When equivalent amounts of protein from each of these fractions were assayed, the P100 fraction was found to be enriched in Ypt32p GDP release activity (Figure A). Treatment of the P100 fraction, prepared from SFNY 1088, with 0.5 M NaCl released the activity from membranes (Figure B). This release factor had exchange activity as it stimulated the uptake of GTPgammaS onto Ypt32p (Figure A). Figure 3 | The Ypt32p exchange factor is associated with the P100 fraction. The Ypt32p exchange factor is associated with the P100 fraction. (A) A cell lysate was prepared from SFNY26-3a and fractionated as described in MATERIALS AND METHODS. Cell fractions or BSA (1 mg/ml) were incubated with [3H]GDP-Ypt32p for various times at 30C. For each time point, the data are expressed as the percentage of label bound to Ypt32p. (B) The P100 fraction, prepared from SFNY 1088, was treated with 0.5 M NaCl or no salt and centrifuged at 100,000 x g for 1 h. The supernatant (s) and pellet (p) were assayed for their ability to stimulate the release of [3H]GDP from Ypt32p. The intrinsic rate of [3H]GDP release from Ypt32p was measured in the presence of BSA, and the value obtained was subtracted as background. Figure 4 | Depletion of TRAPP does not affect the Ypt32p exchange activity. Depletion of TRAPP does not affect the Ypt32p exchange activity. (A) The P100 fraction of SFNY1088 was treated with buffer (control) or 0.5 M NaCl, centrifuged, and assayed as described below. To deplete TRAPP, the salt-treated supernatant (2.4 mg) was incubated with 20 mul of packed IgG-Sepharose or mock-treated with Sepharose 4B as described in MATERIALS AND METHODS. The TRAPP-depleted or mock-treated sample was incubated with [35S]GTPgammaS for 30 min at room temperature in the presence or absence of (His6)-Ypt32p. Samples were then incubated with 20 mul of packed Nickel-nitrilotriacetic acid-agarose (Ni-NTA) beads for 1 h at 4C, and GTPgammaS uptake was measured. The intrinsic rate of GTPgammaS uptake onto Ypt32p measured in the presence of BSA, and the value obtained in the absence of (His6)-Ypt32p was subtracted as background. (B) A P100 fraction, prepared from SFNY1088, was treated with buffer (control) or 0.5 M NaCl and centrifuged. The supernatant was incubated with IgG-Sepharose or mock treated with Sepharose 4B. The supernatants (s) and pellets (p) as well as the TRAPP-depleted and mock-treated samples were assayed at 30C for 30 min for their ability to stimulate the release of [3H]GDP from Ypt1p. The intrinsic rate of [3H]GDP release from Ypt1p was measured in the presence of BSA, and the value obtained was subtracted as background. (C) TRAPP-depleted (0.5 mg) and mock-treated (0.3 mg) samples were resolved by SDS-PAGE and analyzed by Western blot analysis using antibodies directed against Trs120p, Bet3p, Trs33p, Trs31p, Trs20p, and Bet5p. TRAPP is also efficiently extracted from membranes with 0.5 M NaCl ( and Figure B). To determine if the Ypt32p exchange factor we are characterizing is TRAPP, the salt extract of the P100 fraction was depleted of this complex by treatment with IgG-Sepharose beads. These fractions were then assayed for Ypt1p or Ypt32p exchange activity and blotted for the presence of TRAPP subunits. The IgG-Sepharose --treated fraction was devoid of Ypt1p exchange activity (Figure B), whereas the Ypt32p exchange activity was unchanged (Figure A). Furthermore, antibodies directed against Trs120p, Bet3p-PrA, Trs33p, Trs31p, Trs20p, and Bet5p detected each of these subunits in mock-treated but not IgG Sepharose --treated fractions (Figure C). Trs120p is only present in the TRAPP II complex, whereas Bet3p, Trs33p, Trs31p, Trs20p, and Bet5p are present in both forms of the complex . Antibodies recognizing the remaining TRAPP components were not of sufficient titer to detect subunits in the mock-treated sample. These findings demonstrate that the Ypt32p GDP release factor we are characterizing is an exchange factor for Ypt32p. Furthermore, this exchange factor is not TRAPP. The Ypt32p Exchange Factor Is a Putative Effector of Ypt1p | Genetic studies have shown that the overexpression of YPT31 or YPT32 suppresses the dominant YPT1-D124N mutation, which fails to bind guanine nucleotides . Additionally, we found that the overexpression of YPT32 suppresses the trs130ts2 mutant (our unpublished data). As cells defective in trs130 have decreased amounts of active Ypt1p , we reasoned that the exchange factor for Ypt32p might be an effector of Ypt1p. To test this possibility, we purified Ypt1p as a recombinant GST fusion protein on glutathione Sepharose beads. The beads containing Ypt1p were loaded with either GDP or GTPgammaS and incubated with a yeast lysate. As controls, immobilized GST-Ypt51p or GST were preloaded with GTPgammaS and incubated with lysate. Ypt51p is a small GTP-binding protein that regulates membrane traffic on the prevacuolar/endosomal pathway . The beads were washed and assayed for Ypt32p exchange activity. The Ypt1p-GTPgammaS coated beads accelerated the release of GDP from Ypt32p (Figure A) and the uptake of GTPgammaS onto Ypt32p (Figure B). This activity was not dependent on TRAPP, as the Ypt32p GDP release activity was unaffected when TRAPP-depleted cytosol was incubated with the Ypt1p-GTPgammaS coated beads (our unpublished data). These results imply that the Ypt32p exchange factor is an effector of Ypt1p. This factor preferentially binds to the GTP-bound form of Ypt1p and is specifically recruited by Ypt1p, but not Ypt51p. Furthermore, this exchange factor appears to be specific because it does not stimulate the release of GDP from Ypt1p (Figure C). Figure 5 | The Ypt32p exchange factor is a putative effector of Ypt1p. The Ypt32p exchange factor is a putative effector of Ypt1p. (A) Lysate, prepared from a wild-type yeast strain, was incubated with beads that contain GST-Ypt1p, GST-Ypt51p, or GST. The beads were either preloaded with GTPgammaS or GDP. The treated beads were incubated with [3H]GDP-Ypt32p at 30C for various periods of time. [3H]GDP that bound to protein was measured by by filter-binding, and the data are expressed as the percentage of label bound to Ypt32p. (B) Beads containing 0.8 mg of GST-Ypt1p-GTPgammaS, GST-Ypt1p-GDP, or GST were incubated with a yeast lysate (800 mg) and then assayed for 30 min at room temperature in the presence of 1.6 nmol of (His6)-Ypt32p and 32 nmol of [35S]GTPgammaS. The reaction was stopped by the addition of 1 ml of ice-cold stop buffer. The beads were spun, and the supernatant was incubated with 25 mul of packed Ni-NTA agarose beads for 1 h at 4C. The Ni-NTA beads were washed three times with 1 ml of stop buffer, and the amount of GTPgammaS that bound to the beads was measured by filter-binding. The intrinsic uptake of GTPgammaS onto Ypt32p was measured in the presence of immobilized GST, and the value obtained was subtracted as background. (C) TRAPP-depleted cytosol was incubated with GST-Ypt1p-GTPgammaS immobilized on beads. The beads were then incubated with [3H]GDP-Ypt32p or [3H]GDP-Ypt1p for 20 min at 30C. The intrinsic rate of [3H]GDP release from Ypt32p or Ypt1p was measured in the presence of immobilized GST, and the value obtained was subtracted as background. Uso1p Is not an Exchange Factor for Ypt32p | In general, small GTP-binding proteins have many effectors . To date, only Uso1p has been implicated as a putative effector of Ypt1p. Uso1p, a large protein that contains coiled coil and globular domains, plays a key role in tethering COP II vesicles to the Golgi complex . The proposal that Uso1p is an effector of Ypt1p is based on two observations. First, its orthologue p115 was recently shown to be an effector of Rab1 . Second, the extraction of Ypt1p from membranes by GDI was found to decrease the association of membrane-bound Uso1p . These findings prompted us to test the possibility that Uso1p is a Ypt1p effector with exchange activity on Ypt32p. Uso1p was purified from a strain harboring a 2-mum plasmid that contains a myc-tagged version of Uso1p. Cytosol prepared from this strain was loaded onto a Mono Q column and eluted with a linear salt gradient. The peak of Uso1p was further purified on a Superdex-200 gel filtration column, and the fractions containing Uso1p were pooled and concentrated. The cytosol, Mono Q column flow-through and purified Uso1p were assayed for exchange activity on Ypt32p. Ypt32p preloaded with [3H]GDP was incubated with these fractions, and the radioactivity that bound to protein was measured using a filter binding assay. Although Uso1p was retained on the Mono Q column and none could be detected in the flow-through (Figure B), the flow-through and cytosol had the same Ypt32p GDP release activity (Figure A). Consistent with this observation, purified Uso1p did not display any exchange activity on Ypt32p (Figure A). These findings clearly demonstrate that Uso1p is not an exchange factor for Ypt32p. Figure 6 | Uso1p is not a nucleotide exchange factor for Ypt32p. Uso1p is not a nucleotide exchange factor for Ypt32p. Uso1p was purified from a yeast cytosol using Mono Q ion exchange and Superdex-200 gel filtration columns. [3H]GDP dissociation from Ypt32p was measured by filter-binding for the indicated times in the presence of either 2 mg/ml cytosol, Mono Q column flow-through, purified Uso1p or BSA. Data are expressed as the percentage of label bound to Ypt32p (A). Samples containing 0.25 mg of cytosol or Mono Q column flow-through were resolved by SDS-PAGE and analyzed by Western blot analysis with 9E10 antibody (B). DISCUSSION : Previous studies have shown that preparations of partially purified TRAPP stimulate guanine nucleotide exchange on two different small GTP-binding proteins, Ypt1p and Ypt32p. Although partially purified TRAPP stimulated Ypt1p nucleotide exchange activity robustly, very little Ypt32p exchange activity was observed . These findings suggested that the reported Ypt32p exchange activity was either insignificant or was due to a contaminant. To resolve this issue and to definitively prove that TRAPP is an exchange factor for Ypt1p, we assayed chemically pure amounts of TRAPP for Ypt1p and Ypt32p exchange activity. Earlier work demonstrated that both forms of the TRAPP complex (TRAPP I and TRAPP II) are exchange factors for Ypt1p . Chemically pure amounts of TRAPP were prepared from a strain in which Trs33p, a subunit that is present in both forms of the complex, was TAP tagged. Purified TRAPP was found to have Ypt1p, but not Ypt32p, exchange activity. If TRAPP is not an exchange factor for Ypt32p, lysates depleted of TRAPP I and TRAPP II should still retain Ypt32p exchange activity. Approximately 99% of the TRAPP was depleted from lysates prepared from a strain in which the sole copy of Bet3p is fused to Protein A. Although these lysates no longer had Ypt1p exchange activity, their Ypt32p exchange activity was unaffected. These results imply that TRAPP I and TRAPP II may be the only exchange factors for Ypt1p. Furthermore, they clearly demonstrate that TRAPP is not a major exchange factor for Ypt32p. In an earlier study from another group, TRAPP was reported to have comparable Ypt32p and Ypt1p exchange activities . purified TRAPP from a strain in which Bet3p was overproduced and fused to GST. However, when we assayed purified, GST-tagged TRAPP in the presence of 5 pmol of [3H]GDP-Ypt1p, the specific exchange activity (504 pmol/min/mg) was found to be approximately twofold less than PrA-tagged TRAPP (962 pmol/min/mg). In our hands, very little Ypt32p exchange activity was observed with either preparation of TRAPP and was the same as previously reported . Genetic experiments have suggested a functional relationship between Ypt1p and Ypt32p. The overexpression of either YPT31 or YPT32 was found to suppress the growth defect of the dominant YPT1-D124N mutation . Furthermore, we isolated YPT31 and YPT32 as high copy suppressors of the trs130ts2 mutant, which is defective in the activation of Ypt1p. These suppression studies suggested that the Ypt32p exchange factor may be an effector of Ypt1p. The finding that a factor that stimulates nucleotide exchange on Ypt32 will specifically bind to Ypt1p in its GTP-bound form supports this hypothesis. This Ypt1p effector is not Uso1p and appears to be novel. Our findings imply that the activated form of Ypt1p can influence the activity of Ypt32p by recruiting the Ypt32p exchange factor. Furthermore, our data suggest that Ypt1p and Ypt31p/Ypt32p may interact in a signal cascade that directs traffic to and through the Golgi complex. Interestingly, in a parallel study it was shown that activated Ypt32p recruits Sec2p, the exchange factor for Sec4p. Sec4p is the small GTP-binding protein that regulates membrane traffic from the Golgi to the plasma membrane . Thus, one may speculate that the activated form of each Rab may recruit the exchange factor that activates the Rab that functions at the next stage of membrane traffic. It will be important to see if the activated forms of other small GTP-binding proteins in the Rab family do the same. Backmatter: PMID- 12221138 TI - Purification and Identification of Secernin, a Novel Cytosolic Protein that Regulates Exocytosis in Mast Cells AB - After permeabilization with the pore-forming toxin streptolysin-O mast cells can be triggered to secrete by addition of both calcium and a GTP analogue. If stimulation is delayed after permeabilization, there is a progressive decrease in the extent of secretion upon stimulation, eventually leading to a complete loss of the secretory response. This loss of secretory response can be retarded by the addition of cytosol from other secretory tissues, demonstrating that the response is dependent on a number of cytosolic proteins. We have used this as the basis of a bioassay to purify Secernin 1, a novel 50-kDa cytosolic protein that appears to be involved in the regulation of exocytosis from peritoneal mast cells. Secernin 1 increases both the extent of secretion and increases the sensitivity of mast cells to stimulation with calcium. Keywords: INTRODUCTION : Mast cells are secretory cells found on the mucosal and serosal surfaces of tissues throughout the body where they are involved in the allergic response . The cells can be activated by the cross-linking of high-affinity IgE receptors by antigen-specific IgE that leads to activation of phospholipase C, generating IP3 that subsequently causes a release of calcium from cytosolic stores and thus triggers secretion . The mast cells secrete a variety of inflammatory mediators, including histamine, from granules that contain many lysosomal markers . The presence of lysosomal markers in the secretory granules of mast cells and other secretory cells of hemopoietic lineage (such as basophils, cytotoxic T cells, natural killer cells, neutrophils, eosinophils, and macrophages) has led to the suggestion that these secretory granules are not derived from the classical secretory pathway, but are derived from the lysosomal pathway . Permeabilization of secretory cells with detergents such as digitonin, bacterial toxins such as Streptolysin-O (SLO) or by mechanical disruption allows exocytosis to be triggered by the addition of buffered calcium solutions, bypassing the need for activation of a triggering receptor and associated signal transduction events . In mast cells an absolute requirement for guanine nucleotides has been reported, unlike neuroendocrine cells where guanine nucleotides appear to have a modulatory role . This has led to the identification of a number of GTP binding proteins including Gialpha3 , betagamma subunits , rac , rho , and cdc42 , which all appear to regulate secretion in mast cells. Although the final fusion event is mediated by the SNAP/NSF/SNARE system in both mast cells , neurons and neuroendocrine cells , it is possible that the origin of these secretory granules accounts for the differences observed in the regulation of secretion in these cells when compared with the more commonly studied neuroendocrine systems. After permeabilization secretory cells leak cytosolic proteins, leading to a progressive loss of responsiveness to Ca2+ and nucleotides. This loss of response or rundown can be slowed by the provision of exogenous cytosol and has been used as the basis of a bioassay in mast cells , chromaffin cells , and PC12 and GH3 cells to purify cytosolic proteins that regulate secretion from these cells. We have previously shown that a rac/RhoGDI complex isolated from bovine brain cytosol partially restores secretory responsiveness in Streptolysin-O --permeabilized mast cells . A number of other activities were partially purified during the purification of the rac/rhoGDI complex. Here we report the full purification and identification of a second brain cytosolic protein, Secernin 1 (Secern is an archaic English term for secrete), which is also capable of regulating exocytosis in permeabilized mast cells. MATERIALS AND METHODS : Frozen bovine brains were purchased from First Link UK (Brierley Hill, West Midlands, UK). Male Sprague Dawley rats were purchased from B&K Universal Ltd. (Hull, UK). GTP-gamma-S and bovine serum albumin was purchased from Roche Diagnostics Ltd. (Lewes, East Sussex, UK). Streptolysin-O (Murex formulation) was purchased from Corgenix Biotech Limited (Temple Hill, Dartford, Kent, UK). Ceramic hydroxyapatite column was purchased from (Hemel Hempstead, UK), and all other chromatography columns were purchased from Amersham-Pharmacia (Amersham, UK). Donkey anti-rabbit horseradish peroxidase (HRP) antibody was purchased from Amersham-Pharmacia. The GATEWAY cloning system and ThermalAce DNA polymerase kit was obtained from Invitrogen, Life Technologies (Paisley, UK). The human cDNA clone of the gene KIAA0193 was obtained from the Kazusa DNA Research Institute (Kisararazu, Chiba, Japan) and inserted in the pBluescript SK+ vector. All other chemicals used were of the highest quality available from standard commercial sources. Experimental Procedures | Secretion Measurements. | Cells were obtained by peritoneal lavage of male Sprague Dawley rats (>300 g), and mast cells were purified to >98% purity by centrifugation through Percoll as previously described . Cells, suspended in assay buffer (137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, 20 mM piperazine-N,N'-bis(2-ethanesulfonic acid), pH 6.8) supplemented with 1 mg/ml bovine serum albumin (BSA) were incubated with metabolic inhibitors (0.6 mM 2-deoxyglucose and 10 muM antimycin A) for 5 min at 37C and then cooled to ice temperature and added to SLO (1.6 IU/ml) in the presence of 0.1 mM EGTA. After 5 min, cells were washed free of unbound SLO and contaminating impurities by dilution and centrifugation at 4C. Permeabilization and hence rundown of the secretory response was initiated by transferring the cells to prewarmed (37C) assay buffer containing 1 mg/ml BSA, 0.3 mM Ca/EGTA buffer (10 nM Ca2+), 100 muM MgATP, and proteins under test in 96-well microtiter plates. After allowing predetermined times for rundown (generally between 5 and 20 min), the cells were stimulated to secrete by addition of solutions containing Ca/EGTA buffers formulated to regulate 10 muM Ca2+(or 100 nM Ca2+ for controls) to a final concentration of 3 mM and GTP-gamma-S to a final concentration of 100 muM (or zero for controls) with sufficient MgATP to maintain the concentration at 100 muM. After 20 min the reactions were quenched by addition of ice-cold buffer supplemented with EGTA (10 mM), and the cells were sedimented by centrifugation. The supernatants were sampled for measurement of secreted hexosaminidase as previously described . Calcium/EGTA buffers were prepared by mixing solutions of EGTA and end-point --titrated CaEGTA made up at identical concentrations and adjusted to pH 6.8, according to a computer program, as previously described . Secretion is expressed as the percent of total cellular hexosaminidase released, calibrated by reference to appropriate reagent blanks and the total cell content released by 0.1% Triton X-100. Stimulated secretion is calculated as the difference in the amount of hexosaminidase released in response to 100 nM Ca2+ or 10 muM Ca2+ + 100 muM GTPgammaS. All determinations were carried out in quadruplicate unless otherwise stated. Purification of Secernin 1. | All chromatography was carried out on a Biologic liquid chromatography system at 4C. Frozen bovine brains, 500 g, were thawed at 4C before homogenization in a Waring blender in 1 liter homogenization buffer (137 mM NaCl, 2.3 mM KCl, 1 mM MgCl2, 1 mM EGTA, 1 muM Pepstatin, 1 muM Leupeptin, 0.1 mM PMSF, 0.02% NaN3, 20 mM Pipes, pH 6.8). The homogenate was then centrifuged for 10 h at 10,000 x g at 4C in a fixed angle rotor. This cytosol extract was subjected to ammonium sulfate precipitation and an active fraction between 60 and 90% (NH4)2SO4 was resuspended in 40 ml homogenization buffer. The active material was subjected to chromatography on Octyl Sepharose FF as previously described , and the active fractions were combined. All column fractions under test were buffer exchanged into assay buffer using NAP-5 columns before assay. DEAE Chromatography. | The active fractions from the Octyl Sepharose column were combined and desalted into buffer A (20 mM diethanolamine, 0.02% NaN3, pH 8.7) in aliquots of 5 ml on a HiPrep 26/10 desalting column (Pharmacia). The desalted material was loaded, using on line dilution via the pump, at 20% protein with 80% buffer A, onto a DEAE Sepharose column (XK26/50, 100 ml, Pharmacia) that had been equilibrated with buffer A. The column was then washed with 36 ml 20 mM diethanolamine, pH 8.7, before elution with a linear gradient of 0 --40% buffer B (1 M NaCl, 20 mM diethanolamine, 0.02% NaN3, pH 8.7) over 372 ml followed by a final elution in 100% buffer B over 120 ml. The column was run at 5 ml/min, and 8-ml fractions were collected. The fractions were assayed for activity, and the active fractions (45 --52) on the third peak were combined. Hydroxyapatite Chromatography. | Active fractions (45 --52) from the third peak of the DEAE chromatography were concentrated to 5 ml on a 50-ml Amicon pressure concentrator (43-mm YM10 membrane) and then desalted into buffer C (50 mM MES, 0.02% NaN3, pH 6.0) on a HiPrep 26/10 desalting column. This was then loaded onto a ceramic hydroxyapatite column (Econo-Pac CHT-II, 1 ml, ) preequilibrated with buffer C. The column was then washed with 2 ml buffer C before elution with a linear gradient of 0 --100% buffer D (500 mM NaCl, 50 mM MES, 0.02% NaN3, pH 6.0) over 16 ml. The column was then washed with an additional 2 ml buffer D, before elution with a 0 --100% linear gradient of buffer E (500 mM KH2PO4, 50 mM MES, 0.02% NaN3, pH 6.0) over 16 ml and finally washed with an additional 4 ml of buffer E. The column was run at 1 ml/min, and 1-ml fractions were collected. The fractions were assayed for activity, and the active fractions (35 --38) in the second peak were combined. Phenyl Superose Chromatography. | The combined fractions from peak 2 of the hydroxyapatite column were diluted with 3.4 M (NH4)2SO4, buffered with 50 mM NaH2PO4, 0.02% NaN3, pH 7.5, to produce a final concentration of 2 M (NH4)2SO4. The protein was then applied to a Phenyl Superose column (Pharmacia, HR5/5, 1 ml) equilibrated in buffer F ((NH4)2SO4, 50 mM NaH2PO4, 0.02% NaN3, pH 7.5) and the column washed with 5 ml 100% buffer F before eluting with a 30 --80% gradient of buffer G (50 mM NaH2PO4, 0.02% NaN3, pH 7.5) over 30 ml, followed by a final 5 ml of 100% buffer G. The column was run at 0.4 ml/min, and 1-ml fractions were collected. Superose 12 Chromatography. | Active fractions from the Phenyl Superose (21 --23) were concentrated to 240 mul on a 10K Microsep centrifugal concentrator (Filtron, Northborough, MA) at 4C and injected onto a Superose 12 column (Pharmacia, HR10/30, 24 ml) equilibrated in homogenization buffer. The column was run at 0.2/ml min, and 0.5-ml fractions were collected. The active fractions were combined and concentrated before use in secretion experiments. Leakage of Secernin. | Purified mast cells were treated with diisopropyl fluorophosphate (2 mM) for 10 min at 4C. The cells were treated with SLO at ice temperature as described above, resuspended at similar1 x 106cells/ml, and then permeabilized by bringing the temperature to 37C. Samples of cells, 100 mul, were removed at intervals and sedimented by centrifugation at 14,000 x g, and the supernatants were harvested. Ice-cold acetone was added to the supernatants to a final concentration of 80%, and the mixture was maintained at -20C for 2 h after which the aggregated proteins were sedimented by centrifugation at 14,000 x g. These were taken up in Laemmli sample buffer and separated on 12% SDS-polyacrylamide gels. Proteins were transferred to nitrocellulose using a wet blot method and probed for Secernin using the polyclonal anti-Secernin antibodies, SK1147 and SK1148, at a dilution of 1:4000. Antibody binding was detected using a donkey anti-rabbit HRP-linked secondary antibody at 1:2000 dilution and an ECL detection kit. Plasmid Construction. | Oligonucleotide primers for the amplification of the KIAA0193 gene were designed with attB1 or attB2 sites for the insertion into the GATEWAY donor vector pDONR201 (Life Technologies) by homologous recombination. Primers with the following sequences were synthesized by MWG, Inc.: KIAA0193 (forward), 5'- GGGGACAAGTTTGTACAAAAAAGCAGGCTTCG A A GGAGATAGAACCATGATAAGCAGACCCGCCTGGCTCT-3'; KIAA 0193 (reverse), 5'-GGGGACCACTTTGTACAAGAAAGCTGGGTC-CTATCACTTAAAGAACTTAATCTCCGTG-3'. The primers were used to generate the attB PCR product using a ThermalAce DNA polymerase kit (Invitrogen) from the pBluescript SK+ vector containing KIAA0193. The PCR products were cloned into pDONR201, and the resulting plasmid pENTR KIAA0193 were used to transfer the gene sequences into pDEST15 (N terminal GST fusion) or pDEST17 (N terminal His fusion) via homologous recombination. The corresponding plasmids, pEXP15 KIAA0193 and pEXP17 KIAA0193, were used for overexpression of the fusion proteins in Escherichia coli BL21-SI. Recombinant Protein Purification. | Cells carrying either pEXP15 KIAA0193 or pEXP17 KIAA0193 were grown overnight in LB broth without NaCl in the presence of 100 mug/ml ampicillin at 30C. The cells were then diluted 1 in 10 in prewarmed LB broth without NaCl, in the presence of 100 mug/ml ampicillin at 30C, and grown to an optical density of 0.6 at 600 nm. Protein expression was induced by the addition of NaCl to 0.3 M, and the cells were grown for a further 3 h. Cells were harvested by centrifugation, and inclusion bodies were purified by the method of . Affinity Purification of Antibody SK1147. | Inclusion bodies of the recombinant GST-tagged KIAA 0193 (human Secernin 1) were solubilized in SDS-sample buffer and separated on 10% SDS-PAGE. After blotting onto nitrocellulose paper the recombinant protein was visualized with Ponceau S staining, and the bands were excised. The antibody SK1148 was affinity purified against this protein by the method of . The antibody was dialyzed overnight against homogenization buffer before use in neutralization experiments. Immunoneutralization of Secernin. | Polyclonal anti-Secernin antibody SK1147 or preimmune serum was diluted 1:50 into assay buffer and added an equal volume of either purified 60 mug/ml Secernin or 6 mg/ml freshly prepared rat brain cytosol in assay buffer. The antibody and proteins were incubated at 4C for 30 min before addition to permeabilized mast cells at a final concentration of 10 mug/ml Secernin and 1 mg/ml cytosol as described above. Varying concentrations of affinity-purified SK1147 was incubated with 3 mg/ml cytosol in the presence or absence of 5 mg/ml inclusion body containing recombinant human Secernin 1-His fusion protein. After incubation for 60 min at 4C, the cytosol was centrifuged at 14,000 x g to remove the inclusion body before addition to permeabilized mast cells at a final concentration of 1 mg/ml cytosol and antibody as indicated. Protein Assay. | Protein concentration was assayed by the method of using BSA as a standard. Protein Analysis. | Purity of protein samples was assessed by electrophoretic separation on 12% SDS-polyacrylamide gels and detection by silver staining . Production of Antisera. | Two polyclonal rabbit antisera (SK1147 and SK1148) were raised against purified Secernin 1 by Abcam Ltd. (Cambridge, UK), using a 30-mug initial injection followed by three booster injections of 30 mug. Mass Spectrometric Analysis. | p50 was alkylated with iodoacetamide in sample buffer (Novex, Encinitas, CA) and run on a 4 --12% SDS-PAGE gel with a MOPS running buffer system (Novex). The gel was stained with Sypro Orange (Molecular Probes, Eugene, OR), and the p50 band was excised and digested with 12.5 mug/ml modified trypsin (Roche) in 20 mM NH4CO3. A proportion of the sample was analyzed by MALDI-TOF MS, and the tryptic peptide ions were searched against NCBI and Swiss Prot databases using the MS-FIT search algorithm from Protein Prospector (UCSF, San Francisco, CA). The remainder of the sample was chromatographed on a 150 x 0.075-mm Pepmap C18 capillary column coupled to an LC Packings Ultima HPLC system (Dionex, Camberley, UK). The column was equilibrated with 2% acetonitrile/0.1% formic acid in water at 0.2 mul/min, and developed with a gradient of acetonitrile/0.1% formic acid. The outlet of the column was connected to a Micromass Q-TOF2 mass spectrometer, equipped with a nanoflow source, and peptide ions were automatically submitted for ms/ms fragmentation. Spectra from ms/ms experiments were interpreted, and the sequences were searched against NCBI nr and dbEST databases using the BLAST search algorithm. Spectral data was also searched against the same databases using the Sonar ms/ms search algorithm . RESULTS : Bovine brain cytosol was prepared and fractionated by ammonium sulfate precipitation followed by Octyl Sepharose chromatography as previously described . The purification of Secernin 1 described in MATERIALS AND METHODS is summarized in Table . Activity from cytosol prepared from frozen bovine brains is only detectable after the Octyl Sepharose column, unlike cytosol from freshly isolated rat brains . The purification is therefore calculated from the pooled activity from this column. Figure A shows that the activity eluted from the Octyl Sepharose can be separated by a DEAE Sepharose column into three distinct peaks of activity by a gradient of NaCl. Peak 1 was found to contain the previously purified rac and rhoGDI, as assessed by Western blotting, so further purification of this peak was not undertaken. Peak 3 appeared to have the highest activity and was therefore subjected to further purification. Table 1 | Purification of secernin 1 Figure 1 | Purification of secernin. Purification of secernin. (A) DEAE chromatography. Active fractions from Octyl Sepharose were desalted and loaded onto a DEAE Sepharose column, the activity was eluted with a rising gradient of NaCl. (B) Hydroxyapatite chromatography. Pooled fractions from peak 3 on the DEAE column were desalted and applied to a ceramic hydroxyapatite column. The activity was initially eluted with a rising gradient of NaCl followed by a second gradient of KH2PO4. (C) Phenyl Superose chromatography. Pooled fractions from peak 2 on the hydroxyapatite column were adjusted to 2 M (NH4)2SO4 and applied to a Phenyl Superose column. The activity was eluted with a falling gradient of (NH4)2SO4. All column activities were assayed for protein concentration and stimulated secretion as described in the MATERIALS AND METHODS. Data shown are mean +- SEM (n = 4); similar results were obtained on at least 10 occasions. Some error bars are smaller than symbols used. The combined fractions of peak 3 from the DEAE column were applied to a hydroxyapatite column and eluted with a rising gradient of NaCl, followed by a rising gradient of KH2PO4 as shown in Figure B. A small peak of activity elutes with NaCl and a larger second activity elutes with KH2PO4. This second activity peak was pooled, applied to a Phenyl Superose column and eluted by a decreasing gradient of (NH4)2SO4 as shown in Figure C. Fractions from the column were then subjected to SDS-PAGE analysis, a single protein of 50 kDa appears to correlate with the activity from the phenyl superose column. To confirm the correlation, the pooled fractions from the Phenyl Superose column were concentrated and applied to a Superose 12 gel filtration column. The fractions were assayed for their ability to retard the rundown of exocytosis in mast cells and subjected to SDS-PAGE, and again a single 50-kDa protein was found to correlate with the activity, as shown in Figure . Figure 2 | Superose 12 chromatography. Superose 12 chromatography. (A) Active fractions from the Phenyl Superose column were concentrated and loaded onto a superose 12 gel filtration column and assayed for protein concentration and stimulated secretion as described in the MATERIALS AND METHODS. Data shown are mean +- SEM (n = 4); similar results were obtained on at least 10 occasions. Some error bars are smaller than symbols used. (B) Active fractions were analyzed by silver-stained 12% SDS-PAGE. To identify the 50-kDa protein, the purified material was subjected to SDS-PAGE and in-gel trypsinization. The peptide mixture was recovered and subjected to MALDI-TOF mass spectrometry, and the resultant mass fingerprint used to interrogate sequence databases. This analysis identified p50 as a protein corresponding to a previously cloned mouse cDNA (GenBank accession no. ) of unknown function . To confirm the identity of p50, the balance of the peptide mixture was chromatographed by reverse-phase HPLC, and individual peptides were delivered to an online Q-TOF mass spectrometer for ms/ms fragmentation and consequent peptide sequence identification. A total of seven peptides were sequenced (Figure B, bold). Of these, three corresponded to peptides previously identified by mass fingerprint, and each showed sequence identity with the putative protein encoded by . One peptide, not identified in the initial mass fingerprint also was identical to the mouse sequence, whereas three additional peptides were closely related, showing >75% homology with it. We conclude that p50 is a novel protein, which we have named Secernin 1, which is encoded by the bovine orthologue of mouse . Figure 3 | Identification of p50 by MALDI-TOF and Q-TOF mass spectrometry. Identification of p50 by MALDI-TOF and Q-TOF mass spectrometry. A tryptic digest of p50 excised from a SDS-PAGE gel was subjected either to fingerprint analysis by MALDI-TOF mass spectrometry or reverse-phase HPLC and online Q-TOF ms/ms fragmentation. (A) MALDI-TOF mass spectrum annotated with the sequence and molecular mass (m/z) of identified peptides (highlighted in gray). Peptides whose sequence was deduced by Q-TOF fragmentation analysis are boxed. (B) The predicted polypeptide sequence encoded by mouse cDNA . Peptide sequences identified by MALDI fingerprint analysis above are highlighted in gray. Peptides whose sequence was deduced by MS/MS fragmentation (in bold font) are aligned. || indicates a match, + indicates a conserved substitution, and (.) indicates a mismatch between the sequences. The outlined area represents the predicted coiled-coil region. Figure shows the effect of increasing concentrations of purified Secernin 1 on the secretion of hexosaminidase from permeabilized mast cells. The optimal concentration of Secernin 1 is 7.1 +- 0.9 mug/ml with an EC50 of 2.21 +- 0.2 mug/ml (mean +- SEM, n = 4), which is comparable to other proteins previously demonstrated to regulate secretion in these cells . This effect on secretion is unlikely to be due to nonspecific protein effects because neither boiled Secernin 1 at 10 mug/ml nor BSA up to 1 mg/ml had any effect on secretion from these cells (O'Sullivan, unpublished observations). Figure shows the effect of the protein at various times of rundown, the protein initially has no effect, but as the cytosol leaks from the cells, there is an increasing effect, which in turn declines as other proteins leak from the cell. However, Secernin 1 is incapable of completely preventing rundown. Figure 4 | Secernin 1 dose response. Secernin 1 dose response. Mast cells permeabilized with SLO in the presence of varying concentrations of Secernin 1. After 15 min the cells were stimulated by addition of solutions containing EGTA buffered to 10 muM Ca2+ plus GTPgammaS (final concentration, 100 muM) or 100 nM Ca2+. After a further 20-min incubation, the cells were sedimented by centrifugation, and the supernatants were sampled for analysis of secreted hexosaminidase. Stimulated secretion is calculated as the difference in the amount of hexosaminidase released in response to either 100 nM Ca2+ or 10 muM Ca2+ + 100 muM GTPgammaS. Data shown are mean +- SEM (n = 4); similar results were obtained on four occasions. Some error bars are smaller than symbols used. Figure 5 | Effect of Secernin on rundown of secretory response. Effect of Secernin on rundown of secretory response. Mast cells permeabilized with SLO in the presence (circle) and absence of Secernin 1 at a final concentration of 3.5 mug/ml. At times indicated, samples were removed and stimulated by transfer to solutions containing EGTA buffered to 10 muM Ca2+ plus GTPgammaS (final concentration, 100 muM) or 100 nM Ca2+. After a further 20-min incubation, the cells were sedimented by centrifugation, and the supernatants were sampled for analysis of secreted hexosaminidase. Stimulated secretion is the difference in the amount of hexosaminidase released in response to either 100 nM Ca2+ or 10 muM Ca2+ + 100 muM GTPgammaS. Data shown are mean +- SEM (n = 4); similar results were obtained on four occasions. Some error bars are smaller than symbols used. To determine that the protein is present in mast cells, a rabbit polyclonal antibody SK1147 was generated against the purified protein. Figure A shows a Western blot of whole mast cells demonstrating a single immunoreactive band at 50 kDa; a similar result was obtained with a second polyclonal antibody SK1148. The results shown in Figure are consistent with a model in which Secernin 1 leaks slowly from the permeabilized cell and the exogenously added Secernin 1 replaces the lost protein. To test this hypothesis, supernatants were collected from cells permeabilized over increasing periods. The supernatants were precipitated by acetone at -20C, separated by SDS-PAGE, and analyzed by Western blot. Figure B shows that the antibody detects an increasing amount of the 50-kDa protein in the supernatants with time after permeabilization, confirming that the protein is cytosolic and leaks from the cell after permeabilization with a similar time course to the rundown of the secretory response. Figure 6 | Presence of Secernin in mast cell cytosol. Presence of Secernin in mast cell cytosol. (A) Western blot of intact mast cells. After transfer to nitrocellulose membranes, the proteins were detected by probing with the rabbit polyclonal antibody SK1147 and a donkey anti-rabbit HRP secondary antibody visualized by ECL. (B) The permeabilized mast cells were sampled at the times indicated and rapidly sedimented. Proteins present in the supernatants were precipitated and Western blotted. Proteins were detected with antibody SK1147. To confirm that that the effects of Secernin 1 were due to the 50-kDa protein the anti-Secernin antibody, SK1147 was used to immunoneutralize the purified protein. Figure A shows that the antibody blocks the effect of the optimal dose of Secernin 1, whereas same concentration of the preimmune sera has no significant effect. This concentration of antibody is also inhibits the effect of freshly prepared rat brain cytosol by 50%, indicating that Secernin is responsible for a significant amount of the activity found in whole cytosol. To confirm that this effect was due to the interaction of SK1147 with Secernin, the antibody was affinity purified against recombinant human Secernin 1. Figure B demonstrates that the affinity-purified antibody inhibits fresh rat brain cytosol and that this inhibition can be blocked by inactive recombinant human Secernin 1 in an inclusion body. Figure 7 | Inhibition of Secernin in cytosol by anti-Secernin antibody. Inhibition of Secernin in cytosol by anti-Secernin antibody. (A) Mast cells permeabilized with SLO in the presence of 10 mug/ml purified Secernin 1 or 1 mg/ml rat brain cytosol, which had been immunoneutralized with SK1147 antisera, preimmune serum, or sham neutralized with buffer. After 15 min the cells were stimulated by addition of solutions containing EGTA buffered to 10 muM Ca2+plus GTPgammaS (final concentration, 100 muM) or 100 nM Ca2+After a further 20 min incubation, the cells were sedimented by centrifugation, and the supernatants were sampled for analysis of secreted hexosaminidase. Stimulated secretion is the difference in the amount of hexosaminidase released in response to either 100 nM Ca2+ or 10 muM Ca2+ + 100 muM GTPgammaS. Data shown are mean +- SEM (n = 4); similar results were obtained on four occasions. (B) Mast cells permeabilized with SLO in the presence or absence of 1 mg/ml rat brain cytosol that had been immunoneutralized varying concentrations of affinity-purified SK1147 antisera, in the presence and absence of inclusion bodies containing recombinant human Sescernin 1. After 15 min the cells were stimulated by addition of solutions containing EGTA buffered to 10 muM Ca2+plus GTPgammaS (final concentration, 100 muM) or 100 nM Ca2+. After a further 20-min incubation, the cells were sedimented by centrifugation, and the supernatants were sampled for analysis of secreted hexosaminidase. Stimulated secretion is the difference in the amount of hexosaminidase released in response to either 100 nM Ca2+ or 10 muM Ca2+ + 100 muM GTPgammaS. Data shown are mean +- SEM (n = 4); similar results were obtained on four occasions. Two possible mechanisms for Secernin 1 to increase the extent of secretion would be to increase either the rate of secretion or to allow secretion to occur over a much longer period. Figure clearly demonstrates that the addition of Secernin 1 to the cells during rundown, before stimulation, increases the rate of secretion from the permeabilized mast cells. If the data are normalized to the response at 25 min, as shown in the inset to Figure , it becomes clear that although the rate of secretion is increased by Secernin 1, the time course of the secretory response is unaffected by the presence of Secernin 1. Figure 8 | Effect of Secernin on the time course of secretion. Effect of Secernin on the time course of secretion. Mast cells permeabilized with SLO in the presence or absence of 3 mug/ml purified Secernin 1. After 15 min the cells were stimulated by addition of solutions containing EGTA buffered to 10 muM Ca2+plus GTPgammaS (final concentration, 100 muM). After a further 20-min incubation, the cells were sedimented by centrifugation, and the supernatants were sampled for analysis of secreted hexosaminidase. Data shown are mean +- SEM (n = 3); similar results were obtained on four occasions. Some error bars are smaller than symbols used. The inset figure shows the same data normalized to 100% response for each of the conditions after 25 min secretion. We have previously shown that the sensitivity of the mast cells to both Ca2+ and GTPgammaS declines as a consequence of the rundown of the secretory response . Therefore, we investigated the possibility that Secernin 1 not only increased the rate and extent of secretion, but also had an effect on the sensitivity of the mast cells to stimulation with both Ca2+ and GTPgammaS. Figure A shows the effect of Secernin on the GTPgammaS sensitivity of cells in the presence of 0.1 and 10 muM Ca2+. Under these rundown conditions, all concentrations of GTPgammaS in the presence of 0.1 muM Ca2+ are incapable of inducing secretion, whether Secernin 1 is present or not. In the presence of 10 muM Ca2+ Secernin 1 increases the extent of secretion by the same amount, independent of the concentration of GTPgammaS. Figure B clearly demonstrates that Secernin 1 partially restores the Ca2+ sensitivity of the rundown mast cells, with some Ca2+-dependent secretion even occurring in the absence of GTPgammaS. Figure 9 | (A and B) Effect of Secernin on the sensitivity of mast cell secretion to Ca2+ and GTPgammaS. (A and B) Effect of Secernin on the sensitivity of mast cell secretion to Ca2+ and GTPgammaS. Mast cells permeabilized with SLO in the presence (filled symbols) or absence (open symbols) of 3 mug/ml Secernin 1. After 15 min the cells were stimulated by addition of solutions containing the following: (A) EGTA buffered to either 0.1 muM (circle, ) or 10 muM Ca2+ plus varying concentrations of GTPgammaS or (B) EGTA buffered to varying concentrations of Ca2+ plus a final concentration of either 0 or 100 muM GTPgammaS . After a further 20-min incubation, the cells were sedimented by centrifugation, and the supernatants were sampled for analysis of secreted hexosaminidase. Data shown are mean +- SEM (n = 4); similar results were obtained on four occasions. Some error bars are smaller than symbols used. DISCUSSION : We have purified a novel cytosolic protein, termed Secernin 1, which regulates exocytosis in permeabilized rat peritoneal mast cells. Database analysis reveals that fragments of Secernin 1 identified by mass spectrometry show 93.5% identity and 95.7% homology with the predicted product of mouse cDNA , which has a predicted MW of 46296, which is consistent with our estimated molecular weight of 50 kDa from SDS-PAGE. Database searching indicates that Secernin 1 is encoded by one of a small family of related genes . The human genome contains three Secernin genes (termed Secernins 1 through 3), which are localized on chromosomes 7 (7p14.3-p14.1), 17 (17q21.3), and 2 (2p14-q14.3), respectively. Analysis of cDNAs corresponding to human Secernin 2 indicates the existence of at least two splice variants (2a and 2b), one of which generates a protein containing a truncated C terminus. Comparison of human and mouse Secernin 1 cDNAs suggests that the former has a truncated N-terminus lacking similar60 amino acids. However, it is also possible that this is a splice variant of a larger protein, because inspection of the 5' noncoding region of the existing cDNA as well as the genomic DNA sequence indicates the presence of a putative exon, which is highly homologous to the relevant 5' coding sequence of mouse Secernin 1. We have also identified two partial cDNA sequences in the TIGR bovine gene index database. These correspond to a bovine homologue of human Secernin 2a . The mouse genes are ubiquitously expressed , which would imply some common role in many cell types rather than a protein specifically involved in mast cell exocytosis. Figure 10 | Identification of Secernin family members in mammalian species. Identification of Secernin family members in mammalian species. The protein sequence for mouse Secernin 1 was used to interrogate the NCBI Human Genome and nonredundant databases using the BLAST search program. Sequences were aligned using the PIMA sequence alignment program and formatted using BOXSHADE. Amino acid identities are shown as black boxes. Conservative changes are shown as shaded boxes. The GenBank accession numbers for Secernins are as follows: hsSES1, ; hsSes2a, XP_054038.1; hsSes2b, XP_032208.1; hsSes3, XP_035198.1; mmSes1, ; mmSes3, ; BtSes2a, TC 80787, Tc80786. The full-length sequence for Secernin 1 shows no significant degree of homology with any protein known to be involved in exocytosis, membrane fusion events, or intracellular signaling. The protein contains no known domain structures apart from a small 21 amino acid region between 358 and 378, which is predicted to form a coiled-coil domain by the COILS program . It is therefore possible that the actions of the protein may be mediated by a direct interaction with another protein. Two rabbit polyclonal antibodies raised against purified bovine Secernin 1 recognize a single immunoreactive band at 50 kDa in mast cells, confirming the presence of Secernin 1 in mast cells. After permeabilization this protein leaks from the mast cells and can be precipitated from the supernatant, confirming that the protein is cytosolic. Secernin 1 leaks from the cell at a rate similar to the decrease in secretory response of the mast cells during rundown, implying that the leakage of Secernin 1 is a major cause of the loss of secretory response. Addition of Secernin 1 to permeabilized cells followed by stimulation within the first 5 min has no effect on secretion, demonstrating a lack of effect of the protein until sufficient Secernin 1 has leaked from the cells to impair secretion. The data are consistent with the hypothesis that exogenously added Secernin 1 is replacing the endogenous protein leaking from the permeabilized cells, thus enhancing the secretory response under these conditions. The addition of Secernin 1 alone is not capable of blocking the loss of secretory response but slows the rate of loss of response. When cells are permeabilized and incubated at 37C before stimulation, there is a decline in sensitivity to both Ca2+ and guanine nucleotide. Secernin 1 partially restores Ca2+ sensitivity, but not sensitivity to guanine nucleotide. We have shown that although Secernin 1 increases the extent of secretion from mast cells, the time course of secretion remains the same. If we assume that any granule that fuses with the plasma membrane releases either all or a constant proportion of its hexosaminidase, then the simplest explanation of this data is that Secernin 1 causes the recruitment of additional secretory granules to the site of exocytosis in a calcium-dependent manner. Alternatively Secernin 1, in the presence of calcium, may be acting to increase the granule swelling, core expulsion and breakdown observed in fused granules , thus increasing the release of hexosaminidase from the cells. The loss of the secretory response is due to a number of protein factors, including Secernin 1, Rac/RhoGDI, and the other partially purified activities shown in Figure , and it is likely that the full reconstitution of the secretory response will require a large number of different proteins. This is confirmed by the finding that inhibition of Secernin in cytosol by immunoneutralization blocks similar50% of the recovery of secretion, indicating that although Secernin is not the only cytosolic component that regulates mast cell secretion, it appears to be an important component of the response. Cytosolic proteins that are capable of regulating exocytosis in permeabilized cell assays appear to fall into four major categories: 1) proteins that directly interact with the fusion machinery, including alpha-SNAP ; 2) GTPases involved in intracellular signaling, such as Arf , rac , rho , and cdc42 ; 3) proteins involved in other intracellular signaling pathways, such as PKC and 14 --3-3 ; and 4) proteins involved in the regulation of PIP2, such as PI kinases and PITP . Our data clearly demonstrate that the cytosolic protein Secernin 1 has a major role in the regulation of exocytosis in mast cells, but further work is required to determine which or if any of these pathways is the site of action of Secernin 1. Backmatter: Abbreviations used: : BSA = bovine serum albumin GTP-gamma-S = guanosine5'-3-O-(thio) triphosphate SLO = streptolysin-O PMID- 12221139 TI - Role of LAMP-2 in Lysosome Biogenesis and Autophagy AB - In LAMP-2 --deficient mice autophagic vacuoles accumulate in many tissues, including liver, pancreas, muscle, and heart. Here we extend the phenotype analysis using cultured hepatocytes. In LAMP-2 --deficient hepatocytes the half-life of both early and late autophagic vacuoles was prolonged as evaluated by quantitative electron microscopy. However, an endocytic tracer reached the autophagic vacuoles, indicating delivery of endo/lysosomal constituents to autophagic vacuoles. Enzyme activity measurements showed that the trafficking of some lysosomal enzymes to lysosomes was impaired. Immunoprecipitation of metabolically labeled cathepsin D indicated reduced intracellular retention and processing in the knockout cells. The steady-state level of 300-kDa mannose 6-phosphate receptor was slightly lower in LAMP-2 --deficient hepatocytes, whereas that of 46-kDa mannose 6-phosphate receptor was decreased to 30% of controls due to a shorter half-life. Less receptor was found in the Golgi region and in vesicles and tubules surrounding multivesicular endosomes, suggesting impaired recycling from endosomes to the Golgi. More receptor was found in autophagic vacuoles, which may explain its shorter half-life. Our data indicate that in hepatocytes LAMP-2 deficiency either directly or indirectly leads to impaired recycling of 46-kDa mannose 6-phosphate receptors and partial mistargeting of a subset of lysosomal enzymes. Autophagic vacuoles may accumulate due to impaired capacity for lysosomal degradation. Keywords: INTRODUCTION : Lysosomes are acidic membrane-bound organelles containing hydrolytic enzymes for degradation of proteins, lipids, nucleic acids, and polysaccharides. Lysosomal enzymes are synthesized in the endoplasmic reticulum and sorted in the trans-Golgi network (TGN) by mannose 6-phosphate receptors (MPRs). MPRs bind the mannose 6-phosphate tag of lysosomal enzymes in the trans-Golgi network (TGN), and the receptor-ligand complexes are transported to endosomes in clathrin-coated vesicles. In endosomes ligands dissociate from the MPRs due to the acidic pH, and receptors may then recycle back to TGN. A small proportion of newly synthesized lysosomal enzymes is secreted to the extracellular medium . Mammalian cells contain two different MPRs, the cation-independent or 300-kDa MPR, and the cation-dependent or 46-kDa MPR. Both receptors are needed for efficient intracellular retention of lysosomal enzymes . In addition MPRs also recycle between early endosomes and the plasma membrane. However, only MPR300 is able to mediate endocytosis of exogenous mannose 6-phosphate containing ligands . The limiting membrane of the lysosomal compartment has multiple functions. It is responsible for acidification of the interior, sequestration of the active lysosomal enzymes , and transport of degradation products from the lysosomal lumen to the cytoplasm . The lysosomal membranes contain several highly N-glycosylated proteins including LAMP-1 and LAMP-2. These two glycoproteins are structurally similar and evolutionary related . Like LAMP-1, LAMP-2 is composed of a large luminal portion, which is separated by a proline-rich hinge region in two disulphide containing domains, a single transmembrane-spanning segment, and a short cytoplasmic tail of 11 amino acids . The latter contains a Gly-Tyr motif critical for transport to lysosomes . LAMP-2 is one of the major carriers for poly-N-acetyllactosamines in cells . Although the ubiquitously expressed LAMP-2 is localized primarily in the late endosomes and lysosomes , under certain circumstances, e.g., after platelet activation, during granulocytic differentiation and activation, in malignant carcinoma cells, and in cytotoxic T lymphocytes, it is also found at the cell surface . LAMP-2 has also been described as a receptor for the selective import and degradation of cytosolic proteins in the lysosome, or chaperone-mediated autophagy . Autophagy is a central mechanism in cellular metabolism that cells use to degrade parts of their cytoplasm and organelles using lysosomal enzymes. The autophagic-lysosomal pathway is known to play an important role in the cellular protein economy . In hepatocytes exposed to nutrient starvation it can account for as much as three quarters of the overall protein degradation . The first step in autophagy is segregation of cytoplasm by a membrane cisterna, which forms a double or multiple membrane-bound vacuole called the autophagosome. Autophagosomes acquire lysosomal membrane proteins, vacuolar proton pumps, and acid hydrolases, presumably by fusing with endosomes/lysosomes. Finally, the contents are degraded by lysosomal enzymes . The autophagic pathway is subject to complex regulation, it is activated in mammalian cells by amino acid deprivation . LAMP-2 deficiency leads to premature postnatal death of about half of all LAMP-2 --deficient mice . In several LAMP-2 --deficient tissues, including muscle, heart, pancreas, and liver, an accumulation of autophagic vacuoles was observed. We also observed a reduced contractile function of the heart muscle. Thus, LAMP-2 --deficient mice represent a valuable animal model for autophagic vacuolar myopathy, or Danon disease, a human disease associated with a mutated LAMP-2 gene . Here we extend the phenotype analysis of these mice using cultured hepatocytes. MATERIALS AND METHODS : Primary Antibodies | The following antibodies were used in this study: Rabbit antiserum against mouse cathepsin D , rabbit antiserum against rat liver cathepsin D purified from lysosomal contents by affinity chromatography on pepstatin A-sepharose as described , affinity-purified rabbit antibody against the cytoplasmic tail of MPR46 (MSC1; ), rabbit antiserum against human MPR46 (II-4; ), rabbit antiserum against rat MPR300 (I-5; ), mouse mAb against gamma-adaptin (Transduction Laboratories, Lexington, KY), and rat monoclonal antibodies against LAMP-2 (ABL93) and LAMP-1 (1D4B; Developmental Studies Hybridoma Bank, Iowa City, Iowa). Preparation of Primary Mouse Hepatocytes | Mouse hepatocytes were prepared from 3- to 6-month-old mice according to a described procedure . Hepatocytes were enriched using a Percoll gradient (58% wt/vol) and plated on collagen-coated or -uncoated plastic culture dishes. Unless stated otherwise, the cells were cultured overnight in RPMI-1640 containing 10% fetal calf serum (FCS) and penicillin/streptomycin (Life Technologies, Rockville, MD) before the experiments. Endocytic Uptake | Six-nanometer gold particles were prepared and coated with bovine serum albumin (BSA). BSA-gold was dialyzed against RPMI-1640 medium and diluted to serum-free RPMI containing 0.2% BSA. After uptake the cells were rinsed and fixed for electron microscopy. Immunofluorescence | Hepatocytes were grown on glass coverslips. The cells were fixed with cold methanol or 4% paraformaldehyde in 0.1 M HEPES, pH 7.4, and permeabilized with 0.5% saponin or 0.1% Triton X-100. The cells were immunostained using antibodies against MPR46. The primary antibodies were detected with goat anti-rabbit IgG conjugated with Texas red or fluorescein (Dianova GmbH, Hamburg, Germany). After embedding in Mowiol (Calbiochem-Novabiochem GmbH, Bad Soden, Germany) containing DABCO, fluorescence was examined using a confocal laser-scanning microscope (LSM 2; Zeiss, Oberkochen, Germany). Western Blotting | Expression of cathepsin-D, MPR46, and MPR300 was analyzed in hepatocyte homogenates (cathepsin D) and membrane fractions (receptors). Hepatocytes were homogenized in Tris-buffered saline (wt/vol; 1:9) at 4C using an Ultra-Turrax, and the homogenate was analyzed for protein content. To obtain membrane fractions cell pellets were resuspended in TBS including proteinase inhibitors, subjected to sonication (3 times, 200 s), and pelleted at high speed (100,000 x g). The resulting pellet was resuspended in TBS/proteinase inhibitors/1% Triton X-100 and subjected to sonication (3 times, 20 s). One hundred micrograms of protein was subjected to SDS-PAGE (5% polyacrylamide in case of MPR300; 10% in case of MPR46 and cathepsin D) under reducing conditions. Proteins were transferred to a polyvinylidene difluoride membrane (Schleicher & Schull, Dassel, Germany), which was subsequently blocked with 10 mM PBS, pH 7.4, 0.05% Triton X-100, 5% milk powder (blocking buffer) for 1 h at 37C. The blot was incubated overnight at 4C with rabbit anti-mouse cathepsin-D, anti-MPR46 MSC1, or anti-MPR300 serum. Membranes were washed six times for 5 min in 10 mM PBS, pH 7.4, 0.1% Tween 20. Subsequently, incubation with horseradish peroxidase --coupled anti-rabbit antibody was performed for 1 h at room temperature followed by washing six times for 5 min in 10 mM PBS, pH 7.4, 0.1% Tween 20. Blots were finally analyzed using the ECL Detection System (Amersham Pharmacia Biotech, Piscataway, NJ). Quantification was performed by densitometry (Scan Jet 4c/T; Hewlett-Packard, Palo Alto, CA; WinCam 2.2). Metabolic Labeling and Immunoprecipitation | Hepatocytes were incubated in methionine-free medium for 1 h and then labeled with 35S-methionine/cysteine (Amersham Life Science, Inc., Rockville, MD) in the same medium containing 5% dialyzed FCS. During the chase, the medium was supplemented with 0.25 mg/ml l-methionine/l-cysteine. Immunoprecipitation from cells and media was carried out as described previously with rabbit antibodies against cathepsin-D, MPR46 (II-4), or MPR300. Densitometric quantification of the bands was done with a phosphoimager (Fuji, Stamford, CT) and the program MacBas. Incubation of Cells with Radioactive Glucosylthioceramide and thioGM3 | An aliquot of a stock solution of the desired glucosylthioceramide and thioGM3 in methanol was dried under a stream of nitrogen. The dried lipid was dissolved by first adding 20 mul of ethanol and then 0.75 ml of RPMI containing 5 mg of defatted BSA under vigorous stirring. The resulting solution was diluted with 6.85 ml of RPMI to yield a 10 muM lipid-BSA complex. Hepatocytes seeded on 6-cm dishes were washed in RPMI and incubated with the [14C]C8-Glc-S-Cer/BSA complex or the [14C]thioGM3/BSA complex in RPMI for 3 h at 37C, washed with PBS, and further incubated for 24 h at 37C in RPMI containing 5% heat-inactivated FCS. The incubation media were saved, and the cells were washed with PBS, harvested with a rubber policeman, and centrifuged at 2000 x g for 10 min. For protein determination, the pellet was suspended in 0.4 ml of H2O, and 5-mul aliquots were assayed. The lipids were extracted with 4 ml of chloroform/methanol (1:1, by volume) for 3 h at 38C. The lipid extracts and the media were desalted according to . Total radioactive lipids of both cells and media were determined by liquid scintillation counting of aliquots. The lipid extracts of cells and media were analyzed by TLC using chloroform/methanol/15 mM calcium chloride (60:35:8, by volume) as the developing solvent. In addition, TLC plates were exposed to x-ray film (Kodak X-Omat XAR-5; Eastman Kodak, Rochester, NY). Northern Blotting | Total RNA of cultured hepatocytes was prepared using the Qiagen Rneasy system (Hilden, Germany). Ten micrograms of total RNA were separated in a formaldehyde agarose gel and processed as described . Filters were hybridized with a MPR46 and MPR300 cDNA probe and a beta-actin probe (, Palo Alto, CA). Hybridization and washing of the filters were performed as described . Electron Microscopy and Autophagosome Quantification | Isolated hepatocytes were fixed in 2% glutaraldehyde in 0.2 M HEPES, pH 7.4, at room temperature for 2 h. The cells were scraped off the culture dish, pelleted, and postfixed in 1% osmium tetroxide in 0.1 M phosphate buffer for 1 h. The cells were dehydrated in ethanol and embedded in Epon. For estimation of volume fractions by stereology , 20 --25 micrographs, primary magnification x10,000, were taken with systematic random sampling from each sample. The cytoplasmic volume fraction of autophagic vacuoles was estimated by point counting. Autophagic vacuoles were classified as early, containing morphologically intact cytoplasm, and late, containing partially degraded but identifiable cytoplasmic material . Statistical significance was estimated using Student's t test. Immunogold Electron Microscopy and Quantitation | Cultured hepatocytes were fixed in 4% paraformaldehyde in 0.2 M HEPES, pH 7.4, at room temperature for 2 h and then stored in 2% paraformaldehyde for up to 4 d. Some samples were fixed by adding 0.1% glutaraldehyde to the initial fixative. The cells were then embedded in gelatin and processed for cryosectioning as described . Cryosections were picked up with 2.3 M sucrose or a mixture of sucrose and 2% methyl cellulose. The sections were labeled with rabbit antibodies against MPR46 (MSC1), mouse anti --gamma-adaptin, or rat anti --LAMP-1 or --LAMP-2. The primary antibodies were detected with goat anti-rabbit-IgG, goat anti-rat-IgG, or goat anti-mouse IgG coupled to 5 or 10 nm gold (British BioCell, Cardiff, UK). Double labelings were done by mixing two primary antibodies, from different species, and the corresponding secondary antibodies, conjugated to different sized gold particles. The smaller gold size was used for the less abundant antigen. Density of LAMP-2 labeling on the limiting membranes of autophagic vacuoles was estimated as described . To quantitate the distribution of MPR46 labeling, the sections were systematically screened under the electron microscope, and each time a gold particle was seen, it was recorded into a certain cellular compartment, as explained in Table , or into an unrecognizable compartment. To estimate MPR46 labeling in endosome associated tubular and vesicular structures , LAMP-1 --positive multivesicular endosomes of 200 --800-nm diameter (Figure , B --D) were systematically scored. Gold particles situated inside and outside, within a distance of 300 --400 nm of the endosomes, were counted. At least 40 endosomal profiles, from two independent samples, were counted, and the results were expressed as a ratio of gold inside/gold outside. Enzyme Activities | Hepatocytes were isolated and cultured overnight. After washing with PBS the cells were incubated in RPMI medium containing 5% heat-inactivated FCS. After 6, 12, and 24 h, samples of the medium were removed and cells were washed, scraped, and homogenized in TBS/1% Triton X-100. Lysosomal enzyme activities were detected using fluorometric and colorometric assays as described . To avoid a possible effect of the medium change leading to nonspecific increase of lysosomal enzyme activities, the enzyme activity in the medium after 6-h culture was withdrawn from the specific activities measured in the medium after 12 and 24 h. In each experiment the enzyme activities were measured from two parallel culture dishes. Statistical significance was evaluated using Student's t test. RESULTS : We observed accumulation of both early autophagic vacuoles (Avi), containing morphologically intact cytoplasm, and late autophagic vacuoles (Avd), containing partially degraded cytoplasmic material, in liver and cultured hepatocytes of LAMP-2 --deficient mice . We performed immunogold labeling of control hepatocytes to examine the presence of LAMP-2 in autophagic vacuoles and detected labeling in both Avi and Avd (for Avi see Figure ). Quantitation of LAMP-2 labeling on the limiting membrane showed that the labeling in Avd was five times higher than that in Avi. This enrichment of LAMP-2 in Avd is similar to that observed for LAMP-1 . Figure 1 | LAMP-2 localizes to autophagic vacuoles in control hepatocytes. LAMP-2 localizes to autophagic vacuoles in control hepatocytes. Isolated hepatocytes were cultured overnight, fixed, and processed for cryosectioning. Thin frozen sections were immunogold labeled for LAMP-2 (10 nm gold) and cathepsin-D (5 nm gold). LAMP-2 localizes to the limiting membrane of the autophagic vacuole, which can be identified by its cytoplasmic contents (a mitochondrion) as an Avi. Cathepsin-D is found in the space between the limiting membrane and the contents of the vacuole. The arrow indicates a vesicle, containing both LAMP-2 and cathepsin-D, that is likely to be in the process of fusion with the autophagic vacuole. Bar, 200 nm. Half-life of Autophagic Vacuoles | Long-lived protein degradation is decreased in LAMP-2 --deficient hepatocytes , suggesting that autophagic vacuoles accumulate because of impaired degradation of autophagocytosed material. To clarify this, we determined the half-life of autophagic vacuoles using quantitative electron microscopy. Isolated hepatocytes were incubated in starvation medium (free of serum and amino acids) for 5 h to induce autophagy. Subsequently the fate of the autophagic vacuoles was followed by shifting the cells to FCS containing medium, supplemented with 10 mM 3-methyladenine (3MA) to prevent formation of new autophagosomes . Samples for electron microscopy were taken before the starvation, after 5-h starvation, and 1- and 3-h chase in full medium with 3MA, as indicated in Figure . The volume fractions of Avi, Avd, and endosomes/lysosomes were estimated using stereology. The latter compartment included all endo/lysosomal vesicles lacking morphologically identifiable cytoplasmic material, i.e., early and late endosomes and lysosomes. In control cells (Figure , A and B) the volume fraction of Avi and Avd increased similar3.8-fold (p = 0.0012) and 1.7-fold (p = 0.078), respectively, during 5-h starvation. After 1-h chase in full medium with 3MA, almost all Avi had disappeared from the control cells (p = 0.000015 compared with the 5-h starvation). This suggests that Avi matured into Avd and these in turn degraded their contents. During the 3-h chase the size of Avd decreased in control cells (p = 0.047 compared with the 5-h starvation), suggesting further degradation of the cytoplasmic material. Figure 2 | Half-life of autophagic vacuoles in control and LAMP-2 --deficient hepatocytes. Half-life of autophagic vacuoles in control and LAMP-2 --deficient hepatocytes. The cells were cultured overnight and then starved for serum and amino acids for 5 h to induce autophagy. Subsequently the cells were shifted to serum and amino acid --containing culture medium (full medium), including 10 mM 3MA. Samples for quantitative electron microscopy were taken as indicated in D. Cytoplasmic volume fractions of early (Avi, A) and late (Avd, B) autophagic vacuoles and endo/lysosomes (C) were estimated by point counting. The results are the mean +- SEM from two independent experiments; a minimum of 40 micrographs were quantitated for each time point. Statistical significance compared with the previous time point on the same graph: *, 0.00122 < p < 0.042; ***, p = 0.000015; o, 0.05 < p < 0.079. Statistical significance compared with the 5-h time point on the same graph: &, p = 0.08311; section, p = 0.04716. Consistent with our previous findings , the volume fractions of both Avi and Avd were much higher in LAMP-2 --deficient hepatocytes. Also the volume fraction of endo/lysosomes was twice as high as in the control cells (Figure , A --C). Starvation did not increase the volume fraction of total Avi/Avd/endo/lysosomal pool in LAMP-2 --deficient cells, suggesting that autophagy cannot be further stimulated by starvation. However, we observed a small increase in Avd volume fraction (p = 0.041) on the expense of endo/lysosomes (p = 0.051). During the 1-h chase in full medium and 3MA, less than half of the Avi volume fraction was consumed (p = 0.016), indicating retarded conversion of Avi into Avd. Approximately corresponding increase was observed in the volume fraction of endo/lysosomes (p = 0.0029), which suggests initial maturation of some Avd into lysosomes. During the 3-h chase no further decrease of Avi volume fraction was observed, whereas the volume of Avd tended to increase (p = 0.08311 compared with the 5-h starvation). This increase in Avd may point to fusion with plasma membrane derived endosomes, which would increase the volume fraction similar to maturation of Avi into Avd. The results indicate that in LAMP-2 --deficient hepatocytes the consumption of Avd is severely retarded and that the half-lives of both Avi and especially Avd are prolonged. Fusion of Endosomes with Autophagic Vacuoles | In LAMP-2 --deficient hepatocytes the pH in Avi and Avd is only slightly higher than in control cells and the limiting membranes of Avd contain substantial amounts of LAMP-1 . This suggests that transport of membrane proteins such as the vacuolar proton pump and LAMP-1 to Avi and/or Avd does take place. To directly demonstrate the convergence of endocytic and autophagic pathways we used 6-nm gold particles coated with BSA (BSA-gold) as a fluid phase endocytic tracer. To check that the knockout hepatocytes are able to perform fluid-phase endocytosis at rates comparable to that of the control cells, the cells were fed with horseradish peroxidase for 10 --30 min and washed, and the peroxidase activity was assayed from cell homogenates. The results revealed that the rate of fluid-phase uptake is comparable in LAMP-2 --deficient and control hepatocytes. BSA-gold was then fed to cells for 1 h, and the cells were chased for 2 h. Electron microscopy was used to score the distribution of gold particles in endo/lysosomes and autophagic vacuoles. Typical fusion profiles of Avd and multivesicular endosomes in LAMP-2 --deficient cells are presented in Figure , A and B. In control cells, similar20% of gold particles were found in Avd, compared with >40% in LAMP-2 --deficient hepatocytes (Figure C). The larger fraction of gold in Avd in LAMP-2 --deficient cells may be due to the prolonged half-life of these organelles, resulting in trapping of the tracer within Avd. The result indicates that multivesicular endosomes are able to fuse with autophagic vacuoles in LAMP-2 --deficient hepatocytes. Figure 3 | Endocytosed BSA-gold is transported to autophagic vacuoles in LAMP-2 --deficient cells. Endocytosed BSA-gold is transported to autophagic vacuoles in LAMP-2 --deficient cells. Cells were fed with BSA-gold for 1 h, chased in full culture medium for 2 h, and subsequently fixed and processed for electron microscopy. (A and B) Typical fusion profiles of multivesicular endosomes and autophagic vacuoles in LAMP-2 --deficient cells. The arrowheads indicate the BSA-gold, which is associated with small vesicles. Autophagic vacuoles were identified by their cytoplasmic contents: electron dense, partially degraded ribosomes or rough ER (Avd) or a mitochondrion (Avi). No BSA-gold was found in early autophagic vacuoles. Bars, 200 nm. (C) Distribution of the endocytic tracer between autophagic vacuoles and endo/lysosomes (end/lys) was estimated by systematically screening the thin sections. A total of 1469 and 2186 gold particles were screened from control and LAMP-2 -/- cells, respectively. Trafficking of Lipids between the Plasma Membrane, Lysosomes, and the Golgi | We studied the uptake of the nondegradable glycosphingolipid glucosylthioceramide ([14C]C8-Glc-S-Cer). A fraction of the internalized glucosylthioceramide is transported to the Golgi as indicated by its elongation to GM2 and GM3 by Golgi resident glycosyltransferases . The glucosylthioceramide was taken up by control and LAMP-2 --deficient hepatocytes. In both genotypes Golgi-specific glycosylation products (e.g., GM2 and GM3) were detected to a comparable level in cells and media (Figure , lanes 1 --4). The appearance of glycosylation products in culture media containing FCS is due to transport of the glycosylation products from the Golgi apparatus to the plasma membrane and to the extraction of the semitruncated glycolipids by protein components of the FCS. Figure 4 | Postendocytotic glycosylation of [14C]C8-Glc-S-Cer and of catabolites of thioganglioside [14C]C8-SGM3 in control and LAMP-2 --deficient hepatocytes. Postendocytotic glycosylation of [14C]C8-Glc-S-Cer and of catabolites of thioganglioside [14C]C8-SGM3 in control and LAMP-2 --deficient hepatocytes. Mouse hepatocytes were incubated with [14C]C8-Glc-S-Cer (lanes 1 --4) and [14C]C8-S-GM3 (lanes 6 --9) as described in MATERIALS AND METHODS. After 3-h labeling the cells were chased for another 24 h. The chase media were saved, and the cells were harvested. The desalted lipids of cells and chase media were separated by TLC using chloroform/methanol/15 mM calcium chloride (60:35:8, by volume) as a developing system. The radioactive lipids were visualized by exposure to x-ray --sensitive film. Lane 5, reference lipids from top to bottom: 1, [14C]C8-Glc-S-Cer; 2, [14C]C8-Lac-S-Cer; 3, [14C]C8-GbOse3Cer; 4, [14C]C8-S-GM3; 5, [14C]C8-S-GM2; 6, [14C]C8-S-GM1; 7, [14C]C6-GD1a with their relative mobilities denoted by standard abbreviations for GSL as explained in . In addition the mobilities of N-glycolyl neuraminic acid containing thiogangliosides GM3 and GM2 denoted as NeuGc-GM3 and NeuGc-GM2 are indicated, respectively. Their endogenous counterparts are the predominant gangliosides in murine hepa-tocytes and are formed by sialylation of lactosylceramide with CMP-NeuGc as the activated sugar. GM3 cannot be transformed directly into NeuGc-GM3 or NeuGc-GM2. Arrowheads indicate the labeled glycosphingolipids that were fed to the hepatocytes. Having shown that hepatocytes allow efficient uptake and trafficking of labeled sphingolipids, we studied the metabolism of the partially degradable thioganglioside ([14C]S-GM3). This compound allows to follow the trafficking of the internalized thioganglioside to endosomes/lysosomes, where it is degraded to glycosylthioceramide, and the trafficking of glucosylthioceramide to the Golgi where it is elongated to GM2 and GM3. GM2 and GM1 originating from elongation of the added GM3 contain neuraminic acid. In contrast, the GM3, GM2, and GM4 originating from glucosylthioceramide generated from the added GM3 in the lysosomes contain the hepatocyte derived N-glycolyl-neuraminic acid and are thereby distinguishable . The uptake, accumulation, and degradation in lysosomes and the subsequent transport of the labeled lipids to the Golgi apparatus and elongation to N-glycolyl neuraminic acid containing thiogangliosides GM3 and GM2 were similar in LAMP-2 --deficient and control hepatocytes (Figure , lanes 6 --9). This indicates that transport of lipids from the plasma membrane to lysosomes and from there to the Golgi apparatus is not affected by LAMP-2 deficiency. Activities and Trafficking of Lysosomal Enzymes | The prolonged half-life of autophagic vacuoles could result from impaired degradation of their content. Impaired acidification as a cause for compromised lysosomal degradation has already been excluded . We therefore determined the activities of lysosomal enzymes in isolated hepatocytes and liver homogenates as well as the trafficking and maturation of newly synthesized cathepsin D, a major lysosomal proteinase. The activities of three representative lysosomal hydrolases were differentially affected in LAMP-2 --deficient hepatocytes. Intracellular beta-mannosidase activity of LAMP-2 --deficient hepatocytes was decreased to similar14% of control (p = 0.022), beta-glucuronidase was unchanged, and beta-hexosaminidase seemed to be increased about twofold, although this increase was not statistically significant . Interestingly, the activity of beta-mannosidase was only moderately affected in liver homogenates of LAMP-2 --deficient mice (0.43 +- 0.03 vs. 0.59 +- 0.08 U/g in control liver). This indicates that the beta-mannosidase activity in nonparenchymal liver cells is normal. The fraction of enzyme activity recovered in the secretion was elevated 2.55 --4.7-fold for beta-mannosidase (p = 0.022) and beta-glucuronidase (p = 0.017) and not affected for beta-hexosaminidase . This indicates mistargeting of a subset of lysosomal enzymes into the culture medium. Table 1 | Lysosomal enzyme activities in control and LAMP-2 --deficient hepatocytes Cathepsin-D was quantified in hepatocytes by Western blotting. The total amount of cathepsin-D was decreased similar2.5-fold in LAMP-2 --deficient hepatocytes. In particular the amount of the catalytically active forms of cathepsin-D, the 46-kDa processing intermediate, and the mature 30- and 14-kDa polypeptides were decreased in LAMP-2 --deficient hepatocytes, whereas that of the 52-kDa precursor form was increased (Figure A). Figure 5 | Intracellular sorting of cathepsin D is impaired in LAMP-2 --deficient cells. Intracellular sorting of cathepsin D is impaired in LAMP-2 --deficient cells. (A) Western blotting of cathepsin-D from control (+/+) and LAMP-2 -/- cells. p, precursor; i, intermediate; and m, mature form of the enzyme. Equal amounts of protein were loaded in each lane. (B) Immunoprecipitation of cathepsin-D. The cells were labeled with [35S]methionine for 1 h and then chased as indicated. Cathepsin-D was immunoprecipitated from the cells and culture medium. The bands were quantitated by densitometry. The 3- to 5-fold higher fraction of beta-mannosidase and beta-glucuronidase in the secretions and the decrease of proteolytically processed cathepsin-D forms in the hepatocytes points to a missorting of newly synthesized lysosomal enzymes into the secretions. To follow the fate of newly synthesized cathepsin-D, hepatocytes were metabolically labeled, and cathepsin-D was immunoprecipitated from cells and media (Figure B). We observed in LAMP-2 --deficient cells an increased secretion of cathepsin-D. After 6-h chase, 47% of cathepsin-D was recovered in the secretions and 21% had been proteolytically processed to the 46-kDa intermediate. In control hepatocytes 14% had been secreted, and 68% were proteolytically processed to the catalytically active intermediate. Taken together these results show that missorting contributes to the lower levels of cathepsin-D in LAMP-2 --deficient hepatocytes and possibly also to the intracellular deficiency of beta-mannosidase as well as to the increased proportion of beta-glucuronidase in the culture medium. Expression of MPR300 | Missorting of cathepsin-D and two other lysosomal enzymes into secretion suggested that LAMP-2 --deficient hepatocytes might have a defect in the function of MPRs. We first investigated the expression of MPR300. Western blotting showed comparable protein levels in control and LAMP-2 --deficient hepatocytes (Figure A). Densitometric quantitation from 17 independent cultures revealed that LAMP-2 --deficient cells had 78.3 +- 22.6% (mean +- SD) of the protein level found in control hepatocytes. Metabolically labeled MPR300 had comparable half-lives in control and LAMP-2 --deficient hepatocytes (Figure B). Figure 6 | Steady state levels of MPR300. Steady state levels of MPR300. (A) Western blotting and of MPR300 in control (wt), LAMP-2 -/-, MPR46-/-, and MPR300 -/- cells. Ponceau-stained bands are shown as a loading control. (B) Immunoprecipitation of MPR300 after 6-h pulse with radioactive methionine followed by chase for 0, 12, and 24 h. Densitometric quantitation of the bands is also shown. Half-life and Localization of MPR46 | We next performed Western blotting of MPR46 (Figure A). Densitometric quantitation of bands from 10 independent cultures revealed that the steady state level of MPR46 was reduced to 27.5 +- 8.9% of controls in LAMP-2 --deficient hepatocytes. However, Northern blotting showed that the level of MPR46 mRNA was comparable to the control (Figure B). Stability of metabolically labeled MPR46 showed MPR46 to have a shortened half-life in LAMP-2 --deficient hepatocytes: 41 --42% of newly synthesized receptor was recovered in LAMP-2 --deficient cells after 12-h chase (Figure C), compared with 96% in control cells. Addition of lysosomal protease inhibitors, leupeptin and pepstatin, to the chase medium rescued the half-life in knockout cells to control level (Figure C), suggesting that degradation by lysosomal proteinases contributes to the shorter half-life. Figure 7 | MPR46 has a shorted half-life and altered subcellular localization in LAMP-2 --deficient hepatocytes. MPR46 has a shorted half-life and altered subcellular localization in LAMP-2 --deficient hepatocytes. (A) Western blotting of MPR46 from control (+/+) and LAMP-2 -/- cells. MPR46-/- and MPR300 -/- hepatocytes are shown as controls. (B) Northern blotting of MPR46 and actin. (C) Immunoprecipitation of MPR46. The cells were labeled with [35S]methionine for 12 h. Leupeptin (100 muM) and pepstatin (100 muM) were added to the chase medium as indicated (Leu/Pep). Quantitation of the bands is shown at the bottom. (D and E) Immunofluorescence of MPR46 from control (D) and LAMP-2 --deficient cells (E). The shortened half-life of MPR46 prompted us to study its intracellular localization. Immunofluorescence showed that in control cells MPR46 was localized to the perinuclear region as vesicular-reticular labeling (Figure D), consistent with earlier results showing that MPR46 is located in the TGN, endosomes, and small cytoplasmic vesicles . In the LAMP-2 --deficient hepatocytes MPR46 labeling was observed in only 42% of cells, whereas 95% of control hepatocytes showed labeling. Thus, there is an apparent inconsistency between the Western blot (27.5% the protein level of controls) and immunofluorescence labeling of MPR46. However, 27% of average protein level could be a result from 1) 27% of cells expressing normal amount of protein, or 2) all cells expressing only 27% of protein, or any combination of these two. In addition, immunofluorescence does not detect labeling if the local concentration of the protein is low, such as the concentration of MPR46 on the plasma membrane. In LAMP-2 --deficient hepatocytes with detectable MPR46, the receptor was found more distributed throughout the cytoplasm and less concentrated in the perinuclear region (Figure E). To investigate localization of MPR46 in the TGN, we performed double immunogold labeling of MPR46 with gamma-adaptin. In control hepatocytes the most concentrated MPR46 labeling was found in the Golgi region, identified by gamma-adaptin labeling and the presence of a Golgi stack (Figure A). In LAMP-2 --deficient hepatocytes the proportion of MPR46 label found in the Golgi region was reduced to 27% of the control . With gamma-adaptin double labeling it was not possible to unequivocally identify the compartments where the rest of MPR46 was located. We thus also performed double labeling of MPR46 with LAMP-1. In control hepatocytes MPR46 was frequently found in small membrane structures surrounding LAMP-1 --positive multivesicular endosomes (Figure B). In LAMP-2 --deficient hepatocytes less MPR46 labeling was observed in these structures (Figure C). The ratio of MPR46 labeling inside versus outside of multivesicular endosomes was 4.8 times higher in LAMP-2 --deficient cells . In addition more MPR46 was found inside autophagic vacuoles in LAMP-2 --deficient cells (Figure , C and D). Quantitation showed that 3.9 --5.2 times more MPR46 labeling was found in autophagic vacuoles . Typically the MPR46-positive autophagic vacuoles were fusion profiles of an Avd and a multivesicular endosome (Figure , C and D). Taken together, our results show that in LAMP-2 --deficient hepatocytes MPR46 accumulates in Avd/endosomes, suggesting that defective sorting out of endosomes leads to faster degradation of the protein. Figure 8 | Immunogold electron microscopy of MPR46. Immunogold electron microscopy of MPR46. (A) Double immunogold labeling of MPR46 (10 nm gold) and gamma-adaptin (5 nm gold, arrowheads) in control cells. The most concentrated labeling is found in the Golgi region. G indicates the Golgi stack. (B --D) Double immunogold labeling of LAMP-1 (10 nm gold) and MPR46 (5 nm gold) in control (B) and LAMP-2 -/- cells (C and D). (B) In control cells MPR46 was typically located in membrane structures (arrowheads) around the endosomal vacuoles (indicated by E), and only some labeling was seen inside the endosome (arrows). (C and D) In LAMP-2 -/- cells the amount of MPR46 in membrane structures around endosomes was smaller (arrowheads), whereas more MPR46 was found inside endosomes (E) or autophagic vacuoles (AVd) (arrows). Note that in C and D the autophagic vacuoles appear to have fused with multivesicular endosomes; stars indicate some of the internal vesicles. Bars, 200 nm. Table 2 | Localization of MPR46 immunogold labelling (% of total) in the Golgi region and endo/lysosomal compartments in control and LAMP-2 -/- hepatocytes DISCUSSION : We have shown that intracellular retention of cathepsin-D, beta-glucuronidase, and beta-mannosidase is impaired in LAMP-2 --deficient hepatocytes. We also could show altered localization of MPR46, including accumulation in Avd, less MPR46 in the Golgi region, and altered localization within endosomes. In control hepatocytes MPR46 was frequently found in small membrane structures surrounding LAMP-1 --positive multivesicular endosomes (Figure B). These structures have been called "endosome associated vesicles and tubules," and they are thought to participate in MPR46 recycling out of the endosomes . We observed less MPR46 in these structures in LAMP-2 --deficient cells (Figure , C and D, and Table ). During targeting of lysosomal enzymes, MPR46 and MPR300 deliver newly synthesized enzymes to early or late endosomes , after which they may recycle back to TGN from late endosomes. Because in LAMP-2 --deficient hepatocytes, endosomes and late autophagic vacuoles readily fuse , both compartments can be assumed to act as starting point for receptor recycling. The ratio of MPR46 in the Golgi region versus MPR46 in endo/lysosomes and AVd was 1.81 in control and 0.28 in LAMP-2 knockout cells . Together with the lower amount of MPR46 in endosome-associated vesicles and tubules, this suggests that the recycling of MPR46 from endosomes back to the TGN is less efficient in LAMP-2 --deficient cells. However, we also observed that the transport of lipids from endo/lysosomes to the Golgi is not affected by LAMP-2 deficiency , suggesting that the recycling defect may be specific to MPR46. The recycling defect may be due to a failure to sort MPR46 in endosomes. Localization to different endosomal subcompartments is known to be the basis of sorting between the endosome-to-plasma membrane recycling and early endosome-to-late endosome transport routes . MPRs are thought to recycle to the TGN from late endosomes , which are also called multivesicular bodies or the prelysosomal compartment . Reduced ability to recycle MPR46 from late endosomes has been shown to lead to transport of the receptors to lysosomes where they are degraded . Because the endocytic and autophagic pathways merge, we observed enrichment of MPR46 in autophagic vacuoles . Impaired trafficking of MPR46 could explain partial mistargeting of a subset of lysosomal enzymes to the extracellular medium, which in turn could be one cause for the impaired capacity for lysosomal degradation of long-lived proteins observed in LAMP-2 --deficient hepatocytes . Thus, rather than by decreased fusion of autophagic vacuoles with lysosomes, the increased half-life and accumulation of autophagic vacuoles can be explained by defective lysosomal biogenesis. In this study we observed impairment of intracellular retention for cathepsin D, beta-mannosidase, and beta-glucuronidase. Intracellular activity of beta-hexosaminidase appeared to be increased but the proportion of activity recovered in the culture medium was comparable to controls . This suggests that the increased intracellular activity is likely due to increased synthesis. Partial mistargeting of a subset of lysosomal enzymes is in agreement with the findings that we observed a shorter half-life of MPR46, whereas the half-life of MPR300 was comparable to controls. Both MPRs are necessary for normal targeting of lysosomal enzymes . In addition, other lysosomal enzyme targeting mechanisms exist besides the mannose 6-phosphate mediated route . We observed a lower steady state level and shorted half-life for MPR46 in LAMP-2 --deficient cells, whereas those of MPR300 were comparable to the control. Although the endosomal sorting and/or trafficking of both receptors may be impaired in LAMP-2 --deficient cells, their susceptibility to lysosomal degradation or acidic pH may vary. Alternatively LAMP-2 deficiency may impair, directly or indirectly, recycling of MPR46 but not MPR300. In spite of being impaired in degrading autophagocytosed cytoplasm, LAMP-2 --deficient hepatocytes are able to acidify autophagic vacuoles . MPR46 dissociates from its ligands in acidic environment . Using the pH indicator drug DAMP, we estimated the pH of Avd to be 5.7 in control and 5.8 in LAMP-2 --deficient hepatocytes . Also the pH in multivesicular endosomes was close to the controls (pH 5.8) in LAMP-2 --deficient hepatocytes (pH 5.9, unpublished observations). This suggests that impaired acidification in endosomes or Avd is not the cause of impaired MPR46 recycling and accumulation in Avd. How could LAMP-2 deficiency lead to impaired recycling of MPR46 from late endosomes to the TGN? One possibility is that LAMP-2 is necessary for sorting of MPR46 inside endosomes, either directly or indirectly by stabilizing factors needed for recycling. LAMP-2 has been shown to mediate transport of a specific set of cytosolic proteins across the lysosomal membrane in chaperone-mediated autophagy . Thus it is possible that LAMP-2 could participate in chaperone-mediated transport of recycling promoting factors from the cytoplasm to the endosomal lumen. Another possibility is that LAMP-2 is needed to prevent transport of MPR46 from endosomes to lysosomes and thus to increase the probability that the receptor will have time to reach the endosomal subcompartment that is destined for recycling to the TGN. LAMP-2 has been shown to recycle between the lysosomal limiting membrane and matrix and thus regulate the rate of chaperone-mediated autophagy . Downregulation of chaperone-mediated autophagy by LAMP-2 levels has been proposed to mediate epidermal growth factor --induced cell growth in renal tubular cells . These findings show that LAMP-2 is not merely a structural component of the lysosomal membrane but has more sophisticated functions. Yet another possible explanation for the decreased intracellular protein degradation in LAMP-2 --deficient cells is that LAMP-2 deficiency leads to selective disturbances in lysosomal functions, including the observed beta-mannosidase deficiency, and beta-glucuronidase and cathepsin-D mistargeting. LAMP-2 has been suggested to play a role in intralysosomal matrix formation . This could in turn lead to secondary effects such as impaired receptor recycling. It is also possible that altered trafficking of lysosomal enzymes and loss of MPR46 are separate events caused by loss of LAMP-2. Still another, although less likely explanation of altered lysosomal enzyme trafficking would be altered activity and/or localization of uncovering enzyme. This enzyme performs the final cleavage step in the biogenesis of mannose 6-phosphate tag . Further studies are needed to differentiate between these possible connections between LAMP-2 deficiency and lysosomal enzyme targeting. MPR46 knockout mice had a normal phenotype, although partial missorting of many lysosomal enzymes into secretion was observed in cells isolated from these mice . Lysosomal storage was not detected in liver by electron microscopy . However, in intact tissues the increased secretion of lysosomal enzymes was shown to be compensated by uptake via carbohydrate-specific endocytic receptors . Our preliminary results suggest that, although fluid-phase endocytic uptake is normal, receptor-mediated endocytic uptake is impaired in LAMP-2 --deficient hepatocytes, suggesting that the compensatory uptake is not functional in LAMP-2 --deficient mice. This could explain the more severe phenotype in LAMP-2 --deficient mice and isolated hepatocytes. Inhibition of the phosphoinositide 3 kinase Vps34 by a dominant negative form of this enzyme has been reported to cause a similar mistargeting of cathepsin D as observed in LAMP-2 --deficient hepatocytes. On the other hand Vps34 has also been shown to be necessary for autophagosome formation . Because we see a profound accumulation of autophagic vacuoles in LAMP-2 --deficient hepatocytes, we can relatively safely conclude that Vps34 is functional in these cells and thus rule out Vps34 deficiency as an explanation of the defect in lysosomal enzyme targeting. Why do we only see accumulation of autophagic vacuoles in some tissues of LAMP-2 --deficient mice, including pancreas, liver parenchyme, heart, muscle, capillary endothelium of kidney, intestinal wall, lymph nodes, and neutrophilic leukocytes , while other tissues such as brain and fibroblasts seem to be normal? Many of the affected tissues in LAMP-2 --deficient mice are those that have a high degree of autophagy in normal animals. Using mainly rat tissues, active autophagy has been described at least in liver , pancreas , muscle , heart , and kidney . This suggests that active ongoing autophagy in normal conditions may be the prerequisite for the observed phenotype of autophagic vacuole accumulation in LAMP-2 --deficient tissues. It should be noted that the structurally related LAMP-1 might compensate in part for the loss of LAMP-2 in LAMP-2 --deficient mice. These compensating functions are supported by the phenotype of LAMP-1/LAMP-2 double-deficient mice. Although the single-deficient mice are fertile and viable , the loss of both LAMP molecules leads to embryonic lethality associated with accumulation of autophagic vacuoles in almost all embryonic tissues (P. Saftig, unpublished data). In summary, instead of being only a structural component of the lysosomal membrane, LAMP-2 plays a more dynamic role than previously anticipated in cellular processes such as macro autophagy, chaperone-mediated autophagy, and receptor trafficking. To further elucidate the postulated LAMP-2 function in receptor sorting and/or trafficking, it will be important to identify possible binding partners of LAMP-2. Coimmunoprecipitation and yeast two-hybrid screen experiments will possibly bring more light to this intriguing question. Backmatter: Abbreviations used: : Avi = early autophagic vacuole Avd = late autophagic vacuole BSA = bovine serum albumin ER = endoplasmic reticulum LAMP = lysosomal associated membrane protein LBPA = lysobisphosphatidic acid 3MA = 3-methyladenine MPR = mannose 6-phosphate receptor TGN = trans-Golgi network PMID- 12221131 TI - A Rab8-specific GDP/GTP Exchange Factor Is Involved in Actin Remodeling and Polarized Membrane Transport AB - The mechanisms mediating polarized delivery of vesicles to cell surface domains are poorly understood in animal cells. We have previously shown that expression of Rab8 promotes the formation of new cell surface domains through reorganization of actin and microtubules. To unravel the function of Rab8, we used the yeast two-hybrid system to search for potential Rab8-specific activators. We identified a coil-coiled protein (Rabin8), homologous to the rat Rabin3 that stimulated nucleotide exchange on Rab8 but not on Rab3A and Rab5. Furthermore, we show that rat Rabin3 has exchange activity on Rab8 but not on Rab3A, supporting the view that rat Rabin3 is the rat equivalent of human Rabin8. Rabin8 localized to the cortical actin and expression of Rabin8 resulted in remodeling of actin and the formation of polarized cell surface domains. Activation of PKC by phorbol esters enhanced translocation of both Rabin8 and Rab8-specific vesicles to the outer edge of lamellipodial structures. Moreover, coexpression of Rabin8 with dominant negative Rab8 (T22N) redistributes Rabin8 from cortical actin to Rab8-specific vesicles and promotes their polarized transport to cell protrusions. The C-terminal region of Rabin8 plays an essential role in this transport. We propose that Rabin8 is a Rab8-specific activator that is connected to processes that mediate polarized membrane traffic to dynamic cell surface structures. Keywords: INTRODUCTION : Cell division, differentiation, and migration are crucial events for the development of multicellular organisms. During these processes cells polarize through reorganization of both external and internal components , such as actin, microtubules, and adhesion receptors. Actin is linked to adhesion molecules that mediate cell attachment, and microtubules modulate the distribution of many internal organelles and structures. In migrating cells, actin and small GTPases of the Rho family control the dynamic formation of lamellipodia, filopodia, and focal adhesions during cell morphogenesis and migration . There is also indirect evidence that inhibiting membrane trafficking affects these processes . Although vesicular transport has been extensively studied in simple polarized cells, like epithelial cells , little is known about the mechanisms that target vesicles to a specific site on the cell surface. Apical and basolateral targeted vesicles must use different microtubules to reach their appropriate surfaces, but it is not known whether they fuse randomly or at distinct locations on the plasma membrane. There are some indications that basolateral transport vesicles are targeted to tight junctions: the mammalian Sec6/s8 complex is localized to tight junctions in polarized cells but are relocalized to vesicles when the polarity is lost . This protein complex may also have a role in targeting membranes to axonal synapse-assembly domains . Targeted membrane transport in motile cells is likely to be more complex than in stationary cells because the cells are behaviorally dynamic: new cell surface domains are being continuously created and destroyed . In such a situation the targeting process is probably under strict control . The Rab small GTPases are likely candidates in controling targeting of vesicles, because they are known to regulate different transport routes in the cell . The role of Rab proteins in the endocytic pathway has been well documented ; Rab5 controls membrane traffic from the plasma membrane to early endosomes and binds to several proteins : RabGDI and Rabex-5 modulate the function of Rab5, and Rabaptin5, EEA1, and Rabosyn5 are Rab5 effectors . Most Rab effectors specifically interact with a particular Rab protein, and only closely related Rabs bind to the same effector molecules . There is a growing list of proteins that interact with Rabs, perhaps indicating a wider function for Rabs than originally assumed. Rab8 modulates polarized membrane transport through reorganization of actin and microtubules, induces the formation of new surface extensions, and has an important role in directed membrane transport to cell surfaces . We have characterized proteins that interact with different forms of Rab8 and reported a Rab8-GTP binding, coiled-coli protein: FIP-2 . This TNF-alpha inducible protein links Rab8 to huntingtin and modulates cell morphogenesis to some extent. Rab8-GTP also binds Rabip8/GCK, a participant in TNF-alpha --mediated processes , indicating a likely role for Rab8 in regulating membrane transport linked to stress responses and differentiation. We also looked for proteins that bind to Rab8-GDP and have identified a novel human protein that we named Rabin8. Rabin8 is closely homologous to rat Rabin3: a Rab3A-binding protein of unknown function . We show that Rabin8 is a guanine nucleotide exchange factor that participates in the polarized delivery of Rab8 vesicles to protrusive structures and that a non --Rab8-binding region of Rabin8's carboxy terminus is essential for targeting Rab8 vesicles to the cell surface. MATERIALS AND METHODS : Constructs | We use PCR mutagenesis to construct a C-terminal deletion mutant of Rab8 (Rab8Delta) where we removed Rab8's lipid modification motif. We also used PCR to construct Rabin8's deletion mutants of Rabin8, which consisted of codons 1 --120, 1 --221, 1 --316, 101 --316, 222 --460, and 306 --460. These deletions were cloned into pB42AD, and all constructs were verified by sequencing. Rabin8 and Rabin8 (1 --316aa) open reading frames were provided with a myc-tag and cloned into pEGFP-N1. The pHis-Rab8 T22N and pHis-Rab8 Q67L were obtained by cloning his-Rab8 T22N and his-Rab8 Q67L, from recently described pGEM-his-Rab8-T22N and pGEM-his-Rab8-Q67L plasmids, into pEGFP-NI . The RhoA and Rab2 genes were amplified by PCR from human HeLa cDNA and the corresponding mutants RhoA (14V), RhoA (19N), and Rab2 (20N) were generated as described previously for Rab8 . Rab3A was amplified from human brain cDNA, and the Rab3A-36N mutant was created as above. All constructs made by PCR were verified by sequencing. For expression of Rabin8 in Escherichia coli, Rabin8 was cloned in pET43 (Novagen, Madison, WI), pGEX2T (Amersham Pharmacia, Piscataway, NJ), and pGAT2 . The rat Rabin3 open reading frame was amplified by PCR from PC12 cDNA and cloned into pET43. The Rab3A, Rab5, and Rab8 genes were also cloned into pET43. Details of the constructs are available on request. Yeast Two-hybrid Screen and Cloning of Rabin8 | Screens were done with the Gal4-based system (, Palo Alto, CA), and subsequent two-hybrid interaction studies were performed in the LexA system . Rab8bDelta-T22N (the dominant negative mutant) was cloned into the pAS2 --1 vector and used as bait to screen a human brain cDNA library expressed from the pACT2 vector . Out of 10 million clones screened, 70 were true positives; 50% of the positives were Mss4, and 10% contained sequences with a high degree of homology to rat Rabin3. We recloned Rabin8 into pB42AD in order to test for binding to different small GTPases in the lexA-based two-hybrid system. Rab8, Rab8-T22N, Rab8-Q67L, Rab8b, Rab8b-T22N, Rab8b-Q67L, Rab3A-36N, and Rab2 --20N were cloned in pGilda and used with pB42AD-Rabin8 or with corresponding deletion mutants (see above; ; ). Full-length human Rabin8 was cloned from a human brain cDNA lambda Triplex library by a PCR-based method . In the first round 64 wells with 8000 phages per well were screened. We performed PCR on extracts from positive phage pools and then ran the PCR product on a 1% agarose gel. The agarose gel was blotted overnight to nitrocellulose and then probed with a Rabin8-specific oligonucleotide. One positive well was found, titered, and subdivided into 64 new pools, now with 150 phages per well. These were screened in the same way, and one positive well was found. At this stage the phages from the positive well were screened by a plaque lift on nitrocellulose filters according to the Triplex manual. Positive plaque were picked and eluted over night in lambda dilution buffer at +4C. The phage DNA was converted to plasmid form according to the Triplex manual, and the inserts of two independent clones were sequenced. The full-length sequence of human Rabin8 was present in both clones. Sequence comparisons and evaluations were done with the Basic Local Alignment Search Tool (BLAST; ). Domain and motif searches were done with the SMART tool . Northern Blot | A human 12-lane Multiple Tissue Northern Blot was hybridized over night in UltraHyb solution (Ambion, Austin, TX) using standard hybridization conditions. A 496-bp SpeI-HindIII fragment from the full-length clone was used as probe. It was labeled with 32P-dCTP (Amersham) with a random labeling kit (Amersham). The control probe (beta-actin), provided with the MTN-filter, was labeled and hybridized to the stripped filter in the same way. The blots were visualized by autoradiography or by phosphoimager. Rabin8 Antibodies | His-GST-Rabin8 expressed from pGAT2 was purified under denaturing conditions and prepared for immunization of rabbits as previously described for Rab8 . Antibodies to Rabin8 were affinity purified by use of nitrocellulose strips containing recombinant Rabin8 . Production of Recombinant Proteins | For nucleotide exchange assays, NusA-His-Rabin8 was expressed from the pET43a vector, induced at 37C for 3 h with 0.5 mM IPTG. The cells were broken by French Press in lysis buffer (20 mM Tris-HCl, pH 8.5, 50 mM NaCl, PMSF), and the fractions were separated by centrifugation at 10, 000 x g for 10 min. The supernatant was loaded onto an anion exchange column, and the fractions containing Rabin8 were pooled and concentrated from 20 to 2 ml. Aliquots of 0.5 ml were loaded onto a gel filtration column, the fractions containing Rabin8 were pooled, and concentrated from 4 to 1 ml. This material was snap-frozen in liquid nitrogen and stored at -70C. NusA-His-ratRabin3 was expressed and purified as described above for Rabin8. Rab8-fusion proteins have been insoluble in all tested expression vectors and conditions, but a Rab8 NusA-fusion is partially soluble. NusA-His-Rab8 was expressed from the pET43 vector at 37C for 3 h with 0.1 mM IPTG. The cells were resuspended in lysis buffer (50 mM phosphate buffer, pH 7.0, 300 mM NaCl, 5 mM MgCl2, 200 mM GDP, 5 mM beta-mercaptoethanol, 0.5% Triton X-100, 10% glycerol, and PMSF) and lysed by a French Press. The fractions were separated by centrifugation in a table centrifuge at 13,000 rpm for 10 min. The fusion protein was bound to Talon resin and washed four times during 40 min with the same buffer not containing PMSF. The beads were then washed briefly in Thrombin buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mM MgCl2, 2.5 mM CaCl2). The beads with bound protein were incubated with thrombin (Sigma, St. Louis, MO) overnight at room temperature in Thrombin buffer. Cleaved, soluble, wild-type Rab8 was recovered from the buffer, snap-frozen in liquid nitrogen, and then stored at -70C. Rab3 and Rab5 were expressed and purified in the same way as Rab8wt. GST-fusion Expression and Binding Assay | We expressed GST-Rabin8 from pGEX-2T at 15C overnight with 200 muM IPTG and purified the fusions according to the manufacturer's instructions (Amersham Pharmacia). The lysate containing GST-Rabin8 expressed from the pGEX2T vector was incubated with glutathione-agarose beads (Sigma) at +4C for 1 h and then washed three times during 30 min with lysis buffer. Control beads with GST alone were done the same way as GST-Rabin8 beads except that the GST protein was induced from the pGEX vector at 37C for 3 h instead of 15C overnight. Rab8-Q67L, Rab8-T22N, Rabin8 (1 --316aa), and full-length Rabin8 were translated in vitro using a TNT Quick kit (Promega) according to the manufacturer's instructions. The in vitro translation products were then incubated with protein-coupled glutathione agarose beads in binding buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 2 mM MgCl2, 1% Triton X-100) on a wheel at +4oC for 1 h. The beads were washed four times with buffer during 30 min. Bound material was eluted from the beads with Laemmli sample buffer and loaded onto a 12% SDS-polyacrylamide gel. As a control, part of the in vitro translation reactions was also loaded onto the gel. The bands were visualized by autoradiography of the dried gel. GDP/GTP Exchange Assays | We assayed [3H]GDP release by preloading purified wild-type Rab8wt (20 pmol), Rab3A (20 pmol), and Rab5 (20 pmol) with 100 pmol [3H]GDP (10 Ci/mmol; Amersham, TRK 335) diluted to in preloading buffer (20 mM HEPES, pH 7.2, 5 mM EDTA, 1 mM DTT) for 15 min at 30C. The reactions were then transferred to ice, and MgCl2 was added to a final concentration of 10 mM. The reactions were started by addition of reaction buffer (20 mM HEPES, pH 7.2, 10 mM MgCl2, 1 mM GDP, 1 mM DTT) with or without purified NusA-Rabin8 (10 pmol), NusA-ratRabin3 (10 pmol) or NusA (10 pmol) to a total reaction volume of 50 mul containing 20 pmol of wild-type Rab8. The reactions were incubated at 30C for varying periods of time. Samples of 5 mul were diluted into 2 ml of ice-cold wash buffer (20 mM Tris-HCl, pH 8.0, 20 mM NaCl, 10 mM MgCl2, 1 mM DTT) and applied to wet nitrocellulose filters. The filters were washed twice with 3 ml ice-cold wash buffer and dried before adding the scintillation fluid (Optiphase High Safe-3; Wallac, Turku, Finland). [35S]GTP binding was assayed in the same way as the GDP release, but preloading was done with 30 pmol cold GDP nucleotide. The reaction buffer contained 1 mM ATP to prevent hydrolysis of [35S]GTP (1117 Ci/mmol; Amersham, SJ 1320). The reaction contained 20 pmol wild-type Rab8 and 22 Ci/mmol of [35S]GTP. Samples were taken in the same way as described above and counted after overnight incubation in scintillation fluid. Cell Culture and Transfections | HeLa cells were cultured in DMEM supplemented with 10% FCS, 2 mM glutamine, 100 U/ml penicillin, and 100 mug/ml streptomycin. For transfection experiments cells were grown overnight on coverslips in 35-mm plates. Cells were transiently transfected using the Fugene 6 transfection reagent (Roche, Indianapolis, IN) with constructs containing Rabin8, EGFP-Rabin8, myc-Rabin (1 --316), his-Rab822N, and his-Rab867L, EGFP-RhoA14V, and EGFP-RhoA19N. Twenty hours after transfection some cells obtained different inhibitors freshly diluted in Optimem; cytochalasin D (1 muM) for 20 min, nocodazole (1 mug/ml) for 60 min, and phorbol 12-myristate 13-acetate (PMA; 100 ng/ml) for 30 min. Confocal Immunofluorescence Microscopy | Cells were prepared for immunofluorescence microscopy by fixing them with 4% paraformaldehyde, permeabilizing them with 0.1% TX-100 and staining them with appropriate antibodies . Goat anti-rabbit IgG lissamine and goat anti-mouse IgG-FITC secondary antibodies were from Jackson Immunoresearch (West Grove, PA). Actin was detected in fixed cells by Alexa488-conjugated phalloidin or Texas-Red XS --conjugated phalloidin (Molecular Probes, Eugene, OR). Fluorescence of fixed cells was observed either with the MRC-1024 confocal system (Hercules, CA) linked to Zeiss Axiovert 135 M microscope (Thornwood, NY) or with an fluorescence microscope (Lake Success, NY). RESULTS : Identification of Rab8-GDP --interacting Proteins | To find potential Rab8 activators we performed a yeast two-hybrid screen with a human brain cDNA library, using dominant negative Rab8b (the T22N mutant) as bait. We found several proteins that interact with this mutant; human Mss4 was 50% of the sequenced clones and is a known exchange factor for a subgroup of Rabs, including Rab8 . Another group of clones contained sequences highly homologous to rat Rabin3 . By Northern blot analysis Rabin3-like mRNA is ubiquitously expressed and at the highest levels in kidney, brain, and heart (Figure B). We cloned Rabin-3-like's full-length gene from a lambda triplex human brain cDNA library. The nucleotide and amino acid sequences were identical to the Homo sapiens hypothetical protein FLJ22548 (NCBI; accession no. ). This Rabin-3-like protein is 90% identical and 93% similar to rat Rabin3 (Figure A). Because of functional specificity of this protein on Rab8, but not on Rab3A, we call it Rabin8 (see below). Specific antiserum against the full-length Rabin8 produced in E. coli was made in rabbits. Affinity-purified anti-Rabin8 was used to probe cell lysates from four different cell lines: HeLa, Jurkat, A431, and endothelial cells (Figure D). The antibodies recognized a 50-kDa protein, which closely matched the theoretical size of the protein. Rabin8 was most abundant in A431 and Jurkat cells, whereas there was very little Rabin8 in human endothelial cells. We could also show that Rab8 was present in all of the examined cells (Figure D). Figure 1 | Amino acid sequence, tissue distribution, and endogenous Rabin8 protein in cell extracts. Amino acid sequence, tissue distribution, and endogenous Rabin8 protein in cell extracts. (A) Sequence alignment of human Rabin8 and rat Rabin3 obtained by ClustalW. Black boxes indicate identity and gray boxes indicate conservative amino acid substitutions (Boxshade software). (B) Northern blot analysis of Rabin8 RNA in different human tissues. (C) Actin RNA in corresponding tissues. (D) The presence of the endogenous Rabin8 protein in whole extracts of Hela, Jurkat, A431, and human endothelial cells was examined by Western blotting using affinity-purified anti-Rabin8. Molecular markers are indicated on the left. (E) Western blotting of same cell extracts detected by anti-Rab8. Characterization of the Interaction between Rab8 and Rabin8 | We performed two-hybrid binding studies with Rabin8 against a panel of different Rabs and their variants and found that Rabin8 interacted with Rab8 and Rab8b in the GDP-bound form (T22N mutant) but not with the corresponding GTP-bound (Q67L mutant) or wild-type forms (Figure A). Rabin8 also interacted with GDP-bound Rab3A (T36N mutant) and surprisingly also with Rab3A-wt, but it did not bind the Rab2 dominant negative mutant (20N) or laminin. These results suggest that Rabin8 interacts with GDP-bound Rab3A and Rab8 proteins. Figure 2 | Characterization of Rab8 binding to Rabin8. Characterization of Rab8 binding to Rabin8. (A) Table showing interactions of Rabin8 in the two-hybrid system. The interactions were assayed by X-gal overlay and by the ability of the yeast to grow in the absence of leucine; -, no interaction; +, weak interaction; +++, strong interaction. (B) Schematic representation of Rabin8 deletions interacting with Rab8-T22N in yeast two-hybrid analysis. The black box indicates the coiled-coli region of Rabin8, and the numbers behind the bars indicate the amino acids encompassing the Rabin8 deletions. Positive interaction, +; negative interaction, -. (C) In vitro binding of Rab8 mutants to recombinant GST-Rabin8. Lanes beneath in vitro indicate input of translated Rab8-T22N and Rab8-Q67L, whereas lanes in the GST-Rabin8 column indicate in vitro bound Rab8-T22N and Rab8-Q67L to GST-Rabin8. The GST column shows the result obtained by GST alone. Molecular weight markers are indicated on left. (D) In vitro association of Rabin8 to GST-Rabin8. In vitro translated full-length (FL) Rabin8 or the Rabin8 1 --316 deletion mutant (DM) seen as in put material in the in vitro column. The lanes beneath the GST-Rabin8 column indicate bound material of full-length (FL) and deleted Rabin8 (DM) to GST-Rabin. The column on right shows corresponding binding to the GST control beads. Positions of protein standards are given at the left of the gel. We then constructed Rabin8 deletions and assayed them in the two-hybrid system (Figure B). Neither the N-terminal half (aa1 --221) nor the C-terminal half (aa222 --460) of Rabin8 interacted with GDP-bound Rab8, but the 1 --316aa and 101 --316aa constructs did bind GDP-bound Rab8. This suggests that the Rab8-specific binding region of Rabin8 resides between aa101 and aa316. This region contains the entire coiled-coil domain (aa149 --244) of Rabin8 and also a potential RhoA-binding HR1 motif; however, neither the two-hybrid analysis nor the in vitro binding assays showed any interactions between Rabin8 and RhoA. To further verify binding specificity of Rabin8 against Rab8, we performed an in vitro binding assay. We bound a GST-Rabin8 fusion protein to glutathione agarose beads and found that it pulled-down in vitro --translated GDP-bound Rab8 (T22N) but did not bind to GTP-bound Rab8 (Q67L; Figure C). Neither of the in vitro translation products bound to GST beads. Under identical conditions Rab3A-T36N was not bound to GST-Rabin8, perhaps indicating that the sensitivity of the two-hybrid system might bring up false positives. Many coiled-coli proteins form homodimers or homomultimers, so we tested whether immobilized GST-Rabin8 could bind to in vitro --translated Rabin8 (Figure D) and found that there is an interaction, indicating that Rabin8 can potentially form dimers or multimers. Rabin8 did not bind to GST under identical conditions. Rabin8 Is a Rab8-specific GDP/GTP Exchange Factor | Rabin8 has some identity with Saccharomyces cerevisiae's Sec2p, a Sec4p-specific exchange factor that raised the possibility that Rabin8 may also be an exchange factor . Rabin8 and especially Rab8 have been difficult to produce in E. coli because of a strong tendency to form inclusion bodies, but we were able to express these proteins in a more soluble form by using a recently introduced expression system based on NusA fusions . Rabin8 was used as a purified NusA fusion protein, whereas Rab8 was cleaved from the purified NusA-Rab8 fusion with thrombin (Figure A). The purity of the proteins ranged between 85 and 95%. Figure 3 | Rabin8 and Rabin3 are nucleotide exchange factors for Rab8. Rabin8 and Rabin3 are nucleotide exchange factors for Rab8. (A) Recombinant NusA-Rabin8 (lane 1), Rab8 (lane 2), and NusA (lane 3) expressed and purified from E. coli lysates. Molecular markers seen at the left of the Coomassie brilliant blue --stained gel. (B) Rabin8 promotes dissociation of GDP from Rab8. Recombinant, wild-type Rab8 was loaded with [3H]GDP and then incubated alone (K), with NusA protein or with NusA-Rabin8. Aliquots of these three reactions were taken out at different time points and filtered on nitrocellulose filters. Note that Rab8 loaded [3H]GDP in the presence of NusA-Rabin8 lost 80% of its radioactivity within 10 min. (C) Rabin8 stimulates GTP association with Rab8. The [35S]GTPgammas binding to recombinant, wild-type Rab8 was measured after incubation at 37C for indicated time periods with NusA-Rabin8, NusA, or Rab8 alone. The data are shown as the fold increase of [35S]GTPgammas binding compared with that obtained with nucleotide alone. (D) Rab3A, Rab5, and Rab8 were subjected to [3H]GDP release reactions in the presence (black bars) or absence (gray bars) of NusA-Rabin8. The percentage of [3H]GDP that remained bound to Rab3A, Rab5, and Rab8 after 20 min is presented. Values are means +- SEM from three independent experiments. (E) Rat Rabin3 is a Rab8-specific GEF. Purified NusA-Rabin3 was used to measure [3H]GDP release on Rab3A and Rab8. The percentage of [3H]GDP that remained bound to Rab3A and Rab8 after 20 min is presented. Values are means +- SEM from three independent experiments. We measured the rates of GDP exchange as the decrease in radioactive GDP bound to Rab8 with time (Figure B). Purified, wild-type Rab8 was preloaded with 3H-labeled GDP, and the exchange assay was then performed with an excess of cold GDP in the buffer. We found that Rab8's intrinsic GDP-exchange rate is low but observed a fast drop in radioactivity when we added purified NusA-Rabin8, indicating that Rabin8 facilitates the release of GDP from Rab8. In a GTP exchange assay, we loaded pure Rab8 with cold GDP and incubated it with 35S-labeled GTP. We found that added NusA-Rabin8 increased the rate of bound radioactivity, indicating that Rabin8 facilitates the exchange of GDP for GTP on Rab8 (Figure C). These results lead us to conclude that Rabin8 is a real GEF that promotes both GDP dissociation and GTP association. In contrast, the mss4 protein can only promote GDP dissociation . Because Rab3A interacted with Rabin8 in the two-hybrid system, we tested whether Rabin8 could function as a GEF for Rab3A as well as for Rab5. However, Rabin8 could not increase the intrinsic rate of GDP exchange for Rab3A or for Rab5, as seen for Rab8 (Figure D). The close similarity between Rabin8 and rat Rabin3 raised the question whether rat Rabin3 is also a GEF for Rab8. We cloned the open reading frame of rat Rabin3 into the same expression vector as used for Rabin8. This Rabin3 was expressed, purified, and tested for exchange activity on both Rab3A and Rab8. Rabin3 was unable to increase the exchange rate of Rab3A, but there was a clear increase in the rate of Rab8 (Figure E). Likewise, could not detect any exchange activity on Rab3A by Rabin3. Together, these results suggest that both Rabin8 and rat Rabin3 are Rab8-specific GEFs. Rabin8 Modulates Actin Organization | We transfected HeLa cells with Rabin8 and determined its localization. Rabin 8 expression reduced the number of stress fibers compared with normal HeLa cells (Figure A) with a concomitant formation of protrusions and ruffles (Figure , C --H). Rabin8 colocalize with cortical actin in these structures (Figure , C --H). Excess Rabin8 in cells caused blebbing, suggesting that Rabin8 overexpression is deleterious to the cell. Figure 4 | Overexpressed and endogenous Rabin8 in Hela cells. Overexpressed and endogenous Rabin8 in Hela cells. HeLa cells were transfected over night with empty vector (A and B) or with a construct containing a myc-tagged Rabin8 (C --H). Actin was localized by Alexa488-conjugated phalloidin (A, C, E, and G) and Rabin8 by anti-myc --specific mAb (B, D, F, and H). Protrusions, lamellae, and tails are indicated by arrows. Note that Rabin8-expressing cells contain fewer actin stress fibers and that the cells obtain ruffles, protrusions, and tails. Endogenous Rabin8 (J) and actin (I) were localized by affinity-purified anti-Rabin8 antibodies and Alexa488-conjugated phalloidin, respectively. Rabin8 staining along peripheral actin filaments are indicated by an arrowhead. Bars, 5 mum. It was recently shown that Sec2p localization is dependent on actin . Thus, we tested the effects of inhibitors affecting the dynamics of actin and microtubules on the localization of Rabin8. We treated Rabin8-expressing cells with nocodazole, a microtubule-disrupting agent. In these cells actin became incorporated in strong stress fibers, and Rabin8 was uniformly distributed along the plasma membrane (Figure , D --F) instead of colocalizing with cortical actin (Figure , A --C). TrioD1, a RhoG-specific GEF, has a similar sensitivity to microtubule depolymerization . We also treated Rabin8 transfectants with cytochalasin D, an actin-depolymerizing agent. In these cells cortical actin was disrupted, and Rabin8 was found in patches together with actin, suggesting their close association (Figure , G --I). However, we were not able to demonstrate a direct in vitro association between purified actin and Rabin8. Figure 5 | Rabin8 and the integrity of cytoskeletal elements. Rabin8 and the integrity of cytoskeletal elements. HeLa cells were transfected over night with a construct containing Rabin8. Then the cells were left untreated (A --C) or incubated next morning for 60 min with 1 mug/ml nocodazole (D --F) or for 20 min with 1 muM cytochalasin D (G --I). Analysis were done by confocal microscopy. Actin was detected by Alexa488-phalloidin (A, D, and G) and Rabin8 by an affinity-purified anti-Rabin8 antibody (B, E, and H). C, F, and I are corresponding merged pictures. Arrowheads indicate actin patches containing both actin and Rabin8 after cytochalasin D treatment. Note that there is no colocalization of actin with Rabin8 after nocodazole treatment. Bars, 5 mum. Polarized Distribution of Endogenous Rab8 Is Linked to Cell Density | Endogenous Rab8 exhibited strong polarized localization to the tips of protrusions in HeLa cells (Figure A). Rab8 was often found in filopodia (Figure D), where it colocalized with the transferrin receptor (TFR; Figure , B and C). However, there was very little colocalization of Rab8 and the transferrin receptor on vesicles in the cytoplasm, indicating that Rab8 is not controling the recycling of TFR . We also observed that some cells lacked a polarized distribution of Rab8, which was instead confined to a reticular structure in the perinuclear region (Figure G), and we determined whether the difference in Rab8 localization was due to cell density. HeLa cells were plated either at low or high density on fibronectin-coated coverslips and fixed on the following day. As shown in Figure I, >80% of the cells plated at low density contained polarized distributed Rab8, but only similar10% of the cells plated at high density displayed polarized Rab8. Figure 6 | Polarized distribution of endogenous Rab8 in HeLa cells. Polarized distribution of endogenous Rab8 in HeLa cells. (A) In cells plated at low density, endogenous Rab8 is localized predominantly to the tip of protrusions (arrows), where Rab8 partially colocalize with the transferrin receptor (B and C). A strong localization of Rab8 to filopodia is often seen (D) where it colocalize with the transferrin receptor (E and F). In the cytoplasm the colocalization of Rab8 and the transferrin receptor is not often seen (F). In confluent monolayers Rab8 is localized to a perinuclear region (G). This perinuclear region is very similar to that seen for expressed Rab8 --22N (H). (I) Polarized distribution of endogenous Rab8 was assessed quantitatively by comparing the staining of Rab8 in cells grown overnight at low density (low) or high density (high). All data is a mean +- SEM of three independent experiments (N = 50 cells). Bars, 5 mum. We then expressed constitutive active Rab8 (Q67L) or dominant negative Rab8 (T22N) in HeLa cells. Rab8-Q67L localized to the plasma membrane and to the tips of cell protrusions, . The GDP-bound Rab8-T22N mutant predominantly localized to the perinuclear, reticular structure, similar to that seen for endogenous Rab8 in confluent cells (Figure H). These results indicate that the endogenous Rab8 in its GDP-form is present in the perinuclear, whereas GTP-bound Rab8 shows a polarized distribution to distal regions of cell protrusions. Activation of PKC Results in Polarized Distribution of Rabin8 and Rab8 | Rab8 is linked to huntingtin by FIP-2, and the expression of FIP-2/NEMO --related protein is regulated by cytokines and phorbol esters . Because phorbol esters activate PKC and modulate actin assembly, we determined their effects on Rab8 and Rabin8's distributions. Treating HeLa cells with phorbol esters resulted in the loss of stress fibers and the appearance of actin into structures resembling lamellipodia . When Rabin8-transfected cells were treated for 30 min with PMA, there was a dramatic redistribution of both Rabin8 and actin to the periphery of lamellipodial structures (Figure , A and B). A similar translocation after phorbol ester treatment has been shown for ARNO, a GEF for the Arf6 GTPase . Figure 7 | Redistribution of Rab8 and Rabin8 after PKC activation. Redistribution of Rab8 and Rabin8 after PKC activation. HeLa cells transiently transfected with Rabin8 were treated at 16 h with 100 ng/ml PMA for 30 min. Cells were fixed, permeabilized, and stained with alexa488-phalloidin to detect actin (B) and anti-Rabin8 (A). Similarly transfected cells were also used to detect both Rabin8 (C) and endogenous Rab8 (D) by anti-myc and anti-Rab8 antibodies. Untransfected cells treated with PMA for 30 min were used to compare tranferrin receptor (G) and endogenous Rab8 (H) distribution. Note that Rab8-specific vesicles obtain a peripheral distribution, whereas the transferrin receptor is localized more to vesicles in the cytoplasm. Arrowheads indicate Rab8-positive vesicles in the lamellipodia (D and F). Bar, 5 mum. We tested whether Rab8's distribution is affected by active PKC by treating myc-Rabin8 --transfected cells with PMA and staining for endogenous Rab8 and myc-Rabin8. Rab8-specific vesicles redistributed from the perinuclear region to lamellipodial structures at the cell's periphery, where they were close to Rabin8's positive area (Figure , C and D). PMA did not redistribute the transferrin receptor to these lamellipodia, supporting our view that the main part of this receptor is not found in Rab8-specific vesicles (Figure , E and F). However, we occasionally saw colocalization of Rab8 with the transferrin receptor on vesicles in the cytoplasm. Our results show that PKC activation promotes a polarized distribution of Rabin8 and Rab8 to dynamic actin-containing structures and that the polarization of Rab8 vesicles to the cell periphery is linked to Rabin8's translocation. Rabin8 Promotes Polarized Transport of Rab8-specific Vesicles | The localization of Rabin8 to the plasma membrane would indicate that the nucleotide exchange takes place at the cell surface. However, Rab8-GDP (T22N) was preferentially found in the perinuclear region. We therefore coexpressed Rabin8 with dominant negative Rab8 (T22N) to determine whether Rab8-GDP meets Rabin8 at the plasma membrane or on intracellular vesicles. In the cotransfectants Rab8-T22N clearly relocalized from perinuclear reticular structures to numerous vesicles, and these vesicles accumulated at the tips of protrusions, which is similar to endogenous Rab8's distribution in polarized cells (Figure , A --C). Some of the vesicles were arranged in rows along the cell boundary, possibly traveling along actin filaments (Figure , G --I). Rabin8 also clearly translocated from the plasma membrane to Rab8-T22N --containing vesicles, suggesting that Rab8-T22N expression recruits Rabin8 from the cell periphery to perinuclear vesicles and that Rab8's activation takes place on intracellular vesicles. Rabin8 probably blocks the negative effect of Rab8-T22N and simultaneously activates endogenous Rab8, which leads to polarized delivery of Rab8-T22N/Rabin8 --specific vesicles to the PM and to protrusion formation. When Rabin8 was coexpressed with Rab8-GTP (Q67L), there was no redistribution of Rabin8 from the PM to vesicles, although Rab8-GTP was found on vesicles in cell protrusions. Thus, recruitment of Rabin8 to vesicles is dependent on the nucleotide state of Rab8, taking place when only Rab8-GDP is present. It also suggests that Rabin8 does not dictate the localization of Rab8-GTP, but that Rab8-GTP is competent to mediate polarized transport by itself. This has also been shown for Sec2p and Sec4p . Figure 8 | Rabin8 promotes polarized transport of Rab8-specific vesicles. Rabin8 promotes polarized transport of Rab8-specific vesicles. HeLa cells were cotransfected with Rab8-T22N and myc-tagged Rabin8 cDNA and stained for Rab8 (A, D, and H) and myc (B, E, and G). Cotransfected cells exhibited protrusions (A) with growth cone like structures (square), with vesicles that contained both Rabin8 and Rab8-T22N (B and C). Numerous vesicular structures appeared in these cells (D, E, and F) that were seen to align along the cell periphery (arrows) and to accumulate in tips of protrusions (arrowhead). When cotransfection of Rab8 --22N and a deletion mutant of Rabin8 (1 --316aa) was done no polarization of Rab8-specific vesicles was seen (J, K, and I). In these cells Rab8 --22N (H) remained in a perinuclear region and most of the Rabin8-deletion mutant in was found on the plasma membrane (G). However, there was partial colocalization of Rab8 --22N and the Rabin8-deletion mutant in the perinuclear region (I). C, F, and I represent merge pictures. Bars, 5 mum. Rabin8 is similar to the yeast exchange factor, Sec2, whose carboxy terminus is needed for polarized delivery of vesicles . So we constructed a similar Rabin8 mutant lacking a C-terminal region (1 --316aa), which still bound dominant negative Rab8-T22N. When the constructs were cotransfected, Rab8-T22N did not move from the reticular structures to vesicles, and Rabin8 (1 --316) did not translocate from the plasma membrane to vesicles (Figure , J --L), showing that Rabin8's carboxy terminus is essential for Rabin8 to move to Rab8-containing vesicles and for polarized delivery of these vesicles to the cell surface. DISCUSSION : Only a few GDP/GTP exchange factors have been documented for Rab proteins. One is Rabex-5, which catalyzes Rab5's guanine nucleotide exchange . Rabex-5 forms a complex with another Rab5-binding protein called Rabaptin-5, which may stabilize Rab5-GTP and promote binding to EEA1 . This complex recruits other components necessary for fusion of endosomes to endosomes and fusion of CCV with early endosomes. Recently, another Rab5-specific GEF (RIN) has also been found that links Rab5 to the Ras small GTPase . Another known Rab-specific GDP/GTP exchange factor is Rab3-GEP, a large protein (178 kDa) that affects Rab3A, Rab3C, and Rab3D but does not affect Rab3B . Rab3-GEP colocalizes with Rab3 at synaptic release sites, where it regulates the activity of Rab3 and the release of neurotransmitters . A homologous Caenorhabditis elegans protein, AEX-3, is essential for neurotransmitter release in many classes of neurons . Rabin8 shows no identity with Rabex-5 or Rab3 GEP but is similar90% identical to Rabin3, a rat protein of unknown function . Although Rabin3 was found through a two-hybrid screen using Rab3 as bait, neither Rabin3 nor Rabin8 exhibited any GEF activity for Rab3 (; this report). Because both Rabin8 and Rabin3 are GEFs for Rab8, it is likely that rat Rabin8 is the rat equivalent of human Rabin8. Moreover, a Rab3-specific GEF (GRAB) has previously been found that shows >60% similarity with Rabin8. Both GRAB and Rabin8 utilize the coiled-coil region to mediate Rab interaction . In addition, GRAB is known to bind inositol hexaphosphate kinase (InsP6K1) through the same coiled-coil region . Whether Rabin8 binds InsP6K1 is not known, but it is possible, because the coiled-coil region between GRAB and Rabin8 is almost identical. GRAB and Rab3 both regulate growth hormone secretion negatively . However, GRAB does not affect cell morphology as shown here for Rabin8 . The membrane traffic route regulated by Rab8 is still far from clarified. There is, however, evidence that Rab8 modulates polarized delivery of membrane proteins through reorganization of actin and microtubules . This is further supported by our results showing also that Rabin8 participates in the process mediating polarized transport of Rab8-specific vesicles to the cell surface. Interestingly, the corresponding Rabin8 homolog in yeast Sec2p, an exchange factor for the small GTPase Sec4, is required for polarized transport of secretory vesicles . The carboxy termini of both Rabin8 and Sec2p are essential in mediating polarized membrane transport , even although this region is not needed for in vitro binding to Rab8 or Sec4p (; see above). The carboxy terminus may target Rabin8 and Sec2p to membrane domains that contain the small GTPases in their inactive form, perhaps via a membrane-bound receptor . We show that Rabin8 localizes to cortical actin at the plasma membrane and that endogenous Rab8 has perinuclear localization in confluent cells, whereas migrating cells have a polarized distribution of Rab8 in the tips of protrusions. Our results with Rab8 T22N and Q67L mutants suggest that GDP-bound Rab8 is perinuclear, and GTP-bound Rab8 is localized to the plasma membrane and on exocytic vesicles . It has been generally accepted that the nucleotide exchange of Rabs takes place on the donor compartment, but Rabin8 is mainly confined to the acceptor domain (plasma membrane). However, coexpression of Rabin8 with Rab8-GDP relocalizes Rabin8 from the plasma membrane to vesicles that move to specific surface domains at the plasma membrane. This favors the idea that activation of Rab8 by Rabin8 is after all taken place on the donor compartment (vesicles). Whether Rabin8 is recruited from the cytoplasm or reach the vesicles through a membrane mediated recycling process is not known. Our hypothesis does not exclude the possibility that Rabin8-mediated activation of Rab8 might also occur on the PM, especially in processes mediating actin reorganization at the PM. It is unclear whether Rabin8 is needed for the initial transport of Rab8-specific vesicles along microtubules to actin filaments. One possibility is that Rabin8 activates Rab8 at several points during transport to the cell surface. This could allow for the recruitment of different Rab8 effectors during vesicle generation, movement, and fusion. Alternatively, other Rab8-specific GEFs may be involved. What is the function of Rab GEFs? Are they only needed for the activation of Rab proteins, or do they have other roles? Rabin8 is a coiled-coil protein that can self-associate, and we found that Rabin8 does not seem to bind actin even although it is closely connected to cortical actin (unpublished data). Rabin8 is likely to be indirectly associated with actin, perhaps through myosins. Interestingly, Rab8 colocalize with the expressed tail of myosin-Vc in HeLa cells . Alternatively, Rabin8 could bind to some lipid or lipid-binding protein at the membrane that participate in actin dynamics . Given that Rabin8 binds to Rab8 at the donor vesicle, it is possible that vesicular Rabin8 could bind an actin-associated Rabin8 at the plasma membrane. Rabin8 self-association could perhaps promote vesicle movement along cortical actin filaments to an appropriate plasma membrane site for fusion. This model is supported by the findings that cytochalasin D promotes Rabin8 aggregation to actin patches that could inhibit appropriate targeting of Rab8-specific vesicles to the plasma membrane. Finally, we do not yet know whether Rabin8-induced actin reorganization is coupled to activation of Rab8 or linked to some other signal pathway. However, we prefer the former alternative, because also Rab8-GTP induces actin reorganization. Several reports now point to an important role of Rab8 in regulating cell morphogenesis and fate. First, expression of activated Rab8 or Rabin8 promotes reorganization of actin and cell shape (; ; ; ; this report). Furthermore, mutant Rab8 causes cell death of transgenic Xenopus rods and depletion of Rab8 from primary neurons inhibits neurite outgrowth . Second, Rab8 interacts with FIP-2 (optineurin) that also modulates cell morphogenesis and links Rab8 to the huntingtin protein . Interestingly, mutations in optineurin cause primary open-angle glaucoma . In ADPKD (autosomal dominant polycystic kidney disease) cells loss of polarity is associated with redistribution of Rab8, suggesting a role of Rab8 in controling epithelial polarity . Why is Rab8 linked to processes mediating cell morphogenesis? A likely explanation is that Rab8 regulates a pathway where a specific membrane traffic route participates in remodeling the cell shape in response to different signals. Rab8 is known to interact with the germinal center kinase, a protein involved in TNF-alpha signaling . There is also a connection of the Rab8-interacting protein optineurin (FIP-2) to TNF-alpha --mediated signals . TNF-alpha is a major mediator of inflammatory responses, and it induces changes in several cellular processes, such as cell migration, differentiation, necrosis, respiratory burst, and adhesion . TNF-alpha -- and phorbol ester --induced respiratory burst is known to increase exocytosis of granules containing NADPH oxidase and adhesion molecules . The fact that phorbol induced actin reorganization is associated with translocation of both Rabin8 and Rab8-specific vesicles to the periphery of lamellipodia argues for a possible role of Rab8 in regulating respiratory-associated exocytosis. Alternatively, in cells lacking respiratory burst, Rab8 could regulate a TNF-alpha --induced membrane traffic route mediating recycling of adhesion receptors that participate in actin reorganization and the formation of new cell surface domains . We show here that a search for potential Rab8 activators resulted in the discovery of a Rab8-specific GEF that like Rab8 participates in processes mediating cell morphogenesis. Future studies concentrating on finding new interacting molecules for both Rab8 and Rabin8 will certainly open the way to unravel the cellular function of the Rab8 modulated transport route. Backmatter: PMID- 12223127 TI - Risk factors for breast cancer in Iran: a case-control study AB - Abstract | Background | Iranian breast cancer patients are relatively younger than their Western counterparts. The objective of the present study was to investigate risk factors for breast cancer in Iranian women. Method | A case-control study was conducted from April 1997 to April 1998 in Tehran, Iran. Demographical data and risk factor related information were collected using a short structured questionnaire. Odds ratios (ORs) and 95% confidence intervals (CIs) were derived from logistic regression analysis. Results | In all, 286 women with breast cancer and 249 control women were interviewed. In multivariate analysis, only marital status (never married: OR 4.24, 95% CI 1.70-10.57 [P = 0.002]; widowed/divorced: OR 1.71, 95% CI 1.05-2.68 [P = 0.03]) and family history (positive family history of breast cancer: OR 2.95, 95% CI 1.15-7.59 [P = 0.02]) were associated with significantly increased risk for breast cancer. Conclusion | The findings of the present study suggest that family history and marital status may have an impact on the incidence of breast cancer in Iranian women. Keywords: breast cancer, case-control, Iran, risk factor, breast cancer, case-control, Iran, risk factor, Introduction : Geographical variations in incidence and mortality rates of breast cancer suggest that the known risk factors for breast cancer may vary in different parts of the world and that environmental factors are of greater importance than genetic factors . For instance, in Iran it has been shown that, even after adjusting for age, young women are at relatively higher risk for developing breast cancer than are their Western counterparts . Thus, study of risk factors for breast cancer in Iranian women is important and might contribute to current knowledge on this important topic. The aim of the present study was to examine the relationship between reproductive and other risk factors for breast cancer in Iranian women. Method : A case-control study was conducted from April 1997 to April 1998 (one complete calendar year) in Tehran, Iran. The cases were incident, newly diagnosed breast cancer patients living in Tehran, and they were entered into the study if they had a confirmed pathological breast cancer diagnosis and were admitted to the Imam University Hospital (a large teaching and general hospital in the central district of Tehran) for breast surgery. The control women were recruited from female patients living in Tehran without any history of breast problems or neoplastic disease admitted to the same hospital for a wide spectrum of general surgical procedures. Women with hysterectomy and artificial menopause were excluded from the study. Although the hospital does not have an ethics committee, in addition to a written permission from hospital authorities, verbal consent was obtained from all participants. All women were interviewed by one investigator (MV). Demographical and risk factor data were collected using a short structured questionnaire, including information on age, educational level, marital status, family history of breast cancer (first-degree relatives), age at menarche, parity, oral contraceptive use, age at first and last full-term pregnancy, menopausal status, and age at menopause. Women were classified as menopausal if they had not menstruated during the 6 months before the date of data collection. A full-term pregnancy was a pregnancy lasting 8 months or longer. Univariate logistic regression analysis was performed to calculate odds ratios (ORs) and to examine the predictive effect of each factor on risk for breast cancer. P < 0.05 was considered statistically significant. Those risk factors that were significantly associated with breast cancer were entered into a forward selection multivariate logistic regression analysis. Results : In all, 286 women with breast cancer (from the total of approximately 300 women who were newly diagnosed as having breast cancer during the study period) and 249 control women were interviewed. The mean +- SD age of cases and controls was 47.5 +- 12.5 years (range 24-81 years) and 44.2 +- 13.3 years (range 23-80 years), respectively. There were no significant differences between cases and controls with regard to age (except for the age group = 60 years), educational level, age at menarche, past oral contraceptive use, age at first and last full-term pregnancy, and age at menopause. The results of univariate logistic regression analysis are shown in Table . Compared with married women, never married women were at higher risk for breast cancer (OR 4.25, 95% confidence interval [CI] 1.71-10.57). Breast cancer risk was significantly greater in women with a family history of the disease (OR 2.87, 95% CI 1.13-7.30) and in nulliparous women (OR 2.78, 95% CI 1.45-5.35). However, other variables did not exhibit a significant association with breast cancer except for age 60 years or older (OR 1.87, 95% CI 1.13-3.10). Table 1 | Risk factors for breast cancer derived from univariate analysis On multivariate analysis, in addition to those factors that were found to carry a significant association with breast cancer in the initial analysis, age was also included. This was done because, although the study was not an age-matched case-control study, older age exhibited a significant association on univariate analysis. The final model indicated that only marital status (never married: OR 4.24, 95% CI 1.70-10.57 [P = 0.002]; widowed/divorced: OR 1.71, 95% CI 1.05-2.68 [P = 0.03]) and family history of breast cancer (positive family history: OR 2.95, 95% CI 1.15-7.59 [P = 0.02]) were significantly associated with breast cancer, and that nulliparity and older age were not significant risk factors for breast cancer. Discussion : Breast cancer patients in Iran are relatively young, and the findings presented here suggest that marital status and a positive family history of breast cancer are risk factors for breast cancer in Iran. This is in accordance with other research findings indicating that a positive family history of breast cancer is a strong risk factor for breast cancer at young age , although this has a comparatively small effect on the absolute lifetime incidence of and mortality from breast cancer . However, with regard to the findings from the present study, one may argue that the relatively high proportion of young breast cancer cases in Iran is most likely due a to a young population structure and to a combination of high age at menarche and low age at first pregnancy, which are protective in later life. Evidence from the USA also suggests that, in some Asian subgroups such as the Vietnamese, women diagnosed with breast cancer tend to be younger than those from other racial or ethnic groups, with half of the diagnoses occurring in women younger than 50 years; this needs further exploration. The findings presented here show that never married women were at higher risk for breast cancer. In most studies single and nulliparous married women were found to have a similar increased risk for breast cancer as compared with parous women of the same age . Thus, one may argue that marital status by itself is not a determining factor for increased or reduced breast cancer risk, and rather the main protective effect is from early first full-term pregnancy. However, in the present study no association with parity emerged when multivariate analysis was performed. Evidence suggests that there is an interaction between marital status and parity , supporting a 'dual effect' of parity on breast cancer risk with pregnancy. Studies have shown that interactions between age, family history of breast cancer and parity might exist . In addition, studies have reported that nulliparity reduces risk for breast cancer at younger age and elevates risk in the elderly . The lack of significant associations between breast cancer and the other variables studied was unexpected. For example, studies have shown that past oral contraceptive use is associated with a somewhat higher OR among young women or women who have a family history of breast cancer . However, although there were clear patterns in ORs for most variables , we suspect that this lack of significant associations may be explained by the limited power of the present study to estimate risk. Finally, one should be aware of the limitations of the present study, including case and control ascertainment and representation. Although the results cannot be generalized, the findings suggest that the associations between some known risk factors for breast cancer may differ in Iran as compared with Western countries, and that familial breast cancer in young Iranian breast cancer patients deserves further investigation. Abbreviations : CI = confidence interval; OR = odds ratio. Backmatter: PMID- 12223128 TI - Impact of false-positive mammography on subsequent screening attendance and risk of cancer AB - Abstract | Background | One area of concern within the largely successful UK National Health Service breast screening programme is the relatively high proportion of women showing mammographic abnormalities who undergo further diagnostic tests that prove negative. Previous studies suggest that, in addition to increasing anxiety, such false-positive mammography is associated with increased risk of subsequent interval cancer. In the present article, we quantify this increased risk, investigate whether it extends to cancers detected at rescreening, and determine whether cancers differ between women who have, and have not, experienced false-positive mammography. Methods | This was a retrospective cohort study of 140,387 women aged 49 --63 years routinely invited for first screening by the East Anglian National Health Service breast screening programme. Proportions reattending, and subsequent risk and pathological attributes of cancer were compared between women who underwent further (negative) assessment following false-positive mammography and women mammographically normal at first screen. Results | At first screen, 108,617 (91.9%) of the screened women were mammographically normal, 4278 (3.6%) were assessed and then judged normal, and 514 (0.4%) underwent benign biopsy. Compared with nonassessed normal women, reattendance was lower among assessed women: 83.1% (95% confidence interval [CI], 82.0 --84.1) versus 85.7% (95% CI, 85.5 --85.9) (odds ratio [OR], 0.82; 95% CI, 0.76 --0.89). Assessed women were at greater risk of interval cancer (rate per 1000 women screened, 9.6 [95% CI, 6.8 --12.4] versus 3.0 [95% CI, 2.7 --3.4]; OR, 3.19 [95% CI, 2.34 --4.35]), and also of cancer detected at second screen (rate per 1000, 8.4 [95% CI, 5.8 --10.9] versus 3.9 [95% CI, 3.5 --4.3]; OR, 2.15 [95% CI, 1.55 --2.98]). More cancers in assessed women measured >= 20 mm (OR, 1.59; 95% CI, 0.99 --2.55). Conclusions | Women undergoing false-positive mammography at first screen were less likely to reattend for subsequent screens than were nonassessed women, yet they were more likely to develop interval cancers or cancers at second screen, and their cancers were larger. Factors predisposing for false-positive mammography require investigation. Women should be encouraged to continue with screening. Keywords: breast screening, false-positive mammography, interval cancer, screen-detected cancer, screening attendance, breast screening, false-positive mammography, interval cancer, screen-detected cancer, screening attendance, Introduction : Effective population-based screening requires adequate compliance among the target population . The goal of the UK breast screening programme, a 25% reduction in mortality, required first-screen compliance to be 70% . This target was exceeded in many areas throughout the UK . Another key prerequisite to maintaining a successful screening programme is acceptability, and the >= 90% reattendance rate among previous screenees indicated high acceptability in UK women . Despite meeting most of the targets set, however, breast screening in the UK has not been an unmitigated success story. When screening was introduced there were reports of increased anxiety among the large numbers of women undergoing assessment following positive screening mammography . Although most women do not show abnormalities on their screening mammograms and are simply asked to return for rescreening in 3 years, the mammograms of 5 --9% of all women indicate possible malignancy. These women must undergo a second 'assessment' phase of the screening process, which involves further procedures such as ultrasound, needle biopsy and, possibly, open biopsy to establish a definitive diagnosis. As expected, malignancy is ruled out for the majority of women assessed, indicating that the result of their mammography was falsely positive. In the current study, the term 'false positive' is thus applied to any woman who is recalled for assessment on the basis of mammographic findings and in whom cancer is not diagnosed. Other studies may limit use of the term only to those women who have undergone open biopsy with no resultant diagnosis of cancer. Although some studies have indicated that false-positive mammography does not deter women from reattending , it clearly has associated financial and psychological costs. Furthermore, ourselves and other workers have found that women judged false positive at first screen are more likely to develop an interval cancer before the second screening is due. To assess the impact of false-positive mammography on screening effectiveness, we investigate whether false-positive mammography affects subsequent reattendance in East Anglia. We quantify the magnitude of the increased risk of interval cancer, and extend these investigations to determine whether false-positive mammography at first screen increases the risk of cancer detected at second screen. Finally, we compare the pathology of cancers presenting in women who have undergone false-positive mammography with that of cancers in women judged normal at the preceding screening. Methods : Screening was introduced in East Anglia over the period 1989 --1991. Women were invited by year of birth, in 5-year age groups . The first round of screening occurred in 1989 --1995, and the second occurred 3 years later (1992 --1998). All women invited to the first round, and eligible for reinvitation to the second, were identified on the breast screening computer system. The majority was aged 50 --62 years. Those aged 63 --64 years were beyond the invited age range at the time of the second screen and were excluded. We only included women invited on schedule (i.e. within 24 months of commencing invitation of the relevant birth cohort) in order to exclude women who failed to respond at first invitation but who attended for the first time on reinvitation 3 years later . Women were followed up from the date of first invitation until invitation to the second screening if within 3.5 years (i.e. 1277 days), or for the mean interscreen interval for the screening unit if no screen occurred within this time (mean +- standard error interscreen interval for all regional screening units, 1106 +- 7.6 days). Interval and screen-detected cancers were identified by matching screening unit and Cancer Registry databases as described previously . Women not attending within 6 months of the first invitation were deemed to have refused screening. With the exception of women attending for the first invitation within 6 months but placed on early recall, only those women whose second screening episode (either accepted or refused) was completed within 3.5 years of the first invitation are considered in the present analysis. Information on prognostic characteristics size (maximum diameter of invasive component), grade , node status and histological subtype was obtained from the Cancer Registry, from screening units and from medical records. Cancers were assigned a prognostic risk group , based on histological type and malignancy grade . Differences in distribution of these attributes, and of the risk of cancer, among study groups were expressed as ORs . Study groups : Table presents the relationship of the study groups to the total population of 203,194 women eligible for invitation to the first screening round. Of these women, 140,387 (69.1%) were identified as first-round invitees who were invited within schedule and were eligible for reinvitation to rescreening. Table 1 | Relationship of study groups to total population of women eligible for and invited to the first screening round in East Anglia Of these 140,387 women, 118,216 (84.2%) completed a satisfactory first screen within 6 months of the first appointment. At first appointment, 108,617 (77.4%) were judged normal on the basis of mammography alone (the 'non-assessed normal' group). Of 6460 (4.6%) women who underwent further assessment to establish definitive diagnosis, 6337 were investigated immediately; the remainder underwent further appointments for technical reasons or through failure to respond to invitations. Of those assessed immediately, 4278 women were judged normal at assessment ('assessed normal' group), and 514 proceeded to open biopsy, the result of which was benign ('assessed benign' group). Women in assessed normal and assessed benign groups returned directly to routine screening. Overall, there were 113,409 women in the nonassessed normal, assessed normal and assessed benign study groups, comprising 80.8% of all women who were invited for a routine first screen within schedule and were eligible for reinvitation to the second round. The mean +- standard deviation age of women was similar in all study groups: 56.0 +- 3.5 years in the entire invited cohort eligible for reinvitation, 55.8 +- 3.5 years in the nonassessed normal group, 56.1 +- 3.5 years in the assessed normal group, and 56.4 +- 3.6 years in the assessed benign group. Of the 6337 women assessed immediately following the first screen, 680 were initially diagnosed with cancer. A further 11 women were diagnosed following initial nonattendance at assessment, and 22 at early recall. This produced a total of 713 women. Of the 680 women diagnosed at initial assessment, 676 were registered at the Cancer Registry. The remaining four women were either recurrences of earlier breast primaries or other malignancies not classified as primary breast cancer (e.g. cytosarcoma phyllodes) . Results : Numbers of cases, rates per 1000 screened, and odds of presenting with interval cancer or having a cancer detected at second screen are presented by study group in Table . In all study groups combined, 375 interval cancers presented and 463 cancers were detected at second screen. The risk of subsequent interval cancer, or cancer detected at second screen, was higher in women assessed at the preceding screen. The OR among all assessed women for an interval cancer arising was 3.19 (95% CI, 2.34 --4.35) compared with the nonassessed normal group, and that for detecting a cancer at second screen was 2.15 (95% CI, 1.55 --2.98). Compared with the nonassessed normal group, the OR of a second round screen-detected cancer in the assessed benign group was lower, but the difference was not significant, possibly due to the small size of the assessed benign group (OR, 0.50; 95% CI, 0.07 --3.55). Table 2 | Likelihood of presentation of interval cancer following an initial screen, and of detection of cancer by screening at the second screening round, by assessment status Numbers and proportions of women with appointments for, and attending, the second screen are presented, by study group and interval cancer status, in Table , which also presents the ORs for the likelihood of recorded appointments and reattendance. Explanations for a lack of recorded reinvitation appointment include deletion of appointments cancelled in advance, and removal of women from the invitation list following moving house, bilateral mastectomy or death. Overall, 90.6% of women had recorded appointments for a second screen within 3.5 years, and 85.6% reattended. Compared with the nonassessed normal group, the proportion with recorded appointments was slightly lower among the assessed normal group (OR, 0.78; 95% CI, 0.71 --0.86) and was lower still among the assessed benign group (OR, 0.51; 95% CI, 0.41 --0.65) (see Table ). Reattendance was similarly slightly lower among women in the assessed normal group (OR, 0.84; 95% CI, 0.78 --0.92) compared with the nonassessed normal group, and was lower still among the assessed benign group (OR, 0.65; 95% CI, 0.52 --0.81). Reattendance among women with recorded reinvitation appointments was similar in all groups (94 --95%, calculated from data in Table ). Table 3 | Numbers and proportions of cases of interval cancer and second round screen-detected cancer, and likelihood of presentation, by study group and interval cancer status Reattendance was much lower among women presenting with interval cancer following the first screen. Combining assessed and nonassessed women, 90.8% of women without interval cancer had reinvitation appointments recorded, and 85.8% actually reattended, whereas only 27.2% of women with interval cancer had reinvitation appointments, and 19.2% reattended (see Table ). Among the 680 women diagnosed with cancer immediately following first screen, similarly low proportions with reinvitation appointments (250 of 680 women [36.8%]) and reattending (224 of 680 women [32.9%]) were observed (data not shown). The effects of the study group on pathological attributes of interval cancers and second round screen-detected cancers are presented in Table . Interval cancers and cancers detected at second screen in assessed women were larger compared with cancers in nonassessed women, with more measuring at least 20 mm (OR, 1.59; 95% CI, 0.99 --2.55). This effect reached statistical significance in the assessed normal group (OR, 1.63; 95% CI, 1.01 --2.64). There were no significant differences between assessed and nonassessed women in risk of high grade (grade 3) cancers. Compared with the nonassessed normal group, there were some indications of fewer high grade interval cancers among those assessed (OR, 0.47; 95% CI, 0.17 --1.28), but this was not statistically significant. Cancers detected at second screen in women who were assessed at first screen showed some indication of increased risk of positive nodes (OR, 1.51; 95% CI, 0.67 --3.39). There was a non-significant increase in numbers of advanced stage (stage 2+) interval and second round screen-detected cancers in assessed women (OR, 1.32; 95% CI, 0.83 --2.11). Interval cancers in the assessed normal group showed a slightly increased tendency to present in the left breast (OR, 1.51; 95% CI, 0.77 --2.95). For interval cancers, the likelihood of a cancer being of high prognostic risk (group 3) was significantly lower in the assessed normal group than in the nonassessed group (OR, 0.34; 95% CI, 0.12 --0.99). Table 4 | Effect of study group on pathological attributes of interval cancers and second round screen-detected cancers Discussion : False-positive mammography is a relatively frequent occurrence within breast screening programmes. It was calculated that, during 10 years in New England, USA, one-third of women screened by mammography and clinical examination underwent false-positive screening . Last year in the UK there were around 11 unnecessary recalls for every cancer detected within the National Health Service breast screening programme (8.3% of women assessed at first screen, 6.7 cancers detected per 1000 screened; 7.6% false-positive at first mammography) . This figure fell to around six recalls for every cancer detected at repeat screening (3.9% of women assessed, 5.5 cancers detected per 1000 screened; 3.4% false-positive at subsequent mammography) . For a woman attending all five screens to which she is currently invited, assuming the result of each screening is independent of any previous result, her cumulative probability of at least one false-positive screening is therefore around 19% {1-(P [No. false positives on first screen] x P [No. false positives on subsequent screen]4)}, i.e. {1-([1-0.0763] x [1-0.0335]4)}. In East Anglia, 4.1% of women screened in the first screening round and eligible for a second screen underwent false-positive mammography at first screen. These women were more than three times as likely as nonassessed normal women to present with an interval cancer before the second screen was due, and more than twice as likely to have a cancer detected at second screening. If the followup period of the current study had been extended beyond the interscreen interval, these women might have continued to show increased risk of cancer beyond the due date for second screen. The rate in lapsed attenders who were false positive at first screen is thus also likely to have been high. For women who undergo false-positive mammography and then present with cancer, the validity of the negative assessment comes into question. Among women in the current study, 12.3% of those presenting with interval cancer after the first screen, and 8.6% of those with cancer detected at second screen, had previously been assessed compared with the 4.2% of women assessed of those not diagnosed with cancer. A review of the original screening films of women with interval cancers has shown that around one-fifth of all East Anglian cases might potentially be prevented through earlier diagnosis at the previous screening . However, quantification of the extent to which failure of diagnosis at assessment has contributed to the interval cancer rate requires detailed comparison of the site and nature of relevant lesions at assessment and diagnosis, which was beyond the scope of the current study. Another possible explanation for the increased risk of cancer in women following false-positive mammography might be that a characteristic of women's breasts which makes them difficult to interpret mammographically and predisposes them to the risk of a false-positive result is actually itself a risk factor for breast cancer. Such a link between risk of false-positive mammography and risk of cancer might be hormone replacement therapy (HRT). Laya et al. in 1996 , and others since , demonstrated that current use of HRT by women aged 50+ years is associated both with decreased screening sensitivity (increased interval cancers) and decreased specificity (increased false positives). Furthermore, long-term use of HRT has been shown to increase breast cancer risk . It has been proposed that these effects of HRT are mediated through high risk (i.e. dense) mammographic patterns , which themselves have been shown to be associated both with reduced screening specificity and with increased risk of breast cancer . Unfortunately, information on HRT use among women in the current study was not available to test this hypothesis. If found relevant, however, consideration of a woman's individual risk profile might prove helpful in deciding a strategy for subsequent rescreening. In addition to being at increased risk of cancer, women who experienced false-positive mammography were 18% less likely than those who were genuinely screen normal to reattend for a second screening. Women who underwent benign biopsy were even less likely (35%) to return. This finding was in contrast with reports from previous studies when false-positive mammography either had no effect on subsequent attendance or, alternatively, actually increased by 20% the likelihood of future reattendance , even in those who underwent a negative biopsy . As previously suggested , however, regional variations in handling of assessment procedures and false-positive cases may explain differences in uptake of screening reinvitations. Such findings suggest that the figures may be amenable to improvement through a change in practice. The reduced reattendance among women with false-positive mammography was accompanied by a reduced likelihood of reinvitation appointments recorded on the screening computer system. Followup of the 85 assessed benign women who did not have reinvitation appointments recorded on the computer system revealed that at least 35 of them (41%) had been offered reinvitation appointments, which they had cancelled in advance. The reduced reattendance among assessed women would not have been apparent had the analysis investigated the proportions of women with recorded reinvitation appointments who returned, since reattendance among those with recorded reinvitation appointments was similarly high in all women (around 94%), regardless of assessment history. Furthermore, such analysis would have failed to reveal the correct magnitude of the increased risk of cancer among women undergoing false-positive mammography, since the OR for an interval cancer arising in all assessed women versus that in nonassessed women was 3.19 (95% CI, 2.34 --4.35). This OR fell to 1.21 (95% CI, 0.49 --2.97) when only those women with recorded reinvitation appointments within 3.5 years were considered (data not shown). Regardless of whether the lack of recorded reinvitation appointments is due to an active choice on the part of women not to attend or due to a failure of the system to reinvite them, the fact that they are not being rescreened yet they are at increased risk of cancer is clearly of great concern. False-positive mammography, leading to unnecessary assessment of disease-free women, has associated costs. First, there are the psychological costs of inconvenience and increased anxiety in women unnecessarily recalled . Second, there are the direct financial costs to the health service of unnecessary procedures . Third, there are the overall costs to the invited population of the reduced effectiveness of screening . East Anglian women undergoing false-positive first mammography were more likely to present with cancer and less likely to reattend for a second screen. The impact of nonreattendance on potential subsequent mortality reduction is not insignificant since, for those who do not reattend, the stage of any subsequently diagnosed cancers will be shifted from earlier, when detected by screening, to later, when presenting symptomatically in lapsed attenders. For women who experienced false-positive mammography and then presented with an interval or screen-detected cancer, the current study indicated that these cancers were more likely to measure 20+ mm and to be of a higher stage than those in nonassessed women, and second screen-detected cancers were more likely to be node positive. This indicates a poorer prognosis for women who presented with cancer having undergone false-positive mammography, although when assessed by histological subtype and grade such cancers were apparently of lower prognostic risk than those in nonassessed women. The reasons for this inconsistency are unclear. The current findings relate to women assessed at first screen within the East Anglia screening programme, carried out over the period 1989 --1995. With technical improvements to the programme and women's increased familiarity with it, reduced reattendance and increased risk of cancer may be less associated with false-positive mammography at first screen. Furthermore, the impact of false-positive mammography may be lower after a second or subsequent screening. Fewer women are assessed at second and subsequent screens , possibly due to the increase in specificity associated with the availability of previous films for comparison . Finally, it should be noted that, while there was increased risk of subsequent cancer associated with women assessed at first screen, such women contributed only a modest proportion of all interval and second round screen-detected cancers. Conclusions : False-positive mammography in the first screening round in East Anglia was associated both with increased risk of interval cancers and cancer detected at second screening, and with reduced reattendance at subsequent screens. If these associations persist within the screening programme, then efforts must be made to identify the factors predisposing certain women to false-positive mammography, to encourage continued participation, and to detect any subsequent cancers at the earliest possible opportunity. Abbreviations : CI = confidence interval; HRT = hormone replacement therapy; OR = odds ratio. Backmatter: PMID- 12223126 TI - Tumour Fas ligand:Fas ratio greater than 1 is an independent marker of relative resistance to tamoxifen therapy in hormone receptor positive breast cancer AB - Abstract | Background | The objective of the present study was to examine the prognostic and predictive significance of the apoptosis-related marker Fas ligand (FasL):Fas ratio in breast cancer. Methods | Tumour biopsies from 215 primary invasive breast cancer patients were examined for the expression of FasL and Fas mRNA transcripts by quantitative real-time RT-PCR. Their prognostic and predictive impact on patient survival was determined in univariate and multivariate survival analyses. Results | Using a cutoff value of 1, a FasL:Fas ratio greater than 1 was found to have significant prognostic value for disease-free survival among the total population (median follow up 54 months). It was associated with a significantly decreased disease-free survival (P = 0.022) and with a tendency toward increased mortality (P = 0.14) in univariate analysis. Hormone receptor positive women exclusively treated with tamoxifen (n = 86) and with a FasL:Fas ratio greater than 1 had a significantly decreased disease-free survival (P = 0.008) and overall survival (P = 0.03) in univariate Kaplan --Meier analysis. Furthermore, tumour size and FasL:Fas ratio were of independent predictive significance in the multivariate model for disease-free and overall survival in that subgroup. Among postmenopausal patients (n = 148) both of those factors retained independent prognostic significance in the multivariate model for disease-free survival. In contrast, FasL:Fas ratio had no significant predictive value in patients exclusively treated with chemotherapy. Conclusion | The data presented indicate that FasL:Fas ratio may be useful not only as a prognostic factor but also as a predictive factor for projecting response to the antioestrogen tamoxifen. The results strongly support a correlation between FasL:Fas ratio greater than 1 and lack of efficacy of tamoxifen in hormone receptor positive patients. Keywords: apoptosis, breast cancer, Fas ligand, predictive factor, tamoxifen, apoptosis, breast cancer, Fas ligand, predictive factor, tamoxifen, Introduction : The Fas receptor (CD95/APO-1) is a widely expressed glycosylated cell surface molecule of approximately 45 --52 kDa. It is a type I transmembrane receptor and can also occur in several soluble forms. Under physiological conditions Fas-mediated apoptosis is triggered by Fas ligand (FasL/CD95L), the natural ligand of the receptor. FasL was shown to be a tumour necrosis factor related type II transmembrane molecule, and its expression is more restricted than that of Fas . FasL is predominantly detected in activated T cells and activated natural killer cells, but it is also found on some macrophages/monocytes and on cells of immune privileged tissues . Although there is evidence that some cancers express FasL on their cell surface, creating an immune privileged site , a number of laboratories have failed to confirm these observations . These findings, in addition to a report that tumours transfected with genes that express FasL (which was originally predicted to accelerate engraftment) were unexpectedly rejected , have confounded interpretation of the earlier studies and the role of FasL in immune recognition of malignancies . In invasive breast carcinomas the expressions of Fas and FasL appeared to be inversely related . A significant association was observed between FasL and the presence of metastatic lymph nodes and larger tumour size, whereas Fas expression was correlated with node-negative status and smaller tumour size . Furthermore, high FasL mRNA and protein expression appears to be positively correlated with histological grading , suggesting that a neoplastic Fas-negative/FasL-positive phenotype may be linked to breast cancer progression. However, Ragnarsson et al. concluded that FasL is expressed intracellularly in both normal and malignant breast epithelium and is unlikely to be important for the immune evasion of breast tumours. We previously established the FasL:Fas ratio in tumour tissue as a prognostic factor in breast carcinomas . An extended follow up allows a re-evaluation of our database regarding the prognostic and predictive value of FasL:Fas ratio in breast carcinomas. In particular, we investigated for the first time whether the FasL:Fas ratio may predict response to cytotoxic or hormonal therapies, in the same way that the presence of hormone receptors predicts response to hormonal treatment. Using data obtained by analysis in various subgroups, the findings presented here contribute to the recent controversial discussion on the effects of tumour-expressed FasL in vivo. Materials and method : Patients and tissue collection | Tissue samples from 215 unselected primary breast carcinomas were collected during surgery at the Department of Obstetrics and Gynaecology of the University of Rostock between 1994 and 1998. None of the patients had received chemotherapy, radiotherapy or immunotherapy before surgery. The mean age for breast cancer patients was 58 years (range 29 --90 years). The majority of breast tumours (n = 199 [92.5%]) were invasive ductal carcinomas. The median follow-up time was 54 months (range 2 --77 months). The main classifying prognostic factors are shown in Table . A total of 200 patients (93%) received postoperative adjuvant therapy, consisting of chemotherapy (n = 52), hormone therapy (n = 96), or both (n = 52). Among patients undergoing endocrine treatment 90 women (93.7%) were treated with tamoxifen, receiving an oral dose of 20 mg/day tamoxifen for a maximum of 5 years. Among the tamoxifen-treated group, we excluded four patients with hormone receptor negative tumours from further survival analysis. Of the remaining 86 patients, 41 blood samples were available for detection of serum oestradiol concentrations using the SR1 enzyme immunoassay (BioChem Immunosystems GmbH, Freiburg, Germany). Table 1 | Clinicopathological characteristics of 215 breast cancer patients and univariate analysis of established prognostic factors for disease-free and overall survival Real-time reverse transcription polymerase chain reaction | The RT-PCR procedure and preparation of RNA standard were described previously . In brief, total RNA of freshly frozen breast tissue was prepared in accordance with the acid guanidium thiocyanate-phenol-chloroform protocol. All specimens were tested by analysis of glyceraldehyde-3-phosphate dehydrogenase housekeeping gene expression using conventional RT-PCR. First-trimester placenta (FasL) from women undergoing legal abortions and liver mRNA (Fas) served as positive controls. The primer pairs and probes were designed using the Primer ExpressTM 1.0 program (PE Applied Biosystems, Foster City, CA, USA) and were obtained from Applied Biosystems GmbH (Weiterstadt, Germany). The primers yielded RT-PCR products of 82 (FasL) and 105 (Fas) nucleotides. For calibration of the FasL and Fas TaqManTM assays, two RNA standards were generated using an in vitro T7-polymerase transcription system (RiboMAXTM Large Scale RNA Production System; Promega, Madison, WI, USA). The TaqMan EZ RT-PCR Kit (PE Applied Biosystems) was used for RT and amplification of both targets and standards. Production of cDNA and PCR amplification were carried out in a single-tube, single-enzyme system, without addition of subsequent enzymes or buffers. All RT-PCR reactions were performed in duplicate, with a final volume of 25 mul. Statistical analysis | Clinical, histological and biological parameters were compared using the most appropriate test from among the Spearman correlation coefficient, chi2 test, Mann --Whitney U test and Kruskal --Wallis H test. Disease-free and overall survival were analyzed using the Kaplan --Meier method, and comparison of study groups was performed using the log-rank test. The Cox regression model was applied over both univariate and multivariate analyses, with the associated likelihood ratio test used for tests of trend differences. In multivariate analysis, a backward stepwise selection procedure was used. The Statistical Package for Social Science (SPSS Inc., Chicago, IL, USA) was used to conduct statistical analysis. In all tests P < 0.05 was considered statistically significant and all were two-tailed tests. Results : Analysis of survival among the total study group | The impact of established prognostic parameters regarding disease-free and overall survival for the total population is listed in Table . Tumour size (P = 0.001) and lymph node status (P = 0.009) were significant factors for disease-free survival, whereas tumour size (P < 0.001) and histological grading (P = 0.003) were significant prognostic parameters for overall survival within univariate analysis. Using the cutoff value of 1, a FasL:Fas ratio greater than 1 was found to be a significant prognostic factor for disease-free survival in the present update analysis. It was associated with a significantly decreased disease-free survival (P = 0.022; Fig. ) and with a tendency toward increased mortality (P = 0.14; Fig. ). The mean disease-free survival time for patients with a ratio of 1 or less was 67 months (95% confidence interval [CI] 63 --70 months) versus 59 months (95% CI 53 --65 months) for patients with a ratio greater than 1. Figure 1 | (a) Disease-free and (b) overall survival curves for breast cancer patients (n = 211) with respect to Fas ligand (FasL):Fas ratio. (a) Disease-free and (b) overall survival curves for breast cancer patients (n = 211) with respect to Fas ligand (FasL):Fas ratio. Survival was analyzed using the Kaplan --Meier method, and comparison of study groups was performed using the log-rank test. All variables that achieved significance in the univariate survival analysis were incorporated in a multivariate analysis . Tumour size (P = 0.004) and lymph node status (P = 0.04) were of independent significance in the multivariate disease-free survival analysis (n = 208). Furthermore, tumour size (P = 0.001) and grading (P = 0.023) were independent prognostic parameters for overall survival in the multivariate model (n = 208). Table 2 | Multivariate Cox regression analysis for disease-free and overall survival, including prognostic factors that were significant on univariate analysis Analysis of survival among tamoxifen-treated patients | Ideally, a predictive factor should be evaluated only in the setting of individuals who have not received different systemic therapies. For this reason we divided the total population into hormone receptor positive tamoxifen-treated (n = 86) and chemotherapy-treated (n = 52) subgroups. Among the tamoxifen-treated subgroup 75 patients (83.3%) were node negative. The FasL:Fas ratio values were positively correlated with histological grading (P = 0.048) and serum oestradiol concentrations (P = 0.04). Women with a FasL:Fas ratio greater than 1 had significantly higher serum oestradiol levels than did women with a FasL:Fas ratio of 1 or less (P = 0.002). No associations between FasL:Fas ratio and other factors such as tumour size (P = 0.129), lymph node status (P = 0.378), age (P = 0.822), body mass index (P = 0.174) and use of exogenous hormones (P = 0.732) were observed. Women exclusively treated with tamoxifen (n = 86) and with a FasL:Fas ratio greater than 1 had a significantly decreased disease-free survival (P = 0.0084) and overall survival (P = 0.03) on univariate Kaplan --Meier analysis . The mean disease-free survival time for tamoxifen-treated patients with a ratio of 1 or less was 71 months (95% CI 67 --75 months) versus 57 months (95% CI 48 --67 months) for patients with a ratio greater than 1. The corresponding data for overall survival were 74 months (95% CI 70 --77 months) for a ratio of 1 or less and 66 months (95% CI 58 --73 months) for a ratio greater than 1. Tumour size (P = 0.013) and FasL:Fas ratio (P = 0.013) were significantly related to reduced disease-free survival in univariate Cox regression analysis . Tumour size (P = 0.023) and FasL:Fas ratio (P = 0.027) were also of independent prognostic significance in the multivariate model (n = 86). The relative risk for occurrence of breast cancer related relapse in tamoxifen-treated patients was 3.0 (95% CI 1.14 --7.9) for women with a FasL:Fas ratio greater than 1. Figure 2 | (a) Disease-free and (b) overall survival curves for tamoxifen-treated, hormone receptor positive breast cancer patients (n = 86) with respect to Fas ligand (FasL):Fas ratio. (a) Disease-free and (b) overall survival curves for tamoxifen-treated, hormone receptor positive breast cancer patients (n = 86) with respect to Fas ligand (FasL):Fas ratio. Survival was analyzed using the Kaplan --Meier method, and comparison of study groups was performed using the log-rank test. Table 3 | Cox regression analysis for disease-free and overall survival in tamoxifen-treated, hormone receptor positive patients Corresponding analyses for overall survival in tamoxifen-treated patients revealed prognostic impact of tumour size (P = 0.002), tumour grading (P = 0.015) and FasL:Fas ratio (P = 0.042) in the univariate setting. The multivariate analysis confirmed all three of those parameters as independent markers for overall survival in tamoxifen-treated patients . Analysis of survival among postmenopausal patients | The tamoxifen-treated subgroup (age range 50 --90 years) comprised 94.5% postmenopausal and 5.5% premenopausal patients, and we therefore re-analyzed the data for the postmenopausal population (n = 149). On univariate Kaplan --Meier analysis, postmenopausal women with a FasL:Fas ratio greater than 1 had decreased disease-free survival (P = 0.01) and overall survival (P = 0.182) as compared with postmenopausal women with a ratio of 1 or less . The mean disease-free survival time for postmenopausal patients with a ratio of 1 or less was 67 months (95% CI 62 --71 months) versus 56 months (95% CI 49 --63 months) for patients with a ratio greater than 1. The corresponding data for overall survival were 69 months (95% CI 65 --73 months) and 65 months (95% CI 59 --71 months), respectively. Tumour size (P = 0.001) and FasL:Fas ratio (P = 0.01) were significantly related to reduced disease-free survival on univariate Cox regression analysis . Figure 3 | (a) Disease-free and (b) overall survival curves for postmenopausal breast cancer patients (n = 149) with respect to Fas ligand (FasL):Fas ratio. (a) Disease-free and (b) overall survival curves for postmenopausal breast cancer patients (n = 149) with respect to Fas ligand (FasL):Fas ratio. Survival was analyzed using the Kaplan --Meier method, and comparison of study groups was performed using the log-rank test. Table 4 | Cox regression analysis for disease-free and overall survival in postmenopausal patients Both factors (tumour size P = 0.003; FasL:Fas ratio P = 0.028) retained independent prognostic significance in the multivariate model (n = 148). The relative risk for occurrence of breast cancer related relapse in postmenopausal patients was 2.04 (95% CI 1.14 --7.36) for women with a ratio greater than 1. Among the postmenopausal subgroup no associations between FasL:Fas ratio and other factors such as tumour size (P = 0.096), tumour grading (P = 0.108), lymph node status (P = 0.339), age (P = 0.55), oestrogen receptor (ER; P = 0.741), progesterone receptor (P = 0.748), body mass index (P = 0.92) and use of exogenous hormones (P = 0.26) were observed. Analysis of survival among patients undergoing cytotoxic therapy | In contrast to the findings summarized above, FasL:Fas ratio had no significant prognostic value in patients treated exclusively with chemotherapy. Using Kaplan --Meier analysis, disease-free (P = 0.94) and overall survival (P = 0.765) were not related to FasL:Fas ratio. Similarly, no significant associations were found between FasL:Fas ratio and survival time in premenopausal patients (data not shown). Recurrence pattern | Because of the significance of FasL:Fas ratio for disease-free survival in postmenopausal or tamoxifen-treated women, we analyzed the pattern of locoregional and distant relapse. Among the 149 postmenopausal patients, 39 (26.2%) had recurrent disease. The relapse rate was 39.3% (n = 22) in 56 patients with a ratio greater than 1 and 18.3% (n = 17) in 93 women with a ratio of 1 or less. Skeletal (41.2%) and pulmonary metastases (23.5%), and locoregional progression (23.5%) were frequently observed in postmenopausal patients with a ratio of 1 or less. Liver (33.3%) and skeletal metastases (22.2%), and locoregional progression (22.2%) occurred predominantly in patients with a ratio greater than 1. However, no specific recurrence patterns were detected with respect to FasL:Fas ratio. Discussion : At present there are only two factors with clinical level 1 evidence supporting a predictive role for response to breast cancer therapy. First, optimal adjuvant hormonal therapy is 5 years of tamoxifen for anyone with a tumour that is hormone receptor positive, even if at a low level. Patients whose cancers express neither ER nor progesterone receptor will not be helped by tamoxifen administration . Second, recent data suggest that adjuvant treatment with trastuzumab will prove ineffective for tumours with low levels of HER-2/neu expression . The new data presented here indicate that the FasL:Fas ratio may be useful not only as a prognostic factor but also as a predictive factor for projecting response to the antioestrogen tamoxifen. In patients with hormone receptor positive primary tumours, a FasL:Fas ratio greater than 1 defines a subgroup of patients who are less likely to respond to tamoxifen. Women exclusively treated with tamoxifen and with a FasL:Fas ratio greater than 1 had a 14-month shorter mean disease-free survival and a relative risk of 3.0 for relapse of disease as compared with patients with a FasL:Fas ratio of 1 or less. Furthermore, a ratio greater than 1 was associated with a significantly shorter overall survival (P = 0.043) and a relative risk for death of 3.65 in that subgroup. No correlation was seen between FasL:Fas ratio and survival time in patients receiving cytotoxic treatment. The unexpected finding that FasL:Fas ratio is a prognostic factor in postmenopausal and a predictive factor in tamoxifen-treated patients may be accounted for by hormone dependency. Various studies suggest that the FasL/Fas system may play an important role in the regulation of cell death in response to hormonal changes. Nilsen et al. showed that oestrogen induces apoptosis in neurone-like cells that express ER-beta, and that this is mediated by the FasL/Fas system. Selam et al. suggested that oestradiol and progesterone may play a role in regulating maternal immunotolerance for the implantation of a semiallograft embryo by inducing FasL expression in human endometrium. Recent findings support the hypothesis that oestrogen-induced thymic atrophy occurs as a result of apoptosis and is mediated by oestrogen-induced FasL expression . Finally, maintenance of endothelial FasL expression by oestradiol may represent a mechanism for the apparently antiatherogenic effect of oestrogen . Novel findings further link the FasL/Fas system and hormones such as oestrogen and lactogenic hormones to the biology of the normal mammary gland and the development of breast cancer . Treatment of LTED cells, which were derived by growing MCF-7 breast cancer cells under long-term oestrogen-deprived conditions, with oestradiol increased the expression of FasL . Mor et al. showed that FasL in breast tissue is functionally active, and that oestrogen and tamoxifen regulate its expression. Those investigators identified an oestrogen-recognizing element-like motif in the promoter region of the FasL gene, suggesting direct effects of oestrogen on FasL expression. This effect is receptor mediated because tamoxifen blocked the oestrogenic effect. The indication for use of antioestrogens in mammary carcinoma arises from the observations that oestrogen is a mitogen in breast cancer and that approximately one-third of patients will respond to endocrine therapy. The main therapeutic antioestrogen that has been used is tamoxifen. The present data indicate that the antitumour effect of tamoxifen varies according to the FasL:Fas ratio of the breast tumour, and values of the ratio are positively correlated to serum oestradiol levels. The use of tamoxifen in patients with a FasL:Fas ratio greater than 1 is less effective than in those with higher ratios, leading to the hypothesis that the recently developed aromatase inhibitors anastrozole, letrozole and exemestan are more appropriate in these patients, or that additional cytotoxic treatment will be of benefit in this subgroup. The role of FasL expressed by breast tumour in the induction of apoptosis of infiltrating Fas-expressing immune cells, and thereby conferring immune privilege on the tumour, is at present unclear . The expression of FasL by tumours implies not only that cancers have acquired defensive strategies (Fas resistance) but also that they can take the offensive (counterattack) . On the basis of the findings presented here, we hypothesize that the association between high-FasL/low-Fas phenotype and shorter disease-free survival in older breast carcinoma patients may be due to a decline in immune function, which results in increased escape of tumour cells from immune surveillance. Of special interest are age-related changes in tumour-infiltrating lymphocytes. Cytotoxic lymphocytes from aged mice are less able to bind targets, although they appear to be equally effective in destroying their targets . Assuming that FasL-positive tumour cells interact with tumour-infiltrating lymphocytes, our clinical data support indirectly the counterattack model. Immune surveillance among older patients is insufficient in breast carcinomas with a FasL:Fas ratio greater than 1, which leads to an increased rate of locoregional or distant metastases and to a shorter disease-free survival in postmenopausal patients. However, there is little evidence to support direct causal links between immune senescence and most malignancies, although data from in vitro and in vivo animal and human studies demonstrate clear age-related alterations in both cellular and humoral components of the immune system . Conclusion : The present findings confirm the prognostic significance of FasL:Fas ratio in breast carcinoma in a re-evaluated database over 54 months of follow up. Particularly in postmenopausal patients, the FasL:Fas ratio retained independent prognostic significance in multivariate analysis for disease-free survival. Additionally, the data strongly support a correlation between FasL:Fas ratio greater than 1 and lack of efficacy of tamoxifen in hormone receptor positive patients. However, given the important therapeutic repercussions of this issue, further studies are required before evaluation of FasL:Fas ratio can be routinely used to select patients who are likely to benefit from tamoxifen administration. Abbreviations : CI = confidence interval; ER = oestrogen receptor; FasL = Fas ligand; PCR = polymerase chain reaction; RT = reverse transcription. Backmatter: PMID- 12193273 TI - Delineation of RAID1, the RACK1 interaction domain located within the unique N-terminal region of the cAMP-specific phosphodiesterase, PDE4D5 AB - Abstract | Background | The cyclic AMP specific phosphodiesterase, PDE4D5 interacts with the beta-propeller protein RACK1 to form a signaling scaffold complex in cells. Two-hybrid analysis of truncation and mutant constructs of the unique N-terminal region of the cAMP-specific phosphodiesterase, PDE4D5 were used to define a domain conferring interaction with the signaling scaffold protein, RACK1. Results | Truncation and mutagenesis approaches showed that the RACK1-interacting domain on PDE4D5 comprised a cluster of residues provided by Asn-22/Pro-23/Trp-24/Asn-26 together with a series of hydrophobic amino acids, namely Leu-29, Val-30, Leu-33, Leu-37 and Leu-38 in a 'Leu-Xaa-Xaa-Xaa-Leu' repeat. This was done by 2-hybrid analyses and then confirmed in biochemical pull down analyses using GST-RACK1 and mutant PDE4D5 forms expressed in COS cells. Mutation of Arg-34, to alanine, in PDE4D5 attenuated its interaction with RACK1 both in 2-hybrid screens and in pull down analyses. A 38-mer peptide, whose sequence reflected residues 12 through 49 of PDE4D5, bound to RACK1 with similar affinity to native PDE4D5 itself (Ka circa 6 nM). Conclusions | The RACK1 Interaction Domain on PDE4D5, that we here call RAID1, is proposed to form an amphipathic helical structure that we suggest may interact with the C-terminal beta-propeller blades of RACK1 in a manner akin to the interaction of the helical G-gamma signal transducing protein with the beta-propeller protein, G-beta. Keywords: Rolipram, Protein kinase C binding protein, signalling scaffold, cyclic AMP, Rolipram, Protein kinase C binding protein, signalling scaffold, cyclic AMP, Background : cAMP is a ubiquitous second messenger that regulates numerous key physiological processes . Its levels are determined both by controls on its rate of synthesis through adenylyl cyclase activity and its rate of degradation through cAMP phosphodiesterase (PDE) activity . A large multigene family encodes many proteins that exhibit PDE activity . Of these, the PDE4 cAMP-specific phosphodiesterases have attracted considerable attention as inhibitors that are selective for them have behavioral, anti-inflammatory and smooth-muscle relaxant activity in humans [,,,-]. PDE4 enzymes can be differentiated from other cyclic nucleotides phosphodiesterase (PDE) families on the basis of sequence differences in their catalytic region and by their ability to be specifically inhibited by the drug, rolipram . They are also characterized by unique regulatory regions located in the amino-terminal half of the proteins, called Upstream Conserved Regions 1 and 2 (UCR1 and UCR2) . 'Long' PDE4 isoforms exhibit both UCR1 and UCR2, whilst 'short' isoforms lack UCR1 and 'super-short' isoforms lack UCR1 and have a truncated UCR2. In long PDE4 isoforms, UCR1 and UCR2 appear to interact with each other in order to form a regulatory module that mediates the functional outcome of phosphorylation by PKA and ERK . The mammalian PDE4s comprise a large family of isoforms, encoded by four different genes (PDE4A, PDE4B, PDE4C, and PDE4D), with additional diversity being generated by alternative mRNA splicing and the use of alternative promoters . A characteristic feature of these different isoforms is their unique N-terminal regions, which are believed to play a key role in the intracellular targeting of these isoforms . For example, the different N-terminal regions of PDE4A1 and PDE4A4/5 determine distinct patterns of intracellular targeting. Thus PDE4A1 is an exclusively membrane-associated species that is targeted to the Golgi, whilst PDE4A4/5 is localised to perinuclear and cell cortical regions in COS7 cells. PDE4D5 is one of five different isoforms encoded by the PDE4D gene and is found in a variety of tissues and cell types, including the brain . PDE4D5 can be distinguished from other PDE4D isoforms by the presence of a unique amino-terminal region of 88 amino acids, which is highly conserved among mammals . We have recently demonstrated that this PDE4D isoform can interact with the WD-repeat signalling "scaffold" protein, RACK1 . This was done using independent methods to demonstrate that RACK1 and PDE4D5 interact; namely two-hybrid screening, pull-down assays with recombinant RACK1, binding studies with purified recombinant proteins, and also by co-immunoprecipitation of native, endogenously expressed RACK1 and PDE4D5 . Native forms of these two proteins thus interact in a variety of different cell types . RACK1 is a 36 kDa WD-repeat protein that was first identified as a protein that could bind to certain protein kinase C (PKC) isoforms subsequent to their activation by either diacylglycerol or phorbol esters such as PMA . However, RACK1 has also been shown to interact with other proteins such as the beta-subunit of integrins , the common beta-chain of the IL-5/IL-3/GM-CSF receptor , the Epstein-Barr virus cytoplasmic A73 protein and SRC family tyrosyl protein kinases . Thus RACK1 appears to act as a scaffold or anchor protein as do other WD-repeat proteins such as Gbeta and beta '-COP . Previously, it has been shown that a small cluster of 4 amino acids in the unique N-terminal region of PDE4D5 is needed for its interaction with RACK1 . Here, using a program of deletion and mutagenesis strategies we show that a much more extensive surface within the unique N-terminal region of PDE4D5 is needed, involving a series of hydrophobic amino acids extending from Leu-29 to Leu-38. We call this the RACK1 Interaction Domain, RAID1 and suggest that it forms a helical region able to interact with RACK1. Results : Long PDE4 isoforms are characterised by the presence of two blocks of sequence, called UCR1 and UCR2, located between the conserved catalytic unit and the extreme N-terminal region that is unique to each isoform . The unique amino-terminal region of PDE4D5 consists of a block of 88 amino acids that has no homology with any of the other four PDE4D isoforms . We have demonstrated previously that this region of PDE4D5 is necessary and sufficient for it to bind to RACK1, thus explaining why PDE4D5 uniquely amongst the PDE4D isoforms is able to interact with RACK1. Within the PDE4D5 N-terminal region we discovered that the Asn-22/Pro-23/Trp-24/Asn-26 grouping of amino acids was crucial in allowing PDE4D5 to bind to RACK1. The discovery of this essential group of amino acids was achieved through a strategy of progressive N-terminal deletions and subsequent point mutations within the 88 residue N-terminal region of PDE4D5 . Figure 1 | The unique N-terminal region of PDE4D5 The unique N-terminal region of PDE4D5 This shows schematically the domain structure of the PDE4D5 long isoform together with the amino acid sequence (GenBankTM accession number AF012073) of its unique 88 residue N-terminal region. Indicated by a horizontal arrow is the sequence of both the 16-mer and 38-mer peptides that were assessed for binding to RACK1. In bold typeface are shown the residues of the essential Asn-22/Pro-23/Trp-24/Asn-26 grouping and also those that form the essential hydrophobic ridge, Leu-29, Val-30, Leu-33, Leu-37 and Leu-38. Also indicated are the positions of the various truncated species made with associated residue numbers. Mapping of the site for RACK1 interaction within the unique N-terminal region of PDE4D5 | Using an ELISA assay, we confirmed previous observations , that PDE4D5 interacted with RACK1 with a Ka of 4 -- 12 nM (range; n = 5 separate experiments). In contrast to this, using both ELISA and TNT competition assays, we failed to observe any binding to RACK1 of a 5-mer peptide (sequence = NPWLN) that contained the Asn-22/Pro-23/Trp-24/Asn-26 grouping , which we had previously identified as being pivotal for interaction of PDE4D5 with RACK1. Indeed, we were further surprised to discover that a 16-mer peptide (PEVDNPHCPNPWLNED), representing residues 13 to 28 of PDE4D5 and thus also containing the Asn-22/Pro-23/Trp-24/Asn-26 grouping , showed little if any interaction with RACK1 (Ka > 1 muM; n = 5). These data suggested to us that amino acids in addition to those within the Asn-22/Pro-23/Trp-24/Asn-26 grouping might be involved in allowing the efficient binding of PDE4D5 to RACK1. We thus set out to try and identify such determinants. The discovery of the Asn-22/Pro-23/Trp-24/Asn-26 grouping was based upon a strategy of progressive N-terminal deletions of the 88 residue N-terminal region of PDE4D5, followed by scanning mutagenesis of amino acids Pro-21 through Glu-27. Thus any additional RACK1-interacting residues must be C-terminal to the Asn-22/Pro-23/Trp-24/Asn-26 grouping. To evaluate this we instigated a strategy of progressive C-terminal truncation of the unique N-terminal region of PDE4D5 in a 2-hybrid screen. We have previously confirmed the fidelity of such an approach by using a variety of independent biochemical analyses, including co-immunoprecipitation, pull down assays with GST-RACK1 and the binding of purified components. The full N-terminal region of PDE4D5 interacted with RACK1, as did the 12 --88 N-terminal truncate in this 2-hybrid analysis . However, the C-terminally truncated construct, encompassing residues 12 through 28 of PDE4D5, failed to interact with RACK1 , despite the fact that such a region contained the Asn-22/Pro-23/Trp-24/Asn-26 grouping. Such a result indicates strongly that additional residues, located C-terminal to the Asn-22/Pro-23/Trp-24/Asn-26 grouping, are required for PDE4D5 to bind to RACK1. Further truncation analyses demonstrated that a C-terminal truncate, containing residues 12 through 49 of PDE4D5, was able to bind to RACK1 . This suggests that certain amino acids, located within positions 27 through 49 in the sequence of PDE4D5, are also required for RACK1 interaction. Figure 2 | Truncation studies within the unique N-terminus of PDE4D5 used to probe its interaction with RACK1 using a 2-hybrid screen Truncation studies within the unique N-terminus of PDE4D5 used to probe its interaction with RACK1 using a 2-hybrid screen. Plasmids encoding fusions between the DNA-binding domain of LexA and various amino terminal regions of PDE4D5 were tested for their ability to interact with RACK1, expressed as a fusion with the GAL4 activation domain (left column "pGADN-RACK1"). The identical LexA fusions were tested for their ability to interact with the GAL4 activation domain alone (right column "pGADN"). The regions of PDE4D5 included in the various constructs are annotated. The positions of the portions of PDE4D5 used in this study are shown schematically in Figure . The interactions between these components were evaluated with the filter beta-galactosidase assay described previously by us . The bottom two patches serve as internal positive (being the interaction between the oncoproteins RASv12 and RAF) and negative (vectors with inserts) controls, respectively. Previously we have shown that Glu-27 was not involved in RACK1 interaction. Here then we set out to determine which specific amino acids in the 28 through 49 region of PDE4D5 are necessary for the interaction with RACK1. Again, a 2-hybrid assay was used to test for the effect of individual mutations on the interaction. Each amino acid in this region was separately mutated to alanine except for the hydrophobic amino acids that were changed to aspartate and Ser-32 that was changed to both alanine and aspartate. This analysis highlighted a single residue, Leu-33, whose mutation to either aspartate or glutamate (not shown) ablated the interaction with RACK1. No other single mutations in the 28 through 49 region of PDE4D5 served to ablate interaction with RACK1 (data not shown), although we did observe a marked reduction in interaction with the Arg34Ala mutant . Figure 3 | Identification of residues involved in the interaction between PDE4D5 and RACK1 using a 2-hybrid screen Identification of residues involved in the interaction between PDE4D5 and RACK1 using a 2-hybrid screen. Shows a typical experiment of one done at least 3 times, where mutations of individual or pairs of residues in the indicated N-terminal region of full length PDE4D5 were expressed as LexA fusions of the non-mutated PDE4D5 (1 --746), pGADN-RACK1 and LexA alone ('vector'). The interaction between the oncoproteins RASv12 and RAF served as a positive control We noted, however, that the region C-terminal to the Asn-22/Pro-23/Trp-24/Asn-26 grouping contained a series of three leucine residues interposed by three amino acids in the repeat sequence Leu-29, Xaa, Xaa, Xaa, Leu-33, Xaa, Xaa, Xaa, Leu-37. This is reminiscent of hydrophobic repeat motifs that are found in amphipathic helical regions where coiled-coil forms of protein -- protein interactions occur . This repeat in PDE4D5 contains Leu-33, which we show here to be essential for PDE4D5 to interact with RACK1. As Leu-29 and Leu-37 both have hydrophobic residues immediately adjacent to them, namely Val-30 and Leu-38, we surmised that our inability to ablate interaction with RACK1 using single mutations might be due to contributions from these adjacent hydrophobic residues. We thus set out to evaluate whether mutation of these hydrophobic pairs would disrupt interaction with RACK1. Indeed the mutation of the Leu-29:Val-30 pair, to either aspartate or to glutamate (not shown), served to ablate interaction with RACK1. Similarly, mutation of the Leu-37:Leu38 pair, to either aspartate or to glutamate, also ablated interaction with RACK1 . This suggests to us that these repeated hydrophobic residues are important in the binding of PDE4D5 to RACK1. We did note, however, that Leu-25, within the Asn-22/Pro-23/Trp-24/Asn-26 grouping, is also located three residues N-terminal to Leu-29 and thus falls within a repeating Leu, Xaa, Xaa, Xaa, Leu unit. However, previously we have shown that, unlike other residues within the Asn-22/Pro-23/Trp-24/Asn-26 grouping, the mutation of Leu-25 to alanine did not ablate binding to RACK1. In this study we additionally mutated Leu-25, to either aspartate or to tryptophan, with no ablation of RACK1 interaction as assessed by 2-hybrid analysis (data not shown). Thus Leu-25 does not appear to be essential for interaction with RACK1 in the way that either the surrounding residues in the Asn-22/Pro-23/Trp-24/Asn-26 grouping are or as Leu-33 has been shown to be . We also explored various other pairs of amino acids through their simultaneous mutation to alanine; predominantly to evaluate residues that might be involved in charge-charge or hydrogen bonding interactions. These were the pairs Arg-34/Glu-35, Gln-39/His-40, Glu-41/Lys-42, Ser-43/Thr-45 and Arg-47/Lys-48. Of these, only the double mutant, involving Arg-34/Glu-35, attenuated the interaction. This reduction in signal was clearly less than the ablated signal seen, for example, with the Leu33Asp mutation . However, the apparent reduction in signal was comparable to that seen with the single Arg34Ala mutant (Fig. ; right panel). In contrast to this, the single Glu35Ala mutant showed a similar signal to the control, wild-type PDE4D5 construct . This suggests that Arg-34 and not Glu-35 plays a role in mediating the interaction between PDE4D5 and RACK1. We wished to provide independent confirmation that residues in addition to those forming the Asn-22/Pro-23/Trp-24/Asn-26 grouping were important in the interaction between PDE4D5 and RACK1. To do this, the various Leu33Asp, Leu37Asp:Leu38Asp, Leu37Glu:Leu38Glu and Arg34Ala mutants were created in a VSV epitope-tagged version of PDE4D5 and transiently expressed in COS7 cells. Extracts from these cells were then subjected to a pull-down assay with GST-RACK1 as done before by us . This analysis demonstrated that all of these mutations disrupted the interaction between PDE4D5 and RACK1. These data are consistent with the two-hybrid results (Figs , ) in highlighting an additional region of interaction within the unique N-terminal region of PDE4D5 that is essential for interaction with RACK1. Figure 4 | Pull down assays to probe the interaction between PDE4D5 and RACK1 Pull down assays to probe the interaction between PDE4D5 and RACK1. Pull down assays between RACK1 and PDE4D5 mutants were done as described in Methods. Shown are immunoblots for VSV-epitope tagged forms of full length PDE4D5 detected using an anti-VSV monoclonal antibody. Tracks 1 to 5 are immunoblots of cell lysates showing that equal amounts of immunoreactive material were taken for these pull-down studies. Tracks 6 to 10 are blots showing the 'pull-downs', namely material that is found associated with glutathione agarose-immobilized GST-RACK1. Full length PDE4D5 forms were used as either wild type (tracks 1, 6) or the mutant forms Leu33Asp (tracks 2, 7), Arg34Ala (tracks 3, 8), Leu37Asp:Leu38Asp (tracks 4, 9) and Leu37Glu:Leu38Glu (tracks 5, 10). These data are typical of experiments done at least three times. Whilst the 12 --38 truncate of PDE4D5 contained both the Asn-22/Pro-23/Trp-24/Asn-26 grouping and the essential hydrophobic ridge, it failed to bind to RACK1 in the 2-hybrid screen . It should be noted, however, that the final two residues of this truncated segment, namely Leu-37 and Leu-38, are together essential for allowing interaction with RACK1. It is thus highly likely that either steric influences or disruption of appropriate helix formation through replacement with 'foreign' residues, immediately C-terminal to Leu-37 and Leu-38 in the 2-hybrid construct, underpins this observation. A 38-mer peptide, comprising residues 12 --49 of PDE4D5, contains both the previously identified Asn-22/Pro-23/Trp-24/Asn-26 grouping and also the novel leucine repeat unit shown also to be crucial for RACK1 to bind to PDE4D5. This peptide bound to RACK1 with similar affinity1 (Ka of 6.4 +- 1.2 nM (Mean +- SD; n = 3 separate experiments) to that reported for PDE4D5 . This is consistent with the 2-hybrid screen done in this study , which indicated that the construct expressing amino acids 12 to 49 of PDE4D5 interacted with RACK1 in a similar fashion to full length PDE4D5. Thus the core PDE4D enzyme does not have any profound effect on the functional capability of the N-terminal region of PDE4D5 to interact with RACK1. Molecular modeling | Molecular modeling studies , as well as the protein structure prediction programs, PHD and SSpro (data not shown) , suggest that the RACK1 Interacting Domain 1, RAID1 that we have identified here in PDE4D5 is likely to form a helical structure. In this the essential hydrophobic residues (Leu-29, Val-30, Leu-33, Leu-37, Leu-38) is suggested to provide a distinct hydrophobic ridge along one face of an amphipathic helix . This hydrophobic ridge appears to comprise 3 key sticky 'patches', with 'patch 1' formed by Leu-29 and Val-30, 'patch 2' by Leu-33 and 'patch 3' by Leu-37 and Leu38. Destruction of any one of these patches in its entirety suffices to ablate interaction with RACK1. However, single mutations, to the charged amino acid aspartate, within either patch 1 or patch 3 are not sufficient to ablate interaction of PDE4D5 with RACK1. Figure 5 | Modeling RAID-RACK1 interactions Modeling RAID-RACK1 interactions. (a) Surface representation of RAID1 core segment, Asn-22 to Leu-38, as a helix, with sidechains curtailed at the Cbeta atom for clarity. Residues implicated in RACK1 binding are colour-coded on a yellow helical background; blueAsn-22, Pro-23, Trp-24, Asn-26, Arg-34, and purple Leu-29, Val-30, Leu-33, Leu-37, Leu-38. These residues tend to align along one side of the helix. (b) Gbeta (blue) -- Ggamma (yellow) interactions from protein data bank coordinate set '1got,' with a molecular surface (white) drawn for the WD repeats 5 --7 of Gbeta (lacking the initial beta-strand of WD repeat 5 that is associated with the previous propeller blade). (c) Electrostatic potential surface of comparative modeled RACK1 WD repeats 5 --7 (equivalent part and in the same orientation as that drawn for Gbeta in panel b). Red denotes negative, blue positive, and white is non-polar. (d) RACK1 mutations that affect RAID binding are drawn in green on a molecular surface for WD repeats 5 --7. (e) Model for RACK1-RAID interactions, with the core RAID segment of panel a forming the right-hand helix, that connects through a loop to a positive region of RAID that is modeled (left-hand helix) into a negative part of the RACK1 surface. Colour coding for the RACK1 surface follows that in panel 2, and that for the core RAID segment follows panel a. Arg-34 is at the back of the helical ribbon and not visible in this view. The orientation matches panels b,c, and d, so that general features can be compared with Gbeta-Ggamma. At the N-terminal 'head' of this proposed hydrophobic ridge is found the Asn-22/Pro-23/Trp-24/Asn-26 grouping that is also essential for interaction of PDE4D5 with RACK1. Leu-25 appears at first glance to form a simple continuation of this hydrophobic ridge into the Asn-22/Pro-23/Trp-24/Asn-26 grouping. However, its mutagenesis to a range of charged and aromatic amino acids did not affect the ability of PDE4D5 to interact with RACK1, suggesting that Leu-25 has little or no role in determining the interaction between these two proteins. Thus the Asn-22/Pro-23/Trp-24/Asn-26 grouping and the proposed hydrophobic ridge that encompasses Leu-29, Val-30, Leu-33, Leu-37 and Leu-38 appear to form two discrete units within RAID1 that are each essential for PDE4D5 to bind to RACK1. Discussion : We have defined here a sub-domain, which is located within the unique 88 amino acid N-terminal region of PDE4D5, that allows PDE4D5 to bind to the signaling scaffold protein, RACK1. This region was defined using independent methodologies of 2-hybrid assays and biochemical pull-down assays. We propose calling this region RAID1, for RACK1 Interacting Domain 1. RAID1 is located in the N-terminal half of this 88 amino acid isoform-specific region of PDE4D5, extending from Asn22 to Val30. It consists of a small cluster of amino acids, namely the Asn-22/Pro-23/Trp-24/Asn-26 grouping , together with a series of hydrophobic amino acids, namely Leu-29, Val-30, Leu-33, Leu-37 and Leu-38 in a 'Leu-Xaa-Xaa-Xaa-Leu' repeat. The WD-repeat protein Gbeta, illustrated in Fig. , interacts with the Ggamma protein through coiled-coil interactions between the N-terminal regions of each subunit, and through helical and non-helical segments of the Ggamma C-terminus, interacting largely with propeller blades 5 --7 at the C-terminus of Gbeta. Hydrophobic surfaces play a large role in these interactions. Using a truncation strategy coupled with a reverse 2-hybrid screen, we have demonstrated that the C-terminal region of RACK1 plays a crucial role in conferring the binding of PDE4D5. Indeed, various single amino acid mutations in this region of RACK1 serve to ablate the binding of PDE4D5 . Using colour coding of modeled RACK1 repeats 5 --7 we identify here regions of particular polarity as well as those that have been implicated through mutagenesis in binding PDE4D5 . Whilst RACK1 lacks an equivalent to the N-terminus of Gbeta, which mediates the helical coiled-coil interactions with Ggamma, these mutagenesis data and surface properties indicate that PDE4D5-RACK1 interactions may occur in a manner generally akin to that for the C-terminal regions of Ggamma and Gbeta. We have attempted to explore this possibility in the model shown in Fig. . The starting point for the model is placement of RAID1, modeled as a helix , adjacent to the mutagenesis-implicated groove on RACK1 . Of the other major surface regions highlighted in Fig. , that to the far right, by comparison with Fig. , forms internal interactions and the upper central green feature is due to a single extended residue, R245, which would appear to be sufficiently flexible as to allow, in principle, an interaction with E35 in the PDE4D5 N-terminal region. Having placed the core segment of RAID1 in the model, we note that a second helical segment, containing a positively charged region of the PDE4D5 N-terminus, could be positioned to interact with a negatively charged region of RACK1 . Thus Fig. shows both segments of the model, with a short connecting loop that mediates a change in helical direction. We note that in overall terms the model has similarities with Gbeta-Ggamma binding , using helical segments binding into grooves and intervening turns. The reverse direction of the RAID1 polypeptide chain relative to Ggamma, on the propeller surface, results from docking the positively charged RAID1 segment into a negatively charged part of the RACK1 model. Since the role of this part of RAID1 in binding is not clear, we regard this element of the modeling as secondary to the primary location of the amphipathic RAID1 helix in the RACK1 groove having the identified multiple binding mutants. This primary docking could in fact be accomplished with either direction of polypeptide chain, so that we assign less weight to our suggestion of RAID1 running opposite to Ggamma than to the proposed binding groove for the core RAID segment. Interestingly, Sondek and Siderovski , based on the established interaction between Gbeta and Ggamma, used 3D-modeling analyses to propose that protein-protein interaction involving the C-terminal region of beta-propeller proteins may be directed by putative G-protein Gamma-Like (GGL) motifs on the binding partner. Indeed, they went on to suggest that the Asn-22/Pro-23/Trp-24/Asn-26 grouping, in the unique N-terminus of PDE4D5, might resemble such a GGL domain and therefore bind to the C-terminal region of RACK1 in a similar region to that where Gbeta and Ggamma have been shown to interact. Thus, by analogy with the GGL model , Asn-22/Pro-23/Trp-24/Asn-26 may represent a common core motif as seen with homologous proteins. If this is the case then the specificity of interaction must come from additional structural motifs that either enhance or reduce interaction with particular beta-propeller proteins. On this basis, we suggest that the leucine-rich region, which we have identified in the present study, may serve primarily to direct specificity for interaction of RAID1 with RACK1 rather than other beta-propeller proteins. The existence of this type of "structural conditioning" would mean that a particular family of beta-propeller proteins could have different specificity with regards to their protein-binding partners. Additionally, for each beta-propeller protein there may be a family of proteins that can even interact at one 'site'. From the 3D-models presented in this study we can see that WD 5 --7 of RACK1 contains a range of different interaction sites and therefore may accommodate a range of possible interactors; including those that bind weakly by only interacting with a small part of the surface, and those that interact strongly, like RAID1 , which interact with multiple determinants over an extended surface. We also noted, however, that mutation, to alanine, of the positively charged residue, Arg-34 ablated PDE4D5 interaction with RACK1 in pull-down studies and severely attenuated interaction in 2-hybrid screens. This indicates that Arg-34 is also important in defining the binding of PDE4D5 to RACK1. Arg-34 is proposed as being located within the helix that contains the hydrophobic ridge and is adjacent to the negatively charged Glu-35. In our model it is possible that Arg-34 could either interact directly with RACK1, perhaps through Asp-294, adjacent in the model, or perhaps by virtue of non-polar interactions mediated by the aliphatic part of an arginine sidechain. Whilst, potentially, it could also stabilise a bound helical segment through interaction with Glu-35 we consider this to be unlikely as the Glu35Ala mutation interacts with RACK1 akin to wild-type PDE4D5. Conclusions : Thus, an amphipathic helical conformation is suited to the part of PDE4D5 that we show here is essential for its binding to RACK1, namely Leu-29, Val-30, Xaa, Xaa, Leu-33, Arg-34, Xaa, Xaa, Leu-37, Leu-38. This helix may provide the driving force for an initial interaction between PDE4D5 and RACK1 that is then further stabilized through the correct presentation of the essential Asn-22/Pro-23/Trp-24/Asn-26 grouping. The structural model that we propose may describe such an interaction suggests that binding of PDE4D5 to RACK1 may involve interactions that are akin to those seen for the binding of the Ggamma protein to Gbeta in terms of helix/groove nature and overall location. This interaction may thus allow PDE4D5 to be recruited into a signaling scaffold complex where RACK1 acts as an adaptor. Our definition of the interaction domain, RAID1 should aid in the design of reagents aimed at disrupting the interaction between PDE4D5 and RACK1 so as to use them to gain insight into the physiological function of such interaction in intact cells. Certainly, the targeting of PDE species is likely to contribute to the generation of compartmentalised cAMP responses . Indeed, there is currently much interest in identifying anchor proteins for protein kinase A (AKAPs) that, seemingly, serve to sample gradient of cAMP within cells, leading to the selective activation of specific PKA isoforms that are associated with particular signalling functions . Methods : Materials | An antibody to RACK1 was obtained from Transduction Laboratories. A monoclonal antibody to the VSV epitope tag was from Sigma. The polyclonal antiserum specific for PDE4D was as described before by us . Peptides were from Cambridge Biosciences (Cambridge, UK) and were determined to be >98% pure with composition verified by HPLC and mass spectrometry. Two-hybrid constructs and assays | Various segments of the PDE4D5 cDNA (GenBankTM accession number AF012073) were cloned into the Not I site of the vector pLEXAN, to generate fusions between the various amino-terminal portions of PDE4D5 and the DNA binding domain of the E. coli LexA protein . pLEXAN is a derivative of pBTM116, with a Not I site inserted into the polylinker. These constructs were prepared by the addition of Not I sites to the cDNA regions by the use of PCR, as described previously by us . Similarly, the full open reading frame (ORF) of human RACK1 cDNA (ref. ; GenBank accession number M24194) was cloned into the Not I site of pGADN, to produce pGADNRACK1. pGADN is a derivative of pGADGH, but with a Not I site inserted into the polylinker. This vector expresses proteins as in-frame fusions with the trans-activation domain of the S. cerevisiae GAL4 protein. Two hybrid assays were performed in the S. cerevisiae strain L40 using methods that we have described in detail previously . Generation of bacterial expression constructs | Various portions of the PDE4D5 cDNA were cloned into the Not I site of pMALN, using PCR as described above for the LexA fusions. pMALN is a derivative of pMALC2 (New England Biolabs), with a Not I site inserted into the polylinker. All PCR-generated or mutant constructs were verified by sequencing prior to use. Generation of cDNAs encoding mutant forms of PDE4D5 | To generate mutations of one or two individual amino acids in PDE4D5, the full-length PDE4D5 cDNA was subjected to site-directed mutagenesis with the QuikChange site-directed mutagenesis kit (Stratagene). All PCR-generated or mutant constructs were verified by sequencing prior to use. Expression of glutathione S-Transferase (GST) and maltose binding protein (MBP) fusion proteins in E. coli | The generation, expression and purification of fusion proteins between the full ORF of RACK1 and both glutathione-S-transferase (GST) and maltose-binding protein (MBP) have been described in detail previously by us . RACK1 pull-down assays | The construction of pcDNAPDE4D5VSV has been described previously by us . It contains the full ORF of PDE4D5, with a vesicular stomatitis virus (VSV) glycoprotein epitope at the carboxyl terminus , cloned into pcDNA3 (Invitrogen). Transfection of COS7 cells and pull-down assays were performed as described previously by us . Briefly, COS7 cells were transfected with 20 mug of control plasmid pcDNA3 or with 20 mug of plasmid pcDNAPDE4D5VSV encoding wild type VSV-tagged PDE4D5, or with 20 mug of plasmids pcDNAPDE4D5L33DVSV (Leu33Asp), pcDNAPDE4D5R34AVSV (Arg34Ala), pcDNAPDE4D5L3738DVSV (Leu37Asp:Leu38Asp), pcDNAPDE4D5L3738EVSV (Leu37Glu:Leu38Glu) encoding various mutant forms of VSV-tagged PDE4D5 (as indicated in parentheses). 72 h after transfection, cells were harvested in lysis buffer (55 mM Tris-HCl, pH 7.4, 132 mM NaCl, 22 mM sodium fluoride, 11 mM sodium pyrophosphate, 1.1 mM EDTA, 5.5 mM EGTA, 0.1% (v/v) Triton X-100 plus protease inhibitors) and incubated for a further 1 h with rotation at 4C. Cell debris was removed by centrifugation at 13,000 g for 10 min at 4C. 400 mug of cleared lysate was incubated with 50 mug of GST or GST-RACK1 and 60 mul of glutathione-Sepharose beads for 2 h at 4C with rotation. Beads were pelleted by centrifugation at 2000 g for 5 min at 4C, washed three times in lysis buffer and then boiled in 75 mul Laemmli buffer. 50 mul samples from beads and 50 mug each of cleared lysate were analysed by SDS-PAGE followed by immunoblotting with an anti-VSV monoclonal antibody . SDS polyacrylamide gel electrophoresis and immunoblotting | These were as described previously . In brief, samples were resuspended in Laemmli buffer and boiled for 5 min. Membranes were blocked in 5 % (w/v) low-fat milk powder in TBS (10 mM Tris-HCl, pH 7.4, 150 mM NaCl) overnight at room temperature. They were then incubated with anti-VSV monoclonal antibody diluted in 1:5000 (v/v) in 1 % (v/v) low-fat milk powder in TTBS (TBS plus 0.1 % (v/v) Tween 20) for 3 h at room temperature. Detection of the bound antibody was with anti-mouse IgG peroxidase (Sigma) and the enhanced chemiluminescence (ECL) system (Amersham). Measurement of protein concentrations | Protein concentrations were measured by the method of Bradford , using BSA as a standard. ELISA binding assay | A modification of the method described previously by us was used. High binding ELISA plates were treated with 100 mul of MBP-PDE4D5, MBP alone or various peptides at a concentration of 1 muM in 50 mM carbonate-bicarbonate buffer (Sigma C-3041) for 16 h at 4C. The plates were then blocked in 5% (w/v) low-fat milk powder in above buffer for 1 h at room temperature. Plates were washed three times with 200 mul/well wash buffer (TBS + 0.1% (v/v) Tween-20) before being incubated for 2 h, at room temperature, with 100 mul GST-RACK1 solution in TBS. The GST-RACK1 solution was formed from a range of dilutions over the concentration range 50 pM -- 2 muM. Plates were then washed as above with protein complex formation being detected by the addition of anti-RACK1 monoclonal antibody followed by peroxidase conjugated anti-mouse IgM. These antibodies were both used at 1:2500 in TBS for 1 h at room temperature with a wash step in between. Immunoreactivity was visualized by adding ABTS solution (0.4 mM 2,2'-azino-bis [3-ethylbenz-thiazoline-6-sulfonic acid] in 50 mM sodium citrate pH 4.0 plus 36 mul 30% (vol/vol) H2O2 just prior to use) and quantified using an MRX plate reader set at a wavelength of 405 nm. Molecular modeling | A comparative model of RACK1 was constructed with the program Modeller , using bovine transducin Gbeta as the structural template. The Gbeta subunit was extracted from a G protein heterotrimer structure with coordinate file identifier 1 got in the protein structural data bank. The underlying sequence alignment between RACK1 and the Gbeta template followed that given previously , except for WD repeats 6 and 7 where the sequence match is less clear and structural factors were used to make small adjustments. In particular, putative beta-strands in RACK1 were identified through alignment of stretches of non-polar residues with the known beta-strands of the Gbeta subunit, which are also predominantly non-polar. This adjustment aligned W339, N340 of Gbeta with W310, Q311 of RACK1. Analysis of the RACK1 comparative model in terms of binding the N-terminal region of PDE4D5 focussed on WD repeats 5 --7. Molecular surfaces were drawn for WD repeats 5 --7 that lack the start of WD repeat 5, since this region would form the outer beta-strand of an otherwise absent propeller 4, and is therefore unlikely to exhibit structural order in the experimental WD repeat 5 --7 construct. The N-terminal region of PDE4D5 implicated in binding to RACK1 by mutational analysis was modelled as mainly alpha-helical polypeptide, with the potential for non-helical linker regions. Binding surfaces on RACK1 and PDE4D5 were assigned from the mutation affects on binding, and their combination in a general model for complexation used these data as well as overall complementarity of surface shape and polarity. Computer programs QUANTA (Accelrys) and Swiss-PDB Viewer were used for manipulation and display. Authors' contributions : MDH and GBB conceived, directed and managed the study. GBB and MRS performed the 2-hybrid analyses. AM performed the biochemical pull-down studies and had input into the direction of the study. SJW performed the ELISA studies and had input into the direction of the study. JW performed the molecular modeling analyses. Backmatter: PMID- 12184817 TI - Low agreement for assessing the risk of postoperative deep venous thrombosis when deciding prophylaxis strategies: a study using clinical vignettes AB - Abstract | Background | Several clinical practice guidelines (CPG) on antithrombotic prophylaxis in surgical patients help to decide about the prophylaxis strategy based on the patient risk of deep venous thrombosis (DVT). However, the physician risk estimates of DVT could have little inter-observer reproducibility, which could lead to different individual prophylaxis practices. Methods | Physicians were asked to evaluate DVT risk in eight clinical vignettes, describing actual patients cared for in our hospital. The vignettes included all possible levels of DVT risk. Results | The degree of prophylaxis strategies accuracy was 63% (95% CI 523 --75%). Overall agreement was 0.32 (z = 7.61, p < 0.001) and for each level of risk kappa was 0.38 (z = 6.50, p < 0.001); 0.1 (z = 1.65, p < 0.049) and 0.5 (z = 8.45, p < 0.001) for small, moderate and high risk group respectively Conclusions | Our results showed that there is poor agreement when physicians have to evaluate the risk for postoperative DVT, and in the cases of low and moderate risks of DVT there is the smallest agreement. In addition, the data also showed that the overall accuracy of DVT prophylaxis strategy was only moderate and the risk evaluation did not correlate to the selection of the strategy. The issue of inter-observers variability should be taken into account when CPG performance are analysed, especially when considering the risk-evaluation to choose the appropriate actions. Keywords: Background : Deep venous thrombosis (DVT) prophylaxis is an effective intervention in the postoperative period . Several clinical practice guidelines (CPG) on the topic have been published, with the aim to increase the utilisation and appropriateness of this practice. Usually, these guidelines established the choice of the prophylaxis strategies based on the type of surgery, as well as on the presence of specific patient's risk factors To establish an individual patient risk level is a complex task that could lead to substantial variations between observers. The integration of several cognitive inputs and patient data under real work pressures is a task where the physicians' performance is sub-optimal . In order to determine if there are variations in the postoperative DVT risk perception among physicians, we performed a study using clinical vignettes. Methods : Study design | Clinical practice guideline | Our guideline on DVT prophylaxis is mostly based in the 5th ACCP consensus conference recommendation , where risk-assessment approaches are used to decide the prophylaxis strategy. In our guideline, the postoperative DVT risk is evaluated basing on the items that follow: the age, the presence of venous thrombosis risk factors and the duration of surgery. With this information we could classify the patients in low, moderate and high-risk groups. Early mobilisation or graded compression stockings are prescribed in the case of low risk group patients. Moderate risk patients are prescribed fixed subcutaneous calcium heparin and graded compression stockings could replace heparin if high haemorrhagic risk exists. High-risk patients received nadroparin but when it was deemed inappropriate due to a high haemorrhagic risk, intermittent pneumatic calf compression (IPCC) was indicated. Both nadroparin and IPCC may be used in selected high-risk patients. The guideline has been widely publicized throughout the institution and it is available in the Hospital Intranet. Clinical Vignettes | We have written eight comprehensive clinical vignettes including patients that cover the entire risk spectrum in order to evaluate the inter-observer variability about the risk assessment,. The information was mostly taken from actual patients cared for at our hospital. Specific details were changed in order to respect patient confidentiality and to avoid clueing physicians about the appropriate decisions. The vignettes are summarised in the table. No consent is required in our hospital to use clinical record data if no breach to patient confidentiality is possible. One of us (DG) independently reviewed the clinical vignettes for accuracy and validation regarding the study objectives. Table 1 | Clinical Vignettes summary with risk level and proposed prophylaxis strategyin the CPG. We randomly chose 10 physicians from the Internal Medicine and the Intensive Care Unit to read the vignettes. Then we asked them: 1) to estimate the inhospital complications risks (nosocomial infection, cardio-respiratory events, DVT risk and hospital mortality) and to classify them in three levels: low, moderate and high risk; and 2) to order a complete diagnostic and therapeutic plan for each vignette. Statistical analysis | Percentages, along with 95% confidence intervals, were calculated for the accuracy of the strategy selection compared to the intended level of risk for the vignette, as decided by the investigators. The inter-observer agreement was evaluated using the kappa statistic for multiple raters and for multiple categories. Statistical calculations were performed with STATA 6.0. Results : The degree of appropriateness of prophylaxis strategies used were 63% (95% CI 52 --75%). There was no correlation found between the accuracy of the prophylaxis strategies and the DVT risk evaluated by the physician (chi2 for trend 0.54, p = 0.46). Although no statistically significant, the low risk strata showed a higher rate of strategy inappropriate selection compared to the moderate and high risk ones (72%, 61% and 61.5 respectively, x2 = 0.7, p = 0.7). Overall agreement was 0.32 (z = 7.61, p < 0.001) and kappa by risk category was 0.38(z = 6.50, p < 0.001); 0.1(z = 1.65, p < 0.049) and 0.5 (z = 8.45, p < 0.001) for the low, moderate and high risk group respectively. All the DVT risk wrong estimates in the low risk group were rated as "moderate" and in the moderate risk group as "low risk". In the high-risk group, 70% of the wrong estimates were "low risk" and the rest were classified as "moderate risk". Discussion : Our results showed a poor agreement for assessing the risk of postoperative DVT, especially in the low and moderate risk groups. Wrongly classified patients who belong to moderate risk groups were classified in lower risk category. In addition, the data also showed that overall DVT prophylaxis strategy was not good enough and the risk evaluation did not correlate to the strategy selection. As stated by Weed et al , our example on DVT risk assessment showed that integration of cognitive inputs and patient data must be systematically explored for different clinical settings since considerable variation may occur, especially when the physician behaviour is assessed. The evaluation of the physician performance is a difficult task, as several factors from physicians themselves, healthcare structure and case complexity could modify the evaluation. Clinical vignettes are a validated measurement of the physician performance , as compared both to utilisation review and simulated patient-doctor encounters. This method permits controlling case-mix, an important confusing variable when the physician performance is evaluated. A standardised way of risk evaluation could improve the quality of a key point in the decision process of the CPG, such as risk assessment. Computerized decision support systems have shown benefits in increasing usage and accuracy of DVT prophylaxis . Conclusions : The issue of inter-observers variability should be taken into account to analyse variations in CPG performance, especially when considering risk-evaluation to select the appropriate actions. Competing interests : None declared Authors' contributions : MO y DG designed the study and created the clinical vignettes. MO analyzed the data and wrote the manuscript. KL and PM interviewed the physicians. All authors read and agreed with the final manuscript. Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12153701 TI - Accuracy of responses from postal surveys about continuing medical education and information behavior: experiences from a survey among German diabetologists AB - Abstract | Background | Postal surveys are a popular instrument for studies about continuing medical education habits. But little is known about the accuracy of responses in such surveys. The objective of this study was to quantify the magnitude of inaccurate responses in a postal survey among physicians. Methods | A sub-analysis of a questionnaire about continuing medical education habits and information management was performed. The five variables used for the quantitative analysis are based on a question about the knowledge of a fictitious technical term and on inconsistencies in contingency tables of answers to logically connected questions. Results | Response rate was 52%. Non-response bias is possible but seems not very likely since an association between demographic variables and inconsistent responses could not be found. About 10% of responses were inaccurate according to the definition. Conclusion | It was shown that a sub-analysis of a questionnaire makes a quantification of inaccurate responses in postal surveys possible. This sub-analysis revealed that a notable portion of responses in a postal survey about continuing medical education habits and information management was inaccurate. Keywords: Background : Postal questionnaire surveys of physicians are a popular instrument to gather information . They are often used for studies about continuing-medical education (CME) habits and information management since they are relatively inexpensive and easy to handle . A major problem of such surveys is the low response rate . Besides this non-response bias there are a number of other factors which restrict conclusions from postal surveys. Alreck and Settle for example describe ten such sources of response bias . But it seems that there are even more biases. Most important are: the tendency to socially desired responses (especially in surveys on sensible subjects like drug abuse or sexual habits) , acquiescence or the tendency for only yes- or no-responses , failure in self-perception or (technically) inaccurate statements (e.g. because of willful lies or inaccurate memories) . Most of the studies about the problem of potential biases are restricted to questionnaires for patients. Therefore a MEDLINE search revealed only articles dealing with the accuracy of statements by physicians in general, but there were no satisfactory results for articles about potential inaccuracies in postal surveys about CME habits or information management (search terms -- MeSH: "Physicians", "Reproducibility of Results", "Questionnaire", "Bias", "Quality Control". Title words: "Questionnaire", "Postal Survey*", "Validity", "Bias", "Inaccura*", "Accura*") . This is also represented by the fact that authors of postal surveys in this field often do not discuss the problem of inaccuracies [e.g. ] or if it is discussed it is not quantified [e.g. ]. For this reason a sub-analysis of a questionnaire survey about CME and information habits of German diabetologists was performed. It should primarily determine how accurate the information in this study was and if the responses were credible. Furthermore I tried to evaluate if these inaccuracies could be attributed to the socially desired response bias. The following report focuses on the sub-analysis and not on other results from the survey which are (partly) reported elsewhere . Methods : Survey | Questionnaire | The data used for this sub-analysis was collected by an explorative survey about information management and CME habits (for details see ). For this survey a new questionnaire had to be developed. Initially a preliminary questionnaire was developed considering three already published surveys . It was discussed with members of the research group and sent to experts requesting comments (practicing diabetologists, experts in evidence-based medicine, technology assessment, survey methodology, and continuing medical education). After incorporating these comments the questionnaire consisted of 92 items. It can be divided into the following sections : CME in general, therapeutic decision making and behavior of problem solving, use of databases, reading habits, knowledge of technical terms and critical appraisal, personal data. 3 of 92 items were asked open. Sample and send out | The sample comprised of 461 diabetologists in the northern part of Germany. It was selected from a database of German diabetologists (Diabetologe DDG) (URL: ). The sample represented 29% of all 1585 diabetologists in the database. Sample size was calculated with regard to confidence intervals for estimated population frequencies (95% CI): a maximal margin of error for proportions of +- 6.25% for questions answerable dichotomously was considered narrow enough (i.e. the maximum width of the 95% CI for proportions should be 12.5% for questions with only two response categories e.g. yes/no). Given the population of 1585 diabetologists this required a sample size of 213. Response rates of prior surveys ranged from 50% to 70% which results in a sample size of at least 416 persons. For technical reasons it was not possible to draw a random sample. Therefore the sample was determined by the first figure of the zip code (code 1 --3). In October 2000 the questionnaire had been distributed for the first time. One week later a reminder postcard was sent to all participants and after three weeks a new questionnaire was sent to all non-respondents. A cover letter as well as a metered self-addressed envelope were enclosed. Coding by numbers for response control was explicitly mentioned but the analysis was fully anonymous. Sub-analysis | Variables used | The analysis is mainly based on contingency tables of answers to logically connected questions. Non-consistent responses were denoted "positive". The following variables were used (see original questions in the additional file 1. Did respondents, who stated that systematic reviews/meta-analyses had a strong influence on their therapeutic decision making, report that they knew these two terms? 2. Did respondents, who stated that published clinical trials and systematic reviews/meta-analyses had a strong influence on their therapeutic decision making, report that they read these kind of articles? 3. A question about the knowledge of technical terms was asked (as suggested by McColl and colleagues ). A contingency table was created with answers to the term absolute risk reduction (ARR) and number needed to treat (NNT). Respondents who stated that they could explain the number needed to treat but could not explain absolute risk reduction were labeled positive (the number needed to treat is the reciprocal of the absolute risk reduction). 4. There was a question on the knowledge about a fictitious technical term (the McNemar-Quality-Scale; explanation of terms was not required). Respondents who stated that they knew this scale were labeled positive. 5. Did respondents, who stated that they appraised the scientific value of an article by evaluating its methods section (as suggested by Williamson and colleagues ) report that they read this section of an article? Test for socially desired response bias | The assumption was that the tendency to socially desired responses would be the most dominant response bias in this survey. It was also presumed that this would be most prevalent in the question about technical terms. Two tests were used to support these assumptions: 1. A knowledge-score was calculated for each respondent using responses to the question about technical terms (All items/technical terms were included except the McNemar-Quality-Scale. Every cross at category: I understand this term and could explain it to others was valued with one point. Every cross at category: I have some understanding was valued with a half point. The sum was rounded. Therefore maximum score was twelve points). This knowledge-score was cross tabled with the positive answers of variable 5 (knowledge of the McNemar-Quality-Scale). 2. A contingency table with answers to the fictitious McNemar-Quality-Scale and the most unknown technical term was created. This term was the Alpha-error/Type-I-error. Only 50% (117/233) of all respondents knew this term. Statistical Analysis | Descriptive statistics were mainly used. The chi2-test was used for comparison of categorical data (Yates continuity corrected for comparisons with 1 degree of freedom). Fisher's exact test was used if the expected cell values were less than 5. The mediantest was used for a comparison of the knowledge-scores because distributions were neither normal nor comparable . Two-sided p-values < 0.05 were attributed as significant. Analyses were performed with the use of EpiInfo 2000, version 1.0.4 and KyPlot, version 2.0. Results : Of the 461 questionnaires distributed, 45 (10%) were returned because they were undeliverable. In this group the proportion of hospital-based physicians was significantly higher (33/45: 73% vs. 199/416: 48%; chi2 = 9.564; p = 0.002) whereas the proportion of practicing physicians was significantly lower (9/45: 20% vs. 187/416: 45%; chi2 = 9.349; p = 0.002) than in the remaining sample. 239 (52%) questionnaires were eligible for analysis (if undeliverable questionnaires are disregarded response rate is 57%). Table compares the characteristics of the respondents, all German diabetologists (Diabetologen DDG), and the whole sample. Table compares the respondents and non-respondents. Table 1 | Characteristics of respondents, sample, and all german diabetologists Table 2 | Characteristics of respondents and non-respondents Understandability of the questionnaire was good. 56/235 (24%) of respondents stated questions were easy to understand, 160/239 (68%) found them rather easy to understand, and 19/235 (8%) found them rather difficult. Nobody found questions difficult to understand. 1. Knowledge of influential factors | 15% (35/232) and 23% (53/230) respondents who stated that meta-analyses and systematic reviews respectively have a strong or very strong influence on their therapeutic decision making, had no or only a rough understanding of the meaning of these types of articles. 2. Reading and influence of different article types | The rates of respondents, who stated that the different article types have strong or very strong influence on their therapeutic decision making, but do not read these articles were very low. Rates were 3% (7/235) for clinical trials, 0% (1/235) for systematic reviews/meta-analyses, and for narrative reviews there was no discrepancy. 3./4. Knowledge of the NNT and a fictitious term | 16% (38/234) could explain the number needed to treat but could not explain absolute risk reduction . Overall 7% (17/234) of the respondents allegedly had at least some understanding of the McNemar-Quality-Scale of which one stated that he/she could explain this scale to others (24 respondents (10%) reported that they knew the scale but did not understand it). Table 3 | Knowledge of the number needed to treat and absolute risk reduction* 5. Examining and reading the methods section of articles | Table shows whether respondents who reported that they evaluated article-quality by examining the methods section actually read this part of an article. 13% (22/172) of responses were contradictory if they were interpreted strictly. Categorized in two groups (always/often and seldom/never) 8% (14/172) of contradictory responses remained. Table 4 | Examining and reading the methods section of articles* Test for socially desired response bias | The median knowledge-score in the group positive for variable 4 (knowledge of the McNemar-Quality-Scale) was 10 (IQR: 8 --11.5; range: 6 --12; mode 10 and 12). In comparison the median knowledge-score of the other respondents was 6 (IQR: 4 --8; range: 0 --12; mode: 6). This difference was significant (Fisher's exact test: p = 0.001). A comparison of responses to the terms McNemar-Quality-Scale and Alpha-Error/Type-I-Error revealed that only 2 (12%) of the 17 respondents allegedly knew both terms. The difference to the responses negative to the McNemar-Quality-Scale (2/17 vs. 115/216) was significant (chi2 = 9.249; p = 0.002). Discussion : Methodological issues | Because selection of the sample was not randomized systematic biases are possible. Demographic characteristics of all German diabetologists are subject to limited availability. Therefore an assessment of the representativity of the sample is restricted. The different proportion of general practitioners and pediatricians can be considered as bias. But whether this is of relevance for this analysis remains questionable (an association of positive responses and specialty could not be found; checked for variables 3, 4, and, 5; data not shown). The response-rate lies under the average of other surveys . But no major differences in the four available demographic characteristics could be detected between the respondents and the sample . The relatively higher rate of undeliverable questionnaires among hospital-based physicians is certainly negligible since the number of persons is too small. Non-response bias may be another problem but its relevance seems as well questionable because an association between proportions of positive answers and sex, work place, or location of work place could not be found (checked for variables 3, 4, and 5; data not shown) (see also ). Nevertheless caution should be applied when generalizing the results of this survey and rates or numbers should be interpreted as a trend rather than at face value. Another limitation lies in the methodology of this analysis. Since actual procedures of physicians were not observed (e.g. how they read journal articles) it is only possible to determine inaccuracies indirectly. Though it would be preferable to conduct such a study it is not feasible for practical reasons. Furthermore this analysis allows no extensive conclusions about the nature of the inaccuracies . Although it was tried to evaluate the tendency for socially desired response it is not possible to definitely conclude which biases may contribute to the inaccurate responses. Qualitative methods would be needed for these kind of studies. Interpretation of findings | 1. Knowledge of influential factors | The rate of physicians who ascribed a high impact on their therapeutic decision making to factors not well known was very high with values of 15% and 23% respectively. These rates decreased to 2% and 4% respectively if one concedes that factors which were only roughly known can also have a strong influence. 2. Reading and influence of different article types | The rate of respondents who stated that published clinical trials had a strong influence on their therapeutic decision making but who read such articles only infrequently was very low. But it should be taken into account that virtually all surveyed physicians read this kind of articles always or often if they appear in journals they had subscribed (207/237; 87%). 3./4. Knowledge of the NNT and a fictitious term | The rate of respondents who allegedly could explain the number needed to treat but could not explain absolute risk reduction was very high. As McColl and colleagues did not perform an analysis like this a comparison between both studies is restricted. For such an analysis raw data are required. But the data in their publication indicate that there were also inconsistencies. In their survey 35% of respondents could explain the term number needed to treat but only 31% could explain the term absolute risk . The alleged knowledge of the McNemar-Quality-Scale was lower than the knowledge of the NNT. But the value was also around 10%. One might argue that positive respondents confused the fictitious term with McNemar's statistical test or that they thought the researchers had been confused. But this seems not very likely since nobody during the development of the questionnaire referred to this potential problem. Moreover somebody who knows a statistical test would know the term Alpha- Error/Type-I-Error which was not the case in the majority of the positive respondents. The proportion of inaccurate responses to this knowledge-question should be viewed as a very conservative estimate. A recently published study found that virtually nobody who stated that he/she allegedly understands the technical terms of the questionnaire developed by McColl et al. actually did so . 5. Examining and reading the methods section of articles | As for the other variables, the proportion of positive answers was about 10%. Test for socially desired response bias | Given the other results of this analysis and the kind of response categories in this survey it seems reasonable to assume that the tendency for socially-desired responses would be the most prominent response bias. To attribute the alleged knowledge of the McNemar-Quality-Scale to the tendency for socially desired responses it must be interpreted in association with the knowledge-score and the responses to the most unknown term. If the knowledge-scores are low among those respondents who allegedly knew the McNemar-Quality-Scale, this response behavior can not be interpreted as socially desired. Other explanations have to be considered instead of. But the analysis showed that their knowledge-scores were well above the other respondents. This could lead to the conclusion that these 17 respondents (7%) have had a tendency for socially desired responses. Their knowledge of the Alpha-Error/Type-I-Error indicates on the other hand that these physicians were by all means willing to admit knowledge-gaps because they reported a lack of knowledge or understanding more frequently than the others. Therefore it seems unlikely that the inaccurate answers can be attributed to the socially-desired response bias. Acceptance acquiescence as another potential and important response bias is also unlikely due to the other findings of this analysis and the response categories in the questionnaire. The tendency for only yes- or no-responses can be ruled out as only 7 questions with a yes/no-response-category were asked. Thus it is believed that the inaccuracies in this survey are rather a problem of careless reading/answering (Which again might have been resulted from the long questionnaire or busy respondents although an association between the weekly hours of work and positive responses could not be found. Checked for variable 4; data not shown) or a failure in self-perception/overestimation of competency. Furthermore misunderstanding of questions or about specific terms might also have contributed to the inaccuracies as was shown in a recent study . Conclusions : As a result of this analysis the proportion of inaccurate or illogical responses in a survey about CME habits and information management of physicians was around ten percent. Although some researchers try to correct such inaccuracies it has to be determined how accurate such methods are. It seems unlikely that respondents had a significant tendency for socially desired responses. The analysis indicates that it rather seems to be a problem of careless reading/answering of questions, a failure in self-perception or a misunderstanding about specific terms or questions. However in order to understand response biases and the processes involved qualitative studies are needed. The method described is considered appropriate and feasible for evaluating the accuracy of responses in surveys but further research is necessary to validate it. It should be applied to future questionnaire surveys about CME habits and information management of physicians to enable appropriate assessments of such studies. Competing interests : None declared. Authors' contributions : Sven Trelle conceived, designed, conducted, analyzed, and wrote the study. Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12213183 TI - Funding source, trial outcome and reporting quality: are they related? Results of a pilot study AB - Abstract | Background | There has been increasing concern regarding the potential effects of the commercialization of research. Methods | In order to examine the relationships between funding source, trial outcome and reporting quality, recent issues of five peer-reviewed, high impact factor, general medical journals were hand-searched to identify a sample of 100 randomized controlled trials (20 trials/journal). Relevant data, including funding source (industry/not-for-profit/mixed/not reported) and statistical significance of primary outcome (favouring new treatment/favouring conventional treatment/neutral/unclear), were abstracted. Quality scores were assigned using the Jadad scale and the adequacy of allocation concealment. Results | Sixty-six percent of trials received some industry funding. Trial outcome was not associated with funding source (p= .461). There was a preponderance of favourable statistical conclusions among published trials with 67% reporting results that favored a new treatment whereas 6% favoured the conventional treatment. Quality scores were not associated with funding source or trial outcome. Conclusions | It is not known whether the absence of significant associations between funding source, trial outcome and reporting quality reflects a true absence of an association or is an artefact of inadequate statistical power, reliance on voluntary disclosure of funding information, a focus on trials recently published in the top medical journals, or some combination thereof. Continued and expanded monitoring of potential conflicts is recommended, particularly in light of new guidelines for disclosure that have been endorsed by the ICMJE. Keywords: Background : There has been increasing concern regarding the potential effects of the commercialization of research . This concern has coincided with the reduced availability of public research funds which has, in turn, translated to scientists' increased reliance on industry support. For example, in the United States, approximately 70% of clinical drug trials are now funded by industry . Industry funding of biomedical research can be viewed as a conflict of interest because the sponsor of the study has a vested financial interest in its outcome. Since even the perception of a conflict can undermine the public's trust, due diligence, in the form of recognizing and managing potential conflicts, is warranted. This is not to suggest that industry abandon its sponsorship of biomedical research. Indeed, many medical discoveries may not have occurred without industry funding. Nevertheless, fundamental errors have been noted in the design of industry-funded trials . These methodological deficiencies, perhaps operating in conjunction with the well-documented phenomenon of publication bias, have led to a preponderance of published trials that have received funding from for-profit entities and whose conclusions favour industry . Concern over the increasing commercialization of research also extends to the quality of reporting. Low quality reports have been noted among published trials and reports of low quality trials have been found to exaggerate an intervention's effectiveness . Because relationships between funding source, trial outcome and reporting quality may bias study results, making informed decision making about the merits of an intervention more difficult for clinicians and consumers alike, it is important to examine the extent to which a trial's source of funding influences its results. Methods : A convenience sample of 100 randomized controlled trials (RCTs) was identified by hand-searching recent issues of five peer-reviewed, high impact factor general medical journals (Annals of Internal Medicine, British Medical Journal, Journal of the American Medical Association, The Lancet, The New England Journal of Medicine). Issues published between January 1999 and October 2000 were searched until 20 RCTs/journal were identified. To be eligible for inclusion, the RCT needed to be published as a full report. Interventions were restricted to pharmaceuticals; medical devices, surgical procedures and methods of medical management (e.g., lifestyle counseling) were excluded. No attempts were made to limit the selection to any particular RCT design, number of treatment arms, comparator (e.g., placebo, active control, alternate dosing, herbal therapy), study population or disease category. Relevant data, including funding source(s) and primary outcome, were abstracted from each eligible RCT. Reporting quality was assessed using both a composite (overall score on the Jadad scale ) and a component (individual items on Jadad scale and adequacy of allocation concealment ) approach. The Jadad scale consists of a total of five items; two items relate to blinding, two items relate to randomization and one item assesses the description of withdrawals/drop-outs. When using the Jadad scale to score the quality of a trial report, each of the five items receives a "yes" or a "no," resulting in an overall/composite quality score that can range from 0 to 5; higher scores reflect better methodological quality . Allocation concealment was rated as adequate, inadequate or unclear in the manner proposed by Schulz et al . Allocation concealment refers to the process that prevents foreknowledge of treatment assignment and thus, shields those who enroll participants from being influenced by this knowledge. For example, a trial was rated as having "adequate" concealment if allocations were performed using central randomization; numbered/coded bottles/containers; serially numbered, opaque, sealed envelopes or if the formulations were prepared by a pharmacy. Allocation was classified as "inadequate" if assignments were made on an alternating basis or via reference to case record number or date of birth. Trials that received an "unclear" rating would have failed to provide sufficient information regarding the allocation process on which to base our decision. In all cases, reporting quality was evaluated by two independent, experienced reviewers (DM, TC). No formal training was conducted prior to evaluating the RCTs using either of the quality assessment scales since both raters have extensive experience using these methods. Moreover, since any disagreements in quality ratings were resolved by consensus, we did not undertake assessment of inter-rater reliability. In order to examine the relationship between trial outcome, funding source and reporting quality, SPSS-PC software was used to conduct statistical analyses in the form of Fisher's exact test or ANOVAs, as appropriate. The odds, and corresponding 95% confidence intervals, of unclear allocation concealment, by funding source and trial outcome, are also presented. Trials were classified according to their funding source(s) in a manner similar to that used by Rochon , permitting comparison of trials across four levels of funding: entirely industry, entirely not-for-profit, mixed and not reported. A trial was classified as having "mixed" funding if it received support from at least one industry source and at least one not-for-profit source. Because this study was restricted to RCTs that examined pharmaceuticals, industry funding is synonymous with pharmaceutical company funding. The primary outcome was defined as the one stated as such by the authors or, if there was no such statement, the one that was most clinically relevant (i.e., mortality over morbidity). If one outcome was not more clinically relevant than the others, the outcome contributing the most patients was used. On the basis of statistical interpretation of results, rather than reliance on authors' interpretations presented in discussion/conclusion sections, the outcome of trials was classified as favoured new treatment, favoured conventional treatment, neutral (i.e., non-significant) or unclear. Results : Of the 100 trials reviewed, sixty-six were funded in whole or in part by industry; 6 did not disclose their source of funding. Of these same 100 trials, 67 favoured the new therapy, 6 favoured the conventional treatment, 19 reported neutral findings while, in eight cases, the outcome was classified as unclear owing to ambiguity between the defined outcome of interest and data presented in the results section. Results for reporting quality varied according to the dimensions of quality that were measured: the overall/composite score on the Jadad scale indicated that 74% of the trials were of higher quality (scores of 3 --5) with a mean score of 3.31 (SD 1.19) while allocation concealment was found to be unclear in close to 60% of trials. The results of bivariable analyses examining associations between allocation concealment and funding source and between allocation concealment and trial outcome are presented in Table . Odds ratios and corresponding 95% CI for these associations are also presented. Data reflecting the comparison between funding source and trial outcome are presented in Table . In none of these cases were the associations found to be statistically significant. A one-way ANOVA also failed to demonstrate statistical significance between the overall/composite score on the Jadad scale, measured as a continuous variable, and funding source (F = 1.853; df = 3; p = .143), trial outcome (F = 1.003; df = 3; p =.395) and allocation concealment (F = 1.319; df = 2; p = .272). The absence of a statistically significant association between reporting quality and funding source and between reporting quality and trial outcome persisted when individual components of the Jadad scale (i.e., blinding, randomization, and description of withdrawals/drop-outs) were examined. Table 1 | Allocation concealment by funding source and by trial outcome Table 2 | Funding source by trial outcome Discussion : The results of this pilot study failed to document any association between funding source, trial outcome and reporting quality among a sample of RCTs that were recently published in the top five general medical journals. It is not known whether this finding (which has been observed by some but not by others ) reflects a true absence of an association or, instead, represents an artefact, due to limitations inherent in this, a pilot study. For example, our failure to detect any significant associations may result from a type 2 error that indicates inadequate statistical power. Although our results do not even hint at a trend (and, perhaps, reflect the emphasis now placed on disclosure at the journals included in our sample ), the potential for type 2 error is real, as suggested by the width of the confidence intervals presented in Table . Our estimates of association were based on a sample size that, while relatively small, is typical of the initial phases of a program theme, such as that upon which we were embarking; future works will accrue larger samples to reduce the likelihood of this error. Admittedly, the limitations of our study extend beyond issues of statistical power. For example, we depended on authors' disclosure of a trial's funding source(s) and the subsequent publication of this information. Recent work suggests that failure to disclose personal financial conflicts is widespread . Although this is different than reporting a trial's source of funding, it supports the notion that failure to abide by journal disclosure requirements is common. Moreover, since journals, themselves, do not always abide by their own disclosure rules , our categorization of funding source(s) may be biased. In addition, because this study focussed on recent publications of the top five general medical journals, our results may not generalize to journals that differ in their impact factor, disclosure requirements and/or reporting policies. It is important that our results be viewed in this context, particularly since some research suggests that researchers submit their "best" work to the "best" journals . Future research, relying on a modified design that has been used to address similar questions might allow more comprehensive exploration of these relationships. For example, future works, accruing a larger sample of RCTs from a wider variety of journals, should adjust for covariates (e.g., number of sites involved, number of treatment arms, sample size) in multivariable models and may find it fruitful to examine statistical outcomes in the context of effect size. It is also possible that the discordance seen between our findings and others stems, in whole or in part, from differences in the operational definition of "industry-sponsored research." More specifically, some have categorized RCTs that receive any industry funding to be "industry-sponsored" and have compared this category with those RCTs that are wholly funded by not-for-profit monies. In our study, we treated RCTs that received funding from one or more corporate sponsor and one or more not-for-profit sponsor as receiving "mixed funding" while the category "industry only" was reserved for those RCTs that did not receive any not-for-profit funds. It is important that authors acknowledge the potential for differences in operational definitions of funding source between studies to produce discordant results; this can be best accomplished by ensuring that the scheme for classifying trials according to their funding source(s) be made explicit in each report. As a result of these limitations, we are unable to conclude, with a high degree of confidence, that the absence of an association between funding source, trial outcome and reporting quality that was documented in this study reflects a benefit of strengthened disclosure requirements at the top general medical journals. Our results do, however, suggest the benefits of one particular standardized reporting requirement, CONSORT . Endorsement of CONSORT by the International Committee of Medical Journal Editors (ICMJE) may have contributed to the majority (~75%) of trial reports receiving moderate-high quality composite scores on the Jadad scale. This has not always been the case, with almost uniformly poor quality found amongst trials published prior to widespread adherence to CONSORT [,-,]. There is still room for improvement, however. Recent revisions to CONSORT and its adoption by increasing numbers of journals should address the alarming dominance of "unclear" allocation concealment seen among trials examined in this report and elsewhere [,,-]. The persistence of a preponderance of trial reports favoring novel treatments , however, remains a challenge. Conclusion : Concerted and continued efforts to monitor the reporting quality of RCTs, to ascertain the best method(s) for its evaluation and to encourage the mandatory registration of trials [,,,-] are recommended. Given that the ICMJE recently strengthened its requirement for disclosure of information as to the role(s) of study sponsor(s) in all aspects of study design, conduct and publication while, at the same time, the New England Journal of Medicine has announced that they will be relaxing their longstanding rules on conflict of interest , the question posed by this study should be revisited to allow for more definitive determination of the impact of industry sponsorship on biomedical research. We encourage journal editors to continue to work together in order to reach consensus as to the particulars of reporting requirements. Competing Interests : None declared. Authors' Contributions : TJC and DM have participated sufficiently in the work to take public responsibility for the whole content. TJC has made substantial contributions to the intellectual content of the paper as they relate to this study's conception and design, the acquisition of data, its analysis and interpretation. TJC was responsible for the initial draft of the manuscript and all subsequent revisions. NB assisted in the analysis and interpretation of the data and provided critical revisions of the manuscript for important intellectual content. DM provided the impetus for the study's conception and design, participated in the analysis and interpretation of data, provided feedback on initial drafts of the manuscript and, via administrative supervision, also permitted TC and NB the time to carry-out this work. All authors have read and approved the final manuscript. Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12225617 TI - Patient attitudes toward using computers to improve health services delivery AB - Abstract | Background | The aim of this study was to examine the acceptability of point of care computerized prompts to improve health services delivery among a sample of primary care patients. Methods | Primary data collection. Cross-sectional survey. Patients were surveyed after their visit with a primary care provider. Data were obtained from patients of ten community-based primary care practices in the spring of 2001. Results | Almost all patients reported that they would support using a computer before each visit to prompt their doctor to: "do health screening tests" (92%), "counsel about health behaviors (like diet and exercise)" (92%) and "change treatments for health conditions" (86%). In multivariate testing, the only variable that was associated with acceptability of the point of care computerized prompts was patient's confidence in their ability to answer questions about their health using a computer (beta = 0.39, p = .001). Concerns about data security were expressed by 36.3% of subjects, but were not related to acceptability of the prompts. Conclusions | Support for using computers to generate point of care prompts to improve quality-oriented processes of care was high in our sample, but may be contingent on patients feeling familiar with their personal medical history. Keywords: reminders, preventive services, quality, computer, primary care, reminders, preventive services, quality, computer, primary care, Background : Improving health care quality remains an elusive goal as many barriers exist to changing: 1) health care systems, 2) physician behaviors and 3) patient behaviors . Prior studies have shown that computerized reminders and decision support systems consistently increase the delivery of preventive services and improve the quality of outpatient care . For example, in a randomized trial of 39 primary care physicians, McPhee and colleagues found that computerized reminders led to significant increases in fecal occult blood testing, pap testing, smoking cessation counseling, and diet counseling . Although obstacles exist to incorporating reminder systems into routine care , to our knowledge, no study has examined patient acceptance of computerized reminders to improve routine health service delivery. The goal of this study was to examine patient acceptability of point of care computerized prompts as a method of improving quality of care among a diverse group of primary care practices in Rhode Island. We hypothesized that patients would support the use of computerized reminders for health screening tests and behavioral counseling, but not for prompting physicians to make changes to their treatment plans. We also hypothesized that interest in using computerized reminders would be strongly related to an individual's general level of comfort with computers. Methods : Methods: measures | Computer Acceptability | A series of four focus groups were held with primary care patients to understand the important issues in designing and implementing patient-centered computer systems in primary care settings. Focus groups included 7 --11 subjects, recruited from posters placed in public areas of the hospital and via emails send over the hospital intranet. Subjects were paid $30 for participating in the focus groups, and one of the investigators (CNS) served as the facilitator. Attitudes and concerns about using computers to improve health care quality that resonated among group members were considered for inclusion in the survey instrument. Instrument items that were analyzed for this paper, reflecting domains identified in the focus groups, included: 1) patient attitudes toward the possible uses of computer technology to improve health care quality ("to prompt your doctor to do health screening tests", "to prompt your doctor to counsel you about your health behaviors", and "to prompt your doctor to change your treatments for conditions like high blood pressure and diabetes") and 2) concerns about data security (".. would you be concerned that someone would be able to steal your personal information" from the computer"). Attitude questions were asked on a 5-point Likert-type scale from "definitely no" (1), to "definitely yes" (5). A summative variable representing acceptability of using computers to improve health services was created by summing responses to the three questions examining the acceptability of using computers for different uses ("prompt your doctor to do health screening tests", "prompt your doctor to counsel you about your health behaviors", "prompt your doctor to change your treatments for conditions like high blood pressure and diabetes") from Table . The variable had a minimum of 3 (least accepting) to 15 (most accepting). Table 1 | Attitudes toward using computers to improve health services Two other themes that emerged from the focus groups were 1) ability to use computers and 2) ability to answer questions about their health. Ability to use computers was assessed with the following item: How comfortable do you feel using computers, in general?" measured on a 6 point scale from "very comfortable" (6), "somewhat comfortable" (5), "neither comfortable nor uncomfortable" (4), "somewhat uncomfortable" (3), "very uncomfortable" (2) and "I have never used a computer" (1). Ability to answer questions about one's health was assessed using the following item "Do you think you could use a computer to answer questions about your health?" and was scored a 5-point Likert-type scale from "definitely no" (1), to "definitely yes" (5). The questionnaire item regarding computer acceptability was adapted from the Pew Internet Tracking Project . Patient assessment included demographic information. Brief screening questions for age, gender, educational attainment, chronic health conditions (e.g., diabetes, depression) and health risk factors were adapted from the year 2000 Behavioral Risk Factor Surveillance Survey and the 1995 National Health Interview Survey. Methods: subjects | As part of a larger study to examine the feasibility of using a tailored message computer program to enhance physical activity and smoking cessation, one hundred and thirty letters were sent to a random sample of primary care providers in the Providence area. Eleven physicians expressed interest and were enrolled. After recruitment, one physician, a general internist, later dropped out of the study after moving his practice. The ten remaining community-based primary care practices enrolled in the study included family practice (4), internal medicine (5) and obstetrics (1). From 3/1/2001 to 6/1/2001, two research assistants approached 148 consecutive adult outpatients (10 --15 per office) to recruit 124 patients (83%) to complete the survey after their visit with their doctor, but before they had left the office. The introduction to the survey described the planned use for the computerized reminder system in which physician's offices would have computers installed and "each patient would spend 5 --10 minutes answering questions on the computer before every visit with their doctor". The introduction specified that the computerized prompts would be based on data entered by patients, but did not specify whether or not additional data sources (e.g., electronic medical records) would be used to supplement or verify data entered by patients. The protocol was approved by Institutional Review Board of The Miriam Hospital. Methods: data analysis | All data analysis were carried out using SPSS for Windows, version 10.1.0. Groups were compared using chi-square testing for categorical variables and analysis of variance testing for continuous variables. Due to the presence of some missing data, some row totals are less than 124. The Student's t-test (variables with 2 categories) and Analysis of Variance testing (variables with > 2 categories) was used to examine group differences among categorized variables. Variables significantly associated with acceptability of using computers to improve health services were included in a multiple linear regression model to test for an association between attitudes toward computers and acceptability of using computers to improve health services, while adjusting for potential confounders. Results : The average age of participants was 42.8 years and nearly 75% were female (see Table ). Approximately half of participants reported using a computer at least weekly (49.2%) while 12.7% had never used a computer. Over 85% of patients answered "yes" (either "definitely yes" or "probably yes") when asked about their support for using computers to prompt their doctor to either do health screening tests (92.2%), do health behavior counseling (92.2%) or change patients' medical treatments if necessary (85.5%). See Table for results of bivariate testing. Males, patients who were more comfortable using computers and patients who were more comfortable using a computer to answer questions about their health were more accepting of using the point of care computerized prompts to enhance the quality of care. For example, those who were the most comfortable in their ability to answer questions about their health using a computer had a mean acceptability score of 13.7 (of 15) compared to 9.7 for those least confident (p = .001). Age, education level, minority status, concern about data security and presence of chronic illnesses were not associated with acceptability of using computers to improve health services. Table 2 | Background characteristics by interest in using computers in the doctor's office to improve health services In multivariate testing, the only variable associated with acceptability of using point of care computerized prompts to enhance the quality of care was a patient's comfort in their ability to answer questions about their health using a computer (beta = 0.39, p = .001). Discussion : The main findings in our study were 1) point of care computerized prompts to enhance the quality of care, specifically preventive services, health behavior counseling and chronic disease management, were highly acceptable to primary care patients, 2) patients' ability to answer questions about their health, but not patients' experience in using computers, was significantly associated with acceptance of the computerized prompts and 3) concerns over data security, while present, were not associated with acceptance of the computerized prompts. As the use of medical informatics becomes more widespread, computerized reminders and decision support systems may play an important role in improving the delivery of preventative services and outpatient care. An important component of the success of these programs is patient acceptability, yet no study to our knowledge has examined this. Over 85% of patients answered "yes" (either "definitely yes" or "probably yes") when asked about their support for using computers to prompt their doctor to either do health screening tests (92.2%), do health behavior counseling (92.2%) or change patients' medical treatments if necessary (85.5%). We suspect this support for computerized reminders systems reflects a general belief that computerized feedback is accurate, and possibly more accurate than the decisions made by their physicians. In their 1999 publication, "To Err Is Human", the Institute of Medicine concluded that as many as 98,000 patients die each year from medical errors . This book has received much press coverage and may have enhanced consumer awareness of the prevalence of errors and, therefore, the need for computer systems that help physicians make complex medical decisions. In our study, acceptability was related to patients' confidence in their ability to answer questions about their health, even after adjusting for patients' self-reported experience in using computers. We feel that this is due mainly to patients' lack of knowledge about their health, specifically their knowledge of test results, as they would be required to enter these results into the computer to use the proposed system. Not being aware of one's medical history, such as blood pressure or cholesterol values, would prevent them from participating in such as system. This is consistent with the findings of Branch and colleagues, who noted that patients who were unaware of their test results, despite being tested, were less likely to report receiving other health promotion services . Among 4298 patients who reported being tested for hypertension and hyperlipidemia in the previous year, 449 (10.4%) did not know either test result, and a presumably greater percentage did now know at least one of the test results. The lack of knowledge of one's health history may be related to an individual's health literacy which, though not measured in the current study, has been associated with poorer health outcomes, such as hospitalization, and knowledge of one's chronic disease . Though the computer system described in the introduction to the survey was located in a doctor's office, more than a third (36.3%) of the patients in our sample had concerns that their personal information would be stolen from the computer. Though the Health Insurance Portability and Accountability Act (HIPAA) of 1996, which includes standards for privacy of individually identifiable health information, may allay these concerns over data security , the present study shows that these concerns were not related to acceptability of the computerized prompts. This is consistent with the history of internet commerce which has grown at an impressive rate, despite individual's concerns over the security of their credit card and other personal information . The study has several noteworthy limits. First, the small sample size and geographical range limit the generalizability of the findings. However, the rates of health conditions including high cholesterol (23.5%) and high blood pressure (20.7%) are similar to other larger samples . These findings will need to be repeated in other populations and in larger sample sizes. Second, the brevity of the survey, due to concerns over subject burden in the primary care setting, leaves open the possibility that an unmeasured variable, such as health literacy may be a source of residual confounding. Future studies should include measures of health literacy, such as the Test of Functional Health Literacy in Adults . Conclusions : Computerized reminders and decision support systems have repeatedly been shown to improve the quality of outpatient care. Though the personal computer and the Internet have tremendous potential for changing the way that health care is delivered, both have failed, as yet, to facilitate the dissemination of computerized reminder systems. This study shows that acceptability of point of care computerized prompts to enhance the quality of care is high. To use such systems, however, patients will need to be more knowledgeable about their health, to enable them to feel comfortable answering questions that will be used to create the computerized prompts. List of abbreviations : Health Insurance Portability and Accountability Act (HIPAA) Competing interests : None declared Author's contribution : CNS participated in the collection of data, the statistical analysis and in the preparation of the manuscript. JD participated in the interpretation of results and in the preparation of the manuscript. PM participated in the design of the study and in the statistical analysis and interpretation of results. Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12204094 TI - Pro-domain removal in ASP-2 and the cleavage of the amyloid precursor are influenced by pH AB - Abstract | Background | One of the signatures of Alzheimer's disease is the accumulation of aggregated amyloid protein, Abeta, in the brain. Abeta arises from cleavage of the Amyloid Precursor protein by beta and gamma secretases, which present attractive candidates for therapeutic targeting. Two beta-secretase candidates, ASP-1 and ASP-2, were identified as aspartic proteases, both of which cleave the amyloid precursor at the beta-site. These are produced as immature transmembrane proteins containing a pro-segment. Results | ASP-2 expressed in HEK293-cells cleaved the Swedish mutant amyloid precursor at different beta-sites at different pHs in vitro. Recent reports show that furin cleaves the pro-peptide of ASP-2, whereas ASP-1 undergoes auto-catalysis. We show that purified recombinant ASP-2 cleaves its own pro-peptide at ph 5 but not pH 8.5 as seen by mass spectrometry, electrophoresis and N-terminal sequencing. Conclusion | We suggest that ASP-2 processing as well as activity are influenced by pH, and hence the cellular localisation of the protein may have profound effects on the production of Abeta. These factors should be taken into consideration in the design of potential inhibitors for these enzymes. Keywords: Background : Alzheimer's disease is a common age-related dementia which is characterised pathologically by the appearance of brain senile plaques composed primarily of aggregated forms of Abeta. These are 39 --43 residue peptides released following proteolytic processing of the transmembrane precursor glycoprotein, APP. The amyloidogenic pathway requires the APP to be sequentially cleaved by beta and gamma secretases . beta-Secretase cleaves APP close to the membrane to produce betaAPPs (secreted), and the 12-kDa, C100 transmembrane stub, subsequently cleaved by gamma-secretase to produce the Abeta peptide and a cytoplasmic fragment with very short half life. alpha-Secretase cleaves APP within the Abeta sequence thus preventing its formation producing the N-terminal alphaAPPs domain and the 10-kDa membrane-localised C-terminal stub, C83. As aggregated Abeta is thought to promote neuronal death , the secretases represent potential drug targets for the treatment and/or prevention of AD. Presenilin-1 was suggested to be the ideal candidate for gamma-secretase whereas alpha-secretase has been characterised as ADAM10 disintegrin and metalloprotease . Recently, several groups used expression cloning, genomic search, or purification and proteomic analysis to clone and identify beta-secretase as an aspartic endopeptidase (EC 3.4.23) named BACE (beta-site APP clevage enzyme), ASP-2 (aspartic protease 2), or memapsin 1. An additional candidate, ASP-1, BACE-2 or memapsin-2, has also been cloned . The ASP-2 gene codes for a signal peptide, a pro-peptide (22-TQHGIRLPLRSGLGGAPLGLRLPR-46), followed by the catalytic domain, a transmembrane segment and a cytoplasmic C-terminal tail. Several cysteine residues are present, six of which are in the lumenal domain which may form intramolecular disulphide bridges contributing to the folding of the active site . Both ASP-1 and -2 are extensively glycosylated and phosphorylated , and contain S-palmitoyl groups which may aid membrane anchorage. Nearly all aspartic proteinases (EC 3.4.23.X) are synthesized as zymogens which are converted to active enzymes at acidic pH by proteolytic cleavage of the pro-segment . This process is autocatalytic for some pro-enzymes such as pepsinogen and cathepsins D and E . Furin is thought to cleave the ASP-2 pro-domain , though other pro-protein convertases were effective as well . In contrast, ASP-1 has been demonstrated to cleave its own pro-peptide . Although the pro-sequences of ASP-1 and ASP-2 are dissimilar we investigated whether ASP-2 also cleaves its pro-peptide. The activity of ASP-2 was assessed using the Swedish mutant form of APP which is more readily cleaved by beta-secretase than the wild type . We investigated the effects of pH on ASP-2 processing and activity. We report that pH affects the pro-domain removal of ASP-2 in vitro as well as its site of beta-secretase cleavage in APP to produce Abeta. Results : Expression and Ni2+-affinity purification of recombinant ASP-1 and ASP-2 in HEK293 cells | HEK293 cells were transfected with the beta-secretase candidates ASP-1 or ASP-2 pcDNA3.1mycHis clones and selected for stable transfection with Geneticin (800 mug/ml). HEK293 cell lines have been used by others to overexpress these enzymes and have also been used to study the effects of beta-secretase cleavage on recombinant APP . Cell lysates from positive clones were analysed by SDS-PAGE and western blotting with an anti-myc antibody to detect the recombinant ASP-1 and ASP-2 proteins. The myc epitope is encoded in frame downstream of the C-terminal of the proteins followed by a 6-Histidine tag for affinity purification. Both recombinant ASP-1 and ASP-2 were expressed and detected by western blotting as 55 --70-k and 65 --80-k bands, respectively, higher than the estimated molecular weights of the pro-apoproteins (53.4 and 53.7-kDa including the pro-peptide domains), mainly due to glycosylation as predicted by their amino acid sequences and shown in recent reports by Charlwood et al., 2001. Recombinant ASP-2 was purified by Ni2+-affinity chromatography. Silver-stained SDS-PAGE confirmed the purity of ASP-2 from other contaminants as seen in figure ; lower bands (dotted arrows) seen in the purified preparation are due to artefactual materials also present in the negative control sample (note the presence of higher molecular weight forms of ASP-2). Immunoblotting of the same preparation of ASP-2 detected with the anti-myc antibody, (figure ) confirms purification of ASP-2. The most prevalent of these ASP-2 proteins was a 75-k band. We used an affinity-purified antibody raised against the C-terminal peptide of ASP-2 (anti-ASP-2) to characterise the recombinant proteins. This antibody was specific for ASP-2 and did not cross-react with ASP-1 or other cellular proteins as characterised by ELISA and western blotting. Figure 1 | Ni2+-affinity purification of ASP-2. Ni2+-affinity purification of ASP-2. A detergent extract was prepared from a total of 50 x 106ASP-2-transfected cells as described in the methods, in the presence of protease inhibitors and loaded on an equilibrated Ni2+-charged agarose column, following dilution of the lysate to 0.5% (wt/vol) detergent and pure protein was eluted in fractions of 1 ml with 300 mM imidazole solution with 0.1% detergent. Eluates were loaded on a SDS-PAGE and proteins were detected by silver staining, (panel A) and anti-myc antibody developed by ECL (Panel B). Molecular weight standards are indicated in the first lane whereas the last lane shows a sample of elution buffer in loading buffer without protein. Typically, 20% yield was recovered following purification, amounting to approximately 40 --60 mug of protein for a preparation from 50 x 106 cells. Solid arrows indicate the positions of the ASP-2 protein; the double solid arrow designates a pitative ASP-2 dimer. Removal of N-linked oligosacharides increased the relative mobility of ASP-2 from 65 --80-k to 50 --60-k, a size resembling the estimated molecular weights of the core protein confirming glycosylation as reported by others either by suppression of N-glycosylation or deglycosylation . When increased amounts of protein were loaded (as in figure ) the anti-ASP-2 IgG detected ASP-2 bands of higher Mw, larger than 200-k or at Mr~140-k, shown with triple and double arrows, respectively; the 140 k band was also visible by silver staining as indicated by the double arrow in figure . These correspond in size to three or two ASP-2 molecules, respectively, implying ASP-2 oligomerisation into a trimer and dimer. The sensitivity of the putative dimer (140-k) to PNGase resulted in a reduction of size by ~30-k, to yield a band migrating as 100-k, equivalent to the behaviour of 2 linked monomers. Reduction with DTT and beta-mercaptoethanol (figure ) almost completely removed the ~200-k protein and increased the intensity of the 75-k band, but failed to affect the appearance of the putative dimer (140-k) suggesting strong intramolecular interactions. The oligomers were visible in freshly produced and purified ASP-2, though the 200-k protein was more prominent (data not shown) upon longer storage and at higher protein concentrations indicating that this was partly due to non-specific aggregation. An additional observation was the delayed mobility of ASP-2 in reducing conditions (figure ) which is possibly due to loss of globular structure during reduction of intramolecular disulphide links described previously by Haniu et al., 2000. Figure 2 | ASP-2 post-translational modifications; glycosylation and disulphide linkages ASP-2 post-translational modifications; glycosylation and disulphide linkagesPanel A: ASP-2 samples were denatured by boiling in the presence of SDS and prior to treatment with PNGase, according to the manufacturer's protocol. Glycosylated and deglycosylated samples are depicted. A reduction in size of ~15-k is seen in the presence of PNGase due to the removal of N-linked sugars. Note the presence of higher ASP-2 bands, indicated with double arrows, which are similarly, deglycosylated. Panel B: ASP-2 samples were denatured in the presence or absence of a mixture of 10 mM DTT and 2% (vol/vol) beta-mercaptoethanol designated as reducing agent. A reduction in the amount of the largest ASP-2 protein (marked with 3 arrows) is noted in the presence of reducing agent as well as a reduction in the mobility of ASP-2 proteins. Ni2+-affinity purified ASP-2 was used at a concentration of 40 nM. Western blots were detected with anti-ASP-2. Bold arrows designate larger putative, dimer and multimer. ASP-2 undergoes a two-step pro-peptide cleavage at acidic pH | We investigated the processing of ASP-2 at different pH values, since self-activation of other aspartic proteases such as pepsinogen has been shown to be sensitive to pH. Pure ASP-2 was dialysed into a buffer of ph 5 or pH 8.5, immediately following purification. Incubation of the protein at room temperature was carried out prior to PAGE analysis. A sample before dialysis was also analysed alongside the ASP-2 at pH 8.5 and ph 5. A western blot analysis of these samples, depicted in figure , reveals a small shift in molecular weight of 3 k for the monomeric band and ~8 k for the dimer at ph 5. This was the first indication that cleavage of the pro-peptide (2.56 k) was occurring, which was also mirrored, by increased mobility in the putative dimer. PNGase treatment of the ph 5 and pH 8.5 ASP-2 samples revealed the deglycosylated proteins to be resolved in 3 and 2 bands (figure ), respectively. The bands of the ph 5 sample were smaller (~3-k) than the ASP-2 bands in the pH 8.5 sample. The fact that PNGase did not reduce the ASP-2 protein to a single band of the unmodified protein is probably due to other modifications mentioned earlier, such as palmitoylation and phosphorylation . The triplet at ph 5 is consistent with deglycosylated proteins with the pro-segments removed (processed), with and without additional modifications. The highest band is possibly non-processed deglycosylated protein, because its intensity decreases upon incubation at 37C. Figure 3 | Reduction in size of ASP-2 upon incubation at acidic pH proposes autocatalytic pro-peptide cleavage Reduction in size of ASP-2 upon incubation at acidic pH proposes autocatalytic pro-peptide cleavagePanel A: Pure ASP-2 samples were dialysed at ph 5 and pH 8.5 immediately following purification and subsequently incubated for 2 hrs at room temperature as described in the methods. The initial non-dialysed sample was also incubated before all the samples were all analysed by SDS-PAGE and western blotting. A reduction of the size of ASP-2, ph 5 was observed compared to the initial eluate and the pH 8.5 samples. Note the presence of higher Mw bands ~140-k which also displayed increased mobility at ph 5. Panel B: The same ASP-2 samples dialysed at different pH were treated with PNGase for better resolution of the protein bands. The samples dialysed at pH 5 were either not incubated (first two lanes) or pre-incubated before PNGAse treatment. These samples needed to be titrated to pH 8.5 before deglycosylation. Similarly samples at pH 8.5 were treated or not with PNGase as indicated on the figure. Compared to the respective sample at pH 8.5, the ph 5 deglycosylated ASP-2 reveals the lowest ASP-2 band visible on this blot at ~50-k, possibly mature protein without the pro-peptide and sugars, suggesting maturation of ASP-2. Pre-incubation of the sample at 37C increased the intensity of this band compared to the samples that were not pre-incubated implying that its appearance is time-dependent on incubation of the protein at a certain pH. For both panels sizes of molecular weight standard proteins are indicated on the left. Both panels were probed with anti-ASP-2 IgG and developed with ECL. Pure ASP-2 that had been incubated at ph 5 after adjustment of the pH with acetic acid was analysed by mass spectrometry to further detect the cleaved pro-peptide (figure ). The sample dialysed at ph 5 could not be used since the small, cleaved fragments would have been dialysed out of the 4 k-exclusion membrane. Compared to pH 8.5 (profile B), the sample at ph 5 (profile A) displays two distinct peaks of size 1094 and 1352 Da. The peak at 1094 corresponds closely to the estimated m/z of the N-terminal fragment of the pro-sequence namely, QHGIRLPLR (1089Da) (the N-terminal fragment from the Pro-sequence from which the N-terminal T had been removed-as found by N-terminal sequencing-see below, P1) and the peak at 1352 corresponds to a dehydrated form (-18 mass) of the penultimate N-terminal fragment of the Pro-sequence, P2 SGLGGAPLGLRLPR (1364 Da) respectively. Figure 4 | Mass spectrometry of ASP-2 samples at ph 5 and 8.5 Mass spectrometry of ASP-2 samples at ph 5 and 8.5 Pure ASP-2 samples that were prepared as described were titrated to pH 5 and pH 8.5 and incubated at room temperature. The first panel shows a profile following incubation at pH 8.5 and the second panel, at pH 5. The pro-peptide has an estimated Mw = 2535-Da. the ph 5 contains two peaks of 1094.4 and 1352.3 that are not present at pH 8.5. These correspond closely to the estimated mw of two fragments of the pro-sequence namely, QHGIRLPLR (1089) and SGLGGAPLGLRLPR (1364). Solid lines display the detected peaks in each sample. Dotted lines indicate the estimated positions corresponding to the pro-peptide fragments as calculated above; these estimated values of these peaks are boxed. The profiles seen indicate percentage of intensity against mass/charge ratio. The mass spectra indicated firstly that pro-peptide cleavage occurs readily at ph 5 at two sites within the pro-domain. N-terminal sequencing of these protein preparations revealed the presence of two sequences in the pH 8.5 preparation, namely ETDEEPEE (mature protein starting at E46), and QHGIRL (immature protein with an intact pro-domain starting at Q23). The presence of protein without its pro-domain, within the pH 8.5 sample, is not surprising as this protein was purified from the cells which themselves are expected to process nascent ASP-2. In contrast, Edman degradation of the preparation at pH 5 gave the sequence ETDEEPEE, predominantly matured protein, which was produced by processing in vitro following purification and dialysis -at low pH, suggesting that pro-peptide cleavage of ASP-2 can occur at two sites by autocatalysis at ph 5 in vitro. To further ascertain autocatalytic cleavage we added EDTA to inhibit endogenous cleavage by furin-like proteases as well as the BACE inhibitor (H-Lys-Thr-Glu-Glu-Ile-Ser-Glu-Val-Asn-Sta-Val-Ala-Glu-Phe-OH, Bachem, UK) to the incubations at ph 5. Mass spectrometric analysis of those samples (figure ) showed that EDTA had no effect on the pro-peptide cleavage of ASP-2 as the previously identified peaks of 1094 and 1352 Da are still present in this sample as they are in the control which did not contain any inhibitor. Most importantly incubation of ASP-2 with the BACE inhibitor resulted in the disappearance of these peaks suggesting that pro-peptide cleavage is a result of autocatalysis. Figure 5 | BACE inhibitor reduces pro-peptide cleavage of ASP-2 at pH 5 BACE inhibitor reduces pro-peptide cleavage of ASP-2 at pH 5 Pro-peptide cleavage of ASP-2 results in two peaks of 1094.4 and 1352.3 not present in the sample incubated with the BACE inhibitor (Mw = 1651 g/mol). These peaks are present when ASP-2 was incubated with EDTA. Samples of pure ASP-2 were prepared as described. Titration to ph 5 was carried out at 4C and immediately EDTA (5 mM) or BACE inhibitor (20 muM, according to the protocol in Hussain et al. 2001) were added. Control ASP-2 at ph 5 contained no inhibitor. The peak of 1750 seen in the BACE inhibitor sample is also present in the inhibitor profile (not shown) and in agreement with its molecular weight is the inhibitor itself; additional larger bands are also present the inhibitor control sample. Following incubation samples were analysed by mass spectrometry as described in the methods. Diverse cleavage of APPswe by ASP-2 at pH 5 and pH 8.5 | The activity of recombinant ASP-2 was assessed using the Swedish mutation APP (APPswe) purified using the anti-CT15 IgG from stably transfected HEK293 cells. Initially [35S]-labelled C-terminal APPswe polypeptides were isolated by immunoprecipitation with anti-CT15 IgG and protein-A-agarose and resuspended in reaction buffer (pH 8.5) prior to the addition of ASP-2. Figure , depicts an autoradiogram of a 16.5% Tris-Tricine gel where APPswe was incubated with or without ASP-2. In the absence of ASP-2 we noted the presence of a small band of ~9.25 (+- 0.6, n = 3)-k corresponding to the C-83 APPswe fragment produced by endogenous alpha-secretase cleavage. The addition of ASP-2 resulted in an additional band at ~12.4-k consistent with the estimated size of the C-terminal stub after cleavage of beta-secretase at the Abeta (1 --40/42) site. To confirm the identity of this band the same reactions were carried out with non-radiolabelled substrate and the proteolytic fragments were detected by western blotting using either the anti-CT15 antibody (figure top panels) or the 6E10 (figure bottom panels) antibody, specific for the N-terminal of Abeta. These reactions were carried out both at optimum pH for aspartic protease activity, pH 5, as well as at pH 8.5 for comparison and the whole sample was analysed. ASP-2 was also pre-incubated at the respective pH. The arrows indicate the alpha-secretase cleaved fraction to be present in both pH 8.5 and ph 5. Two fragments (12.39 (+- 0.24, n = 3) and 10.43 (+- 0.45, n = 3)-k, designated beta1-C100 and beta2-C90, respectively) appear from ASP-2 cleavage at ph 5, consistent with fragments occurring by beta-secretase cleavage at the Abeta (1 --40/42)-producing beta1 site starting with D1, and the E11, beta2-site resulting in Abeta (11 --40/42). Incubation of the purified recombinant ASP-2 with the APPswe at ph 8.5 also produced cleavage at the beta1-site as only the 12.39 k band is visible. The 12.4-k band was also labelled by the 6E10 antibody further confirming that it is the result of APPswe cleavage at the beta1-site. These results reveal that the activity of ASP-2 is influenced by pH which apparently governs the site at which ASP-2 cleaves the APPswe protein. Figure 6 | In vitro cleavage of APPswe by pure recombinant ASP-2; differential cleavage occurs at acidic or alkaline pH In vitro cleavage of APPswe by pure recombinant ASP-2; differential cleavage occurs at acidic or alkaline pHPanel A: anti-CT15 IgG immunoprecipitated (35S)-labelled APPswe (2 muM APPswe) was incubated with either just buffer (lane 1) or with pure ASP-2 at 6 nM (lane 2). The total reaction was analysed on a16.5% Tris-Tricine gel, stained and fixed, and amplified as described in the methods. The gel was then dried and exposed on a Kodak-X-Omat film. Panel B: Immunoprecipitated APPswe samples as above were incubated with or without ASP-2 as described above at pH 5 or ph 8.5 as indicated in the figure. The blots were developed with anti-CT15 IgG that is specific for the C-terminal of APP. The bottom panels are the same reaction probed with the 6E10 clone antibody, which was raised specifically against the first residues of Abeta and thus only recognises beta-secretase cleaved fragments that produce Abeta. Panel C: A typical gel calibration standard curve for the low molecular weight standards (carbonic anhydrase, 34.3 kDa, soybean trypsin inhibitor, 26 kDa, Lysozyme, 17.9, kDa, Aprotinin, 8 kDa, insulin, 4 kDa) plotted as the Log10 of molecular weight against Rf, for size estimation of peptides. For all panels alpha- and beta-secretase cleavage are designated by arrows and the letters alpha or beta for the respective secretase cleavage. beta1 and beta2 indicate cleavage at the Aspartate (1) and the Glutamate (11) sites of Abeta, producing the C-100 and a C-90 fragment respectively. Discussion : Understanding what governs the activity of ASP-2 toward the APP substrate at a cellular and molecular level could facilitate the discovery of compounds that could inhibit the development of AD. It was therefore important to study the maturation, processing and activity of ASP-2 and relate that to the proteolytic events that lead to Abeta production. In agreement with several other reports glycosylated proteins were successfully produced in HEK293 cells displaying mobility of ~55 --70 and 65 --80-k, for ASP-1 and ASP-2, respectively. The broadness of these bands suggested a heterogeneous protein population, as a result of variable translational modification which was mainly due to glycosylation and possibly pro-processing and as also detected by others in the fully glycosylated, endoglycosidase H-resistant ASP-2 forms . Using a new anti-ASP-2 IgG, we detected the presence of higher molecular weight proteins in the purified preparations, by analysis under denaturing SDS-PAGE, which corresponded in size to ASP-2 homo-multimers. Oligomerisation has also been identified for pepsin and cathepsin E . A putative dimer was clearly noted at ~150-k, corresponding to (ASP-2)2, which was also sensitive to deglycosylation giving rise to a change in mobility which corresponded to two monomers. The appearance of the largest ASP-2 forms was partly dependent on concentration and length of storage suggesting this is partly due to non-specific aggregation. In contrast, the putative dimers (~140 kDa) were evident even in fresh samples and were mainly insensitive to reducing agent providing evidence for the existence of strong non-disulphide protein-protein interactions not broken by high salt and non-ionic detergents used during purification or the strong denaturing conditions of SDS-boiling used during SDS-PAGE. Recently a laboratory in Germany reported the isolation of native active ASP-2 dimers from human brain homogenates, (Multhaup, G. and colleagues from Germany, unpublished data). These native dimers were similar to those observed in our preparations as they were SDS-resistant. Activation of acidic proteases by removal of a pro-peptide occurs either by autocatalysis as in the case of pepsinogen and other aspartic proteases, or the action of other proteases, like Cathepsins S and L which activate Cathepsin C . Several reports by other laboratories, showed furin to successfully cleave the ASP-2 pro-segment also demonstrating that recombinant ASP-2 does not cleave its own synthetic or recombinant fusion pro-segments. However, Hussain et al. (2001), have recently shown that minimal cleavage of ASP-1 pro-domain is achieved by ASP-2 whereas ASP-1 displayed a unimolecular pro-peptide auto-removal. Here we demonstrated in vitro cleavage of the ASP-2 pro-peptide in full-length, pure, recombinant proteins instigated by incubation of ASP-2 at acidic pH, implying autocatalysis. We propose that cleavage occurs at two sites RLPLRS32 and RLPRE46 within the pro-domain, suggesting a two-step cleavage resembling the mechanism of maturation of other aspartic proteinases by autocatalysis (for a review see Richter et al., 1998). We have not investigated to see if the pro-sequence removal from ASP-2 is a unimolecular reaction, or due to the attack of a second ASP-2 molecule. APP cleavage at both the Asp1 (beta1-site) as well as the Glu11 (beta2-site) positions shows that ASP-2 has a preference to cleave N-terminally to an acidic residue as is the case for the pro-peptide cleavage. However, although the RXXR found in the ASP-2 pro-sequence is thought to be a minimal sequence required for a pro-peptide convertase furin prefers to cleave C-terminal to two adjacent basic residues. Recent results showed that ASP-2 processing takes place in furin-deficient cell lines and suggested that other pro-peptide convertases are involved whereas Bennett et al., 2000 observed that a proportion of recombinant ASP-2 is processed in cells untransfected with furin. Further to those findings we suggest that ASP-2 can remove its pro-domain by autocatalysis in vitro, as the BACE inhibitor inhibits this process whereas EDTA does not, so that an additional pro-peptide convertase may not be required. Hence, it is possible that autocatalysis could be preferred under conditions of acidic pH, or in cells or cellular subcompartments where pro-peptide convertases are not expressed. Although other reports have shown that ASP-2 does not cleave its pro-domain in vitro the difference between those and our observations could be due to differences in the protein preparations. We used full-length pure proteins for enzyme and substrate whereas, synthetic peptides as well as truncated fusion proteins were employed in those studies. These factors in addition to differences in expression systems (and modifications) could influence the results observed herein as encountered for pepsinogen, which displays different activation kinetics as a fusion than as a whole protein . To conclude we suggest that ASP-2 may cleave its pro-peptide in a two-step mechanism in the absence of any other protease. Multiple-step removal of pro-sequences has been seen with other aspartic proteinases, where complete activation sometimes requires cleavage at more than one site . We showed in vitro beta-secretase activity of ASP-2, producing a C-terminal APP fragment, Mw 12-kDa, consistent with APPswe cleavage at the beta1-site. This cleavage occurs efficiently at ph 8.5 although the optimum pH of ASP-2 activity is acidic. We must mention though that a proportion of ASP-2 in the ph 8.5 preparation is processed (as purified from the cells) and could contribute to the beta-cleavage of APP we observe at ph 8.5. Cleavage of APPswe at ph 5, however was different to pH 8.5 as it resulted in two C-terminal APP fragments each corresponding to cleavage at the Abeta-producing site (beta1), and the beta2-site described by Creemers et al. (2000), to be the most favourable producing N-terminally truncated, highly cytotoxic , Abeta peptides (11 --42) found in patients with the Flemish mutation. Creemers et al. (2000) suggested that cleavage at these sites is governed by the expression levels of ASP-2, in vivo, rather than its maturation. In addition to that we report here that our results show that pH can also determine the site of APPswe beta-cleavage by ASP-2. Results from our laboratory have localised beta-cleavage of APP in the TGN and early endosomes, and recently Huse et al. (2000) showed TGN localisation for ASP-2 indicating that these compartments are loci for beta2-site cleavage of APP as their pH is acidic. On the other hand, localisation in the endoplasmic reticulum and early Golgi imply processing at neutral to alkaline pH that according to our results gives cleavage at the beta1-site and production of Abeta (1 --40/42). Conclusions : Taken together, our results of ASP-2 pro-peptide processing and diverse APP proteolysis, both happening at ph 5, imply an additional mode of cellular control of APP beta-cleavage whereby subcellular localisation of APP and ASP-2 in different compartments where the pH is either acidic or alkaline influences the proteolytic processing of both ASP-2 as well as of APP by ASP-2 itself and consequently the production of Abeta. Thus APP cleavage at ph 5 is relevant to Alzheimer's in patients with the Flemish and Dutch mutations who demonstrate increased deposition of N-terminally truncated Abeta, resulting from APP processing at the beta2-site. Methods : Materials | Culture media and antibiotics were from Life Technologies, Paisley, UK. All culture plasticware, were from Nalge Nunc International, Loughborough, UK. Amino Acids for peptide synthesis were from Nova Biochem, Nottingham, UK. Freunds Adjuvant, standard chemicals, 9E10 clone ascites fluid (anti-myc), Triton X-100, Nonidet P40, and Kodak X-Omat film were purchased from Sigma-Aldrich Company Ltd, Dorset, UK, 6E10 (anti-Abeta) from ID labs, P.O. Box 3556, Glasgow, Scotland, United Kingdom. C-18 ZIP tips, Centricon-30 concentrators, PVDF membrane were purchased from Millipore. Automated N-terminal peptide sequencing was carried out by AltaBiocscience, University of Birmingham, Edgbaston, UK. All secondary horseradish peroxidase-linked anti-species IgG, Enhanced chemiluminescence reagents, ECL Hyper film, Superdex 200HR, Sephadex G-10, were from Amersham Pharmacia Biotech, UK Ltd, Buckinghamshire, UK, whereas the Ni2+-affinity resin NiNTA was from Novagen, Nottingham, UK. Pfx2 Lipids were initially purchased from Invitrogen Life Technologies, Paisley, UK. Ready-made Tris-Tricine gels and low molecular weight Kaleidoscope Polypeptide Standards were from BIORAD, Hemel Hempstead, UK. Prestained protein molecular weight markers and Acrylamide mixture (30% acrylamide: 0.8% (wt/vol) bisacrylamide) from National Diagnostics, East Riding of Yorkshire, UK. Protease Inhibitors and N-glycosidase-F were purchased from Roche Molecular Biochemicals, East Sussex, UK. cDNA constructs and transfections | pCEP4APP695 Hygromycin resistant, APPswe Neomycin resistant. pcDNA3.1MycHis (A) ASP-1 and 2 were kindly donated by the GlaxoSmithKline laboratories, Harlow, UK. The clones were engineered in frame with the Histidine tag and the Myc epitope, which in that order are downstream of the coding sequence of the Aspartic proteases (ASP). Cell manipulations, expression protein purification | For the production of stable colonies HEK293 cells were transfected using pFX2 lipids as described by the manufacturers for 2.5 x 104 cells per transfection reaction. Positive transfectants were identified and selected using Geneticin. Colonies were screened by SDS-PAGE and Western Blotting analysis of cell lysates prepared by direct suspension of the cells in SDS, 0.1% (wt/vol) followed by sonication and suspension in gel loading buffer. HEK293 cells were cultured and propagated as described in Frears et al., 1999. Cell lysates for affinity- or immuno-purification were prepared by lysis of the cells with frequent agitation for 2 hrs at 4C in phosphate buffered saline (PBS: 150 mM NaCl 2.7 mM KCl, 10 mM Na2HPO4 and 1.75 mM KH2PO4, at pH7) supplemented with protease inhibitors, 0.5 mM EDTA, 1 muM Leupeptin, 1 muM Pepstatin and 1 mM PMSF, and Triton X-100, 1% (wt/vol) and NP40, 1% (wt/vol). For affinity purification of ASP-2 EDTA and pepstatin were omitted. Prior to purification the lysate was diluted to 0.5% (wt/vol) detergent with the above buffer/inhibitor solution. Ni2+ -- affinity purification was carried out as recommended by the manufacturers using a pre-charged Ni2+ column, 1 ml of settled bed resin for 50 x 106 cells with minor alterations; all binding and washes (with increasing imidazole concentrations, pH 7.4) were carried out in PBS buffer supplemented with 0.2% (wt/vol) NP40 to reduce non-specific binding. Elution was achieved with 300 mM imidazole/0.2% (wt/vol/) detergent. Protein incubations | Freshly purified ASP-2 was dialysed immediately into 50 mM CH3COONa, 20 mM NaCl, 0.2% (wt/vol) NP-40, at either pH 8.5 or 5, at 4C overnight, in the presence of 1 muM Leupeptin and 1 mM PMSF. The protein solution was then incubated at room temperature for two hours before SDS-PAGE, in vitro cleavage reactions, or N-terminal sequencing. For Mass Spectrometry the proteins were first dialysed in the above buffer at pH 8.5, and then the pH was adjusted with dilute acetic acid to ph 5. For N-terminal peptide sequencing the samples were immobilised on a PVDF membrane. Immunoprecipitations and in vitro cleavage assays | Anti-CT15 (2 mug/ml) was used to immunoprecipitate (Stephens and Austen 1996) APPswe from HEK293 cell lysates/extracts prepared as above. The pellet was resuspended in reaction buffer (20 mM CH3COONa, ph 5, 50 mM NaCl, 0.2% NP40, 1 mM PMSF and 1 muM Leupeptin) prior to the addition of ASP-2 (purified and pre-incubated in reaction buffer at designated pH). Reactions were carried out for 2 hours at 37C. Samples were analysed by SDS-PAGE on a 16.5% Tris-Tricine gel as described by the manufacturers followed by western blotting and detection with anti-CT15 IgG (0.5 mug/ml) or 6E10 (4 mug/ml). Mass spectrometry | Purified ASP-2 which had been dialysed at pH 8.5 as described above was pre-incubated in the reaction buffer titrated at the designated pH without detergent prior to absorption to a C-18 ZIP tip (Millipore) equilibrated with 0.1% (vol/vol) trifluorocetic acid, (TFA). Peptides were eluted with 50% (vol/vol) acetonitrile and analysed on an Axima-CFR KRATOS Mass Spectrometer after addition of an equal volume of 10 mg/ml alpha-cyano-4-hydroxycinnamic acid (Sigma) in 50% (vol/vol) acetonitrile. Angiotensin 1, the dimeric form of alpha-cyano-4-hydroxycinnamic acid, and neuropeptide Y were used as external calibrants spotted on an adjacent spot on the chip. Peptide synthesis and IgG production | The C-terminal peptide of ASP-2 (CLRQQHDDFADDISLLK residues 482 --501) was synthesised on a Milligen 9050 synthesizer using Fmoc N-terminal protection and, after deprotection and release from resin was purified by HPLC on a column of Vydac C4 with gradients of acetonitrile in 0.1% TFA. In brief, bovine thyroglobulin was activated with succinimidyl 4-(N-maleimido-methyl) cyclohexane-1-carboxylate, desalted on a Sephadex G-10 column and coupled to the HPLC-purified synthetic peptide utilising its free-cysteine as described in . The coupled peptide was used to immunise rabbits and its immunoreactivity was screened by ELISA using immobilised antigenic peptide. Reactive serum was purified on a peptide-conjugated Sepharose column, aliquoted and stored at -20C. Pure anti-ASP-2 IgG was tested for its specificity towards ASP-2 by western blotting of extracts of recombinant clones as well as mock-transfected cells using purified IgG pre-incubated or not with excess antigenic peptide. Metabolic labelling and detection of recombinant proteins | 80% Confluent flasks of HEK293 APPswe-trasfected cells (~5 x 106 cells) were depleted of Methionine by incubation in Met-free media supplemented with FCS and Glutamax and pyruvate for 1 hr. The same media was then supplemented with [35S]-Met at 50 muCi/ml in 5 mls, and incubated for 3 hrs. Cells were first rinsed with PBS and then lysed and extracted as described earlier. The [35S]-labelled protein was immunoprecipitated as described earlier and reactions with ASP-2 were carried out. These were analysed by 7% SDS-PAGE, soaked in AmplifyTR for 15 minutes and dried before exposure to Kodak X-Omat film. SDS-PAGE and Immunoblotting | Proteins were first denatured by boiling in SDS sample buffer (0.25 M Tris-HCl, pH 8.8, 2.2% (wt/vol) SDS, 10% (v /v) Glycerol, 0.05% (wt/vol) bromophenol blue) with or without reducing agent (1% beta-mercaptoethanol and/or 10 mM DTT) and electrophoresed on SDS-PAGE gels prepared as described by the manufacturers of the acrylamide solution (National Diagnostics) and run using a BioRad Mini-Gel system at 20 mA per gel. The gels were subsequently transferred onto PVDF membrane using a Biorad semi-dry blotter in 25 mM Tris, 192 mM Glycine, and 20% (vol/vol) methanol. Antibody detection was carried out by immunoblotting as described in Stephens and Austen (1996). For silver staining, gels were fixed with methanol, 40% (vol/vol) and acetic acid 10% (vol/vol), for one hour, before they were soaked for 30 min in DTT (0.5 mug/ml). The gels were rinsed twice with water and soaked for one hour in AgNO3, 0.1% (wt/vol) prior to development with a solution of Na2CO3, 3% (wt/vol) and formaldehyde, 0.0185% (vol/vol). PNGase treatment | Cellular extracts or purified ASP-2 were first heat-denatured in the presence of 0.5% (wt/vol) SDS in PBS, pH8 and the rest of the procedure was carried out as described by the manufacturers. Abbreviations : Abeta,beta amyloid peptide; APPswe, Mutations in APP in which KM is replaced by NL in Swedish family with familial Alzheimer's dementia; DMEM, Dulbecco's modified eagle medium; ECL, Enhanced chemiluminescence; HRP, Horseradish peroxidase; NP40, Nonident P 40; Octylglucoside, Octyl-beta-D-glucopyranoside; PMSF, Phenylmethylsulfoxide; PNGase F, N-Glycosidase-F; PVDF, Polyvinylidene Fluoride; SDS-PAGE, Sodium dodecyl sulphate polyacrylamide gel electrophoresis Authors contributions : CS carried out the molecular and cell biology as well as the design, biochemical analyses of this study and drafted the manuscript. CL carried out protein purification and cell manipulations. BMA conceived the study, participated in its design and coordination and also carried out the mass spectrometry analysis. Backmatter: PMID- 12351358 TI - Using standardised patients to measure physicians' practice: validation study using audio recordings AB - Objective | To assess the validity of standardised patients to measure the quality of physicians' practice. Design | Validation study of standardised patients' assessments. Physicians saw unannounced standardised patients presenting with common outpatient conditions. The standardised patients covertly tape recorded their visit and completed a checklist of quality criteria immediately afterwards. Their assessments were compared against independent assessments of the recordings by a trained medical records abstractor. Setting | Four general internal medicine primary care clinics in California. Participants | 144 randomly selected consenting physicians. Main outcome measures | Rates of agreement between the patients' assessments and independent assessment. Results | 40 visits, one per standardised patient, were recorded. The overall rate of agreement between the standardised patients' checklists and the independent assessment of the audio transcripts was 91% (kappa=0.81). Disaggregating the data by medical condition, site, level of physicians' training, and domain (stage of the consultation) gave similar rates of agreement. Sensitivity of the standardised patients' assessments was 95%, and specificity was 85%. The area under the receiver operator characteristic curve was 90%. Conclusions | Standardised patients' assessments seem to be a valid measure of the quality of physicians' care for a variety of common medical conditions in actual outpatient settings. Properly trained standardised patients compare well with independent assessment of recordings of the consultations and may justify their use as a "gold standard" in comparing the quality of care across sites or evaluating data obtained from other sources, such as medical records and clinical vignettes. What is already known on this topic | Standardised patients are valid and reliable reporters of physicians' practice in the medical education setting However, validating standardised patients' measurements of quality of care in actual primary practice is more difficult and has not been done in a prospective study What this study adds | Reports of physicians' quality of care by unannounced standardised patients compare well with independent assessment of the consultations Keywords: Introduction : Standardised patients are increasingly used to assess the quality of medical practice. They offer the advantage of measuring quality while completely controlling for variation in case mix. Although standardised patients have long been used to evaluate medical students and residents, their use in actual clinical settings is relatively new. In medical education standardised patients are introduced into a carefully controlled setting: typically they are directly observed, work in a designated room, and evaluate students from a single school or training programme. Under these controlled conditions standardised patients have been validated to ensure that they perform consistently. Well trained standardised patients effectively and convincingly imitate medical conditions and are remarkably consistent performers, showing high inter-rater agreement and excellent operating characteristics. Validating the use of standardised patients to measure quality in the actual practice setting is, however, challenging and to our knowledge has not been done. Direct observation in the clinic is difficult for a variety of reasons, including cost, a potential Hawthorne effect (physicians performing better under observation), and ethical problems linked to informed consent (J Peabody, sixth European forum on quality improvement in health care, Bologna, March 2001). We did a validation study to determine whether standardised patients perform as well in the clinic as they do in medical education settings. We introduced unannounced standardised patients into clinics and compared their reports of a physician's practice with a covert audio recording of the same visit. Box 1 | Clinical scenarios portrayed by standardised patients Clinical scenarios portrayed by standardised patients Methods : Setting | The study sites were four general internal medicine primary care clinics in California. All staff physicians, teaching physicians, and second or third year residents were eligible. Of the 163 eligible physicians, 144 consented to see standardised patients at some time during the 1999-2000 and 2000-1 academic years. We used the sampling function of Stata to randomly select consenting physicians to whom standardised patients would present with one of eight different clinical cases, two cases each for four common outpatient conditions: chronic obstructive pulmonary disease, diabetes, vascular disease, and depression (box ). Training of standardised patients | We trained 45 professional actors, approximately six per case scenario, as standardised patients. The training protocol involved several steps and is described in detail elsewhere. We prepared detailed scripts for each case scenario and assigned each actor to one of the eight cases. Actors, in groups of three, underwent five training sessions. They were trained how to act as a patient and to observe and recall the physician's actions during the visit. The actors were trained to complete a checklist of 35-45 items that might be performed or discussed by the physician (box ). The actors completed the checklist immediately after the visit by marking each item as done or not done. Checklist items were based on quality measurement criteria derived from national guidelines on specific conditions and were arrived at by expert panel review and a modified Delphi technique (a formal method to determine the extent of consensus). Box 2 | Checklist for evaluating quality of a consultation for chronic obstructive pulmonary disease with a mild exacerbation and history of hypertension Checklist for evaluating quality of a consultation for chronic obstructive pulmonary disease with a mild exacerbation and history of hypertension Duration of dyspnoea Severity of dyspnoea Any similar previous episodes History of asthma, emphysema, or chronic obstructive pulmonary disease Medications taken Presence of fever Presence of cough Quality of cough History of hypertension History of high cholesterol concentrations Exposure to allergens or other irritants at workplace Smoking history Alcohol use Last flu or tetanus shots Last Pneumovax Marital status Job or other social history Blood pressure (both arms) Palpation of jugular vein distension or point of maximal impulse Chest auscultation Lung auscultation Examination of digits for cyanosis or clubbing Examination of lower legs for oedema Peak flow evaluation Pulse oximetry (or arterial blood gas analysis) Rectal/prostate examination Diagnosis of chronic obstructive pulmonary disease, emphysema, or bronchitis Discussion of severity of chronic obstructive pulmonary disease Diagnosis of hypertension or discussion of need to continue taking medication for hypertension Consulted with an attending physician Verified proper use of inhaler(s) Told to drink more fluids Told to call or return if symptoms worsen Told needed oxygen or intravenous fluids or to go to the emergency room (not necessary) Wanted to admit patient to hospital (not necessary) Discussed smoking cessation Counselled on diet Counselled on exercise Recommended, advised, or referred to have colon cancer screening Follow up appointment recommended Audio recording of visits | Of the 45 trained actors we recorded 42, using a digital "pen" recorder concealed on the actor. Three actors left the study before completing their recorded visits. Each actor was recorded once. Two recordings were unusable because of difficulties with the recorders. In 27 of the 40 successfully recorded visits the physicians reported that they had detected the standardised patients. The number of visits was similar across study sites, conditions, and physicians' level of training. To minimise potential variation in performance, we asked the actors to wear the recorder for visits that were not recorded. A single transcriptionist transcribed all recordings. A trained medical records abstractor then scored each transcript using the same quality criteria as in the standardised patients' checklist. A second trained medical record abstractor reviewed each transcript against the recording. Analysis | A total of 1258 quality measurement items were compared. The items were aggregated into four domains corresponding to stages of a visit: history taking, physical examination, diagnosis, and treatment and management. An additional 287 items in the physical examination domain were recorded on the standardised patients' checklists but not compared, because they were only visually observed and could not be verified in the audio recordings. We calculated the percentage of items in agreement between the standardised patients' checklists and the recording transcripts. We calculated kappa values to further quantify the degree of agreement. Percentage agreement and kappa values were disaggregated by condition, site, physicians' level of training, and domain. A calibration curve was constructed to assess variation across actors. Sensitivity and specificity were calculated for each visit and for all visits combined, taking the audio recording as "truth" in the calculation. A receiver operator characteristic curve was then constructed by plotting sensitivity and specificity for each visit and choosing the most conservative spline that circumscribed all data points. Results : The overall rate of agreement between corresponding items on the standardised patients' checklists and the recording transcripts was 91% (kappa=0.81) (table ). Agreement rates for the four conditions ranged from 88% for depression (kappa=0.75) to 95% for diabetes (kappa=0.89). Agreement rates for individual sites ranged from 90% (kappa=0.78) to 93% (kappa=0.81). Agreement rates and kappa values also varied little by physicians' training level. Agreement rates were similar for history taking (91%; kappa=0.81), diagnosis (89%; kappa=0.69), and treatment and management (93%; kappa=0.85). Table 1 | Agreement (%) between standardised patients' assessments and audio recordings of consultations Figure shows the variation among standardised patients. This calibration curve plots the percentage of checklist items done by the physician as noted by the standardised patient against the corresponding percentage indicated by the audio recording of that visit. Points cluster closely along the plotted regression line, which has an intercept of 0.4% and a slope of 1.03. (Perfect calibration would yield a line with intercept of 0% and a slope of 1.00.) Figure 1 | Percentage of items on checklist done by physician, as rated by standardised patients and as indicated by audio recordings of visits Percentage of items on checklist done by physician, as rated by standardised patients and as indicated by audio recordings of visits Sensitivity of standardised patients' assessments, compared against the audio recording transcripts, was 95%, and specificity was 85% (table ). Table also shows that about two thirds of the items where the two methods disagreed were reported as done by the standardised patient but determined to be not done according to the transcript. Table 2 | Sensitivity and specificity of standardised patients' assessments, with respect to audio recordings of consultations Figure shows the operating characteristics of standardised patients. Each data point represents the sensitivity and specificity values for one recorded visit. The area under the resulting receiver operator characteristic curve is 90%. Figure 2 | Receiver operator characteristic curve for standardised patients with respect to audio recordings Receiver operator characteristic curve for standardised patients with respect to audio recordings Discussion : Although patients and physicians alike desire improved quality, accurate measurement of quality remains problematic. Comparisons of quality across physicians and sites are hampered by imperfect adjustments for variation in case mix. Also, the underlying data on quality are of uncertain validity, because of logistical and ethical difficulties in directly observing physicians while they care for patients. Measurement of quality has therefore relied largely on medical records, which at best are incomplete and at worst falsified. Standardised patients, despite being costly to train and implement, overcome the first problem by providing presentations that are perfectly adjusted for case mix. They may also be able to overcome the second problem, if their validity in the outpatient setting can be shown. Many studies have turned to standardised patients when highly accurate measures of quality are needed. Standardised patients are particularly well suited for cross system comparisons, such as comparing general practice with walk-in care or for assessing quality for potentially sensitive conditions such as sexually transmitted infections and HIV. Standardised patients are already considered the criterion standard for evaluating competence in specialties and have become part of national certification examinations in the United States. And while the accuracy of standardised patients is assumed to be high, it has not been prospectively evaluated. We found that standardised patients were well calibrated to actual recordings of clinical encounters. No apparent systematic bias was seen by medical condition, site, level of physicians' training, or domain of the encounter. Intermethod reliability was uniformly high. Standardised patients showed excellent sensitivity, specificity, and operating characteristics. We observed higher sensitivity than specificity ---that is, the false positive rate of standardised patients' assessments exceeded the false negative rate. Given the inherent trade off between sensitivity and specificity, we attribute this finding to our explicit instructions to, "when in doubt, give the provider the benefit of the doubt." Alternatively, although the technical quality of the recordings was generally high, some false positives could be attributed to unclear speech (if doctor and patient spoke at the same time). The design of our study helped mitigate technical issues that might have degraded the audio recording data. Although the physicians' informed consent meant that some standardised patients were detected, we received no reports by physicians or standardised patients that the concealed recorder itself was detected. The actors were coached in precise placement of the recorder, particularly as they undressed during the visit. The accuracy of the transcript was ensured by the use of an experienced transcriptionist as well as a trained abstractor who independently reviewed each transcript against the recording. Limitations of the study | We assessed only verbal communication. In future studies doctors may consent to unannounced visits that are video recorded. We did not measure within-actor variation. In the medical education setting such variation is managed by using standardised physicians to calibrate the standardised patients. Such results show that performances by a standardised patient are consistent from visit to visit. We believe from anecdotal evidence that this was the case in our study as well but have not measured it objectively. Another issue that merits further study is how accurately standardised patients can measure quality through a single encounter ---or even a short series of visits. Some studies suggest that a "first visit bias" may skew assessment of quality, since chronic diseases typically necessitate several visits and ongoing follow up. We deliberately used clinical scenarios that required immediate interventions, and we are separately analysing those items (particularly preventive care) that could be postponed to a future visit. Future research might assess how well standardised patients' measurements of quality for a few selected cases can comprehensively assess an individual physician's overall competence. We used explicit checklists of quality criteria to measure physicians' performance. Other studies involving standardised patients have used different analytic approaches, such as global rating scales. While checklists and rating scales have different emphases ---for example, technical versus interpersonal skills ---some researchers argue that both these types are valid and reliable. We did not use rating scales because of our concerns over the potentially more subjective nature and lower inter-rater reliability of global ratings. Setting standards | Using standardised patients to measure quality raises the question of how to set standards for what is considered adequate clinical competence. Panels of expert judges have been shown to be reliable for setting standards. The expert judges seem to use a compensatory model, where very good performance on some cases compensates for performing poorly on other cases. Analysis of the receiver operator characteristics of standardised patients has also been used to set standards in performance assessments of students at examination level. Receiver operator characteristic analysis shows that standardised patients can differentiate between disparate levels of competence ---for example, accurately discriminating between second and fourth year medical students. Conclusions | Standardised patients' assessments seem to be a valid measure of the quality of physicians' care for a variety of common medical conditions in actual outpatient settings. Concealed audio recorders were effective for validating standardised patients' assessments. Properly trained standardised patients should be considered for comparative measurements of quality of care across sites when validity is essential. As the criterion standard, standardised patients can be used to evaluate the validity of data obtained from other sources, such as medical records and physicians' (self) reports. We believe standardised patients are particularly useful to validate innovative methods of quality measurement, such as computerised clinical vignettes. Vignettes, like standardised patients, inherently control for case mix variation; and, once validated against actual clinical practice, vignettes can be more widely used because they are cheaper and do not require subterfuge. Ultimately, accurate and affordable measurements of clinical practice underlie any effort to provide better quality for patients. Backmatter: PMID- 12351359 TI - Spontaneous talking time at start of consultation in outpatient clinic: cohort study AB - Keywords: null: null Participants, methods, and results : We investigated a sequential cohort of patients from the outpatient clinic of the department of internal medicine at the university hospital in Basle. The study protocol was approved by the university's ethics committee. Inclusion criteria were sufficient knowledge of the German language, first contact with the outpatient clinic, and mental competence. We informed doctors about the purpose of the study and told patients that we were interested in their opinion concerning the service provided. We asked doctors to activate a stop watch surreptitiously at the start of the communication and press it again when patients indicated that they wanted the doctor to take the lead (for example, by saying: "What do you think, doctor?"). Patients did not know that a timer was being used. Doctors were trained for one hour in basic elements of active listening, such as waiting, use of facilitators like "hmm-hmm," nodding, or echoing. They were told not to ask questions during the initial phase of the consultation. To comply with their consultation schedule they were advised to interrupt if a patient talked for more than five minutes. Within three months 406 out of a total of 1137 patients fulfilled the inclusion criteria; 33 were later judged as not correctly classified. Of the remaining 373, 20 patients did not give informed consent; for nine patients doctors did not register talking time; and data on talking time were lost for nine patients. We analysed spontaneous talking time in 335 patients who had been seen by 14 doctors. Of the 330 patients who provided sociodemographic data, 176 (53%) were female, mean age was 42.9 years (SD 18.2 (95% confidence interval 17 to 84) years). The sociodemographic characteristics were typical of patients seen at this hospital. The 11 male and three female doctors had worked a mean of 58 (26) months in the clinical field, with a mean of 38 (19) months spent in internal medicine. Figure | Spontaneous talking time of 331 patients at start of consultation in outpatient clinic Spontaneous talking time of 331 patients at start of consultation in outpatient clinic Mean spontaneous talking time was 92 seconds (SD 105 seconds; median 59 seconds; figure), and 78% (258) of patients had finished their initial statement in two minutes. Seven patients talked for longer than five minutes. In all cases doctors felt that the patients were giving important information and should not be interrupted. No other sociodemographic variable (education, income, civil status, type of employment, and sex) had a significant influence on spontaneous talking time except for age (rs=0.41; P<0.001; 17-29 years: 77 (105) seconds; 30-49 years: 92 (93) seconds; 50-87 years: 108 (114) seconds). Comment : Doctors do not risk being swamped by their patients' complaints if they listen until a patient indicates that his or her list of complaints is complete. Even in a busy practice driven by time constraints and financial pressure, two minutes of listening should be possible and will be sufficient for nearly 80% of patients. We gathered data in a tertiary referral centre that is characterised by a selection of difficult patients with complex histories. Patients in less selected groups might need even less time to complete their initial statement. Backmatter: PMID- 12351360 TI - Women's attitudes to the sex of medical students in a gynaecology clinic: cross sectional survey AB - Keywords: In Tomorrow's Doctors the General Medical Council recommended that medical schools construct a list of procedures in which students should show competence by the time they qualify. There is general acceptance that such core skills include passing a speculum, taking a smear, and performing a competent pelvic examination. Anecdotal evidence from medical students, particularly male students, is that experience in this area is difficult to obtain. This is not a problem confined to the United Kingdom. In response to a similar perception among their male students, staff at the University of California studied patients' views on the involvement of medical students in the women's visits in an outpatient gynaecological and obstetric setting. They found that 81% of patients accepted the involvement of students during a gynaecological visit, with no preference for a particular sex. However, the study did not directly address the issue of intimate examinations. We surveyed women attending a gynaecology clinic in an inner London teaching hospital to examine women's experience of and attitudes to the sex of medical students. Table : Numbers (percentages) of women responding to the question "Would you allow a student to do an intimate examination?" Methods and results : We surveyed women attending a gynaecology clinic in the academic year 1999-2000. Women were approached only when a student was working with the doctor they had seen. Questionnaires were given out by nursing staff after the consultation. Two hundred questionnaires were distributed and 181 were returned. The age range of respondents was 17-79 years (mean 40 (SD 13) years). Just under a quarter (44) of the women were attending a gynaecology clinic for the first time. Ten women had never been sexually active, and 64 had no children. In the sample 166 women had interacted with students. Six women who saw more than one student at the same consultation were omitted from the analysis. Ninety seven women had interaction with male students and 63 with female students. Students had low levels of interaction with patients. Just under half (73) of the women reported that students asked questions, 25 that students did general examinations, and 31 that students did intimate examinations. There was a trend towards female students being more actively involved in examination: in 12 of the 63 visits (19%) involving female students the student did a general examination, compared with 13 of the 97 visits (13%) involving a male student, and the corresponding figures for intimate examinations were 14 (22%) for female students and 15 (15%) for male students. The women were asked to consider the potential involvement of a student during a consultation. Their attitudes differed according to the sex of the student, with a preference for female students in all types of interaction. More women said they would allow a female student than a male student to observe their genital area (140 v 93 of the 181 women; chi2=45, P<0.001), and more said they would allow a female student than a male student to do an intimate examination (114 v 72; chi2=63, P<0.001). From a practical perspective we were interested in ways of identifying women who would agree to intimate examination by students of either sex. This would reduce the difficulty of the encounter and embarrassment for patient and student. We chose parity and age as easily identifiable markers, both suggested during the questionnaire design process. Older women were more likely than younger women to agree to intimate examination by students of either sex, as were women who had had children, compared with women who had not (table). Although older, parous women were more accepting of the involvement of students, the difference according to sex of the student was maintained. Comment : Our findings support the claim of male medical students that it is more difficult for them than for female students to get experience of gynaecological examination. Some women attending this outpatient clinic were agreeable to examination by students of either sex. It may be necessary to target such women for involvement with student education. It may be appropriate to use different teaching methods and settings for different aspects of teaching gynaecology: the teaching of consultation skills could be confined to the outpatient clinic, while pelvic models and volunteers could be used to teach clinical skills. Backmatter: PMID- 12351361 TI - UK senior doctors' career destinations, job satisfaction, and future intentions: questionnaire survey AB - Keywords: null: null Participants, methods, and results : We mailed questionnaires in 1998, asking about employment history and current job. In all, 77% (1717/2217) replied (72% (1223) of men and 80% (494) of women). Of respondents, 97% (1673) were in medical employment in the United Kingdom or abroad, and 85% (1460) worked in the UK NHS. The table shows the distribution of their specialties. In NHS general practice, 98% (488) of men and 85% (183) of women were principals; of these, 39% (72) of women and 5% (25) of men worked part time. In NHS hospital practice, 97% (469/485) of men and 76% (114/150) of women whose main paid post was in the NHS (not in a university) were consultants; of these, 26% (30) of women and 11% (50) of men worked part time. Table | Profile of 1673 respondents (graduates from 1974) in medical employment in 1998: specialty and job satisfaction. Values are numbers (percentages of all medical posts) unless stated otherwise We asked respondents to score five statements about job satisfaction, from "strongly agree" (score 1) to "strongly disagree" (5). The statements were "I find enjoyment in my current post"; "I am doing interesting and challenging work"; "I feel dissatisfied in my current post"; "most days I am enthusiastic about my work"; "I am often bored with my work"). We calculated a job satisfaction score by totalling the scores for all five statements: 20 or more represented a positive response, on average, to all statements, and we suggest that this shows a high level of satisfaction. Respondents rated their job satisfaction reasonably highly (table), with some significant but small differences between occupational groups. Hospital doctors were more satisfied with their jobs than general practitioners; women general practitioners were more satisfied than men (group median scores 19.7 and 18.8 respectively); and part time general practitioners were more satisfied than those working full time (19.8 and 18.9) (all P<0.001). This last finding was mainly explained by the comparatively lower job satisfaction of male general practitioners working full time. Respondents practising medicine in the United Kingdom were asked whether they intended to continue doing so for at least another five years. Most respondents definitely intended to do so; more NHS hospital doctors than general practitioners were definite (85.8% (580/676) and 77.9% (553/710); P<0.001). Reasons cited for considering leaving UK medicine included dissatisfaction or disillusionment with their job or with the NHS; ill health; stress or pressure associated with the job; and a desire for change, travel, other interests, or to work in a developing country. Comment : The results provide benchmark findings on the career destinations of doctors who qualified in the 1970s, with which career profiles of more recent generations can be compared. Only a quarter of the 1974 graduates were women. Women from the 1974 cohort were less likely than men to hold consultant or principal posts, and few women were surgeons. Many more women than men worked part time. In the future NHS, many more of the senior posts will be filled by women because more women than men now enter medical school, reversing the former situation in which men substantially outnumbered women. As the government recognises, doctors' working lives need to be balanced with their expectations about raising a family and other interests. Replies about short term intentions portray a workforce committed to working in the NHS, with stable careers over the next few years. Levels of job satisfaction were generally high. Nevertheless, many respondents commented that they were working close to the limits of what they regarded as a reasonable commitment of their time, and many do not intend to work to the age of 65. Backmatter: PMID- 12351362 TI - Patients' perceptions of entitlement to time in general practice consultations for depression: qualitative study AB - Objective: | To investigate patients' perceptions of entitlement to time in general practice consultations for depression. Design: | Qualitative study based on interviews with patients with mild to moderate depression. Setting: | Eight general practices in the West Midlands and the regional membership of the Depression Alliance. Participants: | 32 general practice patients and 30 respondents from the Depression Alliance. Results: | An intense sense of time pressure and a self imposed rationing of time in consultations were key concerns among the interviewees. Anxiety about time affected patients' freedom to talk about their problems. Patients took upon themselves part of the responsibility for managing time in the consultation to relieve the burden they perceived their doctors to be working under. Respondents' accounts often showed a mismatch between their own sense of time entitlement and the doctors' capacity to respond flexibly and constructively in offering extended consultation time when this was necessary. Patients valued time to talk and would often have liked more, but they did not necessarily associate length of consultation with quality. The impression doctors gave in handling time in consultations sent strong messages about legitimising the patients' illness and their decision to consult. Conclusions: | Patients' self imposed restraint in taking up doctors' time has important consequences for the recognition and treatment of depression. Doctors need to have a greater awareness of patients' anxieties about time and should move to allay such anxieties by pre-emptive reassurance and reinforcing patients' sense of entitlement to time. Far from acting as "consumers," patients voluntarily assume responsibility for conserving scarce resources in a health service that they regard as a collective rather than a personal resource. What is already known on this topic | A widespread concern is that pressure of work is reducing the length of general practice consultations and that doctors can't deal adequately with patients' problems in the time available Little is known about patients' experience of time in consultations What this study adds | Patients with depression feel under such acute pressure of time that they are often inhibited from fully disclosing their problems, preventing them making best use of the consultation There is often a disparity between patients' sense of time shortage and the amount of time their doctors are willing and able to provide Doctors should be more aware of patients' anxieties about time and allay these anxieties by providing pre-emptive reassurance as a means of reinforcing patients' sense of entitlement to consultation time Keywords: Introduction : Shortage of time in general practice consultations is widely acknowledged to constrain the quality of care of patients. It has become emblematic of the scarcity of basic resources in a national health system perceived to be strapped for cash and struggling to cope. Technological innovations expand the boundaries of clinical need, and bureaucratic complexity increases the professional workload. With an average slot of 5-8 minutes per patient, doctors protest that there is not enough time for the increasing number of tasks involved in routine consultations. Shortage of time is considered a major obstacle to the realisation of a more patient centred medical practice that actively involves patients in treatment decisions. Discussion of time has focused on the professional and organisational perspective, rather than patients' experiences and views. References to patients' dissatisfaction with consultation time occur in accounts or analysis of satisfaction with health care. However, little is known about patients' perceptions of the adequacy or quality of time spent with their general practitioners. We report on part of the findings of a study of patients' and professionals' understandings of depression and its treatment, in which time emerged as an important topic. Methods : We used a qualitative interview to compare the views of patients and general practitioners on the nature of depression and the effectiveness of its treatments. We recruited 19 general practitioners and two counsellors from eight practices in the West Midlands varying in size, type of location, and catchment. A convenience sample of 32 patients with recently diagnosed mild to moderate depression was recruited by general practitioners from their lists. All 32 patients had an initial interview, and 30 of the patients took part in a follow up interview six months later. General practitioners varied in their strategies for recruiting patients: some did an audit of practice records; others identified suitable patients during consultation. All patients received a written personal invitation from their general practitioner. Judgments about mild to moderate depression were left to individual doctors. We asked the general practitioners to recruit patients with "typical" depression of the kind they would routinely deal with in general practice, and none had any problem in recognising such a category. However, the doctors had a fairly elastic notion of "recent" diagnosis. Although we initially aimed to recruit patients whose depression was diagnosed in the preceding six months, some patients turned out to have had depression for rather longer, while five remained uncommitted to the diagnosis (table). Table | Duration of depression in general practice patients and Depression Alliance respondents We wanted to compare the views of relatively new patients with those of people who had more experience of dealing with depression and had accessed information and support outside the formal health service. An additional group of 30 respondents was recruited from the regional membership of the Depression Alliance. These respondents were interviewed once. The two groups differed in some ways, but for the purpose of analysing perceptions of time they constituted a uniform set. This paper deals specifically with all relevant material from the three sets of interviews. The response rate in both groups was predictably low, at around 50%. We assumed that patients who were more severely ill, had opted out of treatment, or were of lower socioeconomic status were less likely to take part. As this was a qualitative investigation respondents do not constitute a representative sample, but their accounts illustrate the range of views and experiences of patients with depression. Respondents varied in age, occupation, socioeconomic group, and marital status. The two groups were similar in terms of age, sex, marital status, and occupation. Women outnumbered men by nearly three to one (46 to 16). A more detailed account of the study design and recruitment is given elsewhere. All but three of the interviews were taped and fully transcribed. We used the NUD*IST software package to analyse the content. Untaped interviews were typed up from extensive written notes and entered into the NUD*IST database. The research was approved by the local research ethics committee. All patients gave signed consent at the start of interviews. We interviewed the respondents in their homes. The interviews were wide ranging discussions in which we responded to the patients' own concerns and emphasis and focused on picking up on and exploring aspects of thought and experience of particular salience to the patients. We developed an initial coding frame that we then jointly extended through an iterative process of comparison and evaluation. Results : Time emerged as an important theme in the interviews. The general practice patients and the Depression Alliance respondents did not differ in the nature of the issues concerning time, although the topic seemed to have greater salience for patients whose onset of depression was more recent. A concern with time featured in 23 of the 32 first interviews with general practice patients, compared with nine of the 30 follow up interviews and 13 of the 30 interviews with Depression Alliance respondents. Sense of entitlement to time with the general practitioner | The respondents commonly expressed concern about "wasting" their doctor's time by presenting with inappropriate or trivial complaints. Lacking a diagnostic or interpretive framework for their experience, many patients were uncertain about the appropriateness of approaching their doctors for help and the legitimacy of their "medical" need. Respondents regarded general practitioners as very busy people whose time was scarce and valuable. They were concerned not to take up more than their fair share and, particularly, not to disadvantage other patients whose needs they perceived as being more urgent (box ). Box 1 | Entitlement to time with the general practitioner Entitlement to time with the general practitioner Anxiety about inappropriate consulting and wasting doctor's time There are lots of posters in the GP's surgery about depression . . . and it has got like, I think, . . . 5 or 10 points, and it says, if you suffer from these you may have depression. See your GP now. I kind of looked at them and thought, half of them possible. Probably the others, but not every day, not something that happens all the time. You know, I do go out. I don't lock myself in my room. I do sleep some days, but not others. Yes, so it was like: I can't really go there [to general practitioner] with this, so I just felt I couldn't really go there because it was almost like wasting his time (patient 112) . . . and I know that the last time [I saw the general practitioner], by the time I had got in I had been sitting there for an hour and a half, and I felt really terrible at going in there. I thought, "Oh God, I am wasting his time . . . he has probably been rushed out and had to speak to somebody who is incredibly ill (patient 310) Inadequate time in normal 5-10 minute appointment You have a 5 minute appointment. You can't just say to them in 5 minutes, "Look, I am so desperate. Help me." They just won't take it, will they? (patient 415) Respondents often referred to a time entitlement of no more than 5 or 10 minutes with their doctor. Patients had internalised an unspoken norm about how long a consultation should last, irrespective of the complexity of the problem. At the same time, they recognised depression as requiring more time than a physical ailment. With a physical problem it was possible to plunge straight into describing the symptoms without introduction or background, whereas explaining mental health problems was often a more difficult and lengthy process. Time management | Respondents reported taking on themselves the pressure of time they felt their doctor to be under and felt some responsibility for helping the doctor to manage time. Respondents were concerned to play their part in keeping the appointment system running to time, a concern that was reinforced when other patients were in the waiting room. Because they felt under pressure of time, many patients thought that they did not make best use of the consultation. They felt unable to express their concerns or respond constructively to what the doctor was saying, and thus their presentation of symptoms and concerns was often heavily edited (box ). Box 2 | Time management Time management Taking responsibility for managing the doctor's time I think that GPs are . . . far too busy, you know, they have such short consultations . . . that also worries me, you see, that I am running over time in a consultation. I mean, that should be worrying the doctor, not the patient, but, you know, "I must go. I mustn't keep him, he will be late, and his next patient will be running late," and it is . . . silly things that I think about and worry about (patient 413) Rationing time and limiting disclosure of information I didn't feel that I could go in and let him have the whole lot. I mean, he's got other patients waiting, and I feel guilty that I've taken enough time up (patient 301) I think he knew that I had got problems, but I was not sort of being very open, you know. I just felt that I couldn't talk . . . or I hadn't got time, or wasn't ill enough. There were lots of people far worse than what I was sitting in the waiting room. And in sort of 5 minutes you couldn't really discuss half of what was going off in your head (Depression Alliance patient 113) . . . you feel when you go to the surgery that there is a room full of other people and they [doctors] haven't got long . . . The GP is excellent. If she has got the time she will give you the time, but sometimes you sense that she is already running an hour late, or whatever, because I have waited, and you daren't, you feel that you dare not say too much because . . . if she is worth her salt she is going to say, ask you the question, the right question that is going to start you opening up, and you are also aware that she is probably an hour behind (patient 310) The responsibility that many patients assumed for the management of their doctor's time inhibited them from extending the consultation even when they knew that more time was available or being offered. Sometimes they recognised a question to be too probing to answer fully in the time they felt to be available. Other respondents described how they would select only the most pressing issues to discuss with their doctor or would reveal concerns gradually over a series of visits, because there was not enough time in a single consultation. The patients saw their entitlement to time as running across consultations, and they assessed this entitlement in terms of the demand resulting from the totality of their health problems, not just depression. There was a limit to the number of issues that could be raised within a certain period ---another factor in patients' rationing of their consultation time. Respondents were not critical of the shortage of time. They did not hold their doctors responsible for what they perceived to be a consequence of the scarcity of resources in the NHS. They sympathised with the pressures they perceived their doctors to be under and tried to help by exercising restraint in the demands they made on the system. In contrast with the many patients who referred to their restraint in taking time with their doctors, only one respondent described going into a consultation with the intention of securing more than her normal time entitlement. Even then, this was not for herself but on behalf of a relative. She reported this encounter in fighting terms, apparently feeling that she had to steel herself to flout conventions to get what she wanted. Her surprise when the doctor responded positively to her stance indicates the strength of the time rule and the pressure patients felt to keep within the bounds of "normal" time. Extended time | Despite the sense of time pressure expressed by many respondents, it was not uncommon for patients to have a longer consultation than the standard 5-10 minutes. Giving more than the standard time was seen as a mark of the doctor's quality and care. Extended time was most likely to occur in first consultations. In some cases, when presented with extreme distress, doctors responded immediately and somehow found the means to provide an extraordinary amount of time (box ). Doctors' ability to respond to urgent need, even during a normal busy surgery, was deeply appreciated by patients. The effects of such a response could be experienced as momentous and far reaching. Several respondents described them as literally "life saving." Box 3 | Extended time Extended time First consultations often take more time The first time I saw him was about 10 to 15 minutes, that's all, and the second time was just a standard medical consultation sort of thing, you know (patient 414) Sometimes a recognised need is met with extraordinary time I just booked an appointment, and I was so nervous about going and telling him that I think I needed something to help me, as I was under such a black cloud, but he was great . . . I was in there for a good 30 minutes, which is really good for a doctor. And he was so kind, he was concerned (patient 411) . . . even now, I often think of all these poor patients waiting to see him . . . You're only supposed to have 10 minutes ---an hour and a half later, I was still there. He took his phone off the hook, and he was brilliant (Depression Alliance patient 111) Quality time | Having the time and opportunity to talk about their problems was widely appreciated by patients and experienced as intrinsically therapeutic. However, longer consultations were not necessarily felt to be better. How time was used was crucial, and even short consultations could be experienced as effective (box ). The converse was also true. It was not so much the amount of time that mattered as the subjective perception of the quality of time spent with the doctor. The doctor's manner could be as important as the amount of time available. A too brisk consultation could send a strong message about how the doctor viewed the legitimacy and substance of the patient's problems. Doctors varied in their ability to convey interest, empathy, the feeling of not being under pressure, and the feeling of taking problems seriously. Box 4 | Quality time Quality time Standard consultations could be effective . . . but, due respect to him . . . he got like I said to you before, he got more out of me in 10 minutes than I have said to anybody for four years (patient 414) Importance of doctor's manner in mediating patients' subjective experience of time I know they've not got a lot of time, anyway, for each patient or whatever, but at the same time, I think . . . with some of them, the [way] some of them have like reacted . . . I felt like I was just a burden (Depression Alliance patient 104) So he never once gave me the impression that he'd run over time, you know, "I've got other patients to see." It was as though his body language and his conversation sort of really . . . I supposed reinforced his initial statement that I am a GP who considers depression to be an illness and not a figment of somebody's imagination (patient 401) It was just, you know, he would see me for 10 minutes. "How are you?" "I'm not too bad." "Right, carry on taking the tablets." You know, that was it, and I'd say, "Well, why am I like this?" "Well, you're depressed." And you would think, "Yes, OK, but I don't feel depressed as such." "Well, you are." It was really the communication side of things that wasn't there (Depression Alliance patient 202) Time for talking | Perceptions of time entitlement influenced how patients evaluated the service they got. They made allowance for what could be realistically achieved in the time available and adjusted their expectations accordingly. Time for talking could be delegated through referral to counselling agencies and specialist services. However, a sense of time pressure was so deeply ingrained that some patients carried it over into the longer sessions sometimes explicitly scheduled for extended talking ---they felt that time that was bounded was intrinsically constraining (box ). Patients felt they had little or no control over time and were the passive recipients of a resource someone else had allocated and which had to be shared among all patients. Furthermore, such referrals constitute an appropriation of patient time. Box 5 | Talking time Talking time Delegating time to talk I think, to be able to talk to someone who really understands, you know. I mean, your GP hasn't got time to sit down for hours talking to you. I mean, having somebody like a therapist or one of the qualified nursing staff that work in those situations (Depression Alliance patient 106) Anxiety about time could constrain patients' ability to use it effectively But again at those [counselling] consultations I knew that I had three quarters of an hour and I was desperately worried that I was going to run over . . . I couldn't see the clock, but there was a clock in the room, but it was in a position that I couldn't see it and I was desperate. I didn't want to appear rude, but I desperately wanted to make sure that I was within three quarters of an hour [laughing] (patient 413) Interviewer: Did you find [psychiatrist outpatient appointments] helpful? Respondent: I didn't, no. I mean he was . . . you know, we sort of talked, but I was always aware of the time . . . The time was always ticking away. I suppose when I got comfortable, you know, it was almost time for the next appointment in another couple of months, whatever, so I never felt that I addressed the problems, or I just really never felt ---I can't remember the exact questions, but I just never felt that they got anywhere (Depression Alliance patient 112) Some patients felt that extended consultations were a waste of time I think that most GPs I've seen . . . it would be nice to have more time if you felt that what was being discussed in that period was useful. I have seen psychiatrists in the past, and the last psychiatrist I saw just put me off any sort of psychiatric help ever again . . . it was a complete and utter waste of time (Depression Alliance patient 116) I did feel the sort of things she was coming out with were the sort of things you would say to yourself as a matter of course . . . I didn't feel I was getting from her anything additional to what I already knew in my own mind . . . I just thought that was a complete and utter waste of time, you know (patient 412) Although patients generally acquiesced in the time allocated to them, they were active in appraising the quality and outcome of the service. Across a range of consultations with different professionals, there was no clear relation between the length and perceived quality of time. It was not necessary for a consultation to be long for it to be effective. Indeed, a number of patients felt that their meetings with professionals had been a waste of time. Discussion : Patients' anxiety about time, awareness of pressure of time, and feelings of low entitlement to consultation time were among our most striking findings. Patients' reluctance to consult and fear of wasting professional time reflect a norm that professional help should be sought only in cases of extreme and genuine need. Whether and when to seek medical advice are judgments that are particularly difficult for people experiencing psychological distress. Our respondents' comments expose a tension between patients and the medical system in setting boundaries of entitlement and access to care. The comments reinforce other findings that patients' perceptions of low entitlement to consultation time and anxiety about wasting professional time with "trivia" discourage them from making appointments and inhibit their disclosure of problems. The patients took upon themselves the time pressure they felt general practitioners to be working under and also the responsibility to ration time, by not overrunning the 5-10 minutes they thought was the standard allocation. Crowded waiting rooms and long delays to see a doctor intensified their already fragile sense of entitlement. As other studies have also shown, other patients' needs are routinely judged to be more pressing than one's own. It is patients rather than doctors who take the initiative in rationing time. In effect, our patients often failed to capitalise on the resource that was on offer, feeling unable to use the time their doctor was willing to extend to them. In consequence, they left the surgery with questions unasked and issues unexplored. This must have a negative impact on general practitioners' capacity to provide effective support for patients presenting with depressive disorders. It also inhibits the development of more patient centred and concordant consultations. Paradoxically, it may be patients' concern to keep within their perceived time entitlement that causes time to be "wasted" in ineffective and superficial consultations from which they derive little benefit. Despite their anxieties about time, patients quite commonly reported receiving more than they expected, especially during a first consultation for depression. Where patients presented with severe distress, some doctors managed to find up to an hour or more to extend the consultation. Patients were deeply appreciative of this kind of response. These instances show what can be achieved even within the current constraints of general practice and the scope for flexible and creative use of time that still obtains (though doubtless reinforcing the stress of other patients in the waiting room). Although much of the debate about time scarcity in general practices focuses on the "average" length of 5-8 minutes, research also shows a wide range of consultation time. Perhaps the preoccupation with average length of consultation is misplaced. Doctors' ability to allocate time flexibly and according to individual need is what is really critical. The quality of patients' relationships with their doctor is fashioned over a series of consultations, and it is important to take account of the extension of time across these consultations as well as what happens in discrete episodes. Some consultations will require more time than others. Shorter consultations may still be effective where they build on support and shared understandings developed through past encounters. Conclusions | Perceived quality of time, rather than just quantity, is critical to patients' experience of the consultation. Doctors, by conveying the impression of having time for the patient, can express a powerful message about their interest and concern. In this they legitimate patients' problems, validate their decision to consult, and confirm their need for treatment and support. Positive reinforcement of this kind is essential for patients' sense of entitlement to be increased to the point where they are can contribute more effectively ---and more concordantly ---to the consultation. Backmatter: PMID- 12351363 TI - Patient centredness in the MRCGP video examination: analysis of large cohort AB - Keywords: The examination for membership of the Royal College of General Practitioners (MRCGP) consists of four modules. The consulting skills module, introduced in 1996, is normally taken by submitting a videotape of seven consultations, selected by the candidate to show "competency" in each of 15 "performance criteria." These criteria are explained in the examination literature and on the college's website. The module is based on a competency model of assessment, which defines outcomes, in a hierarchical framework. The performance criteria (see table) are grouped into five areas: discovering the reason for the patient's attendance, exploring the problem(s), tackling the problem(s), explaining the problem(s), and making effective use of the consultation. Table : Success of 2094 candidates in meeting each performance criterion in none, one, two, three, four, or all of the first five consultations submitted. Values are percentages (numbers) of candidates, unless stated otherwise Participants, methods, and results : Doctors may submit videotapes for examination in May or November. This report is based on submissions in May and November 1999 and May 2000, which together comprised 2094 candidates. Each candidate was assessed by seven examiners, each rating a different consultation on the tape. We have described the assessment method elsewhere. The examination is primarily a "competency hurdle" ---that is, candidates have to achieve competency across a range of criteria. Three criteria (numbered in the table as 4, 10, and 11) were found in pilot studies to be rarely achieved. However, as we considered these to be markers of "good practice" and, in particular, of patient centredness, we designated them "merit" criteria ---for awarding the merit grade to candidates who had already passed on the other 12 criteria. The table shows the extent to which each criterion was met in the first five consultations assessed because at that time pass-fail judgments were made initially on the first five consultations; the last two were considered only if the candidate had not clearly passed on the first five. (From 2002 all seven consultations are considered.) The column headed "mode" indicates the commonest frequency, whereas the "mean" column allows comparison between the criteria. "Patient centredness" has been well defined and characterised, comprising five dimensions: a biopsychosocial perspective, the "patient-as-person," sharing power and responsibility, the therapeutic alliance, and the "doctor-as-person." We explicitly intended that our performance criteria should reflect this and have identified aspects of patient centredness in criteria 2, 3, 4, 10, 11, and 13. Competency in exploring the patient's own beliefs about the illness (criterion 4), using those beliefs in explaining the illness (10), and checking the patient's understanding after the explanation (11) were not seen in 14%, 31%, and 45% of doctors respectively. The related competency of involving patients in decision making (criterion 13) was not seen in 14% of doctors, and only 36% (762) managed to show it in three or more of the first five consultations (the stated target for a pass). These four criteria all had modes of 2 or less (table). In contrast, 69% (1442) candidates were able to meet the remaining performance criteria in at least three of the first five consultations (modes of 3, 4, or 5 in the table). Comment : On the basis of their "best" five recorded consultations, doctors nearing completion of a three year postgraduate training in general practice showed only limited ability to achieve patient centred outcomes. The ability to elicit patients' ideas, concerns, and expectations is fundamental to good consulting, but our results suggest that few doctors regularly use this ability, even in a highly selected set of consultations. Likewise, the checking of understanding, and the involving of patients in decision making ---both likely to improve concordance ---are rarely demonstrated. Patient centredness may not be appropriate for every consultation, but these abilities are held by the Royal College of General Practitioners to be necessary markers of good general practice. Backmatter: PMID- 12351364 TI - Competency based medical training: review AB - Keywords: The competency approach has become prominent at most stages of undergraduate and postgraduate medical training in many countries. In the United Kingdom, for example, it forms part of the performance procedures of the General Medical Council (GMC), underpins objectively structured clinical examinations (OSCEs) and records of in-training assessment (RITA), and has been advocated for the selection of registrars in general practice and interviews. It has become central to the professional lives of all doctors and is treated as if it were a panacea ---but there is little consensus among trainees, trainers, and committees on what this approach entails. I aim to explore the origins and development of the competency approach, evaluate its current role in medical training, and discuss its strengths and limitations. Summary points : The birth of the competency movement : The competency approach did not result directly from recent scandals of incompetent doctors. It originated from parallel developments in vocational training in many countries, such as the national qualifications framework in New Zealand, the national training board in Australia, the national skills standards initiative in the United States, and the national vocational qualifications (NVQs) in the United Kingdom. This movement was driven largely by the political perceived need to make the national workforce more competitive in the global economy. For example, in Britain, the national vocational qualifications were developed as a set of standards each broken down into elements by which performance in the workplace can be assessed. This approach has since been adopted for training across other areas, particularly the technical and vocational fields. How does competency based training work? The basic essential elements consist of functional analysis of the occupational roles, translation of these roles ("competencies") into outcomes, and assessment of trainees' progress in these outcomes on the basis of demonstrated performance. Progress is defined solely by the competencies achieved and not the underlying processes or time served in formal educational settings. Assessments are based on a set of clearly defined outcomes so that all parties concerned, including assessors and trainees, can make reasonably objective judgments about whether or not each trainee has achieved them. Potential benefits of this approach include individualised flexible training and transparent standards. This approach has attracted several criticisms. Firstly, functional analysis of occupational roles is problematic. It is difficult to identify a range of competencies that truly cover work roles in their broadest sense and to represent adequately the types of knowledge relevant to the competency identified. Secondly, the assessment of competencies is by no means value free, and people who use it shape its meaning. Thirdly, the competency approach is based primarily on the behaviourist framework, which attempts to break down work roles into small discrete tasks. It ignores the connections between individual tasks and the meaning underlying each task. It therefore cannot represent the complex nature of situations in the real world. The danger is that these narrowly defined competencies will dominate the curriculum, which would not be suitable for learning in higher education. The approach using checklists and passing or failing candidates is superficial and often proves demotivating, as it encourages trainees to do the right thing to pass rather than to think critically and excel. The parties concerned ---trainees, employers, professional bodies, and the government ---may have different views about which aspect of the occupation is regarded as the most important. The process of developing competencies is at least partly political because it allows the government to influence what are included as important competencies and to allocate resources based on outcomes of performance. A recent review of published evaluative studies of competency based training found an increase in administrative burden but no convincing beneficial effects on motivating students, work performance, or relevance to the needs of industry. The rise of "holistic" varieties : As this behaviourist approach to learning would be even less appropriate for professions requiring complex skills, a range of broader competency approaches flourished. In 1991, the general national vocational qualifications, which include core skills such as numeracy, communication, and problem solving, were developed to supplement the NVQ framework, although doubts exist about whether such generic skills transferable to all context actually exist. An integrated approach acknowledges competency as a complex combination of knowledge, attitudes, skills, and personal values. A holistic approach takes into account the cultural and social context in assessing competence and focuses on how personal attributes are used to achieve outcomes in real life scenarios. A competency of a higher order ---meta-competency ---has been used to describe the general ability to learn and apply competencies effectively in many different aspects of a person's activities. These approaches attempt to make the competency based model less reductionist in nature. Current scene in medical training : Traditionally, the framework of medical training was time based, and students were assessed periodically to determine their grades. Equal weight was given to both process and outcome of learning. Emphasis was given to the understanding of basic concepts and principles, and skills were evaluated globally. Recently, competency based approaches have gradually taken over. Although the behaviourist approach may occasionally be used for training in areas where rigid protocols exist, such as the advanced life support course, holistic varieties of the competency based approach are used more widely. In Australia, criterion referenced procedures to set standards have been used to define and measure competency for the graduate entry medical programme. In the United Kingdom, the Royal College of General Practitioners distinguishes between clinical competence (what doctors can do) and clinical performance (what doctors do do) and defines competencies as a combination of knowledge, skills, and attitudes which, when applied to a particular situation, lead to a given outcome. Competency based medical training is usually developed in four steps: determine what the appropriate competencies are, devise training programmes, devise appropriate assessment methods, and set minimum pass standards. Appropriate competencies can be determined in several ways, such as the GMC's Good Medical Practice for its performance procedures, postal questionnaire surveys of examiners and the committee of trainee members for the part 2 of the examination for membership of the Faculty of Public Health Medicine. Competencies for general practitioners have been defined by using triangulation of results from focus groups with general practitioners, behavioural coding of general practitioners' consultations with patients, and interviews with patients. There is little evidence, however, that addressing each of these competencies separately is a more effective form of training and assessments than the traditional global approach. Figure | Competency based medical training is prominent in undergraduate medical education Competency based medical training is prominent in undergraduate medical education Based on the competency approach, the objective structured clinical examination using checklists and standardised patients was initially thought to be more reliable and objective and gradually replaced the traditional long case. A recent review has found, however, that, for equal testing time, it is slightly less reliable than the long case. Several possible reasons for this surprising finding were given: standardisation of what happens within a case does not eliminate the variability of performance across clinical problems, and the use of ratings in long cases may achieve higher reliability than checklists. Perhaps another reason is that checklists including attributes such as attitudes and personal values may achieve lower reliability than behavioural outcomes. If this were the case, the exclusive focus on outputs that is often perceived to be the key advantage of the competency based approach does not necessarily result in objective and reliable assessments. In their summary assessments, general practice registrars need to submit a video of seven consultations to demonstrate each prescribed competency at least four times. Some candidates find such an exercise exceedingly time consuming and think that it might hinder other educational opportunities and enjoyment of general practice. Leading royal colleges set criterion referenced minimum pass standards by a panel agreeing on the probable scores of borderline candidates for both the written examinations and the objective structured clinical examination. Although these procedures can be used to set standards for excellence, they currently tend to focus on the minimum acceptable standards. In other examinations and assessments, the pass standards may be more arbitrary. Other issues are important. Firstly, a key advantage of the competency approach is its focus on competencies achieved rather than time served, so that trainees can progress at their own pace. But the training period for undergraduate and postgraduate medical training is currently fixed. Secondly, the competency approach ignores the learning process, although the process is important for lifelong learning. Thirdly, with the focus of the competency approach on skills and attitudes rather than a solid understanding of the basic concepts and principles, the risk is that "medical education" may give way to "medical training." An evaluation : Compared with the traditional approach, the competency based approach potentially leads to individualised flexible training, transparent standards, and increased public accountability. If applied inappropriately, it can also result in demotivation, focus on minimum acceptable standards, increased administrative burden, and a reduction in the educational content. Higher order competencies need to be defined and developed more robustly. We should be cautious of adopting the competency based approach universally across stages of medical training for which well defined and validated competencies are unavailable. After all, it is just one of many potentially useful approaches that may have a role at various stages of the educational progress. Backmatter: PMID- 12364376 TI - Clinical and Taxonomic Status of Pathogenic Nonpigmented or Late-Pigmenting Rapidly Growing Mycobacteria AB - The history, taxonomy, geographic distribution, clinical disease, and therapy of the pathogenic nonpigmented or late-pigmenting rapidly growing mycobacteria (RGM) are reviewed. Community-acquired disease and health care-associated disease are highlighted for each species. The latter grouping includes health care-associated outbreaks and pseudo-outbreaks as well as sporadic disease cases. Treatment recommendations for each species and type of disease are also described. Special emphasis is on the Mycobacterium fortuitum group, including M. fortuitum, M. peregrinum, and the unnamed third biovariant complex with its recent taxonomic changes and newly recognized species (including M. septicum, M. mageritense, and proposed species M. houstonense and M. bonickei). The clinical and taxonomic status of M. chelonae, M. abscessus, and M. mucogenicum is also detailed, along with that of the closely related new species, M. immunogenum. Additionally, newly recognized species, M. wolinskyi and M. goodii, as well as M. smegmatis sensu stricto, are included in a discussion of the M. smegmatis group. Laboratory diagnosis of RGM using phenotypic methods such as biochemical testing and high-performance liquid chromatography and molecular methods of diagnosis are also discussed. The latter includes PCR-restriction fragment length polymorphism analysis, hybridization, ribotyping, and sequence analysis. Susceptibility testing and antibiotic susceptibility patterns of the RGM are also annotated, along with the current recommendations from the National Committee for Clinical Laboratory Standards (NCCLS) for mycobacterial susceptibility testing. Keywords: INTRODUCTION : The species of rapidly growing mycobacteria (RGM) capable of producing disease in humans consist primarily of the Mycobacterium fortuitum group, the M. chelonae/abscessus group, and the M. smegmatis group. Key features for identification of these groups are the presence of typical long-chain fatty acids known as mycolic acids, growth of readily visible colonies on primary isolation within 7 days on multiple types of solid media, arylsulfatase activity within 3 days or 2 weeks, and the absence or slow appearance of any pigmentation . Historically, the M. fortuitum group has been composed of two known species and a taxon which has been reported to include more than one species . The species include M. fortuitum (formerly M. fortuitum biovar fortuitum), M. peregrinum (formerly M. fortuitum biovar peregrinum) and the taxon known as the unnamed third biovariant complex . Several additional taxa to be discussed later are candidates for inclusion in this group. M. chelonae (formerly M. chelonae subsp. chelonae) and M. abscessus (formerly M. chelonae subsp. abscessus) , along with the newly recognized species M. immunogenum , are members of a group known collectively as the M. chelonae-abscessus group. Finally, the M. smegmatis group contains M. smegmatis sensu stricto and two newly described species: M. goodii, and M. wolinskyi . Prior to recent molecular reevaluations, the taxa in the M. chelonae-abscessus and the M. fortuitum group were considered "subspecies" or "biovariants," respectively. However, the introduction and evolution of 16S ribosomal gene (rDNA) sequencing provided strong evidence that these biovars and subspecies were in fact separate species. Gene sequencing also permitted much easier recognition of new taxa, since investigators could use data banks for strain comparisons rather than performing the much more technically difficult genomic DNA-DNA pairing experiments with all potentially related taxa. The genomic DNA-DNA pairing experiments that show <70% homology to other species still remain the "gold standard" for recognition of new species, but such studies are performed infrequently and many new species of mycobacteria are based genetically on 16S rDNA comparisons only. In this review, we update the taxonomy of these nonpigmented RGM and point out many of the changes brought about by such newer technologies as 16S rRNA gene sequencing; high-performance liquid chromatography (HPLC) of mycolic acid esters, including fluorescence-HPLC; and PCR restriction fragment length polymorphism (RFLP) analysis (PRA) of a 439-bp fragment (referred to as the Telenti fragment) of the 65-kDa heat shock protein-encoding gene (hsp65) . Clinical disease caused by these groups, together with their drug susceptibility, and most effective drug treatment are also addressed. The disease syndromes caused by these organisms are listed in Table . EPIDEMIOLOGY : The nonpigmented RGM are extremely hardy and thrive in even the most hostile of environments (, , -). Some of the taxa ---such as a subgroup of M. peregrinum, some members of the unnamed third biovariant complex of M. fortuitum, and most isolates in the M. smegmatis group ---are able to grow at 45C . Additionally, some species, such as the M. chelonae/abscessus group and M. mucogenicum, resist the activity of disinfectants and biocides such as organomercurials, chlorine, and alkaline glutaraldehyde . These hardy species of RGM are commonly seen in municipal tap water . One study by Carson et al. showed that 55% of the incoming city water in hemodialysis centers throughout the United States contained RGM. Some outbreaks of human infection related to these organisms have involved hospital water systems as the microbial reservoirs. Recently, the presence of acid-fast mycobacteria in up to 90% of biofilms (the slime layer present at water-solid interfaces) taken from piped water systems has been described . The presence of acid-fast mycobacteria in these biofilms probably serves as a major environmental reservoir for organisms such as M. kansasii, M. mucogenicum, M. simiae, M. xenopi, and M. gordonae. Because of the ubiquity of the RGM, human infections have been reported from most geographic areas in the world and species of RGM have been recovered from 30 to 78% of soil samples throughout the United States . Most nosocomial (hereafter referred to as health care-associated) outbreaks and pseudo-outbreaks have occurred in the United States and seem to be concentrated mainly in the South. M. senegalense, originally found in Africa, has never been described elsewhere. Such a localized distribution among the RGM seems to be rare, however. Health care-associated outbreaks and pseudo-outbreaks commonly involve exposure to tap water or water sources such as ice, ice water, and water-based solutions . Contaminated ice machines are a relatively important hospital reservoir for the RGM, especially M. fortuitum. Reported disease outbreaks have included sternal wound infections , surgical wound infections following plastic surgery , and postinjection abscesses . Catheter infections also have been associated with the RGM, including M. mucogenicum . Pseudo-outbreaks of disease, defined as clusters of false infections or artifactual clustering of real infections, have been associated with contaminated bronchoscopes and automated endoscopic cleaning machines with tap water as the source of the organism . Localized infection in sporadic community-acquired disease usually occurs after a traumatic injury followed by potential soil or water contamination . Such injuries as stepping on a nail, motor vehicle accidents, compound fractures, etc., are typical of the clinical histories seen in patients with RGM disease . TAXONOMY AND CLINICAL SIGNIFICANCE : M. fortuitum Group | The M. fortuitum group has historically included three taxa: M. fortuitum, M. peregrinum, and the unnamed third biovariant complex. Here, we propose the addition of six more species to the group: M. mucogenicum (for reasons given below), M. senegalense, M. mageritense, and three recently described (M. F. Schinsky, M. P. Douglas, A. G. Steigerwalt, R. W. Wilson, M. M. Floyd, M. I. Daneshvar, B. A. Brown-Elliott, R. J. Wallace, Jr., M. M. McNeil, D. J. Brenner, and J. M. Brown, Abstr. 12th Int. Symp. Biol. Actino., abstr. P-117, 2001) species: M. septicum, M. houstonense, and M. bonickei. Few studies have accurately separated these eight taxa, and they were often referred to simply as "M. fortuitum," as if they were a single species. For purposes of this review, "M. fortuitum group" is used when no subgrouping was performed or the data include all subgroups. Names of the specific taxa are used when the isolates were so identified by current methods of carbohydrate utilization or PRA of the hsp65 gene . Common features within this group are a positive 3-day arylsulfatase, the absence of pigmentation, a positive nitrate reductase, a positive iron uptake, and susceptibility to multiple drugs including polymyxin B, sulfonamides, and the newer fluoroquinolones. Most of these taxa grow better at 30 than 35C. Their 16S rDNA sequences generally differ by 15 bp or less. Historical perspective. | (i) M. fortuitum. | M. fortuitum was the designation given by da Costa Cruz to a strain of RGM (ATCC 6841T) isolated from a human postinjection abscess in 1938. Subsequently, Stanford and Gunthorpe determined that the isolate was identical to an established species known as M. ranae, with the type strain isolated from a frog by Kuster in 1905 . Runyon challenged the name M. ranae, in part because many isolates identified as M. ranae were subsequently found to be M. smegmatis and in part because the name M. fortuitum was more widely recognized and established in the medical literature. Hence, in 1972, the Judicial Commission of the International Committee of Systematic Bacteriology of the International Association of Microbiological Societies ruled in favor of Runyon's recommendation to maintain the species designation M. fortuitum, which has remained to the present . (ii) M. peregrinum. | M. peregrinum was first proposed in 1962 when Bojalil and colleagues published the first Adansonian or numerical classification of mycobacteria . This analysis separated mycobacteria into 12 different categories or branches based on physiological characteristics. Branch 1 consisted of M. smegmatis, M. phlei, and a group of other strains that showed the greatest metabolic capacity of all the strains analyzed. The name M. peregrinum sp. nov. was proposed for this latter group (Latin adjective meaning strange or foreign) because "they were the only non-pigmented strains in Branch 1" . The type strain is ATCC 14467T. (iii) M. fortuitum third biovariant complex. | In 1966 Bonicke divided isolates of the M. fortuitum group into three subgroups on the basis of differences in acid production from carbohydrates and designated them biotypes A, B, and C . Biotype A had no unique carbohydrates, biotype B was mannitol positive, while biotype C was mannitol and inositol positive. Subsequently, Pattyn et al. renamed these three Bonicke biotypes M. fortuitum biovariant fortuitum, M. fortuitum biovariant peregrinum, and an unnamed third biovariant. In 1984, Tsang et al. used a combination of chemical analysis, seroagglutination, and enzyme-linked immunosorbent assay to compare the glycolipids of the RGM and provided evidence that M. fortuitum biovar peregrinum was a separate species . M. fortuitum biovar peregrinum was later confirmed as an independent species, as M. peregrinum, based on genomic DNA-DNA relatedness studies by Levy-Frebault et al. and Kusonoki and Ezaki that showed <70% genomic DNA-DNA homology of M. peregrinum (ATCC 14467T) to other RGM species. By 16S rDNA sequencing, M. peregrinum (ATCC 14467T) is unique, with a Hamming distance of 9 to 15 bp from other members of the M. fortuitum group. Wallace et al. noted two subgroups within M. peregrinum. One group contained the ATCC type strain (ATCC 14467T), which was pipemidic acid susceptible; while the other, represented by ATCC 35755, was pipemidic acid resistant. The two groups have different PRA patterns with the hsp65 gene . An analysis of the taxonomic status of this second group using phenotypic and genetic analysis is ongoing (M. F. Schinsky, R. E. Morey, M. P. Douglas, A. G. Steigerwalt, R. W. Wilson, M. M. Floyd, M. I. Daneshvar, B. A. Brown-Elliott, R. J. Wallace, Jr., M. M. McNeil, D. J. Brenner, and J. M. Brown, unpublished). M. fortuitum biovar fortuitum was also elevated to species status (M. fortuitum) based on DNA-DNA pairing studies which showed <70% homology of M. fortuitum ATCC 6841T to other taxa (,). By 16S rDNA sequencing, M. fortuitum differs by only 6 bp from the unnamed third biovariant sorbitol-negative group, by 8 bp from M. senegalense, and by 15 bp from M. peregrinum (ATCC 14467T). The unnamed third biovariant, first described by Bonicke in 1966 , was later characterized and then subdivided into two groups by Wallace et al. on the basis of a number of characteristics including sorbitol utilization. The two unnamed third biovariant groups were known as M. fortuitum third biovariant, sorbitol positive, and M. fortuitum third biovariant, sorbitol negative . By 16S rDNA sequencing, these two groups had a Hamming distance of 9 , with proposed representative strains being ATCC 49403 (sorbitol positive) and ATCC 49404 (sorbitol negative) . They differed by up to 20 bp from other M. fortuitum group members. Both sorbitol-positive and negative groups appeared heterogeneous, especially when studied for beta-lactamase alleles but also when compared by PRA of the hsp65 gene . (a) M. houstonense and M. bonickeii (proposed species). Investigations currently under way may better delineate the multiple taxa or species within the third biovariant complex. Currently, at least six new species have been delineated. The majority of the sorbitol-positive group have been renamed M. houstonense, while the majority of the sorbitol-negative group have been renamed M. bonickei (Schinsky et al., Abstr. 12th Int. Symp. Biol. Actino.). Additional details of these species await publication of the entire study. (b) M. septicum.M. septicum is one of the new species within the third biovariant complex. The type strain (ATCC 700731T) was the causative agent of central- line sepsis in a child with metastatic hepatoblastoma . By HPLC analysis of mycolic acids, the isolate was distinctive but closely related to other members of the M. fortuitum group , including M. senegalense. Standard biochemical testing showed the isolate to be similar to members of the unnamed third biovariant complex sorbitol-negative group (i.e., mannitol positive, inositol positive, sorbitol negative). Although initial testing showed the isolate to be arylsulfatase negative, it has been found to be positive in other laboratories (K. Jost, unpublished data). Analysis of the 16S rDNA showed a sequence related to but not identical to M. fortuitum, M. peregrinum, and M. senegalense (Schinsky et al., Abstr. 12th Int. Symp. Biol. Actino.). (c) M. mageritense.The first five reported organisms in the new species M. mageritense were isolated from human sputa in two hospitals in Spain but were not considered clinically significant . Recently, six clinical isolates of this species were recovered in the United States, four of which were associated with clinical disease . Phenotypically, by antibiotic susceptibility patterns and biochemical tests, isolates of M. mageritense resemble sorbitol-positive members of the M. fortuitum third biovariant complex (positive for mannitol, inositol, and sorbitol) . By sequencing of its 16S rDNA, however, genetically, M. mageritense is more closely aligned with members of the M. smegmatis group . M. mageritense differs by only 9 bp from the type strain of M. wolinskyi (ATCC 700010T), by 16 bp from the type strain of M. goodii (ATCC 700504T), and by 18 bp from the type strain of M. smegmatis sensu stricto (ATCC 19420T). It generally differs by 23 to 28 bp from members of the M. fortuitum group. Future phenotypic and molecular studies of M. mageritense that include larger numbers of isolates may provide a more complete and accurate taxonomic placement of this species. (iv) M. mucogenicum. | M. mucogenicum has been recognized as a species since 1995. The organism was first called M. chelonae-like organism (MCLO) in 1982, when it was reported as the etiologic agent in a peritonitis outbreak involving two peritoneal dialysis units . It was given the designation MCLO because the outbreak strain, just as M. chelonae, was nitrate negative and growth was inhibited on 5% NaCl. In 1993, a large number of sporadic clinical isolates were evaluated by Wallace and colleagues (200) by using biochemical reactions, HPLC of mycolic esters, and antibiotic susceptibility patterns. Subsequently, in 1995, Springer et al. proposed the name M. mucogenicum for this organism group, reflecting the highly mucoid character of the isolates . This species has always been grouped with the M. chelonae-abscessus group, but is unlike that group in that approximately 50% of isolates are nitrate positive, have a weak but positive iron uptake, and are much more susceptible to antibiotics including the fluoroquinolones, amoxicillin-clavulanic acid, polymyxin B, and cephalothin. In addition, by 16S rDNA sequencing M. mucogenicum is more closely related to the M. fortuitum group than to the M. chelonae-abscessus group . The type strain (ATCC 49650T) of M. mucogenicum differs by 11 to 18 bp from members of the M. fortuitum group but differs by 35 bp from M. abscessus and by 38 bp from M. chelonae. Hence, we propose to add this species to other members within the M. fortuitum group. (v) M. senegalense. | M. senegalense was originally described by Chamoiseau in 1973 as a subspecies of M. farcinogenes . Later, however, it was recognized as a different species closely related to M. fortuitum . Like M. peregrinum, M. senegalense is positive only on mannitol, when tested on common sugars, but has a unique PRA pattern with the hsp65 Telenti fragment . Its 16S rDNA sequence differs by 4 bp from the unnamed M. fortuitum third biovariant complex (sorbitol-positive) strain ATCC 49403; by 5 bp from the unnamed third biovariant complex (sorbitol-negative) strain ATCC 49404; and by only 8 bp from M. fortuitum (ATCC 6841T). By DNA comparison studies, however, it is a species apart from these other organism groups. Type of disease. | (i) Community-acquired disease. | The M. fortuitum group accounts for 60% of cases of localized cutaneous infections caused by RGM but is a rare cause of chronic mycobacterial pulmonary disease . Localized cutaneous disease generally occurs in previously healthy hosts, and drug-induced immune suppression appears to result in minimal increase in this risk. Wallace et al. studied 123 patients with extrapulmonary disease caused by RGM and reported that 76 (63%) of these infections were due to the M. fortuitum group. Griffith et al. studied 154 patients with pulmonary disease due to RGM and reported that only 16% of the infections encountered during the 15-year study were due to the M. fortuitum group . The M. fortuitum group was a common lung pathogen (50% of cases) only in the setting of chronic aspiration secondary to underlying gastroesophageal diseases such as achalasia. Approximately 25% of M. fortuitum group infections based on one study have been associated with a variety of diseases other than skin or soft tissue infections including cervical lymphadenitis, mastoiditis, and meningitis. . Species in the M. fortuitum group are relatively rare causes of disseminated disease compared to other pathogenic RGM species, especially M. chelonae and M. abscessus. (ii) Health care-associated disease. | The M. fortuitum group is responsible for the majority (60 to 80%) of cases of postsurgical wound infections and catheter infections caused by the RGM . Most of the responsible organisms are M. fortuitum and are detailed below in the specific section on M. fortuitum . Geography. | Cutaneous disease caused by the M. fortuitum group, although reported from all over the United States and worldwide, has been recognized most commonly in the southeastern United States . A report in 1989 indicated that about 80% of wound isolates related to cardiac surgery were from seven southern coastal states: Texas, Louisiana, Georgia, Maryland, Alabama, Florida, and South Carolina. A second report published in the same year found that 92% of 37 identified cases of surgical wound infections following augmentation mammaplasty were in patients from southern coastal states, with the majority being from Texas, Florida, and North Carolina . Approximately 80% of isolates in both studies combined belonged to the M. fortuitum group. Individual taxa. | (i) M. fortuitum. | (a) Community-acquired disease. In essentially all series of community-acquired or health care-associated disease attributed to the M. fortuitum group, most or all of the cases are due to M. fortuitum. In a study of 154 patients with RGM pulmonary disease, Griffith et al. reported 13% of infections were due to M. fortuitum and only 3% were due to other members of the M. fortuitum group. In a series of cases of extrapulmonary disease cased by the M. fortuitum group, Wallace et al. reported that almost 80% of the infections were due to M. fortuitum. Disseminated infections with M. fortuitum are rare. The first case report appeared in 1990, when Sack described a patient with a history of intravenous IV drug abuse and AIDS, who had cutaneous lesions from which M. fortuitum was isolated. Cultures of specimens from lymph nodes, urine, pleural effusions, and feces all yielded M. fortuitum. (b) Health care-associated disease. M. fortuitum has been implicated in numerous outbreaks of hospital or health care-associated infections . These include sternal wound infections; postinjection abscesses related to electromyography needles ; and a respiratory disease outbreak in Washington, D.C. . M. fortuitum has also been recovered from sporadic cases of surgical wound infections and catheter-related infections and is the most common RGM species in women with surgical wound infections following augmentation mammaplasty . Details of some of the outbreaks are listed below. Cardiac disease outbreaks. Isolates of M. fortuitum have been found in sternal wound infections in Hong Kong, Colorado, Nebraska, and Texas . Interestingly, three of the four involved multiple RGM species or multiple strains of the same species. The first outbreak occurred in 1976 in a hospital in Colorado. Of nine patients who underwent cardiovascular surgery within a 2-week period, four became infected with a single genetic strain of M. fortuitum. Despite intense infection control efforts to recover the organism from environmental sources and subsequent molecular analysis of the isolates recovered, no environmental source was identified . In 1981, Preheim and colleagues (L. C. Preheim, M. J. Bittner, D. K. Giger, and W. E. Sanders Jr., Program Abstr. 22nd Intersci. Conf. Antimicrob. Agents Chemother., abstr. 564, 1982) reported an outbreak in a Nebraska hospital that involved five patients who developed sternotomy infections following coronary bypass surgery. All had multiple debridements. One patient died (Preheim et al., 22nd ICAAC). Later analysis by plasmid profiles , multilocus enzyme electrophoresis (MEE), and pulsed-field gel electrophoresis showed the presence of two strains of M. fortuitum. Another outbreak in 1981 involving M. abscessus and M. fortuitum, was reported in a hospital in Corpus Christi, Tex. . Of 51 patients undergoing surgery over a 6-month period, 6 (11.3%) developed infections with one of these RGM. One patient with sternal wound infection due to M. fortuitum died as a result of complications from sternectomy and antimicrobial therapy. A subsequent study was performed in which culture media were inoculated with samples taken from environmental sites including the municipal water system, water from the cold water tap in the operating room, ice machines, swabs of lamps, oxygen tanks, suction apparatus, and commercial bone wax. Multiple sites were culture positive for M. fortuitum strains which, by MEE and subsequent pulsed-field gel electrophoresis, were identical to the disease strains . This is the first and only cardiac surgery outbreak in which the piped-water system has been implicated as the source of the pathogenic RGM. In this case, the cardioplegia solution was cooled in ice made from tap water contaminated with the outbreak strains. The fourth and latest outbreak of M. fortuitum infection following cardiac surgery came from Hong Kong . Both M. fortuitum and M. peregrinum were implicated. The outbreak spanned the years 1987 to 1989, and 7 (33%) of 21 wound infections were due to M. fortuitum. (Details of the M. peregrinum strains are presented in the M. peregrinum section, below.) Subsequent ribotyping and susceptibility testing suggested that multiple strains were involved . No source for the prolonged Hong Kong outbreak was identified. Postinjection abscess disease. Most outbreaks of postinjection abscesses caused by RGM have involved M. abscessus. However, Nolan et al. reported an outbreak of M. fortuitum in five (83%) of six patients undergoing electromyography at a facility in Washington state. The outbreak was traced to a break in a manual procedure used for sterilizing the reusable needle electrodes. When subsequent sterilizations were performed using an autoclave, no further cases were reported . Plastic surgery-related disease. No large outbreaks of wound infection following plastic surgery due to M. fortuitum have been reported, but sporadic wound infections following plastic surgery have been well documented . Many of these infections, involving M. fortuitum, occurred after augmentation mammaplasty . A study by Wallace et al. identified 37 cases of surgical wound infection with RGM following augmentation mammaplasty. Of these isolates, 26 (70%) were identified as M. fortuitum. Several surgeons had more than one case, suggesting that local environmental factors were important to disease development. Six cases of M. fortuitum disease associated with spontaneous breast abscesses with no history of surgery or trauma were also reported. Pseudo-outbreaks. One pseudo-outbreak involving M. fortuitum occurred in 1987 in Houston, Tex., and involved ice contaminated with M. fortuitum. Although M. fortuitum was recovered from specimens from four patients who underwent bone marrow aspiration, none of the patients had evidence of disease due to M. fortuitum. It was later noted that each of the patients whose aspirates yielded M. fortuitum on culture also had viral cultures performed, and this required that the syringe containing the aspirate be plunged into ice for transport. Samples taken from the ice machine located on the same floor as the patients with positive aspirate samples were also positive for M. fortuitum. Samples from ice machines on other floors did not grow M. fortuitum . No molecular typing was performed. (ii) Unnamed third biovariant complex including M. septicum, M. mageritense, M. houstonense (proposed), and M. bonickei (proposed). | Pattyn et al. published one of the earliest studies of the unnamed third biovariant complex approximately 30 years ago. The authors commented that most of the isolates they studied were environmental strains and none had a definite disease association. (a) Community-acquired disease. In 1991, Wallace et al. , characterized 85 clinical isolates of unnamed M. fortuitum third biovariant complex, all of which were disease associated. These represented 16% of 410 isolates of the M. fortuitum group submitted to a Texas laboratory and 22% of 45 isolates submitted to the Queensland, Australia, state laboratory. Over 75% of the infections involved skin, soft tissue, or bone. Clinical histories were available for 52 patients with skin and soft tissue infections, and the type of injury responsible for infection was reported for 42 patients. Of these 42 patients, 29% had osteomyelitis confirmed by bone biopsy. Most infections occurred following puncture wounds or compound (open) fractures. Only two patients (children) had no history of trauma leading to their infection. Metal puncture wounds (48%) or motor vehicle accidents (26%) were the most common histories given, and approximately 40% of the injuries involved the foot or leg. Stepping on a nail was the classic scenario . No cases of disseminated disease were observed in this study, and, to date, none have been reported by other authors. This study also identified 26 isolates of the third biovariant complex from pulmonary sources. Clinical significance was not determined, but almost certainly some of these were disease producing. None of these isolates were studied by molecular methods that would identify them as one of the recently described species within the complex (i.e., M. houstonense [proposed], M. bonickei [proposed], M. septicum, or M. mageritense) (190; Schinsky et al., submitted). (b) Health care-associated disease. A small number of cases have been reported in association with hospital-acquired disease in other studies, including wound infections following cardiac surgery and augmentation mammaplasty . (c) Geography. The original description by Pattyn et al. of this group was based on isolates from Europe and Africa. Later in 1983, Levy-Frebault et al. characterized 23 additional isolates of M. fortuitum from France and found 6 environmental isolates that were identified subsequently as M. fortuitum unnamed third biovariant complex . A later study by Wallace et al. included 10 isolates from Australia. The remaining 70 isolates were from the United States, of which approximately 70% were from Texas, 6% were from Florida, and 5% were from Alabama and Georgia . Four percent or less of the third biovariant isolates identified in this study were from other states including Tennessee, Louisiana, New Hampshire, South Carolina, Arkansas, and Connecticut (our unpublished data). (iii) M. peregrinum. | Currently, there is no published series or review evaluating the clinical significance of M. peregrinum. No case of disseminated infection due to M. peregrinum has been reported. However, a small number of cases of sporadic infections have been reported and have been associated with diseases similar to other members of the M. fortuitum group. These include chronic lung disease , sternal wound infections , and cutaneous disease . In general, they represent only 1 to 2% of sporadic community-acquired or health care-associated infections due to RGM. M. peregrinum has been reported as a cause of a pseudo-outbreak of respiratory disease due to a contaminated ice machine and as the etiologic agent of 67% of cases in an outbreak of sternal wound infections from Hong Kong that occurred from 1987 to 1989 . (iv) M. mucogenicum. | M. mucogenicum , formerly MCLO, was originally discovered in two outbreaks of peritonitis in 1976 and 1978, in 5 (18%) of 22 and 5 (63%) of 8 patients, respectively, undergoing intermittent chronic peritoneal dialysis . Infections were traced to the use of contaminated automated chronic peritoneal dialysis machines in two dialysis centers. Seven sporadic cases of peritonitis due to M. mucogenicum from the same two centers were also diagnosed. The investigators' findings suggested that ineffective disinfection of the equipment followed by colonization of the machines by this newly described species was responsible for the outbreaks. Three studies showed that M. mucogenicum, like other pathogenic species of RGM, is relatively resistant to formaldehyde and glutaraldehyde disinfectants, which are typically used to disinfect dialysis equipment . In 1985, Bolan et al. described a mycobacterial disease outbreak in a hemodialysis center in Louisiana. Of 26 identified isolates, 25 were M. abscessus and 1 was M. mucogenicum. The single factor common to all patients was their exposure to hemodialyzers (artificial kidneys) that had been ineffectively treated with concentrations of disinfectant (formaldehyde) that were below effective levels for these RGM species . Although M. mucogenicum is often recovered as a laboratory contaminant, these three outbreaks alert us to the potential significance of this waterborne organism. In 1993, Wallace et al. evaluated 87 sporadic isolates of M. mucogenicum. Of these isolates, 54 (62%) were respiratory, and only 2 (4%) of them (both from patients with AIDS) were clinically significant. For the remaining 33 nonrespiratory isolates, significant clinical diseases included posttraumatic wound infections and catheter-related sepsis. Recovery of M. mucogenicum from skin, wound, or blood cultures was most often associated with clinical disease. In contrast, a single positive sputum culture is almost never clinically significant. Goldblatt and Ribas recently reported the first case of a patient with granulomatous hepatitis caused by M. mucogenicum. No cases of disseminated cutaneous disease due to M. mucogenicum have been reported. The frequent presence of this organism in tap water, including ice machines , may contribute to the transient colonization or contamination of sputum samples . In a study of 113 mycobacterial isolates from tap water samples from various geographic sites, the most frequently occurring nontuberculous mycobacterium (41%) was M. mucogenicum . This study underscores the potential health risks of these ubiquitous organisms. M. mucogenicum was first described by Band et al. in the peritonitis outbreak in Washington state. Since that time, isolates have been recovered from Texas, Arizona, Maryland, Delaware, Illinois, Pennsylvania, Missouri, Florida, Washington D.C. and California (R. J. Wallace, Jr., unpublished data). (v) M. senegalense. | M. senegalense is an etiologic agent of farcy, a disease of skin and superficial lymphatics in African bovines . It has not been reported from environmental or clinical cultures in the United States or Europe. (vi) M. septicum. | The type strain of M. septicum (ATCC 700731T) was recovered in Australia from three separate blood cultures and a central venous catheter tip after its removal . No other isolates have been reported. M. chelonae-abscessus Group | The M. chelonae-abscessus group contains three species: M. chelonae, M. abscessus, and M. immunogenum. When researching the literature, it is important to recognize the difficulty in establishing which species is responsible for cases of disease associated with this group. This is particularly cogent for diseases during the period from 1972 to 1992 (generally prior to 1990), when M. chelonae and M. abscessus were considered to belong to the same species ("M. chelonei" or "M. chelonae"). Although they have been recognized as subspecies since an international collaborative study published by Kubica et al. in 1972, little effort was made to separate them. Furthermore, interpretation of the literature after 1992 from investigators unaware that these organisms are, in fact, two species continues to be a problem. For purposes of this review, the description "M. chelonae-abscessus group" is used for cases when no subgrouping was performed or the data include all subgroups; the specific species names are used when an isolate was characterized by its ability to utilize citrate or by genetic methods (eg., PRA of the hsp65 Telenti gene fragment) . Common features within this group are a positive 3-day arylsulfatase, the absence of pigmentation, better growth at 30 than 35C, a negative nitrate reductase, a negative iron uptake, and resistance to polymyxin B and most other antimicrobial drugs except amikacin and clarithromycin. Historical perspective. | In 1953, Moore and Frerichs recovered an RGM (now ATCC 19977T) from a knee abscess. The authors thought the isolate was distinctive biochemically and morphologically from other RGM and identified it as a new species, M. abscessus. This name was selected because of the ability of the organism to produce deep subcutaneous abscesses . Stanford et al. first reported studies on clinical isolates of what was then known as M. borstelense . Among the isolates they studied were isolates from postinjection abscess outbreaks in Holland and England, human strains from other parts of Europe and Africa, and environmental strains. This study resulted in the official adoption of the name "M. chelonei" for these isolates, a name which was later changed to the more correct Latin, M. chelonae. Many early investigators, however, believed M. chelonae and M. abscessus to be the same organism because they showed almost identical biochemical features. A cooperative numerical phenotypic study by the International Working Group on Mycobacterial Taxonomy (IWGMT) published in 1972, however, demonstrated that the two taxa were sufficiently different to be classified as subspecies and renamed them M. chelonae subspecies chelonae and M. chelonae subspecies abscessus . Using genomic DNA-DNA hybridization studies, M. chelonae subspecies abscessus (ATCC 19977T) was later shown to be a separate species on the basis of <70% genomic homology with other RGM taxa, including M. chelonae subspecies chelonae (ATCC 35752T) . In 1992, Kusunoki and Ezaki elevated M. abscessus comb. nov. to species status, and M. chelonae subsp. chelonae once again became M. chelonae . Interestingly, by 16S rDNA sequencing, M. chelonae and M. abscessus differ by only 4 bp and are examples of the few different nontuberculous mycobacterial species that have identical 16S rDNA hypervariable region A sequences . M. immunogenum, formerly M. immunogen, is a newly described RGM first recognized in contaminated metalworking fluids . It is closely related to M. chelonae and M. abscessus but readily distinguished by genetic methods . The organism was named for its potential relationship to cases of hypersensitivity pneumonitis in factory workers (metal grinders) who used mycobacterium-contaminated metal-grinding fluids for lubrication and cooling of their machines. By 16S rDNA sequencing, the ATCC type strain ATCC 700505T differs by only 8 bp from M. abscessus and by 10 bp from M. chelonae. It is morphologically similar to M. abscessus but has a different drug susceptibility pattern and a different PCR restriction analysis pattern of the hsp65 Telenti fragment . Type of disease. | (i) Community-acquired disease. | For several years prior to the current molecular microbiology era, the M. chelonae-abscessus group was referred to collectively as M. chelonei or M. chelonae without further differentiation of species (see "Taxonomy and clinical significance above"). The M. chelonae-abscessus group has been associated with a variety of different diseases. The most common clinical disease is probably chronic lung disease, usually in elderly women with bronchiectasis or young adults with cystic fibrosis (CF). The M. chelonae-abscessus group is responsible for approximately 95% of disseminated cutaneous infections caused by the RGM. Unlike patients with a localized infection, patients with disseminated cutaneous disease have multiple painful draining small abscesses that involve the arms and legs. Localized cellulitis, osteomyelitis, and small-joint arthritis are also commonly associated with the M. chelonae-abscessus group. (ii) Health care-associated disease. | Sporadic (single) cases of otitis media, following tympanostomy tube placement, catheter infections, and postsurgical wound infections following a variety of surgical procedures (especially plastic surgery) also have involved this group of RGM . The M. chelonae-abscessus group has been involved in several health care-associated disease outbreaks including post-cardiac surgery sternal wound infections and vein graft site infections . Other outbreaks of M. chelonae-abscessus group infection have involved plastic surgery , hemodialysis, and miscellaneous outbreaks including wound infections following laparoscopy, liposuction , and post-tympanostomy tube placement . Additionally, postinjection abscess outbreaks following the use of multidose vials or contaminated biologicals also have been reported. Vaccine-related outbreaks involving M. chelonae-abscessus as contaminants also are recorded . (Isolates from most outbreaks since 1980 were restudied at a later date and shown to be M. chelonae or M. abscessus. These outbreaks are detailed under the specific species.) In addition to these true outbreaks of infection, several health care-associated pseudo-outbreaks have been described in conjunction with contaminated or malfunctioning bronchoscopes (, , , ; K. Petersen, N. Bus, V. Walter, and C. Chenoweth, Abstr. Infect. Control Hosp. Epidemiol., abstr. S-32, 1994), automated endoscope-cleaning machines , and contaminated laboratory reagents . Geography. | The M. chelonae-abscessus group is a collection of ubiquitous organisms found in soil and water worldwide. Outbreaks of M. chelonae-abscessus group disease have occurred primarily in the United States in southern coastal states and have been reported in North Carolina , Louisiana, Georgia, Florida, and Texas . Generally, almost all of the states in the southern United States have reported disease with M. abscessus . Outside the United States, isolates have been recovered from Hungary , Japan, Germany , Canada , France, Italy, Sweden, Australia, Belgium, Switzerland, Colombia, South America , and the United Kingdom (; unpublished data). Individual taxa. | (i) M. chelonae. | M. chelonae is one of the most antibiotic- resistant species of the pathogenic RGM. Like M. abscessus, M. chelonae is involved in several different types of community-acquired infections. (a) Community-acquired disease. Pulmonary disease. Unlike M. abscessus and M. fortuitum, M. chelonae is only rarely a cause of chronic lung disease. In the series of 154 patients with chronic lung disease due to RGM reported by Griffith et al. , only 1 of 146 isolates identified to species was an M. chelonae. Disseminated disease. M. chelonae causes three basic types of cutaneous disease (see Table ). The most common type is disseminated cutaneous disease, which occurs when the host is chronically immunosuppressed . Wallace et al. reported that 53% of 100 clinical skin and/or soft tissue and/or bone isolates of M. chelonae were from patients with disseminated cutaneous infections . These infections were seen in patients receiving long-term corticosteroids and/or chemotherapy, primarily because of underlying organ transplantation, rheumatoid arthritis, or other autoimmune disorders . Chronic lung disease, solid-tumor malignancies and other disorders were less frequently associated with this disease . McWhinney et al. described three cases of M. chelonae in febrile neutropenic patients receiving chemotherapy. Infections with M. chelonae have occurred predominantly in patients with drug-induced immunocompromised status. In contrast, disease states which lead to immune suppression, such as AIDS, have not been significant risk factors for the development of disseminated M. chelonae infection. Localized infections. The second type of infection seen with M. chelonae involves community-acquired localized infections following trauma . These infections range from localized cellulitis or abscess to osteomyelitis. In the series by Wallace et al., 35% of the infections caused by M. chelonae were of this group (i.e., localized wound infections) . (b) Health care-associated disease. Sporadic localized wound infections following medical or surgical procedures including needle injections can occur with M. chelonae but are rare compared to infections with M. fortuitum and M. abscessus. Health care-associated outbreaks due to M. chelonae are also rare and have been observed only following injection with contaminated syringes or needles, the implantation of contaminated porcine heart valves , and, recently, the use of liposuction . In this last outbreak, the organism was recovered from tap water connected to the suction cannulas. The third, and least common, type of infection caused by M. chelonae, yet the most common type of health care-associated disease, is that of catheter-related infections . In 1992, Wallace et al. reported that 8 of 100 clinical isolates of M. chelonae were associated with intravenous catheters, an additional 3 involved chronic peritoneal dialysis catheters, and 1 involved a hemodialysis shunt . They found that both the use of corticosteroids and renal failure were risk factors for these catheter-related infections . (ii) M. abscessus. | M. abscessus and M. chelonae are probably the most antibiotic resistant species of the pathogenic RGM. Like M. chelonae, M. abscessus is involved in a variety of different types of community-acquired infections. (a) Community-acquired disease. Pulmonary disease. Pulmonary disease accounts for most clinical isolates of this species . According to Griffith et al. , the majority (82%) of the 146 disease-associated pulmonary RGM isolates identified to species over a 15-year period by a Texas reference laboratory were M. abscessus. In patients with M. abscessus pulmonary disease, underlying diseases included bronchiectasis, CF , gastroesophageal disorders, and prior granulomatous disease such as sarcoidosis or tuberculosis. The analysis by Griffith et al. of M. abscessus pulmonary disease emphasized striking similarities to pulmonary M. avium complex lung disease of the type known as nodular bronchiectasis. The latter presents as an indolent course, occurs predominantly in elderly nonsmoking female patients, and exhibits a possible geographic disposition (i.e., southern coastal states have 69 to 75% of the cases). These similarities suggest a common pathogenicity or host susceptibility . Like patients infected with M. avium complex, most patients with pulmonary disease due to M. abscessus have underlying bronchiectasis of a type known as nodular bronchiectasis . Approximately 20% of patients with M. abscessus infection will also develop infection or disease due to M. avium complex , again emphasizing similar if not identical risk factors. There is controversy about whether M. abscessus can be a "colonizer" in the lungs. These authors believe that true colonization does not exist and that patients with minimal symptoms just have minimal disease. Repeated isolation of M. abscessus from the respiratory tract is usually associated with significant lung disease. Pulmonary disease in patients with CF. M. abscessus patients with underlying CF deserve some special comments. The recovery of M. abscessus from the respiratory tracts of patients with CF is being noted with increasing frequency. Patients with CF are predisposed to airway and parenchymal infections for several reasons, including the nature of CF disease and the usual associated bronchiectasis . The primary risk factor that makes patients with CF more susceptible to mycobacterial disease is thought to be bronchiectasis. Evaluation of the significance of the mycobacterial infection can be complicated because isolation of other organisms such as Pseudomonas aeruginosa often makes isolation and interpretation of the clinical significance of the RGM difficult. After M. avium complex, M. abscessus is the second most common species of nontuberculous mycobacteria recovered from respiratory specimens in patients with CF . Lung transplantation may be considered a therapeutic option in some CF patients. However, the posttransplantation immunosuppressive therapy increases the risk of both the development and the dissemination of nontuberculous mycobacterial infections. Patients with CF and M. abscessus lung disease carry the risk of developing disseminated infections, including cervical adenitis, following transplantation . Extrapulmonary disease. After M. fortuitum, M. abscessus is the second most common RGM species in clinical specimens; it also produces a wide variety of extrapulmonary diseases. Wallace et al. studied a series of 59 nonrespiratory isolates belonging to the M. chelonae-abscessus group and found that M. abscessus cases outnumbered M. chelonae cases more than 2:1 (30 and 12 cases, respectively). Among the 30 cases of nonpulmonary disease caused by M. abscessus, 43% were postsurgical or postinjection wound infections, 23% were localized community-acquired wound infections, 20% were disseminated cutaneous infections, and 13% were miscellaneous types of infections including keratitis and prosthetic valve endocarditis. Of the 23% of nonpulmonary disease cases resulting in localized infection, most characteristically developed following a break in the skin surface and subsequent direct contact with contaminated water or soil. Localized trauma with a resulting pyogenic abscess is sometimes followed by a sporotrichoid appearance of ascending lymphadenitis predominantly in immunocompromised patients . Other examples of localized M. abscessus wound infections include a soft tissue infection of the cheek following an insect bite and a case of vertebral osteomyelitis . (b) Health care-associated disease. M. abscessus and M. fortuitum are the most common mycobacterial species causing nosocomial disease, especially sporadic and clustered outbreaks of surgical wound infections. As noted above , surgical wound infections represented 43% of clinical cases of nonpulmonary infections due to this species. Disease outbreaks have been described after augmentation mammaplasty, facial plastic surgery, cardiac surgery, injections of alternative medicines, steroid injections, and miscellaneous types of surgery . In the largest outbreak of RGM-mediated postinjection abscesses, which occurred in an alternative medicine clinic in Colombia, South America , 205 (59%) of 350 of patients developed localized cutaneous abscesses or cellulitis due to M. abscessus. Another large M. abscessus outbreak in the United States resulted from the injection of an unlicensed product sold as adrenal cortex extract (ACE) . Of 140 persons known to have received the ACE injections, 87 subjects (62%) from 16 states were identified as infected. M. abscessus was cultured from seven vials of ACE, six of which were unopened. Isolates from both patients and opened and unopened vials of ACE were typed by MEE and pulsed-field gel electrophoresis and shown to be identical . Pseudo-outbreaks related to contaminated bronchoscopes have also been attributed to M. abscessus . Although disseminated M. abscessus disease is relatively unusual, it is serious. Most cases have occurred in chronically immunosuppressed patients receiving corticosteroids, and the disease has no apparent portal of entry. The disease presents as multiple draining cutaneous nodules, usually involving the lower extremities. Patients with disseminated infection have rarely included detectable bacteremia and endocarditis , and these cases can occur as a complication of localized infections. This is especially true in patients on hemodialysis. Bolan et al. reported 25 infections due to M. abscessus in a hemodialysis center in Louisiana (see the section on M. mucogenicum [above] for details). Nine of these patients had widely disseminated disease. Subsequent molecular studies using random amplified polymorphic DNA-PCR showed that the M. abscessus strains from the water supply and the clinical isolates were identical . This hemodialysis outbreak not only served to show the potential virulence of RGM disease in this setting but also pointed out the relative resistance of these organisms to commonly used disinfectants, a fact which increases the risk of health care-associated infections. In the Louisiana outbreak, investigators discovered that formaldehyde concentrations lower than 2% were used in disinfecting the reusable hemodialyzers. Failure to maintain a 2% concentration probably played a large role in this outbreak because this concentration had been previously established as the minimum concentration to which M. abscessus was susceptible in vitro . Five years later, Lowry et al. reported M. abscessus infection in five patients receiving dialysis with reusable dialysis tubing at another outpatient hemodialysis clinic. Again, the disinfectant used (2.5% Renalin) appeared to play a role, since at this concentration it did not completely kill the M. abscessus recovered from the patients and from the dialyzers that were manually reprocessed . (iii) M. immunogenum. | In 2000, Moore et al. , described an outbreak of hypersensitivity pneumonitis among workers in an industrial plant that was undergoing extensive remodeling and renovation. The workers utilized cutting, drilling, and grinding machines and worked with a semisynthetic metalworking fluid that was sprayed on the machines to keep them cool. Part of the outbreak investigation involved performance of cultures of the metalworking fluid for mycobacteria. Twenty- five isolates were recovered from different samples throughout the plant that were similar to M. chelonae-abscessus complex but with a unique hsp65 PRA pattern. This finding launched a search for other M. chelonae-abscessus-like RGM isolates with the same RFLP pattern. Isolates with this PRA pattern were identified from unrelated nosocomial pseudo-outbreaks involving contaminated endoscopes and from patients with serious infections. Although these strains exhibited overlapping biochemical and HPLC features with M. chelonae and M. abscessus, they differed from clinical and reference strains of both these species and most isolates had a unique susceptibility pattern of resistance to both cefoxitin and tobramycin . Molecular examination that included DNA homology studies showed that these isolates belonged to a separate species, which has been proposed as M. immunogenum . M. immunogenum organisms are able to grow and survive in degraded metalworking fluid , although it has not yet been established whether these organisms can metabolize any of the constituents of the fluid or additive materials for nutrition. The presence of other microorganisms (especially aerobic gram-negative bacilli) in degraded metal-grinding fluids and the use of biocides probably facilitates fluid degradation and subsequent growth of this species . (a) Clinical disease. In the only detailed study of clinical disease, 11 isolates of M. immunogenum were identified from patients . Three came from cultures of blood from patients with catheter- or pacemaker-related sepsis; two came from cutaneous cultures of samples from a liver transplant recipient and an infant with severe combined immunodeficiency syndrome with disseminated cutaneous infections; two came from catheter exit sites; and one each came from fluid a septic joint in a hand, bronchoalveolar lavage fluid from a patient with chronic pneumonia, a cornea from a patient with suspected keratitis, and urine from a patient with an unknown diagnosis . Additionally, two pseudo-outbreaks have been reported from Kentucky and Missouri, involving contaminated automated bronchoscopic washing machines which ultimately led to contaminated bronchoscopes , retrospect, these infections were found to be due to M. immunogenum . M. smegmatis Group | The M. smegmatis group currently is composed of M. smegmatis sensu stricto and the recently described M. wolinskyi and M. goodii . For purposes of this study, isolates identified specifically as one of the three recent taxonomic groups are referred to by their current species name(s). Strains not so recognized are referred to as the M. smegmatis group. Characteristic features of the M. smegmatis group include a negative 3-day arylsulfatase, growth at 45C, a positive nitrate reductase, a positive iron uptake, often a very smooth colony type, utilization of mannitol, inositol, and sorbitol as carbon sources, and a unique characteristic PRA pattern of the Telenti fragment of the hsp65 gene with BstEII . A late (7- to 10- day) yellow to orange pigmentation (most but not all isolates of M. smegmatis sensu stricto and M. goodii; the M. wolinskyi isolates are nonpigmented) on Middlebrook 7H10 agar is often seen. Isolates of some of these taxa have been recovered from the environment, and all have been recovered from patients, most of whom had clinical disease . One important distinguishing feature of isolates of the M. smegmatis group, in contrast to the M. fortuitum group and the M. chelonae-abscessus group, is their general lack of susceptibility to the new macrolides, including clarithromycin . Since clarithromycin has been considered the cornerstone of antimicrobial therapy for RGM disease, it becomes vital to identify RGM isolates to exclude groups like the M. smegmatis group and the M. fortuitum third biovariant complex sorbitol-positive group, which are intrinsically resistant to this class of drugs. Historical perspective. | The M. smegmatis group, first isolated by Lustgarten in 1885, was named for the genital secretions (smegma) from which it was recovered in a patient with a penile ulcer . The first well-described case of human disease caused by the M. smegmatis group involved the lungs and pleura of a patient with underlying exogenous lipoid pneumonia and was reported less than 15 years ago . Type of disease. | (i) Community-acquired disease. | The first series of clinical patients was reported by Wallace et al. in 1988 when they characterized 22 clinical isolates. The authors noted that the isolates were heterogeneous, and fell into three groups with different antibiotic susceptibility patterns. Later, in 1999, these three groups were studied in greater detail, including DNA homologies, and were found to be three distinct species: M. smegmatis sensu stricto, M. wolinskyi, and M. goodii . The last two names honored Emanuel Wolinsky and Robert Good, two early leaders in the field of nontuberculous mycobacteriology . The three species are separated with approximately 90% accuracy on the basis of tobramycin susceptibility. M. smegmatis sensu stricto is tobramycin susceptible (MIC, <=1mug/ml; agar disk diffusion zone, >30 mm). M. goodii has intermediate susceptibility to tobramycin (MIC, 2 to 8 mug/ml; agar disk diffusion zone, 11 to 30 mm), and M. wolinskyi is resistant to tobramycin (MIC, >8 mug/ml; agar disk diffusion zone, <=10 mm) . The HPLC patterns produced by M. smegmatis sensu stricto, M. wolinskyi, and M. goodii can be differentiated from those produced by members of the M. fortuitum group (; K. C. Jost, Jr., S. H. Chiu, R. B. Dunlap, L. B. Elliott, B. A. Brown, V. A. Steingrube, R. W. Wilson, and R. J. Wallace Jr., Abstr. 99th Gen. Meet. Amer. Soc. Microbiol. 1999, abstr. U-36, 1999), and the three species have different patterns. The overlap between patterns of all mycobacterial species makes identification of the individual M. smegmatis species difficult, however, when evaluating individual clinical isolates. The most accurate separation of the three species within the M. smegmatis group is achieved by molecular techniques including PRA of the Telenti fragment of the hsp65 gene and 16S rRNA gene sequence analysis . Until a case of lung disease proven by lung biopsy was reported in 1986 , the M. smegmatis group was considered to be an environmental saprophyte of no clinical significance. Community-acquired disease due to M. smegmatis group is now known to involve cellulitis, localized abscesses, and/or osteomyelitis of a wound site following a traumatic event. Newton et al. , reported that the M. smegmatis group was the causative agent of two cases of infection following motor vehicle accidents, with cellulitis and extensive soft tissue and periosteal necrosis evident at the time of surgical debridement. A few cases of lipoid pneumonia (pneumonia resulting from inhalation or aspiration of lipid-containing medicinals or food particles) with secondary mycobacterial infection have also been reported to be caused by the M. smegmatis group . No case of disseminated cutaneous disease due to the M. smegmatis group has been reported to date. (ii) Health care-associated disease. | Health care-associated infections involving the M. smegmatis group have included sporadic cases of catheter sepsis, infected pacemaker site, sternal wound infection with possible osteomyelitis following cardiac surgery, and infections following plastic surgery (breast reduction surgery and a face-lift) . No health care-associated disease outbreak or pseudo-outbreak due to the M. smegmatis group has yet been reported. Geography. | Isolates of the M. smegmatis group have a wide geographic distribution. Isolates have been recovered in the United States from Texas, Alabama, California, Florida, Illinois, Indiana, Massachusetts, Minnesota, Mississippi, Missouri, North Carolina, Ohio, Oklahoma, South Carolina, Utah, and Wyoming . Outside the United States, isolates have been reported from Australia, Russia, Canada, and Switzerland . Individual taxa. | (i) M. smegmatis sensu stricto. | In 1988, Wallace et al. , reported a series of 21 patients with infections due to the M. smegmatis group. In the latter taxonomic study of these isolates published in 1999 , 52% of the 21 original clinical isolates matched the type strain and additional ATCC reference strains of M. smegmatis and hence were renamed M. smegmatis sensu stricto. With expansion of the number of clinical isolates which met the criteria for the M. smegmatis group to 71, 49% were M. smegmatis sensu stricto . These isolates had a unique mycolic acid pattern and were susceptible to tobramycin agar disk diffusion (zones, >30 mm with a 10 mug commercial disk). Additionally, 16S rRNA gene sequence analysis and PRA of the 439-bp hsp65 gene sequence were unique to this species . Isolates of M. smegmatis sensu stricto have been reported from several states including Florida, Wyoming, South Carolina, Texas, Mississippi, and Illinois and, outside the United States, in Australia . M. smegmatis sensu stricto has been incriminated in community-acquired cases of lymphadenitis, cellulitis, osteomyelitis, wound infections and, rarely, respiratory disease, usually associated with exogenous lipoid pneumonia . It has been recovered from health care-associated infections, including sternal wound sites following cardiac surgery, bacteremia from intravenous catheter placement, and breast abscess following augmentation mammaplasty . (ii) M. goodii. | As mentioned previously, the Wallace study in 1988 launched a second study, published in 1999 , which identified 8 of 21 isolates reported in the 1988 publication and 20 new isolates in the second study which proved to be M. goodii. These 28 isolates represented 39% of the 71 total isolates of the M. smegmatis group studied. They had a mycolic acid pattern that differed from the other two species in the group, were intermediately susceptible to tobramycin by agar disk diffusion (zones, 11 to 30 mm), had a unique 16S rRNA gene sequence, and a unique PRA pattern . M. goodii is the second most frequently isolated species within the M. smegmatis group. Isolates of M. goodii have been recovered from California, Texas, Florida, Alabama, Minnesota, Utah, Oklahoma, Missouri, Indiana, Ohio, North Carolina, and Massachusetts . Outside of the United States, isolates have been reported from Russia , Australia and Canada (Sylvia Chomyc, Provincial Laboratory of Public Health, Alberta, Canada, personal communication). Isolates of M. goodii have been recovered from cases of cellulitis, bursitis, and osteomyelitis after open (compound) fracture or penetrating trauma . A few cases of respiratory disease due to M. goodii have been reported. Most have been associated with underlying exogenous lipoid pneumonia with pulmonary infiltrates, similar to M. smegmatis sensu stricto . M. goodii has been involved in several types of sporadic health care-associated disease, including bacteremia with catheter sepsis , cardiac bypass infection with osteomyelitis, infected pacemaker site, and infection following breast reduction surgery . (iii) M. wolinskyi. | Of 21 isolates from the 1988 study by Wallace et al. , 2 matched 6 other isolates in the second study in their 16S rRNA gene sequence patterns, PRA patterns, and mycolic acid patterns. These eight strains collectively were named M. wolinskyi and represented 11% of the total 71 clinical isolates identified as belonging to the M. smegmatis group. Isolates of M. wolinskyi have been recovered from Texas, California, and Switzerland . Clinical histories were available for seven of the eight reported isolates of M. wolinskyi . Four (57%) of these seven isolates were associated with community-acquired infections that included cellulitis and osteomyelitis following a traumatic event and cellulitis and localized abscess following a motor vehicle accident . The remaining three isolates (43%) of M. wolinskyi were associated with sporadic health care-associated infections. These included sternal wound infection and osteomyelitis following cardiac surgery, surgical wound infection following facial plastic surgery , and an infected arteriovenous shunt in a patient on hemodialysis. Other nonpigmented RGM of uncertain clinical (human) significance. | M. chitae was described by Tsukamura in 1966 . Four strains of this species were recovered from soil samples collected near manure heaps. The organisms are not known to be associated with disease. M. agri was described by Tsukamura . The only isolate was recovered from an alkali-treated soil sample and was not known to be associated with disease. Differential characteristics of both these species may be found in Bergey's Manual of Systematic Bacteriology, vol. 2 . M. porcinum also was described by Tsukamura et al. in 1983, when the authors characterized 10 strains recovered from pigs with submandibular lymphadenitis. TREATMENT OF INFECTION : Antimicrobial Treatment | General. | Antimicrobial therapy for RGM, unlike chemotherapy used for most slowly growing mycobacterial diseases, may vary depending on the nature of the disease. For example, single-drug therapy for localized or minor disease due to RGM is often sufficient, with minimal risk of development of mutational drug resistance. In contrast, disseminated cutaneous disease and pulmonary disease usually require multiple antimicrobials, including both intravenous and oral medications. The newer oral antimicrobials linezolid and gatifloxacin offer great promise as alternatives to injectable medicines, but clinical experience with them is very limited (; B. A. Brown-Elliott, R. J. Wallace Jr., and C. J. Crist, Program Abstr. 41st Intersci. Conf. Antimicrob. Agents Chemother., abstr. E-723, 2001). Table summarizes antimicrobial treatment of the frequently encountered pathogenic RGM. M. fortuitum group. | Since the M. fortuitum group is much less drug resistant than M. abscessus and M. chelonae, treatment has been much easier and generally more effective. The usual therapeutic drugs recommended for infections with the former group include amikacin, cefoxitin, imipenem, sulfamethoxazole, and fluoroquinolones . Doxycycline has proven to be an effective oral agent , but only about 50% of isolates of M. fortuitum are susceptible to <=4 mug/ml . In vitro susceptibility to clarithromycin within the M. fortuitum group is variable . Approximately 80% of isolates of M. fortuitum are susceptible to <=4 mug/ml. Isolates of (proposed species) M. bonickei (sorbitol-negative third biovariant group) and M. peregrinum are all susceptible to <=4 mug/ml, while isolates of the M. fortuitum third biovariant sorbitol- positive group (e.g., M. houstonense [proposed] and M. mageritense), are uniformly resistant to clarithromycin. Wallace et al. have recommended a sulfonamide or doxycycline as acceptable agents for oral monotherapy of localized wound infections caused by the M. fortuitum group (based on in vitro susceptibilities). A major problem associated with quinolone monotherapy, however, was the development of mutational resistance with treatment failure or relapse . Therefore, if a quinolone is used for therapy, an additional antimicrobial agent is usually necessary. Acquired mutational resistance of the M. fortuitum group to clarithromycin, doxycycline, and trimethoprim-sulfamethoxazole has not been reported, and we have not observed it (unpublished observations). For serious disease with M. fortuitum, the aminoglycoside amikacin, combined with a beta-lactam (cefoxitin or imipenem) or a quinolone (ciprofloxacin or l-ofloxacin), has been used for initial therapy . Linezolid offers excellent potential as an oral or intravenous (i.v.) therapeutic agent, since 96% of isolates of all members of the M. fortuitum group (M. fortuitum, M. peregrinum, and the proposed new species M. houstonense and M. bonickei) are susceptible or intermediate in vitro . However, there is no reported clinical experience with this drug for the M. fortuitum group. M. chelonae-abscessus group. | The treatment of localized infections due to M. chelonae or M. abscessus is currently managed by using the newer macrolide clarithromycin as the cornerstone of therapy . Azithromycin appears to work as well, but there is much less clinical experience with it. Acquired mutational resistance to the macrolides has not been observed when treating localized infections . However, more serious disease should be treated, for at least the first 2 weeks, with clarithromycin in combination with one of the injectable agents . For serious, extensive extrapulmonary disease or disseminated infections involving M. chelonae, the injectable agents tobramycin plus imipenem have been used for the first 2 to 6 weeks in combination with clarithromycin to avoid or minimize the development of drug resistance to the macrolide . For M. chelonae, cefoxitin is not used, since isolates of M. chelonae are resistant to this agent (MIC, >128 mug/ml), and tobramycin is preferred to amikacin because of its greater in vitro activity . Potential alternative oral agents are available which make injectable therapy less mandatory than in years past. The tetracycline analogues minocycline and doxycycline are effective against about 20% of M. chelonae . Newer potential oral agents being considered for use in combination with clarithromycin are the 8-methoxyfluoroquinolones (e.g., gatifloxacin) and/or linezolid. However, there is little experience with these newer agents (see "Newer drugs" below). For serious, extensive extrapulmonary disease or for disseminated disease due to M. abscessus, amikacin plus cefoxitin or imipenem is used for the first 2 to 6 weeks in combination with clarithromycin. Approximately 90% of M. abscessus isolates are susceptible or intermediate to amikacin (32 mug/ml) and cefoxitin (64 mug/ml) and about 50% are susceptible or intermediate to imipenem (8 mug/ml) and linezolid (16 mug/ml) (,). Whether linezolid could be used for this species has not yet been studied. For patients with pulmonary disease due to M. abscessus, treatment options are few. Generally, the only oral antimicrobials to which M. abscessus is susceptible are the new macrolides clarithromycin and azithromycin. M. abscessus also is usually susceptible in vitro to amikacin, cefoxitin, and imipenem , i.v. medicines whose administration is limited by cost, toxicity, and the need for frequent administration (the beta-lactams). (As with extrapulmonary disease, the potential usefulness of linezolid has not been studied.) The optimal treatment for M. abscessus lung disease is probably combination therapy using a macrolide and parenteral antibiotics. Most patients improve when given parenteral therapy for 2 to 4 weeks, but few can tolerate these medicines for longer periods. Unfortunately, to date, permanent sputum conversion along with permanent symptomatic improvement when using the combination of oral clarithromycin and parenteral agents for M. abscessus is rare. Only surgical resection (with localized disease) of the lung has produced long-term conversion of sputum cultures to negative and complete resolution of symptoms in patients with M. abscessus infectious . Better drugs are clearly needed for this species. M. smegmatis group. | Treatment of disease has generally involved the same drugs as for treatment of the M. fortuitum group, with doxycycline and trimethoprim-sulfamethoxazole being the most common oral agents. Injectable agents have usually included amikacin and/or imipenem. All isolates of the M. smegmatis group are uniformly susceptible to sulfonamides, doxycycline, imipenem, and amikacin . They exhibit intermediate susceptibility to the older fluoroquinolones (ciprofloxacin and ofloxacin) and variable susceptibility to cefoxitin and clarithromycin. The M. smegmatis group is the only nonpigmented RGM pathogenic for humans that is susceptible to ethambutol. With this exception, none of the other first-line antituberculosis drugs is efficacious against any of the RGM . Newer drugs. | Several new alternative antimicrobials have emerged for the management of RGM disease, with the most active in vitro being linezolid, the 8-methoxyfluoroquinolones moxifloxacin and gatifloxacin (; B. A. Brown-Elliott, R. J. Wallace Jr., and C. J. Crist, Program Abstr. 41st Intersci. Conf. Antimicrob. Agents Chemother., abstr. E-723, 2001), and tigecycline (formerly GAR-936) (R. J. Wallace, Jr., B. A. Brown- Elliott, C. J. Crist, L. Mann, and R. W. Wilson, submitted for publication). The oxazolidinone linezolid (Zyvox; Pharmacia) offers great potential in the treatment of RGM disease, primarily against isolates of the M. fortuitum group, the M. smegmatis group, M. mucogenicum, and M. chelonae . Wallace et al. reported that 96% of 74 isolates of M. fortuitum group were susceptible to <=16 mug of linezolid per ml, with a modal MIC of 4 mug/ml. Most (94%) of 50 isolates of M. chelonae had linezolid MICs of 4 to 16mug/ml, with a mode of 8 mug/ml. In contrast, more isolates of M. abscessus were resistant, with only 24 of 98 (48%) isolates being susceptible or intermediate (MIC, <=16mug/mL) . A single patient with disseminated cutaneous disease due to M. chelonae resistant to clarithromycin was successfully treated with a 3-month course of linezolid monotherapy . The recently FDA-approved 8-methoxyfluoroquinolones gatifloxacin and moxifloxacin have greater activity against the M. fortuitum group and M. chelonae than do the older fluoroquinolones such as ciprofloxacin. In a study by Brown-Elliott et al. (41st ICAAC), 100% of 26 isolates of M. fortuitum group and 1 isolate each of M. smegmatis, M. wolinskyi, M. goodii, and M. immunogenum were susceptible, with gatifloxacin and ciprofloxacin MICs for 90% of isolates an (MIC90) of <=0.12 L and 1 mug/ml, respectively. In the same study, 96% of 27 isolates of M. chelonae were susceptible or intermediate to gatifloxacin (MIC, <=4 mug/ml), in contrast to only 8% susceptible or intermediate to ciprofloxacin (MIC, <=2 mug/ml). No isolates of M. abscessus were susceptible or intermediate to ciprofloxacin, and only 10% (of 20 isolates) were susceptible or intermediate to gatifloxacin. The new glycylcycline (a tetracycline analog) tigecycline (formerly Gar-936) is in phase III clinical trials. In vitro studies with RGM have shown all pathogenic species of the nonpigmented RGM to be highly susceptible to this agent, with MICs of <=0.25 mug/ml (R. J. Wallace, Jr., et al., submitted). Duration of therapy. | Generally, the length of treatment with any of the current antimicrobials for most RGM skin, soft tissue, or bone disease has been 4 months for mild disease and 6 months for serious disease. Treatment with injectable agents is usually limited to the first 2 to 6 weeks of therapy, in an attempt to minimize cost and drug toxicity. The exception for total length of therapy is that pulmonary disease is usually treated for at least 12 months. However, microbiological cure has not been possible in most cases of pulmonary disease caused by M. abscessus, even after lengthy antimicrobial treatment . Surgical Treatment | Minor wound infections with RGM may resolve spontaneously or after surgical debridement. In an early study by Wallace et al. , 13 (17%) of 76 and 3 (6%) of 47 extrapulmonary cases of infection with the M. fortuitum group and the M. chelonae-abscessus group were successfully treated by surgery alone. The authors also reported that surgery alone for more serious wound infections, without antimicrobial therapy, was followed by healing but that relapse often occurred within 4 to 6 weeks of the surgery. In contrast, surgical excision and/or debridement of the wound site combined with appropriate antimicrobial therapy resulted in healing without relapse . Likewise, when a foreign body such as breast implant, percutaneous catheter, etc., is involved, removal of the foreign body appears to be essential to recovery . In a 1985 report , the success rate for the treatment of M. fortuitum wound infections was about 90% even though the macrolides, the newer quinolones, and imipenem were not yet available. In 1985, that was not the picture for wound infections due to M. chelonae or M. abscessus even with surgical intervention. Until the arrival of the newer macrolides (in approximately 1993), serious disease caused by M. abscessus or M. chelonae was difficult to treat, even with a combination of antimicrobials and surgery, and infection with M. chelonae-abscessus could persist for years . No highly active oral agents were then available for treatment. The available drugs (amikacin, tobramycin, and cefoxitin) were too toxic and/or too expensive to continue long enough to cure most serious infections, and their effects were worse than those of minor disease. The introduction of clarithromycin almost certainly improved the response rate for M. chelonae and M. abscessus and has decreased the need for routine surgery in patients infected by these species. In a series of two patients infected with the M. smegmatis group, Newton et al. expressed the opinion that extensive surgical debridement of soft tissue and bone followed by skin grafting has been necessary for cure of these organisms. We would agree with surgical debridement in patients with evidence of tissue necrosis or extensive disease, but we have rarely had to resort to skin grafting for the M. smegmatis group or other RGM. Lung infections caused by M. abscessus have remained difficult to treat and incurable with currently available drugs. Patients with localized unilateral disease, who were surgical candidates, were treated with amikacin and cefoxitin before surgical resection was performed. Unfortunately, most patients with M. abscessus lung disease have bilateral disease and are not surgical candidates. CLINICAL FEATURES OF INFECTIONS : Posttraumatic Wound Infections | The M. fortuitum group and the M. chelonae-abscessus group are most commonly associated with skin and soft tissue infections, usually following some type of penetrating trauma. Accidental puncture wounds, especially due to stepping on nails contaminated with soil or water, are often seen in this group of infections . Of the 34 RGM infections discussed by Wallace et al. , 12 (35%) directly involved the foot and 9 of these 12 (75%) were associated with stepping on a nail. In a study exclusively of M. chelonae, localized cellulitis, subcutaneous abscess, or osteomyelitis usually followed similar types of traumatic injury . Use of corticosteroids predisposed this group to this type of infection. (Among the pathogenic RGM species, this species is the one most commonly associated with chronic corticosteroid use.) Other species of RGM which have been associated with posttraumatic wound infection include M. wolinskyi and M. goodii . Brown et al. reported that four (50%) of eight isolates of M. wolinskyi were associated with cellulitis following a traumatic event. Two of the four patients also progressed to osteomyelitis. Brown et al. reported that 7 (25%) of 28 patients had infections with M. goodii that were associated with posttraumatic cellulitis. Of the seven patients, four were thought to have complicating osteomyelitis. For mild localized posttraumatic wound infections, monotherapy with an oral agent given for 4 to 6 months has been very successful. For extensive disease, surgical debridement combined with initial combination drug therapy based on in vitro susceptibilities followed by oral therapy to complete 6 months of treatment has also proven effective. Bone and Joint Infection | Bone and/or joint disease is not an infrequent complication of infection with the RGM. As with bacterial disease, osteomyelitis may follow open bone fractures, puncture wounds, and hematogenous spread from another source. The most common setting is an open fracture of the femur, often followed by orthopedic correction that includes open reduction and internal fixation. The most common pathogen in this setting is M. fortuitum and the M. fortuitum third biovariant complex (proposed species M. houstonense and M. bonickei) (; Schinsky et al., submitted). Osteomyelitis due to M. chelonae or M. abscessus may be seen in patients receiving long-term steroid therapy. Maxson et al. reported a case of osteomyelitis due to M. abscessus in a patient with systemic lupus erythematosus who had been treated with prednisone for 10 years. The patient presented with lumbar pain and an area of caseous abscess formation, and compression fractures were discovered after several incorrect diagnoses were made. Operative cultures revealed M. abscessus. Previously, Wallace et al. had reported at least two cases of multifocal osteomyelitis with disseminated M. chelonae cutaneous disease in patients receiving corticosteroids . In a later series involving 76 patients, Wallace reported another 14 cases of osteomyelitis. A case of vertebral osteomyelitis in a patient with a history of intravenous drug abuse was described by Sarria et al. computed tomography guided aspiration yielded a pure culture of M. abscessus. Although RGM rarely cause vertebral osteomyelitis. Sarria and colleagues identified 15 cases and found clinical information on six of the cases . Four of the six cases of RGM vertebral osteomyelitis were due to M. fortuitum . All of the patients had some type of underlying condition such as systemic lupus erythamatosus, chronic granulomatous disease, achalasia, back injury, mental retardation, and/or IV drug abuse. Osteomyelitis secondary to a puncture wound is probably the second most common cause of osteomyelitis after open fractures and has been reported with M. peregrinum and M. fortuitum . Both et al. reported two cases of M. fortuitum infections in patients with septic arthritis associated with joint prostheses. Both patients responded favorably to surgical drainage and removal of the prosthesis. Later, Herold et al. reviewed five other cases of M. fortuitum involved in prosthetic knee and joint infections. Brown et al. recently reported that 13 (36%) of 36 patients with infection caused by two new species (M. goodii and M. wolinskyi, belonging to the M. smegmatis group) were diagnosed with osteomyelitis. These new species were more often associated with osteomyelitis than were infections involving M. smegmatis sensu stricto. Generally, cures of osteomyelitis due to RGM have been accomplished by surgical wound debridement combined with drug therapy (based on in vitro susceptibilities of the isolate) given for a minimum of 6 months . Surgical debridement is of greater benefit to patients with disease due to the M. chelonae-abscessus group since drug therapy is so much more difficult. Postsurgical Wound Infections | The RGM have been recognized for over 20 years as causative agents of sporadic nosocomial or health care-associated infections including disease that involves renal dialysis; punch biopsy surgery; augmentation mammaplasty; other forms of plastic surgery including face-lifts and liposuction; sternal wound infections following cardiac surgery; and postinjection abscesses . In one series of 44 postsurgical RGM infections, 31 cases involved M. fortuitum and 13 cases involved the M. chelonae-abscessus group . RGM infections following augmentation mammaplasty are well known. In a 3-year time span, Clegg et al. collected 17 cases of infections with M. fortuitum group and M. chelonae-abscessus group following implantation of breast prostheses. All of these infections remained localized, and most were unilateral. In 1989, Wallace et al. reported an additional 37 cases of RGM wound infections that occurred after augmentation mammaplasty. Most of these infections (70%) were caused by M. fortuitum. Successful therapy of these infections involved removal of the infected implant, 6 months of drug therapy, then reimplantation of the breast if desired. Shorter causes of therapy (e.g., 4 months) might be effective but have not been studied. In addition, RGM are well recognized as a cause of cardiac surgical infections. These may involve the sternal wound site, the saphenous vein site, or even an inserted prosthetic valve. Of 36 isolates in one series of patients with sporadic infection, 26 were M. fortuitum and the remaining 10 were placed in one of several groups: the M. chelonae-abscessus group (4 cases), M. smegmatis group, later identified as M. wolinskyi (4 cases), and 2 cases of pigmented RGM . Thus, two (25%) of eight isolates of the newly proposed M. wolinskyi were seen with postsurgical infections . At least 5 (18%) of 28 other cases of the newly proposed M. goodii were associated with postsurgical infections including infected pacemakers, breast reduction surgery, and cardiac surgery . Additionally, one case of M. abscessus associated with an infected pacemaker was reported by Cutay et al. . Treatment of sternal wound infections (osteomyelitis) has invariably included surgical debridement, an initial period of i.v. combination antibiotics that usually includes amikacin and a beta-lactam (cefoxitin or imipenem), and subsequent oral therapy (when possible) to complete at least 6 months of therapy . Patients with M. abscessus sternal wound infections often undergo removal of the sternum (sternectomy) because drug therapy for this species is so much more difficult than for the M. fortuitum group or the M. smegmatis group. Treatment of prosthetic valve endocarditis requires removal of the infected valve. RGM infections in hemodialysis patients have also been recognized. Bolan et al. reported on 27 of 140 patients infected with RGM in a center in Louisiana. Three had soft tissue infections, 9 had disseminated disease, 1 had an access-graft infection, and 14 had bacteremia alone. Of 26 identified isolates, 25 (96%), were M. abscessus, and 1 (4%) was identified as M. mucogenicum. Environmental cultures of the water system showed heavy contamination with RGM. Later, Lowry et al. , described five patients with M. abscessus who were dialyzed at an outpatient clinic in California. Four of the five patients had arteriovenous graft infections, and two of them died during antimicrobial therapy. These authors stressed the need for complete graft removal to ensure total recovery. These studies antedated the availability of many of the newer antibiotics, so optimal drug therapy for these types of infection is not known. Postinjection Abscesses | RGM also have been associated with a large number of outbreaks of postinjection abscesses. Two of the largest case studies recently were published. The first large outbreak in Baranquilla, Colombia, involved the isolation of M. abscessus subsequent to local injections of lidocaine in a physician's office. Over a 5-month period, 350 of about 2,000 patients (18%) developed localized abscesses or cellulitis. Therapy with a combination of surgical excision and 3 to 6 months of clarithromycin administration was successful in 95% of the treated patients. Fewer than one-third of the patients who received either surgical therapy alone or clarithromycin alone were cured. An identical outcome was also seen in an earlier single case of M. chelonae infection following excision and closure surgery for a basal cell carcinoma (,). Galil et al. reported the largest outbreak of M. abscessus infection in the United States, which followed injections of ACE. From 1995 to 1996, 87 persons were identified with postinjection abscesses attributable to M. abscessus from contaminated ACE. Similar to the Colombia outbreak, surgical excision or drainage combined with clarithromycin therapy for 3 to 6 months appeared to be the optimal therapy for M. chelonae or M. abscessus postinjection abscesses. Catheter-Related Infections | Catheter-related infections are a relatively common form of RGM disease and the most common form of health care-associated disease. The most common mycobacterial pathogen is M. fortuitum. However, M. chelonae, M. abscessus, M. immunogenum, the M. smegmatis group, M. peregrinum, M. mucogenicum, and some pigmented species have also been associated with catheter-related infections . Most clinically significant cases of M. mucogenicum disease have involved catheter-related infections . The organisms may produce exit site infections, tunnel infections (where the catheter is tunneled under the skin), and/or bacteremia. Rarely, hematogenous dissemination to sites such as the lungs or liver (granulomatous hepatitis) also occurs . In a large study by Wallace et al. of M. chelonae infections, 12 of 100 cases (12%) were catheter related. Immunosuppressive disorders such as transplantation, leukemia, and corticosteroid therapy are most commonly associated with RGM catheter-related infections . Most commonly, long-term central indwelling catheters such as Hickman or Broviac catheters are incriminated . Examples of these infections are provided by Burns et al. , who described two cases of RGM catheter infection. The first was a 5-year-old boy with end-stage renal disease who had received peritoneal dialysis for 4 years. M. chelonae was cultured from a peritoneal fluid aspirate. The peritoneal dialysis catheter was removed, and treatment with amikacin, cefoxitin, and clarithromycin was instituted. After 3 weeks, cefoxitin and clarithromycin administration was stopped and the child was treated successfully for 3 months with intravenous amikacin. The second case involved a 7-year-old girl with a congenital defect in her gastrointestinal system that had required total parenteral nutrition since birth. She had utilized multiple subclavian Hickman catheters over the years. She ultimately developed a tunnel site infection associated with one of the catheters. The Hickman catheter was removed with implantation of a Broviac catheter but with no resolution of symptoms. Cultures made from the catheter site grew out M. chelonae. The infected tissue in the chest wall was excised down to the muscle. A new central line was placed using a new tunnel and exit site. The patient recovered after the debridement and administration of multiple antibiotics over 6 months, including ciprofloxacin, amikacin, and clarithromycin for 2 months followed by an additional 4 months of clarithromycin therapy. Successful therapy of these catheter-related infections involves removal of the catheter and antimicrobial therapy, usually for 2 to 4 months. Although disease due to M. fortuitum may resolve if the catheter is removed, reinsertion of another catheter in a similar location without drug therapy usually results in disease recurrence (as in the above case). Disseminated Cutaneous Disease | Disseminated cutaneous disease associated with RGM is unusual but serious. More than 90% of patients with disseminated cutaneous disease have identified risk factors such as chronic renal failure, renal transplantation, and, especially, a history of chronic corticosteroid therapy . Only rarely is the patient with disseminated cutaneous disease due to RGM also human immunodeficiency virus positive . Most documented cases of RGM disseminated cutaneous disease have been associated with the M. abscessus-chelonae group. In a series of 100 nonpulmonary isolates of M. chelonae, more than 50% were associated with disseminated cutaneous disease . Similar data for M. fortuitum and M. abscessus are much more limited. Of nine cases reported by Wallace et al. , six (67%) were caused by M. abscessus, two (22%) were caused by M. fortuitum, and one (11%) was caused by M. chelonae. Of these nine patients with disseminated RGM disease, four had multiple, recurrent episodes of chronic skin and soft tissue abscesses, usually involving the upper extremities , but did not appear systemically ill. The other five patients were acutely ill, and organisms were recovered from multiple sources including blood. Only two of the nine patients were not immunosuppressed, and one died . Later, it was noted that the development of clarithromycin mutational resistance with clinical relapses of M. chelonae-abscessus was rare except in the setting of disseminated disease . Tebas et al. reported a case of clarithromycin resistance leading to multiple relapses in a patient with disseminated disease with M. chelonae following heart transplantation. An unusual group of 16 cases of M. chelonae-abscessus disseminated disease that included chronic cervical lymphadenitis were reported from a university hospital in northeastern Thailand. Most patients also had serious infections with pathogens other than the RGM . The duration of symptoms in the 16 patients persisted from 1 month to 5 years, with a mean of 15.6 months, before diagnosis of the RGM was established. There were 12 patients who had other multiple organ systems involved including sinuses (6 patients), lungs (4 patients), liver (4 patients), spleen (3 patients), skin (3 patients), bone and joint (2 patients), and tonsils (2 patients). Eleven patients had reactive skin disease based on pathology . In the 1992 series of 53 patients with disseminated skin disease due to M. chelonae, most patients had multiple nodular, subcutaneous, draining lesions. Evidence of spread to other parts of the body was rare. Only three patients had positive blood cultures; no patients had any associated systemic disease. Unlike post-traumatic wound infections, in which a portal of entry for the infection is found, no source was evident in the patients. In addition, osteomyelitis was rare in this group, with only two patients having multifocal osteomyelitis of the extremities and draining lesions secondary to primary bone involvement. Disseminated disease with M. fortuitum, unlike the M. chelonae-abscessus group, is rare but does occur. An unusual case of disseminated M. fortuitum in a 76-year-old male with chronic lung disease, in which the RGM infection began in the lungs and spread to the bone and skin, was described by Burns et al. . Additionally, Horsburg and Selik reported five cases of disseminated M. fortuitum infection in AIDS patients. No further details were documented. Until 1979, only 11 cases of disseminated M. fortuitum disease were documented, and exact taxonomy of the RGM was not very good at that time, making it likely that some of these isolates may have actually been M. chelonae-abscessus . Treatment of disseminated cutaneous disease involves drainage of abscesses and, for the M. chelonae-abscessus group, the use of clarithromycin for at least 6 months. Wallace et al. , studied 14 patients with cutaneous infections due to M. chelonae, most of which were disseminated. Of 11 patients who completed therapy, all had complete resolution with no relapses within 6 months of clarithromycin monotherapy. Because of the risk of the development of clarithromycin resistance with monotherapy of disseminated disease (estimated to be about 10 to 20% in this setting), initial therapy for the first 3 to 6 weeks should include other drugs based on in vitro susceptibilities whenever possible. Pulmonary Disease | Approximately 80% of the chronic pulmonary disease caused by the RGM is due to M. abscessus . Although some heterogeneity occurs, the typical patient with M. abscessus lung disease is a nonsmoking female in her 60s, with disease symptoms that have existed for at least 2 years. Many patients have a diagnosis of bronchiectasis with chronic lung disease . The symptoms most frequently observed are cough and chronic fatigue, although sputum production, hemoptysis (coughing up blood), and weight loss also occur. As mentioned above, the cases are seen clustered in the southeast coastal United States . CF patients comprise about 10 to 15% of patients with M. abscessus lung disease . The CF patient is at risk for developing infections due to M. abscessus and M. avium complex because of existing obstructive airway disease and underlying bronchiectasis . They are most often identified because of episodic fever that is poorly controlled with anti-pseudomonal therapy. Treatment of M. abscessus lung disease remains a disappointment, since the use of amikacin, cefoxitin, and clarithromycin often produces clinical improvement but is insufficient to cure the disease. There is a desperate need for newer and better oral drugs. Linezolid offers potential to some patients, but the MICs are relatively high and there is little clinical experience. The new glycycycline, tigecycline, offers exciting potential, but is not yet FDA approved and is an i.v. drug. The remaining 20% of cases of lung infection not due to M. abscessus are due to the less commonly encountered M. smegmatis group and the M. fortuitum group. Of 28 cases due to the newly reported species M. goodii, 6 (21%) were pulmonary disease . M. fortuitum is also associated with pulmonary disease, with approximately 50% of cases associated with esophageal achalasia (a disease of severe dilation and loss of function of the esophagus), lipoid pneumonia, and diseases with chronic vomiting and aspiration . Rare cases of lung abscess and infections in CF patients as a result of M. fortuitum have also been seen, although accurate species identification for this species has been readily available only in recent years . Although M. fortuitum may be isolated frequently from respiratory specimens, it is usually a single positive culture and, compared to M. abscessus, only rarely is a significant pathogen . Treatment of M. fortuitum lung disease with such antimicrobials as the quinolones, sulfonamides, doxycycline, amikacin, and cefoxitin has usually been successful . Central Nervous System Disease | Central nervous system disease involving RGM is rare but serious. Most cases have been associated with M. fortuitum . The first report of central nervous system infection with M. fortuitum was in 1970. An 8-year-old boy was thought to have "aseptic meningitis" after routine cerebrospinal fluid (CSF) cultures were negative. However, after repeated spinal taps with negative cultures, surgery was performed and a vertebral abscess was discovered and drained. Cultures subsequently grew M. fortuitum. The patient was treated and recovered . Dalovisio et al. described a case of M. fortuitum meningitis in a child with a history of chronic otitis media who subsequently developed an obstructive hydrocephalus and had a ventriculoperitoneal shunt placed. His condition improved when amikacin injections were administered every 48 h directly into the ventricles through the shunt. Cultures of spinal fluid at 8 months were negative for M. fortuitum. The patient's illness resolved except for residual decreased hearing. Another case involved infection of a ventriculoatrial shunt inserted for spontaneous cerebral hematoma with obstructive hydrocephalus in a 60-year-old woman. After placement of the shunt, the patient had persistent fever and her CSF grew M. fortuitum. The patient's condition continued to deteriorate, in spite of therapy with amikacin and ofloxacin, until the shunt was removed . A study from India reviewed 50 cases of brain abscesses in which one specimen grew M. fortuitum . Details of this case were not given. Another isolation of M. fortuitum from spinal fluid occurred in a teenage boy who had been in a motor vehicle accident. During the accident, he sustained a wound in the sacral region, which harbored a foreign body. The patient ultimately responded to trimethoprim-sulfamethoxazole and surgical drainage . A case of M. fortuitum meningitis was reported in a patient with AIDS. M. fortuitum was recovered from skin biopsy specimens, bone marrow, and CSF. Unfortunately, therapy with amikacin and doxycycline was not successful, and the patient died . The only other RGM which have been reported to cause meningitis are M. mucogenicum and M. goodii. In a series of 20 cases of M. mucogenicum, 1 patient with AIDS had multiple CSF samples that were positive on smear and culture for M. mucogenicum. No other details were given . Similarly, M. goodii was recovered from the CSF of one patient but was of unknown clinical significance . Finally, Flor et al. recently reviewed the literature for cases of nontuberculous mycobacteria associated with meningitis and found that 12% of 52 isolates were identified as M. fortuitum. Treatment of CNS infections due to RGM has been difficult and often requires prolonged therapy (6 to 12 months) with multiple drugs. Miscellaneous Diseases | Various other diseases have been associated with the RGM. A few of these unusual situations are discussed here. Otitis media. | Lowry et al. reported an outbreak of 17 cases of otitis media caused by M. abscessus in two ear-nose-throat clinics. It is not clear if the index patient from the first clinic "spread" the strain into the other clinic and contaminated the equipment in the second clinic or if the strain was already present. However, all but 1 of the 14 isolates obtained from the 17 patients were aminoglycoside resistant, as were isolates obtained from the suction tubing and suction-sink water in both clinics. The resistance was secondary to a mutation in the 16S rRNA gene, which implies that the original outbreak isolate was from a patient with chronic otitis rather than from the environment. In addition to surgery, all patients who had tympanostomy tubes had them removed and received erythromycin (premacrolide era). Of the 14 patients, 13 had clinical resolution of drainage . Franklin et al. later detailed clinical information on 21 sporadic cases of chronic otitis media with local drainage caused by RGM, which followed ear tube placement. Some of these patients also had mastoiditis. Of these, 20 of 21 cases (95%) were caused by M. abscessus. Therapy included surgical debridement, removal of the tympanostomy tubes, and antibiotic therapy. Initial therapy included amikacin (if susceptible) and either cefoxitin or imipenem for 3 to 6 weeks, followed by long-term (6 months) oral therapy with erythromycin (pre-newer macrolide era) or clarithromycin . Approximately 50% of the isolates from these isolates were also aminoglycoside resistant secondary to the use of chronic aminoglycoside ear drops. Only a few cases of M. fortuitum chronic otitis media or mastoiditis have been reported . The earliest report of otitis media due to M. fortuitum was described by Austin and Lockey . A 63-year-old man with mastoiditis also had a subperiosteal abscess at the surgical site from which M. fortuitum was cultured. Dalovisio et al. reported finding M. fortuitum in a 10-year-old boy with chronic recurrent otitis media and chronic mastoiditis, who also had a subdural empyema and meningitis. The patient was treated with amikacin for 4 months and underwent drainage procedures and placement of ventriculoperitoneal shunt before the infection resolved. Plemmons et al. reported a case of persistent otitis media in a 14-year-old girl who underwent right-side mastoidectomy. Prior to surgery, she had purulent otorrhea unresponsive to topical and systemic antibiotics and steroids. During surgery, a specimen was obtained from which M. fortuitum was grown. Cultures of fluid draining from the ear after surgery also grew M. fortuitum. She was given clarithromycin, and later, since the site was still draining, oral trimethoprim-sulfamethoxazole was added. The otitis media resolved and her mastoidectomy site completely healed after a 12-month course of antibiotics. Corneal infections (keratitis). | Ocular infections due to RGM have become more prevalent over the last 20 years. A comment must be made that before 1978, all of the cases reported were identified as "M. fortuitum." This probably reflects the inadequacies in taxonomy that existed at that time. Often in pre-1978 literature, "M. fortuitum" was used to designate M. fortuitum complex, and most publications did not give sufficient data to classify the organisms to species (as M. fortuitum, M. chelonae, or M. abscessus). We have attempted here to ascertain species (i.e., M. abscessus versus M. chelonae) when possible. This often was evident if susceptibility patterns of the organisms were reported. Before 1978, when Gangadharam et al. , reported a case of ulcerative keratitis caused by M. chelonae, the few prior reports had involved only M. fortuitum . Gangadharam et al. detailed the ophthalmological, bacteriological, and histopathological examinations of a puncture wound in a patient who was misdiagnosed as having herpes simplex keratitis. The patient received unsuccessful treatment for 2 months with topical, subconjunctival, and systemic antibiotics including gentamicin, kanamycin, and erythromycin. The ulcer regressed only after keratoplasty. After the Gangadharam et al. report, several other cases of keratitis caused by "M. chelonae" were noted . In most cases, it was still unclear whether the authors actually differentiated between M. chelonae and M. abscessus. Several reports of keratitis due to M. abscessus were described , and the organism was identified to species. Other authors only identified their isolates as M. chelonae (or the former term M. chelonei), but from the text (susceptibility patterns), we were able to differentiate the organisms as M. abscessus . Additionally, Bullington et al. did an extensive review of the literature involving keratitis with nontuberculous mycobacteria from 1965 to 1992. He found that 21 of 38 (55%) isolates were M. fortuitum, 16 of 38 (42%) were M. chelonae-abscessus group, and 1 (2%) was identified only as group IV mycobacteria . The most widely used antimicrobial treatment in the patients with keratitis in this review involved the use of topical aminoglycoside and/or systemic amikacin, gentamicin, kanamycin, or neomycin. Almost half of the patients required some type of surgical intervention including keratoplasty, keratoectomy, corneal graft, and several surgical debridement procedures . Since the review by Bullington et al. , a few additional cases of keratitis due to RGM have been recorded. Recently, an interesting case concerned a soft-contact-lens wearer with a corneal infiltrate. The patient had received oral steroids, systemic antibiotics including ampicillin and co-trimoxazole, and also topical chloromycetin, tobramycin, and miconazole. Subsequent corneal scrapings revealed M. chelonae. The isolate was initially misidentified as Nocardia and treated with ciprofloxacin, erythromycin, and fusidic acid. About 6 months later, the patient experienced a recurrence of the infiltrate, which initially improved with amikacin eye drops and systemic erythromycin but subsequently recurred and necessitated keratoplasty. Cultures of the excised cornea were still positive for M. chelonae. The patient was treated again with amikacin, erythromycin, and co-trimoxazole for 2 months. Eighteen months later, the graft had cleared and no evidence of infection remained . This patient's case, as others, was presumably complicated by the use of steroids . Since the early 1990s, other reports of eye disease due to RGM have been recorded, including postkeratoplasty and following laser (LASIK) surgery for correction of myopia . Medical therapy often is ineffective, due to delayed or incorrect diagnosis, difficulty in drug penetration (the cornea has no blood vessels and antibiotic entrance is only by surface diffusion), resistance to conventional antibiotics, and, rarely, emergence of drug resistant strains. Often, lamellar keratoectomy or keratoplasty (removal of the cornea) is the suggested treatment for patients with nonresponsive disease. These studies illustrate that, although the RGM are unusual causes of keratitis, infection can be severe and the outcome is often unfavorable. Cervical lymphadenitis. | The most common mycobacterial cause of lymphadenitis is M. tuberculosis in adults and M. avium complex in children. However, other mycobacteria including M. fortuitum, have been reported to cause lymphadenitis. Most RGM disease has involved the cervical lymph nodes, and has followed a dental procedure. Although M. fortuitum is a rare cause of lymphadenitis in adults or children, at least 19 cases have been reported in the literature. Cervical adenitis caused by M. fortuitum has been reported in patients with AIDS . Butt described two cases of lymphadenitis in patients with AIDS who were successfully treated with initial surgical drainage and antibiotics. Details were also extracted from seven other patients with of lymphadenitis caused by M. fortuitum, including one patient with generalized lymphadenopathy, two with bilateral involvement, and one each with unilateral submandibular nodes, submental lymph nodes, supraclavicular lymph nodes and cervical lymph nodes . Three of the nine patients listed (33%) had a history of dental procedure or extraction up to 6 months pre-dating the diagnosis of M. fortuitum. Four of the nine patients (44%) were treated with surgical incision and drainage followed by administration of combination antibiotics. Only two of the nine patients (22%) were treated with excision alone. Six of the nine patients (67%) improved or resolved their symptoms, and three (33%) died, with two of the three deaths being unrelated to the infection. At autopsy, however, disseminated disease with M. fortuitum was discovered in the third patient . Generally, it is accepted that abscess formation in lymphadenitis due to M. fortuitum is best treated by incision and drainage followed by combination antibiotics that includes amikacin (initially) and concludes with one or more oral antibiotics based on in vitro susceptibilities. Therapy should continue for at least 6 months . LABORATORY ASPECTS OF THE RAPIDLY GROWING MYCOBACTERIA : General | Traditional clinical laboratory identification of the RGM involved relatively few tests and was based mainly on growth rate, selected biochemical tests, pigmentation, and colony morphology. The RGM are defined as mycobacteria that grow within 7 days (most within 3 to 4 days) . Standard biochemical tests include iron uptake, nitrate reductase activity, tolerance to 5% NaCl, and arylsulfatase reaction. All members of the M. fortuitum group and the M. chelonae-abscessus group exhibit strong arylsulfatase activity at 3 days . Members of the M. smegmatis group are similar in growth rate but do not exhibit arylsulfatase activity at 3 days. The latter group is the only one of the three that produces pigmentation . Approximately 95% of isolates of M. smegmatis sensu stricto and 78% of isolates of M. goodii develop yellow-orange pigmentation after prolonged (7 to 10 days) incubation on Middlebrook 7H10 agar . (Because the pigmentation is so late and often occurs only on select media, it is often missed. It should always be sought in RGM with a negative arylsulfatase that are not obviously pigmented.) These relatively uncomplicated methods have proved inadequate for recognition of some older species (e.g., M. peregrinum) and many of the newer species (e.g., M. mucogenicum, M. immunogenum, M. goodii, and members of the M. fortuitum third biovariant complex such as the proposed species M. houstonense). Carbohydrate utilization tests and molecular studies have enabled a more accurate laboratory identification of the RGM species and groups. Unfortunately, in most clinical laboratories and many reference laboratories, identification of RGM to the species level has been relegated to low priority. Isolates often are identified to the group level only (e.g., M. chelonae-abscessus group, M. fortuitum-smegmatis group). However, it is no longer acceptable among good clinical and reference laboratories to fail to identify disease- producing RGM isolates to the species level, especially the separation of M. chelonae from M. abscessus. Types of clinical disease and antimicrobial susceptibilities often differ for individual species of RGM. The most common taxonomic error is the failure to separate the two distinct species, M. chelonae and M. abscessus. Although once thought to be subspecies within the species M. chelonae, they are, in fact, two distinct species. M. chelonae (formerly M. chelonae subspecies chelonae) is most often associated with disseminated skin and soft tissue infections with multiple painful draining lesions in immunosuppressed persons and is rarely a cause of chronic lung disease. Although M. abscessus is a cause of skin and soft tissue infections in patients (some of whom are immunosuppressed), it is also responsible for more than 80% of the chronic lung disease caused by RGM . The aminoglycoside preferred for treatment of M. chelonae is tobramycin, while amikacin is the preferred aminoglycoside for M. abscessus. Resistance to cefoxitin is one of the key differences between these two species, since M. chelonae is highly resistant (MIC, >=256 mug/ml) while M. abscessus is intermediately susceptible, with modal cafoxitin MICs of 32 mug/ml. M. chelonae appears to be much more susceptible to the newer antimicrobials linezolid and gatifloxacin than is M. abscessus (see the discussions of the two species above). Thus, the two species differ in clinical disease presentation as well as in susceptibility to drugs and in optimal therapeutic regimens, hence the importance of separating the two species in the laboratory. Implementation of additional laboratory methods including carbohydrate utilization and molecular diagnostics are now required to enable accurate species identification of the nonpigmented and late-pigmenting RGM (Tables and ). Although highly accurate at identifying slowly growing nontuberculous mycobacteria to species, HPLC has proven to be ineffective in identifying these organisms to the species level. Biochemical and Phenotypic Identification | After establishing a clinical isolate as an RGM, the best combination of traditional tests for recognition of the most commonly encountered species include the 3-day arylsulfatase test, iron uptake, nitrate reductase, and utilization of the carbohydrates mannitol, inositol, and citrate . A number of additional nonmolecular tests have also been utilized. A disk diffusion test using polymyxin B can also distinguish between the M. fortuitum group and the M. chelonae-abscessus group. Isolates of the M. fortuitum group exhibit a partial or complete zone of growth inhibition of 10 mm or greater around the polymyxin disk, whereas isolates of the M. chelonae-abscessus group show no partial or complete zone of inhibition . A previous IWGMT study showed that growth in 5% NaCl could reliably differentiate strains of M. abscessus (100% positive) from M. chelonae (17% positive). The citrate test was also found to be another useful biochemical test in that approximately 80% of M. abscessus isolates were citrate negative and 100% of M. chelonae isolates were citrate positive . (In our hands, the citrate utilization test has proven highly reliable.) Additionally, of the M. fortuitum group, only the unnamed third biovariant complex is positive for inositol. Utilizing molecular methods as the standard of identification, positive citrate tests with M. abscessus are rare . Many of the published studies of biochemical testing of RGM have utilized in-house prepared media or tests. There have been few studies of currently available commercial test systems to see if they are equivalent to these other methods. Knowing this, commercial systems could be used but require careful in-house validation. For example, Conville and Witebsky noted problems with such systems in identifying isolates of M. mucogenicum . High-Performance Liquid Chromatography | HPLC of mycobacterial cell wall mycolic acids also is used routinely in many reference laboratories as a means of identifying isolates of RGM . Recently, a comparison of the HPLC patterns obtained from the pathogenic members of the RGM was performed by Jost et al. (Abstr. 99th Gen. Meet. Am. Soc. Microbiol.). A standardized modified fluorescence detection (FL-HPLC) method was used for the analysis. FL-HPLC and UV detection methods (UV-HPLC) were analogous, and the study concluded that only under standardized conditions of culture medium, incubation time, and temperatures could "most" isolates and "most" species of the RGM be differentiated to species level by either method. The use of standardized methods is stressed by Chiu et al. (S. H. Chiu, K. C. Jost, Jr., D. F. Dunbar, and L. B. Elliott, Abstr. 98th Gen. Meet. Am. Soc. Microbiol. 1998, abstr. U-76, 1998), because differences in growth conditions (e.g., medium, temperature, and harvest time) can cause a diversity of patterns of the mycolic acid peaks and present difficulties in species identification (; Chiu, et al., Abstr. 98th Gen. Meet. Am. Soc. Microbiol. 1998). Thus, under nonstandardized conditions, even group identification of the RGM by HPLC is clearly problematic, in contrast to the general ease by which this method is able to differentiate the slowly growing mycobacteria into species. By routine HPLC and FL-HPLC, most isolates of M. chelonae and M. abscessus cannot be differentiated from one another. The new species M. immunogenum is also not separable by routine HPLC from the other two species in the M. chelonae-abscessus group . Similarly, members of the M. fortuitum group including M. fortuitum, M. peregrinum, and the unnamed third biovariant complex are not separable from each other or from the M. smegmatis group. They are generally grouped as the M. fortuitum-smegmatis group. Although the three species of the M. smegmatis group can be separated from each other and from M. fortuitum , the pattern overlap of all the M. fortuitum group members is too great to identify an unknown isolate to species. Thus, current studies show that HPLC is an acceptable method for separation of isolates of the M. fortuitum-smegmatis group from those of the M. chelonae-abscessus group but cannot identify isolates accurately to species. HPLC also works well when comparing a new taxon with already established one, since minor differences between the taxa may be readily apparent. It can be helpful for identification of members of the RGM only when used in conjunction with other methods and when used under specifically standardized testing conditions. Thus, identification of the RGM based solely on HPLC is not adequate. Molecular Identification | Nucleic acid probes. | No commercial DNA or RNA probes are currently available in the United States for any of the RGM. A kit (INNO-LIPA Mycobacteria; Innogenetics, Ghent, Belgium) based on reverse hybridization, in which the mycobacterial 16S-23S internal transcribed spacer region is amplified by PCR and amplicons are subsequently hybridized with probes for several species of RGM, is currently available in Europe but not in the United States . However, molecular methods for the identification of mycobacteria have been evolving rapidly and are now used in some specialized reference laboratories. Hybridization techniques with species-specific nucleotide probes, PRA, or direct sequencing of PCR-amplified products based on the polymorphism of the 16S rRNA gene have been useful in the identification of slowly growing mycobacterial species. However, because of the low level of variability within the 16S rRNA gene between some RGM species (e.g., M. chelonae and M. abscessus differ by only 4 bp in the entire 16S gene and have an identical hypervariable region A), a more variable gene sequence such as the hsp65 gene has proven helpful to distinguish between closely related species such as M. chelonae and M. abscessus . PCR-restriction enzyme analysis. | Several investigators have evaluated the hsp65 gene, present in all mycobacteria, for its value in the identification of RGM . Ringuet et al. found a less than 2% difference between the three most common RGM pathogenic species (M. abscessus, M. chelonae, and M. fortuitum) when the base sequences of the three type strains were studied. The base pair diversity was still much greater compared to the 16S rRNA gene. For example, the M. chelonae and M. abscessus sequences differ by almost 30 nucleotides, whereas their 16S rRNA genes differ by only 4 nucleotides. Thus, the hsp gene sequence is much more advantageous for the accurate identification of these two species than is 16S rRNA gene sequencing. The hsp65 sequences are highly conserved within a species and thus can be used for taxonomic studies. Telenti et al. demonstrated that a 439-bp portion of the hsp65 gene could be used for PRA and showed the patterns for most slowly growing mycobacteria and selected RGM. Steingrube et al. provided the most detailed PRA study to date of the RGM. They reported the PRA patterns from the 439-bp Telenti segment of the hsp65 gene for 129 clinical and reference strains of RGM belonging to 10 taxonomic groups. The authors found that among 24 endonucleases evaluated, PRA patterns produced by HaeIII and BstEII gave the best separation. More than half of the RGM were differentiated using HaeIII digestion alone. Single unique patterns were observed using both HaeIII and BstEII for M. fortuitum, M. smegmatis, M. mucogenicum, the sorbitol-negative third biovariant of M. fortuitum (100%), M. abscessus (96%), and M. chelonae (94%). Using another restriction endonuclease, AciI, RFLP patterns among clinical isolates of the M. smegmatis group supported the presence of the recently named two new species within the M. smegmatis group (M. goodii and M. wolinskyi) . Currently, the Telenti fragment of the hsp65 gene is the most widely used sequence for PCR-based identification of the RGM and is highly accurate for the M. fortuitum group, the M. chelonae-abscessus group, and the M. smegmatis group. It has not been studied for the identification of pigmented RGM. Vaneechoutte et al. devised another system of enzymatic amplification and restriction analysis using the entire 16S rRNA gene sequence. They studied 18 different species of Mycobacterium including strains of M. fortuitum and M. chelonae. They used different restriction enzymes (CfoI, MboI, and RsaI) from Telenti et al. and Steingrube et al. (both of whom used BstEII and HaeIII) and called their method ARDRA (amplified rDNA restriction analysis). Other target sequences have been studied for the identification of mycobacteria by using PRA or sequencing. These include the 32-kDa protein gene , the internal transcribed spacer of the 16S-23S rRNA gene , the superoxide dismutase gene , and the DNA J gene . However, to date, only selected slowly growing species have been extensively studied using these gene sequences and only one or two isolates of RGM have been tested. Preliminary data suggest that for the RGM, most of these gene sequences are much more variable and perhaps less useful for species identification than is the hsp65 gene. PRA seems particularly useful for identifying clinical isolates which gave equivocal results between compared species when other identification techniques were used. Although computerized analysis of PRA patterns has been recommended, visual inspection of the profiles is satisfactory when appropriate or comparative control strains are used . Recently, a related method for the identification of mycobacteria (including nontuberculous mycobacteria and M. tuberculosis), using amplification of the internal transcribed spacer regions with genus- and species-specific PCR primers, was described . A total of 87 isolates of nontuberculous mycobacteria from 10 species, including M. fortuitum, M. chelonae, and M. abscessus, were evaluated and identified using this system. The authors commented that the spacer sequences of the slowly growing mycobacteria are approximately 75 nucleotides shorter than those of RGM. Detailed studies of multiple strains of each RGM were not done and, to date, have been done only with the Telenti et al. 439-bp fragment of the hsp65 gene sequence . Plasmid profiles. | One of the earliest molecular tools used to help differentiate RGM was plasmid profiling. DNA probing with a plasmid-associated probe has also been performed and found to be a potentially useful guide for comparison of strains of RGM . Genetic relatedness of plasmids also may be studied further by RFLP of the plasmid or hybridization with defined or repetitive sequences. However, because these methods focus on extrachromosomal DNA, they do not provide conclusive evidence that strains are related. In fact, isolates having similar plasmid profiles may belong to different biovariant groups and vice versa . Thus, the use of plasmid profiles for species identification is currently of limited value, since some isolates of RGM do not contain detectable plasmids, plasmid profiles may change with time, and completely different plasmids may be of the same size . In general, other molecular methods that focus on chromosomal DNA have replaced this early type of analysis. Ribotyping. | Ribotyping is another potential molecular tool for strain comparison. However, it is probably more useful in delineating taxonomic rather than epidemiological relatedness of strains. Yew et al. stated two caveats to remember when applying this method. First, mutational changes can alter restriction patterns, and second, strains with small numbers of bands require restriction enzyme analysis with at least two different endonucleases. This method has been applied to only one major outbreak of RGM, and so knowledge of its usefulness is limited. Hybridization. | Another genetic technique which has been applied to the identification of some species of RGM is DNA amplification and oligonucleotide-specific hybridization . The system is based on selective amplification by PCR using mycobacterial DNA primers based on genes coding for 16S rRNA. During PCR, a label (digoxigenin-11-dUTP) is incorporated with the amplicon. After amplification, the amplicon is hybridized with species-specific oligonucleotides. After specific hybridization, enzyme immunoassay is used to show the specifically bound completer and thus identify the mycobacterial isolate. De Beenhouwer et al. reported that four of five M. chelonae-abscessus group strains were positive with their prepared probe using this technique and the fifth strain was positive with genus-specific probes but negative with all species-specific probes. None of the M. fortuitum strains studied hybridized with the species-specific probe. Hybridization with species-specific probes requires the development and availability of probes for every species. Thus, although this method has been useful for studying some species of mycobacteria, a detailed analysis of the RGM has been problematic. Patel et al. also studied a method which specifically hybridized a 5'-fluorescein-labeled strand of DNA to a species probe and was colorimetrically detected with an antifluorescein-enzyme conjugate. The method was able to correctly identify 10 species of mycobacteria, including some species of RGM. Direct detection in paraffin-embedded tissue. | Detection of nontuberculous mycobacteria by direct detection of nontuberculous mycobacteria in paraffin-embedded tissue by using amplified nucleic acid probes is an important advance in the diagnosis of mycobacterial disease. The technique is especially useful in clinical areas where the diagnosis is uncertain. Additionally, this technique has been advantageous when the mycobacteria fail to grow in culture because of small numbers or the fastidious nature of the organism, as well as in cases wherein mycobacterial disease was not initially considered. Moreover, an amplification assay can potentially give a result much faster (within 2 to 3 days) than can culture (up to 6 weeks) . PCR for detection of M. tuberculosis, from such clinical specimens as sputum, fluid aspirates, and tissue homogenates, has also been helpful in establishing a more rapid diagnosis of tuberculosis . Recently, human tissue samples stored as formalin-fixed, paraffin-embedded blocks have been used together with PCR methods to detect and identify the mycobacteria present . Briefly, as described by Marchetti et al. , DNA is extracted from formalin-fixed paraffin-embedded tissues. This method was originally developed to detect M. tuberculosis but, with some modification, can also detect nontuberculous mycobacteria. Paraffin is removed by adding xylene, vortex mixing, incubating at room temperature, and centrifuging. To facilitate pelleting and hydration of samples, ethanol is added and the supernatant fluid (xylene layer) is removed from the sedimented pellet. The pellet is air dried and resuspended in a special digestion buffer. Proteinase K is inactivated, and DNA is extracted from the emulsified tissue by adding phenol, vortex mixing, and centrifuging the mixture. A nested PCR of four assays, which uses three different concentrations of DNA, is performed. When a sample yields a positive result when amplified with primers homologous to sequences shared by a variety of mycobacterial species other than M. tuberculosis, it can be identified presumptively as containing a nontuberculous mycobacterial species. Shafran and Chui described a similar method of DNA extraction from paraffin-fixed skin biopsy tissue using a modified protocol described by Telenti et al. for detection of M. tuberculosis. No mention was made of detection of nontuberculous mycobacteria by this technique. Earlier reports by Ghossein et al. , described a method in which amplified fragments from paraffin-embedded tissue, as well as cultures of M. tuberculosis, M. avium complex, and "saprophytic mycobacteria," were identified by PCR of a 383-bp segment of the gene encoding the 65-kDa mycobacterial surface antigen and subsequent digestion with NarI. Unfortunately, the "saprophytic mycobacteria" were not further identified by the authors . Later, Perosio and Frank described a proteinase K digestion using a freeze-fracture extraction method developed by Ghossein and associates which enhances detection of mycobacterial DNA in clinical samples. Subsequently, a nested PCR with primers for the mycobacterial 65-kDa antigen gene was performed. Their primers were within a genus-specific region conserved among several nontuberculous mycobacteria including M. avium complex, M. gordonae, M. kansasii, and M. fortuitum. By comparing their primers with those used in previous studies and using available sequence data for different species, they predicted that their primers would amplify DNA from most pathogenic Mycobacterium species . Using PCR, mycobacterial DNA was detected in 7 of 7 wedge specimens and 9 of 18 transbronchial biopsy specimens. Restriction enzyme digestion of the amplified PCR product differentiated the species. Bascunana and Belak developed another nested PCR technique to amplify a 424-bp segment of the gene encoding the 65-kDa surface antigen of mycobacteria by using a restriction enzyme analysis procedure. The authors found that the location of the binding sites of the PCR primers in highly conserved parts of the 65-kDa antigen gene was important in the detection of all mycobacterial species without sacrificing the sensitivity of the test . The authors stated, however, that identification of similar restriction patterns such as with the RGM is difficult without special equipment and computer programs. Furthermore, nonspecific bands, which may be present with clinical samples, may interfere with the restriction enzyme analysis. Fluorescence in situ hybridization assay. | Another useful technique for direct detection of mycobacteria in situ was reported by Stender and colleagues (H. Stender, O. F. Rasmussen, K. Lund, K. H. Petersen, P. Hongmanee, H. Miorner, and S. E. Godtfredsen, Abstr. 29th World Conf. Int. Union TB Lung Dis., abstr. 170-PPDisc, 1998). This procedure, known as fluorescence in situ hybridization assay, uses peptide nucleic acids to penetrate mycobacterial cell walls and hybridize specifically to target rRNA. The authors state that the procedure includes a probe for a "range of other mycobacteria" including M. avium complex, M. gordonae, M. kansasii, and M. tuberculosis, but no specific mention of RGM is made. In summary, PRA using the hsp65 gene sequence and direct sequencing of the 16S rRNA gene to include the hypervariable regions, especially region A, are the best genetic methods at present for RGM species identification. PRA is the more practical and cost-efficient of the two methods. Susceptibility Testing for Taxonomic Purposes | The RGM may also be differentiated taxonomically using some standard antimicrobial susceptibility results. Both broth microdilution MICs and agar disk diffusion may be useful. Overall, the most useful agents have been polymyxin B, relative susceptibility to amikacin and kanamycin, and susceptibility to cefoxitin . The M. fortuitum group is easily separated from the M. chelonae-abscessus group by polymyxin B disk susceptibility. As noted above, members of the M. fortuitum group are inhibited by polymyxin B whereas the M. chelonae-abscessus group is resistant to polymyxin, with no complete or partial zone of inhibition . The uniform susceptibility of the M. fortuitum group to the sulfonamides, the fluoroquinolones, amikacin (with low MICs), and other drugs is distinctly different from the situation for the very resistant M. chelonae-abscessus group and also helps in separating the two groups. Furthermore, in the separation of M. chelonae from M. abscessus, susceptibilities to both cefoxitin and tobramycin are useful. As discussed above, isolates of M. chelonae have cefoxitin MICs of >256 mug/ml and tobramycin MICs generally of <=4mug/ml. In contrast, isolates of M. abscessus have cefoxitin MICs in the range from 16 to 64mug/ml, with 32 mug/ml being the modal MIC. Also, tobramycin MICs tend to be higher (usually >=16 mug/ml) than those of amikacin. Unlike the M. fortuitum group, all wild strains of M. chelonae and M. abscessus are clarithromycin susceptible. The newly proposed species M. immunogenum is similar to M. chelonae in that isolates are resistant to cefoxitin (MICs, >256 mug/ml). However, in contrast, the MICs of tobramycin are also high, usually >=8 mug/ml, which is more like those for M. abscessus. When using the agar disk diffusion susceptibility test, it was noted that the diameters of the zones of inhibition of both amikacin and kanamycin were equivalent for M. immunogenum, while M. abscessus and M. chelonae are more susceptible to kanamycin than to amikacin . Another species of RGM which has equivalent zone sizes for amikacin and kanamycin is M. mucogenicum. This species is highly drug susceptible, as are other members of the M. fortuitum group, but the MICs of amikacin are much lower than those of kanamycin for the other members (M. fortuitum, etc.). Isolates of M. mucogenicum are distinguished from the other nonpigmented RGM by the presence of a zone of inhibition of cephalothin for 90% of clinical isolates . Finally, tobramycin susceptibility is a useful test for the two newly proposed species within the M. smegmatis group compared to each other and M. smegmatis sensu stricto. By disk diffusion, isolates of M. smegmatis sensu stricto have zones of inhibition of >30 mm, M. goodii isolates have zones of inhibition of 11 to 30 mm, and isolates of the other new species, M. wolinskyi, have no zones of inhibition of tobramycin. Likewise, by broth microdilution, these three members of the M. smegmatis group have tobramysin MICs of <=1, 2 to 8, and >8 mug/ml, respectively. Isolates of the M. smegmatis group, unlike most of the other species of RGM, are usually resistant to clarithromycin . Susceptibility Testing for Clinical Purposes | For almost 20 years, susceptibility testing of RGM has been used as another tool for taxonomic separation of the RGM. Most species of RGM have a unique drug susceptibility pattern, and these patterns not only are important for therapeutic reasons but also can be used for taxonomic purposes to help identify the organism. Four different methods have been used for suspectibility testing isolates of RGM. These methods are agar disk diffusion, broth microdilution, E test, and agar disk elution. Each method has proved useful, but until recently, none of the methods has been well standardized and each method has both advantages and disadvantages that must be considered. In December 2000 , members of the NCCLS (National Committee for Clinical Laboratory Standards) Mycobacterial Subcommittee on Antimicrobial Susceptibility recommended the use of MIC determinations by using broth microdilution as the "gold standard" for susceptibility testing of the RGM. The eight antimicrobials initially recommended for MIC testing were amikacin, cefoxitin, ciprofloxacin, clarithromycin, doxycycline, imipenem, sulfamethoxazole, and tobramycin. These selections were made by the Mycobacterial Subcommittee after a review and statistical analysis of two interlaboratory studies that involved multiple susceptibility test methods and representative test organisms . Since some laboratories may perform susceptibility testing using one of the other methods, we briefly address each method. Agar disk diffusion. | Agar disk diffusion applies the Kirby-Bauer technique to the testing of RGM. Briefly, a suspension of the test organisms is prepared to match a McFarland 0.5 standard and inoculated onto plates of Mueller-Hinton agar supplemented with 5% oleic acid-albumin-dextrose. Commercial antibiotic disks are then applied, and plates are read after 72 h of incubation at 30C . This method has not been recommended for several years due to inherent technical problems. The major problems with this method are the absence of ready availability of some disks (e.g., sulfamethoxazole and doxycycline) and interpretation of "partial zones" of inhibition, which are observed when the concentration of the drug is near the MIC breakpoint between susceptibility and resistance to the drug. For example, the cefoxitin disk contains 30 mug of drug. However, the intermediate breakpoint is 32 to 64 mug/ml, so that the amount of drug in the disk is near the breakpoint and thus an area of colony growth often occurs within the primary zone of inhibition. Another major disadvantage is that susceptibility by disk diffusion for many of the newer drugs (e.g., fluoroquinolones, clarithromycin, and imipenem) has not been validated against an MIC method. The major advantages of the method are its ease and simplicity of set-up, plus the opportunity to look at colony morphology and to exclude the presence of a mixed culture. However, the above-mentioned problems outweigh the advantages, and thus the agar disk diffusion method is no longer recommended as anything other than a "screening tool" or taxonomic test. Therapeutic decisions should not be made on the basis of this method alone . Agar disk elution. | Agar disk elution has been used mostly by laboratory personnel who test limited numbers of isolates and do so only infrequently. This method uses commercial susceptibility disks from which the drug is eluted into the oleic acid-albumin-dextrose and then mixed with melted agar to produce specific drug concentrations. Failure to elute the drug prior to addition of the melted agar results in uneven drug distribution and often growth of susceptible organisms at the edges of the agar, where the drug concentration is the lowest. This method is optimally performed using six round-well tissue culture plates which hold about 5 ml each . Susceptibility testing is then performed by applying the method of proportions with a inoculum that averages 100 to 400 CFU/ml . The advantages of this method are that it utilizes materials and commercial drug disks readily available in most susceptibility laboratories and that the plates may be prepared on demand. It is not advisable to store plates for more than 3 days because of antimicrobial degradation . The method also correlates well with MIC tests performed in broth or agar for older drugs such as amikacin, doxycycline, cefoxitin, and sulfonamides. Some disadvantages of this method include the following: (i) preparation of the plates is tedious; (ii) the inoculum suspension must be carefully adjusted so that overinoculation of the wells does not occur (this is especially important and easily detected with drugs like the sulfonamides); (iii) there have been no MIC correlation or validation studies with agar disk elution and such newer antimicrobials as the fluoroquinolones, imipenem, linezolid, and clarithromycin; (iv) some desired concentrations for some drugs are not attainable due to the amount of drug in the commercial disks; (v) "trailing end points" can be a major problem with erythromycin and presumably the newer macrolides, because many strains that are susceptible in broth may produce fine, dysgonic colonies on agar or the drug may be slowly bactericidal (or bacteristatic), and since the end point for agar disk elution is either "growth" or "no growth", this means that isolates could be misinterpreted as resistant by the agar disk elution method when they appear susceptible by broth microdilution . E test. | The E test (AB Biodisk), or gradient MIC test, actually combines the agar diffusion technique with an exponential gradient of antimicrobial dilutions to produce an MIC result. The main advantages of this system include ease of set-up and use of standard agar media . However, recent interlaboratory studies by members of the Mycobacterial Subcommittee for Antimicrobial Susceptibility Testing of the NCCLS indicated that interpretation and reproducibility of the E-test MICs with RGM were often difficult. Isolates that were susceptible in broth were often interpreted as resistant by the E test in the four separate laboratories involved in the study . Diffuse elliptical edges and trailing growth or end points often made the determination of a precise MIC difficult for several drugs including ciprofloxacin, clarithromycin, imipenem, and cefoxitin. Because no standard RGM strain was found that exhibited reproducibility for all drugs tested, no breakpoints could be established and no control strains could be suggested for use in this method. Therefore, the NCCLS decided that further studies using the E test with the RGM were necessary before a recommendation for its usage could be endorsed . Broth microdilution MIC. | The broth microdilution method is the only method currently recommended by the NCCLS for antimicrobial susceptibility testing of RGM . The drug dilutions are prepared in cation-supplemented Mueller-Hinton broth using serial twofold dilutions of each drug . Several colonies of an isolate of RGM are suspended in Mueller-Hinton or Trypticase soy broth to reach a turbidity equivalent to the 0.5 McFarland standard. The organisms are diluted and inoculated into the drug wells of a 96-well microtiter plate. The panels are covered and incubated at 30C for 3 days in room air. End point MICs of all drugs except sulfonamides, which are read at 80% inhibition, are read as the first well in which there is no growth. The plates may be prepared in-house by the user with 96-well microtiter plates and automated dispensing equipment such as the Mini-Quick Spense System (Dynatech Inc., Chantilly, Va.). Commercially prepared custom-made panels for mycobacteria are also available from Trek Diagnostic Systems, Inc. (Columbus, Ohio). Commercial MIC panels for routine bacterial susceptibility testing, can be used but do not provide the optimal combinations and concentrations of antimicrobials that should be tested. For example, the concentrations of cefoxitin are too low (<=16 mug/ml) and panels that contain clarithromycin or doxycycline are not readily available. The eight drugs recommended by the NCCLS for the panel include those previously listed . A number of recommendations about test results were also made . As noted in Table , tobramycin should be reported only for isolates of M. chelonae since it has been recommended for therapy and validated by interlaboratory study only for this species. Any isolate of M. abscessus with an amikacin MIC of >=64 mug/ml should be retested and/or sent to a reference laboratory if the repeat result is the same since validated resistance is unusual (although mutational resistance involving the 16S rRNA gene does occur) . Imipenem MICs should not be reported for isolates of M.chelonae and M. abscessus because the results are not reproducible. Also, from that same study, it was decided that isolates of the M. fortuitum group with imiperen MICs of >8 mug/ml should also be repeat tested, with careful attention being paid to use of a maximum incubation time of 3 days, since all isolates should have imiperen MICs of <=8 mug/ml and the drug is notoriously unstable over time. Isolates of M. chelonae and M. abscessus do not exhibit good reproducibility in tests with imipenem and thus are not recommended for testing against this drug. This lack of reproducibility among the four laboratories was one of the major findings in the 1999 NCCLS study reported by Woods et al. . Another problem which was noted during the study was the trailing end point for M. fortuitum isolates when tested against clarithromycin. Therefore, this study recommended that isolates of M. fortuitum with clear end points should have their clarithromycin MICs reported; isolates that exhibit trailing end points to macrolides should currently be considered resistant until a better or different method is available. If confirmation is necessary, the isolate should be sent to a qualified reference laboratory. Finally, as stated above, MICs of sulfamethoxazole are read using 80% inhibition of growth as the susceptibility end point, not the 100% inhibition used for other antimicrobials. The 80% growth inhibition is usually the well in which a marked, definite decrease in the growth button is observed. Because few (if any) isolates of the M. fortuitum group are resistant to sulfonamides, testing of the resistance of this group of organisms to sulfonamides may not be necessary. Overinoculation of the drug panels is often most obvious with the sulfonamide wells. An inexperienced laboratorian may interpret the sulfonamide MIC as resistant when, in reality, the inoculum was too heavy; in this case, the MIC test should be repeated with a lower inoculum. In the same NCCLS study , the remaining antimicrobials ---cefoxitin, ciprofloxacin, and doxycycline ---exhibited few discrepancies in reproducibility and accuracy among laboratories. Additionally, Woods et al. proposed several breakpoint changes from the current NCCLS criteria for aerobic bacteria. The newly recommended RGM resistance breakpoint for cefoxitin is >=128 mug/ml, compared to its breakpoint of 32 mug/ml for other aerobic bacteria . For doxycycline the intermediate breakpoint is 8 mug/ml for other aerobic bacteria; however, for the RGM, the recommended susceptible MIC is <=1 mug/ml, with an intermediate range of 2 to 8 mug/ml. The resistance breakpoint is unchanged at >=16mug/ml. The newly recommended NCCLS breakpoints for the RGM are shown in Table . Some additional drugs not yet approved by the NCCLS should also be considered for testing. These include the new 8-methoxyfluoroquinolones (gatifloxacin and moxifloxacin), cefmetazole (not currently available in the United States), levofloxacin, vancomycin, and linezolid (; C. J. Crist, R. J. Wallace, Jr., B. A. Brown- Elliott, and L. B. Mann, Abstr. 101st Gen. Meet. Am. Soc. Microbiol. 2001, abstr. U-35, 2001). For these agents, the breakpoints are those for aerobic organisms in NCCLS M100- S11 , except for linezolid, for which tentative breakpoints for RGM were recently proposed . These values are listed in Table . CONCLUSIONS : The RGM have continued to emerge as important human pathogens that can cause a variety of diseases from localized cutaneous infections to disseminated disease. The RGM have been responsible for a number of health care-associated outbreaks and pseudo-outbreaks. The recent advances in antimicrobial therapy, including the new macrolides, fluoroquinolones, and oxazolidinones (linezolid), have improved the therapeutic options for the clinician and the prognosis of disease due to these organisms for the patient. This is especially true for M. chelonae and the M. fortuitum group. There is still, however, an urgent and compelling need for the development of better, more effective, and safe oral antimicrobials for the treatment of disease caused by the RGM, especially M. abscessus. M. abscessus lung disease, for example, is still generally incurable with currently available drug therapy. Susceptibility testing of the RGM is essential to choose optimal effective drug therapy and to monitor for the development of mutational drug resistance, which may occur with prolonged therapy. NCCLS tentative standards for such testing by broth microdilutions were published for the first time in December 2000. Taxonomically, the pathogenic RGM have undergone dramatic changes in the past few years. Multiple new species such as M. goodii, M. immunogenum, and M. houstonense (proposed) have been introduced, and some former subspecies or subgroups have attained species status. These taxonomic advances have been attributed primarily to the use of HPLC and the new molecular techniques such as PRA, ribotyping, hybridization, and 16S RNA gene sequence analysis. These molecular methods, especially PRA, are rapidly replacing conventional biochemical testing in the large reference laboratory. Identification of the RGM by these molecular diagnostic methods may not only improve the correct recognition of current RGM species as well as identity previously uncharacterized species, but will also decrease the traditional laboratory delays in species identification and hence will lead to a more rapid and accurate diagnosis of disease. This should result in earlier and more effective therapy for many of these infections. Hopefully, time will prove both of these speculations to be true. TABLE 1 : Species or taxonomic group and their most frequently recognized clinical disease syndromes TABLE 2 : Antimicrobial activity for the three most frequently isolated species of nonpigmented RGM TABLE 3 : Currently recognized taxa or species of nonpigmented (or late-pigmenting) RGM TABLE 4 : Laboratory phenotypic features of the 12 most clinically important species of nonpigmented or late-pigmenting RGM TABLE 5 : Suggested broth microdilution breakpoints for susceptibility testing of RGM Backmatter: PMID- 12364370 TI - Evolutionary and Historical Aspects of the Burden of Malaria AB - Malaria is among the oldest of diseases. In one form or another, it has infected and affected our ancestors since long before the origin of the human line. During our recent evolution, its influence has probably been greater than that of any other infectious agent. Here we attempt to trace the forms and impacts of malaria from a distant past through historical times to the present. In the last sections, we review the current burdens of malaria across the world and discuss present-day approaches to its management. Only by following, or attempting to follow, malaria throughout its evolution and history can we understand its character and so be better prepared for our future management of this ancient ill. Keywords: INTRODUCTION : This article is about the nature and impact of malaria; it is not a review of scientific research on malaria. It explores how, when, and where malaria parasites, their mosquito vectors, and humans, may have interacted and with what effect on ourselves, their human hosts. This is not a straightforward task. It is hampered by difficulties of collecting appropriate and reliable data and by problems of their interpretation. Inferences or conclusions may vary from well-founded and widely accepted to tentative or controversial. But, while the picture remains blurred in a number of places, the many faces and features of malaria and its imprint on the human species emerge, usually clearly and unmistakably. Many published discussions have dealt with the experiences of malaria, its nature, and its effects (see, e.g., references , , , , , -, , , , , , , , , , , , , , , , , , , ). They are informative and revealing, especially because each is a product of the medical and human health context, general outlook, and knowledge base of its time. Our present times also offer a unique perspective on this subject. It is only within the past half century that we have experienced malaria against a background of human health and health services which are, for the most part, vastly better than in all previous generations. In the same period, we have witnessed both the achievements of the first globally coordinated health delivery programs and also their problems. Within recent decades, and especially the last, biochemical and molecular genetic technologies to investigate distant events in the evolution and coevolution of malaria parasites and humans have become available. The following, therefore, is a brief reconstruction of the evolution and history of malaria and its burdens as we may perceive them at the start of the 21st century. MALARIA PARASITES OF HUMANS : Malaria is due to blood infection by protozoan parasites of the genus Plasmodium, which are transmitted from one human to another by female Anopheles mosquitoes. Four species of malaria parasite infect humans . The two which almost certainly achieved the widest global distribution are Plasmodium vivax and Plasmodium malariae. To the Europeans, these have been known and characterized since historically ancient times as the "benign tertian" (P. vivax) and "quartan" (P. malariae) periodic fevers. "Benign tertian" fevers were so named because they were not associated with the severe and often fatal manifestations of the "subtertian, malignant" periodic fevers (P. falciparum). "Tertian" and "quartan" refers to their characteristic feature of an acute febrile episode, or paroxysm, that returns respectively every third (P. vivax) or fourth (P. malariae) day. Tertian and quartan fevers are referred to with similar frequency in writings from northern Europe through much of the past millennium and from around the shores of the Mediterranean Sea from about the 5th century B.C. onward . Today, P. malariae has lost whatever predominance it may once have had and P. vivax and P. falciparum are the most commonly encountered malaria parasites . P. vivax is still found sporadically in some temperate regions, where in the past it was widely prevalent. It remains, however, very common throughout much of the tropics and subtropics. Because of the temperature limitations on its transmission by its mosquito vectors, P. falciparum is normally present only in tropical, subtropical, and warm temperate regions. In the tropics today, P. falciparum remains widely prevalent. The fourth human malaria parasite is Plasmodium ovale, which, like P. vivax, is the agent of a tertian malaria and which, also like P. vivax malaria today, carries a very low risk of fatal outcome. P. ovale has the most limited distribution of all the malaria parasites of humans. While it is prevalent throughout most of sub-Saharan Africa, it is otherwise known to be endemic only in New Guinea and the Philippines . TABLE 1 | Some representative distributions of the four recognized species of malaria parasites of humans in the world today EPIDEMIOLOGY OF THE BURDEN OF HUMAN MALARIA : Before we attempt to trace the passage and effects of malaria through evolutionary and historical time, we will review the principles which determine how malaria affects individuals and populations. Disease States of Malaria : The experience of malarial infection has many different forms, which we touch on here in outline. The diseases due to all four species of malaria parasite share the characteristic febrile episodes with their tendency to regular periodic paroxyms with chills, rigors, and sweating. They also have many symptoms in common with other infectious illnesses, including body aches, headache and nausea, general weakness, and prostration. Untreated infections of malaria are characterized by enlargement of the spleen. In P. falciparum malaria, severe and life-threatening conditions commonly arise. These cause dysfunction of vital organs, i.e., the lungs, kidneys, liver, and, most famously, the brain during "cerebral malaria." Severe anaemia can occur. These are the conditions which are associated with most of the mortality of acute malaria. Chronic infection with P. malariae can result in a nephrotic syndrome, and this, too, can eventually be fatal. Repeated attacks of malaria due to any species of the parasites over several to many years severely debilitate body and mind. Cachexia, a wasting of body tissues, takes place, and splenic enlargement becomes a constant feature. Lethargic and with sunken and sallow features, spindly limbs, and hard swollen belly is the general description of the condition. In this state the affected individual succumbs to diseases or other hardships that would scarcely threaten a person in reasonable health. Under the burden of chronic malaria, both the quality and duration of life are greatly reduced. An individual's experience of malaria at a particular time is, however, strongly governed by the type and degree of antimalarial immunity that he or she may have attained. Immunity to Malaria : Protective immunity against malaria can be thought of in different categories. There are two types of clinical immunity, one which reduces the risk of death from malaria and another which reduces the intensity of clinical symptoms. A third type of protective antimalarial immunity is antiparasitic immunity, which directly reduces the numbers of parasites in an infected individual. These are epidemiological definitions of immunity. The actual cellular and molecular mechanisms of immunity may overlap to a considerable degree between these categories. The number of malarial inoculations experienced, and the intervals between them, are all-important to the malaria immune status of an individual. In the case of acute attacks of P. falciparum malaria, it is possible that a degree of immunity to some aspects of severe, life-threatening disease may be achieved after only one or two infections . However, clinical immunity to other, non-life-threatening clinical effects of malaria requires more and frequent inoculations of malaria . Effective antiparasitic immunity is achieved only after very many and frequent infections . Two other forms of acquired immunity to malaria should be mentioned here. One is immunity which could protect against the parasites ever becoming established in a human host, in other words immunity against the stages which are introduced into the human body by a mosquito bite. These are the sporozoites and also the early, nonpathogenic stages as they develop in the liver. While immunity against sporozoites and liver stages may eventually develop following long exposure to intense malaria transmission, it is probably not very significant in the crucial early years of exposure. The other form of nonprotective immunity against malaria parasites is that against the sexual blood stages of the parasites. These are the stages which are infective to mosquitoes. Antibodies against antigens in gametocytes and gametes of malaria parasites develop readily during the course of natural infections and probably have some effect on the transmission dynamics of malaria in certain endemic situations. Anti-sexual-stage immunity, however, has no direct relevance to the status of an individual with respect to his or her experience of a current malarial infection. The attainment of protective immunity against the disease-causing asexual blood stages of malaria parasites is a complex process. Not only is such immunity specific to each parasites species, but also there is certainly a major element of parasite "strain" specificity to antimalarial immunity . Much of the slow acquisition of antiparasitic immunity occurs because an individual is usually subjected to a succession of inoculations of parasites which are genetically and antigenically distinct from each other. Therefore, to combat each infection, a human host must mount a new and "strain"-specific immune response to each antigenically distinct parasite inoculum. Only when a sufficiently wide spectrum of such parasite "strains" has been experienced is effective immunity achieved against all the parasites within a locality where infection is endemic. However, not only are the parasites in different inoculations genetically different but also, during the course of each infection, the parasites from a single inoculation undergo clonal "antigenic variation" . To control an infection, the host must mount a new specific immune response to each new antigenic variant as it arises. Through this process, a single malarial infection can be prolonged over many months to years. Because of the time taken to achieve effective immunity to malaria under conditions of endemic infection, antimalarial immunity is often said to be "age dependent." In the sense intended, however, it would be more accurate to say that it is "duration of exposure dependent." There are, nevertheless, truly age-dependent aspects both to the attainment of immunity and to the pathologic responses to malarial infection. Very young children appear to have a poor capacity to acquire effective protective antimalarial immunity of any sort, while older children and adults may so do more readily . Infants and the very young are more prone to malarial anaemia, while cerebral damage due to P. falciparum malaria predominates in slightly older children. Yet other severe conditions, including renal, hepatic, and pulmonary failure, are most commonly seen in adults . Every individual exposed to endemic malaria thus faces a long and dangerous battle to achieve protective immunity against the diversity of malaria parasites and their antigens to which he or she may become exposed. However, having been achieved at such cost, effective immunity is readily lost again. An interval of perhaps half a year to a year without reinfection appears to be sufficient to leave an individual vulnerable, once more, to the full impact of a malarial infection (, , ; A. Lukas, personal communication). Direct evidence for this statement is admittedly hard to find, yet it has long been attested by anecdote and expert opinion. Loss of protective immunity within a few months without reinfection also appears to hold in relation to infection-induced protective immunity against malaria in animals (R. Carter, unpublished observations). If the above statement is indeed correct, then the number, type, and pattern of delivery of malarial inoculations, in other words the type of malarial endemicity, to which an individual is exposed must profoundly affect his or her immune status and hence disease status, as is now discussed. Malarial Endemicity : Following the approach of MacDonald , malarial endemicity may be categorized into three types . One is stable endemic malaria. Stable malaria occurs when a population is continuously exposed to a fairly constant rate of malarial inoculation. The next category is unstable endemic malaria, under which a population is subjected to more or less permanent malaria transmission but under circumstances in which there are large fluctuations in the rates at which malarial inoculations are delivered to individuals within the population. In terms of disease and immunity, these fluctuations become especially significant when they cause individuals to experience intervals of a year to several years between inoculations of malaria. The third type is epidemic malaria. This is, in effect, an extreme form of unstable malaria. It occurs when a population, or even a small group of individuals, is subjected to an increase in malaria transmission rates above that previously or normally experienced. When P. falcaparum is involved, malaria epidemics can be among the most lethal forces of nature. It should be noted that the mean rates of malarial inoculation or transmission intensity are not included in the definitions of any of these three types of malarial endemicity. High or low transmission rates can, in principle, occur for stable, unstable, and epidemic malaria . Nevertheless, the highest natural malaria inoculation rates, those of hundreds of infectious bites per individual per year, probably occur only under stable endemic conditions. Inoculation rates of as few as one or two infectious bites per year are characteristic of unstable conditions but can also be encountered under conditions of stable endemic malaria . Malaria epidemics can, and probably often do, occur under conditions of relatively low to moderate malarial inoculation rates . TABLE 2 | Characteristics of the three categories of malaria transmission How Endemicity and Immunity Interact To Determine the Effects of Malaria : Stable malaria. | Under conditions of stable malaria, because of the regularly delivered inoculations, a strong protective immunity against overt illness and risk of death from malaria is acquired, usually by the age of 4 or 5 years. Before this age there is much morbidity and mortality, especially in the presence of P. falciparum malaria. Historically, and still today, such conditions have prevailed mainly in sub-Saharan Africa. The following quotation is characteristic of the historical picture of malaria in tropical Africa. "In East Africa 50% of native children die before the age of 4 years, mostly from malaria" (Brumpt, 1922, quoted in reference ). The so-called age-dependent pattern of immunity to malaria in sub-Saharan Africa is often attributed to the intense malaria transmission rates in this region. However, we suggest that it is the stability of malaria transmission in tropical Africa, due to biological features of the Anopheles gambiae group of vectors of malaria, their ecology, and their environment (discussed below), that is the more important feature of African malaria transmission. Because, individuals rarely go more than a few months without a malaria challenge under conditions of stable transmission, there is little risk that immunity, once attained, will be lost again. This remains true even in the setting of quite low annual malarial inoculation rates . There is, nevertheless, a clear shift from severe malarial disease in younger children (younger than 5 years) toward relatively increased rates of both mild and severe disease in older age groups at entomological malarial inoculation rates probably below about 10 to 20 infectious bites per year. However, under conditions of stable malaria in Africa, the lifetime risk of severe disease, and hence of direct malaria mortality, due to P. falciparum seems to remain fairly constant across the spectrum of inoculation rates from very high to very low . Unstable and epidemic malaria. : By contrast, where transmission conditions are unstable, low to moderate mean malarial inoculation rates can be highly dangerous. This is because immunity to malaria, while slowly gained, is, as we have suggested, rapidly lost following perhaps half a year to a year without infection. Wherever the rate of delivery of malarial inoculations is both low and highly erratic, extended periods of a year or so without reinfection occur often, while at the same time the risk of eventual reinfection remains high. In these circumstances, individuals are vulnerable throughout life to clinically active malarial infection and are often affected by it. Furthermore, all species of malaria parasite, and not just P. falciparum, become dangerous under these conditions. Recurrent infections with any species of malaria parasite are eventually so debilitating that life expectancy can be reduced to half or less of that in a contemporary malaria-free and otherwise salubrious environment . The situation is illustrated by the following quotation which refers to malarious districts around the shores of the Mediterranean in the early 19th century. "The most fertile portions of (Italy) are a prey to (malaria); the labourer wanders... the ghost of a man, a sufferer from his cradle to his grave; aged even in childhood, and laying down in misery that (brief) life which was but one disease... . Such also is Sicily, such Sardinia and such is classic Greece. To live a living death and to be cut off from even half that life" . Wherever antimalarial immunity has declined, has been lost, or was never achieved in the first place, populations at the margins of malaria transmission zones are vulnerable to epidemic malaria. When P. falciparum is involved in an epidemic, death rates can become very high. A single untreated attack of P. falciparum malaria in a nonimmune individual carries a risk of death that may be anywhere from a few percent to at least 20 to 30% according to circumstances . These higher rates are the kinds of mortality rate that occur during P. falciparum malaria epidemics . The following quotation is representative of these events. "In each house were... three or four patients who complained of chilling, severe headaches, sweating, pain in back and extremities... After four or five relapses, the headaches and pain became unbearable for many patients who then exhibited a muddling delirium with coma, ending in death. Most... were between the ages of 5 and 20 years. Since they are far away from even the simplest clinic, which means no possibility of saving their lives, they are dying like bees in a smoked hive" (from a Field Report from the 1958 malaria epidemic in Ethiopia ). In a season, this very typical malaria epidemic produced about 3.5 million cases of malaria among which both P. vivax and P. falciparum malaria would have been present; it took about 150,000 lives. Similar and even higher malaria mortality rates were experienced by Europeans entering tropical regions before the introduction of quinine in the mid-19th century (, -, , ). Indeed, the consequence of the introduction of nonimmune persons into a malaria-endemic region is, in effect, a form of epidemic malaria. Biological Basis of Stable, Unstable, and Epidemic Malaria : The differences in stability of malaria transmission, notably between tropical Africa and most other malarious regions, are due largely to the behaviors and other biological characteristics of the regional species and subspecies of Anopheles vectors and to their environment. The strong human-biting preferences and highly domestic habits of the tropical African vectors lead to very uniform contact between them and the human blood source in sub-Saharan Africa. The climatic conditions are also highly conducive to malaria transmission, being warm and humid with relatively few fluctuations. This supports longevity of the vector mosquitoes and rapid development of the parasites within them. All of these features combine to a recipe for stable and, indeed, generally intense malaria transmission. Elsewhere, in the tropical, subtropical, and temperate worlds, the females of most Anopheles species have a preference for animal rather than human contact and have less domestic breeding, resting, and dispersal habits than the African vectors. The climates and especially the microclimates experienced by these vector mosquitoes are usually cooler, drier, and more variable than in tropical Africa, leading to more uncertain survival of the mosquitoes and less reliable development of the parasites within them. These are conditions that lead to highly erratic contact between potential vectors of human malaria and the human hosts and to very irregular delivery of malaria inoculations, in other words, to unstable malaria transmission. Epidemic malaria arises in any situation in which new or elevated malaria transmission capacity is suddenly introduced into a population with inadequate levels of immunity with which to absorb it. The circumstances under which this can occur are extremely diverse . Typically they involve nonimmune populations in otherwise malaria-free locations adjacent to regions of endemic malaria. An epidemic occurs when atypical weather, e.g., drought, excess rainfall, and higher than usual temperatures, create conditions that transiently support malaria transmission. However, warfare, pioneering, and malaria control, among other human activities, have all been responsible for the creation of malaria epidemics. Indirect Mortality of Malaria : Under whatever condition of endemicity malaria occurs, much, and according to the evidence of a number of analyses, generally most , of the malaria-related mortality has in the past been an indirect result of the effects of malarial infection combined with other infections and conditions. Wherever interventions have been conducted for the reduction of malaria transmission, reductions in total mortality rates have been several times greater than the malaria-related death rates as estimated prior to the interventions . Some of this discrepancy may be due to failures to detect or account for all the deaths in which malaria would have been an apparent cause. However, most of the excess deaths associated with the presence of malaria, but not obviously due to malaria, can be attributed to delayed or indirect causes. These could include the nephropathy of long-term P. malariae infection . However, more generally, the deaths are due to the predisposing effects of malaria to death from other conditions such as respiratory infections and malnutrition . Nonlethal Health Consequences of Endemic Malaria : In historical times, and as recently as the early to mid-20th century in southern Asia, continual malarial infection and reinfection had devastating effects on the mental, physical, social, and economic conditions of the individuals and communities affected. In addition to its toll in death and general morbidity, there are several features of the burden of malaria that warrant specific mention. Malaria presents particular problems during pregnancy. This is probably due, at least in part, to the "immunosuppressed" status of the pregnant woman. Even in otherwise highly malaria-immune women, the risk of malarial infection in pregnancy is high, with increased risk of low birth weight, miscarriage, infant mortality , and morbidity and mortality in the pregnant woman. "Malaria exerts a considerable influence in reducing the births and increasing the number of still births in a community" (Bentley, 1911, quoted in reference ). Historically, "infantilism" in both sexes and impotence in males were associated with malarious areas . "One of the great evils is the impotency so commonly found in waterlogged villages (and) which results from the deterioration of the health produced by constant attacks of fever and the presence of an enlarged spleen" (Dyson, 1895, quoted in reference ). Together with the problems of malaria in pregnancy and infant mortality, these added up to a large reduction in the fecundity of a population and, in the worst-affected locations, contributed greatly to depopulation under the impact of malaria. "In hyperendemic and severely endemic (for malaria) districts in India the tendency is for the population to decline" . Among the most oppressive and, today, least appreciated of the effects of malaria is that on the mental state of the sufferer. Observers from many different times and places, including Hippocrates , Macculloch , Jones , Anderson , and Sinton , are remarkably consistent in their portrayals of the condition of the inhabitants of malarious regions especially in Europe or Asia. They depict individuals and communities in states of pronounced mental and psychological distress under the influence of endemic and invariably unstable or low-transmission malaria. The following are typical of these. "Anyone who has observed closely cases of malaria cannot fail to have noticed its effects upon the mentality of the sufferer ---mental activity is dulled, irritability of temper is the rule, initiative is lacking, decisions are put off or reached with difficulty, ambition is lost and depression is a prominent symptom" ; "School children in the Transvaal infected with chronic malaria were mentally classed as "feeble-minded"" (Leipoldt et al., 1921, quoted in reference ). In today's much healthier world, situations matching their descriptions are conspicuous by their absence. Nevertheless, and not to be confused with the immediate consequences of malarial infection itself on the mental state, a significant proportion of the survivors of malarial infections still carry mental or neurological deficits . Effects of Malaria on Social and Economic Development : An adequate discussion of the economic effects of malaria is beyond the scope of this article. Populations exposed to the unremitting impact of malaria must always have lived and died in destitution of one degree or another. In historical times, economic enterprise has been difficult and often impossible in the presence of malaria . Malaria alone has often been sufficient to wreck efforts in pioneering, agriculture, and civil engineering, as illustrated by the following representative quotations. "The hyperendemic areas (for malaria), although sparsely inhabited, are often areas where large plantations and large industrial undertakings are situated and which are, therefore, often the site of a considerable immigrant populations that are quickly mown down." (Malaria Commission of the League of Nations, 1930, quoted in reference ); "While there is good land in the Southern United States as in the North, the land in the North sells at about 12 to 20 times the price, the difference being mainly due to malaria" (Carter, 1922, quoted in reference ); "Railway construction in the tropics is nearly always associated with fulminent epidemics of malaria ---"a death a sleeper" is the generalisation on the happenings" (Senior-White, 1928, quoted in reference ). Throughout history and to the present day , wherever and however it may have manifested itself, malaria has always imposed one of the severest of impediments to social and economic development. THE "MALARIA HYPOTHESIS" AND RECENT HUMAN EVOLUTION : It can be no surprise, therefore, that malaria parasites should have had a profound impact on recent human evolution . This is the proposition of the "malaria hypothesis," which posits that certain human genetic polymorphisms, especially those affecting red blood cells (RBCs), have been selected to high frequencies because they have protected against the effects of malarial infections. Indeed, a greater number of identified human genetic polymorphisms meet some or all of the expectations of the malaria hypothesis than can be attributed to selection under any other single agent . The first statement of the malaria hypothesis was made in 1948 by J. B. S. Haldane , who proposed it as an explanation for the high frequencies of thalassemia around the shores of the Mediterranean Sea, where malaria had long been endemic. Probably unknown to Haldane at the time, a possible association between another inherited hemoglobinopathy, the sickle cell trait, and protection against malaria had already been noted. In 1946, in a study of inpatients at a regional hospital in Northern Rhodesia, now Zambia, E. A. Beet had recorded lower rates of malarial infection among carriers of the sickle cell trait than amongst nonsicklers . In the following, we discuss the evidence for, and implications which arise from, the main candidate human polymorphisms to fall within the terms of the malaria hypothesis. Thalassemias : The thalassemias are a class of anaemias which are due to abnormalities in the genes coding for hemoglobin. They involve the effective loss of either the alpha (alpha thalassemias) or beta (beta thalassemias) chain of this molecule . Many different specific mutations in the genes for the alpha and beta haemoglobin chains are represented among the known thalassemias . In Europe, high frequencies of thalassemias are found around the shores of the Mediterranean Sea, as Haldane had noted . Elsewhere, thalassemias occur at elevated frequencies in populations through most of Africa, the Middle East and Central Asia, the Arabian peninsula, the Indian subcontinent, Southeast Asia, southern China, and the islands of the Western Pacific from the Philippines in the north to the Timor Sea in the south and to New Guinea and the islands of Melanesia in the east . Up to about 2,000 years ago, these were the probable global limits within which malaria was endemic. The basis of Haldane's original conjecture has been confirmed and expanded in several studies which show the association of thalassemias with the long-standing exposure of a population to the presence of malaria (see, e.g., references and ). An approximately 50% reduction in the risk of malarial disease has recently been shown for both heretozygotes and homozygotes for certain alpha+ thalassemias . Glucose-6-Phosphate Dehydrogenase Deficiency : Of broadly similar global distribution to the thalassemias are a group of mutations in the sex-linked gene for glucose-6-phosphate dehydrogenase (G6PD) which give rise to the G6PD deficiencies . G6PD deficiencies were originally recognized as a clinical condition known as favism, so called because of the hemolytic crisis that can be caused in those affected, by the consumption of oxidant foods such as fava beans. Today, G6PD deficiency is a recognised hazard in malarious regions because of the associated risk of oxidant stress from taking antimalarial drugs such as primaquine. In heterozygous and hemizygous combinations, G6PD deficiency has now been shown to be associated with a level of protection of about 50% against severe P. falciparum malaria . Therefore, high rates of G6PD deficiency in many parts of the world can probably be accounted for as the result of selection by malaria. Recent evidence indicates that the alleles for G6PD deficiency were selected in African populations between 4,000 and 12,000 years ago . This period coincides with some estimates for the time of emergence of P. falciparum as a major human pathogen at around 4,000 years ago . Equally, however, the selection for G6PD deficiency may have preceded this event and overlapped a time when P. vivax malaria could still have been a major selective force in Africa (see below). So which parasite could have been responsible for the selection of G6PD deficiency in the African populations? It may often appear that the malaria hypothesis should be taken to apply mainly, or solely, to P. falciparum. However, as we shall further explore, the hypothesis is equally valid for P. vivax and perhaps for other species of human malaria parasites as well. That P. vivax malaria may represent a force for the selection of G6PD deficiency is consistent with the presence of elevated frequencies of G6PD deficiency in a population in northern Holland , where P. vivax malaria, but almost certainly not P. falciparum malaria, was prevalent for at least 500 years. We suggest that either or both of P. falciparum and P. vivax could have supplied the force that first selected for G6PD deficiency in African populations. Sickle Cell Trait : Neither the thalassemias nor G6PD deficiency is highly protective against death from P. falciparum malaria, with each appearing to reduce the risk by about 50%. There are, however, human mutations which give at least 90% protection against the mortality of malarial infection. These include the gene for sickle cell hemoglobin, or hemoglobin S. The possible protective effects of the sickle cell trait against P. falciparum malaria were, as mentioned above, first noted by Beet in 1946 . They were next extensively explored clinically and epidemiologically by Allison and subsequently by many others. The sickle cell trait is due to a single point mutation in the gene for the beta chain of hemoglobin in which the glutamate at position 6 is replaced with a valine . It results in a hemoglobin molecule, hemoglobin S, which, in the homozygote, gives rise to sickle cell anemia. Today in the United States, such individuals commonly reach middle adulthood . In the premodern world, however, those affected with sickle cell disease would probably rarely have survived to puberty . In reproductive terms, the cost to a hemoglobin S homozygote would have been a loss of fitness of nearly 100%. Such a mutation could have been sustained only if the heterozygotes, those carrying one hemoglobin S gene and one normal, hemoglobin A, gene, had an advantage great enough to balance the cost of the homozygotes to the population. It has been fully confirmed that this advantage is protection against the risk of death from infection with P. falciparum malaria. Children in West Africa who are heterozygous for the sickle cell gene are at approximately 1/10 of the risk of death from P. falciparum malaria as are children who are homozygous for the normal gene . Hemoglobin S clearly fulfils the conditions of the malaria hypothesis. In many parts of Africa, the frequency of the sickle gene exceeds 30% . The reproductive losses which these frequencies impose through homozygosity for this gene affect at least 1 birth in 10. They reflect the cost of endemic P. falciparum malaria to a human population. Hemoglobin C : The gene for hemoglobin C is allelic with that for sickle cell hemoglobin but codes for a lysine instead of a glutamate (hemoglobin A) or a valine (hemoglobin S) at position 6 in the beta chain of hemoglobin. In the homozygous carrier, hemoglobin C confers a loss of fitness which is much less than that associated with the sickle cell trait (homozygous hemoglobin S), being apparently comparable to that of a mild form of thalassemia . Hemoglobin C is found only within certain West African populations, where the frequency of the allele reaches, and sometimes even exceeds, 10 to 20% . There is now little doubt that this polymorphism has been, and indeed is probably still being, selected under the effects of P. falciparum malaria in West Africa . However, the nature of the protection by hemoglobin C is very different from that associated with hemoglobin S. In the case of hemoglobin S, the heterozygous state confers a high level of protection against death from P. falciparum malaria. In that of hemoglobin C, similar high levels of protection, greater than 90% against infection and therefore against the risk of death by P. falciparum malaria, are achieved only in the homozygote . The protective effects of hemoglobin C in a heterozygous combination are, at about 30%, relatively weak. Now, the rate of selection of an allele whose advantage is expressed mainly in the homogygous state, as is the case for hemoglobin C, is slow compared to that of an allele, such as that for hemoglobin S, which carries a similar advantage but in a heterozygous combination. This principle is discussed more fully in relation to selection for RBC Duffy negativity (see below). Under the influence of the same selective force, therefore, hemoglobin S would be expected at first to achieve higher frequencies than hemoglobin C. However, because the balancing cost of hemoglobin C is low and that of the allelic hemoglobin S is high, hemoglobin C should eventually replace hemoglobin S in populations exposed to selection by P. falciparum. In West Africa this has clearly not yet happened , and it suggests that P. falciparum malaria may have arrived only "recently" within the West African population. When, therefore, might this "recent" arrival of P. falciparum in West Africa have taken place? The relative fitnesses of the different haemoglobin alleles, A, S, and C, in the West African situation can now be given , and the relevant calculations can be made . We suspect that they would show that the hemoglobin C allele should approach population equilibrium within several thousand years, and almost certainly within less than tens of thousands of years, of selection under P. falciparum malaria, as, indeed, has already been suggested . Therefore, by this line of argument, because hemoglobin C has not yet reached its expected equilibrium frequency, P. falciparum has been a selective force in West Africa for less than this time, i.e., probably for less than a few tens of thousands of years. Hemoglobin E : Populations with high frequencies of the sickle cell gene, the S allele of hemoglobin, extend throughout all but the southernmost malaria-free tip of Africa, through parts of the southern and eastern shores of the Mediterranean, notably in Greece, across the Arabian peninsula and the Indian subcontinent, and as far as the eastern border of present-day Bangladesh . Here the prevalence of hemoglobin S ceases. It is immediately replaced, however, by another mutant of hemoglobin, in which the glutamate of position 26 of the beta chain in haemoglobin A is replaced with a lysine . This is the gene for hemoglobin E, and it occurs at high frequencies in populations throughout Southeast Asia. Hemoglobin E has no obvious clinical effect except in combination with certain thalassemia mutations . It does not, therefore, impose severe costs on a carrier population. While there appear to be no reports that directly demonstrate protection by hemoglobin E against P. falciparum, there is evidence suggesting that it may protect against P. vivax infection . It has also been reported that hemoglobin E trait patients clear infections of P. falciparum more rapidly than do others during treatment with artemisinin derivatives although not during treatment with other antimalarial drugs . The geographical region across which hemoglobin E is prevalent is one of the most malarious, both historically and today, and the burden of malaria in this region has therefore long been one of the heaviest in the world. The gene for hemoglobin E is certainly a good candidate to have been selected under the pressure of malaria in accordance with the malaria hypothesis, possibly under the influence of P. vivax as the principal agent of selection prior to the arrival of P. falciparum within these populations. Ovalocytosis : Beyond the Indonesian and Malay inhabited archipelagos on the Western Pacific rim lies Melanesia, including the island of New Guinea, the Solomon Islands, and the islands of Vanuatu. Outside of tropical Africa, malaria transmission intensities here include the highest in the world. Among the people of this region there occurs a polymorphism which affects one of the main structural proteins of the RBC, known as band 3 . A deletion mutation in the gene for this protein has been identified as being at least one genetic determinant of a condition known as ovalocytosis, so named from the abnormal shape of the affected RBCs in the heterozygote . Since, as it appears, no homozygotes for this band 3 gene deletion are born , the homozygous condition is probably uniformly lethal in utero. The reproductive cost to a homozygote for the ovalocytosis mutation would appear, therefore, to be 100%. The malaria hypothesis almost certainly explains the high frequency of the ovalocytosis mutation in these populations. Neither the sickle cell gene nor hemoglobin E appears to be present among Melanesian populations. Other than ovalocytosis itself, only the moderately protective G6PD deficiency gene and some thalassemias are found . Ovalocytosis occurs at high frequencies (up to 20%) in the highly malarious lowlands of New Guinea but virtually disappears in the malaria-free highlands. Carriers of the trait are at reduced risk of infection with P. falciparum and, especially, P. vivax malaria . Two studies of the frequency of the band 3 gene deletion mutation in children with P. falciparum malaria showed that none who carried the gene (in heterozygous combination) developed cerebral malaria . Protection against malarial infection by ovalocytoses of unspecified genotypes has also been demonstrated in populations in nearby Southeast Asia . Ovalocytic RBCs resist invasion by P. falciparum and by Plasmodium knowlesi , a simian malaria parasite with similar RBC invasion characteristics to P. vivax. It is not known, however, if the ovalocytic conditions which protected against malaria in these studies were due to the same band 3 gene deletion that was shown to protect against cerebral malaria . Given the apparent uniform in utero lethality of the homozygote form of the band 3 gene deletion mutation, its high frequency in the lowland populations of New Guinea once again reflects the magnitude of the reproductive cost of malaria to this population. RBC Duffy Negativity : It is striking and long-recognized fact that most members of indigenous populations of West and Central Africa are completely refractory to infection with P. vivax malaria . Almost all members of these populations are also homozygous for an FYnull (RBC Duffy-negative) allele of the gene that controls expression of the Duffy antigen on RBCs . The Duffy blood group antigen system is represented mainly by two serologically distinct forms determined by alleles FY*A and FY*B. Each of these alleles also exists in a mutant, unexpressed, or null form as FY*Anull or FY*Bnull . The Duffy antigen, which has been identified as a chemokine receptor , is also an essential receptor for P. vivax merozoites to be able to enter a host RBC . This accounts for the association between complete refractoriness to P. vivax infection in most West and Central Africans and the almost universal homozygous RBC Duffy negativity in these populations. The molecular genetic basis for RBC Duffy negativity is a single-nucleotide substitution polymorphism (SNP) in the promoter region of the gene for the Duffy antigen. This promoter controls the expression of the Duffy antigen specifically in RBCs. The same SNP is associated with both the FY*A and the FY*B alleles , leading to the FY*Anull and FY*Bnull RBC Duffy-negative alleles. Now, the promoter for a structural gene, such as that for the Duffy antigen, governs its expression on its own chromosome strand only. Therefore, in a heterozygous individual, expression of the Duffy antigen in RBCs should be suppressed only on the strand with an FYnull allele, while from the strand with a Duffy-positive allele, the Duffy antigen should be fully expressed. In accordance with this expectation, Duffy antigen expression on RBCs from individuals from Papua New Guinea who were heterozygous for the FY*Anull allele was approximately half of that on RBCs from homozygous Duffy-positive individuals . In the same human population in Papua New Guinea, a 50%, but statistically insignificant, reduction was noted in the prevalence of P. vivax infections among heterozygous FY*A/FY*Anull individuals compared to those who were homozygous Duffy positive . The characteristics of human RBCs with genetically reduced RBC Duffy antigen expression (FyxFyx, according to serotype nomenclature ) and weak reaction with Duffy antigen antisera was investigated using the simian malaria parasite P. knowlesi, which, like P. vivax, is absolutely dependent on the Duffy antigen for the ability to invade human RBCs. Merozoites of P. knowlesi invaded RBCs from such individuals with about half of the efficiency that they invaded RBCs from individuals with full Duffy antigen expression . However, RBCs from heterozygotes for RBC Duffy negativity were invaded as efficiently as were RBCs from normal RBC Duffy-positive individuals (J. Barnwell, personal communication). Susceptibility to infection with P. vivax malaria of FYnull heterozygotes may be partially reduced, but the effect is probably not very strong, certainly by comparison with the total refractoriness to P. vivax malaria of individuals who are homozygous for an FYnull allele. Throughout West and Central Africa, the frequency of the FY*Bnull allele is, as already noted, close to fixation. In most of these populations, its frequency exceeds 97% , leading to almost universal RBC Duffy negativity in these populations. Not surprisingly, P. vivax malaria is very rare throughout the region . It might seem natural to conclude that P. vivax malaria must have been the selective agent for the near fixation of the FY*Bnull allele in these African populations, to the point that P. vivax was itself virtually eliminated. While this is our view, it is not universally held . Nor is the case, as we will now present it, straightforward. In contrast to the frequently and directly lethal P. falciparum, it is often assumed that P. vivax exerts little or no selective pressure on a human population. On these grounds, therefore, and notwithstanding that the RBC Duffy-negative condition confers no evident disadvantage on a carrier , P. vivax might be discounted as a selective force for the near fixation of RBC Duffy negativity in West and Central Africa. If, on the other hand, P. vivax can, and does, exert a significant selective force, we are confronted with a paradox. Why, in this case, has selection for RBC Duffy negativity not taken place in other parts of the world, in southern Asia and the Western Pacific rim , where P. vivax has certainly been present for several thousand years? It has recently been suggested, in a report of the presence of an FY*Anull allele in a single population Papua New Guinea, that such selection may indeed be happening . However, no previous studies appear to have detected this allele in Papua New Guinea, where FY*A is otherwise at virtual fixation . Moreover, at around 2%, the FY*Anull allele in this single population in Papua New Guinea is still far below the frequencies of the FY*Bnull allele, which, throughout tropical Africa, are rarely less than 50% and more often in excess of 95% . Concerning the first of the points above, there is strong evidence that P. vivax has, in fact, often placed a heavy burden of mortality and loss of fecundity on the populations that it afflicted . Its effects are greatest under conditions of relatively low and unstable malaria inoculation rates. These were the conditions that probably prevailed in Africa before 5,000 to 10,000 years ago and possibly throughout much of the preceding 100,000 years. Had P. vivax been prevalent within human populations in West and Central Africa during this period, and we propose that it was, we would expect that the homozygous RBC Duffy-negative condition would have carried a considerable selective advantage and the heterozygous condition a slight one. We would, however, expect that selection for raised frequencies of an FYnull allele in a population under P. vivax pressure should also take a "long" time. This expectation follows if mainly the FYnull homozygotes are refractory to P. vivax infection but rather little selective advantage is associated with the heterozygous condition. This, as we have just discussed, is quite likely to be the case. Now, the homozygous combination of a rare allele, as the FY*Bnull gene in Africa would at first have been, is itself almost vanishingly rare. Moreover, even when the homozygous combination of such an allele had arisen, it would, in the next generation, almost invariably have been diluted again to the weakly advantageous heterozygous state in combination with one of the high-frequency FY*B or FY*B RBC Duffy-positive alleles. Thus, a selection which involved mainly the persistently rare FY*Bnull/FY*Bnull homozygotes would inevitably proceed "slowly." It would proceed slowly, that is, compared, with the speed of selection for a gene such as the allele for hemoglobin S (Hb*S), in which the heterozygous condition, Hb*S/Hb*A, carries the main selective advantage and which, in the early stages of selection, arises as frequently as the Hb*S allele itself. Moreover, the speed of selection for Hb*S would be further increased above that for the FY*Bnull allele, because P. falciparum is probably much more dangerous than P. vivax under any conditions of endemicity. Hb*S has been estimated to approach a balanced equilibrium after around 2,000 years of selection under P. falciparum malaria . The "longer" time expected for P. vivax malaria to select for the near fixation of the FY*Bnull allele could, therefore, be in the order of 10,000 to several tens of thousands of years. While we may infer that the process of selection for near fixation of the FY*Bnull allele in Africa under P. vivax would have taken several tens of thousands of years, we cannot, from the above line of argument, even hazard a guess as to the times when this process may have begun or ended. These may, however, be partly estimated from other evidence. Modern human migrations out of Africa are believed to have taken place largely, if not entirely, within the past 100,000 years. Now, RBC Duffy negativity is, as already noted, an apparently harmless genetic condition. Therefore, once an FYnull allele has been selected in a human population, there should be little or no loss of frequency of the gene, especially from a population in which it had become fixed. Outside Africa, RBC Duffy negativity is found in declining frequencies through the Arabian peninsular, across the Middle East, and to the edges of Central Asia, but beyond these areas it is, with rare exceptions, virtually absent . Had the high frequencies of the FY*Bnull allele been selected before the main dispersals of modern humans began, this allele should be common in all human populations. Since it is not, its selection in Africa must have been completed only after these dispersals had taken place, i.e., within less than the past 100,000 years. Indeed, since human migrations out of Africa across the Old World and into the Western Pacific have probably continued into much more recent times and have carried almost no trace of the FY*Bnull allele with them, we can probably reduce the time by which selection for African FY*Bnull would have been completed to less than 50,000 years, which is the approximate period of the migrations to Melanesia and Australia. And if, as we do, we take it that the force for the selection of the FY*Bnull allele in Africa was P. vivax malaria, then the fact that at least 5,000 years of exposure to P. vivax in southern Asia and China and the Western Pacific has not led to selection for high levels of RBC Duffy negativity anywhere in this region suggests that the process also takes longer than 5,000 years. We now have a case that the selection for near fixation of RBC Duffy negativity takes at least 5,000 years and that it was completed in Africa less than 50,000 years ago. Unfortunately, we still cannot place an upper limit upon how long the process takes. The P. vivax pressure could, for example, have begun several hundred thousand years ago and reached completion in a final rapid spurt at any time within the past 50,000 years, or it could have begun in Africa only 6,000 years ago and been completed in the last few hundred years, just in time for us to record the outcome. However, we now have corroborating evidence for the rough period within which the beginning of the selection would have taken place. In a study of haplotypes associated with the FY*Bnull allele in African and European populations, Hamblin and Di Rienzo have proposed that a selective sweep towards the near fixation of RBC Duffy negativity in the African populations began between 97,200 and 6,500 years ago within 95% confidence limits. We suggest that the most likely period for the selection of RBC Duffy negativity in Africa to have taken place, and therefore for strong selective pressure from P. vivax to have been still active on these populations, lies somewhere between perhaps 10,000 and 5,000 years ago. This falls at the height of the last major glacial period, when equatorial Africa would have been much cooler than today and when it would have been infested with Anopheles mosquitoes, which were at the time relatively inefficient vectors of malaria . These are conditions which would have tended to support low levels of unstable, and, therefore, severely life-degrading, P. vivax transmission. They are the conditions which would have strongly favored selection for RBC Duffy negativity in the affected human populations in Africa. Some Points from the Malaria Hypothesis : How long does it take malaria to select for human polymorphisms? | Human polymorphisms which meet some or all of the expectations of the malaria hypothesis are found in elevated frequencies usually only in populations which have been exposed to malaria for at least several hundred years. Thus, the indigenous populations of northern Europe (with the exception of part of The Netherlands), northern parts of Asia, Australia, and all but the westernmost islands of the Pacific Ocean, and the Americas, from which malaria either has always been absent or has been present for less than this time, are all without these polymorphisms in significant frequencies. Mutations which are moderately protective against malaria, which protect in heterozygous combination, and which carry a relatively low balancing cost to a population, namely, the thalassemias and G6PD deficiency, are generally the first to reach elevated frequency under selection by malaria. These are the malaria-selected polymorphisms which predominate in the parts of Europe, around the shores of the Mediterranean Sea, where malaria arrived between 2,500 and 2,000 years ago. In northern Europe, where it has arrived more recently, probably mainly between 1,000 and 500 years ago, even these polymorphisms are almost entirely absent. In regions in which more than 2,000 to 3,000 years of selection by malaria has taken place, namely, in Africa, parts of the Middle East, southern Asia, and Melanesia, mutations are found which are highly protective against malaria in heterozygous combination but which also carry a high balancing cost in homozygous combination. These are the genes for sickle cell hemoglobin and ovalocytosis. Longest of all, many thousands or tens of thousands of years, may be the selection for alleles whose protective effects against malaria are expressed only, or mainly, when they are present in homozygous combination. Such is the allele for hemoglobin C, which, in homozygous combination, protects strongly against P. falciparum malaria and which has probably not yet reached equilibrium frequencies even after several thousand years of selection in West Africa. Such, also, is the allele for RBC Duffy negativity, which, only in homozygous combination, protects solely and absolutely against P. vivax malaria. High frequencies of RBC Duffy negativity extend from West and Central Africa and, in declining frequency, toward East Africa and southern Africa as well. These regions, and especially West Africa, are the heartlands from which most, if not all, of today's populations of human malaria may have had their origin and in which, therefore, selection under both P. vivax and P. falciparum has been of longest duration. Did malaria enter most human populations before or after their global dispersal? : In each broadly grouped human population in which they occur, the malaria-protective polymorphisms are usually distinct from each other. Hemoglobin E occurs only in Southeast Asia, and ovalocytosis occurs in Melanesia and also, probably in several different forms, through Southeast Asia. Hemoglobin S is absent from these regions, and hemoglobin C occurs only in West Africa. The variety of different G6PD deficiency mutations found in, and characteristic of, different human populations is large, and for the thalassemias it is even greater . The G6PD deficiency mutations are associated with a number of different haplotypes in different human populations , and the sickle cell gene occurs in association with several different population haplotypes within Africa and with three distinct haplotypes in the Middle East and India . Thus, individual polymorphisms lie within haplotypes which are usually characteristic of the human populations within which the polymorphism is found. These facts can be most easily explained if each indigenous human population had separated and settled before the selective agents, P. falciparum and P. vivax malaria, arrived within it. This, as will be analyzed in the following sections, is almost undoubtedly the general pattern of events. Legacy of the impact of malaria on human populations. : Regardless of its implications for our understanding of the evolution and history of malaria, the malaria hypothesis points to the public health reality that our past contacts with malaria have left a large burden of genetic diseases among us. Today this genetic burden, in the form of the thalassemias, sickle cell disease, G6PD deficiencies, and the ovalocytoses, is of the same order as the residual burden of malaria itself. About one-third of a million to half a million babies are born each year with severe forms of these inherited disorders . ORIGINS AND EVOLUTION OF HUMAN MALARIA PARASITES : Origins of Malaria Parasites | If settlement by most human populations preceded the arrival of malaria within them, then when, where, and how did the ancestral malaria parasites arise and when, where, and how did they enter human populations? The ancestors of the malaria parasites have probably led a parasitic existence almost since there were potential hosts to parasitize, at least half a billion years ago. Of course, the parasitic forms in this lineage must themselves have had a preparasitic ancestor. Molecular genetic evidence strongly suggests that this ancestor was a choroplast-containing, free-living protozoan which become adapted to live in the gut of a group of aquatic invertebrates . This single-celled organism probably had obligate sexual reproduction. Indeed, to the present time, all known members of the phylum Apicomplexa, to which the malaria parasites belong, have retained an obligate sexual stage in their life cycles . Moreover, the environment where gamete formation and fertilization takes place is still always within the midgut lumen of a host species. At some relatively early stage in their evolution, these "premalaria parasites" acquired an asexual, and usually intracellular, form of reproduction called schizogony. By evolving this form of "vegetative" reproduction, the parasites greatly increased their proliferative potential. Schizogony may be defined as growth and subsequent nuclear and cellular division into numerous daughter cells from within the body of a single parent cell. It almost invariably takes place during the intracellular phases of parasite growth in the tissues of a host. The completion of each schizogonic event is usually followed by reinvasion of new host cells by the daughter cells or spores (designated sporozoites or merozoites), leading to another round of schizogonic development. In this way, numerous successive rounds of multiplication can take place. It is during schizogony in the RBCs of humans and other vertebrates that certain descendants of these ancestral parasites now cause the disease that we call malaria. Among the invertebrates to which the ancestors of the malaria parasites became adapted were probably aquatic insect larvae, including those of early Dipterans, the taxonomic order to which mosquitoes and other blood-sucking flies belong. These insects first appeared around 150 million to 200 million years ago. During or following this period, certain lines of the ancestral malaria parasites achieved two-host life cycles which were adapted to the blood-feeding habits of the insect hosts. The malaria parasites of humans are typical products of this line of development since they alternate between human and the blood-feeding female Anopheles mosquito hosts (within whose midguts gamete formation and fertilization still, of course, take place). In the 150 million years since the appearance of the early Diptera, many different lines of malaria and malaria-like parasites evolved and radiated. They came to parasitize members of most major groups of land vertebrates including reptiles, birds, and mammals. In the mammalian orders today, more species of malaria parasite have been identified among primates than in all other mammals combined . Over 25 distinct species of malaria parasites of primates have been named. Four of these are the recognized malaria parasites of humans: P. falciparum, P. vivax, P. malariae, and P. ovale. Origins of Human Malaria Parasites : There has been much speculation concerning the evolutionary relationships of the four human malaria parasite species both to each other and to the malaria parasites of other animals. In the past, these discussions depended on observations on the morphology and biology of the parasites and on the evolutionary relationships among, and the geographic locations and origins of, their various host species. More recently, molecular genetic analysis of the parasites themselves has contributed a new dimension to this detective work . On the basis of such molecular genetic evidence , as well as on classical biological grounds , P. falciparum is very closely related to, but nevertheless evolutionarily divergent from, a malaria parasites of chimpanzees, P. reichenowi. Curiously, molecular evidence also shows that these two parasite species are more closely related to the malaria parasites of birds than they are to those of other mammals as, indeed, had long been suspected by classical biologists . The divergence of the lineage of the bird malaria parasites, including P. falciparum and P. reichenowi, from the line that gave rise to all other known and tested malaria parasites of mammals is an ancient event, possibly dating to around 130 million years ago . This would place their separation near the very origin of the two-host life cycle involving blood-feeding Dipterans and land vertebrates. By contrast, the separation of the lines that led, on the one hand, to P. falciparum and, on the other, to P. reichenowi has been placed at only 4 million to 10 million years ago . This overlaps the period, around 5 million years ago, in which the human line diverged from that of the African great apes . The three remaining species of human malaria parasite, P. malariae, P. ovale, and P. vivax, fall within a single clade that includes all the mammalian malaria parasites, other than P. falciparum and P. reichenowi, on which molecular genetic analysis has been done . Within this clade, all of the primate malaria parasites belong to one or other of the lineages represented by P. malariae, P. ovale, or P. vivax. These three lines appear to have diverged over 100 million years ago , long before the emergence of the lines leading to the distinct mammalian, let alone primate, orders of today. Among the primate malaria parasites, P. ovale is known only as an infection in humans and has no genetically confirmed close relative . P. ovale is therefore the sole known surviving representative of its line . P. malariae, on the other hand, in addition to infecting humans, is found in apparently indistinguishable form as a natural parasite of chimpanzees in West Africa . Moreover, a morphologically indistinguishable parasite, P. brazilianum, infects New World monkeys in Central and South America . Molecular genetic analysis has failed to distinguish P. brazilianum and P. malariae . There can be no doubt that these are the same or almost the same parasite infecting different primate species, namely, humans and New World monkeys. We take the view that P. brazilianum is, in fact, a zoonotic form of P. malariae which has been introduced recently from humans into the New World monkey populations. The converse of this, i.e., that P. malariae has been received by lateral transfer from New World monkeys to humans, has also been argued . However, a New World origin of this parasite does not reconcile with the common prevalence of P. malariae in the Mediterranean region more than 2,000 years ago . Virtually all, if not all, pre-Colombian human migrations of at least the past 50,000 years appear to have been from the Old to the New World, and none have been in the reverse direction. Therefore, had P. malariae originated in the Americas by lateral transfer from monkeys to humans, when, and by what route, could it have entered the Mediterranean region to have become prevalent there at least 2,000 years before any known or likely human traffic with the New World? Except for P. brazilianum, no other primate malaria parasite is known to belong within the P. malariae lineage. Thus, whether as a human parasite or, in essentially identical form, as a parasite of New World monkeys, P. malariae/P. brazilianum, like P. ovale, is the only confirmed and extant representative of its line. In contrast to these lonely survivors of ancient lines, P. vivax belongs within a large and closely related body of malaria parasites which today are found mainly infecting monkeys of southern and southeastern Asia and the Western Pacific rim . Among these parasites, P. cynomolgi and its subspecies have long been considered especially closely related to P. vivax . Because these close relatives of P. vivax are today found only infecting monkeys of the southern Asian regions, there is a view that P. vivax itself must have originated from within these regions and from nowhere else (see, e.g., reference ). However, this conclusion may not necessarily be so simply reached. The most recent estimate for the time of the divergence of P. vivax and P. cynomolgi is in the range of two million to three million years ago (A. Escalante personal communication). At this time, the ancestral hosts of these parasites were probably spread throughout much of the tropical, subtropical, and, indeed, temperate regions of the combined Asian and African landmasses. Therefore, the region(s) where the P. vivax and P. cynomologi lines diverged and hence where P. vivax, in this sense, originated cannot, solely on these grounds, be located more precisely than to somewhere on the continents of the Old World. There are two other parasites of primates which, like P. cynomolgi, are morphologically indistinguishable from P. vivax. One is a parasite of chimpanzees in West and Central Africa and is named P. schwetzi . In contrast to P. cynomolgi, sporozoite infections of which readily induce blood infections in humans, various attempts to infect humans with P. schwetzi have all, apparently, failed . It seems appropriate, therefore, to regard P. schwetzi as a separate species from either P. vivax or P. cynomolgi, although it is probably closely related to both. Unfortunately no molecular genetic data yet exist with which to test these relationships. The other parasite which is morphologically identical to P. vivax is P. simium, a parasite of New World monkeys. Molecular genetic analysis shows that P. simium and P. vivax are, indeed, indistinguishable . We take the view that P. simium is almost certainly an enzootic form of P. vivax which was introduced into New World monkeys following the arrival there of humans. Our argument parallels that given above for P. brazilianum and P. malariae. We also draw on the following. The time of divergence of P. vivax from the closely related parasite of Old World monkeys, P. cynomolgi, is, as already noted, put at two million to three million years ago. Now, South American and Old World simians diverged from each other around 40 million to 50 million years ago and have been geographically isolated from each other probably throughout this time . Therefore, the divergence within the past three million years of two lines, one of which, P. cynomlogi, is unquestionably the parasite of an Old World host species, must itself have taken place in the Old World. It follows that P. vivax had its origin in the Old World. P. vivax could not, therefore, have derived from P. simium; the converse, in fact, must be the case. FIG. 1. | Phylogeny of the malaria parasites of humans and of some other related malaria parasite species. Phylogeny of the malaria parasites of humans and of some other related malaria parasite species. See the text for sources on which this reconstruction is based. EXPANSIONS AND DISPERSALS OF HUMAN MALARIA : In spite of the difficulties in trying to identify the region, or even the continent, of origin of the ancestors of extant species of human malaria parasite, the attempt is worthwhile. In the following, we discuss the general proposition that the modern populations of each of the four species of human malaria parasite grew to prominence mostly in Africa and that it is largely, if not entirely, from out of Africa that these populations have dispersed. Out of Africa : We have argued that P. malariae was introduced from humans into New World monkeys and not vice versa. P. malariae is a natural parasite of African great apes and humans in Africa. It is found in no other Old World primate species within or outside Africa. On these grounds, the simplest hypothesis for the origin of P. malariae is that it was in Africa itself. As a parasite of the ancestor of both humans and African great apes, ancestral P. malariae would have continued to parasitize and cross-infect both host lineages as they diverged around five million years ago. How soon thereafter may P. malariae have been carried by its hominid hosts beyond its African homeland is an open question. P. malariae thrives under both tropical and temperate transmission conditions and is adapted to endemicity in sparse and mobile human populations (see below). Therefore, it could well have been prevalent among hominids migrating from Africa to the warmer regions of Europe and Asia at any time in the past five million years. P. ovale, by contrast, is a strictly warm-climate parasite. Today it is found commonly throughout tropical Africa and in very limited distribution elsewhere in the tropics, notably in New Guinea and in the Philippines . There is no reason not to suppose that it, too, had its origin as a human parasite in Africa. Because P. falciparum and P. vivax each have a close biological relative (P. reichenowi and P. shwetzi, respectively) which is a parasite of African great apes, an argument, similar to that for P. malariae, can be made for the African origins of both of these species. In contrast to the case for P. malariae, however, the argument concerning these two human-ape parasite pairs requires that, as the human and ape lines diverged, so also did the parasites they carried. In one case, this would have led to P. vivax in the human line and to P. schwetzi in apes, and in the other case, it would have led to P. falciparum in humans and P. reichenowi in apes. Molecular genetic data on the present-day populations of P. falciparum support the proposition that they did, indeed, spread from an origin in sub-Saharan Africa . As already discussed, the evidence from human genetic data imply a West African origin of P. falciparum within the past few thousand years. This is also, according to other population genetic analyses of P. falciparum, the time within which modern populations of the parasites had their origin (see below). The case for an African origin of most recent P. vivax populations rests entirely on the evidence from the global distributions of RBC Duffy negativity in human populations. The extremely high frequencies of RBC Duffy negativity in African populations and its almost total absence beyond Africa provide strong evidence that the greatest and/or the longest burden of P. vivax malaria was that experienced by African populations . If global populations of all four species of human malaria parasites derive from African sources, how were their characteristics formed there and why, at particular times, did they disperse beyond Africa? Everything that took place was certainly governed by the climatic events of the past 100,000 years. These were dominated by cooling of the planet and by the massive glaciations which affected much of the northern hemisphere. The last glacial period, which was at its coldest about 20,000 years ago, ended with the rapid warming of the planet to approximately present-day temperatures by about 10,000 years ago. The warmer global climate heralded the beginnings of agriculture. According to one appealing line of argument, the entry of agricultural practice into Africa was pivotal to the subsequent evolution and history of human malaria . The Agrarian Revolution in Africa : Around 4,000 to 5,000 years ago, the Neolithic agrarian revolution, which is believed to have begun in the "Fertile Crescent," southern Turkey and northeastern Iraq, 8,000 years ago, took root in western and Central Africa. In response to the profound changes to the human environment that this engendered, organisms in contact with that environment evolved to adapt. Among these were the Anopheles vectors of human malaria. Adaptations of the vectors of malaria in Africa. : Outside of Africa, wherever malaria is or has been endemic, the female mosquito vectors of malaria are zoophilic rather than anthropophilic; i.e., they prefer to feed on animals rather than on humans . The degree of preference for human blood has important consequences for human malaria transmission . All other factors being equal, the probability of a malarial infection being transmitted from one human to another is proportional to the square of the fraction of blood meals that the vector mosquitoes take on humans . For example, mosquitoes that take only 1/10 of their blood meals on humans will transmit a malarial infection 100 times less frequently than mosquitoes that take all their blood meals on humans. In most parts of the world, the anthropophilic index (the probability of a blood meal being on a human) of the vectors of malaria is much less than 50% and often less than 10 to 20% . By contrast, in sub-Saharan Africa, the vectors of human malaria usually have an anthropophilic index of 80 to almost 100% . This is probably the most important single factor responsible for the stability and intensity of malaria transmission in tropical Africa today. The high human-biting preferences of African Anopheles, as Livingstone was the first to suggest, may have been an adaptation to conditions created by the agrarian revolution in Africa. In adopting an agricultural way of life, human populations in sub-Saharan Africa changed from a low-density and mobile hunting and gathering life-style to communal living in settlements cleared in the tropical forest. This new, man-made environment had two important consequences for the mosquito populations. One was that the numbers and densities of humans began to increase under the new agricultural economy. The other consequence was that the new life-style generated numerous small water collections close to the human habitations. Those who adopted agriculture thus transformed themselves into large, stable, and accessible sources of blood in the midst of abundant mosquito-breeding sites. The new situation provided a strong selective advantage to mosquito populations which became adapted to breed close to human habitation and to feed primarily on human blood. Livingstone argued that this led to the very high anthropophily of the vectors of African malaria and, in large part, their great vectorial efficiency. Agricultural village economies had, of course, also developed throughout the tropics and subtropics of Asia and the Middle East up to several thousand years before those in Africa . Outside Africa, however, malaria vectors have never acquired the same extraordinary preference for human blood. Why should this be? The answer probably lies in the abundance in Asia of animal species whose domestication was possible, and was achieved, during the rise of agriculture. Because there were none in sub-Saharan Africa , malaria vectors with a preference for human blood would have had a strong selective advantage in Africa but nowhere else. Parasite adaptations before and following the agrarian revolution. (i) P. malariae, the "gypsy" parasite. : P. malariae is the human malaria parasite which would have been best adapted to preagricultural conditions. It could have been, and probably was, sustained at very low infection rates among the sparse and mobile human populations of the time. This is because, unlike other human malaria parasites, P. malariae can remain for decades within a human host and in a state potentially infectious to mosquitoes . As a result, infections of P. malariae can be transmitted many years after they have been contracted. It is the perfect adaptation to the parasitization of wandering hosts and mosquito vectors that rarely encounter each other. Because of low, or very low, infection rates, the burdens of P. malariae on early hunters and gathers may have been relatively light. However, in those who were affected, P. malariae could have imposed a severe long-term burden of morbidity and mortality through the eventually severe and often fatal nephropathy that it can cause . (ii) P. vivax, the great survivor, and P. ovale, the "hothouse plant." : As human numbers slowly increased in preagricultural Africa, so too would have the prevalence of P. ovale and P. vivax. Both parasites are more dependent than P. malariae on denser and less mobile host populations. Under the still low and unstable preagricultural transmission conditions, both P. vivax and P. ovale malaria would have begun to impose significant burdens of morbidity and mortality on the affected human populations. Gradually, however, the frequency of the mutation for RBC Duffy negativity would have begun to rise until, as it approached fixation, P. vivax malaria itself declined to very low levels in West and Central Africa. However, it was not, and has not been, extinguished there . Moreover, P. vivax was to spread beyond Africa and throughout almost the entire habitable zones of the planet. Both within Africa and beyond it, P. vivax has proved itself to be the most tenacious survivor of all the malaria parasites of humans . With no equivalent of homozygous RBC Duffy negativity to repel it, P. ovale remains highly prevalent throughout tropical Africa. Elsewhere, however, it is found only in pockets, all within the tropics. In stark contrast to P. vivax, P. ovale has little resilience beyond its ancestral habitat and none at all outside the tropics and subtropics. While both P. vivax and P. ovale remain prevalent across Africa, there is little or no trace of the burden they would once have imposed there. Intense and stable transmission induces strongly protective antimalarial immunity in early life. It thereby prevents the devastating experience of lifelong, lingering infection with malaria. Therefore, the arrival of agriculture and the intense transmission rates that this engendered would have largely eliminated the mortality, and most of the morbidity, due to these parasites in Africa. (iii) P. falciparum, the parasite that "did not exist". : Wherever transmission conditions are poor, P. falciparum is the least competitive of the four species of human malaria parasite. It can be, and has been, exterminated under conditions which have allowed the far more resilient P. vivax to remain endemic . How, therefore, could a parasite as vulnerable as P. falciparum have been sustained at all by inefficient mosquito vectors, among sparse and mobile human populations, and during the climatically cool era that preceded the arrival of agriculture in Africa? The answer, according to some molecular genetic evidence, may be that P. falciparum in its present form had not yet arisen. Synonymous (not associated with a corresponding amino acid substitution) SNPs have been investigated in protein-coding nuclear genes , from intragenic noncoding, nuclear DNA , and from mitchondrial genes in a number of independently isolated and geographically separated lines of P. falciparum. According to evolutionary genetic theory , the number of synonymous SNPs that may be found between genes from among the members of any species of organism increases linearly with time. From an analysis of the densities of such SNPs within its genome, and with appropriate assumptions, including the absence of selection pressures for or against these mutations, it is possible to extrapolate back to a time when an extant species expanded from a small ancestral population. This has been done in a number of studies of P. falciparum. The results of those referred to above suggest that modern populations of P. falciparum emerged within relatively recent times, and possibly within the past 5,000 to 10,000 years, from a tight population bottleneck . According so some , the bottleneck could have been so tight as to have been represented by a single clone of the parasites. Other published molecular genetic evidence points to a longer time since any population bottleneck ---up to half a million years . It has also been suggested that properties of the P. falciparum DNA sequences used as a source of data may strongly preclude the survival of synonymous mutations and that this, therefore, would lead to considerable underestimations of the age of the present-day global population of these parasites . Another line of counterargument has pointed to the existence of the complex alleles of the gene for the MSP1 protein of P. falciparum whose sequences differ at many tens of amino acid positions. Such complex allelic differences in an extant population must be of very ancient origin, it is argued, and would rule out the passage of a haploid organism such as P. falciparum through a recent clonal bottleneck . While a clonal bottleneck may be considered extremely improbable, none of these counterarguments are conclusive in ruling out the possibility that some sort of population bottleneck did occur in the recent history of P. falciparum. There is, moreover, a concordance between the estimates for the more recent emergence of P. falciparum within the past 5,000 to 10,000 years, as offered above, and estimates for the length of time that P. falciparum has exerted selection on human populations in West Africa. As discussed above, this, too, is probably less than 10,000 years. Moreover, the emergence of P. falciparum into the human population in tropical Africa, somewhere within this period, would resolve the enigma of how it could have been sustained under the conditions of preagricultural Africa. In its present form, P. falciparum would simply not have existed. Where, then, did the present-day parasite come from? It has been argued that there must have been in preagricultural Africa an ancestral population of "proto-P. falciparum," perhaps as a human parasite or in another host species. Following the advent of agriculture, mutant parasites, increasingly adapted to conditions which had become so favorable to the transmission of human malaria, would have begun to replace, or sidestep, their progenitors in a succession of epidemic bursts . In this way, via a series of population bottlenecks that would have purged the rapidly evolving parasites of most genetic diversity except at loci where selection retained it, modern P. falciparum emerged and spread across Africa. Ironically, and poignantly, had the ancestral material from which modern P. falciparum was to arise not existed, African agriculture should have reduced, rather than increased, the burdens of malaria in Africa. It would have done this by stabilizing malaria transmission rates, thereby eliminating the lethality of chronic malaria through the induction of early protective immunity. But somewhere on that continent, proto-P. falciparum did exist . In the new environment of intense malaria transmission, this parasite transformed to become the agent of an infection that was frequently lethal in the acute, as opposed to the chronic, condition. The intense transmission that induced early immunity simply concentrated the mortality due to P. falciparum malaria into the youngest age groups. P. falciparum had become a killer of human beings which, historically and still today, has few rivals (Tables , , and ). TABLE 3 | Malaria mortality and human population by geographical region for the 20th century TABLE 4 | Malaria mortality: summary statistics at the beginning and end of the 20th century Beyond Africa : In the Old World. | Prior to the agrarian revolutions, only P. malariae, by virtue of its transmission characteristics, might have been present among the low-density hunter-gatherer populations of the Old World. Here, indeed, P. malariae may already have been present for perhaps hundreds of thousands of years or much, much longer ---perhaps since before the origin of the human line. The failure of RBC Duffy negativity to have reached even moderately high frequencies almost anywhere outside Africa argues against the long presence of P. vivax in human populations beyond this continent. P. vivax may not have existed in Asia or, indeed, anywhere else outside Africa until relatively recently. Indeed, it may not have spread beyond Africa before 10,000 to 30,000 years ago. Perhaps this spread occurred within the 15,000 years since rising global temperatures began the last glacial recession. Perhaps it was not until after the rise of agriculture in the Fertile Crescent around 7,000 to 8,000 years ago. But whenever it was that P. vivax began to spread beyond its African homeland, today, a gradient of declining RBC Duffy negativity, like a serpent's trail in the sand, marks its early contact with Arabia and the Near East . By the time modern P. falciparum emerged into Africa around, as we now conjecture, 4,000 years ago, P. vivax may have already spread far beyond its "native" shores and into the agricultural communities that were expanding in size and numbers throughout the warm southern regions of the Eurasian continent and the Western Pacific. In the same period, P. ovale would also have made its more tentative way beyond Africa into a few other tropical safe havens, such as New Guinea. These same environments, rich in human settlement and sustaining flourishing populations of Anopheles, were also well prepared for, and receptive to, the arrival and transmission of P. falciparum. As it entered, P. falciparum would have struck each new human settlement in its most lethal form, the epidemic. Thereafter, and depending on the local conditions, P. falciparum would have settled in many areas to a state of more or less stable endemicity while continuing to exact high perennial mortality. This, in turn, would gradually lead to the selection of elevated frequencies of thalassemias and G6PD deficiency and eventually to traits such as sickle cell and ovalocytosis that characterize the descendent populations today. In the New World. : Two human malaria parasite species, P. malariae and P. vivax, bear a remarkable morphological and molecular genetic similarlity to two malaria parasites of monkeys of Central and South America, respectively: P. brazilianum and P. simium. We have argued that these parasites were almost certainly introduced into the New World monkeys from human infections, and not the reverse. When, therefore, and by what route did human malaria parasites enter the Americas? The story is well known of how, in the 17th century, native Peruvians introduced Spanish colonists to the effects of Cinchona bark for curing malarial fevers. This story has been offered as evidence that Plasmodium species infecting humans were already present in the Americas before the arrival of Europeans. As evidence for such presence it is, however, far from conclusive. The transfer of the knowledge of the antimalarial properties of Cinchona bark took place more than a century after the first arrival of the Spanish in South America. This is time enough for malaria to have spread far across that continent had it been introduced there by the Europeans. If the bark was already in use among the indigenous people, discovery of its antimalarial properties may have followed rapidly. The pre-Columbian populations of the New World are almost totally devoid of any of the genetic abnormalities which are associated with long exposure to malarial infection . They are therefore unlikely to have had very long experience of P. vivax and almost certainly none of P. falciparum, the two parasite species that we are confident can, and do, select for these polymorphisms. We are unsure whether P. malariae can do this, and we are less confident, on these grounds, that this species was not present in the Americas from pre-Colombian times. P. malariae can persist for up to several decades within the human body. Conceivably it could have survived in individual human carriers arriving from Asia by overland trek or sea voyage of years or decades through the malaria-sterile sub-Arctic north or the Anopheles-free Pacific Ocean. By contrast, entry of P. vivax, let alone P. falciparum, by the northern route must surely have been precluded. Could either of these species have survived a sea voyage or, more probably, a succession of voyages across half of the diameter of the planet from southern Asia or its offshore archipelagos to the eastern coasts of the Americas? It may just be possible. Thus, while it is at least conceivable that P. malariae may have reached the Americas before, and perhaps long before, the arrival there of European and African peoples, it seems unlikely that P. vivax or, especially, P. falciparum could have done so. The post-Colombian landings of infected Europeans and Africans, on the other hand, through the mediation of the many naturally receptive species of American Anopheles, would readily have sparked first local and then regional epidemics of P. vivax and P. falciparum. Both species would eventually spread and become endemic through most parts of the American continents from the Atlantic to the Pacific and from Canada to Argentina. P. malariae, whether it was first on the scene or not, was to become extremely rare in the New World. P. ovale may never have even gained a foothold . In the Western Pacific. : Until its transient introduction into the northern Australian tropics in the 19th and early 20th centuries, neither the Australian continent nor the islands of the Pacific Ocean beyond the "Buxton line" (south of the equator, eastward of 170 east, and south of 20 south, and north of the equator to the Tropic of Cancer, eastward of 130 east, and north of the Tropic of Cancer, eastwards of 135 east) had probably ever experienced malaria transmission. To the west of the Buxton line lie the highly malarious islands of the Western Pacific Ocean, including New Guinea and Melanesia, Indonesia and the Philippines, and the transiently malarious northern Australia. However, elsewhere across the Pacific Ocean in the huge area beyond the Buxton line up to the west coasts of the Americas, not even the Anopheles vectors of human malaria occur . Indeed, the total absence of malaria from these tropical Pacific islands long endowed them with one of the most salubrious environments on Earth . As in the Americas, genetic polymorphisms attributed to selection under malaria are virtually absent from indigenous populations throughout this vast region . MALARIA IN HISTORICAL TIMES : On the basis of general scientific evidence, we have tried, in the preceding sections, to portray human malaria as and where it may have been largely in our prehistoric past. In the following, we attempt to trace it, and its effects, through the period of historical record. From the "Dawn of History" to the Nineteenth Century : Malaria seems to have been known in China for almost 5,000 years. The Nei Ching (The Canon of Medicine) from 4,700 years ago, apparently refers to repeated paroxysmal fevers associated with enlarged spleens and a tendency to epidemic occurrence . This suggests P. vivax. P. malariae is unlikely epidemic material. The absence of reference to "deadly effects" of these epidemics argues against the involvement of P. falciparum, and the date precedes the likely time of the emergence of P. falciparum out of Africa. This evidence indicates, therefore, that P. vivax, whether from out of Africa or from elsewhere, was by this time present across the Old World. From the Indus valley in northern India, Vedic (3,500 to 2,800 years ago) and Brahmanic (2,800 to 1,900 years ago) scriptures contain many references to fevers, some of which are said almost certainly to concern malaria . In the Atharva-Veda there is, for example, an incantation to Takman, the fever demon, which runs as follows. "To the cold Takman, to the shaking one, and to the deliriously hot, the glowing, do I render homage. To him that returns on the morrow, to him that returns for two successive days, to the Takman that returns on the third day, shall homage be" . Tertian (P. vivax) and quartan (P. malariae) fevers are mentioned in the Vedas . They are also said to make reference to autumnal fevers as the "king of diseases" . In much later times, autumnal fever is certainly P. falciparum malaria . Hence, we may infer that P. falciparum had reached India by around 3,000 years ago. References to conditions that bear several of the hallmarks of malaria ---fevers, intermittent and otherwise, and splenomegaly ---have been quoted from Sumerian and Egyptian texts dating from 3,500 to 4,000 years ago and thus from slightly before the earliest Vedic writings. The Sumerian records apparently make frequent reference to deadly epidemic fevers . Could these have been the first onslaughts of P. falciparum from out of Africa? Although there seems to be nothing in the descriptions more definitively to identify these outbreaks as malaria, the marshy lands between the Tigris and the Euphrates, where Sumer stood, would have been prime environments for malaria transmission. Its long-term impact, however, could not have been that devastating. Mesopotamian civilizations flourished, rising and falling in vigorous, mutually inflicted succession, throughout historical antiquity. But the region would certainly have remained generally malarious. Over 2,000 years later, in 323 B.C., beyond Mesopotamia on the route to India, Alexander the Great is said to have died of malaria. If true, in a young, fit and healthy man, it would almost certainly have been P. falciparum malaria. By this time, malaria had already appeared in the writings of the Greeks from around 500 B.C. Characteristically it seems to have coincided with human pioneering and land clearing for agriculture . Before this time, references in Greek writings to recurring fevers seem to have been conspicuous by their absence . The works attributed to Hippocrates (460 to 377 B.C.), however, leave no doubt about the presence of malaria in Greece in its benign tertian (P. vivax), quartan (P. malariae), and malignant subtertian (P. falciparum) forms. Malaria seems not to have reached mainland Italy until the second century B.C. . For a time, the marsh-ridden countryside around Rome, the Campagna, became a virtual desert, and by early imperial times Sicily and Sardinia had become notorious for summer and autumnal fevers. However, under the prosperity of the Roman Empire (circa 50 B.C. to 400 A.D.), by drainage, husbandry, and building development, malaria was excluded for several centuries from the Roman Campagna itself . Then, as the Empire declined and the Campagna fell into ruin, Dark Age and early medieval Rome became surrounded, once more, by the dangerous and largely uninhabitable marshes known and feared by later generations . These would have been hotbeds of P. vivax and especially P. falciparum. It is noteworthy, however, that in the one-and-one-half millennia until the present time, periods of several centuries of high malaria incidence on the Campagna seem to have alternated with similar periods when malaria was apparently absent . These long episodes of absence and then presence of malaria have been associated with corresponding rising and falling agricultural and economic prosperity . They reflect, yet again, the dependence of malaria on prevailing human activity and life-style. By the beginning of the Christian era, malaria was widespread around the shores of the Mediterranean, in southern Europe, across the Arabian peninsular, and in Central, South, and Southeast Asia, China, Manchuria, Korea, and Japan up to 40 of northern latitude. Malaria probably began to spread into northern Europe in the Dark and Middle Ages. Its path would have lain via France and Britain to the countries of the North Sea seaboard and by this route, and also from the Balkans, into central and eastern Europe. From here it would have crossed into southern Sweden and the Baltic states and thence, and also from the south from the Ukraine, the Caucases, and Central Asia, into Russia to the Volga basin and the Urals. Its progress was, however, probably uneven. In Britain, potential malaria hot spots, such as the English fens, were at first protected by Roman technology . Not, it seems, until late medieval times were these and other marshy areas of England, the Somerset levels, and the estuaries of the Rivers Solway and, above all, the Thames and Medway to become severely and enduringly blighted by endemic malaria . Meanwhile the arrival of Europeans and West Africans in the New World at the end of the 15th century A.D. introduced P. vivax, P. falciparum, and, also, perhaps, P. malariae for the first time. At first the malarious heartlands of the Americas were in the Caribbean and parts of Central and South America . From the mid-18th century, however, accompanying the economic growth of the Southern States of North America based on slaves brought from West Africa, malaria took firm hold across the North American continent. Throughout the next 100 years, as colonists moved westward across the United States of America and Canada, breaking the ground for agriculture, malaria sprung from the new-claimed land. By around 1850 A.D., it prevailed through the length and breadth of the tropical, subtropical, and temperate regions of the two American continents. From the time of the voyages of Columbus until the mid-19th century, European trade and colonization in the tropics were marked by enormous losses of life from infectious disease. More deaths were registered as recognisably due to malaria than to any other disease, although in specific circumstances and locations the tolls from yellow fever and dysentery were often greater (, -, , ). In the worst locations, as on the coasts of West Africa, mortality rates often exceeding 50% of a company per year of contact were the norm . From the mid-19th century onward, a chemically purified antimalarial component in Cinchona bark, quinine, entered widespread use among Europeans in West Africa. Overall mortality rates fell rapidly to less than one-quarter of those before the introduction of quinine. It showed the extent to which the previous mortality had, indeed, been the result mainly of malaria . At some time during the 19th century, malaria reached its global limits. In absolute numbers and in the proportion of the humanity now affected, malaria was exacting its highest ever toll of sickness and death. Well over one-half of the world's population was at significant risk from malaria. Of those directly affected by malaria at least 1 in 10 could expect to die from it. The prosperity and well-being of all who lived within its reach were reduced greatly and usually catastrophically. Then, from toward the end of the 19th century, throughout North America and Northern and Western Europe, malaria entered an inexorable decline toward its present extinction in these regions. Elsewhere, however, and especially in Asia, malaria's high tide was yet to come. In the Twentieth Century : The following section is based on material prepared for "The World Health Report, 1999" . Details of sources and the approach to the analysis are given in the Appendix to this article. The data analysis itself is presented in Tables and and in Fig. to . The mortality from malaria has, in recent decades, been due almost entirely to P. falciparum. However, in the early 20th century, repeated untreated infections of P. vivax, and prolonged infections of P. malariae, with their associated nephrosis, would have contributed significantly to the mortality due to malaria in Asia and in parts of the Americas and Europe. Throughout the century, global malaria-related deaths may never have fallen below half a million per year; they were probably at least 3 million per year during most of its first half. These figures suggest that somewhere in the region of 150 million to 300 million people have died from the effects of malaria during the past 100 years. If it is taken that around 6,000 million people have died during this period, malaria may be reckoned to have been a factor in between 2 and 5% of all deaths across the planet in the 20th century. It is possible to attribute up to 10% of global deaths to malaria in the early part of the century. In India it probably accounted for over half . FIG. 2. | Malaria mortality in the 20th century. Malaria mortality in the 20th century. The graph shows the total number of deaths due to malaria per year in Europe and North America ( ---); the Caribbean and Central and South America ( --- ---); sub-Saharan Africa (* --- ---*); China and Northeast Asia (X ---X); the Middle East, South Asia, and the Western Pacific ( --- ---); and worldwide (---) (see Appendix and Table ). FIG. 5. | Malaria mortality in the 20th century. Malaria mortality in the 20th century. The graph shows the number of malaria deaths per 10,000 population per year worldwide minus sub-Saharan Africa (lozenge--lozenge) and sub-Saharan Africa (* --- ---*) (see Appendix and Table ). Malaria in Europe and the Americas. : Within Europe and North America, the malaria problem at the start of the 20th century was already diminishing. The process of the recession of malaria from Europe from the mid-19th century onward has often been represented as something of a mystery. The circumstances are, however, not difficult to explain. Beginning in northwestern parts of Europe, in England and The Netherlands, human health and rural environments and living conditions, and especially housing, were beginning to improve rapidly, so that contact between humans and vector mosquitoes was in decline. In addition, quinine was becoming cheap and widespread. The conditions for malaria transmission were therefore being steadily eroded, while at the same time the human hosts were becoming increasingly fit to control their infections. Nevertheless, as the 20th century began, large areas of Europe and North America were still affected. These included the southern and midwestern United States, much of southern Europe, including the Mediterranean and Balkan countries, parts of northern Europe, still including areas of England and The Netherlands, and much of central Europe and southern Russia. Within the global context, however, the numbers of malaria deaths in Europe and North America, at a few tens of thousands per year, were relatively small (Table , row 1; Fig. and ). Throughout the early part of the century, indigenous human malaria in Europe and North America continued its recession. This, as we have seen, was mainly as a result of the spontaneous decline in contact between human and vector populations, the by-product of rising prosperity. However, it was strongly aided by the vector control measures which began to be implemented from the beginning of the 20th century following the discovery of the role of Anopheles mosquitoes in the transmission of malaria. By the early 1950s, malaria had largely disappeared from North America and from almost all of Europe, including, in large degree, from the European parts of the then Soviet Union (Table , row 1). By the end of the 1960s, all these areas had been formally declared malaria free by the World Health Organization. In Central America and the Caribbean (Table , row 2) the malaria problem in the 20th century had, on a per capita basis, been much more significant than that in North America. However, by the last decades of the 20th century, vector control, by house spraying with the residual insecticide DDT, had achieved vast reductions in the numbers of malaria-related deaths, although complete interruption of malaria transmission was not achieved. In South America, certain countries have long been highly malarious, notably Suriname and the Guianas on the northeastern Atlantic coast. However, most of the population of this continent seem generally to have had little contact with malaria, simply because most people have always lived outside malaria-prone locations. Thus, the absolute amount of malaria-related mortality in the early part of the 20th century appears to have been much less in South America (Table , row 3) than in Central America and the Caribbean. Only locally would it have been intense, as in parts of the Amazon and on the northeastern Atlantic coast. The control of malaria and malaria mortality in the locations of South America with endemic infection remain, however, a considerable problem. This is partly due to the size and remoteness of many of the areas affected. It is also complicated by the development of resistance of the parasites to antimalarial drugs, including chloroquine. Remarkably, chloroquine-resistant malaria parasites had not yet arisen in Central America at the close of the 20th century, in spite of the prevalence of chloroquine-resistant P. falciparum through South America since the early 1960s . A major engine of malaria in South America has been, and remains, pioneering and other human activity in previously uninhabited land. Conducted across the vast Amazon basin, these activities have created one of the most difficult problems of malaria control anywhere in the world. FIG. 3. | Malaria mortality in the 20th century. Malaria mortality in the 20th century. The graph shows the number of malaria deaths per 10,000 population per year in Europe and North America ( ---); the Caribbean and Central and South America ( --- ---); sub-Saharan Africa (* --- ---*); China and Northeast Asia (X ---X); the Middle East, South Asia, and the Western Pacific ( --- ---); and worldwide ( ---) (see Appendix and Table ). Malaria in Asia and the Western Pacific. : From the beginning of the 20th century to its end, Europe and the Americas carried a small, and eventually a tiny, part of the global burden of malaria. By far the greatest part of this burden has always been in Asia and in Africa. In the first half of the 20th century, the malaria problem in Asia overwhelmed all others (Table , rows 4 and 5; Fig. and ). On the Indian subcontinent in the early decades of the 20th century, epidemic malaria could kill tens of thousands to hundreds of thousands within a few months in a single region (-, ). However, in spite of the lethality of the epidemics, their death tolls in South Asia were much lower than the malaria-related mortality that was sustained year upon year by the populations of these regions. Working in India at the turn of the century, Ronald Ross named malaria the "million murdering death" . Malaria, however, did not kill unassisted. Contemporary commentators often linked famine, a constant specter in Asia in the early part of the 20th century, with the huge malaria-related mortality of the Asian regions (-, ). The frequently poor condition of the affected populations certainly contributed much to their susceptibility to the effects of malaria. For almost two decades toward the middle of the 20th century, from the early 1930s to the late 1940s, war across Asia and the Western Pacific stirred the cauldrons of malaria. Records from these and other war zones (the American Civil War, the First World War) show large increases in malaria incidences and deaths among the military . Under the economic and environmental depredations that war produces, civilian populations would have been no less affected. Then, with the relative political stability which followed the end of the Second World War, the burden of malaria across most of this vast region began a period of remarkable decline (Table , row 4; Fig. and ). In the late 1940s and early 1950s, national malaria control campaigns were established in almost all of the affected countries of the region, from the Middle East, through the Indian subcontinent and Southeast Asia, to the islands of the Western Pacific, including those of Indonesia and the Philippines. Under the broad direction and encouragement of the newly formed World Health Organization , and employing the residual insecticide DDT to spray homes, spectacular reductions in malaria incidence and malaria-related mortality were achieved, especially in India and Ceylon (now Sri Lanka) . Crucial to this success was the simultaneous deployment throughout most of the affected areas of health delivery systems providing treatment with highly effective antimalarial drugs, most notably chloroquine. By the mid-1960s, malaria incidence in India and Sri Lanka was so low as to be almost undetectable by the monitors of the now huge national malaria-screening programmes. However, in spite of this powerfully delivered effort, the anticipated goal, the eradication of malaria, was not achieved in any country of the region. The prohibitive economic and political costs of operating the malaria control campaigns, which from 1954 onward became formally instated by the World Health Organization as Malaria Eradication Campaigns , were not sustainable. This, combined with emerging resistance of the parasites and their vectors to the chemicals used to attack them, led, from the early 1970s, to resurgence of malaria transmission throughout southern Asia and the Western Pacific. Most damaging was the emergence of multidrug-resistant P. falciparum, including total resistance to chloroquine. Since the mid-1960s, chloroquine resistance has spread inexorably outward across the tropics of Asia and the Western Pacific and into Africa from a focus of origin in Southeast Asia . These setbacks notwithstanding, however, a return to the previous huge malaria-related mortality rates in this vast section of the human population has never been remotely approached (Table , row 4; Fig. and ). The massive and sustained reduction of malaria-related mortality from the Mediterranean to the Western Pacific in the second half of the 20th century has been a truly outstanding, if precarious, success in the improvement of human health. China followed its own path in the second half of the 20th century (Table , row 5; Fig. and ). The emergence of strong national government and the absence of major warfare must have contributed significantly to the reductions of its malaria burden in the decades that followed the end of the Second World War. However, political turmoil within China in the late 1960s and early 1970s prevented major advances. Then, in the mid-1970s, a determined antimalaria campaign was initiated in which vector control was integrated with rigorous malaria case detection and treatment . Malaria-related deaths in China may now be fewer than several hundred per year, compared with the hundreds of thousands that were sustained through the early decades of the century. The Chinese antimalaria campaign yielded another outstanding contribution. This was the development of the artemisinin derivatives of Quinhaosu, a traditional Chinese herbal medicine, to combat chloroquine-resistant P. falciparum. In Vietnam, in Southeast Asia, in the early 1990s, artemisinin derivatives contributed dramatically to the reversal of a renewed rise in malaria mortality rates in the region . Today these compounds are at the heart of attempts to reformulate global malarial drug policy based on drug combinations which include artemisinin derivatives as a means of combating the emergence and spread of antimalarial drug resistance . Malaria in Africa. : Because of the massive burden of malaria in Asia in the early part of the 20th century, the campaigns against malaria in Asia in its second half were the response to a long-standing crisis. And because of the vulnerability of Asian malaria transmission to vector control measures using DDT and to the relative organizational and political stability that emerged across the region after the Second World War, these campaigns were practical and, as events have shown, to a large extent, if not completely, successful. By contrast, attempts by similar means to control malaria in sub-Saharan Africa had little success, except in its southern and most marginal zones of transmission. Indeed, the malaria problems of Africa were, and are, of an altogether different type from those confronted anywhere else, both in human terms and in the biological factors that underlie African malaria transmission. Above all, the stability and intensity of malaria transmission in Africa presented two huge, actual or potential, problems. When global campaigns for malaria control were being planned at the World Health Organization in the late 1940s and early 1950s, there were many who doubted that even a tool as powerful as DDT could have a significant impact on African malaria because of the intensity of its transmission . Moreover, and by the same token, it was questioned whether it was even safe to try to reduce malaria transmission intensities in tropical Africa because of the eventual loss of immunity in the older population and the consequent risk of epidemic malaria should control measures fail. National malaria control organizations were nevertheless operational in many African countries by the 1950s. It must soon have become clear, however, that whatever may have been being achieved elsewhere by reducing malaria transmission using DDT, rather little effect was served by this approach in sub-Saharan Africa, except in certain limited circumstances and mainly in its southernmost parts. Nevertheless, a determined optimism reigned among the advocates of "global malaria eradication" and a policy of "intent to tackle the problem in Africa upon an eradication basis" seems to have persisted until at least 1996 . In the end, however, and as the goal of "malaria eradication" collapsed in most other malaria-endemic regions of the world , this aspiration for Africa also, and inevitably, died . One location, which is effectively a part of the African region and in which malaria control succeeded virtually to the point of eradication, was in the highlands of the island of Madagascar in the Indian Ocean. Here, for around 20 years from the end of the 1960s to the early 1980s, malaria transmission was almost completely suppressed. Then, in 1986, the worst fears of those who had cautioned against malaria eradication campaigns in environments that naturally and/or potentially sustain intense malaria transmission were realised. With all natural immunity against malaria lost in the population and control measures now at a low level, malaria was reintroduced into the Madagascar highlands. For 2 years a severe epidemic raged, with high death rates in all age groups. Probably many tens of thousands lost their lives . Overall, however, Africa had benefited during the era of "malaria eradication." It has benefited from a new availability of antimalarial drugs, especially chloroquine. Although drug distribution and access to treatment were relatively poor and largely uncontrolled, the effects during this period were real and evident. Malaria-related deaths in Africa showed evidence of relative (per head of population) decline from the 1950s to the early 1980s (Table , row 6; Fig. and ) . But then, from some time in the 1980s, the downward trend in malaria-related mortality appears to have reversed. In relation to total population and in relation to deaths from other causes, the numbers of childhood deaths from malaria in tropical Africa are almost certainly rising again (Table , row 6; Table , row 4; Fig. and ). The most likely cause is the spread throughout Africa, from its seedbeds in Southeast Asia, of chloroquine-resistant P. falciparum . Death from Malaria in the Twentieth Century : The foregoing was a sketch of the general picture of malaria through the course of the last century. In the following section, we discuss some quantitative conclusions from this study. The data referred to are summarized in Table . The first half of the twentieth century. : The relatively huge problem of malaria on and around the Asian continent (the Middle and Near East, the Indian subcontinent, Indo-China, the Western Pacific, China, and Northeast Asia) from the beginning to the middle of the 20th century has already been emphasized. As the century began, total malaria deaths stood at around 3 million per year (Table , row 3), representing around 30 lives taken by malaria per 10,000 population per year, or approximately 10% of all deaths across these combined regions at this time. Of course, only a part of this population, perhaps half and perhaps considerably less, would have resided in significantly malarious locations. In these areas, including much of India, much higher proportions, in excess of 50%, of all deaths were probably due to malaria . In India, malaria deaths rates could reach 150 per 10,000 population per year across entire provinces, e.g., United Provinces, West Bengal . However, because of the highly uneven distribution of malaria across India , many localities within these regions would have surpassed even these rates, to a point at which they became uninhabitable through the presence of malaria . In the same period, the early 20th century, in sub-Saharan Africa, with its intense, stable transmission, mainly of P. falciparum, the total malaria death toll was probably around 200,000 to 300,000 per year (Table , row 6; Table , row 4). This, at the time, represented only about 1/10 of the world malaria death toll . Across tropical Africa, the rate of malaria deaths would have been relatively uniform and almost all concentrated in the very young. At around 20 to 30 malaria deaths per 10,000 total population per year, between 5 and 10% of all deaths in tropical Africa were due to malaria. In absolute and in relative terms, this was certainly a much lower overall toll than in the most malarious parts of Asia. It supports the proposition that low to medium malaria inoculation rates under unstable transmission conditions can be much more lethal than stable malaria transmission at almost any inoculation rate. The annual malaria death toll for the whole of Europe and North America in the first decade of the 20th century was probably around 50,000 to 100,000 per year (Table , row 1). Remarkably, this is as much as one-third of the total malaria deaths in sub-Saharan Africa at the time. In the malarious parts of Europe, mainly around the Mediterranean shores and in the Volga basin, and in North America across almost the entire southern states and the Midwest, malaria mortality rates would have been similar to those in many parts of Asia. However, across Europe and North America as a whole, the overall risk of death from malaria was less than 2 per 10,000 total population per year. It contributed to no more than 1% of all the deaths in Europe and North America at the time. Of all the continental regions, the least populous in the early 20th century were Central and South America, including the Caribbean. Absolute numbers of malaria deaths, at around 40,000 per year, were correspondingly low (Table , row 2). However, the relative risk was moderately high. Across the region as a whole, malaria took the lives of 5 to 10 of every 10,000 population per year. The distribution of malaria deaths would have extended through all ages, as in Asia and Europe. FIG. 4. | Malaria mortality in the 20th century. Malaria mortality in the 20th century. The graph shows the total number of deaths due to malaria per year worldwide minus sub-Saharan Africa (lozenge--lozenge) and in sub-Saharan Africa (* --- ---*) (see Appendix and Table ). The second half of the twentieth century. : Until the middle of the 20th century, the relative balance of malaria mortality in the different continental regions remained approximately the same (Fig. and ). From shortly after the Second World War, however, a huge transformation in the global burden of malaria began to take place (Fig. to ). It is this change which certainly gave rise to the recent common perception that "malaria had gone away." In Asia and the Western Pacific rim, the risk of death by malaria plummeted. At around 0.2 per 10,000 population per year across the entire region (Table , row 3), the malaria death rate is now less than 1% of what it had been at the beginning of the 20th century. Even in the most malarious countries of the region, such as Cambodia and Laos, no more than 1 to 2% under present circumstances, die from malaria. The situation in these two small countries is, today, held to be a continuing heart of malarial darkness. But even here, the lives culled by malaria are a fraction of those of little more than half a century before. By the early 1960s, malaria in Europe and North America was extinct. Today, the chances of a European or a North American dying from malaria, other than through involvement in a malarious war zone, when the risk soars, is probably much less than 1 in 100,000 (Table , row 1), and it arises only through travel. For the world as a whole, outside Africa, the chances of dying from malaria are around 1% of what they were at the beginning of the 20th century (Table , row 5; Fig. and ). Thus, from around 1 death in 10 at the century's start, malaria accounted for only about 1 in every 1,000 at its end (Table , row 5). Within Africa, however, the proverbial "one million children die of malaria every year" was, by the estimates shown here, finally reached in the last decade of the 20th century (Table , row 4) . Much as they were at the beginning of the 20th century, 5 to 10% of all those born in tropical Africa today are destined to die from malaria before they reach the age of 5 years (Table , row 4; Fig. ). STRATEGIES TO MANAGE THE BURDEN OF MALARIA TODAY : In most of the world today, therefore, we no longer experience the worst natural impacts of endemic malaria, its massive death tolls, and its otherwise crippling effects on individuals and communities. The reason is partly that the prevalence of malarial infection is now 1/10 or less of what it was at its height. However, in addition to its lower prevalence, malaria case fatality rates have also been reduced to around 1/10 of previous levels. This is because even in most of the world's poorest countries, access to treatment of malarial infections is sufficiently reliable that mortality from acute P. falciparum infection is usually prevented and prolonged, untreated malarial infection is rarely experienced. But this applies mostly outside tropical Africa. Here, where improvements were, in any case, relatively small, the situation is once again deteriorating (Fig. and ). Elsewhere, across many tropical and subtropical regions, the problems of and created by malaria are also far from over. In this final section, we examine current global approaches to managing the malaria problem today. Lessons from the Past : With the exceptions mainly of Europe and North America, elsewhere in the world the mid-20th century goal of malaria eradication was never realised. What was achieved, however, was an unprecedented reduction in the morbidity, and especially the mortality, due to malaria across vast regions of the tropical and subtropical world. For the most part, the types of tool that are available and are used for malaria control today are the same as those which were available during the "eradication" area. From this experience, both in its successes and in its failures, there are important lessons. The first is that the tools of the last century were, and remain, very effective in reducing the malaria burden. These tools are (i) drugs with which to treat malaria, which not only reduce its immediate morbidity and mortality but also help to contract the human reservoir of malarial infection and hence its transmission, and (ii) methods to reduce human-mosquito contact and thus the transmission of infections, which include house spraying with chemical insecticides, the use of insecticide-treated bed nets and other materials, and other forms of vector control by nonchemical means such as those involving environmental and biological approaches. The malaria eradication campaigns of the 1950s and 1960s were aimed at total interruption of malaria transmission by a strategy which entailed aggressive antimalarial activity over large geographical areas . By their very nature, these campaigns had to be "vertical" in administration and disease specific in operation. The second lesson learnt from that experience is that such highly disease-specific and intensive operations also have major inherent drawbacks. Not the least of these was the difficulty, and eventually the impossibility, of sustaining such efforts in the poorly resourced environments that malaria-endemic countries almost invariably are. With few exceptions, the goal of eradication was never reached, and most of those countries which were involved have been left with continuing burdens of malaria. These have been contained by "malaria control" programs ever since the unsustainable efforts toward "malaria eradication" were abandoned. Another problem of the large, vertical programs was their effects on general health service delivery, which they tended to bypass and, in some respects, to disrupt. However, the malaria eradication campaigns also delivered some great and lasting benefits. These campaigns were the means of setting up countrywide infrastructures, including large numbers of field clinics for the diagnosis and treatment of malaria and offering many poor rural communities their first access to health services of any kind. These are still the mainstay of health delivery in many remote parts of these countries and are an invaluable legacy to the countries which participated in the malaria eradication campaigns. The Malaria Situation Today : From "global eradication" to global gloom. | During the 1970s and 1980s, the health sectors of countries where malaria eradication had been attempted, in Asia and the Americas, went through substantial reforms. They moved from a vertical, disease-specific approach to one of supporting the delivery of a more integrated health service. During this period, countries also devolved many of their systems of administration, including the operation of their health systems, to provinces and districts and adopted primary health care as a key strategy . In parallel with these changes, however, the 1970s also marked the beginning of a period of trauma in the history of the struggle against malaria. The objective of malaria "eradication" had been abandoned . Some disarray followed as new strategies of malaria control and management were attempted. The difficulties were compounded by technical failures resulting from the development of resistance of the parasites to the antimalarial drug chloroquine and of the mosquito vectors to the insecticide DDT, the two "wonder tools" through which the recent previous success had been won . Where, not long before, malaria had been reduced, in some countries almost to the point of extinction, renewed outbreaks began to occur. Although its scale was relatively small compared to that in the first half of the century, the resurgence of malaria, and increasingly in drug-resistant forms, spread global despondency at the failure to resolve this major health issue. The problem at hand. : In Asia and in most of the Americas, as we have thoroughly discussed, the malaria burden of today is a residue of what once prevailed. It is worst in remote, rural areas and in situations where there is civil unrest or other conflict. Thus, in Asia, some of the areas of highest malaria burden are in the Mekong region along international borders and in other conflict-stricken parts of the countries of the region . Similarly, in South America in the Amazon basin, malaria is brought to indigenous people by the incursions of commercial and pioneering activities . Elsewhere, for example in the Central Americas, in India and Sri Lanka, and in the islands of the Western Pacific, the burden of malaria is, by historical standards, not yet excessively high, but it persists at unacceptable levels . With the constant threat of unmanageable drug-resistant malaria, especially P. falciparum malaria, situations fluctuate between worse and better. Should antimalarial drugs fail completely, the global malaria situation would become catastrophic. In tropical Africa, the geographical areas of malaria endemicity have remained largely unchanged for at least the past 100, and most probably for the past several thousand, years. It carries by far the greatest burden of malaria today. A prominent aspect of this situation is a lack of adequate health systems to effectively deliver antimalarial drugs or other interventions to those at risk. Africa is highly vulnerable to international failure to ensure the continuing supply of effective antimalarial drugs. Current Global Approach to Controlling Malaria : The old vertical programs against malaria have been, largely if not completely abandoned. If there is a theme, it is to harness resources, and institutions ---local, national, regional, and international ---to use whatever means there are that can be brought to bear to reduce, or even contain, the burden of malaria. Rather than as separate and free-standing programs, however, these means are, wherever possible, integrated into the general health systems of countries. The focus is now on the areas of the world most in need and not, as in the past, on those which were susceptible to a particular method of control. These areas are the entire tropical region of Africa and, elsewhere in the world, the most improverished sectors of the populations affected by malaria. A key challenge is to achieve the sure sustainability of antimalarial efforts, however they may be formulated. In striking contrast to the outlook in the mid-20th century, there is an implicit assumption that malaria will remain with us indefinitely. This is, for the time being at least, a realistic and a constructive attitude. There are, as there will always be in these circumstances, two goals in the management of malaria. These are (i) to treat those who are infected and sick with malaria as quickly and as effectively as possible and (ii) to reduce to the minimum the numbers of those who are at risk of becoming infected and ill with malaria . Treating the sick. : The first goal, treating the sick, is entirely dependent on the effective use of antimalarial drugs delivered to malaria patients in a timely manner. In areas of intense transmission, this will include regular, intermittent prophylactic treatment of high-risk groups such as infants and pregnant mothers. To achieve this, health delivery systems will have to be vastly improved, especially in most of tropical Africa , and this will involve fundamental rethinking at many levels. The issues of near-the-home versus health facility-based treatment, public versus private channels of health care delivery, and interactions between governments and funding bodies, are all under active discussion between national governments and international and other agencies interested in health care delivery . At present, in most parts of remote, rural Africa, formal health care systems can rarely be accessed in time when a child's life is at risk. Often this is as a result of malarial infection. Saving these lives will entail drug distribution systems operating near the home, through either community resource persons or drug vendors and other informal health care providers. Governments will have to assume a much greater stewardship role, coordinating delivery, ensuring quality of service through regulation and control, and creating an informed public demand for health care through education and societal movements . Such responses to the malaria problem will also help greatly to strengthen general health care delivery systems in these regions of Africa. At an entirely different level is the issue of the development, production, and availability of effective antimalarial drugs at prices affordable to those in need of them in the countries with endemic infection. Drug resistance, including multidrug resistance, has eliminated, or is rapidly eliminating, the cheap and previously very effective "wonder drug," chloroquine , and its increasingly more costly successors. The solution to this problem requires finding ways to engage the expertise and commitment of the research-based but market-driven global pharmaceutical industry, which holds the key to drug development. Until very recently, commercial interest in the development of new and effective antimalarial drugs for highly impoverished markets has been minimal. However, intensive dialogue between global health institutions and drug companies has resulted in public-private venture capital funds being established for drug discovery and development for malaria (, ; R. G. Ridley and W. E. Gutteridge, Proc. Med. Front. Symp., 1999). Still further financial resources will be needed if such drugs are to be effectively deployed at prices affordable to the target populations. Among the future tools against malaria, mention must be made of the immense efforts toward the development of protective vaccines against malaria. In spite of the duration of this struggle toward malaria vaccine development (perhaps 20 years of seriously goal-oriented effort to this point), progress has been inexorably forward. The number of clinically testable malaria vaccine candidates and formulations increases yearly. It is the aspiration that a practical blood stage vaccine would protect an individual against the clinical effects of malarial infection for at least several years. This would reduce the dependence of the individual on the health systems for treatment and care. It would also greatly relieve the burden on the health systems themselves. Reducing the risk of malaria. : The second goal in the management of malaria, reducing the number of individuals at risk of malarial infection and illness, involves reducing human-mosquito contact by whatever means are practical. These means will certainly include, especially in Africa, the expanded deployment and use of insecticide-treated materials, bed nets, and curtains for those at highest risk, namely, infants, young children, and pregnant women. Included in these approaches may be prophylactic interventions, such as can be linked to the Expanded Programme for Immunization initiative. It has been shown in an African setting that intermittent treatment with an antimalarial drug in the first year of life can greatly reduce the numbers of clinical episodes of malaria in infancy without appearing to increase the subsequent risk of infection through loss of early immunity . Vaccines could also play a role in transmission reduction in the future by preventing the infection of mosquitoes (anti-sexual or mosquito stage malaria vaccines) and by preventing vaccinated individuals from becoming infected by mosquitoes (preerythrocytic stage malaria vaccines). Together with insecticide-treated materials and mosquito control measures, these effects of vaccination could greatly impede malaria transmission in a locality. Other examples of reducing human-mosquito contact, which may be relevant especially in locations of low endemicity, mostly non-African locations, involve the appropriate construction and siting of housing and local environmental improvement. Inexpensive, high-technology approaches will apply computerized information management and Geographic Information Systems to identify locations where targeted attack on malaria transmission is called for. Monitoring, including by satellite, of all aspects and features of a malarious situation will be important to the timing and targeting of antimalarial interventions . Complications may arise, however, as a result of reduction of malaria inoculation rates. The relationship of stability and instability of malaria transmission to acquired immunity to malaria will be critical to how this turns out. Under conditions of stable malaria, immunity is both acquired and retained from high to low malaria inoculation rates. However, at very low inoculation rates under conditions of stable malaria, immunity is poor and older age groups become increasingly vulnerable to severe disease. Under conditions of unstable malaria transmission, except at very high malaria inoculation rates, there is generally little acquired immunity anyway, so that reductions in inoculation rate lead mainly to reduced incidence of severe disease. Potential loss of immunity accompanying transmission reduction must nevertheless be constantly monitored in all situations. Should it occur, it will show up as increasing numbers of cases of clinical malaria in older ages groups. It must be managed by ensuring adequate access to treatment with antimalarial drugs. Managing epidemic malaria. : Epidemic malaria will remain a constant threat in regions which border on endemic malaria transmission zones. This is especially the case in highland areas adjacent to such zones and in lowland areas, at risk, for example, of widespread flooding, or, indeed, drought. In these locations, extreme weather and climatic fluctuations will periodically create the conditions for malaria epidemics. It should be noted that wherever malaria control in high-transmission areas begins to succeed, those populations will become vulnerable to epidemic malaria, as occurred in Madagascar in 1986. Monitoring and reliable communications, integrated with regional and global capacity to mount rapid distribution of drugs and other relief, must be ensured . Global role in malaria management. : The global institutions which support the management of malaria remain at the heart of this enterprise. It is through them that those technical, economic, and political initiatives can be developed and implemented by which malaria-endemic countries can be enabled to operate, and to be supplied with the resources and materials needed for, their antimalarial efforts. To this end, international efforts are, for example, engaging the private, for-profit sector in setting up processes to ensure the continued development of effective and affordable antimalarial drugs and eventually of vaccines (, ; Ridley and Gutteridge, Proc. Med. Front. Symp., 1999). Civil society and voluntary organisations are being encouraged to work with national government-run programs and the public sector health system. In their malaria control planning, governments are being encouraged to involve and coordinate a wide range of national initiatives. These include government economic and environmental policy planning, as well as enlisting communications, transport, and civil construction programs to the cause of managing malaria. Many of these are aspects in which current efforts to control malaria differ strategically from previous campaigns (-). The resources required to sustain this multifaceted campaign against malaria can only come from international sources. They are much greater than are currently on-line. Nevertheless, through the publicity already mounted in support of global malaria control, the responses from some of world's largest economies to contribute to this effort through bilateral and multilateral arrangements are already encouraging . CONCLUSIONS : How then, can we sum up the lessons of the impact of malaria, past and present, and how can they help us to deal with the disease today? Overall, we can say that the impact of malaria on our species has been very great. Where and how this has occurred has been, nevertheless, very variable. In one form or another, the burden of malaria continues to this day at an unacceptable level. This burden is composed not only of the direct effects of malaria but also of the great legacy of debilitating, and sometimes lethal, inherited diseases that have been selected under its impact in the past. These inherited illnesses will probably be among us for many generations to come. In the future management of malaria, the tools available, drugs, insecticides, insecticide-treated materials, etc., will be of great importance. They are, individually and collectively, very effective instruments, although they are under constant threat from drug-resistant parasites and insecticide-resistant vectors. Even better malaria control would be achievable if new tools, such as vaccines, were to become available. However, even to maintain old tools, let alone to bring in the new, will require sustained investment of effort and resources from the wealthy nations of the world and on a much greater scale than is taking place at present. Malaria is now one of the great diseases of poverty. Today no wealthy nation is affected by its endemic presence. In Europe, in the United States of America, and parts of China, there are whole regions which could not have achieved their current degree of prosperity had they been unable first to banish malaria. By contrast, in many other countries, and they include the poorest in the world, the magnitude of the malaria problem is simply overwhelming. The level of material support available for the management of malaria, even with the tools available today, is far below what is needed or what could be provided. And yet malaria-endemic countries are, by and large, better prepared than for many decades past to absorb, and to use to good effect, the resources that are needed to manage their malaria problems. Given the necessary global political will, and there is a tide running in its favor, much can be achieved by greater financial and organizational commitment to the battle against malaria. Of the malaria problems in India at the start of the 20th century, Rickard Christophers wrote, "At the back of such colossal manifestations as the (malaria) epidemic of 1908... we shall do well to bear in mind the magnitude of the influence against which we are pitting ourselves, and not be led in foolish vein to reduce the remedy of the whole matter to trifling proportions." The "magnitude of the influence" remained gargantuan for another half century. Yet today, throughout India, that "influence" has been reduced, by comparison, to a small remnant. Thus, in the longer run of historical time, real improvements have taken place where once they may have seemed almost impossible. And, of course, long-term changes have, at other times, been for the worse. Nevertheless, and in final conclusion, it may be worth noting that we could, in principle, be truly rid of malaria. This possibility arises from the basic realities of malaria transmission. It could happen if, and when, life-styles and environments in all the currently malaria-endemic regions have reduced human mosquito contact below the levels that sustain malaria transmission under their prevailing climatic conditions. This, indeed, is the only basis for the permanent elimination of malaria. However, it could happen only in the presence of the necessary economic, political, and social development in all of the affected countries. APPENDIX : Sources of Data and Approaches to Their Analysis | The data presented in Tables and represent an attempt to quantify malaria-related mortality within six major geographical zones at different periods during the 20th century. The time points have been chosen to encompass intervals within which major historical changes have occurred and/or between which there have been significant developments in the malariological situation. The geographical divisions represent ones within which a broad, sometimes a very broad, unity of human and environmental conditions apply and between which clear distinctions can generally be made on the same grounds. The sources and types of relevant data drawn upon in this synthesis are very variable. For all regions outside sub-Saharan Africa, they derive (i) from individual research reports and statements of "expert opinion" concerning malariological situations in different times and regions, (ii) from records of offices of national governments, and (iii) for the second half of the 20th century, from reports of the World Health Organization Regional Offices. For sub-Saharan Africa, estimates of malaria mortality rates in the earlier part of the century were made from the mortality rates implicit in the frequencies of the sickle cell gene in African populations and from a limited number of reports of malaria mortality rates in African for the period. The estimates for this region for the second half of the century depend entirely on contemporary published scientific reports of malaria mortality in children in tropical Africa. These include a number of such reports, including unpublished data, assembled by Robert W. Snow and used during the preparation of the "World Health Report, 1999" . We are grateful for permission to include these data in the present analysis. Limits of Uncertainty of the Estimates of Malaria-Related Mortality : The degrees of uncertainty associated with the estimates of malaria-related mortality as presented here are hard to assess. The estimates of malaria mortality rates in tropical Africa, especially for the second half of the 20th century, have been made from considerations of individual examples of recorded mortality rates. These estimates take into account the age distribution of malaria mortality in tropical Africa and assessments of the locations and sizes of the populations at risk. Given the relatively constant malaria mortality rates that these studies reveal within spans of a decade, the mean values that we have given are probably accurate to within less than a factor of 2. For the first half of the century, the mortality estimates for sub-Saharan Africa are absolute expectations from the recorded frequencies of the sickle cell gene. These compare well with recorded estimates of malaria mortality rates from the period covering the 1930s and 1940s . Outside sub-Saharan Africa, the estimates for the second half of the 20th century use malaria reports of the World Health Organization Regional Offices. The estimated mortality data for each country for this period take account of human population size as well as malariological information. We used internal cross-referencing between computed numerical outcomes for different parameters, e.g., malaria incidence rates, mortality rates, and case fatality rates, in order to retain all of them within the limits of biological plausibility, as we have described previously . The outcomes are probably accurate within a factor of 2 or less. The estimates for Europe and North America for the first half of the 20th century are all from national records of causes of mortality and could be expected to have a similar level of accuracy. For the rest of the Americas for the first half of the century, the situation is difficult to assess; the numbers given here are guesses extrapolated from recorded information from the 1940s and 1950s. For the first half of the 20th century there are few national records of malaria statistics from Asia, including China. The values provided here are extrapolations from extensive and well-documented studies or statements by expert observers (see, e.g. references -, , , and ) and from authoritative analyses (see, e.g., references , , and ). From the population sizes at risk and from representative malaria mortality rates within this region at different periods, the figures for Asian malaria-related mortality for the early part of the century seem unlikely to be in error by factors of more than 2 or 3 at most. This analysis has used published statistics for human population sizes at each of the time points represented . Relative to the estimates of malaria mortality rates, it is assumed that population statistics have a high degree of accuracy and do not, therefore, contribute significantly to uncertainties associated with the analysis. Country Composition of the Six Global Regions : The country names listed below do not always represent their current political designations but are used for convenience to encompass human populations wherever they have been located throughout the century. Europe and North America. Albania, Andorra, Austria, Belgium, Bulgaria, Czechoslovakia, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Luxembourg, Malta, Netherlands, Norway, Poland, Portugal, Romania, Spain, Sweden, Switzerland, Turkey (in Europe), Russia (in Europe), Yugoslavia, United Kingdom, small European states, Greenland, Canada, United States of America. Central America and the Caribbean. Belize, Costa Rica, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama; Antigua, Bahamas, Barbados, Bermuda, Cayman Islands, Cuba, Dominican Republic, Haiti, Grenada, Jamaica, Martinique, St Vincent & Grenadines, Puerto Rico, Trinidad & Tobago, other Caribbean islands. South America. Bolivia, Brazil, Colombia, Ecuador, French Guiana, Guyana, Suriname, Venezuela, Peru, Paraguay, Uraguay, Chile, Argentina. South Asia, Western Pacific, and Middle East. Afganistan, Bahrain, Iran, Iraq, Israel, West Bank, Jordan, Lebanon, Syria, Oman, Qatar, United Arab Emirates, Saudi Arabia, Kuwait, Yemen, Turkey (in Asia), Cyprus, Russia (in Asia)and Republics of the former USSR in Asia, Morocco, Algeria, Tunisia, Libya, Egypt, small states of the region; Pakistan, Bhutan, India, Nepal, Bangladesh, Burma, Sri Lanka, Thailand, Laos, Cambodia, Vietnam, Indonesia including Irian Jaya, Malaysia, Brunei, Philippines, Singapore, Japan, Papua New Guinea, Solomon Islands, Vanuatu, New Caledonia, Australia, New Zealand, Fiji, other Pacific islands. China and North East Asia. People's Republic of China including the Autonomous Region of Tibet, Mongolia, North Korea, South Korea, Taiwan. Africa, sub-Sahara. Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African Republic, Chad, Comoros, Congo, Djibouti, Equatorial Guinea, Ethiopia, Eritrea, Gabon, Gambia, Ghana, Guinea, Guinea-Bissau, Ivory Coast, Kenya, Lesotho, Liberia, Madagascar, Malawi, Mali, Mauritania, Mauritius, Mozambique, Namibia, Niger, Nigeria, Rwanda, Senegal, Sierra Leone, Somalia, South Africa, Sudan, Swaziland, Tanzania, Togo, Uganda, Zaire, Zambia, Zimbabwe. Sources used. (i) Malaria mortality. Europe and North America ; Central America and Caribbean (, , , , ; World Health Organization, unpublished data); South America (, , , , , , ; World Health Organization, unpublished data); South Asia, Western Pacific, and the Middle East (, , -, , , , , , , , , , , , ; World Health Organization, unpublished data); China and North East Asia ; Africa, sub-Sahara; (, , , , , , , , , , , , , , , , , , , ; (Robert W. Snow, personal collection). (ii) World and regional population sizes. : Data obtained from . ADDENDUM IN PROOF : Since the final version of this manuscript was submitted for publication, three relevant studies have been published. In one (S. Tavare, C. R. Marshall, O. Will, C. Soligo, and R. D. Martin, Nature 416:726-729, 2002), the fossil record of primates is reviewed. This report shows that primate species occurred throughout the African and Eurasian continents, including present-day western Europe, from the late Oligocene (30 million years ago) to the late Pleistocene (a few hundred thousand years ago). It covers the period of 10 to 2 million years ago, at some time within which (, , A. Escarante, personal communication), P. vivax and the primate malaria parasites most closely related to it probably diverged. Therefore, an hypothesis that this divergence took place in southern and/or southeast Asia would have no prima face basis if it depended solely upon the fact that most of the P. vivax-related malaria parasites are found in primates of these regions today. From the evidence of the primate fossil record, the parasite speciations which have led to today's human P. vivax malaria could as likely have occurred in Europe or in Africa, as we and others suggest. The second report (J. Mu, J. Duan, K. D. Makova, D. A. Joy, C. Q. Huynh, O. H. Branch, W.-H. Li, and X.-Z. Su, Nature 418:323-326, 2002) concerns the antiquity of P. falciparum. Following the molecular evolutionary approach of previous workers to estimate the time of the most recent common ancestor of modern populations of P. falciparum, the report by Mu et al. involves the largest amount of P. falciparum DNA sequence examined in this context so far. Their results show higher rates of single nucleotide synonymous substitutions among P. falciparum genomes than were found in previous studies . They imply that passage of the ancestors of modern P. falciparum through a tight population bottle neck, if indeed there ever was one, could not have occurred more recently than at least 100,000 years ago. This conclusion is not, however, in conflict with a view that significant recent adaptations may have occurred in populations ancestral to today's P. falciparum. A small number of mutations selected under the influence of the profound environmental and ecological changes that were taking place in the period between 15 and 4 thousand years ago could have swept through the ancestral populations and led to the formation of modern P. falciparum and, moreover, could have done so without perturbing other features of its genome. In the companion article to that by Mu et al., evidence for selective sweeps of haplotypes in very small regions of the P. falciparum genome was demonstrated by Wootton et al. (J. C. Wootton, X. Feng, M. Ferdig, R. A. Cooper, J. Mu, D. I. Baruch, A. J. Magill, and X. Z. Su, Nature 418:320-323, 2002). These haplotype sweeps were in response to the very recent (within the past 50 years) pressures imposed upon P. falciparum by human use of the antimalarial drug chloroquine. Backmatter: PMID- 12364375 TI - Diagnosis and Management of Human Cytomegalovirus Infection in the Mother, Fetus, and Newborn Infant AB - Human cytomegalovirus (HCMV) is the leading cause of congenital viral infection and mental retardation. HCMV infection, while causing asymptomatic infections in most immunocompetent subjects, can be transmitted during pregnancy from the mother with primary (and also recurrent) infection to the fetus. Hence, careful diagnosis of primary infection is required in the pregnant woman based on the most sensitive serologic assays (immunoglobulin M [IgM] and IgG avidity assays) and conventional virologic and molecular procedures for virus detection in blood. Maternal prognostic markers of fetal infection are still under investigation. If primary infection is diagnosed in a timely manner, prenatal diagnosis can be offered, including the search for virus and virus components in fetal blood and amniotic fluid, with fetal prognostic markers of HCMV disease still to be defined. However, the final step for definite diagnosis of congenital HCMV infection is detection of virus in the blood or urine in the first 1 to 2 weeks of life. To date, treatment of congenital infection with antiviral drugs is only palliative both prior to and after birth, whereas the only efficacious preventive measure seems to be the development of a safe and immunogenic vaccine, including recombinant, subunit, DNA, and peptide-based vaccines now under investigation. The following controversial issues are discussed in the light of the most recent advances in the field: the actual perception of the problem; universal serologic screening before pregnancy; the impact of correct counseling on decision making by the couple involved; the role of prenatal diagnosis in ascertaining transmission of virus to the fetus; the impact of preconceptional and periconceptional infections on the prevalence of congenital infection; and the prevalence of congenitally infected babies born to mothers who were immune prior to pregnancy compared to the number born to mothers undergoing primary infection during pregnancy. Keywords: INTRODUCTION : Human cytomegalovirus (HCMV) is the vernacular name of human herpesvirus 5, a highly host-specific virus of the Herpesviridae family. HCMV is the largest virus in the family and is morphologically indistinguishable from other human herpesviruses. HCMV, like all herpesviruses, undergoes latency and reactivation in the host. Although HCMV has been shown to infect a broad spectrum of cells in vivo , the only cells that are fully permissive for HCMV replication in vitro are human fibroblasts. In these cells, virus replication results in the formation of intranuclear and intracytoplasmic inclusion bodies , with the former full of nucleocapsids and the latter containing several dense bodies . Nucleocapsids acquire the envelope from the nuclear membrane or cytoplasmic vacuoles . HCMV is a virus of paradoxes. It can be a potential killer or a lifelong silent companion. These two aspects are confirmed in an exemplary manner by the circumstances, vividly reviewed by Thomas H Weller , surrounding the isolation of the first HCMV strains. In 1956, Margaret G. Smith recovered the first HCMV isolate from the submaxillary salivary gland tissue of a dead infant and the second isolate from the kidney tissue of a baby dying of cytomegalic inclusion disease . The same year, Rowe and coworkers, who recovered adenoviruses by observing cytopathic changes in uninoculated cultures of human adenoids, noted unique focal lesions and intranuclear inclusions primarily in the fibroblast component of cultures of adenoidal tissues from three asymptomatic children . The cytopathic effect of the new virus strain (AD169) very closely resembled that of the Davis strain that was observed 1 year later by Weller and colleagues in human embryonic skin muscle tissue cultures inoculated with a liver biopsy taken from a 3-month-old infant with microcephaly, jaundice, hepatosplenomegaly, chorioretinitis, and cerebral calcifications . The same group of researchers isolated two additional HCMV strains: the Kerr strain from the urine of a newborn with petechiae, hepatosplenomegaly, and jaundice, and the Esp. strain from the urine of an infant with hepatosplenomegaly, periventricular calcification, and chorioretinitis . In the following years, HCMV also showed its pathogenic properties in organ transplant recipients, patients with AIDS, and cancer patients, while it gained the leading position among infectious agents responsible for mental retardation, intellectual impairment, and deafness. Presently, HCMV infection is mostly controlled in immunocompromised patients by available antiviral drugs, yet it continues to maintain its role as the most dangerous infectious agent for the unborn infant. Thus, HCMV infection is still a major health problem, warranting strong preventive measures. The major scope of this review will be to analyze and update the diagnostic and prognostic implications of primary HCMV infections in pregnancy in the mother, fetus, and newborn. Special emphasis will be given to less-investigated issues, such as detection of virus and viral products in the blood of the mother during primary HCMV infection, the presence of clinical signs and symptoms in the mother, prenatal diagnosis of congenital infection in amniotic fluid and fetal blood, maternal and fetal prognostic markers of HCMV infection and disease, and the impact of counseling. Measures of treatment and prevention of congenital HCMV infection will be mentioned briefly. The last part of the review will deal with the most controversial issues, in particular, how the problem of HCMV infections in pregnancy is perceived by the scientific community and public health authorities. Is preconception serologic screening justified, and should HCMV-seronegative women be prospectively monitored? What are the limits of prenatal diagnosis (false-positive and false-negative results)? What is the role of preconceptional and periconceptional infections in HCMV transmission to the fetus? Are reactivated infections significant in transmission of virus to the fetus? Finally, one additional goal is to focus the attention of the scientific community on the problem of congenital HCMV infection and to appeal for international collaboration. We truly need to develop and implement consensus strategies for prevention of congenital HCMV infection, ideally through a vaccine. FIG. 1. | HCMV replication in human embryonic lung fibroblast cell cultures. HCMV replication in human embryonic lung fibroblast cell cultures. (A) HCMV-infected human fibroblast 120 h postinfection (following immunoperoxidase staining with human antibodies), showing intranuclear (IN) and intracytoplasmic (IC) inclusion bodies. (B to D) Electron microscopy of HCMV-infected human fibroblasts. (B) Horseshoe-shaped intranuclear inclusion (IN). (C) Dense bodies (arrows). (D) Maturing virus particles at the level of the nuclear membrane. EPIDEMIOLOGY OF VERTICAL HCMV TRANSMISSION : The term vertical transmission is used here to indicate HCMV transmission from mother to fetus during pregnancy, thus excluding virus transmission from mother to newborn infant. Due to latency following primary infection and periodic reactivation of HCMV replication causing recurrent infections, in utero transmission of HCMV may follow either primary or recurrent infections . It is commonly recognized that primary HCMV infections are transmitted more frequently to the fetus and are more likely to cause fetal damage than recurrent infections . In addition, it seems that primary infection occurring at an earlier gestational age is related to a worse outcome . Initially, the role of recurrent maternal infections in causing congenital infections was supported by three independent reports describing congenital infections in consecutive pregnancies . In all three reports, the first newborn was severely affected and the second one was subclinically infected. Molecular epidemiological studies indicated that in each of three pairs of congenitally infected siblings, the viruses were identical to each other when examined by restriction fragment length polymorphism analysis . However, the first convincing evidence of the possible transmission of HCMV from immune mothers to the fetus came from a prospective study showing that 10 congenitally infected infants were born to immune mothers within a group of 541 infants of women who were seropositive before pregnancy , with a prevalence of 1.9%. Subsequently, similar findings were observed in a geographic area where nearly the entire population was immune to HCMV during childhood, and the prevalence of congenital infection was found to be 1.4% . In 1985, Stagno and Whitley estimated the maternal risk of acquiring either primary or recurrent HCMV infection in pregnancy as well as the risk of intrauterine transmission to the offspring in two groups of women of low or high socioeconomic status. Their estimates showed that the risk of primary maternal infection was about three times higher among the higher-income susceptible women (45%), compared to 15% in the lower-income group . In both groups, transmission to the fetus occurred in about 40% of cases, with delivery of about 10 to 15% symptomatic and 85 to 90% asymptomatic congenitally infected newborns. Among the asymptomatic newborns, about 10% developed sequelae, while about 90% of infants that were asymptomatic at birth developed normally. On the other hand, the rate of congenital infections from recurrent maternal infection was 0.15% in the higher-income group of pregnant women, who were 55% immune, and 0.5 to 1% in the lower-income group, which was 85% immune, i.e., 3 to 7 times higher. However, the rate of clinically apparent disease was low and similar (0 to 1%) in both groups. It is currently accepted that congenital HCMV infection may be the consequence of either a primary or recurrent maternal infection . Recurrent infections may consist of either reactivation of the virus strain causing primary infection or reinfection by a new virus strain. Recently, the incidence of symptomatic congenital HCMV infections in immune mothers has been shown to be similar in primary and recurrent maternal infections . In addition, symptomatic congenital infections appear to be mostly caused by reinfection of immune mothers during pregnancy by a new HCMV strain . This conclusion was based on demonstration of the appearance of antibodies directed against new epitopes of glycoprotein H of HCMV not present in the blood prior to the current pregnancy. Sequencing of the glycoprotein H gene has confirmed the presence of a new virus strain in the reported cases . On the other hand, congenital infections following reactivated maternal infection are mostly asymptomatic . In conclusion, the true frequency and clinical importance of congenital HCMV infections from recurrent maternal infections remain to be determined in long-term prospective studies. However, primary HCMV infection continues to be the major viral cause of congenital infections, with significant morbidity. Recent findings related to the potential role of recurrent maternal infection in symptomatic congenital infection complicate but will not interfere with efforts aimed at developing a safe and efficacious vaccine. FIG. 2. | Characteristics of HCMV infection in pregnancy. Characteristics of HCMV infection in pregnancy. (From S. Stagno and R. J. Whitley , used with permission.) PATHOGENESIS OF CONGENITAL INFECTION : In the case of primary maternal infection, the antiviral immune response begins proximate to virus transmission to the fetus, whereas in the case of recurrent infection, virus transmission occurs in the presence of both humoral and cell-mediated immune responses. As a result, viremia occurs as a rule only in primary infections , whereas it is either absent or undetectable in recurrent infections of the immunocompetent host and common in recurrent infections of immunocompromised patients . Since, following primary HCMV infection, intrauterine transmission occurs in only 30 to 40% of cases, an innate barrier seems to partially prevent vertical transmission . In addition, a similar event seems to occur among infected newborns, less than 15% of whom show clinically apparent infection, in the great majority of cases resulting from primary maternal infection . Finally, in reactivated maternal infections, the risk of symptomatic congenital infection is even markedly lower , as shown by the few symptomatic infants reported in the past to have been born to mothers who were immune before pregnancy. In fact, although existing immunity does not prevent transmission of the virus to the fetus, reactivated infections are less likely to cause damage to the offspring than primary infections . Multiple mechanisms of immune evasion for HCMV could relate to the pathogenic role of the virus. Recently, expression of immune evasion genes US3, US6, and US11 of HCMV in the blood of solid organ transplant recipients has been investigated, showing that, after clinical recovery, transcripts of these genes remain detectable, indicating that persistent low viral activity may have implications for long-term control of HCMV infection . Little is still known about the mechanisms of HCMV transmission to the fetus. It has been reported that about 15% of women undergoing primary infection during the first months of pregnancy abort spontaneously, showing placental but not fetal infection . Subsequently in the course of pregnancy, placental infection has been shown to be consistently associated with fetal infection . Understanding the mechanisms of HCMV transmission to the fetus implies elucidation of some major steps in placental development . The development of the placenta requires differentiation of specialized epithelial stem cells, referred to as cytotrophoblasts, in both floating villi, where they fuse into multinucleate syncytiotrophoblasts covering the villous surface, and anchoring villi, where they aggregate into columns of single cells invading the endometrium and the first third of the myometrium (interstitial invasion). While the syncytiotrophoblast is in direct contact with maternal blood, mediating transport of multiple substances to and from the fetus, the cytotrophoblast columns also invade maternal arterioles (endovascular invasion) by replacing endothelial and smooth muscle cells and thus generating a hybrid cell population of fetal and maternal cells inside uterine vessels . Syncytiotrophoblasts upregulate expression of the neonatal immunoglobulin G (IgG) Fc receptor, involved in transport of maternal IgG to the fetus . In parallel, invading cytotrophoblasts initiate expression of adhesion molecules, such as integrin alpha1beta1, and proteinases, which are required for invasion, besides molecules inducing maternal immune tolerance, such as HLA-G and interleukin-10 . Additionally, in the process called pseudovasculogenesis, invading cells modify the phenotype of their adhesion molecules, mimicking that of endothelial cells by expressing alphavbeta3 integrin, a marker of angiogenic endothelium, and vascular endothelial cadherin, a marker of cell polarization . That placenta behaves as a reservoir in which HCMV replicates prior to being transmitted to the fetus has been experimentally shown in the guinea pig, which, as in humans, has a hemomonochorial placenta with a single trophoblast layer separating fetal from maternal circulation . In experimental infection of the guinea pig with species-specific CMV, the virus disseminates hematogenously to the placenta, from which it is transmitted to the fetus in about 25% of cases. The guinea pig CMV also persists in placental tissues long after virus has been cleared from blood . Recently, a greater understanding of the human placenta has been achieved by using two in vitro models for the study of trophoblast populations lying at the maternal-fetal interface, villous explants and isolated cytotrophoblasts (-). These data, coupled with immunohistochemical studies of in vivo HCMV-infected placentas and recent findings on HCMV latency , have led to new hypotheses for routes of transmission of HCMV to the fetus in primary and reactivated maternal HCMV infection. During primary infection of the mother, leukocytes carrying infectious virus may transmit HCMV infection to uterine microvascular endothelial cells (E. Maidji, E. Percivalle, G. Gerna, S. Fisher, and L. Pereira, Abstr. 8th International Cytomegalovirus Workshop, abstr. p. 31, 2001). These cells are in direct contact with cytotrophoblasts of anchoring villi invading maternal arterioles and forming hybrids of maternal-fetal cells . Infected cytotrophoblasts may in turn transmit the infection to underlying tissues of villous cores, including fibroblasts and fetal endothelial cells , thus spreading to the fetus. An alternative model of transmission, in the case of primary maternal infection, is spreading of infection to the villous stroma by infected maternal leukocytes through breaches of the syncytiotrophoblast layer . A further hypothesis has been raised suggesting possible transportation of the virus as antibody-coated HCMV virions by a process of transcytosis through intact syncytiotrophoblasts similar to that advocated for transport of maternal IgG to the fetus . Finally, syncytiotrophoblasts may be directly infected, but the infection proceeds slowly and remains predominantly cell associated until infected cells are eliminated during the physiological turnover . This hypothesis therefore excludes transmission through virus replication in syncytiotrophoblasts. In the case of congenital HCMV infection following recurrent maternal infection, it must be considered that the placenta is a hemiallograft inducing local immunosuppression in the uterus . This may cause reactivation of latent virus in macrophages of the uterine wall, with HCMV transmission to the invading cytotrophoblasts. Then, virus could spread in a retrograde manner to anchoring villi and subsequently to the fetus . In this regard, HCMV establishes a true latent infection in CD14+ monocytes, which can be reactivated upon allogeneic stimulation of monocyte-derived macrophages from healthy blood donors . Reactivation of latent HCMV is dependent on the production of gamma interferon in the differentiation process . These data await confirmation by other laboratories. As a consequence of placental infection, HCMV impairs cytotrophoblast differentiation and invasiveness, as shown in vitro . This could explain early abortion occurring in women with primary infection. In addition, HCMV infection impairs cytotrophoblast expression of HLA-G, thus activating the maternal immune response against the cytotrophoblast subpopulation expressing this molecule . FIG. 3. | Diagram of a longitudinal section that includes a floating and an anchoring chorionic villus at the fetal-maternal interface near the end of the first trimester of human pregnancy. Diagram of a longitudinal section that includes a floating and an anchoring chorionic villus at the fetal-maternal interface near the end of the first trimester of human pregnancy. The anchoring villus (AV) functions as a bridge between the fetal and maternal compartments, whereas the floating villus (FV), containing macrophages (Mp, Hofbauer cells) and fetal blood vessels, is bathed by maternal blood. Cytotrophoblasts in the anchoring villus (zone I) form cell columns that attach to the uterine wall (zones II and III). Cytotrophoblasts then invade the uterine interstitium (decidua and first third of the myometrium; zone IV) and maternal vasculature (zone V), thereby anchoring the fetus to the mother and accessing the maternal circulation. Zone designations mark areas in which cytotrophoblasts have distinct patterns of stage-specific antigen expression, including integrin and HLA-G. Decidual granular leukocytes (DGLs) and macrophages (Mp) in maternal blood and fetal capillaries in villous cores are indicated. Areas proposed as sites of natural HCMV transmission to the placenta in utero are numbered 1, 2, and 3. (From Fisher et al. , used with permission.) DIAGNOSIS OF PRIMARY INFECTION DURING PREGNANCY : By far the major role in transmitting HCMV infection to the fetus is played by primary infections of the mother during pregnancy. In fact, the rate of vertical transmission was found to be 0.2 to 2.2% in previously seropositive mothers undergoing recurrent infection during pregnancy and 20 to 40% in pregnant women with primary infection . Thus, the ratio of transmitting to nontransmitting mothers is on the order of 1:100 between those with recurrent and those with primary infection. In this respect, diagnosis of primary infection during pregnancy is a major task of the diagnostic virology laboratory. It may be achieved in the majority of cases through concurrent analysis of the following factors: serum antibodies, virus detection in blood, and clinical signs and symptoms. Serology : Seroconversion. | The diagnosis of primary HCMV infection is ascertained when seroconversion is documented, i.e., the de novo appearance of virus-specific IgG in the serum of a pregnant woman who was previously seronegative. However, such an approach is feasible only when a screening program is adopted and seronegative women are identified and prospectively monitored. In this respect, screening programs are not approved by public health authorities of the great majority of developed countries, as reported elsewhere (see Universal Serology Screening). Thus, detection of HCMV-specific antibodies or IgG in the blood of a pregnant woman in the absence of prepregnancy antibody determination does not lead to suspicion of primary infection. HCMV-specific IgM antibody must be determined for this purpose. Although detection of specific IgM is not sufficient per se to diagnose primary HCMV infection (IgM can also be detected during reactivations), primary infection is consistently associated with the presence of a virus-specific IgM antibody response. IgM assays. : Several serologic assays have been used in the past to detect HCMV-specific IgM antibodies both in whole serum and serum fractions obtained by sucrose density gradient centrifugation or column chromatography. These include complement fixation, anticomplement immunofluorescence, indirect hemagglutination, and radioimmunoassay . More recently, enzyme-linked immunosorbent assays (ELISAs) have been more widely used in both the indirect ELISA and the capture ELISA format with either labeled antigen or antibody . The indirect ELISA shows the following potential sources of error when performed on whole serum: (i) competitive inhibition due to the presence of specific IgG; (ii) interference due to rheumatoid factor of the IgM class (IgM-RF) or to IgM-RF reactive only with autologous complexed IgG; and (iii) interference due to IgM antibody reactive with cellular antigens . All these interfering factors could be readily eliminated by mixing serum samples with anti-human gamma chain serum . However, following the development of ELISA technology, most initial IgM indirect ELISAs were replaced by IgM capture assays based on selective binding of IgM antibody to the solid phase. In capture ELISAs, while IgG does not interfere, IgM-RF may cause false results by competing with viral IgM for anti-IgM binding sites on the solid phase, complexing with specific IgG, which in turn binds viral antigens, reacting directly with the labeled viral antibody, and mutual interference with antinuclear antibody. More precisely, in capture ELISAs, the presence of the sole IgM-RF (or IgG-RF) does not cause false-positive results, which have been observed to occur in serum samples containing both IgM-RF and IgG-antinuclear antibody . Initially, capture ELISAs with enzyme-labeled antigen appeared to be the most promising assays . However, after a few years, it was recommended, for specificity control of test results, that human serum samples be tested in parallel with viral and cell control labeled antigens . In addition, false-positive results due to the presence of both RF and antinuclear antibody, as reported above, could be avoided in capture ELISAs employing labeled F(ab')2 fragments of specific antibody instead of the IgG fraction . In order to avoid false-positive results, we developed a capture ELISA IgM assay with a mixture of viral antigen and mouse monoclonal antibody to the nonstructural HCMV major DNA-binding proteins (pp52 or ppUL44) as a detector system. This phosphoprotein is prominent in HCMV-infected cells and is known to be recognized primarily by human IgM during the convalescent phase of a primary HCMV infection . According to this approach, antinuclear antibody of the IgM class bound to the solid phase will not give false reactions because only IgM antibody reactive to pp52 are recognized by the specific monoclonal antibody. Different levels of specificity were determined with this assay. General specificity, determined on a series of unselected IgM-negative serum samples from an adult population, was 100%. Stringent specificity, evaluated on a series of potentially interfering serum samples from patients who had Epstein-Barr virus-related infectious mononucleosis, autoimmune diseases, or rheumatoid factor or who had been treated with radioimmunotherapy based on the use of mouse monoclonal antibody, was 96.3%. Finally, clinical specificity, determined on a series of IgM-negative serum samples drawn prior to onset of primary HCMV infection, was 100%. Thus, the overall specificity was 98.9% (363 of 367 IgM-negative serum samples tested). The sensitivity, assayed on 277 IgM-positive serum samples, was 100%. Comparison of the results obtained by this assay with those given by enzyme-labeled antigen showed that the HCMV p52-specific IgM antibody response paralleled that obtained by using enzyme-labeled antigen, thus representing a major component of it, i.e., a major part of the antibody response within the IgM class. In addition, this study showed that, while HCMV-specific IgM drops sharply in titer in normal subjects within 2 to 3 months after onset of infection and is virtually undetectable within 12 months, in immunocompromised patients such a response persists much longer. Thus, in pregnant women, detection of HCMV IgM antibody may be related to a primary infection occurring during pregnancy when the IgM titer falls sharply in sequential blood samples. The presence of low, slowly decreasing levels of IgM may indicate a primary infection initiated some months earlier and possibly prior to pregnancy. These findings are basically in agreement with previous reports describing a broad HCMV IgM antibody response . An additional risk of HCMV IgM ELISA is a false-positive result due to primary Epstein-Barr virus infection acting as a potent B-cell stimulator and resulting in the production of HCMV IgM antibody in HCMV-immune individuals . Dual HCMV and Epstein-Barr virus infection has also been reported . Recombinant IgM assays. : Besides the lack of standards for HCMV IgM serology, the high level of discordance among commercial assays for detection of HCMV-specific IgM has been attributed to the lack of standardization of the viral preparations used. More recently, in an attempt to improve the specificity of conventional ELISAs and to overcome the discordant results given by commercial kits based on use of crude viral preparations, HCMV IgM immunoassays have been developed based on recombinant HCMV proteins or peptides. The HCMV-coded proteins reactive with IgM antibody are both structural and nonstructural . Major structural proteins include pp150 (UL32), pp65 (UL83), and pp38 (UL80a), while nonstructural proteins include pp52 (UL44) and p130 (UL57). Vornhagen et al. developed a recombinant HCMV IgM ELISA for Biotest (Biotest AG, Dreieich, Germany) with only peptides derived from nonstructural proteins pp52 (amino acids 297 to 433) and p130 (amino acids 545 to 601). In particular, it was found that the indicated portion of the UL57 gene product is a dominant IgM antigen which may be superior in both sensitivity and specificity to fragments from other HCMV proteins for detection of IgM antibodies during primary HCMV infection. Recombinant proteins and their fragments have been studied in a Western blot or immunoblot assay for their reactivity to IgM-positive serum samples prior to being included in an ELISA. The group of M. P. Landini, in close association with Abbott Laboratories (Abbott Park, Ill.), developed two versions of the HCMV IgM immunoblot assay with both recombinant proteins or peptides and viral proteins from purified virus preparations . In the new version of the assay , the viral section of a slot blot contains the entire viral proteins pp150 (UL32), pp82, pp65 (UL83), and pp28 (UL99) purified by gel electrophoresis, while the recombinant section contains only portions of pp150 (amino acids 595 to 614 and 1006 to 1048), p130 (amino acids 545 to 601 and 1144 to 1233), pp52 (amino acids 202 to 434), and pp38 (amino acids 117 to 383). A preliminary evaluation of the new immunoblot assay indicated that 13 of 80 (16%) IgG- and IgM-negative serum samples and as many as 38 of 200 (19%) IgG-positive, IgM-negative serum samples did react with one or more of the viral or recombinant proteins, while 126 of 126 (100%) IgM-positive serum samples reacted variably. In order to render these highly nonspecific results interpretable, an algorithm for reading of test results had to be introduced. Thus, only serum samples reactive with at least one viral and one recombinant protein or serum samples reactive with at least three recombinant protein bands were considered positive for IgM. By using this approach, a sensitivity of 100% and specificity of 98.6% were reached with respect to the consensus of two of the most used commercial ELISAs (Behring AG, Marburg, Germany, and DiaSorin, Saluggia, Italy). This assay was used as a reference test for development of the Abbott AxSYM CMV IgM microparticle enzyme immunoassay, with microparticles coated with the indicated portions of three structural (pp150, amino acids 595 to 614 and 1006 to 1048; pp65, amino acids 297 to 510; and pp38, amino acids 117 to 373) and one nonstructural (p52, amino acids 202 to 434) protein. This assay, when compared to a consensus given by three commercial HCMV IgM immunoassays (discordant results were resolved by immunoblot), showed a relative sensitivity, specificity, and agreement of greater than 95%. In addition, the assay was able to detect seroconversion very early and displayed a higher positive reactivity rate than the commercial assays tested on pregnant women . The level of cross-reactivity was 3.3%. The diagnostic utility of the AxSYM IgM assay in detecting low levels of IgM antibody (not detected by other commercial assays) in some serum samples is stressed by the finding that some of these serum samples contain low-avidity IgG , a marker of primary HCMV infection (see IgG Avidity Assay). At least one additional approach has been reported, with a combination of two HCMV peptides derived from pp150 (UL32, amino acids 1011 to 1048) and pp52 (UL44, amino acids 266 to 293) for IgM detection and a combination of peptides from pp150 (amino acids 1011 to 1048), pp28 (amino acids 130 to 160), and gB (amino acids 60 to 81) for optimal IgG detection . Sensitivity was 96.4% for the IgM assay with respect to a viral lysate-based ELISA. Although the development of immunoassays based on use of recombinant viral proteins or peptide epitopes represents major progress towards standardization of serological assays, these assays do not appear to be reliable from the diagnostic standpoint due to exceedingly high sensitivity and somewhat low specificity. In a recent study, 10 of 42 (23.8%) potentially cross-reactive or interfering serum samples were scored IgM-positive with a commercial ELISA based on use of recombinant HCMV antigens, whereas two commercial ELISAs based on use of viral lysates detected zero and one positive sample, respectively, in the same panel of problematic serum samples . Indeed, false-positive results still represent the major pitfall of HCMV IgM serology. In this respect, in a recent retrospective review of 325 consecutive pregnant women referred to our laboratory over a 2-year period because of a positive IgM result and a suspicion of primary HCMV infection, as many as 188 (57.8%) were found to be IgM negative by two different in-house-developed capture ELISAs in the absence of primary infection . Interpretation of positive IgM results. : Once the specificity of a positive IgM result has been verified, the interpretation of the clinical significance of IgM antibody present in the serum of a pregnant woman begins. We must recall that the IgM antibody response, which is currently detected in primary HCMV infections of both immunocompetent and immunocompromised patients, may also be detected during recurrent infections of the immunocompromised person, but generally not in the immunocompetent host. Thus, IgM detection in the serum of a pregnant woman is likely to be a reliable marker of a primary HCMV infection. However, IgM can reveal different clinical situations which can be related to the acute phase of a primary HCMV infection, the convalescent phase of a primary HCMV infection, or the persistence of IgM antibody. The kinetics of the HCMV-specific IgM antibody response during primary infection may vary greatly among individuals and depends substantially on the test or commercial kit used for testing. However, in general, high to medium levels of IgM antibody (peak titers) can be detected during the first 1 to 3 months after the onset of infection (acute or recent phase), after which the titer starts declining (convalescent or late phase) (Fig. ; M. G. Revello and G. Gerna, unpublished data). By using two capture ELISAs, it was shown that of nine immunocompetent individuals, four became negative for IgM within 6 months, three within 12 months, while two remained IgM positive for more than a year after the onset of primary infection . A recent study compared the sensitivities of the same two in-house-developed IgM capture assays based on use of viral lysates and a commercially available recombinant IgM assay . The kinetics of the IgM antibody response as determined on 213 sequential serum samples from 76 pregnant women with primary HCMV infection was grossly overlapping , showing a low-level IgM antibody response persisting for several months (M. G. Revello, G. Gorini, M. Parea, and G. Gerna, unpublished data). We define persistent IgM antibody response as the detection of stable levels of HCMV-specific IgM antibody for longer than 3 months. Although varying among different individuals, levels of persistent IgM antibody are mostly low, perhaps representing the sustained tail of an IgM response following a primary infection in some subjects . In a recent survey of 137 pregnant women confirmed to be positive for HCMV-specific IgM, only 60 (43.8%) were diagnosed as having primary HCMV infection acquired during pregnancy, whereas 39 (28.5%) had persistent IgM. In 38 (27.8%) of the 137 women, the IgM kinetics could not be determined due to the availability of only a single serum sample . FIG. 4. | (A) Kinetics of IgG, IgM, and neutralizing (Nt) antibody (Ab) response as well as IgG avidity index (AI) in a pregnant woman with primary HCMV infection. (A) Kinetics of IgG, IgM, and neutralizing (Nt) antibody (Ab) response as well as IgG avidity index (AI) in a pregnant woman with primary HCMV infection. (B) Kinetics of infectious virus and different virus products in the blood of the same pregnant woman as in A during the convalescent phase of a primary HCMV infection. Ag, antigenemia; Vir, viremia; DNA, DNAemia; IE mRNA, immediate-early mRNA; +, positive; -, negative; GE, genome equivalents; PBL, peripheral blood leukocytes. (M. G. Revello and G. Gerna, unpublished data.) FIG. 5. | Kinetics of IgM antibody response in 76 pregnant women with primnary HCMV infection as determined in 213 sequential serum samples by using two in-house-developed capture assays in parallel. Kinetics of IgM antibody response in 76 pregnant women with primnary HCMV infection as determined in 213 sequential serum samples by using two in-house-developed capture assays in parallel. IgM assays were based on the use of (thin line) virus lysate and (thick line) a commercial recombinant IgM assay . (M. G. Revello, G. Gorini, M. Parea, and G. Gerna, unpublished data.) IgG avidity assay. : When the presence of HCMV-specific IgM antibody in the serum of a pregnant woman cannot be directly related to a primary infection during pregnancy, an IgG avidity assay can help distinguish primary from nonprimary HCMV infection. This assay is based on the observation that virus-specific IgG of low avidity is produced during the first months after onset of infection, whereas subsequently a maturation process occurs by which IgG antibody of increasingly higher avidity is generated. Only IgG antibody of high avidity is detected in subjects with remote or recurrent HCMV infection. Avidity levels are reported as the avidity index, expressing the percentage of IgG bound to the antigen following treatment with denaturing agents, such as 6 M urea. The utility of the assay in diagnosing a primary infection has been reported for a variety of viruses . Measurement of IgG avidity is also of value in determining the duration of primary HCMV infection . We have shown that mean avidity index values relevant to serum samples collected less than 3 months after onset of primary infection were 21% +- 13%, whereas mean avidity index values for serum samples from subjects with remote HCMV infection were 78% +- 10% . Thus, the presence of high IgM levels and a low avidity index are highly suggestive of a recent (less than 3 months) primary HCMV infection. In a recent study, an avidity index above 65% during the first trimester of pregnancy could reasonably be considered a good indicator of past HCMV infection, whereas in all women with a low avidity index (<=50%), there was a risk of congenital HCMV infection. The risk increased with the gestational age at the time of testing . That is, only 2 of 12 (16.7%) women with a low avidity index during the first trimester of pregnancy transmitted the infection to the fetus, whereas in utero infection of the fetus was found in 6 of 15 (40.0%) women with a low avidity index detected during the second or third trimester of pregnancy , approaching the transmission rate reported by several groups . A negative predictive value of 100% was found when the avidity index was determined to be high or moderate before 18 weeks of gestation . On the other hand, when the avidity index was calculated at 21 to 23 weeks of gestation, it failed to identify some women who transmitted the virus, with a negative predictive value of 90.9% . Figure shows the maturation of HCMV-specific IgG avidity in 560 sequential serum samples from 176 immunocompetent individuals with primary HCMV infection (M. G. Revello and G. Gerna, unpublished data). It can be observed that in the interval between 4 and 6 months after the onset of infection, while most avidity index values are intermediate, a minor portion are either low (<30%) or high (>50%). This implies that in some pregnant women examined during the first trimester of pregnancy, a low avidity index may be related to a primary infection acquired prior to conception (false-positive result with respect to primary infection during pregnancy), while a high avidity index observed in the second trimester of pregnancy does not necessarily exclude a primary infection acquired during pregnancy. Recently, the ability of three IgG avidity assays to detect a primary HCMV infection was found to approximate 100%, whereas the ability to exclude a recent infection was shown to range from 96% to 32%. These data indicate that standardization of the assay is urgently needed . FIG. 6. | Kinetics of IgG avidity index (maturation of HCMV-specific IgG) in 560 serum samples from 176 pregnant women with primary HCMV infection. Kinetics of IgG avidity index (maturation of HCMV-specific IgG) in 560 serum samples from 176 pregnant women with primary HCMV infection. (M. G. Revello, and G. Gerna, unpublished data.) Neutralizing antibody. : It has also been reported recently that determination of HCMV neutralizing antibody may be an additional useful parameter for identification and timing of primary HCMV infection via a single serum sample . A neutralizing antibody response was not detected for 15 weeks (range, 14 to 17 weeks) after onset of primary infection. On this basis, it was concluded that the absence of neutralizing antibody during the convalescent phase of a primary HCMV infection is a reliable marker of primary infection, whereas the presence of neutralizing antibody rules out a primary infection in the previous 15 weeks. However, although it is well known that the neutralizing antibody response is the last to be mounted after a primary HCMV infection , the reported 15-week delay appears too extended, at least for immunocompetent subjects. When we tested 89 serum samples from 22 pregnant women with primary HCMV infection with the same neutralizing assay, we found neutralizing antibodies in 9 of 20 (45%) serum samples collected within 30 days, 20 of 23 (87%) serum samples collected within 30 to 60 days, and in all 46 (100%) serum samples collected >60 days after onset (M. G. Revello and G. Gerna, unpublished data). Thus, the absence of neutralizing antibody in a serum sample from a pregnant woman containing HCMV IgG and IgM may indeed provide additional evidence of recent primary infection. In contrast, the presence of neutralizing antibody is of no help in interpreting a positive IgM result. Conclusions. : The most definitive diagnosis of primary HCMV infection in a pregnant woman is by detection of seroconversion, i.e., the appearance of HCMV-specific IgG antibody during pregnancy in a previously seronegative woman . When this result cannot be achieved, detection of IgM antibody during pregnancy as well as during follow-up (whenever possible) can be used to determine clinically significant primary HCMV infection. Further testing by the IgG avidity test may be of great help in both confirming and clarifying the clinical significance of IgM antibody. When, at the end of the diagnostic algorithm, a primary HCMV infection is either diagnosed or suspected, prenatal diagnosis should be offered to a pregnant woman to verify whether the infection has been transmitted to the fetus. However, prior to performance of prenatal diagnostic procedures, the diagnosis of primary infection may be further confirmed or substantially supported by performing assays for detection of virus or virus products in the blood of the mother . FIG. 7. | Schematic of diagnosis of primary HCMV infection in pregnancy, including both serologic and virologic approaches. Schematic of diagnosis of primary HCMV infection in pregnancy, including both serologic and virologic approaches. AI, avidity index; Ag, antigenemia; Vir, viremia; DNA, DNAemia; IE mRNA, immediate-early mRNA; NT, neutralization test; Ab, antibody; pos, positive. Detection of Virus and Viral Products in Maternal Blood : Following primary infection, HCMV can be recovered from multiple body fluids such as saliva, urine, and vaginal secretions for a variable period of time. However, virus shedding from the same body sites may occur during reactivations and reinfections as well. Thus, the recovery of HCMV from these biological materials does not allow differentiation between primary and nonprimary infections in either immunocompetent or immunocompromised individuals. In the last decade, it has been clearly shown that only detection and quantitation of HCMV in blood has a high predictive value for HCMV disease in immunocompromised patients with either primary or recurrent HCMV infections . In addition, virus detection in blood has been reported to be diagnostic of primary HCMV infection in immunocompetent individuals , whereas in immunocompromised patients it is indicative of both primary and nonprimary infections. During the last decade, several methods have been developed to detect and quantify HCMV in blood. The most widely used assays include determination of viremia, i.e., infectious HCMV in blood; determination of antigenemia, i.e., number of pp65-positive peripheral blood leukocytes; quantification of HCMV DNA in whole blood (DNAemia), leukocytes (leuko-DNAemia), or plasma; determination of immediate-early and late mRNA (RNAemia); and search for the presence of circulating cytomegalic endothelial cells (CEC) in blood. An extended review of the methodological aspects and clinical applications of different assays for quantitation of HCMV has been published recently . Viremia. : Conventional methods for determination and quantitation of viremia are time-consuming because they are based on the appearance of cytopathic effect and include determination of 50% tissue culture infectious doses and plaque assays. These methods have been replaced by the "shell vial" assay, which provides results within 24 h. Following its introduction in the early 1980s , the assay was later rendered quantitative based on the assumption that each p72-positive fibroblast in a human fibroblast monolayer is infected by a single leukocyte carrying infectious virus . The shell vial monolayer is stained with either the immunofluorescence or the immunoperoxidase technique and a monoclonal antibody reactive with the HCMV major immediate-early protein . Then, the number of positive nuclei is counted . Since it was shown that a single monoclonal antibody may not identify virus strains with mutations in the relevant epitope of the major immediate-early protein, virus identification in our laboratory is performed with a pool of monoclonal antibodies reactive to different epitopes of p72 (G. Gerna, E. Percivalle, and M. G. Revello, unpublished data). In immunocompromised patients, the presence of HCMV viremia is commonly associated with a high risk of developing HCMV disease . Thus, its determination represents a useful parameter for initiation of antiviral treatment , monitoring of the efficacy of antiviral treatment , and detection of treatment failure due to emergence of a drug-resistant HCMV strain . However, major disadvantages of the viremia assay are its low sensitivity, the toxicity of peripheral blood leukocyte suspension for fibroblast monolayers, and the loss of HCMV viability in stored clinical samples . FIG. 8. | (A) Viremia, indicating the presence in a shell vial monolayer of HCMV p72-positive fibroblast nuclei following cocultivation with peripheral blood leukocytes carrying infectious virus and immunostaining by fluorescein-conjugated p72-specific monoclonal antibody. (A) Viremia, indicating the presence in a shell vial monolayer of HCMV p72-positive fibroblast nuclei following cocultivation with peripheral blood leukocytes carrying infectious virus and immunostaining by fluorescein-conjugated p72-specific monoclonal antibody. (B) Antigenemia ex vivo, showing immunofluorescent staining with a pool of monoclonal antibodies of pp65-positive peripheral blood polymorphonuclear leukocytes from a patient with AIDS and disseminated HCMV infection. (C) Antigenemia in vitro, showing pp65-positive polymorphonuclear leukocytes from a healthy blood donor following cocultivation with HCMV-infected human umbilical vein endothelial cells and immunofluorescent staining with the same pool of pp65-specific monoclonal antibodies used in B. (D) Circulating cytomegalic endothelial cell with a pp65-positive leukocyte (arrow). Immunofluorescent staining was done with a pool of pp65-specific monoclonal antibodies. Antigenemia. : The antigenemia assay detects and quantifies peripheral blood leukocytes, mostly polymorphonuclear leukocytes and, to a much lesser extent, monocytes/macrophages, which are positive for the HCMV lower matrix phosphoprotein pp65 . This HCMV protein, which was initially believed to be the major immediate-early protein p72 , is transferred to polymorphonuclear leukocytes from infected permissive cells via transitory microfusion events between two adhering cells . The antigenemia assay has been optimized and standardized by using in vitro-infected leukocytes . The methodological aspects of this assay have been reviewed recently . Experience obtained with transplant recipients has shown that antigenemia becomes positive earlier than viremia but later than DNAemia at the onset of infection, and it becomes negative later than viremia but earlier than DNAemia in the advanced stage of a systemic infection ; high antigenemia levels are often associated with HCMV disease; the assay is widely used for monitoring of HCMV infections and antiviral treatment ; and during ganciclovir treatment of primary HCMV infections, antigenemia levels may increase for up to 2 to 3 weeks despite the efficacy of treatment as shown by the disappearance of viremia, prompting clinicians to erroneously change antiviral drugs . A major advantage of the antigenemia assay is rapidity in providing results in a few hours, while major disadvantages are the limited number of samples processed per test run and the subjective component in slide reading . DNAemia. : Detection and quantification of HCMV DNA in blood has become a major diagnostic tool for transplant recipients. To this purpose, two major approaches have been used, PCR and hybridization techniques. For PCR, two main types of competitors have been used in the quantitative-competitive PCR: homologous competitors containing small deletions or insertions with respect to the target sequence , and heterologous competitors having the target sequence for primers as the target nucleic acid but differing in the intervening sequence . In addition to in-house-developed methods, a commercially available method has been developed by Roche (Cobas Amplicor CMV monitor test; Roche Molecular Systems, Branchburg, N.J.) for both detection and quantification of HCMV DNA . Finally, a new and interesting approach to the quantification of viral DNA is the detection and measurement of PCR products as they accumulate, thus overcoming the limited linear dynamic range of the traditional quantitative PCR. This technique, referred to as real-time PCR, is now being tested (Perkin Elmer, Applied Biosystems, Foster City, Calif.) and is based on the release of fluorescent dye molecules at each PCR cycle, the intensity of which is proportional to the amount of DNA in the sample . Among the hybridization techniques amplifying the signal generated rather than the viral DNA itself, two have become commercially available for quantification of HCMV DNA: the Digene hybrid capture system CMV DNA assay (version 2.0; Abbott Laboratories, Abbott Park, Ill.) and the branched DNA assay (Bayer, Chiron Corporation, Emeryville, Calif.). The hybrid capture system is based on the formation of a DNA-RNA hybrid which is captured by a monoclonal antibody specific for the hybrid and is then reacted with the same monoclonal antibody labeled with alkaline phosphatase. The hybrid is finally detected with a chemiluminescent substrate, whose emission is proportional to the amount of target DNA present in the sample . The second-generation hybrid capture system assay has been reported to have increased sensitivity and, thus could be considered for detection of viral DNA in the blood of immunocompetent hosts (see below). The branched DNA assay is based on the use of branched DNA amplifiers (branched probes) containing multiple binding sites for an enzyme-labeled probe. The target DNA sequence binds to the branched DNA molecule, and the complex is revealed by a chemiluminescent substrate whose light emission is directly proportional to the target DNA present in the sample . In immunocompromised patients, HCMV DNA quantification has been shown to be useful for follow-up of disseminated infections and evaluation of the efficacy of antiviral treatment. In addition, it is useful for the diagnosis and local evaluation of the effect of antiviral treatment at special body sites, such as the eye and nervous system . Finally a special application concerns its use for prenatal diagnosis of HCMV infection and for quantification of viral DNA in amniotic fluid samples (see below). RNAemia. : Detection of HCMV transcripts in blood is currently considered a marker of HCMV replication in vivo and late viral transcripts in particular are considered to better reflect active HCMV replication and dissemination . With reverse transcription-PCR, false-positive results may result from the difficulty in differentiating between RNA- and DNA-derived PCR products in the case of unspliced transcripts . Unlike reverse transcription-PCR, detection of mRNAs by the recently introduced nucleic acid sequence-based amplification (NASBA) method, which allows specific amplification of unspliced RNA in a DNA background , appears very useful for different populations of transplant recipients . Recently, two retrospective studies, in which preemptive therapy of both solid organ and hematopoietic stem cell transplant recipients was antigenemia guided, monitoring of HCMV pp67 mRNA (a late viral transcript) by NASBA appeared to be a promising tool for initiation and termination of preemptive therapy for solid organ transplant recipients with reactivated HCMV infection , whereas monitoring of immediate-early mRNA expression appeared to be a useful parameter for initiation of preemptive therapy in hematopoietic stem cell transplant recipients . At this time, prospective studies with NASBA assays are ongoing in transplant recipients, whereas preliminary data on the kinetics of immediate-early mRNA in immunocompetent individuals with primary HCMV infection are already available . Endotheliemia. : The term endotheliemia was introduced to indicate HCMV-infected CEC in the peripheral blood of immunocompromised patients. CEC were first described in 1993 by two independent groups and were shown to be endothelial in origin and fully permissive for HCMV replication. CEC are derived from infected endothelial cells of small blood vessels, which progressively enlarge until they detach from the vessel wall and enter the bloodstream. More recently, CEC have been studied in hematopoietic stem cell transplant recipients and in AIDS patients with disseminated HCMV infection . In recent years, the introduction of highly active antiretroviral therapy for AIDS patients and the adoption of prophylactic and preemptive therapy approaches for transplant recipients have nearly eliminated CEC from blood of these patient groups. However, CEC may still be found in the blood of fetuses and newborns with symptomatic congenital HCMV infection (M. G. Revello, E. Percivalle, and G. Gerna, unpublished data). Virus and viral products in blood of immunocompetent persons as an aid for diagnosis of primary infection. : Although there is an extensive amount of data obtained from studies with immunocompromised patients , very few data are available on the presence of HCMV in the blood of immunocompetent individuals with primary infection . In particular, little has been done to assess the diagnostic value of virus detection in the blood of nonimmunocompromised patients. Recently, an investigation was conducted on the peripheral blood leukocytes of 52 immunocompetent individuals (40 pregnant women) with primary HCMV infection by quantitation of pp65 antigenemia, viremia, and leukoDNAemia . pp65 antigenemia was detected in 12 of 21 (57.1%), 4 of 16 (25%), and 0 of 10 patients examined 1, 2, and 3 months after onset, respectively. Viremia was detected in 5 of 19 (26.3%) patients during the first month only. Finally, leukoDNAemia was detected in 20 of 20, 17 of 19 (89.5%), and 9 of 19 (47.3%) patients tested 1, 2, and 3 months after onset, respectively. Four (26.6%) of 15 patients were still DNAemia positive at 4 to 6 months, whereas none were positive at >6 months. No assay was positive in any of 20 subjects with remote infection or of 9 subjects with recurrent infection. In addition, virus levels were low by all assays. The conclusion of the study was that primary HCMV infection can be rapidly and specifically diagnosed whenever any of the studied virologic markers is detected in blood. On this basis, dating of the onset of infection can also be attempted . Viremia, i.e., virus recovery from blood, allows diagnosis of primary infection in about 25% of cases during the first month after onset. In fact, HCMV could not be recovered from the blood of 86 blood donors and in only one study was HCMV isolation from the blood of 2 of 35 normal donors reported . However, in this case, the interpretation of positive HCMV recovery from the blood of healthy people may hypothetically be referred to the convalescent phase of an unknown asymptomatic primary infection. Antigenemia may allow diagnosis of primary HCMV infection in 50% of patients in the first month and in 25% of patients in the second month after onset of infection. Again, positive antigenemia was never reported in immune healthy subjects or in patients prior to transplantation. Finally, DNAemia and, in particular, leukoDNAemia allow diagnosis of primary HCMV infection in 100% of subjects examined within 1 month after onset of infection and in 98% of those tested within 2 months. The presence of viral DNA in the leukocytes of healthy people is a more controversial issue. While in three reports viral DNA was found in virtually all seropositive healthy adult volunteers and in most seronegative persons when monocytes or peripheral blood leukocytes were examined, other investigators failed to detect viral DNA by PCR in monocytes or peripheral blood leukocytes or reported low (4 to 6%) positivity rates . Recent studies support the concept that viral DNA is not detected in the peripheral blood leukocytes of HCMV-seropositive immunocompetent individuals , emphasizing the utility of viral DNA detection in blood as a parameter for diagnosing primary HCMV infection. More recently, a new virologic parameter has been found to be useful for diagnosis of primary HCMV infection, detection of immediate-early mRNA by the NASBA technology. The clinical specificity of this parameter was first assessed in healthy individuals with remote or recurrent HCMV infection, showing its consistent negativity. Then, immediate-early mRNA was detected in the blood of all subjects with primary HCMV infection in the first month after onset of infection, whereas the proportion of positive subjects declined over time and became negative >=6 months after onset of infection . Thus, immediate-early mRNA kinetics appears to be comparable to that already reported for viral DNA, which suggests that detection of immediate-early mRNA in the blood of immunocompetent individuals can be considered an additional marker of recent primary HCMV infection. However, if immediate-early mRNA detection appears to be slightly more sensitive than DNA detection in diagnosing the early phase of primary HCMV infection , it appears to be slightly less sensitive in detecting the late phase of primary HCMV infection . Clinical Signs and Symptoms : The great majority of primary HCMV infections in the immunocompetent host are clinically silent . In addition, less than 5% of pregnant women with primary infection are reported to be symptomatic, and an even smaller percentage suffer from a mononucleosis syndrome . Thus, a primary HCMV infection cannot generally be diagnosed on clinical grounds alone. However, careful collection of the clinical history may be extremely useful for detecting minor clinical symptoms and dating the onset of infection. Whenever a primary HCMV infection is diagnosed in a pregnant woman, an interview is mandatory. Apart from major clinical findings observed in HCMV mononucleosis (such as fever, cervical adenopathy, sore throat, splenomegaly, hepatomegaly, and rash) which are not commonly detectable, if a pregnant woman is carefully questioned by experienced personnel, minor symptoms typical of HCMV mononucleosis, such as malaise, fatigue, headache, and myalgia, can be recalled, allowing quite precise dating of the onset of infection in the majority of cases . In addition, a slight increase in serum levels of liver enzymes (alanine transaminase, aspartate transaminase) may help in dating the onset of infection. In a survey conducted on 60 pregnant women with primary HCMV infection, mild clinical symptoms and/or liver function abnormalities were detected in as many as 38 (60%) . In a more recent survey (M. G. Revello, and G. Gerna, unpublished data) conducted on 244 pregnant women with primary HCMV infection, clinical symptoms were present in 166 (68.1%), fever (60.2%), fatigue (48.8%), and headache (26.5%) being the most frequent symptoms . In addition, 70 (42.1%) women reported three or more symptoms. The high rate of symptomatic primary HCMV infections in pregnancy may be explained by the careful medical interview. Whether the pregnancy-associated immunosuppression might play a critical role remains to be determined. Dating of primary HCMV infections in pregnancy is crucial for at least three reasons. The first refers to prognosis, in the sense that primary infection acquired just before conception is generally assumed to represent a lower risk than primary infection acquired during pregnancy. The second refers to prenatal diagnosis. It seems to be important to delay prenatal diagnosis as long as possible with respect to the onset of infection in order to minimize the rate of false-negative results . In fact, false-negative results have been reported despite the use of the most sensitive techniques available . Finally, primary infection early in pregnancy implies greater likelihood of congenital disease. TABLE 1 | Clinical and laboratory findings in 244 pregnant women with primary HCMV infection Diagnosis of Recurrent Infection in the Mother : It is generally accepted that transplacental transmission of HCMV infection represents a major risk during primary infection. However, over the past 20 years, data have accumulated that recurrent HCMV infections during pregnancy may also cause congenital infections. Thus, there is a rough correlation between the rate of maternal seropositivity and the rate of congenital infections . The estimated risk of intrauterine HCMV transmission during both primary and recurrent HCMV infections of pregnant women from low-income and high-income backgrounds in the United States is reported in Fig. . Diagnosis of recurrent infection can be accomplished by virus isolation or viral antigen or viral DNA detection in clinical samples, such as samples from the genital tract or cervix or urine, other than blood in the absence of concomitant serologic and virologic markers of primary HCMV infection, i.e., HCMV-specific IgM, low-avidity IgG, absence of neutralizing antibodies, and absence of virus and viral markers in blood. In this respect, it is very important to emphasize that, in order to ascertain the diagnosis of a congenital HCMV infection following a recurrent infection of the mother, at least the following requirements must be satisfied: the mother must be defined as immune to HCMV at least 1 year prior to pregnancy; a prepericonceptional HCMV infection (see below) must be excluded; and HCMV should be recovered from the genital tract. Thus, only prospective well-designed and extended epidemiological studies will be able to define the true impact of recurrent maternal infections (either reactivations or reinfections) in determining congenital HCMV infections in the future. Maternal Prognostic Markers of Fetal Infection : Thus far, no reliable prognostic markers of transmission of HCMV infection to the fetus have been identified in the mother. Recently, no correlation has been found between virus load in blood and the clinical course of HCMV infection in primary infections of immunocompetent subjects , in contrast to immunocompromised patients . In addition, no correlation has been found between virus load in blood and intrauterine transmission of the infection . Similarly, human immunodeficiency virus type 1 has been reported to be transmitted from mother to fetus within a wide range of maternal plasma human immunodeficiency virus type 1 RNA levels . Furthermore, no correlation was found either between persistence of viral DNA in blood for 3 months or >3 months and the risk of fetal infection or between gestational age and the risk of intrauterine transmission . With respect to the last issue, it was found, in agreement with a previous study , that HCMV transmission occurred in 50%, 40%, and 71% of fetuses after maternal infection in the first, second, and third trimester of pregnancy, respectively . In addition, virus recovery during pregnancy from the cervical tract or urine in both primary and recurrent infections is a poor indicator of risk of intrauterine transmission . The neutralizing antibody response has also been investigated as a potential prognostic marker of intrauterine transmission. Lower neutralizing antibody titers were detected in transmitting mothers with primary HCMV infection compared to nontransmitters, suggesting an association of neutralizing activity and intrauterine transmission . In the same study, a significant correlation was also observed between neutralizing activity and antibody avidity, thus suggesting that a maturation of antibody avidity is necessary for production of high levels of neutralizing antibodies, while a defect or delay in avidity maturation may play a role in intrauterine HCMV transmission . Symptomatic congenital HCMV infections have been noted in infants born to mothers with prepregnancy anti-HCMV immunity . Moreover, intrauterine transmission of HCMV from immune mothers to their infants has been related to reinfection with a different virus strain capable of causing symptomatic infections, as measured by the acquisition of new antibody specificities against epitopes of the glycoprotein H of the reinfecting HCMV strain . However, only prospective studies will be able to define the frequency of reinfection in immune pregnant women and its clinical impact on congenital infections. Finally, a lymphoproliferation assay against HCMV has been reported to provide an early marker of fetal infection after primary HCMV infection in pregnancy . In that study, all eight women with positive lymphoproliferative response gave birth to uninfected babies, whereas four of six women with negative responses delivered congenitally infected babies. Those findings suggested that depression of cell-mediated immunity in pregnant women after primary infection may represent a marker of fetal infection. COUNSELING : Once a diagnosis of primary HCMV infection has been achieved, the woman should receive sufficient information to make informed choices about further testing and options. This step is generally indicated with the term counseling. The term itself is vague and, in a way, misleading. Indeed, it is well recognized that the counselor is not supposed to give suggestions, opinions, or advice; rather, his or her role is that of facilitating informed choice by giving information and helping people to make decisions that reflect their value systems. Similarly, many terms such as informed decision, effective decision, and evidence-based choices are used to encompass informed choice. There is a growing tendency to consider informed choice as being "based on relevant knowledge, consistent with the decision-maker's values and behaviorally implemented" . According to this definition, an informed choice to undergo a test, such as prenatal diagnosis, occurs when the woman has relevant knowledge about the test, has a positive attitude towards undergoing a test, and undergoes it. An informed choice to decline a test occurs when the woman has a negative attitude towards undergoing a test, has relevant knowledge about the test, and does not undergo it. As a consequence, whenever the woman does not have relevant information or her attitudes are not reflected in her behavior, her choice should be considered uninformed. With this classification, very recently a model has been developed to provide a measure of informed choice capable of assessing both knowledge and values in relation to Down's syndrome testing in pregnancy . It would be very interesting to prove the validity of this approach for HCMV specifically. One of the major benefits would be to determine whether decisions are informed and, if not, the types of interventions required to increase rates of informed choice. Since no study has so far specifically addressed the issue of counseling of pregnant women for HCMV, data are not available concerning the number of health professionals actually providing the counseling, be it specialists in infectious diseases, virology, microbiology, psychology, obstetricians, or midwives. Similarly, nothing is known about how counseling is structured and performed or about the outcome, i.e., effect of counseling on informed decision making. Finally, it must be stressed that, at least in Italy, very few health professionals have received specific training in counseling, and in most instances, including our own, the counselor is a self-taught health professional with specific knowledge and wide experience. In less fortunate (from the woman's standpoint) but not infrequent cases, the health professional providing the counseling has neither specific knowledge nor experience, which often has disastrous consequences. In our experience, counseling is a complex process that proceeds step by step and is tailored to each individual. The first, most crucial step is the diagnosis of the mother. From a practical point of view, we do not provide extensive information on the possible clinical outcome until a diagnosis is firmly established. In particular, whenever a woman is referred to us because of IgM positivity detected in other laboratories during routine screening, we do explain what HCMV is and how one becomes infected. However, we focus primarily on the possible meaning of the laboratory results and multiple diagnostic options (false-positive result, persistent IgM, cross-reactive IgM due to herpesvirus infections other than HCMV, HCMV-specific IgM to be related to a preconceptional infection or to a primary infection in pregnancy). We do anticipate that only the last alternative may carry some risks to the fetus, and we explain to the woman that extensive information will be given only when the final diagnosis is reached. In this way, sufficient information is given to justify additional blood samplings and the time required for definite diagnosis without overly upsetting the woman. Once an acute or recent primary HCMV infection is diagnosed with certainty or high probability, the woman is given complete information about the risks of transmission, possible clinical outcome for the child, therapeutic possibilities in the case of symptomatic disease at birth, as well as prenatal diagnosis (if gestation time allows this option). All information is given within a framework that is as neutral as possible and in an unhurried fashion. Evidence (research)-based information is tailored to single cases, according to timing of maternal infection, certainty of diagnosis, and time of gestation. Possibilities and limitations of prenatal diagnosis, including the event of a false-negative result, are discussed in detail. If the mother has an acute or recent infection and is still viremic, the possibility of iatrogenic transmission is also discussed. The woman is also informed about the possibility of terminating the pregnancy, but she is referred to her obstetrician for specific counseling. Finally, if the woman undergoes prenatal testing and the fetus is found to be infected, results of prenatal diagnosis are discussed during an additional counseling session in order to provide the woman with the most accurate picture of fetal conditions based on biochemical/hematological, virological, and ultrasound findings. The woman (or the couple) then makes the final decision about continuation or termination of the pregnancy. DIAGNOSIS OF CONGENITAL INFECTION IN THE FETUS : After more than a decade, there are still those who do not favor prenatal diagnosis and those who consider prenatal diagnosis a major achievement in monitoring pregnancy. The main reasons of those who are against prenatal diagnosis are that the predictive value of a negative result is not yet quantified and because there is no specific antiviral treatment during pregnancy, the only clinical decision which can be made following prenatal diagnosis is whether or not to terminate the pregnancy; also, because only 35 to 40% of primary maternal infections are transmitted to the fetus and the great majority of congenital infections are asymptomatic , most pregnant women may prefer not to pursue prenatal diagnosis or termination of pregnancy . Reasons supporting prenatal diagnosis are to study of the natural history of congenital HCMV infection; to better prepare the family to face the health problems of the infant or young child; and to allow identification of prognostic markers of HCMV disease. Prenatal diagnosis may also represent the step preceding the potential introduction of antiviral therapy in the future . Finally, it can assist in decisions about continuing or terminating the pregnancy. Clinical samples currently used for prenatal diagnosis are fetal blood drawn by cordocentesis and amniotic fluid obtained by amniocentesis. Cordocentesis was introduced by Daffos et al. in the early 1980s and allows fetal blood sampling via the umbilical cord. It is usually performed after 17 weeks of gestation and is completed in a few minutes. Complications of cordocentesis, which occur at a low rate, may include transient bleeding, transient fetal bradycardia (7 to 9%), premature delivery (<2.0 to 5.0%), and fetal loss (1.7 to 1.9%) . Amniocentesis was first introduced by Bevis for diagnosis of immune hemolytic anemia and by Davis for diagnosis of congenital HCMV infection. Although rare, complications of amniocentesis may include fetal loss (<1%), leakage of amniotic fluid, and vaginal bleeding . By 1992, 20 cases of congenital HCMV infection diagnosed by amniocentesis were reported, as reviewed by Grose et al. . In subsequent years, the number of reports of congenital HCMV infection diagnosed prenatally increased progressively, with a major contribution provided by European investigators . Major clinical indications for prenatal diagnosis are documented primary HCMV infection in the mother, diagnosed according to the criteria reported above, and ultrasonographic abnormalities, known to be found frequently in fetal HCMV infection (such as intrauterine growth retardation, hydrops or ascites, and central nervous system abnormalities). Fetal Blood : Fetal blood and amniotic fluid samples are often drawn in parallel during procedures for prenatal diagnosis. Fetal blood can be used for both determination of HCMV-specific IgM antibody and quantification of viral load . However, the utility of IgM determination in fetal blood remains to be fully assessed . In addition, while several studies have established the diagnostic and prognostic value of the determination of viral load in the blood of immunocompromised patients, the clinical significance of the presence of virus and viral components in the blood of fetuses exposed to HCMV has never been fully investigated. Fetal blood may allow assessment of biochemical and hematological parameters, such as hemoglobin and platelet counts, and measurement of liver enzymes (gamma-glutamyl transferase, alanine aminotransferase, and aspartate aminotransferase). These nonspecific tests, although per se not very useful as prognostic markers of fetal disease, could help as complementary assays . Among the HCMV-specific assays, IgM antibody, which can be determined after 20 weeks of gestation, may be more helpful, even though this assay is known to possess a limited diagnostic value due to its low (20% to 75%) sensitivity . In two subsequent studies by the same group, the sensitivities of HCMV-specific IgM determination in fetal blood were 55.5% and 57.9%, while the specificity was 100% in both studies . While looking for prognostic markers of symptomatic infection, the authors observed that both frequency and levels of virus-specific IgM antibody were significantly higher in congenitally infected fetuses with ultrasound or biochemical/hematological abnormalities than in fetuses with normal ultrasound and biochemical findings . This finding was shown not to be related to the time of maternal infection, to the interval between onset of maternal infection and time of prenatal diagnosis, or to gestational age at amniocentesis. The finding was interpreted as potentially due to the fact that IgM-negative fetuses were sampled early during the infection on the basis of the following observations: four of seven IgM-negative fetuses had virus-specific IgM at birth, and an additional infant developed IgM antibody 65 days after birth, and the great majority of IgM-negative fetuses had a low viral load in blood and normal biochemical and hematological values. It is reasonable to assume that IgM-negative fetuses at 20 to 23 weeks of gestation could become positive in the advanced stages of pregnancy and thus not differ from fetuses with high-level IgM antibody. However, IgM levels were consistently higher in fetuses with HCMV disease, implying that this parameter represents a true prognostic marker of congenital HCMV disease. The same study addressed the issue of the diagnostic and prognostic value of viral load in fetal blood. While in the past, viremia was found to be negative in all eight HCMV-infected fetuses examined , in the above-mentioned study, evaluation of the prenatal diagnostic value of different tests for diagnosis of congenital infection on fetal blood showed that the sensitivity of antigenemia was 57.9%; of viremia, 55.5%; and of leukoDNAemia, 82.3%, while the specificity was 100% for all assays . Although only antigenemia reached levels of significance, higher levels of all virologic parameters determined were found in the groups of fetuses with ultrasound or abnormal laboratory findings compared to apparently normal congenitally infected fetuses. The following major conclusions were drawn from the study: no assay for detection of virus or virus components in fetal blood was sensitive enough to significantly improve prenatal diagnosis of intrauterine transmission of the virus; however, tests performed on fetal blood are confirmatory of results achieved on amniotic fluid (see below); fetuses with normal biochemical, hematological, and ultrasound findings, low or absent HCMV load in blood, and undetectable IgM antibody at 20 to 24 weeks of gestation may have a more favorable outcome; and taken together, virologic, laboratory, and ultrasound findings may contribute to a better prognostic definition of fetal infection . In a recent prospective study of 237 pregnancies at risk, in which prenatal diagnosis of congenital HCMV infection was achieved or excluded by amniocentesis with or without cordocentesis, the sensitivity of the IgM assay was comparable (51%) to that in previous studies , whereas PCR, which was positive in 17 of 41 cases (sensitivity 41%), and culture, which was positive in 2 of 27 cases (sensitivity 7%), were by far less sensitive . In this study, IgM antibody was unexpectedly detected in one fetus proven uninfected at birth, whereas all the other positive fetal blood samples were in total agreement with positive amniocentesis results. In unpublished studies on 86 pregnant women undergoing prenatal diagnosis for 88 fetuses , specificities and positive predictive values of all assays on fetal blood were 100%, whereas sensitivities and negative predictive values of both DNAemia and immediate-early mRNA were around 85% (M. G. Revello and G. Gerna, unpublished data). Thus, when fetal blood was used for prenatal diagnosis, all uninfected fetuses were correctly detected by all assays, whereas about 15% of them were missed by the most sensitive assays. Finally, there are two anecdotal observations demonstrating circulating CEC in the peripheral blood of two fetuses presenting with high viral load and ultrasound abnormalities (E. Percivalle and M. G. Revello, unpublished data). Circulating CEC have been detected in immunocompromised patients with disseminated HCMV infection , in association with very high levels of antigenemia and viremia and the presence of end organ disease . The finding of these cells in congenitally infected fetuses indicates a disseminated infection comparable to those reported in immunocompromised patients . FIG. 9. | Median levels (horizontal lines with values beside arrows) of HCMV antigenemia, viremia, and DNAemia and IgM ratio in fetal blood of symptomatic (sympt) and asymptomatic (asympt) congenitally infected fetuses and newborns. Median levels (horizontal lines with values beside arrows) of HCMV antigenemia, viremia, and DNAemia and IgM ratio in fetal blood of symptomatic (sympt) and asymptomatic (asympt) congenitally infected fetuses and newborns. All parameters evaluated were significantly higher in symptomatic than in asymptomatic subjects. Fetuses were considered symptomatic if they had ultrasonographic abnormalities, and newborns were considered symptomatic if they were born with clinical symptoms. PBLs, peripheral blood leukocytes; GE, genome equivalents. (M. G. Revello and G. Gerna, unpublished data.) TABLE 2 | Diagnostic value of different assays for prenatal diagnosis of congenital infection in 88 fetuses of 86 mothers with primary HCMV infection in pregnancy Amniotic Fluid : Due to its high sensitivity and absolute specificity (100%), HCMV isolation from amniotic fluid has been recognized as the reference method for prenatal diagnosis . At the beginning of the 1990s, the striking increase in the number of reported cases of prenatal diagnosis of congenital HCMV infection by virus isolation after amniocentesis was partly due to improvement of the tissue culture technology . In fact, the availability of monoclonal antibodies to HCMV major immediate-early protein and shell vial cell cultures allowed diagnosis to be performed within 16 to 24 h after sample collection . However, following the initial enthusiasm generated by findings showing that all cases of congenital infection could be diagnosed by virus isolation from amniotic fluid samples , several studies began documenting false-negative results of amniotic fluid cultures . With the advent of PCR, the question arose of whether the sensitivity of culture could be increased by using the PCR technique for HCMV DNA detection in amniotic fluid samples. In a retrospective study, the sensitivity of PCR for prenatal diagnosis was found to be only slightly superior (76.9%, or 10 of 13 cases detected) to virus culture (69.2%, or 9 of 13 cases detected) by either single-step or nested PCR . In other words, PCR could not avoid of three false-negative results out of 13 intrauterine infections diagnosed at birth. Subsequently, a substantial increase in the sensitivity of the PCR assay for prenatal diagnosis was obtained with a modified protocol for nested PCR . In the new assay, multiple instead of single aliquots and 100 mul instead of 20 mul of amniotic fluid were amplified and tested. By this approach, low DNA levels (1 to 10 genome equivalents) could be detected in a variable number of replicates of six amniotic fluid samples from four fetuses that previously had false-negative results. The specificity of the new assay was 100%, as demonstrated by negative results on 29 amniotic fluid samples from 22 pregnant women with primary HCMV infection who did not transmit the infection. However, the new assay failed to result in a positive prenatal diagnosis in the first amniotic fluid sample from a retrospective case that required two subsequent procedures for final diagnosis and did not rule out an additional false-negative diagnosis 8 weeks after maternal infection when used prospectively. Therefore, although the use of a very sensitive technique such as PCR can increase the sensitivity of prenatal diagnosis of HCMV congenital infection, it is reasonable to assume that a delay in intrauterine transmission of the infection may represent a major obstacle to achieving 100% sensitivity . These results were confirmed by other reports showing that even the combination of the most sensitive assays available, such as viral culture and PCR on amniotic fluid samples, may reach a sensitivity of about 70% to 80% . In a recent prospective study, global sensitivity, specificity, and positive and negative predictive values of prenatal diagnosis for HCMV detection in amniotic fluid and fetal blood (taken together) were 80%, 99%, 98%, and 93%, respectively, while the percentages were nearly overlapping if prenatal diagnosis was based on PCR of amniotic fluid alone . In a series of recent reports from a single group, the sensitivity of viral culture by the shell vial assay was found to be 50% to 62.5% and the specificity 100%, whereas the sensitivity of PCR was 100%, but surprisingly, the specificity was 67.3% to 83.3%, with a positive predictive value of 48% to 48.5% . This means that congenital HCMV infection was documented in only 12 of 27 fetuses or newborns found to be positive in amniotic fluid by PCR or in 16 of 33 fetuses or newborns reported subsequently . Recently, two false-positive PCR results in amniotic fluid in a series of 96 uninfected newborns (specificity, 97.9%) have been reported . On the other hand, false-positive PCR results in amniotic fluid have never been reported by other groups . In fact, when virus was detected by PCR (or culture) in amniotic fluid samples, it was consistently recovered from fetal tissues or excreted by newborns in all reported cases . Therefore, virus detection in amniotic fluid must be considered a marker of fetal and congenital infection . Also, in our experience, the detection of even small amounts of viral DNA (<100 genome equivalents/ml of amniotic fluid) has always correlated with congenital infection at birth . In our recent, as yet unpublished study , when prenatal diagnosis was based on amniotic fluid, the specificities and positive predictive value of all assays were again 100% or close to 100%, while the sensitivities and negative predictive value of the most sensitive assays for detection of viral DNA and mRNA were >=92% (M. G. Revello and G. Gerna, unpublished data). This means that, while nearly all uninfected fetuses were identified by each assay, about 7 to 8% of infected fetuses were missed by molecular assays and thus scored as false negative. Lazzarotto et al. suggested that the high sensitivity of PCR could detect small amounts of virus which could be cleared by the defenses of the mother or fetus. In addition, the authors suggested that the term "rate of intrauterine transmission of CMV" should be applied to indicate the percentage of amniotic fluid-positive samples rather than the percentage of HCMV-infected newborns or fetuses. In this respect, extreme caution must be used in evaluating these thus far unconfirmed results, based on the following considerations: prenatal diagnosis is a very delicate task and, as a rule, irrevocable decisions must be taken on the basis of test results, and PCR assays and the related containment measures must be extensively validated before being used for diagnostic purposes. The risk of HCMV transmission during antenatal diagnostic procedures performed in the presence of maternal DNAemia does not seem to be major, although it cannot be excluded . This conclusion seems to be supported by the observation that transmission rates were not different between women with a single prenatal sampling and women with multiple samplings and were not higher after initiation of prenatal diagnostic procedures compared to historical controls without prenatal intervention . Apart from the most sensitive techniques used, the sensitivity of prenatal diagnosis may be increased by repeated sampling; increasing gestational age at time of amniocentesis; increasing the time between onset of maternal infection and time of amniocentesis; and repeating ultrasonographic examinations. With reference to the first point concerning multiple sampling, it was found that for all undiagnosed infected fetuses except one, only one prenatal sample was collected. On the other hand, of 24 infected fetuses with multiple samples taken at different times during pregnancy, prenatal diagnosis was positive in 23 (96%). Of 44 pregnancies with transmission of virus to the fetus, 12 infected fetuses were diagnosed upon the second sampling . The correlation of gestational age and virus transmission was documented by the finding that prenatal diagnosis showed a sensitivity of 30% (6 of 20) if the first amniotic fluid sample was taken before 21 weeks of gestation, whereas of 35 women tested for the first time after 21 weeks of pregnancy, 26 (74%) were diagnosed as transmitters, 25 (71%) with tests on amniotic fluid and 17 of 28 (61%) with tests on fetal blood . The difference in sensitivity between amniocentesis before and after 21 weeks of gestation was found to be statistically significant. The same data were obtained by other groups . In addition, a correlation was found between time elapsed after onset of maternal infection and time of positive amniocentesis . All infected fetuses were detected when a mean interval of 7 weeks between maternal symptoms and amniocentesis had elapsed . Other authors recommended an interval of at least 4 weeks to avoid false-negative results . However, with these recommendations, a rate of 23% false-negative results would have been obtained, while with a time lapse of 7 weeks, positive antenatal diagnoses could be achieved in all cases . Repeated ultrasonographic examinations may help in only a small minority of fetuses with severe disseminated infection at autopsy . However, frequent ultrasonographic evaluations in pregnancies with evidence of vertical transmission may help to predict fetal damage (such as hydrocephaly, microcephaly, ventriculitis, or brain calcifications), thus identifying fetuses at significant risk of clinical sequelae . Fetal Prognostic Markers of HCMV Disease : As suggested by previous studies , it has recently been documented that fetal HCMV disease is preferentially associated with maternal infection occurring in the first part of pregnancy . By combining data on aborted fetuses with severe ultrasonographic abnormalities and infected newborns with poor outcome, it was observed that fetal HCMV disease was more severe if maternal infection occurred prior to 20 weeks of gestation. In fact, in this case the rate of fetal severe HCMV disease was 26% (10 of 38 fetuses), whereas only 1 fetus of 16 infected after 20 weeks of gestation had a minor sequela of retinitis . A potential role of fetal viral load as a prognostic factor has been advocated . Lamy et al. observed that viral load was very high in fetuses with brain ultrasonographic anomalies. Donner et al. , although not specifically investigating the issue, found that false-negative results in prenatal diagnosis were consistently associated with low viral load and asymptomatic infections both at birth and during follow-up. A correlation of high levels of viral load in the blood and the appearance of HCMV disease has been repeatedly reported in immunocompromised patients, representing the basis for the development of strategies of preemptive therapy in transplant recipients . In addition, recent studies have clarified the dynamics of HCMV replication in vivo and have claimed the predictability of HCMV disease based on a single viral DNA quantification in a blood sample drawn early during HCMV infection . However, the clinical significance of HCMV load in fetal blood and amniotic fluid of congenitally infected fetuses has not been fully investigated until very recently. In fetal blood, all virologic parameters tested to determine viral load, i.e., viremia, antigenemia, and leukoDNAemia, were found to be higher in fetuses with abnormalities than in fetuses with normal findings, although only levels of antigenemia were significantly different . In addition, as reported above, the level of virus-specific IgM antibody was significantly lower in fetuses with normal findings. These data appeared to justify the conclusion that congenitally infected fetuses with normal biochemical, hematologic, and ultrasound findings and low viral load in blood (together with low or undetectable IgM antibody) might have a more favorable outcome . On the other hand, in the amniotic fluid of mothers of 21 congenitally infected fetuses, quantification of HCMV DNA showed that median levels of HCMV DNA were 1.25 x 108 genome equivalents/ml in the group of fetuses with abnormal ultrasound findings at the time of prenatal diagnosis or with symptomatic infection at birth (n = 7) and 3.75 x 106 genome equivalents/ml in the group of fetuses with normal ultrasound findings at the time of amniocentesis and subclinical infection at birth (n = 14). This difference was not significant (P = 0.09), although this could be due to the small number of fetuses tested . In particular, very high levels of viral DNA (108 genome equivalents/ml) could be observed in both asymptomatic and symptomatic fetuses. In addition, some fetuses with asymptomatic infection showed levels of viral DNA in amniotic fluid of <100 genome equivalents/ml. This low level of viral DNA was found not only in fetuses tested in amniotic fluid shortly after maternal infection, but also in fetuses tested several weeks after maternal infection . However, all fetuses infected with low viral DNA levels were asymptomatic. In this respect, it is important to stress that all fetuses with viral DNA in amniotic fluid, including those with <100 genome equivalents/ml, were born with congenital infection. One reason why no correlation was found between HCMV load in amniotic fluid and clinical symptoms may be that viral DNA is accumulating in the amniotic fluid (C. Liesnard, F. Brancart, M. L. Delforge, F. Gosselin, F. Rodesch, and C. Donner, Abstr. 8th International Cytomegalovirus Conference, abstr. p. 15, 2001) instead of being cleared, as also indirectly shown by the lack of degradation of viral DNA in an amniotic fluid sample stored at 37C for at least 6 months (M. G. Revello, unpublished data). The most recent survey of our series indicates that in the fetal blood of symptomatic fetuses or newborns, all virologic parameters (antigenemia, viremia, and DNAemia) as well as IgM antibody levels were significantly higher than those of asymptomatic fetuses or newborns , confirming previous results. The difference in DNA level between fetuses born with symptomatic congenital infection and fetuses born with asymptomatic infection was also found to be significant in the amniotic fluid (M. G. Revello and G. Gerna, unpublished data). However, while levels of viral DNA were >105 genome equivalents/ml in all symptomatic fetuses and newborns, only 11 of 18 (61.1%) asymptomatic fetuses and newborns showed DNA levels greater than 105/ml of amniotic fluid, the remaining 7 showing DNA levels of <=102 genome equivalents/ml. Thus, while the difference in DNA level was not significant between asymptomatic and symptomatic fetuses and newborns when only DNA levels of >105 genome equivalents/ml were considered, such a difference became highly significant (P = 0.0073) when all cases in the asymptomatic group were considered. These results were in agreement with those reported by Liesnard et al. , showing 100% specificity of PCR and culture in detecting HCMV in amniotic fluid, but were in disagreement with those reported by Lazzarotto et al. , who suggested that small amounts of viral DNA were eliminated by the fetus without transmission of the infection. While these data were reported on a qualitative basis, the same group subsequently, using quantitative PCR, reported that levels of viral DNA of 103 to 105/ml of amniotic fluid were necessary to transmit the infection, whereas levels >105 were required to cause HCMV disease in the fetus. Thus, levels of <103 genome equivalents/ml were unable to transmit the infection and were cleared by the fetus during fetal life . Following qualitative PCR on amniotic fluid at 21 to 22 weeks of gestation, specificity and positive predictive value with respect to the presence or absence of HCMV infection in the fetus or newborn were only 67.3% and 48.5%, respectively. This means that as many as 17 of 33 (51.5%) fetuses or newborns had viral DNA in the amniotic fluid and were not infected. However, the authors claim that when viral DNA was quantified by PCR, levels of >103 genome equivalents/ml indicated a high probability of infection (12 of 12 were infected, with positive predictive value of 100%), while levels of <103/ml indicated a low probability of infection (only 4 of 21 were infected, with a negative predictive value of 81%). In addition, levels of viral DNA of >105 genome equivalents/ml were selected as a cutoff to indicate a high probability of disease (9 of 9 were ill, with a positive predictive value of 100%), whereas levels of <105 genome equivalents/ml indicated a low probability of disease (2 of 24 were ill, with a negative predictive value of 91.6%). It has recently been shown that newborns with symptomatic congenital HCMV infection have significantly higher levels of HCMV in the blood at birth and that clearance of virus takes longer than in subclinically infected infants . Moreover, it has been reported that infants with symptomatic congenital HCMV infection excrete larger amounts of virus in the first few months of life than those with asymptomatic infection . These data seem to indicate that quantification of viral load may correlate with clinical conditions, at least after birth. During fetal life, apart from the peculiar physiopathological condition of the fetus in its relationship with the mother, it is possible that viral load might correlate with clinical symptoms or pathological findings. In this respect, viral load in fetal blood as determined by antigenemia is significantly higher in fetuses with HCMV disease . Also, median values of viral DNA in amniotic fluid are markedly higher in fetuses with pathological findings . What is surprising in the above-mentioned studies and, more importantly, not confirmed by data from other laboratories is that viral DNA is often detected in amniotic fluid without being transmitted to the fetus. We conclude that, at this time, multiple difficult to determine variables, such as gestational age at maternal infection, timing of intrauterine transmission of the infection, timing of prenatal diagnosis, and, most important, the unfeasibility of a follow-up of the infection during fetal life represent the major obstacles to identification of a reliable prenatal marker of symptomatic congenital HCMV infection. However, high viral loads in amniotic fluid may be associated with either symptomatic or asymptomatic congenital infections, while low viral loads are consistently associated with asymptomatic congenital infections. FIG. 10. | Comparative median levels (horizontal lines with values beside arrows) of DNA in amniotic fluid of mothers with symptomatic (sympt) and asymptomatic (asympt) fetuses. Comparative median levels (horizontal lines with values beside arrows) of DNA in amniotic fluid of mothers with symptomatic (sympt) and asymptomatic (asympt) fetuses. While in both groups all (symptomatic) or most of the (asymptomatic) fetuses showed high levels of HCMV DNA (>105 genome equivalents [GE]/ml), in the asymptomatic group there were seven fetuses, four of whom had DNA levels of <102 genome equivalents/ml and three with undetectable DNA levels at the time of amniocentesis. (M. G. Revello and G. Gerna, unpublished data.) DIAGNOSIS OF CONGENITAL INFECTION IN THE NEWBORN : At birth, or during the first 2 weeks of life, postnatal diagnosis of congenital HCMV infection is required either to confirm the results of prenatal diagnosis or to investigate transmission of the virus to neonates born to women who experienced a suspected or ascertained primary HCMV infection during pregnancy . The gold standard method for diagnosis of congenital HCMV infection is represented by virus isolation in human fibroblasts in the first 2 weeks of life, because subsequent virus excretion may represent neonatal infection acquired in the birth canal or following exposure to breast milk or blood products . Urine and saliva are the clinical samples of choice for culture. Urine samples may be stored at 4C for 7 days, with the isolation rate dropping to only 93%, whereas storage at room temperature or freezing decreases infectivity dramatically . In the 1980s, methods for rapid virus isolation were developed, based on the use of monoclonal antibodies to the HCMV major immediate-early protein p72 associated with low-speed centrifugation of clinical samples onto monolayers of human fibroblasts grown on coverslips inserted on the bottom of shell vials . The shell vial method was subsequently adapted to 96-well microtiter plates, where it showed a sensitivity of 94.5% and a specificity of 100% compared to standard virus isolation in a series of 1,676 newborn urine specimens . The assay retained the same level of sensitivity and specificity when saliva was tested instead of urine . PCR was first used for HCMV DNA detection in the urine of congenitally infected babies at the end of the 1980s . Urine samples were repeatedly frozen and thawed. When compared with the standard tissue culture isolation procedure, the PCR assay followed by dot blot hybridization showed a sensitivity and specificity of 100%. Obvious advantages of PCR over culture were the small amount of sample required; the short time required for test results (24 to 48 h versus 2 to 28 days); the ability to use frozen specimens with noninfectious virus; and no need for extensive DNA purification measures. These results prompted clinical virologists to test for the presence of viral DNA in the blood of congenitally infected newborns. First, Brytting et al. reported detection of HCMV DNA in the serum of five of five congenitally infected infants tested within 2 weeks after birth, while two of these five newborns were negative for HCMV-specific IgM. In 1995, Nelson et al. reported detection of HCMV DNA in the serum of 18 of 18 (100%) infants with symptomatic congenital HCMV infection, 1 of 2 infants with asymptomatic congenital HCMV infection, and 0 of 32 controls. In 1999, Revello et al. investigated the diagnostic and prognostic value of HCMV load as determined by different assays in the blood of 41 newborns with congenital infection and 34 uninfected newborns with respect to conventional virus isolation from urine. Sensitivities of HCMV DNAemia (by PCR), antigenemia, viremia, and IgM determination were 100%, 42.5%, 28.2%, and 70.7%, respectively, while specificity was 100% for all assays. That study concluded the following: (i) determination of viral DNA in blood by PCR at birth appears to be as sensitive and specific as virus recovery from urine for diagnosis of congenital HCMV infection; (ii) significantly higher levels of HCMV load are detected in infants with congenital symptomatic HCMV infection; and (iii) virus clearance from blood occurs spontaneously in both symptomatic and subclinically infected newborns, even though the process takes longer in symptomatic newborns . A further simplification of the procedure for detection of viral DNA in the blood of congenitally infected infants was proposed in 1994 by Shibata et al. with dried blood spots stored on filter paper, as originally suggested for human immunodeficiency virus type 1 by Cassol et al. . Although Shibata et al. reported an extraordinarily high rate of viral DNA positivity in the blood of healthy Japanese newborns (25.1%), suggesting a possible carryover contamination in the laboratory, the Japanese approach was verified by others . Thus, with dried blood spots collected from babies in the first days of life during routine screening procedures for genetic and metabolic disorders, Barbi et al. reported that eight of eight symptomatic and 11 of 11 asymptomatic congenitally infected babies were positive for HCMV DNA when extraction was done with medium instead of water. Therefore, the method showed 100% sensitivity and specificity with respect to virus recovery by culture. More recently, determination of HCMV immediate-early mRNA in the blood of congenitally infected newborns by NASBA has been used to diagnose congenital HCMV infection . The immediate-early mRNA NASBA assay had 100% sensitivity in detecting 12 congenitally infected newborns examined during the first week of life and previously found to be positive for both HCMV DNAemia and virus recovery from urine. However, immediate-early mRNA was detected for a significantly shorter period of time (median 37 days) than DNAemia (median, 87.5 days; P = 0.04). This trend was attributed to the stricter association of immediate-early mRNA with the early stages of HCMV infection in vivo. Indeed, recently, by using an in vitro model, it has been shown that viral DNA detected in polymorphonuclear leukocytes is transferred from HCMV-infected cells, whereas an aliquot of immediate-early mRNA is synthesized in these cells, which indicates an active, albeit abortive, replication of HCMV . Thus, it seems reasonable to speculate that immediate-early mRNA detected in blood might represent a more reliable marker of active HCMV infection. Finally, it is important to stress that the new assay showed 100% specificity, since no immediate-early mRNA was ever found in healthy newborns. As already mentioned, IgM antibody determination has somewhat limited sensitivity in diagnosis of congenital HCMV infection. The solid-phase radioimmunoassay described by Griffiths and Kangro had a sensitivity of 89% and specificity of 100% . With IgM ELISA, the specificity was nearly 95% and the sensitivity approximately 70% when congenitally infected infants were tested . Similarly, with a capture ELISA method with enzyme-labeled monoclonal antibody, the level of sensitivity for congenital HCMV infection was found to be 70.7% . TREATMENT OF CONGENITAL INFECTION : Although specific antiviral drugs, such as ganciclovir and foscarnet, have been available for several years for treatment of life-threatening or sight-threatening HCMV disease in immunocompromised patients, their use for treatment of congenital HCMV infection remains undefined due to a paucity of data. In principle, two levels of treatment could be considered, prenatal (during fetal life) and postnatal (based on severity of clinical symptoms). Foscarnet is a competitor of pyrophosphate, while ganciclovir acts as a competitor of guanosine during viral DNA synthesis . However, the degree of toxicity of the two drugs must be carefully considered, with special regard to the renal toxicity of foscarnet and the hematologic toxicity of ganciclovir. Anecdotal studies do not support the efficacy of antiviral drug therapy in fetal HCMV infection. The first was reported in 1993 . Ganciclovir was administered in utero for 12 days to a 29-week-old fetus with congenital HCMV infection, thrombocytopenia, and elevated gamma-glutamyl transferase levels. Following therapy, the virus titer in amniotic fluid and fetal urine dropped, viral DNA disappeared from the blood, and the platelet count and gamma-glutamyl transferase level became normal. However, stillbirth occurred at 32 weeks of gestation, and HCMV inclusion bodies were detected in several organs at autopsy. Two subsequent reports concerned the administration of HCMV hyperimmune globulin to HCMV-infected fetuses with the intent of mitigating the damaging effects of HCMV infection . An additional attempt at fetal therapy in a congenitally infected fetus presenting with high gamma-glutamyl transferase values was reported . Three subsequent doses of ganciclovir (100 mg, 50 mg, and 200 mg) were administered intra-amniotically 1 week apart starting at 25 weeks of gestation. Viremia dropped after the first drug administration, becoming undetectable at 29 weeks of gestation. However, levels of antigenemia, DNAemia, and infectious virus in amniotic fluid did not change during follow-up. IgM antibody, which was quite high at 23 weeks, decreased progressively and became negative at 29 weeks. At birth, the baby showed petechiae, microcephaly, hepatomegaly, thrombocytopenia, increased alanine transaminase levels, and hearing impairment. A few anecdotal reports on the use of ganciclovir in congenitally infected infants have also been discouraging . In these reports, indications for treatment were acute symptoms of HCMV organ localization (pneumonia, hepatitis) or generalized congenital disease. However, in at least one case, ganciclovir was administered with the specific aim of preventing further involvement of the central nervous system . Drug dose, duration of treatment, and age at initiation of treatment varied in single reports. However, in all studies a reduction in or temporary cessation of virus excretion was observed during therapy. In a case of ganciclovir therapy of congenital HCMV hepatitis, viremia was the first parameter to become negative, followed by antigenemia during the first 2 weeks of treatment . Clearance of virus from urine required an additional week of treatment. Clinical efficacy was excellent, with improvement of all biochemical parameters by the end of therapy in the absence of side effects. However, 9 days after cessation of therapy, a resumption of virus replication occurred in both blood and urine. On the other hand, more controlled multicenter clinical trials have begun evaluating the use of antiviral drugs for treatment of infants with symptomatic congenital HCMV infection. A phase II study carried out on a group of 47 infants with congenital infection to investigate the efficacy of ganciclovir treatment following intravenous administration at 12-h intervals for 6 weeks showed that ganciclovir administration had to be stopped in eight infants because of toxicity (mostly neutropenia); the most common side effects were neutropenia and elevation of liver enzymes; excretion of virus with urine decreased during treatment, returning to pretreatment levels after drug discontinuation; and the most significant clinical result was hearing improvement, observed in 5 of 30 infants (16%) after 6 months of follow-up or later . The main goal of antiviral chemotherapy would be to treat pregnant women with primary HCMV infection in order to (hopefully) prevent transmission of the virus to the fetus. In this respect, the combination of hyperimmune globulin with antiviral drugs of low or negligible toxicity could represent the best approach to preventing vertical HCMV transmission in the future . Prior to achieving this major goal, it will be highly problematic to identify a truly efficacious treatment because cases of congenital infection are currently diagnosed and, thus, identified weeks or months after virus transmission to the fetus; there are a wide spectrum of congenital HCMV diseases; the natural course of the disease is erratic; and irreversible damage has already occurred before any therapeutic intervention in the fetus or newborn infant can be attempted . PREVENTION OF CONGENITAL INFECTION : Congenital HCMV disease is still a major public health problem, which does not appear to be resolved by means other than active immune prophylaxis, i.e., vaccination . Due to the fact that congenital infection is the leading infectious cause of mental retardation in children, HCMV is considered a prime candidate for eradication from the human population through vaccination . As reported above, while some data point to the role of recurrent infections in causing defects in infants born to mothers with prepregnancy immunity , the overwhelming majority of studies indicate that congenital HCMV disease is the result of primary maternal infection during pregnancy. Thus, the ultimate goal of the HCMV prevention program is to develop a vaccine which can be administered to seronegative women of childbearing age to prevent the occurrence of primary HCMV infection during pregnancy. Over the last 30 years, attempts to develop an HCMV vaccine have been directed at five major strategic approaches: (i) live attenuated vaccines; (ii) recombinant virus vaccines; (iii) subunit vaccines; (iv) peptide vaccines; and (v) DNA vaccines. The first attempts were aimed at preparing a vaccine containing live attenuated virus. However, several problems had to be faced from the beginning: the vaccine virus strain may persist in the body as a latent virus and periodically reactivate; reactivations may not be preventable; safe markers of attenuation of vaccine strains had to be identified ; animal models for HCMV are not available; and HCMV may be oncogenic in vivo, as suggested by its ability to transform both human and embryonic hamster cells in vitro . Live Attenuated Vaccines : The first vaccine was developed by Elek and Stern , using a strain of AD169 that was presumed to have been attenuated by propagating the virus 56 times in human fibroblasts. This vaccine, administered subcutaneously, elicited a good, although transitory, neutralizing antibody response in the absence of virus shedding. Almost simultaneously, a live attenuated virus vaccine was developed at Wistar Institute in Philadelphia, Pa., with the Towne strain, which was recovered from the urine of a newborn with congenital infection and propagated 125 times in human embryonic fibroblasts . Shortly thereafter, the vaccine was shown to induce a significant antibody response . In three subsequent vaccine trials carried out in renal transplant recipients, it was shown that vaccinated seronegative recipients of kidneys from seropositive donors had a significant reduction in disease severity but not infection with respect to the control group receiving placebo; the Towne vaccine induced not only an antibody response but also a cell-mediated immune response, as determined by a lymphoprolipheration assay; and the vaccine strain was not excreted and was not found to undergo latency . Recombinant Virus Vaccines : The recent finding that laboratory-adapted HCMV strains lack a large DNA fragment found in primary HCMV isolates and a low-passage reference HCMV strain (Toledo) has shed some light on the pathogenesis of HCMV infection . In particular, in a DNA fragment referred to as ULb' , as many as 19 open reading frames have been identified, some of which are particularly interesting, such as UL146, coding for an alpha (CXC) chemokine functionally involved in active recruitment of polymorphonuclear leukocytes . It is generally believed that extensive propagation of Towne in human fibroblast cell cultures has been the major factor causing such a significant modification of the viral genome and, potentially, of virus pathogenicity and immune response to vaccine administration. Therefore, new alternative strategies were developed, aimed at inserting the entire genome of Toledo, subdivided into four fragments, in the genetic background of Towne, generating four chimeras, each representing a potential vaccine strain. The efficacy of these four recombinant virus strains in inducing antibody and cell-mediated immune responses in the absence of clinical symptoms is still under evaluation . Subunit Vaccines : Given the teratogenic role of HCMV, over the last decade several investigators have addressed their efforts to the development of a subunit vaccine. In this respect, major antigenic sites of the immune response to HCMV which are potential components of a subunit vaccine are viral glycoproteins gB (UL55) and gH (UL75), the principal targets of the neutralizing antibody response, and viral phosphoproteins pp65 (UL83) and pp150 (UL32), which are the dominant targets of the cytotoxic T-lymphocyte (CTL) immune response . The last is also directed to the nonstructural protein p72 (UL122). One of the most promising subunit vaccines has been based on the use of a recombinant gB molecule which was mutagenized to eliminate a cleavage site and deprived of the transmembrane region prior to being combined with a new adjuvant, MF59, based on an oil-in-water emulsion of squalene . Following administration of three doses of this vaccine to seronegative subjects at 0, 1, and 6 months, levels of neutralizing antibody and antibody to gB 2 weeks after the third dose exceeded those in seropositive control subjects, while a fourth dose induced a prompt rise in antibody level. HCMV gB vaccine was also shown to produce significant levels of antibody at mucosal surfaces . However, the induction of good levels of neutralizing antibody to gB may not be sufficient to prevent fetal infection and disease. Thus, induction of an immune response to other viral proteins which are targets of neutralizing antibody or CTL may well be required . It has been shown that pp65 is largely dominant in generating in vitro HCMV-specific CTL, followed to a much lesser extent by gB, gH, and p72 . In a prevalence study carried out in seropositive healthy individuals, CTL to IE1-exon 4 were nearly as prevalent as to pp65 (76% versus 92%), while gB- and pp150-specific CTL were detected in about one-third of subjects . Based on multiple reports showing that gB is the major target of the humoral and pp65 is the major target of the cell-mediated immune response, noninfectious defective enveloped particles of HCMV referred to as dense bodies have recently been proposed as an ideal natural vaccine immunogen, consisting mostly of pp65 and gB and lacking viral DNA . Dense bodies have been shown to induce neutralizing antibody and T-helper 1- and CTL-mediated immune responses in mice, thus representing a potential basis for the future development of a recombinant nonreplicating vaccine against HCMV . The most immunogenic viral genes have been inserted in different vector systems to generate recombinant subunit vaccines. This line of research, already active in the past , has more recently taken advantage of a nonreplicating canarypox expression vector in which the HCMV gene coding for gB or pp65 has been inserted. Multiple reasons justify the use of canarypox vaccine vectors: avipoxviruses accept large amounts of foreign DNA, thus directing synthesis of foreign proteins, and canarypoxviruses do not produce progeny in mammalian cells and are immunogenic in nonavian species without producing disease . In a group of 20 seronegative adults randomly receiving either a canarypoxvirus (ALVAC) expressing HCMV gB or an ALVAC expressing the rabies virus glycoprotein (controls), with all subjects receiving a dose of Towne vaccine after 90 days, the ALVAC-CMV (gB) was found to prime the humoral immune response to HCMV gB, which was much earlier and higher and persisted longer than in controls . In a subsequent study by Berencsy et al. with a canarypox-HCMV pp65 recombinant in a phase I clinical trial on seronegative volunteers, it was found that pp65-specific CTL were elicited after only two vaccinations and were CD8+, while pp65-specific lymphoproliferative response was detected in vitro, revealing stimulation of CD4+ T cells. In addition, a consistent antibody response to pp65 was elicited. Although these results do not provide evidence of the protective effect of the canarypox-HCMV pp65 recombinant vaccine against HCMV disease, immunization with this vaccine seems to confer immunity similar to that provided by natural infection. The protective effect of CTL could be reinforced by the involvement of CD4+ T cells and by the as yet unexplained role of pp65-specific antibodies . Peptide Vaccines : An important advance in viral immunology has been the finding that peptide fragments of immunogenic viral proteins, referred to as minimal cytotoxic epitopes, when properly selected, bind to MHC molecules with high affinity . Synthetic versions of these peptides, which are commonly 8 to 11 amino acids long, bind to MHC molecules, sensitizing targets to lysis by CD8+ CTL without requiring any further proteolytic processing to act as CTL epitopes. Peptide vaccines have a disadvantage in that they are of limited efficacy due to their limited HLA specificity. However, by using a computer algorithm, the sequence of HCMV pp65 was scanned for HLA-A*0201-binding motif peptides, selecting a nonamer peptide (amino acids 495 to 503) capable of sensitizing target cells for lysis in the absence of activity on HLA-mismatched cells . However, it was observed that minimal cytotoxic epitopes had to be suspended in a strong adjuvant to be able to elicit an efficient CTL response . Lipidated peptides have been shown to confer a good immune response to different pathogens in a safe and effective way . It is known that control of viral infections requires T-helper (TH) besides CD8 cell activity. Thus, efforts have been directed at increasing TH cells by using strong TH epitopes derived from tetanus toxin or synthetic peptides . With transgenic mice expressing both HLA class I (A*0201) and class II DR1 molecules and inoculated with a peptide mixture containing an HCMV-derived class I HLA (A*0201)-restricted CTL peptide epitope (pp65495-503) and tetanus toxin-derived MHC-binding TH epitope, a significant enhancement in CTL response was observed compared to that in transgenic mice expressing only class I molecules .subspecies included in this table. DNA Vaccines : DNA vaccines are based on the in vivo expression of heterologous genes carried by plasmid vectors. Results of DNA vaccination are determined by both the efficiency of delivery and the level of expression of the heterologous gene. The efficiency of delivery has been improved by using liposomes as an adjuvant . The efficiency of expression has been related to the promoter used. In a recent comparative study, the promoter of the HCMV immediate-early gene was shown to be more active than other viral promoters in determining gene expression . In the mouse model, inoculation of a plasmid vector expressing pp89 of mouse CMV (homologous to the immediate-early gene of HCMV) conferred protection against subsequent experimental infection with sublethal doses of the virus, as shown by the decrease in viral load in different organs and by the induction of pp89-specific CTL . Subsequently, a similar degree of protection and reduction in viral titer were observed in mice inoculated with a plasmid expressing the M84 gene (the homolog of HCMV pp65) . Naked-DNA immunization of mice with plasmids expressing gB or pp65 of HCMV has been shown to induce high antibody titers and dose-dependent CTL immune responses, respectively, indicating that both humoral and cellular immune responses to HCMV can be elicited in mice following DNA vaccination . PERCEPTION OF THE PROBLEM : In the last 30 years, more than 800 papers have been published dealing with the epidemiology, diagnosis, and outcome of vertically transmitted HCMV infections. In the introduction of most if not all of these papers, the dreadful scenario of estimated figures of dead and handicapped children (and relevant social and health care costs) due to congenital HCMV infections is duly recounted. Apparently, researchers from different countries on different continents have fully recognized the existence of the problem and its consequences, judging from the huge number of studies funded and performed. Nevertheless, very little has been done in practical terms to face this health problem (see below). A worrisome consideration emerges spontaneously: HCMV appears to be a rewarding topic from a scientific standpoint (let alone a lucrative business for companies), but remains a lingering danger to every pregnant woman. Almost 30 years have elapsed since Elek and Stern reported their study with the compelling title "Development of a Vaccine against Mental Retardation Caused by Cytomegalovirus Infection In Utero" , but no licensed vaccine is available yet. Indeed, the absence of an effective means of prevention has been seen by many investigators as an insurmountable obstacle to sensible screening practices. Due to the suggestion reported in recent papers that recurrent maternal HCMV infection can be as dangerous as primary infections , preconceptional vaccination is no longer considered a solution by some investigators . Thus, it seems that the best and only practical approach suggested by many scientists and health authorities is to ignore the existence of HCMV. However, ignoring the problem does not make it disappear, and it is surprising that in an age in which litigiousness is rising, parents of congenitally HCMV-infected children with severe handicaps have not yet taken legal action against health professionals for not offering the possibility of screening and thus potentially preventing the birth of an affected child. Finally, to the best of our knowledge, no study has ever investigated the awareness of women of childbearing age about possible risks carried by HCMV infection acquired during pregnancy. It is possible that in this era in which access to information is greatly facilitated by the Internet, knowledge among lay persons might have increased over that just a few years ago. Public awareness can be a potentially strong lever for raising awareness of the problem of congenital HCMV infection. In 1989, during a BBC television program in Britain, parental pressure groups called for a program to identify women susceptible to HCMV in order to prevent congenital HCMV infection and to be counseled about how to avoid the infection. However, an editorial published in The Lancet less than a week after the call concluded that screening for HCMV was inappropriate and avoidance of infection impractical (Editorial, Lancet ii:599-600, 1989). A further plea for screening (D. O. Ho-Yen, letter, Lancet ii:803, 1989) was dismissed shortly afterwards (P. M. Preece, letter, Lancet ii:1101, 1989), and to date, no screening program is available in Britain (P. D. Griffiths, personal communication). Although the success of public initiatives may have been limited in the past, it is doubtful that today a similar call would remain unmet. TABLE 3 | Public health impact of congenital CMV infection in the United States UNIVERSAL SEROLOGY SCREENING : Universal screening for HCMV by serology has been and still is a debated issue. In a way, it reflects the same dichotomy in the scientific community's attitude presented above: why study how to diagnose a condition that cannot be treated? An impressive number of papers have been published regarding the diagnosis of HCMV infections with special emphasis on immunocompetent individuals and particularly pregnant women. Different algorithms for HCMV monitoring in pregnant women have been developed and proposed, but none has ever been officially established in any health care program. To our knowledge, routine serologic screening of pregnant women has never been recommended by any public health authority in any country. The only exception was represented by Italy, where, in 1995 to 1998, determination of HCMV antibody was included in the panel of examinations (ToRCH screening) which could be performed free of charge for pregnant and nonpregnant women. An informal survey performed by us for this review confirmed that, presently, HCMV screening is not officially recommended in any of the following European countries: Austria (T. Popow, personal communication), France (L. Grangeot-Keros, personal communication), Switzerland (W. Wunderli, personal communication), Germany (G. Enders, personal communication), Belgium (C. Liesnard, personal communication), United Kingdom (P. D. Griffiths, personal communication), the Netherlands (A. M. van Loon, personal communication), Spain (F. de Ory, personal communication), or Italy (authors' observation), as well as Israel (S. Lipitz and E. Mendelson, personal communication), Canada (M. Chernesky, personal communication), and Japan (K. Numazaki, personal communication). However, most obstetricians do test pregnant women in Italy (authors' observation), Israel (S. Lipitz and E. Mendelson, personal communication), Belgium (C. Liesnard, personal communication) and France (L. Grangeot-Keros, personal communication), whereas determination of HCMV-specific antibodies is performed only on specific request in Austria (T. Popow, personal communication), Switzerland (W. Wunderli, personal communication), Germany (G. Enders, personal communication), and Japan (K. Numazaki, (personal communication) or in case of symptoms in the mother in the United Kingdom (P. D. Griffiths, personal communication) and the Netherlands (A. M. van Loon, personal communication) or in the presence of possible occupational hazards (pregnant nurses working in obstetrics and pediatrics) in Austria (T. Popow, personal communication) and Japan (K. Numazaki, personal communication). Similarly, the United States is not committed. In fact, the National Center for Infectious Diseases Internet site on HCMV , among the recommendations for pregnant women with regard to CMV infection, merely indicates that "Laboratory testing for antibody to CMV can be performed to determine if a woman has already had CMV infection" and that "Pregnant women working with infants and children should be informed of the risk of acquiring CMV infection and the possible effects on the unborn child." It recommends that specific testing should be restricted to "women who develop a mononucleosis-like illness during pregnancy." The Internet site developed by the National Congenital CMV Disease Registry is only apparently more committed: "Every woman of childbearing age should consider knowing her CMV status. Prior to pregnancy, consult your doctor to have a blood sample drawn and a CMV antibody test performed." It is clear that the scientific community will always be divided into those who support and those who do not support universal screening. However, we believe the time has come for considering this crucial issue under a different perspective. Should women of childbearing age be informed about HCMV so that they can make an informed choice about (voluntary) screening? Or, as advocated by many researchers, should women continue to be denied testing (and, consequently, information), until more data are available about the efficacy of hygienic practice and the outcome of recurrent maternal infection or an appropriate vaccine or specific therapy is licensed? In this era in which much emphasis is being put on the issue of appropriate patient information, there is growing evidence that withholding information on possible medical interventions is unethical (and legally risky). Indeed, if we consider that reliable and (relatively) inexpensive assays are available for determination of HCMV immune status; precautionary measures can be suggested to seronegative women; prenatal diagnosis procedures can be performed; and voluntary termination is an option in many countries, we believe that it is no longer acceptable or ethically justified to discourage HCMV screening. However, it must be recognized that we are confronted with two major challenges: educating doctors and offering women the opportunity to know their HCMV immune status prior to pregnancy. The first challenge is the most urgent and basic, as health professionals represent a key factor in screening. In fact, the choice of whether to have a screening test performed is greatly influenced by the personal attitude of the health professionals offering the screening . In addition, health professionals who have limited knowledge and thus are unable to inform patients may ultimately affect the choice . Therefore, family doctors, obstetricians, internists, pediatricians, and allied professional clinicians as well as medical social workers, nurse practitioners, and physician assistants need to be informed and educated so that they can provide relevant, good-quality, and unbiased information. Unless this step is accomplished, HCMV in pregnancy will remain a problematic issue. Second, women of childbearing age should be considered the best target population for HCMV prevention. Determination of HCMV immune status before pregnancy carries many advantages: (i) it may be less expensive, as only specific IgG can be safely determined; (ii) IgG-seropositive women could be reassured and informed that they do not need any further testing (this may account for 50 to 70% of the population in Western countries); (iii) IgG-seronegative women could be properly informed, so that whenever they will become pregnant they will already be aware of the possible risks and preventive measures; (iv) monitoring of IgG-seronegative pregnant women would be easier and cheaper without much need to resort to the armamentarium of assays now necessary for the confirmation or interpretation of laboratory test results (IgM antibody and IgG avidity assays) in women of unknown preconceptional serology. CRUCIAL IMPACT OF CORRECT COUNSELING : Many women receive their first generic (and often misleading) information about HCMV directly from the staff of the laboratory where HCMV-specific IgM is detected, sometimes well before the result is confirmed and correctly interpreted. Consultation with the obstetrician ensues, usually very shortly. Subsequently, depending on how knowledgeable the obstetrician is about HCMV and pregnancy, the woman can be either referred for further testing and counseling or offered the immediate option of terminating the pregnancy. No study has ever investigated how many pregnancies are terminated at an early stage (i.e., <12 weeks of gestation) on the basis of a positive IgM result and inadequate or misleading information ("your baby will be mentally retarded"). Thus, a great responsibility lies on the health professional providing the first information. Indeed, the first communication to parents is crucial since it may affect how information presented later is accepted or even whether it is sought. In addition, since some time may elapse between the first detection of IgM positivity and the final diagnosis or exclusion of primary HCMV infection, a tremendous level of anxiety, often increased by the opinions of additional "experts," can be experienced by the woman in the meantime. In our experience, such levels of anxiety cannot be relieved by any subsequent counseling no matter how careful and complete it may be. On the other hand, falsely reassuring counsel or failure to recognize the potential risk of a positive IgM result ("you are completely protected against the infection because you have both IgG and IgM") can be equally deleterious. Again, no formal study has specifically addressed the psychological consequences for parents of children with unexpected severe HCMV infection in terms of anxiety, depression, stress, and, most important, attitude towards the disabled child. However, our experience indicates that the impact may be disastrous, and the parents of a severely affected child are more likely to blame doctors for the birth of that child and to pursue legal action against them (authors' personal observation), as already shown for parents of children with Down's syndrome (; L. Parsons, J. Richards, and R. Garlick, letter, Br. Med. J. 305:1228, 1992). PRENATAL DIAGNOSIS: IS THERE A ROLE? : In 1992, Pass questioned the usefulness of prenatal testing because predictive values of positive and negative results were unclear and the absence of prognostic markers and fetal therapy prevented both obstetricians and pregnant women from making informed decisions whenever an intrauterine infection was diagnosed . After a decade, some of the points raised are still unanswered (fetal therapy is still lacking as well as reliable prognostic markers of fetal disease) and some issues have become even more debated, such as the predictive value of positive results obtained with molecular techniques. On the other hand, the predictive value of negative results is now quite well defined, thanks to the contributions of many European investigators. Similarly, different diagnostic approaches and techniques have been evaluated and compared, a great deal of information has been obtained, and the overall reliability of prenatal diagnosis has definitely improved over time. Moreover, since both the limitations and benefits of prenatal testing have been better defined, counseling of pregnant women has similarly improved. Thus, prenatal diagnosis has assumed a crucial role in the management of pregnancies complicated by primary HCMV infection. However, the exact role of prenatal diagnosis in the management of pregnancies complicated by primary HCMV infection has not been defined. We reviewed the records of 179 women with a definite diagnosis of primary HCMV infection and known outcome who were examined at our institute during the period from 1990 to 2000 (M. G. Revello and G. Gerna, unpublished data). In Italy, pregnancy can be terminated in the first 12 weeks of gestation upon the woman's request. Voluntary termination in the period from 13 to 24 weeks is allowed only if continuation of pregnancy will severely affect the mental or physical health of the woman, while voluntary termination is not allowed beyond 24 weeks of gestation. Women included in the survey were thus divided into three groups according to the time of pregnancy at which primary HCMV infection was diagnosed . In the group of 73 women in whom primary HCMV infection was diagnosed at <=12 weeks of gestation, 17 (23.3%) asked for elective termination, 35 (47.9%) decided to undergo prenatal testing at 20 to 22 weeks of gestation, and the remaining 21 (28.8%) women chose to continue the pregnancy without any invasive procedure. In the second group of 68 women with primary HCMV infection diagnosed at 13 to 23 weeks of gestation, as many as 45 (66.1%) chose the option of prenatal testing, 22 (32.4%) continued the pregnancy, and only 1 (1.5%) woman aborted without resorting to prenatal testing. Finally, 3 of 38 (8.5%) women in whom HCMV infection occurred beyond 24 weeks of gestation underwent prenatal diagnosis. Thus, altogether, the option of prenatal diagnosis was chosen by 80 of 141 (56.7%) women with primary HCMV infection diagnosed at <=23 weeks of gestation. However, the most interesting finding comes from examination of the results of prenatal diagnosis and the subsequent behavior of the women who chose antenatal testing. In fact, while all 46 pregnancies with a prenatal diagnosis negative for fetal infection went to term (data not shown), only 14 of 37 (37.8%) pregnancies with documented intrauterine transmission of HCMV infection were terminated. These data suggest the following. (i) Prenatal testing represents a very important option in the case of primary infection during pregnancy, as documented by the acceptance by the majority of women to whom it was presented. (ii) Prenatal testing appears to be very beneficial in terms of overall reduction in the number of terminated pregnancies, since none of the pregnancies with negative antenatal results and less than half of those with fetal infection were terminated. In fact, although we have no data for comparison, we believe that the toll in terms of voluntary abortions would have been (much) higher without the option of prenatal diagnosis. (iii) Finally, the observation that the majority of women continued the pregnancy despite the knowledge of fetal infection clearly indicates that the main reason for undergoing prenatal testing was not (or not only) selective termination in case of a positive result, but rather eagerness to know whether the infection had been transmitted. Indeed, some women in this group carried long-sought pregnancies or did not contemplate the option of termination for ethical reasons, while we believe that some women changed their mind over time, eventually choosing to continue their pregnancy. TABLE 4 | Management and outcome of 179 pregnancies complicated by primary HCMV infection according to gestational age at time of diagnosis PRECONCEPTIONAL AND PERICONCEPTIONAL HCMV INFECTIONS : While the risks associated with primary HCMV infection during pregnancy have been well assessed and pregnant women can be properly counseled and make informed decisions, no information is available about the outcome of pregnancies complicated by primary HCMV infection acquired shortly before or around the date of conception. Similarly, there is no information about how long conception should be delayed following primary infection. We performed a retrospective study aimed at defining the risks associated with primary HCMV infection acquired <=3 months before the last menstrual period (preconceptional infections) or in the first 4 weeks after the last menstrual period (periconceptional infection) (M. G. Revello, M. Zavattoni, and G. Gerna, Abstr. 8th International Cytomegalovirus Conference, abstr. p. 66, 2001). Back records for 162 consecutive pregnant women with ascertained diagnosis of acute or recent primary HCMV infection and known outcome of pregnancy were thoroughly reviewed. Diagnosis of primary infection was based on either decreasing levels of virus-specific IgM and rising levels of IgG avidity, as determined from sequential serum samples, HCMV detection in blood, or both. Since dating of the infection was crucial for the purpose of the study, only women with a well-defined clinical history were considered. In each case, the kinetics of virologic parameters were considered in relation to symptoms and/or laboratory findings compatible with a primary HCMV infection in order to define the onset of the infection. By using these strict inclusion criteria, 26 women were identified in whom primary HCMV infection occurred either 2 to 11 weeks (median, 4 weeks) before the last menstrual period (10 women) or 1 to 4 weeks (median, 2 weeks) after the last menstrual period (16 women). In the group of 16 women with periconceptional HCMV infection, 5 (31.2%) elected to terminate their pregnancy before 12 weeks of gestation, while spontaneous abortion occurred in 2 (12.5%) women at 7 weeks of gestation . Products of conception were not examined for HCMV. On the other hand, five (31.2%) women underwent prenatal diagnosis at 17 to 23 weeks (median, 20 weeks) of gestation, and all five fetuses were found to be negative for HCMV infection. On the whole, nine pregnancies went to term, and seven of nine (77.8%) newborns examined at birth were found to be uninfected (including the five newborns examined during fetal life), whereas the remaining two (22.2%) were congenitally infected. Of these, one newborn presented with neurologic and metabolic alterations and developed mild sequelae (neuromuscular deficits of the left arm) at 6 months of age (the mother had had HCMV-specific IgM at 6 weeks of gestation and 1 day of moderate fever at 3 weeks of gestation). The other newborn was asymptomatic at birth, and the mother was found to be positive for HCMV-specific IgM at 12 weeks of gestation, while reporting a debilitating and sustained asthenia concomitant with the last menstrual period. In the group of 10 women with preconceptional infection, no spontaneous or induced abortion occurred, and two women underwent prenatal diagnosis at 18 and 23 weeks of gestation . Neither of the two fetuses was infected. All 10 pregnancies went to term, and 9 of 10 newborns (90%) were found to be HCMV free, while one newborn was subclinically infected. The mother of this newborn infant was sent to our center because of a positive IgM result at 6 weeks of gestation. Retrospective data revealed fever, asthenia, headache, and upper respiratory tract symptoms 8 weeks before the last menstrual period. A prenatal diagnosis procedure performed at 18 weeks of gestation, while the woman was still positive for HCMV DNAemia, gave negative results, as HCMV was neither isolated from nor detected in amniotic fluid or fetal blood. However, at birth the virus was isolated from the urine, and DNAemia was present in neonatal blood together with low levels of HCMV-specific IgM. Both DNAemia and IgM were still positive at 6 months of age, while the infant remained free of symptoms. Possible explanations for the discrepant results obtained antenatally include delayed transmission, iatrogenic transmission, and performance of prenatal procedures too early during pregnancy. All these factors have been repeatedly shown to affect the reliability of prenatal diagnosis, as discussed above. Although it must be pointed out that the true incidence of intrauterine transmission in our series could not be precisely defined because materials from spontaneous or induced abortions could not be examined for HCMV, the risk of fetal infection following maternal infection acquired before or immediately after conception appears to be substantially lower (10% and 22%, respectively) than that generally reported for infections acquired during gestation (40 to 50%). As for the prognosis of congenitally infected infants born to mothers with periconceptional infection, the numbers are far too small to allow us to draw a sound conclusion. These findings were confirmed in a more extended series . These results, albeit partial and preliminary, still can be useful in facilitating informed decisions by pregnant women. In particular, from a practical point of view, pregnant women with documented or suspected primary infection acquired before conception can be reassured and counseled to continue their pregnancy without resorting to antenatal testing unless required by parental anxiety. On the other hand, the option of prenatal diagnosis should be offered to women with periconceptional infection in view of the slightly higher incidence of transmission and the uncertainty of clinical outcome. In addition, as more information about the risks to the fetus becomes available from prospective monitoring, the proportion of women electing to undergo voluntary abortion at early stages of gestation will hopefully be reduced. Finally, as for how long to wait between primary HCMV infection and conception, no definite answer is available yet. However, on the basis of the data reported above, the observation that about 20% of immunocompetent subjects with documented primary HCMV infection are still DNAemia positive at 6 months after onset , and the consideration of DNAemia as a marker of potential infectivity, we suggest that at least 6 months should elapse prior to initiation of pregnancy. It must be stressed, however, that this is a general recommendation, and preconceptional primary HCMV infection should not be considered an indication for termination of pregnancy. TABLE 5 | Management and outcome of 26 pregnancies complicated by preconceptional or periconceptional primary HCMV infection CONGENITAL INFECTION FROM IMMUNE MOTHERS AND CLINICAL OUTCOME : The overall incidence of congenital HCMV infection varies from 0.24% to 2.2% . This percentage includes newborns born to mothers with primary infection during pregnancy as well as newborns born to mothers with existing immunity. The contribution of either component depends on the seroprevalence in a given population. It is generally recognized that primary HCMV infection carries the highest risk for symptomatic (including sequelae) congenital infection. However, much attention has been paid recently to the possibility that in utero transmission following recurrent maternal infection may result in adverse fetal outcome more frequently than previously thought. The epidemiologic correlation of seroprevalence and incidence of congenital infection supports this observation. Over time, individual cases of symptomatic infants born to apparently normal (K. Ahlfors, S. Harris, S. Ivarsson, and L. Svanberg, Letter, N. Engl. J. Med. 305:284, 1981; D. Rutter, P. Griffiths, and R. S. Trompeter, letter, Lancet ii:1182, 1985) or variably immunocompromised HCMV-immune mothers have been published, indirectly testifying to the rarity of the event. On the other hand, two large independent studies reported that asymptomatic infants born to immune mothers may develop long-term neurologic sequelae, particularly deafness, in 8% and 22% of cases. Moreover, in the study by Williamson et al. , the frequency of sequelae in infants born to immune mothers was comparable to that observed in subclinically infected children born to mothers with primary infection. In 1999, groups from Sweden, Belgium, and the United States reported long-term prospective studies of congenital HCMV infection. In the Swedish study , transient symptoms were noted at birth in an identical number (n = 6) of newborns born to mothers with confirmed primary or recurrent infection, whereas neurological symptoms at follow-up were recorded in four and two children born to mothers with confirmed primary or recurrent HCMV infection, respectively. The Belgian study reported that one of three severe congenital infections was due to a recurrent maternal infection and that one of two infants with late-onset hearing impairment was born to a mother with existing immunity. Finally, the U.S. study reported that of 20 newborns with symptomatic congenital HCMV infection at birth and defined maternal infection, 8 were born to mothers with primary infection, 8 to mothers with confirmed recurrent infection, and 4 to mothers with presumed recurrent infection. No difference was found in the severity of symptoms at birth or in long-term sequelae between children born to mothers with primary or nonprimary HCMV infection. In earlier studies performed by the same group in the same geographic area and on a comparable number of children, no symptomatic infection was observed at birth in newborns of immune mothers and a lower incidence of sequelae was found in newborns from mothers with nonprimary infection (8% versus 25%). No clear explanation for these discrepancies was given by the authors. Because of the potential implications of these studies as far as screening and vaccination policies are concerned, it is important to stress that all studies mentioned above suffered from a major intrinsic methodological weakness, i.e., none of them, least of all the U.S. study, was specifically designed to investigate the risk of severe HCMV-related abnormalities after recurrent maternal infection. In addition, in none of the quoted studies were mothers of congenitally infected children investigated for other possible causes that might have been responsible for the adverse outcome of pregnancy, such as smoking or drinking habits or illicit-drug consumption. Moreover, since the definition of primary versus recurrent infection relied only on the presence or absence of IgM (except whem seroconversion could be demonstrated or immunity was documented before pregnancy), the type of maternal infection remained uncertain in a fair number of cases in both the Swedish and U.S. studies. Finally, additional assays were not performed to rule out the possibility of a primary infection in the periconceptional period. Indeed, very large prospective studies of women known to be immune before pregnancy should be performed in order to define this issue. Although we doubt that such studies will ever be funded and carried out , still, defining the exact impact of recurrent maternal infections on neonatal disease should be the top priority for those interested in developing and implementing vaccine strategies. While it is possible that the contribution of recurrent maternal infection to symptomatic congenital infection might have been underestimated because of the difficulty in diagnosing it, the available data are still insufficient to allow modification of counseling of immune pregnant women. On the other hand, one wonders whether it would be beneficial to screen newborns for HCMV in view of identifying asymptomatic babies to be enrolled in long-term follow-up. This could be justified because it has been shown that asymptomatic congenitally infected children born to mothers with either primary or recurrent infection face a significant risk of developing sequelae, particularly hearing defects ; in these silently infected children, hearing loss has been shown to progress silently over time ; unrecognized hearing loss has a significant negative impact as far as language development, school performance, and communication skills are concerned; and early diagnosis of hearing impairment is mandatory for early intervention. Thus, much effort should be directed to developing rapid, inexpensive, and simple assays for the detection of HCMV in urine so that HCMV screening may become an additional routine test for all newborns. Finally, since it has been recently proposed that recurrences and unfavorable outcome might be related to reinfection with a new HCMV strain rather than reactivation of the endogenous strain, controlled studies of molecular epidemiology are much needed. TABLE 6 | Incidence of congenital HCMV infection in relation to rate of maternal immunity CONGENITAL INFECTION IN TWINS : A fascinating, albeit little investigated aspect of the complex relationship between HCMV and pregnancy concerns congenital HCMV infection in twins. So far, only 12 documented cases of congenital HCMV infection in twins and one case in a quadruplet pregnancy have been described. Two main observations derive from these reports. The first concerns transmission of the infection. In monozygotic twins with a monochorionic placenta, congenital HCMV infection has been observed to occur in both children, while in dizygotic twins with a dichorionic placenta, only one of the twins was generally infected. On the other hand, in three pairs of dizygotic twins with fused placentas, both twins were found to be infected in two pairs, whereas only one twin was infected in the remaining pair . Histopathological examination of the placentas showed inflammatory signs in cases of fetal infection, whereas they were apparently normal in noninfected siblings . In the case of congenital infection in a quadruplet pregnancy, HCMV immediate-early antigens were detected in three available placentas in the absence of viral inclusions . Our own experience is in keeping with these findings. We observed two twin pregnancies complicated by primary HCMV infection in the mother and subsequent transmission of the infection. In one case, the mother suffered from primary HCMV infection at 16 weeks of gestation. Two baby girls were vaginally delivered at 38 weeks of gestation. HCMV was isolated from urine collected 3 days after birth from one newborn, whereas the second newborn was found to be uninfected. Both newborns were asymptomatic. The placentas were dichorionic-diamniotic and separate. In the second case, the mother had primary HCMV infection at 10 weeks of gestation. The placenta was monochorionic-diamniotic. Both twins were found to excrete HCMV in urine at birth in the absence of signs or clinical symptoms (M. G. Revello and M. Zavattoni, unpublished data). The second observation is relevant to the clinical outcome of congenitally infected twins. While in the reported cases of congenitally infected monozygotic twins, both members were either severely affected or subclinically infected, the clinical outcome of dizygotic twins appeared more variable. In particular, in one set of congenitally infected twins who appeared asymptomatic at birth, one member was normal at follow-up, while the other had bilateral deafness, was restless, and had poor attention . More recently, the variable outcome of three infected survivors of a quadruplet pregnancy has been reported . One infant had cholestatic jaundice at birth and died of liver failure at 3 months of age, one infant showed no signs or symptoms at 18 months of age, and the remaining infant had hearing loss and delayed development. The observed variability in HCMV transmission and clinical outcome in infected twins clearly indicates that twin fetuses may react differently to maternal HCMV infection and that, as anticipated by Ahlfors in 1988 , the placenta seems to play a key role in the transmission of or protection from the infection. Moreover, the markedly different clinical outcome of congenital HCMV infection among infants of a multiple pregnancy suggests that genetic determinants might be involved. In conclusion, these data cast some doubts on the value of some maternal factors, such as humoral and cellular immune response, as possible prognostic markers of intrauterine transmission, as postulated by some researchers . They should also be taken into consideration when testing the hypothesis that reinfection with a different HCMV strain may be responsible for the unfavorable outcome observed in some infants born to mothers with existing immunity . Backmatter: PMID- 12364373 TI - Tularemia AB - Francisella tularensis is the etiological agent of tularemia, a serious and occasionally fatal disease of humans and animals. In humans, ulceroglandular tularemia is the most common form of the disease and is usually a consequence of a bite from an arthropod vector which has previously fed on an infected animal. The pneumonic form of the disease occurs rarely but is the likely form of the disease should this bacterium be used as a bioterrorism agent. The diagnosis of disease is not straightforward. F. tularensis is difficult to culture, and the handling of this bacterium poses a significant risk of infection to laboratory personnel. Enzyme-linked immunosorbent assay- and PCR-based methods have been used to detect bacteria in clinical samples, but these methods have not been adequately evaluated for the diagnosis of pneumonic tularemia. Little is known about the virulence mechanisms of F. tularensis, though there is a large body of evidence indicating that it is an intracellular pathogen, surviving mainly in macrophages. An unlicensed live attenuated vaccine is available, which does appear to offer protection against ulceroglandular and pneumonic tularemia. Although an improved vaccine against tularemia is highly desirable, attempts to devise such a vaccine have been limited by the inability to construct defined allelic replacement mutants and by the lack of information on the mechanisms of virulence of F. tularensis. In the absence of a licensed vaccine, aminoglycoside antibiotics play a key role in the prevention and treatment of tularemia. Keywords: INTRODUCTION : In 1911, an outbreak of a plague-like disease in rodents in Tulare County, California, provided the first isolation of a small gram-negative bacterium which was named Bacterium tularensis . Subsequently, natural infections with Francisella tularensis have been reported in a range of vertebrates including mammals, birds, amphibians, and fish, and even in invertebrates . Tularemia occurs only in the northern hemisphere and most frequently in Scandinavia, northern America, Japan, and Russia . However, tularemia has recently been reported from Turkey, Yugoslavia, Spain, Kosovo, and Switzerland , indicating that tularemia is even more widely distributed than was previously thought . In many parts of the world, colloquial names such as rabbit fever, hare fever, deerfly fever, and lemming fever have been used to describe the disease . Tularemia is not a disease which is notifiable to the World Health Organization, and the worldwide incidence of disease is not known. Nevertheless, some data on the incidence of disease are available. In Sweden, the annual number of reported cases of disease in humans over the period from 1973 to 1985 ranged from less than 5 cases to over 500 . In Japan 1,355 cases of tularemia were reported over the period 1924 to 1987 . In Turkey, 205 cases were reported over the period 1988 to 1998 , while in Slovakia 126 cases of disease were reported during the period 1985 to 1994 . In some parts of the world, the number of cases of tularemia has declined markedly during the twentieth century: in the United States, the annual number of reported cases of tularemia declined from several thousand in the 1930s to several hundred in the 1980s . It is likely that these figures are a gross underrepresentation of the true incidence of tularemia, because many infections are not diagnosed as a consequence of the relatively benign nature of disease caused by some strains of F. tularensis and by some difficulties in laboratory testing. F. tularensis has been an organism of concern as a biological threat agent since the large state-funded biological weapons programs of the 1950s, when the United States first evaluated the organism as a biological weapon, and it was subsequently incorporated into weapons by the U.S.S.R. (reviewed in reference ). Now that the emphasis has shifted towards defending against biological terrorism, public health and medical management protocols following a release of tularemia have been reviewed . FIG. 1. | World map showing areas where F. tularensis World map showing areas where F. tularensis is endemic (shaded). EPIDEMIOLOGY : F. tularensis is thought to be maintained in the environment principally by various terrestrial and aquatic mammals such as ground squirrels, rabbits, hares, voles, muskrats, water rats, and other rodents . In regions where tularemia is endemic, antibodies to F. tularensis are frequently detected in the sera of trapped wild animals . Outbreaks of disease in humans often parallel outbreaks of tularemia in wild animals. For example, in Sweden a clear correlation between peaks in vole and hare populations and outbreaks of tularemia in humans have been reported , and outbreaks of tularemia in humans in the former Soviet Union have been linked to epizootics of disease in ground voles . However, it is not clear whether these animal species are the true reservoir of the bacterium in the environment. A wide range of arthropod vectors have also been implicated in the transmission of tularemia between mammalian hosts. In central Europe, the ticks Dermacentor reticulatus and Ixodes ricinus are important vectors. In areas of the Czech Republic and Austria where natural foci of tularemia occur, between 2.1 and 2.8% of D. reticulatus ticks analyzed contained F. tularensis . In the United States, biting flies are the most common vectors in Utah, Nevada, and California , while ticks are the most important vectors east of the Rocky Mountains. In the former Soviet Union, the bacterium is transmitted by both mosquitoes (Aedes, Culex, and Anopheles species) and the Ixodes species of tick . Such arthropod vectors play a role both in the transmission of the disease within wild animal populations and in the transmission of disease to humans. It therefore follows that rural populations, especially individuals who spend time in endemic areas such as farmers, hunters, walkers, and forest workers, are most at risk of contracting tularemia . There is evidence that the bacterium can persist in watercourses, possibly in association with amoebae . Beavers and muskrats in North America and lemmings and beavers in Scandinavia might also play a role in the maintenance of the bacterium in watercourses . During a prolonged period in water (140 days in the study reported), the number of directly culturable bacterial cells declined to below detectable levels . However, on the basis of rhodamine 123 staining, which is indicative of metabolic activity, at least 30% of these cells were judged to be viable. Although this finding suggests that F. tularensis can persist in a viable but nonculturable form in water, the bacteria were not able to cause disease when injected into mice . These findings might have implications for techniques which rely on the direct culture of F. tularensis to identify bacteria in watercourses. TAXONOMY OF F. TULARENSIS : The taxonomic position of F. tularensis is complex and has changed frequently. F. tularensis was originally included in the genus Bacterium, later in the Pasteurella genus, and subsequently provisionally placed in the Brucella genus . In 1947, the proposal was made that the bacterium should be the sole member of a new genus called Francisella . The most recent issue of Bergey's Manual of Systematic Bacteriology indicates that F. tularensis and F. novicida are the two species in the genus Francisella. F. novicida has been differentiated on the basis of inability to produce acid from sucrose, the relative ease of culture, and the lack of virulence for humans and rabbits . However, several workers have since questioned whether F. novicida should be considered a separate species. F. novicida is now known to be capable of causing a tularemia-like illness in humans, and on the basis of DNA hybridization, F. novicida was not distinguishable from F. tularensis . More recently, the 16S ribosomal DNA (rDNA) sequences of F. tularensis and F. novicida have been shown to have a high degree of sequence similarity (99.6%) . On the basis of these studies, both Hollis et al. and Forsman et al. proposed that F. novicida should be considered a subspecies of F. tularensis. More recently, a detailed analysis of several strains of Yersinia philomiragia revealed their fatty acid compositions to be typical of the genus Francisella . In addition, the DNA from these strains showed a high degree of relatedness to F. tularensis DNA . These observations indicated that Y. philomiragia should be included in the Francisella genus and renamed Francisellaphilomiragia . This proposal is supported by the finding that the 16S rDNA sequences of F. tularensis and F. philomiragia confirm both its placement in the Francisella genus and its identity as a separate species . F. philomiragia strains can also be distinguished from F. tularensis because the former are oxidase positive in the Kovacs modification of this test, are often able to hydrolyze gelatin, and are relatively easily cultured. In addition, F. philomiragia is considered virulent only in immunocompromised individuals or in those who have recently had a near-drowning experience . Therefore, the current consensus position appears to be that F. tularensis and F. philomiragia are the sole members of the Francisella genus within the Francisellaceae family. The 16S rDNA sequences of the members of this family contain the consensus sequence for members of the gamma-subclass of Proteobacteria . This family also includes Wolbachia persica , a parasite of arthropods and worms, the Dermacentor andersoni symbiont, which was originally isolated from Rocky Mountain wood ticks , and Ornithodorous moubata symbiont B . Most of these arthropod endosymbionts are poorly characterized. However, W. persica is known to exist in the Malpighian tubules of the soft tick Argas arboreus and is not considered a pathogen of mammals. For F. tularensis, several subspecies other than novicida have been proposed . Originally, F. tularensis strains were identified as belonging to either subsp. tularensis (also known as type A or subspecies nearctica) or subspecies palaearctica (also know as type B or subspecies holarctica) principally on the basis of virulence, citrulline ureidase activity (conversion of l-citrulline to ornithine), and acid production from glycerol . Subspecies holarctica is now the most widely used terminology in place of subspecies palaearctica . A more detailed description of the properties of subsp. tularensis and subspecies holarctica can be found in Olsufjev and Mesheryakova . Three biovars of F. tularensis subspecies holarctica have been suggested ; biovar I (erythromycin sensitive) , biovar II (erythromycin resistant) , and biovar japonica . Strains of F. tularensis subsp. tularensis are considered the most virulent for humans, with an infectious dose of less than 10 CFU and a mortality of 5 to 6% in untreated cases of cutaneous disease . F. tularensis strain Schu S4 is the proposed subsp. tularensis type strain . Strains of subspecies holarctica do not cause disease in rabbits, and the mortality rate associated with cutaneous disease in humans is less than 0.5% . Until recently, F. tularensis subsp. tularensis was thought to be found only in North America. However, the recent isolation of this subspecies in Europe suggests that it may be more widely distributed than was previously thought . F. tularensis subspecies holarctica is found mainly in North America and in Eurasia . A fourth subspecies, F. tularensis subsp. mediaasiatica, is found predominantly in the central Asian republics of the former USSR . These strains possess citrulline ureidase activity and are able to ferment glycerol, but are less virulent than strains of F. tularensis subsp. tularensis in rabbits . FIG. 2. | Evolutionary distance tree, based on 16S rDNA sequences, showing the relationship of F. tularensis Evolutionary distance tree, based on 16S rDNA sequences, showing the relationship of F. tularensis with other putative members of the Francisellaceae and with other closely related members of the gamma subclass of the Proteobacteria. Agrobacterium tumefaciens was included as an outgroup. Reproduced from Forsman et al. with the kind permission of the International Union of Microbiological Societies. TABLE 1 | Taxonomy and characteristics of the subspecies of F. tularensis GENOMIC ANALYSIS OF F. TULARENSIS : The genetic makeup of F. tularensis is poorly understood, with sequences of only 34 genes deposited with GenBank. Two different cryptic plasmids (pOM1 and pNFL10) are reportedly found in the live vaccine strain (LVS) and in F. novicida, respectively (GenBank accession AFO55345). This paucity of information has prompted two projects to sequence the F. tularensis genome. The genome sequence of a strain of F. tularensis subsp. tularensis (strain Schu S4) is currently being determined by a consortium of laboratories in Europe and the United States . A project to sequence the genome of LVS, originally derived from a virulent F. tularensis subsp. holarctica strain, has recently commenced in the United States. Preliminary results from the strain Schu S4 genome sequencing project suggest a total genome size of <2 Mbp, making this one of the smaller bacterial genomes . Strain Schu S4 does not appear to possess either plasmid pOM1 or pNFL10 . The G+C content of the genome is approximately 34% . A preliminary annotation has identified 1,804 candidate open reading frames which are characterized by a slight shift in G+C content. Of these, 1,289 are thought to encode proteins, of which 413 had no database match , meaning that in comparison with many other bacterial genomes, F. tularensis contains a high proportion of unique genes . On the basis of database matches with known proteins, genes encoding putative transport/binding, gene regulation, energy metabolism, and cellular processes appeared to be underrepresented in the F. tularensis genome . FIG. 3. | Preliminary annotation of the F. tularensis Preliminary annotation of the F. tularensis strain Schu S4 genome sequence; comparison of the number of F. tularensis genes in 15 functional categories (solid bars). The mean number of genes in these categories in 20 other bacterial species is shown as open bars, with error bars indicating the lowest and highest numbers of genes in each category. Reproduced from Prior et al. with the kind permission of Blackwell Scientific Limited. DETECTION AND DIAGNOSIS : F. tularensis is a fastidious organism which requires enriched medium for growth. Traditionally, cysteine glucose blood agar has been the growth medium of choice. However, enriched chocolate agar (cysteine heart agar supplemented with 9% heated sheep red blood cells [CHAB]) and nonselective buffered charcoal yeast extract agar also support the growth of the organism and may be used for isolation . The Centers for Disease Control and Prevention guidelines recommend the use of CHAB once growth on the general microbiological agars, such as sheep blood agar, chocolate agar, and Thayer-Martin agar, which are routinely used in U.S. laboratories for the isolation of bacteria from clinical specimens, indicates the pathogen to be present . A heavy inoculum on appropriate medium will yield visible growth in 18 h, but the appearance of individual colonies may require 2 to 4 days of incubation . F. tularensis grows slowly at 37C and poorly at 28C, and this can be exploited to distinguish F. tularensis from Yersinia pestis, F. philomiragia, and F. tularensis subsp. novicida, all of which grow well at 28C . On CHAB, colonies are 2 to 4 mm in size, greenish-white, round, smooth, and slightly mucoid, while on media containing whole blood there is usually a small zone of alpha-hemolysis surrounding colonies . Gram staining of cultured material reveals small (0.2 to 0.5 mum by 0.7 to 1.0 mum), mainly single gram-negative coccobacilli which stain weakly. Chemically defined media capable of supporting the growth of F. tularensis do exist, but growth on such media is slower and the colonies are smaller than on traditional rich agars. F. tularensis does not grow well in liquid media even when the medium is supplemented with cysteine and requires a large inoculum to obtain visible growth within 24 h. Clinical laboratories routinely inoculate clinical specimens into broth, usually brain heart infusion or Trypticase soy broth, for the recovery of bacteria. However, F. tularensis requires enriched media for growth, supplemented with cysteine. The organism has been cultivated in modified Mueller-Hinton broth and thioglycollate broth. In Mueller-Hinton broth, the addition of 0.025% ferric pyrophosphate appeared to enhance the growth of F. tularensis. Growth in liquid media is slow, requiring up 3 to 7 days of incubation if the broth is shaken or a minimum of 10 days in unshaken broth to produce visible growth. In static thioglycollate broth, growth is seen first as a dense band near the top which diffuses throughout the broth as growth progresses . The most widely used synthetic medium used for the growth of tularemia was devised by Chamberlain . However, defined media are not used for routine diagnosis of tularemia. Direct isolation of bacteria is frequently achieved from ulcer scrapings, lymph node biopsies, or sputum . The organism is rarely cultured directly from blood, although this is becoming feasible with the development of sensitive blood culture systems . The isolation of bacteria from urine or feces is not frequently done , but antigen detection in urine and RNA hybridization of wound specimens have been reported . In stained tissue sections, bacteria may be found both intra- and extracellularly . Enzyme-linked immunosorbent assay (ELISA)-based tests can also be used to detect the bacteria in clinical samples. For example, a capture ELISA using monoclonal antibodies against F. tularensis lipopolysaccharide (LPS) recognized all strains of F. tularensis tested other than those of subspecies novicida, with no cross-reactivity with other bacterial species tested. The sensitivity was 103 CFU/ml in phosphate-buffered saline (PBS) and 104 CFU/ml in spiked human sera . An immunochromatographic hand-held assay has also been developed based on a polyclonal and a monoclonal antibody to LPS of F. tularensis LVS; the detection limits of this assay were 106 CFU/ml PBS and 106 to 107 CFU/ml in spiked human sera. This assay is principally designed for field use, being compact and easy to use and giving results in 15 min, but the relatively low sensitivity means that a negative result does not exclude tularemia . Because of the difficulty in culturing F. tularensis, most cases of tularemia are diagnosed on the basis of clinical picture and/or serology . Serological tests for the diagnosis of F. tularensis infection are attractive because diagnostic work involving culture procedures carries a risk of infection for nonvaccinated laboratory staff . The diagnosis of human cases of tularemia is usually confirmed by the demonstration of an antibody response to F. tularensis, which occurs about 2 weeks after the onset of the disease . The detection of serum antibodies is most frequently achieved by agglutination or an ELISA . A latex agglutination test commercially available from BBL (Becton Dickinson, Franklin Lakes, N.J.) has been used by some workers to identify individuals with antibodies to F. tularensis: reactions at dilutions greater than 1:20 are considered specific and significant. Using this test, 2% of trappers in Quebec were found to be seropositive (at serum dilutions of 1:20 to 1:2,048) for F. tularensis . The only association between this population and the development of antibodies appeared to be the trapping of muskrats. Commercially available antigens can also be used with standard tube agglutination tests. A fourfold increase during illness or a single titer of 1:160 or greater is considered diagnostic . However, while serological tests are frequently used for diagnosis, strains of F. tularensis have occasionally been isolated which fail to agglutinate commercially available F. tularensis antigens . A range of PCR-based assays have been reported for the detection of F. tularensis or for the diagnosis of tularemia. These PCR assays have all used primers directed against genes encoding outer membrane proteins such as fopA or the 17-kDa outer membrane lipoprotein . These PCR assays offer high specificity, failing to generate an amplicon using template DNA from a range of other bacterial pathogens . When used with pure cultures of F. tularensis as the source of template DNA, the assays also offer high levels of sensitivity. The PCR assay reported by Grunow et al. allowed the detection of 102 CFU/ml in PBS , and the use of a nested PCR assay allowed the detection of 1 CFU of F. tularensis . However, samples such as blood contain compounds capable of inhibiting the PCR, making detection of such low levels of bacteria in clinical samples impossible. In addition, blood samples invariably require some form of processing to allow high-sensitivity PCR tests to be carried out . Nevertheless, the PCR assay reported by Grunow et al. was shown to detect 103 to 104 CFU/ml of spiked serum , while the nested PCR assay reported by Fulop et al. was able to detect 102 CFU/ml of spiked blood . Even in their current forms, these assays do appear to offer advantages over the direct culture of bacteria. In mice, 83% of blood samples taken 24 h after experimental infection were positive by PCR, while bacteria could be cultured from only 48% of the samples . This advantage appears to extend to clinical samples: in a study with swab samples taken from the lesion site in 40 human cases of ulceroglandular tularemia, a PCR assay was positive for 73% of the samples, whereas bacteria were cultured from only 25% of the samples . Moreover, PCR identified F. tularensis in one sample which had not been identified by culture or serology in a parallel study . Subsequently, a number of other workers have reported the use of PCR assays for the diagnosis of tularemia or for the analysis of environmental samples . These PCR-based tests might also be safer than tests which involve the culture of bacteria . Improved methods for the isolation and processing of tissue or blood samples and for the transport of samples to the laboratory might further improve the sensitivity and utility of the PCR test . Indeed, a specially formulated filter paper designed for sample collection has been proposed for the rapid preparation of template DNA from clinical or field-collected tick vectors . When linked with a TaqMan 5' nuclease assay or a PCR-enzyme immunoassay, sensitivities of <100 CFU were reported . While PCR assays have allowed the rapid detection of a range of F. tularensis strains, other workers have developed molecular methods to discriminate between Francisella strains . PCR-based methods have been evaluated for their potential to identify F. tularensis and discriminate between the different subspecies. Long random sequence oligonucleotide primers and primers specific for repetitive extragenic palindromic (REP) sequences and enterobacterial repetitive intragenic consensus (ERIC) sequences were evaluated . REP-PCR has been applied to specifically identifying strains of F. tularensis subsp. novicida, but patterns from subspecies holarctica and tularensis were found to be similar . A one-base difference in the 16S rRNA sequences of F. tularensis subsp. tularensis and F. tularensis subsp. holarctica has been demonstrated, and on this basis a PCR has been developed which can differentiate the two subspecies . In one study, PCR analyses based on the use of ERIC, REP or long random sequence primers yielded reproducible banding patterns of similar complexity and allowed differentiation of strains at the subspecies level, but the methods investigated do not meet the criteria for typing of individual isolates . More recent studies suggest that PCR amplification of tandem repeat regions may be highly discriminatory and a useful tool in strain typing . A specific PCR has been developed which produces amplicons of different lengths (target unknown), which has been used in combination with the 17-kDa lipoprotein PCR and can distinguish F. tularensis subsp. holarctica from strains of other F. tularensis subspecies . CLINICAL DISEASE : Tularemia in humans can occur in several forms, depending to a large extent on the route of entry of the bacterium into the body. Although tularemia can be a severely debilitating disease, especially when caused by F. tularensis subsp. tularensis, many cases of disease caused by lower-virulence strains are undiagnosed. The most common form of the disease is ulceroglandular tularemia, which usually occurs as a consequence of a bite from an arthropod vector which has previously fed on an infected animal . Some cases of ulceroglandular tularemia occur in hunters and trappers as a consequence of the handling of infected meat, with infection via cuts or abrasions. After an incubation period of typically 3 to 6 days , the patient experiences the sudden onset of flu-like symptoms, especially chills, fever, headache, and generalized aches . An ulcer, which can persist for several months, forms at the site of infection, usually on the lower limbs in the cases of tick-borne disease (a disease with similar symptoms but without the appearance of an ulcer is termed glandular tularemia). Bacteria are disseminated from this site via the lymphatic system to regional lymph nodes. The enlargement of these lymph nodes often resembles the classical bubo associated with bubonic plague. From this site, bacteria may be disseminated to other tissues such as the spleen, liver, lungs, kidneys, intestine, central nervous system, and skeletal muscles. However, the bacteremic phase of the infection is transient and occurs relatively early in the infection process. Recovery from the disease can be protracted, but the ulceroglandular form of tularemia is rarely fatal (the mortality rate from ulceroglandular tularemia is typically less than 3% ). Even without treatment, ulceroglandular and glandular tularemia are rarely fatal, but may take a significant length of time to resolve. On the other hand, an acute form of the disease produced by F. tularensis subsp. tularensis, typhoidal tularemia, is typified as a septicemia without lymphadenopathy or the appearance of an ulcer and carries a mortality rate of 30 to 60% . In addition to the symptoms described above, the patient may be delirious and shock may develop. A rare variation of ulceroglandular disease is oculoglandular tularemia, where the conjunctiva is the initial site of infection, usually as a result of the transfer of bacteria on the fingertips . The disease is marked by the appearance of ulcers and nodules on the conjunctiva, and without treatment the infection spreads to the local lymph nodes The ingestion of infected foodstuffs or of bacteria in drinking water can result in oropharyngeal or gastrointestinal tularemia, depending on the site of colonization of host tissues. The former is often described as a painful sore throat with enlargement of the tonsils and the formation of a yellow-white pseudomembrane. This is most often accompanied by swollen cervical lymph nodes . Depending on the infecting dose, gastrointestinal tularemia ranges from mild but persistent diarrhea to an acute fatal disease with extensive ulceration of the bowel. Probably the most acute form is disease is associated with the inhalation of bacteria, although pneumonic disease can also occur as a complication of ulceroglandular, glandular, oculoglandular, or oropharyngeal tularemia . The clinical and roentgenographic features of pneumonic tularemia are quite variable, making diagnosis difficult . On occasion, pneumonic tularemia may even occur without any overt signs of pneumonia. The occasional naturally occurring cases of inhalation tularemia often arise from farming activities which involve the handling (and the subsequent generation of dusts) of hay which has previously been the site of residence of infected rodents. . FIG. 4. | Enlarged lymph node in a tularemia patient. Enlarged lymph node in a tularemia patient. ANIMAL MODELS : The pathogenesis of the LVS strain in mice has been used as a model to study the behavior of intracellular pathogens. A study published in 1946 compared the sensitivities of various animal models to strain Schu S4. It was shown that Schu S4 was fully virulent in the mouse, guinea pig, rabbit, hamster, and cotton rat. The last was shown to have a high degree of individual variation in response to Schu S4, and accordingly this model has received little attention. Strains from all of the subspecies of F. tularensis are reported to be virulent in the murine and guinea pig models of disease, but only strains of F. tularensis subsp. tularensis are considered virulent in the rabbit model of disease . Most studies on the behavior of F. tularensis in vivo rely on the infection of inbred mice, and this murine tularemia system has become a model to study the behavior of intracellular pathogens. There is some evidence that strains with differing virulence can be differentiated by using this model. For example, strain Schu S4 is fully virulent regardless of the route of administration. However, the virulence of LVS has been shown to be dependent on the route of delivery; LVS is fully virulent in the mouse, with a median lethal dose of <10 CFU when delivered intraperitoneally, but is attenuated in the mouse when delivered intradermally . TABLE 2 | Virulence of F. tularensis biotypes PATHOGENESIS AND HOST RESPONSES TO INFECTION : In typical human ulceroglandular tularemia, a skin lesion first appears at the site of infection 3 to 5 days after infective exposure . During this initial phase, T cells appear to play little role in combating infection. The systemic administration of tumor necrosis factor or gamma interferon (IFN-gamma) is capable of reducing the severity of tularemia , indicating the key role these cytokines play in response to infection . During the early stages of disease, studies with other pathogens have suggested that keratinocytes are a likely source of tumor necrosis factor, while natural killer (NK) cells may produce IFN-gamma . A transient bacteremia occurs, during which the pathogen must resist lysis by complement. Protection appears to be due to the presence of capsule, as a nonencapsulated mutant was susceptible to killing by nonimmune serum . The bacteremic phase allows the organism to be seeded throughout the body, infecting all reticuloendothelial tissues. Within days, a wide range of cytokines are expressed in the reticuloendothelial tissues. In the liver, for example, the bacteria are able to invade and multiply within hepatocytes , and tumor necrosis factor alpha, interleukin-10, interleukin-12, and IFN-gamma are produced within 48 h . However, although this cytokine response, assumed to evoke the expansion of a protective Th1 response, was capable of mounting a defense against the LVS strain, it was unable to prevent a more virulent strain from mounting a lethal infection in mice . Infection also induces a stress response in infected host cells; hsp72 was induced in peritoneal exudate cells of infected mice which were susceptible to tularemia for 3 days following infection . Later during infection, T cells appear to play a major role in protection. Mice depleted of CD4+ and/or CD8+ T cells were capable of controlling a primary LVS infection, but were not able to resolve it , indicating a scenario of complex interactions of different cell types in primary infection. Studies with knockout mice infected with strain LVS confirm these observations, but CD4-, beta2-microglobulin-deficient/CD8-, and gammadelta T-cell receptor-negative mice were all able to resolve an infection . However, alphabeta T-cell receptor-negative knockout mice succumbed to infection, indicating that while either CD4 or CD8 T cells are individually sufficient to resolve infection with strain LVS, alphabeta T-cell receptor cells are required for protection. In mice challenged with a high-virulence strain (Schu S4), both CD4 and CD8 T cells appeared to play key roles in controlling disease . Circulating gammadelta T cells are well known for controlling bacterial intracellular infections; for example, they are essential in controlling Listeria monocytogenes infections . A significant expansion in the Vgamma9Vdelta2 T-cell subpopulation was observed in a Japanese patient with tularemia . This effect was subsequently shown to occur within a week of infection, but could persist for over a year . This response appears to be due to exposure to phosphoantigens expressed by F. tularensis in vivo . Phosphoantigens are powerful stimuli for Vgamma9Vdelta2 cells, as has been shown for Mycobacterium tuberculosis and Plasmodium falciparum . Interestingly, vaccination with strain LVS did not result in an expansion of the Vgamma9Vdelta2 T cells similar to that observed in tularemia patients. Neutrophils also appear to play a role in defense by ingesting and killing microorganisms, lysing infected hepatocytes and acting as a source of cytokines . Different strains of mice show different degrees of susceptibility to infection, indicating that resistance involves multiple genetic loci . One locus which appears to play a role in the natural resistance to primary infection is Bcg (Nramp1) . Expression of this allele has many pleiotropic effects associated with activation of macrophages by IFN-gamma or LPS, and mutation of the allele can confer susceptibility to infection by a range of intracellular pathogens , including F. tularensis . The pathogenicity of intracellular bacteria depends on their ability to survive within macrophages, although other cell types such as hepatocytes may play an important but as yet poorly defined role . As such, most of the work on the cellular interaction of F. tularensis has concentrated on the macrophage. INTERACTION WITH HOST CELLS : Although a facultative intracellular organism in vitro, F. tularensis has been described as an "obligate intracellular pathogen of macrophages in vivo" . Early work showed the organism multiplied in murine macrophages , guinea pig hepatic cells, endothelium , and gut endothelial cells isolated from ticks . In various artificial in vitro culture systems, such as chick embryos , HeLa cells , and mouse fibroblasts , intracellular multiplication has also been demonstrated. That the organism is an intracellular pathogen in vivo as well as capable of intracellular growth in vitro is supported by the similarity of the histopathology of tularemia to that seen in infections produced other intracellular pathogens such as M. tuberculosis (reviewed in reference ). In addition, resistance to highly virulent strains of F. tularensis is dependent on lymphoid tissue rather than serum , further supporting the case that F. tularensis is an intracellular pathogen. F. tularensis enters macrophages using a cytochalasin B-insensitive pathway without triggering the respiratory burst . However, opsonized F. tularensis has been shown to be actively phagocytosed by polymorphonuclear leukocytes, which are capable of killing the bacteria by oxidative killing mechanisms . A protein, AcpA, has been identified in F. tularensis which has an acid phosphatase function . AcpA is capable of inhibiting the respiratory burst more efficiently than previously described acid phosphatases of other intracellular pathogens, such as Leishmania donovani and Legionella micdadei . Bacteria live within the macrophage in a phagosome which does not fuse with lysosomes , but acidification of the phagosome does occur and is essential for growth of F. tularensis and acquisition of iron . In some animals, nitric oxide (NO) production is produced by macrophages to limit infections by intracellular bacterial pathogens (reviewed in reference ). NO production by the infected host appears to have a nonspecific role in protection against F. tularensis infection . Phase variation of F. tularensis LPS has been observed, and different forms of the LPS appeared to affect NO induction, thus modulating the innate immune response . Unlike peritoneal macrophages , NO has been shown not to be involved in killing of F. tularensis by alveolar macrophages , which can behave differently from other resident macrophage populations (reviewed in reference ). Alveolar macrophages activated by IFN-gamma were able to kill F. tularensis, and this killing was resistant to inhibitors of NO production, although the inhibition of growth correlated with nitrite production by the cell. Cytokines known to regulate the effector functions of activated macrophages (tumor necrosis factor alpha, interleukin-10, transforming growth factor beta 1, and IFN-alpha) also did not affect the IFN-gamma-induced killing by alveolar macrophages , indicating that the IFN-gamma-induced responses of alveolar and peritoneal macrophages are fundamentally different. Bacteria must modify gene expression to survive in hostile environments. | F. tularensis was shown to upregulate the expression of four proteins during growth in a macrophage cell line . Compared to the stress response induced in other intracellular pathogens such as Salmonella enterica serovar Typhimurium , Legionella pneumophila , and M. tuberculosis , F. tularensis demonstrates a very low level of response on encountering an adverse environment. Although DnaK, GroEL, and GroES have all been shown to be upregulated in vitro in response to exposure to heat and hydrogen peroxide stress , these were not upregulated in the macrophage . The four proteins which showed increased levels of expression relative to broth-grown cells had molecular masses of 20, 23, 55, and 70 kDa. Subsequent separation of protein extracts by two-dimensional gels revealed the 23-kDa protein to have a pI of 5.8, but the other three proteins of interest appeared to have pI values outside the range of 4 to7 and thus were not visible on the gels . The 23-kDa protein had no amino acid sequence homology with any known protein, and its function has not been elucidated to date, although it appears to play a role in response to stress, as it is upregulated under conditions of oxidative stress but not heat shock . One operon shown to be essential for growth of F. novicida inside macrophages is mglAB , which is also present in F. tularensis . MglAB shows high similarity to the SspAB of Escherichia coli and is therefore likely to be a transcriptional regulator. In E. coli, SspAB regulates the expression of a range of proteins in response to nutritional stress, and in F. novicida inactivation of the operon resulted in a change in expression of a number of proteins; most notably, four proteins of 33, 38, 20, and 70 kDa were absent in the mutant compared to the parent strain . Phenotypic assays also indicated that MglAB may regulate expression of at least one exported phosphatase, a group of proteins which includes AcpA. Another locus described as necessary for survival in macrophages is minD . MinD is a 29-kDa protein and thus is not one of those described above that are upregulated intracellularly. Two roles for MinD have been proposed based on possible actions of the first-described minD in E. coli. Initially, the E. coli MinD was suggested to have a role in septum formation during cell division, but subsequently homology with a heavy-metal ion pump indicated an alternative role. Thus, Anthony et al. proposed that the MinD in F. tularensis may be essential for intracellular growth either because it acts as a pump for toxic or radical ions or because abnormal septum formation in the minD mutant results in loss of cell wall integrity, allowing bactericidal agents into the cell . After multiplication within the macrophage, F. tularensis induces cell death by apoptosis . This releases the bacteria from the cell, allowing infection of fresh cells. Interestingly, relatively large numbers of bacteria were required to be present within the macrophage before apoptosis could occur and required a longer time from infection to induction of apoptosis compared to Salmonella, Shigella, Yersinia, and Legionella species. This probably reflects the slow growth rate and the obligate intracellular lifestyle of F. tularensis in vivo. VIRULENCE DETERMINANTS : Few virulence factors have been identified for F. tularensis. However, studies with F. novicida and the availability of genome sequence data for F. tularensis may help to dissect the pathogenic mechanisms of this enigmatic organism. One of the problems in analyzing the contribution of a specific gene to virulence is that isogenic mutants in F. tularensis have not yet been produced, although methods have been described for producing mutants in F. novicida . Transposon mutagenesis has been employed successfully and has been used to identify the genes mentioned above involved in macrophage survival and growth, but so far this technology has not been used to specifically target virulence factors. The capsule, although essential for serum resistance , is not required for survival following phagocytosis by polymorphonuclear leukocytes . Noncapsular mutants possess higher neuraminidase activity than capsular wild-type strains . The reason for this is not known, but the authors proposed a role for neuraminidase in colonization, as the enzyme was active in degrading natural mucins but not glycoproteins. The LPS of F. tularensis does not exhibit the properties of a classical endotoxin. It fails to induce interleukin-1 from mononuclear cells and only poorly induces tumor necrosis factor and NO production from macrophages . The inability of F. tularensis LPS to antagonize a range of endotoxin-induced cellular responses seen with most LPS molecules indicates that F. tularensis LPS does not interact with LPS receptors . LPS from F. tularensis has been shown to undergo phase variation, which affects both antigenicity due to variations in the O antigen and the NO response of macrophages due to variation in the lipid A moiety. The phase variation of the lipid A has been demonstrated to affect the organism's ability to grow intracellularly. In one phase, reduced NO induction results in bacterial growth, while in another phase, increased NO production suppresses growth . This growth restriction was observed only in rat macrophages and not in mouse macrophages . Studies on F. novicida have also shown a role for LPS in macrophage growth. The valA gene encoded an ABC transporter possibly required for transport of LPS to the outer membrane , and mutants defective in this gene were unable to grow in macrophages and showed an increased susceptibility to serum killing . As can be seen, very few classical virulence factors have been identified for this pathogen. Some of these, such as secreted toxins, are not produced by the organism , while others await discovery. For example, tularemia can be contracted by drinking contaminated water , but how the organisms invade from the gut to produce infection is not known and no invasin has been identified for this pathogen. Probing the genome sequence data will help to identify putative virulence factors such as adhesins, but the inability to create specific allelic replacement mutants will delay research into their role until this problem is overcome. DEVELOPMENT OF LIVE TULAREMIA VACCINES : Initial efforts to develop a live attenuated tularemia vaccine began in the former Soviet Union prior to the Second World War. Attenuation of strains was achieved either by repeatedly subculturing fully virulent strains in media supplemented with antiserum or by drying the strains . In 1934, El'bert et al. inoculated animals with a weakly virulent tularemia culture. Protection was demonstrated when these immunized animals were challenged with a virulent culture, and it was suggested that the same immunization procedure might be applicable to humans . Strain Moscow was reported to show weakened virulence and high immunogenicity and was used as a live vaccine in humans in 1942. The effectiveness of vaccination was successfully demonstrated in volunteers, and several thousand individuals were reportedly vaccinated before the strain was apparently lost . Strain 15 was later identified and shown to have reduced virulence in guinea pigs while retaining virulence for mice. These strains were administered subcutaneously into humans in clinical trials, and it was concluded at this time that vaccination of humans with attenuated F. tularensis strains was harmless . In subsequent years, a number of mass vaccination programs using strain 15 were carried out in areas of the former Soviet Union where tularemia outbreaks were prevalent. As many as 60 million individuals were immunized with live vaccine preparations in the former Soviet Union until 1960 . In subsequent years it was shown that strain 15 had become so attenuated that it was no longer virulent in mice . The strain was passaged in animals and a variant, strain 15 restored, was derived. Another attenuated vaccine, strain 155, was also developed at this time, and both strains were produced as live vaccines at the Gamaleya Institute in Moscow. These strains were transferred to the United States in 1956 . However, cultures grown from reconstituted ampoules showed that both vaccine strains segregated into the two colony types, designated blue colony variant or grey colony variant depending on their appearance when viewed microscopically under oblique light. The blue colony variant was shown to be more virulent and immunogenic in small animals than the grey colony variant . Mice immunized with the blue colony variant vaccine were protected against subsequent challenge with the fully virulent strain Schu S4. Guinea pigs immunized with the blue colony variant showed increased resistance to challenge. Lyophilized preparations of the blue colony variant were prepared, and a live vaccine strain (LVS) was derived after five passages through mice. The LVS appeared to be an effective vaccine, protecting immunized mice and guinea pigs against an inhalation challenge with F. tularensis strain Schu S4 . These studies were sufficiently encouraging to warrant an extension of studies to humans. It was shown that clinical tularemia could be induced in nonvaccinated individuals by inhalation of approximately 10 to 50 CFU . Treatment of infected volunteers occurred at the earliest indication of systemic disease, employed either streptomycin or tetracycline, and resulted in the complete recovery of infected individuals. In a subsequent study, 18 volunteers were vaccinated with LVS prior to an inhalation challenge with strain Schu S4. Whereas 8 of 10 controls showed evidence of infection, only 3 of 18 LVS-immunized individuals showed evidence of infection . It was concluded from this study that immunization with LVS induced significant protection against respiratory challenge with F. tularensis . A subsequent investigation involved volunteers who had been immunized with LVS approximately 1 year before challenge by the aerosol route with strain Schu S4. The challenge dose ranged from 200 to 20,000 CFU. Results showed that the majority of vaccinees challenged with up to 2,000 organisms escaped clinical illness. Immunized volunteers challenged with 20,000 organisms showed modified disease symptoms compared to nonimmunized volunteers infected with a similar dose of Schu S4 . Several routes of immunization with F. tularensis LVS have been evaluated over the past 40 years. Airborne administration of LVS was investigated in nonhuman primates and guinea pigs in order to enhance the immunity provided by the live tularemia vaccine and was sufficiently encouraging for aerogenic vaccination of humans to be initiated. Initial studies demonstrated that aerogenic immunization of animals gave at least comparable protection against tularemia infection as dermal immunization. When these studies were repeated in humans, there was an apparent greater level of protection after respiratory immunization compared to the conventional intradermal method of administration . Similarly, oral administration of high doses of F. tularensis LVS was reported to induce protection against aerosol challenge . In the United States, the initial batches of the LVS vaccine were produced by the National Drug Company in 1959. In 1977, a retrospective study was published that showed the effectiveness of the LVS vaccine in the prevention of laboratory-acquired tularemia . The period from 1950 to 1959, prior to LVS vaccination, was compared with the period 1960 to 1969, during which LVS vaccination was used routinely. Figures showed that the incidence of typhoidal tularemia fell from 5.7 to 0.27 cases per 1,000 at-risk employees. The incidence of ulceroglandular tularemia remained unchanged in the two periods, but the clinical signs and symptoms of this form of the disease were moderated in vaccinated individuals. An application to license the LVS vaccine was submitted to the Federal Drug Administration. In this instance, LVS was derived from NDBR 101, lot 9, and grown under fermentation conditions. The new vaccine contained fewer of the immunogenic blue colony variant types than the parent strain and also had a higher residual moisture content . The immunogenicity of the new vaccine lot was evaluated in 19 human volunteers after administration by scarification . Tests were conducted to evaluate humoral and cell-mediated immune responses in volunteers between 7 and 63 days after immunization. Immune responses were assayed by ELISA and lymphocyte proliferation techniques. Results showed that by day 63, a positive IgG, IgA, and IgM response towards an ether-extracted antigen was evident in 100% of volunteers. Lymphocyte proliferation assays using the same antigen demonstrated a positive response in 40% of volunteers 7 days after immunization, and 80% of volunteers gave a positive result by day 63. In the early 1960s, the vaccine was approved by the Food and Drug Administration only for use in clinical trials under investigational new drug status. However, some properties of the LVS vaccine may give cause for concern and make licensing difficult. For example, the protective response induced by the vaccine has not been characterized. In addition, the basis of attenuation of the LVS strain is not known, and studies to examine the virulence of LVS in the mouse model have shown that the strain is fully virulent when delivered intraperitoneally, with a median lethal dose of less than 10 CFU. However, when delivered intradermally, the strain is avirulent in the mouse. The median lethal dose for LVS after intravenous and subcutaneous delivery is reported to be approximately 103 and 105 CFU, respectively. However, these results were derived after repeated passage through mice to increase the virulence of the strain. Studies in mice with ampoule-derived LVS that has not been animal passaged have shown the median lethal doses to be 105 CFU after intravenous administration and 107 CFU when delivered subcutaneously. The LVS vaccine remains the only effective vaccine against tularemia developed to date. However, this vaccine is not currently available, though work to fully license this vaccine is under way in the United States. The finding that an attenuated mutant of F. tularensis can induce protective immunity suggests that this approach to vaccine development is feasible. In a range of other pathogens, the introduction of defined mutations into genes required for growth of the pathogen in vivo has yielded safe and effective vaccines. The construction of a defined attenuated mutant of F. tularensis could provide a safe, effective, and licensable tularemia vaccine. The aromatic amino acid and purine biosynthesis pathways have already been identified from genome sequence information as targets for the construction of a defined attenuated mutant . However, the utility of this approach is limited because, as outlined in a previous section of this review, work to date has failed to devise methods for the construction of allelic replacement mutants of F. tularensis. DEVELOPMENT OF NONLIVING TULAREMIA VACCINES : Prior to the development of the LVS vaccine, immunologically based therapies against tularemia were reported in the 1930s by Lee Foshay, who suggested that immune serum could be administered to favorably modify the clinical course of tularemia in humans. This finding stimulated Foshay to work to develop a killed tularemia vaccine that induced humoral immunity . A number of techniques were employed to prepare the killed bacterial cells, including heating, acetone, and phenol treatment, and the Foshay vaccines were administered to human volunteers with variable results. There were reports of lesions at the site of administration and that the killed vaccine caused severe local reactions when administered to immune humans. These reactions were reduced somewhat if the vaccine was prepared with acetone rather than phenol killing of bacterial cells. Subsequently, vaccination was preceded by a simple skin test to determine whether immunity was present, and the vaccine dose regimen was modified accordingly to limit severe reactions . The efficacy of such killed whole-cell vaccines appears to be questionable. The phenol-killed vaccine was able to protect nonhuman primates against very low systemic challenges with F. tularensis strain Schu S4, although immunized nonhuman primates were as readily infected as controls . When tested in mice, the Foshay vaccine afforded a similarly low level of protection against virulent strains. However, although these animal studies suggest that killed whole-cell vaccines induced only low levels of protection against disease, studies in humans indicated that immunization with these vaccines reduced the number of infections and considerably modified the course of the disease . Administration of the LVS vaccine has been demonstrated to induce a variable cell-mediated immune response in humans . The nature of the protective response to tularemia is generally believed to be T-cell mediated . Previous studies with killed vaccines generated a predominantly humoral immune response that was nonprotective and was believed to have failed to generate a sufficient cell-mediated immune response. An ongoing strategy towards subunit vaccine development has been to identify those antigens of LVS that are capable of inducing a protective immune response. Up to 23 cytoplasmic and envelope antigens have been identified in F. tularensis LVS . The next step was to identify whether any of the identified antigens could individually elicit the activation of T cells . Cell-mediated immune reactions can be readily demonstrated by the lymphocyte stimulation test, and in 1987 a study was published that investigated T-lymphocyte stimulation by membrane proteins from F. tularensis . It had previously been shown that in LVS-vaccinated individuals, lymphocytes reacted with protein antigens and antibodies were produced against carbohydrate antigens present on the capsule. Protein antigens that elicited T-cell reactivity in LVS-immunized humans were investigated by using the capsule-deficient mutant of LVS. Several polypeptide antigens were identified, and four major ones were purified, estimated at 61, 37, 32, and 17.5 kDa. It was believed that the latter two proteins were present on the bacterial surface. All four polypeptides caused lymphocyte proliferation in cells from vaccinated individuals. Subsequently, the cell-mediated immune response to LVS of individuals immunized by natural infection with F. tularensis was examined . Several membrane polypeptides of LVS were recognized, including the 40-kDa protein as well as the four polypeptide antigens identified in 1987. Among the conclusions of this study was that the polypeptides relevant to the inducement of cell-mediated immune are well conserved in the live vaccine strain. There was high immunological specificity of these proteins, because T cells from nonimmunized and uninfected individuals did not show evidence of proliferation against them. Additionally, Surcel and coworkers identified two heat-modifiable proteins of 17 and 40 kDa. The 17-kDa protein showed strong T-cell proliferative activity in subjects immunized with LVS, whereas the 40-kDa protein did not induce proliferation of T cells, although a strong antibody response to it was evident. A similarity between this 40-kDa protein and the OmpA-like outer membrane proteins in other bacterial species was also suggested. The significance of the lack of T-cell proliferation was explained because OmpA proteins have very little alpha-helical structure in contrast to major T-cell epitopes, which are alpha-helical peptides. A further study investigated whether immunization with another protein, FopA, could induce a protective response in the mouse model of infection . The protein was delivered in Salmonella cells to enhance the cell-mediated immune response, and although a strong immune response was generated, there was no evidence of protection against an LVS challenge. In addition to FopA, another membrane component has been the focus of a number of studies. LPS from F. tularensis is reported to have low endotoxicity in the Limulus amoebocyte lysate assay in comparison to LPS from other bacterial species . Studies in mice have demonstrated that immunization with LPS purified from LVS offers protection against both LVS and Schu S4 challenge . This protection has been shown to be antibody mediated, as demonstrated by serum passive transfer experiments. This protective role for LPS identifies it as one possible component of a subunit vaccine. Analyses of immune responses after LVS immunization have identified the 17-kDa membrane protein as a suitable subunit candidate because of its strong T-cell proliferative activity. A DNA fragment containing two genes, one of which encoded the 17-kDa protein, was cloned into an attenuated S. enterica serovar Typhimurium strain . After oral immunization with the construct, mice were challenged with LVS. Immunized mice showed lower viable counts of LVS in tissues after challenge compared to controls, and it was suggested that this might involve a T-cell-mediated mechanism. Further studies with this construct showed that the 17-kDa protein-mediated protection was not as high as LVS-mediated protection in the mouse model . Incorporation of the 17-kDa protein into immunostimulatory complexes again elicited a strong immune response but did not confer any protection against LVS infection . Although a subunit tularemia vaccine that offers protection against challenge with a fully virulent strain has not yet been identified, progress towards identifying protective antigens has been made. Sequencing of the F. tularensis strain Schu S4 genome will facilitate the identification of protective antigens through bioinformatics. Analysis of the immune response to the LVS vaccine has shown a heterogeneity of immunogenic epitopes recognized in humans, and this indicates it is likely that a subunit vaccine will be composed of a number of protective antigens to provide protection against virulent strains. CHEMOTHERAPY : The aminoglycosides streptomycin and gentamicin are bactericidal against F. tularensis and are currently the drugs of choice for the treatment of tularemia infections . Alternative therapies have been proposed, although generally there is a lack of supporting clinical data . The fluoroquinolones have been shown to have good bactericidal activity against F. tularensis in in vitro systems , and both ciprofloxacin and doxycycline have been found to be effective in treating F. tularensis infection in mice . In a recent human epidemic outbreak in Spain, ciprofloxacin was the antibiotic with the lowest level of therapeutic failure and with the fewest side effects . Ciprofloxacin was also shown to be suitable for the treatment of tularemia in children and in a case where relapse was evident after initial gentamicin therapy . The efficacy of levofloxacin has also been demonstrated in the treatment of two immunocompromised patients diagnosed with tularemia . Tetracycline and chloramphenicol are bacteriostatic against F. tularensis and have been used to treat tularemia. However, treatment failures have been associated with these antibiotics, although the chances of relapse are reduced with longer treatment regimens . In a study of in vitro susceptibility of F. tularensis strains isolated from humans and animals, all isolates tested were found to be resistant to beta-lactams and azithromycin . Also, evidence of in vitro activity for ceftriaxone against F. tularensis did not correlate to successful treatment when used clinically . The ketolide telithromycin has been shown to be bactericidal for F. tularensis both in axenic medium and within a cell culture system and is known to be highly active against other intracellular pathogens, including Chlamydia and Legionella spp. Since there is no effective control of this disease in nature, public awareness of the ubiquitous presence of this organism and the potential for human infection should be maintained. In areas where tularemia is endemic, the handling of dead or moribund animals should be avoided, and the possibility of insect bites should be reduced. The chlorination of municipal drinking water has virtually eliminated epidemics from that source, but untreated water should be considered when other routes are not evident . CONCLUSIONS : In spite of the clinical significance of F. tularensis, little is known about the genetic makeup of the bacterium or its mechanisms of pathogenicity. While it is known that the bacterium can invade a range of cell types, and it is clear that the bacterium is primarily an intracellular pathogen, the mechanisms allowing invasion and growth within host cells are not known. This lack of information is hindering research aimed at developing improved vaccines against tularemia. One obvious starting point for these investigations would be to determine the mechanisms of attachment to host cells. The availability of genome sequence information should facilitate the identification of candidate adhesins and pili. Second, the ability of F. tularensis to survive and grow within the endocytic vacuoles of macrophages appears to be critical for its lifestyle. The mechanisms by which this pathogen prevents phagosome-lysosome fusion should be addressed. The completion of the ongoing genome sequencing projects will certainly provide a wealth of information, and the challenge for the future will be to interpret this information. Microarrays should allow this information to be fully exploited, for example, by identifying genes which are upregulated in macrophages. However, it is unlikely that significant progress will be made in understanding the molecular basis of virulence unless efficient methods for the construction of defined allelic replacement mutants can be devised. The development of such methods therefore remains one of the highest priorities for workers in this field. Equally intriguing is the relationship between Francisella strains of differing virulence and between F. tularensis and the related arthropod endosymbionts such as W. persica. Are these arthropod endosymbionts, with a reduced host range and limited metabolic ability, possible ancestors of F. tularensis, which has acquired additional DNA sequences? Alternatively, the arthropod endosymbionts may have evolved from F. tularensis and their constrained lifestyles might reflect genome downsizing. The ongoing genome sequencing projects and the use of microarrays to probe the genetic makeup of F. tularensis strains belonging to different subspecies should allow these questions to be answered. However, again it will be difficult to prove the relationship between genetic makeup and virulence in these different subspecies without the ability to generate defined allelic replacement mutants. Although only limited genetic information is available, it has been very effectively exploited to provide modern methods for the detection of F. tularensis and for the diagnosis of tularemia. However, in spite of the availability of these tools, we are still not sure of the life cycle of the bacterium or the true reservoirs of the bacterium in the environment. In this respect, the association with amoebae is intriguing and certainly merits further attention. Similarly, the finding that the bacterium can enter a viable but nonculturable state should be investigated further. For example, can such bacteria be recovered after multiple passages in animals? The availability of genome sequence information is now likely to provide methods for strain typing and for detailed epidemiological analyses which might support these studies. The LVS vaccine is known to be effective in preventing tularemia in humans, and since an improved vaccine is likely to be at least a decade away, additional studies should be undertaken to characterize this vaccine, with a view towards licensure. In parallel, work should continue to attempt to identify protective subunits (this work will also serve to support licensing if the LVS vaccine). In the absence of an effective vaccine, antibiotics are the only available treatment or therapy, and the need to monitor for the appearance of antibiotic-resistant variants strains remains critical. Backmatter: PMID- 12364374 TI - Interactions among Strategies Associated with Bacterial Infection: Pathogenicity, Epidemicity, and Antibiotic Resistance AB - Infections have been the major cause of disease throughout the history of human populations. With the introduction of antibiotics, it was thought that this problem should disappear. However, bacteria have been able to evolve to become antibiotic resistant. Nowadays, a proficient pathogen must be virulent, epidemic, and resistant to antibiotics. Analysis of the interplay among these features of bacterial populations is needed to predict the future of infectious diseases. In this regard, we have reviewed the genetic linkage of antibiotic resistance and bacterial virulence in the same genetic determinants as well as the cross talk between antibiotic resistance and virulence regulatory circuits with the aim of understanding the effect of acquisition of resistance on bacterial virulence. We also discuss the possibility that antibiotic resistance and bacterial virulence might prevail as linked phenotypes in the future. The novel situation brought about by the worldwide use of antibiotics is undoubtedly changing bacterial populations. These changes might alter the properties of not only bacterial pathogens, but also the normal host microbiota. The evolutionary consequences of the release of antibiotics into the environment are largely unknown, but most probably restoration of the microbiota from the preantibiotic era is beyond our current abilities. Keywords: INTRODUCTION : The mechanisms involved in the virulence (defined as the relative capacity of a microbe to cause damage in a host [, ]) of pathogenic bacteria as well as those determining antibiotic resistance are important and widely studied topics in clinical microbiology. However, they have rarely been analyzed and integrated as we intend to do in the present review. In terms of evolution and ecology, antibiotic resistance and virulence determinants share some basic characteristics. Since these determinants have been acquired by horizontal gene transfer from other organisms, many are examples of what has been termed "evolution in quantum leaps" . Also, most determinants serve to escape the action of antibacterial defense systems that have developed either by natural (host anti-infectious mechanisms) or cultural (antibiotics) evolution to prevent infections. Both the natural anti-infective defenses and antibiotic treatments lead to stringent conditions for bacterial growth. In biology, any limiting condition for the majority is a golden opportunity for the minority. Those bacteria that are capable of surviving and multiplying under these conditions will gain access to organic spaces in which competition with other microorganisms is avoided (exclusive environments). We might then assume that both virulence and antibiotic resistance are formally similar adaptive mechanisms selected to survive under stress conditions (either host invasion or antibiotic treatment). From an ecological point of view, both infective conditions and antibiotic treatments are evolutionary bottlenecks that tend to reduce microbial biodiversity, so that only a very specific subset of bacteria are capable of colonizing the host under those conditions . There are several bacterial species that are able to grow at 37C and are tolerant to the oxygen tension present in different parts of the human body. The fact that environmental microorganisms that are unable to produce disease in the healthy host frequently infect immunocompromised patients indicates that many organisms are ecologically compatible with the physicochemical conditions within the human body. The human body and its physicochemical conditions are then an ecological space that can be colonized by several microorganisms , frequently with an environmental origin . In the normal host, this potential for colonization is limited by the immune system, which actively impedes colonization of the human body by opportunistic pathogens. In the immunodepressed host, only antibiotic treatment maintains a small colonizable space in the human body (see below). We want then to go one step beyond. Is there any evolutionary relationship between resistance and virulence? If modern medicine has limited the spread and maybe the evolution of bacterial pathogens, this has been done at the expense of increasing antibiotic resistance. Apparently, a decrease in the size of pathogenic populations and an increase in the number of antibiotic-resistant microorganisms have characterized the evolution of infectious diseases. In pure theory, when the number of pathogens decrease to a critical value, antibiotics should be less required, and restoration of antibiotic susceptibility could be expected to occur. We know, however, that this is not true. The widespread dissemination of antibiotic resistance among bacterial populations has maintained or even increased the number of harmful bacteria involved in infections. In fact, and in spite of previous claims, infectious diseases are among the most prominent health problem nowadays , in part as the consequence of the increasing number of antibiotic resistance phenotypes, which make bacterial infections untreatable by current therapeutic protocols . If new antibiotics and protocols are required for fighting infectious diseases, we must first understand the relationship between virulence, transmissibility, and antibiotic resistance. Thus, two essential topics are first reviewed in this work: the effect of acquisition of antibiotic resistance (and antibiotic treatment) on bacterial virulence, and analysis of whether pathogenicity and antibiotic resistance might prevail as linked phenotypes in the world of the future, if pathogens become antibiotic resistant as a consequence of intense antibiotic selective pressure. FIG. 1. | Infection and antibiotic treatment are both stringent growing conditions. Infection and antibiotic treatment are both stringent growing conditions. Several bacteria are able to grow at the temperature and oxygen tension of and using the nutrients present in the human body. However, only some are able to produce infection; this is shownin panel A. Several bacterial species (ovals are pathogens, circles are environmental ones) coexist outside the host. Some species are able to displace the commensal flora, traverse through different epithelia, resist the action of macrophages, antibodies, defensins (with squares), and all the anti-infectious mechanisms of the human body, to finally reach the target cells where the disease is produced. At any of these sequential bottlenecks, only some bacteria are selected, and their population is further amplified, so that, from the high variability of bacteria that could potentially produce the disease, only some are really pathogenic. In panel B, the same situation is analyzed but under antibiotic treatment (spheres). In this case, only a small proportion of the cells belonging to the infectious species (the antibiotic-resistant ones; blue ovals with a red line) are able to produce infection, so that the antibiotic-treated host is a more stringent ecosystem for the growth of bacteria. However, in the case of a debilitated person (panel C), the situation is somewhat different. In these patients, the indigenous microflora might be removed because of antibiotic use, the epithelial integrity may be impaired (intubated patients), the immune system can be abolished (immunocompromised patients), and even the target cells can change. Under these circumstances, some environmental species can produce infection (opportunistic pathogens), because growth conditions are less stringent. However, at least in the developed world, those patients are usually under heavy antibiotic treatment, frequently with a combination of antibiotics (golden and blue spheres), so that only those species with an intrinsically antibiotic-resistant phenotype can produce the infection. Under these circumstances, the main selective force is antibiotics, so that antibiotic treatment becomes a risk factor for some pathogens. GENETIC LINKAGE OF RESISTANCE AND VIRULENCE : Virulent bacteria have acquired their phenotype through a long evolutionary course in close contact with their natural hosts. Most virulence determinants are either located in chromosomal gene clusters (pathogenicity islands) or harbored in genetic accessory elements such as plasmids and phages. This suggests that evolution from an avirulent way of life to pathogenicity frequently implies the acquisition of foreign pieces of DNA . Nevertheless, for the organism to be a real pathogen, these pathogenicity factors should be inserted in an organism ecologically compatible with the potential host. Moreover, in some cases it is not an acquisition but a deletion (virulence-associated "black holes") which is needed to become a pathogen . Indeed, any change in lifestyle has a biological cost, as functions needed in one habitat may cause a burden in another habitat and could therefore be counterselected. Acquisition of a virulence phenotype might then require the acquisition of some different pathogenicity islands and the loss of some chromosomal DNA regions. Therefore, the construction of a pathogen by the acquisition of specific pathogenic elements in ecologically compatible host-adapted bacterial genomes has probably occurred over a long evolutionary time. In this regard, the record of infectious diseases (mainly epidemic ones) in human history and bioarchaeological analysis of the paleontological record indicate that we have been in contact with pathogens similar to those currently producing infections for a long time. In this way, it has been demonstrated that infectious diseases such as syphilis, tuberculosis, and other infections due to bacteria, viruses, and parasites were common in prehistoric men , particularly in the Neolithic age, when agriculture and farming ensured close contact between humans and between humans and animals. In some cases, however, such evolution has occurred in historical times, as the evolution from Yersinia pseudotuberculosis to Y. pestis, which was shortly followed by the first pandemic plagues . Obviously, the evolutionary outcome of pathogenic organisms tends to be limited in modern times by the anti-infectious repertoire developed by humans that includes hygienic measures, epidemiological controls, vaccines, and, significantly, antibiotics. Conversely, acquisition and further spread of antibiotic resistance genes among pathogenic bacteria is a phenomenon that has occurred just in the last 50 years as a consequence of extensive antibiotic use for human therapy and animal farming. At first glance, pathogenicity and resistance should be unlinked phenomena. However, several examples indicate that this is not the situation for several bacterial pathogens. Antibiotic resistance and virulence genes can be linked (and then coselected) in the same replicon, or eventually a single determinant can be involved in both virulence and resistance. Some examples of these situations will be reviewed. Common Mechanisms Involved in Virulence and Resistance : In this section, we analyze whether a mechanism which has been selected on the basis of a bacterial virulence phenotype might also be relevant for antibiotic resistance and vice versa. Let us first examine the effect of the spatial localization of bacteria in the human body. Several virulent bacteria base their pathogenic characteristics in an intracellular way of life during infection . Internalization may be required for the induction of inflammatory cytokines and produce tissue damage through the induction of either necrotic or apoptotic responses. A remarkable example is the apoptosis of macrophages induced by pathogens such as Shigella, which both avoid the antibacterial effect of those cells and trigger inflammation . In some well-characterized examples, bacteria are able to travel from cell to cell without any significant contact with the extracellular milieu , thus avoiding the immune system as well as contact with antibiotics that do not enter mammalian cells . This provides the first clue as to how well-characterized virulence phenotypes can be related to antibiotic resistance. Mammalian cells are poorly permeable to or easily exclude several families of antibiotics . Even if the antibiotic enters the mammalian cells, intracellular growth might induce a transient antibiotic-resistant phenotype, a situation that has been described for Legionella pneumophila . The location within the host may alter in an unspecific way the susceptibility of the bacterial organism to antibiotics. Haemophilus influenzae grown in animals undergoes modifications in penicillin-binding proteins, as peptidoglycan metabolism is directly affected by the environment , and Salmonella peptidoglycan is dramatically altered for bacteria growing intracellularly . Although the effect of these changes on antibiotic susceptibility has not been analyzed extensively, it is possible that those metabolic changes may alter the activity of antibiotics such as beta-lactams against bacteria growing during infection. It is then possible that intracellular localization may allow bacteria to maintain a phenotype of antibiotic resistance. In this respect, treatment with some antibiotics might select for intracellular clones. This could be the case for species in which some clones are able to invade mammalian cells and others are noninvasive; one such species is Pseudomonas aeruginosa . On the contrary, some antibiotics, such as macrolides and fluoroquinolones, are actively accumulated by human cells . In this case, intracellular bacterial localization might eventually increase the antibiotic effect against bacterial pathogens, leading to better eradication and favoring the less invasive strains. A similar situation of phenotypic antibiotic resistance can be observed for bacteria growing in biofilms. Bacterial biofilms are frequent in persistent infections , such as those associated with cystic fibrosis , chronic bronchitis, osteomyelitis , and foreign-body-associated infections. Bacteria growing in biofilms are more resistant than those under a planktonic way of life to the action of phagocytic cells' antibacterial activity as well as the action of antibiotics . As in the previous case, antibiotics might select biofilm-forming organisms, thereby increasing the prevalence of chronic infections. Other mechanisms of virulence could prevent the action of antibiotics against bacteria. For instance, toxins leading to local necrosis or abscess formation certainly will decrease the local availability of antibiotics due to reduction in antibiotic arrival to the foci or because of local inactivation of the drugs by altered pH, free proteins, or DNA. In these cases, the mechanisms of pathogenicity may finally serve as mechanisms for antibiotic resistance. A final example of a virulence mechanism with a role in antibiotic resistance is presented by Bordetella pertussis, the etiological agent of whooping cough. The cell wall of the virulent strains of this microorganism is infrequently susceptible to autolysis triggered by beta-lactam antibiotics; only avirulent B. pertussis strains are known to be lysed. It was demonstrated that this phenotypic tolerance and the antibiotic-induced autolytic activity are controlled by the vir locus, which determines phase transition in B. pertussis, a key element in the virulence properties of this bacterial species . In previous examples, we have seen that a "virulent way of life" can contribute to an antibiotic resistance phenotype. Could antibiotic resistance determinants also contribute to bacterial virulence? Some examples suggest that this could be the case. Recently, one area of intense study of antibiotic resistance is the analysis of multidrug resistance (MDR) efflux pumps . These determinants are able to extrude an ample range of substances that include antibiotics, solvents, dyes, and quorum-sensing signals . Extrusion by the same pump of compounds as diverse as erythromycin, quinolones, beta-lactams, and ethidium bromide is a rule more than an exception. Might these MDR determinants extrude compounds involved in the host defense be contributing to bacterial virulence? The answer is yes. For instance, adaptation to growth in the presence of bile salts is a prerequisite for any pathogen to colonize the intestinal tract. It has been reported that Escherichia coli extrudes bile salts through the acrAB system, which was first characterized as an MDR determinant ; the same occurs in Salmonella enterica serotype Typhimurium. In fact, it has been suggested that extrusion of bile salts present in the intestinal ecosystem to which these bacteria are adapted is the function for which those MDR determinants were selected . Analysis of S. enterica serotype Typhimurium mutants with high susceptibility to bile salts has demonstrated that this higher susceptibility was a consequence of inactivation of the MDR determinant acrB. This inactivation leads to a reduced ability to colonize the intestinal tract in a murine infection model , which indicates that acrAB is involved in both antibiotic resistance and virulence. Some MDR systems, such as Mtr from Neisseria gonorrhoeae and SapA from S. enterica serotype Typhimurium, can actively extrude both defensins and antibiotics, thus contributing to reducing antibiotic susceptibility as well as to increasing the number of virulence determinants of the organisms. In fact, it has been described that resistance to defensins is required for a full virulence phenotype in the case of Salmonella . The mechanisms of action of defensins resemble at least in part that of peptidic antibiotics currently used in clinical practice . Thus, it is conceivable that they might share the same mechanisms for resistance. This situation has been described for polymixin B. The sensitivities to polycationic peptides were compared between two groups of Yersinia enterocolitica isolates, one with an environmental origin and the other from infections . Data indicated that pathogenic strains of Y. enterocolitica were resistant to defensins and also to polymixin B, showing a clear cross-linkage between antibiotic resistance and virulence as a consequence of the expression of a resistance determinant. Not only MDR pumps but also biocide efflux determinants may play a role in bacterial virulence. It has been reported that the QacA pump of Staphylococcus aureus, involved in resistance to several organic cations (biocides included), has a role in resistance to thrombin-induced platelet microbicidal protein . Thus, the presence of the pump contributes to the survival of strains carrying this determinant at sites of endothelial damage as well as in experimental endovascular infections. We want to stress here that the mechanisms of extrusion of antibiotics are similar to those translocating some proteins involved in pathogenicity. The hemolysin export apparatus in Escherichia coli comprises the outer membrane channel trimmer protein TolC that is also involved in the AcrA-B multidrug efflux transport complex . The periplasmic proteins of both systems, AcrA and HlyD, have a similar structure and are able to interact with TolC . In all previously discussed examples, a mechanism selected for bacterial virulence can also produce an antibiotic resistance phenotype. Selection for the most virulent strains might select for antibiotic resistance, and conversely, selection for antibiotic resistance might select the most virulent microorganisms. However, the opposite situation can also be found: antibiotic resistance may decrease bacterial virulence (see below). For instance, the KatG catalase-peroxidase activity is important for the survival of Mycobacterium tuberculosis in the host . Mutations that eliminate this activity prevent the activation of isoniazid and are the major cause of resistance to this drug in Mycobacterium spp. . It might then be predicted that isoniazid-resistant Mycobacterium mutants might be less virulent than wild-type isoniazid-susceptible strains, at least in the case of M. tuberculosis and Mycobacterium bovis . However, when isoniazid-resistant isolates were analyzed, such reduced virulence was not found. It turned out that M. tuberculosis pathogenic isolates accumulate compensatory mutations in the gene encoding the alkyl peroxidase AphC, which can substitute for KatG for survival inside the host . Costs of and Compensations for Virulence and Resistance : It is largely assumed that acquisition of novel genetic determinants may have a cost for the bacterial host. That may happen because of partial incompatibility of previous and acquired lifestyles, or because of the extra energy required to maintain the genetic vectors carrying the new genes. The acquisition of plasmids or antibiotic resistance-virulence genes might have an effect on bacterial fitness . As previously stated, pathogenicity islands, acquired through horizontal transfer of large gene arrangements, have a major role in bacterial virulence . Also, large vir plasmids are needed for some pathogens, such as Shigella and Yersinia , to express a pathogenic phenotype. We might speculate that acquisition of these determinants implied a cost in bacterial fitness when they were included in the genome of the previously nonvirulent bacteria. However, acquisition of resistance genes by pathogenic bacteria has occurred over the last 50 years, but acquisition and further evolution of structures as pathogenicity islands occurred thousands of years ago. Thus, the evolutionary time needed to acquire and optimize compensatory mutations to alleviate the effect on fitness of the acquisition of pathogenicity determinants has been much longer. Also, for plasmids carrying vir determinants, the costs of the acquisition of such determinants might be compensated for by mutations in other loci. Of note, it has been shown that the vir plasmid can be easily lost in vitro in Shigella . However, growth conditions in vivo should preserve the plasmid's presence during infection, otherwise Shigella would not be virulent anymore. Acquisition of vir plasmids ensures the ability to colonize a different biological compartment, thus evading competition with other bacterial species and eventually reducing the potential fitness cost imposed by the new genetic element. The same has probably occurred for enteroinvasive E. coli isolates . Recent work suggests that the invasive ability of those isolates has evolved in different chromosomal backgrounds, probably through the spread of plasmid-borne invasion genes, and the maintenance of invasive phenotypes in separate lineages suggests that this ability confers a selective advantage to invasive strains . Although the Ewald hypothesis is still controversial, it can be speculated that sanitation procedures, vaccination, and widespread antibiotic use, impeding host-to-host spread of vir plasmid-based pathogens, may reduce the overall pathogenic power of microorganisms. Due to the rapid acquisition and fixation of compensatory mutations by bacteria, experiments to evaluate the biological cost of the expression of novel antibiotic resistance and virulence determinants are difficult to perform. Antibiotic resistance and virulence can be acquired either by horizontal transfer of antibiotic resistance genes or by mutation . The presence of new plasmids and transposons in the bacterial genome has a relevant cost for the recipient bacteria . However, it has been demonstrated that under these circumstances, the cost is compensated for in few generations as the consequence of genetic change by the host, getting the specific plasmid-bacterial strain association more fit than the previous non-plasmid-containing bacterium . Also, in the case of mutations leading to antibiotic resistance, some examples demonstrate the possibility of fitness reduction, which may also reduce bacterial virulence (see last example in previous section). For instance, current evidence supports the idea that highly fluoroquinolone-resistant Salmonella strains could be counterselected in the field . A substantial reduction of the virulent characteristics has been also described for antibiotic-resistant S. enterica serotype Typhimurium isolates . Although this may occur for some bacteria, it does not mean that it will always occur. In fact, an outbreak of quinolone-resistant S. enterica serotype Typhimurium DT104 has recently been described , and an increase in quinolone-resistant Campylobacter jejuni infections has been reported in Minnesota . The emergence of compensatory mutations rapidly alleviates the biological cost of antibiotic resistance in S. enterica serotype Typhimurium isolates , and this could be the situation for the aforementioned outbreaks of quinolone-resistant bacteria. Noteworthy, the compensatory mutations that are selected in vivo and in vitro are different, reflecting the different growing conditions for bacteria during infection compared with microorganisms growing in vitro . This indicates that the metabolic requirements needed for infection are different from those for surviving in the environment. Therefore, acquisition of the characteristics required for infection and antibiotic resistance might make them less proficient for surviving in the environment and to be transmitted among different hosts. It seems then that acquisition of novel virulence and antibiotic resistance traits has a cost in bacterial fitness, but the cost is rapidly compensated for due to the high plasticity of bacterial genomes and the huge populations of bacteria from which compensatory mutants can be selected . However, this situation may not always occur. Possibly, only those mechanisms that can be compensated for by mutations in other loci are selected because these are the only ones that can avoid fitness reduction. As stated by other authors , the effect of antibiotic resistance (and in a higher grade of virulence determinants) on bacterial fitness has been properly addressed in only a few studies. This topic needs to be investigated so that the biological potential of novel virulent and antibiotic-resistant bacteria may be predicted. Evolution and Dissemination of Genes Involved in Virulence and Resistance : As previously stated, both antibiotic resistance genes and virulence determinants have been acquired by horizontal gene transfer in most cases . One intriguing question that has not been completely resolved resides in the origin of these determinants. It is widely accepted by the plasmids and transposons that carry antibiotic resistance genes may have originated in antibiotic-producing organisms in order to avoid the deleterious effect of the antibiotic on themselves . Eventually, they could have further evolved in organisms in an ecological consortium with antibiotic producers. More recent works indicate that this hypothesis about the origin of resistance is only half true . For instance, it is difficult to believe that chromosomal beta-lactamases as well as some aminoglycoside-inactivating enzymes and a plethora of MDR determinants which are present in all isolates of a given bacterial species originated in the antibiotic producers. In fact, all these determinants must have a functional role other than antibiotic resistance, and this phenotype will only be a consequence of their primary physiological function (see reference for a review of this concept). In the case of virulence determinants, the current paradigm indicates that pathogenicity islands acquired through horizontal gene transfer are frequently responsible for the pathogenic properties of virulent bacterial species . Different from antibiotic resistance genes, the pathogenicity islands are more difficult to explain, because pathogenic bacterial ancestors carrying such gene clusters have not been detected. It has been proposed that pathogenicity islands might be relevant for the biodegradative properties of microorganisms (including decomposition of dead bodies), to kill living cells (to obtain food and reduce competition), and to live inside eukaryotic cells (such as amoeba, protozoa, plants, and animals) in natural environments . Antibiotic resistance genes and virulence determinants might then play a different role in the original organisms from which they were transferred to pathogenic bacteria. However, once those determinants have been selected in the heterologous host, they confer a selective advantage and are fixed in the pathogenic bacterial populations, where they can evolve further and eventually be transferred to other bacterial species. Mutation and recombination. : Two different processes account for evolution of antibiotic resistance-virulence phenotypes: gene mutation and gene recombination (horizontal gene transfer included). The role of mutation in antibiotic resistance is well known . In the case of virulence, fewer examples of such a role have been described. However, mutational activation-inactivation of intrinsic genes might also contribute to a virulent phenotype in the case of opportunistic pathogens. As an example of cryptic virulence determinants activatable by mutations, laboratory isolates of E. coli are hemolytic because of mutations in the hns and fnr genes. Whether this is just a laboratory curiosity and might actually have a role in the course of nosocomial infections by E. coli is a matter of speculation. Some virulence determinants of P. aeruginosa, such as alginate production and cytotoxicity, are downregulated. However, mutants in which the expression of such determinants is derepressed are frequently found during infection . This indicates that mutation probably plays a major role in the emergence of the different virulent phenotypes shown by P. aeruginosa clinical isolates. Pathogenic P. aeruginosa isolates can be broadly classified into two groups, those with a cytoinvasive phenotype and those with a cytotoxic phenotype. However, both kinds of isolates contain the genes required for invading epithelial cells. Invasive and cytotoxic strains differ in the expression of the genes under the control of an activator called ExsA . Mutations in the gene exsA, which encodes the activator, lead to not a cytotoxic but only an invasive phenotype . Although both types of bacteria are virulent, they occupy different environments (inside and outside epithelial cells), so that one or the other phenotype might be selected in vivo by, for the moment, unknown processes. Recombination also has an important role in the evolution of antibiotic resistance determinants. Progressive clustering of genes, leading to an operon structure, may have occurred to optimize and regulate the expression of ancient genes producing a resistance (frequently low-level resistance) phenotype . In a similar way, the construction of a pathogenicity island requires the recruitment of different genes that recombine to produce a single genetic element containing several genes with a common role, virulence. In other cases, intragenic recombination is essential to produce a resistant phenotype. The most conspicuous example of recombination in chromosomal DNA leading to antibiotic resistance is beta-lactam resistance produced by recombination of penicillin-binding-protein genes in Neisseria gonorrhoeae and Streptococcus pneumoniae . Similar recombinatorial processes may have influenced the evolution of virulence. However, unlike for antibiotic resistance, few examples of recent evolution of bacteria to a virulent phenotype have been described. One might be recombination between different sets of genes, which causes the rearrangements that can be observed in the capsular antigens of S. pneumoniae (-, ) and can be considered a mechanism of defense against host immunity. Another could be the emergence and further dissemination of enterohemorrhagic E. coli O157:H7 strains . This E. coli serogroup has emerged as a relevant pathogen in the last 20 years. Enterohemorrhagic E. coli isolates contain virulence plasmids and pathogenicity islands similar to those found in Shigella spp. . Although it was a matter of speculation whether these genetic determinants were acquired recently by E. coli, the recent sequencing of the genome of an E. coli O157:H7 strain demonstrated that it contains as many as 1,387 new genes in comparison with the previously sequenced nonpathogenic laboratory strain E. coli K-12 . We must be aware that acquisition of pathogenicity determinants by previously nonpathogenic organisms might occur in a similar way as it happens with antibiotic resistance determinants. For instance, the acquisition of virulence plasmids by bacteria forming part of the human indigenous flora and thus already well adapted for surviving inside their host (such as E. coli) might lead to the emergence of novel pathogens. In a similar way, the acquisition of virulence determinants by environmental microorganisms should produce novel pathogenic bacteria. Although the evolution of novel pathogens is possible, the selective pressure in favor of the selection of virulence determinants is not as strong as in the case of antibiotic resistance determinants. Therefore, an explosion of novel pathogens, such as has occurred with antibiotic-resistant bacteria, will most likely not happen in the near future, not only because the selective pressure for a pathogenic phenotype is less strong than for an antibiotic resistance phenotype, but because selection for pathogenicity in today's world might decrease with sanitation if the Ewald hypothesis is true. A critical point to discuss here is that virulence and/or antibiotic treatment might increase the rate of bacterial variation. In the case of infection, this situation might contribute to increased mutation rates and even to the selection of hypermutator strains (see below), thus enhancing antibiotic resistance mutation . In a similar way, antibiotic challenge might also produce hypermutable phenotypes , increasing the possibility of mutants overexpressing virulence determinants. Mechanisms favoring hypermutation may also facilitate recombination . Moreover, in vivo transfer of plasmids carrying antibiotic resistance genes and/or virulence determinants and recombination are probably favored during infection due to host signals that enhance gene transfer . In this respect, the effect of antibiotics in inducing the transfer of plasmid and transposons has been demonstrated in vitro . It has been described that the same genes required to initiate infection of human macrophages by Legionella pneumophila are involved in plasmid mobilization . Recent results in our laboratory also suggest that bacterial expression of factors involved in cell-to-cell DNA transfer (likely antibiotic resistance plasmids) in some organisms, such as S. pneumoniae, may be triggered by inflammatory products (M. R. Baquero, unpublished results). It can then be suggested that bacteria are expected to evolve more rapidly inside the host and under antibiotic selective pressure, so that an infected patient under antibiotic therapy may act as an evolutionary accelerator. Plasmids. : It is well known that plasmids are major vectors for the dissemination of both antibiotic resistance and virulence determinants among bacterial populations. It is also clear that the presence of virulence and antibiotic resistance determinants in the same genetic element will produce coselection of both types of determinants. This applies as well for genes present in transposons, phages, and cassettes, which are discussed below. For bacteria carrying those linked determinants, the selection of an infective population will select for antibiotic resistance, and antibiotic selective pressure will select the virulence trait. Table shows some published examples regarding this type of gene linkage. Among them, antibiotic resistance plasmids carrying, alone and in combination, genes encoding the synthesis of bacteriocins , siderophores , cytotoxins , and adhesion factors have been described. General plasmid-encoded transfer functions (present in all transferable resistance plasmids) may be relevant by themselves for bacterial virulence. One of these functions is related to the traT genes, involved in plasmid conjugation . The other function is the production of pheromones , also involved in conjugation of plasmids and transposons in gram-positive bacteria. Conjugative plasmids have a relevant role in the dissemination of antibiotic resistance. Plasmid conjugation is encoded in the tra region. It has been demonstrated that expression of traT genes might be involved in bacterial resistance to serum , biofilm development , and phagocytosis , therefore contributing to the virulent properties of bacteria carrying conjugative (frequently antibiotic resistance) plasmids. The plasmid transfer proteins are sometimes chromosomally encoded, but they can have a role in both plasmid transfer and virulence. This could be the case for Legionella pneumophila. It has been described that the dot virulence genes encode a large putative membrane complex that functions as a secretion system that is able to transfer plasmid DNA from one cell to another . Mutations in the dot genes reduced both virulence and plasmid transfer. In the case of gram-positive bacteria, aggregation substances encoded by pheromone plasmids are involved in the clumping required for an efficient transfer of DNA by conjugation. It has also been described that Enterococcus faecalis aggregation substance promotes resistance to killing by human neutrophils in spite of phagocytosis and neutrophil activation and enhances pathogenicity in a rabbit model and thus, as said by R. Wirth, is "more than a plasmid collection mechanism" . The analysis of plasmids from the preantibiotic era demonstrated that there have not been major changes in the families of plasmids present in pathogenic bacterial populations, but just recruitment of antibiotic resistance genes by formerly "antibiotic-susceptible" plasmids . Since plasmids encoding adaptive traits (eventually involved in the pathogenic lifestyle) certainly preceded antibiotic resistance, it could be suspected that many current resistance plasmids contain genes with a role in bacterial colonization and virulence. In fact, this has probably been the origin of plasmids containing virulence determinants and antibiotic resistance genes such as those listed in Table . We do not know, however, if this association is the rule for antibiotic resistance plasmids, because only a very small fraction of these determinants have been entirely sequenced. TABLE 1 | Examples of virulence determinants encoded by antibiotic resistance plasmids Transposons. : Transposons are also relevant for the dissemination of antibiotic resistance genes, either by integration in transferable plasmids or by direct conjugation and further integration in the bacterial chromosome. Also, transposons containing virulence determinants have been described. One example is the aerobactin operon . Aerobactin is a siderophore produced by several bacterial species . It has been proposed that aerobactin is a virulence factor with relevance for iron acquisition at the site of infection . Aerobactin genes have been found in the chromosomes of different bacterial species and in several different plasmids, a situation which indicates easy mobilization. The aerobactin operon forms part of a transposable element , and it has been proposed that it is part of a pathogenicity island in Shigella flexneri . Other virulence determinants found in transposable elements are the E. coli enterotoxin STII gene , the Shiga toxin operon, which has been found in a putative composite transposon in Shigella dysenteriae 1 , and the toxic shock toxin, carried by a family of mobile pathogenicity islands in Staphylococcus aureus . Little is known about the association between antibiotic resistance and virulence determinants in the same transposon. However, there is no reason for this lack of association. Indeed, like other complex gene arrangements (see below), transposons frequently have a mosaic structure in which highly recombinogenic regions are included , so that the acquisition of novel traits is a common occurrence in transposon evolution . Phages. : The presence of virulence determinants in phages infecting different bacterial species has been described . Also, a phage origin has been proposed for at least some pathogenicity islands (see below). Bacteriophage-associated transduction of antibiotic resistance determinants has been described as well . Nevertheless, the presence of antibiotic resistance genes together with virulence determinants in the same phage has not been reported. A possible explanation for this phenomenon is the size requirements for phage DNAs. Phage DNA needs to accommodate inside the head of the phage particle, so that its length must fit a fixed range of sizes. In such circumstances, the gain of some genes must be accompanied by the loss of others. Because of this, the possibility of combining antibiotic resistance genes and virulence determinants in the same bacteriophage is lower than in the case of plasmids, which have less stringent requirements for the incorporation of novel DNA fragments. TABLE 2 | Examples of virulence determinants encoded by phages Gene cassettes. : In the last few years, several works have shown the presence of gene cassettes capable of integrating novel genes from different DNA sources, offering possibilities for the horizontal dissemination-linked antibiotic resistance-virulence genes. The most relevant family of these gene cassettes is the integron family . Integrons are the primary systems for the capture of antibiotic resistance genes in gram-negative bacteria and have also been described in gram-positive bacteria. Integrons are formed by gene cassettes located downstream of a recombinase-encoding conserved sequence that includes a strong promoter . This organization allows the formation of large arrays of gene cassettes which can eventually be transferred as a whole between different replicons. Most of the integrons described to date are formed by antibiotic resistance genes . However, several different genes, including not only antibiotic resistance but also virulence determinants, have been found in integrons. Prototypic of composite integrons containing virulence genes are the VCR cassettes found in the chromosome of Vibrio cholerae. VCRs (from Vibrio cholerae repeated sequences) are a family of 123- to 126-bp sequences of imperfect dyad symmetry, highly repeated in the chromosome of V. cholerae (up to 60 copies). Analysis of VCR clusters demonstrated that the gene-VCR organization is identical to that of the antibiotic resistance cassettes present in integrons . Also, the formation of the VCR islands likely occurred by means of an integrase-mediated process. This indicates that VCR islands might have a role in gene capture similar to the proposed role of integrons in the acquisition of antibiotic resistance genes. It cannot be dismissed that some genes in VCRs may play a role in antibiotic resistance, perhaps to still undiscovered antibiotics. Another type of gene cassette with a key role in the establishment of the infective process are pathogenicity islands. These DNA regions found in the chromosomes and plasmids of bacterial pathogens are composed of clusters of genes (typically from 15 to 25, although more are possible) which have been acquired by horizontal gene transfer . It has been shown that several of these gene arrangements have been introduced into tRNA genes through a phage-mediated transfer mechanism . However, a role for insertion sequences has also been suggested in other cases. This is the case for the pathogenicity island required for growth of Y. pestis in iron-deprived environments . This island is flanked by direct repeats of IS100, an insertion sequence which is present in the chromosome and plasmid of this bacterial species and thus may contribute to recombination of genes from different replicons. The presence of sequences with a role in DNA mobility either inside or flanking pathogenicity islands is quite common . However, few of these gene clusters have been demonstrated to be mobile. One example is the aforementioned family of pathogenicity islands carrying the gene encoding toxic shock toxin in Staphylococcus aureus . Stabilization of pathogenicity islands in the bacterial genome probably requires inactivation of the functions involved in mobility. In fact, the "mobility elements" present in pathogenicity islands frequently carry a large number of mutations, often leading to stop codons , which abolish the expression of such functions. Antibiotic resistance determinants have not been detected in pathogenicity islands, but that is an open possibility. The transfer of gene clusters (contained in either plasmids, transposons, bacteriophages, or gene cassettes) is extremely relevant to bacterial adaptation to novel environments because it allows "bacterial evolution in quantum leaps" . As previously stated, both bacterial virulence and antibiotic resistance can be considered strategies to explore and colonize novel environments in which bacterial competitors are scarce. Bacteria will make use of all the tools (mutation, gene transfer, recombination) which allow the high plasticity shown by bacterial genomes . In the treated host, adaptation of pathogens will require both antibiotic resistance elements and virulence determinants, so that both adaptive mechanisms must evolve together to produce the explosive spread of virulent and antibiotic-resistant microorganisms that we are now facing. FIG. 2. | Structure of an integron. Structure of an integron. Integrons are site-specific recombination elements that mediate the acquisition and spread of genes among bacterial populations. Integrons are formed by an integrase gene followed by one primary att recombination site (dark grey box) and several cassettes, each including one gene and one 59-bp recombination site (white square). The transcription of the system is controlled by a strong promoter located upstream of the integron. This structure favors the arrangement of genes in tandem, which are transferred as single elements among bacterial populations. Phenotypic Adaptation in Virulence and Antibiotic Resistance : Over the course of an infection, bacteria must face several different environments , so that phenotypic adaptation to all of them is an important trait for the final pathogenicity result. In this respect, both virulence and antibiotic resistance determinants can be environmentally regulated. The signaling network that regulates the expression of bacterial virulence and antibiotic resistance determinants during infection is a hot topic in current research in microbial pathogenesis. Environmental signaling in virulence and antibiotic resistance. : Bacteria are constantly sensing the environment in order to respond to changes in its composition. In the case of bacterial pathogens, constant cross talk occurs between the microorganisms and their hosts which modulates the expression of several genes in both the bacterial and host cells. Physiological signals such as calcium concentration , low pH , high temperature , and low iron concentration among others, trigger the expression of several virulence genes. However, little is known about whether such signals might also trigger the expression of antibiotic resistance genes. Nevertheless, we know that the antibiotic susceptibility of bacteria is modulated by several factors which include growth phase , pH , carbon dioxide , temperature , bile salts , and low iron . It is then clear that certain conditions during infection might induce the expression of both antibiotic resistance and virulence genes, and a common regulation of both types of determinants might then occur. Most probably this regulation does not always involve the same regulatory network, but in some cases virulence and antibiotic resistance genes can be part of the same regulon. In this regard, it has recently been described that the MarA protein (from multiple antibiotic resistance) of E. coli regulates the expression of more than 60 chromosomal genes. These genes include not only MDR determinants , but also genes with a potential role in virulence, such as those involved in oxidative stress and iron metabolism . In general, a linkage between antibiotic resistance and virulence gene regulation might occur in global regulons, which modulates the expression of stress genes (see below). This type of regulation, coupled with phenotypic resistance of bacteria growing in biofilms, inside cells, and even in resting cells which are resistant to antibiotics, might produce situations of in host resistance at the site of infection that is impossible to predict by routine laboratory susceptibility tests (reviewed in reference ). Antibiotics as effectors of bacterial virulence. : Common models of bacterial virulence analyze host-pathogen interactions under a defined set of conditions that usually do not include antibiotic therapy. Although the first stages of an infection usually occur without treatment, once a diagnosis is available, most infectious diseases evolve under antibiotic treatment. We want to stress here that antibiotics are not only bacterial killers, but also modulators of bacterial and even host cell transcription of different genes . In such a way, antibiotics modulate the interaction between host cells and bacteria . It has also been suggested that they may have an important role as modulators of the interactions among natural environmental bacterial populations . For instance, quinolones alter DNA supercoiling , and the expression of many genes, including those involved in virulence, is dependent on DNA supercoiling . More recently, the role of a natural quinolone on quorum sensing in Pseudomonas aeruginosa has been demonstrated . Quinolone treatment may alter the expression of several bacterial genes. In fact, a decrease in the expression of virulence determinants upon exposure to subinhibitory concentrations of quinolones has been described in some cases . In contrast, ciprofloxacin and other SOS-inducing antimicrobial agents cause Shiga toxin-encoding bacteriophage induction and enhanced Shiga toxin production from E. coli O157:H7 both in vitro and in animal models . Therefore, quinolones clearly affect the expression of virulence determinants, but their role as either enhancers or inhibitors of bacterial virulence depends on the model analyzed and must be established for each pathogenic bacterium. On the other hand, resistance to quinolones as a consequence of mutations in bacterial topoisomerases in some cases alters DNA supercoiling , leading to changes in the level of expression of several bacterial proteins, including virulence determinants. For instance, treatment of quinolone-resistant S. aureus with subinhibitory concentrations of quinolones increases the expression of fibronectin-binding proteins, contributing in part to their emergence and maintenance in clinical settings . Other antibiotics besides quinolones can affect the expression of virulence determinants in different ways. For instance, Suerbaum demonstrated that low antibiotic concentrations increased serum sensitivity in E. coli strains protected by the K1 antigen , and low macrolide concentrations appeared to reduce the adhesive properties of some gram-negative rods . Also, antibiotic treatment may trigger the release of bacterial products, including lipopolysaccharide vesicles and endotoxin , that interact with host cells, changing the virulence characteristics of pathogenic bacteria. Exposure to antibiotics can produce global changes in the metabolism of bacteria, as is possibly the case with Mycobacterium tuberculosis. Exposure to low concentrations of antibiotics induces expression of the alternate sigma factor sigF, which is involved in the regulation of several genes, including those involved in virulence . In light of these data, a role of chemotherapy in the persistence of this important pathogen has been suggested. Recently, it has been shown that tetracyclines are iron scavengers . Iron deprivation induces the expression of several virulence-associated proteins in bacteria , and it is known that free iron is scarce during infection . In this way, tetracyclines might trigger the expression of several bacterial genes just as a consequence of their iron-chelating activity. The influence of subinhibitory concentrations of antibiotics on virulence is then dependent on the bacterial species-antibiotic combination. The response of resistant strains to antibiotics will be different from that of susceptible ones. For instance, if subinhibitory concentrations of antibiotics reduce the expression of virulence determinants, resistance to drugs (even low-level resistance ) would restore the original virulence levels. Major efforts must be made to analyze the effect of antibiotics on the metabolism of antibiotic-resistant and -susceptible bacteria. One such approach is the recently published analysis of isoniazid-induced alterations in M. tuberculosis gene expression . The authors showed that treatment with this drug produces a dramatic change in the expression of several bacterial genes, some of them related to the drug's mode of action, but others with no apparent relationship to the antibiotic killing mechanism, which may eventually influence the outcome of infection. The current availability of tools for whole-genome expression analysis may allow a better understanding of the role of antibiotics on regulation of bacterial metabolism and thus potentially virulence. Cross talk between virulence and antibiotic resistance regulons. : The environmental regulation of the expression of antibiotic resistance and virulence determinants has already been discussed. The possible cross talk of both types of regulons will now be reviewed. The most relevant example to emerge in the last few years is the effect of quorum-sensing signals on bacterial virulence and the effect of antibiotic resistance on quorum sensing. Several bacterial species are able to determine the local concentration of bacterial cells, a process known as quorum sensing . Quorum-sensing signals trigger the expression of several genes, some of which are involved in bacterial virulence . Quite interestingly, it has recently been described that mutations leading to multidrug resistance in P. aeruginosa as a consequence of overproduction of the MexAB-OprM efflux pump also affect the quorum-sensing response . This response could be due to a direct effect on the extrusion of quorum signals by this MDR pump in P. aeruginosa . More recently, it has also been described that overexpression of the MexEF-OprN multidrug efflux system affects cell-to-cell signaling in Pseudomonas aeruginosa . Since quorum-sensing signals trigger the expression of several virulence-associated genes, such as those involved in production of elastase, rhamolipid, pyocianin, proteases , and type III secretion system, antibiotic resistance might strongly influence virulence through the differential processing of quorum-sensing signals in antibiotic-resistant and antibiotic-susceptible bacterial populations. Indeed, recent work in our laboratory has shown that MDR P. aeruginosa mutants are impaired in their virulence properties . Conversely, elements with a role in virulence might also be involved in the regulation of antibiotic resistance. As previously stated, iron deprivation induces expression of the MDR efflux pump MexAB-OprM. Also, the expression of MDR determinants can be induced by salicylate . Salicylate is a relevant signal molecule in plant-bacteria interactions . It is also an intermediate in the synthesis of different siderophores and is itself a siderophore in Pseudomonas spp. . Bacterial siderophores are relevant virulence factors that are produced during infection as a consequence of the small amount of free available iron that is present in the human body . Since P. aeruginosa might produce salicylate during infection and this compound induces the expression of MDR determinants, salicylate might thus induce a phenotype of antibiotic resistance for bacteria growing under iron-deprived conditions, such as occur during infection . Mutations in the genes encoding some histidine kinases, which are involved in signal transduction mechanisms in Streptococcus pneumoniae, cause a phenotype of tolerance to the killing ability of glycopeptide antibiotics , and these mutants are also less virulent when tested in animal models . Deletion of the ciaR gene, which is involved in regulation of competence and cefotaxime resistance in S. pneumoniae, led to a 1,000-fold attenuation of bacterial growth in vivo. Deletion of the vncSR gene pair, which is involved in tolerance to glycopeptide antibiotics in this bacterial species , also produces growth attenuation, although the effect is smaller. These signal transduction systems are thus pleiotropic regulators that may influence antibiotic susceptibility, virulence, and even horizontal gene transfer. Therefore, antibiotics may select for variants unable to respond to a number of environmental signals. If that is the case, the environmental signals that induce pathogenicity traits may eventually be altered. Signal transduction systems also have a role in antibiotic resistance in P. aeruginosa. phoP-phoQ mutants showing increased resistance to aminoglycosides, cationic peptides, and polymyxin have been described . PhoP-PhoQ is a two-component regulatory system that has been studied mainly in S. enterica serovar Typhimurium. In this bacterial species, PhoP-PhoQ regulates the expression of at least 40 genes in response to magnesium concentrations and has an important role in both virulence and resistance to polymyxin . Although a role of P. aeruginosa PhoP-PhoQ in virulence has not yet been demonstrated, it is highly expressed under magnesium starvation , which indicates a role similar to that previously described in the case of S. enterica serovar Typhimurium. PhoP-PhoQ may then coordinately regulate both virulence and antibiotic resistance in response to the magnesium concentration. We have addressed some examples of cross talk between antibiotic resistance and virulence regulons. However, genes encoding both types of determinants might also be part of the same regulon. As previously stated, this is the case for the Mar regulon in E. coli and S. enterica serovar Typhimurium, which is involved in multiple antibiotic resistance , superoxide resistance , organic solvent tolerance , and bile salts resistance , among others. The same situation probably occurs for several other MDR determinants, so that the possibility that the physiological signals that bacteria receive during infection might trigger a global response including the regulation of both antibiotic resistance and virulence genes should be carefully analyzed. THE CROSSROADS OF VIRULENCE, EPIDEMICITY, AND ANTIBIOTIC RESISTANCE : The pathogenic relevance of a bacterium relies mainly on two different properties, virulence and epidemicity (the ability to produce epidemics). A compromise between the two properties is required to cause a major disease because a nonvirulent bacterium obviously will not produce any disease, whereas a nontransmissible bacterium might cause a dangerous disease but only in a limited number of people. Since bacteria are under antibiotic pressure during infection (see above), highly efficient pathogens must be not only virulent and epidemic but antibiotic resistant as well. The problem from the microorganisms' perspective is whether or not the acquisition of one of these traits could have a detrimental effect on the expression of the others. In such a case, an "optimization strategy" will be required. We will review whether the acquisition of these abilities is detrimental to the expression of the others and, conversely, reinforces the possibility of acquisition of a full proinfective repertoire, virulence, epidemicity, and antibiotic resistance. Are Antibiotic-Resistant Bacteria More or Less Virulent? : When antibiotic resistance occurs as a consequence of a genetic alteration in a housekeeping gene (for instance, those involved in ribosome function, cell wall construction, biosynthetic pathways, and the DNA replication machinery), conventional wisdom suggests that an evolutionarily optimized mechanism has deviated from its functional optimum. Therefore, resistance should have a "direct" cost in bacterial fitness, and resistant organisms should exhibit suboptimal behavior. The location of antibiotic resistance determinants in accessory genetic elements (plasmids, transposons, integrons) probably reduces this "direct" cost, but at the expense of the "indirect" cost of carriage of the elements themselves . In both cases, antibiotic resistance should have a fitness cost , which might reduce bacterial virulence. Although it is difficult to find examples of the effect of antibiotic resistance on the virulence of clinical bacterial isolates, epidemiological data support the idea that, in some cases, antibiotic-resistant organisms may show a decrease in pathogenicity. For instance, in urinary tract infections, fluoroquinolone-resistant E. coli are easily found in cystitis but are isolated at very low frequencies as causative agents of pyelonephritis . Recent studies also suggest that penicillin-resistant S. pneumoniae strains may be less pathogenic than susceptible ones, at least in animal models , and isolates from cases of bacteremia are frequently less resistant than those isolated from mucosal infections and even in carriers . In this case, a possible explanation is that the capsulation involved in serum resistance eventually interferes with acquisition of foreign DNA encoding resistant penicillin-binding proteins. Nevertheless, the pathogenic process might provide an advantage for the bacteria, such as permitting access to exclusive habitats. For instance, a reduction in the rate of protein synthesis because of a ribosomal resistance mutation may be deleterious for a bacterial organism competing with the wild-type susceptible strain in a rich medium, where the effect of such a deficiency is maximized, while under slow-growth conditions, the difference would be minimal. The expression of genes involved in pathogenicity is frequently regulated by stationary-phase signals , so that the slow growth due to a reduction in protein synthesis might eventually trigger the expression of genes involved in pathogenesis. In this case, the growth defect produced by antibiotic resistance could be compensated for by the increased virulence of the pathogen, and bacteria could easily accumulate compensatory mutations which restore their original pathogenic abilities totally or in part . Although antibiotic resistance usually reduces bacterial fitness, the opposite situation might also occur. For instance, the bleomycin resistance gene contained in transposon Tn5 confers improved survival and growth advantage on Escherichia coli . The increase in the virulence properties of antibiotic-resistant bacteria is a disturbing possibility, because in these cases, revertants that become susceptible by means of mutations and loss of antibiotic resistance plasmids may be at a selective disadvantage relative to the ancestor resistant population, so that the resistant population will prevail even in the absence of antibiotic selective pressure. In general, antibiotic resistance obviously increases the overall pathogenic potential of the organism in the treated patient, because the susceptible organisms are killed by the antibiotics and the infection does not progress. For instance, the mortality rate of S. pneumoniae meningitis in South African children is significantly higher for antibiotic-resistant strains . However, in the absence of antibiotics, antibiotic resistance may influence pathogenicity in different ways, depending on the antibiotic resistance determinants involved and the possible accumulation of compensatory mutations. Are Antibiotic-Resistant Bacteria More or Less Epidemic? : The term epidemic is usually restricted to widespread outbreaks of pathogenic and/or resistant microorganisms. Thus, the actual dimension of the link between transmissibility and resistance may be difficult to evaluate without extensive studies on the population biology of microbes associated with humans. Although the epidemiology of nonvirulent and nonresistant organisms remains to be evaluated, some important observations indicate that in countries such as Sweden, where antibiotic-mediated selection is low, epidemics of antibiotic-susceptible but pathogenic bacteria may be, as in the preantibiotic era, the usual case. Studies of commensal bacteria in wild animals also suggest that the gut flora of wild populations with little human contact (and thus no contact with antibiotics) shows very low levels of antibiotic resistance , whereas those from the guts of animals that have been in contact with humans show higher levels of antibiotic resistance . On the other hand, the most relevant increases in antibiotic resistance at the local and global levels are due to the spread of a relatively small number of resistant bacterial clones, such as with the "international" resistant clones of S. pneumoniae , methicillin-resistant S. aureus , and vancomycin-resistant Enterococcus faecium . The extreme rapidity with which some of these resistant organisms spread in remote countries suggests that their colonizing potential is nearly optimal for a broad range of hosts. Several reasons can be given in favor of the existence of an evolutionary link between antibiotic resistance and host-to-host transmission. The diminution in bacterial diversity due to antibiotic treatment will result in overgrowth of resistant bacterial populations that are in the minority under normal competitive circumstances. Antibiotic treatment is then a bottleneck that is crossed by the few antibiotic-resistant bacteria that are present in the population. However, once this bottleneck is crossed, the best colonizers among the remaining antibiotic-resistant bacteria will have an advantage for recolonizing the host. If the challenge is repeated several times, it could be expected that a certain within-host evolution of bacteria to optimize colonization will occur . Considering that all members of the host population are very similar in their ability to be colonized by each particular organism, success in colonizing the host will be reflected in a corresponding success in between-host transmission ability. Large numbers of resistant cells within the host should facilitate host-to-host transfer, particularly to other treated hosts. This perspective may have some exceptions. Some bacterial populations are strictly linked to particular environments, including a dependence on other local bacterial populations . That situation is not infrequent in mucosa-associated communities. In these cases, the antibiotic-resistant population may remain confined and even disappear under antibiotic therapy just because the other members of the community are eradicated. The possibility that a resistant population will be enriched under antibiotic treatment thus requires the ability of this species to independently exploit the available habitat. This ability is common among organisms capable of crossing ecological barriers, including the more epidemic ones. In serial-passage experiments with mixed populations, the strain with the greatest number of cells in the transferred inoculum has a selective advantage. The driving force may be within-host competition and selection for increased parasite growth rate . In this regard, the ecological abundance of specific clones in natural populations of Staphylococcus aureus might be linked to the virulence of such clones . The clones should then have been selected because they are good colonizers as well as being highly transmissible and highly virulent . This statement is true in general, except perhaps for organisms with high virulence, such as Mycobacterium tuberculosis, for which a small inoculum may be sufficient to cause the illness. Epidemicity ensures constant high multiplication rates that may be needed for the acquisition of resistance. Therefore, these bacteria are simultaneously more prone to becoming resistant, to be selected for by antibiotics, to increase their absolute numbers, and consequently to be transmitted efficiently. There is an obvious corollary: antibiotics should be prescribed in a way that ensures eradication of the bacterial pathogen (as said by Paul Ehrlich, "hit hard and hit early"), as any survival gives the organism an opportunity to evolve and spread in a more efficient way. The relationship between epidemicity and pathogenicity may also help to modulate the development of resistance. Within-host growth usually induces the expression of virulence determinants in several pathogens, thereby increasing their virulence. Increased between-host bacterial transfer may also increase virulence. It is also well known that serial passage of a given organism among susceptible hosts frequently increases its pathogenic power in the new host, although it is attenuated for the former host . Bacterial virulence might then be increased in densely clustered host groups (camps, schools, hospitals), with close contact and high transmission rates, particularly in the early stages of the spread. Antibiotics preferentially exert their selective pressure on clinically ill patients, who are treated more frequently than carriers. If extinction of the pathogen does not occur as a result of treatment, those epidemic strains will tend to be subjected to increased antibiotic pressure, and antibiotic resistance will be more easily achieved. Although we think that antibiotic resistance usually helps microorganisms in their transmissibility, this is not the case for all. For instance, comparative studies of mutations in animal isolates and experimental in vitro- and in vivo-selected mutants of Salmonella spp. suggest that in some cases, a counterselection of highly fluoroquinolone-resistant strains might occur in the field . This indicates that fluoroquinolone-resistant Salmonella spp. probably have a strong reduction in fitness outside of the host, so that they cannot be maintained in the environment. Alternatively, it could be that the compensatory mutations accumulated during infection are deleterious for bacteria growing in the environment (see above). The same probably occurs for opportunistic pathogens such as P. aeruginosa, Burkholderia cepacia, and Stenotrophomonas maltophilia. These bacterial species are intrinsically resistant to several antibiotics , but, during therapy, mutants with higher levels of resistance and displaying resistance to other antibiotics are selected . It is noteworthy that cross infection is not frequent for these opportunistic pathogens, and the resistant clones selected during therapy are not frequently disseminated. In the case of fluoroquinolone-resistant Salmonella strains, this suggests that these resistant opportunistic pathogens are probably impaired in their dissemination in natural environments (mainly water and soil). Of course, as we want to stress throughout this review, this is not a black and white situation, and outbreaks of nosocomial infections due to the transmission of resistant clones have been described. Relevant examples of this situation are vancomycin-resistant E. faecium , the selection of such antibiotic-resistant bacteria in avoparcin-exposed farms , and some outbreaks of infections with multidrug-resistant Acinetobacter baumanii . Are Virulent Bacteria More Resistant to Antibiotics? : Because virulent organisms are able to produce clinical symptoms of infection in human and animal hosts and should therefore be exposed more frequently to antimicrobial drugs, the risk of development of resistance is expected to be higher. That is expected to produce a link between higher virulence and higher resistance. Despite this prediction, the opposite is frequently the case. Pathogens such as Neisseria meningitidis, Salmonella enterica serovar Typhi, Shigella dysenteriae, Bordetella pertussis, Leptospira icterohaemorragica, and Brucella melitensis are rarely resistant to antibiotics. Some reasons can be given for this phenomenon. First, as stated previously, the possible cost (in fitness) of antibiotic resistance should be tolerable by the bacterial host. Usually this tolerance involves the acquisition of cost-compensating mutations that must be tolerated as well. This is not a critical problem for many opportunistic pathogens that are able to survive in various alternative environments. But in several highly specialized pathogens, the genetic structure cannot be extensively modified by mutation without risk of local extinction as a consequence of an increased number of lethal mutations . Second, in contrast to less-specialized pathogens, some of these highly virulent organisms have probably evolved in protected environments less open to competition and therefore less exposed to natural antibiotics. Third, the absolute number of many pathogens is reduced compared to their commensal counterparts, so that its chances of being exposed to antibiotics and acquiring antibiotic resistance should be lower. This view has been proposed as an explanation for the differences in the ability to develop mutational resistance to penicillins in S. pneumoniae and the viridans group streptococci versus S. pyogenes . Fourth, the particular niche occupied by highly pathogenic organisms may be excluded or almost excluded from the action of antimicrobial agents (see below), so that they are not subjected to antibiotic selective pressure. Are Epidemic Bacteria More Resistant to Antibiotics? : Epidemic bacteria are considered here as those able to spread in an efficient way between hosts in a short period of time (note that we do not mention whether epidemic bacteria are pathogens or not). This property is also frequently expressed as organisms having a "high reproductive rate" (R0) . By definition, epidemic bacteria are frequently encountered in a large number of individuals and thus reach high population sizes throughout the host community. Because these populations frequently face variable environments in terms of antibiotic selective pressure (patients under therapy and not, different therapeutic regimens), the chances for acquisition of antibiotic resistance increase . Additionally, bacteria with a high dispersal (migration) rate are in contact with more types of microorganisms occupying different ecological habitats from which antibiotic resistance genes can be acquired by horizontal transfer. These characteristics make epidemic bacteria more prone to acquire antibiotic resistance. Once an antibiotic resistance phenotype is acquired, an epidemic microorganism can be widely disseminated. In addition, antibiotic selective pressure in treated patients will convert such an organism into a predominant clone in this bacterial species . This is probably the cause of the worldwide distribution of antibiotic-resistant clones of species such as S. pneumoniae (see above). Epidemicity implies that the microorganism can colonize a large number of hosts, so that the overall number of bacterial cells to which selection can be applied is also large. Thus, the probability of selecting an antibiotic-resistant mutant is higher than in the case of nonepidemic microorganisms. On the other hand, if an epidemic organism becomes resistant in one host and is transferred to other hosts, the use of antibiotics in each treated host ensures the removal of the susceptible bacterial populations and its replacement by the epidemic population. At this stage, the antibiotic-resistant microorganisms increase in number, and all the possible variants within this population compete for establishment in the novel hosts. As a result of this competition, the best colonizers of the epidemic, antibiotic-resistant microorganism are now selected. Selection of epidemic and antibiotic-resistant bacteria is thus based on a feedback mechanism which links the two properties, unless the cost of antibiotic resistance severely impairs the effectiveness of host-to-host spread (see above). This cost will depend on the specific bacterial species-antibiotic resistance mechanisms. In this way, studies which integrate virulence properties and the behavior of antibiotic-resistant microorganisms in the environment are needed to establish the best protocols for avoiding the spread of virulent antibiotic-resistant bacteria. FIG. 3. | Epidemic bacteria can easily acquire an antibiotic-resistant phenotype. Epidemic bacteria can easily acquire an antibiotic-resistant phenotype. An epidemic bacterial strain (green) spreading between hosts (each individual blue square) increases its population size, facilitating the emergence of antibiotic-resistant mutations (yellow arrow) and the acquisition of resistance genes by horizontal transfer (brown arrow). Acquisition of resistance is immediately amplified by the epidemicity of the strain, even in the absence of antibiotic selection (resistant variants, yellow and brown). Interactions between different mechanisms of resistance may occur, eventually leading to multiresistance (red). If antibiotics are almost ubiquitously present in the hosts, as in an intensive care unit (black squares), the spread of the resistant bacteria is favored. RESISTANCE AND VIRULENCE: AN INTEGRATED VIEW : The human host has acquired several different mechanisms in the course of evolution for fighting infections, including the immune system, fever, production of antimicrobial peptides , and iron-scavenging , among others . Another mechanism of defense against colonization and further infection is the presence of an indigenous microflora which has coevolved with its host . This defense constitutes the very first biological barrier against subsequent colonization and infection with external potentially pathogenic microorganisms. Very recently in terms of evolutionary time, a new line of defense has been added to the armamentarium of humans in their fight against infection, that being antibiotic therapy. Based on the effect that they produce on infectious bacteria, pharmaceutically produced antibiotics play the same anti-infectious role as antimicrobial compounds produced by either the host or its allied microbiota. All these types of substances are able to inhibit bacterial growth and thus constitute a part of the defense of the human host. The only difference between these substances resides in the fact that conventional (industrial) antibiotics are a product of the so-called "cultural evolution" , whereas the other mechanisms for defense against infections are the result of the natural "genetic evolution" of the human species. In opportunistic infections, where the genetic defense strategy is severely impaired, the only line of defense that remains is the antibiotic cultural defense. With this approach, antibiotic resistance is hard to distinguish from other mechanisms involved in microbial pathogenicity. Resistance as a Colonization Factor in the Treated Host : Acquisition of an antibiotic resistance phenotype will have two effects on bacteria during colonization and infection of the treated host. First, resistance is required for evading cultural host defenses (in this case, antibiotic treatment). Therefore, it is equivalent to other systems for evading other host defenses, such as production of siderophores , resistance to oxidative stress , resistance to defensins , and resistance to bile salts for intestinal pathogens. These characteristics are needed, although not sufficient, for the development of infection. Second, resistance in the presence of antibiotics in the treated host will favor colonization. As stated previously, the very first line of defense against infections is the indigenous microflora , which form a colonization barrier against exogenous microorganisms. However, this microflora is severely affected by antibiotic therapy. Therefore, antibiotic therapy debilitates the defenses of the host by altering the protective biological defenses. Only if the infecting microorganism is resistant to antibiotics can it make use of this advantage for better colonization of the host. We might thus conclude that antibiotic treatment promotes further colonization and infection by antibiotic-resistant bacteria which otherwise may not be able to effectively compete with indigenous microflora. In this way, antibiotic resistance is a relevant colonization factor for the microorganisms interacting with the treated patient. In fact, epidemiological studies have demonstrated that antibiotic therapy is a relevant risk factor for colonization by antibiotic-resistant bacteria such as S. pneumoniae, methicillin-resistant Staphylococcus aureus, Stenotrophomonas maltophilia , Acinetobacter baumannii , and Campylobacter jejuni . Opportunistic Infections : Opportunistic infections occur in individuals whose natural defenses against bacterial infection are severely impaired . In the preantibiotic era, these people were mainly infected by microorganisms belonging to their own microflora. Currently, even when the genetic defense has collapsed, the second line of defense, antibiotics, keeps these hosts free from infection, but obviously only if antibiotic-susceptible organisms are involved. As research on antibiotics has been focused for years on frequent human pathogens, many antibiotics are still active on microbes associated with humans. Because of this, many opportunistic pathogens today come from the environment, with one of their more salient characteristics being high intrinsic resistance to a broad range of antibiotics . Indeed, it is this ability that allows opportunistic pathogens to invade the compromised host under antibiotic therapy, since environmental opportunistic pathogens do not need to harbor specific pathogenicity determinants for humans (as the natural defense has been broken), but only antibiotic resistance. In fact, opportunistic pathogens are unable to produce infections in nondebilitated patients. Therefore, environmental and clinical isolates of opportunistic pathogens such as P. aeruginosa and Stenotrophomonas maltophilia are genetically indistinguishable. Analysis of two collections of P. aeruginosa isolates from environmental and clinical settings has demonstrated that they were functionally equivalent in traits relevant for both their pathogenic (host-oriented) and biodegradative (environment-oriented) abilities . The healthy host is quite an "extreme" environment in terms of stringent growing conditions for microorganisms. Although few bacteria have evolved to efficiently colonize this environment, many more are potentially able to grow using the resources of human body (for instance, those involved in corpse decomposition and putrefaction). In other words, the dead body implies a rupture of some ecological constraints of the living one. Also, because conditions for bacterial growth are less stringent in the ill host , a greater variety of bacteria can eventually produce an infection. However, when antibiotic treatment is begun, there is a new ecological constraint on the growth of the bacteria, the antibiotic, so that under these conditions only antibiotic-resistant bacteria will be able to maintain the infection. Finally, because opportunistic pathogens are not the result of evolution of an infectious bacterial species in contact with the host, the term "emergent pathogens" has been inappropriately used for their description. These organisms, having acquired the ability to colonize different habitats, can infect the sick host. Their selection is the result of cultural human evolution, since antibiotic treatment has selected a previously naturally resistant bacterial species. The replacement of traditional opportunistic pathogens with novel, more resistant ones has occurred over the last few years . Thus, antibiotic resistance linked to antibiotic therapy might be considered the primary factor influencing the selection of such novel antibiotic-resistant opportunistic pathogens. FIG. 4. | Colonization space of opportunistic pathogens. Colonization space of opportunistic pathogens. Colonization space can be defined as the combination of the different physicochemical parameters (including the space) in which an organism can survive. In this regard, stringent growth conditions (for instance, absence of oxygen) reduce the ecological space. As discussed, conditions for infection are quite stringent, so that only a few bacteria can colonize the host's ecological space (dark gray square) compared to bacteria that colonize the environment and, eventually, the dead host (lightly shaded square). In the case of the sick host, however, stringency for colonization is lower, so that some environmental bacteria can now colonize this larger ecological space (intermediately shaded square in panels b and c). These bacteria with an environmental origin are opportunistic pathogens. Once infected with these bacteria, the host is usually treated with antibiotics. Again, antibiotic treatment restricts the ecological space so that only antibiotic-resistant bacteria can grow under these conditions (oval in C). In all cases, small ovals indicate human-associated bacteria and small circles indicate environmental bacteria. White, antibiotic-susceptible bacteria; black, antibiotic-resistant bacteria. Does Virulence Affect Exposure to Antibiotics? : As previously stated, virulent organisms producing clinical symptoms are more frequently treated with antibiotics than nonvirulent ones, and therefore virulence should increase antibiotic exposure. Antibiotic treatment will affect both the infecting bacteria and indigenous microflora. However, pathogenic bacteria will be treated as soon as they are detected (and their presence suspected from clinical symptoms), and thus antibiotic selective pressure is exerted quite frequently on these organisms. Indigenous microflora are challenged with antibiotics only as an unwanted side effect of the therapy directed against the pathogens, so that most time is maintained without antibiotics. Nevertheless, several bacterial pathogens may behave as commensals over long periods of time. Many bacterial infections are controlled by the host in many cases without clinical symptoms . In this case, even pathogenic organisms may not be frequently challenged by antibiotics. In some cases, host lesions induced by pathogenic microbes may occur but with few if any clinical symptoms (subacute infections). In these cases, these organisms with "low virulence" will also be less exposed to antibiotics. One might suggest that antibiotic therapy could be considered a cost and a penalty associated with virulence. If this were the case, it is conceivable that some pathogens could evolve to low pathogenicity after prolonged use of antibiotics in human populations. In general, close relatives to pathogenic bacteria are now members of the normal flora. Are they the result of the evolution from pathogen to harmless variant to escape host defense mechanisms? In particularly severe infections, the host may quickly die, reducing bacterial exposure to antibiotics. Once the host dies, the microorganisms may either die as well (if they need to grow inside the host) and be released into the environment, where the antibiotic pressure is low. It is thus possible that the most virulent bacteria are perhaps less exposed to antibiotics than nonvirulent ones. We can conclude that bacteria with intermediate levels of bacterial virulence have a greater probability of being exposed and developing resistance to antibiotics than less virulent and highly virulent organisms , in the first instance because they are not always detected and thus not treated, and in the second, because the host dies without giving them enough time for the development of antibiotic-resistant variants. Finally, epidemic organisms may evade a short antibiotic exposure because, by definition, they move to another host (usually nontreated) more easily than a nonepidemic microorganism. Thus, mechanisms facilitating between-host transfer may in some cases reduce exposure to antibiotics, particularly if the organism is not highly virulent. In the long run, however, epidemic organisms will confront higher levels of antibiotics, as the hosts are treated after the infection is detected (see above). FIG. 5. | Infectious bacteria under risk of antibiotic treatment. Infectious bacteria under risk of antibiotic treatment. As stated in the text, in some cases infections are extremely acute (black line), so that the host is damaged (and even killed) before an antibiotic treatment is implemented. Other infections present no symptoms and have subclinical manifestations, so that they are not treated even if they occur for long periods of time (light gray line). Only infectious bacteria that produce clinically relevant symptoms and are present long enough to be treated (dark gray line) are under antibiotic selective pressure. In host spatial location and exposure to antibiotics. : The location of microorganisms in tissues during infection is an important factor in their exposure to antibiotics. In some cases, organisms are refractory to treatment even if they are classified as susceptible by standard in vitro tests . Virulent organisms may invade spaces in which the achievable concentration of antibiotic is low, such as in otitis, sinusitis, and meningitis . Also, the pharmacodynamic properties of antibiotics influence in a critical way the inhibitory and killing effects of these drugs. For instance, if a given concentration of a beta-lactam is not maintained in contact with the organism over a critical period of time, the antibacterial effect (predicted by the AUC/MIC ratio and time above the MIC) will be reduced. In other cases, there is a critical concentration of antibiotic (Cmax) that is affected by changes in the medium during infection. For example, low pH is characteristic of inflammatory fluids as well as some intracellular compartments . Several antibiotics are less active in acidic conditions, so that the environment of the infection site may contribute to lowering the antibacterial effect of antibiotics . As previously discussed, several bacterial pathogens can invade host cells and are thus less exposed to antibiotics . Biofilms (see above) also produce an antibiotic resistance phenotype due at least in part to the structure of the biofilm, which precludes the efficient transfer of antibiotic to all the bacteria present in the biofilm . More recently, it has been described that resting cells are as resistant as bacteria growing in biofilms, and the effect of biofilms on antibiotic susceptibility might be due to a stationary-like growth phase of the bacteria growing in the biofilms Obligate intracellular pathogens such as Chlamydia, Lawsonia, and Rickettsia spp. are infrequently resistant to antibiotics. For these pathogens, the MICs of antibiotics such as tetracycline and erythromycin that are distributed either extracellularly or in intracellular compartments that are different from those in which bacteria are located have remained constant over decades This suggests that some intracellular compartments serve as refuges against antibiotics. It is obvious that intracellular location also prevents easy acquisition of resistance elements such as plasmids and transposons, from other organisms. These elements may occasionally be present in bacteria, as Coxiella and Listeria spp., which do not have an absolute requirement for an intracellular location for growing. Other obligate intracellular pathogenic organisms such as Brucella spp. are infrequently resistant to antibiotics and rarely contain plasmids. Comparison of the genomes of different bacterial species has demonstrated that in organisms with small genomes, horizontal gene transfer is very infrequent . It has been suggested that because of the extreme specialization of the small genome of intracellular organisms, any new gene or gene modification may disrupt the delicate equilibrium of vital genes, leading to extinction of the microorganism . In other words, resistance genes might reduce the viability of intracellular organisms. Moreover, the altered metabolism of these organisms might reduce their ability to utilize resistance genes. For instance, the energy needed for expression of resistance to tetracyclines and macrolides by efflux mechanisms may not be accessible in organisms dependent on host cell energy sources . In general, intracellular bacteria might tolerate only mutational changes that lead to energy-independent mechanisms of resistance and not reduce intracellular fitness. The possibility that some bacteria could have developed mechanisms of intracellular invasion in the environment, as an escape from natural antibiotics, cannot be discarded. Facultative intracellular pathogens such as Legionella spp. have environmental protozoa as natural hosts, and the pathogenicity of these organisms in humans may only reflect an accidental change leading to the invasion of another type of eukaryotic cell . Something similar may have occurred in Listeria. Interestingly, acquired antibiotic resistance is a very unusual event in both genera, which indicates that intracellular location is a effective way to escape from antibiotics. In summary, we want to stress that the location of an organism in a particular body site is influenced by its colonizing and pathogenic abilities. Therefore, these properties determine the cell-antibiotic interaction. On the other hand, this interaction is critical for the eventual development of antibiotic resistance. Normal microbiota versus pathogenic bacteria. : The normal microbiota is the first line of defense against infections due to its role as a colonization barrier . However, once colonization occurs, normal bacterial flora might also protect virulent bacteria from antibiotic action. Protection of a susceptible population by a resistant one has been described in vitro for beta-lactam antibiotics . Beta-lactamase-producing bacteria rapidly inactivate beta-lactam antibiotics. If a mixed population of beta-lactamase producers and nonproducers is exposed to beta-lactams, the nonproducers will begin to die. However, once the antibiotic is inactivated by the beta-lactamase producers, the remaining susceptible bacteria that have not been killed by the beta-lactam can grow. This effect is frequently seen in the laboratory, when ampicillin-susceptible colonies exhibit satellite growth around an ampicillin-resistant clone in an ampicillin-containing plate. It has been demonstrated that minute amounts of beta-lactamase-producing strains can protect fully susceptible strains from the action of beta-lactams . During infection, the same situation might occur in vivo. Antibiotic-resistant populations, either as normal colonizers or in the case of coinfection by different bacteria, can then protect virulent, susceptible bacteria from antibiotic action. This effect has been claimed as relevant when beta-lactam therapy failed to eradicate fully susceptible bacteria from colonized environments . Normal microbiota can also contribute to the acquisition of antibiotic resistance determinants by bacterial pathogens because in vivo transfer of antibiotic resistance determinants has been reported . Thus, acquisition of antibiotic resistance genes by normal, nonpathogenic microbiota is a concern because the final outcome of antibiotic resistance in bacterial pathogens might be compromised. Indeed, the possibility of maintaining hospitals with a reduced proportion of resistant bacteria depends on the local reduction the antibiotic selection, but also on the dilution of the resistant organisms of the hospital caused by the admission of patients from the community with susceptible normal flora . POPULATION STRUCTURE : The population structure of clinically relevant bacteria is an important epidemiological topic . If only one or a few clones of a bacterial species are relevant for infection and antibiotic resistance, efforts must concentrate on these clones. Also, preventive strategies can be more easily designed if the prevalence and spread of the various clones involved in disease are known. A more comprehensive analysis of the infection must involve the study of bacterial populations from both the ecological and evolutionary perspectives. This analysis requires the use of molecular biology tools for analyzing the structure and evolution of bacterial pathogens. Bacterial Variation Facing Stressful Environments : As was stated above, bacterial adaptation is essentially the consequence of mutation and gene recombination. Mutational events are particularly important when a large population of cells is confronted with critical, abrupt, and nonspecific changes in the environment, eventually permitting rapid but not always optimal adaptation. In fact, selection of bacteria with a hypermutator phenotype (see below) has been suggested to be a good adaptive strategy to sudden environmental changes , and mutational adaptation requires large bacterial populations , as this evolutionary strategy implies the death of most members of the population . On the other hand, recombination allows the acquisition of several different traits in a single step . In this regard, horizontal transfer of genes and even large chromosomal regions is frequent in all bacterial species , contributing to the acquisition of several physiological functions in a single step (see above). However, recombination between different strains of the same bacterial species occurs frequently in only a subset of bacterial species that are highly specialized to grow in constant and frequently in unique habitats (such as Helicobacter pylori, Neisseria meningitidis, Neisseria gonorrhoeae, Haemophilus influenzae, and S. pneumoniae). In these cases, recombination between members of the same species is important for combining the beneficial mutations present in the entire bacterial population and eliminate the deleterious ones. In fact, it has been stated that a high recombination rate could be selected when several mutations (present in different members of the population) are needed for adaptation . Bacterial adaptation can increase under conditions of stress such as infection and antibiotic selective pressure. In the last few years, it has been shown that bacteria have stress-regulated systems which allow them to mutate at a higher rate under stress conditions. Two different situations are relevant for increased mutation rates in vivo in bacterial populations. Stress-induced transient increases in mutation rates. : Bacteria increase their mutation rate as a consequence of stress, such as starvation , antibiotic challenge , and possibly during infections; this process is termed adaptive mutation . A situation of stress triggers a bacterial response that involves changes in the expression of several proteins. These changes are a consequence of the recognition by the bacteria of "stress signals" that activate a global response based in the concomitant transcriptional regulation of several genes . Some aspects of these responses are specific to the type of stress involved, but other changes are common to different stress situations. Little is known about the bacterial responses to pathogenic and antibiotic-induced bacterial stresses. However, it is conceivable that these types of stress trigger responses similar to those in other bacterial stress situations. For instance, treatment with antibiotics at bacteriostatic concentrations provokes a stationary phase-like response, and it has been demonstrated that subinhibitory concentrations of antibiotics targeting the ribosome induce the synthesis of either heat shock or cold shock proteins, depending on whether the A site of the ribosome was empty or occupied . In any case, adaptive mutation to stress is a regulated phenomenon by means of which bacteria can increase their mutation rate severalfold . The primary factors in this process are stress-responsive, error-prone DNA polymerases V (umuCD) and IV (dinB), whose expression transiently increase the rate of bacterial mutation . Eventually, bacterial stress over the pathogenic process may increase the probability of emergence of antibiotic-resistant mutants; conversely, antibiotic stress may increase bacterial adaptation to the host. Hypermutable (mutator) strains. : Hypermutation can be the consequence of a regulated increase in mutation rates (see below) or the presence of subpopulations of "mutator" bacteria within "normal" bacterial populations . Some works have demonstrated the presence of such mutator populations in infecting bacteria . These cells have mutation rates that can be increased as much as 10,000-fold, usually as the consequence of inherited defects on the methyl-directed mismatch repair system . These strains should have increased abilities to survive antibiotic challenge and antibacterial host defenses. Among strains lacking the methyl-directed mismatch repair system, the ability to recombine (homologous recombination) should increase , which may facilitate the acquisition of both antibiotic resistance and bacterial virulence in the same organism. Mutator strains are supposed to be less fit than wild-type strains outside the stress situation (adaptive bottleneck). Induction of a transient hypermutation under stress serves to generate mutants which can escape from transient stress situations, whereas stable mutator strains are fixed in the populations only when bacteria are subjected to frequent bottlenecks. This situation probably occurs in chronic infections, such as those associated with cystic fibrosis. In fact, P. aeruginosa isolates from cystic fibrosis patients frequently present a mutator phenotype, a characteristic feature that does not occur in P. aeruginosa isolates obtained from blood and sputum cultures of non-cystic fibrosis patients . Recently, the fitness of mutator E. coli strains has been analyzed in an in vivo model system . A high mutation rate was shown to be beneficial because it allowed faster adaptation. However, the benefit disappeared once adaptation was achieved, because mutator strains accumulated mutations that, although neutral in the analyzed model system, can be deleterious in a secondary environment. Thus, mutator strains may have short- and long-term advantages that justify high competitiveness of mutator bacteria during infection, but low competitiveness during transmission and recolonization of other hosts. In summary, stressful growth conditions in the host, particularly under antibiotic therapy, are important elements for the evolution of bacterial populations, influencing gene transfer and increasing mutation rates. Stress associated with the pathogenic process and antibiotic exposure not only increases bacterial variation, but also strongly selects particular bacterial variants. In this sense, pathogenesis and antibiotic exposure can be considered "accelerators" of bacterial evolution. Clonal Structure : Molecular epidemiology has demonstrated that a few clones may account for the worldwide dissemination of virulent and antibiotic-resistant bacteria for several bacterial species. This view of the genetic structure of bacterial species is known as the clone concept . Further analysis has demonstrated that the degree of clonality varies depending on the species analyzed. In that way, pathogenic species such as E. coli and Salmonella spp. are highly clonal, whereas species such as Neisseria gonorrhoeae present a very high rate of recombination and have a panmictic (species that are not subdivided into discrete phylogenetic lineages ) genetic structure . Even for microorganisms with a predominantly clonal structure, analysis of complete genome sequences has demonstrated that horizontal transfer due to recombination is frequent in bacterial populations . The high rate of recombination in several bacterial species might put in question the validity of the epidemiological analysis of clinically relevant bacteria under the clonal structure approach. However, even for bacteria for which recombination occurs very frequently, this approach can be extremely useful. When analyzing the genetic structure of these bacterial populations, evolution can be traced over time, following the routes of transmission. For highly epidemic species such as Neisseria meningitidis, clones emerge that disseminate further worldwide. Vaccination against the most frequent serotypes may produce the replacement with less common ones (see below). Also, the replacement of antibiotic-susceptible with antibiotic-resistant clones has occurred, and these replacements are easily followed by using molecular typing methods . In this case, a transient "quasi-clonal" population can be observed at a fixed time depending on the geographic localization . This situation is the consequence of the emergence of clonal lineages which are more proficient in their epidemicity, virulence, and antibiotic resistance than others. The more successful lineages will then predominate. However, once they become the predominant population, they will produce novel divergent lineages as a consequence of recombination with other clones and may then expand in different environments. The presence worldwide of different environmental habitats for bacterial populations guarantees this divergence, as has been demonstrated in different model systems . The clonal hypothesis is also relevant for understanding the evolution of genetic elements that are transmitted in bacterial populations. In fact, the variability of antibiotic resistance genes with clinical relevance is not very high, which indicates a "clonal" origin for several. Thus, most antibiotic resistance genes currently acquired by horizontal transfer were originated by transfer of an ancestor gene to a single or a few lineages of bacteria and then disseminated among bacterial populations. For instance, it has been demonstrated that the gene mecA, responsible for resistance to methicillin in S. aureus, was acquired only once by this bacterial species . Something similar has probably occurred for widely distributed genes such as plasmid-borne beta-lactamase genes of the TEM group . Even for highly recombinogenic bacterial species, a clonal origin of their genetic resistance determinants can eventually be traced. An example is the evolution of the genes encoding the synthesis of penicillin-binding proteins in Neisseria and Streptococcus spp. The analysis of pbp genes from different isolates has demonstrated that they are very uniform among penicillin-susceptible strains of N. meningitidis, N. gonorrhoeae, and S. pneumoniae, whereas those from penicillin-resistant strains consist of a mosaic structure that arose from the horizontal transfer of blocks of DNA from other, more naturally penicillin-resistant, commensal species by genetic transformation . Analysis of beta-lactam-resistant populations has demonstrated that these novel pbp genes have occasionally been selected, as single gene "clones," and afterwards, either the antibiotic-resistant bacterial clone or the pbp "resistant gene clone" disseminates and evolves further among the bacterial population. The same notions applied for analyzing the evolutionary biology of bacterial populations can be applied when analyzing accessory genetic elements (mainly plasmids and transposons). The most successful plasmid will be one with a good compromise between plasmid "epidemicity," defined as the ability to transfer between different species (number of different species or strains to which the plasmid can be transferred), "infectivity," defined as the ability to infect a specific host (frequency of entrance in the bacteria); and presence of selective traits (see above). As discussed above, antibiotic resistance may reduce bacterial fitness , and the same occurs with the acquisition of plasmids . If the fitness reduction is too high, the resistant bacteria will be unable to compete with the wild-type susceptible ones. Thus, most probably, we only detect those antibiotic resistance genes that do not produce an unaffordable fitness reduction on bacteria. We think that there are probably many more antibiotic resistance genes than those currently detected in bacterial populations. As stated previously, accessory genetic elements of clinically relevant bacterial populations also follow the rules of the clonal hypothesis. The same vir plasmids are found in all members of a single Yersinia species and are quite similar when one species is compared to another . This indicates both the clonal origin for these determinants and their successful maintenance in Yersinia populations for very long periods of time. Also, dissemination of epidemic antibiotic resistance plasmids is very frequent, and the acquisition of novel genetic determinants by a clonally originated plasmid has been documented . In fact, analysis of bacterial collections from the preantibiotic and postantibiotic eras demonstrated that they contain plasmids belonging to the same families, although the ones from the preantibiotic era do not contain antibiotic resistance genes . These results indicate the ability of plasmids to evolve as evolutionary units through the acquisition of antibiotic resistance determinants. The clonal hypothesis is then a useful tool for analyzing relationships between bacterial strains as well as mobile elements and even between genes involved in clinically relevant phenotypes. The use of this analysis of the lineages of such different elements can give a complete picture of the evolution of bacterial pathogens in response to novel selective pressure (such as antibiotic treatment). This approach is also relevant from the epidemiological point of view and can offer new light for understanding the future evolution of bacterial pathogens in terms of virulence, epidemicity, and antibiotic resistance. The application of the clonal hypothesis to the study of both organisms and genes will eventually serve to explore the hypothesis of possible convergent evolution of antibiotic resistance and bacterial pathogenicity, eventually leading to increasing numbers of organisms that are both resistant and virulent (see below). Resistance and virulence decrease bacterial diversity. : Virulent bacteria must face several environmental changes over the course of infection. To infect the host, bacteria must enter the organism, traverse through different epithelia, avoid the action of different host defense mechanisms against infection, and finally reach the target organ(s), where they produce the disease . It is intuitively appealing that the challenge of bacterial populations by these successive selective bottlenecks should reduce genetic diversity, at least in the short term. Changing environments are very stringent conditions for evolutionary adaptation, so that only a few organisms can survive those changes. As stated in the legend to Fig. , this is the case for the process of infection. The diversity of commensal bacteria in humans is extremely high, which shows that many different bacteria are adapted to grow on each of the different host surfaces (intestine, skin, oropharynx). On the contrary, the diversity of virulent bacteria is much lower . Virulent bacteria need to traverse and eventually colonize different environments as well as survive stringent and changing conditions (see Fig. ). The requirement for multiplication under all these stringent conditions might explain the low diversity of very specialized pathogens. A release from some of the stringent conditions required for growth in the healthy host will increase the possible diversity of potential pathogens. In this regard, a diverse array of opportunistic pathogens are able to infect the immunosuppressed and severely ill host, which presents a much less stringent environment (see above). A stringent selection also occurs for antibiotic resistance. In this case, rapid changes in the environment, produced by antibiotic exposure, rapidly lead to lethal conditions of selection. Only resistant clones will survive, replicate, and subsequently colonize the previously complex ecosystem. In fact, it is well known that many antibiotic treatments reduce the bacterial diversity of the human-associated microflora. Selection is even stronger with conditions of combined and successive antibiotic therapy, because bacteria are challenged by simultaneous and consecutive lethal bottlenecks. Therefore, only a few types of bacteria will survive, and diversity will be strongly reduced as a consequence of the treatment. Taxonomic Implications of Changes in Resistance and Virulence : The development of antibiotic resistance and the emergence of new pathogens have two implications for the taxonomy of microorganisms. Misclassification as new species. : The phenotypic changes associated with the acquisition of an antibiotic-resistant phenotype may alter the characteristics of the well-known human pathogens, producing mistakes in the classification of these microorganisms. For instance, optochin-resistant S. pneumoniae could be easily classified with the viridans streptococci , and penicillin-resistant pneumococci manifest major changes in peptidoglycan structure . If classic approaches to phenotypic identification are used without knowledge that these bacteria are antibiotic-resistant derivatives of pneumococci, a new species defined as "an antibiotic-resistant gram-positive organism, very close to S. pneumoniae but with some different features, such as resistance to beta-lactams and a characteristic peptidoglycan structure" could have been described. Emergence of novel microbial species. : The emergence of novel virulent species may occur by the acquisition of pathogenicity islands, which changes the pathogenic characteristics of the bacterial species. For instance, we know now that Y. pestis, the cause of plague, is a recently emerged clone of the much less virulent Y. pseudotuberculosis . Also, acquisition of a new host specificity could lead to novel species, such as occurs with the human immunodeficiency virus (for instance, from chimpanzees to humans ) and several mycobacteria (from ruminants to humans [, ]). In fact, it has been suggested that speciation of M. tuberculosis occurred 15,000 to 20,000 years ago at the time of ruminant domestication. Finally, new bacterial species (and strains) for the moment unknown might emerge linked to novel diseases. For instance, enterohemorrhagic E. coli has been recognized as a relevant health problem just in the last two decades. However, the large divergences found upon comparison of the sequence of the whole genome of the enterohemorrhagic E. coli strain O157:H7 with the nonpathogenic laboratory strain K-12 demonstrates that these strains diverged a long time ago. FUTURE EVOLUTION IN AN ECOLOGICAL PERSPECTIVE : Major Factors | The future evolution of bacterial virulence is expected to be shaped by at least a dozen major factors. TABLE 3 | Factors that might influence the future evolution of virulence and antibiotic resistance Sanitation and hygienic measures (mainly drinking water and sewage management). : The rise in the world population coupled with demographic changes (megacities, displacement of the population from the countryside to the cities, increased numbers of refugees) produces an increased requirement for noncontaminated water reservoirs. Control strategies must be implemented to guarantee the quality of water and food resources used in agriculture and animal and human consumption in order to prevent the emergence and dissemination of potentially relevant novel diseases. Preservation and eventual bioremediation of normal ecosystems in human, animal, and environmental microbiota. : Industrial contamination produces severe changes in natural ecosystems. Heavy metals, pesticides, solvents, and even antibiotics are among the most conspicuous contaminants. The effect of this type of contamination on indigenous bacterial populations has been little analyzed; however, the above-mentioned deleterious compounds may have a profound impact on natural microbial populations, including the selection of antibiotic-resistant variants , increased recombination, and, in general, qualitative changes in the microbial species and clones that colonize different habitats. Bioremediation of contaminated environments may help to reduce the potential effect of pollution on microbial ecosystems . Number of contacts between microbes from infected and noninfected hosts (human and animal) and between infected people and the environment. : The demographic structure of the human population increasingly facilitates urban crowding, favoring host-to-host contact and the dissemination of infectious diseases among the population. Global environmental changes. : Global environmental changes, including increases in warming and UV exposure, altering the biogeographical limits of bacterial species and thus mixing them with new hosts might serve to accelerate bacterial evolution and increase the number of diseases with previous geographic restrictions. For instance, global warming might increase the prevalence of tropical infectious diseases in areas such as the United States and the European Union. Human sociological movements. : Human sociological movements, including commercial interchanges , emigration, travel, war, and establishment of underdeveloped social islands in developed societies might increase contact between people infected with pathogens with a previous closely restricted geographical allocation (see the example of Yanomami Indians in reference ). Increase in susceptible hosts. : The increase in the number of elderly people, immunosuppressed patients, and children, particularly if they are densely clustered (day care centers, nursing homes), will increase the role of opportunistic pathogens in infectious diseases. New technologies. : New aggressive medical technologies for diagnosis and therapy (not anti-infectious) might reduce the incidence of pathologies such as cancer and diseases of the elderly at the cost of increasing the problem of infections due to opportunistic pathogens . Intensive farming. : Intensive farming, which reduces the diversity of animals (the same animal breeds are used in different farms worldwide), might also reduce the diversity of microorganisms colonizing such animals, producing an ecological alteration in the populations of microbes that can come into contact with humans . Also, the use of antibiotics for farming purposes might increase the emergence or spread of antibiotic resistance . Viral and virus-like diseases. : Viral and virus-like diseases affecting humans and animals, such as AIDS, might evolve and alter the host response against infections. New antimicrobial strategies. : The discovery and use of new antimicrobial strategies, including antibiotics, antivirals, conventional vaccines, pro- and prebiotics, antipathogenic therapies, and modulators of biological response modifiers in the infected host, may be useful for fighting infectious diseases. Vaccination. : New technologies in vaccination (in some cases food-associated; see below) and gene or sequence therapy may be useful for reducing susceptibility to some infections . Bioterrorism. : Eventually, bioterrorism will affect antibiotic resistance and virulence, with dangerous and unpredictable consequences . Interestingly, all these factors will also influence antibiotic resistance. Over the last two decades, resistance to antibiotics has crossed the limits of the patient and hospital to become a global environmental phenomenon in which infectious diseases of wildlife and domestic animals might have a deep impact on the emergence and further spread of new human pathogens . Everything that promotes bacterial spread among different natural ecosystems and among different hosts (protozoa, animals, plants, and humans) will also potentially affect the spread of resistance. A common theme resulting from all these factors is the possible progressive reduction in bacterial diversity by alteration of the complexity of current ecosystems (see above). In this regard, intensive farming procedures may have consequences for the future evolution of virulence and resistance. In fact, farming procedures reduce microbial diversity . The collapse in the diversity of animal and plant natural variants (races) within agronomically relevant species has already started, because only those variants with high productivity rates are being used worldwide. In fact, most crops are obtained from the seeds produced by a few companies. The result of animal cloning procedures, by which the offspring will be genetically identical to the progenitor, will produce an even larger reduction in the biodiversity of farming animals. Genetic diversity among hosts is probably a natural barrier to infections, since multiple-passage experiments have shown that adaptation to a new host frequently involves attenuation for the former one . From this perspective, a reduction in host diversity will allow bacterial organisms infecting a particular host to spread with greater efficiency to other closely related hosts (identical in the case of clonal farming), eventually increasing virulence. The effect of a possible change of host (for instance, from animals to humans) may result in bacterial attenuation for the former animal host with an unexpected increase in virulence for the new human host . The economic and health consequences of these possible changes in bacterial transmission have not been evaluated. The possible outcome is an increase in certain bacterial populations that will alter their potential virulence as a result of interactions with new hosts and then become targets for antibiotic intervention. If antibiotics are used over the next 50 years as they have been used in the previous 50 years, there is no doubt that the number of replicons and genetic elements carrying antibiotic resistance genes will increase continuously on Earth, and the cost of harboring these genes and vectors will probably be minimized , particularly in the more common bacterial populations, thus ensuring their perpetuation in the microbial world. Strategies against Virulence May Reduce Antibiotic Resistance : Antivirulence strategies tend to reduce the number of bacterial pathogens and may act on the expression of pathogenic mechanisms. In some cases, if the pathogenic mechanism is essential for the lifestyle of the bacteria, the elimination of the pathogenic factor will reduce the overall presence of the pathogen. For instance, nonsystemic toxins, such as necrotic toxins and hemolysins, contribute to the successful growth of the bacterial pathogen inside the host, so that antitoxin vaccination will not only reduce the toxic effect of the toxin, but also contribute to eradicating the bacterial species. In this regard, strategies against virulence may reduce antibiotic resistance, because lower number of pathogenic bacteria implies less antibiotic exposure, a smaller number of antibiotic-resistant mutants, and fewer possibilities of acquisition of antibiotic resistance genes from other bacterial pathogens. Reducing host-to-host transmission: hygiene and vaccination. : Any strategy that leads to a reduction in the number of infections in the host population will reduce the probability of emergence of antibiotic resistance for two reasons: a reduction in the number of bacteria under antibiotic selective pressure, and a reduction in the number of selective environments (different hosts). This results in a decreased probability of selecting different antibiotic-resistant mutants . In this regard, hygiene and vaccination are among the best systems for avoiding infections and resistance. The use of conjugate vaccines against those S. pneumoniae serotypes that are more frequently resistant to antibiotics have significantly decreased the rate of penicillin resistance in closed communities . As predicted by mathematical models, this decrease is related to the selection of nontargeted and poorly targeted serotypes . Nevertheless, in some instances, the more susceptible serotypes may also be the more invasive . In this case, if these serotypes are not included in the vaccine, vaccination could increase pathogenicity. Vaccination against Haemophilus influenzae type b reduces resistance is this species, as resistance is found more frequently in this serotype . In this particular case, both pathogenicity and resistance have dropped simultaneously, as no clear replacement by other types has been documented to date. In general, the vaccine-mediated reduction in the number of pathogenic organisms and infected hosts reduces the likelihood that resistance will evolve in these bacteria, as acquisition of resistance by mutation and horizontal transfer is a density-dependent phenomenon (see above). Obviously, the principle applies not only to human but also to animal infections and may have an important impact on the development of resistance in the food industry, where antibiotics are frequently used to prevent infections. In the next few years, we will probably face an explosion in the number of potentially effective novel vaccines. Such a trend is based on the growing use of genomic information together with in silico prediction of useful antigens and massive production of recombinant proteins to be used for immunization of animal models . This has been called reverse vaccinology and will doubtless offer a great number of antigenic candidates for the development of a new generation of vaccines. We must insist, however, that when vaccines are developed, the local microbial ecology should be considered. We have mentioned that antibiotic-susceptible but more virulent bacterial types may emerge because a vaccine is directed against resistant organisms. But it is also true that some vaccines may eventually be more effective in eliminating some antibiotic-susceptible types and therefore select for antibiotic resistance. This possibility should be considered when a vaccine is introduced into a given human habitat . A more optimistic perspective is to consider that prolonged use of antibiotics has reduced the diversity of some bacterial pathogens, so that some pathogenic species may be less able to adapt to novel challenges. If an effective vaccine is available for reducing the number of antibiotic-resistant clones, some pathogens in some parts of the world may become locally extinct (see below). Antivirulence-directed therapeutic approaches. : Antivirulence is considered here as interventions directed against virulent organisms and/or the expression of the virulence determinants. One of the best-known side effects of antibiotic treatment is the replacement of the antibiotic-susceptible commensal flora with resistant organisms. The release of antibiotics in the environment, as well as the use of antibiotics as growth promoters in animal feed, may have a profound impact on bacterial populations. One theoretical possibility to avoid this side effect is to develop antibiotics directed exclusively against virulent organisms. A "magic bullet" is one that only kills virulent bacteria in such a way as to preserve the antibiotic-susceptible phenotype of nonpathogenic bacteria. In this way, commensal bacteria will remain a colonization barrier against infection and will not acquire antibiotic resistance genes. Acquisition of antibiotic resistance genes by nonpathogenic bacteria is detrimental for two reasons. First, these bacteria will constitute a reservoir of antibiotic resistance genes (and antibiotic resistance vectors) that may be transferred to virulent bacteria. Second, antibiotic-resistant bacteria can protect the susceptible ones (eventually pathogenic) from the action of antibiotics (see above). Although this magic bullet will obviously be desirable, it is not easy to find a lethal target present only in virulent bacteria. Some types of antibiotics act and have been designed in a way that antibiotic resistance will automatically reduce bacterial fitness and consequently the pathogenicity of the resistant organism. For instance, naturally occurring albomycins , metalloporphyrins , and synthetic derivatives of Fe3+ siderophores with antibacterial activity are taken up across the bacterial outer membrane by transport systems for Fe3+ siderophores and heme transport systems. Indeed, some antibiotics have been developed that include a siderophore moiety linked to the antibiotic molecule . Those antibiotics make use of the siderophore transport system for entry into bacteria. Resistant bacteria can arise as a consequence of mutations in this transport system. However, loss of the iron transport system is detrimental for bacterial growth at the site of infection, so that treatment with this type of antibiotic may severely affect bacterial survival, either by the antibiotic effect itself and by allowing host defenses to cope with the infection due to a mechanism of resistance that has an excessive biological cost for the bacteria. Vaccination directed against virulence determinants has also been proposed as an effective antivirulence approach to prevention of infectious diseases. Such as all strategies not based on antibiotic treatment, this may help reduce the emergence and spread of antibiotic-resistant bacteria. One example of this strategy is the development of novel vaccines against S. aureus. In this organism, the expression of virulence factors is triggered by the RNA III activating peptide (RAP). This peptide is continuously secreted by bacteria, and its activity resembles some other quorum-sensing signals which are also inducers of the expression of virulence determinants . Vaccination with RAP protects mice from S. aureus pathology . This indicates that anti-quorum-sensing strategies might be useful to fight infection. A different approach for antivirulence vaccination was used by Mason et al. . A transgenic potato containing the gene for E. coli heat-labile enterotoxin was obtained and used to feed mice. Feeding with such potatoes produced higher levels of antitoxin antibodies than the whole-bacteria vaccine. Thus, this opens the possibility of developing transgenic plants containing several virulence factors that could be used as multisubunit edible vaccines . Strategies Directed against Resistance May Reduce Bacterial Virulence : It has been suggested that the search for and further use of antibiotics active only against antibiotic-resistant bacteria might reverse the problem of antibiotic resistance. The hypothesis of antibiotics able to select susceptible organisms was suggested some years ago . For instance, to fight the spread of beta-lactamase-producing bacteria, it may be possible to develop an antibiotic drug whose activation depends on the breakage of a beta-lactam ring. Because only beta-lactamase-positive resistant bacteria would be susceptible to such an antibiotic, beta-lactam-susceptible organisms will be selected. This and other approaches are currently under investigation. Obviously, any new antibiotic able to eliminate (because of the absence of cross-drug resistance) resistant organisms to other drugs has at least the possibility of eradicating resistance. This effect is only possible if the resistant organisms are in a minority with respect to susceptible ones, so that recolonization after therapy will occur from within the host and from other hosts with susceptible bacteria. However, until now, the best (and maybe only) strategy against resistance is a good policy for the prescription and use of antibiotics . Rotation in antibiotic use has been suggested and even implemented, in some cases, with an improvement in antibiotic resistance rates . However, the levels of resistance never drop to zero, so that rotation of antibiotics is useful only transiently . Antibiotics must be used only when needed , and the consumption of antibiotics for purposes other than the treatment of infectious diseases must be avoided as much as possible. For instance, the ban on the use of avoparcin in animal feed has curbed the development of vancomycin resistance in the European Union . It has been discussed whether the ban would be followed by a reduction in the productivity and an increase in the morbidity and mortality of the animals as a consequence of an increasing number of infections . However, neither a reduction in productivity nor an increase in the number of infections has been observed after discontinuing the use of antimicrobial growth promoters , probably because hygienic measures were implemented at the same time as the ban . Strategies against the emergence and further spread of antibiotic resistance genes in bacterial populations will obviously reduce virulence because infections will be easily treated and the disease will disappear. We cannot forget that during the 1980s, it was common to believe that infections would not be a relevant clinical problem in the future of mankind. However, it became clear in the 1990s that this was not the case. Some infections became untreatable by conventional chemotherapy, and this produced a reemergence of the problem of infectious diseases . This renaissance has not been a consequence of the introduction of novel bacteria with a higher virulence potential, but mainly as a result of emergence of the same families of bacteria which now are resistant to antibiotics. Because antibiotics are the very last line of defense incorporated along human evolution in its fight against infections, breaking down this defense will restore the problem of infectious diseases to the situation of the preantibiotic era. Therefore, antiresistance strategies will clearly diminish the overall virulence of bacterial pathogens, because bacteria will remain susceptible to treatment so that they will not produce disease. To achieve this goal, novel antibiotics are required. Drug diversification. : Although several antibiotics are currently being used in therapy, all belong to a few structural families, so that resistance to one antibiotic frequently produces resistance to other members belonging to the same family. Novel antibiotics belonging to different structural families than those currently in use are then needed in order to fight against antibiotic-resistant bacteria Two approaches can be combined in the search for novel antibiotics. One is the classical search for novel inhibitors of lethal bacterial targets either by using natural sources or by screening of synthetic compounds. The other is the search for inhibitors of resistance systems, with the aim of recovering a susceptible phenotype in a previously resistant population. Examples of the latter strategy are the beta-lactamase inhibitors currently used in clinical practice, and the inhibitors of MDR pumps which are under development by some pharmaceutical companies. Concerning the search of inhibitors of classical targets, major efforts are currently under way in order to use the knowledge derived from the sequencing of whole microbial genomes . The advantages derived from the use of combinatorial chemistry and rational design by computerized molecular modeling of crystal structures of potential targets , the use of new methods for the screening of natural compounds , and the increasingly refined methods of high-throughput screening will be useful for developing novel families of antibiotics in the future. Nevertheless, we must still wait for the introduction of new antibiotics developed by using these methodologies. In the meantime, new fluoroquinolones such as moxifloxacin , gemifloxacin , and gatifloxacin , new tetracyclines such as glycylcycline , streptogramins such as the combination quinupristin/dalfopristin , macrolide-related agents such as ketolides , and a new family of antimicrobials, oxazolidinones , have been launched and/or in the last phases of development. Some new families and resistance inhibitors are under development. These include the novel inhibitors of class C beta-lactamases such as boronic acids and inhibitors of MDR efflux pumps The diversity of clones within a given bacterial species is severely reduced by antibiotics. In some instances, the species may maintain a certain degree of diversity when the resistance determinant disseminates by horizontal transfer into different clones, and if a mechanism of diversification is available, for instance, capsular transformation in S. pneumoniae. Nevertheless, even in the case of species able to exchange DNA, a progressive simplification of the population structure is expected to occur. In S. pneumoniae, prolonged challenge with beta-lactams has selected a relatively small number of "international resistant clones," certainly at the expense of reducing the number of susceptible clones. It has been discussed that this reduction in diversity may lead to the potential extinction of some highly virulent but antibiotic-susceptible clones. In this sense, the overuse of antibiotics may have reduced the overall pathogenicity of certain organisms. In any case, the antibiotic-driven reduction in diversity of the bacterial clones may cause extinction of some of these bacterial pathogens, using either vaccines against the reduced number of resistant clones or new antibiotics to which the resistant bacteria are susceptible, or both strategies simultaneously. Implementation of barriers to transmission will increase the possibility of reaching extinction. It seems that a low-diversity population structure will reduce the possibility of maintaining a pathogen under these new pressures. INFLUENCE ON EVOLUTION OF THE HOST : The sentence "I am myself and my circumstances" from the Spanish philosopher Ortega means, in biological terms, that not only genetics but also the environment is relevant to phenotype. In the case of humans, it might also mean that the human body is formed by 10% mammalian cells and 90% prokaryotic cells (our normal microbiota) , each of which can influence the evolution of the other. The human host is also surrounded by a plethora of environmental microorganisms, some of which may interact transiently with the human body. Therefore, it is clear that the ability of bacteria to grow and damage human tissues has most probably had an effect on human evolution. For instance, the build-up of the extremely complex system of immunological responses involved in the defense against infection that was selected during the evolution of animals was necessary (and, so, selectable) only because of the existence of virulent microorganisms. In this regard, it has been suggested that infectious diseases might be the major evolutionary force driving the high diversity displayed by the major histocompatibility complex . Body temperature, which is higher than the mean environmental temperature, limits the growth of microorganisms in humans to those that can grow in a narrow range around 37C. In fact, it has been demonstrated that pathological hypothermia favors infection by environmental bacteria such as Pseudomonas putida . We thus speculate that body temperature is a very first line of defense that precludes infection by environmental microorganisms. It is thus possible that this anti-infective function together with several other selective pressures has contributed to select 37C (a temperature that requires that constant metabolic activity be maintained) as the human body temperature. These examples illustrate that virulent microorganisms might have had a relevant effect on host evolution, as suggested by Haldane more than 50 years ago . Since microorganisms can overcome human anti-infective defenses and kill the human host, infection might select the most resistant (in terms of infection) genetic backgrounds in the host population . In such a way, the prevalence of specific infectious microorganisms in different geographic allocations might be a driving force in human evolution. A very well known example of this viewpoint is the prevalence of sickle cell anemia in regions with a high rate of malaria infections . Sickle-shaped hemoglobin is protective against malaria. Thus, infection by Plasmodium falciparum has produced a bias in the number of sickle cell heterozygotes. The percentage of sickle cell homozygous individuals did not show such a relevant increase because these people have a high mortality risk caused by complications of sickle cell disease. A more recent example is the prevalence of cystic fibrosis in Caucasian populations, which is higher than predicted by genetic analysis. Cystic fibrosis is a genetic disease caused by the lack of activity of a chlorine channel (CFTR). CFTR is also the receptor for some bacterial pathogens, such as Salmonella enterica serovar Typhi. Therefore, it has been suggested that cystic fibrosis could prevent infections by this bacterial species . In such instances, an increased risk of infection by S. enterica serovar Typhi might select populations with a bias in the percentage of cystic fibrosis patients. In a similar way, it has been found that resistance and susceptibility to virulent M. tuberculosis is a complex genetic trait , so that different genetic backgrounds make people more and less susceptible to developing clinical tuberculosis once they have been infected with M. tuberculosis. It has been described that Yanomami Indians of the Brazilian Amazon have diminished cell-mediated immune responses against M. tuberculosis compared with control individuals of European extraction living within the same region. Yanomami Indians remained isolated until the mid-1960s and have not been in contact with M. tuberculosis until recently. As stated by the authors of that work, M. tuberculosis infections exerted a powerful selective pressure, "resulting in the elimination of a significant proportion of highly susceptible individuals over their reproductive age" . This selection occurred before the availability of antituberculosis treatment in the 1950s and might be occurring now if the treatment is not effective due to the emergence of antibiotic-resistant M. tuberculosis. The same occurs for other infectious diseases. We must to keep in mind that the contact between Europeans and native populations in America produced dramatic effects on the Native Americans due to the dissemination of "European" infectious diseases among them. Whereas Europeans had been in close contact with these infectious agents, with a resulting selection against the most susceptible individuals occurring a long time ago, the first encounter between the native population and the European "alien" infectious agents caused the death of a great number of people. It has also been suggested that infections in the New World have been an important force for selecting the mixed race (mestizo) produced by the mating between Spaniards and Indians. As suggested by the authors, mestizos should be immunologically more capable of defending themselves against the different infectious agents brought over from the Old World and maintaining their lower susceptibility to New World infections, so that they should be more adapted than Europeans and Native Americans to the new conditions produced by the introduction of novel pathogens in the environment. These examples illustrate that infections have been important driving forces in human evolution. We have to keep in mind that several genetic changes might have been selected over human evolution just because they are protective against infections. However, we have only detected those changes which, like cystic fibrosis and sickle cell disease, are associated with a relevant genetic disorder. More recently, the study of infection-resistant phenotypes by classic human genetic analysis is shedding new light on the potential role of infectious diseases as driving forces in host evolution. The use of antibiotics and the development of antibiotic resistance mechanisms might also have an effect on host evolution, because they will change the risk for infection. To illustrate, we can speculate that the introduction of effective treatments (antibiotics, prophylaxis, vaccines) against S. enterica serovar Typhi might eventually reduce (in the long term) the prevalence of cystic fibrosis to the percentage of population predicted by purely genetic criteria. Also, as stated previously, effective antituberculosis treatment will allow the highly susceptible population to survive, because M. tuberculosis would not exert any selective pressure. Antibiotic treatment will then have the effect of removing selective pressure due to infections, and antibiotic resistance will have the effect of returning such selective pressure. We want to stress here that, in any case, the time scale for a detectable effect of antibiotic treatment on human evolution is a lot longer than the time scale needed for bacterial evolution due to the extremely different lifetimes of the two types of organisms (from minutes to hours for bacteria, several decades for humans), so that a direct effect of antibiotics in human evolution is unsuitable unless antibiotics are maintained as the main choice for treatment of infections over several centuries. However, the most rapid and evident effect of antibiotic treatment will be on the composition of normal microflora because antibiotic treatment will select for resistant variants not only in the treated virulent microorganisms, but in the commensal microflora also (see below). Changes in Normal Host Microbiota : Might virulent and/or antibiotic-resistant bacteria change host microflora? It is clear that a bacterial pathogen needs to successfully compete with normal host microflora in order to produce an infection, so that a temporal displacement of commensal bacteria might occur over the very first steps of infection. On the other hand, virulent bacteria might either select "resistant" hosts (see above), in which they can multiply without producing any relevant disease, and attenuated bacterial variants might also be selected. In any of these cases, the previously virulent bacteria might finally be incorporated as a commensal microorganism into the normal host microflora. Antibiotic treatment will also displace normal host microflora, allowing the recolonization by exogenous antibiotic-resistant bacteria as well as by antibiotic-resistant mutants belonging to normal microflora. This effect will probably be buffered in the long run, because human populations are not always under antibiotic treatment, so that the risk for selection of resistant variants will be lower for commensal microflora than for virulent microorganisms (see above). Nevertheless, it cannot be ruled out that because of the huge consumption of antimicrobials in the world, some bacterial members which form the normal host microbiota will be replaced by others. For instance, highly sporulating gram-positive bacilli may replace (because spores are antibiotic resistant) less-sporulating organisms. Some Lactobacillus species are intrinsically less resistant to antibiotics than others and will be progressively selected. In general, bacteria with lower growth rates (less susceptible to most antibiotics) may also be selected. It can be predicted that those changes will produce effects, for the moment largely unknown, on human health. Even for the same bacterial species, a replacement of human-adapted strains by animal-adapted strains might occur. For instance, natural populations of E. coli are organized in an ecotypic structure where adaptation to the host plays an important role . As discussed before, the use of antimicrobials as growth promoters selects antibiotic-resistant clones in the commensal microbial flora of farming animals . Those clones that are not good colonizers of humans might replace human-adapted antibiotic-susceptible clones in the case of antibiotic treatment, thereby breaking down the close association between bacterial clones and their host. It is true that current bacterial diversity will probably offer "almost equivalent" resistant organisms to replace those that may be under danger of extinction. Nevertheless, most probably our intestinal flora will change, and therefore our microbial heritage for the future of humankind (the "historical host" in the future) will be adulterated. Again, the evolutionary consequences are totally unknown. Microbial evolution is not a reversible process. The release of huge quantities of antibiotics on the Earth, together with other environmental toxic compounds, and the changes in the population structures of human and animal hosts have profoundly changed the interactions of microbes and humans. The possibility of wide restoration of a wild antibiotic-susceptible bacterial flora in nature is beyond our current abilities. Backmatter: PMID- 12364379 TI - Malaria Rapid Diagnostic Tests: One Size May Not Fit All AB - Keywords: null: null Backmatter: PMID- 12364369 TI - Atopic Dermatitis and Fungi AB - Atopic dermatitis (AD) is a chronic, itching, inflammatory skin disease which is associated with asthma and/or hay fever and a familial occurrence of these conditions. Genetic factors are important in the development of AD, but the exact hereditary pathway is still unknown. Dry skin and the weakened barrier function in patients with AD is very important for the patient's reactions to irritants and other external trigger factors including microorganisms. The standard treatments are topical corticosteroids, topical immunomodulating agents, and emollients. If AD cannot be controlled by this type of treatment, systemic immunomodulating agents may be used. UVB, UVA, or psoralen-UVA may also be used for widespread severe lesions. However, some patients do not respond to these standard treatment, and then it is important to consider the role of microorganisms, house dust mites or food. The role of the Malassezia yeasts in AD, especially AD located to the head and neck region, is now documented in several papers. There are also several papers indicating the role of Candida as an aggravating factor in AD. Patients with AD also develop chronic dermatophyte infections more easily, and patients with AD and chronic dermatophyte infections may show improvement in their AD when treated with antifungal drugs. Keywords: INTRODUCTION : Atopic dermatitis (AD) is a chronic, itching, inflammatory skin disease which is associated with asthma and/or hay fever and a familial occurrence of these conditions . AD is the most common chronic skin disease in children . Approximately 18% of 7-year-old children have or have had AD. In 90% of patients the disease starts before 7 years of age, and usually it starts before 1 year of age. The majority of the patients improve before or around puberty, and it is estimated that approximately 2% of the adult population has AD . However, AD may start in adult life, although this is rare. There is a documented increase in the incidence of AD in all countries with an Western life-style, but the exact reason for this is unclear (the frequency of AD in 7-year-old children is 2 to 3% for those born before 1960, 4 to 8% for those born between 1960 and 1970, 9 to 12% for those born between 1970 and 1980, and 15 to 20% for those born after 1980) . Genetic factors are important in the development of AD, but the exact hereditary pathway is still unknown . If both parents have had AD, the risk for the children is approximately 70%. Infiltrates of activated T-helper cells are found in the skin of patients with AD . There is an imbalance between Th1 and Th2 cells toward an increase in cells of the Th2 type. Th2 cells have a cytokine profile with an increased production of interleukin 4 (IL-4), which stimulates the plasma cells to an increased production of immunoglobulin E (IgE) . Another interesting observation in the skin of AD patients is the presence of a dendritic epidermal cell population (Langerhans' cells and mast cells) that carry IgE receptors and specific IgE . Antigen presented by these cells may stimulate Th2 cells to synthesize IgE and hence contribute to the inflammatory changes in the skin . When the skin of patients with AD is examined, it is often drier than the skin of an individual without AD . It also reacts differently in response to stimulation. The white dermography appears when the skin is lightly scratched and is characterized by blood vessel constriction instead of dilation . The threshold for itch is also lower than in healthy individuals. Dry skin and the weakened barrier function in patients with AD are very important in the reactions of the patients to irritants and other external trigger factors including microorganisms . The pathophysiology and etiology of AD have been reviewed and discussed in a supplement of the American Journal of Dermatology . Staphylococcus aureus can be cultured from both involved and uninvolved skin in 85 to 90% of patients with AD . S. aureus may aggravate AD due to a secondary infection . However, of greater importance is probably the presence of exotoxins and other substances from S. aureus that may act as allergens or, more importantly, as superantigens . Viral infections, especially with herpes simplex virus, may also aggravate AD . The yeasts Malassezia and Candida may aggravate AD due to an allergic reaction . In particular, the role of the Malassezia yeasts is now documented in several papers . Malassezia species are members of the normal human cutaneous flora, and a defect in the skin barrier may facilitate the contact of these yeasts with the immune system . Candida yeasts are members of the normal flora of mucous membranes, and the major contact with the immune system is through the gastrointestinal tract and, in women, even through the vagina . Chronic dermatophyte infections are more common in patients with AD, and dermatophytes, especially Trichophyton rubrum, may act as allergens . The role of other hyphomycetes in AD has not been studied in detail . Other important factors are food, house dust mites, topically applied irritants, psychological factors, and climate . The diagnosis of AD is usually based on many variables, including history and clinical and laboratory findings. Rajka was the first to make a diagnostic list consisting of major and minor criteria . These criteria were later revised and are known as the Hanifin and Rajka criteria . These criteria have been scientifically evaluated and found to work fairly well in the clinical situation even though many of the criteria have no precise definition, some are nonspecific, and some are uncommon. However, Williams et al. developed a minimum list of reliable criteria for the diagnosis of AD that could be used in clinical as well as in epidemiological studies. The obligatory criteria are an itching skin condition or report from parents that a young child itches or scratches as if itching. Thereafter, at least three of the following criteria should be fulfilled: history of involvement of the skin folds on the flexor sides of the elbows and knees, anteriorly on the ankles, on the throat, and on the cheeks in children younger than 10 years; history of asthma or hay fever in the patient or of atopic disease in the immediate family of children younger than 4 years; history of general skin dryness during the last year; visible eczema present on the flexor sides of the elbows and knees, anteriorly on the ankles, and on the throat (or cheeks, forehead, and extensor side of the extremities in children younger than 4 years); and onset before the age of 2 years (this criterion should not be used if the child is younger than 4 years) . The clinical picture varies with age and may be described in 3 different phases: infant phase (younger than 2 years), childhood phase (2 to 12 years), and adolescent/adult phase . Important differential diagnoses are seborrheic dermatitis, psoriasis, rosacea and perioral dermatitis, fungal infections, ichtyosis vulgaris, scabies and contact dermatitis. Accurate assessment of the extent and severity of AD is essential for quantitating the baseline clinical disease burden and the effectiveness of treatment regimens being tested . One assessment system that is videly used is the SCORAD method , and another method is the EASI . It is beyond the scope of this article to discuss which one may be superior. Treatment of AD is topical with emollients, corticosteroids, or immunomodulating agents . The use of topical tacrolimus is a new approach to the treatment of AD . In severe AD, systemic treatment with corticosteroids and immunomodulating drugs is needed . UV treatment is also an effective alternative . Topical and systemic antibacterial treatment is used in patients with signs of bacterial infections. Antifungal therapy is used in patients with signs of a fungal infection. However, a more widespread use of antimycotics based on the role of the most yeasts as common allergens is still not commonly recommended. The prognosis of AD is good; 60 to 70% of patients recover during childhood . The prognosis is less positive in women, in patients who had severe AD as children, and in those with concurrent allergic rhinitis and/or asthma or a positive family history of AD. FUNGI INVOLVED IN ATOPIC DERMATITIS : Not only bacteria, especially S. aureus, but also fungi may play an important role as aggravating factors in AD. The majority of studies in AD and fungi deal with the Malassezia yeasts, probably because these yeasts are the most important fungi in AD. Malassezia : Historical. | The earliest report of Malassezia-like yeasts was made by Eichstedt in 1846 . In 1853, Robin described round cells on the skin of patients with dandruff . In 1873, Rivolta described the presence of round double-contoured budding cells in a patient with psoriasis . In 1874, Malassez noticed budding cells of various shapes in the stratum corneum of patients with various skin diseases, and the name Malassezia furfur was proposed by Bailon in 1889 . However, Malassez was never able to culture the microorganism. Von Sehlen and Unna were probably the first to culture the organism. The genus name Pityrosporum was proposed by Sabouraud in 1904 . In 1913, Castellani and Chalmers were able to culture the organism and characterize the growth properties . They introduced the name Pityrosporum ovale, and their description of the organism is based on the characteristics of the fungus both in skin scales and in the laboratory . The name P. pachydermatis was introduced by Veidman in 1925 for a nonlipophilic species of the genus Pityrosporum isolated primarily from animals . In 1951, Gordon isolated a round, double-contoured yeast that produced spherical to oval buds from both pityriasis versicolor scales and normal skin . The name he gave the organism, P. orbicualre, is based primarily on the micromorphology. However, later studies have shown that the round and oval forms of the organism are not stable and that the two forms can change from one to the other on special culture media or under special culture conditions and when experimental infections are induced in rabbits . The genus Malassezia has in a taxonomic revision been placed in the kingdom Fungi, division Basidiomycota, class Hymenomycetes, order Tremellales and family Filobasidium uniguttulatum . Molecular biology using rRNA sequence analysis and nDNA comparisons, as well as G+C content in extracted DNA, has clearly divided the genus Malassezia into seven different species . Today this genus consists of M. furfur , M. pachydermatis , M. sympodialis , M. globosa, M. slooffiae, M. restricta, and M. obtusa . Morphology and growth requirements. : The seven Malassezia species exhibit typical morphological and physiological features . They reproduce by unilateral budding, leaving prominent scars on the mother cells . The yeast cells may be bottle-shaped, globose, ovoid, or cylindrical. Buds may be formed on a narrow or wide base. The most interesting feature of the genus Malassezia is the presence of a thick multilayered cell wall whose inner surface is corrugated. The protoplasmic membrane is firmly attached to the inside of the cell wall. The major components of the cell wall are mannoproteins (75 to 80%), lipids (15 to 20%), and chitin (1 to 2%) . It is thought that the immunomodulatory capacity of the Malassezia yeasts are related to cell wall lipids. Using various physical, chemical and metabolic characteristics, the Malassezia species can be identified . We have recently developed a modification of previous identification systems by which the seven different Malassezia species easily can be identified . The advantages of this system are that it is easier and faster and that Tween reaction, which is often difficult to read, is not necessary. We always isolate the Malassezia yeast using the media described by Leeming and Notman . This medium contains, among other ingredients, ox bile, glycerol, glycerol monostearate, Tween 60, and whole-fat cow's milk. Compared to other culture media, this was the most effective for isolation of the Malassezia yeasts . Various morphological and physiological criteria can be used to characterize the Malassezia yeasts . M. pachydermatis is nonlipophilic and can be cultured on lipid-free culture media. Cells are small and ovoid with a very broad base, and they never produce filaments. M. furfur can vary in micromorphology, producing cells that are long cylindrical ovoid or even globose. A few strains may produce filaments spontaneously or during special culture conditions. M. furfur is catalase positive, does not split esculin, grows on Dixon agar at 38C, and is capable of growing in the presence of Cremophor. M. sympodialis has small ovoid cells. It is catalase positive, splits esculin, grows on Dixon agar at 38C, and is not capable of growing in the presence of Cremophor. M. slooffiae has short cylindrical cells appearing ovoid under the light microscope with buds formed on a broad base. It is catalase positive, does not split esculin, grows on Dixon agar at 38C, and is not capable of growing in the presence of Cremophor. M. obtusa has large cylindrical cells with a broad base. It is catalase positive, splits esculin, does not grow on Dixon agar at 38C, and is not capable of growing in the presence of Cremophor. M. globosa has stable spherical cells. Buds are formed on a narrow base and might elongate to form germination tubes or even short filaments. It is catalase positive, does not split esculin, does not grow on Dixon agar at 38C, and is not capable of growing in the presence of Cremophor. M. restricta has small spherical or ovoid cells with a relatively narrow base. It is catalase negative, does not split esculin, does not grow on Dixon agar at 38C, and is not capable of growing in the presence of Cremophor. FIG. 1. | Typing system for the Malassezia yeasts. Typing system for the Malassezia yeasts. Distribution and isolation. : The Malassezia yeasts are members of the normal human cutaneous flora . However, there is great variation in the density and presence in various skin locations , in children compared to adults , and in normal skin compared to diseased skin . The Malassezia yeasts are also associated with several skin diseases such as pityriasis (tinea) versicolor, Malassezia folliculitis, seborrheic dermatitis, atopic dermatitis, and some forms of confluent and reticulate papillomatosis (Gougerot-Carteaud syndrome) . They may even be involved in systemic diseases . In 1951, Gordon isolated lipophilic yeasts from both normal healthy volunteers and patients with pityriasis versicolor . Using direct microscopy or culture, Roberts found what he called P. ovale on the scalp and on the chest of 97% and 92 to 100% normal healthy adults, respectively . He also isolated the yeast in large numbers from patients with pityriasis versicolor . In a quantitative culture study, using a modification of the Williamson-Kligman scrub technique, Malassezia yeasts were cultured from clinically normal skin of the on the chest, back, upper arm, lower leg, and dorsal aspect of the hand . The highest count was found on the back, and the lowest was found on the hand. An easy semiquantitative method is to use contact plates . This was first described in detail in 1987 . However, better results have later been obtained using a modification of the culture medium developed by Leeming and Notman . Leeming et al. were able to culture the Malassezia yeasts from 20 different body sites, with the largest numbers being found on the scalp and upper trunk . We found in a culture study of adults that the density of the Malassezia yeasts decreased with increasing age . Although the number of yeast cells is larger in adults, the yeasts can also be cultured from children . Several studies have demonstrated that the Malassezia yeasts can be cultured in newborns. The Malassezia yeasts can be cultured not only from healthy skin but also from patients with various skin diseases . The number of yeasts is not always increased, but other mechanisms may be important in the pathogenesis. The yeast may induce immunological changes or behave as irritants, and then a reduction even of a relatively small number of yeast cells will result in an improvement. Quantitative or semiquantitative cultures can therefore be used to monitor the effect of treatment or changes in predisposing factors. Only a few culture studies have been done using the new classification of the Malassezia yeasts . However, it is important to look at the distribution of the various Malassezia species not only on healthy skin but also on the skin of patients with Malassezia-associated skin diseases. Nakabayashi et al. isolated various Malassezia species from the skin of healthy individuals and from patients with pityriasis versicolor, seborrheic dermatitis, and atopic dermatitis in Japan . The yeasts were cultured qualitatively using swabs and Dixon's agar medium, and specimens for culture were taken from the scalp, face and trunk. Cultures were negative in 14% (pityrasis versicolor) to 50% (healthy individuals) of cases. M. globosa was the most commonly isolated species from healthy volunteers, patients with pityriasis versicolor, and nonlesional skin of patients with AD. M. furfur was the most commonly isolated species from patients with seborrheic dermatitis and lesional skin of patients with AD. In a study by Kim and Kim, the Malassezia yeasts were isolated qualitatively from various body regions in normal healthy volunteers from South Korea . All six lipophilic species were cultured. However, there was a great variation in the incidence of these yeasts, with M. restricta being the most commonly isolated species from the scalp and forehead and M. globosa being the most commonly species isolated from the chest and the back. In two reports from Canada, Gupta et al. have isolated the various Malassezia yeasts from both healthy individuals and patients with various Malassezia-related skin diseases . In the first study, the Malassezia yeasts were isolated qualitatively, in patients with pityriasis versicolor, by scraping skin flakes into a paper bag, and later the flakes were transferred to Leeming and Notman agar and to Littman agar supplemented with olive oil . A total of 111 samples were obtained from patients in Ontario, Canada, and 18 were from patients in other countries (South Africa, Hong Kong, Portugal, Brazil, and Hawaii). In the patients from Canada, M. sympodialis was isolated in 59%, M. globosa was isolated in 25%, and M. furfur was isolated in 11%. In patients from other countries, M. globosa was the dominating species, being cultured in 62%. In the other study, the Malassezia yeasts were isolated semiquantitatively, using contact plates with the Leeming and Notman agar , from both healthy volunteers and patients with pityriasis versicolor, seborrheic dermatitis, AD, and psoriasis. All the lipophilic Malassezia species were isolated. M. sympodialis was the dominant organism in healthy controls and in patients with pityriasis versicolor and AD, while M. globosa was the dominant species in patients with seborrheic dermatitis and psoriasis. The number of yeast colonies was largest in lesional skin from patients with pityriasis versicolor but was not statistically significantly larger than in nonlesional skin or in skin from healthy individuals. However, the number of colonies of Malassezia was statistically significantly larger in both healthy individuals and patients with pityriasis versicolor than in patients with seborrheic dermatitis, AD, and psoriasis. Erchiga et al, cultured the various Malassezia species qualitatively using Dixon agar from both lesional and nonlesional skin of patients with pityriasis versicolor in Spain . Cultures were positive in all patients. M. globosa, either alone or in combination with M. sympodialis, was the dominant species, being cultured in 97% of patients; M. sympodialis was cultured in 32%, either together with M. globosa or alone (only three patients). In nonlesional skin from the shoulders, Malassezia species were cultured in 47% of the patients and M. sympodialis was the dominant species, being cultured in 42% of all cultures. According to these five reports, M. sympodialis, M. globosa, and M. furfur were the most commonly isolated Malassezia species. However, there were differences in the species that was most commonly isolated, not only between healthy individuals and patients with various diseases but also between results from different countries . The Malassezia yeasts as pathogens. : As mentioned above, the Malassezia yeasts are opportunistic pathogens, and under the influence of various predisposing factors, they are involved in several diseases . In pityriasis versicolor, under the influence of exogenous or endogenous predisposing factors, the yeast changes from the blastospore form to the mycelial form . Pityriasis versicolor is a chronic superficial disease that usually affects the upper trunk, neck, or upper arms. It has a worldwide distribution. In tropical areas, it has been found in 30 to 40% of the population. However, the incidence is much lower in temperate climates. The diagnosis is based primarily on a typical clinical picture in combination with bright yellow fluorescence under Wood's light examination and direct microscopy. Direct microscopy is of major importance, and the round or oval budding cells and hyphae can easily be identified. There are numerous ways of treating pityriasis versicolor, both topically and with systemic drugs . However, the high rate of recurrence, affecting 60% of patients after 1 year and 80% after 2 years, is an outstanding problem. Consequently, a prophylactic treatment regimen is necessary to avoid recurrence. Malassezia folliculitis is a chronic disease characterized by pruritic follicular papules and pustules located primarily on the upper trunk, neck, and upper arms . Under the influence of predisposing factors, Malassezia folliculitis may be explained by an extensive growth of Malassezia in the hair follicle. The inflammation may be due both to products of the yeast and free fatty acids produced as a result of the lipase activity of the yeast . The diagnosis is based on a typical clinical picture of itchy papules and pustules as the predominant symptoms, direct microscopy, and the effect of antimycotic treatment . The effect of antimycotic treatment is often dramatic. Most cases responds well to topical treatment, but systemic treatment with e.g. itraconazole or fluconazole has been used in difficult cases . Lesions and itching recur in most patients if treatment is not maintained intermittently. Therefore, a prophylactic treatment schedule, such as topical treatment once or twice a week, is recommended. Seborrheic dermatitis is characterized by red scaly lesions located predominantly on the scalp, face, and upper trunk. There are now many studies indicating that the Malassezia yeasts play an important role in seborrheic dermatitis . Many of these are treatment studies which describe the effectiveness of antimycotics, paralleled by a reduction in number of the Malassezia yeasts, whereas recolonization leads to a recurrence of seborrheic dermatitis. The increased incidence of seborrheic dermatitis in patients with immunosuppressive disorders suggests that the relationship between Malassezia and the immune system is important, but the exact mechanism involved is still unknown. Studies designed to investigate this mechanism in patients with seborrheic dermatitis fall into four areas : the humoral response to Malassezia, lymphocyte transformation response to Malassezia extract, T helper (CD4)/T suppressor (CD8) cell ratios, and cytokine responses. However, studies in all of these areas have given inconsistent results, and the local reaction in the skin may give a better answer. In a recent study, the number and type of inflammatory cells and mediators from lesional and healthy skin in patients with seborrheic dermatitis were been studied . Staining was often more intense when Malassezia yeasts cells were present. An increase in NK- and CD16-positive cells in combination with complement activation indicates that an irritant or nonimmunogenic stimulation of the immune system is important. With the interleukins, an increase in the production of inflammatory interleukins and an increase in the regulatory interleukins for both Th1 and Th2 cells were seen. The immune response described for cutaneous C. albicans infections has similarities to the skin reactions found in seborrheic dermatitis . Atopic Dermatitis and the Malassezia Yeasts : Distribution of the Malassezia yeasts. | Many reports clearly indicate the role of Malassezia in AD, especially AD located to the head and neck area in adults . The Malassezia yeasts have been cultured from patients with AD with the same frequency as from healthy individuals . In one study of children from newborn to the age of 20 years, the Malassezia yeasts were isolated from lesional and nonlesional skin, from the same region, as well as from normal skin of the forehead of patients with AD . Normal skin on the forehead in age- and sex-matched individuals with rhinoconjunctivitis and/or asthma and healthy volunteers served as controls. Cultures did not differ significantly among the three groups or between lesional and nonlesional skin. There was, however, a difference between the age groups. Cultures were positive in 5 to 15% of children aged up 10 years and in 65 to 100% of 11- to 20-year-old persons. In a study of adult patients with AD involving the scalp and neck, Malassezia samples were taken from lesional skin of the forehead and cultured on Leeming and Notman agar in contact plates . Culture was positive in 83% of cases. These two studies were performed before the new taxonomy for the Malassezia yeasts was introduced. In the above-mentioned study by Nakabayashi et al., they were able to classify the Malassezia yeasts . Samples were taken from lesional and nonlesional skin from the scalp, face, and trunk. In lesional skin the most commonly isolated species was M. furfur, which was cultured in 21% of samples. M. globosa was cultured in 14%, M. sympodialis was cultured in 7%, M. slooffiae was cultured in 4%, and cultures were negative in 46%. In nonlesional skin, the most commonly isolated species was M. globosa (33%); M. furfur was cultured in 11% of samples; 22% were unknown species or contaminated cultures, and 33% of samples were culture negative. Gupta et al. cultured the Malassezia yeasts semiquantitatively, using the Leeming and Notman culture medium in contact plates, from patients with various skin diseases and healthy controls . In patients with AD, the total number of isolated colonies was significantly larger from nonlesional skin than from to lesional skin. The number of colonies from both lesional and nonlesional skin was significantly smaller than the number of colonies from healthy individuals, patients with pityriasis versicolor (both lesional and nonlesional skin), and nonlesional, but not lesional, skin of patients with seborrheic dermatitis. In AD there were no differences in the percentages of the various Malassezia species isolated from lesional and nonlesional skin. M. sympodialis was the most commonly isolated species, being cultured in 51% of the patients, M. globosa was cultured in 18%, M. furfur and M. obtusa were each cultured in 10%, M. restrticta was cultured in 8%, and M. slooffiae was cultured in 3%. The majority of Malassezia yeasts were cultured from the forehead and the trunk. In an English abstract from a paper otherwise written in Russian, Arzumanian stated that M. sympodialis was the most commonly isolated species isolated from both patients with AD and healthy individuals . In a recently completed multicenter study in Sweden, involving the Departments of Dermatology in Gothenburg, Lund, and Stockholm and the Department of Clinical Immunology, Karolinska Hospital, Stockholm, we cultured the Malassezia yeasts from 125 patients with AD, 16 with seborrhec dermatitis, and 30 healthy controls . Cultures were performed using the Leeming and Notman agar in contact plates from normal-appearing skin of the upper back and, in AD and seborrheic dermatitis patients, also from lesional skin. Cultures were positive in 56% of patients (70 of 125) with AD, 88% of patients with seborrheic dermatitis, and 83% of healthy controls. In patients with AD, M. sympodialis was cultured in 40%, M. obtusa was cultured in 25%, M. globosa was cultured in 23%, M. slooffiae was cultured in 6%, M. furfur was cultured in 4%, and M. restricta was cultured in 2%. The same frequency of the Malassezia yeasts was seen in healthy controls. However, in patients with seborrheic dermatitis, M. obtusa was cultured in 30% and M. sympodialis was cultured in 30%. Immunology. (i) Skin prick test for type I hypersensitivity. : A positive skin prick test (SPT) was first reported in 1958 among a group of AD patients . In 1983, Clemmensen and Hjorth described a positive effect of oral ketoconazole in the treatment of adult patients with AD localized to the head and neck area and with positive SPTs to Malassezia . In 1985, Waerstad and Hjorth investigated 741 patients with AD and found a positive Malassezia SPT only in patients with active eczema and most frequently in patients with head and neck dermatitis (28%) . Patients with other atopic manifestations such as rhinitis or asthma but without AD were negative in SPT to Malassezia. Kieffer et al., in collaboration with our group in Gothenburg, reported a positive SPT to Malassezia in 79% of adult patients with AD localized to the head and neck area . The SPT was positive in 45% of patients with AD but without the head and neck distribution, and patients with seborrhoeic dermatitis were negative. These percentages are much higher than those obtained by Waersted and Hjorth. However, several children were included in the report by Waerstad and Hjorth. The Malassezia extract used in this and the two earlier studies is a protein extract prepared by ALK Laboratories, Copenhagen, Denmark. The same extract has also been used for a histamine release assay (Lucotest-HR; Lundbeck Diagnostics, Copenhagen, Denmark) measuring the release of histamine from basophils after stimulation with the antigen. The results of histamine release after stimulation with Malassezia extract paralleled the results obtained with the SPT . Rokugo et al. found that the results of the SPT vary according to the age of the patients . Among AD patients younger than 10 years, 39% were positive in SPT, while 64% of patients older than 10 years were positive. Wessels et al. found positive SPTs to Malassezia in 84% of adult patients with AD . In a study by Broberg et al., children with AD, rhinitis and/or asthma without AD, and healthy controls were investigated . A total of 60 children in each category were included, and the children were divided into four groups: 0 to 5 years, 6 to 10 years, 11 to 15 years, and 16 to 21 years, with 15 in each group. SPTs to Malassezia were positive in 23% of patients with AD, 0% of patients with rhinitis and/or asthma, and 8% of healthy controls (three subjects). Among the patients with AD, 1 of 15 patients in the age group from 0 to 5 years was SPT positive compared to 6 of 15 (40%) in the age group from 16 to 21 years. Kim et al. have performed an SPT with an M. furfur extract in 80 young adult patients with the head and neck distribution of AD and compared the results to those obtained for controls without AD . A total of 45% showed positive SPT responses, and the severity score of the head and neck lesions in the M. furfur SPT-positive group was significantly higher than in the M. furfur-negative group. In another study, Devos and van der Valk studied patients with AD, with and without the head and neck distribution, using SPT with a Malassezia extract and other procedures . In a total of 92 patients with AD (the majority were young adults), the SPT was positive in 40%. However, only 6% of the patients with the head and neck distribution of AD were positive. This is contradictory to all earlier studies, in which patients with the head and neck distribution of AD showed the largest number of positive SPTs to Malassezia . In the above-mentioned Swedish multicenter study, we also performed SPT with extracts of Malassezia, the recombinant allergens Mal f 1, Mal f 5, Mal f 6, and a mix of these recombinant allergens in patients with AD, patients with seborrheic dermatitis, and healthy controls; 74% of the AD patients had the head and neck distribution. We found that 57% of the AD patients were positive to the extract or recombinant allergens; 52% were positive to the Malassezia extract. However, 56% of patients with the head and neck distribution of AD were positive to the extract compared to only 36% of the patients without the head and neck distribution of AD. None of the patients with seborrheic dermatitis or the healthy controls were positive. All readings were done in a controlled double-blind design. In summary, positive SPT to Malassezia is found primarily in adult patients with AD and especially in adult patients with the head and neck distribution of AD . TABLE 1 | SPT reactions to the Malassezia yeasts in patients with AD (ii) Specific IgE antibodies to the Malassezia yeasts. : The first paper describing specific IgE antibodies to the Malassezia yeasts was published in 1991 . Specific antibodies were found especially in patients with AD and less commonly in patients with other atopic manifestations. This was later verified by other groups using various extract from Malassezia . In the study by Broberg at al., specific IgE antibodies were found in eight children with AD, five in the age group from 16 to 21 years, and only one patient with rhinitis and/or asthma . In another study with children (4 to 16 years old), 22% of patients with AD had IgE antibodies against Malassezia . In these patients, the AD was generally worse than in those without specific Malassezia IgE antibodies. The head and neck distribution of AD was also more common and the onset of AD occurred earlier in life than in specific IgE-negative children. Back et al. found in a study of 74 adult patients with AD that 44% had specific Malassezia IgE antibodies . No antibodies were found in patients with seborrheic dermatitis or healthy controls . The presence of specific Malassezia IgE antibodies correlated with elevated total IgE levels. In the study by Kim et al., 68% of patients with the head and neck distribution of AD had specific IgE antibodies against Malassezia . The clinical severity and total IgE level in serum for these patients were greater in patients with Malassezia-specific IgE antibodies than in patients with a negative response. Devos and van der Valk also studied the presence of specific IgE antibodies in patients with AD . However, In contrast to their SPT results, mentioned above, they found that 100% of patients with the head and neck distribution of AD had Malassezia-specific IgE antibodies compared to only 14% of AD patients without the head and neck distribution. Savolainen et al. recently described specific and nonspecific humoral cellular and cytokine responses to Malassezia and Candida in AD patients . Fifteen AD patients and 7 healthy controls were included. The total IgE level in serum was clearly higher in AD patients than in controls. The Malassezia- and C. albicans-specific IgE level was also significantly increased in AD patients compared to controls. Malassezia was a stronger inducer of IgE responses, and C. albicans more strongly induced IgG responses. However, there were no significant differences in Malassezia- and C. albicans-specific IgG levels between the groups. The rest of the study is discussed in the section on cellular and interleukin responses (see below). Mayser and Gross have studied the presence of specific IgE antibodies against various Malassezia species . The presence of specific antibodies against M. sympodialis, M. furfur, and the commercially available Malassezia extract was studied in patients with AD, seborrheic dermatitis, pityriasis versicolor, and hymenoptera allergy and healthy controls. The allergens alone as well as various combinations were included. It was found that 35% of AD patients had specific IgE antibodies against one or several allergens, 12% of patients with seborrheic dermatitis were positive, 8% of patients with hymenoptera allergy were positive, and none of the patients with pityriasis versicolor or the healthy controls were positive. For patients with AD, the largest number of positive results was seen in patients with the head and neck distribution of AD. There was also strong correlation between specific IgE antibodies and the total IgE level. Differences in the type and extent of sensitization were also found among the Malassezia yeasts. M. sympodialis and the commercially available Malassezia extract were almost equally positive, while the frequency of sensitization to M. furfur was lower. The commercially available Malassezia was typed using the new classification and found to be M. sympodialis. M. sympodialis is more common on the skin of patients with AD and it therefore a better allergen for testing than M. furfur. In the Swedish multicenter study, mentioned above, specific serum IgE anribodies to the Malassezia extract was found in 45% of patients with AD; again, patients with the head and neck distribution of AD had the highest number of reactions (55%); 40% of the AD patients had both specific Malassezia IgE antibodies and positive SPT. Patients with seborrheic dermatitis and healthy controls were negative. RMal f 1, rMal f 5, and rMal f 6 specific serum IgE antibodies were found in 19, 28, and 25%, respectively. Specific serum IgE -antibodies to the recombinant Malassezia antigens, alone or together, were found in 40% of the patients. In contrast to the extract, moderately strong correlations were found between SPT (in millimeters) and specific serum IgE level for rMal f 1, rMalf 5, and rMal f 6. Specific Malassezia IgE antibodies were found in 20 to 100% of patients with AD. The prevalence was lower in children, higher in adults, and highest in adult patients with the head and neck distribution of AD. Table summarizes the specific Malassezia serum IgE antibodies found in various studies. TABLE 2 | Specific IgE serum antibodies to the Malassezia yeasts in patients with AD (iii) Characterization of antigens. : Several IgE binding components of Malassezia have been isolated . Three major allergen components from Malassezia were later identified using monoclonal antibodies . Two of them, a 67-kDa component and a 37-kDa component, were proteins, and one, a 14-kDa component, was probably of carbohydrate origin. In other studies, allergen cross-reactivity between Malassezia and C. albicans has been found . According to Zargari et al., the 14-kDa component may be found in both yeasts . Jensen-Jarolim et al. have also tried to characterize various components of Malsssezia, and they described 9- and 15-kDa components that probably are nonprotein carbohydrate components of the cell wall . However, Savolainen et al. have described protein components common to both Malassezia and C. albicans . In 1997, Lintu et al. have described IgE antibodies to both protein and mannan antigens of Malassezia in sera from patients with AD . Lintu et al. found other major components from those found by Zargari et al. In another paper, Lintu et al. described cross-reacting IgE and IgG antibodies to both Malassezia, C. albicans, Saccharomyces cerevisiae, and, to some extent, Cryptococcus albidus . In the IgE response, the main cross-reacting pattern was the mannan region and Malassezia was the most potent inhibitor of the IgE-binding components. In contrast, C. albicans was found to be the most important inducer of IgG antibodies. There are several explanations for the different results obtained. The technique used is important, and the strain of Malassezia used is also important. The length of the culture period may also be of importance . The protein content of the extract was highest after 2 days but gradually declined after day 4. On the other hand, the carbohydrate content remained fairly constant. The stability of the extracts is also influenced by the storage period and conditions . The stability of Malassezia protein extracts stored, even at 4C, for more than 1 month was poor. The 9-kDa component was more stable than the 20-kDa component and, the 96-kDa bands in particular were far more labile. In none of these reports were the Malassezia yeasts classified according to the new taxonomy. Antigenic bands against both M. sympodialis and M. furfur were found in sera from patients with AD . The molecular masses of the most important bands were 15, 22, 30, 37, 40, 58, 79, 92, 99, and 124 kDa for M. sympodialis and 15, 25, 27, 43, 58, 92, 99, and 107 kDa for M. furfur. Control sera were negative. When M. furfur-positive sera were preincubated with M. sympodialis, the clear band of 15 kDa disappeared, indicating a possible cross-reaction of the antibodies. However, this study clearly demonstrates that there are differences between different strains of Malassezia and that differences in results may be due to the use of different strains. Schmidt et al. have identified the complete cDNA sequence and expression of one major allergen protein of Malassezia . According to the WHO/IUS nomenclature, this recombinant allergen is named Mal f 1. Yasueda et al. have isolated two additional major recombinant allergens, named Mal f 2 and Mal f 3 , and Onishi et al. have isolated the Mal f 4 recombinant allergen . Recently, Rasool et al. isolated and characterized Mal f 7, Mal f 8, and Mal f 9 . However, like the allergen Mal f 1, these three recombinant allergens were prepared from the ATCC strain 42132 of M. furfur (formerly P. orbiculare). This Malassezia strain is not M. furfur but M. sympodialis . I was the one who donated this strain to the ATCC in 1979, and I have now retyped it according to the new taxonomy. According to this, I have also retyped it as M. sympodialis. These recombinant allergens should therefore be named Mal sy (1 etc.). The reference strain used by Yasueda et al. and by Onishi et al. are from another reference laboratory (strain 2782 from Teikyo Institute for Medical Mycology) and is named M. furfur. I have not personally confirmed that this species is M. furfur. Recently, Koyma et al. isolated three major components of M. globosa from extracts of this yeast . These antigens were named Mal g 46a, Mal g 46b, and Mal g 67. Mal g 46b was found to be the major antigen, and it reacted only with sera from patients with AD and with none of the sera from patients with seborrheic dermatitis, pityriasis versicolor, or chronic urticaria or healthy controls. Mal g 46b reacted with 69% of the sera from patients with AD. Mal g 46b reacted with 83% of sera containing IgE antibodies to M. globosa. In a lectin blot with concanavalin A, both Mal g 46 a and Mal g 46 b reacted with concanavalin A, indicating that these antigens were glycoproteins containing mannose chains in the molecules. When species specificity was tested, the antibody reacted strongly with M. globosa. A slight reaction was seen with M. restricta, but no reaction was seen with M. sympodialis, M. furfur, or M. slooffiae. Also, no reaction was seen with C. albicans. In the Japanese population, this antigen may be more important that antigens against, e.g., M. sympodialis, because M. globosa is more commonly cultured from AD patients in Japan . (iv) Cellular and cytokine responses. : Both in vitro and in vivo studies of cellular and cytokine responses have been performed. In parallel to SPT, several skin patch test experiments have been performed for investigation of a delayed-type hypersensitivity to the Malassezia yeasts. Rokugo et al. found many patients with AD who were patch test positive to a Malassezia extract . However, they used the chamber scarification method, and Kieffer et al., using a standard test procedure without scarification, were not able to reproduce these large numbers of positive results . Patch test reactions are often difficult to read in patients with AD. To evaluate the results of primarily aeroallergens, a special technique called the atopy patch test (APT) was developed . Briefly, the stratum corneum is stripped approximately 15 times with cellophane tape to achieve a slight increase in its permeability. The allergens are then applied on paper disks in Finn chambers (8 mm; Epitest Ltd Oy, Tuusula, Finland) on normal-appearing skin of the back and covered with a nonirritating tape. The disks are removed after 48 h, and the skin reaction is evaluated 24, 48, and 72 h after that. In a study by Tengvall-Linder et al., using this technique, 8 of 15 adult patients with AD were patch test positive to a Malassezia extract . However, none of 8 patients with seborrheic dermatitis or 15 healthy volunteers were positive. The AD patients with positive APT reactions had significantly higher specific Malassezia serum IgE antibody levels than did AD patients with negative reactions. Malassezia-specific IgE, but not total IgE, levels correlated with the degree of APT response at 48 h in the whole group of AD patients. However, specific serum IgG levels did not differ significantly between the two groups of AD patients. In the above-mentioned Swedish multicenter study, patients with AD, patients with seborrheic dermatitis, and healthy controls were also subjected to APT with both the crude Malassezia extract and extracts containing the recombinant allergens Mal f 1, Mal f 5, and Mal f 6, both alone and in combination. Positive APT reactions were found in 38% of patients with AD, 7% of patients (1 patient) with seborrheic dermatitis, and none of the healthy controls. For the AD patients, positive APT reactions to rMal f 1, rMal f 5, and rMal f 6 were found in 6, 15, and 10%, respectively. No positive reactions were seen in patients with seborrheic dermatis or in healthy controls. The APT reactions differed among these three recombinant allergens. rMal f 5 gave the strongest reactions, and 75% of these patients were also SPT positive and had specific serum IgE levels. Kroger et al. have measured the synthesis of various lymphokines as well as the IgE synthesis from peripheral blood mononuclear cells after stimulation with a Malassezia extract . They found that IL-4 and IL-10 synthesis was increased and IL-2 and gamma interferon (IFN-gamma) synthesis decreased in patients with AD and with RAST+-specific serum IgE antibodies compared to healthy controls. IgE synthesis in the supernatants was only increased in patients with AD and with RAST+-specific serum IgE antibodies. The proliferative response in peripheral blood mononuclear cells (PBMC) to a Malassezia extract has been studied in 10 AD patients with specific serum IgE antibodies to Malassezia and in 6 healthy controls with no specific serum Malassezia IgE antibodies . The PBMC response to Malassezia was significantly higher in the AD group than in the control group. A response to a C. albicans extract was either not measurable or very low. T cells were obtained from peripheral blood and the skin by cloning from one AD patient with severe lesions and a high titer of specific Malassezia serum IgE antibodies. Several T-cell clones (TCC) from lesional skin showed a strong response to a Malassezia extract, whereas only a few TCC from peripheral blood were positive and none from nonlesional skin were positive. The response to a C. albicans extract was very low. In the same study, freshly isolated PBMC produced only low cytokine levels of IL-4, IL-5, and IFN-gamma after 12 days of incubation with a Malassezia extract. When blood- and skin-derived Malassezia-reactive CD4+ TCC were tested for their production of IL-4, IL-5, and IFN-gamma after stimulation with phytohemagglutinin (PHA) and an anti-CD3 antibody, a variation of cytokine profile was produced. Th1, Th2, and Th0, cytokine profiles and a Th2/Th0 cytokine profile were found. The Th2-like profile of Malassezia-reactive TCC derived from lesional skin indicates that Malassezia may play a role in maintaining IgE-mediated skin inflammation in AD. Incubation of TCC with a Malassezia antigen for 12 days induced measurable production of IL-5 but not IL-4 or IFN-gamma. Nonstimulated and Malassezia-stimulated PBMC expressed similar levels of IFN-gamma mRNA at all time points. Stimulation of PBMC with Malassezia induced a time-dependent increase in gene expression for both IL-5 and IL-13, with maximal expression at 24 h. Gene expression for these cytokines was hardly detectable in nonstimulated PBMC. Stimulation with anti-CD3 antibody induced a profound increase in the expression of genes for IFN-gamma, IL-5, and IL-13 at 16 to 24 h. The same group has compared the immune response of AD patients to a Malassezia extract with that of healthy controls . The proliferative response of PBMC was significantly higher in the AD patients than in the healthy controls. Significantly higher levels of IL-5 was produced by PBMC from AD patients compared to healthy controls. Malassezia-reactive T-cell lines obtained after stimulation of PBMC by Malassezia antigen produced significantly higher levels of IL-4 and IL-5 after stimulation with anti-CD3 antibody and showed a higher IL-4/IFN-gamma ratio in AD patients than in healthy controls. These results indicate that the Malassezia yeasts may play a role in maintaining skin inflammation in AD. Recently, Buentke et al. looked at the possible interaction of immature and mature antigen-presenting monocyte-derived dendritic cells (MDDCs) with Malassezia yeast cells and different Malassezia allergenic components . Internalization of Malassezia yeast cells and yeast components by immature MDDCs was found, and the results showed that a median of 94% of the immature CD1a+ MDDCs were Malassezia whole-cell extract positive, 81% were positive to the recombinant antigen rMal f 5, and 93% were mannan positive. However, mature MDDCs were significantly less positive. The binding of the Malassezia extract and mannan was inhibited in a dose-dependent manner by methyl-alpha-d-mannopyranoside, suggesting uptake via the mannose receptor. This study demonstrated that human immature CD1+ MDDCs can efficiently take up Malassezia and allergenic components from the yeast in the absence of IgE, implying that sensitization of AD patients to Malassezia can be mediated by immature dendritic cells in the skin. In their paper, the authors claimed that the Malassezia species used was M. furfur. However, the species was the above-mentioned ATCC isolate 42132, which both by myself and by Mayser and Gross , has been retyped as M. sympodialis. Specific IgE antibodies to M. sympodialis are more commonly seen in AD patients than are specific IgE antibodies to M. furfur . Savolainen et al. have recently tried to characterize the Malassezia- and C. albicans-specific and -nonspecific humoral, lymphoproliferative, and cytokine responses in AD patients compared to healthy controls . Malassezia- and C. albicans-specific IgE levels and Malassezia-induced PBMC proliferation were elevated in AD patients. Both Malassezia- and C. albicans-specific IgE antibodies were detected in AD patients but not in healthy controls. The median titer of specific IgE response was significantly higher for Malassezia than for C. albicans. The IL-4/IFN-gamma ratio induced by Malassezia was higher than that induced by C. albicans. The PHA-induced IL-2 and IL-4 responses and the C. albicans-induced IL-5 response and IFN-gamma response were elevated in AD patients. Several correlations were found between total and yeast-specific IgE levels in serum, Malassezia-specific lymphoproliferation, PHA-induced IL-2, IL-4, and IL-5 levels, and C. albicans-induced IL-5 levels. The cytokine profiles found in this study for Malassezia were in parallel with and close to the results found by Tengvall-Linder et al. , and they support the role of a combination Th0-, Th1-, and Th2-responding Malassezia cells in the pathogenesis of AD. Differences in Malassezia responses obtained may occur because different Malassezia species were studied. In the study by Savolainen et al., Malassezia appears to be associated more with IL-4 responses and C. albicans appears to be associated more with IFN-gamma responses . So far, the results of cytokine production of lymphocytes have been mentioned. In a paper by Watanabe et al., cytokine production by human keratinocytes after stimulation with cells of various Malassezia species has been described . The Malassezia species studied were M. pachydermatis, M. furfur, M. slooffiae, and M. sympodialis, and the cytokine production studied was that of IL-1beta, IL-6, IL-8, monocyte chemotactic protein-1 (MCP-1), and tumor necrosis factor-alpha (TNF-alpha). Cytokine production by keratinocytes after coculture with M. furfur was low or undetectable. In the supernatants cocultured with the other Malassezia yeasts, IL-1beta, IL-6, IL-8, and TNF-alpha were detected whereas the MCP-1 levels were low or undetectable. The cytokine levels were highest with M. pachydermatis. M. pachydermatits is cultured primarily from animals, and the dermatitis seen in animals is often more inflammatory than the human Malassezia-related diseases. There were even differences between the Malassezia yeasts in the number of yeast cells that were needed to start cytokine production. Further studies are needed to explain differences in results between AD and other skin diseases as well as differences between the various Malassezia yeasts. A number of results for lymphocyte stimulation responses and cytokine production have been reported in the literature. Kroger et al. found an increase in IL-4 and IL-10 production and a decrease in IL-2 and IFN-107 production in AD patients after stimulation with Malassezia extract; Tengvall-Linder et al. and Savolainer et al. found an increase in IL-4 and IL-5 production. Treatment. : In a study by Gupta et al., the in vitro susceptibility of the seven different Malassezia species against the azoles, ketoconazole, itraconazole and voriconazole and the allylamine terbinafine was tested . All of the Malassezia yeasts were highly sensitive to all three azole drugs. However, only M. pachydermatis and M. sympodialis were highly sensitive to terbinafine. A double-blind crossover study by Clemmensen and Hjorth has shown that in patients with AD, oral ketoconazole is statistically more effective than placebo . In an open study by Back et al. ketoconazole at 200 mg once daily for 2 months followed by 200 mg twice weekly for 3 months resulted in an improvement in clinical score in AD patients . There was also a reduction in the levels of both total IgE and specific IgE for Malassezia and C. albicans. There was a correlation between the reduction in the levels of total IgE and specific IgE for Malassezia but not between reduction in clinical score and the level of specific IgE for Malassezia. However, for specific C. albicans, there was a correlation between a reduction in the specific IgE and total IgE levels and clinical score. Today oral ketoconazole should not be used for long-term treatment due to the risk of liver toxicity. If ketroconazole is used for more than 14 days, liver enzymes should be monitored before the start of treatment and every 14 days during treatment. The newer azoles, the triazoles fluconazole and itraconazole, have less risk of side effects; although they are more expensive, they should be the drugs of choice. In a double-blind placebo-controlled study, Back and Bartosik have treated 29 AD patients with oral ketoconazole at 200 mg once daily or placebo for 3 months . In the ketoconazole group, the levels of total IgE and specific IgE to both Malassezia and C. albicans decreased significantly. The patients were allowed to use topical corticosteroids during the study period, and the clinical score decreased equally in both groups. However, the improvement was correlated with the use of topical corticosteroids in the control group but not in the ketoconazole group. Lintu et al. have, in a double-blind, controlled study, treated 80 AD patients with oral ketoconazole at 200 mg or placebo once daily for 30 days . Yeast cultures from the skin and pharynx, SPT and specific IgE antibodies for Malassezia, C. albicans, and S. cerevisiae, the total IgE level in serum, and the clinical severity score was assessed at baseline and after 1 and 3 months. A significant improvement was seen in the clinical score after 30 days of treatment in the ketoconazole group but not in the placebo group. The number of positive Malassezia cultures was also reduced significantly in the ketoconazole group but not in the placebo group. The clinical response was highest in female patients with positive cultures for Malassezia and/or C. albicans. In this study there was a slight, but not statistically significant, reduction in the total IgE level after 30 days of treatment with ketoconazole. For all other parameters studied, there were no significant difference between baseline and the 2- or 3-month follow-up visits. This may be due to a shorter treatment period than in earlier trials. In one study, the positive effect of oral itraconazole in the treatment of patients with AD has been mentioned . Topical treatment with bifonazole, in an unknown number of patients, has also been effective in the treatment of AD . In an open study, the combination of hydrocortisone and miconazole was effective in the treatment of AD . However, in a double-blind controlled study of the effect of hydrocortisone and miconazole compared with hydrocortisone alone in the treatment of AD, no statistically significant difference was found and both treatments were effective . Oral treatment may be more effective because it more effectively eradicates the Malassezia yeasts located deep in the follicle. The number of treatment studies using antifungal therapy is small, and new well conducted, controlled and double-blind trials are needed to confirm if antifungal therapy is beneficial in the treatment of some patients with AD. Candida : More than 50 different Candida species have been described . Many species have been isolated from human sources, but eight species are dominant these are C. albicans, C. glabrata, C. guilliermondii, C. kefyr, C. krusei, C. parapsilosis, C. tropicalis, and C. doblinensis. All the pathogenic Candida species multiply primarily by the production of buds from blastospores. The sizes and shapes of blastospores are sometimes characteristic for a species, but the gross microscopic appearance is more or less similar. The cell wall is multilayered, and the cell membrane is a typical eukaryotic phospholipid bilayer with frequent invaginations. The most commonly encountered pathogenic Candida species usually grow well in aerobic cultures base on rich or poor nutrient media with a pH in the range of 2.5 to 7.5 and at a temperature in the range of 20 to 38C. Distribution on normal skin. : In persons working in hospitals, in patients with skin diseases, in cancer patients, in patients with immune defects, and in individuals living in a warm and humid climate C. albicans may be cultured even from glabrous skin. Other Candida species, e.g., C. parapsilosis, are often cultured from the toe cleft. Distribution on mucous membranes. : Isolation of Candida from the gastrointestinal tract or, in women, the vagina does not automatically imply disease because many healthy persons are colonized with Candida in these areas . An prevalence of 3 to 65% in asymptomatic persons has been reported . Under the influence of various predisposing factors, the prevalence may increase . The most commonly cultured Candida species is C. albicans, but C. glabrata, C. tropicalis, C. parapsilosis, and C. krusei may also be found. The Candida yeasts as pathogens. : Infections with Candida occur not only due to the presence of individual or environmental predisposing factors but also due to the virulence of the infecting organism . Important predisposing factors are human immunodeficiency virus infection, other immunosuppresive disorders, diabetes mellitus, age, pregnancy, hormonal dysfunction, contraceptive pills, and the use of drugs, e.g., oral antibiotics, oral corticosteroids, and immunosuppresive drugs. The presence of skin diseases often predisposes to skin infections with Candida. Theoretically, Candida may be involved in diseases due to infections or to allergic reactions . In the host defense against Candida, both nonspecific immune mechanisms (e.g., an intact skin barrier or mucous membrane, presence of other microorganisms, polymorphonuclear leukocytes, macrophages, complement, and NK- and/or CD16-positive cells) and specific immune mechanisms (e.g., serum antibodies and cell-mediated immunity) are involved . Chronic mucocutaneous candidiasis and systemic candidiasis are both associated with defects in the cell-mediated immune response . Virulence factors in Candida that are important for an infection are, e.g., adhesion to epithelial or mucosal cells, proteinases, hypha formation, propensity to switch phenotype, and surface hydrophobicity. Infections by various members of the genus Candida have been reported from almost all anatomical sites ; they can be divided into superficial and deep infections. It is not within the scope of this article to give a detailed description of Candida infections. The most commonly seen skin infections are intertrigo, cheilitis, diaper rash, balanitis, paranychia, nail infections (most commonly secondary to paranychia), and folliculitis. Superficial infections on mucous membranes include vaginitis, balanitis (which may involve both skin and mucous membranes), stomatitis, glossitis, and other superficial infections of the gastrointestinal tract. In deep candidiasis, almost all organs may be involved. Many times it is impossible to make a distinction between a primary systemic Candida infection and disseminated candidiasis. C. albicans contains several antigenic components capable of stimulating immediate hypersensitivity responses. However, although Candida, and especially C. albicans, often is mentioned in relation to various allergic reactions, the majority of these reports are anecdotal and not reports of evidence-based studies. In patients with cutaneous Candida infections, sterile eczematous lesions may occur in other areas of the skin . This type of reaction, often called "id" reaction, is also seen in dermatophyte infections, especially tinea pedis due to Trichophyton mentagrophytes (var. interdigitale). Although not evidence based, this type of reaction is called an allergic reaction to fungal material because it disappears when the fungal skin infection heals. The role of an allergic type I reaction to C. albicans has also been mentioned for other skin diseases, e.g., chronic urticaria and psoriasis. Other reports, also of anecdotal origin, in the literature claim that Candida allergy may be a major cause of manifold feelings of lack of well-being, e.g., anxiety, depression, diarrhea, premenstrual tension, poor intellectual function, and many more . These reactions should be due to the presence of Candida in the gastrointestinal tract or in the blood. These reports are not evidence based, but they may be harmful because they often recommend antimycotic therapy. Non-evidence-based antimycotic therapy should of course be avoided due to the risk of allergic drug reactions, drug interactions, and the risk of development of fungal drug resistance. Atopic Dermatitis and the Candida yeasts : Distribution of Candida. | There is little information about the colonization of the skin in patients with AD. The available literature describes that Candida species, and especially C. albicans, have been cultured more frequently from both normal and lesional skin in patients with AD than from healthy individuals . However, there are no reports about any correlation between the presence of C. albicans on the skin, the activity of AD, and/or the effects of antifungal therapy. Candida has been cultured more frequently from the gastrointestinal tract in patients with AD than in healthy controls . In a study by Buslau et al., Candida species were cultured more frequently from feces of patients with AD and those with psoriasis ; C. albicans was cultured in 90% of the cases where Candida species were found. The authors cultured yeasts in 70% of patients with AD compared to 54% of healthy controls . In a paper by Savolainen et al., C. albicans was cultured from the nasopharynx in 31 of 52 patients (59%) with AD . Skin cultures for fungi were negative. A correlation between the severity of AD and the levels of specific IgE antibodies to C. albicans was seen only in AD patients with positive nasopharyngeal cultures. Immunology. (i) Skin prick test, with C. albicans and other yeasts, for type I hypersensitivity. : In a study of 117 adult patients with AD, 28 patients with allergic rhinitis, and 39 healthy controls, 52% of patients with AD were SPT positive to a C. albicans extract compared to 22% of patients with allergic rhinitis and 5% of healthy controls . Positive SPTs were more commonly seen in patients with severe AD. Positive SPTs to S. cerevisiae, Malassezia, and molds showed a specific association with the presence of specific C. albicans IgE antibodies. In a study from Japan, approximately 30% of patients with AD were SPT positive to a C. albicans extract and approximately 40% of AD patients were SPT positive to a Malassezia extract . There was no difference between AD patients with or without simultaneous atopic respiratory disease. However, only approximately 10% of healthy controls were SPT positive. There was a positive correlation between positive results in SPT and specific RAST for both C. albicans and Malassezia. In another study by the same group, AD patients who showed larger SPT responses to a C. albicans extract tended to display lower lymphocyte proliferative responses . In a study by Morita et al. , the role of C. albicans in AD was evaluated and various immunological parameters were compared in patients with AD , patients with allergic rhinitis , and nonatopic subjects . The SPT response was significantly larger in the AD group than in the non-AD group. Nissen et al. prick tested 15 patients with the head and neck distribution of AD or secondary infected AD with both culture filtrates and cellular extracts of C. albicans and Malassezia . Culture filtrates contain metabolites and cell wall components, whereas cellular extracts contain predominantly cell walls and probably fewer metabolites. Of 15 patients, 4 (27%) were SPT positive to C. albicans cellular extract and 2 or 13% were SPT positive to the culture filtrate. Of 15 patients, 9 (60%) were SPT positive to a commercially available Malassezia extract. The authors also included the leukocyte histamine release test in this study. The results were more or less comparable to the SPT results. Kortekangas-Savolainen et al. have studied the SPT results to an extract of S. cerevisiae in 226 patients with AD, 50 patients with allergic rhinitis and/or asthma, and 173 nonatopic controls . A positive SPT result was seen in 94% of patients with severe AD, 76% of patients with moderate AD, and 25% of patients with mild AD. Patients with allergic rhinitis and/or asthma and nonatopic controls showed positive results in 8 and 2% of tests, respectively. A significant correlation was found between total serum IgE levels and SPT results. There was also a parallel between SPT result with Malassezia and C. albicans, indicating a cross-reactivity. Positive SPT results to Candida were found in 22 to 94% of AD patients were lower in healthy controls. The largest number of positive reactions was seen in patients with severe AD. Previous studies have found positive SPT reactions to C. albicans in 52% , 25 to 94% , 30% , and 27% of patients with AD. (ii) Specific IgE antibodies to C. albicans and other yeasts. : One of the first reports of production of specific IgE antibodies against C. albicans was presented by Okudaira et al. in 1983 . They found the production of specific C. albicans IgE antibodies in patients with AD and/or asthma but not in controls. In the above-mentioned paper by Savolainen et al., the severity of AD was associated with the production of C. albicans-specific IgE antibodies . In patients with the head and neck distribution of AD, there was a significant correlation between specific IgE antibodies and severity of AD only in patients with gastrointestinal growth of C. albicans. This pattern was also the same for AD in general but not for AD located only to the body or extremities. The specific summarized IgE staining responses to all allergens (both mannan and proteins) in immunoblot analyses, as well as the total serum IgE levels, were highest in patients with severe AD. In a study by Tanaka et al., the production of specific IgE antibodies against C. albicans and Malassezia was significantly higher in patients with AD than in healthy controls, and the levels were also higher in patients with AD (49 patients) alone than in patients with both AD and allergic respiratory disease (48 patients) . In another study, Tanaka et al. found that AD patients with higher total IgE levels in serum as well as higher specific C. albicans IgE antibody levels tended to display lower lymphocyte proliferative responses than did patients with lower levels, although the correlation was not statistically significant . Back et al. found specific IgE antibodies to C. albicans in 21 of 69 patients with AD (33%) but none in patients with seborrheic dermatitis or in healthy controls . The correlation between specific IgE antibodies to C. albicans and Malassezia was strong: 20 of 28 patients with specific Malassezia IgE antibodies had IgE antibodies to C. albicans, and 20 of 21 patients with specific C. albicans IgE antibodies also had IgE antibodies to Malassezia. There was also a correlation between the total IgE level and the levels of specific IgE antibodies to both C. albicans and Malassezia Matsumura et al. studied total IgE and production of specific IgE antibodies to C. albicans in 46 patients with AD, 41 patients with allergic respiratory disease, and 42 healthy controls . Both patients with AD and those with allergic respiratory disease showed a significantly higher total IgE than did controls. However, specific C. albicans IgE antibody levels were significantly higher in patients with AD than in patients with allergic respiratory disease. Patients with AD also reacted to a larger number of allergens than did those with allergic respiratory disease. Doekes et al. found a highly significant correlation between levels of anti-Malassezia IgE and IgE reacting with extracts of C. albicans . Of 128 sera from AD patients, 34 sera reacted positively with both yeast extracts, 38 reacted with Malassezia but not with C. albicans extract, and only 1 of the 56 anti-Malassezia-negative sera showed a very week reaction with C. albicans. The correlation was due to a marked cross-reactivity, as shown by inhibition enzyme-linked immunosorbent assay. However, Malassezia allergens were more potent inhibitors than were the corresponding C. albicans components, suggesting than the IgE antibodies in AD result from sensitization to Malassezia and then cross-react with C. albicans. In a report by Nermes et al., a RAST test for determination of IgE antibodies to C. albicans mannan was described . Of 78 patients with AD, 53 (68%) showed elevated levels compared to healthy controls, and there was a significant correlation with the severity of AD. Of 30 patients with asthma 16 (53%) had elevated IgE mannan C. albicans antibody levels in serum. However, 12 of these 16 patients with a positive result also had AD. Of 32 patients with allergic rhinitis but without AD, 12 (38%) had elevated mannan IgE antibody titers. This RAST method is a sensitive assay for determination of polysaccharide-specific IgE C. albicans antibodies in patients with AD. As part of a more extensive immunological study of patients with AD, Kawamura et al. measured total IgE and specific C. albicans IgE antibody levels in serum in 15 adult patients with AD . All patients had elevated total IgE levels, and 12 of 14 patients showed elevated specific C. albicans IgE antibody levels. Savolainen et al. studied the presence of IgE-, IgA-, and IgG-specific yeast antibodies (total) and yeast mannan-specific IgE, IgA, and IgG antibodies in patients with atopic diseases . Of 20 patients, 14 were C. albicans SPT positive, 1 was C. utilis SPT positive, and 1 was Rhodotorula rubra SPT positive. A total of 12 patients with AD and 2 with rhinitis had specific C. albicans IgE antibodies in immunoblot analyses, and 10 (9 with AD and 1 with rhinitis) of 14 tested patients had C. albicans-specific mannan IgE antibodies. Nine patients had specific IgE antibodies to C. utilis, 8 patients had specific IgE antibodies to R. rubra, 4 patients had specific IgE antibodies to S. serevisiae, and 4 patients had specific IgE antibodies to Cryptococcus albidus. IgA-specific antibodies were detected in 25 to 85% of the patients; the highest level was for C. albicans, and the lowest level was for R. rubra. IgA mannan-specific antibodies were detected in 18 to 100%; the highest level was for C. albicans and the lowest level was for Cryptococcus albidus. IgG-specific antibodies were detected in 20 to 95%; the highest level was for S. cerevisiae, and the lowest level was for C. utilis. IgG mannan-specific antibodies were detected in 74 to 100%; the highest level was for R. rubra, and the lowest level was for S. cerevisiae. Cross-reactivity between C. albicans and the other yeasts may probably explain several of the positive reactions to yeasts other than C. albicans. In a study by Nissen et al., IgE-specific antibodies to C. albicans and Malassezia were studied using immunoblotting . Of AD patients, 93% had specific Malassezia IgE antibodies, 47% had specific IgE antibodies to a C. albicans cellular extract, and 60% had specific IgE antibodies to a culture filtrate extract of C. albicans. Savolainen et al. studied specific IgE and IgG antibodies to a protein and mannan antigen from C. albicans in sera from 10 patients with AD and 7 healthy controls . Levels of both mannan- and protein-specific IgE antibodies were elevated in AD patients compared to controls. The anti-mannan IgG responses were higher than the anti-protein IgG responses. However, no significant differences between AD patients and controls were seen. Morita et al. found that the total IgE level was significantly higher in patients with AD (37 patients) than in both patients with allergic rhinitis (23 patients) and healthy controls (42 subjects) . C. albicans-specific IgE was found in 85% of AD patients compared to only 9% of patients with allergic rhinitis and 5% of healthy controls. Scalabrin et al. investigated specific IgE responses to various fungi and house dust mites in 73 patients with AD (including 16 children), 156 patients (54 children) with asthma, and 212 non-asthmatic patients (102 children) . Patients with AD had significantly higher total IgE levels than did those in the other two groups. Specific IgE antibodies to both C. albicans and Malassezia were detected in 75 and 54%, respectively. However, for Malassezia, significantly larger numbers of adult patients (65%) than of children (13%) had specific IgE serum antibodies. There was also a statistically significant correlation between the total serum IgE level and the level of specific IgE to both C. albicans and Malassezia as well as between specific C. albicans and Malassezia IgE antibodies, indicating a cross-reaction. Adachi et al. found that the percentage of AD patients with specific C. albicans IgE antibodies was significantly higher for patients with severe symptoms and high total serum IgE levels . This was also found in patients with specific Malassezia IgE antibodies. Patients with the head and neck distribution of AD had higher specific IgE titers to both C. albicans and Malassezia. Kimura et al. found high specific C. albicans IgE antibody levels in patients with AD ; specific IgE antibodies were found in 87% of patients with AD compared to only 15% of patients with asthma. In a study by Broberg et al. of children with AD, specific C. albicans IgE antibodies were found in only 16% (9 of 57) . Lindgren et al. measured the levels of total IgE and specific IgE antibodies to a variety of external allergens in 119 children with AD . Of these children, 47% had specific C. albicans IgE antibodies and 34% had specific Malassezia IgE antibodies. RAST positivity to all allergens increased with age. However, this was most pronounced for Malassezia, where 43% of 10-year-old children had specific Malassezia IgE antibodies. The RAST-positive children generally had worse AD than did the RAST-negative children. Children who were RAST positive to C. albicans and/or Malassezia not only had a more severe AD than did RAST-negative children but also more often had AD with a head and neck distribution. The frequency of specific serum Candida IgE varied from study to study. In general, the number of patients with specific antibodies was larger for adults than for children, for patients with severe AD than for those with mild AD, and for patients with a high total IgE level. Another explanation for the variation in the results obtained in various studies is the difference in the antigen preparation. Table summarizes the specific IgE antibody responses found in the literature. Nermes et al. have studied the antibody response to C. albicans mannan in children younger than 5 years with and without AD . There was no difference in the production of C. albicans IgG- and IgA-specific antibodies between the two groups. IgM anti-mannan levels were higher in 5-year-old children with AD. The authors concluded that children are not more susceptible to bacterial infections on the basis of poorer ability to produce antibodies to polysaccharide antigens. TABLE 3 | Specific IgE serum antibodies to C. albicans in patients with AD (iii) Characterization of antigens. : Several of the papers dealing with characterization of C. albicans antigens have already been mentioned in the section on Malassezia antigens due to the cross-reactivity between these two yeasts. Ishiguro et al. have identified various antigens from C. albicans that reacted with IgE antibodies from sera of 57 patients with allergic diseases . The 175-, 125-, 46-, 43-, and 37-kDa antigenic components reacted most frequently. The 46-, 43-, and 37-kDa antigens were recovered in cytoplasmic fractions and had homology to the S. cerevisiae glycolytic enzymes enolase, phosphoglycerate kinase, and aldolase, respectively. The 175- and 125-kDa antigens were not recovered in any fraction. This result suggests that the levels of IgE antibodies against only small parts of their epitopes are elevated in the allergic patients and that, e.g., enolase may be a major immunodominant protein in both allergies and fungal infections. Savolainen et al. found the most dominant protein allergens from C. albicans to be 27- and 46-kDa proteins . However, these allergens changed when the same patients were studied 5 years later and the IgE responses were found to be directed mostly against allergen bands with molecular masses of 34, 29, 20, and 13 kDa. IgA and IgG responses were directed mainly against the polysaccharide mannan. Kortekangas-Savolainen et al. have tried to characterize the IgE binding components of S. cerevisiae . The found that the most frequent staining was found with a 48-kDa band. When the staining pattern of the 48-kDa band, a 45-kDa band, and mannan was compared with C. albicans allergens or purified S. serevisiae enolase, a simultaneous binding was seen with the 48-kDa band of S. cerevisiae and the 46-kDa band of C. albicans and enolase whereas the 45-kDa band was not associated with the 46-kDa band of C. albicans or the purified enolase. Nermes et al. analyzed cross-reactivity between C. albicans and S. cerevisiae mannans . IgE binding to all three studied mannans from both yeasts was seen simultaneously. However, the strongest response was seen to C. albicans mannan. These results indicate that in AD patients, simultaneous IgE responses to various yeast polysaccharides occur, that the major sensitizer is C. albicans, and that the S. cerevisiae mannan is cross-reacting. Savolainen et al. studied the major protein profiles of extracts from C. albicans, C. utilis, R. rubra, and Cryptococcus albidus . The greatest similarity in the protein profiles was seen between C. albicans and C. utilis, which both showed a dominant protein staining of 46 kDa. In IgE immunoblotting, prominent IgE binding to the 46-kDa band was seen in one serum sample not only to C. albicans and C. utilis but also to S. cerevisiae and to a lesser extent to the corresponding weak bands of R. rubra and Cryptococcus albidus. The IgE immunoblotting also revealed the largest number of IgE binding bands in C. albicans followed by C. utilis, S. cerevisiae, R. rubra, and Cryptococcus albidus. When cross-reactivity was studied with immunoblotting inhibition, the C. albicans inhibition of C. utilis reached 80% inhibition. The enolase staining was almost as strongly inhibited with C. albicans as was the autoinhibition. S. cerevisiae enolase inhibited C. albicans staining to 55% and C. utilis staining to only 15%, while the same concentration gave an enolase autoinhibition of 98%. An important IgG and IgA response was seen with high-molecular-weight mannan-like components. (iv) Cellular and cytokine responses. : Several papers dealing with cellular and cytokine responses have already been mentioned in the Malassezia section. The delayed-type hypersensitivity response to C. albicans was found in several studies to be different from the result seen with Malassezia and other environmental allergens. Using the chamber-scarification patch test technique, 97 patients with AD were investigated for delayed-type hypersensitivity against extracts from C. albicans, Malassezia, and other environmental allergens . Healthy controls had a higher incidence of positive patch test results (86% positive) against C. albicans than did patients with AD (49%). For the other environmental allergens tested, including Malassezia (approximately 10% positive), the incidence of positive results in patients with AD was equal to or higher than the reaction seen in healthy controls. A negative correlation between the patch test result for C. albicans and both total IgE and specific RAST to C. albicans was found in patients with AD. The authors conclude that the persistent exposure to C. albicans allergen in patients with AD leads to the development of immediate hypersensitivity, with a subsequent decrease or total loss of delayed-type hypersensitivity to C. albicans allergen. This is in parallel with the results of experimental dermatophytosis experiments in patients with AD . The persistent carriage of C. albicans in the oral cavity and intestine may induce a Th2-type T-cell response that facilitates the IgE response on the one hand but reduces the delayed-type hypersensitivity on the other . In another study by the same group, the incidence of positive patch test results in respons to a C. albicans allergen was again significantly lower in patients with AD than in healthy controls . The results of a lymphocyte stimulation test with a C. albicans allergen, using PBMC from AD patients, gave a significant lower stimulation index than for healthy controls. The results with Staphylococus enterotoxin A and B were similar for AD patients and healthy controls. Matsumura et al. performed patch test (scarification) experiments using various environmental allergens in 46 patients with AD (without any respiratory allergy), in 41 patients with allergic respiratory disease but without AD, and in 42 healthy controls . AD patients showed a significantly lower incidence of positive responses to C. albicans (34%) than did both patients with allergic respiratory disease (90%) and healthy controls (84%). When the patch test results were compared to specific RAST results the percentage of those with positive RAST results and negative patch test results was higher in the AD group, while the percentage of those with negative RAST results and positive patch test results was strikingly higher in the group with allergic respiratory disease. In a study by Kawamura et al., PBMC from both patients with AD and healthy controls were incubated with a crude extract of C. albicans . PBMC from both AD patients and healthy controls showed a significant increase in proliferation. However, PBMC from healthy controls showed a significantly higher stimulation index than did PBMC from AD patients. When PBMC stimulated with C. albicans were costimulated with monoclonal antibodies to CD40-, CD54-, CD80-, or CD86-positive cells, the proliferation was suppressed in a significantly larger number of subjects with anti-CD54 than with any of the other antibodies. This effect of anti-CD54 was seen in both AD patients and healthy controls. Kimura et al. have studied the production of IFN-gamma and IL-5 by PBMC after stimulation with an extract of C. albicans in patients with AD, patients with bronchial asthma and healthy controls . The amount of IFN-gamma produced was significantly smaller in AD patients than in both patients with bronchial asthma and healthy controls. The production of IL-5 was again significantly lower in patients with AD than in patients with bronchial asthma but comparable to the production in healthy controls. The production of both INF-gamma and IL-5 was significantly lower in patients with severe AD than in to patients with mild to moderate AD. Therefore, decreased IFN-gamma and IL-5 production seems not to be an accessory phenomenon but appears to be related to an essential immunological abnormality which results in AD. The authors conclude that it remains to be elucidated whether the decreased cytokine production by PBMC, after stimulation with C. albicans, from patients with AD is due to decreased cytokine production per cell or to a decrease in the number of cytokine-producing cells in PBMC. These results obtained by Kawamura et al. are in conflict with results obtained by Savolainen et al. . They found that both a mannan and a protein extract from C. albicans induced an higher proliferative response in patients with AD than in healthy controls. In general, mannan was a clearly stronger inducer of lymphoproliferation and cytokines. Mannan, but not protein, induced higher IL-2 response in AD patients than in healthy controls. Both mannan and protein induced the same level of IL-4 response in AD patients and healthy controls. Mannan induced a higher IFN-gamma production in AD patients than in healthy controls. In AD patients there was a correlation for both protein and mannan C. albicans extract between IL-2, IL-4, and IFN-gamma production, total serum IgE level, and specific C. albicans IgE level. However, Kimura et al. found again, in a study performed in 2000, that the production of IFN-gamma in patients with AD was significantly lower than that in both patients with bronchial asthma and healthy controls . They found no significant difference in the production of IL-4 among these three groups. In contrast to Savolainen et al. they found a significant negative correlation between specific serum C. albicans IgE and IFN-gamma production by C. albicans-stimulated PBMC. In a recently published study by Savolainen et al., the proliferative response by PBMC, after stimulation with both C. albicans and Malassezia extracts, was significantly higher in AD patients than in healthy controls . For C. albicans, IL-5 production and IFN-gamma production was significantly higher in AD patients than in healthy controls. C. albicans-induced IFN-gamma production was significantly higher than for Malassezia. Malassezia-induced IL-4 production was significantly higher than for C. albicans. Total IgE and PHA-induced IL-2 and IL-4 levels in serum were elevated. A network of correlations was seen between the total IgE level in serum and the yeast-specific IgE level, Malassezia-specific proliferation, PHA-induced IL-2 and IL-4 levels, and the C. albicans-induced IL-5 level, suggesting that these are associated with AD. According to the authors, hypersensitivity to yeasts is an important pathogenic factor in some cases of AD. Both C. albicans and Malassezia can induce IgE, lymphoproliferation, IL-2, IL-4, IL-5, and IFN-gamma responses. However, according to the authors, Malassezia is a stronger inducer of Th2-type response and C. albicans is a stronger inducer of Th1-type responses. There are several explanations of the differences obtained in different studies. Kimura et al. studied children, and Savolainen et al. studied mainly adults. Differences in the preparation of extracts, the length of the culture period, and the methods used for the experiments are also important factors. The lymphocyte stimulation responses and cytokine responses found in the literature can be summarized as follows: Kawamura et al. found increased lymphocyte stimulation but less than in healthy controls; Kimural et al. in 1999 found decreased production of IL-5 and IFN-gamma and in 2000 found decreased production of IFN-gamma; Sarolainer et al. in 1999 found increased production of IL-2 and IFN-gamma and in 2001 found increased production of IL-5, IL-2, and IFN-gamma. Treatment. : Several studies have already been mentioned in the section dealing with Malassezia yeasts and treatment of AD . In the above-mentioned study by Morita et al., five patients with AD were treated orally once daily with 50 mg of ketoconazole and five patients with AD were treated orally once daily with 50 mg of fluconazole for 3 months . In both treatment groups, four patients were evaluated as improved after 3 months of treatment, and C. albicans-specific IgE levels in serum were significantly decreased in both groups. In a study by Adachi et al. , the effect of oral antifungal treatment was evaluated in 140 patients with treatment refractory AD and with a positive specific RAST result to C. albicans. According to the authors, good or excellent results were obtained in 60% of patients receiving fluconazole, 35% receiving itraconazole, 31% receiving oral amphotericin B, and 28% receiving nystatin. The number in each group, the dosage, and the length of treatment are not mentioned in the English abstract of this Japanese paper. That oral amphotericin B and nystatin both work indicate that C. albicans present in the gastrointestinal tract may be important as an allergen in AD. Although some studies may indicate that antifungal treatment of both Malassezia and Candida may be effective in the treatment of some patients with AD, we need further controlled studies to really prove this. Table summarizes the results of treatment studies mentioned in the literature. TABLE 4 | Treatment of AD patients with antifungal drugs Atopic Dermatitis, Dermatophytes, and Other Hyphomycetes : There are conflicting data in the literature about the association of AD with chronic dermatophyte infections . However, if a patient with AD develops a chronic dermatophyte infection, it is often more severe and more difficult to eradicate. The first extensive report about a link between AD and chronic dermatophyte infections was presented by Jones et al. . They found that 14 (40%) of 25 patients with chronic dermatophyte infections were atopic (AD, asthma, hay fever, or all three). A comparison of the frequency of chronic dermatophyte infection showed that 41% (14 of 34) in the atopic group had chronic dermatophyte infections compared to only 14% (21 of 146) in the nonatopic group. They also found that there was no correlation between chronic dermatophyte infection and any clinical condition except atopy. Of the 14 AD patients with chronic dermatophyte infections, 11 (79%) manifested immediate sensitivity to Trichophyton. However, the reason for the failure of immunity in persistent dermatophyte infections and its relationship to chronicity are still not well understood. There is an association between the presence of atopy and chronic dermatophytosis. A high proportion of those with persistent disease had atopic diseases (most commonly asthma or hay fever) as well as immediate-type hypersensitivity and raised total IgE levels . It has been suggested that modulation of T-lymphocyte activity either locally (in the skin) or systemically may be responsible. Possible mechanisms include activation of a Th2 pathway, which might explain the spectrum of antibody responses. It has also been found that dermatophyte antigens, including those that contain mannose residues, can reversibly suppress lymphocyte proliferation but not the expression of human leukocyte antigen (HLA-DK) . Patients with persistent infection have detectable levels of circulating antigen. Both are possible factors in the regulation of immunity in dermatophytosis. In a study by Rajka and Barlinn, patients with AD and chronic dermatophyte infections showed not only positive immediate reactions to Trichophyton but also to Penicillium and Cladosporium . Even patients with AD who showed no sign of dermatophyte infections had this cross-reactivity. However, this was absent in nonatopic patients with chronic dermatophyte infections, and healthy controls were negative to all the three extracts. These findings indicate that immediate reactivity to Trichophyton in AD patients is not necessarily a sign of contact with dermatophytes, but a cross-response based on mold reactivity. Patients with dermatophytosis are usually otherwise healthy. However, chronic and extensive infections have been reported to be more common in patients with immunodeficiency . In addition, there is an increased incidence of atopy in those with chronic infections, suggesting that host factors may well determine the clinical course. Dermatophytes may also have an impact on the severity of atopy, especially asthma . Platts-Mills et al. reported on a patient with severe asthma, tinea pedis, and tinea cruris who showed positive results of immediate skin test and RAST to Trichophyton . When the patient was treated with an effective antifungal drug, his asthma improved significantly. Ward et al. have reported 12 patients with allergic respiratory diseases with positive immediate reactions to Trichophyton, whose disease became worse when they were challenged with Trichophyton antigen . Eight of these patients improved when treated with specific systemic antifungal therapy. Even AD patients with dermatophyte infections may show improvement of their AD when treated with specific antifungal therapy . Wilson et al. have described a patient with tinea unguium, recalcitrant AD, and specific IgE Trichophyton serum antibody who showed improvement of her AD when treated with oral antifungal treatment. Klein et al. have also described a patient with AD and tinea pedis and tinea unguium culture positive for T. rubrum . The patient showed improvement of her AD when treated with oral ketoconazole for her tinea infection. Scalabrin et al. have studied the presence of total IgE and specific IgE to various fungi in AD patients . The majority of AD patients had both elevated total IgE and specific IgE antibodies directed against Aspergillus fumigatus and A. alternata (fungi that are characteristically inhaled) as well as against C. albicans, M. furfur, and T. rubrum (fungi that colonize the skin or mucous membranes). There were extensive correlations among specific IgE for each of the fungi. The authors conclude that fungi may be important allergens in several patients with AD and that antifungal therapy may improve AD in many patients. Lindgren et al. found that 18% of children with AD had IgE antibodies to Trichophyton . These patients also had specific antibodies to other fungi, house dust mites, and other environmental allergens. Seven children with positive RAST results to both T. rubrum, Malassezia, and C. albicans had severe AD. CONCLUSION : Do yeasts and other fungi play any role in AD? As shown in this review, there are conflicting data in the literature. It is important to remember that fungi are not the causative agents of AD. However, in some patients especially in adult patients who do not respond to traditional treatment, Malassezia and Candida may play a role. Although some studies indicate that antifungal treatment of both Malassezia and Candida may be effective in the treatment of some patients with AD, we need new, well-conducted, double-blind controlled studies to really prove this. It is also important to remember that a possible allergic or nonimmunogenic role of fungi is just an aggravating factor in AD and that traditional treatment with emollients and anti-inflammatory agents is still necessary. In the literature, conflicting results are presented for both immediate hypersensitivity (SPT and IgE), delayed hypersensitivity (APT), lymphocyte stimulation assays, and IL assays. There are several explanations for this: adult patients compared to children, the severity of AD, the presence of other atopic manifestations, differences in the preparation of extracts, the length of the culture period used, and the method of the assay. Backmatter: PMID- 12364368 TI - Acute Septic Arthritis AB - Acute septic arthritis may develop as a result of hematogenous seeding, direct introduction, or extension from a contiguous focus of infection. The pathogenesis of acute septic arthritis is multifactorial and depends on the interaction of the host immune response and the adherence factors, toxins, and immunoavoidance strategies of the invading pathogen. Neisseria gonorrhoeae and Staphylococcus aureus are used in discussing the host-pathogen interaction in the pathogenesis of acute septic arthritis. While diagnosis rests on isolation of the bacterial species from synovial fluid samples, patient history, clinical presentation, laboratory findings, and imaging studies are also important. Acute nongonococcal septic arthritis is a medical emergency that can lead to significant morbidity and mortality. Therefore, prompt recognition, rapid and aggressive antimicrobial therapy, and surgical treatment are critical to ensuring a good prognosis. Even with prompt diagnosis and treatment, high mortality and morbidity rates still occur. In contrast, gonococcal arthritis is often successfully treated with antimicrobial therapy alone and demonstrates a very low rate of complications and an excellent prognosis for full return of normal joint function. In the case of prosthetic joint infections, the hardware must be eventually removed by a two-stage revision in order to cure the infection. Keywords: SOURCE OF INFECTION : Most septic joints develop as a result of hematogenous seeding of the vascular synovial membrane due to a bacteremic episode . Although a rare cause, acute septic arthritis may also occur as a result of joint aspiration or local corticosteroid joint injection . In addition, bacterial arthritis may arise secondary to penetrating trauma (such as human or animal bite or nail puncture) or after trauma to a joint without an obvious break in the skin. The direct introduction of bacteria during joint surgery has increasingly been a source of bacterial arthritis, particularly in association with knee and hip arthroplasties. When a bone infection breaks through the outer cortex and into the intracapsular region, a joint infection may also result, especially in children . In infants, small capillaries cross the epiphyseal growth plate and permit extension of infection into the epiphysis and joint space . In children older than 1 year, osteomyelitis infection presumably starts in the metaphyseal sinusoidal veins and is usually contained by the growth plate. The joint is spared unless the metaphysis is intracapsular. The infection spreads laterally, where it breaks through the cortex and lifts the loose periosteum to form a subperiosteal abscess. In adults, the growth plate has resorbed and the infection may again extend to the joint spaces. MICROBIOLOGY : Virtually every bacterial organism has been reported to cause septic arthritis. The microorganisms responsible for bacterial arthritis are largely dependent on host factors (see "Risk Factors" below). The most common etiological agent of all septic arthritis cases in Europe and all nongonococcal cases in the United States is Staphylococcus aureus . The representation of S. aureus is more pronounced in patients with either rheumatoid arthritis or diabetes. After S. aureus, Streptococcus spp. are the next most commonly isolated bacteria from adult patients with septic arthritis . While one study had a high representation of Streptococcus pneumoniae , Streptococcus pyogenes is usually the most common streptococcal isolate, often associated with autoimmune diseases, chronic skin infections, and trauma . Groups B, G, C, and F, in order of decreasing preponderance, are also isolated, especially in patients with immunocompromise, diabetes mellitus, malignancy, and severe genitourinary or gastrointestinal infections . Gram-negative bacilli account for approximately 10 to 20% of cases . Patients with a history of intravenous drug abuse, extremes of age, or immunocompromise display a higher prevalence of infection by gram-negative organisms. The most common gram-negative organisms are Pseudomonas aeruginosa and Escherichia coli. Anaerobes are also isolated in a small percentage of cases, usually in diabetic patients and patients with prosthetic joints. Approximately 10% of patients with nongonococcal septic arthritis have polymicrobial infections. Historically, Haemophilus influenzae, S. aureus, and group A streptococci were the most common causes of infectious arthritis in children younger than 2 years. However, the overall incidence of H. influenzae as a cause of septic arthritis is decreasing because of the H. influenzae type b (Hib) vaccine now given to children . A recent study of 165 cases of acute hematogenous osteomyelitis or septic arthritis treated in the years before and after the advent of the Hib vaccine demonstrated that musculoskeletal infections due to this bacterial species were reduced to nearly nonexistent levels . Therefore, the coverage of H. influenzae as part of the empiric antibiotic coverage may no longer be needed in the management of acute septic arthritis in Hib-vaccinated children. While H. influenzae has lost its predominance as the most commonly identified gram-negative pathogen in pediatric populations, the normal oropharyngeal resident of young children, Kingella kingae, may have taken its place, specifically in patients younger than 2 year . In fact, a recent study found that the nearly half of the clinical isolates from patients younger than 2 years with acute septic arthritis were K. kingae . However, these results have yet to be repeated in other regions. Clinical data suggest that the organism may gain access to the bloodstream in the course of an upper respiratory infection or stomatitis . In children older than 2 years, S. aureus, streptococci, H. influenzae, and N. gonorrhoea have usually been isolated , although H. influenzae may have also lost its predominance in patients in this age group . Microbiological associations exist with concomitant disease states. Septic arthritis following cases of infectious diarrhea may be caused by Shigella spp., Salmonella spp., Campylobacter spp., or Yersinia spp. . However, these cases may reflect a form of reactive arthritis. A rare form of migrating polyarthritis may be caused by Streptobacillus moniliformis. In human immunodeficiency virus (HIV)-infected patients, S. aureus continues to be the most common isolate (approximately 30%) . However, increased numbers of opportunistic pathogens are isolated from this patient subset, including S. pneumoniae, mycobacterial species, and fungal species . While relatively rare in Western Europe, the diplococcus gram-negative bacterial species Neisseria gonorrhoeae is the most common cause of septic arthritis in United States . The number of cases of gonorrhea decreased by 72% between 1975 and 1997, and this decrease was correlated with a reduction in disseminated gonococcal infection and arthritis . However, the reported rate has increased by 9.2% between 1997 and 1999 and now stands at 133.2 cases per 100,000 per year . Specifically, the rate of gonococcal infection in men who have sex with men has demonstrated an alarming increase. These increased incidence rates may also cause larger numbers of observed gonococcal arthritis cases. PATHOGENESIS : The pathogenesis of acute septic arthritis is multifactorial and depends on the interaction of the host immune response and the invading pathogen. By taking into account the steps of bacterial colonization, infection and induction of the host inflammatory response, one may gain a greater understanding of this joint disease. Nongonococcal Arthritis : Since S. aureus has been extensively studied with regard to its role in septic arthritis and causes the majority of cases in most nations (and the majority of nongonococcal cases in the United States), we will use this bacterial species as the "typical" pathogen in the discussion of acute nongonococcal septic arthritis. Joint colonization and bacterial adherence. : The synovial membrane has no limiting basement plate under the well-vascularized synovium; this allows easy hematogenous entry of bacteria. As mentioned above, bacteria may also gain entry into the joint by direct introduction or extension from a contiguous site of infection. Once bacteria are seeded within the closed joint space, the low fluid shear conditions enable bacterial adherence and infection. Colonization may also be aided in cases where the joint has undergone recent injury. In this environment, the production of host-derived extracellular matrix proteins that aid in joint healing (e.g., fibronectin) may promote bacterial attachment and progression to infection. The virulence and tropism of the microorganisms, combined with the resistance or susceptibility of the synovium to microbial invasion, are major determinants of joint infection. S. aureus, Streptococcus spp., and N. gonorrhoeae are examples of bacteria that have a high degree of selectivity for the synovium, probably related to adherence characteristics and toxin production. Aerobic gram-negative bacilli such as Escherichia coli rarely infect the synovium except in the presence of an underlying and compromising condition. The virulence of the organism once inside the joint varies. In rabbits, intra-articular injection of 10 5S. aureus organisms into the knee joint resulted in major joint destruction but identical injections of N. gonorrhoeae or S. epidermidis caused no joint inflammation . S. aureus has a variety of receptors, termed microbial surface components recognizing adhesive matrix molecules (MSCRAMMs), for host proteins that mediate adherence to the joint extracellular matrix or implanted medical devices . Some of the host matrix proteins include fibronectin and laminin (adherence proteins), elastin (imparts elastic properties), collagen (structural support), and hyaluronic acid (a glycosaminoglycan that is rich in the joints and the matrix and provides cushioning through hydration of its polysaccharides). A number of adhesin genes have been determined and include genes encoding fibrinogen binding proteins (fib, cflA, and fbpA) , fibronectin binding proteins (fnbA and fnbB) , a collagen receptor (cna) , an elastin binding protein (ebpS) , and a broad-specificity adhesin (map) that mediates low-level binding of several proteins including osteopontin, collagen, bone sialoprotein, vitronectin, fibronectin, and fibrinogen . Also, S. aureus possesses a number of other host protein binding receptors whose genes have not yet been determined. These include a laminin binding protein (52 kDa) , a lactoferrin binding protein (450 kDa) , and a transferrin binding protein (42 kDa) . The staphylococcal receptor that binds laminin may be used in extravasation . These receptors specific to S. aureus were absent from the noninvasive pathogen S. epidermidis . The lactoferrin and transferrin receptors bind to host iron acquisition proteins and may be used as adhesins and/or as iron acquisition mechanisms. Increasing evidence supports the importance of staphylococcal surface components as virulence determinants by enabling initial colonization. In a number of studies, mutations in these receptors strongly reduced the ability of staphylococci to produce infection. In a murine septic arthritis model, inoculation of mice with mutants containing mutations of the collagen adhesin gene showed that septic arthritis occurred 43% less often than in the corresponding wild type . Also, vaccination with a recombinant fragment of the S. aureus collagen adhesin was able to reduce the sepsis-induced mortality rate to 13%, compared with 87% in the control group . However, the role of collagen adhesion of S. aureus as a major virulence factor has recently been questioned since approximately 30 to 60% of clinical isolates do not display collagen binding in vitro or the cna-encoded collagen adhesin . Staphyloccal fibronectin binding proteins (FbpA and FbpB) may play a major role in the colonization and virulence of septic arthritis. In a recent study, all of the tested clinical isolates (n = 163) contained one or both of the coding regions for these binding proteins and 95% of these strains had a comparable fibronectin binding capacity to that seen in a staphylococcal reference strain known to efficiently bind fibronectin . In addition, an in vivo study of endocarditis in a rat model showed that mutants deficient for fibronectin binding protein were 250-fold less adherent to traumatized heart valves . Also, S. aureus adherence to miniplates from iliac bones of guinea pigs was three times higher for the wild-type strain than for the adhesin-defective mutant strain . It is likely fibronectin binding proteins play an important role in joint infections, especially those associated with joint trauma or implanted medical devices . These receptors may play an additional role in an intracellular immunoavoidance strategy. S. aureus survives intracellularly after internalization by cultured osteoblasts . Staphylococci have demonstrated internalization into other cultured mammalian cells as well as osteoblasts; these include bovine mammary gland epithelial cells, human umbilical vein endothelial cells, and pulmonary epithelial cells isolated from a cystic fibrosis patient . Initial adherence to glandular epithelial cells is mediated by S. aureus fibronectin receptors , possibly using fibronectin as a bridge between the host cell and the bacterial receptors. Following adherence, bacteria may be internalized by host mechanisms involving membrane pseudopod formation (seen in established bovine mammary epithelial cell lines) or through receptor-mediated endocytosis via clathrin-coated pits (seen in mouse osteoblasts and epithelial cells) . In either case, the dependence on the action of host cytoskeletal rearrangements through microfilaments is evident. Following internalization, staphylococci may induce apoptosis (via a host caspase-dependent mechanism) or survive intracellularly . Induced apoptosis may exacerbate the host cell damage seen in septic arthritis. Also, staphylococci may escape clearance by the immune system and antimicrobial therapy by persisting within these host cells. This survival was recently demonstrated in vivo when S. aureus cells were found in the cytoplasm of embryonic chicken osteoblasts and osteocytes in mineralized bone matrix . In another study, S. aureus was found within polymorphonuclear neutrophils in an in vivo infection model, and these infected host cells were able to establish infection in naive animals . Therefore, this pathogen may utilize invasion as an immunoavoidance technique during the host inflammatory response. After the downregulation of the adaptive immune response through T-cell apoptosis (mediated by superantigens, other toxins, and invasion), fulminant and/or persistent infection may result. Joint infection and the host immune response. : Once colonized, bacteria are able to rapidly proliferate and activate an acute inflammatory response. Initially, host inflammatory cytokines, including interleukin 1-beta (IL-1beta) and interleukin 6 (IL-6), are released into the joint fluid by synovial cells . These cytokines activate the release of acute-phase proteins (e.g., C-reactive protein) from the liver that bind to the bacterial cells and thereby promote opsonization and activation of the complement system. In addition, there is an accompanying influx of host inflammatory cells into the synovial membrane early in the infection. Phagocytosis of the bacteria by macrophages, synoviocytes, and polymorphonuclear cells occurs and is associated with the release of other inflammatory cytokines that include tumor necrosis factor alpha (TNF-alpha), IL-8, and granulocyte-macrophage colony-stimulating factor, in addition to increasing the levels of IL-1beta and IL-6, which are already present. It was demonstrated in a recent clinical study that IL-6 and TNF-alpha concentrations were persistently high even 7 days after treatment was initiated while IL-1beta levels decreased significantly after 7 days . Many of these cytokines and the associated immune response have been shown in animal models to be required for bacterial clearance and the prevention of mortality due to bacteremia and septic shock . Nitric oxide, a common mediator of inflammatory cytokines, is also required . The T-cell mediated (Th1) and humoral (Th2) adaptive immune responses may also play a role in the clearance and/or pathogenesis of acute septic arthritis. T cells enter the joint within a few days following infection . The role of CD4 + T cells in joint destruction has been demonstrated by showing that their in vivo depletion resulted in a considerably milder course of staphylococcal arthritis . These lymphocytes are specifically activated by bacterial antigens in association with host antigen-presenting cells or nonspecifically in the case of bacterial superantigens (e.g., toxic shock syndrome toxin 1 [TSST-1]). The cytokine, gamma interferon (IFN-gamma), produced by these activated T cells reduced the level of mortality and joint destruction in a mouse model of group B Streptococcus when delivered 18 h after bacterial inoculation . However, when S. aureus was used as the infecting organism in this model, IFN-gamma increased the frequency and severity of septic arthritis while simultaneously protecting mice from septicemia . Also, it has been found in a recent study in mice that a high level of IFN-gamma (a Th1 cytokine) plays a detrimental role in staphylococcal infection and that IL-4 and IL-10, both being Th2 cytokines, are involved in host resistance to infection through regulation of IFN-gamma . However, the necessity of the Th2 response to clear S. aureus infection has lately been questioned in a study utilizing IL-4-deficient mice . It seems that a Th2 response is required for S. aureus infection clearance only in certain mice, depending on their genetic background. Therefore, the exact role of T cells in host tissue damage and infection clearance is still being elucidated. Joint damage. : Under most circumstances, the host is able to mount a protective inflammatory response that contains the invading pathogen and resolves the infection. However, when the infection is not quickly cleared by the host, the potent activation of the immune response with the associated high levels of cytokines and reactive oxygen species leads to joint destruction. High cytokine concentrations increase the release of host matrix metalloproteinases (including stromelysin and gelatinase A/B) and other collagen-degrading enzymes. When monoclonal antibodies or steroids attenuate these cytokines, cartilage degradation is minimized. The joint is further damaged by the release of lysosomal enzymes and bacterial toxins . Host proteoglycans are initially degraded, and this is followed by collagen degradation. In fact, the polymorphonuclear response with subsequent release of these proteolytic enzymes can lead to permanent destruction of intra-articular cartilage and subchondral bone loss in as little as 3 days. Metalloproteinases and the antigen-induced inflammatory response may persist and continue to damage the joint architecture even after the infection has been cleared . The infectious process induces a joint effusion that increases intra-articular pressure, mechanically impeding blood and nutrient supply to the joint. Thus, increased pressure destroys the synovium and cartilage. Because of the proximity of the epiphyseal growth plate to the joint, direct extension of a joint infection to any of the articulating bones may lead to decreased bone growth in infants and children . While bone mineralization is preserved, cartilage destruction causes joint space narrowing and erosive damage to the cartilage and bone if left untreated . In addition, the infection can spread to surrounding soft tissue, form sinus tracts, and disrupt ligaments and tendons in the untreated state . Bacterial products and their pathogenic role. : While bacterial attachment proteins promote colonization and initiate the infectious process, a number of bacterial products activate the host immune response and increase tissue damage in cases of septic arthritis. S. aureus has a large variety of factors that have been implicated in host virulence. Many of these factors have been tested for their ability to increase the morbidity and mortality associated with acute septic arthritis. Most studies evaluating the potential role of these bacterial products have been performed using the murine model of septic arthritis . During acute septic arthritis, the innate immune system responds to the presence of the peptidoglycan wall (via N-formylmethionine proteins and teichoic acids) of S. aureus to produce proinflammatory cytokines (such as IL-1beta, IL-6, and TNF-alpha) and C-reactive protein. Bacterial DNA (specifically unmethylated CpG motifs) also elicits an intense inflammatory response . When bacterial DNA from S. aureus or E. coli or synthetic, unmethylated oligonucleotides containing CpG motifs were injected into the knee joint of mice, arthritis developed quickly and lasted up to 14 days, while methylated DNA had no significant effect. Also, the affected tissue was characterized by monocyte and macrophage influx with the release of their associated cytokines and chemokines and the absence of T cells. It has been noted that bacterial superantigens such as staphylococcal TSST-1 and enterotoxins may play a major role in the potent activation of the host inflammatory response, thereby increasing the mortality rates and exacerbating host inflammatory-cell invasion, cytokine release, and joint degradation . Most animals infected with strains of S. aureus isogenic for TSST-1 or enterotoxins (A through D) developed frequent and severe arthritis . However, most animals (80%) infected with strains devoid of these toxins had no symptoms and the animals with symptoms had only mild or transient arthritis infections . Also, vaccination with a recombinant form of staphylococcal enterotoxin A devoid of superantigenicity was able to generate significant protection from S. aureus sepsis in mice . Superantigens act by binding to the conserved lateral regions of the host major histocompatibility complex class II molecule and T-cell receptor. While only approximately 1 in 10 4 T cells are activated during normal presentation of a nonself antigen, a superantigen may activate 2 to 20% of all T cells . These activated T cells are then able to increase the release of a number of cytokines, such as IL-2 , IFN-gamma, and TNF-alpha . This upregulated production of cytokines causes a significant systemic toxicity and suppression of the adaptive immune responses and inhibits plasma cell differentiation. Also, the stimulated T cells proliferate and then rapidly disappear, apparently due to apoptosis . Therefore, immune suppression may be due to generalized immunosuppression and T-cell deletion. Human B cells are also stimulated by these staphylococcal superantigens. Besides the role that superantigens play in the mortality and morbidity associated with septic arthritis, other staphylococcal toxins may also contribute to the disease process. One study was able to demonstrate that alpha-hemolysin was a significant mediator of virulence in arthritis . Alpha-hemolysin is secreted as a monomer that attaches to host membranes and polymerizes into a hexameric ring channel . While this hemolysin binds to human erythrocytes in a nonspecific manner, it can still mediate significant host cell lysis when produced in high concentrations in the infection environment . Also, alpha-hemolysin promotes significant blood coagulation by neutrophil adhesion , platelet aggregation (via a fibrinogen-dependent mechanism) , and its nonlytic attack on human platelets . In addition, this hemolysin can form channels in nucleated cells (e.g., endothelial cells) through which calcium ions freely pass . The calcium influx is responsible for the vasoregulatory process and inflammatory- response disturbances seen in severe infection . Lastly, alpha-hemolysin interferes with lymphocyte DNA replication . These multiple effects of alpha-hemolysin on the host contribute to the vascular disturbances and immunodeficiency seen in staphylococcal infections. The pathogenic properties of alpha-hemolysin were recently found to only occur when another staphylococcal toxin, the leukocyte-specific gamma-toxin, was also present in the infecting strain . Gamma-hemolysin (HlgAB and HlgCB) and a related S. aureus leukocidin (LukSF-PV) specifically lyse leukocytes. Each of these toxins is composed of an interchangeable two-component system. The active toxin is formed by taking one protein from the S-component family (LukS-PV, HlgA, and HlgC) and one from the F-component family (LukF-PV and HlgB) . The S component is most probably responsible for the specific cytopathic effect of each of the toxins, while the F component is responsible for the common leukocyte binding activity. While LukF and HlgA proteins show very strong similarity, they are encoded on different gene loci . Since these cytotoxins specifically interact and lyse leukocytes, they contribute to the inhibition of infection clearance by the host immune system, thereby enabling staphylococcal species to persist. Therefore, it is the combined effects of the hemolysin and leukotoxins that increase the ability of S. aureus to cause acute septic arthritis. These factors enable the host to mount a protective inflammatory response that contains this pathogen and often resolves the infection. However, when the infection is not cleared by the host innate immune system, S. aureus is well equipped to persist by possessing a number of virulence factors and strategies, including but not limited to invading and surviving in mammalian cells, hiding within a biofilm, or producing a thick, antiphagocytic capsule. The difficulty in treating septic arthritis and the ability of the bacteria to evade clearance by the host immune response reside in a number of staphylococcal defense mechanisms. Such characteristics are expressed at both the cellular and matrical levels. As mentioned above, protein A is bound covalently to the outer peptidoglycan layer of their cell walls. This receptor binds to the Fc portion of immunoglobulin G and presents the Fab fragment of the antibody to the external environment. Therefore, the Fc portion is unable to either bind complement or signal polymorphonuclear leukocytes, thereby interfering with staphylococcal opsonization and phagocytosis. This interference has been demonstrated in vitro and in animal models with subcutaneous abscesses and peritonitis. Also, protein A coats the staphylococcal cell in a coat of host Fab fragments, and the ability of the immune system to recognize the pathogen as nonself is hindered. The importance of protein A in S. aureus septic arthritis was demonstrated in a recent study in which strains that obtained this virulence factor caused greater inflammation and cartilage destruction . Capsular polysaccharide may interfere with opsonization and phagocytosis. Among the 11 reported serotypes, capsule types 5 and 8 (microcapsule producers) comprise the vast majority (75 to 94%) of clinical isolates . The capsule of these two serotypes is much smaller than the capsule of other serotypes of S. aureus (such as capsule type 1) or pathogenic species such as Streptococcus pneumoniae. Unencapsulated and microencapsulated strains demonstrated a high rate of serum clearance compared to fully encapsulated strains. Therefore, the role of capsular polysaccharide in opsonization and phagocytosis was questioned . However, the thin capsule may be necessary in early bone infection stages in order to allow the interaction of staphylococcal adhesion factors with host proteins (such as fibrin and fibronectin). In one study, it was shown that a small capsule was necessary for fibroblast attachment by protein A of S. aureus and that a fully encapsulated strain reduced the binding efficiency . In another study, the thin capsule was shown to be necessary for binding to bone collagen type 1, since high capsular expression actually inhibited binding . Once these microorganisms adhere to solid surfaces (such as bone), both in vitro and in vivo, staphylococci produce larger quantities of cell-associated capsule than do those grown in liquid cultures . Specifically, type 5 and type 8 capsule production is strongly upregulated during postexponential growth (i.e., after adhesion and colonization) . This upregulated capsule production makes them resistant to antimicrobial treatment and host immune clearance. Therefore, once staphylococcal adherence proteins establish the infection, the pathogen enters the postexponential growth phase and begins producing a thicker capsule that covers and hides the highly immunogenic adherence proteins. This thicker type 5 and type 8 capsule is serum resistant through inhibition of phagocytosis and opsonization . The effect of the staphylococcal polysaccharide microcapsule in murine arthritis was recently explored. In this study, strains expressing type 5 capsule were shown to cause a higher rate of mortality, a higher frequency of arthritis, and a more severe form of the disease compared to capsule mutants . In a clinical trial, a vaccine (Staph Vax) that consists of isolated type 5 and 8 capsular polysaccharides was able to significantly reduce infection rates (by 57%) in a high-risk population for as long as 10 months . As mentioned above, S. aureus also survives intracellularly after internalization by cultured osteoblasts . Type 5 capsule production by in vivo-grown S. aureus (i.e., internalized in cultured osteoblasts) was recently shown to be upregulated compared to that by S. aureus grown in vitro . Therefore, the capsule may not only resist phagocytosis and opsonization but may also contribute to intracellular survival. In summary, S. aureus infects and elicits a strong native immune response, cytokine release, and high T-cell activation. This pathogen is able to use a number of immunoavoidance strategies during this time while the host immune system simultaneously causes damage to "self" tissues and blood vessels in the area of infection. This damage may cause local circulatory and immune system compromise. The high T-cell activation eventually results in apoptosis and a weakened immune system, enabling the pathogen to effectively produce a sustained and destructive infection. While the bacterial products discussed above have been shown to increase joint damage in acute septic arthritis, many more S. aureus virulence factors have not yet been tested. Therefore, we would expect that number of factors implicated as playing a role in septic arthritis would undoubtedly increase, and their relative roles will be more clearly elucidated in future studies. Bacterial clearance versus joint damage. : The interaction of the bacteria and host is of utmost importance in the initiation and prolongation of infection and cartilage damage. There is a subtle balance between an effective immune response to eliminate the infecting organism from the host and the overactivation of this response that causes the majority of infection-related joint destruction. Therefore, care must be exercised and further studies must be performed in regard to using agents that suppress the inflammatory response in the treatment of septic arthritis. Gonococcal Arthritis : Gonococcal arthritis occurs in approximately 42 to 85% of patients with disseminated gonococcal infection (DGI) and begins with a localized mucosal infection . DGI-producing strains are unusually sensitive to in vitro killing by penicillin G and possess unique nutritional requirements for arginine, hypoxanthine, and uracil. N. gonorrhoeae possesses a number of virulence factors. It is the combined effects of these factors, their phase and antigenic variation, and properties of the host immune response that enable this pathogen to persist and allow the localized infection to become DGI. Gonococcal virulence factors. : N. gonorrhoeae possesses a number of cell surface structures that have been implicated in virulence. Initial attachment to host epithelium is mediated by long, hair-like protein projections called pili Phase variation, i.e., the question of whether this membrane structure is assembled (Pil +) or not (Pil -), is determined by posttranslational proteolytic cleavage, variations in homologous recombination, and slipped-strand DNA replication resulting in frameshift mutations . In addition, the antigenic character of the pili is altered by homologous recombination between coding regions for the various pilin subunits. Protein I is the main protein on the outer membrane. It is a porin that is expressed in two different forms, a protein IA variant that is almost always associated with disseminated infection and a protein IB variant that is associated with strains causing localized infections. Strains that are able to cause a disseminated infection in hosts with a normal immune system display serum resistance . Protein IA enables stable serum resistance by binding the host factor H. This bacterially- bound host factor efficiently inactivates C3b (a central factor in both the classical and alternative complement cascades) into iC3b , thereby reducing the efficacy of the host complement system. This porin may also be responsible for the prevention of phagolysosomal fusion in polymorphonuclear leukocytes and a reduced oxidative burst, thereby enabling survival within these cells. Another extracellular gonococcal protein is protein II, which is also called Opa since colonies expressing protein II on their surface have a more opaque appearance. This protein is thought to cooperate in the more intimate attachment following initial pilus interaction. In addition, protein II is able to attach to the lipooligosaccharide (LOS) of other N. gonorrhoeae organisms, thereby enabling the cells to bind to one other and form microcolonies. These microcolonies may also aid in the initiation of mucosal surface attachment. Protein II is capable of avoiding clearance by the host immune system by phase and antigenic variation . Phase variation occurs through slipped-strand synthesis that produces a frameshift mutation and produces a prematurely terminated form of the protein. In addition, multiple variants of the protein II gene exist, and therefore the antigenic character of protein II can be changed by homologous recombination between these variants. While this protein is important for mucosal infections, most isolates from patients with DGI are missing protein II from their outer membrane and grow to form transparent colonies. Protein III is another porin that is prevalent on the bacterial surface. The antibodies directed against protein III are not bactericidal, and they sterically inhibit antibody binding to protein I and unsialylated LOS that would probably result in bactericidal action . Therefore, the generation of these blocking antibodies may prevent serum bactericidal action. LOS is like the lipopolysaccharide of other gram-negative bacteria except that its carbohydrate portion does not have the complex structure of the repeating O side chain. LOS has endotoxin activity and is largely responsible for the synovial damage produced in gonococcal arthritis . While stable serum resistance is due to protein IA, unstable resistance is mediated by the ability of some gonococcal strains to covalently attach activated forms of host sialic acid to the galactose residues on LOS . This covalent attachment coats the bacterial cell in host proteins and avoids complement activation. In addition, opsonization by complement components and the formation of the membrane attack complex of the complement system are inhibited. N. gonorrhoeae also produces an immunoglobulin A protease that may aid in colonization. However, the relevance of this potential virulence factor in gonococcal pathogenesis needs further study. Host factors. : The host may contain a gonococcal infection through the action of the innate immune response, with particular dependence on the complement system. This system is largely responsible for attracting polymorphonuclear leukocytes and the resulting cascade of inflammatory cytokines and chemokines. However, during periods surrounding early pregnancy, puerperium, and menstruation, the accompanying alterations in vaginal pH, cervical mucus, and genital flora and the endometrial exposure of submucosal vessels may predispose the female patient to N. gonorrhoeae invasion and DGI . As mentioned above, defects in the complement and/or reticuloendothelial systems may also inhibit the host's ability to contain gonococcal infection. RISK FACTORS : Besides the obvious risk of septic arthritis associated with age older than 60 years and recent bacteremia, certain medical conditions predispose joints to nongonococcal infection. Degenerative joint disease, rheumatoid arthritis, and corticosteroid therapy are the most common predisposing conditions. Specifically, patients with rheumatoid arthritis have an approximately 10-fold-higher incidence of septic arthritis than does the general population . Patients with diabetes mellitus, leukemia, cirrhosis, granulomatous diseases, cancer, hypogammaglobulinemia, intravenous substance abuse, or renal disease and patients undergoing cytotoxic chemotherapy also have an increased incidence of septic arthritis . Total joint arthroplasties are susceptible to intraoperative or hematogenous seeding and subsequent prosthetic joint infections. While patients infected with HIV demonstrate a higher prevalence of musculoskeletal infections than does the general population (approximately 60 and 2 to 10 cases per 100,000 persons per year, respectively), it is unclear if this higher occurrence is due to the common septic arthritis risk factors due to intravenous drug abuse and multiple transfusions in this patient population . In 0.5 to 3% of gonorrhea infections, the pathogen is able to gain access to the bloodstream from the primary mucosal site of infection and produce DGI . A number of risk factors have been epidemiologically associated with the development of DGI; these include infection with transparent, piliated N. gonorrhoeae strains capable of phase variation; diagnosis delay (especially in females due to asymptomatic nature of the infection); complement system deficiency; systemic lupus erythematosus; menstration, pregnancy, and puerperium; male homosexuality; urban residence; promiscuity; and low socioeconomic and educational status. Females are four times as likely to develop DGI as males . This prevalence in women may be due to the asymptomatic nature of gonorrhea infections in women and the associated delay in diagnosis, thereby providing time for the bacteria to gain access to the bloodstream. In addition, many affected females are either pregnant or menstruating at the time of the infection . Also, since the clearance of gonococcal infection depends on an effective complement-mediated immunity and a functional reticuloendothelial system, complement deficiencies and systemic lupus erythematosus are risk factors in this patient subset. DIAGNOSIS OF NONGONOCOCCAL ARTHRITIS : Nongonococcal septic arthritis is a medical emergency that can lead to serious sequelae and mortality. Therefore, prompt recognition and treatment are critical to ensuring a good prognosis. Clinical Presentation : The classical presentation of acute nongonococcal septic arthritis includes recent onset of fever, malaise, and local findings of pain, warmth, swelling, and decreased range of motion in the involved joint . A significant number of patients have mild fever and may not demonstrate localized heat and erythema around the affected joint . The clinician should obtain a detailed history with special emphasis on determining the presence of any risk factors discussed above. However, the diagnosis of infectious arthritis rests on the isolation of the pathogen(s) from aspirated joint fluid. While any joint can become infected, the most commonly involved joints in nongonococcal septic arthritis are the knee and hip, followed by the shoulder and ankle . The hip may be more frequently involved in children. Also, infectious nongonococcal arthritis is monoarticular in 80 to 90% of cases . Atypical joint infection, including the sternoclavicular, costochondral, and sacroiliac joints, may be common in intravenous drug users. Also, penetrating trauma, including human or animal bites, and local corticosteroid therapy may cause septic arthritis in atypical joints. Polyarticular septic arthritis is usually accompanied by a number of risk factors (see above). Laboratory Findings : Peripheral blood leukocyte counts are usually elevated in children but are often within normal limits in adults. Most patients display elevated C-reactive protein levels and erythrocyte sedimentation rates. Synovial fluid analysis is also very important and usually reveals turbid, low-viscosity fluid with leukocyte counts usually in excess of 50,000/mm3. However, nonbacterial inflammatory processes, such as acute crystalline joint disease or reactive arthritis, may have counts above this level while gonococcal and granulomatous arthritis may have counts below 50,000/mm3. In nongonococcal arthritis, the fraction of polymorphonuclear leukocytes approaches 90% . Even though low joint fluid glucose levels (<40 mg/dl or less than half the serum glucose concentration) and high lactate levels are nonspecific, they are suspicious for bacterial arthritis. Normal joint glucose and lactate levels are usually found in patients with viral arthritis . Synovial fluid from any adult with monarticular arthritis should be examined by compensated polarizing light microscopy for negatively birefringent (uric acid) and positively birefringent (calcium pyrophosphate dihydrate) crystals in order to rule out crystalline joint disease. However, simultaneous bacterial infection and crystalline disease has been reported . Gram stains of synovial fluids may support the diagnosis of septic arthritis. In addition, it may differentiate between infections by gram-positive and gram-negative bacteria, thereby directing initial antimicrobial therapy before antibiotic sensitivity results are obtained. The synovial fluid should be sent for aerobic, anaerobic, mycobacterial, and fungal culture prior to the initiation of antimicrobial therapy. In addition, antibiotic sensitivities should be determined. Cultures are positive in nongonococcal arthritis approximately 90% of the time, while Gram stain is effective only 50% of the time . These cultures may be negative in patients in whom treatment has already been initiated. Once aspirated joint samples are obtained, it is imperative that they be quickly transported to clinical microbiology and not be allowed to stand for a long time without processing or culturing. In one study, it was found that by directly inoculating the aspirated sample into blood culture tubes, even very small numbers of viable bacteria in infected fluid could be detected . However, an increase in false-positive results due to general skin or other contaminants may also occur with this technique. If fluid cultures are sterile but the suspicion of septic arthritis persists, tissue samples of the synovial membrane may also be cultured for microbial isolation and identification. Sputum, urine, and blood cultures are also often required. Around half of all patients with nongonococcal arthritis show positive blood cultures . A summary of the diagnosis and management of acute nongonococcal arthritis can be seen in Table . TABLE 1 | Principles of diagnosis and management of acute nongonococcal septic arthritis Imaging Studies : Imaging studies of septic arthritis can be used only to support or dissuade a clinical suspicion of the disease; they should not be used as an absolute diagnostic indicator. Because the approaches and techniques are both numerous and diverse, there is confusion about which technique is most effective. Radiographic images are usually not revealing in the first few days of infection since they are usually normal or show only preexisting joint disease. However, swelling of capsule and soft tissue around the affected joint, fat pad displacement, and in some cases joint space widening due to localized edema and effusion may be seen. Also, the initial radiographic image may be used to determine associated conditions, such as osteoarthritis or simultaneous osteomyelitis, or may be used as a baseline image in monitoring the response to treatment. As the infection progresses, radiographic detection of diffuse joint space narrowing due to cartilage destruction is possible. Radiographs can also evaluate late, inadequately treated stages of septic arthritis in which generalized joint destruction, osteomyelitis, osteoarthritis, joint fusion, calcifications in the periarticular tissues, or subchondral bone loss followed by reactive sclerosis are seen. Ultrasonography is capable of showing both intra- and extra-articular abnormalities not apparent by plain radiography and is a very powerful tool to detect early fluid effusions and to guide initial joint aspiration and drainage procedures . Even small collections of fluid (1 to 2 ml) can be accurately detected . Non-echo-free effusions (due to clotted hemorrhagic collections) are very characteristic of a septic joint. It has been suggested that the presence of only an echo-free effusion (caused by transient synovitis and fresh hemorrhagic effusions) may rule out the diagnosis of septic arthritis . This imaging technique is also useful for detecting collections of fluids in deep joints, including the hip. In addition, the status of the intra-articular compartment, joint capsule, bony surface, and adjacent soft tissues and the patient's response to therapy can be monitored. Since ultrasonography is also noninvasive, inexpensive, easy to use, and devoid of irradiation or any other known complications, more clinicians should use it in the diagnosis of septic arthritis in the future. To diagnose ambiguous cases of septic arthritis or to determine the extent of bone and soft tissue infections, computed tomography (CT), magnetic resonance imaging (MRI), and radionuclide scans may be obtained. In most cases, these diagnostic tests are not required for septic arthritis. Like radiographs, CT scans have limited use during the early stages of septic arthritis. However, they may enable the visualization of joint effusion, soft tissue swelling, and para-articular abscesses. In addition, CT is more sensitive than plain radiography in the imaging of joint space widening due to localized edema, bone erosions, foci of osteitis, and scleroses. This scanning technique may be useful in the diagnoses of arthritis cases that are difficult to assess, including infections of the hip, sacroiliac, and sternoclavicular joints. In addition, it may assist in guiding joint aspiration, selecting the surgical approach, and monitoring therapy in these difficult infections . MRI has become a useful diagnostic tool for the early determination of musculoskeletal infection and its extent . As with CT, MRI may be particularly useful in aiding the diagnosis of joint infections that are difficult to access, such as sacroiliitis . MRI displays greater resolution for soft tissue abnormalities than CT or radiography and greater anatomical detail than radionuclide scans. The spatial resolution of MRI makes it useful in visualizing joint effusion and differentiating between bone and soft tissue infections. Furthermore, patients do not have to be exposed to ionizing radiation. The main disadvantages to MRI are high cost, lack of universal availability, imaging interference due to metal implants, and lower resolution of calcified bone structures and the cortex . Initial MRI screening usually consists of a T1-weighted and T2-weighted spin-echo pulse sequence. In a T1-weighted study, edema and fluid are dark while fat is bright. In a T2-weighted study, the reverse is true. Therefore, joint effusions, abscesses, and soft tissue edema generate a high signal on T2-weighted images. As with the other imaging techniques, MRI is nonspecific and is unable to differentiate between infectious and noninfectious inflammatory arthropathies . Radionuclide scans are often able to detect localized areas of inflammation. The 99mTc methyldiphosphonate scan demonstrates increases isotope accumulation in areas of osteoblast activity and increased vascularity . However, this radionuclide scan may be normal in the early stages of septic arthritis. A second class of radiopharmaceuticals used for the evaluation of septic arthritis includes 67Ga citrate and 111In chloride scans. 67Ga citrate and 111In chloride attach to serum proteins, including transferrin, lactoferrin, haptoglobin, and albumin, that leak from the bloodstream into areas of inflammation. Gallium and indium scans also show increased isotope uptake in areas of concentrated polymorphonuclear leukocytes, macrophages, and malignant tumors. While these scans are more specific and sensitive in the detection of active infection than 99mTc methyldiphosphonate , they do not show bone or joint detail well, and it is often difficult to distinguish between bone, joint, and soft tissue inflammation. Three-phase 99mTc methyldiphosphonate scans may help resolve this problem. In 111In-labeled leukocyte scanning, a sample of the patient's leukocytes is isolated, labeled with 111In, and injected back into the patient. These radiolabeled leukocytes localize into areas of acute infection according to host inflammatory cytokine and chemokine gradients. While this scan is positive in approximately 60% of patients with septic arthritis, false-positive results may occur in patients with synovitis secondary to active osteoarthritis . Differential Diagnosis : Preexisting joint infection. | A number of associated arthropathies should be considered in the differential diagnosis of acute septic arthritis. Patients with underlying chronic joint disease (including rheumatoid arthritis, osteoarthritis, and other connective tissue diseases) have a poor prognosis when suffering from acute septic arthritis . The poor prognosis associated with this patient population is mainly due to diagnostic delays since clinicians incorrectly ascribe symptoms to the preexisting arthropathy . Also, these patients are often afebrile and demonstrate an indolent presentation . Therefore, a diagnosis of septic arthritis must be entertained whenever a sudden onset of inflammatory arthritis in one or two joints occurs in these patients. Endocarditis. : Patients with infective endocarditis also demonstrate relatively high incidence rates (23 to 44%) of musculoskeletal abnormalities . Specifically, many of these patients have sterile myalgias and althragias, and the joint symptoms are usually polyarticular and symmetric, affecting both the large and small joints. In addition, septic arthritis is seen in approximately 5 to 15% of these patients, especially in intravenous drug abusers. Cases of endocarditis-associated septic arthropathies are usually mediated by Streptococcus spp. or S. aureus. Chronic infectious arthritis. : The diagnosis of mycobacterial or fungal arthritis should be entertained when a patient presents with chronic monarticular arthritis. Both of these arthritides have increased in prevalence, largely due to increased incidence rates seen in HIV-infected patients . Synovial fluid cultures for acid-fast bacteria and fungi should be considered for any patient who is immunocompromised or receiving immunosuppressive therapy or who has a persistent effusion. A culture of a synovial biopsy specimen should be done for fungi and acid-fast organisms in any person with a chronic monarticular involvement whose synovial fluid cultures are negative . In mycobacterial arthritis, macrophages may predominate in the synovial fluids. The delayed onset and insidious progression of this disease are markedly different from those of acute septic arthritis. Therefore, a thorough history of the illness in the patient, appropriate culture, and monitoring of clinical progression should provide reasonable clues to distinguish acute infectious joint arthropathies from cases of fungal or mycobacterial arthritis. Lyme disease may also present with chronic monoarticular arthritis. About 60% of untreated patients with Lyme disease in the United States have intermittent attacks of joint swelling and pain, especially in the knee, even months after the onset of illness . Demonstration of the typical and slowly expanding erythema migrans skin lesion at the site of the tick bite and the development of influenza-like symptoms (including malaise, fatigue, headache, fever, regional lymphadenopathy, and migratory polyarthragia) lasting weeks to months are highly indicative of Lyme disease . This is especially true in patients living in areas where this disease is endemic. While culture of Borellia burgdorferi (the spirochete responsible for Lyme disease) from specimens in Barbour-Stoenner-Kelly medium permits a definitive diagnosis, determination of an antibody response to B. burgdorferi by enzyme-linked immunosorbent assay is also possible . Mycoplasma spp. have also been repeatedly isolated in cases of chronic erosive septic arthritis, particularly in patients suffering from hypogammaglobulinemia . Viral arthritis. : Patients with viral arthritis usually present with polyarthritis, fever, lymphadenopathy and characteristic rash . Also, synovial fluid samples reveal an abundant presence of mononuclear leukocytes, and normal joint glucose and lactate levels are usually found . Clinical and epidemiological clues often lead the clinician to perform appropriate serological studies via antibody titers. Crystal-induced arthritis. : Gout and pseudogout may mimic many of the symptoms associated with septic arthritis. Therefore, synovial fluid samples should be examined by compensated polarizing light microscopy for the presence of negatively birefringent (uric acid) and positively birefringent (calcium pyrophosphate dihydrate) crystals to rule out crystalline joint disease. However, it must be noted that simultaneous bacterial infection and crystalline disease has been reported . Reactive arthritis. : An inflammatory joint response to extra-articular rather than intra-articlar presence of microorganisms may be defined as reactive arthritis . Therefore, while infection can be demonstrated at a distant site, joint inflammation occurs without traditional evidence of sepsis at the affected joint(s). Most cases are associated with patients with the major histocompatibility complex antigen HLA-B27. Also, patients usually have recent microbial infections in distal sites that include the gastrointestinal (e.g., Shigella spp., Salmonella spp., Campylobacter spp., or Yersinia spp.), genitourinary (e.g., chlamydiae and mycoplasmas), and respiratory (e.g., Streptococcus pyogenes) tracts . Patients present with a sterile, inflamed joint and may also demonstrate enthesopathy, uveitis, conjunctivitis, or skin and mucous membrane lesions . Specifically, poststreptococcal reactive arthritis can follow group A streptococcal infection and presents with nonmigratory arthritis, lack of response to aspirin or nonsteroidal anti-inflammatory agents, and the presence of extra-articular manifestations, including vasculitis and glomerulonephritis . Recent studies utilizing immunofluorescence, immunohistochemical, and PCR techniques have detected persistent microbial antigens within joints affected with reactive arthritis . These results may be explained by one hypothesis that describes the presence of bacteria and/or their antigens as a reflection of the persistence of small numbers of latent nonculturable microbes in the joint space. This hypothesis may be valid in specific cases (e.g., chlamydia-triggered reactive arthritis), since the early administration of tetracycline therapy may reduce the length of disease and the associated articular damage . However, antibiotics are usually ineffective, especially when given at later stages of reactive arthritis. Therefore, another hypothesis is that the detection of microbial products may just reflect the natural filtering action of the synovium and the subsequent concentration of these products, thereby stimulating inflammation. DIAGNOSIS OF GONOCOCCAL ARTHRITIS : Clinical Presentation | Gonococcal arthritis may present as part of a disseminated infection or as arthritis and usually affects young, healthy, and sexually active individuals. It is important to obtain a complete patient history in order to identify the presence of individual risk factors (a full list of potential risk factors for DGI is given in "Risk Factors" above). The presenting symptoms in DGI may include migratory arthralgias, moderate fever, chills, dermatitis, and tenosynovitis. The large majority of these patients have asymptomatic genital, anal, or pharyngeal gonococcal infections . The classic skin lesion manifests as small erythematous papules which progress to vesicular or pustular lesions and are often limited to the extremities and the trunk. If the papules are present on the affected joint, there are typically 5 to 10 lesions. The tenosynovitis is characterized by pain, swelling, and periarticular erythema. Some patients develop septic gonococcal arthritis without prior polyarthralgia, tenosynovitis, or dermatitis. In fact, while most patients with DGI present with tenosynovitis, only 21% of patients with confirmed suppurative arthritis display this clinical sign . Therefore, properties of the host and the serological properties of the infecting N. gonorrhoeae strain may be responsible for determining whether the DGI will result in tenosynovitis and dermatitis or will produce arthritis. In the absence of the characteristic dermatitis or overt genital infection, septic gonococcal arthritis is often clinically indistinguishable from other forms of septic arthritis . In contrast to nongonococcal arthritis, distal joints including the fingers, wrists, elbows, knees, and ankles are most often affected in gonococcal arthritis. Also, migratory asymmetric joint pain followed by polyarticular infection is common in this patient population. Laboratory Findings : Peripheral leukocytosis and elevated erythrocyte sedimentation rates are present in more than half of these patients. Also, N. gonorrhoeae is isolated from synovial cultures in only approximately 50% of patients with gonococcal arthritis, and Gram stains are even less reliable . Therefore, there is a high dependence on clinical presentation, accurate history, and positive cultures from affected sites for the diagnosis of this disease. Cultures derived from the uterine endocervix are positive in approximately 90% of women, while urethral, pharyngeal, and rectal mucosal cultures are positive in approximately 50 to 75, 20, and 15% of men, respectively. Blood and skin lesion cultures are rarely positive. It is extremely important for the bedside clinician to be aware of the specific requirements for correctly culturing N. gonorrhoeae from patient samples. Briefly, blood and synovial fluid samples should be plated immediately on prewarmed chocolate agar while genitourinary, rectal, and pharyngeal samples should be plated on prewarmed Thayer-Martin or modified New York medium with appropriate antibiotic supplementation . The plates should then be incubated at 37C in a moist 5% CO 2 environment within 15 min of sample harvest. It is also important to note that N. gonorrhoeae growth is inhibited in blood culture tubes containing polyanethol sulfate. Since N. gonorrhoeae DNA has been detected in culture-negative synovial samples by PCR amplification , this molecular biology tool may have a future widespread role in accurate gonococcal arthritis diagnosis. The specificity and sensitivity of this technique were 96.4 and 78.6%, respectively, and the false-positive rate was 3.6% . However, it is unclear whether a positive PCR result represents viable but nonculturable bacteria or nonviable bacteria with an associated reactive arthropathy. As with all PCR-based techniques, careful sample preparation and the inclusion of proper positive and negative controls are essential to maintaining the efficacy of this diagnostic tool. However, the generalized use of this molecular technique for routine screening and detection of N. gonorrhoeae will not replace the "gold standard" of culture since PCR-based methods do not yet provide information about antibiotic sensitivity. Imaging Studies : The utility of the imaging studies discussed above also applies to gonococcal arthritis, especially with regard to advanced cases and monitoring treatment success. However, these diagnostic tools are generally not used in the diagnosis of this infection. The extremely rapid clinical response to treatment, the lack of complicating manifestations, and distal joint involvement often make imaging unnecessary. Differential Diagnosis : The symptoms associated with N. gonorrhoeae joint infection can be mimicked by arthritis due to other bacteria. Arthritis due to N. meningitides is nearly indistinguishable from DGI, especially with regard to the musculoskeletal manifestations and arthritis-dermatitis syndrome. Skin lesions, similar to those produced in cases of gonococcal arthritis, are also occasionally induced by infection with other pathogenic species, including H. influenzae, Streptobacillus monoliformis, and Streptococcus pyogenes. However, patients with nonneisserial joint infections usually present with distinct clinical characteristics and laboratory findings (see above). As described above for nongonococcal septic arthritis, preexisting joint infection (including systemic lupus erythematosus, rheumatoid arthritis, and other connective tissue disorders), endocarditis, chronic infectious arthritis, viral arthritis, and crystal-induced arthritis should be considered in the differential diagnosis. Also, cases of gonococcal arthritis and reactive arthritis (specifically sexually acquired reactive arthritis) may be hard to distinguish since both involve sexually active people and common symptoms are seen, including urethritis, conjunctivitis, oral ulcers, and genitourinary manifestations. However, the patients are usually HLA-B27 positive, the onset is often slower and less acute, and the skin lesions (if present) are usually keratoderma blenorrhagicum and circinate balanitis in this form of reactive arthritis. Also, antibiotic therapy is usually ineffective. Preicteric hepatitis may also be confused with DGI since it commonly presents with tenosynovitis, polyarthritis, and skin rash. However, in hepatitis-associated arthritis, the rash usually resembles hives, the concentrations of synovial leukocyte are lower, and the hepatitis surface antigen can usually be detected in the blood. TREATMENT OF NONGONOCOCCAL ARTHRTIS : Acute nongonococcal septic arthritis is a medical emergency that can lead to significant morbidity and mortality. Therefore, prompt recognition and rapid and aggressive treatments are critical to ensuring a good prognosis. The treatment of this form of septic arthritis includes both appropriate antimicrobial therapy and joint drainage . Antibiotic Therapy : Most people with suppurative arthritis respond clinically to appropriate antimicrobial agents after the initial diagnostic joint aspiration. Initial antimicrobial therapy is based on the clinical presentation, a thorough history, initial Gram stain, and joint fluid analysis. The patient's history and clinical course often provide clues to distinguish between gonococcal, nongonococcal, and granulomatous arthritis. Joint culture collection followed by initiation of treatment with an effective broad-spectrum antibiotic should be done as soon as possible. The initial antibiotic therapy is adjusted, if necessary, based on appropriate culture and antibiotic sensitivity results . The usual course of therapy for nongonococcal arthritis is 2 weeks for arthritis due to H. influenzae or Streptococcus spp. and 3 weeks for arthritis due to S. aureus or gram-negative bacilli. Initial antibiotic therapy in children younger than 5 years includes cefuroxime, cefotaxime, or ceftriaxone depending on the blood and joint culture results. Initial therapy for patients older than 5 years is aided by the Gram stain. If clusters of gram-positive organisms suggestive of S. aureus are seen, treatment with intravenous (i.v.) penicillinase-resistant penicillin is begun. If gram-positive organisms in chains consistent with Streptococcus spp. are seen, penicillin G is used for therapy. If the Gram stain is negative, an extended-spectrum or broad-spectrum cephalosporin or semisynthetic penicillin is appropriate. Ceftriaxone is a reasonable initial antibiotic in sexually active adults. The initial antibiotic therapy is adjusted, if necessary, on receipt of appropriate culture and sensitivity results. Few controlled studies exist assessing the optimal duration, dose, or route of administration of antibiotics in nongonococcal arthritis . Most antibiotics achieve excellent bactericidal concentrations in synovial fluid following parenteral or oral administration . Intra-articular antimicrobial administration is usually not necessary and may cause a chemical synovitis. TABLE 2 | Initial choice of antibiotics for therapy of infectious arthritis (adult doses) Antibiotic Administration in the Elderly : Most cases of nongonococcal arthritis occur in the elderly, even though persons older than 65 years account for only 12% of the population; this relative percentage will only increase with our aging population . The choice of antibiotic therapy for these patients must be carefully made due to the decreased organ reserve capacity, altered pharmacokinetics and pharmacodynamics, and polypharmacy with associated drug-drug and drug-disease interactions characteristic of older patients, all of which cause a high rate of adverse drug effects in this patient population. In addition, low compliance with prolonged or complicated oral regimens must be considered. The physician can help to minimize adverse drug reactions and improve outcomes by being aware of the principles of clinical pharmacology, the characteristics of specific drugs, and the special physical, psychological, and social needs of older patients. The most important and best-studied pharmocokinetic alteration that occurs in the elderly is the age-associated decline in normal renal function. The creatinine clearance is a very useful measure of renal function in elderly patients and can be estimated by the Cockroft-Gault equation in which the creatinine clearance (in milliliters per minute) is assumed to equal the percentage of normal renal function: Antibiotic loading and maintenance doses should be estimated and confirmed by measuring peak and trough concentrations in serum after the fourth dose. The loading dose may be calculated by using the ideal body weight to estimate lean mass : The dose may be adjusted upward or downward to compensate for increased or decreased extracellular fluid volume. The maintenance dose should be estimated using ideal body weight and percentage of normal renal function. Prolonged use of aminoglycosides should be avoided if possible because of the increased risk for ototoxicity and nephrotoxicity in elderly patients. It is important to note that for drugs with appreciable renal clearance, such as vancomycin and aminoglycosides, monitoring of plasma drug levels is wise . In Australia, cases of cholestatic hepatitis were found in elderly patients (predominantly women) following 3 weeks of flucloxacillin treatment . Augmentin (amoxacillin-clavulanate) was also associated with instances of cholestatic hepatitis. However, this side effect was noted primarily in elderly men following Augmentin treatment for more than 2 weeks . Due to reports of seizures, the intravenous dose of 0.5 g every 6 h should be reduced in elderly patients with decreased renal function, cerebrovascular disease, or seizure disorders . Cefamandole may increase creatinine levels in the elderly. Seizures due to hypo- or hyperglycemia were noted in four elderly patients being treated with ofloxacin . Quinolones, fluoroquinolones, and tetracycline may have decreased oral absorption when coadministered with aluminum- or magnesium-containing antacids or sucralfate (Carafate). Quimapril (Accupril), a newly released angiotensin-converting enzyme inhibitor, contains a high concentration of magnesium, which may also decrease the oral absorption of fluoroquinolones and tetracycline. The interaction of rifampin with a large number of therapeutic agents requires close patient monitoring and follow-up . It is important to note that potent loop diuretics decrease the extracellular fluid volume, thereby elevating the levels of antibiotics in serum and requiring further reductions in dose levels. Surgical Therapy : There are a variety of methods to drain the purulent fluid from the infected joint. Presented in ascending order of invasiveness, cost, and effectiveness in the thoroughness of drainage, they include needle aspiration, tidal irrigation, arthroscopy, and arthrotomy. There is no set of universally accepted criteria for choosing the drainage method. It is agreed that the specific method of drainage used should be tailored to the clinical situation of the patient. However, some general guidelines can be listed. Patients should be initially treated with needle aspiration if a joint infection is easily accessible, if the vast majority of the purulent fluid can be removed, and if the patient does not suffer from negative prognostic indicators (see below). Although no prospective studies of these methods exist, most retrospective studies suggest that peripheral joints such as the knee, elbow, ankle, and wrist receive needle aspiration initially whereas axial joints, such as the hip, shoulder and sternoclavicular joint undergo open drainage . Repeated needle aspiration for recurrent joint effusions has been used with success during the first 7 days of treatment . If the volume of synovial fluid, the cell count, and the percentage of polymorphonuclear leukocytes decrease with each aspiration, then the combination of antimicrobial therapy and aspiration as needed is probably adequate . Persistence of effusion beyond 7 days is evidence that arthroscopy or open drainage should be performed. Tidal irrigation is as effective as arthoscopy and can be performed at the bedside. This closed-system irrigation method may be useful when needle aspiration results in incomplete evacuation or when multiple synovial fluid samples demonstrate different characteristics, indicating the presence of loculating pockets of infection. Arthroscopic lavage has been increasingly used in the treatment of septic arthritis of the knee. A recent study demonstrated that this method may also be effective for deep joints, such as the hip. Arthroscopy is advantageous in that extensive debridement can be performed with a small incision, thereby allowing for a more rapid and effective rehabilitation period. Further study of the efficacy of tidal irrigation and arthroscopy needs to be performed. Arthroscopy is a less invasive technique than open surgery and provides much better irrigation and visualization than needle aspiration . Aspiration under radiologic imaging or open surgical drainage with vigorous exploration and debridement is recommended for hip infections as well as for joint infections possessing adhesions or loculated areas of abscess . Arthrotomy should be used when an infected joint must be decompressed urgently because of neuropathy or compromised blood supply, when the infected joint is inaccessible by less invasive methods (such as the hip and sometimes the shoulder), when the joint has been damaged by preexisting disease, when bacterial arthritis is complicated by osteomyelitis, and when the less invasive methods of treatment fail. Also, when the isolated pathogen (e.g., P. aeruginosa) can be treated only with aminoglycosides, arthrotomy is often required to overcome the low oxygen tensions and pH of the infected joint. A number of patient factors have also been implicated as negative prognostic indicators in septic arthritis and may increase the need for invasive surgical intervention. Some of these factors include a long duration between symptom onset and treatment, complicated joint site, extremes of age, underlying illness, immunosuppressive drugs, underlying joint diseases, presence of juxta-articular osteomyelitis, and chronic failure of less invasive methods to clear the infection as demonstrated by positive blood or syovial fluid cultures, continued back pain, and restriction of motion. During the acute phase of bacterial arthritis, patient rest and optimal joint position are absolutely required to prevent the occurrence of joint deformation and deleterious contractures. Splints may be used to maintain proper joint position (hip in neutral rotation in some abduction, knee in full extension, elbow in flexion at 90, and forearm in neutral rotation). Isotonic exercise is often helpful in preventing muscular atrophy. Following the acute phase, early physical therapy and aggressive mobilization are vital for optimal recovery . TREATMENT OF GONOCOCCAL ARTHRITIS : The treatment of gonococcal arthritis strongly relies on appropriate antimicrobial therapy, and surgical procedures besides aspiration are rarely indicated. Patients should initially be hospitalized and should remain in this setting until 1 or 2 days following symptom resolution or for the entire length of therapy for patients who cannot be relied on to comply with treatment. The patient should also return 1 week after completion of the prescribed antibiotic regimen for follow-up, and clinicians should obtain and analyze synovial fluid samples of all previously affected joints at this time. In the United States, nearly 30% of all N. gonorrhoeae isolates are resistant to penicillin, tetracycline, or both . Therefore, the Centers for Disease Control and Prevention suggest that patients with gonococcal arthritis should be treated initially with parenteral ceftriaxone (1 g intramuscularly [i.m.] or i.v. every 24 h) . Therapeutically equivalent doses of other broad-spectrum cephalosporins (e.g., cefotaxime 1 g i.v. every 8 h or ceftizoxime 1 g i.v. every 8 h) are effective . The tetracyclines (except in pregnant women) or penicillins may be used if the infecting organism is proven to be susceptible. Skin lesions may continue to develop for up to 2 days following the initiation of antibiotic therapy. These lesions are often due to the localization of host complement complexes in the skin. The treatment may be switched to oral antibiotic therapy with a quinolone (ciprofloxacin 500 mg orally twice a day or ofloxacin 400 mg orally twice a day), except in pregnant women or young children, or cefixime (400 mg orally twice a day) to complete 7 to 10 days of total therapy 48 h after clinical improvement begins . It should be noted that resistance to ceftriaxone and cefixime is rare in the United States. Also, only approximately 1.4% of all N. gonorrhoeae isolates demonstrate intermediate or full resistance to ciprofloxacin . Therefore, these antibiotics are still highly effective in the treatment of DGI. Patients indicating penicillin allergies should be given spectinomycin (2 g i.m. every 12 h). However, the clinician must be aware that this antibiotic shows poor activity against pharyngeal N. gonorrhoeae infection. Therefore, cultures should be performed on these patients 3 to 5 days following treatment. In the Western world, spectinomycin-resistant gonococcal isolates are a rare occurrence . However, resistance rates of up to 10% of isolates for this antibiotic have been demonstrated in a few countries . Alternative antibiotics in the beta-lactam-allergic patient may be ciprofloxacin (500 mg i.v. every 12 h) or ofloxacin (400 mg i.v. every 12 h). Children weighing more than 45 kg should be treated with a single daily dose of ceftriaxone (50 mg/kg and a maximum dose of 2 g, i.m. or i.v.) for 10 to 14 days. For children weighing less than 45 kg, a 7-day parenteral ceftriaxone regimen (50 mg/kg and a maximum dose of 1 g, i.m. or i.v. in a single daily dose) is recommended. In geographic areas with high rates of N. gonorrhoeae and Chlamydia trachomatis coinfection, doxycycline or azithromycin may be added to the antibiotic treatment regimen since the cost of therapy for chlamydia is often lower than the cost of testing . Surgical management of the affected joint is usually not necessary, with the exception of the initial joint aspiration for synovial fluid sample collection at presentation. The diminution of symptoms is often rapid in these patients, and so subsequent joint drainage is often unnecessary. However, in cases of persistent effusion, the affected joint should be repeatedly drained as needed. In rare, very advanced cases, tidal irrigation, arthroscopy, and arthrotomy may play a role in disease resolution. While most patients display a dramatic response to therapy, some patients, especially those with large joint effusions or high erythrocyte sedimentation rates, may require longer hospital stays. A SPECIAL CASE: PROSTHETIC JOINT INFECTIONS : The increased use of implanted prosthetic joints has provided a physiological niche for pathogenic organisms to cause septic arthritis. In fact, prosthetic joint implantation and replacement is the single most common cause of joint infections. The prevalence of infection after total knee or hip arthroplasty is estimated to be approximately 1 to 2%, while in patients with rheumatoid arthritis, the incidence rate can climb to 4.4% . If the infection is of recent onset (<3 months after surgery), it was probably the result of surgical contamination. In this setting, Staphylococcus epidermidis predominates as the major isolate. However, late-onset infection is usually caused by hematogenous seeding, and S. aureus is the most common isolate, followed by Streptococcus spp., gram-negative bacilli, and anaerobes. An inherent problem associated with implants is their propensity to be coated by host proteins such as fibrinogen and fibronectin shortly after implantation . In the short term, fibrinogen and fibrin seem to be the dominant coating host proteins, while fibronectin becomes dominant in the long term since fibrinogen and fibrin are degraded. Implants can then act as a colonization surface to which bacteria readily adhere, like the fibrinogen and fibrin binding receptors of S. aureus. Also, implants are often responsible for reduced blood flow and local immunocompromise by impairing natural killer, lymphocytic, and phagocytic cell activities. These implanted devices have also been linked to decreases in the amount of superoxide, a mediator of bacterial killing within professional phagocytic blood cells . Another mechanism by which implanted medical devices produce local immune compromise is through frustrated phagocytosis . In this case, professional phagocytes may undergo apoptosis when encountering a substrate of a size that is beyond their phagocytic capability. The resulting release of reactive products of oxygen and lysosomal enzymes may cause accidental host tissue damage and local vascular insufficiency, thereby increasing the predisposition of osteomyelitis development. Also, some of the normal phagocytic processes are devoted to removal of the implant foreign material (particularly with metals, methylmethacrylate, and polyglycolic acid), thereby utilizing the energy and resources of the immune system that would normally be used to fight infection . Therefore, prosthetic implants not only provide a substrate for bacterial adherence but also limit the ability of the host to adequately deal with the infection. Once colonized, bacteria (such as staphylococcal species) are able to synthesize a "slime" layer, termed the glycocalyx or biofilm. This layer prevents the inward diffusion of a number of antimicrobials and host phagocytic cells, allowing bacteria to escape from the effects of antimicrobial therapy and host clearance . Once an implant is colonized and osteomyelitis ensues, the only treatment option is implant removal. Finally, it has been shown that nasal carriage of the organism by the patient was the most important risk factor associated with surgical site infection . Therefore, it may be a worthwhile goal to eliminate S. aureus nasal carriage prior to invasive procedures. The risk of implant infection may be increased by a number of factors. First, certain joint replacements are more susceptible to infection because they remain close to the surface and have poor soft tissue coverage (e.g., total elbow arthroplasties) or require prolonged surgery. Second, certain patient populations are at increased risk because of underlying conditions or systemic diseases, including patients with diabetes mellitus and rheumatoid arthritis ; also, patients who are elderly, obese, or malnourished or who have undergone prior surgery at the implantation site are at risk. Third, polymethylmethacrylate bone cement may be inhibitory to the activity of leukocytes and complement function; also, the heat released during polymethylmethacrylate polymerization may kill the juxtaposed cortical bone, thereby creating a nonvascularized area and providing the bacteria with a lush growth environment while being sealed off from the circulating host defenses. The clinical presentation of prosthetic joint infections of early onset (<3 months postimplantation) is much like that of acute septic arthritis and includes joint swelling, pain, leukocytosis, and a febrile response. In contrast, patients with late-onset infections, while demonstrating an elevated erythrocyte sedimentation rate, are often afebrile (50%), lack leukocytosis, and have less pronounced clinical features and gradually progressive joint pain. Imaging studies can be used but are often unable to distinguish between hardware loosening, a noninfectious inflammatory response, and active infection. Therefore, accurate diagnosis depends on collecting periprosthetic and fluid samples by needle aspiration in suspected cases of infected prosthetic knees or by arthrotomy in cases involving infected hips. If caught rapidly, early-onset prosthetic joint infections may be successfully treated with antibiotics alone or in combination with debridement without prosthesis removal, especially when rifampin in combination with another culture-directed antibiotic is used. However, in most cases, the disease has progressed to a state in which the hardware must be eventually removed to cure or arrest the bone infection. A two-stage procedure of implant removal and debridement (stage 1) and reimplantation (stage 2) is recommended. This procedure should focus on the state of the patient rather than the specific organism when determining the interval between stages since recurrence is usually associated with the quality of the initial debridement, not with the infecting organism. Earlier attempts at arthrodesis rather than repeated attempts at reimplantation are recommended. On occasion, patients can be given suppressive oral antibiotic therapy if they refuse implant replacement or if surgery is prohibitive. Nevertheless, prosthesis removal will eventually have to be performed. Another 4- to 6-week course of culture directed antibiotic therapy should be administered following the last major surgery. Since recurrence rates have been found to be up to 60% in patients with rheumatoid arthritis, these patients should be monitored. PROGNOSIS : A permanent reduction in joint function is seen in approximately 40% of patients with nongonococcal septic arthritis but ranges between 10 and 73% (, -). This wide range of observed morbidity reflects the dependence of therapy success on host, bacterial, and diagnostic and treatment factors. The mortality associated with this disease is usually between 5 and 20% and is often a result of the transient or chronic bacteremia that causes most cases of septic arthritis (, -, ). This high rate has not changed significantly over the last 40 years, even with present-day diagnostic and treatment options . The results of treatment vary greatly with the number of indicators of poor prognosis. Patients who start treatment after experiencing symptoms for 7 days or more demonstrate a poor outcome. Therefore, prompt diagnosis and rapid initiation of therapy are of the utmost importance in limiting the morbidity associated with septic arthritis. In addition, early physical therapy and aggressive mobilization are important for optimal recovery . A delay in diagnosis can also lead to a longer time being taken to clear the joint infection with appropriate therapy. An extended time (>6 days) required to sterilize the joint is another indicator of poor prognosis . The outcome in patients with septic arthritis due to some of the more virulent organisms such as superantigen-producing S. aureus and certain gram-negative bacilli is poor in spite of the use of optimal therapy . Elderly patients demonstrate a high mortality (19 to 33%) associated with septic arthritis since they often have preexisting medical conditions (e.g., diabetes mellitus) and joint diseases (e.g., osteoarthritis and rheumatoid arthritis) . In addition, these patients are more susceptible to a number of infections than are younger adults . The decline in natural and induced immunity in elderly patients causes a generalized reduction in the immune response to foreign antigens. The greater susceptibility to infections is due to the effects of age on the immune system and to immuno suppression caused by age-related illnesses. Specifically, the deficient immune response to foreign antigens results from the loss of thymic and T-lymphocyte function (mainly related to the production and response to IL-2) and associated decrease in antibody production by B cells . Underlying joint disease (e.g., osteoarthritis or rheumatoid arthritis) is another indicator that despite optimal treatment, the patient has a poor prognosis despite optimal treatment . This poor prognosis is often due to a delayed diagnosis since the clinical symptoms of septic arthritis are often mistaken for symptoms related to the preexisting joint disease. Patients who present with polyarticular septic nongonococcal arthritis have a very poor prognosis (aa. 30% mortality) due to the associated bacteremia and a reduced ability to resist the infection. Polyarticular septic nongonococcal arthritis may result in even higher rates of mortality when seen in patients infected with staphylococcal species (up to 56% mortality) or those with concomitant diagnosis of rheumatoid arthritis (up to 49% mortality) . The prognosis for patients with gonococcal arthritis is very favorable, with a rapid diminution of symptoms and a full return of joint function. In rare cases of DGI (i.e., 1 to 3% of cases), complications such as endocarditis, pericarditis, osteomyelitis, pyomyositis, perihepatitis, and meningitis may occur. Backmatter: PMID- 12364378 TI - Effects of Global Climate on Infectious Disease: the Cholera Model AB - Recently, the role of the environment and climate in disease dynamics has become a subject of increasing interest to microbiologists, clinicians, epidemiologists, and ecologists. Much of the interest has been stimulated by the growing problems of antibiotic resistance among pathogens, emergence and/or reemergence of infectious diseases worldwide, the potential of bioterrorism, and the debate concerning climate change. Cholera, caused by Vibrio cholerae, lends itself to analyses of the role of climate in infectious disease, coupled to population dynamics of pathogenic microorganisms, for several reasons. First, the disease has a historical context linking it to specific seasons and biogeographical zones. In addition, the population dynamics of V. cholerae in the environment are strongly controlled by environmental factors, such as water temperature, salinity, and the presence of copepods, which are, in turn, controlled by larger-scale climate variability. In this review, the association between plankton and V. cholerae that has been documented over the last 20 years is discussed in support of the hypothesis that cholera shares properties of a vector-borne disease. In addition, a model for environmental transmission of cholera to humans in the context of climate variability is presented. The cholera model provides a template for future research on climate-sensitive diseases, allowing definition of critical parameters and offering a means of developing more sophisticated methods for prediction of disease outbreaks. Keywords: INTRODUCTION : The relationship between human health and climate is not a new concept, nor is it a new subject. In fact at least as far back as Hippocrates, many believed that human health was intricately linked to the seasons, local weather patterns, and other environmental factors . While preceding the advent of Pasteur's germ theory in the mid-1800s, these notions described certain patterns of disease, and often therapies were based directly on changes in the environment. Proponents of miasma, i.e., a poisonous atmosphere arising from swamps and putrid matters, as a source of disease noted that fevers and intestinal ailments were common in warm seasons and were often associated with wet, poorly drained, and humid areas, such as swamps . This association motivated draining of swamps in rural areas and the installation of the first central sewer systems in urban areas in North America . Of course, the therapeutic benefits of these practices had little to do with ridding the environment of miasma, but these practices did reduce the size and number of habitats breeding mosquitoes (disease vectors) and removed raw waste (hence, waterborne pathogens) from immediate contact with humans. In spite of the developments in the mid- to late 1800s demonstrating the role of microorganisms in disease, the theory of weather-borne diseases held firm in early medicine until the late 1800s . Clearly, the relationship between climate and health has been a subject of study for a very long time. However, during the past 2 decades, development of modern tools and technologies has led to fascinating observations sparking new interest in the role of the environment, including weather and climate, in infectious disease dynamics. Scientific interest has been further stimulated by the growing problems of antibiotic resistance among pathogens, emergence and reemergence of infectious diseases worldwide, the potential threat of bioterrorism, and the debate concerning climate change. During the past few years, scientists and several agencies, including the World Health Organization, American Academy of Microbiology, Intergovernmental Panel on Climate Change (IPCC), and U.S. Global Change Research Program, among others, have published papers and issued reports highlighting the topic of climate and health (e.g., see references , , , , , , , , and ). Concerns over the impact of anthropogenic alterations to both terrestrial and aquatic habitats, coupled with a changing global climate, have helped to spawn an expanding cross-disciplinary effort to understand how such changes might affect human health. Progress is under way in using climate factors in predictive models for certain diseases, notably cholera. CLIMATE AND HEALTH : Part of the growing interest in the effects of climate on health is due to concerns about global climate change and variability. In concurrently appearing papers in Science, Barnett et al. and Levitus et al. independently reported warming in the top 3,000 m of the global ocean since 1950. Furthermore, models incorporating anthropogenic forces (e.g., gases) match the observed changes in oceanic heat content . These studies provide more evidence of human-induced climate change than those models that rely on near-surface temperatures, largely because the heat capacity of the oceans gives a more stable parameter than the surface temperatures . Superimposed on global trends or long-term changes in climate are periodic signals in the ocean-atmosphere that result in climate variability at interannual to interdecadal scales. In particular, the El Nino-southern oscillation (ENSO) has been the focus of much attention in the popular press and scientific communities. This phenomenon originates in the eastern equatorial Pacific, offshore from Peru, where every few years near Christmastime (hence, "El Nino," the Christ child) the normally cool waters of this region are suppressed by a layer of warm water arriving from the west. This event triggers changes in weather patterns across the globe and is considered to be among the most-important factors in global climate variability . Climatologists have also begun to look at global impacts of other modes of atmospheric variability associated with large ocean basins, e.g., the North Atlantic oscillation (NAO), where decadal oscillations in the atmospheric pressure difference between Iceland and the Azores affect wintertime weather in Europe and the Atlantic coast of the United States. Research is under way to improve prediction of such climate signals, which is poor in comparison to the better-understood ENSO (J. Marshall, Y. Kushnir, D. Battisti, P. Change, J. Hurrell, M. McCartney, and M. Visbeck, unpublished data ). Although controversy remains over how human activities might be driving climate variability and change, most scientists in the field agree that our climate is changing and there is an increasing need to understand the potential outcomes of such changes on human health. MODES OF DISEASE TRANSMISSION : Vector Borne | Vector-borne diseases are those that require a host (most often arthropods or rodents) to transmit a pathogen. Variability in climate and weather can affect both the vector population and the amplification of the pathogen within the host. Malaria and dengue fever continue to be the most-important vector-borne diseases worldwide . Unlike yellow fever and rift valley fever, which affect humans only incidentally, malaria and dengue are transmitted directly by infected mosquitoes that commonly bite humans. Other mosquito-borne infections, including rift valley fever, yellow fever, and encephalitis, are transmitted between the mosquito vector and a nonhuman animal host under normal circumstances and affect humans only incidentally. Plague and Lyme disease, among others, are transmitted to animals by fleas and ticks, respectively, and hantavirus is transmitted by rodents . While humans are not a normal host for some of these agents, they can all pose a threat when the pathogen uses a vector that bites or cohabitates with humans. Vector-borne diseases remain the most-studied group of diseases in the context of climate variability. The geographical range of the vector population is indicative of ecological and climatic conditions that limit the distribution of both the host and the pathogen. Often, ecological disturbances, including weather-related events, can cause a shift in host vectors or change in habitat that results in a greater likelihood of the pathogen coming in contact with humans, thus increasing the risk of human infection . In border areas where conditions are not optimal for the pathogen and/or vector, small ecological disturbances or changes in weather may also alter conditions for disease transmission by creating a more favorable environment . It is often in such border regions that the most-dramatic results of climate change and/or variability are recognized. In the recent literature, many researchers have documented relationships between changes in weather due to the ENSO and certain vector-borne disease outbreaks (e.g., see references , , , and ). In this review, we have adopted a more inclusive definition of vector-borne disease that encompasses pathogens that do not require an obligate host (vector) for transmission but which are likely to be transmitted to humans by such a route. Below we present how this definition applies to Vibrio cholerae. Airborne | As a whole, little is known about airborne infectious diseases and how they are affected by climate. While there is a growing body of work focused on health impacts associated with air pollution and climate change, these studies largely focus on noninfectious agents (ozone and particulate matter, etc.) or allergens . Influenza- and rhinoviruses, which cause flu and colds, respectively, are recognized as infectious airborne pathogens. Indeed, there does seem to be a regular season for flu and colds, and climatic factors are suspected to contribute in part to these cycles. In the Northern Hemisphere, flu cases peak in the late fall and winter, but there is no significant correlation with mean temperature . Humidity, however, may affect survival of aerosolized virus particles . In the southwestern United States, Coccidioides immitis, a soil fungus responsible for valley fever, causes higher infection rates during dry periods that follow a rainy season, when wind can distribute the pathogen . Dust-borne agents distributed in the air are also affected by large atmospheric phenomena. Relative to normal conditions, large numbers of viable microorganisms have been isolated in Caribbean air samples during "African dust events" . Periodically, massive plumes of dust are transported across the tropical Atlantic from the sub-Saharan region of Africa. Such events have been exacerbated by drought in the Sahel region, overgrazing, and the drying of Lake Chad and may be related to the NAO (Marshall et al., unpublished data ). To date, microorganisms found in the dust particles include Aspergillus spp., Pseudomonas spp., and Sphingomonas spp., among others ; however, it remains unknown if any human infections can be related to the African dust events. Similar phenomena are associated with the Gobi Desert and wind-swept dust from this region of the world. Waterborne | The quality and quantity of drinking water, irrigation water, and environmental and/or recreational waters can be associated with changes in environmental conditions including weather- or climate-related variables. Floods may cause the overflow of wastewater treatment plants, failure of septic systems, or combined sewer overflows, which could contaminate nearby surface waters or wells. Furthermore, there is increasing concern about pathogens in storm water runoff . Maintaining sanitary water conditions is also an issue during drought conditions, when contaminants may become concentrated in available water. Additionally, the likelihood of multiple uses in a water body may increase (e.g., for cleaning, bathing, and drinking) during droughts and consequently enhance the risk of contamination and exposure. Increasingly, the importance of the effect that weather and climate have on water quality, and not simply quantity, is being realized. Severe weather events appear to be correlated with enteric diseases, such as outbreaks of cryptosporidiosis related to excessive demand placed on sewage treatment plants from heavy rains and flooding . A retrospective study of drinking-water-related outbreaks of acute gastrointestinal illness in the United States by Rose et al. revealed that 20 and 40% of groundwater and surface water outbreaks, respectively, between 1971 and 1994 were statistically associated with extreme precipitation. Other work has also demonstrated an ENSO connection to rates of enteric illness in South America, levels of enteric microorganisms in coastal areas of south Florida (United States) , and cholera in Bangladesh . In addition to enteric pathogens, climate also influences the abundance and ecology of nonenteric and other pathogens which are naturally present in the environment. The best examples of this include the pathogenic Vibrio spp., which are autochthonous in estuarine ecosystems. Vibrio parahaemolyticus and Vibrio vulnificus are responsible for a majority of the nonviral infections related to shellfish consumption in the United States and also result in infections of open wounds during recreational exposure, such as swimming and fishing. Along with V. cholerae, these bacteria thrive in warm waters of moderate salinity and are closely associated with aquatic invertebrates . Therefore, with a changing climate, the geographic range of these pathogens may also change, potentially resulting in increased exposure and risk of infection for humans. Furthermore, changes in plankton populations, and other hosts for which vibrios are commensals or symbionts, would similarly alter the ecology of these pathogens that are autochthonous to the aquatic environment. CHOLERA AS A MODEL FOR CLIMATE-RELATED INFECTIOUS DISEASE : Cholera, caused by V. cholerae, lends itself to the study of the role of climate in infectious disease. This disease has a historical context linking it to specific seasons and biogeographical zones. Although cholera is an ancient disease and had disappeared from most of the developed world during the second half of the 20th century, it persists in many parts of the world with serious epidemics, most often centered in tropical areas. The association between the disease and water has been long held in folk tradition but was first established epidemiologically in the 1850s when John Snow made the link between the disease and a shallow well in London, United Kingdom . Subsequently, extensive studies demonstrated that V. cholerae is, in fact, native to coastal ecosystems . Vibrios, including V. cholerae, can be found in virtually any coastal water body, especially in the tropics and subtropics, when appropriate techniques are used. Recently, ecologically based models have been developed which define the role of environmentally, weather-, and climate-related variables in outbreaks of the disease . Data on the association between plankton and V. cholerae have been gathered over the last 20 years and strongly support the hypothesis of a commensal or symbiotic relationship between zooplankton and this bacterium. Because V. cholerae is highly concentrated on zooplankton carapaces and in their gut, the risk of ingesting an infectious dose increases when untreated surface waters are used for consumption. In this way, cholera shares some properties of a vector-borne disease. This hypothesis is currently under study in our laboratory at the Center of Marine Biotechnology and elsewhere (E. Whitcomb, personal communication, 2001). Cholera and V. cholerae offer an excellent example of the effect of climate and weather on infectious diseases and pathogens. Here we discuss some of the recent research on the subjects of V. cholerae, climate, and health and describe a framework for future studies using the cholera model. VIBRIOS AND V. CHOLERAE : During a cholera epidemic in 1854 in Florence, Italy, Pacini first described the comma-shaped gram-negative Vibrio, the "comma bacillus," responsible for cholera, which was subsequently named V. cholerae by Robert Koch. Pacini carried out detailed studies on the etiology of the disease. Pacini studied cholera for about 20 years and published several articles in which he described the destruction of the intestinal mucosa during cholera, ascribing it to the bacterium. Furthermore, Pacini clearly stated that the Vibrio was the "specific" causative agent of the disease and insisted the disease was contagious. However, at this time, the idea that cholera was contagious was under dispute. His data were ignored, and the etiological agent of cholera was rediscovered by Robert Koch in May 1884 (30 years after Pacini's description and 1 year after Pacini's death). Microscope slides from Pacini's studies on cholera are on exhibit in the Department of Anatomy and Histology of the University of Florence . During the same period of time that Pacini did his research, cholera ravaged Western Europe, and John Snow made the link between cholera and a drinking water source in London, establishing cholera as a waterborne disease. Although V. cholerae is the cause of a devastating diarrheal disease, it is also a natural member of the aquatic microbial community, a finding not clearly understood until more than 100 years later . Vibrio, a diverse genus, is the most intensely studied of the Vibrionaceae family and has engendered a great deal of interest in the taxonomy and systematics of Vibrio spp. . Taxonomic debates aside, Vibrio spp. are among the most commonly isolated bacteria in marine and estuarine waters . They significantly affect nutrient cycling in these habitats and often comprise a major portion of the natural flora. Furthermore, several of the members of the genus are pathogenic either for humans or marine animals. To date, ca. 200 serogroups of V. cholerae have been recorded, of which only two (O1 and O139) have been associated with major epidemics. Furthermore, genes for cholera toxin (CT) production are rarely found in serogroups other than O1 or O139. Prior to the seventh pandemic, which started in 1961, outbreaks were traced to the classical biotype of V. cholerae O1. V. cholerae O1 El Tor strains largely replaced the classical biotype by 1961 . Toxigenic V. cholerae O139 first emerged as a pandemic threat in southeastern India and the Bay of Bengal in 1992 and appears to have arisen by genetic exchange with V. cholerae O1 El Tor . Recent evidence also suggests that O139 isolates may have arisen by genetic exchange with non-O1 V. cholerae strains as well as clinical strains of O1 (; S. M. Faruque, M. S. Islam, G. B. Nair, and R. R. Colwell, unpublished data). Before the seventh pandemic, V. cholerae O1 El Tor was considered to be a different species, referred to as Vibrio eltor . Similarly, V. cholerae isolated from the environment and clinical strains that did not agglutinate with O1 antiserum were designated nonagglutinating vibrios and were believed to be a different species from clinical isolates of V. cholerae O1 . Over the last 30 years, biochemical and molecular tests and polyphasic numerical taxonomy have proven that O1 and non-O1 strains of V. cholerae are the same species . Recent work using 16S rRNA gene sequencing has also shown no differences at the species level between V. cholerae O1 El Tor and classical strains or between V. cholerae O1 and non-O1 strains. While environmental and clinical strains of V. cholerae are now known to represent a single species, there is significant genetic diversity among environmental and non-O1, non-O139 strains of V. cholerae . Molecular tools have been used to fingerprint and track toxigenic strains of V. cholerae . These types of analyses have demonstrated that O1 and O139 isolates group closely with one another, forming a tight cluster, while environmental non-O1, non-O139 isolates are generally diverse , as would be expected, given that ca. 200 serotypes have been described. Detailed investigation of the V. cholerae genome and of specific housekeeping genes revealed that recombination may be the major source of this variability, a common feature in the evolution of V. cholerae . Furthermore, the three main pathogenic clones of V. cholerae O1 (sixth pandemic [classical], seventh pandemic [El Tor], and U.S. Gulf Coast isolates) appear to have evolved independently . Evidence indicates that toxigenic strains may arise from environmental, nontoxigenic progenitors in coastal areas . The emergence of toxigenic O139 strains may have a similar history, with confirming evidence now being developed (S. M. Faruque, M. S. Islam, R. R. Colwell, and G. B. Nair, personal communication, 2002). Most V. cholerae strains, especially those from the environment, lack the genes required to produce CT, but the possibility of genetic exchange in the environment allows the potential emergence of new toxigenic clones. While horizontal gene transfer is common among V. cholerae, as well as between and among other bacteria, the significance of this phenomenon was not recognized until widespread patterns of antibiotic resistance in bacteria began to emerge in the 1950s . Since then, similar rapid evolution by acquisition of "foreign" DNA has also been implicated in the transfer of metabolic properties and virulence genes . In particular, horizontal transfer and genetic reassortment have been a key mechanism for the emergence of new toxigenic strains of V. cholerae as well as clonal diversity . The best example of horizontal gene transfer in V. cholerae is the transduction of the elements that code for CT, the CTX genetic element, and the V. cholerae pathogenicity island (VPI). The CTX element includes six genes in its core region: ctxAB, zot, ace, cep, and orfU. The VPI codes for 15 separate open reading frames, including the tcp (toxin-coregulated pilus) gene . Evidence now indicates that the CTX element is actually the genome of a distinct filamentous phage that lysogenizes V. cholerae . VPI may also be part of a phage genome, but the evidence is not as clear. The presence of tcp (part of the VPI) is required as a receptor for CTXPhi. Furthermore, this phage-mediated transfer occurs not only among strains of V. cholerae but also with the very closely related V. mimicus, an occasional source of diarrheal disease in humans . In addition to acquiring toxin-encoding genes, V. cholerae and other gram-negative bacteria have a selective advantage in their ability to enter a dormant stage, termed viable but nonculturable (VBNC), when environmental conditions are unfavorable for active growth and cell division. In 1982, Xu et al. first demonstrated this phenomenon in Escherichia coli and V. cholerae using direct viable count and fluorescent-antibody staining methods. In V. cholerae and other vibrios, this state, in which metabolically active cells cannot be cultured on microbiological media, is induced by changes in environmental conditions, including temperature (reduction) and salinity . Microcosm studies demonstrated that these cells could remain viable in the environment for years and continue to be capable of causing disease . The existence of VBNC cells continues to be debated in the scientific community, as the definition of living, as opposed to dead, bacteria is complex and difficult to articulate. Some maintain that bacterial cells not growing on culture plates are dead. Yet, Gonzalez et al. suggested bacterial cells should be considered dead only when they lose both culturability and cellular integrity. Early studies by Valentine and Bradfield , Postgate , Jannasch , Stevenson , Kurath and Morita , and others provide a framework for the concept of a resting stage for nonsporulating bacteria. Xu et al. concluded that cells that were unable to grow on conventional culture media but that responded to nalidixic acid treatment were VBNC. This phenomenon represents a state of dormancy and allows survival and persistence of bacterial cells in the natural or host environment. To date, at least 18 species undergoing a VBNC-type shift have been documented , and more-sophisticated methods are being employed that confirm continued viability . The ability to adapt to unfavorable conditions may be a key factor in the success of V. cholerae as an opportunistic pathogen and is most likely explained by its commensal existence with zooplankton, which themselves undergo diapause. CLIMATE, V. CHOLERAE, AND CHOLERA : Seasonality | Analysis of the frequency of specific disease cases throughout the year is a logical first-order assessment of the potential role of climate in disease transmission. For many diseases (vector-, air-, and waterborne) there is a clear seasonal trend in the detection or isolation of a pathogen and prevalence of disease. For cholera, a distinct seasonal pattern is evident, particularly in regions of endemicity. In the following sections we discuss environmental and climatic factors that drive the seasonality and occurrence of V. cholerae and the disease cholera. Endemicity. | For most of its known history, V. cholerae was believed by physicians and clinical workers to be only coincidentally associated with environmental waters and detectable only if there was contamination by an infected individual. While we now know that V. cholerae is autochthonous in riverine, estuarine, and coastal waters, as first proposed in 1977 by Colwell et al. , Cockburn and Cassanos reported that as early as the 1800s there was discussion about environmental connections of the disease. By the late 1800s, areas of endemicity had been defined, along with relevant descriptions of the environment, and an association of cholera with monsoons had been noted . V. cholerae was shown to withstand high-pH conditions, and the notion that ponds may be a potential source of infection had been put forward . In retrospective studies of cholera in India during the 19th century, there is clearly a series of epidemic years associated with monsoons and other weather events (Whitcomb, personal communication, 2001). Today, geographical areas once known to have experienced cholera epidemics can be characterized into three levels. Cholera-free communities are defined as having no locally acquired infections. In areas of cholera epidemicity, the disease diminishes after an outbreak. In regions of cholera endemicity, the disease does not disappear after an epidemic peak and returns in successive waves . Of particular interest and relevance to identifying environmental or climate factors that may promote epidemics is the understanding of dynamics of the disease in areas of endemicity. For example, what causes periodic oscillations in cholera outbreaks, and why are some areas more prone to endemism? Torres Codeco incorporated seasonality into a model of endemic cholera and showed that cholera infections occurred 2 to 4 months after the onset of V. cholerae growth in the environment; however, this predictive model was influenced by population size and rate of contact. Furthermore, only transient environmental reservoirs of V. cholerae were required to maintain endemism in poor communities . (i) Ecology and biogeography of V. cholerae. | Detection of V. cholerae, both non-O1 and O1 strains, has been reported from almost every part of the world where there are rivers and coastal regions . In regions no longer with endemic Asiatic cholera in North America, V. cholerae has been isolated from estuarine and coastal water collected along the Pacific, Atlantic, and Gulf coasts of the United States (e.g., see references and ), and it has also been isolated in Australia , Southeast Asia, the island nations of the South Pacific, Africa, Asia, and throughout Europe. V. cholerae has been similarly isolated in areas of endemicity (India, Bangladesh, and Central and South America) . Despite the prevalence of the causative organism worldwide, only certain regions of the world, mainly in the tropics and subtropics, maintain endemicity for the disease. The nature of this phenomenon is not entirely understood but is most likely related to both environmental and socioeconomic factors. Following the sixth pandemic (similar1950), areas with endemic cholera had been reduced to southeastern India and Bangladesh, a significantly narrower geographic region, historically . During the current seventh pandemic, the range of endemicity has expanded and includes areas of Africa and South and Central America . Where cholera is endemic, cases tend to demonstrate distinct seasonal trends. These patterns are strongly related to the ecology of V. cholerae in the environment, where high numbers are observed during times of warm water temperatures and zooplankton blooms . Over the last 25 years, the major cholera epidemics, including the first outbreak of O139 disease, have originated in coastal areas (; T. Ramamurthy et al., Letter, Lancet 341:703-704, 1993). Currently, the main regions of cholera endemicity include the coasts surrounding the Bay of Bengal, both Bangladesh and the Indian subcontinent, and coastal Latin America. In each of these three geographical regions, patterns of disease frequency follow similar trends and are most likely explained by the same physical or environmental drivers (see "Hierarchy in the cholera model" below). In Bangladesh, there is a bimodal distribution in the frequency of cases over an annual cycle. The first, and smaller, peak occurs in the spring. The larger peak, however, follows the monsoon season in the fall. The onset of epidemics coincides with dry weather and the warmest water temperatures of the year (August or September) . The frequency of cases is reduced to background levels of a few cases as the temperature decreases in winter . In Calcutta, India, cholera cases tend to peak in April, May, and June . In South America, where the disease has become endemic since its reemergence in 1991, cases are concentrated in the austral summer months (January and February) . The geographic limits of endemicity are also consistent and suggest a biogeographical pattern of endemic disease that follows the environmental habitat of the causative organism. Although inland countries report cases of cholera and freshwater plankton species harbor vibrios, case frequency and location of regions of endemicity increase near coastlines. Along the Indian subcontinent, there is a distinct boundary approximately 150 miles inland and to the northern part of the peninsula where the frequency of cholera cases rapidly diminishes . In South America, outbreaks start along the coast , and epidemiological studies in Mexico demonstrate that people living in coastal states are at a high risk of contracting the disease (; M. L. Lizarraga-Partida, I. Wong-Chang, G. Barrera-Escorcia, A. V. Botello, A. Flisser, L. Gutierrez, A. Huq, and R. R. Colwell, unpublished data). V. cholerae O1 is also endemic in regions where there are now only sporadic cases of cholera. Such endemic clones are found along the U.S. Gulf Coast and in Australia and are unique, although closely related, based on phage typing and gene sequence analyses . In these regions, sanitation practices are well developed and the waterborne disease occurs rarely, although V. cholerae is often detected in estuaries and shellfish . Consumption of raw or improperly cooked shellfish harboring V. cholerae, and not contaminated drinking water, causes the sporadic cases of diarrhea and occasional septicemia that do occur . Yet, as in the case of the range of areas of cholera endemicity, the geographic distribution of these endemic strains tends to be limited to the subtropical and subtemperate climates. Similarly, along the U.S. Gulf Coast and in Australia, the frequency of isolation of V. cholerae from both the environment and patients peaks during the warmer months of the year . Non-O1 isolates also follow this pattern, with sporadic cases occurring mostly in the late summer in the United States and in the early summer in Bangladesh (prior to the annual fall epidemic) . Europe also reports detection of V. cholerae during the summer and early fall . Seasonality in detection of the presence of V. cholerae in water and sediment in the Chesapeake Bay of the east coast of the United States, a cholera-free region, has been abundantly demonstrated (, , ; V. Louis et al., submitted for publication). (ii) Emergence of toxigenic strains from the environment. | When cholera reemerged in South America during the seventh pandemic in 1991, a new focus of endemicity was initiated. Many investigators have speculated on the reemergence of cholera in South America after its hundred-year absence, and several hypotheses have been put forth. One theory was that an isolated event led to the epidemics ---that is, an exchange of ballast water by a ship previously traveling through areas of cholera endemicity. It has also been proposed that given the similarity between strains isolated in Latin American and Asia (O1 El Tor), cells in the VBNC state may have been transported from Southeast Asia to Peru with a transfer of water associated with the 1991 El Nino event . V. cholerae may have also already been a feature of the nearshore aquatic environment in South America, but some environmental cue triggered the bacterial population numbers past a threshold point, resulting in outbreaks of the disease . Following this, further contamination with infected stool and poor sanitary conditions led to outbreaks and epidemics. However, hospital records from the months prior to the official onset of the epidemic in Peru (January 1991) indicate the occurrence of cases matching the description of cholera as early as October of the previous year . Furthermore, cases were detected at about the same time in several geographically distant locales, suggesting that multiple strains contributed to the epidemic . This scenario supports the notion that a population of V. cholerae already existed along the coast of Peru and South America. Furthermore, data show that both O1 and non-O1 isolates had been detected in South and Central America prior to the onset of epidemics, as early as 1978 . Recent molecular analyses also suggest that multiple strains may have been and continue to be involved in outbreaks or emergence of toxigenic strains. Prior to 1992, the origin of all cholera epidemics had been strains of V. cholerae O1; however, during that year cholera outbreaks were associated with a new toxigenic serogroup, V. cholerae O139, in India and for a short time this group replaced the V. cholerae O1 El Tor . Since the emergence of V. cholerae O139 strains in 1992, there has been much interest in determining the mechanisms by which this pathogenic serogroup could have arisen. Various researchers have now demonstrated that O139 strains associated with epidemics arose via an homologous recombination event. Part of the sequence for O-antigen biosynthesis in an O1 El Tor strain was lost and replaced with sequences derived from a non-O1 strain . This recombination event also led to the expression of a capsule in O139, a feature often associated with non-O1 strains . Furthermore, ribotyping of both toxigenic and nontoxigenic strains of V. cholerae O139 revealed that isolates might have arisen from at least two different progenitors, including non-O1 serogroups; half of the nontoxigenic isolates produced distinctly different ribotype patterns than those produced by V. cholerae O1 El Tor . Moreover, analysis of housekeeping genes in the sixth (classical) and seventh (El Tor) pandemic and U.S. Gulf Coast strains suggests that these pathogenic clones were derived independently from nontoxigenic non-O1 progenitors . Thus, given the likelihood of more recombination events leading to new toxigenic strains, it becomes increasingly clear that there is a need to understand the occurrence of and monitor for V. cholerae non-O1 in the environment. Interestingly, even in areas of endemicity it is often difficult to isolate V. cholerae O1 or O139 from the environment, particularly during interepidemic periods, while V. cholerae non-O1 is more readily detected . The first study to demonstrate the presence of V. cholerae year-round in Bangladesh was conducted in the late 1990s using direct detection methods . In attachment studies using copepods, there was no difference observed between V. cholerae O1 and non-O1 strains. However, other studies have shown that non-O1 V. cholerae strains survive better in the environment and are more resistant than O1 strains to detergents and chelating agents, which may be discharged into the environment . There was early speculation about the possibility of conversion between O1 and non-O1 serogroups, and phage-mediated seroconversion between biovars (Inaba and Ogawa) has been reported . Later, serogroup conversion was demonstrated in laboratory microcosms . Conversion from non-O1 to O1, and vice versa, was observed in all microcosms of both artificial seawater and natural water at various temperatures; however, seroconversion occurred earliest (within 5 days) at a salinity of similar10 and temperatures near 35C . Environmental conditions affect both the overall abundance and, potentially, the serogroup of V. cholerae in the environment. Surprisingly, even in areas of cholera endemicity, O1 strains are detected infrequently in the environment compared to non-O1 strains using traditional culture methods. Evidence for spontaneous seroconversion, in part, may explain this observation . Furthermore, in both endemic and cholera-free areas (e.g., Chesapeake Bay), increased environmental water temperatures correspond with increased detection rates of V. cholerae, which suggests a triggering factor may be responsible for enhancing the number of organisms in a given environment and condition (V. Louis et al., submitted for publication). Contributing to the apparent dichotomy between clinical (toxigenic) and environmental (nontoxigenic) strains is the issue of acquisition of virulence genes in the environment, as most environmental isolates harbor neither tcp nor ctxAB . Recombination and acquisition of foreign DNA appear to be common features among vibrios and V. cholerae . Genes for both TCP and CTX can be readily transduced into recipient strains via temperate phages. While this is a key issue in the emergence of toxigenic strains, how this occurs in the environment and/or intestine is still under investigation. Recent studies show that, in general, phage-mediated transduction may occur at higher rates in the marine and estuarine environment than originally thought. Jiang and Paul showed transduction rates of 1013 transformants year-1 in Tampa Bay Estuary, Fla. Furthermore, in the same subtropical estuary, prophage induction was demonstrated in 52% of surface water samples, and at least 41% of the bacterial isolates were lysogens (carrying a prophage in the genome or as a self-replicating element) . Work to date on the CTXPhi reveals that while there is no induction in the intestine , exposure to sunlight is a key factor in the induction of CTX prophage from the host and subsequent propagation of the phage particle . Environmental conditions, such as pH and temperature, also influence V. cholerae phage infection and lysogenizing of the host or recipient cell (E. K. Lipp, I. N. G. Rivera, M. Talledo, A. Neale, D. Karaolis, A. Huq, and R. R. Colwell, Abstr. 101st Gen. Meet. Am. Soc. Microbiol., abstr. Q427, 2001). Although continued research is needed, it appears that seasonal environmental factors may affect phage-host dynamics and acquisition of virulence genes to a significant degree. Hierarchy in the cholera model. | The spatial and temporal patterns of emerging and reemerging diseases have been changing throughout history due to a variety of factors. Their patterns as well as incidence or prevalence of disease are influenced by a complex interaction of direct and indirect factors. Among physical factors, temperature perhaps has the most direct and significant effect on the ecology of most bacteria. It is not any different for V. cholerae, because warmer temperatures in combination with elevated pH and plankton blooms can influence its attachment, growth, and multiplication in the aquatic environment, particularly in association with copepods . Here we revisit the model for environmental cholera transmission proposed by Colwell and Huq and suggest a hierarchy to interpret the significance of climate and the environment in cholera and V. cholerae dynamics. A schematic representation of the model is presented in Fig. , and examples are detailed in Table . (i) Abiotic factors. | Most Vibrio spp., including V. cholerae, are characterized by an increased growth rate at warm temperatures, which is evident in the higher rates of isolation in the environment during warm months. That is, between epidemics in areas of endemicity in the world it is exceedingly difficult to isolate V. cholerae O1 due to changes in salinity, lower water temperatures, and perhaps seroconversion; however, during periods of warmer water temperatures, success in isolation of V. cholerae O1 rises substantially. Although optimal salinity for growth is between 5 and 25, V. cholerae is one of the few Vibrio spp. that can withstand a salinity of 0, provided Na+ is available . With sufficient dissolved organic matter present, V. cholerae can also grow well at salinities near 45 . Studies conducted by Miller et al. suggested that strains of V. cholerae vary greatly in their survival in the culturable state under low-salinity conditions (0.05). Survival was not related to serogroup, source (clinical or environmental), or geographical origin of the strain . Because V. cholerae can withstand low-salinity conditions, it is well adapted to both fresh and brackish water of areas of endemicity such as Bangladesh. However, periodic intrusion of salt water appears to enhance survival in such environments . V. cholerae also thrives under high-pH conditions, a parameter that has been used to improve isolation of V. cholerae from environmental samples by enrichment (using alkaline peptone water at pH 8.0 to 8.6). Because of its low solubility in environmental water, iron is often a limiting micronutrient for both bacteria and algae. To deal with this, V. cholerae can produce iron-chelating siderophores to take up insoluble iron from the environment. This adaptive feature allows for prolonged survival when soluble iron is unavailable; survival is thus further improved when insoluble iron (Fe2O3) is available under alkaline pH conditions . Environmental conditions may also affect expression of virulence genes in V. cholerae. As cited above, sunlight can induce propagation of the CTXPhi phage , and the viability, i.e., culturability, of V. cholerae remains stable in full sunlight compared to enteric bacteria such as E. coli, which may impart some selective advantage to vibrios at tropical latitudes . Expression of CT is optimal at salinities between 2 and 2.5, independent of cell concentration . Furthermore, moderate levels of introduced iron also increase the expression of CT . Therefore, environmental triggers may become epidemiologically important for prevalence of the organism and its virulence (potentially resulting in shorter onset times and lower infectious doses). (ii) Phytoplankton. | Abiotic factors both affect and are affected by the biotic environment. Sunlight, temperature, and nutrients all influence the growth of phytoplankton and aquatic plants, which in turn alter the dissolved O2 and CO2 content of the water and, therefore, the pH of the surrounding water. Both direct and indirect effects of algal growth subsequently influence the population of V. cholerae in a given environment. Although this is a simplistic description, it does illustrate how environmental factors build upon one another to characterize cholera dynamics. Cockburn and Cassanos first proposed the theory that ponds in Bangladesh were the main source of infection to the community. They proposed that if the pH in ponds were sufficiently elevated, V. cholerae could outcompete other bacteria and reach infectious dose levels. Experimentally they showed a relationship between elevated pH and onset of cholera cases, which was also related to time of year, light, temperature, and precipitation. The driver of the temporal pH changes in their hypothesis was phytoplankton, which elevated the pH of the surrounding water by the sequestration of CO2 during photosynthesis , with the greatest changes occurring under conditions with elevated temperatures, low turbidity, and sufficient light. These insightful observations provided a foundation for the importance of environmental conditions in cholera dynamics but were still lacking critical information on other contributors influencing the ecology of V. cholerae. Other authors have suggested that there may be a more direct relationship between algae, or aquatic plants, and V. cholerae. In general Vibrio spp. are found at high concentrations in zooplankton samples relative to the surrounding water column community. Simidu et al. reported that nonfermenting bacteria such as pseudomonads are found in association with phytoplankton, whereas vibrios and aeromonads are more likely to be associated with zooplankton. Vibrios are also frequently isolated during red tides in Spain . V. cholerae produces mucinase, an enzyme that degrades mucin and mucin-like substances that are often encountered in the gut. Although clearly involved in intestinal colonization during the infection process, this enzyme can also degrade similar substances in plants and algae . Research has identified a close association between V. cholerae, algae (Volvox sp., Rhizoclonium fontanum, and green algae), and the mucilaginous sheath of Anabaena spp. (cyanobacteria) and aquatic plants . Direct detection from the environment compared with plate counts showed that V. cholerae cells were often VBNC . Islam et al. also report that there may be some relationship between virulence and increased toxin expression when V. cholerae is associated with algae (R. fontanum) over time, although the observed relationship may be due to indirect effects such as pH change. A few studies have shown that V. cholerae may increase its survival and persistence in the environment by association with aquatic plants or algae, and some research suggests that blooms of cyanobacteria, or other phytoplankton, may have a moderate effect on the multiplication of V. cholerae . However, epidemics of cholera in Bangladesh have two distinct peaks, one in late spring and one in late fall. Once an algal bloom disintegrates, the bacterial cells require another mechanism of survival and multiplication in the environment to cause later epidemics during a given year . Instead of any direct influence of phytoplankton on V. cholerae, it is more likely that the high nutrient levels arising from the breakdown of phytoplankton via algal viruses or other bacteria may have an indirect influence on bacteria other than those involved in phytoplankton disintegration, an interesting, but unproved hypothesis. High phytoplankton production does, however, produce food for zooplankton grazers, the tertiary stage in the cholera model. (iii) Zooplankton. | In addition to favorably affecting water column pH, increased phytoplankton production also provides additional nutrients for the next level of the food chain, namely, zooplankton . Chitinous organisms, e.g., copepods, amphipods, and other crustaceans, are prevalent among zooplankton populations, and chitin is the second most abundant organic compound found in nature. Approximately 1011 metric tons of chitin year-1 is produced in aquatic environments; >109 metric tons is produced by copepods alone . If chitin were not degraded, high levels of carbon and nitrogen would remain insoluble and inaccessible to most organisms. Using chitinase, marine bacteria including Vibrio spp. and V. cholerae play a significant role in chitin remineralization such that very little chitin can be detected in aquatic sediments . In 1979, Nalin et al. , building on the observations of Kaneko and Colwell (-), demonstrated significant adhesion to chitin by V. cholerae and multiplication of cells in chitin suspensions. In addition to providing a food source for vibrios and enhancing survival under starvation conditions , chitin also offers protection to V. cholerae at low temperatures and under acidic conditions (such as the human gut, which could be important if copepods carrying V. cholerae are ingested in drinking water) . Other vibrios besides V. parahaemolyticus and V. cholerae have been studied with respect to chitin, as all vibrios tested to date are chitinolytic. Mechanisms of adhesion to chitin by Vibrio furnissi have been extensively studied. Yu et al. found that populations of cells exhibited an adhesion-deadhesion behavior that allowed a number of free-swimming cells to relocate when a chitin source was depleted and, thereby, go in search of another source. Bassler et al. subsequently demonstrated that Vibrio cells are also positively chemotactic to products of chitin hydrolysis. In V. cholerae, surface proteins have been identified in the attachment to chitin particles . While the exact mechanisms of chitin colonization in V. cholerae are not known, there is a distinct chitin recognition system . Together, these studies suggest a quorum-type response in the colonization of chitin particles and, therefore, copepods and other chitinous animals. There may be a synergistic effect between phytoplankton and zooplankton in chitin colonization by V. cholerae. An alkaline pH of 8.5, often associated with algal blooms, was found to positively influence the attachment of V. cholerae to copepods . Huq et al. proposed that once cells of V. cholerae attach to zooplankton, they are protected from the external environment and begin to proliferate, taking advantage of the increased surface area and improved conditions of nutrition, the latter derived from the disintegration of phytoplankton and release of nitrogenous products into the water . It was further suggested that during interepidemic monsoon seasons in Bangladesh, when there is a significant alteration of nutritional conditions arising from seasonal changes in the chemical parameters of the water , V. cholerae may become nonculturable . The simple presence of crustacean copepods enhances the survival of V. cholerae O1, as demonstrated in laboratory microcosm experiments . In the field (Bangladesh), VBNC cells of V. cholerae O1 and O139, detected by microscopy using direct fluorescent antibodies, were found attached to chitin particles, an observation providing support for results obtained in laboratory experiments (A. Huq et al., Letter, Lancet 345:1249, 1995). Other investigators also confirmed results of previous studies showing V. cholerae O1 surviving for significantly longer periods of time when attached to copepods . Such protection afforded to VBNC cells by chitin or chitin-containing organisms in the environment and in the human digestive system surely may play a role in the epidemiology of cholera. Scanning electron microscopy (SEM) showed that egg cases and the oral region of copepods were sites of significantly enhanced attachment by V. cholerae . Attachment was found to be strong for V. cholerae but significantly less so for E. coli or Pseudomonas spp.; that is, preparation for SEM requires vigorous washing that did not separate cells of V. cholerae from sites where they were attached but did remove other bacteria. Dumontet et al. found that bacterial cells attached to copepods in the natural environment detached when SEM was employed. This is a particularly important finding because both freely swimming and nonculturable V. cholerae attach to copepod eggs and multiply rapidly . Studies are currently under way in our laboratory to determine if attachment is reversible. According to Oppenheimer et al. , during the Indian Ocean monsoon (rainy) season in June and July, zooplankton populations decrease because of reduced levels of nutrients. During August and September, after the monsoon season, levels of nutrients significantly increase and blooms of phytoplankton followed by zooplankton occur. Kiorboe and Neilson published a report on an extensive study of copepod production, demonstrating two distinct seasons for the production of eggs by several species of copepods. One peak occurs in February through April, and the other occurs during the months of August and September. Interestingly, the timing correlates very well with incidence of diarrhea and cholera cases, particularly in India and Bangladesh. In Japan, a seasonal relationship between culturable V. cholerae and zooplankton was not detected , but Japan, like the United States, enjoys excellent sanitation and safe drinking water; in Japan, epidemics of cholera do not occur, and cases tend to be sporadic and few in number. Nonetheless, results of the research by Oppenheimer et al. and Kiorboe and Neilson in a region of cholera endemicity fit very well into the original hypothesis of Kaneko and Colwell and Huq et al. that copepods play an important role in the survival, multiplication, and transmission of vibrios, including V. cholerae, in the natural aquatic environment. Cells of V. cholerae also colonize the gut of copepods. Singleton et al. proposed that there might be an environmental role for CT production by V. cholerae in this type of association. In the human intestine, binding to the CT receptors of epithelial cells eventually results in an efflux of Na+ and other electrolytes and water, resulting in diarrhea. If CT affects the epithelial cells of crustaceans in a similar manner, it would provide an effective mechanism to regulate the concentration of Na+ when it is not optimal for the host. CT could, thereby, play a role in the osmoregulation of the copepod . Evidence to date suggests that there may be a specific symbiosis between species of copepods and V. cholerae. A recent report stated that V. cholerae may also be associated with Chironomus sp. (midge) egg masses in freshwater , but V. cholerae is known to be more commonly found in association with zooplankton . Our preliminary analyses of data from Bangladesh ponds show only a very few chironomid egg masses, and none yielded V. cholerae cells. Furthermore, we detected no relationship of chironomid egg masses with cholera epidemics (S. Islam, A. Huq, G. B. Nair, and R. R. Colwell, unpublished data, 2002), whereas zooplankton samples from Chesapeake Bay, Peru, and Bangladesh all reveal the presence of V. cholerae when examined using direct staining and molecular genetic methods. Further evidence indicates that V. cholerae may preferentially associate with Acartia tonsa in coastal waters (M. L. Lizzaraga-Partida, personal communication, 2001). (iv) Transmission to humans. | Cholera is primarily a waterborne disease, and an epidemic may be enhanced by secondary transmission. Yet, it has been demonstrated that V. cholerae can be transmitted to humans via the environment, in drinking and cooking water, irrigation water, and shellfish. With active environmental monitoring systems using epifluorescence microscopy and molecular methods for direct detection, the presence of V. cholerae, ctxAB-carrying bacteria, and vibriophages in sewage and surface waters has been shown to precede the onset of cholera outbreaks by 1 to 4 months . Furthermore, cholera is dose dependent, with at least 104 cells required for infection . A single copepod may carry 104 to 106 cells of V. cholerae, and thus incidental ingestion of a few copepods in untreated drinking water could result in an infection . Although, V. cholerae can be transmitted directly in water, it also acts in a vector- or host-dependent manner because the presence of copepods substantially raises the dose of V. cholerae over that in copepod-free water. Preliminary studies show that Bangladeshi households which filter their pond water through sari cloth before drinking reduce their risk of cholera infections by approximately 50% . Extreme Events | Cholera is often considered a reemerging disease, in part because infections are appearing in novel communities or communities where the disease had been absent for many years, and the range of areas of endemicity is expanding. This may be due to changes in the environment or climate that promote favorable conditions for the pathogens and/or its invertebrate host. Because V. cholerae clearly plays a role in coastal ecosystems in association with plankton communities, it is obvious that this opportunistic pathogen cannot be eradicated, but, certainly, the disease can be controlled . Ecological investigations that focus on relating V. cholerae temporally and spatially to easily measured predictors or proxies offer a sound method to prevent exposure and, therefore, cholera. Progress has been made along these lines via historical examination of cholera cases in areas of endemicity and by defining clear seasonal trends and anomalies associated with climate variability and change. Climate variability and change. | Few would disagree that our climate is changing, although the drivers of climate change are a subject of continued controversy. Recent research has focused on the impact of such changes on various ecosystems and public health . The effects of climate variability driven by natural cycles (e.g., ENSO, NAO, and others) or by anthropogenic activities (greenhouse gas emissions, etc.) will superimpose anomalous weather patterns onto typical seasonal trends. Climate or weather may affect each of the levels in the cholera model presented here. Climatologists predict a 1.4 to 5.8C rise in mean temperatures over the next 100 years . Increasing temperatures would be expected to expand the range and increase the prevalence of V. cholerae and cholera both geographically and temporally, if public health measures are not implemented. For example, temperature shifts would alter the latitudinal distribution of plankton species. Should sea level rise, as anticipated with higher temperatures, inland areas would experience greater saltwater intrusion and increased levels of marine and estuarine bacteria, including V. cholerae. Changes in sunlight and/or UV intensity could increase the rates of induction and propagation of the CTXPhi phage and, thereby, increase the potential for the emergence of new toxigenic V. cholerae strains. Introduction of insoluble iron into the environment by agricultural practices, industrial pollution, and/or dust events, such as the transport of African dust to the Caribbean and southeastern United States, may lead to increased survival and virulence in affected populations. The introduction of dust may be related to modes of climate variability such as the NAO. In light of climatic and environmental drivers in cholera disease dynamics, it is necessary to take into account the capabilities of public health systems and improved sanitation systems. It cannot be assumed that the number of cholera cases will increase because the risk of cholera, especially in developing countries, may increase. The disease itself is preventable if sanitation is sufficient to protect public health and measures such as boiling of water can be implemented. While it is impossible to predict precisely the outcome of cholera dynamics with climate change, analyses of cholera and V. cholerae in the context of known variability, e.g., ENSO events, have been undertaken that demonstrate anomalous patterns in disease dynamics superimposed on normal seasonal trends. The ENSO story. | The ENSO is a normal mode of climate variability that, in simple terms, involves pressure shifts across the equatorial Pacific. During the El Nino phase, a zone of high pressure moves eastward, easterly trade winds slacken, and a slow-moving low-frequency wave moves from the west to the east, resulting in depression of the thermocline along the South American coast and warming in the eastern equatorial Pacific. During La Nina episodes the trend is somewhat reversed, and there is a cooling in the eastern equatorial Pacific. Teleconnections from this coupled ocean-atmosphere phenomenon result in anomalous weather patterns around the world. Because of the widespread effects of the ENSO and because it can be accurately forecast, it offers an attractive starting point for building predictive models for disease, such as cholera. Though not necessarily directly related, the frequencies of both El Nino events and cholera outbreaks have been increasing since the 1970s . Cholera dynamics in areas of South America (Peru) and the Bay of Bengal seem to be particularly affected by ENSO-driven anomalies. It has been hypothesized that an El Nino event in 1991 may have been the trigger that resulted in the resurgence of cholera in Peru . Warm waters along the coast, coupled with plankton blooms driven by El Nino rains, may have helped to amplify the population of V. cholerae already in the environment . Warm temperatures in Peru during the 1997-1998 event were also associated with increased numbers of diarrheal cases, including cholera. In Bangladesh, cholera outbreaks are related to the end of the monsoon season on an annual basis and to ENSO on an interannual frequency. Pascual et al. used time series analyses to examine the historical association between cholera in Bangladesh and ENSO (1980 to1998). Both previous disease levels in the community and ENSO were significant predictors in the nonlinear models, and interannual climate variability associated with ENSO clearly plays a role in the year-to-year variability of cholera outbreaks . Because of the recognized connection between ENSO and weather around the Indian Ocean , it is tempting to use ENSO to predict variability in the monsoon season and thus variability in annual cholera outbreaks. However, the ENSO-monsoon connection appears to be bidirectional and therefore cannot be used in a predictive capacity . FIG. 1. | Hierarchical model for environmental cholera transmission (modified from Colwell and Huq ). Hierarchical model for environmental cholera transmission (modified from Colwell and Huq ). TABLE 1 | Influence of environment, climate, and weather on cholera and V. cholerae dynamics CONCLUSIONS : V. cholerae has long been known as a fecal-oral pathogen, and indeed, infection rates are significantly greater in areas with poor sanitation; however, the evidence showing that V. cholerae is naturally present in warm, brackish environments is overwhelming. Evidence collected over the last 20 years indicates a close association of V. cholerae with copepods, such that the bacteria are concentrated on the exoskeleton and gut of the zooplankters relative to the surrounding water. When communities rely on untreated environmental water sources for bathing, cooking, and drinking water, the incidental ingestion of copepods, which carry a high "dose" of V. cholerae, can initiate an infection. Likewise, the likelihood of consuming an infectious dose is higher when a bloom of copepods occurs in the water. Based on this association, a simple and effective measure to filter plankton from water using sari material has been developed, and preliminary evidence shows a significant reduction in cholera cases in those households using this filtration procedure (R. R. Colwell, A. Huq, S. Islam, K. M. Aziz, M. Yunus, N. H. Khan, and E. Russek-Cohen, unpublished data, 2002). Because of the aquatic and oceanic connection of V. cholerae and commensal copepods and the moderating effect of the aquatic environment, outbreaks theoretically follow more predictable patterns than most vector-borne diseases, because local microenvironments largely control the vector population. Drivers such as sea surface temperature, sea surface height, and plankton blooms can be remotely sensed and used to forecast outbreaks . While issues of sanitation and human behavioral adaptation cannot be ignored, we propose that the hierarchical model presented here and elsewhere can serve as a template for future research on similar climate-sensitive diseases to define critical parameters and develop methods for prediction. Backmatter: PMID- 12364377 TI - Current Consensus Guidelines for Treatment of Neurocysticercosis AB - Taenia solium neurocysticercosis is a common cause of epileptic seizures and other neurological morbidity in most developing countries. It is also an increasingly common diagnosis in industrialized countries because of immigration from areas where it is endemic. Its clinical manifestations are highly variable and depend on the number, stage, and size of the lesions and the host's immune response. In part due to this variability, major discrepancies exist in the treatment of neurocysticercosis. A panel of experts in taeniasis/cysticercosis discussed the evidence on treatment of neurocysticercosis for each clinical presentation, and we present the panel's consensus and areas of disagreement. Overall, four general recommendations were made: (i) individualize therapeutic decisions, including whether to use antiparasitic drugs, based on the number, location, and viability of the parasites within the nervous system; (ii) actively manage growing cysticerci either with antiparasitic drugs or surgical excision; (iii) prioritize the management of intracranial hypertension secondary to neurocysticercosis before considering any other form of therapy; and (iv) manage seizures as done for seizures due to other causes of secondary seizures (remote symptomatic seizures) because they are due to an organic focus that has been present for a long time. Keywords: INTRODUCTION : Cysticercosis, the infection caused by the larval stage of the tapeworm Taenia solium, is the most common parasitic disease of the nervous system in humans and the single most common cause of acquired epileptic seizures in the developing world, where prevalence rates of active epilepsy are twice those in developed countries . Before the introduction of modern neuroimaging diagnostic techniques, knowledge of the natural history of human disease was limited and largely based on cases diagnosed either by the presence of subcutaneous nodules, by plain X-rays showing calcifications in the brain or soft tissues, by surgery of cases with intracranial hypertension, or from necropsy data (-, , ). The image of an aggressive, lethal disease arose from this clearly biased (towards more severe infestations) group of cases. During the last two decades, the introduction of computed tomography (CT) and later magnetic resonance imaging (MRI) permitted the identification of mild cases with only a few parenchymal cysts, and the terms benign and malign cysticercosis were coined . Later, studies in India showed that a vast majority of single enhancing lesions, until then attributed to tuberculosis, were in fact degenerating cysticerci . The introduction of praziquantel and albendazole as specific antiparasitic agents was enthusiastically adopted by many segments of the medical community. The value of these agents has been questioned by some authors (, , -, ), and an intense controversy still exists. Unfortunately, this has led to confusion and poorly informed decisions in clinical management, especially in areas where neurocysticercosis is not a daily diagnosis. At a recent meeting on cysticercosis held in Lima, Peru, a panel of experts in different aspects of the disease reached a consensus as to the minimal treatment guidelines for neurocysticercosis. Effort was made to identify about which types of neurocysticercosis there was agreement and a uniformly accepted approach. The panel also attempted to define the range of therapeutic approaches for those forms for which there was no consensus The Parasite | Taenia solium is a two-host zoonotic cestode. The adult stage is a 2- to 4-m-long tapeworm that lives in the small intestine of humans. No other final hosts are known for T. solium tapeworms in nature. As in all cestodes, the gravid proglottids at the terminal end of the worm are full of eggs that are the source of infection with the larval stage, or cysticercosis. The natural intermediate host is the pig, harboring larval cysts anywhere in its body. Humans become infected with cysts by accidental ingestion of T. solium infective eggs by fecal-oral contamination . The Disease | After ingestion of Taenia eggs containing infective oncospheres, the parasites become established in the tissues as larval cysts and reach their mature size in about 3 months . The parasite may locate almost anywhere in the body. The infection burden varies from a single lesion to several hundreds, and lesions may range in size from a few millimeters to several centimeters . Laboratory studies and information from other cestodes suggest that viable cysts actively modulate the host's immune system to evade destruction by it . Symptomatic disease results almost exclusively from the invasion of the nervous system (neurocysticercosis) and the eye and is clearly different in parenchymal neurocysticercosis and extraparenchymal neurocysticercosis. The usual presentation of parenchymal neurocysticercosis is with seizures, which can be controlled with antiepileptic drug therapy. Occasionally, the cysts may grow and produce a mass effect. Extraparenchymal infection may cause hydrocephalus by mechanical obstruction of the ventricles or the basal cisterns, either by the cysts themselves or by an inflammatory reaction (ependymitis and/or arachnoiditis). The so-called racemose variety occurs in the ventricles or basal cisterns and is characterized by abnormal growth of cystic membranes with degeneration of the parasite's head (scolex) . These cases follow a progressive course, and even after ventricular shunting, the membranes or inflammatory cells and proteins frequently block the shunt. In most patients, neurocysticercosis seems to produce symptoms years after the initial invasion of the nervous system by the parasite , by either inflammation around the parasite, mass effect, or residual scarring . There is a clear association between inflammation around one or more cysts and development of symptoms, especially with regard to seizures . The natural history of parenchymal cysticercosis has been studied by pathological examination and imaging studies using CT and MRI . Viable cysts are 10 to 20 mm in diameter, thin-walled sacks filled with clear cyst fluid. On imaging studies, the wall is not visible and the fluid is isodense with the cerebrospinal fluid. There is little or no evidence of perilesional inflammation, and they do not enhance with contrast media on neuroimaging . As the parasite loses the ability to control the host immune response, an inflammatory process begins. Initially, the cysts show slight pericystic contrast enhancement . Later they become markedly inflamed and edematous and appear as ring-like or nodular areas of enhancement after the injection of contrast . This phase has been called "granulomatous cysticercosis, "cysticerci in encephalitic phase," or "enhancing lesions." Finally, the cyst is processed by the cellular response, and its remnants either are not detectable by imaging or become calcified lesions . "Giant" cysts, measuring more than 50 mm in diameter, are occasionally found, located primarily in the Sylvian fissure . Cysticercotic encephalitis is a rare form of the disease in which patients have numerous inflamed cysticerci, leading to diffuse, severe cerebral edema . Extraparenchymal neurocysticercosis includes cysticerci in the ventricles and basal cisterns (racemose cysticercosis, Fig. ). Since the cyst membrane is thin and the fluid is isodense with the cerebrospinal fluid, uninflamed extraparenchymal cysticerci are usually not visible on CT and may only reveal subtle, indirect findings on MRI. Scans may reveal hydrocephalus without noticeable cysts, ependymitis , distorted basal cisterns, or basal meningitis. Diagnosis | The diagnosis of neurocysticercosis is difficult because clinical manifestations are nonspecific, most neuroimaging findings are not pathognomonic, and some serologic tests have low sensitivity and specificity. A set of diagnostic criteria was proposed in 1996 and recently revisited , based on objective clinical, imaging, immunological, and epidemiological data; these criteria consist of four categories that are stratified according to their diagnostic strength . These criteria provide two degrees of diagnostic certainty: definitive diagnosis, in patients who have one absolute criterion or in those who have two major plus one minor and one epidemiologic criteria; and probable diagnosis, in patients who have one major plus two minor criteria, in those who have one major plus one minor and one epidemiologic criteria, and in those who have three minor plus one epidemiologic criteria. This chart of diagnostic criteria for neurocysticercosis has not yet been tested in hospital-based studies. Differential diagnosis between cysticercosis and other parasitic diseases may be difficult on clinical grounds. However, epidemiological data as well as evidence provided by neuroimaging studies and highly specific immune diagnostic tests usually provide useful diagnostic clues. Cystic hydatid disease almost always appears on CT/MRI as a single, large, spherical, and nonenhancing intracranial cyst. This is a very rare form of presentation of T. solium cysticercosis. Also, the current assay of choice, immunoblot, does not cross-react with echinococcosis. Other condition that may resemble T. solium cysticercosis from the clinical and neuroimaging points of view is coenurosis, an extremely rare condition caused by the cestode Multiceps multiceps. Therapeutic Alternatives | Therapeutic measures include antiparasitic drugs, surgery, and symptomatic medication. Antiparasitic drugs. | As demonstrated by experiments in animals, praziquantel and albendazole are effective antiparasitic drugs against T. solium cysticerci. Initial studies with praziquantel noted that doses as low as 5 to 10 mg/kg/day had some effect against cysts, and doses as high as 50 to 75 mg/kg/day were well tolerated. A dosage of 50 mg/kg/day for 2 weeks was adopted by most subsequent studies (, , , , , , , , , , -, , , ), although a single-day regimen of praziquantel has recently been described , with similar rates of cyst disappearance in some groups of patients. No dose ranging studies were performed with albendazole in cysticercosis. Instead, the dose previously used in hydatid disease (15 mg/kg/day) was used for cysticercosis. The initial length of therapy was 1 month, later reduced to 15 days and 1 week . There is limited experience with higher doses of both drugs . Between 60% and 85% of parenchymal brain cysticerci are killed after standard courses of treatment, with most trials showing a higher parasiticidal effect of albendazole. Between the second and fifth days of antiparasitic therapy, there is usually an exacerbation of neurological symptoms, attributed to local inflammation due to the death of the larvae. For this reason, both albendazole and praziquantel are generally given simultaneously with steroids in order to control the edema and intracranial hypertension that may occur as a result of therapy. Praziquantel interacts with steroids, decreasing its serum concentrations , although there is no evidence that this pharmacological interaction affects its parasiticidal properties. Serum levels of phenytoin and carbamazepine may also be lowered as the result of simultaneous praziquantel administration . Albendazole has better penetration into cerebrospinal fluid, its concentrations are not affected when given with steroids , and it is cheaper than praziquantel. Table summarizes most trials of antiparasitic therapy for neurocysticercosis. After the initial descriptions of successful use of praziquantel and albendazole in neurocysticercosis, several cases series noted that some types of parenchymal neurocysticercosis can resolve on imaging studies without being treated with antiparasitic drugs . Since then, an alternative opinion has been voiced that the acute, severe brain inflammation resulting from their use is unnecessary because parenchymal brain cysticercosis follows a benign course and cysts will degenerate and heal by natural evolution of the disease (, , -). The arguments on both sides of the discussion are contrasted in Table . Most published studies on treatment of neurocysticercosis are uncontrolled case series, often purely imaging studies. The main evidence for clinical improvement after antiparasitic treatment of parenchymal neurocysticercosis is based on two independent retrospective studies published in 1992. In one of them, all untreated patients (49 of 49) had seizures in the follow-up, compared to 46% (54 of 118) of albendazole-treated patients (mean follow-up, 3 years) . In the other, 74% (20 of 27) of untreated patients had seizures in the follow-up, compared to 17% (16 of 95) in albendazole-treated patients (mean follow-up, 2.5 years) . In these studies, the rate of seizure recurrence in those refusing treatment was much higher than noted in patients treated with antiepileptic drugs alone . The increased relapse rates in the comparison group could be explained by the fact that patients who did not receive albendazole in these centers were obviously biased towards poor compliance and would have only returned to the hospital in the event of later seizures . To date, there is only one published prospective trial evaluating the clinical evolution of patients with viable neurocysticercosis treated with antiparasitic agents compared to no treatment. Carpio and colleagues studied 138 patients with cystic neurocysticercosis in an open-label study of steroids alone or together with either albendazole or praziquantel . Overall, there were no significant differences in the proportion of patients free of cysts at 6 months or 1 year, in the proportion free of seizures for 2 years, or in the rates of sequelae. There was also a high rate of resolution among patients with single lesions, raising the question of whether patients with inflamed lesions were rigorously excluded from that trial. Disappearance of lesions in the control group only occurred in cases with a single lesion, whereas nearly half of the cases that were free of parasites in the treatment arms had multiple cysts. Thus, this study, while not demonstrating significant differences, suggests benefit for those with multiple lesions. Among many case series (, , , , -, ), only three randomized, controlled trials of albendazole have been reported on patients with single enhancing lesions. Two studies demonstrated no significant difference in radiographic resolution . The other demonstrated more rapid resolution with albendazole but no significant difference in the frequency of clinical events . Surgery. | Prior to the advent of antiparasitic drugs, surgery was the primary therapy for neurocysticercosis, mainly open surgery for excision of large cysts or cysts in the ventricles. The role of surgical therapy in the management of neurocysticercosis has significantly decreased over time and is now mainly restricted to placement of ventricular shunts for hydrocephalus secondary to neurocysticercosis. The main problem in these cases is the high prevalence of shunt dysfunction; indeed, it is common for patients with hydrocephalus secondary to neurocysticercosis to have two or three shunt revisions . The protracted clinical course of these patients and their high mortality rates (up to 50% in two years) were directly related to the number of surgical interventions to change the shunt . According to one report, maintenance steroid therapy may decrease the frequency of shunt blockages . Many authors advocate shunting combined with antiparasitic drugs to further reduce the incidence of shunt failure . Recently, less invasive procedures have been described, specifically the use of neuroendoscopic resection for ventricular cysts . Overall results have been excellent, with much less morbidity than with open surgery . Symptomatic and anti-inflammatory medication. | Corticosteroids are frequently used in patients with neurocysticercosis. The most frequent regimen is dexamethasone at doses of between 4.5 and 12 mg/day. Prednisone at 1 mg/kg/day may replace dexamethasone when long-term steroid therapy is required. Corticosteroids are frequently used to decrease neurological symptoms due to the death of the parasite and are the primary management for chronic cysticercosis arachnoiditis or encephalitis, where up to 32 mg of dexamethasone per day is needed to reduce the brain edema accompanying this condition . Mannitol, at doses of 2 g/kg/day, is also used for acute intracranial hypertension secondary to neurocysticercosis. Other medications commonly used to treat symptoms in neurocysticercosis patients are antiepileptic drugs and analgesics. Seizures secondary to neurocysticercosis usually respond well to first-line antiepileptic. Withdrawal of antiepileptic drugs can be achieved, although residual calcifications on CT scan mark patients for whom the risk of recurrent seizures is high. To use only a short period of antiepileptic therapy in patients with a single degenerating cysticercus, assuming that they have acute symptomatic crisis, has been proposed , but there are no data to support this assumption. FIG. 1. | Life cycle of T. solium Life cycle of T. solium. FIG.2-9. | Viable cysts (contrasted MRI; Fig. Viable cysts (contrasted MRI; Fig. 2), cyst with perilesional contrast enhancement (contrasted MRI; Fig. 3), enhancing lesion (contrasted MRI; Fig. 4), calcifications (noncontrasted CT; Fig. 5), giant cyst (contrasted CT; Fig. 6), cysticercotic encephalitis (contrasted MRI; Fig. 7), basal subarachnoid cysticercosis (contrasted MRI; Fig. 8), and IV ventricle ependymitis (contrasted MRI; Fig. 9). TABLE 1 | Diagnostic criteria for neurocysticercosis TABLE 2 | Studies on anti-parasitic therapy for neurocysticercosis TABLE 3 | Arguments in favor of and against antiparasitic treatment for neurocysticercosis DISCUSSION GUIDELINES AND METHODS : When the panel began to discuss the preferred treatment for neurocysticercosis, it immediately became clear that neurocysticercosis is not a single disease for which one therapy can be recommended. There are marked differences in clinical presentation, pathogenesis, natural history, and treatment options for the different forms. In the following sections, the consensus and disagreements of the panel are summarized. There was general agreement that the number of parasites is an important factor for determining treatment decisions. The panel chose to categorize patients as those having a single lesion, a few lesions, moderate to heavy parasite loads, and massive infections. For the sake of clarity, a threshold of five or fewer parasites was chosen to represent cases with a few lesions, and a threshold of 100 parasites was chosen to represent massive infections. Since the management recommendations were the same for patients with a single lesion and those with a few parasites, these two categories were later collapsed into one. The panel categorized all available information according to level of quality of evidence as follows: I, evidence obtained from at least one properly randomized controlled trial; II-1, evidence obtained from well-designed controlled trials without randomization; II-2, evidence obtained from well-designed cohort or case-control analytic studies, preferably from more than one center or research group; II-3. evidence obtained from multiple time series with or without the intervention, including dramatic results in uncontrolled experiments; and III, opinions of respected authorities, based on clinical experience; descriptive studies and case reports; or reports of expert committees. The information provided below represents the recommended management for patients with the pure forms of the disease. Whenever there was no agreement, the proposed alternatives are described as positions a, b, and c in order of preference, according to the number of experts in the group who favored each option. For patients with mixed presentations of neurocysticercosis, a consideration of risks and benefits must be done to choose the order of interventions. PANEL CONSENSUS ---GENERAL CONCEPTS : (i) Guidelines for treatment of neurocysticercosis must be individualized in terms of number and location of lesions, as well as based on the viability of the parasites within the nervous system (ii) Growth of a parenchymal cysticercus is not a common event and may be life-threatening. A growing parasite deserves active management, either with antiparasitic drugs or by surgical excision. (iii) In patients with intracranial hypertension secondary to neurocysticercosis, the priority is to manage the hypertension problem before considering any other form of therapy. Antiparasitic drug treatment is never the main priority in the setting of elevated intracranial pressure (iv) Antiepileptic drugs are the principal therapy for seizures in neurocysticercosis. In general, seizures should be managed in a similar manner to other causes of secondary seizures (remote symptomatic seizures), since they are due to an organic focus that has been present for a long time. However, after resolution of the parasitic infection with normalization of imaging studies, most patients who are seizure-free can eventually discontinue antiepileptic drugs. Antiparasitic drugs should not be regarded as an alternative for antiepileptic drug therapy. OVERVIEW OF PARENCHYMAL BRAIN CYSTICERCOSIS : Viable Cysts | There was agreement that albendazole and praziquantel are effective antiparasitic agents, destroying most viable cysts. Whereas there was some disagreement with regard to the best management of patients with few cysts, in patients with more than a few parasites, a balance of risks and benefits led the group to reach a consensus in that the risk of cyst growth, ventricular invasion, or multiple episodes of cyst degeneration with the corresponding symptomatic periods outweighs the potential (but theoretical) benefits of milder inflammation and minor scarring by natural evolution. The majority felt that there is no contraindication for the simultaneous use of steroids, but one expert raised the possibility of interfering with the clearance of parasite remnants. Consensus also broke down on whether to use antiparasitic drugs in patients with massive (hundreds of viable cysts) infections, mainly due to a perception of high risk for severe side effects. Enhancing Lesions | The panel felt that patients with a single enhancing lesion are likely to do well with antiepileptic drugs independently of whether antiparasitic therapy is added. While most panel members do not routinely use antiparasitic drugs for single enhancing cysticerci, some felt that faster radiological resolution merits its routine use, and others felt that antiparasitic drugs may be used in selected cases, i.e., those in which antiepileptic therapy cannot be adequately monitored and the risks of adverse outcomes from seizures are high. Conversely, for patients with massive infections (cysticercosis encephalitis), there was agreement that they should not be used because it may exacerbate the inflammatory reaction in the brain parenchyma. There was, however, no consensus on whether to use antiparasitic drugs in these patients after resolution of cerebral edema. Calcifications | For patients with only calcified lesions, there was also consensus that there was no role for antiparasitic agents because the cysts are already dead. The recent description of contrast enhancement and edema around calcified brain cysts led to discussion of whether there will be a role for anti-inflammatory medication in such patients, but no controlled data exist yet . TABLE 4 | Guidelines for use of antiparasitic treatment in neurocysticercosis OVERVIEW OF EXTRAPARENCHYMAL CYSTICERCOSIS : Whenever hydrocephalus or intracranial hypertension is present, its management should be the first priority. It can generally be managed by means of a ventricular shunt. Since extraparenchymal neurocysticercosis is associated with a worse prognosis, there was a consensus towards aggressive management. The panel felt that patients with cysts located in the ventricular system should be treated surgically, especially if proper technology for neuroendoscopical resection is available. The location of the cysts and the presence of ependymitis need to be assessed prior to planning the surgical approach, and the possibility of acute blockage of the cerebrospinal fluid flow during antiparasitic treatment must be borne in mind. As for growing parenchymal cysts, a warning was issued for basal subarachnoid cysticercosis. As far as is known, this type of neurocysticercosis is progressive and has a grim prognosis. There was consensus that cysticercosis of the basal cisterns should be treated with antiparasitic drugs. This is based on the limitations of surgical resection, the poor prognosis with diversion procedures alone, and case series that used shunting plus antiparasitic drugs . The optimal duration of antiparasitic treatment for this type of lesion is not known, but it was felt that therapy should be sustained for longer than is routine for parenchymal disease . There was consensus that all patients with subarachnoid cysticercosis should be managed with corticosteroids as an adjunct to antiparasitic therapy. Large (giant) cysts in the Sylvian fissure seem to respond well to corticosteroids along with antiparasitic drugs . COMMENTS : Cysticercosis should no longer be considered an "exotic" disease. Besides being the leading cause of late-onset seizures in the developing world, it is now also frequently diagnosed in industrialized countries because of increased immigration from areas where it is endemic, with over 1,000 cases diagnosed annually in the United States alone . Probably the best advice on the treatment of neurocysticercosis would be not to generalize but to approach and assess each case individually. However, neurologists and infectious disease specialists who are not familiar with the disease need at least a basic set of principles to follow. In developing such principles, the panel have tried to avoid the trap of futile discussions and to expose, in the clearest way, the current preferred treatment options for each type of neurocysticercosis. The selection of a treatment option must include in the consideration of risks and benefits the economic situation of the patient. In areas where neurocysticercosis is endemic, where most cases are seen, follow-up neuroimaging examinations may not be performed for economic reasons, and patients may not have access to modern surgical techniques or may be managed in centers where intensive care is not available. The standard of care will change, but still the best option has to be defined. The current controversy on whether the use of antiparasitic agents is of benefit in the long-term control of epileptic seizures in neurocysticercosis led to much confusion about whether such agents should be used in any form of the disease. This undesirable situation causes inappropriate use of antiparasitic drugs, i.e., in cases with already calcified parasites or in cysticercosis encephalitis. Even more dangerous, physicians may refrain from using antiparasitic treatment when it is the best treatment option, i.e., for a growing cyst. This "conservative" therapeutic approach may even lead to the death of the patient if the period of inaction extends long enough. A further conclusion of the panel was the inadequacy of study designs and outcomes in the treatment of neurocysticercosis. This does not reflect so much poorly conducted trials (sometimes the case) as lack of agreement that would permit definition of specific research questions, to homogenize treatment groups in similar trials, or to perform multicentric trials to include larger numbers of patients with similar forms of neurocysticercosis. The inclusion of long-term outcomes, including the proportion of residual calcifications and the appearance of hydrocephalus, is mandatory to provide a fair evaluation of risk-benefit analysis. Backmatter: PMID- 12364371 TI - History of Human Parasitology AB - Humans are hosts to nearly 300 species of parasitic worms and over 70 species of protozoa, some derived from our primate ancestors and some acquired from the animals we have domesticated or come in contact with during our relatively short history on Earth. Our knowledge of parasitic infections extends into antiquity, and descriptions of parasites and parasitic infections are found in the earliest writings and have been confirmed by the finding of parasites in archaeological material. The systematic study of parasites began with the rejection of the theory of spontaneous generation and the promulgation of the germ theory. Thereafter, the history of human parasitology proceeded along two lines, the discovery of a parasite and its subsequent association with disease and the recognition of a disease and the subsequent discovery that it was caused by a parasite. This review is concerned with the major helminth and protozoan infections of humans: ascariasis, trichinosis, strongyloidiasis, dracunculiasis, lymphatic filariasis, loasis, onchocerciasis, schistosomiasis, cestodiasis, paragonimiasis, clonorchiasis, opisthorchiasis, amoebiasis, giardiasis, African trypanosomiasis, South American trypanosomiasis, leishmaniasis, malaria, toxoplasmosis, cryptosporidiosis, cyclosporiasis, and microsporidiosis. Keywords: INTRODUCTION : During our relatively short history on Earth, humans have acquired an amazing number of parasites, about 300 species of helminth worms and over 70 species of protozoa . Many of these are rare and accidental parasites, but we still harbor about 90 relatively common species, of which a small proportion cause some of the most important diseases in the world, inevitably, these are the ones that have received the most attention. Since most of these parasitic diseases occur mainly in the tropics, the field of parasitology has tended to overlap with that of tropical medicine, and thus the histories of these two fields are intertwined. There is, however, much more to the history of human parasitology than this, and our understanding of parasites and parasitic infections cannot be separated from our knowledge of the history of the human race. In particular, the spread and present distribution of many parasites throughout the world has largely been the result of human activities, and the advent of AIDS has added a new chapter to the history of parasitology. HUMAN EVOLUTION, MIGRATIONS, CIVILIZATION, AND PARASITIC INFECTIONS : Human evolution and parasitic infections have run hand in hand, and thanks to the spinoffs from the Human Genome Project, we now know much more about the origins of the human race than ever before . Sometime, about 150,000 years ago, Homo sapiens emerged in eastern Africa and spread throughout the world, possibly in several waves , until 15,000 years ago at the end of the Ice Age humans had migrated to and inhabited virtually the whole of the face of the Earth, bringing some parasites with them and collecting others on the way. For the purpose of this review, the parasites that infect humans can be classified as heirlooms or souvenirs. Heirlooms are the parasites inherited from our primate ancestors in Africa, and souvenirs are those that we have acquired from the animals with which we have come in contact during our evolution, migrations, and agricultural practices. The development of settlements and cities facilitated the transmission of infections between humans, and the opening up of trade routes resulted in the wider dissemination of parasitic infections. The slave trade, which flourished for three and a half centuries from about 1500, brought new parasites to the New World from the Old World ; in more recent times, the spread of human immunodeficiency virus HIV and AIDS and the immunodepression associated with these conditions has resulted in the establishment of a number of new opportunistic parasitic infections throughout the world . We are beginning to learn a lot about the past history of parasitic infections from studies of archaeological artifacts, such as the presence of helminth eggs or protozoan cysts in coprolites (fossilized or desiccated feces) and naturally or artificially preserved bodies; from such studies has emerged a new science, palaeoparasitology. Examples of some of these discoveries will be discussed later. So vast is the field of human parasitology, and so many and far-reaching the discoveries made, that it is not possible to do justice to the whole subject. Therefore; only the most significant aspects and the most important parasites are considered under two major headings, the helminth worms and the protozoa. EARLY WRITTEN RECORDS : The first written records of what are almost certainly parasitic infections come from a period of Egyptian medicine from 3000 to 400 BC, particularly the Ebers papyrus of 1500 BC discovered at Thebes . Later, there were many detailed descriptions of various diseases that might or might not be caused by parasites, specifically fevers, in the writings of Greek physicians between 800 to 300 BC, such as the collected works of Hippocrates, known as the Corpus Hippocratorum, and from physicians from other civilizations including China from 3000 to 300 BC, India from 2500 to 200 BC, Rome from 700 BC to 400 AD, and the Arab Empire in the latter part of the first millennium. As time passed, the descriptions of infections became more accurate and Arabic physicians, particularly Rhazes (AD 850 to 923) and Avicenna (AD 980 to 1037) , wrote important medical works that contain a great deal of information about diseases clearly caused by parasites. In Europe, the Dark and Middle Ages, characterized by religious and superstitious beliefs, held back medical progress until the Renaissance, which released a flurry of activity that eventually led to the great discoveries that characterized the end of the 19th century and the beginning of the 20th. These discoveries included the demolition of the theory of spontaneous generation and the evolution of the germ theory by Louis Pasteur, the demonstration by Pasteur that diseases could be caused by bacteria, the discovery of viruses by Pierre-Paul Emile Roux, the introduction by Robert Koch of methods of preventing diseases caused by microorganisms, and the incrimination by Patrick Manson of vectors in the transmission of parasites. The great personalities of this period made discoveries in a number of fields, and their findings and ideas fed off one another. The names of Pasteur, Koch, Roux, and Manson occur time and time again in the history of parasitology and microbiology. DISCOVERY OF THE HELMINTH WORMS : Because of the large size of some helminths, such as the roundworm Ascaris and the tapeworms, it is practically certain that our earliest ancestors must have been aware of these common worms. There is some evidence for this assumption based on contemporary studies of primitive tribes in Sarawak and North Borneo, where Hoeppli found that most people are aware of their intestinal roundworms and tapeworms . Some historians have identified references to helminth worms and their diseases in the Bible, but the relevant passages are open to several interpretations. Among the Egyptian medical papyri, the Ebers papyrus refers to intestinal worms, and these records can be confirmed by the discovery of calcified helminth eggs in mummies dating from 1200 BC. The Greeks, particularly Hippocrates (460 to 375 BC) , knew about worms from fishes, domesticated animals, and humans. Roman physicians including Celsus (25 BC to AD 50) and Galen (Galenus of Pergamon, AD 129 to 200) were familiar with the human roundworms Ascaris lumbricoides and Enterobius vermicularis and tapeworms belonging to the genus Taenia. Somewhat later, Paulus Aegineta (AD 625 to 690) clearly described Ascaris, Enterobius, and tapeworms and gave good clinical descriptions of the infections they caused . Following the decline of the Roman Empire, the study of medicine switched to Arabic physicians, including Avicenna, who recognized not only Ascaris, Enterobius, and tapeworms but also the guinea worm, Dracunculus medinensis, which had been recorded in parts of the Arab world, particularly around the Red Sea, for over 1,000 years. The medical literature of the Middle Ages is very limited, but there are many references to parasitic worms. In some cases, they were recognized as the possible causes of disease but in general, the writings of the period reflect the culture, beliefs, and ignorance of the time. The science of helminthology really took off in the 17th and 18th centuries following the reemergence of science and scholarship during the Renaissance period. Linnaeus described and named six helminth worms, Ascaris lumbricoides, Ascaris vermicularis (= Enterobius vermicularis), Gordius medinensis (= Dracunculus medinensis), Fasciola hepatica, Taenia solium, and Taenia lata (= Diphyllobothrium latum) . Thereafter, more species were described until at the beginning of the 20th century, 28 species had been recorded in humans, a number that has now grown to about 300 species, including accidental and very rare records . Even if some of these are doubtful, at least 280 species are recognized by Ashford and Crewe in their annotated checklist . Ascaris and Ascariasis | Ascaris lumbricoides, the large roundworm, is one of six worms listed and named by Linnaeus; its name has remained unchanged ever since. One billion people are now estimated to be infected with this worm. The adult worm lives in the intestine, and the female produces eggs that pass out with the feces, and the larvae within the eggs develop to the infective stage in soil. Humans become infected when food contaminated with infective eggs is eaten and the larvae emerge in the intestine. The worms do not mature immediately but migrate around the body, reaching the lungs, from which they are coughed up and swallowed and then develop into adults in the intestine. Ascariasis is an ancient infection, and A. lumbricoides eggs have been found in human coprolites from Peru dating from 2277 BC and Brazil from about 1660 to 1420 BC . In the Old World, there are records of A. lumbricoides in a Middle Kingdom Egyptian mummy dating from 1938 to 1600 BC and from China in the Ming Dynasty between AD 1368 and 1644 . The presence of this large worm, which reaches a length of 15 to 35 cm and is often voided in the feces or sometimes emerges from the anus, is very obvious. There are extensive written records including the Egyptian medical papyri, the works of Hippocrates in the fifth century BC, Chinese writings from the second and third centuries BC , and texts of Roman and Arabic physicians . Surprisingly, it was not until the late 17th century that the detailed anatomy of the worm was described, first by Edward Tyson, an English physician , and shortly afterward by the Italian Francesco Redi, who described the worms in his book Osservazioni Intorno Agli Animali Viventi che si Trovano Negli Animali Viventi, one of the first books on parasitology . These two publications, together with that of Tyson on the tapeworms of humans , can be considered to mark the beginnings of the subdiscipline of helminthology, which reached a peak in the 19th century. It was also during this period that the first real attempts were made to understand the infections caused by Ascaris and other worms and how they might be treated . In the meantime, the problem for those studying Ascaris and other parasitic nematodes was how the parasite's eggs infected a new host after leaving the original host. It was not until 1862 that transmission by ingesting eggs was demonstrated by the French medical scientist Casimir Joseph Davaine and later by the Italian scientist Giovanni Battista Grassi, who infected himself with the eggs of A. lumbricoides and subsequently found eggs in his feces . The life cycle in humans, including the migration of the larval stages around the body, was discovered only in 1922 by a Japanese pediatrician, Shimesu Koino, who infected both a volunteer and himself and realized what was happening when he found large numbers of larvae in his sputum . There are good accounts of the history of ascariasis by Grove and Goodwin . Hookworms and Hookworm Disease | Human hookworm infections are caused by two species, Ancylostoma duodenale and Necator americanus, the former originating in Asia and the latter originating in Africa. The life cycles of the two worms are similar. Adult male and female worms live in the small intestine, where they can cause massive blood loss. Eggs pass out with the feces to contaminate the soil, where larvae emerge and molt to become infectious larvae that bore through the skin of a new host. In humans the larvae migrate to the lungs and trachea, from which they are swallowed before maturing into adults in the small intestine. Human hookworm infections have been associated with humans in the Old World for over 5,000 years . The presence of hookworm infections in pre-Columbian America is a fiercely disputed topic. Robert Desowitz has little doubt that hookworms were present before the arrival of Europeans , but Kathleen Fuller suggests that hookworms were introduced into the Americas after 1492 . Palaeoparasitological evidence appears to back Desowitz's ideas since ova identified as Ancylostoma sp. have been found in a human coprolite dated from somewhere between 3350 BC and AD 480 . Larval nematodes, possibly hookworms, have been found in fecal samples dated to about 200 BC from the Colorado Plateau . The introduction of hookworms into the Americas is discussed in more detail elsewhere . The classical signs of hookworm disease are anemia, greenish yellow pallor, and lassitude. None of these symptoms is obvious or unambiguous, and the one distinctive feature exhibited by some individuals, geophagy, is not necessarily associated with disease. Although worms must have been present in many civilizations, most infections have gone unnoticed such that early accounts of the disease interpreted in retrospect must be treated with caution. The greenish pallor called Egyptian chlorosis, first associated with hookworm infections by 19th century scientists, is not recorded in the early Egyptian papyri. It has been suggested that the enigmatic condition aaa that occurs in many papyri including the Ebers papyrus might refer to hookworms , but there is no real evidence for this . This subject is discussed when considering schistosomiasis below. There are references to yellowish pallor and geophagy in the works of Hippocrates and Lucretius, who noted the pallor seen in miners in about 50 BC. There are also references from the third century BC in China to laziness and a yellow disease . During the 18th and 19th centuries, there were increasing numbers of records from the West Indies and South and Central America . Worms were found in a human in 1838 by the Italian physician Angelo Dubini , and the connection between the worms and disease was finally established by Wilhelm Griesinger in 1854 . Although the association between pallor and working in mines had been made by Lucretius, it was not until 1879 that the Italian veterinarian Edoardo Perroncito established the real connection while investigating the diseases of miners in the St. Gothard tunnel . Conditions in mines favor the development of larval hookworms that require warmth and damp. The fact that hookworm larvae enter the body by boring through the skin was not discovered until the end of the 19th century, when Arthur Looss accidentally infected himself . In the early part of the 20th century, hookworm disease was such a serious problem in the United States that the Rockefeller Foundation took on the task of controlling the disease, an activity that subsequently led to the establishment of a number of Schools of Public Health and the creation of the World Health Organization . There are good accounts of the history of hookworm disease by Ball , Foster , and Grove . Trichinella and Trichinosis | Trichinosis, also known as trichinellosis and trichina infection, is caused by the intestinal nematode worm Trichinella spiralis, which requires two hosts in its life cycle. The female worms produce larvae that encyst in muscle, and a new host becomes infected when muscle is eaten. Because human infections are usually acquired by eating pork infected with the encysted larvae, this might have given rise to the Mosaic and Islamic traditions of avoiding pork, a practice that has also been attributed to tapeworm infection (see below). The association between trichina infections and pigs has been long recognized, but the encysted larvae in the muscle were not seen until 1821 and even then were not associated with disease in humans . The discovery of the worm in humans in 1835 was made by James Paget, then a medical student at St. Bartholomew's Hospital in London and later knighted as a distinguished physician, but the definitive report was written by Richard Owen, who played down Paget's role and did not realize that the worm in human muscle was a larval stage of a nematode. The adult worms were discovered by Rudolf Virchow in 1859 and Friedrich Zenker in 1860, and it was Zenker who finally recognised the clinical significance of the infection and concluded that humans became infected by eating raw pork . The importance of these studies lies not only in the field of human parasitology but also in the more general field of parasitology concerned with the transmission of parasites between different animal species and the importance of predator-prey relationships in such transmission. There are good accounts of the history of trichinosis by Bundy and Michael , Foster , and Grove . Strongyloides and Strongyloidiasis | Humans are hosts to two species of Strongyloides, S. stercoralis and S. fuelleborni, of which there are two subspecies, S. f. fuelleborni in Africa and S. f. kellyi in Papua New Guinea. As far as human disease is concerned, S. stercoralis is the more common and important species. Its life cycle is more complex than that of any of the other nematodes discussed so far and involves both parasitic and free-living generations. Adult parthenogenic female worms in the small intestine lay eggs that hatch within the host to produce first-stage larvae, which are passed out in the feces and adopt a free living existence in the soil. Here they molt to produce infective larvae that penetrate the skin and are carried around the body to the lungs and are swallowed and reach the gut in the same way as hookworms. Sometimes the larvae mature to the infective stage in feces on the skin and reinfect the host through the skin (autoinfection), or the larvae may mature to the infective stage without leaving the gut and penetrate the gut wall. Thereafter, in both cases, the infection proceeds as described above. In immunosuppressed individuals, larval stages can be found throughout the viscera. S. stercoralis also has an alternative free-living life cycle in the soil. Given the absence of eggs and the small size of the larvae, combined with confusion with other free living species of nematodes, it is not surprising that S. stercoralis was not recognized until 1876, when the larvae and the disease strongyloidiasis were both discovered by Louis Alexis Normand, a physician to the French naval hospital at Toulon . Normond later found adult worms and, not knowing what they were, sent them to Professor Arthur Rene Jean Baptiste Bavay at the French Conseil Superieur de Sante, who realised that they were the adult worms of the larvae that were found in the feces . In 1883 the distinguished German parasitologist Karl Georg Friedrich Rudolf Leuckart discovered the alternation of generations involving parasitic and free-living phases . The discovery that infection occurred through the skin was made by a Belgian physician, Paul Van Durme, whose studies were based on the work of Looss, mentioned above, who had shown that A. duodenale infects its host in this way . It is now thought that Van Durme was probably working with A. fuelleborni , but the correct mode of infection had been established, and it was Looss who later succeeded in infecting himself by putting larvae of S. stercoralis on his skin and finding larvae in his feces 64 days later . Friedrich Fulleborn, working with dogs in Hamburg, described the phenomenon of autoinfection and discovered how S. stercoralis (and also Ancylostoma spp.) migrates around the body before ending up in the intestine . For over half a century, S. stercoralis received little attention until detailed studies on infections in prisoners of war who had acquired their infections in the Far East in the 1940s revealed disseminated infections in immunosuppressed patients . It was later found that Strongloides infections were more severe in patients infected with human T-lymphotropic virus type 1 and were at one time, but are no longer, regarded as major concomitants of AIDS . Strongyloides infections and strongyloidiasis are not well covered in the literature, but there is a good account by Grove . Dracunculus and Dracunculiasis (Guinea Worm Disease) | The best-documented parasitic disease known from the earliest times is undoubtedly that caused by the nematode worm Dracunculus medinensis. Adult worms live in subcutaneous connective tissue, from which the female worm emerges to release thousands of larvae into water, where they are eaten by intermediate hosts, cyclopodid crustaceans, in which they mature into infective larvae that infect humans when the crustaceans are accidentally swallowed with drinking water. The large female worm, up to 80 cm in length, protrudes from the skin, usually of the leg, and causes intense inflammation and irritation, signs that are so unusual and unambiguous that ancient texts can be interpreted with some certainty. The earliest descriptions are from the Ebers papyrus from 1500 BC and include instructions for treating aat swelling in the limbs; they appear to refer to both the nature of the infection and techniques for removing the worm. This interpretation is widely accepted by most parasitologists , but there are difficulties in interpreting this particular text since the word aat may simply mean a swelling . Nevertheless, confirmation of the presence of this worm in ancient Egypt comes from the finding of a well-preserved female worm and a calcified worm in Egyptian mummies . Dracunculiasis is one of the few diseases unambiguously described in the Bible, and most parasitologists accept that the "fiery serpents" that struck down the Israelites in the region of the Red Sea after the Exodus from Egypt somewhere about 1250 to 1200 BC were actually Guinea worms . The most authoritative interpretation of this biblical text, thought to have been written in the eighth century BC, is that by Gottlob Friedrich Heinrich Kuchenmeister, a parasitologist, theologian, and Hebrew scholar, in his 1855 textbook translated into English as Animal and Vegetable Parasites . Assyrian texts in the library of King Ashurbanipal from the 7th century BC also refer to conditions that are obviously dracunculiasis, and later descriptions of dracunculiasis occur in all the major Greek and Roman texts and works by the Arab physicians the 10th and 11th centuries . Because there is reference to "Medina vein" in the Arab literature, some historians have suggested that the Arab physicians may have thought that the worm was actually a rotten vein, but most informed observers now agree that the Arab physicians were fully aware of the worm-like nature of dracunculiasis but not necessarily the actual cause of the disease . Interest in dracunculiasis reemerged when the condition began to be recognized by European travelers visiting Africa (hence the common name, Guinea worm) and Asia. In 1674, Georgius Hieronymus Velschius initiated the scientific study of the worm and the disease it caused , and in 1819, Carl Asmund Rudolphi discovered adult female worms containing larvae , a discovery that was followed up in 1834 by a Dane known only as Jacobson . In 1836, D. Forbes, a British army officer serving in India, found and described the larvae of D. medinensis in water , and over the next few years several parasitologists, including George Busk , pursued the idea that humans became infected through the skin. It was not until 1870 that the whole life cycle, including the stages in the crustacean intermediate host, was elaborated by the Russian Alekej Pavlovitch Fedchenko . Fedchenko's observations gained wide acceptance after they were confirmed by Manson in 1894 , and the whole life cycle was finally elaborated in 1913 by the Indian bacteriologist Dyneshvar Atmaran Turkhud, who succeeded in infecting human volunteers with infected Cyclops . There are more detailed accounts of the history of Dracunculus by Foster , Grove , and Tayeh . Filarial Worms and Lymphatic Filariasis (Elephantiasis) | Lymphatic filariasis is caused by infection with the nematode worms Wuchereria bancrofti, Brugia malayi, and B. timori, which are transmitted by mosquitoes. The discovery of the life cycle by Patrick Manson in 1877 is regarded as one of the most significant discoveries in tropical medicine, but in the context of the history of parasitology it is better perceived as a logical extension of much that had gone before. Like Dracunculus, the adult filarial worms live in subcutaneous tissues, but unlike Dracunculus, the larvae, called microfilariae, produced by the female worm pass into the blood and are taken up by a blood-sucking mosquito when it feeds. After development in the mosquito, the microfilariae are injected into a new host when the mosquito feeds again. One particular form of the disease that must have attracted the attention of our ancestors is elephantiasis, which is characterized by grotesque swellings of the limbs, breasts, and genitals. These deformities appear to have been described and depicted in drawings from the earliest times, but the interpretation of the early records must be viewed with caution . Lymphatic filariasis was, and is, common along the Nile and, although there are no written records, the swollen limbs of a statue of the Egyptian Pharaoh Mentuhotep II from about 2000 BC suggest that he was suffering from elephantiasis, and small statuettes and gold weights from the Nok culture in West Africa from about AD 500 depict the enlarged scrota characteristic of elephantiasis . Greek and Roman writers were aware of the differential diagnosis of the condition and used the term "elephantiasis graecorum" to describe leprosy and the term "elephantiasis arabum" to describe lymphatic filariasis; the Arabic physicians, including Avicenna, were also aware of the differences between leprosy and lymphatic filariasis . The first definitive reports of lymphatic filariasis only began to appear in the 16th century. Lymphatic filariasis is also known as "the curse of St. Thomas" , and on a visit to Goa between 1588 and 1592, the Dutch explorer Jan Huygen Linschoten recorded that the descendants of those that killed St. Thomas were "all born with one of their legs and one foot from the knee downwards as thick as an elephants leg" . Thereafter, there are numerous references to elephantiasis, especially in Africa but also in Asia, including China, where Manson was later to discover the life cycle of the parasite. Another pathological condition associated with lymphatic filariasis is chyluria, in which the urine appears milky. This condition was recorded by William Prout in his 1849 book On the Nature and Treatment of Stomach and Renal Diseases . The larval microfilariae were first seen in hydroceel fluid by the French surgeon Jean-Nicolas Demarquay in 1863 and, independently, in urine by Otto Henry Wucherer in Brazil in 1866 . It remained for Timothy Lewis, a Scottish physician working in Calcutta, to confirm the finding of microfilariae in urine and blood and to recognize their significance in elephantiasis . The adult worm was described by Joseph Bancroft in 1876 and named Filaria bancrofti in his honour by the British helminthologist Thomas Spencer Cobbold . The elucidation of the life cycle, one of the triumphs of parasitological research, was the work of Patrick Manson in 1877 . This is widely regarded as the most significant discovery in tropical medicine, with implications that went far beyond helminthology into such diverse areas as malaria and the arboviruses. The story of Manson's discoveries has been told many times , but what is often omitted from the history of Manson's discoveries is the fact that he was aware of Fedchenko's earlier studies on the life cycle of D. medinensis and its transmission using an intermediate cyclopodid host (see above). Fedchenko's observations stimulated Manson to seek an intermediate host but also led him astray when he tried to demonstrate that infection was caused by drinking contaminated water. Manson, then working in Amoy in China, found microfilariae in the blood of dogs and humans and hypothesized that these parasites in the blood might be transmitted by blood-sucking insects. Accordingly, he fed mosquitoes on the blood of his gardener, who was harboring the parasites, and found larval stages in the mosquitoes . However, Manson thought that the parasite escaped from the mosquito into water and that humans acquired infection from this contaminated water by drinking the parasi water or via penetration of the skin. The actual mode of transmission was not established until suggestions made by the Australian parasitologist Thomas Bancroft were followed up by Manson's assistant George Carmichael Low, who demonstrated the presence of microfilariae in the mouthparts of mosquitoes in 1900 . The history of lymphatic filariasis is well described in the works already cited in this section . Loa and Loiasis (Eye Worm) and Onchocerca and Onchocerciasis (River Blindness) | Both loiasis, caused by infection with Loa loa, and onchocerciasis, caused by infection with Onchocerca volvulus, are filarial worms with life cycles similar to those described above. It is logical to consider these two conditions together because both affect the eyes and must have attracted the attention of early observers interested in sight and blindness. Surprisingly, there are no reliable early records. In loiasis the adult worm moves across the eye under the conjunctiva, an alarming experience that must have attracted attention of both sufferers and observers. An engraving by J. and T. de Bry made in 1598 was at one time thought to depict the extraction of a worm from the eye, but this has been hotly disputed, and it is now thought that this particular engraving represents a punishment for some offense rather than a treatment . The first definitive record is that of a French surgeon, Mongin, who, in 1770, described the worm passing across the eye of a woman in Santa Domingo, in the Caribbean, and recounts how he tried unsuccessfully to remove it . There are, however, less detailed earlier records of similar cases in 1768 and 1777 in an account of the history of French Guyane and Cayenne by Bertrand Bajon . In 1778, a French ship's surgeon, Francois Guyot, noticed that slaves in transit from West Africa to America suffered from recurrent ophthalmia and successfully removed a worm from one of them . The first English account of the removal of worms from the eye is that by William Loney in 1848; thereafter, there are increasing numbers of similar records . The microfilariae were discovered in 1890 by the ophthalmologist Stephen McKenzie and were sent for identification to Patrick Manson, who speculated that these might be the larvae of Loa loa . Loa infections are not confined to the eye, and there are also sometimes swellings on the arms and legs caused by the worm in its wanderings. These swellings, now known as Calabar swellings, were first recorded by a Scottish ophthalmic surgeon, Douglas Argyll-Robertson, in Old Calabar in Nigeria in 1895 , but it was not until 1910 that Manson suggested that they might be associated with infections by Loa loa , an opinion shared by his colleague George Low . The transmission by biting flies, Chrysops spp., was unraveled by the British helminthologist Robert Thompson Leiper in 1912 . There is an excellent account of Loa and loasis by Grove . Onchocerciasis, caused by the filarial worm Onchocerca volvulus, is found mainly in Africa and in parts of South America and the Arabian peninsula, where it was introduced from Africa, and it was only when these regions were opened up by explorers that the disease was recognized. The most important signs are blindness, an unexceptional condition that might have been due to a number of causes, and scaly, itchy, nodular skin, which was unusual and was known locally in West Africa as kru kru or craw craw. The microfilariae live in the skin and were discovered by the Irish naval surgeon John O'Neill when examining skin snips from patients suffering from craw craw in Ghana in 1874 . Some years later, in 1890, the adult worms were also discovered and identified by Patrick Manson . The role of the microfilariae in causing the skin lesions was established by Jean Montpellier and A. Lacroix in 1920 , and the part played by microfilaria in blindness was finally elaborated by Jean Hissette in the Belgian Congo (now the Democratic Republic of the Congo) in 1932 . O. volvulus is transmitted by sandflies, and their role in the transmission of onchocerciasis was discovered by the Scottish parasitologist Breadablane Blacklock in Sierra Leone in the mid-1920s . There are accounts of the history of onchocerciasis by Grove and Muller . Schistosomes and Schistosomiasis | Schistosomiasis, also known as bilharzia, is caused by infection with trematode worms belonging to the genus Schistosoma, of which the most important are S. haematobium, S. mansoni, and S. japonicum. The adult worms live in blood vessels associated with the intestine or bladder, and the females produce eggs that are passed out with the feces or urine. Larval stages, miracidia, emerge from the eggs when they reach water and bore into the intermediate host, a snail. After a period of multiplication in the snail, the next larval stages, the cercariae, emerge, and these are the stages that infect humans. The cercariae bore through the skin and transform into schistosomula that migrate through the body until they reach their final position in blood, vessels where they mature. The pathological effects of the disease are due mainly to immunological reactions to eggs that, instead of passing to the outside world, become deposited in different tissues; the effects depend on the tissues involved . In this context, it is interesting that schistosomiasis has been associated with carcinomas of the colon and bladder , one of the few examples of parasitic infections causing cancer (the others being the fluke infections opisthorchiasis and clonorchiasis [see below]). There is nothing special about the symptoms of schistosomiasis that might have attracted the attention of early observers except the bloody urine, hematuria, associated with S. haematobium infections, which is discussed below. There is no doubt that schistosomiasis is an ancient disease. In 1910, Marc Armand Ruffer found S. haematobium eggs in two Egyptian mummies dating from the 20th dynasty, 1250 to 1000 BC , a finding that is generally regarded as the beginning of the subdiscipline of palaeoparasitology. Thus, there is direct evidence that schistosomes were present in ancient Egypt, and there have been numerous attempts to find descriptions of this condition in the medical papyri . The most contentious word is aaa, which occurs in over 50 early papyri including the Ebers papyrus. In some medical papyri aaa occurs together with the initial hieroglyph suggesting a penis discharging what has been interpreted as blood . The juxtaposition is the papyri of aaa, antimony-based remedies, and possibly worms in the body suggests schistomiasis haematobia, and this interpretation is widely quoted in historical and parasitological textbooks. However, things are probably not as simple as this because no passages from the papyri link aaa with the bladder or urine and the discharge from the penis might represent semen and not blood. This subject is discussed in more detail by Nunn and Tapp , who abandon aaa as a possible ancient Egyptian word for schistosomiasis. However, since schistosomiasis was almost certainly common and widespread in ancient Egypt, it is curious that the Egyptians did not have a word for it unless it was so common that it was ignored. In this context, it should be mentioned that there have been a number of other suggestions about what aaa might be, including hookworm disease, which is discussed above. If we accept that there is no authoritative description of schistosomiasis in the earliest medical literature, the first definitive record must be that of an epidemic among soldiers in Napoleon's army in Egypt in 1798 by a French army surgeon, A. J. Renoult, who writes that "A most stubborn haematuria manifested itself amongst the soldiers of the French army... continual and very abundant sweats diminished quantity of urine...becoming thick and bloody" . Thereafter there are numerous reports of illnesses characterized by hematuria, particularly among armies including those involved in the Boer War (1899 to 1902). The worm S. haematobium was described by the German parasitologists Theodor Bilharz and Carl Theodor Ernst von Siebold in 1851 . Bilharz, with Wilhelm Griesinger, made the connection with the urinary disease a year later . Although it was known that other flukes employed a snail vector, the search for the intermediate stages in the life cycle of S. haematobium took a long time and a number of experienced parasitologists including Arthur Looss, Prospero Sonsino, and Thomas Cobbold, working at the end of the 19th century, all failed to infect snails ; it was not until 1915 that Robert Leiper demonstrated the complete life cycle in the snail host . Our knowledge of the history of intestinal schistosomiasis caused by S. mansoni dates back to conclusions reached by Manson in 1902 that there were two species of Schistosoma in humans . Even though there had been similar suggestions by other workers, Manson's ideas were not universally accepted, and it was Leiper who firmly established the existence of S. mansoni as a separate species in 1915 . The third important species is the Asian form, S. japonicum. One aspect of schistosomiasis japonica is Katayama disease, an ancient disease that was properly recorded in Japan in the Kwanami district only in 1847 by Dairo Fujii in a report that did not become available until 1909 . Fujii found people, cattle, and horses affected by wasting, abdominal swelling, and severe rashes on the legs, but he did not know the cause. By the time Fujii's paper had become available, another Japanese worker, Tokuho Majima, had found schistosome eggs in patients with Katayama disease , and he associated the pathological changes with the presence of the schistosome eggs. The worm itself, S. japonicum, was discovered and described by Fujiro Katsurada in 1904 , and its development in the snail host was described by Keinosuke Miyairi and M. Suzuki in 1913 , 2 years before Leiper independently described the life cycle of S. haematobium. Fuller accounts of the history of Katayama disease are given by Goodwin and Grove . The 20th century has been marked by the discovery of further species of schistosomes, S. intercalatum and S. mekongi. The history of such an important disease as schistosomiasis involves a great number of observations, events, and individuals; a detailed account of the history is given by Grove , and there are shorter accounts by Foster , Goodwin , and Hoeppli . A full bibliography is given by Warren , and an account of schistosomiasis in the context of British and American imperialism is given by Farley . Liver and Lung Fluke Diseases | Over 100 other species of flukes infect humans either as adults or as larvae, and only the most important ones are considered here. These are Paragonimus westermani, the lung fluke that causes paragonimiasis; Clonorchis sinensis, the liver fluke that causes clonorchiasis; and Opisthorchis spp., which cause opisthorchiasis. Virtually all the important discoveries about the parasites themselves were made during the period 1874 to 1918 as a result of observations on other parasitic flukes such as Fasciola hepatica in sheep and others of zoological rather than medical interest. The life cycles of these flukes are essentially similar to that described for Schistosoma spp. above, with the added complication that in some species, there is an additional intermediate host between the snail and the human in or on which the cercariae encyst. Humans become infected when they eat the infected second intermediate host. The various discoveries were made by a large number of people, often in obscure publications, and no attempt is made here to list the individual achievements; for this, the reader is referred to Grove and Muller . Our knowledge of the pathologic effects of clonorchiasis and opisthorchiasis has emerged gradually , with few historically interesting discoveries except the relatively recent finding of an association with the bile duct cancer cholangiocarcinoma . The history of these infections as diseases begins with the discovery of the worms and continues with the elaboration of the life cycles. P. westermani was discovered in the lungs of a human by Ringer in 1879 , and eggs in the sputum were recognized independently by Manson and Erwin von Baelz in 1880 . Manson also suggested that a snail might act as an intermediate host, and a number of Japanese workers, including Koan Nakagawa, Sadamu Yokogawa, Harujiro Kobayashi, and Keinosuke Miyairi, reported on the whole life cycle in the snail Semisulcospira between 1916 and 1922 . The human liver fluke, C. sinensis, was first recognized by James McConnell in 1875 , and the snail host was recognized by Masatomo Muto in 1918 , but it was the discovery in 1915 by Kobayashi of a second intermediate host, an important food fish from which human infections are acquired, that had the greatest impact on our knowledge and control of this infection . The first records of Opisthorchis infections in humans were made by Konstantin Wingradoff in 1892 , and the snail and fish hosts and their roles in the life cycle were discovered by Hans Vogel in 1934 . Cestodiasis (Tapeworm Infections) | Humans can be infected by about 40 species of adult tapeworms and about 15 larval forms, mainly as accidental hosts . The most important cestodes belong to two groups, the taeniid and diphyllobothriid tapeworms. The characteristic taeniid adults, which can reach a length of several meters, live in the intestine attached by a scolex and shed mature proglottids ("segments") containing numerous eggs, which pass out into soil or water, where the eggs are released. When an intermediate host consumes the eggs, they hatch in the intestine, releasing larval stages, oncospheres, that burrow through the gut wall to reach various tissues of the host, where they develop into encysted cysticerci or bladderworms. The life cycle is completed when undercooked or raw meat is eaten and the cysticerci are released and attach to the gut wall of the final host and develop into adult tapeworms. The two species in humans, Taenia saginata, the beef tapeworm and the larger of the two, and T. solium, the pork tapeworm, use cattle and pigs as their respective intermediate hosts. The scientific study of the taeniid tapeworms of humans can be traced to the late 17th century and the observations of Edward Tyson on the tapeworms of humans, dogs, and other animals . Tyson was the first person to recognize the "head" (scolex) of a tapeworm, and his subsequent descriptions of the anatomy and physiology of the adult worms laid the foundations for our knowledge of the biology of the taeniid tapeworms of humans. Although by this time it had become clear that there were differences between the broad tapeworm (see below) and the other tapeworms that we now know to be taeniids, the distinctions between T. solium and T. saginata were not obvious. These worms continued to be confused long after the work of Tyson, and although Goeze (see below) in 1782 had suspected that there were two species , it was not until the middle of the 19th century that Kuchenmeister is credited with recognizing the differences between T. solium and T. saginata based on the morphology of the scolex . In 1784, the first indications that intermediate hosts were involved in the life cycles of taeniid tapeworms emerged from the detailed studies of the pork tapeworm by a German pastor, Johann August Ephraim Goeze, who observed that the scolices of the tapeworm in humans resembled cysts in the muscle of pigs . Some 70 years later, Kuchenmeister, in much-criticized experiments, fed pig meat containing the cysticerci of T. solium to criminals condemned to death and recovered adult tapeworms from the intestine after they had been executed . Shortly afterward, in 1868 to 1869, J. H. Oliver observed that T. saginata tapeworm infections occurred in individuals who had eaten "measly" beef , and this was confirmed by the Italian veterinarian Edoardo Perroncito in 1877 . The adult stages of T. solium and T. saginata rarely cause any overt signs or symptoms, and there are no early descriptions of diseases that might be caused by these tapeworms. On the other hand, humans are host to two important kinds of larval tapeworm, cysticerci of the pork tapeworm T. solium and hydatid cysts of the dog tapeworm Echinococcus granulosus. The encysted larvae, cystercerci, of T. solium in the flesh of pigs, known as "measly pork," were well known to the ancient Greeks and are referred to by Aristotle (384 to 322 BC), who, in the section on diseases of pigs in his History of Animals, gives a detailed and accurate account of "bladders that are like hailstones" . Although the cysts in the muscle cause no obvious illness in humans, cysts in the brain can cause symptoms resembling epilepsy, and these must have been apparent in early civilizations. However, there is nothing in the encyclopedic works of Hippocrates to suggest that the Greek physicians knew that humans harbored such cysts or suffered from any conditions associated with them. There is, however, indirect evidence from different cultures that people were aware of the possible dangers inherent in eating the flesh of pigs. Kuchenmeister comments that infections with cysticerci are not found in those, such as Jews and Muslims, whose religious beliefs forbid the consumption of pork , but as we have already seen, similar arguments have been put forward with respect to Trichinella spiralis infections. There are accounts of what are possibly cysticerci in humans by Johannes Udalric Rumler in 1558, Domenico Panaroli in 1652, and Thomas Wharton in 1656, but none of these observers realized that the structures they described were parasites . The first reliable accounts of cystercerci as parasites of some kind are by Philip Hartmann in 1688 and Marcello (Marcus) Malpighi in 1697 , but the realization that these cysts were the larval stages of tapeworms had to await studies by Johann Goeze in 1784 . The demonstration of the life cycle of T. solium shed new light on the nature of the human condition, cysticercosis, and it became apparent that humans could probably become infected with the larval stages of T. solium when they ingested the tapeworm eggs. Although the conclusive experiments could not be carried out for ethical reasons, many experiments with animals and observations of humans established without doubt by the middle of the 19th century that cysticercosis was caused by the ingestion of the eggs of T. solium . These observations had a massive impact on the control of tapeworm infections in humans by restricting the amount of meat of infected animals available for human consumption. There are brief accounts of the history of cysticercosis by Nieto and more detailed accounts by Foster and Grove . There are also less easily accessible accounts by Vosgien , Henneberg , and Guccione . The most serious human disease caused by a larval cestode is echinococcosis, or hydatid disease, resulting from accidental infection with larval stages of the canid tapeworm, Echinococcus granulosus, which frequently occurs as an adult in dogs and as a larval cyst in wild and domesticated animals including sheep. The massive bladder-like hydatid cysts, particularly in the liver, were well known in ancient times, and there are references to such cysts in ritually slaughtered animals in the Babylonian Talmud and, in animals slaughtered for food, by Hippocrates in the fourth century BC, Arataeus in the first century AD, and Galen in the second century AD . There are also descriptions of hydatid cysts in humans in the Corpus Hippocratorum and in the works of Galen and in later European medical texts, in which they have variously been considered to be sacs of mucus, enlarged glands, distorted blood vessels, lymphatic varices, or accumulations of lymph . Francisco Redi in the 17th century was the first to appreciate the parasitic nature of these cysts , but credit for the hypothesis that these cysts were the larval stages of tapeworms goes to the German clinician and natural historian Pierre Simon Pallas, who showed this in 1766 . It was not until 1853 that Carl von Siebold demonstrated that Echinococcus cysts from sheep gave rise to adult tapeworms when fed to dogs , and in 1863 Bernhard Naunyn found adult tapeworms in dogs fed with hydatid cysts from a human . There are good accounts of the history of hydatid disease by Foster and Grove . Humans also harbor the adults of Diphyllobothrium latum, the broad or fish tapeworm that lives in the intestine. Eggs are passed out in the feces, and the first larval stage, the coracidium, develops within the egg and is eaten by a copepod, in which it develops to the second larval stage, the procercoid. When an infected copepod is eaten by a fish, the procercoid develops into the third larval stage, the plerocercoid, and when a human eats an infected fish, the plerocercoid develops into an adult tapeworm in the gut. The broad tapeworm was well known in antiquity and is mentioned, sometimes indirectly, in the major classical medical writings including the Ebers papyrus, the Corpus Hippocratorum, and the works of Celsus and Avicenna. However, there are no accurate early clinical records because there are few overt signs of the infection apart from abnormal hunger, malaise, and abdominal pain. Early descriptions of the worm tend to be unreliable because, as has already been mentioned, there was considerable confusion with the two common species of Taenia. Nevertheless, by the beginning of the 17th century, it became apparent that there were two very different kinds of tapeworm (broad and taeniid) in humans . It is generally agreed that Diphyllobothrium was first recognized as being distinct from Taenia by the Swiss physician Felix Plater, who also provided the first descriptions of the disease at the beginning of the 17th the century (, 316). The first accurate description of the proglottids was by another Swiss biologist, Charles Bonnet, in 1750 , but, unfortunately, the worm he illustrated had a Taenia scolex, a mistake he remedied in 1777 . By the middle of the 18th century, it was apparent that infections with D. latum occurred in humans whose diet was mainly fish. However, it was not until the life cycles of other tapeworms of zoological interest had been elaborated that further progress became possible, since the existence of three hosts in the life cycle, human, fish, and copepod, confused the issue. An understanding of the life cycle of this parasite began in 1790, when the Dane Peter Christian Abildgaard observed that the intestine of sticklebacks contained worms that resembled the tapeworms found in fish-eating birds ; however, it was some time before there was any significant advance in our understanding of the life cycle of D. latum. In the meantime, there were a number of misleading observations until 1881, when the German zoologist Maximillian Gustav Christian Carl Braun realized that the unsegmented tapeworms common in pike and other fish were the larval stages of D. latum and succeeded in infecting dogs with these plerocercoids; in 1882 he achieved similar results in humans . Braun suspected that this was not the whole story, but it was many years later that two Polish scientists, Constantine Janicki and Felix Rosen, working in Switzerland, incriminated copepods in the life cycle and showed that they fed on the eggs of the tapeworm and were then eaten by fish, which, in their turn, were eaten by humans . There are good accounts of Dipyllobothrium and diphyllobothriasis by Foster and Grove . DISCOVERY OF THE PARASITIC PROTOZOA : Because of their small size, it was not possible to recognize any protozoa until the invention of the microscope and its use by Antonie van Leeuwenhoek toward the end of the 17th century. The study of parasitic protozoa only really began two centuries later, following the discovery of bacteria and the promulgation of the germ theory by Pasteur and his colleagues at the end of the 19th century. Amoebae and Amoebiasis | Humans harbor nine species of intestinal amoebae, of which only one, Entamoeba histolytica, is a pathogen. The life cycle is simple. The amoebae live and multiply in the gut and form cysts that are passed out in the feces and infect new individuals when they are consumed in contaminated water or food. Most infections are asymptomatic, but some strains of E. histolytica can invade the gut wall, causing severe ulceration and amoebic dysentery characterized by bloody stools. If the parasites gain access to damaged blood vessels, they may be carried to extraintestinal sites anywhere in the body, the most important of which is the liver, where the amoebae cause hepatic amoebiasis. Supposed evidence that both the intestinal and liver forms of the disease were recognized from the earliest times is circumstantial because there are so many causes of both the bloody dysentery characteristic of amoebiasis and the symptoms of hepatic amoebiasis that many of these records are open to other interpretations . With these reservations in mind, the earliest record is possibly that from the Sanskrit document Brigu-samhita, written about 1000 BC, which refers to bloody, mucose diarrhea . Assyrian and Babylonian texts from the Library of King Ashurbanipal refer to blood in the feces, suggesting the presence of amoebiasis in the Tigris-Euphrates basin before the sixth century BC , and it is possible that the hepatic and perianal abscesses described in both Epidemics and Aphorisms in the Corpus Hippocratorum refer to amoebiasis . Since epidemics of dysentery by itself are likely to result from bacterial infections and dysentery associated with disease of the liver is likely to be amoebic, later records are easier to interpret. In the second century AD, Galen and Celsus both described liver abscesses that were probably amoebic, and the works of Aretaeus, Archigenes, Aurelanus, and Avicenna toward the end of the first millennium give good accounts of both dysentery and hepatic involvement . As amoebiasis became widespread in the developed world, there were numerous records of "bloody flux" in Europe, Asia, Persia, and Greece in the Middle Ages . The disease appears to have been introduced into the New World by Europeans sometime in the 16th century , and with the later development of European colonies and increased world trade, there are numerous clear descriptions of both the intestinal and hepatic forms of amoebiasis. In the 19th century, several books mainly concerned with diseases in India, including Researches into the Causes, Nature and Treatment of the More Prevalent Diseases of India and of Warm Climates Generally by James Annersley, clearly refer to both intestinal and hepatic amoebiasis , and it is now generally agreed that this book contains the first accurate descriptions of both forms of the disease. The connection between amoebic dysentery and liver abscesses was described by William Budd, the English physician who discovered the method of transmission of typhoid . The amoeba itself, E. histolytica, was discovered by Friedrich Losch (also known as Fedor Lesh) in 1873 in Russia , and Losch also established the relationship between the parasite and the disease in dogs experimentally infected with amoebae from humans. Stephanos Kartulis, a Greek physician, also found amoebae in intestinal ulcers in patients suffering from dysentery in Egypt in 1885 and 1896 and noted that he never found amoebae from nondysenteric cases . Kartulis also showed that cats could be infected with amoebae per rectum and developed dysentery a finding discussed below. The authoritative report by William Thomas Councilman and Henri Lafleur, working at the Johns Hopkins Hospital in 1891, represents a definitive statement of what was known about the pathology of amoebiasis at the end of the 19th century, and much of it is still valid today . It was pointed out above that humans harbor several species of amoebae. The most common are E. histolytica, which has just been considered, and a larger and superficially similar harmless species, E. coli; the presence of these two parasites confused early workers in this field. The first clues that there was more than one species in humans came from the work of Heinrich Iranaus Quincke and Ernst Roos working in Kiel in 1894, who observed that cats could only be infected per rectum or orally with cysts of amoebae that contained ingested red blood cells and not with those that did not, i.e., E. coli . Thereafter, the most contentious arguments relate to the various morphologically identical species and strains of E. histoloytica, and their relationship to disease has only recently been resolved by using biochemical techniques that clearly show that the presence of two common species, E. histolytica, which can cause disease, and E. dispar, which cannot . The history of amoebiasis is well documented. The most comprehensive account of the early history is that by Dobell , and there are also good accounts of the early history by Bray , Foster , Kean , Scott, , and Wenyon and reviews containing more recent information by Craig , Guirola , Imperato , Martinez-Baez, , Stilwell , and Svanidtse . Giardia and Giardiasis | Giardia holds a special place in the science of parasitic protozoology because the parasite Giardia duodenalis, also known as G. lamblia or G. intestinalis, was the first parasitic protozoan of humans seen by Antonie van Leeuwenhoek in 1681 . The life cycle of Giardia is very simple: the parasites multiply in the duodenum and form cysts that are passed out in the feces and infect new individuals when they are swallowed in food or water. Most infected individuals show few or no signs of infection, but in some, particularly children, there may be malabsorbtion, diarrhea, and abdominal pain. G. duodenalis was first seen by Leeuwenhoek and, interestingly, associated by him with his own loose stools. Leeuwenhoek's illustrations are not very informative, and the first good illustrations of Giardia are those of Vilem Lambl in 1859 . The parasite received little attention until 1902, when the American parasitologist Charles Wardell Stiles began to suspect that there was a causal relationship with diarrhea . This was not followed up until the 1914 to 1918 World War, when soldiers with diarrhea were found to pass Giardia cysts that caused similar symptoms when administered to laboratory animals . In 1921, Clifford Dobell suggested that Giardia was a pathogen , and in 1926, Reginald Miller, a physician working in London, conclusively showed that some children infected with Giardia did suffer from malabsorption whereas others acted as unaffected carriers . It was not until 1954, however, that the detailed studies by the American physician Robert Rendtorff produced unambiguous evidence linking the parasite with the disease . In the 300 years since Giardia was first discovered, it has become recognized as a common parasite and potential pathogen worldwide; however, it is still not known how many species infect humans and what role, if any, reservoir hosts play in the epidemiology of the infection. Fuller accounts of the history of giardiasis are given by Wenyon and Farthing . African Trypanosomes and Sleeping Sickness | African trypanosomiasis is caused by infection with two subspecies of trypanosomes, Trypanosoma brucei gambiense, which causes Gambian or chronic sleeping sickness, and T. b. rhodesiense, which causes Rhodesian or acute sleeping sickness. The trypanosomes multiply in the blood and are taken up by tsetse flies when they feed. Within the tsetse fly, there is a phase of multiplication and development resulting in the formation of infective trypanosomes in the salivary glands of the fly, which are injected into a new host when the fly feeds. The infection itself causes a number of symptoms including anemia, wasting and lethargy, and, in some cases, if the parasites pass into the brain and cerebrospinal fluid, coma and death. There are similar parasites in wild and domesticated animals. The first definitive accounts of sleeping sickness are by an English naval surgeon, John Atkins, in 1721 and Thomas Winterbottom, who coined the term "negro lethargy" in 1803 . An appreciation of the real cause of the disease was not possible until Pasteur had established the germ theory toward the end of the 19th century. Trypanosomes had been seen in the blood of fishes, frogs, and mammals from 1843 onward, but it was not until 1881 that Griffith Evans found trypanosomes in the blood of horses and camels with a wasting disease called surra and suggested that the parasites might be the cause of this disease . These observations led to the most important discoveries about human and animal trypanosomiasis shortly afterwards. In 1894, David Bruce, a British army surgeon investigating an outbreak of nagana, a disease similar to surra, in cattle in Zululand, was looking for a bacterial cause and found trypanosomes in the blood of diseased cattle; he demonstrated experimentally that these caused nagana in cattle and horses and also infected dogs. He also observed that infected cattle had spent some time in the fly-infested "tsetse belt" and that the disease was similar to that in humans with negro lethargy and fly disease of hunters . Trypanosomes were seen in human blood by Gustave Nepveu in 1891 . In 1902, Everett Dutton identified the trypanosome that causes Gambian or chronic sleeping sickness (T. b. gambiense) in humans , and in 1910 J. W. W. Stephens and Harold Fantham described T. b. rhodesiense, the cause of Rhodesian or acute sleeping sickness . Although Bruce had shown that trypanosome infections in cattle were acquired from tsetse flies, he thought that transmission was purely mechanical, and the role of the tsetse fly in the transmission of sleeping sickness remained controversial until Friedrich Kleine, a colleague of Robert Koch, demonstrated in 1909 the essential role of the tsetse fly in the life cycle of trypanosomes . The persistence of trypanosomes in the blood and the existence of successive waves of parasitemia were described in detail by Ronald Ross and David Thompson in 1911 , but the actual mechanism of what happens and how the parasite evades the immune response, now called antigenic variation, was not elaborated until the work of Keith Vickerman in 1969 . The story of African sleeping sickness is told briefly by Hoare and in more detail by Foster , Nash , Lyons , Wenyon , and Williams . South American Trypanosomiasis: Chagas' Disease | Chagas' disease is caused by infection with another trypanosome, Trypanosoma cruzi, transmitted by insects belonging to the order Hemiptera or true bugs, commonly known as kissing bugs because of their tendency to bite the lips and face. The transient trypanosome forms circulate in the blood and are taken up by a blood-sucking bug when it feeds. The parasites multiply in the gut of the bug, and infective forms are passed out in the feces while the bug is feeding on a new host and are rubbed into the bite. In the human host, parasites enter and multiply in a variety of different cells and eventually induce what are thought to be autoimmune responses that results in the destruction of both infected and uninfected tissues. The nature of the disease depends on the tissues and organs involved, and the most conspicuous forms are massive distension of the intestinal tract, especially the esophagus and colon, and destruction of cardiac muscle, which can result in death many years after the initial infection. T. cruzi infections are common in many mammals on the American continent, but the human disease now occurs only in South and Central America. The earliest indication that Chagas' disease is an ancient infection in South America comes from the examination of spontaneously mummified human remains from Chile between 470 BC and AD 600 that show clear signs of the characteristic destructive nature of the disease . The use of immunological and molecular techniques has made it possible to detect the presence of T. cruzi without necessarily visualizing the parasites themselves. T. cruzi DNA has been detected in the heart and esophagus of mummified bodies from Peru and northern Chile dating from 2000 BC to AD 1400 and in samples from bodies in museums from northern Chile from about AD 1000 to 1400 . The parasites themselves have also been identified by light and electron microscopy in a Peruvian mummy from the 15 to 16th century AD . The history of T. cruzi and Chagas' disease really begins with a series of discoveries by the Brazilian scientist Carlos Chagas, between 1907 and 1912. Chagas not only discovered the trypanosome T. cruzi and demonstrated its transmission by bugs but also described the disease that affects some 18 million people and now commemorates his name. Chagas' first observation was that the blood-sucking bugs that infested the poorly constructed houses harbored flagellated protozoa and that when these flagellates were injected into monkeys and guinea pigs, trypanosomes appeared in the blood . Chagas later found the same trypanosomes in the blood of children with an acute febrile condition and suspected that blood-sucking bugs might also transmit the parasite to humans, but he thought that the trypanosomes were transmitted via the bite of the insect . It was the French parasitologist Emile Brumpt who demonstrated transmission via the fecal route . The links between infection with T. cruzi and the various signs of Chagas' disease, such as distended colon and esophagus and cardiac failure, were not determined until the work of Fritz Koberle in the 1960s . Exactly how the damage to heart and nerves is caused and what role the autoimmune component plays are still controversial. The history of Chagas' disease has been well documented by Scott , Lewinsohn , Leonard , Miles , and Wenyon . Leishmania and Leishmaniasis | Leishmaniasis, caused by several species of Leishmania, is transmitted by sandflies and occurs in various forms in the Old and New World. The parasites infect and multiply in macrophages and are taken up by sandflies when they feed. In the gut of the sandfly, the parasites multiply and reach the mouthparts, from where they are injected into a new host when the sandfly feeds again. The disease, leishmaniasis, takes a number of forms ranging from simple cutaneous ulcers to massive destruction of cutaneous and subcutaneous tissues in the mucocutaneous forms and the involvement of the liver and other organs in the visceral form. From a historical viewpoint, it is easiest to consider the Old World forms first. Old World cutaneous leishmaniasis, known as oriental sore, is an ancient disease, and there are descriptions of the conspicuous lesions on tablets in the library of King Ashurbanipal from the 7th century BC, some of which are thought to have been derived from earlier texts from 1500 to 2500 BC . There are detailed descriptions of oriental sore by Arab physicians including Avicenna in the 10th century, who described what was (and is) called Balkh sore from northern Afghanistan, and there are later records from various places in the Middle East including Baghdad and Jericho; many of the conditions were given local names by which they are still known . Old World visceral leishmaniasis, or kala azar, characterized by discolored skin, fever, and enlarged spleen, is easily confused with other diseases, especially malaria. Kala azar was first noticed in Jessore in India in 1824, when patients suffering from fevers that were thought to be due to malaria failed to respond to quinine; by 1862 the disease had spread to Burdwan, where it reached epidemic proportions . The cause remained unknown, and several eminent clinicians, including Ronald Ross, were convinced that kala azar was a virulent form of malaria . It was not until the parasite, L. donovani, was discovered in 1900 by Leishman and Donovan (see below) that the true nature of the disease became apparent . The discovery of the parasites responsible for the Old World cutaneous disease is controversial, and a number of observers described structures that might or might not have been leishmanial parasites from oriental sores . Credit for their discovery is usually given to an American, James Homer Wright , although there is no doubt that they were actually seen in 1885 by David Cunningham , who did not realize what they were, and in 1898 by a Russian military surgeon, P. F. Borovsky . The discovery of the parasite that causes visceral leishmaniasis, L. donovani, is less controversial, and it is universally accepted that a Scottish army doctor, William Leishman , and the Professor of Physiology at Madras University, Charles Donovan , independently discovered the parasite in the spleens of patients with kala azar. It is fair to point out that Borovsky's discoveries were unknown to Wright and to Leishman and Donovan. The search for a vector was a long one, and it was not until 1921 that the experimental proof of transmission to humans by sandflies belonging to the genus Phlebotomus was demonstrated by the Sergent brothers, Edouard and Etienne . The actual mode of infection, through the bite of the sandfly, was not finally demonstrated until 1941 . The history of Old World leishmaniasis is described by Garnham , Manson-Bahr , and Wenyon . In the New World, cutaneous and mucocutaneous leishmaniasis cause disfiguring conditions that have been recognized in sculptures since the 5th century and in the writings of the Spanish missionaries in the 16th century . It was originally thought that New World Leishmaniasis and Old World leishmaniasis were the same, but in 1911 Gaspar Vianna found that the parasites in South America differed from those in Africa and India and created a new species, Leishmania braziliensis . Since then, a number of other species unique to the New World have been described. Following the discovery of the sandfly transmission of Old World leishmaniasis, the vectors in the New World were also assumed to belong to the genus Phlebotomus, but in 1922 it was discovered that the genus involved was actually Lutzomyia. Over the last two decades, the complex pattern of species of parasite, vector, reservoir host, and disease has been painstakingly elaborated by Ralph Lainson and his colleagues . Malaria | Malaria is one of the most important infectious diseases in the world, and its history extends into antiquity. The disease is caused by four species of the genus Plasmodium, P. falciparum, P. vivax, P. ovale, and P. malariae. Similar parasites are common in monkeys and apes. It is now generally held that malaria arose in our primate ancestors in Africa and evolved with humans, spreading with human migrations first throughout the tropics, subtropics, and temperate regions of the Old World and then to the New World with explorers, missionaries, and slaves. The characteristic periodic fevers of malaria are recorded from every civilized society from China in 2700 BC through the writings of Greek, Roman, Assyrian, Indian, Arabic, and European physicians up to the 19th century. The earliest detailed accounts are those of Hippocrates in the 5th century BC, and thereafter there are increasing numbers of references to the disease in Greece and Italy and throughout the Roman Empire as its occurrence became commonplace in Europe and elsewhere. Over this period, it became clear that malaria was associated with marshes, and there were many ingenious explanations to explain the disease in terms of the miasmas rising from the swamps . Our scientific understanding of malaria did not begin until the end of the 19th century following the establishment of the germ theory and the birth of microbiology, when it became necessary to discover the cause of the disease that was then threatening many parts of the European empires. The discovery of the malaria parasite and its mode of transmission are among the most exciting events in the history of infectious diseases, and this topic has been reviewed many times, particularly by Bruce-Chwatt , Garnham , Harrison , McGregor , Poser and Bruyn , and Wenyon . The life cycle is a very complex one that begins when an infected anopheline mosquito injects sporozoites, the infectious stages, into the blood of its host. Sporozoites enter and multiply in liver cells, and thousands of daughter forms, merozoites, are released into the blood. These merozoites invade red blood cells, in which another phase of multiplication occurs; this process is repeated indefinitely, causing the symptoms of the disease we call malaria. Some merozoites do not divide but develop into sexual stages, the male and female gametocytes, that are taken up by another mosquito when it feeds; fertilization and zygote formation occur in the gut of the mosquito. The zygote develops into an oocyst on the outside of the mosquito gut, and within the oocyst there is another phase of multiplication that results in the production of sporozoites that reach the salivary glands to be injected into a new host. The parasites in the blood were first seen in 1880 by a French army surgeon, Alphonse Laveran, who was looking for a bacterial cause of malaria and who immediately realized that the parasites were responsible for the disease . The discovery that the mosquito acted as a vector was due to the intuition of Patrick Manson. Manson had already demonstrated that filarial worms, also blood parasites, were transmitted by mosquitoes and postulated that the vector of the malaria parasite might also be a mosquito, partly because of his knowledge of the life cycle of filarial worms and partly because of the known association between the disease and marshy places in which mosquitoes breed . Manson was unable to undertake this investigation himself and persuaded Ronald Ross, an army surgeon, to carry out the work in India. The story of Ross' discoveries has been told many times and is not repeated in detail here, since there are excellent accounts by Ross himself and in the Ross-Manson collected letters and also by Bruce-Chwatt , Garnham , Harrison , Manson-Bahr , Nye and Gibson , Poser and Bruyn , and Russell . In 1897, Ross saw what we now know to be the oocysts of P. falciparum in an anopheline mosquito that had fed on a patient with crescentic malaria parasites (gametocytes) in his blood, but he was unable to follow this up at the time . Turning his attention to a bird malaria, P. relictum, he found all the stages of the parasite in culicine mosquitoes that had fed on infected sparrows . In making this discovery, Ross acknowledged the work of a young Canadian, William George MacCullum, whose studies on the development of the sexual stages of a related avian parasite Halteridium (=Haemoproteus) columbae had led him to the conclusion that these parasites were similar to those in the blood of humans with malaria . In the same year that Ross made his discovery, the Italian malariologists Giovanni Battista Grassi, Amico Bignami, and Giuseppe Bastianelli described the developmental stages of malaria parasites in anopheline mosquitoes; the life cycles of P. falciparum, P. vivax, and P. malariae were described a year later . For nearly 50 years, the life cycle in humans remained incompletely understood and nobody knew where the parasites, which could not be seen in the blood, developed during the first 10 days after infection. In 1947, Henry Shortt and Cyril Garnham, working in London, showed that a phase of division in the liver preceded the development of parasites in the blood . The final brick was put in place when an American clinician, Wojciech Krotoski, in collaboration with Garnham's team, showed that in some strains of P. vivax the stages in the liver could remain dormant for several months . Sadly, the discovery of the life cycle of the malaria parasite eventually led to acrimony between Ross and Manson and between the British and the Italians, something that still rumbles on a century later . Toxoplasma, Toxoplasmosis, and Infections Caused by Related Organisms | Toxoplasmosis is one of the most common and widespread parasitic infections but is relatively little known because in the majority of cases, infections are asymptomatic; however, it can be a serious cause of mortality and morbidity in fetuses and immunodeficient individuals. The parasite that causes the infection, Toxoplasma gondii, was discovered independently by the French parasitologists Charles Nicolle and Louis Herbert Manceaux while looking for a reservoir host of Leishmania in a North African rodent, the gundi Ctenodactylus gondi and by Alfonso Splendore in Sao Paulo in rabbits . At about the same time, Samuel Taylor Darling saw what were probably similar organisms in a human , and the first definitive observation of T. gondii from a child in connection with an infection was made by a Czech physician, Josef Janku, in 1923 . Even then, T. gondii was largely regarded as an interesting curiosity until an association with human congenital disease was recognized in 1937 by Arne Wolf and David Cowen . This association was followed by the realization that T. gondii rarely causes disease even though it is a very common parasite of adults but that in pregnant women the parasite can cross the placenta and can damage the fetus. The early history of the discovery of T. gondii and toxoplasmosis is discussed by Wenyon and Dubey and Beattie . While these developments were taking place, there were increasing numbers of records from virtually all species of mammals and many birds, but the nature of the parasite remained obscure until the life cycle had been worked out. The life cycle of T. gondii is a very complicated one and remained elusive until 1970, when scientists in Britain, Germany, The Netherlands, and the United States independently demonstrated that this parasite was a stage in the life cycle of a common intestinal coccidian of cats. In the most simple form of the life cycle, cats become infected when they swallow oocysts, the resistant infective stages containing sporozoites, which invade and multiply in intestinal cells, where sexual stages are produced, fertilization occurs, and oocysts are produced. However, there is an alternative life cycle. If the oocysts are swallowed by a mouse (or any other nonfeline host), multiplication occurs in the intestinal cells, but instead of sexual stages being produced there follows a disseminated infection during which resistant stages form in the brain and muscle. There is no further development in the mouse, but when the mouse is eaten by a cat, the life cycle reverts to its basic sexual pattern. Humans are infected in the same way as mice if they consume oocysts, but they can also become infected by eating any kind of meat containing the resistant forms. It is therefore not surprising that the life cycle remained elusive until William McPhee Hutchison, working in Glasgow in 1965, showed that the infectious agent was passed in the feces of cats. At the time he thought that it was transmitted with a nematode worm, as happens with the flagellate Histomonas meleagridis and the nematode Heterakis gallinarum in fowl. Hutchison subsequently identified protozoan cysts in the feces as those of a coccidian related to Isospora, a common parasite of cats . In the meantime, other groups were following up Hutchison's 1965 observation of the presence of infectious agents in the feces of cats, and Hutchison's incrimination of the isosporan parasite of cats as T. gondii was independently confirmed by Jack Frenkel and Harley Sheffield in the United States, Gerhard Piekarski in Germany , and J. P. Overdulve in The Netherlands . The discovery of the T. gondii life cycle initiated a massive search for similar phases in the life cycles of other coccidian parasites, with the result that a number of protozoa that had not been properly identified were classified as stages in the life cycle of other poorly understood coccidians and that in many cases transmission depended on a predator-prey relationship . Humans are infected with two related parasites, Sarcocystis hominis and S. suihominis, acquired from beef and pork, respectively, and S. lindemanni, whose source is unknown. The early history of our knowledge of Sarcocystis is covered by Wenyon , and subsequent discoveries are described by Tadros and Laarman . Humans are also hosts to three other species of coccidia, Isospora belli, Cryptosporidium parvum, and Cyclospora cayetanensis, that have in the past been regarded as rare and accidental curiosities but have recently been identified as pathogens in AIDS patients and other immunocompromised individuals. All have simple life cycles initiated by the ingestion of oocysts followed by multiplication and spread within the intestinal cells of the host and the eventual production of sexual stages, as for T. gondii infection in cats. C. parvum was discovered in 1912 by the American parasitologist Edward Ernest Tyzzer in the gastric glands of laboratory mice in which he had previously found another species, C. muris . C. parvum is not very host specific, and the first cases in humans were recorded in 1976 independently by Nime and Meisel . From 1981 onward, numerous new cases began to be recognized in AIDS patients. The oocysts Cryptosporidium are very resistant to chlorination, and the source of these infections is probably drinking water contaminated with cattle feces. Cryptosporidium infections are now known to be very common and have caused a number of epidemics in which the victims have experienced abdominal pain and diarrhea. In immunodepressed individuals, especially those infected with HIV, the infection can become disseminated to the liver, pancreas, and respiratory tract and can be fatal. There is an excellent history of human cryptosporidiosis by McDonald and a short but useful review by Dubey et al. . C. cayetanensis is another coccidian that is associated mainly with AIDS. In 1979, the English parasitologist Richard Ashford found an unidentified coccidian in patients in Papua New Guinea , but it received little attention until it was found again in the stools of patients with HIV by Soave et al. in 1986 . In 1992, this parasite was named Cyclospora cayetanensis , and since then it has been identified as the cause of a number of outbreaks of diarrhea and fatigue in both immunocompetent and immunosuppressed individuals . Cyclospora infections are known to be transmitted in water and on fruit, but the original source is not known. The last of this group of parasites, Isospora belli, discovered by Woodcock in 1915 , is another coccidian frequently found in asymptomatic immunocompetent individuals but associated with diarrhea in AIDS patients. The whole subject of parasitic infections in immunocompromised hosts is discussed by Ambroise-Thomas . Microsporidians | Microsporidians are extemely common spore-forming parasites of vertebrates and invertebrates that were until relatively recently grouped with myxosporidians as cnidosporidians and classified with or close to the Sporozoa. We now know that the myxosporidians are more closely related to the Metazoa than the Protozoa and that the microsporidians are more closely related to the Fungi . Nevertheless, microsporidians are still regarded as the province of parasitologists and have become important as concomitant infections in AIDS patients. The life cycle of microsporidians is quite complex. The most conspicuous stage is the resistant gram-positive spore containing a coiled filament and an infective body, the sporoplasm. The host becomes infected when the spore is ingested or inhaled. The sporoplasm is extruded through the filament and penetrates a host cell, within which the organism multiplies and spreads to other cells; eventually, another generation of spores is produced. There are, however, many variations on this basic pattern. What are now thought to have been the spores of Nosma bombycis were described by Nageli investigating an outbreak of a disease called pebrine in the silkworm Bombyx mori in 1857 and studied in much more detail by Louis Pasteur in 1865 to 1870 . During the 19th century, microsporidians attracted considerable attention mainly as parasites of invertebrates. Our knowledge of human microsporidiosis in the past is limited because of difficulties in interpreting various structures that might or might not have been spores, but from the second decade of the 20th century onward, there have been a number of sporadic reports of what might have been human microsporidial infections. The first case was probably that of Encephalitozoon chagasi in a newborn baby recorded in 1927 , but the first authenticated record was not until 1959, when Hisakichi Matsubayashi and his colleagues in Japan found an Encephalitozoon sp. in boy with convulsions . Thereafter, there were reports of a number of sporadic cases of microsporidian infections in humans , but interest in this group really took off in 1988, when E. bieneusi was found in an AIDS patient . Since then, about 7 genera and 14 species associated with fulminating infections in immunodepressed patients and less serious infections in immunocompetent individuals have been described and the number of cases, particularly in AIDS patients, continues to rise . Despite their importance, very little is known about the transmission and epidemiology of the microsporidians. SUMMARY AND CONCLUSIONS : The history of parasitology is a fascinating one, and parasites have been the subjects of some of the most exciting discoveries in the field of infectious diseases. We now know that many of the important parasites encountered today not only existed but were widespread in their distribution before written records began, and our early ancestors must have been aware of the presence of the largest and most common worms and of some of the diseases caused by parasites. The subsequent history of human parasitology revolves around early descriptions of a particular disease and the identification of the parasite causing the disease, not necessarily in this order; the elaboration of the life cycle; and, finally, the establishment of the causal relationship between the parasite and the disease. In this review, it has been possible only to touch on the major events and some of the personalities involved in these discoveries, but the history of parasitology has been well served in the scientific literature and the interested reader is referred to the appropriate sections in books that are concerned mainly with aspects of medicine, particularly tropical medicine, such as those by Ackernecht , Brothwell and Sandison , Bynum and Porter , Chernin , Cox , Kiple , Mack , Norman , Ranger and Slack , Ransford , and Scott . There are also a number of publications dedicated to the history of parasitology, including those by Cox , Foster , Garnham , Hoeppli , and Warboys . The most comprehensive work on the history of any aspect of parasitology is A History of Human Helminthology , which contains over 800 pages of detailed accounts of all the discoveries in the field of human helminthology. The two-volume Tropical Medicine and Parasitology: Classical Investigations edited by Kean et al. requires a special mention, since it is an invaluable source of information consisting of whole articles, excerpts, and translations of most of the important papers in the history of parasitology. Backmatter: PMID- 12364372 TI - What Happened to the Streptococci: Overview of Taxonomic and Nomenclature Changes AB - Since the division of the Streptococcus genus into enterococci, lactococci, and streptococci in 1984, many changes in the nomenclature and taxonomy of the Streptococcus genus have taken place. The application of genetic comparisons has improved the proper classification of the different species. The Lancefield system of serogrouping the streptococci by the expression of beta-hemolysis on blood agar plates is still very useful for the identification of streptococci for patient management. The Lancefield grouping system cannot be used in itself for accurate identification of specific beta-hemolytic species, but it can be a useful part of the identification procedure. Except for identification of the "Streptococcus bovis group" of species and Streptococcus suis, Lancefield grouping is of little value in identification of the non-beta-hemolytic streptococci and related genera. In fact, identification of the non-beta-hemolytic species is problematic for conventional as well as commercially available identification procedures. A combination of conventional tests and specific chromogenic tests suggested by several investigators is presented and discussed. Tables are included that suggest tests and procedures to guide investigators attempting to identify all the species. Keywords: Introduction : The purpose of this communication is to outline the changes in taxonomy and nomenclature of the Streptococcus genus that have occurred in the past 15 years. These changes are the result of the application of DNA-DNA reassociation, 16S rDNA gene sequencing, and other molecular techniques that help delineate differences in bacterial genera and species. The 1984 Bergey's Manual of Systemic Bacteriology listed only seven genera of faculatatively anaerobic gram-positive cocci (GPC); Aerococcus, Leuconostoc, Micrococcus, Pediococcus, Staphylococcus, Streptococcus, and Stomatococcus . At present there are 17 different genera of GPC. The discussion is limited to the Streptococcus genus and closely related GPC that are catalase negative and display chains in the Gram stain. Technically, Leuconostoc bacteria fit into this category, but discussions on changes in this genus are not included, because iden-tification procedures and clinical relevance of the leuconostocs can be found in a review written in 1995 . Shortly before the publication of Bergey's Manual in 1986, the genus Streptococcus was split into three genera (Enterococcus, Lactococcus, and Streptococcus) . Changes in the Enterococcus and Lactococcus genera are not detailed, but some changes are included to explain changes in the Streptococcus genus. Many new species of streptococci have been added to the genus, and six new genera of GPC that form chains have been established (Abiotrophia , Granulicatella , Dolosicoccus , Facklamia , Globicatella , and Ignavigranum ). The majority of these genera were split off the Streptococcus genus by genetic and phenotypic information. The earliest attempt at differentiating the streptococci was probably made in 1903 by Shottmuller , who used blood agar to differentiate strains that were beta-hemolytic from those that were not. Before 1933, fermentation and tolerance tests were the only tests used for differentiating many of the streptococci. In 1933 Lancefield reported the technique of demonstrating specific carbohydrate "group" antigens associated with the beta-hemolytic strains . In 1937, Sherman proposed a scheme for placing the streptococci into four categories. These categories were organized by hemolytic reaction, group carbohydrate antigens, and phenotypic tests (primarily fermentation and tolerance tests) . Sherman's four divisions were the pyogenic division, the viridans division, the lactic division, and the enterococci. The pyogenic division included the beta-hemolytic strains with defined group antigens (A, B, C, E, F, and G). This division of the streptococci is not appreciably different from that of today's identification systems based on serogrouping. Sherman's viridans division included streptococcal species that were not beta-hemolytic, were not tolerant to high-pH growth conditions, were not salt tolerant, and did not grow at 10C. This group is still known today as the viridans streptococci, and many more species have been added to this classification. Sherman's lactic division included strains that were associated primarily with the manufacture of dairy products. They were not associated with human infections. This group differed from the pyogenic group by not being beta-hemolytic, by having the capacity to grow at 10C but not at 45C, and by failing to grow in broth containing 6.5% NaCl. Sherman's lactic division was reclassified as the Lactococcus genus in the mid-1980s. Lactococcus species that have recently been isolated from human infections have phenotypic characteristics that are not the same as those described by Sherman . Sherman's fourth division was termed the enterococci and included the four species known at that time. Although some of the enterococci were beta-hemolytic, other characteristics such as the capacity to grow in broths at high pH, high salt concentrations, and a wide temperature range (10 to 45C) differentiated them from the other three divisions. The number of Enterococcus species has increased to more than 20. All of the new species have phenotypic characteristics similar to those described by Sherman . The classification and identification of streptococci was severely hampered by a hierarchical dichotomous approach relying on a very limited number of complex characters (colony size, hemolysis, and group carbohydrate antigens) that resulted in species definitions that are often qualified by a number of exceptions. Moreover, because these organisms grow slowly and may require additional factors for isolation and characterization, as well as the somewhat cumbersome classification system, identification of streptococci to the species level is rarely performed in time to be relevant to the treatment of the patient from whom the organism was isolated. One of the most useful tools applied to the revision of the classification system for the Streptococcus genus is the application of 16S rRNA gene sequencing. Figure depicts a phylogenetic tree of the currently available sequences for the species included in the Streptococcus genus. I have chosen to present the identification schemes listed in the following tables based strictly on phenotypic characteristics because I feel that the clinical microbiologist can easily place the species into groups that are phenotypically related with only a few microbiologic tests. These phenotypic groups do not necessarily correlate with the groups shown in the genetic tree in Fig. . FIG. 1. | Phylogenetic relationship among 55 Streptococcus species based on analysis of 16S rRNA gene sequences. Phylogenetic relationship among 55 Streptococcus species based on analysis of 16S rRNA gene sequences. The dendrogram was constructed by the clustal method using the DNASTAR program. The units at the bottom of the tree indicate distance between sequence pairs. BETA-HEMOLYTIC STREPTOCOCCI : For clinical laboratories as well as taxonomists, one of the most useful phenotypic characteristics of streptococci is the reaction of the bacteria on blood agar plates. Hemolysis is used as a guide for managing patients as well as an aid in classification of the bacterium to the species level. J. H. Brown in 1919 first defined the reactions of streptococci on blood agar plates . This monograph is no longer available, but Brown's definitions are accurately shown in reference . It is very important that clinical and physician office laboratories accurately identify the beta-hemolytic reactions of the streptococci. Table lists all beta-hemolytic streptococci known to date. There is one exception included in this table; S. dysgalactiae subsp. dysgalactiae is not beta-hemolytic but is included for taxonomic reasons. Column 1 in Table lists all the species and subspecies of beta-hemolytic streptococci. Column 2 lists the Lancefield group antigens that are associated with each species. Columns 3 to 12 list 10 phenotypic characteristics that can be used to help identify the streptococci to the species level when hemolysis and identification of group carbohydrate antigen fail to do so. The last column gives the most common natural host of the species. With the exception of S. dysgalactiae subsp. dysgalactiae, S. equi subsp. equi, and two recently described species, S. phocae and S. didelphis , all the species listed in Table have been isolated from human infections. Table was constructed based on the molecular and phenotypic characteristics described by Schleifer and Kilpper-Balz , Vandamme et al. , Vieira et al. , and our own unpublished results of testing all reference strains for each of the species. Investigators used a variety of techniques including DNA-DNA reassociation, 16S rDNA sequencing, whole-cell protein analysis, multilocus enzyme electrophoresis, and phenotypic characteristics to help establish the species and subspecies included in this table. Streptococcus pyogenes | S. pyogenes is also known as beta-hemolytic group A streptococcus or Lancefield's group A strep (GAS). It is the most pathogenic bacterium in the genus Streptococcus. All clinicians should be aware that GAS is the agent that causes bacterial pharyngitis, impetigo, and a host of other infections including severe invasive diseases. The recent review by Cunningham is an excellent source of information for interested investigators to familiarize themselves with the severity and diversity of diseases caused by this bacterium . The major (but not exclusive) virulence factor associated with GAS is the M-protein antigen. This surface antigen is the antigen that allows the GAS to avoid phagocytosis and to survive in the human host. The hypervariable N-terminal portion of this protein dictates the type specificity of each antigen. Lancefield prepared type-specific antisera to many of these antigens and used the antisera in a capillary precipitin test to subtype the GAS . This system is still in use in some laboratories after more than 60 years. In recent years, an alternative system called emm typing has been developed, which uses the sequence of the gene that encodes the M-protein hyperviable region. Correlation between the serologic and emm typing systems is very good, and in most cases the emm type reflects the M-protein serologic type. The emm typing system can be accessed through the CDC web site . Methods for emm typing and the sequences of more than 120 emm-types of GAS are described in references and . From Table the reader can determine that S. pyogenes is not the only Streptococcus that may possess the group A antigen. The true incidence of non-S. pyogenes GAS strains found in human infections is unknown, but from the information available to us at the Centers for Disease Control and Prevention (CDC) Streptococcus laboratory, these strains are not common. More information regarding the two species S. dysgalactiae subsp. equisimilis and S. anginosus is given below where these two species are discussed. S. pyogenes is best identified by demonstration of the group A antigen on the cell. Presumptive identification can be made by bacitracin susceptibility or pyrrolidonylarylamidase activity. S. pyogenes strains are the only beta-hemolytic streptococci that are positive in both of these tests. Some of the other beta-hemolytic streptococci can also be positive in one or the other, but not both, of these tests; therefore, these tests alone are not 100% specific for S. pyogenes . Streptococcus agalactiae | S. agalactiae, or Lancefield's group B streptococcus (GBS) is the most common cause of neonatal sepsis . Reference is an excellent starting point for those interested in learning more about the epidemiology and types of diseases caused by this organism. Two major avenues of investigation have been explored for the prevention of GBS disease: the development of vaccines and the screening procedures for the presence of GBS in anal, cervical or vaginal carriers and subsequent antimicrobial management . S. agalactiae is the only Streptococcus species that has the group B antigen. Some other streptococcal species have recently been identified however, that cross-react with commercial slide agglutination tests (see the discussion of S. porcinus, below). GBS can also be presumptively identified by the CAMP and hippurate reactions. Together with the unique hemolytic reaction (very small zone of lysis), these two presumptive tests are very accurate in the identification. Streptococcus dysgalactiae subsp. dysgalactiae | The exact composition of the taxon Streptococcus dysgalactiae has been in a state of flux for the past few years. S. dysgalactiae subsp. dysgalactiae is the only species listed on Table that is not beta-hemolytic. This species is included because of the other subspecies included in this discussion. Isolation of this bacterium from human infections has not been documented. The organism has virulence factors similar to those of S. pyogenes, including M-like proteins . Identification by determining the characteristics in Table should be considered presumptive. Determining that an alpha-hemolytic streptococcus has group C antigen is insufficient for identification; other non-beta-hemolytic streptococci, e.g., viridans streptococal species, may also have group C antigen . Additional phenotypic or genetic characteristics should be used to identify this bacterium . Streptococcus dysgalactiae subsp. equisimilis | S. dysgalactiae subsp. equisimilis is the revised taxonomic epithet for what was previously termed S. equisimilis or Lancefield's group C Streptococcus. Genetic investigations indicated that Lancefield's group C strain (also known as group C human strain or S. equisimilis), Lancefield's group G and L strains, and the species known as S. dysgalactiae were all genetically similar and should be included in one taxon. Since S. dysgalactiae was the oldest officially recognized species, all of these entities were placed in the S. dysgalactiae classification . Later investigations indicated that Lancefield group C strain (S. equisimilis) and Lancefield's group G and L strains should be grouped into one category and that S. dysgalactiae should be placed in different category, thus creating two separate subspecies, S. dysgalactiae subsp. dysgalactiae and S. dysgalactiae subsp. equisimilis. More recently, investigators have shown that some strains of S. dysgalactiae subsp. equisimilis possess the group A antigen . Group A, C, G, and L S. dysgalactiae subsp. equisimilis strains are found in human infections. The true incidence figures are difficult to estimate but in our experience at the CDC, group C and G strains are found much more commonly in human infections than are group A and L strains of this species. Like S. dysgalactiae subsp. dysgalactiae, the strains possessing different group antigens of S. dysgalactiae subsp. equisimilis also have virulence factors similar to S. pyogenes, including emm gene homologs (see the CDC web site above). The group antigen can be used only as an aid in species identification. The phenotypic tests in Table should be used, together with hemolytic reaction and group antigen, to identify the species possessing Lancefield's group C antigen. A total of six different beta-hemolytic streptococcal species or subspecies can have the group C antigen. Streptococcus equi subsp. equi | Streptococcus equi subsp. equi is a beta-hemolytic group C Streptococcus that causes strangles in horses. To my knowledge, this species has not been isolated from humans. S. equi subsp. equi has a protein that induces opsonic antibodies in horses and is thought to be involved in the organism's virulence . Identification of this subspecies is based on the hemolytic reaction, demonstration of the group C antigen, differences in hydrolysis of esculin and starch, and fermentation of sorbitol and trehalose . Streptococcus equi subsp. zooepidemicus | Unlike S. equi subsp. equi, S. equi subsp. zooepidemicus is found in human infections. Most recently, this organism was associated with a large outbreak of nephritis in Brazil . The publication by Nicholson et al. provides references to previous nephritis outbreaks and other infections caused by this bacterium. Most human infections can be traced back to the consumption of contaminated dairy products. This organism is a frequent cause of bovine mastitis. S. equi subsp. zooepidemicus has a surface-exposed protein (Szp) that is antigenically variable, stimulates opsonic protective antibodies, and shares certain structural features with the S. pyogenes M protein . Although this protein does not have significant sequence homology to the M protein, it is tempting to speculate that Szp also plays an antiphagocytic role. S. equi subsp. zooepidemicus is identified by beta-hemolysis, group C antigen presence, hydrolysis of esculin (70% of strains) and starch, fermentation of sorbitol, and occasional fermentation of trehalose. Streptococcus canis | S. canis is an organism that was described in 1986 as having the Lancefield group G antigen and was isolated from animals, most frequently dogs (hence the name "canis") . Extensive phenotypic testing of isolates from dogs was described in 1994 , and the isolation of S. canis from a human with sepsis was reported in 1997 . It is not possible to estimate how frequently S. canis is isolated from humans, because most clinical laboratories do not perform phenotypic tests on group G streptococci other than the hemolytic reaction and group determination. The phenotypic profile given in Table for the identification of S. canis is that of nonhuman strains. It is not known whether human strains will have the same phenotypic characteristics. Streptococcus anginosus Group | None of the beta-hemolytic streptococci has caused more confusion about terminology and classification than the group of organisms listed as S. anginosus . There is no doubt that this group includes three distinct species and more subspecies. S. anginosus, S. constellatus, and S. intermedius were all collectively known as either S. anginosus or S. milleri at one time. The problem with the term "S. milleri" was that it was never accepted by the taxonomist as a confirmed taxonomic entity. DNA-DNA reassociation studies clearly show that the aforementioned species are distinct taxons . There are beta-hemolytic strains of each of the three species, and the strains may possess one of four different Lancefield group antigens or no group antigen. Adding to the confusion is the fact that non-beta-hemolytic strains of the three species are more common than beta-hemolytic strains. Non-beta-hemolytic varieties of each of the three species are grouped into the general classification of viridans streptococci. Whiley et al. have proposed an identification scheme based on degradation of chromogenic substrates for the differentiation of the three species . In addition, although these authors state that hemolysis and group antigens are of little value in differentiation of the species, more isolates of S. constellatus tend to be beta-hemolytic than of either S. anginosus or S. intermedius. Isolates of S. intermedius are rarely beta-hemolytic . These authors also state that there tends to be an association between the clinical sources and the three species. S. anginosus isolates are commonly isolated from urogenital and gastrointestinal sources, S. constellatus is often isolated from respiratory and many other sources, while S. intermedius strains are commonly identified from brain and liver abscesses. The distribution of the group antigens also shows some association with the species. Isolates of S. intermedius rarely have group antigens, while isolates of S. anginosus and S. constellatus often have group F, C, A, and G antigens (in order of frequency). Other investigators have also found similar distributions of the three species using Whiley's scheme and DNA reassociation as reference identification procedures . Very little is known about virulence factors produced by this group of bacteria. Two enzymes that may be considered virulence factors are alpha-N-acetylneuramidase (sialidase) and hyaluronidase; S. intermedius produces both of these enzymes, S. constellatus produces only hyaluronidase, and S. anginosus produces neither. The production of these enzymes is part of the identification scheme proposed by Whiley et al. . Identification of the beta-hemolytic as well as non-beta-hemolytic species are given in Table . Whiley's scheme is considered a standard for phenotypic identification of the three species and potential subspecies. Table has been constructed from the data presented in references , , , , , , , , , , and . The hemolytic reaction and Lancefield group antigens that are commonly associated with the three species are also included. Subspecies are also included. Column 2 under each of the species listed in Table represents the original Whiley description of the species; column 1 under each of the species represents a beta-hemolytic variant of each species; and column 3 represents an official or unofficial subspecies designation of that particular species. This table should be used as a guide for identification of the species, and confirmation of subspecies may require additional molecular documentation. All the species and subspecies listed in this table should have the phenotypic characteristics of the S. anginosus group listed in Table . Column 1 under S. anginosus is the description of the group of strains identified but not proposed for subspecies status by Whiley et al. . Column 3 under S. anginosus is the description of the "motile S. milleri" strains . Additional variants of S. anginosus have been described based on ribotypes, some of which correlate with the entities listed in Table . Commercial identification systems that have three of the seven chromogenic substrates suggested by Whiley et al. are available. The Fluo-Card Milleri (Key Scientific, Round Rock, Tex.) includes beta-d-fucosidase, beta-glucosidase, and alpha-glucosidase . Compared to the Whiley scheme, this system identified 98% of S. anginosus strains, 97% of S. constellatus strains, and 88% of S. intermedius strains. The Rapid ID-32 Strep system (Bio-Merieux, Marcy l'Etoile, France) (not available for clinical microbiology use in the United States) includes N-acetyl-beta-glucosaminidase, beta-glucosidase, and beta-galactosidase . These authors did not use a reference method to compare their identification, and so all 70 strains in their study were placed into one of the three species. Limia et al. evaluated the Rapid ID-32 Strep system as well as the Whiley system against a line blot hybridization assay. There was an 80% agreement between the genotypic method and Whiley's method. The agreement between the genotypic method and the Rapid ID-32 Strep system was 76%. The latter system was particularly inaccurate, with identification of only 57% of the S. intermedius strains. There have been reports that some of the problems with the Rapid ID-32 system is with preparation of the inoculum density and growth conditions . A third commercial system available for identification of the three species is the Becton Dickinson Microbiology Crystal Gram-Positive system . This system has beta-N-fucosidase, beta-glucosidase, and alpha-glucosidase. Very little information is available about the utility of this system. Molecular technology for specific identification of these species has been described. Pulsed-field gel electrophoresis , sequencing of specific genes and 16S rRNA genes , and species-specific probes have all been described. In summary, we know that beta-hemolytic strains of all species have been isolated from human infections (with and without various Lancefield antigens), but we know very little about the incidence and clinical significance of each of the redefined species and subspecies. Streptococcus constellatus subsp. pharyngis | Whiley et al. have recently described a subspecies of S. constellatus that is beta-hemolytic and carries Lancefield's group C antigen . These beta-hemolytic group C streptococci have a predeliction for the human throat and cause pharyngitis. Although still within the parameters of the criteria for inclusion in the species, the DNA-DNA reassociation percentages are different for the subspecies. Phenotypic differences in the chromogenic substrate degradation are also different for S. constellatus subspp. pharyngis and constellatus. A second group of beta-hemolytic group C and one strain of group G streptococci were also described in this report but not formally proposed as a subspecies (see column 1 of Table under S. anginosus). This group was closely related to S. anginosus and was not necessarily associated with pharyngeal infections. The reactions listed in Table in column 3 under S. constellatus subsp. pharyngis lists the tests to identify the subspecies. Streptococcus porcinus | The beta-hemolytic streptococci that carry Lancefield group E, P, U, and V and four new antigens were included in a description of S. porcinus . S. porcinus has been isolated from the human genitourinary tracts of female patients of reproductive age . The incidence and the importance of this finding are inconclusive and are complicated by the fact that most human isolates of S. porcinus react with commercially prepared GBS grouping antisera . Slide agglutination group B reagent has also been reported to react with nonhuman strains . Lancefield extracts of these strains generally do not react with CDC group B antiserum . Most strains isolated from human sources react with antiserum prepared against one of the new group antigens. Careful observation of the hemolytic reaction can be a useful indicator of S. porcinus, and clinical microbiologists should be aware of the potential for misidentification of these isolates as S. agalactiae. Typical strains of S. agalactiae lyse red blood cells weakly, and colonies of GBS are surrounded by a small zone of lysis. On the other hand, typical strains of S. porcinus lyse red blood cells in an agar plate in a much larger area around colony growth . Streptococcus iniae | The original cultures of S. iniae were isolated from freshwater dolphins. S. iniae is also found in aquacultures of fish and can be transmitted to humans via the fish . An outbreak of bacteremic disease in humans who handle contaminated fish has been reported , but sporadic reports of this organism have been rare at the CDC. Beta-hemolytic streptococci without group antigen should be tested for the phenotypic characteristics listed in Table for definitive identification. Streptococcus phocae and Streptococcus didelphis | As noted above, S. phocae and S. didelphis have not been confirmed as being isolated from humans. S. phocae has been isolated from seals , and S. didelphis has been isolated from opossums . The data for identification of these two beta-hemolytic streptococci were obtained from these published reports and our own testing of the type strains for these two species . TABLE 1 | Identification of the beta-hemolytic streptococci TABLE 2 | Identification of the S. anginosus group NON-BETA-HEMOLYTIC STREPTOCOCCI : As stated above, the determination of hemolysis is one of the most useful characteristics for the identification of streptococci. Nonhemolytic variants of S. pyogenes, S. agalactiae, and members of the S. anginosus group are well documented. Whether there are nonhemolytic variants of the other streptococcal species that are normally beta-hemolytic remains to be determined. It is not useful to try to distinguish between alpha-hemolysis and no hemolysis on blood agar plates. The composition of the medium including the type of blood and incubation atmospheres can influence whether "alpha" hemolysis occurs. There is no documented enzyme or toxin that affects red blood cells to produce alpha-hemolysis by streptococci. The greening or partial destruction of red blood cells is produced primarily by the production of hydrogen peroxide by the streptococci. The peroxide destroys some of the red blood cells and releases hemoglobin into the medium surrounding the streptococcal colony that appears green-like. If oxygen is removed from the growth atmosphere, peroxide is not formed, and cultures that appeared "alpha-hemolytic" will be nonhemolytic. The value of identification of Lancefield antigens on non-beta-hemolytic streptococci is also limited. While determination of the group D antigen for identification of S. bovis and determination of group R or other type antigens for S. suis are useful, determination of Lancefield antigens of all other non-beta-hemolytic streptococcal species, including the viridans streptococci, is of very little value for identification . Streptococcus pneumoniae | No changes in the classification have been made in the past few years for S. pneumoniae. This organism is still the leading cause of community-acquired pneumoniae, and the increasing prevalence of multidrug resistance is of great concern . Major developments in the formulation of vaccines for adults and children are encouraging for control of pneumococcal infections. Isolation and identification of S. pneumoniae is still problematic since conventional culture techniques are far from perfect and identification of some strains is refractory. Phenotypic physiologic tests place S. pneumoniae in the viridans Streptococcus, S. mitis group. Ideally, S. pneumoniae is identified by a serologic technique such as omni serum or a slide agglutination test or typing serum . Under most circumstances, if serologic techniques are not used, identification can be accomplished by determining susceptibility to optochin and/or bile solubility. Recent molecular techniques, such as PCR tests for targeted proteins, pneumolysin , and penicillin binding protein genes , have been shown to improve the identification of culture-negative cases of pneumonia and identification of strains. Streptococcus bovis Group: S. bovis, S. equinus, S. gallolyticus, S. infantarius, S. pasteurianus, S. lutetiensis | Table lists the streptococcal species included in this section. DNA-DNA reassociation experiments have clarified the taxonomic classification of this group of streptococci, but the nomenclature of these species is confusing and subject to debate. The reasons for the changes listed in Table for the S. bovis group begin with a report published in 1984 . These investigators examined a collection of strains, most of which were nonhuman isolates, and reported that the phenotypically described type strains of S. bovis and S. equinus were a single DNA group, i.e., the same species. Farrow et al. identified seven different DNA groups in their study. These DNA groups are referred to throughout this section. Other investigators performed DNA reassociation studies on human isolates termed S. bovis biotype I and biotype II (the latter is sometimes referred to as S. bovis variant) and discovered that neither biotype I nor biotype II of human origin was sufficiently closely related to the reference type strains of the S. bovis-S. equinus DNA group to be included in the newly described species . These investigators reported that the S. bovis biotype II isolates of human origin could be divided into two groups based on phenotypic characteristics determined in the Rapid Strep system. These two groups were called S. bovis II/1 and S. bovis II/2. The investigators also reported that the human isolates of S. bovis II/2 were closely related to the S. bovis I strains by DNA reassociation studies. Osawa et al. demonstrated that strains isolated from koalas, dogs, cows, and other animals belonged to the Farrow et al. DNA group 2 and proposed that these strains be called S. gallolyticus based on gallate-degrading capacity. These investigators also included isolates of S. bovis I and II from human infections in their studies. The human isolates of S. bovis I and II/2 joined the S. gallolyticus species based on DNA reassociation studies. Further studies confirmed that no human strains called S. bovis by current identification procedures did in fact join the S. bovis-S. equinus group described by Farrow et al. . All the human isolates of S. bovis I and II/2 were suggested to be officially identified as S. gallolyticus. These experiments were done primarily by whole-cell protein analysis, which correlates with DNA reassociation studies. Nelms et al. reaffirmed these studies using 16S rRNA probes . Recently, Schlegel et al. described yet another species, S. infantarius, and further suggested that two subspecies S. infantarius subsp. infantarius and S. infantarius subsp. coli could be identified. These strains were closely related to the S. bovis group. DNA reassociation experiments showed that these human isolates were unique and corresponded to DNA group 4 described by Farrow et al. . The majority of strains in these studies were isolated from humans. The cultures described by Coykendall and Gustafson and Knight and Shlaes as S. bovis II/1 were a major part of the DNA group, S. infantarius. Strains of S. bovis II/1 are distributed in both subspecies. Even more recently Poyart et al. have suggested that the strains identified as S. bovis II.2 be renamed S. pasteurianus and the strains identified as S. infantarius subsp. coli be named S. lutetiensis . These suggestions were based on a combination of DNA homology, whole-cell protein, and sequencing of the sodA gene. The importance of this change in nomenclature can be translated back to the association of colonic cancer and the isolation of S. bovis from blood cultures of these patients. Ruoff et al. showed that S. bovis I was more commonly associated with colonic cancer patients than was S. bovis II/1. This translates into correct terminology that S. gallolyticus and S. pasteurianus are isolated from blood cultures of patients with colonic cancer more often than is S. infantarius or S. lutetiensis. Whether this terminology is accepted in the medical literature will depend on whether clinical microbiologists adopt procedures that accurately identify each of the new species with the published criteria. The information in Table was summarized from the publications just cited . Hopefully, additional studies will be undertaken to verify these changes in the nomenclature of this group, because cultures previously identified as S. bovis are frequently found in blood cultures of patients with bacteremia, sepsis, and endocarditis. Streptococcus suis | S. suis is the name assigned to streptococci that were formally called Lancefield groups R, S, and T. Kilpper-Balz and Schleifer determined that the strains representing group, R, S, and T and several other serotypes of strains with the same phenotypic characteristics were a single DNA group, which they called S. suis. Although some strains are beta-hemolytic and produce a hemolysin on agar plates containing horse blood agar, all strains of this species are alpha-hemolytic when grown on blood agar plates containing sheep blood . Early reports also included results that some strains of this species contained streptococcal group D antigen. However, later reports indicated that the group R and group D antigens were similar but not identical and the observed reactions were judged as cross-reactions in the group D antiserum. The discovery of additional serotypes of this species led to a change in the nomenclature of how the capsular types were identified. There are now 35 different antigenic carbohydrate types of the species. Group R, the most common strain identified, is type 2, and group S is type 1. The important change to clinical microbiologists is that the only serotype identified from humans has been type 2 (group R). Difficulties with the Lancefield extraction procedure, which is the reference procedure used for demonstrating group antigens in all other streptococci, have led to the recommendation that a capsular reaction be used for identification of the various serotypes. This reaction is similar to the quellung reaction used to type S. pneumoniae. A modified Lancefield extraction may be used to identify type 2 by using group R-precipitating antiserum . Lutticken et al. reviewed the literature and summarized 44 cases of human infections caused by S. suis. There were 39 cases of meningitis and 5 cases of septicemia without meningitis. All cases were associated with the patient having worked with raw pork meat or having contact with pigs; none of the cases were from the United States. Trottier et al. reported the first case of infection cased by S. suis in North America; this was a case of human endocarditis from a Canadian patient . Type 2 S. suis cultures isolated from humans and pigs can be very similar. Chateillier et al. performed random amplification polymorphic DNA (RAPD) analysis of 88 strains of type 2 S. suis including 8 isolates of human origin. Twenty-three different RAPD types were identified. Five of the eight human strains were placed in RAPD type 1, which contained 26 cultures isolated from pigs. By this procedure, the strains isolated from humans may be identical to those causing infections in pigs. S. suis is found in infections among pigs in the United States . Why no infections caused by this organism have been documented among U.S. patients is unknown. It is possible that laboratories may not have the microbiological capacity to identify the isolates. Phenotypically, S. suis resembles the viridans streptococcal species S. gordonii, S. sanguinis, and S. parasanguinis and therefore may be misidentified. Careful microbiological examination of alpha-hemolytic viridans streptococci isolated from human cerebrospinal fluid may result in identification of S. suis in the United States. Typing antisera and group R antiserum can be obtained from the Statens Serum Institute in Copenhagen, Denmark. Viridans Streptococci | In 1997, Bruckner and Colonna listed 15 different streptococcal species that were included in the viridans streptococci . Table includes 26 streptococcal species that have the phenotypic characteristics of typical viridans streptococci. In addition to being catalase-negative, gram-positive cocci arranged in chains, all the species listed in Table have the phenotypic characteristics described for the viridans streptococci listed in Table . These characteristics are leucine aminopeptide positive, pyrrolidonylarylamidase negative, and no growth in 6.5% NaCl broth. Nearly all species, with the exception of those in the salivarius group, are bile-esculin negative. Species not yet identified from human sources are included because transmission of streptococcal species from nonhuman sources (S. iniae, S. porcinus, and S. suis) to humans has caused documented infections and there is no reason to believe that any of the species listed in Table will not be isolated from human infections. The 26 species are arranged according to the six phenotypic characteristics listed in Table . Note that with this system, most individual species cannot not be identified but are placed in one of the six groupings. Not all investigators agree with this grouping of streptococcal species. Some have preferred to group S. hyointestinalis in the pyogenic group and S. alactolyticus in the S. bovis-S. equinus group based on sequence data of the 16S rDNA gene and Fig. . Other investigators have included S. acidominimus and S. uberis in the viridans category . These two species are included in the category termed "other streptococci" because of their phenotypic characteristics (discussed below). The term "viridans streptococci" may not be the best to describe this group of organisms because many of the species do not express the hemolysis-like reaction (described as alpha-hemolysis, as discussed above) on blood agar plates. Many investigators refer to them as the oral streptococci; however, this designation does not truly represent the origin of all the species; some of them originate from gastrointestinal, vaginal, and dairy product sources. At this time, it is proposed that the term "viridans streptococci" be used to include streptococcal species with phenotypic characteristic described in Tables and . There are two excellent reviews that summarize the molecular experiments that define the majority of viridans streptococcal species . In addition, the corrected epithets of several species are given according to the rules of nomenclature . Interested investigators will need to devise a three-tier testing system for the definitive species identification. The first determination is described in Table , the viridans streptococci group; the second determination is described in Table , the viridans streptococcal species group; and the third level needs to be similar to that described in Table for the S. anginosus species group. A table similar to Table using chromogenic substrates should be constructed for each viridans species group to identify definitive species . All the phenotypic characteristics listed in Tables , , , and in this review were obtained from conventional tests. The tables in the papers by Whiley and Beighton and Ruoff et al. include most of the species listed in Table ; however, a few additional species were described after the publication of their reports. Streptococcus mutans Group | Whiley and Beighton give an excellant review of the species origin included in the S. mutans group with the exception of S. hyovaginalis. This species is included in the mutans grouping because of similar phenotypic characteristics . According to previous reports, the most common species isolated from human sources (primarily the oral cavity) are S. mutans and S. sobrinus; S. cricetus and S. ratti are rarely isolated . The identification scheme devised by these authors for differentiation of S. mutans and S. sobrinus included the fermentation of melibiose; S. mutans is positive and S. sobrinus is negative. In a previous study, we found that 88% of blood culture isolates (48 of 54) were melibiose positive, which indicates that the majority of isolates from human noncaries infections are S. mutans . S. hyovaginalis was described by Devriese et al. in 1997 . Several strains of this species were isolated from the genital tracts of female swine and appear to be part of the normal flora. Identification from human sources has not been documented. Streptococcus salivarius Group | The S. salivarius group of bacteria is closely related to the S. bovis-S. equinus-S. gallolyticus group by both 16S rRNA gene analysis and phenotypic characteristics (Tables and ). S. alactolyticus is included in the "S. bovis" group by some investigators . Some strains of each of the species listed in the S. salivarius group react with streptococcal group D grouping antiserum, which would seem to indicate that they possess the group D antigen. This may not be the case. The problem with streptococcal group D antiserum is that this serum is rarely if ever tested for cross-reactions with non-beta-hemolytic streptococci. The Streptococcus grouping antisera are made for testing beta-hemolytic strains, and such sera are evaluated only with preparations of strains that are beta-hemolytic; therefore, group D antiserum may contain antibodies that cross-react with non-beta-hemolytic streptococci. Antigenic homology between strains of Enterococcus faecalis and members of the S. Salivarius group with streptococcal group D antiserum has not been published. Of the species listed in the S. salivarius group, only strains of S. salivarius, S. vestibularius, and S. infantarius have been isolated from humans. S. salivarius is commonly identified from a variety of human infections . S. vestibularius was identified from the human oral cavity , and its association with human infections has not been confirmed. As discussed above, S. infantarius has been isolated from human infections . S. infantarius is listed in both the S. bovis group and the S. salivarius group because some S. infantarius strains are bile-esculin and esculin negative, which excludes them from the S. bovis group. These bile-esculin-negative strains were identified as S. infantarius by DNA-DNA reassociation studies. These strains are misidentified as S. bovis II/1 by the Rapid Strep system. Additional confusion will result in light of the latest findings that S. bovis II/1 can be either S. infantarius or S. lutetiensis. Accurate identification of these four species will establish their role in human infections. S. hyointestinalis isolated from the swine gut, was originally identified as S. salivarius; however, DNA-DNA reassociation studies demonstrated that this organism was unique and deserved species status . The taxonomic status of S. thermophilus has been in question for several years, and some investigators proposed that S. thermophilus should be a subspecies of S. salivarius. However, Schleifer et al. conducted DNA-DNA reassociation experiments under stringent conditions and determined that these strains deserved full species status . Data from the references noted in this section should be compiled to form a identification scheme for species identification. Streptococcus anginosus Group | The species included in the S. anginosus group have been discussed in the section describing beta-hemolytic streptococcal species. Streptococcus sanguinis Group (Formerly Known as S. sanguis) | As stated above, some investigators have preferred to arrange the viridans Streptococcus species differently from the classification listed here. For example, some have included the S. sanguinis and S. mitis groups together based on 16S rRNA sequence analysis , Fig. , and phenotypic characteristics . According to the phenotypic characteristics listed in Table , these two groups can be separated for the most part by their reactions in the arginine and esculin tests. There are exceptions. S. gordonii was split from the S. sanguinis species . Although there are many similarities between the two species, they can be differentiated by immunoglobulin A1 IgA1 protease, amygdalin, and alkaline phosphatase reactions . S. parasanguinis was split from the S. mitis grouping and hence has some characteristics of both the S. sanguinis and S. mitis groups. The authors described characteristics that helped to differentiate S. parasanguinis from S. sanguinis, but the differentiation was based on percent positive reactions in 14 different tests; inulin fermentation was the only test that was positive for all strains of S. sanguinis and negative for all strains of S. parasanguinis. Formation of the extracellular polysaccharide dextran on 5% sucrose agar by both S. sanguinis and S. gordonii but not by S. parasanguinis also helps to differentiate these species. All three of these species are commonly found in viridans streptococcal type infections. S. sanguinis is probably the most frequently identified viridans Streptococcus species from patients with subacute bacterial endocarditis. It is difficult to assess the association of any of the newly described and redefined viridans streptococcal species with specific human infections. Streptococcus mitis Group | The S. mitis group of viridans streptococci has caused considerable confusion for both clinical microbiologists and taxonomists. One of the problems has been the use of invalid species names such as Streptococcus viridans and S. mitior. Although these names may convey a meaning to physicians managing patients, they are troublesome to taxonomists in that there are no official reference strains which to study. Another problem has been the use of the terminology "biotypes." For example, "S. sanguis biotype II" is a common designation . The biotype describes strains with certain phenotypic characteristics but does not have official taxonomic status, and therefore no "type" strains are usually available for study. In addition, the type strain for S. mitis was determined to be inappropriate based on molecular experiments. The addition of S. oralis and S. cristatus , formally called S. crista, added to the complexity of the S. mitis group. Beighton et al. presented a reasonable scheme based on 23 tests (including 10 chromogenic tests) to differentiate the S. sanguinis and S. mitis groups recognized in 1991 . Since then, two more species, S. peroris and S. infantis, have been added to the S. mitis group . S. mitis, S. oralis, S. cristatus, S. peroris, and S. infantis have all been isolated from the human oral cavity. Even more recently, Streptococcus orisratti has been added to the S. mitis group. S. orisratti possesses Lancefield's group A antigen, and this phenotypic characteristic can be used to identify this species once it is determined that the strain in question has phenotypic characteristics consistent with the S. mitis group. Devising a scheme for identification of all six species in the S. mitis group will take additional work. It is not yet possible to correlate the new and revised species in this group with human infections. Whether any of the newly described species are associated with human infections is unknown. Options for Identification of Viridans Streptococcal Species | What are the options for identifying the viridans streptococcal species? It is apparent that the use of conventional tests such as those used at CDC does not differentiate most species. Nearly 40 conventional tests have been used to test all the type strains as well as other reference strains of each species, and successful differentiation of species has not been accomplished. Ruoff et al. in the last two editions of the Manual of Clinical Microbiology included identifications tables that could be used to identify the viridans streptococcal species known at that time . These tables included conventional tests as well as a series of fluorogenic substrates that had to be prepared in house because they were not available from commercial medium sources. Commercially available systems do not have all the species in their databases, so that the identifications generated by those systems are compromised, at least to some extent. Probably the best one could hope for was reported by Kikuchi et al. . When DNA homology studies were used to confirm the identifications, 87% of 156 strains of 13 different species in the database were correctly identified and 56% of 15 strains of 6 species not in the database were incorrectly identified using the Rapid ID 32 Strep system (bioMerieux). In the latter case, a result of "unidentified" would be correct. Unfortunately, DNA-DNA reasociation procedures at this time are not applicable for most clinical laboratories. 16S rRNA gene-sequencing procedures have been used to show relationships between many of the streptococcal species; however, strict criteria for determining species have not been established. For example, the sequence identities of the 16S rRNA genes for the types strains of S. mitis, S. oralis, and S. pneumoniae are greater than 99% similar . This does not leave much room for sequence variability that occurs in many strains. In addition, S. macedonicus, S. wais, and S. gallolyticus appear to be separate species based on 16S rRNA sequences ; however, DNA-DNA reassociation and sodA sequencing studies indicate that the type strains representing the species are a single species, S. gallolyticus . Rudney and Larson used restriction fragment polymorphism of rRNA ribotyping for six species and found that 91% of 53 isolates were reported as one of three species. Whole-cell protein analysis is thought to correlate very well with DNA-DNA reassociation, but Vandamme et al. did not get correlation between their whole-cell protein profiles and species identification. This is in contrast to the results obtained by most other authors using this technique but would seem to point to a need for a standardized procedure. Garnier et al. targeted the d-alanine-d-Alanine ligases of six viridans species, applied PCR reaction technology with specific primers, and successfully identified the six species. However, they were unable to differentiate the S. anginosus group of species. Two investigations have been reported examining the tDNA intergenic spacer length polymorphism (tDNA-PCR), with different results . These investigators used different detection systems, which may explain the differences in the results; however, neither report indicated a completely successful result in differentiating all species tested. The most successful molecular technique reported to identify viridans streptococcal species to date is that reported for examining the sequence of the gene encoding the manganese-dependent superoxide dismutase (sodA int) . Poyart et al. reported differentiation of 29 streptococcal species including 16 viridans species. S. mitis, S. oralis, and S. pneumoniae were clearly differentiated by sodA sequencing, as opposed to the situation for 16S rRNA gene sequencing. Although other technologies that may be used to study strain-to-strain relationships for epidemiologic studies include electrophoretic isoenzyme typing , pulsed-field gel electrophoresis , and various other PCR-based methods , these procedures are not useful for the identification of specific species. Antimicrobial Susceptibilities of Viridans Streptococcal Species | Viridans streptococci are isolated from a variety of infections but most significantly from patients with subacute bacterial endocarditis and from neutropenic patients with cancer. In these two instances, knowledge of the antimicrobial susceptibility of the infecting organisms is useful for good patient management. The question is whether knowledge of the taxonomic identity of infecting bacteria is useful for predicting the antimicrobial susceptibility of the organism. In view of the multiple changes in taxonomy and nomenclature of this group of gram-positive cocci, it is difficult to say exactly if it is possible to predict potential susceptibility problems related to specific species. However, some generalities can be made regarding susceptibilities and species. First of all, antimicrobial resistance is substantial in the viridans streptococci as a group. Penicillin resistance in the viridans streptococci is as high as 48% in strains from the United States, 45% in strains from Canada, and 33% in strains from Latin America . Specific species in this study were not adequately identified, but S. mitis was the most common species group identified and the most likely to be resistant. Other studies with better identification procedures have indicated that S. mitis and S. oralis strains are the most common found in blood cultures of cancer patients and are commonly resistant to beta-lactam antimicrobials . Although not all these investigators used identification procedures that would allow identification to the species level according to today's standards, the identification certainly would be in the S. mitis grouping. The species distribution is somewhat different in endocarditis patients; S. sanguinis, S. oralis, and S. gordonii, in descending order, are the most common strains isolated from cultures of blood of endocarditis patients . Although the incidence is not as high in the S. sanguinis group as in the S. mitis group, antimicrobial resistance is present. One other generality can be made with the S. anginosus group. Resistance to beta-lactam antimicrobial was only at the intermediate level, but resistance to macrolides was found in 17% of the strains . More recently, Tracy et al. used 16S rRNA sequencing to identify the species of the S. anginosus group and found no difference in antimicrobial susceptibilities of the three species (S. anginosus, S. constellatus, and S. intermedius). These authors expressed the opinion that it was unnecessary to identify the infecting organism to the species level and that identification to the "milleri group" was sufficient for patient management. Whether the same can be said for the other viridans groups remains to be established. There appears to be antimicrobial susceptibility differences in some of the other species within these groups. TABLE 3 | Identification of non-beta-hemolytic gram-positive cocci in chains TABLE 4 | Identification of major groups of viridans Streptococcus species TABLE 5 | Identification of Abiotrophia and Granulicatella species NUTRITIONALLY VARIANT STREPTOCOCCI, ABIOTROPHIA AND GRANULICATELLA SPECIES : The nutritional variant group of gram-positive bacteria has been described as satelliting, pyridoxal-dependent, vitamin B 6-dependent, cell wall-deficient (L-form), and finally nutritionally variant streptococci (NVS) before being reclassified as Streptococcus adjacens and Streptococcus defectivus . Kawamura et al. demonstrated that the two NVS species were phylogenetically distant from the streptococci and proposed that they be given new genus status, Abiotrophia adjacens and Abiotrophia defectivus . Recently, three new species have been added to the Abiotrophia genus, A. elegans , A. balaenopterae , and A. para-adiacens . Collins and Lawson further proposed that some of the Abiotrophia species were phylogenetically distinct from each other and proposed the establishment of the genus Granulicatella to include G. adiacens, G. balaenopterae, and G. elegans in this new genus while A. defectiva remains in the Abiotrophia genus . Note the corrected epithets. Table includes all the species of "NVS" and an identification scheme based on the published reports and our own results examining 100 strains of NVS taken from the CDC stock culture collection . Our results indicate that of 101 isolates from 97 patients (58 with endocarditis), 55 were G. adiacens, 43 were A. defectiva, and 3 were G. elegans. Other authors using slightly different species identification criteria found 15 G. adiacens, 13 G. para-adiacens, 9 A. defectiva, and 8 G. elegans strains among 45 endocarditis patients . NVS overall is reported to cause approximately 5% of all cases of endocarditis . These reports indicate that all the species except G. balaenopterae have been isolated from human infections. Patients with endocarditis due to NVS are more difficult to treat than those infected with viridans streptococci. As many as 41% of patients may fail to respond to antimicrobial therapy, and combination therapy is often recommended . UNUSUAL STREPTOCOCCUS SPECIES AND OTHER GRAM-POSITIVE COCCI IN CHAINS : The last group of Streptococcus species and related genera consists of bacterial strains that do not fit into any of the species or species groups discussed above. None of these strains are beta-hemolytic, which eliminates them from the species listed in Table . Note that in Table , the last species-group listed is "other streptococci and genera." The species listed in Table are not viridans streptococci, because all viridans streptococci are leucine aminopeptidase positive and pyrrolidonylarylamylase-negative and do not grow in 6.5% NaCl broth. The species in Table have at least one reaction different from this pattern. These species are bile insoluble and optochin resistant (eliminating the pneumococci) and do not have defined group antigens (eliminating the S. bovis and S. suis groups). The identifications based on the reactions listed in Table are a guide for potential identification. The majority of reactions listed for the species S. acidominimus, S. pluranimalium, S. thoraltensis, S. uberis, S. parauberis, and S. urinalis were taken from published materials . In most cases, only the type strain for each species has been tested in our laboratory. In addition, none of these six species has been isolated from human infections. Streptococcus acidominimis | The type strains of S. acidominimis have been poorly described, and most reference strains have not had the phenotypic characteristics of the original description of the species in 1922. Recently, Devriese et al. examined the reference strains, firmly established the correct type strain for the species, and described better phenotypic criteria for identification . Although it had been reported that S. acidominimus was identified from humans in 1977 , we have recently revised that finding. The human isolates previously identified as S. acidominimus are now identified as Facklamia sourekii . The reactions listed in Table are those obtained from Devriese et al. (-) and our testing of the current type strain for the species. At this time, we have not confirmed any human isolates of S. acidominimus when using the revised identification criteria. Streptococcus pluranimalium | The new species S. pluranimalium was described by Devriese et al. . The strains resembled S. acidominimus, and in fact some of the reference strains of S. acidominimus in culture collections were reidentified as S. pluranimalium. S. pluranimalium has been isolated from bovine mastitis; bovine vagina, cervix, and tonsils; and canary lung and lesions . No human isolates have been confirmed. Presumptive identification can be made on the basis of the reactions listed in Table . Streptococcus thoraltensis | The new species S. thoraltensis was also described by Devriese et al. . Cultures of this species were recovered from the intestinal tracts of swine. Identification of this species from humans has not been documented. Presumptive identification can be made by matching reactions with those listed in Table . Streptococcus uberis and Streptococcus parauberis | The two species S. uberis and S. parauberis are placed in Table because differentiation of the two species by conventional phenotypic tests is not possible. Species-specific probes have been developed for identification of both species . S. uberis can be found in up to 20% of cases of bovine mastitis . Confirmation of isolation S. uberis or S. parauberis from human infections has not been documented. All the isolates of S. uberis previously reported from humans have been reidentified as Globicatella sanguinis (see below). Streptococcus urinalis | The description of S. urinalis was made on the basis of a single strain isolated from a case of human cystitis . Since the publication of that report, a second isolate of S. urinalis has been identified, a blood culture isolate with no additional information. Phenotypically, S. urinalis is similar to S. uberis, G. sanguinis, and enterococci. In addition to the tests described in Table , isolates should be confirmed as not being enterococci by use of the GenProbe Enterococcus probe because S. urinalis has overlapping phenotypic characteristics with the enterococci. Dolosicoccus paucivorans | D. paucivorans was initially described based on one organism isolated from a culture of blood from a patient with pneumonia . Two additional cultures have been isolated from blood samples. One patient had a diagnosis of pneumonia, and the other report did not include clinical information. The appearance of dolosicocci on blood agar plates is similar to that of the viridans streptococci, but the positive pyrrolidonylarylamidase and negative leucine aminopeptidase reactions clearly differentiate this bacterium from the viridans streptococci . Facklamia species and Ignavigranum ruoffiae | Five species of Facklamia have been described, F. hominis, F. ignova, F. languida, F. sourekii, and F. tabacinasalis. F. languida is described as not forming chains; however, in our hands some chains are apparent in Gram stains prepared from thioglyolate broth . All but F. tabacinasalis have been isolated from human infections. Most isolates have been from blood cultures, but very little clinical information was provided with the isolates . Table lists the tests necessary for presumptive identification of the Facklamia and Ignavigranum genera. Identification of the species has been recently reviewed . Some strains exhibit decreased susceptibilies to the beta-lactam and macrolide antimicrobials . Globicatella sanguinis | G. sanguinis (sanguis) was described in 1992 from a collection of cultures from human sources identified as Streptococcus uberis-like . They were called S. uberis-like because of phenotypic similarities. Isolates of this species are also similar to Aerococcus viridans in that both species are leucine aminopeptidase negative and pyrrolidonylarylamidase positive and hydrolyze hippurate. However, the cellular arrangement of A. viridans in the Gram stain is in pairs and clusters, as opposed to chains (G. sanguinis). G. sanguinis is differentiated from S. uberis by the leucine aminopeptidase test (S. uberis is positive) and the bile-esculin test (S. uberis is negative, but G. sanguinis is positive). In a recent study of 28 strains of G. sanguinis, 20 were isolated from cultures of blood from patients with diagnoses including bacteremia, sepsis, and endocarditis. Urinary tract infections were associated with three urine culture isolates . These data seem to indicate that this species may be an important pathogen in humans. G. sanguinis can be identified using the phenotypic tests listed in Table . New Species Not Yet Tested | Five new Streptococcus species, all of which appear to fall into the viridans Streptococcus grouping, and one Facklamia species have been described that have not yet been tested in our conventional test system. S. australis appears to fit into the S. mitis grouping, and because it deaminates arginine, it most closely resembes S. cristatus. S. australis was isolated from the human oral cavity. S. ovis , isolated from vaginal swabs of sheep, perhaps fits into the S. mutans group. S. ovis is reported to form acid in mannitol and sorbitol broths like the other members of this group. S. sinensis appears to fit into the S. anginosus group by both phenotypic and genetic tests. S. sinensis was isolated from a human with endocarditis. S. entericus , isolated from cattle intestines, phenotypically fits roughly into the S. mitis group. Yet another recently described species, S. gallinaceus , isolated from chickens phenotypically resembles species in the S. anginosus group, because the cultures are mannitol positive, a trait rarely observed in the S. anginosus group, they may be identified fairly easily. F. miroungae was isolated from an elephant seal. Phenotypic tests indicate that this species is closely related to F. hominis. An exact position of any of these new species cannot be determined until they are tested with conventional tests, and this has not been reported by the authors of these reports. TABLE 6 | Identification of unusual Streptococcus species and other genera of gram-positive cocci in chains CONCLUSIONS : The changes in the nomenclature and taxonomy of the Streptococcus genus are numerous and varied. Eleven species and four subspecies of streptococci are beta-hemolytic and can be identified by Lancefield grouping and a few phenotypic tests. Nearly all these species and subspecies are isolated from human infections. However, among the non-beta-hemolytic species, 26 different species of viridans streptococci, 5 different species of what was termed S. bovis, 5 different species of nutritionally deficient streptococci, 9 different species of other streptococci, and 3 new genera of gram-positive cocci in chains have been described. The majority of changes in the non-beta-hemolytic streptococci have been the addition of species. These additions for the most part have complicated the recognition of each specific species. It is apparent that conventional tests and commercially available devices and systems have not incorporated all the taxonomic changes into the identification procedures. In my opinion, the development of molecular genetic procedures is required before the non-beta-hemolytic streptococci can be accurately identified. Backmatter: PMID- 12183232 TI - Isoniazid-Induced Transient High-Level Resistance in Mycobacterium tuberculosis AB - An American Type Culture Collection reference strain and eight clinical strains of Mycobacterium tuberculosis, all of which were susceptible to isoniazid (INH) (mean MIC, 0.06 mg/liter) and negative for the Ser315Thr katG mutation, were left in their BACTEC 12B vials (for use with the BACTEC 460-TB method) containing 0.1 mg of INH per liter for periods of up to 28 days after the completion of the antibiotic susceptibility test. Each eventually grew to levels compatible with those of INH-resistant strains. Successive passages in INH-containing BACTEC 12B vials and onto solid media showed that the resistance noted above was maintained. Successive passages of these M. tuberculosis strains in which INH resistance had been induced into BACTEC 12B vials or solid media containing stepwise increases in INH concentrations eventually yielded organisms resistant to 20 mg of INH per liter. Transfer of cells in which INH resistance had been induced to drug-free medium followed by repeated passages in that medium eventually yielded organisms whose susceptibility to INH was identical to that of the original parent strains. The cycle of induced INH resistance could be repeated with these now INH-susceptible cells. The use of M. tuberculosis identification probes and IS6110-based restriction fragment length polymorphism analyses of cultures throughout the induction of INH resistance and the reversal of resistance in drug-free medium eliminated the possibility that the culture was contaminated or that the initial specimen had a mixed type of infection. Induced high-level resistance to INH (20 mg/liter) could be reduced 100-fold with a subinhibitory concentration of reserpine but not with verapamil. These results collectively suggest that high-level resistance to INH can be induced in INH-susceptible M. tuberculosis strains by the induction of a reserpine-sensitive efflux mechanism. Keywords: Introduction : Isoniazid (INH) is highly effective against Mycobacterium tuberculosis , having MICs that range from 0.02 to 0.06 mg/liter . However, the rates of resistance to INH have been escalating worldwide . The mechanism by which the resistance of M. tuberculosis to INH takes place is very complex and has been ascribed to the individual mutation of genes encoding a catalase-peroxidase (katG) , an enzyme of the mycolic acid pathway (inhA) , and a beta-ketoacyl-acyl carrier protein synthetase (kasA) . Because only about 50 to 60% of all strains studied so far have mutations in any of these genes associated with resistance to INH , conventional wisdom dictates that other genetic factors bestowing resistance to INH must be involved. To this extent, resistance to INH has also been related to its neutralization by the overproduction of arylamine N-acetyltransferase , by limitations of NAD+-binding proteins , and by the overexpression of antioxidant enzymes that compensate for the loss of function of the KatG protein . The first and last mechanisms still fail to account for the resistance in at least 20% of all INH-resistant strains . Plasmids play no role in acquired INH resistance in M. tuberculosis . Resistance to INH is not totally accounted for by the mechanisms described above, since recent evidence shows that resistance to INH in Mycobacterium smegmatis can also be mediated by an energy-dependent efflux pump . Efflux pumps that render an organism resistant to one or more antibiotics have been identified in all bacteria that have been subjected to such analyses, including mycobacteria . These pumps are normally present in the plasma membrane of the bacterium and afford protection from toxic substances that are recognized as substrates by the pump. Susceptibility to a given antibiotic that is a coincidental substrate for a pump may be considered the result of overcoming the effectiveness of the pump due to its limited number of units. Because among the isolates of one strain there are members that have different quantitative susceptibilities to a given antibiotic attributed to genotypic variation , it is possible that these differences may be due to a variation in the number of existing efflux pumps. Furthermore, these variations may be due to a mechanism that controls the number of such pump units. The demonstration of antibiotic efflux pumps in mycobacteria has raised the obvious question of whether these or any antibiotic efflux pump is subject to an increase in their number, which in turn renders the organism increasingly resistant to the antibiotic. To this extent we have attempted and succeeded to induce INH sensitivity in M. tuberculosis strains increasingly resistant to INH by mechanisms that do not involve the selection of mutations that bestow resistance to this antibiotic. The mechanism by which this induced resistance to INH takes place might involve an efflux pump. MATERIALS AND METHODS : Antibiotics and chemicals. | INH, chlorpromazine (CPZ), verapamil, and reserpine were purchased from Sigma Aldrich Quimica SA, Madrid, Spain. Rifampin, streptomycin, and ethambutol as well as all materials and equipment used to test the antibiotic susceptibilities of the M. tuberculosis strains tested with the BACTEC 460-TB system were purchased from Becton Dickinson Diagnostic Instrument Systems, Sparks, Md., and were prepared according to the recommendations of the manufacturer. Solid and liquid media used for purposes other than susceptibility testing are identified in the legends to the figures or in the tables and were purchased from Difco, Detroit, Mich. All solutions were freshly prepared on the day of any given experiment. Microorganisms. | M. tuberculosis H37Rv (strain ATCC 27294 from the American Type Culture Collection ([ATCC]) and eight clinical isolates of M. tuberculosis, the latter of which were isolated from untreated patients and previously shown to be susceptible to INH, rifampin, streptomycin, and ethambutol by the BACTEC 460-TB system (S. Siddiqi, BACTEC TB system, product and procedure manual, 1980, Becton Dickinson Diagnostic Instrument Systems, Towson, Md.), were selected for this study and were designated INH-susceptible strains. Drug susceptibility assays by the BACTEC 460-TB radiometric proportion method and determination of INH MICs by the broth (BACTEC) radiometric method and the agar dilution method. | The indirect drug susceptibility assay (the BACTEC 460-TB radiometric proportion method) was conducted with the respective desired final concentrations of INH (0.1 mg/liter), rifampin (2 mg/liter), streptomycin (6 mg/liter), and ethambutol (7.5 mg/liter) by the procedure and with the interpretive criteria recommended by the manufacturer (; S. Siddiqi, BACTEC TB system, product and procedure manual, Becton Dickinson Diagnostic Instrument Systems) and by Rastogi et al. . Absolute control vials (with no drug) and drug-containing vials were inoculated with approximately 105 to 106 CFU of an organism identified as M. tuberculosis by rRNA-DNA probe hybridization (AccuProbe; Gen-Probe-BioMerieux, Lyon, France) . A separate 1/100 dilution of each bacterial suspension served as the proportional control. All vials were incubated at 37C and were evaluated daily until the growth index of the proportional control reached 30, which was within 4 to 12 days. The MIC of INH was determined for the H37Rv (ATCC 27294) control strain and clinical strains by the modified broth radiometric method of Lee and Heifets (; S. Siddiqi, BACTEC TB system, product and procedure manual, Becton Dickinson Diagnostic Instrument Systems). The MIC was defined as the lowest INH concentration that completely inhibited the generation of 14CO2 after an incubation period of 12 days. The determination of the INH MIC by the agar dilution method was performed as described previously . The simultaneous use of Middlebrook 7H11 solid medium for susceptibility testing and MIC determination was necessary because the agar dilution method, the "gold standard" in mycobacteriology , affords accurate determination of the percentage of resistant bacteria in the population present in the broth used by the radiometric method as well as for the isolation of single colonies. Both methods yielded comparable MICs. Inducement of INH resistance, reversal of resistance, and repetition of the cycle of INH-induced resistance. | The vials that were used to determine the susceptibilities of the clinical and INH-susceptible (ATCC 27294) strains to INH (0.1 mg/liter) and that would normally be discarded after a maximum of 12 days (S. Siddiqi, BACTEC TB system, product and procedure manual, Becton Dickinson Diagnostic Instrument Systems) were retained in the incubator for an additional period of time. By the end of 28 days, all of the vials yielded visible and radiometric evidence of growth. The contents of these vials were identified as M. tuberculosis by conventional biochemical methods and rRNA-DNA probe hybridization and were evaluated for INH susceptibility by the BACTEC 460-TB and solid medium methods. IS6110-based restriction fragment length polymorphism (RFLP) analysis (IS6110-RFLP analysis) confirmed that each of the bacterial populations that could now grow in the presence of 0.1 mg of INH per liter was identical to the initial parent population. The MIC of INH for each of these strains was determined and compared to that of the initial respective parent, and aliquots of 0.1 ml from growing cultures containing a concentration of INH just below the MIC were inoculated into fresh BACTEC 12B vials containing the lowest INH concentration that had completely inhibited the growth of the previous population. At the end of 10 to 28 days these vials showed evidence of maximum growth, and a new INH susceptibility test and MIC determination confirmed their capacity to grow in the presence of higher concentrations of INH that had previously been shown to inhibit growth. This sequential process was conducted for five additional consecutive passages, with each passage containing a stepwise increase in the concentration of INH corresponding to the lowest concentration that had inhibited growth in the previous passage. This process was conducted in parallel with solid medium. The latter method provided colonies that could be assayed for identity and antibiotic susceptibility and colonies that could be assayed by other procedures to ensure that what was being observed with liquid medium was also taking place in solid medium. This phase of the study consumed an average of 230 days (see Fig. ) with the radiometric broth medium and yielded M. tuberculosis isolates that were capable of growing in medium containing a concentration of INH as high as 20 mg/liter. The potency of the INH in the medium was assayed throughout the period described above after the removal of mycobacteria by filtration (filter pore size, 0.22 mum; Millipore), and the cell-free medium was reinoculated with the control strain, H37Rv (ATCC 27294). After the sixth passage (when the strains were growing in INH at 20 mg/liter), each derived INH-resistant M. tuberculosis isolate obtained from the initially INH-susceptible strains was subsequently transferred to INH-free medium and incubated for an extended period of 28 days. An aliquot from each vial was then transferred to fresh INH-free medium, and these vials were reincubated for the same extended period; three more consecutive passages in drug-free medium were conducted for each derived strain. The INH MIC for each strain was determined after each passage in drug-free medium. After the fourth passage the susceptibility testing criteria for INH susceptibility were verified by procedures conducted with broth and solid media. Aliquots from the last passage in drug-free medium were subjected to the same antibiotic susceptibility assay described above for the respective initial strains; i.e., they were allowed to remain in the incubator well past their period of antibiotic susceptibility until evidence of growth. The contents were subjected to the same procedures used for the eventual inducement of high-level resistance to INH described above. IS6110-RFLP analysis and analysis of katG mutations by PCR-RFLP analysis. | Before and after the inducement of resistance to INH, strain H37Rv (ATCC 27294) and the clinical strains were subjected to IS6110-RFLP analysis to determine possible contamination or to detect an initial mixed culture, as described previously . Determination of the most common point mutation in katG, consistently found at position 315 (serine to threonine [S315->T]) in clinical isolates with high-level INH resistance, was conducted with the initial ATCC control strain, representatives of the initial INH-susceptible clinical strains, representatives of the progeny in which INH resistance had been induced and obtained after the sixth consecutive passage in INH, and representatives obtained from the last (fourth) consecutive passage in the absence of INH, as described previously . Two mutants previously found to have katG mutations were similarly analyzed and served as positive controls. MICs of putative inhibitors of efflux pumps for the strains of M. tuberculosis used in this study and effects of the inhibitors on induced INH resistance. | The MICs of CPZ, verapamil, and reserpine for strain ATTC 27294 and the clinical strains of M. tuberculosis were determined with the BACTEC 460-TB system described above, except that INH was replaced by separate various concentrations of known inhibitors of efflux pumps: CPZ at 1.25 to 40 mg/liter, verapamil at 8 to 5,000 mg/liter, and reserpine at 0.01 to 160 mg/liter. These inhibitors of efflux pumps were separately added to cultures at concentrations that were previously shown to have no effect on the growth of the initial ATCC and clinical strains of M. tuberculosis used in this study. The concentrations of these compounds used in combination with INH, as well as the concentration of INH, in cultures containing representatives of the initially INH-susceptible strains, representatives from cultures induced to have high levels of resistance to INH, or representatives from cultures whose progeny had reverted in drug-free medium to their initial susceptibility to INH are described in the figure legends and in the tables, where appropriate. FIG. 2. | Time course of induced INH resistance and reversion of the H37Rv (ATCC 27294) reference strain. Time course of induced INH resistance and reversion of the H37Rv (ATCC 27294) reference strain. Culture of INH-susceptible isolates (time zero; MIC, 0.03 mg/liter) in INH-containing medium (0.1 mg/liter) incubated beyond the antibiotic susceptibility assay time frame (4 to 12 days) resulted in evident growth after 28 days. The ability to grow in 0.1 mg of INH per liter was confirmed, the MICs were determined, and the cells were designated as having induced INH resistance (MIC, 0.2 mg/liter) at IL1. These cells were then exposed to INH at concentrations that corresponded to the MIC for the previous passage phase, and after 21 days the growing cells were then tested as described above and the cells were designated as having induced INH resistance (MIC, 0.5 mg/liter) at IL2. The process was continued to IL7, at which time the cells were able to grow in the presence of 20 mg of INH per liter (MIC, 40 mg/liter). Cells at IL7 were repeatedly transferred to drug-free medium. The MICs were determined between passages, and the MICs gradually decreased (IL8, IL9, and IL10), reaching the MIC for the initial parent strain (0.06 mg/liter) at IL11. The cells present in the culture at IL11 served as a source for the repeat cycle of induced resistance to INH (data not shown). RESULTS : Repetition of the antibiotic susceptibility assay with M. tuberculosis H37Rv (ATTC 27294) and the eight clinical strains with the BACTEC 460-TB system indicated that each of these acid-fast-positive strains was susceptible to 0.1 mg of INH per liter as well as to rifampin, streptomycin, and ethambutol. Extending the incubation period for INH-susceptible cultures in INH-containing medium beyond the completion of the antibiotic susceptibility assay yielded evidence of visible growth. By the end of 28 days after the beginning of the INH susceptibility assay, the growth in all of the cultures with initially susceptible isolates reached a level comparable to that for the individual controls grown in drug-free medium . Direct examination of the contents of each vial by the use of identification probes and staining for acid-fast bacilli indicated that all of the organisms were M. tuberculosis. However, unlike their parents, which were all positive by staining with the Ziehl-Neelsen acid-fast stain, the cells from any given INH-containing vial presented with wide variations in acid-fast staining, ranging from negative to positive staining status . This variation in acid-fast staining is probably due to some inhibition of cell wall mycolic acid synthesis . Given that INH resistance is associated with the loss of catalase activity , the contents of the INH-containing vials were tested for this biochemical characteristic. All retained catalase activity . The possibility that the potency of INH in the vials had been affected by prolonged incubation was ruled out since the medium, when filtered free of cells, was as effective as freshly made INH-containing medium against M. tuberculosis isolates with proven INH susceptibility . Inducement of INH resistance. | Once the possibility of contamination and the possibility that the INH-containing medium had reduced potency were ruled out, aliquots of each of the cultures that were then growing in the presence of INH were subjected to a repeat INH susceptibility assay with the BACTEC 460-TB system and solid medium. Separate determinations of the MIC of INH by the modified broth radiometric method yielded similar data (data not shown). For a better understanding of the INH MICs, the values presented in the text and tables refer only to the INH concentrations used with the modified broth radiometric method of Lee and Heifets (; S. Siddiqi, BACTEC TB system, product and procedure manual, Becton Dickinson Diagnostic Instrument Systems). The results presented in Table show that at the first stage of resistance to INH by initially INH-susceptible strains (induction level 1 [IL1]) these organisms were indeed resistant to INH (0.1 mg/liter) and grew on INH-containing agar (0.2 mg/liter). The MICs of INH for these strains were much higher than those for the parents used at the beginning of this study . Aliquots of these cultures with INH-induced resistance at IL1 were transferred to fresh BACTEC 12B vials containing a concentration of INH that corresponded to the new MIC (0.2 mg/liter). These cultures reached full growth by the end of 10 to 21 days and were subjected to INH MIC determination, after which time they served as sources for the next passage in medium containing a concentration of INH that corresponded to that which was shown to be the new INH MIC for the strains in the previous passage. These organisms, referred to as isolates with INH-induced resistance at the second induction level (IL2), were assayed for determination of the INH MIC (the INH MIC at IL2 was then 0.5 mg/liter) and were used as sources for five additional serial passages, with each succeeding passage containing a concentration of INH that corresponded to the previous MIC for the strain with INH-induced resistance at IL2, IL3, IL4, IL5, IL6, and IL7. These passages were also conducted with solid medium containing equivalent concentrations of INH, and the plates were evaluated for the numbers of CFU. Figure describes the increases in the INH MICs for each of the induction levels when the ATCC strain was subjected to stepwise increases in the concentrations of INH, as described above. This procedure yielded M. tuberculosis isolates that were capable of growing in liquid or solid medium containing 20.0 mg of INH per liter. All eight clinical strains were subjected to the same procedures used for the induction of high-level resistance to INH, and similar results were obtained, with little difference in the amount of time required for the manifestation of induced resistance at each sequential level of induced resistance being detected (data not shown due to limitation of space). Reversal of INH resistance. | Although the contents of each culture at the conclusion of each passage phase were shown by the use of identification probes to be M. tuberculosis , there was the real possibility that the increasing level of resistance to INH was due to the selection of mutants containing any one of the known mutations associated with resistance to INH. Since mutations remain true when the isolates are transferred to drug-free medium, aliquots were taken from cultures of isolates that were growing in 20.0 mg of INH per liter (induced resistance to INH at IL7) and transferred to drug-free medium, after which time serial passages were conducted in drug-free medium. Between serial passages, an assay for determination of the INH MIC was conducted. As shown in Fig. , the MICs for isolates from each successive passage (IL8, IL9, and IL10) continued to decrease, eventually reaching that for the initial parent strain (average MIC at IL11, 0.06 mg/liter). The latter progeny that by this time were fully INH susceptible were subjected to prolonged incubation in INH, as were their original parents, and the cycle of induced resistance to INH was again successfully repeated . The induction of INH resistance and its reversibility with repeated passage in drug-free medium, followed by the repeat of the INH resistance induction cycle, argue against the involvement of any chromosomal mutation that may have been selected or the introduction of a contaminant during the manipulations described above. Nevertheless, to eliminate these two possibilities, the probability of which are low, representative strains from each phase of the cycle (susceptible strains and strains with induced resistance) were subjected to IS6110-RFLP analysis and katG mutation identification analysis. The IS6110-RFLP pattern remained identical for each representative throughout the cycle, thus eliminating any chance that the appearance of an increased level of resistance was due to the introduction of a contaminant or to the testing of a specimen of a mixed type (data not shown). The appearance of INH resistance is not due to the selection of katG mutants, which are known to be the most prevalently associated with high-level resistance to INH as a result of the S315->T mutation, since all of the PCR-RFLP analyses that detect this point mutation were negative for all strains tested (data not shown) . Reduction of levels of INH resistance by an efflux pump inhibitor. | The induction of high-level resistance to INH may be due to the activation or induction of an efflux pump mechanism(s). For this purpose INH-susceptible reference strain M. tuberculosis H37Rv (ATCC 27294) and isolates with INH-induced resistance at IL7 (resistant to INH at 20.0 mg/liter [MIC, 40 mg/liter]) were evaluated for the presence of an efflux pump by the use of known inhibitors of bacterial efflux pumps such as CPZ , verapamil, and reserpine . The existence of an efflux pump that renders a bacterium resistant to a given antibiotic is shown by the ability of known inhibitors of efflux pumps to reduce the level of resistance to the antibiotic. The concentration selected for an inhibitor of an efflux pump must be one that has a nominal effect on the growth of the bacterium itself so that the effects related to the action of the pump on the activity of a given antibiotic can be isolated. Table identifies the highest concentrations of CPZ, verapamil, and reserpine that had no effect on the initial INH-susceptible ATCC 27294 reference strain and the clinical strains used in this study, as well as the concentrations that correspond to their MICs. Whereas subinhibitory concentrations of CPZ or verapamil did not affect the ability of the cells with induced INH resistance at IL7 to grow in the presence of 20 mg of INH per liter (data not shown), reserpine at 20 mg/liter reduced the ability of all of the strains at IL7 to grow in the presence of 20.0 to 0.2 mg of INH liter . Complete inhibition of the growth of strains with induced INH resistance at IL7 in the presence of 0.1 mg of INH per liter required an average reserpine concentration of 40 mg/liter. However, although this concentration of reserpine alone had no effect on the growth of the initial INH-susceptible strains , it did inhibit the growth of these cells with induced INH resistance by as much as 15% , an effect which we deemed the nominal effect. The data presented in Table also show that when strains with induced resistance to INH at 20 mg/liter (IL7) were grown in the presence of reserpine at 20 mg/liter and INH at 0.1 and 0.2 mg/liter (with the 0.1-mg/liter concentration being the critical concentration of INH used as the cutoff for the demonstration of resistance of a clinical specimen), their growth was not completely inhibited. This apparent resistance may reflect the large number of efflux pump units that have been induced, with the inhibitory capability of INH overcome by this concentration of reserpine. This interpretation receives support from the data also presented in Table , which demonstrate that an increase in the concentration of reserpine to 40 mg/liter practically eliminates resistance to INH. It may be noted from the data in Table that strains that reverted to INH susceptibility had some demonstrable resistance to concentrations of INH that were slightly above and at the cutoff point used for the interpretation of resistance to INH. We interpret these results to indicate that even after extensive culture in drug-free medium, there is still evidence of a reserpine-sensitive mechanism which is completely inhibited only by the higher concentration of reserpine (40 mg/liter). FIG. 1. | Appearance of INH resistance after completion of INH susceptibility assay. Appearance of INH resistance after completion of INH susceptibility assay. The growth curves of the H37Rv (ATCC 27294) reference strain (*) and the eight clinical INH-susceptible strains in BACTEC 12B vials containing 0.1 mg of INH per liter and the H37Rv (ATCC 27294) reference strain and the eight clinical INH-susceptible strains grown in drug-free medium as controls are shown. The results for the eight clinical strains are recorded at 5-day intervals, with each point being the mean +- standard deviation of the daily readings. All strains were initially susceptible to 0.1 mg of INH per liter, as evidenced by the results at the end of the INH susceptibility assay period (maximum, 12 days). Incubation for a maximum of an additional 28 days yielded a growth index that was indicative of resistance to INH. TABLE 1 | Characteristics of contents of INH-containing BACTEC 12B vials TABLE 2 | Highest concentrations of CPZ, verapamil, and reserpine with nominal effects on growth of M. tuberculosis parental strains (INH susceptible) and the MICs for complete inhibition of growth TABLE 3 | Effect of a known inhibitor of efflux pumps, reserpine, on activities of INH on INH-susceptible strains, strains at IL7 with induced INH resistance, and INH-susceptible revertants at IL11a DISCUSSION : Considerable variation in antibiotic susceptibility is found among the different species of mycobacteria, such that whereas M. tuberculosis is highly susceptible to INH, M. smegmatis and other fast-growing mycobacteria are resistant to the drug . The susceptibility of M. tuberculosis to INH is due to an efficient conversion of the drug INH by the catalase-peroxidase enzyme to a compound that inhibits mycolic acid synthesis . Recently, natural resistance of M. smegmatis to INH has also been ascribed to a verapamil- and reserpine-sensitive efflux pump . The observations made from our study indicate for the first time that INH-susceptible strains of M. tuberculosis exposed to INH for a prolonged period of time can become resistant to 0.1 mg of INH per liter. This induced resistance to INH is reproducible and can gradually be increased to 20 mg/liter by gradual stepwise increases in the concentration of INH. Induced resistance can be effectively reversed by consecutive passage in drug-free medium. The cycle of induced resistance to INH can be repeated by these procedures, probably ad infinitum. The use of identification probes and IS6110-RFLP analysis ruled out any possibility that a contaminant was present or that the initial specimen was of a mixed type. The reversal of induced resistance to INH by simple serial passage in drug-free medium ruled out the possibility of selection of INH resistance mutations that could account for the INH resistance. Direct evaluation of the cultures at each phase of the study for the presence of katG mutations further supported the contention that the induction of resistance to INH was not due to the selection of mutations most commonly associated with resistance to INH. Lastly, although gene amplification can cause increased levels of antibiotic resistance in other species of bacteria by overproduction of an antibiotic-inactivating enzyme , it is unlikely that the high-level resistance to INH induced in this study is caused by a similar mechanism, inasmuch as the resistance observed in our study continues while the organism is replicating in the presence of high levels of INH. The loss of this resistance requires a large number of replications afforded by five prolonged passages in drug-free medium, whereas antibiotic resistance due to gene amplification is much more rapidly lost. In addition, gene amplification has not been observed with M. tuberculosis due to the high degree of inherent stability of the chromosome . As mentioned in the introduction, the INH resistance in only about 70 to 80% of M. tuberculosis strains has been correlated to mutations in any of the genes assigned as potential targets or to alternative metabolic pathways for INH inhibition . Efflux pumps that render bacteria resistant to antibiotics have been identified in all bacteria, including mycobacteria, studied to date . The correlation of antibiotic resistance to an efflux pump is made when the level of antibiotic resistance is reduced in the presence of both antibiotic and an inhibitor of the pump itself. Without exception, each of the studies cited evaluated the organisms for the presence of a preexisting efflux pump . Accordingly, strains of M. tuberculosis that are susceptible to INH would by definition not contain a mutation for INH resistance, nor would they be expected to harbor an efflux pump capable of pumping out INH. The demonstration in our study that high-level resistance to INH can be gradually induced in INH-susceptible strains of M. tuberculosis via a mechanism that does not involve a mutation but that is sensitive to reserpine supports the contention that induced resistance might be due to an efflux pump mechanism. Whether induced resistance is due to activation of a preexisting efflux pump, an increase in the number of efflux pump units, or de novo synthesis of the pump itself was not determined in the present study. The growth inhibition results obtained by the use of the lower concentration of reserpine (20 mg/liter) in combination with 0.2 or 0.1 mg of INH per liter with strains obtained at IL11 (INH-susceptible revertant strains), although consistent with the presence of an efflux pump mechanism, are difficult to interpret, inasmuch as the growth of these strains should be fully inhibited. Although it is possible that the induced resistance to INH in our study was due to the selection of mutants capable of synthesizing an efflux pump, we consider this possibility remote, inasmuch as the ATCC control strain and all eight clinical strains were induced to present the efflux pump. The induction of high-level resistance to INH by continuous exposure to increasing concentrations of the antibiotic has not been described previously. The induction of INH resistance by INH may occur during INH therapy and may account for the remainder of the INH resistance that is not accounted for by mutations and that, due to the limitation of the time interval used for the INH susceptibility assay, is not detected. Backmatter: PMID- 12183224 TI - Evaluation of Antibiotic Susceptibilities of Three Rickettsial Species Including Rickettsia felis by a Quantitative PCR DNA Assay AB - Rickettsiae grow only intracellularly, and the antibiotic susceptibilities of these bacteria have been assessed by either plaque, dye uptake, or immunofluorescence assays, which are time-consuming. We used a quantitative PCR (with the LightCycler instrument) to assess the levels of inhibition of Rickettisa felis, R. conorii, and R. typhi DNA synthesis in the presence of various antibiotics. We established the kinetics of rickettsial DNA during growth and showed that R. conorii grows more quickly than R. typhi in cell culture, with maximum replication occurring after 5 and 7 days, respectively. The MICs of the antibiotics tested for R. conorii and R. typhi by the quantitative PCR assay were similar to those previously obtained by plaque and dye uptake assays. We found that R. felis is susceptible to doxycycline, rifampin, thiamphenicol, and fluoroquinolones but not to gentamicin, erythromycin, amoxicillin, or trimethoprim-sulfamethoxazole. The resistance of this new species to erythromycin is consistent with its current taxonomic position within the spotted fever group. We believe that quantitative PCR could be used in the future to simplify and shorten antibiotic susceptibility assays of other rickettsiae and other strict intracellular pathogens. Keywords: Introduction : Rickettsiae are strict intracellular bacteria belonging to the alpha group of the class Proteobacteria . Rickettsioses are zoonoses that have geographical distributions similar to those of their vectors. The genus comprises the spotted fever group rickettsiae and the typhus group rickettsiae, which includes Rickettsia prowazekii, the agent of epidemic typhus, and R. typhi, the agent of murine typhus . Since they are obligate intracellular bacteria, in vitro studies of their susceptibilities to antibiotics necessitate the use of cell culture systems. Using the plaque assay and the dye uptake assay , we have now described the antibiotic susceptibilities of most of the rickettsiae with the exception of R. felis, which, until recently, was not available to us. The two assays depend on the induction of cytopathic effects and plaque formation in cell cultures by the rickettsiae, but some rickettsiae do not normally cause cytopathic effects in primary cultures . Recently, Ives et al. (-) described a new assay that uses immunofluorescence staining, which avoids the problem of a lack of cytopathic effects. A lack of cytopathic effects occurs with R. felis, the agent of flea-borne spotted fever, which we have now isolated and cultured in a Xenopus laevis cell line (XTC-2 cells) and in Vero cells. To determine the antibiotic susceptibility of this species we developed an original antibiotic assay using PCR with the LightCycler instrument. It is a new, real-time PCR technique which has been used to diagnose several bacterial infections but which has not been used to determine antibiotic susceptibilities. The technique combines PCR amplification and detection of products in a single optically clear glass capillary tube, which enables rapid temperature transitions with hot air and which reduces the PCR time to 45 min . In this report we describe the PCR method with the LightCycler instrument that we developed for antibiotic susceptibility testing of Rickettsia isolates. Furthermore, we report on the antibiotic susceptibilities that we obtained for R. felis, R. conorii, and R. typhi and compare these susceptibilities with those obtained by the plaque assay, the reference method . MATERIALS AND METHODS : Rickettsial isolates. | Four isolates of R. felis from fleas were used in the study: Marseille-URRWFXCal2 I-2363 (Cal2) , Pete2, Marseille, and Baton Rouge . They were cultured in XTC-2 cells with Leibowitz-15 medium (2 mM l-glutamine and amino acids in the l configuration; Gibco) containing 5% fetal calf serum and 2% tryptose phosphate (Gibco), as described previously . Also used were R. conorii (ATCC VR-141), the agent of Mediterranean spotted fever, and the R. typhi Wilmington strain (ATCC VR-144), the agent of murine typhus, which were grown in Vero cells in minimum essential medium at 32C with 2 mM l-glutamine (Gibco) containing 4% fetal calf serum. Antibiotics. | The antibiotics tested were amoxicillin (Beecham-Sevigne, Paris, France), gentamicin (Dakota Pharm, Creteil, France), ciprofloxacin (Bayer Pharma, Sebs, France), erythromycin (Abbott, Rungis, France), rifampin (Cassenne, Puteaux, France), doxycycline (Pfizer, Neuilly, France), telithromycin (Hoescht Marion Roussel, Romainville, France), levofloxacin (Hoescht Marion Roussel), ofloxacin (Diamant, Puteaux, France), trimethoprim-sulfamethoxazole (Roche, Paris, France), and thiamphenicol (Sanofi Winthrop, Gentilly, France). Stock solutions were prepared by dissolving all antibiotics except telithromycin in sterile distilled water; telithromycin was first dissolved in methanol before being diluted in sterile distilled water. Stock solutions were stored at -20C in culture medium, and dilutions were made immediately prior to the tests. The plaque assay for R. conorii and R. typhi was performed as described previously and served as a control for the two other methods that we used, with the results obtained by the plaque assay compared with those obtained by the other two methods. For the plaque assay with R. felis, 24-well microtiter plates with XTC-2 cells were infected with R. felis and incubated at 28C in Leibowitz-15 medium. Similarly, plates containing Vero cells were infected with R. conorii or R. typhi and incubated at 32C with minimum essential medium . The antibiotics at each dilution were added to rows of infected cells; positive controls consisted of infected cells without antibiotics in the medium, and negative controls consisted of uninfected cells with antibiotics in the medium. From day 0 to the end of the experiment, three wells from each row were harvested each day and stored at -20C in sterile tubes for the PCR assay. Also, cells from each row were stained with Gimenez stain, and the number of intracellular rickettsiae was counted at x1,000 magnification. The MIC was defined as the lowest antibiotic concentration that completely inhibited the growth of the rickettsiae. Experiments were performed in duplicate to confirm the results. Real-time PCR was performed with a LightCycler instrument (Roche Biochemicals, Mannheim, Germany), which is a combined fluorimeter and thermocycler that enables rapid (45 min) PCR thermocycling. Samples and the PCR master mixture are placed in 30-mul glass cuvettes, and sample detection is based on the principle of fluorescence resonance energy transfer , with adjacent hybridization probes directed against the intended PCR product. With fluorescein serving as the donor fluorophore and LC-Red 640 (Roche Biochemicals) serving as the acceptor fluorophore, the presence of PCR amplicons can be assessed by detection of the fluorescence of LC-Red 640. Samples can be assayed for the presence of this signal in real time at each PCR cycle, and the cycle number at which the signal is first detected can be correlated to the original concentration of DNA . The specificity of amplification can be confirmed by melting curve analysis. Single melting peaks can be generated by depicting the negative derivative of fluorescence versus temperature (-dF/dT) over the course of gradual melting of a PCR product. Extraction of DNA. | After the harvested cells were thawed, they were centrifuged in a Beckman Avanti 30 centrifuge at 5,000 rpm for 10 min, the supernatant was discarded, and the pellet was washed twice with sterile distilled water before being resuspended in 200 mul of sterile distilled water. Extraction of the DNA was performed with 20% Chelex (biotechnology-grade chelating resin [Chelex 100; Bio-Rad, Richmond, Calif.]) in sterile water . Briefly, 500 mul of Chelex was added to each tube of centrifuged and washed cells before the tubes were vortexed and placed in a boiling water bath for 30 min. After centrifugation at 14,000 rpm for 10 min, the supernatants were harvested and stored in sterile tubes at 4C before PCR was performed. PCR master mixture. | Master mixtures were prepared by following the instructions of the manufacturer and with primers CS877F (5'-GGG GGC CTG CTC ACG GCG G-3') and CS1258R (5'-ATT GCA AAA AGT ACA GTG AAC A-3'), whose sequences are specific for the sequence of the citrate synthase gene and which have been described previously . The 20-mul sample in each glass cuvette contained 2 mul of LightCycler DNA Master SYBR Green (Roche Biochemicals), 2.4 mul of MgCl2 at 4 mM, 1 mul of each primer at 0.5 muM, 11.6 mul of sterile distilled water, and 2 mul of extracted DNA. PCR cycling and melting curve conditions. | After one pulse centrifugation to allow mixing and to drive the mixture into the distal end of each cuvette, the tubes were placed in the LightCycler instrument. The amplification program began with a denaturation step of 95C for 120 s, which was followed by 40 cycles of denaturation at 95C for 15 s, annealing at 54C for 8 s, and extension at 72C for 15 s, with fluorescence acquisition at 54 in the single mode. Melting curve analysis was done at 45 to 90C (temperature transition, 20C/s), with stepwise fluorescence acquisition done by real-time measurement of the fluorescence directly in the clear glass capillary tubes. Sequence-specific standard curves were generated by using 10-fold serial dilutions (105 to 106 copies) of a standard concentration of Rickettsia. The number of copies of each sample transcript was then calculated from the standard curve with the LightCycler software. The MIC was defined as the lowest antibiotic concentration that caused inhibition of growth of the rickettsiae, i.e., that resulted in the number of DNA copies being similar to that on day 0. The experiments were carried out in duplicate. RESULTS : The Gimenez staining assay was easy to perform with R. felis, but counting of the organisms was laborious and time-consuming, as 1,000 cells were examined for each sample. The results obtained with the reference strains (R. typhi and R. conorii) by the Gimenez staining assay were similar to those obtained by the plaque assay . The PCR assay with the LightCycler instrument was rapid and reproducible and allowed the number of DNA copies to be counted precisely. Figures and present the results of the quantitative PCR assay. Figure shows the melting curves obtained with standard concentrations of R. felis and R. conorii, showing the specificity of the PCR product (a single peak was obtained in all PCRs). Figure shows the standard calibration curve obtained with 10-fold serial dilutions of R. felis. The standard curve was determined in each experiment to enable the results of all experiments to be correlated. In Fig. we show the growth kinetics of R. felis, R. conorii, and R. typhi determined by the quantitative PCR assay. In Vero cells, the number of copies of R. conorii DNA increased more rapidly than the number of copies of R. typhi DNA (5 versus 7 days), but the two strains achieved a zenith of about 105 DNA copies after 6 to 7 days. R. felis grown in XTC-2 cells had a maximum number of DNA copies after 5 days. In the antibiotic susceptibility assays, the results obtained by the Gimenez staining method and the real-time PCR assay were consistent with those obtained previously by the plaque assay, which is regarded as the "gold standard." The MICs of the antibiotics for the four isolates of R. felis, R. conorii, and R. typhi determined by the different assays are shown in Table . The four isolates of R. felis had the same antibiotic susceptibility profiles, being resistant to amoxicillin, gentamicin, and trimethoprim-sulfamethoxazole. They had poor susceptibilities to erythromycin, with MICs ranging from 8 to 32 mug/ml; in contrast, R. typhi was highly sensitive (MIC, 1 mug/ml). Doxycycline was the most effective antibiotic against all isolates tested, with MICs ranging from 0.06 to 0.125 mug/ml. Fluoroquinolone compounds had MICs from 0.5 to 1 mug/ml, whereas the MICs of thiamphenicol were between 1 and 4 mug/ml. Rifampin was active, with MICs ranging from 0.06 to 1 mug/ml. The new ketolide compound, telithromycin, was bacteriostatic at concentrations from 0.5 to 1 mug/ml. In previous studies we have found that plaques caused by R. conorii in cell cultures are larger in the presence of trimethoprim-sulfamethoxazole (unpublished data), and in this study we found the number of DNA copies in cultures with or without trimethoprim-sulfamethoxazole to be almost identical (388,500 and 395,000 copies, respectively) at the time of maximum growth on day 7 of culture. FIG. 1. | Melting curve obtained with standard concentrations of R. conorii Melting curve obtained with standard concentrations of R. conorii by PCR with the LightCycler instrument. The specificities of the PCR products are shown by detection of a single peak. FIG. 2. | Standard calibration curve obtained with 10-fold serial dilutions of R. felis Standard calibration curve obtained with 10-fold serial dilutions of R. felis (strain Baton Rouge). Quantification was performed by determining the threshold cycle (cycle number). The threshold is defined as the cycle at which the fluorescence exceeds 10 times the standard deviation of the mean baseline emission. One representative curve and its equation are given. FIG. 3. | Kinetics of R. conorii Kinetics of R. conorii , R. typhi (*), and R. felis growth in cell culture determined by quantitative PCR assay. TABLE 1 | In vitro susceptibilities of rickettsial strains to antibiotics DISCUSSION : Rickettsiae are gram-negative bacteria that are primarily associated with arthropods that are the vectors of human infections. Since they are obligate intracellular bacteria, in vitro studies of their susceptibilities to antibiotics require the use of cell culture systems . In our laboratory we have recently isolated and cultured R. felis, the agent of flea-borne spotted fever. We inoculate crushed cat fleas, in which the DNA of the ELB agent was previously detected, onto XTC-2 cells incubated at 28C . With R. felis no cytopathic effects were observed in XTC-2 cell cultures, and only very small plaques could be seen in Vero cells, with these plaques observed only after 18 days of culture. As it was not possible to use conventional assays to determine the susceptibility of R. felis to antibiotics, we developed an assay using Gimenez staining to determine rickettsial numbers before and after addition of antibiotics to the culture medium. We also developed a new quantitative PCR assay using the LightCycler system to determine the antibiotic susceptibilities of rickettsiae. The Gimenez staining method was too laborious and time-consuming to be a recommended technique. The quantitative PCR assay with the LightCycler instrument has already been shown to be an effective diagnostic tool and has been used to detect antibiotic-resistant strains of Mycobacterium tuberculosis . Our experiments by PCR with the LightCycler instrument were, to the best of our knowledge, the first to determine the MICs of antibiotics by that technique. The quantitative method that we describe enabled us to determine the antibiotic susceptibilities of fastidious Rickettsia spp. rapidly and consistently and may be useful for determination of the antibiotic susceptibilities of many other bacteria. The susceptibility results that we obtained for the reference strains, R. conorii and R. typhi, by the PCR assay were comparable to those obtained by the plaque assay, and for the first time, we were able to describe the kinetics of rickettsial DNA during the rickettsial growth phase. These were consistent with our knowledge of the in vitro culture characteristics of rickettsiae. In cell culture, R. conorii grows rapidly, the majority of cells are infected after 3 to 4 days, and subculturing is generally necessary after 5 days, when all cells have been destroyed. R. typhi, however, grows more slowly and can be maintained for 7 to 9 days before subculturing is necessary. We have observed similar growth kinetics for R. felis in XTC-2 cells, with almost all cells becoming infected after 5 days. The kinetics that we found in our PCR study were consistent with these in vitro culture characteristics of R. felis . Finally, the kinetics of growth of each bacterium enabled us to choose the best time to perform the antibiotic assay, i.e., at day 4 for R. conorii and R. felis and at day 6 for R. typhi. We found that beta-lactams, gentamicin, trimethoprim-sulfamethoxazole, and erythromycin are not active against R. felis. These results differ from those reported by Radulovic et al. , who found that "R. felis" was susceptible to erythromycin (MICs, 0.5 to 1 mug/ml). These were similar to the MICs for R. typhi , and it later appeared that the "R. felis" strain that they studied was in fact a strain of R. typhi . The most effective antibiotics in our study were doxycycline, fluoroquinolones, rifampin, thiamphenicol, and telithromycin. Our results also confirm those of previous experiments that showed that doxycycline, rifampin, and fluoroquinolones are the most effective antibiotics in vitro against all strains of rickettsiae . The MICs found in our study and those reported previously and obtained by the plaque assay were lower than those obtained by Ives et al. by the immunofluorescence assay, especially for fluoroquinolone compounds. The results obtained by the immunofluorescence assay are not consistent with clinical experience, suggesting that the method should be reconsidered. Patients suffering from Mediterranean spotted fever have been successfully treated with fluoroquinolones , and thus, fluoroquinolones may be considered a safe alternative to tetracyclines for the treatment of rickettsial diseases. During pregnancy and childhood, macrolide compounds may represent a safe alternative to doxycycline and fluoroquinolones. There is, however, a high level of heterogeneity in the susceptibilities of spotted fever group rickettsiae to these compounds ; josamycin is the only antibiotic whose efficacy against rickettsial diseases in this population has been tested , whereas erythromycin has not been shown to be effective . We have previously shown that it is possible to distinguish the typhus and spotted fever groups of rickettsiae by their susceptibilities to erythromycin . The relative resistance of R. felis to erythromycin is consistent with the fact that this rickettsia belongs to the spotted fever rather than the typhus group , the members of which are susceptible to erythromycin. In summary, our report describes for the first time the susceptibilities of R. felis to a wide range of antibiotics determined by a new quantitative PCR assay. Our results show that doxycycline, fluoroquinolones, telithromycin, and rifampin are effective against the organism but that trimethoprim-sulfamethoxazole, beta-lactams, and erythromycin are not. This new quantitative PCR assay should be developed and adapted to testing of the antibiotic susceptibilities of other strict intracellular pathogens. Backmatter: PMID- 12183243 TI - In Vitro and In Vivo Synergy of Fosmidomycin, a Novel Antimalarial Drug, with Clindamycin AB - Fosmidomycin acts through inhibition of 1-deoxy-d-xylulose 5-phosphate (DOXP) reductoisomerase, a key enzyme of the nonmevalonate pathway of isoprenoid biosynthesis. It possesses potent antimalarial activity in vitro and in murine malaria. In a recent clinical study, fosmidomycin was effective and well tolerated in the treatment of patients with acute uncomplicated Plasmodium falciparum malaria but resulted in an unacceptably high rate of recrudescence. In order to identify a potential combination partner, the interaction of fosmidomycin with a number of antimalarial drugs in current use was investigated in a series of in vitro experiments. Synergy was observed between fosmidomycin and the lincosamides, lincomycin and clindamycin. The efficacy of a combination of fosmidomycin and clindamycin was subsequently demonstrated in the Plasmodium vinckei mouse model. Keywords: Introduction : In humans, isoprenoids are synthesized via the mevalonate pathway. In contrast, they are synthesized by the nonmevalonate pathway (the 1-deoxy-d-xylulose 5-phosphate [DOXP] pathway, also called the MEP pathway) in a number of bacterial species and inside the plastides of algae and higher plants . Similarly, the enzymes of the nonmevalonate pathway are located inside the plastide-like organelle (apicoplast) of malaria parasites . The antibiotic fosmidomycin, originally isolated from Streptomyces lavendulae, represents a potent inhibitor of DOXP reductoisomerase, a key enzyme of the nonmevalonate pathway . Recently, it was demonstrated that fosmidomycin possesses potent antimalarial activity in vitro and in murine malaria . FR900098, a fosmidomycin derivative, was found to be twice as effective, while the prodrug derivatives had increased oral bioavailability in the mouse model . In a recent clinical study conducted in Gabon and Thailand, 20 patients with acute uncomplicated Plasmodium falciparum malaria were treated with fosmidomycin administered orally (B. Lell, R. Ruangweerayut, J. Wiesner, M. Missinou, A. Schindler, T. Baranek, M. Hintz, D. Hutchinson, H. Jomaa, and P. Kremsner, unpublished data). The treatment was well tolerated and resulted in rapid parasite and fever clearance times, comparable to those obtained with conventional quinoline antimalarial agents. All patients were clinically and parasitologically cured by day 7. By day 28, however, 9 out of 18 evaluable patients experienced recrudescence. A similarly high rate of recrudescence had been observed previously when the hydroxynaphthoquinone antimalarial agent atovaquone was evaluated as a single entity . Subsequently, proguanil was identified as a partner for atovaquone on the basis of in vitro synergistic activity, resulting in a highly effective and well-tolerated fixed drug combination, approved and marketed as Malarone . Using a similar approach, we have investigated the interaction of fosmidomycin with most antimalarial agents in clinical use. MATERIALS AND METHODS : Materials. | Blood components were provided by the local Institute of Clinical Immunology and Transfusion Medicine. Chloroquine, quinine, artemisinin, doxycycline, ciprofloxacin, rifampin, and lincomycin were purchased from Sigma. Mefloquine and halofantrine were gifts from Reto Brun (Basel, Switzerland). Atovaquone was a gift from Peter Kremsner (Tubingen, Germany). Proguanil was a gift from Wallace Peters and Brian Robinson (Harrow, Middlesex, United Kingdom). Lumefantrine was provided by Welding GmbH & Co. (Hamburg, Germany). Azithromycin was extracted from Zithromax tablets (Pfizer). Clindamycin was purchased from Sigma, ICN, and Welding GmbH & Co. The P. falciparum laboratory strains used were 3D7 (The Netherlands), HB3 (Honduras), Dd2 (Indochina), and A2 (Gambia). The Plasmodium vinckei strain was provided by Henri Vial (Montpellier, France). In vitro antimalarial activity. | P. falciparum was cultivated in RPMI 1640 medium (Life Technologies) supplemented with 10% human type O+ serum and 25 mM HEPES. Human type O+ erythrocytes served as host cells . Cultures were kept at 37C under an atmosphere of 5% O2, 3% CO2, and 92% N2. In vitro drug sensitivity assays were carried out on 96-well microtitration plates . Fosmidomycin was dissolved in complete culture medium and sterilized by filtration. The other drugs were dissolved in dimethyl sulfoxide and prediluted with complete culture medium. Infected erythrocytes (0.15 ml per well with 2% hematocrit and 0.4% parasitemia) were incubated in duplicate with a twofold serial dilution of each drug for 48 h. After addition of 0.8 muCi of [3H]hypoxanthine (Amersham Pharmacia) in 50 mul of medium per well, the plates were incubated for another 24 h. Parasites were collected on glass fiber filters with a cell harvester (Micromate 196; Packard), and incorporated radioactivity was measured using a beta-counter (Matrix 9600; Packard). Growth inhibition was expressed as percent 3H incorporation compared with untreated controls. Values were plotted on semilogarithmic paper for extrapolation of 50% inhibitory concentrations (IC50s). In vitro drug interaction. | Drug interaction studies were performed as previously described . Initially, the IC50s of the test drugs alone were determined. Subsequently, drug solutions were diluted with culture medium to initial concentrations of 80 times the predetermined IC50s. These solutions were combined in ratios of 1:5, 1:2, 2:1, and 5:1. Single and combination drug solutions were then introduced into 96-well plates to give duplicate rows of fosmidomycin alone, the test drug, and the four combinations. Finally, the IC50s of the two test drugs alone and in combination were determined. For data interpretation, the IC50s of the drugs in combination were expressed as fractions of the IC50s of the drugs alone normalized to 1. Isobolograms were constructed by plotting the IC50 of one drug against the IC50 of the other for each of the four drug ratios, with a concave curve indicating synergy, a straight line indicating addition, and a convex curve indicating antagonism. To obtain numeric values for the kind of interaction, results were expressed as the sum of the fractional inhibitory concentrations (sum FIC), calculated as (IC50 of drug A in mixture/IC50 of drug A alone) + (IC50 of drug B in mixture/IC50 of drug B alone). Sum FIC values indicate the kinds of interactions as follows: <0.5, synergy; 0.5 to 1, addition; 1 to 2, indifferent interaction; >2, antagonism. Sum FIC values were calculated for the drug ratio resulting in the point closest to the middle of the isobologram. For determination of growth inhibition by fosmidomycin in the presence of constant clindamycin concentrations, 20-ml aliquots of the suspension of infected erythrocytes in culture medium were adjusted to the desired clindamycin concentration from a 2 mM stock solution in dimethyl sulfoxide before being loaded in triplicated rows onto the 96-well plate. Then a dilution series of fosmidomycin was prepared on the plate. In vivo drug interaction. | For in vivo drug testing, mice were inoculated by intraperitoneal injection with approximately 5 x 107 infected erythrocytes from a donor mouse. Fosmidomycin was dissolved in phosphate-buffered saline and administered orally (75 mg/kg of body weight). Clindamycin hydrochloride was dissolved in distilled water and administered by intraperitoneal injection (5 mg/kg). Four mice were used for each treatment group, and three mice were used for the control group. Parasitemia was monitored by Giemsa staining of blood smears. Mice were sacrificed when parasitemia exceeded 40%. The animal experiments complied with all relevant federal guidelines and institutional policies. RESULTS : In vitro drug interaction. | Before we embarked on the identification of a therapeutic partner for fosmidomycin, a control experiment was performed to validate the methodology through assessment of the interaction between atovaquone and proguanil, a proven synergistic drug combination. As expected, the IC50s of the individual drugs in the four different mixtures were significantly lower than the IC50s of the drugs alone, resulting in a concave curve in the isobologram . Subsequently, fosmidomycin was tested in combination with most of the currently used antimalarial drugs . The antifolate drugs, pyrimethamine and cycloguanil, were excluded from the study because of the existence of highly resistant P. falciparum strains in the field . The study was performed with four different strains of P. falciparum including the multidrug-resistant strain Dd2 (resistant to chloroquine, pyrimethamine, and cycloguanil). The absolute IC50s of the drugs used in the study for the different strains are listed in Table . In the interaction experiments, there was no apparent specific tendency for any strain. Typical isobolograms are shown in Fig. , and the sum FIC values for all drug combinations tested are summarized in Table . The interaction of fosmidomycin with all quinoline and aryl-amino-alcohol antimalarial drugss was indifferent, with the exception of quinine, for which the interaction was additive. Also, artemisinin, atovaquone, and proguanil had indifferent effects. In addition, a triple combination of fosmidomycin, proguanil, and atovaquone was tested but also resulted in an indifferent effect (data not shown). Among the antibiotics with known antimalarial activity, doxycycline and azithromycin were additive, and ciprofloxacin and rifampin were indifferent. Synergy was observed only with clindamycin and its natural precursor, lincomycin. Remarkably, the shapes of the corresponding isobolograms were asymmetric, in contrast to the curve obtained in the control experiment with atovaquone and proguanil. To assess whether the synergy of fosmidomycin and clindamycin is of clinical relevance, it should be noted that clindamycin is an effective but very slow acting antimalarial drug. Therefore, the absolute IC50s of clindamycin obtained under our assay conditions were comparatively high . At lower concentrations of clindamycin, the parasites develop normally in the first cycle after exposure, and reinvasion of new host erythrocytes takes place. Growth inhibition finally occurs at the end of the second cycle (the so-called "delayed kill effect") . Consistently, for patients treated with clindamycin, amelioration of symptoms is observed as late as the fourth day of treatment . Fosmidomycin, in contrast, kills the parasites at the end of the first cycle. Consequently, we have investigated whether the synergy of fosmidomycin and clindamycin remains apparent under pharmacologically achievable concentrations of clindamycin. It has been reported that during the course of standard low-dose clindamycin therapy with 5 mg/kg every 8 h, minimal plasma drug levels of 150 to 800 ng/ml are achieved . The broad range reported possibly depends on the methods used for drug determination. Therefore, we investigated the sensitivity of P. falciparum to fosmidomycin in the presence of clindamycin concentrations between 42 ng/ml (0.1 muM) and 850 ng/ml (2 muM). Parasite growth was not affected by these concentrations of clindamycin alone within the assay time, but the parasites were killed when the incubation time was extended to 4 days (data not shown). In the presence of an 850-ng/ml concentration of clindamycin, the IC50 of fosmidomycin for P. falciparum strain HB3 changed from 82 to 48 ng/ml . Remarkably, in the presence of a 42-ng/ml concentration of clindamycin, the IC50 of fosmidomycin was still reduced to 55 ng/ml. In an independent experiment using P. falciparum strain A2, a similar shift to lower IC50s was observed in the presence of clindamycin . These data clearly demonstrate that an increased therapeutic response can be expected from a combination of fosmidomycin and clindamycin. In vivo drug interaction. | Next, the efficacy of fosmidomycin plus clindamycin was investigated in the P. vinckei mouse model. Mice were treated with each drug, administered in doses that were calculated to result in a partial reduction of parasitemia. In parallel, mice were treated with a combination of these drugs in subtherapeutic doses. A relatively high dose of fosmidomycin was chosen because of the anticipated short half-life in plasma . In the first experiment, the mice were infected on day 0 and treated on days 1 and 2. To assess efficacy, parasitemia was monitored on days 3, 4, and 5. On day 3, the parasitemia of mice treated with 75 mg of fosmidomycin/kg or 5 mg of clindamycin/kg was 7.8 or 20%, respectively, in comparison to 42% in untreated control mice . When the mice were treated with a combination of 75 mg of fosmidomycin/kg and 5.0 mg of clindamycin/kg, the parasitemia was approximately 0.1% on day 3, increasing to 0.2% on day 5. A combination of 75 mg of fosmidomycin/kg and 2.5 mg of clindamycin/kg was equally efficacious (data not shown). However, efficacy decreased significantly when the clindamycin dose was further reduced to 1.3 mg/kg (data not shown). An additional study was designed to investigate whether a combination of fosmidomycin and clindamycin would be effective when treatment is initiated in the presence of high parasitemia. This was of particular interest because clindamycin alone, even at high doses, is not able to rescue mice under such conditions. Therefore, treatment was started on day 3 after infection at a parasitemia of approximately 20%. Following treatment with fosmidomycin alone, parasitemia continued to increase for 24 h but then fell to 3.9% by day 5. By day 6, parasitemia had reached a range between 3.6 and 74%, and all mice, even those with only moderate parasitemia, had symptoms of severe anemia, a condition known as postschizontal anemia. Treatment with clindamycin alone did not stop the rise of parasitemia. When fosmidomycin and clindamycin were administered in combination, there was a constant reduction in the level of parasitemia to 0.5% on day 6. Again, a combination of 75 mg of fosmidomycin/kg plus 2.5 mg of clindamycin/kg was as effective as 75 mg of fosmidomycin/kg plus 5.0 mg of clindamycin/kg (data not shown). FIG. 1. | Representative isobolograms of the interaction of fosmidomycin with quinine, artemisinin, proguanil, clindamycin, or lincomycin. Representative isobolograms of the interaction of fosmidomycin with quinine, artemisinin, proguanil, clindamycin, or lincomycin. The interaction of proguanil with atovaquone was assessed as a control experiment (upper left panel). The P. falciparum strain used for each experiment is indicated. FIG. 2. | Dose response of P. falciparum Dose response of P. falciparum growth to fosmidomycin in the presence of different constant clindamycin concentrations. P. falciparum-infected erythrocytes were incubated with a serial dilution of fosmidomycin in the absence (filled circles) or in the presence of clindamycin at 43 ng/ml (diamonds), 210 ng/ml (inverted triangles), or 850 ng/ml (triangles). Parasite growth was monitored by radioactive hypoxanthine incorporation. The clindamycin concentrations selected did not have intrinsic antimalarial activity under the assay conditions. Results obtained by independent experiments with P. falciparum strains HB3 (upper panel) and A2 (lower panel) are presented. FIG. 3. | In vivo efficacy of fosmidomycin plus clindamycin in P. vinckei In vivo efficacy of fosmidomycin plus clindamycin in P. vinckei-infected mice. (A) For suppressive treatment, drugs were administered on days 1 and 2 postinfection and parasitemia was monitored on days 3 to 5. (B) For curative treatment starting with high parasitemia, drugs were administered on days 3 and 4 and parasitemia was monitored from days 3 to 8. Geometric mean values and ranges of observed values are indicated. Fos, fosmidomycin; Cli, clindamycin. TABLE 1 | In vitro IC50s of the test drugs determined for P. falciparum strains 3D7, HB3, Dd2, and A2 TABLE 2 | Interaction of fosmidomycin with other antimalarial drugs against P. falciparum in vitro DISCUSSION : In our in vitro experiments, combining fosmidomycin with commonly used antimalarial drugs resulted in an indifferent effect in most cases, with sum FIC values between 1.04 and 1.46. An antagonistic effect, resulting in sum FIC values higher than 2, refers to loss of activity when drugs are used in combination, indicating that higher concentrations of the individual drugs would be required to produce the same effect as when the drugs are administered singly. Such conditions, however, were not observed in this study. Therefore, even the drugs associated with an indifferent effect in vitro may prove to be useful therapeutic partners for fosmidomycin. In this regard, a combination of fosmidomycin with an artemisinin derivative may be of particular interest. Artemisinin derivatives such as artesunate and artemether have been successfully employed in various drug combinations, since they are active against multidrug-resistant parasites, display a favorable safety profile, and rapidly reduce the parasite load within one replication cycle . The additive effect observed with quinine may become relevant in severe malaria, where intravenous quinine, commonly given in combination with doxycycline, is still the first choice of treatment. Since the use of doxycycline is not appropriate for young children and pregnant women, there would seem to be a case for replacing it with fosmidomycin in view of the more rapid action of the latter. Our studies have provided conclusive evidence of potent in vitro synergistic activity between fosmidomycin and the lincosamide antibiotics lincomycin and clindamycin. This has also been established in a malaria mouse model, even when treatment was delayed pending the development of high parasitemia. While lincomycin as the natural precursor of clindamycin is of only historical interest, clindamycin is widely used for treating infections with gram-positive or anaerobic bacteria. In addition, it is active against parasites of the phylum Apicomplexa, which includes Plasmodium, Theileria, and Toxoplasma . Clindamycin is believed to target the prokaryote-like ribosomes of the apicoplast . By this means, self-replication of the organelle is inhibited, leading to the death of the parasite in the second replication cycle. The fact that fosmidomycin inhibits DOXP reductoisomerase, an enzyme that is localized in the apicoplast, may provide an explanation for the observed synergy with the lincosamides. However, it is not yet clear why other antibiotics that are also believed to impair apicoplast function do not exhibit such synergy. Possibly, clindamycin facilitates the transport of fosmidomycin into the parasite or the apicoplast by an unknown mechanism. Early studies on the effectiveness of clindamycin as an antimalarial agent were very promising . However, in view of its slow onset of action, the use of clindamycin as a single entity is restricted to the treatment of asymptomatic or uncomplicated P. falciparum malaria. It is also useful as a therapeutic partner in antimalarial-drug combinations. In Gabon, all of 38 adult patients receiving 5 mg of clindamycin/kg twice daily for 5 days were cured, with only 1 patient developing a recurrent parasitemia which may have been due to reinfection . The same regimen led to a 100% cure rate for 35 patients in Brazil . A quinine-clindamycin combination was effective against multidrug-resistant malaria in Thailand . Furthermore, a 3-day course of clindamycin plus quinine was curative in the treatment of uncomplicated P. falciparum malaria, compared to 7 days of treatment with quinine alone . Since clindamycin has been used only for a relatively small number of malaria patients, it is not expected that resistant parasites have developed in the field. As a therapeutic partner for fosmidomycin, clindamycin has the advantage of having a similarly short half-life in plasma . As a consequence, repeated dosing will be necessary, but the parasites will be exposed to subtherapeutic drug concentrations for a short time only, thereby deterring the emergence of resistance. Furthermore, the safety of clindamycin as an antibacterial agent has been substantiated through 35 years of clinical experience (, , 23). In addition to its potential use as an effective and affordable medication for uncomplicated malaria, fosmidomycin plus clindamycin may be of particular value for the treatment of severe malaria when patients are not able to tolerate oral medication. Conventional treatment with intravenous quinine may be life-threatening when the required dose is infused too rapidly. Parenteral administration of several highly potent drugs such as mefloquine, halofantrine, and atovaquone is precluded by their poor solubility. Fosmidomycin, in contrast, is freely water soluble, and bolus infusions of as much as 2 g were well tolerated in a phase I volunteer study . Also, clindamycin, in the form of its phosphonic acid ester, is available as an intravenous formulation. Therefore, development of an intravenous formulation of a combination of fosmidomycin and clindamycin for treatment of severe malaria should be technically feasible. However, the role of the combination of fosmidomycin and clindamycin in the treatment of acute uncomplicated P. falciparum malaria will first be established through an extended program of phase II studies. The first clinical studies are currently ongoing in Gabon and Thailand. Backmatter: PMID- 12183254 TI - Antistreptococcal Activity of Telithromycin Compared with Seven Other Drugs in Relation to Macrolide Resistance Mechanisms in Russia AB - The susceptibilities of 468 recent Russian clinical Streptococcus pneumoniae isolates and 600 Streptococcus pyogenes isolates, from 14 centers in Russia, to telithromycin, erythromycin, azithromycin, clarithromycin, clindamycin, levofloxacin, quinupristin-dalfopristin, and penicillin G were tested. Penicillin-nonsusceptible S. pneumoniae strains were rare except in Siberia, where their prevalence rate was 13.5%: most were penicillin intermediate, but for three strains (two from Smolensk and one from Novosibirsk) the MICs of penicillin G were 4 or 8 mug/ml. Overall, 2.5% of S. pneumoniae isolates were resistant to erythromycin. Efflux was the prevalent resistance mechanism (five strains; 41.7%), followed by ribosomal methylation encoded by constitutive erm(B), which was found in four isolates. Ribosomal mutation was the mechanism of macrolide resistance in three isolates; one erythromycin-resistant S. pneumoniae isolate had an A2059G mutation in 23S rRNA, and two isolates had substitution of GTG by TPS at positions 69 to 71 in ribosomal protein L4. All S. pyogenes isolates were susceptible to penicillin, and 11% were erythromycin resistant. Ribosomal methylation was the most common resistance mechanism for S. pyogenes (89.4%). These methylases were encoded by erm(A) [subclass erm(TR)] genes, and their expression was inducible in 96.6% of isolates. The rest of the erythromycin-resistant Russian S. pyogenes isolates (7.6%) had an efflux resistance mechanism. Telithromycin was active against 100% of pneumococci and 99.2% of S. pyogenes, and levofloxacin and quinupristin-dalfopristin were active against all isolates of both species. Keywords: Introduction : The problem of pneumococcal strains resistant to beta-lactams, macrolides, and other compounds has spread worldwide to such a degree that we are in the midst of a pandemic . Across the world, the incidence of pneumococcal strains that are fully resistant to penicillin G is rising relative to strains with intermediate penicillin resistance (, , ; A. L. Barry, P. C. Fuchs, and S. D. Brown, Letter, J. Antimicrob. Chemother. 40:139-140, 1997). The problem is complicated by the fact that the higher the penicillin G MIC for pneumococci is, the higher the prevalence of macrolide resistance is. In the United States, <5% of pneumococcal strains susceptible to penicillin G are macrolide resistant, while approximately 30% of penicillin-intermediate and >50% of penicillin-resistant strains are macrolide resistant . This relationship between penicillin G and macrolide resistance is not the same in all countries. For example, approximately 50% of Spanish strains with intermediate resistance to penicillin G are macrolide resistant, and in Northern Italy, where the rate of penicillin resistance is approximately 5%, 45% of penicillin-susceptible strains are resistant to macrolides . Pneumococcal strains resistant to beta-lactams are a particular problem in Central and Eastern Europe, especially in Hungary, Slovakia, Romania, and Bulgaria . Although Streptococcus pyogenes strains have not yet developed resistance to penicillin G, macrolide resistance in these strains is quite widespread throughout the world, particularly in Europe . Although the prevalence of macrolide resistance in the United States has been described as being low , a recent report has described an increase in macrolide resistance in the San Francisco Bay area . High prevalence of erythromycin resistance in pneumococci and group A streptococci in many Central and Eastern European countries has already been documented by our group . Macrolide resistance among Streptococcus pneumoniae and S. pyogenes is usually because of the ribosomal methylation of 23S rRNA [erm(B) or erm(A)] or active efflux of these antibacterials [mef(A)] . Recently, ribosomal mutations in 23S rRNA and ribosomal L4 proteins that confer macrolide resistance were described in clinical strains of S. pneumoniae . This study compared the activity of telithromycin with activities of erythromycin, azithromycin, clarithromycin, clindamycin, penicillin G, quinupristin-dalfopristin, and levofloxacin against S. pneumoniae and S. pyogenes organisms from centers throughout the Russian Federation. Macrolide resistance mechanisms of erythromycin-resistant isolates were determined. MATERIALS AND METHODS : Bacteria and antimicrobials. | Isolates were isolated consecutively from the various centers during 2000 and 2001 and were screened using optochin and bacitracin disk methodology. For the purpose of this study, the Russian Federation was divided into five regions: (i) central (Moscow, Kazan, Ryazan, Nizhnij Novgorod, and Yaroslavl); (ii) northwest (Smolensk and St. Petersburg); (iii) south (Krasnodar); (iv) Urals (Ekaterinburg and Chelyabinsk); and (v) Siberia (Tomsk, Irkutsk, Tumen, and Novosibirsk) . Organisms were then transported to Hershey Medical Center, Hershey, Pa., where they were stored frozen in double-strength skim milk (Difco Laboratories, Detroit, Mich.) at -70C until use. Before testing, S. pneumoniae and S. pyogenes cultures were checked for purity. The identity of S. pneumoniae strains was confirmed by optochin and, as necessary, by bile solubility testing. The confirmation of identity of S. pyogenes was done by bacitracin disk (Becton Dickinson, Cockeysville, Md.) and, as necessary, by latex agglutination. Telithromycin, quinupristin-dalfopristin, levofloxacin, erythromycin, azithromycin, and clarithromycin were obtained from Aventis Pharma, Infectious Division, Romainville, France. Clindamycin and penicillin G were purchased from Sigma Inc., St Louis, Mo. Susceptibility testing. | Susceptibility testing was performed by agar dilution on Mueller-Hinton agar (Difco) supplemented with 5% sheep blood . Inocula were prepared by suspending growth from overnight cultures in sterile Mueller-Hinton broth to a turbidity equal to a 0.5 McFarland standard. A 1:10 dilution was made in Mueller-Hinton broth so that the final inocula contained 5 x 104 CFU/spot. Plates were inoculated with a Steers replicator using 3-mm-diameter inoculating pins, delivering 1 mul of inoculum, and incubated overnight at 35C in air . The lowest concentration of antimicrobial showing no growth was read as the MIC. Standard quality control strains, including Staphylococcus aureus ATCC 29213 and S. pneumoniae ATCC 49619, were included with each run. Breakpoints used were those approved by the National Committee for Clinical Laboratory Standards for S. pneumoniae and S. pyogenes except for telithromycin, for which European breakpoints were used: susceptible, <=0.5 mug/ml; intermediate, 1.0 to 2.0 mug/ml; and resistant, >2 mug/ml (C. J. Soussy, F. Goldstein, A. Bryskier, H. Drugeon, J. Andrews, F. Baquero, O. Cars, D. Felmingham, B. Olsson-Liljequist, A. Rodloff, G. C. Schito, B. Wiedemann, and R. Wise, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 321, 2000). Determination of macrolide resistance mechanisms and gene sequencing. | Isolates found to be resistant to macrolides were screened for their resistance mechanisms. Primers and PCR conditions used for this work have been described previously: those for mef(A) and erm(B) were described by Sutcliffe et al. , those for erm(A) were described by Pankuch et al. , and those for 23S rRNA and L4 protein genes were described by Tait-Kamradt and coworkers . The PCR products after amplification of domain V of 23S rRNA, and entire genes encoding L4 and L22, were purified using a QIAquick PCR purification kit (QIAGEN, Valencia, Calif.) and sequenced using an Applied Biosystems model 373 DNA sequencer. All erythromycin-resistant strains were screened for the ability to induce resistance by the double-disk diffusion test using erythromycin and clindamycin disks . PFGE. | Pulsed-field gel electrophoresis (PFGE) was performed using a CHEF DR III apparatus (Bio-Rad, Hercules, Calif.) as described previously . FIG. 1. | Geographic locations of Russian centers providing S. pneumoniae Geographic locations of Russian centers providing S. pneumoniae and S. pyogenes isolates for this study. RESULTS : S. pneumoniae. | S. pneumoniae strains were isolated predominantly from respiratory sources (61%). Most of the pneumococci were from sputum (51%), ear (11.5%), and cerebrospinal fluid (11.1%). Patient ages varied from 1 month to 85 years. The peak age distribution was found in children 2 to 10 years old. Susceptibilities of the pneumococcal strains tested are shown in Tables and . Raised penicillin G MICs were found at a prevalence of <=5% in all regions of Russia tested except in Siberia, where the prevalence was 13.5%. In all areas, penicillin G-intermediate isolates for which the MICs were between 0.125 and 1.0 mug/ml far outnumbered penicillin G-resistant isolates for which the MICs were >=2.0 mug/ml. However, three penicillin G-resistant pneumococcal isolates for which the MICs were 4.0 to 8.0 mug/ml were isolated. Two isolates from Smolensk, for which the MICs of penicillin G were 4 and 8.0 mug/ml, were from the lung tissue of a 4-month-old infant and the sinus of a 40-year-old patient, respectively; these strains were resistant to macrolides but susceptible to clindamycin and telithromycin. The other strain, for which the MIC of penicillin G was 8.0 mug/ml, was isolated from the sputum of a 30-year-old patient in Novosibirsk; this strain was susceptible to all other agents tested. The cefotaxime MICs for these three strains were all 1 mug/ml. Macrolide resistance was rare in all areas tested. All pneumococcal strains were susceptible to telithromycin, levofloxacin, and quinupristin-dalfopristin at the breakpoints used. The MICs of the agents tested according to resistance mechanism are shown in Table . Telithromycin was the most active antibacterial tested against all strains of both species, regardless of their macrolide resistance mechanism. The prevalence of erythromycin resistance among S. pneumoniae strains was 2.5% (12 isolates). Most of the macrolide-resistant S. pneumoniae were isolated from sputum (63%). The mechanisms of macrolide resistance were mef(A) (five isolates; 41.7%) and erm(B) (four isolates; 33.3%) genes and ribosomal mutations (three isolates; 25%) in ribosomal protein L4 (two isolates) and in 23S rRNA (one isolate). Two strains had an L4 ribosomal protein gene mutation that caused substitution of the three amino acids GTG by TPS at positions 69 to 71. Interestingly these two strains were two penicillin-resistant strains from Smolensk (northwest Russia). As seen in Fig. the PFGE patterns of these strains after digestion with SmaI were similar. One strain, isolated from the pleural fluid of a 38-year-old patient in Ekaterinburg, had a mutation at position 2059 (Escherichia coli numbering system) in domain V of the gene encoding 23S rRNA, with replacement of A by G. The resistance phenotype of this strain was similar to the ML phenotype of three strains isolated in the United States and had the same mutation . The quinupristin-dalfopristin MIC for this isolate was 0.5 mug/ml, while the MICs of quinupristin and dalfopristin were 8 and >128 mug/ml, respectively. S. pyogenes. | Although most S. pyogenes strains were isolated from the respiratory tract (69%), a number of strains, especially those from the northwest, were isolated from wounds (27%). S. pyogenes isolates were from patients 2 weeks to 81 years old. Most were isolated from children and young adults (63%). Susceptibility patterns of S. pyogenes are presented in Tables and . All isolates were susceptible to penicillin G. Of 600 S. pyogenes isolates tested, 66 isolates (11%) were found to be resistant to macrolides. Erythromycin resistance was mainly seen in the Siberia, northwest, and central regions. The MICs of the agents tested according to resistance mechanism are shown in Table . The majority of macrolide-resistant S. pyogenes strains were isolated from wound (47%) and throat (36%) specimens. Erythromycin resistance was a result of the presence of the erm(A) gene subclass erm(TR) in 59 isolates (89.4%) or the mef(A) gene in five isolates (7.6%) . Except for two strains the expression of erm(A) genes was inducible by the double-disk method (96.6%). The susceptibility rate of telithromycin was 99.2%. Five isolates (7.6%) among the 66 erythromycin-resistant S. pyogenes isolates were intermediate to telithromycin (MICs, 1 mug/ml); two had constitutive erm(A) genes and three had mef(A) genes. The MICs of erythromycin, azithromycin, and clarithromycin for two strains with constitutive erm(A) were >64 mug/ml, and those of telithromycin were 1 mug/ml. Two strains had no known erythromycin resistance genes or mutations in genes encoding 23S rRNA or ribosomal proteins L4 or L22; these strains were from Yaroslavl (central Russia) and were isolated from throats of 22- and 40-year-old patients. For these strains, the MICs of erythromycin, azithromycin, clarithromycin, clindamycin, and telithromycin were 1, 2 to 4, 0.5 0.125, and 0.03 to 0.06 mug/ml, respectively. These strains had the same PFGE pattern after digestion with SmaI . The mechanism of macrolide resistance of these strains is currently under investigation. Among the strains with the erm(A) gene, four PFGE types containing more than one isolate were characterized. Eighty-six per cent of the erm(A) strains had one of these four PFGE patterns after digestion with SmaI . Twenty-two strains were PFGE type A, 5 were PFGE type B, 6 were PFGE type C, and 18 were PFGE type D. Three isolates from Smolensk, two from Tumen, and one from Ekaterinburg had the same PFGE type; these cities are located far from each other . FIG. 2. | PFGE pattern of three S. pneumoniae PFGE pattern of three S. pneumoniae strains with ribosomal mutations after digestion with SmaI. Lane 1, strain with mutation in 23S rRNA; lanes 2 and 3, strains with mutation in ribosomal protein L4. FIG. 3. | PFGE pattern of two S. pyogenes PFGE pattern of two S. pyogenes strains from Yaroslavl with unknown resistance mechanisms. FIG. 4. | PFGE patterns of S. pyogenes PFGE patterns of S. pyogenes strains with the erm(A) gene. TABLE 1 | Agar dilution MICs of agents tested against S. pneumoniae TABLE 2 | Rates of susceptibility of S. pneumoniae to various agents TABLE 3 | MICs according to mechanism of resistance in 12 macrolide-resistant S. pneumoniae isolates TABLE 4 | Agar dilution MICs of agents tested against S. pyogenes TABLE 5 | Rates of susceptibility of S. pyogenes to various agents TABLE 6 | MICs according to mechanism of resistance in 66 macrolide-resistant S. pyogenes isolates DISCUSSION : In the present study, a very low prevalence of drug resistance was found for S. pneumoniae and S. pyogenes except in Siberia (both species) and the northwest and central regions (S. pyogenes). The pneumococci for which the penicillin G MICs were highest (4 to 8.0 mug/ml) came from Novosibirsk, a city in south central Siberia, and two strains from Smolensk. The Siberian isolate was susceptible to other drug classes, but the Smolensk strain was macrolide- and azalide-resistant, with mutation in ribosomal protein L4. The reason for these geographic differences is unclear at present but may be related to travel patterns and/or antibacterial use in the different parts of Russia. This is currently under investigation. Irrespective of resistance to other drugs or drug classes, all pneumococci were susceptible to telithromycin, levofloxacin, and quinupristin-dalfopristin. It is noteworthy that the low prevalence of drug-resistant pneumococci and S. pyogenes in most areas of Russia parallels those found recently by our group in Baltic countries such as Lithuania and Latvia (P. C. Appelbaum et al., Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstracts 2153 and 2154, 2000). Data from the Urals and the southern part of Russia are preliminary and must be confirmed by testing more strains. The prevalence of drug resistance in pneumococci and S. pyogenes throughout the cities in the Russian Federation studied is currently low. However, the apparent presence of a nidus of increased resistance in Siberia, as well as a multiresistant strain from Smolensk, indicates potential for the spread of these organisms in the future. The prevalence of macrolide resistance among pneumococci was generally low in all of the regions of Russia studied (2.5%). However, rRNA and ribosomal protein mutations were significant mechanisms of macrolide resistance among these isolates (3 of 12). The prevalence of macrolide resistance was higher in S. pyogenes than in S. pneumoniae and was associated with four clones. Five strains from three distinct regions of Russia were found to belong to same clone. In summary all pneumococci and 99.2% of S. pyogenes isolates, irrespective of their erythromycin resistance, were susceptible to telithromycin at a breakpoint of 0.5 mug/ml, as well as being uniformly susceptible to levofloxacin and quinupristin-dalfopristin. Telithromycin, therefore, has considerable potential for the empirical treatment for community-acquired infections caused by macrolide-susceptible and -resistant S. pneumoniae and S. pyogenes. Backmatter: PMID- 12183223 TI - Importance of the Fourth Alpha-Helix within the CAP Homology Domain of Type II Topoisomerase for DNA Cleavage Site Recognition and Quinolone Action AB - We report that point mutations causing alteration of the fourth alpha-helix (alpha4-helix) of the CAP homology domain of eukaryotic (Saccharomyces cerevisiae) type II topoisomerases (Ser740Trp, Gln743Pro, and Thr744Pro) change the selection of type II topoisomerase-mediated DNA cleavage sites promoted by Ca2+ or produced by etoposide, the fluoroquinolone CP-115,953, or mitoxantrone. By contrast, Thr744Ala substitution had minimal effect on Ca2+- and drug-stimulated DNA cleavage sites, indicating the selectivity of single amino acid substitutions within the alpha4-helix on type II topoisomerase-mediated DNA cleavage. The equivalent mutation in the gene for Escherichia coli gyrase causing Ser83Trp also changed the DNA cleavage pattern generated by Ca2+ or quinolones. Finally, Thr744Pro substitution in the yeast type II topoisomerase rendered the enzyme sensitive to antibacterial quinolones. This study shows that the alpha4-helix within the conserved CAP homology domain of type II topoisomerases is critical for selecting the sites of DNA cleavage. It also demonstrates that selective amino acid residues in the alpha4-helix are important in determining the activity and possibly the binding of quinolones to the topoisomerase II-DNA complexes. Keywords: Introduction : DNA topoisomerases are enzymes that catalyze changes in the topology of DNA via a mechanism involving the transient breakage and rejoining of phosphodiester bonds in the DNA backbone . Studies in both prokaryotic and eukaryotic cells have demonstrated the importance of topoisomerases in transcription, DNA replication, and chromosome segregation . Type II topoisomerases (Top2p) are conserved among eukaryotes. Gyrase and topoisomerase IV are bacterial Top2 enzymes with significant sequence similarity to eukaryotic Top2 proteins . Eukaryotic Top2 enzymes are homodimeric and cleave one DNA duplex (the G, or gate, segment) using two tyrosine residues (one from each Top2p monomer) to attack phosphodiester bonds on both DNA strands. In this process, each enzyme monomer becomes covalently attached to the 5' end of the cleaved DNA by a phosphotyrosine linkage. Top2p then opens the DNA, bridging the gap in the G segment. A second DNA duplex (the T, or transported, segment) passes through the gap. Thereafter, the G segment is resealed, and the T segment is liberated from the enzyme . Under physiological conditions, the covalent Top2p-DNA complexes (referred to as cleavage complexes) are normally short-lived intermediates in the catalytic cycle of the enzyme. Top2p requires the presence of a divalent cation for catalytic activity, which is Mg2+ under physiological conditions. However, when magnesium is replaced by Ca2+, a higher level of Top2p is trapped in a Ca2+-promoted covalent enzyme-DNA complex, and DNA double- and single-stranded DNA cleavage can be detected . Beyond its vital cellular functions, Top2p is the primary cytotoxic target for some of the most active drugs used for the treatment of human cancers (see below) . In addition, inhibitors of bacterial type II topoisomerases, such as quinolones, are among the most widely prescribed antibacterial drugs . Top2p inhibitors can be divided in two groups: Top2p catalytic inhibitors and Top2p poisons . Top2p catalytic inhibitors ("suppressors") are defined as drugs that inhibit enzyme activity but do not stabilize DNA cleavage complexes. Proflavine and 9-aminoacridine belong to this category . Top2p poisons inhibit the enzyme by increasing the steady-state levels of DNA cleavage complexes . Hence, they convert Top2p into a physiological toxin that creates protein-linked DNA breaks in the genome of treated cells . Top2p poisons can, in turn, be subdivided in two groups: DNA intercalators ---which include doxorubicin, mitoxantrone, amsacrine, and ellipticines ---and nonintercalators, whose main representatives are the demethylepipodophyllotoxins etoposide (VP-16), teniposide (VM-26), and the quinolones , such as ciprofloxacin and norfloxacin, which are specific for bacterial Top2 enzymes . Previous studies showed that substitution of Gly747 to Glu of Saccharomyces cerevisiae Top2p and B show enzyme-specific differences. Cleavage was reduced at several sites for Top2pS740W and Top2pQ743P compared to Top2pWT. For example, cleavage at positions 3124, 3147, 3156, 3226, 2706 or 2707, 2735, 2741, 2745, and 2754 was reduced for Top2pS740W and at that at positions 3094, 3170, 2754, and 2735 was reduced for Top2pQ743P. Most reduction in Ca2+-promoted cleavage was seen for Top2pT744P , which is in contrast to the intensity of the drug-induced cleavage sites generated by Top2pT744P . Cleavage was enhanced shows a comparison of the DNA cleavage patterns observed in the presence of two different quinolones, CP-115,953 and ciprofloxacin. Because the GyrAS83W mutant is quinolone resistant and therefore bound to give less cleavage than gyraseWT, we chose drug concentrations (100 muM) that are approximately three times the 50% inhibitory concentration for supercoiling (data not shown). As for yeast Top2p, the S83W substitution changed the enzyme cleavage pattern. Several cleavage sites were reduced in intensity for GyrAS83W in the presence of CP-115,953 and ciprofloxacin (e.g., 2747, 2744, 2772, 2770, 2991, 3020, 3017, and 3020), whereas other sites were enhanced (e.g., 2722, 2893, 2768, 2794, 2805, and 2962). These results demonstrate that the GyrAS83W mutation markedly affects the DNA sequence specificity of DNA gyrase. FIG. 1. | CAP homology domain and positions of Ser740, Gln743, and Thr744 in yeast Top2p. CAP homology domain and positions of Ser740, Gln743, and Thr744 in yeast Top2p. (A) Illustration of dimeric S. cerevisiae topoisomerase II structure (Protein Data Bank accession number ). Drawings were generated using the program QUANTA (version 97). The helical ribbon representation shows the 92-kDa fragment of the yeast enzyme with a DNA fragment modeled into each of the putative DNA-binding sites . The CAP homology domain of each protomer is highlighted in green. (B) Close view of the putative DNA-binding region, presenting the proposed proximity of the alpha4-helix within the CAP homology domain, including Ser 740, Gln 743, and Thr744 (in a stick model), to DNA. (C) Alignment of protein sequence for the yeast (Sc Top2p), human Top2palpha (Hu Top2palpha) and E. coli gyrase. The mutated residues studied in the present report are indicated by arrowheads. Shaded residues correspond to the alpha4-helix . No shading is shown for human Top2p because of lack of structural data. Boxed regions indicate similarity between the amino acid residues. Ser763 in the human Top2palpha sequence corresponds to Ser740 in yeast Top2p. FIG. 2. | Comparison of Ca2+-promoted DNA cleavage patterns of yeast Top2pWT, Top2pS740W, Top2pQ743P, Top2pT744P, and Top2pT744A as well as gyraseWT and GyrAS83W. Comparison of Ca2+-promoted DNA cleavage patterns of yeast Top2pWT, Top2pS740W, Top2pQ743P, Top2pT744P, and Top2pT744A as well as gyraseWT and GyrAS83W. DNA fragments from the human c-myc first intron (A) and from the junction between the c-myc first intron and first exon (B) were prepared by PCR. For each fragment, 5'-end-labeling was performed with 32P. Yeast Top2p reactions were performed at 37C for 30 min (A and B), and gyrase reactions were performed at 25C for 2 h (C). In all reactions, the CaCl2 concentration was 5 mM. Lanes: Control, without Top2 enzyme; Purine, ladder obtained after formic acid reaction; y WT, yeast wild-type Top2p; y S74 0W, yeast Top2pS740W; y Q743P, yeast Top2pQ743P; y T744P, yeast Top2pT744P; y T744A, yeast Top2pT744A; gyrWT, gyraseWT from E. coli; GyrAS83W, mutant Ser83Trp gyrase. Numbers correspond to genomic positions of the nucleotide covalently linked to the enzyme via the 5' phosphate. FIG. 3. | DNA cleavage patterns generated by yeast Top2pWT, Top2pS740W, Top2pQ743P, Top2pT744P, and Top2pT744A enzymes in the presence of etoposide, the fluoroquinolone CP-115,953, and the intercalator mitoxantrone. DNA cleavage patterns generated by yeast Top2pWT, Top2pS740W, Top2pQ743P, Top2pT744P, and Top2pT744A enzymes in the presence of etoposide, the fluoroquinolone CP-115,953, and the intercalator mitoxantrone. DNA fragments from the junction between the c-myc first intron and first exon between positions 2671 and 3072 were prepared by PCR using one primer labeled with 32P at the 5' terminus. The left-hand panels show labeling of the upper DNA strand at position 2671. The right-hand panels show that labeling of the lower DNA strand was at position 3072. Top2 enzymes are indicated above each lane. Lanes: y WT, yeast wild-type Top2p; y S740W, yeast Top2pS740W; y Q743P, yeast Top2pQ743P; y T744P, yeast Top2pT744P; y T744A, yeast Top2pT744A. Concentrations were 100 muM for etoposide (A) and CP-115,953 (B) and 1 muM for mitoxantrone (C). Reactions were performed at 37C for 30 min in the presence of 5 mM MgCl2. Purine ladders were obtained after formic acid reaction. Numbers correspond to genomic positions of the nucleotide covalently linked to Top2p via the 5' phosphate. Double-headed arrows correspond to DNA cleavage sites with a 4-bp stagger that represent potential DNA double-strand breaks. Comparison of yeast Top2pT744Pand E. coli gyrase DNA cleavage sites generated in the presence of quinolones. : In contrast to the mutation causing Ser740Trp, which confers quinolone resistance, the Thr744Pro substitution causes hypersensitivity to CP-115,953 . Figure compares DNA cleavage produced by CP-115,953, ciprofloxacin, and norfloxacin for GyrAWT, Top2pWT, and Top2pT744P. Remarkably, prominent cleavage was observed for Top2pT744P in the presence of ciprofloxacin and norfloxacin, whose antibacterial activities are due to specific inhibition of bacterial Top2p and have no effect on mammalian Top2p. A number of cleavage sites stimulated by Top2pT744P in the presence of quinolones matched the E. coli DNA gyrase sites (e.g., at positions 3202, 3179, 3145, 3121, 3108, 3066, and 3111). However, other sites were different for DNA gyrase. Enhanced sensitivity to antibacterial quinolones by the Thr744Pro substitution in yeast Top2p was also detected in other DNA fragments (data not shown). FIG. 5. | Generation of DNA cleavage by the yeast Top2pT744P mutant in the presence of antibacterial quinolones. Generation of DNA cleavage by the yeast Top2pT744P mutant in the presence of antibacterial quinolones. A 254-bp DNA fragment from the c-myc first intron was prepared. (A) Labeling of the upper DNA strand at position 3035. (B) Labeling of the lower DNA strand at position 3288. Drugs (100 muM each) and enzymes are indicated above the lanes. Lanes: yWT, yeast wild-type Top2p; Y T744P, Top2pT744P. Numbers correspond to genomic positions of the nucleotide covalently linked to Top 2p. Double-headed arrows correspond to DNA cleavage sites with a 4-bp stagger that represent potential DNA double-strand breaks. Base preference of the CP-115,953-stabilized cleavage complexes for mutant Top2T744P enzyme. : Because we observed that the Thr744Pro substitution enhanced sensitivity to quinolones (including the fluoroquinolone CP-115,953) and produced common DNA cleavage sites with E. coli DNA gyrase, we examined whether this particular amino acid substitution affected the DNA base preference of yeast Top2p in the presence of CP-115,953. Cleavage sites for the three c-myc DNA fragments and one c-jun fragment (see Materials and Methods) were analyzed for both DNA strands. For the wild-type yeast Top2p, CP-115,953 preferentially stabilized sites with C or T at position -1 (C-1 + T-1: 49 out of 65 sites) and A or G at position +1 (A+1 + G+1: 54 out of 65 sites). Complementary preferences (although slightly weaker) were seen for T+4 and G+5. Top2pT744P also demonstrated a preference for T at position -1, which was higher than that for the wild-type enzyme (41 out of 89 sites), in combination with preference for G at position +1 (44 out of 89 sites). However, preferences for C at position -1 and A at position +1 were not detectable in the mutant protein (Thr744Pro). Moreover, complementary preference also changed to A+5. Because bacterial Top2 enzymes are also sensitive to CP-115,953 , we analyzed the base preferences for DNA gyrase. The observed preferences for T at position -1 (41 out of 107 sites) and G at position +1 (50 out of 107 sites) are in agreement with previous reports obtained with the antibacterial fluoroquinolones . These data indicate that enhanced quinolone sensitivity of the mutation causing Thr744Pro is associated with a shift in base sequence preferences, resembling that of DNA gyrase. FIG. 6. | Probability of the observed base frequency deviations at Top2p cleavage sites for the yeast wild-type enzyme, for Top2pT744P and for E. coli Probability of the observed base frequency deviations at Top2p cleavage sites for the yeast wild-type enzyme, for Top2pT744P and for E. coli gyrase in the presence of CP-115,953. Position 0 corresponds to the cleavage site, and positions -1 and +1 to the bases immediately 3' and 5' from the cleavage site, respectively. The panels present the probability of the observed base frequency deviations from expectation for the indicated enzyme. On the y axis, P is the probability of observing that deviation or more, either as excess (above baseline) or deficiency (below baseline), relative to the expected frequency of each individual base . Cleavage sites for the yeast wild-type Top2p (A), the yeast Top2pT744P (B), and E. coli gyrase (C) were analyzed after treatment with 100 muM CP-115,953. DISCUSSION : The A' domains of eukaryotic Top2 enzymes (homologous to the amino terminus of the A subunit of gyrase) comprise the primary DNA binding sites and include the catalytic tyrosine Y782 for the yeast S. cerevisiae , which is homologous to Y122 for E. coli gyrase . DNA footprinting analyses indicate that for eukaryotic Top2p and bacterial topoisomerase IV, similar15 to 35 bp of DNA is tightly bound to the enzyme . In the case of gyrase, a more extensive stretch of DNA (similar130 bp) is protected by the enzyme . However, the central section of this region around the DNA cleavage site is very similar in size to that bound by Top2p and topoisomerase IV . The length of this DNA segment is consistent with the combined length of the A' grooves present in the Top2p and GyrA structures . Within the A' domain of Top2p, a 29-kDa fragment containing the active-site tyrosine and the helix-turn-helix motif from the CAP homology domain can be cross-linked to DNA , and protein footprinting suggest that the binding of DNA protects lysines from chemical modification within the same region . CAP homology domains contain typically a three-alpha-helix bundle backed by beta-sheets. Two adjacent alpha-helices, usually the second helix (referred to as alpha3-helix in yeast Top2p according to reference )and the third helix (referred to as alpha4-helix of the bundle in yeast Top2p according to reference ), are connected by a short turn. This motif known as the helix-turn-helix is responsible for many of the critical contacts between CAP-like proteins and DNA. The second helix of this motif generally inserts into the major groove of DNA with the turn contacting the phosphodiester backbone . Our observations suggest that specific substitutions in the alpha4-helix (Thr744Pro, Gln743Pro, and Ser740Trp of eukaryotic Top2p), which corresponds to the third helix of the CAP homology domain, can change the enzyme-DNA interactions, as reflected by differences in DNA cleavage patterns induced in the presence of Ca2+ or in the presence of intercalating and nonintercalating drugs. We also observed differences in base sequence preferences. We interpret these results as an indication that the alpha4-helix of the Top2p CAP homology domain is critical for DNA sequence recognition prior to cleavage of the G strand, which is consistent with the proposed binding of DNA to yeast Top2p with the alpha4-helix close to the catalytic tyrosine and to the ends of the cleaved DNA . Figure presents the positions of Ser740, Gln743, and Thr744 in the yeast Top2 protein . These conserved residues tend to be buried inside the protein, while the nonconserved residues Ser740 and Gln743 are more solvent accessible . The nonconserved amino acid residues might contribute to the observed differences in DNA sequence selectivity and drug sensitivity between the bacterial, the yeast, and the human Top2 enzymes . Other regions of Top2p must interact with different DNA segments, such as the T segment for the strand passage reaction. The B' domain (homologous to B subunit of gyrase) is also essential for DNA binding and cleavage . Recently it was proposed that DNA breakage and rejoining by Top2p involve the coordinated action of the CAP homology domain in the A' fragment containing the active-site tyrosine together with the Rossmann fold in the B' fragment (, , ; J. G. Heddle, and A. Maxwell, submitted for publication). The Rossmann fold, which is important for DNA binding and cleavage reactions, contains a number of highly conserved acidic residues that probably bind divalent metals . Structural and biochemical studies suggest that the active-site tyrosine in the DNA-binding domain of one protomer cooperates with several residues in the Rossmann fold of the other protomer . Our data suggest that not only the active-site tyrosine and the Rossmann fold domain but also the alpha4-helix of the CAP domain are involved in the concerted molecular actions, including DNA interactions leading to sequence-specific cleavage of the G-strand duplex DNA. We have shown that the Ca2+- and drug-generated DNA cleavage patterns of Top2pT744P and Top2p Q743P are different from those of Top2pT744A and Top2pWT and that Top2pT744A is comparable to Top2pWT. Since the Top2p Q743P substitution does not lead to enhanced drug sensitivity , the change in the observed pattern of Top2p cleavage sites cannot arise solely from global quantitative alterations. Also, since Top2pT744A does not have a pattern of cleavage that differs from the wild-type enzyme, Thr744 probably does not play a crucial role in determining cleavage site selection by the enzyme. Since detailed structural data for any of the mutant proteins are not yet available, we modeled the specific molecular changes that might result from the Thr744Ala/Pro and Gln743Pro substitutions in the yeast Top2p structure . Molecular dynamics simulations of the Thr744Pro mutant Top2p predicted structural changes within the CAP homology domain (data not shown; see also reference ). Compared to wild-type Top2p and Top2pT744A, the Thr744Pro substitution was associated with a kink in the alpha4-helix and a bending angle of approximately 37o. The helical structure was preserved on both sides of the kink (data not shown). These simulations suggest that an alteration in the secondary structure of the alpha4-helix can alter protein-DNA interactions. It is important to note, however, that changes in other parts of the protein also occurred. Therefore, it is premature to conclude that all of the biochemical alterations we observed in the mutant proteins arise solely from changes in the alpha4-helix. A key issue in understanding the mechanism of action of Top2p targeting drugs is the determination of where the drugs bind in the Top2p/DNA complex. Our present findings indicate that the same Top2p domain that presumably interacts with DNA is also critical for drug interaction in the Top2p/DNA/drug ternary complex. In particular, the solvent-exposed residues Ser740, Gln743, and to a lesser extent Thr744, which affect drug activity, are important, directly or indirectly, for drug binding to the Top2p/DNA complex . An intriguing finding arising from this work is that a quinolone-hypersensitive yeast Top2p mutant, resulting in Thr744Pro substitution, changed the DNA cleavage pattern in the presence of quinolones so that it more closely resembled the DNA cleavage pattern of a quinolone-sensitive prokaryotic topoisomerase II. Quinolones act against eukaryotic topoisomerase II by accelerating the rate of DNA cleavage . Recent studies have shown that along with gyrase and topoisomerase IV of E. coli, quinolones can also inhibit religation . A plausible model is that the quinolone-hypersensitive mutant allows quinolones to inhibit religation as well. By this hypothesis, the action of quinolones against Top2pThr744Pro more closely resembles the action of quinolones against prokaryotic topoisomerase II, demonstrating conservation of mechanisms of drug action between eukaryotic and prokaryotic type II topoisomerases. Thus, the quinolone-protein interaction domains between the prokaryotic and eukaryotic enzymes may be comparable, with only limited differences preventing antibacterial quinolones from acting against eukaryotic Top2p. null: null Preparation of end-labeled DNA fragments by PCR. : Three sets of labeled DNA fragments were prepared from the human c-myc gene by PCR. A 254-bp DNA fragment from the first intron was prepared between positions 3035 and 3288 (GenBank accession ) using the following oligonucleotides: 5'-GTAATCCAGAACTGGATCGG-3' for the upper strand and 5'-ATGCGGTCCCTACTCCAAGG-3' for the lower strand (annealing temperature, 56C). A 401-bp DNA fragment from the junction between the first intron and the first exon was prepared between positions 2671 and 3072 using the following oligonucleotides: 5'-TGCCGCATCCACGAAACTTT-3' for the upper strand and 5'-TTGACAAGTCACTTTACCCC-3' for the lower strand (annealing temperature, 60C). A 480-bp fragment from the first exon containing promoters P1and P2 was prepared between positions 2265 and 2745 using the following oligonucleotides: 5'-GATCCTCTCTCGCTAATCTCCGCCC-3' for the upper strand and 5'-TCCTTGCTCGGGTGTTGTAAGTTCC-3' for the lower strand (annealing temperature, 70C). A 213-bp fragment from the human c-jun gene was prepared between positions 5'-TGTTGACAGCGGCGGAAAGCAGS-3' for the upper strand and 5'-CGTCCTTCTTCTCTTGCGTGGCTCT-3' for the lower strand (annealing temperature, 64C). Single-end labeling of these DNA fragments was obtained by 5'-end labeling of the specific primer oligonucleotide. Ten picomoles of DNA was incubated for 60 min at 37C with 10 U of T4 polynucleotide kinase and 10 pM [gamma-32P]ATP (100 muCi) in kinase buffer (70 mM Tris-HCl, pH 7.6; 0.1 M KCl; 10 mM MgCl2; 5 mM dithiothreitol; bovine serum albumin [0.5 mg/ml]). Reactions were stopped by heat denaturation at 70C for 15 min. After purification using Sephadex G-25 columns (Boehringer Mannheim), the labeled oligonucleotides were used for PCR. Approximately 0.1 mug of the c-myc DNA that had been restricted by SmaI and PvuII (fragment spanning positions 2265 to 2745) and XhoI and XbaI (fragment spanning positions 2671 to 3072 and fragment 3035 to 3288) was used as a template for the PCR. Ten picomoles of each oligonucleotide primer, one of them being 5' labeled, was used in 22 temperature cycle reactions (each cycle with 94C for 1 min, annealing for 1 min, and 72C for 2 min). The last extension was for 10 min, and DNA was purified using PCR Select-II columns (5Prime-3Prime Inc., Boulder, Colo.). Overexpression and purification of yeast topoisomerase II and E. coli DNA gyrase. : Wild-type yeast Top2p, Ser740Trp, Gln743Pro, Thr744Pro, and Thr744Ala proteins were overexpressed using YEpTOP2-PGAL1 or YEptop2-S*W-PGAL1 and yeast strain JEL1t1- and purified to homogeneity as previously described (59). The detailed procedure has been described elsewhere . GyrA (gift of N. A. Gormley), GyrAS83W (gift of C. J. R. Willmott), and GyrB were purified as described previously . GyrAS83W was shown to have the expected level of quinolone-resistant supercoiling activity when complexed with wild-type gyrB. Topoisomerase II-generated DNA cleavage reactions. : For the yeast Top2p experiments, DNA fragments (5 x 104 to 10 x 104 dpm/reaction) were equilibrated with drug or with the corresponding 1% dimethyl sulfoxide concentration in 10 mM Tris-HCl (pH 7.5)-50 mM KCl-5 mM MgCl2-2 mM dithiothreitol-0.1 mM Na2EDTA-1 mM ATP-bovine serum albumin (15 mug/ml) for 5 min before the addition of 8 U (80 ng) of purified Top2p in a 10-mul final reaction volume. Unless otherwise indicated, reactions were for 30 min at 37C. Reactions were stopped by adding EDTA and sodium dodecyl sulfate (25 mM and 1% final concentrations, respectively) and were further digested with proteinase K (0.4-mg/ml final concentration for 30 min at 55C). For the DNA gyrase experiments, similar conditions were used except that the DNA cleavage reactions were performed with 10 to 15 U of purified E. coli gyrase for 120 min at 25C . Electrophoresis and base sequence analysis. : For DNA sequence analysis, samples were precipitated with ethanol and resuspended in 5 mul of loading buffer (80% formamide, 10 mM NaOH, 1 mM EDTA, 0,1% xylene cyanol, 0.1% bromphenol blue). Samples were heated to 95C for 5 min and thereafter loaded onto DNA sequencing gels (7% polyacrylamide; 19:1 acrylamide-bisacrylamide) containing 7 M urea in 1x Tris-borate-EDTA buffer. Electrophoresis was performed at 2,500 V (60 W) for 2 to 3 h. The gels were dried on Whatman no. 3MM paper sheets and visualized using a PhosphorImager (Molecular Dynamics, Sunnyvale, Calif.) and ImageQuant software. The determination of preferred bases around Top2p cleavage sites was done as described previously . Molecular dynamics simulation. : Modeling of the Thr744Pro substitution into published crystal structure data of a 92-kDa fragment of yeast Top2p (PDB identification code 1bgw) was performed with the molecular modeling package QUANTA (version 96) run on both an SGI Crimson and an SGI Octane workstation. Energy calculations and related operations, such as energy minimizations and the molecular dynamics simulations, were carried out with QUANTA's associated molecular mechanics program CHARMm, version 23.1, run on a separate computation host, a Digital 2100 AlphaServer 4/275, using the MSI parameter set version 22.0. In order to generate a low-energy starting geometry for the subsequent molecular dynamics simulation, each of the wild-type and mutant structures was energy minimized in 5,000 steps of Adopted-Basis Newton-Raphson minimization. Each of these minimized structures was subjected to a total of 300 ps of molecular dynamics simulation using the Verlet algorithm for integration of the equations of motion. Using a step size of 1 fs, the structures were heated from 0 to 310 K in approximately 10 ps (9,982 steps), then equilibrated at 310 K for 40 ps (40,000 steps) with an equilibration frequency of 200 steps, and finally subjected to a simulation at 310 K for 250 ps (250,000 steps). Snapshot frames containing the current coordinates were written out every 50 steps. RESULTS : Alteration of Ca2+-promoted DNA cleavage by Ser740Trp, Gln743Pro, Thr744Pro, and Thr744Ala substitutions in yeast Top2p and by the Ser83Trp substitution in bacterial GyrA. | It is well known that DNA cleavage by type II DNA topoisomerases can be stimulated by a variety of drugs , but cleavage can also be stimulated when Ca2+ is substituted for Mg2+ in the enzyme reaction . In this case it is likely that the cleavage specificity reflects the natural sequence specificity of the enzyme. The Ca2+-promoted, drug-independent DNA cleavage sites generated by yeast wild-type Top2p (Top2pWT) and four yeast mutant proteins (Top2pS740W, Top2pQ743P, Top2pT744P, and Top2pT744A) were mapped in two different c-myc DNA fragments. Figure To further investigate differences between Top2pS740W, Top2pQ743P, Top2pT744P, Top2pT744A, and Top2pWT, we compared drug-stimulated DNA cleavage patterns. Etoposide results are shown in Fig. . Results for the fluoroquinolone CP-115,953 in are shown in Fig. , and those for mitoxantrone are shown in Fig. . Top2pS740W (as well as the corresponding yeast mutant) has been previously characterized as being resistant to the fluoroquinolone CP-115,953 and hypersensitive to etoposide . Compared with Top2pWT and the other mutant proteins, several cleavage sites stimulated by etoposide were markedly enhanced with Top2pS740W, e.g., at positions 2711 to 2712, 2781, 2983, 3011, 3019, and 3026 . Proline substitutions at positions 744 and 743 also changed the distribution of Top2p-mediated DNA cleavages. These effects were more pronounced for Top2pT744P than for Top2pQ743P . By contrast, Top2pT744A behaved very similarly to Top2pWT . Together, the data shown in Fig. demonstrate that single mutations in the alpha4-helix of Top2p (Ser740Trp, Gln743Pro, and Thr744Pro) changed enzyme-specific DNA cleavage patterns produced by nonintercalating and intercalating drugs. Differences in quinolone-generated DNA cleavage patterns by the mutant GyrAS83W and wild-type DNA gyrase. : We next tested the role of the corresponding alpha4-helix in the bacterial topoisomerase II, DNA gyrase. Figure FIG. 4. | DNA cleavage patterns generated by gyraseWT and mutant GyrAS83W in the presence of the quinolones CP-115,953 and ciprofloxacin. DNA cleavage patterns generated by gyraseWT and mutant GyrAS83W in the presence of the quinolones CP-115,953 and ciprofloxacin. The same DNA fragments presented in Fig. were used. The left-hand panel shows labeling of the upper DNA strand; the right-hand panel shows labeling of the lower DNA strand. Drugs and enzymes are indicated above each lane. Concentrations were 100 muM for CP-115,953 and ciprofloxacin, respectively. Cleavage reactions were performed at 25C for 2 h in the presence of 5 mM MgCl2. Purine ladders were obtained after formic acid reaction. Lanes: Control, no Top2p, no drug treatment; Gyr WT: gyraseWT from E. coli; GyrAS83W, mutant Ser83Trp gyrase. Numbers correspond to genomic positions of the nucleotide covalently linked to Top2p. Double-headed arrows correspond to DNA cleavage sites with a 4-bp stagger that represent potential DNA double-strand breaks. Backmatter: PMID- 12183262 TI - Mutation in 23S rRNA Associated with Macrolide Resistance in Neisseria gonorrhoeae AB - Fifty-six azithromycin-resistant (MICs, 2.0 to 4.0 mug/ml) Neisseria gonorrhoeae strains with cross-resistance to erythromycin (MICs, 2.0 to 64.0 mug/ml), isolated in Canada between 1997 and 1999, were characterized, and their mechanisms of azithromycin resistance were determined. Most (58.9%) of them belonged to auxotype-serotype class NR/IB-03, with a 2.6-mDa plasmid. Based on resistance to crystal violet (MICs >= 1 mug/ml), 96.4% of these macrolide-resistant strains appeared to have increased efflux. Nine of the eleven strains selected for further characterization were found to have a promoter region mtrR mutation, a single-base-pair (A) deletion in the 13-bp inverted repeat, which is believed to cause overexpression of the mtrCDE-encoded efflux pump. The two remaining macrolide-resistant strains (erythromycin MIC, 64.0 mug/ml; azithromycin MIC, 4.0 mug/ml), which did not have the mutation in the mtrR promoter region, were found to have a C2611T mutation (Escherichia coli numbering) in the peptidyltransferase loop in domain V of the 23S rRNA alleles. Although mutations in domain V of 23S rRNA alleles had been reported in other bacteria, including E. coli, Streptococcus pneumoniae, and Helicobacter pylori, this is the first observation of these mutations associated with macrolide resistance in N. gonorrhoeae. Keywords: Introduction : Erythromycin is sometimes used to treat gonorrhea if a patient has a cephalosporin allergy or a history of immediate and/or anaphylactic reaction to penicillins . In some parts of the world, azithromycin, a 15-membered macrolide, is used to treat gonorrhea (1 g orally in a single dose) , but this dosage is not recommended for such use in Canada . In Canada, azithromycin (1 g dose) is used either to treat Chlamydia trachomatis infections or in conjunction with expanded spectrum cephalosporins or quinolones to treat coinfections of gonococcus and C. trachomatis . The increased use of azithromycin in treating other diseases may increase the selective pressure for macrolide-resistant gonococcus. Macrolides such as erythromycin and azithromycin act by binding to the 50S subunit of bacterial ribosomes and restrain protein synthesis by inhibiting the elongation of peptide chains . The mechanisms of resistance to macrolides include efflux of these antibiotics and modification of the ribosomal target by modification of enzymes or mutations to reduce the affinity of the antibiotics for ribosomes. A number of bacterial efflux systems which confer macrolide resistance have been described . The first efflux pump described for Neisseria gonorrhoeae was the MtrC-MtrD-MtrE system, encoded by the mtrRCDE operon, which exports hydrophobic agents including dyes such as crystal violet and macrolides such as azithromycin and erythromycin . A fourth Mtr protein (MtrF) may also be associated with this efflux system . The MtrR protein regulates the expression of MtrCDE, and increased expression of the MtrCDE pump is due to either the lack of the repressor (MtrR) protein (through deletion or insertional inactivation of mtrR) or a mutation of the promoter region of this repressor gene . More recently, another macrolide efflux pump, encoded by the mef gene, which had originally been described for some of the gram-positive organisms , has now been found in clinical strains of N. gonorrhoeae . Mechanisms of resistance related to modification of the ribosomal target by methylases have been described for Streptococcus pneumoniae , Staphylococcus aureus , and N.gonorrhoeae . There are reports of the emergence of azithromycin-resistant gonococci which show cross-resistance with erythromycin . One study showed that strains from Uruguay and Seattle, Washington, had acquired from other bacterial species one or more of the 23S rRNA methylases [erm(B), erm(C) and erm(F)] which modified the antimicrobial targets . Another ribosomal modification, which involves mutations in the peptidyltransferase loop in domain V of 23S rRNA, also confers resistance to macrolides . This mechanism has been described in a variety of organisms: Escherichia coli , Streptococcus pneumoniae , and Helicobacter pylori . Mutants selected for resistance to macrolides have a substitution at nucleotide position 2058 (E. coli coordinates; GenBank accession number ) or at nearby nucleotides, suggesting their involvement in the binding of these macrolides to domain V of 23S rRNA . Point mutations in ribosomal proteins L4 (rplD) and L22 (rplV) conferring resistance to macrolides have been described for E. coli . More recently, mutations of ribosomal protein L4 were also described for macrolide-resistant S. pneumoniae . Proteins L4 and L22 primarily bind to domain I of 23S rRNA, but macrolide resistance mutations in these proteins cause a change in the conformation in domains II, III, and V affecting the action of the macrolides for domain V of 23S rRNA . FIG. 1. | Secondary structure of the peptidyltransferase loop in domain V of 23S rRNA. Secondary structure of the peptidyltransferase loop in domain V of 23S rRNA. Nucleotide numbering is that of E. coli 23S rRNA ; the corresponding N. gonorrhoeae, S. pneumoniae , and H.pylori numbers are in parentheses. The C->T mutation at position 2611 (position 2599 in alleles 1 and 2 and position 2605 in allele 4) in N. gonorrhoeae strains 20869 and 20870, also previously reported for E. coli, is indicated by a square. The other mutations that confer macrolide drug resistance are circled: G2057 (E. coli numbering) and G2141 (H. pylori numbering); A2058 (E. coli numbering), A2142 (H. pylori numbering), and A2060 (S. pneumoniae numbering); A2143 (H. pylori numbering); and A2064 (S. pneumoniae numbering). In 1997, we detected in Canada the emergence of azithro-mycin-resistant N. gonorrhoeae strains (MIC >= 2.0 mug/ml) which also have resistance to erythromycin (MIC >= 2.0 mug/ml). In this study, we determined the phenotypic characteristics of 56 azithromycin-resistant N. gonorrhoeae strains which were isolated in Canada between 1997 and 1999 and investigated their mechanisms of macrolide resistance. MATERIALS AND METHODS : Bacterial strains. | Between 1997 and 1999, a total of 56 erythromycin- and azithromycin-resistant N. gonorrhoeae isolates were collected as part of the National Surveillance of N. gonorrhoeae Program in Canada. The strains used in this study originated from British Columbia, Alberta, Ontario, Quebec, and Nova Scotia. N. gonorrhoeae strains were subcultured on GC medium base (GCMB) (Difco Laboratories, Detroit, Mich.) containing 0.2% BioX (QueLab, Montreal, Quebec, Canada) and incubated for 24 h at 35C in a 5% CO2 atmosphere with or without antibiotics and maintained in brain heart infusion (BHI) medium (Difco Laboratories) containing 20% glycerol and stored at -80C. All other nonfastidious organisms were grown aerobically on BHI medium. Antimicrobial susceptibility testing and strain characterization. | Antimicrobial susceptibilities of N. gonorrhoeae to azithromycin (compliments of Pfizer, Pointe-Claire/Dorval, Quebec, Canada), cefixime (Wyeth-Ayerst Laboratories, Mason, Mich.), ceftriaxone (Roche, Laval, Quebec, Canada), ciprofloxacin (compliments of Bayer, Etobicoke, Ontario, Canada), erythromycin (compliments of Lilly, Indianapolis, Ind.), penicillin (compliments of Novopharm, Scarborough, Ontario), spectinomycin (compliments of Pharmacia & Upjohn, Kalamazoo, Mich.), and tetracycline (Sigma-Aldrich Canada Ltd., Oakville) were performed according to the agar dilution method described by the National Committee for Clinical Laboratory Standards (NCCLS) , but the GC medium base was modified to contain 1% Kellogg's supplement (0.05 g of ferric nitrate, 1.0 g of cocarboxylase, 5.0 g of glutamine, 200.0 g of glucose in 1 liter of H2O) . The MICs of all antibiotics except azithromycin and erythromycin (for which there are no NCCLS MIC breakpoints) were interpreted according to the NCCLS guidelines . For the purposes of this study, the azithromycin and erythromycin MICs at which strains exhibited resistance were defined as >=2.0 mug/ml . N. gonorrhoeae reference cultures ATCC 49226, WHO III, WHO V, and WHO VII were used as quality control strains. Susceptibility to crystal violet was used as an indicator of efflux, and crystal violet susceptibility assays were performed as previously described . Auxotyping of the N. gonorrhoeae strains was performed as previously described on a chemically defined medium to determine their nutritional requirements for leucine (L), ornithine (O), citrulline (C), proline (P), arginine (A), hypoxanthine (H), uracil (U), and methionine (M). Prototype strains that did not require amino acid supplements for growth were designated NR (no requirements). Serotyping of the N. gonorrhoeae isolates was performed as described by Knapp et al. . Plasmid profiles of the N. gonorrhoeae isolates were performed as previously described . PCR and DNA sequencing of the macrolide resistance targets. | PCR was used to determine the presence or absence of methylase genes and to determine mutations in the mtr efflux pump and ribosomal genes. Using PCR with primers and conditions as previously described, clinical isolates were tested for the presence of one or more of the methylase genes encoded by erm(A), erm(B), erm(C), erm(F) , and an efflux pump mef(A) . The following reference strains were used as positive controls for PCR: for erm(A), Staphylococcus aureus RN1389; for erm(B), Streptococcus pyogenes AC1; for erm(C), S. aureus RN4220; for erm(F), E. coli V831; and for mef(A), S. pyogenes 02C1064. In an attempt to identify point mutations, sequences of genes related to macrolide resistance were determined for the mtrR promoter region, domain V of the 23S rRNA, and rplD and rplV (encoding ribosomal proteins L4 and L22, respectively) . The mtrR promoter region of the mtrR efflux pump operon was amplified using the published primers RPMAL#2 (5' -ACTGAAGCTTATTTCCGGCGCAGGCAGGG-3') and KH9#3 (5'-GACGACAGTGCCAATGCAACG-3') and sequenced in an attempt to identify mutations in this region. To identify mutations within domain V (of the peptidyltransferase loop) for each of the four copies of the 23S rRNA gene, a two-step PCR method was developed. The first step used the gonrRNA-F primer, paired with a specific primer for each of the 23S rRNA alleles listed in Table , to amplify specific regions of 23S rRNA in eight strains. PCR conditions were 1 min of denaturation at 94C, 1.5 min at 66C (for alleles 2 and 3 of domain V) or 68C (for alleles 1 and 4 of domain V) for annealing, and 2.5 min at 72C for elongation for 30 cycles. The amplicons obtained were then used as templates in a second PCR using gonrRNA-F and gonrRNA-R2 primers to amplify the peptidyltransferase loop for sequencing determination. The conditions for this PCR step were 94C at 1 min for denaturing, 59C for 1 min for annealing, and 72C at 1 min for elongation for 35 cycles. PCR products were also sequenced in an attempt to identify point mutations in rplD and rplV genes in ribosomal proteins L4 and L22, respectively . All PCR amplicons were purified (Promega PCR Prep Kit; Fisher, Nepean, Ontario, Canada), and both strands were sequenced using an ABI Prism 377 DNA sequencing system (Applied Biosystems, Foster City, Calif.). DNA sequences were aligned using BLAST and GenBank programs to identify the mutations of these genes. Transformation studies. | Total genomic DNA from both clinical strain 20870 and strain 19416 was introduced into piliated N. gonorrhoeae F62 (MICs of erythromycin and azithromycin, 0.25 and 0.125 mug/ml, respectively) by transformation. As a control, a ciprofloxacin-resistant strain with a ciprofloxacin MIC of 8.0 mug/ml and with identified mutations of gyrA and parC was transformed into F62. Briefly, N. gonorrhoeae F62 was suspended in 1 ml of 10 mM Tris-HCl-10 mM EDTA disodium salt (pH 8.0) to an optical density of 0.60 to 0.63 at 600 nm, inoculated on GCMB, and incubated for 1 h at 37C in an atmosphere enriched with 7% CO2. Chromosomal or amplified DNA (1 mug) was then applied to the inoculated plate. After a 24-h incubation period, individual colonies from the GCMB plates were transferred to GCMB containing erythromycin (0.5, 1.0, and 2.0 mug/ml). Potential transformants were again subcultured onto GCMB containing erythromycin (0.5, 1.0, and 2.0 mug/ml) prior to MIC determination and examination of mutations in the 23S rRNA and the promoter region of mtrR. In vitro conversion of resistant strains to susceptibility. | The following procedure was performed to obtain spontaneous susceptible revertants from macrolide-resistant strains. Single colonies of the macrolide-resistant cultures 20869 and 20870 (erythromycin MIC, 64.0 mug/ml; azithromycin MIC, 4.0 mug/ml) were subcultured daily for 7 days onto nonselective GCMB plates. Fifty isolated colonies were randomly selected and replicated onto GCMB plates with and without erythromycin (8.0 mug/ml), and another single colony was also subcultured onto GCMB to obtain isolated colonies. This step was repeated until colonies on replicated plates grew on GCMB but not on erythromycin plates (8.0 mug/ml) after 3 days of culturing. Mutants with decreased susceptibility were selected for MIC determination. Sequencing of the mtrR promoter region flanking mtrC and mtrR and the domain V region of the four alleles of 23S rRNA (rrl gene) of the resistant mutants and of the susceptible parent strains was also performed, and the results were compared. In vitro mutation to erythromycin resistance. | To generate macrolide-resistant mutants, the following procedure was performed. Three susceptible N. gonorrhoeae strains (21855, 21305, and 22831) (erythromycin and azithromycin MICs, 0.5 to 1 and 0.125, respectively) were selected from our culture collection and subcultured onto nonselective GCMB plates. After a 24-h incubation period, the growth from a single colony of each strain was swabbed onto the entire surface of GCMB plates containing a gradient of erythromycin of 0 to 4.0 mug/ml. After a 24-h incubation, the plates were examined for growth and isolated colonies on the highest concentration of antibiotics were selected and streaked onto another GCMB gradient plate with a concentration of erythromycin from 0 to 4.0 mug/ml. After the selection step was repeated seven times, a gradient plate with a higher concentration of erythromycin (0 to 12.0 mug/ml) was used. Colonies in the region of the highest concentration of erythromycin were selected and subcultured onto three plates, namely, a GCMB gradient plate (0 to 12.0 mug of erythromycin/ml), a GMCB plate containing erythromycin at 8.0 mug/ml, and a GCMB plate containing erythromycin at 16.0 mug/ml. Colonies that grew on GCMB containing erythromycin (8.0 or 16.0 mug/ml) were subcultured twice prior to MIC determination. Sequencing of the mtrR promoter region between mtrC and mtrR and the domain V region of the four alleles of 23S rRNA (rrl gene) of the resistant mutants and of the susceptible parent strains was also performed, and the resulting sequences were compared. Nucleotide sequence accession numbers. | The nucleotide sequences of the four alleles of 23S rRNA for strains 20869 and 20870 have been deposited in GenBank under accession numbers to . TABLE 1 | Oligonucleotides used in the amplification of the four alleles of the peptidyltransferase region in domain V of the 23S rRNA RESULTS AND DISCUSSION : Antibiotic susceptibilities and characterization of N. gonorrhoeae isolates. | Our laboratory began monitoring azithromycin resistance in 1996, and the emergence of resistant strains was first observed in 1997 in Canada. During 1997 to 1999, 56 azithromycin-resistant N. gonorrhoeae strains (MICs, 2.0 to 4.0 mug/ml) were isolated in five provinces in Canada (Quebec, British Columbia, Ontario, Alberta, and Nova Scotia). In 1997, 95% (19 of 20) of the azithromycin-resistant strains were isolated from Quebec and only 1 other strain was isolated from British Columbia. In 1998, 52% (13 of 25) of the azithromycin-resistant strains were from British Columbia, 5 from Ontario, 3 from Alberta, 2 from Nova Scotia, and 2 from Quebec. Of the 11 strains isolated in 1999, 8 were from British Columbia and 3 were from Quebec. The 56 strains were classified into 16 types by auxotype and serotype (A/S). The most common classes were type NR/IB-01 (33 strains [58.9%]) and type NR/IB-03 (6 strains [10.7%]). Three A/S classes had two strains each, NR/IA-04, H/IB-01, and H/IB-03. Eleven A/S classes (NR/IB-06, NR/IB-10, NR/IB-14, NR/IB-17, NR/IB-23, NR/IB-32, NR/IB-26, H/IB-20, P/IB-03, P/IB-02, and P/serotype [not typeable]) each had a unique strain. All the strains contained a 2.6-mDa plasmid, except for one P-requiring strain. All azithromycin-resistant strains (azithromycin MIC >= 2.0 mug/ml) were also resistant to erythromycin (erythromycin MIC >= 2.0 mug/ml). Some of these strains (58.9%) also had chromosome-mediated resistance to both tetracycline and penicillin, and the remaining 41.1% were resistant to tetracycline (tetracycline MIC >= 2.0 mug/ml). A total of 96.4% of the azithromycin-resistant strains showed resistance to crystal violet (crystal violet MICs >= 1.0 mug/ml), indicating that their cell envelopes may have had reduced permeability . Detection of mechanisms of macrolide resistance. | Several targets were used to determine different mechanisms of resistance in the azithromycin-resistant N. gonorrhoeae. Forty-six selected strains, including strains from all A/S classes and strains from different geographical locations, were all negative by PCR for rRNA methylases using erm(A), erm(B), erm(C), or erm(F) (data not shown), indicating that our azithromycin-resistant strains did not have the same mechanisms of macrolide resistance as those found in Seattle and Uruguay . PCR amplification of a subset of 14 strains selected randomly also did not show the presence of the mef(A) gene, a macrolide transporter (data not shown). PCRs using genomic DNA of control strains all showed positive results, as expected. In addition, the structural genes rplD (ribosomal protein L4) and rplV (ribosomal protein L22) were examined for point mutations, using nine and eight strains, respectively . The DNA sequences of the rplD and rplV genes were similar to those determined in a susceptible N. gonorrhoeae strain, FA1090 (Gonococcal Genome Sequencing Project, University of Oklahoma ). We concluded that macrolide resistance in these strains was not due to mutations of rplD or rplV or to the acquisition of erm genes. It has been well established that overexpression of the mtr efflux pump can play a role in increasing resistance to macrolides . DNA sequencing of the mtrR promoter region in the 14 azithromycin- and erythromycin-resistant strains revealed that 12 of these strains had a single-base-pair deletion within a 13-bp inverted repeat within the mtrR promoter . The deletion which activates the overexpression of the mtr pump in these 12 strains was the same mutation as that reported for azithromycin-resistant N. gonorrhoeae strains in another study . The other two azithromycin-resistant strains (20869 and 20870) that did not have a mutation in the mtr promoter region therefore have other resistance mechanisms not previously described for N. gonorrhoeae. There are other mechanisms of macrolide resistance, including mutations in the 23S rRNA as found in other bacteria. Therefore, we explored the possibility that changes occur in the 23S rRNA DNA sequence in strains 20869 and 20870 that affect their susceptibility to macrolides. The four copies of the 23S rRNA rrl gene present in N. gonorrhoeae were amplified individually for domain V in 14 azithromycin-resistant strains . Twelve of these strains did not have mutations in the DNA sequence of the 23S rRNA peptidyltransferase region . These 12 strains did have the deletion in the mtr promoter region. In the remaining two strains, 20869 and 20870 (erythromycin MIC, 64.0 mug/ml; azithromycin MIC, 4.0 mug/ml), alleles 1, 2, and 4 were found to contain a C2611T (E. coli numbering) mutation . These mutations correspond to position C2599T for alleles 1 and 2 and position C2605T for allele 4 (N. gonorrhoeae numbering). Allele 4 was found to contain base insertions throughout domain V which account for the change in nucleotide position number. We believe that this mutation is responsible for macrolide resistance in these two strains. The secondary structure of domain V of the 23S rRNA gene shown in Fig. , based on an illustration by Depardieu and Courvalin , revealed a number of point mutations previously identified in other strains which confer resistance to macrolides . These include the following positions and mutations: positions A2058 and A2062 (E. coli numbering) in S. pneumoniae ; positions G2057, A2058, and A2059 (E. coli numbering) in H. pylori ; and positions G2057, A2058, and C2611 in E. coli. These mutations most likely affect the secondary structure of the peptidyltransferase loop in domain V, due to the pairing of different nucleotides. In the N. gonorrhoeae strains 20869 and 20870, the C->T mutation is located in this same region. Further investigation of strains 20869 and 20870 revealed that they did not have the mutation in the mtrR promoter region causing the mtrR overexpression nor did they have mutations in the ribosomal protein L4 or L22. No mutations were observed in allele 3 of these two strains. Transformation studies. | Transformation studies were used to confirm that mutations in the rrl gene within the peptidyltransferase loop of domain V of the 23S rRNA caused resistance to macrolides in N. gonorrhoeae. Transformation of total genomic DNA from macrolide-resistant strain 19416 (erythromycin MIC, 8.0 mug/ml; azithromycin MIC, 2.0 mug/ml) to N. gonorrhoeae F62 (an erythromycin- and azithromycin-susceptible strain) yielded transformants for which macrolide MICs were increased (erythromycin MIC of 4.0 mug/ml). Sequencing of the mtrR promoter region revealed that the transformants had the same deletion (A) in mtr found in the donor strain and did not have mutations in 23S rRNA. Transformation of total genomic DNA from macrolide-resistant strain 20870 (erythromycin MIC, 64.0 mug/ml; azithromycin MIC, 4.0 mug/ml) to F62 yielded resistant transformants (erythromycin MIC, 4.0 mug/ml; azithromycin MIC, 2.0 mug/ml). However, when the regions encoding domain V of the 4 alleles of the 23S rRNA rrl gene in the transformants were amplified, their DNA sequences showed that these strains did not have mutations in the 23S rRNA peptidyltransferase region or the deletion mutation in the mtrR promoter region. The increased macrolide MICs for the transformants remained unexplained in this experiment. The ciprofloxacin-resistant strain (MIC of 8.0 mug/ml) used as a positive control to transform F62 yielded transformants with a ciprofloxacin MIC of 2.0 mug/ml. Subsequent sequencing of the gyrA and parC genes of the transformants revealed that the mutations in the transformants were identical to the mutations of the donor strain (gyrA mutations of 91PHE and 95GLY and parC mutations of 86ASN). Analysis of erythromycin- and azithromycin-resistant and -susceptible spontaneous mutants. | To confirm that the mutations of rrl were associated with macrolide resistance, we attempted to derive mutants with reduced susceptibility from two macrolide-resistant clinical isolates (20869 and 20870) that had mutations in alleles 1, 2, and 4 of the 23S rRNA . Strains LK368 and LK370 were derived from 20869 and 20870, respectively, after subculturing 35 times on GCMB plates. The MICs of erythromycin were determined to be 4.0 mug/ml for LK368 and 32.0 to 64.0 mug/ml for LK370, and the MICs of azithromycin were 1.0 mug/ml for LK368 and 2.0 mug/ml for LK370. Strains LK368 and LK370 were then subcultured an additional 12 times on GCMB plates and yielded two derivatives, LK369 and LK371, respectively . The MICs of erythromycin were determined to be 2.0 mug/ml for LK369 and 16.0 to 32.0 mug/ml for LK371, and the MICs of azithromycin were 0.5 mug/ml for LK369 and 2.0 to 4.0 mug/ml for LK371 . The partial DNA sequence of 23S rRNA from these derivatives showed that allele 1 had a T->C reversion at nucleotide 2611 (E. coli numbering) for two strains, LK368 and LK369, and allele 2 had a T->C reversion at nucleotide 2611 (E. coli numbering) for one strain only (LK369). The sequences for alleles 3 and 4 remained identical to those of the parent strains. DNA sequencing of the mtrR repressor promoter region revealed that the deletion mutation was not present . The two derivatives (LK370 and LK371) for which the erythromycin MIC values remained in the range of 16.0 to 64.0 mug/ml and the azithromycin MICs remained in the range of 2.0 to 4.0 mug/ml also retained the same DNA sequences for all four alleles of the 23S rRNA as the those of the parent strain. These derivatives (LK370, LK371, LK368, and LK369) did not show mutations in the mtrR promoter region . The resistant- to susceptible-phenotype revertants showed the reversion of 23S rRNA sequence to wild type and no DNA sequence changes in the mtrR promoter. We also attempted to generate spontaneous macrolide-resistant mutants from susceptible strains and compared the DNA sequences for the domain V region of the rrl gene. Three macrolide-susceptible N. gonorrhoeae strains, 22831, 21305, and 21855 (erythromycin MICs, 0.5 to 1.0 mug/ml; azithromycin MIC, 0.125 mug/ml), were selected from our culture collection . The sequences of domain V of rrl of the three strains (22831, 21305, and 21855) were the same as that of the susceptible N. gonorrhoeae reference strain FA1090 (Gonococcal Genome Sequencing Project, University of Oklahoma). Initially, six spontaneously resistant mutants were obtained after subculturing the three parent strains 14 times. Strains LK359 and LK360 were obtained from parent strain 21855, strains LK361 and LK362 were from parent strain 22831, and strains LK363 and LK364 were from parent strain 21305. After subculturing an additional nine times, mutants LK366 and LK367 were obtained from parent strain 22831. The MICs and the DNA sequences of alleles 1, 2, 3, and 4 of the 23S rRNA of the spontaneous mutants and those of their parent strains were determined and compared . No substitution mutations were identified in any of the four alleles of the 23S rRNA of LK361 and LK362 (erythromycin MIC, 8.0 mug/ml; azithromycin MIC, 2.0 mug/ml), but the alleles did have the deletion (A) in the mtrR promoter region, causing the formation of mutants with overexpressed mtr efflux pumps . LK363 and LK364 (erythromycin MIC, 32.0 mug/ml; azithromycin MIC, 4.0 mug/ml) had the C->T substitution in all four alleles of the 23S rRNA. They did not have the deletion in the mtrR promoter region, but they did have an insertion (T). The nucleotide insertion (T) in the mtrR promoter region was also identified in one of the parent strains, and this mutation did not seem to confer resistance. The MICs of erythromycin and azithromycin were found to be 64.0 mug/ml and 8.0 mug/ml, respectively, for strains LK359 and LK360. These strains all had the C->T substitution in all four alleles of the 23S rRNA and did not have the deletion in the mtrR promoter region. Mutant LK366 had the C2611T mutation (E. coli numbering) in alleles 1, 2, and 3 of the 23S rRNA. LK367 had the C2611T mutation (E. coli numbering) in alleles 1, 2, and 4 of the 23S rRNA . In comparison with those for the other mutants, the MICs of erythromycin (64.0 mug/ml) and azithromycin (32.0 mug/ml) were found to be the highest for LK366 and LK367. These strains (LK366 and LK367) had mutations in three of the four alleles of rrl, while strains LK359 and LK360, for which azithromycin had a lower MIC (8.0 mug/ml), had mutations in all four alleles of rrl. Therefore, the higher azithromycin MIC does not correlate with the number of alleles with mutations. Both LK366 and LK367 had one deletion (A) in the mtrR promoter region which LK359 and LK360 did not have. This deletion may account for the finding of higher azithromycin MICs for LK366 and LK367 than for LK359 and LK360, which indicates the additive effects of multilocus mutations. On the other hand, other unknown mechanisms could also attribute to higher azithromycin resistance in LK366 and LK367. In conclusion, we have described a novel C2611T (E. coli numbering) mutation in alleles 1, 2, and 4 in 23S rRNA that confers macrolide resistance in two clinical N. gonorrhoeae. The mtr efflux pump contributes to erythromycin resistance in strains for which erythromycin MICs range from 2.0 to 8.0 mug/ml and azithromycin MICs range from 0.5 to 2.0 mug/ml. Higher levels of macrolide resistance (erythromycin MIC, 16.0 mug/ml; azithromycin MIC, 4.0 to 8.0 mug/ml) are associated with mutations of one or more alleles of domain V of rrl. Neither clinical isolates nor spontaneous mutants in this study showed that rplD (ribosomal protein L4) or rplV (ribosomal protein L22) contributed to resistance. The spontaneous mutants derived by in vitro mutation to erythromycin resistance revealed that high-level macrolide resistance (erythromycin MIC, 64.0 mug/ml; azithromycin MIC, 32.0 mug/ml) may require three of the four mutations in rrl as well as the mutation in the mtr loci, showing the additive effects of mutations on resistance levels. Although we observed additive mechanisms of macrolide resistance with in vitro mutants, no clinical isolates with both mechanisms have yet been identified. This study provided further evidence that N. gonorrhoeae is evolving continuously in response to antibiotic selection pressure. Characterization of mechanisms of resistance in these organisms will provide information necessary for the development of new therapies. TABLE 2 | Characterization of 14 clinical isolates of N. gonorrhoeae for which erythromycin and azithromycin MICs are elevated TABLE 3 | Characterization of spontaneous N. gonorrhoeae mutants: group I, susceptible spontaneous mutants from erythromycin and azithromycin-resistant N. gonorrhoeae strains; group II, erythromycin- and azithromycin-resistant N. gonorrhoeae mutants from susceptible N. gonorrhoeae strains Backmatter: PMID- 12183240 TI - Antiviral Activities of MCC-478, a Novel and Specific Inhibitor of Hepatitis B Virus AB - MCC-478 is a newly synthesized 2-amino-6-arylthio-9-phosphonomethoxyethylpurine bis(2,2,2-trifluoroethyl) ester derivative. MCC-478 showed a substantially higher (ca. 80-fold) anti-hepatitis B virus (HBV) activity than that of lamivudine, despite no significant anti-human immunodeficiency virus activity. Since the bis(2,2,2-trifluoroethyl) ester group was used to improve the oral bioavailability of the phosphonomethoxyethylpurine derivatives, two monoester derivatives and one phosphonic acid derivative were also evaluated. It was suggested that these hydrolyzed derivatives, which appeared in animals given MCC-478, have enough anti-HBV activity to contribute to efficacy in vivo. Furthermore, no apparent cytotoxic effects or reductions of mitochondrial DNA content by MCC-478 and its derivatives were observed. These results indicated that MCC-478 may be a new promising anti-HBV agent. Keywords: Introduction : Hepatitis B virus (HBV) is a small, partially double-stranded DNA virus and a prototype member of the hepadnavirus family. HBV is a causative agent of both acute and chronic hepatitis. The World Health Organization lists hepatitis B as one of the 10 leading killer diseases and estimates that 350 million people are chronically infected with HBV. Until 1999 the only therapy approved for chronic HBV infection was alpha interferon, and this treatment was useful only for a small minority of Asian patients . The hepadnavirus has a unique replication cycle. Following entry into a hepatocyte, the viral genome matures to a single covalently closed circular DNA by an unknown enzyme derived from the host cell and/or hepadnavirus polymerase. The covalently closed circular DNA is translocated into the nucleus and serves as the template for several viral RNAs . A hepadnavirus polymerase, which is encoded by a gene in the longest 3.5-kb pregenome RNA, transcribes full-length negative-strand HBV genome from a pregenome RNA within nucleocapsids . Sequentially, the polymerase exerts its DNA-dependent DNA polymerase activity to synthesize the positive-strand HBV genome . Since the replication cycle of hepadnavirus genome is dependent upon the action of its own polymerase, HBV polymerase could be a good target for an antiviral therapy. Recently, lamivudine was approved to treat chronic HBV infection . Lamivudine is a dideoxycytidine analogue that is active against human immunodeficiency virus (HIV) and HBV . It is also shown that lamivudine triphosphate acts as a chain terminator against viral DNA synthesis . However, prolonged lamivudine treatment results in the emergence of lamivudine-resistant HBV mutants in 17 to 46% of patients treated for 1 year and more than 50% of patients within 2 years of treatment . The emergence of drug-resistant HBV emphasizes the need to develop other antiviral agents and therapeutic strategies. One of the candidates, adefovir dipivoxil, which is an oral prodrug of adefovir (PMEA) and is active against HIV and HBV replication , is now in phase III clinical trials to treat HBV infection . Although adefovir dipivoxil is suggested to be active against lamivudine-resistant HBV strains , nephrotoxicity, characterized by changes in renal function laboratory markers, is observed among HIV patients treated long-term with adefovir dipivoxil . We synthesized more than 100 derivatives of phosphonomethoxyethylpurine and evaluated anti-HBV activity. The absorbability of test compounds was also examined using ex vivo samples from mice and rats given the drug orally . Finally, we found 2-amino-6-arylthio-9-phosphonomethoxyethylpurine bis(2,2,2-trifluoroethyl) ester derivatives that showed high antiviral activity and apparent absorption in the animals. In this report, we evaluated the in vitro antiviral properties of MCC-478, 2-amino-6-(4-methoxyphenylthio)-9-[2-(phosphonomethoxy)ethyl]purine bis(2,2,2-trifluoroethyl) ester. Since the bis(2,2,2-trifluoroethyl) ester group was used to improve the oral bioavailability of the phosphonomethoxyethylpurine derivatives and might be hydrolyzed in vivo to give its monoester and free phosphonic acid, the monoester and free phosphonic acid derivatives of MCC-478 were also tested. In order to investigate the toxicological profile of MCC-478, cytotoxicity and reduction of the mitochondrial DNA content was evaluated. It was shown that MCC-478 is a potent inhibitor of HBV replication, and its antiviral profile is specific to HBV. MATERIALS AND METHODS : Compounds. | The following test compounds were chemically synthesized at Mitsubishi Pharma Corporation: MCC-478 (2-amino-6-(4-methoxyphenylthio)-9-[2-(phosphonomethoxy)ethyl]purine bis(2,2,2-trifluoroethyl) ester), monoester derivative 1 (M-1) (2-amino-6-(4-methoxyphenylthio)-9-[2-(phosphonomethoxy)ethyl]purine 2,2,2-trifluoroethyl ester), free phosphonic acid derivative 1 (F-1) (2-amino-6-(4-methoxyphenylthio)-9-[2-(phosphonomethoxy)ethyl]purine), and monoester derivative 2 (M-2) (2-amino-6-(4-hydroxyphenylthio)-9-[2-(phosphonomethoxy)ethyl]purine 2,2,2-trifluoroethyl ester). PMEA (9-[2-(phosphonomethoxy)ethyl]adenine) was synthesized by the method of Holy and Rosenberg . Lamivudine was extracted and purified from tablets of Epivir (GlaxoWellcome). Epivir tablets (60 150-mg tablets) were ground and extracted with hot methanol. Silica gel was added to this methanol solution, and the resultant slurry was stirred for 30 min. Methanol was removed by evaporation, and the residue was transferred to a silica gel column. The desired material was eluted with a 9:1 mixture of chloroform-methanol. The eluate was concentrated and dissolved in methanol-toluene to produce crystals of lamivudine. The yield was 8.14 g. The structure was identified with proton nuclear magnetic resonance, and the purity (>99%) was checked by thin-layer chromatography, high-pressure liquid chromatography, and measurement of optical rotation. The structures of MCC-478, M-1, M-2, F-1, and PMEA are shown in Fig. . Cell culture. | HB611 , a stably transfected cell line which contains three copies of the HBV genome in tandem in its chromosome, and HuH-6, a parent cell line, were kindly provided by K. Matsubara (Institute for Molecular and Cellular Biology, Osaka University, Osaka, Japan; now at the Department of Molecular Biology, Nara Institute of Science and Technology, Nara, Japan). HB611 cells were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 0.2 mg of Geneticin (GibcoBRL) per ml and 10% fetal calf serum (FCS). HuH-6 cells were maintained in DMEM supplemented with 10% FCS. HepG2 cells were obtained from Dainippon Pharmaceuticals (Osaka, Japan) and maintained in MEM supplemented with nonessential amino acids, 1 mM sodium pyruvate, and 10% FCS. All cell lines were cultured at 37C with 5% CO2. Analysis of anti-HBV activities. | Antiviral activities of the test compounds were analyzed by a modified version of the method of Yokota et al. . HB611 cells were plated at a density of 2 x 104 cells per well on 24-well plates. When the cells had grown to confluence, the medium was changed to medium containing a test compound. Test compounds and the final concentrations in the experiments were as follows: experiment 1, MCC-478 at 0.01, 0.03, 0.1, 0.3, 1.0, and 3.0 muM and PMEA and lamivudine at 0.1, 0.3, 1.0, 3.0, and 10 muM; experiment 2, MCC-478, M-1, and F-1 at 0.01, 0.03, 0.1, 0.3, and 1.0 muM; experiment 3, M-1 and M-2 at 0.03, 0.1, 0.3, 1.0, and 3.0 muM. The cells were cultured for 8 or 9 days with the drug-containing medium changed every 3 or 4 days. On the last day of cultivation, the monolayers were washed with phosphate-buffered saline, and the cells were lysed at 37C for 1 h with 0.5 ml of lysis buffer (10 mM Tris-HCl [pH 7.5], 150 mM NaCl, 5 mM EDTA, 1% sodium dodecyl sulfate [SDS], 0.1 mg of proteinase K [Boehringer Mannheim] per ml). The lysates were collected and placed in 1.5-ml Eppendorf tubes and were further digested by proteinase K for 2 days with gentle agitation. After digestion with 0.1 mg of RNase A per ml at 65C for 15 min, the lysate was extracted by phenol-chloroform-isoamyl alcohol. The total cellular DNA was precipitated by ethanol and was digested with HindIII (Boehringer Mannheim) at 37C overnight. The digested DNA was electrophoresed on a 0.9% agarose-1x Tris-borate-EDTA gel and stained with ethidium bromide. The separated DNA samples were treated with 0.25 N hydrochloric acid for 15 min, denatured in 0.5 N NaOH for 30 min, and neutralized with 0.5 M Tris-HCl [pH 7.5] containing 1.5 M NaCl and 1 mM EDTA for 1 h. DNA was transferred to a nylon membrane (Hybond N+; Amersham) using a VacuGene vacuum blotting system (Amersham). The membranes were dried, linked by using a UV illuminator, and hybridized to a digoxigenin (DIG)-labeled HBV DNA at 42C overnight in hybridization buffer (DIG Easy Hyb; Boehringer Mannheim). The hybridized membranes were washed twice with 2x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate) containing 0.1% SDS at room temperature for 5 min and twice with 0.1x SSC containing 1% SDS at 65C for 30 min. The HBV-specific bands were detected by anti-DIG-AP Fab fragments (Boehringer Mannheim) and CDP-star (New England Biolabs). After exposure to X-ray film (XJB; Kodak), the developed film was scanned in a quantitative densitometric manner using ImageQuant image analysis software on a Personal Densitometer (Molecular Dynamics). The densities of the signals derived from integrated HBV and free HBV DNA were measured. HBV DNA content of each sample was determined by the ratio of HBV DNA replication intermediates to integrated HBV DNA. Cytotoxicity. | HuH-6 cells were plated at a density of 105 cells per ml on 96-well plates and cultured at 37C with 5% CO2 for 3 days in DMEM supplemented with 10% FCS and each test compound at final concentrations of 10, 30, 100, 300, and 1,000 muM. 3-(4,5-Dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) (CellTiter96 aqueous nonradioactive cell proliferation assay; Promega) was added at the end of the culture period, and the plate was incubated at 37C for 2 h. The absorbance at 490 nm of each well was measured with a microplate reader (NJ-2000; Intermed). Anti-HIV activity. | Determination of antiviral activity against HIV type 1 (HIV-1) replication was based on inhibition of the HIV-induced cytopathic effect in MT-4 cells, as described previously . MT-4 cells (105/ml) were infected with HIV-1 (HTLV-IIIB strain) at a multiplicity of infection of 0.02. The cells were cultured in the presence of 0.032, 0.16, 0.8, 4, and 20 mug of the test compounds per ml at 37C with 5% CO2. In order to assess the viability of both HIV- and mock-infected MT-4 cells, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was added on day 4 after virus infection and incubated at 37C for 2 h. The absorbance at 540 nm was measured, and the percent inhibition value was calculated. Mitochondrial DNA content. | HepG2 cells (2.5 x 104 cells/well) were plated in a 12-well plate and treated with each test compound for 8 days. On the last day of culture, the cells were washed with phosphate-buffered saline, harvested, placed in a 1.5-ml Eppendorf tube, and stored at -80C until use. Cells were lysed and digested by proteinase K at 37C for one night with lysis buffer. The lysate was digested with RNase A (0.1 mg/ml) at 65C for 10 min and extracted by phenol-chloroform-isoamyl alcohol. The total cellular DNA was precipitated by ethanol and dissolved in distilled water. DNA (2.5 mug) was denatured in 0.5 N NaOH and 1.5 M sodium chloride for 10 min and dot blotted onto a nylon membrane (Hybond N+; Amersham). The membrane was dried, linked by using a UV illuminator, and hybridized to a DIG-labeled mitochondrion-specific oligonucleotide DNA probe (5'-DIG-CTTATATGAT ATGTCTCCAT ACCCATTACA ATCTCCAGCA-3'), which extends from nucleotides 4207 to 4246 of mitochondrial DNA , in the hybridization buffer (DIG Easy Hyb; Boehringer Mannheim) at 42C overnight. FIG. 1. | Chemical structures of MCC-478 and its metabolites in vivo. Chemical structures of MCC-478 and its metabolites in vivo. RESULTS : Antiviral activities of MCC-478 and other nucleoside or nucleotide analogues. | We first attempted to evaluate anti-HBV activities of MCC-478, lamivudine, and PMEA in HB611 cells, which is a stable transfected cell line producing HBV. HB611 cells were treated with various concentrations of MCC-478, lamivudine, or PMEA (both lamivudine and PMEA are used as reference drugs in many studies for anti-HBV agents). The cells in five wells on a 24-well plate were treated independently at each concentration for 8 days. The extent of HBV DNA replication in cells was evaluated from the ratio of the replicative intermediates to the integrated HBV genome. Antiviral effects were reported as percent control values (untreated cells). Figure shows the average percent control values of HBV DNA replication in cells treated with drugs at each concentration. Also, WinNonlin version 1.1 (Scientific Consulting Inc.) was used to calculate the 50% and 90% effective concentrations (EC50 and EC90, respectively) of MCC-478 and lamivudine. The EC50 and EC90 of MCC-478 were 0.027 and 0.24 muM, respectively, while those of lamivudine were 2.2 and 7.8 muM, respectively. The EC50 of PMEA could not be calculated, since PMEA gave percent control values below 50 even at the lowest concentration (0.1 muM). These results clearly suggested that MCC-478 was a potent inhibitor of HBV DNA replication. Comparison of the anti-HBV activities of the hydrolyzed derivatives of MCC-478. | Anti-HBV activity was also observed in sera from mice given MCC-478 orally (data not shown). Since some of the bisalkyl ester prodrugs of PMEA were not cleaved completely to a free phosphonic acid form , two monoester derivatives (M-1 and M-2) and one phosphonic acid derivative (F-1) of MCC-478 were putative metabolites in serum. In order to understand their contribution to efficacy in vivo, they were synthesized and tested in HB611 cells. As shown in Fig. , the anti-HBV activities of M-1 and F-1 were similar to that of MCC-478, and the anti-HBV activity of M-2 was similar to that of M-1. Cytotoxicity and anti-HIV activity of MCC-478. | The 50% cytotoxic concentration (CC50) of each test compound was determined in the HuH-6 cell line, the parent cell line of HB611. MCC-478 and its derivatives were well tolerated . To evaluate the specificity for antiviral activity, anti-HIV activities of MCC-478, PMEA, and lamivudine were determined by examining the inhibition of the HIV-induced cytopathic effect in the MT-4 cell line. Treatment with MCC-478 showed no significant selectivity between the cytotoxicity against mock-infected MT-4 cells and the inhibition of HIV-induced cytopathic effect, e.g., 16 and 11%, respectively, even at the highest concentration. The EC50s and CC50s of the test compounds in MT-4 cells are summarized in Table . These results suggested that anti-HIV activity of MCC-478 was much weaker than anti-HBV activity, while PMEA retained anti-HBV and anti-HIV activity. Effect of MCC-478 on mitochondrial DNA content. | Therapies with nucleoside analogues sometimes cause mitochondrial dysfunction, which could be derived from incorporation of the nucleoside analogue into mitochondrial DNA or reduction in mitochondrial DNA content . In this study, the effects of MCC-478 and its hydrolyzed derivatives on mitochondrial DNA content were examined, using 2',3'-dideoxycytidine (ddC) as a positive control. HepG2 cells were treated with each test compound at a final concentration of 0.1, 1.0, or 10 muM for 8 days. Mitochondrial DNA content was examined by dot blot hybridization. Treatment with MCC-478, M-1, M-2, or PMEA showed no reduction in mitochondrial DNA content, even though treatment with ddC at every concentration reduced mitochondrial DNA content to nearly half that of the untreated control . FIG. 2. | Inhibitory effects of MCC-478, PMEA, and lamivudine against HBV replication in HB611 cells. Inhibitory effects of MCC-478, PMEA, and lamivudine against HBV replication in HB611 cells. Cells were treated with the indicated concentrations of test compound for 8 days and harvested on the last day of culture. After total cellular DNA was prepared, HBV DNAs derived from both replication intermediates and integrated HBV genome were detected and analyzed by Southern blotting. Percent control values were calculated from the untreated controls (U). The bands derived from integrated HBV are indicated by the arrows. The mean value at each concentration is plotted on the graph, and the standard deviations are indicated by the error bars. FIG. 3. | Comparison of the inhibitory effects of MCC-478, M-1, M-2, and F-1 against HBV replication in HB611 cells. Comparison of the inhibitory effects of MCC-478, M-1, M-2, and F-1 against HBV replication in HB611 cells. Cells were treated with the indicated concentrations of test compound for 8 days. Anti-HBV activity was evaluated as indicated in the legend to Fig. . The mean value at each concentration is plotted on the graph, and the standard deviations are indicated by the error bars. FIG. 4. | Effects of MCC-478 and its derivatives on mitochondrial DNA content in HepG2 cells. Effects of MCC-478 and its derivatives on mitochondrial DNA content in HepG2 cells. Cells were treated with the indicated concentrations of test compound for 8 days and harvested on the last day of culture. After total cellular DNA was prepared, mitochondrial DNA content in 2.5 mug of total cellular DNA was evaluated by dot blot analysis. The DNAs from nuclei (Nuc) and mitochondrial (Mt) fractions of HepG2 cells are indicated by the arrows. Percent control values were calculated from the untreated controls. The mean value at each concentration is plotted on the graph, and the standard deviations are indicated by the error bars. TABLE 1 | Summary of the antiviral and cytotoxic effects of the test compounds DISCUSSION : MCC-478 is a novel phosphonomethoxyethylpurine derivative which has an arylthio group at position 6 of the purine base. Also, MCC-478 has 2-aminopurine nucleobase structure, which indicates that it is more closely related to 9-[2-(phosphonomethoxy)ethyl]guanine (PMEG) than to PMEA. However, introduction of the arylthio group at position 6 of the 2-aminopurine base has been demonstrated to endow the purine base with a higher specificity against HBV and low cytotoxicity comparable to that of PMEA , while PMEG is known to have higher cytotoxicity than PMEA . These results suggest that substitution at position 6 of the purine base may alter the biological and toxicological profiles of the phosphonomethoxyethyl purine derivatives. It was reported that adefovir dipivoxil is primarily hydrolyzed to the monoester form and is further metabolized to adefovir (PMEA) in the presence of either cells or serum . This finding may mean that there was an esterase to hydrolyze both bis(pivaloyloxymethyl) and mono(pivaloyloxymethyl) ester forms in both serum and cells. In the case of MCC-478, which is a bis(2,2,2-trifluroroethyl) ester, it was hydrolyzed to only the monoester form, even in the presence of serum from humans or rats (data not shown). In addition, major metabolites in rats and monkeys were found to be the monoester forms of MCC-478 (M-1 and M-2), rather than the free phosphonic acid form (F-1) (Y. Yamaguchi, personal communication). On the other hand, the free phosphonic acid form, PMEA, was identified as the major metabolite in sera from the animals given adefovir dipivoxil orally. This speculated difference of metabolites between MCC-478 and adefovir dipivoxil in vivo suggested the necessity of examining the antiviral activities of these metabolites in order to estimate their efficacy in vivo. Here, not only the monoester forms but also the free phosphonic acid form were recognized to have antiviral activities similar to that of MCC-478 in HB611 cells . The anti-HBV activities of these MCC-478 derivatives were high enough to suggest that they could contribute to in vivo efficacy of MCC-478. These results also indicated that the mono(2,2,2-trifluoroethyl) ester forms of MCC-478 may be hydrolyzed only within the cell and further metabolized to an active form of MCC-478. In addition to the antiviral profile of MCC-478, it was also shown that the cytotoxic effects of MCC-478 and its hydrolyzed derivatives were quite low . It may also be important to evaluate mitochondrial toxicity to understand the biological profile of a nucleotide analogue, since long-term treatment with antiviral nucleoside analogues could give rise to delayed and possibly severe mitochondrial toxicity . Nucleoside and nucleotide analogs can be classified into two types on the basis of their mechanism of mitochondrial toxicity. The first type of nucleoside analogue, such as fialuridine, has a 3'-hydroxyl group (3'-OH), and mitochondrial toxicity is associated with nucleoside incorporation into the mitochondrial DNA . The second type of nucleoside analogue, such as ddC, does not have a 3'-OH group, and mitochondrial toxicity is associated with reduced mitochondrial DNA content by chain termination . Since MCC-478 and its hydrolyzed derivatives do not have 3'-OH, they may belong to the latter type of nucleoside analogues. However, as shown in Fig. , there was no evidence of inhibition of mitochondrial DNA replication. This lack of evidence of inhibition may mean that MCC-478 and its metabolites did not have inhibitory activity against DNA polymerase gamma. A chain-terminating analysis using DNA polymerase alpha also revealed that neither F-1 nor its diphosphate showed chain-terminating activity against DNA polymerase alpha; however, PMEA diphosphate was thought to have chain-terminating activity at the same concentrations (data not shown). These results emphasize the specific antiviral profiles of MCC-478 and its derivatives. Interestingly, the diphosphates derived from MCC-478 were not found intracellularly, so far as we tried. While there is a possibility that the diphosphates might be a very potent and selective inhibitor of hepadnavirus polymerase, MCC-478 and its derivatives might not require phosphorylation for biological activity. More studies of intracellular metabolism and active metabolites would be needed to elucidate the mechanisms of MCC-478 antiviral activity. Lamivudine is the first approved nucleoside analogue for treatment of HBV infection. It is known to have potent activity against both HBV and HIV replication . However, the rapid emergence of viruses resistant to lamivudine in HIV-infected patients is also known, even with short-term therapy, and recently a similar phenomenon was observed with lamivudine monotherapy . Most lamivudine-resistant virus strains have methionine (M) substituted for isoleucine (I) or valine (V) on the YMDD (tyrosine-methionine-aspartate-aspartate) motif in catalytic domains of HIV polymerase and hepadnavirus polymerase . Another drug-resistant substitution of leucine (L) to M in B domain (L528M) is reported from a clinical trial of famciclovir in HBV-infected patients . The L528M mutation frequently accompanies the YMDD mutation, and HBV carrying these two mutations is cross-resistant to lamivudine and famciclovir . In the case of HIV, a steric hindrance between the mutant amino acid side chain and lamivudine triphosphate is suggested from the molecular model analysis of comparisons of crystal structures of the YMDD mutant and wild-type HIV polymerases with or without double-strand DNA . Moreover, this molecular model suggested that a steric hindrance between the mutant amino acid side chain and the lamivudine sugar ring might be expected in nucleoside analogues with beta-l- ring configurations. In the case of HBV, a recent study reporting molecular modeling analysis using the HBV polymerase homology model also supports this concept. Ono et al. evaluated the inhibitory effects of a panel of 11 nucleoside and nucleotide analogues on HBVs with one or two mutations by in vitro full-length HBV DNA transfection and found that only a few nucleoside analogues, including PMEA, are active against lamivudine-resistant HBVs. Since MCC-478 was shown to be a novel nucleotide analogue having potent and selective anti-HBV activity, even if it was a derivative of phosphonomethoxyethylpurine, like PMEA, MCC-478 should be further examined for an inhibitory profile against drug-resistant HBVs, including lamivudine-resistant HBVs. The aim of this work was to evaluate MCC-478 and its derivatives as potent and specific agents for the treatment of HBV infections. Although many studies have to be done to understand the mechanism of action, toxicological pharmocokinetic profiles, and so on, MCC-478 could be a promising new anti-HBV agent. Backmatter: PMID- 12183227 TI - Heterologous Expression of Epothilone Biosynthetic Genes in Myxococcus xanthus AB - Epothilones are potential anticancer drugs that stabilize microtubules in a manner similar to paclitaxel (Taxol). Epothilones are produced from the myxobacterium Sorangium cellulosum, which has a 16-h doubling time and produces only milligram-per-liter amounts of epothilone A and epothilone B. Furthermore, genetic manipulation of S. cellulosum is difficult. To produce epothilones in a more genetically amenable and rapidly growing host, we chose the closely related and best-characterized myxobacteria Myxococcus xanthus. We inserted 65.4 kb of S. cellulosum DNA that encompassed the entire epothilone gene cluster into the chromosome of M. xanthus by a series of homologous recombination events. The resulting strain produced epothilones A and B. Construction of a strain that contained a mutation in epoK, the P450 epoxidase, resulted in production of epothilones C and D. Keywords: Introduction : Epothilones are potent cytotoxic macrocyclic lactones that show promise as anticancer drugs . The mechanism of action is similar to the anticancer drug paclitaxel (Taxol); both bind and stabilize microtubules, which leads to cell death . Epothilones have superior features relative to paclitaxel. First, epothilones are more water soluble. This may enable a formulation without the use of the solubilizing agent cremophor, currently used in formulations of paclitaxel. Cremophor, on its own, can affect cardiac function and cause severe hypersensitivity . Second, epothilones are effective against tumors resistant to paclitaxel . These advantages make epothilones likely successors to paclitaxel. The need for sufficient material is a major obstacle to the development of epothilones as marketable drugs. The total synthesis of epothilone A and epothilone B has been accomplished . However, the number of steps required for synthesis of these molecules precludes this as an economical method of production. Fermentation methods thus remain the favored route of production. Epothilones are produced from the gram-negative myxobacterium Sorangium cellulosum . The reported yields of epothilones from S. cellulosum strain So ce90 are approximately 20 mg of epothilone A per liter and 10 mg of epothilone B per liter . A disadvantage of S. cellulosum is the relatively long doubling time, which is approximately 16 h and is the longest of all myxobacteria. Furthermore, S. cellulosum is difficult to engineer, due to the low efficiency of introducing DNA into the bacterium and the limited number of molecular tools and markers that have been developed. Recently, the epothilone biosynthetic gene cluster was sequenced, and the genes were introduced into Streptomyces coelicolor, a common host used for production of a variety of polyketides from actinomycetes . The heterologous strain produced small quantities of epothilones; production of epothilones in S. coelicolor may have a cytotoxic effect (L. Tang et al., unpublished data). Thus, an alternative heterologous host is desired. We speculated that a superior expression host for epothilones might be another myxobacterium and chose Myxococcus xanthus for several reasons. First, M. xanthus is the best-characterized myxobacterium and is readily amenable to engineering . Second, M. xanthus has a significantly shorter doubling time than S. cellulosum (5 versus 16 h), which would in itself enhance volumetric productivity. Third, expression of the epothilone genes in M. xanthus may not require the engineering of new promoters, because M. xanthus and S. cellulosum are closely related organisms. The epothilones are polyketides that are synthesized by a type I polyketide synthase . These enzymes are large multifunctional complexes organized in a modular fashion and catalyze the successive condensation of carboxylic acid residues from their coenzyme A (CoA) esters, typically malonyl-CoA and methylmalonyl-CoA . Polyketide synthase enzymes are synthesized as an apo form and are converted to the holo form by the addition of a phosphopantetheinyl (P-pant) moiety to a serine residue of the acyl or peptidyl carrier protein (ACP or PCP) domains by an enzyme called P-pant transferase . Thus, a heterologous host must synthesize malonyl-CoA and methylmalonyl-CoA as well as harbor a P-pant transferase. FIG. 1. | Genetic map of the epothilone biosynthetic gene cluster. Genetic map of the epothilone biosynthetic gene cluster. The boxes represent genes, and the modules encoded by each are designated below. The long arrow represents the direction of transcription. Before initiating this work, it was unknown whether M. xanthus contains a P-pant transferase or methylmalonyl-CoA. It is known that strains of M. xanthus produce the polyketide myxovirescins, also known as TA. Thus, it is likely to contain a P-pant transferase, but whether it would function on the ACP and PCP domains of the epothilone polyketide synthase remained to be determined. Although myxovirescins contains methyl groups extending from the ring, they are all derived from S-adenosylmethionine and not from the utilization of methylmalonyl-CoA. Thus, evidence for the production of methylmalonyl-CoA cannot be deduced from the polyketides that are made naturally from this host. Work studying the developmental life cycle of M. xanthus has revealed the presence of the genes for propionyl-CoA carboxylase , an enzyme used to synthesize methyl malonyl-CoA. Thus, M. xanthus is likely to possess required components to synthesize epothilone. Because of the advantages of using M. xanthus as a host for production of epothilone, we introduced the epothilone genes from S. cellulosum into the chromosome of M. xanthus. We demonstrate that the engineered strain produces epothilones. We also constructed mutations in epoK, the CYP450, which is responsible for the formation of the epoxide at C-12-C-13. This strain produces epothilone C and epothilone D. MATERIALS AND METHODS : Bacterial strains. | Escherichia coli strains XL1-Blue and DH10B were used for transformations. M. xanthus strain DZ1 was the recipient for the epothilone genes . It is streptomycin resistant and is a nonmotile strain that is unable to form fruiting bodies Media and growth conditions. | E. coli with plasmids were grown in Luria-Bertani medium containing 0.5% NaCl at 37C supplemented with ampicillin (50 mug/ml), kanamycin (50 mug/ml), or tetracycline (15 mug/ml). DZ1 was grown in CYE at 30 or 32C. For selection of galactose resistance in M. xanthus, cells were plated in 2.5 ml of CYE top agar and poured onto CYE plates containing 1% galactose. The following antibiotics were used for M. xanthus: kanamycin (50 mug/ml) and oxytetracycline (10 to 15 mug/ml). To test for the production of epothilone, cells were cultured in CMM (0.5% Casitone, 0.2% MgSO4, 10 mM morpholinepropanesulfonic acid [MOPS] [pH 7.6]) supplemented with the following amino acids (1 mg/liter): serine, glycine, and alanine. The cells were initially grown in CYE medium, and while in mid-log phase a 5% inoculum was used to inoculate a flask containing production medium. The cultures were grown at 30C for 64 to 72 h. XAD-16 is an absorber resin that was used in production cultures to bind the epothilones. To prepare the XAD-16, 2% was added to a fermentation flask, covered with water, and autoclaved. Afterwards, the water was removed, CMM was added, and the culture was inoculated. Extraction of epothilones. | To extract epothilones from cells and from the culture (cells and medium), an equal volume of acetone was added and shaken at room temperature for an hour. The acetone mixture was extracted twice with an equal volume of ethyl acetate, and the ethyl acetate fraction was dried and resuspended in acetonitrile. To analyze epothilones bound to XAD-16, the culture broth was removed and the resin was washed twice with 5 volumes of MilliQ water. The epothilones were eluted with methanol and analyzed by liquid chromatography-mass spectrometry . DNA methods. | Standard DNA protocols were used for E. coli. The isolation of M. xanthus chromosomal DNA was done as previously described . Electroporation of M. xanthus was described by Kashefi and Hartzell Transposition reactions. | The tetracycline resistance gene was added to the cosmids pKOS35-70.1A2 and pKOS35-79.85 using the in vivo transposition reaction kit from Epicentre Technologies. The transposon used was EZ::TN, and the reaction was performed as recommended by the manufacturer. Resulting tetracycline-resistant colonies were screened for resistance to kanamycin. Those colonies resistant to tetracycline and sensitive to kanamycin were kept, and the DNA was digested with restriction enzymes to verify that no deletions had occurred. The tetracycline resistant versions of pKOS35-70.1A2 and pKOS35-79.85 are pKOS90-38 and pKOS90-23, respectively. Construction of pKOS35-154 and pKOS90-22. | Plasmids that contained regions of the epo genes flanked by M. xanthus DNA were constructed for homologous recombination of the epo genes into the chromosome. Plasmid pKOS35-154 was constructed in several steps. First, the ca. 3-kb BamHI-to-NdeI fragment from KG2 was ligated into the BglII and NdeI sites of pSL1190 (Amersham-Pharmacia). This plasmid, pKOS55-178, contains the kanamycin resistance and galK cassette that is used for positive and negative selection in M. xanthus. Next, a 4.7-kb NotI fragment from cosmid pKOS35-79.85, containing a portion of the coding sequence for module 7, was ligated into the NotI site of pKOS55-178 to create pKOS55-183.b. Plasmid pKOS55-183.b was cleaved with MfeI, and the DNA ends were made blunt with the Klenow fragment of DNA polymerase I and ligated with a 5,088-bp SrfI fragment from pKOS35-79.85. This plasmid, pKOS90-1, contains two regions of the epo gene cluster flanking the kanamycin resistance and galK cassette. To add flanking DNA from M. xanthus to recombine the epo genes from pKOS90-1 into the chromosome, plasmid pKOS35-151 was constructed as follows. Plasmid pBJ130 , which carries the dev locus from M. xanthus, was cleaved with KpnI; the DNA ends were made blunt with the Klenow fragment of DNA polymerase I and ligated with the polylinker from pSL1190 (Pharmacia), which had been cleaved with EcoRI and HindIII; and the DNA ends were made blunt with the Klenow fragment. This plasmid, pKOS35-151, was cleaved with SpeI and EcoRI and ligated with the 12-kb SpeI-EcoRI fragment from pKOS90-1 to create pKOS35-154. Plasmid pKOS90-22 was used for recombining the 5' end of the epo gene cluster into the M. xanthus chromosome. First, a portion of the epo genes encoded by module 6 and module 7 was isolated as an 8.8-kb BglII fragment from cosmid pKOS35-70.4 and ligated into the BamHI site of pKOS55-178 to create pKOS90-12. To isolate a fragment upstream of the start of the epo genes, a 9.5-kb NsiI-to-AvrII fragment from cosmid pKOS35-70.8A3 was ligated into the PstI and AvrII sites of pSL1190 to yield pKOS90-13. Next, pKOS35-151 was cleaved with EcoRI and HindIII, the DNA ends were made blunt with the Klenow fragment, and the plasmid was religated. This plasmid, pKOS90-5, was cleaved with MfeI and AvrII and ligated with the 11.5-kb EcoRI-SpeI fragment from pKOS90-12 to construct pKOS90-17. Finally, pKOS90-22 was constructed by cleaving pKOS90-17 with SpeI and ligating it with the 9.5-kb SpeI-AvrII fragment from pKOS90-13. Construction of an epoK mutant. | To construct an epoK mutant, a kanamycin resistance cassette was inserted into the epoK gene. This was done by isolating the 4,879-bp fragment from pKOS35-79.85, which contains epoK, and ligating it into the NotI site of pBluescript SKII(+). This plasmid, pKOS35-83.5, was partially cleaved with ScaI, and the 7.4-kb fragment was ligated with the 1.5-kb EcoRI-BamHI kanamycin resistance gene from pBJ180-2 (B. Julien and D. Kaiser, unpublished), which had the DNA ends made blunt with the Klenow fragment, to yield plasmid pKOS90-55. Finally, the ca. 400-bp RP4 oriT fragment from pBJ183 (Julien and Kaiser, unpublished) was ligated into the XbaI and EcoRI sites to create pKOS90-63. This plasmid was linearized with DraI and electroporated into the M. xanthus strain K111-32.25 to create K111-40. To create a markerless epoK mutation, pKOS35-83.5 was cleaved with ScaI, and the 2.9- and 4.3-kb fragments were ligated together. This plasmid, pKOS90-101, has an in-frame deletion in epoK. Next, the 3-kb BamHI-NdeI fragment from KG2, which had the DNA ends made blunt with the Klenow fragment and contains the kanamycin resistance and galK genes, was ligated into the DraI site of pKOS90-101 to create pKOS90-105. This plasmid was electroporated into K111-32, and kanamycin-resistant colonies were selected. To replace the wild-type copy of epoK with the deletion, the second recombination event was selected by growth on galactose plates. These galactose-resistant colonies were screened for production of epothilone C and D. This strain is designated K111-72. RESULTS : Construction of an M. xanthus strain containing the epothilone gene cluster. | To reconstruct the epothilone gene cluster in M. xanthus, we chose to introduce the cluster into the chromosome by homologous recombination. This entails placing regions of homology from the epothilone genes into the chromosome of M. xanthus and using these regions as recombination sites for cosmids containing parts of the epo gene cluster. The location on the M. xanthus chromosome chosen for the recombination site was 3' of the dev operon . This region encodes no recognizable genes, and insertions in this region are not predicted to disrupt any transcription. The plasmids used and the strategy for introducing part of the epo genes in M. xanthus are diagramed in Fig. and . First, plasmid pKOS35-154 was constructed. This plasmid contains a 7-kb fragment of M. xanthus DNA, 2.8 kb of the dev operon and 4.2 kb downstream of dev, and two regions from the epo gene cluster: a region from module 7 and a region 4.7 kb downstream of epoK. Between the two epo fragments is the kanamycin resistance gene from Tn5 and the E. coli galK gene. Plasmid pKOS35-154 was linearized and electroporated into M. xanthus to produce strain K35-159 . Kanamycin-resistant colonies arise from a double recombination event: one recombination with the dev locus and another with the region 3' to dev. This introduces regions from the epo gene cluster that serve as recombination sites for a cosmid containing the downstream half of the epo genes. Next, the cosmid pKOS90-23, which contains epoE, epoF, epoK, and downstream DNA, was electroporated into K35-159 . This strain, K35-174.6, results from a single recombination event. To remove the cosmid vector and leave only the region between the two flanking epo regions inserted by plasmid pKOS35-154, a second recombination event was selected by isolating galactose-resistant colonies. The E. coli galK gene renders M. xanthus sensitive to 1% galactose and provides a convenient method for isolating bacteria that have lost the galK gene, which occurs by a second recombination event between the flanking epo region and its homologous DNA on the cosmid. This results in strain K35-175, which contains part of epoE, the complete epoF and epoK genes, and 4.7 kb of DNA downstream . The 4.7 kb of DNA downstream of epoK contains several putative open reading frames (arrows in Fig. ). None of the open reading frames contain homologies to proteins which would suggest a function. However, the first and last open reading frames encode proteins predicted to harbor membrane spanning regions and so may code for transporter proteins. To insert epoA, epoB, epoC, epoD, and the remaining region of epoE, a second round of recombination, similar to the first, was performed. Plasmid pKOS90-22 was linearized and electroporated into K35-175 to construct K111-13.2. Next, cosmid pKOS90-38 was electroporated into K111-13.2 to construct K111-13.22. Finally, to complete the construction of the gene cluster, a galactose-resistant version of K111-13.22 was selected. This final recombinant strain contains the epothilone gene cluster plus 4.7 kb upstream of the translational start of epoA and 4.7 kb downstream of the translational stop of epoK. Eighteen isolates were examined for epothilone production and analyzed by Southern blotting. From this, the two isolates that produced the highest amounts of epothilones were further examined and are referred to as K111-32 isolates 1 and 2. Although the epothilones are secreted into the medium when produced in S. cellulosum, we were not certain whether a specific transporter was needed to transport the epothilones out of the cell. In constructing the M. xanthus epothilone producer, two open reading frames that appear to encode membrane-spanning proteins were contained within the 65.4 kb of DNA. The two isolates of K111-32 were grown in CMM medium for 60 to 72 h, the period of maximum epothilone accumulation (data not shown), and the levels of epothilone were analyzed in the total culture, which contains the cells and medium together; in the medium; and in the cells alone. The results show that the epothilones are secreted into the medium and only trace amounts are retained in the cells. We currently do not know if the two open reading frames, predicted to encode membrane-spanning proteins, are required for the export of the epothilones. Comparison of the epothilones produced in M. xanthus versus those produced in S. cellulosum shows that the levels of epothilone B are about 100-fold lower in M. xanthus under the growth conditions tested. Interestingly, the ratio of epothilone A to epothilone B produced in M. xanthus is approximately 1:10, in stark contrast to S. cellulosum, where the ratio of epothilone A to epothilone B is roughly 2:1. The difference between epothilone A and epothilone B is the presence of a methyl group at carbon 12. This methyl group is derived from methylmalonyl-CoA and not by C methylation, which indicates that the acyltransferase from module 4 is able to accept either malonyl-CoA or methylmalonyl-CoA . Thus, in M. xanthus, there may be a larger pool of methylmalonyl-CoA during epothilone production that increases the production of epothilone B over epothilone A. Construction of an epoK mutant. | Epothilone A and epothilone B are the major constituents of the fermentation of the natural and heterologous hosts. However, epothilone D appears to have the best therapeutic index (-, ). Of the two other candidate drugs currently in clinical trials , epothilone B and aza-epothilone B, epothilone D proved superior in reducing tumor size in nude mice containing MX-1 tumor cells. Epothilone B is synthesized from epothilone D by the formation of the epoxide at C-12-C-13, which is added by EpoK, a cytochrome P450 . This has been demonstrated in vitro by the conversion of epothilone D to epothilone B in the presence of purified EpoK . Thus, a strain that contains an epoK mutation would be a valuable strain for production of epothilone D. Two strains that contain mutations in epoK were constructed. One contains an insertion of the kanamycin gene from Tn5, and the other has an in-frame deletion of epoK that removes 705 bp, or 56% of the gene. Although there are no genes downstream of epoK that are necessary for the synthesis of epothilones, there may be genes important for export, particularly the first open reading frame downstream of epoK that is predicted to contain several transmembrane spanning regions. Therefore, it was important to construct mutations that will not have polar effects. With the insertion, the kanamycin resistance gene is oriented such that its promoter would drive expression of genes downstream of epoK. Strains harboring mutations in epoK were grown, and the amount of epothilones produced are presented in Fig. A. Isolates of K111-40 contain the insertion of the kanamycin resistance gene (epoK::npt), whereas the isolates of K111-72 contain an in-frame deletion (DeltaepoK). The data reveal that the levels of epothilone A from the isolates of K111-32 are similar to the levels of epothilone C from the strains carrying a mutation in epoK and that both types of mutations, the insertion and the in-frame deletion, result in the same levels of production. Interestingly, the levels of epothilone D produced are about fivefold lower than the epoK mutant strains relative to epothilone B produced from K111-32. Because the resin XAD-16 is used in the fermentation of the S. cellulosum strain that produces epothilone, we investigated whether binding of the epothilones to the resin XAD-16 would sequester them from the cells and stabilize the compounds, particularly epothilone D. Figure shows the results of fermentations in the presence XAD-16. The levels of epothilone D produced from the strains harboring the epoK mutation are higher in the presence of XAD and are equivalent to epothilone B levels produced in K111-32. These results suggest that epothilone D is unstable during the fermentation process relative to the other epothilones. Interestingly, the strain K111-32 produced equal amounts of epothilone B and epothilone D in the presence of XAD-16, even though it contains a fully functional epoK. This is not seen in epothilone production from S. cellulosum. In the presence of XAD-16, epothilone C and epothilone D make up less than 10% of the total amount of epothilones produced in shake flask experiments (B. Julien, unpublished data). This suggests that in S. cellulosum, the majority of the epothilone D is converted to epothilone B before release into the medium, whereas in M. xanthus, epothilone D has an equal probability of being released into the medium as being acted upon by EpoK. FIG. 2. | Plasmids used for introduction of recombination sites for epothilone genes. Plasmids used for introduction of recombination sites for epothilone genes. Lightly shaded boxes represent regions of DNA from M. xanthus. Dark boxes represent selectable markers or regions of DNA from the epothilone gene cluster. Arrows represent the direction of transcription. FIG. 3. | Schematic for introduction of a portion of the epothilone genes into the chromosome of M. xanthus Schematic for introduction of a portion of the epothilone genes into the chromosome of M. xanthus. (A) The introduction of recombination sites for the epothilone genes from pKOS35-154. Arrows indicate direction of transcription. Boxes with curved ends represent partial genes or fragments of DNA from S. cellulosum. (B) Introduction of the cosmid pKOS90-23 into the chromosome. (C) Resulting genetic structure after selecting for the second recombination event to remove duplicated regions and regions with the kanamycin resistance and galK genes. Arrows indicate putative open reading frames and their direction of transcription in the 4.5 kb of S. cellulosum DNA downstream of epoK. FIG. 4. | Production levels from various strains in the absence (A) or presence (B) of XAD-16. Production levels from various strains in the absence (A) or presence (B) of XAD-16. TABLE 1 | Epothilone production in K111-32 DISCUSSION : Myxobacteria produce a variety of novel polyketides with novel modes of action, including the epothilones . Current understanding of polyketide synthase genes allows them to be engineered to produce novel compounds that may have superior activity relative to the natural ones. However, engineering of genes in strains of S. cellulosum, as well as in most other myxobacteria, would be extremely difficult, due to the lack of genetics and molecular techniques available. Although Streptomyces expression systems have been developed for production of polyketides , the yields can be low, and unforeseen side effects of heterologously producing a non-actinomycetes compound, particularly from a gram-negative bacterium, can be problematic, as seen with the epothilones (Tang et al., unpublished data). Thus, the work presented here demonstrates that M. xanthus provides an important alternative host for production of polyketides, especially those from other myxobacteria. Although the levels of epothilones reported here are low, through medium development and using a feeding strategy of Casitone and methyl oleate, current levels of epothilone B are higher than those obtained with the epothilone-producing S. cellulosum strain So ce90 . Furthermore, mutants that harbor a mutation in epoK produce 50 to 100 times more epothilone D than does So ce90. From the work presented here, important information was obtained as to the use of M. xanthus as a heterologous host. First, the strain used has a P-pant transferase that is able to modify both ACP and PCP domains. Second, the strain has a pool of methylmalonyl-CoA during the time of epothilone synthesis. Furthermore, during this time period, there may be more methylmalonyl-CoA than malonyl-CoA, as evidenced by the ratios of epothilone A and epothilone B produced, which is approximately 1:10. Finally, the promoter that drives expression of the epothilone genes in S. cellulosum most likely is the same one that is utilized in M. xanthus. Since no specific regulator proteins have been found associated with any of the several polyketide synthase genes sequenced from S. cellulosum or Stigmatella aurantiaca (, , -), there is likely some global regulator that activates production of secondary metabolites at the onset of stationary phase, and this regulation is similar in many myxobacteria. Further support is provided by the fact that the myxothiazole promoter from S. aurantiaca can drive expression of the epothilone genes in M. xanthus (S. Shah and B. Julien, unpublished data). Now that epothilone can be produced in a heterologous host tolerant of its production and amenable to genetic manipulation, the stage is set to make new compounds that can only be made by engineering of the genes and to identify those with increased therapeutic value. Backmatter: PMID- 12183269 TI - Functional Characterization of Brucella melitensis NorMI, an Efflux Pump Belonging to the Multidrug and Toxic Compound Extrusion Family AB - Two putative proteins (NorMI and NorMII) similar to the multidrug efflux protein NorM of Vibrio parahaemolyticus are encoded by the Brucella melitensis 16 M genome. We show that a drug-hypersusceptible Escherichia coli strain overexpressing NorMI displays increased resistance to norfloxacin, ciprofloxacin, gentamicin, tetraphenylphosphonium ion, acriflavine, and berberine. This elevated resistance was proven to be mediated by an energy-dependent efflux mechanism. NorMI belongs to the multidrug and toxic compound extrusion family and is the first multidrug efflux protein identified in Brucella spp. Keywords: Introduction : Although brucellosis is primarily a disease of domestic animals, it is still a common human disease in many developing countries . The pathogens responsible for brucellosis, Brucella spp., are facultative intracellular bacteria. Therefore, to be effective, the treatment of acute brucellosis requires drugs that penetrate macrophages. With conventional drugs such as tetracycline, aminoglycosides, or sulfonamides, high relapse rates have been reported . More recently, increasing resistance of Brucella spp. to co-trimoxazole was reported . This has led to the search for new drugs which penetrate eukaryotic cells more efficiently. The broad-spectrum in vitro activity of fluoroquinolones such as ciprofloxacin, as well as their oral bioavailability and their excellent intracellular penetration, initially made them very promising for treating brucellosis . However, a lack of effective bactericidal activity of fluoroquinolones against Brucella spp. , an appreciable rate of relapse of patients treated with ciprofloxacin , and development of ciprofloxacin resistance associated with cross-resistance to other fluoroquinolones in Brucella melitensis during ciprofloxacin therapy have been also reported . Nothing is known about the mechanism of fluoroquinolone resistance development in B. melitensis. In other bacteria, resistance to fluoroquinolones usually results from multiple mutations in genes encoding their intracellular targets, gyrase and topoisomerase IV . In addition to this mechanism, which is antibiotic specific, a more general mechanism involves multidrug efflux pumps which exclude toxic compounds, including antibiotics, from the cells. Some of these pumps exhibit low specificity and confer resistance to several unrelated antibiotics when they are overexpressed. These bacterial pumps are classified in five large, ubiquitous superfamilies . One of these, the multidrug and toxic compound extrusion (MATE) family, was recently identified. It includes several bacterial members but also eukaryotic proteins from fungi and plants . Among these, only a few have been functionally characterized. Four bacterial members, NorM and VmrA of Vibrio parahaemolyticus, YdhE of Escherichia coli, and BexA of Bacteroides thetaiotaomicron, have been reported to mediate multidrug resistance (, -). NorM and VmrA were also reported to function by a drug:Na+ antiport mechanism . Among the eukaryotic members, the only two functionally characterized proteins are yeast Erc1, which confers resistance to the methionine analog ethionine, and Arabidopsis Alf5, conferring resistance to toxins . Proteins of the MATE family have a common predicted topology with 12 transmembrane helices . A B. melitensis homolog of NorM was recently identified as a potential virulence factor, as strains with mutations in its gene could not be recovered from animals infected with pools of signature tag mutants (M. S. Zygmunt, S. D. Hagius, W. T. Fulton, J. V. Walker, N. J. Booth, and P. H. Elzer, Brucellosis 2000, 53rd Brucellosis Research Conference, abstr. 56, 2000). In the present work, we functionally characterized this transporter and confirmed that it belongs to the MATE family. Identification of sequences belonging to the MATE family in the genome of B. melitensis 16 M. : By searching for homology with the NorM multidrug efflux protein of V. parahaemolyticus (BlastP) in the B. melitensis 16 M genomic database (accession numbers and ), we found that two putative multidrug efflux pumps belonging to the MATE family are encoded by this genome . The first one, which we named NorMI, is encoded by gene BMEI1585 and corresponds to the above-mentioned protein that was recently identified as a potential virulence factor. The second, which we named NorMII, is encoded by gene BMEI1612. NorMI and NorMII share 27.7 and 19.9% identical amino acids (in an overlap of 448 amino acids), respectively, with the NorM protein of V. parahaemolyticus. As NorMI is more similar to NorM than NorMII, and consequently is more likely to have a similar function, we focused this work on NorMI. To confirm that NorMI belongs to the MATE family, we searched (with BlastP) the EMBL redundant protein database to determine whether NorMI is also homologous to other proteins of this family. Besides the homology with NorM, NorMI also possesses 22.2 to 28.6% amino acid identity to the E. coli YdhE protein, the B. thetaiotaomicron BexA protein, the V. parahaemolyticus VmrA protein, the Arabidopsis Alf5 protein, and the yeast Erc1 protein. No homology was found with any other functionally characterized bacterial proteins. A prediction of transmembrane helices by using the TMHMM server (version 2.0) from the Center for Biological Sequence Analysis revealed that NorMI possesses 12 putative transmembrane helices. These data strongly suggest that NorMI represents a new member of the MATE family. NorMI confers resistance to multiple drugs. : To study the involvement of NorMI in drug efflux, we overexpressed this protein in the drug-hypersusceptible E. coli strain AG100A, which was deleted of the major multidrug efflux system AcrAB by Okusu et al. . The norMI gene was cloned into the pUC19 vector under the control of its natural promoter. In brief, norMI was first amplified by PCR from the chromosomal DNA of B. melitensis 16 M with primers P7 (5'-TCGGATCCGGGACGGAAATTTGCGCTTTC-3') and P4 (5'-GGAATTCCCTGAAAGGCTTCGGTGCCGC-3') (1 muM final concentration) and Pfu DNA polymerase (Promega) in an iCycler thermocycler (Bio-Rad) with the following cycling conditions: one cycle of 2 min at 95C; then 35 cycles of 1 min at 95C, 30 s at 55C, and 4 min at 72C; and finally one cycle of 5 min at 72C. The amplified DNA fragment was restricted with BamHI and EcoRI and inserted between the BamHI and EcoRI sites of pUC19. The resulting plasmid (pUC19-NorMI) and pUC19 were then electroporated separately in E. coli AG100A. The capacity of the construct to confer drug resistance to E. coli AG100A was evaluated by comparing the susceptibilities of the strains carrying pUC19 and pUC19-NorMI to increasing concentrations of many unrelated drugs (norfloxacin, ciprofloxacin, gentamicin, tetraphenylphosphonium bromide, erythromycin, nalidixic acid, chloramphenicol, ofloxacin, quinacrine, tetracycline, ethidium bromide, carbonyl cyanide m-chlorophenylhydrazone [CCCP], and sulfamethoxazole). Exponentially growing bacteria (105) were inoculated into 5 ml of Mueller-Hinton medium containing twofold-increasing concentrations of each drug. For each drug concentration, growth was monitored by measuring the turbidity of the culture at 600 nm after 20 h of incubation at 37C under agitation. As shown in Fig. , NorMI confers increasing resistance of strain AG100A towards hydrophilic fluoroquinolones such as norfloxacin and ciprofloxacin, to aminoglycosides such as gentamicin, and to the tetraphenylphosphonium cation. A slight increase of resistance to acriflavine and berberine was also observed. On the other hand, AG100A carrying pUC19 and pUC19-NorMI showed indistinguishable susceptibilities to the other antimicrobial agents tested (data not shown). The substrate specificity of NorMI is highly similar to that of YdhE of E. coli . Indeed, like NorMI, YdhE confers resistance to norfloxacin, ciprofloxacin, acriflavine, berberine, and tetraphenylphosphonium ion. Although the specificity of YdhE resistance to gentamicin was not tested, this protein confers resistance to similar aminoglycosides such as kanamycin and streptomycin. These results indicated that NorMI mediates multidrug resistance. The contribution of NorMI to the drug resistance of B. melitensis was tentatively approached by disrupting norMI in this strain. Briefly, a kanamycin resistance cassette extracted from pUC-4K (Pharmacia Biotech) by EcoRI restriction was inserted into the unique BssHII site of the pUC19-NorMI construct, a suicide vector for Brucella spp. After introduction of this construct in B. melitensis cells by electroporation, kanamycin-resistant clones were selected and tested by PCR to confirm the correct insertion of the kanamycin cassette and the disruption of norMI. A disruption mutant was selected, and its susceptibility to ciprofloxacin, one of the best substrates of NorMI, was compared with that of the wild-type strain. This was done as described above for E. coli, except that 107 cells inoculated into 5 ml of Trypticase soy broth (Difco Laboratories) supplemented with 0.1% yeast extract were used. The mutant and wild-type strains showed indistinguishable susceptibilities to ciprofloxacin (data not shown). A number of other putative efflux pumps (including seven homologs of the major E. coli AcrAB pump, at least three homologs of the E. coli EmrAB pump, and two ABC transporters similar to those involved in drug efflux) are encoded by the B. melitensis genome . Those pumps may well mask the effect of the norMI mutation, and for this reason, we were obliged to analyze the function of NorMI in E. coli cells previously mutated in their major AcrAB pump . FIG. 1. | Effect of drugs on growth of E. coli Effect of drugs on growth of E. coli cells expressing or not expressing the NorMI protein of B. melitensis. E. coli AG100A harboring a vector expressing NorMI from B. melitensis (pUC19-NorMI) or harboring the vector alone (pUC19) was grown for 20 h at 37C (with agitation) in Mueller-Hinton medium containing twofold-increasing concentrations of norfloxacin (1.5625 to 200 ng/ml), ciprofloxacin (0.3125 to 40 ng/ml), gentamicin (31.25 to 4,000 ng/ml), tetraphenylphosphonium bromide (1 to 128 mug/ml), acriflavine (0.25 to 32 mug/ml), or berberine (3.125 to 400 mug/ml). Bacterial growth was monitored by turbidimetry (at 600 nm). The growth index, calculated by dividing the OD600 of the culture in the presence of drug by the OD600 of the culture in the absence of drug, is shown. Multidrug resistance is mediated by an energy-dependent efflux. : To investigate the mechanism by which NorMI confers multidrug resistance, we compared the norfloxacin accumulation in E. coli AG100A cells harboring the pUC19-NorMI construct with that observed in E. coli AG100A containing pUC19 alone. We also measured the effect of CCCP, a proton motive force uncoupler, on the accumulation of norfloxacin inside the cells. Norfloxacin accumulation was assayed by the method of Mortimer and Piddock with some modifications. Briefly, bacteria were grown in Luria-Bertani broth at 37C to the mid-log phase of growth (optical density at 600 nm [OD600] = 0.6), harvested by centrifugation, washed in 50 mM sodium phosphate buffer (pH 7.0), and resuspended in the same buffer to an OD600 of 6.0 (corresponding to 10 mg [wet weight] per ml). After incubation for 20 min at 37C, norfloxacin (final concentration, 100 muM) was added to the bacterial suspension. Fifteen minutes after the addition of norfloxacin, the efflux pump inhibitor CCCP was added to the assay mixture at a final concentration of 100 muM. Samples (1 ml each) were removed at different times after the addition of norfloxacin, centrifuged at 7,000 x g for 30 s at 4C, washed once with 1 ml of the above-described ice-cold phosphate buffer, and resuspended in 1 ml of 100 mM glycine-HCl (pH 3.0). The suspension was shaken for 20 h at room temperature and then centrifuged at 7,000 x g for 5 min. The fluorescence of the supernatant was measured in a Jasco FP-770 spectrofluorimeter at excitation and emission wavelengths of 277 and 448 nm, respectively. The concentration of norfloxacin in the supernatant was calculated by comparison with the fluorescence of norfloxacin standards (0.161 to 5 nmol/ml) in 100 mM glycine-HCl (pH 3.0). As shown in Fig. , at 15 min after addition of norfloxacin, E. coli AG100A cells expressing NorMI accumulated about fourfold less norfloxacin than the control cells. Addition of the protonophore CCCP induced an increase of this accumulation, which reached rapidly a level similar to that observed in control cells. These results indicated that NorMI mediates an active efflux process driven by an electrochemical potential of H+. By comparison with most known bacterial multidrug efflux proteins, it is very likely that NorMI is a drug/ion antiporter. In conclusion, NorMI is the first described protein of B. melitensis that is able to mediate drug resistance by an active efflux mechanism. The existence of such a multidrug transporter indicates that, even if drug resistance is not considered a major problem in treating brucellosis, it is necessary to control the sensitivity patterns of Brucella spp. to ensure appropriate treatment and prevent increasing development of antibiotic resistance in these bacteria. FIG. 2. | Accumulation of norfloxacin in E. coli Accumulation of norfloxacin in E. coli cells expressing or not expressing the NorMI protein of B. melitensis. E. coli AG100A harboring a vector expressing NorMI (pUC19-NorMI) or harboring pUC19 alone was grown in Luria-Bertani broth. Norfloxacin was added to the cell suspensions at a final concentration of 100 muM. After 15 min, CCCP was added to the suspensions at a final concentration of 100 muM. Samples were removed at the indicated times, and the concentration of norfloxacin extracted from the cells was quantified by spectrofluorimetry. Backmatter: PMID- 12183229 TI - The LisRK Signal Transduction System Determines the Sensitivity of Listeria monocytogenes to Nisin and Cephalosporins AB - The Listeria monocytogenes two-component signal transduction system, LisRK, initially identified in strain LO28, plays a significant role in the virulence potential of this important food-borne pathogen. Here, it is shown that, in addition to its major contribution in responding to ethanol, pH, and hydrogen peroxide stresses, LisRK is involved in the ability of the cell to tolerate important antimicrobials used in food and in medicine, e.g., the lantibiotic nisin and the cephalosporin family of antibiotics. A DeltalisK mutant (lacking the LisK histidine kinase sensor component) displays significantly enhanced resistance to the lantibiotic nisin, a greatly enhanced sensitivity to the cephalosporins, and a large reduction in the expression of three genes thought to encode a penicillin-binding protein, another histidine kinase (other than LisK), and a protein of unknown function. Confirmation of the role of LisRK was obtained when the response regulator, LisR, was overexpressed using both constitutive and inducible (nisin-controlled expression) systems. Under these conditions we observed a reversion of the DeltalisK mutant to wild-type growth kinetics in the presence of nisin. It was also found that overexpression of LisR complemented the reduced expression of two of the aforementioned genes. These results demonstrate the important role of LisRK in the response of L. monocytogenes to a number of antimicrobial agents. Keywords: Introduction : The gram-positive pathogen Listeria monocytogenes imposes a significant burden in terms of both human and economic costs. Listeria was responsible for 71% of all recalls of food products due to bacterial contamination in the United States between 1993 and 1998 and, more importantly, is the cause of almost 30% of all deaths caused by food-borne pathogens in the United States every year . As a consequence, developing methodologies to control the survival and growth of Listeria in foods and during infection is a significant research goal. In this respect it is vital that a more complete understanding of how L. monocytogenes responds to the presence of those antimicrobial agents currently used to control this pathogen, both ex vivo (e.g., food grade inhibitors such as bacteriocins in foods) and in vivo (e.g., antibiotics), is developed. This information will be crucial in reducing the human and economic costs associated with Listeria and listeriosis. Our understanding of the mode of action of nisin, the only lantibiotic approved by the U.S. Food and Drug Administration for use as a food grade inhibitor, is increasing . It is now known that nisin functions, at least in part, by the formation of pores in the bacterial cell membrane, with the interaction being largely dependent on the type of lipids present and, most importantly, the charge carried by those lipids . Pore formation is facilitated by the binding of nisin to lipid II, a membrane-bound peptidoglycan precursor, which is thought to function as a docking molecule rather than as a receptor. In addition, the binding of nisin to lipid II is responsible for a secondary mode of action, i.e., inhibition of cell wall synthesis . Thus nisin inhibits bacterial growth by a combination of these two mechanisms. Despite the extent of this knowledge the role of host genes in determining nisin sensitivity and resistance is less clear. In vivo, i.e., in infection by Listeria, ampicillin, alone or in combination with gentamicin, remains the treatment of choice . However, the high level of innate resistance to cephalosporin antibiotics that Listeria possesses may be especially significant as members of this family of drugs are used most frequently for sepsis due to unknown causes. While cephalosporins were found to be efficient inhibitors of penicillin-binding protein 1 (PBP1), -2, and -4 in L. monocytogenes, which are completely blocked at concentrations well below the MIC, the innate resistance to cephalosporins is thought to be due to their lack of affinity for PBP3, the primary lethal target for beta-lactams in the species . Here we report that LisRK, an L. monocytogenes two-component signal transduction system which we previously identified and which has been found to play a role in acid, ethanol, and oxidative stress and in murine virulence , also plays a major role in the retardation of the growth of Listeria in the presence of nisin and in the innate cephalosporin resistance of this pathogen. In addition, for the first time, genes regulated by LisRK have been identified. It was found that overexpression of the gene encoding the response regulator component of the system, lisR, could complement a number of the phenotypic consequences of mutating lisK (the gene encoding the histidine kinase component) as well as restoring expression of two of the three regulated genes. MATERIALS AND METHODS : Strains, plasmids and media. | L. monocytogenes LO28 (serotype 1/2c) is a clinical isolate obtained from P. Cossart, Institut Pasteur, Paris, France. LO28DeltalisK is a mutant from which a portion of the histidine kinase-encoding gene, lisK, has been deleted by splicing by overlap extension (SOEing) PCR . The culture medium used was tryptone soy agar or tryptone soy broth (Oxoid, Basingstoke, Hampshire, England) supplemented with 0.6% yeast extract (Difco) (TSA-YE or TSB-YE) or brain heart infusion (BHI) agar or broth (Oxoid). Plasmid pKSV7, used for SOEing PCR, was a kind gift from Kathryn Boor, Cornell University, Ithaca, N.Y. Plasmids pNZ8048 and pNZ9530 were gifts from Michiel Kleerebezem, NIZO, Ede, The Netherlands. Plasmid pNZ44 was a gift from Stephen McGrath, University College Cork, Cork, Ireland. Growth in the presence of nisin. | The rates of growth of L. monocytogenes LO28 and LO28DeltalisK in the presence of different levels of nisin (2% inoculum in TSB-YE containing 50, 100, 150, 200, or 300 mug of nisin powder [Sigma, St. Louis, Mo.]/ml) were compared by monitoring optical density at 600 nm (OD600) with a Spectra Max 340 spectrophotometer (Molecular Devices, Sunnyvale, Calif.) over a 20-h period. Antibiotic assays. | Assays to determine the sensitivities of LO28 and LO28DeltalisK to a wide range of antibiotics were carried out by agar diffusion. Overnight cultures were diluted to 106 CFU/ml and swabbed onto TSA-YE. Commercially purchased disks (6 mm in diameter; Oxoid) containing 30 mug (unless otherwise stated) of the antibiotics to be studied were then placed on the surfaces of agar plates. Following overnight incubation of the plates at 37C, the diameters of the zones of bacterial growth inhibition surrounding the filter disks were measured. The relative susceptibilities of different strains to the various antibiotics tested were correlated with the sizes of the zones of inhibition, with increased zone size reflecting increased susceptibility. Initially the antibiotics assayed were cefotaxime, cefuroxime, vancomycin, erythromycin, kanamycin, fosfomycin (50 mug), minocycline, polymyxin B (300 mug), streptomycin (25 mug), fusidic acid (10 mug), oxytetracycline, penicillin (10 mug), clindamycin (10 mug), spectinomycin (25 mug), ampicillin (25 mug), novobiocin, rifampin, nalidixic acid, gentamicin, colistin sulfate (25 mug), chloramphenicol (CAM), and tetracycline. In addition to this assay, designed to compare antibiotic susceptibilities in general, further studies involved disks containing antibiotics of the cephalosporin family. These were cefuroxime, ceftazidime, cefaclor, cephalothin, cefoxitin, ceftriaxone, cefotetan, cefoperazone, cephradine, cephalexin, and cefotaxime (all 30 mug). Implementation of the nisin-controlled expression (NICE) system in L. monocytogenes LO28DeltalisK. | A strategy to replace the hemolysin gene, hly, on the L. monocytogenes chromosome with the nisRK genes was devised. Primers were designed to amplify the chromosomal regions flanking the hly gene (primers HSOEA, -B, -C, and -D; Table ). The resultant A-B and C-D fragments were spliced by overlap extension PCR using primers HSOEA and -D to create a single A-D fragment, representing the region surrounding hly but with the gene precisely removed. This fragment was subsequently cloned in temperature-sensitive plasmid pKSV7 . Since primers B and C have built-in restriction sites, we were able to clone the nisRK operon (amplified, by using primers nisRF and nisKB, from pNZ9530 ) into the location formerly occupied by hly. This plasmid construct, pCPL-53, was electroporated into L. monocytogenes LO28DeltalisK, and transformants were selected on BHI agar with 10 mug of CAM/ml (BHI/CAM). Chromosomal integration of the plasmid at 41C was selected by serial passage of a transformant in prewarmed BHI/CAM broth and streaking onto prewarmed BHI/CAM agar. Plasmid excision was accomplished by continuous passage in BHI at 30C, and clones in which nisRK had replaced hly on the chromosome were identified by plating them onto blood agar plates at 30C. Replica plating of nonhemolytic colonies onto BHI and BHI/CAM at 30C identified cured derivatives. PCR and sequencing analysis of one such strain confirmed that the nisRK operon had replaced the hly gene. This strain was designated LO28DeltalisK-NICE. Overexpression of lisR. | PCR primers NZlisRF (containing the lisR start codon) and NZlisRB (containing the lisR stop codon) with incorporated NcoI and XbaI sites, respectively, were used to amplify lisR from LO28. The resultant PCR product was digested with NcoI and XbaI restriction enzymes (Roche) and cloned into similarly digested pNZ8048, resulting in the generation of a translational fusion between the nisin-inducible nisA promoter on pNZ8048 and the lisR gene. This event was confirmed by sequence analysis. This strain was designated LO28DeltalisK-NICE(pNZ8048-lisR). pNZ8048 was also introduced into the LO28DeltalisK-NICE background to create LO28DeltalisK-NICE(pNZ8048). As an alternative to inducible overexpression, lisR (amplified by primer pair NZlisRF and NZlisRB) was also cloned into plasmid pNZ44, a derivative of pNZ8048 in which the Pnis promoter was replaced by a constitutive P44 promoter , resulting in the creation of LO28DeltalisK(pNZ44-lisR). LO28DeltalisK(pNZ44) was also created to serve as a control for subsequent experiments. RT-PCR. | RNA isolation and reverse transcription-PCR (RT-PCR) were carried out as described previously . RNA was isolated from overnight cultures (constitutive overexpression system) or following nisin induction (inducible overexpression system). For induction with nisin cultures were grown to an OD600 of 0.2 and preinduced with 4.5 mug of nisin powder/ml for 1 h, followed by induction with 45 mug of nisin powder/ml (a concentration at which there is no difference in the nisin sensitivities of the two strains) for 30 min and then isolation of RNA. In all cases cDNA was amplified by PCR with specific primers and samples were taken at regular intervals and run on agarose gels. Primers for the 16S rRNA of L. monocytogenes LO28 were used as controls . TABLE 1 | PCR primers used in this study. RESULTS AND DISCUSSION : Nisin resistance of the DeltalisK mutant. | Because deletion of the L. monocytogenes histidine kinase-encoding gene, lisK, results in an altered response to environmental parameters , we examined the susceptibility of L. monocytogenes to the lantibiotic nisin. It was found that the levels of nisin required to inhibit growth of wild-type LO28 are high relative to those required to inhibit the growth of a number of other gram-positive bacteria . This is a feature of this strain which we have previously observed in relation to other bacteriocins such as lacticin 3147 and enterocin A (our unpublished data). Growth curves were carried out to determine whether deletion of lisK affected the response to nisin. In the absence of nisin the growth rates of the parent and mutant strains were identical, but in the presence of increasing levels of nisin differences became apparent, primarily manifested as a significantly longer lag phase in wild-type LO28 . This is most obvious when one observes that the lag period for LO28 in the presence of 300 mug of nisin/ml is 6 h longer than that for LO28DeltalisK . There is no difference in the growth rates once the lag phase has been exited (e.g., both are 0.046 U/h during logarithmic growth in the presence of 200 mug of nisin/ml). Therefore, the LisK mutation results in a culture that is able to rapidly initiate growth in the presence of nisin, whereas the parent apparently needs a significant period in which to adapt to the presence of the inhibitor. We speculate that, during prior growth in the absence of nisin, the LisRK system plays a role in creating a particular cell envelope composition that renders the cell more susceptible to nisin (and also alters its response to ethanol, low pH, and hydrogen peroxide [, ]). In the absence of LisK, the cell must presumably fail to sense an as yet unknown environmental parameter, resulting in an altered envelope composition that is manifested as rapidly initiated growth in the nisin assay. Antibiotic disk assays. | Antibiotic disks were used to determine if LO28 and LO28DeltalisK differed in their responses to other antimicrobial agents. An initial extensive study using a wide range of antibiotics (see Materials and Methods) revealed that only cefotaxime and cefuroxime differentiated significantly between the two strains though a slightly enhanced (though not statistically significant) sensitivity to ampicillin and penicillin was also observed. It was found that for the parent strain, LO28, the diameters of the zones surrounding the cefotaxime and cefuroxime disks were very small (16 and 15.4 mm, respectively). In contrast, for LO28DeltalisK the diameters extended to 26 and 26.4 mm, representing significant 63 and 71% increases in zone size for cefotaxime and cefuroxime, respectively. As a consequence of the dramatic nature of these findings, further antibiotic disk assays were performed using other antibiotics of the cephalosporin family . While, as expected, it was observed that the LO28 background was, in general, more sensitive to narrow-spectrum cephalosporins (e.g., cephalothin, cefaclor, and cephradine), the most significant discovery was the greatly enhanced sensitivity of the LO28DeltalisK mutant in all cases . Our observations are especially significant when one considers that, despite L. monocytogenes susceptibility to a wide range of antibiotics, it is resistant to the cephalosporins , a large and expanding family of drugs based on cephalosporin C which are frequently the initial choice for hospital treatment of bacterial infection resulting in fever due to unidentified organisms. LisRK-regulated genes. | While the physiological changes present in nisin-resistant mutants have been examined on a number of occasions , until the recent study by Gravesen et al. it was not known which genes might be involved. In a high percentage of spontaneous nisin-resistant L. monocytogenes 412 mutants studied by this group increases in the levels of cDNA corresponding to three genes were uncovered by restriction fragment differential-display PCR. Of the putative proteins encoded by these genes one showed high homology to the glycosyltransferase domains of PBPs, another was a histidine kinase (though not LisK), and the third was a protein of unknown function. These findings were especially relevant as a large proportion of these spontaneous mutants also demonstrated enhanced resistance to cefuroxime , leading us to speculate that LisRK may regulate one or more of the same genes. To determine if this was the case, RT-PCR analysis was carried out to investigate whether transcription of these genes varies between LO28 and LO28DeltalisK. Significantly, it was found in all cases that the relative level of expression of these three genes was greatly reduced or eliminated in LO28DeltalisK . The gene encoding the putative PBP in strain 412 corresponds to the open reading frame designated lmo2229 in genome-sequenced strain L. monocytogenes EGD-e . This putative PBP shows highest homology to PBP2a of Streptococcus pneumoniae and PBP1a of Bacillus subtilis , both of which are high-molecular-weight PBPs possessing both glycosyltransferase and transpeptidase domains. The gene encoding the histidine kinase corresponds to lmo1021 and shows homology with the yvqE gene of B. subtilis and llkinD of Lactococcus lactis (genes encoding members of the NarQ/NarX subfamily). In each case the gene encoding the histidine kinase is located between a response regulator gene homolog (lmo1022, yvqC, or llrrD) and a gene encoding a protein of unknown function (lmo1020, yvqF, or tcdsorf1). While mutation of llkinD has not been achieved, it was found that the most significant trait associated with an llrrD mutant was an increased osmosensitivity. However, unlike lisK mutants, this mutant did not exhibit enhanced sensitivity to oxidative stress . The third gene fragment, originally designated fragment C, corresponds to lmo2487 and is homologous to B. subtilis yvlB, the predicted protein of which has not been assigned a function. While we have yet to ascertain what role, if any, these three genes play in the phenotypes associated with the lisK mutation, the variations in the levels of transcript and the observation that at least one other histidine kinase plays a role suggest that LisRK is involved in a complex regulatory pathway. The apparent inconsistency between our findings and those of Gravesen et al. with regard to the relative increase or decrease in transcription of these genes may reflect the growth phases during which cells were studied, i.e., late exponential phase (OD600 = 0.6) and late stationary phase (overnight growth), respectively. Growth phase-dependent variations have previously been reported with respect to the acid resistance of the LO28DeltalisK mutant. We can, however, definitively state that alterations in the levels of these proteins are associated with nisin resistance and cephalosporin sensitivity. In addition to this now-established link between nisin resistance and cephalosporin sensitivity it has also been found that a number of spontaneous L. innocua 4202 mutants resistant to bacteriocin AS-48 displayed enhanced sensitivity to cephradine while lacticin 3147-resistant L. monocytogenes LO28 mutants were more sensitive to cephalexin, cefaclor, and cephradine . The recurring link between bacteriocin resistance and cephalosporin sensitivity is thus significant and merits further study. Intriguing are suggestions that the S. pneumoniae CiaRH two-component system, which is closely related to LisRK and which plays a role in cefotaxime resistance as well as in other phenotypes attributable to the cell envelope , also controls levels of undecaprenol (bactoprenol), a component of lipid II (N-acetylglucosamine-beta-1,4-MurNAc-pentapeptide-pyrophosphoryl-undecaprenol) as well as monitoring the integrity of the cell wall in general . This is especially relevant as lipid II serves as the docking molecule for nisin binding prior to pore formation and is the target for its secondary mode of action, inhibition of cell wall synthesis. Though the nisin resistance of CiaRH mutants has not been reported, it is tempting to suggest, on the basis of the involvement of lipid II, that their resistance would be modified and thus that further examination of this pathway in the appropriate mutants may reveal the mechanism responsible for the correlation between nisin resistance and cephalosporin sensitivity. Implementation of the NICE system in L. monocytogenes LO28DeltalisK. | To confirm that the observed phenotypic and transcriptional changes associated with the deletion in lisK were linked to its interaction with the cognate LisR regulator (and not due to cross talk or a secondary undetected mutation in LO28DeltalisK), we designed a strategy to overexpress LisR in the LO28DeltalisK background. The rationale for this approach stems from the observation that in certain circumstances overexpression of a response regulator alone may mimic activation thereof . This phenomenon may be explained by the observation that a number of response regulators are capable of binding their target DNA when nonphosphorylated, though less efficiently than when phosphorylated . It may also be possible that the high stoichiometry of LisR allows some level of phosphorylation as a result of either cross talk or the possible contribution of another histidine kinase that specifically interacts with LisR (the existence of which has not been ruled out). Therefore, a system to allow overexpression of LisR was necessary to investigate whether excess LisR could complement the absence of LisK. In addition to the need for a very strong promoter it was desirable that the system also be inducible, in the hope that it might ultimately facilitate the identification of additional promoters controlled by LisRK by using a strategy analogous to that used by Soncini et al. to identify PhoPQ-regulated genes in Salmonella enterica serovar Typhimurium . Due to the paucity of controlled expression systems in Listeria and the extent to which it can overexpress proteins, the lactococcal NICE system was employed . The NICE system has the advantage of being extremely responsive to an external stimulus (nisin) and has been successfully used in a number of other gram-positive genera, including Leuconostoc and Lactobacillus and Streptococcus, Enterococcus, and Bacillus . It depends on the presence of a two-component signal transduction system (NisRK) which senses nisin in the external environment and stimulates transcription from the Pnis promoter. Although nisin is both the induction factor for the NICE system and one of the inhibitors under investigation in this study, because of the relative differences in concentration required for induction and for inhibition, we were able to utilize nisin as both inducer and challenge to determine whether overexpression of LisR can overcome the DeltalisK lesion. Original applications of the NICE system used two plasmids, one with a nisA promoter and the relevant gene to be overexpressed and the second on which the nisRK regulatory genes were located in trans . More-recent innovations involve either placing the nisRK genes on the host chromosome, allowing the subsequent use of a one-plasmid system , or placing both nisRK and the nisA promoter on a single plasmid . We used the former approach and introduced the nisRK genes onto the listerial chromosome, replacing the hemolysin (hly) gene and placing nisRK under the control of the hly promoter. This replacement did not impact on the level of resistance of the strain. This particular gene replacement strategy was chosen because transcription of hly in strain LO28 in vitro is constitutively high but can be further increased (by the addition of charcoal to the culture medium) or decreased (by a reduction in growth temperature from 37 to 20C) if required . Second, this has the advantage of creating a much less virulent host with which to perform physiological assays. The strain resulting from this procedure, the construction of which is described in Materials and Methods, was designated LO28DeltalisK-NICE . To overexpress LisR, the corresponding gene was cloned into pNZ8048 behind the Pnis promoter. The resulting plasmid, designated pNZ8048-lisR, was introduced into LO28DeltalisK-NICE. Plasmid pNZ8048 lacking an insert was also introduced into the same background for use as a negative control in subsequent experiments. To confirm that nisin induced overexpression of a lisR transcript, an RT-PCR comparing the relative quantities of lisR mRNA in the two backgrounds was performed . After 15 cycles it was apparent that the levels of lisR mRNA being produced, following induction by nisin (45 mug/ml), were much greater when pNZ8048-lisR, rather than pNZ8048, was present. Complementation of the DeltalisK phenotype by overexpression of lisR. | To determine whether overexpressing lisR reversed the enhanced nisin resistance displayed by LO28DeltalisK, we compared the growth of LO28DeltalisK-NICE containing pNZ8048-lisR to that when it contained pNZ8048. The level of nisin required to affect the growth of the strains (greater than 100 mug/ml) was in excess of that required to induce the NICE system maximally (45 mug/ml), thus allowing the induction of lisR and the examination of the growth kinetics under nisin inhibition. The results confirmed that induction of lisR results in a reversion to nisin sensitivity, overcoming the DeltalisK mutation . As in the parental strain, this sensitivity manifests itself as an increased lag period relative to that of LO28DeltalisK-NICE(pNZ8048), which is 6 h in the presence of 300 mug of nisin/ml . Overexpression of LisR by nisin was also found to result in a reversion to wild-type levels of ethanol sensitivity (data not shown) though a reversion to cephalosporin resistance was not apparent. For fear that the use of nisin was interfering with the ability of the NICE system to complement cephalosporin sensitivity, an alternative overexpression vector, pNZ44, was used. This vector is a derivative of pNZ8048 in which a constitutive P44 promoter replaces the inducible Pnis . It was found that LO28DeltalisK containing pNZ44-lisR exhibited a reversion to cephalosporin resistance, though the degree of reversion varied depending on the antibiotic used . This system was also used to determine if the levels of the three genes associated with nisin resistance increased in response to LisR overexpression . It was found that, while increases in the levels of transcript corresponding to lmo2487 (gene C) and, to a much lesser extent, lmo2229 (PBP) were observed, no change in the level of lmo1021 (histidine kinase) was detected. The variation in the extent to which complementation of the cephalosporin sensitivity phenotype occurred may reflect either the involvement of an alternate response regulator that interacts with LisK or, more likely, variations in the affinity of LisR for the three promoters, analogous to the differential expression of genes within the regulon of the closely related CsrR response regulator in Streptococcus pyogenes . As well as examining the regulation and function of the PBPs, histidine kinase, and gene C we are currently attempting to identify both additional genes under the control of LisR (using the inducible overexpression strategy alluded to previously) and the stimulus which triggers LisK, so that we can gain a more precise understanding of the molecular mechanisms underlying this phenomenon. In conclusion, we have shown that the LisRK two-component signal transduction system, in addition to playing a role in the response of bacteria to acid, ethanol, hydrogen peroxide, and in vivo stresses, has a major role in the growth of this potentially lethal food pathogen in the presence of lantibiotic bacteriocin and in determining the sensitivity to antibiotics within the cephalosporin family. A link between nisin resistance and cephalosporin sensitivity, previously observed by Gravesen et al. , in spontaneous mutants has been confirmed, three genes associated with this phenotype have been shown to controlled through LisK, and it has been shown that complete or partial complementation of the phenotypes displayed by LO28DeltalisK can be reversed by overexpressing lisR. Finally, in the process of determining the role of LisRK in nisin resistance, we have also confirmed that the NICE system functions in L. monocytogenes. The importance of this finding stems from the lack of inducible gene expression systems for Listeria and the adaptability of the NICE system, which permits precise regulation of expression levels and which can facilitate the production of very high levels of recombinant proteins. The relatively high nisin resistance of strain LO28 makes it especially suitable for use with this system to facilitate the identification of the LisRK regulon and potentially genes controlled by other transcriptional regulators in Listeria. It should also prove a useful tool for others interested in expressing foreign proteins in L. monocytogenes. FIG. 1. | Growth of LO28DeltalisK in the presence of antimicrobial agents. Growth of LO28DeltalisK in the presence of antimicrobial agents. Shown is growth of LO28 (circles) and LO28DeltalisK (squares) in TSB-YE with 0 (A), 50 (B), 100 (C), 200 (D), and 300 (E) mug of nisin powder/ml. Growth was determined by using a Spectra Max 340 spectrophotometer (Molecular Devices) over a 20-h period. Error bars, standard deviations from the means of quadruplicate experiments. FIG. 2. | (A and B) RT-PCRs to compare levels of transcripts for genes whose products show homology to histidine kinase (HK), a protein of unknown function (gene C), and a PBP in strains LO28 (L) and LO28DeltalisK (DeltaK) (A) and LO28DeltalisK(pNZ44lisR) (+R) and LO28DeltalisK(pNZ44) (-R) (B). (A and B) RT-PCRs to compare levels of transcripts for genes whose products show homology to histidine kinase (HK), a protein of unknown function (gene C), and a PBP in strains LO28 (L) and LO28DeltalisK (DeltaK) (A) and LO28DeltalisK(pNZ44lisR) (+R) and LO28DeltalisK(pNZ44) (-R) (B). (C) Confirmation of the overexpression of lisR by RT-PCR using lisR-specific PCR primers to amplify cDNA templates of equal concentrations generated from LO28DeltalisKNICE(pNZ8048) or LO28DeltalisKNICE(pNZ8048-lisR) RNA following nisin induction. In all cases control PCR primers were used to ensure the complete removal of DNA from RNA preparations prior to reverse transcription and to ensure that levels of cDNA for samples that were to be compared were equal. FIG. 3. | Complementation of the nisin resistance phenotype of LO28DeltalisK by using the NICE system. Complementation of the nisin resistance phenotype of LO28DeltalisK by using the NICE system. Growth of LO28DeltalisKNICE(pNZ8048) (squares) and LO28DeltalisKNICE(pNZ8048-lisR) (circles) in TSB-YE with 0 (A), 50 (B), 100 (C), 200 (D), and 300 (E) mug of nisin powder/ml. Growth was determined by using a Spectra Max 340 spectrophotometer (Molecular Devices) over a 20-h period. Error bars, standard deviations from the means of quadruplicate experiments. TABLE 2 | Cephalosporin resistance of LO28 and LO28DeltalisK Backmatter: PMID- 12183276 TI - Prevalence of Protease and Reverse Transcriptase Drug Resistance Mutations over Time in Drug-Naive Human Immunodeficiency Virus Type 1-Positive Individuals in Rio de Janeiro, Brazil AB - The prevalence of mutations that confer resistance to protease inhibitors and to nucleoside and nonnucleoside reverse transcriptase inhibitors in 49 blood samples from drug-naive human immunodeficiency virus type 1-infected blood donors living in Rio de Janeiro state, Brazil, in 1998 was evaluated genotypically and phenotypically. Keywords: Introduction : The new generation of drugs targeting the reverse transcriptase (RT) and protease (PR) genes of the pol region of human immunodeficiency virus type 1 (HIV-1), such as nucleoside reverse transcriptase inhibitors (NRTIs), nonnucleoside reverse transcriptase inhibitors (NNRTIs), and protease inhibitors (PIs), has revolutionized treatment of infected individuals . However, not all patients respond to highly active antiretroviral treatment (HAART), and many develop drug resistance, one of the most serious obstacles to sustained suppression of HIV-1 The emergence of amino acid mutations associated with resistance to RTIs and PIs has been extensively characterized , and these substitutions can be classified into primary and secondary mutations. Primary mutations lead to a severalfold decrease in sensitivity to one or more antiretroviral (ARV) drugs . Secondary mutations may not result in a significant decrease in drug sensitivity but are associated with restoration of the original viral fitness in the presence of existing inhibitors . The transmission and dissemination of drug-resistant strains have major public health implications, including disrupting the efficiency of established ARV treatment for HIV-1-infected patients and of prophylaxis against both accidental exposures and perinatal HIV-1 transmission. Prevalences of primary resistance mutations for any drug observed among recent seroconverters range from 5 to 11% in Switzerland (; S. Yerly, E. Race, S. Vora, P. Rizzardi, J. P. Chave, M. Flepp, P. Vernazza, A. Telenti, M. Battegay, A. L. Veuthey, B. Hirschel, L. Perrin, and the Swiss HIV Cohort Study, Program Abstr. 8th Conf. Retroviruses Opportunistic Infections, abstr. 754, 2001), 4 to 17% in France (, ; M. L. Chaix, M. Harzic, B. Masquelier, I. Pellegrin, L. Meyer, D. Costagliola, C. Rouzioux, and F. Brun-Vezinet, Program Abstr. 8th Conf. Retroviruses Opportunistic Infections, abstr. 755, 2001), 13% in Germany , 14% in the United Kingdom , 2 to 26% in North America [, , , , ; S. J. Little, S. Holte, and J. P. Routy, abstract from 5th Int. Workshop HIV Drug Resist. Treatment Strategies, 4 to 8 June 2001, Scottsdale, Ariz. Antivir. Ther. 6(Suppl. 1):21, 2001], and 23 to 26% in Spain . Information on PI- and RTI-associated mutations in HIV-1-infected drug-naive individuals in Brazil is limited to specimens collected before 1996, when free access to HAART was established for AIDS patients by the Brazilian Ministry of Health . These specimens have shown a low prevalence of mutations related to NRTIs, NNRTIs, and PIs. In this paper, we describe the genetic diversity of HIV-1 PR and RT sequences and the prevalence of mutations linked to PI and RTI resistance in treatment-naive HIV-1-infected blood donors living in the state of Rio de Janeiro, Brazil, in 1998. Of the 18,667 blood units collected from January through December 1998 at blood banks distributed throughout the state of Rio de Janeiro, 56 were rejected based on their HIV-1 positivity by enzyme immunoassay (Prism HIV-1/HIV-2 [Abbott, North Chicago, Ill.] and Vironostika HIV Uni-Form plus O [Organon Teknika BV, Boxtel, The Netherlands]). After confirmation by Western blot analysis, the viral RNA was isolated from plasma, and PR and RT genes were RT-PCR amplified and sequenced as previously described . This study was approved by the Brazilian Ethics Committee (CONEP) as an anonymous unlinked study. Phylogenetic analysis was performed using the neighbor-joining method as previously described . Of 44 samples positive for both PR and RT sequences, 40 had concordant subtype assignments in these regions, including 39 (89%) subtype-B specimens and 1 (2%) subtype-C specimen . The remaining four samples revealed discordant subtypes for the PR and RT regions, which may represent recombinant viruses. Finally, of the five specimens with HIV-1 subtype assignment for only one viral region, two PR and two RT sequences were of subtype B and one RT sequence was of subtype F. These results indicate that subtype distributions in our sampling reflect the proportions previously observed in Brazil . TABLE 1 | Presence of PR and RT inhibitor-associated mutations in subtype-B and non-B Brazilian HIV-1 strains collected in 1998 To examine the frequency of primary (D30N, M46I, G48V, I50V, V82A, -F, or T [V82A/F/T], I84V, and L90 M) and secondary (L10I/R/V, K20M/R, L24I, V32I, L33F, M36I, M46I/L, I47V, I54L/M/V, L63P, A71V/T, G73S, V77I, V82I, and N88D) PI-associated substitutions within the PR region, we analyzed these 20 amino acid sites, which are associated in vivo with resistance in HIV-1 subtype B to PIs approved by the U.S. Food and Drug Administration (FDA) . This analysis revealed a lack of primary substitutions in 46 PR sequences. However, secondary mutations were found in 39 (85%) of the 46 viral sequences at six positions. L63P was the most common secondary substitution (54%; 25 of 46), followed by V77I (26%; 12 of 46), M36I (20%; 9 of 46), A71V/T (15%; 7 of 46), V82I (6%; 3 of 46), and L10I/V (4%; 2 of 46). Single amino acid mutations were identified in 23 (50%) of the 46 PR sequences, whereas 34% of sequences harbored dual (30%; 14 of 46), triple (2%; 1 of 46) and quadruple (2%) PI-associated substitutions . To search for changes over time in the PI-associated mutations among the subtype-B infections in Brazil, PI-associated substitutions found in 41 PR sequences collected in 1998 were compared with those found in 38 sequences collected from drug-naive individuals in 1987 to 1994 . There were no primary PI-associated mutations in PR sequences in either collection, and PI secondary mutational patterns were similar in the two sets. However, the frequency of isolates harboring PI-associated secondary resistance mutations was higher (0.025 < P < 0.05 by a one-tailed chi2 test) in the 1998 data set (85%) than in the 1987-to-1994 collection (66%). This finding possibly reflects a natural evolution of the PR gene over time rather than selection of HIV-1 strains harboring PI-associated substitutions resulting from PI treatment in Brazil. FIG. 1. | Percentages of PR subtype-B sequences with PI-associated secondary amino acid mutations at two different time points. Percentages of PR subtype-B sequences with PI-associated secondary amino acid mutations at two different time points. The number of sequences with the indicated mutation is given inside or directly above each bar. Statistical significances of the differences (by a one-tailed Fisher or chi2 test) are given above the bars. Total*, total numbers of sequences with at least one mutation. The results presented in Table indicated that none of the 47 RT sequences from the 1998 collection harbored primary mutations associated with resistance to FDA-approved NRTIs and only one had the secondary mutation V118I (2%; 1 of 47). Analysis of FDA-approved NNRTI-associated mutations indicated that neither primary nor secondary substitutions were present in 47 RT sequences analyzed. Our findings are in agreement with the low proportion of drug-naive subjects infected with viruses harboring primary drug-resistant mutations in Venezuela and Argentina but in contrast with reports of the presence of HIV-1 strains carrying primary mutations in drug-naive individuals in North America [, , , , ; Little et al., Antivir. Ther. 6(Suppl. 1):21, 2001] and Europe (, , , , ; Chaix et al., Program Abstr. 8th Conf. Retroviruses Opportunistic Infections, abstr. 755, 2001; Yerly et al., Program Abstr. 8th Conf. Retroviruses Opportunistic Infections, abstr. 754, 2001) who are infected with subtype-B viruses. A few factors may contribute to these differences, such as the length of time under drug selective pressures and the time period between primary infection and sampling. In addition to specific mutations known to contribute to RTI resistance, a few unusual changes were observed in RT sequences at both the NRTI resistance-associated sites (E44G [6.4%], K70I [2.2%], V75L [2.2%], Q151K [2.2%], T215P/S [8.5%], and K219P/R/T [8.5%]) and the NNRTI resistance-associated positions (A98S [4.2%] and V106/I/L [13%]) . Of note, recent data indicate that the T215S substitution, which was found in two of our RT sequences, is responsible for a more-rapid emergence of zidovudine (AZT) resistance in vitro, and it is speculated that this selection could also take place in vivo (J. G. Garcia-Lerma, S. Nidtha, K. Blumoff, H. Weinstock, and W. Heneine, Program Abstr. 5th Int. Workshop HIV Drug Resistance Treatment Strategies, abstr. 21 and 22, 2001). Sixteen isolates that represented distinct phylogenetic subtypes and a variety of genotypic PI- and RTI-associated mutation patterns were selected for in vitro evaluation of susceptibility to the PIs (saquinavir [SQV], ritonavir [RTV], indinavir [IDV], nelfinavir [NFV], and amprenavir [APV]), NRTIs (AZT, lamivudine [3TC], didanosine [ddI], zalcitibine [ddC], stavudine [d4T], and abacavir [ABV]), and NNRTIs (nevirapine [NVP], delavirdine [DLV], and efavirenz [EFV]) by using VIRCO's Antivirogram assay . This phenotypic analysis revealed moderate resistance to NFV in only one strain (98BRVM15), which harbored quadruple PI-associated secondary mutations: 36I 63P 71V 77I. Three strains (98BRVM15, 98BRVM16, and 98BRVM85) revealed moderate resistance (4.0- to 6.5-fold) to ddI. Finally, strain 98BRVM58 showed levels of resistance above the biological cutoff value for NVP (8.0-fold) . The levels of resistance above cutoff values for ddI and NVP observed in these samples could not be easily explained by the respective viral genotypes, which lacked primary resistance mutations. It is likely that mutations L283I, Y318F, and P236L (codons that were outside the RT fragment analyzed in this study), in combination with other, as yet unidentified NNRTI mutations, are responsible for moderate resistance. For example, the dual mutations I135T L283I were recently found to be associated with intermediate resistance to NVP and DLV . Our study showed that isolate 98BMVM58, which has shown resistance to NVP, had the I135T substitution. Also, it is possible that the presence of leucine instead of valine at codon 106 could contribute to the NVP resistance of this strain . Observed discrepancies between genotype and phenotype, especially with regard to intermediate resistance to NNRTIs, have been reported previously . TABLE 2 | Phenotypic and genotypic profiles of selected samples This study provides the most comprehensive evaluation to date of both the genetic diversity and the PR and RT drug-resistant patterns in treatment-naive HIV-1-infected individuals after the introduction of free access to HAART in Brazil in 1996. Our findings showing a lack of resistant strains circulating in drug-naive subjects do not support the need for genotyping before ARV therapy initiation. Nucleotide sequence accession numbers. : The nucleotide sequences of the HIV-1 PR and RT genes sequenced in this study have been assigned GenBank accession numbers to . Backmatter: PMID- 12183287 TI - Amino Acid Substitutions at Position 69 of the Reverse Transcriptase of Human Immunodeficiency Virus Type 1 Are Frequent in Zalcitabine-Naive Antiretroviral-Drug-Experienced Patients AB - Keywords: Winters and Merigan recently reported interesting results showing that several amino acid substitutions encoded by codon 69 of the human immunodeficiency virus type 1 reverse transcriptase (HIV-1 RT) gene affect susceptibility to nucleoside-nucleotide analog RT inhibitors (NRTIs) . While it is classically considered that the T69D/N mutation confers resistance to zalcitabine (dideoxycytidine [ddC]) , we have demonstrated, and Winters and Merigan have confirmed, that this mutation can be identified in ddC-naive individuals . On the basis of this observation, we have further analyzed the amino acid substitutions at position 69 in the RT sequences obtained routinely in our laboratory; amino acid insertions or deletions at this position have not been considered in this study. After exclusion of the patients tested after treatment interruption, 789 RT sequences were available for analysis: 34 from antiretroviral-naive patients and 755 from antiretroviral-experienced patients. Among the antiretroviral-experienced patients, 50 (6.6%) had received one antiretroviral regimen, 444 (58.8%) had received two or three antiretroviral regimens, and 261 (34.6%) had received four or more antiretroviral regimens; 3 (0.4%) had received one NRTI, 58 (7.7%) had received two NRTIs, 89 (11.8%) had received three NRTIs, 221 (29.3%) had received four NRTIs, 235 (31.1%) had received five NRTIs, 145 (19.2%) had received six NRTIs, and 4 (0.5%) had received seven NRTIs. Amino acid substitutions at position 69 were not found in sequences from the NRTI-naive patients but were found in those from 116 (15.4%) NRTI-experienced patients: 71 patient sequences (61.2%) had T69D mutations, 28 (24.1%) had T69N mutations, 7 (6.0%) had T69S mutations, 4 (3.4%) had T69A mutations, 3 (2.6%) had T69E mutations, and 1 each (0.9%) had T69G, T69I, and T69L mutations. The prevalence of amino acid substitutions at position 69 increased significantly with the number of drug regimens that a patient received (P = 0.0002), since a prevalence of 8.0% was observed in patients who had received one treatment regimen and a prevalence of 21.8% was observed in patients who had received four or more treatment regimens. A significant relationship (P = 0.01) between the prevalence of amino acid substitutions at position 69 and the number of NRTIs that a patient received was also observed. A multivariate logistic-regression analysis showed that the number of antiretroviral drug regimens that a patient received was the sole factor independently related to amino acid substitutions at position 69. However, when only the D and N substitutions were considered, a significant relationship (P = 0.03) between the T69D/N mutation and experience with didanosine (ddI) was observed (data not shown). There was no significant relationship between D/N (P = 0.26) or other amino acid substitutions (P = 0.13) at position 69 and experience with ddC. By univariate analysis, amino acid substitutions at position 69 were significantly associated with mutations at positions 44, 67, 70, 118, 210, 215, and 219, whereas a borderline significance was observed for the association with the L74V mutation . By multivariate analysis, however, only E44D/A (P = 0.02), D67N (P < 0.0001), and V118I (P = 0.0001) remained significantly associated with amino acid substitutions at position 69. By focusing on the D and N substitutions, a significant association was observed with the T69D/N mutation and the Q151M mutation (P = 0.004). These results indicate that amino acid substitutions (mainly T->D and T->N) are common in multi-NRTI-experienced patients, that the prevalence of these mutations increases with the number of NRTI combination regimens received, and that previous ddI rather than ddC experience could be a factor of selection of the T69D/N mutations in multi-NRTI-treated pa-tients. These conditions of selection are also supported by the association observed between amino acid substitutions at position 69 and the E44A/D, V118I, and Q151M mutations, which are also known to be associated with multiple NRTI experience, including experience with ddI . Amino acid substitutions were also closely associated with the D67N mutation, which is located in the same beta3-beta4 hairpin loop (positions 67 to 70) region of the RT; this association suggests that interactions exist between these mutations. Our results, in agreement with those reported by Winters and Merigan , indicate that amino acid substitutions at position 69 may be selected for by ddC-free antiretroviral regimens. These mutations could therefore be involved in resistance to NRTIs other than ddC. This is supported by the findings of Winters and Merigan that show that different substitutions at position 69 confer reduced susceptibility to zidovudine, ddI, and stavudine. The T69D/N mutation is currently taken into consideration for interpreting resistance to ddI (low-level resistance) and stavudine (potential low-level resistance) by using the Stanford HIV RT and protease sequence database algorithm . However, more effort is needed to accurately interpret HIV-1 genotypes that encode amino acid sub-stitutions at position 69. In particular, the effect of the different amino acid substitutions at this position and the interactions of these substitutions with the other NRTI resistance-associated mutations remain poorly documented. TABLE 1 : Association of the amino acid substitutions at position 69 with other NRTI resistance mutations in 789 sequences Backmatter: PMID- 12183249 TI - Human Immunodeficiency Virus Type 1 Genotypic and Pharmacokinetic Determinants of the Virological Response to Lopinavir-Ritonavir-Containing Therapy in Protease Inhibitor-Experienced Patients AB - The response to regimens including lopinavir-ritonavir (LPV/r) in patients who have received multiple protease (PR) inhibitors (PI) can be analyzed in terms of human immunodeficiency virus type 1 (HIV-1) genotypic and pharmacokinetic (pK) determinants. We studied these factors and the evolution of HIV-1 resistance in response to LPV/r in a prospective study of patients receiving LPV/r under a temporary authorization in Bordeaux, France. HIV-1 PR and reverse transcriptase sequences were determined at baseline LPV/r for all the patients and at month 3 (M3) and M6 in the absence of response to treatment. pK measurements were determined at M1 and M3. Virological failure (VF) was defined as a plasma viral load >=400 copies/ml at M3. A multivariate analysis of the predictors of VF, including clinical and biological characteristics and the treatment history of the patients, was performed. The PR gene sequence at M0, including individual mutations or a previously defined LPV mutation score (D. J. Kempf, J. D. Isaacson, M. S. King, S. C. Brun, Y. Xu, K. Real, B. M. Bernstein, A. J. Japour, E. Sun, and R. A. Rode, J. Virol. 75:7262-7269, 2001), and the individual exposure to LPV were also included covariates. Sixty-eight patients were enrolled. Thirty-four percent had a virological response at M3. An LPV mutation score of >5 mutations, the presence of the PR I54V mutation at baseline, a high number of previous PIs, prior therapy with ritonavir or indinavir, absence of coprescription of efavirenz, and a lower exposure to LPV or lower LPV trough concentrations were independently associated with VF on LPV/r. Additional PI resistance mutations, including primary mutation I50V, could be selected in patients failing on LPV/r. Genotypic and pK parameters should be used to optimize the virological response to LPV/r in PI-experienced patients and to avoid further viral evolution. Keywords: Introduction : Treatment of human immunodeficiency virus (HIV)-infected individuals with combination therapy including protease inhibitors (PI) results in a significant suppression of HIV replication and in improvement in clinical outcomes, with marked reductions in HIV-associated morbidity and mortality . However, the efficacy of antiretroviral (ARV) treatment can be impaired by several factors, including poor compliance with treatment regimens, suboptimal antiviral potency and drug concentrations, and selection of ARV-resistant HIV quasispecies . Resistance to PI is driven by the selection of primary mutations located close to the active site of the HIV type 1 (HIV-1) protease, producing significant changes in the affinity of the binding of the inhibitor to the mutant active site and often occurs early during virological rebound. Secondary resistance mutations may be selected later and may compensate for the initial decrease of viral fitness related to the appearance of primary mutations. These secondary mutations tend to be common to all PI, facilitating the emergence of resistance to the whole PI class. Lopinavir (LPV)-ritonavir (LPV/r) is a coformulation of lopinavir, an HIV PI, and low-dose ritonavir, which inhibits LPV metabolism and which enhances plasma LPV levels . LPV/r has shown significant potency in treatment-naive and in PI-experienced patients. Few data concerning the determinants and the emergence of drug resistance in LPV/r-treated patients are available. In the LPV/r arm of a first-line ARV therapy protocol, all virological failures (VF) were shown to correspond to rebounds with wild-type HIV-1 (B. Bernstein, J. Moseley, D. Kempf, M. King, K. Gu, E. Bauer, and E. Sun, Abstr. 8th Conf. Retrovir. Opportunistic Infect., abstr. 453, 2001). A panel of viral isolates from patients failing therapy with other PI were used to show that 11 amino acid mutations in the protease were associated with a reduced sensitivity to LPV . The number of baseline mutations out of the cumulative number of these mutations (LPV mutation score) was shown to be predictive of the virological response to a regimen including LPV/r in PI-experienced, nonnucleoside HIV-1 reverse transcriptase inhibitor (NNRTI)-naive patients (D. Kempf, S. Brun, R. Rode, J. Isaacson, M. King, Y. Xu, K. Real, A. Hsu, R. Granneman, Y. Lie, N. Hellmann, B. Bernstein, and E. Sun, 4th Int. Workshop HIV Drug Resist. Treatment Strategies, 12 to 16 June 2000, Sitges, Spain, abstr. 89, 2000). In this study, the overall virological response was important, since efavirenz (a NNRTI) was systematically coadministered, enhancing the efficacy of the ARV therapy. We thought it important to precisely identify the virological and pharmacological determinants of the virological response to LPV/r-containing regimens in the context of a salvage therapy in multiple-PI-experienced, frequently NNRTI-experienced patients. MATERIALS AND METHODS : Study design. | From May to November 2000, patients followed up at the Bordeaux University Hospital by the Aquitaine Cohort of the Groupe d'Epidemiologie Clinique du SIDA en Aquitaine and receiving LPV/r in the setting of the French Drug Agency temporary authorization for use (ATU) were prospectively enrolled into this study. Patients were PI experienced, and there were strict ATU entrance criteria (CD4 cells < 200/mul and log10 copies of plasma HIV-1 RNA/ml > 4) when enrollment began, with a progressive enlargement to authorize the use of LPV/r in a larger PI-experienced population. At baseline of LPV/r therapy (month 0 [M0]) demographic data, prior and current ARV regimens, HIV-1 RNA, and CD4+ cell count were collected, as well as HIV-1 protease and reverse transcriptase (RT) gene sequences. In two patients, on treatment interruption at M0, preinterruption parameters measured 2 (for one patient) and 3 months (for the other) before beginning LPV/r were considered for the analysis. Patients were followed up at M1, M3, and M6 with viral load and CD4 measurements and additional HIV-1 genotype determination in case of VF, defined as plasma HIV-1 RNA level of >=400 copies/ml. Virological analyses. | Plasma HIV-1 RNA was quantitated using the bDNA Quantiplex assay, version 3.0 (Chiron Bayer, Emeryville, Calif.). The RT and protease gene sequences were determined from plasma samples by the Agence Nationale de Recherches sur le SIDA (ANRS; Paris, France) consensus method with a CEQ L sequencer (Beckman Coulter) as previously described . All individual ARV resistance mutations reported by the International AIDS Society-USA (IAS-USA) panel were considered. For each patient at baseline LPV/r an LPV mutation score was defined as the number of protease mutations out of the following 11 mutations: L10F/I/R/V, K20M/R, L24I, M46I/L, F53L, I54L/T/V, L63P, A71I/L/V/T, V82A/F/T, I84V, and L90M. The virology laboratory of Bordeaux participates to the ANRS quality control assessment of HIV-1 drug resistance sequencing (F. Brun-Vezinet, D. Descamps, V. Calvez, M. L. Chaix, J. Izopet, B. Masquelier, A. Ruffault, C. Tamalet, P. Dehertogh, C. Loveday, L. Perrin, D. Costagliola, and The ANRS Resistance Group, 5th Int. Workshop HIV Drug Resist. Treatment Strategies, 4 to 8 June 2001, Scottsdale, Ariz., abstr. 157, 2001). Pharmacokinetic analysis. (i) Data preparation and pharmacokinetic analysis. | Patient data files were created from observed data for LPV plasma concentrations by using the PASTRX program in USC*PACK PC clinical programs (R. Jelliffe, University of Southern California, Los Angeles, Calif.). Clinical, pharmacokinetic, and demographic data and treatment history relevant to the pharmacokinetic analysis were obtained. At entry and each follow-up visit, the patient filled in a questionnaire seeking information on the dates, times, and doses of nucleoside RT inhibitors (NRTI) and PI taken and recent over-the-counter medication used. Each individual provided one blood sample per visit: a predose trough sample (for trough LPV plasma concentration [Cmin]) or a postdose sample (maximum concentration of the drug in plasma [Cmax]) collected between 2 and 5 h after the dose taken during the visit. Blood samples for the pharmacokinetic study were drawn at M1 (steady state) and then at M3. For the analysis of possible relationships between pharmacokinetic and virologic parameters, areas under the plasma concentration-time curve over a 12-h dosing interval (LPV AUC0-12) were used, with a cutoff value defined as the median LPV AUC0-12 obtained in the study. (ii) Plasma sample analysis. | Plasma LPV/r concentrations were measured by a validated high-performance liquid chromatography method and UV detection. The lower and upper limits of LPV quantification were 0.05 and 50 mg/liter, respectively. For that study, assay correlation coefficients (20 analytical runs) exceeded 0.998 for LPV. Based on quality control samples, interday accuracy for the two analytes ranged from 99 to 101% of a known concentration and interday variability was <8%. In total, 59 blood samples were drawn at M1 versus 65 at M3, with 0.87 sample per patient at M1 and 0.96 sample per patient at M3. (iii) Pharmacokinetic model. | The pharmacokinetic analysis was performed by nonparametric method NPEM2 (R. Jelliffe), which considered only LPV concentrations. NPEM2 used a model having absorptive, central, and peripheral compartments. Parameters included Ka (absorption rate constant), V (volume of distribution), Kcp, and Kpc (constants between central and peripheral compartments), Ke (elimination rate constant), and F (fraction orally absorbed). A one-compartment model with first-order absorption and first-order elimination was used to fit the data. Assay variability was determined, and the absorption rate constant and the ratio of the volume of distribution to the fraction orally absorbed (V/F) were fixed. Individual Ke was calculated, and ratio of clearance (CL) to F was extrapolated from the equation CL/F = Ke x V/F. Individual CL/F was used to determine individual exposure to LPV by calculating AUC0-12 from the equation CL/F = (dose x F)/AUC0-12. Statistical analysis. | For comparison of qualitative variables, we used the Pearson chi-square or Fisher test and Student's test or the nonparametric Kruskal-Wallis test for quantitative variables. All variables associated with the outcome (VF at M3, defined as >=400 copies of plasma HIV-1 RNA/ml) with a P value of <0.25 in the univariate analysis were included in multivariate models by using a descending stepwise logistic regression. The variables significant at the 0.05 level were kept in the final models to study first the significance of the LPV mutation score and then the significance of individual protease mutations and pharmacokinetic parameters for the virological response. Four different models were used because of the colinearity between the LPV mutation score and individual mutations and the lower number of patients with pharmacokinetic measurements. Analyses were performed using Stata (College Station, Tex.) statistical software. RESULTS : Baseline characteristics. | Sixty-eight patients were included in the study. Their biological and clinical characteristics are described in Table . Most of the patients were multiple-PI and NNRTI experienced, with a high viral load at baseline. The HIV-1 genotypes for 66 patients were obtained, the other two patients having a viral load below 500 copies/ml at baseline. The prevalence figures of baseline protease resistance mutations, as reported by the IAS-USA panel are summarized in Fig. in the format of categories of the LPV mutation score. The mean (+- standard deviation [SD]) number of protease resistance mutations was 7 +- 3 (median, 7; range, 1 to 19), the mean LPV score was 5 +- 2 (median, 5; range, 1 to 9). The RT gene sequences for 65 patients were determined; the prevalences of the baseline RT resistance mutations are shown in Fig. . The mean number of NRTI resistance mutations was 5 +- 2 (median, 5; range, 0 to 9), and the mean number of nucleoside-associated mutations (NAMS; formerly zidovudine resistance mutations; M41L, D67N, K70R, L210W, T215Y/F, and K219Q/E) was 3 +- 1 (median, 3; range, 0 to 5). Ten patients had no NNRTI resistance mutation. Description of the LPV/r-containing regimens and virological and immunological responses. | The NRTI combined with LPV/r were zidovudine, didanosine, zalcitabine, lamivudine, stavudine, and abacavir in 5 (7%), 45 (67%), 7 (10%), 32 (48%), 28 (42%), and 18 (27%) of the 68 patients, respectively. Nevirapine and efavirenz were prescribed in eight (12%) and seven (10%) patients, respectively. LPV/r-containing regimens associated three drugs in 53 patients, four drugs in 13 patients, and five drugs in 2 patients. The patients exhibited a median decrease of plasma HIV-1 RNA of 1.16 (0.40 to 2.37) log10 copies/ml at M3 (n = 62) and of 0.90 (0.42 to 1.80) log10 copies/ml at M6 (n = 63) and mean increases of 64 (range, 35 to 107) CD4+ cells/mul at M3 (n = 48) and of 81 (range, 34 to 140) CD4+ cells/mul at M6 (n = 35). The proportion of patients with a plasma HIV-1 RNA below 400 copies/ml was 34% at M3 (n = 65) and 28% at M6 (n = 61). With a less strict definition (plasma HIV-1 RNA of <400 copies/ml and/or a decrease of plasma HIV-1 RNA of >=1 log10 copy/ml), the proportion of virological response reached 55% at M3 and 46% at M6. Between M0 and M3, four patients stopped LPV/r therapy and one died of an AIDS-related syndrome. Two patients stopped LPV/r between M3 and M6. Pharmacokinetic results. | Sixty-eight patients were included in this pharmacokinetic study. At pharmacokinetic steady state, the mean LPV plasma concentrations and times to reach these concentrations at M1 and M3, respectively, were as follows: Cmin, 3.88 +- 1.97 and 4.22 +- 2.66 mg/liter; Tmin (time to reach LPV Cmin), 12.19 +- 2.56 and 12.49 +- 1.99 h; Cmax, 9.79 +- 5.09 and 9.10 +- 5.09 mg/liter; Tmax (time to reach LPV Cmax), 3.26 +- 1.42 and 2.29 +- 0.68 h. The LPV exposures determined from the individual AUC0-12 were 87 +- 21 and 77 +- 53 mg/liter h at M1 and M3, respectively. The median LPV AUC0-12 in the whole study was at 80 mg/liter h and defined the cutoff value for comparisons. The mean calculated Cmin values at M1 and M3 were 3.37 +- 0.86 and 2.93 +- 0.80 mg/liter, respectively. Determinants of the response to LPV/r. | The univariate analysis showed that the following were associated with a poor virological response (P < 0.250): the presence of baseline protease substitutions M46I, I54V, and V82A; an LPV mutation score >5; a large total number of protease mutations; a low LPV AUC0-12; a low LPV Cmin at M1 and M3; prior exposure to didanosine, abacavir, nevirapine, ritonavir, indinavir, or amprenavir; a high number of previous PI, a high number of treatment lines; and prior exposure to NNRTI. On the other hand, the coprescription of efavirenz, the presence of baseline protease mutation V77I, D30N, or N88D, an older age at inclusion, and the female gender were associated with a better virological response. The characteristics of the RT gene sequence (total number of NRTI and NNRTI mutations, number of NAMS) were not associated with the virological outcome at M3. The median LPV AUC0-12 values at M1 in patients with virological response and in patients with VF were 104 (interquartile range [IQR], 97 to 114) and 79 (IQR, 67 to 90) mg/liter h, respectively (P < 0.0001). The median LPV Cmin values at M1 in patients with virological response and in patients with VF were 4.01 (IQR, 3.76 to 4.23) and 3.08 (IQR, 2.65 to 3.81) mg/liter, respectively (P < 0.0001). The results of the multivariate analysis of the determinants of the virological response, according to four different models, are shown in Table . In the first model, a higher number of prior PI, prior therapy with at least one NNRTI, and an LPV mutation score higher than 5 were associated with VF; on the other hand, the female gender and the presence of efavirenz within the LPV/r-including regimen were associated with a better virological response. Four of six patients who received efavirenz with LPV/r had a virological response, with a baseline NNRTI resistance genotype encoding the following mutational patterns: V106A, K101E plus Y181I, wild type, and Y181C plus G190A. The other two patients had VF, one with a G190A mutation and one with a Y181C at baseline. To explain the protective effect of the female gender found in our analysis, we compared the different characteristics of the patients according to gender. No statistically significant difference was observed, but the median LPV Cmin at M1 tended to be higher in women than in men (3.93 versus 3.23 mg/liter; P = 0.08). The second model was constructed to evaluate the role of individual protease mutations. The presence of the I54V mutation was independently associated with a poor virological response at M3. The I77V mutation was associated with a better virological response at M3. The third model included pharmacokinetic parameters and the detailed description of prior PI use. Prior treatment with ritonavir or indinavir was associated with a worse response; a higher LPV AUC0-12 at M1 was associated with a better response. In a model including the number of prior PI but not the details of the prior PI, LPV AUC0-12 was still a predictive factor (data not shown). The fourth model shows that higher calculated LPV Cmin at M1 was also associated with a better response. The LPV mutation score remained predictive of the virological response in all the multivariate models. The virological response according to the LPV mutation score is shown in Fig. , with a reduced response in the patients with six or more mutations, particularly using a strict definition of the virological response. Evolution of the protease genotype in patients failing on LPV/r. | The protease gene sequence could be determined at M3 and/or M6 on an LPV/r-containing regimen for 48 patients without virological response at these points of follow-up. We determined the variations for any amino acid between baseline and M3 and M6. Fifteen patients exhibited no change between M0 and M3 or M6. For the other 33 patients, at least one residue was changed. As shown , different PI resistance mutations appeared on LPV/r, corresponding to mutations included or not included in the LPV mutation score. Interestingly, one primary resistance mutation (I50V) formerly described to be associated with resistance to amprenavir, was selected in four patients, three of whom had no prior amprenavir therapy. Other changes were selected at 13 different positions not involved so far in resistance to PI, with low frequencies (one or two patients for each amino acid); for four patients only changes within these 13 positions were detected. In 38 patients without virological response at M3 and M6 and with a calculated LPV AUC0-12 at M1 or M3, the viral evolution in the protease (with at least one additional PI resistance mutation at M3 and/or M6) occurred in 4 of 12 (33%) patients with low (<80 mg/liter h) LPV AUC0-12 and in 22 of 25 (88%) patients with correct or high LPV AUC0-12 (P < 0.001). FIG. 1. | (A) Mutations in the HIV-1 protease in 66 patients at baseline therapy with LPV/r. (A) Mutations in the HIV-1 protease in 66 patients at baseline therapy with LPV/r. Protease mutations are those reported by the IAS-USA panel . LPV mutations, mutations included in the LPV mutation score, i.e., the number of baseline protease mutations out of 11 possible mutations: L10F/I/R/V, K20M/R, L24I, M46I/L, F53L, I54L/T/V, L63P, A71I/L/V/T, V82A/F/T, I84V, and L90 M; other R mutations, mutations not included in the LPV mutation score; % of patients, percentage of patients with the corresponding mutation at baseline LPV/r. (B) Mutations in the HIV-1 RT in 65 patients at baseline LPV/r. NRTI and NNRTI resistance mutations are those reported by the IAS-USA panel. FIG. 2. | Virological response to LPV/r-containing regimens at M3 according to the LPV mutation score (n = 65). Virological response to LPV/r-containing regimens at M3 according to the LPV mutation score (n = 65). <400 copies/ml, patients with plasma HIV-1 RNA levels of <400 copies/ml at M3; <400 c/ml or -1log10, patients with plasma HIV-1 RNA levels of <400 copies/ml at M3 and/or with a decrease of plasma HIV-1 RNA by >1 log10 unit between M0 and M3; LPV mutation score, number of baseline protease mutations out of 11 mutations: L10F/I/R/V, K20M/R, L24I, M46I/L, F53L, I54L/T/V, L63P, A71I/L/V/T, V82A/F/T, I84V, and L90M. TABLE 1 | Clinical and biological characteristics of the patients at baseline LPV/r therapy, Aquitaine Cohort, 2000 to 2001 TABLE 2 | Multivariate analysis of the factors associated with VF under LPV/r therapy, Aquitaine Cohort, 2000 to 2001 TABLE 3 | Amino acid mutations selected in the HIV-1 protease in 48 patients on LPV/r-containing regimens, Aquitaine Cohort, 2000 to 2001 DISCUSSION : In this study, we described the determinants of the virological response to LPV/r-containing regimens in PI-experienced patients. The response to therapy in our study, using strict criteria (plasma HIV-1 RNA < 400 copies/ml at M3), was lower than in the initial study (Kempf et al., 4th Int. Workshop HIV Drug Resist. Treatment Strategies, 2000, Sitges, Spain, abstr. 89) of the response to LPV/r in PI-experienced patients. This could be explained by the characteristics of our population: highly pretreated and NNRTI and multiple PI experienced, with a high number of PI resistance mutations at baseline. Nevertheless, the previously constructed LPV mutation score was shown to be predictive of the virological response in our multivariate analysis, with a higher risk for VF with a score at six or more mutations. Similar data were obtained in another study (V. Calvez, I. Cohen-Codar, A. G. Marcelin, E. Guillevic, J. Isaacson, R. Rode, B. Bernstein, E. Sun, D. Kempf, and J. P. Chauvin, 5th Int. Workshop HIV Drug Resist. Treatment Strategies, 2001, Scottsdale, Ariz., abstr. 82, 2001), suggesting that the LPV mutation score could be relevant for optimizing the use of LPV/r in salvage therapy. We also explored the implication of specific mutations in the response to LPV/r. The presence of the L54V protease mutation at baseline was associated with a higher risk of VF. Although this mutation can be considered a secondary PI resistance mutation, it seems to play an important role in resistance to LPV. In a study using correlations between reduced phenotypic sensitivity to LPV and the protease gene sequences from clinical isolates (P. R. Harrigan, C. Van Den Eynde, and B. A. Larder, Scottsdale workshop, abstr. 49, 2001) the L54Vmutation was associated with the highest decrease in susceptibility to LPV/r. This mutation also belonged to the mutations appearing on LPV/r in the followed-up patients without optimal virological response in our study. These findings are consistent with the hypothesis that the addition of the I54V mutation on a background of previously established PI mutations enables the virus to acquire high-level resistance to LPV without impairing its replication capacity. Our finding that the use of efavirenz reduced the risk of treatment failure has to be interpreted with regard with the baseline pattern of NNRTI resistance mutations in patients who received efavirenz and had a virological response. The absence of mutation K103N, which is frequently found in efavirenz-resistant isolates, could explain a residual activity of efavirenz. A potential interpretation of these data, despite a limited number of patients, is that efavirenz could contribute significantly to the virologic success of a regimen, even in NNRTI-experienced patients, when administered as part of a sufficiently potent regimen. The observed individual LPV pharmacokinetic parameters such as Cmin, Tmin, Cmax, and Tmax seemed to be stable when the LPV pharmacokinetic was in steady state. This observed stability between M1 and M3 should enable us in the future to monitor individual LPV plasma concentrations and to adapt LPV/r dosages when necessary. All our pharmacokinetic results were similar to previously reported data (R. Bertz, W. Lam, and S. Brun, Abstr. 39th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 327, 1999). However, the Cmin and AUC0-12 levels found in the group of virological responders (about 4 mg/ml and 100 mg/ml h, respectively) suggest that higher LPV concentrations than previously described in first-line drug-treated patients could be necessary to be active on the first-available-PI-resistant viruses. A lower LPV individual exposure (AUC0-12), as well as lower LPV trough concentrations at M1, were associated with a higher risk of VF at M3; interestingly, the predictive value of the baseline LPV mutation score was strengthened in the multivariate models including the pharmacokinetics parameters, suggesting that both genotypic and pharmacokinetic follow-up are of interest for the monitoring of LPV/r-containing salvage regimens. One striking finding was the protective effect of the female gender found in the multivariate analysis. The higher LPV Cmin found in women could partially explain this effect. In a recent study, a lower clearance of saquinavir, resulting in an higher exposure to this drug, was found to be related to the female gender (R. C. Brundage, E. Acosta, R. Aubrich, D. Katzenstein, R. Culick, and C. V. Fletcher, Abstr. 9th Conf. Retrovir. Opportunistic Infect., abstr. 779, 2002). Although no data concerning the adherence to therapy, a putative confounding factor, were available in our study, a similar difference in LPV clearance could account for the observed effect of the female gender. In a subset of patients with uncontrolled plasma HIV-1 RNA on LPV/r, we found evidence for the selection of additional PI resistance mutations. Neither a specific resistance pattern nor any new and frequent mutation could be inferred from these data, which are consistent with the evolution toward high-level resistance to LPV through the accumulation of greater numbers of changes in the protease. However, the appearance of the I50V mutation in four patients was of particular importance, since it was recently shown to mediate the emergence of viruses cross-resistant to both LPV and amprenavir (J. G. Prado, J. G., T. Wrin, J. Beauchaine, L. Ruiz, C. J. Petroupoulos, B. Clotet, R. D'Aquila, and J. Martinez-Picado, Scottsdale workshop, abstr. 67, 2001), precluding the possibility of use of the latter drug in patients on LPV/r in which this mutation is selected. These data also suggest that further studies should be done in order to evaluate the effect of the I50V mutation on the response to an LPV/r-containing regimen in an amprenavir-experienced population. The correlations between pharmacokinetic parameters and the evolution of the HIV-1 protease in the patients on LPV/r without a complete virological response suggested that low plasma LPV concentrations favored the replication of unchanged PI-resistant viruses. In other patients, VF could occur despite normal LPV concentrations, with the selection of additional PI resistance mutations, leading to broader resistance patterns. In conclusion, our results showed that, in multiple-PI-experienced patients, the virological response to LPV/r-containing salvage regimens can be predicted by the LPV mutation score, even if specific mutations can play a major role in LPV resistance. LPV plasma levels should then be optimized by using pharmacokinetics measurements in order to avoid the evolution of the virus toward high-level resistance to LPV and to other PI. Backmatter: PMID- 12183281 TI - Comparative Activities of the Oxazolidinone AZD2563 and Linezolid against Selected Recent North American Isolates of Streptococcus pneumoniae AB - The activity of AZD2563 against 250 highly resistant pneumococci and 267 drug-susceptible isolates was determined. The AZD2563 MICs for 50 and 90% of the strains tested were 1 and 2 mug/ml and 0.5 and 1 mug/ml, respectively, for the two isolate groups. These MICs were within 1 log2 dilution of those of linezolid. Keywords: Introduction : The incidence of penicillin resistance and multidrug resistance in Streptococcus pneumoniae has continued to increase and has altered empirical choices of therapy for invasive pneumococcal infections . Multidrug resistance to three or more drug classes has made therapy for serious infections challenging, especially in severely ill patients. Resistance to traditional choices for therapy of pneumococcal infections (e.g., beta-lactams, macrolides) has prompted investigation of the activities of newer agents (e.g., fluoroquinolones, ketolides, carbapenems, and oxazolidinones) for predictable activity against highly resistant strains. Linezolid was approved by the U.S. Food and Drug Administration in 2000 and represents the first oxazolidinone compound to reach clinical use. It is indicated for therapy of serious infections due to resistant staphylococci or enterococci and also for community-acquired pneumonia due to S. pneumoniae. Oxazolidinones bind to the 50S ribosomal subunit and inhibit protein synthesis by a unique mechanism that does not appear to share cross-resistance with other protein synthesis inhibitors . Although resistance to linezolid has been found in Enterococcus isolates (G. E. Zurenko, W. M. Todd, B. Hafkin, B. Meyers, C. Kauffman, J. Bock, J. Slighton, and D. Shinabarger, Abstr. 39th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 848, 1999) and more recently in Staphylococcus aureus , it has not been found in S. pneumoniae or other streptococci. AZD2563 is a newer member of the oxazolidinone class that is currently under development. Two hundred fifty isolates of S. pneumoniae with previously determined resistance to one or more of the antimicrobial agents currently in use were tested against AZD2563 and selected comparative agents. A group of 267 drug-susceptible, invasive pneumococcal isolates were chosen from among the 1998 to 2000 Centers for Disease Control and Prevention Active Bacterial Core Surveillance/Emerging Infections Program collection. MICs of AZD2563, linezolid, quinupristin-dalfopristin, erythromycin, clindamycin, penicillin, levofloxacin, and vancomycin were determined by the broth microdilution method recommended by the NCCLS . This included cation-adjusted Mueller-Hinton broth supplemented with 3% lysed horse blood prepared as frozen microdilution panels (Trek Diagnostic Systems Inc., Westlake, Ohio), an inoculum of approximately 5 x 105 CFU/ml, and incubation at 35C for 20 to 24 h in ambient air prior to visual MIC determination. S. pneumoniae ATCC 49619 was included with each day's testing as a control organism. Susceptibility results obtained for the entire collection of 517 S. pneumoniae isolates with AZD2563 and the other comparator agents are depicted in Table . Among the drug-susceptible isolates, the MICs of AZD2563 for 50 and 90% of the isolates tested (MIC50 and MIC90, respectively) were 0.5 and 1.0 mug/ml, respectively, versus 1 and 2 mug of linezolid per ml. Among the drug-resistant isolates, the MIC50 and MIC90 were 1 dilution higher and identical to those of linezolid for this group. For both groups of isolates, the MIC90 for AZD2563 was 2 mug/ml. TABLE 1 | Susceptibilities of selected S. pneumoniae clinical isolates to AZD2563, linezolid, and representative agents of various other antimicrobial classes Figure shows the distributions of linezolid and AZD2563 MICs for the drug-resistant and drug-susceptible groups, respectively. The geometric mean MICs of AZD2563 were 0.70 and 0.94 mug/ml for the susceptible and resistant groups, respectively, compared to 0.99 and 1.14 mug/ml, respectively, for linezolid. For 218 isolates, the MICs of linezolid and AZD2563 were different. For 211 (96.8%) of these, the AZD2563 MIC was lower than that of linezolid. Of these, one was lower by 2 dilutions; all others were lower by a single dilution. In seven cases (3.2%), the MIC was lower by a single dilution for linezolid. MICs of linezolid were never lower than those of AZD2563 by more than a single dilution. FIG. 1. | MICs (micrograms per milliliter) of AZD2563 (AZD) and linezolid (LZD) for 250 drug-resistant S. pneumoniae MICs (micrograms per milliliter) of AZD2563 (AZD) and linezolid (LZD) for 250 drug-resistant S. pneumoniae isolates (A) and 267 drug-susceptible S. pneumoniae isolates (B). Because of the rising rate of resistance in pneumococci and the potential for development of resistance to antibiotics that have thus far been universally active, new agents are needed to provide predictable activity against multidrug-resistant strains. Linezolid has been shown to have almost universal activity against multiply resistant isolates of enterococci, staphylococci, and S. pneumoniae . The newer oxazolidinone AZD2563 has been shown in this study to have activity very similar to that of linezolid against highly resistant, as well as susceptible, recent clinical isolates of S. pneumoniae from geographically diverse regions of the United States. While susceptibility breakpoints have not been determined for AZD2563, the MIC50 and MIC90 of AZD2563 appear to be similar to or slightly lower than those of linezolid. Pharmacokinetic data regarding levels of AZD2563 in blood, cerebrospinal fluid, and other bodily fluid are needed to put into perspective this slightly greater potency of AZD2563. The isolates included in this study were representative of all known resistance phenotypes of S. pneumoniae (including beta-lactam, macrolide, lincosamide, fluoroquinolone, and streptogramin resistance), and regardless of the phenotype selected, all isolates were inhibited by AZD2563 at <=2 mug/ml. Thus, as with linezolid, no cross-resistance to AZD2563 and other classes of antibiotics was observed. This may be a result of the unique mechanism of action of the oxazolidinones, which appears to be inhibition of translation of rRNA . Linezolid has been demonstrated to be effective in the treatment of several serious gram-positive infections, including community-acquired pneumonia . Given the similar activities of AZD2563 and linezolid against S. pneumoniae, AZD2563 could also prove to be useful for therapy in patients with severe infections caused by resistant organisms. Additional studies are needed to determine its pharmacokinetics, safety, and in vivo clinical efficacy against invasive pneumococcal infections in humans. Backmatter: PMID- 12183259 TI - 16S rRNA Mutation-Mediated Tetracycline Resistance in Helicobacter pylori AB - Most Helicobacter pylori strains are susceptible to tetracycline, an antibiotic commonly used for the eradication of H. pylori. However, an increase in incidence of tetracycline resistance in H. pylori has recently been reported. Here the mechanism of tetracycline resistance of the first Dutch tetracycline-resistant (Tetr) H. pylori isolate (strain 181) is investigated. Twelve genes were selected from the genome sequences of H. pylori strains 26695 and J99 as potential candidate genes, based on their homology with tetracycline resistance genes in other bacteria. With the exception of the two 16S rRNA genes, none of the other putative tetracycline resistance genes was able to transfer tetracycline resistance. Genetic transformation of the Tets strain 26695 with smaller overlapping PCR fragments of the 16S rRNA genes of strain 181, revealed that a 361-bp fragment that spanned nucleotides 711 to 1071 was sufficient to transfer resistance. Sequence analysis of the 16S rRNA genes of the Tetr strain 181, the Tets strain 26695, and four Tetr 26695 transformants showed that a single triple-base-pair substitution, AGA926-928->TTC, was present within this 361-bp fragment. This triple-base-pair substitution, present in both copies of the 16S rRNA gene of all our TetrH. pylori transformants, resulted in an increased MIC of tetracycline that was identical to that for the Tetr strain 181. Keywords: Introduction : Helicobacter pylori is a spiral-shaped, gram-negative bacterium that causes chronic infections in the gastric mucosa . This infection will persist for life, unless treated with antibiotics. Cure of H. pylori infection results in ulcer healing and may reduce the risk of gastric cancer and gastric lymphoma . The highest cure rates have been obtained with antimicrobial treatments that include two or more antimicrobial drugs, a bismuth component, and/or a proton pump inhibitor . For the treatment of H. pylori infections, tetracycline-based triple or quadruple therapies are often used as a second-line treatment . Until the end of the last century only a few reports were published on spontaneous tetracycline resistance (; P. D. Midolo, M. G. Korman, J. D. Turnidge, and J. R. Lambert, Letter, Lancet 347:1194-1195, 1996), and it was generally accepted that tetracycline resistance (MIC >= 4 mug/ml) in H. pylori is very rare . However, in the last 2 years an increase in the incidence of tetracycline resistance in H. pylori has been reported . Tetracycline inhibits the protein synthesis by binding to the 30S ribosomal subunit . In most bacteria resistance to tetracycline is due to an energy-dependent efflux of tetracycline-cation complexes across the cell membrane by membrane-associated efflux proteins. Export of tetracycline complexes out of the cell reduces the intracellular drug concentration and protects the ribosomes from tetracycline . Overexpression of the efflux genes confers tetracycline resistance, while the sensitivity to tetracycline increases by deletions in these genes. The second common mechanism of resistance is mediated through ribosomal protection proteins. These cytoplasmic proteins confer tetracycline resistance either by a reduction of the affinity of ribosomes for tetracycline or by releasing the bound antibiotic from the ribosome. The ribosomal protection proteins, such as TetM, TetO, and TetS, show homology with the elongation factors EF-G and EF-Tu . Beside these two most common tetracycline resistance mechanisms, two other mechanisms have been described. One is based on enzymatic inactivation of tetracycline by the product of TetX in the presence of oxygen and NADPH, and the other originates from mutations in the 16S rRNA genes that affect the binding site of tetracycline . TABLE 1 | H. pylori genes potentially involved in tetracycline resistance We recently isolated a tetracycline-resistant (Tetr) H. pylori isolate (strain 181), from a 72-year-old male dyspeptic patient. Here we describe the molecular mechanism of tetracycline-resistance in this strain. To achieve this, 12 genes were selected from the published H. pylori genomes as potential candidates, based on their homology with tetracycline resistance genes in other bacteria . These putative tetracycline resistance genes were amplified from the genome of the Tetr strain 181 and used for genetic transformation of the tetracycline-sensitive (Tets) strain 26695 in order to identify the changes responsible for tetracycline resistance. MATERIALS AND METHODS : Bacterial strains and growth conditions. | H. pylori strains used in this study were the Tetr strain 181 and the TetsH. pylori strains 26695 , J99 , SS1 , and ATCC 43504 (American Type Culture Collection). Bacteria were routinely grown on Columbia agar plates (Becton Dickinson, Cockeysville, Md.) supplemented with 7% lysed horse blood (BioTrading, Mijdrecht, The Netherlands) and H. pylori Dent selective supplement (Oxoid, Basingstoke, United Kingdom), referred to as Dent plates. Bacteria were inoculated on these plates and incubated for 48 to 72 h at 37C in a microaerobic atmosphere of 5% O2, 10% CO2, and 85% N2. Bacterial stocks were prepared by suspending bacteria, harvested from culture plates with a sterile cotton swab, in brain heart infusion with 20% glycerol and stored at -80C. Determination of MIC. | The MIC was routinely determined with the E-test (AB Biodisk, Solna, Sweden) . Inocula were prepared from a fresh H. pylori culture grown routinely for 2 days on Dent plates. Columbia agar plates containing 7% lysed horse blood, but no Dent supplement, were inoculated with approximately 2 x 108 CFU in 20 mul of 0.9% NaCl, the plates were dried for 3 to 4 min, and then the E-test strips were applied to the agar surface. The plates were incubated at 37C under microaerobic conditions, and 3 days later the MIC was determined by the intercept of the zone of inhibition with the graded E-test strip. By this method the susceptibility was determined for tetracycline, doxycycline, minocycline, amoxicillin, clarithromycin, and metronidazole. The isolates were considered resistant when the MICs of tetracycline, doxycycline, and minocycline were >=4 mug/ml and when those of amoxicillin, clarithromycin, and metronidazole were >=8, >=2, and >=8mug/ml, respectively . Natural transformation of H. pylori. | Bacteria were transformed with similar1 mug of genomic DNA or similar250 ng of PCR-amplified gene products from strain 181, as described previously . Tetr transformants were selected on Dent plates containing tetracycline (2 mug/ml; Sigma Aldrich Chemie, Zwijndrecht, The Netherlands). As controls, bacteria were transformed with either genomic DNA of the Tetr strain, TE (1 mM Tris-HCl, 0.1 mM EDTA [pH 8.0]), or DNA from the Tets strains 26695, J99, SS1, and ATCC43504. Individual bacterial colonies present on tetracycline-containing plates (2 mug/ml) were selected, and their MICs of tetracycline were determined. PCR. | Oligonucleotide primers (Isogen, Maarsen, The Netherlands) used for PCR amplification were based on the genome sequences of H. pylori strains 26695 and J99 (Table ; Fig. and ) . PCR was performed in an automated thermal cycler (I-Cycler; Bio-Rad), in a final volume of 50 mul, using the PCR-core system I (Promega, Madison, Wis.), with approximately 25 pg of template genomic DNA and 25 pmol of each primer. Sequence analysis. | Direct sequencing of the obtained PCR products was performed by Baseclear Inc. (Leiden, The Netherlands). Sequence data were analyzed with the help of Lasergene (DNAstar, Madison, Wis.), and Sci Ed Central (Scientific & Educational Software, Durham, N.C.) software. Nucleotide sequence accession number. | The 16S rRNA gene sequence of TetrH. pylori strain 181 has been deposited into to the GenBank sequence database, under accession no. . FIG. 1. | Schematic representation of the 16S rRNA genes of H. pylori Schematic representation of the 16S rRNA genes of H. pylori. Alignment of the 16S rRNA genes (rrnA and rrnB) from the Tetr strain 181, the Tets strain 26695, and four Tetr 26695 transformants is shown. Sequence analysis of the 16S rRNA genes revealed only a few base pair substitutions (numbering according to 16S rrnA of H. pylori strain 26695) in the Tetr strain 181 that did not occur in the Tets strain 26695. For the identification of the 16S rRNA region required for tetracycline resistance, the Tets strain 26695 was transformed with smaller overlapping PCR fragments of the 16S rRNA gene of the Tetr strain 181 (only fragments containing mutations are shown). The transformants were selected on tetracycline (2 mug/ml)-containing Dent plates. Primers used for the amplification of the smaller overlapping PCR fragments started at the outside of the fragment and each had a length of 20 bp. FIG. 2. | Both 16S rRNA genes are mutated in H. pylori Both 16S rRNA genes are mutated in H. pylori tetracycline resistance. rrnA- and rrnB-specific sequences were amplified using specific primers based on sequences which are found outside the two 16S rRNA genes. The rrnA-specific primers, F1 and F2, are located at position 1207020 and 1207242 (numbers corresponding to the H. pylori 26695 sequence ), respectively, and the rrnB-specific primers, F3 and F4, are located at position 1510569 and 1510809, respectively. For amplification, primer R1 (located at position 1208293 and 1511828) was used in combination with one of the other primers. All primers had a length of 20 bp. RESULTS : Determination of the MICs of various antibiotics. | MICs of the four antibiotics commonly used in anti-H. pylori therapy, as well as those of two antibiotics that belong to the tetracycline family, were determined by E-test for the TetrH. pylori strain 181 and the H. pylori reference strain 26695 . The MIC of tetracycline for strain 181 was 8 mug/ml (susceptibility breakpoint >= 4 mug/ml), while the MIC for strain 26695 was 0.19 mug/ml. The MICs of the two other tetracyclines, doxycycline and minocycline, were also significantly higher for strain 181 than for strain 26695. For the three other routinely used antibiotics (amoxicillin, clarithromycin, and metronidazole) the MICs varied between <0.016 and 0.064 mug/ml and did not differ significantly between strains 181 and 26695. Transfer of tetracycline resistance by natural transformation. | Transformation of H. pylori strain 26695 (MIC, 0.19 mug/ml) with genomic DNA of strain 181 (MIC, 8 mug/ml) resulted in Tetr colonies with a transformation frequency of 6 x 10-5. The MIC of tetracycline for the 10 randomly selected Tetr transformants (obtained from three independent transformation experiments), determined by E-test was 8 mug/ml , which is identical to that for the TetrH. pylori strain 181. The Tetr transformants also displayed an increase of MIC of the tetracycline derivatives, doxycycline and minocycline . Transformation with PCR products of putative tetracycline resistance genes. | Based on their homology with tetracycline resistance genes in other bacteria, 12 genes were selected from the published genome sequences of H. pylori strains 26695 and J99 . The TetsH. pylori strain 26695 was transformed with the PCR products of the selected tetracycline resistance genes, which were amplified from genomic DNA of the Tetr strain 181. Only transformation with the PCR product of the 16S rRNA genes resulted in Tetr transformants, with a transformation frequency of 4 x 10-5. No Tetr transformants were found after transformation with one of the other selected genes, TE, or DNA from the Tets strain 26695. Similar results were found for the Tets strains J99 (MIC, 0.5 mug/ml), SS1 (MIC, 0.19 mug/ml), and ATCC 43504 (MIC 0.125 mug/ml). For all strains the MIC of tetracycline for 10 randomly selected Tetr transformants (obtained from three independent transformation experiments) determined by E-test was 8 mug/ml, which is identical to that for the Tetr donor strain 181. Comparison of the 16S rRNA gene sequences of the Tetr strain 181, the Tets strain 26695, and four randomly Tetr 26695 transformants (obtained after transformation with genomic DNA of strain 181) revealed several base pair differences in the Tetr strain 181 as well as in the Tetr transformants that did not occur in the Tets strain 26695 . Three Tetr 26695 transformants had incorporated the complete 16S rRNA gene of strain 181, while the fourth transformant contained the first part of the 16S rRNA gene of strain 26695 and the second part of strain 181. The DNA crossover in this transformant occurred after nucleotide 93 and before 127 (numbering according to 16S rrnA of H. pylori strain 26695). For each strain or transformant, only one sequence was obtained for the 16S rRNA genes, indicating that these 16S rRNA genes were identical in these strains. Identification of 16S rRNA mutations involved in tetracycline resistance. | To determine which residues of the 16S rRNA genes were responsible for tetracycline resistance in strain 181, the Tets strain 26695 was genetically transformed with smaller overlapping PCR products of the 16S rRNA gene fragments, amplified from genomic DNA of the Tetr strain 181. Tetr transformants were only observed after transformation with a 361-bp DNA fragment that spanned nucleotides 711 to 1071 (numbering according to 16S rrnA of H. pylori strain 26695), with a transformation frequency of 5 x 10-6. Transformation with the other DNA fragments did not result in transfer of tetracycline resistance. The MIC of tetracycline for 10 randomly selected Tetr transformants determined by E-test was 8 mug/ml, which was identical to that for the Tetr donor strain 181. The only difference found between these Tetr 26695 transformants and the Tets strain was the triple-base-pair substitution AGA926-928->TTC . Both copies of 16S rRNA genes are involved in tetracycline resistance. | The primers that were originally used for amplification of the 16S rRNA genes did not distinguish between the two copies present on the H. pylori chromosome . To assess the involvement of each copy of the 16S rRNA genes in tetracycline resistance, specific oligonucleotide primers were developed . These specific primers are based on sequences which are located approximately 350 to 600 bp outside the both 16S rRNA genes, rrnA and rrnB. This allowed amplification of rrnA- and rrnB-specific sequences. rrnA- and rrnB-containing PCR-fragments were obtained for the Tetr strain 181, the eight 26695 transformants, and Tets strain 26695, and their DNA sequences were determined. As expected, the rrnA and rrnB sequences were identical, where as the sequences outside the 16S rRNA genes were different. While the Tets strain-derived fragments contained the AGA sequence in both genes, both for strain 181 and the eight 26695 transformants, the triple-base-pair substitution AGA926-928->TTC was found in both copies of the 16S rRNA genes. TABLE 2 | MICs for various H. pylori strains as determined by E-test DISCUSSION : Until recently tetracycline resistance in H. pylori was rare , but in the last 2 years, several TetrH. pylori strains have been isolated (, , ; Midolo et al., letter). These Tetr clinical isolates showed, besides tetracycline resistance, cross-resistance to metronidazole (, , ; Midolo et al., letter). The tetracycline resistance present in these strains was always transferred together with metronidazole resistance to a Tets strain . In these Tetr strains it is not clear whether the tetracycline resistance is caused by a known metronidazole resistance mechanism, a multidrug resistance mechanism, or an unknown tetracycline resistance mechanism . In our TetrH. pylori strain 181, no cross-resistance was found against metronidazole, which indicated that the molecular mechanism of tetracycline resistance in strain 181 could be different from that of these earlier-described TetrH. pylori strains. In H. pylori strain 181 resistance to tetracycline is mediated by a single triple-base-pair substitution, AGA926-928 -> TTC (corresponding to bp 965 to 967 of Escherichia coli 16S rRNA), present in both copies of the 16S rRNA gene. Tetracycline has one primary and multiple secondary binding sites within the 30S ribosomal subunit . In the primary binding site, tetracycline binds exclusively to the 3'-major domain of the 16S rRNA. The primary binding pocket for tetracycline is formed by the 16S rRNA residues 1054 to 1056 and 1196 to 1200 of helix 34 and residues 964 to 967 of helix 31 (numbers corresponding to E. coli 16S rRNA) . The residues 1054 and 1196 interact primarily with tetracycline through hydrophobic interactions, but the majority of the interaction with the drug is made through hydrogen bonds and salt bridges between tetracycline and the 16S rRNA residues . In the TetrH. pylori strain 181, the triple-base-pair substitution AGA926-928->TTC is located right in the primary binding site of tetracycline. Mutations in this primary binding site are likely to affect the affinity of the drug-ribosome interaction and thus the efficacy of tetracycline as a translational inhibitor. In E. coli, the nucleotides G966 and C967 are located not only in the primary binding site of tetracycline but also in a functional region of the ribosome, the P site . Mutations in this region may affect protein synthesis , either by a change in binding of tRNA to the P site itself or by blocking the conformational change needed for the tRNA binding to the A site. In H. pylori strain 181 and the Tetr transformants of strain 26695, the triple-base-pair substitution AGA926-928->TTC had no effect on the growth rate of the bacterium in the presence or absence of tetracycline (data not shown). Similar observations were found with E. coli after the substitution of the nucleotides G966 and C967 . This suggests that the triple-base-pair substitution AGA926-928->TTC present in H. pylori strain 181 does not seem to affect protein synthesis of H. pylori. During revision of this work, Trieber and Taylor reported that an identical AGA->TTC substitution mediates tetracycline resistance in an unrelated H. pylori strain (Midolo et al., letter). None of the other mutations found in their isolates (G332->A, and the deletions G733 and G903 [numbering according to 16S rrnA of H. pylori strain 26695]) play a role in tetracycline resistance of the Tetr strain 181, since these mutations were not present in our Tetr isolate. The differences found in the MIC of tetracycline for the triple-base-pair substitution mutant between the study of Trieber and ours are only due to the methods used for the determination of the MIC (data not shown). The finding that in two unrelated H. pylori strains the exact same mutation is responsible for tetracycline resistance opens possibilities for the development of molecular screening tests for tetracycline resistance in H. pylori. FIG. 3. | Schematic representation of the primary binding site of tetracycline, based on the 16S rRNA structure of Thermus thermophilus proposed by Wimberly et al. Schematic representation of the primary binding site of tetracycline, based on the 16S rRNA structure of Thermus thermophilus proposed by Wimberly et al. . The primary binding pocket for tetracycline is formed by the 16S rRNA residues 1054 to 1056 (box A) and residues 1196 to 1200 (box B) of helix 34 and residues 964 to 967 of helix 31 (box C). The interactions between tetracycline and this pocket are formed by hydrophobic interactions, hydrogen bonds, and salt bridges . The triple-base-pair substitution AGA926-928->TTC (corresponding to bp 965 to 967 of E. coli 16S rRNA) is located in box C and is indicated by asterisks. Backmatter: PMID- 12183270 TI - Clinical Isolates of Staphylococcus aureus with Ribosomal Mutations Conferring Resistance to Macrolides AB - Six strains of Staphylococcus aureus isolated from cystic fibrosis patients after treatment with azithromycin were cross-resistant to azithromycin and erythromycin. None of the isolates contained erm or msr(A) genes, but they all carried either A2058G/U or A2059G mutations within the rrl genes, with a majority of the rRNA copies bearing the mutation. One strain displayed an additional mutation in the rplV gene, encoding the L22 ribosomal protein. Keywords: Introduction : Emergence of resistance to macrolides in staphylococci shortly after the therapeutic use of erythromycin has been reported . In most cases, macrolide resistance in clinical isolates of staphylococci has been linked to target site alteration due to methylation of adenosine 2058 (A2058) of 23S rRNA within the large ribosomal subunit . These ribosomal methylases are encoded by erm genes. In some cases, ABC transporters encoded by plasmid-borne msr(A) genes cause active efflux of 14-member-ring (erythromycin, clarithromycin, roxithromycin, and dirithromycin) or 15-member-ring (azithromycin) macrolides . Rare staphylococcal strains have been reported to produce a macrolide phosphotransferase which inactivates some of these antimicrobials . Overall, in several survey studies, drug efflux and ribosomal methylation have been found to be responsible for macrolide resistance in all of the strains studied . The use of macrolides for the treatment of staphylococcal infections is generally limited to uncomplicated soft tissue infections. Recently, it has been inferred from anti-inflammatory and antiadhesion effects of macrolides observed in vitro that these antimicrobials may have a favorable action at low concentrations for treatment of Pseudomonas aeruginosa infections . These indirect effects might be helpful in the case of cystic fibrosis, where the major cause of morbidity and mortality remains respiratory disease, with P. aeruginosa as the most frequently organism isolated, followed by Staphylococcus aureus . We report on six strains of erythromycin-resistant S. aureus, isolated from patients suffering from cystic fibrosis, with unusual mutations of the ribosomal target of macrolides. Bacterial strains and antimicrobial susceptibility testing. : Among 12 S. aureus strains resistant to erythromycin and isolated from cystic fibrosis patients, 6 did not contain erm or msr(A) genes as determined by PCR . These strains, S. aureus UCN13, UCN14, UCN15, UCN16, UCN17, and UCN18, were isolated at the hospitals of Brest and Caen, France, from the sputa of five patients (S. aureus UCN13 and UCN14 were isolated from the same patient) suffering from cystic fibrosis who were treated with azithromycin (10 mg/kg/day) for a minimum of 3 months and a maximum of 1 year. The strains were considered to be genetically unrelated since the patterns of SmaI-restricted DNA differed by more than three fragments after pulsed-field gel electrophoresis analysis (data not shown) . All isolates were resistant to erythromycin as determined by the disk diffusion technique. Macrolide-susceptible S. aureus ATCC 29213 was included as a control. MICs of antibiotics were determined by the agar dilution method with Mueller-Hinton medium. Dalfopristin (RP54476), erythromycin, quinupristin (RP57669), quinupristin-dalfopristin, pristinamycin, and spiramycin were from Aventis Pharma (Romainville, France), and lincomycin was from Pharmacia-Upjohn (Kalamazoo, Mich.). PCR and DNA sequence analysis. : As mentioned above, no rRNA methylase genes [erm(A), erm(B), and erm(C)] or efflux gene [msr(A)] could be detected by PCR with specific primers . Since mutations in genes coding for L4 or L22 ribosomal proteins or in domains II and V of 23S rRNA have been reported to be responsible for macrolide resistance in a variety of bacterial species, we hypothesized that similar mutations might account for resistance in the strains studied . Portions of rrl genes for domains II and V of 23S rRNA and the genes for ribosomal proteins L4 and L22 were amplified by PCR from total genomic DNA with the oligonucleotides shown in Table . The amplification primers were designed after analysis of the sequence of S. aureus COL obtained from The Institute for Genomic Research website . Mutations were screened for by PCR-SSCP, as described previously . After heat denaturation, the single-stranded PCR products were separated by nondenaturing polyacrylamide gel electrophoresis. Fragments with mobilities different from those of susceptible controls were sequenced. We also determined the copy numbers of the rrl genes carrying 23S rRNA mutations. A strategy was developed to amplify the different copies of rrl domain V individually. Sequence analysis of S. aureus COL showed that the strain carried six copies of the rrl gene. We used primers complementary to unique sequences downstream from each rrl gene (V LA to V LF in Table ) and a primer common to the six alleles and complementary to a region upstream from the peptidyl transferase region in domain V (V U1). Internal primers (V U2 and V L2) were then used to amplify a 144-bp fragment encompassing the domain V region, which was subsequently sequenced. TABLE 1 | Oligodeoxynucleotides used for the amplification of fragments of the 23S rRNA gene and of ribosomal protein genes Susceptibility to antimicrobials. : The six isolates were highly resistant to erythromycin and azithromycin (MIC of >128 mug/ml) . Five strains were also resistant to the 16-member-ring macrolide spiramycin, whereas the spiramycin MIC for strain UCN16 was lower (8 mug/ml). MICs of quinupristin, a streptogramin B, and clindamycin were more widely distributed. MICs of dalfopristin, a streptogramin A, were similar to those for susceptible strains . All strains except strain UCN15 were susceptible to quinupristin-dalfopristin and pristinamycin. TABLE 2 | MICs of macrolides, lincosamides, and streptogramins Identification of ribosomal mutations. : Sequencing showed that SSCP mobilities of fragments different from those of the controls amplified from S. aureus ATCC 29213 were associated with point mutations or a deletion. Four strains, S. aureus UCN13, UCN15, UCN16, and UCN17, carried an A2058G (E. coli numbering) transition; S. aureus UCN14 contained an A2058T transversion; and S.aureus UCN18 had an A2059G transition. No mutation was detected in domain II of 23S rRNA or in the rplD gene (encoding the L4 protein). In strain UCN15, a deletion of nine nucleotides which would result in the deletion of three amino acids at position 101 of the deduced amino acid sequence of protein L22 was associated with the A2058G mutation. Mutations at positions A2058 and A2059 were associated with macrolide-lincosamide-streptogramin B and macrolide-lincosamide phenotypes, respectively, similar to those already reported for other organisms . Relative to other rRNA mutations, A2058G gives the highest level of resistance to 14-member-ring macrolides and confers macrolide-lincosamide-streptogramin B resistance, defined as high resistance to all of the antimicrobials in this group. However, MICs of clindamycin and quinupristin were lower than expected for strains UCN14, UCN15, UCN16, and UCN18. This might be due to the fact that these staphylococci grew slowly on agar, as has been reported for most S. aureus strains isolated from cystic fibrosis patients, which frequently yield small-colony variants . The A-to-U or A-to-G substitutions gave a similar level of resistance. The A2059 mutation gave the macrolide-lincosamide phenotype, with moderate resistance to clindamycin and no resistance to streptogramins B, as previously reported for Helicobacter pylori and Streptococcus pneumoniae . Resistance by mutation in 23S rRNA has generally been reported for bacteria with few copies of rrn operons, such as H. pylori, Mycoplasma pneumoniae, Mycobacterium intracellulare, and Mycobacterium avium. However, mutations in 23S rRNA have been reported for S. pneumoniae, which contains four rrn operons . In PCR experiments carried out to determine the copy numbers of mutated rrl genes, one copy, named B, could not be amplified from five of our strains . In silico analysis of strain N315 DNA, available at the National Center for Biotechnology Information website , revealed that it also contained only five rrl copies and that the copy at a position similar to that of the B copy was missing. In every erythromycin-resistant strain, three or four copies were mutated, confirming that mutation in a majority of rrl operons was associated with significant resistance, as reported for S. pneumoniae . Resistance to streptogramins was observed in strain UCN15, which combined a rplV mutation with an A2058G mutation. Recently, resistance to quinupristin-dalfopristin in staphylococcal strains selected under treatment with this antibiotic has been explained by similar mutations in the conserved 3' end of the rplV gene ). Isolation of staphylococci with resistance to macrolides conferred by ribosomal mutations is unusual. It may be related to the specific context of cystic fibrosis. Half of 12 erythromycin-resistant strains of S. aureus isolated from cystic fibrosis patients in our institutions were mutants. Indeed, patients suffering from this genetic disorder receive multiple courses of antibiotics. In addition, administration of macrolides at low doses aimed at preventing Pseudomonas infections might favor the emergence of mutants, although no definitive relationship could be proven in this study. Finally, colonization of cystic fibrosis patients by hypermutable strains of P. aeruginosa has been reported . A similar possibility for the staphylococcal isolates is currently under investigation. TABLE 3 | Mutations of the rrl genes Backmatter: PMID- 12183226 TI - Oxidative Stress Increases Susceptibility of Mycobacterium tuberculosis to Isoniazid AB - Isoniazid is a first-line antibiotic used in the treatment of infections caused by Mycobacterium tuberculosis. Isoniazid is a prodrug requiring oxidative activation by the catalase-peroxidase hemoprotein, KatG. Resistance to isoniazid can be obtained by point mutations in the katG gene, with one of the most common being a threonine-for-serine substitution at position 315 (S315T). The S315T mutation is found in more than 50% of isoniazid-resistant clinical isolates and results in an ~200-fold increase in the MIC of isoniazid compared to that for M. tuberculosis H37Rv. In the present study we investigated the hypothesis that superoxide plays a role in KatG-mediated isoniazid activation. Plumbagin and clofazimine, compounds capable of generating superoxide anion, resulted in a lower MIC of isoniazid for M. tuberculosis H37Rv and a strain carrying the S315T mutation. These agents did not cause as great of an increase in isoniazid susceptibility in the mutant strain when the susceptibilities were assessed by using the inhibitory concentration that causes a 50% decrease in growth. These results provide evidence that superoxide can play a role in isoniazid activation. Since clofazimine alone has antitubercular activity, the observation of synergism between clofazimine and isoniazid raises the interesting possibility of using both drugs in combination to treat M. tuberculosis infections. Keywords: Introduction : Entering the 21st century, Mycobacterium tuberculosis, the causative agent of tuberculosis (TB), remains a major cause of morbidity and mortality worldwide. TB causes 1.9 million deaths annually among a pool of infected individuals close to 2 billion people . Isoniazid (INH) is one of the most effective and widely used therapeutic agents for the treatment of TB. M. tuberculosis shows exceptional sensitivity to INH, being several orders of magnitude more sensitive than most other bacterial species. Much is known about the mechanism of INH activation, including that fact that it is a prodrug requiring activation by the catalase-peroxidase hemoprotein, KatG , in a process that requires molecular oxygen . It has been further shown that the activated form of INH forms a covalent adduct with NAD+ to generate a potent inhibitor of the InhA protein of M. smegmatis, an enoyl-acyl carrier protein reductase important in mycolic acid biosynthesis . Recent investigations have led to the hypothesis that the sensitivity of M. tuberculosis to INH may be governed by the level of expression of the alkyl hydroperoxide reductase, AhpC, since M. tuberculosis does not express the ahpC gene due to inactivation of the oxidative stress regulatory gene oxyR . Nevertheless, specific details regarding the mechanism of action of INH, such as the chemical nature of the activated form of INH, have yet to be determined. For example, InhA as the primary target for INH in M. tuberculosis has been questioned following the discovery of a covalent complex of INH, acyl carrier protein (AcpM), and the beta-ketoacyl acyl carrier protein synthase, KasA . Rates of resistance to INH and other antibiotics have been increasing such that now approximately 13% of all TB cases in the United States are resistant to at least one first-line drug (INH, rifampin, pyrazinamide, ethambutol, and streptomycin) . INH-resistant M. tuberculosis has been associated with deletions or point mutations in the katG gene , with the threonine-for-serine substitution at position 315 (S315T) in katG being one of the most common mutations found in clinical isolates. A comparison of M. tuberculosis KatG with cytochrome c peroxidase, a member of the same catalase-peroxidase superfamily, suggests that S315 occupies a position near the active site of KatG, and therefore, the S315T mutation could affect the enzymatic activity of KatG . Besides exhibiting catalase and peroxidase activities, several other enzymatic activities have been associated with M. tuberculosis KatG, including Mn(II)-dependent peroxidase , peroxynitritase , and cytochrome P450-like monooxygenase activities. The KatG protein and the KatG protein with the S315T mutation [KatG(S315T)] have comparable catalase and peroxidase activities, and KatG(S315T) is able to oxidize INH at equivalent rates using a hydroperoxide as oxidant, suggesting that the peroxidase activity of KatG may not be relevant for the in vivo activation of INH . In the presence of dioxygen but the absence of a hydroperoxide, INH is oxidized more slowly by KatG(S315T) than by KatG, a phenomenon that has been shown to involve superoxide and that may reflect why the S315T mutation confers INH resistance . Superoxide is formed during INH oxidation and is thought to be involved in the activation process (-). INH oxidation via this route is thought to involve a monooxygenase pathway in which an oxyferrous KatG intermediate is formed by either dioxygen binding to ferrous KatG or superoxide binding to ferric KatG . These biochemical observations support the hypothesis that a mechanism other than the catalase-peroxidase route plays an important role in INH resistance. Direct evidence for this was provided by Drlica and colleagues , who showed that the superoxide generator plumbagin resulted in an increase in the bacteriostatic activity of INH against M. smegmatis, a strain normally resistant to INH. In the study described in this report we tested the hypothesis that superoxide plays a role in KatG-mediated INH oxidation in M. tuberculosis. Radiometric analyses were conducted to determine MICs and the inhibitory concentrations of drug that cause a 50% decrease in strain growth (IC50s) of INH alone and INH in combination with superoxide-generating substances, plumbagin or clofazimine, for an INH-sensitive M. tuberculosis strain and an INH-resistant strain containing the katG S315T mutation. MATERIALS AND METHODS : Bacterial strains. | The M. tuberculosis strains used in this study, strains H37Rv and TBC3, were grown on Middlebrook 7H10 slants at 37C. TBC3 is an INH-resistant clinical isolate. Drugs and reagents. | Dimethyl sulfoxide (DMSO), INH, clofazimine, and plumbagin (technical grade) were purchased from Sigma Chemical Co. (St. Louis, Mo.); shrimp alkaline phosphatase and exonuclease I were provided by U.S. Biochemicals (Cleveland, Ohio). INH was recrystallized from boiling methanol by using activated charcoal decolorization and hot filtration through Whatman no. 1 filter paper (Whatman International Ltd., Maidstone, United Kingdom). The crystals were vacuum filtered, washed with ice-cold methanol, and air dried, yielding white needles with a melting point of 172C. INH was dissolved in water and filter sterilized through a 0.22-mum-pore-size filter. For stock solutions of clofazimine (16.9 muM) and plumbagin (1.6 mM), 4% DMSO was used as the solvent. Appropriate controls were included to ensure that the inhibitory effects of the drugs were not due to the addition of DMSO, although there are data indicating that this solvent enhances the potencies of the drugs . DNA sequencing and sequence analysis. | The presence of a wild-type katG gene in strain H37Rv and a katG gene harboring the S315T mutation in strain TBC3 was confirmed by restriction fragment length polymorphism analysis . Sequencing of the DNA of the entire katG gene from TBC3 was also carried out to confirm the presence of the S315T mutation and the absence of secondary mutations. Briefly, the katG gene was amplified by PCR, and the genomic DNA was isolated from TBC3. Genomic DNA was isolated by inoculating a 10-mul loopful of a culture into a tube containing 100 mul of water. Approximately 2 ml of alkaline wash solution (0.05 M sodium citrate, 0.5 M sodium hydroxide) was added. After centrifugation (at 28,800 x g for 2 min) the supernatant was discarded. Approximately 500 mul of 0.5 M Tris HCl (pH 8.0) buffer was added. After centrifugation the supernatant was discarded and 100 mul of sterile distilled water was added. The tube was incubated at 95C for 15 min. The following oligonucleotide primers whose sequences are specific for sequences flanking the katG gene (GenBank database accession number ) were used as primers in the PCR: KatG 5' (5'-CCGACACTTCGCGATCACATCCGTGATCACAGCCC-3') and KatG 3' (5-GGTGCTGCGGCGGGTTGTGGTTGATCGG-3'). The PCR was carried out with a 50-mul reaction volume containing PCR buffer, genomic DNA (2 mul), 1.1 mM magnesium diacetate, 200 muM deoxynucleoside triphosphates, each primer at a concentration of 0.2 muM, and 1 U of rTth DNA polymerase (GeneAmp XL PCR kit; Applied Biosystems, Foster City, Calif.). The mixture was overlaid with AmpliWax PCR Gem (Applied Biosystems) and heated for 1 min at 94C, followed by 35 cycles of 1 min at 94C and 2.5 min at 72C. A single 2,200-bp band corresponding to the amplified product was detected by ethidium bromide staining after electrophoresis on 0.9% (wt/vol) agarose. The PCR product was prepared for sequencing by treatment with shrimp alkaline phosphatase to dephosphorylate all remaining deoxynucleoside triphosphates and with exonuclease I to degrade all residual single-stranded DNA present in the PCR mixture. The entire katG gene was sequenced with the ABI PRISM Big Dye Terminator Cycle Sequencing Ready Reaction kit with AmpliTaq DNA polymerase on an ABI PRISM 377 DNA sequencer with XL Upgrade and 96-well Upgrade (Perkin-Elmer Applied Biosystems, Foster City, Calif.) by using eight primers whose sequences were specific for sequences that spanned the length of the gene (the primers and their sequences are presented in the Appendix). The sequence data assembled with the Sequencher program (Gene Codes Inc.) confirmed the presence of the S315T mutation and the absence of secondary mutations. Susceptibility testing. | The susceptibilities of M. tuberculosis H37Rv and TBC3 to INH alone and INH in combination with plumbagin and clofazimine were determined by using the radiometric criteria of the BACTEC 460 system (Becton Dickinson, Sparks, Md.) . All procedures were carried out in a biological safety cabinet inside a biosafety level 3 biocontainment facility. Studies involving combinations of drugs were carried out with weakly inhibitory concentrations of clofazimine and plumbagin, as determined by use of the radiometric criteria. According to the instructions of the manufacturer (S. H. Siddiqi, BACTEC 460TB system, product and procedure manual, 1995, Becton Dickinson, Sparks, Md.), a suspension of an actively growing culture was made in water, and 0.1 ml was transferred to a BACTEC 12B bottle. These bottles were incubated until the appropriate growth index (GI) was obtained, after which 0.1-ml aliquots were removed and transferred to fresh BACTEC 12B bottles containing drugs. The bottles were incubated at 37C and analyzed daily until the GI of the growth control diluted 1:100 was >=30. Resistance was determined by comparing the change in GI over that of the previous day (DeltaGI) between the control vial (GIcontrol) and the vials containing drugs (GIdrugs). The results are interpreted as follows: if DeltaGIcontrol is >DeltaGIdrug, the strain is referred to as "susceptible"; if DeltaGIcontrol is =30, and corresponds to the daily concentration that resulted in >99% inhibition of the bacterial population growth. The IC50 corresponds to the concentration of drug that causes a 50% decrease in the GI. The IC50s of INH were obtained by least-squares fitting of the data to the Langmuir isotherm equation: B = 1/[1 + (IC50/[INH])], where B is equal to percent GI/100%. The percent reduction of the IC50 of INH in the presence of clofazimine and plumbagin is determined by the following equation: {[IC50(INH) - IC50(X)]/IC50(INH)} x 100, where IC50(INH) and IC50(X) represent the IC50s of INH in the absence and presence of plumbagin or clofazimine, respectively. The relative error was calculated by applying the basic theory of error: [IC50(X)/IC50(INH)] x {[DeltaIC50(INH)/IC50(INH)] + [DeltaIC50(X)/IC50(X)]}, where DeltaIC50(INH) represents the error of determination of the IC50 of INH, and DeltaIC50(X) is the error of determination of the IC50 of INH in the presence of plumbagin or clofazimine. RESULTS : The effect of the superoxide-generating compounds plumbagin and clofazimine on the activity of INH against M. tuberculosis strains H37Rv and TBC3 was assessed by use of the BACTEC radiometric criteria. DeltaGI values for cultures of these strains in the presence of INH alone were determined. The range of INH concentrations chosen allows accurate determination of the MIC of INH for a strain with wild-type katG (strain H37Rv) and a strain with katG harboring the S315T mutation (strain TBC3) . The MIC of INH for TBC3 was 5 mug/ml, 200-fold higher than the MIC for H37Rv (0.025 mug/ml), confirming the efficacy of the S315T mutation in conferring resistance to INH . It should be noted that TBC3 is not an isogenic mutant of H37Rv. Therefore, genotypic differences other than the S315T mutation that could contribute to the increased MIC cannot be excluded. The susceptibilities of H37Rv and TBC3 to clofazimine and plumbagin alone were determined as well . For both strains, the MIC of clofazimine was >0.1 mug/ml but <=0.2 mug/ml, but a decrease in GI on day 7 was observed with clofazimine at concentrations >=0.1 mug/ml. This is in agreement with previous studies, which showed a MIC of 0.12 mug/ml . A concentration of 0.1 mug of clofazimine per ml was therefore chosen as the subinhibitory concentration to be used in combination with INH (see below). Plumbagin also caused a decline in the GI values for both strains, with noticeable decreases in the GI values at 10 to 20 muM. TBC3 appeared to be slightly more sensitive to plumbagin, with a slight (~30%) decline in the GI value by day 7 in the presence of 10 muM plumbagin, whereas for H37Rv, 20 muM plumbagin led to a similar, modest decrease in the GI value . As a result, the subinhibitory concentrations of plumbagin were chosen to be 20 muM for H37Rv and 10 muM for TBC3. The influences of subinhibitory concentrations of clofazimine and plumbagin on susceptibility to INH were assessed by analysis with the BACTEC system. Both compounds led to significant increases in the susceptibility of wild-type strain H37Rv to INH (Fig. ; Table ). The MIC of INH decreased from 0.025 mug/ml (with INH alone) to 0.012 mug/ml in the presence of 0.1 mug of clofazimine per ml and to 0.008 mug/ml in the presence of 20 muM plumbagin . Susceptibility was also assessed by estimating the IC50 of INH . The IC50, being the concentration of drug that causes a 50% decrease in GI, is a more sensitive measure of the effects of these agents than the MIC, which is the lowest concentration of drug that results in a >99% decrease in cell growth. IC50s were determined by a least-squares fit of the data presented in Fig. , as described in Materials and Methods. The IC50 of INH was found to be 0.015 +- 0.007 mug/ml (Fig. , top), decreasing twofold (to 0.0068 +- 0.0015 mug/ml) in the presence of 20 muM plumbagin (Fig. , middle) and decreasing about 30% (to 0.0099 +- 0.0026 mug/ml) in the presence of 0.1 mug of clofazimine per ml (Fig. , bottom). The activity of INH alone or in combination with plumbagin (10 muM) or clofazimine (0.1 mug/ml) against TBC3 was also investigated (Table ; Fig. ). Both the MIC and the IC50 of INH also appeared to decrease in the presence of these compounds. With INH alone, the IC50 was 2.1 +- 0.7 mug/ml, whereas in the presence of plumbagin or clofazimine, the IC50s were 1.3 +- 0.5 and 1.5 +- 1.2 mug/ml, respectively. Compared to the reductions in the IC50s of INH in combination with plumbagin or clofazimine for strain H37Rv (55 and 34%, respectively), the reductions in the IC50s for TBC3 were lower. For TBC3, the corresponding decreases were 38 and 28% . FIG. 1. | INH susceptibilities of M. tuberculosis INH susceptibilities of M. tuberculosis (Mtb) strains H37Rv (A) and TBC3 (B) using BACTEC radiometric analysis. The control reaction represents that in a vial containing a starter culture diluted 1:100 relative to the inoculum in the vials containing INH. The asterisks indicate the MICs of INH for each strain, as defined in Materials and Methods. GI data were obtained on a daily basis following inoculation of BACTEC vials, but data for only selective days are shown for clarity. FIG. 2. | Susceptibilities of M. tuberculosis Susceptibilities of M. tuberculosis (Mtb) H37Rv (A) and TBC3 (B) to plumbagin and clofazimine. FIG. 3. | Effects of plumbagin and clofazimine on INH susceptibility. Effects of plumbagin and clofazimine on INH susceptibility. The percent GI is plotted versus the log[INH] (p[INH]), expressed in units of nanograms per milliliter, alone (top) or in the presence of plumbagin (middle) or clofazimine (bottom), for M. tuberculosis H37Rv (A) and TBC3 (B). Weakly inhibitory concentration of clofazimine (0.1 mug/ml) and plumbagin (20 muM in panels A, 10 muM in panels B) were included in each vial. IC50s were calculated for each graph as the concentration of INH resulting in a 50% decrease in the GI value relative to that for the control vials, as described in Materials and Methods. The vertical dotted line in each set of panels represents the IC50 of INH alone (0.015 mug/ml in panels A, 2.1 mug/ml in panels B). (C) Reduction of IC50 of INH for H37Rv and TBC3 by plumbagin and clofazimine. The percent GI is expressed as the average value +- the standard error for three experiments. TABLE 1 | MICs of INH used alone and in combination with plumbagin or clofazimine for M. tuberculosis strains H37Rv and TBC3 DISCUSSION : Although a role for KatG in mediating INH activation is firmly established, the mechanism of activation and the chemical nature of the activated product are still unresolved. Initially described as a catalase-peroxidase on the basis of the homology of its amino acid sequence to that of the Escherichia coli hydroperoxidase I enzyme , the enzyme has subsequently been shown to exhibit a number of different catalytic activities. Three of these, the catalase, peroxidase, and Mn2+-dependent peroxidase activities, share a common mechanism in which a hydroperoxide substrate (e.g., H2O2 and tert-butylhydroperoxide) oxidizes the Fe3+ form of the enzyme and forms an intermediate, compound I, which is 2 equivalents more oxidized than the resting enzyme. Compound I, in turn, oxidizes a second substrate (e.g., INH), returning the enzyme to the resting ferric state. The initial hypothesis that the S315T mutation might confer resistance to M. tuberculosis by affecting the catalase-peroxidase mechanism was tested by comparing these activities for recombinant forms of the wild-type and S315T enzymes . These studies demonstrated that the catalytic efficiency of KatG(S315T) is comparable to or only slightly lower than the catalytic efficiency of the wild-type enzyme for all three activities and does not correlate with the ca. 200-fold difference in the MICs of INH for strains carrying wild-type versus S315T katG alleles, as measured in this and other studies . It has been shown that KatG-mediated INH activation is probably not by direct peroxidation, as the enzyme catalyzes INH oxidation in the absence of peroxide . An alternative mechanism for KatG-dependent INH turnover, in which dioxygen is used as the oxidant, has gained attention in recent years (-, ). The mechanism for this reaction is not completely understood but is thought to proceed in a manner analogous to those for cytochromes P450. In this reaction, dioxygen binds to the Fe2+ form of KatG to form an oxyferrous intermediate, which goes on to form a compound I-type species after an additional redox reaction (-, ). The oxyferrous enzyme is a resonance form of the ferric-superoxo enzyme. Consequently, the oxyferrous intermediate can also form by reaction of superoxide anion with the Fe3+ form of KatG . On the basis of the observation that KatG-dependent activation of INH generates reactive oxygen species (, -), it has been hypothesized that superoxide, formed via a one-electron reduction of dioxygen concomitant with INH oxidation, can bind to the ferric heme to form oxyferrous KatG and bring about further INH oxidation and activation. Support for the physiological relevance of this mechanism was provided by Wang et al. , who showed that the superoxide generator plumbagin resulted in an increase in the bacteriostatic activity of INH against M. smegmatis, a strain normally resistant to INH. Other supporting experiments have shown that KatG(S315T) has a reduced ability to oxidize INH to isonicotinic acid when superoxide is used as the oxidant . In this work, the role of superoxide anion in INH oxidation in vivo was investigated by radiometric analysis of INH susceptibility in the presence versus the absence of plumbagin and clofazimine. Plumbagin is a known redox cycling agent expected to increase intracellular superoxide concentrations . The precise mechanism of action of clofazimine is unknown, but it has been suggested that it generates intracellular hydrogen peroxide and superoxide . Clofazimine has been used for the treatment of leprosy and has excellent in vitro inhibitory activity against M. avium-M. intracellulare complex strains, with MICs that range from 0.1 to 5 mug/ml. The drug also has potent in vitro activity against M. tuberculosis, but there is little or no information on its in vivo activity . The M. tuberculosis strain carrying the wild-type katG allele, strain H37Rv, exhibited increased susceptibility to INH when INH was used in combination with either plumbagin or clofazimine, as evidenced by a decrease in both the IC50s and the MICs of INH in the presence of these reagents. Hence, a superoxide-dependent mechanism appears to be important for KatG-mediated INH activation in this strain. KatG(S315T) also appears to utilize superoxide to activate INH, as evidenced by the fact that INH turnover by the purified enzyme is completely abolished in the presence of a catalytic quantity of superoxide dismutase (SOD) . Indeed, although the potency of INH was increased by a putative increase in superoxide production by plumbagin and clofazimine in INH-resistant strain TBC3, our data indicate that the katG S315T mutation attenuates this activation. Thus, KatG-mediated INH oxidative activation is superoxide dependent in M. tuberculosis and the S315T mutation may confer an INH resistance phenotype through a reduced activity toward superoxide. Although the results implicate superoxide potentiation of INH toxicity in both strains, H37Rv appeared to be more sensitive to increased levels of superoxide than TBC3. The hypothesis that superoxide participates in the KatG-mediated INH oxidation is intriguing because M. tuberculosis contains both ferric and Cu,Zn SODs. Superoxide-dependent activation may be possible only because all of the Cu,Zn SOD and 76% of the ferric SOD are exported from the cell, presumably to act as extracellular defenses against the host respiratory burst . The potential chemical implications of these findings are intriguing. It is possible that clofazimine in combination with INH may result in synergistic activity in vitro against M. tuberculosis. This synergistic activity may be more for the wild-type M. tuberculosis strain than for strains harboring the katG S315T mutation. Further animal and human studies are required to validate these assumptions. FIG. 4. | The INH shunt, a proposed mechanism for INH oxidation by ferric superoxo-KatG. The INH shunt, a proposed mechanism for INH oxidation by ferric superoxo-KatG. : The following primers were used for sequencing: KatG 5', CCG ACA CTT CGC GAT CAC ATC CGT GAT CAC AGC CC; KatG 2, ATG ACC ACC TCG CAG CCG; KatG 3, GGC TTC GGC CGG GTC GAC; KatG 4, GTC GGC CCC GAA CCC GAG; KatG 5, CTG CGG GTG GAT CCG ATC; KatG 6, CGT GGT AGC GAC AAG CGC; KatG 7, TTT GCC GTG CTG GAG CCC; KatG 8, AGT GGC AAG GTG AAG TGG; and KatG 2' reverse primer, CCG TAG TCG GCG GGC CAC CAC GGC T. Backmatter: PMID- 12183260 TI - The Antifungal Echinocandin Caspofungin Acetate Kills Growing Cells of Aspergillus fumigatus In Vitro AB - Caspofungin acetate is an antifungal antibiotic that inhibits synthesis of 1,3-beta-d-glucan, an essential component of the fungal cell wall. While caspofungin causes cell death in yeasts and dimorphic fungi such as Candida albicans, its effect on Aspergillus fumigatus is less well understood. We used the fluorescent dyes 5,(6)-carboxyfluorescein diacetate (CFDA) and bis-(1,3-dibutylbarbituric acid) trimethine oxonol (DiBAC), which stain live and dead cells, respectively, to further characterize the antifungal activity of caspofungin. For comparison, compounds whose mode of action was either fungistatic (fluconazole, itraconazole) or fungicidal (amphotericin B) were also evaluated. A correlation between caspofungin-induced loss of viability, decreased CFDA staining, and increased DiBAC staining was established first with C. albicans. For A. fumigatus, caspofungin caused similar dye-staining changes, which were quantified by fluorimetric analysis of stained hyphae grown in a medium that promoted dispersed growth. The minimum concentration of caspofungin required to produce these changes also decreased the level of growth-dependent reduction of the indicator dye Alamar Blue. We observed a differential effect of caspofungin as a function of cell position: 88% of apical cells and 61% of subapical branching cells failed to stain with the viable dye CFDA, but only 24% of subapical cells were unstained. Complementary results were seen with germlings from DiBAC-stained, caspofungin-treated cultures. Extended incubation of A. fumigatus with a single dose of caspofungin affected the same proportion of apical and subapical branching cells for up to 72 h. The dye-staining patterns illustrate that the cells at the active centers for new cell wall synthesis within A. fumigatus hyphae are killed when they are exposed to caspofungin. Keywords: Introduction : Caspofungin, the first clinically used echinocandin, is a member of a new class of antifungal antibiotics that inhibit the synthesis of 1,3-beta-d-glucan. The echinocandins are potent against many pathogenic fungi while maintaining an excellent safety profile. Inhibition of cell wall synthesis represents a novel mechanism of action compared to those of existing antifungal agents, which target fungal cell membranes. Amphotericin B (AMB), a polyene, is fungicidal in vitro and in vivo against Candida albicans and Aspergillus fumigatus, while the azoles fluconazole (FLC) and itraconazole (ITC) demonstrate fungistatic activities against these organisms. The consequences of inhibiting yeast cell wall formation, specifically glucan synthesis, are cell morphology changes and loss of viability. Treatment of C. albicans with echinocandin caused growing cells to lyse , and caspofungin has been shown to be fungicidal against numerous C. albicans isolates . Ernst et al. reported that C. albicans cells treated with caspofungin at a level that inhibits growth by 80% (MIC80) were highly enlarged compared to control cells; when treated with caspofungin at the 100% inhibitory level (MIC100), all cells lysed. For filamentous fungi, inhibition with caspofungin results in profound changes in the growth, morphology, and cell wall structure of hyphae . Caspofungin and other echinocandins have been shown to inhibit 1,3-beta-d-glucan synthesis activity in the filamentous ascomycetes A. fumigatus and Aspergillus nidulans . The FKS gene encodes a subunit of the 1,3-beta-d-glucan synthase complex in several fungal species, including Saccharomyces cerevisiae , A. nidulans , C. albicans , and Cryptococcus neoformans . Beauvais et al. have shown that the A. fumigatus Fks protein localizes to the hyphal apex. Likewise, fluorescence is most intense at the tips of hyphae stained with a beta-glucan-specific fluorochrome (aniline blue). In A. nidulans, new cell wall formation occurs exclusively at the hyphal apices , and mutations affecting this process cause aberrant growth . Caspofungin acetate has been approved for use in the treatment of invasive aspergillosis in patients who are refractory to or intolerant of other therapies. Caspofungin has been shown to have efficacy in animal models of disseminated and pulmonary (; E. M. Bernard, T. Ishimaur, and D. Armstrong, Abstr. 36th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F39, 1996) aspergillosis, and liquid broth microdilution assays demonstrate growth reduction resulting from caspofungin treatment . However, it has not been clear that caspofungin can kill A. fumigatus cells. In both animal models and in vitro cultures, echinocandins often produce little to no reduction in the numbers of A. fumigatus CFU . Characterization of the nature of the activity of echinocandins against A. fumigatus is further confounded by the inability to quantify changes in cell mass or observe individual germ tubes in liquid culture due to the intertwined, filamentous growth of the organism. A recent report described the use of fluorescent probes to detect the fungicidal activity of AMB against C. albicans . At concentrations of AMB that significantly reduce the number of CFU, the level of fluorescent staining with the dye 5,(6)-carboxyfluorescein diacetate (CFDA), which stains live cells, decreased, while the level of staining with the dye bis-(1,3-dibutylbarbituric acid) trimethine oxonol (DiBAC), which stains dead cells, increased. The esterase activity of viable cells cleaves the acetate moieties from CFDA to produce free carboxyfluorescein, which is retained in cells whose membranes are intact . DiBAC is a member of the oxonol class of dyes that exhibit intense fluorescence upon binding to phospholipids . Normal membrane potential, which collapses with mortal injury or cell death, prevents access of DiBAC to the inside of the cell. We sought to further understand the activity of caspofungin against A. fumigatus using these dyes. Our initial experiments performed with C. albicans established a strong correlation between caspofungin-induced changes in dye staining and reductions in the number of viable CFU. With minor modifications, including addition of the polyacrylate Junlon PW110 (Junlon) to the medium to promote dispersed growth, we were able to stain germlings of A. fumigatus strain MF5668 with CFDA and DiBAC and detect fluorescence qualitatively by microscopy and quantitatively by liquid broth microtitrations. Our results demonstrate that caspofungin can cause lysis of A. fumigatus cells. Direct counting of stained caspofungin-treated germlings demonstrated a preferential effect of the drug on cells at the tips and branch points of growing A. fumigatus hyphae. (This work was presented in part at the 40th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Ontario, Canada, 19 September 2000.) MATERIALS AND METHODS : Drugs. | Caspofungin was dissolved in sterile distilled water. AMB (Sigma Chemical Co., St. Louis, Mo.), FLC (Pfizer, Groton, Conn.), and ITC (Janssen Pharmaceutica, Titusville, N.J.) were dissolved in 100% dimethyl sulfoxide. Media. | C. albicans cultures were grown in RPMI 1640 medium with l-glutamine buffered to pH 7.0 with 0.165 M MOPS [3-(N-morpholino)propanesulfonic acid; BioWhittaker, Boston, Mass.]. A. fumigatus cultures were grown in RPMI 1640 medium containing MOPS (pH 7.0) and 0.15% (wt/vol) Junlon (Nihon Junyaku Co., Ltd., Tokyo, Japan). Briefly, powdered RPMI 1640 with l-glutamine (Gibco BRL, Bethesda, Md.) was resuspended in distilled H2O, MOPS (Fisher Scientific) was added to a final concentration of 0.165 M, the pH was adjusted to 7.0, the volume was brought to 800 ml, and the medium was sterile filtered. To prepare Junlon, 1.5 g was added to 200 ml of distilled H2O and heated in a microwave until no discernible clumps were seen. The Junlon was autoclaved and cooled to room temperature before the RPMI 1640 medium was added. Organisms and preparation of inocula. | C. albicans strain MY1055 from the culture collection of Merck (Rahway, N.J.) was spread onto YPAD (1% yeast extract, 2% peptone, 2% dextrose, 4 mg of adenine per ml) plates (Difco, Detroit, Mich.), and the plates were incubated for 2 to 4 days at 35C. A. fumigatus strain MF5668 (ATCC 13073) is a clinical isolate from a human pulmonary lesion . Conidia were inoculated on Sabouraud dextrose agar (SDA) slants (BBL, Cockeysville, Md.) and grown at 37C, with the caps ajar, for 5 to 7 days to allow sporulation. Five milliliters of phosphate-buffered saline (Gibco BRL) containing 0.05% (vol/vol) Tween 20 was added, and the conidia were harvested by gently rubbing the surface with a sterile cotton swab and decanting the liquid. A stock of 1 x 106 to 5 x 106 conidia/ml was prepared by adding drops of the concentrated spore suspension to 5 ml of sterile saline (Becton Dickinson) until the optical density (OD), read in a MicroScan turbidity meter (Dade Behring, West Sacramento, Calif.), equaled that of a 0.5 McFarland standard (Remel, Lenexa, Kans.). Cell growth and drug treatment. | Several colonies of C. albicans strain MY1055 were inoculated from YPAD plates into 20 ml of buffered RPMI 1640 medium and grown to saturation (ca. 40 h) at 37C with shaking (220 rpm). The cells were subcultured to an initial A600 of 0.002 in 65 ml of fresh medium, and incubation was continued until the cultures reached a density of 1x 106 to 2 x 106 cells/ml (determined by counting with a hemocytometer; ca. 9 h). Aliquots were removed, serially diluted in buffered RPMI 1640 medium, and spread onto SDA plates (BBL) to determine the number of CFU per milliliter. Compound (caspofungin, AMB, or FLC) or vehicle was added, and incubation was continued for 15 h. Following treatment, samples from each culture were evaluated for the number of CFU per milliliter or the total cell count. Cells from each 65-ml culture were collected by centrifugation (10,000 x g, 10 min, 4C) and brought to 6.5 ml with buffered RPMI 1640 medium; 1.5-ml aliquots were transferred to microcentrifuge tubes, and cells were collected by centrifugation (9,500 x g, 5 min, room temperature). For DiBAC staining, cells were resuspended in 1.5 ml of 0.1 M MOPS at pH 7.0 (MOPS 7). For CFDA staining, cells were washed two times with MOPS 7 and resuspended in 1.5 ml of 0.1 M MOPS-50 mM citric acid at pH 3.0 (MOPS 3). A. fumigatus MF5668 conidia in sterile saline suspensions were counted with a hemocytometer and inoculated into 80 ml of buffered RPMI 1640 medium containing 0.15% (wt/vol) Junlon (RPMI-Junlon) to an initial density of 105 conidia/ml. The culture was incubated for 14 h at 37C with shaking (220 rpm). Caspofungin, AMB, ITC, or vehicle was then added, and incubation was continued. To estimate growth inhibition, RPMI-Junlon cultures (20 ml) containing 10% (vol/vol) Alamar Blue (Trek Diagnostics, Westlake, Ohio) were evaluated in parallel. Cell-free supernatants were prepared from 1-ml aliquots by centrifugation (3,200 x g, 5 min, 22C), the absorbance at 603 nm was measured, and the Alamar Blue reduction index was calculated from the formula OD603 untreated/OD603 treated. For staining with fluorescent probes, duplicate 15-ml aliquots from each RPMI-Junlon culture were collected by centrifugation (3,200 x g, 5 min, 4C), which produced three phases: a supernatant containing no Junlon and no detectable hyphae, a Junlon layer containing a high density of trapped germlings, and a pellet containing larger clumps of hyphae. Analysis of total protein (see below) in the three layers illustrated that 60% of the total protein was in the pellet, with the remaining 40% in the Junlon layer (data not shown). The Junlon layer was transferred to a new tube, and buffer (MOPS 3 for CFDA staining, MOPS 7 for DiBAC staining) was added to a final volume of 5 ml. Staining with fluorescent probes. | CFDA staining was performed with cell suspensions in MOPS 3 by adding CFDA (Molecular Probes, Eugene, Oreg.) from a 5-mg/ml stock in 100% dimethyl sulfoxide to a final concentration of 50 mug/ml. Samples were incubated in the dark at 37C for 45 min with gentle agitation and then stored on ice until analysis. For DiBAC staining, cell suspensions in MOPS 7 were used. DiBAC (Molecular Probes) from a 1-mg/ml stock in 100% ethanol was added to a final concentration of 2 mug/ml; samples were incubated in the dark with shaking at room temperature (ca. 22C) for 1 h, washed twice with MOPS 7, and stored on ice. Washing of C. albicans samples (1.5-ml aliquots) was performed by centrifugation at 9,500 x g for 5 min; A. fumigatus germlings in the Junlon layer were collected and washed by centrifugation at 3,200 x g for 10 min. To account for background fluorescence in DiBAC-stained A. fumigatus samples, an aliquot of uninoculated RPMI-Junlon medium was processed in parallel and the fluorescence value was subtracted. Fluorometry. | Fluorescence was measured with a Fluoroskan II spectrofluorometer (Labsystems, Helsinki, Finland) by using a fluorescein isothiocyanate (FITC) filter pair (excitation wavelength = 485 nm; emission wavelength = 538 nm) and 96-well Optiplates (Packard, Meriden, Conn.). Samples of each stained cell suspension (400 mul) were added in triplicate to wells of the first Optiplate row and were serially diluted twofold to the last row of the plate in 200 mul of either MOPS 7 (DiBAC-stained samples) or MOPS 3 (CFDA-stained samples). Normalization. | C. albicans fluorescence values were normalized to cell density. Plots of fluorescence versus cell density yielded curves whose values for CFDA fluorescence per 106 cells or for DiBAC fluorescence per 107 cells were near the middle of the linear portion of the curve. The cells in an aliquot of each CFDA- or DiBAC-stained C. albicans culture were counted in a hemocytometer, and the relative fluorescence per 106 cells (CFDA) or 107 cells (DiBAC) were recorded. For A. fumigatus cultures, fluorescence values were normalized to the amount of cellular protein. To determine protein concentration, stained germlings were disrupted with 0.5-mm-diameter glass beads (BioSpec Products, Bartlesville, Okla.) with a Bead-Beater homogenizer (BioSpec Products). Samples (1 ml) from each culture were dispensed into 2-ml screw-cap tubes (Sarstedt, Numbrecht, Germany), glass beads were added to fill the tube, and germlings were broken with three 30-s bursts at 5,000 rpm. The homogenates were collected by centrifugation at 800 x g for 5 min at 4C, and the protein concentration was determined with MicroBCA reagents (Pierce, Rockford, Ill.), with a standard of bovine serum albumin, according to the recommendation of the manufacturer. Fluorescence values were normalized per 1 mg of protein per ml for both dyes. A cell staining index was used to measure the fold change in fluorescence due to drug treatment. Normalized fluorescence values from drug-free cultures of C. albicans or A. fumigatus, stained with either CFDA or DiBAC, were compared with values obtained from cultures incubated with drug. The cell staining index was calculated from the formula normalized fluorescencetreated/normalized fluorescenceuntreated. Microscopy. | Photomicrographs of cells were taken with a Zeiss Axiophot microscope equipped for both Nomarski optics and epifluorescence (with an FITC filter). Magnifications ranged from x400 to x2,000, and the film speed was either ASA 400 or ASA 1600. The exposure time for light micrographs was determined by the camera controller; for fluorescent micrographs, the exposure time was determined by the camera controller or manually, as noted where appropriate. In vitro susceptibility. | A suspension of A. fumigatus MF5668 conidia was prepared as described above and was adjusted to an OD equal to that of the 0.5 McFarland standard by using a MicroScan turbidity meter. A concentration of approximately 1 x 106 to 5 x 106 conidia/ml was verified by counting with a hemocytometer. Susceptibility testing was performed by NCCLS protocol M38-P . Cell counting. | Photomicrograph slides were scanned, and two-panel images, consisting of identical light and fluorescent micrographs of the same field, were printed at 600 pixels/in. The background color was subtracted from the fluorescent images so that only the green color of the fluorescent cells was visible; these panels were printed on transparencies and laid over the corresponding Nomarski image to ensure that the cells were appropriately identified as fluorescent or nonfluorescent. Septa between cells were identified on the basis of a published description . Cells were scored as apical (tip), subapical, or subapical branching. Cells whose boundaries were outside the field were not included in the totals. A single person performed all counting to reduce potential differences in the definitions of septa and cell position. RESULTS : We sought to better understand the physiological effects of caspofungin treatment using CFDA and DiBAC, which are fluorescent indicators of cell viability . First, C. albicans strain MY1055 was grown in buffered RPMI 1640 medium and challenged with caspofungin to determine the appropriate conditions and timing for cell killing. Exposure of cells in the early logarithmic phase to 0.25 mug of caspofungin per ml reduced the number of CFU by 2 log10 in 15 h (data not shown). Cultures grown under these conditions were incubated with 0.15 mug of AMB per ml, 0.2 mug of caspofungin per ml, or 2.5 mug of FLC per ml or without drug. The cells were stained with either CFDA or DiBAC and were examined microscopically . Nearly all cells from drug-free cultures stained with CFDA but not with DiBAC. Conversely, very few cells from AMB- or caspofungin-treated cultures stained with CFDA, while nearly all cells stained with DiBAC. C. albicans cells incubated with FLC showed a unique staining pattern; CFDA fluorescence was uniform but less intense than that seen for untreated cells, and only a few FLC-treated cells stained faintly with DiBAC. The effect of AMB, caspofungin, or FLC on dye staining of live or dead cells was quantified by fluorometry. Because fluorescence intensity is a function of both the physiological state of the cells and the culture density, we normalized each signal to the cell number determined by counting the cells with a hemocytometer. Titrations were performed with AMB, FLC, or caspofungin, and cells from drug-treated cultures were either stained with CFDA or DiBAC or spread onto agar plates to quantify the CFU . At AMB concentrations >=0.15 mug/ml, the number of CFU was reduced and there was a decrease in the amount of CFDA fluorescence per cell and a nearly 70-fold increase in the level of normalized DiBAC staining. Similarly, concentrations of caspofungin >=0.04 mug/ml reduced the number of C. albicans CFU and caused a 4-fold decrease in the level of CFDA staining and a greater than 140-fold increase in the level of DiBAC staining compared to those for the controls. Concentrations of caspofungin or AMB that did not reduce the CFU counts also did not produce changes in staining with either dye. Exposure of C. albicans cells to FLC at concentrations >=0.5 mug/ml caused a decrease in the level of CFDA staining, which we attribute to decreased fluorescence intensity in all cells rather than the all-or-none staining evident in caspofungin- and AMB-treated cultures . There also was a <10-fold increase in the level of DiBAC staining of FLC-treated cells, but there was no reduction in the number of CFU at any concentration of FLC tested. Having established that staining with CFDA and DiBAC reflects a loss of viability of caspofungin-treated C. albicans cells, we adapted this method to further investigate the activity of caspofungin against A. fumigatus. When A. fumigatus is grown in buffered RPMI 1640 medium, it forms tangled hyphal masses; aliquots removed by standard methods are not representative of the total culture. We modified the medium by adding the polyacrylate Junlon to a final concentration of 0.15% (wt/vol), which reduced mycelial masses and produced a culture of small, punctate clumps within a turbid suspension of individual germlings . The presence of Junlon did not have a deleterious effect on the amount of fungal growth, as evidenced by equivalent biomass (measured by determination of the total amount of protein) in cultures grown with or without the polyacrylate (data not shown). Samples of CFDA-stained cultures of A. fumigatus, grown in medium with or without Junlon, were serially diluted in the same medium, and the fluorescence was measured. The linear titration evident in the presence of Junlon contrasted with the scattered, nonlinear fluorescence titration observed with cultures grown in RPMI 1640 medium alone. Individual hyphae of A. fumigatus had a similar appearance when grown in medium with or without Junlon (data not shown), and the addition of Junlon did not affect the susceptibility of A. fumigatus MF5668 to caspofungin, AMB, or ITC in standard liquid broth microdilution assays . We determined the effect of drug treatment on A. fumigatus hyphal morphology, cell staining, and CFU counts. Spores of A. fumigatus strain MF5668 inoculated into RPMI-Junlon and incubated with shaking at 37C germinated and grew into small germlings within 14 h. Compound (either 0.15 mug of AMB per ml, 0.3 mug of caspofungin per ml, or 2.5 mug of ITC per ml) or vehicle was added to the actively growing cultures at this point. After 6 h of incubation with drug or vehicle, aliquots of the cultures were stained with either CFDA or DiBAC and examined microscopically. As with untreated C. albicans cultures, the vast majority of untreated A. fumigatus cells stained with CFDA but not with DiBAC . In contrast, none of the AMB-treated A. fumigatus cells was stained by CFDA, but all AMB-treated A. fumigatus cells were stained by DiBAC. ITC treatment produced a forked or club-like morphology among a small number of apical cells, some of which failed to stain with CFDA or stained with DiBAC. Of the CFDA-positive ITC-treated cells, there was a trend toward decreased fluorescence of all cells, similar to the findings for FLC-treated C. albicans. Caspofungin-treated germlings had a high proportion of cells at hyphal tips and branch sites that did not stain with CFDA or that did stain with DiBAC . Inspection of caspofungin-treated, DiBAC-stained hyphae at a higher magnification revealed plumes of stained material apparently outside of transparent cells at the tips . This is most likely cellular phospholipid debris which was trapped in the viscous Junlon-containing medium after individual cells lysed and released their contents. The high prevalence of damaged cells at hyphal tips and the branch points of caspofungin-treated germlings was not reflected by a change in the number of CFU . A. fumigatus germlings in RPMI-Junlon were exposed to 0.3 mug of caspofungin, AMB, or ITC per ml or vehicle and were monitored for the number of CFU during a 15-h incubation. AMB treatment produced an similar2 log10 reduction in the number of A. fumigatus CFU within 3 h, and ITC or caspofungin treatment did not affect the number of CFU. Cultures grown in the absence of drug displayed obvious increases in hyphal mass without a commensurate increase in the number of CFU. The dose dependence of drug-induced changes in A. fumigatus staining with CFDA or DiBAC was evaluated by titrating each compound and measuring the fluorescence quantitatively. To relate any changes in staining to cell growth and viability, parallel cultures were monitored for the reduction of Alamar Blue in the presence or the absence of drug. A change in the color of the Alamar Blue dye results from the production of reducing equivalents during fungal growth; there is little to no change in dye color if cells die or stop growing. AMB-treated A. fumigatus cells exhibited a threshold for increased DiBAC staining and decreased CFDA staining at AMB concentrations >=0.15 mug/ml . This threshold correlated with the reduction of Alamar Blue and was consistent with the staining patterns observed in photomicrographs . Caspofungin-treated A. fumigatus cultures exhibited a continuum of normalized fluorescence changes and Alamar Blue reduction across a broad range of concentrations, with no clear threshold. A. fumigatus treated with 0.012 or 0.06 mug of caspofungin per ml displayed moderate DiBAC staining; the maximal response was at caspofungin concentrations >=0.3 mug/ml. This correlated with a decreased level of Alamar Blue reduction. ITC at concentrations >=0.5 mug/ml prevented Alamar Blue reduction, with little quantitative change in fluorescent dye staining. At these concentrations, ITC appears to stop the production of reducing equivalents by A. fumigatus while minimally inducing the physiological changes reflected by changes in CFDA or DiBAC staining. The preponderance of staining changes among cells at the tips and branch points of caspofungin-treated A. fumigatus germlings prompted an analysis of the effects of caspofungin as a function of cell position. A. fumigatus cells in photomicrographs were categorized as either apical, subapical, or subapical branched and were scored as fluorescent or nonfluorescent. As seen in Fig. and Table , AMB-treated A. fumigatus cells failed to stain with CFDA and stained with DiBAC, regardless of their position in the germ tube. Among caspofungin-treated germlings, only 12% of apical cells stained with CFDA, and 72% stained with DiBAC, while 39% of subapical branching cells stained with CFDA and 56% stained with DiBAC . For subapical cells within caspofungin-treated germlings, 76% stained with CFDA and 10% stained with DiBAC. ITC treatment led to a staining pattern that was intermediate in its cell specificity; there was a modest preference for apical and subapical branching cells that was not as pronounced as the selective staining pattern seen with caspofungin-treated germlings. We explored the effect of extended incubation in the presence of each antifungal drug on dye staining. Aliquots removed at 6, 24, 48, and 72 h after drug addition from cultures grown in RPMI-Junlon were stained and evaluated by microscopy, and parallel cultures were grown in the presence of Alamar Blue. There was substantial growth in the untreated culture during the course of the experiment (data not shown), and the CFDA-stained hyphae fluoresced strongly at all time points. There was a slight increase in DiBAC staining of untreated germlings with incubation for >=24 h, although this was primarily due to staining of conidia and conidium-bearing structures . AMB at concentrations >=0.27 mug/ml caused changes in staining by CFDA and DiBAC, regardless of the cell position within the germlings, over the entire 72-h time course (data not shown). Of interest, 0.09 mug of AMB per ml had antifungal activity at 6 and 24 h, but there was significant growth by 48 h. This was confirmed both by staining with CFDA and DiBAC and by Alamar Blue reduction. At all time points, caspofungin preferentially affected actively growing cells at the tips and branch points of A. fumigatus hyphae, producing staining changes in roughly 75% of apical cells (Fig. and data not shown). The extent of hyphal branching became more pronounced with prolonged incubation, but the proportion of fluorescent apical and subapical branching cells was consistent at all time points. With as much as 72 h of incubation with caspofungin at concentrations >=0.06 mug/ml, there was no evidence of hyphae with the slender, tapered morphology of untreated A. fumigatus germlings. Exposure to ITC caused only subtle changes in the CFDA and DiBAC staining patterns at all concentrations tested, although there was growth inhibition and no detectable Alamar Blue reduction at ITC concentrations >=0.5 mug/ml (data not shown). Morphological changes associated with sporulation were observed with prolonged incubation in both drug-free and caspofungin-treated cultures. This phenomenon is distinct from microcycle conidiation (direct production of conidiophores without prior mycelial formation), but like microcycle conidiation, it was most likely a consequence of growth at 37C in shake flasks . Without drug, abundant conidiophores formed from submerged hyphae within 24 h, and their morphology resembled that of structures produced by aerial mycelia. In contrast, aberrant conidiophores formed by caspofungin-treated A. fumigatus appeared only after 72 h, they were much less prevalent, and they appeared transparent and flattened . At least one-half failed to stain with CFDA or stained with DiBAC, and none appeared to produce any mature conidia. FIG. 1. | Photomicrographs of untreated and drug-treated C. albicans Photomicrographs of untreated and drug-treated C. albicans stained with the fluorescent dyes CFDA and DiBAC. C. albicans cells from an untreated culture or from cultures treated with AMB (0.15 mug/ml), caspofungin (0.2 mug/ml), or FLC (2.5 mug/ml) were harvested after 15 h, stained with the dye CFDA, which stains viable cells, or the dye DiBAC, which stains nonviable cells, and photographed at x1,000 magnification. Exposure times for epifluorescence of untreated samples stained with DiBAC, AMB-treated samples stained with CFDA, FLC-treated samples stained with CFDA, and FLC-treated samples stained with DiBAC were 20 s; all others were less than 1 s (as determined by the camera controller). Left panels, Nomarski optics; right panels, epifluorescence with an FITC filter set. FIG. 2. | Dose-dependent changes in normalized CFDA or DiBAC fluorescence of C. albicans Dose-dependent changes in normalized CFDA or DiBAC fluorescence of C. albicans. Cultures were challenged with serial dilutions of AMB (A), caspofungin (B), or FLC (C). Aliquots of these cultures were stained with CFDA or DiBAC or were spread onto agar plates to determine the number of CFU. Fluorescence values were normalized per 106 cells (CFDA) or 107 cells (DiBAC), as described in Materials and Methods. FIG. 3. | In vitro growth of A. fumigatus In vitro growth of A. fumigatus in RPMI-Junlon. (A) Tubes containing RPMI 1640 medium (the two leftmost tubes [tubes 1 and 2]) or RPMI-Junlon (the two rightmost tubes [tubes 3 and 4]) were either inoculated with 105A. fumigatus conidia (tubes 2 and 4) or left uninoculated (tubes 1 and 3) and were incubated for 15 h at 37C. The tubes were swirled briefly prior to being photographed. (B) A. fumigatus conidia (105) were grown in either RPMI-Junlon or RPMI 1640 medium alone , incubated for 24 h at 37C, and stained with CFDA. Stained samples were serially diluted twofold across a 96-well plate, and the fluorescence was read (excitation wavelength = 485 nm, emission wavelength = 538 nm). FIG.4. | Photomicrographs of untreated and drug-treated A. fumigatus Photomicrographs of untreated and drug-treated A. fumigatus stained with the fluorescent dyes CFDA and DiBAC. A. fumigatus germlings incubated for 6 h with 0.15 mug of AMB per ml, 0.3 mug of caspofungin per ml, or 2.5 mug of ITC per ml or without drug were stained with the dye CFDA, which stains viable cells, or the dye DiBAC, which stains nonviable cells, and photographed at x400 magnification. Fluorescent micrograph exposure times for untreated samples stained with DiBAC, AMB-treated samples stained with CFDA, and ITC-treated samples stained with CFDA or DiBAC were 20 s; all others were less than 1 s (as determined by the camera controller). Left panels, Nomarski optics; right panels, epifluorescence with an FITC filter set. FIG. 5. | High-magnification photomicrographs of caspofungin-treated, DiBAC-stained A. fumigatus High-magnification photomicrographs of caspofungin-treated, DiBAC-stained A. fumigatus. A. fumigatus germlings incubated for 6 h with 0.3 mug of caspofungin per ml were stained with DiBAC and photographed at x2,000 magnification. Images from two separate germlings from the same culture are shown. Left panels, Nomarski optics; right panels, epifluorescence with an FITC filter set. FIG. 6. | A. A. fumigatus killing curves. A. fumigatus conidia (105) were inoculated into RPMI-Junlon and incubated at 37C for 14 h before 0.3 mug of drug per ml (AMB , caspofungin , or ITC ) or vehicle was added. At 3-h intervals after drug addition, aliquots from each culture were spread onto agar plates for determination of the number of CFU. FIG. 7. | Quantitative fluorescence of dye-stained A. fumigatus. A. fumigatus Quantitative fluorescence of dye-stained A. fumigatus. A. fumigatus cultures were grown as described in the text and incubated with serial dilutions of AMB (A), caspofungin (B), or ITC (C) for 6 h. Aliquots of these cultures were stained with CFDA or DiBAC; parallel titrations were performed with cultures containing 10% (vol/vol) Alamar Blue. The values for CFDA and DiBAC staining were normalized to the total amount of protein, and the fold change relative to the amount for the untreated control (staining index) was determined as described in Materials and Methods. FIG. 8. | Classification scheme for cells within A. fumigatus Classification scheme for cells within A. fumigatus germlings. An A. fumigatus culture was treated with 0.3 mug of caspofungin per ml, stained with the viable dye CFDA, and photographed at x1,600 magnification. Left panel, Nomarski optics; right panel, epifluorescence with an FITC filter set plus bright-field illumination. Septa (indicated by arrows) were identified as described in the text, and the cells were classified as apical (e.g., cells 1 and 6), subapical (e.g., cells 3 and 5), or subapical branching (e.g., cells 2 and 4). FIG. 9. | Effect of extended drug treatment on A. fumigatus Effect of extended drug treatment on A. fumigatus fluorescent dye staining. (A) Untreated, DiBAC-stained germlings harvested 24 h after vehicle addition. Magnification, x400. Arrows indicate conidiophores. Caspofungin-treated germlings (0.3 mug/ml) harvested 72 h after drug addition were stained with CFDA (B) or DiBAC (C), and photographed at x800 magnification. In panels B and C, arrows with single tails indicate lysed apical cells and arrows with double tails indicate lysed conidiophores. The fluorescent micrograph in panel A required a 20-s exposure. Left panels, Nomarski optics; right panels, epifluorescence with an FITC filter set. TABLE 1 | In vitro susceptibility of A. fumigatus MF5668 in the absence or presence of Junlon TABLE 2 | Drug-induced staining changes in A. fumigatus as a function of cell type DISCUSSION : The effects of antifungal compounds on fungal pathogens are frequently tested by spreading samples from either treated cultures or organs from infected animals on agar plates and enumerating the CFU. However, for A. fumigatus, the filamentous nature of growth in both liquid culture and infected tissues makes it difficult to assess activity based on the number of CFU. Increases in cell mass are not usually reflected by increases in the number of CFU , and a CFU by definition reflects the viability only of the entire germling, not of individual cells within the germling. Because of the limitations of CFU measurements for A. fumigatus, alternative methods have been developed which utilize dye staining to assess various aspects of cell viability, most notably, membrane potential and intracellular enzymatic activity. Examples of these stains include tetrazolium dyes and the halogenated dye FUN-1 . Another dye, propidium iodide, has been used in flow cytometric analysis of C. albicans and C. neoformans , but its use with A. fumigatus to date has been limited to conidia . A previous report described the use of CFDA and DiBAC to characterize the response of C. albicans to AMB. Among other things, these dyes have been used to study the membrane potentials of mammalian cells and the viabilities of both bacteria and S. cerevisiae . By using at least two different dyes, which provide information about independent aspects of cell physiology, drugs with distinct mechanisms of action can be compared. It is important to relate preliminary dye-staining results to other viability measurements, including CFU quantitation or Alamar Blue reduction, to ensure that the staining changes correspond to true changes in cell vitality. The mycelial masses which form as A. fumigatus grows in liquid culture confounded our ability to withdraw representative samples or to focus on a field of cells under the microscope. We found that the addition of the polyacrylate Junlon promoted a more homogeneous culture of A. fumigatus in shaking flasks. After drug treatment and dye staining, a linear titration was obtained when germlings grown in RPMI-Junlon were diluted serially and assayed for fluorescence. Junlon has been used to force dispersed growth of other organisms, including Streptomyces spp. and Aspergillus niger . Additionally, by selecting germlings trapped in the Junlon layer, we focused on a form of A. fumigatus which not only fit entirely within a single microscope field but which was also representative of the total culture. The ability to photograph entire germlings and discriminate individual cells within these germlings was especially useful for assessing the preferential lysis of growing A. fumigatus cells by caspofungin. The abundance of 1,3-beta-d-glucan in cell walls formed during different stages of the A. fumigatus life cycle is not well characterized. During vegetative growth, the focus of new cell wall synthesis is the hyphal apex , and inhibition of 1,3-beta-d-glucan synthesis has profound effects on cell wall structure in A. fumigatus . As predicted from work with A. nidulans mutants, whose altered cell wall composition led to hyphal "balloons" which lysed without osmotic protection , we found that cells at the hyphal tips and branch points were more susceptible to lysis than subapical cells with mature cell walls. Even with prolonged incubation (up to 72 h), the focus of caspofungin killing was on apical and subapical branching cells. As aberrant, conidiophore-like structures formed in the presence of caspofungin, their staining pattern with CFDA or DiBAC was also consistent with cell death. Therefore, our results demonstrate that inhibition of 1,3-beta-d-glucan synthesis during either germ tube or conidiophore formation can be a lethal event. Understanding the effect of glucan synthesis inhibition as spores germinate, which is another key developmental step for A. fumigatus, will require further study. Because filamentous fungi grow by apical extension, continued killing of cells at the hyphal tips would be expected to have a profound impact on the growth of the organism in vivo. Petraitiene et al. reported that caspofungin treatment improved survival compared to that for the untreated controls in a rabbit model of pulmonary aspergillosis. This improvement in survival was at least as good as that seen with AMB therapy. Caspofungin treatment reduced several clinical measures of disease, including total lung weight, pulmonary infarcts, and pulmonary injury (determined by computed tomography), despite an increase in the number of CFU. In a recent report, caspofungin treatment yielded a reduced fungal burden (assessed by quantitative PCR) compared to that in untreated animals, and the reduction in the fungal burden was comparable to that determined for an AMB-treated group . The drugs used in these studies (AMB, FLC, ITC, and caspofungin) have been categorized as either fungicidal or fungistatic. AMB is accepted as fungicidal against both C. albicans and A. fumigatus, based on the reductions in the number of CFU. Conversely, azoles such as FLC and ITC are considered fungistatic against C. albicans and A. fumigatus, respectively. It is interesting that both FLC- and ITC-treated cells exhibited weak but widespread staining with CFDA in our study. Perhaps azole-mediated ergosterol inhibition causes increased permeability of the cell membranes, such that azole-treated cells only partially retain free carboxyfluorescein following esterase cleavage of CFDA. Caspofungin has been shown to be fungicidal against C. albicans , and the fluorescence patterns observed in cells stained with CFDA and DiBAC correlate with decreases in the number of CFU in a manner comparable to that seen with AMB-treated cells. The activity of caspofungin against A. fumigatus is more difficult to categorize. While there was no reduction in the number of CFU from standard killing curve measurements, we did see changes in the patterns of staining by dyes that stain viable and nonviable cells analogous to those observed with caspofungin-treated C. albicans. Plumes of DiBAC-stained debris at hyphal tips suggest caspofungin caused lysis of the actively growing A. fumigatus cells. Whether lysis occurs in vivo remains to be determined. Work with persistently neutropenic mice illustrated that caspofungin prevents significant mortality in A. fumigatus-infected animals, even for up to 21 days after therapy is withdrawn . If caspofungin is strictly fungistatic against A. fumigatus, a reemergence of the infection should lead to mortality in the face of continued immunosuppression. Perhaps significant injury to cells at the hyphal tips renders the fungus unable to invade tissue beyond a focus of infection. As subapical cells within the mycelium form branches, their requirement for de novo synthesis of 1,3-beta-d-glucan may render them susceptible to lysis by caspofungin. The inability of A. fumigatus to sustain polarized growth in the presence of multiple doses of caspofungin could lead to significant fungal cell death in tissues. Backmatter: PMID- 12183282 TI - Effect of the Echinocandin Caspofungin on Expression of Candida albicans Secretory Aspartyl Proteinases and Phospholipase In Vitro AB - Although the echinocandin caspofungin primarily inhibits the synthesis of cell wall 1,3-beta-d-glucan, its fungicidal activity could also potentially perturb the expression of virulence factors involved in the ability of Candida albicans to cause infection. Expression of the C. albicans secretory aspartyl proteinase (SAP) and phospholipase B (PLB) virulence genes was determined by reverse transcription-PCR after the addition of caspofungin to cells grown for 15 h in Sabouraud dextrose broth. In cells that remained viable, expression of SAP1 to SAP3, SAP7 to SAP9, and PLB1 was unaltered after exposure to fungicidal concentrations (4 to 16 mug/ml) of caspofungin over a period of 7 h. However, expression of SAP5 increased steadily beginning 1 h after exposure to caspofungin. These results indicate that caspofungin is rapidly fungicidal against C. albicans, before any suppression of SAP or PLB1 gene expression can occur. Keywords: Introduction : The cyclic lipopeptide pneumocandins and echinocandins and the glycolipid papulacandins are members of a new class of antifungal agents which exert their activity by noncompetitive inhibition of fungal 1,3-beta-d-glucan synthase . This enzyme is essential for the synthesis of cell wall glucan which provides structural integrity and osmotic stability for fungi but is not found in cells from higher eukaryotes including humans. Disruption of cell wall structure by inhibition of glucan synthesis results in osmotic instability and lysis of the fungal cell . Caspofungin (MK-0991) is a water-soluble semisynthetic amine derivative of the natural product pneumocandin Bo , which in turn is a fermentation product derived from the fungus Glarea lozoyensis . Caspofungin was developed as a potential antifungal and anti-Pneumocystis agent . In vitro, caspofungin is fungicidal against Candida species, including azole-resistant species, and is fungistatic against Aspergillus species . However, it is inactive against Fusarium, Rhizopus, Trichosporon, and Cryptococcus neoformans . In addition to prolonging survival in mouse models of disseminated candidiasis and aspergillosis , caspofungin has recently shown promising clinical activity for the treatment of life-threatening infections caused by Candida and Aspergillus species (; J. Maertens, I. Raad, C. A. Sable, A. Ngai, R. Berman, T. F. Patterson, D. Denning, and T. Walsh, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 1103, 2000; A. Villanueva, E. Gotuzzo, E. Arathoon, L. M. Noriega, N. Kartsonis, R. Lupinacci, J. Smietana, R. S. Berman, M. J. Dinubile, and C. A. Sable, Abstr. 41st Intersci. Conf. Antimicrob. Agents Chemother., abstr. J-675, 2001). Because caspofungin is fungicidal, it could potentially affect a diversity of cellular processes which contribute to the ability of Candida albicans to cause infection. The primary mechanism of action of pneumocandins and echinocandins is inhibition of synthesis of cell wall 1,3-beta-d-glucan, and these compounds do not inhibit other C. albicans membrane-bound enzymes, such as chitin synthase or serine palmitoyltransferase . However, echinocandin antimycotics inhibit the incorporation of glucan-associated enolase into the growing cell wall of C. albicans . In addition, although the synthesis of C. albicans RNA is only modestly inhibited at the MIC of the pneumocandin L-733,560, a close analogue of caspofungin, marked inhibition occurs at 250 times the MIC . Fungal virulence factors are attracting attention as potential targets for drug development . C. albicans secretory aspartyl proteinases (Saps), under the control of a multigene family (SAP1 to SAP9) expressing distinct isoenzymes which are regulated differentially at the mRNA level in vitro and in vivo , are implicated in the breakdown of several host substrates . There is evidence that phospholipase B expressed by at least two genes (PLB1 and PLB2) also contributes to the pathogenesis of candidiasis by the degradation of host tissues . The potential impact of antifungal therapies on expression of C. albicans virulence factors is exemplified by the enhanced production of Sap by azole-resistant C. albicans isolates from a patient infected with human immunodeficiency virus after growth in subinhibitory concentrations of fluconazole . In the present study, we sought to determine whether caspofungin could potentially exert added anticandidal activity by inhibiting the expression of C. albicans SAPs and PLB1, as a consequence of nonspecific effects on transcription. Candida strains. : Sap- and phospholipase-producing C. albicans LAM-1 (serotype A) was originally isolated from the blood of a patient with systemic candidiasis . The isolate was stored at -80C in a solution containing 65% (vol/vol) glycerol, 10 mM Tris (pH 7.5), and 10 mM MgCl2. Prior to use, Candida was cultured at 37C on Sabouraud dextrose agar (SDA) (Becton Dickinson and Company, Cockeysville, Md.). Antifungal susceptibility testing. : The susceptibility of C. albicans strain LAM-1 to antifungal agents was determined by the broth microdilution version of the M27-A method defined by the National Committee for Clinical Laboratory Standards . The method was performed in RPMI 1640 medium (Gibco BRL, Gaithersburg, Md.) buffered to pH 7.0 with 0.165 M morpholinepropanesulfonic acid (MOPS) buffer (MOPS-buffered RPMI; ICN Biochemicals, Cleveland, Ohio) with readings at 48 h, except for amphotericin B, which was tested in antibiotic medium 3 and read after 24 h . The MICs of the azoles and flucytosine were defined as the lowest concentrations giving prominent growth reduction. For the other drugs, the MIC was the lowest drug concentration giving an optically clear well. Candida parapsilosis ATCC 22019 and Candida krusei ATCC 6258 were included as quality control isolates . After 48 h, 20 mul of each clear well was plated on SDA, and the minimum fungicidal concentration (MFC) was the lowest concentration giving no growth (>98% killing). Time-kill curve procedures. : Caspofungin (MK-0991) was provided as a pure white powder by the Merck Research Laboratories (Rahway, N.J.) and stored at -80C. A stock solution at a concentration of 400 mug/ml was prepared in sterile distilled water for each experiment. Three to five colonies of C. albicans grown for 24 to 48 h on SDA were suspended in 5 ml of 0.01 M phosphate-buffered saline (pH 7.2), and the cells were counted using a hemacytometer. An appropriate volume of fungal suspension was then added to Erlenmeyer flasks containing 70 ml of either Sabouraud dextrose broth (SDB; Difco Laboratories, Detroit, Mich.) or MOPS-buffered RPMI to yield a starting inoculum of approximately 103 CFU/ml. Caspofungin was added to individual cultures with resulting concentrations ranging from 0.125 to 16 mug/ml, either at the start of culture or after 15 h of incubation with rotary agitation at 37C. Culture medium without drug served as a growth control. At predetermined time points, duplicate 100-mul volumes were withdrawn from each culture and serially diluted (10-fold), and a 100-mul aliquot from each dilution was plated on SDA for determination of CFU. The results were expressed as the averages of the duplicate CFU determinations. When colony counts were anticipated to be less than 103 CFU/ml, duplicate 0.1-ml aliquots of the cultures were plated directly on SDA without dilution. A 1-ml aliquot of each culture was also removed at the same time points and centrifuged, and the pellet was immediately frozen at -80C for analysis of C. albicans SAP and PLB1 gene expression. RT-PCR. : Isolation of RNA from the pelleted cells and reverse transcription-PCR (RT-PCR) were performed as previously described . In order to assess gene expression by the portion of C. albicans which remained viable during the time-kill experiments, RT-PCR for each gene and time point was conducted using an amount of RNA originating from 1,400 CFU of C. albicans. For the detection of gene expression, individual digoxigenin-labeled probes were used to specifically detect each of the RT-PCR products by Southern blotting . Positive controls for expression of each of the individual SAP and PLB1 genes were included in the RT-PCR assay, using RNA from C. albicans LAM-1 grown in vitro under conditions known to induce their expression. The EFB1 gene was amplified as a positive mRNA control and to detect any contaminating genomic DNA by the presence of a 365-bp intron . A negative control for expression of each gene was obtained by omitting the RNA in the RT-PCR assay . The MICs for C.albicans LAM-1 were typical of those for other C.albicans isolates (flucytosine, 0.125 mug/ml; amphotericin B, 1 mug/ml; fluconazole, 0.25 mug/ml; itraconazole, 0.125 mug/ml; posaconazole, 0.031 mug/ml; voriconazole, 0.031 mug/ml; caspofungin, 1 mug/ml; micafungin [FK-463], 0.031 mug/ml; and anidulafungin [LY-303366], 0.031 mug/ml). This strain was thus susceptible to all classes of antifungal agents tested, including the echinocandins. The MFC for caspofungin (1 mug/ml) was equivalent to the MIC, demonstrating fungicidal activity against C. albicans and in agreement with reported in vitro susceptibility data of C. albicans to this antifungal . When caspofungin was added at concentrations (0.125 to 1 mug/ml) at or below the MIC to late-log-phase C. albicans grown in SDB for 15 h , the apparent rate of reduction of the number of CFU per milliliter was unexpectedly inversely proportional to the drug concentration. This effect was correlated with increasing macroscopic and microscopic clumping of C. albicans cells as the concentration of caspofungin decreased. Interestingly, aggregates of C. albicans blastoconidia have been previously observed in an enriched synthetic liquid medium after 4 h of treatment with caspofungin, L-733,560, as well as acidic terpenoid (1, 3)-beta-d-glucan synthase inhibitors, at concentrations at or below the MIC or MFC (; M. Kurtz, personal communication). In another report, however, caspofungin concentrations below that required to inhibit 80% of the strains resulted in growth curves similar to that for the control in the absence of caspofungin for all C. albicans isolates tested and no report of aggregation . The apparent enhanced killing of C. albicans at caspofungin concentrations below the MIC may have resulted from the formation of increasingly large aggregates, each giving rise to a single colony and consequently being counted as 1 CFU. Aggregation was not prevented by adding 0.01% (wt/vol) Tween 20 or 0.0025% Tween 80 (Fisher Scientific, Fair Lawn, N.J.) to the SDB culture medium (data not shown). In addition, time-kill curves constructed by a proposed standardized method in RPM1 1640 buffered with MOPS also resulted in aggregation at values at or below the MIC and in a rate of killing comparable to that in SDB at concentrations (4 to 16 mug/ml) above the MIC or MFC (data not shown). As expected, these higher concentrations of caspofungin (4 to 16 mug/ml) all produced approximately 90% killing of C. albicans within 7 h in SDB medium with no visible aggregation but abundant cellular debris microscopically . The absence of additional killing effect against C. albicans at increasing concentrations of caspofungin (4 to 16 mug/ml) has been observed by other investigators . Taken together, the results of the time-kill studies suggest that inhibition of synthesis of cell wall glucan by caspofungin primarily alters the surface properties of C. albicans at concentrations below the MIC, causing the cells to aggregate, while killing at concentrations above the MIC or MFC disrupts the cells and aggregation is thus not observed. Interestingly, caspofungin at concentrations below the MIC induces a high-affinity fibronectin receptor that may promote the adhesive properties of C. albicans (M. L. Pendrak, T. J. Walsh, and D. D. Roberts, Prog. Abstr. 6th ASM Conference on Candida and Candidiasis, abstr. 124, 2002). Time-kill curves were also constructed by adding caspofungin to a smaller inoculum of C. albicans (103 CFU) at the start of culture in SDB and quantitating CFU at determined time points over 20 h of incubation. The apparent reduction in CFU was again inversely proportional to the concentration of caspofungin because of aggregation, but at least 99% killing was achieved within 9 h at higher concentrations . However, a persistent fraction of C. albicans was not killed, and growth of the fungus quickly resumed until the end of incubation. It is hypothesized that this rebound in CFU may be the result of decreased drug stability under the extended study conditions . Monitoring of the pH at each time point revealed pH values of 5 to 5.5, making it unlikely that caspofungin was inactivated by deviating from the recommended pH range of 5 to 7 (Merck Research Laboratories). Expression of the SAP1 to SAP9 and PLB1 genes of C. albicans was determined by RT-PCR after the addition of caspofungin (4 to 16 mug/ml) to cells grown for 15 h in SDB. In comparison to a control culture without caspofungin, expression of SAP1 to SAP3, SAP7 to SAP9, PLB1, and the EFB1 positive-control gene was unaltered at these drug concentrations over a period of 7 h . Interestingly, however, expression of SAP5 increased steadily beginning 1 h after exposure to these concentrations of caspofungin . Although SAP5 is expressed in vitro by the hyphal form of C. albicans at neutral pH , expression has been found using in vivo expression technology in a murine model of esophageal candidiasis when only blastoconidia are observed . Although regulation of expression of the SAP genes has not been completely elucidated at the molecular level, it could be hypothesized that induction of expression of SAP5 may have resulted from signaling events occurring in response to cell wall damage by caspofungin. The C. albicans transcriptional regulators CPH1 and EFG1 mediate the activation of the SAP5 gene and may have been up-regulated after exposure to caspofungin. Alternatively, peptide hydrolysis products of the cyclic hexapeptide caspofungin may have induced expression of SAP5, because peptides up-regulate expression of the SAP genes in vitro . Because the RT-PCR analysis was conducted on the portion of C. albicans cells exposed to but not yet killed by the antifungal agent, inhibition of 1,3-beta-d-glucan synthesis by caspofungin was rapidly fungicidal against C. albicans before the expression of virulence genes could be perturbed. The absence of suppression of expression of SAP genes, PLB1, and the control EFB1 gene by caspofungin concurs with the highly specific inhibition of glucan synthase by the close analogue L-733,560, which begins to show nonspecific effects at 25x MIC . In conclusion, it is unlikely that suppression of SAP and PLB1 gene expression by caspofungin occurs at the caspofungin concentrations in plasma found in clinical treatment of candidiasis , and such an indirect mechanism would not be expected to contribute to clearance of C. albicans during treatment with caspofungin. FIG. 1. | Plots of the number of C. albicans Plots of the number of C. albicans treated with different concentrations of caspofungin over time. The mean values for log10 of the number of CFU per milliliter versus time for C. albicans strain LAM-1 tested against different concentrations of caspofungin. The following concentrations (in micrograms per milliliter) of caspofungin were used: 0 (control) , 0.0625 , 0.125 , 0.25 (*), 1 , 4 , 8 , and 16 . Caspofungin was added to cultures in SDB after 15 h of incubation (A) or at the start of culture (B). Representative results are shown from four (A) and two (B) independent experiments. FIG. 2. | Expression of the SAP2, SAP5, and EFB1 genes of C. albicans Expression of the SAP2, SAP5, and EFB1 genes of C. albicans over a period of 7 h after the addition of caspofungin (4 to 16 mug/ml) to cells grown for 15 h in SDB. Representative results are shown from two independent experiments. EFB1 was used as a positive control. Expression of the remaining SAP genes and PLB1 was similar to that of SAP2. Backmatter: PMID- 12183283 TI - Development of a Yeast Assay for Rapid Screening of Inhibitors of Human-Derived Pneumocystis carinii Dihydrofolate Reductase AB - Human-derived Pneumocystis carinii dihydrofolate reductase (DHFR) was expressed in a Saccharomyces cerevisiae strain whose growth depends on complementation by this enzyme. We utilized a quantitative assay to measure the sensitivity of this yeast strain to DHFR inhibitors. This assay should be useful for identifying new inhibitors of human-derived P. carinii DHFR. Keywords: Introduction : Current treatment of Pneumocystis carinii pneumonia (PCP) is complicated by frequent toxic and allergic side effects, emerging drug resistance, and limited therapeutic options. There is a clear need to discover new agents against P. carinii. Dihydrofolate reductase (DHFR) is a well-described target for antimicrobial chemotherapy, and its inhibitors, like trimethoprim, pyrimethamine, and trimetrexate, are commonly used drugs for PCP therapy or prophylaxis. Genes encoding DHFR of P. carinii from various mammalian host species have been cloned . The amino acid sequence of DHFR of human-derived P. carinii differs from that of rat-derived P. carinii by 38%, with some of the amino acid changes occurring in the active sites of the enzyme . Thus, the ideal target for designing potential antifolate drugs for the treatment of humans is human-derived P. carinii DHFR . The expression of recombinant human-derived P. carinii DHFR enzyme has permitted kinetic characterization of the enzyme and reassessment of the inhibitory potency of several commonly used antifolate drugs by in vitro enzyme assays . The goal of the present study was to develop a rapid, complementation-based drug screening system by using the yeast Saccharomyces cerevisiae expressing human-derived P. carinii DHFR, as has been previously reported for other organisms . The yeast model system used in this study was derived from an S. cerevisiae strain (TH5; Mataura3-52 leu2-3,112 trp1 tup1 dfr1::URA3) (kindly provided by Carol Hopkins Sibley) in which the endogenous DHFR gene is inactivated . The human-derived P. carinii DHFR coding region was amplified by PCR from a plasmid we constructed previously , cloned into an expression vector containing S. cerevisiae sequences that permit replication of the plasmid and expression of a heterologous gene in yeast . and transformed into TH5 cells by the lithium acetate method. The transformants were selected on plates containing tryptophan-deficient synthetic medium supplemented with 100 mug of dTMP (Sigma, St. Louis, Mo.) per ml and plated on rich yeast extract-peptone-dextrose medium without dTMP to test the function of the construct . TH5 yeast strains expressing rat-derived P. carinii DHFR or human DHFR have been described previously . To determine the sensitivity of the engineered yeasts to several selected antifolate drugs, assays of the concentration of the drug required to inhibit cell growth by 50% (IC50 assays) were conducted with 1 mM sulfanilamide, as previously described . Each drug was tested in triplicate in at least two separate experiments. Figure and Table illustrate the inhibition patterns of selected DHFR inhibitors for different yeast strains. For the human-derived P. carinii DHFR yeast strain, trimethoprim and pyrimethamine were both weak inhibitors, with IC50s in the micromolar range; trimetrexate was about 10-fold and 40-fold more potent than trimethoprim and pyrimethamine, respectively. WR99210, a triazine compound (Jacobus Pharmaceutical Company, Princeton, N.J.), has been found to be highly effective against malaria DHFR but has not been tested for activity against human-derived P. carinii DHFR. WR99210 was an even more potent inhibitor, with an IC50 in the 10-8 M range. In comparison with the rat-derived P. carinii DHFR strain, the human-derived P. carinii DHFR strain showed very similar sensitivities to pyrimethamine and WR99210 but was about 10-fold more sensitive to both trimethoprim and trimetrexate. WR99210 exhibited excellent selectivity for both human- and rat-derived P. carinii DHFRs compared to that for human DHFR. While trimethoprim showed a relatively favorable selectivity, trimetrexate and pyrimethamine appeared not to be selective in this assay system. Table shows a comparison between the IC50s from the yeast assay and those from an in vitro enzyme assay, based upon previously published data (for trimethoprim, pyrimethamine, and trimetrexate) or data generated in this study by using previously described methods and conditions (WR99210) . Based on three experiments, the mean IC50 (+- standard deviation) of WR99210 for human-derived P. carinii DHFR was 10.0 (+-1.0) nM. A comparison of the relative inhibition profiles in the yeast assay and those in the in vitro assay supports the usefulness of the yeast assay as an initial step for identifying new inhibitors of human-derived P. carinii DHFR. The yeast assay is simple, fast, and inexpensive and could readily be adapted for automation. The usefulness of this system in screening for antimicrobial drugs has been well demonstrated in previous studies . The IC50 obtained by the yeast assay is the result of a complex function of the level of expression of the target enzyme in the yeast, the penetration of the drug to the cellular location of the target enzyme, and the intrinsic ability of the drug to inhibit the enzyme . Furthermore, the addition of 1 mM sulfanilamide in the yeast assay will also inhibit, to some extent, the growth of the yeast cells . Therefore, the actual IC50 in the yeast assay cannot be compared directly with that obtained from an in vitro assay, which reflects a direct interaction between the drug and the target enzyme under standardized conditions. As shown in Table , the absolute IC50s of the four drugs tested showed some differences between the two assay systems. In a given recombinant strain, enzyme expression should not account for the relative differences among drugs in the yeast assay and those in the in vitro assay, which suggests that the abilities of the drugs to penetrate the yeast cell wall may differ. Given the uncertainty about the ability of various drugs to penetrate P. carinii compared to that of yeast, application of these observations to potential therapeutic benefit in the treatment of PCP must be done cautiously. It is of particular interest that the human-derived P. carinii DHFR yeast strain showed an approximately 10-fold increase in sensitivity to trimetrexate and trimethoprim compared to that of the rat-derived P. carinii DHFR yeast strain. Since the host yeast strains are identical, differences in penetration should not account for the relative differences among drugs demonstrated by the yeast assay and in vitro assay. Furthermore, given that the sensitivities to pyrimethamine and WR99210 were similar, it is unlikely that these differences represent differential levels of enzyme expression in the two yeast strains. Rather, as suggested by the in vitro enzyme inhibition data, this likely reflects the effects of the primary sequence differences between human- and rat-derived P. carinii DHFRs . DHFR inhibitors have been among the most studied classes of drugs for treatment of PCP. While animal studies suggest that trimethoprim plays a minimal role in the treatment of PCP (consistent with its relatively poor inhibition seen in this and other studies), DHFR inhibitors such as trimetrexate clearly have potent anti-P. carinii activity, even as single agents . The present assay should facilitate identification of additional potent and selective agents. FIG. 1. : Inhibition of DHFR from human-derived P. carinii Inhibition of DHFR from human-derived P. carinii , rat-derived P. carinii , and humans , by using S. cerevisiae complemented by the corresponding DHFR genes. The results of a representative IC50 assay are shown for each drug tested. The yeast growth in the control without drug was scored as 100%, and the growth of yeast cells at each drug concentration was divided by that of the control to determine the relative growth. Each point represents the mean value of triplicate data. TABLE 1 : Comparison of the IC50s for DHFR inhibitors obtained from the yeast complementation assay with those from the in vitro assay Backmatter: PMID- 12183246 TI - High Incidence of Cefoxitin and Clindamycin Resistance among Anaerobes in Taiwan AB - Susceptibilities to 16 antimicrobial agents were determined by measurement of MICs for 344 isolates of anaerobic bacteria recovered from patients with significant infections. Resistance rates varied among antimicrobial agents and the species tested. The beta-lactams were more active in gram-positive than in gram-negative anaerobes. Resistance to meropenem was low (<1%). For beta-lactam-beta-lactamase inhibitors, piperacillin-tazobactam was most active for all species (resistance, <6%). The rates of resistance to cefoxitin (31 to 65%) and clindamycin (50 to 70%) for non-Bacteroidesfragilis species of the B. fragilis group were higher than those for B. fragilis (4% resistant to cefoxitin and 33% resistant to clindamycin). Among members of B. fragilis group, Bacteroides thetaiotaomicron was the most resistant to clindamycin (70%) and cefoxitin (65%). Rates of susceptibility to imipenem and metronidazole for B. fragilis continue to be high compared to those from a previous study 10 years ago. However, resistance to metronidazole was found recently in five strains of B. fragilis. We analyzed the genetic relationships among the metronidazole-resistant B. fragilis strains by pulsed-field gel electrophoresis. The metronidazole-resistant B. fragilis strains showed genotypic heterogeneity, excluding the dissemination of a single clone. Keywords: Introduction : Over the past 10 years, there has been a significant problem with increasing resistance to antimicrobial agents among anaerobic bacteria (-, -). Antimicrobial resistance is becoming less predictable and may fluctuate from one medical center to another, as well as from one geographic region to another (-, -). The increasing resistance among several species emphasizes the need to survey the susceptibility patterns of these organisms. Data on susceptibilities of anaerobes are very limited in Taiwan except those from a previous report of 10 years ago . The objective of this study was to determine the susceptibility profiles of clinical isolates of anaerobes in Taiwan and to monitor susceptibility changes over time. Organisms tested included gram-positive and gram-negative anaerobes which were clinically commonly encountered. The test antimicrobial agents included old and new agents. Among the older agents, clindamycin, cefoxitin, and piperacillin are commonly used as the initial empirical treatment for B. fragilis group infections. However, resistance to these agents has been shown to increase in North America, Europe, and other countries during the past decades (-, -, ). The 5-nitroimidazole molecules are very potent anaerobicidal agents commonly used to treat or prevent Bacteroides infections. Although resistance to metronidazole in Bacteroides fragilis strains has been reported in several countries , resistance to metronidazole in B. fragilis strains has not been reported in Taiwan before. The emergence of metronidazole-resistant B. fragilis strains (MIC, >32 mug/ml) in Taiwan is reported in this study. MATERIALS AND METHODS : Bacterial strains. | A total of 344 clinical isolates of anaerobes were collected between 1998 and 2000 from the Bacteriology Laboratory, National Taiwan University Hospital, a 2,000-bed teaching hospital in northern Taiwan. These isolates were recovered from blood, pus from the intra-abdominal cavity, abscesses, soft tissue, head or neck wounds, and others. Only one isolate per patient was included. Antimicrobial susceptibility testing. | The antimicrobial agents used for susceptibility testing were as follows: penicillin, ampicillin, piperacillin, sulbactam, tazobactam, cefoperazone, clindamycin, chloramphenicol, and metronidazole (Sigma Chemical Co., St. Louis, Mo.); ticarcillin and clavulanic acid (SmithKline Beecham, Philadelphia, Pa.); cefoxitin and imipenem (Merck Sharp & Dohme, West Point, Pa.); meropenem (Sumitomo Pharmaceuticals, Osaka, Japan); cefmetazole (Sankyo, Tokyo, Japan); and moxifloxacin (Bayer Corporation, West Haven, Conn.). Antimicrobial susceptibility was tested by an agar dilution method in accordance with guidelines of the National Committee for Clinical Laboratory Standards (NCCLS) . An inoculum of 105 CFU per well was applied with a Steers replicator onto brucella agar supplemented with vitamin K1 and 5% pooled sheep blood. Plates were incubated in an anaerobic chamber for 48 h at 35C. The MIC was defined as the concentration at which there was a marked change in the appearance of growth, compared with that in the control plate. Reference strains of B. fragilis ATCC 25285 and Bacteroides thetaiotaomicron ATCC 29741 were used for quality control of the susceptibility tests. PFGE. | Pulsed-field gel electrophoresis (PFGE) was performed to genotype five metronidazole-resistant B. fragilis strains. Each strain was grown overnight at 37C in 10 ml of brain heart infusion broth in an anaerobic chamber. The preparation of DNA was performed as described previously . After appropriate preparation, the DNAs in each plug were digested with 20 U of XbaI (New England Biolabs, Hitchin, United Kingdom) at 37C for 4 h. The plugs were applied to a 1% agarose gel. Electrophoresis was performed in 0.5x Tris-borate-EDTA buffer at 14oC by using a CHEF-DR III apparatus (Bio-Rad Laboratories, Hercules, Calif.), with pulse times ranging from an initial value of 4 s to a final value of 30 s, for 16 h at 200 V. RESULTS : Antimicrobial susceptibilities. | The MIC ranges, MICs at which 50% of the isolates were inhibited (MIC50s), MICs at which 90% of the isolates were inhibited (MIC90s), and the percentages of 344 clinical isolates of anaerobes that were susceptible, intermediate, or resistant to various antimicrobial agents are summarized in Table The rates of susceptible, intermediate, and resistant isolates were determined by using the NCCLS breakpoints . B. fragilis group isolates were the most encountered clinically significant isolates among the gram-negative anaerobes. The B. fragilis group isolates were uniformly resistant to penicillin and ampicillin. Most isolates were susceptible to imipenem or meropenem. Only one isolate of B. fragilis was found to have intermediate resistance to imipenem. Comparison of the susceptibilities of the individual species of the B. fragilis group showed that 40% of B. fragilis isolates were resistant to piperacillin, 44% were resistant to ticarcillin, and 17% were resistant to ampicillin-sulbactam. Four percent of B. fragilis isolates were resistant to cefoxitin; however, 22% had intermediate susceptibility. Other members of the B. fragilis group were more resistant to cefoxitin, with resistance rates between 31 and 65%. B. thetaiotaomicron was the species with the greatest resistance to cefoxitin (65%) and cefmetazole (80%) among the B. fragilis group. Resistance to clindamycin varied among the species from 33 to 70%, with the highest resistance rate occurring in B. thetaiotaomicron (70%), followed by Bacteroides caccae (67%). Chloramphenicol and metronidazole were active against >90% of isolates of the B. fragilis group. Only 3% of B. fragilis isolates were resistant to metronidazole and 2% were intermediately resistant to metronidazole. All Fusobacterium isolates were susceptible to ampicillin-sulbactam, ticarcillin-clavulanic acid, cefoxitin, and imipenem. The majority of Fusobacterium isolates (95%) were susceptible to cefmetazole. Thirty-five percent of Fusobacterium isolates were resistant to clindamycin. More than half of Prevotella species isolates (62%) were resistant to penicillin and ampicillin. Rates of resistance to cefoxitin, cefmetazole, and clindamycin were 6, 12, and 31%, respectively. All of the Prevotella isolates tested were susceptible to imipenem, meropenem, and chloramphenicol. Among Veillonella isolates, 70% were resistant to penicillin and ampicillin. Five percent of Veillonella isolates were resistant to piperacillin, ticarcillin, piperacillin-tazobactam, and ticarcillin-clavulanic acid. Ten percent were resistant to cefoxitin, 55% were resistant to clindamycin, and 20% were resistant to metronidazole. All isolates were susceptible to ampicillin-sulbactam, imipenem, meropenem, and chloramphenicol. Of the gram-positive isolates, the 20 Clostridium perfringens isolates were susceptible to all of the agents tested. Other gram-positive anaerobes showed various degrees of resistance to penicillin (12 to 16%) and ampicillin (13 to 25%). Among Peptostreptococcus species isolates, 16% were resistant to penicillin. All isolates were susceptible to piperacillin-tazobactam and meropenem. Of Peptostreptococcus species isolates, 3% were resistant to cefoxitin and imipenem but 55% were resistant to clindamycin. Clostridium species other than C. perfringens were more resistant than C. perfringens, with 12% of the isolates resistant to penicillin and ticarcillin, 25% resistant to ampicillin, and 31% resistant to metronidazole. The MIC50s and MIC90s of moxifloxacin for all species ranged from 0.12 to 2 mug/ml (MIC50s) and from 0.25 to 8 mug/ml (MIC90s). Trend of cefoxitin and clindamycin resistance in B. thetaiotaomicron. | Figure shows the annual rates of susceptibility to six routinely tested agents for all of the Bacteroides species isolates at National Taiwan University Hospital from 1977 to 2000. The rates of susceptibility to cefmetazole and clindamycin decreased. Susceptibility testing was performed by the disk diffusion method before 1990 and by the breakpoint agar dilution method after 1991. A stepwise increase in the rates of resistance to cefoxitin and clindamycin resistance was noted. PFGE of metronidazole-resistant B. fragilis isolates. | Five metronidazole-resistant strains of B. fragilis were found in this study. Since no metronidazole-resistant strains of B. fragilis were reported before in Taiwan. We analyzed the genetic relationships among these strains by PFGE. The results are shown in Fig. . Lanes 1 to 6 show metronidazole-nonsusceptible B. fragilis isolates; lanes 2 and 3 show isolates from the same individual. Lanes 7 to 9 show metronidazole-susceptible B. fragilis isolates. Two strains (lanes 2 or 3 and 4) were similar. Other strains have distinct patterns. These isolates showed genotypic heterogeneity, suggesting that they correspond to a highly heterogeneous population rather than to the dissemination of a single clone. FIG. 1. | Rates of susceptibility to six routinely tested agents in all of the Bacteroides species isolates at National Taiwan University Hospital from 1977 to 2000. Rates of susceptibility to six routinely tested agents in all of the Bacteroides species isolates at National Taiwan University Hospital from 1977 to 2000. Susceptibility testing was performed by the disk diffusion method before 1990 and by the breakpoint agar dilution method after 1991. SAM, ampicillin- sulbactam. FIG. 2. | PFGE analysis of XbaI-digested genomic DNA from B. fragilis PFGE analysis of XbaI-digested genomic DNA from B. fragilis isolates. Lane M, lambda ladder molecular size markers; lanes 1 to 6, metronidazole-nonsusceptible B. fragilis isolates (lanes 2 and 3, isolates from the same individual); lanes 7 to 9, metronidazole-susceptible B. fragilis isolates. TABLE 1 | In vitro susceptibilities of clinical isolates of anaerobes DISCUSSION : In this study the antimicrobial susceptibilities of 344 isolates of anaerobes to various agents were determined. In agreement with other reports, the susceptibility results varied among genera and species. Compared to a previous study by our group of 100 B. fragilis isolates done 10 years ago , the increased resistance of the B. fragilis group to cefoxitin and clindamycin is noted. Rates of resistance to cefoxitin for B. fragilis increased from 1 to 4% and rates of resistance to clindamycin increased from 29 to 33% over the period from 1991 to 2000. The level of chloramphenicol susceptibility remained unchanged. As expected, the beta-lactams were more active in gram-positive than in gram-negative anaerobes. According to NCCLS guidelines, members of the B. fragilis group are presumed to be resistant to ampicillin. Peptostreptococcus species have been considered fully susceptible to several beta-lactam drugs, including penicillin G. In the present study, isolates resistant to penicillin within Peptostreptococcus species were mostly Peptostreptococcus anaerobius. In Korea, the rate of resistance to penicillin of P. anaerobius was also high, while those of other species were lower . Addition of a beta-lactamase inhibitor generally reversed the resistance. However, Veillonella displayed similar susceptibilities to piperacillin and piperacillin-tazobactam and ticarcillin and ticarcillin-clavulanic acid. The rate of resistance to ampicillin-sulbactam in B. fragilis isolates increased from 8 (1991) to 17% (this study). Resistance to piperacillin-tazobactam was low (<6%). Data collected from our hospital's clinical microbiology laboratory reveal the decrease in susceptibility to cefmetazole in Bacteroides species from 1987 to 2000. Cefoxitin was more active than cefmetazole against most species. Rates of resistance to cefoxitin varied greatly with species and country. In our institution, the percentage of resistance to cefoxitin rose from 1% in 1991 to 4% for B. fragilis and from 10 to 70% for B. thetaiotaomicron. The rate of resistance to cefoxitin among the B. fragilis group was 12.8% in Spain , 11.3% in Canada (1997) , and 1.7 to 14.2% depending on the species in the United States (1995 and 1996) , 2.9 to 34.5% in Japan (1990 to 1992) , and 5% in South Africa . MICs for many isolates were 32 mug/ml, which was considered by the NCCLS as intermediate in susceptibility. Fusobacterium isolates remained susceptible to cefoxitin. The high rates of resistance to cefoxitin in non-B. fragilis Bacteroides species were unusual. The use of cefoxitin and cefmetazole has not increased in the past 10 years. Therefore the reason for the high incidence of resistance to cefoxitin is unclear. For all species, high rates of susceptibility to imipenem and meropenem were observed. Resistance to meropenem was low (<1%). One isolate of B. fragilis and one isolate of Peptostreptococcus displayed intermediate susceptibility, and another Peptostreptococcus isolate displayed resistance, to imipenem. In general, The MIC90s of imipenem and meropenem were similar except that some isolates showed discordant susceptibilities to imipenem and meropenem. For example, six isolates of B. fragilis (five intermediate and one resistant) were not susceptible to meropenem, but only one isolate was not susceptible to imipenem. Eight percent of B. thetaiotaomicron isolates were resistant to imipenem but susceptible to meropenem. Compared to the data of our previous report, the rate of resistance to imipenem for B. fragilis has not increased, but the MIC50 (from 0.12 to 0.5 mug/ml) and MIC90 (from 1 to 4 mug/ml) have increased slightly. Resistance to the carbapenems has been occasionally and infrequently recorded . Clindamycin has long been considered the drug of choice for treatment of anaerobes. However, over the past 20 years, there has been a significant increase in the rate of resistance to clindamycin among isolates of the B. fragilis group in many areas (, , , -). In our institution, the overall activities of clindamycin against the B. fragilis group were poor (33 to 70%). The high prevalence of resistance to clindamycin in B. fragilis group isolates has been described previously in several reports. For example, a high prevalence of resistance to clindamycin (49%) in the B. fragilis group was observed by Betriu et al. in Spain . In Korea, in 1994, the rates of resistance to clindamycin for B. fragilis, B. thetaiotaomicron, and other Bacteroides spp. were 38, 45.5, and 69%, respectively . However, in several areas the rate of resistance to clindamycin for the B. fragilis group remained low. In South Africa, only 5% of isolates were resistant to clindamycin in the B. fragilis group . It was also reported that clindamycin resistance is associated with hospital-acquired infections . Rates of resistance for isolates varied greatly with species and country. In the present study, high rates of resistance to clindamycin were found for the following organisms: B. thetaiotaomicron (70%), B. caccae (67%), Bacteroides uniformis (56%), Veillonella and Peptostreptococcus spp. (55%), Fusobacterium spp. (35%), and B. fragilis and Prevotella spp. (31%). B. thetaiotaomicron isolates were also more resistant to cefoxitin than other species. In agreement with other reports, the resistance rates for non-B. fragilis species of the B. fragilis group were found higher than that for B. fragilis (33%). Aldridge et al. reported that Bacteroides distasonis and Bacteroides ovatus were more resistant to clindamycin than other species . Since B. thetaiotaomicron is usually the second most frequently encountered Bacteroides species, rapid detection and identification are important. We recently described a PCR assay which provided a rapid and accurate method for identification of B. thetaiotaomicron . Among gram-positive anaerobes, C. perfringens was the most susceptible. Other Clostridium species were less susceptible to penicillin, ampicillin, ticarcillin, and metronidazole. This result is similar to the finding of a study done in Korea but is different from that of a study done in South Africa . In the present study, the MIC results for moxifloxacin confirm the broad spectrum of its activity against gram-positive and gram-negative anaerobic bacteria. Although no interpretation standard is available for anaerobes, for the anaerobic bacteria tested, the moxifloxacin MIC50 varied from 0.12 to 2 mug/ml and MIC90 varied from 0.25 to 8 mug/ml. Many new fluoroquinolones have been tested for in vitro activities against gram-positive, gram-negative, and anaerobic bacteria previously . Similarly, other studies have demonstrated the good activity of new fluoroquinolones against various anaerobic species. Moxifloxacin is, therefore, a potentially useful antibiotic against anaerobes. The rates of resistance to metronidazole for several gram-positive anaerobes, Peptostreptococcus (32%) and Clostridium species (31%) other than C. perfringens, were higher than those for gram-negative anaerobes Fusobacterium (25%), Veillonella (20%), Prevotella (13%), and Bacteroides species (<3%). Previous reports also showed that peptostreptococci are generally less susceptible to metronidazole than gram-negative anaerobes. A similar percentage of resistance to metronidazole for peptostreptococci in Korea was described by Lee et al. . Resistance to metronidazole among B. fragilis isolates in Taiwan is first documented in this report. By PFGE analysis, five strains (four patterns) showed genotypic heterogeneity, suggesting that they correspond to a heterogeneous population rather than to the dissemination of a single clone. The results suggest that the emergence of these resistant strains may be sporadic. The development of antibiotic resistance in anaerobic bacteria has a tremendous impact on the selection of antimicrobial agents for empirical therapy. It suggests the need to monitor antibiotic susceptibility patterns of anaerobes related to geographic regions periodically. Backmatter: PMID- 12183233 TI - Identification and Analysis of Amino Acid Mutations in Porin IB That Mediate Intermediate-Level Resistance to Penicillin and Tetracycline in Neisseria gonorrhoeae AB - PenB is the third resistance determinant in the stepwise acquisition of multiple resistance genes in chromosomally mediated resistant Neisseria gonorrhoeae (CMRNG). Alterations in porIB, one of two alleles at the por locus that encodes the outer membrane protein porin IB (PIB), were recently reported to be responsible for the increased resistance to penicillin and tetracycline conferred by penB, but the specific mutations conferring antibiotic resistance were not identified experimentally. To determine which amino acids in PIB confer increased resistance, we transformed a recipient strain with chimeras of the porIB genes from strains FA1090 and FA140 (penB2). These studies revealed that two amino acid changes, G120D and A121D, were both necessary and sufficient to confer increased resistance to penicillin and tetracycline. Site-saturation and site-directed mutagenesis of Gly-120 and Ala-121 revealed that both a single mutation, G120K, and the double mutations G120R A121H and G120P A121P also conferred antibiotic resistance to the recipient strain. The identical mutations in PIA increased penicillin and tetracycline resistance either moderately or not at all. Analysis of porIB genes present in the GenBank database from 51 clinical isolates demonstrated that lysine and aspartate mutations at positions 120 and/or 121 also occur in nature. These studies demonstrate that charged amino acids at positions 120 and 121 in PIB are highly preferential for conferring resistance to penicillin and tetracycline in N. gonorrhoeae. Keywords: Introduction : From 1945 to 1988, penicillin was the antibiotic of choice for treatment of gonococcal infections. During this time, however, the resistance of isolates gradually increased until treatment failure with penicillin became widespread and penicillin was discontinued as a first-line antibiotic. Resistance to other antibiotics, including erythromycin and tetracycline, also increased during this time. The antibiotics currently recommended for treatment of gonococcal infections are expanded-spectrum cephalosporins (i.e., ceftriaxone and cefixime) or fluoroquinolones , but resistance to fluoroquinolones threatens to make these antibiotics obsolete as well . Resistance to penicillin and tetracycline in gonococci can be either plasmid mediated or chromosomally mediated . Plasmid-mediated resistance to penicillin is due to the production of a TEM-1-like beta-lactamase, whereas plasmid-mediated resistance to tetracycline is due to expression of the TetM determinant acquired from Streptococcus pneumoniae . In contrast, chromosomally mediated resistance to penicillin and tetracycline in Neisseria gonorrhoeae results from the acquisition of multiple resistance genes, each of which confers an incremental increase in resistance until the cell becomes refractory to clinically achieved levels of the antibiotic. As demonstrated by the work of Sparling and others , these resistance genes can be transferred in the laboratory in a stepwise manner from a resistant strain to a susceptible strain by DNA uptake and homologous recombination. The first determinant in the stepwise acquisition of penicillin resistance is the penA gene, which encodes altered forms of penicillin-binding protein 2 (PBP 2) that have a lower rate of acylation by penicillin . Insertion of a single aspartic acid residue preceding Asp-345 (Asp-345a) in PBP 2 is the major cause of the decreased rate of acylation by penicillin, although additional mutations in PBP 2 also contribute . The second resistance determinant is the mtr (multiple transferable resistance) locus, which increases resistance to diverse hydrophobic agents, including erythromycin and detergent-like fatty-acids . Resistance to these agents is most commonly due to a single base pair deletion in the regulatory region of the mtrCDE locus that results in increased expression of the energy-dependent Mtr efflux pump. The third resistance determinant in chromosomally mediated resistant N. gonorrhoeae (CMRNG) is penB, which increases resistance to both penicillin and tetracycline. Previous reports showed that the penB genetic locus was closely linked to the por locus, which encodes an outer membrane porin protein through which small molecules and solutes diffuse into the periplasmic space . N. gonorrhoeae has two porins, PIA and PIB, encoded by alleles of a single por locus . In 1980, Cannon and colleagues observed a 98% cotransformation frequency of porIB and the penB gene, suggesting that por was separate from but closely linked to penB in the genome . However, Gill et al. recently reported that resistance conferred by the penB gene in the intermediate-level penicillin-resistant strain FA140 is due to mutations within the porIB gene. The authors speculated that alterations in putative loop 3, which in the crystal structures of several porins is located within the beta-barrel and forms the constriction of the pore, were responsible for increased resistance to antibiotics, but no experimental evidence was provided to support this hypothesis. Studies with other porins have shown that mutations in other regions can have significant effects on pore properties . In order to delineate more clearly the molecular mechanisms of chromosomally mediated antibiotic resistance in N. gonorrhoeae, we set out to determine the specific amino acid alteration(s) in PIB that confers resistance to penicillin and tetracycline and to define the types of amino acids at these positions that are capable of mediating resistance. MATERIALS AND METHODS : Strains. | N. gonorrhoeae strain FA1090 and the third-level transformant FA140 (FA19 x FA48) have been described previously. Strain FA19 penA4 mtr was derived from FA19 by stepwise transformation with genomic DNA from FA6140 . FA19 penA4 mtr (PIB) was constructed as described below. Escherichia coli MC1061 cells were transformed with porIB-containing plasmids, and colonies were grown on Luria-Bertani (LB) agar containing 50 mug of kanamycin/ml, 350 mug of erythromycin/ml, or 34 mug of chloramphenicol/ml. Construction of strain FA19 penA4 mtr (PIB). | Attempts to transform the coding sequence of porIB into a strain containing the porIA gene, i.e., FA19 penA4 mtr, were unsuccessful. Therefore, we generated a recipient strain of N. gonorrhoeae that contains the first two resistance determinants, penA4 and mtr, and the FA1090 porIB gene to facilitate homologous recombination between donor and recipient DNAs. The porIB gene was amplified from FA1090 genomic DNA with Taq polymerase and transformed into strain FA19 penA4 mtr. The 5' primer (por-S1, 5'-CGAGCTCGCCGTCTGAACCATCTACCGCGCCGACCTTAC-3') was complementary to a region similar1,000 bp upstream of the porIB coding sequence and contained at its 5' end a SacI restriction enzyme site (boldfaced) and the specific 10-bp uptake sequence (US) (underlined) necessary for uptake of DNA into gonococci . The 3' primer (por-U1, 5'-CGTCTGAGGCCGTCTGAATATGGATAGATTCGTCATTCCCGC-3') was complementary to a region similar300 bp downstream of the porIB gene and contained both an US and an XbaI restriction site at its 5' end. The 2.3-kb PCR amplification product was subcloned into the SacI and XbaI sites of a modified pBluescript vector, pBSC-KS, with the porIB gene in the direction opposite from the lac promoter. pBSC-KS has the chloramphenicol resistance gene from pACYC184 in place of the beta-lactamase resistance gene and a deletion of the BglI site. The kanamycin resistance gene was subsequently cloned into the unique BglI site located 40 bp downstream of the porIB stop codon. FA19 penA4 mtr was transformed with this plasmid, and transformants were selected on GCB agar plates containing 50 mug of kanamycin/ml. Replacement of the endogenous porIA gene with the FA1090 porIB gene was confirmed via sequencing of the por gene amplified by PCR from genomic DNA (University of North Carolina ---Chapel Hill [UNC-CH] Sequencing Facility). Generation of porIB mutant constructs. | The mature coding sequences (i.e., lacking the region encoding the 19-amino-acid leader sequence) of the porIB gene from strains FA1090 and FA140 plus 300 bp of downstream sequence were amplified from genomic DNA with por-U1 and a 5' primer (por-S2, CGAGCTCGCCGTCTGAAGATGTCACCCTGTACGGTGCCATCAAA) complementary to bases 58 to 84 of the porIB gene. The 5' primer also contained an US and a SacI restriction site at its 5' end. The PCR products were digested with SacI and XbaI and subcloned into similarly digested pBSC-KS. Silent unique restriction sites (HindIII, EcoRI, and ClaI) were incorporated into the FA1090 and FA140 porIB genes at codons 110, 187, and 274, which surround the codon differences between the two genes. porIB chimeras were constructed by digestion and DNA fragment exchange. Constructs containing single or double mutations were created via four-primer PCR with the FA1090 porIB gene as a template. To aid in selection, the erythromycin resistance gene (ermC) was inserted into the unique BglI site of the plasmids as described above for the kanamycin resistance gene. All constructs were verified by sequencing. A library of porIB genes containing randomized codons at positions 120 and 121 was also generated by four-primer PCR. The porIB gene from strain FA1090 containing the silent restriction sites was used as the template in the first set of reactions. The external primers were por-S2 and por-U1. The internal down primer (5'-384GGATTCCCAAGCATTGACGTTNNNNNNGGTGTTTTTCAGGGGGCTGTT337-3') contained an equal mixture of all four nucleotides at the six base positions encoding amino acids 120 and 121 in PIB and 21 bases of porIB sequence on either side. The internal up primer (5'-364AACGTCAATGCTTGGGAATCC384-3') was complementary to 21 bases of por DNA on the 5' end of the randomized sequence. Sequencing of an aliquot of the final PCR product verified that all four nucleotides were represented equally at each of the six randomized positions. The PCR products were either used directly to transform gonoccocal cells or ligated into pBSC-KS. For the latter, E. coli MC1061 cells were transformed with the ligation mixture, and a plasmid library was selected in LB broth with chloramphenicol. Sequencing of an aliquot of the library again confirmed the presence of all four bases at roughly equal intensities at each of the six randomized positions. PCR amplification of the porIB gene from transformants. | Colonies were resuspended in 50 mul of distilled H2O and boiled for 10 min, and 5-mul aliquots were used as the DNA template for PCR with Taq DNA polymerase. PCR primers were por-S1 and por-U1. Amplification conditions were 94C for 5 min; 30 cycles of 94C for 30 s, 59C for 30 s, and 72C for 1 min, 15 s; and a final extension of 72C for 7 min. For all recombinants, the entire amplified gene was sequenced to verify homologous recombination of the mutation(s). Genetic transformation. | N. gonorrhoeae was transformed as described previously . Briefly, cells were passaged on GCB agar, and a single piliated colony was streaked onto a fresh GCB plate and allowed to grow overnight. Cells were scraped from the plate and gently resuspended to a cell density of 108/ml (optical density at 560 nm, 0.18) in prewarmed GC broth containing 10 mM MgCl2 and supplements I and II . Following addition of NaHCO3 to a final concentration of 10 mM, aliquots (900 mul) of diluted cells were mixed with 100 mul of 5- to 50-mug/ml donor DNA in 1x SSC (0.15 M NaCl plus 0.015 M sodium citrate) and incubated for 5 h at 37C under a humidified 5% CO2 atmosphere. Cells were plated on GCB agar with the appropriate antibiotic and grown for 24 to 48 h at 37C under 5% CO2. Alternatively, cells at 104, 103, or 102 CFU/ml were incubated with 5 to 50 mug of DNA on GCB agar for 16 to 20 h and then transferred to selection plates . MIC assays. | N. gonorrhoeae colonies were suspended in GCB broth to a density of 104/mul, and 5 mul was spotted onto GCB agar plates containing increasing concentrations of penicillin (0.06, 0.125, 0.3, 0.35, 0.4, 0.5, 0.75, and 1.0 mug/ml) or tetracycline (0.156, 0.313, 0.45, 0.625, 0.75, 1.0, and 1.5 mug/ml). The plates were then incubated at 37C under 5% CO2 for 24 h. Alternatively, cells were resuspended in GCB broth, and 1,000 colonies were spread to GCB agar plates containing increasing concentrations of either penicillin or tetracycline. The MIC was defined as the minimal concentration of antibiotic at which no more than 5 colonies were observed after 24 h. The two methods gave very similar results. MICs are presented as average values from at least four separate experiments. Growth assay. | N. gonorrhoeae cells were plated onto GCB agar plates from frozen stocks and grown overnight at 37C under 5% CO2. Nonpiliated colonies were passaged on GCB plates and grown for 16 to 20 h. Cells were gently scraped from the plates and suspended at a final density of similar14 Klett units per 10 ml of GCB broth plus supplements I and II. Side-arm flasks containing the bacterial suspensions were shaken at 37C under 5% CO2, and cell densities were measured every hour with a Klett reader. SDS-PAGE and Western blotting. | N. gonorrhoeae cells were scraped from GCB agar plates and diluted to 5 x 107 CFU/ml in GC broth. The cells were pelleted, suspended in 45 mul of 1x Laemmli buffer, and heated at 80C for 5 min. Four microliters of the cell lysates was loaded onto two 10% polyacrylamide gels, and total proteins were separated via sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). One of the gels was stained with Coomassie brilliant blue R-250, while the other gel was electroblotted onto nitrocellulose membranes in transfer buffer (25 mM Tris, 192 mM glycine, 10% methanol, 0.01% SDS [pH 8.3]). The membrane was probed sequentially with H5 monoclonal antisera, which recognize FA1090 porin protein (a gift from Janne Cannon, UNC-CH), and an anti-mouse secondary antibody conjugated to horseradish peroxidase. Porin proteins were visualized by incubation with Supersignal chemiluminescent substrate (Pierce/Endogen, Rockford, Ill.). Outer membranes. | N. gonorrhoeae outer membranes were purified as described by Lamden and Heckels . Briefly, cells were scraped from GCB agar plates and suspended in 5 ml of lithium acetate buffer (0.2 M lithium chloride, 0.1 M sodium acetate, 0.01 M EDTA [pH 5.8]). The cells were broken by 15 passages through a 22-gauge needle. Cell debris was pelleted at low speed (11,000 x g), and outer membranes were pelleted at 100,000 x g for 2 h. Protein levels were determined via a modified Bradford assay. Four micrograms of outer membrane proteins in 1x Laemmli sample buffer was incubated either at room temperature or at 80C for 5 min and subjected to SDS-PAGE. Western blotting was performed to detect monomeric and trimeric forms of the porin proteins by use of the H5 monoclonal antisera to the FA1090 porin protein as described for whole-cell lysates. RESULTS : Identification of specific mutations in FA140 PIB that confer intermediate-level resistance to penicillin and tetracycline. | The PIB proteins of strains FA1090 and FA140 (penA2 mtr penB2) differ by 10 amino acids , but only the porin from FA140 confers intermediate levels of penicillin and tetracycline resistance. To determine which of these 10 amino acid differences is (are) responsible for conferring resistance to penicillin and tetracycline, we created silent unique restriction sites surrounding the sequences encoding these amino acids to allow construction of porin chimeras and to aid in identifying strains transformed with the recombinant genes . Chimera 1 was created by fusing codons 1 to 187 of FA140 porin (containing three alterations) to codons 188 to 349 (containing seven alterations) of FA1090 porin, while chimera 2 was the reverse, beginning with the PIB sequence from FA1090 and switching to the FA140 sequence at codon 188 . A novel strain of N. gonorrhoeae, FA19 penA4 mtr (PIB), which contains the first two resistance determinants from FA6140 and the porIB gene from FA1090, was constructed to serve as the recipient strain for transformation experiments with the chimeras (as described in Materials and Methods). FA19 penA4 mtr (PIB) was transformed with the porIB chimeras, and transformants were selected on GCB agar plates containing 0.55 mug of tetracycline/ml. Transformation with genomic DNA isolated from strain FA140 (positive control) gave rise to resistant colonies at a frequency of 8 x 10-5. Transformation with chimera 1 (FA14019-187/FA1090188-349porIB) resulted in the growth of tetracycline-resistant colonies at a frequency of 4 x 10-5, whereas transformation with chimera 2 (FA109019-187/FA140188-349porIB) failed to generate resistant colonies. Thus, only transformation with the chimera encoding the FA140 porIB N-terminal sequence resulted in tetracycline-resistant colonies on selection plates. Nine of 10 porIB genes amplified by PCR from separate transformants showed the expected digestion pattern with HindIII and EcoRI, confirming that genomic recombination had occurred. Sequencing of the PIB gene from the one clone that did not show the expected digestion pattern revealed that it differed from the PIB of the recipient strain by only 2 amino acids, G120D and A121D. MICs of penicillin and tetracycline for each of the 10 transformants were the same as for strains obtained by transformation with the complete FA140 sequence . Thus, alteration of the amino acids at positions 120 and 121 (Gly and Ala, respectively) to aspartic acid is sufficient to confer the same level of resistance to penicillin and tetracycline as the PIB from FA140. To verify that other residues in FA140 PIB do not contribute to the penB phenotype, FA19 penA4 mtr (PIB) was transformed with a plasmid containing the FA1090/FA140 porIB chimera (chimera 2) and a linked erythromycin resistance gene (ermC) to aid in the selection of transformed colonies. The ermC gene was incorporated into the plasmid downstream of the porIB stop codon and was flanked on its 3' end by 300 bp of N. gonorrhoeae sequence. Recombination within the por gene in the erythromycin-resistant transformants was confirmed by PCR amplification and restriction digestion. The resistance to penicillin and tetracycline of the resultant strain was essentially identical to that of the recipient strain . Thus, these data further confirm that mutations at positions 120 and 121 are the only changes capable of conferring increased resistance. To determine whether aspartic acid mutations at both residues were necessary to confer full resistance to penicillin and tetracycline, FA1090 porIB constructs containing either a single G120D or a single A121D amino acid alteration and the flanking ermC gene were generated and used to transform FA19 penA4 mtr (PIB) to erythromycin resistance. Following confirmation by sequencing that the mutations had recombined into the genome, MIC analysis revealed that mutation of either Gly-120 or Ala-121 in PIB to aspartic acid conferred a moderate increase in resistance to penicillin and tetracycline over that of the recipient strain . However, resistance to either antibiotic was 1.5-fold lower than that of strains harboring aspartic acid mutations at both positions. Therefore, both the G120D and the A121D mutation are necessary to confer full resistance to penicillin and tetracycline. Isolation of penicillin- and tetracycline-resistant strains transformed with porIB DNA containing randomized codons at positions 120 and 121. | To determine the importance of specific amino acid mutations in PIB proteins at residues 120 and 121, we randomized the codons at positions 120 and 121 by multiplex PCR and used these constructs to transform the recipient strain to increased penicillin or tetracycline resistance. Transformants were selected on GCB plates containing either 0.57 mug of tetracycline/ml or 0.37 mug of penicillin/ml, and 20 isolates were chosen for further analysis. The MIC of penicillin for each transformant was 0.75 mug/ml, identical to that for the strain with G120D and A121D double PIB mutations. However, the MIC of tetracycline for all 20 transformants was 1.5 mug/ml, compared with 1.0 mug/ml for the strain with the double aspartic acid PIB mutations. The porIB genes from each of the 20 resistant transformants were amplified by PCR and sequenced. Eighteen porIB genes harbored a lysine codon at residue 120, while one other had an arginine at that position . Eight different charged and uncharged amino acids were found at residue 121. We also isolated one transformant with prolines at both residues 120 and 121. Overall, these data suggest that mutations at positions 120 and/or 121, particularly to charged amino acids, confer intermediate-level resistance to penicillin and tetracycline. Analysis of charged and uncharged amino acid mutations at positions 120 and 121 in PIB. | We also mutated Gly-120 and Ala-121 to several charged amino acids to investigate further the role of charged side chains in conferring resistance to penicillin and tetracycline. FA19 penA4 mtr (PIB) was transformed with porIB constructs containing either a G120H, G120E, G120K, or A121K mutation and the linked erythromycin resistance gene, and cells were selected on plates containing 6.0 mug of erythromycin/ml. The MICs of penicillin and tetracycline for the strain containing the G120H PIB mutation were essentially identical to those for the recipient strain, whereas the MICs for the strain with the G120E PIB mutation were similar to those of strains harboring either the G120D or the A121D single PIB mutation . Mutation of Gly-120 to Lys conferred the same level of resistance to the recipient strain as each of the random mutants containing a G120K mutation, whereas mutation of Ala-121 to Lys conferred only moderate resistance to both antibiotics . These data demonstrate that a single mutation of G120 to lysine in PIB is the only change needed to confer full intermediate-level resistance to penicillin and tetracycline, which is consistent with the results from our randomized screen. To investigate the consequences of side chain charge versus size in mediating resistance to penicillin and tetracycline, PIB constructs with uncharged amino acids at positions 120 and 121 (G120L A121L or G120N A121N) were transformed into the recipient strain. Leucine or asparagine residues were chosen because their sizes closely resemble those of their charged counterparts lysine and aspartic acid, respectively. The MICs of penicillin and tetracycline for the strains with these double PIB mutations were essentially identical to those of the recipient strain (data not shown), demonstrating that the charge of the side chain, and not its size, is the most important attribute for conferring resistance to the two antibiotics. Rate of growth of N. gonorrhoeae strains. | Because mutations in residues 120 and 121 apparently alter the flux of antibiotics across the outer membrane, we determined whether these alterations might also alter the passage of metabolites. Therefore, we compared the growth rate of the recipient strain, FA19 penA4 mtr (PIB), with those of isogenic strains containing either the G120K or the G120D A121D PIB mutations. All strains exhibited similar growth curves and growth rates in GCB broth over three separate experiments (data not shown). Doubling times were 1.27 +- 0.06 h for the strain with the FA1090 PIB, 1.13 +- 0.06 h for the G120D A121D PIB mutant, and 1.18 +- 0.03 h for the G120K PIB mutant. Therefore, mutations at residues 120 and/or 121 in PIB do not appear to compromise the growth rate of the cell, at least in GCB medium. Determination of porin expression levels in penB strains. | To determine whether the mutations in PIB altered either the level of porin expression or the propensity of the porin monomers to form trimers at the cell surface, we compared the levels of PIB protein expressed in strain FA19 penA4 mtr (PIB) to those in strains expressing PIB with G120K, G120D A121D, G120K A121R, or G120P A121P mutations. Coomassie blue staining and Western blotting (data not shown) of whole-cell lysates showed no detectable differences in PIB expression among the tested strains over several experiments. To test the effects of the mutations on the integrity of porin trimers, outer membranes from the recipient and penB strains were incubated in SDS-PAGE loading buffer either at room temperature or at 80C and subjected to SDS-PAGE. Protein staining with Coomassie blue indicated that porin trimers were present in equal amounts in all of the strains examined when they were incubated at room temperature and completely dissociated into monomers at 80C . These results were further confirmed by immunoblotting (data not shown). Taken together, these data suggest that resistance to penicillin and tetracycline in penB strains is not a result of altered porin levels or the propensity of the porin to form trimers at the cell surface but is presumably due to alterations in the porin channel itself. The G120K mutation in PIA confers increased penicillin and tetracycline resistance. | N. gonorrhoeae has two porins, PIA and PIB, whose genes are alleles of a single locus . To date, increased resistance to penicillin and tetracycline in penB isolates has been correlated only with alterations in PIB and not with alterations in PIA. Since these proteins are 67% identical, we investigated whether mutation of Gly-120 to lysine in loop 3 of PIA, which is the residue equivalent to Gly-120 in PIB , could also confer resistance to penicillin and tetracycline. FA19 penA4 mtr was transformed with porIA mutant DNA, and transformants were selected at a frequency of 1.6 x 10-5 on GCB plates containing 0.24 mug of penicillin/ml. MIC analysis of the transformed strain revealed that while the G120K mutation in PIA increases both tetracycline and penicillin resistance over that of the recipient strain, it does not increase resistance to the same levels as an isogenic strain harboring the G120K mutation in PIB . This was especially evident in the levels of tetracycline resistance. FA19 penA4 mtr was also transformed with porIA G120D G121D DNA. The resultant strain showed only a small increase in resistance to penicillin and no increase in resistance to tetracycline . These data suggest that structural differences in loop 3 or potentially in the barrel regions between PIB and PIA are critical for these mutations to increase resistance. Identification of PIB amino acid mutations at residues 120 and 121 in clinical strains. | To determine the prevalence of naturally occurring charged amino acids at PIB positions 120 and 121, we examined the loop 3 sequences of 51 clinical isolates whose por sequences are present in the GenBank database . These strains represented 14 distinct PIB serovars and separate geographical locations, including the United States, Singapore, Great Britain, and Kenya. Of these, 15 strains (29.4%) had lysine and aspartate codons at positions 120 and 121, respectively, 6 strains (11.8%) had an aspartate residue at position 120 or 121, and 3 strains (5.9%) harbored aspartate codons at both positions. We also analyzed the loop 3 sequences of 45 PIA proteins. At residue 120, 21 strains (46.6%) had an aspartate residue whereas the rest of the isolates had a glycine. Only a glycine was seen at position 121. These data indicate that mutations of residues 120 and/or 121 are common in clinical isolates, especially mutations to charged amino acids. FIG. 1. | FA1090 and FA140 PIB proteins. FA1090 and FA140 PIB proteins. (A) HindIII, EcoRI, and ClaI silent restriction sites were inserted at codons 110, 187, and 274, respectively, flanking the amino acid differences between the PIB proteins of strains FA1090 and FA140. Residues 1 to 19 constitute the cleavable signal sequence. (B) porIB chimeras were generated via DNA sequence exchange at the EcoRI site. FIG. 2. | MICs of penicillin and tetracycline for N. gonorrhoeae MICs of penicillin and tetracycline for N. gonorrhoeae strains harboring porIB chimeras. Solid bars represent MICs for the recipient strain, FA19 penA4 mtr (PIB), and the isogenic strain, FA19 penA4 mtr penB, which contains the PIB from strain FA140. Strains with PIB chimeras (striped bars) and strains with a single aspartate mutation at position 120 or 121 (open bars) were created by transformation of strain FA19 penA4 mtr (PIB) with mutant porIB DNA. MICs shown are averages from five experiments. Error bars, standard deviations. Dashed lines denote the respective MICs for the recipient strain. FIG. 3. | MICs of penicillin and tetracycline for N. gonorrhoeae MICs of penicillin and tetracycline for N. gonorrhoeae strains harboring single-amino-acid mutations in PIB. Strains with the single-amino-acid mutation G120H, G120E, G120K, or A121K in PIB were created by transformation of strain FA19 penA4 mtr (PIB) with altered porIB constructs. MICs are averages from five experiments. Error bars, standard deviations. Dashed lines denote the respective MICs for the recipient strain, FA19 penA4 mtr (PIB). FIG. 4. | SDS-PAGE analysis of PIB expression. SDS-PAGE analysis of PIB expression. (A) N. gonorrhoeae strains (5 x 107 CFU/ml) containing either the FA1090 (WT) or a mutant (K120, D120 D121, or P120 P121) PIB protein were incubated at 80C for 5 min in Laemmli loading buffer and subjected to SDS-PAGE. (B) Approximately 4 mug of outer membranes prepared from the strains described in the legend to panel A was incubated in Laemmli loading buffer at either 23 or 80C for 5 min prior to electrophoresis. Proteins were stained with Coomassie brilliant blue. Arrows indicate porin monomers (M) and trimers (T). FIG. 5. | Alignment of PIA and PIB protein sequences. Alignment of PIA and PIB protein sequences. PIA and PIB proteins from strains FA19 and FA1090, respectively, were aligned with the Clustal-X program . The figure was generated from the aligned sequences by the Boxshade program (available on the World Wide Web). FIG. 6. | MICs of penicillin and tetracycline for N. gonorrhoeae MICs of penicillin and tetracycline for N. gonorrhoeae strains with PIA mutations. Strains harboring PIA with either a G120D A121D double mutation or a G120K single mutation were created via transformation of strain FA19 penA4 mtr with the altered porIA DNA. MICs are averages from four experiments. Error bars, standard deviations. MICs for the strain with the G120K PIB mutation are shown for comparison. Dashed lines indicate the respective MICs for the recipient strain, FA19 penA4 mtr. TABLE 1 | Amino acid residues at positions 120 and 121 in PIB proteins from strains isolated from transformation of FA19 penA4 mtr (PIB) with porIB DNA harboring randomized codon mutations at these positions DISCUSSION : The focus of this study was to investigate the structure-function aspects of porin-mediated resistance in N. gonorrhoeae. We showed that alteration of residues 120 and 121 of PIB to aspartic acids confers intermediate-level resistance to penicillin and tetracycline. We also showed that a single-amino-acid mutation in PIB, G120K, is sufficient to confer full intermediate-level antibiotic resistance when expressed in a penA mtr recipient strain. Moreover, lysine and aspartic acid residues were found at positions 120 and/or 121 in the PIB proteins of approximately 47% of clinical isolates in the GenBank database, indicating that these changes occur in nature. Our data demonstrated that mutations in PIB at positions 120 and/or 121, particularly to charged amino acids, are capable of conferring antibiotic resistance in N. gonorrhoeae. A single Asp mutation at either position 120 or 121 conferred only partial resistance to the antibiotics, showing that Asp mutations at both these positions were necessary for conferring the same level of resistance as the FA140 PIB. Moreover, mutation of Gly-120 to Glu conferred similar levels of resistance to the recipient strain as PIB containing a single Gly-120-to-Asp mutation. Although we did not examine the effects of double Glu mutations at positions 120 and 121 on resistance, we suspect that this mutant would show levels of resistance similar to those of the double Asp mutant. In contrast, only a single G120K mutation was necessary to confer the same level of penicillin resistance and even higher levels of tetracycline resistance to the recipient strain. The latter observation explains both the abundance of lysine and the lack of aspartates in the randomized library screen. The expected frequency of selecting a single lysine codon at position 120 is 1 of 32 (since 2 of 64 codons encode lysine) compared to 1 of 1,024 for transformants harboring aspartic acid codons at both positions [the product of the frequency of aspartate codons at each position, (1/32)2]. Gram-negative bacteria have developed two main porin-mediated mechanisms to increase resistance to antibiotics. One of these mechanisms, in which the bacteria become resistant to antibiotics following a loss of porin expression, is observed in many gram-negative bacteria, including Pseudomonas aeruginosa and Klebsiella pneumoniae . Because certain antibiotics diffuse selectively through a particular porin, loss of that porin species increases resistance to these antibiotics. The second porin-mediated mechanism, as described in this study, utilizes structural alterations in porin channels that presumably decrease the flux of antibiotics across the outer membrane. The latter mechanism is utilized by N. gonorrhoeae because it expresses only a single porin allele, and gonococci lacking porin are not viable . Interestingly, Enterobacter aerogenes has been reported to increase resistance to antibiotics via both mechanisms . Previous protein sequence alignments of N. gonorrhoeae PIB with other porins placed residues 120 and 121 in loop 3, which folds into the lumen of the barrel and constricts the porin channel . Recently, the structure of the Comomonas acidovorans Omp32 porin, which has higher homology to gonococcal PIB than other porins whose structures are known, was solved at a resolution of 2.1 A . The amino acids of the C. acidovorans Omp32 monomer that correspond to N. gonorrhoeae PIB residues 120 and 121, as determined from an alignment of these two porin protein sequences, are Thr-102 and Ser-103. In the crystal structure of Omp32, these two residues are located on a small alpha-helix in loop 3 with their side chains pointing into the channel, in an ideal position to affect the permeation of antibiotics. Gill et al. predicted that residues in loop 3 of PIB were important for mediating resistance to antibiotics, although no supporting evidence was presented. We have confirmed and extended their work, and we show that residues 120 and 121 in loop 3 are responsible for conferring porin-mediated resistance to both penicillin and tetracycline. Crystal structures of Rhodobacter capsulatus and E. coli porins were the first to show that loop 3 folds into the barrel and constricts pore size . The importance of loop 3 in the permeation, ion conductance, and ion selectivity of several porins has been investigated in detail. These studies have related alterations in the constriction loop to changes in pore properties. Specifically, studies with the phosphate starvation-induced E. coli porin PhoE revealed that mutation of a single lysine residue (Lys-125) to glutamic acid in loop 3 reverses ion selectivity from anionic to cationic and inhibits interaction with polyphosphate . Deletions and site-directed mutagenesis of loop 3 in E. coli OmpF and OmpC resulted not only in changes in ion selectivity but also in changes in sugar permeation and antibiotic susceptibility . As with the PIB of N. gonorrhoeae, mutations in loop 3 of E. aerogenes porin conferred reduced susceptibility to cephalosporins . In vitro studies with E. coli porins have shown that mutations in the barrel also confer altered antibiotic susceptibilities and pore properties . It is not clear whether residues other than 120 and 121 in PIB contribute to antibiotic resistance in clinical strains of N. gonorrhoeae; however, our data suggest that mutations in loop 3 are the predominant means by which intermediate-level resistance to penicillin and tetracycline is conferred in penB strains. Antibiotic resistance in both laboratory and clinical strains of N. gonorrhoeae has been linked to PIB serovars and not to PIA . This may be because an aspartate residue at position 120 in PIA confers only modest resistance to penicillin and no resistance to tetracycline over that of wild-type PIA. However, the mechanism underlying the differential susceptibility to tetracycline and penicillin observed in this mutant is currently unknown. The PIA sequences from clinical isolates in the GenBank database show only a glycine or an aspartate at this position. However, since a G120K mutation in PIA confers modest levels of resistance to penicillin and tetracycline over those of wild-type PIA, it is intriguing that a lysine mutation has not developed naturally in PIA proteins. Reasons for this may include biological differences between PIA and PIB proteins, which may contribute to differing selection pressures, including infection localization and serum resistance properties . The mechanism(s) by which the mutations in loop 3 of N. gonorrhoeae PIB confer antibiotic resistance is presently unknown. Gill et al. speculated that aspartic acid residues in loop 3 of the FA140 PIB decrease antibiotic flux via anionic repulsion. However, we believe this mechanism is unlikely, since we show in this study that a basic lysine residue at position 120 also mediates intermediate-level resistance to both penicillin and tetracycline. Our data suggest that mutations at residues 120 and/or 121 in PIB alter the conformation of loop 3 to decrease the flux of antibiotics into the periplasmic space. This model is further supported by the recovery of a penB transformant with proline residues in this region, which are known to perturb protein structure due to the incorporation of a secondary amide linkage into the peptide backbone. Previous studies also have shown that expression of the penB phenotype in the gonococcus depends on prior acquisition of the mtr resistance gene . Electrophysiological and biochemical studies of purified PIB proteins to address the mechanism of resistance to penicillin and tetracycline conferred by mutations in PIB and the apparent synergism between mtr and penB are currently in progress. Backmatter: PMID- 12183279 TI - Incidence of High-Level Evernimicin Resistance in Enterococcus faecium among Food Animals and Humans AB - Six high-level evernimicin-resistant Enterococcus faecium isolates were identified among 304 avilamycin-resistant E. faecium isolates from animals and 404 stool samples from humans with diarrhea. All four animal isolates, and one of the human isolates, were able to transfer resistance to a susceptible E. faecium strain. The resulting transconjugants all tested positive for the presence of emtA, a gene encoding a methyltransferase previously linked with high-level evernimicin resistance. The four transconjugants derived from animal isolates all carried the same plasmid, while a differently sized plasmid was found in the isolate from humans. This study demonstrated a low incidence of high-level evernimicin resistance mediated by the emtA gene in different E. faecium isolates of animal and human origin. Keywords: Introduction : There is a growing concern regarding the emergence of multiply antimicrobial-resistant enterococci as important nosocomial pathogens . This has increased the interest in either developing new antibiotics or modifying older antibiotics with activity against multiply resistant staphylococci and enterococci. One of these agents is evernimicin, an oligosaccharide antibiotic with activity against a broad range of gram-positive pathogenic bacteria including glycopeptide-resistant enterococci, methicillin-resistant staphylococci, and penicillin-resistant streptococci . Evernimicin inhibits protein synthesis in Staphylococcus aureus by binding with high affinity to a single site on the 50S subunit . The evernimicin binding site overlaps the binding site of another oligosaccharide antibiotic, avilamycin. Avilamycin has been used as a growth promoter for food animals in the European Union for several years, and as a consequence avilamycin resistance has been observed among Enterococcus faecium isolates from broilers in Denmark and other countries . High-level (>64 mug/ml) avilamycin resistance is mediated by mutations in the gene encoding ribosomal protein L16; these strains also exhibit low-level (MIC, 2 to 12 mug/ml) cross-resistance to evernimicin. . Recently, a new methyltransferase, which confers high-level (MIC, >64 mug/ml) evernimicin resistance through methylation of 23S rRNA, was cloned from an avilamycin-resistant E. faecium strain isolated from a broiler in Denmark . The present study was conducted to determine the occurrence of high-level evernimicin resistance among avilamycin-resistant enterococcal isolates from food animals and humans in Denmark. A total of 304 avilamycin-resistant (MIC, >=16 mug/ml) E. faecium and 14 Enterococcus faecalis isolates were obtained from the continuous monitoring for antimicrobial resistance among broilers and pigs in Denmark . These isolates were screened for high-level (MIC, >128 mug/ml) evernimicin resistance by the E-test according to the manufacturer's guidelines (AB Biodisk, Solna, Sweden); four positive isolates were identified. In March 1998 a total of 254 human stool samples, submitted for examination of diarrheal pathogens, were examined for the presence of avilamycin-resistant enterococci by plating a loopful of fecal material on Mueller-Hinton II agar plates containing 20 mug of avilamycin/ml. All patients sampled had a history of diarrhea but no history of either a recent hospital stay or antimicrobial treatment. One avilamycin-resistant E. faecium isolate was identified; this isolate also exhibited high-level evernimicin resistance. In June 2001 an additional 150 fecal samples from humans were, in relation to another study, screened for the presence of enterococcal isolates resistant to erythromycin and tetracycline with Slanetz and Bartley agar plates containing erythromycin (20 mug/ml) and tetracycline (10 mug/ml). Isolates obtained from these samples were also screened for avilamycin resistance; one isolate was recovered which also tested positive for high-level evernimicin resistance. All six isolates tested positive in PCR for the emtA gene with the following oligonucleotides as primers: 5'-GGTCAGCAGATCACTTGTTT-3' and 5'-TGAACAATTCTAAGTCCTCG-3'. The six evernimicin-resistant isolates were subjected to pulsed-field gel electrophoresis (PFGE) typing with the restriction enzyme SmaI. DNA purification and enzyme digestion were performed as previously described . Four different PFGE types were identified: three of the four broiler isolates belonged to the same PFGE type, and the two isolates from humans gave different PFGE patterns . We screened all six isolates for the ability to transfer evernimicin resistance to E. faecium BM4105 (resistant to rifampin and fusidic acid) by the filter mating procedure described previously . Transconjugants were isolated from within the inhibition zone of an E-test strip laid down on Mueller-Hinton II agar plates containing rifampin (50 mug/ml) and fusidic acid (10 mug/ml). All four animal isolates and one of the human isolates yielded transconjugants. Transconjugants were screened for susceptibility to avilamycin, bacitracin, chloramphenicol, erythromycin, gentamicin, kanamycin, penicillin, streptomycin, quinupristin-dalfopristin, tetracycline, and vancomycin as previously described . Resistance to avilamycin was cotransferred, but no other drug resistance markers appeared to be cotransferred. Plasmid DNA was extracted from the five transconjugants with the Qiagen Plasmid Midi kit (Qiagen, Valencia, Calif.) and restricted with either EcoRI or PvuII. EcoRI digestion of the plasmids from the four broiler isolates yielded three different bands, while PvuII digestion yielded five bands. The plasmids from the four broiler isolates were approximately 36 kb and indistinguishable even though the four isolates fell into two different PFGE patterns. In contrast, the human isolate harbored a plasmid of >100 kb. EcoRI digestion yielded 22 bands, and PvuII digestion yielded 18 bands. To determine if high-level evernimicin resistance was linked to the presence of the emtA gene , we performed a Southern blot analysis on the restricted plasmids . A digoxigenin-labeled emtA fragment was prepared by PCR with the oligonucleotides described above. The emtA probe hybridized to DNA fragments of approximately 6.9 and 12.5 kb in the EcoRI and PvuII restriction digests, respectively, of the transconjugants from the four broiler isolates (data not shown). In contrast, the emtA probe hybridized to fragments of 3.3 and 4.5 kb in the EcoRI and PvuII digests of the transconjugant of the human isolate. FIG. 1. | PFGE profile of six high-level evernimicin-resistant E. PFGE profile of six high-level evernimicin-resistant E. faecium isolates from humans and broilers. Lanes 1 and 8, 50-kb ladder; lane 2, isolate 1 from humans; lane 3, isolate 2 from humans; lanes 4 to 7, isolates 1 to 4, respectively, from broilers. The present study demonstrated that high-level evernimicin resistance has not been widely disseminated among E. faecium strains isolated from the animal or human population in Denmark. In the four isolates from broilers, three of which appeared clonal, the emtA gene was carried by a similarly sized plasmid in each case. In contrast, a plasmid of different size was observed in the single human isolate capable of transferring evernimicin resistance. Since the emtA gene is part of a transposon , it is possible that it was transposed from one plasmid to a different plasmid. Several studies have indicated that bacteria from animals and humans share the same resistance genes and that exchange probably occurs fairly frequently . This includes resistance genes in enterococci where the genes encoding resistance to important human antibiotics such as vancomycin (vanA) and quinupristin-dalfopristin [vat(E) and erm(B)] probably have spread from the animal reservoir to humans and thereby perhaps shortened the life span of these antibiotics . Backmatter: PMID- 12183222 TI - Resistance to Macrolides and Related Antibiotics in Streptococcus pneumoniae AB - Keywords: null: null THE MACROLIDES : Macrolides have a common structure formed by a large lactone ring. Erythromycin is a mixture of antibiotics that includes erythromycin A, which is the active compound and which has a 14-membered lactone ring with two sugars, l-cladinose and an amino sugar. Other commercially available macrolides derived from erythromycin A include clarithromycin, dirithromycin, roxithromycin, and azithromycin, which has an enlarged, 15-membered ring resulting from a nitrogen insertion. The structural modifications of erythromycin A resulted in improved pharmacokinetic profiles and better tolerance, but cross-resistance between members of this class of antimicrobial agents was still observed. Certain 16-membered macrolides are also available in a few countries (spiramycin, josamycin, midecamycin, and miocamycin) or for veterinary use (tylosin). The recently developed ketolides telithromycin and ABT773 are derived from clarithromycin and have two major modifications, replacement of l-cladinose by a keto function and an 11- to 12-carbamate extension with an arylalkyl modification in telithromycin, the latter of which may partially explain its increased intrinsic activity and activity against erythromycin-resistant strains, as discussed below . In telithromycin and ABT773, modification at the C-6 position prevents inactivation of the molecule in acid medium. RIBOSOME BINDING SITE AND MODE OF ACTION OF ERYTHROMYCIN : The ribosome structure and contact points between the ribosome and erythromycin A were recently identified by crystallography studies . The bacterial ribosome is formed by a small, 30S subunit and a large, 50S subunit. The latter is composed of 23S rRNA and of a minimum of 30 proteins. The secondary structure of 23S rRNA is folded due to base pairing and forms six domains numbered I to VI, while the tertiary structure of the molecule is maintained by its interactions with proteins. Stoichiometric binding of erythromycin A to the 50S subunit causes inhibition of protein synthesis. The binding site of erythromycin is composed of domain V sequences near the peptidyltransferase center, where the polypeptide chain is synthesized. Hairpin 35 in domain II is in the vicinity of this binding site . High-resolution X-ray structures of the 50S ribosomal subunit of Deinococcusradiodurans complexed with erythromycin A showed that the 2'-OH group of the desosamine sugar of the antibiotic appears to form three hydrogen bonds with adenines at positions 2058 and 2059 (Escherichiacoli numbering) . The dimethylamino group of the desosamine sugar also appears to interact with A2505. The 6-OH of the lactone ring may form a hydrogen bond with A2062, the 11-OH and 12-OH may form one hydrogen bond with U2609, but the cladinose sugar does not seem to be involved in interactions with 23S rRNA. Although footprinting experiments have implicated adenine at position 752 (domain II) in the binding of erythromycin, no direct interaction has been shown between the two structures, at least in the ribosome of D. radiodurans . The binding site of erythromycin A is located within the tunnel that serves as a channel for the growing peptide. The surface of this tunnel is formed by domains I to V of 23S rRNA, by several ribosomal proteins including the globular structures of ribosomal proteins L22 and L4, and by a beta hairpin of L22 . Erythromycin does not inhibit the peptidyltransferase activity but prevents the extension of the peptide chain by blocking the polypeptide exit tunnel and provokes the premature release of peptidyl-tRNA . Moreover, erythromycin also prevents ribosomal assembly at an early stage of protein synthesis . MECHANISMS OF RESISTANCE TO MACROLIDES : A common mechanism shared by bacteria for becoming resistant to antimicrobial agents is the diminution of the affinity of the antibiotic for its target. This effect may result from enzymatic detoxification of the drug or, conversely, from target modification. A third possibility is diminished access to the target secondary to active efflux or decreased uptake of the molecules. The resistance of Streptococcuspneumoniae to erythromycin is due to modification of the ribosomal target by methylation or mutation and active efflux of the drug; drug modification has not been reported in this species. RIBOSOMAL METHYLATION: THE MLSB RESISTANCE PHENOTYPE : As already mentioned, ribosomal modification by methylation was the first mechanism of resistance to erythromycin elucidated and remained unique for decades. It is secondary to the acquisition of an erm gene (erythromycin ribosome methylase) usually carried by transposable elements in pneumococci. This gene encodes a ribosomal methylase which dimethylates pneumococcal 23S rRNA at a single site, adenine at position 2058 . As previously alluded to, the A2058 nucleotide is a key nucleotide for the binding of erythromycin. The modification markedly reduces the affinity of erythromycin for its target, probably by preventing direct access to the target or by modifying the conformation of the binding site. Cross-resistance to macrolides, lincosamides, and streptogramin B antibiotics , which gave its name to the MLSB resistance phenotype, is due to the overlapping binding sites of the drugs . TABLE 1 | Macrolide-lincosamide-streptogramin B resistance in S. pneumoniae due to gene acquisition erm DETERMINANTS : The erm(B) determinant, initially called erm(AM), was first characterized on plasmid pAM77 in Streptococcussanguis A1 isolated from dental plaque in 1978 . The gene is widely distributed, not only in S. pneumoniae but also in a variety of other streptococcal and enterococcal species, in enterobacteria, and in staphylococci, indicating easy exchange of genetic information even between phylogenetically remote species. In pneumococci, the gene is borne by conjugative transposons related to Tn1545, Tn1545-like elements, or a Tn917-like element that is part of a larger composite transposon, Tn3872 . Transposition occurs from chromosome to chromosome of strains of S. pneumoniae. Both clonal spread of resistant strains and horizontal transfer of the element account for the high prevalence of the erm(B) gene in erythromycin-resistant pneumococci in certain countries. In one study, sequences homologous to the structural gene for the integrase of Tn1545, an enzyme required for the movements of the element, were found in all 36 S. pneumoniae strains resistant to erythromycin studied . Strains belonging to the 23F or 6B lineage appear to have erm(B) as part of Tn3872 or a modified form of Tn916 and Tn1545. Tn1545-like elements may also be exchanged between pneumococci by transformation. However, this mode of transfer, which is considered essential for the spread of beta-lactam resistance by alteration of genes for penicillin binding proteins in pneumococci, has not been shown for erythromycin resistance. Although widely predominant, erm(B) is not the only representative of the erm gene class in pneumococci. The presence of an erm(A) gene has been reported for a single strain, isolated in Greece, to which it conferred cross-resistance to erythromycin and clindamycin and for one strain with a resident erm(B) gene . This determinant, first detected in Streptococcuspyogenes, was initially designated ermTR and was subsequently included in the erm(A) gene class because of its close relatedness to erm(A) in Staphylococcusaureus . REGULATION OF erm(B) EXPRESSION AND THE MLSB RESISTANCE PHENOTYPE : The methylase encoded by erm(B) may be constitutively or inducibly synthesized. When expression is constitutive, the erm(B) mRNA is active, and its translation by the ribosomes allows constitutive methylation of the ribosomes, probably while they are synthesized . When resistance is inducible, erm(B) mRNA is synthesized, but in an inactive conformation, and becomes active only in the presence of inducing macrolides. Although for erm(B) the mechanism of induction has not been thoroughly studied, a model which can be inferred from the translational regulation model of erm(C) in S. aureus has been proposed and can be summarized as follows. The 5' end of erm(B) presents a series of inverted repeats which are responsible for the lack of methylase synthesis in the absence of erythromycin . Fourteen pairs of repeats have been identified which could form alternative stem-loop structures by base pairing . As shown in Fig. , one of these stem-loops sequesters the ribosome binding site and initiation codon for the methylase. Thus, the methylase cannot be produced, since the initiation motifs for translation of the enzyme are not accessible to the ribosomes. Induction is related to the presence of sequences coding for a small leader peptide of 36 amino acids upstream from the methylase gene. In the presence of low concentrations of erythromycin, binding of the antibiotic to a ribosome translating the leader peptide causes the ribosome to stall, in turn destabilizing the pairing of the inverted repeats and inducing conformational rearrangements in the mRNA. In particular, displacement of the stem-loop shown in Fig. unmasks the initiation sequences for the methylase, allowing synthesis to proceed by the ribosomes that are not complexed with erythromycin or by those that are methylated. Methylation of some ribosomes might occur through transient rearrangements of the stem-loop structures, which would lead to the synthesis of a basal level of the methylase. For a given erm gene, the inducing capacity of the macrolides depends on the antibiotic structure. The global structure of the drug, rather than the number of atoms in the lactone ring, determines the inducing capacity of a macrolide. As an example, erythromycin is an inducer for the production of most Erm methylases, whereas ketolides, which have a similar lactone ring, are not. A lack of inducing ability of ketolides has been related to the replacement of one of the erythromycin sugars, l-cladinose, by a keto function . It is likely that the intimate mode of action of a macrolide determines its capacity to act as an inducer, since proper ribosome stalling is required for the induction of methylase production. For erm(B), the commercially available macrolides (including the 14-, 15-, and 16-membered macrolides), lincosamides, and streptogramin B antibiotics are inducers of methylase synthesis to various degrees, leading to cross-resistance to these antimicrobial agents. It has been shown for erm(A) and erm(C), both in laboratory mutants and in clinical isolates, that constitutive expression is due to deletions, duplications, or point mutations in the attenuator sequence leading to derepressed production of the methylase . In pneumococci, the constitutive expression of MLSB resistance is infrequently found . However, despite the fact that the vast majority of pneumococci express erythromycin resistance inducibly, it has been shown by primer extension analysis of five strains that various proportions of ribosomes are methylated even in the absence of erythromycin . This paradox has been explained for certain strains by the presence of mutations in the stem-loop structure that sequester the initiation sequences for the methylase. Fusion of the mutated erm(B) attenuator with a lacZ reporter gene has confirmed that the expression of the methylase can be partly derepressed in certain strains . Other additional features, such as differences in the promoter strength or in the copy number of the erm(B) gene, may also account for the various levels of ribosomal methylation. FIG. 1. | Schematic representation of the structure of the mRNA from the inducible erm(B) gene from pAM77. Schematic representation of the structure of the mRNA from the inducible erm(B) gene from pAM77. The sequences of the control peptide (hatched box) and of the methylase [(erm(B)] are shown. Numbers 1 to 14 indicate inverted repeats with their symmetry axes (solid ovals flanked by broken lines). The secondary structure which is putatively formed by inverted repeat 14 and which would sequester the initiation sequence for the methylase in the absence of erythromycin is shown at the right. RBS, ribosome binding site. MACROLIDE EFFLUX : Physiological pumps conferring erythromycin resistance by efflux have been described for several gram-positive organisms, such as Cmr from Corynebacteriumglutamicum, which belongs to the major facilitator superfamily class of pumps , but not for S. pneumoniae. However, acquired resistance to macrolides conferred by active efflux has been detected recently in this species . The gene responsible for efflux was initially called mefE and was subsequently assigned to the mef(A) gene class because of its close relatedness to the mefA gene in S. pyogenes . The Mef(A) pump belongs to the major facilitator superfamily class. It contains 12 transmembrane domains spanning the cytoplasmic membrane, and efflux is driven by the proton motive force . Few substrates have been identified, and the pump seems to be specific to erythromycin and its derivatives, including azithromycin. Resistance appears to be induced with erythromycin and is expressed at moderate levels, with erythromycin MICs of between 1 and 64 mug/ml (generally between 8 and 32 mug/ml). Because the 16-membered macrolides, the lincosamides, and the streptogramin B antibiotics are not substrates of the pump, these antimicrobial agents remain active, even after induction with erythromycin. Resistance to erythromycin combined with susceptibility to clindamycin, whether the cells are induced or not induced with erythromycin, defines the M resistance phenotype. The mef(A) gene is transferable among pneumococci and is a member of a group of closely related, large transposable elements . Although the 7,244-bp transposon Tn1207.1 is apparently intact, it is defective for transfer , as is the 5.4- to 5.5-kb MEGA element, which is devoid of the transposase gene . Downstream from mef(A) lies a gene that putatively encodes an ATP binding cassette transporter and whose role in the expression of resistance remains questionable. The cloned mef(A) gene alone is sufficient to confer resistance, although it is not possible to exclude the possibility that the pump interacts with other proteins . STREPTOGRAMINS AND TELITHROMYCIN : Both ribosomal methylation and drug efflux alter the activities of erythromycin A and its derivatives. Several strategies should allow MLSB resistance in pneumococci to be overcome: the use of methylase or efflux inhibitors, synergistic combinations with another antimicrobial agents, and the development of noninducing macrolides or of macrolides that have alternative ribosome binding sites or that are not substrates for the efflux pump. Two types of drugs have been developed with activities against MLSB-resistant pneumococci, the streptogramins and the ketolides. The streptogramins (pristinamycin and quinupristin-dalfopristin) are composed of two streptogramin factors, A and B, with synergistic activity resulting from a dual interaction with the ribosome . As mentioned above, Erm methylation of the ribosome affects the activity of the B component. However, synergy is maintained, most probably because of the mode of action of the streptogramins. Although the mechanism for synergy is not fully understood, the binding of factor A to its target may induce a conformational change in the ribosome leading to an increase in its affinity for factor B . The ribosomal alteration must be sufficiently marked to overcome the loss of affinity for the B molecule that results from rRNA methylation. The bactericidal activity of the streptogramin combination against pneumococci is also generally conserved in vitro . The ketolides, like the macrolides, bind to the bacterial ribosome and exert their antibacterial effect by inhibition of protein synthesis. Despite the similarity between the macrolides and the ketolides, in terms of mechanism of action and therefore cross-resistance, recent data indicated that the ketolides have activity against MLSB-resistant pneumococci . This finding appears to be due to two differences from the macrolides: the strength and nature of ribosome binding and the weak ability of the ketolides to act as inducers of macrolide resistance . It has been shown that macrolides interact with two sites within the bacterial ribosome, domains II and V of 23S rRNA, with the interaction at domain II being relatively weak. The ketolides also interact with domains II and V but appear to have a 10-fold higher binding affinity . As discussed previously, MLSB resistance arises when the binding of the macrolides within domain V is compromised, principally through methylation. In contrast, the ketolides retain in part their ability to bind to MLSB-resistant ribosomes probably because of their stronger interaction with domain II. However, as previously mentioned, crystal studies with D. radiodurans 50S ribosomal subunits do not support the notion of direct contacts between the 14-membered macrolides and A752 or any other domain II residue . Alone, the increased ribosome binding property probably does not account for the activity of telithromycin against macrolide-resistant pneumococci. Another additional feature of the ketolides is their inability to induce MLSB resistance. Lack of induction of MLSB resistance with telithromycin is due to the replacement of the l-cladinose moiety at the C-3 position of the lactone ring by a ketone group . The basal production of methylase may affect weakly the activities of telithromycin and ABT773 because of their affinities for domain II. However, constitutive resistance or high-level basal production of methylase remains a stumbling block for the ketolides . Compared to erythromycin, telithromycin is a weak inducer or substrate for the MefA pump. This fact is reflected by the difference in the increase in MICs due to this mechanism, 50-fold versus 500- to 2,000-fold, respectively . RESISTANCE INDUCED BY RIBOSOMAL MUTATIONS : In vitro selection of E. coli mutants highly resistant to erythromycin has been of considerable value for characterization of the site of binding of this antibiotic to the ribosome. The clinical importance of this mechanism was recognized several years ago for microorganisms such as Helicobacterpylori and Mycobacteriumavium but only recently for pneumococci . Studies with pneumococcal mutants obtained in the laboratory have revealed that several structures participating in the binding of macrolides, domains V and II of 23S rRNA and proteins L22 and L4, can display mutations responsible for macrolide resistance . Most mutations affect 23S rRNA and are similar to those reported for other bacterial species . S. pneumoniae has four copies of the rrl gene for 23S rRNA, and transformation experiments with mutated rrl have shown that susceptibility to erythromycin decreases as the number of the mutated gene copies increases . Since high-level erythromycin resistance can be achieved only when at least two copies are mutated, this finding may explain why resistance conferred by RNA mutation is rare in pneumococci compared to H. pylori or M. avium, which contain only one or two copies of the rrl gene. The resistance phenotype conferred by alterations in the 23S rRNA target varies not only according to the number of mutated copies but also according to the nature of the substituted base . Point mutations at position A2058 or A2059 are associated with phenotypes similar to those previously reported for other organisms. A2058G and A2058U substitutions confer the highest level of MLSB resistance, with MICs of erythromycin and related macrolides of between 32 and >200 mug/ml . Telithromycin appears to be moderately affected (MICs of 0.06 to 1 mug/ml), probably because of the alternative interaction with domain II. Streptogramins retain activity, since synergy between the A and B factors is maintained. The A2059G mutation confers a high level of resistance to erythromycin, azithromycin, and 16-membered macrolides, a moderate level of resistance to clarithromycin and clindamycin, but no resistance to streptogramins, defining the ML resistance phenotype . Mutations at position 2611 destabilize the base pairing G2057-C2611 in the single-strand structure of the central loop . However, the C2611U substitution generally has a weak impact on the MICs of macrolides. Tait-Kamradt et al. found higher levels of resistance to streptogramin B antibiotics conferred by C2611A and C2611G substitutions . The C2610U change has been reported only for pneumococci and yields a slight increase in the MICs of macrolides and clindamycin . While telithromycin activity is only moderately altered by mutations in domain V, mutation of the adenine at position 752 in hairpin 35 (domain II) has a deleterious effect on the activity of the drug. A mutant combining a deletion of this base and a domain V mutation is resistant not only to 14- and 15-membered macrolides but also to telithromycin (MIC, 4 mg/liter), confirming the importance of domain II in the mechanism of action of this antibiotic . Various mutations in the rplV (L22) and rplD (L4) genes have been shown to play a role in resistance in laboratory mutants and in transformants of a susceptible S. pneumoniae strain obtained with mutated genes . The mutations in the L22 protein are located in a beta-hairpin extension at the C terminus of the protein . They confer resistance to streptogramins and low-level resistance to macrolides, whereas clindamycin does not seem to be affected . The MICs of telithromycin are increased but remain below 0.25 mug/ml. The mutations in the L4 protein occur in a region of 32 amino acids highly conserved in various species and interfere with the binding of the protein to rRNA . These mutations generally confer an MSB resistance phenotype. The MICs of macrolides against the mutant strains are moderately increased. Studies by three-dimensional cryoelectron microscopy of erythromycin-resistant ribosomes of E. coli have shown that L4 and L22 mutants have substantial changes in the polypeptide tunnel . The L4 mutant which does not bind erythromycin has a narrowing of the tunnel entrance which probably decreases the capacity of erythromycin to come into contact with its target. In contrast, the L22 mutant has an enlargement of the entrance and could bind erythromycin but in an ineffective way. Many of the mutations selected in vitro have been predictive of those found in clinical isolates . The A2059G mutation confers an ML resistance phenotype . A C2611G mutation was found in an isolate from Finland that was resistant to macrolides and highly resistant to streptogramin B antibiotics . Two types of rplD mutations in clinical isolates have been characterized . Sixteen isolates from Eastern Europe which were resistant to penicillin G and a Finnish isolate contained substitutions of three amino acids (69GTG71 -> 69TPS71) and displayed an MSB resistance phenotype with a high level resistance to macrolides . A Canadian isolate had a six-amino-acid insertion (underlined), 71GREKGTGR72, and displayed a similar phenotype but with a moderate level of resistance to all macrolides, including telithromycin (MIC, 3.12 mug/ml) . Recently, three strains isolated in Japan and for which the MICs of erythromycin were 64 or 128 mug/ml were reported to have an L22 mutation (D. J. Farrell, I. Morrissey, S. Bakker, D. Felmingham, J. Poehlsgaard, and S. Douthwaite, Abstr. 41st Intersci. Conf. Antimicrob. Agents Chemother., abstr. 1811, p. 100, 2001). The recent report of the emergence of an L22 mutant during treatment with azithromycin of fatal pneumococcal pneumonia emphasizes the clinical importance of mutations as a resistance mechanism . In summary, if, as expected, L4 and L22 mutants selected in vivo or in vitro have similar phenotypes, the MICs are surprisingly higher for the clinical isolates. The reasons for this difference are unknown but may be related, at least for L4 mutations, to differences in the types of mutations. Alternatively, other mechanisms of resistance to macrolide-lincosamide-streptogramin B antibiotics may also be present in wild strains. A clinical isolate with an A2062C mutation not obtained so far in vitro had a particular phenotype of a high level of resistance to spiramycin and streptogramin B and a moderate level of resistance to streptogramins A and B and to the combination . It remained susceptible to 14- and 15-membered macrolides, to telithromycin, and to clindamycin. This new phenotype confirms the notion that the binding sites of 14- and 16-membered macrolides are distinct. FIG. 2. | Secondary structures of hairpin 35 in domain II (top) and in domain V (bottom) of 23S rRNA in E. Secondary structures of hairpin 35 in domain II (top) and in domain V (bottom) of 23S rRNA in E. coli. Nucleotides which are protected by erythromycin are circled . Arrows indicate mutations conferring macrolide resistance on S. pneumoniae. The corresponding phenotype is indicated (K, ketolides; L, lincosamides; M, macrolides; M16, 16-membered macrolides; SB, streptogramin B; S, streptogramin A and streptogramin B). Small capital letters denote low-level resistance. TABLE 2 | MICs of macrolides and related antibiotics for ribosomal mutants of S. pneumoniae selected in vitro TABLE 3 | MICs of macrolides and related antibiotics for ribosomal mutants of clinical isolates of S. pneumoniae CONCLUSION : In recent years, both the incidence of macrolide resistance in pneumococci and the variety of resistance mechanism have increased sharply. The emergence of resistance mechanisms conferred by mutational alterations, in particular, is intriguing. This type of resistance may have remained undetected in the past because of a lack of adequate techniques or, alternatively, resistant mutants may have emerged and spread recently. It is conceivable that the use of new, long-acting macrolides with different pharmacokinetics may have contributed to modulation of the selective pressure exerted against pneumococci and to selection of new resistance genotypes. The variety of resulting phenotypes makes it particularly challenging to detect the nature of resistance in clinical isolates and may lead to difficulties in or make impossible the detection of resistance, depending on the individual drug(s) being tested. Backmatter: PMID- 12183235 TI - The Candida dubliniensis CdCDR1 Gene Is Not Essential for Fluconazole Resistance AB - The present study investigated the role of the Candida dubliniensis CdCDR1 and CdCDR2 genes in the development of fluconazole resistance. The C. dubliniensis CdCDR1 gene was 92% identical at the nucleotide sequence level to the corresponding C. albicans gene. However, 58% (14 of 24) of C. dubliniensis genotype 1 isolates tested harbored a nonsense mutation in the CdCDR1 open reading frame that converted codon 756 (TAT) to a TAG translational stop codon. Analysis of five of these C. dubliniensis isolates by Western immunoblotting showed that they expressed a truncated 85-kDa CdCdr1p compared to the full-length 170-kDa CdCdr1p. Expression of CdCDR1 alleles from six C. dubliniensis isolates in a pdr5 Saccharomyces cerevisiae strain revealed that CdCDR1 alleles from three isolates that encoded truncated proteins were unable to confer resistance to drugs and antifungals. However, reassignment of the TAG sequence at codon 756 to TAT (encoding tyrosine) in an allele from strain CD36 conferred the ability to mediate resistance to multiple drugs. Fluconazole-resistant isolates of C. dubliniensis harboring functional alleles of CdCDR1 were found to exhibit two- to ninefold-higher levels of CdCDR1 mRNA than did matched fluconazole-susceptible isolates. By comparison, levels of CdMDR1 expression ranged from approximately 50- to 100-fold greater in resistant isolates. Fluconazole resistance was also identified in isolates harboring nonfunctional CdCDR1 alleles, but resistance in these isolates was only associated with increased CdMDR1 expression. Targeted disruption of two functional alleles of CdCDR1 in a fluconazole-resistant derivative of C. dubliniensis that overexpressed both CdCDR1 and CdMDR1 revealed that although CdCDR1 was important for mediating reduced susceptibility to itraconazole and ketoconazole, there was no affect on fluconazole susceptibility in the double mutant. Evidence presented in this study reveals that CdCDR1 is not essential for the development of fluconazole resistance in C. dubliniensis. Keywords: Introduction : Resistance to azole antifungal drugs in Candida species is now recognized as a major clinical problem . Several studies have shown that a significant proportion of oral Candida albicans isolates recovered from human immunodeficiency virus (HIV)-infected patients are resistant to the azole antifungal drug fluconazole and that some non-C. albicans Candida species, such as Candida krusei, are inherently resistant to this agent . Fluconazole resistance has also been described in Candida dubliniensis, a species that was first described in 1995 . C. dubliniensis is a significant cause of oral disease in the HIV-infected patient population, who routinely receive fluconazole therapy for the suppression of oral candidiasis . Recently, several studies have reported the recovery of C. dubliniensis from the bloodstream, although the true incidence of systemic infection caused by this organism has yet to be determined . We and others have previously reported the recovery of C. dubliniensis isolates with reduced susceptibility to fluconazole from the oral cavities of HIV-infected patients . In addition, we have shown that fluconazole-susceptible C. dubliniensis clinical isolates can readily develop fluconazole resistance when exposed to this agent in vitro . Molecular analysis of fluconazole-resistant C. dubliniensis isolates and in vitro-generated fluconazole-resistant derivatives has shown that in each case, the fluconazole resistance phenotype is associated with increased expression of the CdMDR1 gene encoding the multidrug transporter CdMdr1p . Wirsching et al. demonstrated that targeted deletion of both copies of CdMDR1 in a C. dubliniensis clinical isolate with reduced susceptibility to fluconazole was sufficient to render the null mutant susceptible to fluconazole . A second multidrug transporter-encoding gene, termed CdCDR1, has also been identified in C. dubliniensis, and it is homologous to the C. albicans ABC transporter-encoding gene CDR1 . Almost all isolates of C. albicans (83%) with reduced susceptibility to azoles analyzed to date exhibit increased expression of CDR1 . However, the role of the homologous CdCDR1 gene in C. dubliniensis in determining susceptibility to fluconazole is less clear. Although increased expression of CdMDR1 has been observed in all fluconazole-resistant C. dubliniensis isolates and derivatives analyzed to date, increased CdCDR1 gene expression has only been reported in approximately 50% of fluconazole-resistant C. dubliniensis isolates and derivatives . For these reasons, the objectives of the present study were to functionally characterize the C. dubliniensis CdCDR1 gene and to examine its role in determining azole susceptibility in C. dubliniensis. MATERIALS AND METHODS : Candida strains and culture conditions. | Forty clinical isolates of C. dubliniensis from diverse geographic locations were included in this study . Isolates of C. dubliniensis were routinely cultured on potato dextrose agar (Oxoid) medium, pH 5.6, at 37C . For liquid culture, cells were grown in yeast extract-peptone-dextrose (YEPD) broth, also at 37C. Transformants of C. dubliniensis and Saccharomyces cerevisiae were selected and maintained on minimal agar medium (6.7 g of yeast nitrogen base without amino acids [Difco], 20 g of glucose, 15 g of Bacto agar [Difco], and 50 mg [each] of uracil, lysine, adenine, tryptophan, and histidine [Sigma-Aldrich] per liter). For induction of the SAP2 promoter and excision of the mycophenolic acid (MPA) resistance flipper cassette (MPAR-flipper), cells were grown in YCB-BSA medium (23.4 g of yeast carbon base [Difco], 4 g of bovine serum albumin [Sigma-Aldrich] per liter [pH 4.0]). Antifungal drug susceptibility testing. | Susceptibility testing of C. dubliniensis clinical isolates and their derivatives to antifungal drugs and metabolic inhibitors was performed by using a broth microdilution assay based on the approved NCCLS procedure . Susceptibility tests were carried out in RPMI 1640 medium (10.4 g of RPMI 1640 [Sigma-Aldrich], 20 g of glucose, 34.5 g of morpholinepropanesulfonic acid [pH 7.0] per liter) as described by Moran et al. . Metabolic inhibitors were purchased from Sigma-Aldrich, fluconazole was a gift from Pfizer Central Research (Sandwich, Kent, United Kingdom), and itraconazole and ketoconazole were gifts from Janssen Pharmaceutical (Cork, Republic of Ireland). The following antifungal drugs and metabolic inhibitors were prepared as stock solutions in water at the concentrations indicated: fluconazole, 5 mg/ml; fluphenazine, 20 mg/ml; and rhodamine 6G, 5 mg/ml. Stock solutions of other drugs were prepared in dimethyl sulfoxide at the concentrations indicated: ketoconazole, 1 mg/ml; itraconazole, 1 mg/ml; cycloheximide, 20 mg/ml; cerulenin, 1 mg/ml; and brefeldin A, 5 mg/ml. Crystal violet (2 mg/ml) was dissolved in methanol. Stock solutions of drugs were diluted in RPMI 1640 medium to the following concentrations, from which serial twofold dilutions were prepared in 96-well microtiter dishes (Corning): fluconazole, 64 mug/ml; itraconazole, 4 mug/ml; ketoconazole, 2 mug/ml; cerulenin, 64 mug/ml; rhodamine 6G, 16 mug/ml; cycloheximide, 512 mug/ml; brefeldin A, 512 mug/ml; and fluphenazine, 128 mug/ml. PCR amplification of C. dubliniensis CdCDR1 alleles. | Amplification of CdCDR1 and CdCDR2 genes was carried out by using the Expand high-fidelity PCR system (Roche Molecular Biochemicals). To amplify the entire C. dubliniensis CdCDR1 open reading frame (ORF), the oligonucleotide primer pair 1F-1R (Table ; Fig. ) was designed based on the nucleotide sequence of the C. albicans CDR1 gene. Similarly, for amplification of the CdCDR2 ORF, the primer pair 8F-8R was designed based on the nucleotide sequence of the C. albicans CDR2 gene. These primers contained cleavage sites for the restriction endonuclease XbaI . Template DNA was isolated from C. dubliniensis isolates as described by Sullivan et al. . Sequences flanking the CdCDR1 and CdCDR2 genes were amplified from the C. dubliniensis type strain CD36 by inverse PCR. Briefly, PCR amplification was carried out under the same conditions with self-ligated EcoRI-digested total genomic DNA as template. The 5' flanking region of the CdCDR1 gene was amplified with the oligonucleotide primer pair 3F-3R, which contained the EcoRI restriction endonuclease recognition sequence (Table ; Fig. ). The 3' region of the CdCDR1 gene was amplified with the oligonucleotide primer pair 4F-4R (Table ; Fig. ). Similarly, the 5' and 3' flanking sequences of the CdCDR2 gene were amplified with the oligonucleotide primer pairs 9F-9R and 10F-10R, respectively. PCR products were purified from the reaction mixture by using the Wizard PCR system (Promega). PCR primers were designed with restriction endonuclease recognition sequences at their 5' ends that allowed PCR products to be digested and ligated directly to the plasmid vector pBluescript II KS(-), transformed into Escherichia coli DH5alpha as described previously, and subjected to DNA sequence analysis . Heterologous expression of CdCDR genes in S. cerevisiae. | The oligonucleotide primer pairs 1F-1R and 8F-8R (Table ; Fig. ) were used to amplify the entire CdCDR1 and CdCDR2 ORFs, respectively. The CdCDR1 and CdCDR2 ORFs were amplified from genomic DNA purified from C. dubliniensis isolates CD36, CD57, CD51-II, CD47-IIb, CM1, and CM2 and C. albicans strain CA132A. These products were cloned by using standard techniques into the XbaI site of the expression plasmid pYES and transformed into the Deltapdr5 S. cerevisiae strain YKKB-13 as described by Sanglard et al. . Transformants of YKKB-13 were selected on minimal medium without uracil. In order to induce expression from the GAL1 promoter of pYES, transformants were subsequently maintained on minimal medium without uracil containing 2% (wt/vol) galactose as the sole carbon source. S. cerevisiae transformants were tested for susceptibility to antifungal drugs and metabolic inhibitors on minimal medium containing galactose as described by Sanglard et al. . Briefly, a suspension (2 x 107 CFU/ml) of each transformant to be tested was prepared in sterile saline. This solution was then serially diluted 10-fold, and 5 mul of each dilution was spotted onto plates containing fixed concentrations of each metabolic inhibitor (see Fig. ). Susceptibility to each drug was determined based on the highest dilution of each culture which could grow in the presence of the inhibitor as described by Sanglard et al. . Western immunoblotting. | Crude protein extracts were prepared from C. dubliniensis isolates, resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and subjected to Western immunoblot analysis with anti-Cdr1p polyclonal sera (a gift from D. Sanglard) as described by Moran et al. . Antibody-protein complexes were detected with the SuperSignal West Pico chemiluminescent substrate (Pierce, Rockford, Ill.). Molecular sizes of protein bands were estimated by using the Bio-Rad broad-range SDS-PAGE standards (Bio-Rad) and the GelWorks one-dimensional gel analysis software package (UVP, Irvine, Calif.). Site-directed mutagenesis. | Site-directed mutagenesis was carried out by using the GeneEditor system (Promega). Alleles of CdCDR1 for mutagenization were subcloned into the plasmid vector pGEM-11Z(f+). Point mutations were introduced by hybridization with the 5'-phosphorylated mutagenic oligonucleotide 5'-GCTTACCAATATTATAATTC, which contained a single G to T substitution (underlined), according to the manufacturer's instructions. Mutant alleles were sequenced on both strands in order to confirm the introduction of the desired mutation. Isolation of total RNA from C. dubliniensis and Northern blot analysis. | Total RNA was isolated from C. dubliniensis isolates grown to the mid-exponential phase (optical density at 600 nm, 0.6) in YEPD broth at 37C as described by Moran et al. . RNA was fractionated on 1.2% (wt/vol) agarose gels containing 6% (vol/vol) formaldehyde and transferred to nylon membranes (Osmonics, Westborough, Mass.) by capillary transfer. RNA was hybridized at 42C with DNA probes homologous to CdCDR1, CdMDR1, and CdTEF3 labeled with [alpha-32P]dATP (6,000 Ci/mmol, 220 TBq/mmol; NEN Life Sciences, Boston, Mass.) by random primer labeling as described by Moran et al. . A CdCDR1-specific probe was constructed by PCR amplification of the 5' region of the CdCDR1 gene (nucleotides +1 to +421) with the primer pair 1F-7R . Similarly, a 418-bp CdCDR2-specific probe (nucleotides +1 to +418) was amplified from C. dubliniensis genomic DNA by using the primer pair 8F-7R . A CdTEF3-specific probe was created by amplification of an internal fragment of the CdTEF3 gene with the primer pair TEF3F-TEF3R . Signals from Northern blot autoradiograms were quantified by scanning densitometry and normalized for loading against the CdTEF3 expression signal with the Gelworks one-dimensional software package (UVP). Targeted disruption of the CdCDR1 gene. | Targeted disruption of the CdCDR1 gene in C. dubliniensis CD57R was carried out by using the MPAR-flipper technique . Two different deletion constructs were made to inactivate both alleles of the CdCDR1 gene. To create the first construct, the entire CdCDR1 ORF was cloned in the XbaI site of pUC19. A deletion was created in this clone by inverse PCR with the primer pair 5F-5R (Table ; Fig. ). This allowed amplification of the 5' and 3' ends of the CdCDR1 ORF and the entire plasmid vector . These primers contained SacII and XhoI restriction endonuclease recognition sequences that enabled the insertion of the SacII/XhoI fragment of pSFI1 (containing the site-specific recombinase FLP and the MPA resistance gene) , thereby replacing the central region of the CdCDR1 ORF to create plasmid pCDRDelta1. To create the second construct, the internal AccI/BamHI fragment of CdCDR1 was cloned into pUC19 and subjected to inverse PCR with the primer pair 6F-6R (Table ; Fig. ). This product was also ligated with the SacII/XhoI fragment from pSFI1 to create the plasmid pCDRDelta2. C. dubliniensis was sequentially transformed by using approximately 1 mug of the linear SacI/SphI fragments from plasmids pCDRDelta1 and pCDRDelta2 by electroporation as described by Staib et al. . MPA-resistant transformants were selected on minimal medium agar plates containing 10, 15, or 20 mug of MPA per ml. Isolation of genomic DNA and Southern blotting. | Genomic DNA from C. dubliniensis isolates and derivatives for use in PCR and Southern blotting experiments was isolated by using the method of Sullivan et al. . For Southern blotting, 10 mug of genomic DNA was digested with the appropriate restriction endonuclease and separated on 0.8% (wt/vol) agarose gels. DNA was transferred to nylon membranes (Osmonics) by capillary transfer overnight. Hybridization was carried out under high-stringency conditions by standard techniques with DNA probes labeled with alpha-32P by random primer labeling . FIG. 1. | Restriction map of CdCDR1 (A) and CdCDR2 (B) encoding DNA from C. dubliniensis Restriction map of CdCDR1 (A) and CdCDR2 (B) encoding DNA from C. dubliniensis CD36. The black rectangular boxes represent C. dubliniensis genomic DNA which was amplified by PCR from genomic DNA (see Materials and Methods). The large arrows show the positions and directions of transcription of the CdCDR1 and CdCDR2 ORFs. The asterisk in the CdCDR1 ORF shows the position of the polymorphic codon 756. The approximate positions of the CdCDR1- and CdCDR2-specific PCR primers shown in Table are indicated by small arrows. Primer pairs 5F-5R and 6F-6R shown at the bottom of panel A were used in the inverse PCR experiments performed to disrupt the CdCDR1 gene. Restriction endonuclease cleavage sites: A, AccI; B, BamHI; E, EcoRI; P, PvuII. FIG. 3. | Susceptibility of S. cerevisiae Susceptibility of S. cerevisiae YKKB-13 (pdr5) transformants harboring cloned CDR1 genes to antifungal drugs and metabolic inhibitors. (A) CDR1 alleles from C. albicans and C. dubliniensis isolates were amplified by PCR, cloned into the GAL1 expression vector pYES and transformed into the S. cerevisiae strain YKKB-13. The transformants harbored the pYES plasmid (Vector), cloned CDR1 from C. albicans CA132A (Y132A), and cloned CdCDR1 genes from the C. dubliniensis isolates CD36 (YCD36), CD57 (YCD57), and CD51-II (YCD51-II). A suspension (2 x 107 CFU/ml) of each transformant was serially diluted, and 5 mul of each dilution was spotted onto minimal agar medium plates containing fixed concentrations of the antifungal drug or metabolic inhibitor indicated. (B) Fluconazole susceptibility of YKKB-13 transformants harboring the cloned C. albicans CA132A CDR1 gene (Y132A), the CD36 CdCDR1 gene (YCD36), and the mutated CD36 CdCDR1 gene (YMCD36) in which the premature stop codon has been reassigned. TABLE 1 | Candida strains and derivatives used in this study TABLE 2 | Oligonucleotide primers used in PCRs RESULTS : Isolation and sequence analysis of CdCDR1. | In order to isolate the C. dubliniensis homologue of the C. albicans CDR1 gene, a 6,848-bp region of C. dubliniensis chromosomal DNA was PCR amplified by using a high-fidelity polymerase mixture from genomic DNA isolated from the C. dubliniensis type strain CD36 (CBS 7987). Initially, the primer pair 1F-1R (Table ; Fig. ), which was designed to amplify the entire C. albicans CDR1 ORF, was used to amplify a 4,506-bp region of C. dubliniensis genomic DNA corresponding to the putative C. dubliniensis CDR1 homologue. The nucleotide sequences 5' and 3' to this region were then amplified from C. dubliniensis CD36 genomic DNA by an inverse PCR method with the primer pairs 3F-3R and 4F-4R (Table ; Fig. ), respectively. Using this strategy, a 614-bp region upstream of the putative CdCDR1 ORF and a 1,728-bp downstream region were amplified from C. dubliniensis CD36 chromosomal DNA. The entire amplified chromosomal DNA region from strain CD36 was then sequenced on both strands. The putative CdCDR1 ORF amplified with the primer pair CDR1F-CDR1R was 4,506 bp in length and was 92% identical at the nucleotide sequence level to the C. albicans CDR1 ORF. The 614-bp 5'-flanking region of this sequence shared 46.6% identity at the nucleotide sequence level with the C. albicans CDR1 promoter and contains a putative TATA box at nucleotide position -126 (the first base of the ATG start codon was designated +1). In addition, the C. dubliniensis promoter sequence contains a motif at the nucleotide coordinates -371 to -350 (5'-CGGTTATCGGATATTTTTTTT) matching the drug response element (DRE) in C. albicans . The 3'-flanking region of the putative CdCDR1 ORF also contained sequences homologous to the C. albicans SAP3 gene encoding a member of the secreted aspartyl-proteinase family , which we have termed CdSAP3. This is the same gene order as observed in the C. albicans genome . However, comparison of the C. albicans and C. dubliniensis CDR1 ORFs revealed that the C. dubliniensis CD36 CdCDR1 gene contained a TAG translation stop codon at the nucleotide coordinates +2266 to +2268 (codon 756) which corresponded to a single base difference (G to T) that converted the tyrosine (Y756)-encoding TAT codon found in the C. albicans CDR1 gene to a TAG translation stop signal . The predicted polypeptide encoded by this shorter ORF was 755 amino acids in length and had a predicted molecular mass of 85 kDa; the C. albicans protein is 1,501 amino acids in length and has a predicted molecular mass of 168 kDa. The truncated CdCdr1p was 96.7% identical to the corresponding amino acid sequences of CaCdr1p. In order to confirm that this substitution was not a PCR artifact, this region was sequenced in six clones generated in six separate PCR amplifications and an identical nucleotide sequence was found in each clone. These data strongly suggested that the substitution was present in both alleles of CdCDR1. Identification of polymorphic alleles of CdCDR1. | In order to determine whether other strains of C. dubliniensis harbored the TAG nonsense codon, the region of the CdCDR1 ORF containing codon 756 was sequenced in C. dubliniensis isolates CM1, CM2, CD57, CD51-II, and CD47-IIb . In order to achieve this, a 490-bp region of the CdCDR1 ORF was amplified by PCR with the primer pair 2F-2R (Table ; Fig. ). This amplified product encompassed the nucleotide region +2071 to +2557 of the CdCDR1 ORF, including the TAG codon at the nucleotide coordinates +2266 to +2268 in C. dubliniensis strain CD36. Nucleotide sequence analysis of this region demonstrated that C. dubliniensis isolates CD51-II and CD47-IIb harbor a nonsense mutation identical to that found in the CdCDR1 ORF of isolate CD36. However, isolates CD57, CM1, and CM2 contained a tyrosine codon identical to that found in the C. albicans CDR1 ORF. The entire nucleotide sequence of the CdCDR1 gene from C. dubliniensis CD57 was determined in order to compare the sequence of a gene without the nonsense mutation with that from strain CD36. The nucleotide sequence of the C. dubliniensis CD57 CdCDR1 ORF was 99.7% identical (containing 9 base differences) to the CdCDR1 ORF from strain CD36. The predicted amino acid sequence of the CD57 CdCdr1p contained three amino acid substitutions (H60R, V173M, and S1264L) compared to the CD36 sequence. However, the predicted CdCDR1 ORF from strain CD57 was not interrupted by a premature stop codon at nucleotide positions +2266 to +2268 and was identical in size (4,506 bp) to the homologous C. albicans CDR1 gene . The nucleotide sequence of the CdCDR1 ORF from strain CD57 was 92% identical to the corresponding sequence of the C. albicans CDR1 gene. The CD57 CdCDR1 gene encodes a predicted polypeptide of 169.6 kDa that is 96.5% identical at the amino acid sequence level to the corresponding C. albicans sequence. This highly homologous sequence contained the typical features described in C. albicans Cdr1p, namely the conserved Walker A and Walker B motifs and an ATP binding motif in the N-terminal hydrophilic domain . The tyrosine-encoding TAT at codon position 756 in isolates CM1, CM2, and CD57 was found to be located within the recognition sequence for the restriction endonuclease SspI (5'-AATATT). This SspI recognition sequence was absent in alleles from CD36, CD51-II, and CD47-IIb containing the TAG stop codon (5'-AATAGT). PCR products amplified from isolates CM1, CM2, and CD57 with the primer pair 2F-2R could be digested with the restriction endonuclease SspI to yield two distinct fragments of 200 and 290 bp, respectively, in agarose gels. However, SspI digestion of the amplified product from isolates CD36, CD51-II, and CD47-IIb, which contained the TAG nonsense codon, yielded a single band of approximately 500 bp in agarose gels, indicating that the SspI recognition sequence was absent. We utilized this restriction fragment length polymorphism as the basis of a screening assay to analyze a larger group of C. dubliniensis isolates for the presence or absence of the nonsense mutation. A group of 40 C. dubliniensis isolates (all isolates in Table , excluding derivatives of CD36, CD57, and CD51-II), including representative isolates from the four C. dubliniensis genotypes recently described by Gee et al. , were analyzed for the presence of the TAT codon by SspI digestion of PCR products generated from these isolates with the primer pair 2F-2R . This analysis revealed that all of the isolates tested (n = 16) belonging to C. dubliniensis genotypes 2, 3, and 4 harbored CdCDR1 alleles that contained the TAT tyrosine codon, as SspI digestion of the 2F-2R-amplified region yielded two fragments in agarose gels . However, analysis of 24 genotype 1 C. dubliniensis isolates revealed that 14 (58%) harbored CdCDR1 alleles that could not be digested by SspI at this region. DNA sequence analysis of the 2F-2R-amplified region from these 14 isolates revealed the presence of a TAG stop codon that disrupted the SspI recognition sequence . These data confirm that the premature stop codon is present in both alleles of CdCDR1 in these isolates. Two of these isolates were found to harbor CdCDR1 alleles containing both the TAT and TAG codon sequences, indicating heterozygosity at this locus. In total, 35% (14 of 40) of the C. dubliniensis isolates tested were found to contain CdCDR1 alleles containing the TAG stop codon. Isolation and sequence analysis of CdCDR2. | In order to isolate the C. dubliniensis homologue of CDR2, the primer pair 8F-8R was designed based on the nucleotide sequence of the C. albicans CDR2 gene . This primer pair was used to amplify the complete C. dubliniensis CdCDR2 gene from the genomic DNA of strain CD36. The amplified CdCDR2 ORF was 4,503 bp in length, which is 3 bp longer than the C. albicans CDR2 ORF due to the presence of an additional codon. The C. dubliniensis CdCDR2 gene shared 91% identity with the C. albicans CDR2 ORF at the nucleotide sequence level. This ORF was preceded at the 5' end by a TATA box at position -99 (the first base of the ATG start codon was designated +1). Like CdCDR1, the C. dubliniensis CdCDR2 gene was preceded by a DRE motif at nucleotide coordinates -210 to -189. The C. dubliniensis CdCDR2 ORF encodes a protein with a predicted molecular mass of 168.9 kDa that is 94.4% identical at the amino acid sequence level to the corresponding C. albicans protein. This sequence contained identical Walker A and Walker B motifs and an identical ATP binding motif to that described by Sanglard et al. in the C. albicans Cdr2p . Western immunoblot analysis of CdCdr1p. | Crude protein extracts of C. dubliniensis were analyzed by Western immunoblotting with polyclonal sera raised against the N-terminal region of the C. albicans Cdr1p. In protein extracts from C. albicans CA132A and C. dubliniensis isolates CD41, CD57, CD72, CM1, and CM2 , a protein of approximately 170 kDa was detected, matching the predicted size of C. albicans Cdr1p and the protein encoded by the full-length CdCDR1 ORF . However, in protein extracts from C. dubliniensis isolates CD36, CD51-II, CD47-IIb, CO4, and P30, whose CdCDR1 ORFs were found by sequence analysis to contain a TAG translational stop signal at codon 756, anti-Cdr1p sera reacted with a smaller polypeptide of 85 kDa . This smaller protein is identical in size to the truncated protein predicted from the CdCDR1 nucleotide sequence obtained from C. dubliniensis strain CD36. Functional analysis of the CdCDR1 and CdCDR2 genes. | The CdCDR1 and CdCDR2 ORFs were amplified from genomic DNA recovered from C. dubliniensis isolates CD36, CD57, CD51-II, CD47-IIb, CM1, and CM2 and C. albicans strain CA132A . Amplicons were cloned in the S. cerevisiae GAL1 expression vector pYES and were transformed into the azole-susceptible Deltapdr5 S. cerevisiae strain YKKB-13 . Transformants were cultured on minimal agar medium containing 2% (wt/vol) galactose as the sole carbon source in order to induce expression of the cloned CDR genes. Northern blot analysis of total RNA from galactose-grown transformants revealed high levels of expression of the mRNAs for CdCDR1 and CdCDR2 in the respective transformants (data not shown). Transformants were tested for susceptibility to azole antifungal drugs and metabolic inhibitors known to be substrates for C. albicans Cdr1p and Cdr2p . Cloned CDR1 amplicons from C. albicans CA132A and from the C. dubliniensis isolates CM1, CM2, and CD57 were found to confer resistance to fluconazole and itraconazole and to the metabolic inhibitors rhodamine 6G, cycloheximide, brefeldin A, crystal violet, and cerulenin in S. cerevisiae YKKB-13 (Table ; Fig. ). However, CdCDR1 amplicons from C. dubliniensis isolates CD36, CD51-II, and CD47-IIb, which encoded the truncated CdCdr1p, did not confer resistance to any of these compounds in S. cerevisiae YKKB-13 , despite the fact that high levels of CdCDR1 mRNA could be detected in these transformants by Northern analysis (data not shown). For each of these isolates, at least six separate CdCDR1-encoding amplicons cloned from separate PCRs were analyzed and identical phenotypes were obtained from each clone. In order to unequivocally determine if the premature translational stop signal at codon 756 of the CdCDR1 gene was responsible for the loss of function of the heterologously expressed protein in S. cerevisiae, we carried out site-directed mutagenesis to restore the tyrosine-encoding TAT codon at this position in the cloned C. dubliniensis CD36 CdCDR1 gene. Reassignment of this codon fully restored the ability of the expressed protein to impart a drug resistance phenotype in S. cerevisiae YKKB-13 compared to that of the truncated C. dubliniensis CdCdr1p . The susceptibility of the CdCDR2-harboring transformants to fluconazole and itraconazole was also determined. When compared to S. cerevisiae harboring the plasmid vector alone (fluconazole MIC, 0.5 mug/ml; itraconazole MIC, <0.03 mug/ml), transformants harboring the C. albicans CDR2 gene or the C. dubliniensis CdCDR2 gene were resistant to fluconazole (MIC, 32 to 64 mug/ml) and itraconazole (MIC, 0.5 to 1 mug/ml). Analysis of CdCDR1 and CdCDR2 mRNA expression levels in C. dubliniensis. | Total RNA was isolated from C. dubliniensis isolates and in vitro-generated derivatives with reduced susceptibility to fluconazole in order to ascertain the expression levels of CdCDR1 and CdCDR2 mRNA in these organisms . As reported by Moran et al. , all clinical isolates and in vitro-generated derivatives of C. dubliniensis with reduced susceptibility to fluconazole displayed high levels of CdMDR1 mRNA compared to matched fluconazole-susceptible isolates . CdMDR1 hybridization signals in the resistant organisms CM2, CD57R, and CD51-II were 50- to 100-fold higher than those observed in matched fluconazole-susceptible isolates. Hybridization of these RNAs with a probe specific for the CdCDR1 gene (homologous to the region from +1 to +421 of the CdCDR1 ORF) revealed that several of these isolates and derivatives with reduced susceptibility to fluconazole also displayed comparatively minor increases in levels of CdCDR1 mRNA expression relative to matched fluconazole-susceptible isolates . The oral clinical isolate CM2 (fluconazole MIC, 32 mug/ml) displayed a twofold-higher level of CdCDR1 mRNA compared to the matched fluconazole-susceptible isolate CM1 (fluconazole MIC, 0.5 mug/ml) recovered from the same patient. Isolates CM1 and CM2 both express the full-length CdCdr1p . Also, the in vitro-generated derivative CD57R (fluconazole MIC, 32 mug/ml), which also expresses the full-length CdCdr1p, displayed a ninefold-higher level of CdCDR1 mRNA than the fluconazole-susceptible parental isolate CD57. Interestingly, the oral clinical isolate CD47-IIb (fluconazole MIC, 16 mug/ml), which was found to express the nonfunctional truncated CdCdr1p, expressed a twofold-higher level of CdCDR1 mRNA than the matched clinical isolate CD47-I (fluconazole MIC, 8 mug/ml). However, in vitro-generated derivatives of the fluconazole-susceptible C. dubliniensis isolates CD51-II and CD36 (fluconazole MIC, 0.25 mug/ml), which displayed reduced susceptibility to fluconazole (fluconazole MICs were 32 to 64 mug/ml for CD51-IIR, CD36R1, and CD36R2), displayed levels of CdCDR1 mRNA that were similar to or reduced in comparison to their respective fluconazole-susceptible parental isolates . Both CD36 and CD51-II were shown to express the truncated 85-kDa CdCdr1p . Total RNA from these isolates and derivatives was also hybridized with sequences specific for CdCDR2 (homologous to the region from +1 to +418 of the CdCDR2 ORF). However, expression of CdCDR2 mRNA was not detected in any of the isolates or derivatives tested. Targeted disruption of CdCDR1. | In order to ascertain the role of the CdCDR1 gene in determining susceptibility to azole antifungal drugs in C. dubliniensis, we disrupted both copies of the CdCDR1 gene in a C. dubliniensis strain by using the MPAR-flipper technique described by Wirsching et al. . These experiments were carried out in an in vitro-generated derivative of the clinical isolate CD57 (fluconazole MIC, 0.5 mug/ml), termed CD57R, which exhibits reduced susceptibility to fluconazole (fluconazole MIC, 32 mug/ml). CD57R was originally generated by successive subculture of the fluconazole-susceptible isolate CD57 on fluconazole-containing agar medium . CD57R expresses the full-length CdCdr1p and showed a ninefold increase in CdCDR1 mRNA expression levels relative to a matched fluconazole-susceptible isolate (CD57) and also exhibits increased CdCdr1p levels in Western immunoblots (data not shown). In addition, like all C. dubliniensis isolates with reduced susceptibility to fluconazole, CD57R exhibits an approximately 88-fold increase in expression of CdMDR1 mRNA , and this isolate was chosen in order to determine the influence of a CdCDR1 gene disruption on fluconazole susceptibility in this genetic background. For disruption of the CdCDR1 gene, a cassette (MPAR-flipper) containing the MPA resistance gene and the FLP site-specific recombinase fused to the inducible C. albicans SAP2 promoter , was inserted into the CdCDR1 ORF to create plasmids pCDRDelta1 and pCDRDelta2, thereby deleting the regions from nucleotide coordinates +538 to +4248 and +1468 to +3532, respectively . C. dubliniensis CD57R was first transformed with the insert from pCDRDelta1, and MPA-resistant transformants were analyzed by Southern blotting. In strain CD57R (CdCDR1/CdCDR1), a BglII DNA fragment of 9.2 kb hybridized with the CdCDR1 probe . Insertion of the fragment from pCDRDelta1 into one of the CdCDR1 alleles generated a new BglII fragment of 3.5 kb in strain 57RM1 (CdCDR1/cdcdr1Delta::MPAR-FLIP) due to the presence of a BglII site in the MPAR-flipper. Deletion of the cassette by FLP-mediated recombination resulted in the creation of a new 5.5-kb fragment, 3.7 kb smaller than the fragment in the wild-type strain CD57R; this strain was designated 57RM2 (CdCDR1/cdcdr1Delta::FRT). The insert from pCDRDelta2 was then used in a second round of transformation in order to create a deletion in the second CdCDR1 allele in strain 57RM3 (cdcdr1Delta::FRT/cdcdr1Delta::MPAR-FLIP). The MPAR-flipper was excised again from this transformant to produce a fragment of 7.2 kb in strain 57RM4 (cdcdr1Delta::FRT/cdcdr1Delta::FRT). Disruption of CdCDR1 affects itraconazole and ketoconazole susceptibility. | Heterologous expression of CdCDR1 in S. cerevisiae results in resistance to the azole antifungal drugs fluconazole, itraconazole, and ketoconazole. In order to assess the contribution of the CdCDR1 gene towards resistance to these agents in C. dubliniensis, we compared the azole susceptibility of the cdr1-null mutant strain 57RM4 with that of its azole-resistant parental strain CD57R. Broth dilution susceptibility tests were carried out in RPMI 1640 medium against azole drugs for the azole-susceptible clinical isolate CD57, its in vitro-generated fluconazole-resistant derivative CD57R, and the cdr1 disruptants of CD57R, 57RM2 (CdCDR1/cdcdr1Delta::FRT) and 57RM4 (cdcdr1Delta::FRT/cdcdr1Delta::FRT). Disruption of both copies of CdCDR1 in 57RM4 resulted in an eightfold increase in susceptibility to both itraconazole (change in MIC from 0.5 to 0.06 mug/ml) and ketoconazole (change in MIC from 0.25 to 0.03 mug/ml) . The ketoconazole and itraconazole MICs for the double cdr1 disruptant (57RM4) and CD57, the azole susceptible parental isolate of CD57R, were similar, indicating that expression of the CdCDR1 gene was responsible for the reduction in susceptibility to these drugs in CD57R. However, this disruption did not affect susceptibility to fluconazole . As fluconazole, unlike ketoconazole and itraconazole, is also a substrate for the CdMDR1 multidrug transporter, we hypothesized that the very high levels of expression of the CdMDR1 transporter in CD57R may mask any effect a deletion of CdCDR1 may have on fluconazole susceptibility. In order to determine if CdMDR1 could compensate for the cdr1 disruption in the presence of other drugs, we analyzed the susceptibility of the cdr1 null mutant to a range of other metabolic inhibitors. We did not observe any change in the MICs of other drugs that are substrates for both CdCdr1p and CdMdr1p in the cdr1 disruptant strain 57RM4, including brefeldin A, cerulenin, and cycloheximide . However, we observed increased susceptibility to crystal violet and rhodamine 6G in strain 57RM4, two drugs that are substrates for CdCdr1p but are not transported by CdMdr1p . In addition, we did not observe any differences in fluphenazine susceptibility among the isolates and derivatives tested , indicating that this drug is not a substrate for CdCdr1p. FIG. 2. | Western immunoblot analysis of crude protein extracts from C. albicans Western immunoblot analysis of crude protein extracts from C. albicans CA132A and several C. dubliniensis isolates following SDS-PAGE. Western blots were screened with an anti-Cdr1p polyclonal serum . The uppermost arrow at the right of the figure indicates the position of full-length (170 kDa) Cdr1p, which is expressed by C. albicans CA132A and by C. dubliniensis CD57, CM1, and CM2, respectively. The lower arrow indicates the position of truncated (85 kDa) CdCdr1p, which is expressed by CD36, CD51-II, and CD47-IIb, respectively. FIG. 4. | Northern blot showing expression levels of CdCDR1, CdMDR1, and CdTEF3 mRNAs in matched pairs of C. dubliniensis Northern blot showing expression levels of CdCDR1, CdMDR1, and CdTEF3 mRNAs in matched pairs of C. dubliniensis clinical isolates and in vitro-generated derivatives exhibiting reduced susceptibility to fluconazole. (A) Total RNA was extracted from C. dubliniensis isolates and derivatives grown to the mid-exponential phase in YEPD broth cultures and analyzed by Northern hybridization analysis with [alpha-32P]dATP-labeled DNA probes homologous to CdCDR1, CdMDR1, and the constitutively expressed internal control CdTEF3 gene (see Materials and Methods). (B) Graphical representation of CdCDR1 and CdMDR1 mRNA expression levels. Hybridization signals were analyzed by scanning densitometry and normalized against levels of CdTEF3 expression. FIG. 5. | Inactivation of the CdCDR1 gene by the MPAR-flipper procedure. Inactivation of the CdCDR1 gene by the MPAR-flipper procedure. (A) Restriction map of the CdCDR1 locus from strain CD57R and allelic replacements with the inserts from pDeltaCDR1 (top) and pDeltaCDR2 (bottom). The shaded arrow in the central portion of the figure represents the CdCDR1 coding region; solid lines represent the CdCDR1 flanking sequences. The open box represents the MPAR-flipper cassette. Restriction endonuclease cleavage sites: Bg, BglII; Sc, SacI; Sp, SphI. The solid bar represents the region which was used as a probe in Southern hybridization experiments. (B) Southern hybridization of BglII-digested genomic DNA from C. dubliniensis CD57R and mutant derivatives with the 5' CdCDR1 probe indicated in panel A. The molecular sizes of the fragments are shown in kilobases to the left of the blot, and the identities of the fragments are shown to the right. Lane 1, CD57R (CdCDR1/CdCDR1); lane 2, 57RM1 (CdCDR1/cdcdr1Delta::MPAR-FLIP); lane 3, 57RM2 (CdCDR1/cdcdr1Delta::FRT); lane 4, 57RM3 (cdcdr1Delta::FRT/cdcdr1Delta::MPAR-FLIP); lane 5, 57RM4 (cdcdr1Delta::FRT/cdcdr1Delta::FRT). FIG. 6. | Susceptibility of C. dubliniensis Susceptibility of C. dubliniensis strain CD57, its fluconazole-resistant derivative CD57R, and CdCDR1 disruptants of CD57R (57RM2 [CdCDR1/cdcdr1Delta::FRT] and 57RM4 [cdcdr1Delta::FRT/cdcdr1Delta::FRT]) to fluconazole, itraconazole, ketoconazole, and metabolic inhibitors. MICs were determined by broth microdilution susceptibility testing in RPMI 1640 supplemented with 2% (wt/vol) glucose. TABLE 3 | Susceptibility of S. cerevisiae YKKB-13 CdCDR1 transformants to antifungal drugs and metabolic inhibitors determined by broth microdilution DISCUSSION : The close relatedness of C. dubliniensis to C. albicans and the ease with which fluconazole-resistant derivatives of susceptible C. dubliniensis clinical isolates can be generated in vitro make C. dubliniensis a useful organism to study the development of fluconazole resistance in Candida species . We and others have isolated C. dubliniensis isolates exhibiting fluconazole resistance from the oral cavities of HIV-infected patients, making the investigation of the molecular basis of resistance in this organism worthy of investigation in its own right . Recently, molecular genetic techniques, including targeted gene disruption, have been adapted for use in C. dubliniensis and will greatly aid in the dissection of this organism's biology . Wirsching et al. recently used targeted gene disruption to demonstrate the importance of the major facilitator CdMdr1p in the development of fluconazole resistance in C. dubliniensis . CdMDR1 mRNA is invariably overexpressed in C. dubliniensis strains exhibiting reduced susceptibility to fluconazole, and deletion of this gene in a fluconazole-resistant C. dubliniensis isolate rendered the isolate susceptible to fluconazole. In the present study we investigated the role of the second multidrug transporter, CdCdr1p, in fluconazole resistance in C. dubliniensis. In C. albicans, overexpression of Cdr1p is the most commonly reported mechanism of resistance to azole antifungal drugs . Because of this apparent divergence in fluconazole resistance mechanisms in C. albicans and C. dubliniensis, we decided to investigate in detail the role played by CdCDR1 in fluconazole resistance in C. dubliniensis. Sequence analysis of the C. dubliniensis CdCDR1 and CdCDR2 genes reveals that these genes are highly homologous (>90% identity at the nucleotide sequence level) to their C. albicans counterparts . The promoter regions of the C. dubliniensis genes were also found to contain motifs similar to that of the DRE in C. albicans described by de Micheli et al., indicating that expression of these genes may be regulated by similar mechanisms in C. dubliniensis . However, 35% (14 of 40) of C. dubliniensis isolates examined in the present study were found to contain a nonsense mutation approximately midway through the CdCDR1 ORF. This mutation introduces a premature translation termination signal in the ORF which leads to the translation of a truncated 85-kDa protein (rather than the full-length protein of 169.6 kDa), which we detected in Western immunoblots with an anti-Cdr1p antiserum. We demonstrated that the full-length CdCdr1p and CdCdr2p proteins are functionally equivalent to their C. albicans counterparts, as when they are heterologously expressed in a S. cerevisiae background they render the host strain less susceptible to a similar range of drugs and metabolic inhibitors. However, expression of CdCDR1 ORFs containing the nonsense mutation in S. cerevisiae did not render the host strain less susceptible to any of the drugs tested, indicating that the truncated CdCdr1p is not capable of mediating a multidrug resistance phenotype. Since the C. dubliniensis isolates harboring the nonsense codon containing CdCDR1 ORFs were obtained from clinical sources, this finding suggests that a fully functional, full-length CdCdr1p is not essential for the normal growth of C. dubliniensis in vivo. Furthermore, all of the C. dubliniensis isolates containing this nonsense mutation were found to belong to a group of closely related strains recently identified as C. dubliniensis genotype 1, as described in the epidemiological study of Gee et al. . In the study by Gee et al. , four distinct genotypes of C. dubliniensis were identified based on DNA fingerprint analysis with the C. dubliniensis-specific probe Cd25 and by sequence analysis of the internal transcribed spacer regions of the rRNA genes. Gee et al. found that genotype 1 isolates predominated in an analysis of 98 C. dubliniensis isolates. Genotype 1 isolates are mainly recovered from HIV-infected patients, and DNA fingerprint analysis with the C. dubliniensis-specific Cd25 probe indicates that they are a closely related group of organisms (mean similarity coefficient value of 0.80 +- 0.06) despite being recovered from disparate geographic areas, indicating that these represent a more recent, and therefore a more homogenous subgroup, of C. dubliniensis that has become predominant worldwide . The most likely explanation for why the CdCDR1 TAG mutation is unique to the C. dubliniensis genotype 1 population, is that it probably appeared in this subgroup after these organisms separated from the other C. dubliniensis genotypes and that it has subsequently spread throughout the human population worldwide. In addition, several of the clinical isolates in which the nonsense mutation was identified in the present study were originally recovered from patients with oral candidiasis who had received fluconazole therapy, indicating that the presence of a truncated CdCdr1p does not appear to adversely affect the ability of these organisms to cause disease. In order to investigate whether the nonsense mutation might have any effect on virulence or on strain selection in patients treated with fluconazole, we propose to conduct growth competition experiments comparing the fitness of strains with and without the CdCDR1 mutation by using in vivo models with and without fluconazole. Of greater interest is the fact that two of these isolates (CD36 and CD51-II) were capable of yielding fluconazole-resistant derivatives (MIC, 64 mug/ml) upon exposure to fluconazole in vitro, indicating that the full-length CdCdr1p is not essential for the development of fluconazole resistance. Whether the truncated CdCdr1p protein could carry out some alternative cellular function other than drug efflux is as yet unknown. We have shown that some C. dubliniensis isolates and derivatives exhibiting reduced susceptibility to fluconazole that possess the full-length CdCDR1 ORF exhibit increased transcription of this gene (e.g., CM2 and CD57R) . We have also observed that fluconazole-resistant derivatives of C. dubliniensis isolates possessing the CdCDR1 ORF containing the nonsense mutation often do not exhibit increases in CdCDR1 mRNA expression (e.g., CD36R1 and CD36R2) , and in some cases even exhibit decreased transcription of this gene (e.g., CD51-IIR) . CdCDR2 expression has not been detected in vitro in any of the C. dubliniensis isolates tested to date, indicating that this gene is stringently repressed. In order to further investigate the role of functional alleles of CdCDR1 in the development of fluconazole resistance in C. dubliniensis, we carried out targeted disruption of both alleles of this gene in a derivative of the C. dubliniensis clinical isolate CD57, termed CD57R, which exhibited reduced susceptibility to fluconazole. As CD57R displays a ninefold-higher level of CdCDR1 mRNA expression (in addition to overexpression of CdMDR1), we anticipated that if CdCDR1 is significantly involved in the fluconazole resistance phenotype that its disruption would affect susceptibility to all of the azole antifungal drugs, including fluconazole. We found the double cdr1 disruptant 57RM4 showed increased susceptibility to itraconazole and ketoconazole, and the MICs of these drugs for 57RM4 were similar to those observed for the clinical isolate CD57, the azole-susceptible parent of CD57R, indicating that CdCDR1 overexpression was the main mediator of reduced susceptibility to these drugs in CD57R . This phenotype was expected, as both ketoconazole and itraconazole are transported by the heterologously expressed gene in S. cerevisiae . In addition, the double cdr1 disruptant 57RM4 displayed increased susceptibility to rhodamine 6G and crystal violet , two metabolic inhibitors which are also transported by CdCdr1p in S. cerevisiae . However, the double cdr1 disruption did not affect fluconazole susceptibility in 57RM4 when compared to its parental strain CD57R . However, as the cdr1 mutant also displayed increased levels of CdMDR1 mRNA, which encodes a transporter capable of fluconazole efflux, it is perhaps not surprising that fluconazole susceptibility was not affected in this mutant . It is likely that such high levels of CdMDR1 expression are capable of mediating the fluconazole resistance phenotype alone and may mask the effect of the cdr1 disruption on fluconazole susceptibility. The cdr1 mutant was also unaffected in susceptibility to cycloheximide, cerulenin, and brefeldin A , three compounds which, like fluconazole, are transported by both CdCdr1p and CdMdr1p . The findings presented here, which implicate CdMDR1 gene expression as the most important resistance mechanism to fluconazole and other metabolic inhibitors in C. dubliniensis is supported by the findings of Wirsching et al. . In that study, Wirsching et al. demonstrated that disruption of the CdMDR1 gene in the C. dubliniensis clinical isolate CM2, which exhibits increased expression of CdMDR1 mRNA and also exhibits a twofold increase in CdCDR1 mRNA expression , was sufficient to render the null mutant susceptible to fluconazole . The ketoconazole MICs for the cdmdr1-disrupted strain were still elevated, probably due to the elevated CdCDR1 expression levels seen in this isolate. In the present study, increased expression of a functional CdCDR1 gene in the C. dubliniensis derivative CD57R mediated an eightfold increase in ketoconazole and itraconazole MICs but did not affect the susceptibility of the isolate to fluconazole . Previously, many investigators have suggested that multiple efflux mechanisms may contribute to fluconazole resistance in a single isolate . It has been suggested that simultaneous activation of multiple fluconazole efflux mechanisms in a single cell would lead to an increased rate of fluconazole efflux and therefore increased resistance to this compound. However, the present study demonstrates that the effects of MDR1 and CDR1 overexpression on fluconazole susceptibility in a single strain are not necessarily additive. It is not clear from these data why the Cdr1p and Mdr1p transporters do not have a cumulative effect on fluconazole susceptibility when expressed in concert. Perhaps there is a threshold level of efflux activity, above which the activation of further efflux proteins has no further effect on fluconazole susceptibility. Alternatively, there may be competition between CdCdr1p and CdMdr1p for available substrate (i.e., fluconazole). In this latter scenario, if one transporter has a higher affinity for fluconazole (perhaps in this case CdMdr1p), it may sequester all of the available substrate, rendering transporters with lower affinity for fluconazole redundant. The precise contribution of CdMDR1 overexpression in fluconazole resistance in specific strains (e.g., CD57RM4 [cdcdr1Delta::FRT/cdcdr1Delta::FRT]) will be determined by inactivating the CdMDR1 gene and determining the effect of this mutation on fluconazole susceptibility. At present, it is not possible for us to analyze the role of CdCDR1 as a sole mechanism of fluconazole resistance in C. dubliniensis as we have not encountered any CdCDR1-overexpressing fluconazole-resistant clinical isolates or in vitro-generated derivatives which do not also coexpress the CdMDR1 gene to date. In C. albicans however, several investigators have described isolates in which CDR1 is the sole efflux pump activated, indicating that CDR1-type efflux pumps are capable of mediating fluconazole resistance in this species . However, coactivation of Mdr1p- and Cdr1p-type pumps has been described in fluconazole-resistant isolates of C. albicans and Candida glabrata and conceivably in these species, as in C. dubliniensis, the effects of both pumps on the resistance phenotype may not be additive . The data presented here indicate that the roles of individual transporters in the fluconazole resistance phenotypes observed in C. albicans and C. glabrata species will have to be dissected by targeted disruption of the individual genes in question. Backmatter: PMID- 12183244 TI - Ertapenem versus Ceftriaxone Followed by Appropriate Oral Therapy for Treatment of Complicated Urinary Tract Infections in Adults: Results of a Prospective, Randomized, Double-Blind Multicenter Study AB - The efficacy and safety of intravenous (i.v.) ertapenem (1 g once a day) with the option to switch to an oral agent for treatment of adults with complicated urinary tract infections (UTIs) were compared with that of i.v. ceftriaxone (1 g daily) with the same oral switch option in a multicenter, double-blind, prospective, randomized study. At entry, 592 patients were assigned to one of two strata: acute pyelonephritis or other complicated UTI without acute pyelonephritis. After a minimum of 3 days, patients could be switched to an oral antimicrobial agent. A total of 159 patients in the ertapenem group and 171 patients in the ceftriaxone group were microbiologically evaluable. Approximately 95% of patients in each treatment group were switched to oral therapy. The most common pathogens were Escherichia coli and Klebsiella pneumoniae. At the primary efficacy endpoint 5 to 9 days after treatment, 91.8% of patients who received ertapenem and 93.0% of those who received ceftriaxone had a favorable microbiological response (95% confidence interval for the difference, adjusting for strata, -7.6 to 5.1%), indicating that outcomes in the two treatment groups were equivalent. Microbiological success rates for the two treatment groups were similar when compared by stratum and also by severity of infection. The frequency and severity of drug-related adverse events were generally similar in both treatment groups. In this study, ertapenem was as effective as ceftriaxone for the initial treatment of complicated UTIs in adults, was generally well tolerated, and had a similar overall safety profile. Keywords: Introduction : Urinary tract infections (UTIs) are responsible for over 7 million physician office visits in the United States each year . Most of these are uncomplicated infections of the urinary bladder in otherwise healthy young women and are easily managed with short-term antimicrobial therapy. Treatment of complicated UTIs, however, is generally less successful. Ten to 14 days of therapy with an agent active against a more extensive list of gram-negative bacilli is recommended . For serious infections, initial empirical therapy with a broad-spectrum parenteral antimicrobial agent followed by an oral agent to which the responsible uropathogen is susceptible has become standard therapy (, -). Follow-up urine cultures are important because bacterial eradication is more difficult to achieve, and recurrence several weeks posttherapy is not uncommon. Ertapenem (formerly MK-0826; Merck & Co., Inc.) is a once a day parenteral beta-lactam agent with excellent in vitro activity against many gram-positive and gram-negative aerobes and anaerobes generally associated with community-acquired infections, including most members of the family Enterobacteriaceae, pathogens most commonly responsible for UTIs . Ertapenem is not indicated for Pseudomonas aeruginosa or enterococci, which are more often associated with nosocomial infections. The objectives of this study were to compare the efficacy, tolerability, and safety of 1 g of ertapenem once a day with those of 1 g of ceftriaxone once a day, both followed by optional oral therapy, for the treatment of adult patients with complicated UTIs. (These data were presented in part at the 41st Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, Ill., 17 December 2001.) MATERIALS AND METHODS : Patients. | Patients aged >=18 years with a complicated UTI (i.e., acute pyelonephritis, UTI in men, or UTI associated with obstruction, foreign bodies, or urologic abnormalities) were eligible for the study if they required initial parenteral antimicrobial therapy and if the infection was caused by a pathogen susceptible to the study drugs. Criteria for acute pyelonephritis included fever, flank pain or costovertebral angle tenderness, pyuria (>=10 white blood cells), and positive urine culture (>=105 CFU of a uropathogen/ml) within 48 h of enrollment. Criteria for other complicated UTIs were signs or symptoms of UTI, pyuria, and positive urine culture in a male; female patients were additionally required to have either an indwelling catheter, current bladder catheterization or instrumentation of the urinary tract, or functional or anatomical abnormality of the urinary tract. Patients with any of the following were excluded from the study: pregnancy or lactation in women, history of serious allergy to study therapy (patients with a history of mild rash to beta-lactams could be enrolled), complete obstruction of the urinary tract, perinephric or intrarenal abscess, prostatitis, any rapidly progressive disease, immune-compromising illness or therapy, the need for concomitant antimicrobials, acute hepatic failure, requirement for peritoneal dialysis or hemodialysis, treatment with a systemic antimicrobial agent for >=24 h within 72 h prior to the baseline urine culture, creatinine clearance of <30 ml/min, aspartate aminotransferase or alanine aminotransferase levels of >6 times the upper limit of normal (ULN), bilirubin or alkaline phosphatase levels of >3 times the ULN, absolute neutrophil count of <=1,000 per mul, platelet concentration of <75,000 per mul, hematocrit level of <25%, or coagulation tests of >1.5 times the ULN. Study design and antimicrobial therapy. | This prospective, double-blind (with sponsor blinding), randomized, multicenter study was conducted from April 1998 to February 2000. Written consent was obtained from all patients, and the institutional review board at each participating site approved the protocol. Eligible patients from 31 centers in the United States, Central and South America, and Europe were stratified to pyelonephritis (with or without an abnormality of the urinary tract), or other complicated UTI without acute pyelonephritis. Equal randomization to one of the two treatment arms was accomplished by using a computer-generated random number allocation schedule. Both ertapenem and ceftriaxone were given once a day as a 1-g intravenous (i.v.) dose infused over 30 min. To avoid unblinding due to slight color differences between the study drugs, patients also received a second color-matched saline placebo infusion. After at least 2 days of hospital or clinic-based infusion therapy, study therapy could be completed in the hospital, a clinic, or at home. Consistent with current pyelonephritis treatment guidelines , at the investigator's discretion and after a minimum of 3 days of i.v. therapy, patients could be switched to oral ciprofloxacin (500 mg twice daily) if they were afebrile; nausea and vomiting had resolved; signs, symptoms, and leukocytosis had improved; and a urine culture was obtained. Other oral agents were permitted if the patient could not tolerate ciprofloxacin or if the causal pathogen was resistant. The suggested total duration of i.v. plus optional oral therapy was 10 to 14 days. Clinical assessments. | Patients were evaluated within 24 h of enrollment and daily thereafter while on parenteral study therapy. The clinical response was measured on day 3 to 5 of parenteral therapy, at the discontinuation of i.v. therapy (DCIV), 5 to 9 days posttherapy (test of cure [TOC] visit), and 4 to 6 weeks posttherapy (late follow-up [LFU]). Each infection was assessed prior to unblinding and considered severe if there was bacteremia, signs of sepsis (diastolic blood pressure of <60 mm Hg, altered mental status, or use of vasopressors), or at least three of the following: moderate to severe flank pain, fever of >101C, chills, nausea or vomiting, or a white blood cell count of >=15,000 per mul. Microbiological assessments. | Urine culture with quantitation and blood culture were performed at the baseline. All isolates were identified at the site laboratory, and pathogens were tested for in vitro susceptibility to ertapenem, ceftriaxone, and ciprofloxacin, following the guidelines of the National Committee for Clinical Laboratory Standards . Microbiologic efficacy was assessed at each time point by quantitative urine culture. After at least 48 h of study i.v. therapy, failure was defined as a urine culture with a >=104-CFU/ml concentration of any uropathogen present in the admission culture at a concentration of >=105 CFU/ml. Microbiologic responses during therapy and at the TOC visit were defined as eradication (the urine culture showed that a uropathogen present at a concentration of >=105 CFU/ml at entry was reduced to a concentration of <104 CFU/ml), persistence (the urine culture grew >=104 CFU of an original uropathogen/ml), persistence with acquisition of resistance, superinfection (the urine culture grew >=105 CFU of a uropathogen other than the baseline pathogen during therapy per ml), or new infection (>=105 CFU of a pathogen other than the baseline pathogen cultured after completion of therapy per ml). Additional responses at LFU were recurrence (eradication at TOC but >=104 CFU of an original uropathogen/ml at LFU) and recurrence with acquisition of resistance. Populations for analysis. | The treated population included patients who received at least one dose of study therapy. The microbiologic modified intent-to-treat (MITT) population included patients who received at least one dose of i.v. therapy, had a baseline uropathogen in any quantity, and had a follow-up quantitative urine culture after DCIV. Microbiologically evaluable patients had to meet the following criteria: clinical evidence of a UTI, baseline urine pathogen at a concentration of >=105 CFU/ml (>=104 CFU/ml if the patient was bacteremic), at least one baseline pathogen susceptible or intermediate to both i.v. study antimicrobials, and a urine culture at the TOC visit. The LFU evaluable population included evaluable patients successfully treated at the TOC visit who also had a follow-up assessment within the LFU window. Efficacy variables. | The primary efficacy variable was the assessment of bacteriologic eradication in the microbiologically evaluable population at the TOC visit. Additional assessments were the microbiologic response rates in the MITT population and the clinical and microbiologic responses in evaluable patients at DCIV. Microbiologic recurrence and clinical relapse were also assessed in evaluable patients at the LFU visit. Safety and tolerability assessment. | Patients in the treated population were monitored for adverse experiences daily during parenteral therapy and for 14 days after all study therapy (parenteral plus oral) was completed. The investigator categorized the intensity of the adverse event (mild, moderate, or severe) and the likelihood of its relation to the study drug (definitely not, probably not, possibly, probably, or definitely). The tolerability of the parenteral study drug at the local infusion site was evaluated by the investigator daily. Statistical analyses. | The study was designed to test for equivalence in the efficacy of the ertapenem and ceftriaxone microbiologically evaluable treatment groups. The sample size (minimum of 150 evaluable patients per group) was calculated with Blackwelder's formula and for the following values: alpha, 0.025; beta, 0.20; and the expected response rate in both groups, 90%. In accordance with the Food and Drug Administration's guidance for industry on developing antimicrobial drugs ---General Consideration for Clinical Trials (July 1998), equivalence for this study was determined by the 95% (two-sided) confidence interval (CI) for the difference in response rates between the two treatment groups (ertapenem minus ceftriaxone) . If the observed response rate in the comparator group was >90%, for equivalence to be demonstrated, the CI of the difference had to contain zero and its lower limit could not be less than -10 percentage points. CIs about the difference were calculated by using the normal approximation to the binomial distribution and were adjusted for strata by using the Cochran-Mantel-Haenzel approach . The treatment by stratum interaction was investigated using the Breslow-Day test of homogeneity of odds ratios and the Gail-Simon test, if needed. Kaplan-Meier analyses were also performed to explore the time to DCIV in the microbiologically evaluable population. The SAS, version 6.12, statistical package was used for analysis. No formal tests were performed based on baseline demographics or disease characteristics. RESULTS : Patients. | The distribution of the study patients is summarized in Fig. . A total of 592 patients were randomized: 298 to the ertapenem group and 294 to the ceftriaxone group. A total of 159 (53.4%) patients in the ertapenem group and 171 (58.2%) patients in the ceftriaxone group were microbiologically evaluable. The most common reason patients were not evaluable was failure to isolate >=105 CFU of a uropathogen/ml at baseline. Baseline demographics and disease characteristics of the two treatment groups in the randomized and microbiologically evaluable populations were generally similar . Approximately 30% of the patients in each treatment group of both populations were male. In the evaluable population, 83 (52.2%) patients in the ertapenem group and 92 (53.8%) patients in the ceftriaxone group had one or more urinary tract abnormalities; indwelling bladder catheters or stents were present in 18 (11.3%) and 20 (11.7%) patients in the ertapenem and ceftriaxone treatment groups, respectively. Approximately half of the evaluable patients in each treatment group had severe infection. Therapy. | The duration of parenteral, oral, and total (parenteral plus oral) study therapy in the treated and microbiologically evaluable patients was comparable in the ertapenem and ceftriaxone treatment groups . Of the evaluable patients, 95.6% (152 of 159) of those treated with ertapenem and 94.7% (162 of 171) of those treated with ceftriaxone were switched to oral therapy, most commonly ciprofloxacin (approximately 94% in each treatment group). The Kaplan-Meier curves for the time to the switch to oral therapy were similar for ertapenem and ceftriaxone (data not shown). Most patients in both treatment groups were switched to oral therapy by study day 4. Baseline microbiology. | The distribution of the pathogens in each treatment and their susceptibility profiles were comparable. The most frequently isolated pathogen was Escherichia coli, recovered from 111 (69.8%) patients in the ertapenem group and 117 (68.4%) patients in the ceftriaxone group, followed by Klebsiella pneumoniae, which was cultured from 22 (13.8%) patients in the ertapenem group and 21 (12.3%) patients in the ceftriaxone group. For all other species, there were fewer than 10 isolates per group. All isolates were susceptible to ertapenem, and all were susceptible or intermediate to ceftriaxone, with the exception of two (33.3%) of six enterococcus isolates in the ertapenem group and three Pseudomonas isolates (two of six isolates in the ertapenem group and one of two isolates in the ceftriaxone group). Twenty microbiologically evaluable patients who received ertapenem and 15 patients treated with ceftriaxone were bacteremic at the baseline; the etiologic agent in 27 (75.0%) of these patients was E. coli. Efficacy. | At the primary efficacy endpoint, 91.8% of the microbiologically evaluable patients in the ertapenem group and 93.0% of those in the ceftriaxone group had a favorable microbiologic response assessment (95% CI for the difference, adjusting for strata, -7.6 to 5.1%), indicating equivalence of the two treatments. In the supportive microbiologic MITT analysis, which included 80% of all patients randomized, 92.5% treated with ertapenem and 91.1% treated with ceftriaxone had a favorable microbiologic response (95% CI for the difference, -4.0 to 6.9%), supporting the results of the primary analysis. Overall bacterial eradication rates in the evaluable population are summarized by stratum, subgroup, and time of assessment (DCIV and the TOC visit) in Table . There were no patients in either group who failed treatment prior to receiving three doses of i.v. therapy. Eradication rates by pathogen are shown in Table . One persistent pathogen (E. coli) in the ertapenem group developed resistance to the oral study drug (ciprofloxacin), none had developed resistance to ertapenem. In the ceftriaxone group, one persistent pathogen (Citrobacter freundii) developed resistance to ceftriaxone and two pathogens (Proteus mirabilis and E. coli) had developed resistance to the oral agent (ciprofloxacin). Seventeen of the 20 (85.0%) bacteremic patients in the ertapenem group and 13 of the 15 (86.7%) in the ceftriaxone group had a favorable overall microbiologic response. Failures were a result of persistent bacteriuria; persistent bacteremia was not documented in any of these patients. Bacterial recurrence rates in patients evaluable for the LFU assessment (n = 105 for ertapenem; n = 121 for ceftriaxone) were 7.6% of patients treated with ertapenem and 8.3% of those treated with ceftriaxone (95% CI for the difference, -8.7 to 7.4%), indicating that recurrence rates in the two treatment groups were similar. In the ertapenem group none of these recurrent pathogens had developed resistance to either the parenteral or oral study drug, whereas one recurrent pathogen (E. coli) in the ceftriaxone group developed resistance to the oral agent (ciprofloxacin). Clinical relapse rates at the LFU visit (5.4% of patients treated with ertapenem [n = 114] and 7.9% of those treated with ceftriaxone [n = 135]) were also similar. Nonbaseline, emergent infections included new infections, which occurred in 27 (17.0%) microbiologically evaluable patients in the ertapenem group and 21 (12.3%) patients in the ceftriaxone group, and superinfections, which developed in one patient in each treatment group. The most common pathogens causing new infections in both groups were enterococci, E. coli, and K. pneumoniae, in descending order of frequency. One isolate each of Pantoea agglomerans and P. aeruginosa was responsible for superinfection in the ertapenem and ceftriaxone groups, respectively. Safety and local tolerability. | A total of 293 patients in the ertapenem group and 289 patients in the ceftriaxone group received at least one dose of study i.v. therapy and were evaluated for adverse experiences. During i.v. therapy, one or more drug-related adverse experiences were reported for 86 (29.4%) patients in the ertapenem group and 76 (26.3%) patients in the ceftriaxone group, none of which was considered to be serious. The most common drug-related adverse events in the treated population were diarrhea (ertapenem, 21 [7.2%] patients; ceftriaxone, 19 [6.6%] patients), headache (ertapenem, 19 [6.5%] patients; ceftriaxone, 13 [4.5%] patients), and nausea (ertapenem, 16 [5.5%] patients; ceftriaxone, 10 [3.5%] patients). Parenteral study therapy was discontinued due to a drug-related adverse experience in 8 (2.7%) patients in the ertapenem group and none in the ceftriaxone group. Events resulting in discontinuation of ertapenem were rash in three patients, one of whom reported a prior penicillin allergy, and pruritus, vomiting, paresthesia, diarrhea, and syncope in one patient each. Drug-related laboratory adverse experiences were reported in 27 (9.7%) patients in the ertapenem group and 20 (7.2%) patients in the ceftriaxone group. None of these events was considered serious, and none resulted in discontinuation of study therapy. The most common drug-related adverse laboratory events were elevations of alanine aminotransferase levels (3.8% [9 of 240 patients] in the ertapenem group and 4.6% [11 of 238 patients] in the ceftriaxone group) and aspartate aminotransferase levels (3.7% [10 of 268 patients] and 3.4% [9 of 261 patients] in the ertapenem and ceftriaxone groups, respectively). Twenty-seven (9.2%) patients in the ertapenem group and 20 (6.9%) patients in the ceftriaxone group experienced reactions of moderate to severe intensity at the local infusion site. The most common symptom in both treatment groups was pain followed by tenderness. FIG. 1. | Profile of patient enrollment. Profile of patient enrollment. TABLE 1 | Baseline characteristics and therapy of randomized and microbiologically evaluable patients with complicated UTI TABLE 2 | Favorable microbiologic response assessments in microbiologically evaluable patients with complicated UTI TABLE 3 | Eradication rates at TOC in microbiologically evaluable patients with complicated UTI DISCUSSION : In this double-blind, randomized study, the efficacy and safety of ertapenem were compared with those of ceftriaxone for the empirical treatment of serious complicated UTI judged by the investigator to require initial therapy with a parenteral antimicrobial agent. The study design was consistent with standard clinical practice (-). After a minimum of 3 full days of parenteral study therapy, investigators had the option to switch to oral ciprofloxacin (or other agent) if the patient had clinically improved. Approximately 95% of the evaluable patients in both treatment groups were switched to oral therapy (ciprofloxacin in about 94% of patients in each treatment group), usually by study day 4. The Kaplan-Meier curves for the time to switch to oral therapy in the microbiologically evaluable population appeared similar for both study drugs, indicating that the time to defervescence and improvement of clinical signs and symptoms was comparable for both ertapenem and ceftriaxone. Results of this study show that ertapenem therapy, 1 g once a day, was highly effective in all patient populations, including those with pyelonephritis or severe infection and the elderly, and was equivalent to treatment with ceftriaxone. Over 90% of the evaluable patients in both treatment groups had a favorable microbiologic response assessment at the TOC visit 5 to 9 days posttherapy. These rates are similar to or better than those reported in previous studies of patients with complicated UTIs (-, , , ), and they exceed the cure rates expected when evaluating a new anti-infective drug for the treatment of UTI (>80% for acute pyelonephritis and >65% for complicated UTI) . As expected, success rates were higher at the completion of i.v. therapy than at the TOC visit for both study drugs. This difference was more apparent for patients with other complicated UTIs than for those with acute pyelonephritis, which suggests that the few additional failures posttherapy were due to the underlying urinary tract abnormality or inadequate oral therapy rather than failure of the parenteral antimicrobial agent. Ertapenem has excellent activity against many gram-positive and gram-negative aerobic, facultative, and anaerobic bacteria generally associated with infections acquired in the community. Several studies have shown that most members of the family Enterobacteriaceae are highly susceptible to ertapenem, including isolates resistant to several other antimicrobial agents . In this study, members of the family Enterobacteriaceae, all of which were susceptible to ertapenem, accounted for 93.0% of all isolates from microbiologically evaluable patients in both treatment groups. Similar to this study, other studies have shown that E. coli and K. pneumoniae account for the majority of responsible pathogens in complicated UTIs . Ertapenem is thus a reasonable choice for treatment of complicated UTIs when infection is not caused by P. aeruginosa or enterococci, both of which were recovered from very few patients in this study. The overall safety and tolerability profile of ertapenem in this study was similar to that of ceftriaxone. Drug-related clinical adverse events most frequently reported for both agents were related to the gastrointestinal tract. The most common drug-related laboratory adverse event for both drugs was mild to moderate elevation of transaminase levels, which tended to be transient and without clinical consequence. In summary, in adult patients with moderate to severe complicated UTI requiring initial parenteral therapy, ertapenem (1 g once a day), with the option to switch to an appropriate oral antimicrobial agent after clinical improvement, was generally well tolerated and highly effective both clinically and microbiologically. Backmatter: PMID- 12183285 TI - Relationship between Antibiotic Resistance in Streptococcus pneumoniae and That in Haemophilus influenzae: Evidence for Common Selective Pressure AB - Keywords: We report data from a recent international surveillance study noting significant interspecies relationships between resistance in Streptococcus pneumoniae and Haemophilus influenzae, both common causative agents of respiratory tract infections . Previous studies have shown geographic variation in the prevalence of beta-lactam and trimethoprim-sulfamethoxazole (SXT) resistance among S. pneumoniae and H. influenzae (, , -), which may be attributed to differences in antibiotic consumption and infection control practices between countries . Interspecies relationships of resistance prevalence are not commonly reported. During 1999 to 2000, 5,015 S. pneumoniae and 4,846 H. influenzae isolates were prospectively collected from hospital laboratories in China (three sites), Hong Kong (two sites), South Korea (four sites), Thailand (four sites), France (nine sites), Germany (eight sites), Ireland (one site), Italy (six sites), Spain (six sites), the United Kingdom (seven sites), South Africa (eight sites), Brazil (six sites), and Mexico (five sites). Isolates were submitted to Focus Technologies (Herndon, Va.) for antimicrobial susceptibility testing by NCCLS broth microdilution using NCCLS 1999 interpretive criteria . Relationships between the proportion of resistant isolates in each country to individual antimicrobial agents were tested by measuring the correlation coefficient (r) using a t test analysis. Antimicrobial susceptibility of S. pneumoniae and H. influenzae to all agents tested is shown . Among S. pneumoniae organisms, considerable variations in resistance to beta-lactams, macrolides, and SXT were detected between countries, especially in Eastern regions of the world. Levofloxacin resistance by S. pneumoniae was rare in most countries (except China, 3.3%, and Hong Kong, 8.0%) and undetected in H. influenzae. Ampicillin and SXT resistance among H. influenzae varied considerably by country, with high rates of resistance detected in the same regions as those with higher resistance rates among pneumococci . An interspecies comparison of the proportion of penicillin-resistant S. pneumoniae and ampicillin-resistant H. influenzae by country demonstrated a highly significant relationship (r = 0.928; P < 0.001). A significant interspecific relationship (r = 0.714; P < 0.01) was also recorded for SXT resistance. No significant interspecific relationships (P > 0.05) were found for resistance to amoxicillin-clavulanate, cefuroxime, clarithromycin, or levofloxacin. The interspecific relationship observed for beta-lactam and SXT resistance has not been reported previously. beta-Lactam resistance in S. pneumoniae and H. influenzae is conferred by unrelated mechanisms; respectively, penicillin-binding protein alterations and expression of TEM-1 or ROB-1 beta-lactamases . S. pneumoniae resistance to SXT is chromosome encoded by mutations in folA and sulA . In H. influenzae, mutations in the chromosome-encoded folH have been shown to encode trimethoprim resistance , while sulfamethoxazole resistance is likely chromosome encoded through a sulA analogue. As such the relationship between resistance in these species suggests a response to a common selective pressure with increases in resistance likely driven independently of each other, either by acquisition of a resistance mechanism or through clonal expansion of resistant isolates, a phenomenon well documented in S. pneumoniae but not in H. influenzae. No interspecies correlation between the incidences of clarithromycin, cefuroxime, or levofloxacin resistance was observed. These observations are presumably despite exposure to the same antibiotic selective pressure. The relationship between resistance in these species provides evidence that antibiotic usage drives resistance and suggests that attempts to control the emergence of resistance in S. pneumoniae through prudent antibiotic consumption or improved infection control may impact on resistance prevalence in H. influenzae and perhaps other species. Surveillance of antibiotic consumption and infection control practices is necessary to further explore this phenomenon. (This work was previously presented in part at the 40th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Canada [A. M. Staples, C. Thornsberry, I. A. Critchley, K. S. Murfitt, D. F. Sahm, and M. E. Jones, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. C-1221, 2000].) FIG. 1. : Incidence of resistance to different antimicrobials tested against S. pneumoniae Incidence of resistance to different antimicrobials tested against S. pneumoniae and H. influenzae isolates by country, utilizing NCCLS interpretative criteria. In order by country, the numbers of S. pneumoniae and H. influenzae isolates and the percentage that was H. influenzae beta-lactamase positive were as follows: Brazil (448, 274/10.6%), China (214, 198/10.6%), France (547, 454/5.6%), Germany (560, 639/6.9%), Hong Kong (175, 178/27.5%), Ireland (17, 37/10.8%), Italy (491, 521/10.6%), Mexico (271, 226/17.7%), South Africa (701, 627/7.3%), South Korea (351, 272/60.7%), Spain (492, 559/20.9%), Thailand (206, 305/45.5%), and the United Kingdom (542, 556/11.3%). Backmatter: PMID- 12183286 TI - gyrA Mutations Associated with Nalidixic Acid-Resistant Salmonellae from Wild Birds AB - Keywords: The emergence of quinolone-resistant isolates of salmonellae has been related to the use of antibiotics in veterinary medicine, which causes an important impact on the selection of resistance . It has also been demonstrated that veterinary clones possess the same mechanism of resistance as human clinical isolates . Quinolone resistance in salmonellae is mainly associated with mutations in the quinolone resistance-determining region (QRDR) of the gyrA and parC genes . However, little is known about what the incidence of resistance in wild animals is and which of the mechanisms of resistance these strains have, although it is thought that wild animals are also integrated in the same epidemiologic cycle . The aim of this study was to analyze the prevalence of quinolone-resistant isolates of salmonellae from wild birds and determine the mechanisms involved. A total of 45 Salmonella enterica subsp. enterica isolates were obtained from stool samples from wild birds just arrived to GREFA Wild Life Hospital. Seven strains (15.5%) were selected on the basis of resistance to nalidixic acid. The MICs of nalidixic acid (Sigma Aldrich, Madrid, Spain), ciprofloxacin, and enrofloxacin (Bayer, Leverkusen, Germany) were measured by a twofold agar dilution method and interpreted according to the recommendations of the National Committee for Clinical Laboratory Standards guidelines . To identify gyrA and parC mutations in resistant isolates, PCR and direct DNA sequencing were performed as follows. A 312-bp fragment of the QRDR of the gyrA gene was amplified from a genomic DNA template by using specific primers gyr A-1 (5'-GGTACACCGTGCCGTACTTT-3') and gyr A-2 (5'-TCCACGAAATCCACCGTC-3') corresponding to positions 17 to 137 and 311 to 328, respectively. These salmonella-specific primers were constructed on the basis of Salmonellaenterica serotype Typhimurium gyrA gene sequence data (GenBank accession number ). The primers for the amplification of the QRDR of the parC gene have been described elsewhere . The mutations in the gyrA gene leading to the amino acid substitutions and MICs are shown in Table . TABLE 1 : Susceptibility to quinolone and GyrA substitutions of S. enterica subsp. enterica isolates from wild birds Strains S2, S9, S10, S14, and S20 carried a Cys-284->Ala transversion, resulting in a Ser-83->Tyr substitution in the GyrA subunit. This mutation has been described in Salmonella serotype Typhimurium strains resistant to nalidixic acid and reduced the susceptibility to fluoroquinolones isolated from food-producing animals . Strains S57 and S60 carried a Gly-259->Cys transition, resulting in an Asp-87->Asn substitution in the GyrA subunit. This mutation has also been described in Salmonella isolates of human and animal origin . No mutations in the QRDR of parC were detected. These results may confirm that parC mutations are not necessary to obtain a high level of resistance to nalidixic acid . The mechanisms of resistance described here are the same as the ones described for humans and domestic animals. Therefore, it is possible that wild animals be included in the epidemiologic cycle of the spread of resistance and the dissemination of resistant bacteria among animals and the environment. Further studies are necessary to obtain more information about what is the real role of wild animals in these facts. Nucleotide sequence accession numbers. : The GenBank accession numbers for the partial sequences of the quinolone-resistant strains are , , , , , , and . Backmatter: PMID- 12183284 TI - Prevalence of Oxacillin Resistance in Staphylococcus aureus among Inpatients and Outpatients in the United States during 2000 AB - Keywords: Recent reports documenting community-acquired infections with methicillin-resistant Staphylococcus aureus (MRSA) are cause for concern . However, not all MRSA infections detected within the community were acquired there. In recent years, healthcare practices have shifted MRSA-infected individuals from hospitals into surrounding communities, for treatment and convalescence at home, for example . While the epidemiology of nosocomial MRSA has been well studied, less is known about the dissemination of MRSA within communities surrounding hospitals. Generally, apart from unique community strains, it has been thought that MRSA is transferred from hospital environments, known reservoirs of antimicrobial resistance, into surrounding communities. Additionally, patients may become carriers of MRSA during a stay in a healthcare institution and later, perhaps at home, manifest an infection from this MRSA, blurring the definitions of community- and hospital-acquired infections. If these assumptions were true, one would expect the prevalence of MRSA to be higher in outpatients served by hospitals with higher rates of MRSA. To test this hypothesis, we analyzed data from The Surveillance Network (TSN) Database ---USA (Focus Technologies, Inc., Herndon, Va.). TSN Database ---USA electronically collects daily routine antimicrobial susceptibility testing and patient demographic data from laboratories, which are the data upon which clinical decisions are made . Participant institutions are geographically dispersed throughout the nine U.S. Census Bureau regions. All data are filtered through expert rule algorithms to remove repeat isolates and identify microbiologically atypical results for confirmation or verification before being merged into the final database. The information provided to TSN databases allows us to confidently differentiate inpatients from outpatients but does not allow us to determine whether an infection was community or hospital acquired. Outpatient data contained in TSN databases also include isolates from emergency room visits. For this analysis, we excluded isolates from nursing home and other long-term-care-facility patients, frequently colonized with MRSA. To test our hypothesis that a relationship exists between the prevalence of MRSA in outpatients and inpatients in the same geographic area, we reviewed routine laboratory data comprising 264,687 nonrepeat isolates of S. aureus collected in 2000 from 121 hospitals in TSN Database ---USA, each testing at least 100 isolates. A highly significant relationship (P < 0.0001) was found between the rates of MRSA among outpatient and inpatient isolates of S. aureus in 2000 . This relationship was confirmed by linear regression analysis (r = 0.75). Almost invariably, hospitals with higher rates of MRSA among their inpatients had higher rates of MRSA among outpatient isolates submitted to their laboratories for culture and antimicrobial susceptibility testing. Similarly, hospitals with lower rates of MRSA among their inpatients had lower rates of MRSA among outpatients from their surrounding communities. Multiple drug resistance, defined as concurrent resistance to three or more antimicrobial drug classes, comprised 76.1% and 66.5% of inpatient and outpatient isolates, respectively. The prevalence of MRSA has continued to increase in the United States. A retrospective analysis of TSN Database ---USA showed that the prevalence of MRSA in 33 hospitals participating from 1996 to 2000 increased from 30.1% (4,557/15,143) in 1996 to 45.7% (27,495/60,149) in 2000 (P < 0.0001) in inpatient isolates; among outpatient isolates, the prevalence of MRSA increased from 17.3% (1,781/10,268) to 28.6% (17,858/62,401) (P < 0.0001). Previous reports have noted a growing concern regarding MRSA infections within some community populations, particularly in children and those with dermatological disorders, renal dysfunction, or HIV infection . The clear relationship between the rates of MRSA infection among inpatients and outpatients demonstrated here illustrates that MRSA infections are no longer confined to hospitals and are certainly present in the community. It is likely that these include both distinct community clones and those carried by patients formerly either directly or indirectly exposed to the healthcare environment, as demonstrated in a recent U.S. study . The present report corroborates the need for heightened awareness of MRSA among community health-care practitioners, including up-to-date knowledge of trends in MRSA at local and regional levels. There is also a need for more rapid, reliable methods to detect and report the presence of methicillin resistance among isolates of S. aureus, especially among outpatients, as standard empirical therapies may not be effective in treating infections by MRSA, especially those derived from hospital environments with a greater tendency for multidrug resistance. FIG. 1. : The proportion of MRSA in outpatient samples plotted against the proportion of MRSA in inpatient samples for each of the 121 hospitals able to be evaluated which participated in TSN Database ---USA during 2000. The proportion of MRSA in outpatient samples plotted against the proportion of MRSA in inpatient samples for each of the 121 hospitals able to be evaluated which participated in TSN Database ---USA during 2000. Backmatter: PMID- 12183241 TI - Effects of Neutrophils on Cefazolin Activity and Penicillin-Binding Proteins in Staphylococcus aureus Abscesses AB - Bacteria survive within abscesses despite antimicrobial therapy, usually necessitating drainage. Our previous work showed that bacterial killing is diminished within the neutrophils of animals with abscesses. To further assess the role of neutrophils in Staphylococcus aureus survival and the poor activities of beta-lactams in abscesses, tissue cage abscess-bearing rats were given polymorphonuclear leukocyte (PMN)-depleting antibody prior to and several times following inoculation of the tissue cages with S. aureus. Cefazolin (300 mg/kg of body weight/day) was administered to all animals in appropriately divided doses. After 7 days of antimicrobial therapy, the 17 animals that received anti-PMN serum had significantly fewer abscess neutrophils than the 18 controls and fewer abscess bacteria (5.55 versus 3.79 log10 CFU/ml [P = 0.04]) than the 18 controls. The data were consistent with the premise that cefazolin is more effective in abscesses depleted of neutrophils. To investigate further, S. aureus was incubated with rat peritoneal neutrophils; and bacterial cell membrane proteins were isolated, labeled with biotinylated ampicillin, separated by electrophoresis, blotted onto nitrocellulose, and stained for biotin reactivity. PBP 2 expression was consistently and significantly decreased after a brief, nonkilling PMN exposure. These experiments showed that PMN depletion enhanced the activity of cefazolin in the abscess milieu. Furthermore, altered bacterial cell wall cefazolin targets may be the mechanism by which the PMN diminishes antimicrobial activity, suggesting the importance of the staphylococcus-PMN interaction in the outcome of established infections. Keywords: Introduction : The rapid delivery of neutrophils over the first 4 to 6 h to a site of infection is known to play an important role in controlling an acute infection . Once the infection is established, however, the activities of antimicrobials in a suppurative environment are limited. Most abscesses require surgical drainage for cure. However, some abscesses respond to antimicrobial therapy without drainage , and the mechanisms by which antimicrobial activity is diminished in a suppurative environment are only partially understood. Although the diminished growth rate of bacteria in abscesses is likely important in affecting the reduced effectiveness of antimicrobials in a suppurative environment, we had previously been unable to correlate antimicrobial activity in an abscess with the antimicrobial activity against stationary-phase organisms . We found that beta-lactam antimicrobials had only marginal efficacies, despite adequate antimicrobial concentrations, and that the activities of antimicrobials in polymorphonuclear leukocytes (PMNs) correlated with in vivo efficacy. The PMN may act as a sanctuary since we have previously found that killing by PMNs, but not phagocytosis of staphylococci, (i) is inhibited systemically by the presence of an abscess and (ii) may be further inhibited in the abscess milieu . In this study, we sought to directly determine the role of the PMNs in diminishing the activities of beta-lactams in an abscess. We postulated that neutrophils, although important in the immediate control of pyogenic infections, may inhibit the activities of beta-lactams in an established abscess and that a reduction in the rate of PMN influx into an abscess could enhance the killing effects of beta-lactams. Previous studies have provided some evidence to support this hypothesis, but the question either was not studied directly or conflicting results were presented. Gerding et al. showed improved killing of Staphylococcus aureus by cephalothin in a subcutaneously implanted cellulose tubing chamber which excluded cells or molecules >15 kDa compared to the killing by cephalothin in a tissue cage abscess in the presence of neutrophils. However, the differences in anatomy and blood supply in the control and test models made the precise role of the PMNs less well defined. Calame et al. evaluated the effects of granulocytes on the clearance of S. aureus by cloxacillin over 4 h. They found that when small doses of cloxacillin were used, bacterial clearance was diminished by PMN depletion, but when the largest dose was used, the antimicrobial activity was enhanced by the depletion of neutrophils (although statistical significance was not provided). It is known that neutrophils diminish the penicillin-binding protein (PBP) expression of Escherichia coli . S. aureus has been shown to have altered gene expression on internalization by endothelial cells . We postulated that if neutrophils inhibit the activities of beta-lactams against S. aureus, the inhibition might also be through a mechanism by which bacteria are phagocytized but not killed, with subsequent alteration in the expression of PBPs that are essential targets for beta-lactam activity. To test our postulates, we established a novel in vivo neutropenic model of infection; the animal model used is large enough to allow repeated sampling of an ongoing infection, yet it is small enough to deplete circulating neutrophils by systemic treatment rather than exclusion from a particular site. This report details the quantitative differences in the infective organism, the differences in the host cell profile, and the differences in antimicrobial activity between healthy and neutropenic rats with established infections and offers suggestions, based on additional experiments, on the mechanisms of the differences. MATERIALS AND METHODS : Abscess model. | Thirty-five male rats (weight, >400 g) were anesthetized, and a single table-tennis ball with 300 1.5-mm-diameter holes was implanted in the peritoneal cavity by sterile techniques. Six weeks after implantation, 17 rats were given 0.75 ml of anti-PMN serum of rabbit origin (Accurate Chemical Corp., Westbury, N.Y.) mixed with 1.25 ml of sterile saline 1 day before infection and on days 2, 4, and 6 after infection. Preliminary experiments showed that administration of anti-PMN serum at these times was most effective at depleting peripheral blood neutrophil counts. Eighteen animals served as controls and received control rabbit serum in saline. One day after administration of the initial dose of anti-PMN serum, the capsules in all 35 rats were inoculated with 5 x 107 CFU of an S. aureus isolate initially isolated from a bacteremic patient. The MIC of cefazolin for the organism is 0.5 mug/ml, and the minimal bactericidal concentration is 1.0 mug/ml . One day after bacterial inoculation, all animals were given cefazolin at 300 mg/kg of body weight/day (with one-third of the dose given at 8 a.m. and two-thirds of the dose given at 5 p.m.) for 7 days. The capsules were sampled for determination of bacterial and PMN counts immediately before administration of the 8 a.m. dose on days 2, 4, and 7 after the initiation of antimicrobial therapy. Abscess fluid was subjected to sonication at 90 W (model 300 Sonic Dismembrator; Fisher, Springfield, N.J.) for 15 s to liberate phagocytized S. aureus. Aliquots were serially diluted (10-fold), plated onto blood agar plates, and incubated for 24 h. The lower limit of detection was 10 CFU/ml. Neutrophils were quantified by determination of total cell counts in fresh abscess fluid with a hemocytometer, followed by determination of differential counts of stained cytospin preparations of abscess fluid to determine percent neutrophils. Peripheral blood PMN counts for all animals were determined on the same days that the capsules were sampled. Serum for counts and assay was obtained by collecting blood from the tail tip and placing it in pediatric serum separator tubes (Microtainer; Becton Dickinson Biosciences, Franklin Lakes, N.J.). Antibiotic assay. | The cefazolin concentration was measured by an agar disk diffusion bioassay with Bacillus subtilis ATCC 66333 as the test organism. For preparation of a five-point standard curve, known quantities of cefazolin standard in control rat serum were prepared in plates with antibiotic agar 5 and B. subtilis that were run simultaneously with 20-mul samples of sterile-filtered abscess fluid (or serum). The samples with the cefazolin standard were then added to 6.3-mm filter paper discs, and the discs were incubated for 24 h. The diameter of the clear zone surrounding the disk was used to prepare a standard curve and obtain the numbers of micrograms of cefazolin per milliliter. Rat PMN isolation. | A sterile 1-mg/ml solution of oyster shell glycogen (Sigma Chemical, St. Louis, Mo.) in saline was administered into the peritoneum of an anesthetized rat at a dose of 4 ml/100 g of body weight through an 18-gauge tubing-tipped needle. Four hours later (while the rats were under terminal anesthesia), 5 ml of sterile phosphate-buffered saline/100 g was administered into the peritoneum, and the fluid was drained to yield >108 neutrophils uncontaminated with erythrocytes. The cells were washed two times in Hank's balanced salt solution (HBSS; Sigma) without divalent cations and were counted by use of trypan blue staining to determine viability. Peripheral blood neutrophils harvested from the total blood volume of a single rat have occasionally been used for incubation of S. aureus with no significant difference in effect. These are isolated by standard procedures that include removal of monocytes by centrifugation of blood layered over Hypaque and lysis of erythrocytes with hypotonic ammonium chloride before a wash with HBSS. Incubation of PMNs-bacteria in vitro. | Five hundred milliliters of antibiotic medium 3 was spiked with an overnight growth of S. aureus, and the bacteria were allowed to grow for 6 h until the count was about 2 x 106 CFU/ml. Following centrifugation and a saline wash, half the bacteria (approximately 5 x 108) were suspended in HBSS with opsonizing rat serum and cations, and the suspension was incubated with rat peritoneal neutrophils (ratio of bacteria to neutrophils, 100:1) for 30 min at 37C in a water bath with slow shaking. The control bacteria (which were not incubated with PMNs but to which HBSS was added to form a volume equal to that of the test suspension) were incubated for 30 min at 37C in the same water bath. No lysis or wash to separate intracellular bacteria from extracellular bacteria was performed. Phagocytosis was evaluated in pilot experiments. At 15 min, stained cytospin preparations showed about 20 neutrophils per high-power field, or about 1.3 x 104 neutrophils/mul. About one-third of these contained about 50 bacteria each, which is considered to be very good phagocytosis. The general characteristics of the preparations showed that they resembled cytospin preparations of in vivo abscess fluid very early in infection. By use of bacteria and neutrophils at a 100:1 ratio in the in vitro incubation, it was determined on the basis of the results of pilot experiments that a large number of the S. aureus organisms were internalized. Aliquots of both control bacteria incubated without PMNs and bacteria incubated with PMNs were serially diluted in ice water and plated to determine viability. Methods for PBP isolation. | The S. aureus organisms were resuspended in 1 ml of phosphate buffer and were homogenized in a 5-ml steel blender cup on ice with micro-glass beads (two times for 5 min each time). The cell wall proteins were isolated by harvesting the supernatants obtained by centrifugation at 500 and 5,000 x g and the pellet obtained by centrifugation at 100,000 x g on an Optima LE-80K preparative ultracentrifuge (Beckman-Coulter, Fullerton, Calif.) with a 70.1 Ti rotor at 4C for 30 min. The pellets were each resuspended in 0.05 ml of phosphate buffer. A microassay of the protein in this suspension (Bradford assay; Bio-Rad Laboratories, Hercules, Calif.) showed that 12 to 13 mug of protein was usually recovered. The membrane pellets were incubated with biotinylated ampicillin (30 min at 30C) and recovered from a phosphate buffer wash by centrifugation at 100,000 x g. A total of 0.05 ml of 2x Laemmli sample buffer was added to the resuspended pellets, and the samples were boiled for 2 min and separated on sodium dodecyl sulfate-12% polyacrylamide gels next to prestained molecular weight markers. The gels were blotted onto nitrocellulose overnight. The nitrocellulose was blocked to control nonspecific staining and was then incubated with avidin-alkaline phosphatase (Zymed Labs, South San Francisco, Calif.) in phosphate-buffered saline (1:1,000) to detect the protein bands binding to biotinylated ampicillin. The colorimetric substrate for alkaline phosphatase was 5-bromo-4-chloro-3-indolylphosphate-nitroblue tetrazolium (BCIP-NBT; Kirkegaard & Perry Laboratories, Gaithersburg, Md.). Bands corresponding to important S. aureus PBPs were identified on the basis of a standard curve prepared on each blot from the molecular weight markers in that run. The blots were scanned into a density-quantifying program (ImageQuaNT; Molecular Dynamics, Sunnyvale, Calif.), and the volume and density of each PBP band were determined. Six separate runs of PBPs incubated with PMNs and control PBPs were performed with biotinylated ampicillin as the beta-lactam for the detection of PBPs. Procedural controls. | It was recognized that endogenous S. aureus protein could bind to streptavidin and thus make nonspecific bands, so procedural controls were run to determine the degree of nonspecific background staining. The controls were prepared by preincubation with (i) cold ampicillin before the addition of biotinylated ampicillin, (ii) no ampicillin but the biotin linker, and (iii) no ampicillin and no linker. Two other beta-lactam labels that did not use biotin recognition were also used as procedural controls: digoxigenin-ampicillin and fluorescent penicillin. Separate assays were also performed with washed neutrophils (instead of S. aureus) to determine if the PMN protein nonspecifically bound to the beta-lactam label. Controls were scanned into the ImageQuaNT program, and the densities of the peaks coinciding with S. aureus PBPs were determined. beta-Lactam labels for detection of PBPs. (i) Biotinylated ampicillin. | The biotinylation method of Dargis and Malouin was followed, with slight modification. Ampicillin was linked to biotin-7-N-hydroxysuccinimide, which formed a stable amide bond, by gentle rocking at room temperature for 2 h in a dimethyl sulfoxide-phosphate buffer solution. Unreacted biotin and linker were removed by the addition of Affigel 102 (Bio-Rad) in excess with rocking overnight. Biotinylated ampicillin was recovered by centrifugation and was used only on that day. (ii) Digoxigenin. | An amide bond between ampicillin and digoxigenin- N-hydroxysuccinimide ester (Roche Diagnostics, Indianapolis, Ind.) was made by the methods recommended by the supplier. After cleanup, the digoxigenin-ampicillin was stored at -70C until use (within 2 weeks) as a procedural control. Detection of proteins that bound to digoxigenin-ampicillin was done by blot incubation with anti-digoxigenin-alkaline phosphatase (1:1,500; Roche) in 100 mM Tris-saline. BCIP-NBT substrate for alkaline phosphatase was used for color development. (iii) Fluorescent penicillin. | By using our standard PBP isolation and gel separation procedures, fluorescent penicillin (Bocillin FL; Molecular Probes, Eugene, Oreg.) was used to quantify S. aureus PBPs by the labeling procedures of Zhao et al. . The sodium dodecyl sulfate-polyacrylamide gels were read on a STORM Fluor-Imager (Molecular Dynamics, Sunnyvale, Calif.) and analyzed with the ImageQuaNT program. Statistics. | Comparisons of bacterial and PMN counts and cefazolin concentrations were performed with data for control animals and animals treated with PMN-depleting antibody by Mann-Whitney U tests and other nonparametric tests with Statistica software. Bacterial and PMN counts were log transformed for presentation in graphic form. PBP binding data obtained by using the density values generated with the ImageQuaNT program were analyzed by two-tailed Student's t tests. P values <0.05 were considered significant. RESULTS : Animal model. | Six weeks after implantation of the table-tennis ball in large rats, the ball became encased in connective tissue, developed a blood supply, and became filled with a sterile fluid that had the appearance of serum . The single capsule, which is readily tolerated by larger rats, allowed us to correlate the cell numbers and bacterial counts in a single abscess with the peripheral cell counts, which is not possible with the three-capsule rabbit model in common use . Circulating and abscess PMN counts and cefazolin concentrations. | Use of anti-PMN serum significantly diminished both the circulating and the abscess PMN counts in rats infected with S. aureus during the entirety of the 7 days of treatment with cefazolin (Fig. and B). At day 7 the median circulating PMN counts were 9,549 and 2,137 PMNs/ml for the control animals and the animals treated with PMN-depleting antibody, respectively (P = 0.003), while the median abscess PMN counts were 3,388 and 436 PMNs/ml, respectively (P < 0.001). The cefazolin concentrations were similar in the abscess fluids of the two treatment groups. The abscess fluid of the controls had average cefazolin concentrations of 18.2 +- 12.5 mug/ml, and the abscess fluid of neutropenic rats had average cefazolin concentrations of 23.7 +- 12.8 mug/ml (P = 0.19). S. aureus counts in abscesses. | By days 2 and 4 of cefazolin therapy there was a trend toward lower median S. aureus concentrations in the abscesses of animals that received anti-PMN serum compared to those in the controls . At the end of therapy on day 7 of cefazolin administration, the control animals had a median of 5.55 log10 CFU of S. aureus per ml, whereas the animals receiving the anti-PMN serum had 3.79 log10 CFU of S. aureus per ml (P = 0.04 by the Mann-Whitney U test). In vitro S. aureus viability after PMN incubation. | Following 30 min of incubation with rat peritoneal neutrophils at a ratio of 100 bacteria per 1 PMN under conditions conducive to phagocytosis, the S. aureus cells showed no statistically significant change in viability (1.37 x 108 CFU/ml in the control tubes, compared to 1.30 x 108 CFU/ml in the tubes with PMN-incubated bacteria). PBP expression after incubation in PMNs. | After the in vivo experiments with rats demonstrated significant improvement in the killing by cefazolin in rats with neutropenia, we assayed S. aureus cells incubated with PMNs to look for targets that interact with beta-lactams, specifically, the PBPs. Data from six separate in vitro PMN incubations with isolation of PBPs from PMN-exposed and control S. aureus cells are presented in Fig. . PBP 2 expression (as defined by the density of the bound ampicillin on the separating gel) was decreased significantly by 30 min of incubation of the S. aureus cells with activated rat neutrophils, and the level of PBP 4 expression increased. Although the number of internalized S. aureus cells in the neutrophils were not counted, our earlier studies had shown that good phagocytosis by neutrophils occurs as early as 15 min . Biotinylated ampicillin was used as the PBP marker in these experiments. A representative image density graph of concurrently run control and PMN-exposed S. aureus PBPs is reproduced in Fig. . Figure also shows a control method for detection of biotinylated ampicillin. The PBP was preincubated with an excess of cold ampicillin before biotinylated ampicillin was added, and although the nonspecific band at the top shows intense binding, the cold ampicillin blocks the binding of biotinylated ampicillin to the major PBPs. Figure shows stained electrophoresis gel blots from the ImageQuaNT program (negatives are shown to better demonstrate the bands) of S. aureus PBPs incubated with PMNs and controls in the same run not incubated with PMNs. Several procedural controls were performed with these in vitro experiments. PBPs were detected with fluorescent penicillin (Bocillin FL) and digoxigenin-ampicillin, detected by scanning of a fluorescent gel and with anti-digoxigenin-alkaline phosphatase and alkaline phosphatase substrates, respectively. Procedural difficulties with the scanning and printing of data from the gels, which used the fluorescent penicillin label, halted progress with this beta-lactam marker, although the loss of PBP 2 with incubation with PMNs was demonstrated (data not shown). The digoxigenin-beta-lactam detection method also demonstrated clear bands for PBPs 1, 2, 3, and 4, the first two of which could be blocked by preincubation with cold ampicillin. Again, the loss of PBP 2 with incubation with PMNs was demonstrated (data not shown). When neutrophils alone were subjected to the PBP isolation procedure with a beta-lactam, incubation and blotting demonstrated a lack of important beta-lactam binding. This suggests that the densities of the PBPs in our PMN-exposed S. aureus isolate were unlikely to be enhanced by nonspecific beta-lactam binding by the PMN protein. FIG. 1. | Rat tissue cage model. Rat tissue cage model. A gas-sterilized table-tennis ball into which holes were drilled was implanted by sterile procedures intra-abdominally in large rats (one per rat). The balls were well tolerated when they were infected following 5 weeks of encapsulation. The model was developed so that reagents that were unavailable for the rabbit model could be used. The photograph was made at the time of necropsy of a rat with an 8-day-old S. aureus infection. FIG. 2. | Circulating blood PMN counts (A) and abscess fluid PMN counts (B) during treatment with cefazolin in rats infected with S. aureus Circulating blood PMN counts (A) and abscess fluid PMN counts (B) during treatment with cefazolin in rats infected with S. aureus. Anti-PMN serum of rabbit origin (Accurate Chemical Corp.) was administered 1 day before infection and on days 2, 4, and 6 of infection (n = 17). Total cell counts and differential counts were obtained for cytospin samples made with abscess fluid sampled on days 2, 4, and 7 of cefazolin treatment. The sera of rats treated with anti-PMN serum demonstrated significant drops in PMN cell counts on all days on which the counts were measured compared to the counts in the sera of rats treated with control serum (n = 18). P values were determined by the Mann-Whitney U test. , control treatment; , treatment with anti-PMN serum; bars, maximum and minimum values; the symbols themselves delineate the values for 25 to 75% of the samples; , median values. FIG. 3. | S. S. aureus counts in abscesses. Abscess fluid was sampled on days 2, 4, and 7 of cefazolin treatment. The bacterial count at day 7 was significantly lower in the anti-PMN-treated group (n = 17) than in the control group (n = 18) (P = 0.043 by the Mann-Whitney U test). See the legend to Fig. for definitions of the symbols. FIG. 4. | Effect of incubation with PMNs on the S. aureus Effect of incubation with PMNs on the S. aureus PBPs. Density values (means +- standard deviations for six runs) for PBPs isolated from bacteria incubated with PMNs or not incubated with PMNs were determined by differential centrifugation and identification with biotinylated ampicillin. PBPs were separated electrophoretically, blotted, and stained. Important PBPs were identified by molecular weight on the basis of two standards included in each run and were quantitated with imaging software (ImageQuaNT). After exposure to PMNs, a significant reduction in the level of PBP 2 expression was seen, as defined by a decrease in gel band density with equal amounts of protein (P < 0.01 by Student's t test). The levels of PBP 1 and PBP 3 expression were not significantly changed. The level of PBP 4 expression showed a significant increase in the biotinylated ampicillin series that was not seen by the methods that used fluorescent penicillin. FIG. 5. | ImageQuaNT scans of S. aureus ImageQuaNT scans of S. aureus PBPs (the numbers above each peak). A representative run of viable S. aureus cells incubated or not incubated with viable neutrophils for 30 min is shown. Cell wall proteins were subsequently isolated; incubated with biotinylated ampicillin; and separated on a gel, blotted, and stained. The third scan at the bottom is a methods control: S. aureus membrane proteins were incubated with unlabeled ampicillin in excess for 30 min at 37C before incubation with biotinylated ampicillin for 30 min at 37C, followed by performance of the standard run. FIG. 6. | Photo from the density program of S. Photo from the density program of S. aureus PBPs incubated without or with neutrophils before cell wall isolation and electrophoresis. The cell wall proteins were subsequently isolated, incubated with biotinylated ampicillin, and separated electrophoretically and were blotted and stained with streptavidin-alkaline phosphatase and then the BCIP-NBT substrate. The photo is from the software program used to determine band density. PBPs are indicated to the right of the gel. DISCUSSION : The experiments described here have demonstrated that in vivo depletion of neutrophils allows more efficacious killing of S. aureus by a beta-lactam in a host's established infection and are novel in terms of the model used and the level of quantification. Others have previously considered our suggestion regarding the mechanism of the changes in bacterial cell wall enzymes during phagocytosis by PMNs. Rakita et al. have previously suggested that neutrophils may inactivate the PBPs of bacteria. They found that the levels of seven E. coli PBPs declined over 30 min of incubation with neutrophils and that the decline in the PBP levels correlated with a loss of viability in E. coli. Vriesema et al. showed the potential for genetic changes in S. aureus internalized in cells. Using a transposon-based expression library of S. aureus, they reported that bacterial genes were upregulated by exposure to endothelial cells; among these were genes encoding enzymes leading to cell wall biosynthesis. Our results with S. aureus furnish no genetic data but demonstrate the decrease in the density of PBP 2 after incubation with PMNs. The density of a peak corresponding to PBP 4 increased with incubation with PMNs. However, control experiments with another beta-lactam label suggested that the increased density of a peak corresponding to PBP 4 might have occurred only with biotinylated ampicillin in our assays, so the change in the level of expression of that PBP remains undefined. The change in the level of PBP 2 expression was not associated with an identifiable loss in S. aureus viability, as determined by dilution plating (in ice water) of the PMN-exposed and control bacteria at the end of the incubation. There is some evidence that alterations in PBP 2 could affect the activity of cefazolin. The 50% inhibitory doses of cefazolin for PBP 2 and PBP 4 from susceptible staphylococci were 0.078 and >30 mug/ml, respectively . By using those published values, a decrease in the level of PBP 2 expression may adversely affect the activity of cefazolin. We considered the additional possibility that the PBP modifications that we saw could be artifacts. Since PBPs are enzymes, they are heat and pH sensitive, and perhaps these factors in vitro changed certain PBPs rather than the neutrophils. We consider this unlikely, because buffered HBSS was added to both PMN and control bacteria, and PMN-treated and control bacteria were incubated in an identical fashion. We also considered the possibility that our cell wall preparations included neutrophilic protein only in the preparations consisting of exposed S. aureus cells and not in the control preparations and that such proteins produce nonspecific binding that augments the intensities of some of the PBPs. We found it difficult to control for the phagocytosis of S. aureus by the added neutrophils, as it was very rapid. Subsequently, we tested the entire S. aureus cell wall isolation method using only fresh neutrophils (no bacteria). The PMN protein produced no PBP bands when it was incubated with a labeled beta-lactam and electrophoretically separated, blotted, and stained. Furthermore, if the PMN protein in conjunction with S. aureus PBPs binds to beta-lactam labels nonspecifically, the observed decrease in the level of PBP 2 expression should not have occurred with the S. aureus cells incubated with PMNs. We used the method of Dargis and Malouin for biotinylation of ampicillin for detection of PBPs. The rapid results (compared to the time to retrieval of results by use of isotope labels) are a great benefit from the use of colorimetric labels. A disadvantage, as reported previously (8), are the occasional nonspecific bands which are not consistent with the molecular weights of S. aureus PBP 1, 2, 3, or 4. As controls for the possible nonspecific binding of endogenous bacterial streptavidin to the biotinylated label, we demonstrated that binding of the biotinylated ampicillin label (to high-molecular-weight PBPs) could be blocked by preincubation with cold ampicillin. We subsequently successfully used other labels that did not involve biotin to detect S. aureus PBPs. A few additional beta-lactam binding protein bands, in addition to the PBPs, were occasionally seen even with the fluorescent penicillin label. It is possible that these bands could be isolation artifacts: breakdown products of the cell walls of S. aureus with proteins with low-affinity binding that are visible by the brief (minutes) development procedures used in this study but not expressed over the long development time (weeks) required for isotope labels. Further work on PBP labeling techniques should be performed, but such work was beyond the scope of this study. In previous experiments with a rabbit S. aureus tissue cage abscess model similar to the model used in the present study, we have found that circulating neutrophils from hosts with 14-day-old abscesses had diminished bactericidal capacities and a diminished ability to generate superoxide compared to those for 1-day-old infections . Phagocytosis was relatively well preserved. Additionally, in in vitro assays, abscess fluid further inhibited the bactericidal capacity of normal neutrophils . In a report of a study by Gresham et al. , who used a staphylococcal peritonitis mouse model, the survival of S. aureus within neutrophils was vital to the pathogenesis of the infection, and intracellular survival was found to be regulated by sar, a global regulator that influences the intracellular locale of phagocytized bacteria. Limiting the influx of neutrophils enhanced survival and reduced the bacterial burden after 24 h of infection by decreasing the numbers of intracellular bacteria. There are numerous reports of studies that have evaluated the intracellular concentrations and activities of antimicrobials . Ratios of concentrations in PMNs to extracellular concentrations of up to 0.60 have been reported for beta-lactams, and these would likely provide intracellular concentrations well above the minimal bactericidal concentration for typical staphylococci. Hand and King-Thompson reported an intracellular concentration-to-extracellular concentration ratio of 0.45 for neutrophils that had phagocytized S. aureus but found that the intracellular penicillin did not enhance the killing of phagocytized S. aureus. In their study of penicillin, clindamycin, rifampin, gentamicin, and erythromycin, there was a poor correlation between the intracellular penetration of antimicrobials and intracellular killing activity. These data suggest that the poor activities of beta-lactams against intra-PMN staphylococci may not be secondary to the presence of an inadequate concentration of drug but is due to inadequate activity in the intracellular milieu. We suggest that our measurements for S. aureus-infected, neutropenic rats described in this report have demonstrated that cefazolin kills S. aureus cells in animals with chronic infection more efficiently when neutrophils are depleted. We postulate that the inactivation or diminished expression of PBP 2 after incubation with PMNs ex vivo describes what is happening in our animal model and results in (i) good phagocytosis by PMNs, (ii) a poor bactericidal capacity of PMNs, and (iii) viable but changed S. aureus cells with diminished targets for the cefazolin. Our hypothesis is similar to that of Stevens et al. , who postulated that the diminished expression of PBPs by stationary-phase group A streptococci might explain the diminished efficacies of beta-lactams in the treatment of established group A streptococcal myositis. Increased levels of production of PBP 4 have also been correlated with in vitro resistance to beta-lactams , and the possibility of increased levels of production of PBP 4 after incubation with PMNs deserves further study. In summary, we have found that the partial depletion of neutrophils from our S. aureus abscess model results in enhanced bacterial killing over that for the controls in rats treated with cefazolin. As there were no significant differences between the control and the neutropenic groups in terms of the quantities of drug present in serum or abscess fluid, the data are consistent with the premise that the activity of cefazolin is greater when the host PMN count is lower. We believe that neutrophils, although important in the initial control of a staphylococcal infection, in the milieu of an abscess may engulf staphylococci without effectively killing them. S. aureus cells that have lived and survived in the intra-PMN environment may possess modified target proteins for beta-lactam antimicrobials, a phenomenon that we have detailed ex vivo. While offering a plausible explanation for the diminished efficacies of beta-lactams in an abscess milieu and the need for drainage of purulent infections, such a concept would raise new questions on the role of chronic infections in the genesis of organisms with antimicrobial resistance. Backmatter: PMID- 12183271 TI - Phenylpropenamide Derivatives AT-61 and AT-130 Inhibit Replication of Wild-Type and Lamivudine-Resistant Strains of Hepatitis B Virus In Vitro AB - The phenylpropenamide derivatives AT-61 and AT-130 are nonnucleoside analogue inhibitors of hepatitis B virus (HBV) replication. They inhibited the replication of wild-type HBV with 50% inhibitory concentrations of 21.2 +- 9.5 and 2.40 +- 0.92 muM, respectively, compared to 0.064 +- 0.020 muM lamivudine. There were no significant differences in sensitivity between wild-type and nucleoside analogue-resistant (rtL180M, rtM204I, and rtL180M + rtM204V) HBV. Keywords: Introduction : Chronic hepatitis B virus (HBV) infection, a major cause of hepatocellular carcinoma and liver cirrhosis, affects about 5% of the world's human population . Chronic HBV infection is unresponsive to currently available vaccines and can only be controlled by chemotherapy . In most countries, only two drugs, alpha interferon (IFN-alpha), and lamivudine (LMV) [(-)-beta-L-2'-3'-dideoxy-3'-thiacytidine], are approved for treatment of chronic HBV , but additional drugs are being developed . IFN treatment results in a sustained response in only a minority of cases and may be associated with dose-limiting side effects . LMV, a deoxycytidine analog, is a safer and more effective inhibitor of HBV replication, but its effects are rarely sustained after short-term treatment, and its long-term use frequently results in the emergence of drug-resistant HBV strains . Specific mutations associated with resistance to LMV have been identified by sequence analysis and shown to affect the catalytic YMDD (tyrosine-methionine-aspartate-aspartate) motif of the viral polymerase . Because they are genotype dependent, the numbering systems used to identify these mutations are confusing; as a result, the adoption of a standardized genotype-independent nomenclature has recently been proposed and is used here. The LMV resistance mutations that have been observed most frequently in clinical isolates cause the methionine at amino acid residue 204 in the reverse transcriptase coding region of the polymerase protein (formerly amino acid 539, 549, 550, or 552, depending on the genotype) to be replaced with either isoleucine or valine. These changes are identified as rtM204I and rtM204V, respectively, by the new nomenclature. While the rtM204I substitution alone appears to be sufficient to confer resistance to LMV, rtM204V is rarely observed in the absence of a second mutation that causes an upstream substitution of methionine for leucine at reverse transcriptase residue 180 (rtL108M; formerly identified as polymerase residue 515, 525, 526, or 528) . Interestingly, the rtL180M change, by itself, has been associated with resistance to famciclovir, another nucleoside analogue that has undergone clinical trials against chronic HBV infection . Several independent studies have provided evidence that mutations that confer resistance to LMV and/or famciclovir also confer resistance to other nucleoside analogs (, , , , , ; reviewed in reference ). Cross-resistance may therefore limit the clinical potential of some nucleoside analogs that are being developed as potential anti-HBV agents . The identification of safe and efficacious nonnucleoside inhibitors of HBV replication would greatly improve prospects for overcoming the problems associated with nucleoside analog cross-resistance. Although several low-molecular-weight nonnucleoside anti-HBV compounds have been identified (reviewed in reference ), very few have been tested in controlled clinical trials. The phenylpropenamide derivative AT-61 was recently shown to be a potent inhibitor of in vitro HBV replication in stably or transiently transfected HepG2 cells . AT-61 specifically inhibited the in vitro replication of HBV but had no effect on the in vitro replication of a variety of other viruses, including duck HBV, woodchuck HBV, or human immunodeficiency virus type 1 (HIV-1) . Moreover, AT-61 was found to be equally active as an inhibitor of HBV replication in the HepG2-derived HepAD38 and HepAD79 cell lines . In these cell lines, the expression of stably transfected HBV genomes (wild type [wt] in HepAD38 and rtM204V in HepAD79, respectively) is controlled by a tetracycline-sensitive promoter . It was also reported that combinations of AT-61 and LMV produced synergistic antiviral effects against HBV replication in AD38 cells . These data suggest that the specific anti-HBV activity of AT-61 is due to a mechanism(s) that differs from that of LMV. Furthermore, they imply that mutations that confer resistance to LMV and other nucleoside analogs may not confer cross-resistance to AT-61 or other phenylpropenamide derivatives . FIG. 1. | Structures of AT-61 (a) and AT-130 (b). Structures of AT-61 (a) and AT-130 (b). Mol Wt., molecular weight. AT-130 , a congener of AT-61, differs from AT-61 in that rings A and B contain ortho-methoxy and para-nitroso substituents, respectively; in addition, a bromine atom replaces the chlorine present in AT-61. AT-130 has also been shown to have anti-HBV activity in HepAD38 cells in vitro . The present study sought to determine the relative in vitro potencies of AT-61 and AT-130 as inhibitors of wt and nucleoside analog-resistant HBV replication by using a recently described model that employs recombinant baculoviruses to efficiently transduce HepG2 cells with replication-competent HBV genomes . Recombinant baculoviruses that encode either 1.28-times genome length wt HBV (genotype A, subtype adw2) or one of three drug-resistant mutants were constructed and propagated as described previously . The three drug-resistant viruses harbored point mutations that coded for rtL180M, rtM204I, or rtL180M + rtM204V substitutions in the HBV polymerase sequence, respectively. AT-61 and AT-130 were synthesized as described previously , and stock solutions of each drug were freshly prepared in dimethyl sulfoxide. The final concentration of dimethyl sulfoxide in the cell culture media was always <1% (vol/vol), a concentration that has no measurable effect on HBV replication. The experimental procedures used for drug sensitivity testing were described in detail previously . Briefly, replicate cultures of HepG2 cells were transduced at a multiplicity of infection of 50 PFU/cell with either mutant or wt HBV and then exposed continuously to five different concentrations of AT-61, AT-130, or LMV, beginning immediately after transduction. Culture media were changed on days 2, 4, and 6 posttransduction, and cells were harvested on day 7. Replicating viral DNA was extracted from cytoplasmic core particles and analyzed by Southern hybridization and autoradiography . Image densities from suitably exposed autoradiographs of Southern blots were measured by computer-assisted densitometry (GC-67 scanning densitometer with Molecular Analyst software from Bio-Rad Laboratories, Hercules, Calif.). The amount of viral replication in drug-treated samples was expressed as a percentage of the amount of replication in drug-free controls. Where possible, logistic dose-response curves described by the equation y = a/1 + (x/b)-c were fitted to each set of data and equation parameters were estimated as described previously with the aid of TableCurve2D, a software package from Jandel Scientific (San Rafael, Calif.) . Drug concentrations that inhibited the replication of each mutant HBV by 50 or 90% of the average amount measured in the corresponding drug-free controls (IC50 and IC90, respectively) were estimated from individual dose-response curves. To assess the cytotoxicity of these compounds, additional sets of HepG2 cultures were continuously exposed to AT-61 or AT-130 for 7 days, after which cellular toxicity was assayed colorimetrically by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide cleavage. Cytotoxicity was also tested in Huh-7, another hepatic cell line of human origin. FIG. 2. | Dose-dependent inhibition of replication of wt and nucleoside-analog resistant HBV by phenylpropenamide derivatives. Dose-dependent inhibition of replication of wt and nucleoside-analog resistant HBV by phenylpropenamide derivatives. HepG2 cells were transduced with HBV baculovirus and then exposed to the indicated concentrations of AT-61 or AT-130 for 7 days, after which intracellular HBV replication was analyzed by Southern blotting (see text). Panels A and B show autoradiographs of Southern blots of intracellular replicative intermediates after treatment with AT-61 and AT-130, respectively. After equivalent exposure times, image densities in (drug-free) control lanes were such as to suggest that replication fitness decreased in the order wt > rtL180M > rtL180M + rtM204V > rtM204I, assuming that the transduction efficiencies were equal and that the polymerase changes did not differentially affect virus secretion. The autoradiographs shown were exposed for different times so that the image densities in the control lanes are approximately equal. Panels C and D show a graphical analysis of data derived from panels A and B, respectively. Symbols: *, wt; , rtM204I; , rtL180M; , rtL180M + rtM204V. Preliminary experiments confirmed that treatment with either AT-61 or AT-130 caused dose-dependent inhibition of wt HBV replication in HepG2 cells and that neither compound was measurably toxic to either HepG2 or Huh-7 cells at concentrations of up to 250 muM, the highest concentration tested (data not shown). Results from the recombinant baculovirus assay system produced estimated IC50s of 0.064 +- 0.020, 21.2 +- 9.5, and 2.4 +- 0.92 muM (mean +- standard deviation of triplicate assays against wt HBV) for LMV, AT-61, and AT-130, respectively. These IC50s are consistent with the potency ranking reported earlier but are greater than those reported previously, i.e., 0.6 to 5.7 muM for AT-61, depending on the cell line used for the assay , and 0.13 muM for AT-130 . Subsequent experiments compared the activities of AT-61 and AT-130 as inhibitors of both wt and mutant HBVs. The results, which are summarized in Table , indicate that the calculated IC50s and resistance factors for wt and mutant HBVs were not significantly different following either AT-61 or AT-130 treatment. They also confirmed that, on a molar basis, AT-130 was more effective as an inhibitor of in vitro HBV DNA replication than was AT-61. Calculated IC50s and IC90s were in the ranges of 19 to 27 and 63 to 97 muM, respectively, for AT-61. The corresponding ranges for AT-130 were 1.3 to 5.1 and 11 to 36 muM. To confirm that the phenylpropenamide derivatives were active against LMV-resistant HBV mutants and to confirm that these mutants showed a drug resistance phenotype in vitro, replicate assays were carried out in parallel with LMV as an internal control. The results presented in Table confirm that mutations that confer LMV resistance do not confer cross-resistance to the phenylpropenamide derivatives. TABLE 1 | Sensitivity of wt and drug-resistant HBV to phenylpropenamide derivatives AT-61 and AT-130 TABLE 2 | Relative sensitivities of wt and drug-resistant HBV Results presented here confirm and extend previously reported observations regarding the anti-HBV activities of AT-61 and related phenylpropenamide derivatives . Although the mechanism of action of this group of compounds is unknown, it appears to be independent of interference with the RNA- or DNA-dependent activities of the HBV polymerase. Observing that exposure to AT-61 decreased the amount of RNA-containing cytoplasmic HBV core particles, King and colleagues postulated that AT-61 inhibits the packaging of pregenomic viral RNA , a prerequisite for genome replication . Our observation that the HBV mutants most commonly associated with LMV and/or famciclovir resistance remain sensitive to both AT-61 and AT-130 justifies the further development of these compounds or their derivatives for eventual clinical use. These compounds appear to have the potential to arrest the replication of known drug-resistant HBV strains and thus, in combination with nucleoside/nucleotide analogs and/or other drugs, may help reduce, or perhaps even prevent, the development of drug resistance . Backmatter: PMID- 12183268 TI - beta-Lactamases of Kluyvera ascorbata, Probable Progenitors of Some Plasmid-Encoded CTX-M Types AB - Kluyvera ascorbata produces a beta-lactamase that results in an atypical susceptibility pattern, including low-level resistance to penicillins, cephalothin, and cefuroxime, but this resistance is reversed by clavulanate. Ten nucleotide sequences of the corresponding gene, blaKLUA, were obtained and were found to have minor variations (96 to 100%). Otherwise, blaKLUA was found to be similar (95 to 100%) to some plasmid-encoded CTX-M-type beta-lactamases. Finally, mobilization of blaKLUA on a plasmid was found to be mediated probably by a genetic mobile element like ISEcp1. Keywords: Introduction : Resistance to oxyimino-beta-lactams such as cefotaxime, ceftazidime, or ceftriaxone in gram-negative rods has commonly been associated with the expression of acquired beta-lactamases. Most of them are class A extended-spectrum beta-lactamase (ESBL) derivatives of the TEM-1, TEM-2, or SHV-1 enzyme . Since the 1990s, a new step in resistance to beta-lactams was discovered with the demonstration of the mobilization of a chromosomal beta-lactamase on a plasmid, allowing the wide dissemination of beta-lactamases and conferring the potential for epidemic problems. Many plasmid-mediated AmpC beta-lactamases have been described, such as MIR-1 (Enterobacter cloacae), CMY-2 (Citrobacter freundii), DHA-1 (Morganella morganii), and ACC-1 (Hafnia alvei) . More recently, the discovery of SFO-1 suggested a probable mobilization from the chromosome of Serratia fonticola on a plasmid harbored in an E. cloacae isolate in Japan . To date, that was the only report of the mobilization of a chromosomally mediated class A beta-lactamase. The CTX-M-type beta-lactamases (CTX-M-1 to CTX-M-15, Toho-1 and Toho-2, and UOE-1 and UOE-2), encoded by transferable plasmids, constitute a novel group of class A ESBLs whose origins are still unknown . These CTX-M-type enzymes are not closely related to TEM or SHV ESBLs but share extensive similarities (70 to 75%) with the chromosomal Klebsiella oxytoca beta-lactamases . Nevertheless, Kluyvera ascorbata, a species of the family Enterobacteriaceae that is rarely detected in medical practice , was suggested as another probable progenitor because of the particular susceptibility patterns of some clinical isolates including resistance to cefotaxime and aztreonam but susceptibility to ceftazidime and the high potentiation effect of the reversal of resistance by clavulanate (D. Bertei et al., Abstr. 14th Reunion Interdisciplin. Chimiother. Anti-Infect., abstr. 199, p. 10, 1994). Bacterial strains. : Ten nonduplicate strains of K. ascorbata including the type strain were supplied by the Institut Pasteur Collection (Paris, France), and two strains were isolated at Tenon and Saint-Louis Hospitals (Paris, France), respectively. Each strain was identified on the basis of its behavior in culture and its biochemical characteristics by using the API 20E system (bioMerieux, Marcy-l'Etoile, France) and by using Biotype-100 carbon source utilization strips (bioMerieux) and Recognizer software (Taxolab, Institut Pasteur, Paris, France). Antimicrobial agents and susceptibility testing. : The patterns of susceptibility to antibiotics usually active against members of the family Enterobacteriaceae were determined by the disk diffusion method, as described previously . The MICs of 12 beta-lactams including penicillins and cephalosporins in the presence or absence of beta-lactamase inhibitors (clavulanate at 2 mug/ml or tazobactam at 4 mug/ml) were determined by a dilution method on Mueller-Hinton agar (Sanofi Diagnostics Pasteur, Marnes La Coquette, France) with an inoculum of 105 CFU/spot applied with a multi-inoculation device (Multipoint Inoculator A400; Denleytech, Woking, United Kingdom) . All strains tested a low level of resistance to penicillins (amoxicillin, ticarcillin, piperacillin) but also displayed a low level of resistance to cephalothin and cefuroxime . Significant synergy with clavulanate was observed for amoxicillin and ticarcillin (32- to 512-fold). A low level of synergy was also observed between piperacillin and tazobactam, which correlated with the high level of activity of the ureidopenicillin piperacillin. Finally, all strains tested were highly susceptible to extended-spectrum cephalosporins, moxalactam, cefoxitin, aztreonam, and imipenem. TABLE 1 | MICs of 12 beta-lactams for the 12 clinical K. ascorbata isolates, recipient strain E. coli XL-1, and the E. coli XL-1 transformant producing KLUA-1 Kinetic and IEF analyses. : Subsequently, the resistance pattern was correlated with beta-lactamase production. The beta-lactamase contents were obtained from sonicated extracts of bacteria cultured overnight at 37C in Trypticase soy broth (bioMerieux), prepared as described previously . Analytical isoelectric focusing (IEF) was performed in a polyacrylamide gel as described previously by the iodiometric procedure with benzylpenicillin for detection . We detected beta-lactamases with different pIs, as follows: pI 6.9, two strains; pI 8.0, six strains; pI 8.4, four strains (see Table ). Finally, the substrate and inhibition profiles of a single crude extract (type strain CIP 82.95T) were determined. This extract was produced from 4 liters of culture grown in brain heart infusion (Difco, Detroit, Mich.) and incubated at 37C for 5 h after initial inoculation of 300 ml of a starter culture. Cells were harvested by centrifugation at 5,800 x g for 30 min at 4C. The pellets (about 20 g) were washed by resuspension in 40 ml of ice-cold 0.1 M NaCl plus 0.05 M sodium azide and centrifuged as described above. The washed pellet was resuspended in 40 ml of the same solution and lysed by ultrasound treatment (three times for 10 s each time per gram of pellets; power = 50 W and frequency = 20 kHz with type 20-200 equipment [Alcatel, Paris, France]). The crude extract was cleared by centrifugation at 48,000 x g for 30 min at 4C. Nucleic acids were precipitated by adding 0.2 M spermine (Sigma, Saint-Quentin Fallavier, France) and centrifugation (48,000 x g for 30 min at 4C). The crude extract was then subjected to chromatography on a Sephadex G100 column. The active fractions were pooled, subjected to chromatography on Bio-Rex 70 resin, and then concentrated by ultrafiltration. The kinetic constants Vmax, which was expressed relative to the Vmax for benzylpenicillin (which was set at 100%), and Km were determined by computerized microacidimetric assay in 0.1 M NaCl at pH 7.0 and 37C . One unit of beta-lactamase was defined as the amount of enzyme that hydrolyzed 1 mumol of benzylpenicillin per min at pH 7.0 and 37C. Kinetic analysis showed that the activity of the enzyme was highest with cephalothin, but that cefuroxime and cefotaxime were also good substrates . Conversely, the enzyme displayed a low level of activity with ceftazidime, aztreonam, and cephamycin. Tazobactam was the most efficient inhibitor (50% inhibitory concentration [IC50], 10 nM), closely followed by clavulanic acid (IC50, 20 nM). Sulbactam was the least potent inhibitor (IC50, 100 nM). Finally, chloride ions at a concentration of 0.5 M had no inhibitory effect. TABLE 3 | Genotypic characterization of beta-lactamase contents of K. ascorbata strains and beta-lactamase pIs TABLE 2 | Kinetic constants (Vmax, Km) of beta-lactamase produced by K. ascorbata strain CIP 82. 95T Genetic characterization of blaKLUA. : The identification of only one susceptibility pattern among the 12 strains of K. ascorbata tested suggested that this species possesses a naturally occurring, chromosomally mediated beta-lactamase. Furthermore, this assumption was confirmed by the failure of both conjugation experiments and plasmid extraction with the High Pure plasmid isolation kit (Roche Biochemicals, Neuilly-sur-Seine, France). On the basis of the observed pattern of susceptibility and enzymatic properties, we designed degenerate oligonucleotide primers from the sequences of genes encoding CTX-M-type enzymes (primer MA1-forward [5'-SCSATGTGCAGYACCAGTAA-3'] and primer MA2-reverse [5'-CCGCRATATGRTTGGTGGTG-3']) and used them to amplify an internal fragment of the K. ascorbata bla genes. In each PCR, the cycling conditions were as follows: 35 cycles of 60 s at 94C for denaturation, 30 s at 60C for annealing, and 30 s at 72C for elongation with 2 U of Taq DNA polymerase (Perkin-Elmer, Foster City, Calif.). Amplifications were obtained for all K. ascorbata strains producing a beta-lactamase of pI 8.0 or 8.4. Nevertheless, the two bla genes encoding beta-lactamase of pI 6.9 were not amplified with these CTX-M-specific primers . The PCR products were then purified with Qiagen (Courtaboeuf, France) columns, and the DNA sequences were determined twice on both strands by the procedure of Sanger et al. by using fluorescent dye-labeled dideoxynucleotides, thermal cycling with Taq polymerase, and an ABI 373A DNA sequencer (Applied Biosystems, Foster City, Calif.). All PCR products consisted of a 540-bp fragment, which was found to be highly similar (>97% identity) to the cluster that included the blaCTX-M-2, blaCTX-M-4, blaCTX-M-5, blaCTX-M-6, blaCTX-M-7, and blaToho-1 genes. Finally, other primers more specific for the blaCTX-M gene of this cluster were used to amplify the entire Kluyvera bla gene: primers M3-forward (5'-ATGATGACTCAGAGCATTCGC-3') and M3-reverse (5'-GGGCAATCAGCTTATTCATGG-3') and primers M4-forward (5'-TTGCTCGCTCGTTGGGTGAT-3') and M4-reverse (5'-TATTGCATCAGAAACCGTGGG-3') . Amplifications with primer M3 or M4 were performed by using the same cycling conditions described above. The nucleotide sequences of K. ascorbata genes encoding a beta-lactamase with a pI of 8.0 or 8.4 (blaKLUA-1 to blaKLUA-5 and blaKLUA-8 to blaKLUA-12) were determined . High degrees of homology (>95%) were observed among all these blaKLUA genes coding for class A beta-lactamases. Moreover, blaKLUA genes were also found to share extensive similarities (95 to 100%) with the genes producing the CTX-M-2, CTX-M-4, CTX-M-5, CTX-M-6, CTX-M-7, and Toho-1 beta-lactamases; but they displayed lower levels of identity with the genes for the other CTX-M types (68 to 84%) . However, this high level of similarity observed suggested that the chromosome-encoded beta-lactamase of K. ascorbata is a more probable progenitor of at least some plasmid-encoded CTX-M enzymes than K. oxytoca . Finally, the highest degrees of homology between blaKLUA and chromosome-encoded bla genes were observed for class A enzymes produced by Kluyvera cryocrescens (~77%), S. fonticola (~74%), K. oxytoca (~73%), and Citrobacter sedlakii (~73%) (, , ; C. Humeniuk, unpublished data). Only two bla genes encoding a beta-lactamase of pI 6.9 were not amplified with the CTX-M-specific primers . Characterization of a beta-lactamase(s) produced by both of these strains is in progress. In fact, it appears that K. ascorbata could produce at least two different kinds of beta-lactamase genes, as has been observed for K. oxytoca, in which two main groups of beta-lactamase genes have been distinguished (blaOXY-1 and blaOXY-2; identity, 90%) . This heterogeneity of the K. ascorbata beta-lactamases could be correlated with the diversity observed among CTX-M-type enzymes, which are divided into five clusters . TABLE 4 | Identities of amino acid sequences of K. ascorbata beta-lactamases and CTX-M, Toho, UOE beta-lactamases Genetic environment of blaKLUA-1. : As the chromosomal beta-lactamase genes of K. ascorbata are the probable progenitors of some of the plasmid-encoded extended-spectrum CTX-M-type enzymes, we investigated the genetic organization of blaKLUA-1 by cloning and sequencing the surrounding regions of this gene. Chromosomal DNA was prepared from K. ascorbata CIP 82.95T, partially digested with Sau3A, and ligated into the BamHI site of pBK-CMV (pBK Phagemid Vectors; Stratagene, La Jolla, Calif.), as described previously . The recombinant plasmid was introduced into Escherichia coli XL-1 by the standard CaCl2 procedure. Several transformants were selected on Mueller-Hinton agar (Sanofi Diagnostics Pasteur) supplemented with amoxicillin (40 mug ml-1) and kanamycin (25 mug ml-1) and were further characterized by determination of their antibiotic susceptibility profiles and pIs. The molecular sizes of the inserts were estimated from the results of restriction digestion and electrophoresis in 1 to 3% agarose gels, as described previously . Finally, both strands of the DNA sequence of a 7-kb insert were determined. The BLASTN program on the National Center for Biotechnology Information website (revision date, 20 July 2001; ) was used for database searches . Three open reading frames (ORFs) were found immediately upstream (2,980 bp) and downstream (3,300 bp) of the blaKLUA gene (ORF 2) . The protein corresponding to ORF 1 was 86% identical to an aspartate aminotransferase from E. coli. The deduced amino acid sequence of the protein encoded by ORF 3 was 43% identical to that of a putative protein from E. coli. The protein encoded by the fourth ORF showed various degrees of similarity (28 to 56%) to adhesive proteins (MisL, AidaI, VirG) known to contribute to the pathogenicities of several enterobacteria. Recently, the nucleotide sequence of blaKLUC, which encodes the chromosomal beta-lactamase of K. cryocrescens, and its surrounding regions has been reported . Genes corresponding to an aspartate aminotransferase and a putative protein showing 94 and 66% identities with the proteins encoding by K. ascorbata ORF 1 and ORF 3, respectively, were found upstream and downstream of the blaKLUC gene. In fact, this particular genetic organization could be a characteristic of the genus Kluyvera. Finally, no ampR gene was detected upstream of the blaKLUA-1 gene . Indeed, some class A beta-lactamases closely related to K. ascorbata enzymes are regulated by an ampR gene: cumR (Proteus vulgaris) , sedR (C. sedlakii), and fonR (S. fonticola) (C. Humeniuk, unpublished data). FIG. 1. | Genetic environment of blaKLUA-1 gene encoding the beta-lactamase produced by K. ascorbata Genetic environment of blaKLUA-1 gene encoding the beta-lactamase produced by K. ascorbata strain CIP 82.95T. The regions surrounding the chromosomal blaKLUA-1 gene and the sequences adjacent to the plasmidic genes encoding beta-lactamases CTX-M-2, CTX-M-4, CTX-M-5, CTX-M-6, CTX-M-7, and Toho-1 were compared. The RTR sequence and tnpA correspond to the right terminal repeat and transposase A, respectively, which are typical of ISEcp1. Arguments for mobilization of chromosomal blaKLUA gene. : The analysis of the sequences surrounding the blaCTX-M genes revealed the presence of sequences similar to those for CTX-M-5 on the K. ascorbata chromosome (5 nucleotides [nt] upstream and 179 nt downstream with 80 and 100% identities, respectively, with the chromosome of K. ascorbata), Toho-1 (49 nt upstream and 71 nt downstream with 95.9 and 98.6% identities, respectively), and CTX-M-2, -M-4, -M-6, and -M-7 (29 nt upstream with 100% identity) . However, major differences between the adjacent sequences of the other blaCTX-M genes (CTX-M-1, -M-3, -M-8, -M-12, -M-13, -M-14, and -M-15; Toho-2; UOE-1; and UOE-2) and the chromosome of K. ascorbata were observed. This fact confirms that the beta-lactamase of K. ascorbata is probably the progenitor of only some plasmid-encoded CTX-M-type enzymes. To date, this is the second report of the mobilization of a chromosomally mediated class A beta-lactamase. So far we have investigated the potential mechanism of mobilization of the blaKLUA gene by sequencing surrounding regions of the blaCTX-M-5 gene located on plasmid pCLL3417 . The GenBank accession number for this nucleotide sequence is . Sequencing confirmed the presence of sequences similar to that of the K. ascorbata chromosome (19 nt upstream and 256 nt downstream with 85 and 98.6% identities, respectively) and revealed the presence of the novel insertion sequence, ISEcp1, described by Stapleton (P. D. Stapleton, Abstr. 39th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 1457, 1999) . Subsequently, this element was reported for several plasmid-encoded AmpC enzymes including CMY-5, LAT-1, and ACC-1, which were probably mobilized from chromosomal beta-lactamases of C. freundii or H. alvei . It was also detected on plasmids carrying other blaCTX-M genes (CTX-M-3-like, CTX-M-14, CTX-M-15, Toho-1, Toho-2, and UOE-2) (, , ; GenBank accession numbers , , and ). Given the low level of clinical importance of K. ascorbata , it is important to consider the possible transfer of the bla gene in veterinary practice or in the environment. Several preliminary reports have indeed described Salmonella enterica serovar Typhimurium as producing beta-lactamases of the CTX-M2, CTX-M-4, CTX-M-5, and CTX-M-7 types in various countries (Argentina, Latvia, Russia, Greece) . If K. ascorbata acts as a reservoir for these plasmid-encoded CTX-M enzymes, the sequences of more chromosomal genes must be determined to confirm the existence of a single progenitor due to the heterogeneity observed in the CTX-M cluster . Nucleotide sequence accession number. : The GenBank accession number for the nucleotide sequence of the insert reported in this paper is . Backmatter: PMID- 12183256 TI - Characterization of a Self-Transferable Plasmid from Salmonella enterica Serotype Typhimurium Clinical Isolates Carrying Two Integron-Borne Gene Cassettes Together with Virulence and Drug Resistance Genes AB - An unusual self-transferable virulence-resistance plasmid (pUO-StVR2) was found in nine multidrug-resistant (ACSSuT phenotype) Salmonella enterica serotype Typhimurium clinical isolates that were assigned to four different phage types and a single and distinctive XbaI pulsed-field gel electrophoresis profile. pUO-StVR2 is an IncFII plasmid of about 140 kb in length carrying the spvA, spvB, and spvC (Salmonella plasmid virulence) and rck (resistance to complement killing) genes. It also carries the oxa1/aadA1a (ampicillin resistance and streptomycin-spectinomycin resistance) gene cassette configuration located within a class 1 integron with qacEDelta1/sul1 (ammonium antiseptics resistance and sulfadiazine resistance); the transposon genes merA, tnpA, and tnpR (mercury resistance, transposase, and resolvase of Tn21, respectively); and the catA1 (chloramphenicol resistance) and tet(B) (tetracycline resistance) genes. The insertion of resistance genes into a Salmonella virulence plasmid constitutes a new and interesting example of plasmid evolution and presents a serious public health problem. Keywords: Introduction : Several Salmonellaenterica serotypes harbor plasmids which are essential for virulence. Although virulence plasmids from different serotypes are not identical, the sequences required for virulence are highly conserved . Most of the Salmonellaenterica serotype Typhimurium isolates carry a serovar-specific 90-kb virulence plasmid which belongs to the incompatibility group IncFII. However, irrespective of their plasmid content, serotype Typhimurium isolates have been assigned to different clones, which are usually defined by phage typing and DNA fingerprinting . A particular serotype Typhimurium clonal group which includes phage type 104 (DT104) strains that are, at the least, resistant to ampicillin, chloramphenicol-florfenicol, streptomycin-spectinomycin, sulfonamides, and tetracycline (ACSSuT phenotype) has emerged worldwide and presents a global health problem . Recently, the emergence of other multidrug-resistant serotype Typhimurium clones in which the resistance is plasmid mediated has been reported . In these clones, the resistance genes are often part of transposons and/or integrons, elements that facilitate the intracellular movement of resistance genes. Among the different classes of integrons described, class 1 integrons are the most frequently found in Salmonella as well as in other gram-negative bacteria and they can be found within transposons and be chromosome and/or plasmid located (-, , -, , -, ). In the present work, we report an unusual self-transferable virulence plasmid responsible for multidrug resistance (ACSSuT) in serotype Typhimurium clinical isolates of various phage types sharing a common XbaI pulsed-field gel electrophoresis (PFGE) profile. MATERIALS AND METHODS : Bacterial strains. | In previous works (reference and unpublished data), 83 Salmonellaenterica serotype Typhimurium isolates carrying class 1 integrons were identified. Nine of these isolates were selected for the study presented here. They originated from human feces and were collected at regional hospitals. All were received throughout 1993 to 1999 at the Laboratorio de Salud Publica (LSP) of the Principality of Asturias. The isolates were phage typed in the Centro Nacional de Microbiologia, Instituto de Salud Carlos III, Madrid, Spain, by using Anderson's scheme . As controls in different experiments, the following serotype Typhimurium strains carrying the 90-kb virulence plasmid were tested: LT2 (pSLT90) (Bayer AG Collection), ATCC 14028 (pSLT90) (American Type Culture Collection), and LSP14/92 (pUO-StV1), which is a DT104 (subtype l)-ACSSuT clinical isolate (-). For mating experiments, serotype Typhimurium LSP31/93 was used as plasmid donor and the rifampin-resistant Escherichia coli K-12 strain J53 and nalidixic acid-resistant Salmonellaenterica serotype Panama LSP291/98 were used as recipients. These strains were tested for susceptibility to 15 antimicrobial agents by disk diffusion and to florfenicol by broth microdilution by using a Sensititre semiautomatic system (Trek Diagnostics, East Grinstead, United Kingdom). Results were analyzed according to standards set by NCCLS . Plasmid and PFGE analysis. | Plasmid profiling, plasmid curing, and transfer of antibiotic resistance by conjugation were performed as described in references and . Plasmid relationships were tested by restriction analysis using 5 U of HindIII, ClaI, and EcoRI (Amersham Pharmacia Biotech, Barcelona, Spain) separately and by hybridization using gene-specific probes . PFGE analysis was performed with XbaI as described in reference . PCR amplification, purification, and sequencing of DNA. | The detection of Salmonella virulence plasmid genes, resistance genes, and transposon- and/or integron-related genes was performed with the PCR conditions and primers described previously (, -, ) or designed for this work: repFIIA-F/B, CTGTCGTAAGCTGATGGC/CTCTGCCACAAACTTCAGC; traT-F/B, GATGGTTACACTGGTCAG/TCTGAGATCTGTACGTCG; rck-F/B, TCGTTCTGTCCTCACTGC/TCATAGCCCAGATCGATG; and pefA-F/B, GCACACGCTGCCAATGAA/CACAGACTTGAAGTCACC (accession number ); and [tnpR-F/B], [GGCGACACCGTGGTGGTGCATAGC/CGGTAAGCCCCGCGTTGCTTGGC] (accession number ). Restriction and sequencing of the PCR products were carried out as described in references and , and the sequences obtained were compared to those registered in GenBank. RESULTS AND DISCUSSION : Characterization of a self-transferable virulence-resistance plasmid of Salmonellaenterica serotype Typhimurium. | In previous works (reference and unpublished data), nine serotype Typhimurium isolates were identified which carried a class 1 qacEDelta1/sul1 (encoding ammonium antiseptics resistance and sulfadiazine resistance) integron, generating PCR products of 2,000 bp with 5' CS and 3' CS primers. By using PvuI, TaqI, and BglII endonucleases, the 2,000-bp PCR products yielded the same restriction pattern as the one described previously for an integron-borne oxa1/aadA1a (encoding ampicillin resistance and streptomycin-spectinomycin resistance) cassette array. PCR amplification confirmed the presence of both oxa1 and aadA1a within the 2,000-bp amplicons. The oxa1/aadA1a cassette configuration is characteristic for the transposon Tn2603 . The presence of transposon genes tnpA, tnpR, and merA (encoding the transposase, resolvase, and mercury reductase, respectively) was confirmed by PCR amplification (amplicons of about 327, 240, and 1,232 bp, respectively, were generated) in the nine isolates, but long PCR amplification (data not shown) failed to confirm a link between transposon and integron genes. On the other hand, amplification and sequencing also revealed the presence of the catA1 (encoding chloramphenicol resistance) and tet(B) (encoding tetracycline resistance) genes (accession numbers and , respectively) in the nine isolates. Plasmid analysis showed that the nine serotype Typhimurium isolates carried a large plasmid of about 140 kb together with other small plasmids but not the serotype Typhimurium 90-kb virulence plasmid . However, when these isolates were tested by PCR for virulence plasmid genes (repFIIA [plasmid incompatibility group FII replicons]; traT [conjugative transfer]; spvA, spvB, and spvC [Salmonella plasmid virulence]; rck [resistance to complement killing]; and pefA [plasmid-encoded fimbria]), the expected amplicons (about 288, 483, 600, 1,060, 424, 474, and 442 bp, respectively) were generated with all except the pefA primers. To ascertain if the virulence and resistance genes were linked and whether they were located in the plasmid or chromosome, the isolate LSP31/93 was tested by curing and conjugation experiments. By curing, no segregants were found among the approximately 300 CFU tested. Moreover, the 140-kb plasmid was self-transferable to E. coli K-12 J53 and more effectively transferable to Panama LSP291/98. The transconjugants expressed the same resistance phenotype (ACSSuT) as the parental strain, and their plasmid DNAs generated the expected PCR products and hybridized with probes for the virulence plasmid, transposon and/or integron, and resistance genes (Fig. and ). This plasmid was labeled pUO-StVR2 (plasmid University of Oviedo-Salmonella enterica serotype Typhimurium virulence and resistance). To demonstrate the genetic relationship between pUO-StVR2 from the E. coli CT31 transconjugant and pSLT90 from LT2, both plasmids were tested by restriction analysis with three endonucleases (HindIII, ClaI, and EcoRI) and by hybridization with probes derived from pSLT90 virulence genes . Plasmid restriction analysis showed that each plasmid revealed a distinct restriction profile with some fragments banding at identical positions. The hybridizations made in order to locate the virulence plasmid genes in the HindIII, ClaI, and EcoRI restriction patterns showed that the traT probe always hybridized to matching fragments (18, 6.6, and 14.3 kb, with each endonuclease), the spvC probe also always hybridized to matching fragments (3.6, 6.6, and 3.8 kb), and the rck probe in contrast hybridized to mismatching fragments in all restriction patterns . To locate the transposon and/or integron and resistance genes in the restriction patterns of pUO-StVR2, hybridizations with specific probes were carried out . The most important findings were as follows. (i) The merA probe always hybridized to fragments (16.4, 8.9, and 4.6 and 3.8 kb with HindIII, ClaI, and EcoRI, respectively) different from those to which the other resistance gene probes hybridized. (ii) HindIII analysis showed three gene groups (probes hybridizing to the same fragment): catA1-tet(B) (an approximately 9.2-kb fragment), int1-oxa1 (an approximately 3-kb fragment), and aadA1-qacEpsilonDelta1-sul1 (an approximately 2.8-kb fragment). The two last groups were generated from the integron and were expected on the basis of its DNA sequence (accession number .1). (iii) ClaI analysis revealed a group including all the resistance genes tested: catA1-tet(B)-int1-oxa1-aadA1-qacEpsilonDelta1-sul1 (an approximately 15-kb fragment). (iv) EcoRI analysis showed the int1-oxa1-aadA1-qacEpsilonDelta1-sul1 group (an approximately 6-kb fragment), corresponding to the integron genes, which was separated from catA1 (an approximately 2.7-kb fragment) and tet(B) (about 15.4 and 3.5 kb). By this approach, linkage between the merA and integron genes could not be established. Characterization of the isolates containing the virulence-resistance plasmid pUO-StVR2. | The nine isolates carrying pUO-StVR2 showed the ACSSuT phenotype, but they were florfenicol susceptible. Only one of them was phage type DT104 (subtype b-low), six were non-phage typeable (NT), and one was DT120, and another presented a nonrecognized phage lysis pattern (RDNC). To establish the genetic relationship among the nine isolates as well as their relationship with the serotype Typhimurium control strains, the isolate DNAs were analyzed by XbaI PFGE . PFGE profiles were defined considering only the well-visualized fragments (size, >=18 kb) The nine isolates generated identical profiles (PFGE3) which presented between 11 and 14 matching fragments and between 6 and 11 mismatching fragments regarding the ones from the control strains. These results support the fact that the nine isolates should be classified as members of a single clonal group, well differentiated from the control strains, one of which represented the prevalent DT104-ACSSuT clone (with PFGE2) . Within this new clonal group, there is some heterogeneity with respect to phage type and plasmid content and a relationship between both markers seems to exist . Previous work has shown that changes in phage types can be the result of the loss or acquisition of plasmids . Since not all DT104 strains present the markers described for the so-called DT104-ACSSuT clone or clonal line (in our experience, NT strains could also be ascribed to this clone), it must be taken into account that phage typing alone is not a suitable method for the identification of a clone. Hypothesis concerning the origin of the virulence-resistance plasmid pUO-StVR2. | Some interesting facts regarding the dispersion of virulence and resistance plasmids, as well as of the genes harbored by pUO-StVR2, have previously been published. (i) Serovar-specific virulence plasmids of different sizes have been found in Salmonella subspecies I , but it has also been reported that some serotypes belonging to subspecies II, IIIa, and IV carry spv genes on the chromosome . (ii) At least three other types of virulence-resistance plasmids, each of them encoding different resistance genes, have been detected in Salmonella collected in different countries . (iii) The integron reported here is found forming part of transposon Tn2603 , and a similar integron has been found in a 140-kb IncFI plasmid of serotype Typhimurium non-DT104-resistant isolates collected in Italy . (iv) Both the catA1 and tet(B) genes are widely dispersed among Enterobacteriaceae and can be part of transposons (Tn9 and Tn10, respectively) and be located in plasmids . These facts encourage us to suggest the following possible events leading to the generation of pUO-StVR2. First, in the animal reservoir, a serotype Typhimurium bacterium showing a distinctive PFGE profile and carrying a virulence plasmid (in association with other small plasmids) acquired one (or more) resistance plasmid(s). In a second step, recombination-transposition processes between the virulence and the resistance plasmid(s) could have taken place, generating the hybrid plasmid (pUO-StVR2). This hybrid plasmid belongs to the serotype Typhimurium virulence plasmid family (because it carries the repFIIA, traT, spv, and rck genes) but is of a larger size since it has gained several resistance genes [oxa1, aadA1a, qacEDelta1, sul1, merA, catA1, and tet(B)], some of them integron and/or probably transposon related. Finally, pUO-StVR2-carrying Salmonellaenterica serotype Typhimurium strains spread and established in their reservoir and sporadically reach humans through the food chain. The latter is supported by the isolation of such strains from human feces during the period from 1993 to 1999. In addition to the strains described above, we found in the year 2000 three new serotype Typhimurium clinical isolates carrying the same resistance genes, including the integron-borne oxa1/aadA1a cassette configuration. To date, we have found no other Salmonella isolate presenting this cassette array among about 2,000 isolates (isolated between 1989 and 2000) tested for integrons and/or resistance plasmids (, , , ; unpublished data). The findings presented above support the fact that different serotype Typhimurium clones or clonal groups, with distinctive PFGE profiles, display similar resistance phenotypes based on different resistance genes which can be located on the chromosome and/or on plasmids. In each clone, some resistance genes and their gene arrays seem to be characteristic, frequently including one or more transposon and/or integron structures. Serotype Typhimurium has a wide animal reservoir, composed of mainly cattle and swine, which is linked to humans via the food chain. Consequently, the selection and maintenance of different resistance genes in different serotypes and clones like the one reported here constitute a serious public health problem. FIG. 1. | Plasmid analysis of representative serotype Typhimurium clinical isolates and control strains. Plasmid analysis of representative serotype Typhimurium clinical isolates and control strains. (A) Plasmid profiles. Lanes: C, molecular size standard plasmids; 1 through 7, profiles generated by LT2 (lane 1), LSP14/92-DT104 (subtype l) (lane 2), LSP31/93- DT120 (lane 3), E. coli CT31 (lane 4), LSP28/98-DT104 (subtype b-low) (lane 5), LSP106/94-NT, LSP233/98-NT, LSP349/98-NT; 362/98-NT, LSP436/99-NT, and LSP559/99-NT (lane 6), and LSP21/99-RDNC (lane 7). (B) Lanes: 1 through 4, successive hybridizations of lanes 1 through 4 from panel A, with the probes cited at the top of the figure; lambda, phage lambda DNA digested with PstI. Chr, chromosomal DNA. FIG. 2. | Restriction and hybridization analysis of serotype Typhimurium virulence plasmids. Restriction and hybridization analysis of serotype Typhimurium virulence plasmids. (A) Plasmid restriction profiles generated by HindIII, ClaI, and EcoRI (fragment sizes in kilobases). (B) Successive hybridizations of digested DNA with virulence plasmid gene probes (a, traT; b, rck; c, spvC) and resistance gene probes [d, merA; e, int1; f, oxa1; g, aadA1; h, qacEpsilonDelta1-sul1; i, catA1; j, tet(B)]. Lanes: lambda, phage lambda DNA digested with PstI (fragment sizes in kilobases); 1, pSLT90 from LT2; 2, pUO-StVR2 from E. coli CT31 transconjugant. Chr, chromosomal DNA. FIG. 3. | PFGE analysis of serotype Typhimurium strains representing different clones. PFGE analysis of serotype Typhimurium strains representing different clones. Lanes: L, PFGE ladder (New England Biolabs, Schwalbach, Germany); 1 through 4, XbaI PFGE profiles generated by serotype Typhimurium LT2 (PFGE1), LSP14/92 (PFGE3), LSP31/93 (PFGE4), and ATCC 14028 (PFGE2), respectively. Molecular sizes indicated in the figure correspond to the LT2 XbaI fragment sizes given in reference . The nine clinical isolates under study generated PFGE4. The arrow indicates the fragment corresponding to the 90-kb virulence plasmid. Backmatter: PMID- 12183273 TI - Antibiotic Susceptibilities of Parachlamydia acanthamoeba in Amoebae AB - Parachlamydia acanthamoeba are intracellular bacteria of amoebae and are considered potential etiological agents of human pneumonia. We have determined the in vitro antibiotic susceptibilities of two strains (strain Bn9 and Hall's coccus) in Acanthamoeba polyphaga. The two strains were susceptible to tetracyclines, macrolides, and rifampin, but resistant to fluoroquinolones. Keywords: Introduction : Based on 16S ribosomal DNA sequence comparison, the taxonomic classification of species belonging to the order Chlamydiales has been recently reassessed, and a new family, Parachlamydiaceae, has been proposed . This family now comprises two genera: i.e., the genus Parachlamydia with Parachlamydia acanthamoeba as a type species , and the genus Neochlamydia with Neochlamydia hartmannellae as the type species . Strains belonging to the species P. acanthamoeba include the type strain Bn9 (ATCC VR 1476), isolate Berg17, Hall's coccus , and unnamed isolates (, R. J. Birtles, T. J. Rowbotham, C. Storey, T. J. Marrie, and D. Raoult, Letter, Lancet 349:925-926, 1997), which are all strictly intracellular bacteria with variable Gram staining properties and which display more than 99% 16S rRNA gene similarity (Birtles et al., Letter). Parachlamydia spp. naturally infect Acanthamoeba. Trophozoites of Acanthamoeba hosting chlamydia-like bacteria have been isolated in patients with fever associated with use of humidifiers in Vermont (i.e., Hall's coccus) and from human nasal mucosa (i.e., the Bn9 and Berg17 strains) . Moreover, fourfold rising titers of antibodies directed against Hall's coccus have been detected by an immunofluorescence technique in sera from patients suffering from pneumonia of undefined cause in Ohio and Nova Scotia, Canada (Birtles et al., Letter). These sera did not react with Chlamydia trachomatis, Chlamydophila pneumoniae, or Chlamydophila psittaci antigens (Birtles et al., Letter). More recently, Marrie et al. reported detection of anti-P. acanthamoeba antibodies (antibody titer of 179 [there]phi [there]sigma 121:50) in 8 of 376 patients (similar2%) with community-acquired pneumonia compared with 0 of 511 healthy controls. Thus, a recent medical interest in Parachlamydia strains has arisen because of the potential etiological role in community-acquired pneumonia as well as nosocomial pneumonia, especially in patients with a humidifier . In this respect, the knowledge of their antibiotic susceptibilities could be of primary clinical importance. Especially, it is of particular interest to verify that current first-line recommendations for antibiotic therapy of pneumonia apply to this group of pathogens. Bacterial and amoebal strains. : P. acanthamoeba strain Bn9 was kindly provided by R. Amann (Lehrstuhl fur Mikrobiologie, Technische Universitat Munchen, Munich, Germany), whereas Hall's coccus was a gift from T. J. Robotham (Public Health Laboratory, Leeds, United Kingdom). Parachlamydia organisms were cultured in Acanthamoeba polyphaga, grown in 25-cm2 culture flasks (Becton Dickinson, Le Pont de Claix, France) containing PYG medium until almost complete lysis of amoebae (i.e., 4 days later). Cell supernatants were then recovered and centrifuged at 1,500 rpm (700 x g) for 10 min to remove cell debris. A Parachlamydia inoculum was prepared for each strain tested by diluting supernatants 1:100 in Page's amoebal saline , which corresponded to approximately 108 bacteria/ml. Titration of Parachlamydia was obtained by inoculating 10-fold serial dilutions of the primary inoculum to uninfected amoebal cultures and determining the highest dilution allowing lysis of amoebal monolayers after 4 days of incubation of cultures at 30C. Determination of MICs. : Uninfected amoebae, cultured in PYG medium, were harvested by gentle shaking of monolayers and dispensed (160 mul per well of a 5.105-organism/ml inoculum) in 96-well microtiter plates (D. Dutcher, Brumath, France). Each well received 20 mul of Parachlamydia inoculum (i.e., final inoculum of about 106 bacteria/ml). After a 2-h incubation of infected amoebal cultures at 30C, antibiotics were added (i.e., 20 mul of 10-fold the desired final concentrations). Controls were drug-free uninfected amoebae (as amoebal viability controls), uninfected amoebae with the various antibiotic concentrations tested (as amoebal antibiotic toxicity controls), and drug-free infected amoebae (as Parachlamydia growth controls). We also verified that the P. acanthamoeba strains tested did not grow in Page's amoebal saline in the absence of amoebae. Drug-free controls received 20 mul of saline instead of the antibiotic solution. Cultures were incubated at 30C and observed each day under an inverted microscope at a magnification of x400 until complete lysis of amoebal monolayers in Parachlamydia-infected drug-free controls occurred. By this simple technique, MICs corresponded to the lowest antibiotic concentration that prevented Parachlamydia growth (i.e., destruction of amoebal cultures after 4 days of incubation). Escherichia coli C.I.P. 53.126 and Staphylococcus aureus C.I.P. 103811 were obtained from the Pasteur Institute (Institut Pasteur, Marnes La Coquette, France) and were used to control the antibiotic concentrations tested, with Mueller-Hinton broth as the antibiotic assay medium according to the procedure recommended by the National Committee for Clinical Laboratory Standards . No antibiotic concentration tested displayed a toxic effect against amoebae. MICs for E. coli C.I.P. 53.126 and Staphylococcus aureus C.I.P. 103811 were compatible with those determined by the Pasteur Institute. P. acanthamoeba strains grew well in A. polyphaga cells, with complete lysis of amoebal monolayers in drug-free cultures after 4 days of incubation at 30C. Among the beta-lactams tested, penicillin G, amoxicillin, ceftriaxone, and imipenem were ineffective at concentrations up to 32 mug/ml . Both strains were susceptible to aminoglycosides, macrolides (including the newer ketolide, telithromycin), doxycycline, cotrimoxazole, and rifampin. In contrast, vancomycin, the activity of which is almost restricted to gram-positive bacteria, was ineffective. Thiamphenicol and, more importantly, the fluoroquinolone compounds ofloxacin and ciprofloxacin were not bacteriostatic at the concentrations tested. We have evaluated in vitro susceptibilities of two strains belonging to the species P. acanthamoeba (i.e., Bn9 and Hall's coccus), with A. polyphaga as an in vitro cell system to support growth of these strictly intracellular bacteria. An amoebal system was used because of the impossibility of growing these bacteria in the other cell systems we currently use in our laboratory, including McCoy cells, Vero cells, P388D1 macrophage-like cells, or human embryonic lung fibroblast cells. Our model was based upon inhibition of amoebal lysis due to bacterial multiplication when antibiotics were added to the culture supernatant compared to that of drug-free controls. Thus, it was critical to verify that amoebal lysis was not related to antibiotic toxicity. Despite these technical limitations, our model allowed us for the first time to define the antibiotic susceptibility pattern of P. acanthamoeba and to compare it with those previously reported for C. trachomatis, Chlamydophila pneumoniae, and Chlamydophila psittaci, species that also belong to the order Chlamydiales. P. acanthamoeba strains BN9 and Hall's coccus were found resistant to all beta-lactams tested. The in vitro activity of beta-lactams against C. trachomatis , C. pneumoniae , and C. psittaci has been demonstrated. Although these antibiotics are not considered first-line antibiotic therapy for Chlamydia-related pneumonia , amoxicillin has been used successfully in pregnant women with genital infection due to C. trachomatis . In contrast, we found aminoglycosides to be bacteriostatic against P. acanthamoeba strains, whereas C. trachomatis has been reported to be highly resistant to gentamicin . Cotrimoxazole could inhibit the growth of P. acanthamoeba and is also effective against C. trachomatis . In contrast, C. pneumoniae and C. psittaci are resistant to this antibiotic combination. More surprisingly, P. acanthamoeba strains were found to be resistant to fluoroquinolones, whereas C. trachomatis, C. pneumoniae, and C. psittaci are highly susceptible to these drugs . DNA gyrase is usually the primary target of fluoroquinolones in gram-negative bacteria , and resistance to fluoroquinolones due to mutation in gyrA (the gene encoding the alpha subunit of DNA gyrase) has been reported in C. trachomatis . The possibility of gyrA-mediated natural resistance to fluoroquinolones in P. acanthamoeba should be assessed. Our results should be specifically examined considering the potential role of Parachlamydia spp. as etiological agents of human pneumonia . beta-Lactams are considered first-line antibiotic therapy of Streptococcus pneumoniae-related pneumonia, but are poorly effective against intracellular pathogens responsible for atypical pneumonia, such as Chlamydia spp., Legionella pneumophila, Mycoplasma pneumoniae, or Coxiella burnetii . This may also apply for P. acanthamoeba, a species resistant in vitro to these agents. In contrast, the susceptibility of P. acanthamoeba to macrolides and tetracycline suggests that the current practice of prescribing a macrolide or a tetracycline compound in patients with atypical pneumonia may well apply in case of Parachlamydia infection. The new ketolide compound telithromycin, which is active against erythromycin-resistant S. pneumoniae , as well as against the intracellular pathogens C. pneumoniae , L. pneumophila , M. pneumoniae , and C. burnetii , was found also active against the Parachlamydia strains. Fluoroquinolones, especially ofloxacin and ciprofloxacin, have been advocated as a possible alternative to macrolides in patients suffering atypical pneumonia, although their equivalence to erythromycin in case of legionellosis is still disputed . Interestingly, we found P. acanthamoeba to be highly resistant to these compounds in vitro. TABLE 1 | MICs for Parachlamydia sp., including the Bn9 strain and Hall's coccus, as determined in an A. polyphaga culture model Backmatter: PMID- 12183258 TI - Pharmacodynamic Assessment of Ertapenem (MK-0826) against Streptococcuspneumoniae in a Murine Neutropenic Thigh Infection Model AB - The objective of this study was to determine the susceptibility breakpoint of a new carbapenem, ertapenem (MK-0826), against Streptococcus pneumoniae strains based on bacterial density and survival studies in a murine thigh infection model. Sixteen S. pneumoniae isolates for which MICs ranged from 0.015 to 4.0 mg/liter were tested with neutropenic ICR mice. Animals were infected with bacteria at 105 to 106 CFU per thigh and were treated with ertapenem starting at 2 h postinfection for 4 days. Ertapenem was given subcutaneously at 50 mg/kg of body weight every 6 h, which simulates the human pharmacodynamic profile (in particular, the duration of time that the concentration of free drug remains above the MIC of 2 mg/liter). At 0 and 24 h postinfection, thighs were harvested for bacterial density determination. Survival was assessed during 4 days of therapy and 3 days after the therapy. A protein binding study was conducted with mice by use of the ultrafiltration method. Protein binding in mice was approximately 95%, which is comparable to that in humans. The average change in bacterial density ranged from -0.22 to -4.4 log CFU per thigh over 24 h compared to 0-h controls. The extent of microbial eradication was dependent on the MIC for the S. pneumoniae isolate. Substantial bactericidal activities (i.e., killing of approximately 2 log CFU per thigh) were consistently observed against isolates for which MICs were <=2 mg/liter, which also resulted in nearly 100% survival during the 4 days of drug dosing and 3 days after the therapy. Less-pronounced and highly variable bactericidal activities were detected against isolates for which the MIC was 4 mg/liter. Substantial enhancement in bactericidal activity was observed for CBA/J mice and is attributed to the contribution of the host defenses in the immunocompetent species. Assessment of the effectiveness of ertapenem by bacterial-density reduction over 24 h and by survival over 4 days of therapy in the murine thigh infection model reveals that the drug maintains maximal efficacy against S. pneumoniae isolates for which the MIC of this agent is <=2 mg/liter. Keywords: Introduction : Ertapenem (MK-0826) is a new, long-acting 1-beta-methyl carbapenem with potent antimicrobial activity against a variety of pathogenic species including pneumococci . While this agent apparently has good in vitro activity against Streptococcus pneumoniae, the pharmacodynamic assessment of the susceptibility breakpoint of ertapenem against this important pathogen has not been fully described. The availability of these data will assist not only in optimizing the effectiveness of the prescribed antimicrobial regimen in clinical practice but also in the assessment of appropriate National Committee for Clinical Laboratory Standards (NCCLS) breakpoints for this antimicrobial agent. The purpose of the present study was to determine the susceptibility breakpoint of ertapenem against S. pneumoniae strains based on bacterial density and survival studies using a murine model of pneumococcal thigh infection. MATERIALS AND METHODS : Antimicrobial agents. | Ertapenem (analytical grade standard; lot L-749345-002C089, 98.4% purity) for in vitro and in vivo testing was obtained from Merck Research Laboratory. Ertapenem was reconstituted with a 0.1 M morpholineethanesulfonic acid (MES)-ethylene glycol (1:1, vol/vol) solution for all in vitro testing. For animal dosing, an ertapenem solution was freshly prepared with sterile water for injection prior to drug administration. Bacterial isolates and susceptibilities. | Sixteen clinical isolates of S. pneumoniae were included in this study. The MIC for ertapenem was determined by the microdilution method according to NCCLS guidelines by using cation-adjusted Mueller-Hinton broth (20 to 25 mg of calcium/liter and 10 to 12.5 mg of magnesium/liter) with 5% lysed horse blood in ambient air . Trypticase soy agar with 5% sheep blood was used as the growth medium for S. pneumoniae isolates. Thigh infection model. | Specific-pathogen-free female ICR mice weighing approximately 25 g (Harlan Sprague Dawley Inc., Indianapolis, Ind.) were utilized throughout the experiment. All care and the experiments described in this report were approved by and performed in accordance with guidelines of the Hartford Hospital Institutional Animal Care and Use Committee (IACUC). Mice were rendered transiently neutropenic by two intraperitoneal (i.p.) injections of cyclophosphamide, the first at a dose of 150 mg/kg of body weight 4 days before infection and the second at a dose of 100 mg/kg 1 day before infection . Renal impairment was induced by a single i.p. injection of uranyl nitrate 3 days before infection . Broth cultures of the test organism were grown overnight and subsequently diluted to an inoculum of approximately 106 to 107 CFU/ml. Final inoculum concentrations were confirmed by serial-dilution and plating techniques. Thigh infection with each of the test isolates was produced by injecting 0.1 ml of the inoculum into each mouse thigh 2 h prior to initiation of antimicrobial therapy. Pharmacokinetic studies and dosing-regimen determination. | Pharmacokinetic studies were undertaken using neutropenic ICR mice in order to find an ertapenem regimen for the murine pneumococcal thigh infection model that simulates the pharmacokinetic profile observed in humans receiving 1-g dosing every 24 h (q24h). In an attempt to optimize the pharmacokinetic profile of ertapenem, renal impairment was induced by a single i.p. injection of uranyl nitrate (5 mg/kg) 3 days before infection. Single doses of ertapenem at 20, 50, and 100 mg/kg were administered by subcutaneous injection in a 0.2-ml volume 2 h after pneumococcal thigh inoculation. Blood samples from four mice per time point were collected by intracardiac puncture at 0.25, 0.5, 1, 2, 4, 6, and 8 h after dosing. Serum was separated by centrifugation at 3,000 x g for 10 min and was then transferred into polypropylene tubes and stored at -80C until analysis. Concentrations of ertapenem in murine serum were determined by using a validated reverse-phase high-pressure liquid chromatography (HPLC) assay. The sample extraction method involved liquid-phase extraction. After addition of 50 mul of an internal standard (meropenem, 50 mg/liter) to the mouse serum samples (200 mul), 800 mul of acetonitrile was added. Protein precipitation was achieved by a brief vortexing followed by centrifugation at 2,600 x g for 10 min. Supernatants were transferred into clean labeled tubes, and then 2.5 ml of dichloromethane was added. Again samples were briefly vortexed and subjected to centrifugation. The resultant aqueous solution was separated and transferred into vials for injection into the HPLC system. Chromatography was performed at ambient temperature on a reversed-phase C18 column (5 mum; 100 by 4.6 mm; Keystone Scientific Inc., Belletonte, Pa.) with an injection volume of 30 mul. The mobile phase, consisting of 25 mM phosphate buffer (pH 6.5) solution-methanol (100:9.5, vol/vol), was delivered via a Waters HPLC pump at a flow rate of 1.0 ml/min. Ertapenem and the internal standard were eluted at approximately 10.5 and 14.9 min, respectively. UV detection was performed at a wavelength of 300 nm. The assay was linear over a range of 0.15 to 50 mg/liter. Intra-assay coefficients of variation for the low (1.0-mg/liter) and high (40-mg/liter) check samples were 2.2 and 4.0%, respectively. Interassay coefficients of variation for the low and high check samples were 3.6 and 4.3%, respectively. Individual concentration data from all the dosing regimens were analyzed by using a population approach with the NONMEM computer program, version V, level 1.1; NM-TRAN, version III, level 1.0; and PREDPP, version IV, level 1.0 (NONMEM Project Group, University of California at San Francisco, San Francisco, Calif.). One- and two-compartment structural models with first-order absorption and elimination were compared to fit the data by using the first-order estimation method. Residual error was modeled by using a combined proportional and additive error model. Model selection was based on a likelihood ratio test, Akaike's information criterion, and evaluation of goodness-of-fit plots from the model. The statistical significance level was set a priori at 0.005. Based on the population parameter estimates from NONMEM analysis, the dosing regimen was calculated in such a way as to simulate the drug exposure found in humans following 1-g q24h dosing (area under the concentration-time curve from 0 to 24 h [AUC0-24], 500 to 600 mgh/liter; maximum concentration of drug in serum [Cmax], 140 mg/liter; duration of time that the free drug concentration remains above the MIC of 2 mg/liter [T>MICfree], 20%). Protein binding. | The protein binding of ertapenem was determined by an ultrafiltration (Centrifree) method with ICR mouse specimens. Three concentrations of ertapenem (1, 10, and 100 mg/liter) were tested with freshly collected mouse serum. All spiked serum samples were placed in a 37C shaking water bath for 10 min before being loaded onto Centrifree filters (30,000-molecular-weight cutoff; Millipore). Filtrates were separated by centrifugation at 1,000 x g for 15 min, and samples were measured by a validated HPLC method. Therapeutic efficacy as assessed by bacterial density. | Starting at 2 h after infection, ICR mice were treated with ertapenem at 50 mg/kg q6h by subcutaneous injection in a 0.2-ml volume for 24 h. This dosing regimen was calculated on the basis of the preliminary pharmacokinetic analysis to simulate the drug exposure observed in humans. Control animals received water in the same volume (0.2 ml) and on the same schedule as ertapenem. Untreated control mice (four per group) were sacrificed just prior to antibiotic initiation and after 24 h. Ertapenem-treated animals were euthanized by CO2 exposure followed by cervical dislocation after 24 h. After sacrifice, both thighs were removed and individually homogenized in normal saline. Serial dilutions were plated on Trypticase soy agar with 5% sheep blood for CFU determinations. For the purposes of these studies, efficacy (change in bacterial density) was calculated by subtracting the mean log CFU per thigh of the control mice obtained just prior to antibiotic administration from the log CFU per thigh of each ertapenem-treated mouse at the end of therapy (24 h). In addition, we compared the bactericidal activities of ertapenem against S. pneumoniae in the murine neutropenic and nonneutropenic thigh infection models. An S. pneumoniae strain (SP-129; MIC, 2 mg/liter) was tested with neutropenic ICR mice and immunocompetent CBA/J mice (Charles River Laboratories, Wilmington, Mass.). The animal infection procedure for CBA/J mice was the same as the procedure described above for ICR mice except that the CBA/J mice received no cyclophosphamide. Animals were inoculated at 105 to 106 CFU per thigh and were treated starting at 2 h postinfection for 24 h. At 0 and 24 h postinfection, thighs were harvested for bacterial density determination. The regimens for ertapenem were 50 mg/kg q6h and q24h. Four animals were used in each study arm. Therapeutic efficacy as assessed by survival. | Groups of 20 mice were similarly infected with each test strain for evaluation of survival after 96 h of therapy. Ertapenem therapy (50 mg/kg q6h) was initiated 2 h after inoculation in 15 animals. The remaining five animals received the same volume of water q6h and served as the controls. Cumulative mortality was calculated during 96 h of therapy. Although death has historically been used as an end point for studies of this type, this end point is no longer suitable in the present era of animal research. Therefore, our study methodology has been modified to contemporary standards. We utilized the following approach in order to lessen the duration of pain and suffering during the mortality experiment. (i) Animals were monitored four times daily by members of the study team who have been trained and are experienced in recognizing the signs of illness and abnormal behavior. (ii) Animals which appeared to have substantial alterations in posture (e.g., abnormal posture or head tucked into abdomen), coat, exudate around eyes and/or nose, and breathing or movement were removed from the group housing and were euthanized. The term "mortality" has been used as an end point for this study; however, it should be clearly understood that, when possible, every attempt was made to minimize pain and suffering and that the animals were euthanized prior to naturally succumbing to infection if pain and suffering were observed. For the purposes of this study, death due to the natural infection process and euthanasia were considered the same end point for experimental and statistical purposes. Data analysis. | Spearman's rank correlation coefficient was used to evaluate the correlation between change in CFU and T>MICfree for ertapenem after 24 h of therapy as well as the correlation between mortality and T>MICfree for ertapenem after 96 h of therapy. In addition, the sigmoid Emax model was applied to further evaluate the relationship between these variables. RESULTS : The MICs of ertapenem for the S. pneumoniae isolates incorporated into this study are displayed in Table . The test organisms selected represent a wide range of sensitivities to ertapenem, with MICs ranging from 0.015 to 4 mg/liter. Concentration-independent protein binding levels for ertapenem were observed over the concentrations studied. The average protein binding level in murine serum was 95.5% +- 1.0%, which is comparable to the 94% protein binding at a concentration of 100 mg/liter observed in humans (; product package insert, 2001, Merck & Co., Inc., Whitehouse Station, N.J.). A total of 130 concentrations were used to compute population pharmacokinetic parameters. The pharmacokinetic disposition of ertapenem in mice was best described with a one-compartment model with first-order absorption and elimination; parameter estimates are presented in Table . The population analysis was quite robust, as evidenced by the plot of the observed versus the population predicted concentrations of ertapenem in infected animals . Based on the population analysis, a 50-mg/kg q6h dosing regimen produced an AUC0-24 of 586 mg h/liter, a Cmax of 140 mg/liter, and a T> MICfree of 22%, which are comparable to those found in humans after 1-g q24h dosing. This simulated regimen was then administered to another group of mice to confirm the predicted values of the modeling scheme. As displayed in Fig. , the regimen selected produced a concentration-versus-time profile in this murine model that was comparable to the NONMEM-predicted profile. As presented in Fig. , excellent recovery of bacteria from infected thighs of control animals prior to the start of therapy was obtained for all isolates, with an average bacterial density of 6.09 +- 0.23 log10 CFU per thigh. These data support the accuracy of inoculum preparation and the consistency of the inoculation procedure. Figure displays the growth of each organism in control animals over the 24-h postinfection period. Average bacterial growth in untreated control animals over 24 h was 2.1 +- 0.55 log10 CFU per thigh, with a range from 0.9 to 2.9 log10 CFU per thigh. The mean changes in bacterial density at the conclusion of 24 h of therapy with 50 mg of ertapenem/kg q6h are presented in Fig. . The magnitude of killing ranged from 0.22 to 4.4 log10 CFU per thigh over 24 h, and the extent of microbial eradication appeared to be related to the MIC for the pneumococcal isolate. Substantial bactericidal activities were observed against isolates for which MICs were <=2 mg/liter, with a mean bacterial kill of 3.1 log10 CFU per thigh. Less-pronounced and highly variable bactericidal activities were detected against isolates for which the MIC was 4 mg/liter. Spearman's rank correlation coefficient calculation revealed a significant correlation between the change in CFU and the T>MICfree for ertapenem after 24 h of therapy (P < 0.05). The relationship between T>MICfree and log10 CFU was further described with a sigmoid Emax model . As shown in Fig. , maximal bacterial killing activity appears to be reached once the T>MICfree exceeds 30%. A comparison between the bactericidal activities of ertapenem in neutropenic ICR mice and nonneutropenic CBA/J mice was undertaken. Excellent recovery of bacteria was observed for both ICR and CBA/J mice at 2 h after bacterial inoculation, which is consistent with previous data . However, levels of bacterial growth in untreated control ICR mice and untreated control CBA/J mice were significantly different (P < 0.05) after 24 h of inoculation. Organisms grew approximately 2.5 log10 CFU per thigh over 24 h in untreated ICR mice, while an average reduction in bacterial density of 1.1 log10 CFU per thigh was observed for untreated CBA/J mice . Furthermore, significantly different bactericidal effects were observed for ertapenem-treated ICR versus CBA/J mice. As shown in Fig. , the bactericidal activity observed in CBA/J mice was considerably enhanced over that in ICR mice (P < 0.05). Cumulative mortality over 96 h of ertapenem treatment and 3 days posttherapy is reported in Fig. . Observed mortality in untreated control animals was 100% over this observation period except with one isolate. Animals treated with ertapenem survived the infection during the 4 days of treatment with a 93% survival rate. No significant correlation was detected between mortality and T>MICfree. FIG. 1. | Population predicted concentrations versus concentrations observed by NONMEM analysis. Population predicted concentrations versus concentrations observed by NONMEM analysis. FIG. 2. | Goodness of fit between the NONMEM simulated ertapenem concentration-time profile after 50-mg/kg dosing and the mean observed serum ertapenem concentration-versus-time profile in the murine thigh infection model. Goodness of fit between the NONMEM simulated ertapenem concentration-time profile after 50-mg/kg dosing and the mean observed serum ertapenem concentration-versus-time profile in the murine thigh infection model. FIG. 3. | Densities of S. pneumoniae Densities of S. pneumoniae in the thighs of infected animals at the start of therapy. Each value represents 1 of 16 isolates and is the mean +- standard deviation for 8 thighs. FIG. 4. | Growth in density of S. pneumoniae Growth in density of S. pneumoniae in the thighs of infected control animals over 24 h. Each value represents 1 of 16 isolates and is the mean +- standard deviation for 8 thighs. FIG. 5. | Changes in density of S. pneumoniae Changes in density of S. pneumoniae in the thighs of infected animals after 24 h of ertapenem therapy. Each value represents 1 of 16 isolates and is the mean +- standard deviation for 8 thighs. FIG. 6. | Sigmoid Emax evaluation of the correlation between T>MICfree and the bactericidal activity of ertapenem. Sigmoid Emax evaluation of the correlation between T>MICfree and the bactericidal activity of ertapenem. r2 = 0.88. FIG. 7. | Growth in density of S. pneumoniae Growth in density of S. pneumoniae (SP-129) in the thighs of infected immunocompromised (ICR) and immunocompetent (CBA/J) control animals over 24 h. Values are means +- standard deviations for 8 thighs. FIG. 8. | Changes in density of S. pneumoniae Changes in density of S. pneumoniae (SP-129) in the thighs of infected immunocompromised (ICR) and immunocompetent (CBA/J) animals after 24 h of ertapenem therapy. Values are means +- standard deviations for 8 thighs. FIG. 9. | Percentage of survival of pneumococcus-infected animals after 4 days of ertapenem treatment and 3 days posttreatment (n = 15). Percentage of survival of pneumococcus-infected animals after 4 days of ertapenem treatment and 3 days posttreatment (n = 15). Each set of bars represents one S. pneumoniae isolate. TABLE 1 | MICs of ertapenem for the S. pneumoniae test isolates TABLE 2 | Population pharmacokinetic parameter estimates for ertapenem in mice infected in the thigh with S. pneumoniae DISCUSSION : S. pneumoniae remains one of the leading causes of community-acquired bacterial infections, and severe S. pneumoniae infections such as pneumonia and meningitis have significant morbidity and mortality rates . In the present study, we evaluated the pharmacodynamic profile of ertapenem against 16 clinical isolates of S. pneumoniae with a wide range of sensitivities by using the neutropenic murine model of thigh infection. It has been shown that the most important pharmacokinetic/pharmacodynamic parameter for prediction of the antimicrobial effect of ertapenem is T>MIC (M. L. Van Ogtrop, D. Andes, and W. A. Craig, Abstr. 39th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 999, 1999). In addition, ertapenem has been reported as a highly protein-bound drug, which is in accordance with our observation (95% bound; Merck & Co., Inc., product package insert, 2001). Therefore, the dosing regimen employed in this study was selected based on the drug exposure found in humans following 1-g q24h dosing, in which the T>MICfree is approximately 20%. In the present study, the serum pharmacokinetic profile produced by the 50-mg/kg q6h dosing regimen in our model was quite similar to the target exposure in humans. Over the course of the study, excellent recovery of bacteria from infected thighs was noted for all the isolates prior to the initiation of therapy. However, the growth of organisms in untreated control animals over the 24-h treatment period was quite variable, revealing the inherent variability of in vivo growth of S. pneumoniae, which is consistent with previous findings . Use of the simulated exposure with a number of S. pneumoniae strains for which MICs were varied yielded a wide range of levels of bacterial killing over the 24 h of therapy. The extent of bacterial killing appears to be related to the MIC of ertapenem for the pneumococcal isolate. Substantial bactericidal activities were consistently observed against isolates for which MICs were <=2 mg/liter. In contrast, less-pronounced and highly variable bactericidal activities were detected against isolates for which the MIC was 4 mg/liter. Despite variable bactericidal activity, the overall survival rate was very high (93%) over the 7-day observation period. While inconsistent, this degree of killing of these organisms (MIC, 4 mg/liter) during the initial 24-h bacterial-density studies was sufficient to prevent mortality, an observation which has been made in similar studies with other compounds . Pharmacodynamic evaluation revealed that T>MICfree was closely correlated with efficacy, and the relationship between these two parameters can be adequately described with the sigmoid Emax model. In addition, as shown in Fig. , the T>MICfree required for a static effect is approximately 6%, which is in agreement with previous observations (M. L. Van Ogtrop, D. Andes, and W. A. Craig, Abstr. 38th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-48, 1998). The magnitude of the pharmacokinetic/pharmacodynamic parameter required for a static effect is much less than that observed with penicillin and cephalosporins . Moreover, we compared the bacterial-density determinations for the immunocompromised (ICR mice) and the immunocompetent (CBA/J mice) host with the same pneumococcal strain. A substantial enhancement in bactericidal activity was observed for the CBA/J mice; this is attributed to the contribution of the host defenses in the immunocompetent species. This is in accordance with a previous observation, which also demonstrates the host immunocompetency effect on efficacy . In summary, ertapenem demonstrated substantial bactericidal activity against S. pneumoniae in the murine thigh infection model. The relationship between T>MIC and bactericidal activity can be characterized with the sigmoid Emax model. In the context of the isolates and dosing regimen studied, ertapenem therapy protected animals from death. Enhanced bactericidal activities were observed in the immunocompetent animals. Based on these data and the consideration that the presence of neutrophils would be expected to further enhance the activity of ertapenem, the proposed susceptibility breakpoint for ertapenem would be 2 mg/liter. Backmatter: PMID- 12183228 TI - Different Levels of Genetic Homogeneity in Vancomycin-Resistant and -Susceptible Enterococcus faecium Isolates from Different Human and Animal Sources Analyzed by Amplified-Fragment Length Polymorphism AB - The genetic relationship among fecal vancomycin-resistant Enterococcus faecium (VREF) and vancomycin-susceptible E. faecium (VSEF) isolates (n = 178) from the same populations of pigs, human healthy volunteers, and hospitalized patients (from The Netherlands) and chickens (from The Netherlands and Greece) was studied by amplified-fragment length polymorphism (AFLP). The majority of VREF isolates from pigs, healthy volunteers, and hospitalized patients grouped together (genetic similarity, >=65%). In a previous AFLP study by our group the VREF isolates from hospitalized patients grouped separately, most likely because these were clinical and not fecal isolates as in the present study. Furthermore, VSEF isolates from humans and pigs were found much more genetically diverse than VREF isolates, whereas VREF and VSEF isolates from chickens clustered together in a separate genogroup (genetic similarity, >=65%), a pattern clearly distinct from the patterns for human and pig isolates. The present study suggests that pigs are a more important source of VREF for humans than chickens and that human- and pig-derived VSEF isolates seem much more heterogeneous than VREF isolates. Keywords: Introduction : In countries of the European Union (EU) vancomycin-resistant enterococci (VRE) are relatively frequently found in healthy humans in the community and in farm animals and vancomycin resistance is mostly vanA mediated. This considerable pool of possibly transmissible isolates with vanA-mediated glycopeptide resistance in the EU is very likely caused by the use of avoparcin (an analogue of vancomcyin) as a growth promoter in animal husbandry until April 1997 . Occasionally, genetically related VRE isolates have been found in food animals, meat products , outpatients, and hospitalized patients, suggesting that transmission between animals and humans can occur and may contribute to colonization and subsequent infection in humans . Thus far, molecular comparisons by pulsed-field gel electrophoresis (PFGE) and amplified-fragment length polymorphism (AFLP) analysis of human- and animal-derived enterococci have been done only on VRE strains. These comparisons revealed the existence of a common human and pig genogroup, while poultry-derived vancomycin-resistant Enterococcus faecium (VREF) isolates clustered in a separate genogroup. However, the typing of the Tn1546 transposon showed that identical Tn1546 derivatives were found in humans and poultry, suggesting horizontal spread of the vanA transposon from poultry to humans. Until now, it was not known whether VREF strains constitute a separate population within the E. faecium population and whether host-specific AFLP genogroups are also found among VSEF strains. Previous studies mainly focused on the genetic relationship among isolates from animals and humans within one country . In this study, we used AFLP analysis to study the genetic relationship among 178 VREF and vancomycin-sensitive E. faecium (VSEF) strains isolated from fecal samples from humans and pigs in The Netherlands and from chickens in Greece and The Netherlands. AFLP has the ability to establish genetic relatedness between strains that, by PFGE, would show no similarity at all. PFGE is the reference standard for tracing the transmission of strains in hospital outbreaks but is too discriminatory to determine genetic relatedness among epidemiologically unrelated strains . The AFLP genogroups defined in the present study were compared to our previous results for host-specific genogroups among 255 VREF isolates from hospitalized patients, nonhospitalized persons, and various animal sources (pigs, poultry, calves, dogs, and cats) from nine different countries . Including vancomcyin-susceptible enterococci in molecular epidemiology studies might provide more insight into the composition of the enterococcal intestinal flora in humans and animals and help to further elucidate the transmission routes and persistence of E. faecium strains from animals in the human gut . (Part of the study was presented as a poster at the 41st Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, Illinois, 15 to 19 December 2001 [abstr. 1875].) MATERIALS AND METHODS : Bacterial isolates. | All VREF isolates and a comparable number of VSEF isolates were derived from four different populations of (i) healthy individuals (n = 537), (ii) hospitalized patients (n = 100), (iii) pigs (n = 126), and (iv) poultry (n = 139). The fecal samples from healthy individuals were collected from volunteers living in the north (n = 129), south (n = 171), and west (n = 133) of The Netherlands and were selected by using random addresses from the telephone directory in 1999. From the southern region fecal samples from healthy volunteers were also collected in 1996 (n = 104). The fecal samples from hospitalized patients came from patients admitted to the surgical ward of the University Hospital Maastricht in 1999. The poultry samples came from Greek (n = 50) (kindly provided by K. Sarris, Salonika, Greece) and Dutch slaughterhouses (n = 89). The last group of fecal samples were derived from a Dutch pig slaughterhouse. Multiple sampling from both pigs and poultry from the same farm was prevented by collecting feces from at most every 300th pig and by collecting only one chicken sample in the morning and one in the afternoon after evisceration at the slaughtering line. All fecal samples were diluted 1:10 in 0.9% (wt/vol) NaCl supplemented with 20% (vol/vol) glycerol on the day of arrival at the bacteriological laboratory and were stored at -20C until assayed. After the samples were thawed, 1:10 and 1:1,000 dilutions (40 mul each) were inoculated on KF-Streptococcus agar plates (CM701; Oxoid, Basingstoke, United Kingdom), with and without vancomycin (concentration: 10 mg/liter), by using a spiral plater (Salm en Kip BV, Utrecht, The Netherlands) as previously described . After 48 h of incubation at 42C, one typical enterococcal colony from each sample dilution was randomly chosen from the plates with and without vancomycin, and colonies were identified by generally accepted methods . In total 178 fecal E. faecium isolates were included into the study: 93 VREF isolates and 85 VSEF isolates. Sixty-four isolates were selected from healthy volunteers, of which 37 were VREF and 27 were VSEF. Seven VREF isolates and 12 VSEF isolates were derived from the hospitalized patients. A total of 35 (VREF, n = 16; VSEF, n = 19) and 34 isolates (VREF, n = 19; VSEF, n = 15) came from Greek and Dutch poultry farms, respectively, and 26 isolates were derived from a Dutch pig slaughterhouse (VREF, n = 15; VSEF, n = 11). None of the isolates in the present study were included in our previous study . AFLP analysis. | All 178 isolates were analyzed by AFLP. DNA was isolated as described elsewhere , with the addition of a final ethanol precipitation step to further purify the DNA. The AFLP analysis and the degree of genetic similarity of >=65% for the AFLP patterns to distinguish genogroups were as previously prescribed . This level of 65% was arbitrarily set and is not an absolute limit but distinguished four main genogroups and showed a clear association with the source of the strains. To form a specific genogroup, the minimal number of strains needed was five. The genogroups distinguished were compared to the genogroups of our previous study . Genotypic diversity based on AFLP typing was calculated by the following equation: genotypic diversity (GD) = [n/(n - 1)](1 - Sigmaxi2), where xi is the frequency of the ith identical AFLP type and n is the number of strains . Two isolates were considered to have an identical AFLP type when the similarity of banding patterns was >95%. This cutoff of 95% was based on the previous finding that the degree of similarity between quadruplicate isolates of the same organism was 95 to 99% . When isolates are highly diverse, no two isolates have an identical AFLP type, and the calculated GD is 1. When all isolates are identical the GD is 0. Thus an increase in GD suggests an increased level of genetic heterogeneity. Statistical analysis. | The chi square test (two-sided, P < 0.05) was used to analyze the differences in the distributions of the different experimental groups among the AFLP genogroups. RESULTS : AFLP analysis. | In total 178 strains, 93 VREF strains and 85 VSEF strains, from different animal and human sources, healthy volunteers, hospitalized patients, pigs, and poultry, were subjected to AFLP typing. Four main groups of strains that shared >=65% of their restriction fragments were formed as a result of grouping by AFLP. Most VREF isolates from healthy volunteers (35 of 36) and hospitalized patients (6 of 7) and all VREF isolates from pigs (15 of 15) clustered in genogroup A (Fig. ; Table ), and all of these isolates clustered in our previously reported genogroup A . Compared to the VREF isolates, the VSEF isolates from humans and pigs were genetically more heterogeneous and were dispersed over the entire dendrogram . A significant difference in the distributions of genogroups between the VREF and VSEF isolates from human volunteers and pigs was found (P < 0.05; Table ). New genogroups E and F, mainly consisting of VSEF isolates (11 of 11 and 8 of 9, respectively), were identified in addition to the previous published genogroups . The isolates that clustered in genogroup E were derived from healthy volunteers (9 of 11) and hospitalized patients (2 of 11), and the isolates in genogroup F were from hospitalized patients (2 of 9), pigs (4 of 9), and poultry (3 of 9) (Fig. ; Table ). Fifteen strains were not assigned to any of the genogroups since they did not cluster in a group comprising at least 5 strains. The majority of these strains (12 of 15) were vancomycin sensitive and were isolated from healthy volunteers (10 of 12) and hospitalized patients (2 of 12) (Fig. ; Table ). The genetic heterogeneity of VSEF was also confirmed by the greater genotypic diversity of the vancomycin-sensitive enterococci isolated from human volunteers, hospitalized patients, and pigs (GD, 1.00, 0.97, and 0.95, respectively) than of the vancomycin-resistant isolates (GD, 0.77, 0.87, and 0.84, respectively) . The vast majority of the poultry isolates (65 of 69) grouped in genogroup B , and all clustered within the previously reported genogroup B . The poultry isolates grouped clearly separately from isolates recovered from humans and pigs, irrespective of whether they were VREF or VSEF and of Dutch or Greek origin (Fig. ; Table ). This clustering of poultry-related VREF and VSEF isolates also means that the difference in genotypic diversity between VSEF and VREF isolates from poultry (GD, 0.93 and 0.88, respectively) was smaller than that between VSEF and VREF isolates from humans and pigs . FIG. 1. | Dendrogram of the genetic similarity of all isolates by AFLP analysis. Dendrogram of the genetic similarity of all isolates by AFLP analysis. Groups A, B, E, and F are based on a 65% genetic similarity. The dots indicate the sources, the vancomycin resistance phenotypes, and the countries of origin of the isolates. TABLE 1 | Distribution of isolates of the different experimental groups among the AFLP genogroups TABLE 2 | Genotypic diversity found in the different experimental groups DISCUSSION : In our previous study, 255 VREF strains from different human and animal sources were subjected to AFLP typing. Four genogroups (A to D), with clustering of isolates from nonhospitalized volunteers and pigs in genogroup A, chickens in genogroup B, hospitalized patients in genogroup C, and vealers in genogroup D, were discriminated . In the present study, the VREF strains from (nonhospitalized) healthy volunteers and pigs also clustered in the same genogroup A. However, the strains from hospitalized patients also clustered in genogroup A, whereas in the previous study they formed genetically distinct genogroup C. This observed difference between the two studies may be explained by the fact that in this study the VREF isolates derived from hospitalized patients were not clinical isolates from infections or associated with hospital outbreaks but rather were isolates from the fecal flora. Therefore, it is very likely that the VREF isolates derived from hospitalized patients were already acquired in the community, as also found by Endtz et al. . The VREF strains derived from healthy volunteers in 1996 and 1999 and from different regions in The Netherlands showed no specific clustering of AFLP patterns by year or place and mainly clustered in genogroup A (specific data not shown). The genetic similarity between human and pig VREF isolates did not extend to VSEF isolates. Especially, the VSEF isolates derived from healthy volunteers were genetically more diverse then their resistant counterparts and were spread among genogroups A, B, and E. In addition, 38% did not cluster in one of the four main genogroups at all. Even though the numbers of VSEF strains derived from hospitalized patients (n = 12) and pigs (n = 11) were smaller, these strains also showed more genetic diversity than the resistant strains. In contrast to the clear distinction between the relatively homogeneous VREF strains and relatively heterogeneous VSEF strains derived from both humans and pigs, the vast majority of VREF and VSEF strains derived from chickens clustered in genogroup B and displayed a more or less identical genotypic diversity. Perhaps the antibiotics especially used in poultry farming have already selected for a specific genetic poultry E. faecium population. The finding that strains isolated from Dutch and Greek chickens clustered in genogroup B suggests that the clustering of strains may be more related to the type of animal they were isolated from than their geographic origin. However most broilers in the world not only are genetically related but also descend from only a few breeding centers. Hence a common source of contamination cannot be excluded. The finding of a separate (VREF) poultry group (by AFLP analysis) distinct from human isolates was also reported in our previous study . Also, van den Braak et al. found two major PFGE types of VRE among poultry-derived strains that were not found in the fecal flora of patients . These results suggest that clonal transmission of VREF via the food chain from chickens to humans is less important than transmission from pigs to humans. This might be due to the fact that chicken-specific VREF strains may have difficulty in persisting in the human intestinal tract. Nevertheless, similarity of PFGE and AFLP types between turkeys and turkey farmers and between poultry and poultry farmers has been found, indicating that clonal dissemination of poultry strains to corresponding farmers does occur. For comparison among different typing methods a subset of strains from this study were subjected to a multilocus sequence typing (MLST) scheme developed by our group, and an article describing this MLST scheme for E. faecium has been published . In conclusion, human- and pig-derived VSEF strains are much more heterogeneous than their resistant counterparts and do not form a single genogroup. The VREF strains derived from humans and pigs are genetically different from the majority of the VSEF strains from pigs and humans and VREF and VSEF strains from poultry. These results suggest that pigs are a more important VREF source for humans than chickens by way of clonal dissemination. Why strains derived from chickens do not seem to persist or survive as well as pig strains in the human gut remains to be elucidated. More research on the strain-specific traits of strains derived from different human and animal sources, such as virulence factors and colonization and persistence abilities, is warranted, as they may play a crucial role in the transmission and survival of resistant bacteria in the human gut. Backmatter: PMID- 12183234 TI - Antibiotic Resistance Genes and Salmonella Genomic Island 1 in Salmonella enterica Serovar Typhimurium Isolated in Italy AB - Fifty-four epidemiologically unrelated multidrug-resistant Salmonella enterica serovar Typhimurium isolates, collected between 1992 and 2000 in Italy, were analyzed for the presence of integrons. Strains were also tested for Salmonella genomic island 1 (SGI1), carrying antibiotic resistance genes in DT104 strains. A complete SGI1 was found in the majority of the DT104 strains. Two DT104 strains, showing resistance to streptomycin-spectinomycin and sulfonamides, carried a partially deleted SGI1 lacking the flost, tetR, and tetA genes, conferring chloramphenicol-florfenicol and tetracycline resistance, and the integron harboring the pse-1 gene cassette, conferring ampicillin resistance. The presence of SGI1 was also observed in serovar Typhimurium strains belonging to other phage types, suggesting either the potential mobility of this genomic island or changes in the phage-related phenotype of DT104 strains. Keywords: Introduction : Salmonella enterica is one of the most common pathogens causing food-borne infections in Italy . The epidemic strain serovar Typhimurium definitive phage type 104 (DT104) has been identified as a major cause of salmonellosis in humans and animals, in both Europe and the United States . The majority of the DT104 isolates are characterized by resistance to six drugs: ampicillin (A), chloramphenicol (C), streptomycin (S), spectinomycin (Sp), sulfonamides (Su), and tetracycline (T) . Between 80 and 90% of isolates of DT104 show this resistance type . The resistance mechanism in ACSSpSuT-type DT104 has been recently elucidated, and resistance genes have been described, all located within the same chromosomal locus, designated Salmonella genomic island 1 (SGI1) . SGI1 is a 43-kb genomic island, showing 44 coding sequences, most of them encoding hypothetical proteins. SGI1 shows two 18-bp direct repeats at the external boundaries, which strongly supports the hypothesis that site-specific recombination events may have driven the insertion of the island within the serovar Typhimurium chromosome . The ACSSpSuT resistance genes have been identified within a 14-kb region in SGI1, carried by two class 1 integrons, InC and InD, encoding the aminoglycoside resistance gene aadA2 and the beta-lactamase pse-1 gene, respectively . The intervening region, encompassing the two integrons, contains the floR gene, conferring resistance to florfenicol-chloramphenicol, and the tetR and tetA (class G) genes, conferring tetracycline resistance . InC and InD show a peculiar structure with respect to other class 1 integrons . InC carries the qacEDelta1 gene, conferring resistance to disinfectants, and a truncated, nonfunctional sul1 gene in the 3'-conserved segment (3'-CS). InD shows a deletion of the integrase gene (intI1) which is linked to the groEL gene in the 5'-conserved segment (5'-CS) ; the qacEDelta1 gene, the sul1 gene, conferring sulfonamide resistance, and the two open reading frames of unknown function, ORF5 and ORF6, are contained in the 3'-CS followed by the IS6100 element . The multidrug resistance region in SGI1 is bounded by characteristic 25-bp inverted repeats and flanked by a direct duplication of 5 bp of the target sequence, indicating that it was inserted by a transpositional mechanism (EMBL accession no. ) . The aim of our study was to analyze the molecular basis of antibiotic resistance by searching for, identifying, and characterizing integrons and other resistance genes in multidrug-resistant serovar Typhimurium isolated in Italy from humans and food animals. Strains were also analyzed for the presence and conservation of SGI1. MATERIALS AND METHODS : Bacterial strains. | Fifty-four apparently epidemiologically unrelated serovar Typhimurium isolates of human and animal origin, representative of frequent phage types and recurrent multidrug resistance profiles in Italy , were chosen from the collections of the Istituto Zooprofilattico delle Venezie (Padua, Italy) and of the Istituto Superiore di Sanita (Rome, Italy). Serotypes were determined with anti-O and anti-H antisera obtained from Behringwerke AG (Marburg, Germany). Phage types were determined according to the work of Callow , and antibiotic resistance was determined by the disk diffusion assay on Mueller-Hinton agar with commercial antimicrobial susceptibility disks (Oxoid, Basingstoke, United Kingdom; Becton Dickinson Microbiological Systems, Cockeysville, Md.), according to the recommendations of the National Committee for Clinical Laboratory Standards . The following antibiotics were tested: ampicillin, ceftazidime, chloramphenicol, streptomycin, spectinomycin, sulfonamides, tetracycline, trimethoprim (Tp), kanamycin (K), gentamicin (G), tobramycin (N), amikacin, nalidixic acid, and ciprofloxacin. Strain ST30 was assigned to PT (phage type) U302 by John E. Threlfall (Public Health Laboratory Service, Colindale, London, United Kingdom). Preparation of total DNA and Southern blot hybridization. | Small-scale DNA preparations were made with 3 ml of bacterial liquid cultures grown overnight, as described by Ezaki and colleagues . Four micrograms of total DNA was digested with PvuII-BamHI or XbaI restriction enzymes. Restricted fragments were separated by 1% agarose gel electrophoresis and transferred onto positively charged nylon membranes (Roche Diagnostics, Monza, Italy) by standard methods . Southern blot hybridization was carried out under high-stringency conditions . A specific probe for the intI1 gene was obtained as previously described . The S014 and the PCR8 probes were obtained by PCR amplification with S014-FW-S014-RV and S044-FW-DR-RV primer pairs, respectively . DNA probes were [alpha-32P]dCTP labeled with a random priming kit (Life Technologies, Milan, Italy). PCR amplification, cloning, and sequencing. | Standard PCR amplifications were performed with primers listed in Table and 2.5 U of Taq DNA polymerase (Roche Diagnostics), according to the manufacturer's recommendations. All PCR amplifications were run at 94C for 30 s, 55C for 30 s, and 72C for 3 min, for a total of 30 cycles. Amplification products were sequenced with fluorescent dye-labeled dideoxynucleotides and a 373 automatic DNA sequencer (Perkin-Elmer, Foster City, Calif.). Comparative analysis of nucleotide sequences was performed by the advanced BLAST search program 2.0 within the QBLAST system at the National Center for Biotechnology Information site . Pulsed-field gel electrophoresis (PFGE). | Preparation of total DNA was performed as described by R. K. Gautom except that 0.5 mg of proteinase K per ml was dissolved in the bacterial suspension before addition of 1% melted agarose containing 1% sodium dodecyl sulfate. DNA plugs were digested with XbaI (Roche Diagnostics). Electrophoresis was performed in a CHEF-DRII electrophoresis system (Bio-Rad Laboratories) at 14C with 0.5x Tris-borate-EDTA running buffer and 1% pulsed-field certified agarose (Bio-Rad Laboratories). Electrophoresis conditions were as follows: phase 1, initial switch time, 0.5 s; final switch time, 60 s for 24 h, with 6 V cm-1. For visualization, gels were stained in ethidium bromide, destained in water, and photographed. TABLE 1 | Primers used in this study RESULTS : Detection of integrons in multidrug-resistant serovar Typhimurium isolated in Italy. | Forty-one isolates of DT104, 33 of resistance type ACSSpSuT and 8 with different resistance profiles, and 13 resistant isolates of other phage types were analyzed for the presence of class 1 integrons. The integron search was performed by Southern blot hybridization with the intI1 gene probe , on total DNA digested with PvuII-BamHI restriction enzymes. This hybridization is expected to produce, in DT104 strains carrying SGI1, two integrase-positive restriction fragments of 7,738 bp (band A) and 4,402 bp (band B) . Band A contains InC and the floR, tetA, and tetR resistance genes, from the PvuII site at position 27276 to the PvuII site at position 35014 (EMBL accession no. ). Band B contains InD from the PvuII site at position 35546 to the BamHI site at position 39948 in the SGI1 DNA sequence . As expected, all ACSSpSuT DT104 strains showed both bands, indicating that the SGI1 region containing the two integrons was well conserved . Interestingly, the A and B bands were also observed in a strain belonging to DT1 (ST9) and in a PT U302 strain (ST30), both showing the ACSSpSuT resistance profile . Three DT104 strains showing resistance types SSpSu (ST2 and ST22) and ASSpSuT (ST48) were positive for a PvuII-BamHI fragment of approximately 3,700 bp . This band indicates the presence of an integron located within a different genetic structure from that of the DT104 SGI1. Integron-positive plasmids were not found in these strains, suggesting that integrons carried by ST2, ST22, and ST48 are located on the bacterial chromosome. Strain ST4, showing the ASSpSuGKTp resistance type, was positive for two bands of approximately 4,300 and 800 bp, suggesting the presence of two integrons. All the other serovar Typhimurium strains were negative for the presence of integrons . Antibiotic resistance genes. | Integron-borne gene cassettes were amplified with the 5'-CS and 3'-CS primer pair . All DT104 strains showing the ACSSpSuT resistance profile produced two PCR products of 1,008 and 1,133 bp. By DNA sequencing, these bands corresponded to the well-characterized aadA2 and pse-1 gene cassettes of InC and InD as previously described . The same gene cassettes were amplified and sequenced from strains ST9 (DT1) and ST30 (U302). Strains ST2, ST22, and ST48 produced an amplicon of about 1,000 bp. These amplicons were fully sequenced. The nucleotide sequence of the ST48 integron revealed the presence of the aadA1 gene cassette, encoding streptomycin-spectinomycin resistance. This gene cassette was 99.9% identical, without amino acid changes, to the aadA1 gene cassette identified within the In2 integron carried by the Tn21 transposon (EMBL accession no. ) . Nucleotide sequences of amplicons obtained from both ST2 and ST22 strains showed the presence of the aadA2 gene cassette, encoding streptomycin-spectinomycin resistance. These gene cassettes were 100% (ST2) and 99.9% (ST22) identical to the aadA2 gene cassette carried by InC in SGI1 (EMBL accession no. ). In the ST22 DNA sequence, a single point mutation at position 28211 was identified, resulting in an amino acid change (from glutamic acid to lysine) in the deduced protein sequence. The same substitution was previously described for the DT104 strain H3380 (EMBL accession no. ) . Two PCR products of 2,000 and 850 bp were obtained from strain ST4. The DNA sequence of the 850-bp amplicon revealed the presence of the aadB gene cassette, conferring resistance to kanamycin and gentamicin, while the DNA sequence of the 2,000-bp amplicon revealed the integron-borne oxa1 and aadA1 gene cassettes, conferring ampicillin and streptomycin-spectinomycin resistance, respectively. Serovar Typhimurium strains were then tested for strA, a streptomycin resistance gene that is an alternative to the integron-borne aadA gene cassette . Positive PCR amplifications with the strAF-strAR primer pair were obtained from 21 of the 54 strains , including the integron-negative DT104 strains ST25, ST38, ST10, ST44, and ST49 and the ampicillin-, tetracycline-, and chloramphenicol-susceptible DT104 strains ST2, ST22, and ST48. Analysis of SGI1. | In order to gather more information on the presence and conservation of SGI1 in the DT104 strains, Southern blot hybridization and PCR amplification experiments were performed. SGI1 was initially sought by using Southern blot hybridization with the intI1 probe on total DNA restricted with XbaI, in DT104 strains ST2 and ST22 lacking the A and B PvuII-BamHI bands, and in the ST9 strain (DT1), which tested positive for the presence of SGI1. ST1 and ST13 were used as positive controls and ST25, ST38, and ST3 were used as negative controls in these experiments. As shown in Fig. , ST1 and ST13 produced an integrase-positive band of 11,758 bp (band C in Fig. ), as expected from the XbaI sites located at positions 26038 and 37796 in the SGI1 DNA sequence. This band was also observed in ST9, while ST2 and ST22 showed a different band of approximately 7,000 bp (band E in Fig. ). As expected, band C was missing in the three strains lacking integrons (ST25, ST38, and ST3). The XbaI Southern blot was then hybridized with the S014 probe, located in the left arm of SGI1. This probe recognized the 8,921-bp XbaI restriction fragment (band D in Fig. ) in both ST2 and ST22 as well as in ST1, ST13, and ST9 , revealing that this region of SGI1 was well conserved in these strains. Finally, ST2, ST22, and ST1 BamHI-restricted DNAs were hybridized with the PCR8 amplification product as probe . This probe recognized in both ST2 and ST22 as well as the ST1 positive control strain the 4,699-bp BamHI restriction fragment containing ORF5, ORF6, IS6100 of InD, and part of the retron phage (data not shown). To fully characterize the conservation and structure of SGI1 in ST2 and ST22, PCR was performed with primer pairs listed in Table (PCR amplifications relative to the SGI1 map are illustrated in Fig. ). PCR experiments performed on the left (PCR1) and right (PCR8) direct repeat junctions of SGI1 and on other SGI1 regions (PCR2, PCR3, and PCR4) demonstrated that ST2 and ST22 produced indistinguishable bands with respect to the ST24 control strain . In particular, PCR4, obtained with primers S026-FW and int-RV, demonstrated that integrons in ST2 and ST22 were located within SGI1, linked to the S026 DNA sequence. However, PCR amplifications with flo-FW-flo-RV and tetA-FW-tetR-RV primer pairs (PCR5 and PCR6) were negative in ST2 and ST22, demonstrating that these strains do not carry the floR and the tetA and tetR genes, respectively (data not shown). Finally, ST2 and ST22 produced a PCR product of approximately 2,000 bp with the aadA2-FW-ORF5-RV primer pairs designed on the aadA2 gene cassette and on the ORF5 DNA sequence, respectively (PCR7, Fig. ). This amplicon is about 10 kb shorter than the expected amplicon for complete SGI1 (ca. 12 kb). Southern blot hybridization and PCR amplification results demonstrate that ST2 and ST22 harbor only one of the two integrons of SGI1; this integron (InC1 in Fig. ) carries the same aadA2 gene cassette of InC and shows a 3'-CS identical to that of InD, carrying qacEDelta1, conferring resistance to disinfectants; the entire sul1 gene, conferring sulfonamide resistance; ORF5; ORF6; and the IS6100 element . PCR amplification (PCR1 to PCR8) was then performed on the remaining isolates, demonstrating that, with the exception of strain ST37, all ACSSpSuT DT104 strains carried an intact SGI1 . Strain ST37 was negative for PCR1 and PCR2 but positive for the other PCR amplifications performed on SGI1, suggesting the partial deletion of the left arm of the genomic island. An intact SGI1 was identified in ST9 (DT1) and in ST30 (U302), while it was completely absent in ST48 (DT104), which harbors the aadA1 integron-borne gene cassette. PFGE of macrorestricted genomic DNA. | Strains ST2 (DT104), ST22 (DT104), ST9 (DT1), and ST30 (U302) were analyzed by PFGE following digestion with XbaI. Strain ST24 was used as the DT104 reference control strain. The results showed that the three DT104 strains were highly related at the chromosomal level. ST30 (U302) showed a DT104-related PFGE pattern differing by three bands . PT U302 has been previously shown to be closely related to DT104, and phage conversion of DT104 to U302 has been observed, possibly by the acquisition of plasmids . The presence of a plasmid in ST30 was revealed by PFGE of undigested DNA , suggesting that this plasmid could have caused a DT104-U302 phage conversion in this strain. In contrast, ST9 (DT1) was divergent at the chromosomal level from DT104, differing by at least four bands in the PFGE pattern . In this case the presence of the DT104 resistance traits could be explained by the horizontal transfer of SGI1. FIG. 1. | Schematic representation of the entire and partially deleted SGI1. Schematic representation of the entire and partially deleted SGI1. (A) XbaI (X), BamHI (B), and PvuII (P) restriction map was deduced from the DT104 DNA sequence released under EMBL accession no. . (B) Restriction map of the partially deleted SGI1 was experimentally determined. White boxes represent intI1, qacEDelta1, sul1, floR, tetR, tetA, ORF5, and ORF6 genes. The IS6100 position is indicated between two arrows. The shaded boxes represent gene cassettes inserted into the InC and InD integrons. Black boxes are the 59-base element sites, and DR-R and DR-L boxes represent the direct repeat junctions of SGI1 in the Salmonella chromosome (dotted lines). Thin bars represent Southern blot hybridization bands A, B, C, and D. Heavy lines represent PCR amplicons (PCR1 to PCR8). The positions of the intI1 gene and S014 probes are indicated. FIG. 2. | Southern blot hybridization of serovar Typhimurium strains by the integrase gene. Southern blot hybridization of serovar Typhimurium strains by the integrase gene. Total DNAs restricted with BamHI-PvuII were separated by 1% agarose gel electrophoresis, blotted onto nylon membranes, and hybridized with an intI1 gene-specific probe . Numbers above lanes indicate the designation number for each strain. The positions of bands A and B are indicated. Molecular size standards are shown on the right. FIG. 3. | Southern blot hybridization of the SGI1 island. Southern blot hybridization of the SGI1 island. Total DNAs restricted with XbaI were separated by 0.8% agarose gel electrophoresis, blotted onto nylon membranes, and hybridized with the intI1 gene probe and with the S014 probe. Numbers above lanes indicate the designation number for each strain. The positions of bands C, D, and E are indicated. Molecular size standards are shown on the right of the figure. FIG. 4. | PCR analysis of SGI1. PCR analysis of SGI1. PCR1 to PCR8 were performed on the ST2 and ST22 strains. Strain ST24 was used as DT104 positive control. The amplification product PCR7 (ca. 12 kb) was not obtained from the ST24 DT104 control strains, since no long-run PCR conditions were used in these experiments. A 1-kb marker (KiloBase DNA marker; Pharmacia Biotech, Milan, Italy) was used as a standard. FIG. 5. | Analysis by PFGE. Analysis by PFGE. Serovar Typhimurium DNAs were digested by XbaI and separated by PFGE on a 1% agarose gel. Designation numbers of tested strains are reported above each lane. Lane M, the molecular size lambda ladder ranged from 48.5 to 1,000 kb (Roche Diagnostics). ND, not digested. TABLE 2 | Characteristics of the 54 serovar Typhimurium isolates tested in this study DISCUSSION : Several studies have focused on resistance genes in DT104 strains, and new molecular methods for detection and analysis of the DT104 clone have been proposed elsewhere . Most of these methods are based on PCR amplification of DT104-related traits, such as the 1.0- and 1.2-kb integrons or the floR resistance gene located within the resistance gene cluster . More recently, the cloning and sequencing of the entire genomic island from multidrug-resistant DT104 have opened new possibilities for investigating genetic characteristics of this important and widely diffused Salmonella clone. However, complete information is not available for the conservation of SGI1 among strains circulating in animals and humans, and the frequency of transfer of the genomic island to serovar Typhimurium strains of other phage types has not been quantified. This information could be important to understanding the origin and evolution of the genomic island and potential horizontal mobility of SGI1. To study antibiotic resistance genes carried by integrons and to investigate the presence of SGI1, multidrug-resistant serovar Typhimurium strains of different phage types of both human and animal origin were analyzed by both PCR amplification and Southern blot hybridization. All DT104 strains showing the ACSSpSuT resistance profile harbor the two integrons InC and InD, located within SGI1. Two DT104 strains, isolated in Italy in 1997 (ST2) and in 1998 (ST22) and showing the SSpSu resistance profile, revealed the presence of a partially deleted SGI1. These isolates lack the region of the island encoding resistance to chloramphenicol-florphenicol and tetracycline and also the InD integron carrying the pse-1 ampicillin resistance gene. These two DT104 strains carried the InC1 integron, which shows a 3'-CS, including the qacEDelta1, sul1, ORF5, ORF6, and IS6100 DNA sequences. This integron was found in an XbaI integrase gene-positive band smaller than that observed in DT104 control strains. The ST2 and ST22 resistance island may represent the precursor of the DT104 resistance gene cluster in which InC1 is the unique integron harbored by the genomic island. Our findings suggest that the assembly of the resistance gene cluster within SGI1 could be due to sequential acquisition of resistance determinants. In fact, InC could derive from InC1 by deletion of part of the 3'-CS caused by the insertion of the floR and tet genes. The second integron, InD, could also have been acquired through an independent integration event, leading to the full assembly of the resistance gene cluster. The deleted SGI1 could then represent the precursor of the genomic island before the assembly of InD, floR, and tet genes. However, the genetic structure observed in ST2 and ST22 could also be explained by homologous recombination between InC and InD in the 3'-CS. Recombination between the two sul1 genes could lead to the deletion of a 10-kb DNA region containing the floR, tet, intI1, and pse-1 genes and to the reconstruction of InC1 carrying the 3'-CS of InD. A different deletion in the genomic island was observed in the ST37 ACSSpSuT DT104 strain, lacking the left arm of SGI1. These results demonstrate that several rearrangements could occur in SGI1, suggesting a constant and dynamic evolution of this genetic trait. In another DT104 strain (ST48) an integron that carried the aadA1 gene cassette was found to be located in the chromosome. This strain lacked SGI1, demonstrating that integrons other than InC and InD can be acquired by DT104 strains. Most of the DT104 strains showing different resistance profiles than the ACSSpSuT resistance type lack SGI1 but contain the strA resistance gene, conferring streptomycin resistance. ST2, ST22, and ST48, as well as many DT104 strains harboring the entire SGI1, also carried the strA gene, suggesting that multiple mechanisms of resistance can be simultaneously present in these strains. Of note, the entire SGI1 was found in the chromosome of ST9 (DT1) and in the ST30 (U302) isolate, suggesting the potential capacity of mobilization of SGI1. However, recent studies have demonstrated that strains of U302, DT120, and DT12 of the ACSSpSuT resistance type are probably derived from DT104 by a change of phage sensitivity . This possibility should be taken into consideration. The presence of the DT104 resistance genes has also been previously reported for several serovar Agona strains . However, the multiresistant serovar Agona strains lacked the retron phage at the DR-R boundary of SGI1, while results with serovar Typhimurium strains ST2 and ST22 indicate that in these strains SGI1 is located between the thdf gene and the retron phage, a situation similar to that described for the multiresistant DT104 strain. There is no experimental evidence demonstrating the molecular mechanism of the SGI1 horizontal transfer among Salmonella strains, although transduction experiments with a P22-like phage demonstrated a facilitated transduction of resistance genes to a susceptible Salmonella strain . Multidrug resistance in serovar Typhimurium is now a cause of great concern in both clinical and veterinary medicine. Studies of evolution and dissemination of resistance determinants may help us to better understand the origin of such strains and the transmission from food animals to humans. Backmatter: PMID- 12183266 TI - In Vitro Interaction of Caspofungin Acetate with Voriconazole against Clinical Isolates of Aspergillus spp. AB - The interaction between caspofungin acetate and voriconazole was studied in vitro by using 48 clinical Aspergillus spp. isolates obtained from patients with invasive aspergillosis. MICs were determined by the NCCLS broth microdilution method. Synergy, defined as a fractional inhibitory concentration (FIC) index of <1, was detected in 87.5% of the interactions; an additive effect, defined as an FIC index of 1.0, was observed in 4.2% of the interactions; and a subadditive effect, defined as an FIC index of 1.0 to 2.0, was found in 8.3% of the interactions. No antagonism was observed. Animal models are required to validate the in vivo significance of these in vitro data presented for the combination of caspofungin and voriconazole. Keywords: Introduction : Aspergillus spp. are common causes of nosocomial pneumonia and disseminated infection in immunocompromised hosts such as bone marrow transplant recipients, patients with hematologic malignancies, solid-organ transplant recipients, AIDS patients, and patients with pulmonary diseases. The most common species include Aspergillus fumigatus (approximately 90% of the cases), A. flavus, A. niger, A. terreus, and A. nidulans . The rate of mortality from invasive aspergillosis remains high in severely immunosuppressed patients who receive standard antifungal treatment with amphotericin B. Improved responses have been observed in less immunosuppressed patients sequentially receiving amphotericin B followed by itraconazole and those receiving itraconazole alone . Nevertheless, these drugs are not always effective or tolerated in severely ill patients, and therefore, there is still a need for new therapies and new approaches to improve the outcome of this disease. The echinocandin caspofungin acetate (CAS), formerly MK-0991, belongs to a new class of antifungal drugs that inhibit the synthesis of 1,3-beta-d-glucan, an essential cell wall polysaccharide that represents a selective target present only in fungal cell walls. It has been shown to have in vitro and in vivo activities against many clinically important fungi, including Candida and Aspergillus spp., and has recently been approved for the treatment of refractory invasive aspergillosis (-, , ; A. M. Flatery, P. Scott Hicks, A. Wilcox, and H. Rosen, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 936, p. 369, 2000). Voriconazole (VRC) is an investigational triazole antifungal agent that is similar in structure and spectrum of action to fluconazole and itraconazole, respectively. This agent has demonstrated substantial activity in both in vitro and in vivo models against a variety of fungi such as dimorphic fungi, yeasts, and also opportunistic filamentous fungi, including Aspergillus spp. . It has previously been described that combinations of CAS and amphotericin B against Candida, Cryptococcus, Aspergillus, and Fusarium have additive to synergistic effects against certain isolates, with no evidence of antagonism . In the case of VRC, the combination of CAS and terbinafine has been shown to have synergistic activity against Aspergillus sp., Candida albicans, and Candida glabrata isolates (, ; S. Perea, G. Gonzalez, A. W. Fothergill, D. A. Sutton, and M. G. Rinaldi, Abstr. 10th Annu. Focus Fungal Infect., abstr. 28, p. 69, 2000). The aim of the present study was to investigate the in vitro interaction of CAS and VRC against 48 isolates of Aspergillus spp. isolated from patients with invasive aspergillosis. (This work was presented in part at the 101st General Meeting of the American Society for Microbiology, Orlando, Fla., 2001.) Forty-eight isolates of Aspergillus spp. (24 A. fumigatus, 10 A. terreus, 9 A. flavus, and 5 A. niger isolates) were used throughout the study. All strains evaluated were clinical isolates submitted to the Fungus Testing Laboratory, University of Texas Health Science Center at San Antonio, from patients with invasive aspergillosis. The identities of the clinical isolates were confirmed by standard microbiological procedures, and the strains were stored in sterile deionized water at room temperature until they were used in the study. Paecilomyces variotii (UTHSC 90-459) was used as the control organism in all experiments. CAS (Merck, Rahway, N.J.) and VRC (Pfizer Inc., Sandwich, United Kingdom) were obtained in reagent-grade powder form from their respective manufacturers. Stock solutions were prepared in water (CAS) and polyethylene glycol 400 (VRC). Serial twofold dilutions of each antifungal agent were prepared as outlined in document M38-P of the NCCLS . Final dilutions were made in antibiotic medium 3 (M3; Difco, Detroit, Mich.) supplemented with 2% glucose (M3-2%). Drug interactions were assessed by a checkerboard microdilution method that also included the determination of the MIC of each drug alone by using the parameters outlined in the recommendations of the NCCLS . The final concentrations of the antifungal agents ranged from 0.25 to 64 mug/ml for CAS and 0.03 to 4 mug/ml for VRC. Inocula were prepared spectrophotometrically and further diluted in order to obtain final concentrations ranging from 0.4 x 104 to 5 x 104 CFU/ml. Each microdilution well containing 100 mul of the diluted (two times) drug concentrations of both antifungals (CAS and VRC) was inoculated with 100 mul of the diluted (two times) inoculum suspension (final volume of each well, 200 mul). The trays were incubated at 35C, and the results were read at 48 h visually and spectrophotometrically with a spectrophotometric microtiter plate reader (Dynex Technologies Inc, Chantilly, Va.). MIC endpoints were determined as the first concentration of the antifungal agent, either alone or in combination, at which the turbidity in the well was less than 80% of that in the control well. The geometric mean (GM) MICs and the ranges of MICs were analyzed to evaluate the in vitro activities of both drugs, alone and in combination. Both on-scale and off-scale results were included in the analysis. For computation of the GM values, high off-scale MICs were converted to the twofold concentration just above the highest concentration tested. When the MIC was off the bottom of the scale, the MIC was assumed to be the lowest MIC tested. Drug interactions were classified as synergistic, additive, or antagonistic on the basis of the fractional inhibitory concentration (FIC) index. The FIC index is the sum of the FICs of each of the drugs and is defined as the MIC of each drug when used in combination divided by the MIC of the drug when used alone. The interaction was defined as synergistic if the FIC index was <1, additive if the FIC index was 1.0, subadditive if the FIC index was between 1.0 and 2.0, indifferent if the FIC index was 2, and antagonistic if the FIC index >2. Synergy was further subclassified as marked (FIC index, <=0.50) and weak (FIC index, between 0.50 and 1.0) . The reductions in the GM MICs of the antifungals when they were given in combination compared to the MICs of the drugs when they were given alone were compared by a paired rank test, a nonparametric test for comparison between two related samples. A P value of <0.05 was considered significant. The MICs of CAS for the 48 clinical Aspergillus spp. isolates ranged from <0.125 to >64 mug/ml, (GM MIC, 30.2 mug/ml; MIC at which 50% of isolates are inhibited [MIC50], 32 mug/ml; MIC90, >64 mug/ml). There were marked differences in the activities of CAS against the different species, with A. flavus being the least susceptible (GM CAS MIC, 118.5 mug/ml) and A. niger being the most susceptible (GM CAS MIC, 1.3 mug/ml). VRC MICs ranged from 0.125 to 4 mug/ml (GM MIC, 0.66 mug/ml; MIC50, 0.5 mug/ml; MIC90, 1 mug/ml) . The MICs determined spectrophotometrically were similar to the MICs determined visually. The MICs calculated for the controls were within the acceptable ranges for the two drugs tested. When given in combination, significant reductions in the GM CAS MICs (30.2 to 1.31 mug/ml [P < 0.001]) and VRC MICs (0.66 to 0.16 mug/ml [P < 0.001]) for the clinical isolates were observed. For the combination, the MIC50s and MIC90s were reduced from 32 and >64 to 0.5 and 16 mug/ml, respectively, for CAS and from 0.5 and 1 to 0.25 and 0.25, respectively, for VRC. Synergistic effects were observed in 87.5% (42 of 48) of the interactions. Of those, 45.2% showed marked synergism and 54.8% showed weak synergism. Additive effects were found in 4.2% (2 of 48) of the interactions. Subadditive effects were observed in 8.3% (4 of 48) of the interactions. No antagonism was observed. TABLE 1 | Mode of interaction between CAS and VRC against 48 isolates of Aspergillus spp. We have explored the interaction between CAS in combination with the new triazole VRC against clinical Aspergillus spp. isolates from patients with invasive aspergillosis. A reproducible and clinically relevant method for susceptibility testing of echinocandins has not been fully established yet. Different media (RPMI, M3, and M3-2%) as well as different MIC endpoints for measuring the in vitro activities of these new class of antifungals (MIC50, MIC80, MIC100, and minimum effective concentration) have been published previously . Because of the technical variability, data reported thus far on the activity of CAS against Aspergillus show a wide range of MICs. Pfaller et al. used the NCCLS microdilution methodology and reported MIC90s after 72 h of incubation of 0.12 mug/ml, using an MIC endpoint of a 75% reduction in growth (MIC75). Del Poeta et al. used the macrodilution methodology proposed by the NCCLS for yeasts (M27-A) , an inoculum size of 0.5 x 103 to 2.5 x 103 CFU/ml, and an MIC80 endpoint. The GM MICs for A. fumigatus and A. flavus were <=0.09 and 0.2 mug/ml, respectively . Espinel-Ingroff used the MIC50 and found GM MICs for A. fumigatus, A. flavus, and A. terreus of 2.15, 0.5, and 0.5 mug/ml, respectively. Arikan et al. reported GM MICs, obtained after 48 h of incubation with M3 medium and by the NCCLS M38-P methodology , of 3.7, 2.83, 0.26, and 0.25 mug/ml for A. flavus, A. fumigatus, A. niger, and A. terreus, respectively. Because of the lack of standardization in the methodology used to perform in vitro antifungal susceptibility testing for CAS, we arbitrarily chose to use M3-2% medium and to calculate the endpoint as the lowest drug concentration that showed a significant reduction of growth (approximately 80% reduction). For the other variables (inoculum preparation, preparation of drugs, broth inoculation, and incubation) the NCCLS M38-P methodology was followed . The MICs obtained in the previous studies are lower than the ones obtained in the present study. The CAS MICs obtained in the present study support and extend the findings presented in a previous report of a study which evaluated the activity of CAS against Aspergillus spp. by the same methodology, that is, with M3-2% medium, with reading of the MIC endpoint as the lowest drug concentration that showed a significant reduction of growth (approximately 80% reduction), and by the NCCLS methodology (M38-P) (Flatery et al., 40th ICAAC). In that study, the MIC50 ranges for A. fumigatus, A. flavus, A. niger, and A. terreus were 32, >64, <=0.03, and 32, respectively. With respect to the VRC MICs, our results are similar to those published recently by Espinel-Ingroff when M3-2% medium, reading of the MIC endpoint as MIC50, and the NCCLS M38-P methodology were used. The results obtained in the present study with a combination of agents, which showed either synergy or additivity but no evidence of antagonistic effects when both agents were used in combination, are very encouraging. Although the exact mechanism of the interaction between the two agents is unknown, one theoretical explanation for such an effect would be the simultaneous disruption of the fungal cell membrane by VRC and disruption of the cell wall by CAS, ultimately decreasing the cell stability and leading to the death of the fungal cell. In conclusion, our results indicate that a combination of CAS and VRC might be effective against infections caused by Aspergillus spp. However, improvements to the methodology used to determine the in vitro susceptibility of the fungi to CAS, as well as in vitro-in vivo correlation studies, are required before clinical studies can be conducted. Backmatter: PMID- 12183261 TI - Pharmacokinetic and Pharmacodynamic Profiles of Danofloxacin Administered by Two Dosing Regimens in Calves Infected with Mannheimia (Pasteurella) haemolytica AB - The pharmacokinetics and pharmacodynamics of danofloxacin in calves with induced Mannheimia (Pasteurella) haemolytica pneumonia were evaluated. Calves received either saline as an intravenous (IV) bolus or danofloxacin (0.738 mg/kg of body weight) administered as either a single IV bolus or a 36-h continuous IV infusion. Blood samples and bronchial secretions were collected before and at predetermined times over 48 h following the start of treatment. Calves were assessed clinically throughout, and lung consolidation was assessed at necropsy. Bronchial secretions and lung tissue were cultured for M. haemolytica. Bolus administration of danofloxacin produced a high maximum drug concentration-to-MIC ratio (Cmax:MIC) of 14.5 and a time period of 9.1 h when plasma danofloxacin concentrations exceeded the MIC (T>MIC). Following danofloxacin infusion, the Cmax:MIC was low (2.3), with a long T>MIC (33.3 h). The area under the curve-to-MIC ratios were 43.3 and 49.1 for the bolus and infusion administrations, respectively. The single bolus of danofloxacin was more effective than the same dose administered by continuous infusion, as indicated by a significantly lower (P < 0.05) number of animals with M. haemolytica in bronchial secretions after treatment and lower rectal temperatures in the 24 h after the start of treatment. Thus, danofloxacin exhibited concentration-dependent antimicrobial activity in cattle with respiratory disease caused by M. haemolytica. Keywords: Introduction : Danofloxacin is a fluoroquinolone antimicrobial drug with rapid bactericidal activity against a broad range of pathogens responsible for a number of disease syndromes of economic importance in the commercial rearing of livestock . Since their introduction in the late 1980s, fluoroquinolones have been shown by a number of studies to exhibit concentration-dependent bactericidal activity, whereby the optimal effect is attained by the administration of high doses over a short period . This is a property shared by the aminoglycosides but is in contrast to the predominantly time-dependent bactericidal action shown by the beta-lactam antibiotics , where the time that bacteria are exposed to antimicrobial concentrations exceeding the MIC (T>MIC) is the major determinant of efficacy. These different types of action have been confirmed for danofloxacin and amoxicillin in an in vitro pharmacodynamic model against Actinobacillus pleuropneumoniae . The purpose of this study was to establish the pharmacokinetic and pharmacodynamic properties of danofloxacin in vivo by using an experimental model of calf pneumonia and to determine whether the concentration-dependent activity of danofloxacin in cattle operates under simulated clinical conditions. A fixed equal total dose of danofloxacin was administered either as a single intravenous (IV) bolus or by continuous infusion over a 36-h period, and the clinical and bacteriological outcomes in calves with induced infections of Mannheimia (Pasteurella) haemolytica were compared. The study was conducted in compliance with Good Clinical Practice guidelines , the analysis of samples was conducted in accordance with Good Laboratory Practice guidelines , and the husbandry of all animals was in compliance with the requirements of national legislation and local animal welfare guidelines. The study was conducted under veterinary supervision, with veterinary attention available at all times. MATERIALS AND METHODS : Animals. | Thirty-three male Friesian calves (approximately 11 to 13 weeks of age, with initial body weights of 66.5 to 106 kg) were enrolled in the study and inoculated with M. haemolytica. Prior to enrollment, the calves were free from preexisting medical or surgical conditions and had no history of previous respiratory disease. Following enrollment and prior to inoculation, the calves were allocated randomly to pens and to treatment groups by using an incomplete block design. The calves were housed in straw-bedded pens in a self-contained naturally ventilated calf rearing unit with a common airspace, but divided by solid partitions approximately 1.4-m high to prevent nasal contact between animals, and with 4.4 m2 of floor space per calf. Water was supplied ad libitum, and the calves were maintained on an antibiotic-free concentrate diet following weaning at 6 to 7 weeks of age. On arrival at the study site, at approximately 1 week of age, each calf received an intramuscular injection (20 mg/kg of body weight) of long-acting oxytetracycline (Terramycin LA; Pfizer Ltd., Sandwich, United Kingdom). There was no further antibiotic administration prior to enrollment and treatment with the test materials. Inoculum. | The inoculum was prepared by transferring a 1.0-ml aliquot of M. haemolytica type A1 (reference M7/2) into 9.0 ml of Oxoid nutrient broth no. 2. After incubation at 37C for 16 to 18 h, the starter culture was inoculated into 290 ml of nutrient broth, shaken at 150 rpm, and incubated for 4 h to provide 300 ml of culture. Following incubation, the approximate viable count (CFU per milliliter) was determined by using a McFarland scale. Just prior to inoculation, the broth culture was diluted in 1-liter volumes of sterile phosphate-buffered saline (0.01 M, pH = 7.4, prewarmed to 37C) to give an inoculum with an approximate viable count of 3.3 x 105 CFU/ml. The titer of the inoculum was confirmed pre- and postinoculation by culture on blood agar, following serial 10-fold dilutions, and the CFU per milliliter were calculated by multiplying the number of colonies by the relevant dilution factor. Design. | Animals selected for the study were each inoculated by endobronchial deposition (calves were conscious) over a period of approximately 1 min with 300 ml of the inoculum, representing an inoculum per animal of approximately 108 CFU of M. haemolytica type A1/calf (acceptable range was defined as 5 x 107 to 5 x 108 CFU/calf). A fiber-optic endoscope sterilized with ethylene oxide prior to the start of inoculation was inserted nasally and passed via the nasopharynx into the trachea. At the tracheal bifurcation, the endoscope was pushed approximately 10 cm into the principal bronchus where the inoculum was deposited. Respiratory rates were assessed hourly from approximately 3 h after the inoculation of the first calf. When the respiratory rates of over 72% of the inoculated calves had doubled from those recorded immediately prior to inoculation, pretreatment samples (blood and bronchial secretion) were collected and the animals were administered their allocated treatment. Prior to administration, the commercial danofloxacin 18% formulation (Advocin 180; Pfizer Ltd.) was diluted with 0.9% (wt/vol) sodium chloride to give solutions containing 6.0 and 0.75 mg of danofloxacin/ml for use as test materials. Sodium chloride (0.9% [wt/vol]) was also administered as a negative control. Calves were randomly allocated to one of three treatments (11 animals per treatment): danofloxacin (6 mg/ml) administered as a single IV bolus injection at 0.738 mg/kg; danofloxacin (0.75 mg/ml) administered as a continuous IV infusion over 36 h to give a total infused dose of 0.645 mg/kg, following an initial small IV bolus of danofloxacin (6 mg/ml) at 0.093 mg/kg, to give a total dose of 0.738 mg of danofloxacin/kg; or saline at 0.123 ml/kg as a single IV bolus injection to give a dose volume equivalent to that of the danofloxacin bolus treatment. The dose regimen for the danofloxacin infusion was calculated by pharmacokinetic modeling of previously obtained plasma concentration-time data as the dose which would provide a steady-state concentration in plasma slightly exceeding the MIC for the M. haemolytica strain used in the respiratory infection model (MIC = 30 ng/ml) when infused continuously over a 36-h period. Time zero was defined individually for each calf as the time when treatment administration commenced. Injections were administered with one or two 10-ml disposable syringes via an IV catheter placed immediately prior to treatment or with two 1-ml disposable syringes via the indwelling catheter in animals to be infused. The infusion was delivered by an ambulatory infusion pump (CADD-PLUS model 5400; SIMS Deltec, Inc., Minneapolis, Minn.) attached to the calf's back with a specially designed pouch and harness. An 18-gauge indwelling catheter (Leader-Flex [PUR] Seldinger catheter; Vygon UK Ltd., Cirencester, United Kingdom) was placed in one jugular vein, with local anesthetic being used, and secured by cutaneous sutures. The pump was connected to the IV catheter in the jugular vein via an opaque sterile giving set (1.5-m length of tubing, dead volume of 1.3 ml; Becton Dickinson, Oxford, United Kingdom). The tubing was secured to the animal with sutures and tape and included several loops to relieve tension. Prior to connection of the pump, the initial single bolus was administered via the catheter, followed by administration of 5 to 10 ml of saline. The pump was then connected to the catheter, and the infusion was initiated. The pump was set to deliver the dose volume at a set rate for each animal such that the total dose was delivered over exactly 36 h. The body weight recorded for each animal prior to inoculation was used to determine the dosages administered and the infusion rate. Clinical observations. | Clinical observations for signs of bovine respiratory disease and measurement of rectal temperature were carried out immediately prior to inoculation, immediately prior to treatment administration at time zero, and at 4, 8, 12, 24, 36, and 48 h thereafter. The respiratory rate, rectal temperature, character of respiration, and general demeanor were assessed for each calf. The veterinarian making these observations was unaware of the treatment allocations of the animals receiving the bolus treatments (saline and danofloxacin), although the presence of the infusion apparatus made masking impossible for calves receiving the danofloxacin infusion. Any calf which was recumbent and showed severe depression and/or signs of respiratory distress was immediately euthanatized on welfare grounds. All calves were euthanatized after the final sampling (48 h after the start of treatment). At necropsy, a score of the percentage of lung consolidation was estimated from the extent of visible consolidation, both dorsally and ventrally, as a percentage of the total lung surface area. The necropsies were performed by an experienced pathologist who was unaware of the treatment allocation. In addition, lung samples were collected at necropsy for determination of viable counts of M. haemolytica. Sample collection and handling. | Blood samples were collected immediately prior to treatment and at approximately 15, 30, and 45 min and 1, 1.5, 2, 3, 4, 6, 8, 10, 12, 18, 24, 36, and 48 h following single-bolus administration. For animals receiving the continuous infusion, blood samples were collected immediately before treatment and then approximately 15 and 30 min and 1, 2, 4, 8, 12, 16, 20, 24, 28, 32, 36, 36.5, 37, 38, 42, and 48 h after the start of administration of the small loading bolus. Blood samples (5 ml) were collected by jugular venipuncture of the vein contralateral to that used for treatment administration, except for two samples (at 4 and 10 h) from one calf receiving the danofloxacin single bolus, where the ipsilateral vein was used due to sampling difficulties. Samples were collected using a 1-inch by 20-gauge sterile needle and were then deposited into blood tubes with heparin anticoagulant. The blood tubes were protected from light and were centrifuged at approximately 1,400 x g for 10 min, and the supernatant was transferred by pipette into duplicate plastic tubes. Plasma samples were stored at approximately -20C prior to assay. Measurement of concentrations of danofloxacin and its active metabolite N-desmethyldanofloxacin in plasma was performed by using a validated high-pressure liquid chromatography method with solid-phase extraction and fluorescence detection . The limit of quantification (LOQ) was 10 ng/ml. Bronchial secretion samples were collected from each animal by using an established, well-tolerated method . Samples were collected following endotracheal intubation approximately 16 h before inoculation, immediately before treatment administration began, and at approximately 1, 3, 6, 12, 18, 24, 36, and 48 h thereafter for determination of viable counts of M. haemolytica. An absorbent cotton fabric plug, fixed to the tip of a solid flexible polyethylene rod, was placed inside a sterile disposable stomach tube (one per animal for each sample) and inserted into the trachea orally by using a laryngoscope and a gag. At the tracheal bifurcation, the rod was pushed approximately 10 cm beyond the end of the stomach tube and into the principal bronchus. After a maximum residence time in the bronchus of 2 min, the absorbent plug was withdrawn into the stomach tube and the device was removed from the animal. The absorptive cotton plug was placed in a disposable syringe, and the bronchial secretion was extracted into an inert plastic sample tube by manual pressure. Sample tubes were placed on ice and kept in the dark until transferred to the laboratory within 1 h for immediate processing to determine the viable bacterial cell counts. Following euthanasia, tissue samples of approximately 0.5 g each were excised from eight standard sites in the lung and pooled to give four samples per calf. Each pooled sample was weighed, placed in a separate stomacher bag with 9.0 ml of peptone water, and homogenized. M. haemolytica counts were performed for each bronchial secretion sample and each lung homogenate following serial 10-fold dilutions. Duplicate aliquots of each dilution were cultured overnight at 37C on blood agar, and the mean colony count was used to determine the mean viable count of each sample. The four pooled lung samples per calf were counted separately, and an overall mean was calculated for each calf. Data analysis. | A logarithmic transformation (log [bacteria count + 1]) was applied to the M. haemolytica counts (in both bronchial secretion and lung tissue samples) prior to analysis. Bronchial secretion bacterial counts, respiration rates, and rectal temperatures were analyzed by using a repeated measurement model with the pretreatment value as a covariate. Lung lesion scores and lung tissue bacterial counts were analyzed by using a general linear model. A categorical analysis for repeated measurements was carried out separately for the clinical scores for respiration and demeanor. The proportion of animals completing the study in each treatment group (i.e., treatment successes) was calculated as the number of calves completing the study at 48 h multiplied by 100 and divided by the difference between the number of calves treated and the number of calves removed from the study for reasons not related to respiratory disease. The proportion of animals completing the study (treatment successes) and the proportion of animals with positive M. haemolytica counts in posttreatment bronchial secretions and in lung tissue samples were analyzed by using a logistic model. A priori contrasts were used to assess differences between treatments. The 5% level of significance (P <= 0.05) was used to assess statistical differences for all tests. Pharmacokinetic analysis. | Pharmacokinetic analyses were performed by using WinNonlin version 1.1 (Scientific Consulting Inc., Cary, N.C.). For calves receiving danofloxacin by single bolus or infusion, plasma danofloxacin concentrations were used to determine the concentration in plasma at time zero (Cp0); the maximum observed concentration in plasma [Cmax(obs)]; the total body clearance (CLb), calculated as the dose administered divided by the area under the curve from time zero to infinity (AUC0-infinity); and the terminal elimination rate constant (lambdaz), calculated from regression analysis of log concentrations over time. The elimination half-life (t1/2) was calculated as 0.693/lambdaz. The linear trapezoidal rule was used to calculate the AUC0-t (where t is the last time of measurable plasma concentrations). The MIC for the M. haemolytica strain used as the inoculum was 30 ng/ml; hence, the T>MIC was the time during which plasma danofloxacin concentrations exceeded 30 ng/ml. Plasma AUC0-infinity was calculated as AUC0-t + Ct/lambdaz, where Ct was the last measurable plasma danofloxacin concentration. The ratio of AUC to MIC (AUC:MIC) was calculated as AUC0-infinity/MIC. The plasma Cmax(obs):MIC was also calculated. All pharmacokinetic parameters and concentrations of danofloxacin were calculated for individual animals and are presented as means +- standard deviations (SD), except for t1/2 values, which were calculated as harmonic means. RESULTS : The mean pharmacokinetic values for danofloxacin in treated animals are presented in Table . Two calves receiving continuous-infusion treatment were not included in the calculation of mean pharmacokinetic parameters [except for Cp0, Cmax(obs), and Cmax(obs):MIC], since they were withdrawn from the study on welfare grounds related to severe respiratory disease and the infusion was terminated prior to the 36-h assessment time. When danofloxacin was administered as a single IV bolus, peak plasma drug concentrations were obtained at the first sampling time point, 15 min after bolus administration, with a mean Cmax(obs) of 436 ng/ml and an extrapolated Cp0 of 589 ng/ml . Following rapid distribution, danofloxacin was eliminated, with an overall mean t1/2 of 4.3 h. The continuous IV infusion of danofloxacin was preceded by the administration of an initial small IV bolus to rapidly achieve the target steady-state plasma drug concentration of >=30 ng/ml. Peak plasma danofloxacin concentrations were detected at the first time point, 15 min after administration of the small bolus, with a mean Cmax(obs) of 69.0 ng/ml and an extrapolated Cp0 of 85.8 ng/ml. Danofloxacin concentrations declined rapidly until steady-state concentrations slightly higher than 30 ng/ml were achieved approximately 4 h after commencement of infusion and were maintained at this level over the 36-h infusion period with only minor fluctuations . Following the end of the IV infusion at 36 h, the pump was disconnected and plasma danofloxacin concentrations declined, with a mean t1/2 of 2.3 h. The administration of danofloxacin as a single IV bolus produced a high Cmax:MIC of 14.5 and a relatively short T>MIC of 9.1 h . In contrast, when danofloxacin was administered as an IV infusion over 36 h, a low Cmax:MIC of 2.3 was obtained and steady-state plasma danofloxacin concentrations were maintained above the MIC for a prolonged period, resulting in a T>MIC of 33.3 h. The steady-state concentrations achieved were as intended, i.e., slightly in excess of the MIC for M. haemolytica, 30 ng/ml . The plasma AUC:MIC values for the single IV bolus and the continuous-infusion danofloxacin treatments were similar, 43.3 and 49.1, respectively . N-Desmethyldanofloxacin was not detected at measurable concentrations (LOQ = 20 to 50 ng/ml) at any time point in any animal in the study, and danofloxacin was not detected in any sample from a control animal. The number of animals withdrawn from the study for welfare reasons due to severe respiratory disease was greater in the saline treatment group (5 of 11, 45.5%) than in either the danofloxacin single-bolus (0 of 11, 0%) or the danofloxacin continuous-infusion (2 of 11, 18.2%) treatment group, although the difference was only statistically significant (P = 0.0037) for the comparison between the saline and the danofloxacin single-bolus treatment groups. There were significant reductions in the number of animals with M. haemolytica in bronchial secretions (P <= 0.0173) and in lung tissue samples (P <= 0.0266) for each of the danofloxacin treatment groups compared with the saline treatment group . In addition, the number of animals with M. haemolytica in bronchial secretions was significantly lower (P = 0.0477) for the danofloxacin single-bolus treatment group than for the danofloxacin continuous-infusion treatment group. There were no significant differences in the lung lesion scores between the treatment groups. M. haemolytica was isolated from bronchial secretion samples in some of the animals in the saline treatment group at each assessment time after treatment and in the majority of postmortem lung tissue samples in these animals. In comparison, M. haemolytica was isolated only at a very low count from one animal at 48 h after administration of a single bolus of danofloxacin, and low counts were recovered from five animals in the continuous-infusion treatment group. Compared with the saline treatment, each of the danofloxacin treatments resulted in significantly lower M. haemolytica counts in bronchial secretions (P <= 0.0001) collected from 3 to 48 h, inclusive, and in lung tissue samples (P <= 0.0053); however, the differences between the results for the two danofloxacin treatment regimens were not significant . Animals treated with each of the danofloxacin regimens showed greater clinical improvements (character of respiration and general demeanor) than those in the saline control group (Tables and ), and animals treated with single-bolus administrations of danofloxacin tended to have greater clinical improvement than those treated with the continuous infusion. Compared with animals in the saline treatment group, the animals in both danofloxacin treatment groups showed a significant improvement in demeanor from 0 to 48 h (P <= 0.0097); although the demeanor of the animals in the continuous-infusion treatment group was significantly different from that of the animals in the saline control group from 0 to 24 h (P <= 0.0151), the demeanor of the animals in the single-bolus treatment group was not. However, for improvement in the character of respiration, the single-bolus treatment produced significantly better results than the saline treatment at 24 and 48 h (P <= 0.0416). The difference between the results with continuous-infusion treatment and saline treatment was not significant. There were no significant differences in rectal temperatures between animals receiving the saline and continuous-infusion treatments; however, for those receiving the single-bolus treatment, rectal temperatures were significantly lower than for those receiving saline treatment (P <= 0.0167) at 12 and 24 h and significantly lower (P <= 0.0075) than for those receiving the infusion at 8, 12, and 24 h after the commencement of treatment. Respiratory rates were generally lower in animals in either danofloxacin treatment group than in animals in the saline control group, except at the 36- and 48-h time points. The reduction in respiratory rate was greatest at 8 and 12 h for animals receiving the single bolus, but this result was not significantly different from that seen with animals receiving saline. At 36 and 48 h, the respiratory rates for animals receiving single-bolus danofloxacin treatment were higher than for those receiving either saline or continuous-infusion treatments; however, at these time points, no animals had been withdrawn from the single-bolus treatment group, while four and five calves were withdrawn (at 36 and 48 h, respectively) from the saline treatment group and two calves were withdrawn from the continuous-infusion treatment group because they showed severe respiratory disease. Thus, the interpretation of data between treatments at these time points is difficult as only those calves remaining in the study were included for saline and danofloxacin infusion, so the results from the early time points (0 to 24 h) give a more accurate indication of the clinical response. FIG. 1. | Pharmacokinetics and pharmacodynamics of danofloxacin in cattle with respiratory disease. Pharmacokinetics and pharmacodynamics of danofloxacin in cattle with respiratory disease. Shown are plasma danofloxacin concentrations (means +- SD) following administration as a single IV bolus or as a continuous IV infusion. TABLE 1 | Mean pharmacokinetic parameters for danofloxacin in cattle with respiratory disease following administration either as a single bolus or as a continuous infusion TABLE 2 | Lung lesion scores and classification of animals by presence of M. haemolytica in posttreatment bronchial secretions and in lung tissue samples TABLE 3 | Geometric mean M. haemolytica counts in bronchial secretions and lung tissue samples TABLE 4 | Frequency distribution of clinical signs (assessment of demeanor) TABLE 5 | Frequency distribution of clinical signs (assessment of character of respiration) DISCUSSION : The pharmacokinetics of danofloxacin has been investigated in ruminant species including cattle , sheep , and goats . Following administration of danofloxacin, there is rapid distribution to the lungs and high tissue concentrations are achieved in pneumonic lung, including areas of consolidation . Danofloxacin has a broad range of activity against bacteria and mycoplasmas involved in bovine respiratory disease and is known to have a rapid bactericidal effect in vitro against M. haemolytica . The concentration-dependent killing profile is associated with a relatively prolonged postantibiotic effect . The purpose of this study was to evaluate the pharmacokinetic and pharmacodynamic characteristics of danofloxacin in an in vivo model of M. haemolytica pneumonia in calves. This evaluation was carried out by comparing the clinical and bacteriological outcomes of two regimens with predetermined equal total doses of danofloxacin administered to calves with a respiratory infection, thus establishing whether or not danofloxacin exhibits a concentration-dependent bactericidal effect in cattle. The same total dose of danofloxacin (0.738 mg/kg) was administered either as a single IV bolus injection or as a prolonged continuous IV infusion. The single IV bolus of danofloxacin was predicted to give a high Cmax:MIC and a short T>MIC, while the IV infusion of danofloxacin was predicted to give a low Cmax:MIC and a long T>MIC. As predicted, for those calves receiving the danofloxacin infusion treatment, concentrations in plasma were maintained above the MIC for the majority of the 36-h period (T>MIC = 33.3 h). In contrast, for those calves receiving the single bolus of danofloxacin, concentrations in plasma exceeded the MIC for less than 10 h posttreatment. The initial distribution of danofloxacin following the single IV bolus was rapid, and this was followed by an elimination phase which was monitored for 12 h postadministration, after which concentrations fell below the LOQ . The graph shown in Fig. suggests a multiexponential elimination curve with a terminal elimination phase slope evident after 2 to 3 h postadministration. For calves receiving danofloxacin by continuous infusion, the initial distribution following the loading bolus was also rapid and steady-state concentrations were achieved from 2 to 4 h after the first administration, with only minor fluctuations until the cessation of the infusion. The mean AUC0-infinity estimates for the continuous-infusion (1,472 ng h/ml) and single-bolus (1,298 ng h/ml) treatments represented only small extrapolations over the mean AUC0-t estimates (1,412 ng h/ml and 1,190 ng h/ml, respectively), and the similarities in the estimates of AUC0-infinity for both routes confirm that the total doses administered were equivalent overall. Thus, the AUC:MIC estimates for each treatment regimen were also similar (49.07 for the infusion treatment compared with 43.27 for the bolus treatment), as were the values for CLb for both treatments (507.8 and 578.2 ml/h kg, respectively). However, the t1/2 and lambdaz values for the two regimens were different (2.3 h and 0.3052 h-1, respectively, for the infusion treatment compared with 4.3 h and 0.1597 h-1 for the bolus treatment). These apparent differences in the t1/2 may have resulted from differences in the number of time points used to estimate the terminal elimination rate constant in the bolus treatment group compared with that in the danofloxacin infusion group. In this study, danofloxacin administered either as a single bolus or as a continuous infusion was significantly more effective in the treatment of M. haemolytica infection in calves than the control saline treatment. Overall, the administration of danofloxacin as a single bolus was more effective than administration of the same dose as a continuous infusion, as reflected in a higher percentage of animals successfully completing the study, significantly lower rectal temperatures over the initial 24-h period, and a significantly lower number of animals with M. haemolytica in bronchial secretions. This was in spite of the marginally lower AUC:MIC for danofloxacin following bolus administration. These data establish that danofloxacin exhibits a concentration-dependent antimicrobial activity when administered to the target species, cattle, with respiratory disease caused by M. haemolytica under conditions that closely simulated field conditions. Therefore, these data suggest that maximum therapeutic benefits can be obtained with danofloxacin with the administration of high doses over short periods. In the present study, the Cmax:MIC for the bolus regimen was 14.5. In a previous in vitro pharmacodynamic model, in which danofloxacin was shown to have concentration-dependent bactericidal action against Actinobacillus pleuropneumoniae, danofloxacin showed maximal bactericidal effect and there was no regrowth observed when the Cmax was at least eight times the MIC . The extent of protein binding of danofloxacin in plasma was not determined in the present study, and the concentrations and pharmacokinetic values presented correspond to total danofloxacin. However, the extent of protein binding of danofloxacin was determined in previous studies and can be described as reversible and relatively low, with values of approximately 49% in bovine plasma and 31 and 14% in bovine bronchial secretions and nasal secretions, respectively . In addition, danofloxacin has been shown to achieve concentrations in lungs and in bronchial mucosa that are approximately fivefold and threefold higher, respectively, than that achieved in plasma . Therefore, the steady-state free-drug concentrations achieved in the target tissues (i.e., bronchi and lungs) during the present study would have largely exceeded the MICs for M. haemolytica. Thus, the concentration-dependent activity observed with danofloxacin can be considered as a real effect rather than the result of subtherapeutic concentrations in animals treated with danofloxacin administered as an IV infusion. In addition to the correlation of increased efficacy with high Cmax-to-MIC and AUC-to-MIC ratios, high Cmax-to-MIC ratios have also been shown to minimize the potential for the development of resistance to fluoroquinolones . This characteristic has been established in several studies where the development of resistance to fluoroquinolones could be eliminated or drastically reduced when concentrations of the antimicrobial drug to which the bacteria were exposed exceeded the MIC by at least 8- to 10-fold . The pharmacokinetic and pharmacodynamic evaluation of danofloxacin in this M. haemolytica pneumonia model in calves has demonstrated the concentration-dependent activity of this drug in cattle. The principle of a concentration-dependent approach to therapy has been used to select the commercial dose of danofloxacin in the 18% formulation as 6 mg/kg administered subcutaneously either once or, if clinically required, twice 48 h apart. This selection is based on relating the AUC:MIC and Cmax:MIC results to recently determined MICs for field isolates of susceptible pathogens such as M. haemolytica (with MICs ranging from 0.015 to 2 mug/ml, an MIC at which 50% of isolates are inhibited of 0.06 mug/ml, and an MIC at which 90% of isolates are inhibited of 0.25 mug/ml [data not shown]). It is proposed that this concept will maximize the therapeutic characteristics of this potent molecule while minimizing the potential for the development of resistance and ensuring a high level of treatment compliance. Backmatter: PMID- 12183248 TI - Consumption of Imipenem Correlates with beta-Lactam Resistance in Pseudomonas aeruginosa AB - It is generally assumed that the antibiotic prescription policy of a hospital has a significant impact on bacterial resistance rates; however, few studies are available to support this concept with valid statistical data. During a 3-year period from 1997 to 2000, we monitored the consumption of beta-lactam and other antibiotics with known activity against Pseudomonas aeruginosa in a 600-bed community hospital. Monthly isolations of P. aeruginosa were assessed, and resistance rates were recorded. Partial correlation coefficients between consumption and resistance rates were determined, taking into account possible associations with other variables such as seasonal effects and transfers from other hospitals. A total of 30 +- 7 novel P. aeruginosa strains per month were isolated without epidemic clustering. Prescriptions of imipenem varied significantly during the study period, while prescriptions of other antipseudomonal agents were stable, with the exception of an increase in piperacillin-tazobactam prescriptions. Rates of resistance of P. aeruginosa to the antimicrobial agents used showed a time course similar to figures for imipenem consumption. Monthly rates of resistance to imipenem (partial correlation coefficient [cc], 0.63), piperacillin-tazobactam (cc, 0.57), and ceftazidime (cc, 0.56) were significantly associated with imipenem prescription rates in the same or the preceding month, while consumption of ceftazidime or piperacillin-tazobactam had no apparent association with resistance. Among the variables investigated, imipenem consumption was identified as the major factor associated with both carbapenem and beta-lactam resistance in endemic P. aeruginosa. Periods of extensive imipenem use were associated with significant increases in resistance. Our data support the concept that a written antibiotic policy which balances the use of various antibiotic classes may help to avoid disturbances of a hospital's microbial sensitivity patterns. Keywords: Introduction : Pseudomonas aeruginosa is a major pathogen in nosocomial infections, ranking second only to Staphylococcus aureus in a recent multicenter study of infections in European intensive care units . The organism may be involved in respiratory, urinary, wound, and bloodstream infections, primarily in patients with severe underlying diseases or impaired immune defense . Since the lethality of P. aeruginosa infections is high once the organism has caused disseminated disease, early and adequate antibiotic therapy is essential. Resistance of P. aeruginosa to antimicrobials used for primary treatment has been shown to correlate with an adverse clinical outcome . While the incidence and risk factors for the development of antibiotic resistance in P. aeruginosa have been studied repeatedly in cohort studies or comparative studies evaluating the effects of novel antibiotics , little is known about the influence of a hospital's antibiotic policy on resistance epidemiology in the hospital as a whole. In a 600-bed community hospital, we had the opportunity to observe significant changes in the prescription rates of antipseudomonal beta-lactam drugs over a 3-year period, which coincided with obvious variations in resistance. MATERIALS AND METHODS : Bacteriology. | P. aeruginosa was identified biochemically from routinely obtained specimens by means of the the Vitek ATB Expression System, version 2.7.8 (BioMerieux Deutschland GmbH, Nurtingen, Germany), which uses 32 biochemical reactions. Antibiotic susceptibility testing was performed by the same system, using a microbroth breakpoint dilution technique with an inoculum of 5 x 105 organisms per ml and a test volume of 135 mul in 18-well microtiter trays. MICs of the drugs were read automatically after aerobic incubation at 35C for 18 to 24 h. The susceptibility breakpoints (susceptible and resistant, respectively, in milligrams per liter) for P. aeruginosa were those of the German National Standard: <=2 and >=8 for imipenem, <=4 and >=32 for ceftazidime, and <=4 and >=64 for piperacillin (in the presence of 4 mg of tazobactam/liter) . For comparison, breakpoints of the National Committee for Clinical Laboratory Standards (NCCLS) are as follows: <=4 and >=16 for imipenem, <=8 and >=32 for ceftazidime, and <=64 and >=128 for piperacillin in the presence of 4 mg of tazobactam/liter . Susceptibility data were recorded monthly using a computer-based laboratory documentation system (Medat Software, Munich, Germany). The system is adjusted to count only primary isolates from individual patients, but to include follow-up isolates if the primary isolates show a different pattern of antibiotic resistance. Antibiotic consumption. | Antibiotic deliveries to the hospital were recorded in the clinical pharmacy on a monthly basis using a computer-based pharmacy documentation system. Deliveries could be assumed to reflect usage because the supply of pharmacy items to the wards was maintained continuously by the "Pharmacy Stock Control (top-up)" system. This system is based on an inventory or stock list placed on the ward, describing the contents and optimum stock levels of all trays and baskets, including correct locations. The labels on each tray and basket indicate the contents by bar code and in writing. A member of the pharmacy staff visits the wards three times a week (for wards with known high drug turnover) or twice a week (wards with low drug turnover) and checks the entire cupboard for the presence of bar codes, pharmacy requirements, expiration dates of drugs, and recent changes in prescribing. Thus, in the same way as all other drugs, antibiotics that had been ordered but not used were returned to the pharmacy within the same week, and supplies were delivered within 24 h after the visit of the pharmacist. Gram amounts of antibiotics were converted to "defined daily doses" (DDDs) by using the daily doses most frequently prescribed in this hospital, which were as follows: imipenem, 3 g; piperacillin, 12 g, and tazobactam, 1.5 g; ceftazidime, 6 g; gentamicin and tobramycin, 0.24 g; amikacin, 1 g; ceftriaxone, 2 g; cefotaxime, 6 g; intravenous ciprofloxacin, 0.4 g; oral ciprofloxacin, 1.0 g. Statistics. | The time series of resistance to imipenem, ceftazidime, and piperacillin-tazobactam were cross-correlated with the corresponding time series for use of the drugs in the same period. Associations were quantified using nonpartial and partial correlation coefficients according to Pearson and Spearman, assuming a lag of 1 month between use and resistance, and assuming no delay in between. In this context, a partial correlation between the use of and resistance to a beta-lactam means that seasonal effects, patient transfers from other hospitals, and the use of other antibiotics with activity against P. aeruginosa were eliminated from the association. For example, in estimating the association between ceftazidime resistance and imipenem use 1 month earlier, adjustments were made for the use of ciprofloxacin, gentamicin, tobramycin, piperacillin-tazobactam, and ceftazidime 1 month earlier. P values refer to testing to rule out the possibility that the indicated coefficients were equal to zero. In a further analysis, we modeled the dynamic relationship between antimicrobial use and resistance by using the time series method of Box and Jenkins as described by Lopez-Lozano et al. . Here, each of the antibiotic use series was modeled by an autoregressive low-order time series. Then transfer functions between resistance, as the response variable, and the use series, seasonal effects, and patient transfers from other hospitals were identified. Percentages of resistance among isolates were compared by a two-sided chi-square test. All statistical tests in this paper are exploratory, not confirmatory. RESULTS : Hospital setting. | The 600-bed acute-care community hospital serves a city of 40,993 inhabitants with a surrounding community of more than 200,000 inhabitants. All major medical and surgical disciplines are represented. During the study period, 19,000 to 20,000 patients were admitted every year, and fewer than 3% were transfers from other hospitals. On average, 30 novel P. aeruginosa strains were isolated every month without epidemic clustering . Use of antipseudomonal antibiotics. | Among the antipseudomonal drugs, non-beta-lactam antibiotics were used at a rather constant rate during the study period, with monthly prescription rates (DDDs per month +- 1 standard deviation) of 296.13 +- 59.0 g (ciprofloxacin), 125.8 +- 40.0 g (gentamicin), and 56.7 +- 40.0 g (tobramycin). Netilmicin was not used, and amikacin consumption was close to zero because this drug was used for exceptional cases only. Therefore, figures for amikacin were not taken into account in statistical evaluations. Among the antipseudomonal beta-lactams, ceftazidime use was low and stable during the study period, while use of piperacillin-tazobactam increased. Ciprofloxacin was the only quinolone used in this hospital. Striking variations were noted for the use of imipenem, which was the only carbapenem used during the study period. In fact, even when expanded-spectrum cephalosporins other than ceftazidime were taken into account, imipenem use was the major factor determining overall parenteral antibiotic consumption . Increasing consumption of imipenem during the first half of 1998 coincided with a marketing campaign by the manufacturer, who presented data showing that primary therapy of nosocomial pneumonia, secondary peritonitis, and other infections with imipenem led to significant cost savings . Reprints and summaries of these studies were distributed to hospital physicians by a representative of the manufacturer. Resistance of P. aeruginosa to imipenem. | Shortly after the intensified use of imipenem, we noted a dramatic increase in the resistance of P. aeruginosa isolates to this drug . In the early summer of 1998, the infection control and antimicrobial committees of the hospital expressed concern regarding increasing resistance of P. aeruginosa to imipenem and increasing use of this drug. Analysis of hospital microbiology data revealed no objective need for an intensified use of imipenem, because there was no increase in the number of P. aeruginosa isolates or organisms belonging to other species that would have been sensitive to carbapenems only. Therefore, it was decided that imipenem should not be delivered by the pharmacy unless written prescription forms were filled out stating the reason why other drugs could not be used (e.g., because of microbiologically documented resistance). Furthermore, prescriptions had to be countersigned by the chief or vice-chief of the respective clinical department. Following this decision, imipenem use decreased rapidly, as did resistance rates . At the end of the observation period, after the restrictions had been loosened, the use of imipenem increased again. Over the whole study period, imipenem consumption correlated significantly with imipenem resistance, with a (nonpartial) correlation coefficient of 0.63 (P <= 0.05) . This correlation was seen not only when monthly prescription rates were compared with resistance data from the same month, but also with those of the following month. By contrast, correlation coefficients dropped to insignificant levels when other time lags between use and resistance were used for calculation . Resistance of P. aeruginosa to beta-lactams. | Piperacillin-tazobactam was used more frequently for empirical therapy after July 1998, with monthly prescriptions rising from 28.2 DDDs in July 1997 to 153.8 DDDs in February 1999. Mean monthly prescriptions during the first period of high imipenem use (September 1997 to May 1998) were 68.9 +- 11.8 DDDs per month, compared with 106.0 +- 39.5 DDDs per month during the period of low imipenem use (June 1998 to December 1999). Despite increasing use, no correlation was seen between consumption of this drug and resistance . By contrast, both resistance to ceftazidime and resistance to piperacillin-tazobactam showed time courses similar to that of imipenem consumption, with peak ceftazidime resistance rates of 42% and peak piperacillin-tazobactam resistance rates of 25% during the period of maximum imipenem consumption in 1998 (data not shown). Partial correlation coefficients for imipenem consumption and resistance revealed significant statistical associations between imipenem use and resistance to ceftazidime or piperacillin-tazobactam . Time series analysis. | Results similar to those indicated by the Pearson coefficients were obtained by calculating Spearman's rank coefficients (data not shown) and by using time series modeling according to the work of Box and Jenkins. When modeling imipenem resistance, we found positive regression coefficients quantifying an association with imipenem use in the same month (P < 0.1) as well as with use during the preceding month (P < 0.05). The same was true when we checked the dependence of ceftazidime resistance (P < 0.06) and piperacillin-tazobactam resistance (P < 0.01) on imipenem use during the same month. Neither the use of ceftazidime nor that of piperacillin-tazobactam could be identified as a factor associated with resistance to one of the three antibiotics when intervals of up to 6 months between use and resistance were considered. Hospital epidemiology and hygiene practices. | The hospital had two infection control nurses and a consultant hospital epidemiologist. Microbiology data and printouts detailing the distribution of isolates on individual wards were analyzed monthly by the epidemiologist and summarized in written reports every 2 to 3 months. During the period under study, no epidemic clustering of P. aeruginosa isolates was documented. Cases were randomly distributed over as many as 23 wards, and no apparent breakdown in common hygiene practices was noted. Furthermore, infection control practices and disinfectants were not changed during this period, and the infection control team had been the same for a number of years. Records of microbiological sampling activity were kept during the whole period and showed that between 0.76 and 0.8 cultures were performed monthly per admitted patient, with variations of less than 3% between years and a maximum difference of 10% between corresponding months of successive years (1997 to 2000). Admission rates from other hospitals were no longer available for 1997, but in the beginning of 1998 (up to May 1998) the mean monthly percentage of patients admitted from other hospitals was 1.0%, compared with a mean of 2.6% for the rest of 1998. Mean total patient numbers admitted each month during the periods of low and high imipenem consumption varied by less than 10% (e.g., 1,608 +- 82 [range, 1,533 to 1,792] patients per month from January to May 1998, compared to 1,540 +- 70 [range, 1,428 to 1,613] patients per month for the rest of 1998). Resistance in primary versus secondary isolates. | The frequency of primary versus secondary P. aeruginosa isolates and their patterns of resistance could be evaluated only for the second peak of imipenem consumption in 2000, because earlier data were no more available. From January to May 2000, when consumption figures peaked again to values of >300 DDDs per month , a total of 183 isolates from various wards were isolated, 39 (21.3%) of which were secondary isolates. Thirteen of 144 (9.0%) primary isolates were resistant to imipenem, compared with 21 of 39 (53.8%) secondary isolates (P < 0.01 by the chi-square test). This shows that resistance was not due to an epidemic of primary imipenem-resistant isolates during this period. FIG. 1. | Overall use of major parenteral antibiotics during the 3-year period examined, expressed as DDDs per month for bimonthly intervals. Overall use of major parenteral antibiotics during the 3-year period examined, expressed as DDDs per month for bimonthly intervals. Figures for ciprofloxacin include oral administrations. The term "cephalosporins" refers to ceftriaxone and cefotaxime. FIG. 2. | Correlation between consumption of imipenem and resistance of P. aeruginosa Correlation between consumption of imipenem and resistance of P. aeruginosa to imipenem. (A) Bimonthly resistance rates plotted against bimonthly consumption rates for a 3-year period. (B) Relation between monthly imipenem consumption and resistance in the month of consumption for the whole 3-year period. (C) Nonpartial correlation coefficients between monthly imipenem consumption and resistance in the month of consumption (designated "0") and the 2 months prior to and following consumption. Asterisks indicate statistical significance (P < 0.005). TABLE 1 | Monthly admissions, transfers, and P. aeruginosa strains isolated TABLE 2 | Correlations between antibiotic consumption and resistance of P. aeruginosa DISCUSSION : Although it is often said that the pattern of antibiotic resistance in a given hospital is nothing but the negative imprint of the antibiotic prescription behavior in that hospital, few valid statistical data are available to substantiate this assumption. For some organisms, such as quinolone-resistant Escherichia coli, vancomycin-resistant enterococci, or quinolone-resistant coagulase-negative staphylococci, relationships have been established between consumption rates of the respective drugs and the increasing isolation rates of resistant organisms in individual hospitals . However, the association is less convincing for the use of broad-spectrum beta-lactam agents in relation to the epidemiology of resistance of common gram-negative bacteria. In a recent prospective study involving eight U.S. hospitals, it was not possible to find an association of consumption with resistance for individual antimicrobial-agent-organism pairs . It was speculated that other factors, such as interhospital transfer of resistance due to patient transfers, a community contribution to resistance, or a complex relationship between resistance and the use of a variety of antimicrobial agents, might have acted as confounding variables . Thus, the question of whether the control of antibiotic prescriptions by infectious-disease specialists (; R. P. Wenzel, Editorial, Clin. Infect. Dis. 24:456, 1997) or the scheduled cycling of antibiotics in high-risk areas such as intensive-care units is appropriate for reducing resistance rates is still a matter of controversy . In the hospital described here, several of the confounders just mentioned were virtually absent. Although some specialized outpatient services were offered, these did not include patient groups likely to carry drug-resistant P. aeruginosa. As shown in Table , transfers from other hospitals accounted for less than 3% of total admissions, which excludes a significant impact of antibiotic-resistant strains from other institutions on this hospital's ecology. The epidemiology of P. aeruginosa was characterized by regular isolations of this organism at a low but rather constant rate, reflecting a situation of endemicity. It has been shown previously that about one-half of endemically occurring nosocomial P. aeruginosa infections or colonizations are due to horizontal transmissions of the organism between patients, e.g., via fomites, fingernails of the personnel, or water from contaminated sinks . In the setting described here, horizontal transmissions may have acted as an amplifier of individual molecular events leading to resistance. Nevertheless, a significant association between antibiotic exposures and resistance at a level affecting the whole hospital would not have been detected had there not been enormous variations in the consumption of imipenem during the 3-year period examined. These variations were difficult to explain, because there were no apparent changes in the spectrum of patients treated or the monthly number of admissions. The primary use of carbapenems for empirical therapy of nosocomial infections was supported by various opinion leaders in Germany since the mid-1990s, and studies apparently proving a superior clinical cure rate achievable with imipenem compared to broad-spectrum cephalosporins or penicillins appeared in 1997 and 1998. According to these publications, the less frequent need to switch to other antibiotic regimens and the shorter overall hospital stay of carbapenem-treated patients led to significant cost savings . Issues of drug resistance were not addressed in these reports . Because drug orders in our hospital were served by the pharmacy without requests for authorization, we hypothesize that these publications promoted the intensified use of imipenem for empirical therapy. Although we cannot prove this hypothesis, the immediate decline in imipenem prescriptions by more than 80% after the introduction of written, countersigned prescriptions supported the assumption that ordering behavior during the preceding period had not been dictated by actual medical needs. In addition to the correlation between imipenem use and resistance of P. aeruginosa to the drug itself, our data also revealed an association between imipenem consumption and resistance to ceftazidime and piperacillin-tazobactam . In contrast, rates of consumption of ceftazidime and piperacillin-tazobactam showed no association with resistance to these two drugs. These observations are in accordance with data from Carmeli et al., who analyzed antibiotic treatment courses in individual patients suffering from P. aeruginosa infections . In their study, imipenem therapy was not only associated with emergence of resistance to imipenem (hazard ratio [HR], 44.0 compared to other antipseudomonal drugs; P < 0.001); it also had the strongest association with emergence of resistance to any of the antipseudomonal drugs under study, which included piperacillin-tazobactam and ceftazidime (HR, 2.9; P < 0.008). By contrast, patients treated with piperacillin-tazobactam or ceftazidime had no significantly increased risk of resistance . Similarly, in a case control study, Troillet et al. analyzed 40 patients whose first P. aeruginosa isolate was fully or intermediately resistant to imipenem and compared them to 387 control patients harboring fully susceptible strains. Pretreatment with imipenem was identified as the major risk factor for harboring a resistant strain (HR, 16.9; P < 0.0001), while pretreatment with cephalosporins or any intravenous beta-lactam was not a risk factor . Imipenem-resistant strains had a 3.6-fold-higher rate of concomitant ceftazidime-resistance, and a 2.3-fold-higher rate of piperacillin-tazobactam resistance, than imipenem-sensitive strains . It should be noted, however, that in a recent case control study that included several different control groups, pretreatment with piperacillin-tazobactam also carried a 2.4-fold-increased risk for isolation of an imipenem-resistant P. aeruginosa strain . The rapid emergence of resistance in clinical P. aeruginosa isolates during imipenem therapy has been well described . Commonly, imipenem resistance in this species is due to an interplay between beta-lactamase activity and reduced permeability of the bacterial outer membrane, which results from the loss of a specific membrane protein, the OprD2 porin . Treatment with other beta-lactam drugs could predispose to imipenem resistance by selecting strains with stably derepressed beta-lactamase production, which would then be more likely to lose their porin OprD2 during subsequent imipenem therapy . We did not analyze the mechanisms leading to imipenem and beta-lactam resistance in our strains. However, given the possibility that a proportion of the isolates were harbored by the patients during their prehospital phase, the possibility that these strains had been exposed to beta-lactam agents in the community cannot be excluded. This hypothesis would explain both the rapid emergence of resistance to imipenem and the concomitant beta-lactam resistance seen in our study. Several limitations of our study have to be addressed. Firstly, a more detailed analysis of resistance in primary and secondary isolates would have been desirable, in particular during the first peak of imipenem consumption. During the second peak, the analysis showed that horizontal transfers of resistant strains apparently played a minor part, because only 9% of primary isolates were imipenem-resistant, in contrast to 53.8% of follow-up isolates. Secondly, molecular typing was not done. Demonstration of molecular strain diversity would have excluded a common-source clonal expansion of a single strain during the periods of increased imipenem resistance. Lacking molecular data, our argument against the assumption of a clonal outbreak is that every physician in our hospital is trained to change an initial empirical antibiotic regimen as soon as microbiological tests showing resistance to the empirically used drug(s) are available. Because test results were available within 48 h, an outbreak of primarily imipenem-resistant strains would have led to a decrease rather than an increase in imipenem consumption. Further arguments against an outbreak are the time course between use and resistance , regular visits of an experienced infection control team to all wards, and the absence of clustering of P. aeruginosa isolates on individual wards. Finally, had there been an outbreak due to an unidentified lapse in hygiene precautions, restriction of imipenem prescriptions as the only measure taken would not have resulted in an almost immediate reversion of resistance to normal levels. A third limitation of our study is that overall consumption of antibiotics varied significantly, although this was due mainly to variations in imipenem usage . It is possible that any other agent in the place of imipenem might have been associated with similar phenomena of resistance. Therefore, we believe that our study provides an argument for a regulated, preferably written antibiotic policy which balances the use of individual antibiotics. Most likely, such a policy will prevent the occurrence of major disturbances of a hospital's microbial-sensitivity patterns by unrestricted use of particular agents. Backmatter: PMID- 12183257 TI - In Vitro Interaction of Flucytosine Combined with Amphotericin B or Fluconazole against Thirty-Five Yeast Isolates Determined by both the Fractional Inhibitory Concentration Index and the Response Surface Approach AB - Combination therapy could be of benefit for the treatment of invasive yeast infections. However, in vitro interaction studies are relatively scarce and the interpretation of the fractional inhibitory concentration (FIC) index can be contradictory due to various definitions used; not all information on the interaction study is used in the index, and different MIC end points exist for different classes of drugs. Fitting an interaction model to the whole response surface and estimation of an interaction coefficient alpha (ICalpha) would overcome these objections and has the additional advantage that confidence intervals of the interaction are obtained. The efficacy of flucytosine (5FC) in combination with amphotericin B (AB) and fluconazole (FCZ) was studied against 35 yeast isolates in triplicate (Candida albicans [n = 9], Candida glabrata [n = 9], Candida krusei [n = 9], and Cryptococcus neoformans [n = 8]) using a broth microdilution checkerboard method and measuring growth after 48 h by a spectrophotometer. The FIC index and ICalpha were determined, the latter by estimation from the response surface approach described by Greco et al. (W. R. Greco, G. Bravo, and J. C. Parsons, Pharmacol. Rev. 47:331-385, 1995) by using a computer program developed for that purpose. For the 5FC-FCZ combination, the interactions determined by the ICalpha generally were in concordance with the interactions determined by the FIC index, but large discrepancies were found between both methods for the 5FC-AB combination. These could mainly be explained by shortcomings in the FIC approach. The in vitro interaction of 5FC-AB demonstrated variable results depending on the tested Candida isolate. In general, the 5FC-FCZ combination was antagonistic against Candida species, but for some Candida isolates synergism was found. For C. neoformans the interaction for both combinations was highly dependent on the tested isolate and the method used. Response surface approach is an alternative method for determining the interaction between antifungal agents. By using this approach, some of the problems encountered with the FIC were overcome. Keywords: Introduction : Opportunistic yeast infections, such as candidiasis and cryptococcosis, occur particularly in immunocompromised patients, such as cancer patients and transplant recipients . The treatment of these infections is still problematic. Only a few antifungal agents are currently available, and although therapeutic options have increased considerably during the last decade due to the introduction of new agents , efficacy is not always optimal. Several drugs cause toxicity, and resistance to new agents is on the rise . One way to enhance treatment is to combine antifungal agents. Antifungal agents given in combination may improve efficacy due to synergism, and because the dose can possibly be lowered, side effects could be reduced. Another advantage of combination therapy is the reduction of the development of resistance and possibly a shorter duration of therapy . Flucytosine (5FC) is a drug with a limited spectrum of action that includes Candida spp. and Cryptococcus neoformans . Because resistance to 5FC emerges relatively fast, the drug is preferably given in combination with other antifungal agents. Although 5FC has been given in combination with other agents and has been evaluated in several clinical trials, little is known of the in vitro interaction of 5FC with other drugs . The fractional inhibitory concentration (FIC) index is the most frequently used analysis to quantify drug interaction . Despite the fact that this method has some important disadvantages, it is widely used. The first disadvantage is that one index (the FIC index) is used for many results: the result of every well in the checkerboard is confined in one index, while there may be more refinement necessary since at some concentrations there may be synergism but for others there may be indifference or even antagonism. Another disadvantage is that for some antifungal combinations it is not clear at which MIC end point the combination should be read and, thus, how to determine the FIC index. For amphotericin B (AB) and most antibacterial agents, end points are easily defined (one dilution) and the MIC is read as the lowest drug concentration that showed 100% growth inhibition (MIC-0) . For 5FC and especially for azoles, end points are less sharp (range of twofold dilutions) and exhibit a trailing growth effect. The MIC is read as the lowest drug concentration that showed 50% or more growth inhibition (MIC-2) . When AB is combined with 5FC or an azole, the combination can be read at either MIC-0 or at MIC-2. Finally, a third disadvantage is that there are several different definitions described in the literature for the interpretation of the FIC index , and it is still not clear which of these definitions should be used. To overcome these problems alternative methods have been sought for and developed, especially in the area of antiviral drugs . Instead of the reading of a single well (the MIC), the whole response surface is taken into account. By fitting a model to the whole response surface, an objective criterion for the interaction is obtained. The interaction can be further characterized by statistical analysis and confidence intervals. One of the models based on the response surface approach is the fully parametric model described by Greco et al. which is an Emax-based model (Emax model or Hill equation with variable slope) and fits to the entire data set with a nonlinear regression analysis. The nature and the intensity of an interaction are summarized with a nonunit, concentration-independent interaction parameter that includes the uncertainty in the estimate. We developed a program, called ModLab, which fits the Greco model to the whole response surface and then estimates the model parameters, including the interaction parameter, with their 95% confidence intervals. The aim of this study was to investigate the in vitro interaction of 5FC in combination with AB or fluconazole (FCZ) against a collection of 35 clinical yeast isolates by a broth microdilution checkerboard assay. The interaction was determined using both the FIC index and a response surface approach. (Part of these results were presented at the 40th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Canada, 2000.) MATERIALS AND METHODS : Test isolates. | Thirty-five clinical yeast isolates were tested: Candida albicans (n = 9), Candida glabrata (n = 9), Candida krusei (n = 9), and Cryptococcus neoformans (n = 8). The isolates were part of the private collection of the Department of Medical Microbiology, University Medical Center Nijmegen, Nijmegen, The Netherlands. The isolates were grown on Sabouraud glucose agar at 28C for 24 to 72 h. They were subcultured again on Sabouraud glucose agar for 24 h before preparation of the inoculum. All isolates were tested in triplicate on different days. Candida parapsilosis (ATCC 22019) and C. neoformans (ATCC 90112) were used for quality control in all experiments. Antifungal agents. | 5FC (ICN Pharmaceuticals, Zoetermeer, The Netherlands), AB (Bristol-Myers Squibb, Woerden, The Netherlands), and FCZ (Pfizer, Capelle aan den IJssel, The Netherlands) were obtained as powders. 5FC and FCZ were dissolved in distilled water and AB was dissolved in dimethyl sulfoxide to make stock solutions. The stock solutions were held at -80C until use. Serial twofold dilutions of each antifungal agent were prepared following NCCLS guidelines . Final dilutions were made in RPMI 1640 medium (with l-glutamine, without bicarbonate) (GIBCO BRL, Life Technologies, Woerden, The Netherlands) buffered to pH 7.0 with morpholinepropanesulfonic acid (MOPS) (0.165 mol/liter; Sigma-Aldrich Chemie GmbH, Steinheim, Germany). Interaction of drugs in vitro. | Drug interactions were assessed by a broth microdilution checkerboard method. The final concentrations of the antifungal agents ranged from 0.004 to 512 mug/ml for 5FC, 0.002 to 2 mug/ml for AB, and 0.008 to 512 mug/ml for FCZ. Aliquots of 50 mul of each drug at a concentration four times the targeted final concentration were dispensed in the wells of flat-bottom 96-well microtiter plates (Costar, Corning, N.Y.). Yeast inocula were prepared spectrophotometrically , and further diluted in RPMI 1640 medium. In order to obtain a final concentration of 0.5 to 2.5 x 103 CFU/ml, 100 mul of the yeast suspension was added to each well. The microtiter plates were incubated at 35C for 48 h. Readings were performed spectrophotometrically with a microplate reader (Anthos htIII; Anthos Labtec Instruments, Salzburg, Austria) at 405 nm. The optical densities (ODs) of the blank, which consisted of an uninoculated plate incubated together with the inoculated plates, were subtracted from the ODs of the inoculated plates. The percentage of growth for each well was calculated by comparing the OD of the wells with that of the drug-free control: (OD405 of wells that contained the drug/OD405 of the drug-free well) x 100%. The MIC of AB was defined as the lowest concentration that inhibited growth by 100% compared with that of the drug-free well (MIC-0). The MICs of 5FC and FCZ were defined as the lowest concentrations that inhibited growth by 80% or more compared with that of the drug-free well (MIC-1). For the 5FC-AB combination MIC-0 was taken as an end point, and for the 5FC-FCZ combination MIC-1 was taken as an end point. Definitions. | Drug interaction was determined by the FIC index and by the interaction coefficient alpha (ICalpha) based on the response surface approach by Greco et al. . The FIC index was defined as follows: The interaction was defined as synergistic if the FIC index was <1, additive if the FIC index was equal to 1, and antagonistic if the FIC index was >1 . The response surface approach by Greco et al. was described by the following equation : where and where D1 and D2 are the concentrations of drugs 1 and 2, IC50,1 and IC50,2 are the concentrations of drug 1 and drug 2 resulting in 50% inhibition, E is the measured response, Econ is the control response, m1 and m2 are the slope parameters for drug 1 and 2 in a constant ratio, and alpha is the synergism-antagonism interaction parameter (ICalpha). This model was fitted directly to the entire set of experimental data (percentages of growth for all concentrations of the two drugs alone or in combination) with a nonweighted, nonlinear regression analysis using a computer program (ModLab; Medimatics, Maastricht, The Netherlands). The program estimated the model parameters, and it also calculated the 95% confidence interval for each parameter. Validation of obtained parameters was done using the program Syner, kindly provided to us by G. L. Drusano. If the estimate of alpha is zero, the combination is additive. If it is positive, the interaction is synergistic. If it is negative, the interaction is antagonistic. The estimate of alpha has an associated 95% confidence interval. If the confidence interval does not overlap zero, this provides the statistical significance for the estimate of the interaction. That is, if the 95% confidence interval crosses zero, the interaction is additive. If it does not and alpha is positive, the interaction is significantly synergistic. If it does not and alpha is negative, the interaction is significantly antagonistic . Agreement between the FIC index and the response surface approach. | The interactions found by the FIC index were compared with those found by the response surface approach for both the 5FC-AB and 5FC-FCZ combinations. The percentage agreement between both methods for both combinations was determined. RESULTS : All yeast isolates grew well after 48 h of incubation at 35C. In each batch of broth microdilution tests, the MICs for the quality control strains were within the reference ranges. MIC data. | Table shows the MICs of 5FC, AB, and FCZ for the 35 yeast isolates, using the MIC-0 for AB and the MIC-1 for 5FC and FCZ. FIC index. | Table shows the FIC indices of the 35 yeast isolates for the 5FC-AB and 5FC-FCZ combinations determined by a broth microdilution checkerboard method, together with the results obtained by the response surface approach (see below). 5FC-AB combination. | For the 5FC-AB combination much variation was found between the Candida isolates, both between species as well as within species. Synergism was observed for four C. albicans isolates, and antagonism was observed for five C. albicans isolates, with the FIC index ranging from 0.65 to 2.12 and a median FIC (FIC50) of 1.33. For C. krusei synergism was observed for three isolates, and antagonism was observed for six isolates, with the FIC index ranging from 0.50 to 2.31 and a FIC50 of 1.17. For C. glabrata synergism was observed for six isolates and antagonism was observed for three isolates, with the FIC index ranging from 0.57 to 1.89 and a FIC50 of 0.92. Thus, the FIC50 for all 27 Candida isolates indicate that, overall, there was slight in vitro antagonism for the 5FC-AB combination (FIC50 = 1.14), but this combination appeared to be synergistic for some isolates. In vitro antagonism was observed for all eight C. neoformans isolates, with the FIC index ranging from 1.05 to 3.50 and a FIC50 of 1.83. 5FC-FCZ combination. | For the 5FC-FCZ combination, synergism was observed for five C. albicans isolates and antagonism for four C. albicans isolates, with the FIC index ranging from 0.07 to 9.50 and a FIC50 of 0.29. For C. krusei synergism was observed for one isolate and antagonism was observed for eight isolates, with the FIC index ranging from 0.31 to 2.88 and a FIC50 of 2.17. For C. glabrata antagonism was observed for all nine isolates, with the FIC index ranging from 1.50 to 16.50 and a FIC50 of 5.04. Importantly, for all C. glabrata isolates the mean MIC of the triplicates of FCZ increased from 64 to >128 mug/ml when used in combination with 5FC. In general, the 5FC-FCZ combination was antagonistic against Candida species (FIC50 = 2.17), although marked synergism was observed for some isolates (FIC index down to 0.07). Interestingly, the MIC of 5FC was lowered by at least 1 dilution step when combined with FCZ for all Candida isolates. Synergism was observed for one C. neoformans isolate, and antagonism was observed for seven C. neoformans isolates, with the FIC index ranging from 0.63 to 9.02 and a FIC50 of 1.73. Response surface approach. | Table also shows the interaction coefficient (ICalpha) of the 35 yeast isolates for 5FC-AB and 5FC-FCZ as determined by the response surface approach . Model fits were performed for each isolate and drug combination. An example is shown in Fig. . 5FC-AB combination. | For the 5FC-AB combination much variation was found between the yeast isolates. Synergism was observed for four C. albicans isolates, and antagonism for five C. albicans isolates, with the ICalpha ranging from -0.136 to 0.892 and a median ICalpha (ICalpha50) of -0.071. For C. krusei synergism was observed for seven isolates and antagonism was observed for two isolates, with the ICalpha ranging from -0.286 to 0.996 and an ICalpha50 of 0.255. For C. glabrata synergism was observed for three isolates and antagonism was observed for six isolates, with the ICalpha ranging from -0.164 to 1.239 and an ICalpha50 of -0.077. The ICalpha50 for all 27 Candida isolates indicated slight synergism for the 5FC-AB combination (ICalpha50 = 0.015). Synergism was observed for six C. neoformans isolates, and antagonism was observed for two C. neoformans isolates, with the ICalpha ranging from -0.143 to 0.896 and an ICalpha50 of 0.162. 5FC-FCZ combination. | For the 5FC-FCZ combination synergism was observed for six of the C. albicans isolates and antagonism was observed for three of the C. albicans isolates, with the ICalpha ranging from -0.187 to 30.212 and an ICalpha50 of 6.140. For C. krusei synergism was observed for one isolate and antagonism was observed for eight isolates, with the ICalpha ranging from -0.635 to 0.527 and an ICalpha50 of -0.241. For C. glabrata antagonism was observed for all nine isolates, with the ICalpha ranging from -0.643 to -0.036 and an ICalpha50 of -0.200. In general, the 5FC-FCZ combination was antagonistic against Candida species (ICalpha50 = -0.180) but highly synergistic in some isolates (ICalpha up to 30.212). Synergism was observed for three C. neoformans isolates, and antagonism was observed for five C. neoformans isolates, with the ICalpha ranging from -0.264 to 13.241 and an ICalpha50 of -0.024. Agreement between the FIC index and the response surface approach. | The percentage of agreement in the interpretation of the FIC index results and the response surface approach results was 91% for the 5FC-FCZ combination. For the 5FC-AB combination the agreement between both methods was much lower, namely, 40%: only 14 of the 35 yeast isolates gave the same results when determined by both methods. Thus, there is a discrepancy between the results of the two methods. The effect of several variables on the level of agreement. | Because we found a low level of agreement between both methods for the 5FC-AB combination, we analyzed the effect of several variables on the level of agreement. Interpretation of FIC index. | Different definitions for the interpretation of the FIC index can be found in the literature . Figure shows the percentages of synergism, addition, indifference, and antagonism for 5FC-AB and 5FC-FCZ, determined by the different definitions for the interpretation of the FIC index given in Table . As can be observed from the figures, large discrepancies occur. For instance, for the 5FC-FCZ combination synergism was found between 17 and 20% of isolates, while antagonism was found between 26% (following the definition of Walsh et al. or White et al. ) to 80% (following the definition of Berenbaum or Eliopoulos ) of isolates. Table shows the percentage of agreement between the interpretation of the FIC index results and the response surface approach for both the 5FC-AB and 5FC-FCZ combination depending on the definition chosen. As can be observed, the level of agreement varied between 0 and 40% for the 5FC-AB combination and between 43% and 91% for the 5FC-FCZ combination. Dependence of FIC on MIC end point. | The level of agreement between the interpretation of the FIC index results and the response surface approach results for the 5FC-AB combination was determined with regard to the MIC and the degree of inhibition used in the combination. The level of agreement varied between 43% (15 of 35 isolates) (MIC-0 for AB, MIC-1 for 5FC, and 100% inhibition for the combination) and 60% (21 of 35 isolates) (MIC-0 for AB and 5FC and 100% inhibition for the combination; MIC-0 for AB, MIC-1 for 5FC and 80% inhibition for the combination; and MIC-1 for AB and 5FC and 80% inhibition for the combination). FIG. 1. | Effect surface response of one of the triplicates of Candida krusei 1_25 for 5FC in combination with AB. Effect surface response of one of the triplicates of Candida krusei 1_25 for 5FC in combination with AB. The ICalpha (95% confidence interval) of this isolate was 1.196 (0.853 to 1.676), indicating synergism. FIG. 2. | Interactions found by using different definitions for the FIC index for the 5FC-AB (a) and for the 5FC-FCZ (b) combinations. Interactions found by using different definitions for the FIC index for the 5FC-AB (a) and for the 5FC-FCZ (b) combinations. TABLE 1 | MICs of 5FC, AB, and FCZ for 35 yeast isolates TABLE 2 | In vitro interaction between 5FC and AB or FCZ for 35 yeast isolates as determined by FIC index and ICalpha TABLE 3 | Definitions for interpretation of the FIC index TABLE 4 | Level of agreement between the interpretation of the FIC index results and the response surface approach results DISCUSSION : In the present study we determined the interaction between 5FC and AB or FCZ and showed that the interaction is largely dependent on the tested isolate. We also demonstrate that the use of the FIC index in synergy studies of antifungal agents has some important disadvantages in that the results can be interpreted in different ways. First of all there are different definitions described in the literature for the interpretation of the FIC index . There is no consensus on which definition to use in synergy studies. In this study we compared the interactions found by these different definitions for both the combination of 5FC with AB and that of 5FC with FCZ. Large discrepancies were found, depending on the definitions used. For example, synergism was found between 17 and 20% of isolates, while antagonism varied from 26 to 80% of the isolates for the combination of 5FC with FCZ. Not only the definitions but also the lack of a statistical criterion to define these interactions contribute to these varying results. A second disadvantage is that it is not clear at which MIC end point the combination should be read. This question is particularly important when a combination of drugs with different MIC end points is used. In another study we determined the interaction between AB and itraconazole against 15 Aspergillus fumigatus isolates . Depending on the chosen MIC end point the mean FIC index varied between 1.016 and 2.077 for 10 itraconazole-susceptible isolates or between 0.308 and 1.767 for five itraconazole-resistant isolates. Thus, the interaction found by the FIC index depends not only on the chosen combination but also on the choice of interpretation of the FIC index and, for some drug combinations, on the choice of MIC end point. From this, we conclude that the FIC index is not an optimal instrument to describe interactions between antifungal agents. To overcome these problems a response surface approach as described by Greco et al. can be used to determine the interaction between antifungal agents. The use of a model fit at least allows an objective statistical criterion and it makes an arbitrarily chosen end point unnecessary. Fitting of a model to the whole data surface not only allows the optimal use of information in the data but also allows the determination of error estimates of the interaction coefficient, thereby indicating whether the interaction is significant or not. Since the effect of drug interaction may vary between individual isolates this is an important feature. For the FIC index a statistical approach was described earlier by Drusano, but because of its complexity it is generally not used . Alternatively, not only does the interaction coefficient indicate whether there is significance or not, but the value found also gives an indication of the degree of interaction. This approach was used successfully in synergy studies of antivirals . In fitting the surface response model to the data, two approaches can be taken. The first one, and in most cases the approach of preference, would be to use an inverse variance weighting scheme. In such a circumstance, one would take the three determinations for each strain and calculate the mean and standard deviation of the response for each well. The mean would be the observation. The variance would be calculated from the standard deviation, and the inverse of the observation variance would be taken as the weight. We used this approach initially, but we encountered the problem that the intra-assay variance was much smaller than the interassay variance. Thus, although the surface fit would be very reasonable for each of the three experiments, using the inverse variance for each measurement in each well did not always work out. We therefore chose to fit the surface to the results of the individual experiments. Although this assumes homoscedasticity, the results of the model fit suggest this to be fairly reasonable. To determine the relationship between the two methods, the level of agreement was determined for the various combinations. For the combination of 5FC with FCZ the level of agreement between the interpretation of the FIC index results and the response surface approach results was very high (91%). In contrast, the level of agreement for the combination of 5FC with AB was very low (40%). There may be two explanations for the low level of agreement of this latter combination, namely, the chosen definition for the interpretation of the FIC index and, moreover, the chosen MIC end point (data in Results). This is substantiated by the fact that for the combination of 5FC with FCZ the level of agreement was relatively high. MICs of both 5FC and FCZ are read at the same end point; thus, there is no difficulty in the choice of the end point when calculating the FIC index. In contrast, for the combination of 5FC with AB there is a difference in the reading end points (MIC). The results show that the choice of end point much more influences the level of agreement (data in Results) than the definition itself . The use of the response surface approach precludes all these interpretation problems, by estimating parameters independent of a chosen end point. The clinically most used combination of antifungal drugs is 5FC with AB for Candida and Cryptococcus infections. The superiority of the combination over either drug alone in Cryptococcus infections has been described in both case series and randomized trials . Currently no randomized trials have been conducted to determine whether combination therapy of 5FC with AB is superior to monotherapy for invasive Candida infections. However, combination therapy is recommended in several conditions of invasive candidiasis, such as meningitis, endophthalmitis, endocarditis and peritonitis . Furthermore, it has been shown that human immunodeficiency virus-infected patients with cryptococcal meningitis could possibly benefit from the combination of 5FC with FCZ . However, these studies provide no data on the interactions of the two combinations in vitro. The combination of 5FC with AB showed synergism or indifference against C. neoformans and Candida strains . The combination of 5FC with FCZ showed synergism, indifference or addition, but no antagonism against C. neoformans strains . In this study we found that the in vitro interaction of the combination of 5FC with AB demonstrated variable results depending on the tested Candida isolate. The combination of 5FC with FCZ was generally antagonistic, but for some Candida isolates synergism was found. The in vitro inhibitory activity of 5FC was greatly enhanced by the addition of FCZ. The MICs of 5FC for Candida were decreased by at least 1 dilution step. On the other hand, the addition of 5FC did not greatly enhance the in vitro activity of FCZ. For the C. glabrata strains the mean MIC of the triplicates of FCZ even increased from 64 mug/ml when used alone to >128 mug/ml when combined with 5FC. The interaction found for the in vitro combination of 5FC with AB or FCZ against C. neoformans varied, depending on both the tested isolate and the used method (FIC index or response surface approach). A next step would be to determine whether differences in clinical outcome (responders versus nonresponders of therapy) could be explained using this approach of interaction studies. Our results do not concur completely with previous in vitro findings . This may mainly be due to the fact that different methods (broth macro- and microdilution), different MIC end points and different methods of interpretation of the results were used and possibly also that different fungal strains were tested. In the light of these differences, it is necessary to develop a standard methodology to determine in vitro antifungal interactions. The response surface approach would be a good alternative. It has to be emphasized that the results obtained in in vitro studies do not necessarily correlate with clinical outcome. Although we found antagonism between 5FC and FCZ for several strains, this combination has been found to be very useful in treating people with cryptococcal meningitis. Also AB in combination with 5FC is thought to be useful in the treatment of various forms of candidiasis. Other factors such as prolonged therapy and pharmacokinetic properties of the drugs play an important role in in vivo efficacy. We conclude that the response surface approach is a good alternative for determining the interaction between drugs against yeasts and has considerable advantages over conventional methods. Backmatter: PMID- 12183263 TI - Antibiotic Pharmacodynamics in Surgical Prophylaxis: an Association between Intraoperative Antibiotic Concentrations and Efficacy AB - The objective of this study was to characterize the relationship between gentamicin concentrations during surgery and the development of wound infection following colorectal operations. Despite decades of research in surgical prophylaxis, the relationship between intraoperative antibiotic concentrations and postoperative infection and the concentrations required for effective prophylaxis have not been established. A pharmacodynamic analysis was conducted using data from a previous prospective, randomized, double-blind clinical study which compared two dosage regimens of gentamicin plus metronidazole for prophylaxis in connection with elective colorectal surgery. Univariate and multivariate analyses of risk factors for postoperative wound infection were conducted, and the relationship between intraoperative gentamicin concentrations and surgical outcome was characterized. The gentamicin concentration at the time of surgical closure was one of the strongest independent risk factors for infection (P = 0.02), along with the presence of diabetes mellitus (P = 0.02), stoma (P = 0.04), and advanced age (P = 0.05). Gentamicin concentrations at closure of less than 0.5 mg/liter were associated with an infection rate of 80% (representing 8 of 10 patients with concentrations below that level) (P = 0.003). Receiver operating characteristic curve analysis identified a critical closure concentration of 1.6 mg/liter for effective surgical prophylaxis (P = 0.002; sensitivity, 70.8%; specificity, 65.9%). This study provides new and important information on antibiotic pharmacodynamics in surgical prophylaxis. It demonstrates the critical effect of the antibiotic concentration at closure on wound infection and suggests a significant association between the concentration and other well-established risk factors, like the timing of preoperative antibiotic administration and surgery duration. Keywords: Introduction : Surgical site infections, a significant postoperative complication, can lead to considerable patient morbidity and mortality . Wound infections account for 38% of surgical infections and 17% of all nosocomial infections . The benefits of preoperative antibiotics, which reduce bacterial contamination during clean-contaminated and contaminated operations, are well known . However, the relationship between intraoperative antibiotic concentrations and postoperative infection and the concentrations required for effective prophylaxis have not been established. Over the past decade, pharmacodynamic research has advanced the treatment of infectious diseases by characterizing relationships between antibiotic concentrations and clinical response . Although the application of similar principles to surgical prophylaxis has been suggested, there is a notable lack of supportive study . It is probable that low antibiotic concentrations during surgery due to inappropriate timing of the preoperative antibiotic , prolonged surgery , and patient obesity contribute to the high infection rates associated with these factors. However, the direct effect of intraoperative antibiotic concentrations on surgical outcome has been largely overlooked by clinical studies, which have not included this variable in risk factor analyses. Pharmacodynamic data which characterize effective antibiotic concentrations during surgery could change the approach to surgical prophylaxis. In a previous prospective, randomized, double-blind clinical study, regimens of single high doses of gentamicin (4.5 mg/kg of body weight preoperatively) and of multiple standard doses of gentamicin (1.5 mg/kg preoperatively and at 8, 16, and 24 h postoperatively), both in combination with metronidazole, were compared for prophylaxis in connection with colorectal surgery . Several observations suggested an association between low serum gentamicin concentrations during surgery and clinical failure. First, a trend towards fewer wound infections in the high-dose group suggested improved efficacy when higher antibiotic concentrations were achieved during surgery. Second, a strong association between prolonged surgery, which is a well-documented risk factor, and infection in the standard-dose but not in the high-dose group also supported an association between intraoperative antibiotic concentrations and clinical outcome. Our goal was to conduct a pharmacodynamic analysis of data from the original clinical study to characterize the relationship between intraoperative gentamicin concentrations and the development of wound infection following colorectal surgery. To our knowledge, this is the first such study in the area of surgical prophylaxis. MATERIALS AND METHODS : Previous clinical study. | Data were obtained from a previous prospective, randomized, double-blind clinical study (number of patients, 146) of antibiotic prophylaxis for elective colorectal surgery . Study treatments consisted of either single high doses of gentamicin (4.5 mg/kg) plus metronidazole (500 mg) preoperatively or multiple standard doses of gentamicin (1.5 mg/kg) plus metronidazole (500 mg) preoperatively and at 8, 16, and 24 h postoperatively. Only those patients with serum creatinine levels of less than 150 mumol/liter were enrolled in the clinical study. Gentamicin doses were based on actual body weight or on dosing weight for subjects weighing more than 120% of their ideal body weight. Dosing weights were calculated according to the formula [0.40 x (actual body weight - ideal body weight)] + ideal body weight. The ideal body weight was defined as 50 kg plus 2.3 kg for each inch of height over 5 ft for males and 45.5 kg plus 2.3 kg for each inch of height over 5 ft for females. Metronidazole and gentamicin were infused over 30 min and administered sequentially. All patients received polyethelene glycol electrolyte lavage (Golytely; Baxter Corp., Mississauga, Ontario, Canada) or phosphate soda solution (Phosphasoda; Merck Frosst Canada Inc., Pointe Claire, Quebec, Canada) the day prior to surgery. The primary clinical outcome was the development of surgical site infection within 30 days of surgery. Incisional infections were classified according to the Centers for Disease Control's standard definitions of superficial infections involving skin or subcutaneous tissues and deep infections involving deep tissue, including fascia or muscle . Subjects were monitored for systemic (e.g., fever, chills, or leukocytosis) and local (e.g., erythema, swelling, or purulent drainage) signs of infection. Pharmacokinetics. | Two blood samples were collected from the first 35 participants after the administration of the preoperative gentamicin dose. The first sample was drawn at least 30 min after the infusion, and the second sample was collected in the recovery room. Concentrations of gentamicin in serum were measured by fluorescence immunoassay (TDx; Abbott, Chicago, Ill.). The limit of detection was 0.2 mug/liter, and the coefficients of variation were 10.8% at 2.2 mug/liter and 6.5% at 13.4 mug/liter. The gentamicin assay results were concealed until the end of the study. A reference group (n = 16) of subjects was constructed; for this group, two measurements of serum gentamicin concentrations from appropriately collected blood samples were available. Parameters, including the elimination rate constant (kel) and volume of distribution (V), were calculated by using a one-compartment, linear pharmacokinetic model. Correlations between estimated creatinine clearance (CLCR) and kel and between weight and V were determined by linear regression analysis. A validation group (n = 19) was composed of subjects other than those in the reference group, with at least one measurement of gentamicin concentration. These individuals were excluded from the reference group because samples were not collected in 1 case, drawn too close to the administration of the dose in 12 cases, drawn too closely together in 4 cases, and not documented in 2 cases. The pharmacokinetic model was derived from data from the reference group and was tested with the validation group. The model was then used to construct profiles of serum gentamicin concentration versus time and to predict the concentration at the time of incision, the concentration at the time of closure, and the area under the concentration-versus-time curve during surgery (AUCsurg) for all other subjects . Pharmacodynamics. | Separate risk factor analyses were conducted with the high-dose and standard-dose groups. Univariate analyses were used to test standard variables (i.e., age, gender, weight, inflammatory bowel disease, malignancy, diabetes mellitus, chronic corticosteroid use, surgeon, operation, stoma, intraoperative core temperature, timing of the preoperative antibiotic, and surgery duration) in addition to pharmacodynamics parameters (i.e., gentamicin concentration at the time of incision, concentration at closure, and AUCsurg). Two-tailed t tests were used for continuous data, Wilcoxon rank tests were used for ordinal data, and Fisher's exact or chi-square tests were used for nominal variables (alpha = 0.05). Multivariate logistic regression analysis with backward elimination was used to identify independent risk factors. The polynomial equation from the model was then used to predict the probability of infection as follows: Receiver operating characteristic (ROC) curves were used to further analyze significant gentamicin concentration parameters. All statistical analyses were performed with SPSS for Windows, release 10 (SPSS Inc., Chicago, Ill.). FIG. 1. | Profile of serum gentamicin concentration versus time (Cinc, incision concentration; Cclos, closure concentrations). Profile of serum gentamicin concentration versus time (Cinc, incision concentration; Cclos, closure concentrations). RESULTS : Twelve of the clinical study participants were excluded because the timing of the preoperative antibiotic was not documented. Therefore, 134 subjects, of which 68 received the high-dose regimen and 66 received the standard-dose regimen, were included in the pharmacodynamic study. As shown in Table , there were no significant differences in characteristics between the high-dose and standard-dose regimens. Pharmacokinetics. | In the reference group (n = 16), the mean kel was 0.34 h-1 (95% confidence interval [CI95], 0.29 to 0.39 h-1), which corresponds to a mean harmonic half-life of 2 h, and the mean V was 0.23 liter/kg (CI95, 0.19 to 0.26 liter/kg). The pharmacokinetic model was described by the following equations: In the validation analysis , the pharmacokinetic model showed excellent correlations between predicted and measured serum gentamicin concentrations (r2 = 0.85), with good measures of precision (0.92 mg/liter) and bias (0.031 mg/liter). The gentamicin concentration parameters measured for the reference group and predicted for all other subjects are provided in Table . Pharmacodynamics. | In univariate analyses, several risk factors for infection were identified in the standard-dose group but none were observed in the high-dose group. In the standard-dose group, the gentamicin concentration at the time of closure (P = 0.001), the concentration at incision (P = 0.001), the surgery duration (P = 0.001), the presence of diabetes mellitus (P = 0.003), the presence of stoma (P = 0.03), and the timing of the preoperative antibiotic (P = 0.03) were associated with infection. As detailed in Table , however, only the gentamicin concentration at closure (P = 0.02) and the presence of diabetes mellitus (P = 0.02; odds ratio, 18.2; CI95, 1.7 to 193.8), stoma (P = 0.04; odds ratio, 4.3; CI95, 1.1 to 17.3), and advanced age (P = 0.05) were independent risk factors for infection. A concentration at closure of less than 0.5 mg/liter in 10 participants was associated with an 80% infection rate (P = 0.003; odds ratio, 2.8). The overall probability of infection is described by the following equation: where the values for diabetes and stoma were equal to 1 if present and 0 if absent. Figure simulates the effect of the gentamicin concentration at closure on the probability of surgical site infection in a representative 57-year-old patient population. ROC curve analysis identified a critical value for the gentamicin concentration at closure of 1.6 mg/liter for effective surgical prophylaxis (P = 0.002, sensitivity = 70.8%, specificity = 65.9%). FIG. 2. | Predicted versus observed serum gentamicin concentrations (mg/liter) in the validation analysis (n = 19). Predicted versus observed serum gentamicin concentrations (mg/liter) in the validation analysis (n = 19). FIG. 3. | Probability of postoperative wound infection and gentamicin concentration at closure. Probability of postoperative wound infection and gentamicin concentration at closure. TABLE 1 | Patient characteristics TABLE 2 | Gentamicin concentration parameters TABLE 3 | Independent risk factors for infection in the standard-dose group DISCUSSION : Although the goal of surgical prophylaxis is to maintain adequate antibiotic exposure during surgery, research establishing effective intraoperative concentrations has not been conducted . This study provides new and important information on antibiotic pharmacodynamics in surgical prophylaxis. In the standard-dose group, the risk of wound infection was dependent on the gentamicin concentration at closure and on the presence of diabetes mellitus, stoma, and, to a lesser degree, advanced age. Other well-established risk factors, including the timing of preoperative antibiotic and surgery duration, were identified in univariate tests but were not independently associated with infection. The identification of the gentamicin concentration at closure as an independent variable suggests that a low antibiotic concentration was the predominant risk associated with inappropriate timing of the preoperative antibiotic and with prolonged surgery. Notably, no risk factors for infection were identified in the otherwise-matched, high-dose group. This result could represent a gentamicin dose (i.e., 4.5 mg/kg) that produced adequate concentrations at closure for surgical prophylaxis and which therefore did not characterize the relationship between low antibiotic concentrations and clinical failure. Furthermore, neither timing of preoperative antibiotic nor surgery duration was significantly associated with infection in the high-dose group. This could also indicate that the gentamicin dose achieved sufficient intraoperative concentrations even in cases where the timing of the preoperative antibiotic administration was too early or the surgery duration was prolonged. In the treatment of infectious diseases, antibiotic concentrations are usually related to the pathogen and its susceptibility, as indicated by the MIC. Pharmacodynamic indices such as the peak concentration-to-MIC ratio, time above MIC, and AUC divided by MIC are analyzed to identify those which best correlate with the eradication of microbes or with clinical cure. The application of such principles to the prevention of infection is uncertain . However, this study shows the importance of gentamicin concentrations at surgical closure and identifies a critical value of 1.6 mg/liter for effective prophylaxis. This suggests that standard gentamicin doses, as used in our clinical study, may be suboptimal for colorectal operations. For example, a 1.5-mg/kg preoperative gentamicin dose would require a second dose in 3.8 h to maintain concentrations above 1.6 mg/liter (assuming a patient weighing 70 kg, a kel of 0.35 h-1, and a V of 0.25 liter/kg). A 4.5-mg/kg dose would extend the coverage to 6.9 h, which may be more appropriate for operations with a mean duration exceeding 3 h. The critical gentamicin concentration of 1.6 mg/liter must be interpreted with some caution. First, the value was determined for a specific antibiotic and patient population and may not apply to other operations, for example. Second, it was derived from a retrospective analysis of clinical study data and therefore may have been influenced by the range of intraoperative concentrations available. Although a prospective investigation of preselected and targeted antibiotic concentrations would provide the most complete pharmacodynamic characterization, it would also require the deliberate administration of low, presumably ineffective, doses resulting in clinical failure. This method of study has obvious ethical barriers. One limitation of this study was the application of pharmacokinetic data from a reference group to all other subjects. However, the pharmacokinetic model produced pharmacokinetic parameters that were within expected ranges and that demonstrated excellent predictive performance in the validation analysis . The model described a relatively homogenous population of participants with normal renal function who were admitted for elective surgery. This study was based on the assumption that bacterial wound infections are located in interstitial spaces and that concentrations in serum reflect those in interstitial fluids. First, clean incisional sites and other uninfected tissues are represented by high ratios of surface area to volume and by rapid equilibration of antibiotic levels between serum and wound fluid . The use of gentamicin, with low protein binding and rapid distribution into extracellular fluid, further justifies the use of concentrations in serum to approximate intraoperative levels in tissue. This study demonstrates the critical effect of antibiotic concentration at closure on wound infection following colorectal surgery. The results also suggest a significant association between low concentrations and high infection rates found with other well-established risk factors. Finally, this study shows the value of pharmacodynamic research in surgical prophylaxis and the need for investigations of other antibiotics for colorectal and other operations. Backmatter: PMID- 12183265 TI - Predicting Evolution by In Vitro Evolution Requires Determining Evolutionary Pathways AB - In an early example of DNA shuffling, Stemmer (W. P. C. Stemmer, Nature 370:389-390, 1994) demonstrated a dramatic improvement in the activity of the TEM-1 beta-lactamase toward cefotaxime as the consequence of six amino acid substitutions. It has been pointed out (B. G. Hall, FEMS Microbiol. Lett. 178:1-6, 1999; M. C. Orencia, J. S. Yoon, J. E. Ness, W. P. Stemmer, and R. C. Stevens, Nat. Struct. Biol. 8:238-242, 2001) that the power of DNA shuffling might be applied to the problem of predicting evolution in nature from in vitro evolution in the laboratory. As a predictor of natural evolutionary processes, that power may be misleading because in nature mutations almost always arise one at a time, and each advantageous mutation must be fixed into the population by an evolutionary pathway that leads from the wild type to the fully evolved sequence. Site-directed mutagenesis was used to introduce each of Stemmer's six substitutions into TEM-1, the best single mutant was chosen, and each of the remaining five substitutions was introduced. Repeated rounds of site-directed mutagenesis and selection of the best mutant were used in an attempt to construct a pathway between the wild-type TEM-1 and Stemmer's mutant with six mutations. In the present study it is shown (i) that no such pathway exists between the wild-type TEM-1 and the supereffective cefotaxime-hydrolyzing mutant that was generated by six amino acid substitutions via DNA shuffling (Stemmer, Nature 370:389-390, 1994) but that a pathway to a fourfold more efficient enzyme resulting from four of the same substitutions does exist, and (ii) that the more efficient enzyme is likely to arise in nature as the result of a single mutation in the naturally occurring TEM-52 allele. Keywords: Introduction : In vitro evolution by DNA shuffling and other methods that introduce multiple simultaneous mutations has become a powerful and widely used tool for the engineering of proteins to increase activities and/or to extend substrate ranges (, , , -, ). One of the earliest papers on DNA shuffling used that technique to increase the level of resistance to cefotaxime conferred by the TEM-1 beta-lactamase 32,000-fold . That increase was the result of six amino acid substitutions, three of which occur repeatedly in naturally occurring "extended-spectrum" alleles of the TEM beta-lactamase, plus a promoter mutation that increased the level of expression about 2.5-fold. The measure of the in vivo effectiveness of an antibiotic resistance protein such as TEM is the MIC of the drug substrate; the higher the MIC is, the more effective the protein is. That effectiveness is determined by the catalytic efficiency of the protein, the in vivo substrate concentration, and factors such as stability and folding efficiency that determine the concentration of active protein in the cell. The protein with six substitutions reported by Stemmer conferred a cefotaxime MIC of 640 mug/ml, considerably higher than the cefotaxime MICs conferred by the products of any naturally occurring TEM alleles. Does Stemmer's result predict that TEM alleles that confer cefotaxime MICs of 640 mug/ml will arise in nature as cefotaxime use continues? Blazquez et al. suggest that fluctuating selection as TEM alleles encounter other beta-lactam antibiotics might prevent such supereffective cefotaxime alleles from arising. Fluctuating selection is a special case of the "cost" of mutations that may be advantageous under one selective condition, the presence of cefotaxime, but disadvantageous under another selective condition. A second possible barrier to a supereffective cefotaxime-hydrolyzing TEM beta-lactamase might be the absence of an evolutionary pathway from existing alleles to the supereffective allele. The power of DNA shuffling lies in its ability to speed up evolutionary processes by the simultaneous introduction of multiple mutations into existing sequences. That power may be misleading as a predictor of natural evolutionary processes because in nature mutations almost always arise one at a time and each mutation must be fixed into the population by selection. (For populations as large as microbial and plasmid populations, fixation of any particular neutral mutation that might potentially be advantageous in the presence of another otherwise neutral mutation can largely be ignored.) With respect to increased antibiotic resistance, this means that during selection for antibiotic resistance each mutation must confer a measurable increase in effectiveness in order to be fixed into the population. When the fully evolved allele requires six amino acid substitutions, one can easily imagine two possibilities: (i) there is some order in which those mutations could occur such that each mutation improves the effectiveness of the gene product, i.e., each mutation improves fitness, in which case there is a clear pathway to the fully evolved mutant with six mutations; and (ii) at some point in the pathway, for example, after three advantageous mutations have accumulated, none of the remaining three mutations improves the effectiveness of the enzyme. In that case the evolutionary pathway will terminate with the triple mutant. In the second case, although it is known that the mutant with six mutations can exist and that it is fitter than the triple mutant, there is no pathway that can lead to the mutant with six mutations. EscherichiacolistrainDH5alphaE[F-phi80dlacZDeltaM15Delta(lacZYA-argF)U169endA1recA1hsdR17(r-m+) deoRthi-1phoA supE44 lambda-gyrA96relA1gal] (Gibco) was used as the host for all plasmids. L broth (10 g of tryptone, 5 g of yeast extract, 10 g of NaCl, and 1 g of glucose per liter) was used as a general growth medium. Mueller-Hinton broth (Difco), the standard medium used for the assessment of antibiotic resistance, was prepared according to the instructions of the manufacturer. Plasmids were prepared from overnight cultures grown in L broth containing 15 mug of tetracycline per ml. Plasmids were purified with the Qiagen QuickSpin kit according to the instructions of the manufacturer. Stock solutions of antibiotics were prepared in 0.1 M NaPO4 buffer (pH 7.0), filter sterilized, and stored at -80C in single-use aliquots. Amino acid substitution mutations were introduced into the TEM-1 beta-lactamase gene of plasmid pBR322 by site-directed mutagenesis with the QuickChange PCR mutagenesis kit of Stratagene, Inc., according to the instructions of the manufacturer. E. coli strain DH5alphaE transformants were selected on L-broth plates containing 15 mug of tetracycline per ml, and the TEM genes of the resulting plasmids were sequenced to confirm the incorporation of the desired mutation and the absence of any other mutations. The MIC of each antibiotic was determined by growing the plasmid-bearing strain overnight in L broth containing 15 mug of tetracycline per ml, diluting the culture to about 105 cells per ml in Mueller-Hinton broth (Difco), and adding to 500-mul aliquots of the diluted culture in 48-well microtiter plates twofold serial dilutions of the drug being tested. The plates were incubated for 24 h, and the MIC was taken as the lowest concentration of drug that prevented growth that was detectable by eye as turbidity of the culture. To determine whether there is a pathway from TEM-1 to the mutant with six mutations (mutant ST-3) reported by Stemmer , each of the six amino acid substitutions was introduced by site-directed mutagenesis into the TEM-1 allele borne on plasmid pBR322 . The substitution that gave the greatest increase in resistance to cefotaxime was chosen, and each of the remaining five substitutions was introduced into that allele. Similarly, the best double mutant was chosen, and each of the remaining four mutations was introduced into that allele. That pattern was followed until, after six rounds of selection, the ST-3 allele of Stemmer had been reconstructed. The rationale is that in nature population sizes are large enough that all possible alleles with single mutations are continuously being generated by spontaneous mutations. Given typical spontaneous mutation rates of 2 x 10-10 to 4 x 10-10 base substitutions per site per cell division, bacterial populations as small as 1010 cells will include at least one representative of each possible mutation. When challenged with a beta-lactam antibiotic, if the members of a population are free to compete with each other, the allele that confers the greatest resistance will replace the wild-type allele and subsequent mutations will arise in the new genetic background . Indeed, the selection imposed by 32 mug of cefotaxime per ml selects efficiently for high levels of resistance (MICs, 256 to 512 mug/ml) . Similar results were obtained with three other beta-lactam antibiotics, the MICs of which for the mutants exceeded the concentrations used for selection by 32- to 64-fold . Although neutral and slightly deleterious alleles will also arise and may even persist for some time, their frequencies remain so low that it is very unlikely that a particularly advantageous combination will arise as the consequence of a modestly advantageous mutation occurring in the rare background of such a neutral or slightly deleterious allele. TABLE 1 | Drug sensitivities of mutant TEM-1 alleles The MICs of four beta-lactam antibiotics, ampicillin, piperacillin, cephalothin, and cefotaxime, were determined for each mutant allele. Ampicillin is a widely used penicillin that was introduced into use in 1965, while piperacillin is a more modern penicillin that was introduced in 1981. Cephalothin is a cephalosporin that was introduced in 1964, while cefotaxime is an extended-spectrum cephalosporin that was introduced in 1981. The breakpoint for clinical resistance to ampicillin is an MIC of at least 32 mug/ml, the breakpoint for clinical resistance to piperacillin is at least 128 mug/ml, the breakpoint for clinical resistance to cephalothin is at least 32 mug/ml, and the breakpoint for clinical resistance to cefotaxime is at least 64 mug/ml . Note that more recent NCCLS guidelines define E. coli and Klebsiella isolates that produce any extended-spectrum beta-lactamase as being "resistant," regardless of the MIC conferred by that enzyme . The host, E. coli strain DH5alphaEpsilon, is sensitive to all four antibiotics. The TEM-1 allele confers resistance to ampicillin, piperacillin, and cephalothin but does not detectably increase the level of resistance to cefotaxime . Because the TEM-1 allele does not detectably increase the level of resistance to cefotaxime relative to the level of resistance of the host, the effectiveness of TEM-1 with respect to the hydrolysis of cefotaxime cannot be determined directly from the MIC. The effective MIC can, however, be estimated from the literature. Stemmer showed that the naturally occurring TEM-3 and TEM-15 alleles, both of which include the E104K and G238S substitutions, confer the same cefotaxime MIC (10 mug/ml). Franceschini et al. measured the efficiency with which the TEM-1 and TEM-3 enzymes hydrolyze cefotaxime in vitro. The kcat/Km for TEM-1 was 0.0015 muM-1 s-1 and that for TEM-3 was 1.0 muM-1 s-1, indicating that TEM-1 hydrolyzes cefotaxime 0.15% as efficiently as TEM-3 does and, by extension, 0.15% as efficiently as the mutant with the E104K and G238S substitutions that is equivalent to TEM-15. Table shows that in my hands the cefotaxime MIC for the mutant with the E014K and G2328S substitution is 8 mug/ml, giving an estimated MIC for the TEM-1 enzyme of 0.012 mug/ml. That estimate assumes that the Q39K substitution that is present in TEM-3 and absent in TEM-15 has no effect on the cefotaxime hydrolysis rate. On another substrate, cephaloridine, Q39K by itself increases kcat/Km by 50% relative to that for TEM-1 . A similar effect with respect to cefotaxime hydrolysis by TEM-3 would give an estimated cefotaxime MIC for TEM-1 of 0.018 mug/ml. In the absence of any direct evidence of such an effect, the improvement in the level of resistance of each mutant relative to that of TEM-1 is expressed relative to the estimated MIC of 0.012 mug/ml. The single substitution that confers the greatest increase in the level of resistance to cefotaxime is G238S, with a 42-fold improvement. That allele, however, also reduces the level of resistance to ampicillin, piperacillin, and cephalothin dramatically. Fluctuating selection during which the G238S allele encounters one of those drugs might prevent fixation of G238S in nature. The second most effective allele was E104K, which gives a 21-fold improvement in the level of resistance without suffering any decrease in levels of resistance to the other drugs. Therefore, for round two, both E104K and G238S were chosen, and all other substitutions were introduced into each background. The best double mutant was one with the combination of the E104K and G238S substitutions, a combination that resulted in a 667-fold improvement in resistance relative to that for TEM-1. Therefore, whichever of the two single substitutions was fixed into the population, the same double mutant would be fixed. The best triple mutant was that with the A42G, E104K, and G238S substitutions, in which the level of resistance was improved 5,333-fold relative to that for TEM-1, while the best mutant with four mutations was the one with the A42G, E104K, M182T, and G238S substitutions, which was 85,333 times more effective than TEM-1. The M182T substitution is known to act by affecting the folding pathway for the TEM beta-lactamases, thereby increasing the concentration of active enzyme . Although by itself the M182T substitution has no effect on the cefotaxime MIC, in combination with other substitutions it always increases the level of resistance . Introduction of either of the remaining substitutions, G92S or R241H, reduced the level of resistance to cefotaxime; therefore, neither mutant with five mutations would be fixed into the population by selection. Finally, the mutant with six mutations was exactly as resistant to cefotaxime as the better of the two mutants with five mutations. In the pBR322 background, the mutant with six mutations is resistant to cefotaxime at 256 mug/ml . Stemmer's mutant with six mutations carried an additional promoter mutation that increased the level of expression two- to threefold , and the cefotaxime MIC for the mutant was 640 mug/ml, exactly 2.5 times the level of resistance conferred by the same sequence in pBR322. It was expected either that there would be a clear evolutionary pathway leading to the mutant with six mutations or that the pathway would terminate early, resulting in a failure to achieve the level of resistance conferred by the mutant with six mutations. A third, unanticipated result was obtained: the evolutionary pathway terminated with the mutant with four mutations, but that mutant conferred resistance four times better than the mutant with six mutations. In effect, the introduction of multiple simultaneous mutations by DNA shuffling resulted in leaping over the adaptive peak and landing at a lower adaptive level. This result emphasizes the importance of reconstructing evolutionary pathways when using any in vitro evolution method that introduces simultaneous multiple mutations to predict natural evolution. This result predicts that a supereffective cefotaxime-hydrolyzing TEM beta-lactamase will arise in nature in response to the continued clinical use of cefotaxime. The sequence of the beta-lactamase with the A42G, E104K, M182T, and G238S substitutions, which confers resistance to cefotaxime fourfold more effectively than the best allele of Stemmer does, is just a single mutation (C to G at base pair 119 of the TEM-52 coding sequence) away from the sequence of the naturally occurring TEM-52 beta-lactamase (E104K, M182T, G238S) . The TEM-52 beta-lactamase confers a cefotaxime MIC of 32 mug/ml . Introduction of the A42G substitution not only increases the cefotaxime MIC 32-fold to 1,024 mug/ml, but it also increases the MICs of piperacillin and cephalothin fourfold each. Thus, fluctuating selection in the presence of those drugs will also favor the appearance in nature of a new TEM beta-lactamase that hydrolyzes cefotaxime extremely efficiently. Because selection for high levels of resistance does not require exposure to high levels of cefotaxime , the appearance of a TEM beta-lactamase with the A42G, E104K, M182T, and G238S substitutions is expected even if cefotaxime concentrations in the environment do not reach especially high levels. Backmatter: PMID- 12183245 TI - Method for Estimation of Low Outer Membrane Permeability to beta-Lactam Antibiotics AB - The outer membrane of gram-negative bacteria plays a major role in beta-lactam resistance as it slows down antibiotic entry into the periplasm and therefore acts in synergy with beta-lactamases and efflux systems. Up to now, the quantitative estimation of low outer membrane permeability by the method of Zimmermann and Rosselet was difficult because of the secreted and cell surface-associated beta-lactamases. The method presented here uses the acylation of a highly sensitive periplasmic penicillin-binding protein (PBP) (BlaR-CTD) to assess the rate of beta-lactam penetration into the periplasm. The method is dedicated to measurement of low permeability and is only valid when the diffusion rate through the outer membrane is rate limiting. Cytoplasmic membrane associated PBPs do not interfere since they are acylated after the very sensitive BlaR-CTD. This method was used to measure the permeability of beta-lactamase-deficient strains of Enterobacter cloacae and Enterobacter aerogenes to benzylpenicillin, ampicillin, carbenicillin, cefotaxime, aztreonam, and cephacetrile. Except for that of cephacetrile, the permeability coefficients were equal to or below 10-7 cm/s. For cephacetrile, carbenicillin, and benzylpenicillin, the outer membrane of E. cloacae was 20 to 60 times less permeable than that of Escherichia coli, whereas for cefotaxime, aztreonam, and ampicillin it was, respectively, 400, 1,000, and 700 times less permeable. The permeability coefficient for aztreonam is the lowest ever measured (P = 3.2 x 10-9 cm/s). Using these values, the MICs for a beta-lactamase-overproducing strain of E. cloacae were successfully predicted, demonstrating the validity of the method. Keywords: Introduction : beta-Lactams are the most potent and widely used antibiotics, and many studies have been devoted to understanding how bacteria increase their resistance to these compounds. Gram-negative resistance results mainly from the interplay between four independent factors: (i) the sensitivity of the target enzymes, the penicillin-binding proteins, (ii) the properties and concentration of the periplasmic beta-lactamases, (iii) the permeability of the outer membrane and (iv) the efficiency of the active efflux system . On this basis, a model which allowed a quantitative prediction of the MICs for gram-negative bacteria was developed and applied with success to Escherichia coli and Enterobacter cloacae . However, as demonstrated by Livermore and Davy , this model was not applicable to Pseudomonas aeruginosa, one of the most common opportunistic pathogens. This was due to efflux pumps which also significantly contribute to intrinsic beta-lactam resistance in P. aeruginosa . This was unexpected, because beta-lactam targets are located outside the cytoplasmic membrane and also because in E. coli this mechanism does not produce such a high and broad level of resistance . Li et al., pointed out that the low outer membrane permeability of P. aeruginosa renders this efflux mechanism particularly efficient (15, 16). In fact, while the few permeability coefficients (P) published for P. aeruginosa are low, they are also highly variable. For cephaloridine and nitrocefin, Nikaido's group reported P values of 10-6 and 6 x 10-7 cm/s, respectively , but others have measured values equal to or less than 2 x 10-8 cm/s . All these values were determined on the basis of the method of Zimmermann and Rosselet , in which the hydrolysis of beta-lactam by intact cells is compared to that obtained with a lysate. The analysis is then based on the assumption that a steady state is rapidly established in the periplasm, where the rate of penetration of the antibiotic which obeys Fick's law of diffusion is equal to that of the beta-lactamase-catalyzed hydrolysis, which follows Henri-Michaelis kinetics. If the validity of this assumption has been demonstrated experimentally , the method is not accurate when beta-lactamase activity is present in the medium or on the cell surface. Under these conditions, the error introduced remains negligible if the outer membrane permeability is high, as with E. coli , but when the permeability is low, as with P. aeruginosa, the correction factors range from 4 to 10 depending on the experimental conditions , and in addition it should be carefully verified that steady-state conditions really prevail. It is therefore not surprising that only a few P values have been determined for Enterobacter and Pseudomonas species. In the present study, we have developed a new method based on the properties of a soluble, periplasmic high-affinity penicillin-binding protein (PBP) which can easily be used to measure low outer membrane permeability coefficients. This PBP is the C-terminal domain of the BlaR penicillin sensory transducer involved in beta-lactamase induction in Bacillus licheniformis (M. Swinnen, S. Lepage, A. Brans, B. Granier, J. M. Frere, and B. Joris, submitted for publication). MATERIALS AND METHODS : Bacterial strains and plasmids. | Enterobacter aerogenes 006 is a wild-type strain that produces an inducible beta-lactamase. E. aerogenes 008-4 is a beta-lactamase-deficient strain obtained by subjecting the beta-lactamase-overproducing strain E. aerogenes 008 to chemical mutagenesis. These strains were kindly provided by N. A. C. Curtis . The resistant E. cloacae 908R overproduces a class C beta-lactamase. It was obtained by sequential passages on increasing concentrations of ceftriaxone and was kindly given by R. L. Then (Hoffmann-LaRoche, Basel, Switzerland) . E. cloacae AD2 is a beta-lactamase-negative mutant isogenic to E. cloacae 908R, obtained by insertion of the kanamycin resistance gene in the chromosomal ampR-ampC locus. For details about the construction of this strain see reference and references given therein. P. aeruginosa PAO1 was kindly provided by D. M. Livermore. Plasmids pDML309 and pDML310 were constructed by inserting the 1.2-kb HindIII-BalI fragment of plasmid pRTW8 coding for the 256 residues of the C-terminal domain of the BlaR penicillin sensory transducer of B. licheniformis (BlaR-CTD) in plasmid pIN-III-ompA . This vector allows periplasmic expression of proteins. It contains the strong lppp promoter of the E. coli lipoprotein controlled by the lac-UV5 promoter-operator. The E. coliompA signal peptide is under the control of the two latter promoters and a polylinker is present just at the end of the ompA signal peptide sequence. The ampicillin resistance gene of pIN-III-ompA was replaced by the tetracycline and kanamycin resistance genes, yielding pDML309 and pDML310, respectively (Swinnen et al., submitted). The resulting plasmids were used to transform E. cloacae AD2 and E. aerogenes 008-4, respectively. Plasmid pBL2 was used to produce BlaR-CTD in the periplasm of P. aeruginosa. It was constructed by ligating the 5.8-kb PstI-HpaI fragment of plasmid pDML309 to the 5.9-kb PstI-PvuII fragment of pKT240 (purchased from the Laboratorium voor Moleculaire Biologie Plasmidencollectie culture collection, University of Ghent, Ghent, Belgium). Antibiotics. | The following compounds were kind gifts of various companies: carbenicillin from Beecham (Brentford, United Kingdom), cefotaxime from Hoechst-Roussel (Romainville, France), cephaloridine from Eli-Lilly (Indianapolis, Ind.), aztreonam from Squibb (Princeton, N.J.), benzylpenicillin from Rhone-Poulenc (Paris, France), ampicillin from Bristol-Myers Squibb (Brussels, Belgium), nitrocefin from Oxoid (Basingstoke, United Kingdom), and [14C]cephacetrile (4.5 mCi/mmol) from Ciba-Geigy (Basel, Switzerland). 3H- and 14C-labeled benzylpenicillin (19 Ci and 59 mCi/mmol, respectively) were purchased from Amersham (Little Chalfont, United Kingdom), and tetracycline and kanamycin were purchased from Sigma (St. Louis, Mo.) and Merck (Darmstadt, Germany), respectively. MIC determination. | MICs were determined in Isosensitest broth (Oxoid) with serial doubling dilutions of the antibiotics. For each antibiotic concentration, an inoculum of 2 x 104 CFU taken from a growing culture in Isosensitest broth (A600 = 1), was added to 2 ml of medium. Results were recorded after a 24-h incubation at 37C with shaking (200 rpm). The MIC was taken as the minimal antibiotic concentration inhibiting bacterial growth. Measurement of benzylpenicillin hydrolysis. | An aliquot of cells was centrifuged at 5,000 x g and 4C for 3 min. A 10-mul aliquot of the supernatant was applied to a silica thin-layer chromatography plate (Merck), and benzylpenicillin was separated from benzylpenicilloic acid using chloroform-methanol-acetic acid (88:10:2, vol/vol/vol) as a solvent. SDS-PAGE and fluorography. | Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and fluorography were performed according to standard methods . Cell fractionation and labeling with fluorescent antibiotic. | Enterobacter cells were harvested by centrifugation when the culture had reached an A600 value of 0.8; resuspended in 30 mM Tris-HCl, pH 8, containing 27% sucrose, 5 mM EDTA, and lysozyme (0.5 mg/ml); and incubated at 4C for 20 min. The periplasmic fraction was separated from the spheroplasts by centrifugation at 30,000 x g for 15 min. To lyse the spheroplasts, they were resuspended in 10 mM Tris-HCl, pH 8, containing 10 mM MgCl2 and DNase (20 mug/ml). The lysed spheroplasts are referred to hereafter as the cellular fraction. For P. aeruginosa, the periplasmic fraction was isolated by three freeze-thaw cycles. Labeling of the periplasmic and cellular fractions was performed with 5'-fluoresceyl-glycyl-6-aminopenicillanic acid (5'-Flu-Gly-6APA) as previously described . Determination of beta-lactamase content of cells. | Aliquots of cells growing in Isosensitest broth were withdrawn after various time intervals. A portion (50%) of each aliquot was filtered on a Millipore GVWP filter (pore size, 0.2 mum) for dry weight measurement. The other 50% was kept in dry ice until sonication on ice (three 30-s bursts; amplitude, 6 mum). The beta-lactamase activity was determined in 10 mM HEPES buffer (pH 8.2) and at 30C with 115 muM nitrocefin as a substrate. Determination of outer membrane permeability. | Cells were grown overnight in Luria-Bertani broth containing kanamycin or tetracycline (50 mug/ml) as plasmid selection agents. The resulting culture was diluted 100-fold into the appropriate culture medium without any isopropyl-beta-d-thiogalactopyranoside or antibiotic, and growth was resumed at 37C up to an A600 of 1.5. Cells were harvested by centrifugation at 15 to 20C and resuspended in the appropriate prewarmed medium. The assay was started 3 to 5 min after equilibration at the desired temperature. A sample was also taken for dry weight determination. (i) Determination of outer membrane permeability by the method of Zimmermann and Rosselet. | A portion of cell suspension (cell density [dry weight] between 0.05 and 0.15 mg/ml) was sonicated on ice (three 30-s bursts; amplitude, 6 mum) and used for determination of maximum rate (Vmax). The remaining part of the intact cell suspension was mixed with 400 to 800 muM cephacetrile or 250 to 1,000 muM cephaloridine and transferred to a cuvette with a 1-mm light path, and the absorbance at 260 nm was recorded with the help of an Uvikon 860 device (Kontron Instruments). (ii) Determination of outer membrane permeability by the new method. (a) Preliminary comment. | The protocol described here is optimized for 3H- or 14C-labeled benzylpenicillin and [14C]cephacetrile. However, it can probably be adapted according to the properties of the radiolabeled antibiotic. Optimal results are obtained if BlaR-CTD labeling can be quenched as fast as possible and if the excess of free antibiotic can be almost completely removed. (b) Determination of outer membrane permeability if the studied antibiotic is radiolabeled. | Nine hundred microliters of cells (cell density [dry weight], 5 mg/ml) in the appropriate prewarmed medium was mixed with 100 mul of prewarmed antibiotic (10 muM [3H]benzylpenicillin [see Fig. ]; 4.5 to 18 muM [14C]benzylpenicillin [see Fig. ]; 1.5 muM [14C]cephacetrile). Aliquots of 100 mul were withdrawn after various intervals and mixed to 25 mul of 5 N HCl to instantaneously stop the binding to BlaR-CTD. The excess of free radiolabeled antibiotic was eliminated by three extractions with the same volume of water-saturated butanol. The proteins were then precipitated by adding 25 mul of a mixture of 50% trichloroacetic acid-3.75% tungstosilicic acid (Fluka, Buchs, Switzerland). Samples were left on ice for 10 min and then centrifuged 45 min at 15,000 x g. The supernatant was removed and the tubes were washed once with water to remove the last traces of radiolabeled antibiotic. After centrifugation, the pellet was solubilized in 10% SDS and the radioactivity was determined with a scintillation counter. (c) Determination of outer membrane permeability if the studied antibiotic is not radioactive. | If the studied antibiotic is not radioactive, the unlabeled antibiotic is incubated at 25C for increasing periods of time in a total volume of 1 ml of the cell suspension. The residual free BlaR-CTD is then counterlabeled with 15 muM [14C]cephacetrile. This labeled compound was chosen for its high diffusion coefficient. Under the conditions described here the counterlabeling is completed within 20 s. A higher cell density was used because of the lower specific radioactivity of the counterlabeling antibiotic. Nine hundred microliters of cells (cell density [dry weight], 20 to 30 mg/ml) in the appropriate prewarmed medium was mixed with 100 mul of prewarmed nonradioactive antibiotic. After different time intervals, 100-mul samples were removed and added with 50 mul of 45 muM [14C]cephacetrile. After 2 min of incubation at 25C, the reaction was stopped and the samples were treated as above. The antibiotic concentrations used for permeability measurements were as follows: 7.5, 15, and 30 muM cefotaxime; 50, 100, and 200 muM carbenicillin; 5, 10, and 20 muM ampicillin; and 75, 100, and 200 muM aztreonam. The longest incubation time was 45 min. Model and simulation. | Using the new method, the situation that prevails during the measurement of the outer membrane permeability for a beta-lactamase producing strain is represented by the following model: where Ie represents the external concentration of antibiotic, Ip represents the periplasmic concentration of antibiotic, E1 represents the periplasmic concentration of BlaR-CTD, E2 represents the periplasmic concentration of beta-lactamase, E1I* and E2I* represent the corresponding acyl enzymes, X represents the degradation product of I (all concentrations are micromolar concentrations), kd represents the first-order rate constant for antibiotic diffusion through the outer membrane (per second), kf1 and kf2 represent the second-order rate constants for the formation of E1I* and E2I*, respectively (per micromolar per second), and k3 represents the first-order rate constant for deacylation of the beta-lactamase (per second). The equations describing the variations of Ip, E1I*, and E2I* are Simulations were performed on the basis of these equations with the help of a numerical integration program based on a Runge Kutta method of the fourth order . For beta-lactamase-deficient strains, the same model was used with [E2] and [E2I*] equal to zero. Calculation of the permeability coefficient (P) by the new method. | Since kd is equal to PA/Vp and [E1I*] is equal to nE1I*/Vp, where Vp is the periplasmic volume and nE1I* is the quantity of labeled BlaR-CTD (in picomoles) A is equal to 132 cm2 mg (dry weight)-1, and the unit for [Ie] is nanomolar. d(nE1I*)/dt can be obtained by multiplying the slope of the observed labeling versus time (expressed in counts per minute second-1 milligram [dry weight]-1) by the specific activity of the labeled antibiotic (0.0096 pmol cpm-1 for [14C]benzylpenicillin and 0.123 pmol cpm-1 for [14C]cephacetrile as estimated using [14C]hexadecane [1.1 muCi/g] as a standard for estimating the counting yield). RESULTS : Validity of the new method. | Zimmermann was the first to use the rate of PBP acylation for a qualitative evaluation of outer membrane permeability . However, as pointed out by Nikaido, the method is valid only if the periplasmic concentration of the beta-lactam has not reached its equilibrium value, i.e., if the reaction of the beta-lactam with the PBP is more rapid than the diffusion process . Using the above model for a beta-lactamase-deficient strain, we have searched for combinations of kf1, kd, E1 total concentration (E1tot) and Ie which result in rate-limiting diffusion under these conditions: Simulations were done keeping E1tot between 10 and 40 muM, Ie between 0.5 and 100 muM, and kd at <0.26 s-1. In all cases the acylation of E1 up to 70% saturation (E1I*/E1tot = 0.7) was analyzed by linear regression. The kd values calculated on the basis of equation 2 were then compared to those used in the simulation. From these analyses (data not shown), it appears that diffusion can be considered limiting if This relation gives useful information on the limitations of the method. For example, it will be difficult to study the outer membrane permeability of E. coli because most of its kd values are larger than 0.13 s-1 , and on the basis of equation 3, it would thus be necessary to produce a periplasmic PBP with kf values larger than 30,000 M-1 s-1 (if we assume E1tot = 40 muM). On the contrary, this method applies well to P. aeruginosa because all reported kd values are below 0.13 s-1. BlaR-CTD is therefore an ideal PBP for such studies, because the kf values for various beta-lactams are between 103 and 106 M-1 s-1 (Swinnen et al., submitted). Note that if the value of kf1 for the studied compound is not known, it can be easily measured to demonstrate that diffusion is the rate-limiting factor (equation 3). beta-Lactamase production is another important factor, as these enzymes compete with PBPs for the free antibiotic. Consequently this method is not applicable to beta-lactamase-overproducing strains. For beta-lactamase-inducible strains, which produce low levels of enzyme in the noninduced state, the validity of the method will depend on (i) the relative concentration of the two enzymes, (ii) their relative affinity for the antibiotic, and (iii) the stability of the beta-lactam against hydrolysis. Development of the method. | Because some beta-lactams might cross the cytoplasmic membrane , the cellular localization of BlaR-CTD was analyzed. Periplasmic and cellular fractions were added with 5'-Flu-Gly-6APA, and the labeling was assessed under UV after SDS-PAGE. As expected from the plasmid construction, more than 95% of BlaR-CTD is found in the periplasm, and only a small amount of precursor is detected in the cellular fraction (result not shown). The influence of antibiotics on outer membrane integrity is still largely unknown, and therefore fresh cultures were grown without selective pressure. Under these conditions, 5 to 15% of cells lose their plasmid. Although these cells influence the dry weight but do not participate in the permeability measurement, no correction was introduced when calculating kd. Incubation of E. aerogenes 008-4 intact cells with [3H]benzylpenicillin shows that they bind more beta-lactam if they express BlaR-CTD . Some protein-bound radioactivity is released into the medium during the assay, but it only represents a small percentage of the cell-bound counts and is not expected to influence the slope of the line over the first 15 min. The binding appears to be a three-step process. First, a rapid step of small amplitude occurs during the mixing time (i.e., t = 0 [5 to 10 s in Fig. ]), followed by a second step of larger amplitude over a 15-min time scale and finally a third step with a small amplitude on a larger time scale. The same experiment analyzed by fluorography shows that during the first 15 min, when the highest binding rate is observed, only BlaR-CTD is acylated . Acylation of the membrane-bound PBPs only takes place during the third step of the experiment when BlaR is fully labeled. Because the total amount of BlaR-CTD is constant during the experiment and no significant release into the medium occurs during that time, the labeling observed over the first 15 min after mixing is diffusion dependent and can be used to measure the outer membrane permeability. The binding which occurs during the mixing time remains unexplained, since in the case of E. aerogenes 008-4 devoid of plasmid, no labeled protein is detected when samples are directly denatured and analyzed by SDS-PAGE. This labeling is probably due to nonspecific adsorption or trapping of the antibiotic because of the high cell density and acidic treatment. When pDML310 is present in the cells, in addition to this phenomenon, some BlaR-CTD might be freely accessible because it is inserted into the outer membrane with its active site directed to the medium, as already observed with beta-lactamases . Moreover, whatever its cause, this phenomenon does not alter the linearity of BlaR-CTD labeling during the second phase . For the cells which do not produce BlaR-CTD, the slow labeling is due to the membrane bound PBPs. The method was applied to measure the outer membrane permeability of E. aerogenes 008-4 to [14C]benzylpenicillin. As illustrated by Fig. , during the second phase, the labeling rate increases in proportion with the external antibiotic concentration. The permeability coefficient calculated from the slopes using equation 1 is (2.7 +- 0.5) x 10-8 cm/s. Using the same method, we have measured the permeability coefficient of [14C]cephacetrile. As expected from previous work, its diffusion rate is about 2 orders of magnitude higher than that of benzylpenicillin and is equal to (1.8 +- 0.3) x 10-6 cm/s and (4.2 +- 0.8) x 10-6 cm/s for E. aerogenes 008-4 and E. cloacae AD2, respectively. Because our method is expected to be less accurate for high P values, its validity was tested by measuring the permeability of E. cloacae 908R to cephacetrile with the help of the method of Zimmermann and Rosselet. Using the kinetic parameters of E. cloacae P99 beta-lactamase , the P value was estimated at about (5.2 +- 0.4) x 10-6 cm/s, a value not significantly larger than that measured by the new method. For unlabeled antibiotics, the method is only applicable by counterlabeling with a radioactive compound. Of course, the result is only valid if the diffusion rate of the labeled antibiotic (i.e., kd [Ie]) is at least 10 times larger than that of the unlabeled beta-lactam. Due to its high P value, cephacetrile is adequate in this case. Using the counterlabeling method with [14C]cephacetrile, the P value of E. aerogenes 008-4 for benzylpenicillin was found to be 3.1 x 10-8 cm/s. This value is close to that measured directly with radiolabeled benzylpenicillin, demonstrating that counterlabeling with cephacetrile is accurate. Therefore, the permeability coefficients of E. aerogenes 008-4 and E. cloacae AD2 for ampicillin, carbenicillin, cefotaxime and aztreonam were measured at 25C (Fig. ; Table ). In all cases, the outer membrane permeability of E. aerogenes is lower than that of E. cloacae by a factor of about 2. Application to wild-type beta-lactamase-producing strains. | As mentioned before, the presence of a periplasmic beta-lactamase can be an important limiting factor for the method. Using experimental values obtained with P. aeruginosa PAO1 and the mean P of Enterobacter for benzylpenicillin, simulations were performed to see if a low amount of class C beta-lactamase as encountered in wild-type strains could impair benzylpenicillin permeability measurements. On the basis of the beta-lactamase activity in a total cell lysate of PAO1, a periplasmic enzyme concentration of 0.09 muM was estimated. Table shows that BlaR-CTD is acylated even in the presence of a low amount of class C beta-lactamase. On the basis of the simulated acylation rate (23 muM in 120 s), the calculated kd was 6.3 x 10-3 s-1, which is only 5% lower than that used in the simulation. We therefore measured the outer membrane permeability of two wild-type strains producing low levels of inducible beta-lactamase. The measured P for E. aerogenes 006 to [14C]benzylpenicillin was 2.4 +- 0.5 x 10-8 cm/s. When the method was applied to P. aeruginosa PAO1, a P of 1.3 +- 0.25 x 10-8 cm/s was obtained for benzylpenicillin. In these two experiments, less than 30% of the beta-lactam was hydrolyzed during the experiment as shown by thin-layer chromatography analysis. DISCUSSION : Until now, the only method allowing a direct quantitative measurement of the outer membrane permeability was that of Zimmermann and Rosselet. Although there have been some technical improvements to this method, it still remains inaccurate to measure low permeability values when freely accessible beta-lactamase is present in the medium or on the cell surface. The new method described in this paper uses a periplasmic PBP which exhibits a high affinity for beta-lactams and is best suited for the measurement of low outer membrane permeability values. Indeed, equation 2 is only valid when the diffusion process is rate limiting, a condition more easily fulfilled when P is low. Under our conditions, and as shown by Fig. , only BlaR-CTD was significantly labeled. Interference by another PBP would require a rapidly reacting, overproduced, readily accessible protein, a situation which, to our knowledge, has never been encountered. Another advantage is that experiments can be performed even in complex solutions such as culture media. This is important since various experiments have demonstrated that pH and Donnan potential can modify the P for various beta-lactams . By our method, we have measured the outer membrane P values for two Enterobacter strains and six antibiotics. It should be noted that whatever the species, P values for benzylpenicillin, ampicillin, carbenicillin, and aztreonam had never been measured on whole cells. As shown in Table , with the exception of cephaloridine and cephacetrile, the P values do not exceed 10-7 cm/s. Also note that the permeability for aztreonam is the lowest ever measured and this antibiotic is the only one for which the half-equilibration times are larger than the generation times of the bacteria. This very low diffusion rate can also invalidate the method of Zimmermann and Rosselet, since the steady-state assumption is not verified. It also appears that the outer membrane of E. cloacae is two- to threefold more permeable than that of E. aerogenes, and as reported for other Enterobacter species, P values are lower than those of E. coli . For cephaloridine, cephacetrile, carbenicillin, and benzylpenicillin, E. cloacae AD2 is 20-to 60-fold less permeable than E. coli SNO3. For cefotaxime, aztreonam, and ampicillin, E. cloacae AD2 is, respectively, 400-, 1,000-, and 700-fold less permeable than E. coli. In decreasing order of permeation rate, the classification for E. cloacae is cephaloridine, cephacetrile, ampicillin, benzylpenicillin, cefotaxime, carbenicillin, and aztreonam. These results suggest that the porins of these Enterobacter strains have a different selectivity compared to those of E. coli. They seem to be more restrictive and/or more affected by the drug structure. Note that the only obvious common characteristic of ampicillin, cefotaxime, and aztreonam, the compounds for which the largest difference is observed between E. coli and Enterobacter, is the presence of a free amine on the side chain (in C-6 for the penam, C-7 for the cephem, and C3 for the monobactam). Whether or not this could influence the diffusion through Enterobacter porins remains to be established or confirmed. A similar wider dispersion of the P values in Serratia marcescens when compared to E. coli has been reported by Raimondi et al. . At 37C, the P value of E. cloacae AD2 for benzylpenicillin does not significantly increase when compared to that at 25C, whereas for E. aerogenes 008-4 and cefotaxime, benzylpenicillin, and aztreonam, these values increase by factors of 1.8, 2.4, and 3.5, respectively (data not shown). Consequently it appears that the studied beta-lactams mostly cross the outer membrane through porins. Because the measured diffusion rates are very low, one can wonder if active efflux mechanisms impair the measurement of the P values. The two beta-lactamase-deficient strains studied here are derived from beta-lactamase overproducing strains obtained by a multistep selection process. Based on recent results, one might suppose that these four strains possess an increased efflux activity that can compete with BlaR-CTD for the periplasmic beta-lactam. Various data do not favor this hypothesis. First, E. cloacae 908S (the wild-type beta-lactamase-inducible strain) and E. cloacae 908R (the beta-lactamase-overproducing strain) show the same level of resistance to various non-beta-lactam antibiotics such as chloramphenicol, norfloxacin, nalidixic acid, or erythromycin, and these levels are the same as those observed for Salmonella wild-type strains . Moreover, the P for benzylpenicillin, a beta-lactam sensitive to efflux systems, is the same for E. aerogenes 006, the wild-type strain, and E. aerogenes 008-4. Therefore, the efflux activity of the beta-lactamase-deficient and -overproducing strains used in this study is not significantly different from that of the wild-type strains. To demonstrate that, in the cases we studied, the basal efflux system does not interfere with the method, we have tried to predict the MICs for the beta-lactamase-overproducing strains on the basis of the following equation: MIC = Ipl + VmaxIpl/[PA (Km + Ipl)]. Because its beta-lactamase is still partially inducible (data not shown), the prediction was not done with E. aerogenes 008. In Table , the lethal periplasmic concentration of beta-lactam (Ipl) is the MIC for E. cloacae AD2. These values are higher than expected on the basis of the PBP sensitivity, but as discussed previously , except for cephaloridine, this has little impact on the MICs for E. cloacae 908R, because they are much larger than the corresponding Kms, and therefore the VmaxIpl/PA (Km + Ipl) term simplifies to Vmax/PA. This "abnormal" resistance of E. cloacae AD2 probably results from an efflux system naturally expressed in this strain or from the selection of a mutant overproducing an efflux system during MIC measurements. How ever, while efflux certainly contributes to increase the MICs for AD2, it is unlikely to affect the beta-lactam resistance of 908R (compare the MICs for the isogenic AD2 and 908R strains). If efflux was interfering with P measurement, it would certainly decrease the periplasmic concentration of antibiotic and thus the rate of BlaR-CTD acylation. This would result in underestimated P values and overestimated predicted MICs, which is not the case except for cephaloridine, for which P was determined by the method of Zimmerman and Rosselet, and benzylpenicillin, but in the latter case the MIC might be higher as higher antibiotic concentrations have not been tested. Otherwise the predicted and observed MICs are in fair (benzylpenicillin and carbenicillin) to good (all others) agreement. It is important to note that because the antibiotic concentrations used for the permeability measurements are much lower than those used in the Zimmermann-Rosselet experiments, the new method could be more easily affected by efflux systems since their influence is not masked by a high antibiotic influx. However, as for beta-lactamase, the importance of efflux in the experiment will depend on the relative affinity and expression level of these systems compared to BlaR-CTD. This restriction does not seem to apply to our conclusions, since the rate of BlaR-CTD acylation was high in all cases. Showing that the measured P values are compatible with the MICs for a beta-lactamase-overproducing mutant is thus a useful control of the validity of the method. Thanks to the progress in genetic engineering methods, it is now easy to selectively inactivate chromosomal beta-lactamases or efflux systems in gram-negative bacteria. Therefore, we think that despite some limitations, the method described here will be of great interest to more accurately measure low outer membrane permeabilities encountered in some genera and/or resistant mutants. FIG. 1. : Time-dependent labeling of the 50% trichloroacetic acid-3.75% tungstosilicic acid insoluble fraction of E. aerogenes Time-dependent labeling of the 50% trichloroacetic acid-3.75% tungstosilicic acid insoluble fraction of E. aerogenes 008-4 cell suspensions incubated at 25C with 10 muM [3H]benzylpenicillin. Symbols: , cellular fraction from cells producing BlaR-CTD; , supernatant fraction from cells producing BlaR-CTD; , cellular fraction from nontransformed cells. FIG. 2. : PBP labeling after various incubation periods of time (labeled atop the lanes, in minutes) of 10 muM [3H]benzylpenicillin with whole cells of E. aerogenes PBP labeling after various incubation periods of time (labeled atop the lanes, in minutes) of 10 muM [3H]benzylpenicillin with whole cells of E. aerogenes 008-4 producing BlaR-CTD. Lanes: C, purified BlaR-CTD protein; S, supernatant. (A) Short autoradiography exposure time (24 h). Only BlaR-CTD is detected. BlaR-CTD is not detected in the lane S because the quantity (<0.5 pmol) was too small. In lanes 5 to 30, the quantity of BlaR-CTD is about 3 pmol. (B) Long autoradiography exposure time (9 days). Labeling of membrane-bound PBP becomes detectable after 20 min of incubation. BlaR-CTD is the intense, fast-moving band. FIG. 3. : Progressive labeling of whole E. aerogenes Progressive labeling of whole E. aerogenes 008-4 cells producing BlaR-CTD in the presence of 18 muM , 9 muM , and 4.5 muM [14C]benzylpenicillin. The slopes of the initial parts of the lines were 8.08, 2.8, and 1.6 cpm/s mg (dry weight), respectively, and were thus proportional to the external antibiotic concentration. Errors on the various experimental measurements were below 15% with 18 muM penicillin and below 10% with 9 and 4.5 muM penicillin. Values were obtained at 25C in 10 mM phosphate buffer, pH 7, containing 5 mM MgCl2. FIG. 4. : BlaR-CTD acylation rate in whole cells of E. aerogenes BlaR-CTD acylation rate in whole cells of E. aerogenes 008-4 incubated with 75 muM , 100 muM , and 200 muM aztreonam as detected by counterlabeling with [14C]cephacetrile. The slopes of the initial parts of the lines were -0.133, -0.165, and -0.41 cpm/s mg (dry weight), respectively. TABLE 1 : Outer membrane permeability at 25C of E. aerogenes 008-4 and E. cloacae AD2 for beta-lactams and corresponding half-equilibration times TABLE 2 : Simulation of the influence of a low level of inducible periplasmic class C beta-lactamase (E2) on the acylation rate of BlaR-CTD (E1) by benzylpenicillin TABLE 3 : Predicted and observed MICs of beta-lactams for E. cloacae 908R Backmatter: PMID- 12183264 TI - Prospective Survey of beta-Lactamases Produced by Ceftazidime- Resistant Pseudomonas aeruginosa Isolated in a French Hospital in 2000 AB - In 2000, at the Universite d'Auvergne teaching hospital in Clermont-Ferrand, France, 44 (6.2%) strains of Pseudomonas aeruginosa were found to be resistant to ceftazidime. After genotyping, 34 strains were selected. Nine had an additional beta-lactamase: OXA-21 (n = 6), PSE-1 (CARB-2) (n = 2), or PER-1 (n = 1). Ceftazidime resistance was related solely to the overproduction of the cephalosporinase in 30 strains. Sequencing of five blaAmpC genes encoding cephalosporinases with different pIs showed 99% identity with the ampC gene of P. aeruginosa PAO1. Keywords: Introduction : Resistance to extended-spectrum cephalosporins in Pseudomonas aeruginosa is associated in most cases with the overproduction of a naturally produced cephalosporinase . Derepression of this AmpC beta-lactamase affects all beta-lactams to various extents, which may lead to the misdetection of other acquired beta-lactamases. Enzymes hydrolyzing extended-spectrum cephalosporins of Ambler classes A, B, and D have been reported in P. aeruginosa . Among the extended-spectrum beta-lactamases (ESBLs) of class A, TEM-4 , TEM-24 , TEM-42 , PER-1 , SHV-2a , VEB-1 , and GES-2 have been encountered. Class B enzymes, IMP-1 , VIM-1 , VIM-2 , and VIM-related enzymes , hydrolyze carbapenems in addition to extended-spectrum cephalosporins. The extended-spectrum class D enzymes are OXA-2 derivatives (OXA-15 and OXA-32 ), OXA-10 derivatives (OXA-11 , OXA-14 , OXA-16 , OXA-17 , OXA-19 , and OXA-28 ), and OXA-18 . Other mechanisms implicated in resistance to beta-lactams in P. aeruginosa are impermeability of the outer membrane and increased efflux The aim of this study was to characterize the beta-lactamases produced by ceftazidime-resistant P. aeruginosa isolates. Of the 713 nonduplicate P. aeruginosa strains isolated from the Universite d'Auvergne teaching hospital in Clermont-Ferrand, France, between 1 January and 30 September 2000, 44 (6.2%) were resistant to ceftazidime, as determined by the disk diffusion method on Mueller-Hinton agar according to the recommendations of the Antibiogram Committee of the French Society for Microbiology (zone diameter, <15 mm; ). This prevalence of ceftazidime resistance was higher than the 4% prevalence in 1996 reported in a previous French national multicenter study . Isolates were identified with the ID 32 GN system (BioMerieux, Marcy l'Etoile, France). A duplicate isolate was defined as an isolate with the same antibiotic resistance phenotype as another isolate from the same patient during the same course of infection. The 44 strains were typed by random amplification of polymorphic DNA (RAPD) with two primers, ERIC-2 and 208, as described previously . Isoelectric focusing analysis was performed with 6% polyacrylamide gels containing ampholines (Amersham Pharmacia Biotech, Uppsala, Sweden) with a pH range of 3.5 to 10, as reported previously . Among the 44 strains, 34 were selected on the basis of antibiotic resistance patterns, pIs, and patterns by RAPD analysis . The chromosomal cephalosporinases focused at five pIs: 8.4 (n = 13), 8.0 (n = 8), 8.0 and 8.4 (n = 7), 7.8 (n = 4), and 8.6 (n = 2). Three additional beta-lactamases were found at pIs 7.0 (n = 6), 5.7 (n = 2), and 5.4 (n = 1). The strains were distributed into 33 different RAPD profiles. TABLE 1 | Oligonucleotides used as primers for amplification and/or sequencing TABLE 2 | Characteristics of the 34 ceftazidime-resistant P. aeruginosa isolates The MICs of drugs alone or in combination with a fixed concentration of clavulanate (2 mug/ml) and tazobactam (4 mug/ml) were determined by the dilution method in Mueller-Hinton agar with an inoculum of 104 CFU per spot, as described previously . The MICs of ceftazidime were in a narrow range from 32 to 128 mug/ml. The MICs of ticarcillin ranged from 64 to 2,048 mug/ml. The use of clavulanate did not restore susceptibility to ticarcillin. The MICs of piperacillin were between 16 and 256 mug/ml, and the MICs of aztreonam were between 16 and 128 mug/ml. The strains were moderately resistant to cefepime (MICs, 8 to 64 mug/ml). Fifteen strains were resistant to imipenem (MICs, >8 mug/ml), and of these, three were resistant to meropenem (MICs, >8 mug/ml). Cephalosporinase inhibition, evaluated by the disk diffusion method on Mueller-Hinton agar containing cloxacillin at 250 or 500 mug/ml, was considered significant when the ceftazidime zone diameter increased by >10 mm. P. aeruginosa PAO1 was used as the reference strain . For 25 strains which did not have pI values that might correspond to those for an additional beta-lactamase, ceftazidime activity was partially restored by cloxacillin at 250 mug/ml (n = 15) or only by cloxacillin at 500 mug/ml (n = 10). For these 25 strains without an additional beta-lactamase, the level of production of the AmpC enzymes was measured by determination of their specific beta-lactamase activities. Cultures were grown overnight at 37C in 3 liters of brain heart infusion (Biokar Diagnostics, Beauvais, France). Bacterial suspensions were disrupted by sonication and centrifuged as described previously . The supernatants were dialyzed against 3 liters of 100 mM NaCl overnight at 4C. Then, the specific beta-lactamase activities of the crude extracts were determined by a computerized microacidimetric method described elsewhere with 225 mM benzylpenicillin as the substrate. The total protein concentration in the enzyme preparation was estimated by a protein assay (Bio-Rad, Richmond, Calif.) with bovine serum albumin (Sigma Chemical Co., St. Louis, Mo.) as the standard. One unit of specific beta-lactamase activity was defined as the amount of enzyme which hydrolyzes 1 mumol of benzylpenicillin per min and per mg of protein at 37C and at pH 7. For 21 strains the specific activity was 100- to 1,000-fold greater than that of PAO1, with values ranging from 124 to 15,373 mU mg-1 . Variations in levels of resistance to ceftazidime were weakly related to different levels of enzyme overproduction (i.e., an MIC of 64 mug/ml corresponds to specific activities of 52 mU mg-1 for strain 33 and 15,373 mU mg-1 for strain 5). The ampC genes of the strains that produced five cephalosporinases with different pI values were amplified with primers AMPC-PAC and AMPC-PAD under the following conditions: 94C for 5 min and then 30 cycles of 94C for 1 min, 50C for 3 min, and 72C for 3 min, with a single final extension at 72C for 15 min. The PCR products were sequenced with the primers AMPC-PAD, AMPC-PAG, AMPC-28, and AMPC-PAB1 . Sequence analysis showed a low level of sequence diversity (0.17 to 0.75%) and 99% nucleotide identity with the ampC gene of P. aeruginosa strain PAO1 (GenBank accession no. ). Compared to the PAO1 ampC gene sequence , all ampC genes encoding enzymes with pIs different from 8.0 harbored the mutation A866->G, which led to the substitution Thr105->Ala . TABLE 3 | Substitution in blaAmpC genes according to the pI value for each cephalosporinase and comparison with the sequence of the blaAmpC gene of P. aeruginosa strain PAO1 (pI 8.0) Genes encoding enzymes of pI 8.0 + 8.4 and pI 8.4 harbored a mutation at position 1166 that led to the substitution Val205->Leu. The ampC gene encoding the enzyme of pI 7.8 harbored the mutation G789->A, which led to the substitution Arg79->Gln, and that encoding the enzyme of pI 8.6 harbored the mutation T1080->G, which led to the substitution Leu176->Arg. The Arg79->Gln substitution is associated with a change in charges and may have been responsible for the decrease in the pI value from 8.0 to 7.8, as reported previously . However, enzymes with identical pIs do not necessarily share identical amino acid sequences, and the different pIs suggested that some bands correspond to cleaved derivatives (isoenzymes) of the cephalosporinase, as reported previously . For eight of the nine strains with an additional beta-lactamase, ceftazidime activity was restored only by cloxacillin at 500 mug/ml (strains 18 to 25), and so ceftazidime resistance was very likely due to hydrolysis by the derepressed chromosomal cephalosporinase. For strain 4 the activity of ceftazidime was not restored by cloxacillin at 500 mug/ml. Strain 4 was the only one which had a positive modified disk synergy test result with ceftazidime and clavulanate . The genes of the six beta-lactamases with a pI of 7.0 were amplified with OXA-2 primers (primers OXA-2,3 and OXB-2,3). Nucleotide sequence analysis of the blaOXA PCR products revealed a blaOXA-21 sequence with only one silent mutation (T697->C) in the genes of four isolates compared to the sequence of blaOXA-21 previously reported for Acinetobacter baumannii (EMBL database accession no. ). The six strains that produced OXA-21 had different RAPD patterns (data not shown). To our knowledge, this is the first report of OXA-21 in P. aeruginosa. The blaPSE-1 gene was detected in two strains by PCR with primers OC1 and OC3 , and a protein with a pI of 5.7 conferred a very high level of resistance to ticarcillin (MICs, >2,048 mug/ml) and ticarcillin-clavulanate (MICs, 256/1,024 mug/ml). The blaPER-1 gene of the strain producing the beta-lactamase with a pI of 5.4 was identified by PCR with primers PER-A and PER-B and sequencing. Epidemics caused by strains with this enzyme, previously reported in Turkey and Italy , are now observed in France. The carbapenem resistance observed in 15 strains was not related to enzymatic hydrolysis by a carbapenem-hydrolyzing beta-lactamase. Although several ESBLs were reported in P. aeruginosa, their prevalences were very low in our study. Backmatter: PMID- 12183277 TI - Effect of Lamivudine on Transmission of Human T-Cell Lymphotropic Virus Type 1 to Adult Peripheral Blood Mononuclear Cells In Vitro AB - The effects of lamivudine (3TC) on in vitro infection of peripheral blood mononuclear cells (PBMC) from healthy donors with human T-cell lymphotropic virus type 1 (HTLV-1) were investigated. Direct measures of viral replication (viral DNA, RNA, and protein) all gave similar, very high 50% inhibitory concentrations in comparison with those previously reported for zidovudine. Nevertheless, 3TC inhibited HTLV-1-driven long-term growth of infected PBMC in vitro at concentrations (6.25 muM) which had poor or no direct antiviral effects, suggesting that another mechanism may be playing a role. Keywords: Introduction : Lamivudine (3TC) is a cytosine nucleoside reverse transcriptase inhibitor (NRTI) that is used, in combination with other nucleoside or nonnucleoside reverse transcriptase inhibitors (NNRTIs) or protease inhibitors (PIs), in the treatment of infections caused by human immunodeficiency virus (HIV) . As monotherapy, 3TC is used in hepatitis B virus (HBV) infection . Human T-cell lymphotropic virus type 1 (HTLV-1) is the etiological agent of adult T-cell leukemia/lymphoma (ATL) and of demyelinating diseases characterized by a remarkable inflammatory response, known as HTLV-1-associated human myelopathy/tropic spastic paraparesis (HAM/TSP) . ATL has a poor prognosis and shows high resistance to conventional chemotherapy. Recent studies in vivo have shown the potential therapeutic utility of 3TC in the treatment of pathologies associated with HTLV-1 infection . In particular, 3TC was found to be a relatively safe drug that caused a rapid fall in proviral load in HAM/TSP patients, although phases of decrease in viral load alternated with phases of active replication of the virus . In this study, we have assayed the capacity of 3TC to inhibit infection and long-term growth caused by HTLV-1 in vitro. HTLV-1 transmission was performed by coculturing peripheral blood mononuclear cells (PBMC) isolated from normal adult donors, seronegative for HIV, HBV, and HCV, with lethally irradiated (120 Cy; cesium gamma cell irradiator 1000; Canada Atomic Energy Ltd., Chalk River, Canada) MT-2 cells at a ratio of 5:1 . MT-2 is a cell line chronically infected by HTLV-1 . To assay its antiviral activity, 3TC (Wellcome Research Laboratories, Beckenham, United Kingdom) was added at final concentrations ranging from 6.25 to 100 muM at the onset of the culture and then at half the initial concentrations 3, 7, and 10 days postinfection (p.i.) to ensure a constant level of the drug in the culture medium in the early phase of infection. In preliminary experiments, lower concentrations of 3TC did not exert any antiviral effect against HTLV-1 in vitro. Concentrations of 3TC used were close to the peak levels in serum (0.2 to 20 muM) which are detected in patients treated with pharmacological doses of the drug . The antiviral effects of 3TC at different concentrations were evaluated 4 weeks p.i. by assaying the presence of proviral DNA, viral RNA, and viral protein expression in untreated and treated cell cultures. To detect proviral DNA, 1 mug of DNA was used as a template and was amplified in a standard PCR mix by using a primer pair specific for the pol region (SK54 and SK56) as previously described . To detect viral RNA, total RNA from infected cells was reverse transcribed into cDNA and amplified with primers specific for the HTLV-1 Tax/Rex splicing region . Amplified DNA was analyzed by liquid hybridization and probed with specific 32P-end-labeled oligonucleotides . Three experiments using PBMC from different donors were performed with similar results. Results of one representative experiment are reported in Fig. . The uninfected culture used for this experiment was designated PB-1, while the infected culture was designated PB-1/MT. Addition of 100 muM 3TC caused complete inhibition of proviral DNA , as detected by pol primers and viral RNA , compared to the untreated control. Conversely, 25 and 6.25 muM 3TC partly inhibited the presence of proviral DNA, and 6.25 muM 3TC did not inhibit viral RNA expression. The housekeeping gene glucose-6-phosphate dehydrogenase (G6PDH) was present at equal levels in all the samples for both DNA and RNA. Expression of the virus protein Tax, assayed by Western blot analysis as previously described , was completely inhibited at 100 and 25 muM 3TC, while 6.25 muM 3TC was not inhibitory . Staining with Ponceau dye indicated that equal amounts of proteins were analyzed (data not shown). FIG. 1. | (A) Presence of HTLV-1 proviral DNA in PB-1/MT cell cultures 4 weeks p. (A) Presence of HTLV-1 proviral DNA in PB-1/MT cell cultures 4 weeks p.i. (B) Densitometric analysis of genomic DNA from PB-1/MT cultures. Absorbance values (optical density) refer to amplification of genomic DNA with primers specific for G6PDH. (C) Presence of HTLV-1 viral mRNA in PB-1/MT cultures 4 weeks p.i. (D) Densitometric analysis of total RNA from PB-1/MT cultures 4 weeks p.i. Absorbance values (optical density) refer to amplification with primers specific for G6PDH. (E) Expression of the HTLV-1 protein Tax in PB-1/MT cultures 4 weeks p.i. Western blot analysis of cell lysates was performed. Blots were incubated overnight at 4C with a 1:1,000 dilution of an anti-Tax monoclonal antibody (a generous gift from Elliot Cowan of the U.S. Food and Drug Administration Division of Transfusion-Transmitted Diseases, Bethesda, Md.). (F) Absorbance values (optical density) obtained by densitometric analysis of Tax in PB-1/MT cultures. Lanes for all panels: 1, uninfected control; 2, untreated PB/MT-1 cultures; 3 through 5, PB/MT-1 cultures treated with 100, 25, and 6.25 muM 3TC, respectively. Exposure to HTLV-1 in vitro usually induces proliferation of PBMC, leading in most cases to immortalization of infected cells. Since these are key hallmarks of HTLV-1 pathogenesis, the effects of 3TC on cell proliferation were examined both by a short-term assay and by a long-term growth assessment. To assay proliferation at early times p.i., PBMC were cultivated, alone or with irradiated MT-2 cells, for 7 days in the presence of interleukin 2 (IL-2) without any other stimulation. 3TC was added at the onset of the cultures at concentrations ranging from 2 to 800 muM. Proliferation was measured by [3H]thymidine (Amersham Pharmacia Biotech, Little Chalfont, Buckinghamshire, United Kingdom) uptake during the last 18 h in culture. Figure shows the results of one representative experiment in quadruplicate, out of four performed with cells from different donors. In uninfected cultures, 800 muM 3TC significantly inhibited cell proliferation (P = 0.001 by the two-sided Dunnett's pairwise multiple-comparison t test). There were no significantly different effects at the other 3TC concentrations . Cultures were promptly induced to proliferate following HTLV-1 infection (twofold increase), as shown by comparing thymidine uptake in untreated PB-1/MT versus untreated PB-1 cultures . In infected cultures, high concentrations of 3TC (100 to 800 muM) caused marked inhibition of thymidine uptake (P < 0.001 relative to untreated infected cultures by the two-sided Dunnett's pairwise multiple-comparison t test). In order to examine the consequences of 3TC treatment on the long-term growth of HTLV-1-infected PBMC, cells that were either left untreated or treated with 3TC were maintained in culture for 13 weeks. Cell growth was monitored weekly by evaluating the living-cell count by use of the trypan blue dye exclusion test, calculated as the mean of two independent evaluations. After counting, cell concentrations were adjusted to 1 x 106/ml. Results were expressed as total cell number (TCN), calculated from weekly cell counts as follows. For the first week, TCN was the number of cells (expressed as millions of viable cells per milliliter) detected before the first adjustment. On successive weeks, TCN was calculated as the TCN at the previous weekly passage multiplied by the number of cells (expressed as millions of viable cells per milliliter) detected before weekly adjustment. Figure shows the mean TCNs +- standard deviations (SD) from three different experiments using PBMC from three donors. Uninfected cultures grew poorly when maintained in IL-2 without additional stimuli throughout the time of observation (Fig. , C, and E). Infected cultures treated with 100 or 25 muM 3TC ceased to grow after 5 and 7 weeks, respectively . Infected cultures treated with 6.25 muM 3TC survived to 9 weeks p.i. . Conversely, all infected untreated cultures were still growing at week 13, when the experiments were interrupted. FIG. 2. | Effects of different concentrations of 3TC on the proliferative responses of PB-1 cell cultures left unexposed (A) or exposed to HTLV-1 (PB-1/MT) (B). Effects of different concentrations of 3TC on the proliferative responses of PB-1 cell cultures left unexposed (A) or exposed to HTLV-1 (PB-1/MT) (B). Results are expressed as mean counts per minute of radioactivity +- SD from quadruplicate samples. Bars: 1, untreated PB-1 or PB-1/MT; 2 through 9, cultures treated with 3TC at 2 muM (bar 2), 6.25 muM (bar 3), 25 muM (bar 4), 100 muM (bar 5), 200 muM (bar 6), 400 muM (bar 7), 600 muM (bar 8), or 800 muM (bar 9). Results were obtained from a single donor, designated PB-1, representative of four donors tested. FIG. 3. | Growth curves of uninfected PBMC (A, C, and E) and HTLV-1 infected PBMC (B, D, and F) in the presence of different concentrations of 3TC. Growth curves of uninfected PBMC (A, C, and E) and HTLV-1 infected PBMC (B, D, and F) in the presence of different concentrations of 3TC. Each experimental point represents the mean TCN +- SD determined weekly in cultures from three different experiments using PBMC from three donors. The growth curve of untreated cultures is reported in all graphs, for visual comparison. By one-way analysis of variance, significant differences in HTLV-1-infected PBMC among groups were determined starting by week 3, while no differences in uninfected PBMC among groups were found at any times tested. Different measures of the antiviral effects of 3TC in HTLV-1 infection were compared by calculating the 50% inhibitory concentrations (IC50s) for proviral DNA, viral RNA, and Tax detection, as previously shown for other NRTIs . Means +- SD of three IC50 evaluations, calculated from all data obtained in three experiments performed using PBMC from different donors, were as follows: 29 +- 2 muM for proviral DNA (pol), 24 +- 2 muM for viral RNA (tax/rex), and 14 +- 1 muM for Tax protein (the Pearson coefficient [r2] for the y value was 0.9). IC50s were calculated according to the best-fit curve (y value versus log x, where y is the value of the function examined and x is the drug concentration). These results show that IC50s calculated from proviral DNA and viral RNA assays were about twice as high as the IC50 for Tax expression, suggesting a possible effect of 3TC at the translational level. 3TC completely inhibited proviral DNA, viral mRNA, and protein expression only at a high concentration (100 muM). This concentration is out of the range which can be considered pharmacological on the basis of clinical experience with HIV infection. Moreover, the antiviral potency of 3TC against HTLV-1 transmission is more than 100 times lower than that of zidovudine (AZT), as determined by comparing IC50s presented in this paper with those previously reported . This is different from what has been observed for HIV infection, where inhibitory concentrations of 3TC and AZT are similar. The discrepancy between the behaviors of HIV and HTLV-1 when subjected to 3TC treatment might be due to a different capacity of 3TC to interact with the reverse transcriptase (RT) of HIV and HTLV-1, respectively. This hypothesis is strongly supported by biochemical results showing high-level resistance to 3TC in HTLV-1 RT by use of an enzymatic assay . The YMDD site, the most frequent source of mutations conferring resistance to 3TC in HIV , seems well conserved in the RT of HTLV-1. Therefore, a site other than YMDD could be responsible for a possible low conformational affinity of 3TC to HTLV-1 RT. While 6.25 muM 3TC showed no signs of antiviral or short-term antiproliferative activity, this dose was nevertheless able to arrest growth in the long-term assay of infected cells. This fact suggests that 3TC may counteract long-term growth by mechanisms other than antiviral effect, possibly affecting HTLV-1-infected-cell activity. Inhibition of long-term cell growth in infected cells might be due to relatively high-level incorporation and, possibly, phosphorylation of 3TC in cells chronically infected by the virus. On the other hand, protection against immortalization by 6.25 muM 3TC could be related to the partial, if not complete, inhibitory action on Tax transactivating protein expression caused by the drug at this concentration. Inhibition of long-term cell growth in HTLV-1-infected cells in vitro could explain, at least in part, the effects exerted by 3TC in HTLV-1-infected patients in vivo. Another noteworthy point is that 3TC, in spite of its ability to inhibit long-term growth of HTLV-1-infected cultures, was shown to be less potent than AZT in inhibiting the proliferation of either uninfected or freshly HTLV-1-infected cultures, based on a historical comparison with previous results . This observation indicates a low level of immunotoxicity for this drug, which could be potentially beneficial for patients undergoing therapy. Taken together, these data demonstrate that 3TC has little in vitro antiviral activity against HTLV-1 infection at clinically relevant concentrations and therefore is unlikely to have direct antiviral efficacy in HTLV-1-infected patients. Rather, indirect effects could explain its limited clinical benefit in HTLV-1 infection in vivo. Backmatter: PMID- 12183242 TI - Efficacy of Quinupristin-Dalfopristin in Preventing Vascular Graft Infection Due to Staphylococcus epidermidis with Intermediate Resistance to Glycopeptides AB - A rat model was used to investigate the efficacy of quinupristin-dalfopristin (Q-D) in the prevention of vascular prosthetic graft infection due to methicillin-resistant Staphylococcus epidermidis with intermediate resistance to glycopeptides. The in vitro activity of the compound was compared to that of vancomycin by MIC determination and time-kill study. Moreover, the efficacy of collagen-sealed Q-D-soaked Dacron was evaluated in a rat model of graft infection. Graft infections were established in the subcutaneous tissue of the backs of 120 adult male Wistar rats. The in vivo study included a control group, one contaminated group that did not receive any antibiotic prophylaxis, two contaminated groups that received grafts soaked with 10 and 100 mug of Q-D per ml, respectively, and two contaminated groups that received grafts soaked with 10 and 100 mug of vancomycin per ml, respectively. Rats that received Dacron grafts soaked with 100 mug of Q-D per ml showed no evidence of infection (<10 CFU/ml). In contrast, for rats that received Dacron grafts soaked with 10 mug of Q-D per ml and Dacron grafts soaked with 10 or 100 mug of vancomycin per ml, the quantitative graft cultures demonstrated 2.2 x 102 +- 1.3 x 102, 2.2 x 106 +- 1.9 x 105, and 5.6 x 102 +- 0.3 x 102 CFU/ml, respectively. Taken together the results of the study demonstrate that the use of Dacron grafts soaked with Q-D can result in significant bacterial growth inhibition and show that this compound is potentially valuable for prevention of vascular prosthetic graft infection. Keywords: Introduction : The increasing use of foreign material in many fields of modern surgery is associated with a definitive risk of bacterial infection. Coagulase-negative staphylococci, chiefly the skin commensal Staphylococcus epidermidis, are among the most common pathogens that cause biomaterial infections and the most frequent cause of late-appearing vascular graft infections in humans . Vascular prosthetic graft infection is one of the most feared complications that the vascular surgeon treats; although most cases are resolved by removal of the device and antibiotic therapy, serious complications may result, including organ failure, amputation, metastatic infection, and death. Effective strategies for the prevention of prosthetic infection vary from device to device. The centerpieces of prophylaxis are asepsis and perioperative administration of systemic antibiotics . As adjunctive prophylaxis, in the case of vascular grafts, the use of antimicrobials that bind to prosthetic grafts at high concentrations has been proposed . Since the emergence of methicillin-resistant staphylococci, glycopeptides have often been the only effective drugs. For this reason, the emergence of staphylococcal strains exhibiting reduced sensitivities to vancomycin is of particular concern. The emergence of vancomycin resistance in coagulase-negative staphylococci was described in the 1980s. For this reason, new strategies are needed to treat infections caused by these multidrug-resistant organisms and to reduce the increasing selection pressure of antibiotics on gram-positive pathogens . Various new agents have been demonstrated to have significant in vitro activities against staphylococci. One of these compounds is the new semisynthetic injectable streptogramin quinupristin-dalfopristin (Q-D), composed of two components which may act synergistically: quinupristin, a peptide macrolactone classified as a type B streptogramin, and dalfopristin, a polyunsaturated macrolactone classified as a type A streptogramin, in a 30:70 ratio. It has a focused spectrum of in vitro activity against gram-positive cocci, including multidrug-resistant isolates of staphylococci, streptococci, and Enterococcus faecium . In this study we used one strain of S. epidermidis with intermediate resistance to vancomycin to investigate the in vitro activity of Q-D and its in vivo efficacy when it was bound to a Dacron graft for the prevention of prosthesis infection in a rat model. MATERIALS AND METHODS : In vitro studies. (i) Organisms. | The strain of methicillin-resistant S. epidermidis with intermediate resistance to vancomycin used in this study was isolated from a clinical specimen submitted for routine bacteriological investigation to the Institute of Infectious Diseases and Public Health, University of Ancona, Ancona, Italy. This isolate is described by use of the acronym VISE (vancomycin-intermediate S. epidermidis). Commercially available S. epidermidis ATCC 12228 was used as a quality control strain in the in vitro investigations. (ii) Drugs. | Q-D was obtained from Aventis Pharma, Centre de Recherches, Vitry-Alfortville, France. Vancomycin was obtained from Sigma-Aldrich, Milan, Italy. Laboratory powders were diluted in accordance with the recommendations of the manufacturers. Solutions of drugs were made fresh on the day of assay or were stored at -80C in the dark for up to 20 days. (iii) Antimicrobial susceptibility testing. | The antimicrobial susceptibilities of the strains to Q-D and vancomycin were determined by the broth microdilution method described by the National Committee for Clinical Laboratory Standards (NCCLS) . In addition, the strains were tested for their susceptibilities to vancomycin by the NCCLS reference disk diffusion method with 30-mug vancomycin discs . Experiments were performed in triplicate. (iv) Time-kill studies. | To perform time-kill studies, the VISE strain was grown at 37C in Mueller-Hinton (MH) broth. When the bacteria were in the log phase of growth the suspensions were centrifuged at 1,000 x g for 15 min, the supernatants were discarded, and the bacteria were resuspended and diluted with sterile saline to achieve a concentration of 5 x 1010 CFU/ml saline. The organisms were resuspended in fresh MH broth at approximately 5 x 105 cells/ml and exposed to Q-D and vancomycin (one and four times the MIC) for up to 24 h at 37C. Throughout the experiments, triplicate samples (0.1 ml) were withdrawn after 0, 1, 3, 6, and 24 h of incubation at 37C. Up to seven 10-fold dilutions were made in MH broth from each sample, and, finally, the dilutions were spread onto MH agar plates and incubated for up to 72 h at 37C to obtain viable colonies. The limit of detection for this method was approximately 10 CFU/ml. In preliminary experiments, antibiotic carryover was ruled out by plating samples of bacterial suspensions in the presence or absence of antibiotics. In vivo studies. (i) Rat model. | Adult male Wistar rats (weight range, 250 to 300 g) were studied. The study included a group with no graft contamination and no local antibiotic treatment (uncontaminated control group), one contaminated group that did not receive any local antibiotic treatment (untreated control group), two contaminated groups that received grafts soaked with 10 and 100 mug of Q-D per ml, respectively, and two contaminated groups that received grafts soaked with 10 and 100 mug of vancomycin per ml, respectively. Each group included 20 animals. Rats were anesthetized with ether, the hair on the back was shaved, and the skin was cleansed with 10% povidone-iodine solution. One subcutaneous pocket was made on each side of the median line through a 1.5-cm incision. Aseptically, 1-cm2 sterile collagen-sealed Dacron grafts (Albograft; Sorin Biomedica Cardio, Saluggia VC, Italy) were implanted into the pockets. Before implantation, the Dacron graft segments were impregnated with Q-D and vancomycin, each at concentrations of 10 and 100 mug/ml. Immediately before implantation the grafts were soaked for 20 min in a sterile solution of the agents mentioned above. The pockets were closed with skin clips, and sterile saline solution (1 ml) containing the VISE strain at a concentration of 2 x 107 CFU/ml was inoculated onto the graft surface by using a tuberculin syringe to create a subcutaneous fluid-filled pocket . The animals were returned to individual cages and thoroughly examined daily. All grafts were explanted at 7 days following implantation. This study was approved by the Animal Research Ethics Committee of the Istituto Nazionale Riposo e Cura Anziani Istituto di Ricovero e Cura a Carattere Scientifico, University of Ancona. (ii) Assessment of infection. | The explanted grafts were placed in sterile tubes, washed in sterile saline solution, placed in tubes containing 10 ml of phosphate-buffered saline solution, and sonicated for 5 min to remove the adherent bacteria from the grafts. Quantitation of viable bacteria was performed by culturing serial 10-fold dilutions (0.1 ml) of the bacterial suspension onto blood agar plates. All plates were incubated at 37C for 48 h and evaluated for the presence of the VISE strain. The organisms were quantitated by counting the number of CFU per plate. The limit of detection for this method was approximately 10 CFU/ml. Statistical analysis. | MICs are presented as the modes of three separate experiments. Quantitative culture results regarding the in vivo experiments are presented as the means +- standard deviations of the means. The results were compared by analysis of variance of the log-transformed data by the Tukey-Kramer honestly significant difference test. Significance was accepted when the P value was <=0.05. RESULTS : In vitro data. | According to the broth microdilution method recommended by the NCCLS, vancomycin MICs were 0.25 and 8 mug/ml for S. epidermidis ATCC 12228 and the VISE strain, respectively, while Q-D MICs were 0.50 and 1 mug/ml, respectively. The different patterns of susceptibility were confirmed by the disk diffusion test: S. epidermidis ATCC 12228 showed zone sizes of 18 mm with the vancomycin disc, while the intermediate resistance of the VISE strain to vancomycin was demonstrated by a zone size of 11 mm. Q-D at 2 mug/ml (four times the MIC) produced reductions in bacterial counts of 0.1, 0.3, 1.1, and 1.8 log10 CFU/ml after 1, 3, 6, and 24 h of incubation, respectively. This killing was comparable to that obtained with Q-D at 0.5 mug/ml (one time the MIC). Vancomycin at 32 mug/ml (four times the MIC) produced reductions in bacterial counts of 0.2, 0.5, 1.0, and 1.6 log10 CFU/ml after 1, 3, 6, and 24 h of incubation, respectively, while vancomycin at 8 mug/ml (one time the MIC) exhibited a low level of bactericidal activity, producing reductions in bacterial counts of 0.0, 0.1, 0.4, and 0.6 log10 CFU/ml, respectively (data not shown). In vivo studies. | None of the animals included in the uncontaminated control group had microbiological evidence of graft infection. On the contrary, all 20 rats included in the untreated control group demonstrated evidence of graft infection, with quantitative culture results showing 5.1 x 106 +- 8.8 x 105 CFU/ml. Interestingly, only the group with Dacron grafts soaked in 100 mug of Q-D per ml showed no evidence of staphylococcal infection (<10 CFU/ml). In contrast, the quantitative graft cultures for the rats that received Dacron grafts soaked in 10 mug of Q-D per ml and Dacron grafts soaked in vancomycin demonstrated bacterial growth . None of the animals included in any group died or had clinical evidence of drug-related adverse effects, such as local signs of perigraft inflammation, anorexia, vomiting, diarrhea, or behavioral alterations. There were significant differences in the results of the quantitative bacterial graft cultures when the data obtained for the antibiotic-treated groups were compared with those obtained for the contaminated control group (P < 0.05), with the exception of the group treated with grafts soaked in 10 mug of vancomycin per ml. When data obtained for each group with grafts soaked in an antibiotic were compared with those obtained for any other group with a graft soaked in an antibiotic, the differences were always statistically significant for grafts soaked in 100 mug of Q-D per ml (P < 0.05). Finally, no statistically significant differences were observed between the group with grafts soaked in 10 mug of Q-D per ml and the group with grafts soaked in 100 mug of vancomycin per ml. TABLE 1 | Efficacies of Q-D and vancomycin against a glycopeptide-intermediate S. epidermidis strain causing graft infection in a rat model DISCUSSION : Debate continues over the best treatment for prosthetic vascular graft infections. S. epidermidis is the most important pathogen causing infections related to implanted foreign bodies. It has been recovered from the skin, subcutaneous fat, lymph nodes, and arterial walls of more than one-third of individuals undergoing vascular reconstruction, despite the use of aseptic vascular surgical technique and prophylactic antibiotics. Most important in the pathogenesis of foreign body-associated infection due to this organism is the colonization of the polymer surface by formation of a biofilm . For this reason, prevention through effective antibiotic prophylaxis plays a pivotal role in the control of these infections and has an important impact on patient mortality and the cost-effectiveness of hospital care. The success of prophylactic antibiotics during surgery is dependent on the pharmacokinetics of the antibiotic in tissue and the maintenance of adequate levels of the antibiotic in tissue for the duration of the vascular surgical procedure. Nevertheless, errors in the sterilization procedures and the increases in the incidence and the levels of resistance of S. epidermidis can predispose individuals to prosthesis infections . Glycopeptides such as vancomycin are used parenterally to treat infections caused by gram-positive bacteria, especially staphylococcal infections after the emergence of methicillin-resistant staphylococci. Recently, they have been administered as perioperative antibiotic prophylaxis . Nevertheless, the recent emergence of glycopeptide resistance in coagulase-negative staphylococci heightens concern about the need for other antistaphylococcal agents . Analysis of the data from in vitro studies shows that Q-D and vancomycin had similar activities against the control strain, S. epidermidis ATCC 12228, while Q-D exhibited a higher level of activity than vancomycin against the VISE clinical strain. In fact, the broth microdilution method and the disk diffusion method recommended by the NCCLS showed that vancomycin exerted intermediate activity against the VISE strain, although there are still differences between the current NCCLS interpretive standards and the recommendations made to define categories of susceptibility to glycopeptides in some other countries . The in vivo results were similar to those reported by other investigators , who found that the use of Dacron grafts soaked in an antibiotic can result in significant bacterial growth inhibition, even though high concentrations of organisms were topically inoculated on the Dacron prostheses. Actually, statistical analysis demonstrated that any prophylactic antibiotic treatment except vancomycin at 10 mug/ml was useful. Nevertheless only Q-D at the highest concentration tested inhibited bacterial growth within the limits of detection. Finally, no statistically significant differences were observed between the groups that received grafts soaked with 10 mug of Q-D per ml and those that received grafts soaked with 100 mug of vancomycin per ml. However, it is noteworthy that Q-D and vancomycin are physically different compounds and the quantitative results could be affected by the lack of equivalent abilities of the two antibiotics to coat the artificial materials. It has long been recognized that the treatment of serious bacteremia caused by staphylococci, such as vascular graft infection, typically requires the use of agents with bactericidal activities against the organism. This fact is likely emphasized by the recent emergence in clinical settings of multidrug-resistant bacteria that belong, for the most part, to the staphylococcal species. New therapeutic options are needed, although experience demonstrates that every new antimicrobial agent introduced into clinical practice can be plagued by the emergence of organisms resistant to its effect. The strong in vitro activity and the prophylactic in vivo efficacy demonstrated by Q-D against a staphylococcal strain with decreased susceptibility to the glycopeptides used in the present study make Q-D potentially useful for future topical antimicrobial treatments, such as perioperative chemoprophylaxis in prosthetic surgery. Backmatter: PMID- 12183236 TI - Effect of 5-Iodo-2'-Deoxyuridine on Vaccinia Virus (Orthopoxvirus) Infections in Mice AB - There is a concern that there may be unregistered stocks of smallpox that can be used for bioterrorism or biological warfare. According to the WHO Advisory Committee on Variola Research, there is a need to develop strategies to treat smallpox infections should they reappear. It would also be important to have an effective drug at hand for the treatment of monkeypox disease in humans. We show here that 5-iodo-2'-deoxyuridine (IDU) is a potent inhibitor of vaccinia virus (VV) replication and that IDU inhibits VV DNA synthesis in a dose-dependent way. The in vivo protective effect of IDU was assessed in the VV tail lesion model in immunocompetent mice and in a lethal model for VV infection in SCID (severe combined immune deficiency) mice that had been infected either intranasally, intraperitoneally, or intravenously. Subcutaneous treatment with IDU at 150 and 100 mg/kg of body weight markedly reduced the number of tail lesions in immunocompetent NMRI mice. Untreated intranasally VV-infected SCID mice died at 20.8 +- 3.1 days after infection (mean +- standard deviation). Treatment with IDU (subcutaneously, 150 mg/kg/day [from day 0 to 4] and 75 mg/kg/day [from day 6 to 11]) delayed-virus induced mortality by 15 days (mean day of death +- standard deviation, 35.8 +- 6.7; P < 0.0001). This protective effect was associated with (i) an improvement of lung histology and (ii) a marked reduction in lung viral titers. IDU also delayed VV-induced mortality when mice had either been infected intraperitoneally or intravenously. Even when the start of treatment with IDU (in intraperitoneally VV-infected mice) was postponed until 2 or 4 days after infection, an important delay in virus-induced mortality was noted. Keywords: Introduction : There is an increasing concern that the smallpox virus may be used for biological or terroristic purposes. All vaccination programs against smallpox were discontinued after eradication of the disease. Virtually all children and many adults are now fully susceptible to smallpox. There is only a very limited stock of vaccine available, which may not have been at all properly stored or monitored for potency. It is being studied whether diluting (1/5 or 1/10) the remaining vaccine stocks would be possible so as to increase the number of vaccines available. Novel stocks of vaccine are currently being prepared . If smallpox were used in an act of terrorism or warfare, it could, in a highly mobile and susceptible population, cause a real catastrophe. A World Health Organization (WHO) advisory committee on variola recommended in 1999 the development of drugs to treat human smallpox infections should they reappear (press release WHO/77, 10 December 1999). Human infection with monkeypox occurs sporadically in parts of Western and Central Africa. In 1996 and 1997 an important outbreak of monkeypox occurred in humans in the Democratic Republic of Congo . Also there have recently been more outbreaks of monkeypox disease in humans. It is thus also important to have an effective drug at hand for the treatment of monkeypox disease in humans. Inhibitors of orthopoxviruses are also of interest for the treatment of molluscum contagiosum or disseminated vaccinia (, ; J. C. Guillaume, P. Saiag, J. Wechsler, M. C. Lescs, and J. C. Roujeau, Letter, Lancet 337:1034-1035, 1991). We reported earlier on the potent anti-vaccinia virus (anti-VV) activity of the acyclic nucleoside phosphonate analog cidofovir, both in immunocompetent and immunodeficient mice . VV belongs, akin to smallpox and monkeypox viruses, to the orthopoxviruses and represents a valuable surrogate for the latter viruses . Our findings on the activity of cidofovir in the vaccinia model were corroborated by the recent observation that cidofovir also protects mice from a lethal aerosolized or intranasal cowpox virus or VV challenge, either when given to the mice systemically or when they are intranasally infected . The compound was also shown to inhibit in vitro the replication of 31 different strains of variola (according to reference ) and to rescue nonhuman primates from large quantities (1,000 50% lethal doses) of aerosolized monkeypox (data not shown in reference ). Cidofovir has also proven highly effective against parapoxvirus infections in vitro and in vivo, in particular in a case of orf in an immunocompromised patient , and against molluscum contagiosum (E. G. Davies, A. Thrasher, K. Lacey, and J. Harper, Letter, Lancet 353:2042, 1999). Recently, we demonstrated that the acyclic nucleoside analogue 2-amino-7-[1,3-dihydroxy-2-propoxy)methyl]-purine (S2242) in its diacetyl ester prodrug form protected SCID mice against VV-induced mortality, when given orally . Here we describe the protective effect of 5'-iodo-2'-deoxyuridine (IDU) (under the trade names Herpid, Stoxil, Idoxene, and Virudox), a well-known inhibitor of herpesvirus replication, against VV infections in immunodeficient mice. MATERIALS AND METHODS : Virus, cells, and compounds. | VV (Copenhagen strain) was obtained from the Rijksenstofinrichting (Brussels, Belgium). This virus was formerly used in Belgium as a vaccine against smallpox. Human embryonic lung (HEL) cells were obtained from the American Type Culture Collection. IDU was obtained from Sigma (St. Louis, Mo.). Cidofovir was synthesized at, and provided by, Gilead Sciences (Foster City, Calif.). In vitro antiviral assays. | Cultures of HEL cells grown in 96-well plates were infected for 2 h at 37C with 100 PFU of VV. Following a 2-h adsorption period, the virus inoculum was removed and the cultures were incubated at 37C with different dilutions of the test compounds in modified Eagle medium containing 2% fetal calf serum. At 5 days postinfection, cultures were fixed with 70% ethanol and stained with a 2% Giemsa solution, and virus-induced cytopathic effects (CPEs) were recorded microscopically. The 50% effective concentration (EC50) was defined as the concentration that causes 50% inhibition of virus-induced CPE. All experiments represent mean values for two to four independent determinations. Cytotoxic and cytostatic determinations. | Cytotoxicity was assessed in confluent HEL cell cultures that had been incubated for 5 consecutive days with serial dilutions of the test compounds and was monitored microscopically or by means of the MTS method (Promega). The tetrazolium salt MTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium], when applied to living cells, is converted to a colored formazan. The cytostatic potential of the compounds was assessed in exponentially growing HEL cells. Briefly, HEL cells were seeded in 96-well plates at a density of 4,000 cells per well. At 6 to 8 h after seeding, the cells were incubated with serial dilutions of the test compounds. Cultures were then incubated for 3 days, at which time cell proliferation was determined by the MTS method (Promega). All experiments are mean values for two to four independent determinations. Determination of viral DNA levels. | Confluent cultures of HEL cells grown in 25-cm2 culture flasks were inoculated with VV at a multiplicity of infection of 0.2 for 2 h. After the virus had been removed, cultures were either treated with the test compounds or left untreated. At day 5 postinfection, at which time untreated infected cultures showed 100% CPE, cells were collected, total cellular DNA was extracted (Qiagen blood kit), and 10 mug of heat denatured total cellular DNA was blotted onto a nylon membrane (Hybond-N; Amersham). Following UV cross-linking, prehybridization was carried out for 1 h at 42C. A digoxigenin-11-dUTP-labeled VV-specific probe was generated as described previously . Hybridization was carried out for 18 h at 42C with 30 ng of the digoxigenin-11-dUTP labeled probe per ml. Following washes at high and low stringency, filters were incubated with antidigoxigenin antibody conjugated with alkaline phosphatase (Boehringer Mannheim). Detection of chemiluminescence was performed by standard methods, and films were scanned densitometrically. Inoculation, treatment, and evaluation. | SCID mice (bred at the Rega Institute under specific-pathogen-free conditions) weighing about 15 g, received inoculations intranasally (following sedation), intravenously (in the tail vein), or intraperitoneally with, respectively, 20, 50, or 200 mul of VV (stock of similar105 PFU/ml). IDU or (S)-1-(3-hydroxy-2-phosphonylmethoxypropyl)cytosine (HPMPC) (cidofovir) (for treatment schedules, see Figures and Tables) was administered by subcutaneous injection (in 0.2 ml of phosphate-buffered saline). Mortality was recorded daily. Naval Medical Research Institute (NMRI) mice (Harlan) weighing 17 to 20 g were inoculated in the tail vein with 4 x 103 PFU of VV. Animals were treated by subcutaneous injection with IDU or cidofovir. The number of tail lesions (vesiculae or spots) was recorded at days 6 and 8 postinfection. Statistically significant differences in the mean day of death (MDD) (for SCID mice) or the number of tail lesions (for NMRI mice) was assessed by means of Student's t test. The animal experiments were approved by the Ethical Committee on Vertebrate Animals of the University of Leuven. Determination of viral titers in lungs. | At day 5, 10, 15, 20, or 28 postinfection, lungs from (for each time point) three infected animals that had either been treated with IDU or HPMPC or that had been left untreated were dissected aseptically. Tissue homogenates (10%, wt/vol) were prepared in 2% modified Eagle medium and were titrated (fivefold serial dilutions) on HEL cells. Virus-induced CPE was recorded 5 days later. Histology. | SCID mice (similar15 g) that had been inoculated intranasally with VV and that had either been left untreated or that were treated subcutaneously with IDU (150 mg/kg/day from day 0 to 4 and 75 mg/kg/day from day 7 to 11) or with cidofovir (at 25 mg/kg/day from day 0 to 4 and from day 7 to 11) were killed by ether anesthesia; lungs were dissected, fixed (in Bouin), embedded in paraffin, and sectioned. Sections were stained with hematoxylin-eosin. RESULTS : In vitro anti-VV activity of IDU. | The in vitro anti-VV activities of IDU and HPMPC in HEL cells were compared. As can be derived from Fig. , IDU efficiently inhibited VV-induced CPE formation in fibroblasts with an EC50 of 0.4 mug/ml; the EC50 of HPMPC was 2.6 mug/ml. Viral DNA synthesis was monitored in parallel . A dose-dependent inhibition of VV DNA synthesis was observed, with EC50 values of 0.3 mug/ml for both IDU and HPMPC. The concentration that causes a >=20% reduction in cell metabolism (i.e., mitochondrial activity) as assessed by the MTS method was >100 mug/ml for both IDU and HPMPC. Cytostatic effects of HPMPC and IDU were assessed in uninfected exponentially growing HEL cells. The 50% cytostatic concentrations of IDU and HPMPC were 69 +- 36 and 42 +- 15 mug/ml, respectively (means +- standard deviations). IDU and HPMPC thus yielded comparable selectivity indices (for IDU, 230; for HPMPC, 140) when inhibition of viral DNA synthesis was used as a parameter for drug efficacy. Effect of IDU on VV-induced tail lesions in NMRI mice. | Immunocompetent NMRI mice received treatment with IDU at 150, 100, or 75 mg/kg/day from day 0 to 5 postinfection. Typical tail lesions (first spots, then vesiculae) developed in control animals. The protective effects of IDU and HPMPC were assessed at day 6 and 8 postinfection. On day 6 postinfection, no vesiculae had developed on the tail of mice that had been treated with IDU at 150 or 100 mg/kg/day, whereas all untreated animals showed large numbers of lesions on the tail at that time. IDU still resulted in a marked reduction in the number of vesiculae on day 8 postinfection . Treatment with IDU at 150 or 100 mg/kg/day resulted in a reduction of body weight of, respectively, 17 and 6% on day 8 postinfection (compared to the body weight at the start of the experiment). Yet, the general condition of the animals appeared to be all right. Effect of IDU on VV-induced morbidity and mortality in intranasally infected SCID mice. | SCID mice that had been infected intranasally became sick (ruffled fur, cachexia) at about day 14 postinfection and died at 20 days postinfection (MDD +- standard deviation, 20.8 +- 3.1) . When IDU was administered subcutaneously (150 mg/kg/day from day 0 to 4 and 75 mg/kg/day from day 7 to 11), virus-induced mortality was delayed by 15 days (MDD, 35.8 +- 6.7; p < 0.00005). This particular dosing schedule allowed an "aggressive" induction dose with IDU, followed by a lower dosage that was also effective but that had less of an effect on body weight. HPMPC, which was included as a reference compound at 25 mg/kg/day (same treatment schedule as for IDU), delayed virus-induced mortality by 38 days (MDD, 55 +- 5.9). In a parallel set of experiments viral titers were measured in the lungs of treated and untreated animals at 5, 10, 15, 20, and 28 days postinfection (lungs of three animals were pooled) (data not shown). Titers increased steadily in the lungs of untreated animals from day 5 to 15 postinfection (2.0 x 104, 6.0 x 104, and 2.4 x 105 PFU/g tissue, respectively). None of the untreated animals survived after day 20. In the IDU-treated animals, viral titers were below the detection limit for the first 15 days. Some infectious virus was detected (similar250 PFU/g tissue) at day 20. At day 28 postinfection viral titers in the lungs were comparable (2.8 x 105 PFU/ml) to titers in the lungs of untreated animals at day 15 postinfection. No virus was detectable in HPMPC-treated animals up to 28 days postinfection. Histologically, the lungs of intranasally infected animals (at 18 days postinfection) showed a patchy pattern of mild infiltration of alveolar walls by neutrophils. Exudation of leukocytes into alveolar spaces was minimal or even absent. Some distal bronchioli exhibited an infiltration of the mucosa with neutrophils. Lungs of IDU- or HPMPC-treated animals (at 18 day postinfection) were histologically indistinguishable from lungs of uninfected animals. Effect of IDU on VV-induced mortality in intraperitoneally or intravenously infected SCID mice. | SCID mice that had been inoculated intravenously died at about 18 days postinfection . Mortality of mice that had been treated with IDU at 100 mg/kg/day from day 0 to 4 postinfection and at 50 mg/kg/day from day 7 to 11 postinfection was delayed by 9 days. A similar treatment schedule with IDU at 50 mg/kg/day resulted in some delay in virus-induced mortality, although this was statistically not significant (P = 0.07). SCID mice that had been infected intraperitoneally, and that were left untreated, died about 1 month postinfection (MDD, 33.6 +- 11.6) . Treatment with IDU at 100 mg/kg/day, from day 0 to 4 and at 50 mg/kg/day from day 7 to 11, day 14 to 18, day 21 to 25, and day 28 to 32 delayed virus-induced mortality by almost 1 month (MDD 60.4 +- 20.6). Four out of five animals that had been treated with IDU at 250 mg/kg from day 0 to 2 and at 100 mg/kg from day 7 to 11, day 14 to 18, day 21 to 25, and day 28 to 32 survived the infection until 135 days postinfection, at which time they were sacrificed. No virus was recovered from their organs at that time. Effect of delayed start of treatment with IDU on VV-induced mortality in intraperitoneally infected SCID mice. | Even when start of treatment with IDU in intraperitoneally infected animals was delayed for 2 or 4 days postinfection, a significant delay in virus-induced mortality was still noted, although the protective effect was less pronounced than when treatment was initiated at the day of infection . When the start of treatment was delayed for 7 days postinfection, no significant delay in virus-induced mortality was noted. Effect of IDU treatment on growth of mice. | We documented the effect of treatment with IDU on the growth of young SCID mice . Mice received treatment schedules as used in the experiment depicted in Fig. and in the experiments presented in Tables and 3. Treatment with IDU at 150 mg/kg/day (day 0 to 4) and 75 mg/kg/day (day 7 to 11) had an important effect on the growth of the animals. All mice survived, however, and their body weight slowly increased following cessation of therapy. Lower doses of drugs that still proved antivirally active (Fig. ; Tables and ) were well tolerated and resulted in a less-pronounced growth retardation. FIG. 1. | In vitro anti-VV activity of IDU or HPMPC, as assessed by recording virus-induced CPE (upper panel) or viral DNA synthesis (lower panel). In vitro anti-VV activity of IDU or HPMPC, as assessed by recording virus-induced CPE (upper panel) or viral DNA synthesis (lower panel). Symbols: , IDU; , HPMPC. All data are mean values +- standard deviations (error bars) for two to four independent determinations. FIG. 2. | Effect of IDU or HPMPC on VV-induced mortality in intranasally infected SCID mice. Effect of IDU or HPMPC on VV-induced mortality in intranasally infected SCID mice. Data are mean values from two independent experiments. IDU was administered subcutaneously at 150 mg/kg/day from day 0 to 4 and at 75 mg/kg/day from day 7 to 11, and HPMPC was administered at 25 mg/kg/day from day 0 to 4 and from day 7 to 11. Symbols: , untreated controls (n = 10); *, IDU-treated mice (n = 10); , HPMPC-treated mice (n = 9). FIG. 3. | Change in body weight of SCID mice (five mice per condition) that had been treated twice daily by subcutaneous injection with IDU using the following treatment schedules: IDU at 150 mg/kg/day on days 0 to 4 and 75 mg/kg/day on days 7 to 11 , IDU at 100 mg/kg/day on days 0 to 4 and 50 mg/kg/day on days 7 to 11 (*), IDU at 75 mg/kg/day on days 0 to 4 and on days 7 to 11 ; and IDU at 50 mg/kg/day on days 0 to 4 and on days 7 to 11 . Change in body weight of SCID mice (five mice per condition) that had been treated twice daily by subcutaneous injection with IDU using the following treatment schedules: IDU at 150 mg/kg/day on days 0 to 4 and 75 mg/kg/day on days 7 to 11 , IDU at 100 mg/kg/day on days 0 to 4 and 50 mg/kg/day on days 7 to 11 (*), IDU at 75 mg/kg/day on days 0 to 4 and on days 7 to 11 ; and IDU at 50 mg/kg/day on days 0 to 4 and on days 7 to 11 . Data for animals that were left untreated are also indicated . TABLE 1 | Effect of IDU on VV-induced tail lesion formation in intravenously infected NMRI mice TABLE 2 | Effects of IDU and HPMPC on VV-induced mortality in SCID mice that had been infected intravenously or intraperitoneally TABLE 3 | Effect of delayed start of treatment with IDU on VV-induced mortality in SCID mice DISCUSSION : Until the introduction of vaccines in the early 1800s, poxviruses historically caused serious disease in humans. The last documented case of smallpox occurred in 1977 in Somalia, and the WHO declared that smallpox had been eradicated as of 1980. Before that time the virus had killed millions of people. The public health threat of poxviruses has recently surfaced with (i) the fear for terrorist attacks with smallpox and (ii) the observation of monkeypox infections in humans, especially in the period from 1996 to 1997 in the Democratic Republic of Congo (formerly Zaire) . Monkeypox infection in humans is associated with signs and symptoms similar to those of smallpox, including the development of a pustular rash that is clinically indistinguishable from smallpox. The infection, however, is far less lethal than smallpox . Starting in 1977, the number of laboratories holding stocks of the variola virus was reduced until all known stocks were consolidated in only two WHO collaborating centers: the Centers for Disease Control and Prevention in Atlanta, Ga., and the State Research Center of Virology and Biotechnology (VECTOR) laboratories in Kotsovo, near Novosibirsk, Russia. Although outlawed by the Biological Weapons Convention in 1972, there is a serious concern that there may be unregistered stocks of smallpox that can be used for bioterrorism or biological warfare. A smallpox outbreak occurring today in a highly mobile and susceptible population would likely spread widely before effective measures could be taken (; C. Hooper, Letter, N. Engl. J. Med. 339:2027-2028, 1998). Since there is no approved treatment for smallpox the WHO expert committee on orthopoxvirus infection has recommended that specific antivirals for the treatment of poxviruses should be developed before the destruction of the (officially remaining) stocks of variola (press release WHO/77, 10 December 1999). We previously reported the activity of some nucleoside phosphonate analogs against VV replication. The prototype compound of this family, (S)-9-(3-hydroxy-2-phosphonylmethoxypropyl)adenine (HPMPA), was found to exhibit a broad-spectrum activity against a wide variety of DNA viruses, including VV , and inhibited the development of tail lesions caused by VV . Cidofovir (HPMPC), which is the cytosine counterpart of HPMPA and which has been approved for the treatment of cytomegalovirus retinitis in immunocompromised patients (i.e., patients with AIDS), has an activity spectrum comparable to that of HPMPA but has a more-favorable toxicity profile in vitro and in vivo (i.e., in mice). We have demonstrated that cidofovir possesses potent protective activity against a lethal VV infection in SCID mice . Our observations were corroborated by the findings of Bray and colleagues who demonstrated that cidofovir is also effective in the treatment of lethal aerosolized or intranasal cowpox virus infections in mice . In the same report these authors revealed (their unpublished data) that cidofovir rescued monkeys that had received large quantities (1,000 50% lethal doses) of aerosolized monkeypox and that the compound was active in vitro against several different strains of smallpox. The antipoxvirus activity of cidofovir was also further documented in murine models . Recently, cidofovir-resistant strains of camelpox, cowpox, and monkeypox viruses and VV were generated . Cidofovir has been shown to result in complete regression of lesions associated with poxviruses in humans, i.e., molluscum contagiosum lesions (, ; V. Ibarra, J. R. Blanco, J. A. Oteo, and L. Rosel, Letter, Acta Derm. Venereol. 80:315-316, 2000) and orf (ecthyma infections) in an immunodeficient patient . Recently, we demonstrated that the acyclic nucleoside analogue 2-amino-7-[1,3-dihydroxy-2-propoxy)methyl]purine (S2242) and its diacetylated oral prodrug form HOE961 (the latter when given orally) elicit potent activity against VV infections in both immunocompetent and SCID mice . There were no obvious side effects associated with this treatment. A compound that can be given orally offers, particularly for the treatment of poxvirus infections, an advantage over compounds that must be given intravenously. In this context it is of interest to mention that alkoxyalkyl esters of cidofovir and cyclic cidofovir that may have an improved oral bioavailability profile showed enhanced in vitro inhibition of orthopoxvirus replication . We demonstrate here that the antiherpetic compound IDU (under the trade names Herpid, Stoxil, Idoxene, and Virudox) is able to cause an important delay in VV-induced mortality in SCID mice infected either intranasally, intraperitoneally, or intravenously. The protective activity of IDU (and that of HPMPC) was also reflected by the fact that lung histology in drug-treated infected animals was normal (at 18 days postinfection), whereas focal anomalies (similar to those described in reference ) were noted in the untreated mice. The lung pathology as seen here in the untreated mice is not as severe as the pathology reported in the vaccinia model employed by Smee and colleagues . This may be explained by the fact that in the present study (i) SCID mice were used, which lack functional T and B cells and which may, therefore, have less immunopathological involvement (BALB/c mice were used by Smee and colleagues) and (ii) a strain of virus was used that was different from that used by Smee and colleagues. Also in the cowpox model, lung pathology was reported to be more severe than that observed here . The protective activity of IDU in a nonlethal VV pox tail lesion model was already reported 25 years ago . It has, however, remained unknown whether IDU would cause protection against lethal orthopoxvirus infections. Other compounds, such as trifluorothymidine and arabinofuranosyl cytosine have also been reported to cause protection in the VV tail lesion model or in the VV keratitis model in rabbits . We were, however, not able to demonstrate any protective effect with either arabinofuranosyl cytosine or trifluorothymidine against lethal VV infections in SCID mice (data not shown). The use of IDU for the treatment of herpesvirus infections is restricted to topical use because the compound was found to be too toxic for intravenous administration . However, unlike for herpesvirus infections, there is only one compound, i.e., cidofovir, that could possibly be used for the treatment of infections with variola. In addition, as mentioned above, it has now been shown that cidofovir-resistant strains can be generated . Deliberate release of cidofovir-resistant variola strains would leave us without option for treatment Although one should aim to use compounds that cause as little as possible adverse effects for the treatment of infections with variola, the fact that the fatality rate associated with smallpox may be estimated at 10 to 50% (or even higher in immunocompromised patients) may possibly justify the administration of a relatively toxic compound for a short period of time during the acute phase of the infection. Indeed, the fatality rate of smallpox is much higher than that of many oncological diseases for which patients receive treatment with (relatively) toxic substances. Further studies are warranted to determine whether IDU causes protection against other orthopoxvirus infections (i.e., smallpox, cowpox, monkeypox, camelpox) in vitro and in vivo. Backmatter: PMID- 12183230 TI - Genetic and Biochemical Characterization of CGB-1, an Ambler Class B Carbapenem-Hydrolyzing beta-Lactamase from Chryseobacterium gleum AB - Chryseobacterium gleum (previously included in the Flavobacterium IIb species) is a gram-negative aerobe that is a source of nosocomial infections. An Ambler class B beta-lactamase gene was cloned and expressed in Escherichia coli from reference strain C. gleum CIP 103039 that had reduced susceptibility to expanded-spectrum cephalosporins and carbapenems. The purified beta-lactamase, CGB-1, with a pI value of 8.6 and a determined relative molecular mass of ca. 26 kDa, hydrolyzed penicillins; narrow- and expanded-spectrum cephalosporins; and carbapenems. CGB-1 was a novel member of the molecular subclass B1 of metallo-enzymes. It had 83 and 42% amino acid identity with IND-1 from Chryseobacterium indologenes and BlaB from C. meningosepticum, respectively. Thus, in addition to the previously characterized clavulanic acid-inhibited extended-spectrum beta-lactamase CGA-1 of Ambler class A, C. gleum produces a very likely chromosome-borne class B beta-lactamase. Keywords: Introduction : Metallo-beta-lactamases of Ambler class B are zinc-dependent enzymes that possess the property of hydrolyzing most beta-lactam antibiotics, including carbapenems . Some of these enzymes are emerging worldwide in gram-negative pathogens as a result of plasmid- and integron-location of their genes that are of the blaIMP and blaVIM series . However, several metallo-beta-lactamases have been reported as a source of intrinsic resistance to carbapenems in bacterial species less frequently isolated in clinical microbiology, such as L-1 from Stenotrophomonas maltophilia , Bc-II from Bacillus cereus 569/H , CcrA from Bacteroides fragilis , CphA from Aeromonas hydrophila , FEZ-1 from Legionella gormanii, , and THIN-B from Janthinobacterium lividum . Flavobacteria species are mostly a source of nosocomial infections . Metallo-beta-lactamases have been identified so far in two Flavobacterium species, Chryseobacterium meningosepticum and Chryseobacterium indologenes . C. meningosepticum usually expresses two unrelated metallo-enzymes of the BlaB and GOB-like series that share less than 12% amino acid identity . C. indologenes expresses IND-like enzymes that share less than 45 and 15% amino acid identity with BlaB and GOB-like enzymes, respectively . A search for a potential natural reservoir of acquired carbapenem-hydrolyzing beta-lactamases is interesting, particularly in understanding the mechanism(s) of integration of these genes in an integron. Thus, we proceeded to study the beta-lactamase content of another Flavobacterium species, Chryseobacterium gleum that is a phylogenetically related species to C. indologenes . From C. gleum strain CIP 103039, we have reported a clavulanic-acid inhibited Ambler class A expanded-spectrum beta-lactamase (ESbetaL), CGA-1 . However, since C. gleum CIP 103039 had also reduced susceptibility to carbapenems and CGA-1 hydrolyzes weakly these beta-lactam group, the aim of our study was to analyze the mechanism sustaining carbapenem resistance in C. gleum. We have cloned and identified genetically and biochemically a novel metallo-beta-lactamase that is closely related to metallo-enzymes of C. indologenes. MATERIALS AND METHODS : Bacterial strains. | C. gleum reference strain CIP 103039 was from the Institut Pasteur Collection . Escherichia coli DH10B and nalidixic acid-resistant E. coli JM109 were used for cloning and conjugation assays, respectively. All strains were stored at -70C in Trypticase soy (TS) broth (Becton Dickinson, Le Pont-de-Claix, France) supplemented with 15% glycerol until testing. Antimicrobial agents and MIC determinations. | The antimicrobial agents were obtained in the form of standard laboratory powders and were used immediately after their solubilization. The agents and their sources have been described previously . Antibiotic disks were used for routine antibiograms (Sanofi Diagnostics Pasteur, Marnes-la-Coquette, France). MICs of selected beta-lactams were determined by an agar dilution technique on Mueller-Hinton plates (Sanofi Diagnostics Pasteur) with an inoculum of 104 CFU per spot, as described previously . Cloning and analysis of recombinant plasmids. | Whole-cell DNA of C. gleum CIP103039 was extracted as described previously . All enzymes used in cloning experiments were from Amersham Pharmacia Biotech (Orsay, France). Fragments from whole-cell DNA partially digested with Sau3AI were ligated into BamHI-restricted phagemid pBK-CMV (Stratagene, Amsterdam, The Netherlands), as previously reported . Ligations were performed at a 1:5 vector/insert ratio, in 30 mul of a ligation mixture containing 5 U of T4 DNA ligase at 4C incubated overnight. Recombinant plasmids were transformed by electroporation (Gene Pulser II; Bio-Rad, Ivry-sur-Seine, France) into E. coli DH10B electrocompetent cells (Gibco BRL, Life Technologies, Cergy Pontoise, France). Antibiotic-resistant colonies were selected onto amoxicillin (30 mug/ml)- and kanamycin (30 mug/ml)-containing TS agar plates, and their antibiotic resistance patterns were analyzed according to the results of antibiogram performed by disk diffusion. Recombinant plasmid DNAs were obtained from 100-ml TS broth cultures grown overnight in the presence of amoxicillin (30 mug/ml) at 37C. Plasmid DNAs were recovered by using Qiagen (Courtaboeuf, France) columns. Plasmid mapping were performed after double-restriction analysis . Fragment sizes were estimated by comparison with the 1-kb DNA ladder (Amersham Pharmacia Biotech). One recombinant E. coli strain that had reduced susceptibility to imipenem and harbored recombinant plasmid pCGB-1 with the smallest 2-kb insert was retained for further analysis. Plasmid content, conjugation assay, and Southern hybridization. | Extraction of plasmid DNA from C. gleum CIP 106039 and direct transfer by conjugation of any beta-lactam resistance marker into in vitro-obtained nalidixic acid-resistant E. coli JM109 were attempted as previously described . Southern hybridization was performed as previously described with whole-cell DNA of C. gleum CIP 103039 by using the enhanced chemiluminescence nonradioactive labeling and detection kit (Amersham Pharmacia Biotech) with a 628-bp PCR-obtained probe with primers internal to blaCGB-1 (primer 1: 5'-GCAAACGCCCGGATACAACAG-3'; primer 2; 5'-TTCCATTCATCATGTCCGGG-3') beta-Lactamase purification. | A culture of E. coli DH10B harboring recombinant plasmid pCGB-1 was grown overnight at 37C in 4 liters of TS broth containing amoxicillin (30 mug/ml). Bacterial suspensions were pelleted, resuspended in 60 ml of 30 mM Tris-HCl buffer (pH 8.2), disrupted by sonification (three times at 30 W for 30 s with a Vibra Cell 75022 Phospholyser [Bioblock, Illkirch, France]), and centrifuged for 1 h at 48,000 x g at 4C. the supernatant ultracentrifuged at 100,000 x g for 1 h at 4C and the supernatant was dialyzed overnight against 30 mM Tris-HCl buffer (pH 8.2). This extract was loaded onto a preequilibrated Q-Sepharose column (Amersham Pharmacia Biotech). The enzyme recovered in the flowthrough and determined qualitatively by nitrocefin hydrolysis (Oxoid, Dardilly, France) was dialyzed overnight at 4C against 50 mM sodium phosphate buffer (pH 7.0). The beta-lactamase extract was then loaded onto a preequilibrated S-Sepharose column (Amersham Pharmacia Biotech) preequilibrated with the same buffer. The enzyme was eluted by a 100-ml linear NaCl gradient (0 to 500 mM) in sodium phosphate buffer (pH 7.0). The beta-lactamase was eluted at a concentration of 260 mM NaCl. The fractions containing the highest beta-lactamase activities were pooled, dialyzed overnight against 50 mM sodium phosphate buffer (pH 7.0) containing 50 muM ZnCl2 and stored at -80C. Purity of the enzyme was estimated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) . IEF analysis and N-terminal sequencing. | Purified enzyme preparation from culture of E. coli DH10B harboring pCGB-1 and beta-lactamase crude extracts from 100 ml-culture of C. gleum CIP 103039 were subjected to analytical isoelectric focusing (IEF) on a pH 3.5 to 9.5 ampholine polyacrylamide gel (Ampholin PAG plate; Amersham Pharmacia Biotech) as described previously . In order to determine the cleavage site of the CGB-1 mature protein, the purified enzyme was subjected to an Edman analysis at the laboratory for protein microsequencing at the Pasteur Institute, Paris, France. Purified enzyme and marker proteins were subjected to SDS-PAGE (25 mA, 4 h, room temperature). Proteins were then electrotransferred onto a polyvinyl difluoride membrane (Immobilon-P; Millipore) by using the Mini Protean II transfer cell (8 by 7.3 cm; Bio-Rad) in 50 mM Tris-borate buffer (pH 8.7) at room temperature (3.5 V/cm, overnight). The membrane was then rinsed in distilled water and stained with a solution made of 0.1% Coomassie brilliant blue R-250 in methanol and water (50:40 [vol/vol]). The protein band was then excised with a razor blade and allowed to air dry. The amino-terminal sequence of the mature beta-lactamase was determined with an automated Edman sequencer on a 473A model gas phase sequencer (Applied Biosystems). Kinetic measurements of beta-lactamase CGB-1 and induction study. | C. gleum CIP103039 crude extract was used to detect imipenem hydrolysis in preliminary experiments. Purified metallo-beta-lactamase CGB-1 was used for kinetic measurements performed at 30C in 100 mM sodium phosphate (pH 7.0) containing 50 muM ZnCl2. Determinations of specific activities of purified enzyme and of beta-lactamase extract from C. gleum CIP 103039 were performed with 100 muM imipenem as a substrate as described previously . The total protein content was measured with bovine albumin as the standard (Bio-Rad DC protein assay kit). Km and kcat values were determined with a spectrophotometer ULTROSPEC 2000 (Amersham Pharmacia Biotech) by analyzing the beta-lactam hydrolysis under initial rate conditions by using the Eadie-Hofstee linearization of the Michaelis-Menten equation as previously described previously . Various concentrations of EDTA and clavulanic acid were preincubated with the enzyme for 3 min at 30C before testing the rate of imipenem hydrolysis. Fifty percent inhibitory concentrations (IC50s) were determined for EDTA and clavulanic acid, and results were expressed in micromolar units. Induction experiments were performed with 1 and 4 mug of cefoxitin per ml and 0.1 mug of imipenem per ml as inducer with cultures of C. gleum CIP 103039 and E. coli DH10B(pCGB-1), as previously described . DNA sequencing and protein analysis. | The cloned DNA fragment of recombinant plasmids pCGB-1 was sequenced on both strands, with an Applied Biosystems sequencer (ABI 377). The nucleotide and the deduced protein sequences were analyzed with software available over the Internet as reported previously . The relative molecular mass of the beta-lactamase expressed by a culture of E. coli DH10B(pCGB-1) was estimated by SDS-PAGE analysis, as previously described . Nucleotide sequence accession number. | The nucleotide sequence and deduced beta-lactamase amino acid sequence reported in this work have been assigned to the GenBank and EMBL databases under the accession no. . RESULTS AND DISCUSSION : Cloning experiments. | Preliminary experiments of imipenem hydrolysis with a beta-lactamase extract of C. gleum CIP 103039 suggested the presence of a carbapenem-hydrolyzing beta-lactamase in addition to a previously characterized class A beta-lactamase . Forty recombinant E. coli DH10B clones harboring plasmids with inserts that varied in size (2.0 to 10 kb) were obtained after shotgun cloning of Sau3AI-restricted DNA of C. gleum CIP 103039. Among them, 28 clones had decreased susceptibility to ceftazidime and a beta-lactam resistance phenotype that was susceptible to clavulanic acid inhibition that was consistent with the presence of the previously characterized class A enzyme CGA-1 . Twelve other recombinant E. coli clones had a slight decrease of susceptibility to carbapenems. Among them, E. coli DH10B(pCGB-1) harboring a recombinant plasmid which possessed the smallest 2.0-kb insert was retained. DNA and protein sequence analysis of CGB-1. | DNA sequence analysis of the 2,004-bp insert of pCGB-1 revealed an open reading frame (ORF) (named blaCGB-1) of 729 bp (from nucleotide 461 to 1,189) encoding a 243-amino-acid preprotein . A putative cleavage site was found by computer analysis after the Ala-Asn-Ala motif . N-terminal sequence analysis of the protein confirmed that the first amino acids residues of the mature protein were QDTQ . The resulting mature protein contained 223 amino acid residues. The overall G+C content of this ORF was 36%, which lies within the expected range of G+C ratio of the Chryseobacterium genes (36 to 38%) . Another truncated ORF was detected on this same insert (nucleotide 1,240 to 2,004) of recombinant plasmid pCGB-1 located downstream and in opposite orientation of blaCGB-1 . It encoded 254 amino acids of the carboxyl end of a putative protein that possessed 31.5% amino acid identity with the capreomycin acetyltransferase from Streptomyces capreolus (accession number no. ). A similar capreomycin resistance gene had been identified downstream and in opposite orientation of the class B beta-lactamase gene blaIND-1 from C. indologenes 001 (unpublished data) and in Mycobacterium tuberculosis genome (accession no. ). Thus, the blaCGB-1 gene may be part of a cluster of antibiotic resistance genes. A putative DNA sequence, located from nucleotides 1,187 to 1,229 between the stop codon of blaCGB-1 and the stop codon of the putative capreomycin acetyltransferase gene, might play a role as a transcription terminator . Genetic location of blaCGB-1. | No plasmid was detected in C. gleum CIP 103039, and direct conjugation experiments failed to transfer any beta-lactam resistance marker from C. gleum CIP 103039 to nalidixic acid-resistant E. coli JM109. Using an internal probe for blaCGB-1, Southern hybridization was positive at the chromosomal position of migration of whole-cell DNA of C. gleum CIP 103039 (data not shown) Susceptibility testing. | MICs of beta-lactams for C. gleum CIP 103039 showed that it was resistant to amino- and carboxy-penicillins, narrow-spectrum cephalosporins, cefotaxime, aztreonam, had intermediate susceptibility to carbapenems and remained susceptible to piperacillin . E. coli DH10B(pCGB-1) was resistant to amoxicillin, ticarcillin, some restricted-spectrum cephalosporins, and had a reduced susceptibility to piperacillin, ceftazidime and carbapenems compared to that of E. coli DH10B. MICs of beta-lactams for E. coli DH10B(pCGB-1) were not lowered by addition of clavulanic acid and tazobactam . Resistance to aztreonam in C. gleum CIP103039 as found in other flavobacterial species (-) may be partially explained by expression of the previously characterized clavulanic acid-inhibited CGA-1 enzyme . Other resistance mechanisms such as low outer membrane permeability, efflux, and penicillin binding protein affinity may explain also resistance to this compound. Once cloned onto a plasmid vector and expressed in E. coli, both enzymes, CGB-1 and IND-1, provided a similar pattern of decreased susceptibility to beta-lactams, although in the latter case the degree of resistance was more important . The overall susceptibility of C. gleum to beta-lactams was similar to that reported for C. meningosepticum . As for C. meningosepticum that possesses a class A beta-lactamase (CME) and two class B beta-lactamases (BlaB and GOB) , C. gleum possesses, in addition to the class A beta-lactamase CGA-1, another very likely chromosomal-encoded class B enzyme CGB-1. Biochemical properties of CGB-1. | IEF analysis revealed that cultures of E. coli DH10B(pCGB-1) produced a beta-lactamase activity with a pI value of 8.6. However, only one beta-lactamase activity was detected from crude extracts of a culture of C. gleum CIP103039, at a pI of 8.9 that corresponded to the previously identified class A beta-lactamase CGA-1 . Thus, expression of CGB-1 from C. gleum CIP 103039 might be weak. The specific activity of the purified CGB-1 enzyme was 8.3 mumolmin-1mg of protein-1, determined with 100 muM imipenem as a substrate with a 174-fold purification coefficient. Its purity was estimated to be 95% by SDS-PAGE (data not shown). The mature protein (named CGB-1 for C. gleum class B beta-lactamase) expressed in E. coli DH10B had a relative molecular mass determined experimentally to be ca. 26 kDa (data not shown). Kinetic parameters of CGB-1 revealed a strong activity against amino- and carboxy-penicillins and restricted-spectrum cephalosporins . Surprisingly, cefotaxime was a good substrate for CGB-1 due to a low Km, whereas ceftazidime that possesses a Km superior to 1 mM was a very poor substrate. This difference in hydrolysis activity mirrored the MICs for both beta-lactams for C. gleum CIP 103039 (but not for E. coli DH10B(pCGB-1) . CGB-1 beta-lactamase had poor affinity for imipenem (Km, >1,000 muM), resulting in a weak catalytic efficacy of imipenem . Therefore, its kcat/Km value for imipenem was at least 10-fold lower than that of IND-2, whereas these values for meropenem and for benzylpenicillin were only twofold lower . These results were consistent with the low level of resistance to imipenem of C. gleum CIP 103039 . Hydrolysis of aztreonam was not detectable, as found for all metallo-enzymes . Catalytic efficacy of CGB-1 was lower than that of GOB-1 from C. meningosepticum for all beta-lactams except for benzylpenicillin which were similar . CGB-1 had kcat/Km values which were 4-fold lower for benzylpenicillin, 10-fold lower for cephaloridine and for cefoxitin, 60-fold lower for imipenem, and 3-fold higher for cefotaxime, compared to those of BlaB from C. meningosepticum . The hydrolytic activity of CGB-1 was inhibited by EDTA (IC50, 37 muM) but not by class A beta-lactamase inhibitors such as clavulanic acid (IC50 > 1 mM). These kinetic parameters lead to include CGB-1 in the biochemical group 3a of the Bush beta-lactamase classification for metallo-enzymes . This group includes all characterized metallo-beta-lactamases with a very broad substrate hydrolysis profile except FEZ-1 and Cph-A-like enzymes from L. gormanii and Aeromonas sp., respectively. Induction experiments. | Induction studies with cefoxitin and imipenem as beta-lactam inducers failed to detect induction of beta-lactamase expression with cultures of C. gleum CIP 103039 and of E. coli(pCGB-1) with imipenem as substrate. These results are consistent with the absence of any sequence for regulatory gene, such as LysR-type regulator gene upstream of blaCGB-1 . Amino acid sequence analysis. | The comparison of amino acid sequence of CGB-1 with those of other class B beta-lactamases revealed 83% identity with IND-1, 82% identity with IND-2 and IND-3, and 73% identity with IND-4 from C. indologenes, and only 42 and 12% identity with BlaB and GOB-1 from C. meningosepticum, respectively (-, ). Thus, the metallo-enzyme of C. gleum was most closely related to those of C. indologenes thus paralleling the taxonomic position of both Chryseobacterium species . The six conserved amino acid residues implicated in the Zn2+ or water molecule binding (His116, His118, Asp120, His196, Cys221, and His 220) of metallo-beta-lactamases were found in the CGB-1 sequence . However, an insertion of three amino acids (Asp, Thr, and Gln) was detected at the second, third and fourth position of the N-terminal sequence of the mature protein CGB-1 (positions 22, 23, and 24) compared to IND-like . beta-Lactamase CGB-1 may be classified in the molecular subclass B1 of the structural classification that groups IND-like, Bc-II, CfiA, BlaB, IMP, and VIM beta-lactamases . A phylogenic analysis of CGB-1 from C. gleum, BlaB variants from C. meningosepticum (and not GOB variants from C. meningosepticum also), and IND variants from C. indologenes, showed that they belong to a same phylogenic lineage (data not shown). Comparison of amino acid sequence of CGB-1 with those of IND-1 and IND-2 did not provide evidence of any particular amino acid residue that may explain a lower affinity of CGB-1 for carbapenems. In conclusion, this report identified a novel bacterial species for which beta-lactam resistance is at least mediated by an Ambler class A expanded-spectrum beta-lactamase and an Ambler class B metallo-enzyme, as found in S. maltophilia and C. meningosepticum. FIG. 1. | Nucleotide sequence of the cloned 2,004-bp fragment of recombinant plasmid pCGB-1 containing the blaCGB-1 coding region and a truncated ORF encoding a putative capreomycin resistance determinant. Nucleotide sequence of the cloned 2,004-bp fragment of recombinant plasmid pCGB-1 containing the blaCGB-1 coding region and a truncated ORF encoding a putative capreomycin resistance determinant. The deduced amino acid sequence is designated in the single-letter code below the nucleotide sequence. The putative terminator sequence is indicated by two inverted arrows, and stop codons are indicated by stars. FIG. 2. | Comparison of the amino acid sequence of beta-lactamase CGB-1 with those of IND-1, IND-2, IND-3, and IND-4 from C. indologenes Comparison of the amino acid sequence of beta-lactamase CGB-1 with those of IND-1, IND-2, IND-3, and IND-4 from C. indologenes and BlaB from C. meningosepticum . The BBL numbering scheme is indicated above the sequences . Broken lines and dots indicated identical and deleted amino acid residues, respectively. The vertical arrow indicates the putative cleavage site for the leader peptide of CGB-1. Amino acids that may be involved in binding of Zn2+ or/and water are indicated by a star. TABLE 1 | MICs of beta-lactams for C. gleum CIP 103039, E. coli DH10B(pCGB-1) expressing CGB-1, E. coli DH10B(pSO-1) expressing IND-1 from C. indologenes OO1, and reference strain E. coli DH10B TABLE 2 | Kinetic parameters of purified CGB-1 from cultures of E. coli DH10B(pCGB-1) Backmatter: PMID- 12183237 TI - Penciclovir Susceptibilities of Herpes Simplex Virus Isolates from Patients Using Penciclovir Cream for Treatment of Recurrent Herpes Labialis AB - The antiherpesvirus agent penciclovir (PCV) shares an identical activation pathway and a similar mode of action with acyclovir (ACV). However, since PCV represents a relatively recent treatment option, the clinical resistance profile to PCV is less well known. A susceptibility program was established to assess the resistance profile for serial herpes simplex virus isolates from immunocompetent patients with recurrent herpes labialis obtained throughout a 4-day period of treatment with topical PCV (1% cream) or a placebo. Two isolates (2 of 1,035 [0.19%]), representing 0.34% of the patients (2 of 585), were confirmed to be PCV-resistant (Pcvr) herpes simplex virus type 1 by a plaque reduction assay in MRC-5 cells. These two viruses were highly resistant to PCV (50% inhibitory concentrations [IC50s], >55 mug/ml) and were isolated less than 17 h after the start of patient-initiated treatment. However, subsequent isolates on days 2 and 3 from these patients were completely susceptible to PCV (IC50s, <2.0 mug/ml). Thus, it is not clear whether the resistance to PCV for these two early-treatment isolates was directly associated with the 17 h of PCV treatment; several possible explanations are discussed. In an analysis of the distribution of IC50 differences between the first and last isolates, there were three patients with minor IC50 increases in the PCV-treated population and one in the placebo-treated group, although statistically, only the latter was an outlier. No patients were found to have Pcvr virus at the end of acute treatment, regardless of treatment group. Overall, the prevalence of Pcvr was found to be similar to the 0.3% Acvr reported for immunocompetent, untreated populations. Keywords: Introduction : The use of acyclovir (ACV) and penciclovir (PCV) in treating herpesvirus infections has increased significantly since the introduction of ACV 2 decades ago. These nucleoside analogues of guanine are preferentially phosphorylated to monophosphate by the viral thymidine kinase (TK) and subsequently converted to ACV triphosphate by cellular enzymes. The activated triphosphate inhibits viral DNA polymerase activity and prevents viral DNA elongation . Herpes simplex virus (HSV) resistance to these agents typically develops by mutations in the TK gene, although mutations within the viral DNA polymerase can also confer ACV resistance (-, , ). ACV-resistant (Acvr) HSV variants have been readily isolated in culture after sequential passages in the presence of increasing concentrations of ACV (reviewed in reference ). However, several studies suggest that clinical use of ACV has not been associated with an increased emergence of drug-resistant virus. Sensitivity monitoring surveys have revealed that since the introduction of ACV, the prevalence of resistance in the general population has remained unchanged , and little if any impact on the prevalence of resistant virus in the immunocompetent population has been shown . Clinically significant resistance to ACV has been limited almost exclusively to the immunocompromised population , in which approximately 4 to 10% of patients develop resistance during antiviral treatment . To discern the prevalence of PCV resistance in the immunocompetent population, susceptibility assays were performed on virus isolates from patients participating in two placebo-controlled trials for evaluation of the efficacy of topical PCV for recurrent herpes labialis (; unpublished data). MATERIALS AND METHODS : Isolate sampling and resistance breakpoint. | Topical PCV cream (1%) was evaluated as a treatment for recurrent herpes labialis in two phase III studies conducted according to the same protocol . Patients were instructed to apply PCV or placebo cream nine times per day for 4 days within 1 h of the first sign or symptom of a recurrence. Although treatment was initiated by the patient, the patients visited the study center daily for evaluation and virus isolation. Additionally, each patient kept a daily record of treatment time and compliance. To determine the pattern of virus susceptibility to PCV, antiviral assays were performed on the first and last virus isolates obtained from the patients. In some instances, additional isolates obtained at intervals between the samplings for the first and last isolates were obtained and also tested. The first isolate was defined as the virus isolated within 24 h of the start of topical PCV treatment. The last isolate was defined as the final virus-positive culture obtained. In practice these were obtained during the treatment period or after cessation of treatment. Virus isolates were taken during the vesicle to soft-crust stage. The mean interval between first and last isolates was 1.8 days for the PCV group and 1.9 days for the placebo group. In order to determine the pattern of virus susceptibility, antiviral assays were performed on a random, blinded sample of isolates from each study. An in vitro breakpoint for defining ACV resistance has been set at 2.0 mug/ml, since 50% inhibitory concentrations (IC50s) for isolates from individuals failing to respond to drug therapy were generally above this level . Breakpoints are typically established through a consensus of researchers in the field based on the correlation between clinical treatment failure and in vitro IC50s. Because there is no recognized in vitro breakpoint for defining resistance to PCV, provisional guidelines were set for defining resistance based on the susceptibility of the sensitive control virus strain to PCV, which was determined in each assay. This takes into account the known variation in IC50s between laboratories. Resistance was defined as an IC50 of >=2.0 mug/ml or an IC50 >10-fold higher than the IC50 for the wild-type sensitive control virus within that particular assay . Virological work was performed at two College of American Pathology-approved centers (the clinical virology laboratories of the Children's Hospital of Philadelphia and the University of San Francisco General Hospital), and each center used wild-type HSV type 1 (HSV-1) as a standard virus for susceptibility assays (either strain SC16 or strain F, depending on the center). Statistical analysis. | Analysis of covariance was used to determine whether HSV isolates from PCV-treated patients and placebo-treated patients had different last IC50s after comparison with their respective first IC50s. The analysis was performed using natural logarithms of the IC50s, which provides statistical stringency. It was assumed that within each treatment group, the average difference between last and first isolates was zero, and significant deviation from zero would indicate a trend. Statistically significant differences in the change from the first to the last IC50 between PCV-treated and placebo-treated groups would be indicated by a P value of less than 0.05. Similarly, it was assumed that between the two treatment groups, the average difference between first isolates was zero and the average difference between last isolates was zero; a P value below 0.05 would suggest a trend. Virus culture. | Lesions were swabbed daily throughout the course of the recurrence, during treatment, and often after treatment had ceased, to maximize the chance of identifying Pcvr isolates. Briefly, Dacron swabs moistened with Viral-Chlamydial transport medium (Carr-Scarborough Microbiologicals, Decatur, Ga.) were used to swab the lesions. Swabs were placed in transport medium and either processed immediately for virus isolation or stored at 4C for no more than 48 h. Human diploid fibroblast cells in shell vials were inoculated with the transport medium and incubated overnight at 37C. Resistant virus isolates were typed by staining of viral antigens with type-specific monoclonal antibodies (Dako) and were confirmed to be HSV-1. All other virus isolates were presumed to be HSV-1, although not directly typed. Plaque reduction assay. | Testing of susceptibility to PCV was performed by the plaque reduction assay method in human diploid fibroblast cells (MRC-5) between passages 12 and 20. Briefly, cells were seeded into 12-well microtiter plates at approximately 3 x 105/well in 1.0 ml of Eagle's minimum essential medium containing 10% fetal calf serum or into 24-well plates with approximately one-half the number of cells. Cells were inoculated with 10-fold dilutions (102 to 104) of the viral isolate for 1 h at 37C in a final volume of 0.5 ml of Hanks buffered salt solution. Testing was performed in triplicate in MRC-5 cells by using a series of PCV concentrations over 10 or 11 serial dilutions to provide at least two data points on either side of the IC50. After virus adsorption, the drug, 2x Eagle's minimum essential medium, and 0.8% (wt/vol) SeaPlaque agarose (FMC Bioproducts) were mixed, and 3.0-ml volumes were added to each well of a 12-well plate. After 3 days at 37C, plates were fixed with 1.0 ml of 10% formaldehyde solution for 1 h at room temperature. Cell monolayers were stained with crystal violet after removal of the agarose plugs. Plaque numbers were counted, and IC50s were calculated. Plaque reduction assays on the resistant isolates were performed in D21 cells, a line derived from BUHK-TK cells which constitutively expresses an HSV TK gene, as described in reference . These cells were a kind gift from H. Field (University of Cambridge, Cambridge, United Kingdom). These transformed cells were maintained in modified Eagle's medium with 10% calf serum and HAT supplement (hypoxanthine, aminopterin, and thymidine). For resistant samples, susceptibilities to ACV and foscarnet were also determined by the plaque reduction assay, with compounds obtained from Sigma Chemical Co. (St. Louis, Mo.). TK assay. | Viral TK activity was determined by a modification of the method described by Coen and Schaffer . Human 143 TK-negative cells seeded in duplicate 100-mm dishes were infected at 5 PFU/cell with the HSV preparation in a Beckman G6-CR tabletop centrifuge, in 4.0 ml of serum-free medium. Parallel cell monolayers were mock infected. One hour postinfection, monolayers were rinsed with phosphate-buffered saline, and fresh medium was added for 8 h. Infected cells were then rinsed with phosphate-buffered saline, scraped, and centrifuged for 10 min at 1,000 rpm (4C) and cell pellets were frozen at -80C. Thawed pellets were resuspended in 300 mul of 10 mM sodium phosphate buffer (pH 6.0)-5 mM 2-mercaptoethanol-10% glycerol-50 muM thymidine. Extracts were sonicated on ice and centrifuged to remove cellular debris. This extract (9 mul) was added to a mixture to yield final concentrations of 100 mM sodium phosphate (pH 6.0), 10 mM ATP, 10 mM magnesium acetate, 6 muCi of [3H]thymidine (11 Ci/mmol; NEN Research Products), 50 muM TTP, 25 mM NaI, 0.67 mM dithiothreitol, and 10 mug of bovine serum albumin/ml in a final volume of 30 mul. Reaction mixtures were incubated at 30C. At various times ranging from 0 to 180 min after addition of the cell extract, 5-mul aliquots were removed, added to 20 mul of 1 mM thymidine, and boiled for 2 min. Samples were then spotted onto Whatman DE81 circle filters. After drying, the filters were washed three times with 4 mM ammonium formate and 10 muM thymidine, once with distilled water, and twice with ethanol. Dry filters were placed in scintillation vials with Betafluor and counted. Values from duplicate samples were averaged. Radioactivity from the mock-infected control processed in parallel was used to subtract background. Data points from the linear range of thymidine phosphorylation were used. TK activity for HSV-1 SC16 was set at 100%. The limit of detection was estimated to be 0.3%, in agreement with a previous report . RESULTS : PCV susceptibilities of paired HSV isolates. | Sensitivity testing was performed on a total of 1,035 isolates from 585 patients, of which 864 (83%) represented paired isolates (first and last isolate obtained from each patient). Of the paired isolates, 358 were from PCV-treated patients and 506 were from the placebo-treated group. The IC50 results, including standard deviations and ranges, are summarized in Table . The average IC50s for the isolates at the start of treatment were 1.1 +- 7.4 and 0.28 +- 0.28 mug of PCV/ml for the PCV- and placebo-treated groups, respectively. As discussed below, first virus isolates from two patients were resistant to PCV; subsequent isolates from these patients were sensitive to PCV. Analysis of the first-isolate data after removal of the IC50s for these two isolates (55 and 83 mug of PCV/ml) resulted in similar susceptibility profiles overall for the two treatment groups (0.29 +- 0.25 and 0.28 +- 0.28 mug of PCV/ml for the PCV- and placebo-treated groups, respectively). Lastly, similar average IC50s were reported for all last isolates regardless of treatment (0.32 +- 0.33 and 0.31 +- 0.33 mug of PCV/ml for the PCV- and placebo-treated groups, respectively). The trend analysis on data for paired isolates (first and last HSV isolates tested for PCV susceptibility) from this study indicates there are no statistically significant differences between IC50s for last isolates from PCV-treated and placebo-treated patient populations after accounting for any difference in the IC50s for first viral isolates (P = 0.791 by analysis of covariance) . Furthermore, monitoring changes in IC50s between first and last isolates can serve as a powerful method to identify alterations in the pattern of susceptibility during therapy. The histogram in Fig. illustrates the differences in IC50 (IC50 for last isolate minus IC50 for first isolate) between the paired first and last isolates for each treatment group, up to an increase of 2.5 mug of PCV/ml. Only two isolates, the two Pcvr variants, had IC50 changes greater than 2.5 mug of PCV/ml. Several peaks, representing subpopulations with increasing IC50s, were apparent in the PCV and placebo treatment groups . Clearly, for the majority of samples evaluated, similar profiles were apparent for the two treatment groups. However, to the right of the midpoint, which corresponds to no change in the IC50 between first and last isolates, three minor peaks in the PCV-treated population and one minor peak in the placebo-treated group were present. For three PCV-treated samples, the last PCV IC50 was approximately 0.65 to 0.7, 0.75 to 0.80, or 1.0 to 1.5 mug/ml higher than the IC50 for the first isolate (Fig. , peaks A, B, and C, respectively). For one placebo-treated last isolate, the PCV IC50 was more than 1.5 mug/ml greater than that for the corresponding first isolate (Fig. , peak D). Overall, however, for the majority of isolates, the distribution profiles of the two treatment groups overlap, suggesting that no significant differences in susceptibility are present between treatment groups, although the minor increases in IC50 for four patient samples tested could indicate a trend to increasing resistance or simply represent the inherent variability within the plaque reduction assay. PCV susceptibilities of all HSV isolates. | Table summarizes the susceptibility data for all HSV isolates (paired and nonpaired) collected during the two clinical trials. Once again, average IC50s for the two groups of isolates were very similar, as measured by susceptibility to PCV in the plaque reduction assay (0.29 +- 0.28 and 0.28 +- 0.29 mug of PCV/ml for the PCV- and placebo-treated groups, respectively), when IC50s for the two resistant first isolates were excluded. Inclusion of the two Pcvr isolates in the data set results in an average IC50 of 0.60 +- 4.7 mug of PCV/ml for the PCV-treated group. Inclusion of these highly resistant isolates in the test population provides the most accurate representation of the background frequency of Pcvr, approximately 0.34% (2 of 585 patients). PCV-resistant clinical isolates. | PCV IC50s for virus isolates taken from two patients (patients 024.028.2586 and 024.023.0495) within 17 h of the start of PCV treatment were 83 and 55 mug/ml, respectively, clearly meeting the criteria defining resistance. However, subsequent isolates from these two patients taken from the same episode were susceptible to PCV (PCV IC50s, 1.1 and 0.8 mug/ml, respectively; viruses were isolated on day 3 of treatment in both cases). No other Pcvr viruses were identified from the treatment period or after cessation of treatment with topical PCV. Only for two other isolates were PCV IC50s equal to or greater than 2 mug/ml; both of these were from patients treated with a placebo. For one isolate, obtained on day 3 from patient 024.027.0424, the IC50 was 3.1 mug/ml, but this value was substantially below 9 mug/ml, the breakpoint defined as 10-fold above the IC50 for the wild type. When the isolate was retested against PCV, an IC50 of 0.74 mug/ml was determined, and therefore this isolate was not considered to be resistant to PCV. For another isolate (obtained on day 1 from patient 024.028.2016), the IC50 was 2 mug/ml. The IC50 for the sensitive control strain ranged between 0.9 and 1.6 mug/ml, and therefore, according to the 10-fold criterion, the test isolate was classified as sensitive; however, based on the standard breakpoint of an IC50 of >=2.0 mug/ml, this isolate would be labeled as possibly resistant. Unfortunately, this isolate was not available for further analysis or retesting. Molecular characterization of Pcvr isolates. | Plaque autoradiography on the two Pcvr patient samples (patients 024.028.2586 and 024.023.0495) confirmed that the resistant variants represented the majority of virus present in the virus preparation (data not shown). Plaque-purified isolates from the two patient samples which were confirmed as Pcvr (patients 024.028.2586 and 024.023.0495) were evaluated for the ability to phosphorylate 3H-labeled substrates including thymidine, ACV, and PCV according to previously described methods . These isolates were unable to utilize ACV or PCV as a substrate, compared to phosphorylation by wild-type HSV-1 (set at 100%), and they were also highly impaired in thymidine phosphorylation, suggestive of a TK-negative or TK-partial phenotype . Lastly, PCV IC50s were reduced to below 0.7 mug/ml upon testing in BUHK-TK cells (data not shown), confirming that impaired TK activity was responsible for conferring resistance to PCV. However, the important distinction between TK-negative and TK-partial remains difficult to make based on in vitro TK assays alone. Sequence analysis of the TK coding region identified the presence of frameshift mutations, consistent with the TK-negative phenotype . The isolate from patient 024.028.2586 contained two mutations, a deletion of residue 65 (threonine) and a frameshift (cytosine insertion) in the homopolymeric C run at residue 185. The isolate obtained from patient 024.023.0495 contained three nonconservative changes and a frameshift at residue 270 (cytosine deletion). It was not determined whether the nonconservative residues and/or frameshift was directly responsible for conferring drug resistance. FIG. 1. | Changes in IC50 between first and last HSV-1 isolates. Changes in IC50 between first and last HSV-1 isolates. This parameter was utilized to identify alterations or trends in the pattern of susceptibility during the course of therapy. Differences between the IC50 for the last isolate and that for the first isolate are shown as ranges (in micrograms of PCV per milliliter) on the y axis. Numbers of patients analyzed are shown on the x axis. The distributions of susceptibility are similar for the two treatment groups except for four minor populations. Peaks A through C represent three minor populations in the PCV-treated group with a shift to the right in susceptibility, and peak D corresponds to one minor population in the placebo-treated group with decreased susceptibility to PCV. FIG. 2. | Alignment of mutations within the TK coding sequence. Alignment of mutations within the TK coding sequence. The schematic representation depicts the HSV-1 TK polypeptide and three conserved domains, the nucleotide binding pocket, the thymidine binding site, and the ATP binding site. The homopolymeric hot-spot regions (G7 and C6) are indicated. Below the diagram, the genotypic mutations and resulting residue changes or frameshifts (FS) identified in the two confirmed Pcvr isolates are shown. TABLE 1 | Analysis of PCV susceptibilities of paired HSV-1 isolates TABLE 2 | Analysis of PCV susceptibilities of all HSV-1 isolates, paired and nonpaired TABLE 3 | Biochemical characterization of Pcvr HSV-1 isolates DISCUSSION : This study demonstrates that there is no evidence of reduced PCV sensitivity among viral isolates obtained from patients treating a single episode of recurrent herpes labialis with topical PCV. No statistically significant difference between the PCV and placebo treatment groups was identified in the IC50s for first isolates (P = 0.796) or the IC50s for last isolates (P = 0.757) . Furthermore, the average difference between IC50s for paired isolates in the PCV group was 0.028, and that in the placebo group was 0.025. Thus, there was no statistically significant difference between the last and first IC50s in either the PCV-treated group (P = 0.078) or the placebo-treated group (P = 0.065) . The plaque reduction assay data, along with the histogram analysis presented in Fig. , confirm that acute treatment of recurrent herpes labialis in immunocompetent patients with topical PCV did not result in an increase in the prevalence of resistant HSV. Collecting serial isolates represents a powerful method for examining trends in resistance. Since the distribution profiles of changes in IC50 after treatment significantly overlap for the two treatment groups, acute treatment with PCV cream versus a placebo does not appear to increase the likelihood of selecting for Pcvr HSV in immunocompetent patients with recurrent herpes labialis. However, studies monitoring resistance trends during episodic, prolonged treatment need to be performed in order to assess the development of resistance over time with repeated usage. Two virus isolates for which PCV IC50s were 55 and 83 mug/ml were confirmed by biochemical and molecular analyses to be Pcvr. Both were first isolates, taken within 17 h of the start of treatment with topical PCV. Spontaneous mutations within the HSV genome are introduced by errors during DNA replication and are independent of the presence of an antiviral agent . This natural phenomenon results in the accumulation of 6 to 8 TK-deficient variants per 104 plaque-forming viruses in virus populations that have never been exposed to selective pressure and can explain the coexistence of both resistant and sensitive viruses within all clinical HSV isolates. Consistent with this, there is a low prevalence of resistant HSV among individuals who have not been treated with an antiviral agent . Given the replication cycle kinetics of HSV and the fact that both Pcvr isolates were obtained within 17 h of initiation of treatment, it seems unlikely, although possible, that they resulted from selection pressure as the result of antiviral treatment. Since approximately 50% resistant variants are required within a virus preparation to confer such high IC50s , a preexisting resistant variant would have to represent a substantial proportion of the virus swab and be selectively amplified immediately upon the initiation of treatment. Although the presence of these two resistant isolates from the first isolate swabs may simply reflect the natural heterogeneity of HSV populations , isolates taken on day 2 would be expected to also be resistant, yet these IC50s were below 2.0 mug/ml. Another possible explanation is that these two isolates result from rare cases where residual topical PCV carried from the viral swab facilitated in vitro selection of resistant virus during the isolation process. If this is true, isolates obtained from these patients on days 2 and 3 would be expected to be completely susceptible to PCV. A third explanation which cannot be ruled out is preferential selection for resistant variants on day 1, followed by subsequent selection against these isolates for fitness. Lastly, patient recording of treatment times indicates that lack of compliance was not a factor in the selection for resistance (data not shown). Genotypic characterization of the two confirmed Pcvr HSV isolates described in this report resulted in the identification of a frameshift mutation at residue 185 in one isolate, which was also found in a previously characterized Pcvr isolate and is similar to the mutations routinely found within the homopolymeric region of TK (G7 and C6 hot spots) . Moreover, several mutations were also identified in the second isolate characterized, notably with a frameshift at residue 270, which, like the frameshift at residue 185, could be predicted to disrupt the integrity of the ATP/nucleoside binding pocket. Interestingly, although these isolates were plaque purified several times, they were not absolutely defective in the ability to phosphorylate thymidine, whereas phosphorylation of ACV and PCV was below the level of detection. It is not known whether the minor level of thymidine phosphorylation is due to contamination with wild-type virus undetectable by plaque autoradiography, to ribosomal frameshifting , or to other factors. Most recently, the prevalence of ACV resistance was determined to be approximately 0.4% based on data for HSV isolates obtained from 708 immunocompetent, ACV-naive individuals with genital herpes . Furthermore, the proportion of ACV resistance appears to be relatively stable in immunocompetent patients, even after the increased usage of ACV over the past 2 decades. The present study indicates that the overall prevalence of PCV-resistant HSV isolates, for the immunocompetent population examined, was no greater than 0.19%, in agreement with historical data on ACV. Although PCV and ACV share an identical activation pathway and a similar mode of action, suggesting that the mechanisms of resistance are similar, the widespread usage of ACV for treatment has not increased the prevalence of PCV resistance above that reported for ACV. In conclusion, based on the PCV susceptibilities of sequential isolates from patients with recurrent herpes labialis taken throughout the treatment period in this study, there is no reason to expect a change in the overall prevalence of resistant HSV isolates with use of topical PCV for acute treatment of recurrent herpes labialis in immunocompetent patients. Backmatter: PMID- 12183274 TI - In Vitro Activities of Garenoxacin (BMS-284756) against 170 Clinical Isolates of Nine Pasteurella Species AB - The in vitro susceptibilities of 170 clinical isolates plus 12 American Type Culture Collection strains of Pasteurella species comprising nine species and three Pasteurella multocida subspecies were studied by an agar dilution method. Garenoxacin (BMS-284756), a new des-fluoro(6) quinolone, was active at <=0.06 mug/ml against all isolates, including four beta-lactamase-producing strains, with >90% of the strains susceptible to <=0.008 mug/ml. Garenoxacin was generally 1 to 2 dilutions more active than levofloxacin and moxifloxacin and was the most active agent tested. Cefoxitin required 1 mug/ml for inhibition of 51 of 182 (29%) of strains, and 3 strains (also beta-lactamase producers) were resistant to doxycycline. Keywords: Introduction : Approximately 245,000 to 725,000 (5 to 15% of the 4.5 million people who suffer animal-bite wounds in the United States annually seek medical attention for infection, including 30,000 patients who visit an emergency department for medical treatment and 10,000 who are hospitalized with a bite wound infection . Bite wound infections involve a complex polymicrobial flora , among which Pasteurella multocida is considered an important pathogen by clinicians. Recent advances in molecular methods have suggested a more complex taxonomic structure of the genus Pasteurella, with 11 genetically closely related taxa, the establishment of three subspecies of P. multocida, and a reevaluation of 11 taxa with low levels of genetic relatedness . Many of the genetically closely related species have different ecological niches, different pathogenic potentials, and different propensities for specific tissue invasion. A recent prospective study demonstrated that Pasteurella species are the most common aerobic isolates from both dog bites (50%) and cat bites (75%). However, P. canis is the predominant isolate present in 26% of dog-bite wounds, followed by P. stomatis (12%) and P. multocida subsp. multocida (12%). In addition, P. dagmatis (4%) and P. multocida subsp. gallicida (2%) and P. multocida subsp. septica (10%) are also encountered in dog-bite wounds. In contrast, P. multocida subsp. multocida and P. multocida subsp. septica are more prevalent in cat-bite wound infections (present in 54 and 28% of such wound infections, respectively). Most studies of the antimicrobial susceptibilities of members of the family Pasteurellaceae are often limited because veterinary and human isolates are combined or the full spectrum of Pasteurella species causing human infections is not tested. Pasteurella species are associated with a variety of human infections, most but not all of which are associated with animal contact, especially dog and cat bites. While many patients with bite wounds receive a beta-lactam agent such as amoxicillin-clavulanate, approximately 20% will report a history of an adverse reaction to penicillin or other beta-lactam antibiotics and require an appropriate alternative agent. Fluoroquinolones have been suggested as alternative agents for the treatment of bite wounds. Garenoxacin {BMS-284756; T-3811ME; 1-cyclopro-pyl-8-(difluoromethoxy)-7-[(1R)-1-methyl-2,3-dihydro-1H-5-isoindolyl]-4-oxo-1,4-dihydro-3-quinolinecarboxylic acid methanesulfonate monohydrate} is a new des-fluoro(6) quinolone that lacks the 6-position fluorine that characterizes the previous class of fluoroquinolones. Preliminary data indicate that this drug has a broad spectrum of activity against most gram-positive and gram-negative aerobes including certain strains that are resistant to other fluoroquinolones. We therefore studied the antimicrobial susceptibilities of 170 clinical isolates and 12 American Type Culture Collection (ATCC) strains of the family Pasteurellaceae consisting of the nine species and three P. multocida subspecies most often associated with human infections. Almost all of the strains had previously been isolated from bite wounds (almost all of the strains were isolated from 1995 to 2002) and were identified according to standard criteria . The sources of the isolates were as follows: dog bites, 60 strains; cat bites, 81 strains; tiger bites, 2 strains; bites caused by other animals, 15 strains; and bovine respiratory sources, 12 strains. The three isolates of P. caballi and six of the eight strains of P. haemolytica were from bovine respiratory sources. For comparative purposes, 12 ATCC strains of 10 species (P. aerogenes ATCC 12192, P. canis ATCC 43326, P. dagmatis ATCC 43325, P. haemolytica ATCC 33396, P. multocida subsp. gallicida ATCC 51689 and ATCC 51696, P. multocida subsp. multocida ATCC 7228 and ATCC 12947, P. multocida subsp. septica ATCC 51688, P. pneumotropica ATCC 35149, P. stomatis ATCC 43327, and P. testudinis ATCC 3368) were also tested. Standard laboratory powders were supplied, as follows: garenoxacin, Bristol-Myers Squibb Co., Princeton, N.J.; amoxicillin-clavulanate, SmithKline Beecham Pharmaceuticals, Philadelphia, Pa.; ampicillin-sulbactam, Pfizer Inc., New York, N.Y.; levofloxacin, Ortho-McNeil Pharmaceuticals, Raritan, N.J.; moxifloxacin, Bayer Corp., West Haven, Conn.; cefoxitin, Merck & Co., West Point, Pa.; and doxycycline and penicillin G, Sigma Chemical Co., St. Louis, Mo. Antimicrobial agents were reconstituted according to the instructions of the manufacturers. Serial twofold dilutions were added to the media on the day of testing. Frozen cultures were transferred twice to ensure purity and good growth. Susceptibility testing was performed according to NCCLS standards . The basal medium used was Mueller-Hinton agar supplemented with 5% sheep blood. The agar plates were inoculated with a Steers replicator (Craft Machine Inc., Chester, Pa.) with an inoculum of 104 CFU per spot. Control plates without antimicrobial agents were inoculated before and after the inoculation of each set of drug-containing plates. Escherichia coli ATCC 25922 and Staphylococcus aureus ATCC 29213 were included as controls. Plates were incubated at 35C in an aerobic environment for 18 to 20 h and were then examined; the MIC was defined as the lowest concentration of an agent that yielded no growth or a marked change in growth compared to the growth on the control plate. The results of the present study, in which we tested a large number of clinical isolates of nine Pasteurella species, including the three P. multocida subspecies, are presented in Table . Garenoxacin was active at <0.06 mug/ml against all isolates, including four beta-lactamase-producing strains (P. aerogenes, n = 1; P. caballi, n = 1; P. haemolytica, n = 2), with >90% of the strains susceptible to <0.008 mug/ml. Three of these four strains, including the two strains of P. haemolytica and the single strain of P. aerogenes, were also resistant to doxycycline. Cefoxitin required 1 to 2 mug/ml for inhibition of 51 of 182 (29%) strains. Garenoxacin was generally 1 to 2 dilutions more active than levofloxacin and moxifloxacin on a weight basis and was the most active agent tested. TABLE 1 | Comparative in vitro activities of the des-quinolone garenoxacin (BMS-284756) and seven other agents against 182 clinical isolates of Pasteurella species Pasteurella species are normal inhabitants of 12 to 92% of dogs and 52 to 99% of cats , and most human infections are associated with bites or animal exposure. Many publications do not differentiate or identify the different Pasteurella species, sometimes due to technical difficulties with the interpretation of biochemical tests . Our laboratory has recently reported on the use of PCR fingerprinting and alpha-glucosidase activity as a means of differentiating P. multocida subsp. multocida from P. multocida subsp. septica. Some investigators have suggested different pathogenic potentials and ecological niches for the different species, which increases the clinical importance of differentiating these species in reports of studies. For example, P. multocida subsp. multocida and P. multocida subsp. septica are more frequently recovered from "more serious" infections , including bacteremias, and P. multocida subsp. septica is more frequently isolated from cat-bite wounds and has an affinity for the central nervous system . In a prior study from our laboratory , in which the same methodology described here but a limited number of isolates was used, we found that all Pasteurella multocida subsp. multocida and P. multocida subsp. septica isolates were susceptible to <=0.015 mug of garenoxacin per ml. In general, the quinolones, including gatifloxacin , moxifloxacin , trovafloxacin , and levofloxacin , have consistently been active against P. multocida and other Pasteurella species. On the basis of our in vitro data, garenoxacin offers an alternative for the treatment of human infections associated with all Pasteurella species. Backmatter: PMID- 12183275 TI - In Vitro and In Vivo Antibacterial Activities of DW286, a New Fluoronaphthyridone Antibiotic AB - The in vitro and in vivo activities of DW286, a novel fluoronaphthyridone with potent antibacterial activity, were compared with those of ciprofloxacin, gemifloxacin, sparfloxacin, and trovafloxacin. Against gram-positive bacteria, such as Staphylococcus aureus, Staphylococcus epidermidis, Streptococcus pneumoniae, and Enterococcus faecalis, the in vitro activity of DW286 was stronger than that of any other reference antibiotic. Against gram-negative bacteria, the activity of DW286 was similar to those of trovafloxacin and gemifloxacin but was weaker than that of ciprofloxacin. In a mouse systemic infection caused by three S. aureus strains, including methicillin-resistant S. aureus and quinolone-resistant S. aureus (QRSA), DW286 demonstrated the most potent activity, as found in vitro. Specially, DW286 is >=8-fold more active against QRSA than the other fluoroquinolones. And the 50% protective doses for DW286 were correspondent with the in vitro activities. Keywords: Introduction : The quinolones have evolved from agents used solely for the treatment of urinary tract infections to molecules with potent activity against a wide spectrum of significant bacterial pathogens. Progressive modifications in their molecular configurations have improved both the spectrum and potency of their in vitro activities . These compounds have been successfully used in clinics for a decade. The targets in fluoroquinolone research during the last few years include improving the pharmacokinetic properties, increasing the activity against gram-positive cocci and anaerobes and against fluoroquinolone-resistant strains, and improving activity against nonfermentative gram-negative species . In addition, the increasing pathogen resistance to antimicrobial agents is a cause of concern. This increase in the resistance to fluoroquinolone emphasizes the importance of the continued development of new structural candidates . The use of newer fluoroquinolone derivatives will potentially contribute to decreasing the spread of resistance to other antimicrobial agents by reducing the selective pressure on other antibiotic groups . New candidates have been developed in an attempt to help improve this situation. DW286, 7-[3-(aminomethyl)-4-(methoxyimino)-3-methyltetrahydro-1H-1-pyrrolyl]-1-cyclopropyl-6-fluoro-4-oxo-1, 4-dihydro[1,8]naphthyridine-3-carboxylic acid hydrochloric acid salt, is a novel fluoronaphthyridoneantibacterial agent that was synthesized for this purpose . In this study, we determined the in vitro activity of DW286 in several groups of clinical isolates and compared it with those of ciprofloxacin, gemifloxacin, sparfloxacin, and trovafloxacin. We also compared the in vivo protective efficacy of DW286 with those of the fluoroquinolones against a systemic infection in mice. The studied compounds were obtained as follows: DW286, ciprofloxacin, sparfloxacin, and gemifloxacin were synthesized at the R & D Center, Dong Wha Pharmaceutical Company, Anyang City, Korea. Trovafloxacin was provided by Pfizer Pharmaceuticals, New York, N.Y. Clinical isolates were obtained from hospitals in Seoul, Korea, between 1996 and 2000. The challenge organisms used in the mouse systemic infections were as follows: Streptococcus pyogenes ATCC 8668 was obtained from the American Type Culture Collection (Rockville, Md.), and Staphylococcus aureus Smith was kindly provided by S. Goto of Toho University, Tokyo, Japan. Escherichia coli MB4-01, Klebsiella pneumoniae MB4-02, Streptococcus pneumoniae MB4-21, Pseudomonas aeruginosa MB4-16, methicillin-resistant S. aureus (MRSA) MB4-19, and quinolone-resistant S. aureus (QRSA) MB4-20 were selected through the pathogenicity screening of clinical isolates . MICs were determined by an agar dilution method with Mueller-Hinton agar (Difco Laboratories, Detroit, Mich.) following the National Committee for Clinical Laboratory Standards procedure . Haemophilus influenzae, S. pyogenes, and S. pneumoniae were grown on brain heart infusion (Difco) agar supplemented with 5% defibrinated sheep blood (Difco) at 37 C. The in vitro antibacterial activity of DW286 against the clinical isolates is presented in Table . Its antibacterial activity was superior to those of ciprofloxacin, gemifloxacin, sparfloxacin, and trovafloxacin against quinolone-sensitive S. aureus (DW286, MIC at which 50% of the isolates tested are inhibited [MIC50] = 0.008 mug/ml; MIC90 = 0.016 mug/ml) and QRSA (DW286, MIC50 = 0.5 mug/ml; MIC90 = 8 mug/ml). DW286 was also more active than ciprofloxacin and sparfloxacin and was slightly superior to trovafloxacin and gemifloxacin against Staphylococcus epidermidis (DW286, MIC50 = 0.031 mug/ml; MIC90 = 0.25 mug/ml), S. pyogenes (DW286, MIC50 = 0.016 mug/ml; MIC90 = 0.031 mug/ml), and Enterococcus spp. (DW286, MIC50 = 1 to 4 mug/ml; MIC90 = 4 to 8 mug/ml). Its antibacterial activity against S. pneumoniae was fourfold stronger than those of gemifloxacin and trovafloxacin (MIC90 = 0.125 mug/ml for DW286 and 0.5 mug/ml for gemifloxacin and trovafloxacin). In summary, DW286 exhibited potent in vitro antibacterial activities against gram-positive organisms. Specially, the improved antibacterial efficacy of DW286 against pneumococci was the most prominent difference. Against gram-negative bacteria, including members of the family Enterobacteriaceae, DW286 was less active than ciprofloxacin, but it was as active as trovafloxacin and gemifloxacin. The MIC50 and MIC90 for clinical strains of E. coli, K. pneumoniae, Klebsiella oxytoca, Enterobacter cloacae, Enterobacter aerogenes, and Morganella morganii were 0.016 to 0.25 mug/ml and 0.125 to 0.5 mug/ml, respectively. Its activity against P. aeruginosa (MIC50 = 4 mug/ml; MIC90 = 32 mug/ml) was inferior to that of ciprofloxacin (MIC50 = 0.25 mug/ml; MIC90 = 2 mug/ml) and was slightly inferior to those of sparfloxacin, gemifloxacin, and trovafloxacin. Against Serratia marcescens, Xanthomonas maltophilia, Acinetobacter calcoaceticus, and Proteus spp., DW286 was less active than the other antibiotics. Its activity against gram-negative organisms was comparable to those of trovafloxacin and gemifloxacin but was less than that of ciprofloxacin. TABLE 1 | In vitro antibacterial activities of DW286 and quinolones against clinical isolates The effects of serum on the antibacterial activity of DW286 against S. aureus ATCC 25923, Bacillus subtilis ATCC 6633, and E. coli ATCC 25922 were also evaluated. Heat-inactivated horse serum (Gibco BRL, Gaithersburg, Md.) to a final concentration of 10 or 25% (vol/vol) had no significant effect on the activities of DW286 and the other quinolones tested (data not shown). The protective efficacy of the fluoroquinolones in a systemic infection model in mice was determined. The organisms used were S. aureus Smith, S. pyogenes ATCC 8668, E. coli MB4-01, K. pneumoniae MB4-02, P. aeruginosa MB4-16, S. pneumoniae MB4-21, MRSA MB4-19, and QRSA MB4-20. The organisms were cultured overnight in brain heart infusion broth at 37C. Male ICR mice (body weight, 18 to 22 g; age, 4 weeks) were inoculated intraperitoneally with 0.3 ml of a bacterial suspension adjusted with 3% gastric mucin (ICN Biomedicals, Columbus, Ohio) in saline solution at 100 times the minimal lethal dose. The challenge inoculum was sufficient to kill 100% of the untreated control mice, which died within 48 h postinfection. Each test compound was administered once orally to mice immediately after infection. Five groups of seven mice each were treated with different doses of each antibacterial agent. The 50% protective dose (PD50) was calculated by the method of Litchfield and Wilcoxon from the survival rates on day 7 after infection. The protective effects of DW286 against systemic infections were compared with those of ciprofloxacin, sparfloxacin, trovafloxacin, and gemifloxacin . In this mouse infection model, DW286 showed good in vivo efficacies against gram-positive organisms such as S. aureus, including MRSA and QRSA, S. pyogenes, and S. pneumoniae. In infections caused by E. coli, K. pneumoniae, and P. aeruginosa, its activity was inferior to that of ciprofloxacin. However, its in vivo efficacies were comparable to those of trovafloxacin and gemifloxacin. The PD50s for DW286 correlated with the in vitro MICs. Based on its in vitro activity and in vivo efficacy, DW286 seems to be a promising antibiotic with a broad spectrum of antimicrobial activity, including potent activity against S. aureus and S. pneumoniae. TABLE 2 | Protective efficacy of DW286 on systemic infections in mice The pharmacokinetic studies of DW286 were performed after oral administration in mice, rats, and dogs (C. Lee, Y. Jung, J. Ryu, H. Kwon, and J. Lee, Abstr. 41st Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-555, 2001). The mean maximum concentrations (in micrograms per milliliter) of DW286 in serum in mice were 1.06, 8.55, 24.28, and 96.54, and the total areas under the concentration-time curve (expressed in micrograms hours/milliliter) were 3.06, 30.56, 110.47, and 783.90 at each dose of 10, 50, 250, and 1,000 mg/kg of body weight, respectively. In a single oral administration in rats (40 mg/kg) and dogs (10 mg/kg), the maximum concentrations of DW286 in serum were 9.06 mug/ml (2.0 h) in rats and 3.95 mug/ml (3.33 h) in dogs. The elimination half-life and the total area under the concentration-time curve were 4.5 h and 90.82 mug h/ml in rats and 5.37 h and 34.65 mug h/ml in female dogs. Hence, the pharmacokinetics of DW286 seem to be superior to those of ciprofloxacin. The new fluoronaphthyridone antibiotic, DW286, was absorbed efficiently from the gastrointestinal tract, regardless of the animal species. The pharmacokinetic data explain the correlation between the in vitro activity and the in vivo efficacy. Our findings need to be confirmed in further preclinical and clinical studies of this new fluoronaphthyridone antibiotic. FIG. 1. | Chemical structure of DW286. Chemical structure of DW286. Backmatter: PMID- 12183231 TI - Frequency of Disinfectant Resistance Genes and Genetic Linkage with beta-Lactamase Transposon Tn552 among Clinical Staphylococci AB - A total of 61 strains of Staphylococcus aureus and 177 coagulase-negative staphylococcal strains were isolated from the blood of patients with bloodstream infections and from the skin of both children under cancer treatment and human immunodeficiency virus-positive patients. The MIC analyses revealed that 118 isolates (50%) were resistant to quaternary ammonium compound-based disinfectant benzalkonium chloride (BC). The frequencies of resistance to a range of antibiotics were significantly higher among BC-resistant staphylococci than among BC-sensitive staphylococci. Of 78 BC-resistant staphylococcal isolates, plasmid DNA from 65 (83%), 2 (3%), 43 (55%), and 15 (19%) isolates hybridized to qacA or -B (qacA/B), qacC, blaZ, and tetK probes, respectively. The qacA/B and blaZ probes hybridized to the same plasmid in 19 (24%) staphylococcal strains. The plasmids harboring both qacA/B and blaZ genes varied from approximately 20 to 40 kb. The Staphylococcus epidermidis Fol62 isolate, harboring multiresistance plasmid pMS62, contained qacA/B and blaZ together with tetK. Molecular and genetic studies indicated different structural arrangements of blaZ and qacA/B, including variable intergenic distances and transcriptional directions of the two genes on the same plasmid within the strains. The different organizations may be due to the presence of various genetic elements involved in cointegration, recombination, and rearrangements. These results indicate that qac resistance genes are common and that linkage between resistance to disinfectants and penicillin resistance occurs frequently in clinical isolates in Norway. Moreover, the higher frequency of antibiotic resistance among BC-resistant strains indicates that the presence of either resistance determinant selects for the other during antimicrobial therapy and disinfection in hospitals. Keywords: Introduction : Staphylococci, including Staphylococcus aureus and the coagulase-negative staphylococci (CNS), and enterococci account for approximately one-third of all bloodstream infections and up to 50% of nosocomial bloodstream infections. In the past, CNS were considered to be very rarely involved in disease, but since the 1980s there has been increasing evidence that CNS can act as opportunistic pathogens and can be a frequent and important cause of disease. In recent years there has been a dramatic increase in the prevalence of multiple-drug-resistant strains. Worldwide emergence of antimicrobial resistance in staphylococci is probably a serious and increasing problem, especially in hospitals . Disinfectants based on quaternary ammonium compounds (QACs), e.g., benzalkonium chloride (BC), cetylpyridinium chloride, cetrimide, proceine, and detizor, are frequently used in hospitals to disinfect and to prevent the spread of pathogens. It has been suggested that the widespread use of QACs may impose a selective pressure and contribute to the emergence of disinfectant-resistant microorganisms in these environments . Known qac resistance genes reported in clinical staphylococci, qacA, qacB, and qacC/smr, are generally plasmid borne and are widely distributed in the environment . Large consumption of antibiotics by both humans and animals has resulted in the development and spread of a large number of antibiotic resistance determinants among bacterial populations, thus creating critical public health problems. Concerns have arisen regarding the potential emergence of cross-resistance and coresistance between widely used disinfectants and antibiotics . Resistance genes on transferable genetic elements such as plasmids and transposons may lead to the epidemic spread of resistance between species. On some of the published plasmids (pST6, pSK4, and pSK41) and transposons (Tn552 and Tn4002), qac resistance determinants are located together with antibiotic resistance genes encoding resistance to gentamicin (GEN), trimethoprim (TMP), penicillin, kanamycin (KAN), and tobramycin . Staphylococcal beta-lactamase structural gene blaZ and two closely linked genes, blaI and blaR, that control its expression have been identified on several transposons, e.g., Tn552 , and large plasmids, e.g., pST6 , as well as on the chromosome . Staphylococcal beta-lactamase transposon Tn552 is closely related to staphylococcal transposons Tn4002, Tn3852, and Tn4201 and to Tn4201-like elements. Staphylococcal transposons contains terminal inverted repeats (TIR) at their ends that serve as recognition sequences for transposase in their role of fusing the ends of the transposon with the recipient DNA. The three genes (blaI, blaR, and blaZ) involved in beta-lactamase production constitute the right half of transposon Tn552 (Tn552-like bla gene module). The left half of Tn552 consists of genes p271, encoding a potential ATP-binding protein, p480, encoding a transposase, binL, encoding a resolvase, and a resolution site, resL. Partial and complete genetic structural arrangements of Tn552 in staphylococci from different geographic locations have been described . Some staphylococci harbor only the right or left half of Tn552. Staphylococcal insertion sequence IS257, associated with Tn552 and IS256, has been shown to be an active mobile genetic element . There is limited knowledge concerning the frequency of QAC resistance and the genetic linkage between qac-mediated resistance and antibiotic resistance genes. In this study, we investigated the presence and genetic linkage of known qac resistance determinants and antibiotic resistance genes, along with genetic mobile elements, in staphylococci isolated from clinical environments. MATERIALS AND METHODS : Bacterial strains, culture media, and growth conditions. | A total of 61 isolates of S. aureus and 177 CNS isolates were included in this study . The staphylococci were isolated from patients in four different hospitals. S. aureus (n = 61) and CNS (n = 60) were isolated from patients with bloodstream infections at two hospitals (Molde County Hospital and Norwegian Radium Hospital) during 1991 to 1992 and 1995 to 1996 . The CNS isolates (n = 56) from the skin of human immunodeficiency virus (HIV)-positive patients were from Ulleval Hospital, while the remaining 61 CNS isolates were obtained at two hospitals (Ulleval Hospital and Rikshospitalet University Hospital) from the skin of children under cancer treatment receiving cytostatica and undergoing multiple-antibiotic cures. In all cases, repeat isolates from the same patient were excluded. This strain collection represents staphylococci from different types of hospitals and patients subject to various degrees of antimicrobial treatment. Reference strains were Staphylococcus haemolyticus NVH97A (resistant to BC and penicillin) and S. aureus RN4220, S. haemolyticus DSM20623, and S. aureus ATCC 25923 (sensitive to BC and penicillin). All staphylococcus strains were cultured in Mueller-Hinton (MH) broth or on MH agar plates (Difco Laboratories, Detroit, Mich.) at 37C. Isolates were preserved in MH broth with 15% glycerol at -80C. Antimicrobial agents. | BC was from the Norwegian Medical Depot (Oslo, Norway). Ampicillin (AMP), penicillin G (PEN), methicillin (MET), tetracycline (TET), erythromycin (ERY), KAN, GEN, chloramphenicol, TMP, and ethidium bromide (EBR) were purchased from Sigma Chemical Co. (St. Louis, Mo.), and chlorhexidine (CHX) was from Nycomed (Oslo, Norway). Amphoteric disinfectant Tego 103G was purchased from Otto Olsen AS (Lillestrom, Norway). Neo-Sensitabs disks (Rosco, Medkjemi A/S, Taastrup, Denmark) containing antimicrobial agents were used for susceptibility testing. Susceptibility tests. | MICs of antimicrobial agents and dyes were determined by the microdilution method and by E test using MH medium. The lowest concentration of an antimicrobial agent and dye that totally inhibited bacterial growth after 24 h of incubation at 37C was considered the MIC. Antibiotic susceptibility patterns were also examined by the disk diffusion method (Neo-Sensitabs user's guide, Rosco). The diameters of the inhibition zones around antimicrobial disks were measured in millimeters after 2 days of incubation at 37C and compared to standard MIC breakpoints recommended by "Susceptibility testing standardization groups" (Neo-Sensitabs user's guide, Rosco). The results were interpreted as recommended by the Norwegian AFA Group (Neo-Sensitabs user's guide). Susceptibility tests for the isolates resistant to BC and PEN were repeated two times on different days. MIC testing with disinfectants BC and Tego 103G was carried out by using 1-mug/ml intervals from 0 to 10 mug/ml. For CHX the interval was 0.5 mug/ml from 0 to 4 mug/ml. MIC tests with PEN, AMP, and EBR were performed by using twofold dilutions starting from 320 mug/ml. DNA isolation and analysis. | Cells were lysed with lysostaphin (Sigma-Aldrich, St. Louis, Mo.) at a concentration of 40 mug/ml and incubated at 37C for 30 min. Total DNA and plasmid DNA were isolated by Easy-DNA (Invitrogen, Carlsbad, Calif.) and QIAprep spin (Qiagen GmbH, Hilden, Germany) kits, respectively. After agarose gel electrophoresis, the DNA was stained with EBR and visualized under UV light. Plasmid curing. | Strains were grown in increasing sublethal concentrations of novobiocin in MH broth as described by Heir et al. . Southern blotting and hybridization. | Plasmid DNA and total DNA were transferred from agarose gel to a Hybond-N+ nylon membrane (Amersham Pharmacia Biotech, Buckinghamshire, United Kingdom) along with positive and negative control DNA by vacuum blotting, according to the manufacturer's instructions (Pharmacia, Uppsala, Sweden). PCR products (see below) specific for genes qacA and qacB (common probe), qacC, qacG, qacH, blaZ, blaR, blaI, ermC, cat, tetK, aacA-aphD, and dfrA and insertion sequences IS257 and IS256 were used as probes. The PCR products were purified with a Qiagen PCR product purification kit and sequenced on an ABI PRISM 377 (Applied Biosystems, Perkin-Elmer Cetus Corp., Norwalk, Conn.) before being used as probes. The labeling of probes and hybridization were done with an AlkPhos direct gene image kit in accordance with the manufacturer's instructions (Amersham Pharmacia Biotech). Membrane stripping was done as recommended by the manufacturer. PCR amplification. | The primers used for PCR amplifications were designed from previously known sequences . All the designed primers were first tested for their specificity on known positive and negative control strains. To obtain probes for DNA hybridization, standard PCR was carried out for qacA or -B (qacA/B), qacC, qacG, qacH, blaZ, blaR, blaI, ermC, cat, tetK, aacA-aphD, dfrA, IS257, and IS256. Each PCR mixture contained 50 ng of plasmid DNA, 2 muM (each) primer, 200 muM (each) deoxynucleoside triphosphate, 2.5 muM MgCl2, 1x reaction buffer (Promega Corp., Madison, Wis.), and 2.5 U of Taq DNA polymerase (Promega Corp.) in a total volume of 50 mul. Reaction mixtures were subjected to 30 cycles of amplification. The conditions for each cycle were denaturation for 1 min at 95C, annealing for 1 min at between 50 and 55C depending on the primer set, and primer extension for 2 min at 72C. Finally, reaction mixtures were incubated at 72C for 10 min. The PCR products were separated by electrophoresis in a 1% agarose gel, stained with EBR, and visualized under UV light. XL-PCR. | The primers used for extralong PCR (XL-PCR) amplification are included in Table . The XL-PCR was carried out with a GeneAmp XL-PCR kit (Applied Biosystems) in accordance with the manufacturer's instructions with slight modifications. The GeneAmp XL-PCR kit (Perkin-Elmer Cetus Corp.) was used for amplification of the DNA regions between the beta-lactamase genes and disinfectant resistance gene qacA/B. The primer combinations used were qac IV and blaI R2 and qac IV and blaZ 2F (see Fig. ; Table ). The DNA region between qacA/B and staphylococcal insertion element IS257 was amplified by using primers qac IV and IS257 F (see Fig. ). Each PCR mixture contained 10 ng of plasmid DNA, 2 muM (each) PCR primer, 200 muM (each) deoxynucleoside triphosphate, 1 mM MgCl2, and 1.5 U of rTh DNA polymerase-XL in a total volume of 50 mul. The cycle conditions were preliminary denaturation for 2 min at 93C, followed by 25 cycles of denaturation at 93C for 1 min, annealing for 45 s at between 47 and 50C depending on the primer set, and primer extension for 7 min at 68C. Finally, reaction mixtures were incubated at 68C for 5 to 12 min depending on the primer set. The PCR products were examined by electrophoresis in a 0.7 or 1.0% agarose gel, stained with EBR, and visualized under UV light. DNA sequencing. | The nucleotide sequences of PCR products and XL-PCR products were determined by using an ABI PRISM BigDye terminator cycle sequencing ready reaction kit (Applied Biosystems) with synthetic oligonucleotide primers on an ABI PRISM 377 automatic sequencer (Applied Biosystems) as recommended by the manufacturer. Before application on the sequencer, the extension products were purified as described by the manufacturer. Unknown flanking DNA fragments were isolated by a two-step PCR method (A. B. Sorensen, M. Duch, and F. S. Pedersen, Dynalogue customer report 3:2-3, 1999). The unknown DNA flanking regions were amplified with primers coupled to biotin at the 5' end and degenerate flanking primers in first-step PCR. The PCR products were bound to Dynabead-streptavidin beads and purified with a Dynabeads kilobaseBINDER kit (Dynal, Oslo, Norway). The purified PCR products from the first-step PCR were then used as templates for nested PCR followed by DNA sequencing. Nucleotide sequences were analyzed by using the BLAST website program (National Center for Biotechnology Information, National Institutes of Health), the Sequencher, version 3.0, software package (Gene Codes Corporation, Ann Arbor, Mich.), and the GCG sequence analysis software package, version 8 (Genetics Computer Group, Madison, Wis.). Nucleotide sequence accession numbers. | The nucleotide sequences reported in this study have been assigned accession no. , , and in the GenBank database. FIG. 2. | Genetic organization of qac and beta-lactam resistance genes in staphylococcal isolates Fol24, Fol33, Fol62, Fol89, Fol90, and Fol100. Genetic organization of qac and beta-lactam resistance genes in staphylococcal isolates Fol24, Fol33, Fol62, Fol89, Fol90, and Fol100. Genes involved in qac and beta-lactam resistance are gray and black, respectively. qacA/B encodes QAC resistance; qacR encodes a putative repressor of qacA/B; sin is a putative staphylococcal recombinase gene; p271 encodes a potential ATP-binding protein; p480 encodes a transposase; binR and binL are resolvase-encoding genes; blaZ is a beta-lactamase structural gene; blaR and blaI are beta-lactamase regulatory genes; IS257 is a staphylococcal insertion sequence. Primers for detection of specific PCR products are indicated for Fol33. Vertical bars and boxes downstream of blaI, TIR of transposon Tn552 and a resolution site (res), respectively. TABLE 1 | Screening for phenotypic susceptibility to BC and PEN among clinical staphylococci TABLE 2 | Primers used in PCR and XL-PCR amplification RESULTS : Susceptibility testing. | The 238 staphylococcus isolates were screened for QAC (BC) and PEN resistance . The strains were categorized as BC resistant or sensitive according to the BC MICs. One hundred eighteen (50%) isolates were considered BC resistant (BC MICs between 3 and 8 mug/ml), and 120 isolates were sensitive to BC (BC MICs <= 2 mug/ml). Of the 238 staphylococcus isolates, 143 were found resistant to PEN. One hundred twenty-one CNS isolates from patients with bloodstream infections (n = 60) and from the skin of children under cancer treatment (n = 61) were analyzed for correlation between BC and antibiotic resistance . This analysis showed that the frequency of antibiotic resistance was higher among the BC-resistant strains. As an example, a large number of BC-resistant strains also showed PEN resistance. On the basis of Fisher's exact test , these distributions appear not to be random (P <= 0.01 for PEN, oxacillin, cephalothin, cefuroxime, imipenem, TMP, GEN, doxycycline, and fusidic acid). None of the isolates were resistant to vancomycin, and less than 3% of the staphylococcus isolates were resistant to teicoplanin and rifampin (results not shown). The CNS from HIV-positive outpatients and S. aureus from patients with bloodstream infections were excluded from the analyses of the correlation between BC and antibiotic resistance, because the frequencies of antibiotic resistance among these isolates were relatively low (data not shown). The 118 BC-resistant isolates were screened for susceptibility to a second disinfectant, CHX . Twenty-eight of the BC-resistant isolates showed low resistance to CHX (MIC between 1.5 and 3.0 mug/ml), whereas 90 isolates were sensitive (MIC <= 1.0 mug/ml). Likewise, of 42 BC-resistant staphylococcus isolates randomly selected among strains for which the BC MIC was >=4 mug/ml, 11 and 37 isolates were resistant to disinfectant Tego 103G and dye EBR, respectively . Systematic cross-resistance between BC and EBR and no systematic cross-resistance between BC and CHX or Tego 103G were found in staphylococci. The MIC tests and disk diffusion test revealed that an isolate from a bloodstream infection, Staphylococcus epidermidis Fol62, was resistant to BC (MIC, 5 mug/ml), EBR (MIC, 160 mug/ml), Tego 103G (MIC, 10 mug/ml), PEN (MIC, 160 mug/ml), AMP (MIC, 160 mug/ml), MET (MIC, 50 mug/ml), TET (MIC, 70 mug/ml), KAN (MIC, 7 mug/ml), ERY (MIC, 80 mug/ml) and TMP (MIC, 8 mug/ml). DNA hybridizations. | Of the 118 BC-resistant isolates, 78 isolates for which the BC MICs were high (MICs between 4 and 8 mug/ml) were selected for DNA hybridization studies. Plasmid DNA from these isolates was screened for the presence of known gram-positive bacterium qac genes mediating resistance (qacA/B, qacC, qacG, and qacH), beta-lactamase genes (blaZ, blaI, and blaR), other antibiotic resistance genes (ermC, cat, dfrA, and tetK), and staphylococcal insertion sequences IS257 and IS256 by Southern blotting and hybridization . Sixty-seven of the BC-resistant isolates harbored either qacA/B or qacC on plasmid DNA, and the qacA/B gene was detected by PCR (total DNA was the template) in the remaining 11 isolates. Neither the qacG nor the qacH gene was detected among the isolates. Plasmids from 43 strains also hybridized with beta-lactamase probes. In 19 of these strains, qac and blaZ, blaI, and blaR resided on the same plasmid. Only S. epidermidis Fol62 plasmid pMS62 (32 kb) hybridized with other antibiotic resistance gene probes tested (tetK, IS257, ermC, aacA-aphD, and dfrA) in addition to qacA/B, blaZ, blaI, and blaR. tetK and a copy of IS257 were localized adjacent to blaZ. However, no XL-PCR products were obtained when DNA stretches between tetK or qacA/B and antibiotic resistance gene ermC, aacA-aphD, or dfrA were amplified. The determinant for MET resistance (mecA) in S. epidermidis Fol62 was chromosomally encoded. Genetic organizations of the qac and beta-lactamase genes. | From the 19 staphylococcus isolates harboring disinfectant qac and beta-lactamase genes on the same plasmids, we selected six isolates for a study of gene organization, isolates Fol24 (pMS24, 30 kb), Fol33 (pMS33, 35 kb), Fol62 (pMS62, 32 kb), Fol89 (pMS89, 40 kb), Fol90 (pMS90, 35 kb), and Fol100 (pMS100, 40 kb) (plasmid designations and approximate sizes are in parentheses). These isolates were all identified as S. epidermidis. Isolates Fol24, Fol33, and Fol62 were from patients having bloodstream infections, and isolates Fol89, Fol90, and Fol100 were from the skin of children treated for cancer. Results obtained by XL-PCR and DNA sequencing indicated different structural arrangements of the beta-lactamase genes and the disinfectant resistance qacA/B genes, including variable intergenic distances and different transcriptional orientations of genes located on the same plasmids . Staphylococcal plasmids pMS24, pMS33, and pMS89 harbor a complete copy of transposon Tn552, comprising p271, p480, and binL, encoding a potential ATP-binding protein, a transposase, and a resolvase, respectively, in addition to the beta-lactamase structural gene (blaZ) and regulatory genes (blaR and blaI). Staphylococcal insertion sequence IS257 was found downstream of the blaZ gene in these plasmids. There were sequences homologous to the right TIR (TIRR) of Tn552 downstream of blaZ, a resL site between blaI and binL, and a sequence homologous to TIRL of Tn552 downstream of p271. Isolates Fol24, Fol33, and Fol89 harbored the qacA/B gene together with their regulatory gene, qacR, downstream of the p271 gene. However, the region between p271 and qacR was different for different isolates. Fol24 and Fol89 had very similar regions and contained the sin gene, encoding a putative recombinase, while the sin gene was absent in Fol33. Plasmid pMS62 harbored the incomplete right half of the transposon Tn552-like bla gene module upstream of the binR and sin genes. Also in pMS62, IS257 was found downstream and adjacent to blaZ. The tetK gene, encoding TET resistance, was located approximately 2.4 kb further downstream. In Fol90 and Fol100 the orientation of beta-lactamase gene clusters was inverted relative to the qacA/B gene cluster, so that qacR was situated downstream of blaZ. In pMS90 the intergenic sequence between blaZ and qacR contained only the inverted repeat of Tn552, while in pMS100 this region contained a sin and an additional noncoding 743 bp of DNA. DNA hybridization and PCR confirmed the presence of staphylococcal insertion sequence IS257 on pMS90 and pMS100. When the DNA stretches between the beta-lactamase genes and disinfectant resistance gene qacA/B were amplified by using primer combinations qac IV and blaI R2 and qac IV and blaZ 2F, both primer sets gave rise to PCR products in strains Fol33 and Fol89 . DNA sequencing of the PCR products of qac IV and blaZ 2F revealed that the blaZ and the qacR genes were located very close to each other, with the TIR of Tn552 between the genes. These strains apparently harbored the gene clusters in two orientations. The nucleotide and deduced amino acid sequences derived from complete and incomplete (Tn552-like bla gene module) Tn552 transposons of plasmids pMS24, pMS33, pMS89, and pMS62 were identical to those derived from staphylococcal beta-lactamase transposon Tn552 . The deduced amino acid sequences derived from qacR of plasmids pMS24, pMS33, pMS62, pMS89, pMS90, and pMS100 and qacA/B of pMS33 were similar to those derived from pST6, pSK156, and pSK23. Similarly, the deduced amino acid sequences derived from the resolvase gene (binR) of pMS62 was 100% identical to that derived from pST6 (AY028779) and 99% similar to that derived from pNVH96 (AJ302698). The recombinase genes (sin) of pMS62 and pMS89 were 100% identical and 99% similar to those of pST6 and pNVH97A (AJ400722), respectively. The sin gene of Fol100 is 100% identical to that of S. epidermidis strain SR1 and 90% similar to the sin genes of Fol62 and Fol89. FIG. 1. | Percentages of antibiotic-resistant CNS isolates from patients with bloodstream infections (n = 60) and from the skin of children under cancer treatment (n = 61) among BC-resistant and BC-sensitive isolates. Percentages of antibiotic-resistant CNS isolates from patients with bloodstream infections (n = 60) and from the skin of children under cancer treatment (n = 61) among BC-resistant and BC-sensitive isolates. Antibiotics used are PEN, oxacillin (OXA), cephalothin (CEP), cefuroxime (CEF), imipenem (IMI), cotrimoxazole (COT), ERY, clindamycin (CLI), GEN, ciprofloxacin (CIP), doxycycline (DOX), chloramphenicol (CHL), and fusidic acid (FUS). Susceptibility to CEF, IMI, and CHL was tested only among CNS isolates from patients with bloodstream infections (asterisks). TABLE 3 | BC-resistant isolates were screened for susceptibility to disinfectants and dye by a microdilution method TABLE 4 | Screening for the presence of disinfectant and various antibiotic resistance genes on plasmid DNA among BC-resistant staphylococci DISCUSSION : Little is known about the occurrence and possible genetic linkage of qac and antibiotic resistance in staphylococci. Of the 238 human clinical isolates from Norway investigated in this study, 50% were phenotypically resistant to BC. Plasmid-borne qacA/B and qacC genes were detected in 83 and 3% of the BC-resistant staphylococcus isolates, respectively. The qacA/B genes were also detected in the remaining BC-resistant isolates, indicating that the genes were either chromosomally located or present on large plasmids not obtained by the plasmid isolation procedure. Previous investigators have reported a similar distribution of these three qac resistance genes in clinical S. aureus and CNS although additional staphylococcal disinfectant resistance genes also have been indicated . A lower prevalence of QAC resistance (13%) was reported for staphylococci from food and the food-processing industry . Here, both qacA/B and qacC as well as qacG and qacH were reported, with qacC as the most prevalent determinant. Overall, it seems likely that the presence, maintenance, and widespread dissemination of multidrug efflux qac genes in staphylococci are a consequence of the selective pressure exerted by the use of antiseptics and disinfectants . Many of the staphylococci included in this study were resistant to different antibiotics, but in general the antibiotic resistance was considered low compared to findings in other studies . Interestingly, we observed that staphylococci resistant to BC were generally more often resistant to antibiotics than BC-sensitive isolates . The results are compatible with selective advantages of isolates carrying both disinfectant and antibiotic resistance genes, and the data indicate that the presence of qac genes in staphylococci results in selection of antibiotic-resistant bacteria . Russell also speculated that disinfectant resistance might contribute to antibiotic resistance by coresistance or cross-resistance mechanisms or coselection . This study detected a systematic genetic linkage between resistance to BC and PEN although less than one-half of the plasmid-encoded beta-lactamase resistance was linked to disinfectant resistance genes (19 of 43 isolates). Previous investigators have also reported genetic linkage between disinfectant (qac) and antibiotic resistance genes (blaZ, aacA-aphD, dfrA, and ble) on the same staphylococcal plasmids from clinics , food environments , and an animal clinic (I.-L. Anthonisen et al., unpublished data). To our knowledge, this paper is the first report of closely linked qac, beta-lactamase, and tetK genes on a multiresistance plasmid (pMS62). Plasmid pMS62 also harbored ermC, dfrA, and aacA-aphD, encoding resistance to ERY, TMP, and GEN-KAN-tobramycin, respectively. No genetic linkage between qac and these or other antibiotic resistance genes tested was detected. The observed cross-resistance between BC and EBR is compatible with the ability of all known qac resistance determinants to cause the efflux of structurally similar, monovalent cationic agents. A low-level CHX resistance phenotype probably reflects isolates expressing qacA, which also encodes resistance to divalent cations , while resistance to the amphoteric disinfectant Tego 103G is probably due to other mechanisms. The genetic organizations between qac and beta-lactamase genes in six plasmids (pMS24, pMS33, pMS62, pMS89, pMS90, and pMS100) were studied . The plasmids originated from isolates from different sources and distant geographic areas, and their genetic organizations showed similarities with those of other known staphylococcal plasmids, e.g., pSK1, pSK4, pSK23, pSK57, pSK108, pSK156, pNVH97A, pNVH96, pI258, and pI9789 . Our results generally confirmed previous findings that beta-lactamase-related transposons Tn552 and Tn4002 preferentially insert within TIRL located upstream of the sin gene . Insertion of Tn552-like transposons adjacent to qac genes also occurs frequently, as observed in isolates Fol133, Fol89, Fol90, and Fol100 (this study). Overall, this suggests that genetic organizations containing IS257, Tn552, binR, and sin and/or qacR and qacA/B are widely distributed among staphylococci from clinical as well as food environments. Various rearrangements and DNA inversions generating two alternative staphylococcal beta-lactamase transposons have been reported . We also found the presence of additional organizations of Tn552 and qacA/B within the genomes of strains Fol33 and Fol89 . On pMS62, the resolvase- and recombinase-encoding genes (binR and sin, respectively) were present between qac and the incomplete Tn552-like bla gene module as in S. aureus plasmid pS1 . The truncated Tn552 observed in pMS62 has probably lost its ability to transpose, as the transposase gene (p480) was not detected. Staphylococcal insertion sequence IS257 plays a central role in cointegration events . Large, conjugative plasmid pSK41 has integrated several small plasmids, where all the cointegrated plasmids are all flanked by copies of IS257 . IS257 has been found closely associated with Tn552 , qacB (pSK156) , qacA (AJ400722), and qacC as well as antibiotic resistance determinants, e.g., dfrA, tetK, and aacA-aphD . The presence of IS257 downstream of blaZ on pMS33, pMS24, pMS62, and pMS89, and also detected on pMS90 and pMS100, indicates that IS257 integration events were involved in the formation of the structures of these plasmids. The presence of insertion sequences IS257 and IS256 on some of the plasmids harboring either qacA/B or beta-lactamase genes suggests that these plasmids may act as a source for formation of new multiresistance plasmids . Taken together the above observations indicate that a number of different recombination events have occurred to yield the various genetic organizations. To investigate the stability of the plasmids harboring disinfectant and antibiotic resistance genes, we tried to cure the isolates for plasmids. None of the strains were cured for the resistance plasmids. This indicated that plasmids are stably maintained in these strains. Backmatter: PMID- 12183250 TI - Comparative Study of Mechanisms of Herpes Simplex Virus Inactivation by Sodium Lauryl Sulfate and n-Lauroylsarcosine AB - The mechanisms of herpes simplex virus (HSV) inactivation by sodium lauryl sulfate (SLS) and n-lauroylsarcosine (LS), two anionic surfactants with protein denaturant potency, have been evaluated in cultured cells. Results showed that pretreatment of HSV type 1 (HSV-1) strain F and HSV-2 strain 333 with either surfactant inhibited, in a concentration- and time-dependent manner, their infectivities on Vero cells. SLS was a more potent inhibitor of HSV-2 strain 333 infectivity than LS with respect to the concentration (4.8-fold lower) and time (2.4-fold shorter) required to completely inactivate the virus. No inhibition of both herpesvirus strains infectivities was observed when Vero cells were pretreated with either surfactant. LS prevented the binding of HSV-2 strain 333 to cells without affecting the stable attachment and the rate of penetration into cells, whereas SLS exerted the opposite effect. Both SLS and LS inhibited, in a concentration-dependent manner, the HSV-2 strain 333-induced cytopathic effect, probably by affecting newly synthesized virions that come into contact with surfactant molecules present in culture medium. The pretreatment of HSV-2 strain 333 with specific combinations of SLS and LS concentrations inhibited the viral infectivity in a synergistic manner and resulted in only a small increase in their toxicities for exponentially growing Vero cells compared with that caused by each compound alone. Taken together, these results suggest that SLS and LS, alone or combined, could represent potent candidates as microbicides in topical vaginal formulations to prevent the transmission of herpes and possibly other pathogens that cause sexually transmitted diseases, including human immunodeficiency virus type 1. Keywords: Introduction : The number of individuals infected with human immunodeficiency virus type 1 (HIV-1), herpesviruses and other sexually transmitted pathogens is growing dramatically worldwide. The global incidence, morbidity, and mortality of sexually transmitted diseases (STDs) are very significant. Herpes simplex virus type 1 (HSV-1) and HSV-2 are the most common causes of genital ulceration in developed countries. In the United States only, approximately 500,000 new cases of herpes are reported each year. Genital herpes is lifelong and may result in painful and recurrent lesions, systemic complications, and psychosocial morbidity . HSVs are transmitted by symptomatic lesions and through asymptomatic viral shedding. The high viral titers in symptomatic lesions make this mode of transmission the most efficient. However, the contribution of asymptomatic viral shedding to the genital transmission of herpes in the population may account for as much as 50 to 90% of cases (-, ); the infected partner being unaware that he or she could transmit the virus. Moreover, intrapartum transmission of herpes from the mother to the neonate is also observed during vaginal delivery in 85% of cases and depends on the prevalence of viral shedding at the time of delivery. The transmission of the virus to neonates results in cutaneous, mucocutaneous, or eye infections or in some cases in the spreading of the virus to the brain or internal organs . Latex condoms represent an effective barrier against HIV and other sexually transmitted pathogens, but unfortunately, their use is not generalized. More attention is now given to female-controlled methods to prevent the sexual transmission of these microorganisms, since many women are unable to negotiate condom use with their partners . Nonoxynol-9, a nonionic surfactant with membrane-solubilizing properties, is the most currently used active ingredient in available vaginal formulations. In vitro, this compound inactivates enveloped viruses such as HIV-1 and HSV as well as other microorganisms, including Chlamydia trachomatis and Neisseria gonorrhoeae , but is not effective against nonenveloped papillomaviruses . The activity of nonoxynol-9 is nonspecific and is often associated with adverse effects, including epithelial disruption, genital inflammation, and ulceration, as well as a reduction in the number of lactobacilli , which may explain its lack of protection against HIV-1 and other pathogens causing STDs in clinical trials (, ; L. Van Damme, plenary lecture, XIII Int. AIDS Conf., 2000). There is thus an urgent need to develop new compounds which can be used as microbicides in topical vaginal formulations for women to protect themselves against sexually transmitted infections. Sodium lauryl sulfate (SLS), an anionic surfactant with protein denaturant potency, has been demonstrated to be a potent inhibitor of the infectivities of different strains of enveloped viruses such as HSV and HIV-1 as well as nonenveloped viruses such as papillomaviruses in vitro . We have also demonstrated that n-lauroylsarcosine (LS), another anionic surfactant structurally close to SLS, inactivates HSV infectivity in vitro . Moreover, gel formulations containing SLS or LS were very effective in preventing HSV-2 infection following intravaginal challenge in mice. These formulations were also well tolerated following repeated intravaginal administrations to rabbits. These data suggest that SLS and LS could represent potent candidates for use in topical vaginal formulations to prevent the transmission of HSV, HIV-1, and possibly other pathogens causing STDs. In the present study, we have compared the mechanisms of HSV-1 and HSV-2 inactivation by SLS and LS in cultured cells. We have also evaluated the inactivating potency of combinations of different concentrations of both compounds on herpesvirus infectivity as well as their potential toxicities to cells. MATERIALS AND METHODS : Materials. | SLS and LS were obtained from Sigma Chemical Co. (St. Louis, Mo.). [3H]methyl thymidine was purchased from Amersham Canada, Ltd. (Oakville, Ontario, Canada). Cell lines. | Vero cells (African green monkey kidney cells; American Type Culture Collection, Manassas, Va.) were cultivated in Eagle's minimum essential medium (EMEM) (Canadian Life Technologies, Burlington, Ontario, Canada) supplemented with sodium bicarbonate (0.22%), penicillin-streptomycin (100 U/ml), l-glutamine (2 mM), and 5% heat-inactivated fetal bovine serum (FBS) (Canadian Life Technologies) (EMEM-5% FBS). Cultures were maintained at 37C in a 5% CO2 atmosphere. Virus strains. | HSV-1 strain F (American Type Culture Collection) and HSV-2 strain 333 (kindly provided by Lawrence R. Stanberry, Children's Hospital Medical Center, Cincinnati, Ohio) were propagated in Vero cells in complete EMEM containing 2% FBS (EMEM-2% FBS). Preparation of radiolabeled virus. | HSV-2 strain 333 was labeled with [3H]methyl thymidine according to a previously described protocol with some modifications . In brief, Vero cells were incubated with the virus in EMEM-2% FBS at a multiplicity of infection of 0.1 for 1 h at 37C to allow virus adsorption. The virus was removed, and cell sheets were washed twice with culture medium. Cells were then incubated in EMEM-2% FBS containing [3H]methyl thymidine (25 muCi/ml) for 2 days at 37C. Cells and medium were collected, frozen at -80C, and thawed at 37C. The suspension was centrifuged (600 x g for 10 min at 4C) to pellet cell debris, and the supernatant (10 ml) was layered over a 3-ml cushion of 15% sucrose. Samples were centrifuged (100,000 x g for 2 h at 4C) to pellet the virus and resuspended in 1 ml of phosphate-buffered saline (PBS) (pH 7.4) overnight on ice at 4C. The specific activity of the virus was approximately 0.33 cpm/PFU. Virus inactivation assay. | In a first set of experiments, HSV-1 strain F or HSV-2 strain 333 was first preincubated in PBS (control) or in PBS containing increasing concentrations of SLS or LS for different periods of time (0 to 60 min) in a water bath at 37C. Confluent Vero cells seeded in 24-well plates were then infected with pretreated viruses (approximately 50 to 100 PFU). In a second set of experiments, the virus and SLS or LS were added simultaneously to cells (i.e., in the absence of a pretreatment period). In a third set of experiments, Vero cells were preincubated in PBS (control) or in PBS containing increasing concentrations of SLS or LS for 1 h at 37C prior to the infection with either HSV-1 strain F or HSV-2 strain 333. In all experiments, the plates were immediately centrifuged (750 x g for 45 min at 20C) after the infection of cells to allow virus adsorption. Unbound virus was removed by aspiration, and cell sheets were overlaid with EMEM-2% FBS containing 0.6% SeaPlaque agarose (Mandel Scientific, St-Laurent, Quebec, Canada). Cells were incubated for 2 days at 37C in a 5% CO2 atmosphere. Cells were then fixed with 10% formaldehyde in PBS for 20 min, washed with deionized water, and stained with 0.05% methylene blue. Virus inactivation was evaluated from the determination of the numbers of PFU. Virus inactivation assay in the presence of proteins. | In a first set of experiments, HSV-2 strain 333 was preincubated in EMEM-5% FBS (control) or in EMEM-5% FBS containing increasing concentrations of SLS or LS for 1 h in a water bath at 37C. In a second set of experiments, the virus was preincubated in EMEM containing increasing concentrations of FBS (0 to 60%; control) or with, in addition, 0.5 mM SLS or 1 mM LS for 1 h in a water bath at 37C. In both cases, the residual viral infectivity was determined on Vero cells as described above in the virus inactivation assay. Attachment assay. | The effect of SLS and LS on the attachment of radiolabeled HSV-2 strain 333 to Vero cells was assessed at 4C as previously described . Confluent monolayers of Vero cells seeded in 96-well plates were first incubated in PBS plus 1% bovine serum albumin (BSA) for 30 min at 4C in order to block nonspecific virus adsorption. [3H]methyl thymidine-labeled HSV-2 strain 333 was preincubated in PBS-1% BSA (control) or in PBS-1% BSA containing increasing concentrations of SLS or LS for 1 h in a water bath at 37C. The viral suspension was cooled in a melting ice bath for a few minutes. Cells were then incubated with pretreated virus for 1 h at 4C with gentle agitation (50 to 60 rpm). Cells were washed three times with ice-cold PBS and disrupted with lysis buffer (150 mM NaCl, 10 mM Tris [pH 7.6], 1% Igepal CA-630, 1% Na-deoxycholate). The amount of radioactivity associated with cell lysates was determined by liquid scintillation counting. Postbinding assay. | The effect of SLS and LS on HSV-2 strain 333 stably attached to Vero cells was evaluated as previously described . Confluent Vero cells seeded in 24-well plates were maintained on ice for a few minutes. Cells were first incubated with HSV-2 strain 333 (approximately 50 to 100 PFU) suspended in PBS for 2 h at 4C to allow a stable attachment of the virus without fusion with cell membrane. Unbound viruses were removed by aspiration. Cells were then washed with ice-cold PBS (control) or with ice-cold low-pH citrate buffer (Na citrate, 40 mM; KCl, 10 mM; NaCl, 135 mM, pH 3.0; positive control) or with ice-cold PBS containing increasing concentrations of SLS or LS for 1 min at 4C. Cell sheets were washed once with ice-cold PBS and overlaid with EMEM-2% FBS containing 0.6% SeaPlaque agarose. Cells were incubated for 2 days at 37C in a 5% CO2 atmosphere. Cells were then fixed, washed, and stained as described above. The amount of virus which had penetrated into cells after the shift at 37C was evaluated following the determination of the numbers of PFU. Penetration assay. | The effect of SLS and LS on the rate of penetration of HSV-2 strain 333 into Vero cells was examined as previously described . In brief, confluent monolayers of Vero cells seeded in six-well plates were first incubated with HSV-2 strain 333 for 2 h at 4C. After removal of unbound viruses, the plates were shifted to 37C to allow penetration of virus into cells. At selected times after the temperature shift (0, 15, 30, and 60 min), cells were treated for 1 min with PBS (control) or with PBS containing 300 muM SLS or 500 muM LS prewarmed at 37C. Cells were overlaid with EMEM-2% FBS containing 0.6% SeaPlaque agarose and incubated for 2 days at 37C in a 5% CO2 atmosphere. Cells were then fixed, washed, and stained as described above. The amount of virus which had penetrated into cells was evaluated following the determination of the numbers of PFU. Plaque reduction assay. | Confluent Vero cells seeded in 24-well plates were first incubated with HSV-2 strain 333 (approximately 50 to 100 PFU) suspended in EMEM-2% FBS for 2 h at 37C. Cells were then overlaid with EMEM-2% FBS containing 0.6% SeaPlaque agarose (control) or with, in addition, increasing concentrations of SLS or LS and incubated for 2 days at 37C in a 5% CO2 atmosphere. Cells were then fixed, washed, and stained as described above. Virus-induced cytopathic effect was evaluated following the determination of the numbers of PFU. Analysis of drug combination effect. | The inactivating potency of different combinations of SLS and LS on the infectivity of HSV-2 strain 333 was examined with combinations of various concentrations of the test compounds in a checkerboard design. In brief, HSV-2 strain 333 was first preincubated in PBS (control) or in PBS containing different combinations of SLS and LS concentrations for 1 h in a water bath at 37C. The residual viral infectivity was then determined on Vero cells as described above in the virus inactivation assay. The drug combination effect was analyzed by the isobologram method as described previously . In this analysis, the 50% effective dose (ED50) was used to calculate the fractional inhibitory concentration (FIC). When the FIC index of the combined compounds (i.e., FICx + FICy) is equal to 1.0, the combination is assumed to act in an additive manner; when it is between 0.5 and 1.0, the combination acts subsynergistically; and when it is less than 0.5, it acts synergistically. On the other hand, when the FIC index is between 1.0 and 2.0, the combination is subantagonistic, and when it is greater than 2.0, the combination is antagonistic. Cellular viability. | Vero cells seeded at midconfluency in 24-well plates were incubated in EMEM-5% FBS (control) or with EMEM-5% FBS containing different combinations of SLS and LS concentrations for 72 h at 37C in a 5% CO2 atmosphere. Cells were washed once with complete culture medium and incubated in EMEM-5% FBS containing a 100-mug/ml concentration of a tetrazolium salt [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl]-2H-tetrazolium, inner salt [MTS]; Promega, Madison, Wis.) and phenazine methosulfate (7.5 mug/ml) for 30 min at 37C. The tetrazolium salt is reduced by living cells to yield a formazan product that can be assayed colorimetrically . Reaction mixtures were transferred in a microplate, and the optical density was read at a wavelength of 492 nm. Statistical analysis. | The significance of the differences between (i) the inactivating potency of SLS and LS on herpesviruses' infectivities for susceptible cells and (ii) the toxic effect caused by combinations of SLS and LS concentrations and each compound alone on the viability of Vero cells was evaluated by an unpaired Student t test. All statistical analyses were performed with a computer package (Statview+SE Software; Abacus Concepts, Berkeley, Calif.). A P value of less than 0.05 was considered statistically significant. RESULTS : Viral inactivation. | Figure shows that pretreatment of HSV-1 strain F or HSV-2 strain 333 with SLS for 1 h at 37C decreased, in a concentration-dependent manner, their infectivities for Vero cells. The concentrations of SLS which inhibit 50% (ED50) and 90% (ED90) of the infectivity of HSV-1 strain F were 17.88 +- 5.86 muM and 34.89 +- 10.77 muM, respectively, whereas the corresponding values for HSV-2 strain 333 were 32.67 +- 2.18 muM and 46.53 +- 0.44 muM, respectively (unless otherwise noted, results are presented as means +- standard deviations). In the absence of a pretreatment period with SLS , the ED50 and ED90 values increased to 70.68 +- 5.90 muM and 96.04 +- 11.20 muM for HSV-1 strain F, respectively, and to 54.94 +- 6.65 muM and 82.29 +- 6.75 muM for HSV-2 strain 333, respectively. Pretreatment of Vero cells with SLS concentrations up to 300 muM for 1 h at 37C prior to viral infection did not cause any loss of infectivity for both strains of herpesviruses (data not shown). Similarly, Fig. shows that pretreatment of both strains of herpesviruses with LS reduced, in a concentration-dependent manner, their infectivities for Vero cells. The ED50 and ED90 values obtained following pretreatment of HSV-1 strain F with LS for 1 h at 37C were 194.40 +- 1.11 muM and 329.97 +- 16.54 muM, respectively, whereas the corresponding values for HSV-2 strain 333 were 141.76 +- 4.10 muM and 225.30 +- 12.79 muM, respectively. In the absence of a pretreatment period with LS or when Vero cells were pretreated with LS for 1 h at 37C prior to viral infection (data not shown), no significant inactivation of both herpesvirus strains infectivities could be observed at concentrations up to 500 muM. Time course effect. | Figure shows that the inactivating potencies of SLS and LS against HSV-2 strain 333 infectivity on Vero cells were directly related to the duration of the viral pretreatment period. At equipotent drug concentrations, the times of pretreatment required to inhibit 50 and 90% of viral infectivity were 3.27 +- 0.86 min and 23.10 +- 5.28 min when incubated with 50 muM SLS, respectively, whereas the corresponding values after an incubation of the virus with 250 muM LS were 7.32 +- 2.86 min and 55.41 +- 4.50 min (P < 0.01 compared with SLS), respectively. Effect on binding and postbinding steps. | Figure shows that pretreatment of [3H]methyl thymidine-labeled HSV-2 strain 333 with SLS concentrations up to 300 muM for 1 h at 37C did not reduce its binding to Vero cells when assessed at 4C. In contrast, pretreatment of the radiolabeled virus with LS for 1 h at 37C decreased, in a concentration-dependent manner, the initial attachment of the virus to cell surface heparan sulfate. The 50% reduction of viral binding was observed at an LS concentration of 139.08 +- 34.10 muM, which corresponds with the ED50 obtained for the inhibition of viral infectivity (i.e., 141.76 +- 4.10 muM). We also obtained a correlation coefficient between infectivity and binding data of 0.918 , which suggests that elimination of the viral attachment step is probably the principal mode of inhibition of herpesvirus infectivities by LS. Figure shows that SLS, added to cell cultures after the initial viral binding period at 4C, was able to inhibit the infectivity of HSV-2 strain 333 stably attached to Vero cells. The ED50 and ED90 values were 204.71 +- 25.98 muM and 297.16 +- 24.66 muM, respectively. However, the concentrations of SLS required to inhibit the virus infectivity under these conditions were 3.5-fold higher than that needed to inhibit its infectivity in the absence of a viral pretreatment period . In contrast, LS concentrations up to 300 muM had no effect on viruses stably attached to cells. Treatment of HSV-2 strain 333 stably attached to cells with a low-pH citrate buffer (positive control) for 1 min at 4C reduced to 1.87% +- 0.82% of control values the amount of virus which had penetrated into cells after the shift at 37C (data not shown). Effect on rate of penetration into cells. | Figure shows that SLS, at a concentration of 300 muM, decreases the rate of penetration of HSV-2 strain 333 in Vero cells. The inhibitory effect was more pronounced when the 1-min treatment period with SLS was made 15 or 30 min (approximately 90% inhibition) compared to 60 min (approximately 40% decrease) after the temperature shift at 37C. In contrast, LS, at a concentration of 500 muM, had no effect on the rate of penetration of the virus into cells at any time after the temperature shift at 37C. Effect on virus-induced cytopathic effect. | Figure shows that SLS and LS, added to Vero cells infected with HSV-2 strain 333 after a 2-h incubation period at 37C, reduced in a concentration-dependent manner the virus-induced cytopathic effect. The concentrations of SLS required to reduce the virus-induced cytopathic effect by 50 and 90% were 77.61 +- 4.86 muM and 109.28 +- 20.42 muM, respectively, whereas the corresponding values for LS were 81.36 +- 11.26 muM and 114.81 +- 9.88 muM, respectively. Inhibitory effect of combinations of surfactants. | Table shows the inhibitory effects of various combinations of SLS and LS concentrations on the infectivity of HSV-2 strain 333 for Vero cells following a 1-h viral pretreatment period at 37C. The inhibitory effects of combinations of both compounds on viral infectivity were analyzed by the isobologram method. Different effects were observed depending on the concentrations of surfactants mixed together. Combinations of SLS and LS acted synergistically to inactivate the virus infectivity when LS concentrations between 12.5 and 50 muM were added to increasing SLS concentrations. Combinations of LS and SLS acted synergistically when 12.5 muM SLS was added to increasing LS concentrations, whereas subsynergistic effects were obtained in the presence of 20 or 25 muM SLS. Cellular viability. | Table shows the effects of different concentrations of SLS or LS, alone or combined, on the viability of exponentially growing Vero cells following a 72-h incubation period. SLS and LS alone decreased, in a concentration-dependent manner, the viabilities of these cells. The 50% cytotoxic concentrations (CC50) of SLS and LS to this cell line were 202.04 +- 13.07 muM and 265.61 +- 10.74 muM, respectively. The CC50 of SLS and LS on resting Vero cell cultures incubated with either surfactant for 24 h were 266.30 +- 29.09 muM and 345.84 +- 11.79 muM, respectively (data not shown). Combinations of SLS and LS concentrations similar to those tested for their inhibitory effects on viral infectivity induced a significant decrease of the viability of exponentially growing cells compared with each compound alone. Inhibitory effect of SLS and LS in the presence of proteins. | Figure shows that the concentrations of SLS and LS required to inhibit the infectivity of HSV-2 strain 333 on Vero cells in complete culture medium (EMEM-5% FBS) were about 5.1 and 2.5 times higher, respectively, compared with viral pretreatment in PBS. The ED50 and ED90 values obtained following a viral pretreatment period of 1 h at 37C with SLS in EMEM-5% FBS were 185.41 +- 3.52 muM and 207.63 +- 6.01 muM, respectively, whereas the corresponding values for LS were 384.86 +- 20.71 muM and 490.64 +- 23.08 muM. Figure shows the effect of increasing FBS concentrations on the inhibitory potencies of SLS and LS on the infectivity of HSV-2 strain 333 on Vero cells. Pretreatment of HSV-2 strain 333 with 500 muM SLS for 1 h at 37C inhibited viral infectivity at FBS concentrations of 10% and lower. About 90% of the viral infectivity was inhibited when the virus was pretreated with 1 mM LS in culture medium containing 20% FBS. At higher FBS concentrations, the viral infectivity was not affected by either surfactant treatment, probably because SLS and LS were bound to serum proteins. FIG. 1. | Effect of SLS (A and B) and LS (C and D) on the infectivities of HSV-1 strain F and HSV-2 strain 333 (*) on Vero cells. Effect of SLS (A and B) and LS (C and D) on the infectivities of HSV-1 strain F and HSV-2 strain 333 (*) on Vero cells. In a first set of experiments, confluent cell monolayers were inoculated with viruses pretreated with PBS (control) or with PBS containing increasing concentrations of surfactants for 1 h at 37C (A and C). In a second set of experiments, surfactants and viruses were added simultaneously to cells without any pretreatment period (B and D). In both cases, the plates were immediately centrifuged (750 x g for 45 min at 20C) to allow virus adsorption. Unbound viruses were removed by aspiration. Cell sheets were overlaid with EMEM-2% FBS containing 0.6% SeaPlaque agarose and incubated for 48 h at 37C. Infectivity was expressed as the percentage of PFU compared with that of control cells. Results are the means +- standard deviations (error bars) of three independent experiments. FIG. 2. | Time course effect of SLS and LS (*) on the infectivity of HSV-2 strain 333 on Vero cells. Time course effect of SLS and LS (*) on the infectivity of HSV-2 strain 333 on Vero cells. Confluent cells were inoculated with viruses pretreated with PBS (control) or with PBS containing 50 muM SLS or 250 muM LS for various periods of time (0 to 60 min). The plates were then immediately centrifuged (750 x g for 45 min at 20C) to allow virus adsorption. Unbound viruses were removed by aspiration. Cell sheets were overlaid with EMEM-2% FBS containing 0.6% SeaPlaque agarose and incubated for 48 h at 37C. Infectivity was expressed as the percentage of PFU compared with that observed for control cells. Results are the means +- standard deviations (error bars) of three independent experiments. FIG. 3. | (A) Effect of SLS and LS (*) on the attachment of [3H]methyl thymidine-labeled HSV-2 strain 333 to Vero cells. (A) Effect of SLS and LS (*) on the attachment of [3H]methyl thymidine-labeled HSV-2 strain 333 to Vero cells. Viruses were first pretreated with PBS plus 1% BSA (control) or with PBS-1% BSA containing increasing concentrations of SLS or LS for 1 h at 37C. Confluent cells were then incubated with pretreated viruses for 1 h at 4C. Cells were washed three times with PBS and lysed, and radioactivity associated with cell lysates was determined by liquid scintillation counting. Binding was expressed as the percentage of counts per minute in the control cell lysates. Results are the means +- standard deviations (error bars) of three independent experiments. (B) Correlation calculated between inhibition of viral infectivity and inhibition of viral attachment data obtained following pretreatment of HSV-2 strain 333 with increasing LS concentrations for 1 h at 37C. FIG. 4. | Effect of SLS and LS (*) on HSV-2 strain 333 stably attached to Vero cells. Effect of SLS and LS (*) on HSV-2 strain 333 stably attached to Vero cells. Virus was first adsorbed to the confluent cell monolayer by a 2-h incubation period at 4C. After removal of unbound viruses, cells were treated with ice-cold PBS (control) or with ice-cold PBS containing increasing concentrations of SLS or LS for 1 min at 4C. Cells were washed once with ice-cold PBS, overlaid with EMEM-2% FBS containing 0.6% SeaPlaque agarose, and incubated for 48 h at 37C. Results were expressed as the percentage of PFU compared with that of control cells. Values are the means +- standard deviations (error bars) of three independent experiments. FIG. 5. | Effect of SLS and LS (*) on the rate of penetration of HSV-2 strain 333 in Vero cells. Effect of SLS and LS (*) on the rate of penetration of HSV-2 strain 333 in Vero cells. Virus was first adsorbed to the confluent cell monolayer after a 2-h incubation period at 4C. After removal of unbound viruses, cells were shifted to 37C to allow penetration into cells. At selected times after the temperature shift (0, 15, 30, and 60 min), cells were treated for 1 min with PBS (control ) or PBS containing 300 muM SLS or 500 muM LS prewarmed at 37C. Cell sheets were overlaid with EMEM-2% FBS containing 0.6% SeaPlaque agarose and incubated for 48 h at 37C. Results were expressed as the number of PFU. Values are the means +- standard deviations (error bars) of three independent experiments. FIG. 6. | Effect of SLS and LS (*) on HSV-2 strain 333-induced cytopathic effect on Vero cells. Effect of SLS and LS (*) on HSV-2 strain 333-induced cytopathic effect on Vero cells. Virus was first adsorbed to confluent cell monolayers by a 2-h incubation period at 37C. Unbound viruses were removed by aspiration. Cell sheets were then overlaid with EMEM-2% FBS containing 0.6% SeaPlaque agarose (control) or with, in addition, increasing concentrations of SLS or LS and incubated for 48 h at 37C. Results are expressed as the percentage of PFU compared with that of control cells. Values are the means +- standard deviations (error bars) of three independent experiments. FIG. 7. | Inhibitory effect of SLS and LS (*) on HSV-2 strain 333 infectivity in complete culture medium (EMEM-5% FBS). Inhibitory effect of SLS and LS (*) on HSV-2 strain 333 infectivity in complete culture medium (EMEM-5% FBS). Confluent Vero cells were inoculated with viruses pretreated with EMEM-5% FBS (control) or with EMEM-5% FBS containing increasing concentrations of SLS or LS for 1 h at 37C. The plates were then immediately centrifuged (750 x g for 45 min at 20C) to allow virus adsorption. Unbound viruses were removed by aspiration. Cell sheets were overlaid with EMEM-2% FBS containing 0.6% SeaPlaque agarose and incubated for 48 h at 37C. Infectivity was expressed as the percentage of PFU compared with that of control cells. Results are the means +- standard deviations (error bars) of three independent experiments. FIG. 8. | Inhibitory effect of SLS and LS (*) on HSV-2 strain 333 infectivity in the presence of increasing amounts of fetal bovine serum. Inhibitory effect of SLS and LS (*) on HSV-2 strain 333 infectivity in the presence of increasing amounts of fetal bovine serum. Confluent cells were inoculated with viruses pretreated with EMEM containing increasing amounts of FBS (0 to 60%; control) or with, in addition, 500 muM SLS or 1 mM LS for 1 h at 37C. The plates were immediately centrifuged (750 x g for 45 min at 20C) to allow virus adsorption. Unbound viruses were removed by aspiration. Cell sheets were overlaid with EMEM-2% FBS containing 0.6% SeaPlaque agarose and incubated for 48 h at 37C. Infectivity was expressed as the percentage of PFU compared with that of control cells. Results are the means +- standard deviations (error bars) of three independent experiments. TABLE 1 | Inhibitory effects of different combinations of SLS and LS concentrations on the infectivity of HSV-2 strain 333 on Vero cells after a 1-h pretreatment period at 37C TABLE 2 | Effects of SLS and LS, alone or combined, on the viability of exponentionally growing Vero cells after a 72-h incubation period DISCUSSION : The development of safe and effective topical vaginal microbicides is a high priority for major health organizations. SLS and LS, two anionic surfactants with protein denaturant potency, are potential candidates for use as microbicides in topical vaginal formulations (reviewed in reference ). In the present study, we have examined the mechanism(s) by which SLS and LS inactivate HSV infectivities in cultured cells. We have also evaluated the efficacies of several combinations of both surfactant concentrations to inhibit viral infectivity in relation to their effects on cell viability. As we have previously shown , SLS is a potent inhibitor of HSV-1 strain F and HSV-2 strain 333 infectivities for Vero cells. Howett et al. have also reported that pretreatment of HSV-2 strain 333 with SLS at a concentration of 867 muM for 10 min at 37C completely inactivated the infectivity of the virus for CV-1 cells. The difference between the doses of SLS required to completely inactivate the virus obtained by the two groups probably relates to the cell lines used (CV-1 versus Vero cells). Moreover, the use of complete culture medium containing 10% serum, compared to PBS in our studies, may decrease the surfactant availability due to the high affinity of SLS for the proteins present in the culture medium. In that respect, we showed that, under our experimental conditions, the use of complete culture medium (EMEM-5% FBS), rather than PBS, resulted in a fivefold-increased concentration of SLS required to inhibit viral infectivity. LS, which is structurally close to SLS, is also a potent inactivator of HSV-1 strain F and HSV-2 strain 333 infectivities for Vero cells. The inhibitory potencies of SLS and LS against both strains of herpesviruses were directly related to the duration of the viral pretreatment period. In addition, no inactivation was observed when Vero cells were pretreated with either surfactant rather than the virus, suggesting that these compounds act mainly on viral particles. Compared with SLS, the concentration of LS required to completely inactivate the infectivity of HSV-2 strain 333 following a 1-h viral pretreatment period at 37C was fivefold higher. In addition, at equipotent compound concentrations, LS completely inactivated viral infectivity following an eightfold-longer pretreatment period compared with SLS. The effects of molecules closely related to SLS or LS on the infectivities of herpesviruses have also been investigated. These data demonstrated that naturally occurring anionic surface-active agents present in bile salts, such as taurolithocholic acid-3-sulfate, alone or combined with glycocholic acid, were highly effective against HSV-1, HSV-2, HIV-1, C. trachomatis, and N. gonorrhoeae in vitro with little cytotoxicity . The different steps of HSV entry into cells are targets of choice for microbicides to prevent infection of susceptible cells. The attachment and penetration of HSV in target cells involve a three-step virus-cell interaction . These three stages correspond to (i) the initial attachment to cell surface heparan sulfate, which involves gC and gB (and possibly gD) and is resistant to PBS wash; (ii) the stable attachment to heparan sulfate or another unknown component which involves gD (and possibly gH) and is resistant to heparin wash; and (iii) the fusion followed by virus penetration which involves gH, gB, gL, and gK (and possibly others) and is resistant to low pH citrate buffer wash. Pretreatment of HSV-2 strain 333 with LS decreased the binding of [3H]methyl thymidine-labeled viral particles to Vero cells when assessed at 4C. The good correlation between inhibition of both viral infectivity and binding to cell surface suggests that LS probably acts by abrogating the attachment step of the virus to target cells. In contrast, the viral attachment was not affected by SLS treatment. Previous studies from our laboratory showed that SLS did not prevent the binding of [35S]methionine-labeled HSV-1 strain F to Vero cells . In contrast, SLS, but not LS, added to cell cultures after the initial binding period at 4C, inhibited the infectivity of HSV-2 strain 333 stably attached to cell surface. In addition, SLS decreased the rate of penetration of the virus into cells, probably by affecting the initial steps involved in the fusion process between the viral envelope and the cell surface. In contrast, the rate of penetration of the virus into cells was not affected following a treatment with LS. Both SLS and LS, when added after the infection of Vero cells, also inhibited the virus-induced cytopathic effect, probably by affecting newly synthesized virions that come into contact with the surfactant molecules present in culture medium. Based on the above demonstration, Fig. summarizes our proposed hypothesis for the mechanisms of HSV inactivation by SLS and LS at the different steps of viral entry and infection of target cells. The fact that the modes of herpesvirus inactivation by SLS and LS are different suggests that these surfactants affect different viral target proteins involved in the infectivity process. We have thus tested the efficacy of several combinations of SLS and LS concentrations to inhibit the infectivity of HSV-2 strain 333 on Vero cells following a 1-h viral pretreatment period at 37C. Interestingly enough, some of these combinations acted synergistically to inactivate the infectivity of the virus. The improved efficacy of these combinations was accompanied by only a small increase in their toxicities for exponentially growing Vero cells, especially when fixed amounts of SLS were added to various LS concentrations. For instance, the addition of 15 muM LS to increasing SLS concentrations reduced by 10.5-fold the dose effective at inhibiting the infectivity of HSV-2 strain 333, whereas it only increases by 1.1 times the cytotoxicity following an incubation period of 72 h, suggesting that some combinations of both surfactant concentrations could be highly effective without causing a marked toxicity for the skin or mucosal surfaces. Surfactants, like SLS and LS, first bind and then unfold and extract proteins before solubilizing membrane lipids. These effects greatly depend on the concentration of surfactants used and especially on their critical micellar concentrations. Kragh-Hansen et al. have reported that detergents with strongly hydrophilic heads like SLS only very slowly solubilize liposomal membranes . SLS specifically interacts with the protein component of Ca2+-ATPase membranes, leading to cooperative unfolding and extraction of the protein before solubilization of lipids. This effect results from combinations of hydrophobic and predominantly ionic interactions. LS, which is less hydrophilic than SLS, is thus a less effective solubilizing denaturant . Compared with the effect on proteins, the interaction of SLS with lipids is weaker and requires concentrations close to the critical micellar concentration. Using liposomes, it was shown that SLS extracts membrane lipids from the outer bilayer leaflet, which becomes partially depleted of phospholipids. Thereafter, a reorganization and a redistribution of phospholipid molecules from the inner to the outer leaflet occur. The vesicles slowly open up, leading to fragmentation and finally to solubilization into mixed micelles . We thus propose that at the concentrations used here, which are well below the critical micellar concentration, SLS and LS probably inactivate HSVs by affecting envelope or capsid proteins (reviewed in reference ). As suggested by the data presented in the present work, these proteins may play different roles in the viral replicative cycle . Therefore, the use of these surfactants could represent a convenient tool to study the role of their target proteins in the pathogenesis of viral infection in cultured cells and in animal models. We have demonstrated that SLS and LS are potent inhibitors of HIV-1 strain NL4-3 infectivity on 1G5 cells by eliminating the viral attachment and the fusion steps . In addition, SLS is a potent inactivator of the nonenveloped human papillomavirus infectivity and acts probably by denaturing capsid proteins . Of prime interest, the concentrations required to obtain a complete inactivation of HIV-1, HSV, and human papillomavirus infectivities are below toxic levels for cultured cell lines of different origins. Since the genital tract secretions are rich in proteins, we have evaluated the effect of increasing concentrations of FBS on the inhibitory potencies of SLS and LS against HSV-2 strain 333 infectivity for Vero cells. We showed that increasing the amount of proteins in the culture medium markedly decreased the inhibitory potencies of SLS and LS. However, by comparing the concentrations of surfactants required to completely inhibit the virus infectivity obtained at different concentrations of FBS, we estimated that approximately 50 muM SLS and 30 muM LS are neutralized by 1% FBS. Therefore, we may reasonably expect that under in vivo conditions, a gel formulation containing 2% SLS or LS (70 mM) could deliver sufficient concentrations of surfactants for inhibiting the infectivity of HSV and probably other pathogens. We have previously demonstrated that gel formulations containing SLS or LS are effective in preventing the colonization of mouse genital mucosa by infectious viruses after intravaginal challenge with HSV-2 strain 333 . In addition, repeated administrations of these formulations were well tolerated by the vaginal mucosa in the rabbit susceptibility model. This observation is supported by the fact that toothpastes, mouthwashes, shampoos, and liquid soaps containing SLS, which are sold over the counter and used on a chronic basis are nontoxic for skin and/or mucosal surfaces. On the other hand, microorganism resistance to these compounds is unlikely to happen, contrary to resistance to more-specific targets such as antibodies. We thus believe that SLS and LS could be potential candidates for use as microbicides in topical vaginal formulations to prevent the transmission of HSV, HIV-1, and possibly other pathogens causing STDs. These active ingredient are cheap, and if proven effective in clinical trials, they will allow the production of topical vaginal microbicides affordable to people living in developing countries where HIV-1 epidemic is a major problem. FIG. 9. | Model of SLS- and LS-induced inhibition of herpesvirus infectivity in Vero cells. Model of SLS- and LS-induced inhibition of herpesvirus infectivity in Vero cells. (A) Pretreatment of HSV-2 strain 333 with LS prevents the attachment of the virus to target cells at 4C, whereas SLS has no effect. This step involves the binding of the viral gC and gB (and possibly gD) glycoproteins to cell surface heparan sulfate. (B) Treatment of herpesviruses stably attached to Vero cells at 4C with SLS inhibits the infectivity of the virus, whereas LS exerts no effect. This step implicates the interaction between the viral gD (and possibly gH) glycoprotein to cell surface proteins. (C) SLS decreases the rate of penetration of HSV-2 strain 333 into cells, whereas LS does not interfere with the fusion process. This step involves several viral glycoproteins, such as gB, gH, gL, and gK (and possibly others). (D) Both SLS and LS inhibit the cytopathic effect induced by HSV-2 strain 333 in Vero cells. Both surfactants probably affect newly synthesized virions that come into contact with the compound present in culture medium. Backmatter: PMID- 12183238 TI - Samarangenin B from Limonium sinense Suppresses Herpes Simplex Virus Type 1 Replication in Vero Cells by Regulation of Viral Macromolecular Synthesis AB - Inhibitory effects of ethanolic extracts from 10 Chinese herbs on herpes simplex virus type 1 (HSV-1) replication were investigated. By a bioassay-guided fractionation procedure, samarangenin B (Sam B) was isolated from Limonium sinense; Sam B significantly suppressed HSV-1 multiplication in Vero cells without apparent cytotoxicity. Time-of-addition experiments suggested that the inhibitory action of Sam B on HSV-1 replication was not due to the blocking of virus adsorption. In an attempt to further localize the point in the HSV-1 replication cycle where arrest occurred, a set of key regulatory events leading to viral multiplication was examined, including viral immediate-early (alpha), early (beta), and late (gamma) gene expression and DNA replication. Results indicated that levels of glycoprotein B (gB), gC, gD, gG, and infected-cell protein 5 (ICP5) expression and gB mRNA expression in Vero cells were impeded by Sam B. Data from PCR showed that replication of HSV-1 DNA in Vero cells was arrested by Sam B. Furthermore, Sam B decreased DNA polymerase, ICP0, and ICP4 gene expression in Vero cells. Results of an electrophoretic mobility shift assay demonstrated that Sam B interrupted the formation of an alpha-trans-induction factor/C1/Oct-1/GARAT multiprotein complex. The mechanisms of antiviral action of Sam B seem to be mediated, at least in part, by inhibiting HSV-1 alpha gene expression, including expression of the ICP0 and ICP4 genes, by blocking beta transcripts such as DNA polymerase mRNA, and by arresting HSV-1 DNA synthesis and structural protein expression in Vero cells. These results show that Sam B is an antiviral agent against HSV-1 replication. Keywords: Introduction : Herpes simplex virus type 1 (HSV-1) is an enveloped DNA virus which causes a variety of infections in humans. After primary infection, HSV-1 establishes latency in sensory and autonomic neurons innervating the mucosal tissues, where primary infection takes place, and is reactivated by the proper stimulus to cause recurrence . The period of recurrence is irregular . Immunocompromised individuals and those with cancer are in danger of recurrent HSV-1 infections . The recipients of organ transplantation are at high risk for increased severity of HSV-1 infection . Infection with HSV-1 can lead to life-threatening encephalitis and ocular infections that result in corneal inflammation and scarification. This scarification is a major cause of blindness in developing countries . In addition, HSV-1 has been shown to be a factor for spreading human immunodeficiency virus and causes severe diseases in AIDS patients . One successful replication cycle of HSV-1 is dependent on the completion of a number of steps, including virion entry, subsequent expression of viral immediate-early (alpha) genes such as infected-cell protein 0 (ICP0) and ICP4 genes, early (beta1 and beta2) genes including DNA polymerase and thymidine kinase genes, and late (gamma1 and gamma2) genes encoding glycoprotein B (gB), ICP5, and gC, and unpaired DNA replication . The initial expression of HSV-1 alpha genes depends on the binding of the alpha trans-induction factor (alphaTIF)/C1/Oct-1 multiprotein complex to the TAAGARAT (R, purine; GARAT) sequences of the cis-acting site . Inhibition of any of these stages blocks HSV-1 replication. Nucleoside analogues have been extensively investigated in the search for effective antiherpesvirus agents . Among these acyclovir is widely used for the systemic treatment of HSV infections. It is a highly selective antiviral agent because it is specifically phosphorylated by viral thymidine kinase in infected cells . However, acyclovir-resistant HSV infection in immunocompromised patients such as transplant patients and patients with AIDS has recently been observed . Therefore, it is desirable to develop new anti-HSV agents that substitute for or complement acyclovir. Chinese herbs are potential sources of useful edible and medicinal plants. They are expected to find use as functional foods because of their various biological activities such as immunomodulatory and antitumor functions . More and more people in developing countries utilize traditional medicine for their major primary health care needs . However, ethnopharmacology also provides scientists with an alternative approach for the discovery of antiviral agents. The polysaccharides , anthraquinones , triterpenes , phloroglucinol , flavonoids , and catechin derivatives isolated from medicinal plants have been found to have inhibitory activities against the replication of HSV-1. There has been a promising result for a naturally occurring antiherpetic agent, n-docosanol, which has recently completed extensive clinical evaluation and been approved by the U.S. Food and Drug Administration as a topical treatment for herpes labialis . These findings show that natural products are still potential sources in the search for new antiherpetic agents. In the present study, 10 Chinese herbs which are widely known in folk medicine for the treatment of viral and bacterial infection were selected for an anti-HSV-1 replication assay. The herbs were Ventilago leiocarpa, Ecdysanthera rosea, Ecdysanthera utilis, Hippobroma longiflora, Ardrisia brevicaulis, Selaginella delicatula, Limonium sinense, Ardrisia japonica, Ardrisia violacea, and Andendron benthamianum. The ethanolic extracts that showed appreciable anti-HSV-1 activity were separated by a bioassay-guided fractionation procedure. The effect of active component samarangenin B (Sam B), isolated from L. sinense, on HSV-1 alpha, beta, and gamma gene expression and DNA replication in Vero cells was evaluated. The mechanisms of antiviral action of Sam B were elucidated in vitro. MATERIALS AND METHODS : Preparation of crude extracts for Chinese herbs. | All 10 species of Chinese herbs were purchased from Chinese medicine shops in Taipei, Taiwan, and were identified by Lie-Chwen Lin. Each dried Chinese herb (600 g) was extracted with ethanol (three times in 5 liters). After solvent was removed, the crude extracts were dissolved in dimethyl sulfoxide (DMSO) to a concentration of 100 mg/ml and stored at 4C until use. Sam B isolation from L. sinense. | The method followed procedures described previously . The ethanolic extract of the dried roots of L. sinense was partitioned in succession between H2O and n-hexane, followed by ethyl acetate and n-butanol (n-BuOH). The bioactive n-BuOH fraction was subjected to chromatography on Sephadex LH-20 with gradient methanol-H2O elution (from 0 to 100%) to produce 12 fractions. Sam B was identified from fraction 12 and analyzed with a high-performance liquid chromatography (HPLC) purity program. It was dissolved in DMSO to a concentration of 100 mM and stored at 4C until use. Cell culture and viruses. | Vero cells were cultured in minimal essential medium (MEM; GIBCO, Grand Island, N.Y.) supplemented with 10% fetal calf serum (HyClone, Logan, Utah), 100 U of penicillin/ml, and 100 mug of streptomycin/ml and incubated at 37C in a 5% CO2 incubator. The cells were free from mycoplasma contamination, which was checked by Mycotect kit (Life Technology, Gaithersburg, Md.). To prepare HSV-1 (KOS strain) stocks, Vero cells were infected by HSV-1 at a multiplicity of infection (MOI) of 3 and harvested at 24 h postinfection (p.i.) and centrifuged at 1,500 x g (CR312; Jouan, Nantes, France) at 4C for 20 min. The supernatant was collected and stored at -70C for use. Plaque reduction assay. | The assay followed procedures described previously . Acyclovir was used as a positive control. Vero cells (3.5 x 105/dish) were incubated with 100 PFU of HSV-1, and the test extracts or acyclovir was added to the cells at various concentrations. The viruses were adsorbed for 1 h at 37C, and 1% methylcellulose was added to each well. After 5 days, the virus plaques formed in Vero cells were counted by crystal violet staining. The activities of test extracts and acyclovir for inhibition of plaque formation were calculated. Determination of cell viability. | Approximately 3.5 x 105 Vero cells were cultured in a 25-cm2 flask and incubated with 0.1% DMSO or 25 muM Sam B for 5 days. Total, viable, and nonviable cells were counted three times under the microscope with the help of a hemocytometer following staining by trypan blue. The percentage of viable cells was calculated. The cell viability was also evaluated as lactate dehydrogenase (LDH) release according to the manufacturer's instructions (Roche, Milan, Italy). LDH activity in milliunits per milliliter, where 1 mU is the amount of enzyme required to transform 0.0167 nM NAD per min, was determined. Effect of Sam B on Vero cell growth. | The method was modified from that previously described . For the analysis of total cellular protein synthesis, Vero cells were preincubated with leucine-free MEM for 1 day and then cultured in MEM containing 2% fetal calf serum. The cells (5 x 103/well) were incubated with or without Sam B at various concentrations for 5 days. Then, for detection of DNA, RNA, and protein synthesis in Vero cells, tritiated thymidine, tritiated uridine, or tritiated leucine was added into each well (1 muCi/well; NEN), respectively. After a 16-h incubation, the cells were harvested on glass fiber filters by an automatic harvester (Multimash 2000; Dynatech, Billingshurst, United Kingdom). Radioactivity in the filters was measured by scintillation counting. Extraction of total cellular DNA and RNA. | Cellular DNA and RNA were extracted from Vero cells by the method described previously . Vero cells (5 x 106) were infected with HSV-1 at an MOI of 3 or were not infected in the presence or absence of Sam B (25 muM) and harvested at various times. For DNA extraction, the cells were lysed with 0.2 M Tris-HCl (pH 8.5) containing 100 mM EDTA, 100 mM NaCl, 0.5% NP-40, 1% sodium dodecyl sulfate (SDS), and 100 mug of proteinase K/ml. To extract cellular RNA, Vero cells were washed by cold Tris-saline (pH 7.4) and then suspended in NDD buffer (1% NP-40, 0.5% sodium deoxycholate, 1% dextran sulfate). The DNA and RNA solutions were extracted with phenol-chloroform and precipitated with 100% cold ethanol, and their concentrations were determined by measuring the optical density at 260 nm. Synthesis of first-strand cDNA. | The method followed procedures described previously . Aliquots of 1 mug of RNA were reverse transcribed with the Advantage RT-for-PCR kit from Clontech according to the manufacturer's instructions. The reaction mixture was initially incubated at 42C for 1 h and then at 94C for 5 min to terminate the reaction. Diethyl pyrocarbonate-treated H2O (80 mul) was added to the tube, and the tube was then stored at -20C for use in the PCR. PCR. | The method has been described elsewhere . Briefly, 10 mul of cDNA or total cellular DNA was mixed with 0.75 muM primers, 4 U of Taq polymerase, 10 mul of reaction buffer (2 mM Tris-HCl [pH 8.0], 0.01 mM EDTA, 0.1 mM dithiothreitol, 0.1% Triton X-100, 5% glycerol, 1.5 mM MgCl2), and 25 mul of water in a total volume of 50 mul. The oligonucleotide primer pairs were as follows: for gB (amplification fragment, 341 bp), 5'-CTGGTCAGCTTTCGGTACGA-3' and 3'-CGTTTGTCGACGTGCTGGAC-5'; for HSV-1 DNA polymerase (600 bp), 5'-CGCACCATCTACGACGGCCAGC-3' and 3'-CGCTTTCGTCTAGGCGAGCGCC-5'; for ICP0 (157 bp), 5'-TTCGGTCTCCGCCTGAGAGT-3' and 3'-AGCATACGCCGACCTCCCAG-5'; for ICP4 (670 bp), 5'-CCCGCCGATGCTGCCCTAAAC-3' and 3'-TTCGCCAGACACCTACTCAAG-5'; for beta-actin (1,300 bp), 5'-TTGAGACCTTCAACACCC-3' and 3'-CTCTACTGAAGCTTTTCGACT . The PCR was done at the following settings of the air thermocycler: denaturing temperature of 94C for 1 min, annealing temperature of 53C for 1 min, and elongation temperatures of 72C for 2 min for the first 25 cycles and then 72C for 10 min. The amplified products were quantitated with laser scanning densitometer SLR-2D/1D (Biomed Instruments Inc., Fullerton, Calif.). Northern blot analysis. | The analysis was modified from that previously described . The HSV-1-infected (MOI, 3) Vero cells (5 x 106) were cultured with or without Sam B (25 muM) for 6 h. A 15-mug sample of total cellular RNA was resolved on a 6.66% formaldehyde-agarose gel and then transferred to a nitrocellulose filter (Schleicher & Schuell) with 20x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate). The filter was prehybridized at 42C for 4 h (prehybridization solution of 0.1% sodium pyrophosphate, 0.25 mg of denatured salmon sperm DNA/ml, and 50% deionized formamide). A digoxigenin-dUTP (Boehringer GmbH, Mannheim, Germany)-labeled DNA polymerase probe prepared by PCR was added to the same solution and incubated at 42C for 12 h. The filter was washed with 0.1x SSC, detected by enzyme immunoassay using an antibody conjugated to alkaline phosphatase (Boehringer GmbH), and visualized by the chemiluminescence recorded on X-ray film (Kodak). Western blot analysis. | The experiment followed procedures described previously . The 5 x 106 Vero cells were infected with HSV-1 at an MOI of 3 or were not infected in the presence or absence of Sam B (25 muM) for 16 h. Extracted cellular proteins (20 mug) were dissolved in the dissociation buffer (2% SDS, 5% beta-mercaptoethanol, 0.05 M Tris-HCl, 20% glycerol, pH 7.6) and resolved by SDS-10% polyacrylamide gel electrophoresis (PAGE) then transferred to nitrocellulose filters. After filters were blocked, they were incubated with mouse monoclonal antibodies (Virusys Corporation, East Coast Biologics, Inc., North Berwick, Maine) raised against HSV-1 gamma proteins (gB, gC, gD, gG, or ICP5) and alpha proteins (ICP0 or ICP4). Specific reactive proteins were detected by an enhanced chemiluminescence method employing a rabbit anti-mouse immunoglobulin antibody linked to horseradish peroxidase (Santa Cruz Biotechnology, Inc., Santa Cruz, Calif.). Electrophoretic mobility shift assay (EMSA). | Oligonucleotide sequences 5'-GATCCCGTGCATGCTAATGATATTCTTTGGG-3' and 3'-GGCACGTACGATTACTATAAGAAACCCCTAG-5', corresponding to consensus alphaTIF/C1/Oct-1 site (nucleotides -170 to -143) of the ICP0 gene were used as a probe. Nuclear extracts of HSV-1-infected Vero cells were prepared 5 h after infecting cells at an MOI of 20. Reaction mixtures (15 mul) contained, together with 3 mug of nuclear extracts, 5 x 104 dpm of 32P-end-labeled probe, 2 mug of poly(dI-dC), 5% glycerol, 1 mM EDTA, 100 mM KCl, 5 mM MgCl2, 1 mM dithiothreitol, and 10 mM Tris-HCl, pH 7.4. After incubation for 30 min at room temperature, reaction mixtures were applied to 4% polyacrylamide gels in 0.5% Tris-borate-EDTA buffer. When the effect of Sam B on the binding of the alphaTIF/C1/OTF-1 multiple-protein complex to the probes was determined, nuclear extracts from HSV-1-infected Vero cells were treated with 25 muM Sam B for 5 min prior to the addition of the probes. Statistical analysis. | Data were presented as means +- standard deviations, and the differences between groups were assessed with Student's t test. RESULTS : Effects of ethanolic extracts from 10 Chinese herbs on HSV-1 replication. | As shown in Fig. , ethanolic extracts isolated from 10 Chinese herbs were evaluated for their activities in inhibiting HSV-1 plaque formation in Vero cells. HSV-1 replication was not affected by DMSO treatment. Acyclovir blocked HSV-1 plaque formation in Vero cells. While ethanolic extracts from Ventilago leiocarpa, Ecdysanthera rosea, Ecdysanthera utilis, Hippobroma longiflora, Ardrisia brevicaulis, Selaginella delicatula, Ardrisia japonica, Ardrisia violacea, and Andendron benthamianum had little effect on HSV-1 replication, 100 mug of L. sinense ethanolic extracts/ml significantly suppressed HSV-1 replication. The inhibitory action of L. sinense ethanolic extracts on HSV-1 replication was concentration dependent . The 50% inhibitory concentration (IC50) of L. sinense ethanolic extracts for HSV-1 replication was 25.0 +- 8.7 mug/ml. Sam B identified from L. sinense inhibits HSV-1 replication. | To extract pure active compounds from L. sinense, we used the complete isolation process in each chromatographic cycle and finally HPLC. The compound with the strongest activity was a brown amorphous powder. Nuclear magnetic resonance (NMR) and mass spectrum analyses indicated the structure shown in Fig. . The chemical name of this bioactive component is galloyl-epigallocatechin ([4beta->8, 2'->3"-O]-galloyl-epigallocatechin; C44H31O22; molecular weight, 911). Both mass and NMR spectrum data for the compound were compatible with those previously reported for Sam B by Nonaka et al. . The common name given to this component at that time was Sam B. The purity of Sam B was assessed by an HPLC purity program (column 5C18-MS). The column was eluted with a 35:65 (vol/vol) mixture of methanol and 0.1% H3PO4 and analyzed by a UV detector at 254 nm. Sam B appeared as a single peak at a 3.973-min retention time, and its purity was 99.0%. As shown in Fig. , Sam B blocked virus amplification, and thus its presence resulted in the formation of a cytopathic effect significant lower than that in control infections. The results of plaque reduction assay indicated that the inhibitory effect of Sam B on HSV-1 replication was concentration dependent. At 6.25 muM, Sam B suppressed HSV-1 replication by 24.6% +- 4.2%. The corresponding degrees of inhibition for 12.5, 25, 50, and 100 muM were 58.3 +- 3.0, 88.5 +- 7.3, 100 +- 5.2, and 100% +- 2.2%, respectively, with an IC50 value of 11.4 +- 0.9 muM . Effects of Sam B on viability and growth of Vero cells. | To delineate whether the suppressive effect of Sam B on HSV-1 replication was related to cytotoxicity, we examined the viability of Vero cells after treatment with Sam B for 5 days. DMSO did not affect cell viability. Compared with that of control groups, the viability of Vero cells treated with 25, 50, and 100 muM Sam B was not significantly decreased . Moreover, an evaluation of the cytotoxic effect of Sam B on Vero cells in terms of LDH release indicated that LDH release from Vero cells in the presence of 100 muM Sam B for 5 days was not significantly different from that in the absence of Sam B (3.23 +- 2.2 versus 4.05 +- 1.7 mU/ml). Results indicated that, even at 200 muM, Sam B had no direct cytotoxicity on Vero cells (data not shown). The therapeutic index (50% effective concentration/IC50) for Sam B is higher than 18. Additionally, effects of Sam B on Vero cell growth were determined by tritiated thymidine, tritiated uridine, and tritiated leucine uptake methods. As shown in Fig. , Sam B did not affect DNA, RNA, and protein synthesis in Vero cells. These results demonstrated that inhibitory mechanisms of Sam B on HSV-1 replication were not due to cytotoxicity and did not arrest Vero cell growth. Effect of Sam B on HSV-1 adsorption. | To further elucidate whether Sam B inhibition of HSV-1 replication was related to blocking viral adsorption, we examined the effect of its addition at various times. Cell supernatants were collected at 0, 2, 4, 6, 8, 16, 24, and 36 h p.i., and HSV-1 titers were determined by a plaque forming assay. As shown in Fig. , the virus titers in cell supernatants gradually increased at 8 h p.i. and the highest titer was obtained at 24 h p.i. DMSO did not affect HSV-1 amplification. By contrast, whether 25 muM Sam B was added at the same time as HSV-1 or after HSV-1 adsorption for 1 h, the virus titers in cell supernatants were decreased. The same results were obtained in acyclovir-treated cells. In addition, the pretreatment of Vero cells with 25 muM Sam B for 2.5 h before infection still inhibited virus yield. These results demonstrated that Sam B did not affect HSV-1 adsorption to host cells. Time course analysis of the effect of Sam B on HSV-1 replication. | Time course experiments were performed to determine at what point in the replication process Sam B inhibited HSV-1 replication. Sam B (25 muM) was added at the same time as HSV-1 (0 h p.i.) or at 2, 4, 6, 8, 12, 16, and 24 h p.i. The cell supernatants were collected at 25 h p.i., and a plaque forming assay was performed. The results indicated that addition of Sam B between 0 and 8 h suppressed HSV-1 replication . Addition of Sam B at 12 to 24 h p.i. had only minimal inhibitory effects on HSV-1 replication. The fact that Sam B was inhibitory when added between 0 and 8 hr p.i. suggested that the suppressive effects of Sam B might be related to the blocking of biochemical events or gene expression such as that of the alpha, beta, and gamma genes, which are necessary for HSV-1 replication during this time frame. Effects of Sam B on HSV-1 gamma gene expression in Vero cells. | After adsorption, HSV-1 expresses gamma genes about at 6 to 8 h p.i. or even earlier. The genes are needed for packaging viral particles . We analyzed whether Sam B suppression of HSV-1 replication was related to blocking the synthesis of gamma proteins in Vero cells. After cells were treated with 25 muM Sam B for 16 h, the total cellular proteins were extracted and the expression of viral gamma proteins including gB, gC, gD, gG, and ICP5 in Vero cells was determined by Western blot analysis. As shown in Fig. , while uninfected cells expressed few of these gamma proteins (lane 1), the gB (116 kDa), gC (78 kDa), gD (55 kDa), gG (45 kDa), and ICP5 (155 kDa) proteins were detectable in HSV-1-infected Vero cells (lane 2). However, Sam B suppressed the expression of these gamma proteins in Vero cells (lane 3). When Sam B was added at 12 h to 16 h p.i., the inhibitory action of Sam B was interrupted and the expression of gamma proteins in Vero cells was restored (lane 4). A graphical representation of the ratio of gamma proteins to beta-actin in Vero cells indicated that little signal was expressed in the cells treated with HSV-1 and Sam B at the same time. Because production of gB proteins in Vero cells was attenuated, we determined whether gB mRNA expression in the cells was affected by Sam B. At 16 h p.i., the total cellular RNA was collected from Vero cells in the presence or absence of Sam B (25 muM) and was available for reverse transcription-PCR (RT-PCR). Initially, we examined the dose-response relationship of the PCR amplification of cDNA (data not shown). The exponential phase of amplification was determined by performing 20, 25, and 30 cycles. We found that 25 cycles of PCR were optimal for the virion and beta-actin mRNAs. As shown in Fig. , neither DMSO (lanes 1 and 2) nor Sam B (lanes 3 and 4) affected the levels of beta-actin mRNA in Vero cells. While little expression of gB mRNA was detectable in uninfected Vero cells (lane 1), the levels of gB mRNA were significantly increased in HSV-1-infected cells (lane 2). By contrast, the level of RT-PCR products for gB mRNA amplified from HSV-1-infected Vero cells was decreased by Sam B (lane 3). Laser densitometry analysis demonstrated that the ratio of gB to beta-actin mRNAs was significantly decreased by Sam B. The inhibition of gB mRNA expression due to Sam B activity was blocked when Sam B was added in Vero cells between 12 and 16 h p.i. (lane 4). These results suggested that the decrease of gB proteins caused by Sam B was the result of a deficiency of gB mRNA expression in Vero cells. Effect of Sam B on HSV-1 DNA synthesis. | Because HSV-1 gamma2 genes such as the gC gene are dependent on viral DNA synthesis for expression and because DNA synthesis plays important roles in HSV-1 replication , we further defined whether Sam B has any effect on HSV-1 DNA synthesis in Vero cells. After HSV-1 adsorption, the cellular DNA was harvested at 2, 4, 8, and 16 h p.i. and viral DNA was analyzed by PCR. As shown in Fig. , although HSV-1 DNA could not be detected in uninfected Vero cells (lane 1), HSV-1 DNA synthesis was increased in HSV-1-infected cells (lanes 2 to 5). Graphical representation of the ratio of HSV-1 DNA to beta-actin DNA showed that increasing the signal in HSV-1-infected cells corresponded to increasing the p.i. time. By contrast, 25 muM Sam B suppressed viral DNA synthesis in Vero cells (Fig. , lanes 1 to 4). Additionally, HSV-1 DNA synthesis in Vero cells was restored when Sam B was added to cells at 12 h to 16 h p.i. (lane 5). These results suggested that the blocking of gC protein expression and HSV-1 replication by Sam B was related to impairment of HSV-1 DNA synthesis in Vero cells. DNA polymerase mRNA expression in Sam B-treated Vero cells detected by Northern blot analysis. | The appearance of HSV-1 beta gene products such as DNA polymerase signals the onset of viral DNA synthesis . To elucidate whether Sam B-suppressed HSV-1 DNA synthesis was related to DNA polymerase gene expression, the cells were infected with HSV-1 or were not infected in the presence or absence of Sam B (25 muM) for 6 h and then DNA polymerase mRNA was determined by Northern blot analysis. As shown in Fig. , 3.8 kb of DNA polymerase mRNA was expressed in HSV-1-infected Vero cells (lane 4) but was not expressed in uninfected cells (lane 2). DMSO did not affect DNA polymerase mRNA expression in Vero cells (lane 1). The level of DNA polymerase mRNA was impeded by Sam B treatment (lane 3). However, neither DMSO nor Sam B affected beta-actin mRNA expression in Vero cells. The ratio of the absorbance value for DNA polymerase mRNA to that for the beta-actin mRNA signal was calculated. Results indicated that Sam B significantly attenuated the ratio of HSV-1 DNA polymerase mRNA to beta-actin mRNA . The data demonstrated that Sam B blockage of HSV-1 DNA synthesis was related to a deficiency of DNA polymerase. Effects of Sam B on HSV-1 alpha gene expression in Vero cells. | As we know, beta genes of HSV-1 are not expressed in the absence of competent alpha proteins such as ICP0 and ICP4 . To confirm that Sam B inhibition of DNA polymerase mRNA was related to alpha gene expression, Vero cells were infected with HSV-1 or were not infected in the presence or absence of Sam B. The total cellular proteins was extracted at 4 h p.i., and ICP0 and ICP4 proteins were determined by Western blot analysis. As shown in Fig. , while uninfected cells expressed few virus ICP0 and ICP4 proteins (lane 1), both 110-kDa ICP0 and 175-kDa ICP4 were synthesized in HSV-1-infected Vero cells (lane 4). DMSO did not affect the expression of either protein in Vero cells (lane 2). However, Sam B decreased levels of ICP0 and ICP4 proteins in HSV-1-infected Vero cells (lane 3). Graphical representation of the ratio of ICP0 or ICP4 to beta-actin proteins in Vero cells indicated that little signal from infected Vero cells treated with Sam B was detectable. Furthermore, RT-PCR analysis was applied to determine whether decreasing ICP0 and ICP4 mRNA expression caused the impairment of protein production in Sam B-treated cells. As shown in Fig. , while ICP0 and ICP4 mRNAs were expressed in HSV-1-infected Vero cells (lanes 2 and 4), Sam B decreased the levels of ICP0 and ICP4 mRNAs in the cells (lane 3). A novel virus-induced protein-DNA complex reduced by Sam B. | The formation of alphaTIF/C1/Oct-1/GARAT multiprotein complexes is responsible for induction of alpha genes . To determine whether Sam B suppression of ICP0 and ICP4 transcripts was related to this initial transcriptional trans activation event, the radiolabeled DNA fragment containing GARAT was incubated with Sam B-treated nuclear extracts from HSV-1-infected cells. As shown in Fig. , with extracts from infected cells, a novel virus-induced protein-DNA complex (IEC) was detected (lanes 1 and 4). While DMSO did not influence protein-DNA complex formation (lane 1), a 50-fold excess of unlabeled probe blocked the formation of these complexes (lane 5). The level of IEC was reduced by Sam B treatment (lane 2). These results suggest that the reduction by Sam B of ICP0 and ICP4 mRNAs might be related to an interruption in the formation of alphaTIF/C1/Oct-1/GARAT multiprotein complexes. FIG. 1. | Effects of ethanolic extracts from 10 Chinese herbs on HSV-1 replication in Vero cells determined by plaque reduction assay. Effects of ethanolic extracts from 10 Chinese herbs on HSV-1 replication in Vero cells determined by plaque reduction assay. Vero cells (3.5 x 105/dish) were incubated with 100 PFU of HSV-1 and 10 muM acyclovir, 100 mug of each extract/ml (A), or the indicated concentrations of L. sinense extracts (B). After adsorption for 1 h, 1% methylcellulose was added to each well. On the fifth day p.i., the virus plaques formed in Vero cells were counted by crystal violet staining. Each bar represents the mean of three independent experiments. VL, Ventilago leiocarpa, ER, Ecdysanthera rosea, EU, Ecdysanthera utilis, HL, Hippobroma longiflora, AB, Ardrisia brevicaulis, SD, Selaginella delicatula, LS, L. sinense, AJ, Ardrisia japonica, AV, Ardrisia violacea, AE, Andendron benthamianum. FIG. 2. | The structure of Sam B purified from L. sinense The structure of Sam B purified from L. sinense. FIG. 3. | Cytopathic effect of HSV-1 on Vero cells treated with or without Sam B. Cytopathic effect of HSV-1 on Vero cells treated with or without Sam B. HSV-1-infected (MOI, 3) and uninfected Vero cells were treated with or without Sam B (25 muM). At 22 h p.i., cells were examined by microscopy (magnification, x180). Bar, 50 mum. FIG. 4. | Effects of Sam B on HSV-1 replication and Vero cell viability and growth. Effects of Sam B on HSV-1 replication and Vero cell viability and growth. (A) Inhibitory effects of 10 muM acyclovir and indicated concentration of Sam B on HSV-1 replication were determined by plaque reduction assay. Each bar represents the mean of three independent experiments. (B) Vero cells (3.5 x 105 in 25-cm2 flasks) were treated with medium, 0.1% DMSO, or 25, 50, or 100 muM Sam B for 5 days. Then total, viable, and nonviable cells were counted after staining with trypan blue. Each bar represents the mean of three independent experiments. (C) Vero cells (5 x 103/well of a 96-well plate) were treated with Sam B at the indicated concentrations or were treated without Sam B for 5 days. Then tritiated thymidine, tritiated uridine, or tritiated leucine was added to each well (1 muCi/well) to detect cell growth. After a 16-h incubation, the cells were harvested by an automatic harvester and then radioactivity was measured by scintillation counting. Each bar represents the mean of three independent experiments. FIG. 5. | Kinetics of inhibition of HSV-1 replication by Sam B. Kinetics of inhibition of HSV-1 replication by Sam B. (A) Vero cells (5 x 106) were infected with HSV-1 at an MOI of 3 in the presence or absence of acyclovir (10 muM) or Sam B (25 muM). Treatments were as follows: Sam B was added at the same time as HSV-1 , Sam B was present after the adsorption period , and Sam B was present for 2.5 h and removed before the adsorption phase by washing the cells with medium three times . Then cell supernatants were collected at 0, 2, 4, 6, 8, 16, 24, and 36 h p.i. and the viral titers were determined by a plaque forming assay. Each point represents the mean of three independent experiments. (B) Vero cells (5 x 106) were infected with HSV-1 (MOI, 3), and Sam B (25 muM) was added at the indicated times. Cell supernatants were collected at 25 h p.i., and HSV-1 titers were determined as described in Materials and Methods. Each bar represents the mean of three independent experiments. FIG. 6. | Effects of Sam B on HSV-1 gamma gene expression in Vero cells detected by Western blot analysis and RT-PCR. Effects of Sam B on HSV-1 gamma gene expression in Vero cells detected by Western blot analysis and RT-PCR. Vero cells (5 x 106) were infected with HSV-1 (MOI, 3) or were not infected in the presence or absence of Sam B (25 muM). (A) Lysates (20 mug of protein) were collected at 16 h p.i. and run on an SDS-10% PAGE gel and analyzed by immunoblotting with an anti-gB, -gC, -gD, -gG, or -ICP5 antibody. (B) The total cellular RNA was isolated from Vero cells at 16 h p.i. RT-PCR was done as described in Materials and Methods. Following the reaction, the amplified product was run on a 2% agarose gel. Lane 1, uninfected Vero cells; lane 2, HSV-1-infected cells; lane 3, cells treated with HSV-1 and Sam B at the same time; lane 4, infected cells treated with Sam B at 12 h to 16 h p.i. The graph indicates the ratio of gamma protein or gB mRNA to beta-actin mRNA. Each bar represents the mean of three independent experiments. FIG. 7. | Effect of Sam B on HSV-1 DNA synthesis in Vero cells detected by PCR. Effect of Sam B on HSV-1 DNA synthesis in Vero cells detected by PCR. Cells (5 x 106) were infected with HSV-1 (MOI, 3) or were not infected in the presence or absence of Sam B (25 muM). The cells were harvested at 2, 4, 8, and 16 h p.i., and total DNA was extracted with phenol-chloroform. The PCR was done as described in Materials and Methods. Following the reaction, the amplified product was run on a 2% agarose gel. (A) DNA was extracted from uninfected Vero cells at 16 h p.i. (lane 1) or from HSV-1-infected cells at 2, 4, 8, and 16 h p.i. (lanes 2 to 5). (B) Sam B and HSV-1 were added to the cells at the same time, and the total cellular DNA was extracted at 2 (lane1), 4 (lane 2), 8 (lane 3), and 16 h (lane 4) p.i. Lane 5, total cellular DNA extracted from infected cells treated with Sam B at 12 to 16 h p.i. Each band was quantitated by densitometer, and the ratio of HSV-1 DNA to beta-actin DNA was calculated. Each bar represents the mean of three independent experiments. FIG. 8. | Effects of Sam B on DNA polymerase gene transcription in Vero cells detected by Northern blot analysis. Effects of Sam B on DNA polymerase gene transcription in Vero cells detected by Northern blot analysis. Vero cells (5 x 106) were infected with HSV-1 (MOI, 3) or were not infected in the presence or absence of Sam B (25 muM). (A) Total cellular RNA was isolated from the cells at 6 h p.i. and analyzed in 6.66% formaldehyde-agarose gel and hybridized with digoxigenin-labeled DNA polymerase cDNA or beta-actin cDNA. Lanes 1 and 4, HSV-1-infected cells treated with or without DMSO; lane 2, HSV-1-infected cells treated with Sam B; lane 3, uninfected cells. (B) Each exposed band was quantitated by densitometer, and the ratio of DNA polymerase mRNA to beta-actin mRNA was calculated. Each bar is the mean of three independent experiments. FIG. 9. | Effects of Sam B on HSV-1 ICP0 and ICP4 gene expression and formation of IEC in Vero cells detected by Western blotting, RT-PCR, and EMSA, respectively. Effects of Sam B on HSV-1 ICP0 and ICP4 gene expression and formation of IEC in Vero cells detected by Western blotting, RT-PCR, and EMSA, respectively. Vero cells (5 x 106) were infected with HSV-1 (MOI, 3) or were not infected in the presence or absence of Sam B (25 muM). (A) Lysates (20 mug of protein) were collected at 4 h p.i. and run on an SDS-10% PAGE gel and analyzed by immunoblotting with an anti-ICP0 or -ICP4 antibody. (C) Total cellular RNA was isolated from Vero cells at 4 h p.i. and analyzed by RT-PCR. Lane 1, uninfected Vero cells; lanes 2 and 4, HSV-1-infected cells treated with or without DMSO; lane 3, infected cells treated with Sam B. (B and D) Bar graphs indicating the ratio of ICP0 or ICP4 to beta-actin proteins or mRNAs. Each bar represents the mean of three independent experiments. (E) EMSA was performed as described in Materials and Methods. Nuclear extracts from HSV-1-infected (lanes 1 and 4) or uninfected Vero cells (lane 3) were incubated with a 32P-end-labeled GARAT probe. The effects of Sam B on the formation of IEC in virus-infected nuclear extracts pretreated with 25 muM Sam B for 5 min and then mixed with the probes (lane 2) were detected. Lane 5, results of adding a 50-fold excess of unlabeled probe to the reaction mixture. DISCUSSION : The plaque reduction assay offers a popular system to evaluate the effect of antiviral agents against HSV-1 . In our study, 10 Chinese herbs were screened by this model and L. sinense was found to contain antiherpetic agents, supporting the validity of its use for pharmacological studies. Results shown here indicated that Sam B purified from L. sinense suppressed HSV-1 multiplication in Vero cells without significantly reducing cell viability and growth. The inhibitory effect of Sam B may be attributed to its interference with structural proteins and DNA synthesis, DNA polymerase mRNA transcription, and alpha gene expression of HSV-1. Hence, Sam B suppression of viral replication might have important implications for L. sinense therapeutic activity in microorganism infection. This is the first report of the antiviral action mechanisms of Sam B. Sam B isolated from L. sinense is a polyphenol flavonoid compound . Results showing that Sam B impaired HSV-1 replication in Vero cells were compatible with data reported by Vanden Berghe et al., which indicate that many polyphenols are known for their antiherpesvirus activities . Sam B blockage of HSV-1 replication was probably not related to DMSO because cell viability and growth and HSV-1 replication in Vero cells were not changed by DMSO. The morphology and characteristics of Vero cells treated with Sam B or other Chinese herb extracts were similar, suggesting that inhibitory effects of Sam B were not related to the pH, osmolarity, or other physiology variables in different preparations (data not shown). Although we did not determine whether Sam B bound to virus, results of electron microscopy observations indicated that the morphology of HSV-1 in the presence of Sam B is unchanged from that in its absence (data not shown). A comparison with control groups showed that the HSV-1 titer was not significantly decreased when viral particles were treated with 25 muM Sam B at 37C for 1 h (data not shown). Moreover, pretreatment of cells with Sam B or addition of the drug after viral adsorption produced antiviral activity similar to that when HSV-1 and Sam B were added at the same time. These results suggest that the binding of Sam B to virion or host cells could not be a factor inhibiting virus replication. The preliminary data indicated that pretreatment of Vero cells with Sam B for 24 h and then its removal before infection still inhibit virus yield (data not shown). It is possible that the binding of Sam B to certain membrane molecules of the host cell different from the receptor, resulting in interference with virus penetration into the host cells, such as by interfering with virus-cell fusion, was not related to the inhibitory action of Sam B. In the host cells, HSV-1 replication is coordinately regulated and sequentially ordered in a cascade and is believed to proceed as follows: (i) alphaTIF, a gamma protein packaged in the virion, turns on the alpha genes to be transcribed; (ii) expression of alpha genes regulates the beta genes to be expressed; (iii) both alpha and beta gene expression initiates HSV-1 DNA replication; (iv) gamma genes are synthesized, and then virions are assembled; and (v) HSV-1 is enveloped as it buds through the nuclear membrane . In the present study, we found that Sam B decreased ICP0 and ICP4 gene expression in Vero cells. We are currently attempting to further elucidate whether the mechanisms by which Sam B suppresses ICP0 and ICP4 gene expression are related to the initial transcriptional transactivation events, including formation of alphaTIF/C1/Oct-1/GARAT multiprotein complexes. Our initial experiments using EMSA indicated that retarded species IEC has been decreased in the GARAT probes incubated with Sam B-treated nuclear extracts infected by HSV-1. Thus, the possibility that Sam B inhibited HSV-1 replication through disturbance of alphaTIF/C1/Oct-1/GARAT stable complex formation cannot be excluded. ICP0 has been reported to perform several functions including selection of transcriptional termination sites and stimulation of DNA synthesis. Although ICP0 is not essential for HSV-1 replication in some cell cultures, defects in the associated gene delay the expression of beta and gamma genes and impair viral replication . ICP4 is the major transactivator of HSV-1 genes. Thus, ICP4 and ICP0 play important roles in regulation of beta and gamma gene expression and are essential for HSV-1 replication . Moreover, the present data indicated that Sam B impaired DNA polymerase transcripts and HSV-1 DNA synthesis in Vero cells. It is known that a large array of proteins including DNA polymerase are required for HSV-1 DNA synthesis. We suggest that the decrease in HSV-1 DNA synthesis due to Sam B is related to impairment of DNA polymerase. Recently, we detected thymidine kinase mRNA expression in Vero cells by RT-PCR, and preliminary results indicate that Sam B impeded thymidine kinase mRNA expression (data not shown). We also proved that the production of gB, gC, gD, gG, and ICP5 proteins in Vero cells was attenuated by Sam B. ICP5 is a major capsid protein and is made both early and late in infection. gB, gC, gD, and gG are all involved in the viral envelope structure and play important roles in viral attachment and penetration. The evidence demonstrates that ICP4 is required for expression of gamma genes and that ICP4 binding sites enhance the transcription of the gD gene in vitro . Although effects of Sam B on gC, gD, gG, and ICP5 mRNA expression were not determined, it was found that Sam B decreased gB mRNA expression in Vero cells. HSV-1 DNA synthesis is required for gC mRNA expression. gD, gG, and ICP5 are similar to gB in that their mRNA expression requires ICP4 proteins . Thus, we predict that Sam B attenuates levels of gB, gC, gD, gG, and ICP5 proteins in Vero cells, which may be related to impairment of DNA synthesis and ICP4 and ICP0 production. The attenuation of DNA synthesis and ICP4 and ICP0 production may cause decreases of gB, gC, gD, gG, and ICP5 gene expression in Vero cells. From the present results, we hypothesize that impairment of HSV-1 multiplication in Sam B-treated Vero cells, at least in part, was related to (i) decreases in HSV-1 ICP0 and ICP4 gene expression due to Sam B, which might be related to disturbance of the formation of alphaTIF/C1/Oct-1/GARAT multiprotein complexes, (ii) reduction in DNA polymerase transcriptions in the cells, (iii) inhibition of viral DNA synthesis, (iv) interference with gB protein synthesis due to blockage of gB mRNA synthesis, (v) impaired levels of viral capsid protein ICP5 and envelope proteins gC, gD, and gG, and (vi) no HSV-1 plaque formation in Vero cells. Unlike dextran sulfate isolated from Sargassum horneri, which inhibits HSV-1 replication at the adsorption step , Sam B blocked HSV-1 replication at the immediate-early and early steps. The expression of the immediate-early gene represents one stage of the HSV-1 replication cycle that could be targeted by a novel antiviral therapy to deliver a significant reduction in virus replication in both acute and latent infections. The expression of the immediate-early gene plays important roles in the regulation of all classes of viral genes during lytic infection and is the key initiating event in the process of reactivation of the latent HSV-1 genomes. Low levels of immediate-early transcripts can be identified in latently infected neuron cells. ICP0 mutant viruses that are defective in immediate-early transactivation have been shown to reactivate very poorly from latent infections . While nucleoside analogues have been successful in treating acute infections, they fail to modulate reactivation of latent virus. On the other hand, Sam B lacked the elevated cytotoxicity and antiproliferative properties of interferons . Because Vero cells are notoriously hardy and aneuploid, the effects of Sam B on the growth of human peripheral blood mononuclear cells (PBMC), which are normal diploid cells, were determined. Results indicated that Sam B did not affect DNA, RNA, and protein synthesis in PBMC (data not shown) and that the therapeutic index was about 13.2. Thus, small molecules identified from Chinese herbs such as Sam B which act as inhibitors of HSV-1 immediate-early gene expression may have the potential to impact clinical disease to a far greater extent than currently marketed nucleosides and cytokines. Future experiments with treatment of HSV-1-infected animals with Sam B will be necessary to define whether L. sinense can reduce experimental viral infection injury and prevent recurrent HSV-1 infection. Moreover, this study not only demonstrates that Chinese herbs are potential therapeutic drugs for the viral infection but also supports a model for future protocol design in preclinical studies. Backmatter: PMID- 12183278 TI - Comparative Evaluation of Disk Diffusion with Microdilution Assay in Susceptibility Testing of Caspofungin against Aspergillus and Fusarium Isolates AB - We compared the disk diffusion and broth microdilution methods for susceptibility testing of caspofungin against Aspergillus (n = 78) and Fusarium (n = 22) isolates. Microdilution testing followed the NCCLS M-38P guidelines but was performed in antibiotic medium 3 supplemented to 2% glucose (AM3). Disk diffusion assays were performed on AM3 agar plates with a 2-mug caspofungin disk. By both methods, caspofungin showed favorable activity against Aspergillus isolates and no activity against Fusarium isolates. In the disk-based format, intrazonal growth that was not influenced by the drug concentration gradient was consistently observed for all of the Aspergillus isolates tested. Keywords: Introduction : The development of novel, water-soluble echinocandins is an exciting development in systemic antifungal therapy. These agents exert antifungal activity via inhibition of (1,3)-beta-D-glucan synthesis. Because of the lack of this target in mammalian tissue and the dissimilarity between the mechanism of action of echinocandins and other systemic antifungal drugs (particularly polyenes and azoles), echinocandins appear to be advantageous and promising for the treatment of invasive mycoses . Caspofungin (Cancidas; Merck Research Laboratories) is a novel echinocandin that was licensed in January 2001 in the United States. While caspofungin has proven to be efficacious in the treatment of candidiasis, aspergillosis, and possibly other mycoses (; J. Hiemenz, I. Raad, M. Boogaerts, J. Maertens, A. Saah, C. A. Sable, J. A. Chodakewitz, M. Severino, P. Saddier, R. Berman, M. J. DiNubile, T. F. Patterson, D. W. Denning, and T. J. Walsh, Focus on Fungal Infections 11, abstr. 21, 2001; J. Maertens, I. Raad, C. A. Sable, A. Ngai, R. Berman, T. F. Patterson, D. Denning, and T. Walsh, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 1103, 2000; M. A. Powles, J. Anderson, P. Liberator, and D. M. Schmatz, Abstr. 36th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-42, 1996; C. A. Sable, A. Villanueva, E. Arathon, E. Gotuzzo, G. Turcato, D. Uip, L. Nriega, C. Rivera, E. Rojas, V. Taylor, R. Berman, G. B. Calandra, and J. Chodakewitz, Abstr. 37th Intersci. Conf. Antimicrob. Agents Chemother., abstr. LB-33, 1997), in vitro susceptibility testing method and interpretive test parameters have not been fully established for this novel drug. In addition, there are few data correlating its in vitro activity and the in vivo response to it. Previous studies have focused on broth dilution methods for susceptibility testing of caspofungin against yeasts and molds . When testing caspofungin and other echinocandins against molds, a distinctive microscopic interpretive parameter called the minimum effective concentration (MEC) has been found to be more consistent and also appears to be better correlated with clinical outcome than the conventional macroscopic MIC-2 (, , , ; C. M. Douglas, J. C. Bowman, G. K. Abruzzo, A. M. Flattery, C. J. Gill, L. Kong, C. Leighton, J. G. Smith, V. B. Pikounis, K. Bartizal, M. B. Kurtz, and H. Rosen, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 1683, 2000). Agar-based testing methods are the other obvious avenue of investigation. Some investigators have explored the use of disk diffusion susceptibility testing for caspofungin against yeasts (specifically, Candida isolates) , but data on caspofungin testing against molds by the disk diffusion method are lacking. In this study, we investigated the applicability of a disk diffusion assay in susceptibility testing of caspofungin against Aspergillus and Fusarium isolates and compared the results of this disk diffusion assay with those of the broth microdilution method for these isolates. (This work was presented at the 41st Interscience Conference on Antimicrobial Agents and Chemotherapy [abstr. J-571].) Aspergillus strains (n = 78; 27 Aspergillus flavus, 26 Aspergillus fumigatus, 16 Aspergillus niger, and 9 Aspergillus terreus) and Fusarium strains (n = 22; 18 Fusarium solani and 4 Fusarium oxysporum) isolated from clinical samples were included in this study. One of the Aspergillus fumigatus isolates (strain 2-160) was included in each run of susceptibility tests as a reference strain to validate the reproducibility and quality of the test results. The isolates were defined to the species level by standard methods and stored on Sabouraud dextrose agar slants at -70C until tested. Standard caspofungin powder was provided by Merck Research Laboratories for use in susceptibility tests. Caspofungin susceptibility tests were done by using the broth microdilution and disk diffusion methods. Except as noted otherwise, broth microdilution tests were done in accordance with the NCCLS guidelines for conidium-forming filamentous fungi . Antibiotic medium 3 (lot JD4ZSG; BBL, Becton Dickinson) buffered by addition of 1 g of Na2HPO4 and 1g of NaH2PO4 to each liter and supplemented to 2% glucose (AM3) was used as the test medium. Serial twofold dilutions of caspofungin over a range of 16 to 0.03 mug/ml were prepared in microdilution plates. The results were read after 24, 48, and 72 h of incubation by using two different parameters: the visual MIC-2 (the minimum concentration [micrograms per milliliter] of caspofungin that provides an similar50% reduction in growth compared to the growth in the control well) and the microscopic MEC (the minimum concentration [micrograms per milliliter] of caspofungin that results in the formation of abnormal hyphal growth with short, abundant branchings ). The broth microdilution assay results for the isolates included in this study were published in one of our previous reports . Disk diffusion tests were done by using empirically defined parameters. Caspofungin disks were prepared in house. Blank paper disks (6.3 mm in diameter; Becton Dickinson Microbiology Systems, Cockeysville, Md.) were impregnated with 20 mul of a caspofungin suspension (concentration, 100 mug/ml), resulting in a final concentration of 2 mug/disk. The disks were allowed to dry at room temperature. (This disk concentration was chosen on the basis of the results of preliminary experiments in which blank disks were impregnated with 20 mul from suspensions containing caspofungin at concentrations of 200, 100, 50, and 25 mug/ml. The concentration of 100 mug/ml was chosen because it yielded inhibition zones [IZs] that are wide enough to be measured and sufficiently narrow to be accurately determined.) The strain to be tested was initially suspended in saline, adjusted spectrophotometrically to similar81 or similar70% transmittance (for Aspergillus and Fusarium, respectively), and then diluted 1/100 in distilled water to achieve a final concentration of similar104 CFU/ml. The prepared inoculum was swabbed onto an AM3 agar plate, and the plate was left to dry at room temperature for 20 min. A caspofungin disk was then placed onto the center of the inoculated agar plate. The plates were incubated for 72 h at 35C, and the IZ diameters (millimeters) were measured at 24, 48, and 72 h of incubation. The edges of the IZs were taken as the points of a marked decrease in fungal density. For comparative evaluation of the broth microdilution and disk diffusion methods, the geometric mean (GM) and range of the MICs and MECs and the arithmetic mean and range of the IZ diameters were calculated for each genus-species combination. For computation of GM values, high off-scale MICs and MECs were converted to the twofold concentration just above the highest drug concentration tested. The results obtained by the microdilution and disk diffusion methods at 24, 48, and 72 h of incubation of the test isolates and reference strain 2-160 are shown in Tables and , respectively. While broth microdilution test results could be determined at all of the reading time points (24, 48, and 72 h), disk diffusion test results could not be interpreted at 24 h because of poor growth. At 48 and 72 h, IZs with sharply defined edges were observed on agar plates. TABLE 1 | Microdilution and disk diffusion test results at 24, 48, and 72 h of incubation TABLE 2 | Microdilution and disk diffusion test results obtained for reference strain 2-160 All of the Aspergillus strains generated measurable IZs, and the diameters of these IZs were distributed over a relatively narrow range. Interestingly and in addition, Aspergillus microcolonies were visualized inside the IZs and these colonies were found uniformly distributed right up to and even under the disks without any evident effect of the drug concentration gradient . These colonies were observed for all of the Aspergillus species and strains tested. It is noteworthy that the number of intrazonal colonies observed for A. niger was lower than the number observed for other Aspergillus species. When examined under a microscope, these intrazonal colonies produced short, stubby hyphal branchings and a star-like morphology. This appearance was similar to that observed at the MEC. The colonies outside the zone, on the other hand, showed a normal, elongated, branching hyphal morphology. When the intrazonal colonies were subcultured and retested, they yielded a pattern identical to that seen with the original isolate. On the other hand, confluent growth and absence of an IZ were consistently observed for all of the Fusarium isolates. Comparison of the results obtained by the broth microdilution and disk diffusion assays showed that, as for the Fusarium isolates, the high MECs obtained by the microdilution method correlated very well with absence of IZs on disk diffusion agar plates. Relatedly and as for Aspergillus isolates, relatively lower MECs correlated with the production of measurable IZs around caspofungin disks. FIG. 1. | Aspergillus growth pattern obtained by the disk diffusion test method. Aspergillus growth pattern obtained by the disk diffusion test method. Note the clear zone edges and the microcolonies inside the IZ. The disk contained 2 mug of caspofungin. In the present study, our major goal was to investigate the relevance of the disk diffusion assay for caspofungin susceptibility testing against Aspergillus and Fusarium strains. We thus compared the disk diffusion assay results with our MECs previously determined by the broth microdilution method. Since MICs tended to increase inconsistently, particularly with extended incubation , we focused more on the correlation of MECs with IZ diameters. The comparative evaluation of the two methods showed that while lower MECs corresponded to the generation of IZs, higher MECs were in absolute correlation with the absence of IZs. Relatively lower MECs and measurable IZs were obtained with Aspergillus isolates, while Fusarium spp. consistently generated very high MECs and no IZs. Being less time-consuming and less labor-intensive, the disk diffusion method is preferable to the microdilution method. However, the inability to determine the susceptibility test result at 24 h for an individual isolate appears to be a notable limitation of the disk diffusion assay. Further investigation is required to determine whether higher inoculum concentrations yield satisfactory growth on disk diffusion agar plates at 24 h. Our observation of the growth of Aspergillus microcolonies inside the IZs was noteworthy. Since the intrazonal colonies produced the same growth pattern when retested, the possibility of heterogeneous resistance was ruled out. Thus, this finding appeared to be a special growth pattern. This might originate from the partial inhibitory nature of caspofungin and other echinocandins and correspond to the lack of complete inhibition of growth observed both in microdilution plate wells with high caspofungin concentrations and by examination of the intrazonal colonies under a microscope. The mechanism, meaning, and clinical significance of this observation remain unclear and merit investigation. In this study, we used the disk diffusion assay for susceptibility testing of caspofungin against Aspergillus and Fusarium isolates and compared the results with those obtained by the microdilution method. A preliminary report has been previously published on the use of the disk diffusion assay for candin derivatives L-733560, L-705589, and L-731373 against Aspergillus spp. In that study, potato dextrose agar was seeded with 106 CFU of A. fumigatus spores and poured into petri dishes. Disks impregnated with each of the candins (at concentrations of 128 to 0.06 mug/ml) were placed on the agar plates. IZs were observed after 24 h of incubation, demonstrating the favorable in vitro activity of the compounds against Aspergillus strains. No observation was noted regarding intrazonal growth . We conclude that the disk diffusion and microdilution methods appear to be correlated in the susceptibility testing of caspofungin against Aspergillus and Fusarium isolates. The intrazonal growth of colonies of Aspergillus spp. is of uncertain relevance. Backmatter: PMID- 12183255 TI - Phase I Dose Escalation Trial Evaluating the Pharmacokinetics, Anti-Human Cytomegalovirus (HCMV) Activity, and Safety of 1263W94 in Human Immunodeficiency Virus-Infected Men with Asymptomatic HCMV Shedding AB - 1263W94 [maribavir; 5,6-dichloro-2-(isopropylamino)-1,beta-l-ribofuranosyl-1-H-benzimidazole] is a novel benzimidazole compound for treatment of human cytomegalovirus (HCMV) infection and disease, with potent in vitro activity against HCMV and good oral bioavailability. A phase I study was conducted to determine the pharmacokinetics (PK), anti-HCMV activity, and safety of 1263W94 administered as multiple oral doses to human immunodeficiency virus type 1-infected adult male subjects with asymptomatic HCMV shedding. Subjects received one of six dosage regimens (100, 200, or 400 mg three times a day, or 600, 900, or 1,200 mg twice a day) or a placebo for 28 days. 1263W94 demonstrated linear PK, with steady-state plasma 1263W94 profiles predictable based on single-dose data. 1263W94 was rapidly absorbed following oral dosing, and values for the maximum concentration of the drug in plasma and the area under the concentration-time curve increased in proportion to the dose. 1263W94 demonstrated in vivo anti-HCMV activity in semen at all of the dosage regimens tested, with mean reductions in semen HCMV titers of 2.9 to 3.7 log10 PFU/ml among the four regimens evaluated for anti-HCMV activity. 1263W94 was generally well tolerated; taste disturbance was the most frequently reported adverse event over the 28-day dosing period. Keywords: Introduction : Infection with human cytomegalovirus (HCMV) is common, with seroprevalence ranging from approximately 50 to 60% of adults in Western Europe and the United States to as much as 100% of some adult populations . In immunocompetent HCMV-infected individuals, the virus normally remains latent and does not constitute a major health risk. However, HCMV infection can be a serious complication in immunologically immature individuals, such as neonates, or in immunocompromised individuals, such as solid-organ transplant recipients, bone marrow transplant recipients, or people with AIDS. There are six currently approved therapies in the United States for treatment or prevention of systemic HCMV infection or HCMV retinitis associated with AIDS . Treatment of systemic HCMV infection requires administration of ganciclovir, foscarnet, or cidofovir. After induction therapy with intravenously (i.v.) administered agents or oral valganciclovir, maintenance therapy can be provided by oral ganciclovir or valganciclovir. An intravitreal implant of ganciclovir is available for treatment of HCMV retinitis; however, the implant is insufficient for controlling systemic disease and must be used together with systemic therapy, such as oral or i.v. ganciclovir or oral valganciclovir. Fomivirsen, administered by intravitreal injection, is approved as a second-line therapy for treatment of HCMV retinitis in subjects who cannot tolerate or fail to respond to first-line therapy. Disadvantages of the currently approved therapies include treatment-limiting toxicities, such as bone marrow suppression (ganciclovir) and nephrotoxicity (foscarnet and cidofovir), limited penetration to target sites, and inconvenient i.v. dosing or poor oral bioavailability (oral ganciclovir) . Thus, there is a need for a safe and effective oral therapy for the treatment and prevention of HCMV disease. 1263W94 [maribavir; 5,6-dichloro-2-(isopropylamino)-1,beta-l-ribofuranosyl-1-H-benzimidazole] is a novel benzimidazole compound shown to have antiviral activity against HCMV in vitro . 1263W94 does not require intracellular activation and has demonstrated activity against clinical isolates resistant to ganciclovir or foscarnet . 1263W94 was safely administered as single oral doses of 50 to 1,600 mg to healthy and human immunodeficiency virus (HIV)-infected adults in two separate studies (L. Wang, L., R. Peck, Y. Yin, J. Allanson, R. Wiggs, and M. Wire, unpublished data). In the two single oral dose escalation studies, 1263W94 pharmacokinetics (PK) were dose proportional over the dose range tested, 1263W94 was highly metabolized (similar40% of the dose was recovered in the urine as metabolite, and <2% was recovered as the parent drug), and 1263W94 was highly (similar98.5%) protein bound. Previous in vitro human liver microsomal studies suggested that CYP3A4 was the primary enzyme responsible for 1263W94 metabolism (N dealkylation) (Wang et al., unpublished). Indeed, similar40% of the 1263W94 dose recovered in the urine was recovered as the N-dealkylated metabolite (Wang et al., unpublished). We conducted a phase I dose escalation study to evaluate the PK, anti-HCMV activity, and safety of 1263W94 in HIV-1-infected adult males with asymptomatic HCMV shedding in urine and semen. Quantitative reductions in semen HCMV concentrations are dose responsive and predictive of clinical efficacy (-). MATERIALS AND METHODS : Study population and investigative sites. | The study population consisted of HIV-1-infected males. Eligibility criteria were as follows: all subjects had to be >=18 years old, have a life expectancy of >6 months, and be stable on all chronically administered therapy for HIV infection and opportunistic infections for at least 1 month. Subjects were stratified into two groups, main and satellite, on the basis of semen and urine HCMV culture results prior to study entry. Subjects with a semen HCMV concentration of >=5,000 PFU/ml and with an HCMV-positive urine culture within 30 days of study entry (day 1) were enrolled in the main group for evaluation of the PK, anti-HCMV activity, and safety of 1263W94. There was no requirement for detectable HCMV infection in subjects enrolled in the satellite group for evaluation of 1263W94 PK and safety. Subjects enrolled in the satellite group were required to have a CD4+ cell count of <150 cells/mm3 or <10% of total lymphocytes. For subjects with a CD4+ cell count of <100 cells/mm3 or with signs or symptoms of HCMV disease, an indirect fundoscopic examination was performed by an ophthalmologist to rule out HCMV retinitis. The following exclusion criteria applied to all subjects: active HCMV disease or history of HCMV disease; visual symptoms suggestive of HCMV disease unless HCMV disease was excluded by ophthalmoscopic examination; treatment with ganciclovir, foscarnet, cidofovir, or investigational anti-HCMV drugs within 2 months prior to study entry (day 1); treatment with interferons, immunomodulatory agents, or HCMV hyperimmune globulin within 1 month prior to study entry (day 1); active hepatitis, obstructive hepatobiliary disease, or cirrhosis; gastrointestinal disorders that could interfere with oral dosing or drug absorption or that might indicate HCMV disease; known history of lactose intolerance; diagnosis of chronic diseases that could compromise the safety or compliance of the subject; treatment with radiation therapy or systemic therapy for visceral malignancy within 2 months prior to study entry (day 1), or anticipated need for such treatment during the study period; participation in an investigational trial or treatment with an investigational therapy within 2 months (anti-HCMV therapy) or 1 month (other therapy) prior to study entry (day 1); abnormal laboratory values within 14 days of study entry (day 1), notably, hemoglobin of <8.5 g/dl, a neutrophil count of <750 cells/mm3, a platelet count of <=50,000 cells/mm3, AST, ALT, or alkaline phosphatase levels >4 times the upper limit of normal, total bilirubin of >2 mg/dl, or estimated creatinine clearance of <50 ml/min; or a debilitated condition resulting from HIV disease or associated illnesses or therapies such that the subject was considered unable to complete the study. The study was conducted from 27 August 1996 through 9 July 1997 at three sites in San Francisco, Calif. Quest Clinical Research, University of California San Francisco (UCSF) Mount Zion Medical Center, and San Francisco General Hospital were the three participating centers. The study was approved by the Western Institutional Review Board (Olympia, Wash.) and by the UCSF Committee on Human Research (San Francisco, Calif.) and was conducted under Good Clinical Practices guidelines. All subjects provided written informed consent before any study procedures were performed. The Virology Research Laboratory at the UCSF Mount Zion Medical Center performed the plaque assays and processed samples for shipment to GlaxoWellcome for HCMV DNA PCR analysis, viral sensitivity testing, and measurement of 1263W94 concentrations. Study design. | The study was a phase I multiple-dose, randomized, parallel dose escalation study. Eligible subjects were stratified to the main group or the satellite group on the basis of quantitative HCMV culture in semen and qualitative HCMV culture in urine, as described above. The study was designed to evaluate the PK and safety of 1263W94 in both groups and the anti-HCMV activity of 1263W94 in the main group. Subjects in the main group received open-label 1263W94 at one of the following dosage regimens: 100, 200, or 400 mg three times a day (t.i.d.) or 600 mg twice a day (b.i.d.). Subjects in the satellite group were randomized in a double-blind fashion to receive 1263W94 at a particular dosage regimen (100, 200, or 400 mg t.i.d. or 600, 900, or 1,200 mg b.i.d.) or a matching placebo. Subjects were sequentially enrolled into the escalating-dose cohorts. The study was designed to include a prescreening visit on day -30, a screening visit at day -14, enrollment and initial dosing on day 1, weekly visits on days 7, 14, 21, and 28, and a follow-up visit approximately 4 weeks after the final dosing on day 28 (i.e., day 56). Prescreening assessments of HCMV in semen and urine were performed to determine whether to assign subjects to the main group or the satellite group, as described above. Screening assessments included a test for HIV-1 antibody (enzyme-linked immunosorbent assay), complete medical history, review of inclusion-exclusion criteria, demographic data, ocular examination (if necessary), and clinical evaluations (including physical examination, vital signs, height and weight measurements, assessment of HIV-related conditions, assessment of concurrent medications, and clinical laboratory evaluations, including hematology, CD4+ lymphocyte count, serum chemistry, and urinalysis). On-study assessments included clinical evaluations as described above, review of adverse events, PK evaluations on specified visit days, and HCMV assessments on specified visit days. The study drug was administered on an outpatient basis, without regard to meals, except on days 1 and 28, when serial PK samples were collected following an overnight fast (at least 8 h). A single dose of the study drug was given on day 1, and routine b.i.d. or t.i.d. dosing began on day 2. PK sampling. | Serial plasma samples were collected on days 1 and 28. Subjects were required to fast the evening before dosing on days 1 and 28 and not to take the morning dose (day 28) before arriving at the clinic. Each subject had a cannula inserted into a suitable forearm vein before dosing. A predose blood sample was collected, and the dose was given orally with 480 ml of water. Subjects were required to fast for another 3 h after receiving the dose; afterwards, a regular lunch and dinner were provided at the clinic. Seven-milliliter whole-blood samples were collected in tubes containing EDTA at the following times after dosing: 0.25, 0.5, 1.0, 1.5, 2.0, 3.0, 4.0, 5.0, 6.0, 8.0, 10.0, 12.0, 16.0, and 24.0 h. Each sample was immediately refrigerated or stored on ice until centrifugation. Within an hour after collection, blood was separated by centrifugation, and plasma was aliquoted into two polypropylene storage tubes. Plasma samples were stored upright at -20C. All subjects collected semen samples to bring to the clinic on days 1, 14, and 28; subjects enrolled in the main group collected additional semen samples for the clinic visits on days 7 and 21. Following protocol amendments, subjects in the main group of the 400-mg t.i.d. and 600-mg b.i.d. cohorts also collected semen samples for the day-4 visit. Samples were collected at home in sterile containers, either on the evening before the day of the clinic visit or on the morning of the visit day. The timing of the semen collection was not standardized relative to the time of dosing or relative to the time of plasma PK sampling. Samples collected in the evening were stored in the refrigerator overnight. A 50- to 100-mul semen aliquot was transferred to a polypropylene storage tube and stored at -20C until shipment to GlaxoWellcome for measurement of 1263W94 concentrations by a validated assay. On days 1 and 28, subjects were asked to void before dosing, and urine samples for PK analysis were collected at 0 to 4, 4 to 8, 8 to 12, and 12 to 24 h postdosing. Urine samples were stored in the refrigerator over the collection interval. A 10-ml aliquot from each urine collection interval was transferred to a polypropylene storage tube, and samples were stored upright at -20C until shipment to GlaxoWellcome for measurement of 1263W94 concentrations by a validated assay. Bioanalysis of PK samples. | Plasma, urine, and semen samples were analyzed for 1263W94 by use of separately validated high-performance liquid chromatography-mass spectrometry (MS)-MS methods following solid-phase extraction. For plasma, the validated calibration range was 50 to 6,000 ng/ml, the accuracy (expressed as percent bias) was +-5.1%, and the global precision (expressed as the coefficient of variation [CV]) was 14.1%. For urine, the validated calibration range was 50 to 50,000 ng/ml, the accuracy (percent bias) was +-9.4%, and the global precision (CV) was 10.0%. For semen, the validated calibration range was 50 to 7,500 ng/ml, the accuracy (percent bias) was +-10.5%, and the global precision (CV) was 10.2%. Plasma PK analysis. | PK analyses of plasma 1263W94 concentration-versus-time data obtained following single-dose administration on day 1 and following multiple dosing to steady state on day 28 were conducted by noncompartmental methods, with the log-linear trapezoidal option in WinNonlin Professional, version 1.5 (Pharsight Corporation, Mountain View, Calif.). Values were estimated for the maximum concentration of 1263W94 in plasma (Cmax) and Cmax at steady state (Cmax,ss), time to maximum concentration (Tmax) and Tmax at steady state (Tmax,ss), minimum concentration during a dosing interval at steady state (Cmin,ss), apparent terminal elimination half-life (t1/2), area under the concentration-time curve (AUCinfinity and AUCtau,ss), and average concentration at steady state (Cavg,ss). The daily area under the concentration-time curve at steady state (AUC24,ss) was calculated as AUCtau,ss24 h/tau. Urine PK analysis. | Analyses of urinary excretion data for 1263W94 were performed to determine the percentage of the dose recovered as 1263W94. The urine 1263W94 concentration was multiplied by the volume of urine collected during each urine collection interval, divided by the dose of drug administered, and multiplied by 100. The cumulative percentage of the dose excreted in urine as 1263W94 was determined by summing the percentages of the dose excreted in each urine collection interval. Statistical analysis of 1263W94 PK. | Descriptive statistics were calculated for plasma and urine 1263W94 PK parameters. Log-transformed AUCinfinity and AUCtau,ss were compared within-subject by using analysis of variance (SAS, version 6.08, MIXED procedure [SAS Institute, Cary, N.C.]) to assess the time invariance of 1263W94 PK. The geometric least-squares mean (LSM) for each PK parameter was determined, and the ratio of the geometric LSM for AUCtau,ss to that for AUCinfinity was calculated, along with the associated 90% confidence intervals (90% CI). Dose proportionality was assessed by fitting the data to a power model, relating log-transformed Cmax, AUCinfinity, and AUC24,ss to the log-transformed dose (log-transformed parameter = alpha + betalog-transformed dose), by restricted maximum likelihood using the MIXED procedure. The common slope was estimated, and the associated 90% CI was constructed to examine linearity. The proximity to unity of the slope estimate for AUCinfinity was considered the primary assessment of dose proportionality, and the proximity to unity of the slope estimate for Cmax was considered the secondary assessment of dose proportionality. HCMV sampling. | Stratification of subjects to the main or satellite group was determined by prescreening HCMV assessments that consisted of quantitation of HCMV in semen by using a plaque assay and a qualitative analysis of the presence of HCMV in urine. HCMV assessments for main-group subjects during the study included the following: quantitative evaluation of HCMV in semen (days 1, 4, 7, 14, 21, and 28) by both plaque titration and a PCR-based assay, quantitative evaluation of HCMV DNA in whole blood (days 1, 4, 7, 14, 21, and 28) by a PCR-based assay, qualitative HCMV culture from urine (days 1, 4, 7, 14, 21, and 28), and isolation of HCMV from semen and urine for assessment of the sensitivity of the virus to 1263W94 (days 1, 28, and 56). All samples for HCMV testing were collected prior to dosing on the visit day. Semen samples were collected as described for PK sampling above. Immediately after collection, semen samples were processed by the Virology Research Laboratory for subsequent analyses. Semen lysates containing cell-free virus were obtained for quantitative PCR-based analysis of HCMV concentrations by diluting an aliquot of the semen sample 1:10 in minimal essential medium (MEM; Gibco BRL, Gaithersburg, Md.), lysing the cells by sonication, and filtering the lysate. A 0.5-ml aliquot of the filtered semen lysate was transferred to a polypropylene storage tube and stored upright at -70C until shipment to GlaxoWellcome. Whole-blood samples for quantitation of HCMV by a PCR-based assay were collected in 4-ml EDTA-containing tubes. These samples were stored upright at -70C until shipment to GlaxoWellcome for analysis. For qualitative urine HCMV culture, 10- to 20-ml aliquots from each urine collection were refrigerated and later processed by the Virology Research Laboratory. Cell culture and media. | Laboratory strain AD169 (American Type Culture Collection, Manassas, Va.) was used as the wild-type HCMV reference strain, and strain 2916r was used as a reference 1263W94-resistant strain. 2916r is a derivative of AD169 selected for growth in the presence of the benzimidazole compound 2916W93 . Human foreskin fibroblasts (HFF) and MRC-5 human lung fibroblasts were obtained from BioWhittaker (Walkersville, Md.) and used between passages 20 and 30. Except as noted, reagents and cell media were obtained from Gibco BRL. Cells were cultured in MEM supplemented with 4% fetal bovine serum (HyClone, Logan, Utah), 2 mM l-glutamine, 100 U of penicillin G/ml, 100 mug of streptomycin sulfate/ml, and 20 mug of amphotericin B (Fungizone)/ml. Quantitation of HCMV in semen samples by plaque assay. | Quantitative culture of HCMV from semen samples was performed by the Virology Research Laboratory using a viral plaque assay (titration). Semen samples were initially diluted 1:10 in MEM and sonicated for 30 s for lysis. The sonicate was filtered through a 0.45-mum-pore-size filter, and appropriate dilutions were plated in duplicate in 24-well plates containing HFF. After incubation at 37C for 3 h, the wells were washed and refed with MEM. Cells were cultured for 7 days, after which they were fixed and stained, plaques were counted, and the titer of HCMV in semen (expressed as PFU per milliliter) was determined. Qualitative HCMV plaque assay in urine. | Qualitative culture of HCMV (expressed as positive versus negative results) from urine samples was performed by the Virology Research Laboratory using a viral plaque assay similar to that described above for semen samples. Quantitative HCMV DNA PCR assay. | Quantitation of HCMV DNA was performed using an assay under development at that time by Roche Molecular Systems (Branchburg, N.J.). Reagents for PCR amplification were provided by Roche Molecular Systems, and the assays were performed according to the manufacturer's specifications. The limit of detection of the assay was 16 copies/ml, and the dynamic range was >103 copies/ml. PCR amplification was performed in a 96-well plate by using an ABI Thermal Cycler 9600 (Perkin-Elmer Applied Biosystems, Foster City, Calif.), with the following amplification conditions: 50C for 10 min; amplification at 96C for 30 s and 65C for 30 s for 30 cycles; and a hold at 72C for no more than 2 h. Quantitative detection of the PCR products was based on colorimetric detection of an avidin-horseradish peroxidase complex bound to the biotin moiety of the HCMV DNA primer. Determination of HCMV sensitivity to 1263W94 by plaque reduction assay. | HCMV isolates from semen and urine samples were shipped either as growing, infected cells or as frozen stocks. Infected cells were plated on MRC-5 cells and grown for at least 2 passages until 70 to 90% of cells showed characteristic HCMV cytopathic effects. Infected cell stocks were made and titered. Plaque reduction assays were performed to determine whether clinical isolates had acquired resistance to 1263W94. Clinical isolates were analyzed as cell-associated virus, and AD169 was analyzed as cell-free supernatant virus. Drug concentrations were tested in triplicate, and the mean values were determined for each clinical isolate at each concentration. 1263W94 was assayed at seven concentrations, ranging from 0.01 to 30 muM, as previously described . RESULTS | Subject population. | Seventy-eight subjects were enrolled in the study. Of these, 28 were assigned to the main group and 50 were assigned to the satellite group. The number of subjects included in each of the 1263W94 dose cohorts is shown in Table . The 62 subjects receiving 1263W94 and the 16 subjects receiving the placebo were similar in demographic characteristics, in CD4+ cell count, and in Centers for Disease Control and Prevention (CDC) classification status at baseline, as shown in Table . Of the 78 subjects enrolled in the study, 70 completed the 28-day dosing period of the study. Eight subjects withdrew from the study prematurely. Of these eight, six subjects (three in the 200-mg t.i.d. cohort and one each in the 600-, 900-, and 1,200-mg b.i.d. cohorts) discontinued due to adverse events, consisting of grade-2 rash (five subjects) and sinusitis (one subject). A seventh subject (in the 900-mg b.i.d. cohort) was withdrawn from the study after the day-1 hematology results indicated neutropenia (400 cells/mm3), and an eighth subject (in the 900-mg b.i.d. cohort) withdrew from the study on day 2 due to anxiety about the serial PK sampling. Concomitant medications. | Subjects were stable on all chronically administered therapy for HIV and opportunistic infections for at least 1 month prior to receiving 1263W94. Most of the subjects were receiving concomitant medications that were CYP3A4 inhibitors, such as protease inhibitors and antifungal agents. One subject received efavirenz, a CYP3A4 inducer. PK analysis. | Descriptive statistics for plasma 1263W94 PK parameters are displayed in Table . There was a dose-proportional increase in plasma 1263W94 AUCinfinity, Cmax, and AUC24,ss over the dose range tested, as displayed in Fig. . The slope estimates (and 90% CI) for plasma 1263W94 parameters were as follows: AUCinfinity, 1.12 (0.98, 1.27); AUC24,ss, 0.95 (0.84, 1.06); and Cmax, 0.98 (0.85, 1.10). 1263W94 demonstrated linear PK, with steady-state plasma 1263W94 profiles predictable based on single-dose data as demonstrated by the ratio (and 90% CI) of AUCtau,ss to AUCinfinity. The ratio of plasma 1263W94 AUCtau,ss to AUCinfinity was close to 1, and the associated 90% CI was relatively balanced on either side of 1, for each of the doses administered. The percentage of the dose that was excreted in urine as 1263W94 was minimal; on average, less than 3% of the parent compound was eliminated in urine. Semen 1263W94 concentrations increased with increasing doses. Median 1263W94 concentrations in semen (with the range and number of subjects given in parentheses) were as follows: for the 100-mg t.i.d. cohort, 1.67 mug/ml (0.30 to 3.53; n = 9); for the 200-mg t.i.d. cohort, 2.96 mug/ml (2.54 to 3.09; n = 4); for the 400-mg t.i.d. cohort, 8.41 mug/ml (3.62 to 25.57; n = 9); and for the 900-mg b.i.d. cohort, 11.85 mug/ml (3.42 to 21.77; n = 8). A semen 1263W94 concentration of 6.22 mug/ml was measured for a single subject in the 600-mg b.i.d. cohort. Quantitation of HCMV in semen and whole blood. | Figure presents changes from baseline in semen HCMV amounts (based on plaque assays and PCR assays) for subjects in the main group at days 4 (when applicable), 7, 14, 21, and 28. Based on HCMV titers, subjects in the 200-mg t.i.d. (n = 5) and the 400-mg t.i.d. (n = 6) cohorts experienced the greatest decreases in HCMV amounts, with a mean decrease at day 28 of 3.7 log10 PFU/ml (standard deviations [SD], 0.96 for the 200-mg t.i.d. cohort and 1.08 for the 400-mg t.i.d. cohort), while subjects in the 100-mg t.i.d. cohort (n = 7) experienced a mean (+-SD) decrease of 2.9 (+-0.79) log10 PFU/ml and subjects in the 600-mg b.i.d. (n = 6) cohort experienced a mean (+-SD) decrease of 3.3 (+-1.32) log10 PFU/ml. Quantitative reductions in semen HCMV DNA amounts and the differences between the cohorts were less pronounced than for HCMV amounts as measured by titers. At day 28, mean decreases from baseline in HCMV DNA amounts were 1.1 log10 copies/ml for the 100-mg t.i.d. cohort (n = 7), 1.5 log10 copies/ml for the 200-mg t.i.d. cohort (n = 4), and 1.3 log10 copies/ml for both the 400-mg t.i.d. (n = 5) and 600-mg b.i.d. (n = 2) cohorts. The numbers (and percentages) of subjects in each cohort with >=2-log10-unit reductions from baseline in semen HCMV titers at day 28 were as follows: 6 of 7 subjects (86%) in the 100-mg t.i.d. cohort, 5 of 5 subjects (100%) in the 200-mg t.i.d. cohort, 6 of 6 subjects (100%) in the 400-mg t.i.d. cohort, and 5 of 6 subjects (83%) in the 600-mg b.i.d. cohort. Only a single subject in each cohort had a >=2-log10-unit reduction in semen HCMV DNA concentrations from baseline to day 28. Only one subject, who received 1263W94 at 600 mg b.i.d., had quantifiable HCMV DNA in whole blood at baseline (4.81 log10 copies/ml). In this subject HCMV DNA amounts decreased to below the limit of detection by the PCR assay (<1.2 log10 copies/ml) by day 28. Detection of HCMV in the urine. | The numbers (and percentages) of subjects in each cohort whose qualitative HCMV cultures showed positive results at baseline and negative results in the last two on-treatment measurements (days 21 and 28) were as follows: 2 of 7 (29%) in the 100-mg t.i.d. cohort, 1 of 5 (20%) in the 200-mg t.i.d. cohort, 3 of 6 (50%) in the 400-mg t.i.d. cohort, and 1 of 6 (17%) in the 600-mg b.i.d. cohort. HCMV sensitivity to 1263W94 by plaque reduction assay. | Ninety-two HCMV isolates were obtained from day-1, day-28, or day-56 cultures of semen and urine samples. One subject provided a day-11 semen sample, which was positive for HCMV and was included in the analysis. The 92 isolates analyzed were from 31 subjects in the 100-mg t.i.d., 200-mg t.i.d., and 400-mg t.i.d. cohorts and included 58 semen samples (24 from day 1, 1 from day 11, 15 from day 28, and 18 from day 56) and 34 urine samples (13 from day 1, 9 from day 28, and 12 from day 56). The median 50% inhibitory concentration (IC50) of 1263W94 for the clinical isolates was 0.27 muM (range, 0.05 to 0.88 muM). There were no significant differences in the median IC50 between the day-1 samples and the day-28 or day-56 samples. The median IC50 for wild-type strain AD169, determined from 10 separate plaque reduction assays, was 0.55 muM (range, 0.23 to 0.78 muM). Because the clinical isolates tested might contain mixed populations of drug-susceptible virus and treatment-emergent drug-resistant virus selected by exposure to 1263W94, additional assays were performed to determine the level of resistant virus in a mixed viral population necessary to produce an increase in the IC50. Mixtures of wild-type AD169 and the 1263W94-resistant strain 2916r were assayed for 1263W94 susceptibility by using plaque reduction assays. IC50s did not increase with a mixture containing 10% resistant virus but increased 3.3- and 16-fold with mixtures containing 25 and 50% resistant virus, respectively. These results suggested that emergence of 1263W94-resistant variants comprising as little as 25% of the viral population would have been detected in plaque reduction assays of the clinical isolates. Safety and tolerability. | The majority of subjects enrolled in the study, including 60 of 62 subjects (97%) who received 1263W94 and 10 of 16 subjects (63%) who received the placebo, experienced at least one adverse event during the study period. Most of the subjects reported adverse events that were neurological, such as taste disturbances and headache, or gastrointestinal, such as diarrhea and nausea. Adverse events reported by >=10% of subjects in either the 1263W94 or the placebo group, and the numbers (and percentages) of subjects reporting the event in the 1263W94 and placebo groups, respectively, were as follows: taste disturbances, 51 (82%) and 3 (19%); headache, 13 (21%) and 3 (19%); diarrhea, 16 (26%) and 2 (13%); nausea, 14 (23%) and 2 (13%); rash, 12 (19%) and 1 (6%); pruritus, 12 (19%) and 1 (6%); fever, 7 (11%) and 0; exacerbation of fatigue, 6 (10%) and 2 (13%); vomiting, 5 (8%) and 2 (13%); constipation, 0 and 2 (13%); and upper respiratory tract infection, 2 (3%) and 2 (13%). Diarrhea and taste disturbances appeared to be dose related. After approximately 7 to 12 days of therapy, five subjects receiving 1263W94 prematurely discontinued the study due to a drug-related diffuse maculopapular rash of moderate (grade-2) intensity. Two of these subjects received 200 mg t.i.d., and one each received 600, 900, and 1,200 mg b.i.d. Four of these subjects had a history of allergic reaction to other drugs. After discontinuation of the study drug, the rash resolved within approximately 1 to 4 days without sequelae. Two serious non-drug-related adverse events were reported during the study. The first event was cholecystitis, for which the subject underwent a cholecystectomy. The second event was a single episode of pulmonary thromboembolism. The symptoms of pulmonary thromboembolism (dyspnea and chest pain) resolved within 2 weeks of hospital admission and initiation of appropriate therapy. Both events occurred in subjects who received the placebo. Clinical laboratory values for subjects who received 1263W94 and subjects who received the placebo were not significantly different, and there were no dose-related trends in clinical laboratory values. Decreases in lymphocytes and increases in total protein were consistent with the HIV-infected status of the subjects, and modest decreases in hemoglobin were consistent with the phlebotomy requirements of the protocol. Hemoglobin values increased to screening values by the 4-week poststudy visit. FIG. 1. | Plots of dose versus plasma 1263W94 AUCinfinity (a) and AUC24,ss (b). Plots of dose versus plasma 1263W94 AUCinfinity (a) and AUC24,ss (b). FIG. 2. | (a) Median concentration-time profiles of HCMV in semen, measured in PFU per milliliter by using plaque assays. (a) Median concentration-time profiles of HCMV in semen, measured in PFU per milliliter by using plaque assays. Symbols: , 100 mg t.i.d. (n = 7); , 200 mg t.i.d. (n = 5); , 400 mg t.i.d. (n = 6); , 600 mg b.i.d. (n = 6). (b) Median concentration-time profiles of HCMV DNA in semen, measured in log10 copies per milliliter by using PCR analysis. Symbols: , 100 mg t.i.d. (n = 7); , 200 mg t.i.d. (n = 4); , 400 mg t.i.d. (n = 5); , 600 mg b.i.d. (n = 2). TABLE 1 | Demographic and baseline characteristics TABLE 2 | Plasma 1263W94 PK parameters at days 1 and 28 DISCUSSION : There is a need to develop effective and safe oral therapies for the treatment and prevention of HCMV disease. 1263W94 is a benzimidazole riboside with potent and selective inhibition of HCMV in vitro and with limited cytotoxic and toxicologic effects based on in vitro and in vivo screenings . Pre-clinical studies have indicated that 1263W94 has good oral bioavailability . This study was designed to evaluate the PK, anti-HCMV activity, safety, and tolerability of 1263W94 administered as multiple oral doses to HIV-infected subjects with asymptomatic HCMV shedding in urine and semen over a 28-day treatment period. Anti-HCMV activity was evaluated by measuring HCMV titers in semen and by measuring HCMV DNA amounts in semen and whole blood by PCR. Urine cultures were not used for quantitation of HCMV, because the urinary tract is a closed system, and variations in urine volume may affect accurate quantitation of HCMV. Semen HCMV titration has proven useful in demonstrating in vivo anti-HCMV activity and in selecting clinically useful doses of HCMV therapies. For example, cidofovir was more effective at reducing semen HCMV titers at a weekly dose of 5.0 mg/kg of body weight than at a weekly dose of 3.0 mg/kg , and the higher dose was also more effective in clinical trials Over 28 days of dosing, 1263W94 demonstrated in vivo anti-HCMV activity in semen at all of the dosage regimens tested (100, 200, and 400 mg t.i.d., and 600 mg b.i.d.), with mean reductions in semen HCMV titers of 2.9 to 3.7 log10 PFU/ml among these four regimens. The lowest dosage regimen appeared to have less anti-HCMV activity than the three higher dosage regimens; however, the reductions in HCMV titers for all of the 1263W94 dosage regimens tested compare favorably with results reported for the approved doses of cidofovir (5 mg/kg) . A greater change from baseline in HCMV load following administration of 1263W94 was measurable by culture using plaque titration than by quantitative PCR. An explanation could be that synthesis of DNA continues to some degree in the presence of 1263W94 but that intact viable virus is not produced. Thus, the antiviral effect of 1263W94 could initially be more evident by culture of viable virus than by quantitation of HCMV DNA synthesis. Also, the quantitative reduction in viral DNA may lag behind plaque reduction. If we had continued to assay viral DNA in semen at weeks 5 and 6 instead of stopping at 28 days, we might have seen greater log reductions in HCMV DNA concentrations. Although relatively high concentrations of 1263W94 were detected in semen, the anti-HCMV effect in semen as measured by plaque assay is not attributable to drug carryover into viral cultures, as shown by the following observations. (i) Semen HCMV titers decreased progressively from day 1 to day 28 (see Fig. ); a carryover effect should have been equally apparent in all sequential samples. (ii) HCMV DNA levels in semen also decreased over the 28-day treatment period. The PCR assay is a direct measure of viral DNA at the time of specimen collection and is not subject to inhibition by 1263W94 during the assay. There was a dose-proportional increase in plasma 1263W94 AUCinfinity, Cmax, and AUC24,ss over the dose range tested. 1263W94 demonstrated linear PK, with steady-state plasma profiles predictable based on single-dose data. Several of the concomitant medications that subjects received for treatment of HIV and opportunistic infections were inhibitors, inducers, and/or substrates of CYP3A4, the isoenzyme primarily responsible for 1263W94 metabolism. The effects of these drugs on 1263W94 PK were not assessed. 1263W94 was generally safe and reasonably well tolerated during the 28 days of dosing. Six subjects prematurely discontinued the study drug due to adverse events: five cases of rash and one of sinusitis. Two subjects, both receiving the placebo, reported serious non-drug-related adverse events during the study. Taste disturbance was the most frequently reported adverse event. Taste disturbance, diarrhea, nausea, rash, pruritus, and fever were reported by a higher percentage of subjects receiving 1263W94 versus the placebo, and taste disturbance and diarrhea appeared to be dose related. Overall, 1263W94 showed a favorable profile with regard to safety, tolerability, PK, and anti-HCMV effect. The absence of resistance in isolates obtained at day 28 of treatment (or 28 days later) is encouraging but is similar to previous data for ganciclovir and cidofovir . In order to evaluate the risk of developing resistance, it would be necessary to assess IC50s for isolates from patients receiving 1263W94 for >=90 days . The oral bioavailability of 1263W94, the duration and magnitude of its antiviral effect, and the lack of dose-limiting toxicity make it very attractive as a potential anti-HCMV therapeutic and prophylactic agent. The reductions in HCMV titers for all of the 1263W94 dosage regimens tested compare favorably with results reported for the approved anti-HCMV agents. In addition, the relatively benign toxicity profile of 1263W94 and the specific absence of nephrotoxicity or myelotoxicity suggest a potential role for this agent for patients receiving solid-organ and bone marrow transplants as well as for patients with HIV. Backmatter: PMID- 12183239 TI - Immunomodulatory Effect of Zidovudine (ZDV) on Cytotoxic T Lymphocytes Previously Exposed to ZDV AB - In a previous study, zidovudine (ZDV) was shown to cause a concentration-dependent inhibition of antigen-specific cytotoxic T-lymphocyte (CTL) clonal expansion (S. Francke, C. G. Orosz, K. A. Hayes, and L. E. Mathes, Antimicrob. Agents Chemother. 44:1900-1905, 2000). However, this suppressive effect was lost if exposure to ZDV was delayed for 24 to 48 h during the antigen sensitization period, suggesting that antigen-primed CTL may be less susceptible than naive T lymphocytes to the suppressive effects of ZDV. The present study was undertaken to determine if naive T lymphocytes were more sensitive to the suppressive effects of ZDV than T lymphocytes previously exposed to antigen. The 50% inhibitory concentration (IC50) values of ZDV were determined on naive and antigen-primed T-cell responses in an alloantigen system. Lymphocyte cultures with continuous antigen exposure (double prime) were more resistant to ZDV suppression (IC50 = 316 muM) than were naive lymphocytes (IC50 = 87.5 muM). Interestingly, lymphocytes that were antigen primed but deprived of antigen during the final 7 days of culture (prime/hold) were exquisitely sensitive to ZDV suppression (IC50 = 29.3 muM). The addition of 80 muM ZDV during the initial priming of the single-prime (prime/hold) and double-prime cultures did not select for a more drug-resistant cell population. The differences in ZDV sensitivities are likely a reflection of the physiological properties of the lymphocytes related to their activation state. Keywords: Introduction : Zidovudine (azidothymidine; ZDV), in single or combination therapy, is one of the major drugs used to treat AIDS. However, its positive therapeutic value is countered by clinical side effects and the development of drug resistance . ZDV also has potentially troubling properties affecting immune cell functions, such as suppression of antigen-driven T-cell proliferation , prolongation of the cell cycle , and inhibition of a number of other immunologic responses including lectin- and antigen-induced mitosis, mixed lymphocyte culture reactions, and induction of the cytotoxic T-lymphocyte (CTL) response . Mechanisms by which ZDV influences cell, and specifically immune cell, physiology are largely unknown, and the biologic relevance of these effects in vivo is speculative. However, the potential impact of ZDV on CTL-mediated cytolysis is of particular concern given the importance of these cells in combating human immunodeficiency virus (HIV) infection . In a previous study that used limiting dilution analysis, we reported that ZDV caused concentration-dependent inhibition of clonal expansion of antigen-specific CTL, suggesting that the basis for ZDV-related immunosuppression is stasis of T-cell expansion . In this study the estimated frequency of alloantigen-specific CTLs was profoundly lower in in vitro sensitization assays, where ZDV was present during primary antigen exposure . However, this suppressive effect was lost if exposure to ZDV was delayed for 24 to 48 h during the antigen sensitization period. The results suggested that antigen-primed CTL may be less susceptible to the suppressive effects of ZDV than naive T lymphocytes. The objectives of the present study were (i) to measure the sensitivity of naive T lymphocytes to the suppressive effects of ZDV, and (ii) to determine if T cells sensitized to antigen in the presence of ZDV generated CTL with greater resistance to ZDV suppression. In order to explore this possibility, we used a 50% inhibitory concentration (IC50) assay to measure the relative suppressive effect of ZDV on naive and primed CTL. The results suggest that naive cytotoxic T cells are two to five times more sensitive to the inhibitory effects of ZDV than are activated antigen-primed cells. However, previously antigen-primed T cells that were cultured without antigen but given interleukin-2 (IL-2) were shown to have increased sensitivity to ZDV over that determined for naive CTL. Antigen priming in the presence of ZDV did not generate a cytotoxic T-cell population with greater resistance to ZDV suppression. MATERIALS AND METHODS : Mice. | Animal studies were performed in accordance with the University Laboratory Animal Care and Use Committee and DHEW publication no. NIH 74-23, Guide for the Care and Use of Laboratory Animals. Six- to 8-week-old female DBA/2 (H-2d, Mlsa) and C57BL/6 (B6; H-2b, Mlsb) mice were purchased from Harlan Sprague Dawley Inc. (Indianapolis, Ind.). The mice were housed in a laminar flow cabinet (animal storage isolator; Nu Aire Inc., Plymouth, Minn.) in groups of 5 to 10 animals per cage. Upon arrival, all mice were allowed a 7-day period of acclimation before use. Animals were sacrificed and spleens were collected within 2 weeks following the acclimation period. ZDV. | ZDV was obtained from the AIDS Research and Reference Reagent Program, Division of AIDS, National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH), as a lyophilized powder and stored light-protected at room temperature. ZDV was dissolved in cell culture medium for in vitro studies. Cell harvest, separation, and culture. | Intact spleens from mice euthanized in a standard CO2 chamber were dispersed into single-cell suspensions of splenocytes and washed three times in sterile phosphate-buffered saline. Tissue culture medium was Dulbecco's modified Eagle's medium supplemented with 1.6 mM l-glutamine, 0.27 mM folic acid, 0.27 mM l-asparagine, 0.55 mM l-arginine, 10 mM HEPES buffer, 1.0 mM sodium pyruvate, 100 U of penicillin-streptomycin/ml (Gibco, Grand Island, N.Y.), 5 x 10-5 M beta-mercaptoethanol (Sigma Chemical, St. Louis, Mo.), and 10% heat-inactivated fetal calf serum, with 10 IU of IL-2/ml (Boehringer Mannheim) freshly added. Study design. (i) Animal model. | The experimental model for testing T-cell sensitivity to ZDV is based on the one-way mixed lymphocyte response (MLR) between the major histocompatibility complex-distinct mouse strains C57BL6 (responder) and DBA (stimulator). The product of the MLR is a C57BL6 T-cell culture with a high frequency of DBA-specific cytotoxic T cells . These single-primed cells can be expanded by restimulation with a second round of antigen treatment (cocultivation with irradiated DBA splenocytes [irrDBA]) to generate what we have termed double-prime cells. The single-prime cells from the MLR can be maintained without a second round of antigen exposure by cocultivation with irradiated autologous C57BL6 splenocytes. We refer to these cell preparations as prime-and-hold (prime/hold) T lymphocytes. Thus, depending on the stage of cell stimulation, ZDV sensitivity can be evaluated in three populations of T cells: naive (fresh splenocytes from C57BL6 mice), single primed (prime/hold), or double primed. The rationale for using prime/hold and double-prime CTLs is as follows: alloantigen-activated CTL, in contrast to nonactivated CTL precursors, no longer require contact with specific alloantigens to allow lymphokine-mediated clonal expansion and subsequent detection in microcultures . Exposure to a single round of allogeneic antigen stimulation, however, leads to activation of only a fraction of all antigen-specific cells. Subsequent reexposure to the same antigen (double priming) increases the percentage of antigen-reactive cells by two mechanisms: (i) clonal expansion of already antigen sensitized cells, and (ii) first-time activation of cells which did not respond to the antigen during the first encounter. (ii) Experimental design. | The experimental design had five arms . (a) Arm 1 (naive CTL). | Naive CTL were fresh splenocytes from C57BL6 mice placed directly into the IC50 assay. Briefly, 3 x 104 C57BL6 splenocytes/well (responder cells) were incubated for 7 days with 3 x 104 irrDBA-2 splenocytes/well (sensitizing cells) in U-bottom 96-well plates in the presence of ZDV concentrations ranging between 0 and 1,000 muM. Each ZDV concentration was tested in replicates of six. (b) Arm 2 and Arm 4 (double prime). | For the double-primed CTL cultures, 105 C57BL6 splenocytes were incubated with 105 irrDBA-2 splenocytes for a period of 14 days at 37C in 10% CO2 in the presence (Arm 4) or absence (Arm 2) of 80 muM ZDV . This concentration of ZDV was selected based on previous studies showing suppression of T-cell responses . The cultures were given 25 mul of fresh IL-2 containing growth medium after 9 days of culture. For Arm 4, the growth medium contained 80 muM ZDV. After the 14-day culture period, effector cells were harvested, pooled, and then distributed in U-bottom microculture plates to achieve a final concentration of 3 x 103 effector cells/well. In preparation for the IC50 assay, each well received 105 irrDBA-2 splenocytes (second round of antigen stimulation). The IC50 assay is described below. (c) Arm 3 and Arm 5 (prime and hold). | The prime/hold cultures were the same as the double-prime cultures during the initial 14-day culture period . At that point, however, instead of adding irradiated DBA cells, these cultures received 105 irradiated C57BL6 splenocytes/well as feeder cells in preparation for the IC50 assay. Arm 5 contained 80 muM ZDV during the initial 14-day culture, while Arm 3 was grown without ZDV . IC50 assay. | The IC50 assay followed the procedure outlined in Fig. . As described above, the C57BL6/irrDBA-2 allogeneic mouse system was used as a source of CTL to measure the in vitro effect of ZDV on CTL effector cells. The assay procedure was adapted from a previously described protocol for determining precursor frequency by limiting dilution analysis . The assay measures drug-mediated inhibition of in vitro sensitization and clonal expansion of C57BL6 splenocytes (effector cells) in response to irrDBA-2 splenocytes (sensitizing cells). By calculating an IC50 for this drug effect, it is possible to compare the relative sensitivities of different subsets of CTL to ZDV and other drugs. 51Cr release assay. | A 100-mul volume of cell suspension from each well was transferred to V-bottom microculture plates containing 3 x 103 51Cr-labeled P815 target cells. The assay mixture was incubated for 5 h at 37C in 5% CO2 and then assayed for 51Cr release. P815 cells (DBA-2 origin, H-2d, Mlsa), the target for the CTL assay, were propagated in culture medium consisting of equal parts of Leibovitz L-15 and RPMI 1640, 10% heat-inactivated fetal bovine serum, 100 IU of penicillin/ml, and 100 mug of streptomycin/ml at 5% CO2 and 37C. For 51Cr labeling, a cell pellet of 3 x 106 P815 cells collected from log-phase cultures was resuspended in 650 muCi of Na51CrO4 (New England Nuclear, Boston, Mass.) in a total volume of approximately 100 mul, incubated at 37C, 5% CO2 for 2 h, washed at least three times in phosphate-buffered saline, and adjusted to the required cell concentration. Drug IC50 determination. | To define the IC50 of ZDV, data points from the IC50 assay were fitted to the logistic model by nonlinear regression using the computer program JUMP-IN (SAS Institute Inc.). The logistic model uses the formula alpha = theta1/[1 + theta2 x e(theta3 x log dose)], where alpha was the fitted log dose, and theta1, theta2, and theta3 were adjustable parameters set initially to equal 1 before the fitting iterations, and log dose was the log10 of the ZDV concentration (in micromolar) used in the assay. A fitted curve was plotted on an x-y axis where alpha was the x value and fa was the y value (see Fig. ). fa was defined as the fraction affected by drug, using the formula fa = 1 - [(ZDV-treated CTL count - spontaneous release count)/(untreated CTL count - spontaneous release count)]. The median-effect log concentration and its corresponding 95% confidence limits were determined by plot extrapolation from the fa value of 0.5 on the y axis (50% effect dose) to the x axis (see Fig. ). The IC50 and its 95% confidence limits were then calculated from the anti-log of the values extrapolated from the fitted curve and the confidence band for fa. Determination of cell viability and total cellularity. | Cell viability and cell death were assessed using the trypan blue dye exclusion technique. FIG. 1. | CTL culturing protocol for preparing alloantigen-reactive CTL for analysis of their ZDV sensitivity in the IC50 assay. CTL culturing protocol for preparing alloantigen-reactive CTL for analysis of their ZDV sensitivity in the IC50 assay. FIG. 3. | Drug inhibition concentration (IC50) determination. Drug inhibition concentration (IC50) determination. To define the IC50 of ZDV, data points were fitted to a logistic model by nonlinear regression (see Materials and Methods). The log dose is the log10 of the ZDV concentration (micromolar) used in the assay. A fitted line was plotted on an x-y axis. fa was defined as the fraction of cytolysis affected by drug. The median-effect log concentration was determined from the fa value of 0.5 on the y axis (50% effect dose) to the x axis. The median-effect concentration and 95% confidence limits were then calculated from the anti-log of the extrapolated value. RESULTS : The IC50 assay can be used to determine the relative inhibitory activity of antiviral drugs on CTL sensitization. | An in vitro system to generate CTL was used to titrate the effects of ZDV on the CTL response. Figure illustrates the results of such an assay. A constant number of effector and stimulator cells in replicates of six were incubated with various concentrations of ZDV during antigen priming. As the ZDV concentration increased, the killing capacity of CTL cultures was reduced in a concentration-dependent sigmoid pattern. From these data, the IC50 value for ZDV inhibition of CTL activity was computed by determining the median effective concentration. Figure is an example showing a best-fit plot for determining the IC50 value. Naive CTL are more sensitive to ZDV inhibition than antigen-primed CTL given a second antigen exposure, but less sensitive than antigen-primed CTL held in culture without a second antigen stimulation. | In these experiments we compared the inhibitory effects of ZDV on naive versus double-prime or prime/hold effector cell cultures. Results of five independent experiments are shown in Table . Individual rows indicate experiments performed at the same time. The IC50 values from four independent trials using naive CTL ranged between 79 and 99 muM (mean, 87.5 muM) ZDV (Table , Arm 1). By comparison, the IC50 values for the four trials with double-prime CTL ranged between 62 and 638 muM (mean, 316 muM) ZDV (Table , Arm 2). However, when the antigen-primed C57BL6 CTL were cultured with syngeneic irradiated C57BL6 feeder cells (prime/hold) instead of the double-prime exposure, they were found to be profoundly sensitive to ZDV inhibition, with IC50 values ranging from 29 to 30 muM (mean, 29.6 muM) ZDV (Table , Arm 3). ZDV exposure of CTL during primary antigen stimulation does not select for a more ZDV-resistant CTL population. | The IC50 values of alloantigen-primed effector cells propagated in the presence of 80 muM ZDV were calculated to determine if ZDV treatment during antigen priming selected for a CTL population that was more resistant to the suppressive effects of ZDV. As previously observed , the number of cells collected at the end of the 14-day incubation period in the cultures incubated with ZDV were substantially less than those collected from cultures without ZDV (data not shown). The addition of ZDV to the culture system during the original antigen priming period did not select for a cell population with increased resistance to ZDV suppression. In each of the six paired trials where lymphocytes were primed in the presence of 80 muM ZDV (three in Arm 4 and three in Arm 5), the IC50 values were the same or less than the IC50 values for the lymphocytes primed in the absence of ZDV. It is clear from these assays that preincubation of T cells during alloantigen priming did not produce a more drug-resistant T-cell population. In summary, effector cells which were alloantigen primed for 14 days but not restimulated with DBA-2 alloantigen during the IC50 assay (incubated with the irradiated C57BL6 feeder cell layer) were most sensitive to ZDV, independent of the concentration of ZDV they encountered during the priming period. Unprimed lymphocytes ranked second in sensitivity, followed by alloantigen-primed and restimulated effector cells when treated with ZDV during the priming phase. Alloantigen-primed and restimulated effector cells not treated with ZDV during the priming phase were significantly less sensitive to ZDV suppression than unprimed or primed cells that did not reencounter the antigen. FIG. 2. | IC50 assay. IC50 assay. The influence of ZDV on generation of alloreactive CTL is shown in a representative assay using naive splenocytes. All wells in 96-well culture dishes received the same number of responder cells (C57BL6) and stimulator cells (irrDBA-2). ZDV was present in replicates of six at the indicated concentrations during the 7-day culture period. CTL were measured by 51Cr release. The upper horizontal line represents the mean total release, while the lower horizontal line represents mean spontaneous release from control cultures. TABLE 1 | Effect of alloantigen priming on the IC50 of ZDV-mediated reduction of CTL function DISCUSSION : Recognized limitations of ZDV antiviral therapy are the clinical side effects and the development of drug resistance . In addition, both in vitro and in vivo studies have documented that cells of the immune system are sensitive to therapeutic concentrations of ZDV . Evaluation of immunomodulation by antiviral drugs on lymphocyte effector cell functions may be crucial because drug-induced interference of immune function, already impaired by virus infection, may potentially reduce even further the capacity to respond to the infection . We previously demonstrated a dose-dependent inhibition of CTL clonal expansion by ZDV in an in vitro alloantigenic murine system . In the work described here, using the same alloantigen system, we have introduced a statistical method to determine the IC50 of ZDV on antigen-specific cytotoxic T-cell clonal expansion. Based on previous observations by ourselves and others, we hypothesized that the sensitivity of T lymphocytes to the suppressive effects of ZDV depends on the activation status of those cells (naive versus antigen primed). Our results revealed that 80 to 100 muM ZDV reduced the capacity of naive splenocytes to become CTL effector cells by one-half (mean IC50 = 87.5 muM). By contrast, antigen-primed lymphocytes with continuous exposure to allogeneic cells (double prime) were substantially more resistant to the suppressive effects of ZDV (mean IC50, 316 muM). Antigen-primed CTLs differ from their naive precursors in many regards, including morphology, surface antigen expression, and responsiveness to antigen reexposure and cytokine expression, e.g., IL-2 . It is known that antigen-activated murine CTL do not require additional contact with stimulatory antigen for continued clonal expansion but respond to mitogenic lymphokines for a defined time period after antigen contact . Therefore, substituting syngeneic irradiated C57BL6 splenocytes as feeder cells for alloantigenic DBA-2 splenocytes during the IC50 assay following antigen priming permitted evaluation of cytolytic precursors which became activated during the primary antigen encounter (first 14 days of culture). By contrast, cultures which were primed during the first 14 days and were reexposed to alloantigen during the inhibition assay allowed detection of both primed effector cells and antigen-specific precursors which were not stimulated during the first antigen exposure . With this in mind, we found that alloantigen-primed lymphocytes, reexposed to the same antigen (double prime), were less sensitive to the suppressive effects of ZDV (higher IC50) than primed cells not reexposed to antigen (prime/hold). Since the maintenance of antigen-primed but nonrestimulated effector cells is mainly dependent on IL-2 utilization , ZDV interference with either receptor expression or receptor-ligand interaction would be potential mechanisms explaining the observed difference in sensitivity. Exogenous IL-2 reverses or reduces the in vitro and in vivo suppressive effects of ZDV, consistent with this hypothesis . However, a separate study, which specifically evaluated cytokine production by lymphocytes from HIV-positive patients during ZDV treatment, did not reveal reduced IL-2 production , and another study reported a significant increase in IL-2 receptor (CD25) expression by mitogen-stimulated T lymphocytes from ZDV-treated AIDS patients . Therefore, even though excess exogenous IL-2 was provided in the tissue culture medium of our cultures during the antigen priming period and in the IC50 assay, we cannot exclude the possibility that ZDV interfered with cytokine-dependent T-cell-mediated helper function. We further found that 80 muM ZDV treatment of CTL during antigen priming had a drug concentration-related effect on the number of cells harvested after the 14-day priming period (data not shown) but did not affect the ZDV IC50 values of progeny cells compared to non-ZDV-treated cells. This observation was independent of the type of antigen (syngeneic or allogeneic) used during the culture period. We had previously shown that the concentration-response-related reduction of effector cells harvested following priming in the presence of ZDV was most likely mediated by reduced clonal expansion of alloantigen-stimulated effector cells . However, based on the present study, it appears that previous antigen exposure rendered the effector cell more sensitive to the suppressive effects of ZDV (prime/hold study). This sensitivity was altered by a second antigen encounter. Both in vitro and in vivo studies suggest that the initial antigen-priming period is the most sensitive to interference by ZDV and other nucleoside analogues . This later finding, however, suggests that the activational status of the cell exposed to ZDV is crucially important in terms of drug sensitivity and is independent of previous drug exposure. The double-primed cultures which were antigen primed in the presence of 80 muM ZDV (Table , Arm 4) did not develop greater resistance to ZDV than paired double-primed cultures without ZDV (Table , Arm 2). However, IC50 values from the double-prime ZDV data set (Table , Arm 4) were variable. It is likely that many of the precursor T cells capable of responding to DBA-2 antigen were suppressed by ZDV during the first antigen exposure period, such that they failed to respond and remained in an unprimed state equivalent to naive CTLs. Thus, a second round of antigen exposure may have stimulated a disproportionate number of naive T cells (Table , Arm 4) compared to the cultures which did not have ZDV during antigen priming (Table , Arm 2). A relatively larger number of naive compared to primed T cells responding in these cultures could reduce the IC50 value, since naive cells are more sensitive to ZDV than primed cells, as shown in Table , Arm 1. In addition, because the extent of primary antigen stimulation (percentage of precursors capable of responding) may vary from trial to trial, one might expect variability in IC50 values for double-prime cultures (Table , Arms 2 and 4). In the single-prime cultures (prime/hold), there was no opportunity to recruit additional naive cells, resulting in a more consistent response between either ZDV-treated or untreated cultures (Table , Arms 3 and 5). Further, because the cell concentrations were adjusted after primary antigen stimulation (with or without ZDV), both cultures were presumably the same in terms of the number of potential effector cells. In the subsequent 7-day culture period, both prime/hold and double-prime cultures received IL-2-containing medium, which promoted proliferation of primed cells but not naive cells. Because the single-primed effector cells (prime/hold) from both the ZDV-treated and untreated cultures gave essentially identical IC5o values (Table , Arms 3 and 5), it appeared that ZDV treatment during priming had no propensity to increase T-cell resistance to ZDV suppression. The mechanism by which ZDV suppresses CTL leading to variable IC50 values for naive, prime/hold, and double-prime CTL is not known. Based on previous studies, it is assumed that ZDV suppresses clonal expansion of antigen-stimulated CTL . ZDV has been shown to prolong mitosis , inhibit mitochondrial DNA polymerase , and deplete the TTP pool , all of which affect cell proliferation. In our previous studies , we found ZDV to be cytostatic at concentrations ranging between 15 and 250 muM and cytotoxic at concentrations of 500 muM or greater when tested on naive lymphocyte cultures. These results were gathered using the same C57BL6/DBA-2 alloantigen model . We also observed that a delay in the addition of ZDV reduced its suppressive effect in vitro and in vivo , allowing more effector cells to be produced. Tissue culture cell lines as well as peripheral blood mononuclear cells (PBMC) grown long-term in the presence of ZDV become resistant to ZDV antiviral activity by decreasing the concentration of thymidine kinase (TK), a crucial enzyme necessary for the phosphorylation of ZDV to ZDV-monophosphate (ZDV-MP). The failure of the phosphorylation cascade necessary to convert ZDV to its active anabolite, ZDV-triphosphate (ZDV-TP), renders the cells less sensitive to the antiviral activity of ZDV and less sensitive to the toxic effects of ZDV caused by the ZDV-MP intermediate . Reduced TK expression in ZDV-resistant Jurkat T cells has been linked to methylation of the human TK gene . The observation that PBMC from HIV-infected patients on long-term ZDV have reduced TK is suggestive that ZDV modulates TK expression . However, a more likely explanation is that ZDV therapy selects for a T-cell population of naturally low expressors of TK. Phytohemagglutinin-stimulated human PBMC have higher TK expression and greater sensitivity to ZDV inhibition of HIV-1 infection than nonstimulated controls . These cells are likely to be more sensitive to the cytostatic and cytotoxic effects of ZDV (low IC50). This latter observation is counter to our work, where highly activated T cells (double prime) had greater resistance to ZDV toxicity. An alternative explanation for increased resistance to ZDV by activated T cells is increased expression of the multidrug-resistant (MDR) transmembrane P-glycoprotein (p170), which acts as an efflux pump to remove intracellular ions, toxins, and drugs , rendering the cell resistant to the cytostatic and cytotoxic effects of ZDV . In normal human PBMC, the majority of CD8+ T cells, but less than half of the CD4+ T-cell population, express p170 . Interestingly, phytohemagglutinin stimulation of human PBMC causes an increase in p170 by the CD8+ T-cell subset, and anti-p170 blocks the cytolytic activity of alloantigen-specific cytotoxic T cells . A natural function of p170 may be in aiding the secretion of certain cytokines . Taken together, it appears that p170 expression is easily modulated by immune stimulation and may play an important role in cytotoxic T-cell effector cell function. The upregulation of p170 by antigen-stimulated T cells, as a mechanism for increased drug resistance, would be compatible with our double-prime stimulation results. Further studies will be needed to document the full range of p170 expression in the different activation stages of cytotoxic T cells and to correlate those results with drug sensitivity. Determining the relevance of this work to the plasma ZDV concentrations of humans on ZDV therapy is made difficult by the species differences in ZDV processing. Mouse lymphoid cells form ZDV-TP at a level 16 times higher than that of human lymphoid cells . Human lymphoid cells were reportedly 15 times more sensitive to the cytostatic effects of ZDV and dideoxycytosine than were mouse lymphoid cells . The high level of ZDV-MP formed in human lymphoid cells, which is known to inhibit thymidylate kinase and adenylate kinase , may account for the cytostatic effect of ZDV. Therefore, studies in mouse systems may underestimate the true cytostatic effect of ZDV treatment in human cells by a factor of 15. The concentration of ZDV found to suppress the CTL response of mouse splenocytes by 50% ranged between approximately 30 and >600 muM , while the peak ZDV concentration in patients receiving the recommended ZDV dosage has been calculated to be in the range of 3 to 7 muM. Assuming the 15-fold difference between mice and human cells, one might predict that human naive T cells would be suppressed by ZDV concentrations as low as 2 muM, well within the human peak plasma drug concentration. These assumptions, however, are based on predicted behavior of human lymphocytes and not actual T-cell response data. The IC50 assay used in a mouse alloantigen model will be a useful tool for testing the drug sensitivity of human T-cell responses to HIV antigens. Taken together, our studies show that ZDV reduced cytolytic effector cell function in a concentration-dependent manner and suggest that alloantigen-primed effector cells, when not given continuous exposure or rescued by a second antigen encounter, are more sensitive to ZDV suppression than naive cells. This observation may have relevance in persons with HIV, where a strong CTL immune response is critical for the prevention of disease progression . Administration of ZDV and possibly other nucleoside analogues during the time of initial antigen priming of T cells may reduce the peak CTL response to virus. Possible drug interference with immune function should be considered when determining drug dosage and the time point of ZDV treatment initiation. Backmatter: PMID- 12183247 TI - Characterization of Sparsomycin Resistance in Streptomyces sparsogenes AB - The antitumor antibiotic sparsomycin, produced by Streptomyces sparsogenes, is a universal translation inhibitor that blocks the peptide bond formation in ribosomes from all species. Sparsomycin-resistant strains were selected by transforming the sensitive Streptomyces lividans with an S. sparsogenes library. Resistance was linked to the presence of a plasmid containing an S. sparsogenes 5.9-kbp DNA insert. A restriction analysis of the insert traced down the resistance to a 3.6-kbp DNA fragment, which was sequenced. The analysis of the fragment nucleotide sequence together with the previous restriction data associate the resistance to srd, an open reading frame of 1,800 nucleotides. Ribosomes from S. sparsogenes and the S. lividans-resistant strains are equally sensitive to the inhibitor and bind the drug with similar affinity. Moreover, the drug was not modified by the resistant strains. However, resistant cells accumulated less antibiotic than the sensitive ones. In addition, membrane fractions from the resistant strains showed a higher capacity for binding the drug. The results indicate that resistance in the producer strain is not connected to either ribosome modification or drug inactivation, but it might be related to an alteration in the sparsomycin permeability barrier. Keywords: Introduction : The antitumor antibiotic sparsomycin is a universal translation inhibitor that blocks protein synthesis in all species (for a review, see reference ). The broad spectrum of sparsomycin action indicated that the drug was targeted to a highly conserved component of the translation machinery. The relevance of the target was also supported by the fact that mutations inducing high resistance to sparsomycin have not been reported, and only a moderately resistant strain has been found in Halobacterium salinarium . In fact, it was soon shown that sparsomycin blocks the peptide bond formation . The drug binds and causes important conformational changes in the peptidyl transferase active center. Thus, it was found that sparsomycin can block the binding of substrates at the A-site but enhances binding to the P-site . Actually, sparsomycin has been a very powerful tool in the study of the structure and function of the ribosome . Recently, the antibiotic was found to interact with nucleotide A2602 in the peptidyl transferase center of the bacterial ribosome . The drug was initially developed as a potential antitumor agent, although toxicity soon limited its clinical application . Nevertheless, the synthesis of a series of sparsomycin derivatives with higher inhibitory activities has led to a reappraisal of its potential as an anticancer drug . Sparsomyin is produced by Streptomyces sparsogenes, which is obviously resistant to the drug. Organisms producing toxic compounds, including antibiotics, use different approaches to defend themselves from their own action . Not the least frequent approach is to modify the target, making it insensitive to the drug. Thus, methylation of specific residues within the 16S rRNA makes the ribosomes of many aminoglycoside antibiotic producers resistant to their respective products . The same strategy is used by the producers of macrolides , lincosamides , pactamycin , and thiostrepton . The study of these resistance mechanisms has provided relevant information on the antibiotic mode of action and on the ribosomal binding site at the molecular level. It would be important, therefore, to see whether S. sparsogenes, like other antibiotic-producing streptomycetes, has managed to modify the highly conserved sparsomycin target site to make the ribosomes resistant to the drug. Alternatively, as in the case of other producers, a different resistance mechanism, such as drug inactivation or permeability barrier alterations, might have evolved (for a review, see reference ), perhaps because modification of the target is not possible without seriously affecting its activity. We have approached the study of sparsomycin resistance by directly analyzing the producer and by trying to characterize genetic determinants from S. sparsogenes that provide resistance in Streptomyces lividans, an organism susceptible to the drug. MATERIALS AND METHODS : Bacterial strains and culture media. | Streptomyces strains were S. sparsogenes ISP5356 (ATCC 25498), S. lividans 1326, and S. lividans 3131, which corresponds to S. lividans 1326 transformed with plasmid pIJ702 . Streptomyces spp. were grown in yeast extract-malt extract (YEME) liquid medium or on R5 agar plates . R2YE plates were used for the regeneration of protoplasts after transformation . Escherichia coli DH5alpha was used for manipulating plasmids. Growth conditions for E. coli were as described elsewhere . Sensitivity of Streptomyces spp. to antibiotics. | The sensitivities of the different strains of Streptomyces spp. to antibiotics were assayed either in liquid medium (YEME) or on R5 plates containing the amount of antibiotic indicated below. Inhibition in liquid medium was estimated by monitoring the A550 of the cultures. Inhibition on agar plates was determined either by colony counting or by the size of the halo formed in the plate around 3-mm-diameter paper disks containing the antibiotics at the indicated concentrations. Unless otherwise indicated, thiostrepton was not included in the media, to avoid possible effects on antibiotic resistance . DNA manipulation. | Restriction enzyme digestions, ligations, agarose gel electrophoresis, etc., were performed according to well-established techniques . Restriction endonucleases were purchased from Boehringer Mannheim, MBI Fermentas, New England Biolabs, and Amersham and were used as recommended by the suppliers. T4 DNA ligase, calf intestinal alkaline phosphatase, and the DNA polymerase I Klenow fragment were from Boehringer Mannheim. Standard procedures were used for propagation and subcloning of plasmids in E. coli. Nucleotide sequences of DNA inserts were determined on both strands using a Dye-Terminator cycle sequencing ready reaction kit (Applied Biosystems) with custom-made oligonucleotides as primers. Sequencing reactions were run on an automated DNA sequencer (model 377; Applied Biosystems). Preparation and screening of a genomic library of S. sparsogenes. | Genomic DNA from S. sparsogenes (60 mug), prepared according to the methods described in reference , was partially digested with Sau3A (0.03 U of enzyme per mug of DNA; 15 min at 37C). Six- to 8-kbp fragments were purified by agarose gel electrophoresis, and 10 mug of the sample was ligated to 2 mug of pIJ702 obtained from S. lividans 3131 and purified by cesium chloride-ethidium bromide gradient centrifugation. The plasmid was previously digested with BglII and treated with alkaline phosphatase to prevent recircularization. The ligation mixture was used to transform S. lividans 1326 protoplasts according to standard techniques . The screening of the library was done in two steps. First, the transformants were tested for a plasmid marker (resistance to the antibiotic thiostrepton). The thiostrepton-resistant colonies, numbering approximately 5,000, were allowed to sporulate and the spores were tested for resistance to sparsomycin on plates containing 100 mug of the antibiotic/ml but in the absence of thiostrepton, which has been reported to affect the sensitivity of S. lividans to different translation inhibitors, including sparsomycin . Preparation of cell extracts and ribosomes. | Total cell extracts (S30 and S100 fractions) and ribosomes from the different Streptomyces strains were prepared as previously described . Activity tests. | (i) Binding of 125I-labeled phenol-sparsomycin to ribosomes. Ribosomes (1.0 muM) were incubated for 30 min at 30C with increasing amounts of 125I-labeled phenol-sparsomycin (0.1 muM; 103 cpm/pmol) in 50 mul of binding buffer (10 mM Tris-HCl, 10 mM MgCl2, 60 mM NH4Cl, 6 mM beta-mercaptoethanol). After incubation the samples were diluted with 5 ml of binding buffer and filtered through nitrocellulose filters. After two washes with the same buffer, filters were counted in a gamma counter to estimate the amount of drug bound to ribosomes. (ii) Polyphenylalanine synthesis. | The polymerizing activity of the Streptomyces extracts was estimated by a poly(U)-dependent polyphenylalanine synthesis assay carried out as described elsewhere . Analysis of in vivo sparsomycin modification. | S. sparsogenes and S. lividans cells were grown to mid-exponential phase. Cells were collected by centrifugation (10 min at 10,000 rpm in an SS-34 rotor) and resuspended in binding buffer to an A550 of 5. A 200-mul aliquot of the suspension of cells was incubated with 100,000 cpm of 0.1 muM 125I-labeled phenol-alanine-sparsomycin (104 cpm/pmol) for 1 h at 30C. Cells were recovered by centrifugation, resuspended again in 200 mul of binding buffer, and broken. The antibiotic was extracted from both pellet and supernatant fractions by mixing them with an equal volume of ethyl acetate, and the organic and the aqueous phases were separated by centrifugation. The organic phase was collected and concentrated by evaporation. The radioactive antibiotic was analyzed by thin-layer chromatography on precoated silica gel plates, using chloroform-methanol (80:20 or 80:10) as a solvent. Accumulation of sparsomycin by the cells. | Streptomyces cells were grown until the A550 reached 0.2. Cells from 100 ml of medium were collected by centrifugation and resuspended in 4 ml of binding buffer, and 100 mul of the suspension was incubated with 100,000 cpm of 125I-labeled phenol-alanine-sparsomycin (104 cpm/pmol) for different periods. After the incubation, the cells were diluted with 5 ml of binding buffer and filtered through glass fiber filters. The filters were washed two times with 5 ml of the same buffer and dried, and the radioactivity was estimated by scintillation counting. Binding of radioactive sparsomycin to subcellular fractions. | The different Streptomyces strains were grown in 50 ml of medium to an A550 of 0.2, collected by centrifugation, washed with binding buffer, and resuspended in the same buffer to an A550 of 5.0. Cells were broken in a French press, and the extracts were centrifuged at 15,000 x g for 30 min to obtain a membrane fraction. Ribosomes were prepared by centrifuging the supernatant at 100,000 x g for 3 h. Aliquots containing equivalent amounts of the different fractions were incubated with 100,000 cpm of 125I-labeled phenol-alanine-sparsomycin (103 cpm/pmol) for 30 min at 30C. The bound drug was estimated as described previously. Nucleotide sequence accession number. | The sequence of the S. sparsogenes DNA fragment reported here was submitted to the EMBL nucleotide sequence database and assigned the accession number . RESULTS : Isolation of S. lividans sparsomycin-resistant strains. | S. lividans was transformed with an S. sparsogenes genomic DNA library in plasmid pIJ702 and prepared as described in Materials and Methods. Seven transformants were initially selected which were able to grow on agar plates containing 100 mug of sparsomycin/ml. The parental S. lividans was totally inhibited with 20 mug of drug/ml under similar growing conditions. When spores were prepared from resistant clones and allowed to grow again in the presence of sparsomycin, only one of them, SLT4, showed the same resistant phenotype, as well as conserving the thiostrepton resistance marker of the pIJ702 vector. S. lividans transformed with the empty pJI702 did not show any resistance to the drug in the absence of thiostrepton. The resistance of SLT4 to sparsomycin is associated with the presence of the transforming construct containing S. sparsogenes DNA, since curing the plasmid, which resulted in a loss of thiostrepton resistance, caused a concomitant loss of the sparsomycin resistance phenotype . Unexpectedly, the sparsomycin-resistant strain showed alterations in sensitivity to other protein synthesis inhibitors. Thus, S. lividans SLT4 shows a notable increase in its sensitivity to chloramphenicol and tetracycline while having a higher resistance to lincomycin and puromycin . Characterization of the resistance determinant. | The plasmid present in S. lividans SLT4 was isolated and characterized. A DNA insert of 5.9-kbp was found, which was fragmented using appropriate restriction sites; the fragments were subcloned in the same plasmid, pIJ702. The new plasmids were used to transform S. lividans 1326, and the transformants were tested for resistance to sparsomycin. In this way, a minimal fragment of 3.6 kbp (Fig. , insert 3) was found to be able to induce resistance to the drug. None of the S. sparsogenes DNA fragments subcloned in a pBluescript plasmid were able to provide sparsomycin resistance in E. coli. The fragment was sequenced, and the analysis of the sequence indicated the presence of two ORFs (srd and moxR) and a fragment of a third ORF encoding a transmembrane polypeptide (tmp) at one of the ends. Restriction analysis and subcloning of the original fragment (Fig. , insert 1) indicated that insert 4, comprising the middle ORF ---which showed homology to moxR ---and the incomplete tmp ORF did not induce resistance. In addition, disruption of srd by cutting insert 3 to obtain inserts 4 and 5 resulted in loss of resistance. It seems, therefore, that the sparsomycin resistance determinant, srd, must be associated with the first ORF in the sequenced fragment. The srd ORF carries GTG as a putative initiator codon and seven nucleotides at the 5' end of the fragment. It is probably expressed from a promoter in the vector and encodes a 601-amino-acid-long polypeptide . No significant homology of this S. sparsogenes DNA to any sequence in the data bank could be found. Similarly, no relevant standard functional or structural domain was detected in the polypeptide sequence, except for one weak transmembrane domain. Biochemical characterization of the resistance mechanism. | Three main resistance mechanisms are usually responsible for the resistance of cells to drugs: (i) reduction of the target affinity for the drug; (ii) inactivation of the drug; and (iii) alteration of the permeability barrier affecting the transport. In the first case, the reduction of the affinity can be due to either a direct alteration of the target or to the indirect action of an allosteric effector which increases the dissociation of the drug. (i) Reduction of the target affinity for the drug. | To explore the first possibility, the capacity of sparsomycin to inhibit a poly(U)-dependent polyphenylalanine synthesis cell-free system derived from S. sparsogenes, S. lividans SLT4, and the parental S. lividans 3131 was estimated. As shown in Fig. , all the extracts were equally sensitive to sparsomycin. Moreover, ribosomes from S. sparsogenes and from S. lividans, grown either in the absence (sparsomycin sensitive) or in the presence of thiostrepton (sparsomycin resistant) had a similar capacity to bind the drug . These results indicate that the resistance shown either by the producer or by the resistant S. lividans is not due to either a direct or an indirect effect on the affinity of the target, namely the ribosome, for sparsomycin. (ii) Modification of the drug by the resistant cells. | To test whether the resistant cells are able to inactivate the antibiotic by modifying its structure in some way, S. sparsogenes and S. lividans SLT4 cells were incubated with radioactive sparsomycin, 125I-labeled phenol-alanine-sparsomycin , for 1 h. The cells were broken afterwards and the extracts were centrifuged, yielding a supernatant and a pellet. Both fractions were treated with ethyl acetate to extract the drug, which was then analyzed by thin-layer chromatography using different solvents. No alteration of the drug mobility could be observed in any of the cases using untreated 125I-labeled phenol-alanine-sparsomycin as a control (data not shown). (iii) Accumulation of sparsomycin by the cells. | A possible alteration in sparsomycin transport in the resistant strains was checked by testing the capacity of the cells to accumulate the same radioactive derivative of sparsomycin. As shown in Fig. , the radioactivity found in S. sparsogenes and in S. lividans SLT4 was considerably lower than that found in the sensitive S. lividans 3131 strain. In order to characterize in more detail the different abilities to accumulate sparsomycin of the resistant and sensitive strains, cells were fractionated and the capacities of different fractions to bind radioactive drug were estimated. As summarized in Table , the membrane fractions derived from S. sparsogenes and SLT4 bound more radioactive drug than the equivalent fractions from sensitive cells. FIG. 1. | Susceptibility of S. lividans Susceptibility of S. lividans SLT4 to 90 mug (disk 1) and 60 mug (disk 2) of sparsomycin after (A) and before (B) the curing of plasmid pIJ702 containing the S. sparsogenes DNA insert. FIG. 2. | Effect of antibiotics on the growth of S. lividans Effect of antibiotics on the growth of S. lividans parental 3131 and SLT4 transformant strains. (A) E, erythromycin (60 mug); C, chloramphenicol (50 mug); S, sparsomycin (90 mug). (B) T, tetracycline (30 mug); L, lincomycin (60 mug); P, puromycin (150 mug). The antibiotics were placed on filter paper disks. FIG. 3. | Subcloning of the S. sparsogenes Subcloning of the S. sparsogenes DNA insert. Different fragments of the original insert in the plasmids isolated from S. lividans SLT4 were subcloned on pJI702 as indicated and tested for their capacity to induce sparsomycin resistance in wild-type S. lividans 1326. FIG. 4. | Analysis of the nucleotide sequence of the 3. Analysis of the nucleotide sequence of the 3.9-kbp fragment. Analysis of G+C content identified three putative ORFs in the sequence, marked srd, moxR, and tmp. The region included in the different inserts in Fig. is indicated at the bottom. A vertical line marks the starting point of the S. sparsogenes DNA. FIG. 5. | Nucleotide and deduced amino acid sequences of the srd ORF. Nucleotide and deduced amino acid sequences of the srd ORF. FIG. 6. | Sparsomycin inhibition of polyphenylalanine synthesis in cell extracts from S. sparsogenes Sparsomycin inhibition of polyphenylalanine synthesis in cell extracts from S. sparsogenes , wild-type S. lividans , and S. lividans SLT4 . Polymerizing activity was tested according to standard methods in the presence of the indicated concentrations of sparsomycin. Activity is shown as the percentage of the samples tested in the absence of drug. FIG. 7. | Binding of 125I-labeled phenol-sparsomycin to ribosomes from resistant and sensitive strains. Binding of 125I-labeled phenol-sparsomycin to ribosomes from resistant and sensitive strains. Ribosomes (50 pmol) from S. sparsogenes and S. lividans 3131 grown in the absence or in the presence of 0.5 mg of thiostrepton/ml were incubated with increasing amounts of 0.1 muM 125 I-labeled phenol-sparsomycin (103 cpm/pmol), and the amount of bound antibiotic was estimated by filtration. FIG. 8. | Accumulation of radioactive sparsomycin by sensitive and resistant Streptomyces cells. Cells from S. sparsogenes Accumulation of radioactive sparsomycin by sensitive and resistant Streptomyces cells. Cells from S. sparsogenes , S. lividans SLT4 (*), and S. lividans 3131 grown to mid-exponential phase were collected, washed, and resuspended in binding buffer to an A550 of 5.0. A total of 100,000 cpm of 125I-labeled phenol-sparsomycin was added and the samples were incubated at 30C. Cells were afterwards filtered through glass fiber filters and washed with binding buffer, and the radioactivity was estimated by scintillation counting. TABLE 1 | Binding of 125I-phenol-alanine-sparsomycin by different cellular fractions DISCUSSION : The results from the sparsomycin producer S. sparsogenes, as well as from the resistant S. lividans SLT4 expressing a resistant genetic determinant from the producer, clearly show that the ribosomes from both types of strains bind the drug to a similar extent. These data indicate that the resistance mechanism working in both strains does not involve modification of the drug target, the ribosomal peptidyl transferase center. Similarly, since the drug inhibits the in vitro protein-synthesizing activity of extracts derived from the resistant and sensitive strains equally, the action of an allosteric effector inducing a rapid dissociation of sparsomycin, like the Tet(M) class of proteins in tetracycline resistance , can also be excluded. Sparsomycin is, indeed, a very efficient inhibitor of the peptide bond formation in the Bacteria as well as in Archaea and in Eucarya. Affinity labeling data have shown that the antibiotic seems to bind to nucleotide A2602 in the 23S rRNA peptidyl transferase domain , which is very close to the active center, as revealed by the 2.4 A three-dimensional structure of the 50S ribosomal subunit recently reported . Sparsomycin is, therefore, interacting at a very critical site of the ribosome which, apparently, evolution has not been able to modify in order to produce a sparsomycin-resistant ribosome. In fact, the putative drug binding site, A2602, is a universally conserved nucleotide, which underlines its functional importance and accounts for the broad range of organisms sensitive to the drug. S. sparsogenes does not seem to resist exogenous sparsomycin by modifying the drug either, since radioactive drug incubated with the resistant cells was chromatographically indistinguishable from the untreated controls. Since the target is sensitive to sparsomycin and the drug is not inactivated by modification, the resistance mechanism in the resistant strains must lay at the level of the permeability barrier. Thus, it has been found that S. sparsogenes accumulates considerably less drug than the sensitive S. lividans cells when incubated in the presence of radioactive sparsomycin, suggesting that resistance probably results from differences in the cell permeability. It was possible to transform sparsomycin resistance in sensitive S. lividans cells by using an S. sparsogenes DNA library. The biochemical characteristics of the resistance phenotype in the transformant S. lividans SLT4 cells are similar to those found in the producer, namely, sparsomycin susceptibility of the translation machinery, inability to modify the drug, and reduced cellular accumulation of antibiotic. These results suggest, therefore, that resistance to sparsomycin in the producer might be associated with an alteration in the transport process. The fact that the sparsomycin resistance determinant also affects the sensitivity of the cells to several other inhibitors supports an effect at the level of permeability, although this hypothesis must be experimentally confirmed. An increasing number of antibiotic-producing organisms have been shown to contain active transport systems that efficiently export the antibiotic molecules to the exterior. The mechanisms can be divided into two classes. One of them is connected to proton-driven membrane electrochemical gradients and involves mdr-type proteins . The other one is based on the ABC transporter system, which couples antibiotic transport to ATP hydrolysis . The S. sparsogenes genetic determinant responsible for sparsomycin resistance in S. lividans SLT4 can be associated with srd, an ORF that encodes a 600-amino-acid polypeptide. However, no significant homology of sdr has been found to any sequence in either DNA or protein data banks, suggesting that the mechanism involved in sparsomycin transport is not related to either of the two systems so far described. Nevertheless, since S. lividans SLT4 accumulates sparsomycin at a somewhat higher level than the producer, it cannot be ruled out that the cloned determinant is only a part of the mechanism working in the producer. There is a novel drug-binding-protein-dependent export system that has been described in Streptomyces lavendulae, in which two components, mct, a putative mitomycin transport gene, and mrd, encoding a mitomycin-binding protein, work synergistically to confer resistance to mitomycin in the producer strain . Perhaps srd is only one part of a transport system present in S. sparsogenes. Interestingly, a membrane fraction of the resistant strain, S. lividans SLT4, binds more radioactive sparsomycin than the equivalent fraction from sensitive S. lividans, suggesting that srd might encode a sparsomycin-binding protein, as mrd does in the case of mitomycin . Srd, in spite of not being a typical membrane protein, might be an accessory component of the transport mechanism required for optimal efficiency. In any case, these results must be considered as preliminary, and further studies using the recombinant protein will be required to locate the protein in the cell and to obtain further insight into the real transport mechanism involved in sparsomycin resistance. Backmatter: PMID- 12183251 TI - Application of Real-Time PCR for Determination of Antiviral Drug Susceptibility of Herpes Simplex Virus AB - A quantitative real-time PCR (TaqMan) assay was developed for determination of antiviral drug susceptibility of herpes simplex virus (HSV). After short-time culture of the virus, the antiviral drug susceptibility of HSV isolates for acyclovir (ACV) was determined by measuring the reduction of the HSV type 1 (HSV-1) DNA levels in culture supernatants using real-time PCR. The 50% inhibitory concentration was reported as the concentration of antiviral drug that reduced the number of HSV-1 DNA copies by 50%. A total of 15 well-characterized ACV-sensitive or -resistant strains and clinical isolates were used for assay evaluation. The new assay with real-time PCR readout permitted rapid (3 days), objective, and reproducible determination of HSV-1 drug susceptibilities with no need for stringent control of initial multiplicity of infection. Furthermore, the real-time PCR assay results showed good correlation (r = 0.86) with those for the plaque reduction assay. In conclusion, the real-time PCR assay described here is a suitable quantitative method for determination of antiviral susceptibility of HSV-1, amenable for use in the routine diagnostic virology laboratory. Keywords: Introduction : Extensive use of acyclovir and other antiviral drugs for prophylaxis and treatment of herpes simplex virus (HSV) infections exerts a continuous selection pressure on the HSV virus population. HSV antiviral drug resistance occurs relatively frequently especially in immunocompromised patients such as those undergoing bone marrow (6 to 12%) or solid organ transplantation (similar4%) or AIDS patients (similar6%) and can be associated with serious disease (; B. Lina, D. Thouvenot, C. Danve, F. Morfin, A. Boucher, I. Berlin, and M. Aymard, abstract from European Virology 2000, J. Clin. Virol. 18:47, 2000). The frequency of HSV-resistant infections may increase because of the increasing number of severely immunocompromised patients with chronic or recurrent HSV infections who require prolonged administration of antiviral drugs. In this patient group susceptibility testing is needed to detect drug-resistant HSV strains and to reconsider the antiviral treatment (; F. Morfin, D. Thouvenot, G. Souillet, M. Michallet, and M. Aymard, abstract from Progress in Clinical Virology V, Acta Microbiol Immunol Hung. 46:429, 1999). Safrin et al. has shown a good correlation between the failure of HSV suppression by acyclovir (ACV) in vivo and the determination of ACV resistance in vitro. These observations emphasized the clinical relevance of antiviral resistance determination in the laboratory. Several phenotypic assays have been described, and some of them are used in clinical practice, with the plaque reduction assay (PRA) as the most frequently used drug susceptibility assay. Although this technique is laborious and time-consuming, it still remains the "gold standard" method by which other tests are evaluated . The majority of alternative susceptibility assays is based on reduction in cytopathic effect (CPE), which is either microscopically evaluated or colorimetrically detected . Assays based on enzyme-linked immunosorbent assay (ELISA) include the sandwich ELISA and the microplate in-situ ELISA (MISE-test) . The latter has been shown to correlate well with PRA. Other currently used antiviral susceptibility assays involve the use of DNA hybridization , flow cytometric analysis and transgenic HSV inducible reporter cells . With the increasing numbers of immunocompromised individuals, there is a need for the widespread routine availability of antiviral drug susceptibility assays, which would be rapid, reproducible and clinically relevant. Currently used methods, except for the MISE-test, suffer from certain pitfalls, which preclude their routine use. Most of the assays are time-consuming and labor-intensive; some may have subjective endpoints, require special equipment or trained laboratory personnel. Therefore we set out to develop an assay, which would overcome most of the aforementioned restrictions and could be easily implemented in the diagnostic laboratory. We describe the development and evaluation of a new approach for HSV-1 drug susceptibility determination using quantitative real-time PCR (TaqMan) to measure viral DNA production. MATERIALS AND METHODS : Cells and viruses. | Vero cells (African green monkey kidney) were propagated and maintained in Iscove's modified Dulbecco's medium (Life Technologies, Breda, The Netherlands) supplemented with 5% fetal bovine serum and gentamicin (10 mug/ml; Life Technologies, Breda, The Netherlands). A stock of HSV-1 strain McIntyre counted by electron microscopy (EM) was obtained from Advanced Biotechnologies. Inc., Columbia, Md. The HSV-1 strain KOS and the KOS-derived ACV-resistant mutants (AraAr7, AraAr8, AraAr13, F891C, PFAr5, and PAAr5) were kindly provided by D. M. Coen (Harvard Medical School, Boston, Mass.). The HSV-1 ACV-sensitive strains McIntyre and R39 were generously provided by A. Linde (Swedish Institute for Infectious Disease Control, Solna, Sweden). Well-characterized ACV-resistant clinical isolates HSV 98.25733-MA/3, HSV 98.15779-VA/2 and HSV 98.14742-PE/1 (Morfin et al., Acta Microbiol Immunol Hung. 46:429, 1999) were a gift of M. Aymard (Universite Claude Bernard, Lyon, France). Other HSV-1 clinical isolates were selected from our own collection; two originated from patients after bone marrow transplantation who had recurrent HSV infections not responding to ACV, and one was obtained from a patient with oral HSV infection which resolved spontaneously. Virus stocks were grown in Vero cells and the infectious titer was determined by plaque assay in Vero cells as previously described by Schaffer et al. . Real-time PCR assay for HSV-1: assay setup. | HSV-1-specific PCR primers and a fluorescent probe directed to the HSV-1 glycoprotein G (gG) gene were used for real-time PCR analysis as described by Ryncarz et al. . Each 25 mul of PCR mixture contained 7 mul of a 1:100-diluted culture supernatant, 900 nM concentrations of both forward and reverse primer, and 150 nM probe. Amplification was performed using the Applied Biosystems Sequence Detector 7700 under the following conditions: incubation for 2 min at 50C, and then for 10 min at 95C, followed by 45 cycles of 15 s at 95C and 1 min at 60C. Each PCR run contained two negative controls and a dilution series of HSV-1 DNA (6 x 102 to 6 x 108 copies/ml) derived from EM-counted virus stock (HSV-1, McIntyre), which was used to generate the standard curve. Each sample was analyzed in duplicate. Assay optimization. | The kinetics of HSV-1 DNA replication was examined by measuring the time course of increase in HSV-1 DNA yield in cell culture supernatants. Vero cells in 24-well culture plates were infected at a multiplicity of infection (MOI) of 0.1, 0.01, and 0.001 PFU/cell of HSV-1 strain McIntyre. The development of CPE was monitored, and the levels of HSV-1 DNA were measured by real-time PCR in supernatant samples collected at 12-h intervals after infection. Experiments were also performed to evaluate the effect of ACV in the cell culture supernatant on PCR efficiency. Virus infected cell cultures (MOI, 0.01) were incubated with or without a high concentration of ACV (48 mug/ml) for 48 h to resemble the conditions of the assay described here. Subsequently cell culture supernatants were collected and spiked with HSV-1 DNA. These spiked samples were amplified using the real-time PCR assay, either as undiluted supernatant or as a dilution series. The effect of the MOI on the ACV IC50s was determined in parallel experiments in which cell cultures were infected with a half-log10 incremental range of infectious doses (MOI, 0.001 to 0.5 PFU/cell) of HSV-1 in the presence of serial concentrations of ACV. The levels of HSV-1 DNA were measured by real-time PCR in supernatant samples collected at 24, 48, and 72 h postinfection, and the 50% inhibitory concentrations (IC50s) were determined. The CPE of the virus control was scored at the time of supernatant collection. Real-time PCR assay for HSV-1 antiviral susceptibility testing: final setup. | Virus isolates at an MOI of 0.01 (50 mul) were dispensed in duplicate into wells of the 24-well culture plate containing 450 mul of culture medium with different concentrations of ACV and a suspension of Vero cells (6 x 105 cells/ml). Serial twofold dilutions of ACV (Sigma, Zwijndrecht, The Netherlands) ranging from 0.06 to 32 mug/ml were used. Plates were incubated at 37C for 2 days and monitored for development of CPE. When complete CPE was reached in control wells, 300 mul of culture supernatants was collected, cleared by centrifugation (1,100 x g, 1 min, 4C), and examined in real-time PCR or stored at -70C until assayed. Reference ACV-susceptible (KOS) and ACV-resistant (AraAr8) strains were included as controls in each PCR assay. The IC50 was used to express virus drug susceptibility and was defined as the concentration of antiviral drug that reduced the number of DNA copies by 50% relative to the no-drug virus control. PRA. | The PRA for drug susceptibility determination was performed as previously described by Erlich et al. with minor modifications, using neutral red staining for plaque detection. Briefly, confluent Vero cell monolayers in 24-well culture plates were inoculated with 40 to 60 PFU of virus. After incubation at 37C for 1 h, the viral inoculum was replaced with culture medium containing various concentrations of ACV and 0.5% agar. The same ACV concentrations were used as in the real-time PCR assay. Each drug concentration was tested in quadruplicate. The plates were incubated at 37C for 2 to 3 days until plaques were observed in the control wells without the drug. Subsequently, the monolayers were stained overnight using a second overlay medium containing 0.08% neutral red in 0.8% agar. The same reference control strains were used in each PRA as were used in the in real-time PCR assay. The IC50 was defined as the ACV concentration that reduced the number of plaques by 50% compared to the untreated control wells. Isolates were considered resistant to ACV at IC50 of >=2 mug/ml. Statistical analysis. | Results of real-time PCR assay and PRA were analyzed and compared using Wilcoxon's signed rank test and Spearman's correlation coefficient. RESULTS : Our goal was to develop an easy-to-perform assay for HSV antiviral susceptibility testing, suitable for implementation into the modern routine diagnostic laboratory. To determine and optimize its characteristics, several parameters of the assay were studied in detail, such as the viral replication kinetics, the effect of ACV in culture supernatant on PCR efficiency, and the effect of MOI and incubation time on drug susceptibility values. Once the optimal format of the assay was set, the test was validated on a panel of well-characterized HSV-1 strains and clinical isolates. The sensitivity of the real-time TaqMan PCR to detect HSV DNA was evaluated by using serial dilution of HSV-1 DNA extracted from EM-counted virus stock (McIntyre). The quantification was linear over the range of concentrations examined, from at least 107 to 1,000 DNA copies per ml. Kinetics of HSV-1 DNA replication. | Complete CPE was observed at 36, 48, and 60 h after infection for the cultures infected at an MOI of 0.1, 0.01, and 0.001 PFU/cell, respectively. This corresponded to an HSV-1 DNA yield in the culture supernatant of approximately 9 log10 copies/ml. Culture wells showing less than 30% CPE were found to have DNA levels at the detection limit of the real-time PCR assay . Effect of ACV in culture supernatant on performance of real-time PCR. | Using a viral culture supernatant directly in a PCR analysis may introduce inhibitory substances in the PCR. Particularly the presence of ACV, which is a DNA polymerase inhibitor, in culture supernatants could inhibit the PCR . The presence of ACV at 48 mug/ml in the cell culture supernatants decreased the yield of amplified products approximately sixfold. The inhibitory effect of ACV was completely overcome by diluting the cell culture supernatant 100-fold in water prior to PCR, which was routinely done in later experiments. Effect of MOI and incubation time on IC50. | The IC50s of ACV in the TaqMan PCR assay at different MOIs and incubation times varied between 0.16 and 0.61 mug/ml . Only a slight increase in IC50s at higher MOIs was observed at each time point. At 24, 48, and 72 h of incubation, respectively, a maximum 1.8-, 3.2-, and 2.1-fold difference in IC50 was observed among the cultures infected with a large range of MOIs from 0.001 to 0.5 PFU/cell. For each MOI an increase in IC50 with incubation time was observed . The IC50s detected at 48 and 72 h were on average 2- and 2.4-fold higher than those determined at 24 h after infection. The IC50 results were comparable at each MOI regardless of incubation time as long as the susceptibility was determined when the CPE in the virus control was between 50 and <100%. The same findings were obtained for drug-resistant viruses (data not shown). Based on these results, an incubation time of 48 h and an MOI of 0.01 PFU/cell were subsequently selected for routine use. Antiviral susceptibility testing using real-time PCR assay. | The real-time PCR-based HSV-1 drug susceptibility assay was evaluated in parallel to PRA in a pilot study of nine well-characterized ACV-sensitive and resistant laboratory strains and six clinical isolates . ACV IC50s determined by real-time PCR assay correlated well with those from PRA (r = 0.86; P < 0.0001) . In addition, the observed fold differences in drug susceptibility between the reference strain KOS and each of the tested strains were highly comparable between the two assays (r = 0.99; P < 0.0001) . The absolute IC50s determined by the real-time PCR assay were significantly lower (P < 0.0001) than those from PRA for all ACV-sensitive and resistant laboratory strains and clinical isolates . The average difference in IC50s was 7.7-fold (range, 4.1- to 15.0-fold) between the two assays. Reproducibility. | The reproducibility of the cycle threshold (Ct) values used for IC50 calculation was assessed on two aliquots of the same supernatant samples collected from drug treated and drug control wells during ACV susceptibility testing of three random isolates. The mean intra-assay coefficient of variation calculated from replicate Ct values was 0.71% (range, 0.07 to 1.8%), indicating a high level of reproducibility. To assess the interexperimental variability, the IC50 of the HSV-1 strain KOS was determined in seven repeated experiments. This resulted in a mean IC50 of 0.15 mug/ml (range, 0.10 to 0.25 mug/ml; standard deviation = 0.06 mug/ml). FIG. 1. | Time course of changes in the yield of HSV-1 DNA after infection at different MOIs (PFU/cell). Time course of changes in the yield of HSV-1 DNA after infection at different MOIs (PFU/cell). The DNA levels were measured in culture supernatants by real-time PCR. The percent values represent the percent CPE observed in the wells and are indicated in the graph until the first time point that a 100% CPE was reached for each MOI. FIG. 2. | Fold differences in susceptibility (IC50) between the reference strain KOS and tested strains as determined by real-time PCR assay and PRA (r = 0. Fold differences in susceptibility (IC50) between the reference strain KOS and tested strains as determined by real-time PCR assay and PRA (r = 0.99). The trend line is aligned with the correlation coefficient 1.0. TABLE 1 | Effect of MOI and incubation time on ACV IC50 in Vero cells measured by real-time PCR assay (HSV-1 Mclntyre) TABLE 2 | ACV IC50s for HSV-1 strains determined by real-time PCR assay and plaque reduction assay DISCUSSION : The real-time PCR assay described here could be the basis for a useful novel readout system for antiviral drug susceptibility determination. The assay developed and evaluated for HSV-1 may be generally applicable to other viruses. The assay measures inhibition of HSV-1 DNA production by quantification of viral DNA using the TaqMan technology, whereas in classical PRA the reduction of numbers of virus induced plaques is used to determine the antiviral effect of the drug. Thus, both assays measure the effect of viral replication, though using different read-out parameters. The determination of these different parameters may explain the differences in absolute IC50s between the two assays. Moreover, PRA does not take into account the effect of antiviral agent on the plaque size. In PRA the antiviral effect of the drug is often manifested as a decrease in plaque size without complete prevention of plaque formation . Smaller plaques in drug-treated wells consist of lower numbers of virus-infected cells but are counted equally to plaques of normal size in control wells, which leads to overestimation of viral susceptibility. The real-time PCR assay, however, measures the true reduction of viral DNA production, which is the basic mechanism underlying the antiviral effect of the drug. As such, the real-time PCR assay may give more accurate estimation of the effect of the drug on viral replication. In the real-time PCR-based HSV-1 drug susceptibility assay, the effect of the MOI on the ACV susceptibility was limited, which was demonstrated by only small differences in IC50 among the cultures infected with a large range of MOIs (500-fold difference). The effect of the MOI was small as long as the virus had not infected all the cells. An incubation time of 48 h was routinely used in our assay. However, considering the reported differences in growth rates of clinical isolates, it cannot be excluded that longer incubation times will be needed for particular isolates to reach sufficient amount of CPE (50%) required for reproducible real-time PCR analysis. Therefore, rather than harvesting the virus at a fixed reading time, we would recommend monitoring CPE and subsequent susceptibility testing at CPE levels ranging from 50% to <100%. The real-time PCR assay was evaluated by testing 15 HSV-1 strains for ACV susceptibility and by comparing the results with those from the conventional PRA. The test showed good correlation with PRA on IC50s and also the fold differences in susceptibility between the reference sensitive and tested strains highly correlated. Based on the range of IC50s obtained for sensitive HSV strains and clinical isolates, a cutoff value of 0.3 mug/ml for ACV was considered as a discriminative concentration for sensitive and resistant strains in the real-time PCR assay. For a better and more accurate definition of the in vitro resistance threshold of this assay, a larger number of clinical isolates needs to be analyzed. Nevertheless, as the threshold values defining sensitive and resistant virus can differ depending on the assay utilized , the fold-differences in drug susceptibility compared to a reference strain may provide more relevant information for comparing results generated with different drug susceptibility assays . The real-time PCR assay described here allows the rapid determination of the ACV susceptibility of HSV strains. The test was only mildly affected by variation in the MOI, while quite accurate titration of the clinical isolate is required for PRA. The real-time PCR assay has an objective readout and a good reproducibility, furthermore it is more rapid and easier to perform than the PRA. Full susceptibility testing results from the real-time PCR assay were obtained within 3 days, in contrast to the usual 4 to 6 days required for PRA. This is a considerable improvement and in combination with the technology already available in many routine diagnostic laboratories may render it a useful test for the clinical virology laboratory. At present, real-time PCR-based assays are increasingly implemented into diagnostic clinical virology because of their high sensitivity, high throughput, and ease of use format . The real-time PCR assay described here uses the same PCR components and is performed under the same standard amplification conditions that are routinely used for detection of HSV in clinical specimens. Thus, the assay fits in well with methods already available in the clinical virology laboratory and as such it could be easily implemented in many clinical laboratories. In house availability of antiviral susceptibility testing would enable physicians to obtain results on drug susceptibility in a clinically useful time frame and may help explaining therapeutic failure in patients not responding adequately to treatment. In conclusion, we demonstrated the real-time PCR assay as a suitable method for the determination of antiviral drug susceptibility for HSV-1. Application of the assay for clinical practice needs to be further evaluated. Backmatter: PMID- 12183280 TI - Voriconazole Inhibition of the Metabolism of Tacrolimus in a Liver Transplant Recipient and in Human Liver Microsomes AB - The purpose of this study was to assess the effect of voriconazole on the blood tacrolimus concentration in a liver transplant recipient and to examine the interaction between voriconazole and tacrolimus by using human liver microsomes. Two subjects were enrolled in the clinical study: one received voriconazole, and the other received a placebo. Tacrolimus metabolism was evaluated in human liver microsomes at various concentrations in the absence and presence of various concentrations of voriconazole. Coadministration of voriconazole and tacrolimus resulted in elevated (nearly 10-fold-higher) trough tacrolimus blood concentrations in the liver transplant patient. In the in vitro study, voriconazole at a concentration of 10.4 +- 4.3 mug/ml inhibited the metabolism of tacrolimus by 50%. Clinically relevant concentrations of voriconazole inhibited the metabolism of tacrolimus in human liver microsomes. Close monitoring of the blood concentration and adjustment in the dose of tacrolimus are warranted in transplant recipients treated with voriconazole. Keywords: Introduction : Organ transplant patients are susceptible to invasive fungal infections that necessitate treatment with antifungal agents including the azoles . Azole antifungal agents, e.g., ketoconazole, itraconazole, and fluconazole, are known to inhibit the metabolism of immunosuppressive drugs such as cyclosporine and tacrolimus . Voriconazole is a triazole antifungal agent that is currently undergoing phase III clinical trials for the treatment of a variety of fungal infections. Voriconazole is 4- to 16-fold more active than fluconazole and 2- to 8-fold more active than itraconazole against Candida species, including C. krusei and C. glabrata . Voriconazole is active against a wide range of filamentous fungi including Aspergillus species. In a randomized trial, voriconazole was more effective than amphotericin B as primary therapy for the treatment of invasive aspergillosis (R. Herbrecht et al., Abstr. 41st Intersci. Conf. Antimicrob. Agents Chemother., abstr. J-680, p. 378, 2001). Voriconazole also appears to be a promising agent for the treatment of mycelial fungi that either are innately resistant or respond erratically to amphotericin B, such as hyaline molds and dematiaceous fungi. Availability in an intravenous and a highly bioavailable oral formulation renders voriconazole a potentially valuable drug for the treatment of invasive mycoses in transplant recipients. Antifungal agents are known to inhibit cytochrome P450 3A4/5 (CYP3A4/5) enzymes. CYP3A4/5 is also involved in the metabolism of cyclosporine, tacrolimus, and sirolimus . Preliminary observations indicate that voriconazole at a dose of 200 mg twice a day increases the trough concentrations in blood of cyclosporine in transplant patients (P. Ghahramani, A. J. Romero, A. F. Lant, and M. J. Allen, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 845, p. 24, 2000). We hypothesized that voriconazole will alter the hepatic metabolism of tacrolimus as well. The objectives of the present study were to evaluate the interaction between voriconazole and tacrolimus clinically in transplant patients and in vitro by using human liver microsomes. (These data were presented in part at the 41st Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, Ill., 16 to 19 December 2001.) Clinical study. : The clinical study was designed as an open-labeled, randomized, two-period, two-treatment, placebo-controlled pharmacokinetic study in liver transplant recipients. The study protocol was approved by the Institutional Review Board of the VA Medical Center. Informed consent was obtained from the study participants. Prior to enrollment, patients were to be stabilized for at least 1 week on their present dose of tacrolimus. The first patient was randomized to voriconazole, and the second patient was randomized to placebo. Concurrent medications in patient 1 included fludrocortisone, insulin, and magnesium antacids, and those in patient 2 included felodipine, fludrocortisone, trimethoprim-sulfamethoxazole, and magnesium antacids. Tacrolimus concentrations were measured by a microparticulate enzyme immunoassay (MEIA) . Due to a significant increase in the trough tacrolimus blood concentrations in the first patient during the first few days after administration of voriconazole, the study was discontinued. Table describes the dosing regimen of voriconazole and tacrolimus and the concentrations of tacrolimus in blood in the two subjects who participated in the study. The concentration of tacrolimus increased nearly 10-fold in the patient who received voriconazole. Voriconazole was discontinued and tacrolimus was withheld until the levels in blood returned to baseline values. TABLE 1 | Clinical pharmacokinetic study summary Human liver microsome study. : Human liver microsomes are commonly used to assess drug metabolism and drug-drug interactions. A human liver microsome was prepared by differential centrifugation from liver that was not used for organ transplantation. To evaluate the 50% inhibitory concentration (IC50) of voriconazole, tacrolimus (50 ng/ml) and voriconazole (0 to 100 mug/ml) were incubated with microsomal protein (0.02 to 0.08 mg/ml) in the presence of MgCl2 (10 mM) for 10 min at 37C in 0.1 mM phosphate buffer (pH 7.4). The metabolism of tacrolimus was initiated with the addition of NADPH (1 mM). After 15 min, the reaction was terminated by placing the tubes into ice. Tacrolimus concentrations were measured immediately by a minor modification of the MEIA method used for blood samples. For the calculation of Vmax (maximal velocity of tacrolimus metabolism) and Km values (the affinity of tacrolimus for the enzyme), tacrolimus was incubated at various concentrations (0 to 20 muM) in the absence and in the presence of 200 mug of voriconazole/ml. Tacrolimus with and without voriconazole was incubated with a human liver microsome at a protein concentration of 0.4 mg/ml, in the presence of MgCl2 (10 mM) for 10 min at 37C in 0.1 mM phosphate buffer (pH 7.4). The metabolism of tacrolimus was initiated with the addition of NADPH (1 mM). After 30 min, the reaction was terminated by placing the tubes into ice. The tacrolimus concentration was measured by high-pressure liquid chromatography (HPLC). For HPLC analysis, 5 ml of cold ethyl ether was added to the cooled microsome. Cyclosporine (50 mul of a 0.1-mg/ml concentration in methanol) was added as the internal standard. Tacrolimus and cyclosporine were extracted into ethyl ether, and the ether layer was separated and evaporated under nitrogen. The residue obtained was reconstituted in acetonitrile-H2O (6:4) for HPLC analysis. The change in the concentration of tacrolimus was quantified by reverse-phase HPLC. A C18 column (3.9 by 150 mm; 10 mum; Bondapack; part no. 86684) was equipped with a C18 guard column and maintained at 70C. A mobile phase consisting of acetonitrile-water (6:4) at a flow rate of 1.5 ml/min was used. Tacrolimus eluted at 4.5 min, and cyclosporine eluted at 7.0 min. The column eluent was monitored at 214 nm. The concentration of voriconazole necessary to inhibit the metabolism of tacrolimus by 50% was calculated as the IC50 . IC50, Vmax, and Km were calculated with Prism software (GraphPad Software Inc.). In three different sets of microsomes, the mean (+- standard deviation) concentration of voriconazole needed to inhibit the metabolism of tacrolimus by 50% was 10.4 +- 4.3 mug/ml. The Vmax and Km for tacrolimus metabolism were 1.47 nmol/min/mg of protein and 10.64 muM in the absence and 0.37 nmol/min/mg of protein and 5.43 muM in the presence of voriconazole, respectively. This suggests that voriconazole inhibits the metabolism of tacrolimus by competitive and noncompetitive mechanisms. The in vitro study qualitatively predicted the in vivo interaction between tacrolimus and voriconazole. However, the magnitude of the observed in vivo inhibition of tacrolimus metabolism appeared to be much greater than predicted based on the in vitro studies using concentrations of voriconazole comparable to those documented clinically in the serum. These data suggest that voriconazole may be present in higher concentrations in the gut and may inhibit gut metabolism of tacrolimus, thereby increasing the oral bioavailability of tacrolimus beyond its inhibitory effect on the hepatic metabolism of tacrolimus, as has been shown elsewhere for the interaction between ketoconazole and cyclosporine-tacrolimus . In conclusion, coadministration of voriconazole and tacrolimus resulted in a significant increase in the trough concentrations of tacrolimus in blood. The in vitro human liver microsomal study documented that, at clinically relevant concentrations of voriconazole, the metabolism of tacrolimus was inhibited. Monitoring of blood levels of tacrolimus and adjustment in its dosage are warranted in transplant patients receiving voriconazole. FIG. 1. | Effect of voriconazole on tacrolimus metabolism. Effect of voriconazole on tacrolimus metabolism. Tacrolimus (50 ng/ml) was incubated in the presence of various concentrations of voriconazole (0 to 100 mug/ml). The amount of tacrolimus metabolized per time in the absence and in the presence of various concentrations of voriconazole was measured by MEIA. At a concentration of 6.3 mug/ml, voriconazole inhibits the metabolism of tacrolimus by 50% in one microsome sample. FIG. 2. | Metabolism of tacrolimus in the presence of voriconazole. Metabolism of tacrolimus in the presence of voriconazole. Tacrolimus (0 to 20 muM) was incubated in the absence (*) or in the presence of 200 mug of voriconazole/ml, and the amount of tacrolimus metabolized was estimated by HPLC. The figure shows that voriconazole decreases the maximal velocity of tacrolimus metabolism (Vmax, 1.47 to 0.37 nmol/min/mg of protein) and changes the affinity of tacrolimus for the enzyme (Km, 10.64 to 5.43 muM). Backmatter: PMID- 12183225 TI - N-Alkyl Urea Hydroxamic Acids as a New Class of Peptide Deformylase Inhibitors with Antibacterial Activity AB - Peptide deformylase (PDF) is a prokaryotic metalloenzyme that is essential for bacterial growth and is a new target for the development of antibacterial agents. All previously reported PDF inhibitors with sufficient antibacterial activity share the structural feature of a 2-substituted alkanoyl at the P1' site. Using a combination of iterative parallel synthesis and traditional medicinal chemistry, we have identified a new class of PDF inhibitors with N-alkyl urea at the P1' site. Compounds with MICs of <=4 mug/ml against gram-positive and gram-negative pathogens, including Staphylococcusaureus, Streptococcuspneumoniae, and Haemophilusinfluenzae, have been identified. The concentrations needed to inhibit 50% of enzyme activity (IC50s) for Escherichiacoli Ni-PDF were <=0.1 muM, demonstrating the specificity of the inhibitors. In addition, these compounds were very selective for PDF, with IC50s of consistently >200 muM for matrilysin and other mammalian metalloproteases. Structure-activity relationship analysis identified preferred substitutions resulting in improved potency and decreased cytotoxity. One of the compounds (VRC4307) was cocrystallized with PDF, and the enzyme-inhibitor structure was determined at a resolution of 1.7 A. This structural information indicated that the urea compounds adopt a binding position similar to that previously determined for succinate hydroxamates. Two compounds, VRC4232 and VRC4307, displayed in vivo efficacy in a mouse protection assay, with 50% protective doses of 30.8 and 17.9 mg/kg of body weight, respectively. These N-alkyl urea hydroxamic acids provide a starting point for identifying new PDF inhibitors that can serve as antimicrobial agents. Keywords: Introduction : A largely empirical search for new antibiotics, carried out since the 1940s, has led to an impressive armamentarium of clinically useful antibacterial drugs. However, the actual number of targets interfered with by these agents is very limited. Cross-resistance for antibiotics, resulting from a sharing of the same target, plays an important role in the worsening problem of resistance. It is not surprising that antibacterial drug discovery efforts are now focusing on identifying molecules that block different targets and thus can inhibit the growth of resistant bacteria. Analysis of microbial genomes has revealed an abundance of novel and potentially useful targets , but little has so far resulted from this much heralded effort. On the other hand, many bacterial enzymes that have been well characterized hold promise for the discovery of new antibacterial drugs. One such target that has recently received a great deal of attention is peptide deformylase (PDF; EC 3.5.1.31). Protein synthesis has proven to be a rich source of targets for antibacterial drugs . Many of the known antibiotics target one or more steps of this complex process, including the aminoglycosides, macrolides, tetracyclines, and oxazolidinones. Although the protein-synthesizing machineries of bacterial and mammalian cells exhibit overall similarity, there are sufficient differences to allow for selective blocking of the process in bacteria. One significant difference is the transformylation and subsequent deformylation of the initiating methionine in bacterial translation . Unlike cytosolic protein synthesis in mammalian cells, protein synthesis in bacteria is initiated with N-formylmethionine , which is generated by enzymatic transformylation of methionyl-tRNA. In prokaryotes, the N-formylmethionine of the nascent protein is removed by the sequential action of PDF and a methionine aminopeptidase to afford the mature protein . This bacterium-specific requirement for PDF in protein synthesis provides a rational basis for selectivity, making it an attractive drug discovery target. The possible use of PDF as an antimicrobial target was recently reviewed elsewhere . PDF activity was first reported by Adams in 1968 . However, further attempts to purify the enzyme from cell lysates failed because the activity was not stable. The enzyme was not characterized further until the early 1990s, when the deformylase gene, def, was cloned and PDF was subsequently overexpressed in Escherichiacoli . Bacterial PDF belongs to a new class of metallohydrolases that utilize an Fe2+ ion as the catalytic metal ion . The ferrous ion in PDF is very unstable and can be quickly and irreversibly oxidized to the ferric ion, resulting in an inactive enzyme . Interestingly, the ferrous ion can be replaced with a nickel ion in vitro, resulting in much greater enzyme stability with little loss of enzyme activity . The three-dimensional structures of various PDF molecules, including structures of enzyme-inhibitor complexes, have been solved and published . Although PDF is a ferrous aminopeptidase with a primary sequence very different from those of other metalloproteases, it has been noted that the environment surrounding the catalytic metal ion of PDF appears to be very similar to the active sites of thermolysin and the matrix metalloproteases (MMPs) . The catalytic metal ion of PDF is tetrahedrally coordinated with two histidines from the conserved zinc hydrolase sequence, HEXXH, and a conserved cysteine from an EGCLS motif. A water molecule that presumably hydrolyzes the amide bond occupies the fourth position in the tetrahedron. The fact that PDF is a metalloprotease makes the enzyme a more attractive target for drug discovery. Metalloproteases are among the best studied of the enzyme classes , and there are excellent precedents for the mechanism-based design of their inhibitors. In the last few years, several classes of PDF inhibitors have been reported (, -, , , , , , ). While all of these compounds inhibit PDF activity, most of them do not have antibacterial activity, presumably due to weak potency against PDF and/or an inability to penetrate the bacterial cell. It is interesting that among these compounds, those for which the concentrations needed to inhibit 50% of enzyme activity (IC50s) (or Kis) were greater than 1 muM had no antibacterial activity. Based on mechanistic and structural information, together with an understanding of the general principles of inhibiting metalloproteases, a generic PDF inhibitor structure was proposed. In this structure, X represents a chelating pharmacophore that will be the major component to provide binding energy; the n-butyl group mimics the methionine side chain; and P2' and P3' are regions of the inhibitor that can provide additional binding energy, selectivity, and favorable pharmacokinetic properties. Recently, actinonin, a naturally occurring antibiotic that was first isolated in 1962 from an actinomycete , was shown to be a PDF inhibitor . In this study, we report that N-alkyl urea hydroxamic acids, which also fit this generic structure, represent a new class of PDF inhibitors. The structural characterization and biological evaluation of this class of inhibitors are presented here. FIG. 1. | Generic PDF inhibitor structure derived from the transition state of the deformylation reaction . Generic PDF inhibitor structure derived from the transition state of the deformylation reaction . MATERIALS AND METHODS : Materials. | Actinonin, formate dehydrogenase (FDH), catalase, and NAD+ were obtained from Sigma; E. coli Ni-PDF and Streptococcuspneumoniae Zn-PDF were overproduced and purified as previously described . N-formylmethionine-alanine-serine (fMAS) and thio ester peptide Ac-ProLeuGly-S-LeuLeuGly-OC2H5 were obtained from Bachem. Matrilysin (MMP-7) and angiotensin-converting enzyme (ACE) were obtained from Calbiochem. Hydroxypropyl-beta-cyclodextrin was purchased from Aldrich. All other chemicals were of the highest commercial grade. Preparation of N-alkyl urea hydroxamic acids. | The method of preparation of the urea compounds was recently reported (J. Lewis, J. Jacobs, C. Wu, C. Hackbarth, W. Wang, S. Lopez, R. White, J. Trias, Z. Yuan, and D. Patel, Abstr. 41st Intersci. Conf. Antimicrob. Agents Chemother., abstr. 358, 2001) and will be published elsewhere. Briefly, the urea hydroxamic acids (compound 4) were prepared in eight steps from commercially available amino acid precursors as outlined in Fig. . The synthetic sequence used Fukuyama-Mitsunobu chemistry followed by thiolytic deprotection to provide N-mono-alkyl-glycine methyl esters (compound 1). Acylation of N-alkyl amino esters with excess phosgene under aqueous reaction conditions provided N-alkyl-N-chlorocarbamoyl-glycine methyl esters (compound 2), which were subsequently acylated with the N-alkyl-N-chlorocarbamoyl-glycine methyl esters in pyridine to produce tetrasubstituted ureas (compound 3). Trifluoroacetic acid (TFA) deprotection followed by coupling of the R2 amine with PyBop peptide coupling reagent and treatment with hydroxylamine provided the final product, urea hydroxamic acid PDF inhibitors (compound 4). Enzyme assays. | All absorption measurements were obtained with a SpectraMax plate reader (Molecular Devices). Deformylase activity was assayed by a PDF-FDH coupled assay as previously described . Briefly, the assay was carried out at room temperature with 5 nM E. coli Ni-PDF or 10 nM S. pneumoniae Zn-PDF in a buffer consisting of 50 mM HEPES (pH 7.2), 10 mM NaCl, and 0.2 mg of bovine serum albumin/ml in half-area 96-well microtiter plates (Corning). The reaction was initiated by the addition of a reaction mixture of 0.5 U of FDH/ml, 1 mM NAD+, and 4 mM fMAS at the desired concentration. To determine the IC50s of the desired compounds, PDF was preincubated for 10 min with various concentrations of test compounds prior to the addition of the reaction mixture. The initial reaction velocity was measured as the initial rate of increase in the absorption at 340 nm. Matrilysin (MMP-7) activity was assayed as reported previously by using a thio ester peptide as a substrate, with some modifications. Briefly, 0.12 mug of MMP-7/ml was preincubated at room temperature for 10 min with test compounds at various concentrations in a buffer containing 50 mM Tricine (pH 7.5), 0.2 M NaCl, 10 mM CaCl2, and 0.05% Brij. The reaction was initiated by the addition of 0.05 mM thio ester peptide substrate (Ac-ProLeuGly-S-LeuLeuGly-OC2H5) and 0.1 mM 5,5'-dithio-bis(2-nitrobenzoic acid). Reaction progress was monitored by recording the increase in the absorption at 405 nm. ACE activity was determined with a 96-well format according to the procedure reported by Maclean et al. . The hydrolysis product of the enzyme reaction was detected by derivatization with o-phthaldialdehyde reagent (Pierce) by following the manufacturer's protocol. Briefly, 0.03 U of ACE/ml was incubated with test compounds at different concentrations at room temperature for 10 min in a buffer consisting of 25 mM HEPES (pH 8.2) and 0.3 M NaCl. The reaction was initiated when 2 mM substrate (hippuryl-HisLeu) was added to the mixture. The enzyme reaction was carried out at 35C for 1 h. An equal volume of o-phthaldialdehyde reagent was added, and the reaction signal was detected by recording the change in fluorescence at excitation and emission wavelengths of 360 and 465 nm, respectively. The IC50 was calculated with the following equation: In this equation, yo is enzyme activity in the absence of inhibitor, and [In] is the inhibitor concentration. All data fitting was carried out by using nonlinear least-squares regression with the commercial software package DeltaGraph 4.0 (Deltapoint, Inc). Cytotoxicity assays. | The cytotoxicities of the test compounds were assessed by using human K562 (ATCC CCL-243) and murine P388D1 (ATCC CCL-46) leukemia cell lines. The human cell line K562 was maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum (Gibco BRL) and 1 mM sodium pyruvate. P388D1 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% bovine calf serum (Gibco BRL). The assays were conducted with 96-well microtiter plates (Corning), and test compounds were serially diluted in 10% dimethyl sulfoxide. A volume of 10 mul of each dilution was added to wells 1 to 11 in each row; well 12, used as a control, contained 10 mul of 10% dimethyl sulfoxide solution without drug; and well 12H contained 0.25 mug of puromycin/ml as a no-growth control. Ninety microliters of log-phase cells (5.5 x 104 cells/ml) was suspended in the appropriate assay medium and added to each well. The cells were exposed to the drug for 3 days at 37C in the presence of 5% CO2. On day 4, an indicator solution containing 1 mg of XTT/ml and 7.7 mug of phenazine methosulfate (Sigma)/ml in phosphate-buffered saline was added to each well, and the suspension was reincubated for 4 h under the same conditions. The XTT cleavage product was detected by recording the change in the absorption at 450 nm, and cell growth as a percentage of that in the corresponding control well was used to calculate the IC50 with equation 1. Susceptibility studies and killing curves. | Bacteria used in these studies were stored at -80C and grown at 35C in accordance with NCCLS recommendations . For susceptibility tests, MICs were determined by the broth microdilution method in accordance with NCCLS guidelines but with the following modification. Because of the high frequency of resistance previously reported with other PDF inhibitors , a modified bacterial inoculum size of 0.5 x 105 to 1.0 x 105 CFU/ml was used to determine the relative activities of the compounds. The organisms used are all part of the Versicor strain collection . The MIC was defined as the lowest concentration that yielded no visible growth after 24 h of incubation at 35C; end points were obtained by measuring the optical density at 600 nm (OD600). Minimum bactericidal concentrations (MBCs) were determined for Haemophilusinfluenzae (ATCC 31517; VHIN1004) and S. pneumoniae (ATCC 49219; VSPN1001) . The MBC was defined as the lowest concentration of antibiotic that resulted in a >=3-log10 decrease in the bacterial titer. Time-kill curves were determined by using log-phase cultures of H. influenzae (ATCC 31517), S. pneumoniae (ATCC 49219), and Staphylococcusaureus (ATCC 25923; VSAU1003). The bacteria were exposed to 5 or 10 times the MIC of each compound tested. Cultures were incubated at 35C, and titers were quantitatively determined at 0, 2, 6, 8, and 24 h of incubation. Results were expressed as log10 CFU per milliliter versus time. Expression and purification of E. coli PDF used for crystallographic studies. | The E. colidef gene was cloned for overexpression in E. coli BL21(DE3)/pLysS as described previously . Fermentation was performed by using a Bioflo 2000 fermentor (New Brunswick Scientific) with an on-demand fed-batch strategy and minimal medium. A 4-liter fermentation was maintained at 37C until the culture reached an OD600 of similar8.9. The temperature was then lowered to 18C, and the culture was induced overnight with isopropyl-beta-d-galactopyranoside (IPTG) at a final concentration of 0.4 mM. The bacterial cells (final OD600, similar40) were harvested by centrifugation and yielded similar240 g of paste (wet weight). The cells were frozen as pellets and stored at -80C. All purification steps were performed at 4C unless noted otherwise. Aliquots of cells (similar50 g) were thawed briefly at 37C and resuspended in 250 ml of lysis buffer (20 mM morpholineethanesulfonic acid [MES]-KOH [pH 6.5], 100 mM KCl, 5 mM NiSO4, 5 mM MgSO4, 10 mug of catalase/ml, 100 muM phenylmethylsulfonyl fluoride, 10.0 mug of DNase I/ml, five tablets of complete protease inhibitor cocktail without EDTA [Boehringer Mannheim]). The cells were lysed with four freeze-thaw cycles and then sonicated (four 10-s bursts, full power) with a Virsonic 60 apparatus (Virtus) equipped with a titanium horn. The pellets were collected at 35,000 x g and washed by resuspension in 120 ml of lysis buffer. The supernatants were combined and clarified by centrifugation at 100,000 x g for 30 min. DNA was removed by centrifugation at 35,000 x g following a 15-min treatment with 0.1% polyethyleneimine (final concentration). The supernatant was buffer exchanged with diafiltration against a 5-kDa-cutoff membrane in 20 mM morpholinepropanesulfonic acid (MOPS)-KOH (pH 7.7)-50 mM KCl-1 mM NiSO4 and filtered through a 0.2-mum-pore-size filter. The sample was then loaded at a flow rate of 30 ml/min onto a POROS 20HQ (perfusion chromatography medium; PE Biosystems) column (2 by 16 cm) equilibrated with 20 mM MOPS-KOH (pH 7.7)-1 mM NiSO4 at room temperature. A linear salt gradient, 0 to 300 mM KCl in 20 column volumes, was used to elute bound proteins, and the fractions were immediately placed at 4C. Fractions containing Ni-PDF were identified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis, pooled, and further purified by size exclusion chromatography on a Superdex 75 HiLoad (16/60) column (Amersham Pharmacia Biotech) equilibrated with 20 mM MOPS-KOH (pH 7.7)-100 mM KCl at a flow rate of 1 ml/min. Proteins were pooled and concentrated to similar5 to 10 mg/ml. After the addition of glycerol to 10% and 1 mM Tris-(2-carboxyethyl)-phosphine (TCEP) hydrochloride, the enzyme was stored at -80C. The protein concentration was determined with Bradford reagent (Bio-Rad) and bovine serum albumin as a standard. This highly purified Ni-PDF was used directly for crystallization. Crystallization and structure determination. | An enzyme-inhibitor complex was formed by incubation of a twofold molar excess of PDF inhibitor VRC4307 with 1.5 mg of PDF/ml in binding buffer (20 mM HEPES [pH 7.5], 100 mM NaCl, 0.5 mM TCEP) for 30 min at room temperature. The complex was buffer exchanged four times in binding buffer with 50 muM compound VRC4307 by using a Millipore concentrator (5,000-molecular-weight cutoff) and finally concentrated to 36 mg/ml. Crystals were grown by the hanging-drop vapor diffusion method by mixing equal volumes of protein solution and well buffer [100 mM HEPES (pH 7.0), 2.0 M (NH4)2SO4] at 17C. Crystals were cryoprotected at room temperature by a 15-min, 10-step transfer to cryobuffer solution [100 mM HEPES (pH 7.0), 2.4 M (NH4)2SO4, 25% glycerol] with 5 muM compound VRC4307 and immediately flash-frozen in a 100 K nitrogen gas cold stream (model 700 apparatus; Oxford Cryosystems) for data collection. Diffraction data were collected by using an R-AXIS-IIc image-plate system (Molecular Structures Corp. [MSC]) with Osmic Confocal Max-Flux optics (Blue-3 configuration; Rigaku/MSC) and an RU-2HR generator (MSC) operating at 100 mA and 50 kV. A complete data set was collected from a single crystal (0.2 by 0.18 by 0.06 mm) maintained at 100 K. The images were indexed, and integration was performed by using Denzo/Scalepack . The structure of the enzyme or enzyme complex was determined by molecular replacement with AMoRe as implemented in the CCP4 package with a structure of E. coli PDF (e.g., 1ICJ) as a search model. The structure was refined with the CNX 2000 package (Accelrys Inc.). Isolation of resistant mutants of H. influenzae and S. pneumoniae. | Spontaneous resistant mutants were selected as previously described . Briefly, 1010 CFU of H. influenzae (ATCC 31517) were plated on chocolate agar plates containing 30 mug of antibiotic/ml. In a second experiment, approximately 1010 CFU of log-phase S. pneumoniae (ATCC 49219) were spread on tryptic soy agar (TSA)-5% sheep blood agar plates containing 30 mug of test compound/ml. The plates were incubated for 48 h at 35C in a 10% CO2 atmosphere. Colonies were picked and transferred to drug-free plates; stable mutants that survived this passage were subjected to further characterization. Molecular techniques and sequence analysis. | Primary sequences for the def and fmt homologs of H. influenzae and S. pneumoniae were obtained from public sequence databases and previously published work . The def-fmt operon and flanking DNA (230 bp upstream; 60 bp downstream) from wild-type H. influenzae VHIN1004 and a single PDF inhibitor-resistant derivative were amplified by PCR with respective upstream and downstream primers oPV743 (5'-CGCCAGGTTACCAAATACTTATCTAAG-3') and oPV378 (5'-TTTTCCGCTATAAATTTACCG-3') for def and oPV177 (5'-GCCAGCGCATTAAAGAAAAGTTG-3') and oPV179 (5'-CGAAAGTGCGGTCATTTTTACTG-3') for fmt. The defA, defB, and fmt genes and flanking DNAs (approximately 120, 350, and 60 bp upstream, respectively; 60 bp downstream for all three) from wild-type S. pneumoniae ATCC 49619 and two PDF inhibitor-resistant derivatives were amplified by PCR with the respective upstream and downstream primers oPV369 (5'-CGTGGCTTGGAAGTAACT-3') and oPV370 (5'-GATATACCAGGTCGTTGC-3') for defA, oPV385 (5'-TCCCTTATTGCTCCACTTG-3') and oPV368 (5'-TATCATTTGTTTTCATGCCTC-3') for defB, and oPV383 (5'-CCACAACCTCTATCATTACCAG-3') and oPV384 (5'-ACCGTTCCATAGACCAGC-3') for fmt. All PCR products were purified with QiaQuick DNA purification columns (Qiagen), and DNA sequences were determined by the dideoxy chain termination method (Sequetech). In all instances, mutations were confirmed by analysis of a second independent PCR product. Pharmacokinetic studies. | CD-1 female outbred mice (Charles River Laboratories) were used for pharmacokinetic analyses of selected PDF inhibitors. The compounds were formulated in 20% cyclodextrin (Aldrich) and filter sterilized. After 1 week of acclimation, mice (20 to 25 g each) were dosed either intravenously (i.v.) or orally (p.o.) with antibiotic; at 0.083, 0.25, 0.5, 1, 2, and 4 h after dosing, blood samples were collected from anesthetized mice via cardiac puncture. Groups of four mice were used for each time point. The blood was allowed to clot, and the serum samples were stored immediately at -80C. Serum was extracted with a mixture of acetonitrile, ethanol, and acetic acid and analyzed by liquid chromatography with triple quadropole mass spectrometry (LC-MS/MS) MS (Phenomenex LUNA-C8, 50 by 2 mm, 3 mum, HP1100, Micromass Quattro LC) with an internal standard. The pharmacokinetic parameters, including Tmax (time to maximum concentration), Cmax (maximum concentration measured), t1/2 (terminal half-life), and AUC (area under the curve), were calculated. The p.o. bioavailability was calculated as the ratio of the AUC for p.o. administration to the AUC for i.v. administration. Metabolic stability in liver microsomes. | Preparations of hepatic microsomes from male CD-1 mice, male Sprague-Dawley rats, and human donors were obtained from Xenotech LLC. To assess their metabolic stability, test compounds (1 muM) in phosphate buffer (100 mM, pH 7.4) were incubated at 37C with liver microsomal proteins (0.2 mg/ml) supplemented with NADPH (1 mM), UDP-glucuronic acid (4 mM), and alamethecin (60 mug/mg of protein) for 0, 10, 20, and 30 min. Aliquots were removed at various times, and the reaction was terminated by the addition of acetonitrile. Loss of the parent compound was monitored by LC-MS/MS (PE-SCIEX API3000; triple quadrupole) with reverse-phase high-pressure liquid chromatography separation (Phenomenex LUNA-C18, 2.1 by 30 mm, 3 mum). The viability and activity of the microsomal preparations were separately verified by using a positive metabolic control. Intrinsic clearance was calculated from the slope of the curve of log10 linear concentration versus incubation time. Metabolite identification. | Mouse serum samples from the pharmacokinetic studies described above were extracted with acetonitrile and analyzed by LC-MS/ MS to identify additional metabolites in the serum. In a separate experiment, test compounds (10 muM) were incubated with mouse, rat, and human liver microsomes as described above at 37 C for 20 min, and the resulting samples were extracted before being analyzed for metabolites. Protection from infection in the mouse septicemia model (S. aureus strain Smith). | A standard peritonitis model of infection was used as an initial screen to determine the relative efficacies of the PDF inhibitors (MDS Pharma Services, Bothell, Wash.). Briefly, outbred ICR-derived mice (20 to 25 g each) were inoculated intraperitoneally with a 90 to 100% lethal dose (9.0 x 105 CFU/mouse) of S. aureus strain Smith (ATCC 19636) in 0.5 ml of brain heart infusion broth containing 5% mucin. Compounds were dissolved in 20% cyclodextrin and administered subcutaneously (s.c.) and/or p.o. at 1 and 5 h after infection. Vancomycin was included as a control antibiotic, and animals in this antibiotic regimen were dosed s.c. Groups of five mice were used for each dose level. Mice were monitored daily for 6 days, and cumulative mortality was used to determine the 50% effective dose (ED50) by nonlinear regression with Graph-Pad Prism (Graph Pad Software). FIG. 2. | Synthesis of N-alkyl urea hydroxamic acids. Synthesis of N-alkyl urea hydroxamic acids. The letters in the scheme represent the following solvents: a, NsCl, aqueous NaHCO3, DCM; b, R1OH, DIAD, PPh3; c, PhSH, TBDE-PS, DMF; d, COCl2, aqueous NaHCO3, Et2O; e, l-Pro-OtBu, pyridine, DCM; f, TFA, DCM; g, R2NH2, PyBop, DIEA, DMF; and h, aqueous NH2OH, dioxane. RESULTS : Previously, the alkyl succinate hydroxamic acid VRC3324, (2R)-butyl-N4-hydroxy-N1-[2-methyl-(1S)-(naphthalen-2-ylcarbamoyl)-propyl]-succinamide, was identified as a potent PDF inhibitor with excellent antibacterial activity (C. Hackbarth, S. Lopez, M. Gomez, W. Wang, J. Jacobs, R. Jain, Z.-J. Ni, J. Trias, D. Chen, G. Withers, K. Clark, J. Koehn, B. Weidmann, D. Patel, and Z. Yuan, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 2173, 2000). Similar compounds of this chemical class exhibit broad-spectrum antibacterial activity and in vivo activity in a mouse septicemia model (D. Chen, C. Hackbarth, Z.-J. Ni, W. Wang, C. Wu, D. Young, R. White, J. Trias, D. Patel, and Z. Yuan, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 2175, 2000). In our efforts to discover new pharmacophores that would improve the biological properties of this class of inhibitors, N-alkyl urea hydroxamic acids were identified. VRC3852 represents an early lead in this new class of PDF inhibitors (Lewis et al., 41st ICAAC). Analogs of VRC3852, prepared in both library format and as individual compounds, were screened to identify those with optimized activity and structure. In vitro activity tests. | From more than 200 analogs tested, 21 compounds were purified by high-pressure liquid chromatography and subjected to additional biological characterization. All of these compounds were tested for the inhibition of deformylase activity with the E. coli and S. pneumonia enzymes overproduced and purified from E. coli. In addition, whole-cell activity was assessed with broth microdilution MICs against a panel of 26 organisms, including both gram-positive and -negative bacteria and Candidaalbicans. Twenty of these 21 compounds have IC50s of <=100 nM for E. coli deformylase. Eighteen of these 20 compounds also have excellent activity (IC50s, <=75 nM) against S. pneumoniae PDF. These IC50s corresponded to improved activity against S. pneumoniae. In fact, the two compounds with poor IC50s for the S. pneumoniae enzyme, VRC4347 and VRC4390, also lacked whole-cell activity against this organism, providing further evidence that this class of compounds targets deformylase for the inhibition of bacterial growth. Compared with the MICs of the initial lead in this series, VRC3852, the MICs against the three gram-positive pathogens tested (S. aureus, S. pneumoniae, and Enterococcusfaecium) improved for the majority of the compounds listed. One of the compounds that had poor gram-positive activity, VRC4390, nonetheless maintained its activity against H. influenzae. On the other hand, VRC4234 and VRC4305 were more active against S. aureus and S. pneumoniae but less potent against the gram-negative pathogen H. influenzae. Apparently, these compounds readily undergo efflux in E. coli, since none was active against wild-type E. coli but all possessed low MICs against an E. coliacr mutant which lacks the Acr efflux pump . None of these compounds was active against C. albicans. To determine whether this class of compounds inhibited bacterial growth through the inhibition of deformylase, each of the compounds was tested for its ability to inhibit the growth of an actinonin-resistant S. aureus strain, VSAU6011. This strain harbors a loss-of-function mutation in the formyltransferase-encoding gene, bypassing the need for deformylase activity and permitting growth even when PDF activity is inhibited . S. aureus VSAU6011 was highly resistant to all of the compounds tested, demonstrating that N-alkyl urea hydroxamic acids, like alkyl succinates , inhibit bacterial growth through the inhibition of deformylase. These compounds were also tested for their potential cytotoxicity and inhibition of matrilysin, a member of the mammalian MMP superfamily which closely resembles PDF . Matrilysin activity was not inhibited in the presence of up to and in most instances exceeding 100 muM PDF inhibitor. In addition, the IC50s of VRC4232 and VRC4307 against human ACE, another metalloprotease, were higher than 100 muM. Thus, these N-alkyl urea hydroxamic acids are over 1,000 times more potent against bacterial deformylase than the eukaryotic metalloenzymes tested. When these compounds were tested against two different mammalian cell lines, the IC50s, determined after 72 h of continuous exposure, were at least 10-fold higher than the MICs for the major bacterial pathogens in the screening panel. A closer examination of the structure-activity relationships of these compounds (see Discussion) suggested that the cytotoxicity of these compounds generally correlated with their hydrophobicity. Given their overall in vitro profiles, VRC4232 and VRC4307 were selected as new leads for additional in vitro and in vivo characterizations. X-ray analysis of binding interactions. | Key parameters and summary statistics for the refined PDF model are provided in Table . After independent rigid-body refinement of the three PDF proteins in the asymmetric unit, the electron density for the three inhibitors was clearly seen in each of the active sites. Several rounds of iterative model building (including the addition of water molecules and correcting loops involved in crystal contacts) with the program O and refinement were conducted before the three inhibitor molecules were added to the model. In complex A, the inhibitor could be positioned unambiguously in the electron density. For complexes B and C, there was a weak, broken electron density for the 4,5-dimethylthiazole ring of the inhibitor. Rfree was monitored while additional water molecules were added. The final model contained three PDF-compound VRC4307 complexes as well as 435 water, 3 sulfate, and 5 glycerol molecules, resulting in an R factor of 18.6% (Rfree = 21.6%) in the resolution range of 30.0 to 1.8 A. The structure of E. coli PDF has already been extensively characterized and will not be elaborated on further here . When the PDF structure with bound VRC4307 is examined, the hydroxamic acid is observed to coordinate the active-site Ni2+, as expected; the cyclopentyl ring system of VRC4307 lies in the S1' pocket of the enzyme . The carbonyl oxygen of the proline group of VRC4307 forms a hydrogen bond with the backbone nitrogen of Ile44, maintaining the peptidic interaction observed for the PDF structure with the bound product MAS . Details of the metal coordination, hydrogen bonding, and Van der Waals interactions are shown in Fig. . The electron density for most of the compound is clear and unambiguous. The electron density for the amide-dimethylthiazole (Fig. , green atoms) portion of the inhibitor is weak and broken. Apparently, the inhibitor is able to adopt multiple conformations when bound; this hypothesis is illustrated by superimposition of the three complexes . Surprisingly, the peptidic interaction with the backbone carbonyl of Glu42 is seen in only one of the three complexes . The hydroxamic acid-metal coordination in the PDF-VRC4307 complex maintains a tetrahedral configuration . Although the overall arrangement of the ligand atoms approximates a distorted trigonal bipyrimidal geometry, the carbonyl oxygen of the inhibitor is more distant from the metal than is typically observed in metal-ligand complexes. In all three complexes, the Ni2+ distances are reasonable for the N and O atoms (2.0 to 2.1 A) and for the S atoms (2.3 A); however, the carbonyl O atom of VRC4307 is 2.7 to 3.0 A from the metal. This suboptimal interaction may be offset, in part, by an ideal hydrogen bonding interaction distance between the carbonyl oxygen and the backbone nitrogen of Leu91. The unusual metal-ligand geometry observed appears to be a direct consequence of the relatively planar N compared to an alpha-carbon in substrate/product. The VRC4307 cyclopentyl analog of the methionine side chain is packed on top of His132 of PDF, meaning that the relatively planar configuration of the urea N prevents the hydroxamic acid from fully entering the binding pocket for an optimal interaction with the metal. The favorable hydrogen bond interaction between a carbonyl of VRC4307 and Ile44 is accompanied by a steric clash between the carbonyl and Gly43 . Killing curves and MBCs. | It was previously shown that actinonin, a succinate hydroxamic acid, is bacteriostatic against S. aureus . In this study, killing curves for the urea compounds VRC4232 and VRC4307 were determined at 10 times the MICs for S. aureus, S. pneumoniae, and H. influenzae . Like actinonin, the urea-based compounds are also bacteriostatic against S. aureus, but a killing effect was observed against the other two organisms. For S. pneumoniae, viable counts dropped by greater than 2 log10 units over the initial 12 h of exposure; for H. influenzae, a drop of up to 5 log10 units in viable organisms was observed after 24 h. Both compounds were bacteriostatic over the initial 8 h of exposure; increasing the drug concentration to 20 times the MIC (80 mug/ml) did not improve the initial killing rate (data not shown). These observations were consistent with the MBC determinations for H. influenzae and S. pneumoniae. Unlike S. aureus (data not shown), these two fastidious pathogens are killed by PDF inhibitors, although the reason for the late killing effect is unclear. Resistance and its mechanism. | The in vitro selection of spontaneously resistant H. influenzae and S. pneumoniae mutants occurred at a frequency of 1 in 109 for both VRC4232 and VRC4307 . These resistance frequencies are similar to those observed previously for actinonin against S. pneumoniae . Two independent H. influenzae mutants, VHIN6529 and VHIN6536, were selected for further characterization. Both mutants were highly resistant to all PDF inhibitors tested: the MICs of all such compounds, except actinonin, were >64 mug/ml; the MIC of actinonin was 32 mug/ml . The antibiotic profile of the mutants is comparable to that of the parent strain, suggesting that there is no cross-resistance with other classes of antibiotics and that this resistance is not due to a defect in a transport mechanism that is shared with other antibiotics. The resistant mutants appear on solid medium as small colonies; they grow slowly both in broth and on agar. When these small colonies are streaked on solid agar in the absence of a drug, large, normally sized colonies frequently appear on the plate. These revertants are again as susceptible to deformylase inhibitors as the parent strain, VHIN1004 (data not shown). To better define the mechanism of resistance to compounds of the urea series, the def and fmt homologs from H. influenzae mutant VHIN6529 and from three independent revertants were sequenced and compared to those of the wild type. In comparison with the sequences of the genes from parent strain VHIN1004, no change was observed in the def homolog. However, the fmt gene from mutant VHIN6529 (and each revertant) contained a frameshift mutation: the deletion of a C residue following bp 307 of the open reading frame (ORF). The resulting protein is predicted to diverge from the wild-type protein at amino acid 103, terminating four amino acids later . The def sequences from the three independent revertants were not changed compared to that from the parent strain. However, all three revertants carried, in addition to the original 1-bp deletion in fmt, a T residue inserted following bp 299 of the fmt ORF. This compensatory frameshift mutation restores the fmt reading frame in mutant VHIN6529, such that the revertants express a full-length transformylase with a three-amino-acid substitution at codons 99 through 101 . Two independent S. pneumoniae mutants, VSPN6521 and VSPN6522, were further characterized. Morphologically, the resistant mutant strains resemble the parent strain. The doubling times of the mutant strains (74 to 82 min) are slightly longer than that of the corresponding wild type (65 min), a finding that is very similar to what was found with mutants of S. pneumoniae selected with actinonin . Like that of the H. influenzae mutants, the overall antibiotic profile of the pneumococcal mutants does not vary from that of the parent strain . The defB (PDF), defA (PDF paralog), and fmt (transformylase) genes were amplified from these mutants by PCR, and the DNA sequences were compared to the sequence of the wild-type parent strain. No changes were observed in defA or fmt, but for both mutants, a G-to-T transversion was observed at bp 211 of the defB ORF. This change is predicted to generate a Val (GTT)-to-Phe (TTT) substitution at amino acid 71 of the PDF product, a substitution that is distinct from those observed previously for S. pneumoniae mutants selected with actinonin . One major difference can be observed between the resistant mutants from S. pneumoniae and those from H. influenzae: while H. influenzae mutants selected with urea are cross resistant to both succinate and N-alkyl urea PDF inhibitor classes, mutants from S. pneumoniae are cross resistant to other N-alkyl urea compounds but display increased susceptibility to actinonin, a succinate hydroxamic acid. Pharmacokinetic studies with mice. | The clearance of VRC4232 and VRC4307 in mouse serum was measured after the compounds were administered i.v. and p.o. The concentrations in serum at different times were used to calculate the corresponding pharmacokinetic parameters presented in Table . VRC4232 has a longer half-life (1.1 h) than VRC4307 (0.1 h) after i.v. administration. Both compounds have poor p.o. bioavailability (3.2 and 0.1%, respectively), suggesting that these compounds will not have significant p.o. efficacy. In vitro metabolic stability. | The metabolic stability of the compounds was determined in vitro: the compounds were incubated with mouse, rat, and human liver microsomes and then analyzed for loss of the parent compound over time. As summarized in Table , both compounds were rapidly metabolized by mouse and rat liver microsomes (t1/2, <3 min) but were much more stable in human liver microsomes (t1/2, >19 min). Identification of major metabolites for VRC4232. | The serum samples collected from mice after dosing with VRC4232 were analyzed for major metabolites by LC-MS/MS. Similar analyses were also carried out for samples collected after incubation of VRC4232 with liver microsomes. The major metabolites identified from these samples are summarized in Fig. . Since the authentic compounds were not available, quantitative analysis was not performed. Although these metabolites were identified in all samples, their relative intensities in the LC-MS spectrum varied. The three major peaks found by LC-MS for these metabolites (in order of decreasing relative peak areas) were E, A, and B for in vivo mouse serum; A, E, and F for mouse liver microsomes; D, E, and A for rat liver microsomes; and C, A, and E for human liver microsomes. The majority of the major metabolites correspond to modifications of the hydroxamic acid, including hydrolysis (metabolites A, B, and C in Fig. ) and glucuronic acid conjugation (metabolites E and F in Fig. ). All of these modifications would result in loss of the chelating activity of the parent compound, an effect which should render the compounds devoid of PDF inhibitory activity. Other major modifications identified include hydroxylation of the proline group (metabolite C in Fig. ) and oxidation of dimethylthiazole (metabolites B, D, and F in Fig. ). Mouse S. aureus septicemia model. | VRC4232 and VRC4307 both have MICs of 0.015 mug/ml against S. aureus strain Smith, used in the septicemia model. After s.c. administration of VRC4232, four of five mice survived at 32 mg/kg of body weight and none of five survived at 16 mg/kg, resulting in an ED50 of 29.7 mg/kg. VRC4307 was more efficacious in this model. The survival of three, three, one, and none of five mice was observed after s.c. doses of 32, 16, 8, and 4 mg/kg, respectively, resulting in an ED50 of 17.9 mg/kg. VRC4307 was also tested for its efficacy after p.o. administration; no protection was observed at the highest tested dose of 30 mg/kg (i.e., none of five mice survived). The ED50 for the control antibiotic, vancomycin, was 2.4 mg/kg after s.c. dosing. FIG. 3. | Surface representation of VRC4307 binding to E. Surface representation of VRC4307 binding to E. coli PDF. PDF is represented as a smooth surface created by GRASP; the catalytic Ni2+ is shown as a magenta sphere. (A) The cyclopentyl ring system is positioned in the S1' pocket. (B) The inhibitors of three complexes have been superimposed to illustrate relative disorder in the dimethylthiazole rings of the inhibitors. FIG. 4. | Interactions between E. Interactions between E. coli PDF and VRC4307. (A) Schematic representation of key interactions between PDF and VRC4307. (B) Interactions of the hydroxamic acid with the active-site metal. Although the ligand coordination geometry is a distorted trigonal bipyramid, the fifth ligand, the carbonyl oxygen, is positioned too far from the Ni2+ atom for optimal interactions with three complexes (2.7, 2.7, and 3.0 A). Solid lines indicate proposed interactions; the potential fifth ligand interaction is shown as broken lines. FIG. 5. | Bacterial killing curves. Bacterial killing curves. Log-phase organisms were exposed to either no drug , 10 times the MIC of VRC4232 , or 10 times the MIC of VRC4307 (*). Results are expressed as log10 CFU per milliliter versus time. FIG. 6. | Internal DNA sequence and predicted protein sequence (codons 94 to 109) of the H. Internal DNA sequence and predicted protein sequence (codons 94 to 109) of the H. influenzaefmt ORF from wild-type strain VHIN1004 (top), resistant mutant VHIN6529 (middle), and a susceptible revertant (bottom). Note that the insertion of T at codon 99 restores the fmt reading frame in the revertant. Underlined amino acids indicate divergence from the wild-type formyltransferase protein sequence. FIG. 7. | Modification of VRC4232 in vivo (mice) and in vitro (liver microsomes) and identification of its major metabolites (A to F). Modification of VRC4232 in vivo (mice) and in vitro (liver microsomes) and identification of its major metabolites (A to F). TABLE 1 | Structure and activity of N-alkyl urea hydroxamic acids TABLE 2 | Summary of diffraction data and refinement data for the crystal structure of the E. coli PDF complex TABLE 3 | MICs and MBCs of PDF inhibitors against H. influenzae VHIN1004 and S. pneumoniae VSPN1001 TABLE 4 | Antibiotic profiles of H. influenzae and S. pneumoniae mutant strains TABLE 5 | Mouse pharmacokinetic parameters for VRC4232 and VRC4307 TABLE 6 | Comparison of metabolic stability of VRC4232 and VRC4307, as determined in vitro by incubation with liver microsomes from mice, rats, and humans DISCUSSION : N-alkyl urea hydroxamic acids represent a new class of PDF inhibitors. All previously published PDF inhibitors with reasonable antibacterial activity share the structural feature of a 2-substituted alkanoyl, particularly 2-substituted hexanoyl, at the P1' site . In this study, it was shown that bacterial PDF can accommodate N-alkyl urea at the P1' site. The tested compounds also can effectively inhibit the growth of both gram-positive and gram-negative bacteria. The N-alkyl urea structure allows easy preparation of analogs with different P1' and P3' substitutions to explore the combination of moieties at either site that might provide the best antibacterial activity. Various alkyl groups at the P1' site were explored in a combinatorial library format: branched and cyclic aliphatic groups such as N-cyclopentylethyl and N-isopentyl showed the best antibacterial activity (J. Lewis, unpublished data). Compounds combining these two P1' substitutions with different P3' groups were prepared as purified compounds and tested in a panel of in vitro assays . The data in Table demonstrate that the N-alkyl urea compounds can very effectively inhibit bacterial PDFs with IC50s well below 1 muM. In general, compounds with substituted thiazoles have excellent activity against S. aureus and S. pneumoniae. Overall, all of the N-alkyl urea compounds prepared resulted in moderate inhibition of H. influenzae growth. A comparison of VRC4390, VRC4156, VRC5157, and VRC4234, all of which share the P1' N-isopentyl substitution, suggests that increasing hydrophobicity at the P3' site improves enzymatic and whole-cell activities for this class of compounds. The trend is particularly obvious when a methyl group is incorporated in the P3' site, a substitution which results in a dramatic loss of inhibition of S. pneumoniae PDF but which has only a slight effect on E. coli PDF inhibition. The methyl substitution also renders the compound inactive against S. aureus and S. pneumoniae in whole-cell assays, while activity against H. influenzae and efflux-deficient E. coli is maintained. These differences suggest that VRC4390 lacks activity against S. aureus and S. pneumoniae because of a species-specific loss of activity against the corresponding PDF. This observation further suggests that deformylases from different bacterial species may have different preferences at their S3' sites. The crystal structure of the complex of VRC4307 with E. coli PDF revealed the interactions between the urea hydroxamate and the enzyme. Surprisingly, the hydroxamate did not adopt the expected bidentate interaction that is common with MMP inhibitors and that was observed in the recently published PDF-actinonin structure . The relatively planar N (relative to an alpha-C atom of a substrate or actinonin) coupled with the P2' proline limits the ability of the inhibitor to fully reach the active-site metal. The result is a single metal-ligand interaction with the hydroxamate, while the remaining oxygen atom forms a hydrogen bond interaction with the backbone amide of Leu91. In order to achieve these interactions and maintain the P1' interactions, the peptidic interaction between the carbonyl of the inhibitor and the backbone amide of Ile44 is also distorted relative to the actinonin complex. A good hydrogen bonding distance is achieved, and close contact with the carbonyl and the alpha-carbon of Gly43 results (2.8 A). Due to the structural resemblance of the key catalytic residues between PDFs and MMPs, the 2-substituted alkanoyl can fit very well into the S1' site of the active center of both enzymes . This similarity could result in the inhibition of MMPs by many deformylase inhibitors. In this study, we show that bacterial PDF can accommodate N-alkyl urea in the P1' site. In addition to structural novelty, this alternative P1' substitution also provides excellent selectivity, as demonstrated by the absence of MMP inhibition. The data in Table indicate that none of the compounds exhibits inhibitory activity against MMP-7 at 200 muM; analogous compounds with a 2-substituted alkanoyl substitution at the P1' site usually display complete inhibition of MMP-7 at this concentration (data not shown). Based on the crystal structure, it appears that the selectivity may require both the N-alkyl urea hydroxamate and a P2' proline group. The conformation adopted by the urea hydroxamate places its P2' proline ring at a sterically unfavorable distance from the MMP carbonyl of Pro238 (matrilysin numbering). This proline is conserved in the MMP family, and the orientation of its carbonyl is quite different from that of Gly89, which occupies the equivalent position in E. coli PDF. This structural property of the N-alkyl urea hydroxamic acids should avoid the potential toxicity of hydroxamic acids that is due to the inhibition of MMPs. The urea hydroxamates appear to be bacteriostatic against S. aureus, as was observed previously for other PDF inhibitors against both S. aureus and E. coli . Huntington et al. reported the lysis of Bacillussubtilis by a peptide thiol PDF inhibitor, although the compound appeared to be bacteriostatic when tested with log-phase organisms. In contrast to the situation with S. aureus, VRC4232 and VRC4307 both had a killing effect on S. pneumoniae and H. influenzae, two pathogens that cause upper respiratory tract infections. Notably, neither compound displayed a significant degree of killing of H. influenzae over the initial hours of the experiments, but there was significant death of this species by 24 h of drug exposure. The basis for this delayed killing effect is not clear, but these data and the characterization of the resistant mutants indicate that the inability to remove the N-formyl group from proteins is toxic in H. influenzae. Hypothetically, several cycles of growth in the presence of deformylase inhibitors would deplete the pool of normally processed polypeptides, while N-formylated proteins would accumulate. If, at that late stage, one or more enzymes are required but are nonfunctional due to N formylation (e.g., improper folding), then cell death will ensue. The frequency of resistance to the urea-based inhibitors in both H. influenzae and S. pneumoniae is approximately 10-9, as was previously reported for actinonin, a PDF inhibitor of the succinate hydroxamic acid series. This value is comparable to what was found for a peptide thiol PDF inhibitor against B. subtilis and is 2 to 3 orders of magnitude smaller that what has been reported for the spontaneous selection of resistance to other deformylase inhibitors in S. aureus and E. coli . Two mechanisms of resistance to succinyl hydroxamate-based inhibitors of bacterial PDF have been reported. In E. coli, S. aureus, Moraxellacatarrhalis, and H. influenzae, loss-of-function mutations in fmt have been identified (, , ; P. Margolis, C. Hackbarth, M. Maniar, S. Lopez, W. Wang, R. White, Z. Yuan, and J. Trias, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 2174, 2000). The data presented here indicate that resistance to urea-based PDF inhibitors in H. influenzae occurs through a mechanism similar to that seen for succinyl hydroxamate-based compounds in the same species. As was observed previously for actinonin resistance mutations, resistance to VRC4232 and VRC4307 in H. influenzae apparently is derived from a frameshift mutation that results in a truncated (and presumably nonfunctional) formyltransferase protein. Consistent with previous observations made for E. coli, S. aureus, and H. influenzae, H. influenzaefmt mutant strain VHIN6529 grew more slowly than the wild-type parent strain. Revertants of strain VHIN6529 simultaneously displayed an improved growth phenotype, increased susceptibility to PDF inhibitors, and a compensatory frameshift mutation that restored the fmt reading frame. This correlation strongly supports a model whereby resistance to urea-based PDF inhibitors can occur through bypassing of the formylation cycle. It is not evident why the loss-of-function mutations found in H. influenzaefmt occur at a lower frequency (10-8) than those found with succinate inhibitors in E. coli or S. aureus (10-6). One hypothesis is that the difference in resistance rates may be due to the bactericidal effects of these compounds on H. influenzae, which may lead to a decreased number of viable mutants able to survive the selection process. Alternatively, rather than a bypass mechanism, perhaps the truncated Fmt created by the frameshift mutation remains partly functional, and fmt is essential in H. influenzae. In S. pneumoniae, a second mechanism has been identified; resistance to actinonin results from missense mutations in the defB gene. The distinction in resistance mechanisms observed in different species presumably reflects the fact that fmt is essential in pneumococci . Resistance to the urea-based compounds VRC4232 and VRC4307 in S. pneumoniae also appears to be mediated by an alteration in the target, but the location of the missense mutation in defB (V71F) is distinct from those of two other pneumococcal defB mutations, A123D and Q172K, that were identified previously . The V71F mutation seen in strains VSPN6521 and VSPN6522 is located in the highly conserved box 1 (70GVGLAAPQ77) of S. pneumoniae PDF. As a comparison with the PDF consensus sequence shows, Val71 does not correspond to a strictly conserved residue; an Ile-to-Ala substitution of the equivalent residue in the E. coli enzyme is well tolerated . However, this residue does lie between Gly residues known to be critical for enzyme activity in the E. coli enzyme. It also appears that the defB (V71F) mutants VSPN6521 and VSPN6522 are able to distinguish between inhibitors of the urea and succinate hydroxamate series. These mutants are more susceptible to actinonin as well as to other succinate inhibitors (data not shown), although they are highly resistant to all urea-series PDF inhibitors tested. In contrast, mutants selected with actinonin contain distinct defB mutations and are resistant to both classes of PDF inhibitors (data not shown). A crystal structure of the enzyme containing this V71F mutation is not available; however, the crystal structure of E. coli PDF complexed with VRC4307 can be used to infer the effect of this mutation. The two most common rotamers for Phe71 would place it in either the S1' or the S3' pocket. Given that the V71F mutant is still active, the more likely position would be in the large S3' pocket, which could accommodate this large side chain with only minor perturbations. The urea hydroxamate inhibitors are less likely to tolerate any movement of Phe71 (and the adjacent Gly70) into the active site due to the close contact between Gly70 and the inhibitor. These observations may reflect a species-specific structural aspect of the pneumococcal PDF enzyme. The metabolic instability of the urea-series compounds seen in rodent microsomes is likely a factor in the rapid clearance of the compounds in this species . However, the rate of clearance of both compounds in human liver microsomes is approximately 10-fold slower than that in rodent microsomes, suggesting that this class of compounds should have longer half-lives in humans. The predictive value of these in vitro models has been reported by others and is borne out by the similar spectra of the metabolic products observed in vitro (mouse liver microsomes) and in vivo (mouse serum). Two major modifications, hydroxamic acid hydrolysis and glucuronic acid conjugation, were prominent in both instances, and the sites of modification were also comparable. VRC4232 and VRC4307 were selected for further in vivo studies. In an S. aureus septicemia model, both compounds showed moderate protective activity after s.c. administration, establishing N-alkyl urea hydroxamic acids as potential antibacterial agents. However, when the more potent compound VRC4307 was tested after p.o. administration in the same model, no protective effect was observed up to 30 mg/kg. This lack of p.o. efficacy presumably reflects the poor p.o. bioavailability of VRC4307, as demonstrated by the in vivo pharmacokinetic analysis (absolute p.o. bioavailability, 0.1%). These data suggest the need to incorporate bioavailability tests at early stages of compound screening by using Caco-2 cell in vitro testing or by using cassette dosing. In summary, we have identified N-alkyl urea hydroxamic acids as a new class of PDF inhibitors. This class of compounds has potent whole-cell activity against both gram-positive and gram-negative bacteria but is devoid of MMP inhibition. The potential use of this class of compounds as antibacterial agents is supported by their protective activity in an in vivo infection model. Backmatter: PMID- 12183253 TI - In Vitro Selection of Resistance in Haemophilus influenzae by Amoxicillin-Clavulanate, Cefpodoxime, Cefprozil, Azithromycin, and Clarithromycin AB - Abilities of amoxicillin-clavulanate, cefpodoxime, cefprozil, azithromycin, and clarithromycin to select resistant mutants of Haemophilus influenzae were tested by multistep and single-step methodologies. For multistep studies, 10 random strains were tested: 5 of these were beta-lactamase positive. After 50 daily subcultures in amoxicillin-clavulanate, MICs did not increase more than fourfold. However, cefprozil MICs increased eightfold for one strain. Clarithromycin and azithromycin gave a >4-fold increase in 8 and 10 strains after 14 to 46 and 20 to 50 days, respectively. Mutants selected by clarithromycin and azithromycin were associated with mutations in 23S rRNA and ribosomal proteins L4 and L22. Three mutants selected by clarithromycin or azithromycin had alterations in ribosomal protein L4, while five had alterations in ribosomal protein L22. Two mutants selected by azithromycin had mutations in the gene encoding 23S rRNA: one at position 2058 and the other at position 2059 (Escherichia coli numbering), with replacement of A by G. One clone selected by clarithromycin became hypersusceptible to macrolides. In single-step studies azithromycin and clarithromycin had the highest mutation rates, while amoxicillin-clavulanate had the lowest. All resistant clones were identical to parents as observed by pulsed-field gel electrophoresis. The MICs of azithromycin for azithromycin-resistant clones were 16 to >128 mug/ml, and those of clarithromycin for clarithromycin-resistant clones were 32 to >128 mug/ml in multistep studies. For strains selected by azithromycin, the MICs of clarithromycin were high and vice versa. After 50 daily subcultures in the presence of drugs, MICs of amoxicillin-clavulanate and cefpodoxime against H. influenzae did not rise more than fourfold, in contrast to cefprozil, azithromycin, and clarithromycin, whose MICs rose to variable degrees. Keywords: Introduction : Haemophilus influenzae is a major cause, together with Streptococcus pneumoniae and Moraxella catarrhalis, of community-acquired respiratory infections in children and adults, including pneumonia, acute exacerbations of chronic bronchitis, sinusitis, and otitis media. In countries such as the United States where the H. influenzae type b vaccine is widely used, untypeable H. influenzae strains cause the majority of infections . The major resistance mechanism in H. influenzae is beta-lactamase production (TEM-1, ROB-1). A recent study in the United States has documented the incidence of beta-lactamase production in 1,676 H. influenzae strains isolated throughout the United States to be 41.6% . The incidence of beta-lactamase-negative ampicillin-resistant (BLNAR) strains in the United States is <1%. Of available beta-lactams for treatment of infections caused by this organism, cefixime and cefpodoxime are the most active from a potency as well as a pharmacokinetic and pharmacodynamic point of view, followed by amoxicillin-clavulanate, cefuroxime, and cefdinir. Other oral cephalosporins such as cefprozil, cefaclor, and loracarbef are less active against these organisms . Among macrolides and azalides, azithromycin yields the lowest MIC for H. influenzae, followed by erythromycin and clarithromycin. However, the pharmacokinetic and pharmacodynamic properties of these compounds cast doubt on their clinical efficacy against H. influenzae . It is well known that antimicrobial abuse leads to development of resistance. Baquero has postulated that abuse of azithromycin and clarithromycin has played an important part in the development of macrolide resistance in S. pneumoniae in Spain, Italy, and France . The present study attempts to shed further light on this phenomenon in H. influenzae by using single and multistep resistance selection studies to examine the effect of amoxicillin-clavulanate, cefpodoxime, cefprozil, azithromycin, and clarithromycin. MATERIALS AND METHODS : Bacteria. | Ten recent clinical isolates of untypeable H. influenzae were tested. Of these, five were beta-lactamase positive (by Cefinase disk methodology; BBL Microbiology Systems, Cockeysville, Md.) and five were beta-lactamase negative. Strains were stored in double-strength skim milk (Difco Laboratories, Detroit, Mich.) at -70C before testing and examined for purity throughout the study by culture and Gram stain methods. MIC methodology and antimicrobials. | Microdilution MICs of all strains were determined as recommended by the NCCLS using freshly prepared Haemophilus test medium (HTM), with inoculum checks performed in each case. Standard quality control strains were included in each run. Trays were incubated for 16 to 20 h in ambient air. Drugs were obtained from their respective manufacturers. Multistep mutation studies. | The method previously used by our group for pneumococci was modified as follows. Serial passages in freshly prepared HTM were performed daily for each strain in subinhibitory concentrations of all antimicrobials. For each subsequent daily passage, an inoculum was taken from the tube containing antimicrobial at 1 to 2 dilutions below the MIC. The latter inoculum was used to determine the next MIC. Daily passages were performed until a significant increase (more than fourfold) was obtained. A minimum of 14 passages were performed in each case. For cefprozil, three beta-lactamase-positive strains for which the MICs were 128 were not tested. The maximal number of passages was 50. Control MICs were also determined daily in the original strain without serial passages. Stability of the acquired resistance was determined after 10 daily passages of the clone on chocolate agar (BBL) without antibiotics. Single-step mutations. | The frequency of spontaneous single-step mutations was determined by spreading approximately 1010 CFU/ml in 100-mul aliquots on HTM plates containing one, two, four, and eight times the MIC of each compound. After 24 to 72 h of incubation, resistant colonies were confirmed by replica plating on media with antibiotics. The resistance frequency was calculated as the number of resistant colonies per inoculum. Strains 4, 5, and 7, for which the MIC of cefprozil was 128, were not tested for single-step mutation for this antibiotic. Determination of macrolide and penicillin resistance mechanisms. | Presence of known resistance genes for macrolides mef(A/E), erm(B), erm(A) [subclass erm(TR)], and ere(A) was tested by PCR as described previously . Alterations in the genes coding for 23S rRNA and ribosomal proteins L4 and L22 were verified by sequencing of these genes after amplification by PCR. The sequences for 23S rRNA and ribosomal proteins L4 and L22 in H. influenzae were obtained from The Institute for Genomic Research website . Specific primers used to amplify the region involved in the peptidyl-transferase center of 23S rRNA were HF23-1 (5'CGGCGGCCGTAACTATAACG3') and HF23-2 (5'GATGTGATGAGCCGACATCG3'), which amplify the region from positions 1902 to 2520 in 23S rRNA (H. influenzae numbering). Specific primers were designed to amplify entire genes that code for ribosomal proteins L4 and L22. The L22 gene was amplified with primers HL22-1 (5'CGGCAGATAAGAAAGCTAAG3') and HL22-2 (5'TGGATGTACTTTTTGACCC3'). The L4 gene was amplified with primers HL4-1 (5'TTAAGCCGGCAGTTAAAGC3') and HF4-2 (5'CACTTAGCAAACGTTCTTG3). All PCRs were done as follows: 1 cycle of 94C (5 min); 35 cycles of 94C (1 min), 50C (1 min), and 72C (1 min); and 1 cycle of 72C (7 min). Mutations in the penicillin binding protein PBP 3 were investigated in beta-lactam-resistant mutants by amplification of the ftsI gene as described by Ubukata et al. . The PCR products after amplification of the 23S rRNA, L22, L4, and fstI genes were purified using a QIAquick PCR purification kit (QIAGEN, Valencia, Calif.) and sequenced using an Applied Biosystems model 373 DNA sequencer. PFGE. | All strains were tested by pulsed-field gel electrophoresis (PFGE) before and after resistance selection. PFGE was performed using a CHEF DR III apparatus (Bio-Rad, Hercules, Calif.) as described previously . Transformation. | Total DNA was extracted from mutant H. influenzae strains and used for transformation of the H. influenzae Rd (ATCC 51907) strain as described previously by Barcak et al. . Transformants were selected on brain heart infusion agar (Becton Dickinson, Cockeysville, Md.), supplemented with hemin (2 mug/ml) and NAD (2 mug/ml) (sBHI), that contained erythromycin (15 mug/ml). RESULTS : Multistep resistance selection. | The results of multistep resistance selection experiments are shown in Table . After 50 daily subcultures in amoxicillin-clavulanate, MICs did not rise more than fourfold, and the highest MIC was 4 mug/ml. After 50 days, the MICs of cefpodoxime for two strains rose to the level seen in low-BLNAR strains , with a fourfold increase to 0.25 mug/ml. These strains were not analyzed further, because they did not meet the mutant selection criteria (>4-fold increase) and remained susceptible according to the NCCLS breakpoint of 2.0 mug/ml . Cefprozil could not be tested for three beta-lactamase-positive strains for which the MICs were 128 mug/ml, and one of seven strains tested yielded higher MICs. Only one beta-lactamase-negative strain yielded a >4-fold increase after 32 days and became resistant (MIC >=32 mug/ml). The MIC for this strain was 4 mug/ml and became 64 mug/ml after selection. Azithromycin gave a >4-fold increase in MICs for all 10 strains after 20 to 50 days, and clarithromycin gave a >4-fold increase in MICs for 8 of 10 strains after 14 to 46 days. Parent 1 strain yielded two clarithromycin-resistant clones: one became hypersusceptible to macrolides during the 10 days of daily drug-free subcultures. The MICs of erythromycin, azithromycin, clarithromycin, clindamycin, and quinupristin-dalfopristin for this hypersusceptible mutant were 0.5, 0.06, 0.12, 0.06, and 0.06 mug/ml, instead of 8, 2, 8, 32, and 2 mug/ml, respectively, in the parent strain. All resistant clones were identical to parents by PFGE. MICs of azithromycin for azithromycin-resistant clones were 16 to >128 mug/ml, and those of clarithromycin for clarithromycin-resistant clones were 32 to >128 mug/ml. For strains selected by azithromycin, the MICs of clarithromycin were high and vice versa, but for the one strain selected by a beta-lactam (cefprozil) the azithromycin and clarithromycin MICs were the same as those for the parent strain. Single-step mutations. | Single-step mutation results are shown in Table . The mutation frequencies for cultures grown with MICs of the antimicrobials were 2.0 x 10-3 to 2.0 x 10-9 (clarithromycin), 6.0 x 10-4 to 1.3 x 10-9 (azithromycin), 1.7 x 10-7 to <2.5 x 10-10 (cefprozil for seven strains tested), 1.5 x 10-7 to <1.0 x 10-10 (cefpodoxime), and <6.7 x 10-10 to <1.0 x 10-10 (amoxicillin-clavulanate). Clarithromycin produced the highest mutation rates, followed by azithromycin, cefprozil, cefpodoxime, and amoxicillin-clavulanate. Mechanism of resistance to macrolides. | No parent or macrolide- or azalide-resistant strain had erm(B), mef(A/E), erm(A), or ere(A) genes. Of 18 mutants selected by clarithromycin and azithromycin, 2 had alterations in 23S rRNA, 5 had alterations in ribosomal protein L22, and three had alterations in ribosomal protein L4 (Tables to ). The alignments of amino acid sequence of deduced ribosomal protein sequences of L4 and L22 proteins of mutant resistant strains are shown in Fig. and (H. influenzae numbering). Two mutants, A2 and A7, selected by azithromycin had an A-to-C substitution at position 2059 and an A-to-G substitution at position 2058 in 23S rRNA, respectively (E. coli numbering). The MICs of erythromycin, azithromycin, clarithromycin, and clindamycin for these mutants were >128 mug/ml . Three strains selected by clarithromycin had deletions or one-amino-acid substitutions in the highly conserved region of ribosomal protein L4 . Mutant C3 had a replacement of Gly by Asp at position 65. Mutant 5 had a deletion of two amino acids, Gly and Arg, after position 65, and mutant C10 had a change of Gly to Ala at position 53 and a deletion of Arg and Ala after position 66 in ribosomal protein L4. These mutations in L4 were associated with a >4-fold increase in the MICs of clarithromycin and azithromycin . The alterations in ribosomal protein L22 were in the highly conserved region of this protein by substitution of one amino acid, or by amino acid insertions or deletions . Mutants C1 and A6 had one amino acid change at positions 88 and 93: Arg to Pro and Ala to Glu, respectively. The mutant C2 had seven amino acid insertions after position 72 in L22. Mutants C8 and A3 had one and three amino acid deletions, respectively, after position 82. The increase in MICs of clarithromycin and azithromycin for mutants with L22 alterations was at least eightfold . Total DNAs from C1 with the L22 mutation and C10 with the L4 mutation were used for transformation of H. influenzae Rd. After transformation resistant strains were selected on sBHI agar with erythromycin (15 mug/ml). Two transformants were studied further. Transformants C1/Rd (L22) and C10/Rd (L4) had the same PFGE pattern as the Rd strain after digestion by SmaI. The MICs of erythromycin for Rd strains rose from 2 mug/ml to 64 and 32 mug/ml after transformation with total DNA from mutants C1 and C10, respectively. The sequence analysis of the genes coding for 23S rRNA and ribosomal proteins L4 and L22 were amplified from transformants and the Rd strain and was sequenced. The transformants C1/Rd and C10/Rd had the same mutations as the mutants C1 and C10, respectively . Among 18 resistant mutant strains selected by azithromycin and clarithromycin, no modification was found in ribosomal proteins L4 or L22 or the studied portion of 23S rRNA for eight strains. These strains are currently being studied for their resistance mechanisms. Three randomly selected single-step mutants, selected at two times MIC by azithromycin and clarithromycin, were analyzed for alterations in 23S rRNA, and ribosomal proteins L4 and L22. Strains selected by azithromycin from parents 6 and 8 had deletions in ribosomal protein L22: two amino acids, Arg and Ile, at positions 95 and 96 from strain 6 and Asp and Glu at positions 77 and 78 from strain 8, were deleted. No modification was found in the other strains. Mechanism of resistance to cefprozil. | The gene ftsI that codes for PBP 3 from parent 1 and the mutant strain selected by cefprozil were amplified and sequenced. The analysis of these sequences and alignment of deduced amino acids showed two mutations, which led to Ala-to-Val and Ser-to-Asn substitutions at positions 271 and 357, respectively. These mutations were associated with a 16-fold increase in the cefprozil MIC; however, the activity of amoxicillin-clavulanate and cefpodoxime was not affected . FIG. 1. | Alignment of deduced amino acid sequences of the highly conserved region of ribosomal protein L4 from E. coli, Salmonella enterica Alignment of deduced amino acid sequences of the highly conserved region of ribosomal protein L4 from E. coli, Salmonella enterica serovar Typhimurium, and H. influenzae and parent strains and macrolide-resistant mutant H. influenzae strains obtained by in vitro multistep resistance selection. Alterations are indicated in boldface type. FIG. 2. | Alignment of deduced amino acid sequences of the highly conserved region of ribosomal protein L22 from E. coli, S. enterica Alignment of deduced amino acid sequences of the highly conserved region of ribosomal protein L22 from E. coli, S. enterica Typhimurium, and H. influenzae and the parent strains and macrolide-resistant mutant H. influenzae strains obtained by in vitro multistep resistance selection. Alterations are indicated in boldface type. TABLE 1 | Multistep resistance selection studies TABLE 2 | Frequencies of single-step mutation for 10 H. influenzae strains used for multistep study TABLE 3 | Susceptibilities of parent H. influenzae strains and their selected mutants with alterations in rRNA 23S TABLE 5 | Susceptibilities of parent H. influenzae strains and their selected mutants with alterations in ribosomal protein L22 TABLE 4 | Susceptibilities of parent H. influenzae strains and their selected mutants with alterations in ribosomal protein L4 TABLE 6 | Susceptibility of H. influenzae Rd strains transformed with total DNA from C1 with mutation in ribosomal protein L22 and C10 with mutation in ribosomal protein L4 DISCUSSION : The main mechanism of resistance detected in wild-type H. influenzae strains is the production of beta-lactamase, with an overall prevalence of 13.4% in Europe and up to 41.6% in the United States . In Japan, the prevalence of beta-lactamase production is low (13.9%). However, the prevalence of BLNAR strains is 28.8% , while in the United States and Europe this phenotype is rare (<1%) . This study tested the ability of five antibiotics, three beta-lactams and two macrolides, to select resistant mutants in vitro in H. influenzae. No resistant mutants were selected by amoxicillin-clavulanate or cefpodoxime. Cefprozil MICs rose in one beta-lactamase-negative strain. This MIC increase for cefprozil was associated with two amino acid changes, Ala-271 to Val and Ser-357 to Asn in PBP 3. The contribution of the Ser-357-to-Asn mutation to resistance to beta-lactam antibiotics with decrease of beta-lactam affinity has already been shown with transformation by Ubukata et al. . The low frequency of selection of resistant mutants by the beta-lactam antibiotics tested in this study correlates with the low prevalence of BLNAR strains in the United States. However, in vitro selection of highly resistant mutants by azithromycin and clarithromycin does not correspond with the low prevalence of such resistance levels seen among clinical H. influenzae strains (0.5 and 1.9%, respectively) . Clarithromycin and azithromycin selected a total of 18 resistant mutants from 10 strains tested. Multistep azithromycin and clarithromycin exposure resulted in selection of resistant mutants in all 10 strains tested and 8 of these strains, respectively. Macrolide and azalide MICs for these mutants were 16.0 to >128.0 mug/ml. Such high MICs are not commonly seen in clinical H. influenzae isolates, and their clinical significance is unclear at the present time. To our knowledge, no resistance mechanisms have been described for highly macrolide-resistant H. influenzae strains. In this study we characterized mutations in 23S rRNA and in ribosomal proteins L4 and L22 in H. influenzae mutants for which the macrolide and azalide MICs are high. Macrolide resistance as a result of alterations in 23S rRNA have been found in different gram-negative and -positive species . For two resistant mutants selected by exposure to azithromycin, with A2059C and A2058G mutations in 23S rRNA, the MICs of azithromycin, clarithromycin, erythromycin, and clindamycin were >128 mug/ml. The mutations found in ribosomal proteins L4 and L22 were in the highly conserved region of these proteins (Fig. and ). These mutations probably cause a conformational change in the ribosome that decreases the affinity of macrolides. Resistance as a result of mutations in L4 and L22 have been reported in in vitro mutants of E. coli . L4 mutation has been detected as the cause of resistance in clinical isolates of S. pneumoniae . In laboratory mutants of S. pneumoniae, alterations in ribosomal protein L4 and L22 conferring resistance to macrolides were also shown . This study showed that mutations in ribosomal proteins in L4 and L22 in H. influenzae were associated with increases in the MICs of erythromycin, clarithromycin, and azithromycin and contributed to increases of macrolide MICs in H. influenzae Rd strain, as observed in transformation studies. However MICs of clindamycin and quinupristin-dalfopristin were not raised in all of these mutants. Susceptibility to quinupristin-dalfopristin seems to be affected by alterations in ribosomal proteins L4 and L22 far more than to clindamycin. In some mutants a fourfold decrease in clindamycin MICs was observed. The significance and explanation for these lower clindamycin MICs are unknown at this time. In clinical strains of S. pneumoniae, mutations in L4 do not affect the susceptibility of the strains to clindamycin . The isolation of an H. influenzae clone that was hypersusceptible to macrolides is of great interest. The mechanism of macrolide susceptibility in this hypersusceptible strain is currently under investigation. For eight resistant mutants no mechanism of resistance was detected. The absence of modifications in the studied portion of 23S rRNA and in ribosomal proteins L4 and L22 shows that other ribosomal regions or proteins are likely to be involved in macrolide resistance. Investigation of the mechanisms of resistance are being conducted in our laboratory. In summary, repeated daily subcultures in subinhibitory concentrations of amoxicillin-clavulanate and cefpodoxime did not lead to development of resistant clones, although cefpodoxime MICs of some strains rose to the levels seen in low-BLNAR strains. Only one resistant mutant was selected by cefprozil from a beta-lactamase-negative strain. Sequential subculture in azithromycin and clarithromycin led to extremely high MICs, and in some strains these increases in MICs were associated with alterations in 23S rRNA, or in ribosomal proteins L4 or L22. One initially resistant mutant strain became hypersusceptible to macrolides during subculture without antibiotic. The clinical significance of these findings is not known, and this aspect is currently being investigated by our group. Backmatter: PMID- 12183267 TI - Surveillance for Antiviral-Agent-Resistant Herpes Simplex Virus in the General Population with Recurrent Herpes Labialis AB - In a general population survey in the United States, the prevalence of antiviral-agent-resistant herpes simplex virus was very low among more than 1,000 isolates from individuals with an episode of recurrent herpes labialis not treated with topical antiviral agents. Two isolates had borderline resistance to acyclovir (0.2%), and all were susceptible to penciclovir. Keywords: Introduction : Between October 1998 and February 1999, a study was undertaken in 10 states across the United States to determine the background prevalence of herpes simplex virus (HSV) resistant to acyclovir or penciclovir among isolates from subjects with an episode of recurrent herpes labialis (RHL) which had not been treated with a topical antiviral agent. Results from other studies with immunocompetent patients with HSV infection have shown that the prevalence of resistant virus is low (, , ; M. Reyes, J. Graber, N. Weatherall, C. Hodges-Savola, W. C. Reeves, and the Task Force on Herpesvirus Resistance, 11th International Conference on Antiviral Research, April 1998, Antivir. Res. 37:A44 [abstract], 1998) and stable, despite increasing rates of use of antiviral medication. Topical penciclovir cream was approved by the Food and Drug Administration in 1996 as a prescription product for the treatment of RHL. Topical acyclovir cream for the treatment of RHL is available in many markets, often without prescription, but has not been approved for use in the United States. The survey formed part of a program to support an application to switch penciclovir from a prescription product to a nonprescription product (one that could be sold over the counter [OTC]). It was anticipated that approval for OTC use will increase the levels of use of penciclovir, and therefore, any future change in the prevalence of resistant virus could be tracked against a background prevalence established prior to the switch to OTC status. At the time of implementation, the survey was sponsored jointly by two pharmaceutical companies and was planned in collaboration with the Task Force on Herpes Simplex Virus Resistance. (These data were presented at the Thirteenth International Conference on Antiviral Research, Baltimore, Md., April 2000.) The study was conducted in accordance with good clinical practice and received Essex Institutional Review Board approval. Pharmacists in 47 community-based pharmacies were responsible for the clinical phase of the study . The subjects volunteered for the study in response to advertisements placed in the media . Eligible subjects were aged >=12 years and had perioral herpes labialis at the vesicle or soft ulcer stage. No topical treatment with an antiviral agent was permitted prior to completion of the study, although subjects who received any oral antiviral agent for HSV infection were not excluded. Efforts were made to identify individuals who may have been immunocompromised by asking subjects to complete one page of the case report form, which was then sealed in an envelope by the subject to maintain confidentiality and subsequently transferred to the data management team. Subjects responded to questions on medical history (history of cancer, human immunodeficiency virus infection [HIV] or AIDS, or organ transplantation) and completed a checklist of medications that they were taking, use of which may have implied that the subject was immunocompromised. The checklist specified drugs with immunosuppressant activity or drugs used to treat HIV infection or AIDS or other diseases associated with defects of the immune system. A total of 1,803 subjects were recruited, and each subject was swabbed once. The study population (n = 1,795 subjects with swabs processed for virus isolation) was principally female (67.6%), with a mean age of 37.5 years (standard deviation [SD], 15.34 years). Caucasians formed the largest racial group (89.9%). The preponderance of female Caucasians in the study may reflect that sector of the population that is concerned about RHL and/or that is motivated to seek treatment . The mean +- SD number of episodes experienced during the previous 12 months was 5.2 +- 5.26, and the mean age at the first episode was 13.1 years. The mean +- SD duration of an episode of RHL was reported to be 8.1 +- 4.8 days. Lesions were swabbed by the pharmacist. The swabs were stored in transport medium at 4C and shipped to a central laboratory within 3 to 4 days. Virus was isolated in primary rabbit kidney cells . A cell suspension prepared from each infected culture was used to type the virus by indirect immunofluorescence. Vero cells and MRC-5 cells were used for the plaque reduction assays with acyclovir and penciclovir, respectively. Monolayers in six-well plates were infected with 50 to 100 PFU per well. After adsorption, the inoculum was removed and replaced with culture medium containing methylcellulose (2%) and the test antiviral agent (final concentrations, 0 to 30 mug/ml, tested in duplicate). After incubation for 3 days at 36 +- 1C in 5 to 7% CO2, the monolayers were fixed and stained, the plaques were counted, and the 50% inhibitory concentrations (IC50s) were calculated. Two standard HSV type 1 (HSV-1) reference strains were included in all assays, SC16 (which is sensitive to acyclovir and penciclovir) and DM21 (a thymidine kinase-negative mutant which is resistant to acyclovir and penciclovir). Summary IC50 data for these strains are provided in Table . TABLE 1 | Summary of HSV-1 susceptibility to acyclovir and penciclovir by antiviral use and medical history The 1,803 cultures yielded 1,087 isolates, all of which were typed as HSV-1. A total of 1,002 isolates were tested for their susceptibilities to acyclovir and penciclovir, and 2 additional isolates were also tested for their susceptibilities to penciclovir. The breakpoint for defining resistance to acyclovir was an IC50 >=2 mug/ml . Two criteria were applied to define penciclovir resistance: (i) an IC50 >=2 mug/ml and (ii) an IC50 three or more times greater than the mean IC50 of penciclovir for all isolates in the survey (>=0.84 mug/ml). Two HSV-1 isolates were confirmed to be resistant to acyclovir (IC50s, 3.21 and 2.41 mug/ml, respectively), but resistance was borderline in this assay and in the original assay (IC50s, 2.21 and 4.35 mug/ml, respectively). Both isolates were sensitive to penciclovir (IC50s, 0.38 and 0.34 mug/ml, respectively). The two subjects with acyclovir-resistant HSV-1 isolates were immunocompetent; one subject had no history of antiviral agent use, and the other subject reported prior use of acyclovir (not during the episode evaluated for the present study). All other isolates were susceptible to acyclovir; therefore, the prevalence of acyclovir-resistant HSV-1 isolates was 0.20% (2 of 1,002 isolates tested; 95% confidence interval, 0.02 to 0.72%). A total of 1,004 HSV-1 isolates were tested for their susceptibilities to penciclovir, and none was found to be resistant (0 of 1,004 isolates tested; prevalence, 0.00%; 95% confidence interval, 0.00 to 0.37%). These results are consistent with data from a similar survey conducted in the United Kingdom at a time when topical acyclovir had been available OTC for 5 years and topical penciclovir had been available as a prescription medicine for 2 years . One resistant HSV-1 isolate which was cross resistant to acyclovir and penciclovir was identified (1 of 924 isolates tested for susceptibility to acyclovir; 1 of 915 isolates tested for susceptibility to penciclovir; prevalence, 0.1%) . Other surveys with immunocompetent populations, predominantly, patients with genital herpes, have reported a prevalence of HSV resistance of 0.1 to 0.7% (, ; Reyes et al., Antivir. Res. 37:A44 [abstract], 1998). The historical prevalence of acyclovir-resistant HSV isolates among untreated, immunocompetent patients, as measured by the plaque reduction assay, is 0.3% . Resistant HSV is more common in immunocompromised patients; for example, in a surveillance study by Reyes et al. (Reyes et al., Antivir. Res. 37:A44 [abstract], 1998), 4 of 62 (6.5%) isolates from HIV-infected patients were resistant to acyclovir. IC50 data from the present survey were analyzed to investigate differences in the susceptibilities of isolates from the study subjects to acyclovir and penciclovir on the basis of their reported use of antiviral agents, immunocompetence, or history of RHL . For isolates from subjects who reported use of antiviral medication to treat herpesvirus infections, the median IC50 was similar to the median IC50 for isolates from subjects who had never used antiviral medication both for acyclovir (P = 0.211) and for penciclovir (P = 0.649). For the small number of isolates (n = 49) from subjects who used oral acyclovir, famciclovir, or valacyclovir at the time of the study, the median IC50 of acyclovir (0.31 mug/ml) was significantly lower than that for isolates from subjects who were not using antiviral medication (0.43 mug/ml) (P = 0.025). This difference is not clinically important and may simply have arisen by chance; the parallel comparison for penciclovir showed no significant difference (P = 0.488). No other comparisons reached statistical significance, including that for immune status, although the number of subjects in this survey who may have been immunocompromised was low . In conclusion, the prevalence of resistant HSV-1 strains among isolates from the general population with RHL is low, consistent with earlier surveys with immunocompetent patients with HSV infection, despite the increasing rate of use of herpesvirus-specific antiviral agents over the past two decades. Backmatter: PMID- 12183272 TI - Efficacies of Quinupristin-Dalfopristin Combined with Vancomycin In Vitro and in Experimental Endocarditis Due to Methicillin-Resistant Staphylococcus aureus in Relation to Cross-Resistance to Macrolides, Lincosamides, and Streptogramin B- Type Antibiotics AB - A beneficial effect of the combination of quinupristin-dalfopristin and vancomycin was observed against two methicillin-resistant strains of Staphylococcus aureus harboring or not harboring the ermC gene, which codes for constitutive macrolide, lincosamide, and streptogramin B resistance. The beneficial effect was observed in time-kill studies, in which the drugs were used at inhibitory concentrations, and in a rabbit model of endocarditis, in which the combination was highly bactericidal and more active than monotherapies. Keywords: Introduction : Methicillin-resistant Staphylococcus aureus (MRSA) represents a major cause of nosocomial infection, and glycopeptides remain the standard therapy for the treatment of systemic infections due to such strains. However, the emergence of MRSA strains with intermediate susceptibilities to glycopeptides emphasizes the need for new therapeutic options that may include vancomycin in combination with another antibiotic. Quinupristin-dalfopristin (Q-D) is an injectable streptogramin that combines a type A streptogramin (dalfopristin) and a type B streptogramin (quinupristin). The combination of quinupristin and dalfopristin is synergistic and is active in vitro against MRSA . However, most strains of MRSA are cross resistant to macrolide, lincosamide, and streptogramin B (MLSB)-type antibiotics by methylation of the ribosomal target. The expression of MLSB resistance is more frequently constitutive than inducible in MRSA . When it is constitutive, strains are resistant to quinupristin but remain susceptible to Q-D, although the bactericidal activity of Q-D is reduced in vitro and in vivo . Therefore, the use of Q-D in combination with another antibiotic might be required for the treatment of severe infections due to strains resistant to quinupristin in order to increase the bactericidal activity and to prevent the emergence of resistance in vivo. Since Lorian et al. have shown that Q-D at a subinhibitory dose produced thickening of the cell wall, it was suggested that a positive interaction could be observed between Q-D and cell wall-active agents. Indeed, a synergistic interaction has previously been reported between Q-D and beta-lactams in vitro and in vivo . The combination of Q-D and vancomycin against strains of MRSA might therefore be of clinical interest in order to increase the bactericidal activity of Q-D and to decrease the risk of emergence of resistance to both antibiotics since strains resistant to Q-D have been selected in vivo . Moreover, studies done in vitro and in an infected fibrin clot model study suggested the benefit of this combination. However, the susceptibilities of the study strains to MLSB antibiotics were not mentioned. Two MRSA strains were evaluated in this study: HM1054, a clinical strain that is susceptible to quinupristin, and HM1054R, a strain that is resistant to quinupristin and that was obtained by conjugative transfer to HM1054 of the ermC gene, which codes for MLSB resistance . The two strains remained susceptible to Q-D (MIC, 1 mug/ml), whatever their profile of resistance to MLSB antibiotics (quinupristin MICs, 8 mug/ml for HM1054 and 64 mug/ml for HM1054R). The dalfopristin and vancomycin MICs for the two strains were 4 and 1 mug/ml, respectively. Brain heart infusion (BHI) agar and broth (Difco, Detroit, Mich.) were used for all experiments. Quinupristin, dalfopristin, and Q-D were from by Aventis, Vitry sur Seine, France, and vancomycin was from Lilly, Saint-Cloud, France. Time-kill studies were performed with overnight cultures diluted in fresh BHI broth to yield an inoculum of 106 CFU/ml. Antibiotics were used at concentrations achievable in human and rabbit sera . Q-D was used at concentrations of 1 and 4 mug/ml, and vancomycin was used at concentrations of 1, 8, and 32 mug/ml. After 0, 3, 6, and 24 h of incubation at 37C, serial dilutions of 100-mul samples were subcultured and incubated 24 h at 37C before the numbers of CFU were counted. Bactericidal activity was defined by a decrease in the original inoculum of at least 3 log10 CFU/ml. Aortic endocarditis was induced in rabbits as described previously . Twenty-four hours after catheter insertion, each rabbit was inoculated intravenously with 1 x 106 to 5 x 106 CFU of S. aureus in 1 ml of sterile saline. Untreated rabbits were killed at the start of therapy and served as control animals. For studies with strain HM1054, the killing of control animals and the start of therapy were performed 48 h after bacterial inoculation. For studies with strain HM1054R, this delay resulted in an extremely high concentration of bacteria in vegetations and the death of almost 80% of the rabbits before therapy was given. Therefore, in studies with this strain, the killing of control animals and the start of therapy were performed 36 h after inoculation of the bacteria. This resulted in comparable vegetation weights and bacterial concentrations in vegetations in the rabbits inoculated with the two strains . The animals were treated intramuscularly for 4 days with one of the following regimens: Q-D at 30 mg/kg of body weight every 8 h, vancomycin at 50 mg/kg every 8 h, or the combination of Q-D and vancomycin, which were injected at two different sites. The Q-D dosing regimen provided areas under the curve for quinupristin and dalfopristin comparable to those achieved in humans after intravenous injection of 7.5 mg of Q-D/kg . The vancomycin regimen produced peak (40 +- 8 mug/ml) and trough (12 +- 5 mug/ml) levels in serum comparable to those recommended for humans with severe infections . The animals were killed 8 h after the last antibiotic injection, as described previously . Vegetation homogenates were serially diluted and plated on agar to count the bacteria surviving after 24 h of incubation. To detect resistant mutants, vegetation homogenates were also plated on agar containing two- and fourfold the MIC of Q-D or vancomycin. The comparisons of the treatment effects were performed by nonparametric one-way analysis of rank scores for several independent samples (the Kruskal-Wallis test), followed by the Mann-Whitney test to study differences between the means among the treatments for a given strain. The proportions of sterile rabbits were compared by the Fisher exact test. Q-D at a concentration of 4 mug/ml and vancomycin at a concentration of 8 mug/ml were bactericidal against both strains (Fig. and ). However, the bactericidal activity of Q-D against HM1054R was less than that against HM1054 (reductions of 3.0 and 4.9 log10 CFU/ml after 24 h of incubation, respectively). The combination of Q-D and vancomycin at a low concentration (1 mug/ml) was bactericidal against HM1054, producing a reduction of 3.9 log10 CFU/ml after 24 h of incubation, whereas treatment with Q-D alone achieved a reduction of only 1.5 log10 CFU/ml and treatment with vancomycin alone resulted in bacterial regrowth . The combination of Q-D and vancomycin at 1 mug/ml did not achieve bactericidal activity against HM1054R (a reduction of 2.4 log10 CFU/ml after 24 h of incubation) but was more active than monotherapies, as shown by a reduction of 0.7 log10 CFU/ml with Q-D after 24 h of incubation and bacterial regrowth with vancomycin after 24 h of incubation . The combination did not add any benefit to the monotherapies against the two strains when concentrations of Q-D or vancomycin higher than 1 mug/ml were used in the combination (data not shown) . FIG. 1. | Time-kill experiments of the activities of Q-D at one and four times the MIC (Q-D1 and Q-D4, respectively) and vancomycin at one and eight times the MIC (Vm1 and Vm8, respectively) against S. aureus Time-kill experiments of the activities of Q-D at one and four times the MIC (Q-D1 and Q-D4, respectively) and vancomycin at one and eight times the MIC (Vm1 and Vm8, respectively) against S. aureus HM1054 (A) and S. aureus HM1054R (B). The MICs of Q-D and vancomycin for the two strains were each 1 mug/ml. The activities of the different antibiotic regimens against experimental endocarditis are shown in Table . Q-D and vancomycin alone were bactericidal against HM1054 and displayed similar activities. The combination of Q-D and vancomycin was highly bactericidal, with a reduction of 8.2 log10 CFU/g of vegetation compared with the counts for the controls (P < 0.001), and was significantly more active than vancomycin or Q-D alone (P < 0.05). In addition, the combination sterilized 9 of 10 animals (90%), whereas the monotherapies sterilized 2 of 20 animals (10%) (P < 0.001). TABLE 1 | Activities of vancomycin, Q-D, and their combination against two strains of S. aureus susceptible (HM1054) or resistant (HM1054R) to quinupristin after 4 days of treatment for rabbit endocarditis Q-D and vancomycin were similarly active against HM1054R, as determined by the reduction in growth compared to the growth in the controls, although Q-D did not achieve bactericidal activity in vivo. The combination of Q-D and vancomycin was highly bactericidal, allowing a significant reduction of 5.5 log10 CFU/g of vegetation compared with the control growth (P < 0.001). The combination was more active than Q-D or vancomycin (P < 0.05) and sterilized 5 of 13 animals (38%), whereas the monotherapies sterilized 1 of 22 rabbits (4.5%) (P < 0.05). No subpopulation with reduced susceptibility to Q-D or vancomycin was detected in vivo, regardless of the infecting strain and the regimen used. In the present work, we report that the combination of Q-D and vancomycin is more active than monotherapies in vivo in a severe model of infection due to S. aureus strains with or without the constitutive MLSB resistance phenotype in terms of bactericidal activity and the rate of sterilization. This result correlated with those obtained by in vitro time-kill studies. Nevertheless, it is of major importance to underline the fact that this benefit was observed in vitro only with low concentrations of Q-D and vancomycin (e.g., concentrations that may be achieved in deep foci of infection such as cardiac vegetations) and not with concentrations of Q-D greater than the MIC . We previously demonstrated that the penetration of quinupristin into vegetations is homogeneous, whereas dalfopristin showed a decreased concentration gradient between the periphery and the core of the vegetation . This result could explain why the addition of vancomycin to Q-D might allow an increased effect against both strains of bacteria localized in the core of the vegetation. However, even if the benefit of the effect of the combination was similar against the two strains in vivo, the antibacterial effect of the combination was less pronounced against HM1054R than against HM1054 since both Q-D and vancomycin were less active against HM1054R than against HM1054. This result was expected for Q-D but was not expected for vancomycin. The greater virulence of HM1054R may explain the in vivo phenotypic tolerance observed with vancomycin, as others have described previously . Backmatter: PMID- 12350235 TI - Prolonged exposure of chromaffin cells to nitric oxide down-regulates the activity of soluble guanylyl cyclase and corresponding mRNA and protein levels AB - Abstract | Background | Soluble guanylyl cyclase (sGC) is the main receptor for nitric oxide (NO) when the latter is produced at low concentrations. This enzyme exists mainly as a heterodimer consisting of one alpha and one beta subunit and converts GTP to the second intracellular messenger cGMP. In turn, cGMP plays a key role in regulating several physiological processes in the nervous system. The aim of the present study was to explore the effects of a NO donor on sGC activity and its protein and subunit mRNA levels in a neural cell model. Results | Continuous exposure of bovine adrenal chromaffin cells in culture to the nitric oxide donor, diethylenetriamine NONOate (DETA/NO), resulted in a lower capacity of the cells to synthesize cGMP in response to a subsequent NO stimulus. This effect was not prevented by an increase of intracellular reduced glutathione level. DETA/NO treatment decreased sGC subunit mRNA and beta1 subunit protein levels. Both sGC activity and beta1 subunit levels decreased more rapidly in chromaffin cells exposed to NO than in cells exposed to the protein synthesis inhibitor, cycloheximide, suggesting that NO decreases beta1 subunit stability. The presence of cGMP-dependent protein kinase (PKG) inhibitors effectively prevented the DETA/NO-induced down regulation of sGC subunit mRNA and partially inhibited the reduction in beta1 subunits. Conclusions | These results suggest that activation of PKG mediates the drop in sGC subunit mRNA levels, and that NO down-regulates sGC activity by decreasing subunit mRNA levels through a cGMP-dependent mechanism, and by reducing beta1 subunit stability. Keywords: Background : The soluble form of guanylyl cyclase (sGC) is the main receptor for the signaling agent nitric oxide (NO) . This signaling molecule performs important physiological and biochemical functions in the cardiovascular, pulmonary and neural systems, activating sGC and thus increasing cGMP levels . In certain conditions and disease states such as hypoxia and hypertension, a disturbance in sGC levels and/or activity may play a crucial role in the pathophysiology of these disorders . Moreover, desensitization of sGC may also be involved in tolerance to NO when this compound is used for therapeutic purposes . Purification of mammalian sGC yields a heterodimer comprised of two subunits alpha and beta of which four types exist (alpha1, alpha2, beta1, beta2) . Structurally, each subunit has a C-terminal cyclase catalytic domain, a central dimerization region and a N-terminal portion. This last portion constitutes the heme-binding domain and represents the least conserved region of the protein. Cloning and expression experiments have demonstrated that both alpha and beta subunits are required for sGC to be functionally active . In mammalian cells, two different heterodimeric enzymes, alpha1/beta1 and alpha2/beta1, have been detected, and although alpha1 and alpha2 differ in their primary structure, the two heterodimers were found to be functionally alike . Recently, it has been demonstrated the activation of beta2 subunit of sGC by NO and this enzyme has a monomeric structure . While much has been learned about the regulation of NO synthase , there is scare data on sGC regulation, despite its critical role in the actions mediated by endogenous or exogenous NO . Different reports have shown a decrease in sGC activity after pre-treatment of tissues or cells with NO-releasing compounds, or a higher sensitivity of the enzyme when endogenous NO synthesis is inhibited . However, the mechanism involved in this phenomenon remains unclear. Redox reactions may be a plausible mechanism for enzyme desensitization, as several studies have shown that the redox state of the enzyme bound heme and protein thiol groups has a major role in controlling enzyme activity . This activity can also be regulated by a phosphorylation/dephosphorylation mechanism , and there is emerging evidence that sGC activity is regulated both at the protein and mRNA levels. Several authors have reported that treatments, such as forskolin, dibutyryl-cAMP, 3-isobutyl-methyl xanthine , endotoxin and/or IL-1beta , NO donating compounds or nerve growth factor affect sGC mRNA levels in various cell types. The NO/cGMP pathway has been established as a major controller of several physiological functions of the nervous system . Moreover, the effects of NO/cGMP on neuronal differentiation and survival, and synaptic plasticity suggest that this signal transduction pathway regulates gene expression in the nervous system . Since the role of sGC in transducing inter- and intracellular signals conveyed by NO is pivotal, knowledge of the molecular mechanisms involved in sGC regulation may help our understanding of the physiological and pathophysiological significance of this signal transduction pathway in the nervous system. Compared with findings in vascular tissue, little is known about the effect of prolonged exposure of neural cells to nitric oxide on subsequent NO stimulation of these cells, and the capacity of this agent to elicit cGMP increases. The aim of the present study was to establish whether prolonged treatment of chromaffin cells with low doses of nitric oxide affects sGC activity in a widely used bovine neural cell model. The findings presented suggest that chronic exposure to NO decreases sGC activity by decreasing the availability of beta1 subunit protein levels, and the levels of the mRNAs that encode the alpha1 and beta1 subunits. Results : Desensitization of soluble guanylyl cyclase in cultured bovine chromaffin cells after prolonged DETA/NO treatment | Our first experiments were designed to determine the capacity of DEA/NO to increase cGMP in bovine chromaffin cells in order to select the concentration that produced the maximal stimulation. Figure shows that DEA/NO effectively increased cGMP in chromaffin cells in a concentration-dependent manner with a calculated EC50 value of 0.36 +- 0.02 mumol/L. Figure 1 | Concentration-dependence curves for DEA/NO-stimulated cGMP accumulation on chromaffin cells Concentration-dependence curves for DEA/NO-stimulated cGMP accumulation on chromaffin cells. Cells were incubated for 30 min in Locke's solution containing 0.5 mM IBMX. Then they were stimulated for 10 min with the indicated DEA/NO concentrations and cGMP measured by radioimmunoassay. Results expressed as pmol /106 cells, corresponding to the mean +- standard error of three experiments performed in triplicate, with different cultures. DEA/NO was freshly prepared every day in NaOH 0.01 N and its concentration checked by measuring the absorbance at 250 nm. Data were fit to sigmoidal representation by using Microcal Origing program. To study the effect of prolonged treatment with NO on sGC DETA/NO (50 mumol/L) was added to the culture medium, and at several time points beginning at 2 hours the medium was withdrawn and replaced with Locke's solution containing 0.5 mmol/L of the nonselective phosphodiesterase inhibitor IBMX. After 30 min of incubation, the cells were stimulated with 1 mumol/L DEA/NO or 250 mumol/L YC-1 for 10 min and the intracellular cGMP content determined. As illustrated in Fig. , exposure to DETA/NO resulted in a significant time-dependent decrease in DEA/NO- or YC-1-stimulted cGMP production by the chromaffin cells. Maximal reduction was achieved after 24 --48 hours of treatment. Exposure of the chromaffin cells for 24 or 48 hours to 50 mumol/L DEATA/NO induced no cell toxicity. The percentage of cells that failed to exclude EthD-1 did not differ in chromaffin cells incubated in the presence or absence of DETA/NO. Neither was a difference noted in the fluorescence due to calcein, an indicator of living cells . Figure 2 | Pretreatment of chromaffin cells with DETA/NO causes a time-dependent down-regulation of cGMP accumulation in response to subsequent NO stimulation (A) or YC-1 stimulation (B) without affecting their viability (C) Pretreatment of chromaffin cells with DETA/NO causes a time-dependent down-regulation of cGMP accumulation in response to subsequent NO stimulation (A) or YC-1 stimulation (B) without affecting their viability (C). Bovine chromaffin cells were pretreated with 50 mumol/LDETA/NO for the times indicated. Cells where then incubated with 0.5 mmol/L IBMX for 30 min and stimulated with 1 mumol/L DEA/NO or 250 mumol/L YC-1 for 10 min. cGMP was determined by radioimmunoassay as described in Methods. 1C. Cells were incubated with 50 mumol/L DETA/NO for 48 hours and then incubated for 30 min in Locke's solution in the presence of the viability/cytotoxicity probes EthD-1 and calcein-AM. Images were taken under a fluorescence microscope using an excitation optical filter and two different emission filters: a 530 nm bandpass filter for calcein (upper panels) and a 590 nm longpass filter for EthD-1 (lower panels). Prolonged exposure to NO might alter the intracellular redox state of these cells and sGC contains critical thiol groups in its catalytic site that are sensitive to oxidative inactivation. Reduced glutathione (GSH) is the major intracellular redox buffer in most cell types. To determine whether the reduction in sGC activity caused by NO treatment could be attributed to a drop in reduced intracellular glutathione, a permeable, reduced glutathione (GSH-ME) was added simultaneously with the DETA/NO. As shown in Fig. , the amount of cGMP elicited by DEA/NO in cells treated with GSH-ME was similar to that produced in control cells. Moreover, in cells treated with DETA/NO, the presence of this compound neither modified subsequent DEA/NO stimulation. This indicates that the effect of DETA/NO may not be ascribed to a decrease in intracellular sulfhydryl groups. However, when GSH synthesis was blocked by the addition of 0.5 mmol/L BSO, both DETA/NO-treated and untreated cells showed a clear reduction in the DEA/NO-stimulated cGMP increases (about 35% or 42% respectively). Figure 3 | Effect of intracellular levels of reduced glutathione (GSH) on basal (A) or NO-stimulated (B) cGMP accumulation in control or DETA/NO pretreated cells Effect of intracellular levels of reduced glutathione (GSH) on basal (A) or NO-stimulated (B) cGMP accumulation in control or DETA/NO pretreated cells. Cells were incubated with the membrane permeable reduced glutathione (GSH-ME) (2 mmol/L) or with 0.5 mmol/L buthionine-(S,R)-sulfoximine (BSO) for 48 hours in the presence or absence of 50 mumol/L DETA/NO. The cells were then incubated for 30 min with 0.5 mmol/L IBMX and basal cGMP levels (A), or DEA/NO-stimulated cGMP levels (B) determined by radioimmunoassay. Significant differences from values obtained in DETA/NO-untreated cells (empty column) are marked: ***p < 0.001; **p < 0.01; *p < 0.05 (Paired student's t test). Significant differences from values obtained in control cells are marked: +p < 0.05 (Paired student's t test). DETA/NO treatment reduces sGC beta1 subunit protein levels | Western blot experiments were performed to establish whether the depressed sGC activity in cells treated with DETA/NO could be due to changes in protein levels. Figure shows a western blot incubated with anti-sGC antibody. This antibody recognized two bands in chromaffin cell extracts, a significant band at ~78 --80 kDa was consistently present together with a weaker band at 70 kDa. The immunodetection of these bands was specifically prevented by the presence of the corresponding immunizing peptide (alpha1 or beta1). According to these results these two bands likely corresponded to the alpha1 and beta1 subunits of sGC, although nonspecific binding of the antibody would not be discard. As shown in Fig. , treatment with DETA/NO caused a time-dependent decrease in the amount of sGC beta1 subunit. While in some experiments it was also possible to detect a decrease in the quantity of alpha1 subunit at the longest exposure time, the fall in alpha1 subunit level was not always evident. Thus, taking together the results of all the experiments the changes in alpha1 subunit were not significantly different. The possibility that the observed reduction in sGC protein levels might be attributable to differences in the amounts of proteins loaded was ruled out, since the loading control alpha-actin, remained unchanged in the different cell extracts. The fall in the amount of subunit beta1 was not prevented by the presence of the protein synthesis inhibitor cycloheximide indicating that protein synthesis is not required for the sGC beta1 subunit decrease induced by DETA/NO. On the contrary, as shown in Figure , simultaneous treatment with DETA/NO and cycloheximide reduced the beta1 subunit level drastically making it practically undetectable in our experimental conditions. Figure 4 | DETA/NO pretreatment decreases sGC subunit protein levels in bovine chromaffin cells DETA/NO pretreatment decreases sGC subunit protein levels in bovine chromaffin cells. A.- Immunodetection of sGC in soluble extracts from bovine chromaffin cells. Twenty micrograms of protein and 5 muL of molecular weight markers (prestained precision protein standards from BIO RAD) were electrophoresed and transferred to PVDF membranes and then sGC was detected by using 1 mug/mL of anti-sGC antibody or the antibody plus 80 ng/mL peptide beta1 or antibody plus 60 ng/mL peptide alpha1.) B.- Cells were incubated with 50 mumol/L DETA/NO for the times indicated (0, 2, 4, 8, 24, 48 and 72 hours). Cells were then lysed and equal amounts of proteins (30 mug) corresponding to the soluble fraction of cell extracts fractionated by SDS-PAGE and transferred to PVDF membranes. The two sGC subunits (alpha1 and beta1) were immunodetected using a polyclonal antibody as described in Methods. a-Actin was immunodetected to check for consistency in loading and transfer. Figure 5 | The half-life of sGC activity was diminished in DETA/NO treated cells The half-life of sGC activity was diminished in DETA/NO treated cells. (A) Chromaffin cells were incubated for different times (0, 4, 8,16, 24,32 and 48 hours) with either 10 mumol/L cycloheximide , 50 mumol/L DETA/NO or with both drugs . Cells were then washed preincubated with 0.5 mmol/L IBMX for 30 min and stimulated with 1 mumol/L DEA/NO for 10 min and cGMP levels determined by radioimmunoassay. Data of DEA/NO-stimulated cGMP levels as a time function are represented in a linear model according to the equation logA = k.t/2.3 + logA0, where A is the DEA/NO-stimulated cGMP level at each time and A0 is the DEA/NO-stimulated cGMP level in control cells. From the slopes, the "k " and half-life (t1/2) of sGC activity in each condition can be calculated according to the expression t1/2 = 2.3 log2/k. (B) Immunoblotting analysis of sGC subunits (alpha1 and beta1) or alpha-actin. Cells were pretreated with either: vehicle (control), 10 mumol/L cycloheximide (CHX), 50 mumol/L DETA/NO, or both compounds together. Proteins (30 mug) from each group were subjected to electrophoresis and transferred to PVDF membranes. The membranes were processed as described in Fig. . (C) Cell viability after different treatment. Cells were incubated for 48 hours with either: vehicle (control), 50 mumol/L DETA/NO, 10 mumol/L cycloheximide or 50 mumol/L DETA/NO plus10 mumol/L cycloheximide. After that cells were washed and their viability analyzed as described in Methods. The values are given as the percentage of live cells and are the mean +- SE of three experiments performed in triplicate. Decreased steady-state protein levels can be the result of inhibited protein synthesis or of a decrease in the half-life of a protein. The amount of cGMP produced in response to NO was determined in cells treated with DETA/NO for different lengths of time . Since these determinations were performed in the presence of a broad-spectrum phosphodiesterase inhibitor the variation in NO-stimulated cGMP might reflect changes in sGC activity. A half-life of 30.6 +- 1 hour was calculated for sGC activity in cells pretreated with DETA/NO. In the presence of the protein synthesis inhibitor (cycloheximide), the estimated half-life was 55 +- 3.6 hours and in the combined presence of DETA/NO and cycloheximide this value dropped to 23.4 +- 1.6 hours. This capacity of DETA/NO to induce an even greater reduction in sGC activity when protein synthesis was blocked by cycloheximide indicates that DETA/NO decreases the stability of sGC via a mechanism that is independent of protein synthesis. The reduction in the cGMP increased elicited by NO stimulation may not be explained by a cytotoxic effect of these treatments because as shown in Fig. they did not modify the number of live cells. To explore the mechanisms responsible for the effect of NO on sGC beta1 subunit levels, chromaffin cells were treated with 10 mumol/L ODQ, an inhibitor of sGC, or with 15 mumol/L Rp-8-Br-PET-cGMPS, an inhibitor of PKG, 30 min before exposure to DETA/NO. Inhibition of sGC by ODQ did not prevent the DETA/NO-induced decrease in beta1 subunit levels . However, in the presence of ODQ alone, the beta1 subunit of sGC showed a similar drop in level to that shown by cells incubated with DETA/NO or DETA/NO plus ODQ. Moreover, the presence of a PKG inhibitor partially prevented the beta1 subunit down regulation caused by DETA/NO. These results suggest the involvement of PKG activation in the mechanism underlying the decrease in sGC beta1 subunit levels observed when cells are chronically exposed to NO. To verify that inhibition of the effects of DETA/NO was not caused by the presence of the cGMP analog rather than PKG inhibition, another cGMP analog and PKG activator, Sp-8-Br-PET-cGMPS, was added to the cell culture. This compound failed to prevent the DETA/NO-mediated effect. Figure 6 | Inhibition of PKG partially prevents the DETA/NO-induced down-regulation of beta1 subunit protein levels Inhibition of PKG partially prevents the DETA/NO-induced down-regulation of beta1 subunit protein levels. Cells were incubated with either: vehicle (control), 50 mumol/L DETA/NO, 10 mumol/L ODQ, 10 mumol/L ODQ plus 50 mumol/L DETA/NO, 15 mumol/L Rp-8-BrPET-cGMPS (Rp-PET), 15 mumol/L Rp-8-BrPET-cGMPS (Rp-PET) plus 50 mumol/L DETA/NO, or 10 mumol/L Sp-8-BrPET-cGMPS (Sp-PET) plus 50 mumol/L DETA/NO for 48 hours. ODQ, Rp-PET and Sp-PET were added 30 min before DETA/NO. Proteins (30 mug) from each group were subjected to electrophoresis and blotting. sGC was immunodetected and the band corresponding to the beta1 subunit quantified. Results are expressed as the percentage (%) of the amount of beta1 subunit in cells subjected to different treatment with respect to the amount in untreated cells. The figure shows the results given as mean +- SE of five experiments performed with different cultures. Significant differences compared with non-treated cells (control) is indicated by ***p < 0.001 (Student's t test). ++p < 0.01 represents significant difference between DETA/NO-treated cells and DETA/NO plus Rp-PET-treated cells. DETA/NO treatment causes a drop in mRNA levels corresponding to sGC subunits alpha1 and beta1 | The effect of prolonged exposure to NO on the mRNA levels of both the alpha1 and beta1 subunits of sGC was investigated using a quantitative RT-PCR technique. Fig. presents a representative agarose gel showing the amplified PCR products obtained with the primers designed for the alpha1 and beta1 subunits of sGC and the commercial primers for the 18S rRNA. Each specific primer was used to amplify single 350, 450 and 300 pb fragments. Quantitative PCR experiments revealed that treatment of chromaffin cells with 50 mumol/L DETA/NO for 16 hours leads to a reduction in the mRNAs that code for both sGC subunits (Fig. and ). The presence of two PKG inhibitors (Rp-8-Br-PET-cGMPS and KT5823 at concentrations that only inhibited PKG activity in "in vitro" experiments) effectively abolished the effect of DETA/NO. In cells pretreated with 1 mumol/L KT5823, the DETA/NO-induced effect was totally prevented and levels of alpha1 and beta1 subunit mRNAs were higher than in control cells. The application of Rp-PET-8-Br-cGMPS at three concentrations had a biphasic effect: at a concentration of 3 mumol/L, it effectively blocked the DETA/NO response and increased the levels of the two mRNAs by 1.5 fold with respect to control cells; yet as the concentration of the PKG inhibitor was increased, the blocking effect was still produced but the increases produced in mRNA levels were smaller. At the highest concentration tested (30 mumol/L), blockage of the response to DETA/NO was incomplete; alpha1 and beta1 mRNA levels were 80 and 90% of control levels respectively. Addition of the PKA inhibitor, H-89, at a concentration that selectively inhibits this protein kinase, without affecting others, was unable to prevent the down-regulation of mRNA caused by DETA/NO treatment. These results might suggest the involvement of PKG in DETA/NO-induces down-regulation of sGC subunits mRNAs. Additional experiments where performed in order to discard a possible cross-talk activation of PKA involved in this mechanism. As shown in Fig. the amount of alpha1 and beta1 mRNAs was increased 1.6 and 1.9 fold, respectively when cells were incubated with the cAMP analogue Sp-cAMPS-AM. This effect was effectively counteracted by co-treatment with H-89. Moreover H-89 alone was able to decrease the alpha1 and beta1 mRNAs as compared to control indicating that tonic activity of PKA is necessary to keep the levels of these mRNAs balanced. Nevertheless the cGMP analogue Sp-PET-8Br-cGMPS mimicked the effect observed with DETA/NO and caused a reduction of 40% of both mRNAs supporting the idea that PKG activation down-regulates the mRNAs encoding for sGC subunits (alpha1 and beta1). Figure 7 | DETA/NO treatment causes down regulation of mRNAs coding for the alpha1 and beta1 subunits of sGC DETA/NO treatment causes down regulation of mRNAs coding for the alpha1 and beta1 subunits of sGC. PKG inhibitors prevent this effect. (A) Gel showing the analysis of RT-PCR products for the alpha1- and beta1-subunits of sGC and for 18S ribosomal RNA (used as internal control) by agarose gel electrophoresis stained with SYBR gold. Total RNA (200 ng) was reverse transcribed and amplified for 40 cycles. A marker (M) consisting of DNA fragments in increments of 100 bp (lower band, 100 bp) was used to estimate the size of the PCR products. (B and C) Variation in mRNA content for sGC alpha1 and beta1 mRNAs in cells subjected to different treatments. Chromaffin cells were incubated with a vehicle (control), or 50 mumol/L DETA/NO alone or in the presence of the PKG inhibitor Rp-8-BrPET-cGMPS at the concentrations 3 mumol/L, 15 mumol/L or 30 mumol/L or 1 mumol/L KT5823, or in the presence of a PKA inhibitor (1 mumol/L H89). The protein kinase inhibitors were added 30 min before the DETA/NO. After 16 hours treatment, total RNA was isolated and used for quantitative RT-PCR using specific primers designed for mRNAs encoding bovine sGC alpha1 and beta1 subunits. The figure shows the results of four experiments performed in triplicate given as mean +- SE. Significant differences from control are indicated by **p < 0.01; *p < 0.05 (Student's t test). ++p < 0.01;+p < 0.05 (Student's t test) indicates significant differences between cells treated with DETA/NO alone or with DETA/NO plus PKG inhibitors. Figure 8 | cAMP up-regulates alpha1 and beta1 sGC mRNAs in bovine chromaffin cells cAMP up-regulates alpha1 and beta1 sGC mRNAs in bovine chromaffin cells. Cells were incubated during 24 hours with either: vehicle (control), 100 mumol/L Sp-cAMPS-AM, 100 mumol/L Sp-cAMPS-AM plus 1 mumol/L H-89 (this compound was added 30 min before the addition of cAMP analogue), 1 mumol/L H-89, or 30 mumol/L Sp-PET-8Br-cGMPS. After treatment, total RNA was isolated and used for quantitative RT-PCR using specific primers designed for alpha1 (A) and beta1 (B) subunits. Results are expressed as mean +- SE of two experiments performed in triplicate. Significant differences from control are indicated by *p < 0.05, **p < 0.01. ++p < 0.01 indicates significant differences between cells treated with Sp-cAMPS-AM and cells treated with Sp-cAMPS-AM plus H-89. Discussion : Prolonged treatment of chromaffin cells with DETA/NO led to desensitization of sGC to subsequent NO or YC-1 stimulation. This depressed sGC activity was accompanied by a decrease in sGC protein levels, involving diminished levels of the beta1 subunit and the mRNAs that code for the alpha1 and beta1 subunits. It has been previously described that prolonged treatment with NO causes a reduction in the intracellular levels of reduced thiols and that this can lead to the inactivation of several proteins . However, when the permeable, reduced glutathione was added to the culture medium, the effect of DETA/NO on sGC activity was not prevented. This indicates that the intracellular levels of reduced glutathione in these cells were not likely to be affected by the concentration of NO employed in the present study, which is ten times lower than that used in other studies . However, intracellular levels of reduced glutathione clearly affect sGC activity, because when cells were depleted of the GSH by treatment with an inhibitor of its synthesis (BSO), the NO-stimulated cGMP increase was significantly diminished both in cells treated with BSO alone and in cells treated with DETA/NO plus BSO, compared to their respective controls. The participation of several thiol groups in sGC activity has been described in several reports . The thiol groups of cysteine residues are particular important for the structural and functional properties of proteins. Reducing conditions in the intracellular environment generally prevent the oxidation of sulfhydryl groups, and site-directed mutagenesis experiments have shown that the beta1 subunit has two cysteine residues (Cys-78 and Cys 214) required for heme binding and thus for NO sensitivity . Moreover, the catalytic site of guanylyl cyclase contains critical thiol groups involved in its activation by NO that are sensitive to oxidative inactivation . In this way, it has been demonstrated that NO activation of sGC leads to disulfide formation, which reversibly inactivates the enzyme . In agreement with previous findings, we observed that prolonged treatment of chromaffin cells with NO leads to a reduction in the amount of sGC subunit beta1. In contrast the amount of the protein identified as the alpha1 was not modified in our experimental conditions. This apparent discrepancy could be explained by a mistaken identity of the alpha1 subunit due to a nonspecific binding of the antibody to a protein distinct from this subunit although its detection was almost completely avoided by the presence of the alpha1 peptide used as antigen. On the other hand in the present work as well as in previous work we have shown an apparent molecular weight for the large sGC subunit of 80 kDa. This value is higher to that previously reported by Humbert et al (1990) . The discrepancy between the present and formerly reported molecular weight of the sGC alpha1 subunit might be explained by the existence of different gene products with specific tissue distribution, because alternative splicing for the alpha1 subunit has been demonstrated . The time-course experiments revealed a maximal effect on beta1 subunit levels after 24 or 48 hours of treatment. Nitric oxide appears to decrease the levels of this subunit via a translation-independent mechanism. This conclusion is based on the observation that beta1 subunit levels decreased more rapidly in chromaffin cells exposed to NO than in those exposed to cycloheximide. Moreover, pretreatment of chromaffin cells with cycloheximide did not impede NO altering sGC beta1 subunit levels; on the contrary, a greater decrease was recorded. These results are in clear contrast with several reports showing a transduction- and translation-dependent mechanism involved in the sGC down-regulation caused by NO . A long half-life has been proposed for sGC in physiological conditions. We established a half-life of 55 hours for sGC activity by inhibiting protein synthesis in the present model. It should be noted that the beta1subunit levels but not those of the alpha1 subunit were reduced in extracts from cells treated for 48 hours with cycloheximide, indicating that the estimated half-life reflects that of the beta1 subunit. Further, the half-life estimated for the beta1 subunit in extracts of DETA/NO-treated cells, correlates well with the sGC activity calculated for DETA/NO-treated cells (results not shown). In contrast, whereas NO treatment decreased sGC alpha1 and beta1 subunit mRNA levels, the decrease at the protein level only occurred in the beta1 subunit. This finding might reflect a higher stability of the alpha1 subunit than its counterpart, as previously suggested , at least during continuous exposure to NO. The drop in beta1 subunit levels triggered by NO treatment might be the result of a dual action of NO; decreasing the levels of mRNA that encode the beta1 subunit, on the one hand, and destabilizing the subunit protein making it more susceptible to proteolytic cleavage, on the other. Destabilization may occur through nitrosylation of cysteine or tyrosine residues, but also, prolonged exposure to NO could trigger oxidation of the ferrous component of the heme group and oxidation of different residues, mostly cysteine residues. Any of these modifications could induce conformational changes in the sGC structure making it unstable, although the fact that ODQ treatment also led to a decrease in beta1 subunit levels suggests that oxidation might be the most plausible mechanism. Indeed, it has been demonstrated that ODQ inhibits sGC activity by oxidation of the iron atom of the heme group, which changes from the ferrous to the ferric state . This oxidation leads to heme removal from sGC and thus the enzyme loses its sensitivity towards NO. The long-term effect of ODQ has not been explored to date, but it is known to oxidize different molecules, including sGC. Although this is not likely to be an exclusive mechanism for regulating sGC, it is clear that this enzyme is very sensitive to oxidizing environments . Consistent with this observation is the fact that the antioxidant, vitamin E, attenuates vascular superoxide anion formation and increases sGC beta1-subunit levels in rat aortic tissue . The observed effect of NO on mRNA levels and its partial effect on protein levels was mediated by a cGMP and PKG-dependent mechanism. Two different specific and selective PKG inhibitors, at a concentration that affected no other protein kinase at least in "in vitro "experiments, were able to abolish the effect of NO on the expression of the sGC subunit gene. Other investigators have also suggested that cGMP regulates sGC subunit gene expression. Ujiie et al. observed that incubation of cultured rat medullary interstitial cells with SNP decreased sGC a1 and beta1 subunit mRNA levels. Papapetropoulos et al. also observed a reduction in sGC subunit gene expression after incubation of aortic smooth muscle cells with zaprinast, an inhibitor of cGMP degradation by type V phosphodiesterase. Similarly, Filippov et al. reported decreased levels of mRNAs encoding sGC subunits in rat pulmonary artery smooth muscle cells after NO or 8-Br-cGMP treatment and suggested the involvement of PKG in the underlying mechanism. Here, we showed that the PKG inhibitor, Rp-8-Br-PET-cGMP, was able to attenuate the ability of NO to alter sGC beta1 subunit levels and to fully inhibit its effect on sGC subunit gene expression. At low concentrations, Rp-8-Br-PET-cGMPS also increased sGC alpha1 and beta1 mRNA levels. The same effect was observed with 1 mumol/L KT5823. Several reports have shown that cAMP-increasing compounds and activation of PKA leads to a down-regulation of sGC in smooth muscle cells . In addition, the activation of tyrosine kinases down-regulates sGC levels in PC12 cells . In the present model, the PKA inhibitor (H-89 at 1 muM) was not able to prevent the DETA/NO effect on sGC mRNA levels. On the contrary, this compound caused reduction of the levels of the two mRNAs and counteracted the effect of the cAMP analogue. These findings could indicate that PKA positively modulates mRNA levels of the sGC subunits in these cells and similarly explain the observed biphasic effect of Rp-8-Br-PET-cGMPS (this compound also inhibits PKA with a Ki of 11 mumol/L). At 30 mumol/L in theory PKG should be inhibited by 95% and PKA by around 50%. These results clearly contrast with previous findings showing that PKA negatively modulates sGC levels in smooth muscle cells . PKG I expression is also suppressed by continuous exposure to nitrovasodilators, theophylline, cGMP and cAMP in smooth muscle cells , and activation of either PKG or PKA leads to a similar physiological response in this model. Thus, it is conceivable that if PKG levels fall, PKA activation is the downstream element in the regulatory mechanism triggered by NO . In chromaffin cells, however, these two kinases have opposite effects and PKG levels remaining unchanged in DETA/NO exposed cells. Conclusions : In conclusion, exposure of chromaffin cells to low concentrations of NO leads to reduced cGMP accumulation in response to subsequent exposure to NO. The observation that sGC activity is highly sensitive to conditions of prolonged exposure to NO may have significant implications in neural disorders, such as brain trauma or ischemia, where NOSII is induced and there is a continuous production of NO. While further work is needed to evaluate the extent to which the decrease in sGC level and activity contributes to the pathophysiological response, it may be speculated that a significant reduction in sGC would affect all cGMP-mediated processes. Methods : Materials | IBMX, BSO and anti-alpha-actin (20 --33) were purchased from Sigma Chemical Co. (St. Louis, MO, USA.). KT-5823 and H-89 were from Calbiochem (San Diego, CA, USA). DETA/NO, DEA/NO, rabbit anti-GC (soluble) and guanylate cyclase (soluble) blocking peptide were provided by Cayman Chemical (Ann Arbor, MI, USA). YC-1 was obtained from Alexis Corporation (San Diego, CA, USA.) and dissolved in DMSO. ODQ was from Tocris Cookson (Langford, Bristol, U.K.). Adenosine-3',5'-monophosphorothioate, Sp-isomer; acetoxymethyl ester (Sp-cAMPS-AM), beta-Phenyl-1, N2-etheno-8-bromoguanosine-3',5'-cyclic monophosphorothioate, Sp and Rp isomers (Sp- and Rp-8-Br-PET-cGMPS) were from Biolog (Bremen, Germany). The Live/Dead Viability/Cytotoxicity Kit was from Molecular Probes Europe (Leiden, The Netherlands). TaqMan Reverse Transcription reagents, SYBR Green PCR Core reagents kit, MicroAmp Optical 96-well reaction plates and MicroAmp optical caps were from Applied Biosystems (Foster City, CA, USA). The QuantumRNATM 18S internal standard (relative RT-PCR) was obtained from Ambion (Austin, Texas, USA). Methods | Isolation and culture of bovine adrenal chromaffin cells | Cells were obtained after digestion of bovine adrenal glands with collagenase (EC 3.4.24.3) in retrograde perfusion as previously described . After suspending in DMEM containing: 10% heat-inactivated fetal calf serum, standard antibiotics, 10 mumol/L cytosine arabinofuranoside and 10 mumol/L fluorodeoxyuridine, the purified chromaffin cells were plated onto collagen-treated 24-well Costar cluster dishes at a density of 106 cells/well, onto 6-well Costar cluster dishes at a density of 5 x 106 cell/well or onto white 96-well plate (Porvair) at a density of 2 x 105 cell/well. Cultures were maintained at 37C in 5% CO2/95% air and used during the 3 --5 days following cell isolation. The required DETA/NO concentration was added to the culture medium for different time periods. When the time of treatment exceeded 24 hours, DETA/NO was added every 24 hours. Intracellular cyclic GMP determination | Before cGMP determination, the chromaffin cells were serum-deprived for 24 hours and washed twice in Locke's solution, pH 7.4 (composition in mmol/L: 140 NaCl, 4.4 KCl, 2.5 CaCl2, 1.2 MgSO4, 1.2 KH2PO4, 4 NaHCO3, 5.6 glucose, 0.01 EDTA and 10 HEPES). After 30-min of pre-incubation at 37C in Locke's solution containing 0.5 mmol/L IBMX, the cells were stimulated for 10 min with 1 mumol/L DEA/NO or its vehicle freshly prepared in 0.5 mL of Locke's solution. The cyclic GMP content of the crude extracts was determined using a commercial [3H]-cyclic GMP radioimmunoassay kit (Amersham) as described elsewhere . Experiments were performed in triplicate. The cGMP concentration results are expressed as pmol/106 cells +- standard error respectively. Cell viability | Cells subjected to the different treatments were washed in Locke's solution and incubated in medium containing the viability/cytotoxicity probes, calcein-AM 1 mumol/L and ethidium homodimer (EthD-1) 8 mumol/L for 30 min at 37C as previously described . Cell images were taken with a Slow Scan Cool CCD Chamera Hammamatsu 4880 --80 joined to a Nikon Eclipse TE 200 microscope, using the B-2A FITC filter for calcein and the G-1B TRITC filter for EthD-1. In order to determine the relative numbers of live and dead cells after different treatment cells were plated in 96-well plates. Fluorescence was read in microplate reader coupled to a Perkin-Elmer LS-50C fluorimeter. Fluorescence in cell sample labeled with calcein AM and EthD-1 was read at 645 nm (F(645)sam) and 530 nm (F(530)sam). Fluorescence at 645 nm was read in a sample were all the cells were dead and labeled either with EthD-1 only F(645)max or with calceine AM only F(645)min. Cells were killed by 30 min treatment with 70% methanol. Fluorescence at 530 nm was read in a sample were all (or nearly all) cells were alive and labeled either with EthD-1 only F(530)min or with calcein AM only F(530)max. The percentage of live cells can be calculated as: % Live cells = (F(530)sam - F(530)min/F(530)max - F(530)min) x 100, an the percentage of dead cells can be calculated from: % Dead cells = (F(645)sam - F(645)min/F(645)max - F(645)min) x 100. sGC subunit protein levels | Chromaffin cells subjected to the different treatments were processed at the required times as follows: cells were washed twice in Locke's solution and disrupted in lysis buffer containing protease inhibitors as previously described . Supernatants containing 30 mug of protein and molecular weight protein standards (precision prestained protein standards from BIO RAD) were subjected to 7.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electrophoretically transferred to PVDF membranes. After blocking non-specific binding sites with 5% nonfat dry milk in PBS containing 0.1% Tween-20 at room temperature for 1 hour, the membranes were incubated with anti-soluble guanylyl cyclase at 1 mug/mL or anti-alpha-actin (1:1000) in blocking buffer overnight at 4C, with constant agitation. Once washed (3 x 10 min), the blots were incubated with anti-rabbit-IgG:HRP (1:5000) and developed with the Super Signal substrate (Pierce). Light was directly detected using Bio-Rad Fluor S equipment. Data from western blot were analyzed using the Quantity One software (Bio-Rad). mRNA levels quantification | The mRNA levels of both sGC subunits (alpha1 and beta1) were determined by the quantitative RT-PCR technique. RNA isolation | After subjecting the chromaffin cells to the different treatments, total RNA was extracted using the RNeasy Kit (Quiagen). This method is based on the selective binding of RNA molecules longer than 200 bases to a silica-gel membrane. Biological samples are first lysed and homogenized in the presence of a highly denaturing guanidine isothiocyanate (GITC)-containing buffer, which immediately inactivates RNases to ensure isolation of intact RNA. The RNeasy procedure isolates all RNA molecules longer than 200 nucleotides. RNA quantification | RNA was quantified using the RiboGreenTM RNA Quantification Kit (Molecular Probes) based on an ultra sensitive fluorescent nucleic acid stain. The excitation maximum for RiboGreen reagent bound to RNA is 500 nm and the emission maximum is 525 nm. RNA was determined from the standard curve generated by increasing amounts of purified RNA (range 20 ng/mL to 1 mug/mL). Fluorescence was measured using a microplate reader coupled to a Perkin-Elmer LS-50C fluorimeter. RT-PCR reactions | The RT-PCR reactions were performed in two-steps. First, the first strand cDNA was synthesized with the MultiScribeTM Reverse Transcriptase in RT buffer containing 5.5 mmol/L MgCl2, 500 mumol/L per dNTP, 2.5 mumol/L Random Hexamers, 0.4 U/muL RNase inhibitor and 3.125 U/muL MultiScribe Reverse Transcriptase. Reactions were performed in a final volume of 25 muL containing 1 mug RNA with an incubation step of 10 min at 25C to maximize primer-RNA template binding. The reverse transcription reaction was performed at 48C for 30 min and reverse transcriptase was inactivated by heating the samples at 95C for 5 min before the PCR reactions. Specific primers for both subunits of bovine sGC were designed using published sequences (12, 14) with the help of the primer express software package (Applied Biosystems). These primers were: sGC alpha1-subunit (forward, base position 526 5'-TTTGCAAACTGATTTTCCCA-3', and reverse, base position 876 5'-TATCCAGGCATAGGATGGATG-3'), beta1-subunit (forward, base position 1221 5'-ACGTTACGAGCCCTGGAAG-3' and reverse, base position 1671 5'-CTACCTCCCTGTGTGGATC-3'). 18S rRNA was used as an endogenous control. To amplify a fragment of 324 bp, a commercial mixture of primers and competimers (in a ratio 3:7) from Ambion was employed. PCR reactions were followed in an ABI PRISM 7700 Sequence Detection System (Applied Biosystems) with the SYBR Green PCR reagents. The reaction mixture contained SYBR Green PCR buffer, 1.5 mmol/L MgCl2, 200 mumol/L dATP, 200 mumol/L dCTP, 200 mumol/L dGTP, 400 mumol/L dUTP, 0.625 U AmpliTaq Gold DNA polymerase and 300 nmol/L of each primer for alpha1 and beta1 subunits, and 120 nmol/L primer and 280 nmol/L competimers for 18S rRNA. Reactions were performed with an initial step of AmpliTaq Gold activation at 95C for 10 min and 40 cycles (melting phase 94C for 40 sec, annealing 60C for 30 sec and extension 68C for 45 sec). Direct monitoring of PCR products was performed by measure the increase in fluorescence caused by binding of SYBR Green to double-stranded DNA. The threshold cycle or CT value occurs when the Sequence Detection Application begins to detect the increase in signal associated with an exponential growth of PCR product. Standard curves for 18S rRNA and both subunits of sGC (alpha1 and beta1) were constructed using several dilutions of cDNA and representing the log of the dilution versus CT in a semilogarithmic plot. The CT values for the different samples obtained in triplicate were interpolated in the corresponding plot. Quantifications were always normalized with the endogenous control 18S rRNA to check for variability in the initial concentration, the quality of total RNA and the conversion efficiency of the reverse transcription reaction. Data from the PCR experiments were analyzed using Applied Biosystems software. Authors' contributions : RF carried out cell culture, cGMP measurement, electrophoresis and blotting, participated in RT-PCR experiments. MT carried out RT-PCR experiments, conceived of the study, its design and coordination, performed the statistical analysis and participated in electrophoresis and western blotting. Backmatter: PMID- 12323077 TI - Specificity of DNA triple helix formation analyzed by a FRET assay AB - Abstract | Background | A third DNA strand can bind into the major groove of a homopurine duplex DNA to form a DNA triple helix. Sequence specific triplex formation can be applied for gene targeting, gene silencing and mutagenesis. Results | We have analyzed triplex formation of two polypurine triplex forming oligodeoxynucleotides (TFOs) using fluorescence resonance energy transfer (FRET). Under our conditions, the TFOs bind to their cognate double strand DNAs with binding constants of 2.6 x 105 and 2.3 x 106 M-1. Our data confirm that the polypurine TFO binds in an antiparallel orientation with respect to the polypurine DNA strand and that triplex formation requires Mg2+ions whereas it is inhibited by K+ions. The rate of formation of triple helices is slow with bimolecular rate constants of 5.6 x 104 and 8.1 x 104 min-1 M-1. Triplex dissociation was not detectable over at least 30 hours. Triplex formation is sequence specific; alteration of a single base pair within the 13 base pairs long TFOs prevents detectable triplex formation. Conclusion | We have applied a FRET assay to investigate the specificity of DNA triple helix formation. This assay is homogeneous, continuous and specific, because the appearance of the FRET signal is directly correlated to triplex formation. We show that polypurine TFOs bind highly specifically to polypurine stretches in double stranded DNA. This is a prerequisite for biotechnical applications of triple helices to mediate sequence specific recognition of DNA. Keywords: Background : It has been discovered in 1957 that a homopyrimidine DNA strand (triplex forming oligonucleotide, TFO) can bind to a homopurine/homopyrimide DNA duplex in the major groove by forming Hoogsteen base pairs with the homopurine strand . The Hoogsteen base pairing scheme mediates sequence specific recognition of the double stranded DNA by the TFO where an AT base pair is recognized by a T and a GC base pair by a C that is protonated at N3 (reviews: ). Later it was found that homopurine strands can also specifically form a DNA triplex in which the AT base pair is contacted by an A and the GC base pair by a G . Triple helix formation with purine rich TFOs is less pH sensitive but requires divalent cations like Mg2+. In either case, the two pyrimidine strands and the two purine strands must be arranged in an antiparallel orientation to form a stable triplex (reviews: ). Triple helix formation has been employed for various purposes in biotechnology like gene targeting, mutagenesis and inhibition of gene activity (reviews: ). In this work, we investigate the sequence specificity of triple helix formation, which is the prerequisite for all biotechnological applications. It has been shown that triplex formation with polypyrimidine TFOs is very sequence specific. The energetic penalty of one single mismatch is similar to that observed in double stranded DNA . Similar results were published for polypurine TFOs . However, many analyses of triplex formation were carried out with assays like UV melting, CD melting, differential scanning calorimetry, gel electrophoresis, footprint analysis, affinity cleavage, chromatography and filter binding. Most of these assays are heterophasic and the signal often is not directly correlated to triplex formation. In contrast, Yang et al. have introduced a fluorescence resonance energy transfer (FRET) assay to analyze triple helix formation that is homophasic and allows to study the kinetics and thermodynamics of this process . In this assay, the double stranded DNA is labeled with fluorescein and the TFO carries a rhodamine label. If a triplex is formed, FRET takes place between the fluorescein donor and the rhodamine acceptor . This FRET signal is directly correlated to the formation of triple helices. We have employed this method here to characterize triplex formation of polypurine TFOs and show that it is highly sequence specific. Figure 1 | Oligonucleotides used in this study. Oligonucleotides used in this study. A) Sequences of the TFOs and DNAs used in this study. B) Principle of the FRET assay used in this study. F denotes for 6-FAM, T for TAMRA; the R0-value of this pair for fluorophors is 49 --54 A. Results and Discussion : Detection of triple helix formation by FRET | We have investigated the biochemical and biophysical properties of a stable purine(purine-pyrimidine) triplex , and checked for its sequence specificity. We used a 21 mer double stranded DNA, DS1, that contains a homopurine stretch of 16 base pairs . The pyrimidine strand carries a 6-carboxyfluorescein (FAM) label on its 5' end. The corresponding 13 mer TFO1 is 5' labeled with 5-carboxytetramethylrhodamine (TAMRA). FAM and TAMRA form a FRET pair with an R0 of 49 --54 A . After triplex formation, the donor and acceptor groups are separated by 4 base pairs which according to computer modeling corresponds to a distance of approximately 15 A. Therefore, a direct interaction of the fluorophors, which is unfavorable for the assay, cannot occur, but the probes are well within a distance that should lead to highly efficient FRET. After excitation of DS1 alone at 470 nm, a strong fluorescence of the FAM-acceptor group is detected at 520 nm . In contrast, the emission of the TAMRA group was almost not detectable (580 nm) if TFO1 alone was excited at 470 nm. After mixing DS1 and the TFO1 and annealing at 55C for 10 min, FRET from FAM to TAMRA occurs as illustrated by a decrease in FAM fluorescence and concomitant increase in the intensity of the emission of the TAMRA group. In the absence of Mg2+ no FRET signal was detectable confirming the finding in the literature that triplex formation with polypurine TFOs requires divalent cations . We also did not observe a FRET signal in the presence of K+ions (data not shown) which are known to inhibit triple helix formation of G-rich TFOs by stabilization of quadruplex structures of the TFO . These results show that the FRET assay is well suited to detect formation of DNA triple helices. Figure 2 | Triple helix formation analyzed by the FRET assay Triple helix formation analyzed by the FRET assay A) Fluorescence emission spectra were recorded for DS1 (0.5 muM), TFO1 (2.5 muM) and DS1 annealed to TFO1 in binding buffer. B) Same experiments with DS2 and TFO2. In both panels, a decrease in FAM fluorescence at 520 nm and an increase in TAMRA fluorescence at 580 nm is observed that is indicative of FRET. Figure 3 | Dependence of triplex formation on divalent cations Dependence of triplex formation on divalent cations To analyze the dependence of triplex formation on divalent cations, DS1 (0.5 muM) was annealed to TFO (10 muM) in binding buffer (20 mM HEPES pH 7.5, 50 mM Na-acetate, 10 mM MgCl2) and in buffer not containing Mg2+-ions (20 mM HEPES pH 7.5, 50 mM Na-acetate). In the Mg2+ free buffer, no triplex formation is detectable. To check the general feasibility of this technique, we have used an additional pair of DNA and TFO. To this end, the sequence of TFO1 was changed at four positions to create TFO2 that should specifically bind to DS2. Triple helix formation of TFO2 and DS2 has not been studied so far. As shown in Fig. , the results obtained with the TFO2/DS2 pair are very similar to the TFO1/DS1 pair. This finding demonstrates that triplex formation with polypurine TFOs is a general phenomenon not restricted to certain sequences. Determination of the biophysical properties of the different triple helices | Titrations were carried out with both pairs of double stranded DNA and TFOs in order to measure the binding constants of the TFOs to the DNA. In these experiments, increasing amounts of TFO were added to a constant amount of double stranded DNA. As shown in Fig. , higher concentrations of the TFO led to an increase in the efficiency of FRET as illustrated by a reduction of the donor fluorescence and increasing acceptor fluorescence. Control experiments were carried out to confirm that even at the highest concentrations the TFO alone does not show significant fluorescence (data not shown). The fluorescence of the TAMRA group was analyzed with respect to the equilibrium binding constants of the TFOs to the corresponding DNA. A quantitative analysis of these data shows that TFO1 binds to DS1 with a binding constant of 2.6 x 105 M-1 whereas TFO2 binds to DS2 with Kass = 2.3 x 106 M-1. The finding that triplex 2 is almost 10-fold more stable than triplex 1 suggests that the A-AT triple is thermodynamically favorable in comparison to the G-GC triple by approximately 5 --6 kJ/mol. However, it is possible that quadruplex formation is more efficient with the more G-rich TFO1 . Since quadruplex formation competes with triplex formation, this effect could contribute to the higher apparent stability of triplex 1. Figure 4 | Equilibrium binding constant of TFO2 to DS2 Equilibrium binding constant of TFO2 to DS2 A) To measure the binding affinity of TFO2 to DS2, DS2 (0.5 muM) was incubated with different amounts of TFO2 (10 nM to 10 muM) in binding buffer for 10 hours and the fluorescence spectra recorded. B) The TAMRA fluorescence was determined and fitted to an equation describing a bimolecular binding equilibrium. The total efficiency of the FRET process can be estimated at the highest concentration of the TFO where the DNA is almost saturated with TFO. Under these conditions, the donor fluorescence is quenched by more than 85% . This result confirms that the TFO binds to the DNA in an antiparallel orientation with respect to the polypurine strand , because in a parallel orientation the distance between the donor and acceptor groups would not allow highly efficient FRET. In addition, the kinetics of association and dissociation of both the triplexes were determined. For the association reactions the DNA and different amounts of TFO were mixed and incubated in the cuvette at 37C. At defined time points, fluorescence spectra were recorded. The fluorescence of the TAMRA group was analyzed with respect to a bimolecular binding reaction. As shown in Fig. , association takes place within 30 min under these conditions. A quantitative analysis results in bimolecular rate constants of association of 5.6 x 104 min-1 M-1 for TFO1 and DS1 and 8.1 x 104 min-1 M-1 for TFO2 and DS2. These rate constants are in the range of results obtained with polypyrimidine TFOs and slightly higher than values previously determined for a polypurine TFO using a heterophasic assay . This result shows that the 10-fold increase in the stability of triplex 2 is in part due to a faster rate of formation of this complex. Figure 5 | Kinetics of association of TFO2 to DS2 Kinetics of association of TFO2 to DS2 A) Example of the association kinetics. DS2 (0.5 muM) and TFO2 (2.5 muM) were mixed in binding buffer and incubated at 37C in the cuvette. The fluorescence emission spectra were recorded at various time points. B) The TAMRA fluorescence of the spectra shown in A was determined and the data fitted to an equation describing the kinetics of a bimolecular binding reaction. For the dissociation reactions, triplex formation was performed by annealing at 55C in a small volume of buffer. Then, the triplex was diluted into a solution containing a large excess of TFO not carrying the TAMRA label. Fluorescence was determined over 30 hours, but no significant change in the FRET efficiency was observed, indicating that dissociation occurs very slowly under these conditions. In contrast, rapid dissociation of the triplex was observed after addition of 20 mM EDTA to the binding buffer confirming that the experimental setup was suited to detect triplex dissociation (data not shown). Sequence specificity of triplex formation | The sequences of TFO1 and TFO2 differ in 5 of 13 positions. To check for sequence specificity of triplex formation, we have investigated if TFO1 can also bind to DS2 and vice versa. As shown in Fig. , non-specific triplex formation does not occur indicating that triplex formation is sequence specific. To analyze the degree of specificity, we have used TFO3 that is identical in sequence to TFO1 with the only exception that one G has been changed to T thereby creating a T-GC base triple. Our results show that TFO3 cannot form a triplex on DS1 indicating that triplex formation is very specific, because introduction of one mismatch base abrogates triplex formation. They confirm similar results obtained by Beal and Dervan using the affinity cleavage method for a different TFO sequence . Figure 6 | Specificity of triplex formation Specificity of triplex formation To determine the specificity of triplex formation, different pairs of non-matching TFOs and DNAs were tested for their ability to form a triplex. A) DS1 (0.5 muM) was annealed with 2.5 muM TFO1 or TFO2. B) DS2 (0.5 muM) was annealed with 2.5 muM TFO1 or TFO2. C) DS1 (0.5 muM) was annealed with 2.5 muM TFO1 or TFO3. Triplex formation between non-matching TFOs and DNAs was not detectable indicating a very high sequence specificity of this process. Conclusions : In this work we have investigated formation and specificity of DNA triplexes using a FRET assay. This assay has many important advantages over alternative assay systems to monitor triplex formation: 1) It is very fast, convenient and requires only small amounts of sample. 2) It is a direct assay, because only triplex formation can lead to the specific FRET signal. 3) It is homophasic and does not depend on the separation of free and bound TFO, which always tends to shift the equilibrium. 4) It allows continuous measurements and thus is suited to follow kinetics of triplex formation and dissociation. We show that triplex formation is highly specific, because exchange of one base pair in the double stranded DNA or in the TFO prevents triplex formation. This high specificity is a prerequisite for the usage of triple helices in biotechnology, e.g. for gene targeting, reduction of gene expression or mutagenesis. Methods : Purified oligodeoxynucleotides were purchased from MWG (Eberberg, Germany). The homogeneity of the oligonucleotides was checked on denaturing polyacrylamide gels. Triplex formation was assayed in binding buffer (20 mM HEPES pH 7.5, 50 mM Na-acetate, 10 mM MgCl2). For most experiments, 500 nM double stranded DNA and 2500 nM TFO were mixed, annealed by heating to 55C for 10 min and equilibrated at 37C. Fluorescence was measured using a Hitachi F4500 spectrofluorometer using 50 mul fluorescence cuvettes. Excitation was at 470 nm, emission was determined between 480 and 690 nm. Excitation and emission slits were at 5 and 2.5 mm, respectively. Spectra were recorded at a scanning speed of 2400 nm/min, usually 9 --16 spectra were averaged to improve the signal to noise ratio. To calculate the efficiency of FRET, we used the fluorescence of the acceptor group (TAMRA) which was obtained by averaging the fluorescence emission spectra between 576 and 584 nm. To measure the binding constant of the TFOs to their corresponding double stranded DNA increasing amounts of TFO (10 nM -- 10 muM) were added to a constant amount of DNA (500 nM). The samples were incubated for 10 hours at 37C and the fluorescence analyzed. To determine the equilibrium binding constants, data were fitted as described . Association kinetics were determined using three different concentrations of TFO (1, 2.5 and 4 muM) and a constant amount of double stranded DNA (500 nM). TFO and DNA were mixed, immediately placed in the spectrofluorometer and temperature maintained at 37C. Fluorescence was scanned at defined time points for 30 min. To determine the rate constant of association all data sets were globally fitted to a bimolecular association reaction as described . Backmatter: PMID- 12207826 TI - Optimizing antibiotics in residents of nursing homes: protocol of a randomized trial AB - Abstract | Background | Antibiotics are frequently prescribed for older adults who reside in long-term care facilities. A substantial proportion of antibiotic use in this setting is inappropriate. Antibiotics are often prescribed for asymptomatic bacteriuria, a condition for which randomized trials of antibiotic therapy indicate no benefit and in fact harm. This proposal describes a randomized trial of diagnostic and therapeutic algorithms to reduce the use of antibiotics in residents of long-term care facilities. Methods | In this on-going study, 22 nursing homes have been randomized to either use of algorithms (11 nursing homes) or to usual practise (11 nursing homes). The algorithms describe signs and symptoms for which it would be appropriate to send urine cultures or to prescribe antibiotics. The algorithms are introduced by inservicing nursing staff and by conducting one-on-one sessions for physicians using case-scenarios. The primary outcome of the study is courses of antibiotics per 1000 resident days. Secondary outcomes include urine cultures sent and antibiotic courses for urinary indications. Focus groups and semi-structured interviews with key informants will be used to assess the process of implementation and to identify key factors for sustainability. Keywords: Background : Antibiotic use in long-term care facilities | Antibiotics are frequently prescribed for older adults who reside in long-term care facilities (LTCFs). The reported prevalence of antibiotic use in nursing home residents ranges from 8% to 17% . Prospective studies of antibiotic use in these facilities demonstrate that 50% to 75% of residents are exposed to at least one course of antibiotics over a one year period . There are several important risks associated with the use of antibiotics in residents of LTCFs . First, there is the risk of developing multi-drug antibiotic resistance with exposure to antibiotics . Second, there is the risk of drug-related adverse effects. In a study of antibiotic use in Ontario facilities which provide chronic care, 6% of individuals developed an adverse effect . Because polypharmacy in this population is common , the risk for harmful drug interactions in addition to adverse reactions to antibiotics is high . Third, the increased use of antibiotics in LTCFs results in significant costs. In a study of antibiotic use in Manitoba nursing homes for example, over $257,000 was spent on antibiotics in the 1988 --89 fiscal year for 1000 nursing home residents . Clearly, optimizing the use of antibiotics in this population is an important quality of care priority. Antibiotics for urinary indications | Urinary tract infections are the most common indication for prescribing antibiotics for residents in LTCFs. Urinary tract infections alone account for 30% to 56% of all prescriptions for antibiotics in that population . The diagnosis of UTIs, like respiratory and other infections in residents of LTCFs, is difficult . Clinical symptoms and signs in this population are often vague and non-specific. In the absence of valid diagnostic criteria, it is difficult to develop a strategy to optimize antibiotic use in the institutionalized elderly. Asymptomatic bacteriuria, or the presence of bacteria in the urine in the absence of urinary symptoms, is however an important exception. This condition occurs in up to 50% of older institutionalized women and 35% of institutionalized older men . It is important to note that the term "asymptomatic" includes bacteriuria in the presence of non-specific, non-urinary symptoms (e.g. malaise, fatigue, functional change) . It is recommended that asymptomatic bacteriuria be treated in populations at high risk of developing subsequent infection, such as children or pregnant women. However, there is compelling evidence to support not treating asymptomatic bacteriuria in residents of long-term care facilities. Data from four randomized controlled trials demonstrate a lack of benefit from treating asymptomatic bacteriuria . These trials, conducted in part to validate the finding of an association between asymptomatic bacteriuria and death, found no effect of antibiotic treatment on mortality . Despite clear evidence that supports not treating asymptomatic bacteriuria, institutionalized older adults are frequently treated for it with antibiotics. It is estimated that about one third of all prescriptions for urinary indications in nursing homes are for asymptomatic bacteriuria . In a 12-month antibiotic utilization in chronic care study, 30% of prescriptions for a urinary indication were for asymptomatic bacteriuria . Defining "appropriate" antibiotic use for most bacterial infections is plagued with difficulty due to diagnostic uncertainties. However, antibiotic prescribing for urinary indications is an important exception. Reducing inappropriate antibiotic use for urinary indications may be an important tactic for optimizing the use of antibiotics in LTCFs. Qualitative study on asymptomatic bacteriuria | To help identify strategies for improving the management of asymptomatic bacteriuria in older adults in residential LTCFs, a qualitative study on reasons why antibiotics are prescribed for this condition was conducted . This study revealed that ordering urine cultures and prescribing antibiotics for asymptomatic bacteriuria are largely driven by nonspecific, non-urinary symptoms (e.g. malaise, confusion, agitation). Nurses, who order urine cultures and influence physicians decision to prescribe antibiotics, were key in this process. Education and guidelines for management of asymptomatic bacteriuria and urinary tract infection were viewed by study respondents as an important priority for both physicians and nurses. Some evidence exists to suggest that systematic practise-based interventions are effective in changing physician performance . Therefore, based on the best clinical evidence and our own qualitative data, we have constructed clinical algorithms for managing UTIs in older adults in LTCFs. This paper describes the protocol of an on-going randomized trial to optimize antibiotic use in residents of nursing homes using clinical algorithms. Methods : The primary aim of this study is to determine if an evidence-based clinical algorithm for managing urinary tract infections (UTIs) in older adults in residential long-term care facilities (LTCFs) can reduce the overall use of antibiotics in LTCFs. Secondary study questions include: Does the use of a diagnostic algorithm reduce the number of urine cultures ordered for residents in LTCFs without urinary symptoms? Does the use of a treatment algorithm reduce the number of antibiotic courses prescribed for presumptive UTIs in the target population? Study population | Twenty-two pairs of nursing homes have been enrolled. Only free standing, community-based residential LTCFs are eligible. Other eligibility criteria include the following: 1) the facility has 100 or more residents; 2) the LTCF does not have a stated policy for diagnosis or treatment of urinary tract infections; 3) the LTCF agrees to refrain from introducing new management strategies for antibiotic utilization or clinical pathways for urinary tract infection during the study. To enhance representation for residential LTCFs in the community, the study will be limited to LTCFs not directly associated with tertiary care centres. The design is randomized matched pairs . Within each of the 11 pairs of LTCFs, one was randomized to the intervention (clinical algorithm), the other half to "usual" management. Quantitative outcomes will include 1) the proportion of antibiotic courses prescribed for urinary indications, 2) the total number of courses of antibiotics used, 3) rates of urine cultures ordered, 4) hospitalization rates for urinary tract infections, and 5) mortality rates. Within a LTCF, randomization of individual healthcare providers or residents to the algorithm likely would introduce bias due to contamination. Therefore, for the quantitative component of this study, the nursing home will serve as the unit of allocation and analysis. Figure 1 | Diagnostic algorithm Diagnostic algorithm. This algorithm guides physicians and nurses in the ordering of urine cultures for nursing home residents with suspected infections. Figure 2 | Treatment algorithm. Treatment algorithm. This algorithm allows physicians and nurses to optimize antibiotic use in residents with suspected infections. Intervention: an evidence-based clinical algorithm | Although treatment guidelines for infections are abundant in the literature (e.g community-acquired pneumonia) few diagnostic or treatment algorithms for infections have been systematically evaluated for outcome . A management algorithm for UTIs in nursing homes has been proposed . However, no algorithms for optimizing antimicrobial use or for managing infections have been evaluated in LTCFs. Nursing staff (RNs and RPNs) play a critical role in the clinical management of LTCF residents. Physicians spend relatively little time at the bedside in LTCFs, and must rely heavily on nursing assessments. Therefore, an intervention to change clinical practise in LTCFs ideally must 1) be evidence-based, 2) be feasible to implement, 3) be inexpensive, 4) involve both nurses and physicians, 5) have the potential for strong "buy-in" from both physicians and nurses, and 6) be evaluable in terms of outcomes. We believe that our clinical algorithm for the diagnosis and treatment of UTIs in residents of LTCFs will meet these criteria. A draft diagnostic and a treatment algorithm was developed using the best evidence available, augmented with feedback from primary care physicians and nurses working in residential LTCFs (Figure + ). Since there is no treatment benefit for asymptomatic bacteriuria , the algorithms indicate that urine should not be cultured in the absence of fever or urinary symptoms, nor should antibiotics be prescribed for positive cultures. In the absence of any urinary symptoms, only 10% of LTCF residents with fever and bacteriuria actually have a urinary infection (positive predictive value of urine culture for urinary infection in the setting of fever is 17%) (35). Therefore, prior to ordering a culture other common infections (respiratory or skin and soft tissue) need to be ruled out. When urinary symptoms are present (in the setting of bacteriuria and fever), about 50% of episodes are, serologically, urinary infections. Clinical evaluation for other infections should therefore also be conducted in such instances prior to instituting antibiotic therapy. A negative urine culture effectively rules out a urinary infection (so long as previous antibiotics were not prescribed). Although bacteriuria in the setting of pyuria is often interpreted as a "true infection", studies have shown that over 90% of the institutionalized elderly with bacteriuria also have pyuria . Therefore the presence of pyuria and bacteriuria is not helpful. However, the absence of pyuria suggests the absence of a host response (i.e. absence of infection), therefore bacteriuria in the absence of pyuria indicates that a urinary tract infection is unlikely . Gross hematuria in the institutionalized elderly generally represents an underlying structural abnormality in the genitourinary tract. About 70% of individuals with gross hematuria also have bacteriuria . Since as many as 25% of individuals who develop hematuria subsequently become febrile , treatment of the resident with fever, gross hematuria and bacteriuria is necessary due to secondary invasive infection . Since there is no relationship between the presence or absence of bacteriuria and non-urinary symptoms , only urinary symptoms will be assessed in the diagnostic or treatment algorithm. Adoption of the algorithms | The algorithms were pilot-tested in four nursing homes prior to the start of the actual trial. For the trial, adoption of the intervention has been through in-services with physicians and nursing staff using case-scenarios to explain the use of the algorithms. The algorithms were printed on pocket cards and distributed to physicians and nursing staff at the start of the study. The algorithms are also kept at all nursing stations using large posters. On-site visits are planned to help with adherence to the protocol. Data collection | Demographics of the residents and features about the facilities will be collected. Other data include the name and dose of the antibiotic, route of administration, start and stop date, reason for the prescription, as well as urinary symptoms leading to the prescription, whether a urine culture was ordered, and if so, its result. Information on deaths, all cause hospitalizations, and hospitalizations for urinary sepsis is being collected. Analysis | The unit of analysis for this study is the nursing home. A paired t-test will be used to analyse the within-pair differences between the proportions of antibiotics prescribed for urinary indications in matched pairs of nursing homes. In this way, the fact that the denominator of the proportions is also an outcome is taken into consideration. Differences in rates of overall antibiotic use (antibiotic courses per 1000 resident days) will be compared using a paired t-test. Rates of antibiotic use for urinary indications (antibiotic courses per 1000 resident days and defined daily dosages/1000 resident days), rates of urine cultures obtained (urine cultures per 1000 resident days), rates of hospitalization (per 1000 resident days), and overall mortality rates will be compared using paired t-tests and Wilcoxon signed rank tests. Logistic regression analysis is planned to account for potentially important co-variates such as proportion of residents bed/wheelchair bound and pharmacy automatic stop dates . Sample size calculation | In our 12 month study of antibiotic utilization in Ontario long-term care facilities, 30% of all antibiotic prescriptions were for urinary indications, of which one third were for asymptomatic bacteriuria. We believe that the algorithm will lead to at least a 20% reduction in the overall use of antibiotics, that is, a reduction in the proportion of antibiotic prescriptions for urinary indications from 30% to 10%. To detect this difference, for an alpha of 0.05 and 80% power, a total of 142 prescriptions (71 in each arm) is needed. To adjust for the effect of within cluster dependency, the intracluster correlation coefficient (between home variance for urinary antibiotic prescription / sum of inter- and intra-home variance), was then calculated using data from the 12 month study in Ontario long-term care facilities. The proportion of antibiotics prescribed for a urinary indication was 0.32 (p) and the variance 0.009 (between home variance). The intra-home variance, given by the binomial distribution [(p) (1-p)], was 0.21. Therefore, the intracluster correlation coefficient is 0.04. Donner et al. (41) describe a variance inflation factor given by 1 + (n - 1) tau, where tau is the intracluster correlation coefficient, and n= samples (prescriptions) needed per cluster. Since 23 prescriptions can be obtained per home per month (based on the average LTCF in our Canada-US study), and if data collection is conducted over 11 months, then n = 253. Using the formula given above, the variance inflation factor is 11. Therefore, 1562 (142 x 11) urinary prescriptions are required. Since these represent 30% of antibiotic prescriptions, 5206 prescriptions need to be collected in total. This means that 20 or 10 pairs of nursing homes will need to be followed for 12 months. Since matching, which would improve efficiency, was not accounted for in the sample size calculation, these figures are a conservative estimate. We will recruit another two homes to maintain the target sample size in case a pair of homes withdraws from the study. Qualitative component | To evaluate the process of adopting the proposed algorithms in LTCFs, we plan to conduct focus groups and semi-structured interviews. Two groups of respondents will be interviewed, key clinical administrators in the participating LTCFs (medical directors, directors of nursing, infection control officers), and staff who will implement the algorithms (RN and RPN). Standard methods to ensure that the qualitative data are gathered and analyzed rigorously will be followed throughout the study. These include member checking (asking respondents to review our findings), peer review (asking colleagues to review our research process), and an audit trail (creating documents which outline all decisions made throughout the investigation) . Each of these steps will be taken in this study to ensure that this portion of the study is rigorous and findings are trustworthy. Competing interests : None declared. Authors' contributions : All authors contributed to the development of the protocol of this randomized trial. Mark Loeb wrote the original draft of this paper and all authors offered critical revisions. Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12167173 TI - Combinatorial diversity of fission yeast SCF ubiquitin ligases by homo- and heterooligomeric assemblies of the F-box proteins Pop1p and Pop2p AB - Abstract | Background | SCF ubiquitin ligases share the core subunits cullin 1, SKP1, and HRT1/RBX1/ROC1, which associate with different F-box proteins. F-box proteins bind substrates following their phosphorylation upon stimulation of various signaling pathways. Ubiquitin-mediated destruction of the fission yeast cyclin-dependent kinase inhibitor Rum1p depends on two heterooligomerizing F-box proteins, Pop1p and Pop2p. Both proteins interact with the cullin Pcu1p when overexpressed, but it is unknown whether this reflects their co-assembly into bona fide SCF complexes. Results | We have identified Psh1p and Pip1p, the fission yeast homologues of human SKP1 and HRT1/RBX1/ROC1, and show that both associate with Pop1p, Pop2p, and Pcu1p into a ~500 kDa SCFPop1p-Pop2p complex, which supports polyubiquitylation of Rum1p. Only the F-box of Pop1p is required for SCFPop1p-Pop2p function, while Pop2p seems to be attracted into the complex through binding to Pop1p. Since all SCFPop1p-Pop2p subunits, except for Pop1p, which is exclusively nuclear, localize to both the nucleus and the cytoplasm, the F-box of Pop2p may be critical for the assembly of cytoplasmic SCFPop2p complexes. In support of this notion, we demonstrate individual SCFPop1p and SCFPop2p complexes bearing ubiquitin ligase activity. Conclusion | Our data suggest that distinct homo- and heterooligomeric assemblies of Pop1p and Pop2p generate combinatorial diversity of SCFPop function in fission yeast. Whereas a heterooligomeric SCFPop1p-Pop2p complex mediates polyubiquitylation of Rum1p, homooligomeric SCFPop1p and SCFPop2p complexes may target unknown nuclear and cytoplasmic substrates. Keywords: Background : The ubiquitin/proteasome-dependent proteolysis system has been implicated in a wide variety of cellular regulatory mechanisms, including transcription, signal transduction, and cell cycle control (reviewed in ). The system employs a cascade of enzymatic reactions that lead to the covalent attachment of a chain of multiple ubiquitins to substrate proteins . In many cases, modification by ubiquityl moieties targets proteins to the proteasome, ultimately resulting in their degradation. The ubiquitylation reaction involves a minimum of three enzymes: An E1, which mediates the ATP-dependent activation of ubiquitin, and an E2, or ubiquitin conjugating enzyme (UBC), which, together with an E3 ubiquitin ligase, transfers ubiquitin to the target protein. E3 enzymes are of particular interest, as they mediate the substrate specificity of ubiquitylation reactions. Studies in budding yeast identified SCFCdc4p, an E3 ubiquitin ligase complex that mediates the ubiquitylation of the CDK inhibitor Sic1p . SCFCdc4p consists of at least four proteins: the cullin Cdc53p, the RING domain protein Hrt1p/Rbx1p/Roc1p, the adapter protein Skp1p, and Cdc4p (reviewed in ). Cdc4p contains two sequence motifs, which are conserved in a wide variety of so-called F-box proteins: C-terminal WD-repeats that are involved in binding the substrate Sic1p in a phosphorylation-dependent manner, and a central F-box that interacts with Skp1p . Cdc53p in turn binds to Skp1p and Hrt1p/Rbx1p/Roc1p, which bridges Cdc53p with the ubiquitin-conjugating enzyme Cdc34p/Ubc3p [,-]. In vitro reconstitution demonstrated that SCFCdc4p, Cdc34p, E1, ubiquitin, and ATP are sufficient to mediate Sic1p polyubiquitylation . Components of the SCF system are widely conserved in eukaryotes . In human cells, for example, SCFSkp2 mediates destruction of the CDK inhibitor p27 , while SCFbeta-TRCP targets IkappaB . All of these SCF complexes share homologues of the core components CDC53/CUL1, SKP1, and HRT1/RBX1/ROC1, which associate with different F-box proteins. Several lines of evidence suggest that the SCF pathway is also conserved in the fission yeast Schizosaccharomyces pombe. Pcu1p, a Cdc53p/CUL1-related protein, was shown to associate with two different F-box/WD repeat proteins, Pop1p and Pop2p, when overexpressed . In addition, genetic studies demonstrated that both of these F-box proteins are simultaneously required for efficient destruction of Rum1p and the replication initiator Cdc18p . Rum1p is a Sic1p-analogous CDK inhibitor, which accumulates in G1, but is degraded as cells enter S phase . Failure to degrade Rum1p is the major phenotypic defect of pop1 and pop2 deletion strains, which leads to disturbance of normal cell cycle progression, resulting in polyploidy [-,]. Based on the genetic data and the biochemical observation that Pop1p and Pop2p interact when overexpressed, a putative SCFPop1p-Pop2p complex was proposed, which would contain the heterooligomerizing F-box proteins Pop1p and Pop2p bound to SCF core components . Whether this unusual heterooligomeric SCFPop1p-Pop2p complex exists in vivo and whether it mediates Rum1p ubiquitylation, remained unproven, as not all fission yeast SCF core subunits were identified. In addition, based on overexpression, distinct SCFPop1p and SCFPop2p complexes were proposed to target unknown substrates, but no biochemical evidence of their activity was provided . To address these questions, we cloned two additional subunits of SCFPop and performed a detailed characterization of its activity in vitro and in vivo. Our results indicate that heterooligomeric SCFPop1p-Pop2p mediates Rum1p ubiquitylation whereas distinct SCFPop1p and SCFPop2p complexes target unknown nuclear and cytoplasmic substrates, thereby generating combinatorial diversity of SCF function in fission yeast. Results : Composition of SCFPop1p-Pop2p | Based on the composition of known SCF complexes, we identified in the S. pombe genome database psh1 (pombe skp1 homologue) and pip1 (pop interacting protein 1), two genes encoding proteins with strong similarity to human SKP1 and HRT1/RBX1/ROC1, respectively (data not shown). Consistent with these proteins being components of the putative SCFPop1p-Pop2p complex, they all co-purified with Pop1p, Pop2p, and Pcu1p when overexpressed pairwise (data not shown). Co-immunoprecipitation experiments using affinity purified rabbit antisera confirmed these interactions at the level of the endogenous proteins. While each of the five antisera co-precipitated at least one of the other subunits, Pip1p, Pop1p, and Pop2p antisera co-precipitated all five proteins from wild-type cell lysate . Size fractionation of total cell lysates prior to immunoprecipitation revealed co-elution of Pip1p with Pop1p, Pop2p, Pcu1p, and Psh1p in a high molecular weight complex of approximately 500 kDa, which we refer to as SCFPop1p-Pop2p . The composition of the core complex Pip1p/Pcu1p/Psh1p did not undergo major variations during the cell cycle . We have carefully reexamined potential cell cycle variations of Pcu1p neddylation apparent in the IP/immunoblotting experiment in Fig. . These variations were not seen when samples were denatured in SDS immediately following extract preparation (data not shown), suggesting that they arise from varying degrees of deneddylation presumably occuring during the immunoprecipitation step. In addition, in a separate experiment, Pop1p-Pop2p heterooligomerization was largely constant during the cell cycle . These findings indicated that cell cycle-dependent substrate degradation is unlikely to be controlled at the level of SCFPop1p-Pop2p complex formation. Figure 1 | SCFPop protein interactions SCFPop protein interactions (1A) SCFPop complex formation at endogenous expression levels. Lysate prepared from wild-type fission yeast cells was used for immunoprecipitation with affinity-purified rabbit antisera directed against SCF subunits as indicated. Precipitates were separated by SDS gel electrophoresis and co-precipitating proteins were detected by immunoblotting with the antisera indicated in each panel. Note that Pip1p, Pop1p, and Pop2p antisera precipitated all five SCFPop subunits (lane 1). The control lane (6) contained an immunoprecipitate prepared with antisera depleted of Pip1p antibodies (flowthrough of the affinity column -- FT). Various unspecific bands, some resulting from cross-reactivity of the antibody chains, are indicated by stars. Pcu1p is not resolved into the native and Ned8p-modified forms in this gel system (data not shown). (1B) Co-purification of SCFPop1p-Pop2p subunits in a ~500 kDa complex. Wild-type cell lysate was fractionated by gel filtration on a Sephacryl S300 column, and resulting fractions were immunoprecipitated with Pip1p antisera. Precipitates were analyzed by immunoblotting with the indicated SCF antisera. Size standards are indicated. Immunoglobulin heavy chains are indicated by unfilled asterisks. Other unspecific bands or degradation or modified products are indicated by filled asterisks. (1C) SCFPop protein interactions during the cell cycle. Temperature-sensitive cdc25-22 cells were arrested in G2 by incubation at 37C for 4 h. Cells were released into mitosis and subsequent cell cycle phases by rapidly shifting to 25C. Samples were taken after the indicated times and prepared for immunoprecipitation with Pip1p (top panel) or Pop2p (bottom panel) antibodies. Precipitates were probed for co-purification of SCF subunits with the indicated antisera. Pcu1p is resolved into two bands, presumably representing native and Ned8p-modified species. The percentage of septate cells, which represent cells in S phase, was determined at each time point and blotted against time as illustrated in the graph. SCFPop1p-Pop2p mediates polyubiquitylation of Rum1p in vitro | Genetic experiments suggested that degradation of Rum1p depends on the Pop1p and Pop2p F-box proteins , but also requires phosphorylation of Rum1p on serine 58 and threonine 62 by cyclin-dependent kinase (CDK) . Thus, phosphorylated Rum1p may be a substrate for SCFPop1p-Pop2p-mediated polyubiquitylation. To test this, we first confirmed that Rum1p is an in vitro substrate for the Cdc2p kinase in association with the cyclin Cig1p as judged by a mobility shift on SDS gels . Bacterially expressed Rum1p purified to apparent homogeneity was also efficiently phosphorylated by Cdc2p/Cig1p complexes . Figure 2 | SCFPop1p-Pop2p mediated Rum1p ubiquitylation in vitro SCFPop1p-Pop2p mediated Rum1p ubiquitylation in vitro (2A) Various Cdc2p complexes were affinity purified from cells co-expressing 6 x His-Myc-tagged Cdc2p and HA-tagged cyclins on nickel NTA resin. Kinase complexes were incubated in the presence of ATP with Rum1p produced by combined in vitro transcription/translation in rabbit reticulocyte lysate (lanes denoted with "+"). Reactions only containing buffer were used as negative controls (lanes denoted "-"). Reaction products were separated on SDS gels and detected by fluorography (left panel). Both, Cdc2p/Cdc13p and Cdc2p/Cig1p produced a shift in Rum1p migration indicative of phosphorylation, although it is unclear whether both kinases target the same residues . In a confirmatory experiment, bacterially expressed Rum1p was purified to apparent homogeneity by column chromatography as described (middle panel), and phosphorylated by incubation with Cdc2p/Cig1p complexes in the presence of radiolabeled ATP (right panel). Reaction products were separated by SDS gel electrophoresis and detected by autoradiography. (2B) Phosphorylated Rum1p interacts with Pop1p-Pop2p complexes. 6 x His-Myc-tagged Pop2p and HA-Pop1p were co-overexpressed in wild-type fission yeast, and Pop1p-Pop2p complexes were affinity-purified on Ni-NTA resin. Bound complexes were incubated with bacterially expressed phosphorylated Rum1p. After the incubation, beads were washed extensively, followed by boiling in SDS sample buffer and electrophoresis. Rum1p bound to Pop1p-Pop2p was visualized by autoradiography (lane 1, top panel). As controls, 6 x His-tagged Pop1p or Pop2p individually overexpressed in pop1 pop2 double mutants were employed in the Rum1p binding assay (lanes 2, and 3; top panel). As a further negative control, lysate of the untransformed pop1 pop2 double mutant was absorbed to Ni-NTA beads (lane 4, top panel). The bottom panel shows the affinity-purified Pop1p and Pop2p complexes used in the binding assay. (2C) 6 x His-tagged human E1 and UBC3 were expressed in bacteria and purified by affinity binding to nickel NTA resin, followed by subsequent chromatographic steps (MonoQ and gel filtration). The preparations used in the ubiquitylation assays are shown. (2D) Rum1p ubiquitylation assay. Bacterially expressed, phosphorylated Rum1p, derived from the preparation shown in Fig. was incubated with Pip1p complexes immunopurified from the indicated strains, human E1, UBC3 (hUBC3), ATP, and ubiquitin for 90 min at 30C. Reactions were stopped by the addition of SDS sample buffer, and reaction products were separated by SDS PAGE and visualized by autoradiography. Lanes 2, 4, and 6 included mutant UBC3 in which the catalytic cysteine residue was replaced (UBC3-DeltaCys). Mutant ubiquitin, in which all lysine residues were replaced by arginine was present in lane 7. The assays presented in the right panel (lanes 8 -- 11) were performed with Pip1p complexes purified from csn5 mutants and contained the indicated human or S. pombe UBCs. To determine whether phosphorylated Rum1p interacts with Pop1p and Pop2p, protein lysate was prepared from cells co-overexpressing epitope-tagged combinations of Pop1p and Pop2p. Upon affinity purification on Ni-NTA resin, HA-Pop1p/His-Myc-Pop2p complexes were incubated with bacterially expressed, phosphorylated Rum1p. In this reaction, Pop1p-Pop2p complexes specifically bound phosphorylated Rum1p (Fig. , lane 1). Consistent with the genetic data , His-Myc-Pop1p and His-Myc-Pop2p individually purified upon overexpression in pop1 pop2 double mutants exhibited no Rum1p binding above background (Fig. , lanes 2 and 3). Given our ability to prepare immunopurified SCFPop1p-Pop2p that bound phosphorylated Rum1p, we sought to reconstitute Rum1p polyubiquitylation in vitro. In addition, we required an ubiquitin activating enzyme (E1) and an ubiquitin conjugating enzyme (UBC). While human E1 is highly similar to its fission yeast counterpart (data not shown), inspection of the S. pombe genome revealed fourteen potential UBCs, none of which is an obvious homologue of human UBC3 or budding yeast Cdc34p, since all lack the characteristic C-terminal extension (data not shown). We therefore purified recombinant human E1 and UBC3 (CDC34) upon expression in bacteria . In the presence of human E1, UBC3, ubiquitin, and ATP, SCFPop1p-Pop2p complexes immunopurified with Pip1p antibodies converted a small portion of phosphorylated Rum1p into high molecular weight species . This conversion was dependent on the addition of E1, ATP (data not shown), and wild-type UBC3 (Fig. , lanes 2,4,6). The activity of SCFPop1p/Pop2p was augmented when Pip1p complexes where purified from csn5 mutants (Fig. , lane 5). This mutant accumulates Pcu1p exclusively in a form carrying the stimulatory Nedd8p modification, due to a defect in COP9/signalosome-mediated cullin deneddylation . Replacing wild-type ubiquitin by a mutant lacking all lysine residues prevented the formation of high molecular weight products (Fig. , lane 7), indicating that they represent polyubiquitylated Rum1p species generated in the reaction. Similar polyubiquitylated reaction products were detected upon incubation of phosphorylated Rum1p with Pcu1p immunocomplexes, further suggesting that the activity is mediated by SCFPop1p-Pop2p (data not shown). Moreover, Rum1p ubiquitylation was not obtained with Pip1p complexes prepared from cell lysate of pop2 deletion strains, proving the F-box protein dependency of this reaction (Fig. , lane 3). In addition, the reaction was specific for human UBC3, as no ubiquitylation was obtained with fission yeast Ubc1p, Ubc7p, Ubc11p, or Ubc13p (Fig. , lanes 10,11). Taken together these results strongly suggest that SCFPop1p-Pop2p mediates the polyubiquitylation of CDK phosphorylated Rum1p in vitro. Differential subcellular localization of SCFPop1p-Pop2p subunits | The co-purification of the five identified SCFPop1p-Pop2p subunits and their in vitro activity toward Rum1p suggested that they coexist in a common subcellular compartment. The nuclear localization of the only known substrates, Cdc18p and Rum1p (D.A.W., unpublished), indicated that a substantial portion of SCFPop1p-Pop2p is enriched in the nucleus. To test this assumption, all five SCFPop1p-Pop2p subunits were expressed as fusion proteins with green fluorescent protein (GFP) at low levels from an inducible pRep81 plasmid. While Pip1p, Psh1p, Pcu1p, and Pop2p were present in both the cytoplasm and the nucleus, surprisingly, GFP-Pop1p was largely restricted to the nucleus . These localization patterns were consistently observed in each single cell of an asynchronous population, excluding major variations during the cell cycle. Figure 3 | Subcellular localization of SCFPop subunits Subcellular localization of SCFPop subunits (3A) Pop1p, Pop2p, Pcu1p, Psh1p, and Pip1p were mildly overexpressed in wild-type cells as N-terminally GFP-tagged proteins from pRep81 for 20 h. Cells were fixed in para-formaldehyde and mounted onto poly-lysine coated cover slips. Nuclei were counterstained with DAPI. Fluorescence images were obtained with a Spot CCD camera mounted onto a Nikon E600 epifluorescence microscope. (3B) Immunoblot of cells genetically modified to contain Pop1p or Pop2p tagged with 13 C-terminal Myc epitopes at the endogenous genomic loci. (3C) Differential subcellular localization of endogenous Pop1p and Pop2p. The cells described in (B) were fixed in para-formaldehyde and processed for immunostaining with Myc antibodies as described . Wild-type cells not containing any Myc-tagged alleles were used as specificity controls. DAPI stained cell nuclei are indicated. (3D) The epitope-modified cells described above were fractionated into nuclear and cytoplasmic proteins as described in materials and methods. Equal proportions of both fractions were separated by SDS PAGE and probed with Myc antibodies to detect Pop1p and Pop2p (top panel). Blots were reprobed with antibodies directed against cytoplasmic tubulin (middle panel) and nuclear PCNA (bottom panel) to estimate the efficiency of the fractionation protocol. (3E) Cells harboring 13Myc epitope-tagged Pop1p in a pop2 deficient background (pop1-13myc Deltapop2) or 13Myc-tagged Pop2p in a pop1 mutant (pop2-13myc Deltapop1) were fixed and processed for immunostaining with Myc antibodies. As reference, the strains described above, which contain Myc-tagged Pop1p or Pop2p in a wild-type background are shown. To rule out the possibility that overexpression or N-terminal fusion to GFP affects their localization, Pop1p and Pop2p were modified with 13 C-terminal Myc epitope-tags at the endogenous genomic locus. Immunoblotting proved the expression of correctly sized proteins and, in addition, showed that endogenous Pop1p is approximately twofold more abundant in S. pombe cells than Pop2p . Indirect immunofluorescence staining with Myc antibodies confirmed that Pop1p is predominantly localized to cell nuclei, whereas Pop2p is expressed in both the cytoplasm and the nucleus . To confirm these localization patterns, cells derived from the epitope-tagged strains were biochemically fractionated into cytoplasmic and nuclear components. The efficiency of enrichment of nuclear and cytoplasmic components was estimated by analyzing fractions with antibodies recognizing the nuclear marker PCNA and cytoplasmic tubulin . Although both fractions showed some contamination, Pop1p was detected mostly in nuclear fractions, while Pop2p was apparent in both nuclear and cytoplasmic fractions . Thus, all five SCFPop1p-Pop2p subunits appear to coexist in the nucleus, although all but Pop1p are also present in the cytoplasm. Since Pop1p and Pop2p interact with each other , we asked whether their localization patterns depended on the presence of the respective interaction partner. For this, we created a pop2 deletion strain carrying Pop1p modified with 13 Myc epitope tags at the endogenous genomic locus (pop1-13myc Deltapop2 strain). In addition, we created the reciprocal pop2-13myc Deltapop1 strain containing epitope-tagged endogenous Pop2p in a pop1 deletion background. The distinct localization patterns of Pop1p-13Myc and Pop2p-13Myc were fully maintained in these strains . This observation was confirmed by overexpressing GFP-tagged Pop1p or Pop2p in pop1 pop2 double deletion mutants (data not shown). These data indicate that Pop1p and Pop2p assume their subcellular localization pattern independent of each other, indicating the possibility of distinct nuclear and cytoplasmic homooligomeric SCFPop1p and SCFPop2p complexes. Differential F-box requirements of Pop1p and Pop2p | As shown above and in previous genetic work , SCFPop1p-Pop2p-dependent Rum1p degradation requires two different proteins with highly conserved F-boxes . To better understand the apparent dual F-box requirement for SCFPop1p-Pop2p function, we generated mutants of Pop1p and Pop2p lacking F-boxes (Pop1p-DeltaF, Pop2p-DeltaF; Fig. ). In addition, we prepared a set of mutants, in which the F-boxes of Pop1p and Pop2p were swapped (Pop1p-2F, Pop2p-1F; Fig. ). The mutants were tested for their ability to suppress polyploidy and Rum1p accumulation in the respective pop mutant strains. Figure 4 | Pop1p and Pop2p F-box requirements Pop1p and Pop2p F-box requirements (4A) ClustalW alignment of the F-boxes of Pop1p, Pop2p, budding yeast Cdc4p, and human SKP2 and beta-TRCP. Residues involved in the SKP1/F-box interaction are indicated with arrows . (4B) Schematic of Pop1p and Pop2p F-box mutants. (4C) The mutant proteins displayed in (B) were expressed in pop1 or pop2 deletion strains from pRep81 as indicated. Cells were fixed and processed for flow cytometry. Cells containing the normal 2C DNA content and cells with greater than 4C were quantified and results are presented in the diagram. Note that the polyploidy phenotype of Deltapop2 mutants is mild and, hence, the rescue observed with Pop2p proteins is less pronounced. (4D) The strains described in (C) were induced to express the indicated Myc-tagged Pop1p and Pop2p mutants for 20 h. Protein lysates were probed with antisera against Rum1p and tubulin as a loading control. As described previously , wild-type Pop1p mildly overexpressed from a pRep81 plasmid fully complemented the polyploidization phenotype of pop1 mutants as determined by flow cytometric measurement of the cellular DNA content . In addition, Rum1p accumulation in pop1 mutants was efficiently reversed by wild-type Pop1p . In contrast, Pop1p lacking its F-box (Pop1p-DeltaF) or Pop1p, in which the F-box was replaced by the F-box of Pop2p (Pop1p-2F) were largely inactive in both assays . Thus, as with many other F-box proteins, the F-box of Pop1p is essential for its in vivo functions. In contrast, wild-type Pop2p, the corresponding F-box mutant, and Pop2p containing the Pop1p F-box were equally effective in preventing Rum1p accumulation . The same wild-type and mutant proteins also reversed the mild polyploidy phenotype of pop2 disruptants . Thus, in contrast to Pop1p, Pop2p does not seem, to require its F-box to mediate Rum1p degradation in vivo. To further substantiate this conclusion, we examined Rum1p protein stability in wild-type and pop mutant strains expressing F-box-deleted Pop proteins from the weak pRep81 promoter. Since the sensitivity of our Rum1p sera was insufficient to detect the low levels present in wild-type cells (see Fig. , lane 9), these experiments were conducted in a background where endogenous Rum1p was modified with 13 C-terminal c-Myc epitope tags. Rum1p half-life was increased from ~20 minutes in wild-type to greater than 100 minutes in pop1 or pop2 mutants . While F-box deleted Pop2p expressed from plasmids reduced Rum1p half-life to ~20 minutes in pop2 mutants, F-box-deleted Pop1p was completely defective in rescuing the Rum1p proteolysis defect of pop1 mutants . Instead, expression of Pop1p-DeltaF in pop1 mutants led to even greater stabilization of Rum1p, potentially due to dominant negative interference with the residual activity of Pop2p and/or other F-box proteins. Figure 5 | Rum1p protein stability in pop mutants Rum1p protein stability in pop mutants Wild-type and the indicated pop mutant strains carrying Rum1p modified with 13 c-Myc epitope tags on the C-terminus were incubated with 100 ug/ml cycloheximide (CHX) for the indicated times, followed by preparation of protein lysates. Rum1p-13Myc was detected by immunoblotting with c-Myc antibodies (right panels). For complementation experiments in the bottom two panels, the indicated pop1 and pop2 deletion strains carrying rum1-13myc were transformed with plasmids driving the expression of F-box deleted versions of Pop proteins (pRep81.myc-pop1-DF and pRep81.myc-pop2-DF plasmids, respectively). Expression from plasmids was induced by removal of thiamin for 20 h and 100 ug/ml CHX was added for the indicated periods. Protein lysate was prepared and Rum1p-13Myc was detected by immunoblotting with c-Myc antibodies. Since Pop1p-DF and Pop2p-DF expressed from plasmids are also Myc-tagged, they are detected on these immunoblots as bands migrating above Rum1p-13Myc as indicated. Immunoblots were quantitated using the free imaging software package tnimage for Linux and results are blotted in a diagram to estimate Rum1p half-lifes. F-box independent interaction of Pop1p and Pop2p | The failure of the Pop2p F-box to replace the F-box of Pop1p as well as the finding that it is not essential for Rum1p degradation could be explained, if it was not critically involved in SCFPop1p-Pop2p protein interactions. We therefore tested the possibility that Pop2p can be tethered to the SCF core complex independently of its F-box via an interaction with Pop1p. Consistent with this hypothesis, co-immunoprecipitation experiments of overexpressed proteins revealed that the Pop1p-Pop2p interaction occurs independently of the F-boxes of both Pop1p and Pop2p . Figure 6 | Pop protein interactions Pop protein interactions (6A) Tagged versions of Pop1p and Pop2p F-box deletion and swap mutants were overexpressed in wild-type cells as indicated. 6xHis-Myc-tagged proteins were absorbed to Ni-NTA-agarose and co-purifying HA-tagged proteins were detected by immunoblotting with the indicated antibodies. Control strains only expressed HA-tagged proteins (lanes 1 and 5). Expression of HA-tagged proteins in the total lysate is shown in the lower panel. (6B) The indicated tagged Pop1p and Pop2p fragments were co-expressed in wild-type cells. Cell lysate was prepared and Pop2p immunocomplexes were precipitated with anti-Pop2p antisera. Co-purifying proteins were detected by immunoblotting with the indicated antibodies. Control precipitates were prepared with preimmune sera ("pre-i"). Total lysates are shown in the right panel. The N-terminal fragment of Pop2p from amino acid 2 --241 lacks the F-box. Both Pop1p fragments (2 --402 and 227 --775) contain F-boxes. We had previously mapped the domain of Pop2p that interacts with Pop1p to an N-terminal fragment consisting of the first 241 amino acids and lacking the F-box (, Fig. , lane 8). In co-immunoprecipitation experiments with overexpressed proteins, this fragment also bound to an N-terminal piece containing the first 402 residues of Pop1p (Fig. , lane 12). Thus, the Pop1p-Pop2p interaction is mediated by their N-terminal domains. A further truncation mutant mapped the Pop2p binding domain to a region between residues 228 and 402 of Pop1p (Fig. , lanes 9,10). Individual SCFPop1p and SCFPop2p complexes bearing ubiquitin ligase activity | The apparent dispensibility of the F-box of Pop2p for Rum1p degradation and binding to Pop1p raised the question of why Pop2p does contain an F-box. Based on the subcellular localization data, we considered the possibility that the F-box of Pop2p may mediate the assembly of a cytoplasmically localized SCFPop2p complex, independent of Pop1p. To demonstrate this, we again used the strain in which endogenous Pop2p was modified by 13 Myc epitope tags in a pop1 deletion background. The same experiments were carried out with the reverse stain, which contained Pop1p-13Myc in a pop2 background. As a reference for SCFPop complex formation, we used strains carrying 13Myc epitope-tagged Pop1p or Pop2p integrated into the genome of wild-type cells. Pip1p immunoprecipitates were prepared from lysates of these four strains and appropriate controls, and co-purification of SCF components was determined by immunoblotting. These experiments showed that both F-box proteins, in the absence of their respective heterooligomerization partner, could individually bind to Pip1p in complexes that also contained Psh1p and Pcu1p . These findings indicate the existence of distinct SCFPop1p and SCFPop2p complexes in vivo. Figure 7 | Individual homooligomeric SCFPop1p and SCFPop2 complexes Individual homooligomeric SCFPop1p and SCFPop2 complexes (7A) Lysate from a strain harboring epitope-tagged endogenous Pop1p in a Deltapop2 background or from the reverse strain containing Pop2p-13Myc in a pop1 deletion strain was immunoprecipitated with affinity-purified anti-Pip1p antibodies (lanes 1 and 2). Immunocomplexes were probed for the presence of associated SCF subunits with the indicated antibodies. Strains containing tagged Pop1p and Pop2p in a wild-type background (i.e. in the presence of their respective binding partners) were used as controls (lanes 3 and 4). Wild-type strains did not show any signal with Myc antibodies (lane 5). The control lane (6) contained precipitates prepared in the absence of Pip1p antibodies (bead control). (7B) Lysate from the strains described above were fractionated by gel filtration and fractions were assayed for the elution profile of Pop proteins by immunoblotting with Myc antibodies. The elution profiles for Pcu1p, Psh1p, and Pip1p are shown for reference (bottom three panels). Size standards are indicated. (7C) Ubiquitylation activity of SCFPop1p and SCFPop2p complexes. SCF complexes were immunopurified with Myc antibodies from lysate of the indicated strains. Ubiquitylation assays in the presence of human E1, UBC3, ATP and ubiquitin were performed as described in materials and methods. Note that a substantial amount of the high molecular weight reaction products migrated in the stacking gel. Controls include immunoprecipitates from wild-type lysates (lane 1) and reactions performed in the presence of mutant ubiquitin lacking all lysine residues (lane 6). To further substantiate this conclusion, we used gel filtration to compare the elution profiles of Pop1p and Pop2p in the presence or absence of their respective dimerization partners. If recruitment of Pop2p into a high molecular weight SCF complex required heterooligomerization with Pop1p, its elution profile would be expected to shift to a smaller size in the absence of Pop1p. Consistent with the results presented in Fig. , Pop2p, together with SCF core subunits, eluted in fractions corresponding to 400 -- 600 kDa, irrespective of whether Pop1p was present or not . In the reverse experiment, the elution profile of Pop1p was found to be independent of the presence of Pop2p . While these data support the existence of distinct SCFPop1p and SCFPop2p complexes, individual binding of Pop1p and Pop2p to SCF core components in the absence of their heterodimerizing F-box protein partners does not rule out the possibility that these complexes represent inactive intermediates formed during the normal assembly of functional SCFPop1p-Pop2p complexes. To exclude this possibility, we asked whether distinct SCFPop1p and SCFPop2p complexes bear ubiquitin ligase activity in vitro. To this end, we performed in vitro ubiquitylation assays. Since the substrates of putative SCFPop1p and SCFPop2p ubiquitin ligases are unknown, we adopted a substrate-independent assay originally described by Lyapina et al. . For this experiment, we again used the strains harboring genomically integrated Myc epitope-tagged Pop1p or Pop2p in a background deficient in the respective heterooligomerization partner (Deltapop2 pop1-13myc, Deltapop1 pop2-13myc strains). Pop1p and Pop2p complexes were immunopurified with Myc antibodies and employed in in vitro ubiquitylation assays upon addition of E1, UBC3, ubiquitin, and ATP. High molecular weight products generated in the reaction were detected by immunoblotting with ubiquitin antibodies. As references, we used strains expressing Myc-tagged Pop1p or Pop2p in a wild-type background. The experiment demonstrated that Pop1p and Pop2p each associate with polyubiquitylation activity even in the absence of their respective heterooligomerizing F-box proteins . Thus Pop1p and Pop2p appear to assemble into distinct SCF complexes bearing ubiquitin ligase activity in vitro. Discussion : Molecular architecture of SCFPop complexes | The phenomenon of F-box protein oligomerization is not unique to SCFPop. Although the crystal structure of the SKP1 --SKP2 complex, derived from bacterially expressed proteins, revealed a single SKP2 monomer bound to SKP1 , biochemical studies showed that budding yeast Cdc4p, a close homologue of Pop1p\Pop2p, forms homooligomers when expressed in insect cells (Correll and Deshaies, personal communication). Similarly, beta-TRCP1 and 2, which target IkappaBalpha for degradation as homooligomers, form heterooligomers that each bind SCF core subunits, although no biochemical activity for this heterooligomeric complex was demonstrated . Finally, Pop1p and Pop2p homooligomerize, at least when overexpressed , indicating that both F-box proteins may also be present as homooligomers in individual SCFPop1p and SCFPop2p, respectively. A surprising finding of this study was that the F-box of Pop2p is dispensable for Rum1p degradation and ploidy control, while the F-box of Pop1p is essential for both functions (Fig. and ). The Pop2p F-box is unlikely to be a degenerate, non-functional, and hence dispensable, motif, as it carries all signature residues of the F-box . In addition, Pop2p, in the absence of Pop1p, assembles into a complex containing all SCF core subunits identified here . At present, we cannot exclude that residues outside the F-box mediate binding of F-box deleted Pop2p to Psh1p and other core subunits. Consistent with this idea, biochemical studies based on the crystal structure of the human SKP1 --SKP2 complex revealed cooperation of the SKP2 F-box with an adjacent region in binding of SKP1 . Consistent with this finding, certain truncation mutants of the budding yeast F-box proteins Grr1p and Cdc4p interact poorly with Skp1p, despite the retainment of their F-boxes . On the other hand, the F-box of Pop1p is essential for Rum1p degradation , arguing that residues outside the F-box are insufficient to mediate recruitment of Pop proteins into SCF complexes. Based on the non-essential function of the Pop2p F-box, we propose a molecular architecture of SCFPop1p-Pop2p, in which Pop2p is tethered to the core subunits through interaction with Pop1p. Although we have no direct biochemical evidence to confirm this proposition, which would require Pop2p point mutants deficient in Pop1p binding, our data show that severely truncated N-terminal fragments of Pop1p and Pop2p lacking both F-boxes and WD repeat domains are sufficient to mediate their interaction . Similarly, dimerization of beta-TRCP proteins is mediated by N-terminal "D-domains" lacking binding of SKP1 and other SCF core subunits . According to our model, the F-box of Pop2p would be essential only for the Pop1p-independent activities of Pop2p, for which we provide tentative evidence by demonstrating the in vivo assembly of SCFPop2p complexes . These complexes bear ubiquitin ligase activity in a substrate-independent in vitro assay . To what extent this assay reflects the in vivo activity of SCFPop2p will become testable, once the putative substrates of SCFPop2p are identified. Subcellular compartmentalization of SCFPop as a potential mechanism for substrate selection | Another surprising observation of this study was that Pop1p is primarily localized to the nucleus, whereas Pop2p is present in both the cytoplasm and the nucleus . While, nuclear localization was expected, since both known substrates of SCFPop1p-Pop2p, Cdc18p and Rum1p, are primarily nuclear proteins, the cytoplasmic localization of Pop2p suggests an activity of SCFPop2p directed toward unknown cytoplasmic substrates. In support of this notion, as with human SCF subunits , fission yeast SCF core subunits are also present in the cytoplasm, as shown here for overexpressed GFP-fusions , and for endogenous Pcu1p by immunostaining in a previous report . By analogy, additional nuclear substrates of SCFPop1p may exist. For example, Pop1p is involved in the control of the RNA levels of the cyclin Cig2p . While it is unclear whether this effect is mediated at the level of transcription or mRNA stability, the budding yeast F-box protein Met30p was recently shown to regulate the ubiquitylation and activity, but not degradation, of the transcription factor Met4p . In addition, pop1 mutants display an increased rate of chromosome loss, a phenotype that is not easily explained by accumulation of Rum1p and Cdc18p . Finally, pop1 mutants are sensitive to UV irradiation, whereas pop2 mutants are not (D. Griffiths and D.A.W., unpublished observation). It is therefore likely that other substrates of Pop1p and Pop2p, in addition to their common substrates Cdc18p and Rum1p, do exist. The idea developed above that F-box protein compartmentalization contributes to substrate selection was recently confirmed directly for Cdc4p-mediated degradation of Far1p in budding yeast . Both, Cdc4p and Far1p are nuclear proteins in vegetative cells, owing to the presence of nuclear localization signals (NLS). Fusion of Cdc4p with a nuclear export signal (NES-Cdc4p) prevented its nuclear localization and its ability to direct the degradation of nuclear Far1p. When Far1p also was targeted to the cytoplasm by disrupting its NLS, NES-Cdc4p degraded DeltaNLS-Far1p in the cytoplasm . A putative NLS is also present in the N-terminus of Pop1p located between the F-box and the WD-repeat domain (393PEKIKRC). An N-terminal Pop1p fragment containing this motif, when fused to the WD-repeat region of Pop2p, is targeted exclusively to the nucleus (R.L. & D.A.W., unpublished observation). Like wild-type Pop2p, the reverse Pop2p-N/Pop1p-C chimera localizes to both the nucleus and the cytoplasm, again indicating that Pop1p, but not Pop2p, has a functional NLS in its N-terminus (R.L. & D.A.W., unpublished observation). It is unclear, at present, what regulates Pop2p distribution. One possibility is that Pop2p is co-imported in a complex with other SCF subunits that is preformed in the cytoplasm. Since Pop2p distribution is independent of Pop1p , SCF core subunits are the most likely candidates for such a function. In line with this suggestion, it was previously demonstrated that HRT1/ROC1/RBX1 promotes nuclear accumulation of CUL1 . Conclusion : Our data suggest homo- and heterooligomerization of the F-box proteins Pop1p and Pop2p as a mechanism for generating combinatorial diversity of SCF function in fission yeast. A heterooligomeric SCFPop1p-Pop2p complex mediates polyubiquitylation of phosphorylated Rum1p. In addition, compartmentalization of homooligomeric SCFPop1p and SCFPop2p complexes may direct the ubiquitylation of unknown nuclear and cytoplasmic substrates. Methods : Plasmids and yeast strains | S. pombe genes for psh1 and pip1 were identified in the Sanger Centre S. pombe sequence database based on their homology to the respective human and budding yeast proteins. Complementary DNAs were amplified by PCR, subcloned into pREp81.6xHis-Myc, pRep3.6His-Myc, or pRep4.HA, and sequenced. Deletion strains and epitope-tagged stains were constructed by one-step gene replacement using PCR-generated fragments containing kanamycin or ura4 cassettes . Growth media, flow cytometry, and all other relevant S. pombe techniques were described previously . To generate the pop1::ura4 pop2-13myc-kan strain, an h+pop1::ura4 ura4-d18 leul-32 pRep81.pop1 strain was crossed with a h-leul-32 ura4-d18 pop2-13myc-kan strain, followed by selection of spores on G418/ura- EMM plates. The pRep81.pop1 plasmid required to complement the sterility of the pop1 deletion strain was lost by growth in non-selective media (YES) for several generations. The resulting strain was verified by PCR and immunoblotting. The pop2::ura4 pop1-13myc-kan strain was generated in an analogous fashion. Antibodies | Rabbit antisera were raised at Josman LCC (Napa, CA) against bacterially expressed MBP-Psh1p, GST-Pcu1p, GST-Pip1p, and GST-Pop1p. Sera were affinity purified on affinity matrices containing immobilized GST-Psh1p, MBP-Pcu1p, MBP-Pip1p, and MBP-Pop1p. Column eluates were concentrated to ~1mg/ml and tittered by immunoblotting. Rabbit antisera against Pop2p and Rum1p were described before . Monoclonal Myc and HA antibodies were purified from 9E10 and 12CA5 tissue culture supernatants by binding to protein A. Immunoprecipitation and immunoblotting | Protein lysates for immunoblotting were prepared by bead lysis in a Fastprep device (Bio 101) in the presence of proteinase inhibitors, followed by boiling in SDS sample buffer. Cell lysates for small scale immunoprecipitations were prepared by disrupting cells in immunoprecipitation buffer (20 mM Tris/HC1, pH7.4; 150 mM NaC1; 0.5% Triton X-100, 10 ug/ml leupeptin, 10 ug/ml pepstatin, 17 ug/ml aprotinin, 1 mM PMSF). Lysates were cleared and precipitated with the respective antisera. Immunocomplexes were collected by binding to protein A or G beads, washed and analyzed by immunoblotting as described . Large scale lysates for gel filtration and subsequent immunoprecipitation were obtained by bead beater lysis. Approximately 5 mg of total cell lysates was separated by gel filtration on a 16/60 S300 column (Amersham Pharmacia Biotech), and 1 ml fractions were immunoprecipitated with Pip1p antibodies. Precipitates were fractionated by SDS PAGE and assayed by immunoblotting with the respective antisera. Indirect immunofluorescence | Indirect immunofluorescence staining was performed exactly as described . In vitro ubiquitylation assay | For ubiquitylation reactions, Pip1p complexes were immunoprecipitated from 100 -- 200 ug total cell lysates prepared as described above. Precipitates were washed four times in 20 mM Tris/HCI, pH 7.4; 150 mM NaCl; 0.5% Triton X-100, 10 mug/ml leupeptin, 10 mug/ml pepstatin, 5 mug/ml aprotinin, 1 mM PMSF, and equilibrated in 20 mM HEPES, pH 7.4, 100 mM potassium acetate, 1 mM DTT. A cocktail was added that contained 8 muM ubiquitin, ATP regenerating system (2 mM HEPES at pH 7.4, 1 mM ATP, 30 mM creatine phosphate, 1 mM magnesium acetate, 0.15 mg/ml creatine kinase), reaction buffer (4 mM magnesium acetate, 1mM DTT), 500 nM bacterially expressed 6 x His-UBCs, 100 nM 6 x His-tagged human E1, 0.5 muM ubiquitin aldehyde in a volume of 15 ul. The reaction was started by addition of phosphorylated Rum1p. After 90 min at 30C, the reaction was terminated by the addition 5 x SDS sample buffer. Samples were separated on 11 % SDS-polyacrylamide gels and analyzed by autoradiography. Substrate-independent in vitro ubiquitylation activity was determined by immunoprecipitating Pop1p-13Myc or Pop2p-13Myc with Myc antibodies. The same cocktail as described above was added to the precipitates. Reaction products were determined by immunoblotting with ubiquitin antibodies (Zymed). Subcellular fractionation | Cells grown in YES were harvested and washed in buffer S (1.4 M sorbitol, 40 mM HEPES (pH = 7.2), 0.5 mM MgCl2). Cells were resuspended in buffer S, 1 mM PMSF, 10 mM beta-mercaptoethanol and incubated for 10 min at 30C. Cells were pelleted, resuspended in 4 pellet volumes of buffer S, 1 mM PMSF, Zymolyase (100 ug/ml) and incubated at 30C for 40 min. Cells were diluted in buffer S and pelleted by centrifugation and resuspended in buffer F (18 % Ficoll 400, w/v; 20 mM HEPES (pH = 7.2), 0.5 mM MgCl2 and protease inhibitors. Cells were lysed by homogenization using a teflon pestel fitted into a microfuge tube. Cell lysis was monitored by microscopy. Unlysed cells were pelleted by gentle centrifugation. The lysate was placed on top of buffer GF (7% Ficoll 400, w/v, 20% glycerol, 20 mM HEPES (pH = 7.2), 0.5 mM MgCl2). Nuclei were pelleted by spinning at 7000 rpm in a microfuge. The cytoplasmic fraction was removed and mixed with SDS sample buffer. The nuclear pellet was resolved in an equal volume of SDS sample buffer. Fractions were analyzed by immunoblotting as described in Fig. . Abbreviations : SCF: SKP1/CUL1/F-box protein UBC: ubiquitin-conjugating enzyme HA: hemagglutinin NLS: nuclear localization signal NES: nuclear export signal Authors' contributions : All experiments were performed by VS. with the following exceptions: Building on reagents prepared by VS., CP. performed the experiments shown in Figs. , , and . IS. performed the experiments shown in Fig. , , and . ER. performed the immunostaining experiments shown in Fig. and . RL. performed the experiment shown in Fig. and contributed data not shown. KA. produced recombinant UBCs used in Fig. . CZ. prepared the csn5 mutant used in Fig. , contributed to the preparation of SCFPop antisera used in Figs. , , and , and assisted with study design. DAW. performed the experiment in Fig. , conceived the study, drafted the manuscript, and participated in study design and coordination. All authors read and approved the final manuscript. Backmatter: PMID- 12195021 TI - SAGE Genie: A suite with panoramic view of gene expression AB - Keywords: If the achievement of complete sequencing of the one-dimensional genetic codes of the human genome can be compared with man landing on the moon, the interpretation of genomic instruction in a four-dimensional biological context, such as during development and diseases, will prove to be a much more challenging and daunting task than that of getting man back from the moon to the earth. One of the greatest mysteries of life has been how a fertilized egg, which contains all of the genetic information that defines a living organism, can give rise to so many different tissues, which organize into different organs, as it divides and differentiates. It is clear that to unravel life's mysteries, we will have to rely, at least in large part, on tools that can allow us to determine when and where a gene is to be turned on or off in a cell as it divides, differentiates, and ages. Obviously, such tools are also important for the detection of when and where a seemingly precise interpretation of genomic instruction goes awry, which underlies many disease states such as cancer. Several technologies that show promises of high-throughput and potential for global analysis of gene expression were developed in the 1990s . However, the realization of these promises and potentials has been slow in coming, partly because of the lack of a unified standard for accurate data collection, analysis, and presentation for each methodology. As reported in a recent issue of PNAS , Boon and colleagues have made a major stride in this direction by developing a suite of bioinformatic tools that provides a single platform for compiling, annotating, and interpreting large sets of gene expression data collected by one of these technologies, serial analysis of gene expression, or more widely known as SAGE. SAGE technology, which was originally developed by Kinzler and Vogelsteins' group at Johns Hopkins University , is a clever high-throughput 3 expressed sequence tag (EST) counting methodology. Unlike the original brute-force EST sequencing strategy, where cDNA clones were randomly picked from cDNA libraries, SAGE technology measures the level of gene expression based on the frequency of occurrence of the 3' signature SAGE tags of 10 bases unique to each transcript. Because of the minimal sequence information necessary to define an expressed gene, or messenger RNA (mRNA), many SAGE tags from different genes can be obtained and sequenced at a time, which greatly speeds up the EST counting process. The method has been used successfully and extensively in the past for comparison of gene expression between a pair of RNA samples to identify differentially expressed genes within a given biological context . Such horizontal comparisons mainly focus on SAGE tags corresponding to genes that are either up- or down-regulated, whereas the bulk of the gene expression information, which took a great deal of effort to collect, often sits untapped. SAGE Genie is a logistically laid out suite of bioinformatic tools that allow automatic and reliable matches of SAGE tags to known gene transcripts. This process was accomplished first by filtering out experimentally obtained SAGE tags that had incorrect linker sequences, appeared only once, or were generated by sequencing errors, from millions of tags collected from over 100 different human cell types as part of the National Institutes of Health Cancer Genome Anatomy Project (CGAP). The resulting confident SAGE tags (CSTs) then were used to evaluate and match the virtual SAGE tags predicted from known mRNA transcript (cDNA) sequences of different publicly available databases, including full-length cDNAs or 3' ESTs. The virtual tags were divided into different groups based on the origin of the databases from which the tags were generated, the absence and presence of polyadenylation signals and poly(A) tails, and whether the tags represented differentially spliced or internal (non-3) transcript sequences. The match in percentage of virtual tags to CSTs allows ranking of available databases with known transcript sequences. Reciprocal cross-referencing between virtual tags and CSTs provides not only the best match of a CST to a known gene transcript sequence, but also confirmation that experimentally obtained SAGE tags indeed come from mostly 3' ends of mRNA transcripts. The resulting bioinformatic interface allows automatic tag-to-gene identification, measurement of gene expression normalized to the occurrence of a tag per 200,000 tags collected from a SAGE experiment, and the origins from which a tag is counted. Thus, SAGE Genie provides a computational platform on which not only more than two horizontal comparisons (e.g., normal brain versus brain tumors; Fig. ), but also a nearly infinite number of vertical comparisons (e.g., different tissue or organ types) in gene expression at a global scale can be conducted. The data output can be presented with interfaces such as the Anatomic Viewer, Digital Northern, and Digital Gene Expression Display for any given SAGE tag or gene transcript of interest, thus providing a quick glance at when and where a gene may be expressed. With SAGE Genie, experimentally collected SAGE tags for each biological system can be continuously annotated and inputted into the growing number of unique CSTs. With increasing collections of both CSTs and virtual tags, SAGE Genie could prove to be a very powerful tool for archiving and analyzing the expression profile for any given gene under any biological context. SAGE Genie could prove to be a very powerful tool for archiving and analyzing the expression profile for any given gene under any biological context. Figure 1 : Diagrams of the human brain showing gene expression levels in false-color (center bar) for a normal brain (Left) and or the expression levels for various brain tumors (Right). Diagrams of the human brain showing gene expression levels in false-color (center bar) for a normal brain (Left) and or the expression levels for various brain tumors (Right). The expression levels are determined by counting mRNA transcripts from each of these tissues. Expression levels are archived for nearly any human gene, and are displayed on-line using CGAP's SAGE Genie. In contrast, DNA microarray methodology , which has received much attention recently in the field of gene expression analysis, is still lacking a unified standard for both data collection and analysis . This makes gene expression data archiving and comparison from different research groups difficult. One of the major challenges for microarrays has been determining that a hybridization signal is specific to a known sequence laid on a chip when a complex cDNA probe is used, whereas methodologies such as SAGE are sequence-dependent in gene identification, which is more accurate. In fact, a cDNA probe used for microarrays can be so complex that it consists of as many as 10,000 different species, ranging from a few to thousands of copies per cell. Further compounding the problem in signal specificity for microarrays has been the fact that eukaryotic genes often come in families with many conserved sequences among the family members. Also for microarrays, one is limited to the detection of whatever genes are spotted on a slide, making it a closed system for gene discovery, unlike SAGE and Differential Display , which are open system-based gene screening procedures capable of identifying both known and novel gene transcripts. Although there is no doubt that SAGE Genie has greatly enhanced the utility of SAGE in global analysis of gene expression, challenges remain for the method with regard to the comprehensiveness in gene coverage as a function of the number of tags needed to be counted for each SAGE screen , and SAGE tags that either failed to match any known gene transcript sequences or matched more than one known transcript. Messages that failed to be represented because of the lack of anchoring restriction enzyme site are estimated to be as low as 1%. These, in time, may be overcome by increasing the number of SAGE tags collected for future SAGE screens, and the use of longer SAGE tags or different anchoring enzymes. Such improvements may further increase the power of SAGE Genie in archiving and the analysis of gene expression in model tissues /organs or biological systems. But for biomedical and agricultural research, there seem to be an infinite number of comparisons in gene expression with different biological systems, disease states, developmental stages, drug treatment, and stress conditions, etc., which need to be conducted. Such efforts will still require the use of technologies such as arrays and Differential Display as well as SAGE for custom gene-expression analysis. Nonetheless, with an intuitive web-site-based interface, SAGE Genie offers one of the most comprehensive collections of gene expression data across many different cancer and tissue types, making it a valuable tool for a quick glimpse of expression patterns of any known human gene sequences with the need of only a few strokes on a computer keyboard. In contrast to commercial EST collections and microarray databases, one more added bonus for SAGE Genie is that it is accessible to all, free of charge. Backmatter: PMID- 12185244 TI - Formation of geometrically complex lipid nanotube-vesicle networks of higher-order topologies AB - We present a microelectrofusion method for construction of fluid-state lipid bilayer networks of high geometrical complexity up to fully connected networks with genus = 3 topology. Within networks, self-organizing branching nanotube architectures could be produced where intersections spontaneously arrange themselves into three-way junctions with an angle of 120 between each nanotube. Formation of branching nanotube networks appears to follow a minimum-bending energy algorithm that solves for pathway minimization. It is also demonstrated that materials can be injected into specific containers within a network by nanotube-mediated transport of satellite vesicles having defined contents. Using a combination of microelectrofusion, spontaneous nanotube pattern formation, and satellite-vesicle injection, complex networks of containers and nanotubes can be produced for a range of applications in, for example, nanofluidics and artificial cell design. In addition, this electrofusion method allows integration of biological cells into lipid nanotube-vesicle networks. Keywords: Introduction : The last two decades have witnessed a tremendous development in miniaturization of fluidic devices. The rapid progress in processing hard materials such as silicon and metals , polymeric materials such as polydimethylsiloxane , and parylenes together with advancements in flow regulation have made it possible to manufacture complex chip structures for a wide range of applications, including chemical kinetics , computations , and chemical analysis . The ultimate fluidic device is one that can handle single molecules and colloid particles. Such devices require unprecedented control over transport and mixing behaviors, and to advance current fluidics into the single-molecule regime, we have to develop systems having physical dimensions in the nanometer scale. To create such devices, we can draw much knowledge from biological systems. For example, the Golgi-endoplasmic reticulum network in eukaryotic cells has many attractive features for sorting and routing of single molecules, such as ultra-small-scale dimension, transport control, and capability to recognize different molecular species, and for performing chemical transformations in nanometer-sized compartments with minimal dilution. It is, however, extremely difficult to mimic these biological systems by using traditional microfabrication technologies and materials because of their small scale, complex geometries, and advanced topologies. Furthermore, it is difficult to implement traditional flow regulation methods on nanoscale systems. Our efforts are focused mainly on the development of soft microfabrication technologies for processing of fluid-state liquid crystalline bilayer membranes. These materials have unique mechanical properties allowing creation of nanoscale structures, such as spheres and tubes with extremely high curvatures . The geometry of such structures is governed by both self-assembling and self-organizing properties of the lipid membrane material and can be changed on-line. We want to use these features for development of devices for transport and mixing of extremely small volumes of liquids (10-12 --10-18 liters) containing different reactants . Such systems can be used as generic platforms for fluidic devices with applications in, for example, chemical kinetics, membrane mechanics, chemical analysis, and computation. In earlier work, we have demonstrated techniques for formation of lipid nanotube-vesicle networks (NVNs) , consisting of surface-immobilized vesicle containers interdigitated with lipid nanotubes. A limitation with these techniques is the difficulty in controlling the connectivity of the networks, as they can be used only for formation of genus zero (g = 0) structures terminated by a vesicle container. To form more complex structures of higher-order topologies, vesicles within the networks must be connected by membrane fusion. Fusion of lipid membranes can be stimulated by electric fields and has been performed at the single-cell /liposome level by the use of microelectrofusion techniques . Here we demonstrate a micropipette-assisted electrofusion protocol for formation of NVNs having complex geometries and higher-order topologies. Such structures include circular networks as well as fully connected networks with three-dimensional nanotube layers. In addition, we show that this protocol can be used for connecting biological cells to lipid membrane networks. We also demonstrate how the complexity of these structures can be further increased by incorporating self-organizing branching lipid nanotube networks. Finally, we show how containers within these networks can be modified and differentiated independently with respect to interior contents by using a strategy for delivery of material based on the self-organizing properties of lipid membranes under lateral tension. Materials and Methods : Liposome Preparation. | Preparation of unilamellar soybean lecithin liposomes was achieved by the dehydration /rehydration method described by Criado and Keller with modifications . Briefly, 5 mul of aqueous lipid dispersion (1 mg /ml) was placed on a coverslip and placed in a vacuum dessicator at 25C for 30 min. The partially dehydrated lipid film was then carefully rehydrated with buffer solution (5 mM Trizma base /30 mM K3PO4 /30 mM KH2PO4 /1 mM MgSO4 /0.5 mM EDTA, pH 7.8). After a few minutes giant unilamellar liposomes started to form and a small sample of this liposome suspension was transferred to a droplet of buffer solution placed on a borosilicate coverslip. Microscopy, Fluorescence, and Bright-Field Imaging. | The coverslips with liposome suspension were placed directly on the stage of an inverted microscope (Leica DM IRB, Wetzlar, Germany). The 488-nm line of an Ar+ laser (2025-05, Spectra-Physics) was used for epifluorescence illumination. To break the coherence and scatter the laser light, a transparent spinning disk was placed in the beam path. The light was sent through a polychroic mirror and a x40 objective (Leica PL Fluotar) to excite the fluorophores. The same objective collected the fluorescence and a charged-coupled device camera (C2400-41H, Hamamatsu Photonics, Hamamatsu City, Japan) controlled by an Argus-20 image processor (Hamamatsu Photonics) was used to capture the images. Recordings were made by using a Super VHS (Panasonic S-VHS AG-5700, Stockholm). The same camera set-up was also used for differential interference contrast imaging. Digital image editing was performed by using the Argus-20 system and Adobe premiere and photoshop graphic software. The borosilicate cover slips (24 x 60 x 0.17 mm, Knittel Glaser, Braunschweig, Germany) used for the experiments were cleaned by rinsing in ethanol followed by deionized water. Micropipette-Assisted Formation of Unilamellar Networks. | A carbon fiber microelectrode (5 mum diameter, Dagan Instruments, Minneapolis) and a tapered micropipette, controlled by high graduation micromanipulators (Narishige MWH-3, Tokyo, coarse manipulator: Narishige MC-35A) were used to create unilamellar NVNs with a microelectroinjection technique . Tapered micropipettes were made from borosilicate capillaries (GC100TF-10, Clark Electromedical Instruments, Reading, U.K.), pulled on a CO2 laser puller instrument (model P-2000, Sutter Instruments, Novato, CA). A microinjection system (Eppendorf femtojet) and a pulse generator (Digitimer Stimulator DS9A, Welwyn Garden City, U.K.) were used to control the electroinjections. Chemicals and Materials. | Chloroform, EDTA (titriplex III), magnesium sulfate, potassium dihydrogen phosphate, and magnesium chloride (all pro analysis) were from Merck. Fluorescein (GC-Grade), potassium phosphate (>98%), and Trizma base (>99.9%) were purchased from Sigma. DiO (3,3'-dioctadecyloxacarbocyanine perchlorate) and TransFluoSpheres (40-nm, 488 /605-nm Ex /Em) were from Molecular Probes. Glycerol was from J. T. Baker and deionized water from a Milli-Q system (Millipore) was used. Soybean lecithin (polar lipid extract) was from Avanti Polar Lipids. The polar lipid extract consisted of a mixture of phosphatidylcholine (45.7%), phosphatidylethanolamine (22.1%), phosphatidylinositol (18.4%), phosphatidic acid (6.9%), and others (6.9%). Results and Discussion : Formation of Two-Dimensional Closed Networks by Microelectrofusion. | To construct topologically complex networks by electrofusion, open lipid bilayer NVNs were first formed. These type of networks are topological spheres (g = 0), having multiple surface-immobilized vesicle containers (5 --50 mum in diameter) interdigitated with nanotubes (100 --200 nm in diameter). The networks were made by a micropipette-assisted technique from giant unilamellar, or thin-walled, vesicles attached to multilamellar vesicles as described . With this method, networks can be produced with controlled nanotube length, angle between nanotube extensions, and vesicle container diameter. In brief, a tapered borosilicate-glass micropipette with an outer-tip diameter of 0.5 --1 mum, back-filled with aqueous medium, and mounted onto an electroinjection system , was pressed against the membrane of a surface-immobilized vesicle. By applying dc-voltage pulses of field strengths between 10 and 40 V /cm and duration of 1 --4 ms over the micropipette, the lipid membrane was penetrated (Fig. A). The micropipette was then slowly pulled out and away from the mother vesicle, forming a lipid nanotube connection between the mother liposome and the pipette tip (Fig. B). Aqueous medium was thereafter injected into the nanotube with a pressurized air-driven microinjector, thus forming a small satellite vesicle at the outlet of the micropipette (Fig. C). This newly created vesicle could then be immobilized onto the substrate surface at desired coordinates by the application of an axial force and was subsequently released from the pipette tip. Figure 1 | Schematic sequence showing the formation of lipid nanotube-vesicle circuits. Schematic sequence showing the formation of lipid nanotube-vesicle circuits. Surface-immobilized vesicles are marked with an X, and satellite vesicles that are "free" in solution are drawn as open circles. (A) The membrane of a giant vesicle is penetrated by a combination of mechanical force (F) applied from the micropipette and anodic electric pulses (+V) from a low-voltage pulse-generator (V). As a counter electrode a 5-mum carbon fiber is used. (B and C) A nanotube is created by pulling the micropipette away from the mother vesicle, and a small satellite vesicle is created at the end of the nanotube by injecting buffer solution into the nanotube orifice. (D) The satellite vesicle is positioned in close contact to another vesicle container within the network. Fusion of the vesicle containers is stimulated by application of one or several transient rectangular dc-voltage pulses and mechanical force. (E and F) The micropipette is withdrawn from the daughter vesicle. Lipid tubes adhering to the pipette tip after removal from the daughter vesicle were detached by applying one or several cathodic electric pulses. For creation of a closed network (g = 1), the small satellite vesicle, preferably having a diameter of 4 --6 mum and still adhering to the pipette tip, was positioned in close contact to another vesicle container within the network (Fig. D). To obtain a highly focused electrical field over the two containers, a carbon fiber microelectrode was placed adjacent to the fusion partners. Fusion of the vesicle containers was stimulated by application of one or several transient rectangular dc-voltage pulses of field strengths between 40 and 80 V /cm and duration of 1 --4 ms over the micropipette. Withdrawal of the micropipette could be performed without visible signs of vesicle deformation or leakage and residual lipid tubes still adhering to the pipette tip could be removed by applying one or several cathodic dc-voltage pulses over the micropipette (Fig. E and F). As the vesicle fusion completed, the diameter of the nanotubes instantly increased from an original diameter of 100 --200 nm, up to several mum in diameter for a few seconds before they regained their original diameter. Because liposome fusion is a very fast event that typically is completed in <100 ms, a very rapid change in the surface-to-volume ratio is achieved, causing an instant lowering of membrane tension in the network, and thus a lowering in surface free energy. Because the nanotube radius, rt, is governed by the lateral membrane tension, Tm, of the system according to: where kc is the bending modulus ; the change in lateral tension explains the dramatic modulation of the nanotube diameter. Over time, however, the lowering in surface energy is balanced by vesicle adhesion and a new mechanical equilibrium will be reached . Consequently, the fusion procedure induced a transient instability in the system; however, the structural integrity of the network always remained intact after this perturbation. Detailed studies on the stability of these structures have not been undertaken; these structures, however, appear to be stable for several hours. Over time, the vesicle containers will spread on the substrate surface, ultimately inducing a structural collapse of a network. The electrofusion of two vesicles is simple to perform, and several attempts can be made until fusion is achieved. In our hands single fusion attempts were close to 100% successful. For construction of complex networks, the success rate was somewhat lower. For example, we estimate that closed four container networks (g = 1) can be constructed at a success rate of ~75% by a person skilled in the art. Experience in micromanipulation and micropipette techniques, such as microinjection, is advantageous for successful usage of these protocols. All of the membrane structures shown in this work were completed within 30 min from the start of the respective experiment, and the majority of the structures have been reproduced several times. The protocol presented here is not limited to connection of vesicle containers within a network. The technique also allows incorporation of external membrane structures, such as solitary vesicles and biological cells, into a network. Fig. shows an example where a surface-immobilized vesicle stained with the fluorescent dye DiO was connected through a nanotube to an adherent NG-108 cell. The capability of connecting cells and synthetic vesicles opens up interesting possibilities in integrating and probing biological functions in biohybrid networks. In particular, using methods for delivery of materials in lipid nanotubes exchange of materials between cells and vesicles can be controlled. For example, small cytoplasmic samples can be taken from the cell and delivered through the nanotube to a vesicle containing a reporter system in sensor applications, or cell-affecting agents can be delivered from a vesicle to a cell. Figure 2 | Integration of biological cells. Integration of biological cells. A surface-immobilized giant vesicle was connected through a nanotube to an adherent NG-108 cell by using a combination of micropipette-assisted daughter vesicle formation and electrofusion. (A) A differential interference contrast micrograph of the cell-vesicle system. (B) The corresponding fluorescence micrograph is shown and is pseudo-colored for enhanced visibility. The liposome membrane was stained with the fluorescent dye DiO (0.5 mol%). Both images are composed of a z-directional stack of 16 video frames, explaining the bent appearance of the nanotube. (Scale bar: 10 mum.) If we view surface-immobilized vesicles as vertices in a coordinate system, this method allows connection of any vertex with an edge (nanotube connection) irrespective of the vertex coordinates. Consequently, closed NVNs up to fully connected networks [n containers, and n(n-1) /2 nanotubes], the basis for Hopfield-type architectures , can be created. Fig. A shows a fluorescence micrograph of a fully connected network having four containers and six nanotubes made by using a combination of micropipette-assisted satellite vesicle formation and microelectrofusion as illustrated in Fig. B. This structure represents a g = 3 topology, and in principle, the generic fusion scheme presented here can be used repeatedly on a given network to produce structures of very complex geometries and higher-order topologies. Note that the crossing diagonal nanotubes lie in separate planes as illustrated in Fig. C, making it possible to construct overlaid nanotube network architectures. For example, in fully connected networks the maximum number of layers, NL is given by NL = nt - nc; infinity > nc >= 4 (where nt is the number of nanotubes, and nc is the number of containers) which means that for a six-, seven-, and eight-container network, NL will be 9, 14, and 20, respectively. Figure 3 | Formation of fully connected networks. Formation of fully connected networks. (A) A fluorescence micrograph of a four-container Hopfield-type network constructed by using a combination of micropipette-assisted daughter vesicle formation and microelectrofusion. The lipid membrane is stained with the fluorescent dye DiO (0.5 mol%). (Scale bar: 10 mum.) (B) Schematic illustration of the procedure. The intersecting nanotubes in the middle of the network do not represent a nanotube four-way junction; the paths of two nanotubes are instead crossed in an overlaid fashion as illustrated in C. Geometrical Self-Organization of Networks. | One of the most fascinating features of these continuum membrane structures is their dynamical self-organizing behavior. This feature is governed by the material properties of the fluid-state bilayer membrane. Following the ansatz of Helfrich , the curvature energy per unit area of thin elastic shells is a quadratic function of the principal curvatures, c1, c2: where c0 is the spontaneous curvature, and kc and kc' are the bending modulus and saddle splay modulus, respectively. The equilibrium shape of a free vesicle in solution can be found from minimizing the elastic energy of bending: In our system, vesicles are immobilized on a surface, and an energy adhesion term has to be introduced : where f is the effective potential of adhesion and S* is the area of surface-membrane contact. Because the membrane material located in the tubular segments of a network is trapped in a region of extreme curvature, the tubes are residing in an elastically excited state. In addition, vesicle containers in these networks are in the strong adhesion regime. Consequently, these structures are under lateral membrane tension, further stretching the tubes. Therefore, the tubular segments of a NVN are forced to connect between containers in a way describing the shortest distance to lower the surface free energy of the system. The fluidity of the membrane material allows translation of a nanotube across both nanotube and vesicle surfaces to any position in a network . Therefore, nanotubes emanating from a common vesicle container must be separated by a distance, d, at the vesicle-nanotube interface to preserve the geometry of the system. If two nanotubes are positioned in such a way that there is no separation distance (d = 0), the tubes will coalesce and arrange themselves into the minimum pathway solution of the specific geometry set by the vertex coordinates of the connected containers. Fig. A shows schematically how such structures can be created from a V-shaped three-container network by forcing the nanotubes emanating from the vesicle in the central position to coalesce. As the two nanotubes merge, a three-way junction will form and move to the coordinates describing the minimum pathway. Provided that all of the nanotubes in the structure have the same tension, and thus diameter, simple geometric considerations give that the shortest way to connect three vertices is via an intersection having 120 angles. Figure 4 | Dynamic self-organization of networks. Dynamic self-organization of networks. (A) Schematic showing how nanotube three-way junctions are formed by forcing two nanotubes emanating from the central vesicle in a V-shaped sequential three-container network to coalesce. As the two nanotubes merge, a three-way nanotube junction will form and move to the coordinates describing the minimum pathway arrangement of the nanotubes. (B) Pathway minimization is stimulated in a sequential seven-container network by forcing the tubes marked with black arrows to coalesce. The system was triggered by hydrodynamical displacement of a nanotube by flushing with buffer solution from the micropipette. (C) The final configuration of the network. The system now contains nine tubular segments and three three-way junctions (black arrows), all having separation angles of 120. (Scale bar: 5 mum.) This self-organizing behavior can be used for increasing the complexity of the networks. Fig. B and C shows an experiment where a seven-container network having a total tube length of 120 mum evolves into a lower energy configuration by geometrical rearrangement of nanotubes. The event was triggered by forcing two nanotubes emanating from a mutual vesicle container to coalesce, forming a three-way junction (Fig. B). As this junction moved toward its minimum pathway solution, the nanotubes connected to this junction coalesced with neighboring tubes at their respective vesicle-tube interface, thus causing a domino-like effect where new three-way junctions were formed and transported toward a lower energy configuration. When all of the junctions reached their minimum pathway configuration, the movement of junctions was terminated. The new configuration contained nine tubular segments and three three-way junctions, all having separation angles of 120, and supported a total tube length of 85 mum as illustrated in Fig. C. Nanotube-Mediated Injection of Vesicle Contents and Formation of Complex Contents-Differentiated Networks. | As discussed above, the membrane material located in the nanotubes is trapped in an elastically exited state and therefore acts as mechanical springs that try to pull connected containers together to lower the energy of the system. Thus, satellite vesicles that are released from the pipette tip and that are not allowed to adhere to any surface can be used for highly controlled transport and injection of finite amounts of materials into a specific vesicle container in a network. Satellite vesicles exclusively contain solution originating from the micropipette and can carry volumes as small as a few femtoliters. We loaded satellite vesicles with fluorescein solution that were released from the pipette tip by application of a voltage pulse. After release, the satellite vesicle was transported and subsequently merged with the nanotube-conjugated vesicle into which it released its contents (Fig. A and B). Again, this transport is driven by minimization of elastic energy stored in the membrane material. Figure 5 | Nanotube-mediated injection of vesicle contents. Nanotube-mediated injection of vesicle contents. (A) Series of differential interference contrast micrographs illustrating how a small satellite vesicle created by the micropipette-assisted technique is released from the tip of the micropipette by application of a voltage pulse (40 V /cm, 4 ms). When the vesicle detaches from the pipette, it is immediately pulled toward its tube-conjugated mother where the two vesicles subsequently merge. (B) Series of intensity indexed fluorescence micrographs of the tube-mediated transport and injection of a fluorescein-filled (25 muM in buffer solution) satellite vesicle. Once the vesicles are in contact, coalescence typically occurs within 100 ms. A and B are different experiments. (Scale bar: 5 mum.) (C) Graph showing the pulling force of the nanotube-vesicle system as a function of tube length (mean +- SEM, n = 6). The satellite vesicles were released from a pipette tip into solution as illustrated in A and did not adhere to the substrate surface. This transport exhibits a nonlinear behavior and is composed of three distinct phases (I --III). If frictional forces acting on the nanotube are neglected and the lateral tension in the system is sufficiently high, the satellite vesicle will remain largely spherical, and the pulling force, F, of the system can be approximated from Stokes frictional drag past a sphere . As illustrated in Fig. C, the vesicle transport exhibited a nonlinear behavior composed of three distinct phases. Initially, when a satellite vesicle was released from the pipette tip, the system displayed an acceleration-deceleration peak, an effect most likely caused by the applied electric field used for releasing the vesicle. After this initial response, the pulling force decreased almost linearly with tube length. Before the vesicles merged, there was a distinct decrease in the measured pulling force. This last phase may be caused by alignment of the two nanotube-vesicle junctions. Typical pulling forces in the linear phase were 0.5 --1 pN. In Fig. , we demonstrate how a combination of micropipette-assisted network formation, microelectrofusion, and nanotube-mediated satellite vesicle transport and injection can be used for incorporating closed loops, branched nanotubes, and differentiated containers into a single NVN. The entire sequence for creating this network requires >20 different steps of which 12 are schematically illustrated in Fig. A. In this network, all vesicles and nanotubes were created from a single giant unilamellar-multilamellar vesicle. After the network was formed, a micropipette was loaded with 30 nm fluorescent latex beads and satellite-vesicle injection was used to deliver approximately 50 fl of latex bead solution each into two vesicle containers. As the construction of the network was completed, the nanotube connecting the giant unilamellar-multilamellar vesicle was cut by using a carbon fiber. A fluorescence image of the final network is shown in Fig. B. Figure 6 | Differentiation of networks. Differentiation of networks. (A) Schematic showing the procedure of creating a differentiated network having closed loops and branching nanotubes. Differentiation of the chemical composition of individual liposomes in the network was obtained by nanotube-mediated fusion of satellite vesicles containing red fluorescent 30-nm latex beads. (B) A fluorescence micrograph of the actual structure is shown. The membrane of the NVN is stained with DiO (0.5 mol%) The colors were detected by using separate channels and were overlaid by using adobe software. (Scale bar: 5 mum.) Conclusion and Outlook : We have presented an electrofusion protocol for formation of NVNs of highly complex geometries and advanced topologies, including closed circular networks as well as fully connected networks with three-dimensional nanotube layers. In addition, we have demonstrated how the self-organizing feature of lipid membranes can be used for incorporating branching lipid nanotubes, further increasing the complexity of these networks. Finally, we show how containers within closed circular networks could be modified and differentiated independently with respect to interior contents by using a satellite vesicle transport scheme. Networks produced by us previously all have been topological spheres, which limit their utility for many applications. According to the methods presented here, any vertex (container) can be connected by an edge (nanotube) in a predetermined fashion. We judge this to be an important step toward creating fluid-state membrane devices with applications in nanofluidics , as well as model systems for studies of single-molecule behaviors , synchronized population behaviors of enzymes in confined spaces , and diffusible behaviors of biological molecules . Moreover, cells and cell networks, as well as membrane proteins, can be reconstituted in these systems, enabling a combination of complex function with extreme spatial confinement. Backmatter: Abbreviations : NVN = nanotube-vesicle network DiO = 3,3'-dioctadecyloxacarbocyanine perchlorate PMID- 12186975 TI - SALSA, a variant of yeast SAGA, contains truncated Spt7, which correlates with activated transcription AB - Spt-Ada-Gcn5 acetyltransferase (SAGA) is a previously described histone acetyltransferase /transcriptional coactivator complex in yeast. At promoters of certain genes (HIS3 and TRP3), SAGA has an inhibitory function involving a nonproductive TATA-binding protein interaction mediated by the Spt3 and Spt8 subunits. Related to this, Spt8-less SAGA is a major form of the complex under activating conditions for these genes. In the present study, we purify this activation-specific complex, called SALSA (SAGA altered, Spt8 absent). Besides lacking Spt8, SALSA contains Spt7 subunit that is truncated. Examining the role of this subunit, we find that C-terminally truncated SPT7 resulted in derepressed HIS3 transcription. Furthermore, when grown in rich media (repressing conditions), wild-type cells yielded predominantly SAGA, but Spt7 C-terminal truncations resulted primarily in a form of complex similar to SALSA. Thus, SALSA-like structure and activating function can be partially recapitulated in yeast by truncating the C terminus of Spt7. Overall, these results lead to a model that for a subset of promoters SAGA is inhibitory through Spt3, Spt8, and an Spt8-interacting subdomain of Spt7, whereas SALSA is a form of complex for positive transcriptional regulation. These data clarify a mechanism by which a transcriptional regulatory complex can switch between positive and negative modulation. Keywords: Introduction : Transcription by RNA polymerase II in eukaryotes is highly regulated and involves various proteins in addition to the general transcriptional machinery. Coactivator complexes, which bridge the interaction between activator proteins and the transcription complex, have been shown to harbor various regulatory functions . Among these are activities, such as histone acetyltransferases (HATs) , that counteract repressive chromatin structure . In Saccharomyces cerevisiae, the HAT complexes NuA4 and SAGA (Spt-Ada-Gcn5 acetyltransferase) are coactivators that have been specifically implicated in acetylation and transcriptional regulation in vivo and in vitro . The SAGA complex, which is important for normal growth of yeast and transcription of various genes, contains several groups of previously identified transcription-related proteins . Known subunits include the HAT Gcn5 , Ada adaptor proteins (Ada1 through 5; Ada4 is Gcn5) , the TATA-binding protein (TBP)-related subgroup of Spt proteins (Spt3, Spt7, Spt8, and Spt20, also known as Ada5) , a subset of TBP-associated factors (TafII17, 25, 60, 68, and 90) , and the essential protein Tra1 , the yeast homolog of human coactivator TRRAP. Consistent with its diverse composition, SAGA is known to have a modular structure and contain several discrete regulatory functions : defined subgroups within SAGA include Gcn5 /Ada2 /Ada3 (HAT /adaptor), Spt3 /Spt8 (TBP-related function), and Ada1 /Spt7 /Spt20 (apparently required for SAGA structural integrity). The multiple functions within SAGA and their roles in transcription have been examined experimentally in various studies. The HAT activity of Gcn5 has been extensively studied in vitro and in vivo, and in the context of SAGA it is known to have a positive effect on transcription of certain genes (reviewed in ref. ). The activator-binding ability of SAGA [such as through Tra1 ] is another positive function shown to be critical for activation in transcription assays , and recently SAGA was also demonstrated to have the positive effect of facilitating transcription complex assembly at GAL1 . On the other hand, components of SAGA have also been shown to mediate an apparent negative effect on transcription at certain Gcn4-regulated genes such as HIS3 and TRP3 . Specifically, unscheduled derepression of these genes is promoted in vivo by mutation of SPT3, SPT8, or the TBP-encoding gene (spt15 --21, a point mutant suppressable by point mutant spt3 --401), and SAGA complexes lacking either Spt3 or Spt8 inhibit TBP binding to TATA box DNA in vitro , suggesting that a TBP-related repression mechanism is at work in SAGA at these promoters. The coexistence of positive and negative regulatory features or the interconversion between these functions has also been observed in other complexes involved in transcriptional regulation. For example, TFIID generally supports transcription, and overall, its Taf subunits act as positive cofactors . However, TAF250 homologs (such as Drosophila TAF230 and yeast TafII145) have been shown to have a negative, autoinhibitory effect on TFIID by preventing TBP-TATA interaction , but this may be counteracted by activators or TFIIA, possibly acting through TAF135 . Another example is that human coactivator PC4 can be converted from a repressive to an active form in a stepwise process that may involve phosphorylation by TFIIH and /or TAF250 . In addition, yeast or Drosophila transcription factor NC2 (Dr1-Drap1) can also either activate or repress transcription, in this case depending on the nature of the promoter it occupies . The structure and composition of certain regulatory complexes can also vary, influencing regulatory states. For example, human chromatin-remodeling complexes containing Brg1 can be purified in at least two different forms, largely similar but containing or lacking the regulatory protein mSin3A and several other subunits . In addition, the STAGA complex, a human homolog of SAGA, seems to exist in multiple forms with different functional moieties . Yeast SAGA is also a modular complex with variable subunit composition, and these properties may be directly related to its positive and negative transcriptional effects, mentioned above. The negative regulation of HIS3 and TRP3 apparently occurs through nonproductive interaction with TBP, mediated by the Spt3 and Spt8 subunits. Moreover, there are at least two similar, but structurally and chromatographically distinct, forms of the SAGA complex in yeast cells: classically defined SAGA and an altered form of SAGA lacking Spt8 . The prevalence of these forms shows an interesting correlation with SAGA's inhibitory function in that SAGA is predominant in rich media (repressive for HIS3 and TRP3), whereas the altered form of SAGA is predominant under activating conditions for these genes. Because a major characteristic of this latter complex is lack of the Spt8 subunit, we therefore name it SALSA, for SAGA altered, Spt8 absent. In the present study, we purify the SALSA complex to near homogeneity and further characterize its structural and apparent functional differences with SAGA. Specifically, the Spt7 C terminus has a role in Spt8 interaction with SAGA and may be an important distinguishing feature between SAGA and activation-specific SALSA. Materials and Methods : Yeast Strains and Media. | For large-scale purification of SAGA and SALSA, Ada2 --2FLAG epitope-tagged strain SB345 was used. Wild-type strain FY631 and spt7 mutants FY571 (spt7 --217) and FY569 (spt7 --223) were provided by the Winston lab and used for S1 analysis and conventional purification of complexes. SPT7-deleted strain FY963 was used for pY-SPT7 integrations and subsequent S1 analysis. For SPT7 integrations followed by anti-FLAG immunoprecipitations, haploid strains SB345 and SB303 were crossed to create a diploid, which underwent spt7Delta::LEU2 knockout and was sporulated and tetrad-dissected, followed by identification of an spt7Delta Ada2 --2FLAG-containing haploid. For anti-Myc immunoprecipitations, 13 tandem Myc tags were added to the C-terminal ends of the SPT3, SPT20, or SPT8 genes in the strain FY1531 (from F. Winston, Harvard Medical School, Cambridge, MA) by the method of Longtine et al. ; URA3-plasmid-borne SPT7 was removed by growth on 5-fluoro-orotic acid media , and SPT7 wild-type or Delta213C were integrated as described below. Rich (yeast extract /peptone /dextrose) and synthetic complete media were used for nonselective and selective growth, respectively, at 30C. Plasmids, Sequencing, and S1 Analysis of HIS3 Transcription. | To prepare SPT7 constructs for plasmid integration of spt7Delta strains, 3.1-kb and 5.2-kb ClaI /MluI fragments of pYES2 (Invitrogen) and pFW128 were ligated to produce pY-SPT7; a modification of the QuikChange mutagenesis method (Stratagene) was used to prepare C-terminal truncation constructs of this plasmid. These were digested with ApaI and transformed into FY963 for integration at the URA3 locus; for empty-vector integration (spt7Delta), NsiI-digested pRS306 was used. Sequencing of the spt7 --217 and spt7 --223 mutations was accomplished by preparing genomic DNA from the FY571 and FY569 strains and PCR cloning (TOPO-TA cloning kit, Invitrogen) SPT7 C-terminal regions into pCR2.1. Analysis of in vivo HIS3 transcription was carried out as described , by hybridization of 40 mug of total RNA with specific probes. SAGA and SALSA Purification from Ada2 --2FLAG Yeast Strain. | The complexes were purified as illustrated in Fig. A, a scheme adapted from previous methods . Yeast strain SB345 was grown in 60 liters of synthetic complete medium at 30C to an optical density (600 nm) of 1.5. Sixty milliliters of Ni2+-NTA agarose (Qiagen, Chatsworth, CA) was used for initial binding of whole-cell extract, and eluate was run on an 8-ml Mono Q HR 10 /10 column (Amersham Pharmacia). Peak SALSA and SAGA fractions (monitored by Western blots with antibodies to known shared subunits) were pooled separately, dialyzed against the buffer containing 0.1 M NaCl and 40 mM Hepes, pH 7.5, and loaded onto a Mono S HR 5 /5 column, followed by a 10 column volume, 0.1 --0.5 M NaCl linear gradient. Peak fractions were adjusted to 0.25 M NaCl and incubated with anti-FLAG M2-Agarose (Sigma) overnight. The resin was extensively washed on a column, and bound proteins were eluted with FLAG peptide (Sigma) at 0.2 mg /ml. Samples of SALSA and SAGA were then run on an SDS /PAGE gradient gel (4 --20%; Invitrogen) and stained with colloidal blue (Coomassie G-250; Invitrogen). Figure 1 | Purification and subunit composition of the SALSA and SAGA complexes. Purification and subunit composition of the SALSA and SAGA complexes. (A) Purification scheme for the isolation of SALSA and SAGA from Ada2 --2FLAG cells. IP, immunoprecipitation; cV, column volume. (B) Coomassie-stained gradient gel (SDS /PAGE) run with similar amounts of purified SALSA and SAGA. Molecular mass markers (kDa) are shown at left. Indicated at right are subunits identified by size and /or Western blotting. (C) Western blots of purified SALSA and SAGA with antibodies against various subunits in the complexes. Immunoprecipitation of Complexes and Western Blotting. | Extracts of yeast extract /peptone /dextrose-grown pY-SPT7 integrants of the spt7Delta Ada2 --2FLAG-tagged strain were immunoprecipitated with anti-FLAG antibodies and eluted with FLAG peptide as described . Immunoprecipitations of Myc-tagged strains were performed similarly, but with mouse anti-c-Myc mAb (9E10; Santa Cruz Biotechnology). A total of 5 --20 mul of the elutions was used for Western blots with SAGA subunit antibodies as described . Additional antibodies used in this study were anti-Ada1 (1:1,000 dilution) from L. Guarente (Massachusetts Institute of Technology, Cambridge, MA), anti-Spt7 (1:250) from F. Winston, and anti-Tra1 (1:1,000) from J. Workman (Pennsylvania State University, State College, PA). It should be noted that Saleh et al. previously performed an spt7Delta-extract immunoprecipitation experiment analogous to ours and observed Tra1 in the immunoprecipitate; however, the strain, epitope tags [hemagglutinin (HA)-Ada2 and Myc-Tra1], and immunoprecipitation conditions (anti-HA antibody, and 150 mM salt instead of 350 mM) differed significantly from ours. Results : Purification and Subunit Characterization of SALSA. | Previously, we identified an altered form of the SAGA complex predominant under activating conditions for certain genes [amino acid biosynthesis derepression by 3-aminotriazole (3-AT)], and its primary observed difference with SAGA was lack of the Spt8 subunit and shifted elution from a Mono Q ion exchange column . In the present study, we seek to characterize this altered complex (SALSA) further. To do this, we prepared an extract from epitope-tagged Ada2 --2FLAG cells grown in synthetic complete media [shown previously to result in approximately similar levels of SAGA and SALSA ] and purified the two complexes to near homogeneity by the purification scheme shown in Fig. A. These purifications resulted in bands mostly identifiable by size (Fig. B) and /or Western blot analysis (Fig. C) as known SAGA subunits. A subsequent Superose 6 gel filtration column of these preparations demonstrated that SALSA eluted in similar fractions to SAGA, and therefore the complexes are not discernibly different in size (1.8 MDa) within the resolution of the column (data not shown). Besides the presence /absence of Spt8, the most apparent difference between SAGA and SALSA is the gel mobility of the Spt7 subunit: it has a significantly smaller apparent size in SALSA compared with SAGA. Multiple species of Spt7 had been observed previously , although it was not determined whether the smaller forms of Spt7 were associated with specific Gcn5-containing complexes. It should be noted that aside from Spt7, SALSA and SAGA also show differences in a few faint bands in Fig. B (particularly near the mobility of the 116-kDa marker), but at this point it is unclear whether these are nonspecific, substoichiometric contaminants, altered species of other subunits, or unknown subunits unique to the particular forms of complex. Transcriptional Effects of SPT7 C-Terminal Truncation Mutants. | Because of initial evidence of the existence of C-terminally processed Spt7 in extracts (P.-Y. J. Wu and F. Winston, personal communication), we investigated the possibility that the faster gel mobility of Spt7 in SALSA represented a truncation of the protein. To explore this phenomenon further and determine whether it was related to our previous observations of Spt3 /Spt8-mediated transcriptional repression, we characterized two existing C-terminal truncation mutants of SPT7 ---spt7 --217 and spt7 --223. Previous results had shown that both mutants gave Spt- phenotypes and that the mutant proteins were stable in extracts . Furthermore, spt7 --223 was found to have a larger truncation and more severe phenotype than spt7 --217. To investigate possible transcriptional effects of these truncations and potential relationship with SALSA, gene expression was tested at the SAGA-dependent HIS3 gene by isolating RNA from cells under noninducing and inducing conditions (- and + 3-AT) and performing S1 nuclease assays (Fig. A). As observed in ref. , wild-type cells show a significant amount of HIS3 transcription (focusing on the +13 start-site band) under inducing conditions (Fig. A, lane 2), but display a low basal level of transcription under noninducing conditions (Fig. A, lane 1). In the case of spt7Delta, transcription is low under both conditions (Fig. A, lanes 7 and 8). The more dramatic truncation of Spt7, spt7 --223, displays a transcription profile (Fig. A, lanes 5 and 6) very similar to that of spt7Delta, suggesting that this mutant has lost Spt7 function. However, a different profile is observed in the spt7 --217 samples: basal transcription is elevated ~3-fold (Fig. A, lane 3) compared with wild type (Fig. A, lane 1). This high-basal effect is a hallmark of spt3Delta, spt8Delta, and spt15 --21 mutants, which disrupt the TBP-related negative regulation of SAGA, and the relative values are comparable to those of an spt8Delta strain (about 4-fold over wild-type basal) . The spt7 --217 results therefore indicate that the C terminus of Spt7 may also be involved in the same process. Figure 2 | HIS3 transcriptional analysis and sequencing of spt7 mutants. HIS3 transcriptional analysis and sequencing of spt7 mutants. (A) (Left) S1 nuclease assays of RNA isolated from wild-type and spt7 mutant strains under repressing and derepressing conditions (- and + 3-AT). tRNA was probed as a control for normalization. (Right) The quantitation of basal transcription (minus 3-AT) from HIS3's +13 start site, relative to wild type. (B) (Upper) Sequencing of the spt7 --217 and spt7 --223 mutant genes. Partial DNA sequences and predicted protein sequences (amino acids as one-letter codes) indicate the positions of the mutations (*) and introduction of premature stop codons. (Lower) A schematic diagram of the truncated proteins and the number of residues they contain is shown. Because the spt7 --217 and spt7 --223 alleles were not previously sequenced, we determined where these mutations occurred in SPT7, revealing the extent of the truncations (Fig. B). spt7 --217 had a single-base deletion at codon 1120, leading to six out-of-frame residues followed by a stop codon. The net result is that the last 213 residues of Spt7 are no longer coded for in the altered ORF. In spt7 --223, a single-base substitution occurred, changing codon 843 to a stop codon and removing the last 490 residues of Spt7. Notably, this removed the potential histone fold motif of Spt7, defined as between residues 975 and 1051 . To map the critical regions of the Spt7 C terminus more precisely, we prepared a series of truncations of Spt7, deleting the C terminus at 50-residue intervals (Fig. A), and used them for S1 analysis. Truncations of 50, 100, or 150 residues led to no discernible effect on HIS3 transcription (Fig. B Upper). A 400-residue deletion mutant displayed a profile similar to that of spt7Delta, like spt7 --223; although this truncation was slightly less extensive than spt7 --223, it did also remove Spt7's histone-fold motif. Unlike the smaller deletions, truncations of 200 or 213 residues did give high-basal effects (Fig. B Lower). Delta213C was designed to recapitulate the spt7 --217 mutation (it removes the same residues, albeit by deletion instead of frameshift), and this mutant does in fact display the same high-basal profile as spt7 --217, indicating that this effect is specific to Spt7 truncation and not to anything else in the spt7 --217 strain background (for example, the spt7 --217 strain, FY571, contains four additional auxotrophies compared with the FY963 integrants). In the case of Delta200C, the basal transcription level was nearly as high as that of Delta213C, suggesting that the region responsible for the high-basal effect has its C-terminal border between 150 and 200, and the region may continue beyond 200. Figure 3 | HIS3 transcriptional analysis of Spt7 C-terminal truncation mutant series. HIS3 transcriptional analysis of Spt7 C-terminal truncation mutant series. (A) Diagram of truncation mutations prepared in pY-SPT7 plasmid and integrated into spt7Delta strain FY963. (B) (Upper) S1 nuclease assays of RNA isolated from wild-type and spt7 mutant strains under repressing and derepressing conditions (- and + 3-AT). Lanes at right in the lower panel show direct comparison of +13 basal transcription (minus 3-AT) for repeated samples of wild type, Delta213C, and spt7 --217. (Lower) Quantitation of basal transcription relative to wild type. tRNA was probed as a control for normalization. Analysis of SAGA /SALSA Complexes from spt7 Mutants. | The above results suggest a link between Spt7 C-terminal truncation and the SALSA form of SAGA observed under activated conditions. To address this directly, we performed chromatographic preparations of HAT complexes from yeast extract /peptone /dextrose-grown (noninducing conditions) wild-type and spt7 --217 strains and analyzed the resulting Mono Q fractions for SAGA and SALSA . The wild-type fractions displayed a typical profile for this strain: abundant SAGA in fractions 38 --39 and a very small amount of SALSA about seven fractions earlier. spt7 --217, however, led to a dramatically shifted profile with the predominant form of Spt-Ada complex now eluting at or near the SALSA position. SAGA was still present (Fig. , lanes 38 and 39), albeit at a significantly reduced level. Figure 4 | Subunit analysis of conventionally purified complexes from wild-type and spt7 --217 cells grown in rich media. Subunit analysis of conventionally purified complexes from wild-type and spt7 --217 cells grown in rich media. Western blots of three Spt subunits show the relative amounts of the Spt /Ada complexes and the Mono Q fractions in which they elute. In the case of the earlier-eluting species from spt7 --217, it is unclear whether it is SALSA itself, or a mutant-specific, Spt8-less derivative of SAGA, or a mixture of both. The original chromatographic study of SAGA suggested that SPT7 deletion causes disruption of the complex, but it did not explore the issue of possible partial complexes in spt7 deletion or truncation mutant extracts. To investigate this, we immunoprecipitated Ada2 --2FLAG-containing complexes (which should include at least SAGA, SALSA, and the non-Spt-containing complex ADA) from extracts of cells with wild-type, Delta213C, or Delta400C Spt7 or with SPT7 deleted, and characterized them by Western analysis (Fig. A). In the case of Spt7Delta213C, all tested SAGA subunits were present with the exception of Spt8, again supporting the structural and functional relationship between the Spt7 C terminus and the participation of Spt8 in the complexes. This is further supported by an alternative set of immunoprecipitations (Fig. B), where SAGA /SALSA could be pulled down by Spt8 in the presence of wild type (Fig. B, lane 4) but not Delta213C Spt7 (Fig. B, lane 5), demonstrating that Spt8 is disconnected from the complex in the truncation mutant, even in rich media. However, because the conventional purification from spt7 --217 did show some Spt8-containing SAGA , Spt8 may have a weak interaction with SAGA in that mutant. Figure 5 | Subunit analysis of immunoprecipitated complexes from rich-media extracts of spt7 mutants. Subunit analysis of immunoprecipitated complexes from rich-media extracts of spt7 mutants. (A) Western blots of immunoprecipitated Ada2 --2FLAG complexes (should include ADA, SAGA, and SALSA) showing various Ada, Spt, and Taf subunits and Tra1 recovered from the indicated strains. (B) Western blots of immunoprecipitated complexes from extracts of strains with Myc-tagged SAGA components as indicated. As shown, one strain had C-terminally truncated Spt7Delta213C in combination with Spt8 --13Myc, whereas the others had wild-type Spt7. (C) Size comparison of truncated Spt7Delta213C with naturally processed Spt7 in SALSA. Purified SALSA and SAGA (side lanes) from Fig. were run alongside Ada2 --2FLAG-immunoprecipitated Delta213C and wild-type samples (center lanes) on a 6% SDS /PAGE gel, Western blotted, and visualized with anti-Spt7 antibodies. The migrations of molecular size markers are indicated at left. To verify that the spt7Delta213C mutant does in fact approximate the state of Spt7 in SALSA, we performed a direct size comparison of mutant-derived Spt7 with Spt7 from purified wild-type SALSA. As shown in Fig. C, the truncation mutant does lead to an Spt7 species very similar in apparent size to naturally processed Spt7 in SALSA, but the migration of the mutant protein is slightly faster, indicating that the normal processing site is slightly downstream of the mutant's truncation point. These results are therefore consistent with the mapping of the transcriptional effect, where a 150-residue truncation had no effect, whereas a 200-residue truncation gave a result similar to that of Delta213C (Fig. B). In the case of loss of Spt7 or its histone fold (Delta400C), SAGA /SALSA was disrupted, resulting in an adaptor complex containing just Ada2, Ada3, Gcn5, and Spt20 /Ada5 (Fig. A). Thus, as suggested previously, Spt7 is critical for integrity of the complexes. It may be significant that loss of the Spt7 histone fold correlates with complex disruption, because several other histone fold proteins are contained in SAGA --Ada1, Spt3, and Tafs 17, 25, 60, and 68. One possibility is that loss of a potential histone octamer-like structure could contribute to disruption of the complex. The data in Fig. A also point to a potentially integral relationship between Spt20 /Ada5 and the adaptorHAT module, as demonstrated by its apparent stable association with these subunits even in the absence of the rest of the complex. Spt7, on the other hand, seems to have a closer functional relationship through its C terminus to Spt3 /Spt8, and a structural relationship with the TafIIs, Ada1, and Tra1 through Spt7's histone fold. Discussion : In yeast, many highly inducible promoters are primed to be rapidly activated. Although the level of DNA-bound activators generally rises after induction, there is activator present at many promoters even in the repressed state. In addition, many coactivators are present constitutively in the nucleus. The paradox of needing rapid induction but at the same time preventing association with coactivators at repressed promoters has been solved by negative regulation inherent at many promoters. One mechanism is direct blockage of an activation domain, such as Gal80 for the activator Gal4 . A related mechanism is illustrated by Pho80 binding to the Pho4 activator, at a region distinct from the activation domain and inhibitory to activation . A different mechanism is to build negative regulation into coactivator complexes to prevent their activity, such as the TFIID complex's reversible inhibition of TBP-TATA interaction, described above. We have identified a second example of this latter strategy, within the yeast SAGA complex. Our data indicate that Spt7 and Spt8 within SAGA participate in negative regulation at the HIS3 promoter. Under derepressing conditions, Spt7 within SAGA is processed (perhaps by proteolysis), and Spt8 is not associated, resulting in SALSA, an apparent activated form of the complex. That SALSA's Spt7 is C-terminally truncated is supported by a very recent study that examined Spt7 with epitope tags at both ends and demonstrated that the form that copurifies with SALSA lacks the C-terminal tag, but not the N-terminal tag . So far, it is unclear whether processing of the Spt7 C terminus is a cause or effect of Spt8 loss from the SAGA complex. However, it is clear that a SALSA-like complex and a partial activated state of transcription in cells can be recapitulated in repressing conditions by truncating the Spt7 C terminus. At the molecular level, one obvious possibility is that the C-terminal end of Spt7 may act as part of a docking site for Spt8 to maintain negative regulation of SAGA during gene repression, whereas the rest of the protein, including the histone fold motif, has a more integral function in holding modules of SAGA /SALSA together. A recent study from Wu and Winston characterized Spt7 C-terminal interaction with Spt8 and mapped Spt7's processing site (via an unprocessable, internal-deletion mutant), and these results agree well with ours. Their unprocessable Spt7 construct resulted in reduced amounts of SALSA and little effect on HIS3 /TRP3 activation, interpreted as evidence of a subtle transcriptional role for SALSA. However, their spt7Delta strain also had only minor HIS3 /TRP3 transcriptional effects compared with our observations. Thus, further investigation will be required to reconcile the different results and conclusions of these studies. Recently, a human homolog of Spt7 has been described . The gene product is a component of STAGA, a human counterpart of yeast SAGA /SALSA, but encompasses just the C-terminal third of yeast Spt7, including the histone fold and putative Spt8-interacting region, suggesting that the C-terminal functional regions described above may be conserved throughout the eukaryotes. Although no human homolog of Spt8 (or Spt20) has yet been identified, STAGA's subunit composition and function remain to be fully characterized. Overall, four yeast proteins have been implicated in SAGA-mediated negative regulation of HIS3 ---Spt3, Spt7, Spt8, and TBP (Spt15) ---and mutation of their genes can lead to high-basal effects on transcription (ref. and present study). However, it should be noted that transcription is only partially derepressed by engineered truncation of Spt7 or loss of Spt8, whereas disruption of Spt3, also centrally involved in the inhibitory mechanism, derepresses more fully . Therefore, we hypothesize that Spt3 is likely also to be altered in SALSA (perhaps by covalent modification) to contribute to activating function, although this is not yet apparent in the subunit analysis. It is important to note that the relationship between SALSA and activated transcription has been studied here primarily just in the context of HIS3 and repression /derepression of amino acid biosynthetic genes. So far, only HIS3, and by extension TRP3, have been identified as having Spt7 /Spt8-mediated repression, but potentially, any genes up-regulated by loss of Spt8 or C-terminal truncation of Spt7 may be candidates for this scheme of negative regulation. For example, these may include Swi5-dependent genes such as HO and PCL2, shown by S1 analysis to have heightened expression in spt3Delta or spt8Delta mutants (ref. and D. Stillman, personal communication). In actuality, the expression of many inducible yeast genes depends on Spt /Ada complexes, but different subsets of promoters seem to be regulated in different ways by them, so the question remains as to how these modes are reconciled in the cell. In cases such as GAL1 --10 or PHO5, the complex may act simply as a coactivator, linking an activator with the transcriptional machinery. At GAL1 --10, the HAT activity of Gcn5 is not required, and the Spt3 moiety (inhibitory in the case of HIS3 /TRP3) interacts with TBP to positively, instead of negatively, affect transcription . Thus, SAGA may be the functional form of the complex at some promoters , but at others (such as HIS3) the altered form SALSA is used for activation. The basis for this selectivity is unknown. One possible determinant (like for NC2) is the nature of the promoter: the sequence, position, or environment of the TATA box might dictate a productive or nonproductive interaction between TBP and SAGA. Another possibility is that there is yet another mode or level of regulation that has yet to be described, such as an unknown Spt3 modification, as mentioned above. Alternatively, SALSA may be the active form at many Spt /Ada-complex-regulated promoters, but perhaps certain promoters are not subject to SAGA-mediated repression because they are negatively regulated by other, epistatic means (e.g., the inhibitor Gal80 at GAL1 --10). It may be that many or most SAGA /SALSA-dependent genes (based on known examples such as HIS3, TRP3, GAL1, and PHO5) are inducible, and their standard state, absent of extreme conditions in the cell, is repression (hence the predominance of SAGA over SALSA in rich media). One way this model may be tested is to examine levels of SALSA versus SAGA from cells under inducing conditions other than 3-AT for various known genes (for example, low phosphate media for PHO5 or galactose media for GAL1 --10), to see whether the SAGA /SALSA equilibrium changes in response to which form is needed at these promoters. Although much remains to be learned about transcriptional regulation by coactivator complexes such as SALSA and SAGA, it is clear that the flexibility offered by the multiple levels of their control is an important aspect of gene regulation. Backmatter: Abbreviations : SAGA = Spt-Ada-Gcn5 acetyltransferase SALSA = SAGA altered, Spt8 absent HAT = histone acetyltransferase TBP = TATA-binding protein 3-AT = 3-aminotriazole PMID- 12193647 TI - Using mechanical force to probe the mechanism of pausing and arrest during continuous elongation by Escherichia coli RNA polymerase AB - Escherichia coli RNA polymerase translocates along the DNA discontinuously during the elongation phase of transcription, spending proportionally more time at some template positions, known as pause and arrest sites, than at others. Current models of elongation suggest that the enzyme backtracks at these locations, but the dynamics are unresolved. Here, we study the role of lateral displacement in pausing and arrest by applying force to individually transcribing molecules. We find that an assisting mechanical force does not alter the translocation rate of the enzyme, but does reduce the efficiency of both pausing and arrest. Moreover, arrested molecules cannot be rescued by force, suggesting that arrest occurs by a bipartite mechanism: the enzyme backtracks along the DNA followed by a conformational change of the ternary complex (RNA polymerase, DNA and transcript), which cannot be reversed mechanically. Keywords: Introduction : Escherichia coli RNA polymerase (RNAP) is a highly processive enzyme responsible for the transcription of DNA into RNA. The ternary elongation complex of DNA, RNAP and RNA is extremely stable, with RNAP capable of reaching speeds of 35 nucleotides per second as it translocates along the DNA . Despite this rapid translocation during elongation, RNAP is sensitive to the sequence it transcribes, displaying temporary (pauses) and permanent (arrests) halts to transcription, which are believed to play a role in the regulation of gene expression . Transcriptional pauses and arrests can occur by various mechanisms but share common features. These features include the continued stability of the ternary elongation complex and the displacement of the RNA 3' end from the enzyme active site . Kinetic evidence suggests that both pauses and arrests are kinetically off the main elongation pathway, and that pauses are intermediate between elongation and arrest states : Much of the information about pauses and arrests has come from biochemical footprinting and crosslinking studies on complexes walked to particular template positions. These studies have suggested that core RNA polymerase is, variously, a flexible enzyme, capable of undergoing inchworming motion ; a rigid enzyme, capable of monotonic translocation ; and a "sliding clamp" enzyme, capable of frequent backwards and forwards oscillations . Among these proposed models, the importance given to an RNA:DNA hybrid and to different protein --nucleic acid interactions to explain the stability of the complex varies . Within the sliding clamp model, protein --nucleic acid contacts confer stability to the ternary complex (through a protein clamp enclosing downstream DNA), whereas the RNA:DNA hybrid (of 8 --12 bp) is responsible for accurately positioning the enzyme active site with the 3' OH of nascent RNA . The recently obtained crystal structure of a bacterial RNA polymerase shows evidence of a downstream DNA-binding site, and suggests that an RNA:DNA hybrid of 8 --9 bp is easily accommodated within the enzyme, which is consistent with the sliding clamp model . This model proposes that upon arrest, the entire enzyme backtracks along the DNA template, maintaining both the size of its footprint and the RNA:DNA hybrid, with its backtracking distance constrained by nascent RNA . This severe repositioning of the enzyme upon arrest does not necessitate a flexible enzyme. The postulated backtracking of RNA polymerase at arrest sequences and, to a lesser extent, at pause sites suggests that it might be possible to regulate the degree of transcriptional pausing and arrest by applying mechanical force directly on an RNAP molecule to aid its forward translocation and thus prevent or reduce its backtracking. This type of experiment is possible with single-molecule techniques, which previously have examined transcription by single RNA polymerases subjected to a load opposing the direction of transcription or in the absence of significant external force . In this paper, we use a new "assisting" force geometry to push the enzyme in the downstream (transcriptional) direction. By assisting the enzyme mechanically and by comparing the results with previous results on the same template in the "opposing" force geometry , we seek to determine the role of lateral displacements in the various components of the elongation cycle (RNA synthesis, pausing, arrest). For example, if arrest involves simply a net displacement of the enzyme along the DNA template (with concomitant changes in nucleic acid pairing) without a significant conformational change of the ternary complex, it should be possible to alter both entry into and exit from the backtracked state by applying a mechanical force in the direction opposing backwards displacement. Additionally, it is not known whether backtracking occurs upon arrest during continuous elongation, in the presence of all NTPs, without first artificially stalling transcription. By performing these experiments in the presence of all four NTPs, we can examine the role of mechanical displacement in "natural" arrest states. In the present study, we investigate the effect of mechanical force on pausing and arrest, compare the effect of forces aiding and opposing transcription, and comment on the flexibility of RNAP in the context of the various models of RNA polymerase translocation. Materials and Methods : Previous work in our laboratory studied the effect of an opposing force on pausing and arrest . Our current studies on the effect of an assisting force use the same optical-trapping flow control video microscope and a DNA template containing the same promoter and transcribed sequence. The template is derived from the pPIA2 --6 plasmid by digestions with PspOMI (upstream of the lambda PR promoter) and either StuI or SnaBI to create a blunt downstream end. The upstream end is labeled by filling in the 5' overhanging restriction site with biotinylated nucleotides. Ternary complexes are initially stalled by incubating biotinylated RNAP with the DNA in the presence of a subset of NTPs such that stalling occurs at position +70 . By using an optical trap, we tether a stalled complex between 2-mum streptavidin-coated polystyrene spheres in a flow cell. Then, we turn off the trap and flow in a buffer containing all four NTPs to restart transcription . During transcription, a biotinylated RNAP moves away from the biotinylated upstream end of the DNA and toward its unlabeled downstream end, thus increasing the DNA tether length. By raising the flow speed in the microchamber, we increase the force applied on RNAP relative to the DNA, thus "pushing" the RNAP in the transcribing direction and assisting its forward translocation. In contrast, in the previous opposing-force experiments, the DNA was biotinylated downstream from the promoter and the application of higher forces increased the load opposing enzyme translocation. We obtain the DNA tether length (in bp) from bead-to-bead distance as described . Figure 1 | Experimental setup (not to scale). Experimental setup (not to scale). As RNAP transcribes away from the labeled end of the DNA, the DNA tether between the two beads increases in length. Results and Discussion : Fig. shows representative assisting-force experimental data, illustrating that after the addition of NTPs to the tethered stalled elongation complex, tether length increases with time, which we interpret as transcription. The tether length was never observed to increase or decrease in the absence of NTPs. Fig. also shows temporary halts to transcription, interpreted as pauses. Many molecules also arrest; an enzyme is considered arrested if it does not resume transcription within 10 min of entering a pause. Figure 2 | (A) Change in tethered DNA contour length (calculated from bead-to-bead distance) during transcription by a single RNAP molecule, with pauses (temporary halts to transcription, arrows) and arrest (halt to transcription longer than 10 min; not all of the waiting time is shown here) clearly visible. (A) Change in tethered DNA contour length (calculated from bead-to-bead distance) during transcription by a single RNAP molecule, with pauses (temporary halts to transcription, arrows) and arrest (halt to transcription longer than 10 min; not all of the waiting time is shown here) clearly visible. (B) The transcription rate for the same transcription run, given by the derivative of low-pass filtered length vs. time data. Current models of transcription entail a rate-limiting step that is biochemical (binding or catalysis) rather than mechanical (translocation) at subsaturating NTP concentrations . This assumption is consistent with the results of single-molecule opposing-force experiments, where the transcription rate is force-independent until near the stall force of ~15 --25 pN . We wished to test these models by examining the effect of assisting force on the transcription rate. As can be seen in Fig. , the transcription rate varies along the template and exhibits minima at pause locations. We determined two different measures of transcription rates. The overall transcription rate, which considers only the length transcribed and the time taken to go from start to end, includes pauses. This value for the overall rate is given in Table (voverall) and is strongly force- and NTP-dependent. To determine a transcription rate free of the influence of pausing, we consider only the peak rates of transcription attained between pauses separated by times greater than the resolution of our data-filtration technique (vmax in Table ). These rates are broadly distributed, as is characteristic of single-molecule experiments. The force --velocity relationship obtained from these values for vmax at 200 muM NTPs is shown in Fig. and encompasses both the opposing and assisting force data. This pause-free transcription rate (vmax) is independent of force. It does, however, increase with increasing NTP concentration, which is consistent with a biochemical rate-limiting step in this range of substrate concentrations. Thus, force in either direction does not seem to have a significant effect on translocation along the main transcription pathway. Table 1 | Transcription rates Figure 3 | The pause-free transcription rates (vmax) were determined at different forces for both opposing (negative) and assisting (positive) force at 200 muM NTPs and show no force dependence. The pause-free transcription rates (vmax) were determined at different forces for both opposing (negative) and assisting (positive) force at 200 muM NTPs and show no force dependence. Effect of Force on Pausing. | Although translocation is not the rate-determining step along the elongation pathway, the strong dependence of the overall rate (which includes pauses) on force strongly suggests that mechanical force can act on the off-pathway reaction leading to a pause (kP, Scheme ). Such an effect is consistent with current models for pausing in which a lateral displacement of RNAP relative to the DNA template results in misalignment of the 3' end of RNA from the catalytic center of the enzyme. Thus, transitions into a paused state should be force-dependent if the rate-limiting step to enter a pause involves this lateral movement of polymerase. To determine the effect of mechanical force on pausing, we consider its effect both on the entry into (kP) and exit from a pause (k-P), as can be obtained from the pause density (pauses per kb) and duration. Pauses give rise to minima in the transcription rate; to ensure that these minima correspond to true pauses rather than to artifacts of data sampling and filtering methods, we took only a subset of these minima that met additional criteria, verified by data simulation (see Supporting Text, which is published as supporting information on the PNAS web site, ). Scheme 1 | Two different observations suggest an effect of force on entry into the pause state, kP. First, the mean pause density with assisting force is about two-thirds of that with opposing force . As NTP concentration increases, the pause density with assisting force decreases further, showing that closer to saturation, RNAP pauses less frequently, consistent with results from bulk biochemistry. This decrease in pause density (per unit length) with increasing NTPs is not due solely to a decrease in time spent waiting at each nucleotide position: a time-based pause density (pauses per 1,000 s on the main elongation pathway) also decreases with increasing NTP , suggesting that NTP concentration directly affects kP. Second, the fraction of molecules entering any detectable pauses during transcription decreases by more than a factor of two with an aiding force . Assuming stochastic entry into a pause, a force-independent pausing mechanism would predict a decrease in the percentage of pause-free runs with an increase in length transcribed. Instead, an assisting force increases both the mean transcription length and the fraction of pause-free runs compared with opposing force, providing additional evidence that an assisting force exerts a considerable effect on entry into a pause. As discussed above, force does not affect the kinetics along the main elongation pathway (kN, Scheme ), implying that the reduction in pausing that we observe is not caused by changes in the competing rates of pausing and elongation, but by the direct action of force on the off-pathway transition into the pause state (kP, Scheme ). Table 2 | Pausing statistics The effect of force on exit from a pause to return to the main elongation pathway (k-P) can be studied by examining the pause durations. Again, we have accessed this in two ways. First, we determined the duration of each pause (see Supporting Text), then calculated a mean pause half-life by fitting to first-order decay kinetics. These values for pause half-lives are given in Table and exhibit a slight dependence on force direction. It is difficult, however, to determine precisely the duration of a pause, whereas pause locations are easier to score. Thus, to obtain a more reliable estimate of the pause durations, we used the strongly force-dependent overall transcription rates (voverall) and the force-independent peak transcription rates between pauses (vmax). The time a given polymerase spends actively transcribing a length L is given by L /vmax; the actual time it takes to traverse this length L in an experiment is L /voverall. The difference between the active and overall transcription times gives the total time spent off-pathway in a pause state and reflects both the pause density and pause duration; dividing this total time by the number of pauses gives the mean pause duration. These values for pause durations, listed in Table , appear independent of force. Thus, it seems that the rate of return to the main elongation pathway, k-P, is either very weakly or not at all force-dependent. At first glance, there is an apparent contradiction in the force dependence between entry into and exit from the paused state. Pausing is a reversible process, implying that the forward and reverse reactions traverse the same coordinate. How can traversal of this coordinate in one direction be force-dependent whereas force apparently has little affect on the reverse process? The rate of a force-dependent process scales exponentially with force applied and distance to the transition state . With an asymmetrically located transition state, the effect of force on reaction rate can be orders of magnitude stronger in one direction than in the other (see Supporting Text), easily accounting for the difference in force dependence between kP and k-P. Effect of Force on Arrest. | Although RNAP has been shown to backtrack during arrest in stalled complexes formed by NTP deprivation, there is no evidence for backtracking into arrest during continuous elongation with all NTPs present. Previous single-molecule experiments using the same template as in this study showed that in the presence of an opposing load, all RNAP molecules arrested before arriving to the end of the template (10.7 kb from the transcription start site) at both 200 muM and 1 mM NTPs . Arrest in these experiments is not due to cumulative photodamage of the enzyme, as the optical trap is turned off as soon as a ternary complex is tethered, before resuming transcription. Given the putative large displacement involved in backtracking, arrest of actively transcribing complexes should display significant force dependence, assuming that the rate-limiting step to entry into arrest involves backtracking of the enzyme over this distance. Specifically, mechanical force opposing transcription should enhance the efficiency of backtracking into an arrested state, whereas assisting force should reduce backtracking and, hence, arrest. Indeed, mechanically assisting transcription markedly reduces the incidence of arrest compared with transcription subjected to a load . Even at 200 muM NTPs, fewer than 60% of the molecules arrest under assisting force, and this value decreases to less than 25% as the NTP concentration increases to 1 mM. The arrest density (total arrests per total length transcribed) also exhibits a significant force dependence : the arrest density at 200 muM NTP decreases from 0.87 arrests per kb (opposing force) to 0.30 arrests per kb (assisting force). As mentioned earlier, current kinetic models for transcription elongation suggest that arrests are preceded by pausing intermediates. Given the effect of assisting force in reducing the rate of entry into the pause state, it is possible that the reduction in arrest efficiency with assisting force results simply from reduced pause efficiency and not from a direct effect of mechanical force on entry into arrest (kA, Scheme ). To address this possibility, we computed the effect on pause density (pauses per kb, ~kP) and arrest density (arrests per kb, ~kA) of assisting vs. opposing transcription with force. We found that, at 200 muM NTPs, the pause density with an assisting force is 78% of that with opposing force, whereas the arrest density is decreased much further, to 35% of the opposing force value (Tables and ). The stronger effect of force on arrest density than on pause density indicates that mechanical force acts directly on the rate of entry into the arrested state (kA). The strong effect of mechanical force on arrest efficiency can be rationalized if pushing the enzyme in the direction of transcription tilts the reaction coordinate to energetically disfavor backtracking into the arrested state. Furthermore, this strong force dependence implies that entry into arrest involves a displacement of RNAP, consistent with the backtracked arrest state previously described for stalled complexes. Table 3 | Arrest statistics Having shown that applying an assisting force can prevent the arrest of the molecule during continuous elongation, we asked whether it is possible to rescue mechanically an already arrested ternary complex and return it to the main elongation pathway. For purified ternary complexes under standard biochemical reaction conditions, entry into the arrested state is an irreversible process, as arrested complexes cannot resume transcription without the addition of a transcript cleavage factor such as GreB . This biochemical irreversibility does not imply that the reaction is mechanically irreversible: force could tilt the reaction coordinate, lowering the energetic barrier sufficiently to allow forward displacement of the enzyme. To test this idea, complexes in which elongation had ceased for at least 10 minutes, i.e., arrested by our definition, were subjected to increasing force in the forward direction. These efforts proved largely unsuccessful, with the enzyme resuming transcription in less than 10% of the cases (3 of 41 attempted rescues). Is this inability to rescue arrested complexes due solely to the existence of a large energetic barrier to forward displacement? Forces up to 50 pN were applied to the arrested complexes in the above experiments, which, if acting over a distance of 10 bp (the estimated extent of the backtrack; ref. ), results in ~40 kBT of work. (The maximum of ~50 pN of force available in our instrument is very close to the upper limit of 65 pN that can be applied to double-stranded DNA before it undergoes an overstretching transition to a different structural form; ref. .) If the arrested enzyme is simply mechanically displaced, i.e., if the reaction coordinate to arrest involves only a displacement of the enzyme along the DNA, the failure to rescue arrested complexes implies that an exceptionally high energetic barrier must be overcome to resume elongation, which is consistent with the irreversibility of arrest in bulk assays. However, because RNAP can enter into arrest under thermal conditions, the energetic barrier leading to arrest at those template locations must be on the order of at most a few kBT . Again, assuming that the reaction coordinate for entry into and exit from arrest is the relative position of RNAP along the DNA template, the free energy difference between the active state and the arrested, backtracked state of the enzyme would be a minimum of ~35 kBT. This number is unreasonably large, because it would imply an energetic stabilization upon backtracking by an amount greater than the overall binding energy of RNAP to DNA (~25 kBT; ref. ). This absurd result suggests, instead, that the reaction coordinate for arrest is not simply a displacement of the RNAP along the DNA. Rather, it is likely that a force-independent conformational change of the enzyme or ternary complex occurs after backtracking, and that this conformational change is off the mechanical reaction coordinate, effectively rendering the overall process mechanically irreversible. Figure 4 | (A) Reaction coordinate for polymerase entering arrest by mechanical displacement only. (A) Reaction coordinate for polymerase entering arrest by mechanical displacement only. For illustration purposes, we assume that the barrier to arrest is equal to 5 kBT. (B) Depiction of entry into arrest by backtracking followed by a structural reorganization. What is the nature of this structural reorganization of the ternary complex after backtracking? Zaychikov et al. studied active and arrested ternary complexes and found a significant increase in gel mobility upon arrest, which they interpreted as a change in DNA-bending angle leading to a more extended conformation. Crosslinking studies of arrested and productive elongation complexes have shown a change in interactions between the 3' end of the RNA and polymerase upon arrest and demonstrated that unwound 3'-proximal RNA feeds into the secondary channel of RNAP . It is unlikely that a change in RNA location alone would be sufficient to prevent mechanical repositioning of the polymerase and would account for the large energetic barrier to reactivation of transcriptional elongation. It is more likely that the change in contacts between RNA and the polymerase exerts an allosteric effect on the enzyme, triggering a conformational change away from the active configuration. The rescue of arrested complexes by GreB could be caused as much by the removal of the extruded 3' RNA as by the removal of its allosteric contact with the polymerase. Does the backtracking of RNAP automatically induce this structural reorganization of the ternary complex; i.e., are the two components of arrest tightly coupled? In a tight coupling scheme, translocation of the enzyme is accompanied by the simultaneous structural reorganization of the complex. A number of observations suggest that this is not the case. For example, bulk biochemical studies showed that backtracked enzymes are still capable of resuming active transcription at some locations . Thus, it is plausible that, upon backtracking, the complex is initially in a displaced intermediate state from which it then undergoes a structural reorganization as described above. This last step should be a mechanically irreversible change that prevents mechanical rescue of the complex to the elongation competent form. From the pause durations observed, we conclude that conversion to a transcriptionally inactive form takes place over the course of minutes at room temperature with the NTP concentrations used in this study (200 muM to 1 mM). To test the model of backtracking followed by mechanically irreversible change, we subjected complexes that had ceased transcription for various amounts of time to varying strengths of assisting force. The results of these rescue attempts are shown in Table . Complexes are most likely to be rescued mechanically when they have ceased transcription for less than 5 min. As the delay time is increased before applying greater assisting force on the enzyme, the rescue efficiency decreases, until, finally, for waiting times longer than 12 min, we are unable to rescue any complex mechanically. As controls, the lifetimes of pauses for which we did not increase the force after cessation of transcription were determined. With no increase in initial force, the longest pause observed was 400 s, and the average pause duration was 122 s, whereas with an increase in assisting force, the longest pause rescued was 765 s and the average waiting time before successful rescues was 296 s. Thus, an increase in assisting mechanical force can access and rescue complexes that would otherwise become irreversibly arrested. Most importantly, the ability to rescue an arrested enzyme decreases with waiting time, suggesting that we can mechanically rescue complexes that are paused and possibly backtracked but not those that have undergone an irreversible conformational change into the arrested state . Table 4 | Mechanical rescue of paused molecules Figure 5 | Kinetic scheme for the elongation cycle, with associated force- and NTP-dependences of the rates. Kinetic scheme for the elongation cycle, with associated force- and NTP-dependences of the rates. N-1, N, and N + 1 refer to consecutive steps in the main elongation pathway; P is a pause state, B is backtracked (prearrest), and A is arrest. The proposed changes associated with each step are shown in blue, as discussed in the text. Kinetic Cycle of Transcription Elongation. | The results of this study allow us to determine the roles of mechanical and chemical forces on the various steps of the elongation cycle. Fig. shows the kinetic scheme for RNA synthesis, pausing, and arrest given in our introductory remarks, including the role of force and NTP concentration determined in this study. Consistent with earlier studies, we have found the transcription rate to be force-independent but NTP-dependent. A reduction in NTP concentration leads to more frequent off-pathway excursions, manifest in an increased pause density. This result is consistent with the results of bulk biochemical assays , in which pause efficiencies are enhanced at lower NTP concentrations, and with the scheme of Fig. , in which pausing and elongation compete kinetically. The pause density depends also on the sense (assisting or opposing) of applied force, suggesting that at least a subset of pauses involves displacement of RNAP relative to the DNA template. The stronger effect of force on arrest efficiency lends support to the model of a displaced enzyme in the arrest state and suggests that this displacement is significantly larger than in paused states. The large scale of enzyme backtracking is evidenced by our ability to decrease the incidence of arrest and increase the mean length transcribed by the application of an assisting force and by the fact that even with the smallest of opposing forces (<2 pN), the transcribed length is reduced relative to that with small assisting forces, indicating premature arrest of the polymerase. Interestingly, the process of arrest seems to occur by a bipartite mechanism. First, the enzyme undergoes backtracking into an intermediate state (B, Fig. ), which can be mechanically restored to elongation. This backtracking is followed by the kinetically slower step of a structural reorganization within the ternary elongation complex. Apparently, the conformational transition to the arrested state cannot be affected by force, as we cannot rescue complexes that have been transcriptionally inactive for longer than 12 min. Therefore, it seems that this arrested state can be restored to the active configuration only by biochemical means, after cleavage of downstream RNA by a factor such as GreB . Elongation Models. | The stability of the ternary complex at the end of the template has important implications for models of elongation. At the last position on blunt-ended DNA, there is no interaction between DNA and regions of the polymerase downstream of the active site (the downstream DNA-binding site in the inchworm model, the protein clamp in the sliding clamp model). Thus, the only interactions maintaining the ternary complex are with and within the RNA:DNA hybrid and with upstream DNA- and RNA-binding sites. The majority (84%) of RNAP molecules that reached the blunt downstream end of the DNA in our assisting-force experiments detached from the template. An analysis of the times spent at the end of the template before dissociation gives a half-life of approximately 40 s, a value comparable to the half-life of ternary complexes at the end of a blunt template determined in bulk biochemical experiments . We have found no force-dependent trend to these lifetimes in our study, suggesting that mechanical force exerts no significant effect on the stability of these ternary complexes. Our inability to drastically affect these interactions by the application of an assisting force probably reflects the fact that, in the absence of protein --downstream DNA interactions, lateral forces maintaining the polymerase along DNA are still significant. We have shown that transcriptional arrest can occur during continuous elongation and at NTP concentrations near saturation, even when a mechanical force is acting to prevent backtracking of RNAP. Nonetheless the incidence of arrest is greatly inhibited by an assisting force and by increasing NTP concentrations. Models of a rigid enzyme suggest that entry into the backtracked, arrested state is driven by hybrid energetics. For example, studies that have modified the hydrogen-bonding within the RNA:DNA hybrid affect its stability and the incidence of backtracking . The strong effect of force on arrest suggests that the transition from paused to backtracked complexes involves a large displacement, implying that the enzyme does not populate intermediate states between these two configurations. The experiments outlined in this paper have demonstrated that the application of force can play a unique role in determining the range of polymerase displacement occurring during transcription. That the enzyme can continue not only to grip the DNA but also to transcribe at an NTP-determined rate, even at such high assisting forces as 50 pN, is an impressive feat. It is still not known what the minimal step size of the enzyme is during transcription elongation, which is a crucial detail for discriminating among various models of RNAP translocation. The ability to observe transcription at high forces has direct implications for obtaining high-resolution information on the dynamics of RNA polymerase. More generally, the ability to access a higher range of forces by assisting a motor enzyme may help in gaining more information about the dynamics of other systems. Backmatter: Abbreviation : RNAP = RNA polymerase PMID- 12185245 TI - Loss of microsatellite diversity and low effective population size in an overexploited population of New Zealand snapper (Pagrus auratus) AB - Although the effects of overfishing on species diversity and abundance are well documented, threats to the genetic diversity of marine fish populations have so far been largely neglected. Indeed, there seems to be little cause for concern, as even "collapsed" stocks usually consist of several million individuals, whereas population genetics theory suggests that only very small populations suffer significant loss of genetic diversity. On the other hand, in many marine species the genetically effective population size (Ne), which determines the genetic properties of a population, may be orders of magnitude smaller than the census population size (N). Here, microsatellite analyses of a time series of archived scales demonstrated a significant decline in genetic diversity in a New Zealand snapper population during its exploitation history. Effective population sizes estimated both from the decline in heterozygosity and from temporal fluctuations in allele frequency were five orders of magnitude smaller than census population sizes from fishery data. If such low Ne/N ratios are commonplace in marine species, many exploited marine fish stocks may be in danger of losing genetic variability, potentially resulting in reduced adaptability, population persistence, and productivity. Keywords: Introduction : In 1883, Thomas Huxley, then president of the Royal Society of London, declared that "the cod fishery, the herring fishery and probably all of the great sea fisheries are inexhaustible; that is to say that nothing we do seriously affects the numbers of fish" . Since then, the status of the major fisheries and its perception by fisheries scientists has changed considerably. Several cod and herring fisheries have had high profile "collapses" in recent decades , and in 1997, it was estimated that 60% of the major marine fisheries were either fully exploited or overexploited . Such high levels of exploitation not only affect the abundance of target species but also change the physical and trophic structure of marine ecosystems. Among attempts to ameliorate such immediate far-reaching ecological effects of fishing, there is usually little consideration for more long-term impacts, such as changes in the genetic constitution of exploited species. Although there has been some interest in selective changes in exploited fish populations , their genetic diversity is generally considered to be unaffected by commercial fishing, in an attitude that echoes Huxley's statement of more than a century ago. Indeed, population genetics theory suggests that genetic diversity is significantly reduced only in very small populations , and so even "collapsed" stocks may consist of far too many fish to show declines in genetic diversity measurable with feasible sample sizes . For example, the spawning stock biomass of the Newfoundland cod, whose fishery was so famously closed in 1992, remained at 22,000 t [ref. ; 1 t (tonne) = 1,000 kg], leaving several million fish in the population. Similarly, even after stock crashes and effective cessation of dependent fisheries, the stock biomass of many small pelagic species is generally still several tens of thousands of tonnes, and thus populations typically consist of tens of millions of fish . These population sizes are many orders of magnitude higher than those commonly considered being in danger of losing genetic diversity , and so there appears to be little cause for concern from a genetic perspective. On the other hand, the number of fish in a population (census population size, N) is often much larger than the genetically effective population size (Ne), which determines the genetic properties of a population . The long-term evolutionary Ne is often orders of magnitude smaller than current population sizes, probably because of historic population bottlenecks, "selective sweeps," or colonization histories . However, recent evidence suggests that even the short-term Ne without consideration of such historical events may be very much lower than census population numbers. Especially in marine organisms, high fecundity, a strong bias in reproductive success, large variations in year class strength, and size-dependent fecundity may reduce the effective population size by several orders of magnitude . Millions of individuals may therefore be equivalent to an effective population size of only hundreds or thousands. The notion that collapsed fish stocks may lose genetic diversity is thus not as far-fetched as is often assumed, despite their large spawning stock biomasses. The demonstration of changes in genetic diversity in wild populations is often complicated by the lack of suitable populations for comparison. Most commercially exploited species are fished wherever they occur, and thus comparisons between exploited and unexploited stocks are not possible. Temporal comparisons based on archived material such as scales, otoliths, and bones, on the other hand , are usually restricted by a lack of preexploitation samples, as routine sampling is usually initiated long after the onset of the commercial fishery. Here, we used a collection of scales of two New Zealand snapper (Pagrus auratus) populations, dating back in one population to the beginning of exploitation, to investigate the genetic effects of reductions in stock biomass caused by commercial fishing. The history of the snapper fishery in the north of New Zealand is typical for many other fisheries in the world : yields of the commercial fishery in Hauraki Gulf , which developed in the mid 1800s, increased slowly up to the 1970s, when the introduction of pair trawls raised catches to 12,000 t . By the mid-1980s annual catches had declined to 6,000 t, and stocks showed signs of overfishing. During this period, the spawning stock biomass had decreased from an estimated 280,000 t to 37,000 t, a decline by 87%, although population abundance never fell below 37 million individuals . Because of the importance of the fishery and the drastic decrease in stock abundance, research on snapper biology started soon after World War II, not only in the heavily fished northern populations, but also in hitherto less exploited stocks in the south, such as Tasman Bay in the north of the South Island. There are therefore time series of scale samples from Tasman Bay, beginning in 1950 just after the commencement of the fishery, when spawning stock biomass was essentially at natural levels, and covering the entire exploitation history of the stock, with a reduction in biomass by 85% and in numbers by 75%, and an estimated minimum population size of 3.3 million fish in 1985 (refs. and ; Fig. ; ref. is available at ). There is no, or very little, exchange between the two snapper populations, which thus experienced independent exploitation and demographic histories. Allozyme variation , growth rates , and microsatellite polymorphism show that the Hauraki Gulf population is part of a larger stock along the northeast coast of the North Island, whereas the Tasman Bay fish are differentiated from the Hauraki Gulf stock and appear to be isolated from other populations. A comparison between the two populations in Hauraki Gulf and Tasman Bay therefore allowed an assessment of the genetic effects of different exploitation histories and population sizes. Figure 1 | Map of New Zealand showing the sampling sites . Map of New Zealand showing the sampling sites . Figure 2 | Temporal changes in annual catch (AC; ref. Temporal changes in annual catch (AC; ref. ), spawning stock numbers (N; refs. and ) and biomass (SSB), genetic diversity (mean number of alleles per locus (Na), and mean expected heterozygosity (He); both means +- 95% confidence limits of 30 individuals) in New Zealand snapper (Pagrus auratus). In Hauraki Gulf, estimates of SSB before 1970 are from ref. ; after 1970, from ref. . Materials and Methods : Molecular Work. | Dried scales from 1950 to 1986 were obtained from the Ministry of Fisheries, Wellington, New Zealand, where they had been stored individually at room temperature in paper envelopes. DNA from scales was extracted in a dedicated ancient DNA laboratory by using a previously published protocol . Additional samples were obtained from fresh material collected in 1998, and DNA was extracted by using standard methods. DNA extracts were screened at seven microsatellite loci [Pma1, Pma2, Pma5 ; GA2A, GA2B, GT2, GT4 ] on an ALFexpress automated sequencer (Amersham Pharmacia Biotech). Sample sizes are presented in Table . Table 1 | Measures of genetic variability at microsatellite loci Statistical Analyses. | Multilocus genotypes were tested for deviations from Hardy --Weinberg equilibrium and for linkage disequilibrium by using Fisher's exact test in genepop version 3.2 . Genetic diversity was quantified by using Nei's unbiased heterozygosity (He) and the mean number of alleles per locus (Na). Because the number of alleles strongly depends on the sample size, 30 individuals from each sample were resampled 1,000 times [by using PopTools (add-in for Microsoft Excel, written by Greg Wood, Commonwealth Scientific and Industrial Research Organization, Australia, available at )], so not only standardizing the mean (sampling without replacement), but also providing estimates of the sampling variance (95% confidence limits, obtained with replacement). The statistical significance of temporal trends was tested by calculating slope (b) and Pearson's correlation coefficient (r) of a linear regression of He and Na against year, permuting individuals among samples 1,000 times, and comparing the real parameters with randomized estimates. A linear regression was used as the most parsimonious model because the data suggested a linear relationship and there was no a priori information about changes in Ne as a basis of a more specific nonlinear model. Ne was estimated from both temporal fluctuations in allele frequencies and the decrease in heterozygosity . The small sample of 1980 from Tasman Bay was excluded from the temporal analysis, because the method is sensitive to small sample sizes, and the sample was the only one collected from pair trawls. Calculation of Ne followed sampling plan 2 as the more parsimonious model, because the difference between sampling plans becomes negligible if N is large . Only data of time intervals of 20 years or more were used, because effective population size estimates may be biased in species with overlapping generations if the time interval between sampling events is short . Generation time (G) was estimated following Felsenstein from age frequency data of three consecutive years (1998 --2000) in both Tasman Bay and Hauraki Gulf. Fecundity was estimated from average fish weight in each age class . The second method to estimate Ne used the decrease in heterozygosity . To reduce random noise in the decrease of heterozygosity in a small population , a linear regression between heterozygosity and year was used to estimate effective population size over the entire period. Yearly census sizes of the adult population were obtained from reports of the National Institute of Water and Atmospheric Research in New Zealand . For Ne /N ratios, the harmonic mean of the annual N in specific time periods was used . Results : Deviations from Hardy --Weinberg equilibrium were detected in 6 of 70 tests ; four of these deviations were due to heterozygote deficiencies, and no test was significant after Bonferroni correction. Loci GA2A and GA2B were isolated from the same clone and were significantly linked in 8 of the 10 samples (TB50, TB80, TB86, TB98, HG52, HG72, HG79, HG98) ---therefore, locus GA2A (the locus with fewer alleles and thus with theoretically lower power in detecting reduction in genetic diversity) was excluded a priori from averages. Estimates of genetic diversity and their 95% confidence limits showed a decline of both mean heterozygosity (Ne) and mean number of alleles (Na) in Tasman Bay, whereas there were only random fluctuations of genetic diversity in Hauraki Gulf. Permutation tests showed that this decline of genetic diversity over all loci was significant in Tasman Bay, but not in Hauraki Gulf . Furthermore, linear regressions of individual locus diversity against year show a decrease in genetic diversity of the Tasman Bay population over time in six of the seven loci. In contrast, only one locus decreased in both heterozygosity and the number of alleles in Hauraki Gulf. The effective population size estimated from the linear regression of mean heterozygosity against year in Tasman Bay was 46 individuals. Individual loci showed clear effects of genetic drift, such as fluctuations in allele frequencies and loss of rare alleles . Table 2 | Slopes and Pearson's correlation coefficients of regressions of genetic diversity (number of alleles Na and heterozygosity He) at individual loci, and of the average across loci (excluding GA2A, see text) Figure 3 | Allele frequencies of locus GA2B in Tasman Bay, showing random genetic drift and loss of alleles. Allele frequencies of locus GA2B in Tasman Bay, showing random genetic drift and loss of alleles. The frequency of the most common allele (170) is indicated above the arrow. Temporal changes in allele frequencies in both populations were used to estimate effective population sizes . Mean generation time (G) was estimated at 11.4 years in Hauraki Gulf and 13 years in Tasman Bay, although varying G between 10 and 15 years had only limited effects on the Ne estimate (83 --125% of the point estimate at 13 years in Tasman Bay). In Tasman Bay, the estimated effective population size over the whole time period (1950 --1998), spanning almost four generations, was about 180 individuals, and was significantly different from infinity (95% confidence limits: 80 --720 individuals). With the exception of the first time period, the Ne /N ratio estimated from our genetic data and from fisheries data remained remarkably constant at 1.8 --2.8 x 10-5. In Hauraki Gulf, on the other hand, the estimate over the whole study period (1952 --1998) was not significantly different from infinity, as were all individual estimates with one exception (1972 --1998). Table 3 | Number of generations (G), F values, and estimates of effective (Ne) and harmonic means of census population size (N) in Tasman Bay and Hauraki Gulf Discussion : Our data provide evidence for a loss of genetic diversity in an exploited fish population, despite an estimated minimum census population size of more than 3 million fish. Such a decrease in diversity implies that the effective population size was several orders of magnitude smaller than census numbers. These results suggest that commercial fishing may result not only in selective genetic changes in exploited stocks but also in reduced genetic diversity caused by genetic drift. Subsequently, we will place the evidence into a genetic and demographic context and investigate the wider implications of our study for commercially exploited marine fish species. Loss of Genetic Diversity. | In Tasman Bay, there was a significant decrease in both heterozygosity and the mean number of alleles over the 50 years since the onset of exploitation. There are several possible explanations for such a finding: sampling bias, selection or selective sweeps, immigration, or genetic drift. Sampling bias is unlikely, as the decrease in diversity was observed over five independent sampling occasions and at six of seven loci. It is also unlikely that selective effects occurred at six of seven loci in Tasman Bay. Furthermore, although environmental conditions, in particular temperatures, are unstable and temporally unsuitable for snapper recruitment , growth rates of adult fish are higher than in the north of New Zealand , suggesting a benign environment at least for adult growth. It may be possible that the decrease in diversity was caused by immigration from a population with lower genetic diversity; however, a population survey failed to identify such a low-diversity population in New Zealand . Thus, genetic drift within the Tasman Bay population is the most likely explanation for the loss in diversity, implying a low effective population size and low Ne /N ratio. Genetic diversity was higher in the Tasman Bay population than in Hauraki Gulf in 1950, but lower in 1998. Considering that genetic diversity is usually directly related to population size and that the virgin stock biomass of the Hauraki Gulf population was almost 10 times larger than in Tasman Bay (280,000 vs. 35,000 t), the discrepancy of the 1950 estimates may suggest that genetic diversity had already been lost from the Hauraki Gulf population by 1950. Ryman et al. suggested that large populations decreasing to moderate size might lose more rare alleles than medium-sized populations experiencing a severe bottleneck. The loss of genetic diversity may therefore be particularly pronounced during the initial stages of exploitation, and investigations in the advanced stages of exploitation history may be less likely to detect significant changes in allelic diversity . Effective Population Size. | The effective population size was estimated by using two methods: (i) the decrease in heterozygosity , and (ii) the temporal method . Both methods provided similar estimates of Ne in Tasman Bay between 1950 and 1998 [He: Ne = 46; temporal Ne = 176 (95% confidence limits: 80 --720)]. Furthermore, with the exception of the first time period, the Ne /N ratio in Tasman Bay remained remarkably constant over time, suggesting that Ne is determined by biological factors rather than sampling error or statistical artifacts. The higher Ne /N ratio in the period 1950 --1972 may be because of an overestimate of Ne from the temporal method, as the loss of heterozygosity suggested an Ne of only 35 individuals. However, the general concordance between the estimates of the two methods and the constant relationship to fishery data support the biological reality of Ne estimates. The temporal method assumes discrete generations, as well as samples drawn randomly from the entire generation. In species with overlapping generations and with samples biased toward specific age classes, temporal changes in allele frequencies are dependent not only on Ne but also on genetic differences between cohorts . This downward bias in the Ne estimate becomes smaller with increasing time between sampling occasions, as the contribution of genetic drift to temporal shifts in allele frequencies relative to differences between cohorts increases with the number of generations between sampling occasions . Such a positive correlation between length of time interval and Ne was not apparent in our data from Tasman Bay, as the Ne estimate over the whole time period was similar to the estimates in 1950 --1986 and 1972 --1998, and smaller than in 1950 --1972. This suggests that our estimates of Ne are not overly biased and are likely to be in the correct order of magnitude, despite samples consisting mainly of 3- to 4-year-old fish (P.J.S., unpublished observation). The small effect of biased sampling may be caused by the contribution of many age classes to each cohort in the long-lived snapper, thus minimizing genetic differences among age classes . The temporal method also assumes no immigration, an assumption that is difficult to prove in marine populations. However, there is strong evidence from allozyme and microsatellite data as well as from oceanographic patterns that the Tasman Bay population is isolated from other snapper populations in New Zealand. Furthermore, if the low Ne estimates were caused by immigration, the source population would necessarily have to exhibit considerably lower diversity to produce the observed reduction in genetic diversity in Tasman Bay, and there was no evidence for such a population in a recent population survey . Thus immigration is an unlikely cause of the observed estimates of Ne and genetic diversity. Estimates of Ne were very low in Hauraki Gulf, although not significantly different from infinity, with the exception of the period between 1972 and 1998. Although this was also the time of the lowest spawning stock biomass , the Ne /N ratio was an order of magnitude lower than that in Tasman Bay. As Ne was high (and nonsignificant) over the whole study period (Ne = 1,164) and there was no decrease in genetic diversity , the low Ne value in 1972 --1998 may have been an underestimate because of a statistical artifact, immigration from other populations, or other factors. Collectively, our data strongly suggest an Ne /N ratio of 10-5 in Tasman Bay, which is considerably lower than many previously published empirical and theoretical estimates . However, these published estimates either do not include marine species or fail to consider some of the ecological features of many marine fish. For example, snapper are extremely long lived (up to 50 years) and have indeterminate growth and strong weight-dependent fecundity , and thus a few old fish are likely to contribute disproportionally to overall recruitment. This variation in lifetime reproductive success estimated from age distributions alone leads to a standardized variance in fecundity of 1.9 and an Ne /N of 0.46, whereas variances of more than 1 are usually considered exceptional . Furthermore, older and larger fish produce not only more but also larger eggs , which may improve survival and growth of their offspring , and thus further increase the bias in lifetime reproductive success among individuals. Possibly most importantly, survival of larvae may depend crucially on chance encounters with suitable plankton patches (see match-mismatch hypothesis in ref. ), and much of the early mortality may be family specific and thus decrease Ne . In addition, Tasman Bay is at the southern edge of the species' distribution, and cold conditions during El Nino years often result in complete recruitment failure . Other factors, such as mating behavior and effective sex ratios, may further decrease the Ne /N ratio, although more information on the reproductive ecology of snapper would be required to evaluate the specific biological mechanisms. Discrepancies between effective and census population sizes of two to five orders of magnitude appear to be common in marine species. For example, the female effective population size of red drum (Sciaenops ocellatus) in the Gulf of Mexico, estimated from temporal mitochondrial DNA data, is 4 x 10-3 that of the female census population size . More recent estimates based on microsatellites suggest an Ne /N ratio of 10-4 . Similarly, the Ne /N ratio of vermilion snapper (Rhomboplites aurorubens), an ecologically similar species, has been estimated to be an order of magnitude smaller than the annual commercial catch, and thus the actual number of fish is likely to be several orders of magnitude greater than the effective population size . Low Ne /N ratios thus appear to be characteristic for highly fecund species with high juvenile mortality, and not only are of interest in the conservation of genetic resources but also may provide powerful insights into spawning ecology and recruitment of marine species. Implications. | Is there any adaptive significance to the reduction in genetic diversity demonstrated here? Existing evidence from computer simulations and laboratory studies suggests a complex relationship between molecular and adaptive genetic variability, primarily attributable to the presumed neutrality of molecular markers, contrasting mutation rates, the polygenic control of adaptive traits and the effects of dominance and epistasis . The microsatellite data presented here showed, however, that the effective size of the Tasman Bay snapper population was sufficiently small to cause loss of alleles at neutral and weakly selected genes, which nevertheless may be potentially adaptive in different environments. Such alleles may not contribute significantly to extant quantitative genetic variation, but may be crucial for the long-term adaptive potential of a population under changing environmental conditions. Recent assertions of global warming causing recruitment failure in cold-adapted North Sea cod emphasize the importance of maintaining the adaptive potential of exploited populations. A widely publicized rule of thumb for minimum effective population sizes required to maintain genetic diversity is 50 individuals for short-term conservation of heterozygosity and 500 for more long-term considerations of adaptability . Using the Ne /N estimate of 10-5 in snapper, this translates into census population sizes of 5 million and 50 million individuals, respectively. Therefore, fish stocks of several million individuals may be in danger of losing genetic variability in the long term, and so genetic diversity should become a management consideration in many marine exploited species. Considering that more than half of all marine fisheries are based on fully exploited or overexploited stocks, fishing may have already caused a considerable loss of overall biodiversity. In addition to such genetic concerns, the low Ne /N ratio suggests that only very few fish contribute successfully to the next generation, possibly in part explaining the often poor relationship between the size of the spawning stock and recruitment . The recruitment processes suggested by genetic data also open the exciting possibility of identifying the demographic section of the population most likely to reproduce successfully and to protect these individuals from excessive exploitation. A close collaboration between molecular geneticists and fisheries biologists would be required to carry out research into recruitment processes of marine populations and their possible implications for fisheries management and conservation. Backmatter: PMID- 12186979 TI - Evidence for an ancient selective sweep in the MHC class I gene repertoire of chimpanzees AB - MHC class I molecules play an essential role in the immune defense against intracellular infections. The hallmark of the MHC is its extensive degree of polymorphism at the population level. However, the present comparison of MHC class I gene intron variation revealed that chimpanzees have experienced a severe repertoire reduction at the orthologues of the HLA-A, -B, and -C loci. The loss of variability predates the (sub)speciation of chimpanzees and did not effect other known gene systems. Therefore the selective sweep in the MHC class I gene may have resulted from a widespread viral infection. Based on the present results and the fact that chimpanzees have a natural resistance to the development of AIDS, we hypothesize that the selective sweep was caused by the chimpanzee-derived simian immunodeficiency virus (SIVcpz), the closest relative of HIV-1, or a closely related retrovirus. Hence, the contemporary chimpanzee populations represent the offspring of AIDS-resistant animals, the survivors of a HIV-like pandemic that took place in the distant past. Keywords: Introduction : The MHC present in most vertebrate species studied encodes two clusters of cell surface proteins. In humans these are designated HLA-A, -B, and -C (class I) and -DP, -DQ, and -DR (class II), respectively. The MHC class I and II gene products play a pivotal role in the induction of adaptive immune responses. MHC class I antigens are expressed on virtually all nucleated cells and bind peptides from intracellular origin . Normally they are loaded with self-peptides but in the case of an infection the peptides may originate from viruses (or other intracellular parasites). Cytotoxic T cells (CTL) recognize such MHC class I-peptide complexes as alien, and can trigger the lysis of infected cells. MHC class II molecules, expressed on white blood cells, control antibody production and mediate T-cell help. Apart from self-peptides, they bind peptides usually originating from extracellular pathogens. Polymorphism of the MHC system is mainly confined to the contact residues of the peptide binding site . Different MHC molecules select disparate peptides for T-cell activation, and as a consequence particular MHC molecules /alleles are associated with resistance or susceptibility to different infectious diseases. It has been demonstrated that being heterozygous for particular HLA alleles may represent an advantage . Because of MHC polymorphism, individual variation reduces the chance that one pathogen can sweep through the entire population. In this context, MHC polymorphism could act as an insurance of immunity across a population. Two species of chimpanzees, which shared an ancestor about 2 million years (myr) ago, have been officially recognized: namely, the common chimpanzee (Pan troglodytes, or Patr) and the bonobo (Pan paniscus, Papa) . Based on mitochondrial DNA (mtDNA) variation, common chimpanzees have been divided into at least four subspecies designated P. t. verus (Ptv), P. t. troglodytes (Ptt), P. t. schweinfurthii (Pts), and P. t. vellerosus , although the status of the latter two remains unclear . Humans and chimpanzees display 98.7% similarity at the non-repetitive DNA level and shared an ancestor about 5 --6 myr ago . Chimpanzees display far more variation in their mtDNA than humans do , and similar findings have been reported for other nuclear genes studied . The accepted explanation for these findings is that chimpanzees as a species are older than modern humans and have existed as more subdivided populations, resulting in the accumulation of more variation. In addition, at least one report claims that chimpanzees may have existed at larger effective population sizes than humans . Apart from their younger age as a species, humans appear to have undergone multiple population bottlenecks . The MHC system of chimpanzees probably has a genomic structure similar to that of humans; the MHC class I loci are known as Patr-A, -B, and -C . For each locus, alleles of shared ancestry can be grouped into lineages, which may predate speciation . With regard to the number of MHC class I alleles, chimpanzees seem to display at least as much diversity as humans . However, for the A locus, chimpanzee samples showed only positive typing reactions with particular HLA-A1, -A3, and -A11 alloantisera . Subsequent sequencing studies illustrated that chimpanzees only possess orthologues of the HLA-A1 /A3 /A11 family, whereas alleles grouping into five other HLA-A families appear to be absent . In addition, chimpanzees appear to lack the MHC class II equivalents of the HLA-DRB1*04 and -DRB1*08 lineages . These observations suggest that chimpanzees may have lost certain MHC lineages during evolution. The absence of these particular lineages can be safely considered to represent a loss in chimpanzees rather than a recent gain in humans because of the trans-species mode of evolution of MHC lineages . MHC class I and II sequences encode gene products that were shown to be under frequency-dependent /diversifying selection . This, and the imbalance in sample size between the number of humans and chimpanzees analyzed, can hamper an accurate interpretation of the data with regard to a loss of alleles /lineages because of disease susceptibility. Consequently, we studied the potential influence of negative /purifying selection operating on the MHC class I A, B, and C loci by comparing intron variation in humans and chimpanzees, because introns are known to evolve in a neutral fashion . Materials and Methods : Animals. | The Biomedical Primate Research Centre chimpanzee colony started with 35 founder animals originating from Sierra Leone and belonging to the subspecies P. t. verus (West Africa). Included in the study are five animals of the P. t. troglodytes (Central Africa) and four animals of the P. t. schweinfurthii (East Africa) subspecies. In addition, three animals of the P. t. verus subspecies originating from other colonies were studied. The animals are characterized on the molecular level for MHC class I and II gene polymorphisms . Their offspring have been pedigreed based on segregation of serological specificities (Patr-A and -B) and molecular-defined Patr class II gene polymorphisms. DNA Amplification and Sequencing. | Genomic DNA (gDNA), obtained from Epstein --Barr virus transformed B-cell lines, was used to amplify a +-950-bp fragment [containing complete exon 2 (270 bp), intron 2 (241 bp), and exon 3 (276 bp), and partly intron 1 (+-95 bp) and intron 3 (+-60 bp)], using MHC locus-specific primers: 5AIn /3AIn for the Mhc-A locus, 5BIn /3BIn for the -B locus, and 5CIn /3BCIn for the -C locus . PCR (25 mul) contained gDNA (0.5 mug), 0.2 muM of each primer, 1.5 mM MgCl2 (for the B locus occasionally a concentration of 1.75 mM MgCl2 is necessary), 0.2 mM of each deoxyribonucleoside triphosphate (dNTP), and 0.5 units of Taq polymerase. A total of 33 cycles were run, each cycle consisting of 30 s at 95C, 50 s at 65C, and 30 s at 72C with a final amplification of 8 min at 72C. The PCR reactions were purified using the QIAquick Gel Extraction Kit (Qiagen). Purified PCR products were sequenced directly on the ABI 310 automatic sequencer (Applied Biosystems) by using the above-mentioned locus-specific primers. The products were sequenced from the 5' and 3' ends. Cycle sequencing reactions were carried out with ABI Prism dRhodamine Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems) according to the manufacturer's instructions. At least two independent PCR reactions were performed and /or intron 2 sequences were confirmed by their presence in different MHC typed animals. Phylogenetic Analysis. | The UPGMA phylogenetic tree of intron 2 sequences was constructed by using the pairwise genetic distances method calculated by using the Jukes --Cantor correction for multiple hits and rooted by the midpoint method. This allows translation of genetic distances into divergence time . For the construction, the computer program PAUP* V.4.0B8 for Macintosh was used . The same program was used to define bootstrap values based on 1,000 resamplings. Trees constructed using neighbor joining and maximum likelihood (quartet puzzling) gave very similar topologies in the sense that chimpanzee sequences were only found on relatively few deep clades. Neutrality Test. | Tajima's D test and Fu's Fs test were performed using ARLEQUIN V.2.0, a software for population genetics data analysis (Genetics and Biometry Laboratory, University of Geneva; ). Results and Discussion : Comparative Analysis of MHC Class I Intron 2 Sequence Variation in Humans and Chimpanzees. | Intron 2 (241 bp), situated between the polymorphic exons 2 and 3, is a valuable gene segment for studying the evolution of MHC class I genes. Various HLA and Patr class I alleles, from different lineages, have been analyzed for their neighboring intron 2 sequence. An overview of the number of distinct sequences identified is provided in Table . Table 1 | Number of MHC class I coding alleles and intron 2 sequences in humans and chimpanzees Fig. , summarizing the HLA- and Patr-A, -B, and -C intron 2 sequences, depicts the polymorphic nucleotide positions for the different alleles. For the HLA-A locus most intron 2 sequences appear to be specific for a lineage (Fig. A). The variation in the HLA-A exon 2 and 3 sequences evolved, however, mainly by point mutations and thus reflects diversifying selection in contrast to the apparently neutral evolution operating on the introns. Although to a very minor extent, sharing of identical intron 2 sequences between different HLA-A lineages evidences recombination . Compared with those of humans the chimpanzee A-locus, intron 2 sequences possess fewer unique nucleotide substitutions and are far less heterogeneous (Fig. A). Only ten species-unique nucleotide substitutions are observed in Patr-A versus 23 in the HLA-A intron 2 sequences. Figure 1 | Polymorphic nucleotide positions in HLA and Patr intron 2 sequences. Polymorphic nucleotide positions in HLA and Patr intron 2 sequences. Identity to the consensus sequence (depicted at the top) is indicated by dashes. Substitutions and inserts are depicted by the conventional one-letter code, deletions are marked by "x". For instance, "- /A" indicates that differences in a particular sequence have been reported in the literature. When multiple alleles of a lineage share an intron 2 sequence, only a representative sequence is shown (mentioning only the first two digits in the name). In some cases two allele designations are mentioned when it is not clear which allele couples to this particular intron 2 sequence. The brackets show the division of the alleles into trans-species lineages based on phylogenetic analysis . (A) Mhc-A locus data. (B) Mhc-B locus data. The intron 2 sequences of Patr-B*0101 and -B*0401 are identical to published sequences ChB.Crl2 and ChB.Ch18, respectively . (C) Mhc-C locus data. The figure shows also an intron 2 sequence obtained from a bonobo (Papa-C*0301; ref. ). Table 2 | HLA and Patr class I alleles that share an identical intron 2 sequence The lineage specificity for intron 2 sequences at the HLA-B locus is less strict than for the HLA-A locus (Fig. B). Hence, recombination appears to affect the order of intron --exon sequences, as found for HLA-B exons . Only members of HLA-B*08, -B*27, and -B*35 seem to have lineage unique intron 2 sequences. The Patr-B intron 2 sequences, on the other hand, can be divided into two homogeneous clusters, one of which is characterized by a three-nucleotide deletion (Fig. B). A comparison demonstrates that only ten Patr-B-unique nucleotide substitutions are found, whereas 25 HLA-B-specific substitutions are reported (Fig. B). Twelve unique nucleotide substitutions are left when the HLA-B*7301 sequence is ignored. However, this still indicates a greater diversity in advantage of the human population. The data show that recombination promotes diversification at the Patr-B locus to a lesser extent than in HLA-B . If a selective sweep occurred in chimpanzees, particular Patr-B intron 2 lineages were lost because of the negative selection . Hence, as compared with humans the reservoir of intron 2 sequences in chimpanzees is relatively small and as a consequence recombination should be observed less prominently. Table 3 | Distribution of MHC class I intron 2 lineages and their members in humans (HLA) and chimpanzees (Patr) At HLA-Cw, only a limited set of intron 2 data have been reported (Fig. C). The HLA-Cw*0303 and -Cw*0304 alleles share an identical intron 2 sequence. Furthermore, the intron 2 sequences of HLA-Cw*0701 and -Cw*0702 differ by only one nucleotide, whereas intron 2 of HLA-Cw*0701 also contains an insert of five nucleotides (Fig. C). The present data suggest a situation similar to that observed for HLA-A intron 2 data: namely, that every lineage has its own characteristic intron 2 sequence . The Patr-C intron 2 sequences seem to constitute two main clusters, which allow a further division into four lineages (Fig. C). The intron 2 sequences of Patr-C*0203, -C*1303, -C*02 /*13, and -C*0401 have large parts in common indicating that they arose from one ancestor. The Patr-C*02 /*13 sequences could have resulted from a recombination between the intron 2 sequences of Patr-C*0203, -C*1303, and -C*0401. Twenty unique HLA-Cw substitutions are found in intron 2 versus 16 Patr-C specific nucleotides. The current findings reveal that the HLA-A, -B, and -C intron 2 sequences accumulated or maintained more variation than their chimpanzee counterparts. It is generally accepted that introns of the same gene systems in humans and great apes evolve under identical or nearly identical neutral conditions . This argues against the possibility that humans accumulated more variation than chimpanzees over a relatively short time span. Our observation of low MHC class I intron 2 variation in chimpanzees versus humans sharply contrasts the situation documented for other neutrally evolving genomic segments of these two species . As such it provides evidence that ancestral chimpanzee populations have experienced a selective sweep leading to the loss of particular MHC lineages. The reduced intron 2 variation in chimpanzees versus humans is reflected not only by a limited amount of nucleotide variation but also by a low number of intron 2 sequences. One could argue that our sample size is too small. However, all chimpanzee samples added (8 Ptv, 2 Pts, 1 Ptt) did not result in the detection of novel intron 2 sequences. Moreover, in a comparison between 25 MHC-typed chimpanzees of the Biomedical Primate Research Centre colony and 25 MHC-typed humans of the Dutch population, chi2 statistics shows that the intron 2 variation found in humans is 2.56 times higher [confidence interval (CI) 95% is 0.87 --7.55, P = 0.07] for the A locus and 2.64 times (CI 95% is 1.20 --5.82, P = 0.01) for the B locus. Furthermore, MHC class I intron 3 (586 bp) sequence analysis for particular HLA-Cw and Patr-C alleles indicates that the reduced sequence variation is not intron 2-specific and extends to other gene segments (data not shown). The diversity encountered in the chimpanzee mtDNA and in other genetic systems illustrates, however, that a relatively large population of chimpanzees survived the selective sweep. Phylogenetic Analysis of HLA and Patr Class I Intron 2 Sequences. | Humans and chimpanzees shared a common ancestor ~5 --6 myr ago. The phylogenetic tree illustrates that at least ten Mhc-A (a1-a10), six -B (b1-b6), and nine -C (c1-c9) lineages predate the speciation of humans and chimpanzees . Barring convergent evolution, the existence of alleles of one species with closest relatives in the other species suggests the existence of trans-species lineages (as is illustrated, for instance, by the lineage a2 in Fig. ). Figure 2 | UPGMA phylogenetic tree of human and chimpanzee MHC class I intron 2 sequences. UPGMA phylogenetic tree of human and chimpanzee MHC class I intron 2 sequences. The divergence time was estimated based on a substitution rate for introns of 1.4 x 10-9 per site per year . The bar between 5 and 6 myr highlights the period when humans and chimpanzees shared a common ancestor. The brackets indicate the division in trans-species lineages. The relevant bootstrap values are indicated. The sequences of ChB.Crl1, ChB.Ch39, and ChB.Lena were published . In some cases two allele designations are mentioned when it is not clear which allele couples to the particular intron 2 sequence. B, C, or P indicates that for these alleles differences are reported in the literature. In a test for neutrality (Tajima's D; ref. ), none of the chimpanzee intron 2 sequences considered here showed departures from neutrality as is indicated by D statistics and their P values . We cannot rule out a more recent diversifying selection at the HLA-B locus based on D value, and the high negative value of Fu's Fs test suggests a recent demographic expansion. This should not surprise us, considering that the HLA-B locus is the most polymorphic locus known, with over 400 alleles. Table 4 | Tajima's D values and Fu's Fs values for human and chimpanzee intron 2 sequences For the Mhc-A locus, human alleles represent eight different lineages, whereas chimpanzee alleles represent four lineages . Only the a2 and a5 lineages are shared. These results suggest that chimpanzees have lost the representatives of the a1, a3, a4, a6, a7, and a8 lineages, whereas humans may have lost only the evolutionary equivalents of the a9 and a10 lineages. The alternative interpretation is that modern humans accumulated rapid variation in the MHC class I intron sequences. This reasoning is in conflict with the old age of many human HLA-A intron 2 lineages, as well as with the neutral theory of evolution. At the Patr-A locus, eight distinct intron 2 sequences have been detected clustering into four different trans-species lineages. These intron lineages are coupled to the highly homogeneous Patr-A coding sequences, which all cluster into the HLA-A1 /A3 /A11 family. Only the a5 lineage contains a Patr-A intron linked to the family of HLA-A1 /A3 /A11 exons, whereas other Patr introns cluster into other lineages . This indicates that in the past different chimpanzee introns may have been coupled to exons clustering in different lineages because of recombination. Because of negative selection many exons (except equivalents of the HLA-A1 /A3 /A11 family) and their corresponding introns were lost during chimpanzee evolution. For that reason some of these Patr-A locus introns, which do not show similarities toward HLA-A introns linked to the HLA-A1 /A3 /A11 family, may represent relics from ancient recombination events. At the Mhc-B locus, humans possess alleles from five lineages and chimpanzees only from two lineages, whereas at the Mhc-C locus humans and chimpanzees have alleles from six and four lineages, respectively . Thus, on average, chimpanzees have lost more intron 2 lineages than have humans during ~5 myr of evolution. Moreover, chimpanzees appear to show considerably less intron 2 sequence variation than their human equivalents, as is reflected in the phylogenetic tree by the limited number of distinct clades and /or clades with short branch lengths . These observations are also in agreement with the fact that chimpanzees show less variability at the coding sequences for the classical MHC class I molecules. Time Estimate for a Selective Sweep. | The subspeciation of chimpanzees occurred ~1.5 myr ago . Comparative analyses show that chimpanzee subspecies share identical MHC class I intron 2 alleles, indicating that some of these sequences have been genetically stable over a relatively long time span. Overall, chimpanzee subspecies share MHC class I intron 2 lineages, suggesting that the repertoire reduction took place before the subspeciation. At present, only limited data exist on bonobo MHC class I intron and exon sequences. The consensus view is that bonobos also only have alleles that cluster into the HLA-A1 /A3 /A11 family. This would indicate that the selective sweep also predates the speciation of common chimpanzees and bonobos. The MHC class I intron 2 sequence variation present in chimpanzees appears to be of relatively recent origin. For chimpanzees, the common ancestral intron 2 alleles in the a9, b3, b5, and c1 lineages have ages around 1.5 --3 myr . Taken together, the selective sweep causing the MHC repertoire reduction in chimpanzees must have occurred before the (sub)speciation of chimpanzees. Considering the age of the intron 2 lineages, the selective sweep is dated to have happened ~2 --3 myr ago. The Cause of the MHC Class I Repertoire Reduction: A Hypothesis. | MHC class I molecules play a critical role in the immune defense against intracellular infections caused, for instance, by viruses . Therefore, the MHC class I repertoire reduction in chimpanzees may have resulted from a widespread viral infection in the ancestral populations of the contemporary chimpanzee species. Although the effect is most prominent in the MHC class I region, chimpanzees have apparently lost particular MHC class II lineages as well . This is not surprising, as many viruses have both dominant intra- and extracellular stages of infection and MHC class II-mediated antibody responses are therefore also an important correlate of protection. On the other hand, particular MHC class II lineages may also have been lost because of preferential physical linkage (close proximity on the chromosome) to MHC class I alleles that came under negative selection. The question arises: Which pathogen may be held responsible for the selective sweep in the MHC class I repertoire of chimpanzees? Humans and chimpanzees are the only known species susceptible to infection with pathogens like HIV, HCV, and Plasmodium falciparum. However, both species may show marked differences in pathology after infection . Natural infections with SIVcpz, the closest relative of HIV-1, have been documented in at least six chimpanzees and one free-ranging wild chimpanzee . Furthermore, a recent study estimated the zoonotic event with SIVcpz /HIV-1, which gave rise to the human AIDS epidemic, to have taken place approximately 70 years ago . In this context, the ancestor of SIVcpz /HIV-1 could be considered as a prime candidate for the selective sweep in the MHC class I repertoire of chimpanzees. In the past, worldwide, approximately 150 chimpanzees were infected with various HIV-1 strains, but only one animal was diagnosed with symptoms of AIDS . This particular animal was co-infected with different HIV-1 isolates, and the virus isolated at the time of disease was a recombinant that apparently escaped existing immune responses . The relative resistance of HIV-1-infected chimpanzees to the development of AIDS may be the consequence of an effective immune response controlled, at least in part, by the present set of MHC class I molecules, which are the result of positive selection. We have recently demonstrated that at least some chimpanzee MHC class I-restricted immune responses target conserved epitopes of the HIV-1 virus , resulting in effective control of infection. These Patr alleles are characterized by relatively high frequency numbers . Identical viral epitopes are recognized by human long-term nonprogressors in the context of particular HLA class I molecules associated with resistance . However, humans and chimpanzees recognize such epitopes in the context of MHC class I molecules that group into distinct lineages , thus illustrating that the quality of MHC molecules to bind particular peptides may determine whether an individual is susceptible or resistant to a disease. Some orthologues of the HLA-A1 /A3 /A11 lineage in chimpanzees have HLA-A2 binding motifs . This example illustrates that one has to be careful to extrapolate structural chimpanzee data into consequences for the human situation with regard to the function of MHC class I molecules. For that reason, peptide binding studies have been initiated to determine whether chimpanzee MHC class I molecules indeed preferentially target epitopes mapping to conserved regions of SIVcpz /HIV-1. Santiago et al. suggested that the geographic isolation of P. t. verus predated the infection by the SIVcpz progenitor. The presently described MHC class I repertoire reduction is observed in all chimpanzee subspecies, as well as the resistance to developing AIDS. We therefore put forward the hypothesis that ancestors of today's chimpanzee populations went through a pandemic caused by SIVcpz or a related ancestral retrovirus. As a consequence, contemporary chimpanzee populations have modified MHC repertoires, partially reduced, but able to cope with their natural environment and with retroviral infections such as SIVcpz /HIV-1. The fact that SIVcpz is not readily detectable in wild-ranging chimpanzees may be explained, in part, by an effective cytotoxic T cell (CTL) immune response in resistant animals that eventually contained and controlled virus spread in the population. Backmatter: Abbreviations : myr = million years Patr = Pan troglodytes SIVcpz = chimpanzee-derived simian immunodeficiency virus PMID- 12192093 TI - Functional plasticity of an antigen-specific memory CD4 T cell population AB - The protective nature of memory immune responses is attributed largely to terminally differentiated memory T cells that retain memory of the antigen via the antigen receptor and memory of the effector functions that initially cleared the pathogen. It is not known whether a given population of antigen-specific memory T cells is endowed with functional flexibility to provide protective responses against antigens reencountered in different immunological contexts. Here, we examine functional properties of influenza hemagglutinin (HA)-specific memory CD4 T cells recovered from adoptive hosts that received in vitro-activated HA-specific T cell receptor-transgenic CD4 T cells 2 months to 1 year previously. We demonstrate that this HA-specific memory CD4 T cell population bearing a clonal T cell receptor can produce predominantly T helper 1 or T helper 2 effector cytokines depending on the nature of the recall stimulus. Our findings reveal remarkable functional plasticity within an antigen-specific memory T cell population and have direct implications for modulating memory T cell function in vaccine design and treatments for autoimmune diseases. Keywords: Introduction : The immune response to pathogens previously encountered is more effective than the primary immune response because of an expanded population of antigen-specific "memory" T lymphocytes that efficiently elicit effector functions for antigen clearance. The protective nature of memory responses is attributed largely to terminally differentiated memory T cells that retain memory of the antigen via the antigen receptor and memory of the effector functions that initially cleared the pathogen . It is unknown, however, whether a given memory T cell population can exhibit functional flexibility to potentially provide protective responses against antigens that may be reencountered in an altered immunological context . CD4 T lymphocytes orchestrate an immune response primarily through the types of effector cytokines they produce. After activation, naive CD4 T cells differentiate into T helper types I (Th1) or II (Th2) effector cells producing predominantly IFN-gamma or IL-4 for initiation of inflammatory or humoral responses, respectively . A number of factors such as alterations in antigen dose , antigen affinity for the T cell receptor (TCR) , polarizing cytokines , costimulation , and entry into the cell cycle all have been shown to affect generation of Th1 and Th2 effector cells from naive CD4 T cell precursors. By contrast, little is known concerning regulation of effector cytokine production from memory T cells. Although numerous studies have demonstrated that restimulation of memory CD4 or CD8 T cells with cognate antigen yields the same cytokine profile observed in the primary response , none of these previous studies explored the capacity of the antigen-specific memory T cell population to alter its cytokine profile. Effector T cells generated by in vitro or in vivo activation have been shown to exhibit varying degrees of functional commitment and flexibility. Bulk populations of Th1 and Th2 effector cells generated in vitro under strongly polarizing conditions (exogenous cytokines) maintain their polarity of cytokine production when restimulated with cognate antigen or nonspecifically with phorbol 12-myristate 13-acetate /ionomycin . However, both in vitro- and in vivo-activated effector Th1 and Th2 cells can switch their pattern of cytokine production when restimulated in conditions that drive the opposing polarity . Irreversible commitment has been shown to occur only after repeated antigenic stimulation in vitro or as a result of chronic diseases such as allergy and long-term infections in vivo . Although it has been shown that memory T cells derive from activated /effector cell precursors , it is not known whether the resultant memory T cell population is irreversibly committed for effector cytokine production. In this study, we asked whether a given population of antigen-specific memory T cells could modify their effector response. To address this question, we used an in vivo adoptive transfer system well characterized in this laboratory to generate influenza hemagglutinin (HA)-specific memory T cells bearing a TCR clonotype specific for a single HA peptide /MHC class II complex. We demonstrate that this HA-specific memory T cell population can alter its pattern of cytokine production in response to changes in recall antigen dose and TCR-mediated stimuli. Our results suggest that a proportion of memory CD4 T cells are not terminally differentiated in their ability to produce effector cytokines. Materials and Methods : Mice. | BALB /c mice were obtained from the National Cancer Institute Biological Testing Branch, and HA-TCR-transgenic mice were maintained as heterozygotes in the animal facility at the University of Maryland School of Medicine. RAG2- /- mice on a BALB /c genetic background purchased from Taconic Farms were bred and maintained in the animal facility under sterile and pathogen-free conditions. Antibodies and Reagents. | The following antibodies were purified from culture supernatants from hybridomas maintained in the laboratory: C363.29B (anti-CD3epsilon; ref. ), GK1.5 (anti-CD4; ref. ), anti-CD8 (TIB105, American Type Culture Collection), 212.A1 (anti-I-Ad), anti-Thy1 (TIB238, American Type Culture Collection), anti-Mac-1alpha (TIB128, American Type Culture Collection), and 6.5 (anticlonotype HA-TCR; ref. ) conjugated to biotin (Pierce) according to manufacturer recommendations. The following monoclonal antibodies were purchased from BD PharMingen (San Diego): FITC- and phycoerythrin (PE)-conjugated anti-CD25 (clones 7D4 and PC61, respectively), purified anti-mouse Ly-6G (clone RB6 --8C5), PE-conjugated anti-CD4 (clone GK1.5), purified anti-CD16 /CD32 (clone 2.4G2), FITC- and PE-conjugated anti-IFN-gamma (clone XMG1.2), PE- and allophycocyanin-conjugated anti-IL-4 (clone 11B11), PE- and allophycocyanin-conjugated anti-IL-2 (clone JES6-SH4), and FITC-, PE-, and allophycocyanin-conjugated streptavidin. The HA peptide 110 --119 of the sequence, SFERFEIFPK, was synthesized by the Biopolymer Laboratory, University of Maryland. Cell Purification. | CD4 cells (>90% pure) were isolated from HA-TCR spleen by using immunomagnetic depletion as described . Mitomycin C-treated antigen-presenting cells (APCs) were from BALB /c splenocytes by complement-mediated depletion of T cells as described . In Vitro Generation of Effector Cells. | Effector CD4 T cells were generated from HA-TCR CD4 T cells (1 x 106 cells per ml) by incubating with 5 mug /ml HA peptide in the the presence of APC (3 x 106 cells per ml) in complete Clicks medium (Irvine Scientific) containing 5% FCS (Gemini Biological Products, Calabasas, CA), 50 units/ml penicillin /streptomycin (GIBCOBRL), 2 mM glutamine (GIBCO), 10 mM Hepes (GIBCO), and 50 muM beta-mercaptoethanol in 24-well plates for 3 days at 37C in a 5% CO2 humidified atmosphere as described . The resultant HA-TCR activated /effector CD4 T cells were >95% pure with no residual APC. Proliferation and Cytokine Assays. | CD4 T cells (50,000 per well) and antigen-presenting cells (150,000 per well) were incubated with titrated amounts of HA peptide or anti-CD3epsilon antibody in flat-bottomed 96-well plates in complete Clicks medium. Proliferation was assessed after 72 h by the addition of 1 muCi (1 Ci = 37 GBq) of [3H]thymidine (6.7 Ci/mmol) per well and harvested after 18 h by using a Tomtec 96-well plate harvester (Wallac, Gaithersburg, MD). Radioactivity was quantitated by using a Microbeta Tri-luxe plate scintillation counter (Wallac). The level of IFN-gamma and IL-4 in 48-h supernatants from duplicate cultures was measured by specific ELISA (Endogen, Cambridge, MA) as done previously . ELISA results were analyzed by using MICROPLATE MANAGER software (Bio-Rad). Adoptive Transfers and Cell Purification. | Equal numbers of purified effector CD4 T cells and resting HA-TCR CD4 T cells from naive HA-TCR mice (107 cells per 0.5 ml of PBS) were injected into the tail vein of RAG2- /- mice as done previously . Adoptive transfer recipient mice were killed 2 --12 months posttransfer, and the resultant memory CD4 T cells were isolated by immunomagnetic depletion with anti-Mac-1alpha, anti-FcgammaR, anti-Ly6G, and anti-MHC II antibodies followed by anti-rat IgG-, anti-mouse IgG-, and anti-mouse IgM-coupled magnetic beads as described . Intracellular Cytokine Staining (ICS). | Purified naive or memory HA-TCR CD4 T cells (106 cells per ml) were cultured with HA peptide or anti-CD3 antibody (5 mug /ml) and APC (3 x 106 cells per ml) in 1-ml total volume for 30 --40 h at 37C. Monensin (Golgi Stop, BD PharMingen) was added (4 mul /ml) for an additional 6 h of culture. Cells were harvested, centrifuged through Ficoll, washed with medium to remove dead cells and residual APC, and resuspended in 100 --150-mul staining buffer (PBS /5% FCS /0.05% sodium azide) containing Fc-Block (CD16 /32, PharMingen) followed by surface staining for 6.5 and CD25. For intracellular staining, cells were washed and resuspended in Cytoperm /Cytofix solution (BD PharMingen) for 20 min on ice and subsequently incubated on ice for 30 --60 min with appropriate dilutions of anticytokine or isotype control antibodies in Permwash solution. Cells were washed in Permwash solution before flow-cytometric analysis with the FACScallibur equipped with two lasers for three-color analysis using FITC, PE, and allophycocyanin, and analyzed by using CELLQUEST software (BD, PharMingen). Results : To address the question of functional flexibility of an antigen-specific memory population, we used an adoptive transfer system to generate long-lived memory T cells from TCR-transgenic CD4 T cells bearing a clonotypic TCR (6.5) specific for influenza HA and MHC class II I-Ed (HA-TCR; ref. ). As we previously demonstrated, adoptive transfer of antigen-activated HA-TCR effector CD4 T cells into T cell- and B cell-deficient RAG2- /- mice yields substantial numbers of memory T cells that express the 6.5 TCR, are small in size, lose CD25 expression, and mediate potent recall responses . These HA-TCR memory T cells are similar to those generated in both sublethally irradiated and unmanipulated BALB /c mice (refs. and , and data not shown), yet the use of RAG2- /- hosts has the advantage that all of the T cells recovered after time represent memory cells generated from the input activated /effector T cells. We initially assessed the cytokine profile and activation state of HA-TCR effector CD4 T cells before transfer into adoptive hosts. Effector cells were generated by activating HA-TCR CD4 T cells with HA peptide and splenic APCs for 3 days , in the absence of exogenous IL-2 or polarizing cytokines, by using peptide and APC doses previously determined to optimally stimulate HA-TCR CD4 T cells . The resultant effector T cells were >90% 6.5+, uniformly large in size, and exhibited up-regulation of CD25 (IL-2R) expression when compared with naive 6.5+ precursors (Fig. A; ref. ). Functionally, these HA-TCR effector cells produced high levels of IFN-gamma and low levels of IL-4 as assessed by ELISA (ref. and data not shown). To examine effector cytokine production on the cellular level during effector generation, we used ICS to determine the proportion of activated 6.5+ cells producing IFN-gamma and /or IL-4. Fig. B shows that during the 6-h assay, 6.11% of the activated 6.5+ T cells produced IFN-gamma, 0.84% produced IL-4, and 93% did not produce effector cytokines. These data indicate that the majority of activated HA-TCR CD4 T cells had not fully differentiated to produce effector cytokines in levels sufficient to be detected by ICS. However, the 6.5+ cytokine-producing cells were predominantly Th1-like and produced IFN-gamma with a low proportion of 6.5+ IL-4-producing Th2-like cells. A negligible fraction (<0.2%) of 6.5+ cells were Th0-like, producing both IFN-gamma and IL-4 (Fig. B Lower, FACS plot), consistent with the lack of Th0 generation after antigenic stimulation of CD4 T cells derived from another TCR-transgenic strain . Figure 1 | Phenotype and cytokine profile of HA-TCR effector CD4 T cells. Phenotype and cytokine profile of HA-TCR effector CD4 T cells. (A) CD25 expression profile (Left) and size (forward scatter, Right) of 6.5+ HA-TCR naive and effector CD4 T cells. (B) Production of IFN-gamma and IL-4 during generation of HA-TCR effector CD4 T cells. Intracellular cytokine analysis of IFN-gamma (Upper Left) and IL-4 production (Upper Right) by 6.5+ HA-TCR effector CD4 T cells during their generation in vitro is shown. Monensin was added 2 --3 days after activation of HA-TCR CD4 T cells with HA peptide and APC. IFN-gamma versus IL-4 production is gated on 6.5+ cells (Lower). We transferred similarly activated HA-TCR effector CD4 T cells intravenously into RAG2- /- mice and recovered antigen-specific memory T cells at time points between 8 weeks and 1 year in vivo. We had determined previously that the resultant memory population produced high levels of IFN-gamma and low levels of IL-4 in response to restimulation by antigen . To assess whether the persisting HA-TCR memory CD4 T cells were fixed in their ability to produce high levels of IFN-gamma, we compared the cytokine profile that resulted from activation with HA peptide antigen versus anti-CD3 antibody in the presence of APC (HA /APC versus anti-CD3 /APC). Persisting 6.5+ memory T cells proliferated well to both HA /APC and anti-CD3 /APC (ref. ; see also Fig. ) yet exhibited strikingly different patterns of effector cytokine production . In response to HA /APC, HA-TCR memory T cells persisting 10 weeks to 1 year posttransfer produced predominantly IFN-gamma with low levels of IL-4 (IFN-gamma > IL-4, Fig. ), similar to the pattern of cytokine production by the input HA-TCR effector cells. By contrast, in response to anti-CD3 stimulation, HA-TCR memory cells produced very low levels of IFN-gamma yet substantial levels of IL-4 (IL-4 > IFN-gamma, Fig. ). As expected, CD4 T cells recovered from adoptive hosts that had received naive HA-TCR CD4 T cells 10 weeks or 1 year previously, and freshly isolated naive HA-TCR CD4 T cells proliferated but did not produce significant levels of IFN-gamma or IL-4 in response to either antigen or antibody stimulation . These results indicate that collectively, HA-TCR memory T cells exhibit a Th1-like pattern of cytokine production in response to antigen and a Th2-like pattern in response to anti-CD3 stimulation. Figure 2 | Functional responses of differentially stimulated HA-TCR memory CD4 T cells. Functional responses of differentially stimulated HA-TCR memory CD4 T cells. HA-TCR memory CD4 T cells were recovered 10 weeks and 1 year posttransfer of HA-TCR effector cells into RAG2- /- adoptive hosts. For controls, naive HA-TCR CD4 T cells were transferred into RAG2- /- hosts in parallel. Memory and naive CD4 T cells were restimulated with 1 mug /ml HA peptide (black bars) or 1 mug /ml anti-CD3 antibody (gray bars) in the presence of mitomycin C-treated APC. Freshly purified HA-TCR CD4 T cells (marked Fresh Tg or FTg) were cultured in parallel. IFN-gamma and IL-4 content in 48-h culture supernatants was measured by specific ELISA, and proliferation was assessed by measuring [3H]thymidine incorporation after 72 h (see Materials and Methods). These results are representative of five different experiments. Because anti-CD3 antibody is a potent activator of T cells , the biased production of IL-4 by anti-CD3-stimulated memory T cells could be caused by increased activation strength. We asked whether activating HA-TCR memory CD4 T cells with increased antigen doses would result in a Th2-like pattern of cytokine production, analogous to findings demonstrating that stimulation of naive CD4 T cells with increased antigen dose or affinity favored Th2 effector generation . When stimulated with the optimal dose of 1 mug /ml HA peptide , HA-TCR memory cells produced high levels of IFN-gamma (250 ng /ml) and low levels of IL-4 (<2 ng /ml) ; however, when stimulated with supranormal HA peptide concentrations (100 mug /ml), HA-TCR memory cells produced high levels of IL-4 comparable with amounts produced by anti-CD3 stimulation (17 ng /ml and 15 ng /ml, respectively) (Fig. Upper). High antigen dose also led to greatly increased production of IFN-gamma by HA-TCR memory cells , whereas all doses of anti-CD3 antibody led to biased IL-4 production (data not shown). These data suggest that although memory T cells seem to have varying thresholds for cytokine synthesis, the biased IL-4 production revealed by anti-CD3-mediated stimulation is not caused merely by activation strength. Figure 3 | Cytokine production from HA-TCR memory CD4 T cells in response to increasing antigen doses. Cytokine production from HA-TCR memory CD4 T cells in response to increasing antigen doses. Memory CD4 T cells were isolated from adoptive hosts 6 months posttransfer and activated with the indicated doses of HA peptide or 1 mug /ml anti-CD3 antibody in the presence of APC. IFN-gamma and IL-4 content in culture supernatants was quantitated by specific ELISA as described for Fig. . We therefore asked whether the change in cytokine production by the HA-TCR memory T cell population in response to anti-CD3 was caused by preferential stimulation of a persisting nonclonotype (6.5-) population or a reduced ability of anti-CD3 to activate memory T cells. We thus used ICS to analyze cytokine production on the cellular level in conjunction with TCR and activation markers. Representative ICS results are shown in Fig. , and a summary of ICS results from memory CD4 T cells isolated 2 --5 months posttransfer is shown in Table . Because memory CD4 T cells require restimulation to elicit effector function , we cultured HA-TCR memory T cells before adding a Golgi protein transport inhibitor for intracellular entrapment of cytokines (see Materials and Methods). Figure 4 | ICS analysis of differentially stimulated naive and memory HA-TCR CD4 T cells. ICS analysis of differentially stimulated naive and memory HA-TCR CD4 T cells. HA-TCR memory CD4 T cells isolated from adoptive hosts 3 months previously were reactivated with HA peptide (antigen) or anti-CD3 in the presence of APC followed by incubation with monensin, surface staining for 6.5 and CD25, and intracellular staining for IFN-gamma, IL-4, and IL-2. (A) Production of IFN-gamma from 6.5 TCR-expressing and CD25+ memory T cells gated on large, activated cells. Unactivated T cells showed no IFN-gamma production, and quadrants are designated based on staining with isotype-matched control antibodies for each cytokine. Numbers in quadrants refer to the percentage of total activated cells. (B) Production of IL-4 from 6.5+ and CD25+ HA-TCR memory CD4 T cells. The numbers in parentheses refer to the absolute number of cells in each quadrant. Activation of naive HA-TCR CD4 T cells yielded negligible numbers of IL-4-producing cells (data not shown). (C) IL-4+ and IFN-gamma+ memory T cells shown together. The numbers in parentheses indicates absolute cell number. (D) IL-2 and IFN-gamma production from 6.5+ activated naive and memory CD4 T cells. The numbers in parentheses refer to the absolute numbers of 6.5+ T cells, and data shown are gated on 6.5+ CD4+ T cells. Table 1 | Summary of ICS analysis of memory CD4 T cells isolated from RAG2-/- adoptive hosts that had received HA-TCR 6.5+ effector T cells 2 --5 months previously By ICS analysis, striking differences are seen in the proportion of activated 6.5+ memory CD4 T cells that produce IFN-gamma or IL-4 in response to antigen versus anti-CD3 stimulation in the presence of APC . Stimulation with HA peptide resulted in a much higher proportion of 6.5+ IFN-gamma-producing memory T cells compared with anti-CD3 stimulation (9.04 versus 1.11%, Fig. A), with an average of 8.6-fold more 6.5+ /IFN-gamma+ cells generated by antigen compared with anti-CD3 stimulation (see Table ). Naive 6.5+ T cells, as expected, produced negligible IFN-gamma in response to either stimulus (Fig. A), as did unstimulated memory T cells cultured with APC alone (data not shown). To ensure that these differences in IFN-gamma production were not caused by differences in the extent of activation, we analyzed the activated CD25+ population for IFN-gamma production. As shown in the second row of Fig. A, both antigen and anti-CD3 activated the vast majority of HA-TCR naive and memory CD4 T cells as assessed by CD25 up-regulation. In response to HA peptide, 11% of these activated CD25+ memory cells produced IFN-gamma+; however, in response to anti-CD3 stimulation, only 2% of CD25+ memory cells produced IFN-gamma+ (Fig. A). As expected, CD25+ T cells generated by activation of naive CD4 T cells with either stimulus produced negligible IFN-gamma+ (Fig. A). We also observed differences in the proportion of 6.5+ IL-4-producing cells in response to antigen versus anti-CD3 stimulation. Although a low fraction of 6.5+ IL-4-producing cells were generated in response to antigen (0.45%, Fig. B), a greater proportion of 6.5+ IL-4+ cells (2.28%, Fig. B) were generated in response to anti-CD3 stimulation. This increased IL-4 production derived primarily from 6.5+ memory CD4 T cells, because very few nonclonotype 6.5-IL-4+ T cells were observed (Fig. B). As shown in Table , we consistently observed an increase (average of 2.4-fold) in the proportion of 6.5+ /IL-4+ activated memory T cells produced by anti-CD3 compared with antigen stimulation. Similarly, we found an increase in CD25+ /IL-4+ memory T cells generated in response to anti-CD3 stimulation (3.44% or 270 cells, Fig. B) versus HA stimulation (1.4% or 79 cells, Fig. B). Few dual-producing IFN-gamma+ /IL-4+ memory T cells are observed in response to either stimulus (Fig. C), consistent with enzyme-linked immunospot results demonstrating single cytokine production from polyclonal mouse memory CD4 T cells generated in vivo . The ICS results shwon in Fig. and Table collectively demonstrate that HA-specific memory T cells can produce different patterns of effector cytokines depending on the recall stimulus: In response to antigen, the ratio of HA-specific IFN-gamma versus IL-4-producing memory cells ranged from 5.7 to 20 , indicating a Th1-like profile, and in response to anti-CD3 stimulation, the ratio of IFN-gamma- to IL-4-producing memory T cells was always less than 1.0 , indicating the predominance of IL-4-producing memory cells and a Th2-like profile. Given the functional flexibility of the HA-TCR memory population, we asked whether "nonpolarized" memory T cells producing IL-2 were present within this memory T cell pool. As shown in Fig. D, the proportion of IL-2-producing 6.5+ cells was low for both anti-CD3 and antigen-activated memory T cells (1 --2%). Dual-producing IFN-gamma+ /IL-2+ 6.5+ memory T cells occurred in response to antigen at a low frequency (1.24%) but not in response to anti-CD3 stimulation (Fig. D). Activation of naive CD4 T cells resulted in similar proportions of 6.5+ IL-2+ in response to antigenic or anti-CD3-mediated stimulation (2% for both). These results show that IL-2 producers do not represent a large proportion of the HA-TCR memory T cell pool. Discussion : Our results demonstrate that a memory CD4 T cell population derived from a common activated /effector T cell pool and bearing a clonotype TCR specific for influenza HA can produce either Th1 or Th2 effector cytokines in response to different recall stimuli. In this system, antigen activation led to the predominant production of IFN-gamma, anti-CD3 activation led to the predominant production of IL-4, and high-dose antigen activation stimulated production of both IFN-gamma and IL-4. Our results indicate remarkable heterogeneity and plasticity in cytokine production within a population of influenza-specific memory T cells and suggest that memory T cell function can be modulated in antipathogen immune responses and autoimmune diseases. The overall pattern of effector cytokine production by antigen-activated 6.5+ HA-TCR memory T cells was similar to that of the input activated /effector cells from which they derived, with a higher proportion of IFN-gamma producers than IL-4 producers. This result is consistent with results of others demonstrating that restimulation of memory T cells with cognate antigen leads to the same pattern of cytokine production as the precursor effector cells . In response to anti-CD3 stimulation, known to differ qualitatively from antigen stimulation , there was a dramatic curtailment in IFN-gamma production by activated memory cells, concomitant with an increase in IL-4-producing memory T cells. This biased IL-4 production by anti-CD3-stimulated memory T cells is consistent with findings by Bluestone and coworkers showing that anti-CD3 promotes increased IL-4 production by activated Th2 cells while it triggers inactivation or deletion of activated Th1 cells . A therapeutic potential for using anti-CD3 for in vivo modulation of memory T cell function in diseases is suggested by the recent demonstration that anti-CD3 treatment of patients with autoimmune diabetes [known to involve Th1 effectormemory T cells ] lessened disease severity . When stimulated with high doses of antigen, there was a dramatic increase in both IL-4 and IFN-gamma production, suggesting heterogeneity in activation threshold in this clonotypic 6.5+ memory T cell population. Heterogeneity in cytokine production by polyclonal human memory T cells has been demonstrated in response to different antigen doses , most likely because of memory cells expressing TCR with different antigen affinities. Our findings that high-dose antigen stimulation leads to increased IL-4 production from clonotypic memory T cells are reminiscent of findings that stimulation of TCR-transgenic naive CD4 T cells with supranormal antigen doses leads to increased IL-4 secretion , and suggests that the functional fate of resting memory T cells may be subject to similar influences of antigen dose and signal strength. Our demonstration that the cytokine profile of HA-TCR memory T cells could be altered by the recall stimulus strongly suggests that a proportion of memory cells are uncommitted for cytokine production and therefore not terminally differentiated. Undifferentiated human and mouse memory phenotype cells that produce IL-2 have been identified and can be generated under suboptimal antigen stimulation conditions . Although we did not detect a significant fraction of HA-TCR memory cells solely producing IL-2 (Fig. D), it is likely that the majority of the input activated /effector cells had not fully differentiated to produce effector cytokines despite having acquired an activated cell-surface phenotype. Therefore, we hypothesize that the HA-TCR memory T cell population consists of cells in various stages of functional commitment depending on the differentiation state of the activated /effector T cell from which they derived. For example, the memory 6.5+ cells that produced IFN-gamma in response to both antigen and anti-CD3 stimuli (1 --2%, see Table ) may derive from irreversibly committed effector cells, whereas the activated memory 6.5+ cells that produce IFN-gamma in response to antigen but not anti-CD3 stimulation may derive from uncommitted activated cells. It will be necessary to sort activated T cells expressing different levels of effector cytokines before transfer to precisely determine the origin of these "uncommitted" memory T cells. Mechanisms controlling memory T cell cytokine production are not known, although the molecular basis of Th1 and Th2 effector generation from naive T cells has been elucidated . The transcription factors T-bet and GATA-3 have been found necessary and sufficient for Th1 and Th2 generation, respectively , and transcription of the genes encoding IFN-gamma and IL-4 is marked by reconfiguring of chromatin structure in these gene loci . Assessing the maintenance or loss of these molecular changes during the activated /effector-to-memory transition in vivo is likely to provide insight into mechanisms for the functional commitment of memory T cells. In summary, we have identified both functional heterogeneity and plasticity in a single antigen-specific memory CD4 T cell population. We and others have also identified memory T cell heterogeneity in tissue distribution, homing, and chemokine receptor expression that correlates with functional differences . This memory cell heterogeneity on several levels suggests that the greater effectiveness of the anamnestic response may depend, in part, on the plasticity of the memory T cell population in its ability to home to multiple tissue sites and fine-tune functional responses. A greater understanding of memory T cell complexity can lead to potential therapies for manipulation of the memory immune response in vaccine development and autoimmune, infectious, and malignant diseases. Backmatter: Abbreviations : Th = T helper HA = hemagglutinin TCR = T cell receptor PE = phycoerythrin APC = antigen-presenting cell ICS = intracellular cytokine staining PMID- 12186978 TI - Essential myosin light chain as a target for caspase-3 in failing myocardium AB - Programmed cell death involves the activation of caspase proteases that can mediate the cleavage of vital cytoskeletal proteins. We have recently reported that, in failing cardiac myocytes, caspase-3 activation is associated with a reduction in contractile performance. In this study we used a modified yeast two-hybrid system to screen for caspase-3 interacting proteins of the cardiac cytoskeleton. We identified ventricular essential myosin light chain (vMLC1) as a target for caspase-3. By sequencing and site-directed mutagenesis, a noncanonical cleavage site for caspase-3 was mapped to the C-terminal DFVE135G motif. We demonstrated that vMLC1 cleavage in failing myocardium in vivo is associated with a morphological disruption of the organized vMLC1 staining of sarcomeres, and with a reduction in myocyte contractile performance. Adenoviral gene transfer of the caspase inhibitor p35 in vivo prevented caspase-3 activation and vMLC1 cleavage, with positive impact on contractility. These data suggest that direct cleavage of vMLC1 by activated caspase-3 may contribute to depression of myocyte function by altering cross-bridge interaction between myosin and actin molecules. Therefore, activation of apoptotic pathways in the heart may lead to contractile dysfunction before cell death. Keywords: Introduction : Heart failure is a leading cause of mortality that ensues following the chronic activation of biomechanical stress pathways, resulting from various forms of myocardial injury . Histological evidence of apoptosis has been identified in several cardiovascular disorders leading to congestive heart failure (CHF) . Myocardial apoptosis represents a highly complex cell death program, whose execution is regulated by the caspase family of cysteine proteases. Caspase-3 is a key effector enzyme and cleaves downstream critical cellular targets involved in chromatin condensation, DNA fragmentation, and cytoskeletal destruction, thereby expressing the dramatic morphological changes of apoptosis . Caspase-3 activation has been documented in the myocardium of end-stage heart failure patients , and caspase-3 expression is increased in patients with right ventricular dysplasia, a disease associated with progressive cell loss and sudden death . Recently, myocyte apoptosis, assessed by different biochemical hallmarks, including caspase-3 activity, has been described in pacing-induced heart failure models in animals, and correlates with the time-dependent deterioration of cardiac function . Moreover, we showed that caspase-3 activation directly influences contractile performance of failing ventricular myocytes, and can be corrected via adenovirus-mediated gene delivery of the potent caspase inhibitor p35 with a positive impact on contractility . The molecular mechanism by which activated caspase-3 causes a deterioration of cardiac function has not yet been established. In an attempt to answer this question, we performed a screening for caspase-3-interacting proteins expressed in the heart, using a modified yeast two-hybrid system. We identified vMLC1 (ventricular essential myosin light chain) as a target for caspase-3, and investigated whether a correlation between caspase-3 activation, vMLC1 cleavage, and contractile performance exists in failing myocytes. Materials and Methods : Yeast Two-Hybrid Screening. | Yeast two-hybrid screening using pBTM-casp3-p12p17m as bait vector was performed with a human heart cDNA library fused to the Gal4 activation domain in the pACT2 plasmid (CLONTECH), following the Hybrid Hunter two-hybrid system protocol (Invitrogen) in L40 yeast cells (MATa trp1 lue2 his3 ade2 LYS2::4lexAop-HIS3 URA3::8lexAop-lacZ). A total of 30 x 106 independent clones were screened by selective growth on Trp- /Leu- /His- /Ura- /Lys- /Ade+ synthetic dropout medium plates and expression of beta-galactosidase activity. In Vitro Cleavage of Positive Clone Products by Recombinant Caspase-3. | To construct expression plasmids for positive clones obtained from the two-hybrid screening, EcoRI --XhoI fragments of positive clones were inserted into the EcoRI --XhoI cloning sites of pYES2 /NT-A plasmid (Invitrogen), in which the sequences were under control of the T7 promoter. Biotinylated lysine-labeled proteins were prepared from expression plasmids by using a TNT T7 Quick Coupled Transcription /Translation System (Promega), according to the manufacturer's instructions. Five microliters of biotinylated lysine-labeled protein were incubated for 1 hr at 37C with 15 ng /mul recombinant active caspase-3 (BD PharMingen) and optionally with 25 muM caspase-3 inhibitor DEVD-fmk (fluoromethyl ketone) in a TrisCl reaction buffer, pH 7.5 (6 mM TrisCl, pH 7.5 /1.2 mM CaCl2 /5 mM DTT /1.5 mM MgCl2 /1 mM KCl). Samples were size fractionated by SDS /15% PAGE and blotted to a nitrocellulose membrane. Colorimetric detection of biotinylated products was performed on blots with Transcend Colorimetric Translation Detection System (Promega). Site-Directed Mutagenesis and Expression of Mutant Proteins in COS-7 Cells. | The coding sequence of human wild-type vMLC1 was cloned into the KpnI --XbaI sites of the gfp-expressing pAdTrack vector, between a non-tissue-specific cytomegalovirus (CMV) promoter and a SV40 polyadenylation signal. Site-directed mutagenesis of the Asp at position 132 to Ala (D132A) and of the Glu at position 135 to Ser (E135S) was performed with the QuikChange XL Site-Directed Mutagenesis kit (Stratagene), using the pAdTrack-vMLC1 plasmid and two complementary oligonucleotides containing the desired mutation. Mutated clones were identified by diagnostic StuI and BsrBI digestion (for D132A and E135S, respectively) and verified by sequencing. For transient expression of wild-type vMLC1 and its mutants vMLC1mD132A and vMLC1mE135S, COS-7 cells were transfected with pAdTrack-vMLC1 or pAdTrack-vMLC1m plasmids by polycationic SuperFect Transfection Reagent (Qiagen, Valencia, CA). Forty-eight hours after transfection, cell protein extracts were prepared and analyzed for caspase-3 cleavage by Western blot. Antibodies and Western Blot Analysis. | Protein extracts from rabbit ventricle or from COS-7 cells were prepared by homogenization in TrisCl reaction buffer, pH 7.5. To examine the cleavage by caspase-3, 150 mug of protein were incubated for 1 hr at 37C with different amounts of recombinant human caspase-3, in the presence or absence of the caspase-3 inhibitor DEVD-fmk (25 muM). After size fractionation by SDS /15% PAGE, proteins were electrophoretically transferred to a nitrocellulose membrane and blots were incubated for 1 hr with mouse monoclonal antibodies against vMLC1 (0.2 mug /ml, clone 2c8, BiosPacific, Emeryville, CA; 1:10 dilution, clones F109.16A12 and F109.17A5, Biocytex, Marseilles, France). Bound antibodies were detected with horseradish peroxidase-conjugated antibody against mouse IgG (1:10,000 dilution, Sigma) and visualized by chemiluminescence. Obtained chemiluminograms were evaluated by densitometric analysis using scion image (Scion, Frederick, MD). Determination of Caspase-3 Specificity for DFVE. | Ac-DFVE-AMC and Ac-DEVD-AMC were synthesized by Biosyntan (Berlin) with a purity of 94%. Active recombinant human caspase-3 (BD PharMingen) was added at a final concentration of 3 nM to a reaction mixture containing 20 mM Hepes, 100 mM NaCl, 10 mM DTT, 1 mM EDTA, 0.1% CHAPS, 10% sucrose (pH 7.2), and both substrates in concentrations ranging from 0 to 400 muM. After a 10-min preincubation at 37C, the released fluorogenic AMC was monitored every second minute for 20 min at 460 nm with 380 nm excitation. Km and kcat values were determined from plots of activity versus substrate concentration. Absolute kcat values were calculated using a standard curve determined with AMC. Construction and Purification of Recombinant Adenovirus. | Recombinant (E1- and E3-deficient) adenoviruses (Serotype 5) carrying the green fluorescence protein gfp (Adgfp) or both gfp and the baculoviral apoptotic suppressor p35 (Adp35) were generated as described . Pacing-Induced Heart Failure and Transcoronary Gene Delivery. | Pacemakers were implanted into New Zealand White rabbits, and animals were paced at 340 beats per min for 2 weeks, as described . Transcoronary gene delivery of Adgfp or Adp35 (5 x 1010 plaque-forming units) into rabbit myocardium was performed at the time ventricular pacing was initiated, as reported . The project was approved by the institutional ethics review board. Immunoprecipitation of vMLC1. | Left ventricle lysates from control and 15 days paced rabbit myocardium were prepared by homogenization in denaturing lysis buffer (50 mM TrisCl, pH 7.4 /5 mM EDTA /1% SDS /10 mM DTT /1 mM PMSF /2 mug/ml leupeptin /15 units/ml DNase I), heated at 95C for 5 min, diluted 1:10 with nondenaturing lysis buffer (50 mM TrisCl, pH 7.4 /300 mM NaCl /5 mM EDTA /1% Triton X-100 /10 mM iodoacetamide /1 mM PMSF /2 mug/ml leupeptin /0.02% sodium azide), and centrifuged at 15,000 x g for 10 min at 4C. After dilution to 3.5 mg of protein per ml, supernatants were precleared with excess of protein G-Sepharose beads (Sigma) and incubated for 2 hr at 4C with protein G-Sepharose beads (30 mul of beads per ml lysate), preconjugated with 100 mug anti-vMLC1 monoclonal antibody (clone 2c8). Beads containing the immunocomplex were subjected to SDS /15% PAGE and immunoblotting for vMLC1. Preparation and Culture of Adult Rabbit Ventricular Myocytes. | Single myocytes were isolated from the left ventricle of control and 15 days paced failing rabbits, and cultured in modified M199 medium on laminin-precoated glass slides, as described . Two hours after plating, cells were subjected to detection of activated caspase-3. Activated Caspase-3 Detection and Fluorescence Staining. | Activated caspase-3 was detected in living cells by using CaspaTag Caspase-3 Activity Kit (Intergen, Oxford), according to the manufacturer's instructions. Freshly isolated ventricular myocytes were incubated at 37C (5% CO2) with FAM-DEVD-fmk or SR-DEVD-fmk, carboxyfluorescein- or sulforhodamine-labeled fluoromethyl ketone tetrapeptide inhibitor of caspase-3. After 1 hr incubation, cells were washed, fixed in 4% paraformaldehyde, permeabilized in 100% methanol (at -20C), and subjected to Hoechst 33258 staining and either to immunostaining for vMLC1 /vMLC2 (ventricular regulatory myosin light chain), or to phalloidin staining. vMLCs were detected with mouse monoclonal antibodies anti-vMLC1 (4 mug /ml, clone 2c8) and anti-vMLC2 (1:2 dilution, clone F109.3E1, Biocytex), followed by incubation with Texas red or Cascade blue goat anti-mouse-IgG conjugate (10 mug /ml, Molecular Probes). Polymerized actin fibers were visualized by Texas red-phalloidin (3 units /ml, Molecular Probes). Cell Shortening Experiments. | Fractional shortening was measured in rabbit adult cardiomyocytes isolated from left ventricle of control and 15 days paced failing myocardium, after detection of activated caspase-3. Experiments were performed in a temperature-controlled cuvette (37C), using an electro-optical monitoring system (Scientific Instruments, Heidelberg), as described . Statistical Analysis. | Data represent mean +- SEM and were analyzed by one-way analysis of variance, followed by Scheffe post hoc analysis. Statistical significance was accepted at the level of P < 0.05. Results : Identification of vMLC1 as Substrate for Caspase-3. | In a rabbit model of CHF obtained by rapid ventricular pacing, we previously demonstrated that caspase-3 activation is associated with a reduction in contractile force of failing myocytes. Using in vivo transcoronary adenovirus-mediated gene delivery of the potent caspase inhibitor p35, we could correct caspase-3 activation in failing myocardium with a positive impact on sarcomeric organization and contractile performance . To better understand the mechanism that may cause caspase-mediated sarcomeric disarray, we performed in this study a screening for caspase-3-interacting proteins expressed in the heart. We used a modified yeast two-hybrid system with, as bait vector, the plasmid pBTM-casp3-p12p17m, which has already been successfully used to identify gelsolin as a substrate for caspase-3 . Both large (p17) and small (p12) subunits of active caspase-3 were separately expressed in yeast at equimolar ratios under ADH1 promoters. The small subunit was fused to the LexA DNA-binding domain, and a point mutation in the active site of the enzyme (Cys-163 -> Ser) prevented proteolytic cleavage of interacting substrates. The bait plasmid was cotransfected into yeast with a human heart cDNA expression library fused to the Gal-4 activation domain. By screening 30 million transformants, we obtained 125 positive clones that were divided into 22 groups on the base of inserted fragment size and restriction enzyme digestion pattern. DNA sequencing analysis showed that six of the positive candidates encoded overlapping C-terminal parts (clones 7, 12, and 20) or the complete sequence (clones 3, 9, and 17) of vMLC1. MLC1 is one of the six polypeptide chains of the myosin molecule, and is proposed to function as an actin /myosin tether regulating cross-bridge cycling events . In this study we further analyzed the vMLC1 clones, and the others will be described elsewhere. To examine cleavage of vMLC1 by caspase-3 in vitro, proteins encoded by the cDNAs were produced by in vitro transcription /translation-reaction. As shown in Fig. A, clones 3, 9, and 17, which contained the complete sequence of human vMLC1, were cleaved by human recombinant active caspase-3, and this cleavage was blocked in the presence of its tetrapeptide inhibitor DEVD-fmk, suggesting that vMLC1 is a possible substrate for caspase-3. Immunoblot analysis of protein extracts from left ventricle, incubated with active caspase-3, confirmed this result (Fig. B). An ~20-kDa cleavage product for vMLC1 was already evident with 5 ng /mul active caspase-3. Indeed, other structurally related sarcomeric proteins, ventricular vMLC2 or beta-myosin heavy chain, were not cleaved, demonstrating that cleavage of vMLC1 was not due to a generalized degradation of proteins (data not shown). Figure 1 | In vitro cleavage of vMLC1 by recombinant active caspase-3. In vitro cleavage of vMLC1 by recombinant active caspase-3. (A) SDS /15% PAGE of biotinylated lysine-labeled proteins from three positive clones encoding human vMLC1 after 1 hr incubation with 15 ng /mul human recombinant active caspase-3 in presence or absence of 25 muM DEVD-fmk. (B) Immunoblot analysis of native vMLC1 cleavage in protein extracts from rabbit left ventricle, incubated with indicated concentration of recombinant active caspase-3 for 1 hr. DEVD-fmk was applied in a concentration of 25 muM. Mapping of the Caspase-3 Cleavage Site in vMLC1. | To determine caspase-3 cleavage site of vMLC1, purified human vMLC1 was incubated with recombinant active enzyme (Fig. A). Cleavage of purified vMLC1 resulted in two fragments at ~20 kDa and ~5 kDa. Edman sequence analysis of the cleavage products revealed that caspase-3 cleaved vMLC1 at E135 of the C-terminal motif DFVE135G, which is highly conserved. This result was confirmed by immunoblot analysis, using a monoclonal antibody for vMLC1 (clone F109.16A12) directed against the sequence V134EGLRV139 at the caspase-3 cleavage site. The antibody detected the intact vMLC1 protein but did not detect either of the two cleavage fragments (data not shown). The mapped cleavage site corresponds to the caspase-3 consensus sequence DXXD , with the exception of substituting the last Asp residue for the similar acidic Glu residue at position 135. Caspase-3 is considered to have an almost absolute requirement for an Asp in the P1 and P4 positions . To confirm DFVE135 as the caspase-3 cleavage site of vMLC1, we constructed mutant human vMLC1s replacing the Asp in position 132 with Ala (vMLC1mD132A) and the Glu in position 135 with Ser (vMLC1mE135S) by site-directed mutagenesis. As shown in Fig. B, wild-type vMLC1 was cleaved by caspase-3 in vitro. In contrast, both mutant proteins exhibited resistance to cleavage. This finding points to the caspase-3 cleavage site of vMLC1 as being C-terminal of E135 (Fig. C). Under the same conditions, other executioner caspases, such as caspase-6, -7, and -2, failed to cleave vMLC1 in vitro, suggesting that the protein is a specific substrate for caspase-3 (Fig. D). Figure 2 | Determination of the cleavage site of vMLC1 and its specificity for caspase-3. Determination of the cleavage site of vMLC1 and its specificity for caspase-3. (A) Cleavage of purified human vMLC1 (10 mug) by recombinant active caspase-3 (20 ng /mul, 1 hr, 37C), analyzed by SDS /16.5% PAGE and Coomassie blue staining. (B) Immunoblot analysis of wild-type vMLC1, vMLC1mD132A, and vMLC1mE135S proteins expressed in COS-7 cells, after incubation of the cell extracts with recombinant active caspase-3 (20 ng /mul, 1 hr, 37C). The first line (c) shows extracts from COS-7 cells transfected with control vector plasmid. Mouse monoclonal antibody directed against the residues 1 to 8 (clone F109.17A5) was used for immunodetection. (C) A diagram showing the caspase-3 cleavage site at the C-terminal side of E135 of human, rabbit, rat, and chicken vMLC1. (D) Immunoblot analysis of vMLC1 cleavage in protein extracts from rabbit left ventricle, incubated with different human recombinant active caspases (25 ng /mul) for 1 hr. Because no other known targets for caspase-3 contain the cleavage site DFVE, we characterized the kinetic constants for the DFVE and the classical DEVD substrates in a fluorimetric in vitro assay. Caspase-3 cleaved efficiently both Ac-DFVE-AMC and Ac-DEVD-AMC with a reproducible Michaelis --Menten kinetic. Calculated Km, kcat, and kcat /Km values, which indicate affinity, turnover, and specificity, respectively, are presented in Table . Other caspases did not cleave the Ac-DFVE-AMC substrate (data not shown). Table 1 | Kinetic constants for Ac-DFVE-AMC and Ac-DEVD-AMC cleavage by caspase-3 Cleavage of vMLC1 by Activated Caspase-3 in Vivo. | To determine the functional relevance of vMLC1 cleavage by caspase-3 in the heart in vivo, we investigated the evidence of vMLC1 cleavage products in extracts from rabbit failing ventricular myocardium, where we have previously documented a ~6-fold increase in caspase-3 activity . As shown in Fig. , the intact vMLC1 protein of ~27 kDa was relatively stable in healthy control hearts. In contrast, a main ~20-kDa fragment, corresponding to the N-terminal cleavage product, was present in failing myocardium (% of total vMLC1: 22 +- 5%). Somatic gene delivery of the irreversible caspase inhibitor p35 into failing myocardium reduced vMLC1 cleavage of almost 30% compared with control adenoviral infection, indicating that vMLC1 disruption in failing myocytes is directly mediated by caspase-3 . The lack of a complete abolition in vMLC1 cleavage by p35 infection can be due to the transcoronary gene transfer protocol used, which affects only a third of the whole left ventricular myocardium . Figure 3 | In vivo cleavage of vMLC1 in failing myocardium and its reduction by p35 overexpression. In vivo cleavage of vMLC1 in failing myocardium and its reduction by p35 overexpression. Immunoblot analysis after immunoprecipitation of native vMLC1 cleavage products in extracts of left ventricle from control healthy (c) and failing (CHF) rabbit myocardium, after adenoviral gene delivery of gfp or p35. Shown are representative data from one of three animals in each group. Bar graph represents mean +- SEM of the percentage of the ~20-kDa N-terminal fragment. n = 3. *, P < 0.01; **, P < 0.05 (vs. control and CHF+Adgfp, respectively). Disruption of vMLC1, Sarcomeric Integrity, and Contractile Performance in Failing Myocytes. | Myosin is the major component of the thick filaments of sarcomeres, and consists of two heavy chains (alpha and beta), each associated with two types of light chains, the essential (MLC1) and the regulatory (MLC2). X-ray crystallographic analyses demonstrated that essential and regulatory myosin light chains are spatially close, and are both associated with the neck region of the myosin heavy chain globular head . To examine whether in failing myocytes a morphological disruption of the organized vMLC1 staining of A-bands in sarcomeres occurred, and whether it correlated with caspase-3 activation, single cardiomyocytes from control and CHF hearts were isolated. Fig. A shows confocal laser scanning microscopy of isolated ventricular myocytes after staining for activated caspase-3 and immunostaining for vMLC1 or vMLC2. In cardiomyocytes isolated from control hearts, there was no evidence of caspase-3 activation, and both myosin light chains appeared organized in the sarcomeric units (Fig. A a --b and e --f). In contrast, failing myocytes with activated caspase-3 presented a loss of the characteristic localization of vMLC1 in sarcomeres, and the A-band vMLC2 staining, which was maintained, showed a reduced sarcomeric organization compared with that of control cells (Fig. A c --d and g --h). Sarcomeric disarray in failing cells presenting caspase-3 activation was confirmed by phalloidin staining, which visualizes actin filaments (Fig. A k --l). Figure 4 | In vivo cleavage of vMLC1 in failing cardiomyocytes is prevented by p35 overexpression with positive impact on myocyte contractile performance. In vivo cleavage of vMLC1 in failing cardiomyocytes is prevented by p35 overexpression with positive impact on myocyte contractile performance. (A) Laser scanning fluorescence microscopy of representative ventricular myocytes isolated from the anterolateral wall of control and failing myocardium. Green fluorescence (525 nm) shows activated caspase-3 identified by FAM-DEVD-fmk, blue fluorescence (480 nm) illustrates nuclei by Hoechst 33258, and red fluorescence (620 nm) reflects vMLC1 (a --d) or vMLC2 (e --h), or polymeric actin by phalloidin staining (i --l). (Scale bars = 5 mum.) One hundred cells isolated from three animals were analyzed in each group. (B) Representative ventricular myocytes isolated from failing myocardium after in vivo infection with Adp35 and visualized by laser scanning fluorescence microscopy. Green fluorescence (503 nm) identifies gfp and shows Adp35-infected cells, red fluorescence (580 nm) reflects activated caspase-3 labeled by SR-DEVD-fmk, and blue fluorescence (423 nm) illustrates vMLC1. (Scale bars = 5 mum.) Bar graphs represent the percentage of cells with organized (white pattern) or destroyed (black pattern) vMLC1 staining in the p35-infected and uninfected groups. One hundred fifty myocytes isolated from three animals were analyzed in each group. (C) Contraction amplitude under basal conditions and isoproterenol stimulation (10-8 M) measured in single left ventricle myocytes. White columns, cells from control myocardium; gray columns, cells from failing myocardium negative for activated caspase-3; black columns, cells from failing myocardium positive for activated caspase-3; hatched columns, cells from failing myocardium expressing p35. Data are expressed as mean +- SEM. n = 40 cells from three animals in each group. *, P < 0.005; **, P < 0.001 [in comparison with control or caspase-3-positive failing cells (basal, isoproterenol 10-8 M)]. We have recently demonstrated that adenoviral overexpression of p35 in vivo preserves sarcomeric structure and reconstitutes contractile performance of failing cardiomyocytes . Fig. B illustrates laser scanning fluorescence images of representative single ventricular myocytes isolated from failing myocardium infected with bicistronic Ad-p35 (5 x 1010 plaque-forming units) and stained for active caspase-3 and vMLC1. The typical vMLC1 staining was markedly destroyed in almost 80% of the uninfected failing cells, which presented activated caspase-3 (Fig. B d --f, and adjacent bar graph). On the contrary, in failing myocytes expressing the caspase inhibitor there was no evidence of caspase-3 activation and in 89% of the analyzed cells vMLC1 was localized in the A-bands of the sarcomeric units (Fig. B a --c, and adjacent bar graph). Single-cell shortening experiments in failing cardiomyocytes showed a reduction of basal and isoproterenol-stimulated contraction correlated to the amount of caspase-3 activation in the cytosol of the failing cells (Fig. C). Furthermore, blockade of caspase-3 activity by expression of p35 restored contractile performance of failing myocytes (Fig. C). Adenoviral infection did not alter caspase-3 activation, vMLC1 staining, or shortening characteristics of the cells. Discussion : In this study we used a modified yeast two-hybrid system to screen for caspase-3-interacting proteins of the cardiac cytoskeleton. We identified vMLC1 as a target for caspase-3, and demonstrated that its cleavage in failing myocardium in vivo is associated with a morphological disruption of the organized vMLC1 staining of sarcomeres. Furthermore, caspase-3 inhibition by adenoviral overexpression of p35 prevented vMLC1 cleavage with a positive functional effect on contractility. Interestingly, vMLC1 is cleaved by caspase-3 at a noncanonical cleavage site, DFVE, that has neither been described earlier in other substrates nor identified as a possible caspase-3 cleavage site in combinatorial approaches by using synthetic tetrapeptide substrates . It is widely accepted that caspases are "aspases," enzymes cleaving substrates carboxyl-terminal to an Asp residue, and that caspase-3 consensus motif is DXXD . However, recent evidence indicates that caspase-9 and DRONC, a Drosophila caspase, show activity toward sites with Glu and Asp residues at the P1 position . Furthermore, cleavage of the transcription factor Max by caspase-5, of tumor necrosis factor receptor-I by caspase-7, and of lens connexin 45.6 by caspase-3 have also been identified carboxy-terminal to a Glu . These findings suggest that the specificity of at least some caspases might be more complex than implied previously by studies with small synthetic peptides . Our data clearly confirm that the proposed requirement of caspase-3 for Asp in the P1 position is not absolute. Kinetic constants Km, kcat, and kcat /Km measured for the synthetic tetrapeptides Ac-DFVE-AMC and Ac-DEVD-AMC in vitro showed that caspase-3 has a cleavage preference for the classical site over the noncanonical . However, the differences in enzyme-substrate affinity, turnover, and specificity are in the range of those described for other caspase --substrate interactions . Because of the high substrate /enzyme ratio for vMLC1 and caspase-3 in cardiac myocytes, a lower cleavage specificity may serve as a physiological protection against a massive protein disruption by minimal caspase-3 activation. Moreover, other caspases did not cleave either the native vMLC1 (Fig. D) or the DFVE peptide, indicating that the atypical vMLC1 cleavage site is strictly caspase-3-specific. In a recent study, Communal et al. examined in vitro the cleavage of cardiac myofibrillar proteins by caspase-3. The authors reported that active caspase-3 targets alpha-actin, alpha-actinin, and troponin T, with consequences for myocyte function. In our yeast two-hybrid screening, one of the 22 positive candidates encoded the C terminus of troponin T, but neither the protein obtained by in vitro transcription /translation-reaction nor the native protein in ventricular extracts was cleaved by recombinant active caspase-3. Similarly, we did not detect any cleavage product for alpha-sarcomeric actin in cardiac lysates incubated with the active enzyme (data not shown). These discrepancies could be due to the differing cleavage reaction conditions in vitro or to different antibody sensitivity. In a rabbit model of CHF, we have previously demonstrated that caspase activation contributes to disease progression and influences contractile performance of failing ventricular myocytes by destroying sarcomeric structure . It is plausible that numerous combined biochemical pathways are responsible for caspase-mediated dysfunction and sarcomeric disarray of failing cardiac myocytes, and that several components of the myofilaments can directly be disrupted by caspase-3. In this study, we showed that vMLC1 is cleaved by activated caspase-3 in failing myocardium in vivo and its cleavage can be blocked by p35 overexpression with functional consequences on contractility. These findings support the notion that this sarcomeric protein is a target for caspase-3 and plays a central role in the setting of experimental heart failure obtained by rapid ventricular pacing. Generation of a mouse model carrying a cleavage-resistant vMLC1 by cardiac-specific knock-in techniques would specifically allow evaluation of the importance of vMLC1 proteolysis in cardiac apoptosis. A lot of data suggests that myosin light chains are important for cardiac and skeletal muscle function. Removing MLCs from chicken skeletal muscle myosin reduces the velocity of actin filament movement by 90% in an in vitro motility assay . Furthermore, MLC2 removal has little effect on isometric force, whereas MLC1 removal reduces the isometric force by over 50% . Mutations in the human essential light chain (Met-149 -> Val) or regulatory light chain (Glu-22 -> Lys, Pro-94 -> Arg) of myosin are associated with rare variants of inherited cardiac hypertrophy, characterized by midventricular cavity obstruction, and correlate with disruption of the stretch activation response of the cardiac papillary muscles . In the human heart, two different essential myosin light chain isoforms exist: (i) an atrial specific isoform (aMLC1, atrial essential myosin light chain), which is expressed in the fetal heart and decreases to undetectable levels during early postnatal development in the ventricle, but persists in the atrium for the whole life, and (ii) a ventricular specific isoform (vMLC1), which is the same isoform present in adult slow skeletal muscle . The reexpression of aMLC1 in adult human ventricles has been reported in patients with ischemic or dilative cardiomyopathy . Interestingly, in such patients with end-stage heart failure caspase-3 activation has also been documented . The isoform shift vMLC1 -> aMLC1 correlates with an increase in cross-bridge cycling kinetics, as measured in skinned fibers derived from the diseased muscle . Postsurgical return to a normal hemodynamic state decreases aMLC1 expression in these patients . The functional significance of this isoform switch is not completely clear, but may be a direct compensatory mechanism to caspase-3-induced vMLC1 cleavage, triggered when the heart attempts to maintain normal cardiac function. Taken together, these data clearly illustrate that minute changes in vMLC1 structure or composition can have a dramatic impact on myocyte function and heart contractility. The molecular mechanism for MLC1 to affect the cross-bridge kinetics seems to reside in its Ala- and Pro-rich extended N terminus, which has been shown to interact with the C terminus of actin. The extended MLC1 N terminus may provide a tether between the myosin and actin filaments, serving to position the two filament systems for cross-bridge interaction and to amplify small movements of the myosin globular head . The MLC1 C terminus anchors the protein to the myosin globular head . One could speculate that destruction of vMLC1 at the C-terminal motif DFVE135G by activated caspase-3 may alter myosin /actin cross-bridge interactions by modifying myosin head stability and thereby lead to reduced force transmission. In this study, we have demonstrated that vMLC1 is a cellular target for caspase-3. vMLC1 is cleaved, and its localization in sarcomeres is partially lost in failing cardiomyocytes, presenting caspase-3 activation and reduced contractile performance. Adenoviral expression of the caspase inhibitor p35 rescued the myopathic phenotype of the cells. It is plausible that vMLC1 disruption could alter the stiffness of the myosin neck region and therefore reduce the full range of myosin movement during contraction. Our findings suggest that caspase-3-mediated cleavage of vMLC1 may represent a molecular mechanism contributing to the deterioration of cardiac function before myocyte cell death. Backmatter: Abbreviations : CHF = congestive heart failure fmk = fluoromethyl ketone vMLC1 = ventricular essential myosin light chain vMLC2 = ventricular regulatory myosin light chain PMID- 12154231 TI - Restriction of lentivirus in monkeys AB - Retroviruses are able to cross species barriers and have done so many times throughout evolution. Perhaps as a consequence, dominant mechanisms have arisen to block infection by murine retroviruses in mice (restriction factor Fv1) and humans (restriction factor Ref1), as well as in other mammals. Here we describe a block to HIV and simian immunodeficiency virus in monkeys. Like previously described restrictions the block is saturable and gives rise to multiple-hit infection kinetics. Furthermore, like restriction of murine leukemia virus in humans, the block is before reverse transcription. Intriguingly, African green monkey cells are able to block both HIV and simian immunodeficiency virus, and each virus is able to saturate and abrogate the restriction of the other, suggesting that a common factor is responsible. Keywords: Introduction : Recent phylogenetic analysis of retroviral sequences has suggested that interspecies retroviral infection between mammals may have been frequent during their evolution . Certainly HIV types 1 and 2 are derived from simian immunodeficiency virus from chimpanzees (SIVcpz) and sooty mangabees (SIVsm), respectively . Despite these viruses' ability to cross species, several nonimmunological blocks to replication in foreign hosts have been demonstrated. Such a block exists for HIV type 1 (HIV-1) in rhesus macaques, the major HIV-1 primate experimental model . Much work from several groups has focused on the reasons for the block in macaques. Experiments in macaque PBL and simian MAGI cells have shown that the block is generally, although not always, before completion of reverse transcription . However, the block is not due to an inability of HIV-1 env to direct entry into macaque cells as shown by the efficient replication of SHIVs, chimeras of HIV-1 and simian immunodeficiency virus from macaques (SIVmac) encoding HIV-1 envelope, tat, rev, and vpu proteins and SIVmac gag-pro-pol. Attempts to map the determinant for the block more closely have been hampered by the difficulty of making functional HIV-SIV gag-pol fusions. However, a SHIV has been produced with an HIV-1 capsid-p2 domain replacing the equivalent SIVmac sequence . This virus was able to replicate in human cells but not macaque PBL, strongly implying a determinant in capsid. The virus incorporated cyclophilin and was inhibited by a cyclosporin analogue, further demonstrating the functional capsid phenotype of HIV-1. Unfortunately, the converse SHIV, an HIV-1 with an SIVmac capsid-p2 region, was noninfectious, demonstrating the difficulty of obtaining such chimeras . Further experiments on HIV-1 infection of macaque have shown that the coreceptor may somehow be involved. Although expression of human CD4 allows entry of HIV-1 into simian MAGI cells, the infection is nonproductive with a block before or after reverse transcription depending on the HIV-1 strain used. Expression of human coreceptor, however, facilitates their infection by HIV-1, suggesting that certain entry pathways may be able to bypass postentry restrictions . We have recently characterized a block to murine leukemia virus in a range of mammals, which seems to be due to expression of a saturable factor in the resistant cells . This block to infection resembles Fv1-mediated restriction in mice in its saturable nature and its target specificity. Characterization of these restrictions allowed us to develop an assay based on abrogation or saturation of restriction, depending on the ability of a sensitive virus to soak up and overcome the restriction factor. Whereas previous assays have been dependent on measuring relative infection by a pair of viruses, one restricted and one not, this assay uses a single restricting virus. We have used this assay to examine blocks to HIV-1- and SIVmac251-derived vectors in monkey cell lines. Although rhesus and owl monkey cells have been described as being refractory to HIV-1 infection , it has been unclear whether these cells possess factors preventing HIV-1 infection or whether HIV-1 is unable to interact with cell factors supporting virus infection. Here we demonstrate the presence of a saturable factor that is specifically able to block infection of lentiviruses. Materials and Methods : Cell Lines. | FRhK4 and LLC-MK2 were obtained from the Centro Substrati Cellulari, Brescia, Italy; and OMK, TE671, and SIRC cells were from the European Collection of Cell Cultures, Porton Down, U.K. CV1 cells were a kind gift of P. Jat, Ludwig Institute for Cancer Research, London. Cells were maintained as recommended by suppliers. Viral Vector Preparation. | Viral vectors were prepared by transfection of 293T cells by using Fugene-6 (Roche Molecular Biochemicals) as follows. To make HIV-1 vectors, confluent 293T cells were transfected on a 10-cm plate with a mixture of 18 mul of Fugene-6 in 200 mul of OptiMEM (GIBCOBRL) with 1 mug of pMDG [vesicular stomatitis virus envelope protein (VSV-G) expression vector ], 1 mug of p8.91 HIV-1 gag-pol expression vector , and 1.5 mug of retroviral expression vector encoding enhanced green fluorescent protein (eGFP), SIN CSGW, a kind gift of A. Thrasher, Institute of Child Health, University College London, U.K. or SIN CSPW encoding puro (puromycin resistance). To make SIVmac vectors 1 mug of SIV3+ (SIVmac gag-pol expression vector) and 1.5 mug of SIV-eGFP or SIV-beta-galactosidase (LacZ) (SIV vectors encoding eGFP and LacZ, respectively) were used . SIVmac plasmids were a kind gift of Francois-Loic Cosset, Ecole Normale Superieure de Lyon, France. Murine leukemia virus (MLV) vectors were made by using plasmids as described . VSV-G expression vector pMDG was used to supply envelope for most SIV and MLV preparations, whereas a cytomegalovirus promoter-driven expression plasmid, a gift of F.-L. Cosset, was used to produce an HIV-eGFP vector bearing amphotropic MLV Env. Viral supernatant was collected at 48, 72, and 96 h after transfection and stored at -80C. When required, virus was concentrated by centrifugation at 17,000 rpm in an SW28 rotor for 2 h. The pellet was resuspended in 1 /10 volume media. Viral Titer Determination. | All HIV-1 and SIVmac vector titers were measured on the permissive human cell line TE671, and titers are described as TE671 infectious units (i.u.) /ml. HIV-eGFP titers were determined 48 h after infection by a fluorescence-activated cell sorter (FACS) as described . HIV-puro titer was measured by infection of TE671 cells and colony counting after selection in 1 mug /ml puromycin. SIV-LacZ titers were determined 48 h after infection and LacZ staining as described . Abrogation Assays. | Abrogation assays were performed in six-well plates on 105 cells in 1 ml containing 5 mug /ml Polybrene. Cells were exposed to HIV-puro or SIV-LacZ for 4 h. Cells were washed and exposed to HIV-eGFP or SIV-eGFP. eGFP fluorescence was measured 48 h later by FACS analysis on a Becton Dickinson FACScan or LSR by using CELLQUEST software as described . In titration experiments in Figs. and , 50,000 cells were analyzed by FACS to increase sensitivity; in all other experiments 10,000 cells were analyzed. Mutation of Reverse Transcriptase Active Site. | Quikchange site-directed mutagenesis (Stratagene) was performed according to manufacturer's protocols by using p8.91 HIV-1 gag-pol expression vector and oligonucleotides forward GT141 CAATACATGGAAGATTTGTATGTAGGATC and reverse GT142 GATCCTACATACAAATCTTCCATGTATTG. The conserved YMDD motif was changed to YMED. TaqMan Quantitative PCR of Viral DNA. | As in the abrogation assays, 105 cells were infected in six-well plates in triplicate. Four hours after the second infection total DNA was extracted by using a DNeasy kit (Qiagen, Chatsworth, CA) from two samples. The third sample was subjected to FACS analysis 48 h later to measure infection. DNA (100 ng) was subjected to TaqMan quantitative PCR essentially as described by using TaqMan 2x quantitative PCR buffer (Applied Biosystems) with primers and probe at 300 nM and 150 nM, respectively. Primer and probe sequences were homologous to GFP and are as follows: GFP forward CAACAGCCACAACGTCTATATCAT, GFP reverse, ATGTTGTGGCGGATCTTGAAG, Probe 5'-FAM-CCGACAAGCAGAAGAACGGCATCAA-3'TAMRA . Results : To measure infection by retrovirus we made high-titer retroviral vectors encoding marker genes. These vectors were pseudotyped with the pantropic VSV-G unless otherwise stated, and are designated by a code describing the virus, HIV and SIVmac, and the marker gene, eGFP, LacZ, and puro. Saturation of HIV-1 Restriction in Monkey Cells. | Abrogation assays were performed on cells from rhesus macaque (FRhK4 and LLC-MK2) and owl monkey (OMK) because cells from these species have been shown to be refractory to HIV-1 infection . We also tested African green monkey CV1 cells and human TE671 cells, which are relatively permissive, and rabbit SIRC cells, which are highly resistant to VSV-pseudotyped HIV-eGFP vector. We were testing for an increase in HIV-eGFP titer after an initial exposure to HIV-puro or SIV-LacZ had saturated the available restriction factor. Cells were exposed to high doses of HIV-puro or SIV-LacZ, and then exposed to a low dose of HIV-eGFP. The titer of HIV-eGFP was increased by 10 times (CV1), even though these cells are relatively permissive, 20 times (FRhK4), 3 times (LLC-MK2), or 30 times (OMK) by preexposure to HIV but not when preexposed to SIVmac vectors . HIV-eGFP titer was increased by a small amount (5-fold) on CV1 cells after exposure to SIVmac and also on SIRC cells after exposure to either SIVmac or HIV vectors (4- and 10-fold, respectively). Figure 1 | (A) Abrogation of HIV-1 restriction in owl monkey cells. (A) Abrogation of HIV-1 restriction in owl monkey cells. FACS plots of side scatter versus GFP fluorescence are shown. Owl monkey cells are uninfected (i), infected with 105 TE671 i.u. of HIV-eGFP (ii), or exposed to 107 TE671 i.u. of HIV-puro for 4 h, washed, and then infected with 105 TE671 i.u. of HIV-eGFP. Region of GFP-positive cells (R1) and percentages of positive cells are indicated. (B) Abrogation of HIV restriction in monkey and rabbit cells. Cells (105) were exposed to 5 x 106 TE671 i.u. of either HIV-puro (square) or SIV-LacZ , incubated for 4 h, washed, and then exposed to 104 (TE671), 105 (OMK, FRhK4), 5 x 104 (LLC-MK2, CV1), and 5 x 106 (SIRC) TE671 i.u. of HIV-eGFP. Percentage infection was measured by analysis of eGFP expression 48 h later by FACS. Fold abrogation is calculated by dividing the increased percentage infection after abrogation by unabrogated control HIV-eGFP infection. Typically 1 --5% of unabrogated controls were infected. Errors are standard error of the mean of three independent experiments. Kinetics of Strong Restriction Are Multihit at High Virus Dose. | Restriction of MLV in mice by Fv1 and in humans by Ref1 has been shown to result in two-hit kinetics of infection at high virus doses , which means that the chance of a cell being infected is related to the square of the virus concentration and the slope of a log(% infection) vs. log(virus dose) plot is 2. To investigate infection kinetics of HIV restriction, serial dilutions of HIV-eGFP were titrated onto permissive and nonpermissive cells. We were seeking a characteristic bend in the titration curve, which appears as a restricting virus begins to soak up factors that inhibit infection, and a slope of 2 or greater at high virus dose. Fig. A shows a titration curve of HIV-eGFP on OMK, and Fig. B shows FRhK4 and LLC-MK2 cells. The characteristic bend, indicating a switch from one- to multiple-hit kinetics, can be seen as between 1% and 10% cells are infected. Rhesus macaque LLC-MK2 cells are infected 10 times more efficiently than FRhK4 cells, show no bend in the titration curve , and are not made significantly more permissive by abrogation . Two-hit kinetics was also seen when FRhK4 and OMK cells were infected with HIV-eGFP pseudotyped with MLV amphotropic envelope, demonstrating that restriction is not specific to VSV-G (data not shown). Fig. C shows titration of SIV-eGFP onto OMK, LLC-MK2, and FRhK4 cells. These cells do not restrict SIVmac and show single-hit kinetics of infection. Figure 2 | Titration of lentiviral vectors onto restricting cells. Titration of lentiviral vectors onto restricting cells. Two-fold serial dilutions of HIV-eGFP were titrated onto 105 OMK cells (A), LLC-MK2 and FRhK4 (circle) cells from rhesus macaque (B). (C) Two-fold serial dilutions of SIV-LacZ were titrated onto OMK (circle), FRhK4 , and LLC-MK2 (square) cells. Two-fold serial dilutions of HIV-eGFP (circle) or SIV-eGFP were titrated onto African green monkey CV1 cells (D) or rabbit SIRC cells (E). Results are representative of two independent experiments. Lines are guides for slopes of 1 and 2. Fig. D shows titration of HIV-eGFP and SIV-eGFP onto CV1 cells. Although HIV-eGFP shows single-hit kinetics, the slope for the SIV-eGFP titration curve was higher than 1, suggesting stronger restriction for SIVmac infection. Abrogation of this restriction by both HIV and SIVmac is further examined below (see Fig. ). Low Permissivity to HIV-eGFP Is Partly Due to Saturable Restriction in Rabbit SIRC Cells. | Rabbit SIRC cells are 3 --4 logs less infectable by HIV vectors compared with human TE671 cells. In Fig. E, these cells are infected with SIV-eGFP and HIV-eGFP. Rabbit SIRC cells are permissive for SIV-eGFP and single-hit kinetics are seen. At very high doses of HIV-eGFP, a switch to two-hit kinetics occurs, but SIRC cells remain at least 3 logs less permissive to HIV than SIVmac. We conclude that the block in SIRC cells is only partly due to a saturable restriction factor. Requirement for DNA Synthesis and Decay of Abrogation. | Abrogation of restriction should be mediated by the incoming virus components, most likely viral proteins. Restriction could be associated with the process of reverse transcription of the viral DNA because it seemed to occur at or before DNA synthesis (see Fig. ). To test whether DNA synthesis was required in the abrogating vector we prepared p8.91 HIV-1 gag-pol expression vector with a D185E change in the highly conserved YMD185D motif of the reverse transcriptase active site. This mutation completely blocks the ability of the virus to reverse transcribe but is unlikely to have any physical effect on retroviral core structure . HIV-eGFP was prepared with this construct and used to abrogate restriction in OMK and FRhK4 cells. This virus produced no infection on TE671 cells at high dose (data not shown). Fig. A shows that this virus was able to abrogate infection to a similar degree as wild-type HIV-puro (Fig. B). Fig. B also shows that restriction is about 20- and 30-fold for FRhK4 and OMK cells, respectively, and that 107 TE671 i.u. /ml of HIV-puro are required for maximum abrogation of restriction. We have measured the ratio of infectious dose to physical particles in vector preparations similar to those used here and estimate that 2 x 104 particles per cell are required to abrogate HIV-1 restriction. Figure 3 | Abrogation by reverse transcriptase mutant and time course of abrogation decay. Abrogation by reverse transcriptase mutant and time course of abrogation decay. OMK (circle) or FRhK4 cells (105) were exposed to serial dilutions of reverse transcriptase-defective HIV-eGFP D185E (A) or wild-type HIV-puro (B) for 4 h. Cells were washed and exposed to 105 TE671 i.u. of HIV-eGFP. Percentage infection was measured by analysis of eGFP expression 48 h later by FACS. Fold abrogation is calculated by dividing the increased percentage infection after abrogation by unabrogated control HIV-eGFP infection. About 1% of unabrogated controls typically were infected. p24 concentrations of wild-type and mutant supernatants were similar with around 500 ng /ml corresponding to a titer of 107 TE671 i.u. wild-type HIV-puro. OMK cells (105) were exposed to 107 TE671 i.u. of HIV-puro for 4 h, washed, and exposed to 105 TE671 i.u. of HIV-eGFP at 4-h intervals (C). Percentage infection and fold abrogation were determined as above. Data are representative of two independent experiments. To investigate the stability of the abrogation of restriction we performed the following experiment. OMK cells were exposed to high-dose HIV-puro for 4 h, washed, and then exposed to 105 TE671 i.u. of HIV-eGFP at various later time points. HIV-eGFP titer after abrogation linearly decreases with a half-life of around 9 h (Fig. C). SIV Restriction and Its Abrogation by Both HIV and SIV in African Green Monkey CV1 Cells. | CV1 data in Fig. B, showing that exposure to SIV-LacZ was able to increase the titre of HIV-eGFP, implied that there is a factor in CV1 cells restricting both HIV1 and SIVmac. Furthermore, a steep titration curve suggested stronger restriction of SIVmac than HIV in these cells . To examine this possibility further we titrated HIV-puro and SIV-LacZ onto CV1 cells and then infected them with a fixed dose of SIV-eGFP. High doses of either HIV or SIVmac vector were able to increase the titer of SIV-eGFP by up to 30-fold . Figure 4 | Abrogation of restriction of SIV-eGFP by SIV-LacZ and HIV-puro in African green monkey CV1 cells. Abrogation of restriction of SIV-eGFP by SIV-LacZ and HIV-puro in African green monkey CV1 cells. TE671 , LLC-MK2 (circle), OMK (x), or CV1 cells were exposed to 5 x 106 TE671 i.u. of SIV-LacZ or 5 x 106 TE671 i.u HIV-puro (CV1 only) for 4 h, washed, and exposed to 5 x 104 i.u. of SIV-eGFP. Percentage infection was measured by analysis of eGFP expression 48 h later by FACS. Fold abrogation is calculated by dividing the increased percentage infection after abrogation by unabrogated control HIV-eGFP infection. Approximately 1% of unabrogated monkey controls and 10% of human TE671 controls typically were infected. Data are representative of two independent experiments. Reverse Transcription Is Blocked in Restricted Cells. | Ref1 restriction of mouse viruses in human cells occurs at or before reverse transcription. To determine whether reverse transcription of HIV and SIV is inhibited in restricting monkeys we performed quantitative PCR to measure viral DNA synthesis on extracts from cells 4 h after infection. Equal virus doses (105 TE671 i.u.) were used to infect a fixed number of cells (105) with and without preexposure to an abrogating virus HIV-puro or SIV-LacZ. This assay indicates the difference between DNA synthesis between restricted and nonrestricted infection in the same cell line and differences in reverse transcription between species. In agreement with previous studies the block to infection in macaque cells is before reverse transcription . Figure 5 | Quantitative PCR of restricted and unrestricted infection. Quantitative PCR of restricted and unrestricted infection. FRHK4, OMK, CV1, or TE671 cells (105) were infected preexposed to 107 TE671 i.u. of HIV-puro for 4 h (square) or left unexposed , washed, and then infected with 5 x 104 TE671 i.u. of HIV-eGFP. For SIV on CV1 cells 106 TE671 i.u. of HIV-puro (square) or 106 TE671 i.u. of SIV-LacZ (striped bar) was used to abrogate and then 5 x 104 i.u. of SIV-eGFP was used to infect cells. Four hours after the second round of infection, total DNA was extracted and 100 ng was subjected to quantitative PCR (see Materials and Methods). Viral template copy number per 100 ng of total DNA was calculated by reference to a standard curve. Approximately 1 --10% of unabrogated monkey control cells were infected, and 35% of TE671 controls. Data are representative of two independent experiments with duplicate PCR. Errors are standard error of the mean. Parallel samples were analyzed by FACS, and levels of restriction were similar to those in Figs. and . Discussion : Saturable factors that restrict MLV in mice (Fv1) and humans (Ref1) have been described. Although rhesus macaques have long been known to have very low permissivity for HIV-1, with a viral determinant in gag-pol, the mechanism of the block to infection has remained unclear. These data demonstrate the presence of a saturable factor or factors able to restrict HIV-1 and SIVmac251 in simian cell lines. Fig. shows that the titer of HIV-eGFP vector can be increased in cells from rhesus macaque, owl monkey, and African green monkeys by preexposure to HIV-puro. Rhesus macaques remain the major primate experimental model, essential in the production of an HIV vaccine. The presence of a restriction factor against HIV-1 in macaque may explain the inability of HIV-1 to cause disease in these monkeys. Where a virus is restricted more than 20-fold, multiple-hit kinetics are seen at high virus dose , which indicates the factor is being soaked up by high-titer incoming virus. Two-hit kinetics implies that exposure to an initial restricted virion facilitates infection by a second restricted virion. In this case the probability of infection is related to the square of the concentration of the virus, and the slope of a log(infection) versus log(virus dose) is 2. These properties of restriction of retroviral infection are reminiscent of MLV restriction by Ref1 in human cells and Fv1 in mouse cells. The abrogation of HIV restriction does not require reverse transcription of abrogating virus particles (Fig. A), and the activity after exposure to abrogating virus has a half-life of 9 h (Fig. C). Rhesus macaque LLC-MK2 cells only weakly restrict HIV-1 (Figs. and ), and consequently, HIV-eGFP has a 10-fold higher titer on these cells than on FRhK4. Both FRhK4 cells, which restrict, and LLC-MK2, which do not, are derived from Macaca mulatta kidney. LLC-MK2 cells may have lost expression of the restriction factor. A precedent exists for cell lines from restricting mice lacking restriction in SC1 cells and 3T3FL cells . Further analysis of factor expression will require isolation of the gene or genes responsible. Rabbit SIRC cells are resistant to HIV-eGFP but not SIV-eGFP (Fig. D). The titer of HIV-eGFP is increased 10-fold by exposure to high-dose HIV-puro , and HIV-eGFP infection has multiple-hit kinetics at a high virus dose . Poor HIV infection in SIRC cells is therefore partly due to a saturable restricting factor(s). However, the low titer of HIV vector on SIRC cells may be largely due to an inability of HIV to use rabbit host cofactors. Difference in retroviral titer between cells from different species are likely to be dependent to some degree on cellular cofactor compatibility. Poor infection because of lack of compatibility is not likely to be saturable by preexposure to retrovirus. The data presented here show that restriction factors are more common than previously thought . The target for restriction is likely to reside in the capsid of the lentivirus. It has been shown that an SIVmac with an HIV-1 capsid-p2 domain has the restriction phenotype of HIV-1 rather than SIVmac and was able to replicate in human cells and not macaque peripheral blood mononuclear cells . Furthermore, the Fv1 /Ref1 target sequence is in the capsid of MLV . Cells from African green monkey are able to restrict SIVmac as well as HIV, and reciprocal saturation of the factor by SIVmac or HIV indicates that a single factor is responsible. The fact that a restriction factor can hit more than one virus suggests that the selecting virus, which forced the host to acquire this restriction factor, may be at least as unrelated to HIV or SIVmac as they are to each other. Although Fv1 has approximately 60% sequence homology with gag from HERV-L and MuERV-L , it is only distantly related to MLV, which suggests that lentiviral restriction is a function of retrovirus-like sequences unrelated to lentiviruses themselves. We have shown that cells from African green monkey are also able to restrict N tropic MLV , and it will be interesting to investigate the relationship between restrictions against C-type retrovirus and lentivirus. The large amount of retroviral sequence in mammalian genomes could serve as a pool from which protective sequence can be selected by pressure from pathogenic retrovirus. Some common aspects of the life cycle of distantly related retroviruses may present a capsid target susceptible to interference from gag-related molecules, which depend on their gag-like nature for their protective function. The further characterization and cloning of these factors will provide insight into early postentry events in the retroviral life cycle and reveal opportunities for therapeutic intervention. Backmatter: Abbreviations : MLV = murine leukemia virus SIV = simian immunodeficiency virus SIVmac = SIV from macaques eGFP = enhanced green fluorescent protein LacZ = beta-galactosidase FACS = fluorescence-activated cell sorter puro = puromycin resistance VSV-G = vesicular stomatitis virus envelope protein i.u. = infectious units PMID- 12185243 TI - Stimulation of the cell cycle and maize transformation by disruption of the plant retinoblastoma pathway AB - The genome of the Mastreviruses encodes a replication-associated protein (RepA) that interacts with members of the plant retinoblastoma-related protein family, which are putative cell cycle regulators. Expression of ZmRb1, a maize retinoblastoma-related gene, and RepA inhibited and stimulated, respectively, cell division in tobacco cell cultures. The effect of RepA was mitigated by over-expression of ZmRb1. RepA increased transformation frequency and callus growth rate of high type II maize germplasm. RepA-containing transgenic maize calli remained embryogenic, were readily regenerable, and produced fertile plants that transmitted transgene expression in a Mendelian fashion. In high type II, transformation frequency increased with the strength of the promoter driving RepA expression. When a construct in which RepA was expressed behind its native LIR promoter was used, primary transformation frequencies did not improve for two elite Pioneer maize inbreds. However, when LIR:RepA-containing transgenic embryos were used in subsequent rounds of transformation, frequencies were higher in the RepA+ embryos. These data demonstrate that RepA can stimulate cell division and callus growth in culture, and improve maize transformation. Keywords: Introduction : Like many mammalian DNA viruses, plant geminiviruses have efficient methods to subvert host cell cycle machinery and facilitate their replication . This occurs through interactions of viral replicase gene products and host cell components. In the Mastreviral subgroup of geminiviruses, which includes maize streak virus and wheat dwarf virus (WDV), two ORFs are differentially spliced, resulting in a mixture of the full-length replicase protein (Rep) and a truncated protein, RepA . Although Rep is required for viral replication, RepA is not . Rep and RepA were demonstrated to participate in many overlapping and nonoverlapping interactions with host functions. Examples include interactions of WDV RepA with developmental genes , transactivation of genes by Rep and RepA proteins , and direct interaction of RepA with host cell retinoblastoma-related (Rb) proteins . Of these, binding to Rb is the most thoroughly studied. Both Rep and RepA proteins have a Rb-binding motif, but it appears that RepA binds Rb more efficiently . The Rb gene family contains critical regulators of the G1 /S transition in animal systems. Rb binds to S-phase transcriptional transactivators, such as members of the E2F-family, and masks their activation domain without disrupting DNA binding at cell cycle-regulated promoters. Rb simultaneously recruits transcriptional repressors, such as histone deacetylases and methylases, and DNA helicases, to promote transcriptional quiescence. Animal DNA viruses encode Rb-binding proteins that relieve this repression, stimulate the cell cycle, and create a permissive environment for viral DNA replication . Similar to Rb-binding proteins in nonplant systems, it is speculated that RepA activates the expression of numerous genes that function in replication and S-phase progression. Although parallels between plant geminiviral and mammalian oncogenic viruses are striking in terms of viral replicase interactions with cell cycle proteins, there is one incongruous aspect to this comparison. The most obvious phenotypic impact of mammalian oncogenic viruses and the characteristic that led to intense efforts to unravel their biology is their stimulation of host cell proliferation . Immunochemical analysis of geminivirus-infected plant cells has identified a stimulation of replication-associated machinery. In terminally differentiated tobacco leaf cells, Rep expression or infection with tomato golden mosaic virus are associated with increases in proliferating cell nuclear antigen , an integral component of DNA polymerase regulated by E2F-binding sites in both animals and plants . Surprisingly, despite their interactions with Rb and activation of replication machinery, cell proliferation as a result of Rep or RepA expression has not been reported for plant cells. Thus, a consensus has developed that Rep and RepA do not induce dedifferentiation and reentry into the cell cycle, but rather up-regulate S-phase functions to facilitate viral genome replication . Plant transformation is impacted by cell cycle progression. In tobacco Bright-Yellow 2 (BY-2) cell cultures, transient beta-glucoronidase (GUS) expression increases when tobacco cells are transformed during G2 or M phase . In addition, transformation of synchronized tobacco protoplasts during S --M phase results in increased recovery of selection-resistant colonies and in higher copy, more complex transgene integrations when compared with nonsynchronized cells. More recently, Agrobacterium-mediated delivery of DNA was found to require transition through S phase . Although cell cycle dynamics can influence transformation, no practical methods have been reported to positively influence transformation by directly manipulating the cell cycle. Here we show that expression of the WDV RepA gene stimulates, and the maize ZmRb1341-866 gene inhibits, cell division in tobacco BY2 cells. RepA stimulated cell division is abrogated by the simultaneous expression of ZmRb1341-866. Furthermore, RepA expression stimulates maize embryogenic callus growth, increases transformation frequency, and eliminates the need for chemical selection of transformed calli. Rapidly growing RepA-expressing maize callus maintains its morphogenetic competency and elite maize inbred lines harboring a RepA transgene display an enhanced transformation phenotype. Materials and Methods : Constructs. | The plasmids in Table 11, which is published as supporting information on the PNAS web site, , were used for either stable or transient transformation of plant materials. For brevity, only the gene components for individual constructs are described here. Visible marker genes, uidA (GUS; ref. ) and green fluorescent protein (GFP; ref. ), and a maize codon-optimized version of GFP (moGFP) were used to identify transformed cells. A plasmid containing the firefly luciferase gene was used to balance the DNA content of particle bombardments in some experiments. For maize cobombardment experiments, a fusion between a maize-optimized phosphinothricin acetyl transferase (moPAT) gene and moGFP was generated (moPATsimilarmoGFP). In retransformation experiments, a fusion between domains of R and C1 (CRC) encoding the protein functions necessary to activate anthocyanin accumulation was used as a marker to score transgenic calli, and the bar gene was used for bialaphos selection. Promoters driving expression of visible marker genes included a double-enhanced cauliflower mosaic virus (CaMV) 35S promoter , the nopaline synthase (Nos) promoter , and the maize ubiquitin (Ubi) promoter . Downstream 3' regions used in expression cassettes included those from a proteinase inhibitor (pinII; ref. ) and CaMV 35S . The coding sequences for the wild-type ZmRb1341-866 and ZmRb1C706G mutant sequences were derived from constructs described by Grafi et al. and are 5' truncations of ZmRb1, consisting of amino acids 341 --866 of the RRB1 sequence (GenBank accession no. ). pWI-11 was the parental plasmid for all WDV Rep and RepA sequences . Two Asp700 restriction sites within the WDV Rep sequence were used to make an internal deletion that removed the intron splice junction sequences. The resulting nucleic acid sequence encodes a RepA protein (RepAAsp700) in which the last five carboxy-terminal amino acids are altered from -PGNGK to -RRGSA. To more completely eliminate intron-splicing and reduce the A /T-content of the intron sequence, additional constructs were made by amplification of the RepA or Rep coding sequences by mutagenic PCR. This did not alter amino acids in Rep, but resulted in a valine to lysine substitution in the RepA coding sequence (at the site of the 5' splice junction). The "intronless" sequences, are referred to as Repm and RepAm. Four constructs were used for Agrobacterium-mediated transformation experiments. In the first (designated p108), a DNA fragment containing the Nos promoter, a gene encoding the CRC fusion protein, and the pinII 3' end was inserted between the T-DNA borders in pSB11. The second, p109, contained two separate T-DNA-border-flanked cassettes, LIR:RepAAsp700 and 35S:bar /Ubi:moGFP:pinII. The third (p111) contained LIR:RepAAsp700, Ubi:moGFP and 35S:bar within a single set of T-borders, and the fourth (p110) contained 35S:bar + Ubi:FLP:pinII within a single set of T-borders. Agrobacterium tumefaciens LBA4404 and vectors pSB1 and pSB11 were obtained from Japan Tobacco (700 Higashibara, Iwata, Japan). Tobacco Cultures and Transformation. | Nicotiana tabacum L. cv. Bright Yellow 2 (BY-2) suspension cultures were used for evaluation of RepA and ZmRb1 expression on cell division. Suspension cells were subcultured every 7 --10 days into fresh Nt1 medium (see Table 12, which is published as supporting information on the PNAS web site, for media formulations) and grown on a gyratory shaker at 150 rpm, 24C in the dark. Three, six, or nine days after subculturing, cells were pipetted onto solidified agar medium (Nt1) and left in the dark for 24 h before bombardment. Plasmid DNA was precipitated on 1-mum gold particles with polyethylene glycol (PEG) . Bombardment was performed with a Bio-Rad PDS-1000 helium gun, 650 PSI rupture discs, a 60-cm Hg vacuum and 8 cm between the stopping plate and Petri dish. Cells were shot once with 500 ng gold and 0.5 mug DNA. All tobacco bombardments used the 35S:GFP expression cassette. Treatments consisted of 35S:RepA, 35S:ZmRb1341 --866, 35S:ZmRb1706G, or combinations thereof. Twenty-four hours after bombardment, cells were monitored for GFP expression and cell division by epifluorescence microscopy. The total number of GFP-expressing foci and whether they comprised single or multiple cells was recorded. Data from the controls and treatments were tested for significant differences by Student's t test . Maize Transformation. | A publicly available maize (Zea mays L.), hybrid high type II (Hi-II) , and Pioneer proprietary inbreds P38 and N46 were grown in the greenhouse at 16-h day length. Immature embryos (1.0 --1.5 mm) were excised from fresh developing kernels and used for transformation. Particle-mediated transformation followed a standard protocol . After 4 or 5 days incubating Hi-II immature embryos on 560P medium (see Table 12 for media formulations) in the dark at 28C, embryos were transferred onto 560Y and cultured scutellum-side-up for 3 h before transformation. The scutellar surface was targeted with the PDS-1000 Helium Gun from BioRad at one shot per sample using 650 PSI rupture disks. Approximately 67 ng of DNA was delivered per shot. A similar number of embryos per ear were bombarded for each treatment in an experiment, in aliquots of 25 embryos per plate. After bombardment, embryos were maintained on 560L medium. Transformants were transferred 2 --7 days after bombardment onto 560R for selection. Plates were maintained at 28C in the dark, and transferred to fresh medium every 2 weeks. GFP+ and /or bialaphos-resistant (BAR) calli were scored at 6 --8 weeks, with each plate of embryos as a replicate. Calli were examined by epifluorescence with a dissecting microscope using a filter set (Chroma no. 41020) for GFP excitation and emission. When colony size was recorded, two perpendicular measurements were taken for each independent, GFP-expressing multicellular colony. The two measurements were averaged, providing an estimate of the diameter for each colony. These values were used to calculate the colony's presumed spherical volume (V = 4 /3pir3). After approximately 10 weeks of selection, BAR, GFP-positive calli were scored. Positive lines were transferred to a Murashige and Skoog (MS)-based medium with reduced sucrose and hormone levels to initiate plant regeneration . After somatic embryo maturation (2 --4 weeks), well-developed embryos were transferred to germination medium for 7 --10 days and placed in light. Developing plantlets were transferred to medium in tubes for 7 --10 days until well established. Plants were then transferred to flats (equivalent to a 6.4 cm pot) containing potting soil, grown for 1 week in a growth chamber and an additional 1 --2 weeks in the greenhouse. The plants were finally transferred to 6-liter pots (catalog no. 14-9674-9; Hummert International, Earth City, MO) and grown to maturity. Mature plants were crossed to untransformed plants of the same genotype for analysis of inheritance and retransformation. Methods for Agrobacterium-mediated transformation of maize followed the general protocol described with the following modifications. Agrobacteria were grown to log phase in liquid minimal A medium containing 100 muM spectinomycin. Cells were transformed by culturing P38 or N46 immature embryos in liquid 700 medium. Embryos were immersed in a log phase suspension of Agrobacterium [5 x 108 colony forming units (cfu) per ml]. Embryos were infected for 5 min by using gentle rotation of the Agrobacterium suspension and then cocultured, embryo axis down, for 7 days in the dark at 20C on 710 medium. Embryos were then transferred to 720E selection medium. Plates were maintained at 28C in the dark and observed for colony recovery with transfer to fresh medium every 2 weeks. After 6 --8 weeks, selection-resistant and /or GFP expressing colonies were transferred to 288J maturation medium to begin plant recovery. Colonies were maintained for 7 days in the dark at 28C on this medium, followed by transfer to medium for plant recovery in the light. Recovered plantlets were transferred to culture tubes containing 272 medium for root development before transfer to the greenhouse. Plants were scored based on continued GFP expression, leaf sensitivity to painting with 1% phosphinothricin (Aventis CropScience, Research Triangle Park, NC), and molecular characterization via PCR and Southern blot analyses. Retransformation. | Primary transformations were done as described above by using the Agrobacterium vectors p111 or p109, and transgenic inbred plants were regenerated and pollinated from nontransformed inbreds. When the "two T-DNA" vector (p109) was used, a high percentage of events was recovered in which the two T-DNA's segregated independently . Pollen from wild-type inbred individuals was carried onto inbred transformants hemizygous for either the LIR:RepAAsp700 /Ubi:GFP /35S:bar or the LIR:RepAAsp700 locus, respectively, and the resultant embryos, segregating 1:1 for the presence of RepA, were subjected to a subsequent round of transformation (termed "retransformation"). For retransformation, Agrobacterium was used to deliver either CRC (into LIR:RepAAsp700 /Ubi:GFP /35S:bar containing embryos) or 35S:barUbi:FLP (into LIR:RepAAsp-700 containing embryos). Stable anthocyanin-accumulating or BAR events were scored 6 --8 weeks after Agrobacterium-mediated delivery of the retransformation expression cassette. For 35S:bar retransformation experiments, data were collected as the number of embryos regenerating BAR, PCR-positive plants, relative to the total number of treated embryos. Presence of RepA in segregating material from which transformants were recovered was confirmed by quantitative PCR . Results : RepA Stimulates and ZmRb1 Inhibits Cell Division in BY-2 Cells. | To test whether RepA influences the plant cell cycle, a transfection-based assay of cell cycle regulatory function was done with BY-2 cell cultures. Bombardment with gold particles was used to deliver 35S:GFP and 35S:RepA, and after 24 h GFP-expressing cells were scored as single or divided (Fig. a). If mitotically active cells (4 days after subculture) were bombarded with 35S:GFP alone, approximately 38% of fluorescent foci divided within 24 h . Codelivery of 35S:RepA and 35S:GFP increased the proportion of divided cells to approximately 60%. These proportions are significantly different when compared by Student's t test (P < 0.01), suggesting that RepA stimulates the cell division cycle in actively growing BY-2 cell suspensions. Figure 1 | Effect of RepA expression on plant cell division and callus growth. Effect of RepA expression on plant cell division and callus growth. (a) Recent cell division in a GFP-expressing BY-2 cell culture clearly showing two daughter cells separated by a newly formed transverse wall. (b) GFP expression 2 weeks after particle delivery of Ubi:moPATsimilarmoGFP:pinII DNA into Hi-II immature embryos; only single cells expressing GFP were observed. (c) GFP expression 2 weeks after particle delivery of Ubi:moPATsimilarmoGFP:pinII and Ubi:RepA; multiple GFP-expressing multicellular colonies were observed in addition to single cells expressing GFP. (Scale markers in b and c = 500 mum.) Table 1 | RepA-stimulates cell division in BY-2 cells BY-2 cells enter a stationary phase 7 days after subculturing, when cell density stabilizes, the mitotic index decreases, and cells arrest in G1 /G0 . In 7- and 14-day-old cell cultures transfected with 35S:GFP, the proportion of divided foci were significantly lower (20 --25%) than that of 4-day-old cell cultures (t test, P < 0.01; Table ). Treatment of these cultures with 35S:RepA significantly increased the proportion of divided cells . In fact, 35S:RepA induced a greater fold-increase in cell division in the stationary cultures, suggesting that RepA is able to overcome a G0 /G1-block in BY-2 cultures. To test the hypothesis that RepA acts by means of a plant Rb gene family member, the 35S:GFP, 35S:RepA, and 35S:ZmRb1341-866 cassettes were codelivered to 4-day-old BY-2 cell cultures. Bombardment with 35S:GFP produced fluorescent foci with a similar proportion of multicellular events as observed in the previously described experiments (compare Tables and ). The codelivery of 35S:ZmRb1341-866, but not a mutant 35S:ZmRb1C706G, significantly decreased the proportion of divided cells, demonstrating that the ZmRb1341-866 gene product, which includes the pocket domain of ZmRb1 , is sufficient for cell cycle arrest. The codelivery of 35S:RepA and 35S:GFP increased the number of fluorescent foci with multiple cells to approximately 54.3%, similar to proportions seen previously (Tables and ). When 35S:RepA and the mutant 35S:ZmRb1C706G were codelivered, the number of fluorescent foci was indistinguishable from 35S:RepA treatments (t test, P < 0.01). If GFP, 35S:RepA and 35S:ZmRb1341-866 cassettes were simultaneously delivered, fluorescent foci were found to contain proportions of single and multiple cell clusters indistinguishable from control treatments (t test, P > 0.05) and significantly different from either construct alone (P < 0.01). Thus, RepA and ZmRb1341-866 appear to influence the cell cycle in opposition. Table 2 | ZmRb1 opposes RepA in BY-2 cells RepA Stimulates Early Growth of Maize Transformants. | To investigate whether RepA can increase cell division in maize embryos, they were transformed with moPATsimilarGFP and Ubi:RepA or Nos:RepA. When compared by fluorometric quantitation of GUS activity, the Ubi promoter is 5- to 10-fold more active than Nos in immature embryos (data not shown). By 16 days after bombardment, GFP-expressing cells were observed on the surface of scutellar-derived tissue. Only single GFP-expressing cells were observed (Fig. b) in the majority of control embryos. At this time, single GFP-expressing cells and macroscopic GFP-expressing multicellular clusters were apparent in RepA treatments (Fig. c). The proportion of embryos with GFP-containing colonies and the colonies observed per embryo in the Nos:RepA and Ubi:RepA treatments were greater than controls . If each embryo producing a transformant was counted, the transformation frequencies for the control, Nos:RepA or Ubi:RepA treatments were ~3, 16% and 20%, respectively. However, if GFP colonies were scored as independent events, the control, Nos:RepA and Ubi:RepA treatments contained 6, 59, and 80 transformants with transformation frequencies of 3%, 29%, and 46%, respectively. Macroscopic colony size was also sensitive to promoter strength as Ubi:RepA colonies were significantly larger than Nos:RepA colonies (Table ; Ranked t test, P < 0.01). Table 3 | RepA increases GFP colonies per embryo in maize Table 4 | RepA stimulates multicellular cluster growth in maize RepA Increases Maize Transformation Frequency. | As Tables and show, RepA stimulated early growth in GFP-expressing cell clusters and the recovery of transformed calli. To determine whether the Rep protein had a similar impact on maize transformation, Nos:RepAm or Nos:Repm were cobombarded with moPAT-GFP and compared with control treatments. Nos:RepAm significantly increased the recovery of BAR transformants relative to the moPATsimilarGFP control treatment, whereas Nos:Repm treatments were indistinguishable from moPATsimilarGFP controls . To extend this observation and determine whether transformation frequency, like callus growth, is sensitive to promoter strength, embryos were cobombarded with Nos:RepAm or Ubi:RepAm and moPATsimilarGFP. When compared with moPATsimilarGFP controls, both Nos:RepAm and Ubi:RepAm significantly increased transformation frequencies . Furthermore, Ubi:RepAm treatments displayed significantly higher transformation frequencies relative to Nos:RepAm (t test, P < 0.05). Table 5 | Nos:RepA, but not Nos:Rep, increases transformation efficiency in maize Table 6 | Maize transformation is sensitive to RepA promoter strength As RepA promoted growth, we investigated its utility for identification of transgenic calli in the absence of chemical selection. Visual screening for GFP fluorescence recovered one-fourth the number of transformants obtained with bialaphos selection . Cobombardment of embryos with Ubi:RepA and moPATsimilarGFP substantially increased the frequency of recovering BAR events. In the absence of selection, the inclusion of RepA significantly increased the efficiency of transformant recovery by visual screening for GFP , such that it was indistinguishable from that obtained by chemical selection in the presence of RepA (t test, P > 0.05). Table 7 | Ubi:RepAm increases maize transformation efficiency without chemical selection RepA+ Inbred Germplasm Exhibits Enhanced Transformation. | All experiments described thus far used embryos from the transformation-competent Hi-II maize hybrid. Transformation of agronomically elite maize inbreds would represent a substantial advancement in maize biotechnology. Our initial attempts to extend these results to elite maize inbreds used LIR:RepAAsp700 introduced, along with 35S:bar and Ubi:moGFP (p111), into maize embryos via Agrobacterium, followed by phosphinothricin selection. As seen in Table , although we recovered transformants in both the control and RepA treatments for inbreds N46 and P38, LIR:RepAAsp700 did not significantly enhance the transformation frequencies. However, the transformation frequency for Hi-II embryos was increased by LIR:RepAAsp-700 (t test, P < 0.05). The transgenic plants were regenerated, grown to maturity in the greenhouse and crossed to their respective parents. In progeny, LIR:RepAAsp-700, Ubi:moGFP, and 35S:bar segregated as a single locus that was followed by fluorescence and PCR (data not shown). Agrobacterium-mediated gene transfer was used to deliver Nos:CRC, a cell autonomous stimulator of anthocyanin production , to T3 embryos harvested from segregating ears. Calli were screened for multicellular pigmented clusters in the absence of chemical selection. As shown in Table , GFP-expressing and nonfluorescent embryos from GFP /bar hemizygotes had similar frequencies of anthocyanin-expressing calli. However, for both inbreds, GFP-positive embryos from ears hemizygous for RepA /GFP /bar exhibited a significantly higher transformation frequency (t test, P > 0.01). Table 8 | LIR:RepA does not improve primary transformation of maize inbreds Table 9 | Stable integration of LIR:RepA improves subsequent inbred transformation The presence of RepA in the genetic background also improved the efficiency of transformation with chemical selection. Inbreds N46 and P38 were transformed with LIR:RepAAsp-700 and Ubi:moGFP /35S:bar using a "two T-DNA" vector (p109) and backcrossed. After segregating the RepAAsp700 and GFP /bar loci, Southern analyses, fluorescence imaging, and herbicide resistance tests verified that the RepAAsp700 cassette was intact and the GFP /bar locus was no longer present. Hemizygous RepAAsp700 N46-inbred plants (T2 generation) were crossed to nontransformed N46, and the segregating T3 embryos were retransformed by cocultivation with Agrobacterium containing 35S:bar /Ubi:FLP (p110), followed by selection on bialaphos. chi2 tests demonstrated that BAR transformants were more likely to have arisen on RepA transgenic embryos in all transgenic lines tested , indicating that stable RepA expression enhanced transformation. Surprisingly, higher transformation frequencies were also seen for nontransgenic embryos harvested from hemizygous ears than for wild-type embryos on nontransgenic ears . This result demonstrates that the LIR:RepAAsp700 transgene has a maternal effect on embryo transformation competence. Table 10 | Stable transformation with LIR:RepAAsp700 improves maize inbred N46 transformation with chemical selection Discussion : Similarities between interactions of the RepA protein with plant cell proteins and the resultant subversion of host cell cycle machinery led various researchers to suggest a similarity between plant geminiviruses and mammalian oncoviruses . We can now expand these similarities to RepA-mediated stimulation of the cell cycle in BY-2 cells and callus growth in maize, observations consistent with the stimulation of cell proliferation associated with oncoviral infection . The inhibition of cell division by 35S:ZmRb1341 --866 also provides direct evidence in favor of the plant Rb gene family encoding cell cycle regulators. In mammalian cells, oncovirus-induced cell proliferation is typically associated with an undifferentiated cell phenotype. In contrast, despite rapid growth, RepA-treated transgenic calli generally maintain morphogenic competence. However, some of the most rapidly growing Ubi:RepA transformants were less embryogenic and more difficult to regenerate (data not shown). Delivery of RepA to Hi-II derived calli increased transformation efficiency, and increasing RepA expression enhanced this effect. Additionally, RepA transgenes conferred a high-competence transformation phenotype to both Hi-II (data not shown) and elite inbred genetic backgrounds, overcoming genotype-dependence. Furthermore, this effect was manifested by expression of RepA in the embryo and the ear, suggestive of maternal conditioning of transformation competence. Stimulation of plant cell division and growth has not been reported previously in studies on geminiviral Rep (or RepA). This may be caused by alternative splicing in WDV or maize streak virus, the two Mastreviruses whose Rep have been most thoroughly studied. Inefficient splicing of the single intron in plant cells results in a mixture of Rep and the shorter RepA proteins . These two proteins have different properties, and likely have different impacts on cell physiology. Consistent with the two proteins having different effects on plant growth, Nos:RepA expression, but not Nos:Rep, resulted in increased maize transformation. As others have not observed cell cycle stimulation , there may be restricted cell- or tissue-type responsiveness to RepA. Other cell-type-specific responses to geminiviral Rep expression and replication were reported for wheat suspension cultures and scutellar cells . Embryonic and /or meristematic cells may be more readily stimulated to divide given an appropriate stimulus, i.e., the presence of RepA, and subsequent patterns of morphogenesis, differentiation, and position-sensitive controls on cell division might simply be more difficult to reverse. This may be particularly true in cereals, where cellular plasticity appears to be strongly suppressed . Protein interaction between RepA and Rb's , supported by our data demonstrating ZmRb1341-866-mediated suppression of RepA-stimulated cell division, suggest that increased transformation is a consequence of the relief of Rb repression of the cell cycle. The increase in maize transformation could be caused by enhanced transgene integration. If this were occurring, an increase in transgene copy number might be expected. We examined copy number in RepA transgenic material by Southern analyses. Despite an increased transformation frequency, the range of transgene copy number and complexity in RepA T0 transformants was similar to control treatments with no RepA (data not shown) and to typical particle bombardment-generated transformants in the literature . In a study of Agrobacterium-mediated gene transfer, failure of transformed cells to proliferate and T-DNA silencing were more significant barriers to transformation than T-DNA integration . Therefore, we believe the increased transformation frequencies we observed are more likely to reflect increased plant recovery caused by stimulated cell division, rather than enhanced DNA integration. If RepA acts to stimulate cell division in maize, it should function as a positive "selection" marker. Indeed, in the absence of chemical selection, RepA expression increased the efficacy of visual screening for transformants. The inclusion of bialaphos selection did not further increase transformant recovery, supporting this interpretation. Although visual screening without chemical selection was previously reported for both maize and oats , our results with RepA and GFP represent a many fold improvement over previous methods. Our results demonstrate that RepA improves primary transformation in Hi-II maize embryos. However, Hi-II is a model germplasm for maize tissue culture . Although Hi-II transformation has continued to improve, most elite maize inbreds are recalcitrant to transformation. Results from primary transformation of two Pioneer elite inbreds with LIR:RepAAsp700 confirmed this. Nonetheless, in subsequent generations LIR:RepAAsp700 transgenic germplasm was more transformable than siblings lacking RepA. Experiments using CRC as a marker gene demonstrated that RepA integration improved transformation efficiency from 0.5% to ~25%. Subsequent experiments using herbicide selection showed that LIR:RepAAsp700 conditions both higher and more consistent transformation frequencies in the T3 generation. Thus, the positive impact on transformation is a heritable trait. Furthermore, increased transformation was observed for nontransgenic embryos harvested from LIR:RepAAsp700 hemizygous ears, strongly suggesting that RepA expression also exerts a maternal influence on the transformability of embryos. Our observation that RepA-enhanced transformation is heritable has an important implication. First, an initial transformation with RepA can be used to create transformable inbred germplasm. On subsequent retransformation, for example with an agronomically important gene, the resultant plants can easily be crossed and the RepA transgene (which exists as a separate locus) segregated away from the agronomically important transgenes. When this approach is used, the RepA-gene becomes a laboratory and greenhouse tool that is readily left behind before the newly generated transgenic inbreds are moved into the field. Backmatter: Abbreviations : BAR bialaphos-resistant = Hi-II, high type II, BY-2 bright yellow 2 WDV wheat dwarf virus = GUS, beta-glucuronidase GFP = green fluorescent protein moGFP = maize codon-optimized GFP Rep = replicase protein RepA = replication associated protein A Rb = retinoblastoma-related pinII = proteinase inhibitor Ubi = ubiquitin Nos = nopaline synthase moPAT = maize-optimized phosphinothricin acetyl transferase PMID- 12193650 TI - Sleep forms memory for finger skills AB - Practicing a motor skill triggers a process of memory consolidation that continues for hours after practice has ended, and becomes manifest in an improved skill at later testing. We used a sequential motor task (finger-to-thumb opposition task) to show that, in humans, the formation of motor skill memories essentially benefits from sleep. Independent of whether placed during daytime or nighttime, sleep after practice enhanced speed of sequence performance on average by 33.5% and reduced error rate by 30.1% as compared with corresponding intervals of wakefulness. The effect of sleep after learning proved to be stable when retesting was postponed for another night, to exclude effects of sleep loss and to assure that all subjects had sufficient sleep before retrieval testing. Also, the consolidating effect of sleep was specific for the motor sequence learned. It did not generalize to a similar sequence containing identical movement segments in a different order. Retention periods of wakefulness improved performance only moderately and only if placed during daytime. The observations demonstrate a critical role of sleep for storing and optimizing motor skills. Keywords: Introduction : Memory consolidation refers to processes of brain plasticity by which experiences result in more or less enduring changes in adaptive behaviors. In the case of motor skills, practicing a motor task leads to the generation of an internal model representing the different motor outputs in response to the various task stimuli . Development of the internal model does not stop when practice ends, but continues over hours during which the memory traces of the model are strengthened becoming, for example, increasingly resistant to behavioral interference . Most importantly, the internal model becomes also more effective in this process, as indicated by a distinct gain in performance at retesting 24 h later when the task is performed with greater speed and accuracy . A critical amount of practice is considered to initiate plastic processes, probably mainly in the primary motor cortex, which gate a shaping of the motor representations, and thereby improve performance in the absence of any further training . This latent formation of motor memories has been proposed to be linked to a dynamic reorganization of the respective motor neuronal networks , and to require a covert reprocessing of the memory traces . Sleep, characterized by largely suppressed overt motor activity and conjunct sensory input, might represent a condition optimal for this reprocessing. However, its possible role for motor skill formation has not been assessed in depth. Evidence has accumulated supporting the notion that processes during sleep significantly contribute to the formation of different types of memory . Although earlier human studies suggested facilitating influences of sleep mainly on hippocampus-dependent declarative types of memory, recent studies indicated a similar influence on the formation of procedural memories . In contrast to declarative memory, which refers to the knowledge of facts and events, procedural (or "how to") memory for skills does not require the integrity of the hippocampus. Also, acquisition is slower and shows little generalization to similar behaviors . In many cases, procedural learning appears to be linked to discrete changes in low-level representations in the hierarchy of sensory input and motor output processing, taking place in specific sensory and motor cortical areas . Several recent studies in humans have shown that the consolidation of memories for elementary perceptual skills (texture discrimination) critically depends on sleep . Retrieval testing on the texture discrimination task in these studies revealed a significant performance gain only if the learning session was followed by a period of sleep. Deprivation of sleep after practice completely prevented subsequent formation of memory for the trained skill, even when retrieval testing was delayed to allow for recovery sleep, indicating that the first nocturnal sleep period after practice is critical for initiating memory consolidation . However, these previous studies exclusively focused on perceptual skills. The present experiments were stimulated by indications that sleep could play a similar critical role for the slow latent consolidation process of motor skills . Methods : Participants. | A total of 52 healthy young student volunteers (18 --29 years old; mean +- SD: 23.31 +- 2.69 years) who were nonsmokers, right-handed, and had no history of sleep disturbances participated in the experiments. None of the subjects had practiced playing a musical instrument nor was trained as a typist. All subjects regularly obtained 7 --8 h of sleep per night and had no disruptions of the sleep --wake cycle during the 6 weeks before participation. All spent an adaptation night in the sleep laboratory before beginning the experiments. Subjects abstained from caffeine and alcohol the day before the experimental session. They were instructed to get up before 7:00 a.m. and not to take naps during the day. The experiments were approved by the Ethics Committee of the University of Lubeck. Learning Task. | The finger-to-thumb opposition task (Fig. A) required the subject to tap with his nondominant hand (left hand) the finger sequence as rapidly and accurately as possible without looking at his hand. To familiarize the subject with the task he first practiced the task, with visual feedback provided through the monitor indicating the next finger to be tapped within a response interval of 400 ms. When a criterion of 10 consecutive correct reactions was met, the training period proper started, consisting of three 5-min blocks interrupted by two 2-min periods of rest. The procedure of retrieval testing after the retention interval was the same as during the training proper before sleep. During task performance the subject sat in a silent and darkened room. All instructions were presented on a 15-inch monitor. To register movements, each fingertip of the left hand was covered with aluminum foil connected to a personal computer. Two different motor sequences were used to allow testing of the same subject on two occasions. Figure 1 | (A) Finger-to-thumb opposition task. (A) Finger-to-thumb opposition task. The motor skill task was adopted from Karni et al. and demanded the subject to oppose the fingers of the nondominant hand to the thumb in a certain sequence. Two sequences were used on different conditions, which both were composed of the same five movements but in a mirror-reversed manner. In sequence A, the order of the fingers was 4, 1, 3, 2, 4. In sequence B, the order was 4, 2, 3, 1, 4 (finger numbering is from index to little). During both training and retrieval testing, the subject was asked to tap the given sequence as fast and as accurately as possible without looking at his hand for three 5-min blocks. (B) Protocol for main experiments. Subjects received training (left black fields) on a finger sequence before 8-h retention intervals during which they either slept or stayed awake. Thirty minutes after the end of the retention period, retrieval was tested (right black fields). Retention periods were placed either at night (Upper) or during the day (Lower). Design and Procedure. | In the main experiments, subjects (n = 20, 9 female) were trained on the finger-to-thumb opposition task in the evening (at 10:00 p.m.) before a nocturnal 8-h retention interval during which they either slept regularly or stayed awake. Thereafter (at 7:30 a.m.), motor performance was retested. To examine effects of sleep independent of circadian factors, possibly affecting learning and motor performance , on a further condition learning took place in the morning (at 10:00 a.m.) before 8-h retention intervals of sleep and wakefulness placed during daytime. Retesting took place at 7:30 p.m. (Fig. B). Awakening from sleep was always 30 min before retrieval testing. Each subject was assigned to one nocturnal and one daytime retention condition, performing on each of these occasions one of the two motor sequences, respectively. The order of retention conditions and of motor sequences was balanced across subjects. Experimental conditions for each subject were at least 1 week apart. Subjects stayed awake the night before the daytime sleep condition to enable daytime sleep. Three supplementary studies were performed to examine (i) effects of sleep loss (n = 16, 6 female), (ii) effects of sleep versus wakefulness on a delayed retrieval (n = 6, 2 female), and to test (iii) for the specifity of the memory effects for the particular motor sequence (n = 10, 4 female). Data Reduction and Analysis. | Sleep recordings were visually scored according to standard criteria . Performance in the finger-to-thumb opposition task was measured in terms of performance rate (mean number of correctly completed sequences per 30 s) and accuracy (mean number of errors per 30 s). The initial 30-s period of each block served as an adaptation period and was not included in the analysis. Changes in performance rate and accuracy across the retention intervals were also transformed to percentages, with the individual performance value at learning set to 100%. Repeated-measures ANOVA including a "before /after" and a "sleep /wake" factor with subsequent pairwise contrasts were used to analyze performance rate and errors. A P value < 0.05 was considered significant. Results : Total sleep time was closely comparable during both the nocturnal and daytime retention intervals . Expected circadian influences on sleep expressed themselves in a decreased time spent in stage 2 sleep (P < 0.01) and a tendency toward increased time in stage 1 sleep and rapid eye movement (REM) sleep (P < 0.1) during daytime sleep in comparison with nocturnal sleep. Initial learning of the sequence before the 8-hour retention intervals was comparable for all four conditions. Performance rate, i.e., the number of correctly completed sequences per 30 s, at learning was 12.96 +- 1.06 for the nocturnal sleep condition, 14.17 +- 0.8 for the nocturnal wake condition, 13.13 +- 0.72 for the daytime sleep condition, and 13.26 +- 0.58 for the daytime wake condition (P > 0.4, for pairwise comparisons). Also, error rates at learning did not differ among the conditions (nocturnal sleep: 7.68 +- 0.8; nocturnal wake: 6.51 +- 0.98; daytime sleep: 6.46 +- 0.65; daytime wake: 6.59 +- 0.98; P > 0.3 for pairwise comparisons). Table 1 | Sleep parameters Performance rates were generally improved at retesting after the 8-h retention interval, with this improvement strongly depending on sleep versus wakefulness during the retention interval [F(1,36) = 167.0, P < 0.001 for main effect before /after and F(1,36) = 31.81, P < 0.001 for before /after x sleep /wake ANOVA interaction; Fig. A]. Indeed, the benefit for performance rates was considerably stronger after retention intervals of sleep than that revealed for the corresponding nocturnal or daytime retention intervals of wakefulness [F(1,18) = 19.40, P < 0.001 and F(1,18) = 12.93, P < 0.002, for before /after x sleep /wake interaction, respectively, for the nocturnal and daytime retention intervals]. Performance rate improved on average from 13.04 +- 0.63 to 17.12 +- 0.64, i.e., 33.47% +- 3.73% across the sleep intervals, and from 13.72 +- 0.49 to 15.32 +- 0.52, i.e., 12.49% +- 2.79% across the wake retention intervals . The performance gain for retention intervals of sleep did not differ between nighttime (29.68% +- 3.72%) and daytime sleep (37.26% +- 6.45%; P > 0.3). However, improvements across wake retention intervals were stronger when placed during the day (17.56% +- 2.62%) than at night (7.41% +- 4.49%; P < 0.05, for separate pairwise comparisons). In fact, these analyses revealed that only performance gains across the daytime wake interval were significant (P < 0.001), whereas changes across the nocturnal wake interval were not (P > 0.2). Figure 2 | Performance gains on the finger-to-thumb opposition task are indicated by the difference between training and retrieval testing (A) for performance rate (mean number of correctly completed sequences per 30 s) and (B) error count (mean number of errors per 30 s). Performance gains on the finger-to-thumb opposition task are indicated by the difference between training and retrieval testing (A) for performance rate (mean number of correctly completed sequences per 30 s) and (B) error count (mean number of errors per 30 s). (Left) Mean (+- SEM) differences are indicated for 8-h retention intervals of sleep (black bars) and wakefulness (gray bars) placed during daytime and at night. (Right) In additional experiments, effects of a 48-h retention interval were tested which was filled either with two nights of regular sleep (black bars) or a first night of sleep deprivation followed by a night of recovery sleep (gray bars). *, P < 0.05. **, P < 0.001 for tests against zero and for differences between the effects of the retention intervals. Figure 3 | Progression of performance on the finger-to-thumb opposition task as indicated by the number of correctly completed sequences sampled at 30-s intervals. Progression of performance on the finger-to-thumb opposition task as indicated by the number of correctly completed sequences sampled at 30-s intervals. Three blocks each of 5 min duration were run before (Training) and after (Retrieval) 8-h retention intervals during which subjects slept (open circles), or remained awake (filled circles). Mean (+- SEM; adjusted to first block of training) are shown collapsed across both daytime and nighttime condition (refer to text). Performance rates also improved within the learning sessions, on average by 0.73 sequences per block (P < 0.001, for a comparison between first and third block). Thus, it could be argued that performance gains at retesting after the retention interval reflect a mere repetition effect. This was excluded in supplementary analyses taking into account the estimated repetition effect by linearly extrapolating the individual performance gains within the learning session. This analysis confirmed a significant improvement in performance rates after the retention intervals [F(1,36) = 57.31, P < 0.001 for main effect before /after], in particular when filled with sleep [F(1,18) = 16.54, P < 0.001 and F(1,18) = 8.68, P < 0.01, for before /after x sleep /wake ANOVA interaction for nocturnal and daytime retention intervals, respectively]. Also, separate analysis of the wake retention interval confirmed significance selectively for the daytime retention condition (P < 0.01). Accuracy of performance, as indicated by a decrease in the number of errors per 30 s, also improved only after retention intervals of sleep [on average by 30.07%, F(1,19) = 14.41, P < 0.001], but remained unchanged across retention intervals of wakefulness regardless of whether placed during the night or daytime [F(1,19) = 0.09, P > 0.8, Fig. B]. The improving effects of the sleep intervals on performance accuracy were closely comparable for the nocturnal and daytime sleep conditions (P > 0.9). For the retention sleep conditions, we also assessed the relationship between time spent in the different sleep stages and performance gains (across daytime and nighttime conditions) by using Pearson's correlation. Improvement in performance rate was proportional to the time spent in REM sleep (r = 0.61, P < 0.004). Correlation with time in slow wave sleep (SWS; r = 0.01, P > 0.9), stage 2 sleep (r = -0.37, P > 0.1) and stage 1 sleep (r = 0.14, P > 0.5) remained nonsignificant. Also, dividing sleep time into four quarters did not reveal evidence that performance gains were correlated with the amount of early SWS or late REM sleep, or with stage 2 sleep during any of these intervals. Comparing effects of sleep and wake retention intervals could be in principle confounded by unspecific effects of sleep loss on motor performance. Self-reports of mood and feelings of activation revealed that in comparison to normal nocturnal sleep, subjects after nights of sleep deprivation, felt more tired (P < 0.001), less activated (P < 0.01) and less concentrated (P < 0.05) as revealed by an adjective check-list . Nevertheless, a substantial contamination of motor performance by these subjective feelings seems unlikely in light of the fact that initial performance at learning was comparable for all retention conditions, regardless of whether this period was preceded by sleep or wakefulness. To further rule out unspecific motor effects of tiredness, particularly on retrieval tested after nocturnal wakefulness, two additional groups of eight naive subjects each practiced the finger motor sequences in the morning at 7:30 a.m. after a night of regular sleep and sleep deprivation. Performance on the two conditions was indeed closely comparable with regard to both performance rate (regular sleep: 13.34 +- 0.45; sleep deprivation: 13.38 +- 0.34; P > 0.9) and errors (regular sleep: 7.61 +- 0.96; sleep deprivation: 5.63 +- 0.6; P > 0.1). We also investigated whether the selective gain in performance after nocturnal sleep as compared with a retention period of nocturnal wakefulness is preserved after an additional night, in which sleep deprived subjects had recovery sleep. For this purpose, six additional subjects were trained, as in the main experiment, on one of the finger sequences, and retested after a 48-h retention interval that was filled either with two consecutive nights of regular sleep or with a first night of sleep deprivation followed by a second night of recovery sleep. Initial learning before the retention intervals did not differ between the sleep and sleep deprivation condition with respect to both, performance rates (P > 0.3), and error rates (P > 0.8). At retrieval testing 48 h later, the improvement in performance was distinctly more pronounced when subjects had slept the night after training (from 14.92 +- 1.28 to 18.94 +- 0.96, i.e., 28.99% +- 5.41%) than when they stayed awake in this first night [from 16.20 +- 1.89 to 17.92 +- 1.86, i.e., 11.19% +- 3.03%; F(1,5) = 45.98, P < 0.001, for before /after x sleep /wake interaction; Fig. ]. Errors also decreased only when subjects were retested after two regular nights of sleep (from 7.97 +- 2.28 to 7.0 +- 2.44, i.e., -15.78% +- 13.07%), and increased when subjects had stayed awake the night after training (from 7.61 +- 1.53 to 8.23 +- 1.74, i.e., 12.76% +- 18.96%) with these differences, however, not reaching significance. Finally, we tested the specificity of the improving effect of sleep on motor memories. Before an 8-h interval of nocturnal sleep, 10 other subjects were trained on one of the two finger sequences, as in the main experiment. At retrieval testing after the retention interval subjects were first tested on the untrained sequence and, 1 h later, on the sequence trained before the retention interval. As expected, performance improved across sleep only for the trained sequence (from 13.95 +- 1.1 to 17.52 +- 0.93, P < 0.001), whereas performance on the untrained sequence was similar to that at training before sleep (13.95 +- 1.1 versus 14.48 +- 1.17, P > 0.3). Discussion : Results indicate an improvement in finger motor skills that is distinctly greater and more consistent across retention periods of sleep than of wakefulness. Finger skills after a time of wakefulness were improved only with regard to performance rate, but not to error rate, and only when the retention period took place during daytime. Effects of tiredness do not explain our findings because performance on the task used here was shown to be unaffected by prior sleep deprivation. Moreover, memory for the trained motor sequence was still superior after sleep than after a vigil on the night immediately after training even when retrieval testing was postponed for another 24 h, including a night of sleep for all subjects. Because subjects in these experiments had sufficiently slept before retrieval on both conditions, fatigue and other factors induced by sleep deprivation can be safely ruled out as possible confounds. Importantly, this finding indicates that sleep enhances the formation of memory for the motor skill only within a critical time frame after training. Sleep occurring after training, rather than sleep before recall, appears to be effective. This conclusion complements results of a most recent study employing a similar tapping task . Donchin et al. failed to find an influence of sleep versus sleep deprivation on the memory for reaching movements. This outcome could point to an effect of sleep depending on the type of motor memory. Alternatively, sensitivity of different performance measures to the influence of sleep may vary, though this cannot be decided on the basis of the available data. The present result of a sleep-dependent enhancement in motor memories adds to previous evidence indicating a similar essential role of sleep in the formation of perceptual discrimination skills . Performance improvements observed during daytime wakefulness but not during nighttime wakefulness suggest both that the wake state per se is not sufficient to promote memory formation for the trained finger skill, and that circadian factors play a role in consolidation. Further evidence of a circadian influence comes from our observation that, across both the sleep and wake retention conditions, performance gains on average were greater for the retention intervals positioned during daytime than during the night. Moreover, it is conceivable that the circadian influence on the formation of motor memories during wake retention periods differs in quality from that during sleep. After daytime wake periods, the improvement in finger sequence tapping remained restricted to performance rate and was not paralleled by a decrease in errors, suggesting that only sleep leads to changes in internal representations improving motor accuracy. Also, previous studies indicated that performance on motor skills (reaching movements), though becoming more resistant to interference from similar behaviors , did not improve in accuracy across wake periods of 6 hours . Thus, consolidation during daytime wakefulness appears to spare certain aspects of the internal motor representations which are enhanced only by sleep. Skill learning is characterized essentially by rather discrete changes in low-level representations within the hierarchical organization of motor systems and, importantly, shows little generalization . Here, we found that the enhancing effect of sleep on memory for a motor sequence is highly specific with regard to the trained finger sequence and does not generalize to a control sequence containing the identical finger movements in a mirror-reversed order. This further excludes effects not specifically linked to the task, e.g., on motor fluency, but speaks for a direct influence of sleep on forming the internal model for this particular finger sequence. At the neuronal level, the consolidation of motor skill memories has been considered to involve a reorganization of motor representations residing predominantly in the primary motor cortex (M1) . Human studies using functional magnetic resonance imaging and transcranial magnetic stimulation have shown that the amount of motor training in our subjects is sufficient to trigger plastic neuronal changes in M1, whereby the initially fragile motor representations become increasingly stabilized . Moreover, findings with positron emission tomography suggested that consolidation of a serial motor task is based on a covert reactivation of brain structures already activated at training, with the signs of reactivation being most obvious during REM sleep . In the present study, exploratory calculation of correlation coefficients indicated greater performance gains in subjects with high amounts of REM sleep. Although, in light of the limited size of the subject sample, this result needs to be considered with caution, it would also point to a particular relevance of REM sleep in procedural memory formation. Nevertheless, the covert reprocessing of newly acquired motor representations during sleep is a concept which could explain the considerable performance gain in the finger motor sequence task seen after sleep. This gain, consisting of increased speed and a reduced number of false reactions, cannot arise from a nonselective strengthening of connectivity within the acquired representations, but implies a reorganization that enhances correct reactions to the exclusion of false ones. This reorganization, at the cellular level, probably involves processes such as synaptic long-term potentiation and depression , as well as synaptogenesis in the motor cortex, which may particularly benefit from the specific orchestration of neurotransmitters in the different sleep stages . However, the synaptic factors critically involved in sleep-dependent formation of skill memory remain to be identified. Demonstrating an essential role of sleep in the formation of memory for motor skills, our data extend previous observations of a similar function of sleep with regard to perceptual skills. In generalizing these observations to skills of everyday life (such as learning a musical instrument or sport), we would conclude that sleep is required to achieve optimum performance on any of these skills. Backmatter: Abbreviations : REM = rapid eye movement SWS = slow wave sleep PMID- 12370088 TI - MPN+, a putative catalytic motif found in a subset of MPN domain proteins from eukaryotes and prokaryotes, is critical for Rpn11 function AB - Abstract | Background | Three macromolecular assemblages, the lid complex of the proteasome, the COP9-Signalosome (CSN) and the eIF3 complex, all consist of multiple proteins harboring MPN and PCI domains. Up to now, no specific function for any of these proteins has been defined, nor has the importance of these motifs been elucidated. In particular Rpn11, a lid subunit, serves as the paradigm for MPN-containing proteins as it is highly conserved and important for proteasome function. Results | We have identified a sequence motif, termed the MPN+ motif, which is highly conserved in a subset of MPN domain proteins such as Rpn11 and Csn5/Jab1, but is not present outside of this subfamily. The MPN+ motif consists of five polar residues that resemble the active site residues of hydrolytic enzyme classes, particularly that of metalloproteases. By using site-directed mutagenesis, we show that the MPN+ residues are important for the function of Rpn11, while a highly conserved Cys residue outside of the MPN+ motif is not essential. Single amino acid substitutions in MPN+ residues all show similar phenotypes, including slow growth, sensitivity to temperature and amino acid analogs, and general proteasome-dependent proteolysis defects. Conclusions | The MPN+ motif is abundant in certain MPN-domain proteins, including newly identified proteins of eukaryotes, bacteria and archaea thought to act outside of the traditional large PCI/MPN complexes. The putative catalytic nature of the MPN+ motif makes it a good candidate for a pivotal enzymatic function, possibly a proteasome-associated deubiquitinating activity and a CSN-associated Nedd8/Rub1-removing activity. Keywords: Background : Many regulatory proteins are removed from the cell in a timely and specific manner by a large multi subunit enzyme called the proteasome . For proteins to be recognized by the proteasome, they are usually first covalently attached to a polyubiquitin chain via a cascade of ubiquitinating enzymes. This ubiquitination process is reversible. Specific cysteine proteases known as DUBs (deubiquitinating enzymes) can hydrolyze the amide bond between the Carboxy-terminus of ubiquitin and an amino group on the substrate . Proteolysis takes place within the 20S core particle (CP) of the proteasome, while the 19S regulatory particle (RP) binds polyubiquitinated substrates, unfolds, and translocates them into the 20S CP for proteolysis. The discovery that the 19S regulatory particle of the proteasome (RP) can be separated into two discrete subcomplexes, the lid and the base, suggests that they have distinct roles in preparing a substrate for degradation . The base contains six ATPase subunits, Rpt1-6, as well as the two largest non-ATPase subunits Rpn1 and Rpn2, and plays a role in anchoring the substrate, unfolding it and gating the channel leading into the 20S CP . The lid complex consists of eight non-ATPase subunits whose functions have not been defined. All subunits of the lid subcomplex contain one of two structural motifs: six contain a PCI domain (Proteasome, COP9, eIF3), while the other two (Rpn8 and Rpn11) contain an MPN domain (Mpr1, Pad1 N-terminal) . These domains are found in members of two other eukaryotic macromolecular assemblages as well: the COP9 signalosome (CSN) and the eukaryotic translation initiation factor 3 (eIF3). The functions of these domains are not known, but they are necessary for proper interactions between subunits of these complexes . The lid appears to be required for the degradation of polyubiquitinated substrates but not for hydrolysis of unstructured or short polypeptides . Thus, one possibility is that the lid is required in one way or another for proper interactions with polyubiquitinated chains. At 66% identity between the human and yeast forms, the MPN domain protein Rpn11 is the most highly conserved non-ATPase subunit of the 19S RP, on par only with the highly conserved ATPase subunits, suggesting that it too may play an enzymatic role within the RP . Mutations in RPN11 cause cell cycle and mitochondrial defects, temperature sensitivity, and sensitivity to DNA damaging reagents such as UV or MMS, underscoring the importance of this subunit in proteasome function . Rpn11 is one of a minority of proteasome subunits that exhibit dominant phenotypes upon overexpression. High dosage of human or S. pombe Rpn11 orthologs confer multidrug and UV resistance . These effects may be linked to the stabilization of c-Jun observed upon overexpression of the Rpn11 subunit in Schistosoma, SmPOH1 . In another case, however, overexpression of Rpn11 can suppress an srp1 mutation, the yeast homolog of importin alpha, and enhance degradation of a proteasome substrate , illustrating that effects of Rpn11 are pleiotropic. Together, these results may suggest that Rpn11 embodies an intrinsic enzymatic activity. In order to gain insight into the functions and evolutionary history of MPN domain proteins such as Rpn11, we performed extended database searches using the generalized profile method . An improved MPN domain profile detected a large number of novel significant matches including some prokaryotic proteins from archaeal and eubacterial origins, which until now have not been known to contain MPN domains. In addition to being structurally related to the published eukaryotic MPN domains, all prokaryotic MPN domains contain an additional pattern of five polar amino acids, which are conserved in a highly correlated fashion. This motif is embedded within the conventional MPN domain, and is also conserved in some -- but not all -- eukaryotic MPN proteins. We therefore term it the MPN+ motif. The polar nature and coordinated conservation of the MPN+ residues suggest a catalytic and/or metal-binding function. Since Rpn11 is one of the eukaryotic MPN domain proteins harboring the MPN+ motif, we used mutational analysis to assess the importance of these conserved amino acids for the function of Rpn11 in S. cerevisiae. Rpn11 also contains a highly conserved cysteine residue that is not part of the MPN+ motif but is present in a number of its close paralogs such as the COP9 signalosome subunit, Csn5. It has been suggested that this may correspond to the active site cysteine of a catalytic DUB motif . Therefore, we also analyzed mutants of Cys116, and compared them to mutations in the MPN+ residues of Rpn11. Results : Extending the scope of the MPN domain | In order to identify distantly related members of the MPN domain family, we constructed generalized profiles of previously established MPN proteins. Included in the profile construction were the proteasome lid components Rpn8 and Rpn11, the COP9 signalosome components Csn5 and Csn6, and the translation initiation factor 3 components eIF3f and eIF3h, all from various eukaryotic species. After scaling of the profile , significant hits were found in protein database, including the STAM-interacting protein AMSH , and the uncharacterized human proteins C6.1A and KIAA1915. Subsequently, iterative profile refinement was used to make the profile searches more sensitive. For that purpose, the newly identified MPN proteins were included in the profile construction process. After four iteration cycles, a stable set of significantly matching bona fide MPN proteins was identified. Representative members of this superfamily are shown in Figure . Figure 1 | The MPN+ motif. The MPN+ motif. Sequence alignment of representative MPN domain protein, only two conserved blocks containing the MPN+ motif are shown. The proteins are grouped in seven categories: A, bacterial; B, archaeal; C, proteasome lid components; D, CSN components; E, eIF3 components; F, Prp8-like proteins; G other eukaryotic MPN proteins. In the eukaryotic groups, representative sequences from human (HS), Arabidopsis (AT), S. pombe (SP) and S. cerevisiae (SC) are shown. Prokaryotic species shown are lambda phage (BPL), Yersinia pestis (YP), Synechocystis sp. (SS), Mycobacterium tuberculosum (MyT), Pseudomonas aeruginosa (PA), Pyrococcus horikoshii (PH), Archaeoglobus fulgidus (AF), Methanobacter thermoautotrophicum (MT). Residues invariant or conservatively substituted in at least 50% of all sequences are shown on black and grey background, respectively. The MPN+ motif residues are shown in red. The top line indicates the amino acids constituting the MPN+ motif; the rightmost column indicates whether a protein is considered an MPN+ protein or a plain MPN protein. In addition to the previously known MPN proteins, a number of prokaryotic members were identified, including the phage tail assembly protein K from the bacteriophage lambda and its closely related homologs from other phages and prophages. Even archaeal members of the MPN domain family were found, including predicted proteins from Archaeoglobus fulgidus, Methanobacter thermoautotrophicum and from various Pyrococcus species . These proteins are the shortest MPN protein identified so far and most likely correspond to the structural core region of the domain. Since the finding of prokaryotic MPN proteins was highly unexpected, the validity of the assignment was confirmed by profile searches starting from the bacteriophage proteins that resulted in the same stable set of significantly matching proteins. Definition of the MPN+ motif | Overall, there are no residues that are invariant throughout the MPN domain superfamily. However, while analyzing the alignment of the newly identified proteins, it became apparent that there are a number of polar residues that are conserved in a highly coordinated fashion in a subset of MPN domains (highlighted in Red in Fig. ). These amino acids form a pattern, referred to herein as the 'MPN+ motif', which is part of the conventional MPN domain and embedded within it. The MPN+ motif contains a well-defined pattern of 'H-x-H-x-S-x-D', where x and x indicate stretches of seven and two non-conserved residues, respectively. In addition to this conserved arrangement of four polar residues, there is an additional glutamate residue in a more N-terminal region of the domain, whose conservation is perfectly correlated with the occurrence of the motif . An aromatic residue, preferentially a tryptophane, is found two positions upstream of the conserved serine in most but not all MPN+ proteins. Thus, it should not be considered a part of the core MPN+ motif. With the possible exception of Prp8, all MPN domain proteins shown in Figure can unambiguously be classified as either belonging to the MPN+ or to the 'plain' MPN class. The validity of this distinction is underscored by the fact that invariably all observed orthologs of MPN+ proteins also belong to the MPN+ class. For instance, the proteasome lid complex from all eukaryotes contains one MPN+ protein (Rpn11/S13) and one plain MPN protein (Rpn8/S12). The same is true for the analogous CSN complex in multicellular eukaryotes, where the MPN+ protein is Csn5 and the plain MPN protein is Csn6. In fission yeast, Csn6 appears to be absent. The same is true for the recently identified CSN-like complex of budding yeast; it contains only Csn5 but not Csn6 (Hofmann and Glickman; submitted for publication). The two MPN proteins of the eIF3 complex are both of the plain MPN type in multicellular eukaryotes and fission yeast, while in budding yeast they appear to be missing altogether. Notably, all newly identified prokaryotic MPN proteins are of the MPN+ type; the same is true for the remaining unassigned 'orphaned' MPN proteins in eukaryotes . It should be stressed that the MPN+ motif is not a stand-alone motif to be found independently of the MPN domain; thus, all MPN proteins can be classified as either MPN+ or plain MPN proteins. Mutants in the MPN+ motif of Rpn11 exhibit severe growth defects | In order to assess the importance of the MPN+ residues for efficient Rpn11 function, we performed site-directed mutagenesis of His111, Ser119, Asp122, which are part of the conserved HxH --S --D sequence that defines the MPN+ motif. In addition, Cys116, which is not part of the MPN+ motif but is highly conserved between Rpn11 and Csn5 orthologs might be a candidate for an active site residue, was mutagenized too. As seen in Fig. , alanine substitutions in the MPN+ motif of Rpn11 cause severe growth defects, temperature sensitivity, and sensitivity to amino acid analogs such as canavanine. Substitution mutations at different locations in the MPN+ motif display similar, but slightly distinct phenotypes. Specifically, the His111Ala mutant is extremely slow growing at 25C and lethal when plated at elevated temperatures, even as low as 30C. The Asp122Ala and Ser119Ala mutants are viable but slow growing at 25C and 30C, and lethal when shifted to 37C or when exposed to amino acid analogs. By contrast, substitutions of the conserved Cys at position 116 showed no growth defect, and the cells appeared normal under all conditions tested. Figure 2 | Growth of rpn11 Mutants under Different Growth Conditions. Growth of rpn11 Mutants under Different Growth Conditions. WT and mutant strains were streaked on YPD at 25, 30 and 37C. Cells were also plated on complete minimal media containing 1 mug/ml of the amino acid analog canavinine instead of arginine. Plates were photographed after 3 --5 days. Three single amino acid substitutions in the MPN+ motif of Rpn11 were studied: his111ala, Ser119ala, and asp122ala. For comparison, a substitution of a highly conserved cysteine residue (cys116ala) that is not part of the MPN+ motif (see Fig. ) was included as well. At 25C, all MPN+ mutations are viable but slow growing. Hiss111 is extremely temperature sensitive, showing no appreciable growth even when shifted to 30C, or in the presence of even 1 mug/ml of the amino acid analog canavinine. Asp122 and Ser119 substitutions show lethality under elevated temperature (37C) or exposure to canavanine. In comparison, Cys116 substitutions show no growth defects under these conditions. Stress conditions such as exposure to elevated temperature or amino acid analogs are known to promote accumulation of damaged proteins, which must be removed by the proteasome . Since defects in the MPN+ motif of Rpn11 cause heightened sensitivity to such conditions, it is reasonable to assume that proteasome function is jeopardized in these mutants. We conclude that the MPN+ motif is critical for the proper function of Rpn11 within the proteasome, but that Cys116 does not appear to be an essential residue in Rpn11. Multiubiquitinated proteins accumulate in MPN+ motif mutants | In order to test whether proteasome function is indeed hampered in these rpn11 mutants, we checked the effects of mutations on the ubiquitination pattern of cellular proteins. Whole cell extracts from rapidly growing WT or MPN+ motif mutants were separated by SDS PAGE and immunoblotted with anti-Ub antibody . High molecular weight polyubiquitin-conjugates are not detected at appreciable levels in extracts from yeast containing normal proteasomes since they are rapidly turned over. High molecular weight polyubiquitinated proteins do accumulate, however, in rpn11 MPN+ mutants, indicating defective proteasome activity. In this assay, all mutations in MPN+ residues behave similarly. His111 mutants also accumulate polyubiquitinated proteins when grown at 25C, but since this strain is extremely slow growing it was difficult to prepare extracts for comparison (not shown). No such accumulation is seen in cells containing Rpn11 with Cys116 substitutions. Figure 3 | Accumulation of multiubiquitinated proteins in MPN motif mutants. Accumulation of multiubiquitinated proteins in MPN motif mutants. Total cell extracts from WT yeast and from strains containing the MPN+ motif substitutions were separated on an 8% SDS gel and blotted with anti-ubiquitin antibodies. Cells were rapidly lysed in presence of 12% TCA in order to inhibit post lysis enzymatic activity. Accumulation of high MW polyubiquitinated proteins is detected in the Ser119 and Asp122 mutants, but not in WT or in the Cys116 mutants. His111 mutants accumulate polyubiquitinated proteins as well (not shown). A protein band migrating at around 20 kDa that is detected with the anti Ub antibody is used as an internal loading control. It appears that the MPN+ motif defines the role of Rpn11 in the proteasome but that the conserved Cys116 residue is not critical for proteasome function. The steady state levels of polyubiquitin-protein conjugates are influenced by the rate of ubiquitination on the one hand, and by rates of deubiquitination or proteasome proteolysis on the other. Accumulation of polyubiquitinated proteins in rpn11 mutants could therefore be due to a slowdown in either proteasome associated deubiquitination or proteasome-dependent proteolysis. Stabilization of short-lived proteasome substrate | In order to test whether the accumulation of polyubiquitinated conjugates is directly correlated with a defect in proteasome function, we checked whether mutations in the MPN+ motif of Rpn11 bring about stabilization of proteasome substrates. The steady state levels of known short-lived proteins were measured in WT and in a representative rpn11 mutant strain. In order to estimate the generality of the effect, two different substrates were used: a protein that is ubiquitinated by enzymes of the UFD ubiquitination pathway , and a protein that is ubiquitinated by enzymes of the N-end rule pathway . WT Cells, or those harboring the S119A substitution in the MPN+ motif of Rpn11 were transformed with plasmids expressing Arg-beta-galactosidase (an N-end rule substrate) or Ub-pro-beta-galactosidase (A UFD substrate). Arg-beta-gal and Ub-Pro-beta-gal are short-lived in wild-type yeast with half-lives of ~2 and ~6 minutes, respectively . Steady state levels of these proteins were compared to the levels of a stable, long-lived protein, Met-beta-galactosidase. As expected, WT cells accumulated high levels of the stable Met-betagal but only low levels of the rapidly degraded Arg-beta-gal and Ub-Pro-beta-gal . In contrast, the S119A mutation in rpn11 lead to dramatic stabilization and accumulation of both these short-lived proteins, such that the steady state levels of all three substrates were similar . Stabilization of these short lived proteins was noted also in the D122A mutant (not shown). From these results we conclude that the MPN+ motif of Rpn11 is essential for proper proteolysis of ubiquitinated substrates by the proteasome. The importance of Rpn11 is independent of the ubiquitination pathway. Figure 4 | Stabilization of short-lived proteasome substrates. Stabilization of short-lived proteasome substrates. WT and the S119A mutant strains containing the Ub-Met-, Ub-Arg-, and Ub-Pro-betagal constructs on multi-copy plasmids under the GAL promotor were tested for LacZ activity after galactose-induction. LacZ activity is indicative of steady-state levels of the reporter protein. WT cells accumulate high levels of the stable Met-betagal, but rapidly degrade Arg-betagal and UB-Pro-betagal (left panel), whereas the MPN+ mutation shows dramatic stabilization and accumulation of both short-lived fusion proteins (right panel). The possibility that these deficiencies in proteasome dependent proteolysis were caused by improper incorporation of mutated Rpn11 into the lid, or that proteasome structure was hampered in rpn11 mutants was addressed by native gel electrophoresis. Cell extract from WT and rpn11 mutants was resolved by nondenaturing PAGE and no gross structural changes were observed . Overall levels and proteolytic intensities of proteasomes from MPN+ mutants were indistinguishable from WT, indicating that the substitutions do not alter the structure of peptidase function of the proteasome. In addition, we found no evidence for natural abundance WT Rpn11 outside of the proteasome (data not shown), indicating that the phenotypes associated with mutated rpn11 are unlikely to be due to unincorporated protein. Since peptidase activity of mutant proteasomes was similar to WT , the effects of rpn11 mutations is most likely due to a defect in proteolysis of ubiquitinated substrates . Figure 5 | Migration pattern of proteasome from WT and rpn11 mutants by non-denaturing PAGE. Migration pattern of proteasome from WT and rpn11 mutants by non-denaturing PAGE. Cell extracts were prepared from Logarithmically growing WT or the rpn11 MPN+ mutants that were brought to an identical OD600 level. Extracts were clarified by centrifugation, and identical amounts of total protein were separated on native gel. The gel was then incubated for 10 minutes with the fluorescent peptide LLVY-AMC at 30C, and photographed under UV light (380/440 nm). Fluorescent bands indicate migration of Doubly capped and Singly capped 26S proteasome forms . There is no noticeable change in migration pattern or overall levels of proteasomes from the different rpn11 mutants, indicating that mutated Rpn11 is properly incorporated into the lid and no gross structural changes in proteasome composition or amounts are due to the mutations. Discussion : We have identified an arrangement of five residues, which are perfectly conserved in a subclass of MPN domains, while none of them shows appreciable conservation outside of this subclass. The highly correlated conservation of these residues suggests that they participate in a common structural element and/or a common function that is critical for the proteins in which they are found. This idea is corroborated by the polar nature of the conserved MPN+ residues: glutamate, histidines, serine and aspartate are all amino acids frequently found in the active site of enzymes or as the coordinating ligands in metal-binding proteins. Certainly, these two possibilities are not mutually exclusive. Several classes of enzymes, particularly metal containing hydrolases and proteases, harbor bound metal ions such as Zn2+ as part of their catalytic center . However, while the motif does bear some resemblance to that found in metalloproteases, the specific organization of E --HxH --S --D residues does not correspond to the metal ligands in any of the known Zinc proteases , thus positive identification of Rpn11 as a metalloprotease awaits definite proof. The proteasome complex from a number of sources is known to contain a ubiquitin hydrolyzing activity , while the CSN complex is known to cause the hydrolytic removal of the Nedd8/Rub1 ubiquitin-like molecule from the cullin subunit of the SCF ubiquitin ligase (E3) complex . Thus, the MPN+ containing subunits of the proteasome lid (Rpn11) and the COP9 signalosome (Csn5) would be prime candidates for such a hydrolytic function. The architecturally related eIF3 complex, for which no such enzymatic activity has been described, is conspicuously devoid of MPN+ proteins. That the CSN from yeast lacks the plain MPN protein (Csn6) retaining only the MPN+ subunit (Csn5), while the plain MPN proteins found in eIF3 from eukaryotes appear to be missing altogether in the yeast complex, emphasizes that the MPN+ residues are likely to be the catalytic residues, with plain MPN subunit playing a redundant structural role in complexes in which they are found. Even though purified recombinant Rpn11 does not appear to exhibit DUB capabilities (data not shown), it is possible that once incorporated into their respective complexes, Rpn11 and Csn5 confer the documented hydrolase activities onto the 19S RP and CSN. In this case, Rpn11 would belong to a unique class of enzymes, as all other known DUBs are cysteine proteases. Of note, the conserved cysteine residue common to both Rpn11 and Csn5, which is not part of the MPN+ motif, is not important for the function of Rpn11. Single site substitutions in various Rpn11 MPN+ residues exhibit similar phenotypes supporting the identification of MPN+ as a discrete functional motif. Due to the severe growth phenotypes and attenuated ability of the proteasome to proteolyze polyubiquitinated substrates in these mutants, it appears that the MPN+ motif defines the role carried out by Rpn11 in the lid. As we show, these defects arise from an intrinsic activity of Rpn11 within the context of the proteasome, and not due to a gross structural effect upon incorporation of mutated Rpn11. As mentioned above, the lid, where Rpn11 is situated, is critical for proteolysis of polyubiquitinated substrates. So far, and somewhat surprisingly, all ubiquitin binding activity has been mapped to the Base of the 19S RP. Two subunits in the base can interact directly with ubiquitin chains, Rpt5 and Rpn10 . A number of proteins, such as Rad23 and Dsk2, can also bind ubiquitin and interact with the proteasome, presumably with the base, thus they are thought to serve as shuttles of polyubiquitinated substrates to the proteasome . It is possible that the lid rather than bind ubiquitin, serves to cleave or trim polyubiquitin chains once attached to the Base. While this manuscript was under review, two independent papers substantiated our findings by characterizing a novel deubiquitinating activity in the proteasome, and showing that the MPN+ motif residues of Rpn11 are largely responsible for this activity . These independent studies identified DUB activity associated with the 19S RP of the proteasome, which is lacking upon lethal substitutions in Rpn11. Interestingly, all substitution mutations that we studied in the MPN+ motif of Rpn11 are viable. Quite possibly, Rpn11 is not the sole proteasome-associated DUB, and a number of DUBs play partially overlapping functions. For instance, Doa4/Ubp4 interacts weakly and substoichiometrically with the proteasome and may serve to release ubiquitin and regenerate the proteasome for the next catalytic cycle . The Ubiquitin-like domain (UBL) containing deubiquitinating enzyme, USP14, has been found to interact with the proteasome from mammalian sources . Ubp6, the budding yeast homolog of USP14, interacts with the proteasome as well , and plays a role in proteasome-associated deubiquitination . It has also been reported that the Drosophila DUB p37a and its homologs UCH37 (H. sapiens) and Uch2 (S. pombe), may be responsible for the polyubiquitin chain editing function associated with purified proteasomes . However, as budding yeast lacks an obvious ortholog of UCH37/p37a, other DUBs must play a greater role in proteasomes from this organism. Finding MPN+ motif proteins in prokaryotes will help elucidate the origins of proteasome evolution and the function of Rpn11 in particular. The proteasome probably evolved from self-compartmentalized macromolecular proteases found in prokaryotes. Thus, proteasomal subunits, or proteins with motifs related to proteasomal subunits, are present in archaea and certain eubacteria. These "lidless" prototypes include the 20S CP subunits, and homologs of the base ATPases (Rpt subunits) . However, proteins relating to the ubiquitination process, such as ubiquitin itself, ubiquitin activating conjugating or ligating enzymes, and even lid subunits, are missing from prokaryotes. Identification of a motif from Rpn11 -- a protein that is linked to processing of polyubiquitinated proteins -- in prokaryotic proteins is an interesting development. It is possible that during evolution, the proteasome recruited an existing enzyme as the lid was forming into a regulatory module of the proteasome; likewise for the analogous CSN complex. Studying the prokaryotic MPN+ motif proteins should aid in elucidating what this motif does. The fact that these proteins are the shortest MPN proteins and correspond to the structural core region of the domain should greatly aid in enzymatic and structural studies, especially in comparison to members of the family such as Rpn11 and Csn5 that are naturally found only incorporated into complexes. Interestingly, all identified prokaryotic MPN domains, both from archaea and from eubacteria, also contain the MPN+ motif. It thus appears likely that the ancestral MPN domain was of the MPN+ type, and that the extant 'plain' MPN proteins have lost this motif later on. It is possible that the plain MPN proteins play a structural role in complexes that additionally contain a MPN+ subunit. So far all known MPN proteins are incorporated into complexes that also contain PCI proteins . Interestingly, several of the prokaryotic organisms harboring MPN+ proteins appear to lack PCI domain proteins and therefore probably could not form large PCI/MPN complexes related to the lid and CSN particles. An interesting case is the tail assembly protein 'K' of the bacteriophage lambda (vtak) and its many homologs from other phages and prophages . Little is known about the specific function of the protein but none of the other tail assembly factors contain a PCI or MPN domain (K.H unpublished results). Our work suggests that the K-protein plays an enzymatic role in tail assembly rather than a merely structural one, although whether the analogy to Rpn11 is relevant to vtak is still enigmatic. A similar situation exists for the orphan MPN proteins of eukaryotes. It is noteworthy that complex eukaryotic organisms have far more orphan MPN protein than orphan PCI proteins. However, the known PCI/MPN complex particles have stoichiometries that require more PCI components than MPN components. This apparent paradoxon can be resolved by a number of different but non-exclusive assumptions. i) Additional PCI proteins have eluded detection due to high sequence divergence allowing for more than three PCI/MPN complexes to exist, ii) alternative MPN proteins can interact with existing PCI complexes, iii) some MPN proteins have (enzymatic) functions outside of PCI/MPN complexes. Currently, there is no data available supporting the first two possibilities, although there is some promiscuity in the PCI protein interactions . The third possibility is likely to be true in prokaryotes and could also explain why all eukaryotic orphan MPN proteins belong to the MPN+ class. This could also explain how supra-stochiometric amounts of Rpn11 exhibit dominant phenotypes; possibly it can function outside of the proteasome as well. Conclusions : The MPN+ motif is abundant in certain MPN-domain proteins, including newly identified proteins of eukaryotes, bacteria and archaea thought to act outside of the traditional large PCI/MPN complexes. The putative catalytic nature of the MPN+ motif makes it a good candidate for a pivotal enzymatic function, possibly a proteasome-associated deubiquitinating activity and a CSN-associated Nedd8/Rub1-removing activity. The importance of the MPN+ motif for the efficient function of the proteasome component Rpn11 is compatible with this idea, though the fact that none of the single amino acid substitutions are lethal indicates that either the function of Rpn11 is not strictly essential for proteasome function, or more likely, it is partially redundant with other subunits. Methods : Bioinformatics | All database searches were performed with a nonredundant data set constructed from current releases of SwissProt, TrEMBL, and GenPept . Generalized profile construction and searches were run locally using the pftools package, version 2.1. (program available from the URL ). Profiles were constructed using the BLOSUM45 substitution matrix and default penalties of 2.1 for gap opening and 0.2 for gap extension. The statistical significance of profile matches was derived from the analysis of the score distribution of a randomized database as described . Database randomization was performed by individually inverting each protein sequence, using SwissProt 34 as the data source. Single amino acid substitutions | Haploid yeast strain with a chromosomal knock out of rpn11 was constructed, in which a single copy URA-marked plasmid expressing the RPN11 gene complements the chromosomal knock out (MY71). The heterozygote RPN11/rpn11Delta diploid was purchased from EUROSCAF and transformed with a single copy CEN plasmid with URA3 selection (ycplac33) expressing RPN11 from its own promotor (M82). Growth of MY71 was identical to the isogenic WT strain from EUROSCARF. Plasmids expressing the single site substitutions in Rpn11 were generated using PCR site directed mutagenesis on a similar CEN plasmid with the LEU2 marker for selection (ycplac111). In this manner the following plasmids were constructed RPN11 (M134), rpn11-C116A (M134), rpn11-H111A (M143), rpn11-S119A (M144), rpn11-D122A (M145). Those plasmids were then transformed into the above yeast strains, and the URA3-marked WT rescue plasmid was then forced out of the cells by the presence of 5'FOA in the medium. FOA shuffling was done at 25C as most rpn11 mutants were severely temperature sensitive. In this manner we got viable yeast rpn11 mutant strains. Phenotypes and mutant characterization | Single colonies grown on YPD at 25C were streaked onto YPD and shifted to various temperatures. Plates were photographed after 3 --5 days. For canavinine sensitivity, plates containing complete minimal media containing 1 mg/ml canavinine in replace of arginine were used and growth measured at 25C. Stabilization of known proteasome substrates | Steady-state levels of the Met-, Arg-, and Ub-Pro-betagal fusion proteins were measured by testing LacZ activity . Wild type and rpn11 cells harboring multi-copy URA3-marked plasmids containing each construct were grown to late log in minimal media containing raffinose for carbon source, and induced with 2% Galactose for 4 hrs. The cells were then lysed and LacZ activity upon introduction of the substrate ONPG was calculated when taking into account the time, total protein concentration and amount of product obtained (measured by 405 nm absorbance). Non-denaturing PAGE and peptidase activity detection | Yeast cells were lysed in buffer A (10% glycerol, 50 mM tris7.4, 1 mM ATP, 1 mM MgCl2, 1 mM EDTA) using glass beads, and clarified by centrifugation at 20 x g. Non-denaturing gel was run as previously described . The gel was then incubated for 10 minutes in 10 ml of buffer A and 0.1 mM of the fluorescent peptide LLVY-AMC at 30C. Under UV light, appearing bands show peptidase activity indicate the migration pattern of the 26S proteasome. Abbreviations : RP: 19S regulatory particle of the proteasome CP: 20S core particle of the proteasome CSN: COP9 signalosome eIF3: eukaryotic initiation of translation factor 3 Rpn: regulatory particle non-ATPase subunit Rpt: Regulatory particle triple-A ATPase subunit Ub: ubiquitin DUB: deubiquitinating enzyme beta-gal: beta-galactosidase Authors' contributions : Kay Hofmann and Michael Glickman initiated the project and participated in its design. Kay Hofmann performed bioinformatic characterization of the MPN+ motif (fig. ). Noa Reis and Vered Maytal designed and constructed the MPN+ amino acid substitutions. Vered Maytal carried out all characterization of Rpn11 mutants shown in Figures ,,,. All authors read and approved the manuscript. Backmatter: PMID- 12361483 TI - The PEST sequence does not contribute to the stability of the cystic fibrosis transmembrane conductance regulator AB - Abstract | Background | Endoplasmic reticulum retention of misfolded cystic fibrosis transmembrane conductance regulator (CFTR) mutants and their rapid degradation is the major cause of cystic fibrosis (CF). An important goal is to understand the mechanism of how the misfolded proteins are recognized, retained, and targeted for degradation. Results | Using a web-based algorithm, PESTFind, we found a PEST sequence in the regulatory (R) domain of CFTR. The PEST sequence is found in many short-lived eukaryotic proteins and plays a role in their degradation. To determine its role in the stability and degradation of misprocessed CFTR, we introduced a number of site-directed mutations into the PEST sequence in the cDNA of DeltaF508 CFTR, the most prevalent misprocessed mutation found in CF patients. Analysis of these mutants showed that the disruption of the PEST sequence plays a minor role in the degradation of the CFTR mutants. Multiple mutations to the PEST sequence within the R domain of CFTR inhibit maturation of CFTR and prevent the formation of a 100 kDa degradation product. The mutations, however, do not improve the stability of the mutant DeltaF508 CFTR. Conclusion | These observations show that disruption of the structure of the R domain of CFTR can inhibit maturation of the protein and that the predicted PEST sequence plays no significant role in the degradation of CFTR. Keywords: Background : Mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene cause cystic fibrosis (CF), the most prevalent fatal recessive genetic disease in the Caucasian population . CFTR is a polytopic integral membrane protein synthesized in the endoplasmic reticulum (ER) and normally expressed on the apical surface of epithelial cells where it functions as a phosphorylation-stimulated and ATP-dependent chloride channel. The majority of CF patients express processing defective CFTRs that fail to mature to the cell surface; instead, the processing defective CFTRs are retained in the ER and are targeted for rapid degradation . The retention of processing defective CFTR is a response of the ER quality control system to misfolded proteins, which prevents the progression of misfolded or misassembled membrane and secretory proteins into later compartments of the secretory pathway . During synthesis, nascent CFTR polypeptide chains are translated from ER membrane-bound ribosomes and are inserted into the ER membrane . Various classes of chaperones associate with the nascent polypeptide both in the lumen of the ER and in the cytosol to aid in folding . Upon proper folding, the properly folded CFTR dissociate from the chaperones and are packaged into transport vesicles for export to a post-ER compartment in the secretory pathway, the Golgi. Many of the missense mutations in CFTR retard the folding process. This leads to prolonged association of the nascent chains with the molecular chaperones and prevents the nascent chains from exiting the ER through the default secretory pathway. Instead, the misfolded polypeptides are retrotranslocated across the ER membrane, into the cytosol, and targeted for degradation by the ubiquitin-proteasome pathway . Although much of the molecular mechanism of the ubiquitin-proteasome system has now been elucidated (reviewed in ), the precise mechanism and determinants of recognition of the misfolded polypeptides remain unclear . As proposed by Chang et al. , the retention of misfolded CFTR is most likely due to the exposure of short sequence motifs specifically recognized by components of the ER quality control system or vesicular transport system; the mutations may cause localized misfolding leading to global misfolding to expose or bury motifs that signal for degradation, retention or exportation from the ER. Indeed, it has been shown that the removal of multiple arginine-framed ER retention/retrieval trafficking signals overcomes misprocessing of DeltaF508 CFTR, the most prevalent processing defective CF mutation . Furthermore, attempts to promote maturation of the processing defective mutants by shutting down the cytosolic proteasomes via proteasome inhibitors have led to the speculation of the existence of other systems responsible for the retention and degradation of these processing defective CFTR . Treatment of cells expressing wild-type (WT) CFTR with MG-132, an inhibitor of the 26S proteasome in the ubiquitin-proteasome pathway, leads to inhibition of maturation of the CFTR polypeptide . The resulting maturation-hindered WT CFTR polypeptide exhibit similar stability, structural, and functional properties to misprocessed CFTR mutants such as the prevalent DeltaF508 CFTR . PEST sequences are found in many rapidly degraded proteins. These sequences have been suggested to serve as signals for proteolytic degradation. From a survey of the amino acid sequences of 10 short-lived eukaryotic proteins, Rogers et al. found the proteins to contain one or more regions rich in proline (P), glutamic acid (E), serine (S), and threonine (T). These regions are often flanked by positively charged amino acids. They named these regions PEST regions. Based on their observation, they developed an algorithm that would search for such regions in a given protein sequence. In the present study, we examine the role of a PEST sequence, found in CFTR, in the stability and degradation of CFTR. Results : PEST sequence of CFTR | The algorithm, PESTFind, searches for hydrophilic regions of 12 or greater amino acids that contain at least one P (proline), one E (glutamic acid) or D (aspartic acid), and one S (serine) or T (threonine), flanked by K (lysine), R (arginine), or H (histidine) residues. The algorithm assigns a score to each possible PEST sequence found. The score ranges from -50 to +50, with a score above zero denoting a possible PEST region while a value greater than + 5 being of particular interest. Using the algorithm, PESTFind , we found a highly conserved PEST region in the regulatory domain of the CFTR protein (figure ). The CFTR PEST region (residues 716 -- 734) scored + 6.91 and had a hydrophobicity index of 32.63. Within the 19 residues PEST region, 8 are charged (i.e. 1 lysine, 1 arginine, 4 glutamic acid, and 2 aspartic acid), 4 polar (i.e. 1 threonine, 1 serine, 1 glutamine, and 1 asparagine), and 7 non-polar (i.e. 2 proline, 2 leucine, 1 methionine, 1 glycine, and 1 isoleucine). The low hydrophobicity suggests the region may be surface accessible to proteases or for protein-protein interaction with other proteins such as molecular chaperones, trafficking proteins, or components of proteolytic systems. Figure 1 | Alignment of the PEST region of CFTR Alignment of the PEST region of CFTR. The top line is the human PEST sequence used in the study. A cross-species alignment of the CFTR amino acid sequences from residues 716 to 734 of the human CFTR shows great conservation in the region. Discrepancies with the human CFTR are noted in italics. Significant PEST residues are noted with asteriks. The PEST region is in the Regulatory (R) domain of CFTR. Partial sequences were found using the PubMed . Accession number and references for each species are as follows: Human (P13569; ), Rhesus (AAC14012; ), Macaque (AAF80467; unpublished; direct submission), Baboon (AAD46907; unpublished; direct submission), Rabbit (Q00554; ), Sheep (Q00555; ), Bovine (P35071; ), Rat (1901178A; unpublished; direct submission), Mouse (P26361; ). The lower panel shows a model of CFTR and location of the predicted PEST region in the R-domain (R). The cylinders represent predicted transmembrane domains and the squares represent the nucleotide-binding domains (NBD1 and NBD2). Construction of PEST mutants | To examine whether the PEST sequence directly affects the processing or proteolytic degradation of CFTR, various CFTR PEST disrupting mutants were constructed into the DeltaF508 CFTR cDNA using site-directed mutagenesis (table ). An A52 epitope was also appended at the COOH-terminus for detecting transfected CFTR as opposed to endogenous CFTR. Table lists the mutants with their sequences and PEST scores as determined by the algorithm, PESTFind. The mutants' PEST scores range from being poor PEST candidates (+4.07 for S728A and -26.19 for poly-valine) to being invalid PEST candidates (i.e. no score available). We were particularly interested in the stretch of polar and PEST-significant region, residues 725 to 731. For that region, we made a poly-valine mutant where we mutated the 6 consecutive polar/charged residues (725 -- 730) to valine, a relatively small, neutral residue. This minimized the hydrophobicity of the PEST region and removed the majority of the significant PEST residues; the proline residue 731 was not altered because it may be important structurally (i.e. for formation of a beta-turn). The E725K/E726K mutant was included to test if alteration of charges would affect the function of the PEST sequence. Since serine and threonine are often implicated in functional group modification such as phosphorylation, their role in the PEST sequence was examined. S728 and T717 were both altered to alanine, a small, neutral residue; S728A has PEST score of +4.07. However, since both S728A and T717A mutant have a high PEST score (T717A mutant alone has PEST score of +3.87), we added an additional E725K mutation to the T717A mutant just to further disrupt the PEST sequence. The mutants were also made in WT background. The WT-background constructs were included to ensure that the mutations themselves did not contribute to additional deleterious effects on the protein. Table 1 | Mutations introduced into the predicted PEST sequence of CFTR Expression of PEST mutants | The cDNAs encoding the PEST mutants were transiently transfected into COS-1 cells. The transfected cells were harvested and lysed 3 days post-transfection, separated by electrophoresis and subjected to Western blotting using a mouse monoclonal antibody against the A52 tag attached at the COOH-terminus of the protein. The two N-glycosylation sites on the putative extracellular loop 4 of CFTR (figure ) enable us to determine the maturation state of CFTR. During its synthesis in the ER, two carbohydrate moieties are covalently attached onto the two N-glycosylation sites; this is the immature, core-glycosylated CFTR. The carbohydrate moieties are modified only upon the nascent polypeptide's maturation and progression through the Golgi apparatus onto the cell surface; this is the mature, fully-glycosylated CFTR. The modification of the carbohydrate moieties lead to a shift in the protein's apparent molecular mass: the core-glycosylated ER-resident CFTR has an apparent molecular mass of ~140 kDa while the fully-glycosylated CFTR expressed on the cell surface has an apparent molecular weight of ~170 kDa . While glycosylation of CFTR is not crucial in the maturation CFTR, it serves as an indicator of CFTR's maturation state . As shown in figure , the mutants had no apparent effect on the maturation of CFTR with the exception of the poly-valine mutant. The WT background constructs, except the poly-valine mutant, all have similar levels of expression and maturation as the WT CFTR. The poly-valine construct in WT background is processing defective, similar to the DeltaF508 mutant. All of the DeltaF508 background mutants had no apparent effect on the maturation state of DeltaF508 CFTR and are all immature. An over-exposure of the immunoblot (figure ), however, suggests removal of a protease site by the poly-valine mutations. Since we do not see any larger fragments (>100 kDa) in the poly-valine mutants, it is also possible that the poly-valine mutations introduced additional cleavage sites in the R domain. As shown in figure , the 100 kDa degradation product that is present in DeltaF508 is absent in the poly-valine constructs, both of the WT and DeltaF508 background. There is, however, no improvement in maturation or in the level of expression. Figure 2 | Expression of PEST mutants in COS-1 Cells Expression of PEST Mutants in COS-1 Cells. A) COS-1 cells were transiently transfected with cDNA encoding the various PEST mutants in both WT and DeltaF508 background. Whole cell extracts were subjected to SDS-PAGE and immunoblot analysis (monoclonal antibody against the A52 tag) as described in 'Materials and Methods'. B) An over-exposure of the immunoblot showed the disappearance of the CFTR 100 kDa degradation product with the introduction of the poly-valine constructs. Biogenesis of PEST mutants | It is possible that the effect of disruption of the PEST region may be too subtle to be detected from a simple expression blot. Effects of the disruption of the PEST region may be only apparent by examining its stability in a pulse-chase experiment. To examine whether the disruption of the PEST sequence affected the stability of the various mutants, the mutants' biosynthetic maturation were examined kinetically in pulse-chase experiments (figure ). COS-1 cells were transiently transfected with the various constructs for 3 days prior to the pulse-chase experiments. For the pulse-chase experiment, the cells were starved for 30 minutes in sulfur-minus (S-) media depleted of methionine and cysteine. The cells were then pulsed in S- media supplemented with radiolabeled methionine for 30 minutes. The cells were then chased for 0, 2, 4, 6, 8, 12, and 24 hours in plain media. At each chase point, the cells were harvested and frozen in plain media containing 10% DMSO and stored in a -70C freezer. After harvesting all the time points, the cells were lysed, immunoprecipitated with mouse anti-A52 antibody, separated by electrophoresis and detected by autoradiograph. As shown in figure , all of the WT constructs, with the exception of the poly-valine constructs, were still stable after 12 hours of chase. All of the misprocessed constructs (DeltaF508 background and the poly-valine in WT background), however, were completely degraded by the end of the 6 hours chase. There was no apparent difference in the stability of the constructs. Figure 3 | Biosynthetic maturation of CFTR PEST mutants. Biosynthetic Maturation of CFTR PEST Mutants. A) Pulse-chase radiograph for non-processing defective constructs: WT, WT/E725K/E726K, and WT/T717A/E725K; results for WT/S728A not shown, but is similar to that of WT. B) Pulse-chase radiograph for processing defective constructs: DeltaF508, DeltaF508/poly-valine, DeltaF508/E725K/E726K, and DeltaF508/T717/E725K; results for WT/Poly-valine and DeltaF508/S728A not shown but they show similar results as DeltaF508. Pulse-chase experiments and immunoprecipitation were performed as described in the "Materials and Methods" section. Global structural comparison of PEST mutants | To ensure that the disruption of the PEST region did not cause significant structural differences in the protein, we compared the global structures by examining their sensitivity to trypsin digestion (figure ). The trypsin-sensitivity assay is based on the rationale that trypsin-sensitivity of the various constructs depend on the exposure of trypsin-sensitive sites. It has been previously established that neither the subcellular location nor the glycosylation state of CFTR has a significant effect on its trypsin sensitivity . Crude membranes were made from transiently transfected cells. The membranes were treated with various concentrations of trypsin at room temperature for 5 minutes. The reactions were stopped by the addition of lima bean trypsin inhibitor. The samples were separated by SDS-PAGE and subjected to Western blot with anti-A52 antibody. As shown in figure , mature CFTR were about 100 --1000 fold more resistant to trypsin than the immature ER-resident CFTR. Thus, with the exception of the poly-valine mutant with WT background, the PEST sequence-disrupting mutants did not have any apparent effect on the folding pattern of the CFTR protein. Figure 4 | Trypsin sensitivity of CFTR PEST mutants. Trypsin Sensitivity of CFTR PEST Mutants. Membranes were prepared from COS-1 cells transiently transfected with the cDNA of the various constructs. The membranes were treated with various concentrations (0 --1000 mug/ml) of TPCK-treated trypsin. The reactions were stopped by the addition of lima bean trypsin inhibitor. Equivalent amounts of protein were subjected to immunoblot analysis (monoclonal antibody against the A52 tag). Effect of MG-132 on PEST mutants | We then examined the effects of proteasome inhibitors on the expression of these mutants to test our hypothesis that the PEST region may act as a backup proteolytic system to the ubiquitin-proteasome system; as the ubiquitin-proteasome system has been shown to be the prevalent proteolytic system of misprocessed CFTR, it is possible for it to overshadow whatever proteolytic effects of the PEST region may possess. COS-1 cells transiently transfected with various cDNA of the PEST constructs were treated with 2 muM of MG-132. MG-132 is a potent peptide aldehyde inhibitor designed to enter mammalian cells to inhibit the ubiquitin-proteasome pathway . As shown in figure , MG-132 blocks the maturation of WT CFTR and WT-background mutants. Inhibition of maturation by MG-132 has also been observed with P-glycoprotein . The level of expression, however, remains similar to that of the immature CFTR in untreated cells. With the DeltaF508 CFTR, poly-valine mutant and the other DeltaF508-backgound mutants, however, there was lower expression in the presence of MG-132. Figure 5 | Treatment with proteasome inhibitors. Treatment with Proteasome Inhibitors. Transiently transfected COS-1 cells were treated with (+) or without (-) 2 muM MG-132. The cells were subsequently lysed and subjected to immunoblot analysis (monoclonal antibody against the A52 tag). Discussion : Increasing the stability and promoting the maturation of misprocessed CFTR mutants has been a key interest in CF research. So far, not much is known about the actual mechanism of how these misprocessed CFTR are retained in the ER and targeted for degradation. Many lines of evidence have suggested the ER-associated degradation (ERAD) pathway, which includes the ubiquitin-protease pathway, as the dominant pathway for the disposal of misfolded CFTR in mammalian cells . Misfolded or misassembled proteins, such as processing defective DeltaF508 mutant CFTR, are recognized and retained in the ER by the quality control system in the ER; although the exact mechanism for recognition is yet to be elucidated, evidence indicate that the prolonged association with molecular chaperones may play a role in the retention of CFTR . Ubiquitination, the covalent attachment of ubiquitin to a lysyl epsilon-amino group of the polypeptide, is necessary for degradation by the 26S proteasome . Ubiquitin is a highly conserved polypeptide of 76 amino acids. CFTR may be both co- and post-translationally ubiquitinated in cell-free systems . Although the precise mechanism is still unclear, ubiquitination and ER retention eventually lead to the retrotranslocation of the misfolded protein back into the cytosol via the sec61p translocon . The misfolded protein is then targeted for degradation by the 26S proteasome . There are still many questions about the ERAD pathway and the maturation process of WT CFTR: 1) Although the mechanism for the ubiquitination reaction is well understood , the recognition determinants for ubiquitination are still unknown. 2) How does the ERAD pathway recognize the misfolded protein? 3) Although both WT and DeltaF508 CFTR are ubiquitinated, how is some WT CFTR still able to progress onto the post-ER compartments of the secretory pathway? In a recent report, CHIP, an Hsc70 co-chaperone was actually found to target immature CFTR for proteasomal degradation . It is likely that there are other chaperones that function in a similar manner. So, what are the selective molecular determinants for chaperones such as the CHIP to target only immature CFTR but not the mature form? Furthermore, it is the general consensus that there must be a number of criteria for exiting the ER and progressing through the secretory pathway for expression on the cell surface. Peptide motifs and signals may be exposed or buried depending on the folding state of the protein, thus providing the cell with a means of quality control. The perfect example is the four arginine-framed trafficking signals in CFTR discovered by Chang et al. . The arginine-framed trafficking signal, however, may just be one of the many types of signals. There are a number of peptide motifs identified in rapidly degraded proteins that target proteins for rapid degradation: PEST regions , lysosome-targeting KFERQ motifs , and the cyclin destruction box responsible for eukaryotic cell cycle . The PEST region is of particular interest to us because it has been found in many rapidly-degraded proteins and shown to contribute to their degradation . Although the presence of PEST motifs does not necessarily lead to constitutive degradation of the protein, there were a number of reasons why we thought it was worth investigating. Aside from its proteolytic roles, PEST motifs have also been found to be involved in protein-protein interactions: direct interaction with ubc9 and ligand recognition . The PEST region of CFTR lies within its R domain, which is not present in its processing efficient sister protein, P-glycoprotein . Thus, apart from a potential proteolytic site, the PEST region in CFTR also has the potential to interact with ubiquitination proteins, other proteases, or molecular chaperones. We were interested in seeing if it affects the maturation, folding, and stability of the CFTR protein. To this end, a number of constructs were made via site-directed mutagenesis aiming to disrupt the PEST region and examine the stability and folding pattern of the constructs. To narrow the scope of our study in this paper, we tested the PEST sequence's potential role as a proteolytic site in misprocessed CFTR. A number of mutants were made to disrupt the PEST sequence in the processing defective DeltaF508 CFTR mutant in an attempt to rescue it. The mutants' stability and maturation was examined via pulse-chase experiments and their global structure compared by testing their trypsin sensitivity. Overall, it was found that perturbation to the PEST sequence (residues 716 -- 734) of CFTR made no apparent improvement on the stability of DeltaF508 CFTR. In our survey of PEST-disrupting mutants, we included the WT-background mutants as well as the processing defective DeltaF508-background mutants as a control to ensure the missense mutations themselves did not cause other deleterious effect on the protein. The misprocessed poly-valine construct proves this precaution to be necessary. The observation that poly-valine caused defective processing came as a surprise since the R domain has been previously shown to be predominantly random coil . Although the boundaries of the R domain are still not precisely defined, the R domain extends approximately from residues 634 --708 at the NH2-terminus to approximately 835 at the COOH-terminus . Various studies have shown that the deletion of the R-domain yields cell surface expressing, functional CFTR . Furthermore, of the 19 disease-associated mutations in the R-domain examined by Vankeerberghen et al. , none of the clinical missense mutations studied within the boundaries of the R domain (residues 634 -- 835) was processing defective. Thus, it appeared that the R-domain did not contain significant structural information. However, the fact that the poly valine mutant used in our study leads to such dramatic disturbance as to cause the mutant protein to not mature suggests otherwise. Using a web-based secondary structure prediction program, NNPREDICT , the PEST region and surrounding regions (residues 712 --737) contain turning elements, with 3 helical elements. With the PEST region's low hydrophobicity, the region is likely to be surface-accessible. The poly-valine mutant, however, introduces beta strand elements in the middle of the turning elements. This could have led to disruption of a turn, and thus disrupting global structure. If the primary sequence around the PEST region does indeed carry structural information, perturbation of the PEST region by the other mutations might have also led to structural change and the trypsin assay might not be sensitive enough to detect them. These structural changes might have led to exposure of other proteolytic sites to counter the loss of proteolytic effect caused by alternation of the PEST sequence, hence, no apparent improvement in stability. An alternative test of structural change would be to find cysteine cross-linking pairs in CFTR and test if they still exist in the mutants . This would act as an additional structural check of the mutants. The PEST region may still interact with other proteins. The c-Myb transcription factor is a highly regulated nuclear phosphoprotein involved in the regulation of proliferation, differentiation, and apoptosis of immature hematopoietic cells. Its activity and proteolytic stability are, in part, regulated by the ubiquitin-proteasome system. c-Myb contains a PEST motif in its regulatory domain at its COOH-terminus. The PEST motif was found to directly interact with Ubc9. Ubc9 covalently attaches SUMO-1 protein, an ubiquitin-like protein, onto the lysyl residue near the PEST motif . Similarly, Ubc9 interacts with the PEST region of another nuclear SUMO-1 target protein, HIPK2 . Deletion of the PEST motif, however, does not increase the stability of the protein, suggesting other additional factors may contribute to the proteolytic stability of the proteins. There are a number of lysyl residues to the NH2-terminus of the PEST region and may act as potential candidates for ubiquitination. If CFTR's PEST motif also acts as a binding site for ubiquitin-conjugating enzymes like Ubc9 was for SUMO-1, that would explain why no difference was observed in the treatment of proteasome inhibitors between the WT and PEST mutants; the PEST motif would be part of the ubiquitin-proteasome pathway and not a separate proteolytic pathway. A potential candidate may be the Ubc6, an ER-resident ubiquitin-conjugating enzyme. In a recent report, mutation in Ubc6 was able increase the stability of DeltaF508 CFTR by 2 fold . Conclusions : Our results suggest the PEST region in the R domain of CFTR plays no discernible role in degradation of CFTR and that there may be structural information encoded in the R domain as mutations in the region may lead to misprocessing of the protein. The issue of whether CFTR's PEST motif can act as a recognition site, for other proteins, however, has not yet been addressed. As retention and degradation of misfolded CFTR is likely a collective work by many protein complexes, a survey of components of the proteolytic systems, molecular chaperones may yield more insight into the whether PEST motif plays a role in the degradation of CFTR. Methods : Construction of mutants | The CFTR point mutations were constructed and inserted into the mammalian expression vectors pMT21 as previously described . The full-length cDNA of all constructs were modified to encode the epitope for monoclonal antibody A52 and 6 histidine residues at the C-terminus . The sequence at the COOH terminus that would normally end DTRL then became DTRRAISLISNSCSPEFDDLPLAEQREACRRGD (His)6PRQ. Briefly, a BssHII restriction site was constructed at amino acid residue 1480 via site-directed mutagenesis using the DNA oligomer AAGATACAAGGCGCGCGAGAGCAGCAT to alter the Leu and stop codon to Arg and Ala. The DNA fragment encoding the A52 tag with a 5' BssHII site was then ligated to the 3' of the cftr gene. Expression of wild-type CFTR and mutants | Subconfluent COS-1 cells were transfected with the cDNA constructs with 1 mug/ml of the various vector constructs using a calcium phosphate precipitation method adapted from Chen and Okayama . The media was replaced with plain media at 20 hours post-transfection. The cells were harvested after another 48 hours. Isolation of microsomal membrane vesicles for trypsin digestion | The crude membranes were prepared in the same manner as optimized by Loo and Clarke for trypsin digestion assays. Briefly, for each CFTR construct, 20 (100 mm) tissue culture plates of transfected COS-1 cells were harvested and washed with PBS (phosphate-buffered saline). The cell pellets were subsequently resuspended in 3 ml of low ionic strength buffer (10 mM Tris-HCl, pH 7.5, 0.5 mM MgCl2). After incubating on ice for 20 min., the samples were homogenized using 40 strokes in a loose fitting Dounce homogenizer, followed by 20 strokes after addition of 3 ml of sucrose buffer (10 mM Tris-HCl, pH 7.5, 500 mM sucrose, 0.3 M KCl). The cell debris was removed by centrifugation at 3000 x g for 7 min. at 4C. The microsomes were collected from the supernatant by centrifugation at 44,000 x g for 45 min. at 4C and resuspended with 300 mul of TBS (Tris-buffered saline; 10 mM Tris-HCl, pH 7.5, 150 mM NaCl). The membrane vesicles were immediately used for the trypsin digestion. Trypsin digestion | As described previously , membrane vesicles from various constructs were treated with different concentrations of TPCK-trypsin (L-1-tosyl-amido-2-phenylethylchloromethyl ketone-treated trypsin; Sigma, 12,000 BAEE units per mg) at room temperature for 5 min.; the final concentrations of TPCK-trypsin were: 0, 0.1, 1, 10, 100, and 1000 mug/ml. The reactions were stopped by the addition of 2 mg/ml lima bean trypsin inhibitor (Worthington). The vesicles were solubilized with 2X sample buffer (3% SDS, 5% beta-mercaptoethanol, 10% glycerol, 62.5 mM Tris-HCl, pH 6.8). The samples were then subjected to SDS-PAGE and immunoblot analysis using mouse A52 monoclonal antibody as the primary antibody and a horseradish peroxidase labeled goat anti-mouse antibody as the secondary antibody. The signals were detected by chemiluminescence (enhanced chemiluminescence, Pierce). Pulse chase | Using methods adapted from Loo and Clarke , COS-1 cells expressing various CFTR constructs were radiolabeled and biogenesis of CFTR was followed. For each CFTR construct, 8 (100 mm) tissue culture plates of subconfluent COS-1 cells were transfected. At 24 hours post-transfection, the transfection media were replaced with plain media. At 72 hours post-transfection, the cells were starved for 40 minutes in Dulbecco's Modified Eagle Medium (DMEM) without L-methionine or L-cystine (Invitrogene Life Technologies). After aspirating the sulfur-deficient medium, the cells were labeled for 30 minutes in sulfur-deficient DMEM supplemented with 40 muCi/ml [35S] L-methionine and [35S] L-cystine (ICN Radiochemicals). The samples were then chased with plain media. One (100 mm) tissue culture plate from each construct was harvested at various chase time point, resuspended in 10% dimethylsulfoxide-containing media, and frozen in -70C freezer. Immunoprecipitation of CFTR | After thawing, each harvested pulse-chase sample was washed in PBS and subsequently resuspended in 100 mul PBS. The samples were then lysed in 1 mL buffer I (25 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 % (v/v) Triton X-100, 0.5 % (w/v) deoxycholic acid, 1 mM EDTA, pH 8.0) with protease inhibitors (50 mug/ml AEBSF, 10 mug/ml aprotinin, 25 mug/ml benzamidine, 1 mug/ml E64, and 0.5 mug/ml leupeptin). After removing the cell debris via centrifugation at 16,000 x g for 5 minutes, the supernatants were incubated with 11 mug of monoclonal A52 antibody at 4C for 1.5 hour. 50 mul 50 % (v/v) slurry of Protein A beads in TBS (Protein A sephrose 4 Fast Flow; Amersham Pharmacia) was added to each sample and incubated at 4C for 2 hours. The Protein A beads were then washed 3 times with buffer I. The CFTR proteins were then eluted with 50 mul 2X Sample Buffer with 2% beta-mercaptoethanol. The samples were then subjected to separation on 6% SDS-acrylamide gel via SDS-PAGE. The gels were fixed in 10% acetic acid for 30 minutes and the 35S signal amplified in Amplify solution (Amersham Pharmacia) for 30 minutes. The gels were then dried and exposed to X-OMAT AR film (Kodak) for 12 hours -- 1 week @ -70C. Author's Contributions : E.Y.C. carried out the experimental aspects of the studies and the preparation of the manuscript. D.M.C. conceived of the study and participated in its design and coordination. Backmatter: PMID- 12370087 TI - Inhibition of the MEK1/ERK pathway reduces arachidonic acid release independently of cPLA2 phosphorylation and translocation AB - Abstract | Background | The 85-kDa cytosolic phospholipase A2 (cPLA2) mediates arachidonic acid (AA) release in MDCK cells. Although calcium and mitogen-activated protein kinases regulate cPLA2, the correlation of cPLA2 translocation and phosphorylation with MAPK activation and AA release is unclear. Results | MEK1 inhibition by U0126 inhibited AA release in response to ATP and ionomycin. This directly correlated with inhibition of ERK activation but not with phosphorylation of cPLA2 on Ser505, which was only partially inhibited by ERK inhibition. Inhibition of AA release by U0126 was still observed when stoichiometric phosphorylation of cPLA2 on Ser505 was maintained by activating p38 with anisomycin. Translocation kinetics of wild-type cPLA2 and cPLA2 containing S505A or S727A mutations to Golgi were similar in response to ATP and ionomycin and were not affected by U0126. Conclusions | These results suggest that the ability of cPLA2 to hydrolyze membrane phospholipid is reduced by inhibition of the MEK1/ERK pathway and that the reduction in activity is independent of cPLA2 phosphorylation and translocation to membrane. The results also demonstrate that cPLA2 mutated at the phosphorylation sites Ser505 and Ser727 translocated with similar kinetic as wild-type cPLA2. Keywords: anisomycin, arachidonic acid, calcium, cytosolic phospholipase A2, ERK, fluorescent proteins, phosphorylation, MEK, U0126, translocation, anisomycin, arachidonic acid, calcium, cytosolic phospholipase A2, ERK, fluorescent proteins, phosphorylation, MEK, U0126, translocation, Background : Cytosolic PLA2 specifically hydrolyzes sn-2 arachidonate from phospholipid providing the precursors for many different lipid mediators including prostaglandins and leukotrienes . These lipid metabolites play a role in acute inflammatory responses and also regulate normal physiological processes. Certain prostaglandins are required for female reproduction and kidney function . Because of its important role in controlling levels of arachidonic acid (AA), much attention has been focused on the regulation of cPLA2 activation, with particular emphasis on the role of its phosphorylation and Ca2+-mediated translocation . cPLA2 is regulated by controlling its cellular localization and access to membrane-phospholipid substrate. An amino terminal, calcium-dependent lipid binding (CaLB or C2) domain regulates Ca2+-mediated cPLA2 translocation to intracellular membranes . In vitro, membrane docking via the C2 domain is necessary and sufficient for catalysis and release of AA . Binding of calcium ions by the cPLA2 C2 domain is essential for the lipid association in vitro and translocation in vivo . In response to an increase in [Ca2+]i, cPLA2 translocates to the Golgi and ER, however translocation to Golgi occurs at a lower [Ca2+]i. Protein kinase pathways play major roles in cPLA2 activation, and regulation by the mitogen-activated protein kinase kinase (MEK) /extracellular-signal regulated kinase (ERK) signaling pathway has received particular attention. cPLA2 is phosphorylated by mitogen activated protein (MAP) kinases, including p42/p44 ERKs and p38, on Ser505 in vitro and in response to receptor stimulation [,-]. In addition to phosphorylation by MAP kinase, it has been shown that cPLA2 is also phosphorylated on Ser727 by MAPK-interacting kinase I (MNKI) and on Ser515 by calcium/calmodulin-dependent protein kinase II . Phosphorylation of these sites may also play a role in regulating cPLA2 function in certain cell models. Phosphorylation of Ser505 has been extensively studied because it is readily detected due to a characteristic electrophoretic mobility shift when analyzed by SDS-PAGE . The importance of Ser505 phosphorylation in regulating cPLA2 has been demonstrated in different cells and in vitro models by using cPLA2 containing a S505A mutation . However, the mechanism whereby Ser505 phosphorylation regulates cPLA2 function has been elusive. In vitro studies have demonstrated that dephosphorylated cPLA2 is catalytically active and that Ser505 phosphorylation increases activity by only ~30 percent . In contrast, cells expressing the cPLA2 S505A mutation fail to release AA in response to a low dose of calcium ionophore, but release similar amounts of AA as cells expressing wild-type cPLA2 in response to high dose ionophore . From these studies, it has been suggested that cPLA2 Ser505 phosphorylation may have a role in regulating translocation . A previous study demonstrated translocation of cPLA2 S505A in response to Ca2+ ionophore, but did not address the kinetics of translocation, translocation in response to a physiological agonist, or differences in targeting . To better understand the regulation of cPLA2 by the MEK1/ERK pathway and Ca2+, we investigated the effect of MEK inhibitors on AA release, cPLA2 phosphorylation of Ser505, cPLA2 translocation kinetics, and [Ca2+]i increase in Madin-Darby canine kidney (MDCK) cells. We found that inhibition of MEK1 by U0126 significantly inhibited AA release and this was correlated with inhibition of ERK activation. However, MEK inhibition only partially affected cPLA2 phosphorylation and had no effect on the kinetics of Ca2+-mediated cPLA2 translocation to membrane. In addition, using cells expressing wild-type cPLA2 and cPLA2 with S505A or S727A mutations, it was found that translocation kinetics and membrane targeting in response to ATP or ionomycin was similar to wild-type cPLA2. These data suggest that MEK1 inhibition reduces cPLA2 catalytic activity and AA release independently of phosphorylation and translocation. Results : Effect of MEK inhibition on AA release, ERK activation, and cPLA2 Ser505 phosphorylation | To study the role of the MEK1/ERK pathway in cPLA2 activation, quiesced MDCK cells were treated with the MEK1 inhibitor U0126, and the effect on AA release, ERK activation, and cPLA2 gel shift determined . For equivalence with the imaging studies, cells expressing EGFP-cPLA2 were used in all experiments. EGFP-cPLA2 was expressed to similar levels as endogenous enzyme but did not contribute significantly to AA release in stably transfected cells. However, EGFP-cPLA2 is functional since it dose-dependently catalyzes release of AA when expressed in cells that lack endogenous cPLA2, such as Sf9 cells and immortalized mouse lung fibroblasts from cPLA2alpha knock-out mice . In order to enhance AA release in cells containing endogenous cPLA2, it is necessary to over-express the enzyme several fold as previously reported . Cytosolic PLA2 has been shown to mediate Ca2+-induced AA release in MDCK cells treated with ATP and IONO in experiments using the group IV cPLA2alpha-specific inhibitor pyrrolidine-1 . To measure cPLA2 mediated AA release, EGFP-cPLA2-transfected MDCK cells labeled with [3H]-AA were incubated with 0.3, 1 or 10 muM U0126 for 15 min prior to stimulation with 100 muM ATP, 1 muM IONO, or 10 muM IONO. AA release was measured at 3 min because we have shown that ATP- and IONO-stimulated AA release peaks between 3 to 5 min post-stimulation . Agonist-induced AA release was inhibited dose-dependently by U0126 with the highest U0126 concentration used (10 muM) reducing AA release by 72 --80% with all agonists . This inhibition was independent of the total amount of AA released, since AA release stimulated by 10 muM IONO was 3-fold greater than release stimulated with 1 muM IONO or 100 muM ATP, but the percent inhibition by U0126 was similar. Treatment of MDCK cells with 30 muM PD098059, a less potent inhibitor of MEK , resulted in a ~50% reduction in AA release in response to 100 muM ATP, 1 muM IONO, and 10 muM IONO (data not shown). Thus, in MDCK cells, MEK1 inhibition significantly reduces the ability of cPLA2 to hydrolyze AA from membrane phospholipids. Figure 1 | Effect of MEK inhibition on AA release, ERK inhibition, and cPLA2 Ser505 phosphorylation Effect of MEK inhibition on AA release, ERK inhibition, and cPLA2 Ser505 phosphorylation (A) EGFP-cPLA2-transfected cells were incubated with the indicated concentrations of U0126 for 15 min prior to stimulation with 100 muM ATP, 1 muM IONO, or 10 muM IONO. AA release was assayed at 3 min. Results show the average +- SD of 3 experiments (error bars are down for ATP for clarity). Protein extracts from EGFP-cPLA2-transfected cells incubated with the indicated concentrations of U0126 for 15 min prior to stimulation with 100 muM ATP, 1 muM IONO, or 10 muM IONO for 3 min were subjected to immunoblotting using (B) anti-phospho-ERK and (C) anti-cPLA2 polyclonal antibodies. Results are from 3 to 5 independent experiments. Table 1 | Inhibition of AA release by 10 muM U0126 See "Experimental Procedures" for experimental conditions and methods to determine AA release. Percent inhibition of AA release after 3 min of agonist treatment are average +- S.D. for N experiments preformed in duplicate. The effect of MEK1 inhibition on activation of p42/p44 ERK measured by immunoblot analysis using phospho-specific antibodies in cells treated with U0126 and stimulated as above was determined . Work in our laboratory has shown that recognition of ERK by anti-phospho-ERK antibodies correlates with an increase in ERK activity . Interestingly, the anti-phospho-ERK immunoblots revealed that ERKs were constitutively activated in untreated, quiesced MDCK cells and activation was not enhanced further by ATP or IONO (Fig. , left panel). ERK activation was diminished by increasing concentrations of U0126 and was quantitatively inhibited after 15 min incubation in 10 muM U0126. U0126 decreased ERK activation following ATP or IONO stimulation in the same fashion as in unstimulated cells. Consequently, there was a direct correlation between the reduction of AA release and inhibition of ERK activation in MDCK cells treated with U0126. Because cPLA2 is a target of the MEK1/ERK signaling cascade, we assayed the effect of MEK1 inhibition by U0126 on cPLA2 phosphorylation by analyzing gel shift of cPLA2. Phosphorylation of Ser505 results in a retardation of its electrophoretic mobility (gel shift) . In unstimulated cells, EGFP-tagged and endogenous cPLA2 were nearly completely gel shifted, indicating that most cPLA2 was phosphorylated on Ser505 , which is consistent with the observation that ERKs are constitutively activated. Incubation with U0126 resulted in a partial reversal of the gel shift although, at 10 muM U0126, approximately half of cPLA2 remained phosphorylated on Ser505. Thus, unlike the quantitative effect of U0126 on ERK activation, inhibition of MEK1 with U0126 only partially reversed the gel shift of cPLA2. The reversal of the gel shift was similar in cells treated with ATP and 1 and 10 muM IONO. Due to the increased molecular weight of the EGFP-tagged cPLA2, the two forms of cPLA2 did not separate as well, making the gel shift more difficult to visualize, but generally mirrored the gel shift characteristics of the endogenous cPLA2. These results suggest that, in response to MEK1 inhibition, there is a quantitative, dose-dependent decrease in AA release that correlates well with the loss of ERK activation, but not with the extent of cPLA2 Ser505 phosphorylation. To further investigate whether the MEK1/ERK pathway played a role in regulating AA release independent of Ser505 phosphorylation, we treated cells with anisomycin, which activates the MAPK homolog p38, but not the MEK1/ERK pathway . Activation of p38 in response to anisomycin treatment was analyzed by immunoblotting using an anti-phospho-p38 antibody . The immunoblots demonstrate that 30 min treatment in 25 ng/ml anisomycin resulted in phosphorylation of p38 in unstimulated MDCK cells and in cells treated with ATP or IONO. ATP and ionomycin treatment in the absence of anisomycin only weakly increased p38 phosphorylation. As expected, the MEK inhibitor U0126 did not significantly affect anisomycin-stimulated p38 phosphorylation. In control experiments, anisomycin treatment did not induce ERK activation nor interfere with inhibition of ERK activation by 10 muM U0126 treatment . We have previously reported that p38 is also selectively activated in anisomycin-treated macrophages . Importantly, pretreatment of cells for 30 min with anisomycin resulted in maintenance of the cPLA2 gel shift in the presence of U0126 in unstimulated cells and in cells stimulated with ATP and 1 and 10 muM IONO . AA release assays show that, in MDCK cells treated with anisomycin, U0126 resulted in an AA release reduction of ~67 --76% similar to the inhibition observed without anisomycin. These results demonstrate that activation of the MEK1/ERK pathway is required for AA release even under conditions where cPLA2 Ser505 phosphorylation is maintained, suggesting an alternative role for the MEK1/ERK pathway in regulating cPLA2. Figure 2 | Effect of anisomycin and U0126 on AA release, ERK inhibition, and cPLA2 Ser505 phosphorylation Effect of anisomycin and U0126 on AA release, ERK inhibition, and cPLA2 Ser505 phosphorylation EGFP-cPLA2-transfected cells were incubated with 25 ng/ml anisomycin for 30 min, 10 muM U0126 for 15 min, or with both drugs prior to stimulation with 100 muM ATP, 1 muM IONO, or 10 muM IONO. Protein extracts were subjected to immunoblotting using (A) anti-phospho-p38, (B) anti-phospho-ERK, and (C) anti-cPLA2 antibodies. (D) AA release in response to 100 muM ATP, 1 muM IONO, or 10 muM IONO in the presence and absence of the inhibitors. Anisomycin did not stimulate AA release in unstimulated cells. The results show average +- SEM for 3 to 5 independent experiments. [Ca2+]i increase is independent of MEK1/ERK pathway | One explanation for the decrease in AA is that U0126 inhibits [Ca2+]i mobilization in response to ATP or IONO, thereby preventing translocation of cPLA2. In chick ventricular myocytes, inhibition of MEK1/ERK by PD98059 inhibits zinterol-mediated AA release, but also inhibits zinterol-induced stimulation of [Ca2+]i cycling in electrically stimulated cells . In MDCK cells, extracellular ATP acts via P2Y2 receptors to elicit an IP3-mediated [Ca2+]i increase and IONO acts to increase [Ca2+]i by permeabilizing cell membranes to Ca2+. To determine the effect of U0126 on intracellular Ca2+ mobilization by 100 muM ATP or 10 muM IONO, we utilized single-cell fluorescence microscopy on cells loaded with the calcium indicator Fura2. Analysis of the [Ca2+]i increase in individual cells reveals the heterogeneity in the response to ATP, although most cells exhibited [Ca2+]i spikes of similar magnitude and duration (Fig. and , thin lines). Analysis of the [Ca2+]i increase in several cells (Fig. and , thick line) revealed that although the duration of the [Ca2+]i increase elicited by ATP in control cells was the same as in the U0126-treated cells, approximately 3 --4 min, the amplitude of the [Ca2+]i increase was slightly higher (~20%) in the U0126-treated cells. IONO elicited a sustained, supraphysiological [Ca2+]i increase in cells that was also slightly enhanced by U0126 (Fig. and ). These experiments demonstrate that U0126 does not decrease [Ca2+]i mobilization, and the inhibition of AA release by U0126 cannot be ascribed to a failure in [Ca2+]i mobilization. Figure 3 | Effect of U0126 on ATP- and IONO-induced [Ca2+]i change Effect of U0126 on ATP- and IONO-induced [Ca2+]i change MDCK cells were loaded with the fluorescent Ca2+ indicator Fura2, incubated with 10 muM U0126 or vehicle for 15 min, and stimulated with 100 muM ATP or 10 muM IONO. Graphs show changes in [Ca2+]i (expressed as the 340/380 ratio of Fura2 fluorescence) with respect to time from 8 individual cells (thin traces) and from the average of 16 cells (thick trace) treated with (A) U0126 and ATP, (B) vehicle and ATP, (C) U0126 and IONO, and (D) vehicle and IONO. Results are representative of 4 independent experiments in each condition and analysis of several cells per experiment. cPLA2 translocation is independent of MEK1/ERK pathway | The effect of inhibition of the MEK1/ERK pathway by U0126 on translocation of cPLA2 was investigated. Although there is no inhibition of [Ca2+]i release by U0126 and little effect on Ser505 phosphorylation, it is possible that MEK1 inhibition by U0126 prevents cPLA2 translocation by another mechanism. To investigate this possibility, cells were transfected with a wild-type cPLA2 fused to EYFP (EYFP-cPLA2) and the distribution of EYFP-cPLA2 was imaged in response to [Ca2+]i transients elicited by ATP and sustained [Ca2+]i elevations elicited by IONO, in the presence and absence of U0126. Following stimulation with 100 muM ATP, there was a rapid translocation of EYFP-cPLA2 to Golgi that was unaffected by U0126 . In response to physiological agonists that elicit transient [Ca2+]i changes, only a small fraction of the cPLA2 translocates. This observation is consistent with our previous results and has been demonstrated by Hirabayashi et al. . Most studies of cPLA2 translocation have utilized ionophore, which elicits a large, supraphysiological sustained increase in [Ca2+]i[,,,-], or agonists that produce a sustained [Ca2+]i increase. Under these conditions, a large proportion of cPLA2 binds to membrane. These studies show extensive translocation to the endoplasmic reticulum (ER), nuclear envelope and Golgi . We found that U0126 also failed to alter extensive EYFP-cPLA2 translocation to Golgi and ER in response 10 muM IONO . These results demonstrate that MEK inhibition has no effect on cPLA2 translocation. Figure 4 | Effect of U0126 on translocation of EYFP-cPLA2 following ATP or IONO stimulation Effect of U0126 on translocation of EYFP-cPLA2 following ATP or IONO stimulation Cells expressing EYFP-cPLA2 were incubated with 10 muM U0126 (B, D) or vehicle (A, C) for 15 min prior to stimulation with 100 muM ATP (A, B) or 10 muM IONO (C, D). Representative frames from time-lapse images show the distribution of EYFP-cPLA2 fluorescence before and after stimulation. Results are representative of 4 independent experiments and analysis of several cells per experiment. Translocation of phosphorylation site mutants S505A or S727A is similar as wild-type cPLA2 | Phosphorylation of cPLA2 on Ser505 has been hypothesized to play a role in Ca2+-mediated translocation since Ser505 phosphorylation is required for cPLA2-mediated AA release in response to low-dose, but not high-dose, ionophore . Translocation of cPLA2 S505A in CHO cells has been reported in response to ionophore stimulation , but the effect of Ser505 phosphorylation on the kinetics of translocation, targeting, and in response to a physiological agonist was not investigated. MDCK cells were co-transfected with EYFP-cPLA2 and a cPLA2 with a S505A mutation fused to ECFP (ECFP-cPLA2S505A). Using dual EYFP/ECFP imaging, we were able to directly compare translocation of both constructs in the same cell. The resting distribution of EYFP-cPLA2 was similar to that of ECFP-cPLA2S505A and, in response to ATP followed by IONO, the pattern of translocation of EYFP-cPLA2 was similar to ECFP-cPLA2S505A (Fig. panels A and B, D). Analysis of the increase in fluorescence at the Golgi with respect to time demonstrates that the rates of translocation of cPLA2 and cPLA2S505A elicited by ATP followed by IONO are very similar . As previously reported , the cPLA2 S727A mutation has a similar phenotype on AA release as the S505A mutation and Ser727 was found to be phosphorylated in tandem with Ser505. Imaging experiments were performed using EYFP-cPLA2 and ECFP-cPLA2S727A and we found that the distribution of EYFP-cPLA2 was identical to that of ECFP-cPLA2S727A before and after stimulation with 10 muM IONO (Fig. 7 panels A and B, D). Analysis of the increase in fluorescence at the Golgi with respect to time demonstrates that the rates of translocation of cPLA2 and the cPLA2S727A elicited by IONO are very similar . Figure 5 | Translocation of ECFP-cPLA2S505A following ATP and IONO stimulation Translocation of ECFP-cPLA2S505A following ATP and IONO stimulation Cells expressing ECFP-cPLA2S505A and EYFP-cPLA2 were stimulated with ATP and IONO. ECFP and EYFP fluorescence images were taken at 3 s intervals. Representative frames from time-lapse images show the distribution of EYFP-cPLA2 and ECFP-cPLA2S505A fluorescence before and 30 s after addition of 100 muM ATP, followed by addition of 10 muM IONO at 1 min (panels A and B). (C) Fluorescent protein (FP) fluorescence from the area of the Golgi (inset) was analyzed with respect to time before and after ATP and IONO addition (arrows). (D) Overlay of the 75 s images from panels A and B. Results are representative of 5 independent experiments and analysis of several cells per experiment. Figure 6 | Translocation of ECFP-cPLA2S727A following IONO stimulation Translocation of ECFP-cPLA2S727A following IONO stimulation Cells expressing ECFP-cPLA2S727A and EYFP-cPLA2 were stimulated with IONO. Representative frames from time-lapse images show the distribution of EYFP-cPLA2 and ECFP-cPLA2S727A fluorescence before and 13 s after addition of 10 muM IONO (panels A and B). (C) Fluorescent protein (FP) fluorescence from the area of the Golgi (inset) was analyzed with respect to time before and after IONO addition (arrow). (D) Overlay of the 77 s images from panels A and B. Results are representative of 4 independent experiments and analysis of several cells per experiment. Discussion : The MEK1/ERK pathway regulates cPLA2 and ERKs phosphorylate cPLA2 on Ser505. The results of this study demonstrate that this pathway is required for cPLA2-mediated AA release independent of Ser505 phosphorylation and extend our previous work in macrophages by demonstrating that this alternative role of the MEK1/ERK pathway is not involved in regulating [Ca2+]i change or cPLA2 translocation kinetics or targeting, but is required for optimal hydrolytic activity and AA release. The results shown here demonstrate that inhibition of MEK with U0126 quantitatively inhibits both ERK phosphorylation and AA release in MDCK cells in response to [Ca2+]i mobilization. The MEK inhibitor PD098059, a less potent inhibitor , also inhibited ATP- and IONO-induced AA release (data not shown). ERK was found to be constitutively phosphorylated in our study using MDCK cells from ATCC, in contrast to what has been reported previously for MDCK-D1, a subclone of MDCK selected for adrenergic receptor expression . However, a side-by-side comparison of MDCK cells from ATCC and the MDCK-D1 subclone (kindly provided by Dr. Paul Insel, UCSD) demonstrated constitutive activation of ERKs when both are grown at low density (not shown). However, at high density ERKs are less active and can be further activated by phorbol ester (not shown). In MDCK-D1 cells, without constitutively active ERKs, AA release is delayed after [Ca2+]i mobilization, and is temporally correlated with ERK activation , whereas in MDCK cells with constitutively active ERKs, AA release is rapid, with significant AA release measured 30 s after [Ca2+]i increase . This temporal correlation between AA release and ERK activation has also been reported in CHO cells in response to PAF stimulation . These results support the observations made here that ERK activity is required for phospholipid hydrolysis independently of cPLA2 translocation. cPLA2-mediated AA release must be preceded by translocation of the enzyme to its membrane substrate which is a Ca2+-dependent process and is a function of the calcium-dependent lipid-binding (C2) domain. cPLA2 translocates primarily to Golgi in response to a transient [Ca2+]i changes and to Golgi and ER in response to a sustained [Ca2+]i increase . The reduction in AA release by MEK inhibition did not involve a failure in [Ca2+]i release or translocation. These results show that translocation is necessary but not sufficient for optimum hydrolytic activity. Measuring cPLA2 translocation is not a trivial matter when investigating mechanisms of AA release. For example, cPLA2 constructs with a S505A mutation have long been recognized as unable to support AA release in response to physiological agonists or low-dose ionophore when transfected in cells. In contrast, cPLA2 S505A is active in vitro and phosphorylation only modestly increases the activity of the enzyme . Interestingly, the inhibitory effect of the S505A mutation on AA release is obviated by a high [Ca2+]i increase. In light of these observations, it is possible that Ser505 phosphorylation may alter the [Ca2+]i sensitivity of the enzyme, its rate of translocation, its intracellular targeting or, as has been previously suggested , the ability of cPLA2 to release from a non-membrane sequestering agent. Although one report has shown that cPLA2 S505A translocates in CHO cells in response to ionophore, we were able to directly compare rates of translocation between cPLA2 and cPLA2 S505A to a physiological agonist and found no difference between translocation rates or intracellular targeting. We also demonstrated that there was no difference in translocation rates or intracellular targeting between wild-type cPLA2 and cPLA2 S727A, which has the same phenotype as S505A with regard to AA release when transfected into cells. Thus, the role of cPLA2 phosphorylation in mediating AA release remains unclear. Although the alternative mechanism whereby the MEK1/ERK pathway regulates cPLA2 is not known, it is possible that it affects membrane properties and/or cPLA2 conformation that promotes optimal hydrolytic activity. It is also possible that the alternative mechanism is due to phosphorylation of cPLA2 on a novel site by a kinase that is downstream of the MEK1/ERK pathway or phosphorylation of a regulatory protein. Conclusions : Translocation to membrane is a critical regulatory step for the action of cPLA2 because it is necessary for access to substrate. In this study we demonstrate, however, that association of cPLA2 with membrane when phosphorylated on Ser505 is not sufficient for its full activity in vivo. This is demonstrated by the results showing that inhibition of the MEK1/ERK pathway significantly blocks AA release but has no effect on [Ca2+]i mobilization or cPLA2 translocation and targeting. Diminution of AA release by MEK1/ERK is also independent of cPLA2 phosphorylation on Ser505. Consequently, our results demonstrate in living cells that the translocation process and subsequent catalytic activity on the membrane are two independently regulated steps. Materials and Methods : Fluorescent protein-cPLA2 fusion constructs | DNA encoding the full-length human cPLA2 was cloned into the vector pEGFP-C3 (Clontech) to create pEGFP-cPLA2, as previously described . The XbaI/PstI fragment from a cPLA2alpha clone containing S505A or S727A mutations was inserted into an XbaI/PstI site in pEGFP-cPLA2 to generate pEGFP-cPLA2S505A and pECFP-cPLA2S727A. Different fluorescent-protein tagged constructs were produced by exchanging the NheI/BsrGI fragment containing the fluorescent protein coding sequence between EGFP, EYFP, and ECFP. All constructs were confirmed by sequencing. Cell culture | MDCK cells obtained from ATCC were cultured in DMEM containing 10% FBS, 100 U/ml penicillin, 100 mug/ml streptomycin, 0.292 mg/ml glutamine (growth medium) in 5% CO2 at 37C. Subconfluent cells (5 x 103 cells/cm2) were transfected with 2 mug of the relevant plasmid using Fugene-6 (Boehringer Mannheim) in DMEM containing 0.2% BSA, 100 U/ml penicillin, 100 mug/ml streptomycin, 0.292 mg/ml glutamine (serum-free medium) following the manufacturer's protocol. Stable lines expressing EGFP-cPLA2 were generated by growing transfected cells in growth medium for 3 d, supplementing the growth medium with 5 mg/ml Geneticin (antibiotic G418-sulfate), and culturing for an additional 2 wk in Geneticin. Cells expressing EGFP fluorescence were selected using a fluorescence-activated cell sorter. The EGFP-positive cells were maintained in growth medium supplemented with 5 mg/ml Geneticin. For imaging studies, MDCK cells were plated on glass-bottomed 35 mm culture dishes (MatTek) at 5 x 103 cells/cm2 in growth medium and incubated overnight, transfected with the relevant plasmid(s), changed into serum-free medium to quiesce the cells, incubated overnight, and used the next day. Immunoblotting | Stable EGFP-cPLA2 transfectants were grown on 100 mm dishes at 5 x 103 cells/cm2 in growth medium for one day, then quiesced in serum-free medium overnight. Cells were scraped into ice-cold lysis buffer: 50 mM HEPES, pH 7.4, 150 mM sodium chloride, 1.5 mM magnesium chloride, 10% glycerol, 1% Triton X-100, 1 mM EGTA, 200 muM sodium vanadate, 10 mM tetrasodium pyrophosphate, 100 mM sodium fluoride, 10 mug/ml leupeptin, and 10 mug/ml aprotinin. Lysates were centrifuged at 15,000 x g for 15 min, and protein concentration of the supernatant was determined by the bicinchoninic acid method. Laemmli electrophoresis sample buffer (5x) was added to the lysates, and SDS-polyacrylamide gel electrophoresis and immunoblotting were performed using 35 mug lysate protein, phospho-specific antibodies for ERK and p38, and rabbit polyclonal antibody for cPLA2. Dual imaging microscopy of fluorescent protein translocation | In order to compare the characteristics of full-length cPLA2 and cPLA2S505A or cPLA2S727A translocation, while controlling for cell-to-cell heterogeneity, we used a dual CFP/YFP imaging approach. EYFP-cPLA2/ECFP-cPLA2S505A- or pECFP-cPLA2S727A-transfected MDCK cells grown on MatTek plates were quiesced overnight in serum-free medium, washed with and incubated in Hank's balanced salt solution (HBSS) additionally buffered with 25 mM HEPES pH 7.4 (HHBSS). Cells were imaged using an Olympus inverted microscope equipped with a 60x, 1.25 NA oil immersion objective, CFP and YFP emission filters (Chroma) in a Sutter filter wheel, a dual CFP/YFP dichroic mirror, and a TILL Imago CCD camera (TILL Photonics). Excitation light of 430 and 510 nm for CFP and YFP, respectively, was provided using a Polychrome IV monochromator (TILL Photonics). TILLvisION software was used for acquisition and analysis. Bleach values for ECFP and EYFP were calculated by determining the background-corrected fluorescence for the entire cell with respect to time and normalizing each value to the initial value. ECFP/EYFP fluorescence changes with respect to time for regions of interest corresponding to an area of Golgi membrane were determined by calculating the Ft/F0, where Ft is the background- and bleach-corrected ECFP or EYFP fluorescence at time = t and F0 is the background-corrected ECFP or EYFP fluorescence at time = 0 s. Fluorescence was normalized to the F0 value, which resulted in Ft/F0 representing the fraction of total cell fluorescence at Golgi. Final images were produced using Adobe Photoshop. Calcium imaging | MDCK cells grown on MatTek plates were quiesced overnight in serum-free medium, washed with HHBSS containing 1 mM probenecid and incubated with 5 muM Fura2-AM (Calbiochem) in HHBSS, 1 mM probenecid, and 1% DMSO for 45 min at 37C. Cells were then washed with HHBSS containing 1 mM probenecid and imaged after a 30 min incubation for de-esterification of the Fura2-AM. Single-cell imaging was performed on the Olympus system described above, but using a 40x, 1.35 NA oil immersion objective and a Fura2 dichroic mirror and emission filter (Chroma). Fura2 image pairs illuminated at 340 and 380 nm were taken at 1 Hz. The [Ca2+]i increase is expressed as the ratio of the background-corrected Fura2 fluorescence at 340 and 380 nm . Measurement of AA release | The protocol for determining AA release is essentially as described . MDCK cells stably expressing EGFP-cPLA2 were plated in 12-well plates at 5 x 103 cells/cm2 and incubated in growth medium overnight. Cells were then washed twice with serum-free medium and incubated with 0.25 muCi [3H]-AA/well in serum-free medium overnight. U0126 or vehicle was added to each well and the cells were then incubated for 15 min at 37C in 5% CO2. Cells were washed to remove unincorporated [3H]-AA and then incubated in HHBSS supplemented with 0.05% BSA with either U0126 or vehicle. Cells were stimulated with the agonist of choice and the medium was collected at appropriate time points. The medium was centrifuged at 500 g for 5 min, and the amount of radioactivity in the supernatant was determined by scintillation counting. Cells were scraped in 0.5 ml 0.1% Triton X-100 for determining the total cellular radioactivity. Authors' contributions : JHE carried out the Ca and FP imaging studies, participated in the design and coordination of the study, and drafted the manuscript. DJF performed the AA release and Western blot studies and participated in the design and coordination of the study. CCL conceived of the study, participated in its design and coordination, and participated in writing the draft. All authors read and approved the final manuscript. Backmatter: PMID- 12361478 TI - Socio-demographic factors and self-reported funtional status: the significance of social support AB - Abstract | Background | The aim of the present work was to investigate the relative importance of socio-demographic and physical health status factors for subjective functioning, as well as to examine the role of social support. Methods | A cross-sectional health survey was carried out in a Greek municipality. 1356 adults of the general population were included in the study. Personal interviews were conducted with house-to-house visits. The response rate was 91.2%. Functioning has been measured by five indexes: 'The Social Roles and Mobility' scale (SORM), 'The Self-Care Restrictions' scale (SCR), 'The Serious Limitations' scale (SL), 'The Minor Self-care Limitations' scale (MSCR) and 'The Minor Limitations in Social Roles and Mobility' scale (MSORM). Results | Among the two sets of independent variables, the socio-demographic ones had significant influence on the functional status, except for MSORM. Allowing for these variables, the physical health status indicators had also significant effects on all functioning scales. Living arrangements and marital status had significant effects on four out of five indexes, while arthritis, Parkinson's disease, past stroke and kidney stones had significant effects on the SCR and SL scales. Conclusions | These results suggest that socio-demographic factors are as important as physical health variables in affecting a person's ability to function normally in their everyday life. Social support appears to play a significant role in explaining differences in subjective functioning: people living alone or only with the spouse, particularly the elderly, seem to be in greater risk for disability problems and should be targeted by preventive programs in the community. Keywords: functional status, social support, living arrangements, functional status, social support, living arrangements, Background : Over the last years, the strong proportional growth of the oldest age groups and the prevalence of chronic diseases, have raised two important issues for contemporary societies: coping with disability and ameliorating people's quality of life until the latest years of their lifespan. Understanding the factors that contribute to disability, may help clinicians and all those who participate in community care in preventing it or mitigating its impact and controlling the consumption of costly health care services. Previous findings, based mainly on the assumptions of the Andersen model, have widely confirmed that functional status is one of the physical status variables leading to the use of health services . It has been shown that functional status is a very important factor for a person's evaluation of each overall health status as 'good' or 'bad' and it constitutes in many cases the key concept for individuals, especially the old, prior to their decision to use health services . 'Functioning' -- and/or 'functional status' -- refers to a person's ability to perform the usual activities of everyday life. It is usually summarizing the concepts of 'disability' and 'social handicap' , as they were defined by the three levels of the "International Classification of Impairments, Disabilities and Handicaps" (ICIDH) scheme, i.e. impairment, disability and social handicap or disadvantage, in order to better understand the consequences of disease . 'Impairment' is any reduction in physical or mental capacities and mainly reflects disturbances at the organ level. 'Disability' refers to deficiencies or restrictions of expected activity performance and behaviour, i.e. the individual loses its ability to perform a function in a manner considered normal for a human being. Finally, a person is considered 'socially handicapped' when he/she has lost its ability to perform normal social roles. Because of the fact that functioning is a multi-dimensional concept, different measurement tools have been created through the years, depending primarily on the scope of the study and the conceptual model used . Some of the most widely used indices are the 'BADL' and 'IADL' scales. The 'BADL' (basic Activities of Daily Living) scales assess an individual's ability to perform 'primary biological activities', i.e. eating or dressing ability, and are more relevant for institutionalized patients and/or elderly individuals with severe disabilities . In addition, in order to assess for example the ability of an old person with minor health problems to live independently in the community, the 'IADL' indexes (Instrumental Activities of Daily Living) are preferred to the previous ones. These measurement tools include more typical activities of daily living, e.g. gross mobility, home chores, role performance and, this way, cover partially the 'handicap concept' . It must also be noted that functional status does not always reflect the physician-related health state of the individual. The existence of diseases and/or physical impairments are not sufficient causes for disability and/or social handicap or do not lead to the same levels of disability. As Siegel pointed out, less than half of older people over the age of 65, with medically diagnosed declining health, reported restrictions in their activities . On the other hand, in spite of the absence of a chronic disease or impairment some individuals are -- or feel -- disabled in some ways and have difficulty in performing some 'BADL' or 'IADL' tasks . Thus 'subjective functioning' or 'self-perceived functional status', i.e. a person's evaluation about its functional abilities, is a key issue in our understanding of the path from early symptomatology to disability and plays an important role in the decision making process that leads to the use of health services. Therefore, factors that go far beyond an individual's state of physical health, may also influence functioning. Different conceptual models, in addition to the widely used ICIDH scheme, have been developed, in order to better analyze the disablement process and the factors influencing it . It appears that, demographic and psycho-social variables mediate the relationship between health status and activities of everyday life. In particular, demographic characteristics have an impact on levels of daily functioning and seem to play a role even after controlling for a variety of health variables . There is also a general consensus among researchers that social support plays a positive role for mental and physical health, in spite of different theoretical or methodological approaches . Additionally, different patterns of living arrangements seem to have an impact on functional status and health in general . Particularly in Greece, no information exists on how different social support patterns affect self-perceived functional status. It is, however, believed that the strong social ties among family and community members cover efficiently the needs of people, especially in small communities, suggesting that living arrangements do not play a major role and thus do not affect in a significant way health status and self-perceived functional status in particular. The assessment of a person's ability to function normally in the community requires information not only on the level of disability, but also on other predisposing and environmental factors or the quantity and quality of social support that may be available. The aim of this study was to identify socio-demographic factors that seem to contribute to a better understanding of the disability process and to distinguish groups of people, who, because of certain characteristics, are at greater risk for disability and social handicap. The main objectives were the following: a) to test if there is an association between functioning and the use of health services, b) to investigate the importance of some socio-demographic factors for the functional status of individuals and in particular the social support variables, c) allowing for the influence of socio-demographic factors, what would be the relative effect of the physical health factors on functioning, d) to identify those variables that may better explain variations in functioning, e) to identify possible differences in the relative importance of socio-demographic and need factors in predicting different levels of functioning. Methods : Sample and research setting | The data used in this paper derived from a cross-sectional health survey carried out in Archanes, a medium size municipality of Crete, Greece. The broader aim of this survey was to describe and analyze the use of health services by the inhabitants of the community. The district of Archanes was divided in ten sectors (based on the municipal ordnance plan of house addresses) and a sample of 65 --70 houses was drawn from each sector using tables of random numbers. Ten interviewers visited all the housing units, one interviewer for each sector. A total of 678 houses were visited; in 57 of them the inhabitants either refused to participate or were 'unavailable'. In 3 houses it was impossible to contact the inhabitants, even after two repeated visits at different hours during the day. The response rate based on the selected households was 91.2%. The final sample comprised 2097 persons of all ages. Only the 1356 adults, aged 17 years or older, were included in the present analysis. Measures | Data were obtained through personal interviews with the adults of the household, using a questionnaire specially designed for this study (see : Questionnaire.pdf). A pilot survey preceded the main survey in order to detect and correct different problems of the questionnaire and its administration. This pilot study was carried out in a smaller sample of another municipality presenting similar socio-demographic characteristics to those of Archanes. The interviews were conducted by specially trained interviewers and 35 minutes on average were needed in order to complete all the questions. According to the main objectives of the present study, three groups of factors might be related with functioning: socio-demographic factors, physical health variables and use of health services. Functioning | Functioning was defined as a person's ability to perform his/her everyday typical activities. More particularly, subjective functioning was measured by using two questions on the performance of usual social roles, i.e. housework/work (Q1) and social contacts (Q2) and three questions on physical disability (gross body movements and self-care), i.e. independence in moving inside and outside the house (Q3), in dressing (Q4) and in eating (Q5). For the three functional disability questions the individual was asked not whether he/she 'can' perform the activity, nor what he/she 'does' do, but instead whether he/she 'has any difficulty doing it'. This intermediate phrasing was preferred in order to overcome possible biases related to the 'capacity' and/or the 'performance' wording . As it has been emphasized by previous research, an index using the fist phrasing (can you do the activity) may overestimate the healthiness of the respondent, i.e. the individual may think that he can do the activity, while he cannot. On the contrary, when a person is asked if he does do the activity, he may answer negatively because of reasons not related exclusively to health problems, such as psychological and/or external circumstances etc. Usually scales using the 'performance' phrasing underestimate the healthiness of the respondent . Responses for each of the five questions were given on a four-point scale: 'no restrictions', 'minor restrictions', 'severe restrictions' and 'completely restricted'. The two last categories were regrouped in one, because very few respondents of the sample were 'completely restricted'. Socio-demographic characteristics | Socio-demographic variables included: gender, age, level of education, employment status, profession, marital status, total number of persons living in the house and living arrangements. The last three variables were used as potential measures of social support. The different patterns of living arrangements describe the relationship of the individual with the 'head of the household'. Also, in order to make the distinction between 'active' and 'non active' members of the community, the six employment status categories were regrouped in two broader categories: 'employed' and 'unemployed'. As for marital status, the 'divorced/separated', represented only by six persons, were omitted from the final analyses. Physical health | Physical health status was measured by general health indicators and specific health measures. The generic health indicators included the assessment of general health status in the last 12 months, existence of chronic disease, limitations due to the chronic condition, reported acute morbidity and restriction in daily activities because of acute illness. As specific health indicators, a list of some of the most frequent medical conditions or diagnoses was given. Also the person was asked if he/she had experienced in the past some of the following conditions: stroke, cardiac infractus, asthma crisis or G.I. ulcer crisis. Other measures of physical status were 'surgery in the last 3 years' and a list of 24 symptoms present during the last two weeks. Use of health services | Utilization was measured by 'visits to the doctor during the last two weeks', 'use of medicines in the last two weeks', 'use of a group of health services during the last year' and 'hospitalization during the last three years'. To each one of these health services the respondent had to give a 'yes'/'no' answer. Statistical analyses | Principal Component Analysis | "Principal Component Analysis" (PCA) was used in order to identify a smaller number of underlying factors, which may have generated the dependence structure and the variation in the responses to the initial set of questions on functioning. The number of extracted factors were determined by the magnitude of eigenvalues (should exceed 1.0). Factors were subjected to Varimax rotation . Questions were considered as contributing to a factor if their factor loadings were greater than 0.40. Two PCAs were carried out. The first on the five questions on functioning with their initial three-point scales (1 = 'no restrictions', 2 = 'minor restrictions' and 3 = 'severe restrictions'), and the second with the variables transformed to binary (each question transformed to two binary representing 'minor restrictions' and 'severe restrictions', respectively). Although the two methods are similar, the second was used in order to study further relationships between categories of variables as well as relationships between variables. Analysis of Variance and analysis of covariance | The Analysis of Variance was used to find out if there were statistically significant differences of functioning (in terms of the functioning indexes provided by the two PCA methods) between the groupings of the factors in our hypothesis. . Standard assumptions of normality and constant variance were tested by the Shapiro-Wilk statistic and Levene homogeneity-of-variance test, respectively. Differences between more than two groups were further assessed by the Neuman-Keuls multiple comparison test . Also, in order to account for possible age and sex effects on these differences between functioning and the other groups of factors, we controlled for these two variables using Analysis of Covariance . Multiple Regression Analysis | In order to reveal the variables constituting the best predictors of changes on functioning scales, we used the 'stepwise' (backward elimination) method of multiple regression analysis . Therefore, socio-demographic and physical health status characteristics were the independent variables with dependent variables the functioning indexes of the two PCA methods. Categorical independent variables were transformed in binary (dummy) variables: a categorical variable with k values was transformed into k - 1 dummy variables. For such a transformed variable all dummy variables were included or excluded in the regression as a block. A separate regression was applied in order to find out the relative importance of each set of variables in explaining the variance of the five indexes; in particular, the importance of physical health status variables having allowed for the influence of a series of socio-demographic indicators. This was carried out allowing the sets of variables to sequentially enter the regression model as a block. Four sets of variables were processed: 1st = sex and age; 2nd = education and employment status; 3rd = the social support variables; and 4th = the physical health status variables. Standard assumptions regarding the distribution of errors were tested by examination of residuals . Results : Description of the sample | Table presents the distributions of some of the socio-demographic, physical health status and use of health services variables. 44.6% of the sample were men and 55.1% women. Most of the people of the sample were married (75%). Also, 68.2% of the participants had at best finished primary school. Only a small percentage (4.5%) was living completely alone; these were almost exclusively women. A significant part of the overall sample reported suffering from at least one chronic disease (41.3%) and this percentage was more pronounced among the elderly (76.1%). 17% of the study sample reported visiting a physician during the two weeks prior to the interview. Additionally, 48.1% of the sample used at least one medicine during the same period. Table 1 | Descriptive Statistics of Study Variables (%) Also, in Table are presented separately the distributions of the five functioning questions. 30.8% of the respondents reported restrictions on their main activity due to health reasons and 20.9% on social activities. Females have more problems than males and these differences between the two genders remain important in all age-groups (see : Functioning.pdf). Only for the elderly (65 and over) disparities in functioning between the two sexes are minor. Table 2 | Distribution of functioning questions by sex (% of those who reported any restriction -- level 2,3 and 4*) Principal component analyses | Two functioning indexes were yielded by the first method ; they both had eigenvalues greater than 1.0 and explained 85.1% of the total variance (of the five functioning variables). The communalities and factor loadings, after Varimax rotation, are shown in Table . The first index was characterised by high loadings of the functioning questions 1, 2 and 3 ('social roles and mobility scale' (SORM)), while the second index was loaded on the other two questions, 4 and 5 ('the self-care scale' (SCR)). All functioning questions had high communalities (> 0.80). Higher scores on both scales suggest increasing functioning problems and/or increasing seriousness of activity restrictions. Table 3 | Principal component analysis of functioning variables by two methods. Varimax rotation. All factor loadings are shown, those greater than 0.40 are typed in bold The second method yielded three indexes , all with eigenvalues greater than 1.0 and explaining 68.5% of the total variance. The communalities and factor loadings, after Varimax rotation, are shown in Table . Note that the first index grouped the serious limitations in all five activities, while the other two factors grouped only minor restrictions, in self-care (second index) and in social roles and gross mobility (third index). Therefore the three corresponding scales represent 'the serious limitations scale' (SL), 'the minor self-care limitations scale' (MSCR) and 'the minor limitations in social roles and mobility scale' (MSORM). Higher scores on the 'SL' scale suggest serious problems in functioning, while higher scores on the other two scales suggest minor restrictions in daily activities. Functioning and use of health services | Analysis of variance indicated a strong correlation between the functioning indexes and the four 'use of health services' variables . Again, in all analyses both hypotheses of normality and homogeneity of variance were accepted (p > 0.10). On all five indexes those who responded positively had significantly higher mean scores from those who answered not using these services. In some cases the magnitude of the difference between the mean scores of 'yes' and 'no' answers was quite large. For example, the mean scores on 'SL' and 'MSCR' scales are almost 8 times higher for those who used medicines 'during the last two weeks,' as opposed to those who responded negatively. Similar large differences were also observed between those who used 'one or more health services' during the last year than those who did not ('use of health services index'). Almost all of these differences, in all five scales, remained statistically significant even after adjusting for the effect of age . Table 4 | Means (Analysis of Variance) and Adjusted means (Analysis of Covariance with control variables age and gender) of the 'use of health services variables' for the 5 functioning indexes Functioning and socio-demographic variables | The relationship between functioning and socio-demographic characteristics was studied next. Results of analysis of variance and analysis of covariance are presented in Table . Again, in all analyses both hypotheses of normality and homogeneity of variance were accepted (p > 0.10). Sex, age, education and employment | Female participants reported more function problems and increasing seriousness of activity restrictions than their male counterparts. Only for the two indexes related to self-care restrictions (SCR and MSCR), the differences in mean scores between men and women were not significant. Furthermore, functioning is deteriorating progressively with age, with the highest mean scores for the fourth and fifth age groups. However, differences in mean scores between the elderly and the other age-groups are not so prominent for the 'SCR' index, as it is the case for the 'MSCR' scale. Education differentiated the mean scores on all indexes at a statistically significant level. Those who did not finish primary school (1st level) had the worst functioning profile contrary to those who had completed the highest level of education (5th level). Between these two groups, mean scores decrease from the lower to the higher education levels, although this decrease is not gradual from one level to the other. After controlling for age and sex, differences between mean scores remained statistically significant for all indexes, except the 'MSORM'. Similarly, employment status correlated significantly with functioning. There is a clear difference in mean scores on all scales between those who were employed and those who were 'unemployed' (non active), with the former reporting less problems in their daily functioning. It is worth noting that for index 4 (MSCR) the mean score is 17 times higher for the 'unemployed' compared to the 'employed'. The analysis of covariance corroborated the previous observations, although score differences between the two groups were decreased. The social support variables | The analysis of variance showed that, all three social support variables (marital status, living arrangements and size of household), were significantly associated with functioning (p < 0.001). After controlling for age and sex differences, marital status and living arrangements still contributed significantly to the variability of functioning scores. In particular, the 'unmarried' and more the 'widowed' had more problems in self-care restrictions (SCR and MSCR), and more serious limitations in all five activities of daily living (SL). Also, among the different living arrangements, the 'living alone', 'the couples living alone' and 'the parents of the head of the household' were the three groups with the poorest MSCR functioning and serious limitations (SL). Size of household was the only factor that did not influence functioning indexes at a statistically significant level when controlled for age and sex differences. Functioning and physical health status | All general health indicators were significantly correlated to the functioning indexes (see :Physical health 1.pdf). Accounting for age and sex differences did not change the results for three indexes (SORM, SL and MSORM) for most of the generic physical health status variables (see :Physical health 1.pdf). The most severely limited were those who had a stroke in the past, followed by the 'Parkinson', 'asthma' and 'kidney stones' sufferers (see :Physical health 2.pdf). Moreover, the analysis of covariance revealed that differences in the 'SL' scale between 'prevalent' and 'healthy' status of almost all 15 medical conditions of our list were significant; (see :Physical health 1.pdf). Socio-demographic and physical health status variables as predictors of functioning | Results of the multiple regression analyses are presented in Tables and . In both analyses all standardised residuals were not significant and the plots did not indicate any deviation from the standard assumptions. Living arrangements and marital status were among the most important socio-demographic predictors of functioning. In particular, 'living alone' was the only variable that 'explained' significantly the variance of four indexes, and its positive sign indicates more problems in social roles and mobility (SORM), in self-care restrictions (SCR and MSCR), and more serious limitations in all five activities of daily living (SL) . It was also one of the three variables that contributed to the variance of the 'minor self-care limitations' index (MSCR) the other two being 'unemployed' and 'limited because of acute illness'. Among the marital status categories, the 'widowed' and the 'married' were very strongly correlated with the 'SCR' and 'SL' indexes. The widowed had worse functioning profile (beta positive) compared to the 'unmarried' participants, while the married had the best. From the other socio-demographic variables, age had a significant effect only on the 'SORM' scale (older people tend to have more functioning problems); education on the 'SCR' and 'SL' scales (lower education associated with more problems); and employment status on the 'SORM' and 'MSCR' scales (unemployed status associated with more problems). Among the physical health status variables, 'arthritis', 'Parkinson's disease', 'past stroke' and 'kidney stones' had significant effect on the 'SCR' and 'SL' scales (their prevalence is associated with more functioning problems and more serious limitations in all five activities of daily living (SL). Table 5 | Means (Analysis of Variance) and adjusted means (Analysis of Covariance with control variables age and gender) of the socio-demographic variables for the 5 functioning scales Table 6 | Multiple Regression Analysis ('Stepwise-backward elimination') with dependent variables the 5 functioning scales (only significant regression coefficients (beta) are shown)a Table 7 | Amount of variance (R2) in functioning explained by socio-demographic variables and physical health status variables (multiple regression analysis with sequential order of entered variables) According to the results of the regression with all independent variables included in the model, physical health status variables and socio-demographic factors, the overall results yielded significant amount of variances explained for the functioning indexes, ranging from 32.6% to 63.9% . The socio-demographic factors alone (allowed first in the model), explained significant amounts of variance for all functioning indexes except MSCR (15.1%), ranging from 21.7% to 37.1%. The physical health status variables (allowed second in the model) increased significantly the explained variance of all functioning indexes, but by different amounts. That is, allowing for socio-demographic influences, physical health variables had larger effect on the SCR and SL indexes, than the SORM, MSCR and MSORM. Discussion : Self-perceived functional limitations constitute a central determinant leading to the utilization of health services. The principal goal of this cross-sectional study was to identify factors that are associated with subjective functional status variations in the adult population of a Greek semi-rural community. Moreover, we investigated the relative importance of two sets of variables, socio-demographic and physical health status variables, in influencing functional status. As mentioned above, the present findings were derived from a cross-sectional study and the observed relationships between the studied variables and functional status are valid for the social and demographic context of this small community of Archanes. They can possibly be generalized only for the similar rural/semi-rural population of Greece. However, this analysis, referring at a single point in time, is an initial approach in understanding which factors, other than disease and physical impairments, affect subjective functioning and contribute to disability. Firstly, we examined the relationship between health services use and functioning. It was confirmed by our data that those who reported poorer self-perceived functional status, were also more frequent users of health services. However, very few among the general and/or specific physical health status indices were important in explaining functional status variations. 'Self-perceived health', for example, is considered as a sensitive reflection of the threat that an individual feels about illnesses, affecting also its motivation to perform daily tasks . In previous research 'subjective health' was one of the strong predictors of functional decline . In the present study, it had a marked effect (just failing to be significant) only on 'the Social Roles and Mobility' index (p = 0.0563). Additionally, among the chronic conditions of our list, those predicting significantly more problems and serious limitations in functioning were the most disabling chronic illnesses, e.g. Parkinson's disease, arthritis, past stroke and kidney stones. Although our results suggest that age, education, employment status and the three social support measures -- marital status, size of the 'inner' family and living arrangements -- are associated with different levels of functioning, there are differences between these socio-demographic variables with regard to their effect on functional status. It is of note that gender was not included in the group of strong predictors of functioning status and age had only a significant effect on the 'social roles and mobility' index (older people tend to have more functioning problems of this type). Education had a significant effect only on two indexes: the 'self-care' and the 'serious limitations' index, with lower education associated with more problems. Similarly, participation in the workforce affects positively functional status. In fact, employment status was one of the strong predictors of variance in functioning for two scales: the 'social roles and mobility' and the 'minor self-care limitations' index. From the three proxy social support measurements, marital status was one of the strong predictors of variations in functioning levels. Among the five functioning indexes, marital status was significantly correlated with the 'SORM', 'SCR', 'SL' and 'MOSCR' indexes. More particularly, the 'single/unmarried' group had worse functioning profile than the 'married', whereas the 'widowed' were the least advantaged. In order to better clarify the above finding, we repeated the comparison among the three family status categories only for the older individuals, so as to avoid any distortion due to the small number of 'widowed' among the younger age groups. The results were similar, indicating that among the older people, being widowed in this Greek community has more disadvantages for health and functional status than being married, suggesting the existence and support of the broader social network (e.g. spouse, children). Other possible explanations for the worst functioning profile of the widowed compared to the married among the elderly could be the loss of support (e.g. after spouse's death) and the additional negative effect of living alone. Although the above findings need further investigation they are similar to those of other studies, where the widowed and the never married reported worse health profile than the married. In general, it has been shown that people who live with a partner have lower morbidity rates than those who live alone . Moreover, in previous studies relating marital status with the progression of functional limitations in patients suffering from disabling diseases, the rate of progression of functional disability was generally higher among unmarried participants compared to married patients . Additionally, the size of the 'inner circle' of the family network -- number of persons in the house -- was not among the variables having significant effect on functioning for none of the indexes. This result is consistent with those of past researchers, who stated that the quality and not the quantity of social support is probably the main factor with a positive impact on health . The functional status of those living in different living arrangements was also studied. After controlling for age and sex, the patterns of living arrangements are related significantly with the 'SL' and 'MSCR' indexes, indicating both serious disability problems. Also those who live alone, mainly older women, presented the worst functional status from all the other groups and this variable had a significant effect on functioning for four out of the five indexes. It appears that in this sample the 'living alone' group is the most vulnerable or perceives itself as the most vulnerable among the other living arrangements. Previous studies have not found any association of this group with poor psychological well-being and/or social isolation . Moreover Sarwari found that, unless severely physically impaired, women living independently have less deterioration in functional health when compared with peers living with spouses or nonspouse others . One explanation of the 'vulnerability' of the living alone in the present study might be that, even if they confront the same medical conditions as others, the absence of another adult on a daily basis makes it more difficult for them to cope with the incapacitating effects of illness. An additional reason might be that, in this small Greek community being old and cohabiting with the family is more socially acceptable than being old and alone. Therefore, this pattern of living arrangement probably magnifies one's disability problems. Another interesting finding was that the couples living in a household without children or others ranked equally high, very close to those who live alone. This pattern of living arrangement was among the strong predictors of the variance of the 'self-care scale' and the 'serious limitations scale'. It seems that the disadvantages of the cohabitation outweigh the advantages when we compare individuals of older age groups. On the contrary, being 'head of the family without the presence of a spouse', does not necessarily mean poor functional status, as Anson suggested . The absence of a spouse may be compensated by the support of other family members, such as older parents, children in adulthood, relatives. Additionally, being a lone-parent in this community probably results, not from personal preferences, but from previous life events, widowhood being the most frequent. Thus, it is a socially accepted pattern of living arrangement that may benefit from the support of the whole community. It is also possible that, because strong positive evaluation of health is necessary for keeping a lone-parent family, some functional problems may be underestimated and underreported. In conclusion, we found that the group of socio-demographic factors was an important predictor of the variation in functioning levels in addition to the physical health status variables. Within the first group, the social support variables accounted for a large proportion of the variance of all functional scales, except the 'MSORM'; the latter index had the lowest overall proportion of explained variance . Furthermore, many of the socio-demographic variables were retained as significant in the overall stepwise regressions and in particular two of the three social support measures (marital status and living arrangements) had significant effects on most scales. With respect to living arrangements, space, facilities and quality of relationships were not measured in the present study. However, even if we had included more qualitative data might improve our level of understanding, their absence does not diminish the strength of the association observed. Among the variables assessing physical health status, generic measures of health needs, such as suffering from a chronic condition or having reported 'poor self-perceived health', did not contribute significantly to the understanding of the variability in functioning scales. On the contrary, the existence of a serious disabling chronic condition, such as Parkinson's disease, was more important for functional status. We did not include in our model psychological factors that, undoubtedly, play an important role in offsetting or exacerbating disability . An index referring to possible psychological distress symptomatology during the two last weeks had a significant (negative) effect only on the 'SCR' scale. Conclusions : These findings support the theory that socio-demographic factors are as important as serious physical health needs, in affecting the ability of individuals to function normally in their daily life. As a result they should be taken seriously into consideration, when planning interventions for the prevention of disability in the community. Additionally, even in a medium size semi-rural community as this one, where social ties remain still strong among its members, the 'live alone' or the 'live only with the spouse' patterns of living arrangements were identified as high-risk groups for disability. In Greece, a country with a fast growing elderly population, more consistent and coordinated measures of community care should be considered in the future in order to meet effectively the needs of those groups. Authors' Contributions : SK participated in the design of the study, carried out the study and drafted the manuscript, IGV participated in the design of the study, performed the statistical analysis and drafted the manuscript, AP conceived of the study and participated in its design and coordination. Competing interests : None declared Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12398783 TI - Effect of hemofiltration filter adsorption on circulating IL-6 levels in septic rats AB - Abstract | Introduction | Hemofiltration may modulate the inflammatory response in sepsis through a variety of mechanisms. We sought to distinguish clearance from adsorption as the principal mechanism responsible for reducing circulating IL-6 levels with hemofiltration. Materials and methods | Nine hours after cecal ligation and puncture in 18 adult male Sprague --Dawley rats, we divided the rats into three groups (6 animals each) and placed groups 2 and 3 on a hemofiltration circuit connected between the right carotid artery and femoral vein using an AN69 membrane. In the hemofiltration group (group 2), ultrafiltrate was replaced with lactated Ringer's solution; in the recirculation group (group 3), the ultrafiltrate was reinfused into the animal. A sham group (group 1) had an arteriovenous circuit inserted but no hemofiltration. Blood was obtained for measurement of IL-6 and tumor necrosis factor (TNF) at the start of hemofiltration and after 5 and 11 hours of treatment. Results and discussion | IL-6 levels increased only in the sham-treated animals (20.4 +- 11.3 at baseline to 62.3 +- 16.8 pg/ml at 11 hours, P = 0.03) (differences between groups 1 and 2, P = 0.015, and groups 1 and 3, P = 0.028). TNF levels were highly variable but not significantly different among the three groups. Conclusion | Hemofiltration-associated reductions in circulating IL-6 levels appear to be secondary to adsorption of mediators to the filter membrane. We do not know whether this is due to direct adsorption of IL-6 per se or to the absorption of other mediators with secondary downregulation of IL-6 production or release. In addition, we could not exclude an interaction between adsorption and hemofiltration. Keywords: continuous renal replacement therapy, cytokines, hemofiltration, interleukins, sepsis, tumor necrosis factor, continuous renal replacement therapy, cytokines, hemofiltration, interleukins, sepsis, tumor necrosis factor, Introduction : Numerous studies have shown that hemofiltration using porous synthetic membranes can and does remove a wide range of substances that mediate inflammation from the plasma . Limited evidence supports the notion that this treatment can also influence circulating plasma concentrations of various mediators and mounting evidence suggests that these manipulations have important biologic effects . Despite these advances, there are no randomized clinical trials demonstrating that hemofiltration improves outcome in patients with sepsis. Furthermore, the exact mechanisms responsible for immunomodulation with hemofiltration remain uncertain. Modern hemofiltration filters may adsorb mediator substances as well as permitting them to pass through the membrane. Determining which of these mechanisms is dominant will be essential to advancing the design of materials and methodology suited to this form of therapy. If adsorption is the primary effect, the surface area of the filters must be increased, perhaps by using beads rather than filaments. Conversely, if mediator substances are removed mainly by sieving, increases in ultrafiltration rates will be required. We undertook these experiments to determine whether adsorption alone could affect the inflammatory state of rats with sepsis in a manner similar to standard hemofiltration. We chose to measure interleukin-6 (IL-6) as a marker of the activation status of the cytokine network because it is known to persist for several hours in the circulation and reflects the influences of several other cytokines . Although IL-6 is not cytotoxic, circulating IL-6 is reproducibly detectable in patients with sepsis, and higher concentrations portend a poor outcome . For these reasons, IL-6 has increasingly been used a marker of the inflammatory response in clinical trials of investigational agents in sepsis . We also measured tumor necrosis factor (TNF) in an effort to correlate changes in IL-6 with TNF activity during the same period. Materials and methods : Surgical preparation | After approval by the Animal Care and Use Committee of the University of Pittsburgh Medical Center, we anesthetized 20 adult, male Sprague --Dawley rats (mean weight 486 g) with pentobarbital sodium (50 mg/kg intraperitoneally). We performed a midline laparotomy, exteriorized the cecum, and placed a ligature inferior to the ileocecal valve using 4 --0 silk. We punctured the cecum three times using a sterile 18-gauge needle, placing one puncture site on each of the three antimesenteric surfaces. We then returned the cecum to the abdominal cavity and closed with 2 --0 suture. We then administered a subcutaneous bolus of saline (50 ml/kg) as fluid resuscitation and returned the animals to their cages and allowed food and water ad libitum. In two animals, we obtained serial blood samples each hour for 24 hours and measured IL-6 and TNF concentrations. The results from these animals were used to plan the experiments for the remaining groups. We randomized the remaining 18 rats into three groups. Seven hours after the cecal ligation and puncture (CLP), we reanesthetized each animal and intubated it with a beveled, 16-gauge angiocatheter and ventilated it with room air using a Harvard rodent ventilator (Holliston, MA, USA) at a tidal volume of 10 ml/kg and a frequency sufficient to maintain an arterial PCO2 between 35 and 45 mmHg. We isolated the right carotid artery and the left femoral vein by dissection and cannulated each with 1.27-mm PE-90 tubing. The tubing was formed into a catheter by inserting a beveled 20-gauge needle into one end. We flushed each cannula with heparinized saline (2000 units/liter). We placed a 3-way stopcock on each catheter so that blood sampling could be achieved using the carotid arterial catheter and fluid could be administered through the venous catheter. Experimental procedure | Nine hours after CLP, we connected the arterial and venous catheters for animals in groups 2 and 3 to a clamped hemofilter with an AN-69 membrane (Miniflow-10, Hospal, Lyon, France), surface area = 0.042 m2, fiber internal diameter 240 mum, fiber wall thickness 50 m. The filter, which had been flushed with sterile saline only, was flushed with 5000 unit/liter heparinized saline immediately before the circuit was opened. For animals in group 1, we connected the arterial and venous cannulas to each other and we removed 3 ml of blood to control for the deadspace of the filter, and replaced this volume with heparinized saline. The circuits were driven by the arterial blood pressure. The replacement fluid consisted of the following: 1:1 replacement of blood with 6% Hetastarch in saline, and 1:1 replacement of ultrafiltrate production with lactated Ringer's solution. We infused the lactated Ringer's solution as predilution into the arterial limb of the hemofilter. We gave the Hetastarch through the catheter in the femoral vein. We collected the ultrafiltrate and measured it twice an hour to maintain an accurate replacement rate (average 30 ml/hour). We began replacing the ultrafiltrate after the first half-hour. For animals in group 3, we reinfused ultrafiltrate as replacement fluid. We chose to study reinfusion rather than ultrafiltrate clamping because sieving may augment adsorption by increasing the effective surface area of the membrane by exposing more of the inner matrix of the filter to plasma. Blood was drawn from all the animals at hours 9, 14 and 20. The initial blood sample included blood for blood gas analysis (1.0 ml total). All other blood samples were 0.75 ml. We measured the hemoglobin concentration and oxygen saturation in each blood sample using an OSM3 Hemoximeter (Radiometer America, Inc, Westlake, Ohio, USA) to ensure that the animal was neither hypovolemic nor hypoxic. The blood sample was divided into two 0.4-ml chilled tubes containing EDTA (1 mg/ml) and centrifuged immediately at 1000 g for 10 minutes. The plasma was drawn off, placed into 0.5-ml labeled microtubes and stored at -70C for later cytokine analysis. The animals remained anesthetized and on the ventilator until the 20th hour after the laparotomy, when we killed them humanely. Measurements and calculations | Plasma concentrations of IL-6 and TNF were determined by rat-specific ELISA according to manufacturer instructions (Biosource International, Camarillo, CA, USA, and Endogen, Woburn, MA, USA), respectively. The TNF assay was linear from 0 to 2500 pg/ml, and the IL-6 assay was linear from 0 to 2000 pg/ml; coefficients of variation on repeated samples were less than 10% for both assays. Our primary analysis was based on the change in IL-6 concentrations from the beginning to the end of the hemofiltration (9 hours versus 20 hours) across the three groups. The absolute changes in IL-6 concentrations were compared using one-way analysis of variance (ANOVA) followed by the Student-Newman-Keuls (SNK) multiple comparison test. Our secondary analysis of TNF concentrations was performed after calculating the area under the curve (AUC) formed by three measurements (at 9, 14 and 20 hours) and zeroed by the baseline (9 hours) levels. Mean areas were compared for each group using ANOVA/SNK analysis as per IL-6 above. The absolute changes in IL-6 were correlated across all groups with the TNF AUC, using linear regression. All statistical tests were performed using MedCalc (v 4.2, Mariakerke, Belgium) software and statistical significance was assumed for P < 0.05. Unless specified otherwise, all results are presented as means +- standard deviation. Results : In the two animals used for mapping IL-6 and TNF responses to CLP, we found results consistent with those of other investigators . Namely, plasma IL-6 concentrations rose steadily, peaking between 18 and 20 hours after CLP, whereas no consistent pattern emerged for plasma TNF levels. The results for the three groups are shown in Table . IL-6 levels were not significantly different among the three groups at baseline and at 5 hours. However, IL-6 levels increased in the sham-treated animals (group 1) from baseline to 11 hours (see Table ; P = 0.03) (differences between groups 1 and 2, P = 0.015 and groups 1 and 3, P = 0.028) . TNF levels were highly variable between and within animals. TNF AUC analysis was not significantly different among the three groups, although the trend was in the same direction as for IL-6, in that group 1 tended to have the highest TNF concentrations (453 +- 509 pg/ml per 11 hours), group 2 the lowest (160 +- 443 pg/ml per 11 hours) and group 3 intermediate (359 +- 377 pg/ml per 11 hours). The relation between the change in IL-6 and the TNF AUC is shown in Fig. . The weak correlation between the two as shown by linear regression was not statistically significant. Figure 1 | Plasma IL-6 concentrations in rats before (light bars) and after (dark bars) 11 hours of hemofiltration started 9 hours after cecal ligation and puncture (CLP). Plasma IL-6 concentrations in rats before (light bars) and after (dark bars) 11 hours of hemofiltration started 9 hours after cecal ligation and puncture (CLP). Animals in group 1 underwent CLP with sham hemofiltration. Animals in group 2 underwent CLP and hemofiltration. Animals in group 3 underwent CLP and hemofiltration with reinfusion of the ultrafiltrate intravenously. Heights of bars correspond to mean IL-6 levels; error bars = SEM. Significant differences were found for group 1 before versus after, and after hemofiltration between groups 1 and 2 and 1 and 3 (P < 0.05). SEM, standard error of the mean. Figure 2 | Scatter plot of tumor necrosis factor area-under-the-curve (TNF AUC) versus the change in IL-6 before and after hemofiltration in rats that had undergone cecal ligation and puncture. Scatter plot of tumor necrosis factor area-under-the-curve (TNF AUC) versus the change in IL-6 before and after hemofiltration in rats that had undergone cecal ligation and puncture. The regression line was determined by linear regression (y = 0.0289x + 2.1158), r = 0.3217 (95% confidence interval for r = -0.1709 to 0.6856). Table 1 | Cytokine concentrations (pg/ml; mean +- SEM) found in rats after cecal ligation and puncture followed by hemofiltration Discussion and conclusion : Our findings show that the inflammatory response induced by CLP and characterized by a steady increase in circulating IL-6 levels over 18 --24 hours can be inhibited to a similar degree by hemofiltration whether or not the ultrafiltrate is reinfused. These findings provide strong evidence that immunomodulation seen with hemofiltration is predominantly the result of adsorption of circulating mediator substances to the filter membrane. Future studies of hemofiltration in sepsis might be more successful if steps are taken to maximize adsorption. Our findings also confirm those of our previous trial in humans and work from other groups showing that biologically significant immunomodulation can and does occur with continuous hemofiltration. However, the clinical significance of these changes remain in question. Unlike therapies designed to target specific mediators, hemofiltration has the theoretical advantage of being at once selective and nonspecific . Hemofiltration is selective in the sense that it will affect circulating substances in direct proportion to their concentrations in the plasma, but nonspecific in that it will remove multiple substances, both proinflammatory and anti-inflammatory . This feature makes hemofiltration quite different from other forms of immunomodulation that target specific mediators and reduce their concentration regardless of the baseline levels. Hemofiltration has been shown remove a wide variety of biologically active substances, many of which are known to be involved in the regulation of host defense, inflammation and tissue repair. However, controversy exists as to the primary mechanism responsible for this effect. We have previously shown, in patients with sepsis, that time-weighted mean TNF concentrations decreased during continuous venovenous hemofiltration but increased if the same patients were treated with continuous venovenous hemodialysis . Despite these changes in circulating TNF levels, only trace amounts of TNF were recovered from the ultrafiltrate. Subsequently, other groups have found similar results in patients treated with continuous venovenous hemofiltration on cardiopulmonary bypass or in patients with septic shock . However, both of these recent studies showed only early, transient reductions in plasma cytokines, suggesting that filter saturation occurs after several hours of therapy. The various membranes currently available differ significantly from one another in their adsorptive capacity for mediators . Indeed, the majority of studies of hemofiltration in sepsis have failed to find any change in circulating mediator concentrations. These observations have led to the development of combined hemofiltration/hemoadsorption techniques, which are beginning to show promise in animal studies . Our results do not exclude a possible role for mechanisms other than adsorption. Several investigators have found an association between increased ultrafiltration rates and improvement in hemodynamic stability in both animals and humans with septic shock . It may be that these improvements in hemodynamic stability do not involve cytokine removal. Furthermore, filtration appears to augment adsorption , so that combined filtration/adsorption might be more effective than adsorption alone, at least when hollow-fiber dialyzers are used. Other methods of increasing adsorption might be even more effective . A major limitation to our study is the fact that we cannot distinguish between increased IL-6 removal and decreased IL-6 production as the cause for the observed attenuation in IL-6 response to CLP. We chose to use IL-6 as our marker of the inflammatory response because, unlike many other cytokines, its plasma concentrations have been directly correlated with risk of death in intra-abdominal sepsis in humans, and measurements of this single parameter predicted outcome in these cases with remarkable (82.9%) accuracy . Furthermore, although IL-6 is not directly cytotoxic, there is evidence that it modulates other inflammatory cytokines , and plasma concentrations have been found to be 69% (95% confidence interval 30%-108%) higher in nonsurvivors than in survivors with sepsis . Many studies have shown an association between mean plasma IL-6 concentrations over time and mortality , and persistent elevations in IL-6 appear to be more important than initial or peak levels . However, plasma IL-6 levels reflect both clearance (endogenous and exogenous) and production, and IL-6 production is largely determined by the activity of other proinflammatory cytokines (especially IL-1 and TNF). Thus, hemofiltration may attenuate the sepsis-induced rise in IL-6 by adsorption of IL-6 or by adsorption of other mediators or both. Indeed, it seems likely that our findings resulted from a downregulation in the inflammatory response to CLP. This explanation is speculative, however, because we could not demonstrate a corresponding decrease in plasma TNF. Analysis of plasma TNF concentrations is more complex than of IL-6, because the TNF response is already waning by 9 hours after CLP and is much more variable. Furthermore, we measured total TNF and do not know the proportion that was biologically active at the time. We chose to analyze TNF using AUC because of these concerns and because the effects of TNF over time presumably relate to outcome and to IL-6 expression more than to a concentration at a single time point. Unfortunately, the smallness of the animals precluded multiple sampling and we selected only three time points in order to obtain the 'curve' used for AUC analysis. These limitations and the small sample size perhaps explain why we were only able to show a weak correlation between TNF AUC and changes in IL-6 concentrations over time. Finally, we must acknowledge that our animal model of sepsis, like sepsis in humans, is associated with a variable inflammatory response . However, both absolute and relative changes in IL-6 levels appear similar with hemofiltration with or without recirculation of ultrafiltrate. These results suggest that adsorption is the primary, though perhaps not the sole, mechanism responsible for changes in IL-6 levels during hemofiltration. Key messages : Changes in IL-6 seen with hemofiltration are predominantly the result of adsorption of circulating inflammatory mediator substances to the filter membrane Biologically significant immunomodulation can and does occur with continuous hemofiltration at blood flow rates currently in use in clinical practice We cannot distinguish between increased IL-6 removal and decreased IL-6 production as the cause for the attenuation in the IL-6 response to sepsis observed with hemofiltration Future studies of hemofiltration in sepsis might be more successful if steps are taken to maximize adsorption Competing interests : None declared. Abbreviations : ANOVA = one-way analysis of variance; AUC = area under the curve; CLP = cecal ligation and puncture; ELISA = enzyme-linked immunosorbent assay; IL = interleukin; SNK = Student-Newman-Keuls; TNF = tumor necrosis factor. Backmatter: PMID- 12398784 TI - Decreased levels of dehydroepiandrosterone sulphate in severe critical illness: a sign of exhausted adrenal reserve? AB - Abstract | Introduction | Dehydroepiandrosterone (DHEA) and its sulphate (DHEAS) are pleiotropic adrenal hormones with immunostimulating and antiglucocorticoid effects. The present study was conducted to evaluate the time course of DHEAS levels in critically ill patients and to study their association with the hypothalamic --pituitary --adrenal axis. Materials and method | This was a prospective observational clinical and laboratory study, including 30 patients with septic shock, eight patients with multiple trauma, and 40 age- and sex-matched control patients. We took serial measurements of blood concentrations of DHEAS, cortisol, tumour necrosis factor-alpha and IL-6, and of adrenocorticotrophic hormone immunoreactivity over 14 days or until discharge/death. Results | On admission, DHEAS was extremely low in septic shock (1.2 +- 0.8 mol/l) in comparison with multiple trauma patients (2.4 +- 0.5 mumol/l; P < 0.05) and control patients (4.2 +- 1.8; P < 0.01). DHEAS had a significant (P < 0.01) negative correlation with age, IL-6 and Acute Physiology and Chronic Health Evaluation II scores in both patient groups. Only during the acute phase did DHEAS negatively correlate with dopamine. Nonsurvivors of septic shock (n = 11) had lower DHEAS levels (0.4 +- 0.3 mumol/l) than did survivors (1.7 +- 1.1 mumol/l; P < 0.01). The time course of DHEAS exhibited a persistent depletion during follow up, whereas cortisol levels were increased at all time points. Conclusion | We identified extremely low DHEAS levels in septic shock and, to a lesser degree, in multiple trauma patients as compared with those of age- and sex-matched control patients. There appeared to be a dissociation between DHEAS (decreased) and cortisol (increased) levels, which changed only slightly over time. Nonsurvivors of sepsis and patients with relative adrenal insufficiency had the lowest DHEAS values, suggesting that DHEAS might be a prognostic marker and a sign of exhausted adrenal reserve in critical illness. Keywords: adrenal insufficiency, dehydroepiandrosterone sulphate, multiple trauma, hypothalamic --pituitary --adrenal axis, sepsis, adrenal insufficiency, dehydroepiandrosterone sulphate, multiple trauma, hypothalamic --pituitary --adrenal axis, sepsis, Introduction : Dehydroepiandrosterone (DHEA) and its sulphate (DHEAS) are the most abundant steroids secreted by the adrenal cortex . The concentration of DHEA in the blood oscillates in parallel with cortisol, in response to levels of adrenocorticotrophic hormone (ACTH), but without feedback control at the hypothalamic --pituitary level. The physiological role and biological actions of DHEA(S) are not well known but studies in humans suggest a positive impact on sense of well-being , and DHEA has recently been recognized as a potent modulator of the immune response . DHEA improved host defences by restoring immune cell function and reversed susceptibility to infection . Serum DHEA(S) concentration was low in patients with primary adrenal insufficiency, and short-term oral DHEA replacement improved the clinical condition of these patients . Functional or relative adrenal insufficiency frequently occurs in critically ill patients, with possible fatal consequences, although diagnostic criteria for this entity still pose problems . We hypothesize that the serum DHEAS level has utility as a diagnostic tool and a prognostic marker in such patients. Furthermore, low serum concentrations of DHEAS might be a more sensitive marker of hypothalamic --pituitary --adrenal (HPA) hypofunction than is glucocorticoid secretion. The present study was conducted to evaluate the time course of DHEAS levels (an immunostimulator) as compared with those of cortisol (an immunosuppressor), ACTH (an inducer of DHEAS) and cytokines (stimulators of the HPA axis) in patients with critical illness and in age- and sex-matched control patients. Materials and method : Study design | This was a prospective, observational, clinical and laboratory study conducted in a 20-bed medical/surgical/neurosurgical intensive care unit. Patients | Approval for the study was obtained from our institutional Human Subjects Research Committee, and written informed consent from first-degree relatives was mandatory. Thirty-eight consecutive patients who were admitted to the intensive care unit with septic shock (n = 30) or severe multiple trauma (Injury Severity Score >20; n = 8) were included in the study within 6 hours after admission. Exclusion criteria were as follows: age under 18 years; use of corticosteroids, DHEA or other drugs that affect the HPA axis; pre-existing adrenal insufficiency or known abnormalities of the HPA axis; and presence of diabetes mellitus or congestive heart failure. Age- and sex-matched control patients (n = 40) were patients without acute medical illness who were admitted to the medical department for routine diagnosis and treatment. Data collection | In patients with septic shock or multiple trauma, Acute Physiology and Chronic Health Evaluation (APACHE) II and Sequential Organ Failure Assessment scores, and intensive care unit mortality were used to assess the severity of disease. In addition, use of dopamine (cumulative dose in milligrams per day) was recorded. We obtained serial blood samples for measurement of serum cortisol and DHEAS, and plasma ACTH, IL-6 and tumour necrosis factor (TNF)-alpha once a day between 0700 and 0800 h. Follow up was conducted for 14 days, or until death or discharge from the intensive care unit. Blood samples were stored at -70C until use. Immunoreactive DHEAS, cortisol, ACTH, TNF-alpha and IL-6 concentrations were measured using commercially available chemiluminescent enzyme immunoassays with the Immulite Automated Immunoassay System (Diagnostic Products Corp, Los Angeles, CA, USA). In patients with clinical suspicion of (relative) adrenal insufficiency (unexplained hypotension and resistance to inappropriately high doses of vasoactive drugs) a low-dose ACTH (1 mug) stimulation test was performed. A normal cortisol response was defined as greater than 550 nmol/l after stimulation and an increase of 150 nmol/l or more . Statistical analysis | Values are expressed as means +- SD. Qualitative data were analysed using the chi2 test. Groups were compared using the Kruskal-Wallis test, with Dunn's test for multiple comparisons. The Spearman rank order correlation coefficient (rs) was used to evaluate relations for individual data. P < 0.05 was considered statistically significant. All analyses were performed using a statistical software package (SPSS 9.0.1; SPSS Inc, Chicago, IL, USA). Results : The clinical and laboratory characteristics of the patients and control groups on admission are summarized in Table . On admission, patients with septic shock had significantly lower DHEAS levels (1.2 +- 0.8 mol/l) than did those with multiple trauma (2.4 +- 0.5 mol/l, P < 0.05) and control patients (4.2 +- 1.8 mumol/l, P < 0.01). There was a significant negative correlation between DHEAS and age (rs = -0.55, P < 0.01) in the patient groups (pooled), but there was no significant difference in DHEAS concentrations between male and female patients. We found a negative correlation between DHEAS and dopamine only during the acute phase (septic shock: rs = -0.60; trauma: rs = -0.55; P < 0.01). Also, the correlation between DHEAS and IL-6 (septic shock: rs = -0.61, P < 0.01; trauma: rs = 0.47, P < 0.05) was more pronounced during the acute phase. These correlations were lost during prolonged illness (after 5 days). Table 1 | Clinical and laboratory data on admission in patients with septic shock or multiple trauma, and age- and sex-matched control patients The time course of DHEAS during the observation period of 14 days is shown in Fig. . In both septic and trauma patients, DHEAS concentrations were consistently lower than those in control patients over time. In contrast, cortisol levels were persistently elevated at all time points, in both septic and trauma patients. On admission, nonsurvivors of septic shock (n = 11) had lower DHEAS levels (0.4 +- 0.3 mol/l) than did survivors (1.7 +- 1.1 mol/l, P < 0.01). In addition, DHEAS was inversely correlated with APACHE II (rs = -0.52, P < 0.01) and Sequential Organ Failure Assessment (rs = -0.45, P < 0.05) scores. Figure 1 | The time course of serum dehydroepiandrosterone sulphate (DHEAS) levels in patients with septic shock , patients with multiple trauma , and age- and sex-matched control patients . The time course of serum dehydroepiandrosterone sulphate (DHEAS) levels in patients with septic shock , patients with multiple trauma , and age- and sex-matched control patients . In eight septic patients there was clinical suspicion of (relative) adrenal insufficiency, which was confirmed by a blunted response to low-dose ACTH in four cases. The patients with adrenal insufficiency had lower DHEAS levels at the time of testing (0.35 +- 0.3 mol/l) than did those with 'normal' adrenal function (1.3 +- 0.8 mol/l, P < 0.05). Discussion : We found a clear dissociation between high blood levels of cortisol and extremely low levels of DHEAS in critically ill patients in both the acute and prolonged phases. Parker and coworkers demonstrated such a divergence in adrenal steroid secretion; in that study serum cortisol was increased in adult men with burn injuries, whereas serum DHEAS was reduced. Luppa and coworkers studied serum androgens in a large group of critically ill patients and also found markedly decreased DHEAS levels in both males and females, mainly in those patients with a prolonged clinical course. These data indicate a shift in adrenal steroid synthesis away from mineralocorticoids and androgens and toward excessive cortisol production. The dissociation between blood levels of cortisol and DHEAS appears to be a contradiction because both hormones are synthesized and secreted mainly by the adrenal cortex. However, DHEAS is produced mainly in the zona reticularis of the adrenal cortex, possibly indicating that a differential alteration in the cortical zone is responsible for DHEAS deficiency during severe critical illness. The sustained hypercortisolism, as opposed to the marked DHEAS depletion, during severe critical illness could theoretically result in an imbalance between immunosuppressive and immunostimulatory pathways, and may therefore play a role in susceptibility to infectious complications . Interestingly, we found the lowest DHEAS and the highest cortisol levels in nonsurvivors and the most severely ill patients, indicating that the DHEAS : cortisol ratio might be a prognostic indicator for outcome of critical illness, in particular septic shock. Interpretation of cortisol levels measured in seriously ill patients is difficult. Serum cortisol levels that are regarded as high in control individuals may be inappropriately low in patients who are severely ill. We recently showed that functional or relative adrenal insufficiency can be present in critically ill patients despite 'high' initial serum cortisol levels . By using the low-dose ACTH stimulation test and Thorn test, we demonstrated the relative lack of adrenocortical response to extra stimulation by ACTH in some critically ill patients, because their HPA axis is already maximally stimulated. In the present study the low-dose ACTH test identified four patients out of eight with a clinical suspicion for relative adrenal insufficiency. These patients had very low serum concentrations of DHEAS, which may also be a sign of limited adrenocortical reserve arising during the course of critical illness . In both septic and trauma patients we found a similar degree of stimulation of the HPA axis; however, patients with septic shock were more severely ill than were patients with multiple trauma, as reflected by their APACHE II scores, and TNF-alpha and IL-6 levels. One could argue that this also indicates a state of exhausted adrenal reserve. We found a relation between dopamine use and DHEAS levels, but only during the acute phase. Therefore, acute depletion of DHEAS might reflect the liberal use of dopamine . In addition, we found a negative correlation between IL-6 and DHEAS during the acute phase. In healthy persons, serum IL-6 correlated inversely with DHEAS, and DHEA administration led to inhibition of IL-6 secretion from monocytes, indicating a functional link between DHEAS and IL-6 . IL-6 can act synergistically with ACTH on the adrenal glands to release cortisol . Therefore, IL-6 may be an important regulator of DHEAS in (acute) critical illness. However, the dopamine dosage and IL-6 levels decreased significantly over time whereas DHEAS concentrations remained low, suggesting different mechanisms for the prolonged DHEAS depletion during critical illness. We found a negative correlation between age and DHEAS concentrations. DHEAS concentrations exhibit a biphasic time course following the onset of adrenarche, reaching a peak between the ages 20 and 30 years, and with the greatest decline occurring by age 50 --60 years . This dramatic age-related reduction might be caused by a specific defect in the desmolase activity in the reticular zone of the adrenal gland. Most of patients studied here, including control patients, were aged approximately 50 --60 years. In conclusion, we found extremely low DHEAS levels in virtually all critically ill patients, in both septic shock and multiple trauma. DHEAS depletion was associated with a worse outcome and represents a prognostic marker. Acute depletion of DHEAS is probably related to the use of dopamine and high IL-6 levels. The prolonged depletion of DHEAS might reflect an exhausted adrenal adaptation. Whether DHEA should be administered in DHEAS-deficient states remains to be elucidated. However, theoretically, beneficial effects on immunity, susceptibility for infections and well-being may be expected. Key messages : Critically ill patients exhibit a remarkable depletion in DHEAS in both acute and chronic phases, suggesting an exhausted adrenal adaptation There is a clear dissociation between DHEAS (decreased) and cortisol (increased) levels in critically ill patients, indicating a disturbed balance between immunostimulatory and immunosuppressive factors DHEAS appears to be a prognostic marker, because nonsurvivors of septic shock have extremely low DHEAS levels The acute depletion of DHEAS is probably related to the use of dopamine and the high IL-6 levels during the acute phase of critical illness The prolonged depletion of DHEAS during critical illness is unexplained, but supports the hypothesis of an exhausted neuroendocrine system, and necessitates an interventional study with substitution doses of DHEA, considering the expected beneficial effects on immunity and well-being Competing interests : None declared. Abbreviations : ACTH = adrenocorticotrophic hormone; APACHE = Acute Physiology and Chronic Health Evaluation; DHEA(S) = dehydroepiandrosterone (sulphate); HPA = hypothalamic --pituitary --adrenal; IL = interleukin; TNF = tumour necrosis factor. Backmatter: PMID- 12398785 TI - Clinical evaluation of the Life Support for Trauma and Transport (LSTAT) platform AB - Abstract | Introduction | The Life Support for Trauma and Transport (LSTATTM) is a self-contained, stretcher-based miniature intensive care unit designed by the United States Army to provide care for critically injured patients during transport and in remote settings where resources are limited. The LSTAT contains conventional medical equipment that has been integrated into one platform and reduced in size to fit within the dimensional envelope of a North Atlantic Treaty Organization (NATO) stretcher. This study evaluated the clinical utility of the LSTAT in simulated and real clinical environments. Our hypothesis was that the LSTAT would be equivalent to conventional equipment in detecting and treating life-threatening problems. Methods | Thirty-one anesthesiologists and recovery room nurses compared the LSTAT with conventional monitors while managing four simulated critical events. The time required to reach a diagnosis and treatment was recorded for each simulation. Subsequently, 10 consenting adult patients were placed on the LSTAT after surgery for postoperative care in the recovery room. Questionnaires about aspects of LSTAT functionality were completed by nine nurses who cared for the patients placed on the LSTAT. Results | In all of the simulations, there was no clinically significant difference in the time to diagnosis or treatment between the LSTAT and conventional equipment. All clinicians reported that they were able to manage the simulated patients properly with the LSTAT. Nursing staff reported that the LSTAT provided adequate equipment to care for the patients monitored during recovery from surgery and were able to detect critical changes in vital signs in a timely manner. Discussion | Preliminary evaluation of the LSTAT in simulated and postoperative environments demonstrated that the LSTAT provided appropriate equipment to detect and manage critical events in patient care. Further work in assessing LSTAT functionality in a higher-acuity environment is warranted. Keywords: medical devices, patient simulations, transportation of patients, medical devices, patient simulations, transportation of patients, Introduction : Transport of critically ill patients to, between, and within hospitals can be associated with potentially adverse events . Researchers have shown that increased vigilance, appropriate equipment, and well-trained personnel can lead to improved safety while critically ill patients are being transported . Prior work evaluating the potential problems associated with transport of patients to intensive care units has led to the establishment of guidelines for the transport of the critically ill . The need for safe and effective patient care en route has also been a goal of the United States military. Because initial emergency life-saving surgery and prompt, aggressive resuscitation may have to be performed under austere field conditions to render a patient transportable and since resource limitations or adverse conditions may lead to delayed evacuation or prolonged evacuation times, the United States Army has developed a new critical care transport platform called the Life Support for Trauma and Transport (LSTATTM). Design goals for the LSTAT were solicited from medical personnel who have been deployed to combat zones, military and civilian medical personnel who transport and care for critically ill patients, and experts in military logistics with regard to medical equipment. Design goals included the following: weight limit of 120 pounds, volume not to exceed 22 x 72 x 13 inches (56 x 183 x 20 cm), battery power for up to 60 minutes, computer linkage of all the diagnostic and therapeutic equipment, capability of sending physiologic data to remote sites, and ability to generate pressurized gases for the ventilator. To meet these design constraints, the diagnostic and therapeutic equipment contained within the LSTAT had to be significantly reconfigured and miniaturized. The purpose of this study was to evaluate how modified equipment, configured to fit within the LSTAT, may affect (1) the identification and management of life-threatening physiologic derangements, using a patient simulator, and (2) ongoing monitoring of vital signs in a recovery room setting. Methods : Equipment | The LSTAT (model number 9602, Integrated Medical Systems, Inc, Signal Hill, CA, USA) consists of a pan and a stretcher. The pan contains commercially available equipment that has been reconfigured to fit in the 5-inch-deep (13-cm-deep) pan . The pan fits beneath and is attached to a NATO stretcher and has a head fairing that extends 7 inches above the stretcher . This equipment includes a transport ventilator, a 480-liter oxygen tank, a three-channel infusion pump, a defibrillator, a blood gas and blood chemistry analyzer, a suction device, a vital signs monitor, a computer, a power converter, and a battery power supply. The computer within the LSTAT continuously transmits physiologic data over a wireless network to a fixed large display called the clinical display and to a handheld notebook-computer-based display called the secondary display. Figure 1 | The United States Army's new critical care transport platform the LSTATTM (Life Support for Trauma and Transport) pan, without a NATO stretcher. The United States Army's new critical care transport platform the LSTATTM (Life Support for Trauma and Transport) pan, without a NATO stretcher. Figure 2 | The LSTATTM (Life Support for Trauma and Transport) head fairing. The LSTATTM (Life Support for Trauma and Transport) head fairing. Phase I: Evaluation of the LSTAT by clinicians using a patient simulator | After internal review board approval at the University of Utah and the Army Surgeons Human Subjects Research Review Board, 25 anesthesiologists and 6 recovery room nurses served as consenting volunteer clinicians to compare the clinical utility of the LSTAT with conventional monitoring systems, using a patient simulator. Each volunteer clinician was presented with a scripted description of the study methods and equipment to be used. The study description was read verbatim by the study proctor. The study proctor was the same person for all study participants. Equipment included the LSTAT standard equipment used for physiologic monitoring (Protocol Systems Inc, Model Propaq Encore, Beaverton, OR, USA), an E cylinder filled with oxygen (holding 660 l of oxygen at 2200 psi), a semiopen ventilation circuit (Vital Signs Inc, Resuscitation Circuit Model No. 5105 RV, Totowa, NJ, USA), an anesthesia machine (North American Drager, Model Narkomed AV2+, Telford, PA, USA), and a defibrillator (Hewlett Packard, model number 43110 A, McMinnville, OR, USA). The anesthesia machine contained a ventilator and a suction device. After reading the scripted instructions, each volunteer clinician was allowed to ask questions about the use of each piece of equipment. The study did not proceed until sufficient answers to all questions were given as determined by the volunteer clinician. A comparison was made of the training time required for volunteer clinicians to feel ready to use the LSTAT versus conventional monitors. The training time was defined as the time required by the study proctor to read the instructions plus the time required for each volunteer clinician to ask questions about its use. After having been trained, each volunteer clinician was presented with four scenarios in turn , using a patient simulator (Medical Education Technologies Inc, Sarasota, FL, USA). During two scenarios, the clinicians used the LSTAT and during the other two scenarios they used conventional equipment. The clinicians were randomly assigned to one of two groups. Group A used the LSTAT with scenarios 1 and 2 and conventional equipment with scenarios 3 and 4. Group B did the opposite. Table 1 | Simulation scenarios and key therapeutic maneuvers used to compare the Life Support for Trauma and Transport (LSTATTM) with conventional monitoring equipment Each volunteer clinician was required to state the diagnosis and the treatment needed during each scenario. Incorrect answers were ignored. If a clinician required more than 5 minutes to state the correct diagnosis or treatment, the scenario was stopped and recorded as incorrect. When the correct diagnosis was stated, the time was recorded. The time to treatment was recorded when the volunteer clinician stated the appropriate treatment and demonstrated the appropriate use of the equipment needed to implement it. The times to diagnosis and treatment with the two monitoring systems were compared using the Mann --Whitney test. After the simulations were over, each volunteer clinician completed a survey about use of the LSTAT. The survey asked about the alarm systems, ability to detect critical changes in vital signs, and utility of the LSTAT to manage patients if no other equipment were available. Phase II: Evaluation of the LSTAT in a routine postoperative setting | Nursing staff received a 45-minute training seminar in the use of the LSTAT, in which the system was placed on a modified wheel system (Stryker Medical, Big Wheel No. 1001, Kalamazoo, MI, USA). Ten consenting adult patients were each placed on an LSTAT in the operating room after surgery. A pulse oximeter probe, electrocardiogram leads, and blood pressure cuff were attached to the LSTAT. Each patient received supplemental oxygen via facemask. Patients were then transported to the recovery room. The course in the recovery room was noted for all events requiring intervention (e.g. deteriorating respiratory function requiring acute management of the airway, episodes of hypotension, hypertension, arrhythmias, postoperative nausea and vomiting, and inadequate pain control) as detected by the LSTAT. Measures of performance included both the number of postoperative events requiring intervention that were detected using the LSTAT's physiologic monitors and a survey of its utility taken from recovery room nursing staff, reviewing functionality, problems, and potential problems observed during clinical use. Results : Phase I: Evaluation of the LSTAT by clinicians using a patient simulator | Thirty-one volunteer clinicians participated in the simulator evaluation of the LSTAT. All of them had been trained in Advanced Cardiac Life Support (ACLS), 71% had up-to-date ACLS certification (recertification within the preceding 2 years), and 29% had been trained in Advanced Trauma Life Support (ATLS). The time required to provide instruction to each volunteer clinician ranged from 2 to 7 minutes. They required about 1 to 2 minutes more to learn how to use the LSTAT than to use conventional monitors. Table compares conventional monitoring equipment and the LSTAT with regard to the time required to reach a diagnosis and the time required to choose an appropriate treatment for anesthesiologists. The size of the group of recovery room nurses was not large enough to merit statistical analysis and therefore only the results from the survey are reported. Table 2 | A comparison of the time required to reach a diagnosis and proper treatment between the Life Support for Trauma and Transport (LSTATTM) and conventional monitoring equipment In scenarios 1, 2, and 3, there was no significant difference between the LSTAT and conventional monitors in the time required to reach a diagnosis or treatment. In scenario 4 (pulseless ventricular tachycardia), the time required to make the diagnosis and treatment was shorter than with the other scenarios regardless of which equipment (LSTAT or conventional monitors and equipment) was used; 23 out of 25 anesthesiologists made the diagnosis in less than 1 minute. The time required to treat the pulseless ventricular tachycardia was less with the conventional monitors than with the LSTAT. In scenarios 1, 2, and 3, the number of anesthesiologists unable to provide a correct diagnosis or treatment within 5 minutes was evenly distributed between the conventional monitor and the LSTAT groups. In scenario 4, all anesthesiologists provided the correct diagnosis and treatment within 5 minutes. Table shows the clinicians' response to the survey regarding the clinical usefulness of the LSTAT after completing four simulations. All the participating clinicians reported that they were able to properly manage the simulated patients using the LSTAT. All participating clinicians except one who abstained reported that if no other medical equipment were available in a remote setting, they would be able to provide appropriate care using the LSTAT. All of the survey respondents indicated that it was safe to proceed to the next phase of the study, in which the LSTAT would be used in a clinical setting. Table 3 | Summary of survey results collected from clinician volunteers who used the Life Support for Trauma and Transport (LSTATTM) in managing critical cardiopulmonary events using a patient simulator Survey questions aimed at exploring how useful the LSTAT was in managing critical events revealed that all of the participating clinicians were able to properly manage the simulated patients and 27 reported that they were able to detect critical changes in vital signs in a timely manner. Three clinicians, however, reported that they were not able to detect critical changes in vital signs because of difficulty seeing physiologic data on the display screens and the location of the displays. Several clinicians indicated that they would have liked more time to become familiar with the equipment before assuming patient care. A majority of the clinicians reported that suction and capnography would be useful during transport and that the controls on the LSTAT were easy to operate. In addition, the survey respondents reported that the configuration of equipment, as an intrinsic part of the stretcher, did not obstruct access to the patient. Features that were noted to be useful by respondents were the compactness of all the equipment in the LSTAT and the integration of a ventilator into the LSTAT to facilitate transport of ventilator-dependent patients. Phase II: Evaluation of the LSTAT in a routine postoperative setting | Ten patients were monitored on the LSTAT during their recovery from surgery. Complications experienced by this patient group associated with recovery from surgery included postoperative nausea and vomiting, inadequate pain control, hypertension, hypoxia, and tachycardia. Nine nurses who cared for patients using the LSTAT in the recovery room completed surveys and the results are presented in Table . Table 4 | Summary of survey results collected from recovery room nurses who used the Life Support for Trauma and Transport (LSTATTM) in managing patients recovering from surgery All nine of the recovery room nursing staff reported that the LSTAT provided adequate equipment to properly care for patients recovering from surgery. Five of the nine reported that it was easy to operate. The remaining four reported that they would have liked a more extensive in-service training before using the LSTAT. All those nurses who responded reported that all critical changes in vital signs were detected and addressed in a timely manner using the LSTAT. Limitations cited by the recovery room nursing staff included difficulty reading displays of vital signs on the secondary display and muted audible alarms that were difficult to hear in a recovery room environment. Three of the nine nurses reported that the large clinical display improved their ability to detect changes in vital signs and all but one reported that if no other equipment were available in a remote setting, they would be able to resuscitate a patient with the LSTAT. During transport of recovery room patients, no critical events were reported. Two of the nurses reported that the LSTAT provided an advantage during transport within the hospital and most reported that the secondary display was useful for monitoring vital signs during transport. Three reported that if the secondary display was not available, it would be difficult to monitor vital signs during transport if the patient was placed on the LSTAT so that the head fairing containing the physiologic monitoring and ventilator displays were at the foot of the bed. All the nurses that responded reported that the maneuverability of the LSTAT was adequate to enhanced. Overall comments by recovery room nursing staff suggested that the LSTAT would be helpful in patients with more highly acute conditions (e.g. in the intensive care unit, trauma bay, and prehospital settings) and that features that set the LSTAT apart from conventional equipment include compactness, readily available suction, capnography, defibrillator, and on-board oxygen tank. Finally, 22% of the recovery room nursing staff found the integration of the monitoring and therapeutic equipment into a stretcher for transport very useful, 67% found it to be useful, and 11% found it to be somewhat useful. Discussion : The LSTAT contains equipment typically found in an intensive care unit, integrated and miniaturized to fit beneath a transport stretcher. In addition, many of the components found within the LSTAT were reconfigured and made more rugged to meet design standards for use in military aircraft (minimize electrical emissions, withstand large changes in the ambient temperature, tolerate excessive vibration, and be insensitive to external electromagnetic interference). With these configuration changes in mind, we examined the clinical utility of the LSTAT in simulated and real clinical environments. Our hypothesis was that when equipment typically found in an intensive care unit was condensed to fit in a small space underneath a patient stretcher, the reconfigured equipment embodied in the LSTAT would be an equivalent tool to conventional equipment in detecting and treating life-threatening problems. Our results confirmed our study hypothesis. Our most important finding is that anesthesiologists and recovery room nurses, when asked to manage simulated critical events and care for patients after surgery, were able to provide appropriate care using the LSTAT. Phase I: Evaluation of the LSTAT by clinicians using a patient simulator | In this simulation phase of the study, the time required for clinicians to detect and treat life-threatening physiologic derangements was nearly identical using conventional equipment versus the LSTAT. In the simulations of tension pneumothorax, severe adult respiratory distress syndrome, and cardiac tamponade, the number of anesthesiologists unable to make a correct diagnosis or provide the correct treatment within 5 minutes was evenly divided between the two study groups. In the simulation of pulseless ventricular tachycardia, the median time to diagnosis and treatment was significantly shorter using conventional monitors than using the LSTAT. Although pulseless ventricular tachycardia is a life-threatening arrhythmia and merits immediate attention, the differences between the LSTAT and conventional monitors are subtle and may not be clinically important (49 versus 41 seconds in the time to diagnosis and 60 versus 44 seconds in the time to treatment). One potential source of delay for the time to treatment was that the defibrillator used in the conventional monitoring simulation was similar to one currently used in our hospital operating rooms, whereas the defibrillator incorporated into the LSTAT is not. It is important to point out that if a defibrillator is not readily available (the defibrillator was readily available in our simulations), the time required to locate one and treat the patient could be much longer than that reported in our simulation. In the tension pneumothorax and the adult respiratory distress scenarios, the simulated patient required mechanical ventilation. In the LSTAT group, ventilation was accomplished using the transport ventilator contained within the LSTAT. For the group using conventional monitors, ventilation was accomplished using a semiopen ventilation circuit that required manual operation. One difference reported by the volunteer clinicians was that they wanted to take the patient off the ventilator to hand ventilate the patient in order to validate their diagnosis, despite already having the peak airway pressures, delivered tidal volumes, and end tidal carbon dioxide levels readily available. This issue is not unique to the LSTAT ventilator. None of the commercially available transport ventilators has an auxiliary ventilator circuit that allows manual ventilation. Mechanical ventilation during transport of critically ill patients has been found to be advantageous over hand ventilation in meeting oxygenation and ventilation goals and in minimizing the acid-base disturbances that may lead to hemodynamic instability . In addition, even though hand ventilation was not available during simulated transports with the LSTAT ventilator, there was no difference in the time to diagnosis or treatment for either the tension pneumothorax or the severe adult respiratory distress scenarios. One potential criticism of the LSTAT is that it is too sophisticated and will require excessive training to teach clinicians how to use it. Our results did not validate this concern. Both recovery room nursing staff and the anesthesiologists required approximately 2 minutes more training time with the LSTAT than with conventional monitors. The overall training time never exceeded 7 minutes for the LSTAT. These results may be influenced by several factors. Because the patient transports were simulated, the volunteer clinicians may not have felt that they needed to pursue all the nuances about the LSTAT's equipment that they otherwise would have if they had been caring for a real patient. Secondly, the volunteer clinician group studied has significant experience with various types of patient monitors, ventilators, and defibrillators and may not have required as much teaching time as would other clinicians who are not as routinely involved with these items. Phase II: Evaluation of the LSTAT in a routine postoperative setting | The LSTAT was judged by nursing staff to be adequate for the management of patients recovering from surgery. Complications experienced by the patient group were typical of complications associated with recovery from surgery. Features that set the LSTAT apart from routine monitoring of patients in the recovery room included the readily available defibrillator, availability of suction and capnography for transport, a built-in oxygen source, the fixed large clinical display of the patient's vital signs in the recovery room, and the mobile secondary display which reported the patient's vital signs for use during transport. After having used the LSTAT in the recovery room, the nursing staff was asked to critique the use of the LSTAT in managing critical events. Data visualization and visual and auditory alarms were of primary concern both with the physiologic monitor and the ventilator. Some nurses were concerned that they would not be able to detect critical changes in vital signs because of these limitations. This problem may be the result of two conflicting design goals: the military needs (low sound and low light emissions) and the needs of the intensive care unit (visual physiologic data presentation and loud auditory signals and alarms). The screens contained within the LSTAT were selected to reduce power consumption and minimize light emission. Potential solutions to this concern include enhanced training with the LSTAT to improve clinicians' comfort with the existing data displays and alarm systems as well as exploration of alternatives for data and alarm presentation to improve the clinician's awareness of a patient's status. Five of the nine recovery nurses reported that they did not notice any particular advantage of the LSTAT during transports within the hospital. In this phase of the study, the LSTAT was used to transport patients from the recovery room to the ambulatory surgery discharge area or to a hospital bed, a transport routinely done without any patient monitoring. Thus in less acute transport settings, it is reasonable that the LSTAT would not provide any significant advantage. An expressed concern of some of the military product developers was that the weight of the LSTAT pan and the configuration of the LSTAT as dictated by the size constraints to fit in military aircraft would make the LSTAT awkward to use during patient care. The survey results did not support this potential concern. For example, respondents indicated that the LSTAT was easy to operate, access to the patient was not obstructed, and the LSTAT was easy to maneuver. Many respondents reported that consolidation of all the physiologic monitoring equipment and incorporation of a transport ventilator were all advantages for patient transport. After the simulator study, all respondents indicated that it was safe to proceed to the next phase of the study, in which the LSTAT would be used in a clinical setting. Furthermore, volunteer clinicians reported through their surveys that they were able to properly manage the simulated and real patients using the LSTAT. All but one clinician reported that if no other equipment were available in a remote setting, they would be able to resuscitate a patient with the LSTAT. This report represents a preliminary evaluation of the LSTAT in a clinical setting. The goal was to validate the functionality of the LSTAT before its evaluation in more acute settings such as intensive care units, emergency departments, transports within hospitals, within medical evacuation vehicles during transport between hospitals, and eventually in remote areas where medical resources are limited or unavailable. A logical next step is to evaluate the LSTAT during the initial management of critically injured trauma patients as they present for evaluation in an emergency room trauma bay. This might be best accomplished in a facility designed for and staffed by specialists trained in trauma patient care. The same study hypothesis might be: does equipment typically found in an intensive care unit, condensed to fit in a small space underneath a patient stretcher, serve as an equivalent tool to conventional equipment in detecting and treating life-threatening problems? Additional questions may incoude the following: Does the LSTAT reduce the personnel and resources needed for intra-hospital transport for emergency imaging (e.g. computer tomography scans or angiography studies), rapid transfer to the operating room, or transfer to the intensive care unit? Does remote monitoring of a patient during intrahospital transport improve the clinician's vigilance in detecting life-threatening problems that may develop during transport? Can personnel other than anesthesiologists and recovery room nurses learn to use the medical devices contained within the LSTAT effectively? And finally, does the integration of physiologic data, ventilator data, arterial blood gas and chemistry data, and clinical data into an optimized computer-based display help clinicians evaluate patients more efficiently and make more informed decisions when caring for patients with multiple life-threatening injuries ? The clinical relevance of this line of investigation is a function of the prevalence of trauma in our world today and the need to provide life-saving intervention quickly after injury. Experience in major metropolitan areas where evacuation times are quick and state-of-the-art surgical and resuscitative resources and well-trained personnel are readily available has established the benefit of early surgical intervention and resuscitation on survival . The LSTAT was designed to provide equipment for underserved areas where conventional intensive care resources may be unavailable. Thus, the LSTAT may serve as a critical resource to a highly mobile surgical team because it can be placed very near the site of injury in an effort to reduce the time from injury to life-saving intervention. The LSTAT provides the equipment necessary for appropriate postoperative care of a critically injured patient, for transport to tertiary care facilities, or for holding patients until evacuation is feasible. The LSTAT can also serve as a resource to resuscitate patients who do not require surgery but who do require intensive care. Key messages : Medical equipment reconfigured and miniaturized into a stretcher-based portable intensive care unit (called the Life Support for Trauma and Transport [LSTATTM]) was evaluated using a patient simulator and during patient care in a recovery room setting In the simulation phase of the study, volunteer clinicians compared the LSTAT with conventional monitors while managing critical events In the recovery room phase of the study, nurses critiqued the LSTAT while caring for patients after surgery In both the simulated and postoperative environments, the LSTAT provided appropriate equipment to detect and manage critical events in patient care Competing interests : This study was supported in part by a grant from the United States Army Medical Research and Material Command. LSTAT is a trademark of the United States Army. Abbreviations : LSTATTM = Life Support for Trauma and Transport [platform]; NATO = North Atlantic Treaty Organization; psi = pounds per square inch. Backmatter: PMID- 12398786 TI - Antithrombin III in patients admitted to intensive care units: a multicenter observational study AB - Abstract | Introduction | The administration of antithrombin III (ATIII) is useful in patients with congenital deficiency, but evidence for the other therapeutic indications of this drug is still uncertain. In Italy, the use of ATIII is very common in intensive care units (ICUs). For this reason we undertook an observational study to determine the pattern of use of ATIII in ICUs and to assess the outcome of patients given this treatment. Methods | From 20 May to 20 July 2001 all consecutive patients admitted to ICUs in 20 Italian hospitals and treated with ATIII were enrolled. The following information was recorded from each patient: congenital deficiency, indication for use of ATIII, daily dose and duration of ATIII treatment, outcome of hospitalization (alive or dead). The outcome data of our observational study were compared with those reported in previously published randomized controlled trials (RCTs). Results | Two hundred and sixteen patients were enrolled in the study. The clinical indications for using ATIII were sepsis (25.9%), disseminated intravascular coagulation (23.1%), and other clinical conditions (46.8%). At the end of the study, 65.3% of the patients were alive, 24.5% died and 10.2% were still in the hospital. Among the patients with sepsis (n = 56), 19 died during the observation period (33.9%; 95% confidence interval 22.1 --47.5%). Discussion | Our study described the pattern of use of ATIII in Italian hospitals and provided information on the outcome of the subgroup treated with sepsis. A meta-analysis of current data from RCTs, together with our findings, indicates that there is no sound basis for using this drug in ICU patients with sepsis. Keywords: antithrombin III, disseminated intravascular coagulation, sepsis, septic shock, antithrombin III, disseminated intravascular coagulation, sepsis, septic shock, Introduction : Antithrombin III (ATIII) is a recognized treatment for patients with congenital ATIII deficiency (see also the approval of this indication by the Food and Drug Administration); in contrast, the evidence supporting its use for other clinical indications is uncertain . In Italian hospitals this drug is widely used in patients admitted to intensive care units (ICUs), who are generally given ATIII for the treatment of sepsis or disseminated intravascular coagulation (DIC). The approval of ATIII by the Italian Ministry of Health was granted nearly 10 years ago (before the profound reform of the Drug Regulatory Agency made by the Italian Ministry of Health in 1993) and has remained unchanged since then. This approval of ATIII was rather generic and included 'congenital deficiency of ATIII and all clinical conditions that can cause an acquired deficiency of ATIII'. Three small randomized studies and one large international trial assessed the effectiveness of ATIII in sepsis, but none of these trials found a significant benefit in terms of reduced morbidity or mortality. As regards congenital deficiency, the effectiveness of ATIII is fairly well documented , but these patients are rare. The other clinical indications (such as acute thrombosis or thromboembolism, prevention of DIC in hepatic coma, and treatment of bleeding episodes in cirrhosis) are supported by a small series of very preliminary studies (see, for example, the Drugdex databank, CD-ROM Drugdex, volume 110; Micromedex, Englewood, Colorado, USA). To achieve a better definition of the current use of ATIII in Italian hospitals and to generate naturalistic data (based on routine practice) about the outcome of this treatment, we undertook a multicenter observational study. Methods : Design of the study and aims | The study was based on a multicenter observational design. From 20 May to 20 July 2001 all consecutive patients admitted to ICUs in 20 Italian hospitals and treated with ATIII were enrolled in the study. The study had the following aims: (1) surveying the use of ATIII in patients admitted to ICUs; (2) determining the outcome of patients treated with ATIII; and (3) comparing the results obtained from our observational study with those previously found in the randomized controlled trials (RCTs). A meta-analysis was also conducted to summarize the information deriving from four RCTs that studied the effectiveness of ATIII in sepsis. Data collection | The following information was recorded from each patient enrolled in the study: (1) demographic characteristics (age, sex, weight); (2) congenital deficiency (y/n); (3) baseline ATIII level; (4) ward of first admission in the hospital; (5) clinical indication for using ATIII (sepsis or DIC or any other clinical condition); (6) daily dose and duration of treatment with ATIII; (7) outcome of hospitalization (alive or dead); and (8) concurrent administration of antibiotics and/or heparin. Analysis | The information collected from each patient was analyzed by standard descriptive statistics. In the subgroup of patients with sepsis, the in-hospital mortality rate observed in our study was compared with that previously reported by the four RCTs. All rates were presented together with their 95% confidence interval (CI), which was calculated by using Equations 1.26 and 1.27 of Fleiss . Results : The overall number of patients who were admitted to ICUs during the study period was 1648. Of these patients, 216 (13%) were enrolled in our study. The characteristics of these 216 patients are presented in Table . Table 1 | Characteristics of the 216 patients enrolled in our observational study and of the subgroup of 56 patients treated for sepsis The clinical indication for using ATIII was sepsis (n = 56), DIC (n = 50), or other (n = 101). Table also reports separate information for the subgroup of 56 patients treated for sepsis. The duration of ATIII therapy did not differ at levels of statistical significance between patients treated for different clinical indications (P = 0.57 according to an analysis of variance). The daily dose of ATIII showed a difference between sepsis and other indications . Table reports the outcome of hospitalization according to clinical indication. With regard to the use of ATIII in patients with sepsis, Figure shows the percentage mortality rate (with 95% CI) observed in our study, together with the rates found in four previous studies . Figure 1 | Percentage mortality rate (with 95% CI) of patients with sepsis: comparison between the results of our observational study and those reported in the four RCTs previously published. Percentage mortality rate (with 95% CI) of patients with sepsis: comparison between the results of our observational study and those reported in the four RCTs previously published. Solid lines, treatment groups; broken lines, control groups; dates of publication: 1993, Fourrier et al. ; early 1998, Eisele et al. ; late 1998, Baudo et al. ; 2001, Warren et al. and our study. Table 2 | Relationship between clinical indication for the use of ATIII and outcome of hospitalization Subgroup analyses within the patient cohort of our study did not identify any relationship between mortality and patient characteristics. The administration of heparin, which Warren et al. found to have some implications for outcome, did not influence mortality in our patient series: mortality was 19.6% in the 107 patients who received heparin, compared with 30.5% in the 95 patients who did not receive this drug (P = 0.10) by Fisher's exact test; mortality was 28.6% in the 28 patients with sepsis who received heparin, compared with 42.3% in the 26 patients with sepsis who did not receive this drug (P = 0.39). Discussion : The main scientific value of our observational and prospective study lies in its naturalistic design; the population of patients that we studied was in fact drawn from the everyday practice of more than 20 hospitals and was intentionally free from specific exclusion criteria. In interpreting our outcome data, one disadvantage is that the group treated with ATIII was not compared with any reference group observed prospectively within our research; neither did we include any retrospective control group not treated with the drug. However, historical retrospective controls would have raised profound problems of matching the retrospective data with the prospective ones. A prospective enrollment of controls not treated with ATIII was not feasible because the therapeutic policy of the ICUs involved in our study was to administer ATIII to virtually all patients with a diagnosis of sepsis or DIC. Regardless of our statistical indexes, a 'first-look' comparison between the data on sepsis produced by the previous RCTs (including four treatment groups and four control groups) and those observed in our naturalistic study indicates a complete overlap of the various survival rates and of their respective 95% CIs. This qualitative impression is in agreement with the meta-analysis shown in Figure (see Appendix2 for details of its methodology). Figure 2 | Comparison of death rates between patients given ATIII and patients given placebo in the four RCTs that met the inclusion criteria of out meta-analysis. Comparison of death rates between patients given ATIII and patients given placebo in the four RCTs that met the inclusion criteria of out meta-analysis. The odds ratios of the individual studies and of our meta-analysis are denoted by dots and by a diamond, respectively; each horizontal bar indicates the 95% CI for the odds ratio, and the vertical dotted line represents the identity line. From top to bottom, datasets are the trials of Warren et al. , Baudo et al. , Eisele et al. , and Fourrier et al. ; the bottom dataset is our meta-analysis. In the four RCTs, the crude death rates in the treatment group and in the control group, respectively, were as follows: Fourrier et al. , 7 of 17 versus 9 of 18; Eisele et al. , 5 of 20 versus 9 of 22; Baudo et al. , 31 of 50 versus 33 of 48; Warren et al. , 450 of 1157 versus 448 of 1157. This meta-analysis gave the following results: summary odds ratio 0.98; 95% CI 0.83 --1.15, P = 0.80; chi2 for heterogeneity 1.86; 3 degrees of freedom; P = 0.60. In this meta-analysis, the large-scale trial by Warren et al. outweighed the other three small RCTs in that Warren's trial included 93% of the overall cohort of the four RCTs. In the light of the above data, there seems to be no clinical benefit in administering ATIII to critical patients with sepsis; in this context, one crucial point is that the most recent large-scale trial gave very clear results and was negative. The other clinical indications reported in our patients' series were more difficult to interpret because of the nearly complete lack of previous controlled studies exploring these therapeutic issues. There has been a lively debate in the literature on the relative merits of observational studies and RCTs in providing useful evidence of clinical effectiveness . Although the great majority of researchers stick to the concept that RCTs are the gold standard, common sense suggests that having information both from RCTs and from observational studies is better than having information from RCTs only. In this framework, our study advances knowledge about the use of ATIII in critical patients. In conclusion, our findings based on an observational prospective study and on an updated meta-analysis of the previous RCTs do not support the use of this drug in ICU patients with sepsis. Key messages : Antithrombin III (ATIII) is a recognized treatment for patients with congenital ATIII deficiency; in contrast, the evidence supporting its use for other clinical indications is uncertain. In Italian hospitals this drug is widely used in patients admitted to intensive care units (ICUs), who are generally given ATIII for the treatment of sepsis or disseminated intravascular coagulation. Three small randomized studies and one large international trial have assessed the effectiveness of ATIII in sepsis, but none of these trials has found a significant benefit in terms of reduced morbidity or mortality. Our findings, based on an observational prospective study and on an updated meta-analysis of the previous randomized controlled trials, do not support the use of this drug in ICU patients without congenital deficiency. Competing interests : In 2001 our research group received a grant from Eli-Lilly (Italy) to conduct an original study on factors influencing length of stay in critical patients with sepsis. In Italy, anti-thrombin III is marketed by Aventis-Behring and by Baxter. Appendix 1: Gruppo di Studio sull'antitrombina III (The Antithrombin Study Group) : The Antithrombin Study Group includes the study coordinators (A Messori, F Vacca, M Vaiani, S Trippoli, Laboratorio di Farmacoeconomia, c/o Azienda Ospedaliera Careggi, Firenze) and a total of 51 participants. The names and addresses of the participants involved in the project were the following (all located in Italy): R Banfi, M Cecchi, E Cini, D Dupuis, T Falai, R Fornaini, A Ipponi, ML Migliaccio, F Pelagotti, L Rabatti, I Ruffino, R Silvano, E Tendi (Firenze, four hospitals); P Becagli, M Monciatti (Empoli); B Bozzone, R Casullo, F Cattel, S Pardossi, R Passera, S Stecca, U Tagliaferro (Torino, two hospitals); P Di Bartolomeo, T Faggiano, M Lattarulo (Bari); N Caboni, A Cannas (Cagliari); A Plescia, M Sorci (Rimini); L Bonistalli, M Puliti (Prato); B Ciammitti, M Costantini, F Mammini (Terni); L De Cicco, G Mazzaferro (Napoli); P Marrone, R Tetamo (Palermo); P Beneduce, MG Celeste, P Fiorani, S Galeassi, G Guaglianone, A Pecere, L Ragni (Roma, two hospitals); SM Germinario (Andria); O Basadonna, L Todesco (Camposampiero, Padova); R Calle-gari, M Pegoraro (Asolo); E Lamura (Ancona). Appendix 2: Methodology of the meta-analysis : A MedLine search (PubMed, ) was performed to cover the period from January 1980 to November 2001. The search was limited to the studies published in English and was based on four index terms combined with the following Boolean syntax: "antithrombin III" AND (sepsis OR septic shock OR "disseminated intravascular coagulation"). This search was supplemented by examining the Drugdex databank (CD-ROM Drugdex, volume 110; Micromedex, Englewood, Colorado, USA). Eligible studies were included if they met the following criteria: patients were admitted to an ICU; randomized design; diagnosis of sepsis, septic shock or DIC; assessment of survival. The odds ratio was used as the main index to assess the treatment effect within each trial and to generate the overall results of the meta-analysis. The calculation of the summary odds ratios was based on a random-effect model . Heterogeneity was assessed as described previously . Abbreviations : ATIII = antithrombin III; CI = confidence interval; DIC = disseminated intravascular coagulation; ICU, intensive care unit; RCT = randomized controlled trial. Backmatter: PMID- 12398788 TI - Case report: Survival after deliberate strychnine self-poisoning, with toxicokinetic data AB - Abstract | Introduction | Strychnine poisoning is uncommon, and in most severe cases, the patient dies before reaching hospital. The management of strychnine poisoning is well documented, although there are few data on the kinetics of elimination of strychnine after overdose. Case report | A 42-year-old man presented shortly after ingestion of an unknown quantity of strychnine powder. After a respiratory arrest, with intensive supportive management requiring admission to an intensive care unit, he survived. Eight serum samples were taken over the first 5 days and analysed subsequently for strychnine concentrations. Results | The initial concentration at 1.5 hours after ingestion was 4.73 mg/l, falling to 0.38 mg/l at 74 hours postingestion. Serum concentrations followed a monoexponential elimination curve with a calculated elimination half-life of 12 hours. Discussion and conclusion | Strychnine poisoning presents with classical features, and with early diagnosis and supportive management, the patient can survive. The initial serum concentration of 4.73 mg/l is the highest reported concentration in a patient who has survived. Previous reports of the elimination half-life have suggested it is between 10 and 16 hours, which conforms to the elimination data in our case. Keywords: poisoning, respiratory arrest, strychnine, survival, toxicokinetics, poisoning, respiratory arrest, strychnine, survival, toxicokinetics, Introduction : Strychnine is a bitter, white, powder alkaloid derived from the seeds of the tree Strychnos nux-vomica. It has had an interesting past, initially having been introduced in the 16th century as a rodenticide, and until recently it was used as a respiratory, circulatory and digestive stimulant . It is no longer used in any pharmaceutical products, but is still used as a rodenticide. Strychnine is also found as an adulterant in street drugs such as amphetamines, heroin and cocaine . Strychnine poisoning is uncommon, and in most severe cases, the patient dies before reaching hospital . The management of strychnine poisoning is well documented, but there is less information available about the elimination kinetics of strychnine. We report the toxicokinetics of strychnine in a patient who survived a deliberate strychnine ingestion. Case report : A 42-year-old man with no significant past medical history presented approximately 1 hour after ingestion of a bottle of wine together with some "white powder" from his garden shed (this was later confirmed to be strychnine). He was able to walk in to the Emergency Department, but he was agitated and ataxic, in keeping with his ethanol intake. Within a few minutes of his arrival, his condition rapidly deteriorated and he developed a marked tremor and muscular spasms and shortly after this had a respiratory and secondary cardiac arrest. At this stage he was intubated and ventilated, and cardiac output returned after 5 minutes. He was transferred to the intensive care unit (ICU) and the National Poisons Information Service (London) was contacted for further advice on management. He continued to have marked muscle spasms and so was paralysed with 0.1 mg/kg pancuronium given intravenously. He was given 50 g activated charcoal via a nasogastric tube. Postarrest, his blood pressure was 85/40 mmHg, heart rate 96 beats/min and temperature 38.2C. Arterial blood gases showed a severe metabolic acidosis (pH6.51, PaO2 9.5 kPa, PaCO2 2.6 kPa, HCO3 3.7 mmol/l, base excess [BE] -18) and he was given 3 mmol/kg 8.4% sodium bicarbonate; his metabolic acidosis improved over the next 8 hours (pH7.26, PaCO2 5.35 kPa, PaO2 13.4 kPa, HCO3 18 mmol/l, BE -9). He remained hypotensive despite fluid resuscitation and over the first 24 hours he required norepinephrine to maintain his blood pressure (maximum dose 900 g/h). His temperature rose to 39.9C on day2 but settled after simple cooling measures and rehydration. His creatinine kinase peaked at 8218 IU/l, (although there was no evidence of myoglobinuria) and his creatinine peaked at 194 mol/l on day 2. He was extubated on day3, initially with some persisting twitching and muscular spasms, requiring boluses of midazolam and diazepam and an alfentanil infusion for analgesia. By day5 he was asymptomatic and his renal function had returned to normal. He was discharged from hospital on day10 after psychiatric assessment. Results : Eight serum samples were obtained over the first 5 days and subsequently analysed on a Hewlett-Packard 6890 gas chromatograph (Stockport, Cheshire, UK) equipped with a nitrogen phosphorous detector for strychnine concentrations. A Restex RTX 1 column was used with an isothermal oven temperature of 270C. The injector and detector temperatures were set at 250C and 300C, respectively. The strychnine concentrations are shown in Fig. ; the patient's initial concentration at 1.5 hours after ingestion was 4.73 mg/l, falling to 0.38 mg/l and to 0 at 74 hours and 100 hours postingestion, respectively. The data conformed to a monoexponential elimination curve and the calculated elimination half-life was 12 hours. Figure 1 | Serum strychnine concentrations in a patient with severe strychnine poisoning. Serum strychnine concentrations in a patient with severe strychnine poisoning. Discussion : Strychnine poisoning is an uncommon but potentially fatal poisoning and most patients die before reaching hospital. In the case presented here, the patient presented early after ingestion. He was initially asymptomatic but rapidly developed severe muscle spasms leading to respiratory arrest. With meticulous supportive care in the ICU he made a complete recovery. The mechanism of action causing strychnine toxicity is well understood. This poison is a competitive antagonist of the inhibitory neurotransmitter glycine at receptors in the spinal cord, brain stem and higher centres . It results in increased neuronal activity and excitability, leading to increased muscular activity. The classical features of strychnine poisoning occur from 15 to 30 minutes after ingestion and include heightened awareness, muscular spasms and twitches and hypersensitivity to stimuli. In large ingestions, these can progress to painful generalised convulsions, during and after which the patient retains consciousness. The cause of death is usually respiratory arrest secondary to respiratory muscle spasms , although prolonged muscular spasm can lead to hyperthermia, rhabdomyolysis and associated renal failure due to myoglobinuria . Severe metabolic acidosis can occur due to increased lactate levels following repeated muscular activity, although the lactate is rapidly removed once the muscular spasms have been controlled . Prompt recognition of poisoning and initiation of treatment are required to prevent deterioration and death. Control of convulsions and muscular spasms requires either high doses of benzodiazepines, such as 1 mg/kg diazepam , or, in resistant cases, paralysis with neuromuscular blockers such as pancuronium . In addition, meticulous supportive care is required, and careful monitoring and observation for complications such as acute renal failure. Clinical effects of strychnine ingestion have been seen at doses as small as 2 mg in a child and 20 mg in an adult . Survival after ingestion of large amounts has been reported: 480 mg (24 mg/kg) in a child and 3750 mg in an adult . Death from strychnine ingestion has been reported in adults after doses as low as 5 to 10 mg . Since severe, life-threatening features have been reported after the ingestion of small amounts and patients can deteriorate rapidly, strychnine ingestion should be regarded as potentially fatal and the patient should be admitted to an ICU for close observation and appropriate treatment. In this case, as with many other reported cases, both the concentration of strychnine in the preparation ingested and the amount taken were unknown. The serum strychnine concentration of 4.73 mg/l at 1.5 hours postingestion is the highest reported concentration that a patient has survived. In previous reports, concentrations of 2.1 mg/l at 3 hours and 3.8 mg/l at 0.5 hours caused death. The potential fatal serum concentration of strychnine is not known and clinical features rather than the serum concentration should guide management. At a postmortem, less than 4 hours after an intentional ingestion of strychnine, a serum concentration of just 0.33 mg/l was found . A postmortem study showed that a strychnine concentration of 3.32 mg/l in the blood of the inferior vena cava was associated with concentrations of 11.4 mg/l, 2.42 mg/kg, 2.32 mg/kg and 98.6 mg/kg in bile, brain, skeletal muscle and liver, respectively . Of the 50 reported cases of strychnine poisoning in the past 36 years, only two nonfatal cases have included data on serum strychnine concentrations . Both of these patients had much lower concentrations (1.6 mg/l at 4 hours after ingestion and 2.17 mg/l at 6 hours after ingestion, respectively) than that found in our patient. The elimination and metabolism of strychnine are poorly understood. Strychnine is rapidly absorbed from the gastrointestinal tract, with symptoms occurring within 10 to 20 minutes of ingestion . It is also absorbed through other mucous membranes, as in poisonings when strychnine has been mistaken for cocaine . Strychnine has been shown in animal studies to be metabolised in the liver by cytochrome P-450 microsomal enzymes , but 1 --30% is excreted unchanged in the urine . The proportion of strychnine excreted unchanged appears to be smaller when larger amounts are ingested . In our patient, elimination of strychnine obeyed first-order kinetics with a half-life of 12 hours. Only three other papers have reported on the elimination kinetics of strychnine, and these provide conflicting views of its toxicokinetics. Edmunds and colleagues were the first authors to describe the elimination of strychnine and theirs is the only other report on the kinetics of strychnine in a patient who survived . They reported a 42-year-old man who ingested an unknown quantity of strychnine and developed severe muscle cramps and convulsions and had a respiratory arrest 35 minutes after ingestion. In that case, the elimination of strychnine obeyed first-order kinetics with a half-life of 10 hours, based on 19 blood samples collected over a 53-hour period, the first of which was taken at 4 hours after ingestion. Palatnick et al. also reported first-order kinetics with a half-life of 16 hours, based on 18 serum concentrations over a 51-hour period in a 34-year-old man after ingestion of 125 ml of a 2% strychnine solution . Heiser et al. reported zero-order kinetics in a fatal case in a 51-year-old man who presented after ingesting 4.8 g of strychnine . His serum concentrations were measured in only five samples, ranging from 3.5 mg/l at 0.5 hours after ingestion to <0.1 mg/l at 43 hours after ingestion. The differences in the kinetics of strychnine found in our patient from the kinetics in the above three cases almost certainly reflect the influence of clinical parameters such as different periods of hypotension, metabolic acidosis and renal impairment, all of which would influence hepatic metabolism and renal elimination of strychnine. However, the calculated half-life of 12 hours in our patient is comparable to the half-lives previously reported (10 hours and 16 hours), suggesting that the true elimination half-life of strychnine lies between 10 and 16 hours. Conclusion : We describe a case of severe strychnine poisoning with a favourable outcome in which the strychnine half-life was 12 hours. The patient presented early, his symptoms were recognised and meticulous supportive care was instituted without delay. All of these factors were probably important in preventing what would otherwise have been a fatal ingestion of strychnine. Key messages : Strychnine overdose is rare but potentially fatal With prompt recognition of the characteristic features and rapid supportive care, survival is possible Ingestion of small amounts and/or low plasma strychnine concentrations are associated with fatality Strychnine has a toxicokinetic half-life in overdose of 12 hours Elimination of strychnine in our patient obeyed firstorder kinetics, as has previously been reported Competing interests : None declared. Abbreviations : BE = base excess; ICU = intensive care unit. Backmatter: PMID- 12398787 TI - Prone positioning does not affect cannula function during extracorporeal membrane oxygenation or continuous renal replacement therapy AB - Abstract | Introduction | Prone positioning in respiratory failure has been shown to be a useful adjunct in the treatment of severe hypoxia. However, the prone position can result in dislodgment or malfunction of tubes and cannulae. Certain patients receiving extracorporeal membrane oxygenation (ECMO) or continuous renal replacement therapy (CRRT) may also benefit from positional therapy. The impact of cannula-related complications in these patients is potentially disastrous. The safety and efficacy of prone positioning of these patients has not been previously reported. Materials and methods | A retrospective chart review evaluated ECMO or CRRT cannula location, and displacement or malfunction during positional change or while prone. The study was set in a General Surgery and Trauma Intensive Care Unit. The subjects were all patients at our institution who simultaneously underwent ECMO or CRRT and prone positioning from July 1996 to July 2001. There were no interventions. Results | Ten patients underwent ECMO and 42 patients underwent CRRT during the study period. Seven patients underwent simultaneous prone positioning and either ECMO (4/10) or CRRT (4/42). A total of 68 turning events (prone to supine or supine to prone) were recorded, with each patient averaging 9.7 (range, 4 --16) turning episodes. Turning was performed with sheets and extra nursing personnel; no special mechanical assist devices were used. No patients experienced inadvertent cannula removal during turning. Two patients had poor flow through their cannulae. In one patient, this occurred in the supine position and required repositioning of the cannula. In the second patient, cannulae were changed twice and flow was poor in both the supine and the prone positions. All ECMO and CRRT patients received venous cannulae. Cannula location (seven internal jugular and 11 femoral) did not the affect risk of malfunction. Discussion and conclusions | Patients with venous cannulae for ECMO or CRRT can be safely placed in the prone position. Flow rates are maintained in this position. Potential cannula complications of ECMO and CRRT are not a contraindication to prone positioning in severely ill patients. Keywords: continuous renal replacement therapy, extra-corporeal membrane oxygenation, positional therapy, prone positioning, renal replacement therapy, safety, continuous renal replacement therapy, extra-corporeal membrane oxygenation, positional therapy, prone positioning, renal replacement therapy, safety, Introduction : Prone positioning for respiratory failure has recently gained popularity as an adjunct for the treatment of respiratory failure and adult respiratory distress syndrome. High-risk patients who may benefit from prone positioning include some patients with large-bore, high-flow-access cannulae. This includes patients on extracorporeal membrane oxygenation (ECMO) and continuous renal replacement therapy (CRRT), such as continuous venovenous hemofiltration and dialysis. Cannula-related complications in these patients are potentially disastrous and the safety of turning patients with these types of cannulae has not been previously demonstrated. Materials and methods : All patients admitted to the intensive care unit from July 1996 to July 2001 who underwent prone positioning while receiving either ECMO or CRRT were evaluated. Demographic data were recorded as well as the number of turns, the location of the cannulae and cannula displacement or malfunction as related to positioning. The turning technique used for all of these patients requires only sheets and extra personnel (Figs ,,,,); no mechanical assist devices are used. Access cannulae and tubing are brought off the ends of the bed to provide coaxial rotation. An intensivist, a respiratory therapist, and multiple nurses are present for all turning events. Vital signs are monitored closely before and after the turn to ensure that the patient is tolerating the position change. Figure 1 | A typical patient at our institution undergoing prone positioning. A typical patient at our institution undergoing prone positioning. The abdomen is open due to a gunshot wound. The patient requires multiple vasopressors, continuous venovenous hemofiltration and inhaled nitric oxide. Figure 2 | Adequate padding, especially of the face, is mandatory. Adequate padding, especially of the face, is mandatory. An operative pillow with a cutout for the endotracheal tube is used. Prior to disconnecting the endotracheal tube, it is clamped to prevent loss of positive end expiratory pressure. Figure 3 | The patient is tightly rolled in two sheets and is moved to the far side of the bed, away from the ventilator. The patient is tightly rolled in two sheets and is moved to the far side of the bed, away from the ventilator. Figure 4 | The patient is rolled into an extreme lateral position, facing the ventilator, with close monitoring of the hemodynamics. The patient is rolled into an extreme lateral position, facing the ventilator, with close monitoring of the hemodynamics. As the turn is completed, transverse rolls are place under the chest and pelvis to allow free abdominal excursion. Figure 5 | Positioning is completed with chest and pelvis rolls in place, and the arms flexed at the elbows and in the neutral position at the shoulders. Positioning is completed with chest and pelvis rolls in place, and the arms flexed at the elbows and in the neutral position at the shoulders. The arm position is changed every 2 hours and automated bed rotation is continued in the prone position. The feet are elevated with ankle rolls to prevent pressure breakdown. Results : During the study period, 10 patients underwent ECMO and 42 patients underwent CRRT. Seven patients underwent simultaneous prone positioning and either ECMO (n = 3) or CRRT (n = 4). Table demonstrates the demographics, the disease process and the outcome of the patients. A total of 68 turning events (prone to supine or supine to prone) were recorded, with each patient averaging 9.7 (range, 4 --16) turning episodes. No patients experienced inadvertent cannula removal during turning. Two patients had poor flow through their cannulae unrelated to prone position or turning. In one patient, this occurred in the supine position and required repositioning of the cannula. In the second patient, cannulae were changed twice and flow was poor in both the supine and the prone positions. All ECMO and CRRT patients in this series received venous cannulae. The cannula location (seven internal jugular and 11 femoral) did not affect the risk of malfunction. One internal jugular cannula was repositioned, and one was replaced for poor flow. One femoral cannula was replaced for poor flow. Table 1 | Demographics, disease process and outcome of patients Discussion : Prone positioning for respiratory failure has been shown to increase oxygenation when used as adjunctive therapy for respiratory failure and adult respiratory distress syndrome. This has resulted in an increase in the use of prone positioning worldwide, with numerous studies of its effects. Recent studies, however, have not demonstrated a decrease in mortality with this modality . The act of turning patients prone, and the prone position itself, is not without risk. These patients tend to require high levels of ventilatory and hemodynamic support, and are dependent on endotracheal tubes and monitoring cannulae, as well as on intravenous inotropic infusions. In addition, these patients are heavily sedated and often paralyzed, resulting in their inability to shift position to prevent pressure necrosis or neurologic injury from poor positioning. Hence, the choice to use prone positioning as a therapy must be weighed against the potential risks of the turning and the position. Prone positioning and turning have been reported to result in complications in 32% of prone cycles. Most of these are related to skin pressure necrosis. Inadvertent extubation and central line decannulation are two of the more disastrous complications that have been reported . Our group has previously reported the safety of prone positioning in high-risk patients, such as those with open abdomens . Other similarly high-risk patients with large-bore vascular cannulae may not undergo prone positioning due to fear of cannula complications, including patients on ECMO and CRRT. The safety of turning patients with these types of cannulae has not been systematically evaluated. There are reports of individual cases of prone positioning in patients with continuous venovenous hemofiltration therapy . The present results indicate that prone positioning with these cannulae can be carried out safely and does not significantly affect the function of the high-flow systems. This again expands the patient population in which prone positioning is potentially beneficial. The location of high flow catheters is not related to complication or malfunction rate, thus all sites can be safely used for access. The outcome of the present group of patients was poor, with 57% mortality. This is not surprising given the severity of illness necessitating both prone positioning and therapy with ECMO or CRRT. There were no deaths related to turning, to the prone position or to cannula malfunction. This series is too small to offer any predictions regarding survival with the multimodality therapy used. Conclusions : Using our technique, prone positioning with large-bore venous access is safe and does not result in cannula complications. Flow rates are maintained in the prone position. Potential cannula complications of ECMO and CRRT are not a contraindication to prone positioning in severely ill patients. Competing interests : None declared. Key messages : * Prone positioning is an important adjunct in the treatment of respiratory failure * Some patients with severe respiratory failure, who are receiving ECMO or CRRT may also benefit from prone positioning * It is safe to position patients prone with high flow venous catheters if a co-ordinated method of turning is used with care to avoid dislodgment of the access lines Abbreviations : ECMO = extracorporeal membrane oxygenation; CRRT = continuous renal replacement therapy. Backmatter: PMID- 12202582 TI - Malignant Catarrhal Fever-Like Disease in Barbary Red Deer (Cervus elaphus barbarus) Naturally Infected with a Virus Resembling Alcelaphine Herpesvirus 2 AB - Eight Barbary red deer (Cervus elaphus barbarus) developed clinical signs suggestive of malignant catarrhal fever (MCF) over a 28-day period. These animals were housed outdoors with four other species of ruminants. Affected red deer had lethargy, ocular signs, and nasal discharge and were euthanatized within 48 h. Lesions included ulcers of the muzzle, lips, and oral cavity associated with infiltrates of neutrophils and lymphocytes. Serologically, six of seven red deer tested during the outbreak were positive by competitive enzyme-linked immunosorbent assay for antibodies to a shared MCF virus antigen. PCR using oligonucleotide primers designed for a conserved protein of alcelaphine herpesviruses 1 (AlHV-1) and 2 (AlHV-2) and for conserved regions of a herpesvirus DNA polymerase gene was positive for tissues from all eight clinically affected animals and negative for eight out of eight red deer without clinical signs of MCF. DNA sequencing of PCR amplicons from the diseased red deer indicated that they were infected with a novel herpesvirus closely related to AlHV-2; immunohistochemistry using polyclonal anti-AlHV-2 serum and in situ hybridization demonstrated the presence of virus within salivary glands adjacent to oral lesions of affected animals. A survey of other ruminants near the outbreak subsequently showed that normal Jackson's hartebeest (Alcelaphus buselaphus jacksoni) that were cohoused with the diseased red deer were infected with the same virus and were shedding the virus in nasal excretions. These findings suggest that a herpesvirus closely related to AlHV-2 caused the MCF-like disease epizootic in Barbary red deer and that the virus may have originated from Jackson's hartebeest. Keywords: Introduction : Malignant catarrhal fever (MCF) is a systemic disease of ruminants caused by herpesvirus. Infection with an MCF herpesvirus is usually fatal, although some ruminants, such as wildebeest, hartebeest, and domestic sheep, are well adapted to particular strains of virus and function as clinically normal reservoirs . MCF has been recognized in at least 13 species of deer, including mule deer (Odocoileus hemionus), Pere David's deer (Elaphurus davidianus), white-tailed deer (Odocoileus virginianus), and red deer (Cervus elaphus) . Among these, molecular characterization of the causative virus has been done only for the white-tailed deer outbreaks from 1998 to 2001 . In recent years the use of PCR primers for conserved regions of the herpesvirus DNA polymerase gene has facilitated identification of previously unknown viruses . Among ruminants, these include bovine lymphotropic herpesvirus , caprine herpesvirus 2 (CpHV-2) , caprine lymphotropic herpesvirus (CpLHV) , and a herpesvirus of unknown origin in white-tailed deer (Odocoileus virginianus) . Two of these, CpHV-2 and the herpesvirus found in white-tailed deer, have been provisionally classified as members of the MCF group under Gammaherpesvirinae on the basis of a common antigenic epitope and DNA sequence similarity within the DNA polymerase gene . CpHV-2 has been associated with chronic dermatitis and weight loss in sika deer (Cervus nippon) , while the virus found in white-tailed deer produced a disease with typical MCF clinical signs but had a prolonged course and chronic histologic lesions . These viruses are now included with alcelaphine herpesvirus 1 (AlHV-1) and ovine herpesvirus 2 (OvHV-2) as members of the MCF virus group with known pathogenic potential. AlHV-1, first described and linked to MCF in 1960 , can presently be detected by PCR or by serologic methods (, -) in clinical MCF cases or herd screening. However, unambiguous identification of AlHV-1 may require sequencing of PCR-amplified DNA. OvHV-2 is similarly identifiable by a virus-specific PCR assay and by the same competitive-inhibition enzyme-linked immunosorbent assay (CI-ELISA) that detects antibodies to AlHV-1 . In contrast, two members of the MCF group have not been associated with disease: hippotragine herpesvirus 1 (HiHV-1) and alcelaphine herpesvirus 2 (AlHV-2) . The latter has been isolated several times from clinically normal topi or Cape hartebeest, members of the bovid subfamily Alcelaphinae , but experimental infections of cattle produce no discernible effects; moreover, inoculation of cattle with AlHV-2 does not elicit antibodies protective against subsequent AlHV-1 challenge . AlHV-2 can be detected by AlHV-2-specific PCR as well as by a PCR that will amplify both AlHV-1 and AlHV-2 and by CI-ELISA . We now report evidence that a virus closely related to AlHV-2 caused disease resembling MCF in eight Barbary red deer (Cervus elaphus barbarus) and that the virus may have derived from cohoused, clinically normal Jackson's hartebeest (Alcelaphus buselaphus jacksoni). MATERIALS AND METHODS : Animals and samples. | Eight of 33 Barbary red deer (Cervus elaphus barbarus) housed in a large enclosure at the Wild Animal Park in Escondido, California, with Ankole cattle (Bos primigenius f. taurus), Jackson's hartebeest (Alcelaphus buselaphus jacksoni), sand gazelle (Gazella subgutturosa marica), and scimitar-horned oryx (Oryx dammah), developed clinical signs of MCF and were euthanatized, necropsied, and analyzed in this study. Serum, peripheral blood leukocytes (PBL), nasal swabs, or tissues were collected and analyzed from the diseased Barbary red deer and from selected unaffected animals at the San Diego Wild Animal Park and the San Diego Zoo. The unaffected animals were divided into three groups on the basis of geographic proximity to the diseased red deer: group 1 animals were housed in the exhibit where the Barbary red deer disease outbreak occurred and included three Jackson's hartebeest, a Soemmerring's gazelle (Gazella soemmerringi soemmerringi), a southeastern crowned duiker (Sylvicapra grimmia caffra), a scimitar-horned oryx, and an Ankole. Two of the Jackson's hartebeest were housed in the enclosure during the outbreak, and one was housed in the enclosure 3 years prior to the outbreak and died from trauma. Group 2 animals were from other exhibits or areas of the Wild Animal Park whose deaths coincided with the outbreak in Barbary red deer and included four Sudan Barbary sheep (Ammotragus lervia blainei), two wild mule deer (Odocoileus hemionus), and a Pere David's deer (Elaphurus davidianus). Group 3 animals were housed by themselves at the San Diego Zoo (approximately 50 km south of the Wild Animal Park) and included eight clinically normal Barbary red deer and a stillborn Barbary red deer whose death coincided with the Barbary red deer outbreak. Additional serum samples banked prior to the outbreak were available for six of the eight clinically affected Barbary red deer and two of the three cohoused, clinically normal Jackson's hartebeest. Frozen tissue culture supernatant of AlHV-2 isolate 840412 obtained from a topi (Damaliscus lunatus jimela) in 1984 was analyzed and used for comparison to Barbary red deer virus sequences. Frozen tissue culture supernatant of AlHV-1 isolate WC11 obtained from blue wildebeest (Connochaetes taurinus taurinus) was used as a negative control for Southern blot hybridizations and for the Barbary red deer AlHV-1/AlHV-2 PCR described below. Serology. | Serum samples were submitted to the Washington State University and U.S. Department of Agriculture Animal Research Unit MCF Testing Laboratory in Pullman, Washington, for MCF virus antibody analysis that utilizes a CI-ELISA to detect the presence of antibody (i.e., immunoglobulin M [IgM], IgG, and IgA) to a viral antigen shared by all recognized MCF virus isolates . Necropsy. | Complete necropsies were performed on all eight diseased Barbary red deer and eight unaffected animals from groups 1, 2, and 3. Complete sets of tissues were immersion fixed in 10% neutral buffered formalin for histology. Fresh tissue samples of conjunctiva, intestine, kidney, liver, lymph nodes, lung, nasal mucosa, oral mucosa, and spleen were collected from all eight affected Barbary red deer as well as two Sudan Barbary sheep, two mule deer, one Pere David's deer, and one stillborn Barbary red deer. Fresh tissue samples of lymph node were taken in 1998 from a Jackson's hartebeest that died from trauma; placenta was taken from a Sudan Barbary sheep that gave birth during the outbreak at the Wild Animal Park. Fresh tissues were frozen at -80C or were immediately processed for DNA extraction as described below. Histopathology. | Formalin-fixed tissues were routinely processed, sectioned at 5 mum, and stained with hematoxylin and eosin for microscopic evaluation. DNA and RNA extractions. | PBL, solid tissues, and nasal swabs were extracted with a Qiagen Tissue Kit according to the manufacturer's tissue sample protocol, except that the recommended amounts of sample were first placed with the lysis buffer in 1.5-ml screw-cap FastPrep vials containing ceramic beads and were lysed by agitation in a FastPrep shaker (Q-BIOgene, Carlsbad, Calif.) at a speed of 4 to 5.5 for 40 to 60 s, after which the lysate was transferred to a clean Eppendorf tube for continuation of the Qiagen protocol. DNA was also extracted with the Qiagen Tissue Kit as above from frozen tissue culture supernatant of AlHV-1 isolate WC11 from blue wildebeest (Connochaetes taurinus taurinus) and AlHV-2 isolate 840412 obtained from a topi (Damaliscus lunatus jimela). Total RNA was extracted from selected tissues of all eight affected Barbary red deer by using TRIzol (Gibco BRL Life Technologies, Grand Island, N.Y.). Immunohistochemistry. | Five-micromillimeter tissue sections mounted on Platinum slides (Mercedes Medical, Sarasota, Fla.) were deparaffinized, hydrated, washed in phosphate-buffered saline (PBS), blocked with Peroxo-Block (Zymed Laboratories, South San Francisco, Calif.), washed in PBS, and heated in 0.1 M citrate buffer, pH 6.0, at 96C for 20 min and then allowed to cool to room temperature (RT). Sections were washed in PBS, blocked with avidin-biotin blocking kit (Zymed Laboratories), washed in PBS, blocked with CAS block (Zymed Laboratories), and then treated with a 1:10 dilution of the primary antibody, a custom-made polyclonal rabbit antiserum to psoralen-inactivated AlHV-2 isolate 840412 (LEE BioMolecular, San Diego, Calif.) at RT for 1 h. Control sections were treated with preimmune rabbit serum (LEE BioMolecular) at 1:10 dilution. Sections were washed in PBS and treated with a 1:400 dilution of biotinylated goat anti-rabbit antibody (Zymed Laboratories) for 30 min at RT, washed in PBS, treated with streptavidin-peroxidase (Zymed Laboratories) at 1:400 dilution for 30 min at RT, washed in PBS, and then treated with diaminobenzidine tetrahydrochloride (Vector Laboratories, Burlingame, Calif.) at RT for 5 min and washed in distilled water, counterstained with Gill's hematoxylin, dehydrated, and mounted. In situ hybridization. | Five-micromillimeter sections of selected formalin-fixed, paraffin-embedded tissues on Platinum slides (Mercedes Medical) were deparaffinized, hydrated, treated with trypsin at 37C for 10 min with Digest-All (Zymed Laboratories), washed in Tris-buffered saline (TBS), heated at 98C for 12 min in TBS, placed immediately in 4C TBS, and prehybridized with DIG Easy Hyb Granule solution (Roche Diagnostics Corp., Indianapolis, Ind.) at 42C for 1 h. A 139-bp cloned and sequenced DNA region from the AlHV-1/AlHV-2 PCR from Barbary red deer was labeled with digoxigenin by using a PCR incorporation method according to the manufacturer's instructions (Roche Diagnostics Corp.) and was used in DIG Easy Hyb Granule solution for hybridization at 42C for 16 h at 50 pmol of labeled probe per ml of hybridization solution. Slides were washed in 2x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate) with 0.1% sodium dodecyl sulfate twice for 5 min at RT followed by two washes in 0.5x SSC with 0.1% sodium dodecyl sulfate at 68C for 15 min. Slides were blocked in blocking solution (Roche Diagnostics Corp.) for 30 min, washed in TBS, treated with anti-digoxigenin antibody (Roche Diagnostics Corp.) diluted 1:5,000 (150 mU/ml) in TBS for 1 h at RT, washed in TBS, and treated with nitroblue tetrazolium chloride-5-bromo-4-chloro-3-indolyl phosphate, toluidine salt (Roche Diagnostics Corp.) for 10 min, washed in TBS and distilled water, counterstained with Gill's hematoxylin (Surgipath Instrumentation Inc., Richmond, Ill.), and mounted with Crystal Mount (Biomeda Corp., Foster City, Calif.). Duplicate control slides received identical treatment, except no labeled probe was added to the hybridization solution. Virus isolation. | Isolation was attempted by overnight cocultivation of whole blood from one Barbary red deer with clinical signs, using African green monkey kidney (Vero) cells as previously described . PCR. | The following 11 PCR assays for potential viral etiologies were performed on DNA or cDNA derived from tissue and nasal swab samples. (i) AlHV-1. PCR specific for a region of AlHV-1 but not of AlHV-2 was done with 20 to 1,000 ng of DNA from tissues or leukocytes. DNA was added to a 25- or 50-mul reaction mixture containing 15 mM Tris (pH 8.0), 50 mM KCl, 2.5 mM MgCl2, and 200 muM each of dATP, dCTP, dGTP, and dTTP. AmpliTaq Gold DNA polymerase (Applied Biosystems, Foster City, Calif.) was used at a final concentration of 0.05 U/mul. Thermal cycling conditions were 9 min at 95C followed by 35 cycles of 94C (30 s), 55C (1 min), and 72C (1 min), followed by a final extension at 72C for 10 min. (ii) Deer MCF. PCR specific for the deer MCF agent was used in primary and secondary reactions of either a 25- or 50-mul total volume as previously described . (iii) OvHV-2. OvHV-2 virus-specific PCR was used as previously described , except that only a single reaction was carried out with the oligomers designed to produce a 422-bp product. The reverse primer (5'-GTCTGGGGTATATGAATCCAGATGGCTCTC-3') was modified slightly at its 5' end to match the reported sequence of the plasmid Bp4a1 , kindly supplied by H. W. Reid and used as a positive control. (iv) AlHV-1/AlHV-2. Primers that amplify a homologous region partially within the major capsid protein of AlHV-1 or AlHV-2 were used as previously described , with the minor modifications in the reaction and cycling conditions listed above for the AlHV-1 PCR. (v) Barbary red deer AlHV-1/AlHV-2. Primers (forward, 5'-TTTATTGAAGAAGTGGCTC-3'; reverse, 5'-CCATTTTGTTTTGTCCTG-3') that were designed to the Barbary red deer virus sequences generated by the AlHV-1/AlHV-2 PCR described above were used for nested reamplification of all samples that were negative with the AlHV-1/AlHV-2 PCR. One microliter of PCR product from the AlHV-1/AlHV-2 PCR was used as template. Thermal cycling conditions for the inner primer pair were 6 min at 95C followed by 30 cycles of 94C (30 s), 47C (30 s), and 72C (20 s), followed by a final extension at 72C for 7 min. (vi) Herpesviridae DNA polymerase. PCR using degenerate primers that target consensus regions of herpesvirus DNA polymerase genes for amplification of all herpesviruses was conducted as previously described . (vii and viii) Bovine herpesvirus 1 (BHV-1) and bovine herpesvirus 5 (BHV-5). PCR on DNA from selected tissues for detection of BHV-1 and BHV-5 was done as described previously . (ix to xi) Bovine viral diarrhea virus (BVDV), vesicular stomatitis virus (VSV), and foot-and-mouth disease virus (FMDV). Reverse transcriptase-PCR for BVDV, VSV, and FMDV was done on total RNA from selected tissues as described previously . DNA manipulation and sequencing. | PCR products of the expected size from representative tissues and cases were purified and either direct sequenced or cloned by using the TOPO TA Cloning Kit (Invitrogen). Sequencing reactions were performed by using the CEQ DTCS (dye terminator cycle sequencing) Quick-Start Kit (Beckman Coulter, Fullerton, Calif.). Sequences were acquired by using a CEQ 2000XL capillary sequencer (Beckman Coulter). Sequence analysis and alignments were conducted by using the MacVector v. 7.0 and AssemblyLIGN v. 1.0.9 software packages (Accelrys, San Diego, Calif.). Sequence data were compared to the GenBank database with the basic local alignment search tool. Southern hybridization and detection. | Southern blotting of all PCR products was performed by capillary transfer or by use of a PosiBlot pressure apparatus (Stratagene, La Jolla, Calif.) with positively charged nylon membranes (Amersham Pharmacia Biotech, Piscataway, N.J., or Millipore, Bedford, Mass.). DNA was fixed to the membrane by using a Stratalinker (Stratagene). DNA probes were generated by labeling with digoxigenin by using either the DIG Oligonucleotide 3'-End Labeling Kit or the PCR DIG Probe Synthesis Kit (Roche Molecular Biochemicals, Indianapolis, Ind.) and specific oligomers or PCR-amplified sequence from control plasmids or plasmids containing cloned and sequenced initial amplification products. Hybridization and detection were performed with reagents of the DIG High Prime DNA Labeling and Detection Kit (Roche Molecular Biochemicals) and exposure of the treated blots to Kodak X-Omat LS X-ray film (Eastman Kodak Co., Rochester, N.Y.). Nucleotide sequence accession numbers. The nucleotide sequences obtained in this study from Barbary red deer and Jackson's hartebeest were deposited in GenBank with accession numbers (partial major capsid protein sequence) and (partial polymerase sequence). The sequence from AlHV-1/AlHV-2 PCR on topi AlHV-2 isolate 840412 was also deposited in GenBank with accession number . RESULTS : Over a 4-week period 8 of 33 Barbary red deer in the Wild Animal Park enclosure developed ocular and nasal discharge, drooping ears, coughing, and lethargy. In all cases the affected Barbary red deer were euthanatized within 48 h of the onset of clinical signs. No clinical abnormalities resembling those of the Barbary red deer were seen in other animals at the Wild Animal Park or San Diego Zoo. Serum samples were tested by CI-ELISA that detects the presence of immunoglobulins (i.e., IgM, IgG, and IgA) specific for an antigen shared by all MCF viruses isolated to date . Serum samples taken just prior to euthanasia from six of seven disease-affected Barbary red deer were positive by the CI-ELISA. Three of these positive animals had been seronegative prior to the outbreak, and one was seropositive. Of other group 1 animals, three out of three Jackson's hartebeest and one scimitar-horned oryx were seropositive, while a Soemmerring's gazelle, a southeastern crowned duiker, and a clinically normal Barbary red deer were seronegative. Of group 2 animals, three out of three Sudan Barbary sheep were negative. Seven of the eight clinically normal Barbary red deer located at the San Diego Zoo (group 3) were seronegative. These results are summarized in Table . Necropsy lesions in affected Barbary red deer were of variable severity and distribution. They consisted of ulcers and erosions of the muzzle and lips. Smaller ulcers and erosions were also present in the oral cavity and other areas of the skin. Also present were white-yellow ocular discharge and enlarged lymph nodes. Microscopically, lesions in the skin, oral cavity, and conjunctiva consisted of erosions and ulcers with infiltrates of neutrophils and lymphocytes. Lymphocytes predominated, and infiltrates were most dense in the superficial dermis and submucosa with extension into the overlying epithelium. Some vessels had perivascular inflammatory cells. Lymphoid infiltrates were present in some bronchi and bronchioles and in the trachea. Lymph node sinuses were filled with neutrophils. Vasculitis was not seen in any cases. Immunohistochemistry with polyclonal anti-AlHV-2 serum and in situ hybridization with a Barbary red deer virus-specific probe demonstrated virus within the cytoplasm and nuclei of occasional oral submucosal salivary gland cells located beneath areas of ulceration and inflammation . No staining was seen within lesions or in unaffected tissues. Fifty-eight DNA samples from various organs of the eight Barbary red deer with clinical disease and 70 samples from unaffected animals, including the eight healthy Barbary red deer located at the San Diego Zoo, were tested for AlHV-1, deer MCF, OvHV-2, AlHV-1/AlHV-2, Barbary red deer AlHV-1/AlHV-2, and Herpesviridae DNA polymerase by PCR. PCR products were analyzed by Southern blot hybridization or sequencing. Results are summarized in Table . All samples from affected Barbary red deer and all unaffected animals, including Jackson's hartebeest, were PCR negative for AlHV-1, deer MCF virus, and OvHV-2. Forty-nine of 58 samples (84%) from the eight Barbary red deer with clinical disease were positive for AlHV-1/AlHV-2 PCR. AlHV-1/AlHV-2 amplicons from various tissues for three different Barbary red deer were cloned and sequenced and consisted of identical 139-bp products (internal to the primers). AlHV-1/AlHV-2 PCR was also positive on DNA extracted from the topi AlHV-2 isolate 840412 from 1984 , and amplimers were cloned and sequenced for comparison with findings for Barbary red deer. Sequence analyses demonstrated the Barbary red deer virus products to have highest nucleotide identity (95%) with AlHV-2 isolate 840412 from topi (Damaliscus lunatus jimela) and only 78% identity with AlHV-1. AlHV-1/AlHV-2 PCR was positive on PBL or nasal swabs from two Jackson's hartebeest housed in the enclosure where the outbreak occurred. Clones from PBL were sequenced and were identical to AlHV-1/AlHV-2 amplimers from Barbary red deer . AlHV-1/AlHV-2 PCR was negative on all other unaffected animals. The Barbary red deer AlHV-1/AlHV-2 PCR, which utilized primers designed to the Barbary red deer virus and nested within the AlHV-1/AlHV-2 PCR product, was used to increase the sensitivity of virus detection. All samples that were negative with the AlHV-1/AlHV-2 PCR were tested. Seven out of nine samples from clinically affected Barbary red deer were positive. All of the samples from unaffected animals, except Jackson's hartebeest, were negative. Samples from all three Jackson's hartebeest housed in the enclosure where the outbreak occurred were positive. Clones from four positive samples of PBL and nasal swabs from two Jackson's hartebeest were sequenced and found to be identical to sequences from Barbary red deer samples. Barbary red deer AlHV-1/AlHV-2 PCR was negative on all tissues from all other unaffected animals and on the AlHV-1 isolate WC11 from blue wildebeest . Thirty-five of 58 samples (60%) from the eight Barbary red deer with clinical disease were positive for Herpesviridae DNA polymerase PCR. Clones from six different red deer were sequenced and found to consist of identical 174-bp products (internal to the primers). Barbary red deer sequence had highest nucleotide identity to AlHV-2 isolated from topi (94%) (GenBank ), followed by AlHV-1 (80%) (GenBank ), OvHV-2 (69%) (GenBank ), and deer MCF (67%) (GenBank ). Herpesviridae DNA polymerase PCR was positive on two of three Jackson's hartebeest, including one animal that died 3 years before the outbreak. Cloned sequences had 100% identity with the Barbary red deer virus sequences . Predicted amino acid sequences for the Barbary red deer and Jackson's hartebeest DNA polymerase segment had highest identity to topi AlHV-2 (96%), followed by AlHV-1 (84%), deer MCF virus (73%), and OvHV-2 (71%) . All other unaffected animals were negative by Herpesviridae DNA polymerase PCR. The eight Barbary red deer with clinical disease were PCR negative by visual inspection of agarose gels for all other viral etiologies considered, including BHV-1, BHV-5, BVDV, VSV, and FMDV (data not shown). Virus isolation attempts from affected Barbary red deer were assessed by microscopic inspection of cell cultures and PCR and were determined to be unsuccessful after 20 days in culture. FIG. 1. | (A and B) Photomicrographs of immunohistochemistry for AlHV-2 demonstrating positive staining in buccal salivary gland cells (A) and no staining in negative control (B). (A and B) Photomicrographs of immunohistochemistry for AlHV-2 demonstrating positive staining in buccal salivary gland cells (A) and no staining in negative control (B). Diaminobenzidine tetrahydrochloride with Gill's hematoxylin counterstain, bar = 50 mum. (C and D) In situ hybridization with an AlHV-2-like viral PCR probe from Barbary red deer with a similar pattern of virus detection in submucosal salivary glands (C) and lack of staining in negative control (D). Nitroblue tetrazolium chloride-5-bromo-4-chloro-3-indolyl phosphate was used with Gill's hematoxylin counterstain. Bar = 50 mum. FIG. 2. | Alignment of homologous DNA sequences from the AlHV-1/AlHV-2 PCR for Barbary red deer, Jackson's hartebeest, topi AlHV-2, and previously characterized MCF viruses. Alignment of homologous DNA sequences from the AlHV-1/AlHV-2 PCR for Barbary red deer, Jackson's hartebeest, topi AlHV-2, and previously characterized MCF viruses. Identical nucleotides in the majority of aligned sequences for a given position are enclosed in a box. Sequence identity between the Barbary red deer virus amplimers and other MCF viruses is shown in bold adjacent to the ends of the sequences. The GenBank accession number (and reference) for AlHV-1 is . The source of the AlHV-2 isolate was topi (Damaliscus lunatus jimela), isolate 840412 . FIG. 3. | Ethidium bromide-stained agarose gel (A) and film exposure of Southern blot hybridization (B) of Barbary red deer AlHV-1/AlHV-2 PCR for selected unaffected animals, including Jackson's hartebeest. Ethidium bromide-stained agarose gel (A) and film exposure of Southern blot hybridization (B) of Barbary red deer AlHV-1/AlHV-2 PCR for selected unaffected animals, including Jackson's hartebeest. Lanes 1 and 20, AlHV-1 isolate WC11 from blue wildebeest; lanes 2, 9, and 12, 100-bp ladder; lane 3, Jackson's hartebeest 697068 PBL sampled in October 2001; lane 4, Jackson's hartebeest 697068 nasal swabs sampled in October 2001; lane 5, Jackson's hartebeest 697068 PBL sampled in January 2002; lane 6, Jackson's hartebeest 697068 nasal swabs sampled in January 2002; lane 7, Jackson's hartebeest 601119 PBL; lane 8, Jackson's hartebeest 601119 nasal swabs; lane 10, affected Barbary red deer 696250 nasal swabs; lane 11, Jackson's hartebeest 697064 lymph node sampled in October 1998; lane 13, Ankole PBL; lane 14, scimitar-horned oryx PBL; lanes 15 to 18, mule deer intestine, liver, lymph node, and spleen, respectively; lane 19, no-DNA negative control. FIG. 4. | Alignment of homologous DNA sequences for the Herpesviridae DNA polymerase PCR from Barbary red deer, Jackson's hartebeest, and known MCF viruses. Alignment of homologous DNA sequences for the Herpesviridae DNA polymerase PCR from Barbary red deer, Jackson's hartebeest, and known MCF viruses. Identical nucleotides in the majority of aligned sequences for a given position are enclosed in a box. Sequence identity between the Barbary red deer virus amplimers and other MCF viruses is shown in bold adjacent to the ends of the sequences. GenBank database numbers (and references) for AlHV-1, AlHV-2, OvHV-2, and the deer MCF virus are AF005370 , AF275942 , AF031812 , and AF181468 , respectively. FIG. 5. | Alignment of predicted amino acid sequences from the Herpesviridae DNA polymerase PCR from Barbary red deer, Jackson's hartebeest, and known MCF viruses. Alignment of predicted amino acid sequences from the Herpesviridae DNA polymerase PCR from Barbary red deer, Jackson's hartebeest, and known MCF viruses. Similar residues between aligned sequences for a given position are enclosed in a box. Identity between the Barbary red deer virus predicted amino acid sequence and other MCF viruses is shown in bold adjacent to the ends of the sequences. GenBank numbers (and references) for AlHV-1, AlHV-2, OvHV-2, and the deer MCF virus are AF005370 , AF275942 , AF031812 , and AF181468 , respectively. TABLE 1 | Serology results for MCF antibodies by CI-ELISA TABLE 2 | PCR assays for MCF viruses performed on samples from affected Barbary red deer and unaffected animals DISCUSSION : Deer are known to be susceptible to MCF caused by either AlHV-1 or OvHV-2 . The disease in deer may occur in a peracute form, where animals show no clinical signs and are simply found dead , or it may manifest as the more typical head-and-eye form , with ocular and nasal discharge, conjunctivitis, depression, and hyperemia or epithelial erosion associated with the eyes, oral cavity, or nose . Histologically, lymphocytic vasculitis is a consistent finding in such cases , even when clinical signs are absent and death occurs in 24 to 48 h . In cattle the extent of vasculitis increases with severity of disease caused by AlHV-1 , and the same may be true of deer. The clinical signs seen in the present cases from Barbary red deer were consistent with the classical head-and-eye form of MCF, but the microscopic lesions, while having many features characteristic of MCF, lacked typical vasculitis. Because all of these animals were euthanatized within 48 h of first observance of clinical signs, it is not possible to conclude whether the absence of vasculitis was due to arrested progression of the disease or to species differences in response to the particular etiologic agent in these cases. Data from PCR and serology identified the novel AlHV-2-like virus as the etiology of disease in the Barbary red deer. Other viruses which can cause disease similar to MCF, such as BHV-1, BHV-5, BVDV, VSV, and FMDV, were excluded by PCR. The possibility of a different MCF virus causing the outbreak was also eliminated by extensive sequencing of positive amplicons from the two broader-specificity PCR assays (AlHV-1/AlHV-2 and Herpesviridae DNA polymerase) and negative results of specific PCR assays for AlHV-1, deer MCF, and OvHV-2. The demonstration of viremia in seven out of seven Barbary red deer with MCF and lack of virus infection in every Barbary red deer without MCF, along with the marked contrast in MCF seroreactivity between diseased and healthy Barbary red deer, strongly implicate the AlHV-2-like virus as the primary cause of the outbreak. AlHV-2 has been isolated from two species of topi (Damaliscus korrigum and Damaliscus lunatus jimela) and a Cape hartebeest (Alcelaphus buselaphus caama) . None of these animals had clinical disease. In our study, both immunohistochemistry and molecular data indicated that the virus present in Barbary red deer was closely related to AlHV-2. Positive staining on immunohistochemistry provided evidence that the etiologic agent in the Barbary red deer possesses structural proteins similar to those of AlHV-2, as the primary antibody used in staining was from rabbits immunized with topi AlHV-2 isolate 840412 purified from cell culture supernatant fluid. This polyclonal serum was previously shown by indirect immunofluorescence to bind to AlHV-2-infected cultured cells but not to cells infected with AlHV-1 (R. S. Lahijani, R. B. Klieforth, B. S. Seal, S. M. Sutton, and W. P. Heuschele, 43rd Annual Conference of the Wildlife Disease Association, abstr. 91, 1994). DNA sequences from the PCR. for AlHV-1/AlHV-2 and Herpesviridae DNA polymerase were clearly more similar to AlHV-2 than to any other MCF virus. In regard to the nucleotide differences of 5 to 6% between the red deer virus and AlHV-2 isolate 840412 , we cannot infer from this limited information whether the agent identified in these MCF cases is a strain of AlHV-2 or a closely related, novel virus species. Nucleotide sequence differences between two types of Epstein-Barr virus (EBV) occur at a rate of about 1 to 2% over most of the genome, but four EBV genes show nonsynonymous nucleotide substitutions resulting in differences from 8 to 29% between strains . The paucity of sequence information from isolates of any MCF virus precludes obtaining a range of normal strain variability within this group. If the Barbary red deer virus is indeed a strain of AlHV-2, this is particularly intriguing because AlHV-2 has not previously been implicated in clinical disease resembling MCF in any ruminants. MCF has been previously recognized in red deer (Cervus elaphus) in Scotland and New Zealand. The source of the virus in Scotland was not determined, but sheep were suspected. On the basis of the severity of disease, red deer were believed to be especially sensitive to MCF . In that report, lymphoproliferative changes were extensive but viral particles were not seen ultrastructurally in any of the lesions. Authors of another study of naturally occurring and experimentally induced MCF in cattle described lack of evidence in the literature for immunohistochemical association between the virus and lesion sites and also found no ultrastructural evidence of virus in vascular or systemic epithelial lesions of the cases they examined . Electron microscopy was not attempted for our study. However, immunohistochemistry demonstrated the presence of occasional positive cells in buccal salivary glands adjacent to oral lesions, though not in the lesions themselves. In situ hybridization was also successful, producing a similar pattern of staining in the same tissue. It cannot be determined with certainty whether these results indicate productive infection with release of virus, but salivary glands as sites of MCF virus replication would provide a means for efficient excretion into the environment. Other researchers also used in situ hybridization to demonstrate viral infection of pulmonary alveolar cells of wildebeest (Connochaetes spp.) calves , suggesting another possible site for replication and a mechanism for excretion by aerosolization. The CI-ELISA test proved valuable in the Barbary red deer outbreak not only for corroborating an MCF virus as the etiology but also for helping identify possible reservoirs of the virus. Among our cases there was evidence for seroconversion in three Barbary red deer. One red deer that developed disease was seropositive for two or more years prior to the outbreak. It is possible that this animal suffered from recrudescence of a latent infection with the AlHV-2-like virus and then spread the virus to immunologically naive red deer. Apparent latency and recrudescence of MCF have been documented for Formosan sika deer . Alternatively, seroreactivity in this red deer may have been to a different MCF virus, such as AlHV-1 or OvHV-2, and the AlHV-2-like virus causing the outbreak may have arisen from another species within or adjacent to the enclosure. The Barbary red deer that developed disease in our study had direct or indirect contact with seven other species of ruminants, any of which may have been primary or intermediate reservoirs for the virus that caused the outbreak. Of these animals, only a scimitar-horned oryx and Jackson's hartebeest were seropositive for MCF virus. AlHV-1 seroreactivity has been reported for several species of oryx that were clinically normal , but AlHV-2 has never been described in oryx. The scimitar-horned oryx in our study was negative by PCR of blood for all MCF viruses. Among the other animals that had contact with Barbary red deer, only mule deer have been reported with MCF . The mule deer tested in our study lacked lesions of MCF and were negative by PCR for numerous tissues but, along with the oryx, still cannot be completely excluded as possible intermediate reservoirs. Coke's and Cape hartebeest have long been recognized as carriers of AlHV-1 or AlHV-2 and have been implicated in natural transmission of virus to other ruminants . Our study is the first report of an MCF virus in Jackson's hartebeest. The confirmation of identical AlHV-1/AlHV-2 and Herpesviridae DNA polymerase sequences in Barbary red deer and seropositive, cohoused Jackson's hartebeest points toward the hartebeest as a possible primary reservoir for the virus precipitating the epizootic. It is especially noteworthy that both of the hartebeest tested were viremic and that PCR was positive on nasal swabs, suggesting that virus was being shed in nasal excretions. Barbary red deer are one of only two deer subspecies native to Africa, but they have not shared ranges with hartebeest for several hundred years . It is possible that the lack of contact between modern Barbary red deer and hartebeest resulted in an increased vulnerability of the deer to MCF viruses indigenous to hartebeest, such as the AlHV-2-like virus in this outbreak. The failure to isolate a virus from Barbary red deer is not surprising. It may be that the Vero cells were not permissive for infection by this particular MCF variant or that too much time elapsed between blood collection and inoculation of cell culture flasks. It has been observed for MCF viruses in general that tissues must be collected within 1 to 2 h of death, and attempts at virus isolation must be begun immediately as the viruses are rapidly inactivated in tissues after death . This appears to be a consequence of MCF viruses typically being highly cell associated during in vivo infections . Li et al. have observed that the traditional view of MCF as an "acute highly lethal disease with a short course . . . and a fairly characteristic set of signs and lesions" is rapidly giving way to one of a wider range of disease presentations caused by a greater number of viruses whose common features are a clustering of DNA sequence similarities disjunct from other gammaherpesviruses and a specific antigenic epitope defined by the monoclonal antibody used in the CI-ELISA. In the present cases there is evidence of antigenic similarity to other MCF viruses by CI-ELISA and immunohistochemistry and of DNA sequence similarity to other viruses in the MCF group. The close resemblance to AlHV-2 sequence, in particular, is unique among the viruses associated with MCF-like disease in ruminants to date. Comparison of additional nucleotide and derived amino acid sequences by phylogenetic analysis will provide more information on the relatedness of the etiologic agent in this outbreak to AlHV-2 and other gammaherpesviruses. Backmatter: PMID- 12202561 TI - Detection and Characterization of Hepatitis C Virus RNA in Seminal Plasma and Spermatozoon Fractions of Semen from Patients Attempting Medically Assisted Conception AB - To investigate the risk of transmission of hepatitis C virus (HCV) via semen in assisted reproduction techniques, semen samples from 32 men chronically infected with HCV attending a center for assisted procreation were tested for HCV RNA by a reverse transcription-PCR protocol by using a modified version of the Cobas AMPLICOR HCV assay (version 2.0; Roche Diagnostics). The sensitivity of the test was 40 copies/ml. Four of 32 seminal plasma samples (12.5%) were found to be positive for the presence of HCV RNA. The median HCV load in blood was significantly higher in patients who were found to be positive for the presence of HCV RNA in semen than in those who tested negative (P = 0.02). In one man, seven consecutive seminal plasma samples tested positive for HCV RNA, as did two consecutive motile spermatozoon fractions; the corresponding fractions obtained after migration of the spermatozoa remained negative. Despite the absence of the proven infectivity of virus in semen samples that test positive for HCV RNA, these findings highlight the fact that seminal fluid may exhibit prolonged HCV RNA excretion. The usefulness of HCV RNA detection in both seminal plasma and spermatozoon fractions before the start of a program of medically assisted reproduction in couples in whom the male partner is chronically infected with HCV would need to be evaluated prospectively with a larger population of subjects exhibiting HCV RNA in their semen. Keywords: Introduction : The risk of transmission of hepatitis C virus (HCV) via seminal fluid is still much debated, especially in assisted reproductive techniques (ART). Even though the rate of HCV transmission by the sexual route has been found to be low , the use of spermatozoa from men chronically infected with HCV in ART can lead to a theoretical risk of contamination for the female partner, for the technicians dealing with ART, and perhaps for the artificially conceived embryos of the couple or of other couples treated at the same time or stored in the same container. The management of HCV-infected men enrolled in programs of medically assisted reproduction is highly dependent upon the definition of standardized protocols of detection of HCV RNA in semen. Actually, although previous studies have reported the presence of HCV RNA in seminal plasma of men chronically infected with HCV and coinfected or not coinfected with human immunodeficiency virus (HIV), other investigators have found the opposite pattern . Those contradictory results could be explained, at least in part, by the heterogeneity of the populations studied and by the diversity and the poor standardization of the techniques used for the extraction of RNA from semen and for the reverse transcription (RT)-PCR protocols. Moreover, the cellular fractions of semen, particularly the motile spermatozoa, have been poorly investigated for the presence of HCV. We report herein on the detection of HCV RNA in seminal plasma and, for the first time, in the motile spermatozoon fractions of semen from men chronically infected with HCV and attempting medically assisted reproduction. MATERIALS AND METHODS : Patients. | Thirty-two male patients (mean age, 39 years; age range, 32 to 52 years) entering a program of ART were included in the study after they gave their fully informed written consent. All men tested positive for HCV RNA in blood by RT-PCR and were negative for HIV antibodies and hepatitis B surface antigen. The risk factors for HCV contamination were transfusion of blood products in 4 men (12.5%), hemophilia in 2 men (6.3%), intravenous drug addiction in 8 men (25%), and unidentified in 18 men (56.2%). No patient had received any antiviral drugs for chronic hepatitis C. Sequential samples of semen were obtained from nine men: seven men each gave two samples (mean delay between collection of the two samples, 12.4 +- 8 months), one man gave three sequential samples within 8 months, and one man gave seven successive samples within 11 months. Samples. | Plasma samples were separated from blood by centrifugation and frozen at -80C until use. After 3 days of sexual abstinence, semen samples were obtained by masturbation into a sterile container and were processed within 2 h of ejaculation. One milliliter of the semen sample was centrifuged at 800 x g for 10 min, and the seminal plasma was separated from the cell pellet and stored at -80C until further use (fraction 1). Another milliliter was submitted to centrifugation through a three-layer discontinuous Percoll (Pharmacia, Uppsala, Sweden) gradient (3 ml of 50, 70, and 90% Percoll, respectively); the motile spermatozoa were recovered from the 90% fraction, observed by microscopic examination for number and to check for the absence of white blood cell contamination, and kept frozen at -80C as aliquots of 500,000 cells until use (fraction 2). For one patient, 0.5 ml of sterile synthetic medium (BM1; Ellios Biotek Laboratories, Paris, France) used for in vitro cell culture was layered on the top of the 90% fraction obtained after Percoll selection; after incubation for 45 min at 37C in a 5% CO2 incubator, the upper 0.5 ml was carefully removed and the spermatozoa were counted. This "swim-up" fraction was kept frozen at -80C as aliquots of 500,000 cells until use (fraction 3) . Detection of HCV RNA in blood plasma. | The qualitative detection of HCV RNA in blood plasma was performed by the Cobas AMPLICOR HCV assay (version 2.0; Roche Diagnostics, Meylan, France), according to the instructions of the manufacturer. The quantification of HCV RNA was achieved by the Cobas AMPLICOR HCV Monitor assay (version 2.0; Roche Diagnostics). Detection of HCV RNA in seminal plasma (fraction 1). | The RNA extraction step was performed by a protocol modified from that of the QIAmp viral RNA kit (Qiagen, Courtaboeuf, France), as follows: 250 mul of thawed fraction 1 was diluted (vol/vol) in sterile water and centrifuged at 20,000 x g for 1 h; the supernatant was removed, and the resulting pellet was resuspended in 600 mul of AVL lysis buffer (Qiagen) containing 6.4 mul of the internal control from the Roche Cobas AMPLICOR assay. After an incubation step of 10 min at room temperature, 600 mul of absolute ethanol was added and the whole mixture was centrifuged through a QIAmp column at 6,000 x g for 1 min. After two successive washing steps with AW1 buffer for 1 min at 6,000 x g and AW2 buffer for 3 min at 12,000 x g, the RNA extracts were removed from the column by addition of 200 mul of AVE elution buffer and were centrifuged at 6,000 x g for 1 min. RT and qualitative PCR were done by the Cobas AMPLICOR HCV assay (version 2.0) according to the instructions of the manufacturer. The quantification of HCV RNA in seminal plasma was carried out as follows: the RNA extraction step was similar to that for the qualitative protocol, except that the volume of elution buffer was 400 mul instead of 1,000 mul. RT and quantitative PCR were done by the Cobas AMPLICOR Monitor HCV assay (version 2.0) as recommended by the manufacturer. The final result was expressed as the number of copies per milliliter with reference to the volume of 250 mul used in the extraction step. The sensitivities of the qualitative and quantitative assays were determined by using serial 2-fold dilutions (from 160 to 10 copies/ml) or 10-fold dilutions (from 10,000 to 10 copies/ml), respectively, in HCV-negative seminal plasma of blood plasma from an HCV-infected patient, which had been quantified previously by the Cobas AMPLICOR 4CV Monitor assay; then, each dilution was extracted and tested in six independent experiments. Detection of HCV RNA in spermatozoon fractions. | Twenty-five fractions 2 (motile spermatozoa) from 14 patients and four fractions 3 (swim-up spermatozoa) from 1 patient were tested for the presence of HCV RNA. The RNA extraction was performed by combining a lysis step with proteinase K (50 mg/ml) by use of the RNAeasy Mini kit (Qiagen). With reference to the standard Qiagen protocol, an additional step of elution of the column was performed with the first eluate. The internal control of the Cobas AMPLICOR HCV kit was diluted 1:16 in 50 mul of the column eluate. RT and qualitative PCR were done according to the instructions of the manufacturer. In addition, a test with an extraction control was performed by RT-PCR amplification of an intron-containing region of the protamin 2 gene, as described previously . Use of a threshold of 500,000 cells in the PCR experiments was justified by the reproductive amplification of the control gene at this cell concentration but not at a lower one. HCV genotyping. | The genotypes of the HCV strains in blood and seminal samples were determined by sequencing the 5' noncoding region by the TruGene HCV assay (Visible Genetics, Espinay sur Orge, France). Statistical analysis. | Quantitative and qualitative values were compared by Student's t test and the chi-square test, respectively. The threshold for a significant value was 0.05. RESULTS : Detection of HCV RNA in blood and seminal plasma. | All 32 patients included in the study were positive for the presence of HCV RNA in blood plasma. Their mean viral load was 5.97 +- 0.51 log copies/ml (range, 4.97 to 7.34 log copies/ml). The genotypes of the HCV strains detected in the blood samples of 24 subjects were distributed as follows: 1 (two patients), 1a (seven patients), 1b (four patients), 2 (three patients), 3 (one patient), 3a (six patients), and 4 (one patient). PCR inhibitors were detected in only 1 of 32 (3.3%) seminal plasma specimens, but they were not detected after the seminal sample was diluted 1:2 in RNase-free water. The sensitivity of the qualitative assay, determined in six independent experiments as the lowest dilution found to be positive in all the experiments, was 40 copies/ml (1.6 log copies/ml). By using the same algorithm, the sensitivity of the quantitative assay was 100 copies/ml (2 log copies/ml). Seminal plasma samples from 4 of 32 patients (12.5%) were found to be positive for HCV RNA by the qualitative test. The four patients had been infected through intravenous drug abuse. The seminal viral loads of these four patients are listed in Table . The difference in viral loads between blood and semen ranged from 2.83 to more than 5.34 log copies/ml. The mean blood viral load was significantly higher in patients positive for HCV RNA in semen than in those negative for HCV RNA in semen (6.52 +- 0.55 versus 5.88 +- 0.46 log copies/ml [P = 0.02]) . For the four patients positive for HCV RNA in semen, the same HCV genotype was found in both blood and seminal plasma . Analysis of sequential seminal samples. | Successive seminal specimens were obtained from nine patients. Of the seven subjects who each provided two seminal samples, the detection of HCV RNA in seminal plasma was found to be unchanged over time in six patients (negative in five patients and positive in one patient); in the remaining patient, HCV RNA was detected only in the first seminal sample. For one man, three sequential samples, collected within 8 months, tested negative for HCV RNA. For another man (patient 3 in Table ) who provided seven successive samples within 11 months, all the seminal plasma fractions tested positive for HCV RNA (with the quantities ranging from <100 to 860 copies/ml). Detection of HCV RNA in spermatozoon fractions. | The corresponding motile spermatozoa (fraction 2) collected from the four patients whose seminal plasma was HCV RNA positive were tested for HCV RNA. In addition, 15 fractions 2 belonging to 10 patients whose seminal plasma was negative for HCV RNA were tested. All fractions 2 tested remained negative for HCV RNA with the exception of two of five samples collected from patient 3, in which HCV RNA was persistently detected in seminal plasma from July 2000 to May 2001. Fraction 2 of patient 3 was positive for HCV RNA only for a short time corresponding to the first 2 weeks of April 2001; moreover, the four samples of fraction 3 (swim-up spermatozoa) from patient 3 tested negative during the course of the study. The efficiency of the extraction step with the spermatozoon fractions was evaluated by the detection of protamin 2 gene RNA, and all specimens tested were found to be positive. In addition, cell samples were found to be positive for the internal control, demonstrating the absence of PCR inhibitors in these fractions. Assays after assisted reproduction. | Eleven attempts at medically assisted reproduction were performed for couples for whom the male partner tested negative for the presence of HCV RNA in seminal plasma. Five women became pregnant, with the occurrence of two singleton births and with one set of triplets and two sets of twins being born. The blood of the nine babies tested negative for HCV RNA at birth, and the babies remained uninfected at least 6 months later. For patient 3, given that HCV RNA was detected in all seminal plasma samples tested and in two motile spermatozoon fractions, the in vitro fertilization program with fresh semen was discontinued. When the motile spermatozoon fractions became negative, they were frozen in specific highly secure straws for future ART. For the three other patients whose seminal plasma was positive for HCV RNA, the ART were delayed. TABLE 1 | Characteristics of the four men whose seminal plasma was positive for HCV RNA TABLE 2 | Characteristics of patients whose blood was chronically infected with HCV according to presence or absence of HCV RNA in seminal fluid DISCUSSION : The overall results of this study show that the seminal plasma of 4 of 32 (12.5%) subjects chronically infected with HCV were found to be positive for HCV RNA by the RT-PCR technique. Previous studies on the same topic gave conflicting results: three failed to detect any HCV-positive seminal sample, and two reported anecdotal cases of HCV-positive semen. In addition, two reported that relatively high proportions of patients were positive for HCV RNA in this compartment: 8 of 21 (38%) patients and 4 of 20 (20%) patients in the studies of Leruez-Ville et al. and Pasquier et al. , respectively. The HIV serological status is an important point to be addressed since HIV has been shown to be a factor that dramatically enhances the pathogenicity of HCV in the livers of patients infected with both viruses . Altogether, 35 of the 41 patients tested in the two studies cited above were HIV positive , and 10 of them exhibited HCV RNA in their semen. In comparison, only six HIV-negative subjects (two of whom were positive for HCV RNA in their semen) were tested in the first study and no HIV-negative subjects were tested in the second one. These small numbers of patients do not allow statistical comparison of the prevalence of HCV RNA in the semen of HIV-negative subjects, especially since no data concerning the selection of patients were available. Conversely, the 12.5% rate of HCV RNA positivity for the semen of 32 consecutive HIV-negative subjects in the present study seems more representative of the actual rate of positivity for this compartment. The threshold of the assay reached 40 copies/ml, with a very low rate of false-negative results due to the presence of PCR inhibitors, as demonstrated by the use of a standardized commercial technique that included an internal control. This high rate of sensitivity compared to that of the protocol used in the standard AMPLICOR assay with blood samples can be explained by the addition of a step consisting of high-speed centrifugation of the sample. It was close to the one reported for seminal specimens by Leruez-Ville et al. , who used the same PCR kit used in the present study but a silica-based extraction technique. In addition to previous studies on the same topic, this report provides several new findings: (i) the amount of viral RNA in seminal plasma was larger than 500 copies/ml in three of the four positive patients and was up to 3,000 copies/ml in one patient; (ii) in all four of these patients, the genotypes of the strains in the blood and semen of the same patient were found to be identical; and (iii) a positive correlation between the viral loads in seminal plasma and blood was noted. None of these data describe the compartmentalization of HCV RNA between seminal plasma and blood. They suggest that the semen of patients who are chronically infected with HCV for a long time and who exhibit high viral loads are more likely to be positive for HCV RNA; similar findings have been shown for HIV-infected patients . Another finding original to this study was the detection of HCV RNA not only in seminal plasma but also in spermatozoon fractions used in ART. Relevant controls were used to validate the extraction and amplification steps performed on these cells: for all samples tested the RNA of the protamin 2 gene was amplified and the internal control of the PCR assay was positive. In one patient whose blood and seminal plasma were both chronically infected with HCV, two successive samples of the motile spermatozoon fraction were found to be positive for HCV RNA, whereas the same test performed with spermatozoa corresponding to the swim-up fractions was negative. This observation, which to our knowledge has never been described previously, needs discussion; although the possibility that spermatozoa support HCV replication cannot be definitively excluded, the more valuable explanation relies on the passive adsorption to the cell membrane of HCV RNA or virions present in seminal plasma, as suggested by the intermittent detection of HCV RNA in spermatozoa and by the absence of RNA in the corresponding swim-up fraction. In fact, the adsorption of viral particles to cell membranes could reduce the mobility of the spermatozoa, as described previously for cytomegalovirus . Even though the detection of HCV RNA in fractions of semen does not necessarily imply the presence of replicative virions and the infectivity of spermatozoa via medically assisted reproduction, our results plead for the reinforcement of precaution measures for men whose blood is chronically infected with HCV and who are candidates for ART. It is suggested that seminal plasma be systematically evaluated for the absence of viral RNA in order to ensure that ART are performed with samples free of HCV RNA. In those subjects found to be positive, it is proposed that motile spermatozoa be investigated for the presence of HCV RNA and that only spermatozoon fractions that test negative be used for ART. French authorities have recently recommended a similar algorithm . Since HCV loads can vary dramatically within a few days both in seminal plasma and in spermatozoa, the safest approach to ART for patients exhibiting HCV RNA in seminal plasma would consist of the use of cryopreserved cells that tested negative at the time of sampling. Although the uneventful deliveries of infants to the cohort evaluated in the present study are encouraging, the validation of this guideline would need to be assessed with a larger population of chronically infected individuals. Backmatter: PMID- 12202590 TI - Taxonomic Subgroups of Pasteurella multocida Correlate with Clinical Presentation AB - Pasteurella multocida is a small nonmotile gram-negative coccobacillus that is found in the nasopharynx and gastrointestinal tract of many wild and domesticated animals. In humans it most commonly causes cellulitis and localized superficial skin abscesses following an animal bite or scratch. The respiratory tract is the second most common site of infection for Pasteurella. Of the more than 17 species of Pasteurella known, Pasteurella multocida subsp. multocida and Pasteurella multocida subsp. septica are among the most common pathogens in humans. With the use of molecular techniques, distinction between different subspecies of P. multocida can be made more easily and accurately. We used the sequence of the 16S ribosomal DNA (rDNA) and repetitive extragenic palindromic sequence-PCR (REP-PCR) to characterize 20 strains (14 of P. multocida subsp. multocida and 6 of P. multocida subsp. septica; the 16S rDNA is identical for P. multocida subsp. multocida and Pasteurella multocida subsp. gallicida but differs from that of P. multocida subsp. septica) isolated from various anatomic sites. We found excellent correlation between the 16S rDNA sequence (a marker for a small conserved region of the genome), REP-PCR (a marker for a large portion of the genome), and biochemical tests (trehalose and sorbitol). We also found a correlation between the clinical presentation and the taxonomic group, with P. multocida subsp. septica more often associated with wounds than with respiratory infections (67 versus 17%, respectively) (P < 0.05, Z test) and P. multocida subsp. multocida more often associated with respiratory infections than with wounds (71 versus 14%, respectively) (P < 0.05, Z test). Keywords: Introduction : Pasteurella multocida is a small nonmotile gram-negative coccobacillus that is found in the nasopharynx and gastrointestinal tract of many wild and domesticated animals. It does not often result in disease in its animal hosts. However, in humans it most commonly causes cellulitis and localized superficial skin abscesses following an animal bite or scratch . The respiratory tract is the second most common site of infection for Pasteurella, where it may present as pneumonia, tracheobronchitis, abscess, or empyema. The organism may also cause upper respiratory infections such as sinusitis and pharyngitis . Of these, pneumonia is the most common manifestation of respiratory infection caused by Pasteurella, and patients may acutely or insidiously present with fever, dyspnea, and pleuritis. Those who develop respiratory infection from Pasteurella also tend to be elderly and have underlying chronic lower respiratory tract disease, and the route of infection appears to be through inhalation. The organism may also be opportunistic and affect immunocompromised patients as well, causing pneumonia in patients with AIDS and immunoglobulin A deficiency . Much less frequently, Pasteurella causes osteomyelitis, intra-abdominal infections, septic arthritis, sepsis, and meningitis . Of the more than 17 species of Pasteurella known, Pasteurella multocida subsp. multocida, Pasteurella multocida subsp. septica, Pasteurella canis, Pasteurella stomatis, and Pasteurella dogmatis are the most common pathogens in humans . Cats have the highest rate of oropharyngeal colonization by P. multocida (50 to 90%), followed by dogs (50 to 66%), pigs (51%), and rats (14%) . This distribution is reflected by the greater chance of isolating Pasteurella from cat bites (50%) than from dog bites (20 to 30%), although dog bites are much more common . In most cases of disease, the organism has been acquired either directly through bites or aerosol inhalation or indirectly by contact with fomites contaminated with animal secretions. Interestingly, Pasteurella sp. may also become part of the normal respiratory tract flora in humans. It has been found in healthy veterinary students and animal handlers without any pulmonary symptoms . As for other organisms, with the use of molecular techniques, distinction between different subspecies of P. multocida can be made more easily and accurately . One can detect whether there are any genotypic or subspecies differences between the organisms that cause skin infections and those that cause respiratory infections. Sequencing of the 16S ribosomal DNA (rDNA) and repetitive extragenic palindromic sequence-PCR (REP-PCR) have been used previously to distinguish strains of Francisella, Bartonella, and Mycobacterium (, ; T. Raich and J. E. Clarridge, Abstr. 95th Gen. Meet. Am. Soc. Microbiol. 1995, abstr. U-99, p. 134, 1995). We used these methods to characterize strains of Pasteurella isolated from various human sites. MATERIALS AND METHODS : Identification of strains and correlation with clinical presentation. | The 20 P. multocida isolates were determined from the examining microbiology records for the years between 1985 and 2002 at the Veterans Administration Medical Center in Houston, Tex. These strains were identified originally by conventional microbiological and biochemical methods including Gram stain, colony morphology, oxidase reaction, and RapID NH identification kits (Remel Inc., Lenexa, Kans.). At the time of identification, a stock of each organism was frozen at -70C with glycerol. For this study, the organisms were subcultured on chocolate agar in 8% CO2 for 24 to 36 h. Additional biochemical tests were performed using the API 20E and Rapid ID 32 Strep kits (Biomerieux, Lyon, France). Patients' histories were reviewed from computer records when available. Statistical significance was determined by the Z test. REP-PCR. | The organisms were harvested and resuspended in 0.9% sterile saline to a 3.0 McFarland standard turbidity. The PCR primers REP1R and REP2-1 as well as previously described methods using conserved primers to REP sequences in PCR were used (, ; Raich and Clarridge, Abstr. 95th Gen. Meet. Am. Soc. Microbiol. 1995). The reaction mixture consisted of the following: 2 mul of bacterial suspension, 50 mM KCl, 10 mM Tris HCl, 0.1% Triton X-100, 3.0 mM MgCl2, 10% dimethyl sulfoxide, 200 muM (each) deoxynucleoside triphosphates, 100 pmol of each primer, and 1 U of Taq polymerase (Promega, Madison, Wis.). Amplification was performed on a Perkin-Elmer GeneAmp 9600 PCR system (Perkin-Elmer, Norwalk, Conn.) as follows: 1 cycle of 95C for 5 min; 35 cycles of 94C for 1 min, 40C for 1 min, and 65C for 4 min; 1 cycle of 65C for 10 min; and final holding at 4C until analysis. The PCR products were separated by electrophoresis on a 1.2% agarose-ethidium bromide gel in 0.5% TBE buffer (0.045 M Tris-borate, 0.001 M EDTA) at 80 V. 16S rDNA sequence analysis. | 16S rRNA gene sequence identification was performed using the MicroSeq 500 Gene Kit (Applied Biosystems, Foster City, Calif.) and the 3100 Genetic Analyzer (Hitachi, Tokyo, Japan) according to manufacturer's specifications. Approximately 500 bp in both forward and reverse sense were sequenced for each isolate. Test strain sequences were compared against the MicroSeq 16S rRNA gene sequence database. The database contains sequences from 1,297 different species (1,187 type strains) including type strains from the genus Pasteurella-Haemophilus-Actinobacillus groups. Sequence data obtained from GenBank for nontype strains were also included in the analysis. Sequences were compared in dendrogram form by the neighbor-joining method using the MicroSeq statistical package . RESULTS : Analysis of the REP-PCR products revealed the presence of two predominant patterns which we termed patterns A and B. Although there are variations in the number of bands present among the different specimens belonging to one particular pattern, each group shares common characteristics. Pattern A is distinguished by a closely separated double band at the 1,600- to 2,000-bp region (arrows in Fig. ). However, pattern B is recognizable by the common presence of a band migrating slightly below 1,400 bp (asterisk in Fig. ). The variable presence of some bands may be partially due to differences in the intensity of staining. The separation of the specimens into two distinct groups correlates with the subspecies classification of each organism by 16S rDNA sequence. By comparing 16S rDNA sequence to known sequences of P. multocida subsp. multocida and P. multocida subsp. septica, one can assign each isolate to either subspecies . The sequence of the 16S rDNA showed very high homology between P. multocida subsp. septica and P. multocida subsp. multocida. By this method, the organisms that showed pattern A by REP-PCR also had a 16S rDNA sequence homologous to that of P. multocida subsp. septica. In contrast, those with pattern B had the 16S rDNA sequence of P. multocida subsp. multocida. Of note, P. multocida subsp. multocida and Pasteurella multocida subsp. gallicida have identical 16S rDNA sequences. The subspecies P. multocida subsp. multocida and P. multocida subsp. septica share very similar biochemical profiles. With the identification kits mentioned in Materials and Methods, consistent differences were seen only in the sorbitol and trehalose biochemical tests. In our isolates, P. multocida subsp. multocida was positive for sorbitol and negative for trehalose metabolism, whereas P. multocida subsp. septica was variable for sorbitol and positive for trehalose. We correlated the site and disease process from which each isolate was obtained with the subspecies. Of the 20 strains studied, 14 belonged to P. multocida subsp. multocida subspecies, and the other 6 were classified as P. multocida subsp. septica . The majority (10 of 11, or 91%) of the strains causing respiratory infections were of P. multocida subsp. multocida. In addition, there were two nonwound and nonrespiratory cases caused by P. multocida subsp. multocida: one case of bacteremia and one case of peritoneal cavity seeding. In contrast, four of the six strains of P. multocida subsp. septica were isolated from various wound infections, while only one was seen in a case of pneumonia. The last strain of P. multocida subsp. septica was isolated from the blood. Thus, P. multocida subsp. septica was more often found in infected wounds than in respiratory sites, but the number was not statistically significant. In summary, the relative proportions of wound and respiratory cases were statistically different between the subspecies (P = 0.05, Z test). Of the respiratory cases, P. multocida subsp. multocida was more likely the cause (P < 0.05, Z test). Conversely, P. multocida subsp. multocida had a greater propensity to cause respiratory rather than wound infections (P = 0.05, Z test). Upon review of patient records, most of the patients diagnosed with Pasteurella respiratory infections also had underlying chronic pulmonary disease, including chronic obstructive pulmonary disease and primary lung cancer. These cases often presented with a mild form of pneumonia. Also, most of the patients with cases of wound infection had prior animal contact, and the cases were characterized by a localized skin infection. FIG. 1. | Gel of REP-PCR products from various clinical isolates. Gel of REP-PCR products from various clinical isolates. Lanes 1 and 8 represent molecular size markers (Sigma, St. Louis, Mo.). Lanes 2 to 7 (pattern A) were confirmed as P. multocida subsp. septica by 16S rDNA sequence. In contrast lanes 9 to 18 (pattern B) were demonstrated to be P. multocida subsp. multocida. The double bands marked by arrows are between approximately 1,600 and 2,000 bp in size. The asterisk denotes the band migrating slightly below the 1,400-bp marker. FIG. 2. | Neighbor-joining tree of P. Neighbor-joining tree of P. multocida clinical isolates based on their 16S rDNA sequences. The isolates (those designated with a number) are grouped with either P. multocida subsp. multocida or P. multocida subsp. septica based on their 16S rDNA sequence similarity. Type strains are noted by their full genus and species name and are ATCC 43137, ATCC 19427, and ATCC 29977 for P. multocida subsp. multocida, P. multocida subsp. septica, and P. multocida subsp. gallicida, respectively. Escherichia coli is the outgroup. TABLE 1 | Subspecies distribution of clinical isolates DISCUSSION : By two independent molecular techniques, distinct strain differences were identified between divergent disease processes. In our series, Pasteurella respiratory infections most often involved P. multocida subsp. multocida. Interestingly, P. multocida subsp. multocida itself seems to have a propensity for respiratory sites and infections. However, P. multocida subsp. septica causes respiratory symptoms less often and was found more frequently in infected wounds. The 16S rDNA sequence has been accepted as a reliable barometer of evolutionary change between different bacterial species. It shows distinct differences between subspecies as well, which makes the technique ideal for our purposes. Similarly, in the clinical setting DNA sequencing and not biochemical assays may be the most reliable and efficient method to distinguish between the different subspecies of Pasteurella. The various clinically significant subspecies of Pasteurella share many similar biochemical properties, and results are often equivocal. In contrast to earlier published reports , we found that the metabolism for trehalose is distinctly different between P. multocida subsp. multocida and P. multocida subsp. septica. Furthermore, the process of studying patterns in the genomic DNA by REP-PCR also appears to be accurate and sensitive. With it, we were able to distinguish between P. multocida subsp. multocida and P. multocida subsp. septica. Since REP-PCR can detect differences in the entire genomic DNA by virtue of the various locations of the repetitive extragenic sequences, it may be more sensitive than 16S rDNA sequencing in other organisms. We found excellent correlation between the 16S rDNA sequence (a marker for a small conserved region of the genome), REP-PCR (a marker for a large portion of the genome), and biochemical tests (trehalose and sorbitol). Although the number of the cases is small, we found a statistically significant correlation with the type of disease. Thus, in the case of Pasteurella, genotypic and phenotypic differences between the various subspecies may result in divergent patterns of pathogenicity. Further investigation would be useful to clarify these relationships. Backmatter: PMID- 12202601 TI - Rate of Hepatitis B Virus Infection in Pregnant Women Determined by a Monoclonal Hepatitis B Surface Antigen Immunoassay AB - The rate of HBsAg in 6,976 B-human chorionic gonadotropin (B-hCG)-positive specimens, as determined by the Auszyme Monoclonal assay (Abbott Laboratories, Abbott Park, Ill.), was 0.56% (39 of 6,986 repeatedly reactive [RR] and confirmed-positive specimens). All RR and confirmed specimens were hepatitis B virus positive by at least one additional test, yielding an assay specificity of 99.96%. The findings argue against unique attributes in the pregnant population that might produce inaccurate assay results. Keywords: Introduction : An estimated 20,000 infants are born to HBsAg-positive women in the United States each year . Because these infants are at high risk of perinatal hepatitis B virus (HBV) infection, chronic HBV infection, and chronic liver disease, the American College of Obstetricians and Gynecologists, the American Academy of Family Practices, the American Academy of Pediatrics, and the Centers for Disease Control and Prevention (CDC) Advisory Committee on Immunization Practices have recommended that all pregnant women undergo testing for HBsAg prior to delivery . The objective of this study was to examine the rate of HBV infection in specimens from pregnant females using the Auszyme Monoclonal assay. We investigated whether pregnancy had any potential influence on the specificity of the Auszyme Monoclonal assay results by performing the study under conditions that minimized sample cross-contamination and by using additional HBV marker verification of positive samples. In phase I of this study, all specimens were from females and were B-human chorionic gonadotropin (B-hCG)-positive sera or plasma specimens at the reference laboratory, had a volume of 2 ml or greater, and had not passed through a viral accessioning or testing area. The reference laboratory (Quest Diagnostics, Teterboro, N.J.) aliquoted each sample from the main specimen tube, marked each sample vial with the qualitative or quantitative B-hCG result and a unique identifier number, and shipped the samples by overnight delivery to Abbott Laboratories. The Auszyme Monoclonal assay was performed on all samples in accordance with procedure C (incubation at 40C for 75 min) of the package insert. Initially reactive samples were tested again in duplicate. If neither of the repeat tests was reactive, the specimen was considered negative for HBsAg. If either retest was reactive, the sample was considered repeatedly reactive (RR) and was then tested by the Auszyme confirmatory assay through procedure A. Only those specimens for which RR results were neutralized by the confirmatory procedure were considered positive for HBsAg (HBsAg confirmatory assay package insert [dated 1995], Abbott Laboratories Diagnostics Division, Abbott Park, Ill.). These confirmed HBsAg-positive specimens were then tested by two additional licensed HBsAg assays: the IMx HBsAg assay (Abbott Laboratories), an automated microparticle-based assay with a monoclonal antibody capture phase and an enzyme-linked polyclonal antibody detection phase, and the Ortho Antibody to HBsAg ELISA Test System 2 (Ortho-Clinical Diagnostics, Raritan, N.J.), a microtiter assay using monoclonal antibody capture on the solid phase and an enzyme-linked monoclonal antibody detection phase. Additional tests were performed according to the manufacturer's package insert when there was sufficient sample volume. These tests included the CORAB radioimmunoassay (Abbott Laboratories), which detects HBV core protein-specific antibody; the HBe EIA (Abbott Laboratories), which detects HBeAg; and an in-house research assay for HBV DNA that uses nested PCR. Phase II of the study was carried out by Abbott Laboratories and the Laboratory Corporation of America (LabCorp, Elmhurst, Ill.) reference laboratory. B-hCG-positive specimens provided by New York Biologics, Inc. (New York, N.Y.), were collected using the same criteria employed in phase I of this study, along with a signed patient informed-consent form. Aliquots of the same sample were shipped in parallel both to Abbott Laboratories and to the LabCorp reference laboratory, where the Auszyme Monoclonal assay was performed on all samples. Any initially reactive, RR, or confirmed-reactive sample identified at LabCorp was then tested at Abbott Laboratories using the pristine parallel sample. Discordant samples between the two sites were subjected to the testing described for phase I above. New York Biologics requested that samples showing a low-level reaction, i.e. an Auszyme sample-to-cutoff ratio between 1 and 2, be redrawn from the patients. The redrawn samples were evaluated in the same manner as the initial samples. A population size of 1,286 would be needed to statistically validate an assay showing a 0.4% rate of prenatal HBV infection, which is the lowest rate reported by the CDC (upper and lower limits of 0.0067 and 0.0021, respectively, at a confidence level of 95% with binomial distribution). In case the rates of prenatal HBV infection had dropped significantly since the original CDC report in 1994, we chose to assume the possibility of a 10-fold-lower infection prevalence of 0.04% to establish the number of specimens needed for the study. The targeted study population sizes of 5,000 for phase I and 2,000 for phase II would represent the midpoint of the population size required to validate a prenatal HBV infection rate of 0.04% (6,762 specimens at the 95% confidence level with binomial distribution). Auszyme Monoclonal assay specificity was determined using the following formula: specificity = (number of true-negative specimens/[number of true-negative specimens + number of false-positive specimens]) x 100. During phase I of the study, 4,988 specimens were received and tested. Of these samples, 34 (0.68%) were RR specimens and 31 of the 34 (0.62% of the total sample population) were neutralized in the confirmatory assay and considered positive for HBsAg. All 31 of the Auszyme Monoclonal HBsAg-positive specimens were reactive by the IMx HBsAg assay. Thirty of these specimens were also either reactive by the Ortho Antibody to HBsAg ELISA Test System 2 or reactive by testing for another HBV marker. Therefore, all Auszyme Monoclonal HBsAg-positive specimens were confirmed by at least one additional HBV marker test. Table summarizes the RR and confirmatory rates for the phase I study. The HBV marker profile for the phase I specimens is shown in Table . The rate of RR, nonconfirmed specimens was 0.06% (3 of 4,988 specimens). TABLE 1 | RR, confirmed prenatal specimens from study phases I and II TABLE 2 | Detailed HBV test results for RR, prenatal specimens from study phase I (n = 34) In phase II of the study, 1,998 specimens were tested by the Auszyme Monoclonal assay at the LabCorp reference laboratory. The rates of RR and confirmed-positive specimens for this part of the study are shown in Table . Nine specimens were RR at LabCorp. All nine samples were RR again when they were tested at Abbott Laboratories, indicating an RR rate of 0.45%. Of these nine RR specimens, one specimen was RR and confirmed to be positive at LabCorp but was not confirmed to be a neat specimen at Abbott. This sample had to be tested in the HBsAg confirmatory assay at a 1:25 dilution, due to insufficient sample volume, which may provide an explanation for the lack of confirmation. A second sample was redrawn from this patient 4 months after the initial sample collection. The Auszyme Monoclonal assay result was negative for the redrawn specimen, while the CORAB result was positive, suggesting that the original specimen was truly HBsAg positive. Of the 6,986 prenatal specimens from both phases of the study, the overall rate of RR specimens was 0.62% (43 of 6,986 specimens) and the rate of RR and confirmed-positive specimens was 0.56% (39 of 6,986 specimens) . Under these conditions, the specificity of the Auszyme Monoclonal assay was 99.96% [6,947 of (6,947 + 3) specimens], with a 95% confidence interval of 99.87 to 99.99% by binomial distribution. Licensed HBsAg tests have a very high specificity and sensitivity if reactive tests are repeated and confirmed by neutralization, as is recommended by the manufacturers. However, non-RR or unconfirmed (nonneutralized)-reactive tests do occur . There are a number of possible reasons for such results. First, the patient may have an immune response to a vaccine . Second, there might be positive interference in the mouse monoclonal antibody-based immunoassays caused by the presence of human anti-mouse antibodies. The most common cause of human anti-mouse antibody interference is prior use of mouse monoclonal antibodies for therapeutic or imaging purposes . Third, a technician may perform a laboratory technique improperly, such as washing beads incorrectly or failing to remove particulate matter from specimens prior to testing (Auszyme Monoclonal package insert [dated 1995], Abbott Laboratories Diagnostics Division). Fourth, truly cross-reactive endogenous proteins may be present in the specimen. Neutralized, RR specimens that do not correlate with the negative disease status of a patient exist. In these cases, nonreactive specimens may be contaminated by the transfer of antigen from positive samples, either through soiled equipment or the aerosolized transfer of antigen (Auszyme Monoclonal package insert [dated 1995] and unpublished data, Abbott Laboratories Diagnostics Division). The proportion of high-risk individuals in a previously reported study population might have increased the likelihood of sample cross-contamination, resulting in an unusually high rate of false-positive test results with the Auszyme Monoclonal assay . In rare cases transient (neutralizable) antigenemia has been reported for both adults and infants postvaccination with HBV vaccine . In these cases, assays detect residual vaccine. Due to the above-listed factors, there is some concern that routine prenatal testing may lead to an increase in the number of enzyme immunoassay false-positive results in what is perceived as a low-risk population. This study was designed to determine the rate of HBV infection in a general pregnant population and to evaluate the suitability of the Auszyme Monoclonal assay for widespread prenatal HBV screening. The approach was to use specimens that had never passed through a viral testing laboratory area and were not likely to have been exposed to cross-contamination by HBV-reactive samples. Moreover, several independent markers for HBV were determined for the HBsAg-reactive specimens to provide a profile of the infection stage . Finally, in phase II of this study, the positive samples were evaluated in parallel by two separate testing sites, which yielded equivalent enzyme immunoassay results. The results reported here for the Auszyme Monoclonal assay indicate an overall RR rate of 0.62% and an RR, confirmed rate (determined as recommended by the manufacturer) of 0.56%, with an RR, nonconfirmed rate of only 0.04%. All of the RR, confirmed-positive specimens that had sufficient volume to undergo any further testing were confirmed to be truly positive by at least one additional test from a panel of tests that are indicative of viral load or different stages of HBV infection . Based on these data, the specificity of the Auszyme Monoclonal assay is 99.96%. The overall HBsAg prevalence rate obtained in this study, 0.56%, is in excellent agreement with the 0.50% rate of positive HBsAg tests reported by the CDC when results from several studies involving pregnant women were averaged . Three different HBsAg assay formats and multiple-marker verification of the RR, confirmed specimens suggest that there is no unique factor associated with pregnancy which led to an RR, confirmed specimen that was not truly positive. In conclusion, in the present study, the rate of HBV infection in a total of 6,986 specimens obtained from pregnant females was 0.56%, a value that agrees well with prior CDC estimates for this population. Backmatter: PMID- 12202560 TI - Dichotomy of Glycoprotein G Gene in Herpes Simplex Virus Type 1 Isolates AB - Herpes simplex virus type 1 (HSV-1) encodes 11 envelope glycoproteins, of which glycoprotein G-1 (gG-1) induces a type-specific antibody response. Variability of the gG-1 gene among wild-type strains may be a factor of importance for a reliable serodiagnosis and typing of HSV-1 isolates. Here, we used a gG-1 type-specific monoclonal antibody (MAb) to screen for mutations in the immunodominant region of this protein in 108 clinical HSV-1 isolates. Of these, 42 isolates showed no reactivity to the anti-gG-1 MAb. One hundred five strains were further examined by DNA sequencing of the middle part of the gG-1 gene, encompassing 106 amino acids including the immunodominant region and epitope of the anti-gG-1 MAb. By phylogenetic comparisons based on the sequence data, we observed two (main) genetic variants of the gG-1 gene among the clinical isolates corresponding to reactivity or nonreactivity to the anti-gG-1 MAb. Furthermore, four strains appeared to be recombinants of the two gG-1 variants. In addition, one strain displayed a gG-1-negative phenotype due to a frameshift mutation, in the form of insertion of a cytosine nucleotide. When immunoglobulin G reactivity to HSV-1 in sera from patients infected with either of the two variants was investigated, no significant differences were found between the two groups, either in a type-common enzyme-linked immunosorbent assay (ELISA) or in a type-specific gG-1 antigen-based ELISA. Despite the here-documented existence of two variants of the gG-1 gene affecting the immunodominant region of the protein, other circumstances, such as early phase of infection, might be sought for explaining the seronegativity to gG-1 commonly found in a proportion of the HSV-1-infected patients. Keywords: Introduction : Herpes simplex virus (HSV) is an alpha-herpesvirus with two subtypes, HSV type 1 (HSV-1) and HSV-2, which both establish latent infections with different or similar clinical expressions during reactivation . The ability to discriminate between HSV-1 and HSV-2 infections by serological means has been of importance for several clinical aspects, including diagnosis and treatment of neonatal herpes as well as of complications from the central nervous system . Glycoprotein G-1 (gG-1), a viral envelope glycoprotein that was suggested to contribute to viral entry through apical surfaces of polarized cells , has been utilized as a prototype antigen for HSV-1 type-specific diagnosis due to lack of cross-reactivity with its counterpart in HSV-2, i.e., gG-2 . Several commercial gG-based enzyme immunoassays have been evaluated in clinical settings . Furthermore, gG-1 may be a suitable target for typing of HSV isolates by monoclonal antibodies (MAb) that identify type-specific epitopes. Conservation of the gene coding for gG-1 among clinical HSV-1 isolates might therefore be a prerequisite for a reliable assessment of a type-specific antibody response as well as of correct typing of isolates in individuals infected with HSV-1. In a previous study, several point mutations were found when the gG-1 genes of 11 HSV-1 clinical strains and two reference strains were sequenced , and some of these mutations were present within a recently defined immunodominant region including the epitope of a type-specific anti-gG-1 MAb . Genetic variability of this gene among clinical isolates may therefore be of importance for the immunoglobulin G (IgG) response within the individual host when a predefined gG-1 antigen is used for detection. Previously, a few studies have described a limited genetic variability of genes coding for envelope glycoprotein among HSV-1 strains as shown by DNA sequencing of the gB-gene or the gD gene after PCR amplification of cerebrospinal fluid samples. These studies observed some point mutations which were mostly silent, and the insertion or deletion of one or a few codons compared to a previously published HSV-1 sequence . In contrast to this, we here report the identification of two major genetic variants of the gG-1 gene based on DNA sequencing and investigation of MAb reactivity to a type-specific gG-1 epitope in a large number of clinical isolates. Altogether, 42 strains were found to be unreactive to the anti-gG-1 MAb among the 108 HSV-1 clinical isolates, a finding that was explained by the existence of two genetic variants discovered when the strains were subjected to DNA sequencing of the middle part of the gG-1 gene. However, no differences in IgG reactivity were found when sera from patients harboring either of these variants were compared by gG-1 enzyme-linked immunosorbent assay (ELISA). MATERIALS AND METHODS : Patients and viral strains. | The study material is schematically presented in Fig. . One hundred six patients with reactivated herpetic lesions from different anatomical sites (oral, facial, genital, perianal, and others; see Table ) were randomly chosen for investigation, and 108 virus strains derived from these 106 patients were consecutively received at the Department of Clinical Virology, Goteborg. Green monkey kidney (GMK) cells were used for isolation of the virus strains, which all were typed as HSV-1 by the use of a type-specific anti-gC MAb . Of the 106 patients, documentation of gender was available for 76 subjects, of whom 54 were women and 22 were men. From 53 patients, a serum sample was drawn and included for serological assays. From two of these patients, two isolates were investigated: one strain from the thoracic region and one from penis in one patient, and one isolate from the lip and one from the genital region in another patient. The study was approved by the Medical Ethical Committee in Goteborg, Sweden, (approval no.: S 266-00). MAbs. | The following HSV-1 type-specific MAbs were used: a commercially available anti-gG-1 MAb (Advanced Biotechnologies Incorporated), which was previously mapped by a pepscan technique to the amino acids AFPL at position 110 to 113 , the anti-gC-1 MAb B1C1B4, and the anti-gB-1 MAb 1B11D8 . For typing of HSV-2 strains, the anti-gG-2 MAb O1C5B2 was used. Type-specific serology. | An ELISA was used to determine type-common and HSV-1 and HSV-2 type-specific antibody reactivity in serum samples. In brief, the type-common antigen was based on sodium deoxycholate-solubilized membranes from cells infected with HSV-1 strain F . The HSV-1 type-specific antigen was based on the gG-1 protein (kindly provided by SmithKline Beecham Biologicals), and seroreactivity was analyzed by ELISA as previously described . The HSV-2 type-specific assay was based on Helix pomatia lectin-purified gG-2 . For the type-common and HSV-2 ELISA assays, serum samples were diluted to 1/100, and for the HSV-1 ELISA, serum samples were diluted to 1/50 for further titration. Alkaline phosphatase-conjugated, affinity-purified F(ab)2 fragment goat anti-human IgG (Jackson ImmunoResearch Lab) was used at a 1/3,500 dilution. p-Nitrophenyl phosphate was used as a substrate at a concentration of 1 mg/ml. The A405 value was measured with a reference wavelength of 650 nm against a substrate blank. The cutoff value was calculated using the optical density (OD) value of the negative controls (previously found to be HSV-negative by HSV-1 and HSV-2 Western blot) plus 0.2 OD U for the type-common, as well as for the gG-1, ELISA. In the gG-2 ELISA the OD value of high-titrating HSV-1-positive sera plus 0.1 OD U was used as a cutoff value. When comparing the two groups, the netto-absorbance (absorbance minus background, which was bovine serum albumin) value was used, in order to give consideration to differences that might occur when developing the plates. Indirect immunofluorescence. | To investigate whether subjects with HSV-1 IgG-negative sera experienced a primary infection, IgM antibodies were analyzed by an in-house immunofluorescence assay using HSV-2-infected GMK cells as antigen. The assay was performed as described previously . ELISA with infected cells. | For screening of mutations within type-specific epitopes of gG-1 by the use of a MAb, an ELISA was performed with cells infected with the different virus strains. GMK cells were grown on 96-well plates in Eagle's minimal essential medium supplemented with 1% PEST. The cells were infected with the HSV-1 strains at an infectious dose of 106 PFU/ml. The anti-gG-1 MAb was used at a dilution of 1:500, and the MAbs B1C1B4 (reactive with gC-1) and O1C5B2 (reactive with gG-2) were used at a dilution of 1:50. As conjugate for the MAbs, alkaline phosphate- conjugated F(ab)2 goat anti-mouse IgG was used at a dilution of 1:2,000 (Jackson ImmunoResearch Lab). As the substrate, p-nitrophenylphosphate was used at a concentration of 1 mg/ml. The A405 value was measured with a reference wavelength of 650 nm against a substrate blank. DNA sequencing of the gG-1 gene. | Viral DNA was prepared from stock viruses by using the QIAmp blood kit (Qiagen) method, and the DNA was subjected to PCR amplification. Three overlapping oligonucleotide pairs were used as primers for complete gG-1 gene sequencing as described previously . Since our previous studies showed that the immunodominant region, as well as most of the mutations, was located within the region amplified by the second primer pair, this part was selected for sequencing of all isolates with the exception of three strains. In addition, a group of consecutively received isolates was chosen for sequencing of the complete gG-1 gene, preferentially those for which a corresponding serum sample was collected. Amplified products were separated on a 1% agarose gel prior to extraction of the amplicon bands with the QIAquick gel extraction kit (Qiagen). PCR cycle sequencing was carried out by using the fluorescence-labeled stop nucleotides with dRhodamine terminator cycle sequencing ready reaction kit (Applied Biosystems). The sequencing reaction was carried out in both sense and antisense directions for confirmation and also to act as an internal control. After precipitation with ethanol, the labeled samples were analyzed on an automated sequencer (ABI Prism 310 genetic analyzer; Applied Biosystems). In addition, DNA sequence data from the gG-1 gene in 10 HSV-1 strains from a previous study were included for comparison of deduced amino acid sequences. Phylogenetic analysis. | After alignment of 105 sequences from the present study and 10 database sequences (including KOS321, F, and Syn17+), phylogenetic comparison was done, after bootstrapping to 500 replicates, by distance matrix/UPGMA analysis using the MEGA2 software package available at . Trees were first constructed by comparison of an entire sequence (nucleotides [nt] 136950 to 137238 in Syn17+ according to GenBank) . For evaluation of recombination, separate trees comparing two parts, nt 1 to 100 and nt 101 to 288, was also done . Statistical analysis. | Statistical analyses were performed by Student's t test. Nucleotide sequence accession numbers. | The nucleotide sequences of the viruses were submitted to GenBank and given the accession numbers to . FIG. 1. | Schematic presentation of the virus isolates, DNA sequencing of the gG-1 gene, and availability of corresponding serum samples. Schematic presentation of the virus isolates, DNA sequencing of the gG-1 gene, and availability of corresponding serum samples. *, in two additional strains in each group lacking corresponding sera, the gG-1 gene was also completely sequenced. FIG. 2. | (A to C) Phylogenetic trees created by distance matrix/UPGMA analysis using MEGA version 2. (A to C) Phylogenetic trees created by distance matrix/UPGMA analysis using MEGA version 2.0. Bootstrap values above 70 are displayed. The phylogenetic trees included 30 unique sequences after comparison of all 115 gG-1 sequences. The AF120934 (strain F) sequence was found in 35 further samples, NC_001806 (Syn17+) was found in 18 samples, AF116193 was found in 25 samples, 66 was found in 3 samples, and 49, 55, 64, and 72 were found in 1 additional sample each. The trees in panel A are based on the whole 288-nt sequence. Panels B and C were dated by comparing the segments 1 to 100 and 101 to 288 separately to evaluate suspected recombination in five strains (34, 66, 67, 70, and AF117121), denoted in bold. TABLE 1 | Sites of lesions and anti-gG-1 MAb reactivities of 108 clinical HSV-1 isolates RESULTS : Reactivity of an anti-gG-1 MAb to clinical HSV-1 strains. | When the 108 HSV-1 isolates were investigated for reactivity using an anti-gG-1-specific MAb by ELISA on infected GMK cells, 42 clinical isolates showed no reactivity, and these strains were derived from different anatomical sites . In contrast, all isolates were reactive with the anti-gC-1 MAb, and none were reactive with the anti-gG-2 MAb. As a likely explanation, all of the 42 HSV-1 strains contained nonsynonymous mutations (see below) within the immunodominant region harboring the epitope region of the anti-gG-1 MAb, including the amino acid F111->V mutation situated within the previously mapped AFPL epitope of the MAb . Where the sex of the patient could be identified, 54 strains were isolated from females and 22 were isolated from male hosts, but the reactivity to the anti-gG-1 MAb was unrelated to gender. Of the two patients who each had contributed two strains, one harbored two MAb-negative isolates while the other presented two MAb-positive isolates. DNA sequence variations within the gG-1 gene. | To more accurately define the extent of genetic variability among the MAb-positive and MAb-negative isolates, nucleotide sequences were determined for the middle part (106 amino acids) of the gG-1 gene including the immunodominant region and the type-specific epitope region for the anti-gG-1 MAb in 105 of the 108 strains . The results from DNA sequencing showed that all 42 clinical isolates which were unreactive with the anti-gG-1 MAb showed mutations similar to those that we observed in a previous study for strain KOS 321, which in concordance also was unreactive to the anti-gG-1 MAb. The results, including those from phylogenetic comparison, showed that two main genetic variants of the gG-1 gene exist among clinical and laboratory HSV-1 strains in the Western world (Fig. and ). The complete gG-1 gene was sequenced in 25 consecutively received isolates, 21 strains of which corresponding sera were included in the serological assays. Sequence alignment of these and previously reported sequences suggested the existence of two genetic groups of HSV-1 strains, one represented by KOS321 and the other by Syn17+. This interpretation was supported by high bootstrap values in the phylogenetic tree based on 115 sequences from the middle part of the gG-1 gene . When five putative recombinants were excluded, the bootstrap value was in fact 100% (not shown). The nucleotide difference between the groups was 6.9%, compared to 1.0 and 0.7% within the KOS321 and Syn17+ groups, respectively. From the alignment it was apparent that sequence 34 (Fig. A to C) was similar to Syn17+ in the first part of the sequence (nt 1 to 100) but similar to KOS321 in the second part (nt 101 to 288). The converse was observed for sequences 66, 67, 70, and AF1177121. Phylogenetic analysis of the two segments further supported that these sequences represent recombination, with sequence 34 showing a Syn17+-KOS pattern and 66, 67, 70, and AF117121 showing a KOS-Syn17+ pattern. Detection of gG-1-negative virus due to a single frameshift mutation. | During the sequencing of the 42 HSV-1 MAb-negative strains, one isolate was detected that harbored a single frameshift mutation within the gG-1 gene with a subsequent lack of expression of the type-specific gG-1 epitope. A single insertion of a cytosine nucleotide within a stretch of six cytosines at the position consisting of nt 211 to 216 (amino acids 71 to 72) was observed, which resulted in a different reading frame and a premature stop codon at position 91 . Interestingly, the serum sample drawn from the patient harboring the gG-1-negative isolate showed the presence of IgG antibodies reactive with the gG-1 antigen at a titer of 1/800. Furthermore, this strain was isolated from the genital region. ELISA assay. | Fifty-three serum samples obtained from patients from which 55 HSV-1 strains were isolated were investigated in this study, and the OD values are shown in Fig. . Of these serum samples, 36 sera corresponded to anti-gG-1 MAb-reactive strains and 17 sera corresponded to MAb-unreactive isolates, respectively . In total, 10 sera were negative in the type-specific gG-1 ELISA, of which 7 sera were derived from patients harboring MAb-reactive isolates and 3 sera were derived from patients carrying MAb-unreactive strains. This seronegativity to gG-1 was considered to be due to samples being collected during early seroconversion in seven cases. Of these, two sera were HSV IgM-positive by immunofluorescence. The immunofluorescence test is based on HSV-2 and has a lower sensitivity to HSV-1, which could explain why the test was negative with the other sera. Two sera were reactive in a low titer in the type-common HSV test, which sometimes shows early reactivity, and three additional patients had primary genital HSV infection as judged clinically. Of the three remaining patients with gG-1-seronegative sera, one carried a MAb-reactive isolate and two carried MAb-unreactive isolates. The remaining 43 sera were all positive in the type-common and HSV-1 type-specific gG-1 ELISA. Only two sera were positive in the HSV-2 type-specific ELISA. In the HSV type-common seroassay, mean OD values +- standard deviations were 0.68 +- 0.47 in sera from patients harboring anti-gG-1 MAb-reactive isolates versus 0.73 +- 0.56 in sera from hosts of anti-gG-1 MAb-unreactive isolates . In the HSV type-specific gG-1 seroassay (gG-1), mean OD values +- standard deviations were 0.55 +- 0.50 in sera from patients harboring anti-gG-1 MAb-reactive isolates versus 0.43 +- 0.38 in sera from hosts of anti-gG-1 MAb-unreactive isolates. None of these differences in OD values between the two groups were statistically significant. FIG. 3. | Aligned deduced amino acid sequences of the complete gG-1 gene for 14 variants discovered in 38 clinical and reference strains of which 13 DNA sequences were included from a previously published report (26). Aligned deduced amino acid sequences of the complete gG-1 gene for 14 variants discovered in 38 clinical and reference strains of which 13 DNA sequences were included from a previously published report (26). The number of isolates corresponding to each genetic variant is given to the left. The proposed immunodominant region of gG-1 is shaded, and the epitope of the anti-gG-1 MAb reactivity is depicted in boldface type and underlined. FIG. 4. | Comparison of OD values in ELISA for sera drawn from patients harboring MAb-negative or MAb-positive isolates using a type-common HSV-antigen and an HSV-1 type-specific gG-1 antigen. Comparison of OD values in ELISA for sera drawn from patients harboring MAb-negative or MAb-positive isolates using a type-common HSV-antigen and an HSV-1 type-specific gG-1 antigen. Data are expressed as the mean of two observations. Each mean value for the two groups of sera (those drawn from patients with anti-gG-1 MAb+ or anti-gG-1MAb- strains) is indicated by a horizontal bar and did not differ significantly in either of the two serological assays when analyzed with Student's t test. DISCUSSION : To ensure a firm base for the use of gG-1 as a HSV-1 type-specific antigen, we here determined the genetic variability of the gG-1 gene in a large number of clinical isolates. A recently characterized immunodominant region of the protein overlapping the epitope of an HSV-1 type-specific anti-gG-1 MAb was therefore of special interest. Previously, we found several mutations within this region when the gG-1 gene was sequenced in a limited number of HSV-1 isolates , and reactivity to an anti-gG-1 MAb was early reported by Ackermann et al. to differ among type 1 strains . Here, we utilized a type-specific anti-gG-1-MAb, reactive with amino acids 110AFPL113, for screening of gG-1 variants among 108 clinical isolates. Altogether, 42 (38%) of the clinical isolates showed no reactivity to the anti-gG-1 MAb, and by DNA sequencing and phylogenetic comparison we observed two main genetic variants of the gG-1 gene that correlated to the absence or presence of MAb reactivity. Since these two disparate genetic variant sequences of gG-1 resulted in up to 14 amino acid alterations (see below) and some were localized to the immunodominant region, we suspected that the two variants might also differ in their induction of IgG antibody response to gG-1. If this was the case, one could expect different serological reactions between the two groups in the HSV-1 type-specific ELISA based on gG-1 antigen currently in use . This could be expressed as a low sensitivity in serotesting of patients harboring strains carrying one of the other gG-1 gene variants if the antigen used as a diagnostic tool is based on the sequence of the other genetic variant. A low sensitivity of gG-1-based ELISA kits in clinical settings has been described . To investigate this, we compared the serological reactions in sera from patients infected with isolates representing either variant of gG-1. However, no differences in IgG reactivity were found when sera were tested by the gG-1 ELISA. In addition, only 3 of the 10 sera that were found to be negative in type-specific gG-1 ELISA were drawn from patients harboring MAb-negative isolates. Instead, a possible explanation for the seronegativity to gG-1 could be that these sera were collected early during infection before seroconversion had occurred. In fact, most of the gG-1-seronegative patients showed symptoms and/or serological signs of primary HSV-1 infection. Hence, we suggest that early infection might be a factor contributing to the reported low sensitivity to gG-1 antigen-based ELISA. The finding that only two of the sera were reactive to the gG-2 antigen (lower than expected from seroepidemiological studies ) could be explained by the large proportion of sera derived from what was considered to be primary HSV-1 infections. Furthermore, despite the fact that most of the corresponding viral isolates (28 of 53) were derived from the genital tract, they were typed as HSV-1. A recent epidemiological report from our geographical area has shown that genital HSV-1 infections are increasing , thus explaining why a lower rate of HSV-2 seropositivity might be encountered in such a selected patient group. The clinical HSV-1 strains described herein were isolated from a number of different anatomical sites, such as the lip, ear, eye, other facial regions, genital or perianal regions, and the nervous system. Both genetic variants were proportionally isolated from females and males and were present at all of these different body regions with the possible exception of extralabial and facial manifestations, where none of five isolates were MAb negative. Thus, neither of the two variants could be associated with anatomical site or gender. The dichotomy of the gG-1 DNA sequences found among clinical isolates was also present when reference strains originating from United States and United Kingdom, commonly used for experimental work, were sequenced and included in the phylogenetic analysis. The anti-gG-1 MAb-negative genetic variant was represented by KOS321 as previously described , while both the Syn17+ and F strains were found to belong to the MAb-positive branch of the phylogenetic tree. As a consequence, 14 of 238 (similar6%) of the encoded amino acid residues within gG-1 might differ between reference strains selected as prototype viruses for diagnostic or functional studies. Thus, by using either of Syn 17+ or F together with KOS321 for studies of expression and function of the gG-1 gene, the genetic diversity among clinical isolates is better represented in the experimental setting . Documentation of genetic variability among clinical HSV isolates is limited and was hitherto performed mostly by RE analysis . In comparison, a huge amount of data is available at the nucleotide level for several RNA viruses such as HIV and hepatitis C virus . However, genetic variability in specific genes of HSV-1, such as the gB-1 gene and the gD-1 gene , after PCR amplification performed on clinical samples of cerebrospinal fluid has been reported but could not link any genetic variants to the clinical presentation in the form of encephalitis. As regards HSV-2 , sequence data, derived from clinical isolates, on genes encoding envelope glycoproteins such as gB, gC, gD or gG have been presented . Taken together, these studies showed the occurrence of point mutations, which were mostly silent and coded for insertion or deletion of one or a few amino acids. However, no correlation was found between clinical conditions and genetic variants. From the results of the sequencing of the gG-1 gene, we observed additional mechanisms for generating genetic variability among the isolates: several possible recombinants of the two genetic variants have been described here as well as a frameshift mutant that most likely resulted in a gG-1-negative phenotype. The latter strain was unreactive to the anti-gG-1 MAb but showed only a single insertion of a cytosine nucleotide within a stretch of six cytosines at the N-terminal part of the gG-1 gene, resulting in a different reading frame and a premature stop codon. A similar mechanism explaining a gG-2-negative phenotype in several HSV-2 isolates was also recently described from our laboratory . Sequence alignment suggested that five sequences might represent recombination of two segments of gG-1. Sequence 34 appeared to be a recombinant with a Syn17+-KOS321 pattern. Conversely, the sequences 66, 67, 70, and AF117121 showed a KOS321-Syn17+ pattern, apparently with the recombination point occurring at the same position or a nearby position. Phylogenetic analysis carried out for each segment of the gG-1 gene further supported recombination . In all, 4% (5 of 115) of the strains from clinical samples showed signs of recombination, indicating that this event is relatively frequent and also that these recombinants are viable and circulate in the population. Although recombinants for genes coding for glycoproteins among clinical HSV-1 isolates have not, to our knowledge, been described previously, homologous recombination of the gB gene was reported after coinfection of mice with two different HSV-1 strains . Furthermore, a study of CMV DNA in clinical samples has shown that homologous recombination of the gB gene has occurred and exists in several variants . In conclusion, the earlier-described variability of the gG-1 DNA sequence among clinical isolates and reference strains was here confirmed and further defined as a dichotomy in a large study of HSV-1 clinical isolates. Ongoing studies are investigating whether this form of genetic diversity is also present in other HSV-1 genes. Based on the hypothesis that HSV-1 coevolves with the human host, a DNA sequence comparison with strains isolated from subjects in distant parts of the world would be of interest. Furthermore, recent studies relating gG-1 to apical entry of polarized cells raise the question of whether the two gG-1 variants differ in their function when contributing to HSV-1 infectivity. Backmatter: PMID- 12202588 TI - Actinomyces cardiffensis sp. nov. from Human Clinical Sources AB - Eight strains of a previously undescribed catalase-negative Actinomyces-like bacterium were recovered from human clinical specimens. The morphological and biochemical characteristics of the isolates were consistent with their assignment to the genus Actinomyces, but they did not appear to correspond to any recognized species. 16S rRNA gene sequence analysis showed the organisms represent a hitherto unknown species within the genus Actinomyces related to, albeit distinct from, a group of species which includes Actinomyces turicensis and close relatives. Based on biochemical and molecular genetic evidence, it is proposed that the unknown isolates from human clinical sources be classified as a new species, Actinomyces cardiffensis sp. nov. The type strain of Actinomyces cardiffensis is CCUG 44997T. Keywords: Introduction : The genus Actinomyces embraces a heterogeneous group of anaerobic and aerotolerant, non-spore-forming, non-acid-fast, gram-positive rods with high G+C DNA content . Many Actinomyces species primarily belong to the facultative anaerobic indigenous microflora of human and animal mucous membranes, particularly those of the oral cavity. Some members of the genus are known to cause classical actinomycosis and may also be found in polymicrobial infections, while others are significant in periodontal disease . The genus Actinomyces has undergone considerable expansion in recent years, with a plethora of new species defined, especially from human sources. Recently described Actinomyces species, including Actinomyces funkei , Actinomyces europaeus , Actinomyces graevenitzii , Actinomyces neuii subsp. anitratus and neuii , Actinomyces radingae , Actinomyces radicidentis , Actinomyces turicensis , and Actinomyces urogenitalis , have been isolated from clinical specimens, where they occur as contaminants and/or represent possible opportunistic pathogens. Despite this increase in the number of recognized species, it is clear that knowledge of the habitats, clinical prevalence, and pathogenic potential of many Actinomyces and related organisms is inadequate, and there are indications that much new diversity remains to be discovered from human sources. During the course of study of clinical isolates of Actinomyces, we have characterized a hitherto unknown Actinomyces-like species. Based on both phenotypic and phylogenetic evidence, we propose yet another new species of the genus Actinomyces, namely, Actinomyces cardiffensis. MATERIALS AND METHODS : Cultures and phenotypic characterization. | All eight strains were isolated in the United Kingdom and were presumptively identified as Actinomyces spp. and were referred to the Public Health Laboratory Service Anaerobe Reference Unit, Cardiff, United Kingdom, for confirmation of identity. Strain R10394T (CCUG 44997T) was recovered in 1993 from an intrauterine contraceptive device (IUCD) which was in situ for 7 years in a 35-year-old female in Great Yarmouth, United Kingdom. Strains R13895 (CCUG 45110) and R7770 (CCUG 46084) were also recovered from IUCDs of 26- and 37-year-old females in Farnborough, Hampshire, United Kingdom (2000), and Bury St. Edmunds, West Suffolk, United Kingdom (1994), respectively. The latter patient was described as having the IUCD in situ for a considerable period of time, and Actinomyces-like organisms were seen in her cervical smear. Strain R5999 (CCUG 45108) was recovered from pus from temporal, large parietal, small parietal, and ear abscesses of a 32-year-old male at 4 weeks postmastoidectomy (Cardiff, United Kingdom, 1993). Concomitant organisms in each of the four abscesses were Fusobacterium nucleatum, Bacteroides fragilis, Bacteroides ureolyticus, Porphyromonas endodontalis, Porphyromonas levii, Prevotella oris, Peptostreptococcus micros, Peptostreptococcus sp., Arcanobacterium bernardiae, Eggerthella lenta, and a microaerophilic Streptococcus sp. strain R11374 (CCUG 45109) was recovered from pleural fluid of a 74-year-old male with glaucoma, shortness of breath, and wheezing (Burton upon Trent, East Staffordshire, United Kingdom, 1997). Strain R5571 (CCUG 44995) was recovered from pus from an actinomycotic jaw abscess of a 78-year-old female (Hereford, United Kingdom, 1993). Strain R2037 (CCUG 46083) was recovered from a pericolic abscess, discovered upon resectioning of a colon carcinoma of a 71-year-old female (Lincoln, Lincolnshire, United Kingdom, 1985). Strain R9463 (CCUG 46085) was recovered from a right antral washout of a 48-year-old female with sinusitis (Carlisle, Cumbria, United Kingdom, 1996). Volatile and nonvolatile end products of glucose metabolism were detected by gas-liquid chromatography . For biochemical testing, the strains were cultured on Columbia agar (Difco, Detroit, Mich.) supplemented with 5% horse blood at 37C, incubated anaerobically. The strains were biochemically characterized using both conventional tests and the commercially available API rapid ID32A, API rapid ID32Strep, and API Coryne systems according to the manufacturer's instructions (API BioMerieux, Marcy l'Etoile, France). Amplified 16S rDNA restriction analysis. | Amplified 16S rDNA restriction analyses were performed using HaeIII and HpaII as described previously . 16S rRNA gene sequencing and phylogenetic analyses. | The 16S rRNA genes of the eight isolates were amplified by PCR and directly sequenced using a Taq dye-deoxy terminator cycle sequencing kit (Applied Biosystems, Foster City, Calif.) and an automatic DNA sequencer (model 373A; Applied Biosystems). The closest known relatives of the new isolates were determined by performing database searches. These sequences and those of other known related strains were retrieved from the GenBank or Ribosomal Database Project libraries and aligned with the newly determined sequences using the program PILEUP . The resulting multiple sequence alignment was corrected manually, and a distance matrix was calculated using the programs PRETTY and DNADIST (using Kimura 2-correction) . A phylogenetic tree was constructed according to the neighbor-joining method with the program NEIGHBOR . The stability of the groupings was estimated by bootstrap analysis (500 replications) using the programs DNABOOT, DNADIST, NEIGHBOR, and CONSENSE . Protein profiling. | Polyacrylamide gel electrophoresis analysis of whole-cell proteins was performed as described by Pot et al. . For densitometric analysis, normalization, and interpretation of protein patterns, the GCW 3.0 software package (Applied Maths) was used. The similarity between all pairs of traces was expressed by the Pearson product moment correlation coefficient, which was converted for convenience to percent similarity. Nucleotide sequence accession number. | The GenBank accession number for the 16S rRNA sequence of T is . RESULTS AND DISCUSSION : The eight isolates consisted of slender, straight or curved, pleomorphic, gram-positive rod-shaped cells. Beaded and branched filamentous forms occurred. Cells were non-acid-fast, non-spore-forming, and catalase negative. Strains were facultatively anaerobic but grew better under anaerobic conditions. After anaerobic incubation for 48 h, colonies on Fastidious Anaerobe Agar (LabM; Bury, Lancashire, United Kingdom) with 5% horse blood were pinpoint, convex, smooth, entire-edged, opaque cream to pink, and nonhemolytic. Principal end products of glucose metabolism were small amounts of acetic acid and large amounts of lactic and succinic acids. When subjected to conventional biochemical testing , the isolates formed acid from glucose and sucrose but not from amygdalin, l-arabinose, cellobiose, lactose, mannitol, mannose, d-raffinose, salicin, trehalose, or d-xylose. The level of acid production from d-ribose was variable. The isolates failed to hydrolyze esculin and starch and were lipase and lecithinase negative. They were indole negative. Based upon these findings, the unknown isolates resembled Actinomyces odontolyticus but were differentiated from this species by the fermentation of lactose, dark red pigment, and growth in air, which are characteristic of the latter. Using the commercial biochemical kits, the strains were unidentified. When the API Coryne system was used, the eight strains produced acid from glucose and sucrose and displayed alpha-glucosidase activity. Acid was produced from ribose by five of the eight strains, while seven of the eight strains formed acid from maltose. Nitrate reduction was variable, with two of the eight strains testing positive. None of the isolates produced acid from glycogen, lactose, mannitol, or d-xylose, and activity for alkaline phosphatase, catalase, beta-galactosidase, beta-glucuronidase, N-acetyl-beta-glucosaminidase, pyrazinamidase, pyrolydonyl arylamidase, and urease was not detected. None of the strains hydrolyzed esculin or gelatin. A code of 0/1 0 1 0 1/3 2 1 was generated. When the API rapid 32S system was used, all eight strains showed activity for alanine phenylalanine proline arylamidase and three of the eight strains produced acid from sucrose. However, none of the strains produced acid from d-arabitol, l-arabinose, cyclodextrin, glycogen, maltose, mannitol, melibiose, melezitose, methyl-beta-d-glucopyranoside, lactose, pullulan, d-raffinose, d-ribose, sorbitol, tagatose, or trehalose and none displayed activity for arginine dihydrolase, alkaline phosphatase, beta-glucosidase, alpha-galactosidase, beta-galactosidase, beta-glucuronidase, glycyl tryptophan arylamidase, beta-mannosidase, N-acetyl-beta-glucosaminidase, pyroglutamic acid arylamidase, or urease, using the rapid ID 32S kit. When the API rapid ID32A system was used, all eight strains displayed activity for alanine arylamidase, arginine arylamidase, alpha-glucosidase, glycine arylamidase, histidine arylamidase, leucine arylamidase, leucylglycine arylamidase, proline arylamidase, phenylalanine arylamidase, serine arylamidase, and tyrosine arylamidase. None of the strains displayed activity for alkaline phosphatase, alpha-arabinosidase, arginine dihydrolase, alpha-fucosidase, alpha-galactosidase, beta-galactosidase, beta-galactosidase 6-phosphate, beta-glucosidase, beta-glucuronidase, glutamic acid decarboxylase, glutamylglutamic acid arylamidase, N-acetyl-beta-glucosaminidase, pyroglutamic acid, arylamidase, or urease. Some strains produced acid from mannose (one of eight) and raffinose (two of eight), and three of eight reduced nitrate. None of the isolates produced indole. In this test system, a code of 0 4 0 0/2/4 0/1 7 3 7 0 5 was generated. The cellular morphology and biochemical reactions of the isolates were consistent with their tentative assignment to the genus Actinomyces, but the strains did not appear to correspond to any recognized species of this genus. To investigate the genetic relatedness of the isolates, amplified 16S rDNA restriction analyses was performed. All eight strains produced identical 16S rDNA restriction profiles with HaeIII and HpaII (profile, 001/016), indicating that the strains were highly related to each other. To ascertain the phylogenetic relationships of the unknown organisms, their almost complete 16S rRNA gene sequences (>1,400 nucleotides) were determined. Pairwise analysis showed that the organisms were genetically highly related to each other, displaying >99.5% sequence relatedness. These data, in conjunction with the high phenotypic resemblance of the clinical isolates, strongly support their assignment to a single species. Sequence database searches confirmed that the unknown isolates were most closely related to species of the genus Actinomyces. The highest sequence similarity values were seen with Actinomyces species and phylogenetically related organisms. Other Actinomyces species showed substantially lower levels of relatedness (data not shown). Tree analysis further demonstrated the placement of the unidentified bacterium (as exemplified by strain R10394T = CCUG 44997T) within the genus Actinomyces, with the novel bacterium displaying a specific association to a small subcluster of species which included Actinomyces funkei, Actinomyces radingae, Actinomyces turicensis, and their close phylogenetic relatives . The results of comparative analysis of whole-cell protein profiles of the unidentified strains and reference species of the genus Actinomyces and related genera are shown in Fig. . The eight clinical isolates formed a distinct group (greater than 70% intragroup similarity) that was distinct from all other recognized species. Actinomyces funkei displayed the closest similarity to the unidentified clinical group, joining the latter at about 55% similarity . Other species displayed much lower levels of similarity. In recent years, numerous new Actinomyces species have been isolated from human and animal sources. Most of these new species have come to light as a result of increased interest by clinical microbiologists in the possible role of such organisms as opportunistic pathogens and due to the implementation of improved molecular identification methodologies that permit their differentiation from recognized species. It is evident from the results of the present investigation that the unidentified gram-positive, catalase-negative, rod-shaped organisms represent another hitherto undescribed Actinomyces species from clinical sources. The novel bacterium forms a distinct subline within a subcluster of species within the Actinomyces genus, which includes Actinomyces turicensis, Actinomyces funkei, Actinomyces radingae, Actinomyces canis, Actinomyces georgiae, Actinomyces hyovaginalis, Actinomyces meyeri, Actinomyces odontolyticus, and Actinomyces suimastitidis. Bootstrap resampling, however, showed that the novel bacterium did not possess a particularly significant affinity with any member of this subcluster. Sequence divergence values of 3 to 6% with the aforementioned species further reinforced the distinctiveness of the clinical isolates. Although there is no precise correlation between percent 16S rRNA sequence divergence values and species delineation, it is now universally recognized that organisms displaying values close to 3% or more do not belong to the same species . The observed >3% sequence divergence between the unidentified clinical isolates and all presently defined Actinomyces species is therefore consistent with separate species status. The separateness of the unknown bacterium is also supported by phenotypic evidence. Protein profiling showed that the clinical isolates were phenotypically closely related to each other and were quite distinct from other species within the A. turicensis-A. radingae rRNA subcluster. The biochemical profile of the novel bacterium also readily distinguishes it from all presently described Actinomyces species. Therefore, on the basis of the distinct phenotypic characteristics of the unidentified rod-shaped bacterium and molecular chemical and molecular genetic evidence, we think the clinical isolates warrant classification as a new species of the Actinomyces genus, for which the name Actinomyces cardiffensis sp. nov. is proposed. Tests that are useful in distinguishing Actinomyces cardiffensis from its closest relatives are shown in Table . We believe that the formal description of this new species, together with biochemical tests for its identification, will facilitate its recognition in the clinical laboratory, thereby permitting a future evaluation of its distribution, clinical prevalence, and significance. Description of Actinomyces cardiffensis sp. nov. | Actinomyces cardiffensis (car.dif.fen'sis, L. adj. cardiffensis, pertaining to Cardiff, a city in Wales) cells are pleomorphic, slender, straight-to-curved rods; beaded branching filaments occur. Cells stain gram positive, are non-acid-fast and nonmotile. On Fastidious Anaerobic Agar with 5% horse blood after 48 h of incubation, colonies are pinpoint, convex, smooth surfaced, entire-edged, and opaque cream to pinkish. Nonhemolytic. Facultatively anaerobic and catalase negative. Using conventional testing, acid is formed from glucose and sucrose but not from amygdalin, l-arabinose, cellobiose, lactose, mannitol, mannose, d-raffinose, salicin, trehalose, or d-xylose. Acid production from d-ribose variable. Principal end products of glucose metabolism are lactic and succinic acids, together with small amounts of acetic acid. Hydrolysis of esculin and starch is negative. Lipase and lecithinase are not produced. Using API systems, acid is produced from d-glucose but not from l-arabinose, d-arabitol, cyclodextrin, glycogen, lactose, mannitol, melibiose, melezitose, methyl-beta-d-glucopyranoside, N-acetyl-beta-glucosamine, pullulan, sorbitol, tagatose, trehalose, or d-xylose. Acid may or may not be formed from maltose, mannose, d-raffinose, d-ribose, and sucrose. Gelatin, esculin, and hippurate are not hydrolyzed. Alanine phenylalanine proline arylamidase, alanine arylamidase, arginine arylamidase, alpha-glucosidase, glycine arylamidase, histidine arylamidase, leucine arylamidase, leucylglycine arylamidase, proline arylamidase, phenylalanine arylamidase, serine arylamidase, and tyrosine arylamidase are detected. No activity is detected for alpha-arabinosidase, arginine dihydrolase, alkaline phosphatase, alpha-fucosidase, alpha-galactosidase, beta-galactosidase, beta-galactosidase 6-phosphate, beta-glucosidase, beta-glucuronidase, glycyl tryptophan arylamidase, glutamic acid decarboxylase, glutamylglutamic acid arylamidase, N-acetyl-beta-glucosaminidase, pyrazinamidase, pyroglutamic acid arylamidase, beta-mannosidase, pyrolydonyl arylamidase, or urease. Acetoin is not produced. Nitrate may or may not be reduced. Indole negative. The type strain is CCUG 44997T = CIP 107323T. Isolated from human clinical sources, including pleural fluid, brain, jaw, pericolic and ear abscesses, antrum, and IUCDs. Habitat is not known. FIG. 1. | Unrooted tree showing the phylogenetic relationships of Actinomyces cardiffensis sp. Unrooted tree showing the phylogenetic relationships of Actinomyces cardiffensis sp. nov. (boldface). The tree, constructed using the neighbor-joining method, was based on a comparison of approximately 1,327 nucleotides. Bootstrap values, expressed as a percentage of 500 replications, are given at branch points. FIG. 2. | Similarity dendrogram based on whole-cell protein patterns of Actinomyces cardiffensis sp. Similarity dendrogram based on whole-cell protein patterns of Actinomyces cardiffensis sp. nov. (boldface) and its nearest relatives. Levels of correlation are expressed as percentages of similarity for convenience. TABLE 1 | Results of tests used to distinguish Actinomyces cardiffensis sp. nov. from its nearest phylogenetic relatives Backmatter: PMID- 12202563 TI - Distribution of Environmentally Regulated Genes of Streptococcus suis Serotype 2 among S. suis Serotypes and Other Organisms AB - The occurrence of 36 environmentally regulated genes of Streptococcus suis strain 10 among all 35 S. suis serotypes was determined by using hybridization with the amplified genes as probes. In addition, the distribution of these genes among the virulence phenotypes of serotypes 1 and 2 was assessed. Hybridization was also performed with various other streptococcal species and nonstreptococcal bacterial species which may be present in pigs. Interestingly, probe ivs-25/iri-1, similar to agrA and sapR, hybridized only with S. suis serotype 1 and 2 strains with virulent phenotypes and is therefore suitable as a diagnostic parameter. Only one probe was specific for S. suis. This probe's sequence was identical to the epf gene, a putative virulence factor of S. suis. Probe ivs-31 was similar to a virulence factor of S. suis, namely, a gene encoding a fibronectin- and fibrinogen-binding protein. This probe hybridized only with oral streptococci. Nearly half of the probes (45%) hybridized with the oral streptococci (S. oralis, S. milleri, S. sanguis, S. gordonii, and S. mitis) and with Streptococcus pneumoniae. This indicates a close relationship between S. suis, the oral streptococci, and S. pneumoniae with respect to the selected environmentally regulated genes. One probe only hybridized with gram-negative species and therefore seems to be obtained by S. suis from a gram-negative organism by horizontal transfer. Keywords: Introduction : Streptococcus suis is a major problem in the swine industry. S. suis causes a wide variety of infections in young piglets, including septicemia, meningitis, polyarthritis, and endocarditis , and most often infected piglets do not survive. Occasionally, S. suis causes meningitis in humans . To undertake preventive health measures in the swine industry, it is important to recognize virulent bacteria and to distinguish such organisms from avirulent isolates from carriers. Adult pigs can carry S. suis in their noses and on their tonsils without exhibiting symptoms . From these adult carrier pigs, bacteria are transmitted to young pigs. Within pig herds, carrier rates of up to 100% have been described . Despite this high carrier rate, the prevalence of disease in such herds is less than 5% . The discrepancy between the prevalences of carriership and disease may be related to the differences in virulence between S. suis strains and serotypes . To date, 35 capsular serotypes have been described for S. suis , of which serotypes 1, 2, 7, 9, 14, and 1/2 are most frequently isolated from diseased pigs in Europe . Weakly virulent and avirulent phenotypes are found in serotype 2, and virulent and highly virulent phenotypes are found in serotype 1. These virulence phenotypes are differentiated by the expression of two proteins, muramidase-released protein (MRP) and extracellular factor (EF) . Recently, PCR tests were developed to detect serotypes 1, 2, 7, and 9 and to distinguish virulent serotype 2 strains from avirulent serotype 2 strains . For other serotypes, no reliable diagnostic methods are available. Although there is a strong correlation between the expression of MRP and EF and the virulence of S. suis serotype 1 and 2 strains, MRP- and EF-negative S. suis strains still cause disease . Therefore, other factors must contribute to their virulence. The main goal of this study was to examine the possibility of using environmentally regulated genes for the development of diagnostic tools to detect virulent strains. Since it is known that many important virulence factors are regulated and induced at specific stages of the infection process , we identified environmentally regulated genes of S. suis serotype 2 by using two different selection conditions . To do this, chromosomal DNA fragments of a pathogenic S. suis strain were cloned in a plasmid in front of a promoterless erythromycin resistance gene. The resulting plasmid library was introduced into a pathogenic S. suis strain. Subsequently, we selected S. suis clones in which erythromycin resistance had been induced under iron-restricted conditions and after infection of piglets with the library and treatment of the piglets with erythromycin. Infected animals developed specific signs of disease 3 to 8 days after infection. High numbers of bacteria were recovered from various tissues, and plasmid inserts were isolated from these recovered bacteria. Sequence analysis of these inserts revealed that a limited number of clones had been selected and that these clones were greatly enriched in the affected tissues . In total, 36 genes were identified as being environmentally regulated in S. suis serotype 2 ; among these genes, regulatory genes, virulence factors, and metabolic genes were found. We describe the presence and absence of the selected environmentally regulated genes in all 35 S. suis serotypes with various virulence phenotypes and in various other streptococci and bacterial species which may be present in pigs. Based on these data, we found one gene, homologous to a gene in the database that encodes Agr, that could discriminate virulent serotype 1 and 2 strains from avirulent serotype 1 and 2 strains. This gene is very suitable for the development of a diagnostic test. MATERIALS AND METHODS : Bacteria and growth conditions. | The bacterial isolates are listed in Tables and . Streptococci were grown in Todd-Hewitt broth (Biotrading, Mijdrecht, The Netherlands) and plated on Columbia agar blood base plates (Biotrading) containing % (vol/vol) horse blood. Actinobacillus pleuropneumoniae was grown in brain heart infusion broth (BHI; Biotrading) plus 0.05% NAD (Fluka, Buchs, Switzerland) and plated on Columbia agar blood base plates (Biotrading) containing 5% heated sheep blood and 0.05% NAD (Fluka). Haemophilus parasuis was grown in BHI (Biotrading) and plated on Columbia agar blood base plates (Biotrading) containing 5% heated sheep blood. All other bacterial species were grown in BHI (Biotrading) and plated on heart infusion agar plates (Biotrading) containing 5% sheep blood. Yersinia enterocolitica was grown at a temperature of 30C; all other species were grown at 37C. Chromosomal DNA isolation. | Chromosomal DNA of the various bacterial species was isolated from 50-ml samples of stationary-phase growing cells. Bacterial cultures were centrifuged at 2,500 x g for 10 min. The cell pellet was resuspended in 5 ml of lysis buffer (20 mM Tris-HCl [pH 7.4], 10 mM disodium EDTA, 50 mM NaCl) containing 10 mg of lysozyme (Roche, Mannheim, Germany) per ml (or 10 mug of lysostaphin [Sigma, St. Louis, Mo.] per ml for Staphylococcus aureus) at 37C for 10 min (or at 37C for 30 min for S. aureus). Subsequently, 250 mul of 10% sodium dodecyl sulfate (SDS) and 40 mul of proteinase K (Merck, Darmstadt, Germany; 20 mg/ml) were added and the mixture was incubated at 70C for 20 min (1 h for S. aureus). Suspensions containing the chromosomal DNA were extracted three to five times with equal volumes of phenol-chloroform-isoamyl ethanol (25:24:1) to remove proteins. Extracted DNA was incubated with 15 mul of DNase-free RNase A (Roche; 10 mg/ml) at 37C for 15 min. Subsequently, chromosomal DNA was precipitated with 0.1 volume of 3 M sodium acetate (pH 5.4) and 2.5 volumes of 96% ethanol and then washed with an equal volume of 70% ethanol. The chromosomal DNA was dissolved in ultrapure water. Radiolabeling of DNA probes. | Probes for ivs and iri genes (ivs/iri probes) were amplified by PCR as described by Smith et al. for the analysis of the genomic S. suis library in pIVS-E. GenBank accession numbers for the sequences of the probes are to for the iri gene probes and to for the ivs gene probes . The PCR products were purified by using the High Pure PCR product purification kit (Roche). Approximately 1 mug of DNA was radiolabeled with [alpha-32P]dCTP (3,000 Ci/mmol, 111 TBq/mmol; Amersham) by use of a random-primed labeling kit (Roche) as described by the manufacturers. Detection of ivs and iri genes by dot blotting. | One microgram of chromosomal DNA was spotted onto Genescreen Plus membranes. The membranes were incubated in 0.4 M NaOH-1 M NaCl at room temperature for 10 min to denature the DNA and incubated at room temperature for at least 10 min in 2x sodium chloride-sodium citrate (0.3 M NaCl, 0.03 M sodium citrate [pH 7.0]) for neutralization. The membranes were blocked in (pre)hybridization buffer (0.5 M sodium phosphate buffer [pH 7.2], 7% SDS, 1 mM EDTA) at 65C for at least 30 min. Subsequently, the 32P-radiolabeled probe was added to the membranes in 30 ml of hybridization buffer and the mixture was incubated overnight at 65C. The membranes were washed twice with washing buffer 1 (40 mM sodium phosphate buffer [pH 7.2], 5% SDS, 1 mM EDTA) at 65C for 30 min and twice with washing buffer 2 (40 mM NaPO4 [pH 7.2], 1% SDS, 1 mM EDTA) at 65C for 30 min. The signal was visualized with a STORM phosphorimager (Molecular Dynamics, Sunnyvale, Calif.). As a positive control, a probe specific for S. suis 16S ribosomal DNA (rDNA) was used . TABLE 1 | S. suis strains used TABLE 2 | Bacterial strains used TABLE 6 | Hybridization of ivs and iri genes in several gram-negative and -positive bacterial species RESULTS : Distribution of ivs and iri gene sequences among all known S. suis serotypes. | To examine the presence or absence of the selected ivs and iri genes in all 35 S. suis serotypes, we performed cross-hybridization experiments. The 16S rDNA positive control probe, as well as the probes for iri-7 (similar to rgpG), iri-8 (homologous to gln tRNA), ivs-1 (similar to a transposase gene), and ivs-29 (similar to a hypothetical gene), hybridized with all S. suis serotypes . None of the probes reacted serotype specifically. Most probes hybridized with DNA of most serotypes, except those for ivs-21 (epf gene), ivs-25 and iri-1 (similar to agrA and sapR), and ivs-8 (similar to a transposase gene). Serotypes 32 and 34 did not hybridize with 27 probes. Also, for serotypes 33, 20, 22, and 26, limited hybridization was seen . Probes ivs-21, ivs-25/iri-1, and ivs-8 hybridized with a limited number of serotypes . Distribution of ivs and iri genes among S. suis virulence phenotypes. | We previously showed that serotype 1 and 2 strains differ in virulence . Serotype 2 comprises virulent, weakly virulent, and avirulent strains; serotype 1 comprises highly virulent and virulent strains. The association of virulence with the occurrence of the various ivs and iri genes was studied. A 16S rDNA probe was used as a positive control. One probe, ivs-25/iri-1 (similar to agrA and sapR), discriminated between the virulence phenotypes of S. suis serotype 2 (data not shown). To confirm that there was a relationship between virulence and the presence of the ivs-25 and iri-1 genes, four strains of each virulence phenotype of serotype 1 and 2 were tested, where possible by using strains whose virulence had been tested in a pig model. The tests of four strains of each virulence phenotype invariably resulted in identical hybridization data. As shown in Table , all serotypes with virulence phenotypes hybridized with probe ivs-25/iri-1, while the avirulent serotype 2 strains did not hybridize. Therefore, probe ivs-25/iri-1 can be used to detect virulent S. suis strains among serotypes 1 and 2. Hybridization of ivs and iri genes with various streptococci and other bacterial species. | We next investigated the hybridization of the environmentally regulated genes with other bacterial species. The selected bacterial species from pigs included 17 streptococci besides S. suis, 6 additional gram-positive bacteria and 11 gram-negative bacteria. A 16S rDNA probe of S. suis serotype 2 containing the variable region as well as the conserved region that was used as a positive control showed a strongly positive signal in all cases. Probe iri-8 (homologous to gln tRNA) strongly hybridized with DNA of all streptococci and almost all other bacterial species (Tables and ). Apparently, iri-8 is highly conserved among various bacterial species. Probes for a number of genes, namely, ivs-21 (epf gene), ivs-16 (similar to atlR), iri-31 (cps2A gene), iri-23 (similar to yvyD), iri-11 (similar to nrdD), iri-32 (similar to ruvB), ivs-32 (similar to fliF), ivs-36 (similar to yqeG), and iri-13 (similar to MTCY336_33), did not hybridize to any of the other streptococci or to any of the other species tested, suggesting that these genes might be S. suis specific (Tables and ). However, except for ivs-21 (epf gene), these genes showed homology to sequences in the database of bacteria present in our assay. Therefore, similar genes are present in bacterial species other than S. suis, and therefore the genes are not specific for S. suis. This finding indicates that the homology between the probes used and the genes present in the bacterial species was too low to detect the genes in the hybridization assay used and under the conditions applied. Probe ivs-21 (epf gene) hybridized only to S. suis and showed no homology in the database. Therefore, the epf gene is the only S. suis-specific gene that we found by using the selection procedures described above. Two probes, ivs-23/iri-24 (similar to cpsY and oxyR) and iri-16 (similar to trmU), hybridized to most of the streptococcal species but did not hybridize to other bacterial species except for S. aureus. Apparently, ivs-23, iri-24, and iri-16 are conserved among the various streptococcal species. In contrast to other probes tested, ivs-2/iri-10 (similar to yoaE) hybridized to none of the streptococcal species except for S. suis but did hybridize with four gram-negative bacteria, Escherichia coli, Klebsiella pneumoniae, Salmonella enterica serovar Typhimurium, and Y. enterocolitica (Tables and ). TABLE 3 | Distribution of ivs and iri genes among 35 reference strains of S. suis serotypes TABLE 4 | Distribution of ivs-21, ivs-25, and iri-1 among different virulence phenotypes of S. suis serotypes 1 and 2 TABLE 5 | Hybridization of ivs and iri genes in several bacterial species DISCUSSION : In this study, we examined the distribution of 36 environmentally regulated genes of S. suis strain 10 among all 35 S. suis serotypes in order to improve detection of virulent S. suis strains. The probe ivs-25/iri-1 (similar to agrA and sapR) detected virulent and weakly virulent serotype 2 strains as well as virulent and highly virulent serotype 1 strains. Since this probe hybridized only with virulent strains, it is suitable as a diagnostic parameter to detect virulent S. suis strains among serotypes 1 and 2. ivs-25 was selected in vivo and iri-1 was induced under iron restriction . Both selected fragments are part of one gene in S. suis (unpublished results). This gene showed significant similarity to the sapR gene of S. mutans and to the agrA gene of S. aureus. In S. aureus, AgrA is a very strong regulator of several virulence factors . These data suggest that the agrA gene regulates transcription of genes that are important for virulence. The data showed that one probe, ivs-21, was S. suis specific. Hybridization was not found between this probe and any of the other bacterial species used, and no homology was found in the database. This probe sequence was identical to the epf gene, a putative virulence factor of S. suis encoding EF, which was present in 25 of 35 S. suis serotypes. The ivs-21 probe did not react with serotypes 1, 16, 20, 22, 26, 27, 31, 32, 33, and 34. The ivs-21 probe hybridized with strains of S. suis serotypes 1 and 2 with all MRP and EF phenotypes. Probe ivs-31 contained part of the fibronectin- and fibrinogen-binding protein gene of S. suis serotype 2. Its corresponding protein bound fibronectin and fibrinogen and was involved in the virulence of S. suis serotype 2 . This probe hybridized with four of the oral streptococci, S. oralis, S. sanguis, S. gordonii, and S. mitis. To date, the presence of such fibronectin- and fibrinogen-binding protein in S. oralis, S. sanguis, S. gordonii, and S. mitis has been unknown. ivs-8, similar to a transposase gene, was found to be present in a small number of serotypes, including serotypes 1, 2, 7, 14, and 1/2. Serotype 9, another serotype often isolated from diseased piglets, did not have this transposase gene. Transposases may be involved in the insertion into the genome of foreign DNA-containing genes that are involved in virulence . Therefore, it might be very interesting to determine the genes flanking this transposase gene. The 5' end of ivs-8 shows homology to capsular genes of S. pneumoniae. For S. suis, the capsule is also an important virulence factor . Probe ivs-9, similar to comE of B. subtilis, hybridized only to S. suis, S. pneumoniae, and to the oral streptococci, S. oralis, S. milleri, S. sanguis, S. gordonii, and S. mitis. Havarstein et al. described another competence gene that was present in the same streptococcal species. In B. subtilis, ComE is involved in competence development . For S. pneumoniae, it was shown that genes that are involved in competence are also involved in virulence . Although a natural transformation system has not been described for S. suis, the potential involvement of ivs-9 in competence and in the pathogenesis of S. suis infections needs to be further investigated. A number of genes, namely, ivs-21 (epf gene), ivs-16 (similar to atlR), iri-31 (cps2A gene), iri-23 (similar to yvyD), iri-11 (similar to nrdD), iri-32 (similar to ruvB), ivs-32 (similar to fliF), ivs-36 (similar to yqeG), and iri-13 (similar to MTCY336_33), hybridized only with S. suis DNA. Although these probes were S. suis specific, they showed similarity to sequences in the database of bacteria included in our assay. This means that the genes are not unique for S. suis but that the homology with other bacterial species was nonetheless too low to give a positive hybridization signal under the conditions used. Control hybridization experiments showed that fragments showing 75 to 80% homology on the DNA level will result in positive hybridization signals. Probes iri-3, iri-4, and iri-18 showed no similarity to DNA in the database, but all three probes hybridized with at least one other streptococcal species and therefore are also not S. suis specific. Probe iri-8, homologous to a gln tRNA gene, hybridized to almost all bacterial species tested, including tRNAs of many gram-negative organisms. tRNA genes are very conserved among bacterial species. Most probes hybridized with the majority of the S. suis serotypes, except with serotypes 20, 22, 26, 32, 33, and 34. Chatellier et al. and Brousseau et al. determined the sequences of the 16S rDNA cluster and the chaperonin 60 gene, respectively, of all reference strains of S. suis. They showed that serotypes 20, 22, 26, 32, 33, and 34 were the most divergent serotypes. Two other probes, ivs-23 and iri-16, hybridized with all 35 S. suis serotypes, with nearly all other streptococcal species, and with S. aureus. In the database, probe ivs-23 showed similarity to cpsY and oxyR of various streptococcal species. It is known that transcription regulators, such as cpsY and oxyR, are very conserved sequences. The other probe, iri-16, is similar to trmU, which encodes an RNA methyl transferase that is involved in the modification of nucleosides in bacterial tRNA. The function of this protein and its role in pathogenesis of infections are unknown. The fact that nearly all streptococci hybridized with the probe for iri-16 indicates that this gene is highly conserved among streptococci. Some probes showed hybridization only to S. suis and other streptococcal species. Apparently, such genes are highly conserved among streptococci. Interestingly, nearly half of the probes (45%) hybridized with the oral streptococci (S. oralis, S. milleri, S. sanguis, S. gordonii, and S. mitis) and S. pneumoniae. This indicates a close relationship between S. suis, the oral streptococci, and S. pneumoniae with respect to the selected environmentally regulated genes. Based on 16S rRNA sequencing, S. suis was most closely related to S. bovis and S. equinus . Probe ivs-2 (similar to yoaE) hybridized with four gram-negative species, E. coli, K. pneumoniae, S. enterica serovar Typhimurium, and Y. enterocolitica. This gene had a considerable higher a G+C content than did the other selected ivs and iri genes , suggesting that the yoaE gene of S. suis was obtained by horizontal transfer from E. coli or another gram-negative organism. In conclusion, the most promising candidate for improvement of S. suis diagnostics is probe ivs-25/iri-1, which encodes a putative general virulence regulator. It discriminates between virulent and avirulent serotype 1 and 2 strains. This probe can detect all virulent serotype 1 and 2 strains, while the present diagnostic methods are unable to detect virulent serotype 1 strains. Backmatter: PMID- 12202558 TI - Sensitivity of Three Urinary Antigen Tests Associated with Clinical Severity in a Large Outbreak of Legionnaires' Disease in The Netherlands AB - In 1999 an outbreak involving 188 patients with Legionnaires' disease (LD) occurred among visitors to a flower show in the Netherlands. Two enzyme immunoassays (Binax and Biotest) and one immunochromatographic assay (Binax NOW) were tested, using urine samples from LD patients from the 1999 outbreak. Sensitivity was calculated using positive culture and/or seroconversion as the "gold standard" in outbreak-related patients with radiographically confirmed pneumonia who fulfilled the epidemiological critera. The Binax EIA, Biotest EIA, and Binax NOW assay showed overall sensitivities of 69, 71, and 72%, respectively. When the tests were performed with concentrated urine samples, the overall sensitivities increased to 79, 74, and 81%, respectively. Using multiple logistic regression analysis with backward elimination, a statistically significant association was found between clinical severity and test sensitivity for all tests. For patients with mild LD, the test sensitivities ranged from 40 to 53%, whereas for patients with severe LD who needed immediate special medical care, the sensitivities reached 88 to 100%. These findings have major implications for the diagnostic process in patients with mild pneumonia and suggest that patients with mild pneumonia may go underdiagnosed if urine antigen tests alone are used. Keywords: Introduction : Legionnaires' disease (LD) is an acute pneumonia caused by Legionella, a rod-shaped gram-negative bacillus ubiquitous in (man-made) aquatic reservoirs. Currently 43 Legionella species and 65 serogroups have been described. In the United States, over 90% of Legionnaires' disease cases are caused by Legionella pneumophila, of which 70% of strains belong to serogroup type 1 . Legionella spp. are responsible for 1 to 5% of cases of community-acquired pneumonia (CAP) . Clinically and radiographically, LD cannot be distinguished from pneumonias caused by other microbial pathogens. Because of the high mortality rate in patients with LD requiring hospitalization, early diagnosis to enable adequate antimicrobial treatment is potentially life-saving. Diagnosis of LD in patients with symptomatic pneumonia is based on culture, serologic testing, or antigen detection in urine. Isolation of Legionella from respiratory secretions is not a very sensitive diagnostic test (25 to 75% sensitivity) and has the disadvantage of delay, because a positive result is not available until at least 3 days of incubation. Seroconversion is a diagnostic test with a high sensitivity and a high (serogroup-dependant) specificity, but it is of limited clinical value since it may take up to 9 weeks for patients to develop detectable antibodies . In contrast to the other tests mentioned above, urinary antigen tests combine reasonable sensitivity and high specificity with rapid results. The reported sensitivities of both enzyme immunoassay (EIA) and immunochromatographic test (ICT) show great variation: 50 to 90% . These variations may be explained by differences in patient characteristics, the serogroup with which the patient is infected, the timing of collection the urine sample in the course of illness, and whether the urine is concentrated before testing. To assess the value of the urinary antigen tests in a large outbreak situation , we used three widely used and commercially available tests with urine specimens from patients with outbreak-related LD: the Biotest EIA, the Binax EIA, and the Binax NOW test. MATERIALS AND METHODS : Patients. | In February 1999 an outbreak involving 188 cases of LD occurred in Bovenkarspel, The Netherlands. The outbreak investigation indicated that a whirlpool displayed at the consumer product division of an annual flower show was the most likely source of infection. Genotyping revealed that isolates from 27 patients were identical to one of the environmental L. pneumophila serogroup 1 strains . All 180 hospitalized patients with a confirmed Legionella pneumonia were included in this study after written consent was obtained from patients or their relatives. A confirmed case of LD ("gold standard") was defined as a patient who fulfilled the epidemiological criteria (visitor to the 1999 Bovenkarspel flowershow or member of the exhibition staff) and who suffered from symptoms compatible with pneumonia, who showed radiological signs of infiltration, and who showed laboratory evidence of infection with L. pneumophila. Laboratory evidence included (i) isolation of L. pneumophila from a respiratory sample cultured on buffered charcoal yeast extract supplemented with alpha-ketoglutarate followed by genotyping and subsequent comparison to the environmental strains of the Bovenkarspel outbreak; or (ii) a fourfold rise in the titer of immunoglobulin M (IgM) antibodies to L. pneumophila in paired acute-phase and convalescent-phase sera, with final titers of >=1:32 in accord with the 99% cutoff values found in a serosurvey of healthy volunteers representative of the Dutch population using a microagglutination IgM, serotype 1, antibody assay ; or (iii) seroconversion to positive IgM or IgG antibodies to L. pneumophila in paired acute-phase and convalescent-phase sera, with age-specific titers in accord with the 99% cutoff values found in a serosurvey among healthy volunteers representative of the Dutch population , using a commercial enzyme-linked immunosorbent assay to detect IgM and IgG serotype 1 to 7 antibodies (Serion ELISA; Institut Virion\Serion GmbH, Wurzburg, Germany) . For isolation and genotyping of L. pneumophila from sputum samples, the National Institute for Public Health and the Environment was the reference laboratory; for detection of antibodies against L. pneumophila in serum, the Regional Laboratory of Public Health Tilburg was the reference laboratory. Classification of severity of disease. | To investigate the relation between test sensitivity and severity of disease, the patients were divided into three clinical categories for CAP. Clinical data were collected from the hospital chart by using a standardized case record form. Severity of pneumonia was scored according to the minor criteria for severity of CAP advised by the American Thoracic Society , using the following clinical criteria at hospital admission: (i) respiratory frequency above 30 breaths per minute, (ii) PaO2 below 60 mm Hg or O2 saturation below 92%, (iii) bilateral or multilobar infiltration on chest X-ray, and (iv) systolic blood pressure below 90 mm Hg or diastolic blood pressure below 60 mm Hg. Patients with a radiographically proven unilateral unilobar pneumonia, but without signs or symptoms according to the above-mentioned criteria, were classified as CAP category 1 (mild pneumonia). CAP category 2 consisted of patients with a proven pneumonia who fulfilled only one of the criteria; according to our definition, these patients were suffering from a moderately severe pneumonia. CAP category 3 consisted of patients who presented with two or more criteria; these patients were considered to be suffering from severe pneumonia. Collection of urine samples. | All medical microbiologists who had assisted in the diagnosis and treatment of the pneumonia patients in to this outbreak were asked to send available urine samples from the patients to the Regional Laboratory of Public Health in Haarlem, The Netherlands. After collection, the urine samples were stored in portions at -70C. All but eight available urine samples had been collected during the hospital stay. In four cases, the urine samples had been obtained shortly before admission, and in four cases, they had been obtained after discharge from hospital. Urinary antigen tests. | The presence of L. pneumophila antigens in urine samples was investigated by using the Binax (Portland, Maine) and Biotest (Biotest AG, Dreieich, Germany) Legionella urinary antigen tests, both EIAs, and with the Binax NOW test, a qualitative ICT. All tests were used as specified by the manufacturers. However, to ensure maximum specificity for the ICT , samples giving positive tests were reexamined after 60 min. Urine samples were tested nonconcentrated and, to enhance the intensity of the reaction, after concentration by selective ultrafiltration (Minicon B15; Millipore Corp. Bed Ford, Mass.). This selective ultrafiltration system consists of a permeable membrane that permits the passage of water and substances with molecular weights less than 15,000. Statistics. | Statistical analysis was performed with the statistical program SPSS version 10.0 (Statistical Product and Service Solutions, Chicago, Ill.). Univariate analysis (chi-square test for dichotomous and ordinal variables; independent t test for discrete variables) was used to calculate the association with positive urinary antigen test results for the following variables: age, gender, clinical severity, number of days between onset of symptoms and collection of first urine sample, and number of times a urine sample was collected. Variables that were (borderline) significant were entered in a multiple logistic regression model. Using backward elimination, independent predictors for test positivity were established. Variables were retained in the model if the likelihood ratio test was significant (P < 0.1). RESULTS : Patient selection and classification of disease severity. | In the 1999 outbreak, 188 LD patients were diagnosed, of whom 133 fulfilled the criteria for a confirmed case. Of the patients with confirmed cases, 132 had been hospitalized and were enrolled in this study. A large number of patients in this outbreak (51 patients) were diagnosed by a urinary antigen test alone, leaving 81 patients for evaluation. For 58 (72%) of these 81 patients, urine samples were available. The microbiological diagnosis in these cases was established by culture alone in 11 cases, by culture and serologic testing in 14 cases, and by serologic testing alone in 33 cases. For 55 of the 58 confirmed cases with available urine samples, data on clinical severity could be collected. Women were overrepresented in CAP category 1 (67%), in contrast to CAP categories 2 (25%) and 3 (39%). The median age for women was lower then for men in categories 1 and 2 (64 and 73 years and 57 and 63 years, respectively), in contrast to category 3 (70 and 62 years, respectively). Mean age did not differ significantly between CAP categories, but the age difference between male and female subgroups in the lowest CAP category was significant (independent t test; P = 0.04). All patients who were classified in CAP category 3 needed medical attention in a specialized unit. Nonconcentrated urine samples. | Using nonconcentrated urine, the sensitivities calculated for the three tests were 71, 69, and 72% for the Biotest EIA, Binax EIA, and Binax NOW test, respectively. The differences between the test sensitivities were not significant. When the LD patients were divided in three groups according to their CAP category, it appeared that the average sensitivities for the three urinary antigen tests increased from 42% for patients in the lowest CAP category to 75% for those in CAP category 2 and 93% for those in CAP category 3. The individual sensitivities of the three tests when used with nonconcentrated urine samples are shown in Fig. . Table shows variables which in univariate analysis were significantly associated with test positivity. Age was not associated with test positivity. Multiple logistic regression analysis showed that the CAP category remained associated with test positivity in all tests used . Separate analysis for male and female patients gave identical results, indicating that the association was not modified by gender. Concentrated urine samples. | After concentration of the urine samples for all three tests, a clear but not statistically significant increase in sensitivities was found: to 74, 79, and 81% for the Biotest EIA, Binax EIA, and Binax NOW test, respectively. This increase in sensitivity was small in the Biotest EIA (3%) and more prominent in the Binax EIA (10%) and Binax NOW assay (9%). When test sensitivity results were compared for categories with increasing clinical severity, concentration of urine samples yielded higher sensitivities predominantly for patients in CAP categories 1 and 2 . CAP category, male gender, number of urine samples, and shorter period between onset of symptoms and collection of the first urine sample were associated with a positive test result, but age was not (individual test results are given in Table ). Multiple logistic regression analysis showed that the CAP category was the only factor associated with test positivity in all tests used (see Table ). Separate analysis for male and female patients gave identical results. FIG. 1. | Sensitivity for three urinary L. pneumophila Sensitivity for three urinary L. pneumophila antigen tests in different categories of clinical severity, using concentrated and unconcentrated urine. TABLE 1 | Odds ratios and mean differences with confidence interval for variables associated with urinary L. pneumophila antigen test positivity TABLE 2 | Multiple logistic regression models showing odds ratios with confidence interval for variables associated with urinary L. pneumophila antigen test positivity DISCUSSION : The first urinary antigen tests, based on an ELISA, were described in 1979 . Since then, numerous publications have followed that confirmed the value of urinary antigen detection for the diagnosis of Legionnaires' disease, regardless of the technical configuration of the test (, , , , -). Based on prospective and retrospective studies using data from solitary cases, moderate to high urinary antigen test sensitivities have been described. All reported test sensitivities are based on studies using sporadic LD cases. Most of these studies used a selection of patients or were retrospective ; they are thereby subject to selection bias. Some of them include patients with LD caused by other serogroups than serogroup 1, which leads to underestimation of the test sensitivity. Furthermore, the clinical conditions of the patients described in these studies were not taken into account, which may explain the range of sensitivity values found by different authors. To our knowledge, there are no publications on test sensitivities in an outbreak situation. The 1999 outbreak in The Netherlands provided a unique opportunity to evaluate urinary antigen tests in an outbreak caused by an identified L. pneumophila serogroup 1 strain. A nationwide alert for LD cases ensured optimal case finding, thereby decreasing patient selection bias. Furthermore, the conditions for a gold standard were favorable: two national reference laboratories performed all microbiological tests, and Dutch reference titers were calculated using the distribution of antibodies against L. pneumophila serogroup 1 in a large sample from a national serum bank . In addition, all available clinical data for the LD patients in this outbreak were recorded centrally. Assuming that patient inclusion in an outbreak situation resembles a prospective study design, our data are best compared with the published results of two prospective studies on urinary antigen test sensitivity in LD. One study included a single high titer in the gold standard, inherently lowering the test sensitivity to be measured on the basis of misclassification. In this study, however, a higher sensitivity (84%) was found for patients with LD caused by L. pneumophila serogroup 1. The other prospective study included patients with LD caused by L. pneumophila serogroups 1, 3, and 7, making a lower sensitivity more likely on the basis of a low urine antigen detection capacity for other serogroups than serogroup 1. Indeed, the reported sensitivity in this study was lower (57%). Since the 1999 outbreak in The Netherlands was caused by an L. pneumophila strain of serogroup 1, high sensitivities for the three tests were to be expected. However, overall test sensitivities found in this study were lower than those reported by Ruf et al. and ranged from 69 to 72% for nonconcentrated urine samples. This lower sensitivity may be explained by a difference in the study populations. Active case finding for LD in this outbreak may have resulted in hospitalization of a higher proportion of patients with relatively mild LD compared to a nonoutbreak situation. As our results show, the sensitivity of urinary antigen tests is relatively low for cases in CAP categories 1 and 2. We do not know of a published study in which an association between the severity of disease and the test sensitivity for LD has been demonstrated, although one study hinted at such an association . The association between sensitivity and clinical severity demonstrated in the present study has clinical and diagnostic consequences. Because of the high sensitivity in patients with a severe pneumonia, the early recognition of patients with life-threatening LD can prevent delay in initiating adequate antibiotic therapy. However, the urinary antigen test is less reliable in milder cases of LD, indicating that this diagnostic test, despite its rapid interpretation, cannot replace culture and serologic testing. Therefore, in the setting of persistent clinical or epidemiological suspicion of LD and a negative urinary antigen test result in patients in whom no other microorganism is identified, culture and serologic testing are recommended and treatment must include antibiotic coverage for Legionella. When the detection of antigens in urine during an outbreak is used for epidemiological purposes, one has to keep in mind that 50 to 60% of cases of mild pneumonia will stay undiscovered, depending on whether urine is concentrated (concentration improves the chance of detection). This implies that, due to undiagnosed cases, the size of an outbreak will be underestimated unless complementary diagnostic serologic tests using paired sera are performed in all suspected cases. Because seroconversion can take up to 9 weeks after onset of the disease, a prolonged interval between collection of the two sera is advisable for reliable interpretation of serologic results. Like other researchers , we were unable to demonstrate a significant difference in sensitivity between the two EIAs that were tested and the ICT. Since the latter is very easy to perform without special laboratory equipment and the results are available at short notice even after concentration of the urine samples, this test may be preferable in outbreak situations if serogroup 1 is involved. Previous studies have also demonstrated that a higher sensitivity of urinary antigen detection was found using concentrated urine, regardless of the test used. Concentration by ultrafiltration is easy to perform and can facilitate an early diagnosis, especially in milder cases. In conclusion, in outbreak situations the urinary antigen tests are a useful tool for early diagnosis of LD, especially in patients with severe cases. The ICT scored at least equal to the EIAs and has the advantage of ease of performance combined with rapid test results. Concentration of the urine samples increases the sensitivity, particularly in patients with less severe illness, and is therefore recommended. In outbreak situations, the use of urinary antigen tests alone for evaluation of the incidence rate will lead to underestimation of the actual incidence. Therefore, culture and serologic testing remain necessary diagnostic tools. TABLE 3 | Overview of urinary antigen test sensitivity for sporadic cases of LD Backmatter: PMID- 12202587 TI - Biological and Biochemical Characterization of Sheep Scrapie in Japan AB - Due to the apparent absence of an agent-specific nucleic acid genome, scrapie strains cannot be classified by genome characterization, which is commonly used for the classification of many viruses. However, scrapie strains can be distinguished to some extent by biological properties such as transmissibility to experimental animals and distribution of neuropathological lesions and by biochemical properties such as the molecular mass and relative protease-resistance of the disease-specific isoform of prion protein (PrPSc). In order to preliminarily characterize the scrapie strains that are prevalent in Japan, we analyzed the transmissibility of sheep scrapie isolates to mice and the relative proteinase K (PK) resistance of the corresponding PrPSc. The results indicate that Japanese scrapie strains can be divided into at least three groups based on biological and biochemical properties. The first group includes isolates which cause disease in mice with an incubation period of similar400 days and possess PrPSc with relatively high PK resistance. Isolates of the second group contain PrPSc that is highly resistant to PK digestion but transmit poorly to mice. The final group consists of isolates that cause disease in mice with an incubation period of less than 300 days and are associated with PrPSc with reduced PK resistance. Sheep scrapie has occurred sporadically in Japan since1982, with only similar60 officially reported cases so far. However, the diversity of scrapie strains in the field suggested by our data raises the concern that a scrapie strain similar to the parental agent of bovine spongiform encephalopathy could exist or emerge in Japan. Thus, continuous surveillance for scrapie will be required to prevent the further spread of scrapie, not only among the sheep population but also to other species, and to eliminate any potential risk of sheep scrapie to public health. Keywords: Introduction : Transmissible spongiform encephalopathies (TSEs) are fatal neurodegenerative diseases, which include scrapie in sheep and goats, bovine spongiform encephalopathy (BSE), and Creutzfeldt-Jakob disease (CJD) in humans. Scrapie has existed for more than two centuries, while BSE was first recognized in 1985, followed by a BSE epidemic in the United Kingdom . Epidemiological studies suggested that BSE was primarily caused by feeding meat and bone meal (MBM) contaminated with scrapie agent to cattle . Once BSE appeared, the causative agent spread through the cattle population by the use of BSE agent-contaminated MBM. The appearance of feline spongiform encephalopathy (FSE) in domestic and captive cats and, more recently, variant CJD (vCJD) in humans in 1996 , has raised a global concern for the spread of the BSE agent to other species via the food chain. It is known that there are biologically distinguishable sheep scrapie strains in the United Kingdom ; however, BSE isolates studied so far appear to have comparatively uniform characteristics . The biological and biochemical properties of the BSE agent are very similar to those of the agents of vCJD and FSE, providing strong evidence that cross-species transmission of BSE to humans and felids resulted in vCJD and FSE, respectively . Thus, it is now probable that a particular scrapie strain, which possessed enhanced resistance to heat inactivation, survived the rendering process and was transmitted to cattle via contaminated MBM. Thereafter, the agent passed into humans, possibly via the food chain. Alternatively, a particular strain capable of infecting both cattle and humans might have been selected and amplified during the transmission of the scrapie agent to cattle. One of the characteristics of TSEs is an accumulation of a protease-resistant, abnormal isoform of a host protein, PrPSc, which is specifically detected in the central nervous system and lymphoid tissues of scrapie-affected animals. PrPSc is posttranslationally generated from the host-encoded sialoglycoprotein, prion protein (PrPC). PrPC and PrPSc have the same primary structure but different conformations as detected by analysis of biophysical properties and biochemical properties, such as resistance to protease digestion and solubility in nonionic detergent . The presence of PrPSc usually correlates with scrapie infectivity. Thus, PrPSc is thought to be one of the major components of the scrapie agent, and so detection of PrPSc is often considered an indication of the presence of infectivity. Although the exact nature of the scrapie agent is still controversial, the failure to find an agent-specific genome to date prevents the use of nucleic acid sequencing for strain characterization, a method commonly used for strain typing of viruses and bacteria . However, TSE agents can be distinguished to some extent by incubation periods and distribution of neuropathological lesions on transmission to experimental animals , as well as biochemical properties, such as relative protease resistance and/or molecular mass of PrPSc and differences in the ratio of glycosylated PrPSc bands . There is a sporadic occurrence of scrapie in Japan, and the existence of BSE was disclosed in September 2001. Since BSE is thought to originate from sheep scrapie, surveillance of scrapie-positive sheep and characterization of prevalent scrapie strains in the field are required for the prediction and elimination of a potential risk of scrapie to public health. In order to attempt to characterize scrapie strains present in Japan, we analyzed several isolates for their transmissibility to mice and the biochemical properties of the associated PrPSc. The results showed that at least three different strains of scrapie agent exist in Japan. MATERIALS AND METHODS : Sheep with scrapie. | Eight naturally occurring sheep scrapie isolates (KH2, KU, SB, Y2, Y5, S1, S2, and S3) collected from 1987 to 1996 and three first-passage isolates from experimental sheep scrapie infections (A1, B3, and G1) were used in this study. As a negative control, one sheep (S4), which was defined as negative for scrapie by the absence of both neuronal vacuolation in histopathological examination and detectable PrPSc in central nervous system and lymphoid tissues in immunoblot analysis, was also used. Sheep S1, S2, S3, and S4 were kept on ranch S. Sheep S1, S2, and S3 had been born from two ewes which showed neurological dysfunction several months after delivery and were diagnosed with scrapie by clinical and/or histopathological methods. The sheep S1, S2, and S3 shared the same sire. However, there were no disease-associated mutations in the PrP genes of these sheep, and thus the sheep S1, S2, and S3 were tentatively grouped as cases of endemic scrapie. Sheep KH2, KU, and SB, each from independent ranches, and sheep Y2 and Y5 from ranch Y, were grouped as sporadic cases because there is no immediate blood relationship among these sheep. Sheep A1 and B3 with experimental scrapie received the same brain homogenate of a scrapie-affected sheep that was not included in the natural scrapie group used here, and the results of experimental transmission of this sheep scrapie isolate to mice were reported elsewhere . Sheep G1 was inoculated intravenously with a 10% brain homogenate of scrapie-affected sheep Y4 from the Y ranch. Sheep Y2, Y4, and Y5 were grown on the same farm and developed scrapie within 6 months, but they were born to different ewes. PrP genotyping of sheep. | PrP genotyping of sheep was performed as described previously . Amino acid polymorphisms at codons 112 Met/Thr, 136 Ala/Val, 154 Arg/His, and 171 Gln/Arg/His were basically used for distinction of the PrP genotype. Bioassay. | Twenty microliters of 10% brain homogenates (in phosphate-buffered saline) from scrapie-affected or scrapie-negative sheep were inoculated intracerebrally into 4-week-old female ICR mice (PrP allotype PrPA/A; PrPA encodes PrP with codons 108 Leu and 189 Thr). In some cases I/LnJ mice (PrP allotype PrPB/B; PrPB encodes PrP with codons 108 Phe and 189 Val) were also used. When mice showed clinical symptoms of the terminal stage of scrapie, mice were sacrificed under anesthesia and brains were removed and processed for the detection of PrPSc. The brains of mice that died of unknown causes were also checked for the presence of PrPSc. Sample preparation and PK digestion. | Preparation of brain samples for proteinase K (PK) digestion was carried out as described elsewhere with slight modifications . Brains were homogenized with 7 volumes (wt/vol) of 10 mM Tris-HCl (pH 7.5) and 7.5 mM MgCl2, and the homogenates were incubated at 37C for 1 h with DNase I (40 mug/100 mg tissue). After adding 20% Sarkosyl to a final concentration of 5%, the homogenates were kept at room temperature (RT) for 30 min and then centrifuged at 12,000 x g at RT for 5 min. The resulting supernatants were transferred to new tubes, and then solid NaCl was added to give a final concentration of 10%. After a 16-h incubation at 4C with continuous rotation, the homogenates were centrifuged at 16,000 x g, 4C and, for 40 min, and the resulting pellet was resuspended in 10 mM Tris-HCl (pH 7.5). This suspension was subjected to PK digestion at various PK concentrations and times as indicated in each experiment. Phenylmethylsulfonyl fluoride (PMSF) was added to stop the digestion (final conc. 1 mM), and then the reaction mixture was adjusted to 4% sodium dodecyl sulfate (SDS) and boiled for 5 min. The proteins were precipitated with 10 volumes of ice-cold methanol, and the final pellet was dissolved by SDS-polyacrylamide gel electrophoresis in sample buffer (4% SDS, 5% 2-mercaptoethanol, 5% glycerol, 0.01% bromophenol blue, 62.5 mM Tris-HCl [pH 6.8]). Detection of PrPSc. (i) Immunoblot analysis. | Immunoblot analysis was carried out as described previously with B-103 rabbit serum . Densitometric analysis of X-ray film was performed with a Lane & Spot Analyzer (Atto, Tokyo, Japan). (ii) Dot blot analysis. | The 10% brain homogenates of sheep with scrapie were mixed with an equal volume of cold lysis buffer [0.5% Nonidet P-40, 0.5% sodium deoxycholate, 10 mM EDTA, 100 mM NaCl, 50 mM Tris-HCl (pH 7.4)], diluted serially in twofold steps and blotted onto a nitrocellulose membrane. The membrane was dried and then washed with Tris-buffered saline containing 0.1% Tween 20 (TBST). The membrane was treated with PK (25 mug/ml) in TBST at 37C for 1 h, and the reaction was stopped by adding PMSF to a final concentration of 2 mM. After washing with TBST, the membrane was treated with 3 M guanidinium isothiocyanate for 20 min at RT and washed with TBST again. Immunostaining of the membrane was then performed as described for the immunoblot analysis. RESULTS : Sheep with scrapie. | Characteristics of the sheep used in this study are summarized in Table . The major symptoms of all scrapie cases used here were ataxia and, at the terminal stage, astasia. No sheep except for KU showed obvious loss of fleece, which is thought to be caused by pruritus. The sheep S1, S2, and S3, which were tentatively grouped as cases of endemic scrapie (see Materials and Methods), differed from other sheep with scrapie. They showed hypersensitivity at the onset of disease and died of scrapie at a considerably young age (average, 16.3 months old) compared to sheep with scrapie grouped as sporadic cases here (average, 42.0 months old) or those described by others (2 to 6 years old ). Transmissibility of sheep scrapie to mice. | Transmission of 11 sheep scrapie samples to mice is shown in Table . Eight of 11 successfully transmitted to ICR mice (PrP allotype, PrPA/A), with average incubation periods from 229 to 451 days postinfection (dpi). These eight sheep scrapie samples seem to be divided into two groups based on the incubation periods; one includes A1 and B3, which transmitted to mice with incubation periods of around 230 dpi, and the other includes KH2, KU, SB, Y2, Y5, and G1, which caused disease in mice with incubation periods of similar386 to 451 dpi. There is no significant difference among the incubation periods of KH2-, KU-, SB-, Y2-, and Y5-inoculated mice (P > 0.05 in t test). However, the incubation period of G1-inoculated mice was somewhat shorter than that of the others, with a statistically significant difference between the incubation period of G1-inoculated mice and that of KH2-, SB-, and Y2-inoculated mice (P < 0.05 in t test). The sheep A1 and B3 received the same brain homogenate from a naturally occurring sheep scrapie isolate by intravenous injection. The transmission of this parental isolate of A1 and B3 directly to mice was achieved within 271 to 307 dpi , consistent with the incubation periods observed in A1- and B3-inoculated mice. Thus, the first-pass scrapie samples A1 and B3 and their parental isolate all possessed the ability to transmit to mice with relatively short incubation periods. In contrast, the S1, S2, and S3 scrapie isolates were virtually nontransmissible to ICR mice . One mouse inoculated with the brain homogenate of sheep S1 died at 359 dpi without typical symptoms of scrapie but was positive for PrPSc in the brain. We cannot confirm whether this one case is due to actual transmission or contamination with a mouse-adapted scrapie agent. Several ICR mice inoculated with the brain homogenates of S2 or S3 sheep died without typical symptoms of scrapie during the observation period but were also negative for PrPSc. Therefore, it is obvious that S1, S2, and S3 differ from the other scrapie isolates used here. Furthermore, isolate S3 did not transmit to I/LnJ mice (PrP allotype, PrPB/B), whereas transmission of isolate KH2 to these mice was confirmed by the detection of PrPSc at 275 dpi. The sheep S1, S2, and S3 were born on the same farm and in the same parturient season from two ewes that developed scrapie several months after delivery, and thus the sheep were potentially infected with the same agent. One explanation for the failure of the transmission of these isolates to mice is that the brain homogenates may possess a relatively low degree of infectivity. However, dot blot analysis of serial dilutions of brain homogenates revealed that variations in the amount of PrPSc in the homogenates appear to be less than fourfold among the samples tested , suggesting that the lack of transmission was not due to reduced degrees of infectivity. Therefore, we conclude that the sheep scrapie isolates S1, S2, and S3 have an extremely low, if any, transmissibility to mice. PK resistance of sheep PrPSc. | Since some TSE strains are known to differ in the relative PK resistance of PrPSc , we analyzed the PK resistance of PrPSc in the brains of sheep with scrapie in order to identify any biochemical differences among sheep scrapie isolates. Partially purified PrPSc-containing fractions without PK treatment were first analyzed by immunoblot, and following densitometric analysis to allow normalization of the amount of PrPSc among the samples, the fractions were digested with PK at 0, 8, 40, 200, and 1,000 mug/ml for 2 h . The bands of PrPSc were clearly detected after PK treatment at 40 mug/ml in KH2 and S2, while PrPSc in A1 was undetectable under the same digest conditions. To confirm this difference, the same fractions were treated with PK at 40 mug/ml for various time periods . Consistent with the result in Fig. , the PrPSc bands in the A1 fraction were virtually eliminated by a 2-h treatment with 40-mug/ml of PK, while the PrPSc bands in the KH2 and S2 fractions were readily detected after the same treatment. The PrPSc of B3 exhibited the same PK resistance as A1, while KU, SB, Y2, Y5, S1, and S3 showed levels of PK resistance similar to KH2 and S2 (data not shown). Thus, the sheep scrapie isolates used here appeared to be divided into two groups based on the relative PK resistance of PrPSc. One possesses PrPSc that is highly resistant to PK digestion, as defined by resistance to treatment with PK at 40 mug/ml for 2 h, and the other possesses PrPSc of relatively low PK resistance and is completely degraded under these conditions. Together, the mouse transmission and PrPSc PK resistance data can be used to further divide the sheep scrapie isolates into three groups . The first group includes KH2, KU, SB, Y2, Y5, and G1, which are transmissible to ICR mice with incubation periods of similar400 days and contain PrPSc with relatively high PK resistance. The second group includes S1, S2, and S3, which transmit poorly to mice but possess PrPSc with a relative PK resistance similar to the first group. The last group includes A1, B3, and probably their parental isolate, which are transmissible to ICR mice with incubation periods of similar230 days and possess PrPSc with relatively low PK resistance. PK resistance of PrPSc in mice inoculated with sheep scrapie. | Sheep scrapie isolates which transmitted to ICR mice segregated into two groups based on incubation periods , and interestingly, the grouping of sheep scrapie isolates by relative PK resistance of PrPSc in sheep brains appears to coincide with the grouping by incubation periods . To address the question of whether the phenotype of relative PK resistance of PrPSc in sheep brain is conserved in infected mice, the relative PK resistance of PrPSc in the mouse brains was also examined . PrPSc in the brain of a mouse inoculated with KH2 was resistant to treatment with PK at 200 mug/ml for 1 h, while PrPSc in the brain of a mouse inoculated with A1 was sensitive to this condition . This difference in the relative PK resistance was also confirmed by the experiments in Fig. , in which the samples were treated with PK at 200 mug/ml for the various time periods indicated. PrPSc in the brains of mice inoculated with KU and Y5 showed the same PK resistance as that of a KH2-inoculated mouse, whereas PrPSc in the brain of a mouse inoculated with B3 showed the same PK resistance as that of an A1-inoculated mouse (data not shown). Therefore, the relative PK resistance of PrPSc generated in the brains of mice was similar to that of PrPSc in the inoculum of the corresponding sheep. FIG. 1. | Dot blot analysis of sheep brain homogenates for the presence of PrPSc. Dot blot analysis of sheep brain homogenates for the presence of PrPSc. Serial twofold dilutions of sheep brain homogenates used for inoculation into mice were dotted onto a nitrocellulose membrane and stained with B-103 anti-PrP synthetic peptide rabbit serum and ECL Western blot detection reagent (Amersham). The sheep are indicated at the bottom and dilutions are on the right. In this dot blot analysis, S1 appears to be negative for PrPSc; however, PrPSc was detected in the immunoblot analysis (data not shown). FIG. 2. | Relative PK resistance of sheep PrPSc. Relative PK resistance of sheep PrPSc. (a) PrPSc-containing fractions prepared from sheep brains were treated with various concentrations of PK (0 to 1,000 mug/ml, indicated above the photo) at 37C for 2 h. Molecular mass markers are indicated. PrP was detected by immunoblotting with PrP-specific antibody and ECL. (b) The same PrPSc-containing fractions as used in panel a were treated with PK (200 mug/ml) at 37C for various times (0 to 16 h, indicated above the photo). FIG. 3. | Relative PK resistance of PrPSc generated in mouse brain. Relative PK resistance of PrPSc generated in mouse brain. (a) PrPSc-containing fractions prepared from mouse brains were treated with various concentrations of PK (0 to 1,000 mug/ml, indicated above the photo) at 37C for 2 h. Molecular mass markers are indicated. PrP was detected by immunoblotting as in Fig. . (b) The same PrPSc-containing fractions as used in panel a were treated with PK (200 mug/ml) at 37C for various times (0 to 16 h, indicated above the photo). TABLE 1 | Sheep used in this study TABLE 2 | Transmissibility of sheep scrapie to mice TABLE 3 | Grouping of sheep scrapie samples based on transmissibility to mice and relative PK resistance of PrPSc DISCUSSION : We were interested in characterizing the scrapie strains prevalent in Japan. In the absence of known biological clones (which would take many years to develop), we attempted to gain information about the nature of these strains by examining the biological and biochemical properties of a panel of isolates and showed that at least three types of scrapie strains exist in Japan. It is thought that the scrapie agent entered into Japan along with imported sheep in the1970s . Since the recognition of the first scrapie case early in the1980s, similar60 scrapie cases have been officially reported so far. Thus, our data indicate the possibility that the primary scrapie strain has already exhibited considerable variation during the past quarter century. However, it is unknown if the current repertoire of scrapie strains is derived from a single parental strain that has been changing during its spread through the sheep population, similar to the mutation-like change found on serial transmission of the scrapie agent in experimental animals . Alternatively, multiple infection is likely to occur in natural scrapie , so that it is also possible that multiple strains have entered into Japan with imported sheep. The diversity of sheep scrapie strains in Japan seems similar to that described in the United Kingdom based on transmissibility to mice , because some isolates were essentially nontransmissible whereas others were readily transmissible to PrPA/A mice with incubation periods of >350 days. However, to our knowledge, sheep scrapie which successfully transmits to mice with an average incubation period of similar230 days, like A1 and B3, has an extremely short incubation period in primary transmission. TSE agents have been characterized by transmissibility to experimental animals and distribution of neuropathological lesions . In addition, biochemical characterization of PrPSc seems to be useful for discriminating between TSE strains. For example, two hamster-adapted transmissible mink encephalopathy strains, "Hyper" and "Drowsy," can be distinguished from each other not only by incubation period and neuropathology but also by the relative PK resistance and molecular mass of PrPSc . Here we found that relative PK resistance of PrPSc is likely to be useful for the distinction of sheep scrapie strains to some extent, as the samples we analyzed could be divided into two groups based on relative PK resistance. Recently, the ratio of glycosylated PrPSc bands was used to distinguish vCJD from other types of CJD , and sheep scrapie can be divided into several groups by using this glycoform typing of PrPSc . In general, strain typing approaches using a combination of different properties provide a more reliable means of strain differentiation. Analyses of biochemical properties of PrPSc are less time-consuming than bioassays in mice, and therefore, it is advantageous to use a combination of several biochemical properties such as relative PK resistance, molecular mass, and glycoform typing of PrPSc, to achieve more definitive strain typing. A recent study using 10 sheep with scrapie in the United States showed a perfect correlation between the detection of PrPSc in sheep brain and transmissibility to mice . In contrast, it is well-known that some sheep scrapie strains are difficult to transmit to mice . In this study, we showed that scrapie strains classified into groups 1 and 2 differed in their transmissibility to mice but were indistinguishable by the relative PK resistance of PrPSc . The discrepancy between the presence of PrPSc and the transmissibility to mice may imply that factors other than PrPSc are involved in determining the infectivity or that PrPSc is not the molecule responsible for infectivity . Alternatively, strain-specific conformations of PrPSc have been shown , and so it is also conceivable that subtle differences in biochemical and/or biophysical properties of PrPSc which cannot be detected by PK digestion may influence the transmissibility to mice. It is also possible that the sheep PrP genotype may explain the discrepancy. It has not yet been clarified whether amino acid sequences of sheep PrPSc influence transmission to mice. Here we showed that scrapie occurring in sheep homozygous for PrPMARQ transmitted to mice (e.g., SB, Y2, Y5, A1, and G1). This indicates that sheep PrPSc composed of the product of the PrPMARQ allele could initiate the accumulation of PrPSc in mice by using mouse PrPC as a substrate. However, there are differences in the incubation periods for A1 between these sheep and other sheep possessing the PrPMARQ/MARQ genotype , and more recently, it was reported that one scrapie case occurring in PrPARQ/ARQ sheep (amino acid polymorphisms at codon 112 was unavailable) was virtually nontransmissible to mice , suggesting the amino acid sequence of sheep PrPSc is not the sole determinant of the transmissibility to mice. Among the sheep with scrapie used here, S1, S2, and S3 differed in clinical course from other sheep by death at a young age and hypersensitivity. The simple explanation for this is that the strain(s) infecting these sheep differs from those of the other scrapie cases, and indeed, these agents obviously differed from the others in transmissibility to mice. Scrapie strains adapted to rodents exhibit strain-specific clinical symptoms, neuropathological lesions, and incubation periods . However, it is unclear whether a given sheep scrapie strain determines strain-specific clinical and/or clinico-pathological features. As described in Materials and Methods, these sheep have an immediate blood relationship and were kept on the same ranch, and so other factors such as genetic background, breeds, route of infection, and environment may be involved in defining the type of disease in these particular cases. A particular type of scrapie agent prevalent in the United Kingdom is believed to have initially caused BSE, and later on, this agent passed to felids and human beings through cattle . The diversity of field scrapie isolates in Japan is similar to that reported in the United Kingdom at least in terms of primary transmission to mice . Therefore, at present, we cannot exclude the possibility of the presence of a scrapie strain in Japan that possesses properties similar to the BSE agent. Strain typing using two mouse strains, RIII and C57BL, carrying the PrPA/A allotype appears to be one of the methods to distinguish BSE and BSE-related TSE agents from sheep scrapie . Further analysis will be required to investigate whether sheep scrapie strains in Japan have the potential risk to create new epidemics like the BSE agent. Backmatter: PMID- 12202600 TI - Increasing Incidence of Candidemia: Results from a 20-Year Nationwide Study in Iceland AB - A nationwide study on candidemia was conducted in Iceland from 1980 to 1999. The annual incidence increased from 1.4 cases/100,000 inhabitants/year between 1980 and 1984 to 4.9 cases/100,000 inhabitants/year between 1995 and 1999 (P < 0.0001). Candidemia episodes at university hospitals increased from 0.15/1,000 admissions to 0.55/1,000 admissions (P < 0.0001). Candida albicans was the predominant species responsible (64.4%). The national import of fluconazole increased approximately fourfold during the second half of the study, but increased resistance to this agent was not observed. Keywords: Introduction : The incidence of invasive fungal infections is increasing in many hospitals. The National Nosocomial Infections Surveillance (NNIS) program for U.S. hospitals documented a doubling in the rate of nosocomial fungal infections between 1980 and 1990 . From 1995 to 1996, Candida species was the fourth most common cause of nosocomial bloodstream infections (BSIs) in the United States . In recent years, some studies have reported an increase of candidemia due to Candida non-albicans species, with the threat of increased mortality and antifungal drug resistance . The epidemiology of candidemia has been studied primarily in selected hospitals ; few studies have focused on this problem on a nationwide basis . The purpose of this study was to examine candidemia in Iceland during a 20-year period with respect to epidemiology and clinical mycology. Due to the high quality and accessibility of demographic and medical data, Iceland, a 103,000-km2 island in the north Atlantic ocean with 285,000 inhabitants, is well suited for epidemiological studies. All patients in Iceland with yeast isolated from blood from 1 January 1980 to 31 December 1999 were identified retrospectively by a nationwide search in microbiology databases. During the first decade of the study, a Bactec (Becton Dickinson Microbiology Systems) radiometric system was most widely used. During the second decade of the study, the nonradiometric systems Bactec, Difco ESP (Becton Dickinson), and bioMerieux Vital (bioMerieux) were used. There are 2 university or university-affiliated hospitals and 14 county hospitals in the country. Three clinical microbiology laboratories process blood cultures from all of the hospitals. An episode of yeast BSI was defined as at least one blood culture positive for yeast. Episodes were considered separate if they occurred more than 2 weeks apart. Viable yeast BSIs were subcultured on Sabouraud agar (Oxoid). Species identification was based on germ tube production, distinctive color, and morphology on CHROMagar (Hardy Diagnostics) and sugar assimilation profiles by using the API id32C system (bioMerieux). The MICs of amphotericin B, fluconazole, and itraconazole were determined by using an Etest (AB Biodisk) according to instructions from the manufacturer (Antifungal susceptibility testing of yeasts, Etest technical guide 4, AB Biodisk, Solna, Sweden, 1997). The plates were incubated at 35C for 48 h before reading the MICs. The National Committee for Clinical Laboratory Standards (NCCLS) breakpoint criteria for antifungal susceptibility were used . National import figures on antifungal agents for the period from 1980 to 1999 were obtained from the Icelandic Association of Importers of Pharmaceuticals. The number of defined daily doses (DDD) per packing was calculated, and the import was expressed as DDD per 1,000 inhabitants per year. Information about national demographics, including age distribution, for the period from 1980 to 1999 was obtained from the Bureau of Statistics in Iceland. Information on admissions to pediatric, medical, and surgical wards at the university hospitals for each year of the study was obtained from annual hospital reports. The chi-square test for linear trend was used to compare incidence rates between study periods, and the chi-square test was used to compare fungal species distribution and proportions of blood cultures positive for yeasts between the first and second halves of the study period. In the 20-year period from 1980 to 1999, 172 episodes of BSI with yeasts, predominantly candidemia, were diagnosed in 165 patients in Iceland. Children (<=16 years) comprised 11 (7%) of the patients, and adults comprised 154 (93%) of the patients. Six patients had two or more separate episodes, occurring at least 2 weeks apart. The nationwide annual incidence of candidemia and number of candidemic episodes per 1,000 admissions to the university hospitals is shown in Fig. . A vast majority of the patients (87%) came from the university hospitals. Figure shows the age-specific incidence for the first and second halves of the observation period, respectively. During the second decade of the study, the use of blood cultures at the two university hospitals increased from 71,002 vials between 1990 and 1994 to 93,032 between 1995 and 1999. The proportion of blood cultures positive for yeasts increased slightly, from 0.187% between 1990 and 1994 to 0.204% between 1995 and 1999, but this increase was not significant (P = 0.464). FIG. 1. | Increasing incidence (cases/100,000 inhabitants/year) of candidemia in Iceland during a 20-year period, from 1980 to 1999. Increasing incidence (cases/100,000 inhabitants/year) of candidemia in Iceland during a 20-year period, from 1980 to 1999. The open squares with solid lines represent the population-based incidence, whereas the filled squares with dashed lines depict the incidence as a proportion of 1,000 admissions to the university hospitals in the country. As shown, the nationwide incidence increased 3.5-fold during the 20 years of the study, from 1.4 between 1980 and 1984 to 4.9 between 1995 and 1999 (P < 0.0001). The number of candidemic episodes per 1,000 admissions at the two university hospitals also increased significantly, from 0.15/1,000 admissions between 1980 and 1984 to 0.55/1,000 admissions between 1995 and 1999 (P < 0.0001). FIG. 2. | Age-specific incidence of fungal BSIs in Iceland from 1980 to 1999 calculated for the first and second decades of the study. Age-specific incidence of fungal BSIs in Iceland from 1980 to 1999 calculated for the first and second decades of the study. As shown, the age-specific incidence was similarly distributed in the first (open circles, solid lines) and the second (filled circles, dashed lines) decades, with the exception of the youngest age group (<1 year old). The incidence was low in people aged 5 to 40 years but started to rise among people aged 41 to 50 years. Incidence subsequently increased with advancing age, peaking among patients aged 61 to 70 years (1980 to 1989, 12.7 cases/100,000 inhabitants/year; 1990 to 1999, 19.3 cases/100,000 inhabitants/year). From 1980 to 1989, no children of <1 year of age were diagnosed with candidemia. After 1989, however, the incidence followed a biphasic pattern, with a high incidence occurring in the youngest age group (<1 year, 11.3 cases/100,000 inhabitants/year). Most of the children younger than 1 year old were preterm infants in the neonatal ICU at the time of diagnosis. Table shows the yeast species that were cultured from the blood of Icelandic patients from 1980 to 1999. Species distribution remained relatively stable in the country, with C. albicans causing around two-thirds of infections (range, 63.4 to 65.3%) and Candida non-albicans species causing one-third. There was no significant change in yeast species distribution between the first and second halves of the study (P = 0.35). Species identification of all available isolates revealed Candida dubliniensis in four cases. TABLE 1 | Yeast strains isolated from blood cultures in Iceland from 1980 to 1999 The MICs of amphotericin B, fluconazole, and itraconazole were determined for 99 strains dating from 1991 to 1999 . All were susceptible to amphotericin B (MIC, <=1 mug/ml), 97% were susceptible to fluconazole (MIC, <=8 mug/ml), and 87% were susceptible to itraconazole (MIC, <=0.125 mug/ml). During the study period, the number of isolates classified as susceptible-dose dependent for itraconazole increased from 6.7% (3 of 45) between 1991 and 1995 to 16.7% (9 of 54) between 1996 and 1999, but this increase was not statistically significant. TABLE 2 | In vitro susceptibility of 99 fungal BSIs to amphotericin B, itraconazole, and fluconazole The national import of antifungal agents from 1990 to 1999 is shown in Table . Fluconazole was approved for oral and parenteral use in 1990. During the period from 1991 to 1999, the import of oral formulations increased from 9.4 to 47.6 DDD/1,000 inhabitants/year (406%) and the import of fluconazole for parenteral use increased from 0.5 to 1.8 DDD/1,000 inhabitants/year (260%). TABLE 3 | National import of antifungal agents in Iceland from 1990 to 1999 expressed as DDD per 1,000 inhabitants per year Several studies have shown a substantial increase in the incidence of candidemia in the past 2 decades. Data from the NNIS system on nosocomial BSIs showed an up to fivefold increase in incidence between 1980 and 1989 in the United States . According to another surveillance study from the NNIS program, between 1995 and 1996 Candida species was the fourth most common cause of nosocomial BSI . Our data show that the incidence of candidemia in Iceland has increased steadily and significantly over the past 2 decades. The incidence is somewhat higher than that documented in a nationwide study in Norway from 1991 to 1996 (an average of 0.17 per 1,000 discharges) but similar to that observed in Israel in 1994 (0.50 per 1,000 admissions) . Although the use of blood cultures increased at the university hospitals, the proportion of cultures that turned out positive for yeasts remained stable. The observed increase may be due in part to improved detection, but other factors are likely to have played a role. Among them are the greater use of invasive devices and broad-spectrum antibacterial agents, more extensive surgical procedures, and advanced life support . The length of stay at the university hospitals in Iceland, where 87% of the patients were diagnosed, did not increase in the past 2 decades and therefore does not explain our findings. Many studies on the epidemiology of fungal BSIs have focused on selected hospitals or hospital wards, with different patient populations, and have reported a substantially higher incidence than is reported in this study. According to two studies on candidemia, one in Australia and one in the United States, the incidence was 1.5 and 3.3 episodes per 1,000 discharges, respectively . Another study at the M.D. Anderson Cancer Center reported an incidence of 6 cases of fungemia per 1,000 admissions between 1988 and 1992 . A prospective multicenter study of candidemia at six sites in the United States concluded that the incidence was 9.8 cases/1,000 admissions in surgical intensive care units (ICUs) and 12.3 cases/1,000 admissions in neonatal ICUs . Studies on candidemia and pathogen species distribution from different parts of the world generally agree that C. albicans is still the most commonly isolated fungal pathogen from blood, causing between 50 and 70% of infections . According to our data, approximately two-thirds (64.4%) of fungal BSIs in Iceland were caused by C. albicans and one-third were caused by Candida non-albicans species. These results are comparable to those reported from Norway but somewhat higher than those reported from the United States, Canada, Latin America, and Europe . Among the different fungal species, we identified C. dubliniensis as the pathogen in four cases. This species, which was originally described by Sullivan and coworkers in 1995 , is germ tube positive and can therefore easily be mistaken for C. albicans . The proportion of candidemias caused by Candida non-albicans has not increased in the past 20 years according to our results. In contrast, other studies have reported a shift towards Candida non-albicans species in the past 5 to 10 years . Antifungal prophylaxis with fluconazole may have played a role in this observed shift. In two studies, fluconazole prophylaxis was the single most important determinant for the relative increase in Candida krusei and Candida glabrata infections . Fluconazole prophylaxis is used infrequently in Iceland, which may explain this difference. The antifungal susceptibility patterns revealed that 97% of the Icelandic strains were susceptible to fluconazole, despite an approximately fourfold increase in fluconazole import in the past decade. Little has been published about national consumption of antifungal agents. Our results can be compared to data from the study conducted in Norway, where the use of fluconazole increased from 8.0 to 16.1 DDD/1,000 inhabitants/year between 1991 and 1996 . During the same period, a greater increase in fluconazole import was seen in Iceland: from 9.9 to 24.9 DDD/1,000 inhabitants/year. The use of amphotericin B remained relatively stable. In conclusion, this study has shown that, on a national level, the incidence of candidemia in Iceland has increased 3.5-fold over the past 2 decades, with the highest incidences of infection occurring in the youngest and older age groups. The proportion of infections caused by Candida non-albicans species has remained stable. Fluconazole use has increased approximately fourfold in the past decade, but the majority of strains are still susceptible to this agent. Backmatter: PMID- 12202602 TI - Technical Improvement To Prevent DNA Degradation of Enteric Pathogens in Pulsed-Field Gel Electrophoresis AB - This study used a modified pulsed-field gel electrophoresis (PFGE) method with HEPES as a running buffer to prevent electrophoresis-related DNA degradation of nine Salmonella enterica subsp. enterica serovar Ohio, seven Salmonella serovar Newport, and two enterohemorrhagic Escherichia coli (non-O157) strains. All strains yielded identifiable bands with this method in contrast to a commonly applied PFGE method using Tris buffer. Keywords: Introduction : Indigenously acquired salmonelloses caused by Salmonella enterica subsp. enterica serovar Ohio have been rare in Finland: from 1990 through 2000 only 11 domestic cases were identified, compared with 69 cases of foreign origin. However, in January 2001, a cluster of cases of gastrointestinal disease, 11 of which were microbiologically confirmed, occurred in northern Finland. The same serovar was also found in drinking water taken from a local well. Various phenotypic methods, such as serological typing, phage typing, and antimicrobial susceptibility testing, have traditionally been used in epidemiological studies of Salmonella outbreaks. However, these methods do not always give sufficient information for epidemiological purposes, even in outbreaks caused by a rare Salmonella serotype or phage type. More recently, molecular epidemiology-based techniques analyzing chromosomal DNA or plasmids have been shown to be useful for typing several Salmonella serotypes . Whole-cell DNA analysis by pulsed-field gel electrophoresis (PFGE) has usually proven to be superior to other molecular methods in its discriminatory value. For this study, four outbreak isolates (three isolates from patients and one isolate from water) and five human isolates which were epidemiologically unrelated were chosen for PFGE typing. DNA was prepared as described previously by Gautom , with slight modifications, digested with the restriction enzyme XbaI or BlnI, and separated in 1% agarose, with pulses ramped linearly from 5 to 70 s for 24 h (5.4 V/cm, 14C). Electrophoresis was performed with 0.5x Tris-borate-EDTA (TBE) as a buffer. Only one of the nine Salmonella serovar Ohio strains was typeable by this method, while the DNAs of the other eight strains degraded during the electrophoresis. We have experienced a similar problem with Salmonella serovar Newport isolates and tried to resolve it with formaldehyde fixation, by increasing the incubation times, by varying the concentrations of proteinase K, and by using preincubation with lysozyme prior to the deproteination. None of these methods were useful in protecting the DNA of Salmonella serovar Newport isolates from degradation. Ray et al. have reported that electrophoresis-related, Tris-dependent degradation of Streptomyces lividans DNA was prevented by the use of HEPES buffer (16 mM HEPES-NaOH, 16 mM sodium acetate, 0.8 mM EDTA [pH 7.5]) instead of Tris-containing buffer or the addition of thiourea to the buffer to neutralize a nucleolytic derivate of Tris . Thiourea has also been reported as being useful in PFGE typing of degradation-sensitive Pseudomonas aeruginosa and Clostridium difficile . However, thiourea is a suspected cancer-causing agent, and therefore we decided to use HEPES buffer. After the Tris-containing running buffer (0.5x TBE) was changed to non-Tris-containing HEPES buffer, all Salmonella serovar Ohio strains were typeable. Running buffer conditions had to be modified by reducing the voltage to 4 V/cm to keep the current within the normal range with HEPES, which has a higher ionic strength than does 0.5x TBE. This method was also successfully tested with seven previously untypeable Salmonella serovar Newport strains and two enterohemorrhagic Escherichia coli non-O157:H7 strains. With XbaI as the restriction enzyme, all but one Salmonella serovar Ohio strain shared a common PFGE type. With BlnI, the strains divided into four PFGE types. The four outbreak strains had indistinguishable PFGE patterns with both enzymes, whereas none of the nonoutbreak strains had the PFGE patterns of the outbreak strains . These findings support the association between outbreak and water isolates. Also, as was earlier assumed by Romling and Tummler , the degradation of DNA does not seem to be a clonal trait, because the only strain typeable with Tris shared common XbaI and BlnI PFGE types with a strain that was affected by DNA degradation. FIG. 1. | PFGE banding patterns of Salmonella serovar Ohio isolates digested with the BlnI restriction enzyme. PFGE banding patterns of Salmonella serovar Ohio isolates digested with the BlnI restriction enzyme. Lanes 1, 2, 4, and 5, human isolates from subjects who had recently returned from Spain, the United Arab Emirates, Latvia, and Spain, respectively (year 2000); lane 3, human isolate of domestic origin (year 2000); lanes 6 to 8, human outbreak isolates; lane 9, isolate from drinking water; lanes M, lambda ladder molecular size markers (concatemers of 48.5 kb). All isolates but that in lane 1 were nontypeable with 0.5x TBE as a running buffer. On the basis of this study, the use of HEPES instead of Tris-containing running buffers in PFGE of degradation-sensitive Salmonella and enterohemorrhagic E. coli strains seems to be a convenient, inexpensive, and safe way to ensure typeability in epidemiological investigations. Backmatter: PMID- 12202613 TI - Isolation of Nocardia paucivorans from the Cerebrospinal Fluid of a Patient with Relapse of Cerebral Nocardiosis AB - Nocardia paucivorans represents a new species of the genus Nocardia that has recently been isolated from bronchial secretions of a patient with chronic lung disease. Here, we report on the course of a disseminated infection caused by this species: i.e., cerebral and subsequent meningeal manifestations, isolation from the cerebrospinal fluid, and in vitro susceptibility to various antimicrobial agents. Keywords: CASE REPORT : A 63-year-old man presented to the emergency department of this hospital in July 2000 with a 10-day history of headaches and mild ataxia on the right-hand side. Cranial computed tomography showed a cerebellar mass with surrounding edema, without signs of intracranial bleeding, and the result was confirmed by magnetic resonance imaging. The patient's condition improved with conservative treatment. A few days after admission, a blood pressure crisis occurred, resulting in an impaired neurological status that persisted despite prompt sufficient control of systemic blood pressure. Imaging demonstrated a minor perilesional bleeding in the cerebellum as well as an increase in the size of the original mass. The patient underwent craniotomy, during which the mass, now identified as an abscess, was removed. Histological investigation of the biopsies yielded granulomatous inflammation. Additionally, infiltration of the tissue by dark brown to black branching rods was seen in Grocott-stained sections, a typical feature of nocardiae . Brain abscess material was inoculated in brain heart broth (BHB) and blood (BA) and chocolate (CA) agars and incubated at 36C in an atmosphere enriched with 5% CO2. However, no growth was found after 10 days of incubation. Failure to isolate the pathogen despite its histological demonstration may have been due to sampling error. A decreased number of CD4+ T cells (172 per mul) was found, demonstrating an immunodeficiency of unknown origin, since human immunodeficiency virus types 1 and 2 antibody and p24 antigen tests were negative. The patient received a course of intravenous trimethoprim-sulfamethoxazole, which was replaced by amikacin plus a carbapenem because of systemic side effects. This therapy was administered for 6 weeks, with imipenem given during the first week and meropenem, which was better tolerated, given during the remaining 5 weeks. The patient made a slow recovery and was discharged with minor neurological deficits (tremor and dysdiadochokinesis). The CD4+ T-cell count remained low (maximum count, 414 per mul). Treatment of the nocardiosis was continued with oral minocycline. Antibiotic therapy was discontinued after a duration of 6 months. The patient developed headaches and singultus within 2 weeks, resulting in readmission to the hospital. A lumbar puncture at readmission showed 1,306 granulocytes per mm3 in the cerebrospinal fluid (CSF) without detection of infectious agents upon microscopy. Treatment for suspected bacterial meningitis was initiated with ceftriaxone plus ampicillin. However, the patient's condition did not improve under this regimen, and no fast-growing bacteria could be isolated. Therefore, therapy was changed after 3 days to amikacin plus meropenem, assuming a meningeal manifestation of nocardiosis. The administration of amikacin plus ampicillin plus clavulanate during the 4th week worsened the side effects, leading to reinitiation of the former regimen. After 6 weeks of intravenous therapy, treatment was changed to oral levofloxacin and minocycline based upon susceptibility test results and accounting for multiple drug intolerances. The patient was discharged home with this therapy and since then has been seen regularly in our outpatient clinic. The leukocyte count in the CSF decreased, but it remained mildly elevated for more than 10 weeks before falling to 10/mm3. The pathological findings in magnetic resonance imaging of the brain remained unchanged, but the neurological symptoms have improved. CSF specimens for bacterial culture obtained on readmission were inoculated in liquid media (BHB and the commercially available VITAL blood culture system [bioMerieux S.A., Marcy l'Etoile, France] for aerobic and anaerobic incubation), as well as on BA, CA, and brain heart agar (BHA). Agars and liquid media showed no signs of bacterial growth after 7 days of incubation. Subcultures of the aerobic blood culture medium (not detected as positive by the automatic fluorescence readout system) on BA, CA, and BHA yielded yellowish colonies after 5 days of incubation at 36C in an atmosphere enriched with 5% CO2, whereas there was no bacterial growth in subcultures of the BHB. Growth was most abundant on CA; the colonies developed white aerial hyphae after 3 to 4 days of incubation. On microscopic examination of BHA, microcolonies showed irregular branching. Growth of nocardiae (confirmed by subculture and microscopy) was also found on Loewenstein-Jensen agar after 2 weeks of incubation. The Gram- and Ziehl-Neelsen-stained preparations from single colonies showed mostly gram-positive branched rods and coccoid fragments, as well as partially-acid-fast rods and coccoid fragments, respectively. The morphology of the colonies and the typical Gram and acid-fast stains led to the diagnosis of Nocardia species. Tests for basic biochemical reactions, such as oxidative and fermentative glucose use, gelatin liquefaction, nitrate reduction, and urea hydrolysis, were negative. In addition, the isolate showed remarkable susceptibility to a wide range of antimicrobial agents (disk diffusion test on Mueller-Hinton agar [data not shown]). Subsequent sequence determination of a 600-bp fragment amplified from the 3' end of the 16S rRNA gene identified the isolate as Nocardia paucivorans. The diagnosis was confirmed by the German national consultant laboratory for actinomycetes. Comparative in vitro susceptibility testing including the N. paucivorans type strain (DSM 44386T) was performed on Mueller-Hinton agar at 36C in an atmosphere enriched with 5% CO2 by using the Etest system, and results were read after 24 and 48 h. Strains of Nocardia nova (ATCC 33726) and Nocardia farcinica (ATCC 3308) along with previously published results of broth dilution testing for these organisms were used as controls. The results from our isolate confirmed those obtained by disk diffusion testing . Slightly higher MICs were found for the N. paucivorans type strain, for which no susceptibility data have been available. The results from the control strains were comparable to those published previously . Discussion. | The ubiquitous nocardiae are aerobic actinomycetes found worldwide in soil, sand, and house dust . The organisms are characterized as gram-positive, partially-acid-fast branched rods that form irregular branching colonies on agar; aerial hyphae may be present in some species. The respiratory tract and skin are the sites of entry in most cases of nocardiosis, and nosocomial outbreaks of wound infections have been reported . The course of the disease may be subclinical and self-limiting, but chronic disease can develop. The nocardiae can spread from the site of entry to the brain, kidneys, eyes, bones, joints, and other organ systems. Immunocompromised hosts (e.g., patients with AIDS, malignancies, diabetes, or under immunosuppressive therapy after transplantation or for rheumatological disorders) are more frequently affected, but cases among immunocompetent patients have also been reported (-, , , ). The antibiotics most frequently used are trimethoprim-sulfamethoxazole or minocycline in localized or mild cases and amikacin, imipenem, or cefotaxime in severe cases, such as disseminated infection. Recommendations for the duration of treatment in cases of cerebral nocardiosis vary from 6 to 12 months in immunocompetent hosts, and some authors recommend even longer treatment regimens in immunocompromised individuals . There is no standard method for in vitro susceptibility testing of Nocardia species, but disk diffusion, Etest, and broth and agar dilution methods have been described (, , , -). While high rates of resistance to multiple classes of antibiotics have been reported for various clinical and environmental Nocardia isolates , data on N. paucivorans are not available. Here, we report a case of cerebral infection with the recently defined species N. paucivorans, which is closely related to the Nocardia asteroides complex . Histologically it was diagnosed as a cerebellar nocardial abscess , and microbiologically after its spread to the meninges, representing a rare manifestation of nocardiosis . Different factors may have contributed to the failure to cure the disease and the consequent relapse in this patient despite in vitro susceptibility of the isolate to all drugs administered. Because the specific immune response to nocardiae is mainly T-cell mediated , the patient's immunosuppression is likely to have played a role in addition to drug- and pathogen-related factors. These include the duration of treatment, as well as poor penetration of minocycline into the CSF and its bacteriostatic mode of action. Additionally, L-forms of Nocardiae, which can persist within host tissues, have been described . Ours is the first reported case of N. paucivorans isolated from the CSF of a patient with intracranial or intrathecal manifestation of nocardiosis. It demonstrates that this species can cause severe systemic infections similar to those caused by other medically important Nocardia species, such as N. asteroides, N. farcinica, and N. otitidiscaviarum. The remarkably low level of antimicrobial resistance associated with the few unique biochemical characteristics of both the isolate reported here and the type strain of N. paucivorans necessitates its identification by molecular methods. FIG. 1. | Nocardiae (arrow) in a Grocott-stained histological section of the brain abscess material. Nocardiae (arrow) in a Grocott-stained histological section of the brain abscess material. Magnification, x500. TABLE 1 | MICs of the N. paucivorans CSF isolate, the N. paucivorans type strain N. nova, and N. farcinica as read by Etest on Mueller-Hinton agar Backmatter: PMID- 12202592 TI - Quantitative Multiprobe PCR Assay for Simultaneous Detection and Identification to Species Level of Bacterial Pathogens AB - We describe a novel adaptation of the TaqMan PCR assay which potentially allows for highly sensitive detection of any eubacterial species with simultaneous species identification. Our system relies on a unique multiprobe design in which a single set of highly conserved sequences encoded by the 16S rRNA gene serves as the primer pair and is used in combination with both an internal highly conserved sequence, the universal probe, and an internal variable region, the species-specific probe. A pre-PCR ultrafiltration step effectively decontaminates or removes background DNA. The TaqMan system described reliabAly detected 14 common bacterial species with a detection limit of 50 fg. Further, highly sensitive and specific pathogen detection was demonstrated with a prototype species-specific probe designed to detect Staphylococcus aureus. This assay has broad potential in the clinical arena for rapid and specific diagnosis of infectious diseases. Keywords: Introduction : Currently, the standard method for diagnosing the presence of bacterial pathogens in clinical samples relies on culture techniques. However, active research is under way using new molecular methods to decrease detection time and increase assay sensitivity. PCR has emerged as the molecular method of choice in achieving these objectives. The utility of PCR and other molecular methods is evidenced by the recent guidelines issued by the NCCLS in 1999 encouraging the use of such methods in clinical laboratories performing bacterial identification assays . To detect the presence of any bacterial pathogen in a clinical sample, primers annealing to regions of DNA conserved across a wide range of bacterial genomes have been employed. The design of such universal primers has often focused on the 16S rRNA gene . The presence of multiple copies of this gene within the bacterial genome facilitates its amplification by PCR. Further, sufficient sequence variability allows phylogenetic information to be attained for the purposes of microbial identification. However, up to the present, assays which provide for both universal detection and species identification require a second post-PCR processing step, which can be technically cumbersome and slow the time to reporting of results . Universal PCR-based bacterial detection systems have also been hampered by contamination issues. High sequence conservation of the DNA region chosen for PCR primer annealing coupled with the immense amplification power of PCR results in the amplification of exceedingly minor bacterial contaminants, leading to false positives. Attempts to decontaminate PCR materials have involved nearly all known methods to destroy DNA, including UV irradiation, 8-MOP treatment, and incubation with various enzymes, such as DNase, restriction enzymes, or both in combination . Thus far, none of these methods has been shown to be entirely effective or reproducible. Assessment of bacterial contamination can most reliably be made using real-time detection methods to characterize PCR amplification. Briefly, real-time PCR amplifications are reported by the cycle number at which the PCR product accumulates significantly over baseline levels, as detected by interaction with fluorogenic probes (CT) . Aside from saving time and labor, this technique has been shown to be more objective and consistent than the traditional methods of amplification detection and starting template quantification involving gel electrophoresis . With this more precise technique, Corless et al. found that most decontamination methods decreased PCR sensitivity . The implication of this finding was that the decontamination effect of the aforementioned methodologies could at least in part be explained by a retardation of the sensitivity of the PCR amplification system. The quantitative capacity of real-time PCR has thus redefined the standards by which a decontamination method is measured. Not only will a particular method be required to yield negative for controls under the more precise probe-based real-time system but also the method must be shown to preserve the sensitivity of the PCR assay. In this report, we describe a multiprobe-based real-time PCR system involving the 16S rRNA gene, which allows for simultaneous detection of the presence of eubacterial DNA with species-specific discrimination. In addition, we report a decontamination method for the present PCR system which does not compromise detection sensitivity. MATERIALS AND METHODS : Bacterial species and DNA isolation. | Fifteen common pathogenic microorganisms, all of which were eubacterial except one, Candida albicans, were obtained from the clinical laboratory (Division of Medical Microbiology, Johns Hopkins School of Medicine, Baltimore, Md.). The species and American Type Culture Collection (ATCC) numbers are listed in Table . Microorganisms were grown in standard cultures, and DNA was extracted using the QIAamp DNA kit (Qiagen Corp., Santa Clarita, Calif.). With regard to generating standard curves for starting DNA template quantification, Staphylococcus aureus (ATCC 29213) was grown in Luria-Bertani broth (Gibco/BRL Life Technologies Inc., Gaithersburg, Md.) at 37C with continuous shaking to an optical density at 600 nm of 0.6. Equal aliquots were then plated to determine CFU and subjected to DNA extraction with the QIAamp DNA kit. The isolated DNAs were quantified based on optical density at 260 nm and then serially diluted. Analogous DNA isolation procedures were performed for three other strains of S. aureus (ATCC strains 02131, 15923, and 43300) as well as Staphylococcus epidermidis and Staphylococcus hominis to permit testing of both sensitivity and specificity of the SA probe. Design of primers and probes. | The 16S rRNA gene sequences from a variety of bacterial species were obtained from GenBank. Sequence data were obtained using the program Entrez (see the list below). The sequences were aligned using the program ClustalW from the European Bioinformatics Institute . Two regions of highly conserved sequences, separated by both an internal region of highly variable sequence and another adjacent internal region of highly conserved sequence, were selected as the universal primer annealing sites. The internal highly conserved and highly variable sequences were used as the annealing sites of conserved and species-specific TaqMan probes, respectively . The primers and TaqMan probes were designed according to the guidelines in the ABI Primer Express software program (PE Applied Biosystems, Foster City, Calif.). This program selects probes and primer sets with optimized melting temperatures, secondary structure, base composition, and amplicon lengths. The forward primer (p891F) and reverse primer (p1033R) amplify a fragment of 161 bp spanning nucleotides 891 to 1051 of the S. aureus 16S rRNA gene . The universal TaqMan probe, or UniProbe, was labeled with the reporter dye VIC at the 5' end, and the quencher dye TAMRA was labeled at the 3' end and has the sequence which is the reverse complement of nucleotides 1002 to 1024 of the 16S rRNA gene (see list of sequence accession numbers below) . S. aureus-specific probe, or SAProbe, was designed as the species-specific probe. The SAProbe was labeled with a different reporter dye, FAM, at the 5' end and the same quencher dye at the 3' end, with the sequence which spans nucleotides 946 to 976 of the S. aureus 16S rRNA gene . The probes were designed to anneal to opposite strands of the template DNA. The primers and probes were manufactured by PE Applied Biosystems. PCR master mix and fluorogenic-probe-based PCR (TaqMan assay). | Reactions were performed in 50-mul volumes in 0.5-ml optical-grade PCR tubes (PE-Applied Biosystems). PCR master mix was prepared from the TaqMan core reagent kit (PE-Applied Biosystems). The master mix was comprised of 200 muM (each) dATP, dGTP, dUTP, and dCTP, 0.5 U of AmpErase uracil-DNA glycosylase (UNG), 2.5 mM MgCl2, 1x TaqMan buffer A, 900 nM concentrations of each primer, and 100 nM concentrations of each fluorescence-labeled probe (UniProbe and/or SAProbe). Template DNA, 2 U of AmpliTaq Gold DNA polymerase (PE-Applied Biosystems), and water were added to give a final volume of 50 mul for each sample. The fluorogenic-probe-based PCR, or TaqMan assay, was performed using the ABI 7700 sequence detection system (PE-Applied Biosystems). The cycling conditions used were as follows: 50C for 2 min and 95C for 10 min, followed by 40 cycles at 95C for 15 s and 60C for 1 min each. All PCRs were performed in triplicate. The correct size of the PCR product from each assay was verified by running an amplified sample from each reaction tube on agarose gels stained with ethidium bromide. Ultrafiltration of the PCR mix. | An ultrafiltration step, using the Amicon Microcon YM-100 centrifugal filter device (Millipore Corporation, Bedford, Mass.) was utilized for filtering the PCR mix prior to addition of template DNA. The PCR mix that underwent ultrafiltration included the PCR master mix and AmpliTaq Gold DNA polymerase. This filtration device prevents the passage of potential contaminating double-stranded DNA of 125 bp or greater. The PCR mix was spun through the YM-100 device at 100 x g for 30 min. Post-PCR analysis. | Amplification data were analyzed by the SDS software (PE-Applied Biosystems), which calculates DeltaRn using the equation Rn(+) - Rn(-). Rn(+) is the emission intensity of the reporter divided by the emission intensity of the quencher at any given time, whereas Rn(-) is the value of Rn(+) prior to PCR amplification. Thus, DeltaRn indicates the magnitude of the signal generated. The threshold cycle, or CT, is the cycle at which a statistically significant increase in DeltaRn is first detected. The CT is inversely proportional to the starting amount of target DNA. Amplification plots were generated by plotting DeltaRn versus CT . Nucleotide sequence accession numbers. | The GenBank accession numbers for the sequences determined in this study are as follows: S. aureus, AF015929; S. hominis, AY030318; Enterococcus faecalis, AJ276460; S. epidermidis, L37605; Streptococcus pneumoniae, X58312; Mycoplasma pneumoniae, AF132741; Escherichia coli, AF233451; Haemophilus influenzae, AF224306; Legionella pneumoniae, M59157; Neisseria meningitidis, AF059671; Rickettsia rickettsii, U11021; Borrelia burgdorferi, AF091368; Bacillus anthracis, AF290552; Yersinia pestis, AF366383; Proteus mirabilis, AF008582; and Klebsiella pneumoniae, AF228919. FIG. 1. | Design of primers and probes. Design of primers and probes. The forward (P891F) and reverse (P1033R) primers anneal to highly conserved regions of the 16S rRNA gene. An internal highly conserved region was selected as the annealing site of the UniProbe, and the other internal region of highly variable sequence was selected as the annealing sites of a prototype species-specific SAProbe. TABLE 1 | Sensitivity of the Taqman assay using universal primers and probes TABLE 2 | Oligonucleotide sequences of primers and probes used in the study RESULTS : Sensitivity of Universal TaqMan PCR. | The sensitivity of the primers and probes used for universal amplification of eubacterial 16S rRNA gene was first assessed with genomic DNA extracts from 14 different bacterial species. An isolate from Candida albicans was used as a negative control . In each PCR assay, 5 ng of purified DNA was used. The assay's positivity was determined by examination of the amplification plot (CT versus DeltaRn) generated by the sequence detection software . All 14 bacterial species were correctly amplified and detected, with CT values in the range of 19.2 to 21.8. No amplification (CT > 40) was detected when DNA isolated from C. albicans was used. The assay results were further verified by subjecting reaction products to gel electrophoresis, with visualization of bands of the expected size (162 bp) (data not shown). Theoretical detection limit of TaqMan PCR. | The detection limit of the TaqMan assay was determined by amplifying serial dilutions of eubacterial DNA. The minimal detection limit of the TaqMan system was defined as the amount of template DNA at which the relationship between CT and starting template DNA became nonlinear. Serial dilutions of S. aureus DNA (50 ng to 5 fg) were added to PCRs with universal primers (p891F plus p1033R) and probe (UniProbe). The results are shown in Table . The standard curve in which CT values were plotted against starting template DNA is linear from 50 ng to 5 pg . At DNA levels below 5 pg, this relationship became nonlinear, and the CTs were similar to the CT of the no-template control (NTC). This suggested the presence of contaminating eubacterial DNA in the NTC. The minimal detection limit of the assay was thus 5 pg of S. aureus DNA. As an effort to improve the detection limit of the assay, we implemented a prefiltration step for the PCR mix prior to the addition of template DNA, in addition to the conventional precautionary measures used for reducing contaminating or carryover DNA present in PCR reagents. The filtration device retains contaminating DNA but allows for passage of all components of the PCR mix, including primers, probes, Taq polymerase, and UNG. Addition of this prefiltration step increased the CT of NTC to 40, effectively reducing the amount of contaminating DNA . CTs at DNA levels (50 ng to 5 pg) remained comparable with or without the prefiltration step. Furthermore, the CT values of starting DNA template below 5 pg and those of that greater than 5 pg all fell on the same line, with an r2 value of 0.998. . The efficiency (E) of the prefiltered PCR amplification was calculated to be 1.94 (maximum = 2) based on the following equation: E = e(-1/slope). The minimum detection limit of the assay with prefiltration was 50 fg of S. aureus DNA . Based on the size of the S. aureus genome, which is approximately 2,750 kbp, 50 fg of S. aureus DNA is equivalent to approximately 15 genomes or CFU. This was calculated as follows: 2,750 kb is equal to 1.8 x 106 g/mol; division of this value by Avogadro's number, 6 x 1023, yields 3 fg per S. aureus genome. Comparable results were derived empirically based on counting CFU on plates. The amplified products were subsequently subjected to gel electrophoresis. Visualization of the bands under UV irradiation confirmed the expected amplicon size (data not shown). Multiprobe assay. | In order to demonstrate the assay's ability to simultaneously detect the presence of any eubacterial DNA as well as one or more species of interest within a single reaction tube, a prototype species-specific probe for S. aureus (SAProbe) was designed. The SAProbe was labeled with a different reporter fluorophore so that its signal could be distinguished from that of the universal probe. Real-time PCR assays were performed with the universal primer set, UniProbe, and SAProbe in a single reaction mix. Template DNA samples from each of three closely related Staphylococcus species, S. aureus, S. epidermidis, and S. hominis, were tested in different reactions. For S. aureus, four strains were tested (ATCC 02131, 15923, 29213, and 43300). The assay correctly detected the presence of eubacterial DNA in all samples under the VIC dye layer. When the detection system was reconfigured to detect the FAM dye layer in the same reaction tubes, only the reaction containing the various strains of S. aureus DNA yielded a positive signal . For further experiments involving S. aureus, strain 29213 was used. PCR results for reactions using UniProbe versus those for reactions using SAProbe showed no significant differences in CT values for amplifications with equivalent amounts of S. aureus DNA . Detection equivalence is also described in Fig. with standard curves in which CTs are plotted against starting template DNA. The similar slopes of the two lines indicate equal efficiencies; the coinciding extrapolated axis intercepts indicate equal detection limits. PCR results obtained by using UniProbe with serial dilutions of S. epidermidis template DNA yielded results similar to those obtained with S. aureus template DNA. Specifically, plots of CTs against starting template DNA revealed almost equivalent amplification efficiencies between the two . PCR results obtained by using SAProbe with serial dilutions of S. epidermidis template DNA gave no detection . Finally, PCRs using both probes with a constant amount of S. epidermidis template DNA and serial dilutions of S. aureus template DNA were performed. SAProbe results coincided with the titration in which only S. aureus DNA was serially diluted. UniProbe results differed, however. At concentrations in which S. aureus DNA predominated over that of S. epidermidis, the CTs of the UniProbe results were comparable with those of the SAProbe results. As the amount of S. aureus DNA was diluted and the S. epidermidis DNA became predominant, the CTs leveled off at a value corresponding to the constant amount of S. epidermidis DNA added to each reaction tube . FIG. 2. | Inverse linear relationship of CT versus starting template DNA with and without prefiltration. Inverse linear relationship of CT versus starting template DNA with and without prefiltration. CTs were generated from PCRs with starting template S. aureus DNA serially diluted 1:10 from 50 ng to 5 fg. In the nonprefiltration curve (squares), nonlinearity is encountered at starting template DNA quantities below 5 pg, suggesting background eubacterial DNA contamination of this amount. With prefiltration (triangle), the detection limit was extended to 50 fg. FIG. 3. | Comparison of the standard curves for S. aureus Comparison of the standard curves for S. aureus DNA using UniProbe and SAProbe. PCRs generating the CTs for the curves contained both UniProbe and SAProbe, as well as S. aureus starting template DNA, which was serially diluted 1:100 from 50 ng to 50 fg. That the lines corresponding to UniProbe (square) and SAProbe (triangle) nearly overlay one another indicates comparable amplification efficiencies and detection limits for the different probes used in the same reaction mixtures. FIG. 4. | Comparison of the standard curves for S. aureus Comparison of the standard curves for S. aureus and S. epidermidis DNA obtained by using UniProbe. PCRs generating the CTs for the curves contained UniProbe and either S. aureus or S. epidermidis starting template DNA. S. aureus DNA was serially diluted 1:100 from 50 ng to 50 fg, whereas S. epidermidis DNA was diluted from 20 ng to 20 fg. The near equivalency of the lines indicates that UniProbe has equal detection capacity irrespective of whether S. aureus DNA (triangle) or S. epidermidis DNA (square) is used as the starting template. TABLE 3 | Detection limit of the TaqMan assay with or without prefiltration TABLE 4 | Sensitivity and specificity of the multiprobing PCR assay TABLE 5 | Comparison of multiprobing PCR results in single versus dual infection DISCUSSION : Current methods of universal detection with species identification include PCR amplification with a universal primer set followed by performance of species identification assays, such as oligonucleotide array, restriction digestion, or sequencing . Another variation has been to universally amplify cultured clinical samples and then subject the amplified product to hybridization using different sets of specific probes . Regardless of the methodology, virtually all techniques for universal detection and species identification of bacteria have thus far involved at least two sequential steps. With the probe-based PCR system described here, both steps can for the first time be accomplished simultaneously. The probe-based PCR system we have devised is comprised of a universal primer set, a universal probe, and a species-specific probe. In this way, detection of amplification and extraction of sequence information from amplicons can be performed within the confines of the PCR run, eliminating the need for post-PCR manipulations. This innovation reduces overall assay time to about 2 h or less, depending on the PCR instrument used, while conserving the sensitivity and specificity of the assay. The only present limitation of this system is inherent in the number of fluorophores commercially available and the discriminatory power of the detection instrument itself, which presently can simultaneously differentiate up to four different fluorophores in a single tube . Thus, the number of species-specific probes which may be included in an individual reaction (in addition to the universal probe) is restricted. Simultaneous detection and/or species identification of microorganisms in a given sample has been reported recently with the multiplexing technique, with multiple sets of species-specific primer pairs and probes corresponding to different amplification targets . Our novel PCR design in which multiple probes, including a universally conserved one, exploit regions within the same amplification target does confer several advantages over multiplexing. First, possible competition between multiple PCR primer pairs is avoided with multiprobing, which involves only a single pair of primers. Moreover, when technologies advance to allow for more fluorophores, expansion of the multiprobe system for detection of other templates will require only the addition of another probe. This is in contrast to multiplexing, in which both new primers and probes will have to be added to the reaction mixture. In that circumstance, the new primers may not amplify under the original PCR conditions, which will thus require optimization studies to accommodate all the primer pairs. Finally, even if suitable conditions are ultimately attained, amplification efficiencies may be altered such that standard curves for starting template quantification will need to be rederived. Since multiprobing involves only a single target region, quantification of starting template will in principle rely on only one standard curve, if all probes are designed with a similar melting temperature. With regard to contamination, residual bacterial DNA from various sources has historically prevented widespread use of universal primer sets in PCR-based assays. In our experience, pretreatment of PCR reagents with restriction enzymes followed by heat inactivation did eliminate amplification in the negative controls. However, CT values for positive controls from runs with pretreated reagents were consistently greater than those without pretreatment (data not shown). Whether these findings reflected residual restriction enzyme activity even after heat inactivation, or systemic inhibition of the PCR system by the addition of restriction enzyme, is unknown. The contamination problem was eventually resolved by passing PCR reagents through Microcon YM-100 centrifugal filter devices (Millipore Corporation). Of note, these filters allow decontamination of all PCR reagents, including UNG, Taq polymerase, primers, and probes, which was not possible using other methods, such as DNase treatment. Although Centricons have been employed for decontamination purposes in the past, heretofore their adequacy in the context of real-time PCR systems had not been studied . With prefiltration, the PCR efficiency of the system was not reduced. In addition, with significant reduction in background contamination, the prefiltration step improved the minimum detection limit of the assay from 5,000 to 50 fg of S. aureus DNA. The mean CT of the negative control was 40. In our experience, the CT values of negative controls, although consistently above 35, were variable. These results were not unexpected, since greater sampling errors are encountered at low starting template concentrations . Finally, although the probes used in our real-time PCR system were empirically found to be specific for the organisms tested, there exists the theoretical possibility that the assay will yield a false-positive result when applied more broadly. This is an inherent limitation of design, based on the extent of the sequence search carried out, which in turn is dictated by the intended clinical application of the assay. As such, future studies may require more exhaustive bioinformatic analysis followed by even broader clinical validation of specificity. Nevertheless, the concept of multiprobing in a genomic region comprised of conserved and highly variable stretches remains a valid technique. In conclusion, we have devised a rapid, highly sensitive and specific molecular assay which allows for the simultaneous detection, quantification, and species identification of bacterial organisms. Further development in the technology of species-specific probes will theoretically expand the number of organisms which can be identified using this method, and future research will address this. In principle, the methodology described here could be coupled with DNA extraction protocols using clinical samples derived from any bodily fluid or tissue and thereby provide an assay for diagnosing numerous diseases of infectious etiology. Such an addition to the clinicians' diagnostic armamentarium will better equip physicians to carry out expedited risk assessment and targeted antibiotic treatment. Backmatter: PMID- 12202581 TI - Clinical and Epidemiological Correlates of Genotypes within the Mycobacterium avium Complex Defined by Restriction and Sequence Analysis of hsp65 AB - Species identification of isolates of the Mycobacterium avium complex (MAC) remains a difficult task. Although M. avium and Mycobacterium intracellulare can be identified with expensive, commercially available probes, many MAC isolates remain unresolved, including those representing Mycobacterium lentiflavum as well as other potentially undefined species. PCR restriction analysis (PRA) of the hsp65 gene has been proposed as a rapid and inexpensive approach. We applied PRA to 278 MAC isolates, including 126 from blood of human immunodeficiency virus (HIV)-infected patients, 59 from sputum of HIV-negative patients with chronic obstructive pulmonary disease, 88 from environmental sources, and 5 pulmonary isolates from a different study. A total of 15 different PRA patterns were observed. For 27 representative isolates, a 441-bp fragment of the hsp65 gene was sequenced; based on 54 polymorphic sites, 18 different alleles were defined, including 12 alleles not previously reported. Species and phylogenetic relationships were more accurately defined by sequencing than by PRA or commercial probe. The distribution of PRA types and, by implication, phylogenetic lineages among blood isolates was significantly different from that for pulmonary and environmental isolates, suggesting that particular lineages have appreciably greater virulence and invasive potential. Keywords: Introduction : Species of Mycobacterium, sole genus of the family Mycobacteriaceae, are aerobic, nonmotile, acid-fast bacilli with surprisingly diverse phenotypes related to growth rate, metabolic activity, colony appearance, environmental distribution, and pathogenic potential for eukaryotic hosts. The Mycobacterium avium complex (MAC) comprises slow-growing mycobacteria that are ubiquitous in the environment (soil and water) and capable of infecting diverse species including birds, pigs, and humans, with consequences ranging from asymptomatic infection to clinically significant and even fatal disease. In humans, MAC bacteria have been associated with self-limited cervical lymphadenitis in children, progressive lung infection in patients with chronic obstructive pulmonary disease, and fatal, disseminated infection in AIDS patients . The environmental sources responsible for MAC infection in different populations, the routes of transmission, and the potential for latent infection and reactivation of disease are still incompletely defined. Prior studies have implicated potable water from institutional recirculating hot water systems as one source of acute, disseminated infection in human immunodeficiency virus (HIV)-infected patients . Studies of the epidemiology and pathogenesis of MAC bacteria are complicated by their considerable phenotypic and genotypic diversity. Although several genotyping systems have been proposed, no unique species biomarkers have been identified, and any two genotypic assays may yield conflicting results. Incomplete species identification is not uncommon, particularly for environmental isolates. Currently available methods for species identification include DNA-rRNA hybridization with commercial probes (Accuprobe; GenProbe Inc., San Diego, Calif.) ; analysis of restriction polymorphisms in digests of a PCR-amplified fragment of hsp65 (PCR restriction analysis [PRA]) ; and nucleotide sequencing of the 16S rRNA , the 16S-23S ribosomal DNA internal transcribed spacer sequence , or a 360-bp DNA fragment of the hsp65 gene . The commercial probes, which are currently the most widely used approach for characterizing clinical isolates of MAC, can identify M. avium and Mycobacterium intracellulare. Isolates that react with the complex-specific probe, but with neither of the species-specific probes, are designated "MAC-other" and may represent newly recognized species, such as Mycobacterium lentiflavum , or still-undefined species. Some reports have concluded that further taxonomic clarification of the MAC cluster is needed, with formal identification of new species and possibly redefinition of the boundaries of the cluster itself . During our ongoing clinical and epidemiological studies of MAC, we have applied the hsp65 PRA method of Telenti et al. to over 1,500 isolates of nontuberculous mycobacteria (data not shown). In the course of these studies we identified numerous isolates phenotypically consistent with MAC but with PRA patterns not described by Telenti and colleagues. Moreover, the relative frequencies of these PRA patterns differed among isolates from different clinical and environmental sources. To clarify the species of these isolates and their phylogenetic relationships, we applied several different techniques including Accuprobe hybridization, nucleotide sequencing of the hsp65 gene , and IS1245 hybridization . The results demonstrate extraordinary phylogenetic diversity among isolates within MAC and suggest that there are important genotypic differences in environmental distribution and virulence potential. MATERIALS AND METHODS : Mycobacterial isolates. | Three classes of isolates ---blood, pulmonary, and environmental ---were included in these studies. (i) Blood isolates. | We selected 126 of 1,872 available blood isolates (920A and 920B series) from AIDS patients with disseminated MAC infection . The isolates were chosen to be diverse with respect to multiple factors, including clonality of infection (monoclonal, 37; polyclonal, 22; undefined, 67) (C. F. von Reyn, R. D. Arbeit, T. Barber, R. Brindle, A. Ranki, J. O. Falkinham III, G. O. O'Connor, and the International MAC Study Group, VIIIth Int. Conf. AIDS/IIIrd STD World Congr., p. 145, 1992), geographic origin (Africa, 4; Brazil, 19; Finland, 20; Trinidad, 3; New Hampshire, 28; Massachusetts, 38; Georgia, 14), and date of isolation (1991 to 1997). (ii) Pulmonary isolates. | The pulmonary isolates included 59 consecutive isolates cultured from sputum or bronchoalveolar lavage specimens from men with chronic obstructive pulmonary disease treated at the VA Boston Healthcare System (121B series). Only one patient had a clinical course compatible with chronic MAC infection: the remaining isolates were considered to represent respiratory tract colonization. Five additional pulmonary isolates generously provided by H. Soini et al. (132B series) had unique and/or unusual hsp65 PRA patterns and were included in the sequence studies reported here. (iii) Environmental isolates. | We selected 88 of 4,524 isolates (953A series) cultured from 568 environmental samples (R. D. Arbeit and C. F. von Reyn, unpublished data). These isolates were diverse with respect to geographic origin (New Hampshire, 21; Massachusetts, 43; Washington, 11; Georgia, 9; Florida, 4), source (hot potable water, 40; cold potable water, 47; soil, 1), and date of isolation (1993 and 1994). Species identification by commercial probes. | DNA-rRNA hybridization with commercial probes specific for MAC, M. avium, or M. intracellulare (Accuprobe) was performed per the manufacturer's directions. PCR amplification of hsp65. | To prepare DNA lysates, isolates were grown on 7H10 agar and single colonies were picked, resuspended in 50 mul of sterile water, and boiled for 30 min. Lysates were stored at -70C. PCR mixtures (50 mul) were prepared with a PCR master mix containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 10% glycerol, 200 muM (each) deoxynucleoside triphosphates, 0.5 muM (each) primer, and 1.25 U of AmpliTaq Gold polymerase (Perkin-Elmer, Foster City, Calif.). Samples were amplified as described previously . The primers TB11 (5'-ACCAACGATGGTGTGTCCAT) and TB12 (5'-CTTGTCGAACCGCATACCCT) amplify a 441-bp PCR fragment from the 65-kDa heat shock protein gene . The PCR product specificity was checked by electrophoresis on a 1% agarose gel for 1 h at 100 V. Restriction enzyme analysis of the hsp65 PCR product. | Two separate restriction digests (BstEII and HaeIII) were performed for each isolate . For each restriction enzyme, 10 mul of each PCR mixture was digested. All buffers and restriction enzymes were purchased from New England Biolabs (Beverly, Mass.). Following digestion, samples were electrophoresed on a 1.5-mm 10% mini-polyacrylamide gel electrophoresis (mini-PAGE) gel (Bio-Rad, Hercules, Calif.) for 70 min at 120 V. Fragments were visualized by ethidium bromide staining. Sequence analysis of the hsp65 PCR product. | PCR products were purified with the Wizard PCR Prep system (Promega, Madison, Wis.); both forward and reverse strands were then sequenced on an ABI sequencer (Core Sequencing Laboratory, Department of Medicine, Boston University School of Medicine). To perform dendrogram analysis, maximum parsimony and maximum likelihood analyses were compared with the program PAUP* 4.0 beta2a (Sinauer Associates Inc., Sunderland, Mass.). The final tree was constructed by maximum parsimony (rooted, branch-and-bound) and was topographically identical to one drawn by maximum likelihood analysis. TreeViewPPC version 1.5.3 was used to draw and label the tree. IS1245 Southern blot analysis. | Genomic mycobacterial DNA was isolated as previously described and digested with PvuII (New England Biolabs). Digested DNA was electrophoresed in an 0.8% gel for 20 h at 20 V and vacuum transferred to a GeneScreen Plus nylon membrane (NEN Life Science Products, Boston, Mass.). A digoxigenin (DIG)-labeled IS1245-specific probe was prepared with the PCR DIG probe synthesis kit (Boehringer Mannheim Corp., Indianapolis, Ind.) by using a clinical M. avium isolate with similar20 copies of IS1245 . Blots were probed with the DIG-labeled probe and developed with the DIG luminescent detection kit (Boehringer Mannheim). Nucleotide sequence accession numbers. | The nucleotide sequences described herein have been deposited in the GenBank database under accession no. to and to . RESULTS : PRA. | HaeIII and BstEII digestion of the 441-bp hsp65 PCR amplicon generated fragments ranging from 30 to 441 bp. In our experience these digests were more consistently resolved by gel electrophoresis with 10% polyacrylamide (PAGE) than by that with agarose (data not shown). Although fragments as small as similar30 to 40 bp could be resolved , only fragments larger than similar50 bp were included in the visual analysis. Overall, we visually distinguished 13 distinct hsp65 PRA patterns, including 12 patterns among the 273 isolates from our laboratory, plus one additional pattern (MAC v7) among five additional isolates (132B series) described by Soini et al. . Nucleotide sequence analysis of 18 selected isolates resolved two additional patterns (MAC v5 and v9; see details below) and provided precise sizes for the restriction fragments . GenProbe hybridization studies. | A subset of 27 isolates was selected for further analysis , with one to three isolates for each distinct PRA type, including, when available, one isolate from each different source (blood, pulmonary, and environmental). These isolates were further characterized by hybridization with GenProbe reagents specific for MAC, M. avium, and M. intracellulare and by Southern blot analysis using a probe for IS1245 . The GenProbe probe specific for M. avium hybridized with eight isolates representing three related PRA types, designated Mav v1, v2, and v3. One additional pulmonary isolate (121B-60.1) of type Mav v3 hybridized with the MAC probe but, despite repeated attempts, not the M. avium probe. The GenProbe probe specific for M. intracellulare hybridized with four isolates representing two rather different PRA types, designated Min v1 and v2. There were no other isolates with these PRA types. The remaining 14 isolates in the subset represented 10 PRA types, designated MAC v1 to v10, and included 12 isolates that reacted only with the MAC probe and 2 that were negative with all three GenProbe kits. One of these isolates represented the same PRA type as did an isolate that reacted with the MAC probe; the remaining isolate had a PRA type that was unique but nevertheless appeared related to the other MAC patterns. In summary, among the 15 PRA types detected, at least two were heterogeneous with regard to probe reactivity. In addition, there were several PRA types which included M. avium or M. intracellulare isolates and were very similar to PRA types comprising only MAC-other isolates (e.g., Mav v3 and MAC v10, and Min v1 and MAC v8, respectively). IS1245 Southern blot hybridization. | All eight isolates that hybridized with the GenProbe specific for M. avium also carried IS1245, with copy numbers ranging from 2 to >28. Isolate 121B-60.1, which represented hsp65 PRA type Mav v3 but was negative with the M. avium-specific GenProbe probe, was also negative for IS1245. Nucleotide sequence analysis. | The hsp65 amplicons (441 bp) from 27 isolates, including at least one from each of the 15 PRA types, were sequenced. These data resolved several of the inconsistencies and ambiguities observed with the previous techniques. First, as noted above, sequencing defined precisely the sizes of the PRA restriction fragments . Second, sequence data resolved two additional PRA types among pairs of isolates whose fragments were indistinguishable by PAGE. One such pair was differentiated into MAC v4 and MAC v5, and the other was differentiated into MAC v9 and Min v2. In both cases the BstEII restriction fragments were confirmed as identical, but the HaeIII fragments differed by 6 to 17 bp. Third, nucleotide sequencing clarified those situations in which isolates with the same apparent PRA type were designated as different species by GenProbe. For example, isolate 121B-60.1, which had PRA type Mav v3 but was negative with the M. avium probe and lacked IS1245, had an hsp65 sequence that differed substantially from those of all the M. avium isolates. Conversely, isolate 121B-2, which had PRA type MAC v3 but was negative with the MAC probe, was, nevertheless, clearly confirmed by hsp65 sequence to be closely related to 953A-2033, the other MAC v3 isolate. The sequences were aligned by using a reference M. avium sequence (GenBank accession no. ) with the nucleotide positions labeled according to the hsp65 gene of Mycobacterium tuberculosis (GenBank accession no. ) . Overall, there were 18 different hsp65 alleles among the 27 isolates sequenced (Tables and ). Alleles were designated as "ma," "mi," or "mc" based on the GenProbe reactivity of the isolates. A total of 54 polymorphic sites were identified, including 10 nonsynonymous changes, of which 6 were previously unreported . The phylogenetic relationships among the 18 hsp65 alleles were analyzed by maximum parsimony (branch-and-bound) and rooted with M. tuberculosis (GenBank accession no. ). The resulting dendrogram demonstrated three major clusters (designated A, B, and C) . Cluster A comprised four alleles, including a tight grouping of all three alleles identified among the M. avium isolates (ma.1 to ma.3), plus one closely related MAC allele (mc.5). This cluster was distinct from the other two clusters that included the M. intracellulare allele and all the other MAC alleles. Each of the alleles in cluster A has been previously reported: ma.1 corresponds to hsp65.1 of Swanson et al. and M. avium I of Leao et al. , ma.2 corresponds to hsp65.2 and M. avium II , ma.3 corresponds to M. avium III , and mc.5 corresponds to hsp65.9 . Cluster C comprised three alleles, including two (mi.1 and mi.2) of the three alleles identified among the M. intracellulare isolates, plus one MAC allele (mc.12). Two of these three alleles have been reported previously: the mi.1 allele corresponds to hsp65.3 of Swanson et al. , and mi.2 corresponds to M. intracellulare serotype VII of Bascunana and Belak . The remaining 11 alleles (mi.3, mc.1 to mc.4, and mc.6 to mc.11) formed cluster B, which demonstrated considerable sequence diversity. The single M. intracellulare allele (mi.3) was phylogenetically indistinguishable from the MAC alleles in the cluster. None of these alleles correspond to sequences deposited in GenBank; however, two previously identified M. intracellulare isolates (GenBank accession no. and ) exhibit enough sequence similarity to map within this cluster. Distribution of PRA types among clinical and environmental isolates. | The distribution of M. avium PRA types varied appreciably among isolates from different sources . Among the blood isolates from HIV-infected patients, 69.8% were Mav v1, compared with 13.6 and 4.5% of pulmonary and environmental isolates, respectively; conversely, Mav v2 accounted for only 21.4% of the blood isolates but 40.7% of the pulmonary isolates and 63.6% of the environmental isolates (P < 0.0001, chi-square test). Approximately 13.6% of the environmental isolates, but only 4.8 and 5.1% of the blood and pulmonary isolates, respectively, represented Mav v3. Min v1 was detected for 20% of the pulmonary isolates, 14% of the environmental isolates, and only 3% of the blood isolates. Pulmonary isolates accounted for most of the remaining MAC-associated PRA types observed in this study. FIG. 1. | PAGE analysis of M. PAGE analysis of M. avium and M. intracellulare hsp65 PRA variants digested with BstEII and HaeIII. Lanes M, pBR322/MspI ladder; lanes 1 to 6, BstEII digests of Mav v1 to v3, MAC v10, and Min v1 and v2; lanes 7 to 12, corresponding digestions with HaeIII. FIG. 2. | Algorithm of MAC family variants based on restriction enzyme digestion of the 441-bp hsp65 PCR gene amplicon with BstEII and HaeIII. Algorithm of MAC family variants based on restriction enzyme digestion of the 441-bp hsp65 PCR gene amplicon with BstEII and HaeIII. The variants include 3 variants of M. avium (v1 to v3), 2 variants of M. intracellulare (v1 and v2), and 10 variants of MAC (non-M. avium, non-M. intracellulare) (v1 to v10). FIG. 3. | Nucleotide sequence alignment of 392 bp of the hsp65 gene of 15 unique alleles among 28 different MAC isolates. Nucleotide sequence alignment of 392 bp of the hsp65 gene of 15 unique alleles among 28 different MAC isolates. All 28 isolates shown in Table were sequenced; isolates designated with the same allele had identical sequences. Alleles are grouped as M. avium, M. intracellulare, and MAC-other based on reactivity with GenProbe reagents (see text for details). Nucleotide differences are noted and aligned with the reference M. avium 88-1107 sequence (GenBank accession no. ); the nucleotide sequence numbers refer to the published hsp65 sequence of M. tuberculosis (GenBank accession no. ). Boldface nucleotides represent nonsynonymous changes (n = 10). All of the HaeIII (GGCC) and BstEII (GGTNACC) restriction enzyme sites are underlined; those in the M. avium reference sequence are also labeled. FIG. 4. | Dendrogram analysis of 18 different hsp65 alleles of the MAC family based on 392 bp of the 441-bp hsp65 PCR gene product. Dendrogram analysis of 18 different hsp65 alleles of the MAC family based on 392 bp of the 441-bp hsp65 PCR gene product. Maximum-parsimony analysis using the branch-and-bound option was used to construct the tree. The tree was rooted by using M. tuberculosis (accession no. ). The bar indicates one nucleotide difference. TABLE 1 | GenProbe result, IS1245 copy number, and hsp65 allele of selected MAC isolates representing diverse hsp65 PRA types TABLE 2 | Nonsynonymous nucleotide polymorphisms among hsp65 alleles TABLE 3 | Distribution of hsp65 PRA types among MAC isolates from different sources DISCUSSION : MAC comprises a remarkably diverse set of organisms, which have proven difficult to resolve into distinct species . In this study, we examined MAC isolates specifically selected for diversity in terms of time, place, and source, including disseminated infection in AIDS, respiratory tract of patients with chronic obstructive pulmonary disease, and environment (potable water). Isolates were analyzed for hsp65 PRA type, presence of IS1245, reactivity with GenProbe reagents specific for rRNA, and nucleotide sequence variation within a portion of the hsp65 gene. These studies disclosed additional appreciable diversity among MAC isolates and identified new hsp65 alleles, including seven nonsynonymous changes. Moreover, the data indicate that the distribution of MAC genotypes among isolates causing disseminated infection is significantly different from that of environmental isolates, suggesting that the invasive organisms are highly selected, likely on the basis of still-unresolved virulence factors. Consistent with several previous reports , we found that different methods for species identification of MAC isolates did not always provide congruent results. PRA typing offers several advantages; it is more rapid than biotyping, less expensive than commercial probes, and, in theory, applicable to all species of mycobacteria. However, it is now apparent that within MAC as well as other species of nontuberculous mycobacteria, PRA types are neither unique nor specific. Three distinct PRA types have been identified among M. avium isolates . This study further documents that PRA type Mav v3 comprises both M. avium and MAC-other isolates. Overall, MAC-other isolates demonstrate a potentially bewildering array of PRA types, several of which may be readily confused with types expected to identify species outside of MAC. For example, in silico restriction digestion of hsp65 sequence data for Mycobacterium gordonae (GenBank accession no. ) generates fragments for BstEII (231, 116, and 94 bp) and HaeIII (127 and 112 bp) that would be difficult to differentiate from our MAC v2 pattern (231, 116, and 94 bp and 127 and 103 bp, respectively) by using gel electrophoresis. Nonetheless, PRA typing provided a practical method for screening large numbers of clinical, pulmonary, and environmental isolates. As expected, most clinical isolates of MAC from AIDS patients with disseminated infection represented M. avium, and consistent with the report of Leao and coworkers , the substantial majority demonstrated PRA type Mav v1. However, there were significant differences in the distribution of M. avium genotypes among isolates from different sources. Specifically, Mav v2 was the most predominant pattern among environmental isolates, and even Mav v3 was more common than Mav v1. This novel observation suggests that Mav v1 identifies a genetic lineage that is characterized by an increased propensity for causing invasive, disseminated infection. One hypothesis would be that this lineage has particular virulence factors (currently undefined) by which it is differentiated from the diverse M. avium isolates encountered in the environment. Of note, the sole MAC-other blood isolate detected in this survey had an hsp65 allele closely related to that of M. avium. The distribution of M. avium genotypes among pulmonary isolates was similar to that of the environmental isolates. However, all but one of the pulmonary isolates represented colonization rather than active infection. Further studies will be required to determine the distribution of M. avium genotypes among isolates causing localized pulmonary infection in patients without HIV infection. Although not addressed in this paper, we do have some suggestion that the same strain can be found in the blood, stool, and sputum of individual HIV-infected patients with monoclonal or polyclonal infections . We did not, however, look specifically at stool isolates in this study. In contrast to PRA, GenProbe reagents and nucleotide sequencing of hsp65 provided unambiguous results. Species identifications by those two methods were generally congruent; however, several strains demonstrated significant discrepancies that may be of clinical and biological significance. Both this study and that of Swanson et al. have identified clinical isolates that are nonreactive with the M. avium GenProbe probe but have a particular hsp65 allele (designated mc.5 and hsp65.9, respectively) that is very closely related to that of M. avium. The most likely explanation for this atypical genotype is a recombination event, but its detection in two independent studies suggests that either the particular event or this particular lineage has a selective advantage. This series also identified a strain which reacted with the M. intracellulare probe and had an hsp65 allele (mc.3) that was phylogenetically aligned with those of the MAC-other isolates. Additional genomic analyses will be needed to resolve these conflicting results in species identification. The sequence variation present within the mycobacterial hsp65 gene makes it an attractive locus for species identification and molecular phylogeny. Other mycobacterial loci that have been analyzed for these purposes include 16S rRNA, 16S-23S ribosomal DNA internal transcribed spacer sequence, and gyrB. Including the 12 unique alleles reported in this paper, over 45 hsp65 alleles have been identified within the MAC group based on sequence variation at >72 sites within 360 bp. In comparison, 77 variable sites within a 1,400-bp portion of the 16S rRNA gene have been identified among similar27 different mycobacterial species . In summary, sequence analysis of the hsp65 gene has proven to be a highly robust method for species identification of both slow- and rapid-growing mycobacteria (except within the M. tuberculosis complex) . Among MAC isolates, this approach appears to be particularly useful for identifying genetic lineages of interest for pathogenicity studies and for mapping subclusters that may delineate additional species boundaries within this cluster. Backmatter: PMID- 12202584 TI - Mycobacterium africanum Subtype II Is Associated with Two Distinct Genotypes and Is a Major Cause of Human Tuberculosis in Kampala, Uganda AB - The population structure of 234 Mycobacterium tuberculosis complex strains obtained during 1995 and 1997 from tuberculosis patients living in Kampala, Uganda (East Africa), was analyzed by routine laboratory procedures, spoligotyping, and IS6110 restriction fragment length polymorphism (RFLP) typing. According to biochemical test results, 157 isolates (67%) were classified as M. africanum subtype II (resistant to thiophen-2-carboxylic acid hydrazide), 76 isolates (32%) were classified as M. tuberculosis, and 1 isolate was classified as classical M. bovis. Spoligotyping did not lead to clear differentiation of M. tuberculosis and M. africanum, but all M. africanum subtype II isolates lacked spacers 33 to 36, differentiating them from M. africanum subtype I. Moreover, spoligotyping was not sufficient for differentiation of isolates on the strain level, since 193 (82%) were grouped into clusters. In contrast, in the IS6110-based dendrogram, M. africanum strains were clustered into two closely related strain families (Uganda I and II) and clearly separated from the M. tuberculosis isolates. A further characteristic of both M. africanum subtype II families was the absence of spoligotype spacer 40. All strains of family I also lacked spacer 43. The clustering rate obtained by the combination of spoligotyping and RFLP IS6110 analysis was similar for M. africanum and M. tuberculosis, as 46% and 49% of the respective isolates were grouped into clusters. The results presented demonstrate that M. africanum subtype II isolates from Kampala, Uganda, belong to two closely related genotypes, which may represent unique phylogenetic branches within the M. tuberculosis complex. We conclude that M. africanum subtype II is the main cause of human tuberculosis in Kampala, Uganda. Keywords: Introduction : Mycobacterium africanum is a member of the Mycobacterium tuberculosis complex, which also comprises the closely related species M. tuberculosis, M. bovis, M. microti, and M. canetti . Since its first description in 1968 , M. africanum has been found in several regions of Africa, where it represents up to 60% of clinical strains obtained from patients with pulmonary tuberculosis . Recent surveys show highly variable prevalences of M. africanum in different African regions. For example, M. africanum was found in approximately 5% of patients with tuberculosis in the Ivory Coast and in at least 60% of patients in Guinea-Bissau . Most of the studies presented so far have analyzed small numbers of strains from different regions, and systematic studies of prevalence and geographic distribution of M. africanum are still infrequent. In contrast to M. tuberculosis and M. bovis, M. africanum strains show a higher variability of phenotypic attributes, comprising characteristics common to both M. tuberculosis and M. bovis. This phenotypic heterogeneity of M. africanum complicates its unequivocal identification and may lead to misclassification of clinical strains. According to their biochemical characteristics, two major subgroups of M. africanum have been described, corresponding to their geographic origin in western (subtype I) or eastern (subtype II) Africa. Numerical analyses of biochemical characteristics revealed that M. africanum subtype I is more closely related to M. bovis, whereas subtype II more closely resembles M. tuberculosis . In our recent work, we determined diagnostic criteria, including phenotypic and biochemical characteristics as well as results of the molecular spoligotyping technique, that permit the accurate differentiation of M. africanum subtypes I and II . Spoligotyping is a rapid molecular test based on the detection of various nonrepetitive spacer sequences located between small repetitive units (direct repeats) in the direct repeat locus of M. tuberculosis complex strains. However, spoligotyping does not allow differentiation of M. africanum subtypes from M. tuberculosis without additional routine laboratory procedures. This drawback led us to evaluate the usefulness of gyrB DNA sequence polymorphisms as a further molecular marker for differentiation of the species of the M. tuberculosis complex . We established a rapid PCR-restriction fragment length polymorphism (RFLP) assay that allows the differentiation of M. bovis subsp. bovis, M. bovis subsp. caprae, and M. microti as well as the clear identification of M. africanum subtype I strains. M. africanum subtype II and M. tuberculosis, however, displayed identical gyrB DNA sequences and were indistinguishable in this analysis. Thus, differentiation of M. africanum subtype II from M. tuberculosis continues to be based on phenotypic characteristics such as growth on bromocresol purple medium . This finding, in accordance with previous reports , reiterates the close relationship between M. africanum subtype II and M. tuberculosis and questions the taxonomic status and phylogenetic position of M. africanum subtype II within the M. tuberculosis complex. The present study investigated the population structure of M. tuberculosis complex strains obtained from patients with tuberculosis who were recruited at the Mulago Hospital in Kampala, Uganda, because the presence of M. africanum subtype II in limited study populations in Uganda has been reported . The aim of this study was to assess the genetic relationship of M. tuberculosis and M. africanum subtype II in order to verify a genetic basis for this repeatedly described M. africanum subtype. Furthermore, the intention was to analyze the prevalence of M. africanum in a large study group of well-defined patients with tuberculosis. Based on the results obtained, we hypothesize that the majority of M. tuberculosis complex strains in Uganda belong to M. africanum subtype II and that this subtype contains two distinct genotypes (Uganda I and Uganda II) that may represent two closely related phylogenetic branches within the M. tuberculosis complex. MATERIALS AND METHODS : Strains analyzed. | A total of 234 M. tuberculosis complex strains isolated from sputum samples that had been collected between 1995 and 1997 were analyzed. Sputum samples were obtained from 234 adult patients with newly diagnosed initial episodes of sputum smear-positive pulmonary tuberculosis from the National Tuberculosis/Leprosy Program (NTLP) clinic (the largest tuberculosis clinic in Uganda) at Mulago Hospital. After decontamination (N-acetyl-l-cysteine-sodium hydroxide), the sediment was inoculated on Lowenstein-Jensen slants at 37C as described elsewhere . Nearly all patients were ambulatory, and a few were hospitalized. The households of these patients were subsequently studied in the context of a household contact study. All strains were confirmed as M. tuberculosis complex by spoligotyping . Biochemical tests and susceptibility testing. | Biochemical analysis for differentiation within the M. tuberculosis complex included colony morphology, nitrate reduction on modified Dubos broth, niacin accumulation test (INH test strips; Difco, Detroit, Mich.), growth in the presence of thiophen-2-carboxylic acid hydrazide (TCH, 1 mug/ml), catalase activity at room temperature, and growth characteristics on Lebek and on bromocresol purple medium, as described previously . IS6110 RFLP and spoligotyping analysis. | Extraction of DNA from mycobacterial strains and DNA fingerprinting with IS6110 as a probe was performed according to a standardized protocol described elsewhere . Spoligotypes and IS6110 fingerprint patterns of mycobacterial strains were analyzed with the Gelcompar software (Windows NT, version 4.2; Applied Maths, Kortrijk, Belgium) as described previously . Clusters were defined as groups of patients with bacterial strains showing identical spoligotype and/or IS6110 RFLP patterns. Spoligotyping of strains was performed as described by Kamerbeek et al. . Quality control. | The laboratory participated twice a year in external proficiency testing (national and international). For all test panels, negative and positive internal controls were included. PvuII-digested total DNA of reference strain Mt.14323 (obtained from the National Institute of Public Health and Environmental Protection, Bilthoven, The Netherlands) was used in each Southern blot experiment as an external size standard and for quality control and quality assurance of IS6110 RFLP experiments. M. tuberculosis H37 (ATCC 27294) and M. bovis BCG (ATCC 27289) were used as control strains in each spoligotype experiment performed. The accuracy of each experiment and the normalization procedure performed were controlled by comparing the IS6110 fingerprint patterns or spoligotype patterns of reference strains present on each autoradiogram with those stored in the database. RESULTS : In this study, M. tuberculosis complex strains from 234 patients with pulmonary tuberculosis were investigated by biochemical tests, spoligotyping, and IS6110 RFLP analysis. Species differentiation. | According to their phenotypic characteristics and biochemical test results, the 234 M. tuberculosis complex strains were classified as 157 M. africanum strains, 76 M. tuberculosis strains, and 1 M. bovis strain. M. africanum was identified on the basis of its colony morphology on Lowenstein-Jensen slants (dysgonic growth), microaerophilic growth on Lebek medium (Lebek is a semisolid medium which can be used to test the oxygen preference of mycobacterial strains), low catalase reaction, and lack of induction of a color change of bromocresol purple medium (pH-dependent change of color from blue to yellow, e.g., in the case of M. tuberculosis strains) . The biochemical characteristics of the strains analyzed and of the type strains M. tuberculosis H37 (ATCC 27294), M. bovis (ATCC 19210), and M. africanum (ATCC 25420) are summarized in Table . All strains classified as M. africanum showed resistance to TCH and were therefore differentiated as M. africanum subtype II, which more closely resembles M. tuberculosis . Spoligotyping analysis. | All strains were analyzed by spoligotyping, and the patterns obtained were digitized and analyzed for similarity with the Dice coefficient (position tolerance, 1.0%). A dendrogram was calculated with the unweighted pair group method with average linkage (UPGMA) for the M. tuberculosis and M. africanum strains. The only M. bovis strain found showed the typical absence of spacers 39 to 43 and the presence of spacers 3 to 16. This strain could thus be identified as M. bovis subsp. bovis (pyrazinamide resistant ). In accordance with previous results , all M. africanum subtype II strains showed no hybridization to the M. bovis-derived spacers 33 to 36. Although M. africanum strains were found mainly in two large groups of strains with similar spoligotype patterns, differentiation from M. tuberculosis could not be achieved in the dendrogram based on spoligotyping results . Several M. tuberculosis and M. africanum strains showed only minor differences (one or two spacers) in the spoligotype patterns (see groups on top of dendrogram in Fig. and ), which resulted in adjacent positions in the dendrogram. Considering the clustering rate among the strains analyzed, unique spoligotype patterns were found in only 40 (17%) of the 233 M. tuberculosis and M. africanum strains. The remaining 193 (83%) strains had a spoligotype pattern identical to that of one or more of the M. tuberculosis or the M. africanum strains. Among the M. tuberculosis strains, 57 (74%) were grouped into 17 clusters with identical spoligotype patterns. Each of the clusters contained between two and eight strains. Of the 157 M. africanum strains, 136 (87%) were grouped into 18 clusters, with 2 to 37 strains per cluster. This indicated high genetic homogeneity among the M. africanum strains, an observation that was further supported by the inclusion of more than 50% of all M. africanum strains within the three largest clusters, which contained 17, 29, and 37 strains . IS6110 RFLP analysis. | In order to further analyze the genetic relationship of the strains, DNA fingerprinting with IS6110 as a probe was performed. The IS6110 RFLP patterns of the M. tuberculosis and M. africanum subtype II strains were analyzed for similarity with the Dice coefficient (position tolerance, 1.3%), and a dendrogram was calculated, which is shown in Fig. . In contrast to the dendrogram that was based on the spoligotyping results , RFLP analysis grouped the M. africanum subtype II strains into two closely related genotype families (Uganda I [n = 55] and Uganda II [n = 102]). RFLP patterns among strains of these genotypes showed a similarity of at least 75% and were distinctly separated from the M. tuberculosis strains . Even though M. africanum subtype II and M. tuberculosis strains showed very similar spoligotype patterns, they could be clearly distinguished by IS6110 RFLP typing . Overall, the M. tuberculosis IS6110 RFLP patterns were more variable than those of M. africanum strains, as was depicted by large differences in the IS6110 copy numbers, ranging from only 1 to 17 per strain . In contrast, the IS6110 RFLP patterns among the M. africanum strains were more homogeneous, with copy numbers ranging between approximately 14 and 20 IS6110 bands per strain. Separate evaluation of the spoligotype patterns of strains of the M. africanum subtype II genotypes Uganda I and II showed that the absence of spacer 40 is an obvious marker of both genotypes . In addition, all strains of genotype Uganda I lack spacer 43. In contrast to these findings in M. africanum subtype II, most of the M. tuberculosis strains (44 of 57) showed hybridization to spacers 40 and 43 (data not shown). All of the M. tuberculosis strains, which lack one or both of spacers 40 and 43, were clearly distinguishable from the M. africanum strains by their IS6110 RFLP patterns. This further confirmed our species differentiation based on phenotypic and biochemical characteristics. Thus, lack of spacers 40 and 43 is not an exclusive marker of M. africanum subtype II but might represent a useful additional criterion for M. africanum subtype identification in combination with biochemical test results. When the results from spoligotyping and IS6110 RFLP analysis were combined, rates of strains in clusters with identical spoligotype and IS6110 RFLP patterns were reduced to 47% (110 of 234). Among the 76 M. tuberculosis strains in this study, 37 strains (49%) showed identical IS6110 and spoligotype patterns and were grouped into 13 clusters containing two to seven strains each. Among the 157 M. africanum strains, 72 (46%) were grouped in 28 such clusters with two to seven strains per cluster that consisted mainly (75%) of pairs of strains. Although the fingerprint polymorphism detected by spoligotyping was lower than that of IS6110 RFLP typing, an overall correlation between the two techniques was observed. All strains with identical IS6110 RFLP patterns also displayed identical or very similar spoligotype patterns (data not shown), confirming the genetic relationship of the strains determined by IS6110 RFLP typing. The accurate classification of the M. africanum subtype II genotypes Uganda I and II by IS6110 RFLP typing was further supported by the shared characteristic spoligotype features of the strains. FIG. 1. | Spoligotype patterns of the 233 M. tuberculosis Spoligotype patterns of the 233 M. tuberculosis (darker shading) and M. africanum subtype II (lighter shading) strains. Banding patterns are ordered by similarity in a dendrogram. The position of each spoligotyping hybridization spot is normalized so that banding patterns of all strains are mutually comparable. The scale depicts similarity of patterns calculated with the Dice coefficient and the UPGMA method. FIG. 2. | Spoligotype (a) and IS6110 RFLP (b) patterns of four pairs of M. tuberculosis Spoligotype (a) and IS6110 RFLP (b) patterns of four pairs of M. tuberculosis and M. africanum subtype II strains. M. tuberculosis and M. africanum subtype II strains had very similar spoligotype patterns but were clearly separated by IS6110 RFLP typing. FIG. 3. | IS6110 DNA fingerprint patterns of the 233 M. tuberculosis IS6110 DNA fingerprint patterns of the 233 M. tuberculosis (darker shading) and M. africanum subtype II (lighter shading) strains. Banding patterns are ordered by similarity in a dendrogram. M. africanum subtype II strains were clustered in two closely related strain families (genotypes Uganda I and II) and were clearly separated from the M. tuberculosis strains. FIG. 4. | Representative spoligotype patterns of M. africanum Representative spoligotype patterns of M. africanum subtype II strains of genotypes Uganda I and II (C and D) compared to spoligotype patterns of type strains M. tuberculosis H37 (ATCC 27294), M. bovis (ATCC 19210), M. bovis BCG (ATCC 27289), M. africanum (ATCC 25420), and a collection of M. africanum subtype I (A) and M. africanum subtype II (B) isolates from our previous work . In contrast to M. bovis, all M. africanum strains showed hybridization to several of the spacers 39 to 43 which were derived from the direct repeat (DR) region of M. tuberculosis H37. In the case of M. africanum subtype II, no hybridization was observed to the M. bovis BCG-derived spacers 33 to 36, whereas M. africanum subtype I isolates as well as the M. africanum type strain (ATCC 25420) showed hybridization to at least two of these spacers. All M. africanum subtype II strains showed a characteristic lack of hybridization to spacer 40. Strains of genotype Uganda I lack spacer 43 in addition (arrows). In contrast, M. africanum subtype I strains lack spacer 39. TABLE 1 | Biochemical characteristics of type strains M. tuberculosis H37 (ATCC 27294), M. bovis (ATCC 19210), and M. africanum (ATCC 25420) and the strains analyzed DISCUSSION : This study systematically analyzed the population structure of M. tuberculosis complex strains isolated between 1995 and 1997 from tuberculosis patients living in Kampala, Uganda. Sixty-seven percent of the strains were M. africanum subtype II, suggesting that the main cause of human tuberculosis in Kampala is M. africanum subtype II. We further demonstrated that M. africanum subtype II strains from Kampala, Uganda, belong to two closely related genotypes (Uganda I and II) that share specific spoligotyping characteristics and are clustered into two IS6110 RFLP strain families. Geographic variants of M. africanum had initially been described in studies by David et al. and were more recently noted by Haas et al. . The results of these studies indicated that M. africanum subtype I predominantly originated from West African countries, whereas M. africanum subtype II was found predominantly in East Africa. Systematic studies analyzing larger numbers of strains from one study region are still rare, and their interpretation is complicated by the lack of clear characteristics for the differentiation of M. africanum and its two subtypes. In our own studies , we analyzed a collection of M. africanum strains from western and eastern African countries and found criteria which allowed the accurate differentiation of the two M. africanum subtypes in accordance with the geographic origin of the strains. The main criteria for the differentiation of the two M. africanum subtypes are susceptibility to TCH, hybridization to at least two of the M. bovis-derived spacers 33 to 36, and a specific gyrB DNA sequence for subtype I and resistance to TCH and lack of hybridization to spacers 33 to 36 for subtype II. Recent studies in West African countries have shown M. africanum prevalence rates with high regional variability, ranging from approximately 5% in the Ivory Coast to 61% (biovars 2, 3, and 4) in Guinea-Bissau . Because of their susceptibility to TCH, the M. africanum strains in these two studies were confirmed as M. africanum subtype I. Considering the IS6110 RFLP patterns obtained, an obvious characteristic of the M. africanum subtype I strains in both studies was the presence of an intermediate or small number of IS6110 bands, which has also been observed by Haas et al. for M. africanum subtype I. In accordance with our previous results , the spoligotyping analysis performed by Kallenius et al. confirmed that M. africanum subtype I strains are characterized by a specific spoligotype pattern which is intermediate between those of M. bovis and M. tuberculosis (hybridization to spacers 33 to 36 as well as to spacers 39 to 43). This typical genotype, the combination of an intermediate spoligotype pattern together with a small number of IS6110 bands, was further observed in a very recent study by Viana-Niero and coworkers , who analyzed a collection of M. africanum strains from several West African countries. All these studies verify the presence of M. africanum subtype I in West Africa, which is characterized by certain phenotypic properties as well as a characteristic spoligotype and IS6110 RFLP patterns. Our previous results indicate that this subtype may be identified by a specific gyrB DNA sequence, but this finding still remains to be analyzed for a larger number of M. africanum subtype I strains. In accordance with our preliminary observations during a study of 49 M. tuberculosis complex strains from Kampala , the present study confirms that M. africanum, and particularly its subtype II, represents a major cause of human tuberculosis in this African region. This finding is in contrast to the results obtained in a study performed from 1992 to 1993 in the region of Buluba, Uganda, in which only 16% of the strains analyzed were differentiated as M. africanum . These contrasting results may be simply explained by a variable prevalence of M. africanum subtype II in different regions of Uganda or differences in the sampling procedures applied. A further possible reason is an increase in the prevalence of M. africanum subtype II in recent years, which might have resulted from other contributing factors, such as the increased rate of human immunodeficiency virus type I (HIV-1) in the Ugandan population. In contrast to M. africanum subtype I, subtype II strains were resistant to TCH and showed no hybridization to spoligotype spacers 33 to 36. The most striking finding of this investigation is that the M. africanum subtype II strains from Kampala, Uganda, clustered in two closely related genotypes, which could be clearly separated from the M. tuberculosis strains analyzed by their RFLP pattern. Within both subtype II genotypes, the strains showed very homogenous IS6110 RFLP patterns, but with a large number of IS6110 copies per strain (approximately 16 to approximately 20), clearly differentiating these strains from M. africanum subtype I. A further characteristic of genotypes Uganda I and II is the absence of spoligotype spacer 40 and also the absence of spacer 43 in strains of Uganda I. These results indicate that the strains of these two genotypes are closely related and may have diverged from an M. tuberculosis-like ancestor. In contrast to the homogenous IS6110 RFLP patterns observed for the M. africanum subtype II strains, M. tuberculosis strains from Kampala showed a high variability of IS6110 banding patterns as well of IS6110 copy numbers. One can speculate that the differences in homogeneity patterns between the M. africanum subtype II strains and the M. tuberculosis strains result from closely related indigenous mycobacterial populations in the region of Kampala and a high degree of influx from abroad resulting in highly diverse IS6110 RFLP patterns, respectively. In accordance, Daniel presented an interesting study on the early history of tuberculosis in central Africa, which demonstrates that tuberculosis was present in central East Africa at the time of the earliest European entries in the region of Kampala. The clustering rate obtained by the combination of spoligotyping and IS6110 RFLP analysis was similar for M. tuberculosis and M. africanum subtype II (46% and 49%, respectively) and indicates a high rate of recent human-to-human transmission for strains of both species. Similar clustering rates have recently been measured by IS6110 typing in other African countries such as Botswana (42% ), Namibia (47% ), and South Africa (45% ). Only slightly lower or comparable clustering rates have been reported from other areas of the world with a low incidence of tuberculosis, such as New York (37% ) and The Netherlands (47% ). This somewhat surprising observation may be due to short study periods or limited numbers of patients with pulmonary tuberculosis that were analyzed in these studies. Considering the discriminatory power of both typing methods, the results in this study clearly indicate that spoligotyping alone is not well suited for differentiation of M. tuberculosis complex strains on the strain level in this high-incidence community. Also, spoligotyping did not facilitate an accurate analysis of the genetic relationship of the strains, as M. tuberculosis and M. africanum strains with similar spoligotype patterns were clearly separated by their IS6110 RFLP patterns and biochemical characteristics. In contrast to IS6110 RFLP patterns, for which modifications appear to occur by changes of single bands as a function of time , large alterations of spoligotype patterns seem to be possible in relatively short time periods. Alterations of spoligotype patterns thus do not necessarily represent the overall rate of change of the genome. Hence, spoligotyping appears not to be a useful method for determination of the genomic relatedness of M. tuberculosis complex strains for phylogenetic purposes. In conclusion, the results presented here and in earlier studies clearly confirm the existence of M. africanum subtype I (West Africa) and subtype II (East Africa, Uganda), which have been previously proposed by numerical analysis of the phenotypic characteristics. M. africanum subtype I and M. africanum subtype II represent two unique phylogenetic branches within the M. tuberculosis complex that originates in West and East Africa, respectively. Both M. africanum subtypes have been verified to represent a high portion of M. tuberculosis complex strains in certain regions of Africa, as we confirmed that more than 60% of the tuberculosis cases in Kampala are due to M. africanum subtype II and not to M. tuberculosis. A high prevalence of M. africanum strains in human tuberculosis in Africa might have important implications for tuberculosis control, considering the enormous burden of tuberculosis and HIV-1/AIDS in Africa. Based on the clustering rates observed in our study, no difference in transmission patterns between M. africanum subtype II and M. tuberculosis could be verified. A preliminary result obtained by analyzing 13 patients indicated that presentations and responses to short-course chemotherapy are comparable for M. africanum and M. tuberculosis . A more detailed analysis of the clinical presentation, therapy outcome, and epidemiological characteristics of more than 300 cases of M. africanum- and M. tuberculosis-induced tuberculosis that includes the strains presented in this study is in preparation. Further studies in larger study populations will be needed for more detailed analyses of the regional prevalence and transmission of M. africanum-induced tuberculosis, especially in the context of factors such as coinfection with HIV-1. Backmatter: PMID- 12202578 TI - Fast, Noninvasive Method for Molecular Detection and Differentiation of Malassezia Yeast Species on Human Skin and Application of the Method to Dandruff Microbiology AB - Malassezia fungi have been the suspected cause of dandruff for more than a century. Previously referred to as Pityrosporum ovale, Pityrosporum orbiculare, or Malassezia, these fungi are now known to consist of at least seven Malassezia species. Each species has a specific ecological niche, as well as specific biochemical and genetic characteristics. Malassezia yeasts have fastidious culture conditions and exceedingly different growth rates. Therefore, the results of surveys of Malassezia based on culture methods can be difficult to interpret. We developed a molecular technique, terminal fragment length polymorphism analysis, to more accurately survey the ecology of Malassezia yeasts without bias from culture. This technique involves fluorescent nested PCR of the intergenic transcribed spacer (ITS) ITS I and ITS II region ribosomal gene clusters. All known Malassezia species can be differentiated by unique ITS fragment lengths. We have used this technique to directly analyze scalp samples from subjects enrolled in a demographic scalp health study. Results for subjects assigned composite adherent scalp flaking scores (ASFS) <10 were compared to those for subjects assigned composite ASFS >24. Malassezia restricta and M. globosa were found to be the predominant Malassezia species present in both groups. Importantly, we found no evidence of M. furfur in either group, indicating that M. furfur can be eliminated as the causal organism for dandruff. Both groups also showed the presence of non-Malassezia fungi. This method, particularly when it is used in combination with existing fungal ITS databases, is expected to be useful in the diagnosis of multiple other fungal infections. Keywords: Introduction : Recently, members of the genus Malassezia have become viewed as opportunistic yeasts of increasing importance . They are lipophilic or lipid-dependent yeasts, and at least some belong to the normal cutaneous microflora. Some Malassezia species may act as pathogens when exposed to certain changes in the skin microclimate. For decades the genus Malassezia remained limited to two species, namely, the lipid-dependent Malassezia furfur and the lipophilic M. pachydermatis. In 1995, 28S rRNA gene sequences revealed seven distinct genetic entities , which are now accepted as species (M. furfur, M. obtusa, M. globosa, M. slooffiae, M. sympodialis, M. pachydermatis, and M. restricta) . Malassezia species are exceptionally difficult to cultivate, so additional species may be discovered as DNA-based differentiation techniques are refined and applied to multiple ecosystems. While several of the seven described Malassezia species have been associated with human infection, the pathological role of each species is not fully understood. For example, M. furfur infections have been observed in hospitalized neonates with very low birth weights receiving intravenous lipid emulsions . M. globosa, which corresponds to the original description of Pityrosporum orbiculare and correlates to the former serovar B of M. furfur , may be the most important species in pityriasis versicolor, either alone or in association with other species, such as M. sympodialis . M. restricta, which corresponds to the former serovar C of M. furfur and which visually resembles Pityrosporum ovale, is the species most often associated with seborrheic dermatitis and dandruff. M. pachydermatis, the non-lipid-dependent species, is rarely observed in humans but has been found to cause septic outbreaks . Understanding the clinical role of the individual species has been hampered by the difficulty involved with isolation, cultivation, and identification. Cultivation requirements vary by species . M. furfur is by far the most robust of the Malassezia species in culture and therefore is the organism most frequently isolated. We have found that M. restricta and M. obtusa are the most difficult species to grow in culture. In addition to specific nutrient requirements, we have also found that a constant temperature of 34C is required for growth of M. restricta. Both M. globosa and M. restricta grow much more slowly than M. furfur in culture and would be quickly overwhelmed by any M. furfur present, even if there was initially a much smaller number of M. furfur cells present in the sample. Several approaches have been used to routinely identify Malassezia species. These include determination of the mole percent guanine-plus-cytosine content, DNA reassociation values, cell morphology, growth with different Tween nonionic detergents as the sole lipid supplement, the presence of catalase, temperature requirements , the presence of beta-glucosidase revealed by the splitting of esculin, and selective growth with cremophor EL . In addition, some attempts have been made to use specific molecular methods for the identification of Malassezia isolates, such as pulsed-field gel electrophoresis, randomly amplified polymorphic DNA analysis, sequencing analysis, restriction analysis of PCR amplicons of ribosomal sequences, amplified fragment length polymorphism analysis, and denaturing gradient gel electrophoresis (-, , , , , ; T. Boekhout and B. Theelen, abstract from the 20th International Conference on Yeast Genetics and Molecular Biology 2001, Yeast 18:S332, 2001). While these approaches have met with various degrees of success, most are not well suited for the analysis of complex clinical samples. In addition, all of the methods mentioned above (except denaturing gradient gel electrophoresis) require cultivation to enhance sensitivity, thereby increasing both the potential for culture bias and the turnaround time for analysis. Two methods have been reported to differentiate complex Malassezia communities from skin without cultivation , but these methods require either separate amplification with specific primer sets for each species or restriction digestion. The terminal fragment length polymorphism (tFLP) method uses only three different primer sets, minimizing the potential bias related to amplification efficiency. Because it has been documented that the efficacies of antifungal drugs can vary depending upon the species , timely clinical assessments are critical for the prompt administration of the appropriate therapy, especially when Malassezia yeasts are responsible for nosocomial bloodstream infections . The purpose of the work described here was to develop a specific and highly sensitive molecular method suitable for the rapid and reliable identification of Malassezia species from very small clinical samples. A key objective was to increase the sensitivity of the method to eliminate the need for cultivation and thereby increase the detection rate and eliminate cultural bias in the results. Eliminating the need for cultivation and restriction digestion would also significantly reduce the turnaround time for analysis by at least 4 to 5 days, the typical cultivation time selected. MATERIALS AND METHODS : Preparation of standards and clinical samples. | The Malassezia fungal strains used as standards in this study are listed in Table . These strains were selected for use based on the following rationale. First, two isolates of each Malassezia species known to have been isolated from human scalps were selected. Other common isolates of phylogenetic importance (based on observation of multiple isolates in the laboratories of the Centraalbureau voor Schimmelcultures [CBS], Utrecht, The Netherlands) were included so that representative isolates of the major genotypic groups were considered. Standards were maintained on Leeming and Notman medium (1% [wt/vol] peptone, 0.5% [wt/vol] glucose, 0.01% [wt/vol] yeast extract, 0.4% [wt/vol] desiccated ox bile, 0.1% [vol/vol] glycerol, 0.05% [wt/vol] glycerol monostearate, 0.05% [vol/vol] Tween 60, 1% [vol/vol] high-fat cow's milk, and 1.5% [wt/vol] agar in distilled water) at 30C (or 34C for M. restricta) or were stored at -80C . Stock cultures were stored in modified Dixon liquid medium (3.6% [wt/vol] malt extract, 0.6% [wt/vol] peptone, 2.0% [wt/vol] oxgall, 1.0% [vol/vol] Tween 40, 0.2% [vol/vol] glycerol, and 0.2% [vol/vol] oleic acid in distilled water) with 25% (wt/vol) glycerol at -80C . Standards cultures were supplied by CBS . Serial dilution of standards for determination of sensitivity were prepared by counting the cells on a Coulter Counter and preparing dilutions of known concentration with Dulbecco's phosphate-buffered saline (Invitrogen Corp., Carlsbad, Calif.). Concentrations are reported as the number of cells per milliliter of dosing solution. Swabs were prepared by dosing 50 mul of the counted cell suspension directly onto the rayon tip. Extraction, sample preparation, and PCR were then carried out in the same manner used for the clinical samples. Samples were collected from human subjects enrolled in the U.S. portion of a demographic scalp health study after being graded for scalp flaking severity. Scalp flaking severity was graded on the basis of the adherent scalp flaking scale (ASFS) . By this grading approach, an expert grader, who was also a licensed dermatologist in this case, assigned a numerical grade between 0 and 10 (with 10 representing the most flaking and 0 representing the least flaking) to eight divisions covering the scalp to obtain a composite dandruff score ranging from 0 to 80. Composite ASFSs of less than 10 can be considered normal scalps to scalps with low levels of flaking, whereas composite scores of greater than 24 can be considered scalps with high levels of flaking associated with severe dandruff or seborrheic dermatitis. We hereafter use the term "dandruff" in this report to include subjects with high levels of flaking diagnosed with dandruff or seborrheic dermatitis. Samples from human scalps were collected by rubbing rayon swabs (plain swab, sterile; Copan Diagnostics, Corona, Calif.) back and forth over a 1-in stroke area for 20 strokes while continuously rotating the swab. All samples were collected from human subjects in accordance with federal guidelines and institutional policies. Extraction of Malassezia DNA from swabs. | The DNA extraction procedure described here was optimized in our laboratory for the extraction of Malassezia DNA. All standard, sample, and control swabs were placed in 15-ml conical tubes containing 0.6 ml of 0.01% sodium dodecyl sulfate (diluted from 10% sodium dodecyl sulfate solution [Invitrogen Corp.] with distilled, deionized water) and vortexed (Minivortexer; VWR International, West Chester, Pa.) at medium to high speed for 20 s. The swabs were removed from the tubes and discarded. The samples were then transferred to a 1.5-ml screw-cap tube. One half of the tube cone was filled with 0.5-mm zirconia and silica beads (Biospec Products Inc., Bartlesville, Okla.), and the samples were beaten with the beads at similar75% maximum speed (Minibeadbeater-8; Biospec Products) for 10 s. The samples were allowed to cool for 10 s. This bead beating and cooling procedure was repeated four times for a total of five times to break open the cell walls. A total of 400 mul of a phenol-chloroform-isoamyl alcohol solution (25:24:1 [vol/vol/vol]; Invitrogen Corp.) was then added, and the mixture was vortexed on high for 30 s, followed by centrifugation (model 5415C; Eppendorf, Hamburg, Germany) at 14,000 x g for 10 min. A total of 350 mul of the aqueous phase was removed and placed in a new tube. To this extract was added 35 mul of 3 N sodium acetate (NaOAc; pH 5.2; ISC Bioexpress, Kaysville, Utah). The mixture was vortexed for 2 s, and 0.5 mul of glycogen (20 mug/mul; Invitrogen Corp.) and 963.75 mul of ice-cold 100% ethanol (EtOH; 200 proof; AAPER Alcohol, Shelbyville, Ky.) were added. The volumes of NaOAc and EtOH used were adjusted depending upon the actual amount of the aqueous phase recovered (for NaOAc, 10% of the volume recovered; for EtOH, 2.5 times the total volume of the aqueous phase plus NaOAc and glycogen). The mixture was then vortexed for 5 s and stored at -20C for 3 h or overnight. The samples were then precipitated by centrifugation at 14,000 x g at 4C for 10 min. The supernatant was aspirated, and the pellet and the interior of the tube were washed by addition of 1 ml of 70% ice-cold EtOH (dilution of 200 proof EtOH in distilled, deionized H2O; AAPER Alcohol). Centrifugation and aspiration were repeated, followed by speed vacuum drying (model SC110; Thermo Savant, Holbrook, N.Y.) for 5 min at a medium drying temperature. The pellet was resuspended in 20 mul of 1x TE (1 mM Tris-HCl [1 M; pH 8; Sigma-Aldrich Corporation, Sigma, St. Louis, Mo.], 0.1 mM EDTA [0.5 M; pH 8; Invitrogen Corp.]) to obtain the concentrated DNA. All samples were vortexed for 5 s and centrifuged at 14,000 x g at room temperature for 5 s prior to storage at -20C. Nested tFLP amplification of DNA extracts. | Three sets of primers were used in this work. The first set of primers (first PCR) consisted of the 18S forward primer (5'-AAC TTA AAG GAA TTG ACG GAA G-3') and the 28S reverse primer (5'-GGC AGG AAC CAG CTA CTA G-3'). The second set of primers (ITS I PCR) included the ITS I forward primer (5'-TCC GTA GGT GAA CCT GCG G-3') and the middle reverse primer (5'-TTC GCT GCG TTC TTC ATC GA-3'). The third primer set (ITS II PCR) included the middle forward primer (5'-TCG ATG AAG AAC GCA GCG AA-3') and the ITS II reverse primer (5'-TCC TCC GCT TAT TGA TAT GC-3') . Two primers were prepared and fluorescently labeled (the ITS I forward primer with a D3 label and the ITS II reverse primer with a D4 label) by Research Genetics, Inc. (Huntsville, Ala.) for subsequent fragment analysis. The fluorescent dyes and the linkage chemistry are proprietary to Beckman and are manufactured exclusively by Research Genetics. All other primers were obtained from Invitrogen Corp. The nested tFLP amplification scheme is shown in Fig. . The first step of the nested PCR process involved amplification of parts of all three ribosomal subunits including the variable regions. The first set of PCR primers (the 18S forward and 28S reverse primers) were selectively designed to span the 18S gene through the 28S gene (including both the internal transcribed space [ITS] ITS I and ITS II regions and the 5.8S gene) of fungal rRNA gene. The primers were designed to be as panfungal as possible while not amplifying human or bacterial DNA. The purpose was to increase the sensitivity and, at the same time, to minimize background interference from complex clinical samples. The second two sets of PCR primers were designed to amplify either the ITS I region (ITS I forward and middle reverse primers) or the ITS II region (middle forward and ITS II reverse primers), in which differences in lengths among Malassezia species have been observed . The ITS I forward primer is specific for the 3' region of the 18S gene, and the ITS II reverse primer is specific for the 5' region of the 28S gene. The middle forward and middle reverse primers are complementary primers targeted to the 5.8S gene . The primer locations result in fragments that contain some ribosomal gene sequences but that are short enough for accurate length analysis. The second step provides additional amplification to increase sensitivity (eliminating the need for prior cultivation of clinical samples) and produces two fragments associated with each Malassezia species. Nested PCRs. | The PCR conditions described below were carried out with extracts from standards, clinical samples, and controls. DNA (5 mul) extracted from each sample was added to 45 mul of the PCR master mixture, which consisted of 5 mul of 10x PCR buffer (Applied Biosystems Group, Applera Corporation, Foster City, Calif.), 3 mul of 25 mM MgCl2 solution (Applied Biosystems), 1 mul of a 10 muM deoxynucleotide triphosphate mixture (10 muM each dATP, dCTP, dGTP, and dTTP [Invitrogen Corp.] diluted from individual 100 mM stocks combined into deionized water), 0.35 mul of each primer (18S forward primer [0.5 mug/mul] and 28S reverse primer [0.5 mug/mul]), 0.25 mul of Taq DNA polymerase (5 U/mul; Applied Biosystems), and 35.05 mul of deionized water. PCR was performed in a thermocycler (Touchdown; Thermo Hybaid, Ashford, United Kingdom) with an initial denaturation cycle of 5 min at 94C, 1 min at 60C, and 1 min at 72C, followed by 18 cycles of 1 min at 94C, 1 min at 60C, and 1 min at 72C and a final extension cycle of 1 min at 94C, 1 min at 60C, and 10 min at 72C (for a total of 20 cycles). In the two nested PCRs, 1 mul of the first-round amplification product was added to 49 mul of new reaction mixtures consisting of 5 mul of 10x PCR buffer, 5 mul of 25 mM MgCl2 solution, 1 mul of a 10 muM deoxynucleotide triphosphate mixture, either 1.38 mul of the ITS I forward primer (20 muM) and 0.35 mul of the middle reverse primer (0.5 mug/mul) or 0.35 mul of the middle forward primer (0.5 mug/mul) and 1.38 mul of the ITS II reverse primer (20 muM), 0.25 mul of Taq DNA polymerase, and 36.02 mul of deionized water. Both PCR amplifications were performed in a thermocycler (Touchdown; Thermo Hybaid) with an initial denaturation cycle of 5 min at 94C, 1 min at 50C, and 1 min at 72C, followed by 35 cycles of 1 min at 94C, 1 min at 50C, and 1 min at 72C and a final extension cycle of 1 min at 94C, 1 min at 50C, and 10 min at 72C (for a total of 37 cycles). Fragment analysis. | Prior to fragment analysis, 0.75 mul of the ITS I PCR product and 0.25 mul of the ITS II PCR product were spiked with 0.75 mul of an internal base pair standard (CEQ DNA size standard kit 600; Beckman Coulter) in a 40-mul total volume of freshly deionized formamide (Mallinckrodt Baker, Phillipsburg, N.J.). The internal base pair standard includes DNA fragments ranging in size from 60 to 640 nucleotides. Fragment analysis was then performed with the spiked sample by using a fragment analysis instrument (CEQ 2000 XL DNA analysis system; Beckman Coulter). The standard procedure for fragment analysis, described in the manual that accompanies the system, was followed. Species were identified by size analysis of two unique fragments which contain the complete ITS I region (including some 18S and 5.8S gene sequences) and the complete ITS II region (including some 5.8S and 28S gene sequences). According to the manufacturer, fragment lengths are expected to be reproducible to within less than +-0.27 bp units, but we have found an average reproducibility of +-1 bp to be more typical for these studies. Gel electrophoresis. | All ITS I and ITS II PCR products were analyzed by electrophoresis in a 1.25% (wt/vol) agarose gel by standard procedures, with bands visualized by staining with SYBR Green stain (SYBR Green I Nucleic Acid Gel stain; Molecular Probes, Eugene, Oreg.). FIG. 1. | Structure of ITS gene region and locations of primer sites. Structure of ITS gene region and locations of primer sites. TABLE 1 | Malassezia fungal species and strains used as standards and fragment length obtained by tFLP analysis RESULTS : Analysis of purified strains and mixtures. | Examples of typical fragment analysis results obtained for standards and mixtures of standards are shown in Fig. . Results obtained for a standard swab inoculated with a pure M. restricta (isolate 7877) culture showed two peaks; one represents the ITS I fragment (length, 294 bp) and another represents the ITS II fragment (length, 460 bp). Note that the label at 295.82 is a mislabeled shoulder resulting from the high peak at 294 bp. Base pair standards were used as internal controls by the fragment analysis software to calculate and assign base pair values to standards and unknowns. Fragment analysis results for swabs dosed with standard cultures are summarized in Table . In addition to the seven Malassezia species, several isolates within a single species resulted in unique combinations of fragment lengths. This indicated that several genotypes could be distinguished within one species. For example, all three M. globosa isolate standards (isolates 7874, 7966, and 7990) had unique fragment patterns. Results obtained for a swab inoculated with a mixture of all seven species (including 12 genotypes) of Malassezia showed 10 major ITS I peaks and 10 major ITS II peaks associated with the seven Malassezia species . Assignments of ITS I and ITS II peaks on the basis of the results for the standards were consistent with all 12 of the genotypes present in the mixture: M. furfur (isolate 7982), M. globosa (isolate 7966), M. globosa (isolate 7874), M. globosa (isolate 7990), M. obtusa (isolate 7968), M. restricta (isolate 7991), M. restricta (isolate 7877), M. restricta (isolate 8742), M. slooffiae (isolate 7971), M. slooffiae (isolate 7956), M. sympodialis (isolate 7977), and M. pachydermatis (isolate ATCC 74522). Although 12 genotypes were used to prepare this complex mixture, only 10 ITS I peaks and 10 ITS II peaks were observed in the fragment analysis. This is because there is some overlap in one or the other fragment length value. For example, an examination of M. restricta shows that all isolates tested have identical ITS II peaks, but two genotypes can be distinguished on the basis of differences in ITS I fragment length. Triplicate analysis of swabs inoculated with the same standard mixture resulted in ITS I and ITS II peaks with fragment lengths that were reproducible within +-1 bp, but with various peak heights. These results demonstrate that even for this very complex mixture of Malassezia species, the nested tFLP amplification technique, followed by fragment analysis, is capable of separating and identifying all known species. Evaluation of method sensitivity. | Gel electrophoresis analysis was performed with the nested tFLP PCR products from serial dilutions of each Malassezia standard in order to estimate the limit of detection for the method (see Materials and Methods). The results indicated that the detection limit for each Malassezia standard was on the order of 50 to 100 cells/swab. Analysis of human scalp swab specimens. | Fragment analysis of a representative human scalp swab specimen showed two distinct peaks associated with ITS I fragments and two peaks associated with ITS II fragments, indicating the presence of two Malassezia species . The observation of ITS I fragment peaks at base pair values of 294.87 and 336.64 and ITS II peaks at base pair values of 460.40 and 474.55 indicated the presence of M. restricta (isolate 7877 or 8747 or both) and M. globosa (isolate 7966), respectively. These results demonstrate that the amplification achieved by the nested tFLP approach is sufficient to distinguish Malassezia fragments from human scalp swab specimens. These results also show that the method is capable of distinguishing three different genotypes in the M. globosa species and two different genotypes in the M. restricta species. Samples isolated from a single human scalp swab specimen and analyzed in triplicate showed virtually identical fragment length peaks (+-1 bp), but the peak heights did vary. Results for human scalp swab specimens. | Results for a total of 70 subjects enrolled in the U.S. portion of a scalp health study are shown in Fig. . A total of 24 subjects were assigned composite ASFS of less than 10 and 46 were assigned composite ASFS of greater than 24 by an expert grader. M. restricta and M. globosa were the predominant Malassezia species found in swab specimens from both groups. However, subjects with high composite ASFS were more likely to show the presence of these Malassezia species (M. restricta, 72% with a high composite ASFS versus 50% with a low composite ASFS; M. globosa, 45% with a high composite ASFS versus 33% with a low composite ASFS). M. sympodialis was detected in only a very small percentage of both groups (8% with a low composite ASFS versus 7% with a high composite ASFS). M. slooffiae and M. obtusa were observed only in a very small percentage of the group with a high composite ASFS (4 and 2%, respectively). Importantly, there was no indication of the presence of M. furfur or M. pachydermatis in scalp swab specimens from any subjects. A significant percentage of subjects from both groups (25% with a low composite ASFS versus 28% with a high composite ASFS) showed the presence of non-Malassezia fungal species. Furthermore, scalp swab specimens from subjects with high composite ASFS were more likely to have detectable levels of fungi than those assigned low composite ASFS (no fungi were detected in 15% of those with a high composite ASFS and 29% of those with a low composite ASFS). FIG. 2. | Example of data obtained by tFLP analysis. Example of data obtained by tFLP analysis. (A) Data for ITS I and ITS II DNAs isolated from a swab inoculated with a pure M. restricta culture; (B) all known genotypes of Malassezia inoculated onto one swab, showing that all species can be recovered when dosed in equal proportions. Green, ITS I fragment; blue, ITS II fragment; red, base pair standards. FIG. 3. | Example of data obtained by tFLP analysis for a typical clinical swab specimen showing the presence of M. globosa Example of data obtained by tFLP analysis for a typical clinical swab specimen showing the presence of M. globosa (isolate 7966) and M. restricta (isolate 7877 or 8747 or both). FIG. 4. | Frequencies of species detection on scalps of subjects with low and high composite ASFSs. Frequencies of species detection on scalps of subjects with low and high composite ASFSs. DISCUSSION : Malassezia fungi have been the suspected cause of dandruff for more than a century . Identification of the exact species associated with dandruff has been complicated by several factors discussed previously, i.e., isolation, cultivation requirements, or method of species differentiation. The many nomenclature changes, from Malassezia to Pityrosporum, and the recent identification of at least seven species in the genus Malassezia have also caused some confusion. In early work, microscopic examination of specimens from individuals with dandruff often showed two types of fungi . They were named P. ovale and P. orbiculare on the basis of their morphologies (bowling pin shaped and round, respectively). Later, the original genus name Malassezia was reinstated and two species were generally accepted, namely, the lipid-dependent M. furfur and the lipophilic M. pachydermatis . The two entities associated with dandruff (P. ovale and P. orbiculare) were grouped under the M. furfur species as serovar C and serovar B, respectively. Recently, when seven species were identified, the original M. furfur designation was further delineated into six species. Perhaps due to the changes in nomenclature and the absence of reliable differentiation methods, recent dandruff literature continues to contain references to P. ovale, P. orbiculare, and M. furfur as the suspected causal agent associated with dandruff . Here we report on the development of a novel molecular technique, tFLP analysis, for the rapid differentiation of Malassezia species in complex clinical samples. While the overall approach is based on that of Liu et al. , it differs in that restriction analysis is not required. Instead, this method involves isolation of fungal DNA, followed by nested PCR of the ITS I and ITS II regions of the ribosomal gene cluster with fluorescent primers, followed by fragment length analysis. Results obtained for standards and mixtures of standards show that all known Malassezia genotypes can be identified on the basis of unique fragment lengths, eliminating the need for restriction analysis. Results from this study also show that tFLP analysis is capable of reproducibly assessing the Malassezia species present in complex mixtures and human scalp samples. Importantly, it is specific for fungi and is sufficiently sensitive to allow direct assessments of human scalp swab specimens without the need for prior cultivation. Because clinical assessments can be made without prior cultivation, the results are free from culture bias and the turnaround time for analysis is significantly reduced. Specifically, we have found that a single scalp swab specimen can be analyzed within 2 days of receipt. For survey work, the sample throughput can be up to 75 samples per week. Overall, these results show that the tFLP approach is well suited for routine analysis of both clinical samples and ongoing screening work. The tFLP method will become a very powerful tool when used in conjunction with the new databases containing fungal ITS region data that are becoming available. These include databases by Chen et al. and by Boekhout et al., which is available on CD-ROM (Yeasts of the World 2.0, 2002, ETI Biodiversity Center, Amsterdam, The Netherlands). The utility of the tFLP method was evaluated by analyzing scalp swab specimens from subjects enrolled in a demographic scalp health study . Samples were prepared by extracting DNA directly from scalp swab specimens, without prior cultivation. Results for subjects with low composite ASFSs (ASFS, <10) were compared to those for subjects with higher composite ASFS (ASFS, >24). In general, the overall fingerprint of Malassezia species was found to be similar for both groups , with the group with the low composite ASFSs typically showing a lower percentage of each of the species. The most prevalent Malassezia species found in both groups were M. restricta and M. globosa. Only the group with high composite ASFSs showed a very low incidence of M. slooffiae and M. obtusa. Importantly, neither group showed the presence of M. furfur or M. pachydermatis. On the other hand, the group with the low composite ASFS had a higher percentage of subjects in whom no fungi were detected. In addition, both groups showed comparable levels of non-Malassezia fungi. The identification of non-Malassezia fungal species in the present study is of interest. Cloning and sequencing of the ITS I and ITS II products are under way. It is not surprising that we found no indication of M. pachydermatis in these human scalp swab samples because this species is typically associated with animals. However, the absence of M. furfur in these scalp swab samples has important implications. While we cannot draw definitive conclusions regarding the cause of dandruff from this work, the results strongly indicate that M. furfur, M. pachydermatis, M. slooffiae, M. sympodialis, and M. obtusa can be eliminated as potential causal organisms for dandruff. Several findings from this study are consistent with those previously reported by Sugita et al. and Gaitanis et al. . The results of these studies are expected to be free of potential culture bias. Importantly, Sugita et al. also found evidence for both M. restricta and M. globosa in a high percentage of skin swab specimens obtained from subjects judged to be healthy and those with atopic dermatitis (AD). Skin samples from patients judged to be healthy were also more likely to be free of detectable Malassezia species than samples from subjects with AD. One difference between these two studies is the higher frequency of M. sympodialis and M. furfur detection in both healthy subjects and subjects with AD by Sugita et al. . In addition, the study of Sugita et al. did not detect the presence of non-Malassezia fungi. The former may be because skin samples were obtained from a variety of sites on the body (scalps, napes, and backs), whereas the samples evaluated in the present study were obtained exclusively from the scalp. Furthermore, we were able to identify non-Malassezia fungal species in this study, because the tFLP approach, while validated for Malassezia species in our laboratory, is capable of detecting other fungal species. The method used by Sugita et al. makes use of individual PCRs which are highly selective for individual Malassezia species and, therefore, not applicable for screening of other fungal species. The only other study, to our knowledge, to have dealt with detection and species definition of Malassezia has been reported by Gaitanis et al. . The reported method uses PCR followed by restriction fragment length polymorphism (RFLP) analysis to identify the species present in skin scales collected from 17 sites, including 5 on the human head, but not specifically from the scalp. Because RFLP analysis requires restriction digestion prior to analysis, there would be a loss of speed and method sensitivity (as may be indicated by the detection rate of 44% in samples from patients with disease compared to a detection rate of 85% in the present study). Also, the multiple banding pattern resulting from RFLP analysis complicates the interpretation of complex communities, as are often found in clinical samples. The method reported here requires visualization of only two clearly separated bands per species, which makes identification of the species in complex mixtures much less problematic. Interestingly, and in agreement with our data, they report the detection of only M. restricta, M. globosa, and M. slooffiae on human heads, while M. furfur is confined to the trunk (in pityriasis versicolor). Further evaluation of multiple healthy and diseased skin sites will be necessary to directly compare the two methods. The method described here is expected to be useful in the clinical assessment of the Malassezia species associated with other fungal infections. However, assessments of additional types of clinical samples are indicated to determine the broad applicability of the approach. Backmatter: PMID- 12202608 TI - Evaluation of a New Dot Blot Enzyme Immunoassay (Directigen Flu A+B) for Simultaneous and Differential Detection of Influenza A and B Virus Antigens from Respiratory Samples AB - We report a prospective evaluation of a new dot blot enzyme immunoassay (EIA) method for the direct, rapid, qualitative, simultaneous, and differential detection of the influenza A (IA) and B (IB) virus antigen in different respiratory samples. The EIA method was compared with the shell vial culture system (MDCK cell line) used with the same samples. We studied 160 samples from 93 (58.1%) pediatric patients (hospital emergency room) and from 67 (41.9%) adult patients (sentinel network). Seventy-four(46.2%) samples were considered positive; of them, 46 (62.2%) were from pediatric patients and 28 (37.8%) were from an adult group (P < 0.05), with overall positive values of 49.9% and 41.7%, respectively. All 74 (100%) of the positive samples were isolated in cell culture versus the 68.9% that were detected as positive by the new EIA method (P < 0.05). Of the 41 samples positive for the IA virus, the EIA detected 34 (82.9%) positive samples; of the 33 samples positive for the IB virus, the EIA detected 17 (51.5%) positive samples (P < 0.05). No false-positive reaction was detected with the EIA method (specificity and positive predictive value of 100%). The overall results obtained in the comparison between the new EIA and the shell vial culture had a sensibility of 82.9% and predictive negative values of 92.4% for the IA virus and 51.5% and 84.3%, respectively, for the IB virus. This evaluation shows sensitivity and specificity percentages for the new EIA method that is acceptable for routine use in IA virus detection. The results obtained were worse for IB virus detection, but this new EIA method is actually the only one with the capacity to differentiate between the two influenza viruses. Keywords: Introduction : Influenza is an infection caused by the influenza A and B (IA and IB) viruses, which present as epidemic outbreaks in the winter months. This epidemiological fact is of use in the clinical diagnosis of this infection. Usually, this infection may be considered self-limiting in healthy populations. Nevertheless, in the very young and in immunodepressed patients it may lead to an increase in morbidity and mortality . However, it is necessary to carry out a definitive etiological diagnosis at the beginning of, and during, each epidemiological period in order to establish the prevalence and appearance of new strains or subtypes not included in the recommended vaccine . At the same time, the appearance and the availability of neuraminidase inhibitors requires a rapid (these new antivirals are most effective when given within 48 h of symptoms) and specific diagnosis of influenza virus infection . The diagnosis of influenza infection is largely clinical, but this method has been demonstrated to be both insensitive and nonspecific . The reference method (gold standard) for laboratory diagnosis of influenza is the isolation of the virus. This may be carried out by inoculation in embryonated hens' eggs, in laboratory reference only, or by means of cell culture (classical or shell vial type) . These methods, however, have serious drawbacks in that they are slow and laborious and require from 2 to 7 days to reach the final result. Consequently, rapid techniques based on the detection of viral antigens or of physiological viral activity (neuraminidase) have been developed . Immunofluorescence is a highly sensitive and specific technique, but it requires a minimum number of cells in the sample and an expert technician for good interpretation . The enzyme immunoassay systems (EIA) have provided high sensitivity, high specificity (nucleoprotein as the antigen), rapid diagnosis (less than 15 min), and technical simplicity. The majority of these systems are performed on a solid membrane and are based on an enzymatic reaction with the development of a visual color . The aim of this study was to carry out a prospective evaluation of a new EIA method in the direct, qualitative, simultaneous, and differential detection of the IA and IB virus antigens in different clinical samples of symptomatic patients. From January to December 2001 we evaluated the efficacy of a new commercial rapid EIA method for the differential detection of the IA and IB viruses in clinical samples of two different population groups. Samples were taken from adult patients attended to in the sentinel network (community-based study) by using a throat swab vigorously rubbed on both tonsillar surfaces and the posterior pharynx. In the case of patients attended to in the pediatric emergency room of our hospital (pediatric group), a nasopharyngeal aspirate was taken. Both types of samples were inoculated in compatible liquid transport medium (Earle's minimum essential medium with 0.5% bovine serum albumin) for viruses and sent as soon as possible to the virology laboratory. Each of the samples was subjected to antigen detection against IA and IB viruses with the new rapid differential EIA membrane test (Directigen Flu A+B; Becton & Dickinson Co., Sparks, Md.) following the manufacturer's recommendations. At the same time, each sample was inoculated in two shell vials of the MDCK cell line (Vircell, Granada, Spain), which were incubated for 72 h at 36C, after which the monolayers were stained with a specific monoclonal antibody against the IA (clone IA52/9) and IB (clone IB82/2) viruses (Monofluokit Influenza; Sanofi Diagnostics Pasteur, Marnes la Coquette, France) by an indirect immunofluorescence assay. Statistical analysis was carried out on results of different comparisons by performing the Student's t test on paired data. All P values are two-tailed and considered significant if they are less than 0.05. In this study we analyzed 160 clinical samples from 93 (58.1%) pediatric patients and 67 (41.9%) patients of the sentinel network (adults). Of these samples, 74 (46.2%) were considered positive (detection and/or viral isolation). Of the positive samples, 46 (62.2%) were from the pediatric group and 28 (37.8%) were from the sentinel group (P < 0.05), giving an overall positivity of 49.4% for the pediatric samples and 41.7% for the sentinel group. All 74 (100%) of the positive samples were isolated in cell culture, 68.9% of which were detected as positive by the EIA method (P < 0.05). The IA virus was isolated in 41 samples (25.6%), and the IB virus was isolated in 33 (20.6%) samples. Of the 41 samples positive for the IA virus the EIA method detected 34 (82.9%) positive samples, and of the 33 samples positive for the IB virus the EIA method detected 17 (51.5%) (P < 0.05). Of the 46 positive pediatric samples the IA virus was isolated in 30 and was detected by the EIA method in 26 (86.6%). The IB virus was isolated in 16 samples in this group and was detected by the EIA method in 10 (62.5%). Of the 28 positive samples from the sentinel group, the IA virus was isolated in 11 samples and was detected by the EIA method in 8 (72.7%). The IB virus was isolated in 17 samples in this group and was detected by the EIA method in 7 (41.1%) samples. No false-positive reaction was detected with the EIA method studied, giving us a specificity and positive predictive value of 100%. Table shows the overall results and the results obtained for each group in the comparison between isolation in cell culture and antigen detection (new EIA method). TABLE 1 | Results obtained from the comparison between shell vial culture and the rapid EIA test The new EIA method examined in this study is rapid and simple, and it permits the simultaneous and differential detection of IA and IB virus antigens. The comparison between the cell culture (shell vial) and this new EIA method showed an overall sensitivity of 68.9%, somewhat lower than expected. However, if we separate the two viruses detected we find that the sensitivity for IA was 82.9%, while for the IB virus it was 51.5% (P < 0.05). In a previous study carried out with the same antigen detection EIA method against only the IA virus (Directigen Flu A; Becton & Dickinson), a sensitivity of 84.7% was observed . Therefore, the present method displays a practically identical behavior against the IA virus, maintaining very similar sensitivity values, similar to results of other studies . One of the advantages of the new EIA method is the ability to specifically detect the IB virus. Until now no other EIA method with this capacity had been commercialized, so there are no previous studies with which we can compare our results. Reina et al. have previously reported a sensitivity of 66.6% for an indirect immunofluorescence technique against the shell vial culture in the detection of the IB. This value is slightly higher than that detected by the EIA method evaluated in this study, confirming the idea that, in general, immunofluorescence techniques are usually somewhat more sensitive than the EIA methods for the influenza viruses . One of the main problems when evaluating different methods for the antigenic detection of respiratory viruses is the type of sample studied. For this reason we divided the patients into two different groups, both for reasons of age (children and adults) and for the type of clinical sample analyzed (nasopharyngeal aspirate and pharyngeal swab). The majority of studies concerning other EIA methods with the ability to detect simultaneously, although not differentially, the IA and IB viruses have shown important variations in sensitivity according to the type of sample studied . Thus, in the study of Schultze et al. we find that the optical immunoassay method (Flu OIA) displays an overall sensitivity of 71.8% in pediatric samples and of 51.4% in adult samples. In the same way, Covalciuc et al. reported that, with this same method, the highest sensitivity is obtained with nasal aspirate (88.4%), and the lowest sensitivity is obtained with the throat swab (62.1%). In our study the new EIA method displayed a sensitivity of 86.6% for pediatric samples and 72.7% for adult samples (sentinel network) in the detection of the IA virus. The difference between the two groups and/or types of sample was greater in the case of IB virus (62.5% versus 41.1%; P < 0.05). It seems obvious, once more, that the type of sample and, therefore, the viral load present is what probably determines the sensitivity of the different antigen detection methods against the majority of respiratory viruses . This phenomenon does not affect the cell culture, which scarcely displays differences in sensitivity according to the type of clinical sample (, ,, ). Over the study period we found no false-positive antigen detection with the new EIA method, establishing a specificity and positive predictive value of 100%. Therefore, a positive result with this method provides, with a high degree of certainty, the diagnosis of infection by the influenza viruses. It may be used as a rapid screening method for patients with symptoms of infection by the influenza viruses. A negative result in the test does not exclude the existence of viral infection, especially for those caused by the IB virus. In general, the antigenic detection methods for the IB virus have displayed lower sensitivity percentages, and alternative methods, such as cell culture or reverse transcription-PCR, should be used . The availability of antiviral drugs effective against the IA and IB viruses justifies the need for the rapid and specific detection of infection caused by these viruses . In addition, the efficacy of these drugs is maximum when they were used within the first 48 h of the appearance of the disease. Therefore, physicians require rapid and simple diagnostic methods which they themselves can use in their consulting rooms. However, since sometimes the reading of results obtained by EIA methods can be difficult , a sample should be always sent to the laboratory for confirmation, culture, isolation, and typing the virus for epidemiological studies . In summary, this study shows sensitivity and specificity percentages for the new EIA method which is acceptable for routine use in antigen detection of IA virus, and it is comparable with other no-differential methods. The results obtained are worse for IB antigenic detection, but this new EIA is actually the only one with the capacity to differentiate between the two influenza viruses. At the same time, it was possible to confirm the variations in the behavior of this antigenic method depending on the sample used. It has been shown to be highly effective for nasopharyngeal aspirates from children attended to in hospital emergency rooms and has been shown to be somewhat less effective for the detection of these viruses in adult patients from the sentinel network. Backmatter: PMID- 12202574 TI - Possible Connection between a Widely Disseminated Conjugative Gentamicin Resistance (pMG1-Like) Plasmid and the Emergence of Vancomycin Resistance in Enterococcus faecium AB - A total of 640 vancomycin-resistant Enterococcus faecium (VRE) isolates, which were obtained between 1994 and 1999 from the Medical School Hospital of the University of Michigan, Ann Arbor, were used in this study. Of the 640 strains, 611 and 29 were VanA and VanB VRE, respectively, based on PCR analysis. Four hundred ninety-two (77%) of the strains exhibited resistance to concentrations of gentamicin from 64 mug/ml (MIC) to more than 1,024 mug/ml (MIC). The gentamicin resistance of each of 261 (53%) of the 492 gentamicin-resistant strains was transferred to E. faecium at a frequency of about 10-5 to 10-6 per donor cell in broth mating. More than 90% of vancomycin resistances of the 261 strains cotransferred with the gentamicin resistances to E. faecium strains by filter mating. The conjugative gentamicin resistance plasmids were identified and were classified into five types (A through E) with respect to their EcoRI restriction profiles. The transfer frequencies of each type of plasmid between E. faecium strains or Enterococcus faecalis strains were around 10-3 to 10-5 per donor cell or around 10-6 to 10-7 per donor cell, respectively, in broth mating. Type A and type B were the most frequently isolated, at an isolation frequency of about 40% per VRE isolate harboring the gentamicin resistance conjugative plasmid. The plasmids did not show any homology in Southern hybridization with the pheromone-responsive plasmids and broad-host-range plasmids pAMbeta1 and pIP501. The EcoRI or NdeI restriction fragments of each type of plasmids hybridized to the conjugative gentamicin resistance plasmid pMG1 (65.1 kb), which was originally isolated from an E. faecium clinical isolate, and transfer efficiently in broth mating. Keywords: Introduction : The isolation of vancomycin-resistant enterococci (VRE) was first reported in the United Kingdom (A. H. Uttley, C. H. Collins, J. Naidoo, and R. C. George, Letter, Lancet i:57-58, 1988) and in France in 1988. Shortly after the first reports were made, VRE were detected in hospitals in the United States . Since then, VRE have emerged with unanticipated rapidity and, especially in the United States, are now encountered in most hospitals . Among the acquired glycopeptide resistances of VanA and VanB, VanA resistance has been predominantly isolated from both within and outside the health care environment, from animals, and from the general environment (, ; J. Bates, J. Z. Jordens, and J. B. Selkon, Letter, Lancet 342:490-491, 1993). Most VRE isolates from the health care environment in the United States have multiple-drug resistance, including high-level gentamicin resistance and ampicillin resistance. One of the major factors that have contributed to the dissemination of VRE in the United States and Europe is now evident. In the United States, it is likely that excessive use of glycopeptide antibiotics in the health care environment resulted in the selective increase of VRE in the human intestine , which subsequently spread by nosocomial transmission. In Europe, it is strongly suggested that the use of avoparcin as a growth promoter in animal feed has resulted in the selective increase of VRE in animal intestines, and these VRE subsequently appear in the human community (; Y. Ike et al., Letter, Lancet 353:1854, 1999; M. A. Schouten, A. Voss, and J. A. A. Hoogkamp-Korstanje, Letter, Lancet 349:1258, 1997; A. E. van den Bogaard, L. B. Jensen, and E. E. Stobberingh, Letter, N. Engl. J. Med. 337:1558-1559, 1997). In both cases, the direct selective pressure of glycopeptides is the largest contributor to the selective increase in VRE in different habitats. Besides the direct selective pressure of antibiotics for increasing selectively in drug-resistant bacteria, the genetic transfer system of an organism is essential to the spread of drug resistance in the organism. In Enterococcus faecium, it has been reported that the transferable plasmid or mobile genetic elements encode drug resistance determinants. VanA-type resistance determinant is encoded on transposon Tn1546 borne by nonconjugative or conjugative plasmids that transfer in enterococci by mating on a solid surface (filter mating) . The VanB determinant of E. faecium is encoded on a large mobile genetic element of conjugative transposons such as Tn1547 , Tn1549 , and Tn5382 . Little is known about systems of efficient plasmid transfer in E. faecium. Previously, we described the isolation of the pheromone-independent gentamicin resistance conjugative plasmid pMG1 (65.1 kb), which transfers efficiently among enterococcus strains during broth mating and was isolated from an E. faecium clinical isolate in Japan . In this report we describe the study of VRE clinical isolates derived from a hospital in the United States and show the wide dissemination of a gentamicin resistance plasmid that transferred in broth matings, like pMG1. We also show that these plasmids may contribute to the efficient dissemination of vancomycin-resistance determinants in enterococcus strains. MATERIALS AND METHODS : Bacteria, plasmids, and media. | The laboratory strains and plasmids used in this study are listed in Table . A total of 640 vancomycin-resistant E. faecium clinical isolates were used in this study. They were obtained from different patients who had been admitted to the University of Michigan Medical School Hospital, Ann Arbor, between 1994 and 1999. Of the 640 isolates, strains numbered from 1 to 45, from 46 to 104, from 105 to 164, from 165 to 350, from 351 to 642, and from 643 to 730 were isolated in 1994, 1995, 1996, 1997, 1998, and between January and June of 1999, respectively. Enterococcus strains were grown in Todd-Hewitt broth (THB) (Difco Laboratories, Detroit, Mich.) throughout this study. Mueller-Hinton (MH) broth and MH agar were used for the sensitivity disk agar-N (Nissui, Tokyo, Japan) assay to test the MICs of antimicrobials. Agar plates were prepared by adding 1.5% agar to broth medium. All bacterial strains were grown at 37C. The MICs of the antimicrobials were determined according to the criteria of the National Committee for Clinical Laboratory Standards using MH agar . Overnight cultures of the strains grown in MH broth were diluted 100 times with fresh broth. One loopful (5 mul; about 5 x 103 to 104 cells) of each dilution was transferred to agar plates containing the relevant drug. The antimicrobials used to test the MICs for VRE isolates were vancomycin, teicoplanin, gentamicin, ampicillin, kanamycin, streptomycin, tetracycline, and minocycline. Mating procedures. | Broth matings were performed as previously described with a donor/recipient ratio of 1:10. Overnight cultures of 5 mul of the donor and 50 mul of the recipient were each added to 0.5 ml of fresh THB, and the mixtures were incubated at 37C with gentle agitation for the appropriate times and then vortexed. Unless otherwise described, the mating time of broth mating was 3 h. Portions of the mixed cultures were then transferred to plates of solid media with appropriate selective antibiotics. Colonies were counted after 48 h of incubation at 37C. Mating on a solid surface was performed on agar plates. The mating mixture of donor and recipient was made as described above, and 10 mul of the mixed culture was spread onto THB agar without drug. The plates were then incubated overnight (18 h) at 37C. After the incubation, the bacteria grown on the agar plates were scraped off and transferred to 1 ml of fresh broth, and then 0.1 ml of the suspension was inoculated onto appropriate selective agar plates. Filter matings were performed as previously described with a donor/recipient ratio of 1:4. Overnight cultures were prepared, 0.05 ml of the donor and 0.2 ml of the recipient were added to 4.5 ml of fresh THB, and the cells were then trapped on a membrane filter (Millipore, Bedford, Mass.). The cells on the filters were incubated at 37C for appropriate times and were then suspended in 1 ml of THB. Appropriate dilutions of the mixture were transferred to plates of solid medium containing selective antibiotics. Throughout the mating experiments, the antibiotic concentration used for the selection of gentamicin- or vancomycin-resistant transconjugants was 100 or 12.5 mug/ml, respectively. The antibiotic concentrations for the selection of rifampin- and fuscidic acid-resistant recipient strains or streptomycin- and spectinomycin-resistant recipient strains were 25 and 25 mug/ml or 500 and 250 mug/ml, respectively. Isolation and manipulation of plasmid DNA. | Plasmid DNA was isolated by the alkaline lysis method . Plasmid DNA was treated with restriction enzymes and subjected to agarose gel electrophoresis for the analysis of DNA fragments, etc. Restriction enzymes were obtained from Nippon Gene (Toyama, Japan), New England Biolabs, Inc., and Takara (Tokyo, Japan) and were used in accordance with the suppliers' specifications. Southern hybridization. | Southern hybridization was performed with the digoxigenin (DIG)-based nonradioisotope system of Boehringer GmbH (Mannheim, Germany), and all procedures were based on the manufacturer's manual and standard protocols . Hybridization was performed overnight at 42C in the presence of 50% formamide. The probe for vanA was generated by PCR amplification of DNA from the VanA-type E. faecium FN1 (N. Fujita, M. Yoshimura, T. Komori, K. Tanimoto, and Y. Ike, Letter, Antimicrob. Agents Chemother. 42:2150, 1998) using the PCR DIG probe synthesis kit (Roche Diagnostics, Mannheim, Germany). The nucleotide sequences of the primer pair were as follows: 5'-ATGAATAGAATAAAAGTTGCAATAC and 5'-CCCCTTTAACGCTAATACGAT for vanA ligase and 5'-CCCGAATTTCAAATGATTGAAAA and 5'-CGCCATCCTCCTGCAAAA for vanB ligase . Signals were detected with the DIG chemiluminescence detection kit (Boehringer GmbH). CSPD (Boehringer GmbH) was used as a substrate for alkali phosphatase conjugated to the antidigoxigenin antibody. TABLE 1 | Bacterial strains and plasmids used in this study RESULTS : Drug resistance of VRE isolates. | Six hundred and forty VRE isolates were examined for drug resistance as described in Materials and Methods. Many of the strains exhibited high-level resistance to various antibiotics and multiple drug resistance . There were bipolar distributions of the MICs of antibiotics except for vancomycin in the VRE strains. The distribution of the MICs of vancomycin, teicoplanin, gentamicin, and ampicillin for the 640 VRE strains are shown in Fig. . The 640 vancomycin-resistant E. faecium strains were resistant to vancomycin at levels equal to or greater than 64 mug/ml. Five hundred seventy (89%) of the strains were resistant to teicoplanin at levels equal to or greater than 16 mug/ml. Four hundred and ninety-two (77%) of the strains exhibited resistance to concentrations of gentamicin from 64 mug/ml (MIC) to more than 1,024 mug/ml (MIC), and 608 (95%) isolates exhibited resistance to concentrations of ampicillin from 16 to 512 mug/ml, depending on the strain. About 90 and 70% of the strains were resistant to kanamycin at concentrations of more than 1,024 mug/ml and had a resistance to streptomycin equal to or greater than 512 mug/ml, respectively (data not shown). About 60% of the strains exhibited MICs of tetracycline and minocycline equal to or less than 0.5 mug/ml, and the remainder of the strains exhibited MICs of tetracycline or minocycline between 0.5 and 128 mug/ml or 0.5 and 64 mug/ml, respectively (data not shown). The DNAs of the VRE strains were analyzed by PCR for the presence of the vancomycin resistance gene with each of the vanA- and vanB-specific primers. Of the 640 strains, 611 strains gave rise to the expected 1,029-bp product with the primers specific for the vanA gene, indicating that the strains were VanA-type VRE . Of the 611 VanA-type VRE, 570 strains were resistant to teicoplanin at levels equal to or greater than 16 mug/ml, and for 41 strains the MICs of teicoplanin were less than 16 mug/ml. Of the 640 strains, 29 strains, for which the MICs of teicoplanin were equal to or less than 0.5 mug/ml gave rise to the expected 457-bp product with primers specific for the vanB gene, indicating that the strains were VanB-type VRE . Transferability of high-level gentamicin resistance of VRE isolates. | Of the 640 VRE, 492 (77%) isolates exhibited resistance to concentrations of gentamicin from 64 mug/ml (MIC) to more than 1,024 mug/ml (MIC). The transferability of the gentamicin resistance trait from each of the 492 gentamicin-resistant strains to E. faecium BM4105RF was examined by mating in broth or on a solid surface overnight at 37C. The gentamicin resistance of each of 261 (53%) of the 492 strains was transferred at a frequency of about 10-5 to 10-6 per donor cell, and about 10-2 to 10-4 per donor cell, respectively, in broth mating and mating on a solid surface. The gentamicin resistance of each of 86 (17%) of the 492 strains was transferred at a frequency of about 10-4 to 10-5 per donor cell by mating on a solid surface and was not transferred at a frequency of less than 10-7 per donor cell by broth mating. Of the 492 gentamicin-resistant VRE strains, 145 strains (29%) did not transfer the gentamicin resistance, even by filter mating, at a frequency of less than 10-8 per donor cell. Isolation of the gentamicin resistance conjugative plasmids. | Of the 261 strains that transferred gentamicin resistance by broth mating, 60 strains of VanA VRE were selected at random, and their conjugative plasmids were analyzed. Plasmid DNA was isolated from a representative transconjugant generated from matings using each of the 60 strains. The DNA was digested with EcoRI and examined by agarose gel electrophoresis. A number of transconjugants harbored several plasmids based on a number of different fragments. Plasmid DNAs were studied for homology with the plasmid pMG1 (Gmr) (65.1 kb). The plasmid pMG1 hybridized to specific EcoRI fragments from plasmids of each of the 60 transconjugants. The plasmids isolated from 52 of 60 transconjugants were classified into five types (type A to E) with respect to the EcoRI restriction profiles that hybridized to pMG1 DNA. The plasmids isolated from the remaining 8 transconjugants exhibited different EcoRI restriction profiles that hybridized to pMG1 DNA. Of the 52 strains, a total of 25, 22, 2, 2, and 1 strain(s) harbored the type A, B, C, D, or E plasmid, respectively. Figure shows representative results of Southern hybridization of the plasmids isolated from transconjugants harboring type A or type B plasmid. Each type of plasmid DNA was identified from the transconjugant by repeated transfer experiments between E. faecium BM4105RF and E. faecium BM4105SS by short mating (30-min mating) in either one or two strains of each type. The transconjugant harboring each type of plasmid was resistant only to gentamicin, and each type of plasmid did not encode the VanA determinant based on PCR analysis or Southern analysis with vanA-specific primer or vanA-specific probe, respectively (data not shown). Each type of transferable plasmid DNA was studied to determine the EcoRI or NdeI restriction profiles and homology with pMG1 by Southern analysis. The pMG1 DNA probe hybridized to all EcoRI fragments of type A, B, and C plasmid DNA, and hybridized to specific EcoRI fragments of type D and E plasmid DNA, with the exception of the fragment of about 0.8 kbp . The pMG1 DNA probe hybridized to all NdeI fragments of each type of plasmid DNA . Conjugative transfer of gentamicin resistance plasmid. | Each type of gentamicin resistance plasmid was examined for conjugative transfer in broth or filter mating. As shown in Table , each of the plasmids examined transferred between E. faecium strains, between Enterococcus faecalis strains, and between E. faecium and E. faecalis strains by broth mating. The transfer frequencies between E. faecium strains or between E. faecalis strains were around 10-3 to 10-5 per donor cell or around 10-6 to 10-7 per donor cell, respectively, by broth mating. The transfer frequencies to E. faecium recipient strains were about 2 or 3 orders higher than the E. faecalis recipient strain. The gentamicin resistance plasmid transferred highly efficiently to the E. faecium recipient strain at a frequency of more than 100 per donor cell and to the E. faecalis recipient strain at a frequency of around 10-2 to 10-3 per donor cell by filter mating. Conjugative cotransfer of vancomycin resistance with gentamicin resistance. | Transferability of the vancomycin resistance trait of VRE strains was examined between the donor strain of each of the VRE strains and the recipient strain E. faecium BM4105RF by broth mating or on a solid surface. Of the 261 VRE strains that were resistant to gentamicin and transferred the gentamicin resistance trait at a frequency of around 10-5 to 10-6 per donor cell by broth mating, 255 strains (97.7%) transferred vancomycin resistance at a frequency of around 10-5 to 10-6 per donor cell on a solid surface. Typical results showing the transferability between wild-type strains to the laboratory strain E. faecium BM4105RF by filter mating are shown in Table . Of the 86 VRE strains that were resistant to gentamicin and transferred gentamicin on a solid surface and did not transfer in broth mating, 77 strains (90%) transferred vancomycin resistance at a frequency of around 10-8 to 10-6 per donor cell on a solid surface. Of the 145 VRE strains that were resistant to gentamicin and did not transfer the gentamicin resistance trait even on a solid medium, 70 strains (48.5%) transferred vancomycin resistance at a frequency of 10-8 to 10-7 per donor cell by filter mating on a solid surface. Of the 148 VRE strains that were not resistant (MIC <= 8 mug/ml) to gentamicin, 64 (43%) transferred vancomycin resistance at a frequency of around 10-8 to 10-7 per donor cell on a solid surface. These results implied that when the conjugative gentamicin resistance plasmid is present in a vancomycin-resistant E. faecium, the likelihood of being able to transfer vancomycin resistance is enhanced. DNA-DNA hybridization. | Two types of conjugative plasmids in enterococci have been reported and well analyzed. One type of plasmid is able to transfer at relatively low frequencies on a solid surface, such as during filter mating . These plasmids usually have a broad host range. Macrolide-lincosamide-streptogramin B resistance plasmid pIP501 and pAMbeta1 are representative. The other type of plasmid is mainly found in E. faecalis and is a pheromone-responsive plasmid which transfers between E. faecalis strains at a high frequency of 100 to 10-2 per donor cell within a few hours of broth mating. Among these plasmids, the pheromone-related conjugation systems are well studied for pAD1 , pCF10 , pPD1 , and pAM373 , which confer responses to the sex pheromones cAD1, cCF10, cPD1, and cAM373, respectively. There is homology between the genes involved in the regulation of the pheromone response of these plasmids. Type A and B plasmid DNAs were each studied for homology with that of the pheromone-responsive plasmids pAD1, pPD1, and pAM373 and broad-host-range plasmids pIP501 and pAMbeta1. Each of the type A or B plasmid DNA probes did not hybridize with any EcoRI fragments from these pheromone-responsive plasmids and broad-host-range plasmids and did hybridize with all EcoRI fragments of the type A, type B, and pMG1 plasmid DNAs. These results indicate that the type A or type B plasmid did not contain any sequence homologous with the pheromone-responsive plasmids and the broad-host-range plasmids (data not shown). Restriction endonuclease digestion patterns of the VRE chromosomal DNA. | Pulsed-field gel electrophoresis was used to compare the clinical isolates of gentamicin-resistant VRE harboring the conjugative plasmid. The VRE strains used were clinical isolates corresponding to the transconjugants shown in Fig. that harbored type A and type B gentamicin resistance conjugative plasmids, respectively. The patterns of a total of 14 strains harboring type A conjugative plasmids showed eight different patterns (data not shown). Of these 14 strains, seven strains showed identical bands or differed by one or two bands, indicating that these strains were identical or were related . Another seven strains showed different patterns. The patterns of the 14 strains harboring the type B conjugative plasmid showed six different patterns (data not shown). Of these 14 strains, two groups of two strains and one group of six strains showed identical patterns or differed by one or two bands, indicating that strains of each group were identical or were related . The other four strains showed different patterns. These results indicated that the gentamicin resistance conjugative plasmid could disseminate to different E. faecium strains. FIG. 1. | Distribution of MICs of antimicrobials used in this study. Distribution of MICs of antimicrobials used in this study. The MIC of various antimicrobials (vancomycin [A], teicoplanin [B], gentamicin [C], and ampicillin [D])for 640 isolates were examined by agar dilution methods as described in Materials and Methods. FIG. 2. | Agarose gel electrophoresis of restriction endonuclease-digested plasmid DNAs and hybridization with plasmid pMG1. Agarose gel electrophoresis of restriction endonuclease-digested plasmid DNAs and hybridization with plasmid pMG1. (A1) Agarose gel electrophoresis of EcoRI-digested plasmid DNAs isolated from gentamicin-resistant E. faecium BM4105SS transconjugants harboring the type A gentamicin resistance conjugative plasmid. (A2) The gel was Southern blotted and hybridized to pMG1. (B1) Agarose gel electrophoresis of EcoRI-digested plasmid DNA isolated from gentamicin-resistant E. faecium BM4105SS transconjugants harboring the type B gentamicin resistance conjugative plasmid. (B2) The gel was Southern blotted and hybridized to pMG1. (A) Lanes: 1, HindIII-digested lambda DNA; lanes 2 to 15, EcoRI-digested plasmid DNA isolated from transconjugants of strains 161, 200, 210, 253, 300, 306, 311, 317, 376, 469, 494, 537, 581, and 692, respectively; lane 16, EcoRI-digested pMG1. (B) Lanes: 1, HindIII-digested lambda DNA; lanes 2 to 14, EcoRI-digested plasmid DNA isolated from transconjugants of strains 70, 87, 133, 166, 205, 247, 282, 411, 445, 526, 587, 619, 666, and 725, respectively; lane 15, EcoRI-digested pMG1. FIG. 3. | Agarose gel electrophoresis of restriction endonuclease-digested DNA of each type of gentamicin resistance conjugative plasmid and hybridization with pMG1. Agarose gel electrophoresis of restriction endonuclease-digested DNA of each type of gentamicin resistance conjugative plasmid and hybridization with pMG1. Agarose gel electrophoresis of EcoRI-digested plasmid DNA (A1) and the Southern hybridization with pMG1 (A2). Agarose gel electrophoresis of NdeI-digested plasmid DNA (B1) and Southern hybridization with pMG1 (B2). Lanes: 1, HindIII-digested lambda DNA; 2, pMG1; 3, pG200 (type A); 4, pG445 (type B); 5, pG566 (type C); 6, pG700 (type D); 7, pG120 (type E); 8, pAD1. TABLE 2 | Antimicrobial drug resistance pattern of vancomycin-resistant E. faecium isolates TABLE 3 | Transfer frequency of each type gentamicin resistance plasmid in Enterococcus TABLE 4 | Transfer frequency of gentamicin and vancomycin resistance of VRE strain harboring each type of gentamicin resistance plasmid DISCUSSION : The data shown in this report indicate that many VRE clinical isolates (about 80% of VRE isolates) from a major teaching hospital in the United States have high-level gentamicin resistance and that the gentamicin resistance determinant of at least half of the gentamicin-resistant strains is encoded on conjugative plasmids that efficiently transfer in broth mating at a frequency of about 10-3 to 10-5 per donor cell and transfer efficiently on a solid surface at a frequency of about 100 to 10-1 per donor cell. The conjugative plasmids were classified into five types (A through E) with respect to their EcoRI restriction profiles. Types A and B were the most frequently isolated, at an isolation frequency of about 40% per VRE isolate harboring the gentamicin resistance conjugative plasmid. The EcoRI or NdeI restriction fragments of each type of plasmid hybridized to the plasmid pMG1, indicating that each type of plasmid was similar to pMG1. pMG1 does not show any homology in Southern hybridization with that of the pheromone-responsive plasmids of E. faecalis or broad-host-range plasmids such as pAMbeta1 and pIP501 of enterococcal plasmids . Each of the type A and type B plasmid DNAs also did not show any homology in Southern hybridization with that of the pheromone-responsive plasmids and the broad-host-range plasmids. These results indicate that each type of plasmid was similar to pMG1 with respect to the efficient transferability in broth mating and the nonhomology with the pheromone-responsive plasmids and broad-host-range plasmids. Some VanA-type VRE isolates exhibited low-level teicoplanin resistance (i.e., MIC <= 8 mug/ml). The mechanism of low-level teicoplanin resistance of each of these VanA-type VRE isolates is not yet known. There are reports that amino acid substitutions in the VanS protein or defects in vanZ of the VanA-type determinant, which consists of VanRSHAXYZ genes, result in low-level teicoplanin resistance . The MICs of gentamicin resistance were distributed between 64 and >1,024 mug/ml. The MICs of 64 and 128 mug/ml were relatively low for gentamicin resistance. The gentamicin resistance of many of the strains for which the MICs were 64 or 128 mug/ml were transferred to recipient E. faecium BM4105 strains, and the MICs of gentamicin for the transconjugants were more than 512 mug/ml. These indicate that the gentamicin resistance levels of the clinical isolates depend on each of the isolates. Systems of efficient plasmid transfer are not well known among gram-positive bacteria in general. However, enterococci possess potent and unique abilities to transfer plasmids among themselves, and some of these transfer to other genera . Before the identification of the conjugative plasmid pMG1, two types of conjugative plasmids by which enterococci naturally transfer genetic elements were known and were well characterized in enterococci. One type consists of the narrow-host-range and pheromone-responsive plasmids . These plasmids transfer between E. faecalis at high frequencies (100 to 10-2 per donor cell) in broth and on solid surfaces and also in vivo . The other type represent broad-host-range plasmids (e.g., pAMbeta1 or pIP501) that transfer on a solid surface at low frequency . Transfer of these plasmids requires stable contact between donor and recipient cell on a solid surface. The identification of pMG1 shows the existence of a new system of plasmid conjugative transfer in enterococci that differs from other known conjugative plasmids. At first, pMG1 was thought to be unique to the E. faecium clinical isolate from Japan. However, as described above, gentamicin resistance conjugative plasmids that transfer efficiently in broth mating were isolated at a high frequency from E. faecium clinical isolates in the hospital in the United States, and Southern analysis implied that these plasmids were similar to pMG1. The gentamicin resistance cotransferred at a high frequency with vancomycin resistance. These results implied that the gentamicin resistance plasmid might cotransfer vancomycin resistance plasmids, which might be conjugative or nonconjugative. The mechanism of cotransfer with gentamicin and vancomycin resistance is not yet known. Although this study is limited to the isolates of one hospital, these results imply that pMG1-like plasmids are widely disseminated among E. faecium and may contribute significantly to the spread of other resistance traits, notably vancomycin resistance. Backmatter: PMID- 12202547 TI - Reactive Nitrogen Intermediates Have a Bacteriostatic Effect on Mycobacterium tuberculosis In Vitro AB - Susceptibility of six isolates of Mycobacterium tuberculosis (CB3.3, CDC1551, RJ2E, C.C.13, H37Rv, and H37Ra) and two isolates of Mycobacterium bovis (Ravenel and BCG) to reactive oxygen intermediates (ROI) and reactive nitrogen intermediates (RNI) was determined by standard in vitro survival assays. After 21 days of incubation, the survival of most strains exposed to either acidified sodium nitrite (ASN) or hydrogen peroxide (H2O2) was significantly lower than the same strains unexposed to these RNI or ROI products. However, after 50 days of incubation, these differences in susceptibility became less apparent for strains exposed to ASN but not for strains exposed to H2O2. The recovery of these strains after exposure to RNI suggests that the effect of RNI on M. tuberculosis is bacteriostatic. The in vitro concentrations of ROI and RNI used in these assays were higher than those expected in vivo. These observations suggest that, in vivo, RNI expression at physiologically achievable concentrations may keep M. tuberculosis from proliferating but that removal of RNI may allow the organisms to proliferate. Furthermore, the ability of some M. tuberculosis strains to cause rapidly progressive disease may relate to their intrinsic levels of RNI and ROI resistance. Keywords: Introduction : Although current estimates suggest that one third of the world's population is infected with Mycobacterium tuberculosis, the majority of such infections remain clinically latent. Factors that contribute to the success of M. tuberculosis as a pathogen include its ability to resist the harsh environment of the host macrophage and to persist within immunocompetent hosts even after a clinical cure of the disease is achieved. The ability of M. tuberculosis to cause disease in the human or animal host may vary according to host susceptibility and to intrinsic biologic differences among clinical strains. The latter may reflect a particular strain's specific adaptation to the antimicrobial defenses of the host macrophage. For example, numerous studies have demonstrated that strains of Mycobacterium differ in susceptibility to reactive oxygen intermediates (ROI) and reactive nitrogen intermediates (RNI) . The role of ROI and RNI in controlling acute infections with M. tuberculosis has been fairly well studied in the murine system . However, little is known about the role of these intermediates during latent infections. Studies have demonstrated that the inducible form of nitric oxide synthase (iNOS) expression is required to control mycobacterial infection in mice . However, the induction of iNOS does not lead to complete elimination of M. tuberculosis in vivo . In addition, Flynn et al. demonstrated that M. tuberculosis reactivation occurs if the production of RNI is inhibited in a murine model of latency . This suggests that reactivation tuberculosis may result from the removal of host defense mechanisms, such as RNI, that keep M. tuberculosis from replicating in vivo. Consequently, it is possible that ROI and RNI may inhibit replication (bacteriostatic) but do not eradicate the bacteria (bactericidal). We decided to test this concept in vitro by using RNI and ROI susceptibility assays applied to both clinical and laboratory strains of mycobacteria. MATERIALS AND METHODS : The mycobacterial strains used in the present study are listed in Table . All strains of Mycobacterium were grown in Middlebrook 7H9 broth with 0.5% glycerol, 0.02% Tween 80, and 10% ADC enrichment to mid-log phase. Single-cell suspensions of each isolate were prepared according to a previously published method and quantified by enumeration of their CFU on Middlebrook 7H11 agar plates. A 0.1-ml aliquot of this suspension was added to 0.9 ml of 7H9 broth (pH 5.3 for ASN assays or pH 7.0 for H2O2 assays) containing various concentrations of either sodium nitrite (NaNO2) or hydrogen peroxide (H2O2). As a control, the culture was added to 7H9 broth at the appropriate pH without NaNO2 or H2O2 (unexposed). Each suspension was incubated at 37C for 24 h, plated onto 7H11 agar, and incubated at 37C for 3 weeks. At day 21 (or day 26 for H37Ra), the CFU recovered from isolates exposed to either ASN or H2O2 were marked, quantified, and compared to the CFU for recovered isolates not exposed to ASN or H2O2; the data were then calculated as the percent survival ([number of CFU exposed at day 21/number of CFU unexposed at day 21] x 100). Incubation at 37C was allowed to continue to day 50, at which time each plate was reexamined to enumerate additional CFU recovered during the interval. The total number of CFU was again compared to the total number of CFU from recovered isolates not exposed to ASN or H2O2 at 50 days of incubation, and the data were converted to percent survival ([number of CFU exposed at day 50/number of CFU unexposed at day 50] x 100). Each strain was tested similarly in triplicate, and each assay was repeated at least twice. Comparison of the mean CFU recovery of each strain was made by Student t test, and a P value of <0.05 was considered to show a significant difference. TABLE 1 | Mycobacterial strains used in this study RESULTS : Each strain of mycobacteria demonstrated similar levels of growth between days 21 and 50 when not exposed to ASN or H2O2 (Fig. and ). At day 21, at a lower concentration of ASN (3 mM), the percent survival of all strains of mycobacteria was <=80% compared to the same strains not exposed to ASN. At day 50, however, the percent survival of all of the clinical strains exceeded 80% . Compared to the laboratory strain H37Ra, at 3 mM ASN all of the clinical strains of M. tuberculosis and M. bovis (Ravenel and BCG) were significantly more resistant at day 21 or 50 of incubation (P < 0.05, Student t test) (Fig. and Table ). In addition, at 3 mM ASN the percent survival at day 21 versus that at day 50 was significantly different for each strain. On the other hand, at a higher concentration of ASN (6 mM), the percent survival of the strains at days 21 and 50 showed no significant difference, except for H37Ra and M. bovis Ravenel . At 6 mM ASN all of the strains, except M. bovis BCG, remained relatively resistant compared to H37Ra . At 2 mM H2O2, H37Ra and one clinical isolate (C.C.13) were most susceptible compared to the other strains tested, but no significant difference was detected between the two time points of incubation for each strain, except with H37Ra . At 5 mM H2O2 all of the strains except H37Rv and a clinical strain (C.C.13) remained relatively resistant compared to the laboratory strain H37Ra . At 5 mM H2O2, no significant differences in the percent survival were observed for each individual strain at day 21 versus day 50 except with H37Ra . Two of the clinical isolates (CB3.3 and CDC1551) consistently exhibited the highest resistance to killing by 5 mM H2O2 or 6 mM ASN compared to all of the other strains tested in the present study (Fig. and ). FIG. 1. | Mycobacterial ASN susceptibility assays. Mycobacterial ASN susceptibility assays. Mycobacterial isolates were incubated for 16 h in 7H9 medium (pH 5.3) and then plated onto 7H11 agar. CFU were determined at 21 and 50 days of incubation. (A) Isolates were incubated at 37C in 7H9 broth at pH 5.3 without NaNO2, and CFU were quantified at day 21 (shaded bars) and day 50 (hatched bars). The results are expressed as the CFU +- the standard error of the mean (SEM). (B) Isolates were exposed to 3 mM (day 21 [black diamonds] and day 50 [solid squares]) and 6 mM NaNO2 (day 21 [gray triangles] and day 50 [gray circles]). The results are expressed as the percent survival based on the CFU recovered from ASN-exposed strains relative to the CFU of unexposed strains +- the SEM. Both graphs are representative of triplicate experiments and show mean values of triplicate cultures for each strain. FIG. 2. | Mycobacterial H2O2 susceptibility assays. Mycobacterial H2O2 susceptibility assays. Mycobacterial isolates were incubated for 16 h in 7H9 medium (pH 7.0) and then plated onto 7H11 agar. CFU were determined at 21 and 50 days of incubation. (A) Isolates were incubated at 37C in 7H9 broth at pH 7.0 without H2O2, and CFU were quantified at day 21 (shaded bars) and day 50 (hatched bars). The results are expressed as CFU +- the SEM. (B) Isolates were exposed to 2 mM (day 21 [black diamonds] and day 50 [black squares]) and 5 mM H2O2 (day 21 [gray triangles] and day 50 [gray circles]). The results are expressed as the percent survival based on CFU recovered from strains exposed to H2O2 relative to the CFU of unexposed strains +- the SEM. Both graphs are representative of triplicate experiments and show mean values of triplicate cultures for each strain. TABLE 2 | Comparison to laboratory strain H37Ra of resistance levels of each strain to ASN and H2O2 DISCUSSION : Our results demonstrate that, at a low concentration of ASN, the tested Mycobacterium strains display a greater level of variation in percent survival at 21 days of incubation compared to that observed at 50 days of incubation. However, at high concentrations of ASN, these differences in the percent survival at both day 21 and day 50 disappeared for each strain. Significant differences in the percent survival of the clinical isolates at the two time points were not observed when they were exposed to either concentration of H2O2. These data suggest that the RNI products, but not H2O2, exert a bacteriostatic effect at low concentrations and a bactericidal effect at high concentrations. Considering that most mycobacterial strains were able to recover from exposure to lower concentrations of ASN when allowed to incubate beyond 3 weeks, it appears that the time of incubation after which CFU data are analyzed is important when similar assays are used to determine mycobacterial susceptibility to any type of stress. Although all of the strains in the present study exhibited some level of recovery to low concentrations of H2O2, the differences in the percent survival between day 21 and day 50 within each strain were not statistically significant, with the exception of one strain (H37Ra). One explanation for this observation is that the concentrations of H2O2 we used are bactericidal to Mycobacterium. Another possibility may be that most strains of mycobacteria, except H37Ra, are more resistant to H2O2 at 2 mM, and thus they are able to grow within 3 weeks. Hence, the concentrations of H2O2 used in the present study may have been either too low (2 mM) or too high (5 mM) to see significant differences among all of the strains when we compared the percent survival values at days 21 and 50. These hypotheses are in accord with other studies that have demonstrated either that H2O2 is bactericidal in vitro or that M. tuberculosis is relatively resistant to the effects of H2O2 in a cell-free system . Since H37Ra has a much slower growth rate than the other strains used, the significant level of recovery seen at all concentrations of ASN and H2O2 may be a consequence of not being able to visualize the colonies until much later. We tried to minimize this effect by determining the initial CFU of H37Ra at day 26 instead of at day 21, thus allowing the organism sufficient time to grow. Collectively, our data indicate that under the in vitro conditions used here exposure to low concentrations of ASN resulted in the stasis of most clinical isolates of M. tuberculosis. In contrast to these data, a study by Chan et al. found that chemically generated RNI at concentrations of between 1.0 and 10 mM are bactericidal to M. tuberculosis in vitro , although it should be noted that only one strain (Erdman) was analyzed and that the incubation period after which CFU were enumerated was not clearly defined. Our results are consistent with those of Rhoades and Orme , who demonstrated in vitro that the antimycobacterial activity of interferon-primed macrophages was bacteriostatic rather than bactericidal . In addition, Rhoades and Orme found that a high concentration (10 mM) of NaNO2 was mycobactericidal in a cell-free system and that lower concentrations (0.1 to 5.0 mM) showed a range of tolerance by clinical mycobacterial isolates . It should be noted that the cell-free assay used in the Rhoads and Orme study, although similar, was not identical to ours. Our study exposed mycobacterial strains for 16 h as opposed to a period of 10 days, and the CFU in the Rhoades and Orme study were determined at between 3 and 4 weeks. Despite the observed bacteriostatic effect of RNI, our results also demonstrate that some clinical M. tuberculosis isolates (CDC1551 and CB3.3) are able to resist even high concentrations of ASN or H2O2. These clinical isolates were previously shown to be associated with large outbreaks of tuberculosis and were found to exhibit high levels of resistance to RNI and H2O2 compared to other clinical isolates or laboratory strains of M. tuberculosis . These observations are consistent with numerous studies that have demonstrated strain-related differences regarding susceptibility to RNI and ROI , and suggest that certain strains of M. tuberculosis have evolved an enhanced level of resistance to the antibacterial mechanisms elicited by the host macrophage. The physiologic concentrations of ROI and RNI within human macrophages in vivo are not well established, although in vitro studies have determined that ca. 3.2 nmol of NO2-/105 cells and 287 nmol of H2O2/mg of protein/h are generated from resident murine macrophages stimulated with IFN-gamma . In addition, other studies have demonstrated that between 34 and 241 nmol of nitrite/106 cells in culture supernatants of human peripheral blood monocytes infected with M. tuberculosis and between 10 and 80 pmol of H2O2/mug of DNA from cells stimulated with polystyrene particles are generated. Hence, the amount of ROI and RNI added in the cell-free system used in the present study is higher than that expected to be generated in vivo. Therefore, relatively resistant M. tuberculosis strains, such as CB3.3 and CDC1551, are less likely to be kept in check within the host. Consequently, one could postulate that such strains are more capable of causing active disease more readily after an initial infection than strains that are relatively susceptible to RNI and ROI. In contrast, the proliferation of RNI- and ROI-susceptible strains may be controlled as long as the host is able to maintain RNI and/or ROI expression. In the absence of these stresses, such M. tuberculosis strains may resume proliferation and cause the reactivation of disease. While there may be multiple other host factors that contribute to bacteriostasis of M. tuberculosis in vivo, the observation of the bacteriostatic effect of RNI made in vitro, as well as the wide range of RNI and/or ROI susceptibilities of the clinical isolates, may provide new clues about the different clinical outcomes after M. tuberculosis infection. Backmatter: PMID- 12202552 TI - Identification of Anaplasma phagocytophila (Formerly Ehrlichia phagocytophila) Variants in Blood from Sheep in Norway AB - A total of 41 blood samples were collected from 40 Anaplasma phagocytophila-infected sheep in 11 sheep flocks from four different counties of southern Norway. The presence and nature of the Anaplasma species were identified by microscopic detection of morulae, PCR, reverse line blot hybridization, and 16S rRNA gene sequencing. A. phagocytophila was identified in all of the samples, and sequencing of the 16S rRNA gene revealed the presence of four variants of A. phagocytophila. Two of these variants have been described before, but two were newly identified 16S rRNA variants of this species. A. phagocytophila variant 1 was found in nine flocks, A. phagocytophila variant 2 was found in four flocks, the A. phagocytophila prototype was found in two flocks, and A. phagocytophila variant 5 was found in one flock. In two flocks, some sheep were infected with A. phagocytophila variant 1, whereas others were infected with A. phagocytophila variant 2, and in three animals a double infection with two variants was registered. Analyses of the blood samples revealed that blood from sheep infected with A. phagocytophila variant 2 contained nearly twice as many neutrophils and eight times as many Anaplasma-infected neutrophils as blood from sheep infected with the A. phagocytophila variant 1. Furthermore, only 43% of the A. phagocytophila variant 2-infected sheep displayed antibody responses in an immune fluorescence assay, whereas 93% of the sheep with the A. phagocytophila variant 1-infected sheep were seropositive. Keywords: Introduction : Tick-borne fever (TBF) in sheep caused by Ehrlichia phagocytophila and transmitted by the tick Ixodes ricinus was the first granulocytic ehrlichial infection to be described and has for decades been a well-known disease in domestic ruminants in several countries in Europe . E. phagocytophila belongs to the same genogroup as Ehrlichia equi and human granulocytic ehrlichiosis (HGE) agent, and natural infection with granulocytic Ehrlichia has now been reported in a variety of animal species . Recently, Dumler et al. reorganized the families Rickettsiaceae and Anaplasmataceae, and E. phagocytophila, E. equi, and the HGE agent were unified into the new species combination Anaplasma phagocytophila. For this reason we use A. phagocytophila as the emended name for this species throughout this study. TBF is a common disease in domestic ruminants along the coast of southern Norway . In 1995, more than 11,000 sheep flocks were treated prophylactically against TBF with tick repellent and/or insecticides, including ca. 40% of all flocks in Norway . In sheep, TBF is characterized by high fever, reduced milk yield, abortion, and reduced fertility in rams. The diagnosis was earlier based on the presence of inclusions (morulae) in circulating neutrophils in Giemsa-stained blood smears . A. phagocytophila infection in sheep is known to produce profound effects on the immunological defense system, which increases susceptiblility to disease and mortality from intercurrent infections such as Staphylococcus aureus pyaemia and Pasteurella haemolytica/trehalosi septicemia . Sheep flocks may suffer heavily on I. ricinus-infested pastures both due to direct mortality and to impairment of growth rate and production . In one flock investigated in Norway, almost one-third of the lambs died on Ixodes-infested pastures due to TBF and secondary infections . Lamb losses on I. ricinus-infested pastures may vary considerably between neighboring farms. The reasons for these variations are unknown but may be caused by differences in virulence between variants of Anaplasma. Such variations have earlier been found in both sheep and cattle . The identification of Anaplasma and Ehrlichia species is difficult because conventional bacteriological methods for cultivation and characterization cannot be used. Morphological and serological methods are also unreliable to differentiate Anaplasma and Ehrlichia species due to morphological similarities and antigen cross-reactivity between species . The purpose of the present study was therefore to identify and compare Anaplasma species from sheep with TBF from different areas of Norway by molecular methods. In addition, we wanted to study the number of neutrophils, their infection rate, and the antibody response in infected sheep. MATERIALS AND METHODS : Animals, blood samples, and hematology. | Blood samples were collected from Norwegian sheep with TBF in different I. ricinus-infested areas in Norway. A number of collaborating sheep farmers were informed before the tick season, and they were instructed to contact the local veterinarian for blood sampling when a suspected case of TBF was found in their flocks. TBF had earlier caused high mortality in all of these flocks, except in one flock (flock D; see below and Table ). The rectal temperatures of the actual sheep were measured, and whole blood and EDTA-blood samples were collected and sent to the Department of Sheep and Goat Research for further analyses. No further information of the animals was available after blood sampling. Hematological values, including total and differential leukocyte counts, were determined electronically from the EDTA-blood samples (Technicon H1; Miles, Inc.), and blood smears were prepared and stained with May-Grunwald Giemsa. A total of 400 neutrophils were examined on each smear by microscopy; the numbers of cells containing Anaplasma inclusions were recorded, and the percentages of infected neutrophilic granulocytes were calculated. The rest of the EDTA-blood was frozen at -20C until further analyses could be performed. Serology. | Serum samples were analyzed for the presence of antibodies to Anaplasma by an indirect immunofluorescence antibody assay . Briefly, twofold dilutions of sera were added to slides precoated with E. equi antigen (Protatek, St. Paul, Minn.). Bound antibodies were visualized by fluorescein-isothiocyanate-conjugated rabbit anti-sheep immunoglobulin (Cappel; Organon Teknika, West Chester, Pa.). Sera were screened for antibodies at a dilution of 1:40. If positive, the serum was further diluted and retested. A titer of 1.6 (log10 reciprocal of 1:40) or more was regarded positive. DNA extraction and PCR amplification. | DNA extraction on blood samples was performed according to Olsson Engvall et al. , with some modifications. Briefly, 400 mul of thawed EDTA-blood was treated with 220 mul of cold lysis buffer (10 mM Tris-HCl [pH 7.4], 100 mM EDTA), 0.5% sodium dodecyl sulfate, and 10 mul of proteinase K (20 mg/ml) and then mixed gently and incubated at 50C for 2 h. The mixture was mingled every 15 min, and after 1 h another 6 mul of proteinase K (20 mg/ml) was added. The mixture was then extracted twice with an equal volume of phenol-chloroform-isoamyl alcohol (25:24:1) and once with an equal volume of chloroform-isoamyl alcohol (24:1). DNA was precipitated by the addition of a 1/10 volume of 2 M sodium acetate (pH 6.5) and 2.5 volumes of cold ethanol (99%) and was then collected by centrifugation. The pellet was washed once in cold ethanol (70%), dried, and resuspended in 50 mul of sterile water, and the DNA concentration was then measured with a spectrophotometer (GeneQuant II; Pharmacia Biotech, Uppsala, Sweden). PCR amplifications were performed in a GeneAmp PCR system 9700 thermal cycler (Applied Biosystems, Nieuwerkerk a/d Ijssel, The Netherlands). The 5' part of the 16S rRNA gene of the Anaplasma species in the sheep blood samples were amplified in 50-mul volumes consisting of 25 mul of HotStarTaq mix (Qiagen, Hilden, Germany), 4 mul of primer 16S8FE (80 pmol), 4 mul of primer B-GA1B (80 pmol), 2 mul of tmpB spike DNA (10 fg) , 10 mul of water, and 5 mul of DNA sample. To minimize nonspecific amplification, a touchdown-up PCR program was used: 15 min at 94C, followed by two cycles of 20 s at 94C, 30 s at 65C, and 30 s at 72C; followed by two cycles under conditions identical to the previous cycles but with an annealing temperature of 63C. During subsequent two cycle sets, the annealing temperature was lowered by 2C until it reached 55C. We then carried out an additional 20 cycles of 20 s at 94C, 30 s at 55C, and 30 s at 72C, followed by 20 cycles of 20 s at 94C, 30 s at 63C, and 30 s at 72C, followed again by the touchdown program. The PCR was ended by an extra incubation for 7 min at 72C. Each time that the PCR was performed, negative (no sample added) and positive (Anaplasma or Ehrlichia DNA) control samples were included. Each sample was spiked with a critical amount (150 copies) of the tmpB spike control DNA to detect any inhibition of the PCR that might lead to false-negative results. When a relatively high concentration of Anaplasma DNA was available, the spike was weak or absent. In order to minimize contamination, the reagent setup, the sample addition, and the PCR and sample analysis were performed in three separate rooms, of which the first two rooms were kept at a positive pressure and had airlocks. Reverse line blot hybridization. | The reverse line blotting technique has been described before . Briefly, solutions with 5' amino-linked oligonucleotide probes were coupled covalently to an activated Biodyne C membrane in a line pattern by using a miniblotter (Immunetics, Cambridge, Mass.). After binding of the oligonucleotide probes, the membrane was taken from the miniblotter, washed, and again placed in the miniblotter with the oligonucleotide lines perpendicular to the slots. The slots of the miniblotter were filled with the biotin-labeled denatured PCR product, and hybridization was performed. The membrane was removed from the miniblotter, washed, and subsequently incubated with streptavidin-peroxidase to detect bound biotin-labeled PCR product. After a washing step, hybridization was visualized by incubating the membrane with enhanced chemiluminescence detection liquid (Amersham International, plc., Den Bosch, The Netherlands) and exposing the membrane to X-ray film. For species identification, the biotinylated PCR product was hybridized with 10 different Anaplasma- and Ehrlichia-specific oligonucleotide probes in the reverse line blot assay. All primers and probes are described in Table . DNA sequencing and data analysis. | Most PCR products were used directly for sequencing, but some were cloned into a TA-TOPO vector (Invitrogen, Groningen, The Netherlands). The plasmids were isolated and purified by using the Qiagen plasmid minikit and used for sequencing. The PCR products used for DNA sequencing were purified with QiaQuick PCR purification kits (Qiagen). Since PCR products were obtained in a PCR that included a spike control, the PCR yielded a mixture of Anaplasma PCR product and the tmpB spike. Therefore, we used sequence primers that were specific for the Anaplasma PCR product only (16SEhrSeq and GA1BSeq). For DNA sequencing reactions, the fluorescence-labeled dideoxynucleotide technology was used (Perkin-Elmer, Applied Biosystems Division). The sequenced fragments were separated, and data were collected with an ABI 3700 automated DNA sequencer (ABI, Applied Biosystems Division). The collected sequences were assembled, edited, and analyzed with the DNAStar package (DNAStar, Inc., Madison, Wis.). Statistics. | Statistical calculations on seroprevalences were performed by using the chi-square contingency test, and a two-sample t test was used for the hematological variables and the antibody titers (Statistix, version 4.0; Analytical Software). A P value of <0.05 was considered significant. Nucleotide sequence accession number. | The 16S rRNA gene sequences of the new variants of A. phagocytophila found in the present study are available in the GenBank database under the accession numbers and . TABLE 3 | A. phagocytophila variants from 11 sheep flocks in Norway identified by reverse line blot hybridization and DNA sequencing TABLE 1 | Oligonucleotide primers and probes used in PCRs and hybridization assays RESULTS : Blood samples. | Altogether 41 blood samples from 40 different sheep were collected. Two samples originated from the same sheep and were drawn 1 month apart. All sampled sheep revealed clinical signs of TBF, such as fever, increased respiration, dullness, and inappetence. Concurrent diseases were not observed. The samples were from 11 sheep flocks in four different counties of southern Norway and were collected from April to October in two consecutive years. The age of the sheep varied from 1 month to 2 years; however, most of the animals (68%) were less than 4 months old. Reverse blot line hybridization and DNA sequence analysis. | In order to confirm the results observed by the reverse line blot assay, the PCR products used in hybridizations were also sequenced. Although the sequence analysis largely confirmed the reverse line blot results, additional sequence variation was found. The blood samples carried A. phagocytophila that displayed minor sequence variation of the 16S rRNA gene and were designated variants. Two samples carried 16S rRNA gene sequences identical to the A. phagocytophila prototype (GenBank accession no. ) and a second group carried 16S sequences identical to the sequence with the accession number that differed at nucleotide position 80 from the prototype sequence and was designated variant 1. The largest group of sheep carried A. phagocytophila that differed at positions 80 and 100 of the 16S rRNA gene, and this type was designated A. phagocytophila variant 2 (accession no. ). One sample contained the new A. phagocytophila variant 5, which differed at position 93 of the 16S gene (accession no. ). The prototype and all variant signature sequences, including some other published sequences, are displayed in Table . A probe was designed to detect the A. phagocytophila variant 2, and all samples were retested on a reverse line blot that included this probe . This analysis was in complete concordance with the sequence analysis and confirmed that the observed sequence variation was not an artifact introduced by the sequencing procedure. When blood samples from a total of 11 sheep flocks were examined, A. phagocytophila variant 1 was found in nine flocks, A. phagocytophila variant 2 was found in four flocks, A. phagocytophila prototype was found in two flocks, and A.phagocytophila variant 5 was found in one flock. In two flocks, some sheep were infected with A. phagocytophila variant 1, whereas others were infected with A. phagocytophila variant 2. In three animals the PCR product reacted with two different Anaplasma probes in the reverse line blot, which might indicate a double infection with two different variants. DNA sequencing of these PCR products revealed ambiguous bases at a few positions in the sequence, supporting the supposition that double infection with two variants had occurred. For this reason the PCR products were cloned into a plasmid, and the inserts of 10 clones of each cloned PCR product were sequenced. This indeed revealed the simultaneous presence of two different variants in these three animals. Both samples of the one animal that was sampled twice carried the same variant of A. phagocytophila. In one flock, where 21 animals were examined, 3 (14%) were infected with A. phagocytophila variant 1, and 18 (86%) were infected with A. phagocytophila variant 2 . To exclude the possibility that A. phagocytophila carried two different copies of the 16S rRNA gene, we performed a Southern blot hybridization with a biotin-labeled 16S rRNA oligonucleotide probe on XbaI- and PstI-digested genomic A. phagocytophila DNA. This revealed the presence of a single 16S rRNA gene in the genome (data not shown). This result was not completely unexpected since BLAST searches in the E. chaffeensis genome sequence had also shown that this Ehrlichia species contains a single 16S rRNA gene. Clinical parameters, hematology, and serology. | Clinical variables at the time of blood sampling were obtained from 37 sheep. A marked and significant difference was found in the number of neutrophils. Blood samples from sheep infected with A. phagocytophila variant 2 contained nearly twice as many neutrophils as blood samples from sheep infected with A. phagocytophila variant 1. In addition, blood from A. phagocytophila variant 2-infected sheep carried eight times as many neutrophils with Anaplasma inclusions as blood from sheep infected with A. phagocytophila variant 1. The clinical parameters and hematology in sheep infected with different variants of granulocytic Anaplasma are shown in Table . Antibody titers to E. equi measured at the day of blood sampling are shown in Table . Only 24 of 39 (62%) of the acute Anaplasma-infected animals were found to be seropositive at the time of sampling. Remarkably, 93% of the A. phagocytophila variant 1-infected animals carried antibodies reacting with the E. equi antigen, whereas only 43% of the sheep infected with A. phagocytophila variant 2 were seropositive (P < 0.02). However, the mean antibody titer (log10) was not significantly different between sheep in these two variant groups. FIG. 1. | Reverse blot analysis of PCR products obtained from blood samples of A. phagocytophila Reverse blot analysis of PCR products obtained from blood samples of A. phagocytophila-infected sheep. The oligonucleotide probes are shown as lines in the horizontal direction, and the biotin-labeled PCR products are perpendicular in the vertical direction. Samples 1 to 3, samples from A. phagocytophila variant 1-infected sheep; samples 4 to 8, samples from A. phagocytophila variant 2-infected sheep; B, blank controls (no DNA added); P, A. phagocytophila prototype-positive PCR control; V1, A. phagocytophila variant 1-positive PCR control. TABLE 2 | The 5' end of the 16S rRNA gene sequences (bp 81 to 126) of different Anaplasma and Ehrlichia strains were determined and compared with similar sequences from GenBank TABLE 4 | Clinical variables (geometric mean +- SD) and antibody titer to E. equi antigen in 38 sheep infected with different variants of A. phagocytophila DISCUSSION : We found four 16S rRNA gene sequence variants of A. phagocytophila in blood from sheep suffering from TBF. To our knowledge, three of these variants have not earlier been identified in sheep, and two of them have not been identified in any other study before. Nucleotide differences at 16S rRNA level in A. phagocytophila have been found in isolates from rodents, deer, and Ixodes ticks . However, whether all variants can cause disease in humans and animals remains to be determined. Therefore, the importance of these sequence differences remains to be elucidated. In the study presented here, at least two of the variants found seem to cause classical A. phagocytophila infection in sheep. The sampled sheep were more than 1 month old. Age resistance in lambs older than 1 month and variation in clinical symptoms among Norwegian sheep breeds have not been found in experimentally A. phagocytophila-infected lambs . In the present study, the number of neutrophils, the number of infected neutrophils, and the serological response differed significantly between sheep infected with either A. phagocytophila variant 1 or 2. In the flock with few disease problems, 86% of the variants were of the A. phagocytophila variant 2 type. One possible reason for this difference could be that the A. phagocytophila variant 2 is better equipped to elude the immune system by inhibiting antibody response, resulting in more proliferation within granulocytes. This theory is supported by the observation in mice that pathology due to host immunity seems to play a more important role than pathogenicity of Anaplasma itself during infection with HGE . However, the role of host immunity in the pathogenesis of TBF in sheep has to be elucidated. Alternatively, the differences in morbidity and antibody response may be explained by sampling in the later phase of the infection in case of A. phagocytophila variant 1-infected sheep. However, later sampling may also have been caused by less-apparent acute disease manifestations in the A. phagocytophila variant 2-infected animals. The time point of sampling is important since earlier studies have shown that, in the later phase of the acute infection, both the number of neutrophils and the rate of infected neutrophils decrease . Different clinical and serological responses between variants of Anaplasma have earlier been observed in experimental infections in cattle, horses, and sheep . In the present study it was difficult to compare different clinical and serological values since only single point measurements were available. However, a recent experimental inoculation study in lambs of a single breed revealed a significant difference in the clinical, hematological, and serological responses between these two variants of A. phagocytophila (S. Stuen, K. Bergstrom, M. Petrovec, I. Van De Pol, and L. M. Schouls, unpublished data). The present serology results indicate that only 61% of the acute Anaplasma-infected animals were seropositive at the time of sampling. This result is in accordance with an earlier study in which 22 of 30 (73%) of Anaplasma-positive animals were found to be seropositive . An earlier experimental needle inoculation trial in sheep with A. phagocytophila-infected blood indicated that infected neutrophils may be found by Giemsa-stained blood smears examination several days before seroconversion appears . In addition, some Anaplasma-infected lambs have been found to remain seronegative for up to 6 weeks after the primary inoculation with a species similar to the HGE agent . Serologic investigation is therefore not reliable as the only diagnostic tool to detect acute Anaplasma infection in sheep. There has been much debate about the species definition and nomenclature of the group of granulocytic Ehrlichia. Recently, Dumler et al. clarified this by unifying E. phagocytophila, E. equi, and the HGE agent into a single species: A. phagocytophila. However, there are minor differences in the 16S rRNA gene of the latter species, and these differences can be used to differentiate particular groups within the species A. phagocytophila. Typing within species will require more polymorphism than the limited variation found in the 16S rRNA gene. Therefore, analysis of particular highly polymorphic sequences or of a number of housekeeping genes, such as those used in the multilocus sequence typing, are required for reliable discrimination of strain types. Variations in other genes, especially those coding for surface proteins, are more likely to affect properties such as virulence, host range, and interaction with arthropod vectors. Recently, it has been shown that sera from mice with high concentrations of antibodies that bind to the P44 proteins of the HGE agent or monoclonal antibodies specific to these proteins partially protect mice from the infection and that this protein may be located in the outer membrane of the HGE agent . This suggests that the P44-protein specific antibodies may play a role in the immunity against this infection and that the genes encoding the P44 outer membrane proteins may have a role in pathogenesis and immunresponse in A. phagocytophila infection in mice . In conclusion, the present study shows the existence of different A. phagocytophila variants in sheep: within different flocks, within each flock, and also within a single animal. Variants of A. phagocytophila causing TBF in sheep may accordingly exist simultaneously on the same pasture and may cause differences in both clinical and immunosuppressive reactions within each flock. However, the clinical and epidemiological consequences of these findings have to be further elucidated. Backmatter: PMID- 12202565 TI - Detection of Trichomonas vaginalis on Modified Columbia Agar in the Routine Laboratory AB - Broth culture of Trichomonas vaginalis is considered the "gold standard" for the diagnosis of trichomoniasis. Two studies were carried out to evaluate modified Columbia agar (MCA) for the isolation of T. vaginalis from clinical samples. Study I compared isolation on MCA to that on liquid medium with 889 vaginal samples. Out of 63 samples positive for T. vaginalis (7.1% of total), MCA identified 62 (98.4%) and broth identified 58 (92.1%). In study II, trichomoniasis was diagnosed within the scope of a screening program for a total of 39,585 men and women by culture on MCA and direct microscopy. Culture on MCA detected 199 (98.5%) and Gram staining detected 163 (80.7%) of 202 positive specimens. Wet-mount preparations used for symptomatic patients identified 103 (92.8%) of 111 cases. Culture of T. vaginalis from clinical samples on MCA is highly sensitive and reliable, as well as timesaving, and therefore suitable for screening of symptomatic and asymptomatic individuals. Keywords: Introduction : The parasitic protozoan Trichomonas vaginalis is a common pathogen that causes trichomoniasis and has been linked to preterm birth, acquisition of human immunodeficiency virus, infertility, and nongonococcal urethritis . Males are most frequently considered to be asymptomatic carriers, which represent an important vector and reservoir . According to the population studied, the prevalence of infection in females varies from 5 to 50% in patients attending sexually transmitted disease (STD) clinics . Although often observed in women with vaginal symptoms , trichomoniasis may be asymptomatic in up to 50% of infected individuals . Hence, diagnosis has to be based on laboratory procedures , such as direct microscopy and culture. Most frequently, the saline wet-mount preparation is used for observation of motile organisms. Different staining techniques, including those using Gram stain, Giemsa stain, Papanicolaou (Pap) smear, and acridine orange , and diverse molecularly based diagnostic methods, such as hybridization assay and PCR, have been employed to detect T. vaginalis but vary widely in their sensitivities and specificities (, , -, ). Thus, broth culture is still considered the "gold standard" for diagnosis of trichomoniasis . For this purpose, various liquid culture media have been developed . However, identification of positive samples requires frequent microscopic observations for up to 7 days. To evaluate the suitability of a solid medium cultivation technique in the routine laboratory, two studies were carried out. In study I, modified Columbia agar (MCA) was compared to a commercially available liquid medium in its ability to support trichomonal growth from clinical specimens. Subsequently, in study II, culture on MCA and direct microscopy were used to screen a low-risk population for T. vaginalis. Finally, the reliabilities of the methods employed were determined. MATERIALS AND METHODS : Study I. | A total of 889 women attending the Outpatients' Center in Vienna, Austria, because of symptoms and/or signs of infections of the genital tract, or for contact tracing, were included in the first study period between 1994 and 1995. Four swab samples from the posterior vaginal fornix were collected from each woman and processed at the bedside. The first two swabs were used for wet-mount preparation and Gram staining, respectively. Trichomonas medium and finally MCA plates were inoculated with the third and fourth samples, respectively. To investigate the ability of broth culture and MCA plates to grow different concentrations of T. vaginalis, an in vitro experiment was performed. Standardized inocula of 100, 101, 102, 103, and 104 organisms per ml (final concentrations) were inoculated into 5 and 17 ml of trichomonas broth as well as onto MCA plates. For quality control, T. vaginalis ATCC 30001 was processed in parallel and with all lots of media produced. Study II. | Altogether, 39,585 individuals (13,195 men and 26,390 women) presenting at the STD center in the years between 1996 and 2000 for routine microbiological examinations and for suspected infections of the genital tract with or without symptoms were screened for T. vaginalis. Trichomoniasis was diagnosed by culture on MCA and by Gram staining of vaginal samples from women and urethral specimens from men. For symptomatic patients, immediate wet-mount preparations were used. Specimens were sampled and processed as described above. Cultivation and identification. | Trichomonas medium (Oxoid Ltd., Basingstoke, Hampshire, England) was dissolved, autoclaved, and mixed with 10% sterile, inactivated lamb serum (PAA Laboratories, Linz, Austria), 1% dextrose (Sigma Chemical Co., St. Louis, Mo.), 1% malt extract (Oxoid Ltd.), and 1 ampoule of chloramphenicol selective supplement (Oxoid Ltd.). A 17-ml volume was dispensed into tubes with screw caps to ensure anaerobic growth conditions and to keep organisms viable until the end of incubation. MCA was prepared as follows: 13 g of Columbia agar (Oxoid Ltd.) was dissolved in 320 ml of distilled water, and the solution was adjusted to a pH of 6, cooked for 20 min, autoclaved at 121C for 5 min, and cooled before the addition of 50 ml of sterile, inactivated lamb serum (PAA Laboratories) containing 4 g of dextrose (Sigma Chemical Co.), 4 g of malt extract (Oxoid Ltd.), 15 ml of Baneocin (250 U of bacitron/ml and 5,000 U of neomycin/ml; Biochemie, Vienna, Austria), 10 ml of nystatin solution (10,000 U/ml; Biochrom KG, Berlin, Germany), 4 ml of penicillin-streptomycin solution (10,000 of penicillin/ml and 10,000 mug of streptomycin/ml; Gibco BRL, Life Technologies, Paisley, Scotland), and 1 ampoule of chloramphenicol selective supplement (Oxoid Ltd.). Approximately 10 ml of the resulting mixture was poured into 6-cm-diameter petri dishes. Both media were stored at 4C for no longer than 1 week and allowed to reach room temperature before use. Immediate incubation of inoculated broth and agar plates, which were held under anaerobic conditions (Anaerocult P; Merck KgaA, Darmstadt, Germany), was carried out at 37C for 6 days. Identification of T. vaginalis was achieved by observation of motile organisms in wet-mount preparations and by use of Gram-stained smears. Macroscopic evaluation of all cultures for turbidity in the case of trichomonal growth and of slide preparations from broth cultures was performed regularly. MCA was examined microscopically (magnification, x100), and results were confirmed by using a saline wet mount. Colonies of T. vaginalis on MCA cause a change of pH resulting in a muddy appearance of the plates. The characteristic puzzle-like structures caused by trophozoites lying close to each other can be observed most easily at the edge of the colony. RESULTS : Study I. | Out of the 889 women enrolled in study I, a total of 63 (7.1%) were positive for T. vaginalis by direct microscopy and/or culture. Growth of T. vaginalis was observed in 58 (92.1%) of 63 samples in broth culture and in 62 (98.4%) of 63 specimens in culture on MCA, resulting in an overall accordance of 93.5%. Altogether, 10 (17.2%) of 58 positive cultures in trichomonas broth showed low numbers of organisms, whereas 2 (3.2%) of 62 positive MCA plates showed low numbers of organisms. One positive sample was detected by direct microscopy only, while this method missed a total of six infected samples that were identified by culture on MCA. With broth culture, three of these samples were positive and another two showed low numbers of organisms, leaving the infection in one specimen not detected. A total of five infections would have been missed if broth culture only had been performed. No sample positive on liquid medium but negative on MCA plates was observed. In vitro experiment. | MCA plates were positive with all standardized inocula within 6 days of incubation. After 24 h, plates that had been inoculated with 102 trichomonads per ml or more were positive, while inocula of 101 trichomonads gave positive results after 72 h. All broth cultures inoculated with 100, 101, or 102 trichomonads per ml stayed negative for the whole examination period, regardless of the volume used. Both the 5-ml volume and the 17-ml volume showed positive results with inocula of 104 and 103 trichomonads after 24 and 48 h, respectively. Study II. | Screening for T. vaginalis performed with a total of 39,585 men and women attending the STD center identified 202 infected individuals (7 men [0.05%] and 195 women [0.74%]), revealing a prevalence of 0.5%. Table shows the results obtained for all specimens collected. Gram staining of smears and culture on MCA were performed to detect trichomonads in both asymptomatic and symptomatic patients. Wet-mount preparations were used for samples from 111 symptomatic patients (3 men and 108 women), or 55% of infected individuals. Culture on MCA detected 199 (98.5%) of 202 infections in symptomatic as well as asymptomatic individuals, with sensitivities of 100 and 97.3%, respectively . Infections in three samples could be detected by direct microscopy only, that in one specimen was detected with a wet-mount preparation as well as Gram staining and those in two samples were detected by using wet mount exclusively. Gram staining identified 163 (80.7%) of 202 infected persons, with infections in 12 symptomatic and 27 asymptomatic individuals not detected. For 111 samples from patients with clinical symptoms, wet-mount preparations detected infections in 103 (92.8%) specimens but failed to detect 8 positive specimens, and only 1 of these samples was correctly diagnosed by Gram staining. For 91 samples from T. vaginalis-positive patients, wet-mount preparations were not used due to the lack of clinical symptoms. Performance of wet-mount microscopy in addition to culture increased the detection rate from 97.3 to 100%. TABLE 1 | Identification of T. vaginalis in a total of 39,585 individuals by Gram staining, culture on MCA, and wet-mount microscopy TABLE 2 | Sensitivities of different diagnostic techniques in symptomatic and asymptomatic individuals with trichomoniasis DISCUSSION : A low rate of infection with trichomonads, only 0.5%, was detected in this study population, and this rate is in accordance with observations in other countries . However, T. vaginalis should be included in STD screenings, not only because it is associated with gonorrhea and Chlamydia trachomatis infection , but also because screening for T. vaginalis is considered an achievable strategy to reduce the incidence of human immunodeficiency virus . The requirements of diagnostic techniques for large-scale screenings with respect to sensitivity, specificity, and duration of test results are very high. Although enzyme immunoassays, PCR, and hybridization-based detection of T. vaginalis meet some of these requirements, performance of these tests is restricted to specialized laboratories because of the high costs involved. Moreover, these tests may not always be commercially available. Although wet-mount microscopy is a cost-saving method, a sensitivity of approximately 40 to 80% reduces its reliability as a diagnostic tool . The data presented in this study reveal a sensitivity of 92.8% for wet-mount microscopy performed on samples from symptomatic patients. However, the reliability of clinical symptoms is low, because in the present study as many as 91 individuals, or 45% of infected patients did not show typical signs of infection. An unfortunate limitation of this study due to the great expense of a large-scale screening is that there are no data available on wet mount results for asymptomatic patients. Most likely, these individuals harbor small numbers of trichomonads and/or a reduction in, or loss of, the parasites' motility causes a negative outcome with this test . Pap smears are often performed in gynecologic screenings and have a reported sensitivity of approximately 60 to 70% . However, an error rate of about 48% due to false-negative and false-positive results has been observed when Pap smears are used as the only diagnostic method . Staining techniques, such as with acridine orange or Giemsa stain, show sensitivities of about 60 and 50%, respectively, while Gram staining is not recommended to diagnose trichomoniasis because organisms may appear similar to polymorphonuclear leukocytes . However, Gram-stained smears are frequently implemented for microbiological examination in the routine laboratory. In this study, results obtained with Gram staining demonstrated sensitivities of 89.2% in the symptomatic patient and 70.3% in the asymptomatic patient. So, in contrast to published data, Gram staining was equivalent to the wet-mount or Pap smear technique and may as well be substituted for other diagnostic techniques if immediate microscopic examination is impracticable. On the other hand, additional Gram staining did not significantly increase the detection of T. vaginalis by culture. Finally, the outcome of microscopic examinations of stained smears is subjective and dependent on the experience of microscopists, and results need to be confirmed by culture. The reported sensitivity of culture in liquid media, such as Diamonds and Trichosel, is 85 to 95% . The most important disadvantage of broth culture is the indispensable and time-consuming slide preparation needed for identification of positive cultures over a period of up to 7 days. Culture systems, such as the InPouch TV test , the plastic envelope method , or the microtiter tray culture described by Hill , have been tried to facilitate examination. Semisolid or so-called pour-plate culture techniques to isolate and/or quantitate T. vaginalis organisms have been described previously . However, these techniques are not suitable if large numbers of clinical specimens are to be examined. MCA has been developed to overcome these problems and is to our knowledge the first solid medium suitable for the isolation of T. vaginalis from clinical samples in the routine laboratory. Data presented in this study reveal that MCA has a sensitivity of 98.5% in detection of T. vaginalis in a low-risk population comprising symptomatic as well as asymptomatic patients. The in vitro experiment demonstrated MCA to be superior to broth culture in detecting low numbers of trichomonads, confirming the results of this study. Compared to broth culture, the agar plate technique reduces the risk of missing positive specimens by making possible the microscopic examination of the whole clinical material obtained. This solid medium is a timesaving culture technique because it eliminates the need for slide preparations, and it is favorable for screening a low-risk population if more than one sample is inoculated onto a single plate. Since wet-mount and stained-smear methods show low sensitivities for detection in asymptomatic individuals, culture of T. vaginalis can be recommended as the only diagnostic method for this patient group and should therefore not be replaced by these tests. However, this culture technique has a turnaround time of 2 to 6 days. Sometimes rapid tests, such as enzyme immunoassays or PCR, are preferred. An evaluation of the methods and MCA will have to be carried out in the future. In summary, MCA performed as a highly sensitive, reliable, and easy-to-handle culture technique in the routine laboratory for diagnosis of T. vaginalis infections in clinical samples from symptomatic as well as asymptomatic individuals and may provide a suitable screening method for trichomoniasis. Backmatter: PMID- 12202550 TI - Development and Evaluation of Rapid Urinary Antigen Detection Tests for Diagnosis of Penicilliosis Marneffei AB - Penicilliosis, caused by the dimorphic fungus Penicillium marneffei, is an important opportunistic systemic fungal infection affecting immunocompromised individuals living in areas where penicilliosis is endemic. We have demonstrated previously that a urinary enzyme-linked immunosorbent assay (ELISA) with purified rabbit polyclonal antibody against killed whole-fission-form arthroconidia of P. marneffei was specific and highly sensitive for the diagnosis of penicilliosis. In this study, a dot blot ELISA and a latex agglutination (LA) test were developed with the same polyclonal antibody and compared with the ELISA for the detection of P. marneffei urinary antigen. Urine specimens from 37 patients with culture-proven penicilliosis and 300 controls (52 healthy subjects and 248 hospitalized patients without penicilliosis) were tested. Antigen was detected in urine from all 37 (100%) penicilliosis patients by the LA test, 35 (94.6%) penicilliosis patients by the dot blot ELISA, and 36 (97.3%) penicilliosis patients by the ELISA. False-positive results were found by the three assays for 2 (0.7%), 8 (2.7%), and 6 (2%) of 300 controls, respectively. The overall sensitivities of the diagnostic tests were as follows: dot blot ELISA, 94.6%; ELISA, 97.3%; and LA test, 100% (specificities, 97.3, 98, and 99.3%, respectively). The LA test is simple, robust, rapid, and convenient and should prove to be an important addition to the existing diagnostic tests for penicilliosis. Keywords: Introduction : Penicilliosis, caused by the dimorphic fungus Penicillium marneffei, is increasingly common in southeast Asia, particularly Thailand . Human infection with P. marneffei was rare prior to the AIDS epidemic in this area, but the number of cases has since risen dramatically. In Thailand, penicilliosis is a major opportunistic infection in AIDS patients, particularly in the north of the country . Infection with P. marneffei usually presents as a disseminated and progressive illness in patients with CD4 counts below 100, with fever, anemia, leukopenia, weight loss, diarrhea, cough, massive hepatosplenomegaly, generalized lymphadenopathy, and characteristic multiple umbilicated skin lesions particularly affecting the face and upper body . Penicilliosis is fatal if it is left untreated but responds well to antifungal agents such as amphotericin B and itraconazole . A presumptive diagnosis of P. marneffei infection is made on clinical grounds in conjunction with microscopic identification of characteristic septate yeast cells in smears of skin lesions, blood, or bone marrow. Definitive diagnosis relies on the isolation of P. marneffei by culture from clinical specimens. However, cultures often require at least 3 days of incubation before definitive identification of the fungus is possible. Identification of P. marneffei by means of histological or immunohistochemical techniques and exoantigen tests has been reported. These tests are specific, but time-consuming and laborious. Serological methods have also been developed. Detection of antibodies to P. marneffei by use of an immunodiffusion test , an indirect fluorescent-antibody technique , and an enzyme-linked immunosorbent assay (ELISA) antibody detection test have each been described, but these tests, while specific, are less sensitive than culture. Antigen detection tests have also been studied . We have shown previously that P. marneffei antigen can be detected at high titers in urine by ELISA with a fluorescein isothiocyanate (FITC)-labeled purified rabbit polyclonal hyperimmune immunoglobulin G (IgG) . This test is highly sensitive and specific. We now report on the development of a simplified dot blot ELISA and a latex agglutination (LA) test for detecting P. marneffei antigenuria by using the polyclonal anti-P. marneffei antibody used previously in the urinary antigen ELISA. The dot blot ELISA and LA tests were clinically evaluated in an area of northeast Thailand where penicilliosis is endemic and were compared with the urinary antigen ELISA for the diagnosis of penicilliosis. MATERIALS AND METHODS : Patients and urine samples. | The study was conducted between June 1995 and July 2000. Adult patients with suspected or confirmed penicilliosis admitted to Sappasitprasong Hospital, Ubon Ratchathani, northeast Thailand, were included in the study. All patients were seen by a member of the study team, and full clinical details were recorded on a standard form. Routine hematological and biochemical tests were performed, and the patients were evaluated for the presence of human immunodeficiency virus (HIV) antibody when indicated. The diagnosis of penicilliosis was confirmed by isolation of P. marneffei from blood, urine, lymph node biopsy, and liver aspirate specimens or throat or skin lesion swab specimens; a presumptive diagnosis was made by direct identification of septate yeast cells in stained smears made from appropriate clinical specimens. A total of 37 urine samples were obtained on admission from HIV-positive patients confirmed to have penicilliosis. Control urine samples were obtained from healthy individuals (n = 52); HIV-seropositive patients with other fungal infections (total, n = 34; cryptococcosis, n = 31; histoplasmosis, n = 1; candidiasis, n = 2); patients with melioidosis (n = 168), other bacterial causes of septicemia (n = 12), and other bacterial infections (n = 7); and inpatients for whom all cultures were negative (n = 27). All urine samples were stored at -30C and thawed only at the time of testing. Antigen preparation. | Antigen was prepared as described previously . Briefly, killed fission-form arthroconidia of a clinical isolate of P. marneffei (grown in brain heart infusion broth) were used to immunize rabbits and as the control antigen for the ELISA, dot blot ELISA, and LA test. Anti-P. marneffei IgG preparation. | Hyperimmune rabbit sera with antibodies against P. marneffei were prepared as described previously . The purified IgG fraction was used in the ELISA, dot blot ELISA, and LA test. ELISA. | Urinary P. marneffei antigen was detected by a sandwich ELISA, as described previously . The ELISA uses a polyclonal anti-P. marneffei antibody conjugated to FITC and an anti-FITC antibody amplification system. A titer of 1:40 or greater was considered a positive result for P. marneffei antigen. Dot blot ELISA. | Optimal incubation conditions and reagent concentrations for the dot blot ELISA were determined by checkerboard titration. Urine samples were boiled for 6 min and centrifuged at 5,000 x g for 6 min to remove any precipitate before testing. A 2-mul aliquot of each urine sample was then dotted onto an 11.5-cm2 nitrocellulose (NC) membrane sheet (Bio-Rad Laboratories, Richmond, Calif.) at 1.3-cm intervals. The blotted NC sheet was dried at 37C for 30 min and subsequently blocked with blocking buffer consisting of 2% bovine serum albumin (Sigma Chemical Company, St. Louis, Mo.) in phosphate-buffered saline (PBS) containing 0.02% Tween 20 (PBS-T) for 20 min at room temperature. After the NC sheet was washed with PBS-T, it was incubated with purified rabbit anti-P. marneffei IgG diluted 1:500 in blocking buffer for 1 h at room temperature. The sheet was washed once and was incubated for 1 h at room temperature with alkaline phosphatase-conjugated swine anti-rabbit immunoglobulin (Dakopatt, Copenhagen, Denmark) diluted 1:500 in blocking buffer. After further washing, the NC sheet was immersed in substrate solution (freshly prepared by mixing 330 mul of nitro blue tetrazolium [10 mg/ml] with 33 mul of 5-bromo-4-chloro-3-indolyl phosphate p-toluine salt [50 mg/ml of 100% dimethyl formamide; Sigma] and diluted to 10 ml with substrate buffer [0.1 M Tris buffer {pH 9.5} containing 100 mM sodium chloride and 5 mM magnesium chloride]) and then incubated for 10 min at room temperature to allow color development before it was rinsed with stopping buffer (0.02 M Tris buffer containing 0.2 M disodium EDTA). The NC sheet was then washed with distilled water and air dried. A purple dot indicated a positive reaction, and a pale brown dot or a clear area was read as a negative reaction. The same controls used in the ELISA were included in this assay. All urine samples and controls were tested in duplicate. Preparation of LA test. | A purified IgG fraction of the pooled rabbit immune serum with antibodies against P. marneffei was adjusted to a concentration of 15 to 20 mg/ml and used to coat the test latex particles. A 1% suspension of latex polystyrene beads (diameter, 0.6 mum; Sigma) in glycine-buffered saline (pH 8.2) was sensitized by mixing of the beads with purified rabbit anti-P. marneffei IgG at 37C for 2 h before the addition of bovine serum albumin to a final concentration of 0.5% (wt/vol). The optimal quantity of immunoglobulin used was the highest dilution that produced a clear agglutination with the highest reactive dilution of P. marneffei culture suspension (starting at a concentration of 106 yeast cells/ml in a box titration). Control latex particles were coated with purified, pooled normal rabbit IgG. The suspensions were stored at 4C and brought to room temperature before use. LA method. | All urine samples were boiled for 6 min, cooled, and centrifuged at 5,000 x g for 6 min to remove any precipitate before testing. The LA test was performed by placing 15 mul of test and control latex suspensions on a clean glass slide. Five microliters of urine was added to the latex particles. After the components were mixed, the slide was rocked gently for 3 min. Agglutination with the test latex but not the control latex indicated a positive result, while a lack of agglutination with either latex indicated a negative result. Agglutination with both the test and the control latexes was interpreted as a nonspecific (negative) reaction. Fungal and bacterial strains. | To evaluate the specificities of the dot blot ELISA and the LA method, Merthiolate-killed brain heart infusion broth cultures (20-ml broth volumes) of various fungal species (Penicillium griseofulvum ATCC 48166, Penicillium chrysogenum ATCC 9480, Penicillium notatum ATCC 9478, Aspergillus terreus, Aspergillus fumigatus BCC 123, Aspergillus flavus BCC 235, Histoplasma capsulatum var. capsulatum, Candida albicans [two strains], Candida kefyr, Cryptococcus neoformans var. neoformans, Cryptococcus neoformans var. gattii, Sporothrix schenkii, and Trichosporon beigelii) were tested at various concentrations, together with seven clinical strains of P. marneffei (three hyphal forms and four yeast-like forms). The cultures were incubated for 48 h at 37C (Candida, Cryptococcus, and Trichosporon), 4 days at 37C (P. marneffei yeast phase), or 4 days at room temperature (P. marneffei hyphal form and all other species). The final yeast cell or spore counts in broth were determined with a counting chamber. Following subculture to ensure purity, the organisms were killed by overnight incubation with Merthiolate (final Merthiolate concentration, 0.2%). Appropriate test dilutions were then made in sterile urine (boiled) from a healthy human volunteer. Strains of Burkholderia pseudomallei (strain 203a), Salmonella enterica serovar Enteritidis, and Staphylococcus aureus, which are common causes of systemic infection in Ubon Ratchathani, were also tested, in addition to a strain of S. enterica serovar Typhi. These fungal and bacterial control strains were all as described previously . All samples were tested in duplicate, and each test was repeated three times. Statistical analysis. | Data were analyzed with SPSS for Windows (version 9.0; SPSS Inc., Chicago, Ill.) computer software. Quantitative variables with a nonnormal distribution were expressed as medians and ranges. Qualitative variables were expressed as frequencies and percentages. The sensitivities (proportions of positive samples correctly identified by the test) and the specificities (proportions of negative samples correctly identified by the test) of the dot blot ELISA and LA test were compared with those of the ELISA and evaluated by using the method of Galen . The sensitivities and specificities of the dot blot ELISA and LA test were compared with those of the ELISA by using the McNemar test. The measures of agreement between any pair of these three tests were assessed by use of the kappa statistic, which was interpreted as the chance corrected proportional agreement between the methods. A P value of <0.05 was considered significant. RESULTS : Specificities of tests for P. marneffei. | Specificity results for the dot blot ELISA, the LA test, and the ELISA are presented in Table . All strains of P. marneffei culture supernatant, from both yeast and hyphal forms, gave positive results by the dot blot ELISA as well as by the LA test. In the LA test, the time for the appearance of agglutination for yeast-form whole-cell culture supernatants of P. marneffei (median, 4 s [range, 3 to 6 s]; n = 4) was significantly shorter than that for culture supernatants of the hyphal form (median, 7 s [range, 7 to 15 s]; n = 3) (P = 0.031). The lowest detection limits for P. marneffei antigen in both the dot blot ELISA and the LA test were 103 yeast cells/ml. Similar titration results were found when either PBS or urine was used as the diluent. These findings were similar to those of the established ELISA. However, the ELISA was more sensitive; it gave positive results at 102 yeast cells/ml. All the other fungal culture supernatants gave negative results in the dot blot ELISA, the LA test, and the ELISA up to concentrations of 108 cells/ml. All bacterial strains except S. aureus were nonreactive in these tests; S. aureus gave positive results in the dot blot ELISA and ELISA (but not the LA test), but only at a concentration of 108 cells/ml. Detection of P. marneffei antigen in urine by dot blot ELISA. | The results of the dot blot ELISA with urine from patients with penicilliosis and the control groups are given in Table . The results were compared with those of the established ELISA. Of the 37 HIV-positive patients with culture-confirmed P. marneffei infection, 35 (94.6%) were positive for P. marneffei urinary antigen by the dot blot ELISA and 36 (97.3%) were positive by the ELISA at a cutoff titer of >=1:40. All 52 samples from healthy volunteers were negative by the dot blot ELISA. Of the 34 samples from patients with other fungal infections, 2 (5.9%) were positive by the dot blot ELISA. False-positive results were also found for 2 (1.2%) of 168 urine samples from patients with melioidosis, 1 (8.3%) of 12 urine samples from patients with other bacterial septicemias, 1 (14.3%) of 7 urine samples from patients with other bacterial infections, and 2 (7.4%) of 27 subjects who were culture negative. Thus, false-positive results were found for 8 (2.7%) of 300 control urine samples overall by the dot blot ELISA. This compares with 6 (2%) of 300 control urine samples that were positive by ELISA. Detection of urinary antigen by the LA test. | All 37 samples from confirmed penicilliosis patients were positive by the LA test. The results are shown in Table . The median time to agglutination was 30 s (range, 15 to 180 s); i.e., the time to agglutination was longer than that observed with whole-cell cultures of P. marneffei. All except 2 of the 300 control urine samples were negative; these 2 (5.9%) samples were from among the 34 patients with other fungal infections (the 2 samples were both reactive by the dot blot ELISA and the ELISA, and both patients were HIV antibody positive). Thus, false-positive results were found for only 2 (0.7%) of 300 control urine samples overall by the LA test. Only one urine sample (from a patient in the "other septicemia" group) gave a nonspecific reaction. Diagnostic sensitivity and specificity. | The sensitivity and specificity of the dot blot ELISA, the ELISA, and the LA test are summarized in Table . By using a cutoff titer of 1:40 in the ELISA, the overall sensitivity for penicilliosis was 97.3% and the specificity, using data for all control groups, was 98% (positive predictive value [PPV], 85.7%; negative predictive value [NPV], 99.7%). The sensitivity of the dot blot ELISA was lower, at 94.6%, with a specificity of 97.3% (PPV, 81.4%; NPV, 99.3%). The sensitivity and specificity of the LA test were 100 and 99.3%, respectively (PPV, 94.9%; NPV, 100%). There were no significant differences between the three tests (dot blot ELISA versus ELISA, P = 0.99; LA test versus ELISA, P = 0.45; and dot blot ELISA versus LA test, P = 0.29). In addition, there was significant agreement between each pair of tests (dot blot ELISA versus ELISA, kappa = 0.825 and P < 0.001, LA test versus ELISA, kappa = 0.902 and P < 0.001, and dot blot ELISA versus LA test, kappa = 0.889 and P < 0.001). These results indicate that although the LA test achieved the highest sensitivity and PPV for detection of P. marneffei antigen in urine, there is little in terms of performance from which to choose among the three tests for the diagnosis of penicilliosis. TABLE 1 | Reactivity of purified rabbit anti-P. marneffei IgG against control organisms in dot blot ELISA, LA test, and ELISA TABLE 2 | Comparison of dot blot ELISA, LA test, and ELISA for detection of P. marneffei antigen in urine specimens TABLE 3 | Sensitivity and specificity of P. marneffei urinary antigen detection by dot blot ELISA, LA test, and ELISA DISCUSSION : Penicilliosis is now recognized as one of the most important opportunistic infections in AIDS patients in Thailand, affecting 6.8% of AIDS patients in the north of the country . At present, the diagnosis of P. marneffei infection relies on the identification of yeast cells in smears of skin lesion, blood, or bone marrow specimens, but there may be confusion with histoplasmosis or disseminated cryptococcosis. Culture is often the only way to confirm the diagnosis, particularly if characteristic skin lesions are absent, but this may take several days. Thus, a rapid diagnostic or confirmatory test would be of clinical value, and its results could be used to direct the appropriate antifungal therapy. Rapid identification of fungal pathogens, once isolated in pure culture, can be achieved with specific DNA probes , but these tests have not yet been applied to clinical specimens for rapid diagnosis. There have been few advances in diagnostic tests for penicilliosis since a previous report . Several specific protein antigens of P. marneffei have been identified to be immunologically reactive in Western blot assays for approximately 50% of serum specimens from patients with penicilliosis and 20% of serum specimens from patients with cryptococcosis or candidiasis . A mannoprotein antibody ELISA has been reported to be positive for 80% of HIV-positive penicilliosis patients , but serum specimens from patients with other fungal infections were not included in that study, and there have been no reports of prospective clinical evaluations. Thus, the usefulness of antibody tests for diagnosis of penicilliosis in areas where penicilliosis is endemic remains uncertain. Antigen detection tests may be more appropriate than antibody detection tests for the diagnosis of penicilliosis in immunocompromised patients, in whom antigen loads are high and antibody responses may be muted . Detection of P. marneffei antigen also suggests active rather than past infection, which is particularly relevant in an area of endemicity. A fluorescent-antibody test for the tissue form of P. marneffei that uses a rabbit polyclonal antibody raised against whole P. marneffei yeast cells has been described , as has an LA test for detection of P. marneffei antigen in serum and urine (this was evaluated with a small number of patient and control serum samples, and only two urine samples from penicilliosis patients were tested) . Prospective clinical evaluations of these tests have not been reported. A cell wall mannoprotein antigen has been demonstrated in the serum of 65% of 26 penicilliosis patients; the sensitivity was increased to 88% if this test was used in conjunction with an antibody detection test . A urinary antigen detection test with a rabbit IgG antibody developed for diagnosis of H. capsulatum var. capsulatum infection has been reported to give positive results for 17 of 18 confirmed penicilliosis patients . More recently, we reported that a rabbit polyclonal antibody raised against killed whole-fission-form arthroconidia of P. marneffei could be used to quantify urinary antigen by ELISA . At a cutoff titer of 1:40, this urinary antigen ELISA had a diagnostic sensitivity of 97% and a specificity of 98%. The polyclonal rabbit antibody used previously in the ELISA was used in the present study for the development of the dot blot ELISA and the LA test. Each of our tests, the dot blot ELISA, the ELISA, and the LA test, appears to be highly sensitive and specific. The LA test detected antigen in all 37 urine specimens from patients with culture-confirmed penicilliosis, whereas the dot blot ELISA and ELISA detected antigen in 35 and 36 urine specimens, respectively. The two specimens with false-negative dot blot ELISA results had ELISA titers of only 1:40 (the cutoff titer). Of the remaining 35 dot blot ELISA-positive specimens, 34 had ELISA titers of 1:320 or greater. One dot blot ELISA-positive undiluted urine specimen had antigen detectable only by ELISA (the patient presented initially with disseminated cryptococcal infection). The LA test was also rapid; agglutination occurred in all 37 positive specimens by 180 s (median, 30 s). Only two positive results were found among 300 control specimens; these specimens were positive by all three tests. Both specimens were from HIV-positive patients with cryptococcosis, and we cannot exclude the possibility that they were also infected with P. marneffei, although we were unable to culture it from suitable specimens. False-positive results by the dot blot ELISA (eight patients) were slightly more common than false-positive results by the ELISA (six patients). Two patients gave false-positive results by both the dot blot ELISA and the ELISA; one of these patients was also HIV positive and presented with a Salmonella septicemia, while a coliform organism was cultured from the urine of the other patient (whose HIV infection status was unknown); again, we are unable to exclude penicilliosis in these patients. The antigens detected by the polyclonal rabbit antibody remain to be characterized. However, it is evident from our studies that detection of antigen in urine with this antibody is highly specific and sensitive whether the antibody is used in an ELISA, a dot blot ELISA, or an LA test. There is little in terms of performance from which to choose among the three tests. However, the LA test is by far the more preferable for routine diagnostic use: it is a simple, convenient, rapid, and reliable method for detecting P. marneffei antigenuria, which should allow its widespread use even where laboratory facilities are very limited. The ELISA and dot blot ELISA do not offer any advantages over the LA test and are considerably more laborious and time-consuming. The LA test reagent can be readily produced by sensitizing latex particles with our rabbit polyclonal antibody, and it should prove useful as a diagnostic test for P. marneffei infection. It should therefore be developed further. We also hope to evaluate this test for the detection of antigenemia and use it to assess responses to antifungal therapy in a study of serial urinary antigen testing during treatment and subsequent follow-up. Backmatter: PMID- 12202595 TI - Molecular Analysis of the Pathogenicity Locus and Polymorphism in the Putative Negative Regulator of Toxin Production (TcdC) among Clostridium difficile Clinical Isolates AB - The pathogenicity locus (PaLoc) of Clostridium difficile contains toxin A and B genes and three accessory genes, including tcdD and tcdC, which are supposed to code for the positive and negative regulators of toxin expression, respectively. Different studies have described variations in C. difficile toxin A and B genes, but little is known about C. difficile variants for the accessory genes. The PaLoc of several C. difficile clinical isolates was investigated by three different PCR methods with the aim to identify variant strains. Of the toxinogenic C. difficile strains examined, 25% showed variations. No correlation between C. difficile variant strains and key patient groups was found. Interestingly, all of these strains showed a variant tcdC gene. Three different tcdC alleles were identified, and one of these had a nonsense mutation which reduced the TcdC protein from 232 to 61 amino acids. It is possible that different TcdC variants affect toxin production differently, a hypothesis with important implications for the pathogenic potential of variant C. difficile strains. Keywords: Introduction : C. difficile is the etiologic agent of pseudomembranous colitis and the most common cause of nosocomial antibiotic-associated diarrhea . Toxins A and B, C. difficile virulence factors, belong to the large clostridial cytotoxins, and both disrupt the actin cytoskeleton . The toxin A and B genes (tcdA and tcdB) are part of the pathogenicity locus (PaLoc), a 19.6-kb genetic locus that also includes three additional tcd open reading frames (ORFs), tcdD, tcdE, and tcdC, and the ORFs for the insertion sequences, cdu-2, cdu-2', cdd-2 cdd-3, and cdd-4 . Sequencing and transcription analysis suggest that TcdD and TcdC are involved in the positive and negative regulation of TcdA and TcdB expression, respectively . Different studies have described variations in C. difficile toxin A and B genes . Despite the fact that variant strains can still be associated with clinical diseases, few epidemiological data on their circulation are reported and little is known about PaLoc accessory gene variants . We analyzed several C. difficile clinical isolates to investigate their PaLocs, identify variant strains, and determine possible correlations with a particular patient population. Three different PCR-based methods were used to detect the PaLoc accessory genes , the variations in the toxin A and B genes , and the presence of the binary toxin genes . The majority of C. difficile strains with dramatic variations in toxin A and B genes harbor the binary toxin genes, so their detection is suggested as a method for a quick identification of these strains . Bacterial strains, DNA extraction, and PCR primers. : A total of 51 C. difficile strains, isolated in different Italian hospitals from 1986 to 1999, were examined. Six strains were representative of six different outbreaks, whereas 24 strains were isolated from sporadic cases and 21 were isolated from asymptomatic patients. C. difficile VPI 10463 and C. difficile ATCC 43597 were used as toxinogenic and nontoxinogenic control strains, respectively. C. difficile strains 51377 and 57267 were used as controls for toxinotypes VI and VII, respectively (a description of the toxinotyping method can be viewed on- line at ). The in vitro production of toxins B and A was assayed by cytotoxicity testing and an enzyme immunoassay method (Immunocard-Toxin A; Meridian Diagnostics, Cincinnati, Ohio), respectively. Five microliters of crude extracts of DNA was used for multiplex PCRs and binary toxin gene detection. One microliter of purified DNA, extracted using a Nucleobond AXG100 kit (Macherey-Nagel, Duren, Germany), was the template for toxinotyping and amplification of the entire tcdC gene. PaLoc PCR primers and their locations are shown in Fig. . Oligonucleotides were synthesized by M-Medical, Florence, Italy. FIG. 1. | Primers used in the PaLoc analysis. Primers used in the PaLoc analysis. (A) Specificity and nucleotide sequences of primers and molecular sizes of the PCR products obtained for each pair of primers. (B) Location of PCR primers on a schematic representation of the PaLoc region. The small arrowheads indicate the orientation of primers. PaLoc genes detection and sequencing of the tcdC gene. : Toxin A and B genes, tcdA and tcdB, were amplified by a multiplex PCR assay. The reaction mixture contained 1x buffer (10 mM Tris-HCl, 50 mM KCl, 1.5 mM MgCl2), 200 pmol of each deoxynucleoside triphosphate, 100 pmol of TA1 and TA2 primers, 25 pmol of TB1 and TB2 primers, and 2.5 U of Takara rTaq (Takara Shuzo Co., Ltd., Shiga, Japan). The template was denatured for 5 min at 94C, and DNA was amplified for 30 cycles consisting of 1 min at 94C, 1 min at 50C, and 1 min at 72C. Multiplex PCR for the PaLoc accessory genes detection was performed as described by Cohen et al. . The same cycling conditions were employed to amplify the entire tcdC gene by using primers C1 and C2 . The reaction mixture was prepared as described for toxin A and B detection. Sequencing was performed with a Perkin-Elmer ABI373A DNA sequencer. The deduced amino acid sequence was obtained by the ORF Finder program, whereas the nucleotide and amino acid sequences were compared with database entries by using the BLAST program. Toxinotyping. : Toxinotyping is a PCR-restriction fragment length polymorphism (RFLP) method consisting of the amplification of two toxin fragments, B1 from tcdB and A3 from tcdA, and of their digestion by specific restriction enzymes to obtain patterns characterizing the different variants of the toxin genes . We followed the method reported in the toxinotyping home page , with some minor modifications. The PCR mixture contained 10 mM Tris-HCl, 50 mM KCl, 1.5 mM MgCl2, 400 pmol of each deoxynucleoside triphosphate, 100 pmol of primers, and 2.5 U of Takara ExTaq (Takara Shuzo Co., Ltd.). After a denaturation of 5 min at 94C, the DNA was amplified by 30 cycles of 1 min at 94C, 1 min at 47C, and 5 min at 72C. At the end, samples were held at 72C for 10 min. PCR fragments were purified with a QIAquick PCR purification kit (Qiagen) and digested. Binary toxin gene detection. : Internal regions of binary toxin genes were detected as previously described . Two specific primers, BIN5 (5' AAT ATT GGG AGG GAG AAT AAA TG 3') and BIN6 5' (TGT ATT TTC ATT GTT TCT CCT CC 3'), were designed to amplify the entire ctdA gene, which codes for the enzymatic toxin component, and two other primers, BIN7 (5' ATT GTT GAT GCA ACA TTG ATA CC 3') and BIN8 (5' AAT ATA TAT TGT ATT GAG GGG AC 3'), were designed to amplify the entire cdtB gene, which codes for the binding toxin component. The reaction mixture and the cycling conditions were the same as described for toxin A and B detection. Different studies have demonstrated a great heterogeneity in C. difficile toxin A and B genes . This characteristic has been successfully used for C. difficile strain typing, in addition to the other methods already known . On the other hand, there are few data on C. difficile strains with variant PaLoc accessory genes and on their circulation among patients. We examined the PaLoc of several C. difficile strains isolated from clinical samples to acquire further information. A total of 51 C. difficile strains were examined, and 32 tested as toxinogenic. The analysis of the PaLoc accessory genes demonstrated that 8 (25%) of the 32 toxigenic strains showed a different pattern than that of the control strain, VPI 10463 (Table ; Fig. ). Six strains, Pd5, Pd7, Pd13, Pd16, Pd53, and Pd55, showed a pattern with only four bands, 622, 300, 262, and 162 bp in size, apparently corresponding to the genes cdd3, tcdD, tcdE, and cdu2, respectively. Two single PCRs, for tcdC and tcdD internal fragments, were performed separately on these strains. Unexpectedly, we obtained a product of approximately 300 bp in both PCR assays from all the strains analyzed (data not shown). A deletion in the tcdC fragment explained its comigration with the tcdD fragment. The same deletion was also observed in the tcdC gene of control C. difficile strains 51377 and 57267. Two strains, M7 and Pd3, showed a tcdC fragment smaller than expected . Six C. difficile strains were also recognized as variant strains for the toxin A and B genes . Four strains could be classified as toxinotype V, and two strains could be classified as toxinotype VI. It has already been suggested that toxinotypes VI and V are closely related . In this study, all the strains belonging to these two toxinotypes were isolated from cases which occurred in different departments of the same hospital from 1997 to 1999, suggesting persistent circulation of these variant strains. Few data are reported on the role of variant C. difficile strains in causing severe disease ; therefore, it is interesting that these C. difficile variant strains were responsible both for sporadic cases of antibiotic-associated diarrhea and for an outbreak. The infections caused by this particular group of C. difficile variant strains were not age related, and there was no correlation with a particular patient population. Both genes encoding the binary toxin were detected in seven of the eight variant strains (data not shown), including C. difficile M7, which was recognized as toxinotype 0 . On the contrary, Pd3, the other strain belonging to toxinotype 0 and showing a variant tcdC gene, did not have the binary toxin genes. Specific PCRs for cdtA and cdtB confirmed the absence of the binary toxin genes in all the other toxigenic C. difficile strains. The entire PaLocs of C. difficile M7 and Pd3 were analyzed by PCR-RFLPs to verify the absence of variations in other regions of the toxin A and B genes and in the rest of the genetic unit. C. difficile Pd3 did not show further variations, whereas three fragments of the C. difficile M7 PaLoc (A1, encoding the catalytic domain of toxin A; B2, encoding the translocation domain of toxin B; and PL2, located upstream the tcdB gene) showed different patterns than the reference strain C. difficile VPI 10463 after digestion with specific enzymes (data not shown). These results demonstrate that C. difficile M7 represents a new toxinotype and indicate, in contrast to the data already known , that it is also possible to detect the binary toxin genes in strains with minor modifications in tcdA and tcdB genes. In this study, we demonstrated genetic variability of the tcdC gene, which codes for the supposed negative regulator of toxin A and B gene transcription. Three different tcdC nucleotide sequences were identified and denominated types A, B, and C . The deletions characterizing the tcdC variant genes are located in a DNA region featuring repeated sequences that code for eight 3-amino acid repeats of an acidic or basic nature . These DNA regions show higher mutation frequencies, due to recombination events between repeats . tcdC type A shows a deletion of 39 bp, whereas types B and C show a deletion of 18 bp. A transition from cytosine to thymine in tcdC-A determines a nonsense mutation, so the tcdC protein has only 61 amino acids compared to the 232 expected. Type B and C tcdC genes, in spite of the different nucleotide sequences, code for an identical amino acid sequence. This protein of 226 amino acids is characterized by the deletion of 6 amino acids, determining the loss of the basic repeat KKA and the partial loss of the acidic repeats EEA and EEQ . It is noteworthy that all the strains belonging to toxinotypes V, VI, and VII that were examined in this study showed a TcdC of only 61 amino acids. A truncated protein, with a sequence of 22 amino acids, has been previously observed in C. difficile strain 8864 . It has been hypothesized that this variant TcdC probably lacks its function and that it contributes to the extreme cytotoxicity of strain 8864 . All the C. difficile strains with major variations in toxins A and B examined in this study showed high levels of cytotoxicity (data not shown). This result seems to confirm the possibility that the dramatic modifications observed in TcdC could also lead to an altered function of the protein in these strains, contributing to the high level of toxin expression. A second TcdC variant was identified both in C. difficile Pd3 and in C. difficile M7. The level of cytotoxicity in vitro of M7 was significantly lower than that observed for the other variant TcdC strains (data not shown). Further studies should be performed to determine the influence of the variant TcdC on M7 toxin gene transcription and to investigate the functionality of the mutated toxins of this strain. It is possible that different TcdC variants have a different functionality and diversely affect toxin production, a hypothesis with important implications for the pathogenic potential of C. difficile strains. Therefore, it could be very interesting to extend these studies to all the other C. difficile toxinotypes and to investigate the influence of variant TcdC proteins on the virulence of this pathogenic microorganism. FIG. 2. | Detection of PaLoc accessory genes cdd-3, tcdC, tcdD, tcdE, and cdu-2 by multiplex PCR in eight C. difficile Detection of PaLoc accessory genes cdd-3, tcdC, tcdD, tcdE, and cdu-2 by multiplex PCR in eight C. difficile strains with variant PaLoc genes identified in this study. Lane 1, Pd 5; lane 2, Pd7; lane 3, Pd13; lane 4, Pd16; lane 5, Pd53; lane 6, Pd 55; lane 7, M7; lane 8, Pd3; lane 9, DNA molecular weight marker IX; lane 10, C. difficile VPI10463; lane 11, ATCC 43597; lane 12, C. difficile 51377; lane 13, C. difficile 57267. FIG. 3. | Toxinotyping of eight C. difficile Toxinotyping of eight C. difficile strains with variant PaLoc genes and of the reference strains C. difficile VPI 10463, C. difficile 51377, and C.difficile 57267. (For descriptions of the other toxinotypes, see Rupnik et al. [, ]). The PCR-RFLP patterns of A3 and B1 fragments are shown for each strain. A3 fragments were digested with EcoRI (E), and B1 fragments were digested with HincII (H) and AccI (A). M, 100-bp DNA ladder (BioLabs). Pd7, Pd13, Pd16, and Pd55 (C. difficile strains), toxinotype V; Pd5 and Pd53, toxinotype VI; Pd3 and M7, toxinotype 0 (C. difficile M7 represents a new toxinotype, as demonstrated by PCR-RFLP analysis of the entire PaLoc). C. difficile VPI 10463, C. difficile 51377, and C.difficile 57267 are the reference strains for toxinotypes 0, VI, and VII, respectively. FIG. 4. | Comparison of the TcdC nucleotide and amino acid sequences of the C. difficile Comparison of the TcdC nucleotide and amino acid sequences of the C. difficile reference strain VPI 10463 with those of the variant TcdC proteins identified in C. difficile clinical isolates examined in this study. Dots and dashes indicate identical bases and gaps, respectively, for the different tcdC alleles (tcdC-A, -B, and -C). The termination codon in tcdC-A is underlined. Only the amino acid changes are indicated for each TcdC variant. The eight 3-amino-acid repeats of the VPI 10463 TcdC are indicated by open (acidic in nature) and grey (basic in nature) boxes. TABLE 1 | Molecular characteristics of C. difficile strains with variant PaLoc genes analyzed in this study Nucleotide sequence accession numbers. : The nucleotide sequences of the tcdC genes of C. difficile strains Pd5, Pd3, and M7 were assigned EMBL numbers AJ428941, AJ428942, and AJ428943, respectively. Backmatter: PMID- 12202598 TI - Are Three Sputum Acid-Fast Bacillus Smears Necessary for Discontinuing Tuberculosis Isolation? AB - To evaluate the efficacy of three sputum acid-fast bacillus (AFB) smears to rule out pulmonary tuberculosis, sputum AFB smear and culture results were analyzed at two university-affiliated teaching hospitals. The negative predictive value of the smear increased by only 0.2% on days 2 and 3 each, indicating that in low-prevalence populations, there is limited value in requiring three negative sputum AFB smears before discontinuing tuberculosis isolation. Keywords: Introduction : Current guidelines recommend that patients suspected of having active pulmonary tuberculosis (TB) in a health care facility should be placed in a TB isolation room . These recommendations state that isolation can be discontinued when the diagnosis of TB is ruled out or when a determination has been made that the patient is noninfectious. Once a patient has been diagnosed with TB, "isolation should be discontinued only when the patient is on effective therapy, is improving clinically, and has had 3 consecutive negative sputum acid fast bacillus (AFB) smear examinations collected on different days" . In clinical practice, patients suspected of having pulmonary TB are placed routinely in TB isolation. Although most United States hospitals require three sputum AFB smears to discontinue isolation, the published literature reveals limited data to support this 3-day requirement (, -, ). Therefore, we examined the value of this practice at two New Jersey university-affiliated teaching hospitals. (This work was presented in part at the 39th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, Calif., 26 to 29 September 1999 [P. Mathew, B. Vazirani, Y. H. Kuo, R. H. K. Eng, S. M. Smith, and M. P. Weinstein, Abstr. 39th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 850, 1999].) Mycobacteriology laboratory records for all sputum AFB smear and culture tests performed at Robert Wood Johnson University Hospital, New Brunswick, N.J., and the Veterans Affairs Medical Center, East Orange, N.J., were reviewed from September 1993 through September 1998. Specimens from patients who had three separate sputum collections (induced or expectorated) submitted for AFB smear and culture on different days within a 14-day period were included in the analysis. Specimens were not included in the analysis if they were repeat sputum series from a patient already included in the study. Patients who contributed fewer than three sputum AFB specimens within a 14-day period were also excluded. Sputum smears were decontaminated, digested, and concentrated by standard laboratory methods. Smears were screened by using auramine O stain followed by Kinyoun stain confirmation. Sputum sediments were inoculated to BACTEC 12B broth (Becton Dickinson Microbiology Systems, Sparks, Md.) and nonselective and selective Middlebrook 7H11 solid media. Cultures were incubated at 37C in a 5% CO2 incubator for up to 6 weeks. Positive cultures were identified with the AccuProbe DNA hybridization assay (GenProbe, San Diego, Calif.). If more than one specimen was received on a given day, either they were pooled or only the first specimen was included in the data analysis. The sensitivity, specificity, and negative predictive values (NPVs) of the sputum smear examination were calculated by using the sputum culture results as the "gold standard." Confidence intervals (CI) were calculated by the exact method. For the difference between proportions, the estimations of confidence interval were based on the normal approximation. During the study period, 5,336 sputum specimens were submitted for AFB smear and culture from 1,981 patients. M. tuberculosis grew in 78 specimens from 25 patients at Robert Wood Johnson University Hospital and 125 specimens from 31 patients at Veterans Affairs Medical Center, a total of 203 positive cultures from 56 patients. The overall prevalence of culture-positive M. tuberculosis in sputum specimens was 2.83%. Sputum AFB smears and culture results for M. tuberculosis are shown in Table . Sputum AFB smears had a sensitivity of 67.5% (95% CI, 60.6 to 73.9) and specificity of 97.5% (95% CI, 97.0 to 97.9) when the data from the two hospitals were combined. TABLE 1 | Sputum AFB smear and culture results for M. tuberculosis Only 488 patients (24.6%) produced three separate sputum specimens on different days during a 14-day period. Of these, 27 patients had positive cultures. The results of AFB smear and culture categorized by the sequence in which specimens were submitted are shown in Table . Eight of 27 (29.6%) patients were AFB smear negative but culture positive for M. tuberculosis. Of 19 smear-positive patients, 17 (89.5%) were identified as positive on the first smear, 1 (5.3%) was identified on the second smear, and 1 (5.3%) was identified on the third smear. TABLE 2 | Results of AFB smear and culture categorized by the sequence in which specimens were submitted The NPV of AFB smears according to the day of the first positive smear is shown in Table . As is evident, the incremental value of second and third specimens after a first negative smear was negligible. TABLE 3 | NPV of AFB smears according to the day of collection In this study the overwhelming majority of patients whose AFB smears were negative and for whom TB isolation was discontinued were identified with the first sputum AFB smear. The incremental benefits of 0.2% with the second specimen and of 0.2% with the third specimen have to be weighed against the costs of additional patient days in TB isolation. Current recommendations were formulated at a time when there was resurgence in TB . However, there is a paucity of published literature directly supporting the current clinical practice of continued TB isolation until three sputum smears are negative. The 1985 U.S. Public Service Guide for level III laboratories suggested that the number of specimens submitted for culture be determined by the results of early smear examination, and Bates stated that three specimens were sufficient when at least two of the first three smears were positive. Data to support the numbers of smears recommended to diagnose pulmonary TB were derived from field studies in high-prevalence areas in South East Asia or from preselected symptomatic patients with prolonged cough, purulent sputum, and hemoptysis . In these populations, two consecutive smears were sufficient for detecting TB. In a study conducted at the Tuberculosis Chemotherapy Centre, Madras, India, in the 1950s, two immediate (spot) collections were compared with two specimens collected overnight in 348 sputum-positive patients, the great majority of whom had extensive disease . For AFB smears, the percentages reported positive were 66.2 and 76.4% for the spot and collection methods, respectively, and 89.7 and 93.7% for culture. Limitations in the data presented make calculation of sensitivity and specificities impossible. In retrospective studies by MacGregor and Greenbaum et al. , it was concluded that no more than three sputum specimens were necessary for the diagnosis of TB if, in fact, the diagnosis was to be made by sputum examination. The same criteria were recommended for discontinuing isolation. However, neither study specifically provided data with regard to the number of smears. These data have helped form the basis of the current three-specimen requirement for ruling out pulmonary TB. The probable rationale behind this recommendation is that some patients shed mycobacteria irregularly and in small numbers and, thus, increasing the number of specimens would increase the yield . With regard to the recommendation for early morning specimens, the assumption is that M. tuberculosis would be present in maximum concentration in sputum after pooling overnight in the respiratory tract. Two recently published reports have provided results similar to the observations in the present study. Nelson et al. , in a 10-year retrospective study in Minneapolis, Minn., found that the majority of culture-proven pulmonary TB cases were diagnosed from the first or second sputum specimens and that only rarely was the third specimen of diagnostic value. Craft et al. reviewed 4 years' worth of data at the University of North Carolina and concluded that "modifying the smear policy from three to two negative smears would have resulted in no increased risk of spreading TB and would decrease the number of days patients are unnecessarily placed under airborne precautions" in their institution. Our study and those cited above raise the question of whether three sputum AFB smears are necessary before discontinuing TB isolation. In the two hospitals that we studied, it took up to 14 days to obtain a third sputum specimen for 488 of 1,981 patients, further pointing to the impracticality of the recommendations in routine clinical practice. Whether multiple negative sputum AFB smears obtained during a shorter period such as 24 h would obviate the collections on subsequent days has not been studied. Since the incidence of TB is once again declining in the United States , a reassessment of the requirement for three negative sputum smears to discontinue TB isolation seems warranted. However, once the initial AFB smear, or arguably two smears, is negative, the additional specimens for culture can be obtained after isolation is discontinued. Exceptions to this recommendation may be necessary when the pretest probability of TB is high. Backmatter: PMID- 12202586 TI - Comparison of Lysis Filtration and an Automated Blood Culture System (BACTEC) for Detection, Quantification, and Identification of Odontogenic Bacteremia in Children AB - Lysis filtration (LyF) was compared with BACTEC PAEDS PLUS in estimating the prevalence of, and sensitivity for, detection of odontogenic bacteremia. Both real bacteremia and simulated bacteremia (seeded blood or saline samples) were assessed to determine the validity of LyF in estimating bacteremia. The simulated bacteremia was also used to assess the reliability of LyF to estimate intensity of bacteremia in CFU per milliliter of blood. Reference organisms were used to assess the abilities of LyF and BACTEC to isolate known oral streptococci. There was no difference in the number of CFU per milliliter of seeded saline, seeded blood, and drop cultures of the organisms plated directly onto agar. Blood cell volume had a negligible effect on the yield of organisms for simulated bacteremia. When LyF and BACTEC were compared, the time to detection of bacteremia was always significantly shorter for BACTEC. For aerobic cultures, these times were 43.7 and 9.6 h, respectively (P < 0.01). For anaerobic cultures, these times were 45.1 and 9.9 h, respectively (P < 0.01). These differences occurred as well for bacteremia following the extraction of a single tooth, with LyF and BACTEC aerobic cultures taking 78 and 30.5 h, respectively (P < 0.0001). For anaerobic cultures, the times were 90.8 and 45 h, respectively (P < 0.0004). A preextraction bacteremia was detected on 2.1% of occasions with BACTEC compared to 31% of occasions with LyF (P < 0.05) The use of LyF was an effective and reliable means of estimating the intensity of pre- and postextraction bacteremia. The values were 3.6 and 5.9 CFU/ml, respectively (P < 0.4729), and the difference was not statistically significant. In summary, BACTEC is quicker than LyF, but less sensitive. LyF provides additional important information in estimating the intensity of bacteremia. Keywords: Introduction : The role of odontogenic bacteremia in the etiology of bacterial endocarditis (BE) is well established, with oral streptococci comprising between 51 and 63% of the organisms detected . The most commonly used method of detecting bacteria in blood is by an automated broth culture, which has the advantage of rapid growth of any cultivable organisms . The main disadvantage of the broth culture technique is that it does not enable estimation of the intensity of bacteremia . This is considered important, because large inocula are required to induce BE in experimental animals. The 90% infective dose (ID90) for experimental animals is 1 x 106 to 2 x 108 CFU/ml . The intensity of bacteremia detected in humans under clinical conditions is only 10 to 100 CFU/ml . Even when there is established intravascular infection, such as BE, the intensity of bacteremia is approximately 200 CFU/ml, with values ranging from 16 to 310 . Thus an ID90 in animals is of the order of 10,000 to 1,000,000 greater in intensity than the bacteremia commonly occurring in humans. The guidelines for antibiotic prophylaxis of BE take account of the risk to susceptible individuals from dental procedures thought to be hazardous because of the high percentage of prevalence of associated bacteremia . This highlights an emerging problem, because recent work has demonstrated that procedures excluded from antibiotic prophylaxis, such as placement of a rubber dam, a procedure used in restorative dentistry , cause a bacteremia similar to that following dental extractions . It is clear that this recent quantitative evidence conflicts with the current guidelines regarding dental procedures that are recommended for antibiotic prophylaxis in patients with cardiac lesions predisposing them to the risk of developing BE. Other factors, such as perioperative immune modulation, may be a contributory factor . In view of this conflicting information, it has been suggested that the policy on antibiotic prophylaxis for dental procedures be revised . It is apparent that quantitative data on the intensity of odontogenic bacteremia may play an important part in assessing the need for antibiotic prophylaxis for specific dentogingival manipulative procedures . The availability of such data may lead to a reduction in the number of the specific dental procedures that require antibiotic prophylaxis . If this is true, antibiotic use would be reduced. This is an important consideration in view of current anxieties concerning excessive use of antibiotics and the development of antibiotic-resistant organisms . Conversely, should the number of procedures requiring prophylaxis be increased, there would be a rational basis to justify the increased use of antibiotics in children. This issue of antibiotic prophylaxis is important, because there is evidence that patients who receive antibiotics for prophylaxis harbor antibiotic-resistant streptococci in their oral flora . There are no clinical trials on prophylactic antibiotics in the prevention of BE, probably because in excess of 6,000 patients would be required for a robust test . In practical terms, this number is unachievable. Current recommendations for dental procedures requiring prophylaxis are based on indirect data from dental bacteremia studies . The use of blood culture methods for estimating the intensity of bacteremia, as well as the percentage prevalence of positive cultures, would improve the data available to national bodies when considering dental procedures that should be carried out under the protection of systemic antibiotic prophylaxis. Preliminary work has shown that the technique of lysis filtration (LyF) appears to be an effective method of estimating the intensity of bacteremia following dental extractions in adults . There are no data available on the use of LyF with children. The purpose of the present work was to investigate the reliability of LyF in the detection of dental bacteremia by comparing it with the reliability of an automated blood culture system, the BACTEC 9240. MATERIALS AND METHODS : Ethical approval was obtained from The Eastman Dental Hospital Research and Ethics Committee and The Great Ormond Street Hospital for Children NHS Trust. Effect of blood composition and processing procedures on prevalence and intensity of simulated bacteremia. | Blood was obtained by using an aseptic technique from children receiving comprehensive dental treatment under general anesthesia at The Eastman Dental Hospital. The blood was sampled immediately after anesthetic induction and before any dental treatment was carried out. The first 0.5 ml of blood was discarded to void any skin contaminants. This is not normally recommended . However, it has been shown that 10% of these "discards" contain staphylococci, probably picked up as a needle traverses a hair follicle (G. J. Roberts, unpublished M.Phil. thesis). A 14-ml sample of blood was withdrawn and divided into two aliquots of 1 ml each as well as two further aliquots of 6 ml each. One milliliter of blood was used for a standard laboratory leukocyte count and hemoglobin and hematocrit estimates. The first 6-ml aliquot of blood was added to 1.23 ml of sodium polyanetholesulfonate to prevent clotting. This 6 ml of blood was seeded with 100 mul (giving approximately 100 CFU) of Streptococcus sanguis type strain NCTC 7863 prepared from an overnight broth culture. This was processed by LyF . Briefly, each blood sample was added to lysing solution and incubated at 37C for 10 min. The lysed blood was drawn through a 0.45-mum-pore filter by negative pressure of approximately 60 mm Hg. The filter was removed from the unit and cut in half with sterile scissors. Each half-filter was placed onto brain heart infusion (BHI) agar (Difco, United Kingdom). One plate was incubated aerobically and the other was incubated anaerobically for 10 days. Colonies growing on the filter were subjected to Gram staining and identified to the species level by standard laboratory methods . The second 6-ml aliquot of blood was processed by LyF without any added bacteria. Control samples of bacteria suspended in saline were prepared as follows. Ten 12-ml samples of sterile aqueous 0.9% (wt/vol) sodium chloride were prepared, and each was seeded with 200 mul of S. sanguis. The seeded saline was divided into two aliquots of 6 ml each giving 100 mul per 6-ml sample. The first aliquot was added to streptokinase-streptodornase lysing solution and processed by LyF. The other 6-ml aliquot was processed through the micropore filter without the streptokinase-streptodornase lysing constituents. In addition, two single 50-mul volumes of seeded saline, which between them contained approximately 100 CFU/ml, were inoculated onto BHI agar as "drop plates" for aerobic and anaerobic incubation, respectively. This provided a method of direct estimation of the number of CFU per 100 mul of the seeding solution used for LyF. Simulated quantitative blood cultures for BACTEC and LyF with blood seeded with reference microorganisms. | The subject volunteers comprised 50 healthy adults at the Eastman Dental Institute and Hospital. By an aseptic technique, 18 ml of blood was withdrawn from a vein in the antecubital fossa. This was collected in a Vacutainer tube containing heparin to prevent clotting. Each blood sample was divided into two unequal aliquots of 6 and 12 ml. The order of processing was determined with random number tables. The 6 ml was equally divided between aerobic and anaerobic BACTEC PAEDS PLUS bottles (Becton Dickinson UK, Ltd., Oxford, United Kingdom). These were processed with the BACTEC 9240 at The Great Ormond Street Hospital for Children, London, United Kingdom. The 12-ml aliquot was seeded with 100 mul of one of several type strains of oral streptococci. A run of 10 blood samples was carried out for each of the five Streptococcus type strains: S. mutans NCTC 10449, S. mitis NCTC 551, S. intermedius NCDO 2227, S. oralis NCTC 11427, and S. sanguis NCTC 7863. Each species was prepared as for the simulated blood cultures used for the hemoglobin study described above. This seeded aliquot of 12 ml was divided into two 6-ml aliquots. The first was divided between aerobic and anaerobic BACTEC PAEDS PLUS bottles and processed in the same way as the unseeded aliquots. The second 6-ml aliquot of the seeded aliquot was processed by LyF as described above. Bacteria isolated from both BACTEC and LyF were reidentified to check that they were the seeded species used and to identify contaminants. The BACTEC screens for positive cultures automatically every few hours. The filters for LyF were screened twice daily at 0900 and 1400 h with a stereo microscope. In addition, the effects of the processing on the yield of organisms were examined by comparing the simulated blood culture with saline seeded with identical reference organisms. These results were expressed as CFU per milliliter of seeded blood or seeded saline. These were compared with the yield of reference organisms from a "drop" culture of the broth onto blood agar. This represents 100 mul of the broth of S. sanguis type strain NCTC 7863, which gave approximately 100 CFU per plate. The results were expressed as (i) the time to detection of the bacteremia for both the BACTEC and LyF methods, (ii) the percentage yield from the simulated blood cultures of the five reference organisms for both the BACTEC and the LyF, and (iii) the intensity of bacteremia for the seeded cultures and the seeded saline expressed in CFU per milliliter. Detection of bacteremia following extraction of a single tooth. | The subjects were children at The Great Ormond Street Hospital For Children receiving comprehensive dental treatment under general anesthesia. Immediately upon administration of anesthesia and before dental treatment was started, a 23-gauge intravenous cannula was inserted into a vein of the antecubital fossa or the dorsum of the hand. A 12-ml sample of blood was withdrawn before any dental treatment was started (preprocedure sample). A single dental extraction was performed before any other dental treatment. The second 12-ml blood sample was taken 30 s after the maximum movement involved in a single tooth extraction . This was the postextraction sample. Each 12-ml sample was divided into two 6-ml aliquots. The first was divided into two 3-ml volumes that were inoculated into the BACTEC PAEDS PLUS aerobic and anaerobic bottles, respectively. These were then processed in the BACTEC 9240. The second aliquot was processed by LyF (described above). The results were expressed in the same way as those for the simulated bacteremia. Contamination was assessed by processing a pair of BACTEC bottles with every 10th run of samples. The estimation of contamination for the processing of LyF samples was achieved by processing one filter, which was removed from the sterile envelope and put straight onto an agar plate. This was carried out for each set of samples processed. Statistical methods. | Numerical data were tested for normality by using the Shapiro-Wilk test (STATA reference manual, release 3. 5 ed., Computing Resource Center, Santa Monica, Calif.) and found to be not normally distributed. Thus the Mann-Whitney U test was used to test differences between the groups studied. Percent prevalence data were subjected to the chi-square test. Associations and correlations were examined by using the Spearman correlation coefficient. RESULTS : Simulated blood cultures and the effect of human blood composition and processing procedures on the prevalence and intensity of the simulated bacteremia. | The number of CFU per milliliter estimated from the runs of the blood- and saline-simulated bacteremia are shown in Table . This demonstrated that for both aerobic and aerobic culture methods, there was no difference between total counts from the saline-simulated bacteremia and the drop culture. Similarly for the blood-simulated bacteremia, there was no difference between the total count obtained from blood compared with the count from the drop culture. Simulated blood cultures for BACTEC and LyF with blood seeded with reference microorganisms. | As a result of pilot work, there was the suggestion that there was a slightly lower yield of oral streptococci in the LyF. This made it necessary to explore the possibility that LyF was less reliable at detecting oral streptococci than BACTEC. A total of 50 subject volunteers provided blood sampled from the antecubital fossa by an aseptic technique. For each of 10 blood samples selected at random, a type strain was inoculated into the blood and gently mixed. The samples were processed as for the simulated bacteremia (described above). The time to detection was always statistically significantly less for the BACTEC than for LyF . Oral streptococci were recovered from all of the blood samples inoculated into aerobic and anaerobic BACTEC PAEDS plus seeded with type strains S. mutans (n = 10), S. intermedius (n = 10), S. oralis (n = 10), and S. sanguis (n = 10). S. mitis (n = 10) was recovered from eight of the aerobic bottles and nine of the anaerobic bottles . For LyF, the seeded strains were also recovered from all of the samples, except for blood seeded with S. mitis. There was no aerobic or anaerobic growth from four samples . The difference between BACTEC and LyF was not significant (Fisher's exact probability test P = 0.155). The type strains were reidentified with the API Strep 20 system. There was one positive aerobic culture only in the unseeded group, which was an Actinobacillus species, presumably due to contamination. Overall, the prevalence of contamination was less than 1%. Detection of bacteremia following extraction of a single tooth. | The patients were 49 children 2 years 11 months to 14 years 11 months of age, with a mean age of 8 years 3 months. The percent prevalence values of bacteremia both preextraction and 30 s after extraction of a single tooth were greater with LyF . This was statistically significantly different for the preextraction samples (P < 0.05), but not for the postextraction samples. The intensity of bacteremia preextraction was 3.6 (standard deviation [SD], 6.2), and the intensity postextraction was 5.9 (SD, 8.2), a difference that was not statistically significantly different (P = 0.4729) . The mean time for detection of positive cultures was statistically significantly shorter (P < 0.0004) for BACTEC, with both aerobic and anaerobic methods requiring less than half the time taken for LyF . A wide variety of bacteria were isolated with both BACTEC and LyF. These included staphylococci, streptococci, micrococci, and neisseria and corynebacteria . There were slightly more oral streptococci isolated with BACTEC than with LyF: 31.2 and 17.6%, respectively. This was not statistically significantly different (chi square = 1.661, degree of freedom = 1, P = 0.197). TABLE 1 | Number of CFU per 3-ml sample for simulated bacteremia (by seeding blood), simulated saline bacteremia (by seeding saline), and drop cultures TABLE 2 | Time to detection of simulated bacteremia using seeded type strains for BACTEC versus LyF TABLE 3 | Number of positive cultures from simulated blood cultures (oral streptococci) detected by BACTEC and LyF TABLE 4 | Prevalence of positive cultures for BACTEC versus LyF (n = 49) TABLE 5 | Intensity of bacteremia by LyF TABLE 6 | Time in hours to detection of positive blood cultures following a single dental extraction for BACTEC versus LyF in 49 children TABLE 7 | Bacterial species isolated from positive blood cultures following extraction of a single tooth DISCUSSION : The effect of human blood of the lysing solution on the yield of organisms for LyF under both aerobic and anaerobic conditions for the simulated bacteremia was negligible. Neither the natural blood antibacterial factors nor the enzymes used in the lysing process had any effect on the viable count. The differences between the seeded blood and seeded saline confirmed this, with values that are remarkably close. This is reassuring, because there is the possibility that the enzymes used for the lysing process could affect the viability of oral streptococci. In addition, the total sample counts in CFU per milliliter were identical to the counts on "drop" cultures of the same specially prepared broth of reference organisms for both the blood samples and the saline samples. This is clearly an advantage of LyF, because it enables investigators to reliably estimate the intensity of bacteremia. The similarity of CFU counts from the simulated bacteremia, saline bacteremia, and drop cultures gives considerable support to this claim. This is an advantage over the BACTEC, because broth cultures do not enable estimation of the intensity of bacteremia. An advantage of the broth culture is the very rapid response: a simulated bacteremia was detected within 10 h, compared to over 40 h with LyF. This is important clinically, because it enables BACTEC to detect a potentially life-threatening septicemia very rapidly. The importance of this finding is that lysis filtration is not a substitute for BACTEC broth cultures for the detection of clinically important bacteremia. The abilities of both BACTEC and LyF to detect different reference organisms in simulated bacteremia were extremely good. For four out of five species, this detection was 100%. Both systems were less good at detecting S. mitis. This is perplexing, and it is not possible to offer an explanation for this other than the possibility that S. mitis is, in general, less easy to culture reliably. The prevalence of bacteremia detected in child patients was greater with LyF both pre- and postextraction of a tooth compared with the broth culture There is no obvious explanation for the greater recovery of bacteria from the preoperative LyF compared with the preoperative BACTEC. It is possible that with low levels of bacteria, the BHI medium is more effective than broth at encouraging growth. This is in agreement with earlier work and could also be of clinical significance for the prescription of antibacterial drugs, particularly if the broth culture is negative. A preoperative sample is recommended to give a realistic estimate of postoperative bacteremia. This is because some preextraction bacteremia may be as high as 20 CFU/ml. It would be necessary to have a sample size of 157 subjects to achieve statistical significance (STATA reference manual, release 3.5 ed.). The reliability and consistency of BACTEC for the detection of odontogenic bacteremia have been confirmed, because the percentage of positive cultures detected in this study (35.7%) was similar to that in a previous study conducted by the same investigators (43.2%) by a radiometric method for identification of positive cultures . The technique of LyF used in this study on children was not as sensitive as that in a study of adults, in which 86% of blood cultures were positive following a single dental extraction . This may be because the subjects in the present study were children, and the percentage of prevalence-positive cultures following dental treatment in children is almost always less than that in adults . What is clear is that the use of LyF is more sensitive than BACTEC for the detection of the percentage of prevalence of bacteremia following the extraction of a single tooth. The ability of LyF to estimate the intensity of bacteremia was confirmed with values ranging from 1 to 28 CFU/ml following the extraction of a single tooth. This compares well with the data for intensity in adults, for which the values were up to 43 CFU/ml , a mean of 3.8 CFU/ml , and a mean of 21 CFU/ml , respectively. As with the simulated bacteremia, the time to detection of the clinical bacteremia was very much greater with the LyF . This confirms the superiority of BACTEC in detecting odontogenic bacteremia in patients in terms of the rapidity of obtaining a result. The range of organisms isolated following the single dental extraction appears greater with BACTEC than with LyF and probably reflects the ability of BACTEC to detect very low levels of some bacteremia. This is an issue that should be investigated further. In perusing the data, BACTEC appears to be better at detecting staphylococci and some streptococci. In conclusion, LyF is significantly more sensitive than BACTEC at detecting odontogenic bacteremia and has the ability to detect almost the same range of organisms. It has the disadvantage that it is very much slower than BACTEC and, at present, is limited to research. The ability of LyF to estimate the intensity of bacteremia may be important in monitoring the progress of BE once a diagnosis has been made. Backmatter: PMID- 12202562 TI - Development of a Real-Time Reverse Transcriptase PCR Assay for Type A Influenza Virus and the Avian H5 and H7 Hemagglutinin Subtypes AB - A real-time reverse transcriptase PCR (RRT-PCR) assay based on the avian influenza virus matrix gene was developed for the rapid detection of type A influenza virus. Additionally, H5 and H7 hemagglutinin subtype-specific probe sets were developed based on North American avian influenza virus sequences. The RRT-PCR assay utilizes a one-step RT-PCR protocol and fluorogenic hydrolysis type probes. The matrix gene RRT-PCR assay has a detection limit of 10 fg or approximately 1,000 copies of target RNA and can detect 0.1 50% egg infective dose of virus. The H5- and H7-specific probe sets each have a detection limit of 100 fg of target RNA or approximately 103 to 104 gene copies. The sensitivity and specificity of the real-time PCR assay were directly compared with those of the current standard for detection of influenza virus: virus isolation (VI) in embryonated chicken eggs and hemagglutinin subtyping by hemagglutination inhibition (HI) assay. The comparison was performed with 1,550 tracheal and cloacal swabs from various avian species and environmental swabs obtained from live-bird markets in New York and New Jersey. Influenza virus-specific RRT-PCR results correlated with VI results for 89% of the samples. The remaining samples were positive with only one detection method. Overall the sensitivity and specificity of the H7- and H5-specific RRT-PCR were similar to those of VI and HI. Keywords: Introduction : Highly pathogenic strains of avian influenza virus (AIV), which are type A influenza viruses, cause severe disease in domestic poultry, including chickens and turkeys. Because of the high morbidity and mortality associated with highly pathogenic AIV, infection of commercial poultry can lead to substantial economic losses as demonstrated by the 1983 to 1984 outbreak in Pennsylvania . Highly pathogenic AIV is considered to be exotic to the United States, although low pathogenic AIV is routinely isolated from turkeys reared in migratory bird flyways in the midwest and from various avian species in live-bird markets in New York and New Jersey . The AIVs currently circulating in the live-bird markets of New York and New Jersey have been classified as low pathogenic. However, previous epornitics of highly pathogenic AIV have occurred in regions where low pathogenic AIV had increased in pathogenicity after circulating in the poultry population for a period of time . Historically, highly pathogenic AIVs in poultry have only belonged to the H5 and H7 hemagglutinin (HA) subtypes. Therefore, because there is a greater risk for these subtypes to become highly pathogenic, it is important to identify them specifically in surveillance programs. Currently, virus isolation (VI) in embryonating chicken eggs and subsequent HA and neuraminidase subtyping by serological methods constitute the standard for AIV detection and subtype identification. Although VI in embryonating eggs is a sensitive method, it may take 1 to 2 weeks to obtain results, by which time the results may no longer be relevant. Conversely real-time reverse transcriptase PCR (RRT-PCR) can be a rapid assay; results, including subtyping, may be available in less than 1 day. It can also be less expensive on a cost-per-sample basis than VI in embryonating eggs. Standard RT-PCR has been previously applied to the detection of avian influenza virus and each of the 15 HA subtypes . Additionally, an RRT-PCR assay for influenza virus has been developed; however, it is a two-step RT-PCR, multiplex assay based on human influenza virus sequences for the detection of influenza virus types A and B . One-step RRT-PCR with hydrolysis probes, as described by Holland et al. and Livak et al. , has been successfully applied to the detection of various RNA viruses . RRT-PCR offers the advantages of speed and no post-PCR sample handling, thus reducing the chance for cross-contamination versus standard RT-PCR. In this study we describe the development of a one-step RRT-PCR assay with hydrolysis type probes for the rapid screening of clinical samples and live-bird markets for type A influenza virus and for the subsequent identification of the H5 and H7 avian influenza virus subtypes. MATERIALS AND METHODS : RNA extraction. | RNA was extracted with the RNeasy kit (Qiagen, Valencia, Calif.) with a modified protocol for fluid samples recommended by the manufacturer. Briefly, 500 mul of swab material from clinical samples was clarified by centrifugation at 12,000 x g for 2 min, or, for previously isolated viruses, 500 mul of chorioallantoic fluid (CAF) was mixed with 500 mul of 70% ethanol and 500 mul of kit-supplied RLT buffer (Qiagen) and the entire sample was applied to the RNeasy spin column. Subsequently the kit protocol for RNA isolation from the cytoplasm of cells was followed. RNA was eluted in 50 mul of nuclease-free water, and 8 mul per RRT-PCR was used for the template. Hydrolysis probe and primer sets. | An influenza virus matrix gene-specific PCR primer set and hydrolysis probe were designed for a region conserved in all type A influenza virus matrix genes . In addition, H5- and H7-specific primer sets for conserved regions of the H5 and H7 HA gene sequences were developed. However, because of the large sequence variation of the H5 and H7 genes, the probes and primers were primarily targeted to North American H5 and H7 influenza viruses. All probes were labeled at the 5' end with the 6-carboxyfluorescein (FAM) reporter dye and at the 3' end with the 6-carboxytetramethylrhodamine (TAMRA) quencher dye. RRT-PCR. | The Qiagen one-step RT-PCR kit was used with a 20-mul reaction mixture under the following conditions: 0.8 mul of kit-supplied enzyme mixture (including RT and hot-start Taq polymerase), 10 pmol of each primer, 0.3 muM probe, 400 muM (each) deoxynucleoside triphosphate, 3.75 mM MgCl2, and 6.5 U of RNase inhibitor (Promega, Madison, Wis.). The RT step conditions for all primer sets were 30 min at 50C and 15 min at 94C. A two-step PCR cycling protocol was used for the matrix gene primer set as follows: 45 cycles of 94C for 0 s and 60C for 20 s. The H7 PCR cycling conditions were the same as those for the matrix gene except that a 58C annealing temperature was used for 40 cycles. A three-step cycling protocol was used for the H5-specific PCR as follows: 94C for 0 s, 57C for 20 s, and 72C for 5 s for 40 cycles. All temperature transition rates were set at the maximum transition rate of 20. Fluorescence data were acquired at the end of each annealing step. RRT-PCR was performed with the ruggedized advanced pathogen identification device (RAPID) thermocycler (a rapid air thermocycler) and software (Idaho Technologies, Salt Lake City, Utah). Positive and negative results of RRT-PCRs were determined by the RAPID autoanalysis software and rechecked manually. Specificity of primer and probe sets. | RRT-PCR with the matrix gene primer and probe set was performed with template RNA from influenza virus isolates representing all HA subtypes (see Table ) including both avian North American and Eurasian lineage viruses and isolates of human, equine, and swine origin to demonstrate specificity for type A influenza virus. RRT-PCR with the H5 and H7 subtype-specific probe sets was also performed with template RNA from all HA subtypes listed in Table to demonstrate specificity for their respective subtypes. In vitro transcription. | In vitro-transcribed matrix, H5, or H7 gene RNA was used for positive controls and for the determination of the detection limits of the assay. The influenza H5 and H7 HA genes and the matrix gene which had been cloned into the pAMP1 vector (Life Technologies, Rockville, Md.) as previously described were transcribed with the RiboMax (Promega) kit from the T7 promoter in accordance with the kit instructions and quantitated by spectrophotometer. Sequencing. | The entire matrix gene from six samples which were negative by the RRT-PCR assay and positive by VI and another six samples which were positive by RRT-PCR and VI during the original testing were amplified by standard RT-PCR with the Qiagen one-step kit, and primers were directed to the 12 or 13 conserved bases at the ends of each influenza virus RNA segment. The amplicons were subsequently excised from the gel and extracted with the Qiagen gel extraction kit. If no amplicon was visible, bands at the appropriate size on the gel were excised and extracted from the gel. Products of the gel extraction were cloned into the pAMP1 vector (Life Technologies) as previously described . Sequencing was performed with the ABI BigDye terminator system (Applied Biosystems, Foster City, Calif.). Embryo titration. | Virus was diluted in brain heart infusion broth with 10,000 IU of penicillin G, 2,000 mug of streptomycin, 1,000 mug of gentamicin, 650 mug of kanamycin, and 20 mug of amphotericin B/ml. Fertile chicken eggs were inoculated with 100 mul of virus at 10 days of incubation by the chorioallantoic sac route. Five eggs were inoculated per dilution. At 7 days postinoculation CAF was collected and tested for hemagglutination with 0.5% chicken red blood cells (CRBCs) in phosphate-buffered saline. The virus titer was determined by the Reed-Meunch method . RNA for RRT-PCR was extracted from the virus dilutions at the time of egg inoculation with the RNeasy kit (Qiagen) as previously described. Comparison with VI in embryonating chicken eggs. | Fifteen hundred fifty swab samples were obtained from the live-bird markets of New York and New Jersey. Each sample contained a pool of up to five cloacal, tracheal, or environmental swabs collected in 2 ml of brain heart infusion broth. Tracheal and cloacal swabs were obtained from up to five birds of each lot (different birds of the same species and from the same source which entered the market at the same time) present at the time the market was sampled, except waterfowl, from which only cloacal swabs were obtained. Five environmental swabs were taken in the following areas in each market: the office, bird area, slaughter area, and red meat area if present. RNA was extracted from each sample as previously described. Each sample was tested with the influenza virus matrix primer set. Samples positive with the matrix primer set were subsequently tested with the H7-specific primer set. Samples that were positive with the matrix gene primers and negative with the H7 primer set were then tested with the H5 primer set. Isolation of influenza virus from swab samples was performed in embryonated chicken eggs. Antibiotics and antimycotics were added to each sample to the following final concentrations: penicillin G, 10,000 IU/ml; streptomycin, 2,000 mug/ml; gentamicin, 1,000 mug/ml; kanamycin, 650 mug/ml; amphotericin B, 20 mug/ml. Four 9- to11-day-old chicken embryos were each inoculated with 300 mul of sample with antimicrobials by the chorioallantoic sac route. The eggs were incubated for 4 days and candled daily for viability; embryos that died within 24 h of inoculation were discarded as nonspecific. CAF from dead and surviving embryos was tested for hemagglutination of 0.5% CRBCs in phosphate-buffered saline. Samples from dead embryos that were negative for hemagglutination were passaged a second time. HA subtypes of all hemagglutination-positive samples were determined by hemagglutination inhibition (HI) assay . Hemagglutinating CAF was standardized to 4 HA units, and HA was mixed with an equal volume of influenza virus subtype reference serum at a titer between 1:32 and 1:64. Reference serum and CAF were incubated for 30 min at room temperature, and 0.5% CRBCs were added and mixed. The assay was evaluated for HI after incubation at room temperature for 30 min. Results of the RRT-PCR assay were compiled with the results of VI and HI by an independent third party to prevent bias. TABLE 1 | PCR primer and hydrolysis probe sequences TABLE 6 | Results of RRT-PCR with type A influenza virus-, H5-, and H7-specific primer and probe sets by type A influenza virus isolate and subtype RESULTS : Comparison of RRT-PCR with VI and HI. | The sensitivity and specificity of the RRT-PCR assay were compared to those of VI in embryonating eggs with 1,550 clinical swab samples from New York and New Jersey live-bird markets. Of the swab samples tested by both RRT-PCR and VI, 1,183 were negative by both assays and 202 were positive by both assays . Overall the results of the two assays agreed on 1,385 samples (89%) and disagreed on 165 samples (11%). Of the 165 samples on which results for the assays differed, 101 samples were positive by VI and negative by RRT-PCR; RRT-PCR detected 202, or 66.6%, of the 303 samples that were positive by VI. Sixty-four samples were positive by RRT-PCR but negative by VI; VI detected 75.6% of the samples which were RRT-PCR positive. RRT-PCR and VI were also compared for their relative sensitivities and specificities for the detection of positive live-bird markets. A market was considered positive if at least one sample was positive. Of the 109 live-bird markets tested in the study, 37 were negative by both assays and 61 were positive by both assays. Overall the assays agreed on 98 markets (90%) . Among the markets with differing results, four were positive by VI and negative by RRT-PCR and seven markets were positive by RRT-PCR but negative by VI. There was at least one RRT-PCR-positive sample in 61, or 93.8%, of the 65 markets that were positive by VI. RRT-PCR for H5 and H7 was compared to HI for sensitivity and specificity. Two hundred two samples were tested for H7 by both HI and RRT-PCR. There were four samples negative by both assays, and 194 samples were positive by both assays; results for 98% of the samples were in agreement . One sample was positive by RRT-PCR and negative by HI; three samples were positive by HI but negative by RRT-PCR. Because H5 RRT-PCR was performed only on samples that were influenza virus positive and H7 negative, only eight samples were tested by both methods. Both tests correlated on all samples tested for H5; one sample was positive, and seven samples were negative . The ability of the RRT-PCR assay to detect H5 and H7 was also determined for individual live-bird markets. Both HI and RRT-PCR were used to determine the subgroups of AIV present in 60 markets; 56 of the markets were positive by both methods for the presence of H7 subtype virus , and two markets were negative by both methods (overall 96.6% correlation). The remaining two markets were positive for H7 by HI and negative for H7 by RRT-PCR. Four markets were tested with both methods for H5. Both assays correlated for all four markets; one was positive, and three were negative. Assay sensitivity and specificity. | The sensitivity of the influenza virus matrix gene, H5, and H7 RRT-PCR assays relative to virus titer detectable by standard VI in embryonating eggs was determined. The detection limit of the matrix gene assay was determined to be approximately 10-1 50% egg infective dose (EID50) reproducibly by using the RNA extraction protocol described in Materials and Methods. The assay could detect RNA from 101 EID50s of H5 or H7 AIV. Detection limits for AIV- and the HA subtype-specific probe sets were determined by detection of in vitro-transcribed matrix gene, H5 HA gene, or H7 HA gene RNA. The minimum copy number of matrix gene RNA which could be detected was approximately 103 gene copies or 10 fg of in vitro-transcribed RNA. Both the H5 and H7 primer sets had a reproducible detection limit of 103 to 104 gene copies or approximately 100 fg of in vitro-transcribed RNA. The matrix gene primer and probe set was tested with RNA obtained from avian-origin influenza virus isolates representing all 15 HA subtypes and isolates of human, equine, and swine origin. The matrix primer set was able to detect all type A influenza viruses tested including viruses of human, equine, and swine origin . The H5 and H7 primer sets were tested with avian-origin isolates representing viruses of each HA subtype . The H5 and H7 primer and probe sets detected RNA only from virus isolates of their respective subtypes. Sequencing. | Sequencing the matrix genes of six individual samples that were VI positive from the live-bird market clinical samples but RRT-PCR negative revealed no sequence variation in the regions where the M+64 probe is located. One isolate had a single base change of an A to a G, which was located at the 5' base of the M+25 primer, and there was also a single base change of a T to a C at position 7 from the 5' end of the M-124 primer in all isolates sequenced including the six isolates which were positive by RRT-PCR during the original testing. TABLE 2 | Summary of RRT-PCR and VI results for individual samples from live-bird markets tested for type A influenza virus TABLE 3 | Summary of RRT-PCR and VI results for live-bird markets tested for type A influenza virus TABLE 4 | Summary of RRT-PCR and HI results for individual samples from live-bird markets tested for the H7 or H5 HA subtype TABLE 5 | Summary of RRT-PCR and HI results for live-bird markets tested for type A influenza virus and the H7 or H5 HA subtype DISCUSSION : Because mildly pathogenic AIV infection does not cause any pathognomonic clinical signs and is commonly subclinical, the presence of the virus must be determined by a diagnostic test. We have developed RRT-PCR as a rapid alternative to VI for AIV detection and subtyping. In addition to its speed, RRT-PCR reduces the handling of infectious material compared to VI. RRT-PCR also offers several advantages over standard RT-PCR such as speed and elimination of the possibility of cross-contamination of new samples with previously amplified products because the sample tube is never opened after PCR. Additionally, because the RRT-PCR product is detected with a sequence-specific probe, there is confirmation that the correct target was amplified, reducing the chances for false positives. The RRT-PCR assay for AIV was developed to be a rapid screening tool for commercial flocks and live-bird markets. Importantly, in this study, results could be obtained quickly; 28 clinical samples could be processed and tested in approximately 3 h. The assay also performed well compared to VI and HI since 94% of the VI-positive markets were also positive by RRT-PCR and 97% of the markets that were H7 positive by HI were H7 positive by RRT-PCR. However, RRT-PCR did not correlate as well with VI by individual sample tube, as 11% of the samples were positive by only one method. Differences in the detection of AIV between the assays can probably be explained at least in part by what the assays are detecting. VI can detect only live virus, and virus that has been inactivated during shipping or by disinfectants (which may be present in environmental samples) will not be detected, whereas they may potentially be detected by RRT-PCR. Additionally, all influenza viruses may not readily adapt to growth to detectable titers in embryonating chicken eggs within two passages. This may explain why some samples were RRT-PCR positive and VI negative. Interestingly, 26 of 64 (40.6%) RRT-PCR-positive, VI-negative samples were classified as non-H7 and non-H5 subtype isolates. Factors that may adversely affect the sensitivity of the RRT-PCR assay versus VI include RT-PCR-inhibitory substances in the samples, the use of a one-step RT-PCR method, inefficient RNA extraction procedures, and the potential for RNA to rapidly degrade before testing. Importantly, this RRT-PCR test was designed to balance both sensitivity and ease of use so that the procedure could be performed rapidly and on a large scale. Therefore, a single-step RT-PCR was used, although this method is reported to be less sensitive than a two-step RT-PCR procedure . A single-step RT-PCR procedure also greatly reduces the risk of cross contamination because the tubes are not reopened once the template is added. Additionally, VI tests use an equivalent of nine times more volume of the original sample than is used for RRT-PCR; therefore it is possible that alternative RNA extraction methods utilizing a greater sample volume would increase the sensitivity for samples with a low virus titer. Finally, RNA is relatively unstable and RNA-degrading enzymes are ubiquitous; therefore the RNA may have been degraded prior to testing, although efforts were made to reduce this possibility. Sequence variation in the matrix gene may also explain why the RRT-PCR test failed to detect viral RNA in some of the virus-positive samples. However matrix gene sequences from several samples that were positive by VI but negative by RRT-PCR did not have any sequence variation in the probe binding region, and the single base change present in the M-124 primer is present in many other isolates that were detected with these primers, including six RRT-PCR-positive samples from the same study. The 5' base change in one isolate in the M+25 primer binding region is also probably not adequate to prevent primer binding. In general, the region of the matrix gene to which the primers and probe are directed is conserved among type A influenza viruses; therefore this assay should detect most, if not all, type A influenza viruses, as demonstrated by the detection of nonavian isolates with this primer and probe set. Correlation between the assays was better with H7 and H5, although few samples were tested for H5 and H7, and only samples that were positive for influenza virus by both tests can be compared. For example, it is not known if a sample which was positive by VI and H7 HI would have been positive by the H7 RRT-PCR, although it was negative by the influenza virus matrix gene RRT-PCR test. Although the H5 test detected all of the North American avian-lineage H5 isolates tested, too few clinical samples were tested to definitively validate the H5 RRT-PCR test versus HI. A comparison of EID50 with RRT-PCR indicated that the matrix gene primer and probe set was able to detect 10-1 EID50 of virus. The increased sensitivity of the matrix gene RRT-PCR versus EID50 may be due to the ability of the RRT-PCR assay to detect fewer virions than the minimum necessary to cause hemagglutination. RRT-PCR is also expected to detect noninfectious particles, which have been estimated to constitute as much as 90% of some virus preparations . It is unclear why increased sensitivity relative to EID50 was not observed with the H5- and H7-specific assays. We have developed an RRT-PCR assay for the detection of type A influenza virus and the avian H5 and H7 HA subtypes. RRT-PCR is less expensive than VI, and, importantly, results are available much faster. This assay provides a rapid and feasible alternative to VI in embryonating chicken eggs and subtyping by HI as a flock or live-bird market screening tool. Backmatter: PMID- 12202594 TI - Molecular Typing of Papillomatous Digital Dermatitis-Associated Treponema Isolates Based on Analysis of 16S-23S Ribosomal DNA Intergenic Spacer Regions AB - Papillomatous digital dermatitis (PDD), an emerging infectious disease of cattle, is characterized by painful, ulcerative foot lesions. The detection of high numbers of invasive spirochetes in PDD lesions suggests an important role for these organisms in the pathogenesis of PDD. PDD-associated spirochetes have phenotypic characteristics consistent with members of the genus Treponema. Partial 16S ribosomal DNA (rDNA) sequence analysis of clonal isolates from California cattle showed that they comprise three phylotypes which cluster closely with human-associated Treponema spp. of the oral cavity (T. denticola and T. medium/T. vincentii) or genital area (T. phagedenis). The goal of our study was to apply 16S-23S rDNA intergenic spacer region (ISR) sequence analysis to the molecular typing of U.S. PDD-associated Treponema isolates. This methodology has potentially greater discriminatory power for differentiation of closely related bacteria than 16S rDNA analysis. We PCR amplified, cloned, and sequenced the ISRs from six California PDD-associated Treponema isolates and, for comparative purposes, one strain each of T. denticola, T. medium, T. vincentii, and T. phagedenis. Two ISRs that varied in length and composition were present in all the PDD-associated Treponema isolates and in T. denticola, T. medium, and T. phagedenis. ISR1 contained a tRNAAla gene, while ISR2 contained a tRNAIle gene. Only a single ISR (ISR1) was identified in T. vincentii. Comparative analyses of the ISR1 and ISR2 sequences indicated that the California PDD-associated Treponema isolates comprised three phylotypes, in agreement with the results of 16S rDNA analysis. PCR amplification of the 16S-tRNAIle region of ISR2 permitted rapid phylotyping of California and Iowa PDD-associated Treponema isolates based on product length polymorphisms. Keywords: Introduction : Papillomatous digital dermatitis (PDD), also referred to as digital dermatitis or hairy footwart, was first described in cattle in Italy in 1974 . In 1980, Rebhun et al. reported the presence of PDD in dairy cattle in New York. PDD is now found throughout most of the United States. Analysis of data from the National Animal Health Monitoring System Dairy 1996 study indicated that PDD was reported in cattle from 43.5% of U.S. dairy herds . In 78% of the affected herds, the first PDD cases occurred in 1993 or later. PDD typically presents in dairy cattle as lameness episodes of variable severity . It is an acute or chronic ulcerative condition that affects the skin on the bulbs of the heel or the interdigital cleft. PDD is characterized by erosion of the superficial layers of the epidermis, epithelial hyperplasia and hypertrophy, pain, swelling, and a foul odor. Lesions usually occur on the hind feet and are prone to bleeding. Early lesions are circumscribed with a red, granular (strawberry-like) appearance and variable degrees of proliferation of filiform papillae. Mature lesions are more proliferative and may have long wart-like projections. If untreated, PDD can persist for months. Significant economic losses associated with PDD occur due to reduced milk production, impaired fertility, premature culling, and the costs of treatment and control efforts . The response of PDD lesions to antibiotics is strongly suggestive of a bacterial etiology. However, the nature and location of the PDD lesions, coupled with the presence of contaminating fecal and environmental bacteria, have made definitive identification of the etiologic agent(s) difficult. Interestingly, large numbers of spirochetes are consistently found in superficial PDD lesions and also in deeper tissues where other bacteria are rarely observed (-, , , ). The role of these spirochetes in the pathogenesis of PDD is currently unknown. In 1995, Walker et al. reported the first in vitro cultivation of PDD-associated spirochetes from California dairy cattle. The fastidious, anaerobic spirochetes were differentiated into two phenotypes based on morphological, antigenic, and enzymatic properties. Both phenotypes have characteristics that are consistent with spirochetes in the genus Treponema. Read and Walker showed by immunohistochemistry that organisms similar to the California PDD-associated Treponema isolates were present in PDD lesion biopsies obtained from cattle in 16 countries. Spirochetes were absent in cattle with no history of PDD. Comparative 16S ribosomal DNA (rDNA) sequence analysis of several clonal isolates of California PDD-associated Treponema isolates indicated that they represent three phylotypes that cluster most closely with the human-associated Treponema spp. found in the oral cavity (T. denticola and T. medium/T. vincentii) or the genital area (T. phagedenis) (R. L. Walker, D. H. Read, S. J. Sawyer, and K. J. Loretz, Abstr. 79th Annu. Meet. Conf. Res. Workers Anim. Dis., abstr. 17, 1998). Using 16S rDNA sequence analysis, Choi et al. identified five phylotypes of PDD-associated Treponema in pooled punch biopsies of lesions from German cows. Three of the phylotypes clustered with T. denticola, T. vincentii, or T. phagedenis. The remaining two phylotypes clustered with the group IV oral treponemes that are also associated with periodontitis. Moter et al. showed differential distribution of the German PDD-associated Treponema phylotypes in lesions, suggesting that the development of deep lesions may correlate with the presence of a particular phylotype or combination of phylotypes. We hypothesized that analysis of the 16S-23S rDNA intergenic spacer regions (ISRs) of the PDD-associated Treponema isolates would enhance the discriminatory capability of 16S rDNA analysis due to potential heterogeneity in the length and composition of the ISRs . We present here a comparative sequence analysis of the 16S-23S rDNA ISR1 and ISR2 of six California PDD-associated Treponema isolates. Additionally, we show that PDD-associated Treponema isolates from California and Iowa cattle can be rapidly phylotyped based on PCR product length polymorphisms of the 16S-tRNAIle region of ISR2. MATERIALS AND METHODS : Bacterial strains and cultivation. | Eight clonal isolates of PDD-associated Treponema isolates (1-9185-MED, 2-1498, 7-2009, 9-227, 9-3143, 9-3301, 9-3379, and 9-3528), originally obtained from California dairy cows, were grown anaerobically (BBL GasPak Plus; Becton Dickinson, Cockeysville, Md.) at 37C in oral treponeme enrichment (OTE) broth (Anaerobe Systems, Morgan Hill, Calif.) . The first number for each isolate indicates the dairy farm, and the second number indicates the cow from which the isolate was obtained. T. denticola ATCC 35405 (American Type Culture Collection, Manassas, Va.) and T. phagedenis biovar Reiter were grown anaerobically as described previously . T. vincentii ATCC 35580 (American Type Culture Collection) was grown anaerobically in Spirolate broth (Becton Dickinson) supplemented with 0.05% (NH4)2SO4, 10% heat-inactivated normal rabbit serum, and 0.001% thiamine pyrophosphate. T. medium G7201 was grown anaerobically in NOS medium . Spirochete cell numbers were quantitated by dark-field microscopy. Escherichia coli DH5alpha (Life Technologies, Rockville, Md.), which was used as the host strain for preparation of plasmid DNA for DNA sequence determinations, was grown at 37C in Luria-Bertani (LB) broth or on LB agar containing 100 mug of ampicillin/ml. Genomic DNA isolation, PCR amplification, and cloning of the Treponema 16S-23S rDNA ISRs. | Genomic DNA was isolated from ~2 x 109 to 4 x 109 cells of the California PDD-associated Treponema isolates and the human-associated Treponema spp. with the Wizard genomic DNA purification kit (Promega, Madison, Wis.). The 16S-23S rDNA ISRs were PCR amplified (Expand Long Template PCR system; Roche Diagnostics Corp., Indianapolis, Ind.) with 10 ng of genomic DNA as the template with (i) the forward (5'-TTGTACACACCGCCCGTCA-3') and reverse (5'-GGTACCTTAGATGTTTCAGTTC-3') primer set for T. vincentii and the California PDD-associated Treponema isolates 1-9185MED, 2-1498, 9-3143, and 9-3528 or (ii) the forward (5'-CACACCGCCCGTCACACC-3') and reverse (5'-CTATTCTTTCGCTTGACC-3') primer set for T. denticola, T. medium, T. phagedenis, and the California PDD-associated Treponema isolates 7-2009 and 9-3379. To prevent spurious PCR products, the water that was used for PCR was treated with DNase I (Life Technologies) per the manufacturer's suggested protocol. PCR products were electrophoresed on a 0.8% agarose gel and stained with ethidium bromide. A DNA fragment of the anticipated size was eluted, purified, and cloned into the plasmid vector pGem-T or pGem-T Easy (Promega). Recombinant plasmids were transformed into E. coli DH5alpha, and the transformants were identified by the standard blue-white screening procedure (Promega). Recombinant plasmids from selected clones were purified with the Wizard Miniplus DNA purification system (Promega), and the inserts were confirmed by agarose gel electrophoresis following restriction enzyme digestion. DNA sequencing and analysis. | The nucleotide sequences of the Treponema 16S-23S rDNA ISRs were determined with vector-based M13 primers at the University of North Carolina at Chapel Hill Automated DNA Sequencing Facility. Both DNA strands of the inserts from four to six recombinant plasmids for each cloned, PCR-amplified ISR product were sequenced for accuracy. DNA sequence analyses were performed with MacVector 7.0 (Accelrys Inc., Madison, Wis.) and the Wisconsin Genetics Computer Group software package, version 10 (University of Wisconsin Biotechnology Center, Madison, Wis.). Multiple sequence alignments were constructed with PileUp and adjusted manually. Phylogenetic analyses were performed with PAUPSearch (Genetics Computer group). Unrooted trees were generated with maximum parsimony with a heuristic search setting with stepwise addition and tree-bisection-reconnection branch swapping. Bootstrap analysis was performed with 10,000 resamplings. PCR amplification of the 16S-tRNAIle region of ISR2. | Genomic DNA of the California PDD-associated Treponema isolates was prepared as described above. Genomic DNA from four Iowa PDD-associated Treponema isolates (1A, 3A, 4A, and 5B), originally isolated by D. Trott, T. Stanton, and M. Wannemuehler, was provided by R. Zuerner (National Animal Disease Control Center, Ames, Iowa). Approximately 10 ng of genomic DNA of each PDD-associated Treponema isolate was subjected to PCR amplification with the forward (5'-CCGCCCGTCACACCATCC-3') and reverse (5'-CCCCTTCCTTATCAGAGA-3') primers. These sequences are complementary to conserved regions in the 16S rRNA gene and tRNAIle gene, respectively. The amplification program consisted of a 94C hold for 2 min and then denaturation at 94C for 30s, annealing at 52.5C for 30 s, and extension at 72C for 1 min for 30 cycles, with an additional extension at 72C for 3 min. The PCR products were electrophoresed on a 1.7% agarose gel, stained with ethidium bromide, and photographed. Nucleotide sequence accession numbers. | The nucleotide sequences of the Treponema 16S-23S rDNA ISRs were deposited in GenBank under accession numbers to , , , and . RESULTS : Comparison of 16S-23S rDNA ISRs of PDD-associated Treponema isolates and related human-associated Treponema spp. | PCR amplification of the 16S-23S rDNA ISRs of six California PDD-associated Treponema isolates and four human-associated Treponema spp. yielded one predominant product for each of the organisms. Cloning and sequencing of the individual PCR products revealed the presence of two ISRs (ISR1 and ISR2) for each Treponema isolate or species except T. vincentii, for which only one ISR (ISR1) was identified. ISR1 contains a centrally located tRNAAla gene, whereas ISR2 contains a centrally located tRNAIle gene (Fig. and ). Both tRNA genes are 75 bp in length, and each tRNA gene is virtually identical among the Treponema isolates and species included in this study. In contrast, the regions immediately 5' and 3' of the tRNA genes (i.e., following the 3' end of 16S rDNA and preceding the 5' end of 23S rDNA) are variable in nucleotide sequence length and composition between groups of PDD-associated Treponema isolates and among the four human-associated Treponema spp. (Fig. and ). A pairwise comparison of the nucleotide sequence similarity of ISR1 and ISR2 indicated that the six California PDD-associated Treponema isolates comprised three groups: (i) 1-9185MED, 9-3143, and 9-3528; (ii) 9-3379; and (iii) 2-1498 and 7-2009 . The group 2 and group 3 PDD-associated Treponema isolates were more closely related to T. medium/T. vincentii and T. phagedenis, respectively, than they were to the group 1 PDD-associated Treponema isolates. Additionally, two (9-3143 and 9-3528) of the three group 1 isolates had identical ISR1 and ISR2 sequences. These isolates, which were obtained from separate dairy cows on the same farm, may be the same organism. Comparison of 16S-23S rDNA ISRs of California PDD-associated Treponema isolates and other spirochetes. | Although the lengths of the 16S-23S rDNA ISRs differ, the number of ISRs and the tRNA gene content of the individual ISRs were identical for the six PDD-associated Treponema isolates and for most of the human-associated Treponema spp., including T. pallidum (syphilis agent) and T. pertenue (yaws agent) . In contrast to the Treponema spp., Borrelia burgdorferi and Borrelia hermsii contain one large 16S-23S rDNA ISR in which both tRNAAla and tRNAIle genes are present (; C. Ojaimi and B. E. Davidson, Abstr. 99th Annu. Meet. Am. Soc. Microbiol., abstr. D/B-256, 1999). Brachyspira hyodysenteriae and the saprophytic and pathogenic Leptospira spp. do not contain linked 16S-23S rDNA genes and thus lack ISRs . Leptonema illini contains two 16S-23S rDNA ISRs. Only one of the ISRs has been sequenced, and it does not contain a tRNA gene . ISR-based phylogeny of California PDD-associated Treponema isolates. | Phylogenetic analyses were performed with the ISR1 and ISR2 nucleotide sequences of the six California PDD-associated Treponema isolates and the four related human-associated Treponema spp. with the maximum-parsimony method. The tree that was generated from the ISR1 data showed that the California PDD-associated Treponema isolates formed three distinct clusters . These clusters, designated phylotypes 1, 2, and 3, correspond to the three groups identified by pairwise comparisons of the ISR data . Bootstrapping confirmed the robustness of the phylotype assignments. Phylotype 1 isolates 1-9185MED, 9-3143, and 9-3528 were less closely related to the human-associated Treponema spp. than were the phylotype 2 and phylotype 3 isolates. Phylotype 2 isolate 9-3379 grouped with T. medium and T. vincentii. Phylotype 3 isolates 2-1498 and 7-2009 grouped with T. phagedenis. Interestingly, none of the phylotypes was closely related to T. denticola. The tree that was generated from the ISR2 data was essentially identical to the ISR1 tree (data not shown). The clustering of the California PDD-associated Treponema isolates into three phylotypes is consistent with the results of phylogenetic analysis performed with 16S rDNA sequence data (R. L. Walker, D. H. Read, S. J. Sawyer, and K. J. Loretz, Abstr. 79th Annu. Meet. Conf. Res. Workers Anim. Dis., abstr. 17, 1998). Typing of PDD-associated Treponema isolates based on length polymorphisms of PCR-amplified 16S-tRNAIle region of ISR2. | Availability of the 16S-23S rDNA ISR sequences of the California PDD-associated Treponema isolates facilitated the design of primer sets for PCR amplification of the 16S-tRNAAla and 16S-tRNAIle regions of ISR1 and ISR2, respectively. PCR amplification of the 16S-tRNAIle region of the six California PDD-associated Treponema isolates yielded a single, reproducible PCR product that was more readily interpretable based on intergel comparisons than the product of the 16S-tRNAAla region. Isolates of the same phylotype yielded products of virtually identical size (i.e., phylotype 1, ~330 bp; phylotype 2, similar260 bp; and phylotype 3, similar300 bp) . We PCR amplified the 16S-tRNAIle region of two additional California PDD-associated Treponema isolates (9-227 and 9-3301) and four Iowa PDD-associated Treponema isolates (1A, 3A, 4A, and 5B) to test the utility of this approach for phylotyping. Isolates 9-227 and 9-3301 yielded PCR products of similar260 and similar300 bp, respectively, which assigned them to phylotypes 2 and 3, respectively. All four Iowa isolates yielded a PCR product of similar300 bp, which assigned them to phylotype 3. Cloning and sequencing of the PCR products from two Iowa isolates (1A and 4A) indicated that the sequence of their 16S-tRNAIle region is identical and that it is >99% identical to the 16S-tRNAIle region of the California phylotype 3 isolates 2-1498 and 7-2009 (data not shown). The phylotype assignments of the eight California and four Iowa PDD-associated Treponema isolates are in agreement with the results obtained with comparative 16S rDNA sequence analysis (R. L. Walker, D. H. Read, S. J. Sawyer, and K. J. Loretz, Abstr. 79th Annu. Meet. Conf. Res. Workers Anim. Dis., abstr. 17, 1998; D. Trott, personal communication). Our results indicate that clonal isolates of the California and Iowa PDD-associated Treponema isolates can be rapidly phylotyped based on the length polymorphisms of their 16S-tRNAIle region PCR product. FIG. 1. | Multiple sequence alignment of the 16S-23S rDNA ISR1 of the six California PDD-associated Treponema isolates and the related human-associated Treponema spp. Multiple sequence alignment of the 16S-23S rDNA ISR1 of the six California PDD-associated Treponema isolates and the related human-associated Treponema spp. The 3' end of the 16S rDNA, the tRNAAla gene, and the 5' end of the 23S rDNA are indicated. Abbreviations: Tmed, T. medium; Tvin, T. vincentii; Tpha, T. phagedenis; Tden, T. denticola. FIG. 2. | Multiple sequence alignment of 16S-23S rDNA ISR2 of the six California PDD-associated Treponema isolates and the related human-associated Treponema spp. Multiple sequence alignment of 16S-23S rDNA ISR2 of the six California PDD-associated Treponema isolates and the related human-associated Treponema spp. The 3' end of the 16S rDNA, the tRNAIle gene, and the 5' end of the 23S rDNA are indicated. Abbreviations: Tmed, T. medium; Tpha, T. phagedenis; Tden, T. denticola. FIG. 3. | Unrooted phylogenetic tree (phylogram) inferred from the 16S-23S rDNA ISR1 sequences (within the region bracketed by stars [*] in Fig. Unrooted phylogenetic tree (phylogram) inferred from the 16S-23S rDNA ISR1 sequences (within the region bracketed by stars [*] in Fig. ) of the six California PDD-associated Treponema isolates and the related human-associated Treponema spp. Bootstrap analysis (10,000 resamplings) supported the placement of the California PDD-associated Treponema isolates into three phylotypes. Abbreviations: TMED, T. medium; TVIN, T. vincentii; TPHA, T. phagedenis; TDEN, T. denticola. FIG. 4. | Agarose gel analysis of PCR amplification products of 16S-tRNAIle region of eight California (C) and four Iowa (I) PDD-associated Treponema isolates. Agarose gel analysis of PCR amplification products of 16S-tRNAIle region of eight California (C) and four Iowa (I) PDD-associated Treponema isolates. Lanes: 1, 1-9185MED; 2, 9-3143; 3, 9-3528; 4, 9-3379; 5, 9-227; 6, 2-1498; 7, 7-2009; 8, 9-3301; 9, 1A; 10, 3A; 11, 4A; and 12, 5B. The positions of size standards (in base pairs) are indicated on the left side of the gel. TABLE 1 | ISR1 and ISR2 nucleotide sequence similarity values for the six California PDD-associated Treponema isolates and the related human-associated Treponema spp. TABLE 2 | Comparison of 16S-23S rDNA ISRs of six California PDD-associated Treponema isolates and other spirochetes DISCUSSION : Several studies have shown that the 16S-23S rDNA ISR is a useful target for differentiating species or in some cases strains of various bacterial pathogens of humans and animals . Our studies represent the first application of this methodology to PDD-associated Treponema isolates and to four related human-associated Treponema spp. Our results indicated that the PDD-associated Treponema isolates have two 16S-23S rDNA ISRs of variable nucleotide length and composition. ISR1 contains a tRNAAla gene, while ISR2 contains a tRNAIle gene. The copy number and tRNA gene content of the 16S-23S rDNA ISRs of the PDD-associated Treponema isolates are identical to those of the related human-associated Treponema spp. with the exception of T. vincentii, for which only one ISR was identified. Based on comparative 16S-23S rDNA ISR sequence analysis, six California PDD-associated Treponema isolates were differentiated into three phylotypes. These results are in agreement with those of 16S rDNA sequence analysis (R. L. Walker, D. H. Read, S. J. Sawyer, and K. J. Loretz, Abstr. 79th Annu. Meet. Conf. Res. Workers Anim. Dis., abstr. 17, 1998). However, we observed that 16S-23S rDNA ISR sequence analysis provided somewhat higher discriminatory power than 16S rDNA sequence analysis, since it showed that the California phylotype 1 PDD-associated Treponema isolates were not closely related to T. denticola. As noted previously, Choi et al. , using comparative 16S rDNA sequence analysis, identified five uncultured phylotypes of PDD-associated Treponema that were present in pooled lesion biopsies from four German cows. Three phylotypes (DDKL-3, DDKL-13, and DDKL-4) were most similar to T. denticola, T. vincentii, or T. phagedenis, respectively, and presumably correlate to U.S. phylotypes 1, 2, and 3, respectively. The remaining two phylotypes (DDKL-12 and DDKL-20) had no close relative but clustered to group IV human-associated oral treponemes. Schrank et al. isolated a novel Treponema (T. brennaborense) from a PDD lesion of a German dairy cow. This spirochete, which is genetically distinct from the five phylotypes of Choi et al. , is most closely related to T. maltophilum (89.5% 16S rDNA similarity). Based on comparative 16S rDNA sequence analysis of PCR products amplified from PDD lesion biopsies, there is no evidence that the DDKL-12 and DDKL-20 phylotypes or T. brennaborense are present in California cattle (R. Walker, unpublished data). Interestingly, only sequences of PDD-associated Treponema isolates that are related to T. denticola have been PCR amplified from PDD lesions of cattle in the United Kingdom . The availability of the 16S-23S rDNA ISR sequence data from the California PDD-associated Treponema isolates prompted us to develop a rapid PCR-based method for the phylotyping of clonal isolates. We found that eight PDD-associated Treponema isolates from California cows and four isolates from Iowa cows could be assigned to the correct phylotype based on 16S-tRNAIle region PCR product length polymorphisms. This method has advantages over 16S rDNA and 16S-23S rDNA ISR sequence analysis because it reproducibly yields a single, readily visible PCR product for each isolate and does not require the expense and time associated with sequencing and analyzing cloned PCR products. Although the total number of isolates that were examined was relatively small, we anticipate that PCR amplification of the 16S-tRNAIle region will be applicable to additional clonal isolates of PDD-associated Treponema isolates. Furthermore, modification of the 16S-tRNAIle region PCR should allow the development of a PCR-enzyme-linked immunosorbent assay to directly identify PDD-associated Treponema phylotypes in lesion biopsies, obviating the need for culture. The results of Moter et al. suggest that the presence of certain PDD-associated Treponema phylotypes may correlate with more invasive disease. Thus, the ability to detect and differentiate Treponema phylotypes in lesion biopsies would be a useful adjunct to in situ hybridization for studies of the epidemiology and pathogenesis of PDD. Backmatter: PMID- 12202549 TI - groESL Sequence Determination, Phylogenetic Analysis, and Species Differentiation for Viridans Group Streptococci AB - The full-length sequences of the groESL genes (also known as cpn10/60) of Streptococcus anginosus, Streptococcus constellatus, Streptococcus gordonii, and Streptococcus sanguis and the near full-length sequence of the groESL genes of Streptococcus intermedius, Streptococcus bovis, Streptococcus mitis, Streptococcus mutans, Streptococcus oralis, and Streptococcus salivarius were determined. The lengths of the groES genes from the 10 species listed above ranged from 282 to 288 bp, and the full-length sequences of groEL determined for 4 species (S. anginosus, S. constellatus, S. gordonii, and S. sanguis) revealed that each was 1,623 bp. The intergenic region (spacer) between the groES and groEL genes varies in size (15 to 111 bp) and sequence between species. The variation of the groES sequences among the species tested was greater (62.1 to 95.1% nucleotide sequence identities) than that of the groEL sequences (77.2 to 95.2% nucleotide sequence identities). Phylogenetic analysis of the groES and groEL genes yielded evolutionary trees similar to the tree constructed by use of the 16S rRNA gene. The intraspecies variation of the spacer was minimal for clinical isolates of some species. The groESL sequence data provide an additional parameter for identification of viridans group streptococcal species. Keywords: Introduction : Viridans group streptococci (VGS) are the most common etiologic agents in subacute infective endocarditis and are known to be capable of causing many serious pyogenic infections . Since the clinical significance of VGS may differ between species, it is important to identify the individual species associated with diseases and to recognize their pathogenic traits . Moreover, increases in rates of antimicrobial resistance have been noted among VGS . The difference in susceptibilities between species of VGS indicates the importance of accurate identification. At present, the species of VGS can be divided into five major groups according to their 16S rRNA sequences . These are (i) the anginosus group (also called the milleri group), which includes Streptococcus anginosus, Streptococcus constellatus, and Streptococcus intermedius; (ii) the bovis group, which includes Streptococcus bovis, Streptococcus equinus, and Streptococcus alactolyticus; (iii) the mitis group, which includes Streptococcus mitis, Streptococcus oralis, Streptococcus pneumoniae, Streptococcus sanguis, Streptococcus parasanguis, and Streptococcus gordonii; (iv) the mutans group, which includes Streptococcus cricetus, Streptococcus downei, Streptococcus mutans, and Streptococcus sobrinus; and (v) the salivarius group, which includes Streptococcus salivarius, Streptococcus thermophilus, and Streptococcus vestibularis . No single system of classification suffices for species identification of this heterogeneous group of bacteria. At present, species identification of VGS is based on physiological and biochemical characteristics determined by conventional methods, which are time-consuming . Many clinical laboratories rely on manual or automated phenotypic test systems. However, there have been variations among physiological reactions within the same species, and misidentification has occurred, particularly for some species. S. mutans strains and strains of the anginosus and mitis groups are the most problematic . Differentiation of species within the same group is often difficult . Another approach to species identification may be the use of molecular methods. Several DNA-based techniques have been developed for the identification of VGS to the species level . The target genes have included 16S rRNA genes, the tRNA gene intergenic spacer, 16S-23S rRNA spacers, the gene for d-alanine-d-alanine ligase, and the gene for manganese-dependent superoxide dismutase. The groESL genes (also known as cpn10/60 or hsp10/60), which encode 10-kDa (GroES) and 60-kDa (GroEL) heat shock proteins, are ubiquitous and evolutionarily highly conserved among bacteria. Amplification of the partial Cpn60 (or GroEL) gene segment has been used for identification of many bacteria . Goh et al. developed reverse checkerboard hybridization to identify Staphylococcus and Enterococcus species on the basis of amplification of partial Cpn60 gene sequences. Recently, we have successfully determined the Enterococcus faecalis groESL full-length sequence and developed a PCR-restriction fragment length polymorphism analysis assay for the differentiation of Enterococcus species . The goals of this study were to obtain the full-length (or nearly full-length) sequences of the groESL genes of VGS and provide another approach for species identification. MATERIALS AND METHODS : Bacterial strains. | The bacterial strains used in this study consisted of 10 reference strains obtained from the American Type Culture Collection (ATCC; Manassas, Va.) and clinical isolates which were obtained between 1997 and 1999 from the Bacteriology Laboratory, National Taiwan University Hospital, a 2,000-bed teaching hospital in northern Taiwan. S. sanguis ATCC 10556, S. gordonii ATCC 10558, S. mitis ATCC 49456, S. oralis ATCC 35037, S. intermedius ATCC 27335, S. constellatus ATCC 27823, S. anginosus ATCC 33397, S. mutans GS5, S. bovis ATCC 9809, and S. salivarius ATCC 7073 were obtained from ATCC. The clinical isolates were mostly from blood cultures and deep abscesses (e.g., brain abscesses) and were identified with the API 32 STREP system (bioMerieux Vitek, Inc., Hazelwood, Mo.). DNA preparation. | Genomic DNA was isolated and purified from the VGS with a DNA isolation kit (Puregene; Gentra Systems, Inc., Minneapolis, Minn.), according to the instructions of the manufacturer. PCR amplification and nucleotide sequence determination of groESL genes. | The strategy used to determine the groESL gene sequences of the VGS was similar to the method described previously . A partial sequence (590-bp internal fragments of the groEL genes) from the species listed above was first determined with degenerate primers (primers 590F and 590R), which were described before . The forward primer was 5'-GGNGACGGNACNACNACNGCAACNGT-3' (where N = A+C+T+G), corresponding to positions 255 to 280, and the reverse primer was 5'-TCNCCRAANCCNGGYGCNTTNACNGC-3', corresponding to positions 844 to 819. After determination of the sequence of this fragment in each species, the unknown sequences of the 3' and 5' ends of the gene were amplified with the LA-PCR in vitro cloning kit (Takara Shuzo Co. Ltd., Tokyo, Japan). Briefly, genomic DNA was digested with restriction enzymes. DNA fragments were ligated with a cassette adapter. The ligation mixture of the adapter and genomic DNA fragments was used as a template for PCR. The amplification was performed with one cassette primer (cassette primer C1 for the first PCR and cassette primer C2 for the second PCR) supplied by the manufacturer and a target gene-specific primer to walk downstream on the DNA sequence. For the amplification, 2.5 U of TaKaRa LA Taq was mixed with 10 pmol of each primer, each deoxynucleotide triphosphate at a concentration of 20 mM, 5 mul of 10x buffer (containing MgCl2, provided in the kit), and 5 mul of template DNA in a final volume of 50 mul. The amplification was performed in a thermal cycler (MJ Research, Inc., San Francisco, Calif.) with the following program: an initial 5-min denaturation step at 94C; 35 cycles of denaturation (94C for 30 s), annealing (50C for 60 s), and extension (72C for 2 min); and a final 7-min extension step at 72C. The amplified fragments were subsequently sequenced on a model 377 sequencing system (Applied Biosystems, Foster City, Calif.) with the Taq BigDye-Deoxy Terminator Cycle Sequencing kit (Applied Biosystems), according to the instructions of the manufacturer. Determination of other groESL sequences. | The LA-PCRs were not successful for six of the species tested. Sequences of nearly full length were obtained from these species by combining two overlapping fragments of amplified products with two pairs of primers, with one pair targeting the upstream sequence of groES (primer Strep-ES-UP [5'-GACTATTTCTGACCAAGTGAT-3']) and the 5' region of groEL (primer Strep-EL-120-100 [5'-CTCAAGAACAACRTTRCGDCC-3']) and the other pair targeting nucleotide positions -8 to 20 (forward primer [5'-TCGAATTCATGTTNAARCCNTTNGG-3']) and 1623 to 1603 (reverse primer [5'-YTACATCATNCCNCCCATCAT-3']). Phylogenetic analysis. | DNA and amino acid sequences were aligned by using Gene-Works software (IntelliGenetics, Mountain View, Calif.). The phylogenetic relationships among species were analyzed by the neighbor-joining method listed in the MEGA (molecular evolutionary genetic analysis) analytical package . For the neighbor-joining analysis, the distance between the sequences was calculated by using Kimura's two-parameter model. Levels of similarity were determined among species. Bootstrap values were obtained for 500 randomly generated trees. Amplification and partial sequencing of groESL for examining intraspecies variations among clinical isolates. | Intraspecies variations in the groES, spacer, and partial groEL sequences were examined for clinical isolates. The sequences were determined by PCR and sequencing. The primers used to amplify groES and the spacer were Strep-ES-UP and Strep-EL-120-100, described above. The primers used to amplify the partial groEL fragment were 590F and 590R. Nucleotide sequence accession numbers. | The nucleotide sequences of the groESL genes determined in this study were deposited in the GenBank sequence database. The accession numbers for the full-length groESL genes are as follows: S. anginosus, AF378195; S. constellatus, AF378196; S. gordonii, AF338228; and S. sanguis, AF378197. The accession numbers for the nearly full-length groESL genes are as follows: S. bovis, AF389514; S. intermedius, AF389515; S. mitis, AF417589; S. mutans, AF389516; S. oralis, AY38047; and S. salivarius, AF389517. RESULTS : Nucleotide sequences of the groESL genes. | After the sequence of the 590-bp partial groEL fragment was determined, the full-length sequences of the groESL genes from S. anginosus, S. constellatus, S. gordonii, and S. sanguis were subsequently determined by the LA-PCR method. The sequence data revealed that the first open reading frame (groES homologue) from these four species was 282 bp in length and the second open reading frame (groEL homologue) was 1,623 bp in length (the deduced amino acid residues consisted of 540 amino acids). A putative promoter (TTGACT [-35]-Nx-TACAAT [-10]) (where Nx represents different numbers of nucleotides) was seen upstream of the groES genes. In addition, a perfect match of a putative CIRCE element (TTAGCACTC-Nx-GAGTGCTAA) between the putative promoter and the groES start codon was observed in these four species. Since the sequences upstream of the groES sequences of the species listed above exhibited high degrees of similarity, primers Strep-ES-(-35) and Strep-spacer-1R were designed to amplify the entire groES sequence and the spacer region. The amplification product coupled with a downstream fragment (the nearly full length of the groEL genes) obtained with another pair of primers was used to determine the nearly full-length groESL sequences of the other five species. The lengths of the groES genes of these species were as follows: S. intermedius, 282 bp; S. bovis and S. mitis, 285 bp; and S. mutans and S. salivarius, 288 bp. Comparisons of the groES, spacer, and groEL sequences among reference strains of species. | Table presents the pairwise nucleotide identities of reference strains of the species tested calculated from the nucleotide sequences of the groES genes. The groES sequence similarities among 10 reference strains of the species tested ranged from 62.1 to 95.1% for nucleotide sequence identity and 55.7 to 95.9% for amino acid sequence identity. On the basis of the deduced GroES amino acid sequences, S. mitis (or S. oralis) and S. pneumoniae displayed the highest degree of similarity (95.9% similarity), followed by species within the anginosus group (92.8 to 93.8% similarity). The similarities between other pairs were usually less than 90%. The groEL sequence similarities among 10 reference strains of the species tested ranged from 77.2 to 95.2% at the nucleotide sequence level and 87.4 to 99.1% at the amino acid level . The spacer length ranged from 15 to 111 bp . The lengths and sequences of the spacers were found to be species specific except for S. constellatus and S. intermedius (32 bp) and S. mitis (or S. oralis) and S. pneumoniae (15 bp). The spacers of S. mitis (or S. oralis) and S. pneumoniae were the same size but had one nucleotide sequence difference. Phylogenetic relationships. | The phylogenetic relationships derived from comparisons of the groES or groEL sequences are presented in Fig. . The phylogenetic analysis revealed that the nucleotide sequences of the groES and groEL genes from 10 reference strains of the species tested were divided into five clusters. The data revealed that species within a group are highly related. For example, S. anginosus, S. constellatus, and S. intermedius are highly related. Similarly, S. mitis, S. oralis, and S. pneumoniae are highly related. S. sanguis and S. gordonii are closely related to each other. Intraspecies variation. | To evaluate the general applicability of the groESL sequence-based species identification, the groES, spacer, and groEL sequences from clinical isolates were determined to identify intraspecies variations. Three to five isolates of each species were tested. The results of a comparison of the sequences of clinical isolates and those of reference strains of the same species are shown in Table . The full length of groES and the spacer were tested for all clinical isolates. The full lengths of the groEL sequences of nine isolates of species in the anginosus group were also tested; for the other isolates a partial region of groEL (590 bp) was tested. The identities of the groES sequences among isolates of a species ranged from 94 to 99%. The identities of the groEL sequences among isolates of a species ranged from 92 to 99%. The intraspecies identities of the spacer were quite high. The greatest intraspecies variation was seen among isolates of S. mitis and isolates of S. oralis. For both species, the nucleotide base differences between the sequences of clinical isolates and that of the reference strain were between 84.9 and 98.9% for groES and 84.9 and 99.4% for groEL. The intraspecies similarities of groES and groEL were similar for the S. bovis reference strain and the five S. bovis isolates tested. However, the sequences of the spacer were different between biotypes. Two biotype I isolates had spacer sequences identical to that of ATCC 9809 (biotype I) . Three biotype II/2 isolates had spacer sequences identical to each other, but their spacer sequences were different from that of ATCC 9809 at four nucleotides. FIG. 1. | Phylogenetic relationships of various species of VGS, S. pneumoniae Phylogenetic relationships of various species of VGS, S. pneumoniae, and S. pyogenes based on the nucleotide sequences of the groES and groEL genes. The phylogenetic tree was generated by the neighbor-joining method within the MEGA package. The numbers at the nodes are the percentages of occurrence in 500 bootstrapped resamplings. (A) Phylogenetic tree showing relationships of groES genes; (B) phylogenetic tree showing relationships of groEL genes. TABLE 1 | groES nucleotide and amino acid sequence similarities among reference strains of species TABLE 2 | groEL nucleotide and amino acid sequence similarities among reference strains of species TABLE 3 | Nucleotide sequences and lengths of spacers between groES and groEL among species TABLE 4 | Intraspecies variations of groES, spacer region, and groEL (partial) sequences of clinical isolates compared with those of ATCC reference strains DISCUSSION : The groESL (cpn10/60) gene sequences have previously been evaluated for use in the differentiation of several bacterial species . In the present study, we determined the groESL sequences of 10 commonly encountered species of VGS. Full-length groES sequences were obtained from each species. Four full-length groEL sequences were obtained. The 3' ends of the groEL sequences from the remaining six species were not complete by a few bases. The sequence data revealed that the gene structure of groESL from VGS was similar to those of genes from other genera or species . The lengths of the groES genes from 10 species ranged from 282 to 288 bp. The length of groEL from S. anginosus, S. constellatus, S. gordonii, and S. sanguis was 1,623 bp, which was the same as that of S. pneumoniae but a little shorter than that of Streptococcus pyogenes (1,632 bp). The putative CIRCE element of groESL was found and was shown to be perfectly conserved, as in other gram-positive bacteria . The putative promoter region was also conserved among these species. Since part of the upstream regions of the groES sequences of the species tested exhibited high degrees of similarity, primers Strep-ES-(-35) and Strep-spacer-1R were designed to amplify the entire groES gene, the spacer region, and the 5' end of the groEL gene of other species. We also tested this pair of primers with Staphylococcus, Enterococcus, and Escherichia coli and found that it could amplify the DNAs of most streptococcal species but not those of other genera. Similar to other gram-positive bacteria, the C terminus of GroEL does not consist of three tandem repeats of GGM. Instead, the sequence is PG(S)MMGGM(F). When the sequences that were determined (including the S. pneumoniae groESL sequence from GenBank) were compared to each other, the overall groES nucleotide sequence identity among reference strains ranged from 62.1% (S. mitis and S. salivarius) to 95.1% (S. mitis and S. pneumoniae), and the groEL nucleotide sequence identity ranged from 77.2% (S. mutans and S. sanguis) to 95.2% (S. constellatus and S. intermedius). For groES or groEL, more than 90% nucleotide sequence identity was found between pairs of species of the same group. The nucleotide sequences of pairs of species of different groups usually showed <90% identity. The deduced amino acid sequences also show considerable differences. For GroES, the pairs showed identities of 55.7% (S. oralis or S. pneumoniae and S. salivarius) to 95.9% (S. mitis or S. oralis and S. pneumoniae). For GroEL, the pairs showed identities of 87.4% (S. pneumoniae and S. mutans or S. salivarius) to 99.1% (S. mitis and S. oralis). The results show the higher degree of divergence of groES genes than of groEL genes and the usefulness of groES gene divergence in investigations of the relationships of closely related species. The intergenic (spacer) region between the GroES translation termination codon and the putative translation start codon for GroEL vary in size (15 to 111 bp) and sequence between species. However, the spacer sizes were identical between S. constellatus and S. intermedius (32 bp) and S. mitis (or S. oralis) and S. pneumoniae (15 bp). The spacers of S. mitis (or S. oralis) and S. pneumoniae were the same size, with only one nucleotide sequence difference. The results of this study and data from GenBank indicate that the spacer length usually varies among different species. The spacer length is 57 bp in E. faecalis , 75 bp in Staphylococcus aureus, 87 bp in Lactococcus lactis, 36 bp in Lactobacillus zeae, 46 bp in Bacillus subtilis, and 45 bp in E. coli. The phylogenetic analysis revealed that the nucleotide sequences of the groES and groEL genes from the 10 species tested were divided into five clusters. The phylogenetic tree based on the deduced amino acid sequences of either groES or groEL (data not shown) was similar to the tree based on the nucleotide sequences. Evolutionary trees derived from groES or groEL sequences demonstrate remarkable similarities to those derived from 16S rRNA gene sequences . Phylogenetic analysis of the groES or groEL gene and the 16S rRNA gene revealed a general likeness in the relative position of each species within the tree. Similar to other genes, three species of the anginosus group clustered closely together . The five species in the mitis group were further divided into two subgroups. S. mitis, S. oralis, and S. pneumoniae are highly related. S. sanguis and S. gordonii are more closely related to each other than to the other species. Of the 10 species examined, the greatest diversity was shown for the species in the mitis group. In previous studies, sufficient heterogeneity was revealed by analyzing the tDNA-intergenic spacer length polymorphism within this group . The species in the mitis group are not always clustered together. The phylogeny determined by analysis of groEL shows that S. sanguis and S. gordonii are more closely related to the species of the anginosus group than to the other species of the mitis group. Intraspecies genetic variation of groEL has been noted in Bartonella species . In order to determine the utility of groEL, groES, or the spacer for species identification, we evaluated the intraspecies variations of these genes in clinical isolates. Despite a certain degree of intraspecies polymorphism, the sequences of strains of the same species were more similar to each other than to those of strains that belonged to a different species. The sequencing results revealed that the overall intraspecies variation is low compared to the interspecies variation. Although the interspecies variations of the groES gene sequences were higher than those of the groEL sequences, the intraspecies variations were similar. The slight divergence of the groES or groEL sequences seen within members of the same species may be useful for strain typing. The levels of intraspecies divergence of the groES or groEL sequences vary from one species to another. The degree of intraspecies variation of the groES sequence was low (less than 1%) for S. intermedius, S. constellatus, and S. mutans. Species in the mitis group are more heterogeneous than the species in the other groups. Although it is easy to differentiate between S. sanguis (or S. gordonii) and S. mitis (or S. oralis), it is difficult to differentiate S. mitis and S. oralis by groESL sequence analysis alone. S. mitis and S. oralis are known to be difficult to differentiate by both biochemical methods and 16S rRNA gene sequencing . S. mitis isolate 7593 had a low level of identity with the ATCC reference strain. Two possibilities might exist. One is that strain 7593 is not really an S. mitis strain. Another possibility is that the groESL gene in S. mitis is more divergent than that in other species. The levels of intraspecies divergence in S. mitis and S. oralis suggest that they are more heterogeneous than the other species of VGS. Therefore, no single system of classification is satisfactory, and determination of the sequences of multiple targets in combination with phenotypic testing may perhaps be able to solve the problem. The spacer length and sequence were found to be species specific in the ATCC reference strains tested. For clinical isolates of S. anginosus, S. mutans, S. sanguis, and S. gordonii, there was a perfect correlation between the spacer sequence and the result of the API 32 STREP phenotypic tests. The groES and groEL sequences of the S. bovis reference strain and five clinical isolates tested were similar. The sequences of the spacers were different, however. The S. bovis reference strain used in this study was biotype I. S. bovis strains are mainly divided into two groups, biotype I and biotype II, according to their fermentation of mannitol and glucan synthesis. Biotype II is further divided into type II/1 and type II/2 . Recently, the S. bovis group has been defined on the basis of 16S rRNA gene sequences, ribotyping, and whole-cell protein electrophoresis patterns. Isolates with different biotypes may have different clinical significance. Clarridge et al. reported that S. bovis biotype II/2 isolates form a separate genospecies and are the most common isolates in adult males. Therefore, it is not surprising that the spacer sequences of three of the clinical isolates tested (type II/2) were different from those of the ATCC reference strain and the other two isolates (type I). More isolates need to be tested to determine whether the spacer sequences are consistent with particular biotypes. In clinical laboratories, S. bovis is also confused with S. salivarius when phenotypic identification is used. However, these two species can easily be distinguished by comparing their groES, groEL, or spacer sequences. In conclusion, the groESL sequence was less conserved than the 16S rRNA gene sequence. Therefore, it is potentially useful in the differentiation of closely related species. The groESL sequences of VGS were shown to be useful for the differentiation of species, with a few exceptions. The groESL gene sequence can provide an additional parameter for species identification of VGS, particularly when 16S rRNA sequences share high degrees of similarity. Backmatter: PMID- 12202553 TI - Comparison of Rapid, Automated Ribotyping and DNA Macrorestriction Analysis of Burkholderia pseudomallei AB - An automated ribotyping device (RiboPrinter) was used to determine the ribotypes of a collection of Burkholderia pseudomallei isolates. In a preliminary evaluation with the restriction enzymes BamHI and EcoRI, the protocol with EcoRI was more discriminating. The reproducibilities of the ribotypes obtained with EcoRI (EcoRI ribotypes) were determined by testing three levels of bacterial loads. The performance of the manufacturer's software was assessed by comparing the machine-optimized ribotypes with the type determined from the original gel image analyzed with Bionumerics software. The library of B. pseudomallei EcoRI ribotypes was then compared with the ribotypes obtained by DNA macrorestriction analysis of XbaI digests by pulsed-field gel electrophoresis. The typeability of B. pseudomallei by EcoRI ribotyping was 100%, and the discrimination index was 0.94. The slightly greater discrimination provided by DNA macrorestriction analysis (0.96) was achieved at the expense of a significantly longer processing time of 6 days, although the method was only half the cost of automated ribotyping. Typeability by macrorestriction analysis was lower (97%) unless a thiourea step was added to neutralize the action of Tris-dependent endonucleases. The digital record of B. pseudomallei isolates analyzed thus far provides a useful resource for future epidemiological studies and will help shorten the response time in the event of a further melioidosis outbreak or the deliberate release of B. pseudomallei as a biohazard. Keywords: Introduction : Burkholderia pseudomallei, the soil- and waterborne bacterial species that causes melioidosis, is a member of the taxonomically complex genus Burkholderia. The genus has gained many additional species since it was formed from Pseudomonas RNA group II in 1992 . Identification of B. pseudomallei isolates in the diagnostic laboratory can be difficult due to the misleading results generated by conventional phenotypic identification systems such as substrate utilization panels . Genetic typing methods are increasingly being used to clarify the relationship between and within Burkholderia species. Ribotyping has been used extensively to analyze B. pseudomallei and other clinically important Burkholderia species . The technical demands and time required to complete a single analysis restrict ribotyping to centers with a Burkholderia research interest. The more accessible method of DNA macrorestriction analysis (pulsed-field gel electrophoresis [PFGE]) is widely used for molecular typing of B. pseudomallei . Both methods have been used to investigate suspected Burkholderia sp. outbreaks , but the time and expertise needed have prevented more widespread adoption of either method. When an acute melioidosis outbreak occurred in Western Australia in late 1997, no molecular typing method was available for Burkholderia species locally . Clinical and environmental isolates had to be dispatched out of the state for molecular typing. Once a PFGE method had been established at this center, molecular typing results could be obtained about 1 week after receipt, providing that no higher-priority epidemiological investigations were already under way. Shortly after we obtained an automated ribotyping device (RiboPrinter; Qualicon, Inc., Wilmington, Del.), a European group published its comparison of automated ribotyping methods with the restriction enzymes EcoRI and PvuII with DNA macrorestriction analysis for typing of Burkholderia species . The European study concentrated on B. cepacia and did not include an analysis of B. pseudomallei. In the present study we sought to establish whether an automated ribotyping method could be used to subtype B. pseudomallei isolates and how it would compare with DNA macrorestriction analysis. MATERIALS AND METHODS : Storage, selection, and recovery of bacterial strains. | Bacterial strains are maintained in the Western Australian Culture Collection in 20% glycerol broth at -70C. All stored isolates were identified with a substrate utilization panel (API 20NE system; BioMerieux, Marcy l'Etoile, France), and their identities were confirmed by PCR-based nucleic acid amplification with B. pseudomallei-specific primers . B. pseudomallei strains were obtained from the Western Australian Culture Collection and included a collection of 11 isolates from the Western Australia melioidosis outbreak and 20 other strains from unrelated, distinct geographic locations. The preliminary ribotype analysis and subsequent comparisons of cluster identifications were performed with the entire collection of outbreak-related isolates. Other analyses performed for determination of typeability and the discrimination index used only one isolate from the outbreak collection. The isolates chosen for reproducibility assessment were picked at random from among those in the unrelated strain collection used for typeability and discrimination analyses. The strains were resuscitated by inoculation of 5% horse blood agar and incubation for 24 h at 37C in air and were checked macroscopically for purity (for details, see Table ). Automated ribotype analysis. | Ribotyping was performed with an automated ribotyping device (RiboPrinter; Qualicon, Inc.) and proprietary reagents (Qualicon, Inc.). Bacterial strains were streaked on 5% horse blood agar and incubated for 24 h at 37C in air to produce single-colony growth. The primary inoculum was touched with the end of a proprietary inoculation device (Stickpick; Qualicon, Inc.), which was used to inoculate 200 mul of sample buffer. Thirty microliters of the mixture was transferred to the sample carrier and heated to 80C in the RiboPrinter heating station (Qualicon, Inc.). Five microliters of each lysing agent was then added, and the sample carrier was transferred to the automated analyzer. The remainder of the procedure was conducted in the automated analyzer over 8 h. The results were then transferred to a dedicated microcomputer and interpreted with the proprietary software, as described below. In the first series of analyses, ribotyping with the restriction enzymes BamHI and EcoRI (BamHI and EcoRI ribotyping) was performed with B. pseudomallei isolates to determine which enzyme was most suited to our needs. The results were compared with those obtained with Bionumerics software (Bionumerics version 2.5; Applied Maths, Kortrijk, Belgium), as described below. In the second series of analyses, the EcoRI ribotyping procedure was repeated with a random selection of B. pseudomallei isolates picked at three increasing inoculum densities to determine the reproducibilities of the machine-generated ribotypes. The remaining B. pseudomallei isolates were processed by the EcoRI ribotyping protocol, and the resulting software-optimized patterns ("riboprints") were compared with the unprocessed ribotype gel patterns obtained with the Bionumerics software. DNA macrorestriction analysis. | PFGE was performed on all B. pseudomallei strains with XbaI and double digestion of bacterial DNA by a previously reported method . Gels were scanned by using Quantity One software and a Geldoc scanner (Bio-Rad). The results were analyzed by direct visual inspection and the gel analysis component of the Bionumerics version 2.5 software. The PFGE type and the ribotype were compared with the unprocessed ribotype gel images. The Bionumerics analytical software was used to compare the EcoRI ribotype with the pulsotype to produce a composite dendrogram and to enable three-dimensional cluster analysis. Analysis of molecular typing gel data. | Dendrograms were produced with the Bionumerics software by using a band-based similarity index (Dice coefficient) with equal weighting for each typing system. The discrimination index and the typeability were calculated from the formula recommended by Hunter and Gaston . Epidemiological concordance was analyzed by comparison of the clustering of band patterns from epidemiologically related isolates. The principal outbreak cluster and adjacent isolates linked to the cluster at 90% relatedness or greater were identified on each dendrogram image. Fisher's exact test was applied to the two-by-two contingency table of ribotype cluster or not versus macrorestriction cluster or not. Fisher's exact test was performed with Prism version 2.01 software (GraphPad Software Inc., San Diego, Calif.). TABLE 1 | B. pseudomallei isolates used in this investigation RESULTS : Comparison of BamHI and EcoRI. | Fewer ribotype bands were produced per isolate by the protocol with BamHI than by the protocol with EcoRI. The reduced diversity of ribotypes produced by the protocol with BamHI and the fact that the automated ribotyping device was optimized for the protocol with EcoRI led us to concentrate on using EcoRI. EcoRI ribotype reproducibility. | During early RiboPrinter runs with EcoRI the reproducibility of the results was questioned. Three separate analyses were conducted with each selected isolate. Repeated EcoRI ribotyping of several isolates of B. pseudomallei generated distinct ribotype reference codes on the second or subsequent analysis. Advice from the manufacturer's technical support service identified the strength of the optical signal generated by detection of the probe as a likely cause. After postanalysis optimization of ribotype patterns and application of the merge and split functions of the analytical software, a more reproducible result was obtained. The repetition of EcoRI ribotyping with a series of isolates at increasing inoculum densities from one pick in 200 mul, two picks in 200 mul, and two picks in 100 mul showed that the optimal inoculum for DNA extraction was two picks suspended in 100 mul. These optimal conditions were used to complete the remainder of the study. Unprocessed versus machine-optimized ribotype comparison. | The proprietary RiboPrinter software does not generate a dendrogram . The dendrogram of machine-generated optimized ribotypes analyzed with Bionumerics software did not group the related isolates as well as the gel images analyzed with Bionumerics software. The cluster of Bionumerics software-analyzed gel image results for epidemiologically related isolates showed 90% relatedness or better, whereas the RiboPrinter-optimized gel images showed 80% relatedness or better. Comparison of ribotyping with DNA macrorestriction analysis. | Comparison of automated EcoRI ribotyping of B. pseudomallei with XbaI DNA macrorestriction analysis confirmed that PFGE was more discriminating than ribotyping (Fig. and ). While all isolates were typeable by the EcoRI ribotyping method, several were not typeable by the conventional PFGE method. The level of typeability by PFGE was raised to 100% by the addition of thiourea to reduce the level of DNA degradation by Tris-dependent endonucleases, as reported recently for Pseudomonas aeruginosa . Completion of PFGE by this method took a minimum of 6 days from a live culture start. Automated ribotyping, in contrast, took just over 8 h at a total cost of about A$120 (A$1 = US$0.55) per isolate, compared to a total cost of about A$60 for macrorestriction analysis. Both EcoRI ribotyping and DNA macrorestriction analysis identified the outbreak isolate collection as a distinct cluster among the larger collection of epidemiologically unrelated strains (indicated by a dot alongside the corresponding band pattern). Of the 11 isolates from the outbreak cluster, 8 were found to be closely linked by both methods. The Bionumerics software placed two isolates (isolates 10 and 22) from the ribotype cluster alongside the rest of the outbreak isolates but linked them only at the 82% level, despite a visibly similar appearance. Twenty-one isolates unconnected with the outbreak were correctly placed outside the outbreak cluster by both methods. Ten of the outbreak isolates were linked at the 100% level by EcoRI ribotyping, whereas only five of the outbreak isolates were linked at the 100% level by macrorestriction analysis. FIG. 1. | Riboprinter output (Copyright 1993-2000. Riboprinter output (Copyright 1993-2000. Qualicon, Inc., a DuPont Company. All rights reserved, used under permission of Qualicon.) showing results for B. pseudomallei isolates including an epidemiologically related cluster (uppermost four isolates). FIG. 2. | Composite of results for all B. pseudomallei Composite of results for all B. pseudomallei isolates analyzed by automated EcoRI ribotyping with the corresponding dendrogram, with epidemiologically related isolates indicated (black dots). FIG. 3. | Composite of results for all B. pseudomallei Composite of results for all B. pseudomallei isolates analyzed by XbaI DNA macrorestriction analysis with the corresponding dendrogram, with epidemiologically related isolates indicated (black dots). DISCUSSION : In this analysis we demonstrated the feasibility of using an automated method to ribotype B. pseudomallei. It is said that ribotyping analyzes about 0.5% of the total genome, while PFGE examines about 45% . The level of discrimination achieved by the automated EcoRI ribotyping method compared favorably with that achieved by the lengthier PFGE method. The much faster automated ribotyping method produced an acceptable approximation of the clustering of epidemiologically related isolates achieved by PFGE. The BamHI ribotyping protocol was less discriminating than the EcoRI ribotyping protocol. Once postanalytical result optimization had been mastered, the EcoRI ribotyping protocol successfully typed 100% of the isolates analyzed. Conventional ribotyping of B. pseudomallei has been performed with EcoRI . BamHI has also been used in recent ribotyping studies of B. pseudomallei . A combination of restriction endonucleases has previously been used to ribotype B. pseudomallei by a conventional, nonautomated procedure . Both the automated ribotyping system and the new analytical Bionumerics software should make multiple-enzyme analyses more easily attainable in future. The reproducibility of the ribotype data is particularly important when the data are stored in digital form for comparison with ribotype results from a later analysis or another laboratory. We were surprised at first by the generation of different ribotype reference codes by the manufacturer's software when specific B. pseudomallei isolates were ribotyped on a second or subsequent occasion. Having established how the manufacturer's software can be used to correct variations in background noise on the gel images, we recognize that the results are more reproducible than we originally thought. Important causes of weak bands and high levels of background noise are low and high bacterial DNA loads, respectively, as confirmed by our analysis of a small collection of strains processed repeatedly. It can be expected that the manufacturer's software will continue to generate new ribotypes until a much larger collection of epidemiologically unrelated strains has been incorporated into the ribotype library. We used a live culture start for the ribotyping protocol. Work with B. pseudomallei dictates careful attention to safety in order to avoid laboratory-acquired infection . This required performance of preparation steps for ribotyping in a biological safety cabinet. In our hands DNA macrorestriction analysis with XbaI produces clear and highly discriminating results with B. pseudomallei isolates. A proportion of isolates were untypeable by the previously published method. These isolates were successfully typed following the addition of thiourea, as originally described to prevent DNA degradation by Tris-dependent endonuclease in P. aeruginosa . This step ensured the 100% typeability of B. pseudomallei isolates in our collection. PFGE is a lengthier and more labor-intensive typing process but adds discriminatory power and in combination with ribotyping enhances the accuracy with which clustering can be delineated. PFGE is likely to remain the benchmark molecular typing method in service laboratories for some time to come and can be used as a dissimilar confirmatory method and as a first-line method when rapid turnaround is of little consequence. Previous studies have demonstrated multiple ribotypes of B. pseudomallei, some of which appear to predominate in the main area where melioidosis is endemic . Comparison with our results suggests that the Western Australia melioidosis outbreak was caused by a strain belonging to the commonest group reported. It is a matter of concern that the commonest ribotype of B. pseudomallei should be capable of apparent waterborne dissemination. In view of the association between BamHI ribotype 4 and high rates of mortality , more detailed work with BamHI is required, despite its lower discriminatory power in the present study. Given the correlation between B. pseudomallei ribotype and virulence, the generation of ribotype analyses of clinical and environmental isolates can be expected to assist future investigations into the pathogenesis and ecology of melioidosis. In conclusion, this comparison of automated ribotyping with DNA macrorestriction showed that an EcoRI ribotyping protocol can be used to obtain discriminating molecular typing data on all isolates analyzed. Optimal discrimination was obtained by analyzing gel images of automated EcoRI ribotype patterns obtained with Bionumerics software in combination with the results of DNA macrorestriction analysis. Our experience suggests that automated ribotyping can be applied to the investigation of melioidosis, particularly for a rapid response to suspected common-source incidents, epidemiological surveillance, and biopreparedness. Further work in collaboration with other centers is now required to generate an internationally representative database of B. pseudomallei ribotypes to add to the preliminary collection. Backmatter: PMID- 12202597 TI - Evaluation of GenoType and LiPA MYCOBACTERIA Assays for Identification of Finnish Mycobacterial Isolates AB - Two DNA strip assays, INNO-LiPA MYCOBACTERIA and GenoType Mykobakterien, were evaluated for identification of 81 Finnish mycobacterial isolates. The LiPA assay correctly identified 89.4% of the 66 isolates studied, and the GenoType assay identified 95.1% of 81 isolates. The GenoType assay had a wider selection of species and less stringent temperature requirements. Keywords: Introduction : Of the more than 100 mycobacterial species identified to date, at least 21 are pathogenic to humans and frequently isolated from clinical samples . Immunocompromised patients are especially vulnerable to opportunistic infections caused by mycobacteria other than Mycobacterium tuberculosis (MOTT). MOTT infections usually occur in developed countries, where the incidence of tuberculosis is low, whereas in developing countries M. tuberculosis remains the most common cause of mycobacterial disease . The increasing incidence of MOTT infections has made it important to rapidly identify mycobacteria at the species level, as treatment varies according to the species responsible for the infection. Molecular biological methods such as DNA sequencing , PCR-restriction fragment length polymorphism (RFLP) assays , and commercial tests such as the AccuProbe (Gen-Probe Inc., San Diego, Calif.) have replaced conventional biochemical tests for the identification of mycobacteria. The new methods have greatly improved both the speed and accuracy of mycobacterial diagnostics . However, the methods have their limitations: DNA sequencing is rather time-consuming and requires expensive equipment, while the differentiation of mycobacterial species by PCR-RFLP requires the use of several restriction enzymes. The drawback of the AccuProbe test is the limited number of species that can be identified. Recently, DNA strip technology, based on the reverse hybridization of PCR products to their complementary probes, has been applied to simultaneous detection and identification of mycobacteria. Currently, two DNA strip assays, INNO-LiPA MYCOBACTERIA (Innogenetics N.V., Ghent, Belgium) (LiPA) and GenoType Mykobakterien (Hain Lifescience GmbH, Nehren, Germany) (GenoType), are commercially available. Both assays provide probes for the M. tuberculosis complex, Mycobacterium avium, Mycobacterium intracellulare, Mycobacterium kansasii, Mycobacterium chelonae, Mycobacterium gordonae, Mycobacterium xenopi, and Mycobacterium scrofulaceum. In addition, the LiPA strip can identify members of the M. avium complex and differentiate between the three M. chelonae and three M. kansasii subgroups. The GenoType strip has additional probes for Mycobacterium celatum, Mycobacterium malmoense, Mycobacterium peregrinum, Mycobacterium phlei, and two subgroups of Mycobacterium fortuitum. The performance of the LiPA test has been assessed using BACTEC 12B bottles on a panel of clinical isolates from the United States and on clinical isolates collected from Brazil and Italy . To our knowledge, the GenoType test has not been evaluated previously. Since intraspecies variation within mycobacteria isolated from different geographical regions has been reported , we wanted to assess the capability of the two assays to correctly detect and identify mycobacterial isolates obtained from patients living in Finland. Further, the two tests were compared for cost-effectiveness, ease of use, and interpretation of results. Bacterial strains. : Eighty-one clinical mycobacterial isolates were selected from the strain collection of the Mycobacterial Reference Laboratory, National Public Health Institute, Turku, Finland . The strains were isolated in 1990-2001 from patients living in Finland. The isolates were chosen to represent the mycobacterial species identified by the two tests. The isolates had been identified to species level either by the AccuProbe test (Gen-Probe Inc.) or by 16S ribosomal DNA (rDNA) sequencing and phenotypic characteristics. Members of the M. tuberculosis complex had been further differentiated using the nitrate test and allele-specific amplification . The bacteria were cultivated on Lowenstein-Jensen medium and incubated at 37C. For the GenoType and LiPA assays, DNA was prepared according to the LiPA manufacturer's instructions by boiling at 100C for 10 min and centrifuging at 13,000 rpm (centrifuge 5415D; Eppendorf AG, Hamburg, Germany) for 5 min. TABLE 1 | Identification of mycobacterial isolates by LiPA and GenoType assays LiPA and GenoType assays. : The assays were carried out according to the manufacturer's instructions, using the reagents provided with the LiPA and GenoType kits. Both protocols consisted of PCR amplification, hybridization of the PCR products to the strips, and detection and interpretation of the results. PCR amplification. : Table summarizes the results of both assays. The results of PCR amplification were always confirmed by gel electrophoresis. LiPA PCR, targeting the 16S-23S rRNA spacer region, yielded 400- to 550-bp amplicons for 78 of the 81 isolates studied. The three isolates, which remained negative in LiPA PCR, despite repeated attempts, had been identified as M. chelonae by 16S rDNA sequencing. The same DNA preparations of the three isolates were successfully amplified by GenoType PCR, which targets the 23S rDNA, and identified by the GenoType assay as M. chelonae. With GenoType PCR, the approximately 200-bp amplicons were detected for 80 of the 81 isolates studied. The one isolate that constantly remained negative in GenoType PCR had been identified as M. intracellulare by the reference methods. The same DNA was successfully amplified in LiPA PCR and correctly identified as M. intracellulare. In both assays, all isolates that were PCR positive hybridized to the Mycobacterium genus probe. LiPA assay. : As the LiPA strip has no species-specific probes for M. celatum, M. fortuitum, and M. peregrinum, the assay was evaluated using the 66 strains with species in the identification range of LiPA. Moreover, the LiPA assay has no species-specific probe for M. malmoense, but this species is positive with the M. avium-M. intracellulare-M. scrofulaceum complex (MAIS) probe. If identification at the MAIS complex level is considered correct, the LiPA assay correctly identified 59 of the 66 (89.4%) strains. However, when the nine M. malmoense strains were excluded from the study, the LiPA assay correctly identified 50 of the 57 (87.7%) isolates at the species level . The seven strains that the LiPA test failed to identify correctly included the three M. chelonae strains that remained PCR negative, two M. intracellulare strains, and two M. scrofulaceum strains. M. intracellulare is detected by positive hybridization to two probes, MAIS and MIN. Of the two M. intracellulare strains, one reacted with the MAIS probe alone and the other with the MIN probe alone. The M. scrofulaceum strains are detected by positive hybridization to two probes, MAIS and MSC. Of the two M. scrofulaceum strains, one reacted with the MAIS probe alone and the other with the Mycobacterium genus (MYC) probe only. Interestingly, the two M. intracellulare strains that the LiPA assay failed to identify correctly could not be identified by the GenoType assay either. GenoType assay. : The GenoType assay correctly identified 77 of the 81 (95.1%) strains when compared to the reference methods . One M. intracellulare strain remained PCR negative. In addition, two M. intracellulare isolates and one M. avium isolate were not correctly identified. In GenoType strips, M. intracellulare is detected by positive hybridization to probes 9 and 11. Of the two M. intracellulare strains, one reacted with probe 9 alone and the other with probe 11 alone. M. avium is identified by positive hybridization to probes 3 and 11. The incorrectly identified M. avium strain reacted only with probe 11. Representative examples of the strips are shown in Fig. . FIG. 1. | Examples of the results of the LiPA (A) and GenoType (B) line probe assays. Examples of the results of the LiPA (A) and GenoType (B) line probe assays. The results for the same isolate are shown for each of the species. The position of the probes is shown on the right. Lanes: 1, M. scrofulaceum; 2, M. malmoense; 3, M. avium; 4, M. chelonae; 5, M. tuberculosis; 6, M. kansasii; 7, M. intracellulare; 8, M. xenopi; 9, M. gordonae. M. kansasii and M. chelonae subtypes. : The LiPA assay further differentiates M. kansasii strains to subgroups I, II, and III-V Mycobacterium gastri. Of the six M. kansasii strains studied, three belonged to subgroup II, two to subgroup III-V, and one to subgroup I. The LiPA test also differentiates M. chelonae strains to subgroups with three MCH probes, namely MCH-1 (groups I, II, III, and IV), MCH-2 (group III), and MCH-3 (group I). All five M. chelonae strains reacted with the MCH-1 probe, but not with MCH-2 or MCH-3. Thus, the strains represent group II and/or group IV. Controls. : In addition to the species-specific probes, both assays have a conjugate control line on the strip to ensure that reactive conjugate and substrate have been added. The LiPA assay also has hybridization temperature controls. During this study, LiPA strips constantly indicated that the hybridization temperature was too low by giving faint bands at probes MCH-1 and MKA-2, despite prewarmed reagents and correct temperature in the water bath. However, this did not lead to misidentification of isolates. Two novel line probe assays, LiPA and GenoType, were evaluated for identification of 81 mycobacterial isolates obtained from Finnish patients. The assays were compared to reference methods AccuProbe and 16S rDNA sequencing. The LiPA assay correctly identified 89.4% (59 of 66) of the clinical isolates within the identification range of the test, while the GenoType correctly identified 95.1% (77 of 81). Both assays were rapid, reliable, and easy to perform. The different targets may explain the differences seen in the performance of the three tests. The targets of the AccuProbe system (16S rRNA) and the GenoType assay (23S rDNA) contain conserved and variable regions, but the LiPA target (16S-23S spacer) is known to be more polymorphic (-, ). Balance between suitable variation for species differentiation and stability for successful long-term performance is crucial. In previous reports, the LiPA assay has correctly identified more than 99.4% of isolates studied . The performance of the GenoType assay has not been evaluated before. In our study, the LiPA assay correctly identified 89.4% of the strains and GenoType correctly identified 95.1% of the strains. This is the first study with LiPA PCR failing to amplify some (3 of 81) of the strains and a similar defect was found in the GenoType assay (1 of 81 strains). The difference between the results of our study and previously reported LiPA test performance probably reflects the genetic variation observed in mycobacterial subspecies isolated from different geographical areas . Although all strains were not identified at the complex or species level, as planned, no interspecies cross-reactivity was found. Although we did not evaluate the sensitivity of line probe assays, the PCR amplification step clearly makes line probe assays more sensitive than the AccuProbe test. Further, as Tortoli et al. have pointed out, the LiPA assay as well as the GenoType assay have the advantage of targeting stable DNA, whereas the AccuProbe test targets unstable rRNA and therefore requires a substantial amount not only of bacteria but also of viable organisms. LiPA strips can further differentiate M. kansasii and M. chelonae strains into subtypes. As the clinical differences of the five M. kansasii subtypes are known , this information is valuable for appropriate patient management. The clinical importance of M. chelonae subtypes is not known and, therefore, the additional information provided by the LiPA is merely of epidemiological value. The fact that LiPA PCR only amplified 62.5% (5 of 8) of M. chelonae strains included in this study shows the intraspecies heterogeneity of the 16S-23S rRNA spacer region for this species. All M. chelonae isolates were amplified and correctly identified by the GenoType assay, possibly reflecting the suitable genetic stability of the 23S rDNA. Both strip assays performed very similarly in the laboratory. The protocol takes about 5 to 6 h to complete. The cost of the GenoType test ($13/test) was slightly lower than that of LiPA ($20/test). In our laboratory, an additional limitation of the LiPA assay was the requirement for highly stringent reaction conditions, also pointed out by Tortoli et al. . When processing a large number of samples (>20), temperature does not remain optimal during the manual pipetting steps despite prewarming of reagents. This results in faint bands in hybridization temperature controls, indicating that the test is not carried out properly. The GenoType test has no temperature control, but if the hybridization temperature was too low, several bands would be seen on the strips. Since GenoType results were always easy to interpret, even with large numbers of samples, we believe that it is not as sensitive to temperature changes as the LiPA assay. In 2000, a total of 775 mycobacterial isolates were obtained from clinical specimens in Finland; 51.5% of them belonged to the M. tuberculosis complex, while 48.5% were MOTT. The LiPA test would have covered 89.3% (692 of 775) and the GenoType test 90.7% (703 of 775) of the isolates. If the current identification methods were replaced with a strip-based method, only 10% of the isolates would need to be identified by sequencing. Even though the overall difference between species coverage and performance of the two strip tests is narrow, two clinically important species, M. fortuitum and M. malmoense, are identified only with the GenoType test, rendering this assay more suitable for our laboratory. However, the AccuProbe test is still much faster and easier to perform than the strip assays. This is especially important for rapid identification of the members of the M. tuberculosis complex. In our laboratory, the strip assays would thus be best suited for the identification of MOTT. We conclude that both line probe assays were rapid, reliable, and specific, with easy-to-understand, straightforward test protocols. The GenoType test was found more suitable for the identification of mycobacteria isolated from Finland, due to its wider strain selection, less stringent reaction conditions, cost-effectiveness, and better performance. Studies are under way to evaluate this test in clinical practice. Backmatter: PMID- 12202615 TI - Disseminated Infection Due to Blastobacter denitrificans following Routine Appendectomy in an Adolescent AB - Until now, Blastobacter denitrificans has not been mentioned in the context of human infections. A case of severe complication caused by B. denitrificans after routine appendectomy in a young girl is described and confirms this organism to be an opportunistic human pathogen. Keywords: CASE REPORT : A 12-year-old girl was initially admitted to a general hospital for a routine appendectomy. After the operation, she presented with local abscesses and pneumonia accompanied by severe effusions of the pleura and pericardium . The girl is the first child of healthy parents and grew up normally. She suffered from bronchial asthma, but no immunological disorders were apparent. There are two more sisters and two twin brothers. One of her sisters became deaf following an episode of meningitis. With routine diagnostic methods, we found no evidence for bacterial, fungal, or viral infections to explain the disseminated disease after the appendectomy. Clinical specimens during her stay at the Center of Pediatrics at the University Hospital (Magdeburg, Germany) were reanalyzed for microorganisms by molecular techniques . We performed a broad-range PCR with universal primers of the eubacterial 16S rRNA gene and subsequently sequenced this gene . Only Blastobacter denitrificans could be identified in two clinical specimens, one derived from pneumonia-related effusion and the other from the tip of the central venous catheter. The analysis was performed as described previously; the alignment was 100% (BLAST search, gb/946917.1/S46917) . Thereafter, an aliquot of the pneumonia-related effusion that had been stored was plated onto agar medium prepared as described for the medium 1521 PYGV (American Type Culture Collection, Manassas, Va.;)and incubated at 30C in air. After 2 weeks of cultivation, small colonies were evident on plates. The bacteria were gram-negative rods and were positive for urease, nitrate reduction (dissimilatory), esculin hydrolysis, oxidase, and catalase; the test for indole production was negative, and acid production from d-glucose was not found. The phenotyping strongly indicated that the organism was an alpha-proteobacterium. The identity of the isolate was subsequently confirmed by sequencing of its 16S rRNA gene . Until now, the interest in B. denitrificans was in the synthesis of pharmaceutical chemicals for agriculture or in chiral synthesis of compounds in organic chemistry. There are no clinical reports about B. denitrificans as a human pathogen in a healthy or immunocompromised person. This may be due to the failure of routine diagnostic methods in its detection, since culture of this organism is difficult. B. denitrificans is a member of the alpha-2 subgroup of the Proteobacteria and shows a very close phylogenetic relationship with the genus Afipia . In addition to Bartonella henselae, the microorganism Afipia felis has been correlated with cat scratch disease . Endocarditis is well known to be a complication in Bartonella infections, and infections with Afipia occur especially in immunocompromised patients. Also, after reanalysis of the clinical specimen, there was no indication of an infection with Bartonella spp. There is only some information about the antibiotic susceptibility of this group of bacteria. For A. felis, which is a facultative intracellular bacterium and multiplies in macrophages, only aminoglycosides were found to be bactericidal . Bradyrhizobium strains showed a wide diversity with regard to intrinsic antibiotic resistance . In cases of complicated cat scratch disease caused by Bartonella henselae, treatment with trimethoprim-sulfamethoxazole, ciprofloxacin, or azithromycin is recommended, with gentamicin being reserved for severely ill patients . In the reported case, therapy lasted for about 1 month and included abdominal reoperation (cecal abscess formation) and pleural and pericardial drainages after appendectomy. Several antibiotic regimens were implemented. In this regard, the antibiotic therapy was started in the general hospital 1 week after routine appendectomy, when the signs of fever, local abscesses, and pleural empyema appeared. A combination of ceftriaxone, metronidazole, and ampicillin was applied initially and was changed to ceftazidime, gentamicin, and vancomycin 5 days later. After another 5 days, the child was transferred to the Center of Pediatrics at the University Hospital (Magdeburg, Germany). The antibiotic therapy was changed to a combination of imipenem, gentamicin, and vancomycin for a period of 2 weeks. Ultimately, the girl left the hospital in good condition. To get some information about the antibiotic resistance pattern of the isolated B. denitrificans strain, we performed agar diffusion tests on 1521 PYGV agar plates; the incubation time was 10 days. With regard to the applied antimicrobial substances, the strain was sensitive only to imipenem. This is the first report of severe complication caused by B. denitrificans in a healthy person and confirms the potential of environmental bacteria to cause human disease. FIG. 1. | The abdominal computer tomography of the patient is shown. The abdominal computer tomography of the patient is shown. On the right side of the patient (R) an organized abscess with necrotic tissue and fluid is visible. The arrow indicates the drainage on the right side; the drainage lies in the abscess ending near the intestine, which is filled with contrast medium. Backmatter: PMID- 12202554 TI - Testing Conditions for Determination of Minimum Fungicidal Concentrations of New and Established Antifungal Agents for Aspergillus spp.: NCCLS Collaborative Study AB - Standard conditions are not available for evaluating the minimum fungicidal concentrations (MFCs) of antifungal agents. This multicenter collaborative study investigated the reproducibility in three laboratories of itraconazole, posaconazole, ravuconazole, voriconazole, and amphotericin B MFCs for 15 selected isolates of Aspergillus spp. After MIC determinations for the 15 isolates in each center by the NCCLS M38-A broth microdilution method with four media, standard RPMI 1640 (RPMI), RPMI with 2% dextrose, antibiotic medium 3 (M3), and M3 with 2% dextrose, MFCs were determined for each isolate-medium-drug combination. MFCs were defined as the lowest drug dilutions that yielded <3 colonies (approximately 99 to 99.5% killing activity). The highest reproducibility (96 to 100%) was for amphotericin B MFCs with the four media. Although reproducibility was more variable and medium dependent for the azoles (91 to 98%), agreement was good to excellent for itraconazole, ravuconazole, and voriconazole MFCs with RPMI and M3 (93 to 98%). For posaconazole, the agreement was higher with M3 media (91 to 96%) than with RPMI media (91%). These data extend the refinement of testing guidelines for susceptibility testing of Aspergillus spp. and warrant consideration for introduction into future versions of the M38 document. The role of the MFC under these standardized testing conditions as a predictor of clinical outcome needs to be established in clinical trials. Keywords: Introduction : Aspergillus fumigatus and other Aspergillus spp. are responsible for the majority (85 to 90%) of clinical manifestations of severe infections caused by the filamentous fungi (moulds), especially in the immunocompromised host . The increased incidence of fungal infections and the development of new antifungal agents have underscored the importance of the laboratory's role in the selection and monitoring of antifungal therapy. The National Committee for Clinical Laboratory Standards (NCCLS) Subcommittee on Antifungal Susceptibility Tests has developed a reproducible reference testing procedure for the antifungal susceptibility testing of moulds (the M38-A document ). The recommendations described in the M38-A document for determination of MICs include the use of the standard RPMI 1640 broth (RPMI), which contains 0.2% dextrose . However, the document does not describe testing conditions for determination of minimum fungicidal (or lethal) concentrations (MFCs). Whether MICs are the best in vitro predictors of in vivo or clinical response to antifungal therapy is uncertain. Although standard conditions are not available for determination of fungicidal activities for either yeasts or moulds, it has been demonstrated that MFCs may be better predictors than MICs of therapeutic failure of amphotericin B in trichosporonosis and candidemia . The fungicidal activities of the new triazoles have also been evaluated during the last few years by nonstandardized methods . The purpose of this collaborative study was to investigate the interlaboratory reproducibility of MFCs following determinations of the MICs of five antifungal agents obtained with four media for each Aspergillus isolate (15 isolates) in three laboratories. The drugs evaluated were the conventional agents amphotericin B (Bristol-Myers Squibb Pharmaceutical Research Institute, Wallingford, Conn.) and itraconazole (Janssen Pharmaceutica, Titusville, N.J.) and three new triazoles: voriconazole (Pfizer Pharmaceuticals, New York, N.Y.), posaconazole (Schering-Plough Research Institute, Kenilworth, N.J.), and ravuconazole (Bristol-Myers Squibb Pharmaceutical Research Institute). The four media evaluated were (i) RPMI, (ii) RPMI with 2% dextrose (RPMI-2%), (iii) conventional antibiotic medium 3 (M3), and (iv) antibiotic medium 3 with 2% dextrose (M3-2%). This study was conducted simultaneously with the NCCLS multicenter (eight-center) study that further investigated optimal testing conditions for the determination of the MICs of itraconazole and three new triazoles for Aspergillus spp.; the results of that study have been reported elsewhere . MATERIALS AND METHODS : Study design. | Three laboratories participated in this study, and each laboratory received the same panel of 15 coded (isolates 1 to 15) strains of Aspergillus spp. and two control isolates. The MICs of amphotericin B, itraconazole, posaconazole, ravuconazole, and voriconazole for each isolate were obtained in each of the three centers by the broth microdilution method following a standard protocol. The standard protocol included the susceptibility testing guidelines described in the NCCLS M38-A document for MIC determination with four different media and a detailed description of the testing parameters to be evaluated for MFC determination. Following MIC determination with the four media, MFCs were determined for each isolate-drug-testing medium combination in the three centers. Isolates. | The set of isolates evaluated and the available in vitro and in vivo data are documented in Table . These 15 isolates belonged to the culture collections of the University of Manchester, Salford, United Kingdom; the University of Texas; the Medical College of Virginia; the National Cancer Institute; and the University Hospital Nijmegen, Nijmegen, The Netherlands. Each isolate was maintained as a suspension in water at approximately 25C until testing was performed. The reference isolate of A. flavus ATCC 204304 and the quality control strain Candida parapsilosis ATCC 22019 were included as controls. For C. parapsilosis ATCC 22019, there are well-established microdilution MIC ranges of the five agents evaluated in this study . Reference MIC ranges also have been established for the isolate of A. flavus ATCC 204304 based upon repeated testing in prior studies ; values for amphotericin B and itraconazole are listed in the M38-A document , and those for the new triazoles have been reported elsewhere . MIC ranges for the quality control and reference isolates were within established values . Broth microdilution method (M38-A). | Standard Excel spreadsheets were developed to serve as recording forms and to enter and analyze data. The standard protocol supplied to each laboratory for the broth microdilution method provided detailed instructions concerning the testing conditions. A single lot of each of the four liquid medium formulations was provided, ready to use, to all participants by Trek Diagnostic Systems Inc. (Westlake, Ohio). Each batch of standard RPMI and RPMI-2% was supplemented with 0.3 g of l-glutamine per liter and 0.165 M MOPS (morpholinepropanesulfonic acid) buffer (34.54 g/liter) to pH 7.0 +- 0.1 at 35C and did not contain sodium bicarbonate. Neither batch of M3 or M3-2% was buffered, and the pH of both media was 7.0 +- 0.1. The five antifungal agents amphotericin B, itraconazole, posaconazole, ravuconazole, and voriconazole were provided by the manufacturers as assay powders. Additive drug dilutions were prepared as described in the NCCLS M38-A document and in previous reports (-) in a central facility (Trek Diagnostic Systems) and shipped frozen to each participant. As recommended in the M38-A document , stock inoculum suspensions were prepared in sterile saline (provided by Trek Diagnostic Systems) containing 1% Tween 80 from 7-day-old colonies grown on potato dextrose agar slants (provided by Remel, Lenexa, Kans.). The actual stock inoculum suspensions by inoculum quantification ranged from 0.9 x 106 to 4.5 x 106 CFU/ml for 95% of the inoculum densities evaluated. Low (2.5 x 106 CFU/ml) and high (5.9 x 106 CFU/ml) densities were reported for three isolates. On the day of the test, each microdilution well containing 100 mul of the diluted (twofold) drug concentrations was inoculated with 100 mul of the diluted (two times) conidial inoculum suspensions (final volume in each well, 200 mul). The microdilution trays were incubated at 35C and examined after 21 to 26, 46 to 50, and 70 to 74 h of incubation. The MIC endpoints were the lowest drug concentrations that showed absence of growth or complete growth inhibition (100% inhibition). MFC determination. | The in vitro fungicidal activities (MFCs) were determined for each drug-isolate-medium combination as previously described . After 72 h of incubation, 20 mul was subcultured from each well that showed complete inhibition (100% or an optically clear well), from the last positive well (growth similar to that for the growth control well), and from the growth control (drug-free medium) onto Sabouraud dextrose agar plates; the contents of the wells were not agitated prior to removal of the specified volumes. This step was performed by subculturing either one (two laboratories) or more than one (one laboratory) 20-mul volume on each plate. The plates were incubated at 35C until growth was seen in the growth control subculture (usually before 48 h). The MFC was the lowest drug concentration that showed either no growth or fewer than three colonies to obtain approximately 99 to 99.5% killing activity. Data analysis. | Both on-scale and off-scale MICs and MFCs were included in the analysis. The MICs and MFCs from the three centers for each drug-isolate-medium combination tested were compared to obtain data regarding reproducibility. As previously analyzed (-), values were considered in agreement when the differences among the values were within three dilutions (three wells). A measurement of agreement was then determined as the percentage of endpoints within three dilutions (i.e., 1, 2, and 4 mug/ml) for each combination of drug, isolate, and medium. In addition, each MIC was compared to its corresponding MFC to assess the differences between the in vitro fungistatic and fungicidal endpoints obtained with each medium for each isolate in each laboratory. TABLE 1 | Selected isolates evaluated in this study RESULTS AND DISCUSSION : Reproducibility of susceptibility endpoints and detection of in vitro resistance are the main concerns when establishing standard testing guidelines for an antimicrobial susceptibility test. Although a recent NCCLS collaborative study optimized MIC testing guidelines for detection of resistance to itraconazole and perhaps to three new triazoles in Aspergillus spp., the clinical relevance of those refined parameters has yet to be determined. Because aspergillosis is a severe and commonly fatal disease, especially in the deeply immunocompromised host , an agent with fungicidal activity has therapeutic advantage over an inhibitory agent. The fungicidal activities of three new triazoles against Aspergillus spp. and other moulds have been evaluated and compared to those of amphotericin B and itraconazole . However, collaborative studies have not been conducted to detect optimal testing parameters for MFC determination. Our study represents the first interlaboratory investigation of testing conditions for determination of the MFCs of amphotericin B, itraconazole, and three new triazoles (voriconazole, posaconazole, and ravuconazole) for three species of Aspergillus. As for any other antimicrobial procedure, the issue of reliability should be addressed first. Table presents the summary of interlaboratory agreement for MFCs that were obtained with four media for the 15 isolates of Aspergillus spp. stratified by antifungal agent. Excellent reproducibility (96 to 100%) was demonstrated for amphotericin B MFCs with the four media; 48% of the values were within one dilution, and 20% had the same value. In contrast, for the triazoles, the reproducibility was more variable and dependent on the medium formulation and azole tested. Reproducibility was good to excellent (91 to 98%) with RPMI: 51 (voriconazole) to 6.2% (ravuconazole) of the results had the same value, and 61 (itraconazole) to 29.6% (posaconazole) were within one dilution. Addition of 2% dextrose to either M3 or RPMI medium did not appear to enhance reproducibility. For MIC testing, the other NCCLS study demonstrated that overall interlaboratory reproducibility was higher with RPMI than with the other media for three of the four triazoles. In both studies, differences in agreement among the media were small. The reproducibility of MFC endpoints was also evaluated by repetitive testing (at least three times in each laboratory) of the control isolate of A. flavus ATCC 204304; the MFCs of the five antifungal agents for this isolate were consistently within the expected three-well range with the four media. Table summarizes amphotericin B MIC and MFC endpoints with RPMI and M3. Our amphotericin B values with both M3 media were higher than those routinely obtained. In one of three centers (the University of Texas), ranges for the MIC and the MIC at which 50% of the isolates tested are inhibited (MIC50) of 0.5 to 8 and 1 to 2 mug/ml, respectively, have been obtained with M3 broth for 759 isolates of these three species (with MICs above 2 mug/ml only for 21 Aspergillus terreus and 3 Aspergillus flavus isolates). The reasons for this discrepancy may be the variability among M3 lots that has been reported when testing Candida spp. . As previously reported , amphotericin B MICs and MFCs were usually the same or the MFCs were no more than one to two dilutions higher than the corresponding MICs for A. fumigatus and A. flavus. Our amphotericin B MIC and MFC ranges were narrow, which is also one of the disadvantages of amphotericin B MICs for Candida spp. Although amphotericin B MICs had a relatively wide range (1 and 2 to >8 mug/ml) with M3 media for A. fumigatus, most values were >2 mug/ml or beyond safe concentrations in serum achieved with this agent. Peak levels in plasma that were below the MIC and MFC results in this study correlated with in vivo response in experimental pulmonary aspergillosis caused by isolate 8 . This response was directly related to reduced tissue burden (measured in log CFU per gram) in the lungs. Lack of correlation of in vitro results with in vivo response has been reported in a murine model of invasive A. fumigatus infection when the animals were infected with isolate 4; this isolate was recovered from a patient who responded incompletely to high-dose amphotericin B . In the present study, both media were unable to distinguish this isolate from the others. Both RPMI broths yielded two distinct amphotericin B fungicidal levels for the three A. terreus isolates , while with M3, the results were uniformly high for all isolates. Sutton et al. have demonstrated the potential discriminatory value of amphotericin B MFCs for A. terreus with M3 broth, and amphotericin B inhibitory activity was found to be superior (geometric mean MICs, 1.7 and 3.37 mug/ml) to its fungicidal activity (geometric mean MFCs, 7.4 and 13.4 mug/ml) for this species in two studies . Because M3 is widely used for testing amphotericin B, it is important that laboratory personnel be aware of the problem posed by M3 lot variability as demonstrated in this study for Aspergillus testing. In our study, amphotericin B had fungistatic activity beyond safely achievable concentrations in serum for two isolates (modal MFC, 4 mug/ml), while the MFCs (1 to 2 mug/ml) for the third isolate were within safely achievable levels in serum when tested with RPMI. Although amphotericin B had the same inhibitory activity for all three isolates (MIC range, 0.5 to 2 mug/ml), it had fungicidal activity for only one . It is noteworthy that the E-test can also yield distinctive data for these three isolates (0.5 to 1 and 4 mug/ml). Table summarizes azole MFCs and MICs that were obtained with RPMI and M3 media for the 15 isolates stratified by species and according to established azole susceptibilities . The MFC data with RPMI-2% and M3-2% were similar (data not listed in Table ). The MFCs of most triazoles were consistently higher (one to four dilutions) than the corresponding MICs for A. fumigatus and A. terreus isolates with RPMI media. In two previous studies with larger numbers of isolates, a substantial difference was found between voriconazole MICs (geometric means, 0.22 and 0.63 mug/ml) and MFCs (geometric means, 17.4 and 6.8 mug/ml) for A. terreus . Differences between MICs and MFCs for the three isolates of A. terreus were less pronounced in this study. It has been reported that voriconazole MICs and MFCs were comparable for A. fumigatus and A. flavus (MIC90s of 0.5 to 1.0 mug/ml versus MFCs at which 90% of isolates are killed [MFC90s] of 0.5 to 2 mug/ml), while itraconazole MFCs for A. fumigatus tended to be three to four dilutions higher than MICs (MIC90s of 0.5 mug/ml versus MFC90s of 4 mug/ml). The differences between fungicidal and fungistatic activities of the four triazoles against A. flavus were less substantial: 4.2 to 8.3% of MFCs were more than two dilutions higher than MICs, and 17 to 33% of MFCs for the other species were higher. Fungicidal data for posaconazole and ravuconazole are more scarce. Our results confirm previous reports in which posaconazole in vitro fungicidal and fungistatic activities appeared to be superior to those of the other triazoles. Posaconazole MICs and MFCs were higher for the two itraconazole-resistant isolates than those for most of the other isolates . In a temporarily neutropenic model of invasive aspergillosis , there was an similar100-fold difference between the fungal counts in lungs and kidneys of animals infected with isolate 5 (Tables and ) and those in animals infected with isolate 3 with a lower MIC (0.01 mug/ml). Although these results suggest that posaconazole MICs of >=0.5 mug/ml may indicate potential in vitro resistance to this agent, the MFCs for these isolates were similar. Ravuconazole MFCs with RPMI for isolates 5 and 6 indicated that that agent had similar fungicidal activities for these isolates but different fungistatic activities. The clinical significance of these discrepant results has yet to be determined. In conclusion, because standard RPMI is a chemically defined medium and MFC reproducibility is good to excellent for the five agents, RPMI appears to be a suitable testing medium for determination of fungicidal susceptibilities of Aspergillus spp. to these five agents. Although the reproducibility of M3 medium was similar to that of RPMI, its use is limited until the problem of lot-to-lot variation is investigated and resolved. The results of this NCCLS collaborative study extend the refinement of testing conditions for susceptibility testing of Aspergillus spp. to amphotericin B and four triazoles. However, the role of the MFC as a predictor of clinical outcome for Aspergillus spp. by following this standardized procedure should be established in either clinical trials or experimental infections. The introduction of these optimal testing conditions for MFC determination in a more advanced version of the NCCLS M38 document is warranted. TABLE 2 | Interlaboratory agreement of azole and amphotericin B MFCs for 15 Aspergillus isolates in three laboratories TABLE 3 | Amphotericin B MFCs and MICs for 15 Aspergillus isolates in three laboratories TABLE 4 | Azole MFCs and MICs for 15 Aspergillus isolates in three laboratories Backmatter: PMID- 12202583 TI - Practical Approach for Typing Strains of Leishmania infantum by Microsatellite Analysis AB - Currently the universally accepted standard procedure for characterizing and identifying strains of Leishmania is isoenzyme analysis. However, in the Mediterranean area, despite their very wide geographical distribution, most Leishmania infantum strains belong to zymodeme MON-1. In order to increase our understanding of polymorphism in strains of L. infantum, we developed PCR assays amplifying 10 microsatellites and sequenced PCR products. The discriminative power of microsatellite analysis was tested by using a panel of 50 L. infantum strains collected from patients and dogs from Spain, France, and Israel, including 32 strains belonging to zymodeme MON-1, 8 strains belonging to zymodemes MON-24, MON-29, MON-33, MON-34, or MON-80, and 10 untyped strains. Five of the microsatellites were polymorphic, revealing 22 genotypes, whereas the five remaining microsatellites were not variable. In particular, MON-1 strains could be separated into 13 different closely related genotypes. MON-33 and MON-34 strains also gave two additional genotypes closely related to MON-1, while MON-29, MON-24, and MON 80 strains exhibited more divergent genotypes. Among the foci examined, the Catalonian focus displayed a high polymorphism, probably reflecting isoenzyme polymorphism, while the Israeli focus exhibited a low polymorphism that could be consistent with the recent reemergence and rapid spread of canine leishmaniasis in northern and central Israel. The strains originating from the south of France and the Madrid, Spain, area displayed significant microsatellite polymorphism even though they were monomorphic by isoenzyme analysis. In conclusion, microsatellite polymorphism exhibits a high discriminative power and appears to be suitable for characterization of closely related strains of L. infantum in epidemiological studies. Keywords: Introduction : Protozoan parasites of the genus Leishmania cause a spectrum of diseases, ranging from self-limiting, self-curing cutaneous leishmaniasis (CL) to disseminating, fatal visceral leishmaniasis, and they infect various mammalian hosts. Leishmania infantum may cause either simple CL , debilitating visceral leishmaniasis, or asymptomatic cases. Currently, the universally accepted standard procedure for characterizing and identifying strains of Leishmania is isoenzyme analysis . However, this is performed only in a few laboratories and is, depending on the number of enzymes examined, very labor intensive and time consuming. Unlike some species of Leishmania, e.g., Leishmaniatropica and Leishmania major, which exhibit extensive enzymatic polymorphism , most L. infantum strains isolated in Mediterranean foci belong to the sole predominant zymodeme MON-1, despite their very wide geographical distribution. Indeed, except in some foci, where significant enzymatic polymorphism has been found among strains of L. infantum , in most Mediterranean foci, such as Provence in southern France or Israel, nearly all of the strains of L. infantum belong to the zymodeme MON-1 . In these foci, the insufficient discriminative power of isoenzyme typing methods prevents researchers from establishing correlations between clinical feature, preferential host (dogs, immunocompetent children, and human immunodeficiency virus [HIV]-infected patients), and particular group of strains. In the same way, the lack of discrimination between strains also prevents genetic studies on parasite populations. Several molecular biological typing methods have been developed to improve the discriminative power of typing methods for the genus Leishmania. These include amplification of a parasite DNA sequence by either a specific PCR or a random amplified polymorphic DNA (RAPD) PCR or detection of restriction fragment length polymorphisms (RFLPs) by Southern hybridization with DNA-specific probes . The last two methods have drawbacks. RFLP analysis is a time-consuming technique and large amounts of purified DNA are needed , whereas RAPD analysis requires strict conditions to obtain reproducibility between different laboratories and generates complex patterns . In contrast, specific PCR-based methods are attractive because of their rapidity and because culturing parasites can be avoided . However, in most cases, the level of polymorphism found with coding or repeated noncoding PCR-amplified sequences is not refined enough to distinguish between closely related strains . Microsatellite DNA sequences, tandem repeats of a simple nucleotide motif, are distributed abundantly in the eukaryotic genomes and may reveal important strain polymorphisms. However, until now only two microsatellites showing size polymorphism have been identified and characterized for L. infantum. Usually, to study microsatellites, investigators screen a genomic DNA library and then evaluate the microsatellite size polymorphism by PCR amplification and electrophoresis on acrylamide gels . In the present study, we looked for new microsatellites in the genome of Leishmania without performing genomic library screening. Microsatellites were selected from leishmanial DNA sequences published in data banks and particularly from the L. major chromosome 1 genome. The polymorphism of the microsatellite DNA sequences was evaluated by comparative analysis of the PCR product sequences of 50 strains of L. infantum collected in four Mediterranean regions of endemicity: Catalonia and Madrid, Spain; Provence, France; and northern and central Israel. MATERIALS AND METHODS : Strategy. | Potential targets for PCR amplification were selected from the leishmanial DNA sequences available in the GenBank database. Since most of them were located on chromosome 1 of L. major MHOM/IL/81/Friedlin , we used a pragmatic strategy to obtain the corresponding sequences from the genome of L. infantum. First, we selected, among sequences possessing microsatellites, those of less than 1 kb presenting coding sequences at both extremities. Second, we selected primers that recognized the coding part of the sequences and performed PCR, with DNA from L. infantum as the template. Third, we sequenced PCR products to locate the microsatellites and design new primers closer to these regions. Finally, we selected three DNA targets in the genome of L. infantum. Two of these targets corresponded to sequences originally found in the genome of L. major, one which contains three microsatellites (the Lm2 sequence) and another which contains four microsatellites (the Lm4 sequence). The third target was the internal transcribed spacer (ITS) region, which contains three additional microsatellites. Strains. | Fifty strains were used in this study . They were isolated from 48 hosts in three Mediterranean countries: Spain, France, and Israel. Forty of them had been characterized and identified previously as L. infantum by isoenzyme analysis. In two cases, two strains were isolated from separate samples collected from the same human host (BCN 143/BCN 167 and BCN 224/BCN 226) during successive episodes of leishmaniasis. Fourteen strains were collected in Provence, France. This region represents a homogeneous focus in terms of leishmanial parasites because all the strains collected from it so far have belonged to the zymodeme MON-1, including those isolated from patients coinfected with HIV. Twenty strains were isolated in Catalonia, Spain, from dogs and both immunocompromised and immunocompetent patients. The Catalonian focus is more heterogeneous in its leishmanial parasites and includes strains belonging to the predominant zymodeme MON-1 and other zymodemes (MON-24, MON-29, MON-33, MON-34, and MON-80). Additional Spanish strains, all belonging to MON-1, were collected in Madrid from dogs. Ten strains came from central and northern Israel. Canine leishmaniasis has recently reemerged in central Israel after an apparent absence of more than 40 years . All 10 strains were, essentially, isolated from dogs. One of them, strain LRC-L760*, was isolated from a colony-bred sandfly, a South American Lutzomyia longipalpis, that had fed on a local dog undergoing treatment with allopurinol during xenodiagnosis. Its sibling strain, LRC-L760, isolated from the same dog before treatment directly into rabbit blood agar semisolid medium was identified as L. infantum zymodeme MON-1 by isoenzyme analysis. The other nine Israeli strains were not characterized by isoenzyme analysis. However, 13 other Israeli strains of L. infantum from dogs and humans sent to Montpellier, France, for isoenzyme analysis were all zymodeme MON-1. In addition, the primers used in this study were tested on one strain of L. major (LPM 193) (data not shown). This strain belongs to zymodeme MON-26 and was isolated in 1999 from a patient from Saudi Arabia with CL. Isolation and growth of parasites. | The strains were isolated and grown as promastigotes on Novy-MacNeal-Nicolle biphasic culture medium. DNA preparation. | The total DNA was extracted from the promastigotes for PCR amplification. Samples of 100 mul of culture were washed in 0.3% NaCl. The pellet was lysed by heating at 96C for 20 min with 400 mul of a mixture containing 1% Tween 20 (Sigma, St. Louis, Mo.), 1% Nonidet P-40 (Sigma), and 20% Chelex resin (Bio-Rad, Hercules, Calif.) made up in sterile distilled water. The mixture was then centrifuged at 14,000 x g for 10 min at room temperature. The supernatant was collected and was either used immediately for PCR amplification or stored at -20C until used. PCR amplification. | Three PCRs were performed. The primers A2 (5'-GGGAGAAGCTCTATTGTG-3') and B1 (5'-ACACTCAGGTCTGTAAAC-3') were used for analyzing the ITS region, and primers Lm4A1 (5'-CGGTGCACATTCGACCGCTA-3') and Lm4B1 (5'-ATGGCACGGTGCACGCTTCC-3') were used for amplifying the Lm4 sequence. Two pairs of primers were used to amplify the Lm2 sequence. First, we amplified the Lm2 sequence with Lm2A1 (5'-TGACGCGACGTGGCAAGTCA-3') and Lm2B1 (5'-CCGTGAAGTACTCGGACGCT-3'), primers that recognize a 900-bp-long fragment. Then we used internal primers, Lm2A3 (5'-AAAAAGCGAGGAATGAAAGAA-3') and Lm2B3 (5'-TAGAGGCGTGGCAGAGAC-3'), which were localized on either side of the most variable microsatellite, for sequencing the complete Lm2 fragment in both directions. All the primers were synthesized by Eurogentec (Seraing, Belgium). Each reaction mixture (50 mul) contained a 0.05 mM concentration of each deoxynucleoside triphosphate, a 0.5 muM concentration of each primer, 1x REDtaq buffer, 1 U of REDtaq polymerase (Sigma), and 5 mul of total DNA extract. After a 2-min incubation at 94C, amplification of either the ITS region or the Lm2 and Lm4 sequences was performed for 40 or 35 cycles, respectively, in an automated thermocycler PTJ 100 (MJ Research, Inc, Watertown, Mass.). The conditions used for the ITS region were denaturation at 94C for 20 s, primer annealing at 53C for 30 s, and extension at 72C for 1 min. The conditions used for the other sequences were denaturation for 30 s at 94C, primer annealing for 30 s at 63C, and extension for 1 min at 72C for the Lm4 sequence; denaturation for 30 s at 94C, primer annealing for 30 s at 62C, and extension for 90 s at 72C for the Lm2 sequence with the Lm2A1 and Lm2B1 primers; and denaturation for 30 s at 94C, primer annealing for 30 s at 51C, and extension for 1 min at 72C for the Lm2 sequence with the Lm2A3 and Lm2B3 primers. After amplification, 10 mul of PCR products was visualized by electrophoresis on a 1.5% agarose gel containing ethidium bromide. DNA sequencing. | Amplified DNA was purified by using the QIAquick PCR purification kit (Qiagen SA, Courtaboeuf, France) according to the manufacturer's recommendations. The purified DNA for each strain was then sequenced in both directions with a fluorescent labeling kit (Thermo Sequenase II kit; Amersham Pharmacia Biotech Europe, Orsay, France) with specific primers on an automated DNA sequencer 373A (Perkin Elmer Applied Biosystems, Foster City, Calif.). In some cases, the results obtained by direct sequencing of Lm2 PCR products were confirmed after cloning the PCR products into the vector pCR 2.1 (TA Cloning kit; Invitrogen, Leek, The Netherlands). Phenetic dendrogram. | The ITS, Lm2, and Lm4 sequences obtained for each different genotype were aligned, and a dendrogram was calculated with the fastDNAml software (version 1.2.2) . TREEVIEW , a tree drawing software, was used to produce the tree. Reproducibility of PCR and stability of the genotypes. | Reproducibility was assessed by repeating the preparation of the DNA from the two strains UCM 1 and LPM 191 six times followed by amplification and sequencing of all of the samples. In addition, the primers A2 and B1 were used repetitively, six times, to amplify the ITS sequence from three DNA extracts of the strains BCN 1, BCN 143, and BCN 226. Similarly, six amplifications were performed with DNA isolated from the five strains BCN 1, BCN 143, BCN 235, BCN 258, and LPM 191 with the primers Lm4A1 and Lm4B1. We also verified the stability of the strains in vitro and in vivo. After culturing the promastigotes of strains BCN 1, BCN 224, and LPM 195 in six different batches of Novy-MacNeal-Nicolle medium over 3 months, their DNAs were isolated and PCRs were performed for the three molecular targets. In parallel, total DNA was extracted at each stage from the leishmanial promastigotes of strain BCN 83 grown directly after isolation from a dog, regrown after passage through and reisolation from a hamster, and then regrown after passage through and reisolation from another dog. All of these DNA extracts were submitted to PCR amplification and sequencing. TABLE 1 | Main features of strains of L. infantum collected in three Mediterranean foci RESULTS : Sequence analysis of the PCR-amplified ITS region with primers A2 and B1. | All strains gave an amplified fragment of about 650 bp. The A2B1 part of the ITS sequence contained three microsatellites exhibiting a moderate degree of size polymorphism: the number of AT repeat units was 5 or 6, the cluster TA was repeated 4 or 5 times, and the number of G repeat units was 5, 7, or 8 . Three different sequences were found among the strains from Catalonia, whereas all strains from Madrid, Provence, and Israel had the same sequence . Sequence analysis of the Lm2 and Lm4 fragments. | Amplification of the Lm2 region generated an amplified fragment of about 900 bp for L. infantum and one of about 750 bp for L. major. The Lm2 sequence exhibited four microsatellites. Three of them corresponding to a C7 microsatellite, a (CA)8 microsatellite, and an imperfect (TG)8-CG-(TG)5 microsatellite were identical among all strains. The last one, corresponding to a poly-TG, was highly variable; the dinucleotide was repeated 17, 22, 23, 24, 25, 26, or 27 times . No differences were found between sequences obtained directly after PCR or after cloning of the PCR products. Finally, sequence analysis of the Lm2 PCR products in all strains identified seven profiles only due to the size variation of the poly-TG . Six of them were found for the strains originating from Catalonia. Three different Lm2 sequences were identified among strains collected in Provence and Israel. Amplification with primers Lm4A1 and Lm4B1 produced a fragment of 560 bp for each species, L. infantum and L. major. The Lm4 region contained three microsatellites: poly-A, poly-T, and poly-TA. Two of them were identical among all strains (A10 and T9), while the (TA)n microsatellite exhibited a high level of size polymorphism. The TA unit was repeated 9, 10, 11, 12, 13, or 14 times . Finally, sequence analysis of the Lm4 PCR products in all strains identified six profiles due only to the size variation of the poly-TA . Strains from Israel had the same Lm4 sequence while strains from Madrid, Catalonia, and Provence gave three, four, and five different sequences, respectively. Reproducibility of PCR and stability of the genotypes. | The ITS, Lm2, and Lm4 sequences obtained for each strain were identical when six different preparations of the same DNA were used for the PCRs or when the fragment was amplified six times by using the same DNA. In addition, the ITS, Lm2, and Lm4 sequences showed no variations even after multiple generations in vivo and in vitro. Results comparison. | The ITS-, Lm2-, and Lm4-associated sequences generated 22 genotypes for 50 strains of L. infantum belonging to six zymodemes . In particular, the MON-1 strains gave thirteen different genotypes that were closely related to each other . In addition, DNA typing of strains MON-33 and MON-34 gave two different genotypes that were closely related to the MON-1 genotype. On the other hand, strains MON-29, MON-24, and MON-80 showed more-divergent genotypes. Even when more than one strain was isolated from the same patient, the strains still had identical genotypes. Strains collected in the Provence region, which all belong to the zymodeme MON-1, gave nine different genotypes, while the 10 L. infantum strains originating from northern and central Israel resulted in only three genotypes. FIG. 1. | Complete nucleotide sequences of genomic DNA regions containing microsatellites from strains of L. infantum Complete nucleotide sequences of genomic DNA regions containing microsatellites from strains of L. infantum. Microsatellites are shown in bold type. The superscripts indicate the minimum and maximum numbers of repeat units found for each microsatellite. (A) Nucleotide sequence of PCR product determined with primers A2 and B1. (B) Nucleotide sequence of PCR product determined first with primers Lm2A1 and Lm2B1 and then with primers Lm2A3 and Lm2B3. The internal primers, Lm2A3 and Lm2B3, are underlined. (C) Nucleotide sequence of PCR product determined with primers Lm4A1 and Lm4B1. FIG. 2. | Dendrogram calculated with fastDNAml (version 1. Dendrogram calculated with fastDNAml (version 1.2.2), which is based in part on Felsenstein's nucleic acid sequence maximum likelihood method (version 3.3), and elaborated with TREEVIEW tree drawing software. Each genotype is characterized by its microsatellite combination (ITS-Lm2-Lm4), its zymodeme, and its geographical origin. C, Catalonia; M, Madrid; P, Provence; I, Israel; #, the zymodeme is not available for all of the samples of the genotype . TABLE 2 | Comparison of isoenzyme analysis and DNA typing by microsatellite size polymorphism with L. infantum DISCUSSION : We have developed a reproducible, discriminating molecular typing system that was able to characterize and distinguish closely related strains of L. infantum collected in three Mediterranean countries: France, Spain, and Israel. It is based on the comparative analysis of PCR products from three nuclear noncoding DNA sequences containing polymorphic microsatellites. The partial screening of a leishmanial genomic library was not needed to identify these microsatellites. Instead, polymorphic targets were selected from among complete DNA sequences published in data banks . Since many such sequences are available in the gene data banks, this strategy can be applied to the molecular typing of other eukaryotic microorganisms. One major advantage of this approach is its simplification of Leishmania strain typing in the Mediterranean region. By employing parasite-specific primers to polymorphic microsatellites, parasite culturing may be avoided in many cases. PCR could be carried out directly with infected host tissue, since nonspecific amplification of human or dog DNA with these primers was not observed (data not shown). In addition, this approach for typing strains of Leishmania is rapid. It takes just a few days to perform DNA extraction, amplification by PCR, and sequencing, whereas isoenzyme analysis, which is still the "gold standard," may take several months from strain isolation through culture to provide a full enzyme profile and zymodeme designation. The usefulness of isoenzyme analysis for epidemiological and taxonomic studies has been widely proven, but microsatellite analysis seems to be more discriminating with the 15 different genotypes found for MON-1 in our study. This genetic heterogeneity of strains belonging to one zymodeme has been observed for other approaches based on DNA microheterogeniety, such as RAPD and RFLP PCR methods . The new typing approach presented here is very reproducible; repeat testing of the same isolate after multiple generations in vivo and in vitro gave identical sequences. However, in some cases, sequences of very long microsatellites, such as the poly-TG found in the Lm2 sequence, were quite difficult to read, and ambiguities were observed at the end of the microsatellite. These ambiguities could be resolved by sequencing the reverse DNA strand. Since sequences obtained after PCR cloning were strictly identical, we think that this problem does not reflect a sequence polymorphism within the population of Leishmania in the sample. This problem is more likely due to DNA polymerase misreading during PCR amplification. Among the three sequences (ITS, Lm2, and Lm4) analyzed, only the microsatellite region was variable and size polymorphism was the only variation noted in the microsatellites. No exchanges of bases were seen in or around microsatellite clusters. This is probably due to the fact that all strains of the panel are closely related. As microsatellites exhibit only size variation and sequencing of very long microsatellites is sometimes difficult to perform, it should be possible to improve this technique further. Indeed, size polymorphism of microsatellites could be evaluated by acrylamide gel electrophoresis of fluorescein-labeled PCR products . This would save time and expense because DNA sequencing of the PCR products could be avoided. The microsatellites analyzed here displayed variable levels of polymorphism that may be related to the number of repeat units and/or the composition of the microsatellite. To verify this hypothesis, analysis of more microsatellites would clearly be useful. These data could be used to select more polymorphic microsatellites and increase the discriminatory power of such typing methods. In addition, it is important to note that our results are not in contradiction with the isoenzyme typing data . Between the different foci of leishmaniasis analyzed, the polymorphism observed with the five microsatellites varied. The Catalonian focus displayed the highest genomic polymorphism and probably reflects the isoenzyme polymorphism observed in that region. The Israeli focus displayed less genomic polymorphism among its strains, perhaps correlated with the recent reemergence and rapid spread of canine leishmaniasis caused by L. infantum in northern and central Israel . The strains originating from the south of France displayed significant microsatellite size polymorphism even though they were monomorphic by enzyme analysis. Concerning the Madrid focus, its relative stability (1 zymodeme and 4 closely related genotypes) seems to be correlated with the smaller area in which the strains were obtained. To date, we have only studied strains of L. infantum isolated from humans and dogs. However, it should be possible to use microsatellite variation to study the genetic diversity of leishmanial parasites from all of the possible different situations, i.e., different clinical presentations, human (immunocompromised or immunocompetent children) and animal (dogs and wild canids) hosts, and sandfly vectors existing in the same area of endemicity. Microsatellites might also prove useful in determining whether or not there is any relationship between genotype and the pathogenicity of Leishmania strains. This technique could be useful in discerning whether specific leishmanial genotypes confer host specificity and if leishmanial parasite diversity is consistent in different hosts. In conclusion, microsatellite polymorphism is simply and rapidly detected by PCR, displays a high level of reproducibility, and exhibits a high level of discrimination. It is suitable for characterizing closely related strains of L. infantum. Since many sequences are published in existing data banks, the strategy used here could be applied to the molecular typing of other eukaryotic microorganisms. Backmatter: PMID- 12202548 TI - Detection of Simian Immunodeficiency Virus in Diverse Species and of Human Immunodeficiency Virus Type 2 by Using Consensus Primers within the pol Region AB - Human immunodeficiency virus type 2 (HIV-2) is the result of cross-species transmission of simian immunodeficiency virus (SIV) from sooty mangabey monkeys to humans. Primer pairs (intHIV-2/SIV) based on a region of integrase that has considerable homology across HIV-2 and SIV lineages were designed to develop a broadly cross-reactive molecular assay to detect lentivirus infection in primates. The intHIV-2/SIV primers detect HIV-2 and simian viruses SIVcpz, SIVsmm, SIVsyk, SIVagm, and SIVmnd. The primers are also capable of amplifying some HIV-1 strains. Additionally, sequences from the integrase amplicons were of sufficient genetic diversity to permit not only phylogenetic clustering of all simian viruses to their respective lineages but also HIV type and group classification. Thus, the primers described here provide a method to detect primate lentiviruses from diverse species of nonhuman primates, as well as from persons infected with HIV-1 and HIV-2. Keywords: Introduction : Zoonotic transmission is an important factor in the emergence of retroviruses and other infectious agents in humans . At least 20 different nonhuman primate species in Africa have been shown to harbor simian immunodeficiency viruses (SIVs) . Thus, African primates represent an extremely large reservoir of lentiviruses that can potentially infect other species including humans . Indeed, the extensive phylogenetic relatedness among many strains of human immunodeficiency virus type 1 (HIV-1) and HIV-2 and primate lentiviruses has elucidated the simian origin of AIDS . These studies have further established that HIV-1 infection, the main cause of the worldwide AIDS pandemic , was the result of cross-species transmission of SIV from chimpanzees (Pan troglodytes troglodytes) to humans (, , -). Likewise, the HIV-2 epidemic seems to have emerged via cross-species transmission of SIV from sooty mangabey monkeys (Cercocebus sp.) . Current evidence indicates that the SIV counterparts of HIV-1 and HIV-2 were introduced into the human population multiple times (at least seven transmission events have been suggested) . Yet, the HIV-1 group M viruses appear to have arisen from just one such zoonotic transmission . Thus, while cross-species transmission of primate retroviruses to humans occurs relatively frequently, the subsequent spread of the retroviruses in the human population is rare . Nevertheless, this scenario represents a unique opportunity to study both the emergence of new human retroviruses and the genetic diversities of these human and simian viruses. The primate lentiviruses for which full-length genomic sequences are available fall into five major equidistant phylogenetic lineages: (i) SIVcpz from chimpanzees, together with HIV-1; (ii) SIVsmm from sooty mangabey monkeys together with HIV-2; (iii) SIVagm from four species of African green monkeys; (iv) SIVsyk from Sykes monkeys; and (v) SIVmnd and SIVlhoest from mandrills and l'Hoest monkeys, respectively (, , , , -, ). Recently, an additional SIV (SIVcol) from Guereza colobus monkeys (Colobus guereza), representing a sixth lineage of primate lentivirus, has been identified . Therefore, studies are needed to determine whether transmission of simian lentiviruses other than SIVcpz and SIVsmm to humans is occurring, particularly in regions where SIV infection in nonhuman primates is highly prevalent. These studies require the development of molecular detection assays that can detect a wide range of lentivirus infections in both nonhuman primates and humans. We have recently described serologic and molecular diagnostic assays that permit detection of highly divergent HIV-1 strains and their simian counterpart, SIVcpz . Here we report a sensitive and broadly reactive PCR-based molecular screening tool for detecting HIV-2 and SIVs from diverse species of primates. More importantly, sequences generated from these PCR products can be reliably used for phylogenetic classification within the HIV-2 and SIV lineages, thus permitting identification of the zoonotic source of infection. MATERIALS AND METHODS : Primer design. | Primers were designed based on the consensus sequences in the pol region of HIV-2 and SIV and are designated intHIV-2/SIV primers. For reverse transcription (RT) and primary PCR, the primers were INT-F1 (forward; 5'-ATAGAACCAGCACAAGAAGAACAT; nucleotides 2241 to 2264 based on HIV-2 ROD) and INT-R1 (reverse; 5'-ACTGCTCCTTCACCTTTCCA; nucleotides 2956 to 2975). The primers for the nested PCR were INT-F2 (forward; 5'-AATGTCAACAGAAAGGAGAAGCTATACAT; nucleotides 2354 to 2383) and INT-R2 (reverse; 5'-CCCCTATTCCTCCCTTCTTTTAAAAT; nucleotides 2780 to 2806). Reference clones. | The sensitivity of the intHIV-2/SIV primers was tested by using known copy numbers of cloned material representing HIV-1, HIV-2, and SIVcpz. For HIV-2, a cloned fragment of a Centers for Disease Control and Prevention HIV-2 isolate (GB122) was generated by directly cloning an integrase PCR fragment into the pCR-TOPO vector in accordance with the manufacturer's protocol (Invitrogen, Carlsbad, Calif.). The following previously described molecular clones were used for SIVcpz and HIV-1 : 92UG037.1 (subtype A), 93TH233.3 (CRF01-AE), 92RW009.6 (subtype A/C), 92NG083.2 (CRF02-AG), BCSG3 (subtype B), 93BR029.4 (subtype B/F), 92BR026.8 (subtype C), 94UG114.1 (subtype D), 93 BR020.1 (subtype F), and 90CR056.1 (subtype H). Known copy numbers of the partial GB122 clone, the infectious SIVcpz clone, and the HIV-1 molecular clones were used for sensitivity determination. For all samples, multiple dilutions were run in duplicate and each sample was subjected to PCR amplification at least twice. Viral isolates and plasma specimens. | HIV-1, HIV-2, and SIV isolates were generated by previously described cocultivation procedures . DNA or RNA or both from cultured material were used to test these primers on a wide variety of primate lentivirus isolates. The SIV specimens included 7 SIVsm (sooty mangabey) specimens, 1 SIVstm (stumptail macaque) specimen, 1 SIVrcm (red-capped mangabey) specimen, 4 SIVagm (African green monkey) specimens, 1 SIVmnd (mandrill) specimen, 1 SIVsyk specimen, 1 SIVcpz specimen, and SIVhu (an isolate derived from a human accidentally infected with SIVB670) . Additionally, 22 HIV-2 specimens, including uncultured peripheral blood lymphocytes from 10 HIV-2-infected persons and 12 previously cultured isolates, were included in the study . A plasma panel from known HIV-1-seropositive individuals (Boston Biomedical Inc., Boston, Mass.), including 1 specimen each from Argentina, Canada, China, and Mozambique, 2 from the United States, 3 from Thailand, 5 from Zimbabawe, 6 each from Uganda and Ivory Coast, and 8 from Ghana, and an HIV-2-seropositive panel (Boston Biomedical Inc.), including 14 specimens from Ivory Coast, were also used for viral RNA detection by RT-PCR analysis using the intHIV-SIV primers. PCR. | DNA preparation was done by proteinase K digestion in a Tris-Triton buffer, and RNA was extracted by using the QIAamp viral RNA kit according to the manufacturer's protocol (Qiagen Inc., Valencia, Calif.). Protocols for RNA extraction and conditions for RT-PCR and PCR are described elsewhere , except that an annealing temperature of 50C was used for both the primary and secondary intHIV-2/SIV primer sets. All reactions were carried out with appropriate negative controls to detect possible contamination. Sequence and phylogenetic analysis. | Selected nested-PCR products were purified with the QIAquick PCR purification kit (Qiagen) and sequenced with Big-dye terminators (Perkin-Elmer, Foster City, Calif.) on an automated 377 DNA sequencer (Applied Biosystems, Foster City, Calif.). Sequences were aligned with CLUSTAL W (version 1.74) after editing, and phylogenetic trees were constructed by the neighbor-joining method using the PHYLIP, version 3.5c, package. Nucleotide sequence accession numbers. | The GenBank accession numbers for the sequences obtained in this study are to . RESULTS AND DISCUSSION : Analysis of a cloned fragment from HIV-2 (GB122) revealed that the intHIV-2/SIV primers reliably amplified as few as 25 copies of HIV-2 per PCR . Moreover, these primers also detected SIVcpz with the same sensitivity . We next examined the cross-reactivity of these primers with HIV-1. Amplifications were performed with HIV-1 group M molecular clones with inputs of 100, 50, and 25 copies per PCR. All of the HIV-1 group M subtypes tested with the exception of subtypes C and D were amplified at 25 copies per reaction ; subtypes C and D could only be amplified at 100 copies per reaction. Thus, HIV-2, SIVcpz, and all HIV-1 subtype molecular clones were consistently amplified at 100 copies per reaction. In addition, RNA from the culture supernatant of an HIV-1 group N virus was amplified . We next tested the amplification efficiency of the primers for 12 HIV-2 primary isolates. The intHIV-2/SIV primers amplified all samples, which included nine HIV-2 subtype A specimens, two subtype B specimens, and one subtype A/B recombinant specimen . As cultured isolates from most of the specimens were not available, we performed PCR amplification on DNA derived from uncultured peripheral blood mononuclear cells (PBMC). Analysis of primary PBMC from 10 HIV-2-infected individuals revealed that DNA of those from 8 was amplified by using these primers , giving an amplification efficiency of 80%. These results are comparable to those for the primers identified in the protease region, which are highly specific for HIV-2 detection only . The high sensitivity of HIV-2 DNA detection from primary PBMC led us to test the amplification efficiency of intHIV-2/SIV for detection of plasma viremia from HIV-2- and HIV-1-infected individuals. Results from duplicate experiments showed that the intHIV-2/SIV primers were able to amplify from RNA 5 of the 14 HIV-2 plasma samples tested (35.7%) . Amplification with RNA from the same samples with protease primers, previously shown to amplify as few as 10 cDNA copies from PBMC , resulted in only 43% amplification efficiency (data not shown). Further, amplification of RNA from 34 HIV-1 plasma samples by using intHIV-2/SIV primers resulted in amplification of RNA from 19 (56%) samples. In comparison intM-Z primers, previously shown to amplify HIV-1 and SIVcpz efficiently , amplified RNA from all specimens (100%) . Previous studies have developed HIV-2-specific primers that allow detection of HIV-2 from DNA in persons dually infected with HIV-1 and HIV-2 . More recently, qualitative and quantitative PCR-based assays for specific detection of HIV-2 have been developed . The intHIV-2/SIV primers described here were designed to be as cross-reactive as possible, so that any new variants of HIV-2 or HIV-1 or cross-species transmission of SIVs from nonhuman primates to humans could be detected. The intHIV-2/SIV primers detected 80% (10 of 12) of the PBMC from HIV-2-infected persons but detected viremia in only 36% (5 of 14) of plasma samples. While the sensitivity of amplification from primary DNA was good, one of the limitations of the intHIV-2/SIV primers is that they are not sensitive for detection of HIV-2 plasma viremia. However, using previously characterized HIV-2 protease primers resulted in comparable efficiency of amplification of HIV-2 from viral RNA . The lower sensitivity of HIV-2 detection from plasma may be due to low copy numbers of HIV-2, and indeed several studies have established that HIV-2-infected persons have significantly lower viremia than those infected with HIV-1 . Because of the highly conserved nature of the integrase gene between HIV-2 and SIVs, it was predicted that the intHIV-2/SIV primers could amplify SIVs from diverse primate species. We next verified this hypothesis by using viral DNA or RNA from culture material of diverse SIVs from sooty mangabeys (SIVsmm3, -9, -21, -54, -55, -74, and -156), red-capped mangabeys (SIVrcm), stumptail macaques (SIVstm), mandrills (SIVmndBK-12), Sykes monkeys (SIVsyk), and African green monkeys (AgmKenya, AgmTYO1, AgmTAN1, and Agm1584) . In addition, the primers amplified an SIVhu and SIV from chimpanzees (SIVcpz), the species that has been shown to be the natural reservoir for HIV-1 . These data suggest that intHIV-2/SIV primers are sensitive for detecting DNA and RNA from HIV-2, as well as diverse lineages of SIVs. Since the primers had broad specificity for detection of HIV-2, HIV-1, and SIVs, we next examined if phylogenetic analysis of amplified sequences would permit distinction of HIV-2, HIV-1, and SIV lineages. The products amplified by using intHIV-2/SIV from cultured isolates of HIV-2 and SIVs, plasma of HIV-2-infected persons, and selected HIV-1s were sequenced, and a 420-bp segment was used for phylogenetic analysis. Figure depicts the relationship of HIV-2 and SIV sequences derived from the amplified integrase products. The overall topology of the tree remains the same as those previously demonstrated for full-length integrase sequences. HIV-2 and several SIVs amplified and sequenced from this study revealed distinct cluster patterns similar to those of the previously characterized reference clusters. As expected, HIV-2 and SIV from sooty mangabeys formed a monophyletic cluster. Sequences from SIVrcm, SIVmnd, SIVagm, and SIVcpz were highly divergent and represented distinct lineages . Similar phylogenetic analysis of sequences amplified from persons infected with HIV-1 group M or group O revealed clustering with their respective group M or group O sequences , although the subtype designation within the group M sequences was not reliable (data not shown). Thus, despite the small fragment size, the phylogenetic analysis of this region provides an adequate clustering pattern to identify the correct lineages of SIVs, as well as the HIV type and group. The intHIV-2/SIV primers were able to amplify specimens from five major lineages of primate lentiviruses. Additionally, HIV-1 and HIV-2 sequences were correctly identified. However, the integrase region was not sufficient to provide clade designation for either HIV-1 or HIV-2. In summary, the primers described here provide a quick method to amplify primate lentivirus genomic sequences from diverse species of primates, as well as PBMC from HIV-1- and HIV-2-infected persons. This assay should provide a useful tool for recognizing a wide range of lentivirus infections from nonhuman primates and humans, both for identification of HIV variants worldwide and for early detection of potential cross-species transmission. FIG. 1. | Detection of HIV-2, SIVcpz, and diverse HIV-1 molecular clones by using intHIV-2/SIV primers. Detection of HIV-2, SIVcpz, and diverse HIV-1 molecular clones by using intHIV-2/SIV primers. Known copy numbers (100, 50, and 25 copies per PCR) of reference clones of HIV-2 (GB122), SIVcpz, and HIV-1 subtypes A (92UG037), B (BCSG3), C (92BR026), D (94UG114), F (93BR020), H (90CR056), AC (92RW009), CRF01-AE (93TH233), CRF02-AG (92NG083), and B/F (93BR029) were amplified by using intHIV-2/SIV primers. FIG. 2. | Phylogenetic positions of HIV-2 sequences from this study (boldface and underlined) and reference HIV-2 and SIV sequences in the integrase region. Phylogenetic positions of HIV-2 sequences from this study (boldface and underlined) and reference HIV-2 and SIV sequences in the integrase region. Trees were derived from nucleotide sequence alignment of 413 bp by using the neighbor-joining method. Numbers at the nodes represent the percentages of bootstrap values (only values >70 are shown). FIG. 3. | Phylogenetic position of HIV-1 sequences from this study (boldface and underlined) and previously reported HIV-1 and SIVcpz strains in integrase regions. Phylogenetic position of HIV-1 sequences from this study (boldface and underlined) and previously reported HIV-1 and SIVcpz strains in integrase regions. The topology shows an overall branching order consistent with previously reported phylogenies for full-length sequences. Trees were derived from nucleotide sequence alignments (consensus lengths of 412 bp) using the neighbor-joining method. Horizontal branch lengths are drawn to scale, with the bar indicating 0.10 nucleotide substitution per site. Numbers at the nodes indicate the percentages of bootstrap values (out of 500) in which the cluster to the right is supported (only values 70% or above are shown). TABLE 1 | Amplification of HIV-2 and HIV-1 isolates with intHIV-2/SIV primers TABLE 2 | Comparative sensitivities of intHIV-2/SIV and intM-Z primers for PBMC DNA and plasma RNA detection TABLE 3 | Detection of diverse SIV isolates with intHIV-2/SIV primers Backmatter: PMID- 12202575 TI - Detection and Quantification of Oral Treponemes in Subgingival Plaque by Real-Time PCR AB - Oral treponemes have been associated with periodontal diseases. We developed a highly sensitive and specific method to detect and quantify cultivable oral treponemes (Treponema denticola, Treponema vincentii, and Treponema medium) in 50 subgingival plaque samples from 13 healthy subjects as well as 37 patients with periodontal diseases using real-time PCR assays with specific primers and a TaqMan probe for each 16S rRNA sequence. The specificity for each assay was examined by using DNA specimens from various treponemal and other bacterial species. The TaqMan real-time PCR was able to detect from 103 to 108 cells of the oral treponemes, with correlation coefficients as follows: T. denticola, 0.984; T. vincentii, 0.991; and T. medium, 0.984. The frequencies of occurrence of these three oral treponemes in subgingival plaque samples were as follows: T. denticola, 68.0%; T. vincentii, 36.0%; and T. medium, 48.0%. In addition, the number of T. denticola, T. vincentii, and T. medium cells in plaque samples detected by real-time PCR ranged from 3 to 15,184, 1 to 447, and 1 to 7,301 cells/pg of plaque DNA, respectively. Increased numbers of T. denticola cells were detected in plaque samples from deep periodontal pockets, and T. medium was also detected in deep pockets. On the other hand, T. vincentii was mainly found in shallow pockets. These results suggest that various oral treponemes are associated with the formation of each stage of periodontal disease. Keywords: Introduction : Periodontal disease is clinically observed as an inflammatory condition of the tooth supporting structures that leads to a progressive degradation of periodontal tissues and then tooth loss . The oral flora found in patients with chronic periodontitis consists of a variety of oral bacterial species. Among these organisms, oral treponemes, which are gram-negative, anaerobic, motile, and helical rods, have been shown to be closely associated with various types of periodontal diseases such as gingivitis, acute necrotizing ulcerative gingivitis, and adult periodontitis . Human immunodeficiency virus-positive subjects with gingivitis and adult periodontitis also have elevated numbers of oral treponemes in their subgingival plaque . Oral treponemes have been observed to adhere to and invade connective tissues, as well as gingival epithelial cells , and also induce cytokine production from human gingival fibroblasts . Therefore, it is suggested that they play an important role in causing periodontal diseases. Real-time PCR assays have recently detected a variety of microorganisms including clinical bacteria and viruses . In addition, a real-time PCR analysis method for the detection of Toxoplasma gondii, which causes one of the most prevalent parasitic diseases, toxoplasmosis, has been shown to be effective . Real-time PCR with the TaqMan system allows continuous measurement of products throughout the reaction in a closed tube and exploits the 5' to 3' exonuclease activity of Taq polymerase in conjunction with fluorogenic DNA probes . In this method, a TaqMan probe, designed to hybridize to the target PCR product, is labeled with a fluorescent reporter dye and a quencher dye. During PCR amplification, the probe is digested by Taq polymerase, separating the dyes, resulting in an accumulation of reporter fluorescence along with a corresponding increase in fluorescence intensity. In the present study, we used species-specific PCR methods to identify the oral treponemes Treponema denticola, Treponema vincentii, and Treponema medium in human subgingival plaque samples. Additionally, we quantified the numbers of each organism using real-time PCR (TaqMan) assays. MATERIALS AND METHODS : Bacterial strains. | T. denticola strain ATCC 35404, T. vincentii ATCC 35580, and T. medium ATCC 700293 were grown anaerobically in trypticase-yeast extract-gelatine-volatile fatty acid-rabbit serum broth containing 5% rabbit serum at 37C for 72 h . The morphologies and motilities of the cultures were observed by dark-field microscopic observation. For analysis, each culture was centrifuged at 1,500 x g for 20 min and then washed twice with phosphate-buffered saline (PBS; Sigma, St. Louis, Mo.). The cells were resuspended in PBS, and bacterial cell counts were estimated by phase-contrast microscopy and with a Petroff-Hausser bacterial counter (Hausser and Son, Philadelphia, Pa.). A 10-fold aliquot of each cell suspension was prepared and serially diluted with PBS. The following reference strains of bacteria were used in this study: Treponema socranskii subsp. buccale ATCC 35534, Treponema phagedenis biovar Kazan ATCC 27087, Treponema pectinovorum ATCC 33768, Porphyromonas gingivalis 381, Prevotella nigrescens ATCC 33563, Actinobacillus actinomycetemcomitans SUNY 7185, Escherichia coli ATCC 25922, Fusobacterium nucleatum subsp. polymorphum ATCC 10953, Streptococcus mutans GS5, Streptococcus oralis ATCC 10557, and Streptococcus salivarius ATCC 9222. Plaque samples. | Thirteen healthy subjects and 37 patients of Asahi University Hospital with periodontal disease (28 males aged 18 to 81 years [mean age, 55.0 +- 16.8 years] and 22 females aged 15 to 73 years [mean age, 42.1 +- 19.1]) were studied. Clinical diagnosis was made as described by Paster et al. . Briefly, the healthy subjects (5 males aged 18 to 37 years [mean age, 27.6 +- 9.0 years] and 8 females aged 15 to 72 years [mean age, 32.6 +- 21.7 years]) had no pocket depths greater than 3 mm and no attachment loss greater than 2 mm at any site in the mouth, while those with periodontal disease (23 males aged 34 to 81 years [mean age, 61.0 +- 11.2 years] and 14 females aged 20 to 73 years [mean age, 47.6 +- 15.8 years]) had at least 20 teeth, at least eight sites with a pocket depth of at least 4 mm, and eight sites with attachment loss of at least 3 mm. The patients had not received professional cleaning or antibiotic medication within 3 months. All were informed of the study, and each signed the informed consent form approved by the Ethics Committee of Asahi University Hospital. Prior to subgingival plaque sampling, each tooth was isolated with cotton rolls and air dried, and then the pocket depth was measured by probing. Supragingival plaque was first removed with sterile cotton, and then two absorbent paper points were inserted into the periodontal pockets. After 30 s these paper points were removed and placed into 200 mul of sterilized distilled water in a 1.5-ml tube. The sample solutions were gently dispersed with a vortex mixer for 30 s and stored at -20C until they were used in the assays. In this study, we defined a pocket depth of 4 to 5 mm as a shallow pocket and a pocket depth greater than 6 mm as a deep pocket in patients with periodontal diseases. DNA extraction. | The extraction and purification of bacterial and plaque sample DNA were performed with a GFX genomic blood DNA purification kit (Amersham Pharmacia Biotech Inc., Little Chalfont, United Kingdom), and the purified DNA was resolved in 200 mul of sterilized distilled water. The concentrations of plaque DNA were calculated with a PicoGreen dsDNA quantification kit (Molecular Probes, Inc., Eugene, Oreg.), and the results were determined with a standard curve prepared for each assay. Fluorescence was determined at an excitation of 485 nm and emission of 538 nm with a microplate reader (Fluoroscan Ascent; Dainippon Pharmaceutical Co., Osaka, Japan). Design and synthesis of PCR primers. | The primers designed to detect the target species are listed in Table . These primers were determined as described below. The 16S rRNA sequences downloaded from the GenBank database also included those from the 33 bacteria listed in Table . A multiple alignment of these sequences was constructed by using the GENETYX-MAC Multi-sequences program (version 11.0; Software Development, Co., Ltd., Tokyo, Japan). The primers were designed to target signature sequences unique to T. denticola, T. vincentii, or T. medium and searched for online with the BLAST family of programs to ensure their specificities. A total treponemes primer was designed to target common regions in the 16S rRNA sequences of Treponema species. A ubiquitous primer was quoted from reference . These primers were synthesized commercially (Rikaken, Nagoya, Japan). PCR primers and amplification. | Five microliters of DNA from the samples was amplified with 0.2 muM sense and antisense primers specific for the target genes in a 25-mul reaction mixture containing 1.25 U of Taq polymerase (Takara Biomedicals, Shiga, Japan), 200 muM deoxynucleoside triphosphates, the doses of MgCl2 indicated in Table , and reaction buffer. After an initial denaturation at 94C for 2 min, 30 cycles of denaturation (94C for 45 s), annealing at the temperatures indicated in Table for 1 min, and extension (72C for 1 min) for the respective target genes were performed with an iCycler system (Bio-Rad Laboratories Inc., Hercules, Calif.) . Following PCR, 10 mul of the total amplified products was electrophoresed on ethidium bromide-stained 1% agarose gels and visualized under UV fluorescence. A 100-bp ladder (Takara Biomedicals) was used as the molecular size standard. Real-time quantitative PCR. | The TaqMan probe sequence for the 16S rRNA of T. denticola (887 to 907 bp), T. vincentii (948 to 968 bp), and T. medium (891 to 911 bp) was 5'-6-carboxyfluorescein-GAC GGG GGC CCG CAC AAG CGG-6-carboxytetramethylrhodamine-3'. The probe oligonucleotide was synthesized commercially (Sawady Technology, Co., Ltd., Tokyo, Japan). Five microliters of DNA, which was extracted from each plaque sample or the indicated doses of T. denticola, T. vincentii, or T. medium, was amplified for the target genes in a 25-mul reaction mixture containing 0.25 muM sense and antisense primers, 0.25 muM TaqMan probe, 1.25 U of Taq polymerase (Takara Biomedicals), 200 muM deoxynucleoside triphosphates, 4.0 mM MgCl2, and reaction buffer. After an initial denaturation at 95C for 3 min, 50 cycles of denaturation (95C for 30 s), annealing (60C for 1 min), and extension (72C for 1 min) for the respective target genes were performed with an iCycler iQ detection system (Bio-Rad Laboratories Inc.). Statistics. | The comparative frequencies of occurrence of bacterial DNA in subgingival plaque samples were analyzed by Fisher's exact probability test. A difference with a P value of <0.05 was considered statistically significant. TABLE 1 | PCR primers for detection of oral treponemes TABLE 2 | GenBank accession numbers for 16S rRNA sequence RESULTS : Specificities and sensitivities of PCR primers and TaqMan probe. | The specificities of the primers specific for T. denticola, T. vincentii, T. medium, and total treponemes based on the 16S rRNA sequences was determined by using various oral treponemes as well as other bacteria. The primers for T. denticola, T. vincentii, and T. medium demonstrated specific amplification of each bacterial species, as shown in Fig. . The primer for total treponemes amplified treponemal species but none of the other bacteria tested in the experiments. A positive PCR result gave a single band of the expected size, as assessed by electrophoresis. The sensitivities of the individual TaqMan PCR assays were measured with DNA extracted from a 10-fold dilution series of T. denticola, T. vincentii, or T. medium. For the TaqMan probe, quantification was found to be linear with quantities of from 103 to 108 cells of these microorganisms, with correlation coefficients as follow: T. denticola, 0.984; T. vincentii, 0.991; and T. medium, 0.984 . Detection of oral treponemes in subgingival plaque samples. | Four primer sets were simultaneously subjected to a PCR assay for each clinical sample. All of the plaque samples from all 50 subjects were positive in assays with ubiquitous primers, and positive PCR results for the total treponemes primer were observed for 48 samples (96.0% of the total subjects). Each primer for T. denticola, T. vincentii, and T. medium successfully detected a single band of the expected size from the subgingival plaque samples (data not shown). As shown in Table , among the three treponemes, T. denticola was detected at the highest frequency (68.0%), followed by T. medium and T. vincentii, which were found at frequencies of 48.0 and 34.0%, respectively. For two of the samples, from which no band was amplified by the total treponemes primer, no band was detected by the primers for T. denticola, T. vincentii, and T. medium (data not shown). Relationship of oral treponemes with clinical parameters. | Table shows the relationship between the healthy subjects and patients with periodontal diseases regarding the prevalence of T. denticola, T. vincentii, and T. medium. A significant difference (P < 0.05) for T. denticola and T. vincentii was seen between the healthy subjects (pocket probing depth, <=3 mm) and the patients (pocket probing depth, 4 to 5 or >=6 mm). In contrast, T. vincentii was mainly detected in shallow pockets, and a significant difference (P < 0.05) was seen between healthy subjects and patients with shallow pockets (4 to 5 mm) but not those with deep pockets (>=6 mm). The results were also analyzed for a correlation between gender and the frequency of occurrence of oral treponemes; however, there were no significant differences in the frequencies of T. denticola, T. vincentii, and T. medium between males and females. Furthermore, none of these three treponemal species were detected in samples from the group of individuals who were young (age, <=19 years), whereas the total treponemes primer detected treponemes in 83.3% of the samples from this group. The frequencies of occurrence of T. denticola but not T. vincentii between the samples from subjects younger than 20 years of age and those from subjects ages 20 to 39, 40 to 59, and more than 59 years exhibited significant differences. In addition, significant differences were seen regarding the occurrence of T. medium between those younger than age 20 years and those ages 40 to 59 and more than 59 years. Quantitative analysis of oral treponemes in subgingival plaque samples. | The amount of each oral treponeme in subgingival plaque samples was determined by real-time PCR assays. The DNA concentration of each plaque sample was evaluated by a fluorescence technique, and the number of treponemes was represented as the number of cells per picogram of plaque DNA. As shown in Table , the number of T. denticola cells ranged from 3 to 15,184 cells/pg of plaque DNA in the subgingival plaque samples (n = 34), as detected by real-time PCR analysis, while the number of T. vincentii and T. medium cells ranged from 1 to 447 cells/pg of plaque DNA (n = 18) and from 1 to 7,301 cells/pg of plaque DNA (n = 24), respectively. Samples for which bands were not detected by electrophoresis of PCR products were also not positive by real-time PCR. T. denticola and T. medium were detected in plaque samples obtained from patients with relatively wide ranges of pocket probing depths (4 to 8 mm), whereas T. vincentii was primarily found in plaque samples from subjects with pocket probing depths of 4 to 5 mm . In contrast, these organisms were rarely detected in the healthy subjects. FIG. 1. | Electrophoresis evaluation of PCR products amplified with primers for T. denticola Electrophoresis evaluation of PCR products amplified with primers for T. denticola (A), T. vincentii (B), T. medium (C), and total treponemes (D) and with a ubiquitous primer (E). Lanes: M, molecular size marker (a 100-bp DNA ladder); 1, T. denticola; 2, T. vincentii; 3, T. medium; 4, T. socranskii; 5, T. phagedenis; 6, T. pectinovorum; 7, P. gingivalis; 8, P. nigrescens; 9, A. actinomycetemcomitans; 10, E. coli; 11, F. nucleatum; 12, S. mutans; 13, S. oralis; and 14, S. salivarius. The expected sizes are noted by arrows. FIG. 2. | Standard curves generated by analysis of a known number of T. denticola Standard curves generated by analysis of a known number of T. denticola (A), T. vincentii (B), and T. medium (C) cells by real-time PCR. Linearity is observed from 103 to 108 cells of these treponemes (correlation coefficients, 0.984 for T. denticola, 0.991 for T. vincentii, and 0.984 for T. medium). TABLE 3 | Distribution of oral treponemes detected in plaque samples by PCR analysis TABLE 4 | Number of oral treponemes detected in plaque samples by real-time PCR DISCUSSION : Periodontal microorganisms and the changes in the microbial flora have been associated with the development and progression of periodontal diseases. Among the various bacterial species observed in subgingival plaque samples, P. gingivalis, A. actinomycetemcomitans, F. nucleatum, and Bacteroides forsythus have been implicated as periodontal pathogens . Moreover, the oral treponemes T. denticola, T. vincentii, and T. medium have also been identified in subgingival plaque . However, although most putative periodontopathogenic organisms have been cultured under anaerobic conditions, they have demonstrated fastidious growth behaviors, being selective regarding the medium used and requiring long processing times for culture of particular organisms. Furthermore, it is difficult to identify oral treponemal species by culture methods. In the present study, we identified T. denticola, T. vincentii, and T. medium in subgingival plaque samples by PCR analysis with primers specific for the 16S rRNA of each organism . Dewhirst et al. showed that various Treponema species could be detected by partially sequencing the cloned spirochetal 16S rRNA genes that were amplified from the DNA in subgingival plaque samples. Willis et al. also identified Treponema species in dental plaque by a nested PCR assay. We found T. denticola and T. medium in patients with periodontal diseases (pocket probing depth, >=4 mm), whereas T. vincentii was mainly detected in shallow pockets (pocket probing depth, 4 to 5 mm) . Sato and Kuramitsu did not detect T. vincentii in subgingival plaque from patients with advanced periodontitis (pocket probing depth, >7 mm). In the present study, T. denticola, T. vincentii, and T. medium were rarely detected in subgingival plaque samples from healthy subjects (pocket probing depth, <=3 mm), whereas total treponemes were detected in 84.6% of healthy subjects. Moreover, Willis et al. found that among the organisms T. amylovorum, T. denticola, T. maltophilum, T. medium, and T. socranskii, one or more species were detected in subgingival plaque samples from subjects without periodontal diseases (pocket probing depth, <4 mm). We also estimated the amounts of the three species of oral Treponema using real-time PCR with plaque from subjects with and without periodontal disease. The detection responses were linear from 103 to 108 cells for each Treponema species, and there were strong correlations . Lyons et al. determined the amounts of total bacteria in plaque samples using real-time PCR with a universal primer as well as a fluorescent probe, although it is difficult to estimate the number of total bacteria in plaque samples by real-time PCR with a ubiquitous primer because the PCR amplification reaction varies for each organism in a complex microbial mass. In the present study, we used a method that normalized the number of each Treponema species using the concentration of DNA from each plaque sample and found it to be a relatively easy means for comparing the amount of each organism. The number of T. denticola cells tended to increase in deep pockets, and the number of T. vincentii cells tended to decrease in deep pockets . These results suggest that diverse treponemal species are involved with each stage of periodontal disease. Finally, the present study demonstrated that a TaqMan real-time PCR assay is a highly sensitive and specific assay for detecting quantitatively oral treponemes such as T. denticola, T. vincentii, and T. medium in subgingival plaque samples by using specific primers and a TaqMan probe. Backmatter: PMID- 12202556 TI - Comparison of C18-Carboxypropylbetaine and Standard N-Acetyl-l-Cysteine-NaOH Processing of Respiratory Specimens for Increasing Tuberculosis Smear Sensitivity in Brazil AB - Techniques to improve the sensitivity of smear microscopy would facilitate early tuberculosis (TB) diagnosis and disease control, especially in low-income countries where the positive predictive value is high. C18-carboxypropylbetaine (CB-18) is a zwitterionic detergent that helps to compensate for the innate buoyancy of mycobacteria, potentially enhancing recovery by centrifugation. Previous data suggest that CB-18 may increase the sensitivity of smear, culture, and molecular amplification diagnostic testing. The goal of the present study was to evaluate if the sensitivity of the smear technique using light microscopy could be improved by treating respiratory samples with CB-18. In the first phase, respiratory specimens were collected consecutively from patients with suspected pulmonary tuberculosis in a tertiary-care hospital in Rio de Janeiro, Brazil (236 specimens were analyzed). After protocol modifications, another 120 respiratory specimens were evaluated. The standard technique was N-acetyl-l-cysteine with sodium hydroxide (NALC-NaOH) treatment, smear concentration with centrifugation, and Ziehl-Neelsen staining. Culture on Lowenstein-Jensen slants was performed on all specimens for use as the "gold standard." No specimens from patients undergoing active TB treatment were included. The initial protocol for CB-18 processing resulted in a sensitivity of 59.6% and specificity of 96.8% compared to standard processing with a sensitivity of 66.0% and specificity of 96.8%. Using the modified protocol, the sensitivity of CB-18 increased to 71.4% with a specificity of 97.0% versus standard processing with a sensitivity of 61.9% and a specificity of 99.0%. The diagnostic yield of acid-fast bacillus smear with CB-18 in the absence of fluorescence microscopy and PCR compared to standard processing with NALC-NaOH was not significantly different, although the power to detect a difference by the modified assay was low. Keywords: Introduction : Tuberculosis (TB) remains one of the deadliest diseases worldwide, with an estimated 8.4 million new cases and over 2 million deaths each year . Smear microscopy is currently the most feasible microbiological method for diagnosis of pulmonary TB in developing countries due to its rapidity, low cost, and high positive predictive value for Mycobacterium tuberculosis . However, sputum smear microscopy with Ziehl-Neelsen staining is only 60 to 70% sensitive for the diagnosis of pulmonary TB compared with sputum culture . Among human immunodeficiency virus-infected persons the sensitivity of sputum Ziehl-Neelsen smears has been reported to be even lower . Delayed or missed diagnoses contribute to M. tuberculosis transmission and mortality due to TB . Improvements in the sensitivity of smear microscopy would allow earlier diagnosis. TB programs could then optimize disease control from a public health standpoint as well as improve individual patient management . In previously published reports, homogenization and concentration of respiratory specimens by centrifugation have been shown to improve acid-fast bacillus (AFB) smear sensitivity . Recently, another method for processing respiratory specimens using N,N-dimethyl-N-(n-octadecyl)-N-(3-carboxypropyl)ammonium inner salt, also known as C18-carboxypropylbetaine (CB-18), has been reported to result in a higher sputum smear sensitivity. CB-18 is a zwitterionic detergent that is taken up by viable mycobacteria, making them more dense and thereby increasing their recovery by centrifugation. It has been shown that CB-18 smear processing combined with the use of fluorescent stains significantly improves the sensitivities of AFB smears for M. tuberculosis (-). The purpose of this study was to compare light microscopy results of respiratory specimens processed by standard techniques and by CB-18 treatment for the diagnosis of pulmonary tuberculosis in a high- prevalence setting. MATERIALS AND METHODS : Subjects and setting. | Study participants were drawn from the Outpatient and Inpatient Units of the Hospital Universitario Clementino Fraga Filho (HUCFF), Rio de Janeiro, Brazil. The first phase of the study included samples collected between October 1999 and March 2000, and the second phase included samples collected between July and September 2000. Consecutive respiratory samples sent to the laboratory from adult patients with suspected active pulmonary TB were included in the study. The reference standard for the diagnosis of pulmonary tuberculosis was culture growth of M. tuberculosis. Approval for this study was granted by the Institutional Review Boards of HUCFF, the Brazilian Ministry of Health, and the Johns Hopkins Medical Institutions. Written informed consent was obtained from all study subjects. Specimen collection. | Respiratory specimens submitted to the mycobacteriology laboratory with culture request were included. A total of 236 specimens were analyzed during the initial phase, and an additional 120 specimens were analyzed using the modified protocol. Each specimen with a volume of 5 ml or greater was split and processed by the standard digestion-decontamination method that combines N-acetyl-l-cysteine (NALC) and sodium hydroxide (NALC-NaOH) and by the CB-18 method. NALC-NaOH processing. | The respiratory samples were incubated for 20 min with 2.0% NALC-2.0% NaOH, and then centrifuged at 3,000 x g for 20 min. The samples were decanted, and the sediment was resuspended in the remaining supernatant. All specimens were then stained for AFB and cultured as described below. CB-18 processing. | The CB-18 smear kit provided by the manufacturer (Integrated Research Technology, LLC, Baltimore, Md.) included 20x Tris-citrate buffer, CB-18 stock, and NALC. In the initial manufacturer's protocol, respiratory specimens were treated with a buffered solution of 1 mM CB-18 in Tris-citrate buffer (50 mM Tris-HCl, 12.5 mM citrate [pH 7.6], 1.5 mM NaCl) with 15 mM NALC. The samples were mixed with buffered CB-18 and immediately subjected to centrifugation at 3,000 x g for 20 min at 25C. The samples were decanted, resuspended in the remaining supernatant, and then used to prepare smears for diagnostic evaluation. In the second phase of the study, the manufacturer introduced modifications that included (i) increasing the concentration of CB-18 to 4 mM, (ii) lowering the pH of the buffer to pH 6.0, and (iii) incubating specimens in the CB-18 solution for 90 min. In addition, in the second phase of the study, samples were initially homogenized in an equal volume of 15 mM NALC-50 mM NaOH-2.9% sodium citrate for 15 min prior to addition of the CB-18 solution. Following the incubation, samples were again subjected to centrifugation at 3,000 x g for 20 min, decanted, and then resuspended in the remaining supernatant for smear analysis. AFB smears. | Smears made from the NALC-NaOH and CB-18 protocols were heat fixed and stained using the Ziehl-Neelsen (ZN) technique for AFB staining. A small number of specimens (85 specimens) processed with CB-18 were also stained with the Kinyoun staining technique as well as ZN. The smears were scanned at x1,000 magnification with a light microscope. The results were quantified in accordance with published standards . For quality assurance, an experienced scientist in the Mycobacteriology Laboratory at HUCFF reread 10% of the negative smears and all the positive smears. Culture. | A portion of each processed sediment was removed for smear analysis. The specimen was then resuspended in 2.5 ml of distilled water. Lowenstein-Jensen slants were inoculated with approximately 0.4 ml of the resuspended sediment and incubated at 37C for 8 weeks. All specimens that were culture positive for mycobacteria were tested by standard biochemical methods to distinguish M. tuberculosis from nontuberculous mycobacteria, as described elsewhere . Statistical analyses. | Analyses were carried out using STATA 6.0 statistical software. Sensitivity, specificity, and positive and negative predictive values were calculated for each method. Fisher's test and McNemar's test were used for comparisons of results. Agreement between the two AFB staining methods was evaluated using the kappa score. RESULTS : During the course of the study, 494 samples collected from patients suspected of having active pulmonary tuberculosis were processed by the NALC-NaOH and CB-18 processing methods. In the initial phase of the project, 368 specimens were collected; of those, 132 were not included in the analysis. Reasons for exclusion were flaking and washing away of the smear during staining after CB-18 processing (n = 53), anti-TB treatment started more than 4 weeks before sample collection (n = 18), culture contamination (n = 15), and failure to culture (n = 46). For the modified protocol, 126 specimens were collected; of those, 6 were lost and were not included in the analysis: flaking and washing away of the smear during staining with CB-18 processing (n = 1) and anti-TB treatment started more than 4 weeks before sample collection (n = 5). The 417 specimens evaluated included different types of respiratory specimens, including spontaneous sputum (54%), induced sputum (28%), bronchoalveolar lavage (13%), bronchial lavage (4%), and bronchial aspirate (1%) . Samples other than spontaneous sputum showed no significant differences among techniques (standard, original and modified CB-18 protocols [data not shown]). Overall, 68 samples (16%) had cultures that grew M. tuberculosis. Comparison of processing methods. | In the first phase, M. tuberculosis culture growth was identified in 47 samples, and in the second phase another 21 samples were culture positive. The initial CB-18 protocol was slightly less sensitive than routine ZN smears, 59.6% (28 of 47) versus 66.0% (31 of 47) (P = 0.29). During the first phase of the study, there was no significant difference in the positive predictive value between NALC-NaOH decontamination and the CB-18 processing method (89.8 and 88.7%, respectively) using culture as the "gold standard." The modified CB-18 protocol improved the sensitivity of CB-18 to 71.4% (15 of 21) over that of the routine protocol, which was 61.9% (13 of 21); however, this difference was not significant (P = 0.16). The specificities of both diagnostic techniques (CB-18 and NALC-NaOH) were both 96.8% (183 of 189) in the initial protocol and were 97.0% (96 of 99) and 99.0% (98 of 99), respectively, in the modified protocol. Even with the modification of the CB-18 processing method, the positive predictive value for CB-18 remained close to that of routine ZN due to the high specificity of both methods . The modified CB-18 protocol showed an improvement in smear staining compared to the initial protocol, but the number of positive cultures was small (n = 21). The most notable difference between the two versions of CB-18 processing was in the readability of the slides. In the first phase, 14% of smears were not readable versus 1% (P = 0.00) using the modified protocol. Comparison of AFB staining techniques. | A comparison of two staining techniques used in AFB smear analysis was performed on 85 of the samples processed using CB-18. Kinyoun staining was compared to ZN staining to determine if there was any difference that would affect the performance of CB-18 in diagnostic use. There was fair agreement between the two techniques (overall agreement = 72.9%; kappa test = 0.715). However, 22% of Kinyoun slides and 1% of ZN slides were unreadable due to technical problems . FIG. 1. | Types and distribution of respiratory specimens from 356 patients (236 for the initial phase and 120 for the second phase of the study) at HUCFF. Types and distribution of respiratory specimens from 356 patients (236 for the initial phase and 120 for the second phase of the study) at HUCFF. All samples were processed using both the standard method and the CB-18 method for bacillary recovery. TABLE 1 | Evaluation of diagnostic methods TABLE 2 | Evaluation of AFB techniques DISCUSSION : The CB-18 technique is based on the use of the zwitterionic detergent C18-carboxypropylbetaine. Previous reports of increases in both smear and culture sensitivity were believed to result from a reduced impact on viability, alteration of the buoyant density of the mycobacteria to enhance collection efficiency during centrifugation, and dispersion of those mycobacteria that cord . In a previous study the CB-18 processing method used specimens derived from clinical laboratories in the United States. Those authors reported an increase in smear sensitivity for M. tuberculosis from 69.0% with NALC-NaOH to 93.1% with CB-18 (P < 0.01) . We evaluated this technique under field conditions in a high-prevalence, resource-challenged setting. The initial CB-18 protocol was slightly less sensitive than routine smears, 59.6% (28 of 47) versus 66.0% (31 of 47). The modified CB-18 protocol improved the sensitivity of CB-18 to 71% (15 of 21) over that of the routine protocol, which was 62% (13 of 21). There were no significant differences in the sensitivity or specificity of the two methods for either version of CB-18 processing. The power of this study to detect differences in the two techniques was limited by the number of positive cultures (n = 68), especially in the second phase, where there were only 21 positive specimens. The sensitivity of smear improved when the modified CB-18 protocol was used, but the increased length of time required (90-min incubation for CB-18) limits its routine utility in this setting. These results are consistent with the findings of another recent evaluation of CB-18. Comparison of a sodium dodecyl sulfate-sodium hydroxide processing method with a CB-18 processing protocol that included no incubation step failed to show improvements in smear sensitivity . When a 90-min incubation step was included in the procedure, there was a small improvement in the CB-18-based smear analyses. While CB-18 failed to show previously described improvements in the diagnostic evaluation of TB patients under field conditions, a potential limitation of this study is that M. tuberculosis culture (Lowerstein-Jensen medium) was used as the reference standard for determining diagnostic effectiveness, and culture may be imperfect in identifying all active TB cases. However, due to the high specificity of both techniques among our samples, it appears that using culture as the gold standard did not compromise our field evaluation of CB- 18. The key difference between this study and those that have previously shown a higher smear sensitivity with CB-18 processing is the use of fluorochrome staining methods in the earlier studies. Auramine-rhodamine stain is a more sensitive technique, which could account for the discrepancy between the data presented here and those of Thornton et al. (-). We compared two fuchsin stain procedures and found that ZN staining identified twice as many positive smears as Kinyoun staining. ZN staining proved to be more efficient than Kinyoun staining for analyzing CB-18- processed specimens because many Kinyoun-stained slides were unreadable. Although our study was not able to demonstrate an improvement in smear light microscopy using CB-18 compared to digestion-decontamination with NALC-NaOH followed by centrifugation, CB-18 may be more useful when used in conjunction with nucleic acid amplification, fluorescence staining, or other types of specimens. Published reports have shown an improvement in the recovery of mycobacteria other than tuberculosis among human respiratory specimens and also among animal products such as milk . CB-18 has also been used for preparing samples for amplification and has been associated with an improvement in PCR results compared to other standard techniques . In conclusion, in our setting neither the original nor the modified CB-18 protocol showed a significant increase in diagnostic yield of light microscopy smear compared to NALC-NaOH decontamination and concentration by centrifugation. A larger sample size may have detected a difference, however. CB-18 sample processing may be more useful when used in conjunction with more sensitive techniques for M. tuberculosis detection such as fluorescence microscopy and PCR amplification. Further field evaluation to address these issues is warranted. Backmatter: PMID- 12202577 TI - Leptotrichia amnionii sp. nov., a Novel Bacterium Isolated from the Amniotic Fluid of a Woman after Intrauterine Fetal Demise AB - A novel bacterium was isolated and characterized from the amniotic fluid of a woman who experienced intrauterine fetal demise in the second trimester of pregnancy. The bacterium was a slow-growing, gram-negative anaerobic coccobacillus belonging to the genus Leptotrichia. Unlike Leptotrichia sanguinegens, the isolate did not grow in chopped-meat glucose broth or on sheep blood agar upon subculturing. The isolate was characterized by sequencing and analyzing its 16S rRNA gene. The 1,493-bp 16S ribosomal DNA sequence had only 96% homology with L. sanguinegens. Several phylogenetic analyses indicated that L. amnionii is a distinct species and most closely related to L. sanguiegens. Keywords: Introduction : Molecular-based diagnostic and identification methods for fastidious or uncultivable bacteria have resulted in the recognition of many new pathogenic microorganisms . One of the most successful methods is PCR amplification and sequencing of the bacterial 16S rRNA gene. This method has been successfully applied to environmental as well as clinical samples . The large rRNA sequence databases at GenBank and at Ribosomal Database Project II allow for a quick comparison of 16S ribosomal DNA (rDNA) sequences and accurate identification of bacteria that are difficult to identify on the basis of phenotypic properties alone . The use of this method has greatly expanded the list of indigenous microbial flora of humans and has helped in recognizing the numerous opportunistic pathogens that cause infections related to severe physiological stress and immunosuppression due to chemotherapy. Leptotrichia species are slow-growing, gram-negative anaerobic flora of the oral cavity and genital tract . Colonization by Leptotrichia species has been reported in over 40% of children less than a year old . Leptotrichia buccalis, which is considered indigenous oral flora, has been associated with endocarditis in patients with Down's syndrome and bacteremia in neutropenic children and adults . They seem to colonize permucosal implants of edentulous patients and, not surprisingly, are often considered contaminants if isolated from clinical specimens. Leptotrichia sanguinegens has recently been proposed as an agent of postpartum and neonatal bacteremia . It has not been identified from a healthy individual. We describe an isolate that is related to the species L. sanguinegens, but is different in its genotypic properties and nutritional requirements . For this isolate, we proposed the name "L. amnionii sp. nov." (from "amnion," the extraembryonic membrane enveloping the embryo in utero and containing the amniotic fluid), to signify its source of isolation. CASE REPORT : A 27-year-old previously healthy, multiparous female in the second trimester of pregnancy presented to the emergency room with severe headache, neck and back pain, and a temperature of 102F. The abdominal examination demonstrated no guarding or rebound. A purulent vaginal discharge was noted, but the physical examination was otherwise normal. Fetal heart tones were present. The patient was hospitalized. Initial laboratory values demonstrated a leukocyte count of 7,800 with the differential showing 1 metamyelocyte, 10 band forms, 85 segmented neutrophils, and 3 lymphocytes. The hemoglobin level was 11.5 g/dl, and the C-reactive protein level was 11.4 mg/dl. A urinalysis was unremarkable, showing no evidence of infection. A wet preparation of the vaginal discharge demonstrated no abnormal organisms or evidence of significant vaginal infection. A PCR test for Chlamydia trachomatis was negative. Cerebral spinal fluid evaluation was normal. The symptoms gradually resolved, and the patient was discharged. Six days later, the patient was seen in the outpatient clinic. No fetal heart tones could be heard, and an ultrasound confirmed an intrauterine fetal demise. The patient was admitted for uterine evacuation by labor induction. Prior to induction, an amniocentesis was performed. The amniotic fluid was turbid and brown in color and had a distinct foul smell. A gram stain of the amniotic fluid demonstrated gram-negative coccobacilli. A slow-growing, gram-negative anaerobic coccobacillus was recovered. Scant growth of Bacteroides fragilis and Propionibacterium acnes was observed in cultures of the placenta. The mother was given amoxicillin-clavulanic acid and had an uneventful recovery. MATERIALS AND METHODS : Microbiology. | The amniotic fluid and the placenta tissue specimens were cultured on blood and chocolate agar under both aerobic and anaerobic conditions at 37C. The template DNA for 16S rDNA PCR was prepared from a few colonies that were isolated on the prereduced blood agar incubated anaerobically. The DNA was extracted with a Qiagen DNA extraction kit (Qiagen, Inc., Valencia, Calif.). Broad-range prokaryotic PCR primers and nested sequencing primers were used to amplify and sequence the 16S gene rRNA. The methodology has been described previously . Phylogenetic analysis. | The rDNA sequence of the L. amnionii sp. nov. was aligned with a database of archaeal, bacterial, and eucaryal SSU rRNA sequences (ca. 10,000 sequences in total) by using the ARB software package . Both BLAST analysis and the parsimony insertion tool of ARB tentatively placed the L. amnionii sequence within the bacterial division of Fusobacteria. Consequently, a subset of the ARB alignment, which included the Leptotrichia species of the division Fusobacteria (including the species of the genus Leptotrichia), as well as members of other, outlying bacterial divisions, was selected for phylogenetic analysis. Both full-length data sets and sequence alignments minimized by the use of the Lane mask were analyzed. The sequences of Methanococcus jannaschii and Sulfolobus acidocaldarius were selected as out-groups for phylogenetic analysis. The dendrogram presented in Fig. was constructed by evolutionary distance analysis (neighbor joining with Olsen correction) with the ARB package . The robustness of this tree was assessed by bootstrap resampling (>100 replicates) of evolutionary distance trees by using weighted least-squares mean analysis with Kimura two-parameter or maximum-likelihood correction of evolutionary distances (PAUP* version 4.0b2) . Parsimony and maximum-likelihood analyses (ARB or PAUP*) provided results that were substantially similar to those of the evolutionary distance algorithm. Nucleotide sequence accession number. | The 16S rRNA sequence of the L. amnionii sp. nov. was deposited in GenBank and given accession no. . FIG. 2. | Evolutionary distance dendrogram of selected leptotrichial and fusobacterial 16S rRNA sequences, including that of the Leptotrichia-like sp. Evolutionary distance dendrogram of selected leptotrichial and fusobacterial 16S rRNA sequences, including that of the Leptotrichia-like sp. isolate. Two archaeal species, Methanococcus jannaschii (M59126) and Sulfolobus acidocaldarius (D14876), were chosen as out-groups for phylogenetic analysis. Sequences are identified by species name and GenBank accession number. Branch points supported by >90% bootstrap values are indicated by solid circles. Open circles represent branch points with bootstrap values in the range 75 to 89%. Branch points without circles were not resolved (bootstrap values in the range <75%) as specific groups by this analysis. The bar at the bottom indicates the number of nucleotide changes per site. RESULTS AND DISCUSSION : Numerous gram-negative coccobacilli were observed in the amniotic fluid along with numerous neutrophils. Anaerobic culture of the amniotic fluid on blood and chocolate agar resulted in very small gray colonies, <1 mm in diameter, following 72 h of incubation. Gram stain of the colonies revealed gram-negative coccobacilli, including some filamentous forms . There was no growth on blood agar incubated under aerobic conditions, nor was there anaerobic growth on kanamycin and vancomycin or Mueller-Hinton agars upon subculturing. Viral cultures were negative. Since the bacterium resembled L. sanguinegens, it was inoculated into chopped meat glucose (CMG) broth and incubated under aerobic conditions . This medium did not sustain growth, as evidenced by the lack of turbidity of the medium. The organism was extremely fastidious and did not survive beyond the third subculture. There was insufficient growth to perform biochemical or fatty acid analysis. The isolate was identified and characterized by PCR amplification of the 16S rRNA gene by using broad-range eubacterial primers FD1 and RD1 . The PCR product was directly sequenced as described previously . A 1,493-nucleotide consensus sequence was created and edited with DNAsis software (Hitachi Corporation) and compared with the sequences deposited in the GenBank database. The submitted sequence had only 96% homology to L. sanguinegens (GenBank accession no. ). Phylogenetic analysis. | Based on its unique source of isolation, inability to grow on known special media, such as CMG broth, and unique 16S rDNA sequence, it was evident that this bacterium was related to, but different from, L. sanguinegens . The phylogenetic relationship of L. amnionii with other species of the bacterial division Fusobacteria, including species of the genus Leptotrichia, was inferred by evolutionary distance, parsimony, and maximum-likelihood analyses. Figure shows a representative evolutionary distance dendrogram. Parsimony and maximum-likelihood analyses gave qualitatively similar results. Bootstrap resampling of data provided strong support for a specific association of L. amnionii with other members of the genus Leptotrichia. L. sanguinegens was identified as the closest neighbor of L. amnionii (bootstrap values of 99 and 100% for distance and parsimony analyses, respectively). Clinical significance. | Three species of Leptotrichia, L. buccalis, L. trevisanii, and L. sanguinegens (also called Sneathia sanguinegens) , have been associated with human infections . Fifty-nine percent of the patients were immunosuppressed due to malignancy. Four cases of L. sanguinegens bacteremia were associated with pregnancy, and two neonates were infected . L. buccalis has been well characterized and is part of the normal oral flora. It has been isolated from babies <1 year of age, and 40% of babies seem to be carriers . L. trevisanii is a recently identified bacterium that was recovered from a patient with myeloid leukemia. L. sanguinegens has been proposed as an agent of postpartum and neonatal bacteremia. At this time, we suspect that L. amnionii is indigenous to the urogenital tract and is an opportunist in the appropriate clinical situations. Like other Leptotrichia-related clinical cases, this bacterium was isolated from a clinical condition that is physiologically somewhat analogous to having the same stress and immunosuppression as an underlying malignancy. Hanff et al. described the presence of a strong odor from two neonatal cases of infection possibly due to L. sanguinegens , as was detected from the amniotic fluid in our case. It appears that L. amnionii is not a blood-loving microbe like L. sanguinegens, because it failed to grow on blood agar. Additional clinical isolates will help establish its true ecological niche and pathogenic potential. Description of Leptotrichia amnionii sp. nov. | The name "L. amnionii" (am' n.on.e.i. L. gen. n., amnionii) is derived from the word "amnion." The organism is characterized by pleomorphic coccobacillus, long, nonmotile, fusiform cells. Some cells are joined end to end in a filamentous form. L. amnionii grows anaerobically on blood agar after 3 days of incubation and is closely related to L. sanguinegens based on its 16S rDNA sequences. FIG. 1. | Gram stain of amniotic fluid colonies demonstrating gram-negative pleomorphic bacilli. Gram stain of amniotic fluid colonies demonstrating gram-negative pleomorphic bacilli. TABLE 1 | Comparison of clinical features of Leptotrichia species isolated until 2001 Backmatter: PMID- 12202596 TI - Labor and Cost Requirements of Two Commercial Assays for Qualitative Molecular Detection of Hepatitis C Virus AB - The Bayer transcription-mediated amplification (TMA) and the Roche PCR Amplicor version 2.0 molecular assays for the qualitative detection of hepatitis C virus were compared for cost, hands-on time, assay duration, and complexity. The TMA assay compares well to PCR and may be especially useful for laboratories with large numbers of test requests. Keywords: Introduction : Since the cloning of the hepatitis C virus (HCV) genome in 1989 , diagnosis of infection has been performed by various immunologic and molecular genetic assays in addition to the traditional examination of liver biopsy specimens. Direct amplification of HCV RNA provides early evidence of infection independent of a patient's immunological status and is an important tool in the assessment of viral clearance and thus response to therapy . We performed a time-and-motion study and compared cost of reagents and disposables, labor involved in testing and instrument maintenance, time to result, and hands-on-time of the Bayer transcription-mediated amplification (TMA) assay versus the Roche PCR Amplicor version 2.0, our current standard protocol. The Roche PCR assay (Roche Molecular Systems, Inc., Pleasanton, Calif.) utilizes reverse transcription-PCR . In brief, samples are prepared by precipitation of the viral RNA and addition of an HCV internal control RNA for assay quality control. Subsequently, the target RNA is reverse transcribed and then amplified with specific HCV primers which are designed from the highly conserved 5' untranslated region of the viral genome. Target-specific oligonucleotide probes hybridize with the amplified products. The presence of a hybridization product is established by colorimetric analysis on a COBAS AMPLICOR analyzer. The qualitative Bayer assay amplifies RNA by in vitro transcription of a DNA template. The viral nucleic acids in the specimen, as well as those of an internal control, are hybridized to capture oligonucleotides complementary to highly conserved HCV 5'-untranslated-region sequences. The RNA is subsequently captured on magnetic microbeads. After the addition of primers, reverse transcriptase, and T7 RNA polymerase, RNA is amplified from a DNA template during a 1-h incubation at 41.5C. Single-stranded, fluorescently labeled probes are then hybridized to the RNA amplicons, whereas unhybridized probes are chemically degraded. A Leader HC luminometer (GenProbe, San Diego, Calif.) distinguishes the chemiluminescent signals of the target sequence in the samples and the internal control by differential light emission (Bayer Diagnostics Division, Norwood, Mass.). We evaluated 182 patient sera by the TMA assay. The 182 samples included specimens sent to our laboratory for HCV enzyme immunoassay, HCV recombinant immunoblot assay (RIBA), and qualitative or quantitative HCV PCR by the Roche assay. We also included random specimens that were assessed for analytes not related to HCV, such as human immunodeficiency virus, hepatitis A and B viruses, cytomegalovirus, syphilis organisms, and rubella and varicella zoster virus. A side-by-side evaluation of the TMA assay and the PCR assays was performed with 172 of 182 samples, with TMA assay runs of 24, 54, and 94 patient samples in comparison with either 10 (one ring) or 22 (two rings) patient sample runs of the Roche system. We included one positive and one negative control sample in each Roche ring and three positive as well as three negative calibrators in each TMA assay run. Twelve of the 172 samples were duplicates for reasons not related to the assay. The remaining 10 (of 182) samples were not directly compared to the Roche method but were used to evaluate the possibility of performing INNO-LIPA genotyping (Innogenetics Inc., Norcross, Ga.) after the TMA assay. Our study population consisted of 10 samples with indeterminate RIBA results and 65 with HCV antibody-positive results. Of the 170 samples, 25 were originally sent for testing by Roche PCR and an additional 10 were both positive by PCR and subsequently genotyped by the INNO-LIPA system (Innogenetics, Inc.) . Sixty were random samples (tested for HCV a priori in the TMA assay and PCR comparison). A total of 65 of 170 samples were HCV positive by either qualitative or quantitative PCR (Roche Amplicor), and 66 of 170 were positive by the TMA assay. The sample newly identified as positive by TMA analysis remained negative by qualitative PCR on repeat testing. This finding may reflect either a false-positive result, higher sensitivity of the TMA assay, or PCR inhibition of this sample. The sample was subsequently determined to be negative by RIBA (Chiron Corporation, Emeryville, Calif.), but this is a less sensitive assay than PCR or the TMA assay. Therefore, the true status of this specimen could not be determined. Results were obtained for all TMA samples, whereas two PCR samples failed to undergo amplification after several attempts. Results for the 12 samples that were duplicated in the TMA assay were entirely equivalent. The TMA assay requires a sample volume of 500 mul, versus 200 mul for PCR. Although the Roche amplified product can be used directly, subsequent genotyping using the line probe method (INNO LiPA HCV II; Innogenetics, Inc.) from the TMA product is possible only after an additional amplification. On the other hand, the TMA assay is very easy to perform, does not require repeated vortexing as does the Roche assay, and requires minimal instrument maintenance. The time to results and hands-on time per sample were measured once in each individual assay setup . Both are substantially better for TMA analysis, even when a large number of specimens is tested. Consequently, labor costs are lower for the TMA assay. The general cost of disposables is also lower for TMA assay, but these costs remain well below $2 per reportable result in both assays. Comparing list prices for both assays, the reagent cost of Roche PCR analysis is significantly lower than that of the TMA assay, although manufacturers' volume discounts certainly influence final pricing. FIG. 1. | Detailed and global representation of hands-on and hands-off (h-o) time for HCV diagnosis by the Roche PCR Amplicor version 2. Detailed and global representation of hands-on and hands-off (h-o) time for HCV diagnosis by the Roche PCR Amplicor version 2.0 assay and the Bayer TMA assay. "Break" indicates the technologists' time away from the bench. Neither assay was delayed by these breaks. The bars reflect respective amounts of hands-on and hands-off time in both assays, as well as their relative positions within the assay timelines. TTU, test tube unit (a fixed row of 10 test tubes used in the TMA assay). In summary, TMA analysis compares well to PCR amplification for ease of use and reliability of results. Only the Roche PCR test is FDA cleared at this time. Because larger numbers of samples can be processed in a shorter time than for the PCR assay, the TMA assay may be the assay of choice in a laboratory which requires short turnaround times and can accept a larger required sample volume. The TMA assay may be especially useful for laboratories with large numbers of qualitative HCV diagnostic requests. Backmatter: PMID- 12202540 TI - Laboratory Diagnosis of Lower Respiratory Tract Infections: Controversy and Conundrums AB - Keywords: null: null ACUTE BRONCHITIS : Even though acute bronchitis is clearly one of the most common diagnoses made in adult clinical practice, a precise definition does not exist. A cough that lasts 1 to 3 weeks, with or without sputum production, and that is associated with upper respiratory tract and constitutional symptoms is the typical presentation. Symptoms result from inflammation and hyperresponsiveness of the bronchial tree. Table lists the most common pathogens implicated in acute bronchitis. Viruses, especially influenza virus, cause the vast majority of cases in studies that establish an etiology. Respiratory syncytial virus can also cause symptomatic lower respiratory tract disease, especially in elderly patients . Nonviral agents that have been implicated include Mycoplasma pneumoniae, Chlamydia pneumoniae, Bordetella pertussis, and Bordetella parapertussis. The latter pathogens are most frequently seasonal and occur in epidemics . There are no data to suggest that Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis are important pathogens in uncomplicated bronchitis . Diagnosis is usually made clinically. Purulent sputum is not predictive of viral infections versus bacterial infections. Microscopic examination of sputum and culture are not helpful in distinguishing upper airway colonization from lower airway infection . Meta-analyses of numerous prospective studies have not demonstrated the value of microbiological studies in the management of patients with uncomplicated acute bronchitis . Rapid tests for influenza virus are limited by sensitivities equivalent to that of clinical judgment when influenza is circulating in a community (70 and 65 to 80%, respectively) . Perhaps the one exception to the previous statement is the patient with possible pertussis. Diagnostic tests should be performed for the patient with chronic cough in the setting of appropriate epidemiology, since pertussis is indistinguishable clinically from other causes of bronchitis and adults may be vectors of infection in nonimmunized infants or incompletely immunized children. Table lists the available diagnostic tests. Acute exacerbation of chronic bronchitis in patients with underlying lung disease is usually distinguished from the uncomplicated syndrome discussed above. Although Streptococcus pneumoniae and Haemophilus influenzae contribute more frequently to acute symptoms in patients with chronic obstructive pulmonary disease, viral pathogens are likely responsible for a significant number of episodes (30 to 40%) . Gram-stained smears and culture are of limited value because as many as 25% of patients with chronic obstructive pulmonary disease have upper airway colonization with bacterial pathogens in the absence of symptoms . Management of these patients remains controversial. TABLE 1 | Most common pathogens implicated in lower respiratory tract syndromes and their relative contributions TABLE 2 | Diagnostic studies for specific agents of lower respiratory tract infections COMMUNITY-ACQUIRED PNEUMONIA : CAP is usually characterized by fever, chills, dyspnea, cough, and pleuritic chest pain in association with physical findings suggestive of consolidation in persons who become ill outside of a hospital or chronic-care facility. Table l lists some common organisms that are associated with CAP. Streptococcus pneumoniae is still considered the major cause of CAP. The relative contributions of agents such as Mycoplasma pneumoniae and Chlamydia pneumoniae depend upon the published series, whether there was a community outbreak at the time of the study, and the diagnostic method used. The practice guidelines of both the American Thoracic Society (ATS) and the Infectious Diseases Society of America (IDSA) emphasize using the history and physical examination to aid in the selection of nonmicrobiological diagnostic tests for assessment of the severity of illness and as a guide to empirical antibiotic choices. The history may also provide certain epidemiological clues that may be important when considering a particular etiology. For example, hantavirus pulmonary syndrome should be considered in the otherwise healthy patient who presents with a prodromal illness that rapidly progresses to adult respiratory distress syndrome following activities in an area of endemicity that increase the risk of exposure to rodents. Both ATS and IDSA recommend chest radiography to distinguish pneumonia, which requires antibiotics, from acute bronchitis, which is most commonly viral in etiology . The chest radiograph lacks specificity in establishing a microbial cause of CAP, but it may provide clues to the diagnosis of perhaps unsuspected illnesses that may be mistaken for CAP such as tuberculosis and Pneumocystis carinii infection . Detection of an etiologic agent causing infection such that directed therapy is permitted is the role of microbiological tests. Unfortunately, the ideal test for most pathogens does not yet exist. Methods include sputum Gram stain and culture, blood cultures, serologic studies, antigen detection tests, and nucleic acid amplification methods. Table lists the methods available for the detection of most common pathogens associated with CAP. Among these methods, perhaps the most controversial are the sputum Gram stain and culture. The IDSA guidelines recommend these for patients with CAP who require hospitalization, whereas the ATS guidelines do not. While often regarded as a simple test, proper collection of the sputum sample, rapid transport to the laboratory, adequate sampling of the purulent component of the sample, preparation of the stain, and interpretation are all required. The values of the sputum stain and culture results are also dependent upon the pretest probability that the patient has bacterial pneumonia and upon whether the patient has received antibiotics. Add to this a recent study which demonstrated the intralaboratory sampling variability of expectorated sputum in five centers and it is clear why the value of the Gram stain has been challenged. However, proponents argue that when the caveats mentioned above are fulfilled, namely, adequate sputum collection from a patient with productive purulent sputum who has not received antibiotics, the demonstration of a predominant morphotype may be useful in guiding pathogen-oriented antimicrobial therapy . In a recent study by Rozon et al. , the sensitivity and specificity of a Gram stain from a good-quality specimen for the diagnosis of pneumococcal pneumonia and Haemophilus influenzae pneumonia were 57 and 82%, respectively, and 97 and 99%, respectively. Moreover, in those patient samples in whom a predominant morphotype was seen, 95% of patients received monotherapy as opposed to combination therapy , leading to potential cost savings and less antimicrobial agent-related adverse events. If it is decided to send a sample to the laboratory, patients should be given proper instructions. Food should not have been ingested for 1 to 2 h prior to expectoration. The mouth should be rinsed with saline or water, and the patient should be encouraged to breathe and cough deeply and expectorate immediately into a sterile container. Ideally, the sample is then transported immediately to the laboratory, where it is stained and plated as soon as possible upon receipt. At the University of Utah, attempts to standardize collection as part of an institutionwide focus on appropriate management of respiratory tract infections failed because the burden fell to already overworked nurses or respiratory therapists who simply could not perform these time-consuming steps (personal communication). Once the specimen reaches the microbiology laboratory, it has been established that a microscopic screen to exclude those samples that represent upper airway contamination is beneficial and cost-effective. The specimen is viewed under low power (x10 objective), and the numbers of epithelial cells and/or polymorphonuclear leukocytes (PMNs) present establish the degree of contamination. The presence of many epithelial cells and few to no PMNs is suggestive of a poorly collected sample and the sample should not be planted. Multiple specific criteria incorporating PMNs, epithelial cells, mucus stranding, and the presence of bronchial epithelial cells have been published; but the superiority of one method over the others has not been established . If the sample is inadequate, a new one can be requested. In good-quality screened samples, the presence of a predominant bacterial morphotype should also be reported. Screening should not be applied to samples obtained from patients with possible Legionella or Mycobacterium tuberculosis infection . Routine sputum specimens are typically planted on blood agar, chocolate agar, and MacConkey agar. Although patients with Legionella pneumonia rarely produce purulent sputum , a Gram stain that demonstrates abundant PMNs with scant respiratory flora (in a patient not on antibiotics) is cause for suspicion for this pathogen, and the use of a medium selective for Legionella should be considered after consultation with the physician . While the diagnostic yield from blood samples from patients with CAP is low (5 to 16%) , both ATS and IDSA recommend obtaining them from hospitalized patients before antibiotic therapy is administered . The benefits include definitive identification of the etiologic agent and an estimate of a prognosis, which is helpful for patient management. Up to 40% of patients admitted with CAP will have an accompanying pleural effusion . The decision to perform a thoracentesis is a clinical one, but a Gram stain and culture of the fluid with the media discussed above should be performed. Infected fluids are managed aggressively with chest tube drainage, whereas small parapneumonic effusions typically resolve on their own. Antigen detection tests have a role in the establishment of viral etiologies such as respiratory syncytial virus and influenza virus. Direct fluorescent-antibody tests are more sensitive than point-of-care rapid tests . The specimen of choice is a nasal aspirate or wash or a nasopharyngeal swab. Throat swab specimens are less useful. Urinary antigen tests should be performed for patients suspected of having Legionella pneumophila infection, particularly in geographic settings where serogroup 1 predominates. The sensitivity ranges from 70 to 90%, and the specificity is >99% . Recently, a new method for detection of Streptococcus pneumoniae antigen in urine, an immunochromatographic assay (NOW S. pneumoniae urinary antigen test; Binax, Inc., Portland, Maine), has become available. A large study of 420 adults with CAP and 169 control patients demonstrated that the test has a high degree of specificity and a sensitivity of 80% when positivity by blood culture was used for comparison . Serologic studies are usually reserved for the atypical pathogens including Mycoplasma pneumoniae, Chlamydia pneumoniae, and Legionella pneumophila, among others. The relative contributions of these pathogens to cases of CAP vary depending upon the population studied and the diagnostic methods used. Diagnosis of infections caused by these pathogens is particularly problematic because the clinical presentations may be confused with a variety of other infectious agents, and culture, while possible, is either insensitive or slow and requires specialized culture techniques. Unfortunately, the most reliable serologic evidence implicating infection with one of the organisms mentioned above requires a fourfold increase in immunoglobulin G antibody titers between acute- and convalescent-phase serum samples, which confirms but which does not establish the diagnosis early enough to be useful in routine patient management. Given the lack of rapid or reliable methods for the detection of these agents, the practice guidelines of both ATS and IDSA have incorporated empirical therapeutic regimens that are routinely used for the treatment of infections caused by these organisms . Nucleic acid amplification tests have been developed by many laboratories to more rapidly and accurately detect those pathogens that are difficult to culture. A commercial multiplex PCR assay for the detection of respiratory virus infections is available as a test for research use only (application for approval has been submitted to the U.S. Food and Drug Administration). It has excellent sensitivity and specificity but is costly and time-consuming . Likewise, PCR detection of Bordetella pertussis and Bordetella parapertussis has been shown to be more rapid and at least equivalent to culture, provided that calcium alginate swabs are not used for specimen collection . Two U.S. Food and Drug Administration-approved commercial nucleic acid amplification tests for direct detection of Mycobacterium tuberculosis from respiratory samples are available, the AMTDT (Gen-Probe Inc., San Diego, Calif.) and the Amplicor and COBAS (Roche Molecular, Branchburg, N.J.) tests. Optimal protocols for detection of other pathogens have yet to be established. The following parameters should be established before a nucleic acid test is incorporated into routine clinical use for a particular pathogen: optimum specimen type, internal inhibition control, analytical and clinical sensitivity and specificity, and reproducibility (CAP guidelines) NOSOCOMIAL PNEUMONIA : Pneumonia is the most frequent nosocomial infection (30 to 33% of cases) among combined medical-surgical intensive care units participating in the National Nosocomial Infections Surveillance System . In the intensive care unit setting, 83% of cases of pneumonia are associated with mechanical ventilation . Staphylococcus aureus is the most frequently reported isolate at 17% . Fifty-nine percent of reported isolates are aerobic gram-negative species, the most common of which is Pseudomonas aeruginosa (15.6%), followed by Enterobacter species (10.9%) and Klebsiella pneumoniae (7.0%) . Frequently, infection is polymicrobial . The diagnosis of pneumonia in the hospitalized patient is even more challenging than the diagnosis of CAP. When fever, leukocytosis, and purulent tracheal secretions develop in association with an abnormal chest radiograph, the likelihood of pneumonia is high . However, symptoms suggesting pneumonia may be muted in debilitated or elderly patients, and a variety of other noninfectious conditions may mimic pneumonia . Clinical findings alone, then, are not sufficient for a definitive diagnosis. A variety of noninvasive and invasive tests have been proposed as guides for diagnosis and treatment of hospital-acquired pneumonia. The American College of Chest Physicians convened a panel of experts to establish diagnostic recommendations for ventilator-associated pneumonia based upon an evidence-based assessment of the medical literature . The executive summary prepared by that committee concluded that the lack of specificity of clinical findings and the poor reproducibility of chest radiography warrant the performance of additional procedures, such as cultures of specimens from the lower respiratory tract . Although qualitative culture and Gram stain of endotracheal sputum samples are the least invasive tests, they have the same pitfalls for hospitalized patients as for patients in the community, that is, poor predictive values. Both pathogens and nonpathogens alike may be recovered. Bronchoscopy has been advocated by many. Samples that can be obtained by bronchoscopy include bronchial brushings, bronchial washings, bronchoalveolar lavage (BAL) fluid, and transbronchial biopsy specimens . It is important that a standardized approach be followed. Baselski and Wunderink describe in detail appropriate collection and handling techniques. Two diagnostic approaches are described: the serial dilution method, in which two 100-fold dilutions are made, followed by plating of a measured 0.1-ml amount of material on an agar medium, with direct colony counts reported as the number of CFU per milliliter, and the calibrated loop method, which is similar to the method used for the plating of urine samples . Established quantities for contamination versus infection are >103 CFU of a single organism per ml for protected specimen brushes (PSBs) and >104 CFU of a single organism per ml for BAL fluid . The PSB technique involves advancing a double-catheter brush that contains a distal occluding plug through a fiberoptic bronchoscope. After the bronchoscope is wedged, the plug is ejected and distal secretions are sampled via the brush. The brush is then retracted through the inner lumen of the catheter, which in turn is retracted into the outer cannula . A limitation of this procedure is the small volume obtained (similar0.001 ml), which is diluted in 1 ml of transport medium . A criticism of the literature advocating the PSB technique is that the quality of the samples is usually not reported . Mertens et al. suggest that samples obtained by the PSB technique be screened by using cytospin Gram stains. Specimens containing <10 cells per high-power field may reflect poor sampling, indicating unreliable results . Many intensive care specialists prefer BAL fluid because a large number of alveoli (similar106) are sampled. Reported sensitivities of quantitative BAL fluid cultures range from 42 to 93% , with a mean of 73%, and specificities range from 45 to 100%, with a mean of 82% . The specificity is higher (89 to 100%) when intracellular organisms are detected . Results vary due to differences in the population studied, the prior administration of antibiotics, and the reference test compared . Finally, blinded invasive procedures have been advocated by some because of the expense and potential risk of invasive procedures. Some of these methods include mini-BAL, blinded bronchial sampling, and blinded sampling by the PSB technique . The reported sensitivities and specificities are similar to those for invasive techniques . Since the involved portion of the lung may be missed, this technique should probably be reserved for patients too unstable to undergo bronchoscopy . Regardless of the quantitative method used, the American College of Chest Physicians' position is that there are insufficient outcomes data to show that treatment based on the results of quantitative testing ensures a better clinical outcome . In contrast, a large randomized trial among 31 intensive care units in France showed that a management strategy involving invasive procedures was significantly associated with reduced rates of mortality and morbidity and resulted in less antibiotic use . Perhaps the greatest utility of quantitative cultures of specimens obtained by invasive procedures at present may be in reducing antibiotic use for clinically insignificant organisms and for distinguishing between pneumonia and adult respiratory distress syndrome or other noninfectious causes. Also, there is general agreement at this time that the usefulness of repeated quantitative cultures to assess the response to therapy needs to be better studied. PNEUMONIA IN IMMUNOCOMPROMISED HOSTS : Pneumonia is one of the most life-threatening infections in the immunocompromised host. A broad range of pathogens needs to be excluded; and the infectious agents to be considered vary depending upon the type and duration of immunosuppression, past exposures, geographic location, and the nature of the treatments administered. Less controversial than the diagnostic utility of ventilator-associated pneumonia is perhaps the diagnostic utility of fiberoptic bronchoscpy in this setting. BAL protocols which process samples for both viral and bacterial pathogens, Pneumocystis, Legionella, fungi, and mycobacteria as well as cytologic analysis for noninfectious causes may be appropriate. Such protocols require communication between the clinical microbiology laboratory, infectious diseases specialists, pulmonologists, and transplant teams. In summary, lower respiratory tract infections are among the most commonly encountered infectious diseases causing significant morbidity and mortality. The role of the microbiology laboratory in diagnosis remains controversial until better standardization of methods and outcomes data are generated. Empirical treatment approaches are recommended for bronchitis and CAP not requiring hospitalization. In the hospitalized patient, although diagnostic tests are imperfect, they are suggested. This is particularly true for the immunocompromised host, for whom invasive procedures guided by clinical and epidemiological data may reveal unsuspected opportunistic pathogens. Backmatter: PMID- 12202541 TI - Emergence of Klebsiella pneumoniae Isolates Producing Inducible DHA-1 beta-Lactamase in a University Hospital in Taiwan AB - Ten nonrepetitive clinical isolates of Klebsiella pneumoniae exhibiting an unusual inducible beta-lactam resistance phenotype were identified between January 1999 and September 2001 in a university hospital in Taiwan. In the presence of 2 mug of clavulanic acid, the isolates showed a one to four twofold concentration increase in the MICs of ceftazidime, cefotaxime, and aztreonam but remained susceptible to cefepime (MICs, <=0.5 mug/ml) and imipenem (MICs, <=0.5 mug/ml). PCR, sequence analysis, and isoelectric focusing revealed production by these isolates of TEM-1, SHV-11, and DHA-1, a plasmid-encoded inducible AmpC beta-lactamase originally found in a Salmonella enterica serovar Enteritidis strain. Transfer of the resistance by conjugation experiments was not successful, but Southern hybridization showed that blaDHA-1 was located on 70-kb plasmids, suggesting that the blaDHA-1-containing plasmids in the K. pneumoniae isolates were non-self-transmissible. Five isolates were recovered from patients in two surgery wards and two intensive care units. Acquisition of the DHA-1 producers could be traced back to previous hospitalizations 1 to 5 months earlier for the other five patients. Six and seven patterns among the isolates were demonstrated by plasmid analysis and ribotyping, respectively, indicating that the spread of the DHA-1 producers was due to both horizontal transfer of blaDHA-1 and dissemination of endemic clones. Keywords: Introduction : Chromosome-mediated AmpC beta-lactamases have been described in a wide variety of gram-negative bacilli, such as Pseudomonas aeruginosa and Enterobacter spp. . In most genera of the family Enterobacteriaceae, AmpC is inducible and, when overexpressed, can confer resistance to both oxyimino- and 7-alpha-methoxy-cephalosporins and monobactams . Many plasmid-mediated AmpC enzymes, such as CMY-type beta-lactamases, have been found in bacterial species that naturally lack a chromosomal AmpC beta-lactamase, such as Klebsiella pneumoniae, Proteus mirabilis, and Salmonella spp. (, -, -, , , ). It is believed that such beta-lactamases arose through the transfer of chromosomal AmpC genes onto plasmids . Unlike chromosome-mediated AmpC, plasmid-encoded AmpC enzymes are almost always expressed constitutively (-, -, , ). Plasmid-mediated inducible beta-lactamases are extremely rare. DHA-1 from a clinical isolate of Salmonella enterica serovar Enteritidis from Saudi Arabia is the first identified plasmid-encoded inducible cephalosporinase . The counterpart of blaDHA-1 was the chromosomal AmpC gene of Morganella morganii . The inducibility of DHA-1 is due to the presence of a regulator ampR gene, which is also related to that of M. morganii, upstream of blaDHA-1 on the same plasmid . A DHA-1-related beta-lactamase, named DHA-2, was identified more recently from a K. pneumoniae isolate in France . The enzyme also confers an inducible beta-lactam resistance phenotype. Recently, the standard confirmatory test for the detection of extended-spectrum beta-lactamases revealed an unusual ceftazidime and cefotaxime resistance phenotype in clinical isolates of K. pneumoniae in a university hospital in Taiwan. Thus, a retrospective analysis was carried out to characterize these isolates and their various clinical and epidemiological features. We found inducible expression of DHA-1 by these isolates. To our knowledge, this is the first report of the appearance of DHA-1 in the Far East and is also the first report of fairly widespread of DHA-1-producing K. pneumoniae within a health care institution. MATERIALS AND METHODS : Bacterial isolates and patients. | The standard screening and confirmation methods for the detection of extended-spectrum beta-lactamases were routinely performed at the Department of Pathology, National Cheng Kung University Hospital, a 900-bed teaching hospital in southern Taiwan. Between January 1999 and September 2001, 10 nonrepetitive isolates of K. pneumoniae from 10 patients demonstrated reduced inhibition zone diameters for both ceftazidime and cefotaxime in combination with clavulanic acid versus those for ceftazidime and cefotaxime when tested alone (see Table ), suggesting production of beta-lactamases induced by clavulanic acid. All these isolates were identified by conventional techniques and/or the API 20E system (bioMerieux, Marcy l'Etoile, France). The medical records of the patients from whom the isolates were recovered were reviewed. Antagonism testing. | The disk antagonism method initially used to detect inducibility of chromosomal beta-lactamases was performed with a slight modification to test the 10 K. pneumoniae isolates. Disks of inducing agents and disks of cephalosporins were placed on the surface of Mueller-Hinton agar plates and separated by 25 mm (see Fig. ). The cephalosporins used were cefotaxime, ceftazidime, aztreonam, and cefepime. Clavulanic acid (10 mug) and cefoxitin (30 mug) were used as inducing agents. The plates were examined after overnight incubation at 37C. Susceptibility testing. | MICs were determined by the standard agar dilution method . The antimicrobial agents and their sources were as follows: amoxicillin and clavulanic acid, SmithKline Beecham Pharmaceuticals, Surrey, United Kingdom; aztreonam and cefepime, Bristol-Myers Squibb, New Brunswick, N.J.; cefotaxime, Hoechst-Roussel Pharmaceuticals, Inc., Somerville, N.J.; cefoxitin, Sigma Chemical Company, St. Louis, Mo.; ceftazidime, Glaxo Group Research Ltd., Greenford, United Kingdom; and imipenem, Merck Sharp & Dohme, West Point, Pa. The susceptibilities to six non-beta-lactam agents were determined by the standard disk diffusion method . Antimicrobial disks were obtained from Becton Dickinson Microbiology Systems, Cockeysville, Md., including amikacin, ciprofloxacin, gentamicin, ofloxacin, tobramycin, and trimethoprim-sulfamethoxazole. IEF. | Crude preparations of beta-lactamases were obtained from the isolates by sonication and subjected to analytical isoelectric focusing (IEF) as described previously . Cells induced by 16 mug of cefoxitin per ml were incubated for 3 h before harvesting . beta-Lactamase activity was detected by overlaying the gels with 0.5 mM nitrocefin in 0.1 M phosphate buffer, pH 7.0. PCR and DNA sequencing. | Plasmids from the isolates were extracted by a rapid alkaline lysis procedure and used as templates in PCRs. The entire blaDHA-1 gene was amplified with the oligonucleotide primers DHA-1A (5'-CTGATGAAAAAATCGTTATC-3') and DHA-1B (5'-ATTCCAGTGCACTCAAAATA-3'), corresponding to nucleotides -3 to 17 and 1138 to 1119, respectively, of the DHA-1 structural gene . The PCR conditions were as follows: 3 min at 94C; 35 cycles of 1 min at 94C, 1 min at 55C, and 1 min at 72C; and finally 7 min at 72C. The entire sequences of blaTEM- and blaSHV-related genes were amplified with the primer pairs as described previously . The amplicons were purified with a commercial kit (Roche Molecular Biochemicals, Mannheim, Germany) and sequenced on an ABI Prism 310 sequencer analyzer (Applied Biosystems, Foster City, Calif.). Transfer of resistance. | Conjugation experiments were performed as described previously with streptomycin-resistant Escherichia coli C600 as the recipient . Tryptic soy agar plates supplemented with 500 mug of streptomycin (Sigma) per ml and 64 mug of cefoxitin per ml were used to select the transconjugants. K. pneumoniae strain W142 harboring blaCMY-8 was used as the control . Plasmid analysis and Southern hybridization. | Plasmids from the isolates were analyzed by electrophoresis on a 0.8% agarose gel. E. coli strain NCTC 50192 (National Collection of Type Cultures, London, United Kingdom), which contained four plasmids of 7, 36.3, 63.8, and 148.5 kb, was used as a source of molecular size markers. The gel was stained with ethidium bromide (Sigma), visualized under UV light, and subjected to Southern hybridization according to the original protocol . The blaDHA-1-specific probe was a PCR-generated amplicon labeled with [alpha-32P]dCTP (Amersham Pharmacia Biotech) by the random priming technique with a commercial kit (Gibco-BRL Life Technologies, Gaithersburg, Md.). Ribotyping. | The chromosomal DNA of the isolates was extracted and purified as described previously . The genomic DNA was restricted with EcoRI or BstEII (Roche Molecular Biochemicals) . The digests of chromosomal DNA were electrophoresed at 35 V for 18 h in a 0.8% agarose gel, transferred to a nylon membrane (Amersham Pharmacia Biotech), and then hybridized with a [alpha-32P]dCTP-labeled cDNA copy of E. coli rRNA (Roche Molecular Biochemicals) obtained by reverse transcription with avian myeloblastosis virus reverse transcriptase (Gibco-BRL) as described previously . FIG. 1. | Double-disk antagonism tests with 30 mug of cefoxitin (A) and 10 mug of clavulanic acid (B). Double-disk antagonism tests with 30 mug of cefoxitin (A) and 10 mug of clavulanic acid (B). ATM, aztreonam; CAZ, ceftazidime; CLA, clavulanic acid; CTX, cefotaxime; FEP, cefepime; FOX, cefoxitin. TABLE 1 | Susceptibility patterns of DHA-1-producing K. pneumoniae isolates RESULTS : Inducibility of beta-lactamases. | In the standard extended-spectrum beta-lactamases confirmatory test, the reduced zone diameters for ceftazidime and cefotaxime in combination with clavulanic acid versus those for ceftazidime and cefotaxime tested alone among the 10 K. pneumoniae isolates ranged from 2 to 10 mm (mean, 5.6 mm) and 3 to 8 mm (mean, 5.6 mm), respectively, suggesting production of beta-lactamases induced by clavulanic acid . Inducibility of the beta-lactamases was further recognized by the disk antagonism test, which demonstrated blunting of the cephalosporin disks adjacent to the cefoxitin and clavulanic acid disks . Susceptibility testing. | The results of the susceptibility tests are shown in Table . All 10 isolates exhibited high-level resistance to amoxicillin-clavulanic acid and cefoxitin. In the presence of clavulanic acid, a one to four twofold concentration increase in the MICs of ceftazidime, cefotaxime, and aztreonam was noted, while the changes after addition of clavulanic acid in the MICs of cefepime and imipenem were not obvious. Identification of beta-lactamases. | IEF demonstrated that all 10 isolates displayed three bands of beta-lactamase activity with pIs of 5.4, 7.6, and 7.8. The pI 7.6 band probably represented the chromosomal SHV-1 or SHV-11 type beta-lactamase of K. pneumoniae , the pI 5.4 band might represent the TEM-1 beta-lactamase , and the pI 7.8 band might represent the beta-lactamase responsible for the inducible resistance phenotype. A 1,141-bp fragment was amplified by PCR with the blaDHA-1-specific primers for all 10 K. pneumoniae isolates. The amino acid sequences of the PCR products deduced from the sequence analysis were identical to the plasmid-mediated cephalosporinase DHA-1 from S. enterica serovar Enteritidis . The DHA-1 cephalosporinase was consistent with the pI 7.8 beta-lactamase demonstrated by IEF . All the isolates also carried blaTEM-1 and blaSHV-11, which were identified by PCR with the blaTEM- and blaSHV-specific primers and sequence analysis. Conjugation experiments and plasmid analysis. | Conjugation experiments failed to demonstrate transfer of inducible cephalosporin resistance from any of the isolates. Cefoxitin resistance was transferred from the control strain to E. coli C600 at a frequency of 10-3 to 10-4 per donor cell. Six different profiles were demonstrated by plasmid analysis among the 10 isolates . In all isolates analyzed, the presence of a plasmid of approximately 70 kb was detected. Southern hybridization with the blaDHA-1-specific probe showed that blaDHA-1 was located on the 70-kb plasmid (data not shown). Ribotyping. | The genetic relationship among the 10 K. pneumoniae isolates was investigated by ribotyping with two different endonucleases. Patterns with at least two discordant bands were considered different . The results are listed in Table and partially shown in Fig. . Both EcoRI and BstEII generated seven different patterns. Isolates 1490 and 1596, both of which were collected in early 2000, and isolates 197, 274, and 281, which were all collected in late 2000, had identical ribotypes, suggesting that they derived from two endemic clones. Clinical characteristics. | Three isolates were recovered from sputum samples and were considered colonizers. The other seven isolates were associated with three urinary tract infections, two wound infections, one intra-abdominal infection, and one bloodstream infection. The clinical characteristics of the patients infected with or colonized by the DHA-1-producing isolates are summarized in Table . Six patients had undergone hemodialysis due to either chronic or acute renal failure before isolation. Five isolates were obtained >48 h after the patients were admitted to the hospital. Three of the five isolates were from the patients in the surgery wards, and two isolates were from the patients in the intensive care units. Although the remaining five isolates were obtained within 48 h after the current admission, all patients with these isolates had been hospitalized in the teaching hospital 1 to 5 months earlier. Notably, four of them had been on hemodialysis either in the university hospital or at community hospitals. All nine patients for whom complete medical records were available had been exposed to beta-lactam agents within 2 weeks before isolation of the DHA-1 producers. Patients 3, 4, 6, and 8 received no specific antimicrobial agents for the DHA-1 producers, and patient 7 was not treated for the urinary tract infection at the university hospital. Patients 2 and 9 received ciprofloxacin and trimethoprim-sulfamethoxazole, respectively, for 2 weeks, and the K. pneumoniae strains were not isolated from urine samples afterward. Patient 1 had received cefotaxime and ciprofloxacin, but his deep soft tissue infection was not eradicated until debridement was performed 1 month after admission. Patient 5 received cefotaxime after isolation and died of intra-abdominal hemorrhages and multiorgan failure due to his underlying diseases 1 week later. Whether the bacterial peritonitis was persistent before his death is not clear. Patient 10 was cured of the bloodstream infection with 2 weeks of meropenem and netilmicin therapy. Five patients died during hospitalization, and patient 7 died during the other hospitalization; however, none of the deaths were directly due to infections caused by the DHA-1 producers. FIG. 2. | Plasmid profiles of the 10 K. pneumoniae Plasmid profiles of the 10 K. pneumoniae isolates. Lanes 1 to 10, isolates 387, 416, 1490, 1596, 197, 274, 281, 325, 397, and 1067, respectively; lane M, E. coli NCTC 50192, which contains four plasmids with molecular sizes of 7, 36.2, 63.8, and 148.5 kb. FIG. 3. | Ribotypes generated by EcoRI. Ribotypes generated by EcoRI. Lane 1, molecular size marker kit II (Roche Molecular Biochemicals); lanes 2 to 11, isolates 387, 416, 1490, 1596, 197, 274, 281, 325, 397, and 1067, respectively; lane 12, 1-kb ladder (Promega Co., Madison, Wis.). TABLE 2 | Clinical data, plasmid profiles, and ribotypes of the 10 DHA-1-producing K. pneumoniae isolates DISCUSSION : The plasmid-mediated inducible DHA-1 beta-lactamase was first identified in Taiwan in the present study. PCR, sequence analysis, and IEF revealed production of three beta-lactamases, TEM-1, SHV-11, and DHA-1, by all 10 K. pneumoniae isolates possessing inducible resistance to extended-spectrum beta-lactamases. TEM-1 and SHV-11 are restricted-spectrum beta-lactamases. Since DHA-1, originally found in an S. enterica serovar Enteritidis strain was inducible , the enzyme is believed to be responsible for the unusual inducible beta-lactam resistance phenotype of our isolates. Transfer of the resistance by conjugation experiments was not successful; however, the blaDHA-1-specific probe was hybridized to a 70-kb plasmid in all isolates analyzed, suggesting that blaDHA-1 was on non-self-transmissible plasmids. Similar results have been described in reports of DHA-1-producing S. enterica serovar Enteritidis and DHA-2-producing K. pneumoniae . To our knowledge, plasmid-mediated DHA-type beta-lactamases have only been reported in isolates from Europe and the Middle East . Thus, this is also the first report of the appearance of a DHA-type beta-lactamase in the Far East. In the extended-spectrum beta-lactamases confirmatory test, all K. pneumoniae isolates revealed decreased inhibition zone diameters for ceftazidime and cefotaxime in combination with clavulanic acid compared with those of these two agents tested alone, indicating that the test could also be used to screen for inducible beta-lactamase-producing gram-negative bacilli that naturally lack inducible chromosome-mediated AmpC enzymes. All 10 K. pneumoniae isolates showed blunting of the cephalosporin disks adjacent to the cefoxitin and clavulanic acid disks in the antagonism test, indicating that the test can also be used to recognize plasmid-mediated beta-lactamases. In the Enterobacteriaceae, AmpC-hyperproducing derepressed strains appear frequently in infections caused by organisms naturally producing inducible AmpC enzymes when patients are treated with extended-spectrum beta-lactams . Therefore, it has been recommended that the inducible-AmpC-producing Enterobacteriaceae species should be reported as resistant to all extended-spectrum beta-lactams . The use of extended-spectrum beta-lactams should be restricted accordingly. Studies on determining the therapeutic success or failure of extended-spectrum third-generation cephalosporins in treating infections with plasmid-mediated inducible AmpC producers, such as our DHA-1-producing K. pneumoniae isolates, are lacking. Therefore, whether such K. pneumoniae strains, like gram-negative organisms naturally producing inducible AmpC enzymes, should also be reported as resistant to all third-generation cephalosporins is not known and deserves further investigation. The drugs of choice for the treatment of infections with such organisms are also undetermined. Based on MIC data and the confirmatory test for extended-spectrum beta-lactamases, a majority of the DHA-1-producing K. pneumoniae isolates would not have been reported as resistant to all third-generation cephalosporins. However, after induction by clavulanic acid, these isolates showed reduced susceptibilities to these agents. Moreover, all these isolates remained susceptible to cefepime and imipenem even in the presence of clavulanic acid. Thus, fourth-generation cephalosporins and carbapenems could be better choices for the treatment of infections caused by DHA-1 producers. Alternatively, when the presence of inducible DHA-type enzymes is suspected or detected, physicians should be informed, and the use of strong AmpC-inducing agents, such as clavulanic acid and cephamycins, should be avoided. Six plasmid patterns and seven ribotypes were found among the 10 DHA-1-producing isolates (Fig. and ), indicating that the spread of blaDHA-1 was due to both dissemination of endemic clones and horizontal transfer of the resistance gene. Most isolates in the university hospital were obtained from surgery wards and intensive care units. Five isolates were obtained within 48 h after admission; however, all patients from whom the isolates were obtained had been hospitalized in the same university hospital 1 to 5 months before the current admissions. It is not known exactly whether these isolates were from the university medical center or other hospitals. However, since isolates 274 and 281 had a ribotype identical to that of isolate 197, which was obviously from the university hospital, it is very likely that at least patients 6 and 7 had acquired the resistance strain during previous hospitalizations. Six of the 10 patients infected with DHA-1 producers had been on hemodialysis. Since this was a retrospective study, it is not clear whether the nosocomial infections were associated with the hemodialysis systems. In conclusion, sporadic infections with K. pneumoniae possessing an unusual inducible beta-lactam resistance phenotype were found in a university hospital in Taiwan. DHA-1 encoded by non-self-transferable plasmids conferred the resistance phenotype. The spread of the DHA-1 producers was due to dissemination of endemic clones and horizontal transfer of the resistance gene. Backmatter: PMID- 12202591 TI - Genotypic Diversity of Clinical Actinomyces Species: Phenotype, Source, and Disease Correlation among Genospecies AB - We determined the frequency distribution of Actinomyces spp. recovered in a routine clinical laboratory and investigated the clinical significance of accurate identification to the species level. We identified 92 clinical strains of Actinomyces, including 13 strains in the related Arcanobacterium-Actinobaculum taxon, by 16S rRNA gene sequence analysis and recorded their biotypes, sources, and disease associations. The clinical isolates clustered into 21 genogroups. Twelve genogroups (74 strains) correlated with a known species, and nine genogroups (17 strains) did not. The individual species had source and disease correlates. Actinomyces turicensis was the most frequently isolated species and was associated with genitourinary tract specimens, often with other organisms and rarely with inflammatory cells. Actinomyces radingae was most often associated with serious, chronic soft tissue abscesses of the breast, chest, and back. Actinomyces europaeus was associated with skin abscesses of the neck and genital areas. Actinomyces lingnae, Actinomyces gravenitzii, Actinomyces odontolyticus, and Actinomyces meyeri were isolated from respiratory specimens, while A. odontolyticus-like strains were isolated from diverse sources. Several of the species were commonly coisolated with a particular bacterium: Actinomyces israelii was the only Actinomyces spp. coisolated with Actinobacillus (Haemophilus) actinomycetemcomitans; Actinomyces meyeri was coisolated with Peptostreptococcus micros and was the only species other than A. israelii associated with sulfur granules in histological specimens. Most genogroups had consistent biotypes (as determined with the RapID ANA II system); however, strains were misidentified, and many codes were not in the database. One biotype was common to several genogroups, with all of these isolates being identified as A. meyeri. Despite the recent description of new Actinomyces spp., 19% of the isolates recovered in our routine laboratory belonged to novel genospecies. One novel group with three strains, Actinomyces houstonensis sp. nov., was phenotypically similar to A. meyeri and A. turicensis but was genotypically closest to Actinomyces neuii. A. houstonensis sp. nov. was associated with abscesses. Our data documented consistent site and disease associations for 21 genogroups of Actinomyces spp. that provide greater insights into appropriate treatments. However, we also demonstrated a complexity within the Actinomyces genus that compromises the biochemical identification of Actinomyces that can be performed in most clinical laboratories. It is our hope that this large group of well-defined strains will be used to find a simple and accurate biochemical test for differentiation of the species in routine laboratories. Keywords: Introduction : Microbiological identification of Actinomyces to the species level is difficult in the clinical laboratory. Even with biochemical tests such as the RapID ANA II test, whole-cell fatty acid analysis, or gas-liquid chromatography, assignment to a species is difficult . Thus, based only on Gram staining, the catalase reaction, and better growth under anaerobic conditions than aerobic conditions, strains may be assigned to the genus Actinomyces. An excellent paper addressing this problem and seeking an easily accessible scheme for reliable differentiation at the species level has recently been published . It is not clear that assignment to a species is clinically important, although there have been some excellent previous reports of the frequencies of occurrence and site associations of Actinomyces species . Actinomycesneuii has been associated with abscesses; Actinomyces naeslundii has been reported to cause hip prosthesis infection ; and in recent studies of three newly described species, Actinomyces turicensis was reported to be associated with genital, skin-related, and urinary tract infections, whereas Actinomyces radingae was found only in skin-related infections and Actinomyces europaeus was detected in patients with urinary tract infections . Hall et al. examined a large collection of organisms and found that Actinomyces israelii and A. turicensis were most prevalent and were most commonly associated with intrauterine contraceptive devices. In textbooks, A. israelii is usually identified as the cause of actinomycosis, with other Actinomyces cited as causative agents in various unspecified diseases and as normal flora. However, when the species are distinguished, the data from different institutions show variations in species distributions and clinical associations. Sometimes this is because of the type of laboratory (public health, reference, or oral microbiology laboratory) or because of the relative proportion of isolates examined that are strict anaerobes . Thus, the frequency of occurrence of Actinomyces spp. in a routine clinical laboratory and their clinical significance are largely unknown. In the study described in this paper we characterized by genotypic and phenotypic methods a large group of presumptively significant isolates that were isolated in the routine laboratory but not as strict anaerobes. We also reviewed patient data. We thus wished to determine the frequency of occurrence and association with disease of the aerotolerant Actinomyces spp. isolated in a clinical laboratory which were accurately identified by 16S ribosomal DNA (rDNA) sequence analysis. MATERIALS AND METHODS : Organisms. | We sequenced 100 presumptively clinically significant strains of putative Actinomyces spp. that were recovered in a routine clinical laboratory on Trypticase soy agar with sheep blood or Columbia colistin-nalidixic acid agar (BBL, Becton Dickinson, Cockeysville, Md.) by incubation at 35C with CO2 (7%) added but without the use of special anaerobic conditions. Of these strains, 15 strains were not in the Actinomyces cluster and were not further studied; most were in the Bifidobacterium and Gardnerella groups, with at least one strain each of the genera Streptococcus, Erysipelothrix, and Abiotrophia. Thirteen strains were in the related Arcanobacterium-Actinobaculum taxon and were included in the study. In addition, we identified by 16S rDNA sequence analysis six strains as Actinomyces spp. that were originally identified as Lactobacillus, Streptococcus, or coryneform organisms. Thus, we analyzed 91 strains: 82 strains were obtained from the Microbiology Laboratory of the Houston Veterans Affairs Medical Center (VAMC) and 9 strains were from other clinical laboratories and were referred to the Houston VAMC for identification. All strains were stored frozen at -70C. During the periods when we were most alerted to this group of organisms, the noted occurrence was about 15 to 20 cases per year. When available, patient records were reviewed for the clinical significance of the isolates. We made the assumption of possible significance if the isolate had one of the follow features: (i) was cultured from a normally sterile site, (ii) was the predominant organism or a copredominant organism from a wound infection or abscess, (iii) was found from a urine specimen culture at >104 CFU/ml with no more than one other isolate at >104 CFU/ml, or (iv) was the predominant organism isolated from a purulent sputum specimen. A careful review of medical notes was used to confirm clinical significance. The presence of polymorphonuclear leukocytes (PMNs), the use of surgical drainage, and a record of antibiotic treatment were primary indications of clinical significance. Although gram-positive branching rods were frequently noted on the original Gram stain, at times what were later shown on review to be Actinomyces were interpreted as gram-positive cocci in short chains. Biochemical identification. | All strains were grown at 35 to 37C on sheep blood agar plates (Remel, Lenexa, Kans., and BBL, Becton Dickinson) with CO2 added. Presumptive phenotypic identification was performed by Gram staining, evaluation of the colony morphology, and catalase reaction. None of the isolates was a strict anaerobe; all grew in an atmosphere with elevated CO2 concentrations (5 to 8%). Biochemical testing was performed with RapID ANA II identification kits (RapID ANA; Remel, Inc., Norcross, Ga.). The interpretation of tests was done according to the manufacturer's instructions. 16S rDNA sequence analysis. | 16S rRNA gene sequence identification was performed at the Houston VAMC laboratory and MIDI Labs (Newark, Del.) with the MicroSeq 500 gene kit (Applied Biosystems, Foster City, Calif.) according to the specifications of the manufacturer. Approximately 500 bp in both the forward and the reverse senses was sequenced for each isolate. Test strain sequences were compared with sequences in both the MicroSeq and the GenBank 16S rRNA gene sequence databases. The MicroSeq database contains sequences from 1,297 different species (1,187 type strains), including 9 type strains from the genus Actinomyces. Sequence data obtained from the strains in GenBank were included in the analysis. Sequences were compared in dendrogram form by using the neighbor-joining method (J. E. Clarridge, Q. Zhang, and S. Heward, Abstr. 101st Gen. Meet. Am. Soc. Microbiol., abstr., C-42, p. 157, 2001). Nucleotide sequence accession numbers. | The partial 16S rDNA sequence of the type strain of A. houstonensis Houston VAMC 3971 is deposited in GenBank under accession no. . To help compare isolates, we are depositing in GenBank one 500-bp sequence from each of the genospecies 2, 6, 13, and 17 (strains VAMC Ref113, VAMC 3971, VAMC Ref103, and S3672, respectively) as AF457642, AF457638, AF457641, and AF457640, respectively. RESULTS : Genotypic identification and phylogenetic relationships. | The dendrogram based on gene sequencing shows the overall relatedness of the Actinomyces spp. We show the type strains of 21 Actinomyces spp. and 4 related type strains (11 from the MicroSeq database, 13 from the GenBank database, and a challenge strain from the College of American Pathologists [CAP]) and representative clinical strains for each of the major groups. To present the sequences in a single comparative dendrogram, three or more clinical isolates that differed by no more than 5 bp are represented by a single entry. We note the number of isolates in the group; e.g., in genospecies 5 (A. europaeus), there are seven clinical strains. Because the Actinomyces odontolyticus-Actinomyces meyeri group was so diverse, the isolates are presented individually in Fig. . The 91 strains were clustered into 18 major genogroups and 3 minor genogroups (Fig. and Fig. ). Table shows the distribution of species by the number of isolates in each genogroup, designated as in the dendrogram in Fig. . Table shows the same type of data for the A. meyeri-A. odontotyticus groups, as shown in Fig. , except that the data are presented separately for each strain of this diverse group. Twelve genogroups correlated with a known species: A. turicensis (23 strains), A. radingae (13 strains), A. europaeus (7 strains), A. lingnae (4 strains), Actinomyces graevenitzii (4 strains), A. neuii (3 strains), A. odontolyticus (4 strains) A. meyeri (3 strains), A. israelii (3 strains), Arcanobacterium haemolyticum (9 strains), and 1 strain each of Actinomyces funkii and Arcanobacterium bernardiae. These are so designated in both Tables and 2. Nine genogroups (17 strains; 19%) did not correlate closely with a known type strain. For these, the closest strain with at least 95% similarity is listed and the genogroup is called "novel." The most common sources and associated diseases are listed. Phenotypic characterization. | The biochemical reactions obtained with the RapID ANA II system, coded as a biotype number, for the species sequenced are also presented in Tables and 2. Although the biochemical identifications with this system frequently indicate an inaccurate species name (as indicated in the tables), many of the different genogroups show distinctive and reproducible biochemical profiles. For example, 4 A. europaeus strains had similar code numbers (421670, 421671, and 421070), 2 isolates of genogroup 10 generated similar codes (671671 and 671470), 18 of the 21 A. turicensis strains tested generated the same code (020671), and most of the A. radingae isolates generated a code similar to 677671 . The same may be true for other genogroups, but due to the limited number of isolates in these other groups, we do not believe that the data are sufficiently robust for reliable identification. The full description of the biochemical significance of the biotype numbers is presented in the literature accompanying the RapID ANA II system. Briefly, to summarize some important differences found in this study: most isolates shared the last three digits of 671 in their profile numbers, which indicates positive reactions for the cleavage of leucyl-glycine, glycine, proline, phenylalanine, arginine, and serine and negative reactions for alkaline phosphate, pyrrolidine, and indole. A. turicensis was usually additionally positive only for hydrolysis of aryl-substituted alpha-glucoside. A. europaeus was positive for carbohydrate hydrolysis of aryl-substituted arabinoside, alpha-glucoside, and galactoside. Genogroup 10 was additionally positive for beta-disaccharide, o-nitrophenyl-beta-d-galactopyranoside, and beta-glucoside. It is clear that, according to the RapID ANA II system database, many genospecies would be identified as A. meyeri (e.g., genospecies 2, 6, and 15). All strains except A. neuii and genospecies 9 were negative for catalase; A. neuii has previously been reported to be positive for catalase, and genospecies 9, a strain from gallbladder fluid with an unusual biotype profile, which we report here for the first time, was positive for catalase. Genospecies 9 and 10 cluster with the Actinomyces viscosus and A. naeslundii groups, which are variable for catalase . Most isolates in the A. odontolyticus clusters were identified as either A. odontolyticus or A. meyeri by the RapID ANA II system. Sources and disease associations. | Table summarizes the sources and disease associations of the more common Actinomyces species and genospecies. A. turicensis was the most frequently isolated species and was associated with genitourinary tract specimens, often with other organisms and only rarely with inflammatory cells. In a few instances A. turicensis was isolated as a pure culture: from two urine specimens, one specimen from a wound, and one specimen from a breast abscess. It was the only Actinomyces sp. isolated from urine and urethral exudates. A. radingae was usually associated with abscesses; almost all of these were serious, chronic soft tissue abscesses of the breast, chest, and back. A. radingae was often recovered in large numbers and was usually associated with PMNs. A. europaeus was associated with soft tissue abscesses of the neck and genital areas. The newly described species A. lingnae was found in low numbers in respiratory specimens from compromised hosts. Both A. israelii and A. meyeri were associated with pulmonary actinomycosis and sulfur granules in histological specimens. A. israelii was the only Actinomyces sp. coisolated with Actinobacillus (Haemophilus) actinomycetemcomitans. A. meyeri was coisolated with Peptostreptococcus micros. Because these two coisolates are well-known pathogens in their own right, it is interesting to speculate on their contribution to disease. In Table we present our deductions of the normal niches of the various Actinomyces spp. by correlating both the site of isolation (indicated in Tables , 3, and 4) and the known niche of the common coisolates. A. radingae and A. europaeus were isolated with low numbers of coagulase-negative Staphylococcus spp. and Corynebacterium spp., suggesting skin contamination and that the niche for these organisms is the skin. A. lingnae, A. graevenitzii, A. odontolyticus, A. meyeri, and genospecies 10 are associated with respiratory specimens. The two strains in genospecies 10 were cultured from a mandible and a face, and their sequences had 98% similarity to an unnamed oral clone of Actinomyces sp. from GenBank (accession no. ). Streptococcus intermedius and Eikenella corrodens were also found in these cultures, further suggesting an oral source. The A. odontolyticus-like strains (genospecies 17a to 17c) were isolated from heterogeneous sources, some of which suggest hematogenous spread and a genitourinary or gastrointestinal source. In both instances in which two different Actinomyces spp. were recovered in the same specimen, the presumed niches agreed: the respiratory site-associated species, A. graevenitzii and A. lingnae, were recovered together, both in low numbers, from bronchial wash specimens; and the skin- and abscess-associated species, A. europaeus and A. radingae, were recovered together from a back abscess. Two genetically well-separated genospecies, genospecies 6 and 13, had multiple strains. Below we describe genospecies 6 as A. houstonensis sp. nov. FIG. 1. | Dendrogram showing the genospecies (GS) clusters based on our sequence data for clinical isolates and some Actinomyces type strains from MicroSeq and GenBank databases with Bifidobacterium boum as an outgroup. Dendrogram showing the genospecies (GS) clusters based on our sequence data for clinical isolates and some Actinomyces type strains from MicroSeq and GenBank databases with Bifidobacterium boum as an outgroup. All type strains are represented by the name written out in full. We note the type strain sequences from GenBank by GB and the number of base pairs that were used or available for comparison. The single strain from CAP has the suffix CAP, and the strains from the MicroSeq database do not have a suffix. FIG. 2. | Dendrogram showing the genospecies clusters based on our sequence data for clinical isolates within the A. odontolyticus Dendrogram showing the genospecies clusters based on our sequence data for clinical isolates within the A. odontolyticus-A. meyeri branch. In addition to all 14 clinical strains and the A. meyeri type strain, we show one A. odontolyticus type strain and two other sequences that were called A. odontolyticus in GenBank but which have 16S rDNA sequences different from that of the A. odontolyticus type strain. A. turicensis is the outgroup. TABLE 1 | Characteristics of the genogroups of clinically isolated Actinomyces strains TABLE 2 | Characteristics of genogroups 17 and 18, A. odontolyticus, and A. meyeri strains TABLE 3 | Site of isolation or infection associated with Actinomyes spp. for which there were at least three strains TABLE 4 | Discovery of the probable normal niche of Actinomyces spp. approached by using the known niche of specific coisolates and the sources of the Actinomyces strains in this investigation DISCUSSION : Using 16S rRNA gene sequencing to identify clinically derived Actinomyces spp., we determined the frequency distributions of species isolated in a single laboratory and their site and disease associations. As in some other studies, A. turicensis was the most commonly isolated species . However, A. radingae and A. europaeus were the second and third most frequently isolated species, respectively, in contrast to A. israelii and A. naeslundii, respectively, in the study by Hall et al. . Sabbe et al. found twice as many A. europaeus isolates as A. radingae isolates. A. turicensis was most commonly associated with the urinary tract infections and skin-related infections of the lower body. In contrast to Funke et al. , some of the A. turicensis strains were isolated as pure cultures, and several caused abscesses. A. radingae and A. europaeus were associated with soft tissue infections. Our data show a striking correlation of A. radingae with recurrent abscesses of the chest, back, and breast. Infections due to A. europaeus were also found in the genital area, such as scrotal abscesses and labia abscesses. A. meyeri and A. israelii were both associated with actinomycosis, sulfur granules, and a distinct coisolate. The RapID ANA II system biochemical tests for A. turicensis, A. radingae, and A. europaeus were reproducible, yielding codes that were useful for identification; however, the code numbers either corresponded to another organism or were not in the database. The same biotype for A. turicensis was noted by Sabbe et al. . Our data confirm the revised description by Vandamme et al. that A. radingae strains are positive for N-acetylglucosamine and beta-glucosidase, while A. turicensis strains are negative. Our biochemical data do not support the identification scheme proposed by Sarkonen et al. , who used reagents from a different manufacturer. In contrast to biochemical identification, which, as we have shown here, might be ambiguous, 16S rRNA sequence analysis assigns an unknown strain to a reproducible genocluster. Because we sequenced all isolates in a cluster, we were able to discern that some genogroups are more heterogeneous than others. For example, organisms in the A. graevenitzii and A. viscosus-A. naeslundii clusters were genetically heterogeneous, as was also determined from the data of Hall et al. . The heterogeneity in the A. odontolyticus-like groups allowed us to distinguish one group that seemed to have a gastrointestinal source. In contrast, genogroups that showed minimal variations in their 16S rDNA sequences and that were closely related to the type strain are A. turicensis, A. radingae, A. europaeus, and the newly described species A. lingnae. Nine genogroups did not have a corresponding known type strain at the time of submission of the manuscript. However, as many new sequences are being deposited and new species are being described every day, it is probable that other investigators will find similar strains. At present, clinical microbiologists and infectious disease specialists should be cautious in their acceptance of an identification as an Actinomyces species by testing that is usually performed in most clinical laboratories. Case reports based on identifications achieved prior to the use of 16S rDNA-based techniques may not be accurate. Our data document consistent site and disease associations for 21 genogroups of Actinomyces spp. that provide greater insights into the clinical relevance of the genogroups. However, we also demonstrate a complexity within the genus Actinomyces that compromises the biochemical identification of Actinomyces that can be performed in most clinical laboratories. It is our hope that this large group of well-defined strains will be used to find simple and accurate biochemical tests for differentiation of the species in routine laboratories. Description of Actinomyces houstonensis sp. nov. | Actinomyces means ray fungus for the shape of the microcolonies; houstonensis, in honor of Houston, Texas, indicates the place where the bacterium was identified and described. A. houstonensis is facultatively anaerobic and grows on sheep blood agar as alpha-hemolytic, gray colonies 0.2 mm in diameter after 48 h of incubation at 36C with elevated CO2 concentrations (8%). Growth is equal with elevated CO2 and anaerobic conditions. By Gram stain, the cells are nonsporulating, gram-positive pleomorphic rods, which are more robust than those of A. meyeri, and have a tendency to form half circles. The organism is nonmotile and catalase negative. It is esculin, urease, and gelatin test negative. It reduces nitrate and produces positive reactions for the cleavage of aryl-substituted alpha-glucoside, leucyl-glycine, glycine, proline, phenylalanine, arginine, and serine and negative reactions for alkaline phosphate, pyrrolidine, and indole. It ferments glucose and sucrose but not xylose. The type strain was isolated from an abscess from the back of a patient. All three strains were associated with serious subcutaneous abscesses requiring drainage; two strains were associated with other organisms that tend to be associated with the gastrointestinal tract or skin. The type strain of A. houstonensis is Houston VAMC strain 3971 and is deposited in GenBank under accession no. . Backmatter: PMID- 12202543 TI - Replacement of Candida albicans with C. dubliniensis in Human Immunodeficiency Virus-Infected Patients with Oropharyngeal Candidiasis Treated with Fluconazole AB - Candida dubliniensis is an opportunistic yeast that has been increasingly implicated in oropharyngeal candidiasis (OPC) in human immunodeficiency virus (HIV)-infected patients but may be underreported due to its similarity with Candida albicans. Although most C. dubliniensis isolates are susceptible to fluconazole, the inducibility of azole resistance in vitro has been reported. Thus, the use of fluconazole prophylaxis in the treatment of these patients may have contributed to the increasing rates of isolation of C. dubliniensis. In this study, yeast strains were collected from the oral cavities of HIV-infected patients enrolled in a longitudinal study of OPC. Patients received fluconazole for the suppression or treatment of OPC, and isolates collected at both study entry and end of study were chosen for analysis. Samples were plated on CHROMagar Candida medium for initial isolation and further identified by Southern blot analysis with the species-specific probes Ca3 (for C. albicans) and Cd25 (for C. dubliniensis). Fluconazole MICs were determined by using NCCLS methods. At study entry, susceptible C. albicans isolates were recovered from oral samples in 42 patients who were followed longitudinally (1 to 36 months). C. albicans strains from 12 of these patients developed fluconazole resistance (fluconazole MIC, >=64 mug/ml). C. dubliniensis was not detected at end of study in any of these patients. Of the remaining 30 patients, eight (27%) demonstrated a replacement of C. albicans by C. dubliniensis when a comparison of isolates obtained at baseline and those from the last culture was done. For the 22 of these 30 patients in whom no switch in species was detected, the fluconazole MICs for initial and end-of-study C. albicans isolates ranged from 0.125 to 2.0 mug/ml. For the eight patients in whom a switch to C. dubliniensis was detected, the fluconazole MICs for C. dubliniensis isolates at end of study ranged from 0.25 to 64 mug/ml: the fluconazole MICs for isolates from six patients were 0.25 to 2.0 mug/ml and those for the other two were 32 and 64 mug/ml, respectively. In conclusion, a considerable number of patients initially infected with C. albicans strains that failed to develop fluconazole resistance demonstrated a switch to C. dubliniensis. C. dubliniensis in this setting may be underestimated due to lack of identification and may occur due to the impact of fluconazole on the ecology of oral yeast species. Keywords: Introduction : Candida dubliniensis is a causative agent of oropharyngeal candidiasis (OPC) in human immunodeficiency virus (HIV)-infected and AIDS patients (10, 34; D. C. Coleman, D. J. Sullivan, and J. M. Mossman, Letter, J. Clin. Microbiol. 35:3011-3012). It is a newly identified species of Candida, phenotypically similar to (chlamydospore and germ tube positive), but genetically distinct from, Candida albicans (, , 36). Indeed, C. dubliniensis was initially difficult to distinguish from C. albicans and was often misidentified as such in standard clinical laboratory tests, because it shares phenotypic characteristics with C. albicans . Thus, due to the phenotypic similarity of the two organisms, there is a growing body of evidence suggesting that many previous cases of OPC thought to have been caused by C. albicans were in reality caused by C. dubliniensis . Since its identification as an etiologic agent of OPC, C. dubliniensis has received increasing attention. Although it was initially implicated in OPC in HIV-infected patients, particularly those with recurrent infections, C. dubliniensis has since been found to be both an oral carriage organism and a causative agent of oral candidiasis in HIV-negative individuals (-, , , -, , , , , -; Coleman et al., letter). The interaction and ecological relationships of different Candida species in the oropharynx are not completely understood and appropriately constitute a topic of active research. In a recent study that examined the growth competition between C. albicans and C. dubliniensis under broth and biofilm growth conditions, it was shown that C. albicans had a competitive advantage over C. dubliniensis . The clinical significance of this organism is yet to be determined, but it has been demonstrated that, contrary to fluconazole (FLU) resistance in C. albicans, FLU resistance in C. dubliniensis is easily inducible in vitro . Thus, prolonged use of FLU in HIV-infected patients may impact the epidemiology of OPC. The present study was undertaken to further define the complex microbial ecology in the oropharynxes of HIV-infected patients exposed to FLU for the treatment of OPC and to increase our understanding of the role of C. dubliniensis in OPC. MATERIALS AND METHODS : Clinical samples. | Yeast isolates were recovered from swabs and/or oral rinses taken from the oropharynxes of HIV-infected patients enrolled in a longitudinal study to assess the development of FLU resistance. Sixty-four patients were enrolled and monitored clinically for the presence of OPC. Cultures of the oropharynx were obtained at the time of enrollment and either during times of clinical OPC or at quarterly intervals if OPC was not present. All patients were treated initially with FLU at a dose of 100 mg/day. Doses of FLU were increased up to 800 mg/day in those individuals who developed clinical resistance, which is defined as the clinical requirement for increasing FLU doses for response (, -). All oral samples were screened with the chromogenic medium CHROMagar Candida (CHROMagar Company, Paris, France) and initially identified as C. albicans by colony color and a positive germ tube test in serum. Selected yeast isolates included those obtained on both initial and final cultures from these patients. Susceptibility testing. | Candida isolates were tested for susceptibility to FLU in accordance with NCCLS methods . Candida isolates were considered susceptible (MIC, 8 mug/ml), susceptible dose dependent (MIC, 16 to 32 mug/ml), or resistant (MIC, >=64 mug/ml). All antifungal susceptibility testing was performed at the Fungal Testing Laboratory, University of Texas Health Science Center San Antonio, San Antonio. DNA typing techniques. | Electrophoretic karyotype and fingerprinting analyses of EcoRI-digested chromosomal DNA with the repetitive probes Ca3, specific for C. albicans , and Cd25, specific for C. dubliniensis , were performed as previously described . Hybridization with probes Ca3 and Cd25 was also used as a genotypic tool to confirm the identity of the isolates as C. albicans or C. dubliniensis, respectively. In brief, chromosomal DNA from each isolate was prepared in agarose plugs and separated by pulsed-field gel electrophoresis (Bio-Rad, Hercules, Calif.). Restriction fragment length polymorphism patterns were obtained by digestion of chromosomal DNA by using SfiI and EcoRI (Boehringer-Mannheim, Indianapolis, Ind.). Digested DNA present in the restriction fragment length polymorphism gels was transferred to nylon membranes (Nytran; Schleicher and Schuell, Keene, N.H.) and hybridized with probe Ca3 and subsequently with probe Cd25 (after removal of the bound probe Ca3), radioactively labeled by random priming (Random Primers DNA labeling system; GibcoBRL, Gaithersburg, Md.), and exposed to autoradiography film (Du Pont, Wilmington, Del.) overnight at -70C. RESULTS : Of the 64 patients enrolled in the study, initial cultures of the oropharynx specimens from 42 demonstrated FLU-susceptible C. albicans isolates. Of these 42 patients, 12 developed FLU-resistant (FLU MIC, >=64 mug/ml) C. albicans on their final cultures (species identity for all resistant isolates was confirmed at the DNA level) while 30 patients did not. Upon further evaluation of all final isolates from these 30 patients, eight patients whose initial cultures had demonstrated a FLU-susceptible C. albicans isolate showed a change to a C. dubliniensis isolate at end of study, as shown in Table . C. dubliniensis isolates were susceptible to FLU at MICs ranging from 0.25 to 64 mug/ml. The FLU MICs for six isolates were 0.25 to 2.0 mug/ml, while those for two other isolates were 32 and 64 mug/ml, respectively. Strain delineation of isolates was performed by electrophoretic karyotype analysis, as shown in Fig. . All initial isolates demonstrated chromosomal banding patterns consistent with that of C. albicans. These patterns, however, were distinctly different from those obtained for the end-of-study, or final, isolates. Final isolates (Fig. , lanes B) all demonstrated a distinct low-molecular-weight band that was not found for any of the initial isolates, which is characteristic of C. dubliniensis. Fingerprinting analysis with the probes specific for C. albicans (Ca3) and C. dubliniensis (Cd25) are shown in Fig. and , respectively. As shown in Fig. , all initial isolates hybridized with Ca3, as demonstrated by autoradiography, while none of the final isolates demonstrated hybridization activity. In Fig. , however, the converse is true. None of the initial isolates hybridized with Cd25, while each of the final isolates demonstrated reactivity. These hybridization patterns showed that all initial isolates were C. albicans and all final isolates were C. dubliniensis. FIG. 1. | Electrophoretic karyotype analysis results for Candida isolates from eight patients in whom the culture isolate changed from C. albicans Electrophoretic karyotype analysis results for Candida isolates from eight patients in whom the culture isolate changed from C. albicans to C. dubliniensis. Patient numbers are given at the top of the figure. Lanes: A, initial isolates; B, final isolates. FLU MICs are in micrograms per milliliter. FIG. 2. | DNA fingerprinting analysis of EcoRI-digested chromosomal DNA with the radiolabeled Ca3 probe, which is specific for C. albicans. DNA fingerprinting analysis of EcoRI-digested chromosomal DNA with the radiolabeled Ca3 probe, which is specific for C. albicans. Patient numbers are given at the top of the figure. Lanes: A, initial isolates; B, final isolates. FLU MICs are in micrograms per milliliter. FIG. 3. | DNA fingerprinting analysis of EcoRI-digested chromosomal DNA with the radiolabeled Cd25 probe, which is specific for C. dubliniensis. DNA fingerprinting analysis of EcoRI-digested chromosomal DNA with the radiolabeled Cd25 probe, which is specific for C. dubliniensis. Patient numbers are given at the top of the figure Lanes: A, initial isolates; B, final isolates. FLU MICs are in micrograms per milliliter. TABLE 1 | Initial and end-of-study culture isolates from 30 patients and their FLU susceptibilities DISCUSSION : The increased use of antifungal agents (mainly FLU) may be responsible for the development of more highly resistant microorganisms and may impact the ecology of oral yeast species by effecting a shift toward species with innate resistance or those that are more prone to develop resistance to drugs such as FLU. Although most C. dubliniensis clinical isolates are susceptible to azole derivatives , our group and others have reported the isolation of FLU-resistant C. dubliniensis clinical isolates from the oral cavities of HIV-infected patients with OPC and prior exposure to FLU . In addition, derivatives exhibiting a stable FLU-resistant phenotype can be readily generated in vitro from FLU-susceptible isolates , a phenomenon that is not easily observed with C. albicans, thus indicating that C. dubliniensis may display a higher propensity to develop azole resistance than C. albicans. Our results indicate that replacement of C. albicans by C. dubliniensis occurred in a considerable number (27%) of patients treated with FLU who failed to develop FLU-resistant C. albicans. Of note, virtually all of these isolates were initially identified incorrectly as C. albicans by the phenotypic methods routinely used in our laboratory, and only subsequent use of genotypic methods (karyotyping and Cd25 fingerprinting) allowed correct identification of the isolates as C. dubliniensis. For the patients in the present study, the FLU MICs for initial and end-of-study C. albicans isolates from those who did not exhibit a change in species (n = 22) ranged from 0.125 to 2.0 mug/ml whereas the FLU MICs for C. dubliniensis isolates from end-of-study cultures for eight patients ranged from 0.25 to 64 mug/ml. Interestingly, replacement of C. albicans by C. dubliniensis was not observed in patients whose C. albicans strains were able to develop FLU resistance. These results suggest that the antifungal pressure exerted by prolonged FLU treatment influences the oral microbial ecology in these patients. Species that are better able to adapt to the antifungal pressure may persist over those that are effectively suppressed by the treatment. However, persistence of oral microorganisms as commensal or infectious organisms may also depend on other factors not related to resistance. Indeed, the majority of C. dubliniensis isolates recovered at the end of study did not demonstrate decreased susceptibility to FLU . We have previously demonstrated that, in the absence of antifungal pressure, C. albicans predominates over C. dubliniensis in normal growth competition assays. However, adherent populations of C. dubliniensis were able to withstand the competitive pressure of C. albicans . Some factors that may also favor the survival of C. dubliniensis in the oral cavity are its intrinsic cell surface hydrophobicity, its ability to bind to oral bacteria, and its ability to form biofilms . The presence of subinhibitory concentrations of FLU may alter these properties. These factors, together with better identification techniques, may be responsible for the increasing rates of isolation of this newly described species. In summary, in our longitudinal study of FLU resistance in HIV-infected patients, for a considerable number of patients initially infected with C. albicans strains that failed to develop FLU resistance (more than one-fourth), later cultures showed a change in species isolated to C. dubliniensis. Identification of C. dubliniensis required the use of genotypic methods. Thus, C. dubliniensis in this setting may be underestimated due to lack of proper identification and may occur, in part, because of the impact of FLU therapy on the ecology of oral yeast species. Backmatter: PMID- 12202612 TI - A Pilot with Pain in His Leg: Thigh Abscess Caused by Salmonella enterica Serotype Brandenburg AB - Salmonella enterica serotype Brandenburg is one of the more uncommon serotypes isolated from patients with gastroenteritis. Few cases of extraintestinal infections with serotype Brandenburg have been documented. The first case of a serotype Brandenburg-dependent thigh abscess originating from an atherosclerotic pseudoaneurysm of the femoral artery is reported. Keywords: CASE REPORT : During a medium-distance flight, a 57-year-old male Caucasian pilot suddenly experienced sharp pain in his right thigh followed by fever. The patient had hypercholesterolemia and non-insulin-dependent diabetes mellitus regulated by diet. In addition, he had previously experienced mild claudicatio intermittens in his right leg and had thus been prescribed acetylsalicylic acid. On the following day, he was seen by his general physician, who found no signs of deep venous thrombosis and prescribed diclofenac for symptomatic relief. During the subsequent week, the thigh swelling increased and the fever persisted. Five days after the initial symptoms, a tender nonfluctuating mass measuring 10 by 15 cm was found in the right thigh adjacent to the groin. A fine-needle puncture of the mass yielded blood only. A week later, magnetic resonance tomography of the thigh showed an extensive swelling surrounded by soft-tissue edema (Fig. and B). A pseudoaneurysm of the right femoral artery with extensive collateral circulation was identified by intravenous angiography . Empirical antibiotic therapy with difloxacillin (750 mg three times orally) was initiated. However, the patient's condition deteriorated with expansion of the soft tissue mass and persistent fever. The patient was admitted to the hospital. On admission, the leukocyte count was 16.6 x 10 9/liter with a predominance of neutrophils, the erythrocyte sedimentation rate and C-reactive protein level were 84 mm/h and 160 mg/liter, respectively, and the hemoglobin level was 124 g/liter. Liver and renal function tests were normal. Therapy with intravenously administered cefuroxime was initiated, after which the fever disappeared; however, the soft tissue mass remained unchanged. Three weeks after onset of symptoms, an abscess in the thigh muscles with close connection to the femur was incised and drained of pus. The patient's condition improved rapidly after the surgical intervention. Culture of pus from the abscess revealed growth of Salmonella species strain O4. Bone scintigraphy showed no pathological uptake in the femur or elsewhere in the skeleton. The patient was treated with oral levofloxacin for 3 months with no signs of recurrent infection. Interestingly, 2 weeks prior to the onset of disease, our patient had been to Tunisia, where he had experienced transient gastrointestinal discomfort without diarrhea or fever. The bacterial organism was isolated from the abscess fluid after 4 days of incubation using aerobic flasks with liquid medium (BacT/Alert; Organon Teknika). The isolate was subcultured onto supplemented human blood agar plates containing Columbia II agar, l-cysteine, hemin, and vitamin K1. Subtyping according to Kauffmann-White ( and references therein) using specific antisera revealed that the isolate was Salmonella enterica serotype Brandenburg with the O antigens 1, 4, and 12 and H antigens l, v (phase 1), e, n, and z15 (phase 2). The organism was susceptible to ampicillin, piperacillin, cefuroxime, cefotaxime, ceftazidime, imipenem, tobramycin, co-trimoxazole, and the fluoroquinolones, and furthermore was intermediately susceptible to doxycycline as examined by disk diffusion tests (Biodisk). Fecal specimens were negative for growth of salmonella or any other pathogens, including Yersinia, Shigella, and Campylobacter spp. In addition, urine culture and four aerobic and two anaerobic blood cultures were all negative. To monitor the humoral immune response against the Salmonella infection, a serological analysis was done approximately 3 weeks after detection of the first symptoms. Borderline titers of 1/40 (cutoff = 1/10) for typhoid and paratyphoid O antigens were observed. The qualitative agglutination test (Bio-Rad) consisted of antigens recognizing, among others, the serotypes for typhoid O antigen 12 and paratyphoid O antigens 4 and 12, which all cross-react with serotype Brandenburg. Thus, although titers were relatively low, the serological examination supported the specific species serotype Brandenburg. Discussion. | To our knowledge, this is the first report of a muscle infection caused by serotype Brandenburg. Between 30 and 40 cases of muscle infections caused by Salmonella species have been published during the last 40 years (for a review see reference ). Extraintestinal manifestations of serotype Brandenburg are rare, and in a recent paper only two cases of bacteremia with serotype Brandenburg were found out of a total of 32 cases of Salmonella bacteremia . The low incidence of serotype Brandenburg bacteremia is reflected by the few reports on focal extraintestinal infections. However, case reports have been published on isolation of serotype Brandenburg in a ruptured abdominal aorta , in an ovarian cyst due to endometriosis , and finally in acute suppurative thyroiditis . Serotype Brandenburg belongs to genomic group XVIII when defined according to biotype, serotype, and randomly amplified polymorphic DNA typing . Although most cases of serotype Brandenburg are sporadic, the species has also been described in national epidemics . An extensive study in New Zealand including pulsed-field gel electrophoresis and macrorestriction fragment length polymorphism revealed that, among 115 isolates, seven clusters were defined during a 5-year time period . Extraintestinal manifestations of salmonella are found mainly in immunocompromised hosts or in patients with atherosclerosis. Diabetes mellitus and human immunodeficiency virus infection are the most common predisposing conditions for systemic salmonella infections . Salmonella is a well-recognized cause of endovascular infection in different anatomic sites, of which mycotic aneurysm of the aorta is considered to be the most frequent . Among patients above 50 years of age with non-serovar Typhi Salmonella bacteremia, 7 to 10% have been reported to develop arterial infection . In our patient, the muscular abscess most probably originated from an atherosclerotic plaque that had become infected during a transient bacteremic episode. The intracellular bacterium salmonella has developed a sophisticated mechanism for surviving and replicating in macrophages , a phenomenon that might explain the organism's preference for atherosclerotic plaques, which contain numerous phagocytic cells . Our case illustrates that Salmonella infection has to be considered in the differential diagnosis of muscle infections, especially in patients with predisposing conditions such as atherosclerosis. Apart from surgical drainage of abscesses, correct identification of the causative organism is of utmost importance, especially for the long-term outcome of endovascular infections. FIG. 1. | Cross-sectional magnetic resonance imaging of the right thigh and angiography of the arteria femoralis and its area of distribution. Cross-sectional magnetic resonance imaging of the right thigh and angiography of the arteria femoralis and its area of distribution. An abnormal signal intensity of the psoas muscle closely approximating the groin is shown with longitudinal (A) and sagittal (B) sections. Intravenous contrast is accumulated in the infected area. An angiogram of the right arteria femoralis and its collaterals is shown (C). R, righthand side. Scale bars show centimeters. Backmatter: PMID- 12202572 TI - Molecular Epidemiology of Erythromycin Resistance in Streptococcus pneumoniae Isolates from Blood and Noninvasive Sites AB - Erythromycin-resistant isolates of Streptococcus pneumoniae from blood cultures and noninvasive sites were studied over a 3-year period. The prevalence of erythromycin resistance was 11.9% (19 of 160) in blood culture isolates but 4.2% (60 of 1,435) in noninvasive-site isolates. Sixty-two of the 79 resistant isolates were available for study. The M phenotype was responsible for 76% (47 of 62) of resistance, largely due to a serotype 14 clone, characterized by multilocus sequence typing as ST9, which accounted for 79% (37 of 47) of M phenotype resistance. The ST9 clone was 4.8 times more common in blood than in noninvasive sites. All M phenotype isolates were PCR positive for mef(A), but sequencing revealed that the ST9 clone possessed the mef(A) sequence commonly associated with Streptococcus pyogenes. All M phenotype isolates with this mef(A) sequence also had sequences consistent with the presence of the Tn1207.1 genetic element inserted in the celB gene. In contrast, isolates with the mef(E) sequence normally associated with S. pneumoniae contained sequences consistent with the presence of the mega insertion element. All MLSB isolates carried erm(B), and two isolates carried both erm(B) and mef(E). Fourteen of the 15 MLSB isolates were tetracycline resistant and contained tet(M). However, six M phenotype isolates of serotypes 19 (two isolates) and 23 (four isolates) were also tetracycline resistant and contained tet(M). MICs for isolates with the mef(A) sequence were significantly higher than MICs for isolates with the mef(E) sequence (P < 0.001). Thus, the ST9 clone of S. pneumoniae is a significant cause of invasive pneumococcal disease in northeast Scotland and is the single most important contributor to M phenotype erythromycin resistance. Keywords: Introduction : There are two commonly described mechanisms of erythromycin resistance: active drug efflux and methylation of the antibiotic target site. In Streptococcus pneumoniae these result in two major phenotypes, M and MLSB . M phenotype isolates are resistant to macrolides via an active efflux mechanism that requires the presence of the mef(A) gene . This gene was first identified in Streptococcus pyogenes and originally designated mef(A), while a similar gene with 90% identity to mef(A) was later identified in S. pneumoniae and designated mef(E) . More recently, it has been proposed that these two genes are members of the same family and should be referred to by the generic label of mef(A) . The product of the mef(A) gene has not been directly characterized, but its predicted amino acid sequence shows homology with other transporter proteins . Santagati et al. described, in a clinical isolate of S. pneumoniae, a 7.244-kb chromosomal element, Tn1207.1 that contained 8 open reading frames (ORFs), one of which (ORF4) was 100% identical to the original mef(A) sequence of S. pyogenes. Downstream from mef(A), ORF5 coded for a protein that showed homology to MsrA, an ATP-binding protein that mediates resistance to macrolides and streptogramin B in staphylococci. Upstream from mef(A), ORF2 was thought to represent an integrase or site-specific recombinase, although Tn1207.1 was considered a defective transposon because it terminates at the 3' end in a truncated ORF. In the isolate studied, Tn1207.1 was inserted in the pneumococcal genome within the competence gene celB. A further 5.4- or 5.5-kb chromosomal insertion element has recently been described by Gay and Stephens and has been designated the macrolide efflux genetic assembly (mega). mega contains 5 ORFs, of which ORF1 is identical to the original mef(E) sequence of S. pneumoniae. As with Tn1207.1, there is also a homologue of the msr(A) gene downstream from mef(E) designated mel, after the first three amino acids of the predicted protein . The sequences of the two msr(A) homologues found in Tn1207.1 and mega are 98% identical. The five ORFs of mega show a high degree of identity with ORFs 4 to 8 of Tn1207.1, but mega does not contain ORFs with integrase or recombinase homology. PCR studies on 89 mef(E)-positive clinical isolates from Atlanta revealed that there are more than four different insertion sites for mega in the pneumococcal genome. None of the four insertion sites identified were in the celB gene . In the MLSB phenotype, resistance to the structurally unrelated macrolide, lincosamide, and streptogramin B antibiotics is brought about by methylation of 23S rRNA, the common target of these agents . MLSB resistance is determined by members of the erm gene family, and in S. pneumoniae the erm(B) gene is usually carried on the 25.3-kb conjugative transposon Tn1545 along with a separate gene, tet(M), that codes for tetracycline resistance . Transfer of Tn1545 between strains is mediated by the excisionase (xis) and integrase (int) genes . Other transposable elements such as Tn917-like elements and the composite transposon-like structure Tn3872 can also carry erm(B) in S. pneumoniae . The prevalence of erythromycin resistance in S. pneumoniae has increased in several countries over the past few years . However, there are considerable differences between countries in the relative contributions of the M and MLSB phenotypes to the overall prevalence of resistance. Surveys in the United States show a predominance of the M phenotype , while in Italy , Belgium , and Germany the MLSB phenotype is more common. A survey of S. pneumoniae bacteremia carried out in our laboratory showed that the prevalence of erythromycin-resistant isolates was 12% and that 75% of these isolates had the M phenotype . All of the M phenotype isolates belonged to serotype 14 and had very similar profiles on pulsed-field gel electrophoresis . The aim of the present study was to compare the prevalence of the two major erythromycin resistance phenotypes in pneumococcal isolates from blood and from noninvasive sites. We serotyped all erythromycin-resistant isolates and identified the resistance genes and associated insertion elements by PCR and sequencing. MATERIALS AND METHODS : Bacterial isolates. | All bacterial isolates, unless otherwise stated, were cultured from clinical specimens submitted to the routine diagnostic laboratories of the Department of Medical Microbiology, University of Aberdeen, Aberdeen, United Kingdom. Over a 3-year period, from 1998 to 2000, all blood culture isolates of S. pneumoniae and all erythromycin-resistant isolates from other sites were collected. S. pneumoniae was identified by alpha-hemolysis on blood agar and sensitivity to optochin, and erythromycin resistance was detected in the first instance by disk diffusion . Isolates were serotyped by the Scottish Pneumococcal Reference Laboratory, Stobhill Hospital, Glasgow, United Kingdom, by coagglutination ; selected isolates were also characterized by multilocus sequence typing (MLST) by the same laboratory. Isolates from sites other than blood, designated noninvasive, were obtained from the upper respiratory tract, sputum, eye swabs, and ear swabs. We obtained details on all pneumococci isolated during the study period from the computer database of our diagnostic laboratory and thus obtained baseline figures for the number of isolates from sites other than blood. A small number of isolates from invasive sites other than blood, e.g., cerebrospinal fluid, were excluded, and duplicate isolates from the same episode of infection in any one patient were counted only once. Three S. pneumoniae M phenotype isolates (serotype 14) from the South of England (M44, M47, and M58), two from Australia (M231 and M238), and one from Belgium (M27) were kindly supplied by M. C. Enright and B. G. Spratt, Wellcome Trust Centre for the Epidemiology of Infectious Disease, Department of Zoology, Oxford University, Oxford, United Kingdom. Three susceptible isolates (serotype 14) from Australia (M222, M229, and M237) were also obtained from the same source. The isolates donated are all listed in the MLST database as sequence type 9 (ST9). Isolates were stored at -70C in Protect (TSC Ltd., Heywood, United Kingdom) and recovered when required by culture on blood agar plates at 37C in air with 5% CO2. Resistance to erythromycin and the resistance phenotype were confirmed by disk diffusion assay with the disks adjacent to detect inducible resistance . Erythromycin MICs were determined by E-test (AB Biodisk, Solna, Sweden) according to the manufacturer's recommendations. S. pneumoniae strain ATCC 49619 was tested simultaneously as a quality control, and the MIC for this strain was within the manufacturer's recommended range. MICs were also determined by broth microdilution according to the guidelines of the National Committee for Clinical Laboratory Standards (NCCLS) , except that additional antibiotic concentrations of 3, 6, 12, and 24 mg/liter were added to the recommended test range. PCR. | S. pneumoniae cells were harvested from one confluent blood agar plate. Chromosomal DNA was extracted from cell suspensions by the method of Pitcher et al. or with a DNA extraction kit for gram-positive bacteria (Puregene; Gentra Systems, Minneapolis, Minn.). The mef(A) gene was amplified by using primers based on the published sequence of S. pyogenes (5' ATGGAAAAATACAACAATTG [forward] and 5' TTATTTTAAATCTAATTTTCTAAC [reverse]). PCR conditions for amplification of the mef(A) gene comprised an initial denaturation step at 94C for 4 min, followed by 35 cycles of denaturation at 94C for 30 s, annealing at 50C for 30 s, and elongation at 72C for 30 s. After the amplification cycles, a final elongation step at 72C for 5 min was carried out. The primer set used to amplify the erm(B) gene was based on the erythromycin resistance gene carried in the conjugative transposon Tn1545 from S. pneumoniae and consisted of 5' ATTGGAACAGGTAAAGGGC (forward) and 5' GAACATCTGTGGTATGGCG (reverse). PCR conditions for amplification of the erm(B) gene comprised an initial denaturation step at 94C for 4 min, followed by 30 cycles of denaturation at 94C for 30 s, annealing at 57C for 30 s, and elongation at 72C for 1 min. After the amplification cycles, a final elongation step was performed at 72C for 7 min. The homologue of msr(A) was amplified by using primers based on the sequence of this gene contained in the transposable element Tn1207.1 of S. pneumoniae (GenBank accession number ), 5' TGCCTATATTCCCCAGTT (forward) and 5' TTAATTTCCGCACCGACTA (reverse). PCR conditions for amplification of the msr(A) homologue comprised an initial denaturation step at 94C for 4 min, followed by 30 cycles of denaturation at 94C for 30 s, annealing at 50C for 1 min, and elongation at 72C for 1 min. After the amplification cycles, a final elongation step at 72C for 10 min was carried out. To establish whether the transposable element carrying mef(A) in our isolates had the same insertion site as Tn1207.1, specific PCR primers were designed. The forward primer, 5' CTTTCCTTTCTCTATCCA, lies upstream of the known insertion site of Tn1207.1 in the celB gene (GenBank accession number ). The reverse primer, 5' TACATCAACATTACCATCTG, was based on the 5'-end sequence of Tn1207.1 (GenBank accession number ). Amplification conditions were the same as for the msr(A) homologue. Primers for the tet(M) (5' AGTTTTAGCTCATGTTGATG [forward] and 5' TCCGACTATTTGGACGACGG [reverse]) and int (5' GCGTGATTGTATCTCACT [forward] and 5' GACCTCCTGTTGCTTCT [reverse]) genes were as described by Doherty et al. . Primers for the xis gene, 5' AAGCAGACTGACATTCCTA (forward) and 5' GCGTCCAATGTATCTATAA (reverse), were based on the sequence of this gene available in the database (GenBank accession number ). PCR conditions for amplification of tet(M) comprised an initial denaturation step at 94C for 4 min, followed by 35 cycles of denaturation at 94C for 30 s, annealing at 55C for 1 min, and elongation at 72C for 1 min, 30 s. After the amplification cycles, a final elongation step at 72C for 10 min was carried out. Conditions for amplification of int and xis were the same as those used for the msr(A) homologue. All PCR amplification mixtures contained 100 ng of genomic DNA, forward and reverse primers (250 nM), MgCl2 (1.5 mM), deoxynucleoside triphosphates (200 muM; Amersham Pharmacia Biotech UK Ltd., Little Chalfont, United Kingdom), and Taq polymerase (5 U) plus buffer (Bioline, London, United Kingdom). PCR was performed on a Perkin-Elmer (PE) Biosystems (Warrington, United Kingdom) 9700 Thermocycler. PCR products were detected by electrophoresis on agarose gels, followed by staining with ethidium bromide and UV transillumination. DNA sequence analysis. | PCR amplification products were purified by using Centricon C100 columns (Millipore UK Ltd., Watford, United Kingdom). Purified PCR products were sequenced by using the corresponding amplification primers. In addition, a specific region of interest within the tet(M) gene was sequenced in both directions by using internal sequencing primers 5' CGAACTTTACCGAATCTGAA (forward) and 5' CAACGGAAGCGGTGATACAG (reverse); these were based on the published sequence of tet(M) (GenBank accession number ). Sequencing reactions were performed by using the BigDye Terminator cycle sequencing kit (PE Biosystems) according to the manufacturer's instructions. Sequencing products were run on an ABI 377 automated DNA sequencer, and sequencing analysis was carried out with the SeqEd 1.0.3 DNA analysis program (PE Biosystems). Statistical analysis. | The significance of differences in the distribution of isolates between blood and noninvasive sites was determined by the chi-square test, and the significance of differences between erythromycin MICs was determined by the Mann-Whitney test. Both tests were performed with the SPSS statistical package. RESULTS : Erythromycin resistance in S. pneumoniae isolates from blood and noninvasive sites. | Over the 3-year study period, there were 160 isolates of S. pneumoniae from blood and 1,435 from other sites. Nineteen blood isolates (11.9%) were erythromycin resistant, and 18 of these were available for confirmation of phenotype and further study. The prevalence of erythromycin resistance was much lower in isolates from noninvasive sites. Sixty of the 1,435 isolates (4.2%) were erythromycin resistant, and 44 of these were available for confirmation of phenotype and further study. The distribution of resistance phenotypes and serotypes among available isolates from blood and noninvasive sites is shown in Table . The M phenotype was responsible for 76% (47 of 62) of erythromycin resistance overall, largely due to the contribution of M phenotype, serotype 14 (M14) isolates, which accounted for 60% (37 of 62) of all resistant isolates and 79% (37 of 47) of M phenotype resistance. In contrast, MLSB isolates were responsible for 24% (15 of 62) of erythromycin resistance and were distributed across serotypes 6, 14, 15, 19, and 23 . The M14 clone was responsible for 83% (15 of 18) of erythromycin resistance in blood isolates but 50% (22 of 44) of resistance in isolates from other sites. All M14 isolates were penicillin susceptible (data not shown). If we assume that the distribution of resistance phenotypes was the same in isolates which were not available for examination as in those that were, then 10% of blood isolates were M14 (16 of 160) compared with 2.1% of isolates from noninvasive sites (30 of 1,435). Thus, the M14 clone was proportionately 4.8 times more common in blood than in noninvasive sites, and this difference is statistically significant (P < 0.001 by the chi-square test). Resistance genes in erythromycin-resistant isolates. | All M phenotype isolates and one erythromycin-susceptible isolate from blood were PCR positive for the mef(A) gene. Sequence analysis revealed that the mef(A) gene carried by all M14 isolates and by one isolate of serotype 9 was 100% identical to the mef(A) sequence originally described for S. pyogenes (GenBank accession number ), and all these isolates carried the sequence of the msr(A) homologue found in Tn1207.1 (GenBank accession number ) . All isolates with this mef(A) sequence were positive by PCR with primers designed to amplify Tn1207.1 when inserted into the celB gene . All remaining M phenotype isolates of serotypes 9, 19, and 23 carried the mef(E) sequence originally described for S. pneumoniae (GenBank accession number ). These isolates all carried mel, the msr(A) homologue found in the transposable element mega (GenBank accession number ) . Isolates with the mega sequence did not produce a PCR product with primers designed to amplify Tn1207.1 when inserted into the celB gene. All MLSB isolates were erm(B) positive by PCR. Two serotype 19 noninvasive isolates, included in Table as MLSB phenotype isolates, contained both the mef(E) and erm(B) genes . Tetracycline resistance. | Tetracycline resistance is commonly associated with the MLSB phenotype, since the tet(M) and erm(B) genes are both found on the Tn1545 transposon. Fourteen of the 15 MLSB isolates were resistant to tetracycline and were PCR positive for tet(M) . Nine of these 14 isolates were also PCR positive for int and xis, the integrase and excisionase genes commonly associated with Tn1545; 4 isolates were negative for both int and xis, and 1 isolate was positive only for xis. One MLSB isolate of serotype 23 was susceptible to tetracycline. This isolate was positive for tet(M) by PCR, but sequence analysis of the PCR product demonstrated a 10-bp deletion from base 619 to 628 (this deletion was also found in an isolate with the same phenotype collected in our laboratory in 1997). The tetracycline-susceptible isolate was PCR positive for int and xis. Six M phenotype isolates of serotypes 19 (two isolates) and 23 (four isolates) were also tetracycline resistant and were PCR positive for tet(M), int, and xis . Erythromycin MICs for M phenotype isolates. | Erythromycin MICs for all resistant isolates were determined by E-test and by broth microdilution (NCCLS). By broth microdilution, median erythromycin MICs were 12 mg/liter (range, 8 to 24 mg/liter) for 38 M phenotype isolates with the mef(A) sequence and 4 mg/liter (range, 2 to 8 mg/liter) for 9 isolates with the mef(E) sequence. By E-test, median MICs were 20 mg/liter (range, 12 to 32 mg/liter) for 38 M phenotype isolates with the mef(A) sequence and 3 mg/liter (range, 2 to 4 mg/liter) for 9 isolates with the mef(E) sequence. The difference in MICs between mef(A) and mef(E) isolates was statistically significant by both methods (P < 0.001 by the Mann-Whitney test). Comparison with isolates from other geographical locations. | To determine whether the mef(A) sequence found in our Scottish M14 isolates reflected a local phenomenon, we sequenced the gene in M14 isolates from other locations, three from the south of England, two from Australia, and one from Belgium. All six isolates also carried the mef(A) sequence. Erythromycin MICs for these isolates by E-test were in the range of 12 to 32 mg/liter, similar to that of the local M14 clone. It has previously been shown by pulsed-field gel electrophoresis analysis that the profiles of our Scottish M14 isolates are very similar to those of isolates from the south of England . The serotype 14 isolates from England, Australia, and Belgium have all been previously characterized by MLST as ST9 and belong to an M phenotype clone associated with meningitis in the United Kingdom . MLST analysis of nine representative isolates from our local M14 clone confirmed that they were also ST9. TABLE 1 | Distribution of resistance phenotypes and serotypes in erythromycin-resistant S. pneumoniae isolates collected from blood (n = 160) and noninvasive sites (n = 1,435) TABLE 2 | Distribution of resistance-related genes and transposable elements in erythromycin-resistant S. pneumoniae isolates DISCUSSION : This study demonstrates that the M phenotype is the commonest form of erythromycin resistance in northeast Scotland, largely due to the predominance of a serotype 14 clone that has been characterized by MLST as ST9. Thus, the phenotypic pattern of erythromycin resistance observed in the United Kingdom is closer to that of the United States than to that of other European countries, with M phenotype resistance at least three times more prevalent than MLSB resistance overall. We have demonstrated that the ST9 clone is 4.8 times more common in blood than in other sites and is therefore more invasive than the average S. pneumoniae isolate. The ST9 clone has also been identified as an important cause of meningitis throughout the United Kingdom , and this clone is now recognized as a cause of invasive disease in other countries . There are many virulence factors other than the capsule involved in the pathogenesis of pneumococcal infection , and there is evidence that particular strains of pneumococci have a predilection for blood and cerebrospinal fluid . Further characterization of invasive clones such as ST9 will help to explain this process and may in the future offer targets for treatment or prevention of invasive pneumococcal disease. An increase in the expression of a variant form of the enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in our M14 clone, now known to be ST9, has been demonstrated previously . In other species such as S. pyogenes and Staphylococcus aureus, GAPDH is located in the cell wall and is associated with virulence . The role of GAPDH in the pathogenicity of the ST9 clone is therefore worthy of further study. The mef(A) genes of S. pneumoniae and S. pyogenes were found initially to be different and were designated mef(E) and mef(A), respectively. We have shown that an unusual characteristic of the ST9 clone is that it possesses the mef(A) sequence normally associated with S. pyogenes. The original S. pyogenes sequence [mef(A)] was reported for four isolates of S. pneumoniae from Italy , and more recently it has been reported for a further 17 isolates, also from Italy . Conversely, there is a report of the original pneumococcal sequence [mef(E)] in a single isolate of S. pyogenes , and we have also observed this (data not shown). A survey of erythromycin-resistant viridans streptococci found the mef(E) sequence in S. mitis, S. oralis, and S. anginosus, while the mef(A) sequence was found in only one isolate of S. oralis . Thus, the potential for the spread of M phenotype resistance in S. pneumoniae is increased by the widespread presence of the mef(A) gene in other, less pathogenic species. The original characterization of Tn1207.1 showed that it was associated with the mef(A) sequence , and our data confirmed that all pneumococcal isolates with the S. pyogenes mef(A) sequence contained Tn1207.1. In contrast, the nine M phenotype isolates with the pneumococcal mef(E) sequence contained the mega insertion element. It seems likely, therefore, that the two mef(A) sequences are associated with different transposable elements. In support of this conclusion, Del Grosso et al. have also demonstrated an association between mef(A) sequence and genetic element in six Italian isolates . Our results indicate that the erythromycin MIC for pneumococci possessing mef(A) together with the associated transposable element Tn1207.1 is higher than that for pneumococci possessing mef(E) in association with mega. Gay and Stephens demonstrated that in the transposable element mega, the mef(E) gene is cotranscribed with mel, the msr(A) homologue. In staphylococci, msr(A) encodes an ATP-binding cassette that provides energy for the efflux of macrolides and streptogramin B. Thus, the erythromycin MIC may be influenced not only by the Mef(A) protein but also by the actions of the MsrA homologue. Different sites of insertion in the genome may also result in different rates of transcription. However, relatively few of our isolates contained the mef(E) gene sequence, and therefore further studies are required to confirm the MIC difference. Resistance to tetracycline is a common characteristic of the MLSB phenotype because the erm(B) and tet(M) genes can be found on the same transposon, Tn1545. The presence of both resistance genes erm(B) and mef(A) in MLSB isolates has been described previously , and these isolates are expected to be resistant to tetracycline. In contrast, mef(A) in M phenotype isolates is not known to be linked to tetracycline resistance. We have shown that M phenotype isolates of serotypes 19 and 23 can carry the tet(M) gene and other elements of Tn1545, such as int and xis, without possessing the erm(B) gene. Similar findings were reported for Spanish isolates . In this study we have identified a 10-bp deletion in the sequence of the tet(M) gene of an MLSB isolate that was susceptible to tetracycline, relative to the tet(M) sequence in tetracycline-resistant isolates. Susceptibility to tetracycline in MLSB isolates has been reported for Spanish isolates, but these isolates did not possess the tet(M) gene . Our data show that tetracycline resistance is not a reliable guide to MLSB phenotype erythromycin resistance, since we found both M phenotypes that were resistant to tetracycline and MLSB phenotypes that were susceptible to tetracycline. In conclusion, the present study has demonstrated the clinical importance in our region of an M phenotype, serotype 14 clone of S. pneumoniae that has invasive properties and for which MICs are higher than for other M phenotype isolates. The clone has been identified as ST9 and has been isolated in other parts of the world. Further characterization of this clone may yield further insights into both invasiveness and the detailed mechanisms of M phenotype resistance. Backmatter: PMID- 12202555 TI - Polyphyletic Strains of Hepatitis E Virus Are Responsible for Sporadic Cases of Acute Hepatitis in Japan AB - Among 87 patients who were previously treated for acute hepatitis of unknown etiology between 1992 and 2001 at five hospitals in Japan, 11 (13%) patients were positive for immunoglobulin M-class antibodies to hepatitis E virus (HEV) by enzyme immunoassay and had detectable HEV RNA by reverse transcription-PCR with two independent sets of primers derived from well-conserved genomic areas in open reading frames 1 and 2. Clinical HEV infection was significantly associated with male sex (9 of 11 versus 29 of 76 patients [P < 0.01]) and older age (52 +- 11 [mean +- standard deviation] versus 41 +- 17 years [P < 0.05]), and its prevalence differed by geographic region (6 to 25%), with a higher rate in the northern part of Japan. At admission, the 11 patients with HEV-associated hepatitis had elevated alanine aminotransferase levels of 914 to 4,850 IU/liter, and all but 1 had elevated bilirubin levels of 1.5 to 24.0 mg/dl. The 11 HEV isolates were of genotype III or IV and were segregated into three groups with intergroup nucleotide differences of 9.5 to 22.0%. Phylogenetic analysis revealed that four isolates of genotype III were closely related to a Japanese isolate, while the other four isolates of the same genotype were nearest those from the United States. The remaining three isolates were close to known isolates of genotype IV in China and Taiwan but shared less than 88% identity with them. These results indicate that multiple genotypes of HEV cocirculate in Japan and contribute to the development of sporadic acute hepatitis, with the prevalence differing by age, sex, and geographic region. Keywords: Introduction : Hepatitis E, the major form of acute viral hepatitis in adults in many developing countries in Asia, Africa, and Latin America, is caused by hepatitis E virus (HEV). HEV is transmitted primarily by the fecal-oral route, and waterborne epidemics are characteristic of hepatitis E and may occur in any of three forms: large epidemics, smaller outbreaks, or sporadic infections. Sporadic cases have also been reported in areas where HEV is not considered endemic. Many of these cases can be associated with travel to regions where HEV is endemic . Recently, however, accumulating lines of evidence indicate that HEV-associated hepatitis also occurs among individuals in developed countries with no history of travel to areas where HEV is endemic (, , , , -, , , ). The genome of HEV is a single-stranded, positive-sense RNA of approximately 7.2 kb and contains a short 5' untranslated region (UTR), three open reading frames (ORFs; ORF1, ORF2, and ORF3), and a short 3' UTR terminated by a poly(A) tract . The entire genomic sequence of HEV was first published in 1991 for a strain from Myanmar (formerly called Burma) , which shared nucleotide identity of >93% across the genome with the nucleotide sequences of additional isolates obtained from China, India, Nepal, and Pakistan . In addition, the genomic sequence of a Mexican isolate that was implicated in an outbreak that occurred in Mexico in 1986 was published in 1992 . The Mexican isolate (MEX-14) is distinct from the Burmese isolate and constitutes a second genotype. A third group of HEV that is distinct from the Burmese-like and Mexican isolates has been identified in patients with acute hepatitis in the United States and European countries including Austria, Greece, Italy, Spain, and the United Kingdom and in Argentina, where HEV is not endemic (, , -, , , ). Extensive diversity has also been noted among HEV isolates from patients with acute hepatitis in China and Taiwan, which are distinct from the original Chinese isolates, and these isolates constitute a fourth group . Accordingly, HEV sequences have tentatively been classified into four major genetic groups (genotypes I to IV). Worldwide, most HEV infections are caused by genotype I, while only isolated cases of infection with HEV of genotype III or IV have been described . In Japan, clinical HEV infection rarely occurs, and most, if any, cases of hepatitis E observed thus far have been regarded as imported cases of hepatitis . Recently, however, the seroprevalence of antibodies against HEV (anti-HEV) in healthy individuals was reported to range from 1.9 to 14.1%, depending on the geographic area in Japan . In addition, an HEV strain of genotype III (strain JRA1) has been isolated from a Japanese patient with acute hepatitis who had never been abroad , and a swine HEV strain (strain swJ570) with the highest degree of similarity to isolate JRA1 among the known HEV isolates has been isolated from a farm pig in Japan, although their entire genomes shared only 89% identity . These results indicate that HEV infection may be circulating in Japan. Therefore, in the present study, we tested the sera of 87 patients with sporadic acute hepatitis of unknown etiology from five city or university hospitals located in different geographic regions in Japan for the presence of the immunoglobulin M (IgM)-class anti-HEV and HEV RNA and analyzed the HEV strains molecularly to define the region-dependent prevalence of clinical HEV infection and the extent of genetic diversity among the HEV strains that are spreading in Japan. MATERIALS AND METHODS : Sera from patients with sporadic cases of acute hepatitis and blood donors. | Serum samples were obtained from 87 patients (38 males and 49 females; mean +- standard deviation [SD] age, 42.3 +- 16.7 years) who were seen at five city or university hospitals located (from north to south) in Sapporo on Hokkaido Island and in Iwate, Fukushima, Tokyo, and Yamanashi on mainland Honshu in Japan, with a clinical diagnosis of sporadic acute hepatitis of non-ABC etiology. These patients were seen at the respective hospitals during the past 3 to 10 years (from between 1992 and 2001 to between 1998 and 2001) ; each patient was from the same geographic region where the respective hospital is located. They were all negative for anti-hepatitis A virus (anti-HAV) IgM, hepatitis B virus (HBV) markers (anti-HBV core IgM and hepatitis B surface antigen [HBsAg]), and anti-hepatitis C virus (anti-HCV). The sera were evaluated for the presence of anti-HAV IgM, anti-HBV core IgM, HBsAg, and anti-HCV with commercially available kits (HAVAB-M and CORZYME-M [Abbott Laboratories, Abbott Park, Ill.], Mycell [Institute of Immunology Co. Ltd., Tokyo, Japan], and Abbott HCV PHA 2nd Generation [Dainabot, Tokyo, Japan], respectively). Periodic serum samples were collected from the patients with HEV viremia. In addition, control sera from 200 healthy blood donors (100 males and 100 females; age range, 16 to 24 years) at the Japanese Red Cross Yamaguchi Blood Center, which is located in the southern part of Japan, were used for determination of the cutoff values for the anti-HEV IgM and IgG assays, based on the recent report of Li et al. , who described that healthy Japanese individuals of less than 30 years of age living in the southern part of Japan were negative for both anti-HEV IgM and IgG. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the ethics committees of the institutions, and informed consent was obtained from each patient. Production and purification of recombinant HEV ORF2 protein. | A recombinant HEV ORF2 protein whose N terminus was truncated (amino acid residues 111 to 660 of ORF2) was expressed by a recombinant baculovirus by the method described by Li et al. , with the following modifications. The putative capsid gene (ORF2) of an HEV isolate of genotype IV (HE-J1; DDBJ/GenBank/EMBL accession no. ) was amplified by reverse transcription (RT)-PCR with the following set of primers: sense primer, 5'-GGA TCCATG GCT GTG GCC CCG GCC CCT GAT-3'; antisense primer, 5'-GAG CTCATCAAT ACT CCC GGG TTT TAC C-3'. The restriction sites introduced in the primers (BamHI in the sense primer and SacI in the antisense primer) are underlined, and the truncated ORF2 start codon (ATG) and the sequence of TCATCA complementary to two in-frame stop codons (successive TGA codons) are shown in boldface. The PCR product was cloned into pT7BlueT vector (Novagen, Inc., Madison, Wis.) and digested with BamHI and SacI. The resulting 2-kb fragment was inserted into the BglII-SacI site of a transfer vector, pYMG (Katakura Industries Co. Ltd., Saitama, Japan), and sequenced. The 5' truncated putative capsid (ORF2) gene encoding 550 amino acids was cloned into a baculovirus expression vector and expressed in silkworm pupae. The silkworm pupae were lysed in 20 mM PIPES [piperazine-N,N'-bis(2-hydroxypropane-3-sulfonic acid)] buffer (pH 6.6) containing 10% (vol/vol) glycerol, 0.1 M NaCl, 1 mM phenylmethylsulfonyl fluoride, and 10 mM benzamidine and were then homogenated in 10% (wt/vol) Triton X-100, followed by centrifugation at 100,000 x g at 4C for 15 min. The resulting supernatant was treated with polyethylene glycol at a final concentration of 4% (wt/vol). The precipitates were redissolved in 20 mM Tris-HCl (pH 8.0) and purified by anion-exchange chromatography. Following purification, the purified protein was shown to produce one predominant band of 61 kDa by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The protein concentration was determined with a protein assay kit (Pierce, Rockford, Ill.) by using bovine serum albumin as a standard. The purified recombinant capsid protein was used in enzyme-linked immunosorbent assay (ELISA). ELISA for detecting anti-HEV antibodies. | To detect anti-HEV IgM and IgG by using purified recombinant ORF2 protein, ELISA was performed as follows. Wells of microplates (Microlon 600; Greiner Labortechnik GmbH, Frickenhausen, Germany) were coated with 50 mul of the recombinant ORF2 protein (5 mug/ml in phosphate-buffered saline), and the plates were incubated at room temperature for 4 h. One hundred microliters of saline containing 40% (vol/vol) calf serum (GIBCO-BRL, Grand Island, N.Y.) was added. The microplates were incubated at room temperature for 1 h with shaking. The blocking buffer was discarded, and each well was washed five times with saline. To test for anti-HEV IgG, 50 mul of each sample was added to each well at a dilution of 1:100 in saline containing 40% calf serum. The microplates were incubated at room temperature for 1 h with gentle agitation and were then washed five times with washing buffer (0.05% Tween 20 in saline). Fifty microliters of phosphate-buffered saline containing 25% (vol/vol) fetal bovine serum (Medical & Biological Laboratories, Co. Ltd., Nagoya, Japan) and peroxidase-conjugated mouse monoclonal anti-human IgG antibody (G19; Institute of Immunology Co. Ltd.) was added to each well. The microplates were incubated at room temperature for 1 h with gentle agitation and then washed five times with washing buffer. Fifty microliters of tetramethylbenzidine (TMB) soluble reagent (ScyTek Laboratories, Logan, Utah) as a substrate was added to each well. The plate was incubated at room temperature for 10 min in the dark, and then 50 mul of TMB stop buffer (ScyTek Laboratories) was added to each well. The optical density (OD) of each sample was read at 450 nm. For the anti-HEV IgM assay, peroxidase-labeled mouse monoclonal anti-human IgM antibody (M49; Institute of Immunology Co. Ltd.) was used in place of the enzyme-labeled anti-human IgG. Test samples with OD values equal to or greater than the cutoff value were considered positive for anti-HEV IgG or anti-HEV IgM. Detection of HEV RNA. | Total RNAs were extracted from the serum sample with guanidinium thiocyanate and phenol-chloroform with the TRIZOL LS reagent (Invitrogen, Groningen, The Netherlands). The RNA preparation thus obtained was reverse transcribed with SuperScript II RNase H- reverse transcriptase (GIBCO-BRL) and an antisense primer (primer HE040; 5'-CCC TTR TCC TGC TGA GCR TTC TC-3' [R = A or G]) specific for the HEV ORF2 sequence and was then subjected to nested PCR in the presence of TaKaRa Ex Taq (TaKaRa Shuzo, Shiga, Japan). A part of the ORF2 sequence was amplified with the primer pair HE044 (sense primer; 5'-CAA GGH TGG CGY TCK GTT GAG AC-3' [H = A, T, or C; Y = T or C; and K = G or T]) and HE040 in the first round and HE110-2 (sense primer; mixture of three sequences, 5'-GYT CKG TTG AGA CCT CYG GGG T-3', 5'-GYT CKG TTG AGA CCA CGG GYG T-3', and 5'-GYT CKG TTG AGA CCT CTG GTG T-3' [common nucleotides are underlined]) and HE041 (antisense primer; 5'-TTM ACW GTC RGC TCG CCA TTG GC-3' [M = A or C, W = A or T]) in the second round (ORF2 PCR). The PCR amplification was carried out for 35 cycles in the first round (94C for 30 s [an additional 2 min was used in the first cycle], 55C for 30 s, 72C for 75 s [an additional 7 min was used in the last cycle]) and for 25 cycles in the second round under the same conditions used for the first round except that extension was carried out for 60 s. The size of the amplification product of the first-round PCR was 506 bp, and that of the amplification product of the second-round PCR was 458 bp. The amplification products were electrophoresed on a 1.5% (wt/vol) NuSieve 3:1 agarose gel (FMC BioProducts, Rockland, Maine), stained with ethidium bromide, and photographed under UV light. The RT-PCR assay was performed in duplicate, and reproducibility was confirmed. To confirm the presence of HEV RNA, a part of ORF1 was amplified by nested RT-PCR with the primer sets HE090 (sense primer; 5'-GCA GAC CAC RTA TGT GGT CGA YGC C-3') and HE094 (antisense primer; 5'-TGG CGG RMC ATN GCC TCB GCR ACA TC-3' [N = A, G, T, or C; B = G, T, or C]) in the first round and HE092 (sense primer; 5'-TGT GGT CGA YGC CAT GGA GGC CCA-3') and HE095 (antisense primer; 5'-CCR TCR AAR CAG TAA GTS CGG TC-3' [S = G or C]) in the second round under the same conditions as those described above for the ORF2 PCR; they generated amplification products of 567 and 459 bp, respectively. To avoid contamination during PCR procedures, the guidelines of Kwok and Higuchi were strictly observed. Two negative controls and one positive control were included for every 17 test samples. Results were recorded only when false-positive results were not obtained for the negative controls and HEV RNA was detected in the positive control. The negative control was water treated the same way as the serum samples. The positive control was serum from a Nepali patient with a sporadic case of acute hepatitis E caused by HEV of genotype I , used at a dilution of 1:1,000 in anti-HEV-negative human sera obtained from healthy individuals. The nested RT-PCRs used in the present study had sensitivities comparable to that of a reported method for the detection of HEV RNA in the serum from the Nepali patient and representative serum samples from patients with hepatitis E in the present study. The sequences of the primers mentioned above were chosen from well-conserved regions of the entire HEV genome by comparing 19 known human and 2 known swine HEV sequences to develop two universal RT-PCR assays that are capable of detecting HEV strains with significant sequence variations. The HEV sequences were as follows: B1 and B2 in Burma; C1, C2, C3, C4, C5, and C6 in China; I1, I2, and I3 in India; Ne1 in Nepal; and P1 and P2 in Pakistan in genotype I (the abbreviations used for the HEV isolates are in accordance with the recent report by Schlauder and Mushahwar ); MEX-14 in Mexico in genotype II; US1 and US2 in the United States and JRA1 in Japan in genotype III; and T1 in China in genotype IV as well as swine HEV isolates in genotype III (swUS1 in the United States and swJ570 in Japan) (see Table for the DDBJ/GenBank/EMBL databases accession numbers for each isolate). Sequence analysis of PCR products. | The amplification products were sequenced on both strands either directly or after cloning into the pT7BlueT vector by using the BigDye Terminator Cycle Sequencing Ready Reaction kit (Applied Biosystems, Foster City, Calif.). Sequence analysis was performed with Genetyx-Mac software (version 10.1.6; Software Development, Tokyo, Japan) and ODEN software (version 1.1.1) from the DNA Data Bank of Japan (DDBJ; National Institute of Genetics, Mishima, Japan) . Sequence alignments were generated with the CLUSTAL W program (version 1.8) . Phylogenetic trees were constructed by the neighbor-joining method , based on the partial nucleotide sequence of the ORF1 region (287 nucleotides [nt]) or the ORF2 region (301 nt). Bootstrap values were determined on 1,000 resamplings of the data sets . The final tree was obtained with the TreeView program (version 1.6.6) . Nucleotide sequence accession numbers. | The sequences determined in the present study have been deposited in the DDBJ/GenBank/EMBL nucleotide databases under accession nos. to . TABLE 1 | Characteristics of the 87 patients with acute hepatitis of unknown etiology enrolled in the present study TABLE 4 | Comparison of partial nucleotide sequences of 11 HEV isolates in groups A to C obtained in the present study among each other and with those of 21 human and swine HEV isolates whose entire or nearly entire genomic sequence is known RESULTS : Prevalence of IgM-class and IgG-class anti-HEV. | To determine the cutoff value in the anti-HEV IgM assay, 200 control serum samples were used as a panel. The OD values ranged from 0.010 to 0.352, and the value of 0.353, which was calculated to be six SDs above the mean value (0.068), was used as the tentative cutoff value. Similarly, in the assay of anti-HEV IgG, OD values ranging from 0.002 to 0.150 were obtained for the 200 control serum samples; the OD value of 0.152 (mean + 6 SDs) was used as the cutoff value for anti-HEV IgG. By using these cutoff values, serum samples obtained from the 87 patients at admission were tested for anti-HEV. The sera from 11 patients were positive for anti-HEV IgM, with the OD values ranging from 1.638 to >3.0; those for the remaining 76 patients ranged from 0.022 to 0.283. Anti-HEV IgG was detectable in 16 patients including the 11 patients with anti-HEV IgM. The OD values for anti-HEV IgG in 11 patients with anti-HEV IgM exceeded 2.6, those for the 5 patients without anti-HEV IgM ranged from 0.175 to 0.660, and those for the remaining 71 patients ranged from 0.014 to 0.146. The assays were done at least in triplicate, and reproducibility was confirmed. Prevalence of HEV RNA. | The serum samples obtained from all 87 patients at admission were tested for HEV RNA by RT-PCR with ORF2-specific primers. HEV RNA was detected in the serum of all 11 patients with anti-HEV IgM, and it was not detected in the remaining 76 patients who were negative for anti-HEV IgM. A second RT-PCR assay with ORF1-specific primers confirmed these positive and negative results. Of interest, the prevalence of HEV RNA differed by geographic region, ranging from 6% (2 of 35 patients) in the central part of the Japanese mainland (Tokyo and Yamanashi) to 25% (4 of 16 patients) in Sapporo on Hokkaido, the northernmost island of Japan , with the difference being statistically significant (P < 0.05). The 11 patients with HEV viremia had peak alanine aminotransferase (ALT) levels of 914 to 4,850 IU/liter and peak aspartate aminotransferase (AST) levels of 539 to 5,931 IU/liter ; all 11 patients had an elevated ALT and/or AST level of >1,000 IU/liter. Despite the marked elevation of ALT and AST levels at the initial examination, the abnormal liver function test values normalized rapidly within 1 month in nine patients but persisted until 35 or 53 days after admission in the remaining two patients; one of these patients (patient 3) had severe jaundice, and the condition of the other patient (patient 11) was complicated by duodenal ulcer during the hospital admission. Ten patients developed jaundice and had bilirubinemia with an increased bilirubin level of 1.5 to 24.0 mg/dl, indicating that most patients contracted moderate to severe cholestasis. Of note, the bilirubinemia continued in four patients (patients 3, 4, 7, and 11) even after normalization of the ALT and AST levels. When the 11 patients with HEV viremia were compared with the remaining 76 patients who were negative for HEV RNA, the 11 patients with HEV viremia were significantly older (P < 0.05), consisted of a higher percentage of male patients (P < 0.01), and had higher peak total bilirubin and peak ALT levels (P < 0.05, P < 0.01, respectively) . In other words, the patients with HEV RNA were significantly associated with a higher level of peak total bilirubin (>4 mg/dl) (P < 0.01), a higher level of peak ALT (>1,000 IU/liter) (P < 0.005), and a higher level of peak AST (>800 IU/liter) (P < 0.05). Detection of anti-HEV and HEV RNA in follow-up sera from infected patients. | Figure illustrates the anti-HEV IgM and HEV RNA profiles associated with HEV infection in five patients. Five to seven additional serum samples were available from these five patients (patients 2 to 5 and 10) during the follow-up period of 115 to 287 days after admission. HEV RNA remained detectable in serum until 12 to 35 days after admission and even on day 23, when the liver function test values had returned to nearly normal levels in patient 10. The anti-HEV IgM antibody levels were the highest at admission and then decreased rapidly in all patients except one (patient 3), who continued to have an IgM antibody level of greater than 3.0 until 69 days after admission. A low level of IgM antibody was detectable up through the end of the observation period in three patients (147 days in patient 3, 259 days in patient 4, and 287 days in patient 10), but disappeared at 80 days in patient 2 and 141 days in patient 5. The IgG antibody level was as high as 3.0 OD units at admission in all patients and persisted at high levels. There was no discernible reduction in the IgG antibody level through the end of the observation period. Genetic heterogeneity of HEV isolates recovered from Japanese patients with hepatitis E. | The amplification products of ORF1 and ORF2 (both 412 nt; primer sequences at both ends excluded) from 11 viremic patients were sequenced and compared . The 11 HEV isolates, designated HE-JA1 to HE-JA11 , were segregated into three groups (groups A to C; tentatively named only in this paper) on the basis of the nucleotide sequence. Group A comprised HE-JA5, HE-JA6, HE-JA9, and HE-JA11; group B comprised HE-JA4, HE-JA7, HE-JA8, and HE-JA10; and group C comprised HE-JA1, HE-JA2, and HE-JA3. The intragroup nucleotide sequence identities were 88.8 to 98.8% for the ORF1 sequence and 88.3 to 98.5% for the ORF2 sequence, while the intergroup nucleotide sequence identities were only 79.3 to 90.5% for the ORF1 sequence and 78.0 to 89.1% for the ORF2 sequence. When the sequences of the group A isolates were compared with those of HEV isolates whose entire or nearly entire sequence is known, the group A isolates were found to be closely related to human HEV (JRA1) and swine HEV (swJ570) of Japanese origin , and the group B isolates were most homologous to human HEV (US1 and US2) and swine HEV (swUS1) of U.S. origin ; both group A and group B isolates were classifiable into genotype III. The group C isolates were closest to the T1 isolate of genotype IV, which was recovered from a patient in China with a sporadic case of acute hepatitis , but they had ORF1 and ORF2 nucleotide sequence similarities of <90% compared with the sequences of the other isolates. Partial ORF1 sequences of genotype III have been reported for seven human HEV isolates from European countries (Austria, Greece, Italy, Spain, and the United Kingdom) , two human HEV isolates from Argentina , and one swine HEV isolate from New Zealand . Phylogenetic analysis of the common ORF1 sequence of 287 nt confirmed that group A isolates were nearest JRA1 and swJ570 and that group B isolates were closely related to US1, US2, and swUS1 among all genotype III isolates whose entire or partial nucleotide sequences are known . Partial ORF2 sequences of genotype IV are available for 21 human HEV isolates in China and 8 human and 3 swine HEV isolates in Taiwan . The phylogenetic tree constructed on the basis of the partial ORF2 sequence of 301 nt confirmed that group C isolates belonged to genotype IV but that they were clearly separate from known genotype IV isolates from China and Taiwan . Recently, Arankalle et al. reported on 12 swine isolates of genotype IV in western India whose partial ORF2 sequences of 241 to 263 nt shared only 78.7 to 85.5% identities with those of three isolates of group C (HE-JA1, HE-JA2, and HE-JA3), suggesting that group C isolates belong to a new subgroup of genotype IV, being separate from other subgroups to which the Chinese, Taiwanese, and Indian isolates are classifiable. FIG. 1. | Detection of HEV RNA and anti-HEV IgM in initial and follow-up serum samples from five patients with hepatitis E. Detection of HEV RNA and anti-HEV IgM in initial and follow-up serum samples from five patients with hepatitis E. The ALT and total bilirubin (T. Bil) levels are also shown. FIG. 2. | Phylogenetic tree constructed by the neighbor-joining method based on the partial nucleotide sequence of the ORF1 region (287 nt) of 42 human and swine HEV isolates. Phylogenetic tree constructed by the neighbor-joining method based on the partial nucleotide sequence of the ORF1 region (287 nt) of 42 human and swine HEV isolates. In addition to 21 reported human and swine HEV isolates of genotypes I to IV whose entire or nearly entire sequence is known (see Table for the names of the isolates and relevant accession numbers), 10 reported isolates of genotype III whose partial sequence of 287 or 371 nt is available, as well as the 11 HEV isolates found in the present study, which are indicated in boldface type, were included for comparison. The names (accession numbers) of the 10 reported isolates are as follows: Ar1 and Ar2 from Argentina, Au1 from Austria, Gr1 and Gr2 from Greece, It1 from Italy, swNZ from New Zealand, Sp1 and Sp2 from Spain, and UK from the United Kingdom. Genotype designations I to IV are in accordance with the recent report by Schlauder and Mushahwar . Bootstrap values of >70% are indicated for the major nodes as a percentage of the data obtained from 1,000 resamplings. For visual clarity, HEV isolates of Japanese origin, including those described in previous reports (JRA1 and swJ570) , are indicated by shaded circles with the tentative designation of groups A to C (see Table ). FIG. 3. | Phylogenetic tree constructed by the neighbor-joining method based on the partial nucleotide sequence (301 nt) of the ORF2 region of 73 human and swine isolates. Phylogenetic tree constructed by the neighbor-joining method based on the partial nucleotide sequence (301 nt) of the ORF2 region of 73 human and swine isolates. In addition to 21 reported human and swine HEV isolates of genotypes I to IV whose entire or nearly entire sequence is known, 40 reported isolates of genotype III or IV whose partial sequence of 301, 304, 346, 348, 421, 1,402, or 1,497 nt has been determined were included for comparison. They are deposited under accession nos. , , , to , , , , , , to , to , , , , , to , to , , and to . All except one of the human and swine HEV strains isolated in Japan are segregated into one of three groups (groups A to C), as indicated by the shaded circles; the one exception is swine isolate swJ791, which is close to the Spanish strains (Sp1, Sp2, and Sp3). Bootstrap values of >70% are indicated for the major nodes as a percentage of the data obtained from 1,000 resamplings. TABLE 2 | Profiles of the 11 patients with HEV viremia TABLE 3 | Comparison of various features between the HEV RNA-positive and -negative groups among the 87 patients with sporadic, acute hepatitis of unknown etiology DISCUSSION : Infection with HEV is widespread, and hepatitis E is endemic and occasionally epidemic in many developing countries in Asia, Africa, and Latin America. In industrialized countries, although anti-HEV has been detected in 4 to 36% of healthy individuals , sporadic cases of hepatitis E not associated with traveling to regions of endemicity have only rarely been reported . In the present study conducted in Japan, where HEV infection is not considered endemic, 11 (12.6%) of 87 patients who had previously been diagnosed with sporadic acute hepatitis of non-ABC etiology were found to be infected with HEV. The prevalence of HEV RNA differed by age, sex, and geographic region, being consistent with the reported age-, sex-, and region-dependent prevalence of anti-HEV IgG in healthy individuals in Japan . Interestingly, in our present study, the prevalence of HEV-associated hepatitis among sporadic acute hepatitis cases of non-ABC etiology was significantly associated with male sex, older age (>=40 years), and living in the northern part of Japan (Hokkaido and Iwate) (P < 0.01, P < 0.005, and P < 0.05, respectively). When the evaluation was restricted to male patients, patients aged >=40 years, and patients living in Hokkaido and Iwate, the prevalence of hepatitis E among sporadic acute hepatitis cases of non-ABC etiology is estimated to be 9 of 38 patients (24%), 11 of 48 patients (23%), and 8 of 38 patients (21%), respectively. Furthermore, the 11 patients with hepatitis E had markedly elevated serum transaminase levels (ALT levels, 914 to 4,850 IU/liter; AST levels, 539 to 5,931 IU/liter). When the evaluation was restricted to patients having abnormal ALT and/or AST levels of >1,000 IU/liter at disease onset, the prevalence was estimated to be 11 of 49 patients (22%). Therefore, in Japan, the possibility of the presence of clinical HEV infection should be taken into consideration when clinicians are confronted with patients with sporadic acute hepatitis of non-ABC etiology, paying special attention to age, sex, location of residence, and the results of liver function tests. The limited numbers of samples evaluated in this study suggest that a much larger study with more patients is needed to draw a definitive conclusion. The source of the HEV variants in the 11 Japanese patients studied is unclear, and it is not easy to trace the origins of these isolates. Patients 1 to 4, living in Hokkaido, did not report contact with pigs or rats. However, patients 5 and 6, living in Iwate, were a retail meat dealer and a meat-processing trader, respectively, who were engaged in processing raw meat such as beef, pork, and chicken, suggesting that they were at increased risk for zoonotic HEV infection. Of note, the HE-JA5 isolate from patient 5 and the HE-JA6 isolate from patient 6 were both closely related to swine HEV isolates thus far identified in Japan (isolates swJ570 and swJ681) . HE-JA8 was isolated in 2001 from patient 8, who was a farmer working with pigs in his herd in Iwate. Considering the accumulating lines of evidence for zoonotic HEV infection (, -), it is very likely that he was infected with swine HEV from his herd. However, the HE-JA8 isolate was nearer U.S. human and swine strains (US1, US2, and swUS1), with the highest nucleotide identity being 94.4%, than to a human isolate (JRA1) or a swine isolate (swJ570) that are believed to be indigenous to Japan . This result indicates that HE-JA8 may be a U.S. strain. However, the sequence of HE-JA8 shared 98.5% nucleotide identity with that of HE-JA10, which was recovered in 1993 from patient 10, who lived in Tokyo. In addition, HE-JA8 as well as HE-JA4, HE-JA7, and HE-JA10 shared 90.5 to 99.7% nucleotide sequence identities with three recently reported Japanese HEV isolates of genotype III (JHA-Sap, JKN-Sap, and JMY-Haw) which are also classifiable into group B, suggesting that it is more likely that these seven HEV isolates of group B in genotype III are domestic and widespread in Japan. In this context, comparison of the sequence of isolate HE-JA8 from patient 8 and those of HEV strains from pigs in his herd would be informative. HE-JA11 was from patient 11, who had been to Southeast Asian countries including Malaysia and Indonesia 3 and 12 months before the onset of acute hepatitis, to visit branch offices outside Japan as an employee of an electrical equipment manufacturer for periods of about 1 week. The possibility that this patient was infected with HEV in any of these countries that he had recently visited, where hepatitis E is endemic, cannot be ruled out. However, HE-JA11 was closely related to reported human and swine HEV isolates (JRA1 and swJ570) from Japan, as well as to three HEV isolates (HE-JA5, HE-JA6, and HE-JA9) recovered in the present study from patients living in Iwate and Fukushima who had never been abroad, suggesting that patient 11 was infected in his hometown (Yamanashi) or that he was infected outside Japan with a Japanese strain that had been exported from Japan in the past. Sequence analysis of the HEV strains in Southeast Asian countries such as Cambodia, Indonesia, Malaysia, the Philippines, and Vietnam, for which the nucleotide sequences of the circulating HEV strains have not yet been determined, would be of great interest. The increasing globalization of food markets by industrialized countries has the potential of introducing HEV into new areas in the world. Japanese people have a habit of eating raw fish and other uncooked seafood, both those caught in Japan and those imported from many countries in the world including China, Taiwan, and the United States, where HEV isolates of genotype III or IV circulate. In China, an HEV strain of genotype I is also prevalent and was the source of an epidemic HEV infection . However, genotype I HEV was not detected in our patients. As indicated in Fig. and , isolates HE-JA1, HE-JA2, and HE-JA3, obtained from patients 1 to 3, respectively, who lived in Hokkaido, belonged to genotype IV or group C designated in the present study but were located on a new branch and were considered to be in a subgroup separate from those of all known genotype IV isolates from China and Taiwan. The sequences of these three genotype IV isolates of Japanese origin shared less than 85% identity with those of the recently reported swine isolates from western India . In addition, the nucleotide sequences of isolates HE-JA1, HE-JA2, and HE-JA3 of genotype IV had identities of 88.6 to 98.5% with those of the three recently reported Japanese HEV isolates of the same genotype (JAK-Sai, JKK-Sap, and JSY-Sap; HE-JA3 shared 97.9 and 98.5% nucleotide sequence identities with JKK-Sap and JSY-Sap, respectively) , indicating that all these Japanese isolates of genotype IV can be segregated to a subgroup separate from the Chinese, Taiwanese, and Indian strains, although marked sequence variability among the Japanese isolates was recognized as well. Taken altogether, we speculate that Japanese genotype IV isolates are indigenous to and circulating in Japan, although we cannot reasonably rule out the possibility of an outside source for these strains. To elucidate this issue, global molecular epidemiological studies with HEV strains of various genotypes are required. As described above, evidence is accumulating that hepatitis E is zoonotic in countries where hepatitis E is not endemic . In the United States and Taiwan, where hepatitis E is not endemic in humans, the zoonotic spread of HEV is suspected, as the swine and human HEV isolates in each country belong to the same genotype and are closely related to each other and cross-species infection has been documented . In contrast, in India, where hepatitis E is endemic in humans, the human HEV isolates belong to genotype I, whereas the swine isolates belong to genotype IV . The potential zoonotic spread of HEV infection is also supported by several recent reports that veterinarians working with swine were at higher risk for HEV infection than healthy blood donors in the United States and other countries and that anti-HEV antibodies are highly prevalent in commercial swine populations in Australia, Canada, and New Zealand . Recently, evidence for widespread infection of wild rats with HEV in the United States was reported , leading to the speculation that wild rats might be involved in the high prevalence of anti-HEV among some U.S. city dwellers. An HEV-like agent has been recovered from chickens with big liver and spleen disease in Australia and from those with hepatitis-splenomegaly syndrome in the United States, but avian HEV is genetically related to, but clearly distinct from, known human and swine strains of HEV, displaying less than 60% nucleotide sequence identity . In Japan, only three strains of swine HEV of genotype III (swJ570, swJ681, and swJ791) have been isolated from farm pigs . Swine HEV strains of group B of genotype III and group C of genotype IV have not been recovered from pigs in Japan thus far. Therefore, further epidemiological evidence is required to prove the zoonotic spread of HEV from swine to humans, or vice versa, in Japan by means of isolation of completely identical strains from both humans and pigs. The high prevalence of anti-HEV in a number of other animal species may suggest that multiple sources of exposure to HEV may exist in the general population in industrialized countries who are not at apparent risk of exposure to HEV. In conclusion, 11 HEV isolates that are separated into three distinct groups within two major genotypes (III and IV) and that may be indigenous to Japan were identified from 87 patients with sporadic acute hepatitis of non-ABC etiology in Japan, where clinical HEV infection is rare. Our present study revealed that polyphyletic HEV strains of genotypes III and IV cocirculate in Japan and contribute to the development of sporadic acute hepatitis, with higher prevalences in males, in those over 40 years of age, and in patients living in the northern part of Japan. Whether the domestic spread of HEV infection is via zoonosis and whether the HEV genotype affects the pathogenesis and outcome of HEV infection deserve further analysis. Backmatter: PMID- 12202568 TI - Rapid Detection of Methicillin Resistance in Coagulase-Negative Staphylococci with the VITEK 2 System AB - The aim of the present study was to evaluate the accuracy of the new VITEK 2 system (bioMerieux, Marcy l' Etoile, France) for the detection of methicillin resistance in coagulase-negative staphylococci (CoNS) by using AST-P515 and AST-P523 test cards. Analyses of the VITEK 2 oxacillin MIC determination evaluated according to the actual breakpoint (>=0.5 mug/ml) of the National Committee for Clinical Laboratory Standards resulted in a high sensitivity of 99.2% but a moderate specificity of 80%. The newly included oxacillin resistance (OR) test of the VITEK 2 system displayed a high sensitivity and a high specificity of 97.5 and 98.7%, respectively. Concordance between the results of the mecA PCR and the VITEK 2 oxacillin MIC was observed for almost all Staphylococcus epidermidis strains, but the reduced specificity was attributable to higher oxacillin MICs for mecA-negative non-S. epidermidis strains, especially S. saprophyticus, S. lugdunensis, and S. cohnii. Evaluation of alternative oxacillin MIC breakpoints of 1, 2, or 4 mug/ml resulted in improved degrees of specificity of 84, 90.7, and 97.3%, respectively. Only minor changes occurred in the corresponding sensitivity values, which were 98.4, 97.5, and 97.5%, respectively. Methicillin resistance in CoNS was detected after 7 and 8 h in 91.1 and 93.5% of the mecA-positive strains, respectively, by the VITEK 2 OR test and in 86.3 and 89.5% of the mecA-positive strains, respectively, by VITEK 2 oxacillin MIC determination. After 7 and 8 h the VITEK 2 OR test classified 59.2 and 78.9% of the mecA-negative strains, respectively, as susceptible to oxacillin, whereas comparable values were obtained 2 h later by VITEK 2 oxacillin MIC determination. The results of our study encourage the use of the VITEK 2 system, which proved to be a highly reliable and rapid phenotypic method for the detection of methicillin resistance in CoNS. Keywords: Introduction : Coagulase-negative staphylococci (CoNS) rank among the five most frequent causative organisms of nosocomial infections, which are regularly associated with biomedical implants . Antibiotic therapy of infections caused by CoNS is increasingly problematic due to frequent multiple-antibiotic resistance . Most importantly, the majority of clinical CoNS harbor the mecA gene encoding an additional penicillin binding protein (PBP) 2a (PBP 2a) . Phenotypic detection of methicillin resistance in CoNS is difficult due to the heterogeneous expression of mecA . Detection of mecA by PCR is very sensitive and is considered the "gold standard," but it is not feasible for the busy clinical microbiology laboratory. Detection of PBP 2a as a marker for methicillin resistance in CoNS is also rapid and reliable and has therefore been recommended as an alternative for mecA PCR by the National Committee for Clinical Laboratory Standards (NCCLS) . The VITEK 2 system (bioMerieux, Marcy l'Etoile, France) is a fully automated system for rapid identification and antimicrobial susceptibility testing and is reliable for many rapidly growing bacteria (, -, , , , -, , , , ) and yeasts . In the earlier generation of the system, the Vitek system, detection of methicillin resistance was solely based on oxacillin MIC determination. To enhance the sensitivity and the specificity of the system, an oxacillin resistance (OR) test was introduced into the new VITEK 2 system as a second assay. The composition of the VITEK 2 OR test is comparable to that of the oxacillin agar screen. Recently, detection of methicillin resistance in Staphylococcus aureus with the VITEK 2 system was investigated , but detection of methicillin resistance in clinical CoNS isolates has not been studied in detail. In the present study detection of mecA by PCR was compared to the results of the OR test and oxacillin MIC determination with the VITEK 2 system for 200 isolates of CoNS. The time until detection of methicillin resistance was analyzed for both tests in parallel. (Part of this work will appear in the doctoral thesis of M.A.H., Universitatsklinikum Hamburg-Eppendorf, Hamburg, Germany.) MATERIALS AND METHODS : Bacterial isolates. | The CoNS strains (n = 200) belong to a collection of clinical isolates which were consecutively collected at the University Hospital Hamburg-Eppendorf in 1997 and 1998 and include 13 different species (140 S. epidermidis, 16 S. haemolyticus, 10 S. hominis, 9 S. saprophyticus, 6 S. capitis, 4 S. lugdunensis, 4 S. warneri, 4 S. xylosus, 2 S. schleiferi, 2 S. cohnii, 1 S. chromogenes, 1 S. simulans, and 1 S. kloosii isolates), as described previously . S. aureus ATCC 29213 (mecA negative) and S. aureus ATCC 43300 (mecA positive) were used for quality control. The strains were kept at -80C in microbank tubes (Pro Lab Diagnostics, Richmond Hill, Ontario, Canada) and were subcultured twice onto Columbia agar (Difco, Becton Dickinson, Sparks, Md.) plates containing 5% sheep blood in ambient air at 37C before testing. Antimicrobial susceptibility testing with the VITEK 2 system. | The phenotypes of the strains were determined by detection of methicillin resistance by the oxacillin resistance (OR) test (categories of susceptible or resistant) and oxacillin MIC determination. Both tests are included in the AST-P515 antimicrobial susceptibility test cards and in the newer AST-P523 antimicrobial susceptibility test cards. The AST-P523 cards differ from the AST-P515 cards only in their lower oxacillin MIC range (<=0.25 to >=4 mug/ml versus <=0.5 to >=8 mug/ml) and not in the composition of the tests examined. The instructions of the manufacturer for antibiotic susceptibility testing with the VITEK 2 system were essentially followed. Reading of test cards was performed every 15 min by the VITEK 2 system, and the results for the VITEK 2 OR test and the VITEK 2 oxacillin MIC determination were recorded after every reading for up to 10 h for each isolate. Initially, all CoNS isolates were tested with AST-P515 cards. CoNS strains for which the VITEK 2 oxacillin MIC was <=0.5 mug/ml were additionally tested with the AST-P523 card, allowing determination of lower oxacillin MICs. CoNS isolates with discrepancies between the results of the PCR for mecA and the VITEK 2 tests were reexamined. If an initial oxacillin MIC of <=0.5 or 1 mug/ml was noted, retesting was performed with AST-P523 cards, whereas AST-P515 cards were used for strains for which the initial oxacillin MIC was >=2 mug/ml. On retesting the same procedure was used as described above, and neither method was voluntarily changed. PCR for mecA. | PCRs were performed essentially as described previously . RESULTS : MecA was detected in 124 of 200 (62%) strains comprising 99 of 140 (70.7%) mecA-positive S. epidermidis strains and 25 of 60 (41.7%) mecA-positive non-S. epidermidis strains. The VITEK 2 oxacillin MIC was >=8 mug/ml for 119 of 124 mecA-positive strains . Lower VITEK 2 oxacillin MICs ranging from <=0.25 to 2 mug/ml were found for three mecA-positive strains (one S. epidermidis strain, one S. hominis strain, and one S. haemolyticus strain (Tables and ). With the AST-P515 cards, for 63 of 76 mecA-negative strains the oxacillin MIC was <=0.5 mug/ml . For 12 mecA-negative strains, the VITEK 2 oxacillin MICs ranged from 1 to >=8 mug/ml . The 63 strains for which the initial VITEK 2 oxacillin MIC was <=0.5 mug/ml with the AST-P515 cards were additionally tested with AST-P523 cards. An oxacillin MIC of <=0.25 mug/ml was detected for 60 strains, but for 3 mecA-negative strains (1 S. saprophyticus strain and 2 S. lugdunensis strains) the oxacillin MIC was 0.5 mug/ml (Tables and ). The VITEK 2 OR test with the AST-P515 card detected 119 of 124 mecA-positive strains but failed to indicate methicillin resistance in the 3 mecA-positive strains (1 S. epidermidis strain, 1 S. hominis strain, and 1 S. haemolyticus strain), for which VITEK 2 oxacillin MICs were <=0.25 to 2 mug/ml (Tables and ). All mecA-negative strains, except 1 S. kloosii strain for which the VITEK 2 oxacillin MIC was >=8 mug/ml, were correctly classified as susceptible by the VITEK 2 OR test (Tables and ). Discrepancies between the results of the VITEK 2 OR test and the VITEK 2 oxacillin MIC determination were observed for 14 mecA-negative strains (6 S. saprophyticus, 4 S. lugdunensis, 2 S. cohnii, 1 S. xylosus, and 1 S. epidermidis strains), with VITEK 2 oxacillin MICs ranging from 0.5 to 4 mug/ml but with susceptibility indicated by the VITEK 2 OR test (Tables and ). The results for these 14 strains were categorized as indeterminate (Tables and ). Additionally, the above-mentioned mecA-positive S. epidermidis (n = 1) and mecA-positive S. hominis (n = 1) strains for which the VITEK 2 oxacillin MICs were 1 and 0.5 mug/ml, respectively, but which were susceptible by the VITEK 2 OR test were categorized as indeterminate (Tables and ). Retesting of the 16 strains categorized as indeterminate confirmed the initial VITEK 2 OR test result for all strains and the VITEK 2 oxacillin MIC did not vary significantly . Insufficient growth in the AST-P515 cards was observed for three strains including one mecA-negative S. epidermidis strain and two mecA-positive S. saprophyticus strains. On retesting of these strains with AST-P515 cards, one mecA-positive S. saprophyticus strain again did not grow, whereas susceptibility to oxacillin was correctly determined for the other mecA-positive S. saprophyticus strain and the mecA-negative S. epidermidis strain (Tables and ). Neither S. saprophyticus strain grew with AST-P523 cards . Among the mecA-positive strains, resistance to oxacillin was detected after 7, 8, and 9 h by the VITEK 2 OR test in 91.1, 93.5, and 95.2% of the strains, respectively , and by VITEK 2 oxacillin MIC determination in 86.3, 89.5, and 91.9% of the strains, respectively . However, among the mecA-negative strains, susceptibility to oxacillin was detected after 7, 8, and 9 h by the VITEK 2 OR test in 59.2, 78.9, and 86.8% of the strains, respectively , and by VITEK 2 oxacillin MIC determination in 0, 1.3, and 59.2% of the strains, respectively . FIG. 1. | Time course of results of VITEK 2 oxacillin MIC determination and VITEK 2 oxacillin resistance test for mecA-positive CoNS strains (A) and mecA-negative CoNS strains (B). Time course of results of VITEK 2 oxacillin MIC determination and VITEK 2 oxacillin resistance test for mecA-positive CoNS strains (A) and mecA-negative CoNS strains (B). TABLE 1 | Comparison of the oxacillin MIC determination and the OR test of the VITEK 2 system with the results of the mecA PCR TABLE 2 | Discrepant and indeterminate results of VITEK 2 tests compared to the results of the mecA PCR DISCUSSION : In the Vitek system, detection of methicillin resistance was solely based on oxacillin MIC determination. Three recent studies reported high sensitivities (range, 95.7 to 100%) but only a moderate degree of specificity (range, 61 to 85.5%) for the Vitek system with CoNS . To improve the rate of detection of methicillin resistance in staphylococci, the OR test was implemented in the new VITEK 2 system. Recently, VITEK 2 oxacillin MIC determination performed with excellent results for the sensitive and specific detection of methicillin-resistant S. aureus . So far, analyses of methicillin resistance in CoNS have been performed with only a small number of selected mecA-positive S. epidermidis, S. hominis, and S. warneri strains, with overall agreement with the PCR results . When our study was evaluated according to the actual NCCLS oxacillin breakpoint of >=0.5 mug/ml, VITEK 2 oxacillin MIC determination correctly classified 119 of 124 mecA-positive CoNS strains as resistant and 60 of 76 mecA-negative strains as susceptible, which resulted in a high sensitivity of 99.2% but a moderate specificity of 80%. When the older NCCLS oxacillin breakpoint of >=4 mug/ml is used in the evaluation, VITEK 2 oxacillin MIC determination again correctly classified 119 of 124 mecA-positive strains as resistant but 73 of 76 mecA-negative strains as susceptible, resulting in an excellent sensitivity of 97.5% and a specificity of 97.3%. Calculations based on oxacillin MIC breakpoints of 1 and 2 mug/ml yielded sensitivities of 98.4 and 97.5%, respectively, and specificities of 84 and 90.7%, respectively. Detection of methicillin resistance in CoNS solely on the basis of the results of the VITEK 2 OR test was superior to VITEK 2 oxacillin MIC determination, with a sensitivity of 97.5% and a specificity of 98.7%. Remarkably, 14 of the 16 strains with susceptibility by the VITEK 2 OR test but for which the VITEK 2 oxacillin MIC was above the actual NCCLS breakpoint were non-S. epidermidis strains, consisting of 6 S. saprophyticus, 4 S. lugdunensis, 2 S. cohnii, 1 S. xylosus, and 1 S. hominis strains but only 2 S. epidermidis strains. Another characteristic feature of strains with indeterminate results is the predominance of novobiocin-resistant species for 9 of 16 strains. Only 2 strains (1 S. epidermidis strain and the 1 S. hominis strain) classified as indeterminate were in fact mecA positive, but 14 of 16 strains were mecA negative (Table and ). Only one mecA-negative S. kloosii strain displayed a false-positive result, with a VITEK 2 oxacillin MIC of >=8 mug/ml and resistance by the VITEK 2 OR test (Tables and ). The molecular basis of decreased susceptibility to oxacillin in mecA-negative CoNS warrants further investigations, but alterations in PBPs other than PBP 2a have been documented in S. haemolyticus and S. saprophyticus and were suggested as one possible explanation . A false-negative result was observed for a mecA-positive S. haemolyticus strain for which the VITEK 2 oxacillin MIC was <=0.5 mug/ml (oxacillin MIC on retesting, <=0.25 mug/ml) and which was also reported to be susceptible by the VITEK 2 OR test . The results of our study are comparable to those of recent studies in which an excellent specificity of the new breakpoint was found when it was applied to S. epidermidis, S. hominis, and S. haemolyticus, but the new breakpoint was less accurate when it was applied to S. saprophyticus, S. cohnii, S. warneri, S. lugdunensis, and S. xylosus . In the Vitek system, false-positive oxacillin MIC results have been observed predominantly for mecA-negative S. saprophyticus or S. lugdunensis strains for which oxacillin MICs are 0.5 or 1 mug/ml. Divergent results were also mentioned for S. cohnii, S. sciuri, and S. capitis strains . To enhance the specificity of the Vitek system, exclusion of S. saprophyticus and S. lugdunensis strains has been proposed . This proposal correlates well with the results obtained for a collection of 83 mecA-negative S. saprophyticus strains, with the oxacillin MIC for the majority of strains being 0.5 mug/ml . In that study, the oxacillin MIC results obtained with the Vitek system and by the broth microdilution and agar dilution methods were in good agreement. However, on the basis of the actual NCCLS breakpoint (>=0.5 mug/ml), oxacillin MIC determination with the Vitek system and by broth microdilution and agar dilution would have classified as resistant 81 of 83, 76 of 83, and 81 of 83 mecA-negative strains, respectively . In the VITEK 2 system, the OR test leads to the specific detection of methicillin resistance; however, indeterminate results occur with the OR test, indicating sensitivity, and VITEK 2 oxacillin MIC determination, indicating resistance. If indeterminate results occur, we recommend that the results of the VITEK 2 OR test be followed. An alternative strategy would be to deduce susceptibility to oxacillin from the species identification or at least resistance to novobiocin because we observed the most indeterminate results for mecA-negative strains belonging to the S. saprophyticus group and S. lugdunensis (Tables and ). More isolates of the respective species must be analyzed to reach generalized conclusions, but recently, the NCCLS has also emphasized that the current interpretative criteria for the oxacillin MIC could be misleading, especially for S. saprophyticus and S. lugdunensis strains . The VITEK 2 system rapidly reported the resistance results obtained by oxacillin MIC determination and the OR test: after 8 h for about 80% of the strains . Compared to conventional phenotypic methods, which would take up to 48 h, this substantial acceleration of the diagnostic capability with the VITEK 2 system may have a potential impact on the optimal antibiotic management of CoNS infections. Rapid antimicrobial susceptibility testing (RAST) was reported to have an important clinical impact due to decreased rates of mortality . Additionally, financial benefits are attributable to reduced laboratory pharmacy and other general costs, which were significantly lower for the RAST groups . Concerning CoNS infections, rapid and reliable detection of methicillin resistance in CoNS is highly desirable to reduce the overuse of glycopeptides for the treatment of CoNS infections. Widespread use of glycopeptides risks increasing the numbers of infections caused by vancomycin-resistant enterococci and causing the emergence of glycopeptide resistance in S. aureus . In the present study the combination of the VITEK 2 OR test and VITEK 2 oxacillin MIC determination proved to be a highly sensitive, specific, and rapid procedure at least equivalent to other phenotypic techniques for the detection of methicillin resistance in CoNS, like the oxacillin spread plate technique, the method with oxacillin disk on NaCl-supplemented Mueller-Hinton agar, or the broth microdilution method. The results of our study encourage the use of the VITEK 2 system for the detection of methicillin resistance in CoNS. Backmatter: PMID- 12202571 TI - Determination of Novel Borrelia Genospecies in Swedish Ixodes ricinus Ticks AB - A total of 301 adult questing Ixodes ricinus ticks were collected at 15 different locations along the south and east coasts of Sweden to determine the Borrelia genospecies diversity. Thirty-two ticks (11%) were found to be positive by nested PCR with Borrelia burgdorferi sensu lato-specific primers. Species determination was based on partial sequencing of the 16S rRNA gene and the flagellin gene. Five different Borrelia species were found. The nucleotide sequence of the Borrelia DNA found in two ticks differed extensively from the nucleotide sequences of the Borrelia DNA found in the other ticks, and analysis revealed that they were closely related to the relapsing fever borrelia species Borrelia miyamotoi. This is the first report of a B. miyamotoi-like borrelia in I. ricinus and in Europe. Moreover, the Borrelia DNA of two ticks (6%) clustered within the B. valaisiana complex. B. valaisiana has not previously been reported in Sweden. B. afzelii DNA was found in 14 ticks (44%), and B. garinii DNA was found in 10 ticks (31%). B. burgdorferi sensu stricto DNA was found in four ticks (13%). We conclude that all of the known human-pathogenic species (B. garinii, B. afzelii, and B. burgdorferi sensu stricto) and B. valaisiana found elsewhere in Europe are also present in the Swedish host-seeking tick population and that a B. miyamotoi-like Borrelia species seems to be present in I. ricinus ticks in Europe. Keywords: Introduction : Ticks transmit many clinically important pathogens of the genus Borrelia. These spirochetes are responsible for two groups of human disease: Lyme borreliosis (LB) and relapsing fever (RF). LB is the most prevalent tick-borne zoonosis in Europe and North America and affects the skin, joints, heart, and nervous system . Borreliae that cause LB are transmitted by hard ticks of the genus Ixodes. In Europe the principal vector is Ixodes ricinus. The disease is caused by spirochetes of the Borrelia burgdorferi sensu lato complex, which consists of 10 different named genospecies. Three species, all found in Europe, are known to be pathogenic for humans: B. burgdorferi sensu strico, B. garinii , and B. afzelii . Another two species, B. valaisiana and B. lusitaniae , have been isolated from European ticks. The pathogenic capabilities of the last two species are still uncertain, although B. valaisiana DNA has been amplified by PCR from patients with LB . Two additional Borrelia species have been found in European patients with LB; B. bissettii sp. nov. has been isolated from patients in Slovakia , and a novel Borrelia species has been isolated from a patient in The Netherlands . There has been an increasing interest in the clinical and diagnostic implications of the different Borrelia species, since an association between the clinical manifestations of LB and the infective species has been suggested . The infective Borrelia species also influences the immune response . Tick-borne RF, with periodic febrile episodes as the main symptom, is caused by a genetically and ecologically different group of Borrelia species. RF is rarely seen in Europe but is reported in the most southern parts of Europe . Borrelia species that cause tick-borne RF are usually considered vector species specific and are mostly transmitted by soft ticks (family Argasidae) of the genus Ornithodoros . Two RF-associated Borrelia species are exceptions and are found in hard ticks: B. lonestari, which is transmitted by Amblyomma americanum in North America , and B. miyamotoi, which is isolated from Ixodes persulcatus in Japan . Different methods are used for Borrelia species determination. PCR detection and subsequent sequencing of the 16S rRNA gene is considered a sensitive and reliable method . Sequencing of the flagellin gene gives additional taxonomic data . Since information on the Borrelia genospecies present in the tick population is essential to our understanding of the epidemiology, clinical spectrum, diagnosis, and prevention of LB, we conducted this study to determine the diversity of B. burgdorferi sensu lato among ground host-seeking ticks in Sweden. MATERIALS AND METHODS : Study area and tick collection. | During the summer of 1999 questing adult ticks were collected by flagging at 15 different locations with mixed vegetation along the south and east coasts of Sweden. A total of 301 adult unfed I. ricinus ticks were collected. Twenty-one ticks were collected at two locations in the province of Skane, 233 ticks were collected at nine different locations in the province of Blekinge (108 ticks were collected from a single location), 24 ticks were collected at two locations in the province of Kalmar, 16 ticks were collected in the proximity of Stockholm, and 7 ticks were collected farther north at a location in the proximity of Gavle . Of the ticks collected, 165 (55%) were male and 136 (45%) were female. The ticks were placed into coded tubes and stored at -20C until September 2000. DNA extraction. | The ticks were processed individually. Each tick was washed in 70% ethanol and cut in half sagittally on a glass slide with a drop of phosphate-buffered saline. One half was saved for future use, and the other half was crushed and transferred to a test tube (Eppendorf; 1.5 ml) for DNA extraction. A QIAamp tissue kit (Qiagen) was used for DNA extraction according to the protocol of the manufacturer, with a few modifications. Samples were incubated overnight with proteinase K solution and eluted twice with 100 mul of AE buffer in order to increase the DNA yield. Purified DNA was stored at -20C. PCR amplification. | For detection of Borrelia-infected ticks, the 16S rRNA sequence was amplified by a nested PCR. Primers16S-F and 16S-R were used in the first amplification . The master mixture contained 0.5 muM each primer, 10 mM Tris-HCl, 1.5 mM MgCl2, 50 mM KCl (pH 8.3), 0.2 muM each deoxynucleoside triphosphate, and 1.25 U of Taq DNA polymerase. The reaction volume was set to 50 mul containing 5 mul of sample, and the amplification was conducted by using a program of 94C for 2 min, followed by 35 cycles of 94C for 45 s, 55C for 60 s, 72C for 90 s and finally 72C for 7 min in a Perkin-Elmer 9600 thermocycler. Primers LD-1 and LD-2 were used in the second reaction. The master mixture described above was used, except that 0.2 muM each primer was added and 2 mul of the PCR product from the first PCR was used as the template in the reaction. The volume was set to 50 mul, and amplification was conducted by using a program of 94C for 2 min, followed by 40 cycles of 94C for 45 s, 56C for 60 s, 72C for 60 s and finally 72C for 7 min. The products were visualized by electrophoresis in a 1.5% agarose gel stained with ethidium bromide. A negative control and a positive control were included in all PCR runs. Contamination was minimized by performing the different steps in separate rooms. Gloves and filter tips were always used. Nucleotide sequence analysis. | Partial sequencing of the 16S rRNA gene was performed for all positive samples. The products were extracted with a QIAquick gel extraction kit (Qiagen) according to the protocol of the manufacturer and sequenced by the BIG DYE method (Applied Biosystems) by direct sequencing by PCR. Each strand was analyzed with an ABI 3100 instrument (Applied Biosystems) by the Biomolecular Resource Facility at Lund University. Primers 16S-F and 569r (5'-TACGCCCAATAATCCCGAACAAC-3') were used. The flagellin gene in four of the samples was partially sequenced by use of primer flaC (5'-ATTGAAAT AGAGCAACTTACAGA-3') and primer flaL4 (5'-TTATCTAAGCAATGACAAAACATAT-3'). The DNA sequences were compared with the Borrelia gene sequences registered in the GenBank database. BioEdit software (T. A. Hall, Nucleic Acids. Symp. Ser. 41:95-98, 1999) was used for analysis of the results. Nucleotide sequence accession numbers. | The 16S rRNA sequences of the borreliae from the following Borrelia strains determined in this study have been deposited in GenBank and given the indicated accession numbers: Ri11, ; St1, ; To60, ; To72, ; Ku10, ; To76, ; Al10, ; As15, ; Ga2, ; Os8, ; Ri13, ; St12, ; St13, ; St8, ; Tr2, ; Ri9, ; Al12, ; Al16, ; As18, ; As7, ; Ha3, ; Ha5, ; Os10, ; Os2, ; Osk2, ; Osk3, ; Sa5, ; To106, ; To16, ; To89, ; Na34, ; and St4, . The flagellin gene sequences of the borreliae from the following Borrelia strains determined in this study have been deposited in GenBank and given the indicated accession numbers: St4, ; Na34, ; To76, ; and Ku10, . FIG. 1. | Map of Sweden showing the locations of tick collection. Map of Sweden showing the locations of tick collection. RESULTS : Borrelia PCR. | Of the 301 ticks, 32 (11%) were positive by use of the B. burgdorferi sensu lato-specific primers in the nested PCR. The PCR products of two of the tick samples yielded atypical bands by gel electrophoresis, but the two samples were still considered positive. There were no differences in the prevalences of Borrelia spp. between male and female ticks. Borrelia nucleotide sequence analysis. | Species determination was made by DNA nucleotide sequence analysis, and the sequences were compared with those for previously reported strains available in the GenBank database. Partial sequencing of the 16S rRNA gene, with a minimum of 496 bp analyzed, was performed for all 32 Borrelia 16S rRNA gene-positive ticks, and the results are summarized in Table . The flagellin gene was partially sequenced (a minimum of 608 bp was analyzed) in order to confirm and specify the findings for four of the Borrelia-positive ticks (Na34, St4, To76, and Ku10). Five different Borrelia species were identified in the ticks examined. DNA from two of the ticks (Na34 and St4, collected in Blekinge and Stockholm, respectively) yielded atypical PCR bands by gel electrophoresis (the strongest band was about 1,500 bp, and no LD primer-specific band was seen). The sequences of the 16S rRNA and flagellin genes obtained from the two ticks were identical but differed extensively from the sequences of the genes from the other ticks. Analysis indicated that they are closely related, but not identical, to the previously reported sequences of the genes from B. miyamotoi strains . B. afzelii DNA was found in 14 ticks (44%), B. garinii DNA was found in 10 ticks (31%), and both species were widely distributed geographically. B. burgdorferi sensu stricto DNA was found in four ticks (13%) collected in Blekinge and Stockholm. The DNA of the 16S rRNA and flagellin gene sequenced in two ticks (6%), collected in Blekinge (from two different locations), belonged to the B. valaisiana genomic group. A phylogenetic tree was constructed from the DNA sequence data obtained (Fig. and B). FIG. 2. | (A) Phylogenetic tree, based on a comparison of the 16S rRNA sequences (456 bp) of Borrelia species, obtained by the maximum-likelihood method. (A) Phylogenetic tree, based on a comparison of the 16S rRNA sequences (456 bp) of Borrelia species, obtained by the maximum-likelihood method. B. hermsii, B. parkeri, B. recurrentis, B. turicatiae, B. miyamotoi, B. lonestari, B. burgdorferi, B. valaisiana, B. afzelii, and B. garinii and their GenBank accession numbers are included for comparison purposes. (B) Phylogenetic tree, based on a comparison of the flagellin gene sequences (581 bp) of Borrelia species, obtained by the maximum-likelihood method. TABLE 1 | Analysis of signature nucleotide positions of the partial 16S rRNA in the Borrelia strains studied DISCUSSION : The genetic diversity of Borrelia species in the tick population has not previously been studied in Sweden. In this study five different Borrelia species were identified. We obtained Borrelia DNA from two I. ricinus ticks that were not closely related to any of the B. burgdorferi sensu lato genospecies. Phylogenetic analysis of the flagellin and 16S rRNA sequences indicated that they were closely related to the B. miyamotoi genospecies, previously isolated only in Japan from I. persulcatus ticks . B. miyamotoi has not yet been shown to cause disease in humans and is phylogenetically most closely related to B. lonestari among the relapsing fever borreliae . B. lonestari is transmitted by the hard A. americanum tick (classified within the Metastriata subfamily) and can cause an erythema migrans-like rash . B. lonestari is also closely related to the agent of bovine borreliosis, Borrelia theileri, which is transmitted by hard ticks from the Metastriata subfamily . Phylogenetic analysis distinguishes two separate clusters within the group of RF borreliae that are transmitted by ixodid ticks: the Metastriata tick-transmitted species B. lonestari and B. theileri cluster together, and the B. miyamotoi-like borreliae transmitted by Ixodes spp. form a separate cluster. The present study is the first report of B. miyamotoi-like Borrelia species in Europe and in I. ricinus ticks. We found B. miyamotoi-like DNA in two ticks collected 400 km apart, which indicates that the findings were not isolated but, rather, that the species is geographically widespread. The monophyletic cluster of B. miyamotoi-like borreliae thus seems to be distributed on two continents and transmitted by two subspecies of Ixodes ticks. Although genetically classified within the RF borreliae, B. miyamotoi seems to have ecological characteristics of a borrelia of the B. burgdorferi sensu lato complex. Further studies are needed to characterize the biological properties of this Borrelia species. Isolation and more records are needed to confirm its presence in the European tick population. The primers used in this study were not designed to find any borreliae other than those from the B. burgdorferi sensu lato complex. The tick templates in which the B. miyamotoi-like DNA was found reacted with the primers, but the PCR products were of a different size. The published data do not enable analysis of how the 16S rRNA-specific primers correspond to the 16S rRNA sequence of B. miyamotoi. The 3' ends of the LD-1 and LD-2 primers differ at 3 and 2 bp, respectively, from the sequence of the B. miyamotoi type strain (strain HT31). Of the B. burgdorferi sensu lato species found, B. afzelii and B. garinii dominated, with 44 and 31% of the Borrelia isolates being of these two species, respectively, which is consistent with clinical findings in Sweden and with reports from Europe . B. burgdorferi sensu stricto has just recently been found in Swedish LB patients by PCR and has previously been reported from Ixodes ticks collected from migrating birds arriving in Sweden . B. valaisiana has been reported from a number of European countries , but it has never before been reported in Sweden. We found a Borrelia prevalence of 11% in the Swedish host-seeking tick population by use of the PCR technique. Previous studies have used immunofluorescence analysis , phase-contrast microscopy , or dark-field microscopy ; and prevalences ranging between 7 and 29% have been recorded in adult ticks from locations in southern and central Sweden. Significant spatial and seasonal variations in the prevalence of Borrelia-infected ticks have been observed previously ; and the year, the location of sampling, and the method used for analysis may account for the differences between the reported prevalences. We conclude that all of the known pathogenic species (B. garinii, B. afzelii, and B. burgdorferi sensu strico) and B. valaisiana found elsewhere in Europe are also present in the Swedish host-seeking tick population. Of particular interest is our finding of B. miyamotoi-like Borrelia in Europe. Backmatter: PMID- 12202616 TI - Enteroaggregative Escherichia coli Strains among Classical Enteropathogenic Escherichia coli O Serogroups AB - Keywords: Enteroaggregative Escherichia coli (EAEC) is an emerging cause of diarrhea in both developing and developed countries that is defined by the characteristic pattern of aggregative adherence (AA) to cultured epithelial cells . Studies from various laboratories have demonstrated that some enteropathogenic E. coli (EPEC) serogroups are comprised of serotypes mostly consisting of EAEC strains . These serotypes present interesting characteristics: i.e., they have been isolated from cases of acute infantile diarrhea before AA was recognized as a distinct adherence pattern and in most cases express AA only in the 6-h adherence assay . We examined our laboratory collection of E. coli strains that present the AA phenotype to HEp-2 cells and belong to the EPEC O serogroups (L. R. Trabulsi, unpublished data) for the presence of EAEC-associated virulence markers . Thirty-four AA-producing strains of the following serotypes were selected: O86:H2, O111:H4, O111:H10, O111:H12, O111:H21, O125:H6, O125:H16, O125:H21, O126:H27, O128:H12, and O128:H35. The strains were first analyzed for the presence of intimin (eae) and bundle-forming pilus structural subunit (bfpA) genes of EPEC . PCR for eae was performed as described earlier . The primers employed for bfpA detection (forward, 5'-GGTCTGTCTTTGATTGAATC-3'; and reverse, 5'-TTTACATGCAGTTGCCGCTT-3') were based on its published sequence (GenBank accession no. ) in order to amplify a 485-bp fragment. The presence of plasmid (aggA, aafA, aggR, astA, pet, aspU, and shf) and chromosomal (irp2 and pic) EAEC-associated virulence markers and the EAEC probe sequence were also investigated by PCR, as previously described . All strains evaluated were negative for the presence of EPEC virulence genes, except the two O125:H6 strains, which harbored the eae gene. Interestingly, these latter strains lacked all EAEC markers evaluated and presented the AA pattern to HEp-2 cells in the 6-h adhesion assay. All other serotypes studied presented at least one EAEC marker . The most prevalent markers found were irp2, astA, shf, and pic, which have also been described as prevalent among EAEC strains . The aggregative adherence fimbrial I pilin (aggA) gene was the only EAEC marker not detected. This is in accordance with the low prevalence of aggA found in another study . TABLE 1 : Detection of EAEC- and EPEC-associated virulence markers among E. coli strains belonging to EPEC serogroups and presenting the AA phenotype EAEC is very heterogeneous regarding the presence of putative virulence markers, and although some of these markers are frequent in the category, a sensitive gene probe for EAEC identification is not yet available . Accordingly, adherence to epithelial cells remains the "gold standard" diagnostic test for the category . Our results demonstrated that, with the exception of O125:H6, which has been classified as atypical EPEC , the serotypes studied here should be considered to be EAEC, since they presented the AA pattern and the majority of the EAEC-associated virulence genes. Studies in our laboratories are in progress in order to characterize the adhesin mediating the AA phenotype of the O125:H6 serotype and to ascertain whether these strains are able to produce the characteristic attaching-and-effacing lesion of EPEC . Our findings have clinical and epidemiological relevance, since EPEC identification by O serogrouping is misleading, as EPEC serogroups include different pathogens. Therefore, O sero-grouping should be complemented by flagellar (H) typing or, alternatively, by testing adhesion to epithelial cells. Backmatter: PMID- 12202589 TI - Serodiagnosis of Imported Schistosomiasis by a Combination of a Commercial Indirect Hemagglutination Test with Schistosoma mansoni Adult Worm Antigens and an Enzyme-Linked Immunosorbent Assay with S. mansoni Egg Antigens AB - A commercial indirect hemagglutination (IHA) test using erythrocytes coated with Schistosoma mansoni adult worm antigens (WA) and an enzyme-linked immunosorbent assay (ELISA) with S. mansoni egg antigens (SEA) were assessed for their use in serodiagnosis of imported schistosomiasis (hereafter these tests are designated WA/IHA and SEA/ELISA, respectively). The sensitivity of the tests was evaluated with sera from 75 patients with proven S. mansoni infection, 25 with proven S. haematobium infection, and 10 with clinical Katayama fever. The specificity was assessed with sera from 283 patients with various parasitic, bacterial, viral, and fungal infections and sera containing autoimmune antibodies. Sensitivities of the WA/IHA with a cutoff titer of 1:160 (WA/IHA160) in detecting S. mansoni, S. haematobium, S. mansoni and S. haematobium combined, and clinical Katayama fever were 88.0, 80.0, 86.0, and 70.0%, respectively, with a specificity of 98.9%. The WA/IHA with a cutoff of 1:80 (WA/IHA80) showed sensitivities of 94.7, 92.0, 94.0, and 90.0%, respectively, with a specificity of 94.7%. The comparable values of SEA/ELISA were 93.3, 92.0, 93.0, and 50.0%, respectively, with a specificity of 98.2%. Combined use of ELISA and WA/IHA80 gave sensitivities of 100% for S. mansoni, S. haematobium, and S. mansoni and S. haematobium combined and 90% for Katayama fever. The specificity of this combination in detecting schistosomiasis was 92.9%. Combination of SEA/ELISA with WA/IHA160 gave sensitivities of 98.7, 96.0, 98.0, and 80% with a specificity of 97.2%. Our findings suggest that WA/IHA and SEA/ELISA are each sensitive and specific serological tests that are easy to use for the diagnosis of imported schistosomiasis. The combined use of these two tests enabled the serological diagnosis of schistosomiasis to be achieved with very high degrees of both sensitivity and specificity. Keywords: Introduction : Schistosomiasis is a major cause of morbidity and mortality and has been estimated to infect over 200 million people. An estimated 500- to 600 million people worldwide are still at risk of infection. The disease occurs mostly in the tropical regions, particularly in Africa, South America, and the Far East, and is endemic in 74 developing nations . Schistosomiasis is frequently imported into nonendemic areas by immigrants and travelers returning from the tropics . Cases of imported schistosomiasis are on the increase due to changes in travel destinations and habits of travelers while abroad . Most of these patients are asymptomatic, but recent infection can cause serious disease, such as Katayama fever or severe neurological complications involving the spinal cord . Diagnosis of schistosomiasis by detection of specific antibodies is likely to be more sensitive than the traditional method of diagnosis by detection of eggs in stool or urine . In imported infections, with only a few or no eggs being excreted, antibody detection may be the only means to diagnose schistosomiasis. In order to incorporate serodiagnostics in routine clinical laboratory practice, an easy to use, sensitive, and specific serological test is needed. Unfortunately only a few serological tests for schistosomiasis are commercially available, and still fewer have been evaluated for their diagnostic use. These and other difficulties (e.g., those inherent in antigen preparation for such tests) tend to restrict serodiagnosis in general to larger research centers. Serological tests which could, however, be used in routine clinical laboratories are an indirect hemagglutination (IHA) assay with adult Schistosoma mansoni worm antigens (WA) produced by Fumouze Laboratories (Levallois-Perret, France) (hereafter this assay is referred to as WA/IHA) and an enzyme-linked immunosorbent assay (ELISA) with S. mansoni soluble egg antigens (SEA) (hereafter referred to as SEA/ELISA). In this study we evaluated the results obtained with WA/IHA and SEA/ELISA and the combined results of both tests for their sensitivity and specificity for patients returning from the tropics with egg-positive S. mansoni and S. haematobium infections and individuals presenting at the clinic with Katayama fever. The specificity of the test was evaluated with patients with various other infections and autoimmune disorders. MATERIALS AND METHODS : Patients and sera. | Patients incorporated into this study attended The Academic Medical Centre, Amsterdam, The Netherlands; The Harbor Hospital and Institute of Tropical Diseases, Rotterdam, The Netherlands; and The Prince Leopold Institute of Tropical Medicine, Antwerp, Belgium. A total of 393 patients with the following characteristics were used. (i) One hundred patients had egg-proven schistosomiasis (75 patients with S. mansoni and 25 patients with S. haematobium). (ii) Ten patients had Katayama fever. All patients had recently visited a schistosoma-endemic African country. Patients presented with eosinophilia, fever, symptoms of myalgia, arthralgia, persistent dry cough, weight loss, or general malaise. In two cases S. mansoni eggs were observed in stool samples. (iii) Two hundred-forty patients had other infectious diseases. All patients had proven active infections or were recently infected. Included were patients with fascioliasis hepatica (3), hookworm infection (10), trichuriasis (16), strongyloidiasis (11), filariasis bancrofti (24), onchocerciasis (12), loiasis (10), hepatic amoebiasis (7), visceral leishmaniasis (5), malaria (9), toxoplasmosis (11), syphilis (8), borreliosis (9), human immunodeficiency virus infection (11), cytomegalovirus infection (12), Epstein-Barr infection (12), hepatitis A (22), hepatitis B (13), rubella (11), Coxsackie B virus infection (11), and aspergillosis (11). (iv) Twenty-three patients had autoimmune antibodies (12 patients with rheumatoid factor and 11 with anti-nuclear antibodies). (v) Twenty patients were healthy blood donors originating from The Netherlands. WA/IHA. | The IHA test kit sold by Fumouze Laboratories was used according to the manufacturer's instructions, with a modification that the U-shaped plates provided by the manufacturer were replaced with V-shaped microtiter plates (Greiner Laboratories, Alphen, The Netherlands), as the latter allowed better readability of the results. Absence of agglutination was observed with the V-shaped well as a clear, sharp dark spot instead of the more or less wide ring seen with the original U-shaped well. Briefly, the test procedure was as follows. Fifty microliters of a 1:20 initial dilution of each serum was subjected to further twofold serial dilutions, and 10 mul of sheep red blood cells sensitized with S. mansoni adult WA was added to each diluted sample. Positive and negative control sera and nonsensitized red blood cells were included in each test as controls for naturally occurring antibodies. After incubation for 2 h at room temperature the titer in the test serum was recorded as one dilution before that which yielded a clear, sharp dark spot similar to those in the negative control wells. Titers were expressed as reciprocal values. All sera were tested in duplicate. The results were evaluated with a cutoff titer of 1:160 as recommended by the manufacturer and also with a cutoff of 1:80. SEA/ELISA. | SEA was prepared by retrieving eggs from the tissues of mice heavily infected with an S. mansoni isolate from Puerto Rico. The origin of the parasites, details of their laboratory maintenance, and methods for isolation of eggs have been described previously . The eggs were homogenized by disrupting them for 3 min in an ice-cooled glass homogenizer with a mechanically rotated Teflon plunger and were centrifuged at 20,000 x g for 3 h, and the supernatant was used as the antigen in the ELISA. The ELISA for detection of serum immunoglobulin G against SEA was performed as described previously , the concentrations of antigen and the dilution of serum samples and anti-human immunoglobulin G peroxidase conjugate having been determined by checkerboard titration . However, polystyrene high-binding and flat-bottom enzyme immunoassay/radioimmunoassay plates (Costar, Corning, N.Y.) with small wells were used. The reaction volume of each reagent at each step in these plates was 30 mul per well. Wells were sensitized with SEA to make a concentration of 1.3 mug/ml with 0.05 M carbonate-bicarbonate buffer, pH 9.6. The sensitization was performed overnight at 4C. After removal of the unbound antigen, the wells were treated with 1% chicken egg white in phosphate-buffered saline (PBS) at 37C for 30 min. Wells were washed five times with 0.05% Tween 20 in PBS, and serum samples (1:400 dilution) were added to the wells and incubated for 30 min at 37C. Each serum was tested in duplicate. After washing the wells five times they were filled with a 1:1,500 dilution of horseradish peroxidase-conjugated goat antibody to human immunoglobulin G (Nordic, Tilburg, The Netherlands). Following incubation for 30 min at 37C the wells were washed three times and substrate solution (0.01% 5-amino-2-salicyclic acid in phosphate buffer, pH 5.95, with 0.03% H2O2) was applied to the wells for 1 h in the dark at room temperature. The optical density (OD) at 492 nm was read in an ELISA reader designed by Labsystems (Helsinki, Finland). To correct for day-to-day assay variation the results were expressed as ratios between the absorbance value of the sample and that of a well-defined control serum. The cutoff value of 0.222 was defined as the mean OD plus two standard deviations of the serum samples from the 283 controls (240 patients with other infectious diseases, 23 with autoantibodies, and 20 healthy Dutch blood donors). Definitions. | The WA/IHA was considered positive if the titer of the serum was equal to or greater than the cutoff titers of 1:80 or 1:160. The SEA/ELISA was considered positive if the OD value of the serum was equal to or greater than the cutoff OD of 0.222. The combination of both tests was deemed positive if either one or both of the tests gave a positive reaction. Negative reactions in WA/IHA and SEA/ELISA showed lesser values than the cutoff values. A sample was considered negative if both tests had given negative reactions. The sensitivity of the WA/IHA, the SEA/ELISA, and of the combination of both tests was defined as the number of patients who gave a positive test result as a proportion of the total number of patients who had parasitologically proven schistosomiasis or clinical Katayama fever. The specificity of the tests and of the combination of the two tests in detecting schistosomiasis was defined as the number of patients who gave a negative test result as a proportion of the total number of control patients (other infections, autoimmune antibodies, and healthy Dutch blood donors). RESULTS : Sensitivities of WA/IHA, SEA/ELISA, and the combination of both tests in detecting S. mansoni, S. haematobium, S. mansoni and S. haematobium combined, and clinical Katayama fever and the specificities are shown in Table . Individual titers determined by WA/IHA and OD values determined by SEA/ELISA from schistosomiasis cases, clinical Katayama fever cases, and control cases are shown in Fig. and , respectively. Sensitivity of WA/IHA with a cutoff titer of 1:80 (WA/IHA80) for individuals excreting eggs of S. mansoni, S. haematobium, both infections combined, and with clinical Katayama fever ranged from 90.0 to 94.7%. Sensitivity of WA/IHA with a cutoff titer of 1:160 (WA/IHA160) for detection of egg-positive cases and cases with clinical Katayama fever was lower and ranged from 70.0 to 88.0%. WA/IHA gave higher sensitivity in detecting S. mansoni than in detecting S. haematobium. The specificity in detecting combined S. mansoni and S. haematobium by WA/IHA80 was 94.7%. Cross-reactive antibodies were observed in some of the putatively negative control sera: most cross-reactive reactions had a titer of 1:80 and were observed in the filariasis bancrofti and hepatitis A control groups. However, with a dilution of 1:160 as the endpoint in WA/IHA, cross-reactive antibodies were rare and WA/IHA160 had a specificity of 98.9%. The sensitivities of SEA/ELISA ranged from 50.0 to 93.3%. The sensitivities in detecting S. mansoni and S. haematobium were comparable. Specificity in detecting schistosomiasis was 98.2%. The sensitivities of the results for WA/IHA80 and SEA/ELISA combined ranged from 90.0 to 100%, while the sensitivities for the WA/IHA160 and SEA/ELISA combination were 80.0 to 98.7%. The combination of WA/IHA80 and SEA/ELISA gave a specificity of 92.9%, whereas WA/IHA160 and SEA/ELISA combined had a specificity of 97.2%. FIG. 1. | Reciprocals of WA/IHA titers in sera of individuals with S. mansoni Reciprocals of WA/IHA titers in sera of individuals with S. mansoni (SMA), S. haematobium (SHA), Katayama fever (KAT), Fasciolia hepatica (FAS), hookworm infection (HKW), trichuriasis (TRI), strongyloidiasis (STR), filariasis bancrofti (WUC), loiasis (LOA), onchocerciasis (ONC), hepatic amoebiasis (AMO), malaria (MAL), visceral leishmaniasis (VLE), toxoplasmosis (TOX), syphilis (SYP), borreliosis (BOR), human immunodeficiency virus (HIV), cytomegalovirus (CMV), Epstein-Barr virus (EBV), hepatitis A (HEA), hepatitis B (HEB), rubella (RUB), aspergillosis (ASP), rheumatoid factor (RHF), anti-nuclear factor (ANA), and Coxsackie B virus (COB) and in sera of healthy Dutch blood donors (HEC). FIG. 2. | OD values at 492 nm by SEA/ELISA in sera of individuals with S. mansoni OD values at 492 nm by SEA/ELISA in sera of individuals with S. mansoni (SMA), S. haematobium (SHA), Katayama fever (KAT), fascioliasis hepatica (FAS), hookworm infection (HKW), trichuriasis (TRI), strongyloidiasis (STR), filariasis bancrofti (WUC), loiasis (LOA), onchocerciasis (ONC), hepatic amoebiasis (AMO), malaria (MAL), visceral leishmaniasis (VLE), toxoplasmosis (TOX), syphilis (SYP), borreliosis (BOR), human immunodeficiency virus (HIV), cytomegalovirus (CMV), Epstein-Barr virus (EBV), hepatitis A (HEA), hepatitis B (HEB), rubella (RUB), aspergillosis (ASP), rheumatoid factor (RHF), anti-nuclear factor (ANA), and Coxsackie B virus (COB) and in sera of healthy Dutch blood donors (HEC). TABLE 1 | Sensitivity and specificity of the WA/IHA, SEA/ELISA, and the combination of both tests in detecting schistosomiasis DISCUSSION : This study showed that both a commercially available IHA test using S. mansoni adult WA and an ELISA with SEA are sensitive and specific tests for the serodiagnosis of schistosomiasis in travelers from the tropics. However, a combination of these tests is recommended, because the pooled results gave higher sensitivity than either test alone while maintaining high specificity. IHA tests using adult WA have been evaluated previously for serodiagnosis of schistosomiasis. All but one of these IHA tests were homemade systems that are not readily available for use outside the laboratories that developed them (-, , ), the exception being that sold by Dade (Cellognost, Schistosomiasis; Behringwerke AG, Marburg, Germany). In these IHA tests the sensitivity ranged from 71 to 80% and the specificity ranged from 80 to 100%. The IHA test that was produced by Fumouze until 1996 used soluble egg antigen for coating of the erythrocytes instead of the adult WA as in the test kit used for the present study (personal communication). With the former IHA test a sensitivity of 75.3% and a specificity of 96.9% was obtained . The present IHA test of Fumouze (i.e., the WA/IHA that was used here) has not been studied before in a clinical setting. Although earlier studies suggested that IHA gave high specificity, these studies incorporated only a few control serum samples. Because schistosomiasis is often observed in conjunction with other infectious diseases and autoimmune antibodies and because cross-reactive antibodies are frequently observed with other serological tests for schistosomiasis , we included a large series of controls in the present study. Taking the size and variety of our control groups into account, the specificities observed with the WA/IHA of 98.9 and 94.7% for cutoff titers of 1:160 and 1:80, respectively, were impressive. The sensitivities of the WA/IHA with the cutoff titer suggested by the manufacturer (1:160) were relatively low. For diagnosis of imported schistosomiasis a high sensitivity without a concomitant loss of specificity is, however, generally required. By lowering the cutoff from 1:160 to 1:80 the sensitivities of the WA/IHA increased strongly, with only a slight drop in specificity. On the basis of these observations we suggest that the cutoff titer could be lowered to 1:80. It is in any case questionable as to whether all 15 control samples with titers greater than 1:80 were truly false positives. Most of these patients had visited schistosoma-endemic countries, and although parasitological examinations in these cases were negative, coinfection with Schistosoma spp. could not be excluded. Because severe symptomatology is observed in some cases of Katayama fever, accurate diagnosis in this early stage of a schistosoma infection is important. Schistosoma egg production in patients with Katayama fever is often just commencing, and detection of eggs in stool or urine is mostly unsuccessful. Serodiagnosis, therefore, could be important. Our findings suggested that the WA/IHA80 correctly diagnosed 90.0% of the patients with early infections associated with Katayama fever. These findings are in contrast to those of two earlier studies with IHA containing adult WA, a homemade IHA and a commercially available IHA (Dade, Cellgnost, Schistosomiasis, Behringwerke AG) . These studies reported that the IHA was insensitive for detection of cases of early schistosomiasis associated with Katayama fever. In both studies IHA was compared with an indirect immunofluorescent test (IFAT) on gut-associated antigens of adult S. mansoni worms which appeared more sensitive than IHA in early schistosomiasis. Although the precise composition of the antigens used for erythrocyte coating in the commercially available WA/IHA used in our study is unknown, it is possible that the gut-associated antigens of the adult worm have been incorporated in the present IHA coating. The negative IHA results (1:40) for one of the Katayama patients studied is possibly a consequence of the short period between infection and clinical presentation. The IFAT also has a window period of approximately 30 days after infection before becoming positive . The WA/IHA titer of this patient rose to 1:640 3 months after the initial examination. Because of this delay in antibody response, patients clinically suspected of having Katayama fever but with an initially negative serological test result should be reexamined serologically a few weeks later. The data from this study suggest that the WA/IHA performs slightly better for detection of S. mansoni than for detection of S. haematobium. This result agrees with those of others who used IHA tests with adult WA of S. mansoni . In these studies it was demonstrated that the sensitivity of the IHA improved by using homologous rather than heterologous antigens. To overcome the relatively limited sensitivity of the WA/IHA for S. haematobium infections the antigen composition of the IHA could be changed by incorporating antigens of S. haematobium. ELISA with SEA was evaluated in this study as well. Because large amounts of SEA are difficult to obtain, we developed a more economic ELISA method with microtiter plates with smaller wells than the standard plates (i.e., 30 mul as opposed to 300 mul). In this form our ELISA had a sensitivity of 92.0% in detecting egg-proven cases of imported schistosomiasis. This high sensitivity agreed with the results of several earlier studies carried out with standard microtiter plates in endemic and nonendemic areas , although in one recent hospital study in the United Kingdom a lower sensitivity of 72% was found . There was no difference with our ELISA in detecting S. mansoni and S. haematobium, which confirmed the results of other studies . Although the ELISA showed a high sensitivity in detecting egg-proven cases, the sensitivity in confirming clinical Katayama fever was low because only 5 of the 10 patients (50%) showed positive reactions. In this context it has previously been noted that ELISA against egg antigen is unlikely to detect an infection in which schistosoma egg production has just started but in which anti-egg antibody synthesis has not yet commenced. In this stage of infection signs and symptoms of Katayama fever are, however, often already present . Thus, the probability that cases of Katayama fever can be confirmed by ELISA by using egg antigens is relatively low. Both high and low specificity with ELISA have previously been reported . Because of these variable results specificity in the present study was determined by using control sera, including those from patients with parasitic, fungal, bacterial, and viral infections, as well as from people with autoimmune antibodies. Despite this variety of control groups the specificity of the SEA/ELISA in detecting schistosomiasis was 98.2%, which was considered excellent. Cross-reactivity that was earlier observed with, e.g., filariasis and hepatitis, was not observed in our study. Our results thus suggest that ELISA carried out on plates containing small wells is a very economic, sensitive, and specific test to detect schistosomiasis after the onset of egg production. Although this study suggests that WA/IHA and SEA/ELISA are good diagnostic tools, improvement of diagnosis might be accomplished by combining the results of the two tests. In our study sensitivity and specificity were determined after combination of the results of either the WA/IHA80 or the WA/IHA160 with those of the SEA/ELISA. It showed that these combinations had the highest sensitivities of this study. Thus, the combination of WA/IHA80 and SEA/ELISA detected 100% of the schistosomiasis cases and 90% of the Katayama cases and had a specificity of 92.9%. The sensitivity of the combination of WA/IHA160 and SEA/ELISA was lower but still gave 98.0% detection of egg-proven cases and 80% confirmation of Katayama fever with a specificity of 97.2%. Combining the results of both WA/IHA and SEA/ELISA thus improved the sensitivity with maintenance of high specificity. In an earlier study a combination of both IFAT and ELISA has also been claimed to give a higher reliability in diagnosis of schistosomiasis, but the sensitivity and specificity of this combination were not reported . In conclusion, we suggest that a commercially available IHA is a good diagnostic tool for detection of imported schistosomiasis. The test is easily applicable and specific. Both infections with S. mansoni and S. haematobium can be diagnosed, and the test is sensitive for confirmation of Katayama fever. Our SEA/ELISA is also an economic, sensitive, and specific test for detection of schistosomiasis after the onset of egg production. Pooling the results of both the WA/IHA and the SEA/ELISA gave the most reliable outcome because it increased considerably the sensitivity in all stages of infection with maintenance of high specificity, particularly when the results of the WA/IHA160 were used in the combination with those of the SEA/ELISA. Backmatter: PMID- 12202559 TI - Detection and Genotyping of Oocysts of Cryptosporidium parvum by Real-Time PCR and Melting Curve Analysis AB - Several real-time PCR procedures for the detection and genotyping of oocysts of Cryptosporidium parvum were evaluated. A 40-cycle amplification of a 157-bp fragment from the C. parvum beta-tubulin gene detected individual oocysts which were introduced into the reaction mixture by micromanipulation. SYBR Green I melting curve analysis was used to confirm the specificity of the method when DNA extracted from fecal samples spiked with oocysts was analyzed. Because C. parvum isolates infecting humans comprise two distinct genotypes, designated type 1 and type 2, real-time PCR methods for discriminating C. parvum genotypes were developed. The first method used the same beta-tubulin amplification primers and two fluorescently labeled antisense oligonucleotide probes spanning a 49-bp polymorphic sequence diagnostic for C. parvum type 1 and type 2. The second genotyping method used SYBR Green I fluorescence and targeted a polymorphic coding region within the GP900/poly(T) gene. Both methods discriminated between type 1 and type 2 C. parvum on the basis of melting curve analysis. To our knowledge, this is the first report describing the application of melting curve analysis for genotyping of C. parvum oocysts. Keywords: Introduction : Cryptosporidiumparvum is a coccidian protozoan that causes self-limited diarrhea in immunocompetent individuals. In immuncompromised patients and malnourished children the disease is severe, prolonged, and life threatening. Although several immunological and molecular methods for detection of C. parvum oocysts in stool and environmental samples have been developed , immunomagnetic capture methods have found widespread application, particularly for water monitoring . The detection limits achieved with these systems are typically less then 10 oocysts, although recoveries are affected by the complexity of the matrix from which the oocysts are extracted. User-friendly molecular methods for the detection of oocysts in complex mixtures or genotyping of purified oocysts are needed for clinical and epidemiological applications and for water monitoring. Since no specific chemotherapy is available for this organism , early detection of C. parvum infections, particularly in immunosuppressed patients and children, may be critical to provide supportive treatment. Furthermore, the detection of asymptomatic individuals and animals excreting oocysts may be helpful for preventing secondary infections, studying transmission routes, and identifying reservoir hosts. More than 10 years have passed since the first report describing the detection of C. parvum by PCR . PCR has also been integrated into various genotyping procedures, such as restriction fragment length polymorphism analysis , random amplification methods , and methods detecting conformational polymorphisms . These approaches have been instrumental in advancing our understanding of the taxonomy of the genus Cryptosporidium and for studying the transmission of Cryptosporidium species and genotypes between various host species . More recently, a TaqMan real-time PCR assay for the detection of C. parvum based on the amplification of an 835-bp sequence from the small-subunit rRNA was developed . Although the internal probe used in this assay was complementary to a polymorphic region of this gene, the assay did not discriminate among Cryptosporidium genotypes originating from various host species. The anonymous C. parvum locus originally identified by Laxer et al. was chosen for a TaqMan assay by Guilot and Fontaine (E. Guilot and M. Fontaine, Abstr. Water Quality Technol. Conf., 2001). The limit of detection for this assay was five oocysts spiked into water pellets. The ability of real-time PCR to provide quantitative information was used by Kruger et al. to develop an assay to quantitate the amount of C.parvum DNA present in a sample. Finally, a SYBR Green I-based real-time PCR assay targeting the heat shock protein 70 locus of C. parvum was used to develop a commercial C. parvum detection assay (J. W. Czajka, G. D. DiGiovanni, M. E. Schaffer, A. M. Stolzfus, H. K. White, and M. W. LeChevallier, Abstr. 100th Gen. Meet. Am. Soc. Microbiol., abstr. 2-1062, 2000). As with the TaqMan assays, these assays do not differentiate between C. parvum type 1 and type 2. Here we report on the development of real-time PCR assays capable of detecting low numbers of oocysts and genotyping C. parvum. Polymorphic alleles were differentiated by SYBR Green I and fluorescent probe melting curve analysis (MCA) on the basis of single-nucleotide polymorphisms (SNPs). MATERIALS AND METHODS : Parasites. | The isolates of C. parvum used in this study are listed in Table . Oocysts were purified from fecal samples as described previously . Oocyst micromanipulations and DNA extraction. | Individual oocysts of isolate GCH1 were collected with a mouth-controlled pipette . Oocysts were suspended in water at a concentration of 104/ml. A 20-mul portion of this suspension was deposited onto a microscope slide, and the oocysts were viewed on an inverted microscope at x200 magnification. Individual oocysts were aspirated into a manually pulled glass micropipette with a tip with an inner diameter of approximately 15 mum. Individual oocysts were deposited into microcentrifuge tubes containing 10 mul of PCR-grade water. Oocyst samples were then subjected to three cycles of freezing-thawing and were spun at 16,000 x g for 30 s. The supernatant was introduced directly into 20-mul capillaries together with the PCR premixture for PCR amplification. Alternatively, following freezing-thawing, DNA was extracted from the oocyst lysate with a High Pure PCR template preparation kit (Roche Diagnostics, Mannheim, Germany), as recommended by the manufacturer, except that 50 mul (instead of 200 mul) of elution buffer was used. For negative controls, volumes of 0.5 to 1 mul of water containing no visible oocysts were collected from the oocyst suspension and deposited into microcentrifuge tubes containing 10 mul of PCR-grade water and processed as described above for the oocysts. To test the detection limit of the PCR method with oocysts isolated from feces, portions of 0.5-g fecal samples devoid of visible oocysts by microscopic examination of acid-fast-stained fecal smears were spiked with 3,000 oocysts of isolate GCH1 obtained by serial dilution from a stock suspension of 3 x 106 oocysts/ml. Spiked fecal samples were homogenized in 5 ml of 0.9% saline and filtered through gauze to remove large particles. The fecal filtrates were then centrifuged at 800 x g for 5 min. The pellets were resuspended in 600 mul of 0.9% saline solution and divided into three equal parts of 200 mul each. DNA was isolated from individual aliquots by one of the following three methods: (i) by phenol-chloroform extraction followed by ethanol precipitation, (ii) with the High Pure PCR template preparation kit (Roche Diagnostics), and (iii) and with the Magna Pure LC total nucleic acid isolation kit (Roche Diagnostics). In all three methods DNA was recovered in 200 mul of elution buffer. Portions of 1 mul of DNA solution, theoretically equivalent to five oocysts, were used in each PCR mixture. Real-time PCR with SYBR Green I. | PCR amplifications were performed in 20-mul capillary tubes with a LightCycler instrument (Roche Diagnostics). Reaction mixtures contained 1x LC-Fast Start DNA master mixture for SYBR Green I (Roche Diagnostics), 4 mM MgCl2, 10 pmol each of the forward and reverse primers, and 1 to 10 mul of DNA template. The tubes were capped, centrifuged at 700 x g for 5 s, and placed into the LightCycler carousel. Thermal cycling was performed as follows. A denaturation step of 10 min at 95C was followed by 40 cycles of 0 to 1 s at 95C, 4 to 5 s at the respective annealing temperature, and 7 to 22 s (depending on the template) at 72C. Following amplification, the PCR products were identified by MCA by raising the temperature from 45 to 95C at a rate of 0.05C/s. During the initial optimization phase, following MCA, PCR products were analyzed on agarose gels to ensure that products of the correct size were amplified. To this aim, products amplified on the LightCycler instrument were recovered from the capillaries, mixed with loading buffer, and loaded onto 1.5% agarose gels in 0.5x TBE (Tris-borate-EDTA) buffer. The gels were stained with GelStar (FMC, Vallensbaek Strand, Denmark). The following PCR primers were used: for beta-tubulin protocol 1, primer btub1 (ATGCTGTAATGGATGTAGTTAGACA; positions 552 to 576 [GenBank accession no. ]) (unpublished) and primer btub2 (GTCTGCAAAATACGATCTGG; positions 708 to 689) ; for the poly(T) protocol, primer cry44 (CTCTTAATCCAATCATTACAAC; positions 1331 to 1352 [GenBank accession no. ]) and primer cry39 (GAGTCTAATAATAAACCACTG; positions 1649 to 1629) . Real-time PCR with internal probes. | Oligonucleotides directed at the beta-tubulin gene of C. parvum (GenBank accession no. ) were synthesized by TIB Molbiol, Berlin, Germany. A 157-bp fragment was amplified from the beta-tubulin gene exon 2 with forward primer btub1 and reverse primer btub2. Two negative-stranded internal probes, btubLC (positions 639 to 616) and btubFL (positions 665 to 641), were included in the reaction mixture at concentrations of 0.5 pmol each. Probe btubLC was labeled at the 5' end with Red640, and probe btubFL was labeled at the 3' end with fluorescein. The 3' end of probe btubLC was blocked with a phosphate group to prevent extension by the polymerase. Fluorescence at 640 nm was generated by fluorescence resonance energy transfer following the annealing of both probes to their adjacent complementary sequences and the juxtaposition of the fluorescein and Red640 fluorophores. The sequences of the probes were identical to the beta-tubulin sequence found in C. parvum type 2 . Discrimination between the type 1 allele and the type 2 allele was based on the presence of three SNPs, two within btubFL and one within btubLC (T->C at position 616, G->A at position 646, and C->T at position 662 [type 2->type 1 changes]). These mismatches reduced the affinity of the probes for the type 1 allele, thus lowering the melting temperature. PCR cycling conditions were as described above for the SYBR Green I method, except that for MCA the temperature was increased at a rate of 0.5C/s and the reaction was run in 1x LC-Fast Start DNA master mixture for hybridization probes (Roche Diagnostics). The dynamic range and reproducibility of the assay were tested with 10-fold dilutions of C. parvum DNA corresponding to 104 oocysts to 1 oocyst per reaction mixture. The samples containing low numbers of oocysts were tested in triplicate. TABLE 1 | Isolates of C. parvum tested RESULTS : Sensitivity of SYBR Green I real-time PCR method. | In a preliminary comparison, two protocols amplifying different fragments of the beta-tubulin gene of C. parvum were evaluated. It was determined that the method targeting a 157-bp fragment flanked by primers btub1 and btub2 detected single oocysts, whereas a minimum of 1,000 oocysts was required for the other method. The sensitivity of the method with btub1-btub2 was further investigated with small numbers of oocysts introduced directly into the PCR mixture with a glass micropipette or with DNA extracted from oocysts collected individually by the same method. Table summarizes the results from 59 experiments with oocysts collected individually with a micropipette or with negative control samples devoid of oocysts. A total of 37 reaction mixtures were spiked with one, two, three, or five oocysts and 22 reaction mixtures were spiked with blank samples, as shown in the spiked reaction and negative control columns, respectively, in Table . A majority (12 of 20) of the reaction mixtures spiked directly with single oocysts and half (3 of 6) of those spiked with DNA extracted from single oocysts gave a visible amplification product by PCR with btub1-btub2 . Reaction mixtures spiked with DNA extracted from three and five oocysts were positive in three of four trials and three of three trials, respectively. Reaction mixtures spiked with DNA extracted from two oocysts were positive in two of four attempts. To assess the accuracy of the microisolation procedure and specifically to control for the possibility that additional oocysts or DNA may have inadvertently been introduced into the reaction mixtures, negative control samples were collected with the same micropipette system. Portions of 0.5 to 1 mul of water without visible oocysts were aspirated from the same experimental oocyst suspension. In the first series of assays, 3 of 12 negative control reaction mixtures amplified a detectable amplicon, suggesting either that oocysts were inadvertently collected and transferred to the PCR mixture or that DNA in solution was detected by the assay. To distinguish between these two possibilities, the oocysts from the experimental sample were precipitated by centrifugation and the supernatant was removed and replaced with the same volume of fresh PCR-grade water. Samples of 0.5 to 1 mul containing no visible oocysts were again aspirated from this washed oocyst suspension and subjected to PCR analysis. As shown in Table , none of these control reactions produced detectable PCR products. We conclude that the real-time PCR method with primers btub1 and btub2 can detect individual oocysts in a 40-cycle amplification reaction. The negative control experiments demonstrate that microisolation is an accurate method for the isolation of individual oocysts and that DNA present in solution may lead to false-positive results. Specificity of real-time PCR with primers btub1 and btub2. | The specificity of the btub1-btub2 PCR method was investigated by adding DNA from the following protozoa into standard PCR mixtures: Entamoeba histolytica, Leishmania tropica, Toxoplasma gondii, and Giardia lamblia. DNA from none of these protozoa gave detectable amplification products with primers btub1 and btub2. The specificity and the sensitivity of the btub1-btub2 amplification method were also tested with DNA extracted from C. parvum-negative fecal samples and fecal samples spiked with C. parvum oocysts. DNA was extracted from these samples by three methods: (i) by phenol-chloroform extraction followed by ethanol precipitation, (ii) with the High Pure DNA extraction kit (Roche Diagnostics), and (iii) with the Magna Pure DNA extraction kit (Roche Diagnostics). C. parvum DNA equivalent to that from five oocysts was detected in fecal samples from which DNA was extracted by all three methods, as demonstrated by the presence of a diagnostic melting peak at 81C. DNA was also amplified from unspiked fecal samples, but MCA showed that the amplicons were distinct from those originating from C. parvum. We conclude from the results of these experiments that the assay with primers btub1 and btub2 detects low numbers of oocysts in feces. Amplification of non-C. parvum DNA either was not observed or was easily distinguished from the diagnostic amplicon by MCA. Differentiation between C. parvum type 1 and type 2 by real-time PCR and with fluorescent probes. | A pair of probes 24 and 25 bp in length, respectively, were tested in a real-time PCR format with the same btub1 and btub2 amplification primers used for the SYBR Green I method. The purpose of this assay was to design a real-time PCR method capable of differentiating between C. parvum type 1 and type 2. The probes were designed such that their recognition sequence included multiple SNPs between the beta-tubulin alleles found in type 1 and type 2. In this assay, fluorescence originated from fluorescence resonance energy transfer between a fluorescein and a Red640 fluorophore present on the 3' end of the upstream btubFL probe and the 5' end of the downstream btubLC probe, respectively. As shown in Fig. , MCA differentiated between C. parvum type 1 and type 2. As expected from the sequence identity of the probes with the type 2 beta-tubulin allele, the melting temperature of type 2 isolates was higher than that of type 1 isolates. The presence of two SNPs within the btubFL recognition sequence decreased the affinity of the probe for the type 1 beta-tubulin allele by an average of 4.9C. In order to assess the reproducibility of this method for the detection of C. parvum type 1 and type 2, a total of 10 independent experiments were conducted over a period of 5 months. In each experiment the melting temperature of the probes was determined with the btubS and btub2 amplification primers and DNA from isolate GCH1 (type 2) and isolate TU502 (type 1). The mean melting temperatures were 65.27C (standard deviation [SD] = 1.82; n = 10) for type 2 and 60.33C (SD = 0.88; n = 11) for type 1. The difference in melting temperatures between type 1 and type 2 was statistically highly significant by the t test (P < 0.001). Several natural and laboratory type 2 isolates (isolates GCH1, TAMU, and OHIO) and type 1 isolates (isolates UG1544, UG1259, and UG1610) were tested by this method . The melting temperature among type 1 isolates was more homogeneous than that among type 2 isolates, but the origin of this heterogeneity was not further investigated. The results of gel electrophoretic analysis of the amplicons from seven C. parvum isolates obtained with primers btub1 and btub2 are shown in Fig. . Differentiation between C. parvum type 1 and type 2 by SYBR Green I real-time PCR. | To assess the feasibility of using a previously described polymorphic coding sequence for genotyping of C. parvum by SYBR Green I MCA, a 319-bp fragment from the GP900/poly(T) locus was amplified from seven type 2 and four type 1 C. parvum isolates. The amplicon is located within the gene encoding the GP900 sporozoite surface antigen and was previously shown to discriminate between type 1 and type 2 isolates on the basis of a polymorphic RsaI restriction site . MCA revealed different melting temperatures for type 1 and type 2 sequences , although the difference in melting temperatures between types was smaller than that observed by the probe-based assay described above. The mean melting temperature for three type 1 isolates was 83.47C (SD = 0.06), whereas for seven type 2 isolates the average melting temperature was 83.04C (SD = 0.12). Statistical analysis demonstrated that this difference was also significant (t = 5.73; P < 0.001). The sequence polymorphisms responsible for different melting peak temperatures were investigated by cloning and sequencing of the poly(T) amplicon from two type 2 isolates (isolates GCH1 and JRLHIV) and one type 1 isolate (isolate TU502). Nine SNPs between type 1 and type 2 were identified within the 319-bp amplicon, as was a CCA deletion at position 1542 (GenBank accession no. ) in the type 2 sequences. Of the nine SNPs, four were A or T->G or C changes, four were G or C->A or T changes, and one was a T (type 2)->A (type 1) change. Consistent with the tight distribution of the type 2 melting peak temperatures, the two type 2 sequences were identical. Since the observed SNPs do not change the number of A or T and G or C residues, we conclude that the difference in the melting temperature between the two types is caused by the 3-bp deletion in type 2. FIG. 1. | Detection of C. parvum Detection of C. parvum oocysts recovered by micromanipulation. A diagnostic melting peak was detected with DNA extracted from 1, 2, and 40 oocysts with the High Pure PCR template preparation kit (Roche Diagnostics). The template with 0.5 oocyst was obtained by eluting DNA from 10 oocysts with 50 mul of elution buffer and introducing 2.5 mul of eluted DNA into a PCR mixture. In addition to the diagnostic melting peak, low-temperature melting peaks were observed, particularly when DNA from a few oocysts was amplified. These peaks may originate from primers-dimers, but their origins were not investigated. The dashed line shows the melting profile of the negative control reaction. Note that peak height is not always reflective of the amplicon concentration but is a measure of the slope of fluorescence versus temperature. FIG. 2. | Real-time PCR genotyping of C. parvum Real-time PCR genotyping of C. parvum with two fluorescent probes specific for the beta-tubulin sequence. Shown is the MCA for four type 1 isolates and three type 2 isolates. The cursors were manually positioned on the melting peaks of reference isolates GCH1 and TU502. The melting curves for type 1 isolates TU502, UG1544, UG1259, and UG1610 and type 2 isolates GCH1 , TAMU , and OHIO are shown. FIG. 3. | Gel analysis of amplicons from different C. parvum Gel analysis of amplicons from different C. parvum isolates obtained by PCR with primers btub1 and btub2. Amplicons were fractionated on 1.5% agarose. Isolate designations are shown at the top. Lane M, 100-bp ladder. The positions of the 100- and 200-bp markers are indicated. FIG. 4. | Real-time PCR genotyping of C. parvum Real-time PCR genotyping of C. parvum using SYBR Green I fluorescence. A polymorphic 319-bp fragment of the GP900 gene was amplified by PCR in the presence of SYBR Green I. Shown are the melting peaks for type 1 isolates 2066-K , TU502, and CDC728 and type 2 isolates 7C , TAMU , and JRLHIV. The differences in melting peak temperatures between the type 1 and the type 2 isolates were statistically significant at the 0.001 level. The cursors were positioned manually and are labeled with the respective C. parvum type. TABLE 2 | Real-time PCR detection of oocysts isolated by micromanipulation DISCUSSION : Several real-time PCR methods for detection and genotyping of C. parvum oocysts were developed. As with conventional PCR, the sensitivity of the real-time method was strongly dependent on the amplification primers, even when primer pairs amplifying overlapping regions of the same gene were compared. Even though the sequence detected with primers btub1 and btub2 is present at only a single copy , single oocysts isolated by micromanipulation were detected in more than 50% of the PCR tests. The sensitivity of the method in which oocysts were introduced directly into the PCR mixture appeared to be equal to those of methods in which DNA was extracted prior to PCR. Two series of negative control experiments with washed and unwashed oocyst suspensions were performed to investigate the origin of the C. parvum DNA detected in some of the negative control reactions and specifically to assess whether additional oocysts may accidentally have been transferred to the PCR mixture during the micromanipulation procedure or whether DNA was present in solution. The latter possibility was consistent with the fact that the oocysts used in these experiments had been stored in the same suspension at 4C for approximately 3 months. During storage, a fraction of oocysts could have excysted and DNA from degrading sporozoites could have been released into solution. The absence of positive reactions for the second series of negative controls with a newly washed oocyst suspension was consistent with this assumption and ruled out the possibility of accidental transfer of oocysts during micromanipulation. The development of real-time PCR methods capable of discriminating between C. parvum type 1 and type 2 and other host-associated genotypes is relevant to the study of the epidemiology of this parasite and for source tracking. Type 1 and type 2 are thought to be transmitted by different routes since they differ in their host specificities. Type 1 has been detected almost exclusively in human infections, whereas type 2 infects both humans and a variety of animal species . The melting temperatures observed with fluorescent probes and with SYBR Green I were highly reproducible, indicating that these methods are suitable for differentiating C. parvum types, even on the basis of SNPs or small nucleotide deletions. In 11 independent experiments the coefficient of variation (CV) for the melting temperature of isolate TU502 was only 1.4%, and the CV was 3% for 10 replicate experiments with GCH1. CVs for the melting temperature of the poly(T) amplicon observed with SYBR Green I were as low as 0.1%. A possible source of the relatively small variability in melting temperature is the template concentration. As visible in Fig. , higher oocyst concentrations appeared to correlate with slightly lower melting temperatures, a phenomenon also observed in serial dilution experiments with larger oocyst numbers. For genotyping, assays based on fluorescent probes have higher specificities, but the requirement for primers with a fluorescent label increases the cost. The results of our unpublished experiments with type 1 and type 2 DNA mixed in different ratios indicate that probes do not resolve mixtures of templates as well as the SYBR Green I method does. This limitation is a concern in epidemiological studies, in which it is of interest to detect genotypically mixed parasite populations. In contrast, preliminary experiments with cloned microsatellite alleles of C. parvum mixed at different ratios showed that the limit of detection for a minority template was approximately 10%. This compared favorably with the 25% limit that we previously observed by restriction fragment length polymorphism analysis and is similar to the results recently reported by Reed et al. . The present laboratory methods for the diagnosis of C. parvum by microscopy are generally adequate for samples with high concentrations of oocysts, but are insufficient for the detection of cases of cryptosporidiosis in which only small numbers of oocysts are excreted . The high sensitivity of real-time PCR will facilitate the detection of asymptomatic carriers. Such information could be valuable for epidemiological studies of human and animal hosts. The time savings of the present approach are considerable. With the real-time PCR cycler used in these experiments, a 40-cycle amplification was completed in only 35 to 45 min. The fact that no gel analysis is required for routine experiments reduces the time needed by at least 1 h and could facilitate the implementation of high-throughput screening. The fact that the entire amplification and sample analysis are performed in a sealed capillary tube also reduces the risk of contamination from DNA carryover. The full potential of the real-time method for the detection of diagnostic DNA sequences will be achieved only with the simultaneous detection of different species of pathogens by using the capability of most real-time PCR machines to measure fluorescence at different wavelengths. For instance, one could envision an assay in which multiple species of a pathogen are detected with SYBR Green I on the F1 channel and the genotype of one of these species is identified on the F2 channel with an internal probe. In the case of Cryptosporidium, a possible setup would be to identify the species present in water or fecal samples by SYBR Green I MCA and the C. parvum genotype with fluorescent probes. Backmatter: PMID- 12202603 TI - Early Detection of Negative BACTEC MGIT 960 Cultures by PCR-Reverse Cross-Blot Hybridization Assay AB - We evaluated the efficacy of a PCR-reverse cross-blot hybridization assay, a test which permits identification of mycobacteria by means of species-specific probes and a Mycobacterium-specific probe, for early detection of negative BACTEC MGIT 960 mycobacterial cultures. Aliquots of 549 cultures were collected 7 days after the culture media were inoculated with various clinical specimens and tested with the molecular assay. PCR results were compared to those obtained at the end times with the BACTEC MGIT 960 system. Of the 549 specimens analyzed, 484 were found to be negative and 64 were found positive by both methods; one specimen, found to be positive by the BACTEC MGIT 960 system, was identified as negative by the molecular assay. In view of its high negative predictive value (99.8%), the PCR-reverse cross-blot hybridization assay appears to be a valid tool for early detection of negative BACTEC MGIT 960 cultures. Keywords: Introduction : Rapid, accurate identification of mycobacterial species in clinical specimens is an essential step in any strategy aimed at limiting the diffusion of infections caused by these slow-growing organisms. Conventional methods for recovery of mycobacteria have been supported by other techniques, such as liquid culture- and molecular biology-based systems, which can be used alone or in combination to obtain accurate results in a much shorter period of time . With the new BACTEC MGIT 960 (Becton Dickinson) liquid culture detection system, the presence of mycobacteria in a clinical specimen is detected, on average, 2 weeks after inoculation (, -, , , ). Unfortunately, 6 weeks must pass before a specimen can be reliably declared to be negative for mycobacteria, and this delay may lead to inappropriate empirical treatment with antimycobacterial drugs. We evaluated a PCR-reverse cross-blot hybridization assay, previously developed for direct detection and identification of mycobacteria in clinical specimens (-, ), as a potential predictor of negativity in BACTEC MGIT 960 cultures, and we used this molecular assay on aliquots of BACTEC MGIT 960 cultures 7 days after the culture media were inoculated with various clinical specimens. Results were compared to those furnished by the BACTEC MGIT 960 instrument at the end of the 6-week incubation period. Specimen processing and culture by BACTEC MGIT 960. : A total of 549 respiratory and nonrespiratory, consecutively enrolled clinical specimens (sputum, 354 samples; bronchoalveolar lavage fluid, 92 samples; urine, 79 samples; others [cerebrospinal fluid, blood, stool, ascitic fluid, abscess fluid, and biopsy specimen], 24 samples [total]) were collected from 549 patients suspected of having mycobacterial infections. Most of these patients were hospitalized in wards reserved for hematology and the treatment of infectious diseases. Specimens from sterile body sites were used directly to prepare smears for Ziehl-Neelsen acid-fast staining and BACTEC MGIT 960 cultures. Those from nonsterile sites were digested and decontaminated by standard procedures . The supernatant was discarded, the pellet was suspended in a sterile phosphate buffer (final volume, 2 ml), and the mixture was used for smear and culture preparation. For each specimen, we prepared two BACTEC MGIT 960 tubes (designated A and B), each containing 0.8 ml of MGIT oleic acid-albumin-dextrose-citrate enrichment (Becton Dickinson) and 0.1 ml of antibiotic mixture MGIT PANTA (polimyxin B, nalidixic acid, trimethoprim, and azlocillin). The inoculum was added (0.5 ml/tube), and the tubes were incubated inside the device at 37C and automatically monitored for 6 weeks or until an alarm signal indicated mycobacterial growth. In addition, two Lowenstein-Jensen slants were inoculated with 0.25 ml of each suspension, incubated at 37C for 8 weeks, and inspected weekly. When the BACTEC MGIT 960 system indicated positivity, the sample was removed and subjected to microscopic confirmation and species identification of mycobacteria by the AccuProbe confirmation test (Gen-Probe, Inc., San Diego, Calif.), hsp65 PCR-restriction enzyme analysis , or conventional procedures . The mycobacterial colonies grown on the Lowenstein-Jensen slants were identified as described above. Testing of BACTEC MGIT 960 cultures by PCR-reverse cross-blot hybridization assay. : On the seventh day of incubation, a 1-ml aliquot from each A-tube culture was aseptically collected for testing with the PCR-reverse cross-blot hybridization assay (see below) and the remainder of each culture was reincubated for 5 more weeks (total incubation, 6 weeks [MGIT end time]). The B tubes remained undisturbed for 6 weeks after inoculation (or until mycobacterial growth was detected by the BACTEC MGIT 960 system). The PCR-reverse cross-blot hybridization assay was carried out as previously described . The PCR products were then analyzed in a reverse cross-blot hybridization assay with probes specific for the genus Mycobacterium and the following mycobacterial species: M. tuberculosis complex, M. avium, M. intracellulare, M. smegmatis, M. marinum (or M. ulcerans), M. kansasii, M. xenopi, M. triplex, M. malmoense (or M. szulgai), M. gordonae, M. genavense, M. fortuitum, and M. chelonae . The hybridized PCR products were detected by incubation with streptavidin-alkaline phosphatase and a color substrate . At the MGIT end point, BACTEC MGIT 960 results for tubes A and B were compared to those of the PCR assay. If there was any discordance, both of the cultures (tubes A and B) of the specimen in question were subjected to acid-fast microscopic examination and PCR-reverse cross-blot hybridization. As shown in Table , the BACTEC MGIT 960 system detected mycobacterial growth in 65 of 549 (11.8%) specimens (mean time to detection, 11.8 days; range, 8 to 27 days). The same recovery rate was obtained on Lowenstein-Jensen solid medium (data not shown). At the end of the 6-week incubation, the remaining 484 specimens (88.2%) were negative. In the PCR-reverse cross-blot hybridization assay (performed on 7-day MGIT cultures), 485 specimens (88.4%) were negative and 64 (11.6%) were positive. For the 64 specimens found to be positive by both methods, the PCR products hybridized with the probe specific for M. tuberculosis (58 specimens) or with the probe specific for M. avium (6 specimens), and these results were fully concordant with the identification data obtained with the other methods used. The sensitivity, specificity, and positive and negative predictive values of the PCR-reverse cross-blot hybridization assay, compared to those for the BACTEC MGIT 960 system, were 100% for respiratory tract specimens and only slightly lower for the other specimens tested . Overall, the assay was characterized by high sensitivity (98.4%) and a negative predictive value of 99.8%. The only case of discordant results was that of a urine specimen that proved to be negative in the molecular assay. After 6 weeks of incubation, culture tube A of this specimen was also declared to be negative by the BACTEC MGIT 960 detection system, and when contents of this tube were subjected to PCR-reverse cross-blot hybridization at the end time, no hybridization occurred with any of the probes and the microscopic examination failed to reveal any evidence of acid-fast bacilli. In contrast, 18 days after inoculation, M. gordonae was isolated from the corresponding B culture, which was positive for the same microorganism in the molecular assay. In all probability, this result is a reflection of a paucity of mycobacteria in the original clinical specimen. Despite clinical signs and symptoms of mycobacterial infection, the number of bacilli eliminated can be quite low, and this factor is often responsible for false negativity, even with highly sensitive detection methods . The fact that the PCR-reverse cross-blot hybridization assay incorrectly identified this urine specimen as negative for mycobacteria is not a reflection of a lack of sensitivity ---the assay's sensitivity is, on the contrary, known to be quite high ---but rather an artifact of the specimen's having been divided and cultured in duplicate for control purposes. Another possible explanation for this disagreement between the methods is that M. gordonae was a contaminant in this case, not a clinically significant isolate, and the PCR-reverse cross-blot hybridization assay failed to detect it because the A culture either was not contaminated or was not sufficiently contaminated with the organism for it to be detected. The BACTEC MGIT 960 automated detection system can be a great time-saver in laboratories that process a high number of specimens for detection of mycobacterial species. Our objective in the present study was to evaluate the efficacy of a PCR-reverse cross-blot hybridization assay as an adjunct to the BACTEC MGIT 960 culture detection system for more rapid identification of negative clinical specimens. Performed on MGIT cultures after only 7 days of incubation, the PCR-based assay correctly revealed the absence of mycobacteria in all 484 specimens that were reported to be negative by the automated detection system after the full 6-week period of incubation. The assay was equally effective in identifying positive specimens, and species identification data were 100% concordant with the results of the AccuProbe culture confirmation test, the hsp65 PCR-restriction enzyme analysis, and conventional procedures for mycobacterial identification. PCR-reverse cross-blot hybridization is economical enough to be used routinely ($1.50/specimen), and in our opinion, its cost is more than offset by the potential savings allowed by earlier reporting (e.g., the cost of unnecessary antimycobacterial therapy). Naturally, our data need to be confirmed by additional studies, but use of the PCR-reverse cross-blot hybridization assay with the BACTEC MGIT 960 system seems to be a reliable approach to accelerating the assessment and, in particular, the exclusion of suspected mycobacterial infections. TABLE 1 | Results of the PCR-reverse cross-blot hybridization assay and the BACTEC MGIT 960 system for 549 clinical specimens from patients suspected of having mycobacterial infections Backmatter: PMID- 12202604 TI - Persistent Endemicity of Salmonella bongori 48:z35:- in Southern Italy: Molecular Characterization of Human, Animal, and Environmental Isolates AB - From 1984 to 1999, we collected 31 isolates of the rare serovar Salmonella bongori 48:z35:- in southern Italy. Twenty-four of the isolates were from cases of acute enteritis in humans. Pulsed-field gel electrophoresis analysis showed that all but one of our isolates were at least 80% similar. Our findings suggest that genetically related S. bongori 48:z35:- strains are endemically circulating in southern Italy. Keywords: Introduction : Serovars of Salmonella other than subspecies enterica are associated mainly with cold-blooded animals and rarely colonize the intestines of warm-blooded animals. Human infections with serovars of Salmonella bongori or Salmonella enterica subspecies salamae, arizonae, diarizonae, houtenae, and indica are infrequent and are usually the result of contact with reptiles . From 1985 to 1999, we collected 24 isolates from an epidemic cluster and from apparently sporadic cases of acute enteritis caused by S. bongori 48:z35:- in different cities in southern Italy . Seven further isolates of the same serovar were collected from a healthy human carrier; warm-blooded animals, i.e., two apparently healthy pigeons and a dog with diarrhea; and the environment (urban wastewater and food) (see Table ). All but one of our isolates from human cases of diarrhea were from children aged 1 month to 3 years. The first five isolates were identified at the beginning of 1985 from a small epidemic cluster of cases of acute enteritis in children who presented with moderate to severe diarrhea with fever and who recovered 3 to 8 days after rehydration therapy. The only isolate from an adult case of diarrhea was from a human immunodeficiency virus-positive patient in 1999. This patient also recovered rapidly after antibiotic therapy. To our knowledge, no cases of human infections with S. bongori 48:z35:- have ever been reported in other countries; the only recorded isolates of this serovar are the original S. bongori strain isolated in Chad from a lizard in 1966 and four isolates from foodstuffs sent in 1985 from England to the World Health Organization Collaborating Centre for Reference and Research on Salmonella, Institut Pasteur, Paris, France (M. Y. Popoff, personal communication). We subjected all the southern Italy isolates to molecular typing by pulsed-field gel electrophoresis (PFGE) analysis of digested DNA . Two collection strains kindly supplied by the World Health Organization Collaborating Centre for Reference and Research on Salmonella (M. Y. Popoff), the original strain CIP 261-66, isolated in Chad, and the strain CIP 5040-85, isolated in England, were also analyzed. Analysis of digested DNA by PFGE was performed as described by Faith et al. . The XbaI restriction enzyme (Pharmacia LKB Biotechnology AB, Uppsala, Sweden) was used for digestion of genomic DNA. Similarities among restriction endonuclease digestion profiles were calculated by the Dice similarity index with the Taxotron software RestrictoTyper module (Taxolab, Institut Pasteur, Paris, France). A dendrogram tree was constructed by using the Adanson and Dendrograf module of Taxotron software and applying the unweighted pair group method of averages algorithm to the distance matrix resulting from the comparison of the profiles. PFGE of XbaI-digested genomic DNA from the 31 S. bongori 48:z35:- isolates from southern Italy showed 19 profiles (profiles 1 to 18 and 21) differing by two or more fragments (Fig. and ). Profiles 19 and 20 were obtained, respectively, from the original strain isolated in Chad and from the strain isolated in England and differed from each other and from all the profiles of the isolates from southern Italy by several fragments. All but one of these showed identical or more than 79% similar profiles, differing from each other by two to seven bands. Only the profile of a soft cheese isolate apparently not linked to food-borne infection (profile 21) was clearly unrelated (less than 40% similarity) to those of the other isolates. The five isolates from the epidemic cluster collected from Messina in 1985 shared identical profiles with two other strains isolated in the same town from pigeons (profile 12). Three other small clusters of identical profiles were detected, but they all included isolates from different sources in terms of place and/or year of isolation. Considering profile 12 the original ancestor clone, which included the oldest isolates, all but one of the southern Italy isolates showed two- to five-band differences and should be considered either closely or possibly related to the ancestor's profile . FIG. 1. | PFGE restriction patterns of nine representative S. bongori PFGE restriction patterns of nine representative S. bongori 48:z35:- isolates from Sicily and Calabria and two collection strains from Chad and England. Lanes 1 and 15, lambda ladder pulsed-field gel (PFG) marker 340 (New England Biolabs, Inc., Beverly, Mass.); lanes 2 and 14, low-range PFG marker 350 (New England Biolabs); lane 3, CIP 261-66 (Chad); lane 4, CIP 5040-85 (United Kingdom); lane 5, C 285 CT 85; lane 6, CEIM 24662 ME 85; lane 7, CEIM 26851 ME 85; lane 8, C 634 CT 98; lane 9, CEIM 44721 RG 98; lane 10, CEIM 46162 RG 99; lane 11, CEIM 46295 PA 99; lane 12, CEIM 46049 AG 99; lane 13, CEIM 46082 CS 99. The two-digit numbers after the city abbreviation (see Table , footnote b) indicate the year of isolation. FIG. 2. | Dendrogram showing percent similarity calculated by the Dice similarity index of PFGE restriction endonuclease digestion profiles among the 31 S. bongori Dendrogram showing percent similarity calculated by the Dice similarity index of PFGE restriction endonuclease digestion profiles among the 31 S. bongori 48:z35:- isolates from southern Italy and two collection strains from Chad (CIP 261-66) and England (CIP 5040-85). Genetic homogeneity of southern Italy S. bongori 48:z35:- strains had already been shown by ribotyping , which is known to be less discriminating than PFGE. In profiles produced by PFGE, differences of two to three fragments in the banding patterns suggest that a single genetic mutation has occurred while differences of four to six bands are observed when two independent genetic events occur . None except one of the southern Italy S. bongori 48:z35:- isolates differed in their PFGE profiles from the supposed original clone (profile 12) by more than five fragments. This result is consistent with a large circulation due to a prolonged endemic presence in our population. The presence of S. bongori 48:z35:- in southern Italy before 1984 is not documented but nevertheless cannot be excluded. In past years, many rare and new serovars of Salmonella have been identified in Sicily from wild reptiles but S. bongori 48:z35:- has never been isolated from these animals . Transmission of salmonellae from reptiles to children has been repeatedly observed in other countries . Although reptiles are not present in households of southern Italy either as pets or as food, lizards are widely present in rural Italy as well as in urban areas ; thus, children aged less than 3 years might be exposed to lizard droppings while crawling on the floor or being in a playground. The diffuse endemic presence of S. bongori 48:z35:- in our population is confirmed by the finding of isolates from urban wastewaters whose PFGE patterns were identical or closely related to those of isolates from cases of acute enteritis. The significance of the isolation of this serovar from domestic animals as occurred in the city of Messina from pigeons and in the city of Cosenza from a dog is questionable. Although pet animals can be sources of infections, mainly for children, they can also be infected by humans. Finally, considering that the majority of our cases of enteritis occurred in infants in the first months of life, the infection may have been transmitted by person-to-person contact and originated in adult healthy carriers within the family. In conclusion, S. bongori 48:z35:-, a rare serovar that has never been recorded as being responsible for human or animal infections in other countries, proved to be able to cause acute enteritis in children and, occasionally, in immunodeficient adults and in animals in southern Italy. PFGE analysis of the XbaI-digested genomic DNA of isolates showed them to be genetically distant from the only two previously isolated S. bongori 48:z35:- strains from Chad and England. On the contrary, PFGE profiles of all but one of the isolates from southern Italy did not differ substantially from each other. Since these genetically related isolates have been detected over a 15-year period and are mostly epidemiologically unrelated, we can affirm that S. bongori 48:z35:- is endemically circulating in our population. TABLE 1 | Origin and characterization of the 31 human, animal, and environmental S. bongori 48:z35:- isolates from southern Italy (1984 to 1999) used in this study to determine PFGE XbaI-digested genomic DNA profiles Backmatter: PMID- 12202599 TI - Direct PCR Detection of Burkholderia cepacia Complex and Identification of Its Genomovars by Using Sputum as Source of DNA AB - We developed a nested PCR assay that detects the recA gene of the Burkholderia cepacia complex in sputum. The product of the first PCR round is also used to identify the genomovar of the pathogen. The protocol achieves high sensitivity and specificity with simple interpretation of genomovar status. Keywords: Introduction : Burkholderia cepacia is a multiresistant and transmissible opportunistic pathogen that frequently impairs the clinical state of patients with cystic fibrosis (CF) . The group of B. cepacia complex (BCC) organisms consists of eight genomovars associated with different levels of virulence and patient-to-patient transmissibility : B. cepacia genomovar I, B. multivorans (genomovar II), B. cepacia genomovar III, B. stabilis (genomovar IV), B. vietnamiensis (genomovar V), B. cepacia genomovar VI, B. ambifaria (genomovar VII), and B. pyrrocinia . To reduce the probability of BCC spreading among patients with CF, a reliable early test that detects small quantities of the bacteria in clinical samples is needed. Conventional microbiological diagnostics of the BCC based on the results of culture and subsequent biochemical identification is insufficient for this purpose because of the potential risk of misidentification or false negativity . Moreover, these methods are not able to reliably distinguish the genomovars of the BCC . To increase the sensitivity and specificity of diagnostic routines and to identify the genomovars, use of molecular genetic methods is inevitable. Two target genes are commonly used for BCC analysis: the 16S rRNA gene and the recA gene . The recA gene polymorphisms enable both differentiation of the BCC from other closely related bacteria and its sorting into genomovars. Moreover, the differences in the recA sequences within genomovar III led to the establishment of two recA clusters designated III-A and III-B. The present study sought (i) to develop a rapid diagnostic method for early detection of BCC organisms and determination of their genomovars directly from sputum and (ii) to assess the occurrence of the BCC genomovars in Czech patients with CF. From May 2001 to April 2002, we collected 211 sputum samples from 134 consecutive CF patients attending the Prague CF center (55 males, 79 females; age range, 0 to 33 years). All samples were examined for the presence of the BCC by culture and our novel PCR protocol. For cultivation, specimens were liquefied with a homogenization solution (0.9% NaCl, 50 mM KH2PO4, 35 mM NaOH, 1% N-acetyl-l-cysteine), shaken for 20 min at 800 rpm, and cultured on blood, chocolate, Endo, Sabouraud, and MacConkey agars. The last 53 samples were also cultured on the selective B. cepacia agar (Oxoid, Basingstoke, United Kingdom) that came into use during the study. All of the BCC isolates were identified by the API 20 NE system (Biomerieux, Marcy l'Etoile, France) in accordance with the supplier's protocol. For PCR analysis, sputum was mixed with an equal volume of the homogenization solution and shaken for 1 h at 800 rpm. DNA was extracted from 100 mul of liquefied sputum with the AMPLICOR Respiratory Specimen Preparation Kit (Roche, Indianapolis, Ind.) in accordance with the manufacturer's instructions. To check the DNA content in extraction aliquots, real-time PCR quantitation of the human albumin gene as an equivalent of the human DNA content was done on an ABI 7700 system (Applied Biosystems, Foster City, Calif.). Samples with albumin gene quantities below the fifth percentile of the first 100 assays were re-extracted. To detect the open reading frame of the BCC recA gene, DNA extracts were subjected to nested PCRs. For the first PCR round, the Taq PCR Core Kit (Qiagen, Hilden, Germany) was used. The PCRs were carried out in duplicate in a total volume of 20 mul with 1x PCR buffer, 1x solution Q (Qiagen, Hilden, Germany), 2 mM MgCl2, 200 muM (each) deoxynucleoside triphosphate, 14 pmol (each) of primers BCR1 and BCR2 , 1 U of Taq polymerase, and 1 mul of the DNA extract. The PCR program was run on a GeneAmp 9700 thermocycler (Applied Biosystems, Foster City, Calif.): initial denaturation for 2 min at 94C; 30 three-temperature cycles of 30 s at 94C, 45 s at 62C, and 90 s at 72C; and a final extension of 5 min at 72C. The second round of the nested PCR was performed with 1x PCR buffer (Promega), 2 mM MgCl2, 200 muM (each) deoxynucleoside triphosphate, 10 pmol (each) of inner recA primers REC-IN-5 and BCRBM2 , 0.5 U of Taq polymerase (Promega, Madison, Wis.), and 0.5 mul of the first-round PCR product. The amplification profile consisted of 25 three-temperature cycles of 30 s (for denaturation) at 94C; 45 s (for annealing) at 67C (first 5 cycles), 65C (next 5 cycles), and 63C (remaining 15 cycles); and 60 s (for synthesis) at 72C. A final extension step of 5 min at 72C followed. The PCR products of both rounds were run on a 2% agarose gel in 0.5% Tris-borate-EDTA buffer for 20 min at 10 V/cm. TABLE 1 | Primers used in this study In the BCC-positive samples, products of the first round were also used for genomovar status determination. The PCR product was diluted 1:10,000 (for samples positive in the first PCR round) or 1:100 (for samples that yielded a signal only in the second PCR round) with deionized, double-distilled water. The PCR setup was identical to that of the second round; the genomovar specificity of the eight reactions was maintained by the use of respective sequence-specific primers . New primer pairs were designed to identify B. multivorans (C-2), genomovar VI (C-6), and B. ambifaria (C-7). To check the sensitivity and specificity of the assays, we used strains from the BCC strain panel completed with genomovar VI and B. ambifaria strains from the Laboratorium voor Microbiologie, Ghent, Belgium. The recently described species B. pyrrocinia was not tested. Results are shown in Fig. . Each strain was correctly identified; moreover, our new 3' primers eliminated the previously described cross-reactivity between B. multivorans and genomovar VI . We also tested for reactivity with other bacteria possibly present in CF sputum (i.e., Staphylococcus aureus, Pseudomonas aeruginosa, Haemophilus influenzae, Stenotrophomonas maltophilia, Alcaligenes xylosoxidans, Ralstonia pickettii, B. gladioli, Chryseobacterium meningosepticum, and Delftia acidovorans). No reactivity was observed for either the generic C-IN primer mixture (used in the second round of the nested PCR) or the genomovar-specific primer pairs. To determine the sensitivity of the nested PCR, the procedure described by McDowell et al. was applied to strains LMG 16656 (genomovar III-A) and LMG 13010 (B. multivorans). The detection limit of the nested PCR was 103 CFU/ml of sputum. FIG. 1. | Genomovar-specific reactions on BCC reference strains ATCC 25416 (A), LMG 16660 (B), LMG 16656 (C), LMG 14294 (E), LMG 10929 (F), LMG 18941 (G), and LMG 19182 (H) and a clinical sample of genomovar III-B (D). Genomovar-specific reactions on BCC reference strains ATCC 25416 (A), LMG 16660 (B), LMG 16656 (C), LMG 14294 (E), LMG 10929 (F), LMG 18941 (G), and LMG 19182 (H) and a clinical sample of genomovar III-B (D). The lanes contained sequence-specific primer PCR products of the genomovars. Lanes: 1, genomovar I; 2, B. multivorans; 3a, genomovar III-A; 3b, genomovar III-B; 4, B. stabilis; 5, B. vietnamiensis; 6, genomovar VI; 7, B. ambifaria; IN, generic detection of the BCC species. Molecular size markers (100-bp ladder) were run in lane M. Two hundred eleven sputum samples from 134 CF patients were tested. By culture techniques, the BCC organisms were recovered from 89 (42%) of 211 samples. Comparing results of growth on the nonselective agars and selective B. cepacia medium, we found no difference in the number of BCC organisms recovered. By the nested-PCR protocol, 110 samples (52%) were determined to be BCC positive, of which 28 yielded a signal only in the second PCR round. Of the 134 patients examined, the BCC was detected in 49 patients by culture whereas it was detected in 67 patients by PCR. Thus, 18 more CF patients were found to be BCC positive by PCR than by cultivation. A nested or seminested design is known to increase PCR sensitivity. The improvement in sensitivity over the nonnested setup was evident in 28 positive samples that were found to be positive only after the second round of PCR. This represents 25% of the 110 positive samples. Similarly, the superiority of two PCR rounds was demonstrated by Moore et al. , who employed a seminested approach for detection of B. multivorans and genomovar III. Although the sensitivity of the reported seminested strategy was 1 order of magnitude greater than the sensitivity of the present method, our protocol offers the sorting of BCC organisms into seven genomovars. To further minimize the risk of false negativity, we checked the yields of DNA extraction by using real-time quantification of human genome equivalents per microliter of DNA. If human DNA was not present in a sufficient quantity in the extract, failure of DNA extraction was highly probable since sputum contains large quantities of leukocytes. As regards the genomovar identification of BCC organisms, 62 patients (92.5%) were colonized by genomovar III, recA group III-A; 3 were colonized by genomovar III, recA group III-B; and 2 were colonized by B. multivorans. None of the patients was infected by two or more genomovars, and replacement of one genomovar by another was not observed during the study period. In addition to analysis of clinical samples, we examined 44 BCC isolates recovered from sputum samples of 36 Czech CF patients collected from 1997 to 1999 and archived at the National Institute of Public Health, Prague, Czech Republic. All of them were identified as genomovar III, recA group III-A, which has been previously associated with the most severe course of the infection caused by the BCC . The almost absolute uniformity of the BCC genomovar spectrum in the Czech CF community is alarming, and the identical finding among the archived samples suggests a possible epidemic origin of the infections. However, to fully explain this unfavorable phenomenon, further studies based on molecular typing have to be performed. In conclusion, the protocol described herein offers rapid PCR detection of BCC in sputum and identification of its genomovars. Previous assays for direct PCR detection of the BCC in sputum were based on amplification of the 16S rRNA gene or of the 16S-23S spacer region of the rRNA operon . Recently, McDowell et al. described PCR of the recA gene followed by restriction fragment length polymorphism (RFLP) analysis , concluding that the recA gene is a more suitable target than the 16S region. However, the PCR-RFLP assay can yield many different RFLP patterns, demanding attentive comparison . On the contrary, the algorithm presented here with nested PCR and genomovar-specific recA primers achieves high sensitivity and specificity with simple interpretation of genomovar results. Backmatter: PMID- 12202618 TI - Acrophialophora fusispora Misidentified as Scedosporium prolificans AB - Keywords: Comment Letter 1 : Arthur et al. reported an interesting case of human keratouveitis associated with the long-term intraocular retention of a contact lens, in which a filamentous fungus and Staphylococcus aureus were involved . In spite of no antifungal drug being administered, the infection was resolved. Histological examination revealed the presence of fungal elements in the tissue around the lens. The fungus involved in such infection was identified by the authors as Scedosporium prolificans , an emerging hyphomycete which causes severe infections, mainly in immunocompromised patients . However, it was described as having "flask-shaped conidiogenous cells with elongated necks and individual conidia as well as chains..." . Scedosporium prolificans is indeed characterized by flask-shaped conidiogenous cells, often in brush-like arrangement, not single on hyphae as depicted in Fig. 2 of the mentioned article, and by its ovoid conidia in slimy heads, never in chains. This last feature is useful to inexperienced mycologists for distinguishing this species from other morphologically similar fungi such as Scopulariopsis spp. Some species of Scopulariopsis show dark colonies and also dark conidia, which arise from annellidic conidiogenous cells, forming dry, basipetal chains. However, the fungus shown in Fig. 2 does not seem to be a species of Scopulariopsis, because in this genus single conidiogenous cells are not common. They are mainly in branched and sometimes penicillated conidiophores. Furthermore, conidia of Scopulariopsis spp. are more or less spherical and usually with a wide truncate base, while those of Fig. 2 seem to be fusiform or lemon-shaped. It is difficult, merely by observing the mentioned figure, to ascertain the species to which the fungus belongs. However, the images are reminiscent of Acrophialophora fusispora, which has been recovered recently in several clinical cases. Its colonies are greyish brown, and the conidia are pale brown. An important distinctive feature of this species is the ornamentation of the conidia, which is finely echinulate and often with distinct spiral bands. Acremonium sp. could be another possibility due to the presence of single conidiogenous cells (phialides) emerging from ropes of vegetative hyphae (Fig. 2B), which is typical of this genus, although dark colonies are rarely produced. To provide figures of a fungus involved in a reported case, even if the species is common, constitutes a good practice. This allows readers to agree or not, as in this case, with the identification of species involved. We would like to encourage the authors to deposit cultures of clinical isolates in an international culture collection, from where it could then be obtained by interested people for further studies. Without the availability of the isolates, the published etiology is of low value . Backmatter: PMID- 12202573 TI - Rational Design of DNA Sequence-Based Strategies for Subtyping Listeria monocytogenes AB - The ability to differentiate bacteria beyond the species level is essential for identifying and tracking infectious disease outbreaks and to improve our knowledge of the population genetics, epidemiology, and ecology of bacterial pathogens. Commonly used subtyping methods, such as serotyping, phage typing, ribotyping, and pulsed-field gel electrophoresis, can yield ambiguous results that are difficult to standardize and share among laboratories. DNA sequence-based subtyping strategies can reduce interpretation ambiguity. We report the development of a rational approach for designing sequence-based subtyping methods. Listeria monocytogenes was selected as the model organism for testing the efficacy of this approach. Two housekeeping genes (recA and prs), one stress response gene (sigB), two virulence genes (actA and inlA), and two intergenic regions (hly-mpl and plcA-hly) were sequenced for 15 L. monocytogenes isolates. Isolates were chosen from a representative collection of more than 1,000 L. monocytogenes isolates to reflect the genetic diversity of this species. DNA sequences were aligned, and sliding window analyses were performed for each gene to define 600-bp-long regions that were (i) most polymorphic (using ProSeq) or (ii) most discriminatory (using a new algorithm implemented in WINDOWMIN). Complete gene sequences for actA (1,929 bp) and inlA (2,235 bp) provided the highest discrimination (identifying 15 and 14 allelic types, respectively). WINDOWMIN allowed identification of 600-bp regions within these genes that provided similar discriminatory power (yielding 15 and 13 allelic types, respectively). The most discriminatory 600-bp fragments identified in the housekeeping and stress response genes differentiated the isolates into 8 to 10 subtypes; intergenic region sequences yielded 8 and 12 allelic types based on 335- and 242-bp sequences for hly-mpl and plcA-hly, respectively. Regions identified as most polymorphic were not necessarily most discriminatory; therefore, application of the WINDOWMIN algorithm provided a powerful tool for determining the best target regions for DNA sequence-based subtyping. Our specific results also show that inclusion of virulence gene target sequences in a DNA sequence-based subtyping scheme for L. monocytogenes is necessary to achieve maximum subtype differentiation. Keywords: Introduction : Listeria monocytogenes is a food-borne pathogen that causes approximately 2,500 cases of human illness and 500 deaths annually in the United States . Bacterial subtyping methods have improved our ability to detect and track human listeriosis outbreaks and have also provided tools for tracking sources of L. monocytogenes contamination throughout food systems. Application of subtyping methods also provides insight into the population genetics, epidemiology, ecology, and evolution of L. monocytogenes. A variety of conventional, phenotypic, and DNA-based subtyping methods have been described for differentiation of L. monocytogenes beyond the species and subspecies levels . While phenotype-based methods have been used for many years to subtype L. monocytogenes and other food-borne pathogens, DNA-based subtyping methods are generally more discriminatory and amenable to interlaboratory standardization and are thus increasingly replacing phenotype-based subtyping methods . Commonly used phenotype-based subtyping methods for L. monocytogenes and other food-borne pathogens include serotyping, phage typing, and multilocus enzyme electrophoresis (MLEE) . DNA-based subtyping methods include PCR-based approaches (e.g., random amplified polymorphic DNA and amplified fragment length polymorphism), ribotyping, and pulsed-field gel electrophoresis . These DNA-based methods define bacterial subtypes by using either PCR amplification or restriction digestion of bacterial DNA to generate DNA fragment banding patterns. While many of these methods have proven effective for differentiating L. monocytogenes subtypes, DNA fragment size-based subtyping methods have significant drawbacks. For example, despite the existence of software packages for data normalization and analyses , these subtyping methods are often difficult to standardize. As a consequence, the ease of exchanging and comparing subtype data among laboratories can be severely limited. While DNA fragment size-based subtyping methods have been used for cluster analyses, they generally do not provide information amenable to the inference of primary genetic characteristics (i.e., nucleotide sequences) for evolutionary analyses. As long-term studies on the epidemiology, ecology, and evolution of bacterial pathogens require subtyping data that can be used to infer and quantify the genetic relatedness of isolates, DNA fragment size-based subtyping methods have limited utility for these applications. DNA sequencing-based methods are being developed and increasingly used for subtyping and characterizing bacterial isolates. In these methods, complete or partial nucleotide sequences are determined for one or more bacterial genes or chromosomal regions, thus providing unambiguous and discrete data. Sequencing can target a single gene (single locus approach) or multiple genes. The advantages of sequencing methods over DNA fragment size-based typing methods include their ability to generate unambiguous data that are portable through web-based databases and that can be used for phylogenetic analyses . While a variety of DNA sequence-based subtyping strategies targeting virulence genes, housekeeping genes, or other chromosomal genes and regions are feasible, multilocus sequence typing (MLST), which is an extension of MLEE, represents a widely used strategy . MLEE differentiates bacterial strains by detecting variations in the patterns of the electrophoretic mobilities of various constitutive enzymes. Cell extracts containing soluble enzymes are separated by size in nondenaturing starch gels, and enzyme activities are determined in the gels through application of color-generating substrates . While MLEE has been used to study the population genetics of many bacterial pathogens, including L. monocytogenes , this method is difficult to standardize among laboratories. MLST directly determines the allelic variation of multiple housekeeping genes by using DNA sequencing instead of indirectly characterizing these alleles via measurement of the electrophoretic mobilities of the gene products through MLEE . MLST approaches have been developed for several organisms, including group A streptococci, Staphylococcus aureus, Neisseria meningitidis, and Campylobacter jejuni (-). MLST traditionally targets multiple loci that have slowly diversified from each other through accumulation of neutral or near neutral changes, thus providing reliable differentiation without the potentially confounding effects of positive selection that may particularly occur in certain categories of genes, such as bacterial virulence or surface genes . On the other hand, direct sequencing of virulence genes or intergenic regions may provide more-sensitive discrimination for cluster analyses and short-term epidemiological questions. While some studies have explored the suitability of sequencing single genes to differentiate L. monocytogenes strains , the discriminatory power of DNA sequencing strategies that target multiple distinct genes or regions has not yet been reported. Thus, we selected a well-characterized set of L. monocytogenes isolates to determine the suitability of DNA sequencing of housekeeping and stress response genes (sigB, prs, and recA), two virulence genes (actA and inlA), and two intergenic regions (plcA-hly and hly-mpl) to differentiate L. monocytogenes subtypes. The complete sequence information for these genes and regions was also used to define the most-discriminatory 600-bp fragments within these genes that could be used for rapid sequencing-based subtyping. This study also provides a general outline for a rational approach to the selection of target genes for DNA sequence-based subtyping of bacterial pathogens. MATERIALS AND METHODS : Bacterial strains and lysate preparation. | A test set of 15 isolates, which includes 5 isolates from each of the three L. monocytogenes genetic lineages , was selected from a collection of more than 1,000 isolates from foods, environmental samples, and human and animal listeriosis cases. Within lineages I and II, isolates were selected to represent sets of two or three closely related strains to allow us to best determine the ability of different target genes to differentiate closely related genotypes. Lineage III isolates were chosen to represent one set of closely related strains and three diverse strains, as strains classified into lineage III appear more genetically diverse than strains in the other two lineages. Bacteria were grown on brain heart infusion (Difco, Sparks, Md.) agar and in brain heart infusion broth at 37C. All stock cultures were stored at -80C in 15% glycerol. Bacterial cell lysates for use in PCRs were prepared with lysozyme and proteinase K as previously described . Genes selected for sequence typing. | Two housekeeping genes (prs and recA), one stress response gene (sigB), two virulence genes (actA and inlA), and two intergenic regions in the prfA virulence gene cluster were sequenced . The relative location of these genes on the L. monocytogenes chromosome is also shown in Table . The intergenic regions plcA-hly and hly-mpl were chosen because they constitute the largest intergenic regions found within the prfA virulence gene island. PCR amplification conditions. | All PCR amplifications were performed with Thermus aquaticus DNA polymerase (Perkin Elmer-Applied Biosystems, Foster City, Calif.), MgCl2 at a final concentration of 1.5 mM, 1x PCR buffer, and deoxynucleoside triphosphates at a final concentration of 50 muM. PCR primers and conditions are summarized in Tables and , respectively. PCR conditions were not optimized to achieve a single set of PCR conditions for all amplifications. Purification and quantification of PCR products. | For sequencing, PCR products were purified by using the Qiaquick PCR purification kit (Qiagen, Inc., Valencia, Calif.). For eight L. monocytogenes isolates, nonspecific amplification products and/or primer dimers >60 bp were observed with the sigB PCR primers. For these PCRs, we used the Qiaquick gel extraction kit (Qiagen) to isolate the desired PCR product following electrophoresis in a 1.5% low-melting-point agarose gel and excision of the appropriate band from the ethidium bromide-stained gel. DNA concentrations of purified PCR products were estimated by comparison of amplicon band intensity with that of a DNA marker of known size and concentration (pGEM marker; Promega, Madison, Wis.) with LabImage software (Kapelan, Halle, Germany). DNA sequencing and analyses. | DNA sequencing was performed at Cornell University's Bioresource Center with an ABI 3700 DNA Sequencer. With the exception of actA and inlA, PCR primers were also used for DNA sequencing. Following initial sequencing with PCR primers, primer walking with primers designed to match strain-specific internal sequences was used to complete the sequencing of actA and inlA. Nucleic acid sequences were proofread and aligned with Seqman (DNAStar) and MegAlign (Lasergene). Cluster analyses were conducted using MEGA, version 2.1 , and the unweighted pair group method with arithmetic mean (UPGMA) (number of nucleotide differences) model. The resulting clustering data were assessed together with the alignments to assign allelic types for each region or gene. Two sequences were assigned different allele numbers if the sequences differed by at least 1 nucleotide (nt). Sliding window analyses to determine the number of DNA polymorphisms per 600-bp window for each gene were performed with ProSeq software . A new algorithm was developed and implemented for identifying the most-discriminatory 600-bp region within each gene from a set of aligned sequences (WINDOWMIN). In this algorithm, the most-discriminatory region for a sample of size n is defined as the region that maximizes the number of different DNA sequences in a sample. WINDOWMIN allows the user to define the criteria that will classify two DNA sequences as different. For example, if sequencing errors are common, two DNA sequences may be defined as different from each other if they differ at more than 2 nt positions. If sequencing errors can be ruled out, it may be more reasonable to define two DNA sequences as being different if they differ in at least 1 nt position. For this study, two DNA sequences were defined as different if they differed by at least 1 nt. In WINDOWMIN, the number of different DNA sequences distinguishable by at least k nucleotides in a window starting at position i of the sequences is defined as nk,i. A window starting at position j in the sequence is most discriminatory if nk,j = maxi(nk,i) for a particular value of k. If many possible windows fulfill this criteria, the procedure can be iterated to find the set of windows fulfilling nk+1,j = maxi(nk+1,i) among the sequences fulfilling nk,j = maxi(nk,i). Window size must also be defined; for the analyses performed here, the window size was set to 600 bp. As a practical approach, we iterated the algorithm 10 times and then chose the window with the most segregating sites among the set of remaining windows. WINDOWMIN can be obtained by download at . Nucleotide sequence accession number. | The DNA sequences determined in this study have been deposited in the GenBank database and given accession numbers through . TABLE 1 | L. monocytogenes isolates used in this study TABLE 2 | Relevant characteristics of target genes and regions used for DNA sequencing TABLE 3 | PCR primers TABLE 4 | PCR conditions RESULTS : Sequencing results. | The complete open reading frames (ORFs) of two virulence genes (actA and inlA), three housekeeping and stress response genes (prs, recA, and sigB), and two intergenic regions were sequenced for 15 L. monocytogenes isolates (Tables and ). Overall, approximately 7,500 bp of genomic DNA were sequenced for each strain. DNA sequences were deposited into GenBank and are also available through the Cornell PathogenTracker database . Analyses of DNA polymorphisms and allelic types. | DNA sequences for each of the seven selected genes and intergenic sequences were aligned and evaluated. The size of these seven target sequences varied from 242 bp (plcA-hly intergenic sequence) to 2,235 bp (inlA) . Based on the total number of polymorphic sites in each sequence , actA displayed the highest sequence variability (14.3% of the nucleotides were polymorphic) while prs displayed the lowest level of overall sequence variability (4.9%). Only one target sequence (actA) allowed discrimination of all 15 strains characterized. Two strains (FSL J1-022 and FSL J1-047) differed by only a single nucleotide in actA. The other target genes and regions differentiated between 8 and 14 allelic types for each gene . Two sequences were assigned different allelic types (e.g., 1 to 15 for actA) if the sequences differed by at least 1 nt; the allelic types for all strains are summarized in Table . Table also shows how many allelic subtypes were defined based on at least a single nucleotide polymorphism and how many subtypes could be defined based on a cutoff of at least a 2-nt difference. For example, for the full-length prs sequence 10 allelic subtypes were defined based on at least a 1-nt difference and 7 allelic subtypes were defined based on at least a 2-nt difference. Thus, three of the prs allelic types were defined based on only a 1-nt difference. In addition, allelic information for the housekeeping genes prs and recA and the stress response gene sigB was used to assign MLST types (12 MLST types, A through L) based on the allelic types for all three genes. Insertion or deletion polymorphisms in actA, inlA, and hly-mpl. | In addition to single nucleotide polymorphisms, sequence alignments revealed insertion-deletions in actA, inlA, and hly-mpl. Nucleotide sequence alignment for actA revealed a deletion of 105 bp in strains FSL W1-110, FSL W1-111, FSL J2-039, FSL J1-158, FSL J2-003, FSL J2-064, and FSL X1-002. This finding is consistent with a previous study that reported deletion of a proline-rich repeat region among isolates classified into each of the three genotypic lineages of L. monocytogenes . Moreover, isolates FSL W1-111 and FSL J1-158 showed one insertion or deletion at actA nt 323 (a 9-bp insertion compared to the other isolates). The inlA sequence alignment revealed an insertion or deletion at nt 2052 in three strains (a 9-bp deletion in FSL W1-110, FSL J2-068, and FSL J2-003) and another 3-bp insertion or deletion at nt 2227 in strains FSL W1-111 and FSL J1-158 (a 3-bp insertion in these two isolates). Furthermore, isolate FSL J1-158 had a single base pair deletion at nt 2219, which results in a frameshift mutation with 7 missense amino acid residues at the 3' end of the inlA sequence. Sequence analyses of the hly-mpl intergenic region also revealed lineage-specific insertions or deletions. The five lineage I strains show two additional nucleotides at position 49. All lineage II strains have an additional adenine at nt 52. Insertion or deletion polymorphisms at the nt 121 to 129 region also show distinct patterns that differentiate strains in lineages I, II, and III; lineage II isolates have a 1-bp deletion compared to lineage I isolates, whereas lineage III isolates show 9- and 8-bp deletions compared the lineage I and II isolates, respectively. Definition of most-discriminatory gene fragments. | Since complete sequencing of large DNA fragments (e.g., the 1,929-bp actA ORF) is not practical, in terms of cost and time, for large-scale subtyping, we utilized two different approaches to define smaller (600 bp) gene fragments that will provide optimal strain differentiation, ideally with the same discriminatory power achieved by analysis of the complete ORF sequences. The software ProSeq (version 2.8) was used initially in a sliding window analysis to define the 600-bp region within each gene (actA, inlA, recA, sigB, and prs) that showed the highest number of nucleotide polymorphisms. ProSeq calculated the value pi, which is the average number of nucleotide differences per site between two sequences (also termed nucleotide diversity) , for each gene from the nucleotide alignment of the 15 strains. The region with the maximum nucleotide diversity is shown as the highest peak on a plot of pi determined at sliding windows of 600 bp . The locations of the 600-bp regions with the maximum nucleotide diversities for each gene are shown in Table . For actA, inlA, and sigB, the 600-bp window with the maximum nucleotide diversity did not allow the same level of allelic subtype differentiation as was achieved with the full ORF sequences . DNA sequence alignments were also used to perform sliding window analyses for each gene to define 600-bp sections that were most discriminatory for differentiating the aligned sequences into allelic types. An algorithm (WINDOWMIN) was implemented to determine the most-discriminatory 600-bp section. For three (actA, inlA, and sigB) of the five gene alignments, the WINDOWMIN sliding window analysis defined 600-bp regions that showed superior discrimination of allelic subtypes compared to the regions with the highest nucleotide diversities defined by the ProSeq sliding window analysis . Specifically, the 600-bp windows defined by WINDOWMIN allowed subtype discrimination equal to that achieved with the complete gene sequences for four of the genes sequenced (actA, recA, prs, and sigB), which ranged in length from 780 to 1,929 bp. WINDOWMIN was not able to define a 600-bp region that matched the allelic subtype discrimination of the full ORF for inlA, the largest gene sequenced (2,235 bp). The full inlA ORF allowed differentiation of 14 allelic subtypes, whereas the most-discriminatory 600-bp sequence identified by WINDOWMIN allowed differentiation of 13 allelic subtypes. FIG. 1. | Graphical representation of polymorphisms within 600-bp sliding windows in the genes recA (a) and actA (b). Pi denotes the average number of nucleotide differences per site between two sequences, or nucleotide diversity . Graphical representation of polymorphisms within 600-bp sliding windows in the genes recA (a) and actA (b). Pi denotes the average number of nucleotide differences per site between two sequences, or nucleotide diversity . Pi was calculated with the computer program ProSeq (see Materials and Methods). TABLE 5 | Summary of allelic subtypes and polymorphisms in actA, inlA, prs, sigB, recA, hly-mpl, and plcA-hly TABLE 6 | Allelic profiles of the virulence genes (actA and inlA), intergenic regions (hly-mpl and plcA-hly), and housekeeping genes (sigB, prs, and recA) DISCUSSION : We used L. monocytogenes as a model organism to develop and apply a rational approach for designing a DNA sequence-based subtyping scheme. Two housekeeping genes, one stress response gene, two virulence genes, and two intergenic regions were sequenced for 15 L. monocytogenes isolates selected from a collection of more than 1,000 isolates. The DNA sequence data obtained (approximately 7,500 bp/isolate) were used to define both the most-polymorphic and the most-discriminatory 600-bp regions within each gene. This model for designing discriminatory DNA sequence-based subtyping strategies also should be applicable to other target organisms. The development of a new algorithm designed to determine the most discriminatory section within a DNA sequence alignment and the implementation of this algorithm in a new program (WINDOWMIN) will further facilitate the definition of target genes and gene fragments for subtyping applications. Target genes for DNA sequence-based subtyping of L. monocytogenes. | Housekeeping genes are commonly used for conventional MLST as these genes are thought to diversify by neutral or near neutral nucleotide changes due to the vital roles of these gene products in contributing to an organism's survival . While sequence analysis of housekeeping genes has been a valuable tool for studying the population genetics of bacterial pathogens, DNA sequencing of more-rapidly evolving genes may allow more-sensitive subtype discrimination. We thus selected the housekeeping genes recA and prs, the stress response gene sigB, the virulence genes actA and inlA, and the intergenic regions hly-mpl and plcA-hly to determine the relative abilities of these target genes to differentiate closely related L. monocytogenes strains. While prs is located close to actA on the L. monocytogenes chromosome, sigB, recA, and the virulence genes were chosen to represent distinct chromosomal locations. prs was chosen as a target gene that would allow comparison between the discriminatory power of virulence genes and housekeeping genes located in close proximity. actA and inlA were specifically selected as target virulence genes since they are located in two different L. monocytogenes virulence gene islands . The complete gene sequences for actA and inlA provided the highest discrimination among the 15 isolates (15 and 14 allelic types, respectively). In addition, preliminary analyses of our data showed that the actA virulence gene indeed may be under positive selection (dN/dS > 1.0; dN, rate of nonsynonymous substitutions; dS, rate of synonymous substitutions) (unpublished data). The complete gene sequences for the housekeeping and stress response genes provided discrimination into 8 to 10 subtypes, and the sequences for the intergenic regions provided 8 and 12 allelic types based on 335- and 242-bp sequences for hly-mpl and plcA-hly, respectively. When sequence information for the two housekeeping genes and the stress response gene was used to define MLST types, a total of 12 types (A to L) could be differentiated. These results indicate that sequencing of the virulence genes actA and inlA provides the most-discriminatory DNA sequence-based subtyping for L. monocytogenes. While limited sequencing and PCR-restriction fragment length polymorphism analysis of other L. monocytogenes virulence genes has been performed by various groups , the results reported here provide the first comparative evaluation of different target genes for subtyping L. monocytogenes. Our results support previous observations that DNA sequencing of virulence or surface protein-encoding genes that may have been subjected to positive selection pressures can allow more-sensitive strain discrimination than sequencing of housekeeping genes can. To illustrate, Enright et al. showed that emm sequences differentiated more subtypes than single housekeeping gene sequences in Streptococcus pyogenes. However, interpretation of subtyping results based on highly variable sequences of virulence genes (such as actA) or surface proteins may be misleading, particularly if the data are used to probe the population genetics or long-term phylogenetic patterns of bacterial pathogens such as L. monocytogenes. Specifically, high rates of evolution and recombination among these genes may not reflect the true phylogenetic relationships among isolates . To overcome this obstacle, a subtyping scheme that includes sequencing of selected virulence genes in combination with sequencing of housekeeping and/or stress response genes and of regions with little or no selective pressure (such as intergenic regions) may provide the most appropriate approach for subtyping L. monocytogenes and other bacterial pathogens. Initial analyses for recombination within the two housekeeping genes and one stress response gene sequenced here indeed showed that these genes show no (sigB and recA) or weak (prs) indication for recombination (unpublished data). The sequencing of additional housekeeping genes as previously described for MLST approaches (-) for other bacteria may further improve the ability to study the phylogeny of L. monocytogenes. The inclusion of both positively selected genes, such as actA, and neutrally selected intergenic regions (such as the more-discriminatory plcA-hly region) allows maximization of the discriminatory power of a typing scheme though. Maximum discrimination is particularly important for bacterial pathogens such as L. monocytogenes, for which rapid and standardized cluster detection through molecular subtyping represents a critical public health need . The use of virulence gene targets in DNA sequence-based subtyping strategies also creates the opportunity to use pathogen-specific PCR primers to develop integrated PCR-based detection and subtyping strategies that do not require a culturing step . Rational design of high-throughput DNA sequence-based subtyping schemes. | Sequencing of complete virulence, housekeeping, and stress response genes (with ORF lengths between 780 and 2,235 bp) does not provide a suitable approach for high-throughput subtyping. High-throughput sequencing for subtyping purposes generally targets gene fragments between 450 and 600 bp in length, since these fragment sizes can easily be amplified and sequenced with a single set of primers or one set of sequencing primers nested inside the PCR primers . In the past, target regions have been selected without prior identification of specific desirable characteristics, such as an optimum number of polymorphic and discriminatory nucleotide sites. To identify target regions through a rational strategy, the complete ORFs for the five genes sequenced in this study were aligned to define the 600-bp section(s) for each gene that was (i) most polymorphic (with ProSeq) or (ii) most discriminatory (with WINDOWMIN). Not surprisingly, most genes, and particularly the larger virulence genes actA and inlA, displayed considerable differences in the numbers of polymorphic residues found in different regions. Thus, the allelic discrimination achieved with different 600-bp regions within a given gene also differed considerably . Interestingly, the 600-bp regions with the highest proportion of polymorphic residues did not necessarily provide the highest level of allelic discrimination. For three of the five genes sequenced (actA, inlA, and sigB), WINDOWMIN was able to define more discriminatory 600-bp regions than a program (ProSeq) that only determined the most-polymorphic 600-bp region within a gene. We conclude that our newly developed algorithm provides an improved rational approach for the selection of target regions for DNA sequence-based subtyping of bacterial pathogens and other microorganisms. DNA sequence-based subtyping in L. monocytogenes. | A variety of molecular subtyping approaches have been applied to L. monocytogenes, and the application of these techniques has allowed a better understanding of the biology, ecology, and epidemiology of L. monocytogenes and other bacterial pathogens . Studies using subtyping approaches also have suggested that L. monocytogenes subtypes may display heterogeneity in their potentials to cause disease in humans and animals . The data reported here provide the framework for the development and implementation of DNA sequence-based subtyping methods for L. monocytogenes. We have identified specific 600-bp regions that provide the most discriminatory targets for subtyping within different L. monocytogenes genes. The presence of insertion-deletions in some of these regions (e.g., at nt 852 to 1019 in actA) may complicate the interpretation of subtyping results for some isolates and may hamper design of PCR primers that allow reliable amplification of all L. monocytogenes subtypes. Thus, in addition to identifying the most discriminatory gene regions, the presence of insertion-deletions must also be carefully considered when selecting target regions for DNA sequence-based subtyping methods. Interestingly, our results also indicate that targeting specific insertion-deletions (e.g., in the hly-mpl intergenic region) by appropriate PCR assays may allow for sensitive differentiation of the three previously described L. monocytogenes lineages . Phylogenetic analyses of housekeeping, stress response, and virulence gene sequences also confirmed that the 15 L. monocytogenes isolates tested fall into the previously determined three distinct lineages (unpublished data) . The existence of these lineages has previously been confirmed by a variety of subtyping methods and thus appears to be evolutionarily relevant . While current DNA fragment size-based subtyping methods (such as pulsed-field gel electrophoresis and ribotyping) may provide good subtype differentiation, data obtained by these methods typically cannot be used to determine the evolutionary relatedness of isolates. The implementation of DNA sequence-based subtyping approaches for routine characterization of human, animal, and food L. monocytogenes isolates will not only allow for sensitive and standardized subtyping for outbreak detection, but will also provide an opportunity for using subtyping data to probe the evolution of this food-borne pathogen and to track the spread of clonal groups . DNA sequence-based subtyping methods will also provide standardized data that can easily be shared electronically and through the World Wide Web , concomitantly providing public health professionals and laboratories around the world with direct access to the information needed to identify and monitor emerging pathogenic bacteria. The resolution power of DNA sequence-based subtyping methods is unmatched by any other subtyping method. For example, while it is estimated that MLEE requires approximately 26 nt changes in order to determine a new electrophoretic type , one single nucleotide change at a targeted locus will result in a new subtype classification for DNA sequence-based subtyping . The continued development of new technologies for automated high-throughput sequencing and the availability of these technologies at a reasonably low cost will further facilitate widespread implementation of DNA sequence-based subtyping methods. Further application of the DNA sequencing-based subtyping approaches described here on large sets of epidemiologically well-defined isolates will also provide critical validation of the subtyping scheme proposed here. Backmatter: PMID- 12202619 TI - Pulsed-Field Gel Electrophoresis Can Yield DNA Fingerprints of Degradation-Susceptible Clostridium difficile Strains AB - Keywords: Klaassen et al. recently compared amplified fragment length polymorphism (AFLP) with pulsed-field gel electrophoresis (PFGE) as DNA fingerprinting methods for Clostridium difficile. It is known that certain C. difficile strains are problematic to fingerprint by PFGE due to DNA degradation . Corkill et al. suggested that such strains could be typed by PFGE if 50 muM thiourea was added to the electrophoresis buffer . However, Klaassen et al. reported that although DNA degradation was reduced by this new method, analysis of some C. difficile strains remained difficult due to residual DNA breakdown. We agree with the conclusion by Klaassen et al. that strains from hospital A represented clonal expansion of a single degradation-susceptible strain, but we were surprised at their assertion that previous data do not exist to support widespread dissemination of C. difficile isolates. In England and Wales, the Public Health Laboratory Service Anaerobic Reference Unit reported on the most common PCR ribotypes and established that ribotype 1 was present in 33 of 58 hospitals and accounted for 55% of infections in the United Kingdom . A recent epidemiological investigation within our own hospital established the presence of this same United Kingdom epidemic strain. Ribotype 1 represented approximately 80% of all C. difficile isolates and was found to be degradation susceptible during PFGE analysis. We have now modified our PFGE method to allow analysis of formerly degradation-susceptible C. difficile strains (Fig. and B). Our optimized method differs significantly from that used by Klaassen et al. in several key areas, and we suggest that similar modifications might improve the typeability of their strains. C. difficile strains were grown overnight in Schaedler's anaerobic broth as opposed to on agar plates to discourage mature spore formation which contributes to low yields of poor-quality DNA. In addition, more bacteria were incorporated into the agarose plugs (approximately 2.0 McFarland standards) to ensure that sufficient DNA remained present in the event of residual DNA degradation. Lysozyme was incorporated into the agarose plugs (10 mug/ml) and incubated at 37C for 1 h to ensure optimal bacterial lysis. The concentration of proteinase K was increased to 5 mg/ml to allow for the increased bacterial load. It was found that adequate bacterial digestion was achieved if blocks were exposed to proteinase K for a minimum of 16 h at 55C with fresh solutions added after approximately 8 h. We found that 200 muM thiourea must be present in both the agarose gel and the electrophoresis buffer to ensure minimal DNA degradation. We have found that the random amplified polymorphic DNA technique and PFGE have similar discriminatory powers (unpublished results) and therefore consider PFGE to be a valuable tool in epidemiological analysis of C. difficile strains. We hope that our PFGE protocol modifications will allow a full evaluation of PFGE, AFLP, and other C. difficile molecular fingerprinting techniques. FIG. 1. : PFGE DNA fingerprints of 8 representative ribotype 1 C. difficile PFGE DNA fingerprints of 8 representative ribotype 1 C. difficile isolates (lanes 2 to 9) obtained by using an unmodified PFGE protocol and 13 representative ribotype 1 isolates (lanes 2 to 14) obtained by using our modified PFGE protocol . The markers present in lanes 1 and 10 and lanes 1 and 15 are bacteriophage lambda concatemers. Gel images were generated by an ImageMaster VDS camera (Pharmacia) and were sized by using Adobe Photoshop software. Backmatter: PMID- 12202569 TI - New Variants of Vibrio cholerae O1 Biotype El Tor with Attributes of the Classical Biotype from Hospitalized Patients with Acute Diarrhea in Bangladesh AB - The sixth pandemic of cholera and, presumably, the earlier pandemics were caused by the classical biotype of Vibrio cholerae O1, which was progressively replaced by the El Tor biotype representing the seventh cholera pandemic. Although the classical biotype of V. cholerae O1 is extinct, even in southern Bangladesh, the last of the niches where this biotype prevailed, we have identified new varieties of V. cholerae O1, of the El Tor biotype with attributes of the classical biotype, from hospitalized patients with acute diarrhea in Bangladesh. Twenty-four strains of V. cholerae O1 isolated between 1991 and 1994 from hospitalized patients with acute diarrhea in Matlab, a rural area of Bangladesh, were examined for the phenotypic and genotypic traits that distinguish the two biotypes of V. cholerae O1. Standard reference strains of V. cholerae O1 belonging to the classical and El Tor biotypes were used as controls in all of the tests. The phenotypic traits commonly used to distinguish between the El Tor and classical biotypes, including polymyxin B sensitivity, chicken cell agglutination, type of tcpA and rstR genes, and restriction patterns of conserved rRNA genes (ribotypes), differentiated the 24 strains of toxigenic V. cholerae O1 into three types designated the Matlab types. Although all of the strains belonged to ribotypes that have been previously found among El Tor vibrios, type I strains had more traits of the classical biotype while type II and III strains appeared to be more like the El Tor biotype but had some classical biotype properties. These results suggest that, although the classical and El Tor biotypes have different lineages, there are possible naturally occurring genetic hybrids between the classical and El Tor biotypes that can cause cholera and thus provide new insight into the epidemiology of cholera in Bangladesh. Furthermore, the existence of such novel strains may have implications for the development of a cholera vaccine. Keywords: Introduction : New epidemic strains of toxigenic Vibrio cholerae have appeared at least twice in recent human history . Strains of the classical biotype, which had probably been responsible for most of the epidemic disease in the 19th century and much of the 20th century, were largely replaced as the predominant cause of epidemic cholera by strains of the El Tor biotype in most of the regions where cholera is endemic, beginning in 1961. However, the classical biotype strains reemerged as a predominant epidemic strain in parts of Bangladesh in 1982 and coexisted with the El Tor strains, causing disease until 1993. A second new epidemic strain, carrying the O139 rather than the O1 antigen, emerged in southern Asia in 1992 . The O139 and El Tor O1 strains continue to cause epidemics of cholera, and there are indications that the incidence of cholera due to the O139 serogroup is on the rise in parts of India and Bangladesh. The classical and El Tor biotypes of V. cholerae are closely related in their O-antigen biosynthetic genes , although these two biotypes differ in many other regions of their genomes . Thus, O1 El Tor strains might have arisen following transfer of O1 antigen biosynthetic genes into a previously unknown environmental strain. Conversely, O139 and O1 El Tor strains are closely related in most parts of their genomes but carry different O-antigen genes, suggesting the transfer of O139-specific genes from an unknown donor into a recipient El Tor strain . Similar conclusions about gene transfer have emerged from comparisons of serogroups and sequences of diagnostic housekeeping genes of nonepidemic isolates . In this study, we have identified a new variety of V. cholerae O1 that appears to be a hybrid of the classical and El Tor biotypes from hospitalized patients with acute diarrhea. The phenotypic traits that distinguish the classical and El Tor biotypes of V. cholerae O1 and important discriminating genotypic characteristics of these strains are reported here, and the implications of the existence of such novel strains, especially in relation to cholera vaccine development, are described. MATERIALS AND METHODS : Twenty-four strains of V. cholerae isolated between 1991 and 1994 from hospitalized patients with acute diarrhea in the Matlab hospital, 45 km south of Dhaka, Bangladesh, were included in this study. The basis of a retrospective reexamination of these strains was their unusual response to polymyxin B (50 U), chicken cell agglutination (CCA), Voges-Proskauer (VP) reaction, and sensitivity to group IV and V phages, all of which are phenotypic traits commonly used to differentiate between the classical and El Tor biotypes. The 24 strains were reexamined for the above phenotypic characteristics by standard procedures. The presence of the ctxA gene and the variants of the classical and El Tor tcpA genes was determined by a multiplex PCR assay . The expected size of the PCR amplicons was ascertained by electrophoresis in agarose gels. The identities of all PCR products were further verified with specific oligonucleotide probes. The probes for El Tor and classical biotype-specific CTX prophage repressor rstR were SacI-XbaI fragments of pHK1 and pHK2, respectively . The acfB gene probe was prepared from the PCR amplicon with previously reported acfB-specific primers . The rRNA gene probe consisted of a 7.5-kb BamHI fragment of Escherichia coli rRNA clone pKK3535 . Colony blots or Southern blots were prepared with nylon filters (Hybond; Amersham International plc., Aylesbury, United Kingdom) by standard methods . The probes were labeled by random priming with a random-primer DNA labeling kit (Bethesda Research Laboratories, Gaithersburg, Md.) and [alpha-32P]dCTP (3,000 Ci/mmol; Amersham). Colony blots and Southern blots were hybridized with the probes and autoradiographed as described previously (-). RESULTS : The commonly used phenotypic traits used to distinguish between the El Tor and classical biotypes of V. cholerae differentiated the 24 strains into three types , which we classified as Matlab types I, II, and III (named after the place where these strains were first isolated). Matlab type I included two strains belonging to the Inaba serotype that were resistant to both the El Tor-specific group IV and the classical biotype-specific group V phages, negative by the CCA and VP tests (both are classical traits), and resistant to polymyxin B (an El Tor trait). Matlab type II included one strain belonging to the Ogawa serotype that was sensitive to the group IV phage but showed negative responses in the CCA and VP tests and was sensitive to polymyxin B, all of which are classical biotype characteristics. Matlab type III included 21 Ogawa strains that showed the sensitivity to phages and polymyxin B typical of the El Tor biotype but were negative by the CCA and VP tests (both classical biotype traits). Genotypically, all of the strains carried the ctxA gene, a constituent gene of the CTX prophage that encodes cholera toxin, and acfB and tcpA, which are located in different gene clusters (acf and tcp gene clusters) on the V. cholerae pathogenicity island. The type I strains appeared to belong more to the classical biotype because they carried the tcpA gene and the CTX prophage repressor gene (rstR) of the classical type . The tcpA gene of the single type II strain was of the classical type, while the rstR gene was of the El Tor type. The six representative strains of V. cholerae representing Matlab III also carried the tcpA gene of the classical type. Five of the strains had the El Tor-type rstR gene, while one carried both the El Tor and classical rstR types. The ribotypes of the V. cholerae strains examined in this study, compared to those of selected reference strains of the El Tor and classical biotypes, are shown in Fig. . The ribotypes of different strains representing the three Matlab types of V. cholerae were similar to the ribotypes of El Tor biotype strains and different from that of typical classical biotype strains described previously . The ribotypes of two type I strains (lanes 9 and 10) were similar to that of toxigenic El Tor strains 1849 (lane 11), isolated in 2001, and G-3669 (lane 1), isolated in 1969 in Bangladesh. The Matlab type III strains belonged to three different ribotypes (Fig. , lanes 2 through 7), and the single type II strain had the same ribotype as a type III strain. FIG. 1. | BglI restriction patterns of rRNA genes of V. cholerae BglI restriction patterns of rRNA genes of V. cholerae strains compared to those of selected typical strains of the El Tor and classical biotypes of V. cholerae O1. A Southern blot of BglI-digested genomic DNA was hybridized with the 7.5-kb BamHI fragment of E. coli rRNA clone pKK3535. Lanes (including strain designations and relevant characteristics): 1, toxigenic El Tor strain G-3669 (isolated in 1969 in Bangladesh); 2 through 10, strains MH-08 (Matlab type III), MG-117159 (Matlab type III), MG-117086 (Matlab type III), MG-116926 (Matlab type III), MG-116955 (Matlab type III), MG-116025 (Matlab type III), MG-116226 (Matlab type II), MJ-1485 (Matlab type I), and MJ-1236 (Matlab type I); 11, toxigenic El Tor strain 1849 (isolated in 2001); 12, toxigenic classical biotype strain (isolated in 1963 in Bangladesh). TABLE 1 | Phenotypic traits of Matlab types I, II, and III of toxigenic V. cholerae O1 isolated from patients hospitalized with acute secretory diarrhea in Bangladesh TABLE 2 | Genotypic traits of V. cholerae O1 strains isolated from hospitalized patients with acute diarrhea in Bangladesh DISCUSSION : Classical and El Tor strains of V. cholerae are closely related but are not directly derived from each other . El Tor vibrios appeared in Bangladesh, causing the first significant outbreak in 1968, and by 1973, they completely replaced the classical vibrios . In 1982, the classical biotype reappeared as the predominant epidemic strain in Bangladesh . In retrospect, it appears that classical cholera did not completely disappear from Bangladesh during the 1970s or late 1980s, but rather, its frequency varied in different regions of the country . The classical and El Tor biotypes have temporally overlapped for over a decade and are likely to have interacted and exchanged genetic material either in the human intestinal milieu or in the aquatic environment. The strains isolated in this study probably represent the amalgam of such an exchange. It is well recognized that genetic exchange between divergent bacterial lineages can contribute importantly to the success of a species in complex and inconstant environments, such as those in which V. cholerae may reside. Several studies have also pointed to such exchanges as an important factor in V. cholerae population genetics and evolution . On the basis of their phenotypic and genotypic traits, Matlab type I strains appeared to be more like the classical biotype while Matlab type II and III strains appeared to be more like the El Tor biotype. Matlab I strains, however, had altered phage receptor sites, since both of the strains were resistant to group IV and V phages. We assessed the similarity of the hybrid strains with classical and El Tor biotype strains on the basis of previously described ribotype patterns of classical and El Tor strains . Ribotyping demonstrated that the Matlab I, II, and III strains showed minor differences in fragment patterns shown by the El Tor standard strains, suggesting that the hybrids originated from an El Tor-like clone. Therefore, overall, these strains were of the El Tor biotype displaying traits of the classical biotype. It has been proposed that while El Tor and classical strains are not directly derived from each other but appear to be derived from environmental nontoxigenic strains that are El Tor like . Clinical strains might become classical like in some properties simply by loss of function, and this agrees with the finding of the present study. While some genetic exchange has also probably occurred, it appears that the strains have evolved classical biotype properties. With a V. cholerae genomic microarray that displayed more than 93% of the predicted genes of the whole genome sequence of El Tor strain N16961, Dziejman et al. showed that only seven genes were absent solely in classical strains but present in other strains, leading them to speculate that classical biotype strains may be derived from a primordial environmental strain that was more El Tor like than previously thought. Mitra et al. have previously reported the involvement of bacteriophage PS166 in the acquisition of some classical biotype-specific properties by El Tor strains . Insertion of lysogenic phage genomes in the bacterial chromosome leading to the activation or inactivation of certain genes or expression of new phage-encoded genes is a natural phenomenon in the origination of genetic diversity. However, in the present study, the acquisition of classical properties such as classical type tcpA and rstR genes by El Tor vibrios by conversion through phage PS166 seems unlikely. It seems more probable that more than one genetic exchanges were involved in the conversion of these strains. Irrespective of the mechanism involved in the generation of the natural hybrid strains, the existence of strains showing a combination of classical and El Tor biotype properties has evolutionary and epidemiological importance. Interestingly, all of the hybrid strains carried the tcpA gene of the classical type. Recently, the dominance of the classical type tcpA gene among environmental strains of V. cholerae has been reported . The primary structure of TcpA is highly conserved among V. cholerae serogroups and biotypes shown to be pathogenic to humans, with amino acid identities of nearly 100% between strains of a given biotype and about 80% between classical and El Tor biotype O1 strains . We are not certain if El Tor strains with classical tcpA are more efficient colonizers, but there is enough evidence showing that classical biotype strains elaborate abundant amounts of toxin-coregulated pilin when grown in vitro, in contrast to El Tor strains . The strains analyzed in the present study are not only of academic interest but may well represent precursors of other clones that could lead to a pandemic spread since they have all of the genetic features needed to make a V. cholerae strain pandemic. Moreover, these strains were isolated from clinical cases of acute diarrhea. These strains also represent unique natural recombinants that could be judiciously employed in the construction of live-vaccine strains since they have a combination of virulence attributes of both the classical and El Tor biotypes of V. cholerae O1. The classical biotype of V. cholerae O1 is believed to be extinct and has not been isolated in the past several years, even in southern Bangladesh, the last of the niches where this biotype prevailed (A. K. Siddique, unpublished data). In this study, we show the existence of El Tor strains that have lost some of the El Tor phenotypes and acquired classical biotype characteristics. Therefore, even though strains that represent the classical biotype in entirety have been completely displaced, a reservoir of the virulence genes of the classical biotype still exists in nature. Previous molecular analyses of classical strains isolated between 1961 and 1992 in Bangladesh support the contention that classical vibrios were never completely replaced in Bangladesh Thus, a vaccine developed against cholera must take this into consideration and must be targeted against both of the biotypes, failing which the global use of a vaccine exclusively against the El Tor biotype might select against El Tor strains and favor strains carrying the classical attributes, such as those isolated in this study. These hybrid strains of V. cholerae may be more common than currently recognized because phenotypic methods are inadequate to precisely distinguish between the two biotypes and are not routinely used in clinical microbiology laboratories. IS1004 fingerprinting has determined that an O37 strain of V. cholerae that was responsible for a large outbreak of cholera in Sudan in 1968 is closely related to classical O1 strains . This indicates that horizontal exchange of genes has occurred not only between O1 biotypes but also between classical biotype and non-O1 strains, and the Sudan strain is a typical example of how a novel genotype can cause a large outbreak. Although the strains characterized in this study were isolated a decade ago, the inadequacy of phenotypic tests did not permit us to place these strains accurately. However, molecular techniques have revealed that these strains carry traits of the two biotypes and that such strains exist in nature and are associated with sporadic diarrhea in Bangladesh and, presumably, other areas of the world where cholera is endemic. The recognition of such strains and tracking of their global prevalence and spread are important because each of these types possesses all of the traits necessary to initiate a pandemic spread. Backmatter: PMID- 12202570 TI - Comparative Assessment of Genotyping Methods for Epidemiologic Study of Burkholderia cepacia Genomovar III AB - We analyzed a collection of 97 well-characterized Burkholderia cepacia genomovar III isolates to evaluate multiple genomic typing systems, including pulsed-field gel electrophoresis (PFGE), BOX-PCR fingerprinting and random amplified polymorphic DNA (RAPD) typing. The typeability, reproducibility, and discriminatory power of these techniques were evaluated, and the results were compared to each other and to data obtained in previous studies by using multilocus restriction typing (MLRT). All methods showed excellent typeability. PFGE with SpeI was more reproducible than RAPD and BOX-PCR fingerprinting. The discriminatory power of the methods was variable, with PFGE and RAPD typing having a higher index of discrimination than BOX-PCR fingerprinting. In general, the results obtained by PFGE, BOX-PCR fingerprinting, and MLRT were in good agreement. Our data indicate that different genomic-based methods can be used to type B. cepacia genomovar III isolates depending on the situation and the epidemiologic question being addressed. PFGE and RAPD fingerprinting are best suited to addressing small-scale studies (i.e., local epidemiology), whereas BOX-PCR fingerprinting is more appropriate for large-scale studies (i.e., global epidemiology). In this regard, BOX-PCR fingerprinting can be considered a rapid and easy alternative to MLRT. Keywords: Introduction : Cystic fibrosis (CF) is the most common hereditary disease in Caucasian populations. Clinical manifestations of CF result from a disturbance in electrolyte transport that primarily affects the respiratory and digestive systems. The CF lung is particularly susceptible to infection with a variety of opportunistic bacteria , and exacerbations of chronic infection cause significant morbidity and mortality . Among the bacterial species capable of causing infection in CF are those belonging to the Burkholderia cepacia complex, which is currently comprised of nine closely related genomic species or genomovars . Recent work has demonstrated that the majority of infected CF patients harbor either B. cepacia genomovar III or Burkholderia multivorans (genomovar II) . Furthermore, limited data suggest that B. cepacia genomovar III (or perhaps certain specific strains within genomovar III) may be relatively more virulent than other species in this complex . The broad-spectrum antimicrobial resistance, absence of a vaccine, and virulence of certain strains have made prevention of B. cepacia complex infection an important goal in CF patient care . However, much still remains unknown regarding the epidemiology of infection in CF. A number of previous studies have demonstrated transmission of B. cepacia complex strains between persons with CF (for reviews, see references , , , and ). More recent studies indicate that the natural environment is also a likely reservoir for acquisition of B. cepacia complex strains . Better risk assessment of potential sources of infection and the development of optimal infection control policies rely on a more complete understanding of the molecular epidemiology of B. cepacia complex infection in CF. A number of methods have been used to establish relationships between B. cepacia complex isolates, including phenotypic assays, such as serotyping, antimicrobial susceptibility typing, bacteriocin typing, and biotyping . In recent years, phenotypic methods have been largely replaced by genotypic methods, including macrorestriction digestion of chromosomal DNA followed by pulsed-field gel electrophoresis (PFGE) and various PCR-based fingerprinting techniques . Among these, PFGE is generally considered the "gold standard" in bacteriological typing , and a number of studies have applied PFGE in studies assessing B. cepacia complex epidemiology . PCR-based fingerprinting with short random primers or primers directed against repetitive sequences in the bacterial genome are also increasingly being used for typing B. cepacia complex organisms (, , , -, ). Recently, we introduced multilocus restriction typing (MLRT) as yet another method for genotyping B. cepacia complex . In MLRT, genomic diversity is indexed through restriction fragment length polymorphism analysis of several housekeeping genes. MLRT is particularly well suited to studies analyzing B. cepacia complex isolates collected in large-scale epidemiologic studies . In the present study, we compared results obtained with PFGE and repetitive sequence PCR by using a BOX A1R primer (BOX-PCR fingerprinting) for a set of 97 epidemiologically well-characterized B. cepacia genomovar III isolates. A subset of these 97 isolates was also analyzed by using random amplified polymorphic DNA (RAPD) typing. Typeability, reproducibility, and discriminatory power of all techniques were compared to each other and to MLRT data obtained in previous studies. MATERIALS AND METHODS : Bacterial strains, growth conditions, and species identification. | Isolates were obtained from the B. cepacia Research Laboratory and Repository (University of Michigan, Ann Arbor, Mich.) and had been identified as B. cepacia genomovar III by using ribosomal DNA (rDNA)- and recA-based PCR assays, as previously described . Approximately two-thirds (n = 66) were recovered from CF sputum culture from persons receiving care in 23 CF treatment centers throughout North America; the rest of the isolates were recovered from soil. All had been typed previously by one or more genotyping methods, including PFGE or MLRT . Several clusters of isolates from patients receiving care in the same geographic region were included in this set. Isolates from frozen stocks were grown aerobically on Mueller-Hinton broth (Becton Dickinson) supplemented with 2.2% (wt/vol) agar and incubated overnight at 32C. Macrorestriction digest and pulsed field gel electrophoresis. | Single bacterial colonies were removed from an agar plate, suspended in 1 ml of SE buffer (75 mM NaCl, 25 mM EDTA [pH 7.4]), pelleted by centrifugation at 4,300 x g for 3 min, washed 3 times in 1 ml of SE buffer, and resuspended in 0.5 ml of SE buffer. The optical density at 620 nm was adjusted to approximately 1.0. Two hundred microliters of the cell suspension was homogenized with 200 mul of 2% low-melting-temperature agarose (Sigma, St Louis, Mo.) in 0.5x TBE buffer (45 mM Tris-borate, 1 mM EDTA) and poured into a plug mold. After 15 min at 4C, agarose plugs were placed in 10 ml of PEN buffer (1.0% N-lauryl sarcosine, 500 mM EDTA [pH 9.6]) containing 1 mg of protease ml-1. After incubation at 37C in a rocking incubator for 12 to 18 h, plugs were washed four times for 1 h per wash with 1x TE buffer (10 mM Tris-HCl, 1 mM EDTA [pH 8.0]). Two-millimeter-wide plug sections were cut and incubated for 4 h with 5 U of SpeI (Promega, Madison, Wis.) in 115 mul of digestion buffer. DNA fragments were separated in 1% PFGE-certified agarose (Bio-Rad, Hercules, Calif.) by using a CHEF DRIII system (Bio-Rad). Plugs containing digested DNA of B. cepacia strain AU2725 were included on each gel to allow intra- and intergel normalization. A current of 5.0 V/cm was applied for 25 h, with pulse times of 30 to 70 s (linear ramping). Gels were stained with ethidium bromide and visualized with UV illumination. Gel images were digitized using a GelDoc2000 gel analyzer (Bio-Rad) and stored as TIF files. Digitized images were converted, normalized with the reference lanes (containing DNA from strain AU2725), and analyzed by using Molecular Analyst Fingerprinting Plus software (Bio-Rad). The rolling disk background subtraction method was applied, and similarity matrices of densitometric curves of the gel tracks (the first 16% and last 7.5% of data points were excluded from the analysis) were calculated by using Pearson's product-moment correlation coefficient. Cluster analyses of similarity matrices were performed by the unweighted pair group method with arithmetic averages (UPGMA). BOX-PCR fingerprinting. | DNA from each isolate was prepared by heating one colony at 95C for 15 min in 20 mul of lysis buffer containing 0.25% (wt/vol) sodium dodecyl sulfate (SDS) and 0.05 M NaOH. Following lysis, 180 mul of distilled water was added, and the DNA solutions were stored at 4C. Rep-PCR typing with a BOX-A1R primer (5'-CTACGGCAAGGCGACGCTGACG-3') (BOX-PCR fingerprinting) was carried out as described previously . Briefly, 2 mul of DNA solution was mixed with 2 U of Taq polymerase (Gibco BRL, Gaithersburg, Md.), 1.25 mul of 25 mM (each) of deoxynucleotide triphosphate (Gibco BRL), 2.5 mul of dimethyl sulfoxide (DMSO), 0.4 mul of bovine serum albumin (20 mg ml-1) (Promega), 5 mul of 5x Gitschier buffer and 1 mul of primer (0.3 mug mul-1) in a final volume of 25 mul. Amplification was carried out with a PTC-100 programmable thermal cycler (MJ Research, Incline Village, Nev.). After initial denaturation for 2 min at 95C, 35 amplification cycles were completed, each consisting of 3 s at 94C, 30 s at 92C, 1 min at 50C, and 8 min at 65C. A final extension of 8 min at 65C was applied. PCR products were separated on 25-cm-long 1.5% agarose gels in 0.5x TBE buffer (60 mA for 4 h at room temperature). A 1-kb molecular weight marker (Gibco) was used multiple times on each gel to allow normalization. Following staining with ethidium bromide and visualization by UV illumination, gels were analyzed as described above for PFGE (the first 12.5% and last 8.5% of data points were excluded from the analysis). RAPD typing. | DNA was prepared as described above for BOX-PCR. RAPD fingerprinting was performed with primer RAPD-270 (5'-TGCGCGCGGG-3') as described previously . Briefly, 2 mul of DNA solution was mixed with 1 U of Taq polymerase (Gibco BRL, Gaithersburg, Md.), 2.5 mul of 2.5 mM (each) deoxynucleotide triphosphates (Gibco BRL), 0.7 mul of bovine serum albumin (20 mg ml-1) (Promega), 2.5 mul of 10x buffer (100 mM Tris-HCl [pH 8.0], 0.5 M KCl, 30 mM MgCl2, 1.0% gelatin), and 0.4 mul of primer (100 pmol mul-1) in a final volume of 25 mul. Amplification was carried out with a RapidCycler programmable thermal cycler (Idaho Tech, Idaho Falls, Idaho). The first four amplification cycles were each for 1 min at 94C, 1 min at 36C, and 2.5 min at 72C. The following 29 cycles were 30 s at 94C, 30 s at 36C, and 75 s at 72C. A final extension of 2 min at 72C was applied. Numerical analysis was performed as described above for PFGE (the first 15.75% of data points were excluded from the analysis). Statistical analyses. | Two-tailed unpaired t tests were used to compare reproducibility among replicate assays within a given typing method and were calculated by using GraphPad Prism 3.00 (GraphPad Software, San Diego, Calif.). Discriminatory index (DI) is the probability that two isolates randomly chosen from a population of unrelated isolates will be distinguished by a given typing method within the confines of a specific cutoff value. DI is determined by the number and relative frequencies of the different types defined by a given method and is calculated by using Simpson's index of diversity as follows: where N is the total number of isolates and nj is the number of isolates belonging to the jth type. The cophenetic correlation coefficient is the product-moment correlation between all original matrix similarities and all corresponding similarity values derived from the resulting dendrogram. As such, it provides a measure of whether or not a matrix can be faithfully represented as a bifurcating tree. Cophenetic correlation coefficients of the UPGMA dendrograms produced by using each typing method were calculated by using Molecular Analyst software. TABLE 1 | List of isolates used and their molecular profiles RESULTS : PFGE typing. | An illustration of PFGE patterns is shown in Fig . Reproducibility was checked by preparing agarose-embedded DNA from 15 isolates two or more times. Similarity coefficients between these replicates ranged between 90.7 and 97.6% (mean +- standard deviation, 93.84% +- 2.34%). Among the 97 isolates studied, 10 clusters could be delineated (designated P1 to P10) by using a cutoff value of 65% similarity, while 22 isolates occupied separate positions in the dendrogram (Fig. and Table ). Isolates belonging to the same cluster did not differ by more than six bands, although in general, isolates belonging to different clusters differed by more than six bands (data not shown). The DI for PFGE by using a 65% similarity cutoff value was 0.920. The cophenetic correlation coefficient of the UPGMA dendrogram was 79.7%. BOX-PCR fingerprinting. | An illustration of BOX-PCR patterns is shown in Fig. . Reproducibility was assessed by obtaining a pattern from 15 isolates two or more times. Similarity indices between these replicates ranged between 83.6 and 97.2% (mean +- standard deviation, 90.27% +- 3.88%). By using a cutoff value of 70% similarity, eight clusters could be delineated (designated B1 to B8), while 13 isolates occupied separate positions in the dendrogram (Fig. and Table ). The DI of BOX-PCR fingerprinting with a 70% similarity cutoff value was 0.821. The cophenetic correlation coefficient of the dendrogram was 96.3%. RAPD typing. | From the same set of 97 isolates analyzed by PFGE and BOX-PCR, a subset of 31 isolates (recovered from CF patients receiving care in four different treatment centers in the same U.S. state) was further investigated by using RAPD typing. An illustration of RAPD typing is shown in Fig. . Reproducibility was assessed by obtaining a pattern from six isolates two times. Similarity indices between these replicates ranged from 71.0 to 97.4% (mean +- standard deviation, 83.88% +- 10.39%). With a cutoff value of 80% similarity, three clusters could be delineated (designated R1 to R3), while six isolates occupied separate positions in the dendrogram (Fig. and Table ). The DI for RAPD typing with an 80% similarity cutoff value was 0.738. Comparison of methods. | We found a good correlation between PFGE and BOX-PCR fingerprinting. Most clusters delineated in the dendrogram based on BOX-PCR fingerprinting with a 70% coefficient cutoff have corresponding clusters in the PFGE dendrogram with a 65% cutoff (Fig. and Table ). Specifically, isolates grouping in clusters P5, P7, P9, and P10 in the PFGE dendrogram group in clusters B3, B6, B7, and B8, respectively, in the BOX-PCR dendrogram. The 10 isolates in cluster P5/B3 were recovered from patients receiving care in nine Canadian cities; these had previously been identified by PFGE analysis as ET12, the genomovar III strain that predominates among infected CF patients in Canada and the United Kingdom . The 18 isolates in cluster P10/B8 were recovered from CF patients receiving care in the same U.S. state. Isolates in PFGE clusters P1, P2, P3, and P4 group together in the BOX-PCR dendrogram in cluster B5. The isolates in cluster P1 were recovered from patients in two large CF treatment centers in the same region of the U.S. , while those in P2 and P4 were recovered from soil in the same geographic region . Isolates belonging to BOX-PCR cluster B1 either group in cluster P6 or P8 or occupy separate positions in the dendrogram based on PFGE. The 10 isolates in clusters P6 and P8 were recovered from patients in the same U.S. state. Cluster B2 is composed of one isolate belonging to cluster P6 and one isolate that occupies a separate position in the PFGE dendrogram. BOX-PCR cluster B4 is composed of two isolates that occupy separate positions in the PFGE dendrogram. In general, there was also good agreement between PFGE and RAPD typing for the subset of 31 isolates (Fig. and Table ). Isolates grouping in clusters P'1 and P'2 in the PFGE dendrogram group in clusters R1 and R3, respectively, in the RAPD dendrogram with a cutoff of 80%. Isolates AU2718 and AU0604 group together in PFGE cluster P'3; these isolates group together with isolate AU0824 in RAPD cluster R3. The discriminatory powers of PFGE (DI = 0.712) and RAPD (DI = 0.738) were similar for this data set. We also compared the results obtained with BOX-PCR and PFGE with those previously obtained with MLRT analysis of the same set of 97 isolates (; and T. Coenye and J. J. LiPuma, submitted for publication). MLRT allows the grouping of isolates in a hierarchical fashion such that isolates are first grouped by restriction type (RT), consisting of a unique combination of RFLP profiles for five genetic loci. RTs are then grouped into clonal complexes, defined as groups in which each RT is identical to at least one other RT at three or more of the five loci (T. Coenye and J. J. LiPuma, submitted for publication). For the data set used in this study, the DI of MLRT is 0.952 when using the RT as a cutoff; the DI is considerably lower (0.756) when the cutoff is placed at the clonal complex level. As reported previously , there was a good correlation between data obtained with MLRT and data obtained with PFGE . In the present analysis, we also found an excellent correlation between data obtained with MLRT and that obtained with BOX-PCR fingerprinting ; overall, 95.9% of isolates were grouped the same by both methods. FIG. 1. | Illustration of genotyping methods with representative B. cepacia Illustration of genotyping methods with representative B. cepacia complex genomovar III isolates. PFGE, BOX-PCR, RAPD gels, and UPGMA dendrograms were produced as described in the text. Scale bars indicate percent similarity. FIG. 2. | Dendrograms derived from the UPGMA linkage of correlation coefficients between the PFGE patterns (left) and BOX-PCR patterns (right). Dendrograms derived from the UPGMA linkage of correlation coefficients between the PFGE patterns (left) and BOX-PCR patterns (right). Clusters were delineated with a 65% similarity cutoff for PFGE and a 70% similarity cutoff level for BOX-PCR, as indicated by the heavy vertical lines. Corresponding clusters are indicated by connecting arrows. (Solid, dashed, or dotted lines are included for ease of interpretation only.) FIG. 3. | Dendrograms derived from the UPGMA linkage of correlation coefficients between the PFGE patterns (left) and RAPD patterns (right). Dendrograms derived from the UPGMA linkage of correlation coefficients between the PFGE patterns (left) and RAPD patterns (right). Clusters were delineated with 65 and 80% similarity cutoff values for PFGE and RAPD, respectively, as indicated by the heavy vertical lines. TABLE 2 | Subset of isolates investigated by RAPD typing DISCUSSION : Previous studies regarding the molecular epidemiology of infection due to B. cepacia complex species have employed a variety of genotyping methods, including ribotyping, PFGE, RAPD, ERIC-PCR, and BOX-PCR fingerprinting . Among these methods, PFGE and RAPD have emerged as the most widely used in recent studies; however, few studies have compared different genotyping methods in a systematic way. Bingen et al. concluded that RAPD was less discriminative than PFGE for the study of bacteria in the B. cepacia complex. In contrast, Liu et al. showed that these methods had comparable discriminatory power, but noted that PFGE was considerably more reproducible. These investigators also concluded that ERIC-PCR should be considered a valid and reproducible alternative to PFGE. However, in both studies, relatively small sets of B. cepacia complex isolates were evaluated in the comparison of these methods. In a more recent study, we explored the use of MLRT as an alternative genotyping method and showed a strong correlation between PFGE and MLRT in an analysis of a larger set of B. cepacia genomovar III isolates . In the present study, we sought to expand these findings to include other commonly used PCR-based methods, including BOX-PCR and RAPD typing. We also wished to more specifically identify the strengths and weaknesses of these methods for typing B. cepacia genomovar III, to compare these results with MLRT, and to determine which methods are best suited for addressing questions in the settings of local and global epidemiology. Several criteria have been used in evaluating bacterial genotyping methods, including typeability, reproducibility, discriminatory power, and ease of interpretation . Typeability describes the ability of a given method to provide a readable result for each isolate analyzed . All isolates included in this study were typeable by each method used. In previous work, we have noted that a small fraction of B. cepacia complex isolates are refractory to PFGE typing, presumably due to the presence of high DNase activity (data not shown). Reproducibility measures the ability of a technique to yield the same result when replicate assays are performed on the same isolate . In this study, we found that PFGE was significantly more reproducible than BOX-PCR and RAPD typing (P < 0.05). The reproducibility values of PFGE were also in a significantly narrower range than those of BOX-PCR (P < 0.05) and RAPD (P < 0.0001). The high reproducibility of PFGE is in agreement with findings of most previous studies . In contrast, previous studies have provided conflicting data regarding the reproducibility of RAPD typing . Excellent reproducibility was reported by Mahenthiralingam et al. and Segonds et al. , while significant day-to-day variation in RAPD patterns obtained from the same B. cepacia complex isolate was noted by Bingen et al. and Liu et al. . In our study, we similarly found the reproducibility of RAPD profiles to be rather low. Reproducibility of BOX-PCR fingerprinting was intermediate between that of PFGE and RAPD; moreover, the values we obtained were generally in agreement with those reported by others . Discriminatory power defines the ability of a typing method to distinguish different strains. This may be expressed as an index that measures the probability that two unrelated strains will be placed into different groups. Discriminatory power is most conveniently calculated by using Simpson's index of diversity (DI), which takes into account the number of types defined by the method and the relative frequencies of these types . A DI value of >0.90 has been considered adequate in previous assessments of genotyping methods . In our study, only PFGE (with DI = 0.920, with this data set and a 65% similarity coefficient cutoff) would meet this stringent criterion. However, higher discriminatory power does not necessarily always result in a more accurate representation of epidemiologic relatedness. While this may be true in the setting of outbreak epidemiology, in which very high discriminatory power is needed to trace patient-to-patient transmission or nosocomial outbreaks, methods based on such a high DI may very well be less suitable for the analysis of large populations of organisms collected over extended periods of time. This is illustrated by the results previously obtained by using MLRT analysis of this same set of B. cepacia genomovar III isolates. Important relationships among isolates grouped at the clonal complex level (i.e., employing a significantly lower DI) are not apparent based on PFGE analysis. In this regard, MLRT provides greater utility in allowing meaningful analysis of relationships among isolates collected both in larger and smaller scale (both temporal and spatial) studies . From this perspective, it is important to note that, in this study, the DI of BOX-PCR fingerprinting with a 70% similarity cutoff (DI = 0.821) is intermediate between the DIs of MLRT analysis according to either RT (DI = 0.952) or clonal complex (DI = 0.756)-level cutoffs. Of course, the discriminatory power of any given typing method is determined by the data set and by the cutoff values used. The cutoff values used in this study were based on (i) the available epidemiologic data (i.e., geographic location and previously determined epidemiological relationships), (ii) the reproducibility of the methods, and (iii) the cophenetic correlation coefficient. It is possible that with other data sets, these similarity coefficient cutoff values may need to be adjusted to better accommodate these variables. A final criterion to consider in assessment of typing methods is ease of performance. This includes not only performance of the assay, but also interpretation of the resulting data. PFGE is without doubt the more laborious and time-consuming technique among those examined. In contrast to PCR-based methods, PFGE requires a much longer time to perform (4 days with the protocol described above) and more specialized equipment . BOX-PCR and RAPD fingerprinting are significantly less cumbersome; results can be obtained within 1 working day. MLRT occupies an intermediate position in this regard. Whereas consensus guidelines for interpreting DNA restriction patterns generated by PFGE have been published , comparable criteria for interpretation of RAPD and BOX-PCR patterns are not available. However, the PFGE interpretative criteria were intended to be used only as an aid in the visual examination of small sets of isolates related to putative outbreaks of disease. These criteria are impractical and quite limited for analyzing larger sets of isolates where multiple pairwise comparisons are required. Indeed, for all the genotyping methods under consideration, visual comparison of large number of complex fingerprint patterns is not only time-consuming but also highly subjective. The use of equipment to digitize patterns and software to perform numerical analysis of these patterns are necessary for studies involving typing of a significant number of isolates. The data presented in this study indicate that there are strengths and weaknesses among the various genotyping methods that have been used to investigate the epidemiology of B. cepacia genomovar III. Which technique is chosen for a given study depends not only on the preferences of the investigators and the resources available, but most importantly on the specific epidemiologic question being addressed. Our data indicate that the reproducibility of RAPD typing is not sufficient to allow reliable comparisons across large numbers of assays in large-scale studies. This relatively poor reproducibility also limits the portability of results between laboratories. On the other hand, RAPD analysis is well suited to smaller-scale studies, such as investigation of a hospital outbreak, in which a limited number of samples is collected within a narrow time frame. In this setting, PFGE would be a more reproducible and portable, but also more time-consuming and expensive alternative. For more global epidemiological questions, involving larger number of isolates collected over a longer time frame, both MLRT and BOX-PCR fingerprinting might be considered the methods of choice. While data derived from MLRT can be used for population structure analysis , BOX-PCR fingerprinting has the advantage of being a more rapid and less expensive method. Backmatter: PMID- 12202610 TI - Prevalence of Blood-Borne Infectious Diseases in Blood Donors in Ghana AB - Transfusion-transmissible infections among 808 blood donors in Ghana were investigated in 1999. Antibody seroprevalences of 3.8, 0.7, 8.4, and 13.5%, respectively, for human immunodeficiency virus, human T-cell lymphotrophic virus type 1, hepatitis C virus (HCV), and Treponema pallidum were obtained. The seroprevalence of HCV infection was confirmed to be 0.9% after supplementary testing, and the transfusion risk potential of these pathogens was demonstrated. Keywords: Introduction : Although blood transfusion saves millions of lives worldwide each year, recipients of transfusions risk becoming infected with blood-borne pathogens. Each year, up to 4 million blood donations worldwide are not tested for human immunodeficiency virus (HIV) and few are tested for hepatitis B and C viruses (HBV and HCV, respectively). Virtually none are screened for human T-cell lymphotrophic virus type 1 (HTLV-1) or Treponema pallidum, the causative agent of syphilis . Several studies have previously indicated the high prevalence of HBV in Ghana , and prescreening of blood donors for HBV surface antigen (HBsAg) is thus a routine practice. It is also a standard procedure to screen Ghanaian blood donors for HIV. HTLV-1, which causes leukemia , was reported to be associated with HIV-1-seropositive individuals in Ghana (-). HCV is recognized as the primary cause worldwide of transfusion-associated non-A-non-B viral hepatitis and is endemic in West Africa . However, information on HCV seroprevalence in Ghana is limited, and blood donors are not routinely screened for HCV . T. pallidum, the etiologic agent of syphilis , is prevalent in many African countries , but in Ghana, data on T. pallidum seroprevalence are scanty, with antibodies thought to occur as frequently as HBV antibodies . This study was therefore carried out to determine the current prevalence of HTLV-1, T. pallidum, and particularly HCV in Ghanaian blood donors in order to provide information for appropriate policies. The National Blood Transfusion Service of Ghana currently selects blood donors on the basis of a health check questionnaire and prescreening for HBsAg; donated blood is then tested for the presence of HIV antibodies. We studied 3,131 individuals who presented at the National Blood Transfusion Service, Accra, Ghana, between June and August 1999 and who were routinely tested for HBsAg with a latex agglutination test kit (Biotech Laboratories Ltd., Suffolk, United Kingdom). Of these donors who were seronegative for HBsAg, 808 were randomly adopted as study subjects. Five milliliters of blood was collected from each of the 808 donors, labeled, and transported in coolers to the Virology Unit at the Noguchi Memorial Institute for Medical Research. Sera were then analyzed for antibodies to HIV, HTLV-1, HCV, and T. pallidum with SERODIA passive-particle agglutination assay kits (FUJIREBIO Inc., Tokyo, Japan). Qualitative testing protocols were applied according to the manufacturer's instructions, and serum dilutions were 1:16 for HTLV-1, 1:32 for HIV and HCV, and 1:80 for T. pallidum. Supplementary tests were deemed necessary to confirm HCV infection, as the samples were from healthy, asymptomatic individuals. Therefore, 68 samples shown by the SERODIA assay to be anti-HCV positive at a 1:32 serum dilution were retested at a 1:400 serum dilution, subjected to the HCV-SPOT assay (Genelabs Diagnostics Ltd., Singapore), and examined by an enzyme-linked immunoassay (IMUCHECK-HCV C50Ab; International Reagents Corporation, Kobe, Japan). Furthermore, a third-generation recombinant immunoblot assay (RIBA 3; Ortho Diagnostic Systems, Roissy, France) was applied. RIBA 3 detects antibodies to five structural and nonstructural HCV proteins (c100, c33c, c22p, NS5, and superoxide dismutase), enabling the determination of a full immunoblot profile . Test sera were considered positive when at least two of these antibodies were detected. Reverse transcription-PCR (RT-PCR) was also performed to confirm the presence of the HCV genome. HCV RNA in the sera was identified by a nested RT-PCR method using primers derived from the 5' untranslated region as previously described . The majority of the 808 blood donors lived in or around Accra, Ghana. Thirty (3.7%) of the donors were regular voluntary donors, and 778 (96.3%) were replacement donors who were family members of blood recipients. As shown in Table , the 21-to-25-year age group, which included 212 (26.2%) of the donors, was the largest, followed by the 26-to-30-year age group, with 186 donors (23.0%). The smallest group was that of the 56 to 60 year olds, with only 2 donors (0.2%). Overall, 46 donors (5.7%) were female and 762 donors (94.3%) were male. This trend of male bias is a regular feature at Ghanaian blood donation sites and is commonly observed during blood donation campaigns (J. Ansah, unpublished data). TABLE 1 | Age distribution and seroprevalence of anti-HIV, anti-TP, anti-HTLV-1, and anti-HCV in blood donors by the SERODIA assays The HBV seroprevalence rate for the study period was estimated to be 15.0% by routine prescreening, as 469 of 3,131 individuals were seropositive for HBsAg. Seroprevalence rates obtained by screening with SERODIA were as follows: for anti-T. pallidum, 13.5% (109 were seropositive); for anti-HCV, 8.4% (68 were seropositive); for anti-HIV, 3.8% (31 were seropositive); and for anti-HTLV-1, 0.7% (6 were seropositive). The presence of anti-HCV in the 68 sera initially found to be positive was confirmed as follows: 62 sera were confirmed by the HCV-SPOT assay, 7 sera were confirmed by RIBA 3, 5 sera were confirmed by the SERODIA assay (serum dilution, 1:400), and 3 sera were confirmed by IMUCHECK. Two sera were found to be HCV positive by RT-PCR. The reactivity profiles for samples that were found to be HCV positive by either RIBA 3, RT-PCR, or the SERODIA (dilution, 1:400) or IMUCHECK assay are shown in Table . TABLE 2 | Reactivity profiles for seven subjects positive for anti-HCV by either RIBA 3, PCR, the SERODIA assay, or the HCV-SPOT assay The 15% HBV seroprevalence level indicated by the HBsAg prescreening data for the 3-month duration of the investigation is similar to the previously reported HBsAg seroprevalence of 15.8% . The current transfusion transmission risk potential in Ghana for HTLV-1, HCV, and T. pallidum is illustrated by the data presented in Table . Total seroprevalence levels were highest in the age groups (21 to 36 years) corresponding to those described as the most sexually active . The highest seroprevalence observed was for anti-T. pallidum (13.5%). This corresponds with the results of previous studies of sexually transmitted diseases in Ghana, where T. pallidum and HBV were noted as the most frequently occurring pathogens . The seroprevalence of the anti-HTLV-1 antibody was found to be 0.7%, and the antibody was detected in male blood donors under 40 years of age. The low HTLV-1 seroprevalence obtained by our study confirms the earlier observation of low HTLV-1 antibody levels in Ghana . Previously, HTLV-1 antibodies were associated with HIV and HCV infections , and 19% of the dual infections observed in our study involved HTLV-1 (one case with HIV, three cases with HCV). The 3.8% seroprevalence level obtained for HIV vindicates the screening of donated blood for HIV, and the national seroprevalence of HIV was estimated to be 3% in 2001 . HIV was involved in 59% of the multiple infections recorded and was a major dual infection with T. pallidum. Use of the SERODIA or HCV-SPOT assay resulted in a high rate of anti-HCV false-positive results, which were resolved by supplementary assay (especially RIBA 3), and overall, HCV seroprevalence was 0.9%. Other reports on anti-HCV seroprevalence in Ghana, determined by screening assays, found seroprevalence rates of 5.4% in children , 2.8% in adults , and 5.2% in blood donors . Supplemental tests such as RIBA 3 are necessary to confirm the presence of HCV infection in asymptomatic Ghanaians. The presence of HCV in the blood is indicated by positive detection by RT-PCR , and our data showed two active cases of HCV infection among the blood donors. In conclusion, this study illustrates the current transfusion-transmissible risk of T. pallidum, HTLV-1, and HCV in Ghana. It is recommended that routine blood screening prior to transfusion should include tests for anti-HCV and anti-T. pallidum antibodies. Developing appropriate methods for HCV diagnosis will require an evaluation of the cost-effectiveness of general screening and/or supplementary assays of donated blood. Periodic studies to investigate transfusion-transmissible infectious diseases are required to enable safety reviews of the blood supply. Backmatter: PMID- 12202544 TI - Weak Association between SEN Virus Viremia and Liver Disease AB - Recently, a novel DNA virus designated SEN virus (SEN-V), which is thought to be related to posttransfusion hepatitis, was discovered. The aim of the present study was to clarify the relationship between SEN-V infection and the development of liver disease. We examined SEN-V from the sera of 21 patients with non-B, non-C hepatocellular carcinoma (HCC) and 13 patients with non-B, non-C chronic liver disease (CLD) without HCC who were admitted to our hospital between 1995 and 1997. Thirty-two patients without liver disease served as controls and were also examined for SEN-V. SEN-V DNA was detected by the nested PCR method after extraction of DNA from serum. SEN-V DNA was detected in 74% (25 of 34) of patients with CLD with or without HCC who were negative for both hepatitis B virus surface antigen and anti-hepatitis C virus antibody. SEN-V DNA was detected in 69% (9 of 13) of CLD patients without HCC and in 76% (16 of 21) of HCC patients. The prevalence of SEN-V was no higher in patients with liver disease than in patients without liver disease (24 of 32; 75%). There were no significant differences in age, sex, liver function, history of blood transfusion, or amount of alcohol intake between SEN-V-positive and SEN-V-negative CLD and HCC patients. Genetic analysis suggested that SEN-V is closely related to the TT virus family. SEN-V was detected at almost the same frequency in patients with and without liver disease. SEN-V does not seem to contribute either to the pathogenesis of liver disease or to the development of HCC from chronic liver disease. Keywords: Introduction : The number of patients with hepatocellular carcinoma (HCC) has been increasing over the last 30 years in Japan, and approximately 30,000 patients die of HCC every year . The majority of these patients are positive for either hepatitis B virus surface antigen (HBsAg) or anti-hepatitis C virus (anti-HCV) antibody. In a recent study, 11% of HCC patients were positive for HBsAg and 84% were positive for anti-HCV, but the remaining 4% were negative for both anti-HCV and HBsAg . The pathogenic agent in these 4% of patients remains unknown. An RNA virus associated with non-A to E hepatitis named GB virus C (GBV-C)/hepatitis G virus (HGV) and a DNA virus named TT virus (TTV) were recently cloned . However, previous studies showed that GBV-C/HGV and TTV are unlikely to be major etiologic agents of non-B, non-C HCC, and they have not yet been confirmed to be hepatotropic (-, , , , ). A novel DNA virus named SEN virus (SEN-V) was also recently cloned from the plasma of a patient with posttransfusion hepatitis. This patient had elevated transaminase levels but did not have viral markers for hepatitis type A to E viruses or hepatitis G virus . SEN-V is a circular 3,900-nucleotide DNA virus. There are eight SEN-V genotypes, called A to H. SEN-V was classified into the same group as TTV variants (including SANBAN and TUS01). Prototype TTV, TTV YONBAN, and TTV PMV formed an outer group of SEN-V . Of the eight genotypes, genotypes D and H (or genotype C) have been reported to be related to liver dysfunction . Although SEN-V has been observed in patients with acute and chronic liver disease (CLD) of unknown etiology, the role of SEN-V in the pathogenesis of liver disease is not yet known . We screened patients with non-B, non-C CLD and patients without liver disease for SEN-V DNA to determine the possible role of SEN-V in the pathogenesis of liver disease and in the development of HCC, especially in patients with non-B, non-C chronic hepatitis and HCC. MATERIALS AND METHODS : Patients. | We examined the prevalence of SEN-V DNA in 21 patients admitted to the University of Tokyo Hospital between 1995 and 1997 who were diagnosed with non-B, non-C HCC (the HCC patient group) and in 13 patients with non-B, non-C CLD without HCC (the CLD patient group). Thirty-two patients without liver disease were also examined and served as a control group. These patients presented to our hospital with abdominal complaints but had normal transaminase levels. The diagnosis of HCC was made by ultrasonography, computed tomography, magnetic resonance imaging, angiography, and liver biopsy. The clinical features examined included age, sex, history of blood transfusion, history of alcohol intake, serum transaminase levels, platelet count, and indocyanine green (ICG) retention rate (percent) at 15 min (the ICG R15 test). Anti-HCV was identified by a second-generation enzyme immunoassay (Ortho Diagnostics, Tokyo, Japan), and HBsAg was identified by radioimmunoassay (Abbott Laboratories, North Chicago, Ill.). Patients with an ethanol intake exceeding 80 g/day for more than 10 years were considered to have a positive history for alcohol abuse. Patients who were positive for antinuclear antibody or anti-smooth muscle antibody or who had a positive lupus erythematosus test result were excluded. Patients who had previously been treated with antiviral drugs, such as interferon and lamivudine, were also excluded. Serum samples from all patients with HCC were obtained at the first admission after the diagnosis of HCC. The serum samples were stored at -30C. Detection of SEN-V DNA. | DNA was extracted from 50 mul of serum with the SepaGene kit (Sanko Junyaku, Tokyo, Japan) according to the manufacturer's instructions. The extracted DNA was dissolved in 20 mul of Tris-HCl buffer (10 mM; pH 8.0) containing 1 mM EDTA, heated to 95C for 15 min, and quickly chilled on ice. The total amount of extracted DNA was subjected to nested PCR with Ready-To-Go PCR beads (Pharmacia Biotech, Uppsala, Sweden). The first round of PCR was performed with primer sets specific for each genotype. The sense primers specific for genotypes A to D and H were as follows: SEA-S, SEB-S, SEC-S, SED-S, and SEH-S, respectively . The antisense primers specific for genotypes A to D and H were as follows: SEA-AS, SEB-AS, SEC-AS, SED-AS, and SEH-AS, respectively . These primers were used to amplify 336-, 444-, 619-, 510-, and 830-bp fragments from genotypes A to D and H, respectively. The reaction was run in microtubes with the GeneAmp PCR system 9600 (Perkin-Elmer Applied Biosystems, Foster City, Calif.). Amplification was performed for 35 cycles, with each cycle consisting of denaturation at 94C for 1 min, annealing at 54C for 1 min, and elongation at 72C for 1 min. A 10-min final hold at 72C was used to complete strand synthesis. The second round of PCR was carried out with sense and antisense primers specific for genotypes A to D and H. The sense primers were as follows: SEA-S2, SEB-S2, SEC-S2, SED-S2, and SEH-S2, respectively . The antisense primers specific for genotypes A to D and H were as follows: SEA-AS2, SEB-AS2, SEC-AS2, SED-AS2, and SEH-AS2, respectively . These primers were used to amplify 288-, 396-, 577-, 459-, and 783-bp fragments from genotypes A to D and H, respectively. The amplification was performed for 30 cycles, with each cycle consisting of denaturation at 94C for 1 min, annealing at 54C for 1 min, and elongation at 72C for 1 min, followed by a final hold at 72C for 7 min. The amplified products were electrophoresed on 1.5% agarose gels, stained with ethidium bromide, and observed under UV light. Detection of SEN-V in DNA from human WBCs. | To exclude the possibility that we had amplified the host genome by PCR, we compared the detectabilities of SEN-V DNA from sera and from white blood cells (WBCs). We chose four patients and four controls (two patients each had SEN-V type D- and SEN-V type H-positive sera, and two controls each had SEN-V type D- and SEN-V type H-negative sera) and extracted WBC DNA from all of them. Each pair of samples was randomly chosen from groups of patients whose sera had the same SEN-V infection status. DNA was extracted from 3 ml of whole blood with the SepaGene kit (Sanko Junyaku, Tokyo, Japan) according to the instructions of the manufacturer. The extracted DNA was dissolved in 20 mul of Tris-HCl buffer (10 mM; pH 8.0) containing 1 mM EDTA, heated to 95C for 15 min, and quickly chilled on ice. The total amount of extracted DNA was subjected to nested PCR. We tried to detect SEN-V DNA in sera and WBCs by PCR, using the same primers and protocol and under the same conditions. Genetic analysis of SEN-V DNA. | Nucleotide sequencing of detectable SEN-V DNA was performed with an autosequencer (Perkin-Elmer Applied Biosystems, Foster City, Calif.) and the dye termination method as described previously . The direct sequencing reactions were carried out with primers specific for genotypes D and H: SED-S3 and SEH-S3, respectively . The sequences of SEN-V types D and H were compared to the prototype sequences of SEN-V (SEN-V type D, nucleotides 528 to 677; SEN-V type H, nucleotides 402 to 551) or the reported sequence of TTV (TTV nucleotides 701 to 850 and TTV nucleotides 401 to 550). A phylogenetic tree was constructed by the unweighted pair group method with the arithmetic mean with the computer software package GENETYX-MAC (Software Development Co., Ltd., Tokyo, Japan). Nucleotide sequence homologies among the SEN-V DNAs obtained from patients and controls were calculated and analyzed. In order to identify the full-length nucleotide sequence, fragments of SEN-V types D and H were amplified with the following PCR primer sets: SED Full F and SED(1) R, SED S2 and SED AS2, SED(3) F and SED(3) R, SED(4) F and SED(4) R, SED(5) F and SED(5) R, and SED(6) F and SED Full R and primer sets SEH Full F and SEH(1) R, SEH S2 and SEH AS2, SEH(3) F and SEH(3) R, SEH(4) F and SEH(4) R, SEH(5) F and SEH(5) R, and SEH(6) F and SEH Full R . Each primer set was designed to amplify one of six fragments from SEN-V type D (nucleotides 1 to 553, 495 to 953, 868 to 1589, 1439 to 2082, 1990 to 2688, and 2632 to 3264) or one of six fragments from SEN-V type H (nucleotides 1 to 446, 381 to 1163, 904 to 1593, 1376 to 2163, 1975 to 2558, and 2480 to 3292). The full-length nucleotide sequence was identified by direct sequencing with each set of sense and antisense primers. The codon distributions and hydrophobicities of the polyproteins encoded by the open reading frames (ORFs) of SEN-V types D and H were analyzed. Statistical analysis. | The results are expressed as the means +- standard deviations. Comparisons were performed by Student's t test and the chi-square test. Fisher's exact probability test was used to examine the relationship between HCC and SEN-V. A P value of less than 0.05 was considered statistically significant. Nucleotide sequence accession numbers. | The two full-length nucleotide sequences of SEN-V identified from a CLD patient (type D) and a control (type H) have been submitted to the DDBJ database and given accession numbers and , respectively. TABLE 1 | Primers used to detect or sequence SEN-V RESULTS : Incidence of SEN-V DNA in HCC and CLD patients. | SEN-V DNA was detected in 25 of 34 (74%) patients with non-B, non-C CLD, 9 of 13 (69%) patients with CLD, and 16 of 21 (76%) patients with HCC. SEN-V DNA was detected in 24 of 32 (75%) controls. The incidence of SEN-V was not significantly different between patients and controls or between patients with HCC and patients with CLD but without HCC. Clinical features of SEN-V-positive chronic hepatitis and HCC patients. | The clinical features of the patients with HCC and the patients with CLD but without HCC with respect to SEN-V infection status are shown in Table . The clinical features included age; sex; serum albumin, total bilirubin, and transaminase levels; the results of the ICG R15 test; platelet count; blood transfusion history; and a history of heavy alcohol intake. There were no statistically significant differences in these clinical features between SEN-V-positive and SEN-V-negative patients. There were no significant differences in tumor markers, such as alpha-fetoprotein and des-gamma-carboxyprothrombin levels, or in tumor factors (size, tumor-node-metastasis classification) between SEN-V-positive and SEN-V-negative HCC patients (data not shown). A background liver biopsy specimen was obtained from 12 of 16 SEN-V-positive HCC patients. Of the 12 SEN-V-positive HCC patients, 5 (42%) had chronic hepatitis and 6 (50%) had cirrhosis. A background liver biopsy specimen was obtained from four of the five SEN-V-negative HCC patients. Of the four SEN-V-negative HCC patients, one (25%) had chronic hepatitis and three (75%) had cirrhosis. The fibrotic stage of background liver disease was not significantly different between SEN-V-positive and SEN-V-negative patients, irrespective of HCC status (P = 0.89). Both the serum and the WBCs of only one of four SEN-V-positive patients and controls were SEN-V type D positive. However, the WBCs of none of the patients or controls who were SEN-V negative were SEN-V positive. Genetic analysis of SEN-V DNA. | Most of the SEN-V DNA sequences varied between patients, to the extent that we examined them . D1-13 and H11-18 were SEN-V DNAs obtained from the sera of controls, D15-30 and H20-45 were from the sera of HCC patients, and D31-59 and H53-66 were from the sera of CLD patients without HCC. There were no specific regions where mutations accumulated. All the samples from SEN-V type D- and H-positive patients that were sequenced were classified by genotype according to the genetic analysis shown in the phylogenetic tree. Each genotype-specific primer worked as expected . All of the clones of SEN-V types D and H obtained were classified as closely related to the TTV family . The nucleotide sequence homologies among the SEN-V DNAs obtained from patients and controls were calculated. The nucleotide sequence homologies were 80 to 100% (mean +- standard deviation, 93% +- 5%) among SEN-V type D DNAs and 85 to 97% (91% +- 4%) among SEN-V type H DNAs. Two SEN-V type D genomes and one SEN-V type H genome showed comparatively low levels of homology with other sequences, but there were no significant differences among the sequences of SEN-V DNA from controls, HCC patients, and CLD patients without HCC. Two full-length nucleotide sequences of SEN-V were identified from a CLD patient and a control. One was type D and the other was type H. These sequences had high degrees of homology with the sequences reported for SEN-V type D and H DNAs (data not shown). The codon distribution and hydrophobicity profile within the ORFs (SEN-V type D, nucleotides 252 to 725; SEN-V type H, nucleotides 256 to 723) of the SEN-V DNAs obtained and reported were analyzed. Differences in codon usage between the SEN-V DNAs obtained and reported were observed for 17 SEN-V type D codons and 18 SEN-V type H codons. The hydrophobicity profiles of amino acids 100 to 110 of the SEN-V D type ORF and amino acids 90 to 100 of the SEN-V type H ORF differed from those reported for SEN-V types D and H. FIG. 1. | Multiple-sequence alignments of SEN-V DNAs. Multiple-sequence alignments of SEN-V DNAs. The sequences of the samples were compared with the consensus sequence. Only the nucleotides that differed from the consensus sequence are shown. SEN-V type D nucleotides 528 to 677 and SEN-V type H nucleotides 402 to 551 are the consensus sequences. D1, D6, D8, D10, D12, D13, D15, D16, D17, D23, D25, D29, D30, D31, D32, and D59 and H11, H14, H18, H20, H22, H45, H53, and H66 are samples from patients (HCC, CLD) and controls (CNTL). FIG. 2. | Nucleotides 528 to 677 of SEN-V type D were compared with the sequences of the same regions of SEN-V types A to H and TTV nucleotides 701 to 850. Nucleotides 528 to 677 of SEN-V type D were compared with the sequences of the same regions of SEN-V types A to H and TTV nucleotides 701 to 850. Nucleotides 402 to 551 of SEN-V type H were compared with the sequences of the same regions of SEN-V types A to H and TTV nucleotides 402 to 551. The phylogenetic tree was constructed by the unweighted pair group method with the arithmetic mean by using the computer software package GENETYX-MAC (Software Development Co., Ltd.). The sample numbers are as described in the legend to Fig. . TABLE 2 | Clinical features of patients DISCUSSION : Since the discovery of HCV, it has become obvious that this virus causes the majority of cases of non-A, non-B CLD, including cirrhosis and HCC. However, the causative agent of non-B, non-C CLD remains unknown. Recently, the discovery of SEN-V, a novel DNA virus associated with posttransfusion hepatitis, was reported . The prevalence of SEN-V in our study group of patients with HCC or patients with CLD without HCC is similar to the prevalences reported in previous studies, in which SEN-V DNA was detected in 17 to 92% of patients with liver disease (, -). However, our study showed a high prevalence of SEN-V in controls, in contrast to the low prevalence (1 to 10%) in subjects without liver disease reported in previous studies . The difference in the rate of detection of SEN-V DNA between this and previous reports may be due to differences in the quantity of SEN-V DNA in the sera, differences in the PCR primers used, or differences in the sensitivities of the assay systems used. According to the results of sequence analysis, the specificity of the PCR was high enough. The sensitivity of the PCR could not be determined accurately because the concentration of SEN-V DNA varied among samples. The high rate of detection of SEN-V DNA in sera from healthy controls may indicate a high prevalence of SEN-V. To eliminate the possibility that we were detecting human genome sequences, we performed PCR using WBC DNA as a template. We also compared the detectabilities of SEN-V DNA from sera and from WBCs. WBCs from none of the SEN-V-negative patients was SEN-V DNA positive. If the SEN-V DNA detected originated in the human genome, it should mainly be detected in WBCs rather than in serum. Furthermore, the sequences of the DNAs that we amplified varied among the patients. The DNA sequences obtained from SEN-V type D- and type H-positive patients were classified by genotype, and those from SEN-V type A to C-positive and SEN-V type E to G-positive patients were classified in the phylogenetic tree. The intragenotype homologies between the sequences obtained from our patients and those reported for SEN-V DNA were at least 87% (SEN-V type D) and 90% (SEN-V type H). These results suggest that the DNA that we detected was viral DNA rather than DNA from the host genome. We also investigated whether there was a correlation between the presence of SEN-V and the severity of CLD with concomitant HCC. Infection with SEN-V did not have a significant effect on the status of liver disease. The prevalence of SEN-V DNA viremia in our study was similar in patients with HCC (76%) and patients with CLD without HCC (69%). Thus, SEN-V does not seem to contribute to the development of CLD or the development of HCC from CLD. The prevalence of TTV in this study was lower than that reported previously . The difference in the PCR primers used to detect the TTV DNA could explain the lower prevalence of TTV in our study . Genetic analysis of the SEN-V DNA sequences obtained showed no differences among the control, HCC, and CLD without HCC groups. In the analysis of full-length SEN-V type D and H DNAs, we found slight differences in codon distributions and hydrophobicities. In order to estimate whether these differences are significant, we must compare more sequences from strains of the same genotypes. On the basis of the results of this study, it is unlikely that SEN-V contributes to the incidence of liver disease or the development of HCC in Japan. However, a prospective study of the incidence of HCC in a large population of SEN-V-positive and -negative patients will be required to prove whether SEN-V contributes to the increasing tendency of chronic liver disease to progress to HCC in Japan. Backmatter: PMID- 12202617 TI - Importance of Testing Stool Specimens for Shiga Toxins AB - Keywords: Shiga toxins (ST) I and II, elaborated by serotypes of Escherichia coli O157:H7 (ECO157), are well established as the major cause of hemolytic uremic syndrome (HUS) and hemorrhagic colitis. According to the Centers for Disease Control and Prevention (CDC), at least 200 deaths and 20,000 illnesses in the United States are annually attributed to ST-producing E. coli (STEC) organisms belonging to the O157:H7 serotype. However, it is indisputable that serotypes other than O157 also cause clinical illness identical to that of STO157, and more than 50 STECs are believed to exist. The STECs other than O157 (STNO157) are widely prevalent outside the United States but are believed to be at very low prevalence in this country. For this reason, the majority of the laboratories in this country do not search for STNO157. The delayed advent of commercially available assays for ST may also have contributed to a lack of testing for STNO157. It is appropriate to quote Acheson and Keusch in this regard: "We cannot let ourselves be complacent in thinking that E. coli O157:H7 is the only Shiga-toxin producing microorganism that can cause problems." At Inova Fairfax Hospital, we began to test for ST in June 1995. The purpose of this paper is to show that there is a much greater incidence of STNO157 than was previously suspected. All stool specimens submitted for bacterial culture were inoculated onto sorbitol MacConkey agar (SMAC) in addition to media for isolation of the usual enteric pathogens. The ST assays were performed only on specimens exhibiting one of the following characteristics: liquid, semiliquid, mucous, or bloody. Additionally, specimens for ST testing were inoculated into MacConkey broth and incubated overnight. Early in this study, STO157 was identified by biochemical tests, latex agglutination, and fluorescence antibody stain, but the ImmunoCard STAT E. coli O157:H7 (Meridian Diagnostics, Cincinnati, Ohio) later replaced the fluorescence antibody staining procedure. The ST assay was performed by using the Premier EHEC kit (Meridian Diagnostics) following the manufacturer's instructions. The test kit contains microwell test strips coated with monoclonal anti-ST I and II. The toxin assay was performed directly from fresh stools and from MacConkey broth cultures. Isolates of STO157 and STNO157 were submitted to the Commonwealth of Virginia, Division of Consolidated Laboratory Services (DCLS) for confirmation of STO157 or STNO157. The STNO157 isolates were referred to CDC by DCLS for serotyping. In the past six and a half years, the following enteric pathogens (with numbers of cases in parentheses) were isolated from our 660-bed community tertiary care hospital: Salmonella (125), STEC (65), Shigella (58), Campylobacter (46), and Vibrio parahemolyticus (4). All of the STEC isolates referred for confirmation were correctly identified as either STO157 or STNO157. Among the 65 patients with STEC, 45 had E. coli serotypes belonging to STO157 whereas 20 (31%) had isolates that were found to be STNO157 with the following distribution (serotypes in parentheses): 4 (O45:H2), 3 (O26:H12), 3 (O103:H2), 3 (O111:NM), 3 (O153:H2), 2 (O88:H25), 1 (O145:NM), 1 (O96:H9). There were four HUS cases (three children and one adult) with no fatality, and all were caused by STO157. The majority of STO157 patients (84%) exhibited bloody stools, while only 45% of STNO157 patients produced bloody stools . The SMAC plates missed four STO157 specimens, but subculturing the broth enabled the isolation of STO157. There was only one specimen from which broth culture failed to yield a positive ST reaction; however, the SMAC showed a few sorbitol-negative STO157 colonies. The ST assay from direct stools and broth cultures yielded 42 positives of 65 samples (65%) and 64 positives of 65 samples (98%), respectively . TABLE 1 : Profiles of patients and Shiga toxin-producing E. coli In Europe and South America, more than 50 serotypes of STNO157 are known to be associated with outbreaks of HUS and hemorrhagic colitis. In a Belgian study involving 10,242 stool samples, the isolation rate of STEC was 1%, of which 38% were ECO157 and 62% were STNO157 . In addition, STEC was isolated from 13 patients among 468 gastroenteritis cases in a German pediatric hospital in which 2 were identified as ECO157 and 11 as STNO157, whereas in France 6 of 69 HUS cases were caused by STNO157. In the United States, the publicity of STNO157 cases has been limited, contributing to an underestimation of its clinical importance. Moreover, the CDC did not underscore its importance until last year at the 101st Annual Meeting of the American Society for Microbiology. In 1994 the U.S. Department of Agriculture's Food Safety and Inspection Service (FSIS) introduced a ground-beef testing program for E. coli O157:H7. Interestingly, in January 2002, the American Society for Microbiology's Public and Scientific Affairs Board Committee on Agriculture and Food Microbiology submitted a recommendation to FSIS to include the testing for STNO157. We do not know the incidence of STNO157 in this country due to a lack of consistent testing. However, our data from northern Virginia show that 31% (20 of 65) of STECs were attributed to STNO157 and 45% (9 of 20) of those patients produced bloody diarrhea. Without the ST assay, the diagnoses of 25 patients (5 with STO157 and 20 with STNO157) would have been missed. TABLE 2 | Annual incidence of Shiga toxin-producing E. coli In our institution, physicians are immediately notified once ST is detected (3 to 30 h) directly from stools or broth cultures. In such cases, patients are often discharged in 1 to 2 days without antibiotic treatment or invasive procedures unless complications such as HUS and dehydration occur. In 1997, a liver transplant patient presented with a severe hemorrhagic colitis in the late evening. The direct stool ST assay performed the next morning due to the bloody appearance of the stool was strongly positive, establishing the etiology in 10 h. This isolate was identified as STNO157 and was later confirmed as O26:H11. Our rapid diagnosis prevented further exploratory procedures and longer hospitalization. Not every laboratory has sufficient resources to incorporate ST assays in addition to the use of SMAC plates. However, an attempt should at least be made to test for ST when bloody stools or suspected cases do not result in the isolation of O157:H7 on SMAC. Since STNO157 serotypes are sorbitol fermenters and have no unique phenotypic markers, the toxin assay is the only method of detection available to clinical laboratories. Another reason for which we are advocating the use of toxin in addition to the use of SMAC is the relatively low sensitivity of SMAC for the detection of STO157. Visible sorbitol-negative colonies do not always prevail over the sorbitol-positive enteric flora. The presence of only a few colonies is often obscured by the overwhelming number of sorbitol fermenters. During this study, 5 of 45 (11%) specimens did not exhibit sorbitol-negative colonies on SMAC, but the isolation of STO157 colonies was possible by subculturing the ST-positive broth. The sensitivity of SMAC plating reported by investigators ranges from 50 to 82.5% . Since most laboratories are using only SMAC as the primary plate medium, one would not recognize its limitation unless another method, such as the toxin assay or PCR, is employed concomitantly. Generally, the use of SMAC is perceived as the "gold standard" for the isolation of STO157, although our results as well as other findings demonstrate otherwise. The existence and prevalence of the sorbitol-positive STEC have not been identified in this country, for no one has actively searched for it. In the Czech Republic , however, there were two cases of HUS caused by sorbitol-positive O157:H7, and this strain is also believed to be widespread in Germany, with cattle being a possible reservoir. Again, the toxin assay is currently the only method of detecting such strains due to the lack of phenotypic markers. In conclusion, our data clearly demonstrate that STNO157 is more prevalent than we were aware of, leading us to believe that the use of SMAC alone for detection may be insufficient. Backmatter: PMID- 12202579 TI - Detection of Rickettsia prowazekii in Body Lice and Their Feces by Using Monoclonal Antibodies AB - In order to identify Rickettsia prowazekii in lice, we developed a panel of 29 representative monoclonal antibodies selected from 187 positive hybridomas made by fusing splenocytes of immunized mice with SP2/0-Ag14 myeloma cells. Immunoblotting revealed that 15 monoclonal antibodies reacted with the lipopolysaccharide-like (LPS-L) antigen and 14 reacted with the epitopes of a 120-kDa protein. Only typhus group rickettsiae reacted with the monoclonal antibodies against LPS-L. R. felis, a recently identified rickettsial species, did not react with these monoclonal antibodies, confirming that it is not antigenically related to the typhus group. Monoclonal antibodies against the 120-kDa protein were highly specific for R. prowazekii. We successfully applied a selected monoclonal antibody against the 120-kDa protein to detect by immunofluorescence assay R. prowazekii in smears from 56 wild and laboratory lice, as well as in 10 samples of louse feces infected or not infected with the organism. We have developed a simple, practical, and specific diagnostic assay for clinical specimens and large-scale epidemiological surveys with a sensitivity of 91%. These monoclonal antibodies could be added to the rickettsial diagnostic panel and be used to differentiate R. prowazekii from other rickettsial species. Keywords: Introduction : Rickettsiaprowazekii is the causative agent of epidemic typhus, a severe reemerging disease . It is transmitted to humans by the body louse, Pediculus hominis corporis , and has the most serious epidemic potential among all rickettsiae. Epidemic typhus frequently occurs in areas where poverty, lack of hygiene, and cold weather favor the proliferation of lice. Its prevalence reflects the socioeconomic level of a society . Sporadic cases of epidemic typhus have reemerged in areas of North Africa , North America , and western South America ; and outbreak cases have been reported in Russia and Burundi , where the biggest outbreak since World War II was observed in 1997. R. prowazekii and R. typhi, the agent of murine typhus, which is transmitted by fleas, are grouped together on the basis of comparisons of phylogenetic references, antigenic components, and other pathogenic characteristics . Epidemic and endemic typhus can be observed as sporadic cases; therefore, it is critical to differentiate between these two types of typhus since their epidemic potentials differ greatly. Compared with the high rate of mortality from typhus, which can reach 10 to 30% for epidemic typhus, murine typhus usually runs a mild course. Moreover, a sporadic case of epidemic typhus could elicit an outbreak in a louse-infested population . Although molecular biological tests have been applied to detect R. prowazekii in lice and blood , so far the diagnosis of typhus is essentially based on serological assays, which include immunofluorescence analysis (, -), the Weil-Felix test , the latex agglutination assay , dot blot assay , the slide immunoperoxidase assay , and Western blotting . Due to intensive serological cross-reactions between these two typhus species , their differentiation is difficult by serology. Recently, Western blotting and/or cross-adsorption studies have been shown to be definitive techniques ; however, the high costs of such studies limit their use. In an effort to circumvent the problem of diagnosis of epidemic typhus economically, we produced species-specific monoclonal antibodies (MAbs) against R. prowazekii to detect this pathogen from infected body lice by immunofluorescence assay. Although MAbs against R. prowazekii have been described in previous studies , to the best of our knowledge, these MAbs have not been used in practice and their use in a diagnostic assay has never been tested. MATERIALS AND METHODS : Preparation of antigens. | The Breinl strain of R. prowazekii and 27 other reference rickettsial strains (listed in Table ) were cultivated in confluent monolayers of L929 cells (ATCC CCL 1 NCTC clone 929), as described previously (-). R. felis was grown in XTC2 cells . When the cells were heavily infected and estimated to be at a concentration of 104 PFU/ml, as determined by Gimenez staining , they were harvested and stored at -70C. These unpurified antigens were used in the indirect immunofluorescence assay to screen MAbs. For immunization, they were sonicated, centrifuged at 100 x g (CR4, 12 refrigerated Bank-Top centrifuge; Jouan, Winchester, Va.) for 10 min, and concentrated 20 times to an estimated density of 2 x 105 PFU/ml. For sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE), the antigens of R. prowazekii, R. typhi, or R. felis were purified by centrifugation in Renografin density gradients as described previously . The final pellets were suspended in distilled water, adjusted to a concentration of 1 mg/ml , and stored at -80C. Six bacterial human isolates from different phyla (Pseudomonas aeruginosa, Haemophilus influenzae, Streptococcus pneumoniae, Escherichia coli, Klebsiella oxytoca, Staphylococcus aureus) and three other louse-associated bacteria (Bartonella quintana, Borrelia recurrentis, Acinetobacter baumannii) were used to determine the specificities of the MAbs. Production of MAbs. | Six-week-old female BALB/c mice were inoculated intraperitoneally with 106R. prowazekii organisms in 4% formalin (Sigma Chemical Co., St. Louis, Mo). After three injections at 7-day intervals, the mice were given two boosters 7 and 14 days later by intravenous injection of 4 x 103 organisms in 0.1 ml of phosphate-buffered saline (PBS) into the tail vein. The splenocytes from antibody-positive mice were fused with SP2/0-Ag14 myeloma cells by using 50% (wt/vol) polyethylene glycol (molecular weight, 1,300 to 1,600; Sigma Chemical Co.) by the procedure described by Xu et al. . The fused cells were then grown in hybridoma selective medium (Gibco BRL, Paisley, Scotland) containing 20% fetal bovine serum (Gibco BRL) and hypoxanthine-aminopterin-thymidine selective medium (Gibco BRL) for 2 weeks and successively in hypoxanthine-thymidine medium (Sigma Chemical Co.) for 5 days. The supernatants from viable hybridoma clones were screened for antibodies against R. prowazekii by immunofluorescence assay. Positive hybridoma cells were spread and subcloned two to three times by limiting dilution. The immunoglobulin classes and subclasses of the MAbs were determined with an ImmunoType mouse MAb isotyping kit (Sigma Chemical Co.). Immunofluorescence assay. | An indirect immunofluorescence assay was used to screen hybridoma clones and determine the specificities of the MAbs. Unpurified antigens of R. prowazekii and other reference bacteria were deposited on slides with a pen nib. The slides were air dried and then fixed in acetone for 20 min at room temperature. The assay was modified from a previously described procedure . Briefly, the wells were overlaid with 30 mul of supernatants from hybridoma clones, and then the plates were incubated in a moist chamber at 37C for 30 min before they were washed three times in PBS for 3 min each time. After the slides were dried, bound antibody was detected with dichlorotriazinyl amino fluorescein-conjugated goat anti-mouse immunoglobulin G (IgG) plus IgM (Jackson ImmunoResearch Laboratories, Inc., West Grove, Pa.) diluted 1:200 in PBS containing 3% nonfat dry milk and 0.2% Evans blue (BioMerieux, Marcy 1'Etoile, France). The slides were washed as described above and mounted with Fluoprep (BioMerieux) and were then examined under a Zeiss epifluorescence microscope at x400 magnification. Sera of immunized and healthy mice were used as positive and negative serum controls, respectively, in each assay. SDS-PAGE and Western immunoblotting. | The whole antigen of R. prowazekii was dissolved in Laemmli buffer (0.625 M Tris [pH 8.0], 2% [wt/vol] SDS, 5% [vol/vol] 2-mercaptoethanol, 10% [vol/vol] glycerol, 0.002% [wt/vol] bromophenol blue) for 2 h at room temperature and used as native antigen or was boiled for 10 min and used as heated antigen, or the proteins were digested with proteinase K (2 mg/ml) at 37C for 1 h. The antigens of R. typhi and R. felis were only used as native antigen. Electrophoresis was performed at 16 mA for 2 h on a 10% polyacrylamide separating gel with 5% polyacrylamide stacking gel in a Mini Protein II cell apparatus (Bio-Rad, Richmond, Calif.). A mixture of high-range molecular mass standards (Bio-Rad) was used as a marker to estimate the molecular weight of the separated antigen. The resolved polypeptides were transferred onto a nitrocellulose membrane (pore size, 0.45 mum; Trans-Blot Transfer Medium; Bio-Rad) by electrophoresis at 100 V for 1 h in an electrophoretic transfer cell (Mini Trans-Blot; Bio-Rad) with transfer buffer (2.5 mM Tris base, 192 mM glycine, 20% methanol). Then the nitrocellulose sheet was blocked at 4C overnight with Tampon Saturation buffer (0.121% [wt/vol] Tris base, 150 mM NaCl, 0.05% [vol/vol] Tween 20) containing 5% nonfat powdered milk. After three 10-min washes in Tampon Saturation buffer, the membrane was cut into strips. The supernatants of different hybridoma clones were diluted 1:6 in Tampon Saturation buffer with 0.5% nonfat dry milk and were then applied to the strips, and the strips were incubated at 4C overnight on a rocker. The strips were washed as described above and were then incubated for 2 h at room temperature with peroxidase-conjugated F(ab')2 fragment goat anti-mouse IgG (heavy and light chains; AffiniPure; Jackson ImmunoResearch) diluted 1:400 in Tampon Saturation buffer containing 3% nonfat dry milk. After three washes with Tampon Saturation buffer, the bound conjugate was detected by incubation with a solution which contained 0.015% 4-chloro-1-naphthol (Sigma Chemical Co.) in 16.7% methanol in Tampon Saturation buffer containing 0.015% hydrogen peroxide. The sera of the immunized and naive mice were collected, diluted 1:100, and used as the positive and negative controls, respectively, in each test. Detection of R. prowazekii. | For clinical use we selected an MAb specific for R. prowazekii (MAb P11A12). Twenty-two lice from Rwanda and Burundi sampled during typhus outbreaks were tested by PCR as reported previously . We also used 17 lice experimentally infected with R. prowazekii and 17 noninfected lice. These lice were also tested by PCR, as reported previously . The gut lumens of the lice were dissected as reported previously , and their intestines were crushed and smeared on 18-well microscope slides under an enlarge-scope (Stermi 2000; Zeiss) at x2.5 magnification. We also tested feces from two lice infected in the laboratory and eight noninfected lice. Feces were directly smeared onto the slides. The immunofluorescence test was performed by using a blind method. After fixation in methanol for 15 min at room temperature, the rickettsiae were detected by immmunofluorescence assay by using supernatants of MAb P11A12 (an R. prowazekii-specific MAb) diluted (1/100 and 1/1,000) with distilled water. Following performance of the immunofluorescence assay as described above, the intracellular position of visible R. prowazekii in lice was determined with a laser scanning confocal microscope as described previously to determine infection . TABLE 1 | Rickettsial strains studied RESULTS : Production of MAbs. | Initially, we obtained 254 viable hybridoma clones 14 to 18 days after the fusion. Through screening by immunofluorescence assay, 187 clones were found to secrete antibodies against R. prowazekii, and 40 were selected for subcloning by limiting dilution two to three times. Finally, 29 MAbs were selected. SDS-PAGE and Western blotting. | By SDS-PAGE with Coomassie brilliant blue staining of the polyacrylamide gels, the native protein profiles of R. prowazekii showed three major groups of polypeptide bands: bands at 120 and 60 kDa and low-molecular-mass bands at about 17 to 40 kDa, as reported previously . By Western immunoblotting the polyclonal antiserum of mice against R. prowazekii showed several major reactive bands that comigrated to positions that corresponded to the major polypeptides bands seen by SDS-PAGE. Of the 29 hybridoma clones that produced MAbs, 14 recognized a 120-kDa protein. They belonged to the IgG2b and IgG3 subclasses. This reactivity was completely destroyed after antigen treatment by heating and proteinase K digestion . Fifteen other MAbs directed against a lipopolysaccharide (LPS)-like (LPS-L) antigen produced multiple bands (17 to 40 kDa), and the antigen resisted destruction by heating and proteinase K digestion . The MAbs belonged to the IgG3 and IgG1 subclasses. Specificities of MAbs. | Twenty-nine MAbs against R. prowazekii were tested with 27 other strains of rickettsiae. Among the 14 MAbs against the 120-kDa protein, 8 reacted only with R. prowazekii and 6 had weak cross-reactivity with R. typhi and/or R. canadensis. Fifteen MAbs against the LPS-L antigen strongly reacted with R. typhi and R. canadensis but not with R. felis and were classified as group specific. No MAb was found to have cross-reactivity with any strain of the spotted fever group rickettsiae or the other nine nonrickettsial bacteria including the three louse-associated bacteria. We selected MAb P11A12, an anti-120-kDa protein MAb that reacts only with R. prowazekii, to test the lice. PCR analysis indicated that 4 of 22 wild lice were infected, as were all lice infected in the laboratory . Among the 21 infected (naturally or experimentally) lice, the MAb detected infection in 19 lice by detection of several foci of fluorescent bacteria . The feces from the two infected lice were found to be positive. Both dilutions of the MAb (1/100 and 1/1,000) gave the same results. The sensitivity of this assay was estimated to be 91%. A weak positive reaction was found with three other lice, including one louse that was PCR positive. None of the noninfected feces were positive. The negative predictive value of our test was 100%. FIG. 1. | Western blotting profiles of MAbs against R. Western blotting profiles of MAbs against R. prowazekii. Lane 1, molecular size markers (numbers on the left are in kilodaltons); lane 2, Coomassie brilliant blue-stained, native R. prowazekii antigen (SDS-PAGE); lanes 3, 5, and 7, MAb P11A12; lanes 4, 6, and 8, MAb P9G1; lane 9, polyclonal antisera against native antigen (lanes 3, 4, and 9), heated antigen (lanes 5 and 6), and proteinase K-digested antigen (lanes 7 and 8). FIG. 2. | Fluorescent antibody-stained isolate of R. prowazekii Fluorescent antibody-stained isolate of R. prowazekii. (a) Feces from an uninfected louse; (b) feces of an experimentally infected louse; (c) cell smear prepared from intestines of an experimentally infected louse. DISCUSSION : In the present work we first applied species-specific MAbs against R. prowazekii to detect this pathogen in infected lice by immunofluorescence assay. MAbs against R. prowazekii have been described in previous studies that have investigated common cross-reactivity with R. typhi and that have studied the surface antigen . MAbs against the LPS of typhus group rickettsiae have also been used to establish a timely diagnosis of typhus group rickettsiosis by immunohistochemical examination . An investigation of the antigenic specificities of MAbs against R. prowazekii in a large panel of rickettsial species including R. felis has never been reported. In the present study, more than 90 MAbs were produced. Through two to three limiting dilutions, expansions, and determinations, 29 representative MAbs were finally selected. They were shown to be reactive with the epitopes on an LPS-L antigen that may be LPS and a high-molecular-mass protein of 120 kDa that may be rOmpB, an outer membrane protein. Both were found to be predominant immunogens of R. prowazekii by immunoblotting with mouse polyclonal antiserum . Further analysis of cross-reactivity with 27 different rickettsial species showed that MAbs against LPS-L of R. prowazekii reacted by immunofluorescence assay with each member of the typhus group rickettsiae, R. typhi and R. canadensis, but not with members of the spotted fever group rickettsiae. It indicated that this antigen, shared by every member of the typhus group rickettsiae, was group specific and may be the LPS. R. felis , the agent of flea-borne spotted fever, was first observed in cat fleas (Ctenocephalides felis) by electron microscopy and was originally included in the typhus group rickettsiae . More recently, a wealth of genetic studies have discriminated R. felis from the typhus group and have placed it into the spotted fever group . The MAbs against LPS-L obtained in the present work did not react with R. felis, confirming this classification. By Western blotting, 14 MAbs directed against the 120-kDa protein, which may be on the predominant immunogenic surface protein antigen, rOmpB, were considered species-specific MAbs, none of which was found to recognize the heat-denatured form of this antigen. These MAbs were found not to react with other rickettsial species; however, six of the MAbs showed weak cross-reactivity with R. typhi. Finally, one species-specific MAb against R. prowazekii, directed against the 120-kDa protein, detected the organism in samples of both naturally and experimentally infected lice or their feces by immunofluorescence assay. Other louse-associated bacteria were not detected by these antibodies, which further demonstrated their specificities. This report describes a specific, convenient, and reliable diagnostic assay with a sensitivity of 91%. It especially appears to be of practical value when it is required to identify R. prowazekii in lice in a given area, when typhus cases are suspected, or for epidemiological surveys of lice. Backmatter: PMID- 12202545 TI - Cavitary Pneumonia in an AIDS Patient Caused by an Unusual Bordetella bronchiseptica Variant Producing Reduced Amounts of Pertactin and Other Major Antigens AB - Although Bordetella bronchiseptica can infect and colonize immunocompromised humans, its role as a primary pathogen in pneumonia and other respiratory processes affecting those patients remains controversial. A case of cavitary pneumonia caused by B. bronchiseptica in an AIDS patient is presented, and the basis of the seemingly enhanced pathogenic potential of this isolate (designated 814) is investigated. B. bronchiseptica was the only microorganism recovered from sputum, bronchoalveolar lavage fluid, and samples taken through the protected brush catheter. Unlike previous work reporting the involvement of B. bronchiseptica in cases of pneumonia, antibiotic treatment selected on the basis of in vitro antibacterial activity resulted in clearance of the infection and resolution of the pulmonary infiltrate. Although isolate 814 produced reduced amounts of several major antigens including at least one Bvg-activated factor (pertactin), the molecular basis of this deficiency was found to be BvgAS independent since the defect persisted after the bvgAS locus of isolate 814 was replaced with a wild-type bvgAS allele. Despite its prominent phenotype, isolate 814 displayed only a modest yet a significant deficiency in its ability to colonize the respiratory tracts of immunocompetent rats at an early time point. Interestingly, the antibody response elicited by isolate 814 in these animals was almost undetectable. We propose that isolate 814 may be more virulent in immunocompromised patients due, at least in part, to its innate ability to produce low amounts of immunogenic factors which may be required at only normal levels for the interaction of this pathogen with its immunocompetent natural hosts. Keywords: Introduction : Bordetella bronchiseptica, a small pleomorphic gram-negative coccobacillus, is a common upper respiratory tract pathogen in many domestic and wild animals and causes respiratory diseases in a broad variety of mammal species such as cats, dogs, horses, rabbits, and swine, among others . B. bronchiseptica synthesizes a wide array of virulence factors including adhesins like filamentous hemagglutinin (FHA), fimbriae, and pertactin and toxins like the bifunctional adenylate cyclase toxin/hemolysin and tracheal cytotoxin . In B. bronchiseptica, like in other members of the Bordetella genus, a two-component signal transduction system known as BvgAS controls the expression of all protein virulence factors that have been identified. When B. bronchiseptica grows at 37C in the absence of nicotinic acid or magnesium sulfate (modulating agents), it expresses the so-called Bvg-positive (Bvg+) phase, characterized by the expression of virulence factors and the repression of the flagellar apparatus, among other Bvg-repressed phenotypes. B. bronchiseptica cells grown under Bvg+-phase conditions are fully virulent and nonmotile . Conversely, incubation below 30C or addition of modulating agents to the culture medium inactivates BvgAS, thus simultaneously precluding protein virulence factor synthesis and derepressing the flagellar apparatus and other Bvg-repressed phenotypes. B. bronchiseptica cells grown under these conditions (Bvg--phase conditions) are avirulent and motile . In contrast to this reversible transition, termed phenotypic modulation, mutations leading to inactivation of BvgAS result in an irreversible transition to the Bvg- phase. Thus, this genetic event, which occurs at a low frequency (10-6), gives rise to avirulent Bvg--phase-locked mutants. The existence of a tight correlation between phenotypic (Bvg) phase and virulence has received further experimental confirmation with the demonstration that mutants displaying phenotypically intermediate traits (locked in the so-called Bvgi phase) also exhibit an intermediate ability to colonize the respiratory tracts of suitable hosts . Reports of respiratory disease caused by B. bronchiseptica in humans remain scarce and very rarely involve immunocompetent patients and more commonly involve children and immunocompromised patients including, in many instances, AIDS patients. For the last epidemiological group, isolation of B. bronchiseptica from the respiratory tract or from the blood of human immunodeficiency virus (HIV)-infected patients with respiratory diseases has been increasingly reported. This circumstance has prompted some investigators to propose the inclusion of B. bronchiseptica in the list of opportunistic pathogens causing diseases associated with exposure of HIV-infected patients to animals . However, thus far the pathogenic role that B. bronchiseptica plays in those processes has been difficult to ascertain since, in addition to B. bronchiseptica, other respiratory pathogens are usually coisolated from the respiratory tracts of AIDS patients with pulmonary infections. Finally, to assess the involvement of B. bronchiseptica in human respiratory diseases, it seems necessary to determine whether the presence of specific phenotypic traits in certain B. bronchiseptica strains correlates with their pathogenic potential. To our knowledge, this kind of information is lacking. In this article we report on two cases of pulmonary disease in AIDS patients differing in severity and clinical presentation and make an evaluation of the pathogenic role of B. bronchiseptica in the clinical processes. CASE REPORTS : Patient 1. | A 34-year-old woman was admitted to the Hospital Universitario Virgen del Rocio reporting a 2-week history of cough with purulent expectoration, fever, and chills. In addition, she reported having lost 50% of her normal weight in the previous 2 months. Four years earlier she had maintained frequent heterosexual intercourse with an HIV-infected individual. At the time of her admission, she owned a healthy dog. On examination oropharyngeal thrush was observed and rales in the left hemithorax were heard. The total leukocyte count was 13,100/mul, the hemoglobin concentration was 9.2 g/dl, and the platelet count was 256,000/mul. Serum electrolyte and creatinine levels and liver enzyme levels were in the normal ranges. The arterial partial oxygen pressure was 67 mm Hg, the arterial partial carbon dioxide pressure was 39 mm Hg, and the pH was 7.4. Detection of anti-HIV immunoglobulin G antibodies first by enzyme immunoassay and then by Western immunoblotting confirmed HIV infection. The CD4 lymphocyte count was 25/mul, and HIV RNA levels were 1.5 x 106 copies/ml. A chest X ray showed an alveolar infiltrate with areas of cavitation in the left upper lung lobe . Microscopic examination of sputum smears stained with either acid-fast or Gram stain did not reveal the presence of any microorganism. Because of the suspicion of tuberculosis, isoniazid, rifampin, and pyrazinamide were administered. Sputum cultures exclusively yielded growth of B. bronchiseptica in pure culture. This B. bronchiseptica clinical isolate was designated 814 and was sensitive to amikacin (MIC, 16 mug/ml), amoxicillin-clavulanic acid (MIC, 8/4 mug/ml), ciprofloxacin (MIC, <1 mug/ml), ofloxacin (MIC, <2 mug/ml), gentamicin (MIC, 4 mug/ml), imipenem (MIC, <4 mug/ml), tetracycline (MIC, <4 mug/ml), ticarcillin-clavulanic acid (MIC, <16 mug/ml), and tobramycin (MIC, 4 mug/ml). On the 8th day the patient remained febrile and bronchoscopy was performed. Cultures of the samples obtained through the protected brush catheter and bronchoalveolar lavage exclusively yielded growth of >104 CFU of B. bronchiseptica with the same antibiotic sensitivity as the organism recovered from sputum. Treatment was switched to ofloxacin (400 mg oral twice a day) on the 10th day. On the 20th day the patient was asymptomatic, control cultures of sputum were negative, and chest X ray showed progressive resolution of the infiltrate. On the 28th day, the patient was discharged with antiretroviral therapy consisting of stavudine, lamivudine, and indinavir. Antimicrobial treatment was maintained for a total of 4 weeks. Three months after admission the patient remained asymptomatic, the CD4 cell counts were 250/mul, the HIV RNA levels were 70,000 copies/ml, and a control chest X ray appeared normal. Patient 2. | A 26-year-old man who had been diagnosed with AIDS and who had had Pneumocystis carinii pneumonia 2 months earlier was admitted to the Hospital Universitario Virgen del Rocio with a 5-day history of cough, high fever, weight loss, and left costal pain. He was a heroin addict and did not report any contact with household, farm, or wild animals. The physical examination disclosed fever (38.5C), cachexia, tachypnea, oral thrush, enlargement of the cervical lymph nodes, and middle hepatomegaly. The rest of the physical examination was unremarkable. A chest X ray (not shown) revealed a middle diffuse interstitial infiltrate. The total leukocyte count was 10,500/mul, the hemoglobin concentration was 10.6 g/dl, and the platelet count was 364,000/mul. Serum electrolyte and creatinine levels and liver enzyme levels were in the normal ranges. The arterial partial oxygen pressure was 67 mm Hg, the arterial partial carbon dioxide pressure was 39 mm Hg, and the pH was 7.4. The CD4 lymphocyte count was 97/mul, and the viral load was 338,623 copies/ml. Because of the initial suspicion of P. carinii pneumonia, trimethoprim-sulfamethoxazole treatment was started and the fever disappeared. Later, in a computed tomography scan of the thorax, two noncavitating nodules <5 cm in diameter were observed in both the left lower and right upper lung lobes. Rhodococcus equi grew in the blood cultures, and B.bronchiseptica was isolated from the sputum culture. This B. bronchiseptica isolate was designated 222. Sputum cultures were negative for Legionella spp., mycobacteria, parasites, and fungi. A bronchoscopy was performed. Only B. bronchiseptica (>104 CFU) grew in the samples obtained through the protected brush catheter and bronchoalveolar lavage. The second B. bronchiseptica isolate recovered from patient 2 was designated 629 and was sensitive to ciprofloxacin (MIC, <1 mug/ml), ofloxacin (MIC, <2 mug/ml), amikacin (MIC, 16 mug/ml), amoxicillin-clavulanic acid (MIC, 8/4 mug/ml), gentamicin (MIC, 4 mug/ml), imipenem (MIC, <4 mug/ml), tetracycline (MIC, <4 mug/ml), ticarcillin-clavulanic acid (MIC, <16 mug/ml), and tobramycin (MIC, 4 mug/ml) and was resistant to co-trimoxazole (MIC, 2/38 mug/ml). Trimethoprim-sulfamethoxazole was discontinued; and vancomycin, rifampin, and ciprofloxacin were administered for 15 days and the patient was discharged. One week before discharge the patient quit the antimicrobial treatment. Two months later he was admitted to the hospital with fever, cough, and chills. Chest X ray and computed tomography scan images (not shown) showed multiple bilateral nodules and an infiltrate in the left lower lung lobe. R. equi was the only microorganism isolated from blood cultures and samples obtained through bronchoscopy. FIG. 1. | Chest X ray demonstrates cavitary pneumonia in patient 1. Chest X ray demonstrates cavitary pneumonia in patient 1. MATERIALS AND METHODS : Identification of clinical isolates. | Isolation of B. bronchiseptica from bronchoalveolar lavage specimens was performed as follows. The bronchoalveolar lavage fluid was diluted 1:10 and 1:100 prior to its inoculation on enriched blood and chocolate agar plates. Only bronchoalveolar lavage fluid specimens rendering colony counts of at least 10,000 CFU/ml were considered positive and were further processed. Clinical isolates were identified as B. bronchiseptica by using combined sensitivity-identification panels (NEG Breakpoint Combo Type 15; Dade International Inc., West Sacramento, Calif.) and miniature bacterial identification strips (API NE; bio Merieux, Marcy l'Etoile, France). Microscopic examination of Gram-stained smears, extensive phenotypic characterization of the isolates (see Table and Fig. and ), and DNA sequencing of a portion of the bvgS gene of isolate 814 supported the identification of the isolates as B. bronchiseptica. Bacterial strains and growth conditions. | B. bronchiseptica strains RB50 and RB54 were a kind gift of Peggy Cotter (Department of Molecular, Cellular and Developmental Biology, University of California, Santa Barbara). These are well-characterized strains that have been used in several studies on the interaction of B. bronchiseptica with its natural hosts . RB50 is a wild-type strain isolated from the nares of a naturally infected rabbit. RB54 is a Bvg--phase-locked derivative of RB50 . As positive or negative controls for some biochemical determinations, human and ovine isolates of B. parapertussis (strains 12822 and Fr107, respectively ) were used. For the determination of hemolytic activity and colony morphology, all the strains were grown on Bordet-Gengou (BG) agar plates supplemented with 7.5% defibrinated sheep blood. The plates were incubated at 37C for 48 h. Motility assays were performed by stabbing one medium-size colony in Stainer-Scholte (SS) medium supplemented with 0.35% agar. The diameter of the motility halo was measured after 14 h of incubation at 37C. When it was necessary to grow Bordetella under Bvg--phase (modulating) conditions, nicotinic acid and MgSO4 were added to the medium at final concentrations of 5 and 20 mM, respectively. Tyrosine agar plates for determination of the browning activities of the strains were prepared as described by Porter et al. . Whole-cell lysates for sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) analysis were prepared from cells grown overnight in Luria-Bertani (LB) broth at 37C with constant shaking. To ensure the absence of Bvg--phase mutants in the liquid cultures, an aliquot of each culture was plated on BG blood agar, and after 48 h of incubation at 37C, the colonies were visually inspected for hemolytic activity. SDS-PAGE, Western immunoblotting, and densitometry. | SDS-PAGE was performed as described previously . Briefly, bacterial cells suspended in phosphate-buffered saline (PBS) were lysed by the addition of 2x SDS-PAGE sample buffer . Bacterial lysates were boiled for 5 min, stacked in an SDS-4% polyacrylamide gel, and separated in an SDS-10% polyacrylamide gel. Proteins were stained with Coomassie brilliant blue or transferred to polyvinylidene difluoride membranes for immunoblotting. Proteins were probed with either a 1:16,000 dilution of serum from a rat infected with B. bronchiseptica RB50 (RA1 antiserum), a 1:2,500 dilution of antipertactin monoclonal antibody (MAb) BPE3 , or a 1:2,000 dilution of serum from rats infected with either one of the clinical isolates. Antigen-antibody complexes were detected with a 1:5,000 dilution of horseradish peroxidase-conjugated antibody of the appropriate specificity (Amersham International, Little Chalfont, United Kingdom). The immunocomplexes were detected by using an enhanced chemiluminescence assay (Amersham) according to the instructions of the manufacturer. Densitometric analysis was carried out on digitized images obtained with a high-resolution scanner (JX 325; Sharp Corporation, Osaka, Japan) equipped with Amersham Pharmacia ImageMaster software. Experimental animals. | Female Wistar rats were obtained from Harlan Spain (Harlan Interfauna Iberica S.A., Barcelona, Spain) at 3 weeks of age. To confirm that they were Bordetella-free, two animals of each lot were euthanatized and samples of nasal, tracheal, and lung tissue were removed and cultured on BG blood agar. Inocula were prepared from cells grown overnight in LB broth and consisted of 1,000 CFU administered intranasally in 10 mul of PBS while the animals were slightly anesthetized by halothane inhalation. At the designated time points postinoculation, the rats were anesthetized by halothane inhalation, a sample of blood was obtained by cardiac puncture, and then the animals were killed by cervical dislocation. The chest cavity was opened, and 1 cm of trachea and the right lung lobes were removed and placed in PBS for homogenization. The nose was dissected, and the entire nasal septum and the adjacent tissues were removed and placed in PBS. Tissues were homogenized with tissue grinders, and aliquots of the suspensions were plated on BG blood agar for determination of viable counts. Experimental data were analyzed by a two-way analysis of variance test with SPSS software. Prior to this analysis, the interaction between the parameters under investigation (animals of each group, strain type, site of the respiratory tract) was confirmed to be nonsignificant. Animal protocols were approved by the University of Navarra Animal Research Committee (protocol number 039/00). Hemagglutination and ELISAs. | Hemagglutination activity was quantitated in the supernatants of cultures grown in SS medium for 24 h at 37C with constant shaking. The hemagglutination assay has already been described . Enzyme-linked immunosorbent assays (ELISAs) were performed as described by Cotter and Miller with slight modifications. Briefly, an overnight culture of each B. bronchiseptica strain was adjusted to an optical density (600 nm) of 1.0 with LB broth, and the suspensions were diluted 1:10 in coating buffer (carbonate-bicarbonate buffer [pH 9.7]) and sonicated. A total of 100 mul of the suspension was added to each well, and the plates were incubated at 37C for 2 h in a humidified chamber and then overnight at 4C. Nonspecific binding to the wells was prevented by adding 150 mul of 3% bovine serum albumin to each well and incubating the plate for 60 min at 37C. As the primary antibody, a 1:6,000 dilution of RA1 antiserum was added to the first well, twofold dilutions were performed across eight wells, and the plates were incubated at 37C for 60 min. As the secondary antibody, horseradish peroxidase-conjugated anti-rat immunoglobulin G antibody (Amersham) was used at a final concentration of 1:5,000. The A405 was read after a 30-min incubation with the substrate 2,2'-azinobis(3'-ethylbenzthiazoline sulfonic acid), purchased from Sigma. PCR, cloning, and sequencing. | Two oligonucleotides designed to anneal at positions 2556 (GCTGGAATTCATGCGCGTGCTCA) and 3145 (CGATCTTCGCAATGTCCAG), respectively, of the published B. bronchiseptica bvgAS sequence were used as primers for the PCR. PCR was performed as follows: 3 mM MgCl2, 5% dimethyl sulfoxide, 1 U of Taq polymerase (Perkin-Elmer, Wellesley, Mass.), the four deoxynucleoside triphosphates each at a concentration of 250 muM, and 20 pmol of each primer were combined and brought to a total volume of 25 mul. A small portion of a colony was resuspended in the solution as a source of template DNA. A Perkin-Elmer GeneAmp 2400 thermal cycler was used for the reactions. The cycling parameters were as follows: incubation at 95C for 5 min, followed by 30 cycles of 95C for 1 min, 55C for 1 min, and 72C for 1 min and a final incubation at 72C for 5 min. PCR products were cloned into the pCR2.1 vector by using the TA cloning kit (Invitrogen, Carlsbad, Calif.) according to the instructions of the manufacturer. Plasmids containing the cloned PCR products were submitted for sequencing to Sistemas Genomicos, S.L. (Paterna, Spain). Construction of mutant strains by allelic exchange. | The construction and characteristics of plasmid PGMT8 have already been described . The entire bvgAS locus of B. bronchiseptica 814 and adjacent sequences were deleted by allelic exchange with plasmid pGMT8 to generate strain B. bronchiseptica 814-DAS. B. bronchiseptica 814-DAS shares with RB55 the genetic configuration at that chromosomal location and was constructed by the same method . PCR analysis with primers designed to anneal at both sides of the deletion junction resulted in the amplification of a fragment with the expected size, thus confirming that allelic exchange had occurred as intended. Plasmid pGMT40 carries a chimeric bvgAS locus consisting of the whole bvgAS locus from B. bronchiseptica RB50 except for the portion encoding part of the periplasmic BvgS domain. This fragment was swapped with the equivalent segment from B. pertussis, and the resultant allele was shown to confer resistance to phenotypic modulation to B. bronchiseptica . Plasmid pGMT40 was used to return a wild-type bvgAS locus to strain B. bronchiseptica 814-DAS to generate B. bronchiseptica 814-P5. This strain is equivalent to RB52-P5 and was constructed by the same method . Construction of strains carrying lacZ fusions and quantitation of beta-galactosidase activity. | Plasmid pGMT18 carries an internal fragment of the pertactin gene (prn) cloned upstream of a promoterless lacZ gene . Its integration into the chromosome of Bordetella places the transcription of lacZ under the control of the prn promoter. B. bronchiseptica strains carrying transcriptional fusions to prn were constructed by using pGMT18 and by following a method that has been already described. Beta-galactosidase activity was measured as described previously in cells grown in SS medium to the mid-log phase. Statistical significance was determined by using the Mann-Whitney test. To detect the occurrence of spontaneous bvgAS mutants, aliquots from each broth with growth were plated onto BG blood agar. FIG. 3. | Analysis of polypeptide and antigen expression in whole cells of the B. bronchiseptica Analysis of polypeptide and antigen expression in whole cells of the B. bronchiseptica clinical isolates. (A and B) Whole-cell lysates of the indicated strains were run in duplicate SDS-polyacrylamide gels and either stained with Coomassie brilliant blue (A) or transferred to polyvinylidene difluoride membranes for immunoblotting and probed with serum from a rat infected with B. bronchiseptica RB50 (B). The positions of the molecular weight markers (in thousands) are shown at the left. (C) Quantitation of total antigen expression by a whole-cell antigen-based ELISA. The same serum sample used for panel B was used as the primary antibody. FIG. 4. | Expression of pertactin in whole cells of the B. bronchiseptica Expression of pertactin in whole cells of the B. bronchiseptica clinical isolates. Whole-cell lysates of the indicated strains were run in duplicate SDS-polyacrylamide gels and either stained with Coomassie brilliant blue (A) or transferred to polyvinylidene difluoride membranes for immunoblotting and probed with antipertactin MAb BPE3 (B). Band sizes were determined by densitometry and are expressed in units of optical density millimeter-2 with respect to a calibrated standard. The positions of the molecular weight markers (in thousands) are shown at the left. TABLE 1 | Relevant phenotypes of the strains used in this study RESULTS : Rationale for the study and phenotypic analysis of the clinical isolates. | Compared to patient 2, patient 1's respiratory illness was more severe and could be attributed solely to infection with B. bronchiseptica isolate 814, the only microorganism isolated from the respiratory tract of patient 1. Since both the AIDS clinical stage and the immunological status of patient 2 did not significantly differ from those of patient 1, we hypothesized that B. bronchiseptica 814 might be more pathogenic than other B. bronchiseptica isolates including those recovered from patient 2 (isolates 222 and 629). To address this hypothesis, we performed a comparative study of selected phenotypic characteristics of these three isolates. Specifically, we focused on those phenotypes that are known to affect the pathogenic potential of B. bronchiseptica. Bearing in mind that the phenotypic phase of B. bronchiseptica closely correlates with its virulence, we first examined whether all the clinical isolates were phenotypically wild-type (virulent) or avirulent (bvg) variants. Wild-type strains, like B. bronchiseptica RB50, grown on BG blood agar without modulating agents give rise to small, domed, and hemolytic colonies, while those of Bvgi-phase-locked mutants are slightly larger, flatter, and less hemolytic. When modulating agents are added to the medium or when Bvg is not functional (i.e., Bvg--phase-locked mutants like B. bronchiseptica RB54) colonies appear large, flat, and nonhemolytic. As shown in Table , all B. bronchiseptica clinical isolates were phenotypically wild type with respect to colony morphology and hemolysis. The ability to regulate another prominent phenotype, motility, has been shown to be of major importance for the pathogenicity of B. bronchiseptica. Thus, mutants that cannot repress the flagellar apparatus in the Bvg+ phase are unable to colonize the lower respiratory tracts of rats . To determine whether any of the clinical isolates had an altered pattern of regulation of motility, all of the isolates were grown in SS medium with or without modulating agents. As shown in Table , all the clinical isolates exhibited a pattern of regulation of motility expected of a wild-type B. bronchiseptica strain. Experimental respiratory infection of rats. | The ability of B. bronchiseptica to colonize the respiratory tracts of susceptible laboratory animals such as mice, rats, and rabbits has been successfully exploited to compare the pathogenic potentials of B. bronchiseptica strains in the context of a natural host-pathogen interaction. When delivered intranasally in a low-volume (5- to 10-mul) droplet of PBS, wild-type B. bronchiseptica strains like RB50 efficiently colonize the rat and are recovered in high numbers from the nose and the trachea and less consistently from the lungs of the animal by day 14. A strong anti-Bordetella antibody response is readily detectable by day 10. In sharp contrast, avirulent mutants like RB54 are unable to colonize any site of the respiratory tract even at the earliest time points tested, while other mutants carrying less impairing mutations display an intermediate level of virulence and are characteristically unable to colonize the lower respiratory tract . We chose to use the rat model of respiratory infection to detect potential differences in virulence among the clinical isolates. As expected, at day 7 postinoculation RB54 was not recovered from the respiratory tract of any rat. In contrast, isolates 222 and 814 colonized the noses of all animals at a very similar level and were recovered in similar numbers from the lungs of two of five animals . However, compared to isolate 222, isolate 814 showed a reduced ability to colonize the tracheas of the rats at day 7 postinoculation. This defect was transient since at day 14 postinoculation this isolate was capable of establishing tracheal colonization at a level indistinguishable from those of the other isolates (data not shown). Animals mock infected only with PBS remained Bordetella-free throughout the experiment, and isolate 629 was recovered from all sites of the respiratory tract in numbers similar to those for isolate 222 and strain RB50 (data not shown). We did not detect any macroscopic lesions in the respiratory tracts of the animals or any outward sign of respiratory disease at any time during the course of the infection. Quantification of FHA production by the clinical isolates. | A derivative of strain RB50 carrying an in-frame deletion in the gene encoding FHA has been shown to be unable to colonize the lower respiratory tract of rats . Therefore, the reduced ability of isolate 814 to colonize the trachea of rats could be the result of a mutation leading to the deficient production or secretion of FHA. To test this possibility, the amount of FHA in the supernatants of overnight cultures of the isolates was measured. As shown in Table , all clinical isolates produced FHA at levels that were identical to those produced by RB50. Characterization of clinical isolates by SDS-PAGE and Western immunoblotting. | To determine if the defect in colonization exhibited by isolate 814 could result from the loss of potentially antigenic virulence determinants, we compared the antigenic profiles of all the isolates by Western immunoblotting using serum from a rat infected with B. bronchiseptica RB50 (RA1 antiserum). As shown in Fig. , while the antigenic patterns of isolates 222 and 629 were nearly identical to that of RB50, isolate 814 appeared to produce background levels of a large number of antigens. This result was so striking that we performed several control experiments to rule out any possible experimental error. Thus, a second Western blot analysis with a new whole-cell extract of isolate 814 gave similar results. In addition, a duplicate gel was simultaneously run and subjected to Coomassie brilliant blue staining to ensure that equivalent amounts of protein had been loaded in all the gel lanes . Since the polypeptide profiles of all the isolates were almost indistinguishable, the last experiment also indicated that the difference in antigenic profiles detected by Western immunoblotting was not due to a global deficiency in protein expression. Finally, to rule out potential experimental artifacts derived from the gel transfer process, we performed ELISAs using whole cells as the antigen and RA1 antiserum as the primary antibody. As shown in Fig. , isolate 814 produced background levels of the set of antigens detected by RA1 antiserum. Isolate 814 expresses reduced levels of pertactin. | Due to the limited genetic diversity of B. bronchiseptica, it is highly unlikely that an antiserum raised against a certain isolate will not show strong cross-reactivity when it is used to probe a lysate of any other B. bronchiseptica isolate. Nevertheless, to rule out that RA1 antiserum was unable to recognize isolate 814 antigens, we performed another comparative analysis of antigen expression using an antibody of well-characterized specificity: antipertactin MAb BPE3. This antibody was used to probe whole-cell lysates of all the isolates by Western immunoblotting, and then the levels of pertactin were quantitated by densitometry. While the amounts of pertactin expressed by isolates 222 and 629 were very similar to those expressed by RB50, isolate 814 expressed less than one-third of that amount . In addition, the apparent molecular weight of the band detected in the lysate of isolate 814 was slightly higher than that of the rest of the isolates (approximately 70,000 versus 68,000) and matched the molecular weight reported for the B. parapertussis pertactin. Although B. bronchiseptica strains expressing pertactin isotypes with the same molecular weight as that of the B. parapertussis pertactin have recently been isolated , we performed an additional test to confirm that isolate 814 was indeed a B. bronchiseptica isolate rather than an unusual B. parapertussis isolate. As shown in Table , isolate 814 behaved as a typical B. bronchiseptica strain because, unlike human isolates of B. parapertussis, it was oxidase positive, was motile, grew using citrate as the only carbon source, was able to reduce the nitrates to nitrites, and did not produce browning on tyrosine agar. Isolate 814 also differed from ovine isolates of B. parapertussis in several of these characteristics. The epitope recognized by MAb BPE3 has already been identified , and we confirmed that it is unaltered in all the known pertactin isotypes regardless of their molecular weight. Analysis for the presence of Bordetella-specific DNA sequences in isolate 814. | To provide more solid evidence that isolate 814 was indeed a B. bronchiseptica isolate, we decided to analyze the isolate for the presence of B. bronchiseptica-specific DNA sequences. To do that we constructed primers designed to amplify by PCR a 600-bp DNA fragment encompassing the linker-encoding sequences and some of the transmitter-encoding sequences of bvgS. A unique PCR product of the expected size was amplified, cloned, and sequenced by using chromosomal DNA from isolate 814. A search with the BLAST program and the deduced amino acid sequence revealed that the amplified region is 99% identical to the B. bronchiseptica sequence available in GenBank. Compared to the published BvgS amino acid sequence, we found that the sequence of isolate 814 contains 2 amino acid changes (Met-650-Val and Val-677-Ile) over a stretch of 201 amino acids and that both of these substitutions are conservative. Quantification of the transcriptional activity of a Bvg-activated gene in the clinical isolates. | Since prn is transcriptionally activated by the BvgAS signal transduction system, the reduced level of expression of pertactin observed in isolate 814 could be due to a global decrease in BvgAS activity in this isolate. If so, this potential decrease would also affect the expression of other Bvg-activated factors, thus causing an overall reduction in the level of protein expression such as that observed in isolate 814. To test this hypothesis we constructed derivatives of isolates 222 and 814 carrying a promoterless lacZ operon transcriptionally fused to prn and compared the levels of production of beta-galactosidase in the lacZ fusion strains grown under Bvg+-phase conditions. The results of three independent experiments performed in triplicate indicate that the activity of the prn promoter in isolate 814 (mean, 1,692.24 units; standard deviation, 827 units) is approximately 5 times lower than that of isolate 222 (mean, 8,123.75 units; standard deviation, 2,110 units). Statistical analysis revealed that this difference was very significant (P < 0.01). Comparative analysis of the immune responses of rats infected with the clinical isolates. | We hypothesized that whereas B. bronchiseptica may require expression of wild-type levels of antigens involved in virulence for the colonization of healthy immunocompetent hosts, this pathogen may not need to express antigens at those levels for its interaction with immunocompromised patients. Under those circumstances the lack of expression of several major B. bronchiseptica antigens might not be detrimental, and it might even be advantageous for the pathogen if the lack of expression led to inefficient recognition by the immune system. That could precisely be one of the mechanisms enabling isolate 814 to cause a more severe pathology in immunocompromised humans. To test this hypothesis we compared the antibody response induced by experimental infection with isolate 222 with that brought about by infection with isolate 814. Figure shows the results from a representative experiment. Whereas rats infected with isolate 222 mounted a prominent anti-B. bronchiseptica-specific antibody response at day 14 postinoculation, the antibody response elicited by isolate 814 in rats was almost undetectable at the same time point. This difference in immunogenicity between the clinical isolates was long lasting, as revealed by comparison of the antibody responses induced by the isolates at a later time point (49 days) postinoculation . A longer (fivefold) exposure of the films revealed the presence of faint bands in the membranes incubated with serum from animals infected with isolate 814 (data not shown), indicating that the experiment was technically performed as intended. As expected, sera collected both from animals before inoculation and from mock-infected animals at the final experimental time point were confirmed to be devoid of anti-B. bronchiseptica-specific antibodies (data not shown). Replacement of the bvgAS locus of isolate 814 with a wild-type bvgAS allele. | To study the role of the BvgAS system in the phenotype exhibited by isolate 814, we decided to replace its bvgAS locus with a wild-type bvgAS allele and characterize the phenotype of the recombinant strain. To do that, we first deleted the entire bvgAS locus of isolate 814 and adjacent regions by allelic exchange, thus generating strain 814-DAS. PCR analysis (see Material and Methods) revealed that the chromosomal configuration of strain 814-DAS at the deletion site had the intended arrangement (data not shown) We then exchanged the deletion allele present in strain 814-DAS with the wild-type bvgAS allele carried by plasmid pGMT40, which confers resistance to modulation. As a proof that the exchange had taken place as intended, we confirmed that the resultant strain (814-P5) had lost the sensitivity to modulation characteristic of isolate 814 and had become resistant to modulation. Contrary to our expectations, Western blot analyses with either RA1 antiserum or antipertactin MAb BPE3 repeatedly failed to show any quantitative or qualitative difference in protein expression between strain 814-P5 and its isogenic parental strain (data not shown). FIG. 2. | Colonization of the rat respiratory tract by B. bronchiseptica Colonization of the rat respiratory tract by B. bronchiseptica clinical isolates. Wistar rats were inoculated intranasally with 10 mul of PBS containing 103 CFU of the indicated strains. The animals were killed at day 7 postinoculation; and the numbers of CFU recovered from the nasal turbinates, 1 cm of trachea, and the right lung lobes were determined. Each symbol represents a single animal, bars represent the means, and the dotted line represents the lower limit of detection. Statistical significance is designated with asterisks (one asterisk, P < 0.05; two asterisks, P < 0.01). FIG. 5. | Western immunoblot of RB50 whole-cell lysate antigen probed with serum from rats infected with isolate 222 or isolate 814. Western immunoblot of RB50 whole-cell lysate antigen probed with serum from rats infected with isolate 222 or isolate 814. Lanes 1, 3, and 5, serum from different rats infected with isolate 222 for either 14 days (lanes 1 and 3) or 49 days (lane 5); lanes 2, 4 and 6, serum from different rats infected with isolate 814 for either 14 days (lanes 2 and 4) or 49 days (lane 6). The positions of the molecular weight markers (in thousands) are shown at the left. DISCUSSION : We found 11 reports in the English-language literature describing the isolation of B. bronchiseptica from 15 cases of infectious processes affecting AIDS patients . The patients are usually in an advanced stage of HIV infection and in some cases have had contact with pets. Clinical presentation is acute or subacute and includes fever and respiratory symptoms. In most cases, chest X-ray images show interstitial infiltrates that make the physician consider the involvement of pathogens other than B. bronchiseptica. Etiological diagnosis often requires the retrieval of respiratory samples by bronchoscopy (through bronchoalveolar lavage or a protected brush catheter). The concomitant isolation of other microorganisms from six patients hampered the evaluation of the pathogenic role of B. bronchiseptica in the respiratory processes in those patients. This fact is more evident considering that four of the patients were cured with specific treatment against the pathogens concomitantly isolated . On the other hand, it has been emphasized that a poor correlation exists between the results of antibiogram tests and the clinical response to the antimicrobial regimen selected. In fact, the first specific treatment failed for three of the patients, and one patient died, even though the vitro activity of the antimicrobial treatment used was adequate . Treatment duration is also a controversial issue. It varies from 2 to 4 weeks for patients with a good response to 6 months for patients with recurrent symptoms . The increasing rate of isolation of B. bronchiseptica from AIDS patients with pneumonia emphasizes its importance as an opportunistic pathogen and guarantees the consideration of B. bronchiseptica as a potential etiological agent of pneumonia with or without lung cavitation in AIDS patients. Some aspects of the present study are worth mentioning. First, to our knowledge there has been only one report describing lung cavitation as the clinical presentation of pneumonia caused by B. bronchiseptica in an HIV-infected patient. This patient, like patient 1 in the present study, was first mistakenly diagnosed with tuberculosis and was consequently treated with tuberculostatic agents. Second, the isolation of B. bronchiseptica without the concomitant isolation of other pathogens and the good clinical evolution observed for patient 1 after specific antimicrobial treatment indicate that B. bronchiseptica isolate 814 was the causative agent of patient 1's cavitary pneumonia. To the contrary, all the clinical data suggest that B. bronchiseptica infection was not the primary etiological factor in patient 2's respiratory process. In this work we show that isolate 814 produces reduced amounts of a large number of antigens including pertactin. Consistent with the low level of expression of pertactin, our results demonstrate that the level of transcription driven by the prn promoter is greatly reduced in isolate 814. It seems likely that this decrease in transcriptional activity may cause, at least in part, the deficient expression of pertactin observed in isolate 814. However, the fact that the levels of expression of antigens including pertactin were not restored after exchange of the bvgAS locus with a wild-type bvgAS allele indicates that other (BvgAS-independent) mechanisms must also be impaired in this isolate. If these mechanisms, as our results suggest, act at the translational or postranslational level, that would explain why they seem to be dominant over the mere transcriptional activity mediated by BvgAS. Precisely because of this effect, our results do not allow us to draw a conclusion as to whether the bvgAS locus of isolate 814 is fully functional. Whatever the molecular basis of the phenotype displayed by isolate 814 is, it seems to affect the expression of its virulence determinants to different degrees. Thus, while the level of expression of pertactin is reduced in this isolate, the production of FHA seems to be unaffected. Interestingly, it has very recently been described that deep rough mutants of B. bronchiseptica express reduced levels of pertactin and other Bvg-activated factors . Moreover, like isolate 814, these B. bronchiseptica rough variants were originally isolated from a patient with pneumonia . However, unlike isolate 814, deep rough B. bronchiseptica mutants also express reduced levels of FHA. We are investigating whether isolate 814 carries mutations leading to expression of an altered lipopolysaccharide. The reduced levels of antigen production exhibited by isolate 814 may suffice to explain why this isolate was defective for the colonization of the trachea of the rat, which is one of the natural hosts of B. bronchiseptica. In fact, it has been shown that B. bronchiseptica mutants expressing intermediate levels of several virulence determinants are unable to colonize the lower respiratory tract of rats . However, the experimental hosts used for B. bronchiseptica in the present investigation were immunocompetent animals. Therefore, the inability of isolate 814 to efficiently colonize them may not reflect the pathogenic potential of this isolate in an immunocompromised host like patient 1. In support of this notion, it has recently been shown that a B. bronchiseptica rough mutant colonizes the respiratory tract of immunocompromised mice as efficiently as its wild-type isogenic strain, whereas it exhibits a reduced ability to colonize the lower respiratory tract of immunocompetent animals compared to that of the wild-type strain . Our results indicate that, despite its prominent defect, isolate 814 not only was able to colonize the respiratory tract of patient 1 but also caused a respiratory pathology more severe than that brought about by other B. bronchiseptica isolates including isolate 222. This behavior, which is totally unexpected from a seemingly defective strain, led us to hypothesize that isolate 814 might have a selective advantage for the colonization of an immunocompromised host. In support of this hypothesis, our results show that isolate 814 elicits an almost undetectable antibody response even when colonizing an immunocompetent host. In light of patient 1's immunocompromised condition, it is very likely that her immune response against isolate 814 was even weaker. Interestingly, it has been shown that the clinical outcome of pneumonia caused by an opportunistic pathogen in HIV-infected patients was more severe in subjects who had low levels of specific antibodies, whereas patients who recovered had elevated specific antibody levels over time . Nevertheless, to test our hypothesis it would have been necessary to quantitate the anti-B. bronchiseptica-specific antibodies in patient 1's serum; unfortunately, however, no more serum could be obtained for this purpose. Finally, for isolate 814, reduced levels of expression of antigens could have the additional advantage of relieving it from the metabolic burden involved in synthesizing wild-type amounts of a large number of polypeptides. While such a strain would be counterselected in an immunocompetent host, its superior growth capability may turn out to be advantageous in an immunocompromised host. In agreement with this hypothesis, significant reductions in the doubling times of both B. bronchiseptica and B. pertussis have been observed when isolates of these species are grown under conditions that favor even a slight decrease in the production of Bvg+-phase-specific antigens (e.g., under Bvgi-phase conditions) (G. Martinez de Tejada, unpublished observations). To draw a more complete picture of the features enabling B. bronchiseptica to cause severe respiratory diseases in immunocompromised humans, it seems necessary to characterize a collection of isolates recovered from patients with similar pathologies (i.e., cavitary pneumonia). Such a broad analysis will be possible when a relevant number of cases similar to the one reported here come to light. However, the following conclusion can be extracted from our study: appearances can be deceptive when predicting the pathogenic potential in immunocompromised humans of a B. bronchiseptica isolate displaying a major defect in protein expression involving at least one virulence factor. In fact, if that defect could be proven to result in enhanced virulence, this would be, to our knowledge, the first report suggesting that what makes a B. bronchiseptica isolate defective for the colonization of one particular host may constitute an adaptive advantage for the colonization of another one. Backmatter: PMID- 12202542 TI - Hepatitis C Virus Genotyping: Interrogation of the 5' Untranslated Region Cannot Accurately Distinguish Genotypes 1a and 1b AB - Although the 5' untranslated region (5' UTR) is the most conserved region of the hepatitis C virus (HCV) genome, it has been suggested that interrogation of this region is sufficient for determination of the HCV genotype. We compared two methods of determination of the HCV genotype: (i) direct sequencing of the DNA of the NS-5b region and (ii) reverse line probe assay (LiPA; INNO-LiPA HCV II; Innogenetics N.V.) of the 5' UTR. There was 100% concordance between the two methods for genotype but only 80% concordance for subtype. A significant percentage of genotype 1a isolates were misclassified by LiPA as genotype 1b. Sequence analysis revealed that the only consistent difference in the 5' UTR for these genotype 1a isolates misclassified as genotype 1b was a single nucleotide (A/G) at position -99 of the HCV genome. All isolates with discordant results analyzed had a G at this position, consistent with LiPA determination of these samples as subtype 1b. However, sequence analysis of 222 nucleotides in the NS-5b region clearly identified all of these isolates as subtype 1a. Population distribution data from the University of Pittsburgh Medical Center of over 200 samples analyzed by sequencing of the NS-5b region and over 1,000 samples analyzed by LiPA also indicated that INNO-LiPA HCV II cannot accurately differentiate HCV genotype 1a isolates from HCV genotype 1b isolates. We provide evidence that the A/G at position -99 represents a sequence polymorphism in the HCV genome that cannot differentiate subtype 1a from subtype 1b isolates. In conclusion, the 5' UTR is not heterogeneous enough for use in determination of the HCV subtype and cannot be used for differentiation of HCV genotypes 1a and 1b. Keywords: Introduction : Hepatitis C virus (HCV) is an enveloped positive single-stranded RNA virus which is the major cause of chronic hepatitis worldwide . Chronic hepatitis resulting from HCV infection may lead to severe sequelae, including hepatic cirrhosis and hepatocellular carcinoma . HCV demonstrates a high degree of sequence variation throughout its genome and exists in vivo as a group of heterogeneous but closely related quasispecies . However, the levels of heterogeneity differ considerably among the various regions of the virus, ranging from as little as 10% in the 5' untranslated region (5' UTR) to 50% or more within the E1 region . Based on the analysis of variable coding regions in the viral genome, distinct genotypes as well as subtypes have been defined . The distribution of HCV genotypes differs geographically, with subtype 1a being most common in the United States, subtype 1b being most common in Europe and Japan, and other genotypes being prevalent in other parts of the world (for a review, see reference ). Several studies have shown that HCV genotype influences the response to therapy with alpha interferon alone or in combination with ribavirin . In general, HCV type 2 and 3 isolates have higher rates of response to therapy than type 1 isolates. In addition, some studies suggest that genotype alone may predict disease severity; e.g., patients infected with HCV genotype 1a or 1b may exhibit more severe liver disease (, , , -, -). These findings indicate an important role of genotype identification for prediction of the outcome of HCV infection and the selection of patients for treatment protocols. Identification of the HCV genotype by interrogation of several different regions of the HCV genome has been reported. A wealth of phylogenetic information has been derived from sequence analysis of the NS-5 gene of HCV . We have determined HCV genotypes using direct sequencing of the DNA of the NS-5b region for several years and have devised an algorithm of sequence analysis for genotype determination on the basis of the sequence of this region. Although sequence analysis is considered the "gold standard" for HCV genotype determination, it is expensive, time-consuming, and inconvenient for routine use. A commercial assay is available which allows rapid analysis for the HCV genotype via a reverse hybridization line probe assay (LiPA) of the 5' UTR of HCV (INNO-LiPA HCV II; Innogenetics N.V., Zwijnaarde, Belgium). The 5' UTR is the most highly conserved region among HCV strains . This makes this region theoretically superior for sensitivity but limits its usefulness for determining differences between various subtypes. However, phylogenetic analysis of sequence information obtained from the 5' UTR has correlated fairly well with phylogenetic analysis of sequence information obtained from other regions of the viral genome, including the core, NS-3, NS-4, and NS-5 regions . In this study, we compared the HCV genotypes determined by direct sequencing of the DNA of the NS-5b region and LiPA analysis of the 5' UTR for 63 samples and report our experience with genotyping of over 1,000 samples analyzed by one or both methods over 5 years. MATERIALS AND METHODS : Samples. | Fresh plasma or liver biopsy specimens sent to the Laboratory of Molecular Diagnostics at the University of Pittsburgh Medical Center for routine HCV genotype analysis over the past 5 years were analyzed. A total of 247 samples were analyzed by NS-5b sequence analysis from 1996 to 1998; a total of 1,035 samples were analyzed by LiPA from 1998 to 2001. A total of 63 samples that were representative of the genotypes identified in our laboratory were initially selected for comparison of genotyping by the two different methods. Sequencing analysis of both the NS-5b region and the 5' UTR was performed for a number of samples for which LiPA gave discrepant or indeterminate results. Analysis of these samples was done in compliance with federal and institutional review board policies. cDNA preparation. | RNA was extracted from 230 mul of plasma or 3-mm3 liver biopsy samples with the TRIzol reagent (GIBCO-BRL [now Invitrogen], Carlsbad, Calif.), precipitated with isopropanol, and resuspended in 20 mul of diethyl pyrocarbonate-treated water. cDNA was prepared from 3 mul of plasma RNA or 500 ng of liver RNA. The RNA was heated for 5 min at 70C and quick chilled on ice; and then 7 mul of a pre-master mixture containing 22 pmol of random hexamers (GIBCO-BRL), 50 mM Tris-HCl (pH 8.3), 7.5 mM KCl, 3 mM MgCl2, 10 mM dithiothreitol, 0.2 mM (each) deoxynucleoside triphosphate (Pharmacia, Peapack, N.J.), and 20 U of RNase inhibitor (Promega, Madison, Wis.) was added. A total of 100 U of Moloney murine leukemia virus reverse transcriptase (GIBCO-BRL) was added, and the mixture was incubated at 37C for 60 min. LiPA. | RNA purification and cDNA preparation were performed as described above. Analysis of the 5' UTR was then carried out with the INNO-LiPA HCV II kit (Innogenetics N.V.) according to the instructions of the manufacturer. Briefly, biotinylated primers are used to amplify the 5' UTR region of HCV by PCR. The biotin-labeled PCR products are then reverse hybridized to type-specific probes attached to nitrocellulose strips. Development results in a purple or brown precipitate as a positive line on the strip. The HCV type and subtype were deduced on the basis of the patterns of hybridizing bands by using the INNO-LiPA HCV II interpretation chart. PCR of the NS-5b and 5' UTR regions. | PCR amplification was performed with the primers described below in a model 9600 thermocycler (Perkin-Elmer) with 10 mul of cDNA in a total reaction volume of 50 mul containing each primer at 10 pmol, 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, each deoxynucleoside triphosphate at a concentration of 200 muM, and 1.0 U of Taq DNA polymerase (Perkin-Elmer). After an initial denaturation step at 95C for 1 min, 40 cycles of PCR at 95C for 20 s, 56C for 30 s, and 72C for 1 min were carried out, with a final extension of 72C for 10 min. Nested PCR of the NS-5b region was performed with 1 mul of the first-round amplification product if the amplification product was not sufficient for sequence analysis. Twenty-five cycles of the nested PCR were carried out by using the same conditions described above, except that an annealing temperature of 60C was used. First-round amplification of the NS-5b region was done with a sense primer specific to bases 7904 to 7922 (5'-TGG GGT TCT CGT ATG ATA CCC-3') and an antisense primer specific to bases 8295 to 8275 (5'-CCT GGT CAT AGC CTC CGT GAA-3'). Nested PCR primers for detection of the NS-5b region were a sense primer specific to bases 7916 to 7935 (5'-GAT ACC CGC TGC TTT GAC TC-3') and an antisense primer specific to bases 8284 to 8266 (5'-CCT CCG TGA AGG CTC TCA G-3'). PCR primers for detection of the 5' UTR were a sense primer specific to bases -333 to -317 (5'-CCG ATT GGG GGC GAC AC 3') and an antisense primer specific to bases 05 to -14 (5'-CTC ATG GTG CAC GGT CTA C-3'). The nucleotide numbering for the primers is according to the prototype HCV type 1 sequence . All primers were manufactured by GIBCO-BRL. Direct sequencing of PCR products. | PCR products were purified with MicroCon-100 filters and were sequenced with standard or Big Dye terminator kits (Perkin-Elmer) on an automated ABI373 or ABI377XL instrument. The sequencing primer for the NS-5b region was a sense primer specific to bases 7916 to 7935 (5'-GAT ACC CGC TGC TTT GAC TC-3'). Sequencing of the 5' UTR was performed in both directions for 6 samples that gave discrepant results by the two methods and for 10 samples that gave indeterminate results by LiPA by using a sense primer specific to bases -333 to -317 (5'-CCG ATT GGG GGC GAC AC-3') and an antisense primer specific to bases 05 to -14 (5'-CTC ATG GTG CAC GGT CTA C-3'). Sequence analysis of the NS-5b region. | We have devised an algorithm for determination of HCV genotype based on sequence analysis of the NS-5b region (Z. Chen and J. A. Kant, unpublished data). Briefly, by using consensus nucleotide sequences from a 222-bp region within the HCV NS-5b region , three or four key nucleotides that are specifically associated with major types of HCV were first selected. These nucleotides are shared by each subtype within major types. Another 5 to 11 key nucleotides specifically associated with subtypes within major HCV types were then chosen. Determination of the type and subtype of HCV in each patient sample was made on the basis of the presence of these key nucleotides that had been identified. In addition, the alignments of the sequenced 222-bp region with the NS-5b region was determined for each patient sample by using consensus HCV subtype sequences , and the total number of mismatches per 222 bp was determined for each subtype within the assigned type. Typically, we find fewer than 15 mismatches per 222 bp for samples of the same HCV subtype (>93% similarity) and greater than 30 mismatches per 222 bp for samples of different HCV subtypes (<86% similarity). This is consistent with proposed similarities for classification of HCV genotype on the basis of the sequence of the NS-5 gene . For all samples analyzed, determination of genotype based on this algorithm correlated with the results obtained by sequence analysis with the BLAST program. Restriction enzyme digestion. | Restriction enzyme analysis of the 5' UTR was performed for a subset of samples that were genotype 1 by LiPA but whose subtypes could not be determined. A total of 20 mul of purified reverse transcription-PCR (RT-PCR) products from the 5' UTR were digested with Alw26I at 37C for 3 h and analyzed on a 2% agarose gel stained with ethidium bromide. RESULTS : Comparison of HCV genotype by NS-5b sequence analysis and LiPA. | A total of 63 samples were analyzed by both NS-5b sequence analysis and LiPA in a comparative analysis . The genotypes of the samples selected for the comparative analysis were representative of the genotypes identified in our laboratory by NS-5b sequencing in order to evaluate the performance of the INNO-LiPA HCV II kit. There was 100% concordance between the two methods for type and 80% concordance for subtype. LiPA failed to assign a subtype to four (6%) samples. Of the samples that were assigned a subtype by both methods, there were discrepant subtyping results for 9 of 63 (14%) samples. One sample was called subtype 2b by NS-5b sequence analysis and subtype 2a/2c by LiPA. LiPA cannot discriminate between subtypes 2a and 2c. The remainder of the samples with discrepant results (13% of the total) were called subtype 1b by LiPA and subtype 1a by NS-5b sequence analysis. Of the 37 samples identified as subtype 1a by NS-5b sequence analysis, 22% were classified as subtype 1b by LiPA. NS-5b sequence analysis of samples with discrepant results. | Figure shows the results of NS-5b sequence analysis for the eight samples with discrepant results that were subtyped as HCV subtype 1b by LiPA and subtype 1a by NS-5b sequence analysis. All of the samples with discrepant results were sequenced at least twice in order to confirm their classification as type 1a. The eight samples each demonstrated six to seven of the seven nucleotides identified as key nucleotides for genotype 1a. In contrast, none of these samples had any of the 11 nucleotides identified as key for genotype 1b. In addition, the alignments of the sequences with consensus sequences for subtypes 1a and 1b, as described by Simmonds et al. , were compared. The total number of mismatches was significantly higher with subtype 1b consensus sequences than with subtype 1a consensus sequences for all eight isolates analyzed. This corresponds to 95.5 to 98.6% nucleotide identity with the subtype 1a consensus sequence and 77 to 85% nucleotide identity with the subtype 1b consensus sequence, thus confirming the identification of all eight of these samples as HCV genotype 1a. 5' UTR sequence analysis of samples with discrepant results. | In order to investigate the discrepancy further, we sequenced the DNA of the 5' UTR region (the region interrogated by LiPA) for six of the eight samples with discrepant results for subtype 1a and 1b and for control subtype 1a and subtype 1b samples . The only consistent difference in the 5' UTR for these samples was a single nucleotide (A/G) at position -99 of the HCV genome. Determination of a sample as subtype 1a or 1b by LiPA is on the basis of this single A/G difference at position -99, with LiPA probe 5 corresponding to an A residue at nucleotide -99 and LiPA probe 6 corresponding to a G residue at this position . All six samples with discordant results analyzed had a G residue at this position, consistent with identification of these samples as subtype 1b by LiPA analysis of the 5' UTR. However, all six samples were clearly subtype 1a by NS-5b sequence analysis . One of the samples with discrepant results listed in Table was subtype 1a by NS-5b sequence analysis but genotype type 1 by LiPA and could not be subtyped by LiPA. Subsequent analysis of over 1,000 samples by LiPA in our laboratory indicated that a significant proportion (7.7%) of HCV genotype 1 samples cannot be subtyped by LiPA (see Fig. ). Many of these have LiPA band pattern 1, 2, 3, 4, 5, 6. In order to investigate these further, DNA sequencing analysis was performed for 10 samples that were classified as type 1 by LiPA but that could not be subtyped by LiPA. The genotypes of all 10 of these were unambiguously subtype 1a by sequencing of the NS-5b gene (data not shown). However, sequence analysis of the 5' UTR of these 10 samples with LiPA band pattern 1, 2, 3, 4, 5, 6 demonstrated a heterogeneous pattern, with both A and G nucleotides identified at position -99 (Fig. and ). In some samples, almost equal proportions of both nucleotides were present (a heterozygous pattern); in other samples, the peak for one nucleotide was higher than that for the other. Digestion of RT-PCR products of type 1 isolates with Alw26I. | The 5' UTR RT-PCR products of six of the type 1 samples whose sequences are presented in Fig. as well as those of subtype 1a and subtype 1b controls were subjected to restriction enzyme digestion with Alw26I in order to differentiate between a G residue at position -99 (subtype 1b pattern) and an A residue at this position (subtype 1a pattern) . Restriction enzyme digestion of all six type 1 samples with LiPA band pattern 1, 2, 3, 4, 5, 6 demonstrated a heterogeneous pattern. These results confirmed the results of sequencing analysis of the 5' UTR that samples with LiPA banding pattern 1, 2, 3, 4, 5, 6 carry both an A residue and a G residue at position -99 (Fig. and ) and demonstrated that restriction enzyme digestion is a rapid method of analyzing these samples. The patterns observed for HCV samples with both an A residue and a G residue at position -99 could potentially be explained by the presence of a mixed subtype 1a and 1b infection. Mixing of subtype 1a and subtype 1b controls at a 1:1 ratio also demonstrated a heterogeneous pattern by Alw26 I digestion, similar to the pattern for the samples that were type 1 but unsubtypeable by LiPA. However, at a subtype 1a to subtype 1b ratio of 1:4, the uncut pattern for subtype 1a was very difficult to discern . We also analyzed artificially mixed subtypes by LiPA and demonstrated a mixed LiPA 1, 2, 3, 4, 5, 6 banding pattern with a 1:1 ratio of subtype 1a to subtype 1b; however, with a 1:4 ratio the pattern for subtype 1a was very difficult to discern, and with a 1:8 ratio only the pattern for subtype 1b could be detected (data not shown). Distribution of HCV genotypes among samples analyzed by LiPA versus NS-5b sequence analysis. | A total of 247 samples recovered from 1996 to 1998 were analyzed by NS-5b sequence analysis by using the algorithm described above. A total of 1,035 samples recovered from 1998 to 2001 were analyzed by LiPA. Figure shows a comparison of the distribution of genotypes obtained in our laboratory by the two methods. Samples that could not be typed by RT-PCR were eliminated from the analysis: for samples that were known to be HCV positive by either branched DNA testing or RT-PCR, there was a 2.9% RT-PCR failure rate for LiPA and a 1.4% failure rate for sequence analysis of the NS-5b region. The total distributions of the genotypes were very consistent between the two methods. However, there was a marked difference in the distribution of subtype 1a versus subtype 1b. The distribution by NS-5b sequence analysis was 64.4% subtype 1a and 16.6% subtype 1b. In contrast, LiPA analysis resulted in an almost equal distribution of subtype 1a versus subtype 1b (35 versus 30%). Thus, LiPA identified samples as subtype 1b at a 13% greater frequency than NS-5b sequence analysis did. Of the samples that were HCV genotype 1, LiPA identified samples as subtype 1b with a 21% greater frequency. In addition, by LiPA analysis 33 samples (3.2%) were untypeable and 82 isolates (7.9%) were called type 1 but could not be further subtyped. A total of 59 samples (5.7%) were genotype 1, unsubtypeable with band pattern 1, 2, 3, 4, 5, 6 and likely had both an A residue and a G residue at position -99. By subsequent NS-5b sequence analysis, 4 of 4 of the untypeable samples and 15 of 15 of the type 1, unsubtypeable samples were shown to be HCV genotype 1a (data not shown). FIG. 1. | NS-5b sequence analysis of samples with discrepant results. NS-5b sequence analysis of samples with discrepant results. The algorithm used to define HCV genotype by NS-5b sequence analysis is described in Materials and Methods. The key nucleotides associated with subtypes 1a and 1b are numbered and boxed. Sequence results are shown only for these key nucleotides. Nucleotide 01 in the algorithm corresponds to nucleotide 7975 in the HCV type 1 prototype sequence . The rightmost column shows the total number of mismatches of 222 bp analyzed compared to consensus subtype 1a or subtype 1b sequences . FIG. 2. | DNA sequencing analysis of the 5' UTR for control subtype 1a and subtype 1b samples (cases A and B), for six of the samples from Fig. DNA sequencing analysis of the 5' UTR for control subtype 1a and subtype 1b samples (cases A and B), for six of the samples from Fig. with discrepant results for subtypes 1a and 1b (cases 1 to 6), and for 10 samples that LiPA classified as type 1 but that were unsubtypeable by LiPA (cases 9 to 18). All of the samples had identical 5' UTR sequences with the exception of the nucleotide regions at positions -99 and -138. The banding pattern obtained by LiPA is shown in parentheses at the right. C region, core region; F, forward sequencing primer; R, reverse sequencing primer; A/G, both an A-residue peak and a G-residue peak were observed at an approximately equal peak ratio by sequence analysis; A/g, a higher peak was observed for the A residue than for the G residue by sequence analysis. FIG. 5. | Comparison of the distributions of HCV genotypes obtained by LiPA versus those obtained by NS-5b sequence analysis. Comparison of the distributions of HCV genotypes obtained by LiPA versus those obtained by NS-5b sequence analysis. FIG. 3. | Representative sequence analysis for HCV subtype 1a and 1b controls and for a type 1, unsubtypeable isolate with LiPA banding pattern 1, 2, 3, 4, 5, 6. Representative sequence analysis for HCV subtype 1a and 1b controls and for a type 1, unsubtypeable isolate with LiPA banding pattern 1, 2, 3, 4, 5, 6. Position -99 is indicated by the arrows. Samples found to be subtype 1a by LiPA have an A residue at this position; samples found to be subtype 1b have a G residue. The heterogeneous pattern observed at this nucleotide position for cases 9 to 18 from Fig. is demonstrated. Forward (plus-strand) or reverse (minus-strand) sequences are indicated. FIG. 4. | Digestion of genotype 1 RT-PCR products with Alw26I. Digestion of genotype 1 RT-PCR products with Alw26I. Restriction enzyme analysis of the 5' UTR was performed for a subset of samples that were genotype 1 by LiPA but that could not be subtyped by LiPA. 5' UTR RT-PCR products were subjected to restriction enzyme digestion with Alw26I in order to differentiate between a G residue at position -99 (subtype 1b pattern) and an A residue at this position (subtype 1a pattern). Upper panel, diagram showing the position of Alw26I restriction sites within the 340-bp 5' UTR RT-PCR product; bottom panel, 2% ethidium bromide-stained agarose gel of restriction enzyme products. The contents of the lanes are as follows (the lane numbers represent the unnumbered lanes from left to right, respectively): lanes 1 and 12, molecular weight markers; lane 2, subtype 1b control demonstrating 209- and 91-bp bands; lane 3, subtype 1a control demonstrating a 300-bp band; lanes 4 and 8, mixtures of subtype 1a and subtype 1b control RT-PCR products at 1:1 and 1:4 ratios, respectively; lanes 5 to 7 and 9 to 11, genotype 1 samples that were unsubtypeable by LiPA and that were shown to have both an A residue and a G residue at position -99. Sample numbers at the bottom correspond to those in Fig. . TABLE 1 | Comparison of HCV genotype by NS-5b sequence analysis versus LiPA DISCUSSION : It has been suggested that interrogation of the HCV 5' UTR, although less heterogeneous than other regions of the viral genome, correlates well with and is more sensitive than interrogation of other genomic regions of HCV and is sufficient for determining the HCV genotype . Although direct sequencing is more time-consuming and labor-intensive than LiPA, we have not experienced the failure rates that have been reported by other investigators for sequencing of the NS-5b region . Analysis of the 5' UTR by LiPA with the INNO-LiPA HCV II kit provides a fast and easy method for determination of the HCV genotype. Although the 5' UTR is less polymorphic, the results obtained by LiPA correlate well with those obtained by sequence analysis of the NS-5b region for determination of the genotype, with 100% concordance in this study. However, several reports have indicated that analysis of the 5' UTR cannot accurately differentiate HCV subtypes . This is likely because of the relative lack of heterogeneity of the 5' UTR. Consistent with the previous reports, in a direct comparison of 63 samples in this study, LiPA misclassified similar20% of HCV subtype 1a isolates as subtype 1b . Thus, LiPA tends to overcall genotype 1 samples as subtype 1b. Sequencing of both the 5' UTR and the NS-5b gene for the isolates in six of these samples with discrepant results indicated that the isolates were clearly genotype 1a by analysis of 222 bp of NS-5b gene, but because they had a G nucleotide at position -99 they were genotyped as subtype 1b by LiPA. Consistent with other reports , we have also identified by NS-5b gene sequencing a genotype 1b isolate that had an A residue at position -99 but that was misclassified as type 1a by LiPA analysis (data not shown). These results could be explained by (i) a mixed infection of subtype 1a and 1b isolates; (ii) infidelity of the PCR introduced by Taq polymerase, although this is unlikely to occur consistently at a single nucleotide; (iii) recombination between different HCV subtypes; or (iv) an A/G polymorphism that may exist at nucleotide -99 in HCV isolates. In the last scenario, either an A residue or a G residue, or both nucleotides, could be present in subtype 1a or subtype 1b isolates. This would invalidate the use of this single nucleotide to differentiate subtype 1a and subtype 1b isolates. In addition, 10 genotype 1 isolates that could not be subtyped by LiPA and that had the band pattern 1, 2, 3, 4, 5, 6 had both an A residue and a G residue in the nucleotide at position -99 (Fig. to ) but were clearly genotype 1a by NS-5b sequence analysis. These results could be explained by a mixed infection with subtypes 1a and 1b or by an A/G sequence polymorphism at position -99. Although we cannot rule out the possibility that these samples represent mixed infections with genotypes 1a and 1b, 5' UTR restriction fragment length polymorphism analysis and LiPA evaluation of artificial mixtures of isolates indicate that there would have to be an approximately equal mix of types 1a and 1b to obtain the patterns observed (Fig. ; see also Results). A sequence polymorphism is a more likely explanation, as we would have expected to identify the isolates in at least some of these samples as genotype 1b by sequencing of the NS-5b region if they were truly mixed infections with types 1a and 1b in approximately equal proportions. To our knowledge, this is the first report that HCV genotype 1a isolates from a single patient sample can have both an A residue and a G residue at position -99. The isolates in a total of 59 of 1,035 samples tested by LiPA (5.7%) gave the band pattern 1, 2, 3, 4, 5, 6 and likely have both an A residue and a G residue at this position. The data presented here are consistent with what would be expected if the A/G at nucleotide at position -99 was a sequence polymorphism rather than segregating with a particular subtype: either an A residue or a G residue can be present at nucleotide -99 in both genotype 1a and genotype 1b isolates, and in some cases, both nucleotides can be present in a single genotype as part of the group of quasispecies present in an infected patient. The hypothesis that the A/G at position -99 is a sequence polymorphism in the HCV genome that cannot differentiate between subtypes 1a and 1b is further supported by the genotype distribution in our laboratory of the isolates in over 200 samples analyzed by NS-5b sequencing from 1996 to 1998 and over 1,000 samples analyzed by LiPA from 1998 to 2001 . The distribution in our laboratory of subtype 1a versus subtype 1b isolates was 64 and 17%, respectively, by NS-5b sequence analysis. This distribution is consistent with those in reports of studies from various regions throughout the United States, which have indicated that similar60 and similar20% of U.S. isolates are genotypes 1a and 1b, respectively . In contrast, the distributions of subtype 1a versus subtype 1b isolates by LiPA analysis were 35 and 30%, respectively. Thus, there was almost an equal distribution of subtype 1a versus subtype 1b isolates by LiPA, consistent with an inability of the A/G at position -99 to distinguish genotype 1a from genotype 1b. Interestingly, these distribution data are consistent with those in reports of other studies conducted in the United States that have used the 5' UTR for genotyping . Although we cannot rule out the possibility that the relative incidence of type 1b has increased dramatically at our institution in the past 3 years, this is unlikely, as the population seen over that period was very consistent. In addition, distribution data for over 1,000 samples analyzed by LiPA in our laboratory over the past 3 years indicate that similar30% of genotype 1a isolates were likely misclassified by LiPA as either genotype 1b or type 1, unsubtypeable . These data correlate well with those obtained when the results of LiPA and NS-5b sequencing were compared in parallel . Interestingly, these data are consistent with those in the original paper by Stuyver et al. describing the use of reverse hybridization in the 5' UTR for HCV genotyping. Of 10 prototype 1a and 1b sequences used to define reverse hybridization probes for this assay, only 70% had the pattern of an A residue for type 1a and a G residue for type 1b at position -99. Of the type 1a isolates, 30% (one of three) had a G residue at this position and would have been misclassified as type 1b; similarly, of the type 1b isolates, 30% (two of seven) had an A residue at this position and would have been misclassified as type 1a. Despite mounting evidence that analysis of the 5' UTR does not accurately differentiate HCV subtypes, the INNO-LiPA HCV kit II provides interpretation criteria for multiple subtypes of HCV, and many laboratories use 5' UTR analysis for determination of HCV subtypes. In conclusion, LiPA allows the easy and rapid determination of HCV genotypes but fails to determine the subtypes of isolates in many samples and misclassifies a significant percentage of type 1a isolates as type 1b. These isolates are unlikely to be clinically significant, as most studies have identified clinical differences only between those infected with type 1 isolates and those infected with non-type 1 isolates. However, these findings have implications for epidemiologic studies, as well as for studies investigating clinical differences between patients infected with HCV genotypes 1a and 1b . If 5' UTR analysis overrepresents genotype 1b or cannot differentiate between genotypes 1a and 1b, any clinical difference between patients infected with genotypes 1a and 1b will be diluted in outcome studies that interrogate the 5' UTR for genotyping. Indeed, many studies in both the United States and Europe that have reported differences in outcomes between patients infected with genotypes 1a and 1b have used the NS-5 or core regions to define the genotype . On the other hand, several studies that have used the 5' UTR for genotyping have failed to find a difference in clinical outcomes between type 1a- and type 1b-infected patients . Thus, it is possible that some of the differences in studies of the outcomes of genotype 1b infections are due to the method of genotyping used. We conclude that 5' UTR analysis (by LiPA analysis or other methods) is sufficient for determination of the HCV type and can be used for clinical purposes to differentiate type 1 from non-type 1 isolates, but determination of subtype should rely on analysis of more heterogeneous regions of the HCV genome. Backmatter: PMID- 12202576 TI - High Prevalence of Human Papillomavirus (HPV) Infections and High Frequency of Multiple HPV Genotypes in Human Immunodeficiency Virus-Infected Women in Brazil AB - A group of 208 human immunodeficiency virus (HIV)-infected women in Brazil were studied for the presence of human papillomavirus with the general SPF10 PCR primer set. Virtually all (98%) women were found positive for human papillomavirus (HPV) DNA. Genotyping by the reverse hybridization line probe assay (HPV-LiPA) revealed a high prevalence of multiple genotypes (78.9% of the cases), with an average of 3.1 genotypes per patient (range, 1 to 10 genotypes). HPV 6 was the most prevalent genotype and was observed in 80 (39.2%) patients, followed by types 51 (31.9%), 11 (26.0%), 18 (24.0%), and 16 (22.5%). Of the genotypes detected, 40.9% were low-risk genotypes. Twenty-two (10.5%) patients showed normal (Pap I) cytology, 149 (71.6%) patients had inflammation (Pap II), and 28 patients (13.4%) had a Pap III score. The prevalence of high-risk genotypes increased with the cytological classification. There were no significant associations between the number of HPV genotypes detected and the cytological classification, HIV viral load, and CD4 count in these patients. In conclusion, the highly sensitive SPF10 LiPA system shows that a very high proportion of HIV-infected women in Brazil are infected with HPV and often carry multiple HPV genotypes. Keywords: Introduction : Human papillomavirus (HPV) infections are associated with benign and malignant lesions of cutaneous and mucosal epithelia . So far, more than 100 different HPV genotypes have been identified, of which more than 40 have been detected in the anogenital area. Diagnosis of HPV infection is almost entirely based on molecular tools, which are mainly PCR-based. General or consensus PCR primers have been developed that detect a broad spectrum of HPV genotypes in a single PCR . HPV genotypes that have been detected in cervical carcinomas and in precursor lesions , e.g., HPV 16 and 18, are defined as high-risk genotypes and imply a comparatively high risk for invasive disease. In contrast, other genotypes (e.g., HPV 6 and 11) are considered low-risk genotypes because they are associated with a relatively low risk for the development of cervical carcinoma. Only a small proportion of HPV-infected women will eventually develop cervical neoplasia, and the precise etiologic role of HPV and natural history of the infection remain unknown. The status of the immune system is considered a crucial factor in HPV infections and may determine the development of persistence after primary infection, which has emerged in several studies as an important risk factor for cervical neoplasia . HPV-associated malignancies occur at increased rates in human immunodeficiency virus (HIV)-infected persons, and HPV DNA is commonly detected in the genital mucosa of HIV-infected women. The prevalence of infection is generally much higher than in control groups comprising seronegative women of similar sociodemographic characteristics . This may be explained by an HIV-impaired immune system, which permits a high HPV viral load and persistent HPV infection, leading to an increased risk for the development of cervical neoplasia . In addition, individuals practicing unprotected sexual activities have a combined risk of being infected by HIV and HPV because these viruses may have a common mode of acquisition. The aim of the present study was to assess the prevalence of HPV infections in a group of HIV-infected women and to investigate the presence of specific HPV genotypes with the highly sensitive and specific SPF10 HPV DNA detection and genotyping method. MATERIALS AND METHODS : Patients. | HIV-infected women attending two specialized centers in Sao Paulo, Brazil (Instituto de Infectologia Emilio Ribas and Casa da AIDS), for a routine visit to the gynecologist between December 1997 and September 1999 were invited to participate in the study. All patients provided informed consent. Patients with a CD4 cell count of less than 50 per mul were excluded from the study. Enrolled patients (n = 208) underwent a gynecological evaluation, and cervical scrape samples were collected for cytological analysis and HPV DNA PCR. A blood sample was obtained to determine the HIV viral load. A CD4 cell count was accepted for the study purpose if it had been performed within 3 months of the visit date. Otherwise, a blood sample for CD4 determination was also obtained at the visit date. One hundred and sixty-four patients (79%) were using antiretroviral agents (mainly a combination of two nucleoside analogue reverse transcriptase inhibitors and one protease inhibitor). This study was approved by the ethical committees of the participating institutions. Cytology and histology. | Pap smears were examined and classified according to the Pap classification, comprising Pap I (normal), Pap II (inflammation), Pap III (dysplasia), Pap IV (carcinoma in situ), and Pap V (carcinoma). All patients showing cytologically abnormal smears (higher than class II) were referred for colposcopy-directed biopsy. Cervical biopsy specimens were histologically examined and classified according to the cervical intraepithelial neoplasia (CIN) system. DNA isolation. | Cervical samples were collected with the brush provided in Digene's Hybrid Capture II kit (Digene Corp., Gaithersburg, Md.) and transported to the lab within 48 h in the sample transport medium. In the lab, denaturation was performed upon arrival, according to the manufacturer's instructions. Four hundred and fifty microliters were removed, and DNA was precipitated by addition of 45 mul of 3 M sodium acetate and 900 mul of 100% ethanol. Precipitated DNA was washed, dried, and resuspended in 100 mul of 0.1 M Tris-1 mM EDTA (pH 8.3), and 10 mul was used for PCR. A cervical smear was always performed before the collection of cells in hybrid capture sample transport medium. HPV DNA amplification and detection. | HPV DNA was amplified by the SPF10 PCR primer set, and each run was accompanied by several quality control samples. During each PCR run, 18 samples were tested, together with 1 negative control (water) and 1 positive control (HPV 18-containing cells). Amplification products were first tested by probe hybridization in a microtiter plate assay to detect the presence of HPV DNA as described earlier, and this assay also included appropriate negative and positive controls . SFP10-amplimers from HPV-positive samples were subsequently analyzed by reverse hybridization on the HPV reverse hybridization line probe assay (LiPA) . This assay comprises a membrane strip containing type-specific oligonucleotide probes, immobilized as parallel lines. PCR products are hybridized at high stringency to these probes, generating a type-specific hybridization pattern. The HPV-LiPA permits specific detection of 25 HPV genotypes, HPV 6, 11, 16, 18, 31, 33, 34, 35, 39, 40, 42, 43, 44, 45, 51, 52, 53, 54, 56, 58, 59, 66, 68/73, 70, and 74 . HPV 6, 11, 34, 40, 42, 43, 44, 53, 54, 70, and 74 were considered low-risk types, whereas HPV 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68 were considered high-risk types. Part of the beta-globin gene was amplified from each sample as a positive control for DNA isolation. Appropriate negative and positive controls were used to monitor the performance of the method. HIV viral load. | Blood was drawn by venipuncture and centrifuged within 6 h. Plasma was separated and frozen until processed for viral load. The HIV Monitor version 1.0 assay (Roche Molecular Systems, Sao Paulo, Brazil) was employed for viral load determination on all samples, presenting a linear detection range of 400 to 750,000 copies/ml. Statistical analyses. | Data were analyzed with the chi-square test, with a value of 0.05 considered the threshold level for significance. RESULTS : A total of 208 patients were enrolled in the study. The mean age of the patients was 32.1 years (range, 18 to 67 years). HPV DNA was detected in 204 (98%) of the 208 patients by SPF10 PCR. All negative controls and positive controls for PCR amplification, HPV DNA detection, and genotyping yielded the appropriate results. The HPV DNA-positive samples were further analyzed by genotyping on the HPV-LiPA, and results are shown in Fig. . From 100 of the patients, the SFP10 PCR products were analyzed on two independent LiPA strips, and these duplicates yielded the same typing results in all cases. In two (0.9%) patients, HPV DNA was detected, but the HPV-LiPA did not reveal a genotype. Thus, HPV genotypes were available from 202 patients. HPV 6 was the most prevalent genotype and was observed in 80 (39.2%) patients, followed by types 51 (31.9%), 11 (26.0%), 18 (24.0%), and 16 (22.5%). Overall, single genotypes were found in 43 (21.1%) and multiple genotypes were detected in 161 (78.9%) of the 204 HPV DNA-positive patients. In the HPV DNA-positive patients, an average of 3.1 types per patient were detected, ranging from 1 to 10 different genotypes per patient. The distribution of the number of different genotypes per patient is shown in Fig. . Of the 638 genotypes detected, 261 (40.9%) were low-risk genotypes. A total of 29 (14.3%) patients carried only low-risk genotypes and 43 (21.2%) carried only high-risk genotypes, whereas 130 (64.3%) patients carried both high- and low-risk types. Cervical scrapes could be cytologically classified for 199 patients and revealed that 22 (10.5%) patients showed normal (Pap I) cytology, whereas 149 (71.6%) patients showed inflammation (Pap II). Twenty-eight patients (13.4%) presented abnormal (Pap III) smears on cytology, and all of these were histologically confirmed to have a low-grade (CIN II) or/and a high-grade (CIN III) squamous intraepithelial lesion. One patient classified as Pap III was found to have a carcinoma in situ on the histological analysis. The relationship between cytological classification and the number of HPV genotypes detected is shown in Fig. . There was a clear trend for a decrease in the number of genotypes with higher Pap classification. The prevalence of multiple (>1) genotypes was higher in patients with Pap I or Pap II compared to those with Pap III (chi2 test, P = 0.02). The presence of low-risk and high-risk genotypes in the different cytological classes is shown in Table . At least one high-risk genotype was present in 89% of the Pap I samples, in 84% of the Pap II samples, and in 82% of the Pap III samples. These differences were not statistically significant (chi2 > 0.05). Among the 28 patients with squamous intraepithelial lesions, 5 (18%) carried exclusively low-risk genotypes (types 6, 6, 6 and 70, 70, and 44 and 74, respectively), 13 (46%) carried both low-risk and high-risk types, and 10 (36%) carried exclusively high-risk types. The HIV viral load and the CD4 counts were determined for all patients. Since virtually all patients were found to be HPV DNA positive, no relation between the CD4 count and the presence of HPV could be established. Also, there was no significant correlation between the number of HPV genotypes and the CD4 count or the HIV viral load . FIG. 1. | Distribution of HPV genotypes (total, 638) detected by SPF10 and LiPA in 202 HPV-positive patients. Distribution of HPV genotypes (total, 638) detected by SPF10 and LiPA in 202 HPV-positive patients. The percentage indicates the proportion of the patients in which a particular genotype was observed. FIG. 2. | Frequencies of the number of HPV genotypes observed in the patient group. Frequencies of the number of HPV genotypes observed in the patient group. FIG. 3. | Number of HPV genotypes detected per cytology class (Pap I to III). Number of HPV genotypes detected per cytology class (Pap I to III). Black bars indicate the presence of one genotype, white bars indicate the presence of two genotypes, vertically striped bars indicate the presence of three genotypes, and horizontally striped bars indicate the presence of more than three genotypes. The prevalence of multiple genotypes was higher in patients with Pap I or Pap II than in those with Pap III. This difference was statistically significant (chi2 test, P = 0.02). TABLE 1 | Distribution of high-risk and low-risk HPV genotypes in the different cytology classes TABLE 2 | Relationship between the number of HPV genotypes, cytological classification, and CD4 count and HIV viral load DISCUSSION : In the present study, we observed a very high prevalence (98%) of HPV DNA among HIV-infected women. This prevalence is higher than found in earlier studies. Palefsky et al. observed an overall prevalence of HPV DNA of 63% with the My09/11 PCR primers in HIV-infected women in the United States. Ellerbrock et al. observed a prevalence of HPV DNA of 54% in a cohort of 264 HIV-infected U.S. women . Goncalves et al. studied Brazilian HIV-positive patients and found 80.8% of them to be HPV positive, and 45% of these carried multiple HPV genotypes . The high prevalence of HPV DNA of 98% in the present study may be explained by the very high sensitivity of the SPF10 PCR primer set. It is likely that this high sensitivity permits detection of HPV DNA in samples with a low HPV viral load, which would probably be scored as HPV negative with alternative primer sets. The methods used in the present study have been extensively validated. The sensitivity of the SPF10 primer set is high, presumably also due to the small size of the amplimer, as was shown in earlier studies with various clinical materials . Furthermore, the LiPA method used for identification of HPV genotypes has been validated in different patient groups . The LiPA was shown to be highly reproducible and accurate when used with different clinical samples from the same patient or during follow-up studies . Furthermore, direct comparison between the HPV-LiPA and an alternative reverse hybridization assay showed a very high level of agreement, including cases carrying multiple HPV genotypes . In the present study, all negative and positive controls yielded the appropriate results. Taken together, it is unlikely that the results of the present study are due to lack of reliability of the methods used. There are no indications that the selection of patients in the present study was different from that in the studies cited . All 204 HPV DNA-positive patients were further analyzed by HPV genotyping, revealing a particular distribution of HPV genotypes. Low-risk HPV types (mainly HPV 6, 11, 44, and 53) were also frequently detected. Remarkably, genotypes 6 and 51 were more prevalent than HPV 16 and 18. In contrast, Goncalves et al. reported that types 16 and 18 were the most prevalent types and also found that almost 20% of the HPV were untypeable . This may be due to the restriction fragment length polymorphism-based genotyping method used. Palefsky et al. reported that HPV 53, 58, and 61 were the most prevalent types in a large HIV-infected U.S. population and that the distribution of HPV types in HIV-infected women may be different than in HIV-negative women from the same population . The behavioral and socioeconomic characteristics of HIV-infected women may differ from those of women in the normal population. Also, there may be considerable differences in the spectrum and prevalence of HPV genotypes in HIV-infected women from different geographic origins . The majority of patients in the present study carried multiple genotypes (average, 3.1 types per patient; range, 1 to 10 genotypes). This may reflect the frequent exposure of these patients to multiple HPV genotypes due to unprotected sexual contacts. In the United States, Palefsky et al. detected multiple genotypes in 36% of the HIV-infected and 12% of the HIV-negative individuals . Similarly, Ellerbrock et al. observed multiple infections in 12% of HIV-infected women . In a Brazilian population studied by Goncalves et al. , multiple genotypes were found in 45% of the HPV-positive patients. Again, these differences may be due to the genotyping method employed. Another possible explanation for the high prevalence of HPV infection could be that failure of the immune system to clear the infection increases the persistence rate of HPV . Also, HPV replication may be more efficient in an immunodeficient host, which could result in an increased detection rate as well as a higher chance of developing persistent HPV infection. This hypothesis is in agreement with findings in a recent longitudinal study showing that HPV clearance was strongly reduced in HIV-infected women . If exposure to novel infection continues, this may result in accumulation of different HPV genotypes and a higher prevalence of women infected with multiple HPV genotypes. If the frequency of multiple HPV genotypes is related to the degree of immunocompetence of the host, one could speculate that the number of HPV genotypes would be associated with CD4 counts and HIV viral load. However, the present study did not support this hypothesis, since there was no significant association between CD4 cell counts and the number of HPV genotypes. Our findings are different from the data obtained by Palefsky et al. , who found the highest number of HPV DNA-positive subjects and the highest frequency of multiple infections among patients with a CD4 count of <200/mm3. This difference may be due to the very high sensitivity of the detection and genotyping method used in the present study. Also, most HIV-infected Brazilian patients do receive some form of antiviral treatment, which may result in higher CD4 counts (as long as no resistance appears) but may not yield an immediate improvement in the ability of the immune system to clear HPV infections or reduce the viral load. Of the cases without cytological abnormalities, the majority (71.6%) had inflammation (Pap II). This is in agreement with earlier observations in Brazilian HIV-infected patients, where 75.9% of the samples showed inflammation . The number of HPV genotypes appears to be lower in patients with Pap III cytology. The majority of patients with squamous intraepithelial lesions in this population harbored one or two HPV types, whereas only 2% of these patients were infected with more than three HPV types . Conversely, patients with Pap I or II carried significantly more multiple types. These observations would be consistent with the hypothesis that the development of cervical dysplasia has a clonal basis. In conclusion, a very high proportion of HIV-infected women in this study were found positive for HPV DNA. In contrast to immunocompetent women, infection with multiple HPV types is present in the majority of the HIV-infected patients. However, the number of HPV genotypes was not associated with HIV disease markers such as HIV plasma viral load and CD4 cell counts. The findings of the present study are currently being reanalyzed with another sensitive HPV DNA detection method. Preliminary results confirmed both the extremely high prevalence of HPV and the presence of multiple genotypes in many of these patients (J. E. Levi, personal communication). Investigation of HIV-infected patients is of particular value because they show, as demonstrated here, a much higher multiplicity of infection than other patient groups and may manifest the effects of HPV infection earlier and more intensively. Therefore, sensitive detection methods as well as specific genotyping tools are required and may have important consequences for the management of immunocompromised patients. Backmatter: PMID- 12202564 TI - Effects of Various Test Media on the Activities of 21 Antimicrobial Agents against Haemophilus influenzae AB - As considerable variation in the antimicrobial susceptibility of Haemophilus influenzae has been reported, the effects of various test media on the susceptibility of H. influenzae were studied. MICs were determined by three laboratories for 21 antimicrobial agents against a panel of 100 selected isolates. Testing was performed using a reference NCCLS frozen broth microdilution method with Haemophilus test medium (HTM) broth and dried commercial MIC trays rehydrated with the following media: in-house and commercially prepared HTM broth, Mueller-Hinton broth with 2% lysed horse blood and NAD, IsoSensitest broth with 2% lysed horse blood and NAD, and IsoSensitest broth-based HTM. Overall, all results were very reproducible, with the MIC at which 50% of the isolates tested are inhibited (MIC50), MIC90, and geometric mean MIC being within one doubling dilution by all six methods and at all three testing centers for 15 of the 21 agents tested. Interlaboratory differences were more marked than intralaboratory differences or differences among media. Cefprozil, cefaclor, and trimethoprim-sulfamethoxazole results differed the most, while results for ampicillin, amoxicillin-clavulanic acid, cefdinir, cefixime, ceftriaxone, and clarithromycin were the most reproducible. However, these variations in results caused considerable differences in susceptibility rates for agents for which NCCLS susceptible breakpoints were close to the geometric mean MIC, particularly for cefaclor and cefprozil. This was much less of a problem when pharmacokinetic-pharmacodynamic breakpoints were used. Reproducible susceptibility results were obtained for a wide range of agents against H. influenzae in three laboratories using a variety of media that support the growth of this fastidious species. Keywords: Introduction : Assessment of the susceptibility of Haemophilus influenzae to various antimicrobial agents is relatively easy for agents for which defined resistance mechanisms are known and which result in high-level resistance, such as the activities of ampicillin and amoxicillin against beta-lactamase-producing strains compared to those against beta-lactamase-negative strains . When bimodal MIC populations are found, susceptible and resistant strains can be readily differentiated, particularly if the MICs of susceptible strains are below clinically achievable levels of the agents. However, this is not the case for many agents for which unimodal MIC distributions are found and strains against which the MICs are higher are rare or absent . This circumstance is further complicated by the fact that MIC ranges may include values that are close to clinically achievable levels and/or include MIC breakpoint concentrations . Examples of such agents include amoxicillin-clavulanate, cefuroxime, azithromycin, erythromycin, and clarithromycin. For the two beta-lactam examples, the situation is further complicated by the fact that strains for which the MICs are at the high end of the distribution may have non-beta-lactamase-mediated resistance due to altered penicillin-binding proteins or may be spheroplast-producing strains . Considerable variation in the resistance rates of various agents against H. influenzae has been reported . For example, the resistance rates for amoxicillin-clavulanate varied from 0% in a large international study of 2,718 strains to 4.5% in a U.S. study of 1,539 strains . Similarly, the resistance rates for cefuroxime varied from 3.2 to 6.4%. Although these resistance rates are low and differences between studies are not very large, the susceptible breakpoint for these agents (<=4 mug/ml) is close to the MICs of these agents at which 90% of the isolates tested are inhibited (MIC90s) (1 to 8 mug/ml), and normal population distributions could account for some MICs being in the resistant range . This study examined the effects of various media on the susceptibility of this species to a wide range of antimicrobial agents, using media that would be acceptable in a large number of countries. The results are compared with those of the reference NCCLS method using Haemophilus test medium (HTM) in frozen microdilution trays . MATERIALS AND METHODS : Isolates. | The isolates tested were 100 untypeable H. influenzae isolates from stock cultures which have been used for the validation of MIC plates from several surveillance studies. Of the 100 isolates, 43 were beta-lactamase positive and 4 were beta-lactamase negative and ampicillin resistant. Antimicrobial agents. | A total of 21 antimicrobial agents were tested, consisting of 9 beta-lactam agents (ampicillin, amoxicillin, amoxicillin-clavulanic acid, ceftriaxone, cefuroxime, cefaclor, cefixime, cefprozil, and cefdinir), 3 macrolide agents (azithromycin, clarithromycin, and erythromycin), 6 fluoroquinolone agents (ciprofloxacin, gemifloxacin, grepafloxacin, levofloxacin, ofloxacin and trovafloxacin), chloramphenicol, doxycycline, and trimethoprim-sulfamethoxazole. MIC methods. | All strains were tested with frozen trays utilizing the NCCLS broth microdilution method with in-house-prepared HTM (frozen MH HTM [see below]) and with dried commercial MIC trays (Trek Diagnostics, Westlake, Ohio) rehydrated with five different media. Three different testing sites participated in the study. Frozen MH HTM was tested at Case Western Reserve University, Cleveland, Ohio, while the dried commercial trays were tested at all three sites. For the dried commercial trays, all of the sites used the same batch of in-house HTM (MH HTM), which was prepared by the investigators at Case Western Reserve University using Mueller-Hinton broth base (Difco), 0.5% yeast extract (Difco), 15 mug of NAD per ml, and 15 mug of hematin (Sigma) per ml. Case Western Reserve University (site 1) and M. S. Hershey Medical Center, Hershey, Pa. (site 2) also tested the strains with Mueller-Hinton broth (Trek Diagnostics) supplemented with 2% lysed horse blood (Cleveland Scientific, Bath, Ohio) and 15 mug of NAD/ml (MH LHB NAD). In addition, site 1 tested the strains with a commercial formulation of HTM broth (PML, Tualatin, Oreg.) (PML HTM). GR Micro, London, United Kingdom (site 3) also used IsoSensitest Broth (Oxoid Ltd., Basingstoke, United Kingdom) supplemented with 2% lysed horse blood (TCS Microbiology, Botolph Claydon, United Kingdom) and 15 mug of NAD (Sigma-Aldrich UK Ltd., Poole, United Kingdom)/ml (IST LHB NAD) and IsoSensitest broth-based HTM (IsoSensitest broth supplemented with 15 mug of NAD per ml, 0.5% yeast extract [Oxoid Ltd.], and 15 mug of hematin [Oxoid Ltd.] per ml) (IST HTM). Inoculum checks were performed on all strains, and quality control strains (H. influenzae ATCC 49247 and 49766 and Escherichia coli ATCC 35218) were included in each testing run. The results were accepted if inocula were in the range of 3 x 105 to 7 x 105 CFU/ml and the MICs for the quality control strains were within published limits . Data analysis. | Geometric mean MICs, MIC50s, MIC90s, and standard deviations for each method, based on the doubling-dilution (i.e., log2) values, were determined for antimicrobial agents with unimodal distributions and on-scale endpoints for >90% of the strains. Ampicillin, amoxicillin, and trimethoprim-sulfamethoxazole means were not calculated because of bimodal distributions. Results for all methods were compared to the results for frozen MH HTM, and doubling-dilution differences were calculated. Susceptibility rates were calculated based on both NCCLS and pharmacokinetic-pharmacodynamic (PK-PD) breakpoints . The PK-PD breakpoints were based on standard dosing regimens and criteria appropriate to each agent. For beta-lactams, erythromycin, and clarithromycin, these breakpoints were based on drug concentrations in serum present for 40 to 50% of the dosing interval, while for azithromycin, fluoroquinolones, and doxycycline, they were based on 24-h area under the concentration-time curve/MIC ratios exceeding 25 . For trimethoprim-sulfamethoxazole, the NCCLS breakpoint was used, as the PK-PD breakpoint was not available and the NCCLS breakpoint had been validated in bacteriologic outcome otitis media studies . The data were also analyzed statistically using several approaches, with the frozen MH HTM as the reference method where applicable. First, the data were examined with generalized estimating equations (GEE) (SAS, Cary, N.C.), a method of multiple regression analysis of the entire data set, which was used to assess overall differences between test sites and methods. GEE methodology was chosen because it could handle the lack of independence among the observations caused by the same isolates being analyzed multiple times. In effect, each isolate created a family of related observations. Since there will be more variability in MICs across isolates than across methods for a particular isolate, it is important to use analysis methods that recognize the relationship of the observations. The GEE methodology can be used to simultaneously examine the effects of method and sites, obtaining an estimated regression coefficient for each individual effect variable. Since all the effects were categorical variables, each effect would be measured as a difference from a reference method and site. Second, sequential paired t tests were performed to compare data sets, using the frozen MH HTM as the reference method. This analysis was repeated using the site 1 MH HTM as the standard. Since the MH HTM was prepared at site 1 and distributed to each of the other testing sites, it provided a standard for measuring both methodological and within-site differences. Consequently, paired-t-test analysis was also used to examine within-laboratory differences using each site's MH HTM results as the reference method. To account for potential type 1 error resulting from the multiple t tests employed, a Bonferroni correction was used, requiring a critical P value of <0.0003 for statistical significance. Third, the data were examined for overall mean population differences using the t distribution, since the test sample population MICs were normally distributed with similar variances. While this method is not appropriate for testing dependent data, it was thought to provide some useful information, since population mean data are used in microbial surveillance studies and repeated testing of the same sample would be expected to regress to the mean. The laboratory methods examined in this study are those that would likely be used in large-scale surveillance. The critical value used for this testing was that corresponding to a P value of <0.0003. The robustness of some of the statistical differences found was tested by adjusting the data by +-0.25-, 0.5-, 0.75-, and 1-doubling-dilution differences. Since a one-doubling-dilution difference is considered an acceptable level of sample reproducibility in MIC testing, this analysis was performed to determine whether adjusting the data incrementally within the one-dilution range would effect statistical significance . The percentages of MICs within one and two dilutions of the frozen reference values were calculated. The differences between sites and methods were examined by the binomial test for proportions for both one- and two-dilution differences. The differences in agreement with the frozen reference values were examined directly for each site-method combination, as well as for methods within each site, with the site-specific MH HTM difference from frozen reference values used as the baseline for these comparisons. RESULTS : The geometric mean MICs were within one doubling dilution by all six methods and at all three testing centers for all antimicrobial agents except azithromycin, cefaclor, cefprozil, gemifloxacin, grepafloxacin, and levofloxacin . At least 90% of MICs by all six methods and at all three testing centers were within two doubling dilutions compared with the frozen reference except for amoxicillin with IST HTM at site 3; cefuroxime and ampicillin with IST LHB NAD at site 3; cefaclor and cefprozil with in-house HTM at site 2, MH LHB NAD at site 2, IST HTM at site 3, and IST LHB NAD at site 3; gemifloxacin with MH LHB NAD at sites 1 and 2; and trimethoprim-sulfamethoxazole with in-house HTM at site 3, PML HTM at site 1, MH LHB NAD at site 2, and IST LHB NAD at site 3 . An analysis of the differences by antimicrobial agent demonstrates that the cefprozil, cefaclor, and trimethoprim-sulfamethoxazole results differed the most. The cefaclor and cefprozil results for sites 2 and 3 were one to two dilutions lower than those for site 1. The trimethoprim-sulfamethoxazole results were not reproducible within each laboratory at MICs of <0.5 mug/ml. The results for erythromycin and azithromycin for all media were one to two dilutions lower than those for frozen MH HTM; this was seen more frequently with erythromycin than with azithromycin. Clarithromycin agreement with the frozen reference method was very good for all media and only showed very slight differences from the frozen reference method (MICs in dried trays were lower than those in frozen MH HTM trays). The fluoroquinolone results (with the exception of ciprofloxacin and ofloxacin) for all media in the dried trays were lower (one to two dilutions) than those in the frozen MH HTM trays. The majority of fluoroquinolone results were at very low MICs of <0.03 mug/ml, which may explain the variation in results. The gemifloxacin and grepafloxacin MH LHB NAD results at sites 1 and 2 were lower by one dilution than those of HTM- and IsoSensitest-based media. The results for ampicillin, amoxicillin-clavulanic acid, cefdinir, cefixime, ceftriaxone, ciprofloxacin, clarithromycin, and ofloxacin were the most reproducible. Less variation in mean MICs was found for all methods for amoxicillin-clavulanic acid (0.7 to 1.2 mug/ml), which was one of the most reproducible agents tested, than for cefprozil (2.7 to 9.5 mug/ml), which was one of the least reproducible agents tested. Less variation in ranges of mean +- 2 standard error values (approximate 95% confidence limits) was also noted for amoxicillin-clavulanic acid (approximately 1.5 doubling dilutions) than for cefprozil (approximately 2.5 doubling dilutions). Figure provides a graphical example of these better and worse degrees of agreement, with confidence intervals included. Despite the differences in ranges of variation, the patterns of variability are strikingly similar. In addition, it is possible to see more clearly the similarities within sites and the differences among them. In addition to the differences with gemifloxacin and grepafloxacin in MH LHB NAD, the only other discernible medium differences were with amoxicillin and doxycycline for IsoSensitest-based medium results. Amoxicillin IST HTM varied most compared with other media (+6 to -4 dilution difference from the frozen reference). The amoxicillin IST HTM results were lower by one to two dilutions at MICs above 8 mug/ml. Doxycycline IST HTM results were one to two dilutions higher than other medium results, and the doxycycline IST LHB NAD results were also slightly higher by approximately one dilution. Overall, the results for different media were comparable within each testing laboratory. The results for site 2 and site 3 were slightly lower overall than those for site 1. GEE methodology examined the effects of testing sites and methods compared to the baseline of site 1 MH HTM, providing an estimated regression coefficient for each individual effect variable. This coefficient, stated as the number of doubling dilutions from the baseline, describes the degree of shift from the baseline caused by the specified variable. These coefficients and the statistical significances of the differences are summarized in Table . Site 2 and PML HTM were both significant contributors to differences seen with cefaclor and cefprozil, with shifts in MICs of greater than one doubling dilution for both drugs. The site 2 coefficient was greater than one dilution lower than the frozen standard for both drugs, while the PML HTM coefficient shifted the MICs greater than one dilution higher than the standard for both drugs. The quinolones (except for gemifloxacin and ciprofloxacin) and macrolides (except for clarithromycin) were also shifted nearly one dilution lower with the MH HTM than with the frozen medium. Statistically significant (P < 0.05) differences of +-0.25 dilutions are indicated in the table. Table summarizes the paired t test findings from the comparison within sites, using the MH HTM values as a baseline. Within site 1, the PML HTM medium was highly statistically different for azithromycin and also significantly different for cefprozil, chloramphenicol, clarithromycin, erythromycin, and gemifloxacin. The MH LHB medium was highly different for gemifloxacin and grepafloxacin and significantly different for cefaclor, clarithromycin, and trovafloxacin. Within site 2, MH LHB NAD was in good statistical agreement with MH HTM only for levofloxacin. Agreement for ceftriaxone and clarithromycin were borderline not significant, and that for ofloxacin and doxycycline was close to the level of statistical significance with the Bonferroni correction. All other agents were significantly or highly significantly different between the two methods used at site 2. Site 3 IST HTM was highly significantly different from MH HTM for the quinolones, doxycycline, and ciprofloxacin and significantly different for amoxicillin-clavulanic acid. All other agents tested with IST HTM at site 3 were not significantly different from the baseline. The site 3 IST LHB showed significance for only doxycycline, ofloxacin, levofloxacin, and cefdinir compared to MH HTM. Tables and summarize the statistical findings for the between-site comparisons of like media. Site 2 MICs for the MH HTM medium used at all three sites were significantly different from those seen at site 1, except for the quinolones. Site 3 showed some differences from site 1 for cefaclor, cefprozil, and cefuroxime. Compared to each other, sites 2 and 3 differed for cefaclor, cefdinir, clarithromycin, doxycycline, and erythromycin. When LHB media were compared, site 2 again showed greater differences from site 1 than did site 3. Compared to each other, sites 2 and 3 also showed several significant differences. When tested as a population rather than as discretely paired data, the groups showed greater statistical agreement (data not shown). Compared to the frozen HTM, the quinolones did not show the statistically significant differences that were seen when the data were paired, though the macrolides, cefaclor, and cefprozil remained statistically different. The comparison done within sites showed the best agreement at site 3, where only cefprozil was statistically different using the IST LHB NAD medium and cefaclor and doxycycline were significantly different using the IST HTM medium. At site 1 the PML HTM and MH LHB NAD were significantly different from the baseline MH HTM for cefaclor, cefprozil, and clarithromycin, and only the PML HTM was significantly different for erythromycin. At site 2, MH LHB NAD was significantly different from MH HTM for cefaclor, cefprozil, clarithromycin, and erythromycin. The MIC adjustments done on azithromycin, erythromycin, grepafloxacin, levofloxacin, ofloxacin, and trovafloxacin to test for the robustness of the statistical differences required no more than a one-doubling-dilution adjustment to change from a very high degree of statistical significance to none (data not shown). The MIC distributions of these drugs also show a left shift in the normally distributed curve of approximately one doubling dilution. Table summarizes the percentage of MICs found to be within one and two dilutions of the frozen standard. Many statistically significant differences were found when the site-method combinations were compared to the MH HTM site 1 baseline. When comparisons were made within each laboratory, using the site-specific MH HTM difference from the frozen standard as a baseline, fewer significant differences were seen. At site 1, only 4 of the 19 antimicrobial agents tested with PML HTM broth ---azithromycin, erythromycin, gemifloxacin, and trovafloxicin ---were statistically different from the baseline at one dilution difference of agreement. At agreement to two dilutions difference, only amoxicillin and azithromycin were significantly different from the baseline for PML HTM. At site 2, there were a larger number of drugs that did not agree to within one or two dilutions difference for the two methods tested. FIG. 1. | Geometric mean MICs and upper and lower limits of 95% confidence limits (bars) for amoxicillin-clavulanic acid and cefprozil. Geometric mean MICs and upper and lower limits of 95% confidence limits (bars) for amoxicillin-clavulanic acid and cefprozil. Much less variation in mean MICs was found for amoxicillin-clavulanic acid ranges (0.7 to 1.4 mug/ml) for all methods and sites than for cefprozil (2.0 to 9.5 mug/ml). However, similar variations in the ranges of 95% confidence limits were noted for amoxicillin-clavulanate and cefprozil. TABLE 1 | Geometric mean MICs and standard deviations (doubling dilutions) for each method for agents with unimodal populations and >90% of endpoints on scale TABLE 2 | GEE results; estimated regression coefficients and statistical values of significance for comparison to standard TABLE 3 | Paired t test P value results within sites, with results of MH HTM for each site as standard TABLE 4 | Paired t test P value results of MH HTM results between sites TABLE 5 | Paired t test P value results for LHB-supplemented media TABLE 6 | Percentages of MICs within one or two dilutions from reference method based on test site and medium DISCUSSION : The results of this study indicate that MICs of a wide variety of antimicrobial agents for H. influenzae can be reproducibly determined using several media. Most variations in MICs were due to site-to-site differences and not due to differences in media. The site-to-site differences, however, did not significantly impact the susceptibility rates for the majority of the antimicrobial agents tested, as many of the differences occurred at drug concentrations below susceptibility breakpoints (Tables and ). However, considerable variation in the percentages of isolates susceptible to cefprozil and cefaclor occurred, as the geometric mean MICs were close to NCCLS susceptible breakpoints. Less variation occurred when PK-PD breakpoints were used. There were few significant differences as a result of the compositions of media. The results obtained using Mueller-Hinton broth with 2% LHB and 15 mug of NAD/ml and IST with 2% LHB and 15 mug of NAD/ml were equivalent to the results obtained using HTM for determination of the MICs of agents evaluated against H. influenzae. However, HTM is more difficult to manufacture and has a short shelf life at 4C (4 to 6 weeks) before the hematin component degrades, whereas LHB-containing media can be kept for at least 6 months at 4C; both media are stable at -20C (-). Additional advantages of LHB-based media are that endpoints are easier to read and that the same medium can be used to test streptococci. A necessary consideration when examining data with a permissible range of reproducibility difference statistically is at what point a conventionally significant result is scientifically meaningful. MIC tests are generally regarded as being in agreement when they are reproducible within one doubling dilution, while for statistical tests, such a difference is generally considered highly significant; there is consequently an inherent tension between what is statistically significant and what is scientifically significant. For the purposes of this study, a shift of 0.25-doubling-dilution difference was regarded as representing a scientifically noteworthy methodological or site bias. There appears to be a greater degree of difference between testing sites than among methods within one site. Sites 1 and 3, for the most part, had good reproducibility across methods, while site 2 showed more variability. This is especially apparent when looking at the GEE results, where site 2 is more frequently a contributor of significant difference than the other two sites. Site 2 shows a consistent bias in the negative direction, suggesting that the MICs for all drugs tested at that site would be consistently lower than those found at the other sites. Examination of method-specific differences shows only PML HTM to be a significant contributor of difference. In contrast to the site 2 bias, that seen with this medium is in the positive direction. MICs with PML HTM broth would be expected to be higher than those found with the other methods. Examination of the MICs of many of the antimicrobial agents tested by paired t test remained consistent across both methods and testing sites. The quinolones, while very statistically different in the comparison between MH HTM and the frozen standard, did not differ much in the between-site analysis, the GEE analysis, or when tested as population means. Because these drugs are active at very low concentrations, a difference of one dilution is much less important than for drugs with activity at higher concentrations and therefore closer to or crossing breakpoints. The data adjustment to test robustness and the resulting loss of statistical significance reinforced the data summarized in Table , showing that the groups, while statistically different from each other, largely agree within the acceptable range of differences for antimicrobial MIC testing of 95% within one doubling dilution . The results of these tests, as well as the examination of differences in group means, suggest that there is a systematic shift of less than one dilution difference in several of the drugs. While this is within the acceptable range for reproducibility, it could be problematic if the shift occurs at or near a breakpoint. TABLE 7 | MIC50 and MIC90 for each method TABLE 8 | Ranges and mean percentages of susceptible isolates for all methods based on NCCLS and PK-PD breakpoints Backmatter: PMID- 12202546 TI - Detection of Human Papillomavirus DNA in Urine Specimens from Human Immunodeficiency Virus-Positive Women AB - Human immunodeficiency virus (HIV)-positive women may represent one of the fastest-growing populations at risk for acquiring cervical cancer and thus require frequent screening. The purpose of the present studies was to validate a PCR-based urine assay by comparing detection and genotyping of human papillomavirus (HPV) DNA in urine samples and matching cervical swab specimens of HIV-positive women. Despite a difference in amplifiability, the prevalence of any HPV genotype (58% for the cervical swab specimens and 48% for the urine specimens) was not significantly different in this population. The levels of concordance were 70, 71, and 78% for detection of any HPV type, any high-risk HPV type, or any low-risk HPV type in the two specimen types, respectively. While instances of discordant detection were greater for the cervical swab specimens than for the urine specimens, this was not statistically significant. The distributions of HPV genotypes were similar in the cervix and the urine for the majority of types examined. Importantly, detection of HPV DNA in urine was associated with an abnormal Papanicolaou smear to the same extent that detection of HPV DNA in a cervical swab specimen was. These data provide preliminary support for the proposal to use urine testing as a primary or secondary screening tool for cervical cancer in HIV-positive women or as an epidemiological tool. Additional studies with larger sample sizes must be conducted in order to further verify these findings. Keywords: Introduction : Human papillomavirus (HPV) is the cause of the most common sexually transmitted diseases (STDs) of viral etiology worldwide. Approximately 30 to 50% of the general population is positive for HPV DNA . At least 30 known HPV types infect the anogenital region (; Los Alamos National Laboratory HPV Sequence Database , 2001). These types are classified as being of low, intermediate, or high risk on the basis of their in vitro abilities to cause cellular transformation and their clinical association with cervical cancer. High-risk HPV types, like HPV types 16, 18, 31, and 45, are more closely associated with anogenital malignancies and have been implicated in the etiology of most, if not all, cervical cancers. Cervical cancer is the second most common cancer of women worldwide . Importantly, infection with HPV, even high-risk types, is asymptomatic in most people and usually does not lead to cancer. However, more than 35 HPV types have been found to be associated with at least 90% of cases of cervical intraepithelial neoplasia, which is a precursor lesion to cervical cancer . Although the medical standard for the diagnosis of HPV disease is the Papanicolaou (Pap) smear, screening by this type of method has inherent problems. Only 15 to 50% of patients with HPV infections are accurately identified by Pap smears . Additionally, the efficacy of screening by use of the Pap smear relies on repeated (yearly) visits. Approximately 10% of women in the United States have never had a Pap smear, and about 30% of women do not have them on a regular basis . Furthermore, Pap smear screening requires a pelvic examination, which is invasive and uncomfortable for the patient as well as time-consuming for the health care provider . Perhaps the use of urine sampling for routine detection of HPV could provide a preliminary screen for cervical cancer and thus circumvent the need for an annual Pap smear for women who are negative for HPV DNA. Alternatively, detection of HPV DNA in urine could possibly function as a secondary screening technique for cervical cancer in that it could be used to triage women with atypical squamous cells of undetermined significance (ASCUS). Methods for the detection of DNA in patient urine have recently been used to diagnose other common STDs that affect the cervix, such as Chlamydia trachomatis and Neisseria gonorrhoeae infections . The success of these screening programs provides a good preliminary endorsement of attempts to monitor cervical pathology by detecting HPV DNA in urine specimens. One advantage to the detection of these organisms and the detection of HPV from urine specimens may be that a single specimen could perhaps be used to detect all three of these infectious agents simultaneously. Several investigators have attempted to use urine sampling for HPV DNA detection ; however, no study to date has addressed the ability to detect HPV DNA in the urine of human immunodeficiency virus (HIV)-positive women. HIV-positive women demonstrate increased rates of new HPV infections, persistence of high-risk HPV infections, multiple infections, dysplasia, and cervical cancer . Furthermore, once cervical cancer develops in these women, the disease tends to be more aggressive and less responsive to treatment . In consideration of these factors, the importance of cervical cancer screening in this population is evident. The current strategy for baseline screening of HIV-positive women with normal cervical cytology is to repeat Pap smear testing every 6 months . The detection of HPV DNA in the urine of HIV-positive women could identify women at risk for HPV disease in a routine manner that is less invasive to the patient. The purpose of the studies described here was to validate the urine-based HPV DNA detection method that was developed in our laboratory by comparing the results obtained by that method to those obtained by detection of HPV DNA from matching cervical swab specimens for a population of HIV-positive women for whom the ability to access cervical pathology due to HPV infection by urine testing has not been described previously. MATERIALS AND METHODS : Patient population. | All patients included in the study were enrolled in the HIV Outpatient Clinic at the Medical Center of Louisiana in New Orleans, La. The study included 101 patients with urine samples and matching cervical swab samples. At the time of their scheduled visit, which included a routine Pap smear, a cervical swab specimen and a non-clean-catch urine specimen were collected. Although the specimen collection sequence was not strictly enforced, the urine was usually collected prior to the pelvic examination. All study participants signed informed consents, and the study was approved by the Louisiana State University Health Sciences Center Institutional Review Board in compliance with federally regulated guidelines for research involving human subjects. Both cervical swab and urine specimens from all 101 patients were tested for the presence of HPV DNA regardless of Pap smear status. Demographic data were collected by questionnaire and from the Adult Spectrum of Disease database. Cervical swab extractions. | One cervical swab specimen was collected from each patient by use of a Dacron swab; the swab was then placed in 1 ml of Digene transport medium and stored at 4C until use. Sample extraction was based on the protocol by Ting and Manos . The samples were digested at 37C in 50 mM proteinase K digestion solution (Gibco BRL, Gaithersburg, Md.) for 1 to 2 h. The DNA was then precipitated with a solution of 3 M ammonium acetate in 100% ethanol at twice the volume of the original sample. The DNA pellets were resuspended in 50 mul of TE buffer (10 mM Tris-HCl [pH 7.4], 1 mM EDTA [pH 8.0]) and stored at -20C until use. Extraction of DNA from urine. | All urine specimens collected were stored at -20C for a minimum of 24 h. Two milliliters of each urine sample was then concentrated in an Amicon 100 protein concentration filter (Millipore Corporation, Bedford, Mass.) by centrifugation at 2,087 x g for 30 to 45 min at room temperature, according to the instructions of the manufacturer. The resulting filtrate was then resuspended in 200 mul of 1x phosphate-buffered saline (0.2 M phosphate, 1.5 M sodium chloride [pH 7.4]), and the DNA was extracted with a Qiagen DNA Mini kit (Qiagen Inc. Valencia, Calif.). These DNA extractions were performed according to the manufacturer's Blood and Body Fluid Spin Protocol and were stored at -20C until use. PCR. | Urine and matching cervical swab specimens were subjected to 40 cycles of PCR for amplification of the L1 open reading frame of HPV by using the PGMY09-PGMY11 consensus primer system labeled with biotin (final concentration, 1 muM) . The biotin-labeled GH20-PC04 (beta-globin) primer system (final concentration, 0.025 muM) was also included in the reaction mixture. MgCl2 was used at a concentration of 4 mM, along with 1x PCR buffer II, 1.5 U of AmpliTaq Gold DNA polymerase (Perkin Elmer), and 1 muM deoxynucleoside triphosphates (Perkin-Elmer). The DNA types of the HPV isolates from cervical swab specimens and the matching urine samples were determined by using the reverse line hybridization system by the protocol developed by Gravitt et al. . All reagents were provided by Roche Molecular Systems, Alameda, Calif. Briefly, probes for the L1 regions of the 27 HPV genital types were conjugated to a long strip of nylon membrane with bovine serum albumin. Biotinylated PCR products were denatured by adding EDTA (1.6%), sodium hydroxide (0.13 N), and thymol blue dye for 1 h at room temperature. Next, the samples were hybridized to the probes on the strips by using hybridization buffer containing 4x SSPE (1x SSPE is 0.18 M NaCl, 10 mM NaH2PO4, and 1 mM EDTA [pH 7.4]) and 0.1% sodium dodecyl sulfate for 30 min at 53C. The samples were washed once with 1x SSPE-0.1% sodium dodecyl sulfate at 53C and were then incubated for 0.5 h at room temperature with streptavidin-horseradish peroxidase conjugate solution. The strips were then washed twice at room temperature for 10 min. The pH of the samples was then lowered by the brief addition of sodium citrate buffer (3 ml; 0.1 M). Lastly, a developing solution (2 parts of part A [a citrate solution containing 0.01% H2O2] to 1 part of part B [0.1% 3,3',5,5'-tetramethylbenzidine]) was added, and the strips were developed for 5 to 10 min. Finally, sodium citrate buffer (3 ml; 0.1 M) was added to stop the reaction. Data analysis. | Statistical significance was determined with the SPSS program (version 9; SPSS Inc., Chicago, Ill.) for the McNemar test, the Fisher exact test, likelihood, and odds ratios. The Epi6 program (Centers for Disease Control and Prevention, Atlanta, Ga., and World Health Organization, Geneva, Switzerland) was used to determine P values generated from proportion tables. RESULTS : The study population consisted of 101 HIV-positive women recruited from the HIV Outpatient Clinic in New Orleans, La., who donated urine specimens and matching cervical swab specimens. The demographics of patients from the HIV Outpatient Clinic as well as the women in the urine study are presented in Table . There were no obvious differences in the characteristics of the urine study population compared to those of the HIV Outpatient Clinic population except that the urine study contained a significantly higher percentage of women in the 30- to 39-year-old age group (P = 0.05) . The ability to amplify the samples was determined by visualization of the beta-globin housekeeping gene by the reverse line hybridization assay. beta-globin DNA was amplifiable from 87% of the cervical swab specimens and 73% of the urine specimens (P = 0.01). Thus, there was a significant difference in the ability to amplify DNA from urine specimens compared to the ability to amplify DNA from cervical swab specimens. However, the detection of any HPV type, any high-risk HPV type, and any low-risk HPV type between the two types of samples was not significantly different for this population (P = 0.12, 0.26, and 0.18, respectively) . Interestingly, detection of HPV between the two sample types was better when one considered only urine samples (n = 74) from which beta-globin DNA could be amplified or only cervical swab specimens (n = 88) from which beta-globin DNA could be amplified. In this analysis, 48 of 74 (65%) urine samples and 59 of 88 (67%) cervical swab specimens contained any HPV type (P = 0.77); 45 of 74 (61%) urine samples from which beta-globin DNA could be amplified contained any high-risk HPV DNA type, while 52 of 88 (59%) cervical swab specimens from which beta-globin DNA could be amplified contained high-risk HPV DNA types (P = 0.82). Twenty of 74 (27%) urine samples from which beta-globin DNA could be amplified were positive for low-risk HPV DNA types, whereas 28 of 88 (32%) cervical swab specimens were positive for low-risk HPV DNA types (P = 0.50) (data not shown). Lastly, Fig. demonstrates that 27 of 101 (27%) urine specimens and 35 of 101 (35%) cervical swab specimens were infected with more than one HPV type and 21 of 101 (21%) urine specimens and 26 of 101 (26%) cervical swab specimens were infected with more than one of any of the high-risk HPV DNA types. Of the samples infected with multiple HPV DNA types, the average number of low-risk and high-risk HPV types detected in each sample were 3.4 (range, 2 to 4) and 3.2 (range, 2 to 6), respectively, for the urine specimens and 4.1 (range, 2 to 4) and 3.1 (range, 2 to 8), respectively, for the cervical swab specimens. The reverse line hybridization assay detects 27 different types of HPV (18 high-risk types and 9 low-risk types) . None of the HPV DNA detected in this study was nontypeable (agarose gel electrophoresis positive, reverse line hybridization assay negative). When the overall distribution of HPV types detected in the urine specimens was compared with the overall distribution of HPV types detected in the cervical swab specimens, virtually no differences were observed for 21 of the 27 types represented. For types 53, MM4, MM7, 54, and 55, discordance was detected more frequently for the cervical swab specimens (incidence in cervical swab specimens versus incidence in urine specimens, >2.0). For type MM9, discordance was detected more frequently for the urine specimens (incidence in urine specimens versus incidence in cervical swab specimens, >2.0) . However, these comparisons were not statistically significantly different (P >= 0.11 for all comparisons). It is possible that the relative insensitivity of the assay with urine for the detection of cervical HPV types 53, MM4, MM7, 54, and 55 may become a potential limitation of the assay. Concordance is a measure of whether both samples had any HPV type, any high-risk HPV type, or any low-risk HPV type or whether no HPV was present in the paired samples. The rates of concordance were 70% for any HPV type, 71% for any high-risk HPV type, and 78% for any low-risk HPV type. The rates at which HPV occurred more frequently in the cervix than in the urine are as follows for any HPV type, any high-risk HPV type, or any low-risk HPV type: 21 versus 9% (P = 0.04), 19 versus 11% (P = 0.20), and 15 versus 7% (P = 0.13), respectively (data not shown). Although the test with urine specimens detected fewer instances of discordance for any HPV type, it was not statistically significantly different from the instances of discordant results for cervical swab specimens for the detection of any high- or low-risk HPV type in this analysis. Table illustrates the risk of detection of HPV DNA in urine and the cervix in association with an abnormal Pap smear. Of the 97 patients in the study for whom Pap smear data were available, 53 had Pap smears within normal limits (WNL), 17 had Pap smear results indicating ASCUS, and 27 patients had Pap smears graded low-grade squamous intraepithelial lesion (LGSIL) or high-grade squamous intraepithelial lesion (HGSIL) . There were no significant differences between urine and the cervix in terms of the percentage of any HPV type or a high-risk HPV type of the different cytologic grades detected . Figure also illustrates that for women with normal Pap smears, the rate of detection of any HPV type was 15% higher in the cervix than in urine. The ages, races, CD4 T-cell counts, and HIV loads of the patients included in the urine study were examined and were found by multivariate analysis not to be independent risk factors for having either an abnormal Pap smear or an abnormal Pap smear including ASCUS. When these factors were adjusted for, the detection of any HPV type, any high-risk HPV type, multiple HPV types, or multiple high-risk HPV types was significantly associated with having an abnormal Pap smear or an abnormal Pap smear including ASCUS for both urine and cervical swab specimens . Incidentally, neither HPV DNA nor beta-globin DNA could be amplified from 27 urine specimens and 13 cervical swab specimens. The Pap smear results for these patients were as follows: for urine specimens, 3 of 27 (11%) patients did not have Pap smear data available, 16 of 27 (59%) patients had WNL, 4 of 27 (15%) patients had ASCUS, and 4 of 27 (15%) patients had LGSIL; for the cervical swab specimens, 6 of 13 (46%) patients had WNL, 3 of 13 (23%) patients had ASCUS, 3 of 13 (23%) patients had LGSIL, and 1 of 13 (7%) patients had HGSIL. There was no further investigation as to why neither HPV DNA nor beta-globin DNA could be amplified from these samples by the PCR assay in this study. FIG. 1. | Prevalence of any HPV type, any high-risk HPV type, any low-risk HPV DNA type, or the presence of a multiple infection in urine or cervix for entire study population of 101 HIV-positive women. Prevalence of any HPV type, any high-risk HPV type, any low-risk HPV DNA type, or the presence of a multiple infection in urine or cervix for entire study population of 101 HIV-positive women. The amplifiability of HPV DNA from urine and cervical swab specimens is also compared in the first column. White bars, urine samples; thatched bars, cervical swab samples. Significant differences are denoted by an asterisk (P < 0.05). FIG. 2. | Detection of any HPV DNA type or any high-risk HPV DNA type in urine and cervix of women with normal Pap smears (WNL; n = 53), ASCUS (n = 17), or abnormal Pap smears (n = 27). Detection of any HPV DNA type or any high-risk HPV DNA type in urine and cervix of women with normal Pap smears (WNL; n = 53), ASCUS (n = 17), or abnormal Pap smears (n = 27). White bars, any HPV DNA type in urine; bars with vertical stripes, any HPV DNA type in the cervix; bars with horizontal stripes, any high-risk HPV DNA type in urine; bars with angled thatching, any high-risk HPV DNA type in the cervix. The percentage of urine or cervical swab samples positive for HPV DNA is indicated above each bar for each Pap smear classification. TABLE 1 | Demographic and clinical characteristics of current HIV Outpatient Clinic female population and urine study participants TABLE 2 | Distribution of high- and low-risk HPV types in urine and cervix of HIV-positive women infected with single and multiple HPV types TABLE 3 | Association of Pap smear status with detection of HPV in urine and cervix DISCUSSION : The validity of a PCR utilizing urine for the detection of HPV DNA was examined in these studies. The beta-globin housekeeping gene was amplified from 87% of the cervical swab specimens but only 73% of the urine specimens. This demonstrates that despite our best efforts, the rate of amplifiability of HPV DNA from a cervical swab specimen is significantly higher than that from a urine specimen. This could simply be due to the fact that urine samples from healthy women contain relatively few cells [Microscopic examination of sediment, p. 78, H. M. Free (ed.), Modern urine chemistry, Bayer Diagnostics Division, Bayer Corporation, Tarrytown, N.Y., 1996] or that urine contains inhibitors of the PCR . The rates of detection of any HPV DNA type, any high-risk HPV type, and any low-risk HPV type were not statistically different between the urine specimens and the cervical swab specimens for this population. This may imply that a patient whose cervix is infected with HPV is more likely to shed HPV-infected cervical cells that can then be detected in the urine of that patient. This possibility is not particularly surprising when one considers other virus infection systems, like the cytomegalovirus infection system, in which virus is often shed in the urine . Conversely, the fact that significant differences in rates of HPV DNA detection between the urine and the cervix were not detected in the present cohort may simply be due to a lack of power due to the size of the population studied. Future studies with increased numbers of paired samples are needed in order to strengthen these observations. The overall prevalence of detection of any HPV DNA was 48% for the urine specimens and 58% for the cervical swab specimens. These rates are higher than those found in other recent studies. One study conducted by Strauss et al. with 144 women attending a genitourinary clinic demonstrated HPV DNA in 40% of cervical swab samples and 15% of urine specimens by first-round PCR amplification. Another study conducted with 489 patients who were referred to a colposcopy clinic detected HPV DNA in 49% of the cervical swab specimens and 38% of the urine specimens . The differences in prevalence between these studies and the present study may simply be due to the different populations that were evaluated. Because of the sexually acquired nature of their disease, HIV-positive women probably experience increased levels of exposure to HPV as well as other STDs. This would result in a higher overall prevalence of HPV infection in this population. The prevalence rates found in the present study also differ from those found in a recent study conducted by Sellors et al. with a population of women who were attending a colposcopy clinic. The overall prevalence of HPV of any type in the study by Sellors et al. was 35% in urine specimens and 63% in physician-obtained cervical swab specimens. The discrepancy between the two studies can again be explained in part by the different populations examined. The study of women attending a colposcopy clinic contained only women who had an abnormal Pap smear (ASCUS, LGSIL, HGSIL, or adenocarcinoma). Conversely, 52.5% of the HIV-positive population evaluated in the present study had Pap smear gradings within normal limits , which would lead to overall lower prevalence rates in cervical swab samples. Additionally, the higher prevalence seen in the urine specimens in the urine study with HIV-positive women may simply reflect differences in specimen handling and extraction techniques between the two studies. Lastly, the prevalence of HPV DNA in the present study was lower than that found in another recent study conducted by Jacobson et al. . They detected HPV DNA at relatively high rates in urine (75%) and cervical swab (90%) specimens from an inner-city adolescent population. Interestingly, although no patients under 20 years of age were included in the present study with HIV-infected subjects, the rate of detection of any HPV type in urine was highest (79%) for women ages 20 to 30 years (data not shown). Examination of the samples for the 27 different HPV types that are detected by the reverse line hybridization assay revealed the presence of 26 of these types in the patients in the urine study. The diversity of HPV types represented in this study is most likely a reflection of the high-risk behaviors of these women. When the overall distribution of HPV types that were detected in the urine samples was compared with the overall distribution of HPV types that were detected in the cervical swab samples , potential preferences for the detection of types 53, MM4, MM7, 54, and 55 were seen for the cervical swab specimens and a potential preference for the detection of type MM9 was seen for the urine samples. These differences did not reach statistical significance due to the small sample sizes. HPV type MM9 is considered to be of intermediate to high risk as a cause of cervical cancer . Type MM9 may represent a viral type that is tropic for the vagina or that primarily infects the urethra. Types 53 and 54 are considered low risk; and types 55, MM4, and MM7 are thought to be of intermediate or high risk as causes of cervical cancer. Perhaps infections caused by these viral types are more likely to be localized only to the cervix or have very low copy numbers so that the virus is not shed in the urine. This lack of sensitivity for these types could prove to be a limitation of urine testing, which could potentially be compensated for by taking multiple urine specimens from the same patient . Future studies are needed not only to increase the sample size but also to compare the site(s) of anogenital HPV infection in women whose urine is positive for HPV DNA. Finally, HPV type 16 was commonly found in both the cervix and urine, as seen previously . However, the other most common types found in the present study (types 35 and MM7 in the cervix and type 59 in urine) were either not tested for or found less often in the inner-city adolescent population. This most likely reflects the differences in the populations tested between the two studies. When the rates of concordance of the results between the samples was compared, the results of the test with urine were not statistically different from those of tests with cervical swabs for the detection of any high-risk or low-risk HPV types for specimen pairs with discordant results. However, when the detection of any HPV type is considered, HPV DNA of any type was detected significantly more frequently in the cervical swab specimens. This is most likely due to the increased number of comparisons for this group and indicates that the assay with urine specimens may be a slightly less sensitive method for detecting HPV DNA than the assay with cervical swab specimens. This finding may also be a reflection of the natural history of HPV infection, which may be detected more readily in urine as a function of the higher viral loads and increased shedding of infected cells. Determination of the local HPV load would help prove this theory. The inability to detect low-risk HPV types in urine is not believed to be a serious limitation of this assay. Table examined the risk associated with having an abnormal Pap smear when HPV DNA was detected in urine or cervical swab samples. For both sample types, there were significant associations between having an abnormal Pap smear or an abnormal Pap smear result including ASCUS and detection of any HPV type, any high-risk HPV type, multiple HPV types, or multiple high-risk HPV types. Importantly, for these categories the percentage of samples in which HPV DNA was detected did not differ significantly between the urine specimens and the cervical swab specimens (Fig. and data not shown). These data suggest that testing of urine may be comparable to testing of a cervical swab specimen for HPV DNA detection and that, like the cervical swab specimen, the rate of HPV DNA detection increases with the severity of the abnormal Pap smear or the presence of HPV disease. These data corroborate the findings of Vossler et al. , who demonstrated HPV DNA in urine specimens from 13 of 15 women who had evidence of condylomata, dysplasia, or invasive carcinoma but in only 3 of 6 women who had normal Pap smears. A potential limitation of this analysis is that our study lacks biopsy data to verify the findings of an abnormal Pap smear. Such data could be important for determination of the true positives in the group with abnormal Pap smears. Unfortunately, an ASCUS or even a low-grade abnormal Pap smear is not always cause for an immediate biopsy in this HIV-positive population, and women with normal smears are not normally subject to a biopsy procedure. Thus, with so few biopsy data available for this study, Pap smear status was used to stratify the population. Interestingly, in women with normal Pap smears the rate of detection of any HPV type was approximately 15% greater in the cervix than in urine . In the present study, this finding was not significant. However, it be will interesting to monitor this observation in future studies in order to determine if screening by use of a cervical swab specimen compared to screening by use of a urine sample would produce a higher percentage of false-positive results. In summary, we have developed a urine-based assay for detection of the DNA of HPV, which is responsible for nearly all cases of cervical cancer and which is of particular concern in the HIV-positive population. The assay with urine appears to be as adept as the assay with a cervical swab specimen for the detection of any HPV type or any high-risk HPV type. Given that instances of discordant results for HPV DNA detection between the urine specimens and the cervical swab specimens may potentially be rectified by testing multiple urine specimens from the same patient, the urine assay merits further study to assess whether it may be an appropriate cervical cancer screening technique for HIV-positive women. Primary screening of HIV-positive women could involve the identification of patients infected with high-risk HPV types by urine assay and then monitoring of these patients more closely by use of Pap smears (every 6 months). In addition, this assay could potentially be used as a secondary screen for those HIV-positive women with abnormal Pap smears to monitor the course of their HPV infections. Additional studies need to be performed in order to solidify the role of urine screening in the HIV-positive population. Future studies also need to be undertaken to ascertain the role of testing of urine for HPV DNA in women not infected with HIV. Backmatter: PMID- 12202566 TI - Mycobacterium microti Infection (Vole Tuberculosis) in Wild Rodent Populations AB - Mycobacterium microti (vole tuberculosis) infections in small wild mammals were first described more than 60 years ago in several populations in Great Britain. Few studies of vole tuberculosis have been undertaken since then, and little is known about the relationship between M. microti isolates originating from different populations or at different times or of the prevalence of this infection in wild rodent populations, despite human cases of M. microti infections being increasingly reported. In this study, field voles (Microtus agrestis), bank voles (Clethrionomys glareolus), and wood mice (Apodemus sylvaticus) were found to be infected, with up to 8% having external tuberculous signs, in wild populations in Northumberland and Cheshire, England. Spoligotyping applied directly to the clinical material simultaneously detected and typed M. microti bacteria in skin lesions, lymph glands, and internal abcesses. IS6110 restriction fragment length polymorphism typing of cultured bacteria was used to compare these isolates with previously isolated strains from both animals and humans. This demonstrated that although the current rodent isolates were distinct from those isolated from voles in the 1930s in Great Britain, they had a high degree of similarity to these strains and were distinct from the M. microti isolates from humans, a pig, and a ferret from The Netherlands. Thus, M. microti infection seems to be widespread in wild rodent populations, but more studies are needed to understand how M. microti might be transmitted from animals to humans and to determine better the zoonotic risk posed. Keywords: Introduction : Tuberculosis (TB) in small wild mammals, namely, field voles (Microtus agrestis), bank voles (Clethrionomys glareolus), wood mice (Apodemus sylvaticus), and shrews (Sorex araneus), was first reported by Wells and Oxon in 1937 . The causative agent was named Mycobacterium tuberculosis subsp. muris and later Mycobacterium microti (or vole tuberculosis), a member of the M. tuberculosis complex . The other members of this complex are M. tuberculosis, Mycobacterium bovis (including the attenuated BCG vaccine strains), Mycobacterium africanum, and the recently described subspecies Mycobacterium canetti . M. microti has a characteristic pleomorphic microscopic morphology, with sickle-shaped, spiral, or S-like forms being seen in fresh material. This typical curved appearance is, however, usually lost during in vitro culture . It is difficult to distinguish M. microti from other members of the M. tuberculosis complex on the basis of biochemical properties, but they are readily identified by PCR-based spoligotyping and/or IS6110 restriction fragment length polymorphism (RFLP) typing . The latter requires large numbers of cultivated bacteria but facilitates more detailed phylogenetic comparisons. Well's early studies showed a varying prevalence of M. microti of 9 to 31% in field voles, depending on the place and season of capture, and overall prevalences of 11% in bank voles, 2% in wood mice, and 1.5% in shrews . No further data are available on the prevalence of M. microti in wild mammals, and in general, little attention has been given to M. microti as a causative agent of tuberculosis, with only sporadic cases being documented. It was isolated from a rock hyrax (Procavia capensis) in South Africa in 1960 and from a pig in The Netherlands in 1965 . Pattyn et al. described a disseminated M. microti infection in a llama (Llama vicugna molina) from the Antwerp zoo in 1970, and it was isolated from a pet ferret, also in The Netherlands, in 1993 . In the late 1990s, as part of a project to develop a standardized method for the identification and nomenclature of M. bovis, spoligotype patterns of M. tuberculosis complex strains from the Central Veterinary Laboratory in the United Kingdom were compared to those in an international database at the National Institute of Public Health and the Environment in The Netherlands. The patterns of 11 M. tuberculosis complex isolates from the Central Veterinary Laboratory database were identical or highly similar to the spoligotypes of M. microti isolates. These isolates were from eight domestic cats, a cow, a wild badger (Meles meles), and one human , raising the possibility that the prevalence of M. microti is greater than was hitherto assumed. M. microti has long been considered unimportant as a human pathogen . However, the diagnosis of M. microti infections is severely hampered by the slow growth in vitro of the bacteria on primary isolation, and routine diagnostic laboratories that culture for M. tuberculosis bacteria for up to 4 weeks, as is standard, will probably miss M. microti, since at least 8 weeks of incubation is needed . When a database of nearly 6,000 IS6110 RFLP patterns of M. tuberculosis complex isolates from human isolates was searched, two M. microti infections were diagnosed for the first time by van Soolingen et al. . In the same study, two other human M. microti infections were diagnosed by their M. microti-specific spoligopatterns . Three of these four cases were in immuncompromised individuals, and one was in a 39-year-old immunocompetent man. Later, a fifth case caused by M. microti was diagnosed in another immunocompetent patient in the Netherlands (D. van Soolingen, unpublished observation). More recently still, severe forms of tuberculosis caused by M. microti have been diagnosed in two patients in Germany and in one patient in Switzerland (I. Fengels, R. Mayer, B. Frauchiger, D. van Soolingen, B. Villiger, and G. E. Pfyffer, poster from the Congress of the Swiss Society of Pneumology, 2000), all of whom were immunocompetent. Since these M. microti infections were recognized through the application of molecular techniques, it is conceivable that they represent only a small proportion of the true number of human cases. The present study describes the identification, characterization, and distribution of M. microti in wild rodents at two sites in England. Such studies are needed to provide some insight into how M. microti might be transmitted from animals to humans and to understand better the zoonotic risk posed. Furthermore, this paper describes the genetic relatedness of the vole isolates of the present study to the previously published M. microti strains on the basis of IS6110 RFLP. MATERIALS AND METHODS : Isolation of M. microti from rodents. | Samples were collected from Kielder, Northumberland, in northeastern England and Cheshire, northwestern England. In Kielder, field voles were trapped monthly, using Ugglan special mousetraps, at six 0.3-ha sites, located in clear-fell areas of 5 to 10 ha, in March to October 1998 and March to October 1999 (inclusive) and in March and April 2000 (I. M. Graham and X. Lambin, unpublished data). The numbers of voles in Kielder with clinical signs of TB infection, that is, the characteristic skin lesions described by Wells or obvious lymphadentitis , were recorded. Material from skin lesions was sampled from nine live field voles, and enlarged lymph nodes were removed from four field voles and one bank vole that had died in the traps. Additionally, 20 fecal and two urine samples were collected in April 2000 from live field voles with clinical signs of TB. A snap-trap study was undertaken in Kielder from April to July 2000 as part of a separate project. Postmortems were performed on 180 field voles captured, and gross internal lesions typical of TB were recorded. The lungs from nine voles and internal caseous lesions from seven voles were removed for detection and typing of M. microti. In Cheshire, bank voles and wood mice were trapped in woodland on a monthly basis as part of a long-term study of host-pathogen dynamics , but for most of this time TB was not suspected and therefore was not monitored. On the basis of the results from Kielder, however, TB-like skin lesions were removed from five animals (three bank voles and two wood mice) in 1998 and 1999 for laboratory analysis. Since the clinical samples were collected solely for microbiological examination and since lesions at the Cheshire sites were not routinely recorded, prevalence studies were not carried out here. Culture and DNA extraction. | Clinical samples were decontaminated and grown in two different ways. Five samples were homogenized in N-acetylcysteine-NaOH and inoculated into MB/BacT Process bottles. These were incubated at 37C in the MB/BacT mycobacteria detection system (Organon Teknika Corp., Durham, N.C.). Thirteen samples of external lesions were each sonicated in 0.1 ml of phosphate-buffered saline for approximately 1 min using an ultrasonic power unit operating at 1.7 A. Thereafter the samples were decontaminated in 0.3 ml of 6% H2SO4 at room temperature for 10 min and neutralized with 1.5 ml of neutralization reagent. This reagent was prepared by dissolving 89 g of Na2HPO4 2H2O, 68 g of KH2PO4, and 20 mg of phenol red in 1,000 ml of distilled water and adjusting the pH to 11.0 with 5 M NaOH. Of this solution, 0.5 ml was incubated on Lowenstein-Jensen medium supplemented with 12.5 g of pyruvate per liter and on modified Dubos medium at 35.5C. DNA was extracted from all samples of clinical material (with the exception of the feces and urine samples) for spoligotyping by incubation overnight at 60C in digestion buffer (1:1), consisting of 500 mM Tris (pH 9.0), 20 mM EDTA, 10 mM NaCl, 1% sodium dodecyl sulfate, and 0.5 mg of proteinase K per ml, followed by DNA purification using the QIAamp DNA mini kit (Qiagen Ltd., Crawley, United Kingdom) as specified by the manufacturer for purification of tissue material. DNA was extracted from the feces and urine as specified by the manufacturer of the QIAamp mini kit for stool samples. Molecular typing. | Spoligotyping was performed as described previously . Extracted DNA was tested both neat and diluted 1:100. A total of 60 samples were subjected to spoligotyping. They included material from 2 of the cultures of lesions from Kielder field voles; external lesions from 9 Kielder field voles, 3 Cheshire bank voles, and 2 Cheshire wood mice; lymph node material from 4 Kielder field voles (5 samples, since lymph nodes were taken from both the retropharyngeal region and abdomen of one of the voles) and 1 Kielder bank vole; 20 fecal, 2 urine, and 9 lung samples from Kielder field voles; and internal caseous lesions from an additional 7 Kielder field voles. To control for the occurrence of possible PCR inhibitory effects, 10 of the 60 samples were randomly chosen to test for the presence of inhibitory compounds. Each of these 10 samples was subjected to spoligotyping twice: one subsample was tested in the standard way together with the other samples, and the second subsample was spiked with 1 ng of M. tuberculosis DNA. Standard IS6110 RFLP typing, as described previously , was performed on successful cultures of M. microti (three Kielder field voles and one Cheshire bank vole). Computer-assisted analysis of DNA fingerprints. | Using Gelcompar (version 4.1; Applied Maths, Kortrijk, Belgium), the IS6110 RFLP patterns of the current isolates were compared with those of previous M. microti isolates including those isolated from voles in the United Kingdom by Wells in the 1930s . Furthermore, the IS6110 RFLP patterns of all M. microti isolates were compared with the international database consisting of 5,909 IS6110 RFLP patterns of M. tuberculosis isolates originating in 33 different countries worldwide, held at the National Institute of Public Health and the Environment in The Netherlands. Pathology. | Frozen material from four skin lesions was fixed in 10% formalin. In addition, one male field vole, also from Kielder, with a large external skin lesion between the scapulae was killed following removal of lesion material (for molecular typing) and fixed in formalin. At necropsy, axillary, mandibular, and mesenterial lymph nodes, lungs, heart, liver, spleen, kidneys, adrenals, colon, and femur were processed by standard procedures for embedding in paraffin. Paraffin sections were stained with hematoxylin-eosin [HE], Auramine-rhodamine [AR] or Ziehl-Neelsen stain [ZN]. HE and ZN pictures were taken with an SV-MICRO digital microphotocamera (Sound Vision, Wayland, Mass.). AR fluorescence was imaged with a Bio-Rad 1024 confocal laser-scanning microscope equipped with an argon/krypton laser. The 568-nm line was used to excite rhodamine. RESULTS : Prevalence of clinical TB in wild rodents. | Animals were designated as having clinical TB if they had the characteristic skin lesions described by Wells or obvious lymphadentitis. Most sampling from the Cheshire sites was carried out without reference to TB. Thus, prevalence has been estimated only for the Kielder samples. There, 101 cases of clinical TB were recorded throughout the study period from a total of 4,852 field voles examined (ca. 2% prevalence). There was a significantly higher prevalence in 1999 (67 of 2,495 = 2.7%) than in 1998 (12 of 1946 = 0.6%; chi2 = 26.8; 1 degree of freedom; P < 0.001), and this increase appeared, from the two months sampled, to have continued in 2000 (22 of 411 = 5.4%; chi2 = 3.4; 1 degree of freedom; P = 0.066), reaching around 8% (14 of 172) in April 2000. Detailed analysis of the influence of site, season, and vole density on prevalence will be described elsewhere (R. Cavanagh, X. Lambin, T. Ergon, M. Bennett, I. M. Graham, D. van Soolingen, and M. Begon, unpublished data). From the snap-trap study, 13 of 180 voles had clinical signs of TB (7% compared to the 8% prevalence in animals live trapped at a similar time [see above]). However, dissection revealed that a further 25 had internal tuberculous lesions in the form of cream-colored caseous abscesses under the skin or in the abdomen (overall prevalence, 21%). Detection and identification of mycobacteria by spoligotyping. | The characteristic spoligotype pattern for M. microti, in which spacer sequences 37 and 38 are present , was obtained for 24 of the 60 samples tested. The spoligotypes of 13 of the 14 skin lesions clearly indicated M. microti infection, and 1 was negative. The negative result was from a small lesion, less characteristic in appearance than the others. Spoligotyping of all lymph node samples showed the M. microti pattern except for one mesenteric lymph node from a field vole, and five of the internal caseous abscess samples also had the characteristic pattern, as did the two cultures. The fecal, urine, and lung samples were all negative, however. Of the 10 samples spiked with M. tuberculosis DNA, there was complete inhibition of the spoligotyping PCR when the DNA was used neat for 7 samples, and when it was diluted, 2 of the samples were still inhibited. Culture and RFLP typing. | After 12 weeks, cultures were successfully obtained from lesion material from three field voles and one bank vole. Of these, 2 were included in the 13 samples that were incubated on Lowenstein-Jensen and modified Dubos medium. Growth of the two samples was detected on both media, with the Dubos medium showing somewhat better and faster growth. The other two successfully grown strains were from the five samples that were incubated in the MB/BacT mycobacteria detection system. Culture has so far been unsuccessful from other lesion and lymph node material tested in this study. The results of the IS6110 RFLP typing of the four cultured isolates (voles UK 1999) are shown in Fig. . The three field vole isolates from Kielder all displayed identical IS6110 banding patterns, which exhibited 62% similarity to those for the five vole strains from Wells's collection from the 1930s. The bank vole isolate from Cheshire had a different pattern, which exhibited an overall similarity of 55% to those for the other eight vole strains. The comparison of all IS6110 RFLP patterns of M. microti isolates revealed two distinct groups: one of isolates from the United Kingdom and one of isolates from The Netherlands. The isolates within each of these two groups showed a high degree of similarity as can be seen in Fig. . There is more genetic diversity between the two groups, with a similarity of only 45%. The hyrax isolate was the most different of all. Comparison of all IS6110 RFLP patterns of M. microti isolates, from this study and a previous study , with those of the M. tuberculosis complex isolates present in the international database showed that the RFLP patterns of the M. microti isolates were unique: all other M. tuberculosis complex patterns had less than 80% similarity to M. microti. The M. microti isolates formed four different groups, quite distantly related, in the dendrogram of the international database. The first group consisted of the three vole isolates from Kielder, with identical IS6110 RFLP patterns and a similarity of 68% to the RFLP pattern of the llama isolate from the Antwerp zoo. The second group consisted of the scab isolate from Cheshire together with the vole isolates from the 1930s, which matched with a similarity of 56.9%. The third group was formed by the isolates from a pig, a ferret, and four humans from The Netherlands. The RFLP pattern of the hyrax isolate was found to be unique in the international database. Pathology. | All four dermal lesions showed necrotic histiocytosis with some calcification. One lesion showed exudative tuberculosis, with necrotic macrophages and large numbers of intracellular mycobacteria. In other lesions, more proliferative tuberculosis was observed, with fewer bacteria in macrophages. The killed vole appeared to be in a good nutritional state despite the presence of widespread tuberculous lesions. All its lung lobes contained firm light yellow nodules up to 1 cm in diameter. Histological examination revealed granulomatous interstitial and bronchoalveolar lesions. In ZN- and AR-stained sections, mycobacteria were abundant, especially in alveolar lumina bordering necrotic lesions and bronchioles. Periarterial plasmacellular cuffs were prominent. All lymph nodes showed lymphodepletion, plasmacytosis, and a variable number of microgranulomas. Draining axillary lymph nodes were more severely affected, with classic tubercles containing larger numbers of mycobacteria, than were nondraining lymph nodes. With the exception of the intestine and bone marrow, all other organs or tissues showed histopathological abnormalities, such as microgranulomas and calcification and variable numbers of mycobacteria. FIG. 1. | Comparison of the IS6110 RFLP patterns of the M. microti Comparison of the IS6110 RFLP patterns of the M. microti strains identified in this study and those of the M. microti strains from previous studies. DISCUSSION : This study confirms that 60 years after they were first discovered, M. microti infections are prevalent in wild rodent populations, although there are few reports of the condition in the intervening period . We suspect that, as in human cases, M. microti in wild rodents has simply been overlooked, despite increasing interest in M. bovis infections in wildlife . Wells obtained higher prevalences than those reported here, perhaps because he examined all animals postmortem, enabling him to record internal tuberculous lesions, which appear to develop earlier in the pathogenesis of the condition than external skin lesions do. In support of this, our postmortem study of snap-trapped field voles revealed a prevalence of infection in April to July 2000 of 21% rather than the 7% obtained based on external signs alone. The optimal approach to diagnosing infection in the field, without sacrificing animals, would be the detection of M. microti in feces or urine of infected rodents. Wells found a high prevalence of tuberculous lesions in the gastrointestinal and urinary tracts, and feces and urine can be collected directly from captured rodents. In this study, however, we failed to detect M. microti in fecal and urine samples from voles with clinical signs of TB by using spoligotyping. Pathogenesis studies are therefore essential to establish whether M. microti is excreted in feces and/or urine and, if so, whether it is excreted throughout the infection. Further work is also necessary to optimize DNA extraction to remove PCR inhibitors. In addition, oral swabbing might be tried, since Wells also found a high prevalence of pulmonary lesions . The spoligopatterns of M. microti in field voles, bank voles, and wood mice in this study were indistinguishable from those of strains isolated from British voles in the 1930s and, indeed, most other M. microti isolates. However, RFLP typing using IS6110 enabled finer differentiation of the successfully cultured isolates. The fingerprint patterns of isolates from three recently captured Kielder field voles were indistinguishable from each other but different from that of the Cheshire bank vole isolate and from those of the vole isolates from various British sites in the 1930s. The isolates from The Netherlands appear to cluster separately from the British isolates, but they differ according to the geographic source within each of these countries (unfortunately, the information regarding the exact geographic location within Britain of the five 1930s isolates is not known). This suggests long-term evolutionary divergence between the bacteria found in Great Britain and mainland Europe. In contrast, the strains isolated with a time interval of 60 years in Britain showed a high degree of similarity (62%). The half-life of IS6110 RFLP, which was determined on the basis of the rate of transpositions in 546 serial patient isolates, of 3 to 5 years could be in range with this divergence . Least closely related to the other M. microti isolates was that from a South African hyrax: this isolate also had an unusual spoligopattern . In contrast to the findings by IS6110 RFLP, on the basis of spoligotyping alone, all M. microti isolates in the database form a group clearly separate from other members of the M. tuberculosis complex. This indicates that M. microti, regardless of host or geographic origin, constitutes an evolutionarily conserved group of bacteria. Clearly, more work is required to clarify the nature of strain variation with host, region, and time of isolation. Although more isolates from a variety of host species need to be studied, RFLP typing did not provide any evidence for the existence of a specific M. microti strain adapted to the human host, although human-to-human transmission may occur . This, together with the high prevalences of M. microti found in wild rodents, both in this study and by Wells , suggests that these animals are likely to be the reservoir for the disease in humans. Furthermore, there is strong circumstantial evidence for direct zoonotic transmission of M. microti infection from rodents to humans (; Fengels et al., Poster). Sometimes, however, transmission may occur via liaison hosts, such as domestic cats or pet ferrets: such routes can be important for other zoonotic infections of rodents, for example, cowpox virus and plague . Further study of the epidemiology of vole tuberculosis is timely, with the increased recognition of M. microti infections in humans. We suspect that particularly in environments with a high level of human-rodent interaction and less than optimal hygiene conditions, many cases of tuberculosis due to M. microti infections may remain undiagnosed. Backmatter: PMID- 12202557 TI - PCR-Based Identification of Bacteria Associated with Endodontic Infections AB - PCR primers that target the bacterial 16S rRNA genes (or the tuf gene for the genus Enterococcus) were used to identify 10 putative bacterial pathogens in root canals with necrotic pulp. In addition, the associations of these microorganisms with symptoms and a history of diabetes mellitus were investigated. Microbial samples from the root canals of 24 teeth with necrotic pulp were included in the study. PCR with universal bacterial primers identified bacterial DNA in 22 specimens; the remaining 2 specimens were from intact teeth that had been traumatized 6 months prior to treatment. PCR with specific primers showed that preoperative symptoms were significantly associated with the presence of Streptococcus spp. (P < 0.001 by chi-square analysis). There was also a nonsignificant trend for symptoms to be associated with Fusobacterium nucleatum and Porphyromonas gingivalis (odds ratio, >2) and for diabetes mellitus to be associated with P. gingivalis and Porphyromonas endodontalis (odds ratio, >2). Cloning and sequencing of the universal PCR product in one specimen revealed the presence of an organism related to the genus Olsenella, which has not previously been described in endodontic infections. Keywords: Introduction : The presence of bacteria in the root canal leads to the development of periapical periodontitis . Several studies have shown an association between painful exacerbation of periapical lesions and the presence in the root canal of specific bacteria. Black-pigmented bacteria belonging to the genera Porphyromonas, Prevotella, and Bacteroides have been cultured from root canals in a significant proportion of cases in those studies and are frequently present in the same canals as members of the genera Peptostreptococcus and Fusobacterium . However, the findings of different studies based on culturing of canal contents vary significantly. This may be due, at least in part, to the reduced reliability and sensitivity of culturing techniques. The persistence or further expansion of a periapical lesion, despite seemingly adequate endodontic treatment and timely restoration of the tooth, is usually attributed to the persistence of pathogenic microorganisms in the root canal system. Recent investigations have documented that the presence of cultivable bacteria from canals at the time of obturation was critical in predicting failure of treatment . However, the microorganisms most commonly associated with failed endodontic cases are different from those cultured from canals with pulp necrosis. Studies reveal that most of these failed cases have gram-positive strains such as enterococci, streptococci, and eubacteria, with occasional Candida, peptostreptococci, and fusobacteria . Although enterococci were the most prevalent microorganisms in the last three studies, being present in 54, 70, and 38% of the cases, respectively, the percentages of different strains identified again vary significantly among the studies, and in a considerable number of cases there were no cultivable microorganisms. Therefore, sensitive and accurate molecular techniques are necessary to accurately characterize the root canal microbial irritants in order to determine their association with clinical symptoms and the prognosis of treatment. For example, the introduction of molecular methods into analyses of root canal samples has led to the identification of a number of fastidious organisms such as Bacteroides forsythus and Treponema denticola , which have not previously been described in endodontic infections. PCR amplification of the bacterial 16S or 23S rRNA gene (rDNA) or other rDNAs is more sensitive and more efficient than culturing and biochemical identification of endodontic flora. In the root canal microbial environment, PCR was shown to be more accurate than sodium dodecyl sulfate-polyacrylamide gel electrophoresis in differentiating and identifying the two important endodontic pathogens, Prevotella intermedia and Prevotella nigrescens, which could not be differentiated by culturing . Although the use of DNA probes can be more sensitive and more efficient than culturing, it still requires the presence of >104 bacterial cells to ensure detection . The PCR technique can be sensitive enough to detect a few DNA strands of the microorganisms present if adequate primers are used and the PCR conditions are sufficiently optimized. We have recently shown that, after inoculation of three endodontopathogenic bacteria in mouse pulp exposures, PCR was much more accurate than culturing in detecting the inoculated anaerobic bacteria . Several uncultivable species have been identified from dentoalveolar abscesses by PCR . Previous studies have shown that the diabetic host may have an increased periapical lesion size or may develop more serious infections in response to virulent root canal bacteria . Patients with a history of diabetes mellitus and periapical lesions may have significantly reduced healing following endodontic therapy compared with that for the nondiabetic population . The purpose of this study was to determine the presence of 10 putative root canal microorganisms in samples from root canals with necrotic pulp and apical periodontitis by using universal bacterial as well as species- or genus-specific PCR primers. We also determined the association of these organisms with clinical symptoms and with a history of diabetes mellitus. MATERIALS AND METHODS : Patient selection and sample collection. | All patient-related procedures used in this study conformed to protocols approved by the Institutional Review Board of the University of Connecticut Health Center. The purpose and scope of the study were explained to patients presenting for endodontic treatment for a tooth with pulp necrosis and apical periodontitis. Patients who consented to participate in the study and who had not been treated with antibiotics in the preceding 3 months were included in the study. Patients who indicated that they had diabetes mellitus received a free HbA1c test to determine the degree of their glycemic control. Patients who had no history of diabetes were offered a free fasting blood glucose test to verify that they did not have diabetes. The teeth involved had a negative pulp test result, had a periapical radiolucency on a preoperative radiograph, and had not had any previous endodontic procedures. Detailed information regarding signs and symptoms and radiographic and clinical data were collected. Symptomatic patients were defined as patients who had a preoperative visual analogue score of 30 or more on a 100-point scale, had moderate to severe pain to percussion or palpation of the tooth involved, and/or had swelling. The technique for sample collection was as follows: following isolation of the tooth involved with a rubber dam, the field was disinfected with 30% H2O2 and then 5% tincture of iodine. Caries and/or existing restorations, if present, were removed, and then the cavity was wiped with a sterile cotton pellet slightly wet with 1% buffered NaOCl, with care that it did not seep into the canal. The halogen disinfectants were then inactivated with 5% sodium thiosulfate. The pulp chamber was then accessed with a new sterile bur. If purulence or serous fluid was present in the canal, this was directly sampled with three size fine paper points. Otherwise, sterile saline was deposited in the canal, making sure that it did not overflow. A size 15 to 30 file (depending on the canal size) was used to negotiate the canal to the estimated length. If the canal was very calcified, Gates Glidden burs sizes 2 and 3 were used so that the paper point could penetrate to a depth close to the estimated canal length. Three fine paper points were then used to obtain the sample. The last paper point was left in the canal for 30 s. In multicanaled teeth, one paper point sample was obtained from each canal unless the canals were very calcified, in which case sampling of the canal in the root with the largest periapical lesion and the largest canal was done. The paper points were placed in sterile, DNA- and RNA-free vials containing 1 ml of filter-sterilized 10 mM Tris-HCl, 1 mM EDTA (pH 8), and 0.5 g of sterile glass beads (diameter, 0.71 to 1.18 mm). The vials were frozen at -70C until used. DNA extraction. | The vials with paper point specimens were vortexed for 2 min to disperse microbial cellular material into suspension. The suspension was removed from the original vial and transferred to 2-ml sterile vials, which were then centrifuged at 7,500 rpm (all centrifuge procedures were carried out with Eppendorf [Westbury, N.Y.] scientific microcentrifuge model 5417C) for 10 min, and the supernatant was again removed. DNAs were extracted from the cellular pellet by one of two methods. The first method (specimens 1 to 10) used the Chelex extraction and boiling technique . Briefly, this method involved the addition of 75 mul of 15% Chelex 100 resin (Bio-Rad) to the pellet resuspended in 0.5 ml of Tris-HCl buffer and thorough mixing, followed by incubation at 56C for 30 min in a dry heat block. The mixture was boiled in a dry heat block for 10 min and then chilled on ice for 5 min. It was then centrifuged at 12,000 rpm for 2 to 3 min. The supernatant was carefully removed, with the Chelex being avoided. The DNA was stored at -20C until it was ready for use in the PCR. For the last 14 specimens, we switched to the enzymatic extraction method, according to the protocol described for the QIAamp DNA mini kit (Qiagen, Valencia, Calif.), because of the manufacturer's claims of improved purity and yield of the extracted DNA and because it allows the extraction of fungal DNA for use in future research. The pellet was suspended in 180 mul of enzyme solution (20 mg of lysozyme per ml, 20 mM Tris HCl [pH 8.0], 2 mM EDTA, 1.2% Triton) and incubated for 30 min at 37C. Proteinase K (20 mul) and RNase A (4 mul at 100 mg/ml) were added, and the specimen was incubated for 2 min at room temperature. Buffer AL (200 mul) was added, and the specimen was vortexed and incubated at 56C for 30 min and then for 15 min at 95C. Ethanol (200 mul at 96 to 100%) was added, followed by vortexing and brief centrifugation. The mixture was then added to a QIAamp spin column and centrifuged at 8,000 rpm for 1 min. The column was then placed in a clean 2-ml collection tube, 500 mul of buffer AW1 was added, and the mixture was centrifuged at 8,000 rpm for 1 min. The column was again placed in a clean 2-ml collection tube, and 500 mul of buffer AW2 was added, followed by centrifugation at 14,000 rpm for 3 min. Then, buffer AE (200 mul) was added, followed by centrifugation at 8,000 rpm for 1 min. The elutions were combined for a total yield of 400 mul, which was aliquoted in sterile, DNA- and RNA-free conical tubes and frozen at -20C until use. Prior to performing the second extraction method we conducted a pilot experiment to determine if the two different extraction methods affected the yield of extracted DNA from representative stock strains of two gram-positive bacteria and two gram-negative bacteria. This experiment was also run with one clinical sample that was divided into two aliquots, and each aliquot was extracted by one of the two methods. PCR was later run with primers specific for all the bacteria under study. These experiments did not reveal any perceptible differences in DNA yields or PCR results between the different extraction methods, and therefore, the results obtained by both methods are considered together. The yield of extracted DNA was quantified for each of the control stock bacterial strains and clinical specimens by using a Hoefer DyNA200 fluorometer (Amersham Pharmacia Biotech, Piscataway, N.J.). The yield ranged from 2 to 33 ng/mul for the stock bacterial strains and 1 to 19.5 ng/mul for the clinical specimens. Microorganism selection. | We chose to evaluate the root canals for the presence of 10 microorganisms that have frequently been isolated from root canals with necrotic pulp . Our selection of the bacteria was based on the following criteria: organisms that are highly prevalent in root canals with necrotic pulp (black-pigmented bacteria, Fusobacterium nucleatum, Peptostreptococcus micros, and Streptococcus spp.) , organisms that are frequently found in patients with symptomatic endodontic infections (P. intermedia, P. nigrescens, Porphyromonas gingivalis, and Porphyromonas endodontalis) , organisms that have been detected in root canals from patients who have failed endodontic treatment (Enterococcus spp.) , and organisms that are prevalent in patients with severe periodontitis and that have recently been identified in root canals by PCR (T. denticola and B. forsythus) . PCR amplification of rDNA. | Previously published primer pairs were selected for specific PCR amplification of 16S rDNAs (or the tuf gene for the genus Enterococcus) of the microorganisms listed in Table . Initially, a universal eubacterial primer pair was used to detect DNAs from all bacterial species present in the sample. Subsequently, a PCR mixture with oligonucleotide primers specific for rDNAs was used. At least duplicate experiments were run for each specimen. PCR amplification was performed in a thermal cycler (PE9700 or PE2400; Perkin-Elmer Applied Biosystems, Foster City, Calif.). It was carried out in a volume of 50 mul containing 10 mul of extracted sample DNA or 5 mul of extracted control stock bacterial DNA (see below), 5 mul of 10x PCR buffer, 0.25 mul of 5 U of Taq DNA polymerase (Eppendorf, Cologne, Germany) per mul or 0.5 mul of HotStar Taq (Qiagen), 1.5 mM MgCl2, 0.2 mM concentrations of each of the four deoxynucleoside triphosphates (Takara, Otsu, Shiga, Japan), and a 0.5 muM concentration (500 ng) of each (sense and antisense) primer; the balance consisted of sterile ultrapure water. PCR conditions for each primer combination were optimized in pilot experiments. The PCR conditions used were generally as follows: the initial denaturation was at 94C for 2 min for Eppendorf Taq or 15 min for HotStar Taq. This was followed by 30 cycles of denaturation at 94C for 15 s, annealing at a temperature that depended on the primer for 15 s, and extension at 72C for 45 s. The final extension was at 72C for 5 min, and then the products were cooled to 4C until they were removed. The amplification products were analyzed by 2% agarose gel electrophoresis in TAE buffer (40 mM Tris-acetate, 2 mM EDTA [pH 8.3]). The Power Pac 1000 apparatus (Bio-Rad, Hercules, Calif.) was set at 110 mA for 2 h or 95 V for 1 h. The gels were stained with 0.5 mug of ethidium bromide per ml for 30 min and destained with water for 20 min. The PCR products were visualized under UV light with an Alpha Imager (Alpha Innotech Corp., San Leandro, Calif.). For each primer we ran a number of PCR controls. These included the use of DNA from American Type Culture Collection (ATCC) stock strains of the respective bacterial species (Enterococcus faecalis for the Enterococcus primers) as positive controls. The Streptococcus-specific primers were reported as being specific for Streptococcus intermedius, with possible cross-reactivity with Streptococcus milleri isolates . However, our positive control experiments have shown that these primers reacted with S. intermedius, Streptococcus constellatus, Streptococcus anginosus, Streptococcus mutans, Streptococcus sanguis, and Streptococcus bovis, all at a single band at 500 bp. Thus, these primers were considered Streptococcus genus specific. The DNAs of the ATCC stock strains were extracted from spectrophotometrically determined concentrations of 3 x 108 bacterial cells/ml that were cultured under ideal conditions for the particular species. An additional positive control was the universal primer pair specific for bacterial 16S rDNA, with which positive results were obtained with DNA from all bacteria but negative results were obtained with DNA from Candida albicans. In addition, each primer set was run with DNA extracted from all other bacterial species used in the study together with DNA extracted from ATCC stock strains of Eubacterium nodatum and Actinomyces israelii and with water (no DNA) as negative controls. Representative PCR products obtained by use of each of the species- or genus-specific primers with patient specimens were directly sequenced (see "Cloning and sequencing of novel 16S rDNA sequences" below) to determine the published sequence closest to that of the organism amplified. Cloning and sequencing of novel 16S rDNA sequences. | Two specimens, specimens SP05 and SP08, yielded a PCR product with the universal primer pair but no product with the 10 specific primers tested. Specimen SP08 yielded too little product for cloning and will not be described further. Amplification products from specimen SP05 were cloned into the vector pCR 2.1-TOPO TA (Invitrogen, Carlsbad, Calif.) according to the instructions of the manufacturer or were sequenced directly. For cloning, the PCR product was transformed into Escherichia coli One-Shot TOP10. Colonies containing the insert were used to inoculate Luria-Bertani agar (Miller; Fisher Scientific Co., Pittsburgh, Pa.). Plasmid DNA was purified with the QIAprep Spin Miniprep kit (Qiagen) or the Concert Rapid Plasmid purification kit (Life Technologies, Gibco BRL, Rockville, Md.). The purified plasmid DNA from the cloning procedure was sequenced in the University of Connecticut Health Center Molecular Core Facility by using an ABI Prism 3100 genetic analyzer (Perkin-Elmer Applied Biosystems) and reverse primer M13 or T7 (Invitrogen). The universal PCR products were purified with the Concert Rapid PCR purification system (Life Technologies, Gibco BRL) and directly sequenced by using the universal forward and reverse primers . The resulting sequences were used to search databases available through the National Center for Biotechnology Information. PCR products obtained with specific primers from representative patient specimens were partially sequenced directly after purification as described above to verify the identity of the product. All species-specific primers yielded sequences that matched published sequences for the respective species. The two representative products for the genus-specific primers yielded sequences that had close homology with the sequences of S. sanguis, unidentified oral streptococci (GenBank accession no. ), Streptococcus cristatus, and Streptococcus pneumoniae for the Streptococcus genus-specific primers and various Enterococcus spp. for the Enterococcus genus-specific primers. Phylogenetic analyses of the novel 16S rDNA sequence. | Significant database hits were aligned with our unknown sequence by using ClustalW software in MacVector (Genetics Computer Group, Oxford Molecular Co.). A neighbor-joining phylogenetic tree was constructed from the alignment by using MacVector (Genetics Computer Group, Oxford Molecular Co.). A distance matrix was constructed by using a Tamura-Nei model without gamma correction and with gaps distributed proportionately. Neighbor-joining bootstrap values were derived from 1,000 replications and were added to the tree. Data analysis. | The associations between the positive identification of a bacterial species or genus and symptoms or a history of diabetes were analyzed by odds ratio (OR) analysis. OR associations of 2 or more were considered positive associations . These positive associations were further analyzed by a chi-square analysis to determine their statistical significance. Nucleotide sequence accession number. | The sequence that forms a basal lineage in the Olsenella clade detected in this study has been deposited in GenBank under accession number . TABLE 1 | Oligonucleotide primers used RESULTS : Of 24 patients participating in the study, 8 were considered to be symptomatic and 6 had a history of diabetes mellitus (2 with type 1 diabetes mellitus and 4 with type 2 diabetes mellitus) . The HbA1c results revealed that three diabetic patients had moderate glycemic control (7 to 10%) and three had poor glycemic control (>10%). Of the 18 nondiabetic patients, 9 agreed to take the fasting blood glucose test, and all had results below 126 mg/dl, which is generally accepted as the threshold value for the diagnosis of diabetes mellitus . Twenty-two of the 24 specimens tested reacted positively with the universal bacterial primer pair. Two specimens, specimen SP09 (from a nondiabetic individual) and specimen SP12 (from an individual with type 2 diabetes), had no identifiable PCR amplicons with the universal primers. Retrospective analysis of the clinical conditions of the root canals from which these two specimens were sampled revealed that the root canals may not have had bacterial contamination at the time that treatment was initiated (see Discussion). Of the 22 specimens positive for bacteria by PCR, 20 yielded amplicons with 1 or more of the 10 specific primer pairs . Two specimens (specimens SP05 and SP08) were positive for bacteria by PCR with the universal primer pair but did not have a positive reaction with any of the specific primers used. Certain organisms such as F. nucleatum, P. micros, Streptococcus spp., and P. nigrescens were more commonly identified than other organisms . The results of an analysis that used the OR of the association between specific organisms and the presence of symptoms are shown in Table . Streptococcus spp., F. nucleatum, and P. gingivalis were associated with symptoms. Further analysis by a chi-square test revealed that there was a statistically significant association between Streptococcus spp. and symptoms (P < 0.001) but that the remaining associations did not reach statistical significance. It was also evident from Table that the combination of F. nucleatum and Streptococcus spp. was associated with symptoms in 6 of 9 patients, and the absence of either or both organisms was associated with pain in 2 of 13 patients (OR = 11), which was also statistically significant (P = 0.014 by chi-square analysis). Of the eight symptomatic patients, four had localized or diffuse swelling (specimens SP07, SP10, SP14, and SP16). Three organisms had positive associations with swelling: Streptococcus spp. (OR = 7), P. gingivalis (OR = 6.3), and Enterococcus spp. (OR = 3); however, none of these associations were statistically significant (P > 0.05 by chi-square analysis). Six patients in this study had a history of type 1 or type 2 diabetes mellitus, with various degrees of glycemic control . An analysis of the association of diabetes with specific root canal microorganisms was performed. That analysis revealed a positive association between diabetes and P. endodontalis and P. gingivalis ; however, these associations were not statistically significant (P > 0.05 by chi-square analysis). Other patterns related to the type of diabetes or degree of glycemic control could not be established. The association between symptoms or diabetes and the number of different organisms identified in the specimens that were positive with the universal primers was analyzed. The mean number of organisms identified in the 8 symptomatic patients was 3 +- 1.4, whereas the mean number was 2.78 +- 1.85 for the 14 patients with mild or no symptoms. The mean number of organisms identified in the five patients with diabetes (after excluding specimen SP12, which was negative with the universal primer pair) was 3.2 +- 1.64, whereas the mean number was 2.77 +- 1.72 for the 17 nondiabetic patients. However, these differences were not statistically significant (P > 0.05 by the t test). Direct sequencing or cloning and sequencing were performed with the PCR product obtained with the universal primer pair from specimen SP05, a specimen that did not yield a product by PCR with any of the specific primer pairs. The sequence of the directly sequenced PCR product and the sequences of the PCR products of two other clones obtained with the TOPO M13 reverse primer were homologous. This indicates that the sequence represents that of an organism that is predominant in the specimen and argues against the likelihood of the presence of a chimeric sequence. The neighbor-joining algorithm showed that this sequence forms a basal lineage in an Olsenella clade of a phylogenetic tree . One additional clone from specimen SP05 produced a sequence that was closest to that of the S. mutans, Streptococcus gordonii, and S. sanguis group of microorganisms. Direct sequencing of representative PCR products from patient specimens with each of the specific primers produced sequences that matched those from corresponding species or genera. FIG. 1. | Neighbor-joining tree of the sequence from specimen SP05 and related 16S rDNA sequences as determined by BLAST scores. Neighbor-joining tree of the sequence from specimen SP05 and related 16S rDNA sequences as determined by BLAST scores. Clade associations are quantified by bootstrap values. From left to right, the bootstrap values represent node support from neighbor joining. TABLE 2 | PCR results, diabetes mellitus type, HbA1c results, and preoperative pain and/or swelling experience (symptomatic) for 24 patients TABLE 3 | OR analysis of the association of specific microorganisms with symptoms TABLE 4 | OR analysis of the association of specific microorganisms with history of diabetes mellitus DISCUSSION : In this study both universal and specific PCR primers targeting bacterial rDNAs were used to investigate the prevalence of putative pathogenic bacteria in root canals with necrotic pulp. Furthermore, valuable initial findings on the potential association of certain organisms with clinical symptoms or the presence of diabetes were revealed. More extensive studies are needed to provide definitive conclusions on the association of root canal microorganisms with these and other clinical variables. Two specimens had no identifiable bacterial DNA. Both were from patients who had a history of a traumatic injury that occurred about 6 months before treatment. For these two patients, the pulp was not responsive, the patients were starting to have mild symptoms, and there were periapical radiographic changes that were not consistently seen throughout treatment. It is conceivable that at the time of treatment bacteria had not yet invaded the necrotic pulp in these patients. None of the other patients in the patient population studied had a similar clinical presentation. Therefore, these patients could be considered further controls for the adequacy of the sampling technique used. The proportion of specimens positive for bacterial DNA was 22 of 24 (92%). This was higher than that found in a recent report, in which the same universal bacterial primers were used but yielded a positive identification for only 73% of the cases examined . Differences in clinical diagnosis, sampling, DNA extraction, or PCR techniques between the two studies may account for these differences. In this study, Streptococcus spp. were the organisms most strongly associated with endodontic symptoms and with the presence of swelling. This finding is consistent with those of some previous studies but not others . F. nucleatum was the organism most frequently identified in root canals with necrotic pulp. This organism was found to be the organism that was the most prevalent in endodontic infections in previous studies that have used both culturing and PCR methodologies. It was also positively associated with the presence of preoperative symptoms in this study, and the presence of the combination of F. nucleatum and Streptococcus spp. was significantly associated with the presence of preoperative symptoms. F. nucleatum was previously shown to increase the pathogenicities of other organisms in mixed culture, especially those of P. gingivalis and P. intermedia . P. gingivalis was identified in only two specimens in this study. Although the OR analysis showed that this organism was associated with symptoms, swelling, and diabetes, the sample is too small to establish any definitive association. This organism has frequently been associated with severe endodontic symptoms , and its pathogenicity in endodontic infections should be further investigated. In this study, 12 of 22 samples (55%) had one or more members of the black-pigmented gram-negative rods: P. endodontalis, P. gingivalis, P. intermedia, and P. nigrescens. This group of organisms has long been associated with the presence of endodontic symptoms . However, our analysis of this sample revealed no association of black-pigmented gram-negative rods with symptoms (OR = 0.75). In a recent study, in which these four organisms were investigated, one or more of these four organisms were identified in 59% of the specimens . In that study, it was concluded that black-pigmented organisms in the root canal were not associated with symptoms, although they were very prevalent in pus samples from periapical abscesses. Only one sample was positive for P. intermedia, whereas about a third of the samples were positive for P. nigrescens. P. intermedia is an organism that has frequently been identified in endodontic infections . However, more recently it was recognized that this organism is difficult to differentiate from P. nigrescens by traditional culturing methods but that the two are easily distinguishable by molecular techniques . Our results confirm previous findings that P. nigrescens is more prevalent in endodontic infections than P. intermedia . It is of interest that B. forsythus, a gram-negative rod, and the spirochete T. denticola were always associated with one or more members of the black-pigmented gram-negative rods. These two organisms (together with P. gingivalis) have been called the "red complex" bacteria because of their strong association with severe forms of periodontal disease . Our findings with respect to these organisms (except for P. gingivalis) agree with those in a recent report in that the organisms do not seem to be associated with symptomatic cases. A recent report indicated that other oral treponemes such as Treponema maltophilum and Treponema socranskii, which were not included in this study, may be more prevalent in endodontic infections than T. denticola . More studies are needed to discern the contributions of all these organisms to their pathogenicities and their potential association with treatment failure. A genus-specific primer pair that amplified a unique sequence in the Enterococcustuf gene was used to detect Enterococcus spp. This primer pair was shown to detect 14 of 15 enterococcal species and was negative with 73 of 79 other gram-positive and gram-negative organisms tested . There has been a recent increase in emphasis on the presence of this microorganism in association with failed endodontic treatment . Enterococci are resistant to calcium hydroxide , which is an intracanal medicament commonly used in patients with pulp necrosis. E. faecalis and Enterococcus faecium are the common isolates in these cases. Only three specimens in the present study had Enterococcus, one of which had it as the only organism. Bacterial combinations in root canals may be more pathogenic than individual strains . Therefore, it is important to determine the association of bacterial combinations with clinical signs and symptoms or treatment outcome, as well as the association of certain microorganisms with each other. In addition to the association of the combination of F. nucleatum and Streptococcus spp. with symptoms, discussed before, certain other trends in bacterial associations were evident . Table shows the OR extremes infinity and 0 for a number of cases due to the high prevalence of some organisms (e.g., F. nucleatum) or the paucity of others (e.g., P. intermedia). Because of the small number of positive identifications in a number of cases and the relatively small overall sample size, these data are presented because they reveal possible trends and should be further investigated. The cloning and sequencing of amplicons from one specimen obtained by PCR with universal primers specific for bacterial 16S rDNA yielded the sequence of an organism that has not hitherto been described in the root canal environment. Although the sequence is associated with the Olsenella clade by neighbor-joining analysis, it is divergent from known sequences and may be either a new species of Olsenella or a new genus of bacterium that is related to Olsenella. The association of specimen SP05 with the Olsenella sequences is supported by a relatively weak bootstrap value of 75. Olsenella uli was recently shown to be present in the subgingival flora of patients with refractory periodontitis and acute necrotizing ulcerative gingivitis . Additional tests must be conducted to firmly establish the taxonomic placement of this novel organism. The use of the universal bacterial primer pair not only provided screening information on the presence of bacteria within the specimens but also allowed us to determine the presence of a sequence from a hitherto unknown organism in one specimen that did not have any of the 10 putative organisms tested for. Molecular analysis of 16S rDNA sequences is generally supplemented by culture and biochemical analyses to determine a new species or genus of bacterium and to place the organism taxonomically. However, molecular analysis alone may be sufficient to identify organisms that are uncultivable from samples and allow to us to predict their physiology or pathogenesis by phylogenetic associations. It is logistically difficult to perform cloning and sequencing similar to what was done with specimen SP05 with a large number of specimens. A recent study used restriction fragment length polymorphism analysis to screen 50 to 100 clones into which DNA amplified from eight specimens from infected root canals had been inserted . Clones with similar restriction fragment length polymorphism profiles were grouped, and only one representative from each group was sequenced. That study reported on a number of other organisms that have rarely been reported or not previously reported from de novo or refractory root canal infections. Future studies should be directed at expanding the patient population, its clinical characteristics, and the diversity of microorganisms identified in root canals with necrotic pulp. It is also essential to expand research by using cloning and sequencing experiments in order to explore the presence of root canal microorganisms that may be uncultivable or that may not be on a preconceived list of microorganisms commonly associated with endodontic infections. TABLE 5 | OR analysis of the association of specific microorganisms with other microorganisms Backmatter: PMID- 12202611 TI - Evaluation of a Novel Commercial Enzyme-Linked Immunosorbent Assay Detecting Coxiella burnetii-Specific Immunoglobulin G for Q Fever Prevaccination Screening and Diagnosis AB - A novel commercially available enzyme-linked immunosorbent assay (ELISA) for prevaccination screening and diagnosis of Q fever (PanBio Coxiella burnetii immunoglobulin G [IgG] ELISA) was compared to the complement fixation test (CFT), and the indirect fluorescent-antibody test (IFAT) was used to resolve discrepant results between the other two tests. A total of 214 serum samples was tested. The ELISA demonstrated a specificity of 96% (46 of 48 samples) and a sensitivity of 71% (95 of 134 samples). Of the six serum pairs showing CFT seroconversion, three pairs showed a corresponding ELISA seroconversion. No cross-reactivity was observed in the ELISA with serum samples from patients with mycoplasma, brucella, and chlamydia infections. One of the 13 patients with leptospirosis demonstrated a positive result in the ELISA but not in the CFT or the IFAT, and Legionella pneumophila serogroup 4 antibody was found in one of the two sera that were false-positive by ELISA. The results presented in this study suggest that the PanBio Q fever IgG ELISA is a specific alternative method for prevaccination testing and an aid for the diagnosis of Q fever. This test is suitable for use as a screening assay, with CFT and/or IFAT used to confirm negative results. Keywords: Introduction : Q fever is the most common occupational zoonotic disease of livestock handlers and abattoir workers in Australia. The disease is usually acquired by inhalation of contaminated aerosols from animals, mainly cattle, sheep, and goats, infected with the causative agent, Coxiella burnetii. Q fever usually presents as an influenza-like illness, but asymptomatic infection and shedding of the organism into products of conception may occur. Occasionally, a chronic disease form, subacute endocarditis, may develop months or years later . Other chronic complications are granulomatous changes in liver, lesions in other organs, and the post-Q fever fatigue syndrome. Due to possible severe local or systemic reactions, prevaccination screening to assess prior exposure before vaccination is mandatory for occupational groups at risk of Q fever infection. Adverse reactions are rare (<0.05%) in subjects who are screened to assess prior exposure. Even though 25 to 50% of abattoir workers have immune markers after previous clinical or subclinical infection, the risk of adverse reaction in this group is much higher . Screening comprises serological tests for antibodies and a skin test for cellular immunity. Immunoglobulin G (IgG) antibodies may persist for 10 or more years, as measured by the indirect fluorescent-antibody test (IFAT) and the enzyme-linked immunosorbent assay (ELISA), but may occasionally fall below detectable levels over a long period of time . After an attack of Q fever, complement fixation test (CFT) antibodies fall to low levels some 3 years postillness and eventually to undetectable levels . To take account of this eventuality, a skin test was also performed to determine cell-mediated immunity before vaccination. Similarly, as a safeguard against incorrectly performed skin tests, an antibody test was done. Although the skin test, if performed correctly, is highly sensitive for subjects who have previously been infected with C. burnetii, only about 60% of vaccinated subjects will subsequently develop a positive skin test reaction. If either the antibody test or the skin test is positive, the vaccine must not be given . The traditional serological methods for assessing C. burnetii antibody status have been the CFT and IFAT . Both tests are subjective and are not standardized between laboratories. They are also inconvenient for large-scale screening and cannot be automated . These limitations led to the development of ELISAs that detected antibodies to C. burnetii , including a commercial ELISA (PanBio, Brisbane, Australia) for the detection of IgG antibodies . In this study, we compared the PanBio C. burnetii (Q fever) IgG ELISA to the CFT with sera from patients with past or acute Q fever or other infections and used the IFAT to resolve discrepant results between the other two tests. A total of 214 serum samples was included in this study. Of these, 78 specimens were single-serum samples from subjects being investigated for Q fever prevaccination immunity, 92 specimens were from patients investigated for Q fever infection, and 6 paired sera were from patients showing Q fever CFT seroconversion. An additional 32 convalescent-phase sera from patients with serologically confirmed infections other than Q fever were also tested. They comprised sera from patients with infections due to Mycoplasma pneumoniae (n = 6), Chlamydia psittaci (n = 7), Legionella sp. (n = 4), Leptospira sp. (n = 13), and Brucella sp. (n = 2). All 214 serum samples were tested by the PanBio Q fever IgG ELISA according to the manufacturer's instructions. Sera were diluted 1/100 in the serum diluent provided, and 100 mul of each diluted sample was transferred to microwells coated with C. burnetii whole-cell phase II antigen and incubated for 30 min at 37C. The microwells were then washed six times with phosphate-buffered saline (PBS) containing 0.05% Tween 20. After washing, 100 mul of horseradish peroxidase-conjugated anti-human IgG was added to each well and incubated for another 30 min at 37C. The microwells were again washed six times, and 100 mul of tetramethylbenzidine was pipetted into each well. After 10 min, this reaction was stopped by the addition of 100 mul of 1 M phosphoric acid. The microwells were then read in a microtiter plate reader at a wavelength of 450 nm. The results were determined by comparison with a provided IgG reference serum sample which contains a borderline level of Q fever IgG phase II antibody (cutoff calibrator). A positive sample was defined as having a sample absorbance/calibrator absorbance ratio (ELISA ratio) of >=1.0; a negative sample had a ratio of <1.0. IFAT was done as previously described . Phase II antigen (Nine Mile strain; Commonwealth Serum Laboratories, Melbourne, Australia) was diluted, dropped onto the wells of a glass microscope slide, allowed to dry, and fixed with acetone. Five fourfold dilutions of serum (from 1/10 to 1/2,560) in PBS were reacted with antigen on the slides for half an hour at 37C and then washed with PBS. Bound antibody was then detected via a 30-min incubation with fluorescein-labeled sheep anti-human IgG F(ab')2 fragment conjugate (Amersham, Melbourne, Australia). After the slides were washed and dried, they were mounted with a coverslip and examined by using an incident-light fluorescence microscope (Carl Zeiss, Oberkochen, Germany). Antibody titers were defined as the inverse of the highest dilution with definite staining of C. burnetii membranes. A positive IgG result was defined as having an endpoint titer of 10 or greater. CFT was performed as previously described . After the optimal dilutions of C. burnetii phase II antigen, complement, and hemolysin were determined via checkerboard titrations, serial dilutions of serum were prepared in Veronal-buffered saline and 2 U each of antigen and guinea pig complement were added. After an overnight incubation at 4C, sensitized sheep cells (2%) were added and incubated for 45 min at 37C with intermittent shaking. The highest dilution with >=75% fixation was defined as the endpoint . A positive result was defined as having an endpoint titer of >=2.5 for prevaccination sera and >=4.0 for diagnostic specimens . Analysis of variation was used to compare the mean ELISA ratios for different CFT titers. Receiver operator curve (ROC) analysis was performed to compare sensitivity and specificity at different cutoff values . The cutoff for optimal assay performance was determined by using two-graph ROC analysis . Statistics were performed by using InstatR (Graphpad Software Inc., San Diego, Calif.). Of the 214 serum samples, 184 were tested in the Q fever CFT with phase II antigen. These included all sera submitted for investigation of Q fever immunity and infection and two positive sera from the specificity panel. Any serum that showed discrepant results was retested by ELISA and tested in the Q fever phase II IgG IFAT, which is the reference method. In those samples with conflicting CFT and IFAT results, the latter was used to classify the sample. A CFT antibody titer of 2.5 or greater, a positive IgG IFAT result with a titer of 10 or greater, and a positive IgG ELISA result against phase II antigen are suggestive of recent or past infection. In the comparison with the CFT, and with the use of IFAT as the reference method to classify discrepant results, the Q fever IgG ELISA was found to have a specificity of 96%. In contrast, the sensitivity was only 71% . This low sensitivity means that, if the ELISA is used for Q fever prevaccination screening, then there is a risk that some meat workers who have a negative Q fever IgG ELISA result, and in whom the skin test is also negative, may be inadvertently vaccinated. Such vaccinees have a low risk of developing a severe local or systemic reaction. The combined prevaccination antibody and skin test is not an absolute test of immunity. It is done primarily to exclude those individuals who may develop severe reactions when vaccinated at the inoculation site . Minor or threshold levels of immunity or sensitization to C. burnetii may not be detected by the serological and skin tests but only by lymphocyte stimulation testing . TABLE 1 | Comparison of Q fever CFT and IFAT with PanBio IgG ELISA (n = 182) A correlation was shown between the individual Q fever IgG ELISA assay values and the CFT results with the sera submitted for prevaccination screening (r = 0.634, P < 0.0001) and for investigation of active infection (r = 0.2965, P = 0.0012) . An analysis of the distribution of index values in the ELISA showed that 28 of the 39 sera (71.8%) with false-negative ELISA results had absorbance values within 50% of the cutoff. However, lowering the cutoff ratio of the ELISA below 1.0 (F value, 166.73) did not significantly improve its performance. Although the highest F value (174.50) was obtained with a cutoff of 0.7, the specificity was reduced to 91.7%. As the specificity of the test is clinically more important in prevaccination screening, it is preferable to use a cutoff of 1.0 or 0.9, which results in a specificity of 96% and a sensitivity of 71 or 77%, respectively . The low sensitivity of the IgG ELISA for the diagnosis of recent infection using paired sera is of lesser concern. This is because the formation of IgM antibodies is transient and may precede that of IgG antibodies ; consequently, a specific IgM IFAT or IgM ELISA is preferable for the diagnosis of Q fever infection . FIG. 1. | Comparison of Q fever IgG ELISA ratios with CFT titers in sera for prevaccination screening (A) and investigation of Q fever infection (B). Comparison of Q fever IgG ELISA ratios with CFT titers in sera for prevaccination screening (A) and investigation of Q fever infection (B). Mean ELISA ratios are shown by horizontal bars. The cutoff values (ratio = 1.0) are shown by broken lines. Note that in panel A two sera that had CFT titers of <2.5 had IFAT titers of >=10 and that in panel B one serum sample that had a CFT titer of <4.0 had an IFAT titer of >=10. pos, positive. Of the six serum pairs showing CFT seroconversion or rising titers, three pairs showed a corresponding ELISA seroconversion . TABLE 2 | Detection of Q fever in six paired sera No cross-reactivity was observed in the ELISA with serum samples from patients with mycoplasma, brucella, and chlamydia infections. However, cross-reactivity was suggested in a sample from one of the 13 patients with Leptospira interrogans serovar Pomona infection, which gave a false-positive ELISA result. Samples from one patient with confirmed Mycoplasma pneumoniae infection and one with confirmed Leptospira interrogans serovar Hardjo infection were repeatedly ELISA reactive and were confirmed to be positive by CFT and IFAT. This suggests that these patients had past Q fever infections. In a recent study of the PanBio Leptospira IgM ELISA , sera from 3 of 34 patients with Q fever infections were reactive, although it could not be determined whether this was due to cross-reactivity or persistent antibody from a past leptospiral infection. Legionella pneumophila serogroup 4 antibody was found in one of the two sera giving false-positive ELISA results. This may represent cross-reactivity, as certain C. burnetii epitopes have extensive homology with proteins from other prokaryotes . Further studies with larger sample sizes are needed to validate these preliminary findings. The frequency distribution of the absorbance ratios was not bimodal and showed no clear discrimination between positive and negative ELISA values . The dynamic range of positive ELISA values was largely restricted to samples with lower absorbance ratios (between 1.0 and 2.64) , which suggests that some positive values could be incorrectly categorized as negative. Some changes to the ELISA components may be needed to improve the bimodal distribution of positive and negative values and extend the dynamic range of positive values. FIG. 2. | Frequency distribution of PanBio Q fever IgG ELISA absorbance ratios. Frequency distribution of PanBio Q fever IgG ELISA absorbance ratios. Absorbance ratios of <=1.0 are negative (filled columns) and those of >=1.0 are positive (open columns). The results presented in this study suggest that the PanBio Q fever IgG ELISA is a specific alternative method for prevaccination testing and the diagnosis of Q fever. It provides a standardized method with a total incubation time of 70 min, is suitable for large-scale screening, and has the potential for automation. However, this test is suitable as a screening assay provided that CFT and/or IFAT is used to confirm negative results. Backmatter: PMID- 12202593 TI - Characterization of Hepatitis C Virus Quasispecies by Matrix-Assisted Laser Desorption Ionization-Time of Flight (Mass Spectrometry) Mutation Detection AB - Hepatitis C virus (HCV), the causative agent of hepatitis C, frequently causes chronic infection. The mechanisms of viral persistence continue to be the object of investigation. An important aspect of HCV chronic infection is the quasispecies nature of the viral population, which has been particularly well documented in the hypervariable region 1 of the E2 glycoprotein. Recent studies show that characterization of the quasispecies diversity at the amino acid level can help to predict the outcome of HCV infection. Currently the accurate characterization of HCV quasispecies requires the cloning of PCR products, followed by the sequencing of many clones. In this study we present a new method to characterize HCV quasispecies, based on in vitro translation of the amplicons, followed by mass spectrometry analysis of the resulting peptide mix. The assay was used on reference HCV samples and on clinical samples. In principle, this method could be applied to other chronic viral infections in which quasispecies play a role. Keywords: Introduction : The consequences of hepatitis C virus (HCV) infection are well known and include chronic infection in as many as 85% of cases, of which up to 24% will progress to liver cirrhosis, and development of hepatocellular carcinoma . Ever since the demonstration of chronic infections caused by HCV, the mechanisms of viral persistence have been the subject of investigation. A related problem is the lack of protection against reinfection following successful clearance of the primary infection, in both humans and chimpanzees, in spite of the development of both humoral and cell-mediated immunity . The quasispecies nature of HCV in a given host, i.e., the presence of several distinct, but closely related mutants of HCV that are constantly changing due to the low fidelity of the RNA polymerase , may contribute to HCV pathogenesis, for example, through generation of immune escape mutants. In this regard, the 31-amino-acid region at the N terminus of the E2 glycoprotein, referred to as hypervariable region 1 (HVR1), has been particularly well studied, because most mutations accumulate in this domain. Interestingly, mutations in HVR1 introduced significant changes in its predicted secondary structure . In addition, HVR1 has been shown to contain epitopes for neutralizing antibodies . A convincing demonstration of the role of quasispecies diversity of HVR1 in HCV pathogenesis was published recently by Farci and coworkers . In that study it was shown that the outcome of HCV infection is predicted by the change in diversity of HVR1 amino acid sequences at the time of seroconversion, an increase in HVR1 diversity being associated with chronicity. Furthermore, a larger number of nonsynonymous mutations occurring in HVR1 were documented in cases of progressive hepatitis. This is consistent with the demonstration that some mutants in HVR1 are indeed immune escape mutants . The greater variation in amino acid sequence of the HVR1 in the viral population and the progression of disease are thought to reflect the inability of the immune system to contain viral variation . In practice, genetic changes in HCV must be demonstrated by PCR amplification of the HVR1-coding region of the HCV genome from samples obtained before and after seroconversion, followed by cloning and sequencing of at least 10 clones from which the corresponding amino acid sequences are determined . This procedure would entail enormous logistical problems for a clinical microbiology laboratory. In order to address this issue, we have developed a more-streamlined mutation detection method for the HVR1 region of HCV that uses matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometric analysis of in vitro-synthesized peptides as previously described . Amplicons are generated by reverse transcription (RT)-PCR using RNA template extracted from serum samples, and these amplicons are in turn used as a template to synthesize tagged peptides in a coupled in vitro transcription-translation reaction. The tagged test peptides are then purified using the tag prior to mass determination by mass spectrometry. Variations in the nucleic acid sequence of the virus translate into mass shifts in the encoded peptides. A major advantage of this method for quasispecies analysis, in addition to ignoring silent mutations that cannot affect immunogenicity, is that MALDI-TOF mass spectrometry can detect mutant peptides present at a low frequency in a mixture, which is not possible by direct DNA sequencing of the amplicon, hence the need for the tedious process of sequencing multiple cloned PCR products MATERIALS AND METHODS : HCV strains and plasmids. | The HCV strain H-77 in the form of a serum aliquot; the HCV infectious clone pCV-H77C of strain H-77, genotype 1a ; and the infectious clone pCV-J4L6S of strain J4, genotype 1b , were obtained from R. H. Purcell and Jens Bukh, Hepatitis Viruses Section, Laboratory of Infectious Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health. Primers. | Primers were designed to bracket the segment encoding the HVR1, targeting regions conserved between the sequences of pCV-H77C and pCV-J4L6S. Primer pairs were designed for performing nested PCR. The outer pair consisted of primer HCPEP-1 (5'GGTTCTGATTGTGCTGCTACTATTTGC 3'), whose sequence is homologous to the end of the region coding for the C terminus of E1, just ahead of the E1-E2 junction, and HCPEP-2 (5' CTATTGATGTGCCAACTGCCG 3'), whose sequence is homologous to the segment of the E2 gene just after the end of the HVR1 coding segment. The outer pair generates an amplicon of 161 bp. Several different primers were used for the second round of PCR. The sense primer HCFLAG-3 consisted of a restriction site upstream of a T7 promoter consensus sequence; filler DNA to allow for ribosomal scanning; a Kozak consensus sequence; and a start codon in frame with the segment encoding for the Flag tag (Asp-Tyr-Lys-Asp-Asp-Asp-Asp-Lys) followed by a segment (in frame) homologous to HCV sequence, overlapping with HCPEP-1, and internal to it by 3 nucleotides. The antisense primer HCFLAG-4 contained a stretch of 15 Ts to generate a poly(A) tail during transcription, a stop codon in frame with a segment encoding the Flag tag and in frame with a segment homologous to HCV sequence, overlapping with HCPEP-2 and internal to it by 5 nucleotides . HCFLAG-3 and HCFLAG-4 generate an amplicon of 257 bp. For both HCFLAG-3 and HCFLAG-4, versions of these primers without the Flag coding segment were also synthesized and used in some experiments. Finally, primer HCFLAGALA is a modification of HCFLAG-3, containing a codon for Ala between the Met codon and the Flag coding segment . Synthetic peptide. | The synthetic peptide MVLLLFAGVDAETHVTGGNAGRTTAGLVGLLTPGAKQNIQLINTNGSWHIDYKDDDDK was obtained from the Peptide Synthesis Facility, Advanced Protein Technology Center, Hospital for Sick Children. The sequence of the peptide corresponds to that expected from the translation of an amplicon obtained from the infectious clone pCV-H77C, with primers encoding for the Flag tag at the C terminus only. PCR. | When using plasmid DNA as a template, PCR with the outer primer pair was performed in a total volume of 100 mul and included 50 pmol of each primer, a 200 muM concentration of each deoxynucleoside triphosphate, 2.5 mM MgCl2, 0.5 mul of AmpliTaq Gold (Applied Biosystems Inc., Framingham, Mass.), and the buffer supplied by the manufacturer. Reactions were performed in thin-walled PCR tubes (Stratagene), overlaid with 100 mul of mineral oil, in a Robocycler-40 thermal cycler (Stratagene) with the following parameters: one cycle consisting of denaturation at 95C for 10 min, annealing at 50C for 1 min, and elongation at 72C for 1 min 30 s, followed by 35 cycles with denaturation at 95C for 1 min, annealing at 50C for 1 min, and elongation at 72C for 1 min 30 s. Nested PCR with the inner pair was performed by transferring 10% of the first PCR mix into a new PCR mix prepared essentially as described above, except for the use of an inner pair of primers and adjustment of the buffer and MgCl2, taking into account the contributions from the first round. Cycling parameters were as described above, except for the annealing temperature which was changed to 67C. Precautions against PCR contamination. | PCR reagents were prepared before each assay in a master mix that was then aliquoted. The preparation of the master mix, the extraction of the DNA and addition of the template to the PCR, and the thermal cycling were performed in three different, well-separated rooms, each with their dedicated set of micropipettors and gowns. General precautions against contamination, including systematic use of aerosol-barrier protected pipette tips, frequent changes of gloves, and frequent decontamination of surfaces with UV light and sodium hypochlorite, were strictly adhered to. RT-PCR. | HCV RNA was extracted from H77 serum using the Trizol reagent (Life Technologies) as described previously . The RNA pellet was resuspended in 10 mM dithiothreitol and 5% (vol/vol) RNasin (20 to 40 U/mul; Promega) in nuclease-free, double-distilled water (ddH2O). The RNA pellet was heated at 65C for 2 min and put on ice; 10.5 mul of an RT mix consisting of 0.5 mul of RNasin, 2 mul of 10x PCR buffer, 2 mul of 25 mM MgCl2, 2 mul of deoxynucleoside triphosphate (10 mM each), 3 mul of 10 muM HCPEP-2 primer, and 1 mul of avian myeloblastosis virus reverse transcriptase (5 to 10 U/mul, Promega). The reaction mixture was incubated at 42C for 1 h and then added in totality to a PCR mix consisting of 8 mul of 10x PCR buffer, 8 mul of 25 mM MgCl2, 5 mul of each primer stock solution (10 muM) from the outer pair, 0.5 mul of AmpliTaq Gold, and 53.5 mul of ddH2O. Nested PCR was then set up and performed as described above. In vitro translation. | In vitro translation was performed in a reticulocyte lysate system by transferring 7 mul of the second-round PCR mixture to 40 mul of master mix of the TNT T7-Quick for PCR DNA (Promega), to which were added 1 mul of 1 mM methionine and 2 mul of ddH2O, and incubating the reaction at 30C. Incubation times of 10, 30, 60, and 90 min were studied. Magnetic bead preparation. | Biotinylated anti-Flag antibody M2 (Sigma) was bound to magnetic beads coated with streptavidin (Dynabeads M-280; Dynal), following manufacturer's recommendations, by mixing 2.4 mg of magnetic beads with 0.1 mg of antibody. The magnetic bead-antibody preparation was resuspended in 200 mul of phosphate-buffered saline with 0.1% bovine serum albumin and kept at 4C. Peptide purification. | Peptide purification was performed by adding 10 mul of the magnetic bead-antibody preparation to the in vitro-translation reaction and incubating 10 min at room temperature. The beads were then separated with a magnet, washed 3 times with ammonium bicarbonate 25 mM and three times with ddH2O. Finally, beads were resuspended in 20 mul of 0.1% trifluoroacetic acid to elute the bound peptides and then were removed with a magnet. The eluted peptides were recovered in a microcentrifuge tube and lyophilized in a vacuum centrifuge (Speed-Vac). The purified peptide were then resuspended in 3 mul of an alpha-cyano-4-hydroxycinnamic acid-saturated solution in 50% acetonitrile and 0.1% trifluoroacetic acid, prior to MALDI-TOF analysis. MALDI-TOF mass spectrometry. | Mass spectrometry was performed at the Mass Spectrometry Laboratory, MMRC, Faculty of Medicine, University of Toronto. MALDI-TOF mass spectrometry analyses were carried out by using an Applied Biosystems Voyager-DE STR mass spectrometer (Applied Biosystems Inc.), equipped with a pulsed UV nitrogen laser (wavelength, 337 nm; 3-ns pulse) and a dual microchannel plate detector. For protein molecular weight determination, mass spectra were acquired in linear-DE mode, acceleration voltage was set to 20 kV, grid voltage was set at 94% of the acceleration voltage, guide wire voltage was set at 0.050%, delay time was set at 175 ns, and the low mass gate was set at 1,000 Da. The protonated ions of a mixture of peptides (molecular mass, 1,294.69 to 5,734.59 Da) were used for external calibration. For sample analysis, 1.5 mul of sample solution was applied on the MALDI sample plate. Mass spectra were recorded after evaporation of the solvent and processed using Data Explorer software for data collection and analysis. Genotyping of HCV from clinical samples. | RNA was extracted from serum samples as described above. A segment of the HCV genome at the Core-E1 junction was amplified by nested RT-PCR with the primer pairs 493S_H77-978R_H77 and 502S_H77-975R_H77 as described previously . The amplicon was sequenced using the primer 502S_H77 as the sequencing primer. This was performed by the DNA Sequencing Facility, Center for Applied Genomics, Hospital for Sick Children. The genotype was then determined by phylogenetic analysis, comparing the sequence with that of reference strains . This was performed using the Clustal X for Windows (version 1.18) program and the Treeview for Windows (version 1.5.2) program . FIG. 1. | Some of the primers used to generate amplicons suitable for in vitro translation. Some of the primers used to generate amplicons suitable for in vitro translation. Primer HCFLAG-3 is a sense primer containing an EcoRI restriction site upstream of a T7 promoter core sequence, followed by filler DNA (including three Gs immediately after the T7 core sequence) to allow for ribosomal scanning, a Kozak consensus sequence, a Flag tag (DYKDDDDK) coding region, and a region homologous to HCV sequence, in frame with the Flag coding region. Primer HCFLAG-4 is an antisense primer containing a stretch of 15 Ts to generate a poly(A) tail at the time of transcription, a stop codon in frame with a Flag coding region, and a region homologous to HCV sequence. Primer HCFLAGALA is a modified HCFLAG-3 with an additional codon encoding for alanine. RESULTS : Peptide purification and MALDI-TOF analysis. | Our initial design was to have peptides synthesized with the Flag tag at the C termini, unlike the earlier work done on tagged, in vitro-synthesized peptides that used an amino-terminal tag . A C-terminal tag was used to avoid the possibility of purifying peptides synthesized from mRNA that is not full length. The disadvantage of this approach is that C-terminally tagged peptides cannot be used to detect mutations that result in premature truncation of the peptide. However, the HCV genome has a single large open reading frame that encodes a polyprotein, and any premature truncation would generate a replication-defective virus. The present assay is designed to detect mutations resulting in amino acid substitutions; thus, a C-terminal tag is acceptable. To test our purification method, a synthetic peptide with a predicted mass of 6,125 Da was made whose sequence was derived from residues of the pCV-H77C infectious clone fused to a C-terminal Flag tag. After mixing either 100 or 10 pmol of the peptide in either phosphate-buffered saline or TNT-T7 Quick master mix, we could detect the peptide by mass spectrometry following magnetic bead purification (data not shown). Even after incubation for 90 min at 30C in the TNT mix, the peptide could still be recovered; no cleavage products were observed. Optimization of in vitro transcription-translation. | We used a PCR product obtained from the pCV-H77C infectious clone that should generate the same peptide as the synthetic peptide described above. We were unable to recover a peptide with the expected mass of 6,125 Da, although a peptide of approximately 2,394 Da was consistently recovered in these experiments that was not present in negative controls lacking template. This was interpreted to be the result of a cleavage sustained by the nascent peptide to generate the Flag tagged C-terminal fragment IQLINTNGSWHIDYKDDDDK, which has a predicted mass of 2,391 Da. Experiments with primers designed to yield a peptide with a Flag tag at the N terminus were even less successful in that we could not retrieve any peptides. However, when using the primer pair HCFLAG-3 and HCFLAG-4, designed to yield a peptide with the Flag tag at both termini, we consistently recovered a peptide of approximately 7,162 Da . We found no improved yield by incubating longer than 10 min. The mass observed was approximately 42 Da more than the 7,120 Da predicted from the pCV-H77C sequence, the two Flag tags, and the amino-terminal methionine. The precision of the mass spectrometer used in these experiments was expected to be within 0.1%. This difference of 42 Da would be well accounted for by the expected N-terminal acetylation of the peptide . Using these primers, we could also obtain a peptide of the expected mass (including acetylation) with an amplicon obtained from the pCV-J4L6S plasmid (data not shown). In several of these experiments, a second peak approximately 16 Da apart from the first peak was observed , corresponding, in all likelihood, to oxidized methionine . The 5' primer HCFLAGALA was then used to replace HCFLAG-3, and was designed so that the synthesized peptide would begin by the sequence Met-Ala, ensuring the posttranslational cleavage of methionine and the acetylation of the residual peptide . Experiments with amplicons obtained with this primer from the pCV-H77C plasmid confirmed this hypothesis: we consistently recovered peptides with a mass of approximately 7,102 Da which is the mass predicted from the amino acid sequence encoded by the amplicons, followed by methionine cleavage and acetylation . In contrast to the experiments presented above, however, we obtained our best results after an incubation of 60 min at 30C, suggesting that methionine cleavage, which occurs before completion of the nascent peptide , delayed the peptide synthesis. In agreement with our hypothesis concerning N-terminal methionine oxidation, the extra peak having an additional 16 Da was no longer observed when alanine was used to program the removal of the N-terminal methionine . MALDI-TOF analysis of reference HCV reagents. | In addition to the MALDI-TOF analysis of clone pCV-H77C we also used our procedure with the clone pCV-J4L6S. We consistently obtained peptides of the predicted mass of approximately 7,206 Da . Next, we subjected to transcription-translation and mass spectrometry a mix of amplicons obtained from the two infectious clones, and we consistently obtained peptide mixtures of the two expected masses . Lastly, we performed RT-PCR on HCV RNA from the H-77 serum aliquot, whose quasispecies population has been extremely well characterized . As shown in Fig. , we could clearly demonstrate three peptides species with observed masses of 7,052, 7,105, and 7,126 Da, respectively. MALDI-TOF analysis of clinical samples. | Results obtained from clinical samples are shown in Fig. . The first patient presented at our institution at 26 months of age; this male patient was referred for known hepatitis C infection. His mother was also suffering from chronic hepatitis C. The patient was asymptomatic, and his physical exam was unremarkable. Laboratory analysis revealed an elevated serum alanine aminotransferase (ALT) level at 111 U/liter (normal, 0 to 40 U/liter). Repeat clinical examination 6 months later revealed no changes, but the ALT level was within normal values, at 26 U/liter. This patient was infected with HCV genotype 1b. MALDI-TOF analysis on the initial serum sample demonstrated the presence of two main variants in the HCV quasispecies . Remarkably, 6 months later these had disappeared and were replaced by new variants . The second patient was 21 months old at presentation in our institution. This female patient was referred for possible mother-to-infant transmission of HCV. She had been asymptomatic and her physical exam was unremarkable. However, her laboratory analysis revealed a serum ALT level of 244 U/liter. Six months later, the clinical examination remained unchanged, the ALT level remained elevated at 205 U/liter. A percutaneous liver biopsy performed at the time revealed normal liver architecture, no fibrosis, but some evidence of inflammation limited to focal collections of inflammatory cells in the sinusoids with scattered hepatocyte necrosis. This patient was infected with HCV genotype 1a. MALDI-TOF analysis on the initial sample showed a very homogeneous population, with only a single HCV variant demonstrated . Six months later, the HCV quasispecies had not changed . FIG. 2. | Mass spectrometry analysis of peptides obtained by translation of an amplicon obtained from clone pCV-H77C with primers HCFLAG-3 and HCFLAG-4. Mass spectrometry analysis of peptides obtained by translation of an amplicon obtained from clone pCV-H77C with primers HCFLAG-3 and HCFLAG-4. Based on the sequence of pCV-H77C, the predicted mass of the resulting N-terminal acetylated peptide should be 7,162 Da. In several experiments, as illustrated here, a second peak approximately 16 Da apart was observed and was interpreted as resulting from oxidation of methionine. FIG. 3. | Mass spectrometry analysis of peptides translated from amplicons obtained with primers HCFLAGALA and HCFLAG-4. Mass spectrometry analysis of peptides translated from amplicons obtained with primers HCFLAGALA and HCFLAG-4. (A) Peptide obtained from template pCV-H77C. The predicted mass of the peptide, after methionine cleavage and N-terminal acetylation, is 7,102 Da. (B) Peptide obtained from template pCV-J4L6S. The predicted mass, after methionine cleavage and N-terminal acetylation, is 7,206 Da. (C) Peptides obtained from mixing the amplicons of pCV-H77C and pCV-J4L6S. (D) Peptides obtained from the H-77 serum aliquot. FIG. 4. | Mass spectrometry analysis of clinical samples from patients with hepatitis C. Mass spectrometry analysis of clinical samples from patients with hepatitis C. (A and B) These two samples were taken 6 months apart from a patient infected with an HCV strain of genotype 1b. A complete change of the HCV population was observed during that period. (C and D) These two samples were taken 6 months apart from a patient infected with an HCV strain of genotype 1a. No changes in the HCV population could be observed. DISCUSSION : In this study we have developed a procedure to evaluate the HVR1 peptide diversity in the HCV population of a patient's serum sample, without the need of cloning and sequencing of many clones. This is a conceptually straightforward application of mutation detection using MALDI-TOF , but several problems specific to HCV HVR1 had to be addressed, some of which were unexpected. The power of MALDI-TOF to detect mutations relies on the changes in mass caused by amino acid substitutions. Single-amino-acid substitutions produce mass alterations ranging from 0 to 186 Da. If the peptide is short enough, most of these changes can be detected: for example in a 10-kDa peptide, all but 14 of 380 possible amino acid substitutions should be detectable with a mass accuracy of 200 ppm . However, the effective mass range of the MALDI instrument imposes constraints on the size of the HVR1 test sequence that is to be translated to roughly 150 bp. Unfortunately, sequence comparison between all HCV genotypes failed to reveal conserved sequence across all genotypes within such a short region around the HVR1 that could be used as primer binding sites. In this study we therefore used primers that were designed for genotype 1 and which were shown to amplify HCV strains of genotype 1a and 1b. Generalization of the method will require the design of primers specific for each genotype. These different sets of primers might conceivably be used simultaneously in a multiplex PCR to allow amplification of any HCV genotype present in the patient sample in a single reaction. Alternatively, it might be preferable to keep the reactions separated: in cases of infection by more than one genotype it would allow for monitoring the evolution of HCV quasispecies for each genotype separately. We have designed our primer pairs with the capacity to do nested PCR for increased sensitivity, but of course when working with high titer samples (or from plasmids) one round of PCR with the inner pair may suffice. Purification of the synthesized peptides was achieved through the incorporation of the Flag tag. Since the M2 monoclonal antibody can in principle bind to the Flag sequence at either the N or C terminus, the exact location did not seem to matter and our initial design was to put the Flag at the C terminus. Experiments with a synthetic H-77 peptide confirmed that purification was achieved with our method, and no cleavage of the peptide could be observed after prolonged incubation in the reticulocyte lysate. However, with translation from an amplicon we did consistently observe a cleavage, suggesting that the nascent peptide, as it was being synthesized, was exposed to proteases. In hindsight this is not so surprising, since in order to bracket the HVR1 the amplicon must contain the segment coding for the C terminus of the E1 glycoprotein, which is highly hydrophobic and provides a transmembrane anchor to the protein, causing retention of E1 in the endoplasmic reticulum rather than channeling to the Golgi . Presumably, the highly hydrophobic character of the amino terminus of this peptide results in its targeting by the proteasome, which preferentially recognizes exposed hydrophobic amino acid R groups . With the hydrophilic Flag tag on both ends, the peptide is apparently not exposed to protease activity, perhaps because the hydrophobic character of the amino terminus is altered by the highly charged residues of the Flag tag. We have, however, no good explanation for the poor results obtained when the Flag tag appeared only at the N terminus. Using the HCFLAG-3 and HCFLAG-4 primers, which direct the synthesis of the Flag tag on both ends, we consistently obtained good results and obtained peptides whose masses were essentially those predicted for clone pCV-H77C and for clone pCV-J4L6S (not shown). In addition, correct masses were observed for a mix of amplicons obtained from these two clones (not shown) and three main peptide species after translation of the amplicon from the H-77 serum (not shown). In these experiments we did observe a consistent shift from the predicted mass which coincided with the predicted increase of 42 Da caused by N-terminal acetylation . This is not a problem, since in all our experiment this enzyme mediated posttranslational modification was consistent and complete. More problematic for our purpose was the frequently observed second peak shifted by approximately 16 Da, which we attributed to incomplete and inconsistent methionine oxidation , a common chemical modification. To overcome this difficulty we redesigned the sense primer to direct the incorporation of an alanine between the methionine and the Flag tag. As predicted previously , this resulted in a posttranslational cleavage of the methionine followed by N terminal acetylation. We did not observe the 16 Da shifted second peaks when using this primer. Experiments with clones pCV-H77C and pCV-J4L6S yielded peptides of expected masses, alone or in combination . Finally, mass spectrometry measurements from the translated amplicon obtained from the serum aliquot of H-77 yielded at least three distinct species whose masses correspond favorably to those predicted from the sequence of the most frequent clones in the H-77 quasispecies : 7,102 Da for the master clone (present at a frequency of 70 per 104 clones), 7,122 for the second most common clone (with a frequency of 6 per 104 clones), and 7,056 for the third most common clone (with a frequency of 5 per 104 clones), the other 16 clones identified being present at frequencies of 4 per 104 clones or less . Thus, our procedure appears to give as much information as sequencing approximately 20 clones, since it detected the subspecies present at roughly 1 in 20 (mass, 7,056 Da). The MALDI-TOF-based assay was applied to serial samples from two patients monitored for chronic hepatitis C at our institution. The first patient presented initially with elevated ALT, a marker of liver inflammation that is a correlate of immune response against HCV . In a follow-up sample obtained 6 months later, the quasispecies population had completely changed, in all likelihood because of effective immune selection against the initial HCV variants. Numerous studies have documented the role of immune selection in the evolution of HCV quasispecies in HVR1, including the lack of evolution in patients with agammaglobulinemia (reviewed in reference ). It is interesting that at the time when the second sample was taken, the ALT level was normal. One may speculate that an immune response against the new variants had not yet been mounted. The second patient displayed a very homogeneous population, since only one HCV variant could be demonstrated . It is possible that this homogeneity is a consequence of the infection being acquired vertically, since in these cases the HCV population in the infant is usually very homogeneous . This patient also presented with elevated ALT. However, in this patient the immune response failed to eradicate the HCV variant, since 6 months later the quasispecies had exactly the same profile; at that time, the ALT level remained elevated. Currently, the significance of decreased quasispecies diversity at the amino acid level at the time of seroconversion has been demonstrated . One would have expected also that since part of the anti-HCV activity of alpha interferon is immune mediated , diversity decreases should be predictive of treatment outcomes. This has in fact been recently demonstrated . At this point in time it remains unproven whether quasispecies diversity monitoring will also be of benefit in longitudinal follow-up of patients. However, the two examples presented in this study appear similar to a recent report by Curran and coworkers , in which patients with mild disease had an HCV population whose amino acid sequences in HVR1 changed over time, whereas patients with more severe illness had static HCV populations. Although further work will be required on this point, it is clear that the MALDI-TOF assay allows for the monitoring of HCV quasispecies in the clinical setting in a meaningful way, and is expected to contribute to a better understanding of the natural history of the disease. In summary we have developed and streamlined a procedure based on MALDI-TOF mass spectrometry for the characterization of HCV quasispecies in HVR1. This assay is technically less demanding to perform than cloning and sequencing many clones, since in its current state it involves the translation of the (unpurified) amplicon using a commercially available kit, the purification of the peptide mix using a simple and rapid method, and the submission to MALDI-TOF analysis. The procedure described here could also be easily adapted to other chronic viral infections in which quasispecies play a role, for example the quasispecies observed in the V3 loop of the gp120 surface glycoprotein of human immunodeficiency virus . Backmatter: PMID- 12202606 TI - Evaluation of Dipstick Serologic Tests for Diagnosis of Brucellosis and Typhoid Fever in Egypt AB - Two dipstick assays for the detection of Brucella- and typhoid-specific immunoglobulin M, recently developed by the Royal Tropical Institute of The Netherlands, were evaluated by use of 85 plasma samples from Egyptian patients. Both dipsticks were simple and accurate rapid diagnostic assays, and they can be useful adjuncts for the diagnosis of typhoid fever and brucellosis. Keywords: Introduction : Typhoid fever and brucellosis are frequent causes of bloodstream infections in many countries . The diagnosis of these infections is challenging because they can have diverse clinical manifestations with symptoms that overlap with a wide spectrum of other diseases . Cultures of blood or bone marrow are the most definitive diagnostic methods. In many developing countries, both diseases are diagnosed on clinical grounds and treated empirically . Serology assays are often utilized as a diagnostic tool ; however, neither the Widal nor the Brucella agglutination assay is sufficiently sensitive, specific, or practical in areas of endemicity . In this study, we evaluated the sensitivity and specificity of the Royal Tropical Institute of The Netherlands (RTI) dipstick assay. Representative plasma samples were originally collected from patients admitted to five infectious disease hospitals in Egypt according to certain case definitions and laboratory tests using conventional blood culture and serological techniques . Eighty-five plasma samples were screened by Widal and Brucella agglutination assays . The remaining portions were preserved at -70C for evaluation with the dipstick assay, which was performed within 2 to 3 months postadmission. Results were related to the other findings for the purpose of this study. Among those patients, 25 had blood culture-proven typhoid fever, 25 had blood culture-proven brucellosis, 25 had acute fever but negative blood cultures and serology, and 10 had negative blood cultures but were seroreactive to Widal (n = 5) or Brucella (n = 5) agglutination. The RTI dipstick assay is based on the binding of human Brucella- or Salmonella enterica serovar Typhi-specific immunoglobulin M (IgM) antibodies with the respective antigen, which is detected with an anti-human IgM dye conjugate . IgM has been reported to appear a few days after infection, reaching a peak within 3 to 4 weeks, and then it declines gradually over a similar period of time . The test sample was incubated for 2 h at 35C. The built-in internal control band provided a check on the integrity of the detection reagent and the presence of serum. To avoid bias, investigators were blinded to the blood culture results until the dipstick assays were completed. Blood culture results were considered the "gold standard" and were used to determine the positive and negative results by dipstick assays. Of the 30 patients with laboratory-confirmed brucellosis (25 culture positive and 5 positive by tube agglutination test), 28 tested positive by the Brucella dipstick (93% sensitivity). Additionally, the Brucella dipstick was positive for 4 of 55 serum samples that were negative by culture and Brucella agglutination testing (specificity = 93%) . Of the 30 patients with a laboratory diagnosis of typhoid fever (25 culture positive plus 5 Widal positive), 27 had a positive dipstick test (sensitivity = 90%). The typhoid dipstick was positive in only 2 of 55 samples derived from patients without laboratory evidence of typhoid fever (specificity = 96%) . Table shows that readings of the Brucella agglutination test and dipstick assay were the same, recognizing titers >=160 in 92% of samples collected from culture-positive patients. However, in the case of typhoid fever, the Widal test was negative for 8 of 22 (32%) dipstick-positive specimens, suggesting a lower sensitivity or titer (<160). TABLE 1 | Sensitivity and specificity of Brucella dipstick assay TABLE 2 | Sensitivity and specificity of typhoid dipstick assay TABLE 3 | Comparison of typhoid and Brucella dipstick results with serologic assay results Six patients with blood culture-confirmed infections (three typhoid and three Brucella infections) had an initial negative dipstick test. Upon testing with a twofold dilution to rule out a prozone phenomenon, one patient with brucellosis became positive. Of the 25 patients with culture-proven typhoid fever, only 14 (56%) had Widal titers >=1:160 . Ten samples (100%) from patients with a negative blood culture but serologic evidence by Widal or Brucella agglutination reacted with the respective dipsticks. Since the typhoid dipstick assay was based on IgM detection in acute infections, antibodies peculiar to chronic carriers (anti-Vi-specific IgG) could not be detected. Passive hemagglutination or enzyme-linked immunosorbent assays have been advocated for such discrimination in epidemiological surveys . The RTI dipstick test is a simple, fast, and reliable method for the diagnosis of typhoid fever and brucellosis and confirmed results from previous studies . With sensitivity and specificity of both dipsticks in excess of 90%, the kits performed well in settings where both brucellosis and typhoid fever are endemic. Published studies suggest that blood cultures are positive in only approximately 60% of patients with brucellosis or typhoid fever . The findings that culture-negative patients were positive by dipsticks may be useful in settings where antibiotic use is high and no cultures are available. One concern noted with dipstick kits was that 16% of the patients with culture-proven brucellosis reacted with the typhoid dipstick and 8% of the samples from patients with culture-proven typhoid fever reacted with the Brucella dipstick. The reason for both dipstick tests having a positive reaction from a single patient is not clear, but in addition to simple cross-reactivity, potential options could include recent past infection with one organism and current infection with the second organism or persistence of IgM antibodies in some patients. Another interesting finding was the high correlation (100%) between serological results and the dipstick assay results from patients with clinically suspected brucellosis or typhoid fever among patients who had negative blood cultures. However, for patients with positive cultures, the sensitivity of the typhoid dipstick was found to be significantly higher than that of the corresponding serological results. The relatively low sensitivity of the Widal test in this study may have resulted from an undetectable level of IgM antibody, which was probably due to the young age of patients and/or a relatively short length of illness of the typhoid patients . For the purpose of communicable disease surveillance, the Ministry of Health in Egypt indicated that tube agglutination titers >=1/160 are to be regarded as positive for brucellosis and typhoid fever. This is based on extensive clinical and laboratory findings collected over a number of years. Consequently, the dipstick assay would not be expected to show false-positive results, given the results of positive, negative, and control samples of this study. However, the Brucella dipstick assay showed 8% cross-reactivity with typhoid patients who had no evidence of Brucella infection, but the corresponding agglutination titers were considerably lower than the cutoff. Our preliminary results are quite promising and lead us to recommend the use of the dipstick assays in the primary characterization of infection. Being specific, faster, and easier than conventional methods, they could be of value in areas where diagnostic capabilities for culture and serology are poor. Backmatter: PMID- 12202607 TI - Simple and Reliable Method for Detection and Genotyping of Hepatitis C Virus RNA in Dried Blood Spots Stored at Room Temperature AB - We describe a simple, sensitive, and reproducible method for using whole blood collected onto filter paper (dried blood spots) for detection and genotyping of hepatitis C virus RNA that can be useful in large field studies, particularly in settings where collection, preparation, storage, and shipment of samples at controlled temperature can be difficult. Keywords: Introduction : Dried blood spots (DBS) have been used worldwide for the neonatal screening of congenital disorders . Seroepidemiological studies have been conducted on DBS residual to neonatal screening to assess human immunodeficiency virus (HIV) prevalence among childbearing women . Recently, several studies were focused on DBS for detecting drug resistance mutations and for tracking global spreading of HIV type 1 subtypes in proviral HIV DNA. However, RNA is notoriously less stable, and standardization of DBS for viral RNA detection would be of great benefit for application to large field studies, since DBS collection is easy, does not require skilled phlebotomists and expert technicians, and is suitable for storage and shipment to laboratory in settings where these issues are problematic. Recently, dried plasma spots and DBS have been used for HIV RNA detection and quantification, showing good correlation with titers obtained with conventional plasma samples . However, these observations were limited to short storage at room temperature or at 37C , and a loss of viral titers occurred during storage. For hepatitis C virus (HCV) RNA detection on dried spot samples, the available data are much less exhaustive. A complete match between frozen serum and dried plasma spots, though with a loss of titers after room temperature storage, has been observed . This study was aimed at developing a simple, sensitive, and reproducible method for using DBS in HCV RNA detection and genotyping. The study complied with all relevant national guidelines and institutional policies. Residual laboratory samples of EDTA-whole blood of 39 HCV antibody (Ab)-positive and 16 HCV Ab-negative patients, undergoing routine hematological controls, were used. HCV Abs were determined by third-generation assay (Abbott Diagnostics). Among the Ab-positive patients, 34 had HCV RNA levels ranging between 9,640 and 5,100,000 IU/ml (Amplicor HCV Monitor; Roche Molecular Systems Inc.), and 5 were HCV RNA negative (Versant HCV TMA; Bayer Diagnostic Inc.). The HCV genotype was known for eight patients (four had 1b; two had 2a/2c; one had 3a; and one had 4c/4d). DBS were realized within 5 h from venipuncture by carefully spotting, in multiple replicates for each patient, 50 mul of EDTA-whole blood on SS grade 903 filter paper (Schleicher & Schuell Inc.). Two DBS from each patient were pooled and processed for each assay. HCV RNA detection was performed with both in-house reverse transcriptase PCR (RT-PCR) and transcription-mediated amplification (TMA). For RT-PCR, RNA extraction was performed with Boom technology, utilizing silica-based RNA isolation , which was purchased from Organon Teknika. Specifically, DBS were cut, placed in 9 ml of lysis reagent, rocked 2 h at room temperature, and removed; supernatants were extracted according to the manufacturer's instructions. Ten microliters out of the 50-mul RNA eluate underwent the subsequent steps. Reverse transcription was performed with 50 U of Moloney murine leukemia virus RT (Gibco BRL, Life Technologies) in 20 mul. Ten microliters of cDNA was used in the nested PCR, according to reference . The amplified products were analyzed by agarose gel electrophoresis (amplicon size, 197 bp). TMA, provided by Bayer S.p.A. Divisione Diagnostici Italia, is a recently developed HCV RNA qualitative detection method, based on 5' untranslated region-targeted magnetic separation, followed by reverse transcription and T7 RNA polymerase-driven isothermal amplification . DBS from each patient were soaked in 500 mul of distilled water supplemented with 400 mul of Target Capture Reagent and were incubated at 60C for 60 min. Then supernatants were decanted; the subsequent extraction and detection steps followed the standard TMA protocol. A total of 158 DBS pairs were tested (124 from 34 HCV patients; 24 from 19 HCV-negative individuals). On the whole, with both TMA and RT-PCR all the 124 DBS pairs from positive patients tested positive, and all the 24 DBS pairs from the HCV RNA-negative patients tested negative (sensitivity > 99%; specificity > 95%). No genotype restriction was observed. The detection limit was established by limiting dilution analysis. Specifically, blood samples from two viremic patients (viral burdens, 604,000 and 807,000 IU/ml) were serially diluted with blood from an HCV-seronegative individual. Seven fivefold serial blood dilutions were spotted to form multiple replicate DBS strips. Four replicate pairs of each dilution (24 tests per run) were tested with TMA (after 1 and 5 weeks), giving a total of 48 tests performed, and with RT-PCR. The results, shown in Table , indicate that TMA was positive in 100, 85.7, and 75% of DBS containing, respectively, 1,328, 265, and 53 IU of HCV RNA. Similar sensitivity was observed with RT-PCR (not shown). TABLE 1 | Establishment of detection limit of HCV qualitative RNA assay (TMA) applied to DBS To evaluate the cross-contamination risk, a panel of DBS strips was prepared by spotting in close vicinity and, in alternate order, 15 blood samples from HCV RNA-positive patients and 15 blood samples from HCV Ab-negative individuals. Representative RT-PCR results are shown in Fig. , indicating a complete correspondence between expected and observed results. Similar results were obtained with TMA (not shown). FIG. 1. | Detection of HCV RNA in DBS from HCV-positive and -negative individuals. Detection of HCV RNA in DBS from HCV-positive and -negative individuals. To control the risk of cross-contamination in using DBS for HCV RNA detection, DBS strips were prepared with blood from 30 different patients that was expected to be positive (uneven lines) or negative (even lines). The strips were spotted to realize a lattice of alternately HCV RNA-positive and -negative samples. DBS strips were wrapped in their individual paper covers and stored together in the same plastic bag for 1 week at room temperature before the assay. Representative results obtained with RT-PCR are shown. To test the stability of HCV RNA in DBS over time, a set of replicate DBS from 16 HCV RNA-positive patients was stored at room temperature and assayed at intervals of 2 to 4 weeks over an 11-month period. The results indicate that 100% positivity was preserved along the whole observation period. To test the suitability of DBS for HCV genotyping, a replicate set of DBS from the eight patients with known genotype, stored at room temperature for 11 months, was extracted as for RT-PCR and genotyped by using the Line Probe assay (Bayer Diagnostic Inc.). The genotype, determined in all tested DBS, was in agreement with that obtained on frozen serum samples. On the whole, our findings show that DBS, stored at room temperature for prolonged periods, are suitable for HCV RNA detection and are also useful for HCV genotyping. The assessment of HCV genotype distribution at population level may help in defining the epidemiological dynamics of HCV infection . Since HCV genotype may influence the response to antiviral treatment, severity of infection, and performance of diagnostic assays , these aspects are particularly relevant in view of the possibility of using the method described here in large field studies and for application in settings where collection, centrifugation, storage, and shipment can be difficult, as is often the case in developing countries. Backmatter: PMID- 12202605 TI - Comparison of Serological Detection Methods for Diagnosis of Ehrlichia canis Infections in Dogs AB - We determined the value of four serological assays for the diagnosis of canine monocytic ehrlichiosis by comparing them to the indirect fluorescent-antibody assay "gold standard." The specificity of Dip-S-Ticks was significantly lower than that of all of the other tests evaluated. The sensitivity of Dip-S-Ticks was significantly higher than that of Snap3Dx or the Snap Canine Combo. The sensitivity of the rMAP2 enzyme-linked immunosorbent assay (ELISA) was significantly higher than that of the Snap Canine Combo. The accuracy levels of the rMAP2 ELISA, Snap3Dx, Dip-S-Ticks, and Snap Canine Combo were 97.0, 89.8, 85.1, and 82.9%, respectively. Keywords: Introduction : Canine monocytic ehrlichiosis, caused by tick-transmitted Ehrlichia canis, has been reported in the United States and throughout most of the world, causing extensive morbidity and mortality . Clinical and hematologic abnormalities are often nonspecific during E. canis infections, and coinfections with other tick-transmitted agents such as E. chaffeensis may be common ; thus, a definitive diagnosis may be difficult to make. The indirect fluorescent-antibody assay (IFA) is the method most widely used to diagnose E. canis infection and is considered the "gold standard" . However, it can only be performed in specialized laboratories, reading of results is subjective , and it does not differentiate consistently between E. canis and E. chaffeensis infections . There is a tremendous need for other serological assays for the diagnosis of E. canis infection in dogs. A variety of serodiagnostic tests are commercially available, but the diagnostic value of many of these tests remains unevaluated. The objective of the present study was to determine the value of four serological assays for the diagnosis of canine monocytic ehrlichiosis by comparing them to the IFA gold standard. A total of 97 canine serum samples was obtained from the College of Veterinary Medicine, University of Florida, Gainesville, or the College of Veterinary Medicine, North Carolina State University, Raleigh. Eighteen IFA-positive serum samples from five dogs experimentally infected with E. canis during previous studies were used . Thirty-five serum samples were from naturally infected dogs that had clinical signs consistent with canine ehrlichiosis and positive IFA titers (>=1:40). Forty-four IFA-negative serum samples were obtained from clinically healthy dogs during well-patient visits or preinfection from experimentally infected dogs. All serum samples were tested blindly in the following serological assays. All serum samples, diluted 1/300, were tested with the rMAP2 indirect enzyme-linked immunosorbent assay (ELISA) as described previously . The samples were tested for canine ehrlichiosis with the Snap Canine Combo test kit, the new Snap3Dx assay (IDEXX Laboratories, Inc.), and the InDx canine multitest Dip-S-Ticks assay (PanBio InDx, Inc., Baltimore, Md.) in accordance with the manufacturers' recommendations. The data were analyzed by using SigmaStat, version 2.03, for Windows (SPSS Inc.) and calculated as described by Courtney and Cornell . As observed in Table , a sensitivity of 96.2% was obtained with the rMAP2 ELISA. Its specificity was 97.7%, as one false-positive reaction was detected. The latter serum sample was positive with the Dip-S-Tick assay but negative by both Snap tests. The sensitivity of the rMAP2 ELISA was significantly higher than that of the Snap Canine Combo (P = 0.001) . Also, the sensitivity of the Dip-S-Ticks was significantly higher than that of the Snap3Dx (P = 0.003) or the Snap Canine Combo (P <= 0.001). The specificity of the Dip-S-Ticks was significantly lower than that of all of the other tests evaluated (P <= 0.001). TABLE 1 | Comparison of serological assays to detect antibodies in dogs infected with E. canis TABLE 2 | Comparison of the accuracies, sensitivities, and specificities of four serodiagnostic tests with IFA as the gold standard Dip-S-Ticks are semiquantitative assays. On each stick, two windows containing different dilutions of E. canis antigen were present. One window corresponded to E. canis IFA titers of approximately 1:40 to 1:80, whereas the other window represented IFA titers of 1:5,000 to 1:10,000. All of the infected serum samples tested with IFA titers of 1:40 to 1:80 were positive in the 1:40-to-1:80 Dip-S-Ticks window. We observed that 73.9% of the serum samples having IFA titers of >=5,000 were positive in the 1:5,000-to-1:10,000 Dip-S-Ticks window. No false positives were recorded for the second test window. Twenty-eight percent of the serum samples within an IFA titer range of 1:320 to 1:2,560 were reactive in the 1:5,000-to-1:10,000 Dip-S-Tick window. One of the strengths of this study was that it used duplicate sera from the same dogs to compare different serodiagnostic tests. The assays tested in this study varied in the ability to detect E. canis antibodies. The Dip-S-Ticks assay was 100% sensitive in detecting sera from infected dogs but lacked specificity, as a high percentage of false positives was reported. This may be due to the fact that whole cells from E. canis (Jake strain) were used as antigens causing cross-reactivities and thus false-positive reactions. Semiquantitative results can be obtained with the Dip-S-Ticks assay; however, the Dip-S-Ticks titers did not always correspond to the reported IFA titers. Reading of the assay strips was difficult when recording borderline reactions and may have led to misleading interpretations and to the high percentage of false positives recorded. Also, the range between the upper and lower confidence limits was wide because of the limited number of samples available within each of the infected or uninfected groups. Previously, we demonstrated 97.2% overall agreement between the IFA and the rMAP2 ELISA for E. canis antibody detection . In this study, 96.2% sensitivity and 97.7% specificity were obtained with the E. canis rMAP2 ELISA. The rMAP2 ELISA has the advantage of being a quantitative test. Therefore, detection of recent exposure and active infection can be done by comparing paired titers. The use of recombinant proteins, such as p30 and p43, to increase the quality of the test antigen and to eliminate test subjectivity in Western immunoblot and dot blot assays was demonstrated previously . The Snap Canine Combo from IDEXX, which uses whole cells of E. canis strain Oklahoma as the antigen, is being replaced on the market by the new Snap3Dx assay, which uses recombinant analogs of the major outer membrane proteins , specifically, p30 and p30-1. According to our study, the sensitivity and specificity of the Snap3Dx assay were 15 and 2.3% higher, respectively, when recombinant proteins were used as antigens instead of E. canis whole cells. However, these differences were not statistically significant within the limited sample size. The Snap3Dx assay combines specificity and rapidity and can be used in any clinic. The rMAP2 ELISA, even though it is more time consuming and is not applicable to all laboratories, provides high sensitivity, high specificity, quantitative results, and the convenience of the ability to test many samples in a short time. Its accuracy, which reflects the chance of correctly identifying the infection status of an animal , was the highest among the assays tested. In choosing a serodiagnostic test for a laboratory, one should take into consideration many factors, including the cost of the test, its rapidity, its convenience of use, the sample load, and the predictive value of a positive or negative result. The seroprevalence of E. canis infection within different canine populations may vary dramatically. For example, in a recent survey of sick dogs from North Carolina and Virginia, the seroprevalence of E. canis was estimated to be 2.5% . On the basis of our study and an E. canis prevalence of 2.5%, the calculated predictive values of positive and negative results were both 100% for the Snap3Dx test . Therefore, if the prevalence of E. canis infection is expected to be low in the population tested, the Snap3Dx test would be an excellent test to help in the identification of such infections. FIG. 1. | Predictive values of positive (A) and negative (B) rMAP2 ELISA, Dip-S-Ticks, Snap Combo, and Snap3DX results at prevalence rates ranging from 0 to 100% with IFA as the reference standard test. Predictive values of positive (A) and negative (B) rMAP2 ELISA, Dip-S-Ticks, Snap Combo, and Snap3DX results at prevalence rates ranging from 0 to 100% with IFA as the reference standard test. Alternatively, for serodiagnosis during an outbreak or when testing dogs with clinical and/or laboratory findings consistent with canine ehrlichiosis, the chances of finding more E. canis infections are higher and the predictive values of the results obtained in such an environment would be different. Kordick and colleagues reported an E. canis prevalence of 55.6% in a Walker Hound kennel in North Carolina during an outbreak. In this case, the rMAP2 ELISA would be a good choice as its predictive value of positive and negative results would be 98.1 and 95.4%, respectively . It must be emphasized that serodiagnostic assays do not distinguish between current infection and prior exposure. In this study, specificity was determined in relation to that of IFA, the gold standard. The IFA and most of the rapid serodiagnostic tests do not differentiate among E. canis, E. chaffeensis, and E. ewingii infections in dogs, as these species are closely related and have some cross-reacting antigens . It has been observed that the clinical disease, prognosis, and potential drug efficacy associated with E. canis or E. chaffeensis infection may differ . Therefore, further development of tests for canine monocytic ehrlichiosis should address this important problem and try to eliminate cross-reactivities to provide a more accurate diagnosis. Backmatter: PMID- 12202580 TI - Rapid-Cycle PCR and Fluorimetry for Detection of Mycobacteria AB - In this study we used LightCycler PCR amplification and product detection by fluorescence resonance energy transfer probes to identify mycobacteria and differentiate between Mycobacterium tuberculosis complex, Mycobacterium avium, and other nontuberculous mycobacteria. Targeting the 16S rRNA gene, three different probes specific for mycobacteria, M. tuberculosis complex, and M. avium were constructed. As few as five genome copies of target nucleic acid were detected by the probes, illustrating the high sensitivity of the system. All 33 mycobacterial species tested but none of the closely related actinomycetes and other bacteria produced a specific fluorescence signal. A specificity of 100% was also demonstrated for the M. tuberculosis complex-specific probe and the M. avium-specific probe. Within 45 min, the LightCycler method correctly detected mycobacteria and specifically identified M. tuberculosis complex and M. avium without any post-PCR sample manipulation. In view of future clinical studies, we also constructed and tested an internal control which could be used to assure successful amplification and detection of mycobacteria. Monitoring of PCR inhibition will be essential for evaluation of this system for direct detection of mycobacteria in clinical specimens. Finally, we tested our system on sputum seeded with mycobacteria and were able to detect as few as 10 organisms. At present, this system is the fastest available method for identification and differentiation of mycobacteria from culture-positive specimens and offers an excellent alternative to previously established nucleic acid amplification-based techniques for the diagnostic mycobacterial laboratory. Keywords: Introduction : Mycobacterium tuberculosis claims more human lives each year than any other bacterial pathogen. A third of the world's population is thought to be infected with M. tuberculosis. The emergence of multidrug-resistant strains and its association with outbreaks inside and outside hospitals illustrates that rapid diagnosis is essential . In recent years, an increased incidence of tuberculosis in both developing and developed countries and a deadly synergy with the human immunodeficiency virus have been reported . Diseases caused by nontuberculous mycobacteria used to be pulmonary, confined to cervical lymph nodes, limited to skin, or in rare cases, disseminated. The rise in the incidence of nontuberculous mycobacterial disease in AIDS patients has accelerated rapidly since the first reports in 1982, with disease now being predominantly disseminated . In addition, the role of nontuberculous mycobacteria in the worsening of pulmonary disease in patients with cystic fibrosis has been documented . Mycobacterium avium, Mycobacterium intracellulare, Mycobacterium kansasii, Mycobacterium marinum, Mycobacterium fortuitum, Mycobacterium chelonae, and Mycobacterium abscessus are the nontuberculous organisms most commonly encountered in clinical practice . Thus, a diagnostic assay for mycobacteria should ideally encompass identification of both tubercle bacteria and nontuberculous mycobacteria. In the past few years, nucleic acid amplification-based techniques have become accessible to the clinical mycobacteriology laboratory. PCR protocols amplifying a large variety of chromosomal DNA have concentrated on detection of both genus-specific and M. tuberculosis complex-specific DNA regions . Genus-specific protocols target the 16S rRNA gene or the gene encoding the 65-kDa heat shock protein. Subsequent mycobacterial identification is done by using highly discriminating probes , gene sequencing , or restriction enzyme analysis . Commercially available kit-based systems are almost exclusively restricted to the diagnosis of M. tuberculosis complex. Molecular strategies include either target amplification as done by PCR, transcription-mediated amplification , ligase chain reaction , strand displacement amplification , or signal amplification (e.g., Qbeta amplification) . Applying homemade PCR protocols for direct detection of M. tuberculosis complex, overall sensitivities between 77 and 100% and specificities between 88 and 100% were achieved. For the Amplicor PCR system (Roche, Somerville, N.J.), a sensitivity of 87.9% and a specificity of 99.6% were reported. The Amplified M. tuberculosis Direct Test (Gen-Probe, San Diego, Calif.) yielded overall sensitivities between 82 and 97% and specificities between 97 and 100%. Evaluation of other commercially available, kit-based test formats showed similar results . The LightCycler system is designed to increase the time of DNA amplification by reducing transition times between various steps in each cycle. Temperature shifts are achieved by alternating heated air and air of ambient temperature, which is significantly faster than cycling with conventional block or water bath cyclers. Several fluorescence formats are available for detection of amplified DNA. SYBR Green, a double-stranded DNA (dsDNA) binding dye, fluoresces when bound to dsDNA. To add sequence specificity, sequence-specific oligonucleotides labeled with two different fluorescence dyes may be utilized; these dyes generate a fluorescence signal by fluorescence resonance energy transfer (FRET) when the two probes bind to the target sequence. When monitoring the fluorescence while slowly increasing the temperature, the fluorescence will decrease when one of the probes melts off and the two fluorescent dyes are no longer in close contact. The melting temperature is determined not only by the length of the probe and its GC content but also by the degree of homology between the probe and the target sequence. In cases of one or a few mismatches between hybridization probe and target DNA, the probe can still hybridize but will melt off at a lower temperature. Rapid-cycle PCR amplification with an air thermocycler has decreased detection time of M. tuberculosis . Fluorimeter-based analysis has provided a rapid and sensitive method for identification of PCR products. Real-time fluorescence has been applied to diagnosis of M. tuberculosis in sputum using the TaqMan system . LightCycler technology has been used to detect M. bovis in bovine tissues as well as rifampin and isoniazid resistance-associated mutations in M. tuberculosis . In both studies, amplified fragments were typically 200 bp in size. In this study, we demonstrate that amplification of a 1,000-bp fragment of 16S rRNA from a broad spectrum of mycobacteria is achieved by using only five copies of genomic DNA as a template. Amplification of mycobacterial DNA was confirmed by using genus-specific FRET probes, thereby introducing a genus-specific region that had not been used for molecular diagnosis of mycobacteria before. Species-specific FRET probes were used to identify M. tuberculosis complex and M. avium. Finally, we constructed and tested a synthetic internal control that will permit monitoring of successful DNA amplification in future clinical studies. MATERIALS AND METHODS : Strains. | Most bacterial strains were from the American Type Culture Collection or German Collection of Microorganisms and Cell Cultures (Braunschweig, Germany), with the exception of Mycobacterium bovis bacillus Calmette-Guerin (BCG) Pasteur (Pasteur vaccine strain; Statens Serum Institute, Copenhagen, Denmark), M. bovis, Mycobacterium simae, Bacillus cereus and Rhodococcus equi (submitted to us as part of a national quality control test for diagnostic laboratories that is held twice a year in Germany), Mycobacterium paratuberculosis (gift of P. Valentin-Weigand, School of Veterinary Medicine, Hannover, Germany), Mycobacterium smegmatis mc2155 (gift of W. R. Jacobs, Albert Einstein College of Medicine, Bronx, N.Y.), M. intracellulare, Mycobacterium xenopi, M. fortuitum, Streptomyces griseus, and Corynebacterium jeikeium, which were isolated from clinical specimens in our diagnostic laboratory. All strains were colony purified before growing them in an appropriate liquid medium, such as 7H9 supplemented with 0.2% glycerol, 0.05% Tween 80, and 10% ADS (0.5% bovine albumin fraction V, 0.2% glucose, 140 mM NaCl) for all mycobacteria or tryptic soy broth for all other bacteria. DNA extraction. | Bacterial DNA was purified using the Qiamp Mini Kit (catalog no. 51306; Qiagen, Hilden, Germany). To ensure efficient bacterial cell lysis, the protocol was optimized by increasing the concentration of lysozyme stock solution from 20 to 60 mg/ml and extending incubation with the enzyme from 30 min to 2 h. Proteinase K was added to the cells, the cells were incubated at 56C for 30 min, chromosomal DNA was precipitated with 100% ethanol, and DNA was purified by using a column. Fungal DNA was kindly provided by T. Jack (Department of Medical Microbiology and Hospital Epidemiology, Medical School, Hannover, Germany). DNA was quantified using the PicoGreen system (Molecular Probes, Eugene, Oreg.) as recommended by the manufacturer. PicoGreen is a dsDNA quantitation reagent, which becomes intensely fluorescent upon binding nucleic acids . Serial dilutions of genomic DNA were mixed with PicoGreen dye, and fluorescence was analyzed photometrically. Results were compared to known lambda DNA concentrations. Various numbers of genomic copies per PCR mixture were determined by calculation of molecular weight and subsequent serial dilution. PCR primers and probes. | Primers (MWG-Biotech, Ebersberg, Germany) and probes (TIB MOLBIOL, Berlin, Germany) were designed by comparing previously published sequences of the 16S rRNA gene for various mycobacteria. All mycobacterial 16S rRNA sequences currently available at www.ncbi.nlm.nih.gov/Entrez/were included. For amplification of parts of the 16S rRNA gene, a 100-bp fragment was amplified using LC 5 (GGC GGA GCA TGT GGA TTA) (sense) and LC 4 (TGC ACA CAG GCC ACA AGG GA) (antisense), a 300-bp fragment was amplified using LC 7 (GAT AAG CCT GGG AAA CTG) (sense) and LC 8 (CTA CCG TCA ATC CGA GAG) (antisense), and a 1,000-bp fragment was amplified using LC 1 (GAG TTT GAT CCT GGC TCA GGA) (sense) and LC 4 (see 100-bp fragment). The following FRET probes were used: for detection of M. tuberculosis, LC 11 (CGC GGG CTC ATC CCA CAC CG-fluorescein) (antisense) as an anchor probe and LC 12 (LightCycler Red 640-TAA AGC GCT TTC CAC CAC AAG A) (antisense) as a sensor probe; for detection of M. avium, LC 25 (CGC GGG CCC ATC CCA CAC CG-fluorescein) (antisense) as an anchor probe and LC 26 (LightCycler Red 640-AAA AGC TTT CCA CCA GAA GAC) (antisense) as a sensor probe; and for detection of mycobacterium-specific region III, LC 39 (GCA ACG CGA AGA ACC TTA CCT GG-fluorescein) (sense) as an anchor probe and LC 40 (LightCycler Red 640-TTT GAC ATG CAC AGG ACG) (sense) as a sensor probe. All sensor probes were labeled with LightCycler Red 640 as an acceptor for FRET, and all anchor probes were labeled with fluorescein. Standard LightCycler protocol. | After optimization, the following standard LightCycler PCR protocol was applied to all specimens. A commercially available ready-to-use hot start reaction mixture (LightCycler FastStart DNA Master Hybridization Probes) (catalog no. 239272; Roche Molecular Biochemicals) containing FastStart Taq polymerase, reaction buffer, deoxynucleoside triphosphates, and 1 mM MgCl2 was supplemented with 2 mM MgCl2. After supplying primers at 18 pmol (1.1 muM final concentration) per reaction mixture and DNA probes at 2 pmol (100 nM final concentration) per reaction mixture, the mixture was applied to the top of a glass capillary reaction vessel. Following the addition of DNA template, the glass capillary was filled by a very brief centrifugation to move the liquid into the capillary. The amplification program began with a denaturation step of 10 min at 95C, followed by 50 cycles of PCR, with 1 cycle consisting of denaturation (3 s at 95C), "touchdown" annealing (2 s of a temperature ranging from 68 to 62C), and extension (40 s at 72C). For the first five cycles, annealing was performed at 68C (step delay) and then reduced to 62C with 1C per cycle (step size). The temperature transition rate for all cycling steps was 20C per s. The amplification program was followed by a melting program of 95C for 30 s (denaturation), 38C for 30 s (annealing), and then 38 to 80C at a transition rate of 0.2C/s with continuing monitoring of fluorescence. Version 3.5.3 of the LightCycler run profile software automatically adjusted the gain of the F2 channel photometric detector. In addition, all amplification products were visualized by conventional gel electrophoresis. Each LightCycler run included one capillary in which the template was replaced by water to control for cross contamination, which might have occurred at any time during preparation procedures. Construction of internal control. | The entire 16S rRNA gene was amplified by DNA amplification under standard conditions using the following pair of oligonucleotides as PCR primers: forward primer GAGTTTGATCCTGGCTCAGGA and reverse primer AAGGAGGTGATCCAGCCGCA. DNA amplification was performed in 40 cycles using 56.5C for annealing, 72C for elongation, and 95C for denaturation. Amplified DNA fragment was subcloned in pGEM-T (Promega, Madison, Wis.). To introduce one point mutation within the Mycobacterium genus-specific region III, the following mismatch oligonucleotide primers, each complementary to opposite strands of the vector, were constructed (underlining indicates essential mutations): forward primer GGCTTGACATGCACAGGACGC and reverse primer GCGTCCTGTGCATGTCAAGCC (the mismatch nucleotide is shown underlined). To introduce two point mutations within the genus-specific region III, two mismatch oligonucleotide primers (forward primer GGTTTGACATACACTGGACGC and reverse primer GCGTCCAGTGTATGTCAAACC) were constructed using Pfu Turbo Hotstart DNA polymerase (Stratagene, La Jolla, Calif.). PCR was performed in 18 cycles, with 1 cycle consisting of 30 s at 50C (annealing), 10 min at 68C (elongation), and 30 s at 95C (denaturation). PCR product was gel purified, and point mutations were confirmed by sequencing. The plasmid containing one mismatch was named pJL7, and the plasmid containing two mismatches was named pJL6. Saliva preparation. | Saliva was collected from healthy volunteers and stored at 4C. M. smegmatis, a fast-growing mycobacterial species, was cultured in 7H9 medium to an optical density at 600 nm of 1.0 and then stored at 20C, and the titer was obtained by serial dilutions on 7H10 medium supplemented with 0.2% glycerol and 10% ADS. Next, we mixed 285 mul of sputum with 15 mul of serially diluted M. smegmatis, achieving final concentrations of 20,000, 2,000, 200, 100, 50, and 20 bacteria per 300 mul of sputum. One hundred and fifty microliters of seeded sputum was subjected to DNA preparation using a PUREGENE DNA isolation kit (catalog no. 203040; Gentra Systems, Minneapolis, Minn.). This method uses salt as a substitute for toxic organic solvents in the deproteination step . Briefly, after lysis of cells, protein was precipitated with ammonium acetate and DNA was precipitated with isopropanol. Finally, DNA was resuspended in 10 mul of Tris buffer, which was subjected to LightCycler PCR and hybridization with FRET probes following the above protocol. RESULTS : Optimization of light cycler reactions. | For optimization of PCR amplification, we used exclusively M. bovis BCG genomic DNA as a template. SYBR Green I was used at first for comparison of Taq polymerase (LightCycler-DNA Master SYBR Green I) (catalog no. 2158817; Roche Molecular Biochemicals) with Hotstart Taq polymerase (LightCycler-FastStart DNA Master SYBR Green I) (catalog no. 3003230; Roche Molecular Biochemicals) and to compare one-temperature annealing with a touchdown annealing profile. Both Hotstart polymerase and stepwise reduction of annealing temperature stopped formation of primer dimers and improved the sensitivity of amplification to below 100 copies of genomes. All further parameters were evaluated using DNA probes rather than the SYBR Green I format for detection of the amplicon, aiming for detection of as little as five genome copies. In our hands, five genome copies turned out be the lower limit that could be reproducibly achieved by serial dilution. In the process of evaluating various DNA probes, we tested a range of MgCl2 concentrations for each probe (range of 2 to 5 mM tested in 0.5 mM increments) and finally decided on using 3 mM for all probes. A primer concentration of 18 pmol per reaction mixture (1.1 muM) (5 to 50 pmol tested), an elongation time of 40 s (20 to 60 s tested), an annealing time of 2 s (0 to 7 s tested), and an annealing temperature of 62C (58 to 68C tested) allowed detection of five genome copies. Partial amplification of the 16S rRNA. | Earlier studies by Kirschner et al. showed that the mycobacterial 16S rRNA gene includes two species-specific (Fig. , species A and B) and two genus-specific (Fig. , genus I and II) regions for Mycobacterium. Alignment of previously published 16S rRNA sequences revealed a third genus-specific region (Fig. , genus III) that to the best of our knowledge, has not been used for molecular diagnosis of mycobacteria before. Figure shows primers and species- and genus-specific regions. Assuming that amplification of larger fragments is less sensitive, we initially tested amplification of two fragments, a 100-bp fragment and a 300-bp fragment. Subsequently, we included a 1,000-bp fragment. The various fragments are depicted in Fig. . Surprisingly, sensitivity was equally good for amplification of all three fragments. However, only the 1,000-bp fragment contains both genus- and species-specific regions, which is why for all further experiments, the 1,000-bp fragment was amplified. Sensitivity and specificity of mycobacterium-specific detection. | FRET probes specific for genus region III reproducibly detected five copies of M. bovis BCG genome . Next we tested a broad range of mycobacteria using 2.5 ng of genomic DNA, which translates into 500,000 genomes. A melting peak of 61.5C was found for all mycobacteria using probes specific for region III, except for Mycobacterium chelonae; the melting point of the genus-specific probe decreased by 6.5C to 55C for M. chelonae . Further sensitivity testing was performed with only a selection of mycobacterial species. We used as few as five copies (range, 5,000 to 5 copies) of genomic DNA from M. tuberculosis , M. bovis, M. avium, M. intracellulare, M. paratuberculosis, M. kansasii, M. marinum, M. chelonae, and M. fortuitum as templates (data not shown). For all mycobacteria tested, five copies could be detected. To determine the specificity of genus-specific detection, we used 2.5 ng of genomic DNA from various nonmycobacterial organisms . As the reverse amplification primer targets genus specific-region I, most organisms showed no amplification product . Only bacteria of the genus Corynebacterium gave amplification of the 1,000-bp fragment of the 16S rRNA gene . FRET probe hybridization specific to genus-specific region III, however, clearly allowed separation from mycobacteria. The melting points of Corynebacterium jeikeium and Corynebacterium pseudodiphtheriticum differed from those seen with most mycobacteria by more than 15C or that seen with M. chelonae by more than 10C, and Corynebacterium xerosis showed no hybridization signal at all (Fig. and Table ). Sensitivity and specificity of M. tuberculosis complex- or M. avium-specific detection. | Using the same primers discussed above, the 1,000-bp fragment was amplified and analyzed with FRET probes specific for M. tuberculosis complex or M. avium, hybridizing against species-specific region A . As few as five genomic copies resulted in a positive signal (Fig. and ). For specificity, we tested all the mycobacterial species listed in Table and all the nonmycobacterial microorganisms listed in Table . All organisms other than M. tuberculosis complex showed melting points at least 10C lower than M. tuberculosis complex or showed no signal at all . Internal control for successful amplification and detection. | For future clinical studies, we developed an internal control that contains primer regions identical to those of the target sequence but contains an unique probe binding region that differentiates the internal control from amplified nucleic acid. The entire 16S rRNA gene from M. tuberculosis was amplified and subcloned. Using site-directed mutagenesis, one and two mismatches were introduced to genus-specific region III within the binding site of the sensor probe. Replacing T with C at the 3'-prime end of the probe reduced the melting temperature of the genus-specific probe by just 1C . Mutagenesis at the 5'-prime end, replacing G with A and, four nucleotides apart, A with T, however, reduced the melting temperature by 14.5C . Thus, two mismatches allow differentiation between internal control and target when using FRET probes targeting genus-specific region III. Next 50 copies of the plasmid carrying the two mismatches were mixed with genomic DNA from M. tuberculosis. As few as 10 genomic copies were reliably detected when 50 copies of the internal control were present The fluorescence signal for detection of five copies of the target sequence was weak yet visible. Thus, sensitivity of detection of target DNA, even at low copy numbers, remained unaffected. Next we added various concentrations of Escherichia coli genomic DNA to control for inhibition by background DNA. According to the LightCycler Operator's Manual (version 3.5), the total concentration of genomic DNA should not exceed 50 to 500 ng per capillary. In line with these observations, we found that an increasing amount of background DNA gradually reduced specific fluorescence, with 200 ng of background DNA representing the upper limit for detection of 10 copies of mycobacterial DNA and 50 copies of internal control, respectively . Finally, we seeded saliva from a healthy volunteer with various amounts of M. smegmatis and performed DNA preparation followed by analysis with the LightCycler system. As few as 10 bacteria were detected when 50 copies of the internal control were present (Fig. .). FIG. 1. | Physical locations of primers and regions used. Physical locations of primers and regions used. The mycobacterial 16S rRNA gene includes genus-specific regions I, II, and III (red) and species-specific regions A and B (yellow). Primers (LC) for amplification of a 100-, 300-, and 1,000-bp fragment were universal (blue), specific for mycobacteria (red) or specific for the M. tuberculosis complex (yellow). The 1,000-bp fragment includes all regions and was used for further experiments. FIG. 2. | Screen capture of F2 melting peak analysis, gel electrophoresis of amplicons, and sensitivity of FRET probes specific for genus region III. Screen capture of F2 melting peak analysis, gel electrophoresis of amplicons, and sensitivity of FRET probes specific for genus region III. Five copies of the M. tuberculosis genome were reproducibly detected with a melting peak of 61.5C. In this and all following figures, the melting curve analysis is displayed as the first negative derivative of the fluorescence (-dF/dT) versus temperature. F2 refers to channel 2, which is used by the LightCycler's optical unit to measure signals from LightCycler Red 640 at 640 nm. FIG. 3. | Screen capture of F2 melting peak analysis, gel electrophoresis of amplicons, and specificity of FRET probe specific for genus region III. Screen capture of F2 melting peak analysis, gel electrophoresis of amplicons, and specificity of FRET probe specific for genus region III. Bacteria other than mycobacteria were tested. Only corynebacteria were amplified, but they could be clearly distinguished from M. tuberculosis by a lower melting point or no hybridization signal at all . FIG. 4. | Screen capture of F2 melting peak analysis, gel electrophoresis of amplicons, and sensitivity of the FRET probes specific for M. Screen capture of F2 melting peak analysis, gel electrophoresis of amplicons, and sensitivity of the FRET probes specific for M. tuberculosis complex. Five copies of M. tuberculosis genome were reproducibly detected with probes with a melting peak of 64C. FIG. 5. | Screen capture of F2 melting peak analysis, gel electrophoresis of amplicons, and sensitivity of the FRET probes specific for M. Screen capture of F2 melting peak analysis, gel electrophoresis of amplicons, and sensitivity of the FRET probes specific for M. avium. Five copies of M. avium genome were reproducibly detected with probes with a melting peak of 61C. FIG. 6. | Melting analysis of subcloned 16S rRNA gene using genus II-specific probe showed, as expected, 61.5C for the wild type . Melting analysis of subcloned 16S rRNA gene using genus II-specific probe showed, as expected, 61.5C for the wild type . Two mismatches changed the melting point to 47.5C . One mismatch reduced melting peak by just 1C. FIG. 7. | Screen capture of F2 melting peak analysis of serially diluted genomes of M. Screen capture of F2 melting peak analysis of serially diluted genomes of M. tuberculosis mixed with 50 copies of internal control. As few as 10 genomes were detected when 50 copies of internal control (I.C.) were present . FIG. 8. | Screen capture of F2 melting peak analysis of 10 genomes of M. Screen capture of F2 melting peak analysis of 10 genomes of M. tuberculosis combined with 50 copies of internal control and various amounts of background DNA. Two hundred nanograms of background DNA is the upper limit for detection of 10 copies of mycobacterial DNA and 50 copies of internal control (I.C.), respectively. FIG. 9. | Screen capture of F2 melting peak analysis of saliva with various numbers of M. Screen capture of F2 melting peak analysis of saliva with various numbers of M. smegmatis and 50 copies of internal control. As few as 10 bacteria were detected when 50 copies of internal control (I.C.) were present . TABLE 1 | Profile of 16S rRNA gene amplification and probe hybridization DISCUSSION : The intriguing feature of the 16S rRNA molecule is the presence of conserved and variable regions, allowing the amplification of nucleic acids on the genus level followed by confirmation and species differentiation using highly discriminating probes . The LightCycler System achieves high transition times by alternating heated air and air of ambient temperature. Even though rapid equilibration between the air and the reaction components is ensured by a high surface-to-volume ratio of the capillaries, we expected suboptimal amplification, especially of large DNA fragments of mycobacteria, due the organism's high GC content. Therefore, our initial concept was to amplify a small 100-bp fragment in the hope that the hybridization of this fragment with a genus-specific FRET probe could unambiguously establish a diagnosis of mycobacteria. In a second PCR, a 300-bp fragment was to be amplified and hybridized with a M. tuberculosis complex or M. avium FRET probe, thereby separating these species from other mycobacteria. Surprisingly, we found that amplification of a 1,000-bp fragment of the 16S rRNA gene using the LightCycler PCR system was as efficient as amplification of a 100- and 300-bp fragment. The 1,000-bp fragment allows identification of mycobacteria and differentiation of mycobacterial species after a single PCR , so we used the 1,000-bp fragment for all further experiments. Genus-specific regions I and II have been used for molecular diagnosis of mycobacteria by Kirschner et al. before, targeting region I with a genus-specific primer for selective amplification of part of the mycobacterial 16S rRNA gene and confirming diagnosis of mycobacteria by hybridizing a DNA probe homologous to genus-specific region II. By aligning published sequences, we found a third genus-specific region, which we used as a genus-specific probe for mycobacteria in this study. A melting temperature of 61.5C for this genus-specific probe indicates mycobacteria, whereas melting at 55C indicates Mycobacterium chelonae. In a recent study reporting detection of M. bovis with LightCycler technology, only a single oligonucleotide FRET probe was used, because the target, IS6110, was short and GC-rich. Resonance energy was provided from SYBR Green I intercalated between the oligonucleotide and the PCR product . In this study, two fluorescence-labeled oligonucleotide FRET probes, an anchor and a sensor probe, were constructed hybridizing to adjacent regions of target DNA. Using three different anchor probes and three different sensor probes, amplified target DNA from all mycobacteria, from M. tuberculosis complex, and from M. avium was identified. All three sensor probes could be designed to have a GC content of below 50%, which was quite unexpected, given that the average GC content of mycobacterial DNA is between 65 and 70%. Typically, anchor probes had specific melting points 8 to 10C higher than those of sensor probes, ensuring that the latter dissociated first from the target sequence, causing rapid decrease of specific fluorescence. This format optimized detection of mismatches between the sensor probes and target sequences, guaranteeing a high specificity of melting point analysis. Thus, although amplification was nonspecific for C. xerosis, C. pseudodiphtheriticum, or C. jeikeium, all three corynebacteria were easily separated from mycobacteria due to a significantly reduced melting point of the genus-specific probe. Likewise, hybridization of M. tuberculosis complex-specific probes and M. avium-specific probes, to their specific target DNA showed melting points of 64 and 61C, respectively, whereas the closest melting points of any other mycobacterial species were 54 and 53C. In this study, evaluation of rapid-cycle PCR and fluorimetry using LightCycler technology was confined to cultural isolates. At present, the use of 16S rRNA sequence determination for routine identification of mycobacteria from cultural isolates is superior to all other techniques, as it covers a wide range of mycobacterial species . Although LightCycler technology will not replace direct sequencing, it may provide rapid identification and differentiation of mycobacterial species to those who have no access to sequencing facilities. In theory, a panel of species-specific probes could be generated to meet the needs of an individual laboratory. Those dealing with dermatology clinics might include FRET probes specific for M. marinum, while those dealing with cystic fibrosis clinics might include FRET probes specific for M. abscessus. Another application might be early detection of mycobacteria, M. tuberculosis complex, or M. avium in broth culture following a short incubation period of clinical specimens including tissue samples. The most important application, however, will be direct detection of mycobacteria in clinical specimens. Since DNA extracted from clinical specimens contains impurities that inhibit enzyme-based nucleic acid amplification, negative amplification test results do not necessarily indicate the absence of mycobacteria. Therefore, for future studies, we developed a synthetic internal control as a proxy for the primary target; the internal control contained a mutagenized target sequence for the genus-specific FRET probe. We used the internal control at a low concentration of 50 copies per test sample to monitor amplification at the limit of test sensitivity, which in accord with a previous study recommended as few as 20 copies of the internal control to each reaction mixture. A high load of internal control could fail to indicate inhibition as well as compete with target DNA for amplification . The internal control that was tested in this study allowed unambiguous detection of amplified target nucleic acid at low target loads, even if unspecific background DNA was added to the amplification reaction mixture. We also showed that with the internal control present, we could detect as few as 10 mycobacteria mixed with saliva. However, the performance of this test system with actual sputum samples remains to be determined. We demonstrated that LightCycler technology allows diagnosis of the genus Mycobacterium and identification of M. tuberculosis complex and M. avium within 45 min. Thus, identification of mycobacteria and differentiation between M. tuberculosis complex and nontuberculous mycobacteria can be obtained faster than with any other nucleic acid amplification-based technique that is available at present. The LightCycler's optical unit is capable of measuring fluorescence from FRET probes in two separate channels simultaneously. Channel 2 (F2; 640 nm) is used to measure signals from LightCycler Red 640. Channel 3 (F3; 705 nm) is designed for use with LightCycler Red 705. By this means, a genus-specific FRET probe labeled with LightCycler Red 640 and a M. tuberculosis complex-specific FRET probe labeled with LightCycler Red 705 could separate the former from the latter in a single PCR. Future studies will show whether the system introduced in this study can be applied directly to clinical specimens. Backmatter: PMID- 12202551 TI - Molecular Characterization of Multiresistant d-Tartrate-Positive Salmonella enterica Serovar Paratyphi B Isolates AB - Since 1996, the National Salmonella Reference Laboratory of Germany has received an increasing number of Salmonella enterica subsp. enterica serovar Paratyphi B isolates. Nearly all of these belonged to the dextrorotatory tartrate-positive variant (S. enterica subsp. enterica serovar Paratyphi B dT+), formerly called S. enterica subsp. enterica serovar Java. A total of 55 selected contemporary and older S. enterica subsp. enterica serovar Paratyphi B dT+ isolates were analyzed by plasmid profiling, antimicrobial resistance testing, pulsed-field gel electrophoresis, IS200 profiling, and PCR-based detection of integrons. The results showed a high genetic heterogeneity among 10 old strains obtained from 1960 to 1993. In the following years, however, new distinct multiresistant S. enterica subsp. enterica serovar Paratyphi B dT+ clones emerged, and one clonal lineage successfully displaced the older ones. Since 1994, 88% of the isolates investigated were multiple drug resistant. Today, a particular clone predominates in some German poultry production lines, poultry products, and various other sources. It was also detected in contemporary isolates from two neighboring countries as well. Keywords: Introduction : Zoonotic Salmonella enterica serovars are among the most important agents of food-borne infections throughout the world. Poultry, pigs, and cattle rank as the major sources of Salmonella-contaminated food products that cause human salmonellosis. Only a few Salmonella serovars predominate in an animal population or a country at a certain time. Currently there are global pandemics of Salmonella enterica subsp. enterica serovars Enteritidis and Typhimurium DT104 . However, from time to time less common serovars emerge and can cause outbreaks in humans or animals (, , ; World Health Organization Global Salmonella Survey list server message 2000-18, 10 May 2000). One of the rare serovars is S. enterica subsp. enterica serovar Paratyphi B, which can be differentiated by the use of dextrorotatory tartrate. The classical S. enterica subsp. enterica serovar Paratyphi B is d-tartrate negative (dT-) and a virulent human pathogen. The d-tartrate-positive (dT+) variant was formerly called S. enterica subsp. enterica serovar Java . It is generally considered less virulent for humans, but some authors claim that it should not be underestimated as a possible cause of outbreaks of salmonellosis (; World Health Organization Global Salmonella Survey list server [www.who.ch/salmsurv] message 2000-18, 10 May 2000). In recent years, the National Salmonella Reference Laboratory of Germany (NRL) has received an increasing number of S. enterica subsp. enterica serovar Paratyphi B dT+ isolates originating from poultry and poultry products . The aim of this study was to obtain information about the emergence and spread of dT+ variants of S. enterica subsp. enterica serovar Paratyphi B isolates in Germany by determining their phenotypic and genotypic properties. A collection of contemporary and older S. enterica subsp. enterica serovar Paratyphi B dT+ isolates of different origins and sources were characterized by several classical and DNA-based typing methods. The results of this study demonstrate the emergence of distinct S. enterica subsp. enterica serovar Paratyphi B dT+ clones during the 1990s. One of them displaced the older clones successfully and is frequently encountered in German poultry, poultry products, and various other sources. MATERIALS AND METHODS : Bacterial strains and selection of isolates. | The NRL receives about 5,000 Salmonella isolates annually from all Bundeslaender (federal states) of Germany for serotyping. They originate from routine surveys of different investigation centers involved in public health. In general the isolates are identified only as Salmonella and sent to the NRL for confirmation and detailed serotyping, as well as phage, resistance, and molecular typing. The NRL is quality controlled by the European Community Reference Laboratory and the World Health Organization Global Salmonella Survey program. Detailed sources and properties of the d-tartrate-positive strains of S. enterica subsp. enterica serovar Paratyphi B as well as the control strains used in this study are listed in Table . All isolates from the years 1960 to 1995 and the comparative strains were selected on the basis of their availability. Old strains originated from stab cultures of the NRL strain collection. Contemporary isolates were selected from routine submissions in order to represent all different resistance phenotypes, different geographic locations, and possible sources. Therefore, not only poultry but also, if available, bovine, environmental, and other sources were included. In order to avoid the characterization of duplicates, strains from the same geographic location which were isolated at the same time by the same laboratory and which exhibited the same serovar and resistance profiles were excluded from the study. Serotyping was performed according to the Kauffmann-White scheme, and strains were named serovar Paratyphi B when they exhibited the antigen formula 1,4,[5],12:b:1,2 . Identification of the strains as d-tartrate positive or negative was done as described by Dorn et al. . Antimicrobial susceptibility test. | All of the serovar Paratyphi B strains were tested for their susceptibility to 16 antimicrobial agents by agar diffusion test in accordance with the guidelines of the German Institute for Standards with antibiotic disks (Oxoid Ltd., London, England). The antimicrobial agents used were amikacin (30 mug), ampicillin (10 mug), chloramphenicol (30 mug), cefuroxime (30 mug), colistin sulfate (10 mug), enrofloxacin (5 mug), gentamicin (10 mug), kanamycin (30 mug), nalidixic acid (30 mug), neomycin (10 mug), polymyxin B (300 IU), streptomycin (25 mug), trimethoprim-sulfamethoxazole (25 mug), sulfonamides (300 mug), tetracycline (30 mug), and trimethoprim (2.5 mug). Conjugative transfer of resistance markers. | Two S. enterica subsp. enterica serovar Paratyphi B dT+ strains each from group 2 (strains 2504 and 2496) and group 3 (1086 and 2933), harboring plasmids and revealing multiple antimicrobial resistance determinants, were selected for mating experiments. Escherichia coli strain J53, resistant to rifampin, nalidixic acid, and streptomycin, was used as the recipient strain. Liquid and solid mating experiments for suitable resistance markers (ampicillin and/or chloramphenicol, kanamycin, tetracycline, or trimethoprim) were performed at 37 and 22C as described previously . Subsequently, the plasmid profiles and antibiotic resistance patterns of donor and transconjugant strains were compared. Nucleic acid techniques. (i) Plasmid profile typing. | Plasmid DNA was extracted by the alkaline denaturation method of Kado and Liu with minor modifications. Plasmids were electrophoretically separated in 0.7% horizontal agarose gels at 100 V for 3.5 h in Tris-borate-EDTA buffer. The plasmids were stained with an aqueous solution of ethidium bromide (10 mug/ml; Sigma, Deisenhofen, Germany) and photographed under UV illumination. The E. coli reference plasmids R27 (112 MDa), R1 (62 MDa), RP4 (36 MDa), and ColE1 (4.2 MDa) and the supercoiled DNA ladder (Invitrogen Life Technologies, Karlsruhe, Germany) served as size standards for the determination of plasmid sizes. (ii) PCR amplification and amplicon purification. | Salmonella plasmid virulence region (spvC) tests for all serovar Paratyphi B dT+ strains were performed by PCR with the primers and cycling conditions described previously . The assays were performed in 50-mul reaction volumes containing an amplification buffer (20 mM Tris-HCl [pH 8.4], 50 mM KCl, 1.5 mM MgCl2), 200 muM each of the four deoxynucleoside triphosphates (Roche Diagnostics, Mannheim, Germany), 1 muM each primer (TIB MOLBIOL, Berlin, Germany), 1 U of Platinum Taq DNA polymerase (Invitrogen), and 100 ng of template DNA, in a Perkin-Elmer Gene Amp system (model 2400). Amplicons were analyzed by electrophoresis on horizontal 1.5% agarose gels. A 100-bp ladder and the DNA molecular weight marker X (both from Roche Diagnostics) were used as molecular size markers. Control strains were the spvC-positive S. enterica serovars Typhimurium strain 184 and Choleraesuis strain 807 from the NRL strain collection. Class 1 integron tests of strains and transconjugants were performed by PCR with the primers 5'CS and 3'CS . PCR amplification was carried out as described above with the cycling conditions cited previously . For IS200 profiling, a 557-bp internal DNA probe was generated by PCR amplification as described previously with the primers IS200-L2 and -R2 . The amplification reactions were modified and contained 0.1 muM each primer (TIB MOLBIOL), 1 mg of bovine serum albumin per ml, approximately 50 ng of genomic DNA, and 2.5 U of Platinum Taq DNA polymerase (Invitrogen). The PCR product was purified with the QIAquick PCR purification kit (Qiagen, Hilden, Germany). (iii) Macrorestriction analysis. | Genomic DNA for pulsed-field gel electrophoresis (PFGE) experiments was prepared in low-melting-point agarose gel plugs as previously described . Slices of the DNA-containing agarose plugs were incubated in a final volume of 100 mul overnight at 37C with 10 U of XbaI, SpeI, or BlnI (all from Roche Diagnostics). The plugs were melted at 65C, and DNA fragments were subjected to PFGE in 1% (wt/vol) agarose (Seakem GTG agarose; FMC BioProducts) in 0.5x Tris-borate-EDTA buffer on a CHEF-DR III system (Bio-Rad, Munich, Germany) at 10C. Pulse times were ramped from 5 to 50 s during a 24-h run (XbaI), 4 to 40 s during a 24-h run (SpeI), or 20 to 80 s during a 25-h run (BlnI) at 6 V/cm. A lambda ladder (New England Biolabs, Frankfurt, Germany) was used as molecular size markers. DNA fragment patterns were assessed visually, and PFGE profiles were assigned according to published guidelines . A minimum of three band differences in the electrophoretic pattern defined a distinct PFGE profile. Only fragments larger than 40 kb were evaluated. (iv) IS200 profiling. | Genomic DNA was isolated from 1.5 ml of an overnight broth culture either by the method of Wilson or with the Qiagen blood and cell culture DNA mini kit (Qiagen). DNA was digested at 37C for 4 h with the restriction enzyme PstI (Roche Diagnostics), which lacks restriction sites within the IS200 sequence . The resulting DNA fragments were subsequently separated by electrophoresis through 0.7% agarose gels with Tris-borate-EDTA buffer as the running buffer at 45 V for 18 h. A 1-kb ladder (Invitrogen) and the digoxigenin-labeled DNA molecular weight marker II (Roche Diagnostics) served as molecular mass markers. The gels were stained with ethidium bromide and photographed under UV illumination. The DNA fragments were transferred from agarose gels to positively charged nylon membranes (Roche Diagnostics) in 10x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate) with a vacuum blotter (model 785; Bio-Rad) and fixed to the membrane by cross-linking. Labeling of the purified 557-bp PCR-product used as the probe (described above), prehybridization, hybridization under stringent conditions, and colorimetric signal detection were carried out with the digoxigenin High Prime labeling and detection starter kit I (Roche Diagnostics) according to the instructions of the manufacturer. TABLE 1 | Sources and properties of S. enterica serovar Paratyphi B dT+ and comparison strains RESULTS : Incidence of S. enterica subsp. enterica serovar Paratyphi B dT+. | Figure shows the number of S. enterica subsp. enterica serovar Paratyphi B dT+ isolates received at the NRL each year from 1991 to 2000. In the years before 1996, this serovar was only sporadically encountered (less than 10 isolates per year). Since then, a steady increase has been observed, encompassing 213 (5.4%) of the 3,915 isolates received in 2000. Ninety-three percent of the S. enterica subsp. enterica serovar Paratyphi B dT+ isolates originated from poultry and poultry products, and 99% were multiresistant. In order to investigate the clonal descent of these isolates, 55 selected German S. enterica subsp. enterica serovar Paratyphi B dT+ strains and 8 epidemiologically unrelated comparison strains were included in this study. The contemporary strains represented all different resistance phenotypes, different geographic locations, and possible sources of the isolates received. The phenotypic properties and molecular typing results are summarized in Table . Representative results of the molecular typing studies are shown in Fig. and . Altogether, these data show that the strains could be divided into three groups differing in major phenotypic and genotypic properties. Properties of old S. enterica subsp. enterica serovar Paratyphi B dT+ strains (group 1). | Group 1 consisted of 10 old strains which originated mainly from sources other than poultry. All strains were sensitive to the antimicrobial agents tested. Only one strain harbored two small plasmids (3.6 and 1.3 MDa), one strain carried a 3.9-MDa plasmid, and the other strains possessed no plasmid at all. Macrorestriction analysis performed with the three restriction endonucleases XbaI, SpeI, and BlnI revealed heterogeneity, with six distinct PFGE profiles (X1 to X6, S1 to S6, and B1 to B6). XbaI patterns X1 to X6 were characterized by 16 to 20 fragments ranging in size from 20 to 700 kb, SpeI patterns S1 to S6 were characterized by 16 to 21 fragments ranging in size from 20 to 485 kb, and BlnI patterns B1 to B6 were characterized by 7 to 10 fragments ranging in size from 20 to 750 kb. Figure (lanes 1 to 7) shows the PFGE profiles of seven representative old S. enterica subsp. enterica serovar Paratyphi B dT+ strains after digestion with XbaI as an example (BlnI and SpeI data not shown). IS200 (JSP) profiles revealed 1 to 14 IS200 copies per strain, with corresponding fragment sizes of 1.9 to >21.3 kb. Eight distinct IS200 profiles, ISP1 to ISP8 , were observed, confirming the above-mentioned diversity among these older strains. Figure (lanes 1 to 7) shows the IS200 profiles of the same seven representative old S. enterica subsp. enterica serovar Paratyphi B dT+ strains as an example. Properties of mid-1990s S. enterica subsp. enterica serovar Paratyphi B dT+ strains (group 2). | Group 2 was represented by four strains isolated from chicken meat in different German regions in the mid-1990s and two strains from 1999. These strains differ from the first group and represent a new clone. All these strains except 2696 exhibited a core spectrum of antibiotic resistance determinants for trimethoprim, streptomycin, sulfonamides, and tetracycline. Additional resistances to kanamycin/neomycin and nalidixic acid were found in some isolates. Plasmid profiling revealed that all of the strains except 2696 were characterized by a plasmid of 128 MDa and two plasmids of 3.9 and 2.4 MDa. Conjugation experiments showed that the resistance to chloramphenicol, tetracycline, and trimethoprim (strain 2504) and chloramphenicol, kanamycin, tetracycline, and trimethoprim (strain 2496) was transferable at 22C in one linkage group on the 128-MDa plasmid (data not shown). Strain 2696 carried only an 88-Mda plasmid and had lost trimethoprim, streptomycin, sulfonamide, and tetracycline resistance. The macrorestriction analysis revealed PFGE patterns differing by only one to two fragments in the high-molecular-weight range. All these strains were assigned to PFGE profile X7/S7/B7 . Figure illustrates the PFGE pattern of a representative group 2 Paratyphi B dT+ strain in lane 8. The IS200 profiles (ISP8) of these isolates were very homogeneous as well , showing only one IS200 band at 5.4 kb (Fig. , lane 8). This profile has been detected in sensitive strain 290 of group 1 from 1993 as well. Properties of contemporary German S. enterica subsp. enterica serovar Paratyphi B dT+ strains (group 3). | Group 3 included 39 isolates from the second half of the 1990s and from 2000. Thirty-three strains originated from poultry and poultry products . The remaining six isolates were isolated from cattle, fish, sewage, or unknown sources. All 39 isolates were resistant to trimethoprim. Nearly 50% of the strains were resistant to four or more antimicrobial substances, with predominant resistance to sulfonamides (59%), nalidixic acid (54%), ampicillin (49%), and streptomycin (46%) in addition to trimethoprim. Ciprofloxacin resistance (MIC, 0.25 to 2 mug/ml) could be detected in 40% of the year 2000 isolates (B. Malorny, unpublished data). Plasmid analysis revealed the presence of 16 different plasmid profiles, consisting of one to four plasmids of between 1.4 and 112 MDa, with the most common plasmids being 62, 4.0, and 2.4 MDa in size. Five isolates were plasmid-free . None of the isolates in this study carried spvC-related sequences, as determined by PCR. Strain 1086 and strain 2933 were selected as representatives of group 3 for conjugation experiments. Both strains transferred the ampicillin resistance on the 62-MDa plasmid at 37C in liquid matings. Strain 1086 cotransferred sulfonamide resistance on the 62-MDa plasmid as well. Strain 2933 transferred resistance to sulfonamides, tetracycline, and trimethoprim in one linkage group on the 112-MDa plasmid. In PFGE analysis, each of the restriction enzymes yielded in all cases a typical, uniform PFGE profile (X8, S8, B8) . As an example, Fig. (lanes 9 to 15) shows the macrorestriction profile X8 of seven representative contemporary S. enterica subsp. enterica serovar Paratyphi B dT+ strains from group 3. The clonal similarity of the group 3 strains was confirmed by the fact that all strains showed an identical IS200 profile (ISP9), with only one IS200 band of 5.8 kb (Fig. , lanes 9 to 14). Properties of comparison strains. | As shown in Table , contemporary isolates from Belgium (Bru2-1175) and The Netherlands (Bi52 and Bi3742) showed properties similar to those of the contemporary German isolates. With the PFGE profile X8, S8, B8 and IS200 profile ISP9, these isolates could also be assigned to the third group of German strains. Two contemporary isolates from England, W-S6497 and W-S42000, in contrast, were revealed in the macrorestriction analysis and in IS200 profiling patterns (X9, S9, B9; ISP10) to be highly different from the German ones. In addition, both isolates were sensitive to all of the antimicrobial agents tested and did not harbor any plasmids . Strain 2650, a strain originating from a snake, was sensitive to all of the antimicrobial agents tested but possessed four plasmids of 112, 92, 7, and 5 MDa. The patterns (X10, S10, B10; ISP11) obtained by both molecular typing methods were unique and not comparable to any of the other patterns . NCTC 5706, a 1937 monophasic reference strain from the National Collection of Type Cultures in London, England, was sensitive, possessed no plasmid, and showed a unique pattern (X11, S11, B11; ISP12) by PFGE and IS200 profiling as well . Strain 1328, a contemporary serovar Typhimurium strain of the epidemiologically important phage type DT104, was characterized by pentaresistance (ampicillin, chloramphenicol, streptomycin, sulfonamides, and tetracycline), a plasmid of 60 MDa, PFGE profile X12, and IS200 profile ISP13, both typical of serovar Typhimurium DT104 strains. Consequently, the profile of this strain differed from the profiles of the S. enterica subsp. enterica serovar Paratyphi B dT+ strains . Detection of class 1 integrons. | All strains included in this study and all transconjugants were screened by PCR for the presence of class 1 integrons with specific primers targeting the conserved 5' and 3' segments of the integron. No integron was detectable in the old, antibiotic-sensitive strains of group 1. In contrast, in five of the six resistant strains of group 2, a class 1 integron generating a PCR product of about 1,800 bp was found. This integron resided on the 128-MDa conjugative plasmid, because all transconjugants gave rise to the same PCR amplicon as the donor. In addition, strain 2696, which lacked trimethoprim, streptomycin, sulfonamide, and tetracycline resistance, also did not carry the integron. The majority (38 of 39) of the group 3 strains did not carry any class 1 integron. Only strain 2933 carried an integron with an amplicon of about 1,600 bp. It was located on the 112-MDa plasmid, as shown by analysis of the transconjugants. The comparison strain Bi52 from The Netherlands also harbored an integron with an amplicon of about 1,600 bp. Its properties resembled those of the group 3 strains. FIG. 1. | Incidence of S. Incidence of S. enterica subsp. enterica serovar Paratyphi B dT+ strains received at the NRL from 1991 to 2000. FIG. 2. | PFGE profiles of representative S. PFGE profiles of representative S. enterica subsp. enterica serovar Paratyphi B dT+ strains after digestion with XbaI. Lanes M contain molecular size markers (lambda ladder PFG marker; New England Biolabs). Lanes 1 to 15 contain DNA samples from the following strains and groups: group 1: 1, 1285; 2, 83; 3, 163; 4, 1198; 5, 5707; 6, 521; and 7, 290; group 2: 8, 2504; group 3: 9, 232; 10, 1086; 11, 2399; 12, 189; 13, 135; 14, 2675; and 15, 2933. FIG. 3. | IS200 profiles of representative S. IS200 profiles of representative S. enterica subsp. enterica serovar Paratyphi B dT+ strains. Lanes M contain molecular size markers (digoxigenin-labeled DNA molecular weight marker II; Roche). Lanes 1 to 15 contain DNA samples from the following strains and groups: group 1: 1, 1285; 2, 83; 3, 163; 4, 1198; 5, 5707; 6, 521; and 7, 290; group 2: 8, 2504; and 15, 2696; group 3: 9, 232; 10, 1086; 11, 2399; 12, 189; 13, 135; and 14, 2675. DISCUSSION : Only a few studies deal with typing of strains of serovar Paratyphi B because so far this serovar has not played a significant epidemiological role. In 1988, Barker et al. studied variation in a worldwide collection of 338 isolates of S. enterica subsp. enterica serovar Paratyphi B. They defined 13 biotypes, phage types, and ribotypes. On the basis of these properties, strains were assigned to three groups. Selander et al. applied multilocus enzyme electrophoresis to distinguish 14 electrophoretic types (ETs) within S. enterica subsp. enterica serovar Paratyphi B. The authors observed considerable genotypic diversity among ETs and concluded that most dT- strains comprised a globally distributed clone, Pb1, with highly polymorphic phenotypes, whereas dT+ strains represented seven clonal lineages. Ezquerra et al. in their study found 13 unique IS200 profiles among representative S. enterica subsp. enterica serovar Paratyphi B strains and were able to distinguish distinct genotypes of S. enterica subsp. enterica serovar Paratyphi B dT- and dT+. One of the S. enterica subsp. enterica serovar Paratyphi B dT- profiles, Spj-IP1.0, represented a globally distributed clone. Greater diversity was detected within the IS200 profiles of S. enterica subsp. enterica serovar Paratyphi B dT+ isolates. The authors hypothesized that since S. enterica subsp. enterica serovar Paratyphi B dT+ has many animal and human hosts, a variety of selection pressures exist for the evolution of genetic diversity. The increasing incidence of S. enterica subsp. enterica serovar Paratyphi B dT+ strains in Germany and neighboring countries, especially in poultry and poultry products, indicates the growing importance of this pathogen. The study presented was initiated by the need to obtain more information about the phenotypic and molecular properties of this Salmonella serovar and the emergence and spread of clonal lines in Germany. In addition to classical typing methods, molecular typing methods such as plasmid profiling, macrorestriction analysis, IS200 profiling, and PCR detection of integrons were used. These techniques have proven to be useful procedures in subtyping other Salmonella serovars (, , , -, , , ). The data presented here confirmed the genetic diversity of the S. enterica subsp. enterica serovar Paratyphi B dT+ strains described by Selander et al. in 1990 and by Ezquerra et al. in 1993. However, genetic diversity was detected only in a group of older German S. enterica subsp. enterica serovar Paratyphi B dT+ strains originating from the beginning of the 1960s to the beginning of the 1990s and in foreign comparison strains. These drug-sensitive and generally plasmid-free strains showed several diverse patterns by PFGE and IS200 profiling, suggesting the existence of coexisting clones. In subsequent years, the situation changed. The data presented show that in the middle of the 1990s, new clones of S. enterica subsp. enterica serovar Paratyphi B dT+ originating from poultry and poultry products emerged in Germany and spread. They were multiresistant and could have been selected by the increasing use of antimicrobial agents in poultry production due to attempts to manage the serovar Enteritidis crisis. One clone was resistant to chloramphenicol, sulfonamide, tetracycline, and trimethoprim, and some isolates were resistant to kanamycin, neomycin, and/or nalidixic acid as well. The multiresistance was found to be located on a transferable 128-MDa plasmid and to be associated with a class 1 integron. Strains of this clone exhibited a unique PFGE profile (X7) and a unique IS200 profile (ISP8). This clone did not disappear completely, because two isolates could be detected at the end of the 1990s as well. In 1995, a new clonal line emerged and spread successfully. Today it represents the majority of contemporary strains and has replaced almost all other clones. This clonal line is found predominantly in Germany in poultry and poultry products and has been isolated from other sources such as cattle, fish, and sewage as well. Strains of this clonal line are characterized by a variety of plasmid profiles as well as resistance patterns composed of resistance to trimethoprim alone or, in 87% of the cases, in combination with resistance to streptomycin, sulfonamides, nalidixic acid, and/or ampicillin. On the chromosomal level, strains belonging to the dominant clonal line are highly uniform with respect to molecular typing by macrorestriction analysis (PFGE profile X8) and IS200 profiling (ISP9). The data from the study presented demonstrate the emergence of multidrug resistance among S. enterica subsp. enterica serovar Paratyphi B dT+ strains in Germany during the last decade. Several mechanisms involving mobile genetic elements, such as plasmids and transposons, have been shown to contribute to the spread of resistance . Integrons, able to incorporate antibiotic resistance gene cassettes by site-specific recombination, have been identified on these mobile elements . Class 1 integrons, the most prevalent integron class in gram-negative bacteria, are also widespread among many Salmonella serovars . Therefore, it was not unexpected that class 1 integrons could also be found in multiresistant S. enterica subsp. enterica serovar Paratyphi B dT+ strains. Although the contemporary S. enterica subsp. enterica serovar Paratyphi B dT+ clone is multiresistant, integrons are not important for this phenotype. Characterization of the major genes conferring antimicrobial resistance in all S. enterica subsp. enterica serovar Paratyphi B dT+ strains tested is presently being carried out. Backmatter: PMID- 12202614 TI - First Report on Schizophyllum commune from a Dog AB - This report describes the first isolation of Schizophyllum commune from a granulomatous lesion on the neck of a dog. The biopsy specimen from the lesion disclosed granulomatous inflammation with branching fungal hyphae without clamp connections. The clinical isolate was identified as S. commune by mycological examination and analysis of ribosomal DNA sequences. Keywords: CASE REPORT : A 3-year-old male mongrel dog weighing 10 kg was referred to the Minami Animal Hospital in Mie, Japan, with the chief complaint of a subcutaneous nodule on the ventral surface of the neck. This subcutaneous nodule with no fistula and discharge was round, elastic, unfixed to the subcutis, and 6 by 11 cm in size. The dog was depressed and showed dyspnea. Furthermore, chest radiographs revealed a nodule (5 by 5 cm in size) at the anterior mediastinum and a diffuse interstitial pattern in the cranial lobes . Microscopic examination of a biopsy specimen from the subcutaneous nodule disclosed (i) granulomatous inflammation with branching fungal hyphae without clamp connections and (ii) many granules . Because the owner refused surgery or other invasive procedures, the dog was treated with ketoconazole at 10 mg/kg administered orally once a day for 3 months. This regimen was chosen because ketoconazole is an antifungal drug with few adverse effects and is reasonably priced. However, nodules did not diminish in size. Three months later, the dog died. No autopsy was performed. The basidiomycetous fungus Schizophyllum commune is emerging as one of the important agents of sinusitis and pneumonia in humans ; however, reports of its infection of animals have been lacking. The diagnosis of this infection is sometimes difficult, and the identification of the isolate is uncertain. Clamp connections on hyphae in tissues are often missing, and fruiting bodies are usually not formed in vitro. The fungus is identified only when it is a dikaryon capable of basidiocarp production . In recent years, molecular techniques have greatly improved the identification of S. commune. Analysis of the internal transcribed spacer region of the rRNA genes has been utilized for identification of the species . The present report describes the first isolation of S. commune from a dog. The clinical isolate grew on Sabouraud's glucose agar at 24C for 1 week, and its colony was white and cottony . Microscopic examination of the isolate revealed hyaline hyphae 2.5 to 3.0 mum in diameter and confirmed the presence of clamp connections and tubercles , indicating that the isolate was a basidiomycete. To identify the species of the clinical isolate, molecular analysis of its 25S large-subunit ribosomal DNA was carried out. A mycelial sample (about 10 mg) of the isolate was lysed in lysis buffer consisting of 0.1% Zymolyase 100T (Takara, Kyoto, Japan), 0.1 mM EDTA, 1% sodium dodecyl sulfate, 10 mM Tris hydrochloride (pH 8.0) and 0.3% 2-mercaptoethanol at 37C for 16 h. High-molecular-weight DNA was obtained from this mycelial sample by phenol and chloroform extraction. A 100-ng DNA sample dissolved in TE buffer (10 mM Tris-hydrochloride [pH 8.0], 1 mM EDTA) was used for PCR amplification. The genomic DNA sample was amplified by PCR in a reaction mixture (30 mul) containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl, 0.001% gelatin, 200 mM each deoxynucleoside triphosphate, 1.0 U of Taq polymerase (Takara), and 0.5 mug of a pair of primers. We used the 25S large-subunit ribosomal DNA-specific primer for typing . The sequences of the 25S large-subunit ribosomal DNAs of many species of basidiomycetes have been deposited in the DNA data bank of Japan (DDBJ), including those of the ribosomal DNA internal transcribed spacer region. The sequence of forward primer S-1 was 5'-GCA TAT CAA TAA GCG GAG GAA AAG-3', and that of reverse primer S-2 was 5'-GGT CCG TGT TTC AAG ACG-3'. With these primers, a 600-bp fragment containing the coding sequence of the 25S large-subunit ribosomal DNA was expected to be amplified. PCR amplification was carried out for 35 cycles consisting of template denaturation (1 min at 94C), primer annealing (2 min at 55C), and polymerization (3 min at 72C). The PCR product was electrophoresed through a 2% agarose gel and then stained with ethidium bromide. The PCR product from the sample was sequenced by the dideoxy-chain termination method with an ABI PRISM 310 Genetic Analyzer (Perkin-Elmer Corp., Norwalk, Conn.). These procedures were carried out three times. Amplification of the sample DNA with 25S large-subunit ribosomal DNA primers yielded a fragment of about 600 bp, consistent with the size of the 25S ribosomal DNA of fungal species reported previously . To examine homology relationships between the 25S large-subunit ribosomal DNA of a reference strain of S. commune and that of the clinical isolate, we used FASTA database analysis in DDBJ. The nucleotide sequences of the 25S large-subunit ribosomal DNAs of the clinical isolate and a reference strain of S. commune (DDBJ accession no. ) showed 99% similarity . The results were obtained three times with this molecular system, and the isolate was identified as S. commune. As stated previously, the lack of clamp connections in tissues is not an unusual finding and the case reported herein is no exception. Moreover, morphological identification is made even more difficult when strains exhibit an inability to mate. For these reasons, we attempted to identify our isolate by comparing its 25S large-subunit rRNA gene sequence with that of a known strain of S. commune. Our results help to confirm observations by Sigler et al. and Buzina et al. indicating that this method has potential for the rapid identification of clinical isolates of S. commune. The number of cases of S. commune infection is increasing in both immunocompetent and immunocompromised humans . In this case, no immunosuppressive condition or treatment was recognized in the canine patient. However, we feel that steroid treatment, chemotherapy, and immunosuppressive underlying diseases may facilitate subsequent infection with S. commune in dogs. Animal infection with S. commune has not been reported as far as we know, but with the aid of molecular analysis, additional animal cases can be found hereafter. Nucleotide sequence accession number. | The partial sequence of the S. commune clinical isolate 25S large-subunit rRNA gene reported in this paper has been deposited in the GenBank database and assigned accession no. . FIG. 1. | Chest radiograph revealing the nodule at the anterior mediastinum and a diffuse interstitial pattern in the cranial lobes. Chest radiograph revealing the nodule at the anterior mediastinum and a diffuse interstitial pattern in the cranial lobes. FIG. 2. | Microscopic view of a biopsy specimen from a nodule on the neck disclosing granulomatous inflammation with fungal branching hyphae. Microscopic view of a biopsy specimen from a nodule on the neck disclosing granulomatous inflammation with fungal branching hyphae. FIG. 3. | White, cottony colony of the clinical isolate. White, cottony colony of the clinical isolate. FIG. 4. | Microscopic view of the clinical isolate revealing clamp connections and tubercles. Microscopic view of the clinical isolate revealing clamp connections and tubercles. Backmatter: PMID- 12202567 TI - Simultaneous Detection of Anaplasma marginale and a New Ehrlichia Species Closely Related to Ehrlichia chaffeensis by Sequence Analyses of 16S Ribosomal DNA in Boophilus microplus Ticks from Tibet AB - To identify ehrlichial agents in Boophilus microplus ticks, DNA samples of B. microplus collected from the Tibet Autonomous Region and Sichuan Province of China were screened by a nested PCR. Sixteen of 43 (37%) DNA samples of B. microplus from Tibet were positive in nested PCR analysis. All 27 samples from Sichuan were negative. The screen identified two ehrlichial agents based on different 16S rRNA genes that were found after amplifying and sequencing the 5'-end fragments of the 16S rRNA genes. One sequence was identical to that of the gene of Anaplasma marginale, an etiological agent of animal anaplasmosis. The other sequence was most similar to that of the gene of Ehrlichia chaffeensis, an etiological agent of human monocytic ehrlichiosis. The sequence of 1,501 bases from the novel ehrlichial agent was obtained and showed the greatest levels of sequence similarity (97 to 98%) to 16S rRNA gene sequences of the members of the E. canis group of the genus Ehrlichia. Sequence comparison of the 16S rRNA gene with the members of the genus Ehrlichia reveals that the novel ehrlichial agent detected in B. microplus ticks is a new species of the genus Ehrlichia and is most closely related to E. chaffeensis. Keywords: Introduction : Ehrlichia species are known as important pathogens of medical, as well as veterinary, importance. They are intracellular microorganisms residing within the cytoplasmic vacuoles of monocytes, granulocytes, or platelets of humans and animals. Ehrlichia species elicit illnesses with fever, headache, leukopenia, and thrombocytopenia . The genus Ehrlichia is divided into three distinct genogroups based on the similarity of the nucleotide sequence of the 16S rRNA gene: the Ehrlichia canis group (including E. canis, E. chaffeensis, E. muris, and E. ewingii), the E. phagocytophila group (E. phagocytophila, E. equi, human granulocytic ehrlichiosis [HGE] agent, and E. platys), and the E. sennetsu group (E. sennetsu, E. risticii, and the Stellantchasmus falcatus [SF] agent) . The members of the E. sennetsu group are transmitted by aquatic vectors . In contrast, the members of the E. canis group and the E. phagocytophila group are both tick-borne organisms . E. chaffeensis and the HGE agent are tick-borne ehrlichial organisms discovered in the United States. E. chaffeensis, an etiological agent of human monocytic ehrlichiosis, transmitted by the tick Amblyomma americanum, was discovered in 1986, while the HGE agent transmitted by Ixodes scapularis was first reported in 1994 (, -). Recently, serological and PCR-based studies suggested that human monocytic ehrlichiosis and HGE also exist outside the United States, particularly in some European and Asian countries (-, , ). E. muris, originally isolated from a wild mouse and later isolated from Haemaphysalis flava ticks in Japan, was identified as an agent closely related to E. chaffeensis in serological and genetic analyses . Another new Ehrlichia species closely related to E. chaffeensis was isolated from Ixodes ovatus ticks in Japan . PCR-based assays and DNA sequencing detected E. chaffeensis, E. canis, and E. platys in various ticks from southern China, while the HGE agent was identified in Ixodes persulcatus ticks from northeastern China . Boophilus microplus ticks are widely distributed over the world and are recognized as the vectors of Anaplasma marginale, an etiological agent of animal anaplasmosis . A. marginale is considered to be a member of the tribe Ehrlichieae; 16S rRNA gene analysis shows that it is closely related to the members of the E. phagocytophila group . B. microplus ticks are widely distributed in China, where Coxiella burnetii and Borrelia burgdorferi were isolated from the tick species. In this study we used PCR-based assays and DNA sequencing analysis to assess the occurrence of infection by Ehrlichia in B. microplus ticks collected from Sichuan and Tibet, in order to collect data that can be used to provide future estimates of the likelihood of ehrlichiosis in southwestern China. MATERIALS AND METHODS : Ticks. | B. microplus ticks were collected from infested cattle in Sichuan Province and the Tibet Autonomous Region of China. Immediately after collection, the ticks were immersed in 70% ethanol and stored. Preparation of DNA extracts from ticks. | Two ticks were placed in a 1.5-ml Eppendorf tube, and then the tube was baked at 100C for 10 min in an oven. Immediately after baking, the ticks were pulverized in the tube using a glass bar. Following this, 300 mul of Tris-EDTA buffer (10 mM Tris, 1 mM EDTA, pH 8.0), 40 mul of 10% sodium dodecyl sulfate, and 5 mul of proteinase K (20 mg/ml) were added to the tube. The mixture was incubated at 55C for 3 h and was then incubated in boiling water for 10 min. After a brief centrifugation, the supernatant was transferred to a fresh 1.5-ml tube. The supernatant was extracted twice, first with an equal volume of phenol-chloroform-isoamyl alcohol (25:24:1) and then with an equal volume of chloroform. The aqueous phase (top layer) was transferred to a fresh 1.5-ml tube, the DNA in the aqueous phase was precipitated with ethanol according to the standard method, and the DNA pellets were dissolved in 50 mul of Tris-EDTA buffer and stored at -20C. Nested PCR amplification of ehrlichial 16S rRNA genes. | A pair of universal primers, Eh-out1 (5'-TTGAGAGTTTGATCCTGGCTCAGAACG-3', located at positions 1 to 27 in a 16S rRNA sequence obtained in this study [GenBank accession number ]), and Eh-out2 (5'-CACCTCTACACTAGGAATTCCGCTATC-3', at positions 653 to 627), were used for the primary amplification. Primers Eh-gs1 (5'-GTAATAACTGTATAATCCCTG-3', at positions 167 to 187), and Eh-gs2 (5'GTACCGTCATTATCTTCCCTA-3, at positions 448 to 428), which were designed based on the conserved positions in the sequences of 16S rRNA genes of several tick-borne species of Ehrlichia, were used in the nested amplification. PCR amplifications were performed in 25-mul reaction volumes in a thermal cycler. For the primary amplification, a reaction mixture contained 1.5 pmol each of primers Eh-out1 and Eh-out2, 200 mumol of each deoxynucleoside triphosphate, 2.5 mM MgCl2, 2.5 mug of bovine serum albumin (20 mug/ml), 2.5 mul of 1x PCR buffer (50 mM KCl, 10 mM Tris-HCl [pH 8.3]), 1 U of Taq DNA polymerase, and 2 mul of tick DNA. The reaction mixture was overlaid with mineral oil and incubated for 5 min at 95C, followed by 40 cycles of 95C for 45 s, 55C for 50 s, and 72C for 1 min and then a final incubation at 72C for 5 min to allow complete strand extension. For the nested PCR, the components and conditions were similar to those used in the primary amplification, except that 0.6 mumol of Eh-gs1 and 0.6 mumol of Eh-gs2 were used as primers and that 2 mul of the primary PCR product was used as templates . A seminested PCR assay was used to amplify the 5'-end fragments of ehrlichial 16S rRNA genes from the positive specimens of the nested PCR assay. The components and conditions in this assay were similar to that of the nested PCR assay, except that Eh-out1 and Eh-gs2 were used as a pair of seminested primers . The 3'-end fragments of ehrlichial 16S rRNA genes were amplified from the positive DNA specimens by using universal primers Eh-out1 and 3-17 (5'-TAAGGTGGTAATCCAGC-3', at positions 1501 to 1485) in the primary amplification and using a species-specific primer and the 3-17 universal primer in seminested PCR . The PCR products were electrophoresed on a 1.5% agarose gel, stained with ethidium bromide, and visualized under UV light. Quality control included both positive and negative controls that were PCR amplified in parallel with all specimens. Cloning and sequencing of PCR products. | After electrophoresis, the positive DNA fragments were recovered from the agarose gel with a DNA gel extraction kit (NSBC-Sangon, Shanghai, China) according to the instructions of the manufacturer. The fragments were cloned by using the PUCm-T vector (NSBC-Sangon) and competent cells (E. coli DH5alpha) following the manufacturer's protocol. The recombinant plasmids were extracted from the transformed E. coli cells with a plasmid DNA preparation kit (NSBC-Sangon). Sequencing was carried out by fluorescence-labeled dideoxynucleotide technology with an automated DNA sequencer (SM377; Perkin-Elmer, Norwalk, Conn.). The primers used in the above PCR assays, M13 (5'-CCCAGTCACGACGTTGTAAAACG-3'), located in the PUCm-T vector), and two additional primers (5'-TAGTCCACGCTGTAAACG-3', at positions 761 to 778; and 5'-CCCGTCAATTCCTTTGAG-3', at positions 884 to 866), were chosen as sequencing primers. Computer analyses of DNA sequences. | The analyses of sequence data and sequence alignment were performed on a computer using DNAsis software (Hitachi Software, San Bruno, Calif.). The evolutionary distance calculations and phylogenetic tree inference were performed with the software package CLUSTALX. Nucleotide sequence accession number. | The GenBank accession numbers for the 16S rRNA genes used for comparison in this study were E. chaffeensis, M73222; E. canis, M73221; E. muris, U15527; E. ewingii, U96436; E. bovis, U03775; E. phagocytophila, M73220; E. platys, AF286699; E. sennetsu, M73225; E. risticii, M21290; and Ehrlichia sp. strain SF, 34280. The nucleotide sequence of the 16S rRNA gene of Ehrlichia sp. strain Tibet was deposited in GenBank under accession number . FIG. 1. | The different fragments of 16S rRNA genes amplified with different pairs of primers from the tick DNA specimens. The different fragments of 16S rRNA genes amplified with different pairs of primers from the tick DNA specimens. The italicized words represent the names of primers, and the bold lines represent the sizes of the PCR products. RESULTS : Nested PCR screening of the tick DNA specimens. | A DNA fragment of approximately 280 bp was amplified from 16 of 43 (37%) specimens of B. microplus from Tibet. DNA samples from 27 specimens of B. microplus from Sichuan were all negative. The 280-bp fragments from five of the positive specimens were directly sequenced on the automated DNA sequencer by using Eh-gs1 and Eh-gs2 as sequencing primers. All of the sequences are recognizable as ehrlichial 16S rRNA genes, but the fragment analyzed did not permit them to be distinguished at the species level. Analyses of the 16S rRNA gene sequences. | In order to identify ehrlichial agents at the species level, the 5'-end fragments (similar450 bp) of 16S ribosomal DNA containing ehrlichial species-specific sequences were amplified from 10 of the nested-PCR-positive specimens by a seminested PCR. The 5'-end fragments were cloned, and the fragments inserted in the recombinant plasmids were sequenced. Four of 10 sequences were identical to a 16S rRNA gene sequence of A. marginale (GenBank accession number ). The remaining 6 of 10 sequences were most similar to a 16S rRNA gene sequence of E. chaffeensis (Arkansas strain; GenBank accession number ). A fragment of approximately 1,450 bp was amplified by a seminested PCR from the samples that were shown to contain the novel ehrlichial agent as demonstrated by sequences from the 5'-end fragments. This seminested PCR used primer pair Eh-out1 and 3-17 together with another pair of primers (universal primer 3-17 and a species-specific primer [5'-CGAACGGACAATTGTTTATATC-3'] designed based on the highly variable region of the 16S rRNA gene of the novel agent ). A sequence of 1,501 bases was obtained by linking the 5'-end and 3'-end sequences of the 16S rRNA gene of the novel agent, based on overlapping regions of sequence. When the entire sequence of the 16S rRNA gene of the novel agent is compared to those of other species of Ehrlichia, the greatest levels of nucleotide sequence similarity (97 to 98%) were found with the members of the E. canis group of the genus Ehrlichia. The entire sequence was most similar to the 16S rRNA gene sequences of E. chaffeensis but differed by 19 nucleotides (similar1.4%). The entire sequence of the novel agent was aligned with the 16S rRNA gene sequences of the known species of the genus Ehrlichia by using a multiple sequence alignment program of DNAsis, and a 1,403-base portion of the 16S rRNA gene sequence that could be unambiguously aligned for all species of the genus Ehrlichia was used for phylogenetic analysis. The levels of sequence divergences and similarities between the novel agent and the organisms of the genus Ehrlichia are shown in Table . Consistent with the results above, the greatest levels of similarity (98 to 99%) for the 1,403-bp region were found between the novel agent and the sequences of the E. canis group; the highest level of similarity (99%) was found between it and the sequence of E. chaffeensis . The phylogenetic tree obtained from the data is shown in Fig. . It is clear that the novel ehrlichial agent in the ticks from Tibet belongs to the E. canis group within the genus Ehrlichia and was not closely related to the E. phagocytophila or E. sennetsu genogroup (Table and Fig. ). The analyses of the levels of sequence divergence show that divergence between the novel agent and its closest neighbor, E. chaffeensis, is sometimes greater than the levels of sequence divergence obtained in pairwise comparisons of several other Ehrlichia species. FIG. 2. | A highly variable region of sequence located at the 5' end of the 16S rRNA gene revealed by multiple alignment of 16S rRNA gene sequences of Ehrlichia spp. A highly variable region of sequence located at the 5' end of the 16S rRNA gene revealed by multiple alignment of 16S rRNA gene sequences of Ehrlichia spp. in the E. canis genogroup. C. ruminantium, Cowdria ruminantium. FIG. 3. | Phylogenetic relationships between Ehrlichia sp. Phylogenetic relationships between Ehrlichia sp. strain Tibet and other members of the genus Ehrlichia. The tree was constructed using the neighbor-joining program in the software package CLUSTALX. TABLE 1 | Levels of genetic similarity and differences between novel agent (Tibet strain) and ehrlichial species in 16S rRNA sequences DISCUSSION : Ehrlichia spp. are obligatory intracellular pathogens. Isolation of Ehrlichia from cell culture is a classical method for determining their presence in blood and tissue specimens of human or animals. PCR is a sensitive and simple method used to directly detect infectious agents in various specimens, and nested PCR is more sensitive than standard PCR procedures . A nested PCR assay specific for amplification of the 16S rRNA genes of tick-borne Ehrlichia spp. was established using nested 16S rRNA primers specific only for the tick-borne Ehrlichia species. Using this nested PCR analysis, 37% of DNA samples of ticks from Tibet were positive, while the DNA samples of ticks from Sichuan were all negative. Further, sequence analysis of the amplimers showed that they were recognizable as fragments of 16S rRNA genes of tick-borne Ehrlichia. These results suggest that the nested PCR is both highly sensitive and specific for detection of the tick-borne Ehrlichia spp., and it is particularly useful for screening the specimens containing the new species of the genus Ehrlichia. Since the 280-bp amplimers produced by nested PCR did not contain the species-specific signature sequences of Ehrlichia spp., the ehrlichial agents in the positive specimens could not be identified at the species level by sequencing these amplimers. Therefore, the 5'-end fragments of 16S rRNA genes of the ehrlichial agents in the positive specimens were amplified and sequenced. The analyses of 5'-end fragments revealed that the positive specimens contained two agents: A. marginale, an etiological agent of animal anaplasmosis, and a novel ehrlichial agent that was most closely related to E. chaffeensis, an etiological agent of human monocytic ehrlichiosis. However, the question of whether the two agents coexist in B. microplus remains to be studied, because each DNA sample in this study was prepared from two ticks. Sequence comparison of the 16S rRNA gene is recognized as one of the most powerful and precise methods for determining the phylogenetic relationships of bacteria . In this study, our analysis of 16S rRNA sequences revealed that the novel agent found in the Tibet ticks was a member of the E. canis group of the genus Ehrlichia and was most closely related to E. chaffeensis but was also closely related to E. canis, E. ewingii, and E. muris. Since the 16S rRNA gene is known to exhibit a high level of structural conservation and a low evolutionary rate , levels of sequence divergence greater than 0.5% in comparisons between nearly complete 16S rRNA gene sequences of members of the genus Ehrlichia have been considered sufficient to classify organisms as different species . The levels of divergence of the 16S rRNA sequence between the novel Tibetan ehrlichial agent and the members of the E. canis group were approximately 1 or 2% in pairwise comparisons of 1,403-base sequences, and this level of difference should be sufficient to classify the novel ehrlichial agent as a new species of the genus Ehrlichia. Since the novel agent was first detected in the ticks from Tibet, it is temporarily named Ehrlichia sp. strain Tibet. Although it is well known that A. marginale is an etiological agent of bovine anaplasmosis and is transmitted by B. microplus ticks , the detection of any species of Ehrlichia in B. microplus ticks has not been reported before. The identification of A. marginale and Ehrlichia sp. strain Tibet in B. microplus ticks suggests the possibility of coinfection and cotransmission of the two agents in the area where the ticks were collected. Therefore, it is also possible that, in this area, B. microplus ticks and the cattle infested by these ticks are coinfected with the two agents. However, these probabilities remain to be demonstrated by further studies. Backmatter: PMID- 12202609 TI - Molecular Characterization Confirms the Presence of a Divergent Strain of Canine Coronavirus (UWSMN-1) in Australia AB - Canine coronavirus (CCV) UWSMN-1 was originally identified from an outbreak of fatal gastroenteritis in breeding colonies. In this report, we examined whether UWSMN-1 represents a novel divergent strain or is the result of recombination events between canine and feline coronavirus strains. Sequencing of various regions of the spike and polymerase genes confirms that UWSMN-1 is widely divergent from other CCV and feline coronavirus strains. These data raise the possibility that this strain is the first member of a novel third subtype of CCV. Keywords: Introduction : Canine coronavirus (CCV) is a common pathogen of dogs that generally produces symptoms of mild to moderate gastroenteritis (-, ). In combination with other pathogens such as canine parvovirus, or in young or stressed animals, CCV infection causes more severe symptoms or can be fatal . It has recently been demonstrated that CCV is prevalent in dog populations throughout Australia, and persistent reinfection is common in breeding colonies . The spike (S) gene of the coronavirus family is reported to be important in virus-host attachment and is involved in antigenic differences between strains (for a review, see reference ). Coronaviruses are known to undergo frequent recombination events in vitro , and an increasing body of evidence suggests that recombination of coronaviruses occurs in the field . Several canine and feline coronaviruses (FCoVs) have been identified that have an S gene originally derived from other coronavirus species , and CCVs can be separated into two subtypes (feline- or canine-like) on the basis of S gene origin . The frequent recombination events of coronaviruses have been suggested to be an important means of avoiding host immunity and may occur due to the close proximity of domesticated cats and dogs . We have previously reported the identification of a novel Australian strain of CCV (UWSMN-1) from a fatal case of gastroenteritis in pups from breeding colonies by using an S gene nested PCR assay . Initial sequencing of the variable 5' region of the S gene of UWSMN-1 showed that UWSMN-1 formed an intermediate group between FCoVs and CCVs . These data raised two possibilities: (i) UWSMN-1 is a divergent coronavirus strain, or (ii) the differences observed in the 5' region of the S gene are not reflective of divergence across the whole UWSMN-1 genome but may reflect recombination events between FCoVs and CCVs in the S gene only. To address these competing scenarios, we sequenced the conserved 3' region of the S gene as well as part of the polymerase (Pol) gene from UWSMN-1 and compared them to various coronavirus strains on the basis that these usually conserved regions would give good discrimination of the phylogenetic relationships between strains. A 751-bp region from the 3' end of the S gene and a 409-bp conserved region of the Pol gene were amplified by using the primers CCVSF1 (forward primer, 5'-AGCACTTTTCCTATTGATTG-3') and CCVSR1 (reverse primer, 5'-GTTAGTTTGTCTAATAATACCAACACC-3') for the S gene and CCVPF1 (forward primer, 5'-ATGGGATGGGACTATCCTAAGTGTGA-3') and CCVPR1 (reverse primer, 5'-ATCTTTGTTGTAGCACACAACTCCATCATC-3') for the Pol gene, with PCR conditions as previously described . The PCR products were purified by using a Bresa-Clean DNA purification kit (Geneworks, Adelaide, Australia), and DNA was cloned by using pGEM-T Easy Vector system II (Promega, Madison, Wis.). Nucleotide sequencing was performed in both orientations by automated sequencing at the Australian Genome Research Facility (University of Queensland, St. Lucia, Australia ). Alignments of the DNA sequences from different coronaviruses and CCV strains were made by using the PILEUP program of the Genetics Computer Group package through the Australian National Genomic Information Service (Sydney University, Sydney, Australia ). The calculation of percentages of identity, construction of DNA parsimony phylogenetic trees, and bootstrap analysis wereperformed as previously described by using the HOMOLOGIES (Genetics Computer Group), ESEQBOOT, EDNAPARS, and ECONSENSE programs of the PHYLIP package through the website of the Australian National Genomic Information Service. A BLAST search and subsequent DNA parsimony analysis of the sequenced region of the Pol gene from UWSMN-1 showed that this strain is most closely related to group I coronaviruses (CCV, porcine transmissible gastroenteritis virus, and FoCV) . Calculation of the percentages of identity between coronavirus strains demonstrated that the Pol gene of USWMN-1 shares the highest homology (96.1%) with Pol genes of other CCVs, followed closely by 95.4% identity with the Pol genes of both FoCV and porcine transmissible gastroenteritis virus . As previously reported, these group I coronaviruses share relatively poor homology with coronaviruses of other species . Phylogenetic analysis confirmed this relationship, with an unrooted parsimony tree based on the Pol sequences forming two main groups corresponding to the group I (including UWSMN-1) and group II coronavirus serogroups . These data indicate that, within the Pol gene, UWSMN-1 is most closely related to the group I coronaviruses, particularly CCV. FIG. 1. | (A) DNA alignment of Pol genes of UWSMN-1, feline infectious peritonitis virus , and CCV 1-71 . (A) DNA alignment of Pol genes of UWSMN-1, feline infectious peritonitis virus , and CCV 1-71 . (B) DNA alignment of S genes of UWSMN-1, CCV TN449 , feline enteric coronavirus (FECV) , CCV NSVL , and CCV K378 . GenBank accession numbers are noted in parentheses. Shaded regions indicate conserved amino acid residues among the different CCV and FoCV strains. Variable regions of the consensus sequence are indicated by white boxing. FIG. 2. | Percentages of nucleotide identity between the Pol (A) and S (B) gene sequences of CCV strain UWSMN-1 and other coronaviruses, including the CCV and FoCV strains denoted in Fig. Percentages of nucleotide identity between the Pol (A) and S (B) gene sequences of CCV strain UWSMN-1 and other coronaviruses, including the CCV and FoCV strains denoted in Fig. . TGEV, porcine transmissible gastroenteritis virus. FIG. 3. | Phylogenetic trees of coronaviruses, CCV, and FoCV. Phylogenetic trees of coronaviruses, CCV, and FoCV. (A) Pol gene; (B) 3' region of S gene; (C) 5' region of S gene (reproduced from reference 10 with permission). The sequences are based on DNA parsimony analysis using the PHYLIP package as described in Materials and Methods. Bootstrap values indicate the number of times out of 1,000 iterations that a branch point was identified. The sequence of the 3' end of the S gene of UWSMN-1 also showed only a small apparent bias for other CCVs, with relatively similar average identities of 90.3% to comparable sequences of feline-related coronaviruses and of 93.0% to those of other CCVs . In comparison to FoCVs and other CCVs, the Australian isolate (UWSMN-1) had the least homology with the other strains for the 3' region of the S gene . Phylogenetic analysis confirmed the data on percentages of identity, with UWSMN-1 forming a distinct branch on the 3' S gene nucleotide parsimony tree between different CCV and FoCV strains . The finding that the conserved 3' end of UWSMN-1 forms an independent subgroup when compared to FoCVs and other CCVs mirrors our previous findings with the highly variable 5' end of the S gene . The distinctness of UWSMN-1 among CCVs is therefore confirmed here to be a feature that is not confined only to the variable 5' region of the S gene but is likely to be a general feature of this strain. Sequencing of the various regions of the coronavirus genome has been previously used to investigate the phylogenetic relationship of various coronavirus strains . In these studies, recombination events between different CCV, FoCV, and porcine coronavirus strains were identified . However, in the present study, we demonstrate that UWSMN-1 forms an independent group between CCV and FoCV strains even with regard to the conserved regions of the S and Pol genes. Moreover, in comparing the 751 nucleotides in the 3' region of the S gene, it was found that UWSMN-1 had 21 unique sites and that there were 112 sites where UWSMN-1 was different from at least one of the other strains analyzed . There was no obvious pattern to the locations of these differences with respect to the other CCVs or FoCVs, as they appeared to be randomly interspersed . These data demonstrate that our original finding of a divergent 5' region of the S gene in UWSMN-1 is probably not the result of recombination events between FoCVs and CCVs, as would be indicated if the S gene shared blocks of homology with either FoCV or CCV S genes. Rather, UWSMN-1 appears to be generally divergent, indicating an earlier break from a common ancestor and the gradual accumulation of mutations throughout its genome, which may be reflective of its isolated evolution in Australia. To date, no "typical CCVs" have been identified in Australia by using electron microscopy and reverse transcription-PCR as detection methods . These data indicate that CCV UWSMN-1 forms a novel third CCV subgroup that is distinct from the two feline- and canine-like subtypes. Nucleotide sequence accession numbers. : The sequences of the S and Pol genes for CCV UWSMN-1 were deposited in GenBank and assigned accession no. and , respectively. Backmatter: PMID- 12202585 TI - Fluorescent Amplified Fragment Length Polymorphism Analysis of Salmonella enterica Serovar Typhimurium Reveals Phage-Type- Specific Markers and Potential for Microarray Typing AB - Fluorescent amplified fragment length polymorphism (AFLP) was applied to 46 Salmonella enterica serovar Typhimurium isolates of Australian origin comprising nine phage types, by using the restriction enzymes MseI and EcoRI and all 16 possible MseI +1-EcoRI +1 primer pair combinations. AFLP in the present study showed a very good discrimination power with a Simpson index of diversity of 0.98, and 35 different AFLP patterns were observed in the 46 isolates. AFLP grouped most serovar Typhimurium isolates by phage type and enabled differentiation of phage types. Furthermore, 84 phage-type-specific polymorphic AFLP fragments, for which presence or absence correlated with phage type (including 25 with one exception to phage type specificity) were observed in the 46 strains studied. Eighteen phage-type-specific AFLP fragments were cloned and sequenced. Fifteen are of known genes or have a homologue in the databases. Three sequences are plasmid related, eight are phage related, and four relate to chromosomal genes. Twelve of the 18 fragments are polymorphic because the DNA is present or absent as indicated by Southern hybridization, and we see good potential to use sequences of these fragments as the basis for multiplex PCR and development of a microarray-based molecular phage-typing method for serovar Typhimurium. Keywords: Introduction : Salmonella enterica has been divided into over 2,000 serovars based on the combination of antigenic properties of flagellar H1 and H2 antigens and the polysaccharide O antigen . S. enterica serovar Typhimurium is a very diverse serovar, which includes by definition all 1,4,[5],12:i:1,2 isolates. Serovar Typhimurium causes salmonellosis (Salmonella gastroenteritis) among humans and domestic animals worldwide. It was for many years the serovar commonly isolated, but the other major serovar, Enteritidis, has outnumbered serovar Typhimurium in many areas in recent years . Effective epidemiological surveillance and control of serovar Typhimurium requires the accurate subtyping of strains to determine potential sources of infection. A number of different phenotypic and genotypic methods have been used for this purpose, including phage typing, biotyping, plasmid profile typing, and plasmid fingerprinting. Phage typing is a commonly used method that has proved to be very useful in epidemiological surveillance of serovar Typhimurium infections. The phage-typing scheme is based on combinations of resistance or degree of sensitivity of serovar Typhimurium isolates to a series of specific bacteriophages (phages). The Anderson phage-typing scheme being used today distinguishes 207 definitive phage types (DTs) with 34 phages . The use of phage typing has enabled the rise and fall of different forms and geographical distribution to be monitored, with DT104, for example, rising in recent years to dominance in serovar Typhimurium in much of Europe but remaining rare in Australia. However, knowledge of relationships between phage types is very limited, and the genetic basis of phage type variation remains largely unknown. In some cases, one serovar Typhimurium phage type may be converted to a different type via plasmid, transposon, or temperate phage acquisition . Phage type conversion in S. enterica serovar Enteritidis caused by the introduction of a resistance plasmid has also been reported elsewhere . Although phage typing plays an important role in epidemiology, it has some drawbacks. First, a proportion of serovar Typhimurium strains cannot be classified by the present phage-typing scheme. Second, the recording of results is, to a significant extent, subjective, so that misclassification can occur . Molecular typing methods, such as IS200 typing, ribotyping, and pulsed-field gel electrophoresis (PFGE), have also been used to type serovar Typhimurium and/or to define the relationships between and within phage types of serovar Typhimurium . However, IS200 typing and ribotyping detect only variations in part of the genome. PFGE detects variations in the whole genome without requiring knowledge of sequence and has been quite widely used, with reports of its use for both primary discrimination of S. enterica isolates and subdivision of phage types. In recent years, a novel DNA fingerprinting method, amplified fragment length polymorphism (AFLP), has proved itself as a high-resolution genotyping method and a useful tool in taxonomy and epidemiological studies of microorganisms . AFLP is based on the selective amplification of restriction fragments by PCR from digested genomic DNA, with restriction site-adapter-specific primers under stringent conditions . The restriction fragments analyzed are small, and even mutation of 1 bp can be detected. The use of different sets of restriction enzymes or different primer pair combinations can generate large numbers of different AFLP fingerprints without prior knowledge of sequence. Fluorescent AFLP, which uses fluorescent dye-labeled primer, has proved to be reproducible and capable of standardization . It was shown elsewhere to have a discriminatory power equal to that of PFGE in genotyping S. enterica serovars and higher than that of PFGE in genotyping S. enterica serovar Enteritidis phage type 4 . In the present study, fluorescent AFLP was applied to nine phage types of serovar Typhimurium to explore their genetic relationships; to study the genetic variations detected by AFLP, and to assess the potential of using markers identified by AFLP for multiplex PCR or microarray technology as a successor to phage typing. MATERIALS AND METHODS : Bacterial isolates and DNA preparation. | Forty-six serovar Typhimurium isolates from nine phage types were used. The selected phage types, DT9, DT135, DT64, DT44, DT126, DT12a, DT1, DT141, and DT108, have been dominant or frequent phage types in animal and human infections in Australia in recent years . All isolates used in this study were kindly provided by the Institute of Medical and Veterinary Science (IMVS) in Adelaide, Australia. The isolates were from different sources, including human, animal, and environmental sources in different regions of Australia . Genomic DNA preparation was as described previously by Bastin et al. . Plasmid DNA was extracted by the alkaline method of Kado and Liu . AFLP reaction. | The AFLP technique was performed as described previously by Lan and Reeves for radioactive AFLP, and for fluorescent AFLP the primer was labeled with fluorescent dye. Genomic DNA was digested by EcoRI and MseI and ligated to EcoRI and MseI adapters simultaneously (note that ligation to adapters does not generate a full restriction site). Fluorescent AFLP was done with 6-carboxyfluorescein (FAM; blue) fluorescent dye-labeled MseI primer (ABI Perkin-Elmer) with one base selection (MseI +1) and unlabeled EcoRI primer with one base selection (EcoRI +1). Analysis of AFLP fingerprinting patterns. | One microliter of each fluorescent AFLP product was electrophoresed on an ABI 373 DNA sequencer equipped with the ABI PRISM GeneScan 2.0 software at the Sydney University Prince Alfred Macromolecular Analysis Centre. A GeneScan TAMRA-500 internal size standard (Perkin-Elmer) was also loaded with each AFLP sample to enable precise size determination of amplified fragments. The data for each lane were saved as an individual GeneScan file and displayed as an electropherogram: peaks representing AFLP fragments from 40 to 600 bp were visually inspected, and presence was scored as 1 and absence was scored as 0. Most peaks were easily scored for presence or absence, but some low-intensity peaks were judged by comparison with the same peak of other strains. Generally peaks with intensity less than 1/10 of that of the peak with the highest intensity were not scored in this study. Those peaks that were included could be scored as present or absent without ambiguity. The accuracy of scoring was confirmed when the 18 fragments cloned gave the expected pattern when used as a probe for blotting (see Results and Discussion). The Dice coefficient (SD) was used for phylogenetic tree construction with the unweighted pair group method with arithmetic mean in PHYLIP . AFLP was done for all 16 possible MseI +1-EcoRI +1 primer pair combinations. Primer pair MseI + C-EcoRI + A is represented as C/A, etc. The fragments further analyzed were given a number following the primer pair combination. C/T-1 and C/T-2 are polymorphic fragments generated with primer pair MseI + C and EcoRI + T, etc. These numbers do not indicate relative mobility. Cloning and sequencing of polymorphic AFLP fragments. | Polymorphic AFLP fragments were excised from a 6% polyacrylamide gel after radioactive AFLP, with AFLP patterns visualized by autoradiography for 48 h with Kodak MR BioMax film. The excised fragments were cloned into the pGEM-T Easy plasmid cloning vector (Promega) and transformed into Escherichia coli K-12 strain JM109 by electroporation with a Bio-Rad gene pulser. The recombinant plasmid was extracted from the clone with the Wizard 373 DNA purification system (Promega). For each excised fragment, up to 10 recombinant plasmids were screened for the correct polymorphic AFLP fragment by probing unlabeled AFLP polyacrylamide gel blots with DIG Easy Hyb and the DIG luminescent detection kit for nucleic acids (Boehringer Mannheim) with PCR-amplified insert DNA labeled by the DIG DNA labeling kit (Boehringer Mannheim). In each case one or more plasmids were found with an insert that when used as a probe showed the same polymorphism as observed in fluorescent AFLP and radioactive AFLP. One such insert was sequenced for each of the 18 fragments. Plasmid sequencing was carried out by the Sydney University Prince Alfred Macromolecular Analysis Centre using an ABI 377A automated DNA sequencing system and the ABI dye terminator cycle sequencing kit (Perkin-Elmer). Analysis of AFLP fragment sequences. | BlastN and BlastX searches were carried out with ANGIS (Australian National Genomic Information Service), which incorporates several sets of programs . The searches were against the nonredundant nucleic and the nonredundant protein databases compiled by ANGIS. BlastN searches were also carried out against the S. enterica serovar Typhimurium, serovar Paratyphi, and serovar Typhi genome sequences at and . We used the Enterix site, , for sequences that matched genes in the above genomes. The alignment of the genomes for the above three and also the serovar Dublin and serovar Enteritidis (incomplete) genomes allowed us to determine for each gene if it was present or absent in each of the five genomes. Southern hybridization. | Plasmid DNA or EcoRI-digested genomic DNA was electrophoresed on an 0.7% agarose gel and transferred to a Hybond-N+ nylon hybridization membrane with a VacuGene XL vacuum blotting system (Amersham Pharmacia Biotech). Cloned AFLP fragments were amplified by PCR, labeled with the DIG DNA labeling kit (Boehringer Mannheim), and hybridized to the blots at high stringency with DIG Easy Hyb and the DIG luminescent detection kit for nucleic acids (Boehringer Mannheim). Nucleotide sequence accession numbers. | The nucleotide sequences of the cloned and sequenced polymorphic AFLP fragments were deposited with GenBank under accession numbers as follows: C/A-1, ; C/A-2, ; C/A-3, ; C/A-4, ; C/A-5, ; C/A-6, ; C/A-7, ; C/C-1, ; C/C-2, ; C/G-1, ; C/G-2, ; C/T-1, ; C/T-2, ; G/A-1, ; G/C-1, ; G/C-2, ; C/C-4, ; and C/C-5, . TABLE 1 | Serovar Typhimurium isolates used in this study RESULTS AND DISCUSSION : Distribution of AFLP fragments and phylogenetic tree based on AFLP variation. | Forty-six serovar Typhimurium isolates comprising nine phage types were analyzed by AFLP with all 16 possible MseI +1-EcoRI +1 primer pair combinations. For primer pairs C/A and C/T, AFLP was done twice for all isolates studied to test reproducibility, and all the fragments being scored were reproducible. A total of 1,340 fragments were scored from the 16 primer pair combinations. There were 345 polymorphic fragments, of which 145 were found in only one isolate, leaving 200 phylogenetically informative fragments . The distribution of 84 fragments was very closely correlated with phage type (see below). Thirty-five different AFLP patterns were observed in the 46 isolates. AFLP fingerprints were clearly different between phage types. The Simpson index of diversity was calculated as 0.98. A phylogenetic tree was constructed. The tree showed a good correlation with phage type, grouping most isolates of the same phage type together. However, there were exceptions. DT64 isolate M1855 and DT44 isolate M1871 clustered outside the main clusters due to an additional 22 fragments present in these two isolates only. These fragments, most likely of plasmid or phage origin, represent gain of DNA shared by these two isolates. In a tree that was derived after excluding these 22 fragments, M1855 is close to the DT64 and DT9 cluster, and M1871 is close to two other DT44 isolates (Fig. and legend). Even after exclusion of the 22 fragments discussed above, the six DT44 isolates still did not form a single cluster. The seven DT12a isolates fell into two groups. Isolates M1867 and M1868 were clustered together away from the other five DT12a isolates, due to the presence of seven fragments in these two isolates only. Again, in a tree that was derived after excluding these seven fragments, M1867 and M1868 moved into the main DT12a cluster. Phage-type-correlated AFLP fragments. | AFLP analysis revealed many phage-type-specific markers. There were 59 phage-type-specific polymorphic AFLP fragments, for which presence or absence correlated absolutely with phage type, and 25 for which there was only one exception to phage type specificity in the 46 isolates studied (Table These 84 fragments were analyzed in detail, as they have the potential to be used to subdivide serovar Typhimurium in the same way as achieved by phage typing. Clearly the chance of finding an exception to phage type specificity goes up with the number of isolates studied, and there is no reason to think that the 59 fragments with 100% correlation would be very different in distribution than the 25 others in a much larger sample. In general we treat all 84 together but for some purposes distinguish the 25 that have a single exception to phage type specificity. Of the 84 phage-type-specific fragments, if we ignore the single exception for the 25 fragments, 63 are present only or absent only in one phage type. Five are present only and two are absent only in DT9 and DT64, five are present only in DT12a and DT108, three are present only in DT108 and DT135, and one is absent only in DT108 and DT141. Two are present only in DT12a, DT64, and DT108. Two are absent only in DT1, DT12a, and DT141, and one is present only in DT9, DT44, and DT64. Of the exceptions to phage type specificity in the 25 near-specific fragments, 12 involve DT1 isolate M1879; four involve DT44 isolate M1875; three involve DT126 isolate M1878; two involve DT12a isolate M1867; and single fragments involve each of the isolates DT64 isolate M1855, DT108 isolate M1886, DT135 isolate M1850, and DT9 isolate M1848. The nine phage types studied could all be differentiated by the 84 fragments. Nature of phage-type-correlated AFLP fragments. | Eighteen phage-type-specific fragments were cloned and sequenced to determine the genetic basis of AFLP-detected variation. The sequences were used in BlastN and BlastX searches . The 18 fragments were also used to probe EcoRI-digested genomic DNA of the 46 isolates . Fifteen have high sequence similarity to known genes or sequences in databases. Three matched plasmid genes, seven matched lambdoid phage genes, one matched a noncoding region of a lambdoid phage genome, and four matched chromosomal genes or sequences in the available S. enterica genomes. Fragments with genes of plasmid origin. | Three fragments (C/G-2, C/T-2, and C/A-7, all present only in DT1) have 97 to 100% DNA sequence identity to plasmid R64 trbC, traP, and traQ genes . The latter two matched contiguous sequences, being bases 33140 to 33390 and bases 33387 to 33484 of plasmid R64, respectively, connected by an EcoRI site. These three fragments hybridized to large plasmid DNA (data not shown) present in DT1 isolates only. It seems clear that DT1 isolates have a large plasmid not present in the 43 other isolates. Fragments with genes of lambdoid phage origin. | Three fragments (C/A-1, C/G-1, and G/A-1), all present in DT64 only, also hybridized to DT64 only. Their sequences matched those of lambdoid phage genes. C/A-1 has 97% sequence identity to the lambdoid phage P22 DNA at positions 4918 to 5049 on the eac gene . C/G-1 has 31% sequence identity at the amino acid level to phage P22 gene 16, which encodes a DNA transfer protein . G/A-1 has 75% sequence identity at the amino acid level to lambdoid phage HK97 gene 38 . G/A-1 also has 66% sequence identity at the amino acid level to phage P22 ORF-56, a hypothetical 6.6-kDa protein in the eae-abc2 intergenic region . It seems clear that DT64 carried a lambdoid phage not present in the other eight phage types. Fragment C/T-1 is shared by DT64, DT12a, and DT108 isolates and also hybridized to DT64, DT12a, and DT108 only. C/T-1 has 100% sequence identity to the phage P22 DNA at positions 27405 to 27555 on gene 8, which encodes scaffold protein . Fragment C/A-6, present only in DT12a and DT108, also hybridized to DT12a and DT108 only. It has 100% sequence identity to a noncoding region of the lambdoid phage 21 genome at positions 4846 to 4883. The above data indicated the presence of lambdoid phages in DT64, DT12a, and DT108. It also indicated that some genes are common to these phages. Three fragments (G/C-2, C/C-4, and C/C-5) have 94 to 96% DNA sequence identity to DNA of ORF626 of S. enterica serovar Dublin contig 707, and ORF626 has 40% sequence identity at the amino acid level to the phage P1 kilA gene. KilA protein is nonessential for phage replication and lytic development . G/C-2 and C/C-5 both are contiguous with C/C-4 across an EcoRI site. G/C-2, C/C-5, and C/C-4 matched bases 40 to 213, bases 1 to 213, and bases 210 to 321 of ORF626, respectively. The G/C-2 sequence has 97% identity to part of the C/C-5 sequence with an EcoRI site at the junction. The C/C-5 sequence has an additional segment extending from a 5'-TTGA-3' site in place of the 5'-TTAA-3' MseI site of G/C-2. The distribution of the fragments showed that DT12a isolate M1867 and DT141 have none of these three fragments and also did not hybridize with them. All other isolates except DT108 have fragments C/C-4 and C/C-5, while DT108 isolates have fragments C/C-4 and G/C-2. The above data suggested that both C/C-4 and C/C-5 are from a phage-related DNA sequence and that fragment G/C-2 arose through a point mutation that created a MseI site in DT108, generating an AFLP fragment shorter than C/C-5. Four fragments related by mutational change adjacent to chromosomal genes. | Two of the sequenced fragments (C/A-4 and C/A-5) are inversely correlated. C/A-4 is present only in DT9 and DT64, while C/A-5 is absent only in these phage types. The sequences are nearly identical, and the C/A-5 sequence was found in the five publicly available sequences of S. enterica genomes (serovars Typhimurium, Paratyphi, Typhi, Enteritidis, and Dublin). The C/A-5 sequence is 100% identical to serovar Typhimurium LT2 DNA bases 1096650 to 1096833. There is a four-base duplication in C/A-4, relative to C/A-5, containing 5'-GGTAGGTA-3' and 5'-GGTA-3', respectively, in an intergenic region between STM1003 and the pncB gene (nicotinate phosphoribosyltransferase ) at serovar Typhimurium LT2 genome sequence bases 1096771 to 1096774 . pncB is a housekeeping gene and also present in other bacteria, such as E. coli, Klebsiella pneumoniae, Yersinia pestis, and Vibrio cholerae, etc., according to the Enterix site. Another two fragments (C/C-1 and C/C-2) are also inversely correlated. C/C-1 is absent only in DT126, while C/C-2 is present only in DT126. The sequences are nearly identical, and the C/C-1 sequence was found in four of the publicly available sequences of S. enterica genomes (serovars Typhimurium, Paratyphi, Typhi, and Enteritidis) but absent in serovar Dublin. The C/C-1 sequence is 100% identical to serovar Typhimurium LT2 DNA sequence bases 1766735 to 1766881. There is an eight-base duplication in C/C-2 relative to C/C-1, containing 5'-ACAGACCGACAGACCG-3' and 5'-ACAGACCG-3', respectively, in an intergenic region between STM1672 and STM1673 at serovar Typhimurium LT2 genome sequence bases 1766792 to 1766799 . For both pairs of fragments, the larger fragment is in one or two serovar Typhimurium DTs only, while the smaller fragment is present in most serovar Typhimurium DTs and the other three or four sequenced S. enterica genomes in which the gene is found. It seems clear that the mutations were due to duplication of a short sequence and not vice versa. Fragments with genes of unknown origin. | Three phage-type-specific fragments (C/A-2, G/C-1, and C/A-3) did not have sequence similarity with any DNA or protein entry in the database searches. C/A-2 and G/C-1 are present in DT12a and DT108 only and also hybridized to DT12a and DT108 only. C/A-3 is present in DT135 and DT108 only and also hybridized to DT135 and DT108 only. These fragments were most likely polymorphic due to gain of DNA. Predominance of phage and plasmid genes in phage-type-related AFLP fragments. | Of the 18 AFLP fragments cloned and sequenced, 15 comprise or include genes identifiable to some extent. Three are plasmid related, eight are phage related, and four are chromosomal gene related. Six of the fragments comprise three pairs that differ in one or a few base pairs. The pair C/A-4 and C/A-5, the pair C/C-1 and C/C-2 (each pair comprising different forms of the same intergenic region), and the pair C/C-5 and G/C-2 differ by a base substitution that affects an MseI site in a phage-related gene. These six fragments thus correspond to three loci, each with two alleles, giving a total of 15 AFLP-based loci for the 18 fragments. Two of these mutational changes are in intergenic regions adjacent to chromosomal genes. Hybridization at high stringency showed that the six fragments discussed above comprise DNA generally present in all nine phage types, the only exception being the phage-related gene which was absent in DT141 and DT12a isolate M1867. However, the other 12 fragments contained DNA present only in isolates with that fragment. This included all of the phage-related genes and plasmid-related genes. These results suggested that gain or loss of DNA rather than mutational change was responsible for most of the polymorphisms detected by AFLP. It should be noted that our focus has been on fragments that correlated strongly with phage types and that this generalization may not apply to all polymorphic fragments. However, some cross hybridization was observed at low stringency . C/T-1 gave low-level hybridization to several fragments of different sizes from DT1 isolate M1879, suggesting that M1879 carried a phage related to that in DT12a, DT64, and DT108. Three plasmid R64-related fragments (C/G-2, C/T-2, and C/A-7, present only in DT1) and three phage P1-related fragments (C/C-5, G/C-2, and C/C-4) also gave low-level hybridization to one fragment from DT141 isolate M1882. These cases of cross hybridization indicated the presence of related DNA, perhaps on a plasmid related to that in DT1 or on a phage related to P1. The data overall indicated that the major consistent differences between phage types relate to the presence or absence of phages or plasmids, well known as mobile genetic elements. The data are certainly consistent with earlier observations showing the role of plasmids and phages in alteration or determination of phage types . For example, a study of DT49 and DT204 isolates of human and bovine origin in Britain showed a correlation among plasmid content, antibiotic resistance spectra, and phage type . Tetracycline-resistant DT204 strains all carried nonautotransferring plasmids coding for tetracycline resistance only. Spontaneous loss of tetracycline resistance plasmid NTP108 converted a DT204 strain to a DT49 strain. When plasmid NTP108 was reintroduced into the DT49 strain, the phage type changed back to DT204. DT204 differed from DT49 by loss or reduction in sensitivity to specific typing phages, and further experiments showed that plasmid NTP108 restricted growth of these specific serovar Typhimurium typing phages . The identification of phage-type-specific phage and plasmid genes provides a basis for exploration of the genetic basis of phage resistance and phage type variation. However, our data do not address the relationship of the phages and plasmids detected with the various sensitivities of each phage type to the typing phages. It is possible that some of the phages or plasmids whose presence was detected in this study are simply markers for variants of serovar Typhimurium. The fact that phage and plasmid genes were dominant in our observations does not in itself imply that these elements affect sensitivity to typing phages but suggests only that the rate of gain and loss of these genetic elements affects AFLP pattern more rapidly than gain or loss of restriction sites by mutation or lateral transfer of chromosomal genes. This is an interesting topic for further study. While this paper was being written, Tamada et al. reported sequences of two AFLP markers of serovar Typhimurium. One sequence showed homology with a segment of P22 phage, and the other showed homology with a segment of traG, an F-plasmid conjugation gene. These data are consistent with our data. Given the role of phage-related genes in our AFLP tree and the role attributed to carriage of temperate phage in determining phage type, we compared the relationships of phage types revealed by the AFLP tree with the relationships based on the phage sensitivity in the Anderson phage-typing scheme. The sensitivity of each isolate to each typing phage is assessed by number and turbidity of plaques . The lysis pattern for each of 30 phages as updated (Linda Ward, personal communication) was used to grade the sensitivity of each of the nine DTs in steps from 1 to 24 for increasing degree of lysis. These data were treated as 30 loci each with 24 character steps for construction of the tree, such that relationship would be determined by degree of similarity based on the 24-step scale, and a phenotypic tree was constructed by the parsimony method (data not shown). The phage sensitivity tree was very different from the AFLP tree, with, for example, DT12a, DT64, and DT108 closely related in the phage sensitivity tree but not in the AFLP tree. We suggest that the relationships in the AFLP tree are determined to a significant degree by the overlap in genes present in phages and plasmids carried by the strains, whereas the phage sensitivity tree is determined by the degree of sensitivity to specific typing phages, which are apparently not closely related. Comparison of AFLP with PFGE. | Several molecular methods have been tested for use in subtyping of common S. enterica serovars such as Typhimurium. The only one that has achieved a significant level of use is PFGE, which in some cases has also been used in place of phage typing for primary subdivision of serovar Typhimurium. It has also proved to be useful in subdividing a major phage type . We have studied AFLP in large part because the fragments obtained are small and quite easily cloned and sequenced to determine the underlying basis of the distinctions observed. In the present study, AFLP based on 16 primer pair combinations showed a Simpson index of diversity of 0.98, a higher discriminatory power than that for PFGE applied to Spanish serovar Typhimurium isolates with a Simpson index of diversity of 0.87 . AFLP in the present study showed good correlation with phage type, grouping most of the serovar Typhimurium isolates by phage type and enabling differentiation of phage types. However, AFLP is not suitable for routine use for subdivision of a S. enterica phage type, because with such closely related strains one has to use several primer pairs to obtain the high discriminatory power possible, and we still have difficulty with automatic computer scoring of AFLP fragments. However, this study has demonstrated that AFLP can be used to identify segments of DNA that have a distribution that correlates with phage type. We propose that this knowledge be used to establish a subtyping scheme for serovar Typhimurium that offers the advantages of AFLP with potential for great ease of operation as discussed below. PFGE also showed some correlation with phage type when applied to Danish serovar Typhimurium isolates . However, PFGE uses one restriction enzyme, which cleaves infrequently and generates a small number of large DNA fragments. It can detect mutational changes only in the small number of restriction enzyme sites, or changes such as large insertion-deletion events or chromosomal rearrangements, which cause several-kilobase differences in fragment size. It is difficult to detect insertion-deletion events involving only a few kilobases. In contrast, AFLP uses two restriction enzymes and generates small fragments. AFLP is ideal for detecting insertion or loss of segments of DNA, and it is easy to clone such DNA and determine its distribution among phage types as done in this study. Also, such DNA can be used to develop multiplex PCR or microarray typing. AFLP has more potential than PFGE to discriminate among closely related isolates, as the number of AFLP fragments can be greatly increased by using different enzymes or primer sets, to find more phage-type-specific markers. Potential for microarray typing of serovar Typhimurium using AFLP markers. | The use of AFLP has enabled us to identify segments of DNA for which presence or absence correlates with phage type. Two-thirds (12 of 18) of cloned phage-type-specific AFLP fragments were polymorphic, because the DNA is present or absent as indicated by Southern hybridization. These AFLP markers provide a good basis for PCR-based typing or application of microarray technology. A tree was constructed by using only the 84 phage-type-specific AFLP fragments . All nine phage types studied were differentiated, and furthermore, a minimum of six such fragments (C/A-3, C/A-4, C/A-7, C/C-2, C/C-4, and C/T-1) is sufficient to allocate all isolates to their respective phage types. One can envisage the use of multiplex PCR or microarray technology to obtain comparable data for typing, and we suggest that it is entirely practicable to develop subtyping of serovar Typhimurium by using AFLP-defined markers. We draw attention to the advantages of AFLP and multiplex PCR for the extensive preliminary work that is needed and for ongoing development of such a scheme. Fluorescent AFLP permits very accurate size standardization of AFLP fragments, which would allow interlaboratory comparisons for the choice of fragments to be used as the basis for a typing scheme. Multiplex PCR using primers based on selected phage-type-specific AFLP fragments would be ideal for trials of such a typing scheme, which could be run in parallel with phage typing. However, microarray technology is better in handling the number of sequences required. We have looked at only nine phage types and a limited number of isolates from one region of the world. Further work is needed to establish the generality of these observations and to find DNA that gives good correlation with phage typing. We see great advantage in using a scheme that correlates well with phage typing, as the same classification could be followed and the currently named DTs would continue to be recognized. A microarray-based scheme based on DNA fragments shown to correlate with phage type offers major additional advantages. Unlike PFGE or AFLP, it is relatively easy to scale up to include more test sequences with microarray typing. We have already demonstrated the feasibility of a molecularly based replacement for serotyping , and given the huge number of spots that can be fitted onto a microarray, it would be possible to include a full serotyping scheme and several subtyping schemes on one chip. There is the additional advantage that it would be relatively easy to subdivide the frequently occurring DTs, as there is more information available than needed for replacement of phage typing. In this context we should note that phage typing uses a standard system with 16 phages per plate, making scale-up difficult. The phage-typing system of serovar Typhimurium has proved to be very useful for surveillance of infections. However, the present phage-typing scheme requires considerable experience for consistent scoring and to access specific typing phages. It is difficult to propagate the typing phages to obtain stocks with the same characteristics as the original stocks. This is probably due to recombination between the propagating phage and other phages present in the propagating strain, giving a genetically mixed population in the final stock. For this reason the stocks are prepared in the Enteric Reference Laboratory at the Public Health Laboratory Service in London, United Kingdom, for worldwide distribution. It has even been suggested by Schmieger that, when the stocks of original typing phages become exhausted, the new propagated typing phages may affect the consistency of the Anderson phage-typing scheme. Microarray technology has much potential for screening with DNA sequence markers. AFLP provides the means to identify suitable DNA segments for use in such technology to provide a tool for subtyping serovar Typhimurium which could be applied more widely than the present phage-typing system, which is generally confined to major government laboratories. This would have a very positive impact on epidemiological investigation of disease outbreaks and studies of pathogenicity of this important pathogen. Thus, AFLP's real potential for genotyping serovar Typhimurium probably lies in identification of DNA segments that are discriminatory among phage types, followed by multiplex PCR to confirm their value in parallel with phage typing, and then development of microarray technology for routine typing. FIG. 1. | Phylogenetic tree based on AFLP fingerprints with all 16 primer pair combinations of MseI +1-EcoRI +1 from MseI- and EcoRI-digested fragments. Phylogenetic tree based on AFLP fingerprints with all 16 primer pair combinations of MseI +1-EcoRI +1 from MseI- and EcoRI-digested fragments. Bootstrap values are percentages of 1,000 replications and are indicated at the nodes if the value is greater than 50%. Isolates indicated by an asterisk are shown twice, in the position found in the tree as originally derived (solid line) and in the position found when fragments specific to these isolates were omitted (dashed line; see text). The value of the bar represents genetic distance measured by Dice coefficient. FIG. 2. | Tree constructed by using phage-type-specific AFLP fragments of all 16 primer pair combinations of MseI +1-EcoRI +1 from MseI- and EcoRI-digested fragments. Tree constructed by using phage-type-specific AFLP fragments of all 16 primer pair combinations of MseI +1-EcoRI +1 from MseI- and EcoRI-digested fragments. The value of the bar represents genetic distance measured by Dice coefficient. TABLE 2 | Numbers and types of polymorphisms observed in AFLP fragments in the size range from 40 to 600 bp TABLE 3 | Phage-type-specific AFLP fragments TABLE 4 | Sequences of phage-type-specific AFLP fragments Backmatter: PMID- 12398790 TI - The MRC trial of assessment and management of older people in the community: objectives, design and interventions [ISRCTN23494848] AB - Abstract | Background | The benefit of regular multidimensional assessment of older people remains controversial. The majority of trials have been too small to produce adequate evidence to inform policy. Despite the lack of a firm evidence base, UK primary care practitioners (general practitioners) are required to offer an annual health check to patients aged 75 years and over. Design | Cluster-randomised factorial trial in primary care comparing a package of assessments (i) universal versus targeted assessment and (ii) management by the primary care team (PC) or a multidisciplinary geriatric assessment team (GM). The unit of randomization is the general practice. Methods | Older people aged 75 and over eligible for the over 75s health check and excluding those in nursing homes or terminally ill were invited to participate. All participants receive a brief assessment covering all areas of the over 75s check. In the universal arm all participants also receive a detailed health and social assessment by a study nurse while in the targeted arm only participants with a pre-determined number and range of problems at the brief assessment go on to have the detailed assessment. The study nurse follows a standard protocol based on results and responses in the detailed assessment to make referrals to (i) the randomised management team (PC or GM) (ii) other medical services, health care workers or agencies (iii) emergency referrals to the GP. The main outcomes are mortality, hospital and institutional admissions and quality of life. 106 practices and 33,000 older people have been recruited to the trial. Keywords: Background : Indications for possible benefit from regular assessment of elderly people came from early studies several decades ago in the UK which found high levels of undetected problems in elderly people and highlighted the need for a systematic approach to problem detection. Around the same era, an approach emerged which emphasized function and disability in the medical care of elderly people and questionnaires were developed to assist in the assessment of a range of physical, mental and social dimensions . Doubts about the feasibility of routine assessment of elderly people led to the notion of two stage targeted screening to identify those at greatest risk or greatest need . In the 1980s results from three randomised controlled trials, which examined the benefit of socio-medical assessment, were published . Two of these trials took place in the UK in the setting of general practice and one in Denmark among elderly people living in the community . These trials suggested some possible benefits on mortality and hospital admissions (mainly the Danish trial) but none of the trials provided convincing evidence for regular assessment of elderly people. Despite these equivocal results, the Department of Health in 1990 introduced a contract of service for general practitioners, which required them to offer an annual assessment to patients aged 75 years, and over . Although the contract specified the broad areas for assessment , the method, level and nature of assessment were not defined. Different models of initial assessment have been used (e.g. postal, lay person and nurse) and targeted screening has been recommended but none of these strategies have been rigorously evaluated and compared. Models of subsequent management of problems identified through the assessment process also require evaluation. In studies conducted mainly in North America, multidisciplinary teams in the hospital setting (usually geriatrician, nurse, therapist and social worker) appear to offer advantages in terms of survival, functional status and use of hospital services , but the costs and benefits of the multidisciplinary team as a integrated component of multidimensional assessment, when compared with the usual model of primary care, have yet to be established. Table 1 | Department of Health Contract of Service with General Practitioners 1990 The principal objective of the MRC trial of assessment and management of older people in the community is to evaluate a package of multidimensional assessment and management of older people in the context of the 1990 contract of service. The two main components of the package are the method of assessment -- Universal or Targeted- and the method of clinical management -- multidisciplinary geriatric team (GM) or usual primary care (PC). A secondary objective is to compare different methods of administering a brief screening questionnaire (postal, lay or nurse) These three methods were chosen because they that had been advocated as part of an assessment process (14) (with obvious differences in cost implications) but no formal evaluation of their performance in a single trial has been carried out. We were interested in establishing whether response rates, levels of missing information and, for certain health conditions, sensitivity and specificity compared to measurements at the detailed assessment, varied for the three methods. Design : The study is a cluster-randomised trial with a 2 x 2 factorial design i.e. practices are randomised to: Universal assessment (plus randomised to GM or PC) or Targeted (plus randomised to GM or PC) . In addition to the main randomisations, practices are also randomised to one of the three methods of administering a brief questionnaire, which are balanced across the main randomised groups. All randomisation is at the level of the general practice. Figure 1 | Design of MRC trial of assessment and management of older people in the community Design of MRC trial of assessment and management of older people in the community Description of procedures | All practices administer a brief assessment questionnaire to participants in the study either by post, or by a layperson, or by a nurse. Following the brief assessment, practices in the Universal arm carry out a detailed examination by the study nurse in all patients (irrespective of their responses on the brief assessment) while those in the Targeted arm carry out a detailed assessment only in those patients who "trigger" on the brief assessment. The study nurse follows a standard protocol based on results and responses in the detailed assessment to make referrals to (i) the randomised team (PC or GM)(ii) other medical services, health care workers or agencies (iii) emergency referrals to the GP. Methods : Methods of assessment | The brief assessment questionnaire | The brief assessment questionnaire, developed in conjunction with Wallace and Williams (who have published the binary response version of the questionnaire) covers all the areas specified in the GP contract: social environment, activities of daily living, sensory problems, mobility, physical symptoms including continence, mental condition, use of medication. We chose to use a graded response scale based on the results of pilot studies comparing graded with binary responses, which showed that binary responses inflated positive responses and that patients had difficulty making choices on a binary scale. The areas and the corresponding questions and scale are shown in Table . Additional questions on alcohol consumption, cigarette smoking and physical activity have been included for epidemiological purposes. Criteria for triggering to the detailed assessment are 3 or more problems identified from the brief assessment or any one of 4 "serious" symptoms. According to the randomisation, patients are interviewed either by a lay person (usually a member of the practice staff) or the study nurse or the questionnaire is mailed. Patients are invited to attend the surgery for the interview but are given the option of the interview being carried out at home. Identical questions are asked in each version of the questionnaire but for the postal questionnaire the scoring of the triggers is carried out by the study nurse. The postal questionnaire was printed in large font to permit easy reading. Table 2 | The brief assessment questionnaire areas, specified in the GP contract and corresponding questions The detailed assessment | Patients are invited to attend the surgery for the detailed assessment but are given the option of the assessment being carried out at home. The detailed assessment covers the same areas as specified above but in greater depth for example, whispered voice test for hearing , Glasgow Acuity Cards for vision , Mini Mental State Examination for cognitive impairment , the Geriatric Depression Scale . Additional questions include more detailed assessment of symptoms (e.g. Rose chest pain questionnaire for possible angina , respiratory problems, urinary and faecal incontinence, examination of legs and feet and a modified version of a checklist for possible drug interactions . Additional biological measurements include blood pressure, heart rate, and dipstick for blood, protein, and urine. In all patients a blood sample is taken for a full biochemical screen. Additional laboratory investigations are triggered by question responses or abnormal results e.g. faecal occult testing for a positive response to a question on blood in the stools, an MSU for a positive dipstick result for protein or blood. Patients are also assessed for social problems such as financial difficulties, social isolation. The choice of conditions to screen for was based on a review of the available literature . The study nurse follows a protocol, based on results and responses in the detailed assessment, to make referrals to (i) the clinical teams (PC or GM) (ii) other medical services, health care workers or agencies (iii) emergency referrals to the GP. Referrals to PC/GM are for serious clinical problems, such as depression, frequent falls, abnormal biochemical results, severe breathlessness, and severe leg oedema. A full list is given in Table . The conditions for referral to the teams were based on the results of a survey of 90 general practitioners from the MRC General Practice Research Framework (MRC GPRF) and 52 geriatricians, who had indicated an interest in taking part in the study. Other referrals are to specialities or professions for a variety of problems, such as ophthalmology for visual impairment (not due to refractive errors), audiology, continence advisors, community psychiatric nurse, social services . For certain problems, patients are advised to seek a consultation e.g. with an optician for refractive error (VA<6/18 corrected with pinhole), or to a dentist for problems with mastication. Table 3 | Criteria for referral to primary care team (PC) or geriatric evaluation and management team (GM) 1 Table 4 | Detailed assessment: referrals to other health professionals/agencies Management teams | The teams follow their "usual" practice and there was no attempt to impose a formal protocol. Information is collected on investigations, diagnoses, services and treatments that result from the referral and any further referrals. Outcome measures | The principal outcome measures for hypothesis testing are mortality, hospital and institutional admissions and quality of life. Mortality follow-up is achieved by registering all eligible patients with ONS for notification of death, date and cause of death. Hospital admissions are collected for each participant for a 2-year period from the time of the brief assessment using information from hospital discharge letters in the patients' GP records. Information collected includes specialty, dates of admission and discharge, diagnoses, specialty of consultant. The discharge letter is considered to be a reliable source since this is the routine method of providing information to general practitioners from hospital services. Institutional admissions are collected on an ongoing basis for each patient from the date of the baseline assessment. Quality of Life interviews are carried out in the patient's own homes by fieldworkers who are independent of the practice. The interviews take place (i) at baseline and prior to the assessment (ii) 18 months following the baseline interview (iii) 36 months following the baseline. The Quality of Life core interview schedule includes four dimensions from the Sickness Impact Profile (SIP) (mobility, self-care, home-management, social interaction) , and the Philadelphia Geriatric Morale Scale (PGMS), , a 17-item measure of morale specifically developed for use with older people. We used slightly different follow-up periods for outcome collection since we would expect any effect on hospital admissions to be seen at an earlier rather than later stage whereas mortality effect might show some lag. Quality of life is measured both at 18 months and at 3 years, which covers the same period as both the hospital admissions and the mortality. Use of services | Information on use of services is being collected for the economic analysis. Use of public sector services (health and social services) is collected through two sources (i) in the quality of life questionnaire at baseline, 18 and 36 months) (ii) through a postal questionnaire mailed to randomly selected cross sectional samples of patients at 6 monthly intervals across the follow-up period. This permits us to have information both from a longitudinal sample over the 3 years of the trial from participants in the quality of life sample while the cross sectional random sample provides a more representative sample across the follow-up period since it includes those those who may die or leave the practice. Trial hypotheses and sample size | Sample size methods | Sample size calculations for mortality and hospital admissions were calculated from a program provided by Martin Shipley at University College London, Department of Epidemiology, using methods for cluster randomisation described in Shipley et al and under the assumption that the rates follow a Poisson distribution. The program uses the formula where study power is given by 1-beta, Cbeta is the critical value of the normal distribution in the upper tail of the standard normal curve, Calpha/2 is the critical value of the normal distribution for a two sided value of alpha (in our case taken to be alpha = 0.01), lambda is the average value of the underlying event rate, Delta is the difference to be detected between the event rate in the main randomised groups of the trial (universal minus targeted or GM minus PC), 2 sigma2(1-kappa) is the variance of the underlying event rates, nH is the harmonic mean of all 2 m sample sizes and 2 m is the number of clusters. We took the conservative assumption that the level of matching of the practices would be minimal or essentially zero although we did attempt to balance the practices by Jarman score and SMR by stratified randomization across the joint tertiles of SMR and Jarman. Expected mortality rates were based on ONS statistics for England and Wales with an adjustment downwards of 0.7 of the background mortality rate on the basis that the trial population would have lower rates (healthy participant effect). Rates of hospital admission were adjusted down on the basis of an expected "healthy" participant effect and additionally on the basis that hospital admissions might be under reported and that annual rates do not distinguish between persons and admissions. For hospital admission rates we therefore assumed that the rates might be a half of those published. Thus in both calculations we used conservative estimates of the rates. Sample sizes for institutional admissions could not be calculated as there are no national data on rates of institutional admission. The sample size calculations assumed an average of 500 eligible patients per cluster followed for an average of 3 years. Sample size estimates for the quality of life measures used the formula given by Hsieh based on the Z approximation, N = 8(Sb2 + Sw2/m) (zalpha/2 + zbeta)2/d2 where N is the total number of clusters, Sb2 is the between cluster component of variance, Sw2 is the within cluster component of variance, m is the number of individuals within each cluster, and zalpha/2 and zbeta/2 are the critical points of the normal distribution for alpha of 0.01 and 1-beta of 0.9 and d is the difference between randomised groups in the means of qol scores. Estimates of Sw2 and d2 were based on SIP scores from pilot studies of 52 older people attending geriatric out patient clinics and POMS scores from 115 patients on the care of the elderly wards at the Hammersmith Hospital. We expected 75% of baseline responders to provide quality of life data at the 3 year end point (m = 350). Since no data were available for Sb2 (which is expected to be considerably smaller than Sw2) we used varying estimates of the intraclass correlation coefficient (Sb2/ Sb2 + Sw2) (0.01, 0.02, 0.04). Estimates of sample size | 108 practices (with an average of 500 patients aged 75 years and over) are required to detect differences between Stage 1 assessment methods (targeted versus universal) of: at least 15% in mortality assuming a background mortality rate of approximately 60/1000 person years over an average follow up period of 3 years; at least 10% in 2 year hospital admissions assuming a background first admission rate of 150/1000 person years. This sample size is sufficient to detect differences between the two management strategies (GM and PC) of at least 22% in mortality and at least 15% in hospital admissions. Smaller numbers of practices are required to detect differences in quality of life. A random sample of 24 practices is adequate to detect differences between Stage 1 assessment methods of at least 15% in PGMS and SIP and between GM and PC of at least 25% in PGMS and SIP. The size of the effects to be tested were based on previous trials and a realistic expectation of what might be achievable and of public health importance. Trial population | The trial is being conducted in practices recruited through the MRC GP Research Framework with list sizes of between 200 and 700 patients aged 75 years and over and selected to be representative of the joint tertiles of Jarman and Standard Mortality Ratios (SMRs) in UK practices. To be eligible for randomisation in the trial, recruited practices first had to obtain the agreement of the local geriatrician to participate (in order to ensure no selection bias of geriatricians in practices subsequently randomised to GM or PC). Eligible patients were aged 75 years and over in the year the practice undertook the assessments, excluding anyone in long-term care or with terminal illness. Patients in residential or sheltered accommodation were included. Method of allocation to groups | Allocation to groups was by a computer generated randomisation list, stratified by Jarman and SMR tertile as practices were recruited to the trial. Informed consent | The over 75s check is a contractual obligation of general practitioners. The letter of invitation to the study was done in the context of the letter of invitation to the annual check. The letter of invitation informed patients that the check could be done as part of a research study with a brief description of the intervention (this varied according to the practice randomisation). Patients were given the opportunity to have the check carried out in the usual way and reassured that non-participation in the study would not affect their usual care. Ethical approval | Local Research Ethics Committee approvals were obtained for all the practices participating in the trial. Analysis | The principal analysis is of the components of the package of interventions and therefore all analyses will take account of the other main randomisation. The results will be presented for (i) universal versus targettted (ii) GM versus PC. In order to ensure that the method of administering the brief questionnaire does not influence the main randomisations, the analyses will also be adjusted for the 'nurse vs. lay vs. postal' randomised intervention. The primary population for analysis is the "Intention to Treat" population (i.e. all those eligible for the trial and invited to take part irrespective of participation). Pre-specified covariates are age, sex and (at practice level) SMR and Jarman score. There will be no sub-group analyses in the primary analysis. All analyses will take account of cluster randomisation which has the effect of increasing the standard errors of the estimates and therefore it is unlikely that spurious significant effects will be detected. Secondary analyses will be conducted as above on the "per protocol population" Economic analysis | The primary objective of the economic analysis is to determine the relative costs and costs per life year gained of the different components of the assessment and management packages. Cost analysis will compare the direct health and social care costs of assessments and the longer term costs of care and treatment following the interventions (use of services, hospital admissions, and instiututional admissions). Service use data over the follow-up period of the trial will be obtained from longitudinal samples (as part of the quality of life questionnaire) and from the repeated cross sectional samples. Progress on trial | 109 practices were randomised and study staff trained in the study protocol; three practices withdrew before carrying out any of the study interventions (one of these was also randomised to the quality of life collection). It was too late to replace these practices so the final achieved target is 3 practices short. 106 practices and 33,000 patients are participating in the trial. The practices are equally distributed over the Universal (n = 53) or Targeted (n = 53) arms and of the PC (n = 53) or GM (n = 53) arms of the trial. Discussion : This trial is, by far, the largest and most comprehensive study internationally to examine the benefits of multidimensional assessment of older people in the community. Although a number of trials have been published since the introduction of the 1990 contract, including three from the UK, they have, in common with previous trials, been too small to produce results of sufficient precision and certainty to inform policy decisions . Studies conducted in the US are difficult to translate to the UK health care setting. Previous trials have also suffered from a number of other methodological problems: individual patient randomisation in the same healthcare setting (e.g. general practice) which may lead to contamination of randomised groups and dilution of effect; lack of a clearly defined protocol for intervention and referral; and (in the US studies) low participation rates and over-representation of fit older people. One systematic review and three meta analyses of these trials have been published , but have come to different conclusions concerning benefits on major outcomes, such as mortality and hospital admissions. Our trial avoids some of these design problems by employing: randomisation at the general practice level, which reduces the possibility of contamination; clearly defined protocols for the intervention and referrals; minimal exclusion criteria (terminally ill or in institutional care); high power to test hypotheses on major outcomes. These outcomes were chosen because a number of trials had suggested possible benefits on mortality, institutional (or nursing home) and hospital admissions. Individual studies and meta analyses have also looked at outcomes such as functional decline or physical morbidity -- we used comparable domains of health related quality of life as an outome to capture these aspects and also included morale and social interaction as other relevant outcomes. Our trial compares different strategies of multidimensional geriatric assessment. These different levels are analogous to the different intensities of screening described by Stuck and colleagues in their recent meta analysis (33), for example universal assessment with management by a specialist geriatric team in our trial is comparable to their category of "multidimensional geriatric assessment and follow-up" in the meta-analysis. Hence our results are applicable across differing health care contexts. Our trial interventions and referrals across a broad range of health care professionals and other agencies address disability and impairment as well as disease which we consider to be an appropriate balance rather than adopting either a purely medical or functional model of health at older ages. There are a number of limitations of our trial. Cluster randomisation trials require substantially greater numbers than trials of individual randomisation, and are more susceptible to practice effects e.g. loss of all patients to follow-up if a practice withdraws. The large numbers also made certain aspects of data collection prohibitively expensive. For this reason we have not collected any detailed information on services and treatments already being received by trial participants ie prior to the trial interventions, although we will have some limited information on these if part of the criteria for referral. Also, funding did not permit collection on hospital admissions over a longer period. Nor were we able, for costs reasons, to repeat the intervention packages over several years or, to examine the benefits of annual assessment versus some other time period such a triennial assessment. A further limitation of our trial design was that it was not possible to have a control group (i.e. a group randomised to no intervention) because of the contractual obligation of general practitioners to offer an annual health check to the over 75s. It is extremely unfortunate that this policy was introduced (despite the lack of a firm evidence base) because it has meant that a properly controlled study cannot be conducted in the UK. As with other trials, especially in health services research, there are issues of generalisability of the results to other relevant health care professionals (practice nurses, general practitioners and geriatricians) and to patients. The practices in the trial belong to the MRC General Practice Research Framework and, although the patients in those practices were representative of the wider patient population (in terms of deprivation and mortality), the general practitioners have a interest in being involved in research. However we have no reason to believe that the general practitioners or the nurses had any particular expertise in geriatric assessment or management. The geriatricians were recruited to the study by virtue of being the linked local geriatric service to the general practice and were not selected for any particular expertise or experience in research. Ensuring agreement of the local geriatricians to take part in principal, which was a criterion for randomisation of the practice (i.e. before knowledge of whether the practice was randomised to PC or GM), reduced the likelihood of selection bias of geriatricians. Nonetheless it will be important to describe the clinical management through the collection of process information so that, if the trial shows a benefit for a particular package of assessment and management, its components can be described. The trial results are also of direct relevance to the "single assessment " recommended by the National Service Frameworks for Older people , in particular the decision on how and to whom the brief overview assessment should be targeted. The levels of benefit to be tested in the trial, if demonstrated, will result in important reductions in mortality, hospital and institutional admissions and improved quality of life. These results will be relevant, not only to the UK older population but also to many other countries and settings where health care of older people has become a policy priority. Competing interests : None declared Authors contributions : All authors contributed to the design of the trial. AF wrote the manuscript and DJ, CB and AT provided comments. All authors approved the final manuscript. Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12398771 TI - The International Sepsis Forum's controversies in sepsis: corticosteroids should not be routinely used to treat septic shock AB - Abstract | Corticosteroid treatment of severe sepsis has been one of the most controversial clinical issues in critical care. In fact, few agents can claim to have been evaluated in scores of studies spanning 3 --4 decades. Yet, convincing proof that corticosteroids are useful pharmacologic agents in the treatment of this major clinical problem remains elusive. Recently, interest has resurfaced but this time the focus is on a steroid replacement approach for what has now been termed "relative adrenal insufficiency" rather than relying on the pharmacologic effects of steroids. This route holds promise, but proof remains lacking. Keywords: corticosteroids, septic shock, corticosteroids, septic shock, Introduction : The issue of whether to use corticosteroids for treating septic shock has been ongoing for 20 or 30 years. Indeed, there is much clinical and preclinical data that provide ample reasons for using corticosteroids in septic shock. For instance, we know that steroids increase catecholamine-stimulated contractility, vasomotor catecholamine response, and adrenergic receptor density. They also prevent desensitization of beta receptors, and they may provide adrenal replacement. Cortisol status : The normal cortisol range is 5 --24 mug/dl, and during the stress response this rarely goes above 200 mug/dl. During septic shock, classic adrenal insufficiency is rare (0 --3% of cases) and cortisol levels can range from 7 to 400 mug/dl. However, 50% of patients have levels below 20 mug/dl . A series of studies has examined the relative adrenal insufficiency that is sometimes seen in septic shock. In one such study patients were stimulated with adrenocorticotrophic hormone (ACTH) and their blood cortisol levels were measured at 0, 30, and 60 min. High mortality (82%) was associated with high baseline cortisol levels that did not respond to ACTH stimulation, whereas low mortality (26%) was associated with low baseline cortisol levels that did respond to ACTH stimulation. The intermediate values correlated with intermediate mortality. This makes perfect sense and is very logical, which is one of the reasons why I think it has been adopted so quickly even though the data are fairly preliminary and have not been reproduced. A recent study examined a cohort of patients with septic shock, separated them according to the presence or absence of relative adrenal insufficiency, randomized them into a placebo-controlled trial of steroid therapy, and then stratified the analysis according to the adrenal insufficiency variable . At the onset, patients were tested for eligibility using ACTH stimulation (although the results of this were not immediately disclosed) and were then randomly assigned to either hydrocortisone plus a fluorinated steroid or placebo for 7 days, in other words adrenal replacement therapy versus none. The authors found that the group with relative adrenal insufficiency (by the new criteria) were observed to have statistically significantly improved survival. This tells us that there is a select population of patients with adrenal insufficiency during septic shock who improve when treated with steroids. In spite of the fact that the responders (to ACTH) did not show a benefit, steroids worked in the group as a whole. Hence, that categorization did not actually matter. The 'take home message' ought to be that steroids work in septic shock based on the overall study result in all enrolled patients on an intent-to-treat basis. The sticky part is that this population has been studied with steroids many times in the past but without success. Predicting outcome : So, do cortisol levels predict outcome? Several studies have looked at this: five studies found that nonsurvivors have higher levels of cortisol ; one found that survivors have higher levels ; and three found no correlation . Therefore, these reports taken together indicate that cortisol levels do not predict outcome. Does response to ACTH predict outcome? Two studies found no correlation with outcome, whereas two other studies found that a small (inadequate?) response predicts death. Therefore, as yet there is not enough evidence to state that response to ACTH is a critically important determinant of outcome. Moreover, in those studies, as well as in others, cortisol response is poorly reproducible, even on the same day and in the same patient . Thus, even if response to ACTH were a good measure, it is hard to get reliable clinical data. So what do we truly know about cortisol assessment in the intensive care unit? We know that appropriate levels are not yet known, low random values are difficult to interpret, and the concept of relative adrenal insufficiency is still poorly established. Thus, why not study all comers in septic shock, ignoring whether the patient is adrenally insufficient or not? A couple of studies have addressed this. A study by Briegel and colleagues was a double-blind, randomized trial of 40 patients that used the duration of vasopressor therapy as the end-point. The inclusion criterion was patients on vasopressor therapy with a cardiac index greater than 4, in other words those with vasodilated shock. The treatment was 100 mg hydrocortisone over the first 30 min, and then 0.18 mg/kg per hour for as long as the vasopressor was required. For shock reversal there was no significant difference from the control group; the median time on vasopressor was different (but it is not clear what this means because shock reversal showed no difference); and the mortality was identical. In another randomized double-blind study (this time in 41 patients with septic shock requiring vasopressors for more than 48 hours) , the end-points were shock reversal, haemodynamics, and survival. Hydrocortisone 100 mg was given every 8 hours for 5 days, versus placebo. There was a statistically significant increase in the number of patients whose shock had reversed by day 7, and there was a trend toward increased survival, although the numbers were small (6 deaths out of 22 in the corticosteroid group and 13 out of 19 in the control group). Based on those two studies , the ACTH response does not appear to be necessary to see the shock effect. One interpretation of these data would be that stress steroids was the right approach, whereas high-dose pulse steroid therapy in all patients (not just adrenal insufficient patients) with septic shock was the wrong approach. Some have stated that steroids might have worked if they had been administered in a less potent way, for instance 100 --300 mg/day over several days rather than the high doses given in earlier studies, suggesting that replacement of stress levels is what is important. However, this may not be the case (see below). Treating shock versus mortality : This brings us to the study conducted by Sprung and colleagues of high-dose hydrocorticosteroids in septic shock . Here the end-points were survival and reversal of shock. It was a randomized, prospective trial of 59 patients with septic shock and systolic blood pressure below 90 mmHg after adequate fluids. Treatment was industrial strength methylprednisolone 30 mg/kg or dexamethasone 6 mg/kg, and if shock persisted the dose was repeated after 4 hours to a maximum of two doses (i.e. very short pulse therapy). Shock reversal occurred more rapidly in the corticosteroid-treated patients than in the control group, and this lasted until day 14. There was no difference in mortality between the control group and the corticosteroid group at 25 days. Therefore, shock was actually improved more rapidly with treatment but mortality was no different. There may have been some late steroid deaths that were related to treatment. What do we do now? One or two pulses of very high-dose steroids is effective in reversing shock, but with a conflicting effect on mortality. In summary, the pathophysiologic basis for using steroids in septic shock remains very confusing; the field is plagued by a large number of small trials and there is no doubt that steroids carry risk. Thus, if we do not know that something works, then we should not be using it. A large randomized trial is really needed here to answer this question. We do not know who to treat, we do not know when to treat, we do not know the dose to use, and we certainly do not know for how long to maintain treatment. Finally, in a meta-analysis of all major studies conducted since 1963 it was shown that the relative risk for death with corticosteroid treatment is slightly increased; in other words, there is a slight increased risk for death associated with corticosteroids. Therefore, the challenge is for us to come up with a study design that can give us a clear idea of what we should be doing in our everyday practice. For now, the data indicate that we should not administer steroids to patients who are in shock. Competing interests : PST received an honorarium from the International Sepsis Forum for helping to write this commentary. Abbreviations : ACTH = adrenocorticotrophic hormone. Backmatter: PMID- 12398780 TI - Clinical review: Intensive care follow-up -- what has it told us? AB - Abstract | The majority of intensive care practitioners, until comparatively recently, was content to discharge surviving patients to the care of referring primary specialty colleagues who would undertake subsequent inpatient and outpatient care. With the exception of mortality statistics from clinical studies, the practitioners were thus denied the opportunity of understanding the full impact of critical illness on a patient and their family. The concept of the intensive care follow-up clinic has developed more recently, and is run commonly on multidisciplinary lines. These clinics serve a number of purposes, but importantly have drawn attention to broader patient-centred outcomes after intensive care. Investigators are just beginning to identify, and in some cases quantify, the postdischarge burden on patient and family; additional useful data have also come from follow-up of specific disease states. The purpose of the present review is to highlight some of the important issues that impact on recovery from critical illness towards an acceptable quality of postdischarge life. We have concentrated on the adult literature, and specifically on studies that inform us about the more general effects of critical illness. Head and spinal injury are thus largely ignored, as the effects of the primary injury overwhelm the effects of 'general' critical illness. Keywords: critical illness, follow-up studies, neuropsychology, outcome assessment (healthcare), quality of life, critical illness, follow-up studies, neuropsychology, outcome assessment (healthcare), quality of life, The impact of critical illness on life expectancy : The attrition imposed on health reserve by a critical illness is highlighted by a number of long-term follow-up studies. The 5-year mortality rate for intensive care unit (ICU) survivors is reported to be more than threefold higher than for the general population. Actuarial survival curves for ICU survivors only run parallel to the general population around 2 --3 years after hospital discharge. However, marked differences occur depending on the diagnostic category . Medical patients requiring a period in ICU have an estimated hospital mortality of 23.2% compared with 14.1% for surgical patients . The difference between medical patient survival and surgical patient survival appears to be sustained beyond the ICU period. In one investigation, the median survival period for medical patients was reported as 40 days post-ICU admission compared with more than 900 days for other patients . The physical impact of critical illness : Patients often find themselves unable to accomplish even simple physical tasks without exhaustion after a period of critical illness. A variety of aspects of physical status may contribute to impaired physical functioning, and this may have an important impact on quality of life . Table 1 | Common physical problems following a period in the intensive care unit Nutrition | One invariable consequence of a period of critical illness is weight loss. Losses of 2% lean body mass per day have been reported, which can leave the patient with a formidable energy and protein deficit. Keys et al. demonstrated that even healthy volunteers required longer than 1 year to recover lean body mass following a period of prolonged starvation . Weakness, depression, breathlessness and altered taste perception are likely to reduce a patients' appetite and therefore to slow recovery of body mass. The inter-relationships between tissue loss and physiological function are complex and depend on the setting within which malnutrition arises . There are few published data regarding long-term nutritional follow-up of general ICU patients. Glutamine supplementation during the acute phase of illness appears to have a beneficial effect that is measurable at 6 months . In addition, a number of patients suffer mechanical difficulties in swallowing specifically related to healing tra-cheostomy sites . Neuropathy | Critical illness neuropathy is a well-recognized axonal neuropathy occurring in ICU patients . Although recovery can be anticipated within a few months in milder cases, more severe forms can have devastating consequences. In a recent 2-year follow-up of 19 patients suffering severe critical illness neuropathy (characterized by quadriplegia or quadraparesis) associated with sepsis and multiple organ failure, two patients died within 2 months, 11 patients recovered completely, four patients remained quadriplegic, and two patients remained quadraparetic . In that study, three parameters correlated with poor recovery: longer length of stay in the ICU, longer duration of sepsis, and greater weight loss . The cause of critical illness neuropathy remains obscure. Numerous factors have been implicated but, as yet, insufficient evidence is available to attribute cause , and no information is available regarding its prevention or treatment. Respiratory consequences | Breathlessness is a common symptom reported by ICU survivors . The possible causes of breathlessness include muscle weakness, neuropathy, pulmonary fibrosis, progression of premorbid respiratory and cardiac disease, and also psychological factors. One of the best-characterized ICU diseases is acute lung injury and its severe manifestation, the acute respiratory distress syndrome (ARDS). Follow-up studies have included physiology , radiology , quality of life , and psychological consequences . Davidson et al. suggested that ARDS did not increase a patient's risk of death beyond discharge from hospital ; the long-term mortality of ARDS patients correlated with age, underlying risk factors, and comorbidity. In another study, Davidson et al. used the Medical Outcomes Study Short Form (SF-36) instrument (a generic quality-of-life instrument) and the St George's Respiratory Questionnaires (a pulmonary-specific tool) . They assessed 73 ARDS patients matched with control patients having equal severity of disease or injury. Davidson et al. concluded that, 2 years after ICU admission, ARDS survivors had clinically significant reductions in health-related quality of life (HRQL) that appeared to be caused exclusively by ARDS and its sequelae. The reductions were primarily noted in physical functioning and pulmonary disease-specific domains . Continuing respiratory dysfunction thus appears to be associated with a broad impact on well-being, and this observation is unlikely to be specific to ARDS. Most ARDS survivors have abnormal pulmonary function tests in the first 3 months of recovery; this reaches a plateau over the following 3 --9 months . Pulmonary function tests following ARDS generally demonstrate a restrictive ven-tilatory defect and some impairment of diffusion capacity. Alteration of lung parenchymal compliance has been documented using whole body plethysmography and measurement of transpulmonary pressure . However, the number of patients followed up in this way is small and the significance of the residual restrictive deficit remains to be clarified. Reduction in diffusion capacity may be more significant in explaining reduced exercise tolerance in these individuals, some of whom desaturate during exertion. Persistent radiographic changes in survivors of ARDS have been reported. Desai et al. examined computerized tomography scans of 27 survivors of ARDS . Follow-up scans were performed 110 --267 days (mean +- standard deviation, 196.2 +- 41.3 days) after the initial scan performed during ICU admission. The most frequent persistent abnormality, in 23 patients (85%), was a course reticular pattern with an anterior distribution. The extent of this fibrotic appearance was up to 31% of the lung field. A ground glass appearance, which was taken to indicate either oedema or fine fibrosis below the resolution of the scanner, was observed in 17 survivors; the extent ranging up to 56%. The variation in extent of these changes may partly explain the variation in pulmonary function . The importance of these radiological changes to function remains unclear. Cardiac consequences | Although the cardiovascular system often requires support with inotropes and vasopressors, it is assumed that there are no long-term cardiac sequelae in the absence of infarction. Using a canine model of peritoneal sepsis, Cunnion and Parrillo found that cardiac function returned to normal within a few days of recovery from a septic episode . There are no documented adverse effects of general intensive care on cardiac function or on potential symptomatic problems in the post-ICU period but data are very scarce. Postural hypotension , possibly due to a residual autonomic neuropathy, can be problematic; the resulting dizziness decreases mobility and independence. Other physical problems | Reduced mobility in survivors of intensive care is multifactor-ial: there is reduction in muscle mass, muscle weakness, joint stiffness, and poor balance. These problems are generally recognized and the appropriate physiotherapy provided. Less obvious physical problems such as swallowing difficulties may go unnoticed. Weakness and lack of coordination of the pharyngeal musculature can initially be significant, requiring persistent effort by the patient to retrain the muscles. The presence of a tracheostomy physically impedes swallowing. A late complication of percutaneous tracheostomy, tethering of the skin to the trachea, can cause a distressing sensation when swallowing and may need surgical intervention . There are few long-term follow-up studies of percutaneous tracheostomy. Those available suggest a favourable outcome, with a very small incidence of significant complications . The results of long-term follow-up studies are awaited. Interestingly, in a comment relating to his experience of patients attending his follow-up clinic, Griffiths suggests that significant tracheal stenosis is more commonly associated with repeated intubations . Sexual dysfunction | Withdrawal from sexual intimacy can have detrimental effects on relationships and an individual's sense of well-being. Quinlan et al., reporting on the follow-up of 62 patients, found that 26% reported sexual dysfunction at 2 months, 19% at 6 months, and 16% at 1 year . The reasons given by patients for their poor sex life in that study and a subsequent study were having no desire, suffering impotence despite desire, shortness of breath, surgical disfigurement, and concern that sex might precipitate a relapse of illness . Neuropsychological outcomes : The ICU is a stressful and noisy environment with little differentiation between night and day . Patients are often sedated with psychoactive drugs during at least part of their stay, the long-term effects of which are uncertain. Approximately one-third of the patients attending our own follow-up clinic complains of impaired memory, poor concentration, and/or mental slowing 3 months after ICU discharge (unpublished observation). How these symptoms relate to ICU stay is currently uncertain . Table 2 | Common psychosocial problems following a period in the intensive care unit Delirium | The true incidence of delirium in the ICU is unknown. Various terms have been used in the literature to describe this condition, including 'intensive care syndrome' and ICU psychosis. The lack of an evaluative tool that could be used by ICU staff to diagnose the condition in patients who are unable to communicate has hampered the collection of useful data. Estimates reported in the literature suggest that the incidence of delirium within surgical intensive care patients is 40% ; the incidence may be as high as 60% in older patients . Our knowledge of the long-term effects of delirium is extremely limited. The diagnosis of delirium, in all but the most obvious cases, may not be straightforward; apparent quiet withdrawal may be a manifestation as much as agitated confusion. Ely and colleagues have recently developed and validated a tool for the bedside assessment of delirium in ICU patients , which should prove valuable in future long-term studies. Delirium is accompanied with profound amnesia of events occurring during and preceding the confusional state. The amnesia results in patients' memories of their stay in the ICU being fragmented and frequently distorted. Slow wave sleep is reduced in patients in the ICU , which is known to be important for the integration of factual memories . Patients may enter a hypnagogic state during treatment in the ICU, in which control of the boundary between internally generated fantasy and the experience and recognition of external reality is impaired . This state between sleep and wakeful-ness predisposes to hallucinations and creates a mental environment favouring the development of paranoid delusions. During the immediate post-ICU period, the patient begins to become aware of the physical changes to their body with little conscious awareness of what has brought them to this state . They may have only vague memories of pain and uncomfortable procedures . Nightmares, dreams and hallucinations have all been described during recovery . The recall of delusional memories may predominate and there is some evidence that delusional memories, without factual recall, are highly associated with the early development of stress disorders in ICU patients . The lack of memory of the illness producing physical weakness appears to be a cause of subsequent frustration. Affective disorders | Symptoms of anxiety and depression are common. In a postal questionnaire study of general intensive care survivors, Scragg et al. reported that 38 out of 80 respondents (47%) reported clinically significant anxiety and depression as measured by the Hospital Anxiety and Depression scale . Nelson et al. investigated depression in survivors (median period, 15 months) after intensive care for acute lung injury. Sixty-nine percent of patients without a pre-existing history of depression had clinically relevant depressive symptoms , which correlated with the days of sedation and the use of neuromuscular blockade. In a study of a more varied case mix of ICU patients at 3 month follow-up, Eddleston et al. found a lower prevalence of psychological distress (Hospital Anxiety and Depression scale score >8): anxiety, 11.9%; depression, 9.8% . Jones et al. reported clinically significant Hospital Anxiety and Depression scale scores in ICU patients; notably, those who could not recall any factual memory of events during their ICU stay but who retained delusional thoughts . Stress disorders | Post-traumatic stress disorder (PTSD) is a condition that occurs in individuals who have experienced a traumatic event. In this condition, a 'traumatic' event is characterized by its capacity to provoke fear, helplessness, or horror in response to the threat of injury or death . Symptoms include distressing 'flash backs' (re-experiencing of the event), avoidance of situations reminding the individual of the event, increased arousal, and psychological numbing. The incidence of PTSD in patients following an ICU stay varies; patients who have survived ARDS and younger patients appearing to be particularly vulnerable . Schelling and colleagues found that memories of adverse experiences correlated with subsequent PTSD symptoms, and were also related to poor general quality of life scores . By contrast, Jones et al. found that factual memories seemed in some way protective and that delusional memories without factual recall was a situation more associated with early stress symptoms at 8 weeks. The Schelling et al. study did not attempt to differentiate factual from delusional memories and was undertaken years after discharge. These two important studies are thus complementary rather than contradictory. In Scragg et al.'s postal study , 30 survivors (38%) reported symptoms of PTSD, as assessed by the Trauma Symptoms Checklist 33 and the Impact of Events Scale . Using the Experience after Treatment in Intensive Care 7 Item Scale, an additional questionnaire designed specifically for their study, Scragg et al. were able to show that intensive care treatment contributed to the causation of post-traumatic stress . Patients may experience symptoms of an acute stress disorder during the early phase of their recovery. This condition is a recent diagnostic category, being defined for the first time in the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders in 1999 . Acute stress disorder is characterized by the early development of symptoms similar to PTSD, with the inclusion of dissociative symptoms such as detachment, depersonalization, derealization, and dissociative amnesia. Although the description of this condition as a distinct disorder from severe acute PTSD has been questioned , its recognition allows early psychological intervention to facilitate a patient's resolution of the traumatic experience . Disorders of cognitive function | In a landmark study of ARDS survivors, Hopkins and colleagues reported that 100% of survivors experienced cognitive impairment at hospital discharge , including problems of memory, attention and concentration. Although they demonstrated improvements at 1 year follow-up, 30% of patients were still globally impaired and 78% were impaired in one or all of the domains assessed. Hopkins et al. associated the deficits with hypoxaemic episodes periods in the ICU. This added to earlier work, which had demonstrated the relation between hypoxic brain injury, hippocampal atrophy, and memory deficit . Rothenhausler et al. investigated cognitive performance in ARDS patients at a median time of 6 years after ICU discharge . These investigators found that 75% of ARDS survivors did not display signs of cognitive impairment. Of the remaining 25%, the impairment was either subthreshold or mild in the vast majority. The residual mild cognitive impairments demonstrated were found in subtests relating to attention. Although the impairments were described as mild, all patients with cognitive dysfunction were classified as disabled, compared with 22.9% of those without cognitive impairment. The investigators also found that 58.7% of survivors returned to work. These figures, similar to those published by McHugh et al. (56%) and Schelling et al. (61.3%) , indicate a significant critical care achievement. The corollary of this, however, is that there remained a significant proportion of survivors whose employment status was significantly impaired, implying a negative social outcome. Social and family consequences | During the patient's stay in intensive care the family members, and especially the spouse, have been shown to exhibit significant degrees of anxiety and depression . After the patient has been discharged from hospital this anxiety and depression may manifest itself in overprotective behaviour that, coupled with unrealistic expectations that the patient may have of their capabilities, leads to frustration and conflict . Survivors may experience guilt as they become aware of the strain placed on their family, which adds to their psychological burden predisposing them to affective disorders . The Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment investigators collected data to assess the impact of critical illness on the family . They reported that 34% of patients required considerable care-giving assistance. A family member needed to give up work or make major life changes to provide care for their relative in 20% of the total. It is possible that multidisciplinary intensive care follow-up needs to address the needs of the family during the rehabilitation phase . Quality of life | People with the same health status may not necessarily enjoy the same quality of life. Health and well-being are multidimensional, and they include aspects such as individuals' subjective values and preferences that make the direct measurement of quality of life difficult . Item measurement theory proposes that true quality of life can be measured indirectly via a series of questions (items), each of which is a measure of the same concept or construct. The answers are assigned numerical scores that may be combined to provide 'scale' scores (e.g. physical functioning, symptom, distress, or well-being scales). These scores may then be further combined to give 'domain' scores (e.g. physical, psychological, work performance). HRQL measures potentially provide a holistic measure of the impact of critical illness on the patient, a measure of the contribution of disease-specific changes in health to overall patient well-being, and a means to evaluate the health economic impact of intensive care medicine (e.g. the Quality Adjusted Life Year estimates). Kutsogiannis and Noseworthy have comprehensively reviewed the available HRQL instruments , highlighting the complexity of creating and using HRQL instruments appropriately. By way of a caution, Ridley has recommended that the psychometric properties of questionnaires, which may not have been designed specifically for critically ill patients, should be tested and validated in critically ill patients before the results are presented . Mata et al. have published the results of a quality-of-life questionnaire-based study. The questionnaire was completed by patients or a proxy at the time of admission to the ICU, and was compared with the results of the same questionnaire completed by the survivors 12 months later (444 of the initial 606 patients). The quality of life was reduced in 50% of survivors, remained the same in 23% and improved in 27%. A decreased quality of life was most evident in those older than 75 years of age. The pre-ICU quality of life and age were the factors with the greatest influence on post-ICU quality of life . This original seven-item Quality of Life -- Spanish questionnaire has since been modified to create a 15-item quality-of-life questionnaire specifically for critically ill patients . Konopad et al. used Spitzer's Quality of Life Index, which examines the activity level, the activities of daily living, health, support, and outlook . Of the original group of 504 general ICU patients who completed a baseline questionnaire at ICU admission, 293 completed a 12-month follow-up questionnaire. A significant decrease in the level of activity score and the activities of daily living score occurred at 12 months, and these scores were more significant in the older age group . Using their own quality-of-life questionnaire administered to 238 patients 16 months after discharge, Brooks et al. found that survivors have more health problems, are more anxiously depressed, are more dependent on others, and engage in less sexual activity than a random community sample . A number of investigators have used the SF-36 instrument , which consists of 36 questions covering eight domains and has been validated in the intensive care setting . Compared with age-matched and gender-matched controls, Schelling et al. report that survivors of ARDS, at a median follow-up interval of 4 years, experienced a 25% reduction in physical function and described their general health as 15% lower. Mental health dimensions of the SF-36 were also significantly impaired, although to a lesser degree (4% reduction in mental health, 15% reduction in vitality, 10% reduction in social function) . In that and a more recent study by Eddleston et al., which examined a broader ICU case mix, the SF-36 scores of younger survivors were found to be worse; particularly in physical and emotional role limitation, and in social functioning . A report by Flaatten and Kvale on survivors 12 years after ICU treatment described a similar global reduction in SF-36 scale scores, with reductions in six of the eight scale scores . The pattern and scores of the SF-36 in these studies were similar to those described by Ridley et al. , who were also able to show that the quality of life of patients admitted to their ICU was not the same as the normal population. Patients who enjoyed a normal quality of life before admission suffered significant decreases following their illness. Those with pre-existing morbidity showed some improvement in their quality of life 6 months after intensive care admission, principally by improvements in mental health, vitality and social functioning. Final remarks : What have ICU follow-up studies shown us overall? Survival curves do not run parallel to those of the average population until 2 --3 years post-ICU discharge. 'Survival', however, is a poor metric for describing the impact of critical illness. We as yet have an inadequate understanding of how events in the ICU impact on recovery. The physical and psychological consequences can be significant and unanticipated. The neuropsychological picture is beginning to emerge; a subject that warrants much further study. The burden carried by families or future caregivers is potentially great and, as yet, we have little data concerning the fiscal impact of ICU survival. Increasing knowledge in these areas will provide us with the potential to modify our approach inside the ICU to improve patient-centred outcomes in the long term. Competing interests : None declared Abbreviations : ARDS = acute respiratory distress syndrome; HRQL = health-related quality of life; ICU = intensive care unit; PTSD = post-traumatic stress disorder; SF-36 = Medical Outcomes Study Short Form (SF-36) instrument. Figure 1 | Factors impacting on recovery from the intensive care unit. Factors impacting on recovery from the intensive care unit. Backmatter: PMID- 12398778 TI - ICUs worldwide: Results of a nationwide survey of Colombian intensive care units AB - Abstract | Sixty-three of 89 identified intensive care units in Colombia (Evaluation of Intensive Care in Colombia) participated in this voluntary study. A convenience sample of 20 intensive care units, each submitting 200 patients or more, was chosen, from which the following information is presented. The Intensive Care National Audit and Research Center (UK) protocol was used to evaluate patient severity, length of stay, raw and anticipated mortality, intensive care unit patient admission/rejection criteria, and human and technologic resources available. Information was drawn from public and private institutions. Keywords: admissions, Columbia, intensive care, mortality, resources, admissions, Columbia, intensive care, mortality, resources, Introduction : Colombia's first intensive care unit (ICU) was established in the early 1960s at San Juan de Dios Hospital in Bogota. This was soon followed by additional facilities at the Shaio Clinic, Military Hospital, San Jose Hospital and Caja Nacional de Prevision. Similar to the international community, Colombian critical care has expanded to meet the needs of improving the increasingly complex care mandated by improvements in health care delivery, medical technology and vulnerable populations at the extremes of age. The Clinical Epidemiology and Bio-Statistics Unit of the Javeriana Medical School and the Colombian Critical Medicine and Intensive Care Association collaborated in 1997 to establish an Intensive Care Unit/System Delivery evaluation program in parallel with government and private sector efforts to expand medical coverage nationwide . Evaluation of Intensive Care in Colombia was funded through the following organizations: Pontificia Universidad Javeriana, Colciencias, Colombian Ministry of Health; International Development Department, United Kingdom; Rockefeller Foundation International Clinical Epidemiology Network; and the Colombian Association of Critical medicine and Intensive Care. The study was modeled on the Intensive Care National Audit and research Center (UK) questionnaire to obtain reliable information for comparison with external benchmarks . Study objectives : The study was instituted to prospectively analyze information from the following areas : the description of human and technologic resources available in the ICU; ICU demographics, to include severity of illness and case mix; the description of costs and resource consumption in the ICU; the description of staffing patterns, patient care protocols and unit direction; the comparison of the expected and true mortality based on admission severity across selected units; the contrast of Colombian ICUs according to unit management and care processes and mortality; and the comparison of Colombian and English experiences. Identification of participating ICUs : Eighty-nine ICUs were identified in the country, and 63 units participated voluntarily to develop the present study. Twenty ICUs were chosen after evaluation of data submitted from the original questionnaire. Ten ICUs were located in Bogota and the remaining 10 were located throughout the rest of the country . Acute Physiologic and Chronic Health Evaluation (APACHE) II and APACHE III, Simplified Acute Physiology Score (SAPS) II, Mortality Probability Model (MPM) II0 and MPM II24 methods were utilized to evaluate the severity of illness for different patient populations and to calculate the expected mortality . The structure of critical care service delivery in Colombia was categorized in terms of human and technologic resources and outcome, in both public sector and private sector facilities. Results were compared with National Standards for Intensive Care promulgated by the Ministry of Health (Colombia). Factors leading to patient refusal for ICU admission were researched. Results : Initial information was obtained from 63 of the 89 ICUs contacted (72% response) . The public hospital ICU length of stay was statistically longer at 6.1 days than 4.3 days in the private sector in 1996 (P = 0.05). The public versus private maximum average length of stay was 44 days versus 30 days (P = 0.04). Private hospital ICUs admitted more patients per month (P = 0.02) and had a higher bed turnover (P = 0.03) than those in the public sector. No significant differences were found in technology available in all areas. When compared with nationally approved standards, however, the number of available beds consistently exceeded recommended technologic resources. Of the public hospital ICUs, 42% reported that patients requiring ICU services were refused admission one to 10 times per week, versus 14% in the private sector (P = 0.0049). While the lack of available beds was the most frequent cause in over 85% of the ICUs in both sectors, the lack of available nurses was four times more frequently the cause for admission refusal in the public institutions. All ICUs reported the day and night shifts worked by medical staff and nurses. The equivalent average full-time number of nurses for each ICU was four, and no difference was noted between the public and private ICUs surveyed. Only 25% of the nurses stated that they had received some degree of intensive care training. More than 90% of all ICUs reported the presence of a full-time Medical Director: 64% were internists and 27% were anesthesiologists, and 5% reported specialized training/certification in intensive care. Although medical specialists (internal medicine, anesthesiology) and residents in training (internal medicine, anesthesiology, surgery) cover medical care at night in some areas, over 50% of ICUs reported that all care was performed/covered by general practitioners without formal critical care training. This was more frequent in the public sector than in the private sector (73% versus 39%, P = 0.034). There is a significant deficit with respect to available human resources and technologic support in Colombian ICUs. To upgrade the units to comply with published national standards would cost approximately US$10 million for the 63 ICUs from which accurate data is available. Extrapolating these projections, under the assumption that no significant differences would be found in resources, organizational structure and staffing patterns, to the remaining 26 ICUs results in additional costs of US$5 million for the first year. There are 12,987 hospital beds and 460 ICU (3.5%) beds in the reporting hospitals. If it is assumed that 5% (ideally it would be 10%) of available beds will be available for critical care services, there would be an additional deficit of 320 ICU beds (1997 survey data), with requisite staffing and technologic support consuming additional funding. Fixed costs were higher than variable costs in all reporting ICUs and represented 82 --87% of the total costs associated with patient care. Direct variable costs were distributed as follows: medication, 45 --63%; blood bank, 19 --30%; and nutrition, 6 --21%. Clinical laboratories and radiology services required significantly lower budgets. Further analysis of medication costs revealed that sedatives accounted for 30 --50%, antibiotics accounted for 26 --41%, and inotropes and other medications accounted for the remainder. Significant mortality differences in the reporting ICUs were noted utilizing the APACHE II, SAPS II, MPM II0 and MPMII24 methodologies. These data were reported to the directors of the respective units. The observed mortality/expected mortality ratio with 95% confidence limits is shown in Fig. . Figure 1 | Number of deaths observed versus number of expected deaths using the APACHE II method from 22 July 1997 to 2 October 1998 (2615 patients). Number of deaths observed versus number of expected deaths using the APACHE II method from 22 July 1997 to 2 October 1998 (2615 patients). Data taken from . Conclusions : There is a marked difference between public sector and private sector ICUs in Colombia. Private ICUs normally show better results than do those of the public sector. The differences reside primarily in the quantity and training of the personnel, and in the availability of technology. Lack of formal training was more frequently seen in physicians and nursing staff working in the public sector. Two out of four ICUs with high mortality rates did not have invasive monitoring technology. There seems to be a clear correlation between the lack of specialists and trained nursing personnel and the poor results that public ICUs present . Improvements in human resources and technologic support are needed in public sector and private sector facilities. In comparison with English patients, Colombian ICU patients are younger and represent a lower severity of illness. Competing interests : None declared. Abbreviations : APACHE = Acute Physiologic and Chronic Health Evaluation; ICU = intensive care unit; MPM = Mortality Probability Model; SAPS = Simplified Acute Physiology Score. Backmatter: PMID- 12398773 TI - Is insulin an endogenous cardioprotector? AB - Abstract | Stress hyperglycemia and diabetes mellitus with myocardial infarction are associated with increased risk for in-hospital mortality, congestive heart failure, or cardiogenic shock. Hyperglycemia triggers free radical generation and suppresses endothelial nitric oxide generation, and thus initiates and perpetuates inflammation. Conversely, insulin suppresses production of tumor necrosis factor-alpha and free radicals, enhances endothelial nitric oxide generation, and improves myocardial function. It is proposed that the balance between insulin and plasma glucose levels is critical to recovery and/or complications that occur following acute myocardial infarction and in the critically ill. Adequate attention should be given to maintaining euglycemia (plasma glucose <= 110 mg/dl) in order to reduce infarct size and improve cardiac function while using a glucose --insulin --potassium cocktail. Keywords: cardiac failure, cardioprotection, diabetes, free radicals, glucose, hyperglycemia, insulin, nitric oxide, tumor necrosis factor, septicemia, septic shock, cardiac failure, cardioprotection, diabetes, free radicals, glucose, hyperglycemia, insulin, nitric oxide, tumor necrosis factor, septicemia, septic shock, Introduction : Patients with acute myocardial infarction (AMI) exhibit raised blood glucose concentrations . In addition, a positive association between hyperglycemia and mortality from AMI has been reported , although the exact reason for this association is not clear. Intensive treatment with insulin to lower plasma glucose concentrations decreases overall mortality in patients with diabetes mellitus who have AMI. In a prospective, randomized, controlled study involving adults admitted to surgical intensive care units and receiving mechanical ventilation , intensive insulin treatment reduced mortality and morbidity. Intensive insulin treatment reduced the number of deaths from multiple organ failure with sepsis. Markers of inflammation were found to be abnormal less frequently in the intensive insulin treatment group. This suggests that hyperglycemia is harmful, whereas insulin therapy is beneficial not only in AMI but also in critical illness with or without diabetes mellitus. It is likely that lack of insulin associated with hyperglycemia causes a decrease in glycolytic substrate and an increase in free fatty acids. This induces a reduction in myocardial contractility, and promotes cardiac failure and arrhythmias , leading to poor outcomes in such patients. Hyperglycemia is proinflammatory whereas insulin is anti-inflammatory : Capes and coworkers showed that patients with stress hyperglycemia but without diabetes mellitus at the time of AMI are at increased risk for in-hospital mortality and congestive heart failure or cardiogenic shock. Although the exact cause for the poor prognosis is not clear, it was suggested that hyperglycemia (an indirect reflection of relative insulin deficiency) increases circulating free fatty acids, which are toxic to myocardium and induce arrhythmias . Hyperglycemia causes osmotic diuresis, and the resulting volume depletion may further compromise myocardial function. Both in animal models of diabetes and in patients with diabetes mellitus, increased production of reactive oxygen species and consequent lipid peroxidation were noted . Hyperglycemia increases the production of reactive oxygen species inside cultured aortic endothelial cells . Superoxide anion inactivates both endothelial nitric oxide (NO) and prostacyclin produced by endothelial cells, which are potent vasodilators and platelet antiaggregators . Thus, free radicals induce endothelial dysfunction. Normalizing levels of mitochondrial reactive oxygen species was reported to prevent glucose-induced activation of protein kinase C, formation of advanced glycation end-products, sorbitol accumulation, and nuclear factor-kappaB (NF-kappaB) activation . Glucose challenge stimulated reactive oxygen species generation and levels of p47phox (a key protein of the enzyme nicotinamide adenine dinucleotide phosphate [reduced; NADPH] oxidase), whereas alpha-tocopherol levels decreased significantly in polymorphonuclear leukocytes and monocytes, even in normal subjects . High glucose concentrations induced inflammatory events in rats, as evidenced by increased leukocyte rolling, leukocyte adherence, leukocyte transmigration through mesenteric venules associated with attenuation of endothelial NO release, and increased expression of P-selectin on endotheial surfaces . Local application of insulin attenuated these proinflammatory effects. Insulin infusion inhibited reactive oxygen species generation, p47phox and NF-kappaB in mononuclear cells, and reduced soluble intercellular adhesion molecule-1, monocyte chemoattractant protein-1 and plasminogen activator inhibitor-1 production by increasing NO synthesis . These findings suggest that hyperglycemia has proinflammatory whereas insulin exhibits anti-inflammatory actions. The exact mechanism by which glucose stimulates proinflammatory events is not clear, although indirect evidence suggests that it does so possibly by stimulating production of tumor necrosis factor (TNF)-alpha (a proinflammatory cytokine). A diet with a high glycemic load and hyperglycemia induced production of acute-phase reactants . In experimental animal models of diabetes, the activity of NADPH-dependent oxidase and the levels of NADPH oxidase protein subunits p22phox, p67phox and p47phox were significantly increased , which accounted for the increased superoxide production in addition to decreased endothelial NO synthase activity. Similar to glucose, TNF-alpha also enhances free radical generation by augmenting polymorphonuclear leukocyte NADPH oxidase activity, activates NF-kappaB, and increases intercellular adhesion molecule-1 expression in endothelial cells . This similarity in the actions of glucose and TNF-alpha, and the ability of former to enhance acute phase reactants suggests, but does not prove, that glucose may enhance TNF-alpha production and brings about its proinflammatory actions. Tumor necrosis factor-alpha and myocardium : TNF-alpha is secreted by adipose tissue, macrophages and cardiac tissue, and plays roles in the pathogeneses of insulin resistance, type 2 diabetes mellitus, inflammation, and septic shock . Release of TNF-alpha occurs early in the course of AMI and reduces myocardial contractility in a dose-dependent manner . Using anti-TNF-alpha antibody can reduce TNF-alpha-induced myocardial injury and dysfunction . Cardiac cachexia is believed to be due to an increase in the circulating levels of TNF-alpha , and a direct correlation between the clinical features of congestive cardiac failure (CCF) and circulating levels of TNF have been reported. Following cardiac transplantation TNF-alpha levels decrease . This suggests that TNF-alpha is an important mediator in the pathogenesis of CCF. In addition, it causes dysfunction and apoptosis of endothelial cells, and enhances generation of free radicals (including superoxide anion), which in turn quenches NO. Damage to endothelial cells triggers procoagulant activity and fibrin deposition . These events are detrimental to the patient in the long run. In CCF there is increased mesenteric venous pressure, which causes intestinal edema and increased bowel permeability. This causes an increase in endotoxin absorption from the gut. Increase in circulating levels of endotoxin activates macrophages and other cells to produce TNF-alpha . In patients with CCF, CD14 concentrations (which are indicative of endotoxin-cell interaction) are raised in relation to the elevated levels of TNF-alpha and cachexia . These findings suggest that methods designed to reduce TNF-alpha levels could be of significant benefit in inflammation, septicemia, and CCF. Tumor necrosis factor-alpha and insulin : Both the American College of Cardiology and the American Heart Association recommended that intravenous glucose --insulin --potassium (GIK) be given to patients with AMI, especially those who are poor candidates for thrombolytic therapy and in whom the risk for bleeding is high , because the GIK regimen was beneficial in treating AMI . It is generally believed that the GIK treatment improves the integrity and function of myocardial cells once glucose and potassium are transported in by insulin. Previously, I suggested that the GIK regimen in general and insulin in particular suppresses inflammation by inhibiting production of TNF-alpha, macrophage migration inhibitory factor (MIF) and superoxide anion, and by stimulating endothelial NO synthesis . Satomi and coworkers showed that exogenous insulin injection inhibited TNF-alpha production in a dose-related manner in animals after lipopolysaccharide challenge. Addition of insulin to cultures of peritoneal exudate cells from Propionibacterium acnes primed mice blocked TNF-alpha production, whereas in control animals it did not. Fraker and colleagues reported that reduced food intake, decreased body weight gain, severe interstitial pneumonitis, periportal inflammation in the liver, and increases in the weights of the heart, lungs, kidney and spleen observed in TNF-alpha-treated animals reverted to normal levels when insulin was administered concurrently. The pneumonitis seen in these TNF-alpha-treated animals is somewhat similar to the adult respiratory distress syndrome that is seen in patients with septicemia and septic shock, conditions in which concentrations of interleukin-1, TNF-alpha, and MIF are elevated . In addition, insulin suppresses superoxide anion generation and enhances the production of endothelial NO . Thus, the ability of insulin to suppress TNF-alpha production, which decreases myocardial contractility, could be one mechanism by which the GIK regimen is beneficial in AMI. Is it glucose or insulin that is critical to the heart? : Although several studies suggested that GIK regimen preserved systolic and diastolic function in ischemia and reperfusion and protects the myocardium in patients undergoing open heart surgery , this is not without controversy . Why did some studies give positive results whereas others failed to show a benefit from the GIK regimen? On closer examination, it is clear that not all studies were comparable to each other because the concentrations of glucose and insulin used in those studies were not uniform . Studies in which higher concentrations of insulin were used showed better results than did those studies that employed a lesser dose. For instance, studies in which 33% glucose with 120 units of insulin or 30% glucose with 300 units of insulin was used yielded positive results. In contrast, the results reported by those studies that employed a lower dose (Bruemmer-Smith and coworkers used 500 ml of 5% dextrose with 100 units of insulin, and Rao and colleagues supplemented the cardioplegic solution with 10 units/l insulin) were less favorable. This is supported by the observation that stress hyperglycemia or even mild hyperglycemia with myocardial infarction is associated with increased mortality and that intensive insulin treatment to maintain blood glucose levels between 80 and 110 mg/dl is highly beneficial and reduces morbidity and mortality among critically ill patients . It is possible that the negative results obtained with GIK were due to the low dose of insulin used; this invariably resulted in hyperglycemia, which is detrimental to the myocardium. It has been known for several years that continuous intravenous infusion of insulin is superior to subcutaneous administration in terms of glycemic control, especially in patients with diabetes during the perioperative and postoperative periods . During both the infusion period and the entire observation period (day of operation, and first and second postoperative days), GIK regimen resulted in lower blood glucose levels within the intended range of 90 --180 mg/dl (5 --10 mmol/l) as compared with conventional subcutaneous insulin administration. Improved diabetic control is believed to result in fewer wound infections and better wound healing. However, this view may be too simplistic. The beneficial effects of GIK regimen may extend beyond control of hyperglycemia alone . As demonstrated recently , GIK infusion may salvage myocardium, improve cardiac function, and decrease mortality by an absolute 10%, provided that hyperglycemia is prevented. There is reasonable evidence to suggest that this beneficial effect may be independent of glucose . These results are supplemented by those of a large trial conducted in a heterogeneous group of 1548 critically ill patients . In that trial, intensive insulin therapy to avoid hyperglycemia (blood glucose was maintained below 110 mg/dl) in predominantly nondiabetic patients led to a decrease in morbidity and mortality as compared with less intensively treated patients (blood glucose maintained between 180 and 200 mg/dl). Those findings suggest that maintaining blood glucose concentrations at 110 mg/dl or less is critical in obtaining the benefits of insulin administration. This is supported by the observation that cardiac dysfunction induced by endotoxin administration was not related to arterial blood glucose concentrations . Furthermore, infusions of insulin reversed cardiac failure and maintained normal performance in spite of wide ranges in glucose concentrations (5 --120 mg/dl), suggesting that myocardial dysfunction is not precipitated or induced by the hypoglycemia of endotoxin shock. The ability of insulin to improve myocardial performance may be related to its capacity to suppress TNF-alpha, MIF, and superoxide anion generation . Therapeutic administration of high doses of insulin results in an accumulation of myocardial glycogen stores and improvement in glucose utilization. This leads to augmented myocardial adenosine triphosphate provision and maintains cellular energy charge during coronary ischemia, resulting in better tolerance to ischemia and improved myocardial protection . Conclusion : It is evident from the preceding discussion that hyperglycemia is harmful whereas insulin treatment is beneficial. Even mild hyperglycemia is associated with poor neurologic outcome after brain injury and stroke , and burns or surgery in humans . Animal studies revealed that hyperglycemia aggravates endotoxin shock and that insulin treatment decreases mortality . What are the potential mechanisms by which insulin is able to bring about its beneficial actions? Apart from its ability to lower blood glucose and to inhibit production of potentially dangerous proinflammatory cytokines (i.e. TNF-alpha, MIF, and superoxide anion), insulin has the following actions: it stimulates glucose uptake/glycolysis, pyruvate dehydrogenase and energy production; it increases muscle protein synthesis; it inhibits apoptosis and improves repair of damaged tissues; it promotes ischemic preconditioning and lessens ischemia/reperfusion damage (for review ); and it exhibits anti-inflammatory actions . Because hyperglycemia induces apoptosis of myocardial cells , strict control of blood glucose is essential to preserve cardiac function both in diabetic and nondiabetic persons with stress hyperglycemia. The ability of insulin to enhance endothelial NO synthesis is particularly significant when one considers its beneficial action in AMI, stroke, and critical illness . Recent studies suggested that administration of L-arginine (the precursor of NO) improves postischemic recovery of endothelial and vascular smooth muscle functions after cold cardioplegic arrest, and enhances cardioprotection and postischemic functional recovery and reduces infarct size of the myocardium. Hence, some of the beneficial actions of insulin (and therefore those of the GIK regimen) in various conditions could be attributable to an increase in endothelial NO synthesis . In summary, GIK regimen is useful in preserving the myocardium in septicemia and septic shock, and in patients with severe burn injury , provided that blood glucose levels are maintained at 110 mg/dl or below by employing an adequate insulin dose. Thus, insulin when present in appropriate amounts preserves myocardial integrity and function. Competing interests : None declared. Abbreviations : AMI = acute myocardial infarction; CCF = congestive cardiac failure; GIK = glucose-insulin-potassium; MIF = macrophage migration inhibitory factor; NADPH = nicotinamide adenine dinucleotide phosphate (reduced); NF-kappaB = nuclear factor-kappaB; NO = nitric oxide; TNF = tumor necrosis factor. Backmatter: PMID- 12398777 TI - ICU Cornestone: A lecture that changed my practice AB - Abstract | In 1982, the author attended a lecture by Professor Joseph Civetta dealing with the concept that, at times, the goal of care should be comfort rather than cure, and that inappropriate care prolonged dying and suffering. Efforts to improve end-of-life care subsequent to this had effects on care at a local level and at a state level. Intensive care providers should be leaders in the provision of appropriate and compassionate care at the end of life. Keywords: death, terminal care, death, terminal care, Introduction : It was in 1972 that I was first rostered to work in an intensive care unit. From the first day, I never wanted to work anywhere else. The list of changes and improvements is enormous, but one particular event and its consequence, more than any other, changed both my practice and my life. The event was a lecture by Professor Joe Civetta in 1982 in Sydney. In 1980, I heard Joe speak in San Antonio about ways of using nurses' time better. I had just begun running a 'Continuing Education Meeting' in Sydney that was to focus on things I believed we had not thought about enough. I invited Joe to speak at this meeting in 1981, and he informed me that the data and work was that of his wife, Judy. She agreed to attend. Joe also wanted to come. I asked what he would like to talk about and he said he had a lecture on 'Stress, Death and Dying'. When I queried why anyone would want to listen to such a lecture, Joe responded by suggesting that if 1% of my patients developed renal failure and 10% died, then maybe I should know as much about dying as I knew about renal failure. Joe's lecture in March 1982 totally changed my practice and that of many others who attended. The message was simple, and is, in a small way, discussed in one of Joe's papers : * Intensive care leads to situations where continuing treatment prolongs dying and suffering without producing survival. * To determine when this point is reached involves consideration of objective data and emotive aspects such as patient wishes. * The patient's wishes could sometimes be ascertained from speaking to the patient's family. * There is a need to create an environment in which both health care providers and consumers could discuss the appropriateness of treatment and a peaceful death could occur when indicated. * There comes a time when the goals of good medical care should be comfort rather than cure. The concepts are not as dramatic now as they were then, because today there is a virtual industry surrounding the dying patient and the patient's family. Butwe were taken with Joe's concepts. Withdrawal of care was unusual in our unit at that time, and when it did occur it was often covert. Trying to improve the end-of-life care of our patients moved us rapidly to a situation where only 9% of our patients died when we were still trying to prevent dying with all our resources, and 12.2% became brain dead. In the remainder of the patients, treatment was withdrawn or withheld. We began to talk to patients and families about appropriateness of treatment and about withdrawing and withholding treatment. After reading Majorie Sternberg's paper 'The responsible powerless' , we instituted regular meetings with the nursing staff at which the patients and the appropriateness of their care could be discussed in more depth than on ward rounds. We developed a corporate philosophy that we called 'The same team', as the goals of the providers and recipients of care were usually the same. We encouraged families to participate in planning and delivery of care. 'The same team' also meant dressing the same, so the white coats disappeared. Our mantra was taken from a paper by Dunstan : "You should not judge. The success of intensive care is not to be measured only by the statistics of survival, as though each death were a medical failure. It is to be measured by the quality of lives preserved or restored, the quality of the dying of those in whose interest it is to die and by the quality of relationships involved in each death". The process involved acquiring new skills in communication, particularly in listening. It required learning to understand others' ethnic, cultural, and religious beliefs. We learned that the deeper and closer relationships with patients and families meant their hurt and grief was shared. We learned that it was okay to show your feelings to others. The process improved the relationships between all staff as well as between staff and patients. We began to attract the interest of our colleagues and were asked to see patients in the ward to assist with end-of-life care. The hospital changed from a 'Blue Card' system of identifying patients who were not for resuscitation to requiring a proper 'Do Not Resuscitate' order, which stated what treatment was to be withheld, what treatment was to be continued, why the decision had been reached, who was involved in making the decision, and why the patient was not involved (if they were not involved). Our hospital made 'Do Not Resuscitate' orders a quality issue: 80% of patients who die had such an order when the notes where last surveyed. We began to be asked to present in other hospitals, and to lecture our new interns on dying, a subject we learned was not dealt with in undergraduate training. Eventually, with some trepidation, we produced three papers describing our activities, including the policy of not providing treatment that would not influence outcome at the request of patients or families . The papers were taken up by the national press: 98% of the feedback was positive. With the help of Joe Civetta's Withdrawing and Withholding Care Policy from Jackson Memorial Hospital in Florida, we developed a hospital policy on Withdrawing and Withholding Care. This policy went to the Health Department, who referred it to the Legal Section. Their response to the question 'Could somebody acting in accordance with this policy be charged with murder?' was affirmative. Laws were drafted, and were discussed at a public meeting involving a diverse range of groups, from The Right to Life to The Cryonic Preservation Society. It was a meeting I regard as very important and significant. All groups were unanimous in believing that end-of-life care was an inappropriate matter for laws, courts, and lawyers. Such matters should be resolved by doctors, patients, and families. The morass of case law regarding the ethics and practicalities of end-of-life care in the USA convinces me that this was a very mature attitude for a representative group to take. The public of New South Wales wanted guidelines, not laws. One member of the profession at the meeting put the question regarding a charge of murder to the legal expert in a different manner: 'Do you think it is likely that a doctor withdrawing life support from a patient in New South Wales would be prosecuted for murder?' 'If prosecuted, is it likely that a conviction would be entered?' 'If convicted, is it likely that other than a derisory sentence would be given?' The answer to all three questions was 'No'. The New South Wales Health Department Interim Guidelines for Withdrawing and Withholding Care were released in 1992. I suspect they were made 'interim' because of political fear that they may have caused controversy and cost votes. The reverse occurred. There were no dissenting voices from either ends of the spectrum. In two cases, the State Coroner accepted the guidelines as an appropriate standard of behaviour. The Health Department sought public comment. The final version, incorporating those comments, is due before the end of 2002. The increased awareness of consumers, the diversity of families, and the confidence-breaking patients who threaten our ability to prognosticate by surviving well, against impossible odds, tend to make these processes more complicated today than they were when we first set out on this path. But there is no doubt that it made our intensive care unit a better and fairer place to work, and almost certainly made those of us who walked the path better doctors and nurses, and better human beings. We are constantly impressed by the wisdom and dignity of Australians from all ethnic, religious, cultural, and social backgrounds in dealing with end-of-life decision-making for those they love, when empowered to be part of the process. There are studies suggesting to us that this approach may not be favoured in other areas. The SUPPORT investigators found that bringing patients' wishes to the notice of treating physicians did not improve the quality of end-of-life care . In a more recent French study, the patient's family was involved in only 44% of cases . We have no data that the process we have developed is better than any other. We have received two letters of complaint since 1982 related to withdrawal of care against the families' wishes, referred by the New South Wales Complaints Unit. No action against the doctor involved was deemed necessary. In contrast, we have a vast quantity of mail thanking us for the care and consideration shown. In 1967, the art critic John Berger addressed the question 'What is a human life worth?'. His answer was: "I do not claim to know what a human life is worth -- the question cannot be answered by word but only by action, by the active of a more human society" . We believe there are enormous benefits to both consumers and deliverers of health care in the active creation of a more humane intensive care unit, and this is an appropriate area for our speciality to show leadership. Competing interests : None declared. Backmatter: PMID- 12398770 TI - The International Sepsis Forum's controversies in sepsis: corticosteroids should be used to treat septic shock AB - Abstract | The use of corticosteroids in septic shock remains controversial. It has been demonstrated that high doses of steroids (30 mg/kg methylprednisolone) for short periods of time are not beneficial. More recent studies using smaller doses (200 --300 mg/day hydrocortisone) for longer periods of time have shown beneficial effects. These positive effects have included reversal of shock, trends toward decreased organ system dysfunction and decreased mortality. Based on the high proportion of patients who have relative adrenal insufficiency, the benefits of low doses of steroids and the minimal risks, steroids should be used to treat septic shock. Keywords: corticosteroids, relative adrenal insufficiency, septic shock, survival, corticosteroids, relative adrenal insufficiency, septic shock, survival, Introduction : In the early 1980s we were big proponents of using corticosteroids in septic shock, whereas in the late 1980s, following publication of the findings of the Veterans Administration Systemic Sepsis Cooperative Study Group on corticosteroids in sepsis and those of the study of steroids in sepsis conducted by Bone and colleagues , we thought corticosteroids were a bad idea. In fact, subsequently, two meta-analyses concluded that steroids were not beneficial in sepsis and septic shock. Based on recent data, there is enough evidence to suggest that we probably should be using corticosteroids in our septic shock patients. Although we support the use of steroids for treating septic shock, we clearly need a large trial in this area. In fact, we are starting a prospective, randomized, double-blind trial in Europe that will enroll large numbers of patients with septic shock (the CORTICUS study). In 1984, our group showed that there was a short period of time with a difference in mortality between a steroid-treated group and a control group, and that there was a significant reversal in shock between the groups at 24 hours. We stated at that time that perhaps we should be giving steroids for a longer period of time rather than just two doses. This approach was likely to reverse shock in more patients and perhaps improve survival. There has clearly been a change in thinking of how steroids should be used, in treating both septic shock and the acute respiratory distress syndrome. Older studies used large doses for short periods of time, and were given early in the treatment -- typically one or two doses of methylprednisolone (30 mg/kg). This was often done before any organisms were identified, and hence using corticosteroids might be associated with an increased incidence of complications related to superinfections. However, more recent studies recommend smaller doses of steroids, used for longer periods and given even later in the course of the disease. Relative adrenal insufficiency : During septic shock there is an important problem of relative adrenal insufficiency. This causes adrenergic receptor desensitization -- a decreased number of both alpha- and beta-adrenergic receptors. This is either related to sepsis itself or to the use of vasopressors for long periods of time. Corticosteroids may help with the resensitization of these receptors, leading to haemodynamic improvement-decreasing catecholamine requirements, reversal of shock and perhaps even reversal of organ system failure. All of these effects may decrease mortality. Annane and colleagues found the highest 28-day mortality (82%) in patients who had high baseline cortisol levels and who did not respond to adrenocorticotrophic hormone (ACTH) by increasing cortisol levels by greater than 9 mug/dl. If one looks at relative adrenal insufficiency in terms of nonresponders to ACTH (<= 9 g/dl), then the mortality in that group was 72%, whereas in those who did respond to ACTH the mortality was 32% -- more than double. How common is relative adrenal insufficiency? In the study conducted by Annane and colleagues it was 54%, but in some recent data there are suggestions that between two-thirds and three-quarters of patients with septic shock may have relative adrenal insufficiency. Therefore corticosteroids would be helpful in many patients. Clinical data looking at the increase in mean arterial pressure after noradrenaline (norepinephrine) administration show that there is a greater response to noradrenaline in patients with normal adrenal function than in those with impaired adrenal function. However, when one gives low doses of corticosteroids to patients with impaired adrenal function the responses to noradrenaline move closer together, so that there are no longer statistical differences in the noradrenaline effects of increasing blood pressure . Shock reversal : The prospective, randomized, double blind study conducted by Bollaert and colleagues looked at 41 patients who required catecholamines for more than 48 hours and had been in shock for 5 --6 days. By that time the patient's infectious organisms had most likely been identified and they were probably on appropriate antibiotics, so there were few problems in terms of complications of corticosteroids. Patients were given 100 mg hydrocortisone intravenously three times a day for 5 days. Treatment was discontinued if shock did not reverse. If shock reversed, then treatment was continued with half doses for 3 days and quarter doses for another 3 days and stopped. There was significant shock reversal at 7 days, and 28-day mortality showed a trend toward significance (63% versus 32%). Briegel and colleagues performed a prospective, randomized, double-blind, single-centre study in hyperdynamic septic shock patients. They enrolled septic patients who were on vasopressors and had been in shock for at least 3 days. While on vasopressors, patients were given a bolus of 100 mg hydrocortisone or placebo, followed by a continuous infusion of 0.18 mg/kg per hour of hydrocortisone or placebo. In addition to a decreased mean time for shock reversal, there also seemed to be a trend toward decreased organ system dysfunction in the steroid-treated group as compared with the control group. A multicentre, prospective, randomized, double-blind French study involved 300 patients enrolled early after the onset of shock. Among nonresponders (<= 9 g/dl increase in baseline cortisol after ACTH), there was greater survival in the steroid-treated group as compared with the placebo group. Interestingly, there was a trend toward a higher mortality in those patients who were responders and treated with steroids. Therefore, it is important to evaluate corticosteroids in a larger group of patients, evaluating both responders and nonresponders. In addition to efficacy, which is discussed above, safety is also important. The problems associated with corticosteroids occurred in those studies that used high doses (1 --2 g methylprednisolone). When lower doses (200 --300 mg/day of hydrocortisone) were used, superinfection and other complications were not observed. The one study that did have a complication was the study conducted by Briegel and colleagues . Patients developed hyponatraemia and increased alanine aminotransferase levels at day14. The hyponatraemia was probably related to the continuous infusion. Conclusion : What should one do at the present time? For routine use of low doses of corticosteroids in septic shock, there are minimal or no risks. There are data to suggest that relative adrenal insufficiency is a common problem in the majority of patients with septic shock. Therefore, physicians can reverse shock and improve survival with corticosteroids, as was shown in the prospective, double-blind study reported by Annane and colleagues . One should be aware that, in the study by Annane and colleagues, patients were in septic shock with systolic blood pressure below 90 mmHg for more than 1 hour. This is not the usual patient because most respond to fluids and vasopressors within an hour. In summary, in the typical septic shock patient, based on the data we already have, given the frequency of relative adrenal insufficiency in this population and the lack of a downside to the use of lower doses of corticosteroids, the routine use of corticosteroids is probably beneficial. If physicians do not want to use them early, then they should wait a few days so as not to worry about superinfection. Competing interests : PST received an honorarium from the International Sepsis Forum for helping to write this commentary. Abbreviations : ACTH = adrenocorticotrophic hormone. Backmatter: PMID- 12398781 TI - Clinical review: Hypertonic saline resuscitation in sepsis AB - Abstract | The present review discusses the hemodynamic effects of hypertonic saline in experimental shock and in patients with sepsis. We comment on the mechanisms of action of hypertonic saline, calling upon data in hemorrhagic and septic shock. Specific actions of hypertonic saline in severe sepsis and septic shock are highlighted. Data are available that support potential benefits of hypertonic saline infusion in various aspects of the pathophysiology of sepsis, including tissue hypoperfusion, decreased oxygen consumption, endothelial dysfunction, cardiac depression, and the presence of a broad array of proinflammatory cytokines and various oxidant species. The goal of research in this field is to identify reliable therapies to prevent ischemia and inflammation, and to reduce mortality. Keywords: hemorrhagic, hypertonic saline, inflammation, sepsis, shock, hemorrhagic, hypertonic saline, inflammation, sepsis, shock, Introduction : The incidence of septic shock has increased during the past several decades, whereas mortality rates have remained constant or have decreased slightly . Septic shock is associated with high mortality rates of 30 --80% . Sepsis presents with a systemic inflammatory response, peripheral vasodilatation, myocardial depression, intravascular volume depression, and increased metabolism. Despite considerable knowledge of the pathophysiology of the systemic inflammatory response syndrome, clinical trials using interventions such as immunotherapy have yielded negative results . Global tissue hypoxia results in an imbalance between systemic oxygen delivery and demand, and is a key development preceding multiple organ failure and death . Rivers and colleagues demonstrated the importance of goal-directed therapy in septic shock and severe sepsis. An early resuscitation strategy, which was goal oriented with respect to manipulation of cardiac preload, afterload and contractility, reduced the incidence of multiple organ dysfunction and mortality. Hemodynamic management in severe sepsis and septic shock includes rapid restoration of intravascular volume and adequate balance between systemic oxygen delivery and demand. Several liters of fluids (crystalloids or colloids) are usually necessary to normalize preload and filling pressures, with the objective of establishing adequate tissue perfusion and oxygen delivery . The infusion of several liters of fluid is associated with the adverse effect of extravasation into the interstitial space. In sepsis in particular, this may result in pulmonary edema. Nevertheless, adequate volume repletion with hemodynamic normalization may not be sufficient to prevent persistent microcirculatory dysfunction, which may cause ischemia and tissue damage . The observation reported by Velasco and colleagues in 1980 of beneficial effects of 7.5% saline solutions in dogs with severe hemorrhagic shock attracted interest to this field. The short duration of the circulatory effects of hypertonic saline solution (HSS) has been attributed to a rapid equilibrium of the hyperosmotic solute between extracellular and intra-cellular compartments. Therefore, HSS has been combined with colloids (i.e. dextran or hetastarch) in order to achieve a longer intravascular effect. This combination has synergistic effects, by increasing plasma osmolarity and osmotic pressure . Since the 1980s, several studies have been performed that used small volume resuscitation , which is defined as a rapid infusion of HSS (NaCl 7.2 --7.5%), in combination with dextran or hetastarch, at a dose of 4 ml/kg into a peripheral vein . Recent studies have used HSS in the treatment of sepsis and have demonstrated some promising beneficial effects. Hypertonic resuscitation in experimental models of sepsis : There is a decreased susceptibility to sepsis following administration of HSS in hemorrhagic shock. After 24 hours of cecal ligation and perforation, animals that received HSS had fewer bacteria in serum, lower formation of abscesses in liver and lungs, and less pulmonary and hepatic injury . Reduced organ injury in the HSS group might have been related to an improved hemodynamic profile and decreased extrapulmonary volume. Also, effects on microcirculation (i.e. reduction in ischemia and effects related to immune function) might have contributed partly to decreased organ injury. Experimental studies in septic shock have shown beneficial effects similar to those reported in studies of hemorrhagic shock . Studies of HSS alone or combined with het-astarch in sepsis demonstrated hemodynamic improvements, but these effects had a short period of action. However, these findings bring new possibilities to treatment of septic patients, if the treatments are instituted early in course of the disease. Hemodynamic resuscitation per se can reduce the inflammatory response in sepsis, reducing the phenomenon of ischemia/reperfusion . On the other hand, several studies have demonstrated that HSS modulates immune function favorably (i.e. by reducing production and release of proinflammatory cytokines and augmenting interleukin-10 induction; by reducing L-selectin expression in neutrophils; and by reducing the oxidative burst) . Together, those studies indicate that HSS has actions in two important aspects of septic shock: hemodynamics and immunomodula-tion. Notwithstanding the recent intense focus on immunomodulation in sepsis, Rivers and colleagues showed that early hemodynamically centered therapy yielded significant benefits with respect to outcome. Observations from several experimental studies suggest that HSS combined with a colloid solution is able to improve macrocirculation in sepsis . Also, HSS prevented vascular dysfunction and restored microcirculatory blood flow by capillary reopening. This effect resulted in a beneficial redistribution of regional blood flow to heart, kidney, and splanchnic organs. Hypertonic resuscitation in clinical studies of sepsis : The first clinical study to evaluate the effects of small volume resuscitation in severe sepsis was conducted by Hannemann and colleagues . Those authors observed increased oxygen transport, cardiac output, and pulmonary capillary wedge pressure in patients treated with HSS. Except for the increase in pulmonary capillary wedge pressure, none of the cardiovascular changes lasted for longer than 60 min. Plasma sodium levels increased and normalized within 24 hours after HSS infusion. Oliveira and colleagues studied the hemodynamic effects of a hypertonic saline/dextran solution as compared with those of a normal saline solution in severe sepsis. Patients were randomly assigned, in a blinded manner, to receive 250 ml of a solution of either normal saline (n = 16) or hypertonic saline (NaCl 7.5%/dextran 8%; n = 13). Before they received normal saline or HSS, patients had to have been stable (i.e. no requirement for vasoactive drug or volume change) for at least 60 min. Over the 180 min following infusion of normal saline or HSS (i.e. the period of study), the rate of infusion of regular fluid or vasoactive drug was not changed. The cardiac and stroke volume indices increased, and systemic vascular resistance decreased only in the HSS group, without any change in arterial pressure. The increase in plasma sodium levels lasted for 6 hours in the HSS group. Those investigators concluded that hypertonic saline/dextran solution improved cardiovascular performance and resuscitated severely septic patients through a volume effect, but may also directly improve cardiac function. Mechanisms of action of hypertonic saline solution : The main proposed mechanisms of action of HSS are as follows : instantaneous mobilization of fluids from intracellular to extracellular compartments by the osmotic gradient produced by HSS; increased myocardial contractility; reduced endothelial and tissue edema, improving microcirculation; improved blood viscosity due to hemodi-lution; and immunomodulation. Intravascular volume expansion | A rapid increase in mean arterial pressure occurs following HSS infusion. Studies have shown a redistribution of fluids from the perivascular to the intravascular space, and consequent plasma expansion . The hemodynamic effects of HSS infusion have been studied in sepsis . Most of the studies found that HSS infusion caused a rapid and significant increase in oxygen delivery, and elevated cardiac output and increased oxygen extraction, but these effects were transient . Therefore, despite the immunologic background of sepsis and the significant role played by immunologic mechanisms in the disease, hemodynamic resuscitation has also proved important in management of sepsis . HSS may be able to resuscitate septic patients better and more rapidly, during the critical 'golden hours' of the disease. Cardiac contractility effects | Improvement in myocardial contractility may be related to a direct hyperosmolar effect, restoring transmembrane potentials or decreasing myocardial edema . Reported findings indicate that ventricular contractile force is enhanced by moderate degrees of hyperosmolarity and is depressed by severe hyperosmolarity both in vivo and in vitro . HSS has been shown to increase left ventricular dP/dtmax, cardiac output, and stroke work at equivalent or lower atrial filling pressures than with isotonic solutions . Myocardial function is depressed even in the early hyperdynamic phase of sepsis . However, hypertonic solutions have been shown to improve contractility in animal and human studies of sepsis . These improvements in myocardial performance were unrelated to changes in coronary flow or myocardial oxygen consumption . Neural effect: the role of lung innervation | Two forms of experimental evidence indicate that a pulmonary reflex mechanism may participate in the resuscitative effects of hypertonic saline. First, it has been suggested that passage of hypertonic saline through the pulmonary circulation is necessary for resuscitation. In dogs, prepulmonary (right atrial, pulmonary arterial) administration resulted in resuscitation, but postpulmonary (left atrial, aortic) administration did not provide effective resuscitation . However, data from sheep suggested that the site of administration has no influence on the resuscitative effects of hypertonic saline. The differences between the reports may be explained by the different species investigated. The second form of evidence for lung innervation is provided by studies indicating that vagal blockade attenuates the hemodynamic response to hypertonic saline administration in hypovolemic dogs . In contrast to this finding, nearly identical improvements in mean arterial pressure, cardiac output, heart rate, and blood chemistry parameters in response to administration of hypertonic saline into innervated and denervated pulmonary circulations were reported . However, Younes and coworkers conducted studies 7 days after surgery in a model of total lung denervation. In that model, HSS administration produced a sustained hemodynamic improvement in the innervated group as compared with the denervated group. The different preparations may explain those contrasting results . Endothelial effects | In the initial phase of hypovolemia and shock, the hypoxia and activation of polymorphonuclear cells in the endothelium of postcapillary venules produce endothelial cell edema. This leads to capillary lumen narrowing, which can cause complete obstruction of local blood flow and reduction in oxygen transport . During small volume resuscitation, the intracellular fluid is primarily mobilized from microvascular endothelial cells and erythrocytes. This effect is more pronounced in capillaries in which edema is greater. It produces a reduction in hydraulic resistance and an improvement in tissue perfusion. It has been demonstrated that a reduction in endothelial volume of 20% after infusion of HSS/dextran, as well as an increase in sinusoidal perfusion, may occur, resulting in significant improvements in hepatic energetic status and excretory function . Vasoactive mediators | Studies have revealed increased cardiac output and restoration of peripheral blood flow mediated by vasodilating substances released after HSS infusion, especially prostacycline, together with an increase in the 6-cheto-prostaglandin F1alpha :thromboxane B2 ratio . The decrease in total peripheral resistance is the main factor responsible for the hypotension that occurs immediately after infusion of HSS . The neuroendocrine response to 7.5% HSS/dextran after hemorrhagic shock was quantified in pigs . The result of hemodilution in association with plasma volume expansion was decreased plasma levels of adrenocorticotropic hormone, cortisol, and aldosterone. Also, reductions in plasma concentrations of norepinephrine (noradrenaline), epinephrine (adrenaline), lysine, vasopressin, and renin were greater with hemodilution combined with plasma volume expansion than with hemodilution alone, indicating that alterations in hormone release have a role to play in cardiovascular response in this model of resuscitation. Immunomodulatory effects | Hemorrhage and sepsis often initiate a systemic inflammatory response that is accompanied by organ dysfunction, most commonly acute lung injury . Neutrophil sequestration in the lung is a necessary prerequisite for development of lung injury in most models of hemorrhagic and septic shock . Ischemia has been shown to lead to accumulation of neutrophils and other leukocytes in the microvascular bed of many organs . HSS has been shown to reduce lung injury after hemorrhagic shock . Those studies showed that HSS produced the following improvements in the lung: reduction in neutrophil accumulation, less neutrophils recovered on bonchealveolar lavage, reduced albumin leak and a lower degree of histopathologic injury. The mechanism of neutrophil sequestration or adhesion depends on the particular inflammatory condition. The CD11b integrin is a vital component of neutrophil --endothelial interactions, and in this respect Rizoli and colleagues showed that HSS prevented lipopolysaccharide-stimulated expression and activation of CD11b. Corroborating those data, HSS has been shown to decrease neutrophil L-selectin expression and to eliminate neutrophil priming by mesenteric lymph production ; this suggests that HSS reduces lung injury by preventing neutrophil adhesion to endothelium. Also, Oreopoulos and colleagues showed inhibition of ischemia/reperfusion-induced hepatic expression of intercellular adhesion molecule-1 mRNA with HSS as compared with normal saline. Studies into the action of HSS on cellular mechanisms have yielded data that indicate that HSS regulates the expression and release of elastase, cytokines, free radicals, and adhesion molecules. T cells incubated at NaCl levels of up to 180 mmol/l exhibited 100% enhancement of proliferation (these concentrations of salt are similar to the plasma sodium levels that are introduced by the traditional dose of 4 ml/kg of a 7.5% NaCl solution) . Several circulating factors with T-cell suppressive activity have been identified in trauma patients, suggesting that these factors cause a down-regulation of T-cell function after trauma . Prostaglandin E2 is a T-cell suppressor that interferes with calcineurin-dependent signaling pathways, and thereby inhibits interleukin-2 production and T-cell proliferation. Human peripheral blood mononuclear cells were suppressed when incubated with prostaglandin E2 . T-cell proliferation was significantly enhanced when the cells were exposed to HSS. Cell-mediated immune function and splenocyte proliferation is significantly suppressed after hemorrhage, and HSS resuscitation clearly restored splenocyte function and cell-mediated immune function. Hypertonic saline modulates cellular signaling pathways | Most knowledge in this area stems from work with Saccharomyces cerevisiae . Hyperosmotic conditions trigger activation of a mitogen-activated protein kinase (MAPK) termed high-osmolarity glycerol response (HOG)1 in yeast cells. Han and colleagues identified a mammalian equivalent of HOG1 in monocytic cells. This protein, namely MAPK p38, shares approximately 60% amino acid sequence with HOG1. MAPK p38 is tyrosine phosphorylated and activated under hypertonic conditions, suggesting the existence of an osmolarity sensing system in mammalian cells. A human T-cell line (Jurkat cells) was used to investigate whether HSS triggers signaling events through protein phosphorylation . HSS exposure permitted tyrosine phosphorylation of cellular protein in a dose-dependent manner. Sepsis, trauma, and hemorrhage activate neutrophils and can trigger excessive release of cytotoxic mediators, damaging host tissues and resulting in major post-traumatic complications . Clinically relevant hypertonicity suppressed degranulation and superoxide formation in response to N-formyl-methionyl-leucyl-phenylalanine (fMLP), and blocked the activation of the MAPKs ERK 1/2 and p38. HSS did not suppress neutrophil oxidative burst in response to phorbol myris-tate acetate. This indicates that HSS suppresses neutrophil function by intercepting signal pathways upstream of or apart from protein kinase C. Neutrophils incubated in hypertonic saline showed a reduction in platelet-activating factor mediated MAPK p38 signal transduction. Clinically relevant levels of hypertonic saline attenuated platelet-activating factor mediated beta2 integrin expression, superoxide radical production, and elastase release . Recent evidence suggests that cytoskeletal reorganization is critical for receptor-mediated signal transduction. Cytoskeletal disruption prevented attenuation of receptor-mediated MAPK p38 activation by hypertonic saline. Therefore, hypertonic saline alters cell shape, and this is followed by cytoskeletal reorganization with a resultant immunomodulatory effect . Data have been reported that indicate that HSS augments interleukin-10 induction by lipopolysaccharide at the gene level and reduces tumor necrosis factor levels, independent of nuclear factor-kappaB signaling. These actions may explain the lesser degree of injury following HSS administration. However, because HSS reduces but does not completely abrogate proinflammatory pathways, there is an adequate balance between proinflammatory and anti-inflammatory cytokines, thus maintaining the ability to fight bacteria efficiently. Conclusion : The potential beneficial effect of small volume resuscitation with HSS, which has been extensively studied in hypovolemic shock, appears to be reproducible in various models of experimental septic shock. The anti-inflammatory effects of hypertonic saline on neutrophils, oxidative burst, and cytokine release are mediated through the signaling molecule MAPK p38. These effects may reduce the excessive proinflammatory action found in sepsis, reducing the degree of damage to multiple organs. Hemodynamic effects have been widely demonstrated, and recent data showed that early goal-directed therapy is very important in reducing mortality. The vicious circle of ischemia, inflammation, fluid extravasation, and ischemia that occurs in sepsis may perpetuate damage to organs. A therapy that simultaneously blocks both of the damaging components of sepsis, namely ischemia and inflammation, will probably have an enormous impact on our ability to manage this condition. HSS is emerging as a possible preventive therapeutic in sepsis. Competing interests : None declared. Abbreviations : HOG = high-osmolarity glycerol response; HSS = hypertonic saline solution; MAPK = mitogen-activated protein kinase. Backmatter: PMID- 12398776 TI - Pro/con ethics debate: Should mechanical ventilation be continued to allow for progression to brain death so that organs can be donated? AB - Abstract | Organ transplants continue to redefine medical frontiers. Unfortunately, current demand for organs far surpasses availability, waiting lists are long and many people die before the organ they desperately need becomes available. One proposed way to increase organ availability is to admit patients to the ICU with severe neurological injuries, for a trial of therapy. If the injury is irretrievable, discussions would then focus on extending ventilation for potential brain death/organ donation if a prior wish to donate is known or if the substitute decision maker consents. The following debate discusses the ethical dilemmas of waiting for brain death. Keywords: brain death, ethics, organ procurement, transplant, brain death, ethics, organ procurement, transplant, The scenario : The patient is a previously healthy 17-year-old boy who was transferred by air ambulance from a regional community hospital after nearly drowning. During a boat ride in a lake, the boy fell out of the boat and became entangled and submerged under water for an undetermined period of time. The boy was pulled to shore, where cardiopulmonary resuscitation was begun immediately because of absent vital signs. The boy was brought to the community hospital after cardiorespiratory arrest for approximately 20 --30 min. Vital signs were absent when the patient arrived. The patient was given one round of resuscitation medications, after which a heart rate and cardiac output were restored. The patient was transferred to a tertiary care paediatric hospital. On arrival at the referral hospital, the patient had no spontaneous movements, no motor response to pain, fixed and dilated pupils, and absent corneal, gag and cough reflexes. Spontaneous respiratory efforts were detected. There was a delay in the arrival of family members who were out of country at the time. After 24 hours of mechanical support, his neurological condition was unchanged. Given the severity of hypoxic ischaemic brain injury and the patient's dismal prognosis, the family was counselled to withdraw mechanical support. They requested organ donation but were informed that he was not eligible as he did not fulfil brain death criteria in view of the presence of spontaneous respiratory efforts. They agreed to withdrawal of support. The patient died 10 min after withdrawal of mechanical ventilation. Questions : The outcome after out-of-hospital cardiac arrest in children and adults is well known. Did the present child justify intensive care unit admission and why? Twenty-four hours after admission to the intensive care unit (ICU), the family of this teenager requested organ donation. When the patient has expressed a prior wish or families are motivated toward organ donation, is there an option to continue mechanical ventilation for 24 --48 hours to observe for progression to brain death in anticipation of organ donation? What are the risks and benefits associated with this option? What are the ethical issues raised by this? Con: no, mechanical ventilation should not be continued to allow for progression to brain death so that organs can be donated : Michael Parker The number of organs available for transplantation falls well below the number of people who require a transplant . In the United Kingdom, for example, there were 5354 people on the national transplant waiting list at the end of March 2000 . In practice, because of the recognized shortage of available organs, patients are only placed on this waiting list if they have a reasonable chance of receiving a donated organ. This means that the actual number of those who could benefit from a transplant is much greater than the number of people on the list. A range of different ways has been proposed for increasing the number of donors. One of the most controversial of these is the use of 'elective ventilation' of patients in deep coma and close to death who have no possibility of recovering but do not (yet) fulfill the criteria for brain death . In elective ventilation, patients are ventilated for a few hours to allow preparations for the removal of organs after death to take place. There is evidence that this could increase the number of available organs significantly. The practice was used in Exeter, UK between 1988 and 1994 (when it was declared unlawful), and led to a 50% increase in the number of organs suitable for transplantation there . The case described in the present scenario is an example of the kind of situation in which elective ventilation might make available organs that would otherwise be lost. One way of analysing the ethical implications of elective ventilation is to consider the foreseeable harms and benefits of its use. It might seem from this consequentialist perspective, initially at least, that emphasis ought to be placed on increasing the number of organs available for transplant. A more complete consequentialist analysis would of course have to consider other possible harms and benefits. One potential harm that is surely going to be relevant to any such analysis is the small possibility that, rather than dying following elective ventilation, patients may end up in a persistent vegetative state (PVS) . The chances of this happening are small, but from the point of view of a consideration of the patient's best interests it is not only be the size of the risk that will be of importance, but also its seriousness. Another set of factors important in any consequentialist account will concern the opportunity costs of elective ventilation. In the case presently described, the procedure will require an intensive therapy unit bed for up to 2 days. The question of whether this bed, or the equivalent resource spent elsewhere, would save more lives is surely important and not easy to answer without a full consideration of the health economic implications. The most common objection to the use of elective ventilation is not consequentialist, however, but based in respect for patient autonomy. Elective ventilation involves carrying out an invasive procedure on a living patient to which the patient has not consented. Furthermore, ventilation is not being carried out in the interest of the patient. The only purpose of the intervention is to provide organs for others and this is, in effect, treating patients merely as a means for the benefit of others and not as ends in their own right. One possible way of dealing with this objection would be to argue that there are grounds to believe that this is what the patient would have wanted. This is an argument often used by those in favour of a presumed consent model for transplantation. In the case of the patient who is brain dead, this argument has at least some force. A recent poll in Scotland, for example, found that a slight majority of those asked were in favour of a change to presumed consent. More broadly, opinion polls, in the UK at least, consistently reveal that around 70% of those asked would want their organs used on their death . Nevertheless, even if we were to accept this as a justification for moving to a presumed consent model for brain death, we would not be justified in making the same assumption in the case of elective ventilation. It is entirely unclear in the case of elective ventilation whether, if asked whether they would be willing to take the risk of ending up in PVS in order to provide organs for others, most or even many people would answer in the affirmative. Indeed, what evidence there is appears to point in the opposite direction. Many existing advance directives posit PVS as one of the states that patients would most wish to avoid . This is of course an empirical question requiring further research. Nevertheless, if we are to take patient autonomy seriously, we should not carry out invasive procedures on incompetent patients that we do not consider to be in their best interests without compelling evidence to believe that this is what they would have wanted. In the present case, we have no such evidence. Whether from a consequentialist perspective or one based in the principle of respect for patient autonomy, any move to the use of elective ventilation must be based in both good empirical and health economic analysis and in informed public debate. In the meantime, what evidence there is would seem to point in the direction of caution and would indicate that, in the case under consideration, ventilation should not be allowed to proceed. Pro: yes, mechanical ventilation should be continued to allow for progression to brain death so that organs can be donated : Sam D Shemie Is it medically appropriate and/or ethical to extend ventilation in anticipation of brain death for the purposes of organ donation? To clarify the challenges presented by this case, each of the following issues must be discussed individually: admission of severely brain-injured patients to the ICU, extension of ventilation, anticipation of brain death, and brain death for the purposes of organ donation. Admission of severely brain-injured patients to the ICU | The decision to admit severely brain-injured patients to the ICU is complex and is influenced by many factors. Severity of injury and potential for salvage are clearly dominant factors. Distinguished from adult practice, acutely brain-injured paediatric patients are rarely denied admission on the grounds of poor neurological prognosis. Paediatric death outside ICUs is an unusual event, with 85% of deaths at the Hospital for Sick Childrenoccurring in the ICU setting (SShemie, unpublished data). In general, practitioners outside the ICU are not comfortable with paediatric death, with grounds for prognosticating, and with withdrawal of established mechanical ventilation. A trial of ICU therapy is initiated to collect and confirm the facts upon which prognostications are based and to allow families time to adjust and be counselled. Severely brain-injured adults with artificial ventilation already established, in contrast, may be evaluated in the emergency room and, based on the perception of bad prognosis, be denied access to ICU services. Although based in large part on a poor anticipated prognosis, resource limitations (lack of available ICU nurses/bed space) must be acknowledged as a profound influence on these triage decisions. Studies predicting neurological outcome after cardiac arrest suggest that the most reliable predictors are apparent in the range of 24 hours to 72 hours after arrest. One can credibly advocate that a short-term trial of ICU therapy is warranted in any acutely brain-injured patients in order to confirm facts to avoid expedited decisions that occur in emergency rooms. In addition, there are evolving neuroprotective therapies that may benefit patients which traditionally have been perceived to have irretrievable outcomes. This is well supported by the improvements in neurological outcome with the use of hypothermia after cardiac arrest . In the face of advancing techniques of successful neuroprotection, there is concern for the potential self-fulfilling prophecy of the selecting out of severely brain-injured patients by preventing access to ICU care. Extension of ventilation | Limitation of life-sustaining technology has become standard practice for ICU-based end-of-life care, with the majority of deaths in neonatal, paediatric and adult ICUs being preceded by the withdrawal and/or withholding of some form of life support . However, there are inconsistencies in these practices with variation over time, between centres and between clinicians . The criteria for what may be considered futile therapy remains undefined and, for the same acuity of illness, withdrawal practices may vary . Observers of end-of-lifediscussions in ICUs have concluded that although life support technologies are traditionally deployed to treat morbidity and to delay mortality in ICU patients, they are also used to orchestrate dying. The tempo of withdrawal influences the method and timing of death. Decisions to withhold, to provide, to continue or to withdraw life support are socially negotiated to synchronize understanding and expectations among family members and clinicians . Ethical principles dictate that we must act in the interests of the patient first and foremost. In the complex realities of bedside ICU care, however, life support is manipulated in many ways that are not strictly in the primary interest of the patient. This is seen during the family's (or subspecialist's) adjustment phase of understanding the disease and accepting the terminal phase of illness; this communication process may take days to weeks, and sometimes months. There may be conflict or disagreement with the recommendations of the ICU team, resulting in prolongation of life support. This may be under the guise of acting in the interest of the patient but is, in reality, acting in the emotional interests of the family. There are compassionate reasons to extend ventilation; waiting for extended family to arrive from overseas, or not to have the patient die on a special day (e.g. Christmas or a birthday). It is often difficult to separate the interests of the family from the interests of the patient, and this reality is exaggerated in the paediatric sphere. Can one ethically justify extension of ventilation? Certainly, there is clinical precedence for this practice, many reports of which are aforementioned. A majority of Canadian paediatric intensivists are in favour of extending ventilation for organ donation. For example, in response to the survey question 'in the setting of acute brain injury, would you extend the duration of life support for brain death to potentially occur', 68% of respondents said yes, 21% were unsure and only 11% of respondents said no . The issue at stake here, however, is the patient's interest. Is it being compromised by the extension of ventilation? Is it beneficial, harmful or neither? Suffering is an exaggerated concept in comatose ICU patients who have lost consciousness. Whether they spiritually suffer or their dignity is compromised is at best subjective and uncertain. The issue of benefit versus harm rests between the benefit of actualizing the donation of organs from an individual or family who have expressed this intent and the harm of extending ventilation. Anticipation of brain death | Neurological prognosis after devastating brain injury is able to distinguish between extremes of outcomes. Prognostic criteria for outcomes after cardiac arrest, particularly those presenting with asystole to the emergency room, have been well defined in paediatrics and in adults . Bad outcome, however, is defined as death or vegetative survival, and there has been no clear predictive data that distinguishes between brain death and vegetative survival . Clearly, any intervention that may convert a patient destined to die after withdrawal of ventilation into a vegetative survivor is concerning. This has been anecdotally cited by ICU practitioners as the primary issue of concern regarding the extension of ventilation in severely brain-injured ICU patients. Experience in clinical ICU practice dictates some guidelines where the risks of extending ventilation are minimized. The temporal changes in neurological function after brain injury give information about its anticipated evolution. A deteriorating neurological course may be anticipated if the signs of neurological function (motor score, brain stem reflexes) are deteriorating over time. A comatose patient who decorticates to pain on admission to ICU with intact brain stem reflexes may proceed to lose any motor response to pain, followed by gradual loss of brainstem reflexes (e.g. unilateral fixed and dilated pupil) over the ensuing 24 hours. This scenario may increase the chance of proceeding to brain death, and minimizes the risk that extending ventilation would allow recovery of vegetative brainstem function. Conversely, a patient may present with a Glasgow Coma Scaleof 3, with no response to deep pain on arrival at the ICU, and then at 24 --48 hours may develop decerebrate posturing. This scenario reflects an evolving improvement in neurological function. It is not one where extension of ventilation is safe and may increase the chances of vegetative survival. Brain death for the purposes of organ donation | There remains a perception that extending ventilation to allow for brain death for the purposes of organ donation is an act against the interests of the patient. It may in fact serve the interest of the patient if the expressed wishes were known and the risks are small. Consent decisions are primarily influenced by prior knowledge of the deceased individual's wishes . It is well established that the act of organ donation aids in the grieving process for family members . Donation decisions are a function of attitude toward donation and the religious, cultural, altruistic, normative, and knowledge-based beliefs that comprise the attitude. For the individual expressing intent to donate, there is actual and spiritual sustenance that is derived from the decision to give. If the surrogate decision makers are aware of the patient's to wish to donate, it may be in the patient's interest to pursue any reasonable avenue to fulfil the desire to donate. Abbreviations : ICU = intensive care unit; PVS = persistent vegetative state. Backmatter: PMID- 12398772 TI - Paediatric intensive care: out of commission AB - Abstract | Problems with commissioning paediatric intensive care stem both from difficulties in recruitment and retention of nurses, and from incoherent or nonexistent national audit. Pyramidal career structures and patterns of remuneration that concentrate on administrative responsibility over clinical skills underlie the former, whereas poor audit conceals variations in both service quality and demand. Epidemiologically superior data are required if we are to solve commissioning problems. We need to know what happened to every child from a defined population receiving intensive care and whether a lack of resources means that some children are denied intensive care. Keywords: audit, epidemiology, intensive, nursing, paediatric, audit, epidemiology, intensive, nursing, paediatric, Introduction : It was recently revealed that one quarter of paediatric intensive care beds in Holland are closed and that many critically ill children have to be transferred to receive care. This is not an exclusively Dutch problem. Problems organizing paediatric intensive care exist in most health care systems. If we are to avoid crises such as this then we have to solve two fundamental problems: the way in which we staff the intensive care units with nurses; and the lack of information that we have regarding the service that we are trying to commission. Nurse recruitment, retention and funding : Paediatric intensive care nurses are a scarce resource for a number of reasons. First, they are usually required to have specific higher postgraduate training in order to work at a basic grade. Such training may not enhance their pay as compared with other nurses in other disciplines. Second, their career structure is pyramidal, which limits the opportunities for individual progression and increases the appeal of lateral career moves into less stressful environments such as community nursing or (in the UK) National Health Service direct. From the nurse's perspective, salary progression is usually linked purely to administrative responsibility and often fails to recognise significant additional practical and intellectual skills, such as intensive care training itself or additional experience (e.g. in techniques such as haemofiltration or extracorporeal membrane oxygenation). Even when additional recurrent funds are identified to commission a paediatric intensive care service, it can thus prove prohibitively difficult to open intensive care beds. Furthermore, when service provision is inadequate, problems with recruitment suffer from negative reinforcement. Jobs in a stressed service are inevitably anticipated to be stressful themselves. There is no single answer to this problem. Centralization of care in large high volume units enables the best return from any limited resource, and it is tempting to assume that nurses will be the same as any other resource in this respect. The impact of centralization on nurse recruitment, however, has not been determined. Nurses are unlikely to commute large distances to find work in intensive care in preference to a local change in specialty. Lack of information : Effective planning and commissioning of a paediatric intensive care service requires close audit of activity, which we lack. In many cases this lack of information conceals variation in resource provision and differences in performance. The required information is naturally specific. The capacity of a paediatric intensive care service can be described in terms of the number of physical bed spaces present, but this does not reassure us that the beds are accessible. Variations in patient dependency, the numbers of nurses available and their skill mix all have to be taken into account when deciding whether a bed can be used at a given time. Even knowledge of the number of accessible beds tells us nothing about the amount of work being done. Such information can only be gained by looking at patient flow (admission rate, duration of stay, occupancy, readmission rate) and intervention rates. Even then, information regarding quality of care is lacking. A limited view of the quality of care (its effectiveness) can be inferred from standardized mortality ratios generated using mortality prediction models. There are in essence two such models available for paediatrics . They have considerable differences and are not universally applied. However, the use of mortality data in this way has been questioned in paediatric intensive care, where survival rates are greater than 90% and morbidity may be of increased relevance because of the potential longevity of survivors. Most literature and research using standardized mortality is based on the performance of individual units or groups of units. From an epidemiological perspective, however, it is preferable to know what happened to every child from a defined population who received intensive care (irrespective of where it was provided) and whether a lack of resources meant that some children were denied intensive care . Change and variation in standards : There is evidence that increasing numbers of children are receiving or are expected to receive intensive care. In the UK (or at least in Birmingham ) this is occurring without a change in the intubation rate, implying that the change could represent partial resolution of an asserted shortage . Alternatively, the clinical threshold for intervention (intubation) is falling. Paediatric intensive care space is used as it becomes available. Where there are ample resources there is a tendency to provide high dependency care and 'observation' on the intensive care unit, whereas triage otherwise limits this tendency. Hence, when great variation in the incidence of intubation is observed within or between health care systems, one can infer variation in resource provision. The greatest concern must be when refused admissions occur in units with high intubation rates. The appropriateness of intensive care admission or intervention ultimately still has to be judged on a case review basis. Comprehensive audit : The development of the mortality prediction model 'PIM' (Paediatric Index of Mortality) has involved a collaboration that has, among others, incorporated all paediatric intensive care delivered in Australia. The epidemiological superiority of these data will increase the influence of the conclusions drawn from it. In the UK, where we have long suffered from similar problems to those currently affecting Holland , some centres are also contributing to the PIM collaboration . However, a system for centralized audit is also being established. First, this is being achieved through a study designed to assess the relevant mortality prediction models, which is to include morbidity (United Kingdon Paediatric Intensive Care Outcome Study, UKPICOS) . The study has recruited all the major providers of paediatric intensive care in the UK. Second, the English Department of Health has commissioned a continuous audit of paediatric intensive care to follow on from that study called 'PICANET' (Paediatric Intensive care Audit Network), which will include the successful severity model. Hence, commissioners will have access to comparative, risk adjusted, performance data on which to make their decisions. All we have to do now is pay the nurses what they are worth! Competing interests : None declared. Abbreviations : PIM = Paediatric Index of Mortality. Backmatter: PMID- 12398769 TI - Hypothermia and neurologic outcome in patients following cardiac arrest: should we be hot to cool off our patients? AB - Abstract | Hypothermia as a protectant of neurologic function in the treatment of cardiac arrest patients, although not a new concept, is now supported by two recent randomized, prospective clinical trials. The basic science research in support of the effects of hypothermia at the cellular and animal levels is extensive. The process of cooling for cerebral protection holds potential promise for human resuscitation efforts in multiple realms. It appears that, at least, those patients who suffer a witnessed cardiac arrest with ventricular fibrillation and early restoration of spontaneous circulation, such as those who were included in the European and Australian trials (discussed here), should be considered for hypothermic therapy. Keywords: cardiac arrest, cerebral protection, cooling, hypothermia, resuscitation, cardiac arrest, cerebral protection, cooling, hypothermia, resuscitation, Introduction : In two recent issues of New England Journal of Medicine, studies using hypothermia in patients following cardiac arrest (CA) to improve neurologic outcome were presented and debated . Not a new issue, having first surfaced in the 1950s , hypothermia as a treatment strategy is potentially promising as a mechanism to curtail neurologic injury in specific, although not fully defined, patient situations. As resuscitative measures have expanded, the need for options to improve neurologic function after CA is of paramount importance. Two recent trials reported by Holzer and colleagues (conducted in Europe) and Bernard and coworkers (conducted in Australia) yielded statistically significant, positive outcomes . Our goals in the present commentary are to affirm the viability of hypothermia as a therapeutic intervention, to evaluate the European and Australian trials, and to explore the potential of hypothermia as a treatment modality. Table 1 | Comparison of two recent trials of hypothermia in cardiac arrest Hypothermia: the science : Data support the contention that hypothermia is not only biologically plausible as a therapy but also improves neurologic outcome in animals and, now, in humans . The timing and duration of treatment, as well as the degree of hypothermia, were shown to impact on efficacy and outcome . Given different mechanistic etiologies for neurologic injury, different diseases have been shown to respond variably to treatment with hypothermia . Finally, use of mild hypothermia has refuted the previously expected side-effect profile of hypothermia . The basis for the use of hypothermia in cerebral protection (i.e. to attenuate the effects of cerebral ischemia) is supported by animal studies. Cerebral ischemia causes early and late effects. Energy failure, ion pump failure, and release of free radicals and excitotoxic agents occur early, whereas inflammatory mechanisms and release of stress-related proteins progress over hours after reperfusion. Excitotoxicity and free radical formation promote cell damage in ischemic tissue early after reperfusion. Glutamate levels -- a major component of the excitotoxic response -- decrease during hypothermic treatment of ischemia in rabbits . By attenuating release of glutamate, the 'death funnel' of N-methyl-D-aspartate receptor stimulation (opening ion channels, allowing influx of calcium, and thereby producing the cascade of second messengers that activate kinases and proteases, leading to cellular destruction) is lessened . Later responses to reperfusion following ischemia include the production of stress-related proteins, such as heat shock proteins. These proteins, in turn, are believed to influence other gene products. Elevations in heat shock protein-70 in the hippocampus are blunted by hypothermia . Additionally, arachidonic acid products may be involved in an inflammatory cascade, affecting cell survival. In gerbils, hypothermia decreases levels of leukotriene B4 -- a substance linked to cerebral edema. Safar and Leonov, along with their research groups, conducted elegent animal studies in the 1980s and 1990s , verifying improvement in outcome when hypothermia is used as treatment in cardiac arrest models, both with functional ratings and electromyographic improvement. Weinrauch and colleagues demonstrated that hypothermia to 30 and 34C, achieved by a combination of partial bypass flow and surface cooling performed immediately after cardiac arrest, improves both neurologic deficit score and histologic damage scores in dogs . Hypothermia begun hours after the initial insult is not likely to affect the initial ischemic processes. Thus, early hypothermia is likely to be more effective than delayed hypothermia. In a rat model, delays of 15 min and 30 min preserved the beneficial effect of hypothermia, but with delays of 45 min there was no attenuation of infarct volume . In dogs, delaying hypothermia by 15 min obscured the benefit in functional outcome as compared with that with immediate hypothermia ; however, it might have attenuated tissue damage, as detected histologically. In addition to initiation timing, duration has been investigated. Increasing the duration of hypothermia appears to decrease infarct size . In a rat model, hypothermia with durations of 3 and 4 hours was superior to 2 hours in terms of effect on infarct volume, whereas 1 hour appeared ineffective. An early reported human series by Williams and Spence at Johns Hopkins in 1959 employed durations from 24 to 72 hours, with good outcomes at both extremes. Hypothermia: the current state : Although many models of neuroprotection in traumatic brain injury have shown a positive correlation between hypothermia and outcome, several studies in humans have failed to affirm this. In a recent meta-analysis of seven clinical trials conducted from 1993 to 2001 , it was concluded that hypothermia is not beneficial in the management of severe head injury. Those authors did, however, concede that further studies are 'justified and urgently needed'. Design problems exist in both of the two new trials of hypothermia for CA , namely potential bias (the treating physicians were unblinded), the sample sizes were small, and some of the treatment protocol aspects were different (such as the time of initiation of hypothermia [in the field versus hospital] and duration of hypothermia [12 versus 24 hours]). Critics of those studies have expressed concerns over several issues : the hypothermia and normothermia groups may not have been well matched; the sample sizes were small; the subgroups of patients with CA analyzed were small percentages of the total number of CAs (13 --19%); and side effects of hypothermia can be extensive. The responses of the principal investigators to those issues indicate that both groups were well matched, with median Glasgow Coma Scale scores of 3 in both groups and interquartile ranges from 3 to 4 or 5 . They also point out that, although the subgroup of CA in their studies was a small percentage of the whole CA population, future studies may show that hypothermia could confer a benefit in other subgroups . Finally, the two clinical trials from Europe and Australia showed no statistical difference in side-effect profiles between the normothermic and the hypothermic groups. Potential side effects of hypothermia such as arrhythmias, coagulopathies, infection, electrolyte disturbance, and hypothermia-induced polyuria were previously reported in the literature . In mild hypothermia, it appears that the incidence and severity of side effects is diminished. For example, the development of arrhythmia is temperature related; temperatures below 30C are more likely to cause serious arrhythmias such as ventricular fibrillation . Additionally, some of the complications experienced in early studies of hypothermia can be negated using modern intensive care monitoring and treatment plans. Despite the study flaws described above, outcomes show agreement with the relative percentages presented in other studies. In Fig. the numbers of favorable outcomes from the normothermia arms of the two trials can be seen to approximate closely those of other studies over time. Also, despite the small size of the samples, both studies achieved statistical significance . Figure 1 | Difference in hypothermia versus normothermia: study comparisons. Difference in hypothermia versus normothermia: study comparisons. Shown is the percentage favorable outcome, or survival to discharge, compared among the two recent studies of hypothermia as treatment following cardiac arrest , a small series from 1959 (27 patients, 12 treated with hypothermia) , and three nonhypothermic series . The right-most three bars are zero within the hypothermia group because they represent studies that were not designed to test hypothermia as an intervention . Visual comparison reveals the closeness of new data from the two trials with respect to the other studies . *, , These studies were not hypothermic trials; rather, they are included here to give a perspective on relative discharge statistics following cardiac arrest from other series. Hypothermia: our opinion : The evidence suggests that hypothermia reduces neurologic injury in animals and humans through several intricate biochemical and physiologic mechanisms, most of which we are only beginning to understand. The European and Australian trials both show statistically significant and clinically relevant improvement. Thus, we believe that the time has come to conduct extensive, large, multicenter trials using hypothermia to provide neurologic protection after cardiac arrest. The trials should include broader populations of CA patients (i.e. not limited to ventricular fibrillation arrest) and larger study populations, should explore a quicker method of cooling (such as intravascular cooling catheters), and should attempt to establish an effective duration of therapy (12 versus 24 hours versus other durations). Given the high incidence of CA and the speed with which patients typically come to medical attention, the numbers of patients available for recruitment should allow a reasonable study completion time. Additionally, if, as expected, the larger trials support the findings of the recent smaller trials, then this will provide the impetus to examine other causes of hypoxic/ischemic injury, such as acute ischemic stroke. Until such larger trials are conducted, it is our opinion that the evidence, provided in prior feasibility/safety studies as well as in the combined European and Australian trials reported earlier this year, supports employing mild hypothermic therapy in the patient populations studied (those who have suffered witnessed ventricular fibrillation arrest, restoration of spontaneous circulation, etc.). Competing interests : TLS and TPB have previously conducted trials involving temperature control for neurologic conditions other than cardiac arrest. Abbreviations : CA = cardiac arrest. Backmatter: PMID- 12398782 TI - Statistics review 5: Comparison of means AB - Abstract | The present review introduces the commonly used t-test, used to compare a single mean with a hypothesized value, two means arising from paired data, or two means arising from unpaired data. The assumptions underlying these tests are also discussed. Keywords: comparison of two means, paired and unpaired data, t test, comparison of two means, paired and unpaired data, t test, Introduction : Previous reviews in this series have introduced the principals behind the calculation of confidence intervals and hypothesis testing. The present review covers the specific case of comparing means in rather more detail. Comparison of means arises in many different formats, and there are various methods available for dealing with each of these. Some of the simpler cases are covered in this review, namely comparison of a single observed mean with some hypothesized value, comparison of two means arising from paired data, and comparison of two means from unpaired data. All of these comparisons can be made using appropriate confidence intervals and t-tests as long as certain assumptions are met (see below). Future reviews will introduce techniques that can be used when the assumptions of the t-test are not valid or when the comparison is between three or more groups. Of the three cases covered in this review, comparison of means from unpaired data is probably the most common. However, the single mean and paired data cases are introduced first because the t-test in these cases is more straightforward. Comparison of a single mean with a hypothesized value : This situation is not very common in practice but on occasion it may be desirable to compare a mean value from a sample with some hypothesized value, perhaps from external standards. As an example, consider the data shown in Table . These are the haemoglobin concentrations of 15 UK adult males admitted into an intensive care unit (ICU). The population mean haemoglobin concentration in UK males is 15.0 g/dl. Is there any evidence that critical illness is associated with an acute anaemia? Table 1 | Haemoglobin concentrations (g/dl) for 15 UK males admitted into an intensive care unit The mean haemoglobin concentration of these men is 9.7 g/dl, which is lower than the population mean. However, in practice any sample of 15 men would be unlikely to have a mean haemoglobin of exactly 15.0 g/dl, so the question is whether this difference is likely to be a chance finding, due to random variation, or whether it is the result of some systematic difference between the men in the sample and those in the general population. The best way to determine which explanation is most likely is to calculate a confidence interval for the mean and to perform a hypothesis test. The standard deviation (SD) of these data is 2.2 g/dl, and so a 95% confidence interval for the mean can be calculated using the standard error (SE) in the usual way. The SE in this case is 2.2/15 = 0.56 and the corresponding 95% confidence interval is as follows. 9.7 +- 2.14 x 0.56 = 9.7 +- 1.19 = (8.5, 10.9) Note that the multiplier, in this case 2.14, comes from the tdistribution because the sample size is small (for a fuller explanation of this calculation, see Statistics review 2 from this series). This confidence interval gives the range of likely values for the mean haemoglobin concentration in the population from which these men were drawn. In other words, assuming that this sample is representative, it is likely that the true mean haemoglobin in the population of adult male patients admitted to ICUs is between 8.5 and 10.9 g/dl. The haemoglobin concentration in the general population of adult men in the UK is well outside this range, and so the evidence suggests that men admitted to ICUs may genuinely have haemoglobin concentrations that are lower than the national average. Exploration of how likely it is that this difference is due to chance requires a hypothesis test, in this case the one sample t-test. The t-test formally examines how far the estimated mean haemoglobin of men admitted to ICU, in this case 9.7 g/dl, lies from the hypothesized value of 15.0 g/dl. The null hypothesis is that the mean haemoglobin concentration of men admitted to ICU is the same as the standard for the adult male UK population, and so the further away the sample mean is from this hypothesized value, the less likely it is that the difference arose by chance. The t statistic, from which a P value is derived, is as follows. In other words, t is the number of SEs that separate the sample mean from the hypothesized value. The associated P value is obtained by comparison with the t distribution introduced in Statistics review 2, with larger t statistics (regardless of sign) corresponding to smaller P values. As previously described, the shape of the t distribution is determined by the degrees of freedom, which, in the case of the one sample t-test, is equal to the sample size minus 1. The t statistic for the haemoglobin example is as follows. In other words, the observed mean haemoglobin concentration is 9.54 SEs below the hypothesized mean. Tabulated values indicate how likely this is to occur in practice, and for a sample size of 15 (corresponding to 14 degrees of freedom) the P value is less than 0.0001. In other words, it is extremely unlikely that the mean haemoglobin in this sample would differ from that in the general population to this extent by chance alone. This may indicate that there is a genuine difference in haemoglobin concentrations in men admitted to the ICU, but as always it is vital that this result be interpreted in context. For example, it is important to know how this sample of men was selected and whether they are representative of all UK men admitted to ICUs. Note that the P value gives no indication of the size of any difference; it merely indicates the probability that the difference arose by chance. In order to assess the magnitude of any difference, it is essential also to have the confidence interval calculated above. Comparison of two means arising from paired data : A special case of the one sample t-test arises when paired data are used. Paired data arise in a number of different situations, such as in a matched case --control study in which individual cases and controls are matched to each other, or in a repeat measures study in which some measurement is made on the same set of individuals on more than one occasion (generally under different circumstances). For example, Table shows central venous oxygen saturation in 10 patients on admission and 6 hours after admission to an ICU. Table 2 | Central venous oxygen saturation on admission and 6 h after admission to an intensive care unit The mean admission central venous oxygen saturation was 52.4% as compared with a mean of 59.2% after 6 hours, corresponding to an increase of 6.8%. Again, the question is whether this difference is likely to reflect a genuine effect of admission and treatment or whether it is simply due to chance. In other words, the null hypothesis is that the mean central venous oxygen saturation on admission is the same as the mean saturation after 6 hours. However, because the data are paired, the two sets of observations are not independent of each other, and it is important to account for this pairing in the analysis. The way to do this is to concentrate on the differences between the pairs of measurements rather than on the measurements themselves. The differences between the admission and post-admission central venous oxygen saturations are given in the rightmost column of Table , and the mean of these differences is 6.8%. In these terms, the null hypothesis is that the mean of the differences in central venous oxygen saturation is zero. The appropriate t-test therefore compares the observed mean of the differences with a hypothesized value of 0. In other words, the paired t-test is simply a special case of the single sample t-test described above. The t statistic for the paired t-test is as follows. The SD of the differences in the current example is 7.5, and this corresponds to a SE of 7.5/10 = 2.4. The t statistic is therefore t = 6.8/2.4 = 2.87, and this corresponds to a P value of 0.02 (based on a t distribution with 10-1 = 9 degrees of freedom). In other words, there is some evidence to suggest that admission to ICU and subsequent treatment may increase central venous oxygen saturation beyond the level expected by chance. However, the P value in isolation gives no information about the likely size of any effect. As indicated above, this is rectified by calculating a 95% confidence interval from the mean and SE of the differences. In this case the 95% confidence interval is as follows. 6.8 +- 2.26 x 2.4 = 6.8 +- 5.34 = (1.4, 12.2) This indicates that the true increase in central venous oxygen saturation due to ICU admission and treatment in the population is probably between 1.4% and 12.2%. The decision as to whether this difference is likely to be important in practice should be based on the statistical evidence in combination with other relevant clinical factors. However, it is worth noting that the confidence interval excludes 0 (the expected difference if the null hypothesis were true); thus, although the increase may be small (1.4%), it is unlikely that the effect is to decrease saturation. Comparison of two means arising from unpaired data : The most common comparison is probably that of two means arising from unpaired data (i.e. comparison of data from two independent groups). For example, consider the results from a recently published trial that compared early goal-directed therapy with standard therapy in the treatment of severe sepsis and septic shock . A total of 263 patients were randomized and 236 completed 6 hours of treatment. The mean arterial pressures after 6 hours of treatment in the standard and early goal-directed therapy groups are shown in Table . Table 3 | Mean and standard deviation of mean arterial pressure Note that the authors of this study also collected information on baseline mean arterial pressure and examined the 6-hour pressures in the context of these (using a method known as analysis of covariance) . In practice this is a more appropriate analysis, but for illustrative purposes the focus here is on 6-hour mean arterial pressures only. It appears that the mean arterial pressure was 14 mmHg higher in the early goal-directed therapy group. The 95% confidence intervals for the mean arterial pressure in the two groups are as follows. There is no overlap between the two confidence intervals and, because these are the ranges in which the true population values are likely to lie, this supports the notion that there may be a difference between the two groups. However, it is more useful to estimate the size of any difference directly, and this can be done in the usual way. The only difference is in the calculation of the SE. In the paired case attention is focused on the mean of the differences; in the unpaired case interest is in the difference of the means. Because the sample sizes in the unpaired case may be (and indeed usually are) different, the combined SE takes this into account and gives more weight to the larger sample size because this is likely to be more reliable. The pooled SD for the difference in means is calculated as follows: where SD1 and SD2 are the SDs in the two groups and n1 and n2 are the two sample sizes. The pooled SE for the difference in means is then as follows. This SE for the difference in means can now be used to calculate a confidence interval for the difference in means and to perform an unpaired t-test, as above. The pooled SD in the early goal-directed therapy trial example is: and the corresponding pooled SE is: The difference in mean arterial pressure between the early goal-directed and standard therapy groups is 14 mmHg, with a corresponding 95% confidence interval of 14 +- 1.96 x 2.41= (9.3, 18.7) mmHg. If there were no difference in the mean arterial pressures of patients randomized to early goal-directed and standard therapy then the difference in means would be close to 0. However, the confidence interval excludes this value and suggests that the true difference is likely to be between 9.3 and 18.7 mmHg. To explore the likely role of chance in explaining this difference, an unpaired t-test can be performed. The null hypothesis in this case is that the means in the two populations are the same or, in other words, that the difference in the means is 0. As for the previous two cases, a t statistic is calculated. A P value may be obtained by comparison with the t distribution on n1 + n2 - 2 degrees of freedom. Again, the larger the t statistic, the smaller the P value will be. In the early goal-directed therapy example t = 14/2.41 = 5.81, with a corresponding P value less than 0.0001. In other words, it is extremely unlikely that a difference in mean arterial pressure of this magnitude would be observed just by chance. This supports the notion that there may be a genuine difference between the two groups and, assuming that the randomization and conduct of the trial was appropriate, this suggests that early goal-directed therapy may be successful in raising mean arterial pressure by between 9.3 and 18.7 mmHg. As always, it is important to interpret this finding in the context of the study population and, in particular, to consider how readily the results may be generalized to the general population of patients with severe sepsis or septic shock. Assumptions and limitations : In common with other statistical tests, the t-tests presented here require that certain assumptions be made regarding the format of the data. The one sample t-test requires that the data have an approximately Normal distribution, whereas the paired t-test requires that the distribution of the differences are approximately Normal. The unpaired t-test relies on the assumption that the data from the two samples are both Normally distributed, and has the additional requirement that the SDs from the two samples are approximately equal. Formal statistical tests exist to examine whether a set of data are Normal or whether two SDs (or, equivalently, two variances) are equal , although results from these should always be interpreted in the context of the sample size and associated statistical power in the usual way. However, the t-test is known to be robust to modest departures from these assumptions, and so a more informal investigation of the data may often be sufficient in practice. If assumptions of Normality are violated, then appropriate transformation of the data (as outlined in Statistics review 1) may be used before performing any calculations. Similarly, transformations may also be useful if the SDs are very different in the unpaired case . However, it may not always be possible to get around these limitations; where this is the case, there are a series of alternative tests that can be used. Known as nonparametric tests, they require very few or very limited assumptions to be made about the format of the data, and can therefore be used in situations where classical methods, such as t-tests, may be inappropriate. These methods will be the subject of the next review, along with a discussion of the relative merits of parametric and nonpara-metric approaches. Finally, the methods presented here are restricted to the case where comparison is to be made between one or two groups. This is probably the most common situation in practice but it is by no means uncommon to want to explore differences in means across three or more groups, for example lung function in nonsmokers, current smokers and ex-smokers. This requires an alternative approach that is known as analysis of variance (ANOVA), and will be the subject of a future review. Competing interests : None declared. Abbreviations : ICU = intensive care unit; SD = standard deviation; SE = standard error. Backmatter: PMID- 12398779 TI - Recently published papers: New evidence for old debates, new drugs and some timely reminders AB - Keywords: : The present article provides a brief synopsis from the past 2 month's critical care literature. Sepsis : There has been a longstanding debate in the fields of sepsis and multiple organ dysfunction syndrome regarding the relative importance of microcirculatory failure versus mitochondrial failure. This debate stems from the observation that in the presence of supranormal oxygen delivery there is frequently a failure to achieve normal levels of consumption, and there is an association between the magnitude of this discrepancy and outcome. Accordingly, measures of global oxygen consumption, such as mixed (or central) venous oxygen saturations, plasma lactate and base deficit, are widely considered the best markers of the adequacy of resuscitation. Using a novel method of intravital microscopy of the sublingual mucosa, de Backer and colleagues have demonstrated that microcirculatory perfusion is reduced by ~50% despite normal global haemodynamic parameters, and that shunting by larger vessels is commonplace . This confirms previous animal and tissue studies, and despite the excellent pictures (see ) is unsurprising. Of note, the authors found that the degree of severity correlated with the outcome. What is remarkable, however, is the demonstration, using this intravital microscopy technique of the preservation of local vascular responsiveness to the vasodilatory mediator acetylcholine, and indeed the restoration of local microcirculatory function achieved by its administration. Whether this reflects a local imbalance between vasoconstrictors and dilators remains speculative but, of note, 18 of the 21 patients were receiving exogenous catecholamine vasopressors to maintain a target mean arterial pressure. Enthusiastic conclusions were drawn regarding the utility of this intravital microscopy technique in an accompanying editorial , and a proposal was made that semiquantitative measures of the sublingual microcirculation be added to the goals of the adequacy of resuscitation. Intravital holds considerable promise as a research tool but, like gastric tonometry, its universal application appears to lack foundation, for the present at least. In a complementary study, Brealey and colleagues report on their investigations of mitochondrial function in the muscles of patients with severe sepsis . They found convincing evidence of mitochondrial failure and found a strong correlation between the degree of failure and adverse outcome. What these two studies demonstrate is that both the microcirculatory failure and the mitochondrial failure are important pathophysiological processes in sepsis, and indeed it seems probable that the former precipitates the latter. What makes both of these studies noteworthy is that they represent some of the most eloquent demonstrations of these processes in critically ill patients to date. Let us hope that ongoing research identifies successful strategies to modify these processes. One such strategy may be the potentiation of the paracrine messenger adenosine , although previous successes in mice have all too often failed to translate to humans. Similarly, extrapolating the observation that a low serum granulocyte --macrophage colony-stimulating factor is prognostic of a poor outcome in sepsis to the efficacy of exogenous supplementation is an oft trodden path with a poor track record. Finally, the keenly awaited French 'steroid replacement' trial in sepsis has been published . The first important finding was a very high incidence of adrenal insufficiency in 229 of the 299 patients (77%) studied, albeit that interpreting corticotrophin studies in such patients is controversial. Interestingly, corticotrophin responsiveness did not predict outcome. Intensive care unit, hospital, 28-day and 1-year mortality were very high in all groups (53 --77%). In the responder groups, randomisation to the treatment group conferred no benefit. In the nonresponders, treatment conferred a statistically significant survival advantage to 28 days that persisted but ceased to be significant at 1 year (although the study was not powered to test 12-month mortality). There were no detectable adverse events associated with treatment. The regimen of intravenous hydrocortisone (50 mg, 6 hourly) plus daily enteral fluorocortisone (50 mug) is arguably too high, and no doubt future studies will address this. Annane et al. conclude by advocating universal corticotrophin testing followed by treatment, which should be withdrawn in responders. Cardiac failure : There has been growing interest in the natriuretic peptides for sometime. July saw the publication of a paper by Maisel et al. demonstrating the diagnostic potential of a bedside assay for B-type natriuretic peptide in the differential diagnosis of acute dyspnoea . That study and two studies from last year suggest that levels of these peptides correlate with the severity of cardiac failure but, as with troponins as markers of myocardial damage, caution in the interpretation of results is likely to be required . On the subject of troponin, it is encouraging to find a little light being shed on the difficulties of its meaningful interpretation. It has been observed that troponin levels are elevated in the presence of renal impairment, making interpretation and therefore diagnosis of acute coronary syndromes difficult in the critically ill. To investigate this, Aviles et al. studied whether troponin T levels were predictive of outcome, a composite of death and myocardial infarction within 30 days, in a large cohort of patients (n = 7033) with suspected acute coronary syndrome and both normal and reduced creatinine clearance. After adjustment for potential confounding factors, an abnormally elevated troponin T level (>0.1 ng/ml) was found to be predictive of myocardial infarction or death regardless of renal function. It is sobering to be reminded that the use of inodilators in the treatment of acute cardiac failure has never undergone a large-scale, randomised, controlled trial. However, with the advent of the novel calcium-sensitising agent levosimendan, one such trial has now been reported . This trial randomised 2003 patients with acute or acute on chronic heart failure to receive 24 hours of either dobutamine or levosimendan. A higher proportion of those patients who received levosimendan (29% versus 15%) achieved target improvements in haemodynamic parameters. More impressively, a lower proportion of these patients died over the following 6 months (26% versus 38%). Although cardiogenic shock and multiorgan failure were exclusion criteria in this trial, it surely provides evidence to justify similar studies in critically ill patients with low cardiac output. On a related point, a thought-provoking if methodologically suboptimal study of the significant association between the administration of dopamine and the incidence of delirium will hopefully result in a more thorough examination of the neuropsychiatric sequelae of all the commonly used doperminergic drugs. On a final cardiac note, the encouraging results of a trial of ibutilide as second-line therapy in the treatment of atrial fibrillation in the critically ill suggests that this novel class III anti-arrhythmic may provide a much needed treatment for this common condition, which all too often proves difficult to control. Respiratory : Attention to simple details is as important, if not more, than complex and novel interventions. A semirecumbant or upright posture has been repeatedly demonstrated to result in superior respiratory and overall outcome. Despite these demonstrations, however, the implementation of this intervention appears to be poor. Cook and colleagues have published a timely survey on this issue . They conclude that a classic case of system failure exists with regard to patient positioning, and they suggest a number of useful strategies to overcome this and similar problems. At the other extreme of patient positioning, the study by Lee et al. suggests that instigating the prone position early in the course of acute lung injury to patients with a larger shunt and a more compliant chest wall predicts success. Their study contributes further confirmatory evidence of the clinical benefit of prone positioning but does nothing to answer the vital questions regarding how long patients should be in the prone position for maximum benefit, and indeed whether improving gas exchange translates into better outcomes. Against the advancing wave of enthusiasm for face mask positive pressure ventilation (NIPPV), a significant negative study has been published . The authors randomised a heterogeneous group of patients who developed respiratory distress within 48 hours of extubation to receive standard therapy or standard therapy plus NIPPV. They found no benefit from this intervention. To their credit, the authors acknowledge that more experienced units or targeting specific patient groups might have achieved a positive outcome, but they conclude that they cannot recommend NIPPV as an intervention in this setting. We would argue from both personal experience and trial evidence that NIPPV in this setting can prevent the need to reintubate, and that a trial of NIPPV is sometimes warranted, not least as a failed trial is unlikely to harm the patient. Ventilator-associated pneumonia remains a major cause of morbidity and mortality. The optimal diagnostic technique is controversial, with some studies suggesting that invasive methods improve outcome while others have been unable to demonstrate any benefit from these expensive and time-consuming investigations. To further this debate, Wu and colleagues have compared quantitative culture of endotracheal aspirates, protected specimen brushings and bronchoalveolar lavage to investigate whether bacteriological findings correlated in patients with suspected ventilator-associated pneumonia . They demonstrated that the less invasive and less work-intensive technique of quantitative culture of endotracheal aspirates correlated well with both protected specimen brushings and bronchoalveolar lavage findings, allowing early definitive antibiotic choice or cessation of antibiotics in the case of negative results. Criticism of their investigation might include the small study number and lack of any outcome benefit demonstrated, although this was not the aim of the study. However, a significant impediment to this approach is the inability to obtain routine quantitative cultures in many hospitals. With mounting evidence in favour of quantitative assessment of specimens, this problem may need addressing. Magnesium is known to be a bronchodilator and has been anecdotally reported as having beneficial effects in acute severe asthma. However, these findings have not been consistently reproduced in well designed, randomised, controlled trials. In a welcome attempt to address this issue, Silverman and colleagues have conducted such a trial and found that the addition of intravenous magnesium to standard therapy resulted in a significant improvement in FEV1, at 4 hours, in those with the most severe airway obstruction (baseline FEV1 < 25% predicted). Given that magnesium is safe and inexpensive, and that there is now good evidence to support its routine use in the severest cases, it only remains for the optimal dose and duration of therapy to be established. Other noteworthy papers : Nasogastric feeding is found to be at least as good as, if not better than, nasojejunal feeding in the hands of Neumann and DeLegge , essentially because nasojejunal tube placement results in a significant delay in the start of feeding. Surely the pragmatic approach is to start with a nasogastric tube and to only employ the alternatives in cases of failure. The long-term success of renal allografts from nonheart-beating donors is reported by Weber and colleagues . Sadly, adoption of this technique, in the UK at least, awaits the deliberation of the government and judiciary. Finally, if you have been seduced by the attractive qualities of alcohol-based hand rubs to reduce crossinfection while avoiding hand washing, then be warned: your gel may not be all its cracked up to be . In addition to the aforementioned studies, we would also recommend the following commentaries and reviews . Competing interests : None declared. Abbreviations : FEV1 = forced expiratory volume in 1 second; NIPPV =face mask positive pressure ventilation. Backmatter: PMID- 0 TI - Medicine and the Internet: untangling the web AB - Keywords: decision support, Internet, medical informatics, decision support, Internet, medical informatics, Report : The Internet has ridden a rollercoaster of popularity since it became accessible to the general public in the early 1990s, from being viewed as a short-lived fad to being considered the ultimate solution to information and communication management. The Internet is clearly here to stay, and it now provides tangible support to healthcare providers in a number of roles. However, all of us, whether experienced users or novices, have felt overwhelmed, frustrated or entangled in this web. Medicine and the Internet is a comprehensive review of the history, technology and services of the Internet, and is a useful resource for the healthcare provider. This book is a third edition, the previous edition being published 5 years ago. Immeasurable changes have occurred since then, and the present edition utilizes a team of expert contributors to provide clear and accurate topic reviews. The information is completely up to date and, unlike many similar publications, the provided URL (uniform resource locators)1 are current and functional. The layout is easy to read, with section summaries and appropriate graphics. I found the use of information 'boxes' somewhat annoying in a few chapters, particularly when these were more prevalent than the text. The discussion is comprehensive and relevant to both the new as well as the experienced user. There are useful overviews of some technological aspects that may not interest all readers, but there is also practical advice on getting connected and using the web. The scope of the discussion is wide; sections on basic functions such as Internet service providers, e-mail, web browsers, newsgroups, chat rooms and Internet security provide a good background for the review of medical applications of the Internet. I found the clinical applications to be comprehensive and valuable. In fact, the discussion is far broader than the Internet, including summaries of related topics such as evidence-based medicine and medical ethics. The potential roles for the Internet in a clinical context are discussed in a task-oriented structure, and include accessing medical information, medical education, patient education, research, telemedicine, commerce and more. The emphasis is on British resources, but as a North American this opened up a vista of superb new websites to which I had not previously been exposed. Medicine and the Internet is an interesting read and an excellent resource with a good mix of technology and clinical focus. Although it is not directed at the intensivist, most of the book is relevant to our practice. The book may benefit from a parallel website to allow easy access to the many web resources listed. However, a brief Internet search discovered the author's website with all these links and more. Book details : McKenzie BC (ed): Medicine and the internet. Oxford: Oxford University Press, 2002. 296pp. ISBN 0-19-851063-2. Competing interests : None declared. Note : 1Definition, p. 288. Uniform resource locator: a standardized syntax used on the Internet describing the location and method of accessing Internet resources. Each uniform resource locator is composed of several elements: the type of Internet service, the domain name of the host, the port address, and the path name. Backmatter: PMID- 12398774 TI - Hemofiltration, adsorption, sieving and the challenge of sepsis therapy design AB - Abstract | Circulating inflammatory mediators spilling into the circulation from sites of active inflammation are considered the source of remote tissue injury and associated organ dysfunction in sepsis. Hemofiltration has been proposed as a therapy for sepsis based on its ability to remove circulating inflammatory mediators by sieving or by adsorption, or both. Designing devices and methods for sepsis therapy will require optimization of these two mechanisms. In the present issue of Critical Care Forum, Kellum and Dishart report the relative effects of sieving and adsorption on plasma IL-6 following cecal ligation and puncture in rats. The authors conclude that hemoadsorption is the main mechanism of removal, and discuss some possible implications for filter design but hemoadsorption is well dependant on hemofiltration (the so-called hemofiltration filter adsorption/synergistic effect). It is important to recognize the limitations of conventional systems; Kellum and Dishart have extended our knowledge of hemofiltration filter adsorption, which is quite different from conventional hemoadsorption. If sepsis is a manifestation of a nonlinear dynamic control system out of control, then filtration at modest doses with a large pore filter may succeed as well as high-volume hemofiltration with a conventional cut-off filter. In the present paper, we will explore the strengths and the weaknesses of the 'Kellum and Dishart' study and discussing their findings in the light of the current available literature. Keywords: adsorption, hemofiltration, membrane, sepsis, sieving, adsorption, hemofiltration, membrane, sepsis, sieving, Introduction : Circulating inflammatory mediators (IM) spilling into the circulation from sites of active inflammation are considered the source of remote tissue injury and associated organ dysfunction in sepsis. Hemofiltration has been proposed as a therapy for sepsis based on its ability to remove circulating IM by sieving or by adsorption, or both. Designing devices and methods for sepsis therapy will require optimization of these two mechanisms. In the present issue of Critical Care Forum, Kellum and Dishart report the relative effects of sieving and adsorption on plasma IL-6 following cecal ligation and puncture in rats . The authorsconclude that hemoadsorption is the main mechanism of removal, and discuss some possible implications for filter design. Hemoadsorption is dependent on membrane material and filtration operating parameters (e.g. filtration fraction: the so-called adsorption/synergistic effect). If hemofiltration is to be an effective therapy in the complexity of sepsis, then proper design of its material and operational characteristics must be pursued. Adsorption of proteins to membrane materials is well recognized in pharmaceutical manufacturing, food processing and medical filtration. The type and extent of feed solution proteins adsorbed depends on the membrane material, the pH, ionic strength and composition of the feed solution, the pore size, the membrane morphology and the presence of a polarization layer. Membrane materials vary in the extent and type of cytokines adsorbed. Data from in vitro studies reveal tumor necrosis factor adsorption of 30 --32% for polyamide and AN69, and of 0% for cellulose acetate and polysulfone. IL-1 adsorption was 40% for polyacrylonitrile, 0 --11% for polysulfone, 2% for AN69 and 0% for polyamide . Birk et al. found an approximately sixfold difference in total plasma protein adsorption between different membrane materials. Total protein adsorption was negatively correlated with the adsorption of proteins with molecular weight <65 kDa . The feed solution pH and ionic strength significantly affect adsorption and polarization, as shown in vitro. At pH4.8, the sieving coefficient (SC) for albumin and the SC for IgG are ~0.45 and 0, respectively. At pH7.4, however, these SCs are 0.38 and 0.85, a substantial reversal . At equal concentration, mixtures of IgG and albumin reduce filtrate flux and protein diffusivity more than pure solutions; this results from protein --protein interactions . Generally, these factors cannot be manipulated in clinical hemofiltration. Membranes with a higher molecular weight cut-off (MWCO) adsorb more protein than lower MWCO membranes. Uptake of radiolabeled albumin by a 100 kDa MWCO polysulfone membrane was nearly double that of a 30 kDa MWCO membrane . Protein uptake occurred preferentially in larger pores . This pattern of protein uptake has significant implications for molecular sieving. Molecular sieving in AN69 membranes was characterized using polydisperse dextran before and after blood contact. The SC for dextran of molecular weight <5 kDa was reduced by 14%, and the SC for 20 kDa dextran was reduced by 60% . Protein (including cytokine) adsorption and polarization of filtration membranes have been extensively studied. Awareness of these characteristics is essential in designing filtration therapy for sepsis. Some design elements relevant to adsorption are fixed for a given system (e.g. membrane materials, morphology and surface area), and some do not permit manipulation (e.g. patient plasma protein composition, pH and ionic strength). However, as membrane adsorption is rapidly saturated (30 to 50 minutes) . Recent recognition that the intensity of ultrafiltration needs to be adjusted for patient body size and severity of illness supports the need to focus on sieving and filtrate flow as promising points for new designs. The design of blood filtration in sepsis should focus on those characteristics of hemofiltration that permit greater effectiveness in controlling sepsis and that provide operational flexibility so the therapy may be tuned to patient body size and severity of illness. This process begins with recognition of key features of the inflammatory response. The network of IM, acting as a nonlinear dynamic control system, drives the inflammatory response . Network effects make the inflammatory response robust against large narrowly focused changes ; this robustness probably explains the failure of drugs directed against some single IM . Control networks may be manipulated by application of small changes in the activity to many network elements. The more elements (e.g. IM) affected, the smaller the change in their aggregate activity required for system control. Applied to blood filtration in sepsis, either high doses of filtration with a conventional filter, or lower doses with a large pore filter should be effective. In their study, Kellum and Dishart used an appropriate animal model (ceacal ligation and puncture) relevant in its initial insult and delay in treatment. According to the average body weight for adult, male Sprague-Dawley rats reported (486 g) and to the ultrafiltration flow rate(Quf) reported (30 ml/hour), we can conclude that a dose of ~62 ml/kg (which for a 75 kg human being represent 4.5 l/kg per hour) was delivered. This dose is greater than the highest dose used by Ronco et al. (45 ml/kg per hour) and clinically relevant as high-volume hemofiltration is usually defined to be greater than 50 ml/kg per hour . However when looking at literature, usual average body weight for adult, male Sprague-Dawley rats is about 580 g . In the study of Kellum and Dishart, Quf was not controlled or indexed to body weight. Recent demonstrations of dose-response effects of hemofiltration in human acute renal failure and sepsis make indexing Quf to body weight a critical parameter to assess or control. Filter blood flow was spontaneous and not quantified; however the high Quf suggest an high filtration fraction prevailed. Low filtration fraction promotes IM sieving, high filtration fraction promotes adsorption and reduces sieving of IM . What evidences support effective sieving of IM in sepsis? Honore et al. replaced 35 l of ultrafiltratein 4 hours (using high-volume hemofiltration) in 20 patients with refractory septic shock using a polysulfone membrane (Fresenius, MWCO = 35 kDa) . Predicted mortality for the group was 79%, and observed mortality was 55% (P < 0.05). Patients who responded (improved to specified end points) hemodynamically by the end of the 4 hours survived significantly more often (9/11) than those patients that did not respond (0/9 survived 24 hours). Retrospective analysis of the study of Honore et al. revealed that responders were smaller (66.2 +- 8.4 kg) than nonresponders (82.6 +- 13.4 kg) and therefore received a larger dose of filtration (0.53 +- 0.07 l/kg per 4 hours [+- 150 ml/kg per hour of hemofiltration clearance indexed to body weight and time] and 0.43 +- 0.07 l/kg per 4 hours [+- 110 ml/kg per hour of hemofiltration clearance indexed to body weight and time], respectively). Retrospective analysis of the study of Honore et al. suggests that a sufficiently high dose allows ~82% improvement in survival. The same protocol applied to all patients, thus adsorption should be similar in all patients; adsorption should be saturated by 30 to 50 minutes . Thus the dominant mechanism of IM removal should be sieving. Survival was not assess in the study of Kellum and Dishart . A large pore filter (polysulfone, MWCO = 100 kDa) has been used in a swine model of lethal sepsis . In a paired study with a similar conventional filter (MWCO = 50 kDa), and using identical operating parameters (e.g. equal filtration fraction) the 100 kDa filter was associated with a survival time nearly twice that of the 50 kDa filter. A similar filter (polyamide, MWCO = 100 kDa) has been studied in vitro by Uchino et al. using recirculating human blood . The blood was spiked with endotoxin to raise a cytokine response. Selected results are compared in Table with conventional filters . Table 1 | Sieving coefficients of large pore membranes and conventional pore membranes The 100 kDa membrane has two significant advantages. First, for IM sieved by conventional and the 100 kDa membrane, the 100 kDa exhibits higher SC. Second, the 100 kDa sieves cytokine not sieved by conventional membrane. If sepsis is a manifestation of a nonlinear dynamic control system operating at an excessive and injurious level, then filtration at modest doses with a large pore filter may succeed; high-volume hemofiltration with a conventional MWCO should also be effective. A successful blood filtration therapy for sepsis and septic shock will not be found by accident -- it will be designed. It is important to recognize the limitations of conventional systems. Kellum and Dishart have shown that hemofiltration filter adsorption occurs and exhibits meaningful biologic effects. The design of successful blood filtration therapy in sepsis will require recognition of the limitations of existing systems and methods. The Kellum and Dishart study aids this recognition. Recognition that IM are not operative as single agents, but are closely integrated in a self-regulated control network is of key importance to design of therapeutic systems. A blood filtration system which filters a sufficiently wide of IM to be effective in sepsis, and has the operation flexibility to readily adapt to patients of different body size and severity of illness will require careful design. In matters of membrane separation and system control, our engineering colleagues have much to offer. Partnering with industry and engineering should allow new devices and methods to be developed and tested. This should be done before we embark upon a large scale multicentre study . By viewing the whole problem, we can work out the whole solution. Competing interests : None declared. Abbreviations : IL = interleukin; IM = inflammatory mediators; MWCO = molecular weight cut-off; SC = sieving coefficient. Backmatter: PMID- 12202756 TI - Antiproliferative effect in chronic myeloid leukaemia cells by antisense peptide nucleic acids AB - Peptide nucleic acid (PNA) is a synthetic DNA analogue that is resistant to nucleases and proteases and binds with exceptional affinity to RNA. Because of these properties PNA has the potential to become a powerful therapeutic agent to be used in vivo. Until now, however, the use of PNA in vivo has not been much investigated. Here, we have attempted to reduce the expression of the bcr /abl oncogene in chronic myeloid leukaemia KYO-1 cells using a 13mer PNA sequence (asPNA) designed to hybridise to the b2a2 junction of bcr /abl mRNA. To enhance cellular uptake asPNA was covalently linked to the basic peptide VKRKKKP (NLS-asPNA). Moreover, to investigate the cellular uptake by confocal microscopy, both PNAs were linked by their N-terminus to fluorescein (FL). Studies of uptake, carried out at 4 and 37C on living KYO-1 cells stained with hexidium iodide, showed that both NLS-asPNA-FL and asPNA-FL were taken up by the cells, through a receptor-independent mechanism. The intracellular amount of NLS-asPNA-FL was about two to three times higher than that of asPNA-FL. Using a semi-quantitative RT -- PCR technique we found that 10 microM asPNA and NLS-asPNA reduced the level of b2a2 mRNA in KYO-1 cells to 20 +- 5% and 60 +- 10% of the control, respectively. Western blot analysis showed that asPNA promoted a significant inhibition of p210BCR /ABL protein: residual protein measured in cells exposed for 48 h to asPNA was similar35% of the control. Additionally, asPNA impaired cell growth to 50 +- 5% of the control and inhibited completion of the cell cycle. In summary, these results demonstrate that a PNA 13mer is taken up by KYO-1 cells and is capable of producing a significant and specific down-regulation of the bcr /abl oncogene involved in leukaemogenesis. Keywords: INTRODUCTION : Over the past few years, peptide nucleic acids (PNAs) have emerged as one of the most promising new types of therapeutic oligonucleotides in the form of antisense and anti-gene molecules . From the chemical point of view, PNA is both a peptide and a nucleic acid, since it is a hybrid molecule consisting of N-(2-aminoethyl)glycine monomers linked by amide bonds (,). The purine (A and G) and pyrimidine (C and T) nucleobases are attached to this polyamide backbone via a methylene carbonyl linkage. As PNAs are not charged at physiological pH, they bind to complementary nucleic acid strands with higher affinity than do natural nucleic acids . Although hybridisation of PNAs to complementary DNA and RNA obeys the Watson --Crick rules, both parallel and antiparallel duplexes can be formed. The N-terminus corresponds to the 5' end and the C-terminus to the 3' end of a normal oligonucleotide ( --). However, antiparallel PNA --DNA and PNA --RNA duplexes are more stable and form more rapidly than the corresponding parallel duplexes . The PNA sequences used in this study are shown in Figure . They are designed to reduce the expression of the bcr /abl oncogene in chronic myelogenous leukaemia (CML) cells ( --). In most CML patients, leukaemic cells contain the Philadelphia (Ph) chromosome, which derives from a reciprocal translocation of chromosomes 9q and 22q. Such a translocation generates a fused bcr /abl oncogene (with either a b2a2 or b3a2 junction) encoding a protein of 210 kDa (p210BCR /ABL) which is involved in leukaemogenesis (,). Protein p210BCR /ABL has a higher tyrosine phosphokinase activity than the normal ABL protein and its structure allows multiple protein --protein interactions that link p210BCR /ABL to several intracellular pathways, including the ras pathway, which regulates cell proliferation and differentiation ( --). The presence of a unique sequence at b2a2 and b3a2 mRNA junctions and the observation that CML cells require p210BCR /ABL to survive make the bcr /abl gene an attractive target for antisense molecules. Attempts to reduce the intracellular level of bcr /abl mRNA in different cell lines by means of antisense oligonucleotides have been reported . However, the results have not been very satisfactory, because normal and thioate antisense oligonucleotides displayed a number of limits, including nuclease degradation, non-specific effects and insufficient target affinity. PNAs are not susceptible to degradation by endogenous nucleases and proteases and bind to RNA targets with exceptionally high affinity and in a sequence-specific manner. Because of these favourable properties, PNAs are potentially more powerful antisense agents than phosphorothioate oligonucleotides. To date, only a few reports have shown that PNAs, either free or conjugated, produce an antisense or antigene effect in vivo ( --). In the study presented herein we have explored the capacity of 13mer antisense PNAs to inhibit the expression of bcr /abl in leukaemic KYO-1 cells. The PNAs were designed to bind to the b2a2 junction of bcr /abl mRNA and thus block its translation sterically. We found that the designed b2a2 antisense PNAs: (i) are taken up by KYO-1 cells; (ii) specifically reduce the level of p210BCR /ABL; (iii) inhibit completion of the cell cycle; (iv) impair cell growth to 50% of the control. Figure 1 | Sequence of the b2a2 junction of bcr /abl mRNA expressed in KYO-1 cells. Sequence of the b2a2 junction of bcr /abl mRNA expressed in KYO-1 cells. The structures of the antisense and control PNAs used in this study are also shown. MATERIALS AND METHODS : PNA synthesis | The PNAs used in this study were made by solid phase synthesis using a peptide synthesiser (Applied Biosystems 433A peptide synthesiser) and PNA synthesiser (Perspective Biosystems Expedite Nucleic Acid Synthesis System). All solvents (Biosolve) were used as received. The solid phase syntheses were performed on PEG-PS beads as solid support with rink-amide as linker (loading 0.17 micromol /mg). For PNAs conjugated to NLS, first the peptide part was assembled on the peptide synthesiser on the 20 micromol scale using protected amino acids [5 equivalents, Fmoc-Val-OH, Fmoc-Lys(Boc)-OH, Fmoc-Arg(Pbf)-OH and Fmoc-Pro-OH] under the agency of BOP /HOBt as the coupling agent. Only the first coupling, that of valine, was double, and after the synthesis was complete the final Fmoc was left on. The loading was determined by deprotection of a small amount of the resin and measuring the absorbance of the liberated Fmoc protecting group (loading 0.13 micromol /mg, efficiency 98%). Assembly of the PNAs was established using the standard protocol described in the synthesiser manual, on the 2 micromol scale using Fmoc chemistry and monomers in which the exocyclic amines were Bhoc protected (Perspective Biosystems). As solid support either PRG-PS beads with rink-amide as the linker (loading 0.17 micromol /mg) or NLS-functionalised resin (loading 0.13 micromol /mg) was used. After removal of the final Fmoc protecting group the oligomer was either cleaved from the resin and deprotected under acidic conditions (TFA /TIS / H2O, 9:1:1 v /v /v) or first coupled twice with the O-linker [FmocHN(CH2)2O(CH2)OCH2CO2H], then reacted with 5(6)-carboxyfluorescein succinimide ester (Molecular Probes) and subsequently cleaved with concomitant deprotection. RP-HPLC purification and analysis were carried out on a JASCO HPLC system equipped with an Altima C18 column (10 x 250 mm). Gradient elution was performed at 40C by building up a gradient starting with buffer A (0.1% TFA in water) and applying buffer B (0.1% TFA in acetonitrile /water, 3:1 v /v) with a flow rate of 4 ml /min. The PNAs obtained were lyophilised and characterised by matrix-assisted laser desorption /ionisation time-of-flight mass spectrometry (MALDI-TOF MS) and RP-HPLC. Cell cultures | CML cell lines K562 (a gift of Dr M. Giunta, Bone Marrow Transplantation Unit, Medical School, University of Udine, Italy) and KYO-1 (a gift of Dr C. M. Broughton, Department of Hematology, University of Liverpool, UK) were maintained in exponential growth in RPMI 1640 medium containing 100 U /ml penicillin, 100 mg /ml streptomycin, 200 mM l-glutamine and 10% fetal bovine serum (FBS) (Celbio, Milan, Italy), which was heat inactivated at 56C for 30 min. Cell cycle analysis | Cell cycle analysis was performed on KYO-1 cells growing in the presence of 10 microM PNA for 30 h. Briefly, KYO-1 cells were seeded in a 24-well plate at a cell density of 105 cells /ml, 24 h before PNA treatment. Cells (2 x 105) were harvested (240 g, 4C), resuspended in 0.5 ml of PBS and washed twice. While vortexing, 1 ml of ice-cold 70% ethanol was added dropwise to the pellet and the cells were allowed to fix overnight at 4C. Prior to FACS analysis (Becton Dickinson) the cells were stained with a solution containing 0.05 mg /ml propidium iodide and 0.5 mg /ml RNase A. The cells were centrifuged and the supernatant poured off. The cells were vortexed in the residual ethanol and then 0.5 ml of propidium iodide was added to each sample tube. The cells were incubated for 30 min at 37C and then analysed by FACS. Proliferation assay in liquid culture | Aliquots of 2 x 104 exponentially growing KYO-1 cells in liquid culture were seeded in 200 microl of complete medium in 96-well microtitre plates and treated with the PNAs. The PNAs were added to the medium 24 h after seeding. Cellular growth was measured at 24, 48 and 72 h using 0.4 mg /ml 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) . The absorbance was read at 540 and 690 nm using a microtitre plate reader (Titertek Multiscan MCC; Labsystems). The data are expressed in terms of T /C (OD of PNA-treated cells /OD of control cells) as a measure of cell viability and survival in the presence of test PNAs. Confocal microscopy | KYO-1 cells were seeded in 12-well plates at a density of 0.8 x 105 cells in 600 microl of RPMI medium supplemented with 10% heat-inactivated FBS. After 24 h the cells were exposed for a certain time to fluorescein-labeled non-conjugated PNA (asPNA-FL) and PNA conjugated to the basic peptide VKRKKKP (NLS-asPNA-NLS). The cells were centrifuged, washed twice with PBS, spun on a glass slide and fixed with 3% paraformaldehyde (PFA) in PBS for 20 min. After washing with 0.1 M glycine, containing 0.02% sodium azide in PBS, to remove PFA, and Triton X-100 (0.1% in PBS), the cells were incubated with 24 microg /ml propidium iodide and 0.4 mg /ml RNase A for 30 min at 37C in order to stain the nuclei. Then coverslips were mounted on the glass slide with mowiol 4-88 and DABCO (2.5%). The cells were analysed using a Leica DM IRBE confocal imaging system. Diaphragm and fluorescence detection levels were adjusted to minimise any interference between the fluorescein and propidium iodide channels. Hexidium iodide (Molecular Probes) was dissolved in DMSO and 10 mM stock solutions were stored at --20C. RNA purification and reverse transcription --polymerase chain reaction (RT --PCR) analysis | RNA purification. Total cellular RNA was extracted from K562 and KYO-1 cells using RNAqueous-4 PCR (Ambion). This method is based on cell lysis in a solution containing guanidinium thiocyanate. The RNA precipitate was treated with ultrapure DNase I to remove DNA traces and the RNA solutions were stored at --80C. cDNA synthesis. A volume of 5 microl of RNA solution in diethylpyrocarbonate (DEPC)-treated water was heated at 70C and placed in ice. The solution was added to 16.6 microl of cDNA mix [10 microl of 5x buffer, 5 microl of 0.1 M DTT (Life Technologies), 2 microl of a 12.5 pmol primer AZ (5'-CCA TTTTTGGTTTGGGCTTCA-3'; MWG Biotech), 2 microl of a 5 mM solution containing equimolar amounts of dATP, dCTP, dGTP and dTTP (Amersham Pharmacia Biotech), 0.6 microl of 40 U /ml RNase inhibitor (Life Technologies) and 1 microl of 200 U M-MLV reverse transcriptase (Life Technologies)]. The reactions (50 microl /tube) were incubated for 1 h at 37C. As a negative control the reverse transcription reaction was performed with 5 microl of DEPC-treated water. The cDNA was stored at --20C. Polymerase chain reaction. A volume of 10 microl of cDNA, heated at 95C for 5 min, was mixed with 16.5 microl of BCR /ABL mix [10 microl of 10x Taq polymerase buffer (Boehringer), 3 microl of 25 mM MgCl2, 2 microl of 12.5 pmol /microl primer EA500 (5'-TGTGATTATAGCCTAAGACCC-3'; MWG Biotech), 2 microl of 12.5 pmol /microl primer EA122 (5'-GTTTCAG AAGCTTCTCCCTG-3'; MWG-Biotech), 2 microl of 5 mM dNTPs with equimolar amounts of dTTP, dCTP, dATP and dGTP (Amersham Pharmacia Biotech), 0.5 microl of 2.5 U /ml AmpliTaq DNA polymerase (Perkin Elmer)]. Amplification was carried out on an automated DNA thermal cycler (Progene) as follows: 35 cycles of denaturation (94C for 30 s), annealing (60C for 30 s) and extension (72C for 30 s). Abl amplification (annealing at 55C) was carried out with primers EA500 and ABL1A (5'-CCTCTCGCTGGA CCCAGTGA). Two positive cDNA controls were used. First, RNA extracted only from K562 cells was used to amplify the b3a2bcr /abl mRNA. Secondly, the RNA extracted only from KYO-1 cells was used to amplify the b2a2bcr /abl mRNA. Moreover, as a negative control we tried to amplify the controls of the RNA extraction and the control of the cDNA reaction. Western blots | Aliquots of 2 x 105 KYO-1 cells were lysed in 2x Laemmli sample buffer (3.3% SDS, 22% glycerol, 1.1 M Tris --HCl, pH 6, 0.001% bromophenol blue, 10% beta-mercaptoethanol). Samples were heated at 95C for 10 min and loaded (70 microl) in either 7 or 12% resolving, 5% stacking SDS --polyacrylamide gel. Electrophoresis was carried out at 100 and 40 V for 45 min and 18 h, respectively. The relative concentration of the protein lysates was estimated electrophoretically and equal amounts of lysates were transferred to a nitrocellulose membrane (Sartorius) using a Multiphor II Novablot Transfer Unit (Amersham Pharmacia Biotech). The membranes were then cut at the 45 and 210 kDa molecular weight levels. Both membranes were used to detect the BCR /ABL and beta-actin. A double antibody procedure was used to detect the proteins. The membranes were incubated under agitation for 1 h at room temperature in blocking solution (PBS containing 5% dry milk and 0.1% Tween 20), then at room temperature for 2 h in c-ABL (Ab-3) (1:50) and beta-actin monoclonal antibodies (1:20 000) (Calbiochem). The antibody solution was removed and the membrane was washed three times with PBS containing 0.1% Tween 20. The membranes were then incubated (1 h at room temperature under agitation) with horseradish peroxidase-conjugated goat anti-mouse IgG (1:1000) (Sigma) for the Ab-3 and goat anti-mouse IgM (1:2000) (Calbiochem) for the beta-actin antibody. Chemiluminescence was detected immediately as described by the manufacturer (Super Signal West Pico Trial; Pierce). Films were exposed for similar15 min for BCR /ABL and 1 min for beta-actin. RESULTS : Target and antisense PNA molecules | Ph-positive KYO-1 cells express a bcr /abl mRNA which is characterised by a bcr exon 2 /abl exon 2 (b2a2) junction. As this junction is a unique sequence in CML cells, it has been chosen as the target for the antisense PNA molecules used in this study . According to previous observations, free PNAs do not penetrate or hardly penetrate cell membranes in the absence of a specific carrier (,). So, to enhance cellular uptake, the designed antisense PNA was covalently bound at the C-terminus (3') to a short peptide rich in basic amino acids, having the sequence of the SV40 nuclear localisation signal peptide (NLS) (HO2C-VKRKKKP-NH2). Moreover, to perform uptake studies by flow cytometry and confocal microscopy, both free and NLS-conjugated PNAs were covalently linked at their N-terminus (5') to 5(6)-carboxyfluorescein via an amide bond. The PNA 13mers shown in Figure , designed with a sequence complementary to the bcr /abl mRNA b2a2 junction, were assembled following a well-established solid phase protocol . Sequential elongation of the rink-amide-functionalised solid support, either with or without the NLS peptide present, using Fmoc /Bhoc-protected monomers under the agency of the coupling reagent HATU, gave the immobilised constructs. At this stage, the label could be introduced by reacting the deprotected N-terminus with the succinimidyl ester of 5(6)-carboxyfluorescein. Removal of the N-Bhoc protecting groups and concomitant release of the support was affected under acidic conditions, to give, after RP-HPLC purification, the PNA conjugates. The homogeneity and identity of these constructs were established by RP-HPLC and MALDI-TOF MS. Antisense PNA oligomers were synthesised with (NLS-asPNA) and without (asPNA) NLS peptide and both of them also with fluorescein (asPNA-FL and NLS-asPNA-FL). As a control we designed parPNA, a PNA with an inverted sequence compared to asPNA, and two PNAs with scrambled sequences, scrPNA-1 and scrPNA-2 . PNA uptake: confocal microscopy and FACS | It has recently been reported that unmodified, carrier-free PNA injected into rats either directly into the brain or i.p. was able to enter neuronal cells and specifically reduce gene expression . Other recent studies show that unmodified antisense PNA molecules are taken up by protozoa (Entameba hystolitica) and by neurons in culture . These data suggest that a certain amount of PNA is intrinsically taken up by the cells, in contrast to the proposal that PNAs do not pass or hardly pass through the membranes (, --). Thus, we wished to investigate the uptake of free asPNA-FL and peptide-conjugated NLS-asPNA-FL by KYO-1 cells using confocal microscopy. Two cell staining procedures were followed. In one case, after incubation with free asPNA-FL or NLS-asPNA-FL, the cells were fixed on a glass slide and exposed to 0.1% Triton X-100 and propidium iodide. As Triton X-100 permeabilises the cell membrane, it is possible that some of the PNA taken up by the cells is lost during the washing steps. In order to avoid this we also stained the cells directly, i.e. without fixing them on a glass plate, with 3 microM hexidium iodide, a dye that binds to DNA and RNA in living cells and displays a red fluorescence . KYO-1 cells were treated for 2 h with hexidium iodide, transferred to growth medium containing either free asPNA-FL or NLS-asPNA-FL for a certain time and then directly analysed with a confocal microscope. Figures A and B and show images obtained by both staining methods. PNA-treated KYO-1 cells fixed on glass and stained with propidium iodide are shown in Figure A. Figure A shows the nuclei of KYO-1 cells stained in red by propidium iodide, while Figure B shows the green fluorescent light emitted by asPNA-FL (incubation time 24 h). It is clear from these pictures that asPNA-FL is taken up by the cells. A superimposed view of Figure A and B is reported in Figure C. This image clearly shows that asPNA-FL is partly located in the nuclei and partly in the cytoplasm (green crown surrounding the nucleus, Fig. D). The intracellular localisation of asPNA-FL was also confirmed by xz-planes taken at the level of the arrow (Fig. E). The uptake of NLS-asPNA-FL by KYO-1 cells is shown in Figure B. NLS-asPNA-FL is localised in both the cytoplasm and nucleus, where it appears mainly confined to the nucleoli. Figure shows the PNA uptake in living KYO-1 cells stained with hexidium iodide. The images show in an unambiguous way that both asPNA-FL and NLS-asPNA-FL are taken up by the cells. To investigate the mechanism by which PNA is internalised in KYO-1 cells, experiments were performed at 4C. Basically, the cells were first stained with hexidium iodide, cooled to 4C over 1 h and then exposed to 10 microM asPNA-FL or NLS-asPNA-FL for a further 5 h at 4C. After incubation the cells were immediately analysed with a confocal microscope. We observed that both free and NLS-conjugated PNAs were also detected in the cells at low temperature (not shown). An estimate of the fluorescence intensity associated with the KYO-1 cells treated with asPNA-FL and NLS-asPNA-FL was obtained by flow cytometry. The average fluorescence intensity associated with the cells treated with 10 microM asPNA-NLS-FL was 210 +- 20 units, while the fluorescence of the cells treated with asPNA-FL was about three times lower, i.e. 70 +- 15 units. These values did not significantly change over incubation periods up to 40 h. The designed antisense PNA binds the b2a2 mRNA target in vitro | In order to demonstrate that the designed antisense PNAs stably hybridise to their natural target, i.e. the b2a2 mRNA junction, asPNA, asPNA-NLS, parPNA and scrPNA were incubated separately with total RNA extracted from KYO-1 cells, at concentrations increasing from 1 to 50 nM. Both asPNA and NLS-asPNA are complementary to a 13mer stretch of the b2a2 junction of bcr /abl mRNA, with which they can form the antiparallel PNA --RNA heteroduplex shown in Scheme , which could sterically block reverse transcription. In contrast, parPNA is expected to hybridize to the b2a2 junction with a parallel orientation, forming a PNA --DNA duplex which is expected to be less stable than that formed by asPNA (,). After RT --PCR, carried out using the primers EA122 and EA500, the reaction products were separated by polyacrylamide electrophoresis . It can be seen that asPNA inhibits reverse transcription in a dose-dependent manner, completely blocking the enzymatic activity at a concentration as low as 50 nM. Surprisingly, parPNA too showed a remarkable ability to inhibit reverse transcription, under the experimental conditions adopted. In contrast, samples incubated either in the absence of antisense PNA or the presence of srcPNA (a PNA with a scrambled sequence) did not show reverse transcription inhibition. The efficiency of reverse transcriptase was checked by amplifying a 128 nt stretch of the abl gene, using primers ABL1A and EA500, in each reaction tube. It can be seen that in all samples the expected abl fragment was obtained. These results demonstrate that the designed antisense PNAs are able to bind at nanomolar concentrations to the b2a2 mRNA junction with high affinity and completely block the catalytic activity of reverse transcriptase. We also performed RT --PCR experiments with NLS-asPNA and have observed that it produces an inhibitory effect roughly similar to that obtained with the unconjugated asPNA (not shown). The thermodynamic stability of parallel and antiparallel PNA --RNA hybrids was predicted using the nearest neighbour stacking energy parameters reported by Santa Lucia et al. , which allows Tm calculation for DNA --DNA duplexes. Taking into account that PNA --DNA duplexes are more stable than the corresponding DNA --DNA duplexes by similar1.0 K /base and that PNA --DNA duplexes are less stable than PNA --RNA by similar4C , we calculated for the antiparallel asPNA --b2a2 mRNA heteroduplex a Tm of 68C at 1 microM, 52C at 10 nM and 42C at 10 pM. As for the parallel heteroduplex formed by parPNA with the b2a2 target, its stability should be comparable to that of the corresponding DNA --DNA duplex, i.e. 52C at 1 microM, 37C at 10 nM and 26C at 10 pM (,). Antisense PNA binds to the b2a2 junction in vivo and reduces bcr /abl mRNA | In contrast to phosphate and phosphorothioate oligonucleotides, antisense PNAs do not induce degradation of PNA --RNA complexes by RNase H . Despite this, it has been found that antisense PNA depressed the amount of mRNA coding for prepro-oxytocin in neurons . We therefore measured the relative level of bcr /abl mRNA from PNA-treated KYO-1 cells following a competitive PCR strategy as previously described (,). The total RNA extracted from a nearly 1:1 mixture of PNA-treated KYO-1 cells (containing the target b2a2 junction) and untreated K562 cells (containing a b3a2 junction which is not recognised by the PNA, competitor) was transformed into cDNA and amplified by PCR. Target and competitor mRNA share the primer recognition sites, but differ in size by 74 bp. Running the RT --PCR products in a polyacryalmide gel, two bands were obtained: one of 388 bp from K562, the other of 314 bp from KYO-1 cells. The ratio between b2a2 mRNA from PNA-treated KYO-1 cells (314 bp) and b3a2 mRNA from untreated K562 cells (388 bp) provides an estimate of the relative amount of target mRNA compared to the control mRNA from untreated cells. In the presence of 10 microM asPNA, the b2a2 /b3a2 ratio was reduced to 20 +- 5% of the control, while 10 microM NLS-asPNA reduced the b2a2 /b3a2 ratio to only 60 +- 10% of the control . The efficiency of the RT reaction in each reaction tube was checked by amplifying a 128 bp fragment of abl. Depression of p210BCR /ABL level by antisense PNA | The capacity of antisense PNA to reduce the expression of bcr /abl in KYO-1 cells was measured by immunoblotting. KYO-1 cells were cultured in the absence and presence of antisense or control PNAs for 24, 48 and 72 h. After incubation, the amount of protein in untreated and PNA-treated KYO-1 cells was determined by western blot using an anti-BCR /ABL monoclonal antibody. To demonstrate gene specificity, the level of beta-actin was also measured in all the samples. Figure shows typical western blots at 48 h. It can be noted that after 48 h incubation with 10 microM asPNA the level of p210BCR /ABL protein is reduced to similar35% compared to that in untreated cells. Western blots were also carried out at 24 h and the level of p210BCR /ABL compared to that of untreated cells is reported in the enclosed histograms. It can be seen that the maximum reduction of p210BCR /ABL level is observed in the cells exposed to asPNA for 48 h. The data show that 10 microM NLS-asPNA induces at 48 h an inhibition of bcr /abl expression much lower than that observed with asPNA. An antisense effect was also observed with parPNA, as this molecule forms a parallel hybrid with the mRNA target . However, as parPNA has a lower affinity for bcr /abl mRNA, it promotes a weak inhibitory effect at the times considered: residual p210BCR /ABL is 73 and 65% at 24 and 48 h, respectively. We also observed that exposing the cells to the PNAs for 72 h resulted in a non-significant reduction of the level of p210BCR /ABL (not shown). Antiproliferative effect of antisense PNA | The effect on cell growth of the anti-b2a2 PNAs was investigated by the MTT assay. In a first set of experiments cell growth was measured at increasing PNA concentrations. Control and antisense PNAs were added to the culture medium at 5, 7.5 or 10 microM and the cells were grown for 72 h. The ratio T /C between treated (T) and untreated (C) cells as a function of PNA concentration is reported in Figure (top). It can be seen that asPNA causes a strong antiproliferative effect compared to untreated cells (26, 28 and 50% at 5, 7.5 and 10 microM, respectively) after an incubation time of 72 h. NLS-asPNA was found to promote a weaker cell growth inhibition (21, 29 and 28%) at the same concentrations. In contrast, control molecules scrPNA, NLS-scrPNA and parPNA showed a weak inhibitory effect of similar5 --10% at 10 microM. The effect on cell growth as a function of time of 10 microM PNAs is shown in Figure (bottom). It can be seen, in accordance with the data in Figure (top), that the cell growth inhibition promoted by asPNA reaches a level of 50% at 72 h. After 72 h incubation, NLS-asPNA inhibits cell growth by similar25%, while parPNA produced a smaller effect on cell growth. Cell cycle analysis of KYO-1 cells grown in the presence of antisense PNA | To further investigate the biological consequences of asPNA in KYO-1 cells, the cell cycle was assessed by FACS analysis, using propidium iodide staining of the DNA content. KYO-1 cells were grown in liquid culture for different times in the presence of 10 microM asPNA or control PNAs. Figure shows that asPNA, the compound with the highest antisense activity, produced a significant effect on the cell cycle. After 30 h growth, untreated cells or cells treated with parPNA or NLS-asPNA exhibit a DNA content between 2N (similar47 +- 1% of cells) and 4N (12 +- 1% of cells), suggesting that the S phase is active (26 +- 2% cells). In contrast, the cell cycle profile of the cells treated with asPNA shows a reduction in the G2 /M peak from 12 +- 1% to 3.7 +- 0.5%, suggesting that in the presence of asPNA completion of the cell cycle appears inhibited. According to the cell cycle analysis, a fraction (25 +- 4%) of KYO-1 cells treated with asPNA underwent apoptosis. Thus, it is possible that the inhibition of cell proliferation observed with the MTT assay results from the fact that asPNA induces apoptosis. Figure 2 | Confocal microscopy images of PNA-treated KYO-1 cells fixed on glass slides and stained with propidium iodide. Confocal microscopy images of PNA-treated KYO-1 cells fixed on glass slides and stained with propidium iodide. (A --D) Cells incubated for 24 h with 10 microM asPNA-FL. (E) xz-Planes taken at the level of the arrow. (F --H) Uptake of NLS-asPNA-FL in KYO-1 cells stained with propidium iodide. (I) xz-Sections taken at the level of the arrow. Experiments were performed at 37C. Bar 5 microm. Figure 3 | Confocal microscopy of living KYO-1 cells treated for 12 h with asPNA-FL or NLS-asPNA-FL and stained with the vital dye hexidium iodide. Confocal microscopy of living KYO-1 cells treated for 12 h with asPNA-FL or NLS-asPNA-FL and stained with the vital dye hexidium iodide. (Top) Uptake at 37C by KYO-1 cells treated with 10 microM asPNA-FL. (Bottom) Uptake by KYO-1 cells treated with 10 microM NLS-asPNA-FL. Bar 5 microm. Scheme 1. Scheme 1. Figure 4 | In vitro inhibition of reverse transcription by PNA. In vitro inhibition of reverse transcription by PNA. Total RNA extracted from KYO-1 (b2a2 junction) cells was incubated overnight with control or antisense PNAs at increasing concentrations (0 --50 nM). The RNA samples were subjected to RT --PCR, using bcr /abl primers EA122 and EA500, and the products separated in a 12% polyacrylamide gel. The cDNA amplification with primers EA122 and EA500 provides a 314 bp band. RT --PCR is inhibited by asPNA and parPNA, which bind to RNA and block reverse transcription, but not srcPNA, which does not bind to RNA. As a control, the amplification of a 128 bp DNA fragment of the abl gene was amplified. Figure 5 | Effect of b2a2 antisense PNA on the level of bcr /abl mRNA in KYO-1 cells quantified by RT --PCR. Effect of b2a2 antisense PNA on the level of bcr /abl mRNA in KYO-1 cells quantified by RT --PCR. The RNA from an equal number of PNA-untreated K562 (competitor) and PNA-treated KYO-1 cells was extracted. The extracted RNA was subjected to RT --PCR. Using primers EA122 and EA500, two amplified bcr /abl bands of 388 and 314 bp were obtained from K562 (b3a2 junction) and KYO-1 cells (b2a2 junction), respectively. When the KYO-1 cells were treated with asPNA or NLS-asPNA, the intensity of the 314 bp band, but not that of the 388 bp band, was reduced, compared to the band obtained by treating KYO-1 cells with scrPNA, a control PNA which is unable to bind to bcr /abl mRNA. As a control we also amplified a 128 bp DNA fragment of abl in each reaction tube, using primers ABL1A and EA500. As expected, this amplification is not influenced by PNA. Figure 6 | Immunoblot of KYO-1 cell lysates using anti-ABL and anti-beta-actin monoclonal antibodies. Immunoblot of KYO-1 cell lysates using anti-ABL and anti-beta-actin monoclonal antibodies. KYO-1 cells were incubated with 10 microM anti sense or control PNAs and the amount of protein p210BCR /ABL was evaluated by western blot analysis, after the cells were exposed to the PNAs for 24 and 48 h. The level of beta-actin in the PNA-treated cells was also measured. A typical western blot analysis of cell lysates at 48 h after PNA treatment is shown in the figure. The levels of p210BCR /ABL in lysates obtained from KYO-1 cells treated with antisense and control PNAs at 24 and 48 h are shown in the enclosed histograms. The ordinate reports the residual BCR /ABL protein expressed as percent T /C, where T is the p210BCR /ABL /beta-actin ratio of PNA-treated cells and C is the p210BCR /ABL /beta-actin ratio of PNA-untreated cells. The uncertainty on each value is at most 20%. Figure 7 | Antisense PNA inhibits the proliferation of KYO-1 cells in liquid culture. Antisense PNA inhibits the proliferation of KYO-1 cells in liquid culture. (Top) Cell growth in the presence of increasing PNA concentrations. The cells were grown for 72 h in the presence of the indicated PNAs, before a standard MTT assay was performed. (Bottom) Cell growth as a function of time in the presence of 10 microM PNA, measured by MTT assay. The data are reported as T /C (OD of PNA-treated cells /OD of PNA-untreated cells). Figure 8 | Cell cycle analysis of PNA-treated KYO-1 cells. Cell cycle analysis of PNA-treated KYO-1 cells. KYO-1 cells grown for 30 h in medium containing 10 microM PNA were subjected to cell cycle analysis. Cells were stained with propidium iodide and analysed by flow cytometry. Inset numbers indicate the percentage of cells in the phases of the cell cycle. The percentages of dead cells are 18 and 14% in parPNA- and NLS-asPNA-treated cells respectively, 13% in PNA-untreated cells and 45% in asPNA-treated cells; this high value may be due to apoptosis. DISCUSSION : The results reported in this paper show that a 13mer antisense PNA targeted to the b2a2 mRNA junction in KYO-1 cells down-regulates bcr /abl gene expression up to similar35% of that observed in untreated cells and inhibits cell growth by 50% in a 3 day liquid culture assay. Another finding of this study is that the major biological effect is observed with antisense PNA in the free form and not with its peptide-conjugated analogue. The biological activity in vivo of non-conjugated PNA is rather controversial because it is generally believed that this neutral molecule is unable to penetrate cell membranes. However, recent data disagree with this proposal, because: (i) free PNA administered i.p. was able to cross the blood --brain barrier and to reduce the expression of neurotensin and mu receptor genes; (ii) free PNA added to cells in culture and targeted to a particular region of the non-coding RNA Xist caused disruption of X-chromosome inactivation ; (iii) free PNA administered i.v. in rat and targeted against the delta-opioid receptor gene reduced gene function . In addition to these studies, our own confocal and flow cytometry data show that a certain amount of free PNA can indeed penetrate the cell membrane of KYO-1 cells. We followed two staining procedures. In one case the cells were fixed on a glass slide, permeabilised with Triton X-100 and stained with propidium iodide. In the other case, unfixed cells were treated with hexidium iodide, a vital dye staining both the cytoplasm and nucleus of living cells. This last procedure avoids several washing steps and has the advantage of showing the uptake by intact and living cells. When the PNA was conjugated to NLS, its capacity to penetrate the membranes increased about three times, according to flow cytometry data, in keeping with previous findings . The observation that the uptake of NLS-asPNA was still detected at 4C strongly suggests that this peptide-conjugated PNA is translocated into the cells through a receptor-independent mechanism, which might be based on direct fusion of the neutral PNA with the cell membrane. The cluster of positive charges on NLS should favour uptake by disrupting the structure of the outer membrane layer, acting on the negatively charged phospholipids . Moreover, since asPNA is also internalised in the cells at 4C, its uptake should not be mediated by endocytosis, as occurs for phosphate and phosphorothioate oligonucleotides . The capacity of the designed PNAs to hybridise to the b2a2 RNA target was checked in vitro by RT --PCR. The fact that asPNA (and NLS-asPNA) completely blocked reverse transcription suggests that these molecules can form very stable complexes with the b2a2 target. It is noteworthy that parPNA also forms a sufficiently stable parallel PNA --RNA duplex capable of inhibiting reverse transcription. If we assume that only a small fraction, for instance 10 --3, of the PNA added to the culture medium (usually 10 microM) is taken up by the cells, the intracellular PNA concentration will be similar10 nM. At this low concentration the antiparallel PNA --RNA hybrid should still be characterised by a Tm of similar52C, which is compatible with a strong biological activity. For instance, an antisense PNA injected i.p. into rats was found to reduce neurotensin receptor expression at concentrations as low as 2.4 ng /whole brain . In contrast, 10 nM parPNA, which forms a heteroduplex with a predicted Tm of similar37C with the b2a2 mRNA target, should produce a lower antisense effect, which is in keeping with the experimental results. Another question about the use of antisense PNAs that remains to be clarified regards their mechanism of action. It is known that PNAs do not activate RNase H, so the antisense effect of this molecule should be ascribed to steric blocking of translation. It has been shown in a cell-free system that when the target is near the translation initiation site, hybridisation of the PNA to this target hinders the assembly of the translation initiation complex. When the target is far from the translation initiation site, as at the b2a2 junction, the PNA should induce production of a truncated protein . Interestingly, this and other studies have reported that PNA can promote a reduction in mRNA. To rationalise this finding, Aldrian-Herrada et al. have hypothesised that when mRNA is blocked by an antisense PNA, it may enter a metabolic pathway that leads to degradation. It is also possible, however, that the reduction in mRNA is simply a consequence of the fact that RNA was quantified by RT --PCR. In fact, since PNA --RNA hybrids are resistant to phenol extraction and urea denaturation , the total RNA extracted from the PNA-treated cells will probably contain b2a2 mRNA bound to PNA. It follows that the RNA which is bound to PNA will not be amplified by RT --PCR. In this case, the observed reduction in b2a2 mRNA level would provide indirect evidence that antisense PNA added to the culture medium enters the cells and binds to the corresponding intracellular target. The fact that PNA --RNA heteroduplexes do not induce cleavage of the mRNA target by RNase H suggests that any non-specific effects may be limited, as the possible interaction of asPNA with other mRNA molecules will only result in a temporary arrest of translation. In contrast, non-specific binding of phosphate and phosphorothioate antisense oligonucleotides that induce RNase H may result in irreversible degradation of critical cellular mRNAs with consequent serious side effects. We were surprised to observe that NLS-asPNA appeared to be a weaker antisense agent than free asPNA. In using NLS-asPNA we reasoned that this molecule, by virtue of its ability to penetrate the nucleus, would be able to hybridise with the b2a2 RNA target at both the nuclear and cytoplasmic levels and efficiently block translation. Since this is not supported by the experimental data, it is possible that NLS-asPNA, due to the positive charges of Lys residues, interacts non-specifically with cellular proteins. The reduction in p210BCR /ABL appears stronger at 24 and 48 than at 72 h. This is an obvious consequence of the fact that the antisense effect that we observed follows a single PNA administration. This behaviour is in keeping with the results of a previous study in which the amount of p210BCR /ABL was reduced by circular antisense oligonucleotides directed against the b2a2 junction . Moreover, a significant inhibition of p210BCR /ABL was also reported after 36 h incubation in the presence of antisense oligonucleotides . As p210BCR /ABL has a half-life of similar48 h , after 24 h exposure to asPNA one would have expected a reduction in protein level of 25 --30% and not of 60%, as experimentally observed by us and other researchers. How can this apparent discrepancy be rationalised? Possible explanations are: (i) as the reported half-life for p210BCR /ABL was measured by inhibiting total protein synthesis with cycloheximide (proteases included), a value of 48 h is probably an overestimate; (ii) asPNA may bind non-specifically to proteins and reduce their half-life; (iii) p210BCR /ABL, being involved in signal transduction pathways, is usually complexed with adaptor proteins so that its half-life refers to the complexed state. When the level of p210BCR /ABL is reduced by asPNA, its interaction with adaptor proteins is not favoured and p210BCR /ABL dissociates. It is possible that the free protein has a shorter half-life than the complexed protein. Finally, there is evidence that p210BCR /ABL activates mitogenic pathways and confers a growth advantage to Ph-positive cells over normal haematopoietic cells ( --). In particular, p210BCR /ABL activates, through an interaction with adaptor proteins, the ras pathway, which promotes cell proliferation. The observation that KYO-1 cells treated with asPNA exhibit an incomplete cell cycle and slower cell proliferation is in keeping with down-regulation of the bcr /abl gene by the antisense PNA. Backmatter: PMID- 12202776 TI - Dictionary-driven protein annotation AB - Computational methods seeking to automatically determine the properties (functional, structural, physicochemical, etc.) of a protein directly from the sequence have long been the focus of numerous research groups. With the advent of advanced sequencing methods and systems, the number of amino acid sequences that are being deposited in the public databases has been increasing steadily. This has in turn generated a renewed demand for automated approaches that can annotate individual sequences and complete genomes quickly, exhaustively and objectively. In this paper, we present one such approach that is centered around and exploits the Bio-Dictionary, a collection of amino acid patterns that completely covers the natural sequence space and can capture functional and structural signals that have been reused during evolution, within and across protein families. Our annotation approach also makes use of a weighted, position-specific scoring scheme that is unaffected by the over-representation of well-conserved proteins and protein fragments in the databases used. For a given query sequence, the method permits one to determine, in a single pass, the following: local and global similarities between the query and any protein already present in a public database; the likeness of the query to all available archaeal /bacterial /eukaryotic /viral sequences in the database as a function of amino acid position within the query; the character of secondary structure of the query as a function of amino acid position within the query; the cytoplasmic, transmembrane or extracellular behavior of the query; the nature and position of binding domains, active sites, post-translationally modified sites, signal peptides, etc. In terms of performance, the proposed method is exhaustive, objective and allows for the rapid annotation of individual sequences and full genomes. Annotation examples are presented and discussed in Results, including individual queries and complete genomes that were released publicly after we built the Bio-Dictionary that is used in our experiments. Finally, we have computed the annotations of more than 70 complete genomes and made them available on the World Wide Web at . Keywords: INTRODUCTION : The automatic elucidation of protein function directly from sequence has been the focus of research activity for many years. Such an elucidation has an obvious appeal for it tries to minimize the amount of associated manual labor by reducing a large number of possibilities to one or a handful of choices. This is typically achieved by tapping into repositories of previously accumulated knowledge and by trading computation (i.e. in silico approaches) for typically tedious manual analysis. The discovery of protein function directly from sequence, in an automated or semi-automated manner, has become a fundamental question as thousands of unknown proteins and increasing numbers of complete genomes are made available daily in the public domain. Of course, one should not lose sight of the fact that protein annotation is the first step in the attempt to fully describe a particular organism through characterization of its metabolic pathways and transcription regulation networks. During the last three decades, numerous methods have been proposed for determining protein function from sequence, all of which are essentially instances of a 'guilty by association' approach to solving this problem. Depending on the nature of the information exploited and the manner in which the information is used, these methods can essentially be divided into a handful of well-differentiated categories. The chronologically earliest examples of protein annotation methods rely on the determination of a local or global similarity between a query protein and proteins with known annotation that are contained in a database ( --). If two sequences of comparable length share a large portion of their extent, the previously uncharaterized sequence will inherit the function of the annotated one. The validity of this scheme relies on the implicit assumption that two sequences that 'look the same' at the sequence level also have the same function and structure. This is a reasonable assumption, but counter-examples such as the dehydrogenase /z-crystallin case have also been documented in the literature (for a discussion of this particular case see for example ). The methods in this category are known as 'similarity-based' or 'homology-based' and are numerous. The approach we present in this paper belongs to this category as well. A recurrent situation within the similarity-method category that is pertinent to our discussion relates to the inclination of annotators to use either the first or the best 'hit' from the output of a database search that is carried out by one of the similarity search algorithms such as FASTA , BLAST , Smith-Waterman , etc. In the presence of domains that are shared by numerous proteins , choosing the first or the best hit may not be optimal. As a matter of fact, the multi-domain organization of proteins can lead to incorrectly annotated database entries (Fig. shows a characteristic example of such a misannotation). Use of a domain scan and the exploitation /analysis of the generated output when annotating a query can substantially improve the results. Such a domain scan can be implemented, for example, with the help of the PROSITE, PRINTS, PFAM, BLOCKS or PRODOM databases ( --). For a description of other sources of potential concern in protein annotation and some recommended solutions the reader is referred to previous publications ( --). A second category of methods has become known as the Rosetta stone approach (also known as the domain fusion method) ( --). Here, one seeks to determine groups of proteins that are distinct in a given organism but appear as a single product in another organism, the result of an assumed fusion event. The distinct proteins in the original group are assumed to be physically interacting and, depending on the specifics of each case, this information can be helpful in determining protein function. The methods in the third category seek to determine groups of proteins that repeatedly appear as neighbors of one another in the chromosomes of different organisms. The proteins involved are assumed to have a functional relationship (this methodology is similar in flavor to the Rosetta stone approach but distinct with respect to the type of information that it uses). Exploitation of this idea has found a best fit in the case of prokaryotic genomes where proximal gene organization is manifested in the form of operons, and it has been used successfully to guide functional annotation . It is not evident, however, whether this idea carries over to eukaryotic organisms due to the fact that, in general, the latter lack operons. A closely related variation, which does extend to eukaryotic organisms, operates on the assumption that if an organism is in need of a specific pathway then the organism will carry all or most of the genes comprising the pathway, and vice versa. For example, the work of Marcotte and similar work done by others attempt to define function in terms of the pathways and complexes in which the protein participates, rather than suggest a specific biochemical activity: in this framework a protein is associated with a function via its linkages to other proteins. Finally, in recent years, a fourth category has emerged. Here, one tackles the problem of protein function elucidation through the analysis of correlated mRNA expression of the type that is encountered in the context of DNA- and microarray-chip experiments. The underlying assumption is that functionally related proteins will exhibit correlated mRNA expression levels under multiple experimental settings. Consistent participation of a previously uncharacterized protein in clusters comprising proteins with a well-understood function imposes constraints on the unknown protein's possible behavior by restricting its candidate memberships within the context of a metabolic pathway . In principle, this method can help resolve a protein's function. A more recent variation of this general approach measures levels of protein expression (instead of mRNA) with the help of mass spectrometry or 2D gel electrophoresis in an attempt to determine clusters of strongly co-expressed proteins: these clusters can be used to determine the function of uncharacterized proteins. We next present and discuss a new approach to the problem of protein annotation. At the center of our approach is the Bio-Dictionary, an exhaustive collection of amino acid patterns, heretofore referred to as seqlets, that completely covers the natural sequence space of proteins to the extent that the latter is sampled by the currently available biological sequences. In earlier studies, we showed that the seqlets can capture both functional and structural signals that have been reused during evolution within and across families of related proteins. Our approach relies on the seqlets contained in the Bio-Dictionary and the associated information that is available in a well-maintained database such as SwissProt /TrEMBL , derives from an earlier prototype system we built to carry out similarity searching (,) and employs a weighted, position-specific scoring scheme that is not affected by the over-representation of well-conserved proteins and protein fragments that are present in the public databases. Although similar in intent to systems like GeneQuiz , our method goes beyond simply stating the presence of local and global similarities between a query and one or more database sequences: in fact, we also report information about the secondary structure characteristics of the query, the presence of known domains, signal peptides, active sites, post-translationally modified sites, cytoplasmic /extracellular behavior, the similarity of the query to each of the three phylogenetic domains as a function of amino acid position, etc. Figure 1 | An example of incorrect protein annotation as a result of multiple domain sharing. An example of incorrect protein annotation as a result of multiple domain sharing. The sequences have been aligned in a manner that shows their common domains. The 'd-arabino 3-hexulose 6-phosphate formaldehyde lyase' function has been propagated from the HUMS_BACSU sequence to the rest of the sequences in this group. Instances of the same domain have been shown in the sequences that contain it using the same color and shading scheme. MATERIALS AND METHODS : Background | The Bio-Dictionary idea was introduced and discussed in earlier works ( --); therein we described how one can use the Teiresias pattern discovery algorithm (,) to process a very large public database of amino acid sequences and fragments and derive a collection of amino acid patterns that, by design, appear within as well as across family boundaries. These patterns are referred to as seqlets and have been shown to capture functional and structural signals; moreover, they have the very important property of completely describing the processed input at the amino acid level. Following are some seqlet examples that include the name of the represented feature or of the represented protein family, taken from Rigoutsos et al. : GDG[IVAMTD]ND[AILV][PEAS] [AMV][LMIF]..A (cation-transporting ATPases), V.I.G. G..G...A (NAD /FAD-binding flavoproteins), G..G.GK[ST]TL (ATP /GTP-binding P-loop), KMSKS[LKDIR][GNDFQ]N (class I aminoacyl-tRNA synthetases), H.....HRD.K..N (serine /threonine protein kinases), etc. In terms of the notation used, [LKDIR] means a choice of exactly one among L, K, D, I and R, whereas '.' denotes a single position wild-card character that can replace any of the symbols in the alphabet. The derived collection of seqlets can thus be treated as a (redundant) vocabulary for the natural sequence space of proteins to the extent that the latter is uniformly sampled by the currently available biological sequences. Associating the entries of this vocabulary-like collection with the annotation information contained in a typical entry of the SwissProt database allows us to convert the collection into a dictionary. We have been using the term Bio-Dictionary to refer to the collection of seqlets that has been augmented so as to include the 'annotation meaning' for each of the entries. The key elements behind the Bio-Dictionary, and details of its construction, can be found in Rigoutsos et al. ; analysis of the 3D structural properties associated with the seqlets of a dictionary built out of 17 complete archaeal and bacterial genomes are given in Rigoutsos et al. ; finally, a discussion and description of potential uses for it appears in Rigoutsos et al. . In more recent work, we applied the Bio-Dictionary to in silico prokaryotic gene finding and built a system with exceptional performance : unlike approaches that are based on Markov models where each distinct genome requires that a different model be built, our gene finding system is universal in that a single instance of it is used across all archaeal and bacterial genomes. The earlier work | By carrying out pattern discovery on a given sequence database D, we can use the generated pattern collection C to carry out similarity searches for instances of a query or its fragments in D as follows: a pattern from the derived collection C of patterns that matches a region of the query under consideration pinpoints a potential local similarity between the matched region of the query and all of the sequence fragments from the input database that the pattern represents (recall that by the definition of pattern discovery, patterns appear k or more times in the processed input, with k >= 2). In earlier work, we used the Teiresias algorithm to process Release 34 of the SwissProt database and built a prototype system for similarity searching using only a subset C' from the derived collection C of patterns. A given query sequence was examined for matches of patterns contained in C' and the query and database regions corresponding to the matches were aligned, scored, and finally sorted according to the computed score. Following the sorting, one could proceed in one of two distinct ways: (i) the user was presented with the collection of patterns that matched the query and was asked to identify those of biological importance, then alignments were generated using those patterns alone; (ii) those alignments that resulted from patterns whose database instances carried an associated annotation (namely the 'FT' line) were reported to the user for further study (,). This early system was meant to be a proof of concept. Consequently, complete coverage of the input database by the patterns in the collection C' was neither achieved nor pursued. As a matter of fact, this early system used a mere 565 432 patterns which covered similar50% of the processed SwissProt database at the amino acid level. Neither the existing over-representation of various protein families in the database nor the system's real-time performance were design considerations at that time. However, this early excursion provided an invaluable learning experience that helped guide us toward the system which we present next. The method: description | The first and foremost consideration of the new approach is the achievement of a complete coverage of the natural sequence space as it is currently known. To that end we used as our domain of knowledge the 14 May 2001 release of SwissProt /TrEMBL, a large, publicly available and curated database . This particular release comprised 532 621 amino acid sequences and fragments with a grand total of 170 762 058 amino acids. We processed this input database in two phases. First, we ran Teiresias using L = 8, W = 8 and K = 2 and generated variable length seqlets that contained no wild-cards. For each one of these seqlets, we located and masked all of its instances in the input database except for the one that appeared in the longest among the sequences containing instances of the seqlet. We then reran Teiresias on the masked input, but this time using L = 6 and W = 15. For more information about this scheme and other methodological details the reader is referred to Rigoutsos et al. . The processing required similar45 CPU days worth of computation using IBM RS64III processors with a clock speed of 450 MHz. With the help of a parallel implementation of Teiresias that we have developed for shared memory architectures, we completed this computation in 2 days on a 24 processor IBM S-80 supercomputer. The two pattern discovery phases generated the Bio-Dictionary that we used in our analysis and which contained a combined total of 42 996 454 seqlets [compare the size of the current collection with the 565 432 patterns used in Floratos et al. (,)] that accounted for 98.2% of the processed input at the amino acid level (this degree of coverage essentially implies that, on average, a mere five amino acids per protein sequence cannot be accounted for by this seqlet collection). The length and density distributions of these seqlets match closely the ones shown in Rigoutsos et al. , whereas the average length of a seqlet is similar12 --13 amino acids. It should be noted that this Bio-Dictionary contains redundant seqlets, i.e. a given amino acid position in the processed input will typically participate in and is covered by multiple seqlets; this redundancy of representation is a desired property which we exploit during annotation. Each seqlet in the collection carries along the meaning(s) associated with the regions of the proteins that contained an instance of and gave rise to the seqlet. Instead of a static description of each seqlet's meaning(s) in the manner that we described in Rigoutsos et al. , we opted for composing the full entry of each seqlet during the run time as required. We currently derive labels for meanings from only the DE, OC and FT lines of the respective SwissProt /TrEMBL entry; obviously, we can tap into any database containing complementary information and attach additional meanings to each seqlet. One obvious choice is the PDB (,): in previous publications (,) we described how 3D structure can be associated with seqlets and are currently in the process of extending the approach presented herein in order to reconstruct local 3D structure using the structural hypotheses generated by partially overlapping seqlets (D.Platt, I.Rigoutsos, Y.Gao and L.Parida, submitted for publication). Recall that the DE or 'description' line of SwissProt contains general descriptive information about the respective entry. Similarly, the OC or 'organism classification' line contains the taxonomic classification of the source organism. And the FT or 'feature table' line contains a simple and precise means for the annotation of the sequence data: a fixed abbreviation with a defined meaning relating to the feature being reported is followed by the residue numbers indicating the end points (extent) of the named feature; the line ends with a description containing additional information about the feature. Of the available SwissProt /TrEMBL abbreviations contained in an FT line we only make use of the ones listed in Table . When presented with a query Q to annotate, we carry out the steps outlined in Figure , a markedly different approach than the one used in our early prototype. First, we generate the 'intersection' of the Bio-Dictionary with the query sequence to find those seqlets that match somewhere in the query. For each of the seqlets in this intersection, we examine the corresponding SwissProt /TrEMBL entries for all of the sequences that gave rise to the seqlet during the Bio-Dictonary formation, thus building the corresponding dictionary entry 'on-the-fly' by dynamically attaching to the seqlet all the meaning(s) extracted from those entries. The extracted meanings essentially 'color' each seqlet and by extension the region of the query where the seqlet matches. Note that a given seqlet can carry multiple 'colors', i.e. attributes. Con sequently, a region of the query can be associated with multiple attributes. If the seqlet under consideration is attached to an attribute that has not been encountered before, then a new attribute vector is introduced: the attribute vector has the same length as the query and initially contains zeroes everywhere; the current seqlet assigns its contribution CONTRIB(.,.) to this new attribute vector at precisely the region corresponding to the seqlet's match in the query. If the seqlet carries an attribute that has been encountered before, the seqlet adds its contribution CONTRIB(.,.) to the appropriate region of the already existing attribute vector. Multiple seqlets that carry the same attribute will add their individual contributions to the attribute's vector: the regions to which the seqlets contribute may or may not be overlapping. The manner in which we decide what amount a seqlet will contribute to an attribute vector is described in detail below. After all seqlets in the intersection have been exhausted, and separately for each attribute category (e.g. DE, FT, etc.), the attribute vectors are sorted and ranked based on the accumulated support and the top T ranking vectors of the category are reported (T is a user-specified parameter). Each of these vectors will contain non-zero values at precisely those regions that were matched by possibly overlapping, distinct seqlets that carried the same attribute. Before we describe the scoring scheme it is important to stress some points that are particular to our work. In general, the Bio-Dictionary should not be seen as a collection of seqlets each of which necessarily captures a specific feature such as a kinase domain, a metal binding site, etc. Seqlets that can act as predicates for a feature or protein family do exist in the Bio-Dictionary, but, by design, seqlets may also carry multiple meanings. This is different from the one-to-one correspondence that the reader may be accustomed to and which is typical of predicate-containing databases such as PROSITE, PRINTS and INTERPRO (,,). As we showcased previously , a seqlet can cross functional and structural boundaries and can thus be associated with multiple meanings. Clearly, those of the seqlets that are associated with a unique meaning can function as predicates, but a significant number of them will capture and correspond to multiple meanings. Similarly, the Bio-Dictionary also contains multiple seqlets all of which capture the same meaning. These seqlets can also have instances that overlap with one another, as indicated by the fact that the product of the number of seqlets contained in the Bio-Dictionary times the seqlets' average length is a multiple of the actual length of the processed input . Thus, by design, a given position of a processed query will in general be covered by multiple seqlets. Each of the seqlets covering a position within the query will in general carry one or more meanings that are used to 'color' the corresponding region of the query. Let a given query position be covered by M distinct seqlets. In order for an attribute, e.g. 'metal-binding site', to rank high in the reported results, a large portion of those M seqlets must carry this attribute. Recall that, by definition, each of the seqlets of the Bio-Dictionary appears in at least two places in the processed database (SwissProt /TrEMBL in our case): thus, if M seqlets cover a given position in the query to annotate, then the following two properties will simultaneously hold: *there exists a total of F sequence fragments corresponding to all of the instances of the M seqlets in the processed input database; clearly, these fragments will be similar to the amino acid neighborhood surrounding this query position; *the F sequence fragments in the database will agree on the amino acid identity of the literals (recall that the seqlets contain both literals and wild-cards, i.e. 'dots') contained in each of the M seqlets. These database sequence fragments may or may not agree on the annotation of the query position under consideration. If the annotations for N of these F database sequence fragments state that this site is a metal-binding site then through application of the 'guilty by association' approach, our belief that this query position is also a metal-binding site will be proportional to N /F. This very idea is applied to every attribute and attribute category that is attached to a seqlet. The direct implication of this is that a seqlet can be useful and able to contribute to a specific annotation without having to span an attribute (e.g. protein kinase domain) in its entirety. Moreover, the seqlet does not have to be a predicate for an attribute either. Figure graphically depicts this situation. For discussion purposes, let us assume that when we searched the query Q with the Bio-Dictionary, we determined that seqletK is present in Q in the region [qfrom, qto]. Let us also assume that during the Bio-Dictionary formation, seqletK was determined to have three instances in the processed SwissProt /TrEMBL database. After following these three backpointers to the full entries of the sequences that contain the three instances of seqletK, we determine that in one of the sequences the seqlet instance spans an interval [pfrom, pto] that has a non-empty intersection with a specific region [featfrom, featto] of that sequence that is annotated as np_bind atp, i.e. as an ATP-binding site. Let [ifrom, ito] denote the intersection of the intervals [pfrom, pto] and [featfrom, featto]. In this example, seqletK will corroborate the presence of a partial ATP-binding domain in the query that is being annotated by incrementing the support at the locations [qfrom + (ifrom -- pfrom), qfrom + (ito -- pfrom)] of the np_bind atp attribute vector. It should now be clear why any given seqlet does not have to serve as a predicate for the attribute(s) that it corroborates. The term 'attribute' is overloaded in our discussion and should be interpreted rather loosely: it can mean a local similarity, a global similarity, an active site, a phylogenetic domain, a post-translationally modified location, etc. If the query being annotated contains a true instance of a given attribute, then each one of the numerous seqlets that will cover the region spanned by the attribute more than once will cumulatively and independently provide support for the attribute at the respective positions: as the accumulated support for the attribute increases so does the likelihood of its presence in the query. If the query is a true member of a known protein family, then we expect the attribute vector for this family to obtain support along its entire length from practically every single one of the seqlets that match in this query. If a query contains a known domain, then the attribute vector for the domain will have non-zero support over the region of the query that corresponds to the domain's instance. In an analogous manner, if the processed query shares only a local region with a well-characterized family or an individual protein, then the corresponding attribute vector will have non-zero support only over the shared region. The manner in which we use the seqlets and accumulate scores has proven particularly useful in situations that, among others, include the following: the query is a fragment of a known sequence; the query contains one or more known domains in a novel order; the query has been assembled using an incorrect exon collection (e.g. one or more true exons are missing, introns have been mislabeled as exons and included in the assembly, exons that correspond to distinct genes have been assembled together, etc.). Moreover, the fact that our seqlets have lengths that typically span between 6 and 18 amino acids (for a detailed discussion see ) permits us to easily and correctly process very short input queries, e.g. 8 --10 amino acids, without the thresholding constraints and limitations that one typically encounters when using heuristics-based similarity search algorithms (,). In real-world applications, situations arise where the query represents or contains only a fragment from a known domain, for example a query involving the first few tens of amino acids from say a 'protein kinase domain'. In order to alert the user to this situation, we also include, wherever applicable, the 'minimum', 'average' and 'standard deviation' values for the span of each of the T top ranking reported attribute vectors. This permits easy determination of whether the query represents a complete instance of the stated attribute or only a fragment. The method: scoring | How much to contribute. Above we described how we determine the extent of an attribute vector region to which a seqlet matching the query will contribute. We now discuss how we determine the amount that the seqlet will contribute. Let seqletK be present in query Q and let qi1qi2qi3...qil and pj1pj2pj3...pjl be its instances in the query and in some database sequence d, respectively; let {i1, ... il} and {j1, ... jl} denote the indices of the positions spanned by the seqlet in the query Q and sequence d, respectively. For simplicity, we will assume that the instance of seqletK completely spans an annotated region of d that corresponds to an attribute A. SeqletK brings together two sequence fragments with lengths equal to the span of the seqlet; one fragment comes from the query that is being analyzed while the other is from the sequence d of the database. Obviously, the more similar these two fragments are the more likely it is that upon completion of the annotation process the attribute A that is associated with the database region pj1pj2pj3...pjl will be carried over to the query region qi1qi2qi3...qil through the 'guilty by association' approach. There is a rather straightforward manner in which seqletK can contribute to the vector for attribute A; we simply use one of the available scoring matrices and generate contributions in a position- and content-dependent manner as follows: for m = 1 to l {attribute_vector[i1 + m -- 1] += f(scoring_matrix[qi1 + m -- 1][pj1 + m -- 1])} (the symbol += is shorthand notation for 'increment by amount shown on the right of the = sign'). In other words, the seqlet will contribute to the (i1 + m -- 1)th position of the attribute vector an amount that relates to the degree of similarity between the amino acids occupying the positions qi1 + m -- 1 and pj1 + m -- 1, respectively. A good choice for the function f(.) above is f(x) = 2x + constant; with regard to the scoring matrix to use one can employ any of the standard PAM or BLOSUM scoring matrices (,). In order to avoid the over-counting that would be the consequence of a given protein family or fragment being over-represented in the SwissProt /TrEMBL database, we impose the additional constraint that a given seqlet cannot contribute to the same attribute vector and vector position(s) more than once. In other words, if seqletK captures a very well-conserved region appearing in a large number of SwissProt /TrEMBL sequences, only one of the seqlet's numerous annotated database instances will contribute to the respective attribute vector. How to normalize. As mentioned already, a given seqlet with distinct possible meanings will contribute in turn to each of the attribute vectors that correspond to those meanings. And these contributions will depend on how well a known database instance of the attribute matches its alleged instance in the query. Different attribute vectors will accumulate different amounts of contribution and these contributions will also depend on the position within the attribute vector. During the annotation of the query we maintain a book-keeping array, total_contrib, with length equal to that of the query; for every seqlet with an instance qi1qi2qi3...qil in the query, we update total as follows: for m = 1 to l {total_contrib[i1 + m -- 1] += f(scoring_matrix[qi1 + m -- 1][pj1 + m -- 1])} In other words, the ith position of total_contrib is a measure of the number of seqlets that contribute to it, with each contribution weighted by the degree of similarity between the amino acids in the query and their input database counterparts. The function f(.) is the same as in the previous section. Note that at all times during processing, the value of total_contrib[i] is greater than or equal to the maximum value one will encounter in the ith position of any of the active attribute vectors for this query. Once all of the seqlets matching the query have been examined, we normalize the contents of the ith position of each attribute vector by dividing by the value of total_ contrib[i]. Multiplying this normalized value by 100 gives us, for each attribute vector, a measure of the fraction of the total contribution that this attribute has received, as a function of position within the query. Well-conserved attributes will have values close to 100% whereas less conserved attributes will have fewer seqlets contributing to them and thus will have smaller values. Note that this particular way of normalizing has the additional property of alleviating the situation where equal length regions of the query receive disproportionately different contributions due to differences in the number of contributing seqlets: this normalization will permit all regions in the query to have equally 'strong voices'. How to rank. Once the contents of the attribute vectors have been normalized, we sort them based on their received contributions and report the top T of them. We have implemented a scheme that will rank a narrow, well-conserved region higher than a wider region which is not as well conserved. This permits us to report attributes such as well-conserved active sites or post-translationally modified sites among the top ranking positions of the results. Finally, when we report local similarities, we further require of the attribute vectors pertaining to similarities that any set of consecutive non-zero positions be at least X positions wide; the value of X is user-defined and typical values range in the interval [10, 20]. The method: how to find matches in the query | In order to efficiently implement the above method we need to be able to quickly determine which of the Bio-Dictionary seqlets match where within the query. A simplistic approach would require that, for the similar43 000 000 seqlets and every single query position, we check whether there is a match; this would of course be very slow. The problem of identifying such matches is complicated by the presence of wild-cards ('don't care characters') in the seqlets that we use. To deal with this situation, we have designed and implemented a novel and very efficient method for solving precisely this problem: our method makes use of a very efficient hashing scheme that subselects among the Bio-Dictionary seqlets prior to using the ones that survive in conjunction with a modified version of the Aho-Corasick algorithm . The resulting scheme permits us to fully annotate a 300 amino acid protein in similar10 s on a single IBM RS64III processor running at a clock speed of 450 MHz. The description of this matching algorithm extends beyond the scope of this presentation and will be given elsewhere (M.Lewenstein, T.Huynh and I.Rigoutsos, in preparation). Figure 2 | Pseudo-code showing the computational steps of the method. Pseudo-code showing the computational steps of the method. Figure 3 | Accumulating seqlet contributions. Accumulating seqlet contributions. Seqlets do not have to span a feature in its entirety in order to corroborate an attribute. Nor do they have to be specific enough to act as predicates for the attribute. See also text for more details. Table 1 | FT line labels used in our work (see also text) RESULTS : We next showcase the capabilities of our approach by annotating a carefully selected collection of example queries and discussing the results obtained. All of the results we report in this section can be reproduced using the Web-based implementation of our method, available at . The underlined text in the figures generated by the Web-based tool is in fact hyperlinks which permit the user to issue a search request to SwissProt /TrEMBL and retrieve all of the database entries with the property stated by the text. This capability is meant to facilitate cross-comparisons and verification of the reported results. Moreover, upon completion of an annotation, the user can view the Bio-Dictionary patterns that matched within the query, as well as each pattern's estimated log probability and the actual position within the query where the match begins. Example 1. UBIQ_HUMAN | As our first example, we examine the annotation of the 76 amino acid human ubiquitin, UBIQ_HUMAN. Some of the results of the analysis are shown in Figures --. As can be seen from these figures, the SwissProt /TrEMBL database contains enough information for our method to correctly determine the secondary structure of the fragment: notice the localization of the helices, strands and turns and their interweaving in Figure . It is not always the case that there will be enough information in SwissProt /TrEMBL for us to be able to make statements about the local secondary structure of a query. This limitation can be alleviated in one of two ways: (i) we can rely on SwissProt /TrEMBL's continuing augmentation and updates ---as the database becomes bigger and more enriched, our capability to annotate local structure will also improve; (ii) we can make use of the information in the PDB database in the manner that we have described (,). The seqlets' meanings will be enriched by incorporating structural information from the much more comprehensive PDB; we are currently in the process of augmenting our annotation method so that it will include this component. Finally, note how our method correctly determines the nature and position of seven sites that are relevant to the function of ubiquitin as well the presence and extent of the ubiquitin domain. Example 2. A very short fragment | For our second example, we have selected the 8 amino acid fragment VVVTAHAF, a fragment that is too short to be used with heuristics-based similarity search algorithms such as FASTA and BLAST /PSI-BLAST. As shown in Figure , when we process this fragment with our method we can correctly determine that: (i) it is an amino acid combination encountered only in the eukaryotic domain; (ii) it belongs to a cytochrome c oxidase; (iii) it is part of a transmembrane domain; and (iv) it has a metal-binding site (iron) at the sixth position from the beginning, i.e. at the position of the histidine. Example 3. ACTR_BOVIN | Another capability of our system is the determination of cytoplasmic, transmembrane and extracellular regions in a given query. We showcase this using ACTR_BOVIN, an adrenocorticotropic hormone receptor protein from Bos taurus. Figure shows the plots for the cytoplasmic and extracellular behavior of the query as determined by our method: note that the regions of the query that are not accounted for by these two plots correspond precisely to the seven transmembrane domains of the ACTR_BOVIN (the corresponding transmembrane plots are not included in this figure). Example 4. UL78_HCMVA | The next example is a sequence that comes from the human herpesvirus 5 . In particular, it is the 431 amino acid sequence UL78_HCMVA. In Figure we show the output of both RPS-BLAST and PSI-BLAST on this specific query sequence: as can be seen, the only detectable similarity is with the rhodopsin family and is confined in the region [60, 170]; no other similarities can be determined outside this region (the PSI-BLAST hit at the second position is with an uncharacterized sequence from the Tupaia herpesvirus and thus is not informative). One possible interpretation is that there is no single sequence in the SwissProt /TrEMBL database that resembles UL78_HCMVA (other than the query itself and its Tupaia herpesvirus counterpart), either in terms of the order of any domains that may be present or in terms of its composition. When we process UL78_HCMVA with our method, we discover weak similarities that relate UL78_HCMVA mainly to hypothetical proteins in a manner that is similar to what is shown in Figure . However, further inspection of our results provides us with enough information to appropriately categorize the query. In Figure we show the plot for the query's transmembrane behavior as reported by our method: seven very distinct regions are immediately apparent, thus permitting us to conjecture that this sequence is a G protein-coupled receptor homolog. The seven regions correspond to the intervals 45 --61, 74 --95, 112 --132, 154 --174, 202 --222, 236 --256 and 280 --300, respectively, and have well-delineated boundaries. Notably, a similarity search in the GPCRDB database using UL78_HCMVA as the query currently generates no hits to known families. In Table we show the alignment for the best ranking hits obtained when we search the GPCRDB database subset that contains only SwissProt entries (but not TrEMBL) using each of the seven putative transmembrane regions as a query. If the UL78_HCMVA putative transmembrane regions correspond to transmembrane helices of a GPCR homolog one will expect to see them matching known transmembrane regions from sequences in GPCRDB. This is indeed the case, as this table shows: the regions of the GPCRDB /SwissProt hits that are labeled as transmembrane regions of a G protein-coupled receptor are shown underlined bold next to each of the seven queries. The only exception is the query corresponding to the putative TM helix 3, where the best match is to a transmembrane region from PSAB_ANTMA, a photosystem protein. In all seven cases the quality of the conservation is notable. It should also be noted that several sequences other than UL78_HCMVA were reported as GPCR homologs when the analysis of the complete genome of the human herpesvirus 5 was first published . Example 5. Comparison with the annotations of recently published /updated genomes | We next showcase the capability of our approach by processing the complete genomes of three organisms whose sequences were published after 14 May 2001 and compare our results with the annotations that accompanied the release of the respective genomes. Since the Bio-Dictionary that we used for our experiments was built using the SwissProt /TrEMBL release of 14 May 2001, the results from these comparisons are indicative of our method's ability to extrapolate and annotate novel sequences. Additionally, we annotated two genomes whose sequences were released into the public domain prior to 14 May 2001. Obviously, the sequences of these organisms are already contained in the input database from which we built the Bio-Dictionary used in these annotations. However, the GenBank database entries for these genomes and the respective annotations were updated several months after 14 May 2001: it is these more recent annotations that we use for our comparative study and not the annotations that accompanied the original genome submission. The purpose behind these comparisons is to determine the extent of agreement between our predictions and the original annotators' updated predictions when using a sequence database that has been substantially augmented since the genomes under consideration were originally deposited. It should be stressed that any such comparisons can only provide estimates of what a user can expect when using our method to annotate a genome. Indeed, the very notion of an automated comparison of different annotation collections is, to a certain degree, ill-defined. The following observations will make this last statement clear. First, the published genomes are sequenced, annotated and released by different research groups which employed different automated tools in conjunction with generally distinct, although overlapping, knowledge bases of annotated biological sequences. Once the automatically obtained gene annotations become available, they are typically curated manually during a 'genome annotation jamboree' by a different team of scientists each time and using non-standard nomenclature and abbreviations. As a result of this manual curation, the annotations that accompany a newly published genome contain much more that simply the result of applying a 'guilty by association' automated approach. This last observation puts us at a distinct disadvantage when carrying out the comparisons that we report below. Independently of the annotation approach that is used, there is always the issue of what it means to have 'annotated a protein'. Even ignoring disagreements in the annotations of individual proteins, several levels of detail are possible when making an annotation. As an example, Table shows valid, non-contradicting annotations for a fictitious protein: the thing to notice here is the different amount of information that is conveyed by each annotation statement. Ideally, one seeks the most detailed description possible for the available knowledge base. The possibility of different levels of annotation detail adds an extra degree of difficulty and can result in annotation disagreements when lists of annotations that have been reported by different groups at different points in time are automatically compared with one another (,, --). Even if one ignores the above difficulties, differences can still arise as a result of using different guidelines and criteria each time, thus leading to substantial variations in the claimed percentage of genes that can be annotated in a newly sequenced genome based on sequence similarity with known proteins. Generally speaking, the current state of the art permits one to report functional hypotheses for similar70% of the predicted genes in a given prokaryotic genome ( --). The fraction for eukaryotic genomes is typically much lower, although in the case of specific eukaryotic chromosomes, notable exceptions exist . In light of the above observations, we decided to generate our figures by manually comparing, for each and every one of the involved genes, the annotations reported during the release of the genome with those generated by our method. The results are given in Table . The first three genomes, namely Rickettsia conorii Malish 7 , Staphylococcus aureus Mu50 and Streptococcus pneumoniae TIGR4 were published and made available in the Fall of 2001. The last two genomes, namely Chlamydia pneumoniae J138 and Buchnera sp. APS were published in June and September 2000, respectively, but their GenBank records were updated in the Fall of 2001. For each genome we report the number and percentage of genes that fall into each of the following categories: (i) the latest GenBank annotation and our annotation agree; (ii) the GenBank annotation contains a 'hypothetical protein' entry whereas our system proposes a functional hypothesis; (iii) the GenBank annotation lists a functional hypothesis whereas our system reports a 'hypothetical protein'; and (iv) the GenBank annotation and our annotation disagree. As shown in Table , for the two genomes that were updated recently, the agreement between our automated predictions and the latest GenBank annotations reaches a level of 98% over the entire genome. It should be noted that this figure also includes those genes for which there is no functional hypothesis (i.e. they are listed as 'hypothetical proteins'). For the three novel genomes, the agreement between the predictions ranges between 88 and 92%. It is worth reiterating that the annotations that are included in the GenBank entries for the various genomes are the result of manually curating the output of multiple automated tools, whereas our scheme generates annotations in an entirely automated manner using a single unified framework. In recent collaborative work with colleagues from several European laboratories the complete genome of Chlamydia trachomatis serovar D was re-annotated using (i) manual means, (ii) traditional automated tools and (iii) our method. As described in detail (I.Iliopoulos, S.Tsoka, M.A.Andrade, A.J.Enright, M.Caroll, P.Poullet, V.Promponas, T.Liakopoulos, G.Palaios, C.Pasquier, S.Hamodrakas et al., submitted for publication), the annotations that were obtained through manual means and through our Bio-Dictionary-based method achieved the best overall performance reaching an annotation agreement on 862 of the 893 processed sequences, i.e. 96.5% of the entire genome. Of the remaining 31 sequences, 13 could be annotated manually but could not be annotated by our method, whereas the other 18 could be annotated with our method but could not be annotated manually. Example 6. Annotations on the World Wide Web | Similarly to the previous example, we have annotated the sequences of more than 70 complete genomes across the three phylogenetic domains, including: Methanococcus jannaschii DSM 2661 , Halobacterium sp. NRC-1 , Sulfolobus solfataricus P2 , Mycoplasma genitalium G-37 , Synechocystis sp. PCC 6803 , Escherichia coli K12-MG165 , Helicobacter pylori 26695 , Borrelia burgdorferi B31 , Aquifex aeolicus VF5 , Myco bacterium tuberculosis H37Rv , Chlamydia trachomatis serovar D , Chlamydophila pneumoniae CWL029 , Thermotoga maritima MSB8 , Deinococcus radiodurans R1 , Yersinia pestis CO92 , Saccharomyces cerevisiae S288C , Caenorhabditis elegans , Drosophila melanogaster , Homo sapiens (,) and Mus musculus. The annotations of these genomes are available on the World Wide Web and can be viewed and interactively explored by visiting . The system that we make available on the World Wide Web provides the user with several options. Within a specific genome, if the accession number of a gene is known, then it can be used to locate and view the annotation of the gene. Alternatively, one can search the results in the DE and FT attribute categories of the genome using regular expressions that can be entered with the help of a graphical user interface. For example, when run against the DE results, the regular expression -[1-3]].*calcium.*bind will locate and report all the sequences in the genome under consideration that 'share any similarities with calcium binding sequences and are ranked in the top three positions'. Analogously, when run against the FT results, the regular expression -[1-9]].*domain.*bh[1234].* will permit the user to search for sequences that 'contain one or more of the cell apoptosis-associated domains BH1, BH2, BH3 and BH4 and are ranked in the top nine positions'. To list the three top ranking functional hypotheses for each gene in a genome, one can use the regular expression -[1-3]] to search through the DE results. At the user can find information on how to form these regular expressions and the permitted keywords, as well as several specific examples with explanations. Additionally, we have enabled and made available cross-genomic comparisons /searches: through a graphical user interface, one or more genomes can be selected and their annotations searched for similarities with a specific family (e.g. elongation factor, tRNA-aminoacyl synthetase, etc.) or the presence of a specific feature (e.g. hydrogen bond donor, calcium-binding domain, helix --turn --helix, etc.) with the help of regular expressions similar to those used to analyze individual genomes. Figure 4 | Some results from processing the human ubiquitin UBIQ_HUMAN by our method. Some results from processing the human ubiquitin UBIQ_HUMAN by our method. Figure 5 | Additional results from processing the human ubiquitin UBIQ_HUMAN by our method. Additional results from processing the human ubiquitin UBIQ_HUMAN by our method. Figure 6 | More results from processing the human ubiquitin UBIQ_HUMAN by our method. More results from processing the human ubiquitin UBIQ_HUMAN by our method. Figure 7 | Some of the results obtained from processing the fragment VVVTAHAF with our method. Some of the results obtained from processing the fragment VVVTAHAF with our method. Figure 8 | Partial results from processing the adrenocorticotropic hormone receptor protein ACTR_BOVIN by our method. Partial results from processing the adrenocorticotropic hormone receptor protein ACTR_BOVIN by our method. Figure 9 | Results of RPS-BLAST and PSI-BLAST using the sequence UL78_HCMVA as an input. Results of RPS-BLAST and PSI-BLAST using the sequence UL78_HCMVA as an input. Default parameter settings were used. Figure 10 | Partial results from processing the sequence UL78_HCMVA from human herpesvirus 5. Partial results from processing the sequence UL78_HCMVA from human herpesvirus 5. See text for more details. Table 2 | Best GPCRDB /SwissProt hits when using UL78_HCMVA regions as the query (see also text) Table 3 | Annotations for a fictitious protein that are non-conflicting with one another but correspond to varying degrees of conveyed annotation detail Table 4 | Results from manually comparing our predictions with the annotations that have been reported for several genomes DISCUSSION AND FUTURE DIRECTIONS : In this paper, we have presented and discussed a new method for the automated annotation of amino acid sequences. The method quickly, objectively and exhaustively determines local and global similarities between a given query and any protein already present in a public database, the likeness of the query to all available archaeal /bacterial /eukaryotic /viral sequences in the database as a function of amino acid position within the query, the secondary structure character of the query as a function of amino acid position within the query, the cytoplasmic, transmembrane or extracellular behavior of the query, the nature and position of binding domains, active sites, post-translationally modified sites and signal peptides, etc. The key concept underlying our method is that of the Bio-Dictionary, which we presented and discussed in earlier work. By design, the presented method is extendable and can make use of any type of attribute that would be of interest to the end user. It can also make use of multiple databases. Through a carefully selected collection of examples, we have demonstrated the capabilities of our method and the quality of the annotations that it generates. Our system automatically generates results whose quality matches that of publicly available annotations; recall that such annotations are typically the product of a manual curation that has followed the application of automated processes. In terms of actual annotation speed, our system can annotate a 300 amino acid query in similar10 s on a single IBM RS64III processor running at a clock speed of 450 MHz. We are currently in the process of enhancing our system with several new components. One extension involves the automatic determination and reporting of all the PubMed references pertaining to the query sequence that is annotated. For each of the reported results in the DE category we will be making available links to all PubMed articles that are relevant for the study of the query sequence and the family described by the caption. This is currently work in progress. A second extension, which we have already described above, involves the automated generation of local 3D structure through 'meanings' that are derived from the contents of the PDB database. This is also work in progress. Finally, an important topic that we will be studying relates to the fact that the SwissProt /TrEMBL database has up to now used non-standardized nomenclature to label database entries. For example, the following are some of the DE lines that are associated with aldose reductases: aldose reductase (ec 1.1.1.21) (ar) (aldehyde reductase) aldose reductase alcohol dehydrogenase [nadp+] (ec 1.1.1.2) (aldehyde reductase). When our system is presented with an aldose reductase as a query, e.g. ALDR_HUMAN, then multiple attribute vectors will be reported, one for each of these seemingly distinct (but in reality identical) attributes. A planned future release of our system will alleviate this problem through the use of standardized names. Backmatter: PMID- 12202762 TI - Determination of base and backbone contributions to the thermodynamics of premelting and melting transitions in B DNA AB - In previous papers of this series the temperature-dependent Raman spectra of poly(dA)poly(dT) and poly(dA --dT)poly(dA --dT) were used to characterize structurally the melting and premelting transitions in DNAs containing consecutive AT and alternating AT/TA base pairs. Here, we describe procedures for obtaining thermodynamic parameters from the Raman data. The method exploits base-specific and backbone-specific Raman markers to determine separate thermodynamic contributions of A, T and deoxyribosyl-phosphate moieties to premelting and melting transitions. Key findings include the following: (i) Both poly(dA)poly(dT) and poly(dA --dT) poly(dA --dT) exhibit robust premelting transitions, due predominantly to backbone conformational changes. (ii) The significant van't Hoff premelting enthalpies of poly(dA)poly(dT) [DeltaHvHpm = 18.0 +- 1.6 kcalmol --1 (kilocalories per mole cooperative unit)] and poly(dA --dT)poly(dA --dT) (DeltaHvHpm = 13.4 +- 2.5 kcalmol --1) differ by an amount (similar4.6 kcalmol --1) estimated as the contribution from three-centered inter-base hydrogen bonding in (dA)n(dT)n tracts. (iii) The overall stacking free energy of poly(dA) poly(dT) [ --6.88 kcalmolbp --1 (kilocalories per mole base pair)] is greater than that of poly(dA --dT) poly(dA --dT) ( --6.31 kcalmolbp --1). (iv) The difference between stacking free energies of A and T is significant in poly(dA)poly(dT) (DeltaDeltaGst = 0.8 +- 0.3 kcal molbp --1), but marginal in poly(dA --dT)poly(dA --dT) (DeltaDeltaGst = 0.3 +- 0.3 kcalmolbp --1). (v) In poly(dA) poly(dT), the van't Hoff parameters for melting of A (DeltaHvHA = 407 +- 23 kcalmol --1, DeltaSvHA = 1166 +- 67 calK --1mol --1, DeltaGvH(25C)A = 60.0 +- 3.2 kcalmol --1) are clearly distinguished from those of T (DeltaHvHT = 185 +- 38 kcalmol --1, DeltaSvHT = 516 +- 109 calK --1mol --1, DeltaGvH(25C)T = 27.1 +- 5.5 kcalmol --1). (vi) Similar relative differences are observed in poly(dA --dT) poly(dA --dT) (DeltaHvHA = 333 +- 54 kcalmol --1, DeltaSvHA = 961 +- 157 calK --1mol --1, DeltaGvH(25C)A = 45.0 +- 7.6 kcal mol --1; DeltaHvHT = 213 +- 30 kcalmol --1, DeltaSvHT = 617 +- 86 calK --1mol --1, DeltaGvH(25C)T = 29.3 +- 4.9 kcalmol --1). The methodology employed here distinguishes thermodynamic contributions of base stacking, base pairing and backbone conformational ordering in the molecular mechanism of double-helical B DNA formation. Keywords: INTRODUCTION : An important objective in nucleic acid research is to understand the forces that contribute to the stability of DNA at physiological conditions. Calorimetric, hydrodynamic, mechano-optical and spectroscopic methods have been extensively employed for this purpose ( --). A common aim of these studies is to assess the dependence of thermodynamic melting parameters, such as changes in free energy (DeltaG), enthalpy (DeltaH) and entropy (DeltaS) on specific structural or environmental factors, including nearest neighbor interactions of the bases, backbone conformation, phosphate electrostatic environment, degree of helix hydration, superhelical density, protein binding, small molecule ligation, and the like ( --). Calorimetric (DeltaHcal) and van't Hoff (DeltaHvH) enthalpies of DNA melting and derived thermodynamic parameters have been extensively reported ( --). Current data tabulations rely primarily on the results of differential scanning calorimetry (DSC) and ultraviolet (UV) absorption or circular dichroism (CD) measurements (,,). Recently, temperature-dependent Raman spectra have also been used to calculate thermodynamic parameters of DNA melting . The Raman results were found to compare favorably with those obtained from calorimetric and optical spectroscopic determinations. An advantage of the Raman method is that it offers a diversity of spectral bands for thermodynamic analysis. Typically, the Raman spectrum of DNA comprises several dozen well resolved bands, each originating from a specific and highly localized normal mode of vibration of a base, sugar or phosphate moiety . The temperature dependency of a given band directly reflects changes in the local conformation or interactions of the vibrating group to which it corresponds. In principle, Raman melting profiles of double-stranded (ds) DNA can provide thermodynamic parameters governing changes in (i) Watson -- Crick base pairing, (ii) base stacking, (iii) phosphate -- counterion interactions and (iv) phosphodiester conformation. Whereas changes in (i), (ii) and (iii) with increasing temperature are thermodynamically unfavorable to double helix formation, changes in (iv) are favorable, owing to the positive entropy change with increasing torsional freedom of phosphodiester linkages. Raman spectroscopy has the potential to resolve each of these factors and the related thermodynamic constants (,,). DNA structural perturbations that precede the onset of strand separation, or premelting , can also be probed by the Raman approach (,). The use of Raman spectroscopy to investigate thermally induced structure transformations of nucleic acids was initiated in several laboratories in the early 1970s ( --). Since that time, dramatic improvements in the versatility and sensitivity of Raman instrumentation have greatly enhanced the potential of the method (,,). In recent years, studies of thermally induced structural changes of DNA have combined the data of Raman spectroscopy with other approaches, including DSC , sequence analysis , ultraviolet-resonance Raman (UVRR) spectroscopy , CD and Fourier-transform infrared (FTIR) spectroscopies and superheating of solutions at high pressure . In this paper, we describe methods for thermodynamic analysis of temperature-dependent Raman spectra of double-helical B DNA. The methodology, which has the capability to delineate enthalpic and entropic contributions to DNA melting and premelting transitions, is applied here to the sequence isomers, poly(dA --dT)poly(dA --dT) and poly(dA)poly(dT). High-resolution Raman spectral data are available for both of these DNA structures (,), and the premelting and melting phases of their denaturation processes are of considerable interest (, --). The thermodynamic parameters calculated from the Raman data are compared with results obtained from previous DSC and optical (UV and CD) spectroscopic analyses. The present procedures are based only upon the assumption that thermally induced changes in Raman band intensities (or wavenumber values) reflect structural modifications in the DNA subgroups to which the spectral bands are assigned. With reliable assignments this approach can be adapted to other DNA structures as well as to complexes of DNA with proteins and other ligands. MATERIALS AND METHODS : Sample preparations | Poly(dA --dT)poly(dA --dT) and poly(dA)poly(dT) were purchased as sodium salts from Amersham Pharmacia Biotech (Alameda, CA) and used without further purification. Weighed samples were dissolved to 30 --40 mg/ml in H2O containing 100 mM NaCl at pH 7.0 +- 0.1. Aliquots (similar6 microl) of the polydeoxynucleotide solutions were degassed, sealed in glass capillaries (Kimax No. 34502) and maintained at a constant specified temperature during data collection protocols . Raman spectroscopy | Raman spectra were excited with the 514.5-nm line of an argon laser (Innova 70; Coherent Inc., Santa Clara, CA) using similar200 mW of radiant power at the sample. Spectra were collected in the 90 scattering geometry using a single monochromator spectrograph (Spex 500M; ISA, Edison, NJ) of high spectral resolution (+-3 cm --1) and signal throughput. The instrumentation utilizes a holographic bandpass filter to eliminate interfering laser emissions, a notch filter to reject Rayleigh scattering and a liquid-nitrogen cooled charge-coupled device detector. Typically 7 --12 accumulations of 10 s each were averaged to generate the spectral data from which thermodynamic constants were calculated. In terms of overall signal-to-noise ratio, the Raman spectra of poly(dA --dT) poly(dA --dT) and poly(dA)poly(dT) were improved by similar102 over previously published data (,). Further details of the spectrometer and data collection protocols have been described (,,,). Raman spectra were collected at intervals of 5C in the range 5 --95C on samples maintained to within +-0.5C of the temperature indicated, which reflects the solution temperature near the focus of the laser beam as measured by a calibrated thermocouple insert. Raman intensities were normalized using the peak height of the band at 1092 cm --1, which is assigned to the PO2 -- symmetric stretching mode of the polydeoxynucleotide phosphate group. The peak height of the 1092 cm --1 band is a reliable intensity standard for both native and synthetic DNAs of the B conformation throughout the temperature range 10 --90C (,). The invariance of the 1092 cm --1 band intensity was verified independently in the present study using the 980 cm --1 band of SO42 -- (Na2SO4, added as an intensity standard). Digital subtractions of spectra for measurement of intensity differences of DNA were carried out as in previous work (,). Thermodynamic parameters | Using the formalism of Breslauer and co-workers (,,) and others (,,), we assume a reversible transition between ds and single-stranded (ss) DNA. At temperature T the fractions of molecules present as ssDNA and dsDNA are alpha(T) and 1 -- alpha(T), respectively. For a transition profile that is symmetrical with respect to the median melting temperature (Tm), alpha(T) = x/2a, where x is the absolute value of the difference between the ordinate at T and at a temperature corresponding to the duplex (alpha = 0), and 2a is the maximum change in ordinate between the two temperatures, as shown in Figure . Equilibrium melting properties are extracted from the thermal profile of a Raman band fitted to a plot of alpha(T) versus T . The van't Hoff transition enthalpy is derived from the equilibrium constant K(T) for dsDNA dissociation as follows: DeltaHvH = RT2{[d ln K(T)]/dT}T=Tm = --R{d[ln K(T)]/d(1/T)}T=Tm1 In terms of alpha(T), Tm and the reaction stoichiometry (n = 2), DeltaHvH is given by: DeltaHvH = (2n + 2)RT2m(partialalpha/partialT)T=Tm = 6RTm2(partialalpha/partialT)T=Tm2 The standard-state free energy change (DeltaG) is given by: DeltaG = --RT ln K(T) = --RT ln {[alpha(T)XT]/[1 -- alpha(T)]n}3 where XT = (n/CT)n --1 for non-self-complementary strands and XT = (1/nCTn --1) for self-complementary strands, and CT is the total strand concentration . The van't Hoff entropy (DeltaSvH) is given by: DeltaSvH = DeltaHvH/Tm4 Also, the size of the cooperative unit for the two-state transition, , can be inferred by comparing DeltaHvH with the calorimetric enthalpy, DeltaHcal : = DeltaHvH/DeltaHcal5 Data analysis | Curve fitting of Raman melting profiles and calculation of the van't Hoff enthalpy and entropy of melting. Figure illustrates non-linear curve fitting of an idealized temperature-dependent Raman band exhibiting hypochromic behavior. Such a two-state melting profile can be fitted satisfactorily by a simple sigmoidal function f(T), given by: f(T) = {(2a + b)exp[g(T -- Tm)] + b}/{exp[g(T -- Tm)] + 1}6 where Tm is the melting transition temperature, and a and b are positive constants defining the lower and upper limits of the ordinate (Raman spectral band intensity). The steepness of the transition is represented by the factor g, which is positive for a hypochromic band and negative for a hyperchromic band. In accordance with the definition of alpha(T), equation 6 can be rearranged to yield: alpha(T) = [f(T) -- b]/2a and (partialalpha/partialT)T=Tm = g/4 for a hypochromic band7 and alpha(T) = [2a + b -- f(T)]/2a and (partialalpha/partialT)T=Tm = --g/4 for a hyperchromic band8 With equations 7 and 8, the thermodynamic parameters obtained from a particular Raman band are: DeltaHvH = (3/2)RTm2|g|9 DeltaSvH = (3/2)RTm|g|10 Curve fitting and derivation of thermodynamic parameters were performed using the SigmaPlot 6.0 software package (SPSS Inc., Chicago, IL). Calculation of the base stacking free energy. The base stacking free energy (DeltaGst) of dsDNA is defined as the lowering in free energy that results from stacking of Watson --Crick base pairs. A typical Raman melting curve exhibits variable slope with a maximum at the mid point (Tm) of the transition . At Tm, the normalized Raman intensity is b + a and the slope is ag/2. DeltaGst can be calculated from the slope of the Raman melting curve at T = Tm : Tm(partialalpha/partialT)T=Tm = 0.13tau2/3 (for self-complementary strands)11a or Tm(partialalpha/partialT)T=Tm = 0.13(tau/2)2/3 (for non-self-complementary strands)11b where tau = exp[ --(DeltaGst/RTm)]12 In terms of the experimental parameter g of the Raman melting curve (equation 7 or 8), we obtain from equation 12: DeltaGst = [ --(3RTm/2)] ln [(|g|Tm)/0.52] (for self-complementary strands)13a or DeltaGst = [ --(3RTm/2)] ln {[(|g|Tm)/0.52](1/0.63)} (for non-self-complementary strands)13b The slope of the melting transition also provides information on the average length () of non-melted segments of dsDNA at Tm : = (tm/0.302)[(partialalpha/partialT)T=Tm] = (|g|tm)/1.20814 where tm is the melting temperature in C units. Finally, the melting cooperativity coefficient (sigma) can be calculated from through the relationship : sigma = (1/)215 Figure 1 | Idealized melting profile for a Raman band of dsDNA that is intrinsically hypochromic with respect to double-strand formation. Idealized melting profile for a Raman band of dsDNA that is intrinsically hypochromic with respect to double-strand formation. The abscissa is the temperature (T) and the ordinate is the normalized intensity (Isigma) of the Raman band at wavenumber sigma. The low temperature limit (left asymptote) corresponds to the duplex structure and the high temperature limit (right asymptote) corresponds to the random-coil single strand. Labels refer to the parameters of the empirical fitting function of equation 6, the definition of alpha(T), and the slope at the mid-point (arrow) of the transition (see text). The melting temperature Tm is defined as the temperature at which alpha(T) = 1/2 and Isigma = a + b. RESULTS : Derivation of thermodynamic constants from Raman spectra | Detailed Raman band assignments and structural interpretations for the temperature-dependent Raman spectra of poly(dA --dT)poly(dA --dT) and poly(dA)poly(dT) have been given previously (,). At physiological temperature, both sequence isomers adopt a conformation of the B-DNA type, and both exhibit well-defined premelting and melting transitions. For poly(dA --dT)poly(dA --dT), the temperature domains of premelting (10 < t < 66C) and melting (66 < t < 75C) are similar5 --6C lower than those of poly(dA)poly(dT) (10 < t < 70C and 70 < t < 80C, respectively). At temperatures above 80C, no further structural changes are evident from the Raman spectra. Although both poly(dA -- dT)poly(dA --dT) and poly(dA)poly(dT) exhibit Raman markers of the B-DNA conformation, their Raman spectra are not identical at any temperature prior to the onset of melting. A distinctive Raman signature persists for each duplex throughout the premelting phase; conversely, the melting signatures of poly(dA --dT)poly(dA --dT) and poly(dA)poly(dT) are very similar. The Raman bands of poly(dA --dT)poly(dA --dT) and poly(dA)poly(dT) that exhibit measurable temperature dependency are listed in the first columns of Tables and , respectively. The bands are grouped from top to bottom within the table in accordance with their assignment to adenine only, thymine only, or the sugar --phosphate backbone. Other columns of the tables list parameters relating to equations 6 --8 and 11 --15. Thermodynamic melting parameters obtained from equations 1 --5 and 9 --10 are given for both duplexes in Table . Similar analyses applied to the pre melting phases of poly(dA --dT)poly(dA --dT) and poly(dA) poly(dT) yield the data of Table . In the foregoing analyses, we have focussed on the use of Raman bands that represent highly localized vibrations and exhibit minimal spectral overlap. Nevertheless, the intrinsic cooperativity of DNA melting imposes some degree of coupling between structural transitions monitored by Raman markers of A and T. Accordingly, the thermodynamic parameters derived from Raman markers of each base represent primarily, but not exclusively, the specific contribution of that base. More detailed discussions of these tabulations are given below. Melting and premelting transitions | To characterize the melting transitions of poly(dA --dT) poly(dA --dT) and poly(dA)poly(dT), we obtained temperature profiles (melting curves) for all Raman bands of the deoxyadenosine (dA) and thymidine (dT) residues that exhibited behavior consistent with a two-state transition (Fig. and equation 6). Several such bands were identified by Movileanu et al. . Non-linear curve fitting was applied to each band to determine the midpoint of the transition (tm) and the corresponding steepness factor (g) (Tables and ). Another Raman indicator of melting cooperativity is the temperature range (Deltatm) over which 90% of the total change in Raman band intensity occurs, in accordance with equation 6 and Figure . The calculated Deltatm values for each band of poly(dA --dT)poly(dA --dT) and poly(dA)poly(dT) are included in Tables and , respectively. The data indicate for poly(dA --dT)poly(dA --dT) an average melting temperature = 71.9 +- 1.2C, and for poly(dA)poly(dT) = 76.2 +- 0.6C. Temperature profiles for representative Raman bands of the dA and dT residues of poly(dA --dT)poly(dA --dT) are shown in Figure . Corresponding data for poly(dA)poly(dT) are shown in Figure . Each band has been assigned to a vibration localized largely within the base residue and each exhibits Raman intensity that is sensitive to base unstacking with increasing temperature (,). Interestingly, the band of dT near 1144 cm --1 is hyperchromic; all others are hypochromic with respect to base stacking. Despite large intensity changes in the melting domain, all Raman bands depicted in Figures and show little or no intensity change in the premelting domain. Conversely, certain other Raman bands exhibit temperature dependence that departs substantially from the idealized two-state transition behavior of Figure (; data not shown). Such bands generally occur within the spectral interval 600 --900 cm --1 and have been assigned to vibrational modes localized mainly in the deoxyribose-phosphate moiety . Several are included in the bottom sections of Tables and . The temperature-dependent behavior of these bands is considered typical of non-cooperative structural change. Base stacking (melting) free energy | The base stacking free energy (DeltaGst) is strongly correlated with the melting temperature (tm) and steepness factor (g) of the melting transition (equation 13) (,). Although DeltaGst may be calculated from the melting profile of any Raman band assigned to a base vibration, those exhibiting a relatively narrow melting interval (Deltatm < 20C) and precisely determined steepness factor (large g) are expected to provide the more accurate determinations. Values of DeltaGst determined from various Raman bands of poly(dA --dT)poly(dA --dT) and poly(dA)poly(dT) are listed in Tables and . Reliable values for the adenine stacking free energy (DeltaGstA) are obtained from dA markers at 728, 1209 and 1579 cm --1. Excellent results are also obtained from the band at 1301 cm --1, which is due predominantly to dA residues. Similarly, the thymine stacking free energy (DeltaGstT) is reliably obtained from thymine markers at 1016, 1182, 1236 and 1375 cm --1. The average adenine and thymine stacking free energies in poly(dA --dT)poly(dA --dT) are DeltaGstA = --6.46 +- 0.33 and DeltaGstT = --6.17 +- 0.17 kcalmolbp --1 (kilocalories per mole base pair). For both base types combined, DeltaGstA,T = --6.31 +- 0.29 kcalmolbp --1. The apparent difference between adenine and thymine stacking free energies in poly(dA --dT) poly(dA --dT) (DeltaDeltaGst similar 0.3 kcalmolbp --1) may be significant. The average adenine and thymine stacking free energies in poly(dA)poly(dT) are DeltaGstA = --7.25 +- 0.28 kcalmolbp --1 and DeltaGstT = --6.53 +- 0.24 kcalmolbp --1. For both base types, DeltaGstA,T = --6.88 +- 0.45 kcalmolbp --1. We find that DeltaGst values of poly(dA)poly(dT) are consistently larger than their counterparts in poly(dA --dT)poly(dA --dT), even though the differences observed between the two duplexes are rather close to the limits of experimental uncertainty (Tables and ). Greater stacking energy in poly(dA)poly(dT) is con sistent with its higher melting temperature compared with poly(dA --dT)poly(dA --dT). Raman markers near 1144 (dT), 1262 (dA), 1513 (dA) and 1673 cm --1 (dT), although affected by base stacking (,), are not well suited for DeltaGst determinations either because of inherent band broadness, low intensity, significant shift of the band center with temperature, overlap with one or more other temperature-dependent Raman bands, low cooperativity (small steepness factor) or a combination of these characteristics. The van't Hoff melting enthalpy, entropy and free energy | The van't Hoff enthalpy (DeltaHvH) of melting, which is defined by equation 1, can be determined from the temperature dependency of the Raman band melting parameter alpha given in equation 2. Table lists values of DeltaHvH determined for poly(dA --dT)poly(dA --dT) and poly(dA)poly(dT) from Raman bands of the bases that exhibit the appropriate properties noted in the preceding section. The enthalpy of melting of poly(dA)poly(dT) clearly exceeds that of poly(dA --dT)poly(dA --dT). Interestingly, in both duplexes Raman bands of dA yield a significantly higher van't Hoff enthalpy than bands of dT. The van't Hoff entropy (DeltaSvH) of the melting transition is given by equation 4. Table lists the values obtained from several Raman bands. The results indicate similar overall melting entropies for the two duplexes, although the dA residues consistently exhibit a higher van't Hoff melting entropy than dT residues. The van't Hoff free energy change for the melting transition at 25C , which is obtained from the relation DeltaGvH(25C) = DeltaHvH -- 298.15DeltaSvH, is also listed for each band in Table . In accord with the DeltaHvH and DeltaSvH results, the computed DeltaGvH(25C) value for dA greatly exceeds that for dT in each duplex. Thus, in the case of poly(dA --dT)poly(dA --dT) we find DeltaGvH(25C)A = 45.0 +- 7.6 kcalmol --1 (kilocalories per mole cooperative unit) and DeltaGvH(25C)T = 29.3 +- 4.9 kcal mol --1, and for poly(dA)poly(dT) we obtain DeltaGvH(25C)A = 60.0 +- 3.2 kcalmol --1 and DeltaGvH(25C)T = 27.1 +- 5.5 kcal mol --1. The number of base pairs in the cooperative melting unit, , can be obtained from the van't Hoff (DeltaHvH) and calorimetric (DeltaHcal) transition enthalpies using equation 5 and previously reported calorimetric data (,). Results are included in Table . For both duplexes, the overall cooperative melting unit falls within a relatively narrow range, roughly 20 < < 45 bp. The apparent cooperative melting unit for dA residues is significantly larger than that for dT residues. The presently measured range for compares favorably with the range indicated for , the average helix length at tm (Tables and ). The Raman-based length parameters and are also in accord with values determined by calorimetric methods . The parameter provides the basis (equation 15) for calculation of the cooperativity parameter sigma, listed in Tables and . Thus, for the dA and dT Raman markers noted above, a modest range of values is observed for the cooperativity parameter, consistent with previous determinations (,). The van't Hoff premelting enthalpy | Table shows that the median premelting temperature (Tpm) of poly(dA --dT)poly(dA --dT) is 32 +- 4C, while that of poly(dA)poly(dT) is 39 +- 1C. Averaging of the van't Hoff premelting enthalpy data for poly(dA --dT)poly(dA --dT) gives DeltaHvHpm = 13.4 +- 2.5 kcalmol --1, and for poly(dA)poly(dT) DeltaHvHpm = 18.0 +- 1.6 kcalmol --1. These results suggest that in each duplex the van't Hoff premelting enthalpy is similar20 times smaller than the van't Hoff melting enthalpy. Figure 2 | Melting profiles for selected Raman bands of dA (A) and dT residues (B) of poly(dA --dT)poly(dA --dT). Melting profiles for selected Raman bands of dA (A) and dT residues (B) of poly(dA --dT)poly(dA --dT). The ordinate DeltaIsigma represents the Raman intensity difference for the band at the indicated wavenumber value. The intensity scale is arbitrary, but the relative intensity changes for the respective bands are accurately represented. Thus, in (A), the hypochromicity of the 1579 cm --1 band is approximately twice that of the 1301 cm --1 band and approximately two-thirds that of the 728 cm --1 band . Normalized melting profiles plotted as alpha(T) versus temperature are shown in (C) and (D). Figure 3 | Melting profiles for selected Raman bands of dA (A) and dT residues (B) of poly(dA)poly(dT). Melting profiles for selected Raman bands of dA (A) and dT residues (B) of poly(dA)poly(dT). Conditions are as given in the legend of Figure . Normalized melting profiles plotted as alpha(T) versus temperature are shown in (C) and (D). Table 1 | Temperature dependency and derived thermodynamic constants for selected Raman bands of poly(dA --dT)poly(dA --dT)a Table 2 | Temperature dependency and derived thermodynamic constants for selected Raman bands of poly(dA)poly(dT)a Table 3 | Thermodynamic parameters for melting transitions of poly(dA --dT)poly(dA --dT) and poly(dA)poly(dT)a Table 4 | Van't Hoff premelting enthalpies and entropies of poly(dA --dT)poly(dA --dT) and poly(dA)poly(dT)a DISCUSSION : Previous spectroscopic and calorimetric studies of dsDNA have established a range of premelting temperatures within which the double helix is conformationally altered but not dissociated into single strands (,,,). These earlier studies show that DNA premelting is not sequence specific, although the phenomenon has been most extensively investigated for DNA molecules containing (dA)n(dT)n tracts (,,,,,). A comparison of previous and present results on DNA containing (dA)n(dT)n tracts is given in Table . In oligonucleotide X-ray crystal structures, the (dA)n(dT)n tract exhibits highly propeller-twisted base pairs, which are compatible with three-centered hydrogen bonds involving an N6H2 donor in the adenine strand with two C4O acceptors in the opposing thymine strand ( --). The X-ray structure suggests a simple mechanism to account for premelting in DNA containing (dA)n(dT)n tracts; namely, the conversion of the three-centered AT hydrogen bonds to conventional Watson --Crick hydrogen bonds. While such a mechanism is consistent with the premelting evidenced in Raman spectra of poly(dA)poly(dT) , it does not account for the similar observation on poly(dA --dT)poly(dA --dT) . Three- centered hydrogen bonds are not observed in oligonucleotide crystal structures containing the alternating d(AT)n tract . We conclude that the premelting transitions monitored by temperature-dependent Raman band profiles do not simply reflect the elimination of three-centered hydrogen bonding between A and T in (dA)n(dT)n tracts. Figure A compares temperature-dependent Raman intensity profiles of the 924 cm --1 bands of poly(dA --dT) poly(dA --dT) and poly(dA)poly(dT). The 924 cm --1 marker, which is assigned to the deoxyribosyl moiety, is very sensitive to premelting in each duplex. For both poly(dA --dT) poly(dA --dT) and poly(dA)poly(dT) the band suffers significant intensity change throughout the premelting range (10 < t < 60C), typical of a non-cooperative structural transition. The van't Hoff premelting enthalpy DeltaHvHpm and premelting entropy DeltaSvHpm associated with the change in deoxyribosyl backbone conformation in each DNA is obtained from the corresponding slope of Figure B. Comparable results for other temperature-dependent backbone markers of the two DNAs are listed in Table . Although the average premelting parameters for poly(dA --dT)poly(dA --dT) (DeltaHvHpm = 13.4 +- 2.5 kcalmolbp --1, DeltaSvHpm of 43.8 +- 8.6 calK --1mol --1) are measurably lower than those for poly(dA)poly(dT) (DeltaHvHpm = 18.0 +- 1.6 kcalmolbp --1, DeltaSvHpm = 57.7 +- 5.4 calK --1mol --1), it is clear that robust premelting transitions occur in both DNAs. The premelting phenomena of poly(dA --dT)poly(dA --dT) and poly(dA)poly(dT) are similar with respect to temperature interval, non-cooperativity, Raman band sensitivity and enthalpic and entropic costs. This suggests a similar molecular mechanism for premelting. We propose that the premelting detected by Raman spectroscopy represents a change in the state of hydration of the double helix and that this change in hydration state is accompanied by a conformational adjustment to the deoxyribosyl-phosphate backbone. We note that a similar conclusion was reached by Herrera and Chaires , who attributed premelting to the disruption and release of an ordered 'spine of hydration' in the relatively narrow minor groove of poly(dA)poly(dT). The alternating d(AT)n sequence of poly(dA --dT)poly(dA --dT), which can also adopt a conformation characterized by a narrow minor groove and an ordered spine of hydration (,), may therefore undergo a structurally and thermodynamically similar premelting transition. This is in complete accord with the similar backbone conformations of poly(dA)poly(dT) and poly(dA -- dT)poly(dA --dT) at low temperature and with the remarkably close agreement between DeltaHvHpm determinations by different methods for DNAs containing (dA)n(dT)n tracts . Although the present results support qualitatively similar premelting mechanisms for poly(dA)poly(dT) and poly(dA -- dT)poly(dA --dT), the two transitions are not identical, as evidenced by the higher enthalpic cost of poly(dA) poly(dT) premelting (DeltaDeltaH similar 4.6 kcalmol --1, Table ). Recently, UVRR and CD spectroscopy have been exploited to probe differences in inter-base hydrogen bonding in oligonucleotides containing (dA)n(dT)n and d(AT)n tracts (,). The UVRR data suggest that three-centered inter-base hydrogen bonding specific to the (dA)n(dT)n tract may contribute similar20% of the total van't Hoff premelting enthalpy (measured by CD as similar15 --20 kcalmol --1, Table ), i.e. similar3 --4 kcalmol --1. The previous UVRR and present off-resonance Raman analyses of van't Hoff premelting differences are thus in excellent agreement. We emphasize, however, that because the premelting-sensitive Raman markers of both poly(dA --dT)poly(dA --dT) and poly(dA) poly(dT) include bands assigned to vibrational modes of the deoxyribose ring and deoxyribosyl-phosphate linkages the conformational change accompanying premelting cannot be considered localized to inter-base hydrogen-bonding sites. As first suggested in molecular modeling computations and later supported by biochemical experiments , hydration and conformation are intimately linked in DNA. Disruption of the ordered spine of hydration affects both backbone helical geometry and inter-base interactions, leading to changes in local (propeller twist) and global (groove dimension) aspects of DNA structure. Raman-based determinations of premelting (tpm) and melting (tm) temperatures, stacking free energy (DeltaGst) and van't Hoff melting parameters (DeltaHvHpm, DeltaHvH, DeltaSvH, DeltaGvH(25C), ) provide independent confirmation that poly(dA)poly(dT) forms a thermodynamically more stable secondary structure than poly(dA --dT)poly(dA --dT). Stacking free energies and melting temperatures are in excellent accord with the scant data available on related structures (,, --). Derived values for the parameters > 25 bp and sigma similar 10 --3 --10 --4 bp --2 are indicative of highly cooperative melting in both polynucleotide duplexes, also consistent with previously reported findings for other double-helical B DNAs (,,,). Figure 4 | (A) Premelting (t < 65C) and melting (t > 65C) profiles of the deoxyribosyl Raman marker at 924 cm --1 in poly(dA --dT)poly(dA --dT) (squares) and poly(dA)poly(dT) (circles). (A) Premelting (t < 65C) and melting (t > 65C) profiles of the deoxyribosyl Raman marker at 924 cm --1 in poly(dA --dT)poly(dA --dT) (squares) and poly(dA)poly(dT) (circles). (B) Semi-logarithmic van't Hoff plots [ln K(T) versus 1/T] for the premelting data of (A). Table 5 | Premelting parameters for DNA containing (dA)(dT) tracts determined by different spectroscopic methods CONCLUSIONS : We have described procedures for the determination of thermodynamic parameters governing base stacking (DeltaGst), van't Hoff premelting (DeltaHvHpm, DeltaSvHpm) and van't Hoff melting (DeltaHvH, DeltaSvH, DeltaGvH(25C), ) transitions in DNA of defined base sequences using the data of temperature-dependent Raman spectra. Applications to poly(dA)poly(dT) and poly(dA --dT)poly(dA --dT) illustrate the capability to differentiate thermodynamic contributions of A, T and deoxyribosyl-phosphate interactions to the structural transformations of these DNA duplexes. The results affirm and quantify melting and premelting phenomena in DNAs containing (dA)n(dT)n and d(AT)n tracts. The findings also suggest that A and T melting events may be less coupled in the homopurine and homopyrimidine tracts of poly(dA)poly(dT) than in the alternating A/T tracts of poly(dA --dT) poly(dA --dT). Importantly, these studies reveal robust premelting phenomena for both (dA)n(dT)n (DeltaHvHpm = 18.0 +- 1.6 kcal mol --1) and d(AT)n (DeltaHvHpm = 13.4 +- 2.5 kcalmol --1) sequences. The observed difference (DeltaDeltaHvHpm = 4.6 kcal mol --1) is proposed as the enthalpic contribution from three-centered inter-base hydrogen bonding in (dA)n(dT)n tracts. This estimate is in good agreement with a recent independent determination employing UVRR spectroscopy . The present analysis confirms that poly(dA)poly(dT) is significantly more thermostable (tm = 76.2 +- 0.6C) than poly(dA --dT)poly(dA --dT) (tm = 71.9 +- 1.2C). Within experimental uncertainty all Raman bands of a given DNA, irrespective of assignment, yield the same melting temperature (Tables and ). The enhanced thermostability of poly(dA)poly(dT) vis-a-vis poly(dA --dT)poly(dA --dT) is reflected in more robust base stacking interactions (DeltaDeltaGst similar 0.5 kcalmolbp --1) and appreciably greater van't Hoff melting parameters (e.g. DeltaDeltaGvH(25C)A,T similar 6 kcalmolbp --1). We also find that in poly(dA)poly(dT) the stacking free energy of A exceeds that of T by an amount (DeltaDeltaGst similar 0.8 kcalmolbp --1) comparable to the stacking free energy difference between the two duplexes. This suggests that efficient stacking of adenines along the poly(dA) strand is sufficient to account for the higher van't Hoff melting enthalpy of poly(dA)poly(dT). It should be noted that in the present treatment the thermodynamic parameters for the melting transitions have been computed for the temperature range pursuant to premelting. The combination of premelting (10 --40C) and melting (40 --85C) further underscores the greater overall stability of poly(dA)poly(dT) vis-a-vis poly(dA --dT) poly(dA --dT). Thus, the sum of van't Hoff premelting and melting enthalpies for poly(dA)poly(dT) is 335 kcalmol --1 versus 285 kcalmol --1 for poly(dA --dT)poly(dA --dT). We have shown that temperature-dependent Raman spectra of poly(dA)poly(dT) and poly(dA --dT)poly(dA --dT) have the capability to distinguish contributions of A and T bases and of backbone moieties to the thermodynamic stability of B DNA. The results delineate three molecular mechanisms contributing to DNA stability: base stacking, which is highly cooperative and extensively perturbs vibrational states of base and sugar --phosphate moieties; base pairing, which perturbs vibrational states of localized base sites and is also coupled to base stacking; and backbone conformational ordering, which is highly uncooperative and dominates the premelting interval. We anticipate that the methods used here should be applicable to other DNA sequences and to specific DNA --ligand complexes. Backmatter: PMID- 12202770 TI - Defects in interstrand cross-link uncoupling do not account for the extreme sensitivity of ERCC1 and XPF cells to cisplatin AB - The anticancer drug cisplatin reacts with DNA leading to the formation of interstrand and intrastrand cross-links that are the critical cytotoxic lesions. In contrast to cells bearing mutations in other components of the nucleotide excision repair apparatus (XPB, XPD, XPG and CSB), cells defective for the ERCC1-XPF structure-specific nuclease are highly sensitive to cisplatin. To determine if the extreme sensitivity of XPF and ERCC1 cells to cisplatin results from specific defects in the repair of either intrastrand or interstrand cross-links we measured the elimination of both lesions in a range of nucleotide excision repair Chinese hamster mutant cell lines, including XPF- and ERCC1-defective cells. Compared to the parental, repair-proficient cell line all the mutants tested were defective in the elimination of both classes of adduct despite their very different levels of increased sensitivity. Consequently, there is no clear relationship between initial incisions at interstrand cross-links or removal of intrastrand adducts and cellular sensitivity. These results demonstrate that the high cisplatin sensitivity of ERCC1 and XPF cells likely results from a defect other than in excision repair. In contrast to other conventional DNA cross-linking agents, we found that the repair of cisplatin adducts does not involve the formation of DNA double-strand breaks. Surprisingly, XRCC2 and XRCC3 cells are defective in the uncoupling step of cisplatin interstrand cross-link repair, suggesting that homologous recombination might be initiated prior to excision of this type of cross-link. Keywords: INTRODUCTION : The anticancer drug cisplatin reacts with the N7 atom of purine bases and forms several types of DNA adduct, including DNA interstrand cross-links (ICLs) and (1,2- and 1,3-)intrastrand cross-links . Although cisplatin has been widely used in cancer chemotherapy for many years, the critical cytotoxic DNA lesion induced by this drug is still debated. Evidence that both 1,2-intrastrand cross-links and ICLs are the critical DNA lesion has been presented. The repair of cisplatin intrastrand adducts has been studied in detail using substrates containing defined site-specific cross-links, and cell extracts or purified repair proteins ( --). The three major cisplatin intrastrand cross-links, 1,2-d(GpG), 1,2-d(ApG) and 1,3d(GpNpG), are all substrates for the mammalian nucleotide excision repair (NER) apparatus . The 1,2-cisplatin adducts are, however, much more poorly recognised, adding support to the argument that they are a critical cytotoxic lesion (,). The less efficient repair of 1,2-intrastrand adducts is believed to result from the smaller degree of helical distortion these cross-links induce (,). In contrast, very little is known about the repair of cisplatin ICLs, which form between guanines on the complementary DNA strands. A number of recent in vitro and in vivo studies indicate that specialised repair reactions might incise one strand of the cross-linked DNA, and therefore uncouple the ICLs formed by the nitrogen mustard mechlorethamine and those induced following psoralen plus UVA treatment (,), initiating repair. These reactions are dependent on the action of the XPF and ERCC1 proteins, but not other members of the mammalian NER apparatus ( --). ERCC1 and XPF form a tight heterodimeric complex that possesses endonuclease activity at the junctions between single-stranded and double-stranded DNA . Two different reactions have been described in biochemical studies of the action of this complex on DNA containing defined ICLs. The first reaction involves the 3'->5' degradation of the DNA surrounding the ICL on one strand by XPF --ERCC1 (in the presence of RPA), leaving a single uncoupled adducted nucleotide attached to the complementary strand. The second involves incisions by the XPF --ERCC1 complex 5' and 3' to the ICL, uncoupling the cross-link . This second reaction requires that the ICL bears an unpaired region located immediately 3' to the ICL and it was suggested that such structures arise during replication and contribute to ICL repair during the S phase of the cell cycle. Very recently it has been demonstrated that human MutSbeta might be involved in the recognition and uncoupling step of psoralen ICL repair . Chinese hamster ovary (CHO) cell lines defective in XPF and ERCC1 are extremely sensitive to cisplatin (,), in contrast to cells defective in other components of the NER apparatus. For agents such as mechlorethamine and psoralen plus UVA this seems likely to arise, at least partly, from an inability to uncouple the ICLs produced by these agents by the pathways outlined above ( --). However, another possible explanation is that ERCC1 and XPF cells are sensitive because they have a dual defect in both the excision repair of cross-links and in homologous recombination. Homologous recombination is known to be an important factor determining survival following treatment with cross-linking drugs . Importantly, the Saccharomyces cerevisiae homologues of XPF and ERCC1 (Rad1 and Rad10) are required for the single-strand annealing subpathway of homologous recombination and there is now good evidence for a role for the mammalian ERCC1 --XPF complex in homology-driven recombination ( --). We sought to understand the extreme sensitivity of mammalian XPF and ERCC1 cells to cisplatin. Measure ments of the repair capacity for the two candidate critical lesions, ICLs and 1,2-intrastrand cross-links indicated that, although both are reduced in XPF and ERCC1 cells, these processes were also defective in other NER mutants. Consequently, excision repair defects do not seem to account for the extreme sensitivity of ERCC1 and XPF cells to cisplatin. This is in contrast to other cross-linking agents such as the nitrogen mustard mechlorethamine, where specific defects in ICL uncoupling do appear to be responsible for the sensitivity of ERCC1 and XPF cells . MATERIALS AND METHODS : Cell lines and culture conditions | The cell lines used in this study are listed in Table . The AA8, UV23, UV42, UV61 and UV96 cell lines were obtained from Dr M. Stefanini (Istituto di Genetica Biochimica et Evoluzionistica, Pavia, Italy), while UV135 was purchased from the American Tissue Culture Collection. The V79, irs1, irs1SF, CHO-K1 and xrs5 cell lines were kindly provided by Prof. J. Thacker (MRC Radiation and Genome Stability Unit, Harwell, UK). All cell lines were grown as monolayers in F12 --Ham HEPES medium (Sigma, Poole, UK) supplemented with 2 mM glutamine and 10% foetal calf serum (FCS). Cells were grown at 37C in a 5% CO2 incubator. Trypsin /versine solution was used to detach cells. Chemicals and enzymes | Cisplatin (100 mg /100 ml injectable aqueous stock solution containing 900 mg /100 ml sodium chloride and 100 mg /100 ml mannitol) was obtained from David Bull Laboratories (Australia). Analytical grade mechlorethamine (nitrogen mustard or HN2) was purchased from Sigma. All enzymes used were purchased from Promega UK. Cytotoxicity assay | Cytotoxicity was determined using the SRB growth inhibition assay, described in detail previously . Determination of DNA interstrand cross-linking using the Comet assay | This was determined using a modification of the Comet assay (,). Exponentially growing cells were treated with the desired concentrations of cisplatin in FCS-free medium for 1 h at 37C to reduce binding to serum protein. The medium was replaced with fresh complete medium and incubation continued for the required post-incubation time. Cells were then trypsinised and diluted to a density of 2.5 x 104 cells /ml and kept on ice. All drug-treated samples, plus one control, were subjected to 12.5 Gy X-irradiation on ice, and an unirradiated control was included. Microscope slides were pre-coated with 1% (w /v) Type-IA agarose and 0.5 ml of cells were mixed with 1 ml of 1% (w /v) Type-VII agarose and spread over a pre-coated slide in duplicate. A coverslip was added and the agarose was allowed to solidify. The coverslips were removed and the slides were placed in lysis solution (100 mM Na2EDTA, 2.5 M NaCl, 10 mM Tris --HCl, pH 10.5) containing 1% Triton X-100 at 4C and incubated for 1 h in the dark. The slides were subsequently washed with ice-cold water for 15 min and this was repeated three times. The slides were then transferred to an electrophoresis tank containing ice-cold alkaline solution (50 mM NaOH, 1 mM Na2EDTA, pH 12.5) and incubated for 45 min in the dark. Electrophoresis was carried out for 25 min at 18 V (0.6 V /cm), 250 mA in the dark. Slides were removed and 1 ml of neutralising solution (0.5 M Tris --HCl, pH 7.5) was added and incubated for 10 min. Each slide was rinsed twice with 1 ml of phosphate-buffered saline (PBS) and allowed to dry overnight at room temperature. Slides were stained with 2.5 microg /ml propidium iodide and comets were analysed using a Nikon DIAPHOT TDM inverted epifluorescent microscope (consisting of a high pressure mercury vapour light source, a 580 nm dichroic mirror, a 510 --560 nm excitation filter and a 590 nm barrier filter) at 20x magnification. Fifty cells were analysed per slide using Komet Assay Software (Kinetic Imaging, Liverpool, UK). The degree of DNA interstrand cross-linking present in a drug-treated sample was determined by comparing the tail moment of the irradiated drug-treated samples with irradiated untreated samples and unirradiated untreated samples . The level of interstrand cross-linking is proportional to the decrease in tail moment (DTM) in the irradiated drug-treated sample compared to the irradiated untreated control. The DTM is calculated by the following formula: % DTM = [1 -- (TMdi -- TMcu) /(TMci -- TMcu)] x 100 where TMdi is the mean tail moment of the drug-treated, irradiated sample, TMci is the mean tail moment of the irradiated control sample and TMcu is the mean tail moment of the unirradiated control sample. Measurement of cisplatin intrastrand cross-links by ELISA | Exponentially growing cells were treated with cisplatin for 1 h and incubated in drug-free medium for various times or assayed immediately. Cells were washed with PBS twices and genomic DNA was isolated using a Qiagen Genomic DNA Tip. Genomic DNA was digested with HindIII, ethanol precipitated and resuspended in water. The DNA concentration was determined fluorimetrically. The competitive ELISA developed by Tilby et al. (,) was used with certain modifications. ELISA plates (high-bind 96-well flat bottom; Greiner) were prepared the day before by coating the wells with platinated DNA. Denatured calf thymus DNA treated with cisplatin to give an adduct level of 35.2 microM /g DNA (provided by M. J. Tilby, University of Newcastle, UK) was diluted 1:2000 in coating buffer (1 M sodium chloride, 50 mM sodium phosphate, 0.02% sodium azide) and 50 microl of this solution was incubated in each well at 37C overnight. ELISA wells were then blocked with 150 microl /well BSA solution (1% w /v in PBS) for at least 30 min at room temperature. Assay samples and the standard (kindly provided by M. J. Tilby) were incubated in boiling water for 5 min to increase immunoreactivity and serially diluted in DB buffer (50 mM sodium chloride, 50 mM sodium phosphate, pH 7.0). ICR4 antibody was diluted 1:60 000 in PBS containing 0.2% BSA, 90 mM sodium chloride, 0.2% Tween-20 and 0.2 mg /ml phenol red and 55 microl of this solution was mixed with 55 microl of serially diluted assay samples and the standard. After 30 min incubation at 37C, 50 microl aliquots were transferred to coated 96-well ELISA plates and incubated for 1 h at 37C. All samples were assayed in duplicate. Each assay plate included wells without sample (only antibody) and wells without antibody (only sample /standard). Following five washes with PBS containing 0.1% (v /v) Tween-20, bound antibody was determined using 50 microl /well biotinylated goat anti-rat antibody (Sigma) diluted 1:2500 in PBS containing 1% BSA, 0.2% Tween-20. Following 30 min incubation at 37C, plates were washed three times with PBS containing 0.1% (v /v) Tween-20 and wells were incubated with 50 microl of streptavidin --beta-galactosidase reagent (Boehringer) (stock dissolved in 1 ml of PBS and diluted 1:10 000 in PBS containing 10 mM MgCl2, 1% BSA and 0.1% Tween-20). Following 30 min incubation at 37C, plates were washed seven times and incubated with 50 microl /well 4-methyl-umbelliferyl beta-d-galactoside (Sigma) (80 microg /ml solution made in PBS containing 10 mM MgCl2 and azide). Plates were covered and placed in the dark for 2 --4 h. To measure the fluorescence of each well, plates were read at 360 nm emission and 465 nm excitation using a Tecan Spectrafluor Plus plate reader. The mean background reading was subtracted from all the readings and the percentage inhibition of maximum fluorescence was calculated for each serial dilution of the assay samples and the standard. Cisplatin intrastrand adduct levels were then calculated as described by Tilby et al. . Analysis of double-strand breaks (DSBs) by pulsed field gel electrophoresis (PFGE) | Cells growing in a monolayer were treated with cisplatin for 1 h, washed with 10 ml of PBS and incubated with fresh medium for the required repair time. Cells were trypsinised, 3 x 106 cells were harvested and PFGE plugs were prepared using the Bio-Rad Mammalian CHEF Genomic Plug Kit, as instructed by the manufacturer. PFGE was performed with a 0.7% gel (Pulse Field Certified Agarose; Bio-Rad) in 0.25x TBE buffer using a Biometra Rotaphor Type V apparatus. Electrophoresis runs were for 120 h at 14C with the following parameters: interval 5000 --1000 s log, angle 110 --100 linear, voltage 50 --45 V linear. On completion the gels were stained with 2 microg /ml ethidium bromide for 1 h, destained overnight with water and photographed. Semi-quantitative data were obtained by measuring the absolute integrated optical density of each lane using Gel Pro Analyser (Media Cybernetics) and calculating the percentage of DNA released from the DNA plug. Table 1 | Chinese hamster cell lines used in this study RESULTS : ERCC1 and XPF cells are highly sensitive to cisplatin | Exponentially growing cells were treated with increasing concentrations of cisplatin for 1 h and growth inhibition measured using the SRB assay. As shown in Figure , the UV47 and UV96 cell lines, defective in XPF and ERCC1, respectively, showed extreme sensitivity to cisplatin compared to their isogenic parent cell line AA8. When IC50 values were compared, XPF and ERCC1 mutants were 40- and 37-fold more sensitive to cisplatin than their parent cell line, indicating the crucial role of this structure-specific nuclease in the repair of cisplatin-induced DNA damage. In contrast, the XPB, XPD and XPG mutant cells were only 1.4-, 1.3- and 3.1-fold more sensitive than the parent cell line, respectively, suggesting a minor role for these NER proteins in the response to critical cisplatin lesions. Induction and repair of DNA ICLs | DNA ICLs are a potentially cytotoxic , but quantitatively minor, adduct (similar1% of the total lesions) formed by cisplatin . Using a modification of the Comet assay, which permits detection of ICLs at the single cell level in vivo, the induction and repair kinetics of cisplatin-induced ICLs were studied. Prior to cell lysis the cells receive a fixed dose of X-rays to induce random strand breakage. The presence of ICLs retards migration of the irradiated DNA during electrophoresis, resulting in a reduced tail moment compared to the non-drug-treated control. Figure A shows the percentage decrease in tail moment with increasing concentrations of cisplatin for the parental, repair-proficient AA8 cell line. A decrease in tail moment occurs linearly up to 30 microM, but no further significant increase was observed at 50 microM. No single-strand breaks were seen in unirradiated samples treated with increasing doses of cisplatin (data not shown). The kinetics of ICL induction and repair in the NER mutants were examined following a 1 h treatment with 50 microM cisplatin (Fig. B). This dose gives 50% inhibition of growth in AA8 cells and induces a readily detectable level of ICLs (similar40% decrease in tail moment). As previously reported , ICL levels peak at similar8 h, followed by gradual repair that is still incomplete at 72 h post-treatment incubation. AA8 and its isogenic NER mutant derivatives (XPB, XPG, XPF and ERCC1) were treated with 50 microM cisplatin for 1 h and the level of ICLs assessed following 0, 24, 48 and 72 h post-incubation in drug-free medium (Fig. C). A comparable level of ICLs was observed in all these cell lines immediately following the 1 h treatment, demonstrating that the differences in sensitivity are not due to differences in the induction of ICLs. Consistent with their extreme sensitivities, the XPF and ERCC1 mutants were defective in the uncoupling step, or incision, of cisplatin-induced ICLs. Surprisingly, the much less sensitive XPB and XPG mutants were also defective in the uncoupling of ICLs. The lack of correlation between sensitivity to cisplatin and ICL incision capacity implies that the differential sensitivities of the NER mutants to cisplatin are not due to a difference in their ability to initiate repair of the ICLs induced by this drug. Induction and repair of cisplatin 1,2-intrastrand cross-links | Since no definite correlation was observed between cisplatin sensitivity and the uncoupling of ICLs in any of the NER mutants, the induction and repair of intrastrand cisplatin adducts was investigated in the parent and NER mutant cell lines. These adducts were measured using a competitive ELISA method developed by Tilby et al. employing a monoclonal antibody which recognises cisplatin 1,2-intrastrand cross-links (; M.Tilby, personal communication). In Figure A, the AA8 parental cell line was treated with increasing doses of cisplatin for 1 h and a linear increase in the induction of cisplatin intrastrand adducts was observed. In subsequent experiments 50 microM cisplatin, the IC50 value in the AA8 parent cell line, was used as it gave a readily measurable level of adducts (9.5 fmol/microg DNA). The repair kinetics of cisplatin intrastrand cross-links in the AA8 parent cell line are shown in Figure B. Following a 1 h drug treatment, the level of intrastrand adducts peaked after 4 h post-treatment incubation, consistent with previous reports . At 24 and 48 h following drug removal 64 and 74% of the adducts were eliminated, respectively. Figure shows the post-treatment elimination of cisplatin 1,2-intrastrand cross-links in the NER mutants. The XPG mutant was most severely defective in the elimination of these adducts. At 48 h following drug removal, 3.7 times more intrastrand adduct remained in the XPG-defective cells than the parent cell line. The XPF and ERCC1 mutants were also defective (both 2.3-fold). As with cisplatin ICLs, it is again striking that there is little correlation between ability to eliminate the cisplatin lesion, this time the intrastrand cross-link, and cellular sensitivity amongst NER mutants. The extreme sensitivity of the ERCC1 and XPF cell lines strongly suggests that they have a dual deficiency in excision repair and another pathway vital for the repair and /or tolerance of critical cisplatin --DNA adducts. Importance of homology-driven recombination in response to cisplatin | There is good evidence that homologous recombination plays a role in the cellular response to cisplatin in lower organisms (,). Therefore, defects in recombination as well as NER may account for the extreme cisplatin hypersensitivity of ERCC1 and XPF mammalian cells. XRCC2 (irs1) and XRCC3 (irs1SF) cells, defective for two Rad51-related recombination factors, are derived from the V79 and AA8 parent cell lines, respectively, and xrs5 cells [XRCC5 mutant cells, defective in the Ku86 protein, required for non-homologous end joining (NHEJ)] are derived from CHO-K1 cells. The cisplatin sensitivities of these mutants, together with their parents, are shown in Figure . Confirming published data , the XRCC2 and XRCC3 mutants showed extreme sensitivity to cisplatin, having 50- and 38-fold higher sensitivities compared to their parent cell lines (Fig. A and B, respectively). These values are similar to those obtained with the XPF and ERCC1 cells. In contrast, the NHEJ mutant xrs5 showed only a slight increase in sensitivity (1.7-fold) compared to the parent cell line, suggesting that the involvement of the NHEJ pathway in the repair of cisplatin DNA damage is minor (Fig. C). In S.cerevisiae the repair of psoralen- and mechlorethamine- induced DNA ICLs involves the formation of DSBs in replicating cells (,,). Similarly, we have clearly demonstrated that DSBs also arise in mammalian (CHO) cells during processing of mechlorethamine-induced ICLs . However, despite these suggestions, the induction of DSBs during the processing of cisplatin-induced DNA damage has not been investigated in detail. To address this, the induction of DSBs following cisplatin treatment was examined using PFGE. Exponentially growing AA8 cells were treated with increasing doses of cisplatin for 1 h and analysed by PFGE. DNA from cells treated with an IC50 dose (15 microM) of mechlorethamine (nitrogen mustard) was included as a positive control. No DSBs were observed following cisplatin treatment at 150 microM (three times the IC50) (Fig. A). As expected, the ICLs induced by nitrogen mustard resulted in the accumulation of DSBs, producing a lower molecular weight smear of DNA released from the plugs. To investigate whether cisplatin DSBs arise over longer times, AA8 cells were treated with 100 microM cisplatin for 1 h and DSBs were monitored over a 0 --24 h period. No DSBs were observed at any of the time points analysed (Fig. B). The XRCC2 and XRCC3 homologous recombination mutants are highly sensitive to nitrogen mustard and this correlates with an impaired ability to repair HN2-induced DSBs . XRCC2 and XRCC3 mutants are also extremely sensitive to cisplatin. If the level of induction of DSBs by cisplatin is very low and if they are readily repaired in recombination-proficient AA8 cells, it is possible that this PFGE approach is not sensitive enough to detect these DSBs. If the high sensitivity of XRCC2 and XRCC3 mutants to cisplatin is associated with impaired repair of low levels of DSBs, these mutants should accumulate DSBs following cisplatin treatment. To investigate this possibility, the induction of DSBs was examined in the wild-type (AA8 and V79), XRCC2 and XRCC3 mutants over a 24 h time period, following 1 h treatment with up to 100 microM cisplatin. These experiments failed to demonstrate the occurrence of any DSBs (data not shown). These results suggest that DSBs do not occur in the processing of cisplatin-induced DNA damage, in contrast to other common agents that induce ICLs. Finally, we determined whether a recombination mutant (XRCC3, irs1SF) is competent to initiate ICL repair, specifically at the cross-link uncoupling stage. Cells were treated with 50 microM cisplatin for 1 h and ICL uncoupling followed during 72 h post-treatment incubation in complete medium, using the modified Comet assay. XRCC3 cells were defective in the uncoupling of ICLs (Fig. C). This surprising result indicates that the uncoupling stage of cisplatin ICL repair might be dependent upon the formation of early recombination intermediates. This contrasts with nitrogen mustard, where XRCC3-defective cells have been shown to be capable of incising the ICLs induced by this agent . Figure 1 | Cisplatin sensitivity of parental AA8 cells and NER mutants XPB (UV23), XPD (UV42), XPF (UV47), ERCC1 (UV96) and XPG (UV135). Cisplatin sensitivity of parental AA8 cells and NER mutants XPB (UV23), XPD (UV42), XPF (UV47), ERCC1 (UV96) and XPG (UV135). Exponentially growing cells were treated with increasing concentrations of cisplatin for 1 h, incubated in drug-free medium for 3 days and stained with SRB. Fraction of growth inhibition was calculated for each dose as described in Materials and Methods. All results are the means of at least three independent experiments and error bars show the standard error of the mean. Figure 2 | (A) Decrease in comet tail moment (TM) with increasing concentration of cisplatin. (A) Decrease in comet tail moment (TM) with increasing concentration of cisplatin. AA8 cells were treated with increasing doses of cisplatin for 1 h and analysed immediately. For each drug concentration, the TM of 50 comets was measured and the mean values were calculated. The percentage decrease in TM in comparison with untreated samples was then calculated (Materials and Methods). The results shown above are the means of three independent experiments and error bars show the standard error of the mean. (B) The early repair kinetics of ICL repair in AA8 cells treated with 50 microM cisplatin for 1 h and samples removed at increasing times for Comet analysis. The results presented are the means of at least three independent experiments and error bars show the standard error of the mean. (C) Induction and repair of cisplatin ICLs in parental and NER mutant cell lines. WT (AA8), XPB (UV23), XPG (UV135), XPF (UV47) and ERCC1 (UV96) cells were treated with 50 microM cisplatin for 1 h and incubated in drug-free medium for 24, 48 and 72 h to allow repair. Percentage decrease in TM in comparison with untreated samples was calculated as above. The results presented are the means of at least three independent experiments and error bars show the standard error of the mean. Figure 3 | (A) Induction of cisplatin intrastrand cross-links following treatment of AA8 cells with increasing doses of cisplatin. (A) Induction of cisplatin intrastrand cross-links following treatment of AA8 cells with increasing doses of cisplatin. Exponentially growing AA8 cells were treated with cisplatin for 1 h and the level of intrastrand adducts were analysed immediately using competitive ELISA. (B) Kinetics of cisplatin intrastrand adduct removal in AA8 cells. Exponentially growing AA8 cells were treated with 50 microM cisplatin for 1 h, incubated in drug-free medium for the stated times and the level of intrastrand adducts quantified using competitive ELISA. The results are the means of five independent experiments and the error bars show the standard error of the mean. Figure 4 | Elimination of cisplatin intrastrand cross-links in wild-type (AA8) cells and NER mutants XPG (UV135), XPF (UV47) and ERCC1 (UV96). Elimination of cisplatin intrastrand cross-links in wild-type (AA8) cells and NER mutants XPG (UV135), XPF (UV47) and ERCC1 (UV96). Cells were treated with 50 microM cisplatin for 1 h and adduct levels quantified immediately following drug removal (0 h) and 24 and 48 h later using competitive ELISA. The results presented are the means of between three and five individual experiments and error bars are the standard error of the mean. Figure 5 | Cisplatin sensitivity of CHO recombination mutants. Cisplatin sensitivity of CHO recombination mutants. (A) XRCC2 (irs1) and its parental cell line V79; (B) XRCC3 (irs1SF) and parental AA8 cells; (C) the NHEJ mutant XRCC5 (xrs5) and its parental cell line CHO-K1. Cells were treated with increasing concentrations of cisplatin for 1 h, incubated in drug-free medium for 3 days and stained with SRB. All results are the means of at least three independent experiments and error bars are the standard error of the mean. Figure 6 | (A) PFGE analysis of DNA from cells treated with increasing doses of cisplatin. (A) PFGE analysis of DNA from cells treated with increasing doses of cisplatin. Exponentially growing AA8 cells were treated with the stated concentrations of cisplatin or 15 microM nitrogen mustard for 1 h and immediately analysed for the presence of DSBs on pulsed field gels. (B) PFGE analysis of DNA from cisplatin-treated cells subject to further post-treatment incubation. Exponentially growing AA8 cells were treated with 100 microM cisplatin for 1 h and incubated in drug-free medium for the stated time prior to PFGE analysis. Lane C contained DNA from mock- treated cells subject to 24 h post-treatment incubation. (C) ICL uncoupling in recombination-defective cells. irs1SF (XRCC3 mutant) cells and their isogenic parent AA8 were treated with 50 microM cisplatin for 1 h and incubated in drug-free medium for 24, 48 and 72 h to allow repair. Percentage decrease in TM in comparison with untreated samples was calculated as above. The results presented are the means of at least three independent experiments and error bars show the standard error of the mean. DISCUSSION : It has been established that both NER and homology-driven recombination activities are important for the elimination of cisplatin DNA adducts in Escherichia coli (,) and yeast (,) cells. In agreement with previous reports (,), the results obtained in this study confirm that XPF- and ERCC1-defective mammalian cells are extremely sensitive to cisplatin. In contrast, the other NER mutants considered here (XPB, XPD, XPG and CSB) do not possess this extreme sensitivity . This implies that the XPF --ERCC1 structure-specific nuclease plays a key role in the repair of critical cisplatin lesions or repair intermediates. To address whether the excision repair of a particular lesion is specifically reduced or absent in XPF- and ERCC1-defective cells we have measured the capacity of a set of NER mutants to eliminate two cisplatin adducts which constitute critical DNA damage induced by this agent; the ICL and 1,2-intrastrand cross-link. Both the highly sensitive XPF and ERCC1 mutants and the slightly sensitive XPB and XPG mutants were defective in the uncoupling of ICLs, indicating that a full complement of NER proteins is required for the excision of cisplatin ICLs. No clear correlation was observed between cellular sensitivity and ICL uncoupling capacity. Although the XPB and XPG mutants were completely defective in the uncoupling step of ICL repair, they were only 1.4- and 3.1-fold more sensitive to cisplatin than the parent cell line, compared to a 37 --40-fold increase in sensitivity observed in XPF and ERCC1 mutants. Meyn et al. demonstrated that another ERCC1 mutant CHO cell line, UV20, was extremely sensitive to cisplatin and defective in the uncoupling of interstrand cross-links, and this was taken to demonstrate a direct relationship between sensitivity and ICL repair. However, the ICL repair capacity of the less sensitive NER-defective cell lines was not investigated. Here, for the first time, we show that, in addition to XPF --ERCC1, other components of the NER machinery are essential for the excision of cisplatin ICLs in vivo. The data suggest that ICLs are not, by themselves, the critical DNA lesion induced by cisplatin since an inability to repair this lesion only leads to a very modest increase in cellular sensitivity. These results are very different from those obtained with nitrogen mustard , where a direct correlation between mutant sensitivity and the ability to uncouple ICLs was evident. In this case the XPB and XPG cells were only slightly drug sensitive and were proficient in the uncoupling of nitrogen mustard-induced ICLs, whereas the highly sensitive XPF and ERCC1 cells were severely defective. In addition, recent biochemical studies employing purified proteins (,) have shown that the uncoupling step of psoralen ICL repair can be achieved by the XPF --ERCC1 nuclease in the absence of other factors (except RPA in the former study). A further cross-link-specific excision reaction has been identified by Bessho et al. . Here the complete human excision nuclease makes dual incisions 5' to a psoralen cross-link on one strand of the damaged DNA only. Although the authors state that this intermediate might act as an initiation signal for recombination, such a reaction does not release the cross-link and therefore probably does not represent the uncoupling reactions seen in our study. Further biochemical studies employing defined cisplatin ICL-containing substrates are required to elucidate the molecular nature of this reaction. A surprising observation was that the XRCC2 and XRCC3 mutants were unable to uncouple the cisplatin ICLs (Fig. C). This raises the possibility that homologous recombination is initiated prior to the incisions which uncouple the cisplatin cross-link and that the incision reactions depend upon an early recombination intermediate being formed. Perhaps a homology search, requiring the Rad51 family of proteins (which includes Xrcc2 and Xrcc3), precedes the uncoupling reaction. While this is possible, it should be emphasised that studies of the enzymology of mammalian homologous recombination have yet to identify such reactions. Nevertheless, given that the uncoupling incisions for cisplatin ICLs appear significantly different to those observed for nitrogen mustard ICLs this coordination might help explain the differing requirements for NER proteins in the uncoupling step. The removal of cisplatin 1,2-intrastrand cross-links was reduced in all the NER mutants tested. Our results are consistent with those obtained using purified proteins to reconstitute the excision repair of cisplatin intrastrand cross-links in defined substrates ( --). We showed that XPG, ERCC1 and XPF mutant cell lines were all significantly defective in the repair of this class of adduct. Again, the absence of a correlation between sensitivity and excision repair capacity is apparent. This strongly suggests that the ERCC1 --XPF complex plays an additional role in the repair of cisplatin DNA damage. Confirming published data , XRCC2 and XRCC3 mutants were extremely sensitive to cisplatin, demonstrating that homology-driven recombination plays a major role in the repair of cisplatin DNA damage. Several other emerging homologous recombination components, including Rad51, Brca1 and Brca2, have also been implicated in the mammalian response to cisplatin DNA damage ( --) and mutations in these genes render cells sensitive to cisplatin. Brca1 promotes assembly of subnuclear Rad51 foci following treatment of mouse embryonic stem cells with cisplatin . The recently generated Rad51 family mutant Rad51B -- / --, derived from chicken DT40 cells, shows hypersensitivity to cisplatin . ERCC1 and XPF must now also be considered as candidates for a role in the recombinational processing of cisplatin adducts. A recent study by Zdraveski et al. demonstrated that E.coli mutants defective in both daughter-strand gap and DSB homologous recombination pathways are sensitive to cisplatin, providing evidence that DSBs may arise during the repair of cisplatin ICLs in this organism. Surprisingly, however, no DSB formation was observed during the processing of cisplatin-induced DNA damage in this study, even in XRCC2 and XRCC3 cells. Previous studies in our laboratory, and the work of others, has indicated that the induction of DSBs by psoralen and nitrogen mustard ICLs might be due to the collapse of replication forks (,,). Significantly, mammalian cells possess enzymes able to replicate DNA containing cisplatin adducts. DNA polymerase beta can perform extensive bypass synthesis on platinated DNA in vitro and DNA polymerase eta, which is defective in XPV patients, can also bypass cisplatin intrastrand cross-links efficiently and with high fidelity . Therefore, the absence of DSBs following cisplatin treatment could be a result of the bypass of cisplatin adducts or indicate that when replication forks encounter cisplatin cross-links DSBs do not result. It is conceivable that the ERCC1 --XPF nuclease plays some role at this stage, perhaps in the recombinational restart of replication. Collectively, these studies show that ERCC1 and XPF are vital to the mammalian cell response to this important anticancer agent. It appears that the major contribution made by these factors is largely independent of their role in excision repair. It is highly unlikely that an additional, as yet uncharacterised, critical cisplatin adduct exists which relies entirely upon XPF --ERCC1 for repair. Future work will focus on identifying defects in cisplatin-induced recombination pathways and end-points resulting from loss of ERCC1 and XPF. Defined intrachromosomal repeat recombination substrates have already proved valuable in defining spontaneous recombination defects in ERCC1 and XPF cells (,) and the use of these constructs for the analysis of recombination events induced by different DNA-damaging agents (and the requirement for individual mammalian recombination factors) should be explored. The mechanisms by which homology-driven recombination pathways process cisplatin lesions need to be elucidated since the ability to selectively sensitise cells to cisplatin by inhibiting this pathway could substantially enhance the therapeutic efficacy of this drug. Backmatter: PMID- 12202766 TI - Definition and prediction of the full range of transcription factor binding sites ---the hepatocyte nuclear factor 1 dimeric site AB - In animals, transcription factor binding sites are hard to recognize because of their extensive variation. We therefore characterized the general relationship between a specific protein-binding site and its DNA sequence and used this relationship to generate a predictive algorithm for searching other DNA sequences. The experimental process was defined by studying hepatocyte nuclear factor 1 (HNF1), which binds DNA as a dimer on two inverted-repeat 7-bp half sites separated by one base. The binding model was based on the equivalence of the two half sites, which was confirmed in examples where specific modified sites were compared. Binding competition analysis was used to determine the effects of substitution of all four bases at each position in the half site. From these data, a weighted half-site matrix was generated and the full site was evaluated as the sum of two half-site scores. This process accurately predicted even weak binding sites that were significantly different from the consensus sequence. The predictions also showed a direct correlation with measured protein binding. Keywords: INTRODUCTION : Despite extensive knowledge of DNA binding by transcription factors, determination of binding sites requires detailed experimentation, an approach that is incompatible with the extraordinary amount of new genomic sequencing. Computer prediction will be essential, but will require more exact prediction methods than those now in use. The current methods frequently predict sites that do not bind . Conversely, they also miss many important sites . Computer prediction is especially inaccurate for relatively weak binding sites. These weaker sites differ from the consensus and are hard to find; nevertheless, they can be important. The most dependable widely used computer prediction systems utilize weight matrices, two-dimensional tables compiled from alignments of sequences in a database (reviewed in ,). Such databases, however, are biased towards easily recognized strong binding sites. There are a few examples of alternate approaches where matrices have been derived from direct experimental data, e.g. functional assays of transcription or DNA binding . Like this latter study, we started with measurement of DNA binding. To develop a general predictive algorithm derived from a factor's DNA-binding properties, we focused on a single well characterized factor, hepatocyte nuclear factor 1 (HNF1). Our analysis took advantage of two resources: the large number of HNF1 sites that have been experimentally defined (reviewed in ), and a study by Cereghini et al. that provided detailed competition analysis for a set of binding sites. HNF1 binds DNA as a homo- or heterodimer of two peptides, HNF1alpha and HNF1beta (or vHNF1), and its binding sites are critical in liver gene regulation (,). The albumin promoter site at --60, for example, is essential for both direct promoter activity (,) and regulation by distant enhancers . HNF1alpha and beta are members of the homeobox family of transcription factors. Their divergent homeodomains are highly similar to each other and bind to sites that are experimentally indistinguishable . The HNF1 DNA-binding domain also includes a region derived from the first two alpha-helical regions of a POU DNA-binding domain, on the N-terminal side of the homeodomain. In an intact POU domain, this region binds to the DNA backbone but without any base-specific contacts , suggesting that this extra HNF1 domain enlarges the region of DNA --protein interaction without affecting binding-site specificity. HNF1alpha and beta also have a highly conserved dimerization domain at the N-terminal. The current study was undertaken with three goals: to fully characterize the binding interactions of HNF1 as a resource for the study of liver gene expression; to define an algorithm that can rigorously predict HNF1 binding to DNA at any sequence, identify even weak new HNF1 sites and predict the strength of their binding interactions; and to work out a straightforward laboratory process that can generate similar algorithms for other DNA-binding factors starting from limited information about their binding sites. MATERIALS AND METHODS : Plasmids and HNF1 purification | Plasmid 6HIS /HNF1DB encodes the dimerization and DNA-binding domains of HNF1alpha (residues 1 --281) fused to a six-histidine peptide in pET-14b (Novagen), with MW = 32 215 Da. Following transfection into Escherichia coli strain BL21DE3pLysS (Novagen), peptide was affinity-purified on Ni-NTA resin (Novagen) according to the supplier's protocols. Protein was quantified using the Bradford method (Biorad) and purity was verified on SDS --acrylamide gels. Oligonucleotide binding sites | All binding site oligonucleotides had the same design except for substitutions at numbered positions in the two 7-bp HNF1 half sites (underlined): top strand, 5'-tcgaTGTG1G2T3T4A5A6T7GA9T10T11A12A13C14C15GTT-3'. Unpurified complementary oligonucleotides, as provided by the supplier, were dissolved, annealed at 68C for 15 min and allowed to cool slowly to room temperature. The complementary oligonucleotides were designed with seven asymmetric base pairs to minimize self-annealing. A few of the annealed mixtures were analyzed on 20% acrylamide /6 M urea gels, which showed a single duplex band that contained 80 --90% of the detectable DNA, with very weak bands of unannealed full-length and shorter oligonucleotides. In one experiment, annealed oligonucleotides were purified from the gels and used in competition assays. These competitions were not significantly different from those that used unpurified oligonucleotide duplexes. Labeling of four-base 5'-protruding ends (lower case) by fill-in with Klenow DNA polymerase I and other details of gel-shift assays were carried out as previously described . Gel-shift analyses | Protein-binding reactions were carried out in a 10 microl volume containing 25 mM HEPES, pH 7.9, 100 mM NaCl, 9 mM MgCl2, 0.25 mM EDTA and 18% glycerol. Reactions were incubated for 10 min at room temperature before loading the electrophoresis gel. Electrophoresis, in 6% acrylamide gels and 0.5x Tris-borate --EDTA running buffer, was also carried out at room temperature. Gels were dried and the signals quantified by densitometry of autoradiograms. DNA competition analysis used 4 ng of HNF1 protein, a constant amount (2 ng) of labeled oligonucleotide duplex as the primary binding site and a variable ratio (0.25 --800-fold) of an unlabeled competitor binding site. Data were fitted to sigmoidal curves and the ratio of competitor to labeled probe at 50% competition (C1 /2) was determined by interpolation. The C1 /2 determinations represent data accumulated from two to four experiments. For analysis of protein binding and dissociation , 1 ng of labeled DNA was incubated with varying amounts of recombinant HNF1alpha protein. Binding fraction was plotted versus protein concentration and fitted to sigmoidal curves. To compensate for the presence of a non-binding oligonucleotide component, the binding and non-binding fractions were measured and binding was calculated as the ratio of binding to binding at saturation . For Scatchard analysis (,), values below 80% were used in double logarithmic plots, which were fitted to linear equations for slope determination. Matrix compilation and weighting | Using Excel (Microsoft), a program was written to calculate a total matrix score for a specific binding site and normalize these values to the range of possible values (minimum 0, maximum 100%). The matrix scores were plotted in semilogarithmic plots versus 1 /C1 /2 and the fit was measured by linear regression analysis. For scanning DNA sequences, a computer program was written to calculate two matrix scores at a specified interval, combine them as a sum or product, normalize the scores to the maximum possible value, and report all sites detected above a specific cut-off. Methods for weighting specific values in the matrix are described below. RESULTS : Preliminary analysis | A database of 65 HNF1 binding sites was compiled and the sequences aligned according to an idealized HNF1 binding site, GGTTAAT N ATTAACC , hereafter referred to as the model site. Sites in the database showed a wide range of patterns. Half sites ranged from 7 /7 to as few as 1 /7 matches with the model; stronger half sites averaged 6.3, weaker half sites 4.5. Full sites combined similar or divergent half sites in all possible combinations. The average site matched 10.8 of the 14-base model site, but functional sites with as few as 7 /14 matches have been characterized. To provide a more discriminating approach (,), the relative frequency of nucleotides at each position was tabulated in a simple matrix . Scores were calculated from the matrix by adding the value for the base in each position. These scores were correlated with data obtained from a study of a HNF1 binding site competition published by Cereghini et al. . C1 /2 values were calculated from their competition analysis of 12 binding sites. These comparisons (data not shown) showed a nearly linear relationship between 1 /C1 /2 and matrix score for the four strongest sites, but wide divergence for the weaker ones. Since the matrix was generated from a database, it appeared likely that the values could be modified to give a better correlation with measured binding. To simplify this problem, we modeled dimeric binding by using a half-site matrix to represent the monomer and combined two scores to represent the dimer. The seven columns on the left of the full count matrix were used as the initial half-site matrix. Use of the half-site matrix reduced the number of potential variations to be considered by four orders of magnitude, from 414 (2.7 x 108) to 47 (1.6 x 104). Two processes were compared. Scores from the two half sites (separated by one base and on opposite DNA strands) were added (the sum algorithm) or multiplied (the product algorithm) together. The product algorithm was directly analogous to the behavior of dimeric second-order binding and was more sensitive to the contributions of individual half sites than the sum algorithm. It was possible to alter a few numbers in the matrix by trial and error so that either algorithm gave a linear correlation with 1 /C1 /2. Adjustment was relatively easy for these 12 binding sites, because only a few bases differed from the model site, but the analysis demonstrated the feasibility of a more extended approach. This extension, however, required binding analysis from a comprehensive set of sites. Measurement of binding affinity by competition analysis | Binding competition analysis was carried out as described in Materials and Methods, using recombinant HNF1 protein, a labeled oligonucleotide and varying ratios of an unlabeled competitor. Initial studies utilized the model binding site (site 10) as the labeled probe, but this site was too strong to demonstrate competition by weaker binding sites. In contrast, the weaker site 2 demonstrated competition by both stronger and weaker sites over a range of three orders of magnitude and was used for subsequent comparisons of binding competition. For all oligonucleotides in the study, the measured values for C1 /2 were determined as the ratio of competitor that gave 50% competition of site 2 . A representative experiment characterized the effects of C substitutions in positions 2 and 3 and demonstrated that combining the two substitutions caused similar reduction of binding strength whether they were in the same or opposite half sites. Matrix adjustment to reflect binding | Starting with the frequency matrix, the half-site matrix and the half-site matrix that was modified to fit the Cereghini et al. binding data, DNA sequences were searched and test sites selected for competition analysis. These searches all demonstrated a large number of potential binding sites and it was evident that those with higher scores were likely to bind as predicted. Such analysis demonstrated two new binding sites in known regulatory regions (sites 5 and 8), but the discrimination of weaker sites appeared limited. A more systematic approach was carried out in several stages. Each stage improved the discrimination of both the sum and product algorithms. Overall, binding sites were selected from the list of known sites to provide a wide range of variations (sites 1 --8), and other sites were designed (sites 9 --31) so that all possible individual base changes and some specific combinations were included in the analysis . Two processes were used to convert the contributions of individual base changes to a value in the matrix. For single substitutions (sites 9 and 11 --20), a set of C1 /2 values for a series of oligonucleotides was plotted versus matrix score. This set included the model site 10, an oligonucleotide that contained a single base substitution, and several others without the substitution. The degree of correlation was evaluated by least-squares linear regression. The value for the substituted base was systematically varied and the value that gives a maximum R2 value was substituted into the matrix. The same approach was effective for the few sites in our series that compared two substitutions, e.g. sites 12, 14, 17 and 18 in Figure . The process could have been extended to all sites in the matrix, but had several limitations. First, it would have required 28 different competition comparisons to provide only a single measurement for the effects of each base. Secondly, it did not measure the effects of multiple base substitutions that might interact with each other. Thirdly, it could not utilize the vast majority of natural sites, because they contained multiple substitutions. For a set of binding sites with multiple base changes, independent optimization of some bases improved the overall fit, while others reduced it. However, if an adjusted value was substituted into the matrix before the next modification, then the correlations were successively improved and the values eventually converged. This behavior suggests that the effects of single substitutions were dependent and could not be isolated from each other in the mathematical analysis. The procedure was standardized to the order left-top to right-bottom and repeated several times through the entire matrix until the values converged. Binding site prediction | Once the adjustment system was worked out, additional oligonucleotides were designed and analyzed to complete a representative training set (sites 1 --30). The training set matrix was adjusted so that the correlation with 1 /C1 /2 fitted a line with an R2 value of 0.99 (Fig. , top). Sites 28 --30 were not included because they did not show measurable binding. This weighted matrix was then evaluated for detection of weak binding sites and prediction of the binding strength. Two extended DNA sequences were searched, a 70-kb sequence consisting of the rat albumin --alpha-fetoprotein (AFP) locus (J. Locker, unpublished results) and a comparable region of the human genome (GenBank accession no. AC008076). The searches indicated a large number of binding sites (see Discussion), of which eight were analyzed for binding. Seven of these sites (sites 35, 37 --39, 41 --43) were selected because their weak binding could discriminate predictions made with sum and product algorithms, while the eighth was an unusual high-scoring site containing only A:T base pairs (site 32). Four additional sites were studied: two recently described sites (sites 33 and 36) that critically regulate the PAH gene and two sites (sites 34 and 38) reported as non-binding but predicted to bind by our analysis. The sum (Fig. , bottom) but not the product algorithm (data not shown) scores showed a very strong correspondence with predictions based on the training set. Moreover, these predictions were accurate even for weak binding sites with as many as seven bases differing from the 14-base model site. A final weighted half-site matrix was calculated from the entire set of competition data , with R2 = 0.95 and the scores obtained with this matrix are listed in Table . In screening an extended sequence, the final matrix was approximately twice as selective as the training matrix at higher cut-off scores , increasing discrimination above a selected cut-off and accuracy of binding site predictions. Nevertheless, the effect was relatively limited and tended to push selected scores up by only a few points. At the medium strength cut-off, the discrimination was between somewhat weaker or stronger binding, not between binding and non-binding. Inspection of the 50 rat Alb --AFP locus sites over this cut-off indicated that the vast majority resembled the model HNF1 site and would be expected to bind with close to the predicted strength. The training set was tested with more divergent sites which were mostly well below the medium strength cut-off. Since the final matrix was calculated from only 40 sites, analysis of more sites could further improve the correlations. However, this improvement might be limited, because several base substitutions under-represented in the training matrix were better represented in the final compilation, where each single-base substitution in the half site was represented by at least two different oligonucleotides . The correlation with C1 /2 values was very strong but obviously not perfect. Some of the small deviations presumably reflected the accuracy of the experimental measurements. Alternatively, the averaging process could not fully compensate for synergistic or complementary effects of multiple base substitutions. The final training set matrix (data not shown) was very similar to the training matrix ---the main difference was that the training set matrix had more extreme values in the first column. The new matrix provided a considerable improvement in detection of HNF1 sites. In comparison, scores obtained with the simple frequency matrix or a similar matrix downloaded from TRANSFAC showed only moderate correlation with the full set of competition values (R2 = 0.59 --0.60) and little correlation with the prediction set (R2 = 0.19). Relationship to protein binding affinity | Further studies evaluated the relationship between matrix scores and HNF1 DNA-binding properties. The competition analysis provided C1 /2 values that were related to the dissociation constant, KD, but these were based on simplified experimental conditions and empirical comparison with a single binding site. For this further analysis, four binding sites (sites 2, 6, 9 and 24) were chosen with C1 /2 values ranging over more than two orders of magnitude, from 0.74 to 84.8, and gel-shift assays were carried out with a range of protein concentrations (Fig. A). P1 /2, the molar concentration of HNF1alpha at 50% binding, ranged from 2.3 x 10 --7 to 3.5 x 10 --6. For the two stronger sites (sites 9 and 2), Scatchard binding analysis demonstrated a slope of close to 2 (Fig. B) indicating second-order binding as a dimer (n = 2). For these sites, KD was calculated from the relationship KD = (P1 /2)n, as 5.1 x 10 --14 and 8.9 x 10 --14, respectively . The other two sites (sites 6 and 24) had slopes significantly <2, suggesting that their binding was intermediate between first and second order. In both cases, one half site was much weaker than the other. In site 6, the G in position 9 greatly weakened that half site, while site 24 had four substitutions in one half site and none in the other. The values suggested that monomer binding on the stronger half site was a significant fraction of the total binding at these weakened asymmetric sites. If binding at all four sites was considered second order, the calculated KD values ranged from 5.1 x 10 --14 to 1.2 x 10 --11, almost three orders of magnitude. As an alternative reflecting composite second- and first-order binding, the measured slopes (1.9 --1.4) from the Scatchard plots were used as values for n and gave a five order of magnitude range for KD, from 2.3 x 10 --13 to 2.3 x 10 --8. Because of this ambiguity, the relationship of the matrix score to binding was compared in several ways. The relationship between P1 /2 (or KD = P1 /22) and matrix score was nearly linear (Fig. C), with an R2 of 0.92 when fitted to a simple linear equation. Moreover, a plot of matrix score versus alternative KD values calculated using the Scatchard slopes also gave a linear correlation with the same R2 value of 0.92. In each of these cases, the modified half-site matrix was directly related to the chemical parameters of protein binding. Since binding competition was demonstrated over a larger range than direct protein binding, the ability of weak sites to bind HNF1 was further characterized in a gel-shift assay at room temperature . Binding of a very weak site (site 43, C1 /2 = 512) was easily detected using experimental conditions of medium sensitivity. The biological significance of such weak binding is unclear, but it can be detected, quantified and predicted with the systems described here. Figure 1 | HNF1 site and half-site matrix. HNF1 site and half-site matrix. The matrix was compiled by aligning 65 binding sites to a model binding site with greatest similarity on the left, and tabulating the percent frequency of bases in each position. The consensus sequence shows the most frequent bases; those with a frequency of <50% are in lower case. The half-site matrix comprised the first seven columns of the full-site matrix and had a consensus identical to the model half site. Figure 2 | Representative competition analysis showing the equivalent effects of a substitution in either half site. Representative competition analysis showing the equivalent effects of a substitution in either half site. Analyses of five unlabeled binding-site oligonucleotides are illustrated, in competition with labeled site 2. The experiment examined single base changes to C in positions 2 (site 12) or 3 (site 14), and a combination of these changes in the same (site 18) or opposite (site 17) half sites. The plot shows the values from the single illustrated experiment, while the values in Table were averaged from multiple experiments. Figure 3 | Site matrices and competition analysis. Site matrices and competition analysis. Top, a weighted half-site matrix was derived from the competition values of 27 training set oligonucleotides using the sum algorithm, as described in Materials and Methods. Bottom, this training set-derived matrix was then used to generate scores for 13 oligonucleotides in the prediction set, which were plotted against measured competition values and displayed with a dotted line fitted to the training set values. C1 /2 is the ratio of competitor to labeled site 2 oligonucleotide at 50% competition. Figure 4 | Correlation using the final weighted half-site matrix. Correlation using the final weighted half-site matrix. The training and prediction set competition values were combined and used to generate a more accurate half-site matrix. The matrix scores were compared with C1 /2 as in Figure . Figure 5 | Protein binding relationships. Protein binding relationships. (A) Protein binding curves. The relationship between protein concentration and binding was demonstrated for four binding sites. The illustrated data are from a single experiment. P1 /2, the molar concentration of HNF1 at 50% binding, was calculated by interpolation. Recombinant HNF1alpha MW = 32 215. Values were the average of two determinations. (B) Scatchard plots. Binding values of <80% were plotted in semilogarithmic plots. x, the binding fraction; CH, the molar concentration of HNF1; CO, the molar concentration of binding-site oligonucleotide; m, the measured slope. (C) Relationship between measured dissociation constant (KD) and binding competition (C1 /2). P1 /2 was plotted against scores obtained with the final weighted half-site matrix . The line was a simple linear regression of the values. Figure 6 | The limit of detectable protein binding. The limit of detectable protein binding. Five oligonucleotides, representing the intermediate to low range of C1 /2, were labeled and bound to increasing amounts of protein. The dried gel was exposed for 24 h at room temperature to XAR-5 film using a Lightning Plus intensifying screen (Dupont). Table 1 | HNF1 sites Table 2 | Relative discrimination of the training and final matrix Table 3 | Base frequency in tested half sites DISCUSSION : Experimental characterization of binding sites | Our studies were intended to establish a routine laboratory procedure that could experimentally determine the entire range of binding by a factor and describe this binding by a computer algorithm that could reliably search other sequences. The competition assays were carried out using expedient experimental conditions that enabled us to efficiently analyze a large number of sites. Unpurified commercial oligonucleotides were annealed and directly labeled or used as competitors, usually without further characterization. The gel-shift and competition conditions were experimentally optimized to provide uniform easily measured binding, short film exposures and consistent competition curves . As our algorithms developed, the Cereghini et al. data were displaced and fell on a parallel line, presumably because of differences in the way their competitions were carried out. Thus, the model also fit their data, but the most essential experimental parameter was to make the measurements internally consistent. A selected group of sites were analyzed under more rigorous conditions to determine P1 /2 and KD. As expected, these values showed a simple linear relationship to C1 /2 (data not shown). It is recognized that gel-shift analysis, though frequently used, is not the most accurate way to determine binding constants , but the purpose of our analyses was to determine how the matrix predictions correlated with true binding, which did not require the most precise determination of KD. Even so, the measured KD of 5.1 x 10 --14 was consistent with other transcription factors. For example, the dimeric HLH factor MASH-1 had a KD of 1.4 x 10 --14 , while a monomeric homeodomain factor, Dlx3 had a KD of 7.8 x 10 --8 . In principle, the analysis can start from a single binding oligonucleotide, which can be randomly substituted to generate a preliminary set of competitors. The preliminary assessment can then be followed by more systematic variation of bases. This process differs significantly from generation of consensus sequences or matrices using verified natural sites, because these are biased towards strong, easily recognized sites. Similarly, the widely used 'selected and amplified binding' (SAAB) procedure , in which transcription factor-bound DNA segments are amplified by PCR, is also biased toward the most strongly bound sites. For our approach, it is useful, but not essential, to start with a model of the binding site, which can facilitate the subsequent analysis. Analysis of HNF1 was simplified by the features already established for this factor, dimeric binding at two seven-base half sites on opposite DNA strands, with a fixed separation of one base. The use of recombinant protein also facilitated analysis and enabled direct correlation with KD. Except for calculation of specific parameters like KD, however, the same analysis can be carried out on factors present in crude cell extracts. Comparison of highly purified to unpurified oligonucleotides showed little difference. Moreover, although we averaged multiple determinations of C1 /2, the minor variation from one determination to another also did not significantly affect the overall weighting of the matrix. It there fore does not appear necessary to carry out the competitions with the highest precision. There were two other features that facilitated our experimental approach. First was the use of a half-site matrix. For HNF1, the equivalent additive contribution of half sites was experimentally verified. It is possible, however, that another class of dimeric factor might bind differently. For example, initial binding of a monomer at a strong site might recruit a second monomer to dimerize and bind a weak site in a manner that would make the two sites non-equivalent. As a further advantage, the binding of heterodimeric factors can be described by combining two different half-site matrices. Like most dimeric transcription factors, HNF1 monomers bind only at a fixed distance, though a few factors, like CTF /NF1, bind as dimers with variable spacing between half sites. For such complex binding, Roulet et al. have described an algorithm that includes two half-site matrices and an additional factor that compensates for the effect of half-site spacing. A second feature of our analysis was simultaneous variation of multiple bases in a single oligonucleotide competitor, which allowed comparison of a larger number of base variations per oligonucleotide. This process presumed the approximate independence of each position in determining the score matrix. However, independence was not observed for another DNA-binding protein, the bacteriophage repressor Mnt, where substitution at one of two positions significantly modified the binding of the other base . Such interactions probably also affected our analysis, suggested by the inability to perfectly linearize the correlation. The small magnitude of the effect may reflect specific properties of the HNF1 DNA-binding domain. Alternatively, the contributions non-equivalent base combinations had only a fractional effect on the entire matrix score. We did not test all combinations of two or more base substitutions and cannot rule out a larger effect for untested combinations. For searching DNA sequences, we wrote a computer program to use two (half-site) matrices in a single search. As an alternative for use in other DNA analysis programs, the weighted half-site matrix was combined with the corresponding reverse-complement matrix into a single matrix that gave identical searches and scores. To make it compatible with programs that do not allow negative numbers, the matrix was also translated into a positive numbers by adding 230 to each position. This translated matrix also gave identical searches, but with different cut-off scores. A model of the HNF1 DNA-binding domain | The weighted half-site matrix is a profile of the specific binding interactions. By convention, the binding site was written as a single-strand sequence, but the actual protein DNA contacts were presumably on both strands. For HNF1, the strongest interactions were through an A:T base pair in position 5 and a T:A base pair in position 7. Though the latter was designated as a T, the main interaction was with A on the opposite strand, because bacterial methylation of this A in one half site greatly reduced binding (,). The strongest disruptions occurred through introduction of a G:C base pair in position 6 and G:C or C:G base pairs in position 7. Thus, like other homeodomain binding sites , there was a significant bias toward A:T base pairs. Rather strong, but previously cryptic, HNF1 sites could be entirely composed of A:T base pairs (e.g. site 32). In addition to detecting cryptic sites, the weighted half-site matrix can also be used for rational mutagenesis. For example, the introduction of G:C and G:C into positions 6 and 7 of one half site will effectively inactivate a binding site, even while matching the model site at 12 of 14 base pairs. Conversely, sites that differ at seven of 14 base pairs may still have significant binding (e.g. site 36). Biological relationships | The cut-off point for biologically significant binding was unclear. The weakest functionally characterized site we found was in the human alpha1-microglobulin gene enhancer (site 7), which had a matrix score of 81 and was approximately two orders of magnitude weaker than the strongest binding sites. However, binding was easily demonstrated to much weaker sites, and it would be possible to demonstrate still weaker binding with modified experimental conditions: higher protein concentrations, longer film exposures, binding and electrophoresis at lower temperatures, and competition using binding sites weaker than site 2. We carried out searches of extended DNA sequences and detected very large numbers of sites. Extrapolation of these detections to the entire genome was limited by G+C content and other special local sequence properties, as well as the fact that the matrix was compiled from only 40 binding sites. Even so, the vast majority of sites above the medium-strength cut-off would be likely to show significant binding and the extrapolated number of such sites in the genome was several orders of magnitude greater than the number of HNF1 molecules in the cell. Thus, despite the limited accuracy of the extrapolation, the range of the predictions had important implications for the mechanism of transcriptional regulation by HNF1, and for the interpretation of binding sites detected by any algorithm. Analysis of a 70 000 bp (35% G+C) region containing the rat Alb --AFP locus provided typical results . A search with the cut-off at a matrix score of 92, the maximum that detected the strong albumin promoter site, detected two sites (0.03%, which extrapolated to 1 x 105 per genome). Detection of the weaker functional AFP promoter site required a cut-off of 84, and detected 50 sites. Nevertheless, the matrix algorithm can be considered highly selective. Binding predicted at the limit detected by gel-shift assay (matrix score = 79) corresponded to only 0.44% of all DNA sequences, i.e. 1 x 107 sites per haploid genome. In contrast to this large number of potential sites, Lichtsteiner and Schibler calculated only 1 --2 x 104 HNF1 molecules per cell. Clearly, simple scanning of DNA sequences cannot discriminate even strong sites that actually regulate genes. Sites that function in vivo must be distinguished by context, that is, their direct relationship to other regulatory elements. This makes it difficult to establish a lower limit for functional binding sites. In the right context, even a very weak site might function in a regulatory module if adjacent factors stabilized HNF1 binding. Overall, our approach has greatest utility for characterizing known regulatory regions and their interactions with a single transcription factor. Identifying a new regulatory region is a more difficult problem because demonstration of a site for a single factor, no matter how accurate the binding prediction, is simply not enough to predict that a region is functional. Even so, the solution to such higher-order prediction may not be far away. The required level of prediction might be accomplished by combining accurate analysis of only a few kinds of sites, along with experimentally determined 'grammatical' rules for their association. Figure 7 | Estimation of genomic binding sites. Estimation of genomic binding sites. The weighted half-site matrix was used to detect potential HNF1 binding sites in a 70 000 bp sequence from the rat Alb --AFP locus. The results are presented as the number of sites in 5% matrix score intervals (interval fraction) and the total number of sites detected in the same intervals (cumulative fraction). The plot was cut off at 57%, the score for the lowest possible positive matrix score. Representative values were extrapolated to a genome of 3 x 109 bp. Backmatter: PMID- 12202774 TI - Prediction of consensus structural motifs in a family of coregulated RNA sequences AB - Given a set of homologous or functionally related RNA sequences, the consensus motifs may represent the binding sites of RNA regulatory proteins. Unlike DNA motifs, RNA motifs are more conserved in structures than in sequences. Knowing the structural motifs can help us gain a deeper insight of the regulation activities. There have been various studies of RNA secondary structure prediction, but most of them are not focused on finding motifs from sets of functionally related sequences. Although recent research shows some new approaches to RNA motif finding, they are limited to finding relatively simple structures, e.g. stem --loops. In this paper, we propose a novel genetic programming approach to RNA secondary structure prediction. It is capable of finding more complex structures than stem --loops. To demonstrate the performance of our new approach as well as to keep the consistency of our comparative study, we first tested it on the same data sets previously used to verify the current prediction systems. To show the flexibility of our new approach, we also tested it on a data set that contains pseudoknot motifs which most current systems cannot identify. A web-based user interface of the prediction system is set up at . Keywords: INTRODUCTION : Transcriptional regulation is an important topic in bioinformatics. Much effort has been made to develop useful analysis tools to accelerate the progress in this research. An equally important but much less studied topic is post-transcriptional regulation. Similar to transcriptional regulation, post-transcriptional regulation is often accomplished by the binding of proteins to specific motifs in mRNA molecules ( --). Unlike DNA binding proteins, which recognize motifs composed of conserved sequences, RNA protein binding sites are more conserved in structures than in sequences. The motif prediction algorithms that only consider conserved sequence profiles ( --) may fail to identify RNA motifs. A set of post-transcriptionally coregulated RNAs can be characterized by base-pair interactions that organize the molecules into domains and provide a framework for functional interactions. If a new sequence is found to contain the common motifs, it may have the same characteristics as those coregulated RNAs. We are interested in finding the consensus motifs in a family of coregulated RNA sequences. There has been much work on RNA secondary structure prediction. The current main approaches include free-energy minimization ( --) and comparative sequence analysis ( --). Although they show positive results of predicting secondary structures of a single sequence, it is questionable to use these methods to find common motifs in a set of sequences. Other approaches such as stochastic context-free grammar, e.g. COVE , and genetic algorithms (GAs) have been applied to multiple sequences, but they are aimed at finding a global alignment instead of consensus motifs. A dynamic programming approach called FOLDALIGN, which takes into account both sequence similarity and structure constraints, was first developed to discover RNA motifs in a set of sequences . However, its time complexity is too high for practical use. Recently, a new system called SLASH has been developed. By combining FOLDALIGN and COVE, the time complexity of SLASH is acceptable for real applications, but it is currently limited to finding stem --loop motifs. In this paper we introduce a new approach called genetic programming for RNA motifs (GPRM), which is capable of discovering structural motifs more complicated than stem -- loop structures. To prove GPRM is comparable to the latest approaches, we tested it on the same data sets as used in the experiments of SLASH. Furthermore, we tested GPRM on a published pseudoknot data set to demonstrate its capability that most current prediction methods lack. MATERIALS AND METHODS : Motif prediction can be seen as a concept learning problem, that is, learning a target concept (i.e. motifs) from a set of training examples (i.e. biosequences) . According to its objective and the training examples given, concept learning can be regarded as supervised or unsupervised. From pre-classified training examples, supervised learning is to learn a discriminative concept to distinguish between examples of different classes. On the other hand, unsupervised learning is to learn a characteristic concept to describe a set of unlabeled training examples. We consider motif prediction a supervised learning problem. Unlike most current approaches, we use both positive and negative examples. Positive examples are a family of coregulated RNA sequences; negative examples are the same number of sequences randomly generated based on the observed frequencies of a sequence alphabet in positive examples. We learn the motifs that can be used to distinguish the given coregulated sequences from the random sequences. As RNA motifs may vary in both sequences and structures, we need an expressive representation to describe a wide variety of motifs, and an effective strategy to search a large problem space for the right motifs. Genetic programming (GP) operates on a population of concept hypotheses. Individuals in the population can be described by linear structures, trees or graphs . Unlike conventional GAs, GP does not require an encoding scheme to encode putative solutions into bit strings before the evolutionary process, or a decoding scheme to decode finally converged bit strings back to an interpretable representation. By GP, the hypotheses can converge to the comprehensible target concept through evolution. Because of its generality and effectiveness, we adapt GP to develop GPRM. Since RNA secondary structures are typically formed by base-pairing interactions, GPRM is focused on finding Watson --Crick complementary base pairs. There are three components in GPRM. The first is a population of putative structural motifs. The second is a fitness function that measures the quality of each motif. The third is the genetic operators that simulate the natural evolution process. The details are described in the following sections. Representing individuals in a population | Each individual in a population is a putative motif. We use two kinds of segments to describe structural motifs. A segment is either a Watson --Crick complementary segment or a non-pairing segment. A Watson --Crick complementary segment is a helix, and it may also contain GU wobble pairs. A non-pairing segment, on the other hand, is single-stranded. With different combinations of segments, a wide variety of RNA motifs can be easily represented. For example, in Figure A, we show an interior loop in AUACUCCCAACUAGUUCCUUGUAAC, CGGCCGUCAAGGUAUACGACCACGGCGA and ACAAAAGGAAUUAAUUAAAGAAAUGAAA. The common motif is composed of four complementary segments and two non-pairing segments. It is described as [2:5(4)]-{4:6}-[3:4(3)]-[3:4(2)]-{4:5}-[2:5(1)]. In this representation, we use brackets to indicate a complementary segment; braces, a non-pairing segment. The range of segment length is kept inside the brackets and braces, e.g. {4:6} means the length of the non-pairing segment is between 4 and 6 nt. The number within parentheses in a complementary segment is the index of its corresponding pairing segment, e.g. (4) in complementary segment [2:5(4)] means that this segment is paired with the fourth complementary segment in the motif. Similarly, in Figure B, we show a hairpin motif in AAGGGACUCCUCAGUCCCCA, GGAGGACCCCUCGUCCCA and CCGAGACCCCUUGUCUCCAAA. To find the motifs from a family of RNA sequences, the user of GPRM is required to specify the maximum number of segments and the range of segment length allowed in a motif. According to the specification, GPRM generates the initial population of putative motifs. The number of segments and the segment length in each motif are randomly assigned but conform to the user's specification. The pairing relation between complementary segments is determined at random. After the initial population is created, GPRM applies genetic operators to the population to generate a better population of motifs. This evolution process is repeated until no improvement can be found. Fitness function | In GP, the fitness function is used to measure the quality of individuals in a population. The higher the fitness of an individual, the better its chances of survival to the next generation. We are interested in the motifs that can reflect the characteristics conserved in a family of coregulated RNA sequences, e.g. the RNA protein binding sites. We design a fitness function that assigns higher values to those motifs commonly shared by the given set of RNAs, and rarely contained in random sequences. Our fitness function is derived from the F-score used in the field of information retrieval with the aim to balance the importance of two measures, recall (i.e. sensitivity) and precision (i.e. positive predictive value). Given a positive example set and a negative example set, we define the fitness function as follows: The motif that will be chosen to participate in the genetic operation, e.g. mutation, is dependent on fitness. Motifs with higher fitness have better chances of being selected. We adapt the tournament selection mechanism. It parallels the competition in nature among individuals for the right to take part in evolution. Unlike fitness-proportionate selection, tournament selection does not need a centralized calculation of the average fitness of the population, and it is somewhat faster than rank selection (,). It first randomly picks two individuals from the population. Then the one with better fitness gets selected for the genetic operation. Genetic operators | Reproduction models the self-replication process in nature. Instead of selecting one motif at a time, and passing it to the next generation, GPRM accelerates the reproduction process by passing the better half of the population sorted by fitness from generation to generation. Similar to the mutation operation in nature that causes sporadic and random alterations in the genetic materials, GPRM's mutation operator changes the segment configuration of a motif selected from the population. It first randomly picks a segment of the motif for alteration. If a complementary segment is selected, its corresponding pairing segment and its length range are then randomly changed. For example, a segment of 5 --7 nt in length previously paired with the third complementary segment may be changed to a new segment of 4 --6 nt in length now paired with the fourth complementary segment. On the other hand, if a non-pairing segment is chosen, only its length range is changed. Note that the segment length can only be randomly altered within the range specified by the user. Unlike mutation, the crossover operation is performed on two individuals. Its purpose is to exchange the segment configuration between two tentative motifs to generate two offspring. After two motifs are selected from the population, either a pair of complementary segments or a non-pairing segment is chosen at random for exchange. Implementation | GPRM is an optimization procedure that iteratively applies genetic operators to improve the fitness of tentative solutions. After the creation of the initial population, GPRM goes through three basic steps in each optimization cycle. The three steps are fitness evaluation, individual selection and population generation. The process is repeated until no improvement of fitness can be found, or it reaches the limit of generations. The pseudocode of GPRM is shown in Figure . By masking out the motifs already found in sequences, we can repeatedly apply GPRM to find multiple motifs if they exist in a set of RNA sequences. The first step of GPRM is to find all possible pairs of complementary segments in the training examples, and put them in POOL. Suppose the lower and the upper bound of complementary segment length are l and u, and let f = u -- l + 1. The time complexity of finding all possible pairs of complementary segments is O(fL2N) = O(L3N) if f ~ L, where L is the maximum sequence length and N is the number of total sequences. GPRM computes the fitness of each putative motif in the population by iteratively comparing each complementary segment of the motif with the entire POOL. Let the maximum number of complementary segments in a motif be m, which is specified by the user. The time complexity of fitness computation is O(m2L3NP) where P is the constant population size. Compared with the fitness evaluation, the time complexity of crossover and mutation operations is negligible. The total time complexity of GPRM is thus O(L3N) + O(m2L3NPG) = O(m2L3N), where P is the constant population size, and G is the constant generation limit. In the current version of GPRM, P and G are set to be 1000 and 50, respectively. Note that m is the maximum number of complementary segments specified by the user. Given a family of coregulated RNAs, we are interested in the common motifs for the RNA regulatory protein binding sites instead of a global alignment. Therefore, the number of complementary segments in a motif is relatively small. If m << L and m << N, the total time complexity can be reduced to O(L3N). Figure 1 | (A) A common internal loop motif of three sequences, AUACUCCCAACUAGUUCCUUGUAAC, CGGCCGUCAAGGUAUACGACCACGGCGA and ACAAAAGGAAUUAAUUAAAGAAAUGAAA. (A) A common internal loop motif of three sequences, AUACUCCCAACUAGUUCCUUGUAAC, CGGCCGUCAAGGUAUACGACCACGGCGA and ACAAAAGGAAUUAAUUAAAGAAAUGAAA. The internal loop is composed of four complementary segments and two non-pairing segments. (B) A common hairpin motif of three sequences, AAGGGACUCCUCAGUCCCCA, GGAGGACCCCUCGUCCCA and CCGAGACCCCUUGUCUCCAAA. The hairpin is composed of two complementary segments and one non-pairing segment. Figure 2 | Pseudocode of GPRM. Pseudocode of GPRM. RESULTS : There are two purposes of our experiments. The first is to demonstrate that GPRM is competitive with current RNA motif prediction systems. The second is to show that GPRM can identify complicated motifs that most current systems cannot find. It is important to use the same data sets in experiments to keep the consistency of a comparative study. As SLASH is the latest RNA motif prediction system, we first tested GPRM on the same data sets as used in SLASH's experiments to show GPRM's competitive performance. Moreover, we used a published pseudoknot data set to demonstrate GPRM's flexibility that is lacking in most current systems, including SLASH. These data sets are described in the following sections. Data sets | The first data set is one of the data sets used to test SLASH. It contains 34 archaea 16S ribosomal sequences . This data set was originally derived from a set of 311 sequences extracted from the SSU rRNA database . The archaea set of 311 sequences was further reduced to 34, filtering out the sequences that miss base assignments or are >90% identical. The final 34 sequences are the following, where the number in parentheses is its GenBank accession number: Acidianus brierleyi (D26489), Caldococcus noboribetus (D85038), Cenarchaeum symbiosum (U51469), Desulfurococcus mobilis (M36474), Metallosphaera sp. (D85508_D38776), Pyrobaculum aerophilum (L07510), Pyrodictium occultum (M21087), Stygiolobus azoricus 2 (D85520), Sulfolobus metallicus 2 (D85519), Sulfolobus solfataricus 2 (D26490), Sulfurisphaera ohwakuensis (D85507_ D38775), Thermofilum pendens (X14835), Thermoproteus tenax (M35966), Archaeoglobus fulgidus (X05567_Y00275), Bacterial sp. 34 (X92171), Bacterial sp. 36 (X92172), Haloarcula vallismortis (U17593), Halobacteriaceae gen. sp. 2 (AJ002946), Halorubrum sodomense (D13379), Natrono bacterium magadii (X72495), Methanobacterium sp. (AF028690), Methanobacterium thermoautotrophicum 5 (AE000940_AE000666), Methanothermus fervidus (M32222), Methanococcus jannaschii 3 (U67517_L77117), Methano coccus vannielii (M36507), Methanoculleus marisnigri (AF028693), Methanosarcina frisius (X69874), Methano spirillum hungatei (M60880), Methanothrix soehngenii (X16932_X51423), Pyrococcus sp. 2 (Z70247), Thermo coccus mexicalis (Z75218), Thermococcus stetteri (Z75240), Ferromonas metallovorans (AJ224936) and Thermoplasma acidophilium (M38637_M20822). To ensure that the sequences can only be aligned locally, Gorodkin et al. further randomly truncated each sequence at both ends by up to 20 nt. The second data set is another data set used in the experiments of SLASH. It is the ferritin IRE-like data set (iron response element) constructed by Gorodkin et al. . They first obtained 14 sequences from the UTR database . Since the selected IRE regions are significantly conserved not only in structure but also in sequence, even sequence motif finding algorithms can identify them within the UTRs. Therefore, they modified the IREs and their UTRs to make the search more difficult. By iteratively shuffling the sequences and randomly adding 1 nt to the IRE conserved region, they obtained a set of 56 IRE-like sequences from the 14 IRE UTRs. The new structure motifs are as shown in schemes and below. where the parentheses indicate base pairing, N {A, G, C, T}, W {A, U}, H {A, C, U}, and X is a random nucleotide. They are highly variable in sequence, but with conserved structure. The third data set includes 18 viral 3'-UTRs each of which contains a pseudoknot. Seven of the RNA sequences are the soil-borne rye mosaic viruses; the others are the soil-borne wheat mosaic viruses. We first retrieved the pseudoknot sequences from PseudoBase . Their accession numbers in PseudoBase are listed as PKB183 --PKB189 and PKB194 -- PKB204. The pseudoknot sequences and base pairings are presented below. As the pseudoknots are relatively short, to make the search for the pseudoknots more challenging, we randomly include the flanking of 5 --70 nt at both ends of each pseudoknot sequence. All the data sets above are downloadable from . Evaluation | We applied the Matthews correlation coefficient to quantify the agreement between the predicted motif and the actual structure assignment. For each sequence in the data set, two secondary structure assignments were compared by counting the number of true positives Pt (base pairs exist in actual assignment and are predicted), true negatives Nt (base pairs do not exist in actual assignment and are not predicted), false positives Pf (base pairs do not exist in actual assignment but are predicted) and false negatives Nf (base pairs exist in actual assignment but are not predicted), respectively. The Matthews correlation coefficient can then be computed as: Given that the sequence length is sufficiently large, the Matthews correlation coefficient can be approximated in the following way : With the published /curated alignments, we can evaluate the performance of our approach by calculating the Matthews correlation coefficient. Due to its stochastic characteristics, GPRM was repeatedly tested 30 times on each of the data sets. The correlation coefficients averaged over 30 runs are presented in Table , and Table is the GP tableau for the RNA secondary structure prediction problem. Owing to limited space, a partial result of the predicted RNA motifs is shown in Figure , and a complete GPRM sample output can be found at . The crossover rate and the mutation rate can affect GPRM's performance. As the two operators may interact with each other, in order not to complicate the experiments, we fixed one rate at 0.5 and varied the other from 0.5 to 0.9 to measure its effect on the Matthews correlation coefficient. Figure shows the change of correlation coefficients along with varying crossover rates, and Figure presents the results when applying different mutation rates. According to Figure , the correlation coefficient for the viral 3'-UTR data set is more sensitive to the change of crossover rate. Its SD is 0.007. Compared with the viral 3'-UTRs, GPRM is more stable when applied to the other two data sets. The SDs of their correlation coefficients are 0.001 and 0.003, respectively. Similarly, Figure shows that the performance of GPRM on the viral 3'-UTRs varied with different mutation rates more noticeably than on the other two data sets, 0.008 compared with 0.0009 and 0.003. Currently, GPRM uses a random negative set of the same size as the positive set. To investigate the effect of the negative set size on the correlation coefficient, we repeated our experiments with negative sets of different sizes, varying from 1 to 10 times of the positive set size. The correlation coefficients are presented in Figure . It shows that the performance for the viral 3'-UTRs data set was affected the most. In addition, we examined the effects of varying complementary segment length ranges on convergence. We fixed the minimum length to 8 bp, and varied the maximum length from 15 to 20 bp for the 16S rRNAs data set. We recorded the fitness values at different generations to see how the segment lengths affect GPRM's convergence behavior. The result is shown in Figure A. Similarly, for the IRE-like and the viral 3'-UTR data sets, we fixed the minimum length to 3 bp, and varied the maximum from 10 to 20 bp. The results are presented in Figure B and C, respectively. Figure indicates no significant effects of varying segment lengths. For each test data set, GPRM's fitness values converged before 50 generations. Similar experiments were also performed on varying non-pairing segment length ranges. The results also showed no significant differences (data not shown). Figure 3 | A partial result of the predicted RNA motifs. A partial result of the predicted RNA motifs. The numbers above the sequences are the indices of the nucleotides. The predicted and the published motifs are both shown for reference. A complete sample result can be found at . Figure 4 | The correlation coefficients for different crossover rates. The correlation coefficients for different crossover rates. We fixed the mutation rate at 0.5, and varied the crossover rate from 0.5 to 0.9. For all experiments, the size of the negative data set was set the same as the positive data set size. Figure 5 | The correlation coefficients for different mutation rates. The correlation coefficients for different mutation rates. We fixed the crossover rate at 0.5, and varied the mutation rate from 0.5 to 0.9. For all experiments, the size of the negative data set was set the same as the positive data set size. Figure 6 | The correlation coefficients for negative sets of different sizes. The correlation coefficients for negative sets of different sizes. We varied the set size from 1 to 10 times of positive set size. For all experiments, we fixed the mutation rate at 0.9 and the crossover rate at 0.5. Figure 7 | The fitness values for different base-pairing segment length ranges at different generations. The fitness values for different base-pairing segment length ranges at different generations. For 16S rRNAs, we fixed the minimum length to 8 bp, and varied the maximum length from 15 to 20 bp. The fitness values at different generations are shown in (A). For both IRE-like data and viral 3'-UTRs, we fixed the minimum length to 3 bp, and varied the maximum length from 10 to 20 bp. The result of the IRE-like data set is presented in (B), and the viral 3'-UTRs result is illustrated in (C). In each experiment, we fixed the mutation rate at 0.9 and the crossover rate at 0.5. Table 1 | The experimental results of GPRM on three data sets Table 2 | Tableau for RNA secondary structure prediction problem DISCUSSION : We developed a GP approach to finding common structural motifs in a set of coregulated RNA sequences. Those methods designed to identify only consensus sequences are not reliable to find RNA motifs. With flexible GP operators and structural motif representations, our new method, GPRM, is able to identify general RNA secondary motifs. To show GPRM is comparable to the latest RNA motif prediction systems, we tested it on the same data sets previously used in order to maintain consistency. We first tested GPRM on a set of archaeal rRNA sequences that contain locally aligned stem --loop regions. By comparing them with the curated database alignment, we were able to evaluate our new approach quantitatively by the Matthews correlation coefficient. We obtained a 0.87 correlation coefficient between the predicted structural alignment and the curated database alignment. This is similar to the published experimental results . We also tested GPRM on the ferritin IRE-like data set created by Gorodkin et al. , and obtained a 0.99 correlation coefficient. GPRM was further tested on a a set of viral 3'-UTR pseudoknot regions extracted from PseudoBase . We used this data set to demonstrate its capability that current RNA motif finding algorithms lack. We obtained promising correlation coefficients from 0.75 to 0.83 as shown in Figure . GPRM can be further improved in two directions. First, the current fitness function of GPRM is only based on motif occurrences in training examples. We plan to enhance the fitness function by incorporating background knowledge such as thermodynamic (,) or phylogenetic information. Secondly, GPRM is currently limited to find base-pairing structures. We will extend the motif representation and the genetic operators to deal with more complex structures, e.g. multiple compound stem --loops or structures with multi-branch loops. Backmatter: PMID- 12202760 TI - Sequence-specific protection of plasmid DNA from restriction endonuclease hydrolysis by pyrrole --imidazole --cyclopropapyrroloindole conjugates AB - The pyrrole --imidazole (Py --Im) triamide --cyclopropa pyrroloindole (CPI) conjugates ImPyImLDu86 (7) and ImImPyLDu86 (14) were synthesized and their alkylating activities and inhibitory effects on DNA hydrolysis by restriction endonucleases were examined. Sequencing gel analysis demonstrated that conjugates 7 and 14 specifically alkylated DNA at 5'-CGCGCG-3' and 5'-PyGGCCPu-3', respectively. Agarose gel electrophoresis indicated that incubation of a supercoiled plasmid, pSPORT I (4109 bp), with conjugate 7 effectively inhibited its hydrolysis by BssHII (5'-G_CGCGC-3'), whereas conjugate 14 had no effect on this hydrolysis. These results suggest that conjugate 7 sequence-specifically inhibits the hydrolysis of DNA by BssHII. Sequence-specific alkylation by the Py --Im triamide --CPI conjugates was further confirmed by inhibition of the Eco52I (5'-C_GGCCG-3') hydrolysis of conjugate 14-treated pQBI PGK (5387 bp). In clear contrast, hydrolysis of pQB1 PGK by DraI (3'-TTT_AAA-3') was not inhibited by 5 microM conjugate 14. That ImImPy did not inhibit the hydrolysis of pQB1 PGK indicates that covalent bond formation is necessary for inhibition. A similar experiment, using linear pQBI PGK, achieved the same extent of protection of the DNA with approximately half the concentration of conjugate 14 as was required to protect supercoiled DNA from hydrolysis. Keywords: INTRODUCTION : Small sequence-specific DNA-binding molecules, exemplified by N-methylpyrrole (Py) and N-methylimidazole (Im) polyamides, uniquely recognize each of the four Watson --Crick DNA base pairs in the minor groove according to a simple set of pairing rules dictated by the side-by-side binding of the aromatic amino acids. They have attracted intense interest as powerful tools in molecular biology and human medicine ( --). Trauger et al. have demonstrated elegantly that such polyamides bind to predetermined DNA sequences with affinities and specificities similar to those of DNA-binding proteins. Py --Im polyamides that uniquely recognize each of the four Watson --Crick base pairs can be used as novel recognition components of sequence-specific DNA alkylating agents ( --). We have demonstrated that hybrid molecules constructed from segment A of duocarmycin A and Py --Im diamides or from hairpin polyamides specifically alkylate at predetermined nucleotide sequences within a 450-bp DNA fragment (,). Recently, we found that insertion of a vinyl linker (L) between the polyamide and cyclopropapyrroloindole (CPI) groups alters the location of the reactive cyclopropane ring of the alkylating moiety and allows highly sequence-specific cooperative double-strand alkylation of DNA . More recently, we developed a novel DNA interstrand cross-linking agent that cross-links double strands only in the presence of ImImPy, at the 9-bp sequence, 5'-PyGGC(T /A)GCCPu-3' . Targeting specific sequences in the human genome with such Py --Im polyamides should provide a powerful gene-regulating tool. In fact, Gottesfeld et al. and Dickinson et al. demonstrated that hairpin polyamides inhibit basal and activated transcription from the promoters of RNA polymerases II and III by disrupting specific transcription factor --DNA interactions. The inhibition of transcription by hairpin polyamides is achieved by competitive binding to the binding sites of regulatory proteins. Therefore, the binding of Py --Im polyamides to the coding region cannot effectively inhibit transcription by RNA polymerase . Alkylating Py --Im polyamides that form covalent bonds with DNA should have more potential to regulate transcription and can target, not only the regulatory region, but also the coding region. Inhibition of the hydrolysis of plasmid DNA catalyzed by restriction endonucleases has been used effectively to examine sequence-specific binding of peptide nucleic acid . As a first step in evaluating the potential of alkylating Py --Im polyamides in the regulation of specific genes, we examine here the alkylation of plasmid DNA and its inhibitory effects on the hydrolysis of DNA by restriction endonucleases. MATERIALS AND METHODS : Materials | Polyacrylamide gel electrophoresis was performed on a Hitachi 5500-S DNA Sequencer. Ex Taq DNA polymerase and Suprec-02 purification cartridges were purchased from Takara Co.; the Thermo Sequenase core sequencing kit and loading dye (DMF with fuchsin red) were from Amersham Co. Ltd; 5'-Texas Red-modified DNA oligomer (18mer) was synthesized by Kurabo Co. Ltd; and 50% Long RangerTM gel solution was obtained from FMC Bioproducts. Plasmids pSPORT I and QBI PGK were obtained from Gibco BRL and Takara Shuzo Co. Ltd, respectively. P1 nuclease and calf intestinal alkaline phosphatase (AP, 1000 U /ml) were purchased from Roche Diagnostics. Restriction endonucleases, BssHII, Eco52I, DraI and HindIII were purchased from Takara Shuzo Co. Ltd. Syntheses of Py --Im --CPI conjugates ImPyImLDu86 (7) and ImImPyLDu86 (14) are described in the Supplementary Material. 7: 1H NMR (DMSO-d6) delta 12.39 (s, 1H), 10.57 (s, 1H), 10.29 (s, 1H), 9.76 (s, 1H), 7.60 (s, 1H), 7.56 (d, J = 14.5 Hz, 1H), 7.43 (s, 1H), 7.40 (d, J = 2.0 Hz, 1H), 7.26 (brs, 1H), 7.20 (d, J = 2.0 Hz, 1H), 7.00 (d, J = 14.5 Hz, 1H), 4.20 (m, 2H), 3.95 (s, 3H), 3.87 (s, 3H), 3.81 (s, 3H), 3.74 (s, 3H), 3.48 (m, 1H), 2.47 (s, 3H), 2.11 (dd, J = 4.0 and 4.5 Hz, 1H), 2.02 (s, 3H), 1.36 (t, J = 4.0 Hz, 1H). ESIMS m /e calc. for C34H33N10O4 (M-H) 693.7, found 693.6. 14: 1H NMR (DMSO-d6) delta 12.35 (s, 1H), 10.34 (s, 1H), 10.29 (s, 1H), 9.33 (s, 1H), 7.57 (d, J = 15.5 Hz, 1H), 7.57 (s, 1H), 7.50 (s, 1H), 7.41 (d, J = 2.0 Hz, 1H), 7.01 (d, J = 2.0 Hz, 1H), 6.83 (brs, 1H), 6.56 (d, J = 14.5 Hz, 1H), 4.29 (d, J = 10.0 Hz, 1H), 4.19 (dd, J = 4.5 and 10.0 Hz, 1H), 4.00 (s, 3H), 3.97 (s, 3H), 3.72 (s, 3H), 3.71 (s, 3H), 3.45 (m, 1H), 2.46 (s, 3H), 2.04 (s, 3H), 2.02 (dd, J = 4.0 and 4.5 Hz, 1H), 1.29 (t, J = 4.0 Hz, 1H). ESMS m /e calc. for C34H34N10O7 (M + H) 695.3, found 695.3. Preparation of 5'-Texas Red-end-modified 450-bp DNA fragments | The 5'-Texas Red-modified 450-bp DNA fragments pUC18 F378*-827 and pUC18 R1861*-2310 (they are complementary) were prepared using a PCR method and the 5'-Texas Red-modified 18mers 5'-TGTAAAACGACGGCCAGT-3' (pUC18 forward 378 --396), and 5'-TGCTGGCCTTTTGCTCAC-3' (pUC18 reverse 1861 --1879) as primers, and purified by filtration using Suprec-02. Product concentrations were determined by ethidium bromide staining. The asterisks (above) indicate Texas-Red modifications, and the nucleotide numbering starts at the replication site. High resolution gel eletrophoresis | The 5'-Texas Red-labeled DNA fragment (50 microg) was alkylated with various concentrations of conjugates 7 or 14 in 10 mM Na phosphate buffer (10 microl, pH 7.0) containing 10% DMF at 25C overnight. The reaction was quenched by the addition of calf thymus DNA (5 mM, 1 microl) with heating for 5 min at 90C. DNA was precipitated with ethanol. The pellet was dissolved in 8 microl of loading dye (formamide with fuchsin red), heated at 94C for 30 min, and then immediately cooled to 0C. A 2-microl aliquot was separated electrophoretically on a 6% denaturing polyacrylamide gel using a Hitachi 5500-S DNA sequencer. Preparation of linear DNA | Linear DNA was generated from plasmid DNA (pQBI PGK, 20 microg) by treatment with HindIII (30 U) in 10 mM Tris --HCl buffer (200 microl, pH 7.5) containing 7 mM MgCl2 and 60 mM NaCl at 37C for 1 h. After phenol /CHCl3 treatment, linear DNA was isolated by ethanol precipitation. DNA concentrations were determined by the absorbance of double-stranded DNA (dsDNA) at 260 nm (1 A260 unit = 50 microg /ml for dsDNA). Agarose gel electrophoresis | Reaction mixtures (2.5 microl) containing plasmid DNA (75 ng), and the indicated amounts of conjugates 7 or 14, 10 mM Tris --HCl buffer (pH 8.9), 3 mM MgCl2, 100 mM NaCl and 10% DMF were incubated at 25C for 2 days. After incubation, restriction endonuclease (4 U) was added to the reaction mixture (total volume 5 microl including 0.01% bovine serum albumin), and the solution was incubated at 37C for 2 h. The reaction mixture was directly loaded onto a 1.2% E-gelTM (Invitrogen) containing 0.5 mg /ml ethidium bromide with 2 microl of loading buffer (1% sodium dodecyl sulfate, 50% glycerol and 0.05% bromophenol blue) and 13 microl of H2O. Electrophoresis was carried out under a constant electric field of 70 V for 30 min at 25C. DNA bands were detected and photographed under UV light (302 nm). RESULTS AND DISCUSSION : Sequence-specific DNA alkylation by Py --Im triamide --CPI conjugate | The detailed synthesis procedures of Py --Im triamide --CPI conjugates ImPyImLDu86 (7) and ImImPyLDu86 (14) are described in the Supplementary Material. Sequence-specific DNA alkylation by conjugate 7 was investigated on both strands of a 5'-Texas Red-labeled 450-bp DNA fragment prepared by PCR using pUC18 as the template. As shown in Figure , conjugate 7 at a concentration of 400 nM produced highly selective alkylation of the G in the sequence 5'-CGCGCG-3' /3'-GCGCGC-5' on both strands (Sites 1 and 2) within a 450-bp DNA fragment, with high efficiency. The results are fully consistent with Dervan's pairing rule and clearly indicate that the imidazolyl acrylic part of the molecule plays the same key role as its pyrrolyl counterpart in the dialkylation system . The simultaneous double-strand dialkylation was further confirmed at the oligonucleotide level by HPLC product analysis. Importantly, this agent can dialkylate double DNA strands without the slipped binding mode, which was observed with ImPyLDu86 as a minor alkylation site . Similarly, sequencing gel electrophoresis of conjugate 14-treated labeled DNA fragments demonstrated that conjugate 14 specifically alkylated DNA at 5'-TGGCCG-3' /3'-ACCGGC-5' (Sites 1 and 2) on both strands, by forming a homodimer according to the pairing rule of Py --Im polyamides as shown in Figure . The double-strand alkylation by conjugate 14 was observed in a self-complementary sequence, 5'-CAACGGCCGTTG-3', indicating that alkylation by conjugate 14 occurs at either G or A (data not shown). Sequence-specific protection of plasmid DNA by Py --Im triamide --CPI conjugates | We next investigated the alkylation of supercoiled plasmid DNA (4000 --5500 bp) by Py --Im triamide --CPI conjugates 7 and 14. To examine the sequence specificity of conjugate 7 in supercoiled plasmid DNA, the inhibitory effects of conjugate 7 were examined on the hydrolysis of DNA by the restriction endonuclease BssHII, which hydrolyzes dsDNA at 5'-G_CGCGC-3'. If conjugate 7 selectively alkylates supercoiled plasmid DNA at the 3' end of the 5'-CGCGCG-3' sequence, which overlaps the recognition sequence for BssHII, an inhibition of hydrolysis would be anticipated. Figure shows the separation by agarose gel electrophoresis of the hydrolysis products of pSPORT I (4109 bp), which contains two BssHII cleavage sites and gives rise to 3241- and 868-bp fragments (Fig. B, lane 3). Interestingly, incubation of pSPORT I with conjugate 7 for 2 days (pH 8.9, 25C) effectively inhibited the hydrolysis by BssHII in a concentration-dependent manner (5 microM to 1.0 nM, Fig. B, lanes 4 --11). Almost complete inhibition of hydrolysis was observed in the presence of 5 microM conjugate 7 (Fig. B, lane 4), whereas conjugate 14, which alkylates DNA at the 3' end of the 5'-CGGCCG-3' sequence, had no effect on the hydrolysis of pSPORT I by BssHII (Fig. B, lane 12). The results clearly demonstrate that conjugate 7 sequence-specifically alkylates supercoiled plasmid DNA and that this alkylation effectively inhibits the hydrolysis of DNA by BssHII. The sequence-specific alkylation of supercoiled plasmid DNA by the alkylating Py --Im triamide 14 was further confirmed by the selective inhibition of the hydrolysis of pQBI PGK (5387 bp) by Eco52I (5'-C_GGCCG-3'). Figure shows the separation by agarose gel electrophoresis of the hydrolysis products of pQBI PGK: 4599- and 788-bp fragments generated from the two Eco52I cleavage sites (Fig. B, lane 3). Incubation of DNA with conjugate 14 inhibited the hydrolysis of DNA by Eco52I in a concentration-dependent manner (1 microM to 1 nM, Fig. B, lanes 4 --10). However, hydrolysis by DraI, which hydrolyzes dsDNA at 5'-TTT_AAA-3' and cleaves pQBI PGK into 4676-, 692- and 19-bp fragments, was not inhibited by incubation with 5 microM conjugate 14 (lane 12). The results clearly indicate that conjugate 14 selectively alkylates the target sequence in supercoiled plasmid DNA and sequence-specifically inhibits the hydrolysis of DNA by Eco52I. It is important to note that 5 microM ImImPy did not inhibit the hydrolysis of pQBI PGK, even though this concentration is 3-fold higher than the concentration of conjugate 14 that completely inhibited the hydrolysis . These results clearly indicate that covalent bond formation between DNA and Py --Im polylamides is essential for the inhibition of hydrolysis. Because alkylation in the minor groove influences base pairing and therefore also leads to substantial changes in the major groove, the hydrolytic activity of restriction endonucleases that use major-groove contact might be effectively inhibited. Even though the protection afforded by the alkylating Py --Im triamide relative to that afforded by ImImPy is dramatically different, the concentration required for complete protection was rather high (1 --5 microM) relative to the strong binding of the Py --Im polyamide. Part of the reason may be derived from the nature of the protection assay. Since complete inhibition in protection experiments requires complete alkylation of DNA at the target site, unreacted alkylating agent and /or it's hydrolysis products possibly interfere with the completion of alkylation. Furthermore, the Py --Im --CPI conjugates in the present study do not possess a positively charged terminal end, which is known to enhance the binding affinity to DNA. Incorporation of such a group may further enhance the alkylation activity, and as a result, it may provide more significant protection of the DNA by the alkylating Py --Im polyamide relative to that provided by the non-covalent Py --Im polyamide. Effect of supercoiling on the protection by Py --Im triamide --CPI conjugates | To evaluate the influence of DNA supercoiling on the protection afforded by conjugate 14 against enzyme hydrolysis, a similar experiment was performed using linearized pQBI PGK with HindIII (5'-A_AGCTT-3') digestion of supercoiled pQBI PGK. The linear DNA (5387 bp) was treated with conjugate 14 and subjected to Eco52I digestion. Figure shows the agarose gel electrophoretic separation of the hydrolysis products of linear pQBI PGK by Eco52I. In the absence of conjugate 14, the linear DNA was digested by Eco52I into three DNA fragments of 2887, 1712 and 788 bp. More efficient inhibition was induced by incubation of linear DNA with conjugate 14 than when supercoiled DNA was incubated with conjugate 14. In the former case, almost complete inhibition of hydrolysis was observed in the presence of 500 nM conjugate 14 (Fig. B, lane 5). Even at 50 nM conjugate 14, full-length linear DNA was detected. Densitometric analysis of the unhydrolyzed DNA indicated that approximately half the concentration of conjugate 14 was required to achieve the same extent of inhibition with linear DNA as with supercoiled DNA . Although supercoiling slightly retarded the polyamide binding to the target sequences in the minor groove, it did not affect the sequence-specificity of conjugate 14. These results might be consistent with the recent observation that hairpin polyamides sequence-specifically bind to the nucleosome core particle . In conclusion, we have demonstrated that the alkylating Py --Im triamide --CPI conjugates 7 and 14 alkylate supercoiled plasmid DNA, that this sequence-specific alkylation effectively protects DNA from restriction endonuclease hydrolysis, and that supercoiling of DNA has a slightly negative effect on the DNA alkylation. These results indicate that sequence-specific DNA alkylating agents acting in the minor groove strongly affect DNA --protein interactions. Sequence-specific alkylation by Py --Im polyamides may control gene expression, not only in promoter /enhancer-binding regions, but also in the coding regions of genes. Figure 1 | Thermally induced strand cleavage of the alkylated 5'-Texas Red-labeled 450-bp DNA duplex by ImPyImLDu86 (7). Thermally induced strand cleavage of the alkylated 5'-Texas Red-labeled 450-bp DNA duplex by ImPyImLDu86 (7). Results using 5'-labeled top strand (pUC18 F378 --827) (A) and 5'-labeled bottom strand (pUC18 R1861 --2310) (B) DNA fragments are shown. These two DNA fragments are complementary sequences containing Site 1 and Site 2. Lane 1 contains the DNA control; G, C, T, A are the Sanger sequencing standards. Concentrations of conjugate 7 are indicated in lanes 2 and 3. Sequences containing dialkylation sites (Site 1 and Site 2) are represented (C). The arrows indicate the sites of alkylation. Figure 2 | Thermally induced strand cleavage of the alkylated 5'-Texas Red-labeled 426-bp DNA duplex by ImImPyLDu86 (14). Thermally induced strand cleavage of the alkylated 5'-Texas Red-labeled 426-bp DNA duplex by ImImPyLDu86 (14). Results using 5'- labeled top strand (pUC18 R2207 --2632) (A) and 5'-labeled bottom strand (pUC18 F56 --481) (B) DNA fragments are shown. These two DNA fragments are complementary sequences containing Site 1 and Site 2. Lane 1 contains the DNA control; G, C, T, A are the Sanger sequencing standards. Concentrations of conjugate 14 are indicated in lanes 2 --5. Sequences containing dialkylation sites (Site 1 and Site 2) are represented (C). The arrows indicate the sites of alkylation. Figure 3 | Sequence-specific protection of pSPORT I plasmid DNA from restriction endonuclease BssHII hydrolysis by incubation with ImPyImLDu86 (7). Sequence-specific protection of pSPORT I plasmid DNA from restriction endonuclease BssHII hydrolysis by incubation with ImPyImLDu86 (7). (A) A BssHII restriction map of pSPORT I DNA and schematic representation of the binding mode of conjugate 7 at BssHII hydrolysis sites. The Py and Im rings are represented by open and closed circles, respectively; the triangle represents segment A of DU-86. (B) Agarose gel (1.2%) electrophoresis of triamide --CPI conjugate-treated pSPORT I DNA after BssHII (2.5 U) digestion (lanes 3 --12). Lane 1, 250-bp DNA marker; lane 2, pSPORT I DNA control; lane 3, no drug treatment; lanes 4 --11, 5 microM, 1 microM, 500 nM, 250 nM, 100 nM, 50 nM, 10 nM, 1 nM concentrations of conjugate 7; lane 12, 5 microM conjugate 14. sc, supercoiled DNA; li, linear DNA; oc, open circle DNA. Figure 4 | Sequence-specific protection of pQBI PGK plasmid DNA from restriction endonuclease hydrolysis by ImImPyLDu86 (14). Sequence-specific protection of pQBI PGK plasmid DNA from restriction endonuclease hydrolysis by ImImPyLDu86 (14). (A) A pQBI PGK DNA restriction map for Eco52I and DraI, and schematic representation of the binding mode of conjugate 14 at the Eco52I hydrolysis sites. (B) Agarose gel (1.2%) electrophoresis of conjugate 14-treated pQBI PGK after Eco52I (lanes 3 --10) and DraI (lanes 11 and 12) digestion. Lane 1, 250-bp DNA marker; lane 2, pQBI PGK DNA control; lanes 3 and 11, no drug treatment; lanes 4 --10, 1 microM, 500 nM, 250 nM, 100 nM, 50 nM, 10 nM, 1 nM concentrations of conjugate 14; lane 12, 5 microM conjugate 14. sc, supercoiled DNA; li, linear DNA; oc, open circle DNA. Figure 5 | Hydrolysis of pQBI PGK by Eco52I in the presence of ImImPy. Hydrolysis of pQBI PGK by Eco52I in the presence of ImImPy. (A) The chemical structure of ImImPy. (B) Agarose gel (1.2%) electrophoresis of the products of Eco52I hydrolysis of pQBI PGK DNA in the presence of various concentrations of ImImPy (lanes 2 --5). Lane 1, supercoiled pQBI PGK control (sc); lane 2, 5 microM ImImPy; lane 3, 2 microM; lane 4, 1 microM; lane 5, 500 nM; lane 6, linear pQBI PGK control (li). oc, open circle DNA. Figure 6 | Sequence-specific protection by ImImPyLDu86 (14) of linear pQBI PGK plasmid DNA from restriction endonuclease hydrolysis. Sequence-specific protection by ImImPyLDu86 (14) of linear pQBI PGK plasmid DNA from restriction endonuclease hydrolysis. (A) A pQBI PGK DNA restriction map for HindIII and Eco52I. (B) Agarose gel (1.2%) electrophoresis of conjugate 14-treated linear pQBI PGK after Eco52I (lanes 3 --10) digestion. Lane 1, 250-bp DNA marker; lane 2, pQBI PGK DNA control (li); lane 3, no drug treatment; lanes 4 --10, 1 microM, 500 nM, 250 nM, 100 nM, 50 nM, 10 nM, 1 nM concentrations of conjugate 14. Figure 7 | The amount of pQBI PGK plasmid DNA [supercoiled (sc) and linear (li)] protected from Eco52I hydrolysis by different concentrations of conjugate 14. The amount of pQBI PGK plasmid DNA [supercoiled (sc) and linear (li)] protected from Eco52I hydrolysis by different concentrations of conjugate 14. The amounts of uncleaved supercoiled and linear DNA are given as dark and light shaded bars, respectively: at 1000 nM conjugate 14, 100% sc, 100% li; at 500 nM, 47% sc, 100% li; at 250 nM, 38% sc, 61% li; at 100 nM, 22% sc, 39% li; at 50 nM, 0% sc, 30% li; at 10 nM, 0% sc, 0% li. SUPPLEMENTARY MATERIAL : Supplementary Material is available at NAR Online. Backmatter: PMID- 12202752 TI - Drosophila melanogaster RECQ5 /QE DNA helicase: stimulation by GTP binding AB - The Drosophila melanogaster RECQ5 /QE gene encodes a member of the DNA helicase family comprising the Escherichia coli RecQ protein and products of the human Bloom's, Werner's, and Rothmund-Thomson syndrome genes. The full-length product of RECQ5 /QE was expressed in the baculovirus system and was purified. Gel filtration experiments indicated that RECQ5 /QE was present in an oligomeric state. The RECQ5 /QE protein hydrolyzed ATP and even more actively GTP in the presence of single-stranded DNA. ATP drove the DNA helicase activity of RECQ5 /QE, whereas GTP had little effect. GTP exhibited a stimulatory effect on DNA unwinding when it was used together with ATP. This effect was more apparent with non-hydrolyzable GTP analogs, such as GTPgammaS and GMPPNP. These results indicate that GTP binding to RECQ5 /QE triggers its DNA helicase activity. GTP binding increased the rate of strand separation without affecting the S0.5 (Km) values for the substrates during the DNA helicase reaction. The data collectively suggest that the RECQ5 /QE protein is activated upon GTP binding through the ATP-binding site. Keywords: INTRODUCTION : Helicases are ubiquitous enzymes involved in almost all aspects of nucleic acid metabolic pathways. Separation of double-stranded DNA (dsDNA) or base-paired regions in single-stranded DNA (ssDNA) molecules is a prerequisite for basic genetic processes, such as genome replication, repair, recombination and expression at multiple stages. The loss of helicase function causes a number of disorders in organisms. Three hereditary disorders (Werner's syndrome, Bloom's syndrome and a subset of Rothmund-Thomson syndrome) in humans are associated with the loss of function of the respective RecQ homologs [BLM, WRN and RTS ( --)]. The eukaryotic RecQ homologs are thus named due to their similarity to the Escherichia coli RecQ helicase, which participates in the bacterial RecF genetic recombination pathway . The RecQ helicase domains of BLM, WRN and RTS are similar to those of the E.coli RecQ enzyme. However, the mammalian proteins are larger, due to the presence of additional flanking domains. To date, five RecQ homologs have been identified in the human genome [RecQ1 /QL, BLM, WRN, RTS and RecQ5 ( --)]. In contrast to BLM, WRN and RTS, RecQ1 /QL and RecQ5 possess only short N-terminal regions preceding the helicase domain. The functions of these short N-terminal type RecQ homologs are hitherto unknown. Recently, we isolated a Drosophila RecQ5 homolog, RECQ5 /QE, specifically expressed in early embryos . The results of the Drosophila Genome project revealed that Drosophila melanogaster contains three RecQ homologs, Blm, Rts and RECQ5 /QE, in addition to a Werner exonuclease protein family member (,). The only short N-terminal-type RecQ protein in D.melanogaster appears to be RECQ5 /QE, since no RecQ1 /QL homolog has been identified so far. In both humans and Drosophila, RecQ5 exists as three isoforms produced by alternative splicing . Small isoforms of human RecQ5 localize to the cytoplasm, while the large isoform is nuclear . The RECQ5 /QE protein isolated from Drosophila by our group represents the large isoform, and localizes to the nucleus (S. Maruyama and K. Kawasaki, unpublished results; ). The predicted RECQ5 /QE gene product is a 1058 amino acid protein that contains a helicase domain comprising seven helicase motifs. We propose that RECQ5 /QE encodes an active helicase, based on the results of studies on the in vitro translational product . In addition, a small RECQ5 isoform may function as a 3' to 5' DNA helicase . The presence of multiple RecQ homologs in a single organism suggests that each enzyme has a distinct role . Therefore, to elucidate the mechanistic and functional characteristics of RECQ5 /QE, it is necessary to determine the biochemical properties of the protein. To achieve this goal, we have employed an insect-based expression system that permits the preparation of homogeneous RECQ5 /QE protein with a reasonable yield. We report here that the RECQ5 /QE protein is a ssDNA-stimulated ATPase and an ATP-dependent DNA helicase. Furthermore, we demonstrate that RECQ5 /QE is a ssDNA-dependent GTPase, and that GTP binding stimulates the ATP-dependent DNA helicase activity of the protein. MATERIALS AND METHODS : Construction of BmNPV --RECQ5 /QE | The full-length DraI --NcoI RECQ5 /QE cDNA ( --16 to +3487) was cloned into the SmaI --NcoI sites of the pBm31 transfer vector under the control of the polyhedrin promoter. A recombinant virus, BmNPV --RECQ5 /QE, was generated by homologous recombination in BmN cells by co-transfection with polyhedrin-deficient Bombyx mori nucleopolyhedrovirus (BmNPV-abb) . Purification of the RECQ5 /QE protein | Typically, 26 plates (150 mm dishes) of BmNPV --RECQ5 /QE- infected BmN cells were washed with cold phosphate-buffered saline (PBS), harvested, and stored at --80C, 2 days after infection. Infected cells (6.7 g) were suspended in 33 ml (5 ml /g cell) of lysis buffer [50 mM Tris --HCl, pH 8.5, 10 mM dithiothreitol (DTT), 1 mM phenylmethylsulfonyl fluoride (PMSF), 1% Nonidet P-40 (Calbiochem) and protease inhibitor cocktail (Complete; Roche)]. After 30 min on ice, the lysate was centrifuged at 10 000 g for 10 min. The pellet was extracted with 0.45 M KCl in lysis buffer for 30 min on ice, and was subjected to further centrifugation. The majority of the RECQ5 /QE protein was recovered in the supernatant, which was further diluted to an electronic conductivity equivalent to 0.3 M KCl with Buffer Q (50 mM Tris --HCl, pH 7.5, 1 mM EDTA, 1 mM DTT, 1 mM PMSF, 10% glycerol), and was subjected to chromatography on DEAE -- Sepharose (11 ml; Pharmacia). The flow-through fraction from DEAE --Sepharose was diluted to a KCl concentration of 0.1 M, and was loaded onto a heparin --Sepharose CL6B column (7 ml; Pharmacia) pre-equilibrated with 0.1 M KCl in Buffer Q. A linear KCl gradient (0.1 --1.0 M KCl) was used for elution. The RECQ5 /QE protein eluted at 0.44 M KCl, as determined by SDS --polyacrylamide gel electrophoresis (PAGE) of fractions and immunoblotting with anti-RECQ5 /QE antibodies . The DNA-dependent ATPase and GTPase activities were measured. The active fraction (8 ml) was loaded onto a Mono S column (1 ml HR5 /5; Pharmacia) pre-equilibrated with 20 mM Tris --HCl, pH 7.5, 0.1 M NaCl, 0.5 mM EDTA and 1 mM DTT, and was eluted using 10 vol of a linear NaCl gradient (0.1 --1.0 M) in the same buffer. The RECQ5 /QE protein eluted at similar0.35 M NaCl. The purified, active protein was stored at --70C. The heparin --Sepharose fraction was also analyzed using Superose 6 (HR10 /30; Pharmacia) pre-equilibrated with 50 mM HEPES, pH 7.5, 300 mM KCl, 1 mM DTT and 1 mM EDTA. The protein concentration was determined with a BioRad protein assay kit, using bovine serum albumin as the standard. Preparation of extracts from Drosophila Schneider cells | Drosophila Schneider S2 cells were harvested from a confluent 100 mm dish. The collected cells were washed once with PBS and were resuspended in 0.5 ml of 20 mM HEPES --NaOH pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1 mM PMSF, 1 mM DTT, and protease inhibitor cocktail. After 20 min on ice, the DNA was sheared through a 25G needle. The suspension was centrifuged at 10 000 g for 20 min. The supernatant was filtered through a 0.45 microm membrane. A 0.2 ml aliquot of supernatant was loaded on the Superose 6 column. DNA-dependent ATPase and GTPase assays | DNA-dependent ATPase assays were performed as described previously . The reaction mixture (20 microl), containing 20 mM Tris --HCl, pH 7.5, 2 mM DTT, 90 microg /ml bovine serum albumin, 50 microM [nucleotides (nt)] M13mp18 virion DNA and 1.3 mM [14C]Mg2+ --ATP (14 MBq /mmol; Amersham), was incubated at 27C for 30 min. For the measurement of the GTPase activity, 2.6 mM [35S]Mg2+ --GTP (360 MBq /mmol; NEN) was employed, instead of labeled ATP. The reaction was stopped by the addition of 10 microl of 3 mM ATP, ADP, AMP, and 25 mM EDTA on ice. Aliquots were spotted onto a polyethyleneimine sheet (Polygram CEL300PEI; Macherey-Nagel) and were developed in 0.5 M LiCl, 1 M HCOOH. Radioactivity was quantitated using a BAS2500 Imaging plate reader (Fuji). Preparation of helicase substrates | A 69mer (5'-CCA AGC TTG CAT GCC TGC AGG TCG ACT CTA GAG GAT CCC CGG GTA CCG AGC TCG AAT TCG TAA TCA TGG-3') was labeled with [gamma-32P]ATP using T4 polynucleotide kinase (TaKaRa), and was purified by Sephadex G-50 chromatography (Probe-Quant; Pharmacia). The labeled oligomer was annealed to M13 mp18 virion DNA. Alternatively, a 17mer (5'-GTAAAGACCGACGGCCAGT-3') was annealed to M13mp18 virion DNA and was labeled with [alpha-32P]dCTP and dGTP to produce a 20mer, using Klenow fragment (TaKaRa). The labeled product was purified by Sephadex G-50 chromatography. Next, the annealed substrates were purified by Sepharose CL6B gel filtration. For the determination of polarity, a 48mer (5'-GTG CCA AGC TTG CAT GCC TGC AGG TCG ACT CTA GAG GAT CCC CGG GTA-3') was labeled with [alpha-32P]dCTP using Klenow fragment after annealing to M13mp18 virion DNA, or with [gamma-32P]ATP using T4 polynucleotide kinase before annealing to virion DNA. After purification using Sephadex G-50 chromatography, the annealed substrates were digested with HincII and were purified by Sepharose CL6B gel filtration. Helicase assays | Helicase assays measure strand displacement activity, whereby a partially dsDNA substrate is converted to its component single-strand products. The helicase substrate, 1.0 microM (nt) M13mp18 ssDNA annealed with 32P-labeled oligomer (20mer or 69mer), was incubated with purified RECQ5 /QE in a reaction mixture (20 microl) comprising 20 mM Tris --HCl, pH 7.5, 0.1 mM EDTA, 2 mM DTT, 90 microg /ml bovine serum albumin, 2 mM MgCl2 and 2 mM ATP. After an incubation at 27C for 10 min, the reaction was stopped by the addition of 5 microl of 75 mM EDTA, 5% Sarkosyl, 0.1% Bromophenol blue, 30% glycerol, and 2 microl of phenol /chloroform (1:1) on ice. The reaction products were separated on a 1% agarose gel (for the 69mer-labeled oligonucleotide) in buffer containing 40 mM Tris acetate, pH 8.0, 1 mM EDTA, as described elsewhere , or on a 12% polyacrylamide gel (for the 20mer-labeled oligonucleotide) in buffer containing 89 mM Tris borate, pH 8.3, 2 mM EDTA. Polyacrylamide gels were dried on Whatman DE81 paper and agarose gels were dried on GelBond PAG film (BMA). Dried gels were analyzed using the BAS 2500 Imaging plate reader, or were exposed to X-ray film for autoradiography. Experiments were repeated at least twice. The results were reproducible with little gel-to-gel variation, thus allowing quantitation among samples on different gels. Quantitative data from a single gel are shown. SDS --PAGE, immunoblotting and silver staining | SDS --PAGE was performed according to the procedure of Laemmli . Rainbow molecular weight markers (Amersham) or prestained molecular weight markers (NEB) were used as molecular weight standards. Immunoblotting was performed as described previously . Silver staining of the gel was achieved using the Silver Stain Plus kit (BioRad). RESULTS : Purification of full-length RECQ5 /QE protein | Initial attempts to purify RECQ5 /QE expressed in E.coli were unsuccessful because of the poor expression and the insolubility of the protein. To circumvent these problems, the recombinant virus, BmNPV --RECQ5 /QE, containing the full-length RECQ5 /QE under the control of the BmNPV polyhedrin promoter, was generated (Materials and Methods). Cells infected with BmNPV --RECQ5 /QE yielded a soluble product that migrated on SDS --PAGE with an apparent molecular mass of 120 kDa. This corresponded to the expected molecular mass determined from the open reading frame of RECQ5 /QE. Immunoblotting of crude lysates of infected cells with anti-RECQ5 /QE antibodies revealed a single 120 kDa band, confirming that the overexpressed protein was the desired product. The majority of the RECQ5 /QE protein was recovered from a 0.45 M KCl extract and not the cytosol, suggesting nuclear localization. The protein was not detected in BmN cells prior to infection or by mock infection (data not shown). The 120 kDa protein was purified to near homogeneity by successive chromatography steps, as monitored by immunoblotting and DNA-dependent ATPase and GTPase assays (Materials and Methods; Fig. A). The RECQ5 /QE protein is oligomeric | We examined the native molecular mass of the RECQ5 /QE protein. Superose 6 gel filtration of the RECQ5 /QE protein in the presence of 0.3 M KCl revealed a peak at 480 kDa (Fig. B, C and E). This apparent molecular mass corresponds to a tetramer of the 120 kDa RECQ5 /QE protein. We did not detect RECQ5 /QE in smaller molecular weight fractions corresponding to monomers or dimers. The RECQ5 /QE protein in cultured cells showed the same molecular mass (Fig. D), suggesting that the protein exists in an oligomeric form. The RECQ5 /QE protein is a DNA-dependent ATPase | 'Walker Box' motifs, which predict ATPase activity, were observed in the primary sequence of RECQ5 /QE. The ability of the RECQ5 /QE protein to hydrolyze ATP was therefore examined. Figure shows that RECQ5 /QE is associated with an ATPase activity that is strongly dependent on the presence of ssDNA. The substrate turnover rate was calculated as 18.2 s --1 with M13mp18 virion DNA as a co-factor, and 3.23 s --1 without DNA. Moreover, the ATPase activity required divalent ions. Mg2+ could be replaced by Mn2+ or Ca2+, but not Zn2+ . This specific requirement of divalent cations for activity is similar to that observed with the E.coli RecQ protein . Monovalent ions inhibited the ATPase activity (50% inhibition with 65 mM NaCl). The ATPase activity of the purified protein was stimulated by ssDNA, while dsDNA had no effect . The RECQ5 /QE protein is a DNA helicase | To determine if the ATP hydrolysis by RECQ5 /QE is coupled to the helicase activity, we examined whether the purified enzyme could displace a 32P-labeled oligodeoxynucleo tide from single-stranded circular M13mp18 DNA. The RECQ5 /QE protein separated the oligonucleotide from the ssDNA ring in the presence of ATP and Mg2+ . A time-course of RECQ5 /QE-mediated strand displacement is displayed in Figure A. Over 50% of the labeled fragment was displaced within 3 min, and the reaction was close to completion within 10 min. The quantity of the labeled fragment separated from the single-stranded circular DNA depended on the amount of RECQ5 /QE protein added (Fig. B). The amount of protein required to separate the 69mer from the complementary ssDNA was twice that needed for the 20mer separation. RECQ5 /QE separated the 20mer, 69mer and the Y-shaped short DNA region from single-stranded circular DNA, but not long duplex or blunt duplex DNA (data not shown). Therefore, we suggest that the RECQ5 /QE DNA helicase has low processivity. Since blunt-ended duplex DNA was not a substrate for the RECQ5 /QE helicase, we prepared single-stranded linear DNA with 3' or 5'-end labeled double-stranded regions at both ends (Materials and Methods). As observed for the E.coli RecQ helicase, displacement of DNA strands by RECQ5 /QE occurred in the 3' to 5' direction with respect to the ssDNA flanking the duplex (Fig. C). The RECQ5 /QE DNA helicase activity was dependent on ATP hydrolysis . No helicase activity was observed when the ATP was replaced with the non-hydrolyzable ATP analog, adenosine 5'-O-(thiotriphosphate) (ATPgammaS). Furthermore, in the presence of 2 mM Mg2+ --ATP, the addition of the same concentration of Mg2+ --ATPgammaS partially inhibited the strand separation activity. ATP (2 mM) was replaceable by dATP, although other common nucleotide triphosphates either could not, or only weakly supported the DNA strand separation catalyzed by RECQ5 /QE. These cofactor requirements are similar to those observed for E.coli RecQ and the small isoform of RECQ5 (,). The RECQ5 /QE protein exhibits GTPase activity in the presence of ssDNA | In the presence of GTP, the RECQ5 /QE protein sustained weak DNA strand separation . The purified protein was tested for its ability to hydrolyze GTP in the presence and absence of DNA. Unexpectedly, a ssDNA-dependent GTPase activity that was several fold higher than the ATPase activity was readily detected . The GTP hydrolysis required the presence of ssDNA, but was not activated by dsDNA . The substrate turnover rate of GTPase using M13mp18 virion DNA as a co-factor was 75.8 s --1 (3.63 s --1 without DNA). The GTPase activity required divalent ions. We additionally noted that Mg2+ could be replaced by Mn2+ and partially by Ca2+, but not Zn2+ . The GTPase activity as well as the ATPase activity was absent in the corresponding fraction position in chromatography when the extracts were prepared from mock-infected cells. Both ATPase and GTPase activities are intrinsic to the RECQ5 /QE protein | The ATPase and GTPase activity rates of RECQ5 /QE were plotted as functions of the ATP and GTP concentrations, respectively (Fig. A). The apparent Km values of ATP and GTP were 0.39 and 0.88 mM, respectively. Further characterization of the enzyme revealed that the ATPase activity was inhibited competitively by GTP and its analogs (Fig. B). Conversely, the GTPase activity was strongly inhibited in a competitive manner by ATP and its analogs (Fig. C). However, UTP and CTP did not inhibit the ATPase and GTPase activities as strongly. The optimum reaction conditions, in terms of pH, metal ions and salt, were quite similar for both activities. The purified protein exhibited both the ATPase and GTPase activities in the presence of ssDNA, but not dsDNA. Both the ATPase and GTPase activities were concomitantly inhibited by specific antibodies against RECQ5 /QE (Fig. D). These data suggest that the active centers for the ATPase and GTPase activities are the same or indistinguishable in the RECQ5 /QE protein. We conclude that both the ATPase and GTPase activities are intrinsic properties of the RECQ5 /QE protein, based on the following observations: (i) the purified protein possesses both ATPase and GTPase; (ii) a significant correlation is noted between the ATPase and GTPase activities and the presence of the RECQ5 /QE protein throughout the purification steps in multiple preparations, while the corresponding activities were absent from extracts prepared from a mock infection of virus; (iii) GTP and GTP analogs competitively inhibit the ATPase activity; (iv) ATP and ATP analogs competitively inhibit the GTPase activity; (v) both activities display almost identical characteristics in terms of their optimal reaction requirements; (vi) GTP, albeit a weaker binding substrate than ATP, is also used in the DNA helicase activity of RECQ5 /QE; and (vii) both the ATPase and GTPase activities were neutralized by anti-RECQ5 /QE antibodies. GTP stimulates the helicase activity of RECQ5 /QE | Since ATP and GTP bind to either the same or an indis tinguishable site of RECQ5 /QE, we explored the possibility that GTP affects the strand separation activity of the protein. GTP resulted in weaker DNA helicase activity than ATP . The GTPase activity of the RECQ5 /QE protein was higher than that of the ATPase . To date, we have not established a function of RECQ5 /QE that specifically requires GTP hydrolysis. GTP supported only weak helicase activity (Fig. , lane 7), but did not inhibit the ATP-dependent DNA helicase activity (Fig. , lane 6). Surprisingly, stimulation of the ATP-dependent helicase activity by GTP was observed in the presence of a limited amount of RECQ5 /QE (Fig. , lane 5 versus 6). Under these conditions, ATP strongly inhibited the ssDNA-dependent GTPase activity of the RECQ5 /QE protein, while GTP reduced the ssDNA-dependent ATPase activity to approximately half the original value. Since guanosine 5'-O-(thiotriphosphate) (GTPgammaS) inhibited the ssDNA-dependent ATPase of RECQ5 /QE to the same extent as GTP, a non-hydrolyzable GTP analog, GTPgammaS, was employed to distinguish whether the stimulation of activity was caused by GTP binding or hydrolysis. We observed that GTPgammaS stimulated the ATP-dependent DNA helicase activity to a greater extent than GTP (Fig. , lane 4). Therefore, we propose that the binding of GTP to RECQ5 /QE specifically stimulates its DNA helicase activity. GTP binding to RECQ5 /QE is needed for stimulation of the helicase activity | GTPgammaS neither supported the helicase reaction itself (Fig. , lane 12) nor induced the DNA helicase activity in the presence of GTP (Fig. , lane 9). Moreover, ATPgammaS did not support the helicase reaction (Fig. , lane 11), was inhibitory to this reaction (Fig. , lane 3), and did not induce the DNA helicase activity in the presence of GTP (Fig. , lane 8). GTPgammaS was slowly hydrolyzed by the GTPase, while guanosine 5'-(beta,gamma-imidotriphosphate) (GMPPNP) was not. GMPPNP activated the DNA helicase activity in the presence of ATP, in contrast to GDP . We noted that the stimulation was GTP analog-specific and concentration-dependent (data not shown). Furthermore, GTP hydrolysis was not necessary for triggering the activity, although the requirement for GTP binding was evident. Although non-hydrolyzable GTP analogs inhibited the ATPase activity of the RECQ5 /QE protein, the remaining activity associated with strand separation was not only sufficient, but was actually enhanced. Helicase activity stimulation by GTP analogs is caused by an increase in the strand separation rate in the RECQ5 /QE helicase reaction | GMPPNP and GTPgammaS reduced the DNA-dependent ATPase activity of RECQ5 /QE to approximately half of that observed under stimulating conditions. In the helicase assay, a labeled 20mer was annealed to single-stranded circular M13mp18 DNA (7249b). The overall level of ATP hydrolysis reflected the presence of a large ssDNA region. The stimulatory effects of GMPPNP or GTPgammaS were specific for the helicase reaction. This stimulation may be caused by the efficient recognition of the substrate, ATP or DNA, or an increase in the strand separation rate during the RECQ5 /QE helicase reaction. GMPPNP did not affect the Km value for DNA (0.5 microM) or the S0.5 value for ATP (0.28 mM) in the DNA helicase reaction, but led to an increase in the Vmax value of the DNA helicase reaction . The addition of GTP and GTP analogs did not affect the polarity of the RECQ5 /QE helicase (data not shown). Therefore, GTP binding possibly stimulates the strand separation rate and /or the processivity of the DNA helicase. Figure 1 | The RECQ5 /QE protein has a molecular mass of 120 kDa and is oligomeric. The RECQ5 /QE protein has a molecular mass of 120 kDa and is oligomeric. (A) Coomassie blue-stained 10% SDS --polyacrylamide gel showing the purified RECQ5 /QE fraction (1.5 microg). The protein was purified as described in the Materials and Methods. The position of the RECQ5 /QE protein is indicated on the left. Molecular weight markers are specified on the right. (B) Size of the RECQ5 /QE protein, following Superose 6 gel filtration chromatography of the heparin --Sepharose fraction. The panel displays an SDS --PAGE analysis using silver staining. Vo, void volume. The numbers at the top represent the sizes of native molecular weight markers. Denatured molecular weight markers are shown on the right. (C) The panel illustrates immunoblots, using anti-RECQ5 /QE antibodies. (D) Size of the RECQ5 /QE protein. Extracts were prepared from Drosophila Schneider cells as described in the Materials and Methods. The RECQ5 /QE protein was analyzed after Superose 6 chromatography, by immunoblotting as in (C). (E) The molecular weight calibration to the retention time, using protein molecular weight standards (closed squares), including Blue Dextran (2000 kDa), thyroglobin (670 kDa), ferritin (440 kDa), catalase (232 kDa), aldolase (158 kDa), ovalbumin (43 kDa) and RNaseA (13.7 kDa). The arrow and the open circle indicate the position of the RECQ5 /QE protein. Figure 2 | RECQ5 /QE is a ssDNA-dependent ATPase and GTPase. RECQ5 /QE is a ssDNA-dependent ATPase and GTPase. Reaction mixtures (20 microl) containing purified RECQ5 /QE protein and [14C]ATP or [35S]GTP were incubated at 27C, as described in the Materials and Methods. Aliquots of the reaction mixture were taken at various time-points to determine the initial velocity of the reaction. The initial reaction rates are indicated for the ATPase in the absence (closed triangles) and presence (open triangles) of single-stranded circular M13mp18 virion DNA or double-stranded circular M13mp18 RFI DNA (open diamonds), and for the GTPase in the absence (closed circles) and presence of ssDNA (open circles) or dsDNA (open squares). Figure 3 | Helicase activity of the RECQ5 /QE protein. Helicase activity of the RECQ5 /QE protein. To analyze whether the purified RECQ5 /QE protein possesses helicase activity, a partially double-stranded substrate comprising single-stranded M13mp18 DNA annealed to a radiolabeled single-stranded 20mer (A) or 69mer (B) oligonucleotide was prepared. The substrate was incubated with the purified RECQ5 /QE protein (0.8 pmol), and the reaction products were separated on a polyacrylamide gel (A) or an agarose gel (B). The labeled oligonucleotide was further detected by autoradiography, as described in the Materials and Methods. (A) Time-course of RECQ5 /QE helicase activity. Samples were taken after incubation times of 0, 3, 10, 30 min (lanes 3 --6). Lane 1, heat denatured DNA; lane 2, no RECQ5 /QE protein. (B) Effect of the RECQ5 /QE protein on the helicase reaction. Lane 1, heat-denatured DNA; lane 2, no RECQ5 /QE protein; lanes 3 --6, 0, 0.2, 0.4 and 0.8 pmol of RECQ5 /QE protein. (C) RECQ5 /QE is a 3' to 5' DNA helicase. Helicase assays were performed using the substrate for determining the polarity of unwinding, as described in the Materials and Methods, using a 3'-labeled substrate (lanes 1 --3, top) or 5'-labeled substrates (lanes 4 --6, bottom) with RECQ5 /QE protein (lanes 3 and 6, 0.8 pmol). Lanes 1 and 4, heat-denatured DNA; lanes 2 and 5, no RECQ5 /QE protein. Figure 4 | RECQ5 /QE is an ATP-dependent DNA helicase. RECQ5 /QE is an ATP-dependent DNA helicase. Helicase assays were performed using a 20mer annealed to M13mp18 single-stranded circular DNA as the helicase substrate. The helicase activity of the RECQ5 /QE protein was measured as described in the Materials and Methods. ATP (open triangles) is a better substrate than GTP (open circles) for the DNA helicase activity. Closed squares, no NTP. Figure 5 | RECQ5 /QE ATPase activity is dependent on ATP concentrations. RECQ5 /QE ATPase activity is dependent on ATP concentrations. (A) The ssDNA-dependent GTPase (open circles) or ATPase (open triangles) activity of RECQ5 /QE (0.8 pmol) as a function of ATP or GTP concentrations at a constant Mg2+:ATP (or GTP) ratio of 1.0, in the presence of 2 mM Mg2+. Excepting this variation, reactions were carried out under standard conditions, as described in the Materials and Methods. (B) Competitive inhibition of the RECQ5 /QE ATPase by GTP. The DNA-dependent ATPase activity of RECQ5 /QE (0.8 pmol) was measured as described above, in the absence (open triangles) or presence (closed diamonds) of 2 mM Mg2+ --GTP. (C) Competitive inhibition of the RECQ5 /QE GTPase by ATP. The DNA-dependent GTPase activity of the RECQ5 /QE protein (0.8 pmol) was measured as described above, in the absence (open circles) or presence (closed squares) of 0.034 mM Mg2+ --ATP. Double-reciprocal plots are displayed (B and C). The derived Km values for ATP and GTP are 0.39 and 0.88 mM, respectively. (D) Both the ATPase and GTPase activities are inhibited by anti-RECQ5 /QE antibodies. The DNA-dependent ATPase (open and closed triangles) and GTPase (open and closed circles) activities of RECQ5 /QE (0.8 pmol) were measured as described above, in the presence of rabbit IgG (open triangles and open circles) or antibodies against RECQ5 /QE (closed triangles and closed circles). Figure 6 | GTP stimulates the RECQ5 /QE DNA helicase. GTP stimulates the RECQ5 /QE DNA helicase. A 32P-labeled 20mer annealed with M13mp18 single-stranded circular DNA was used as the helicase substrate. A limited amount of RECQ5 /QE protein (0.05 pmol) was employed in the helicase reaction in the presence of 2 mM Mg2+ --ATP (lanes 3 --6) and /or 2 mM Mg2+ --GTP (lanes 6 --9), or in their absence (lanes 10 --12). Two millimolar Mg2+ --ATPgammaS (lanes 3, 8 and 11) or Mg2+ --GTPgammaS (lanes 4, 9 and 12) was added to the reaction mixtures, as indicated at the top of the panel. Products were analyzed on polyacrylamide gels, as described in the Materials and Methods. Lane 1, heat-denatured substrate; lane 2, no incubation. Figure 7 | GTP binding is required for stimulation of the RECQ5 /QE DNA helicase. GTP binding is required for stimulation of the RECQ5 /QE DNA helicase. A 32P-labeled 20mer annealed to M13mp18 single-stranded circular DNA was used as the helicase substrate. A limited amount of the RECQ5 /QE protein (0.05 pmol) was used in the helicase reaction in the presence (lanes 7 --10) or absence (lanes 3 --6) of 2 mM Mg2+ --ATP. Two millimolar Mg2+ --GTPgammaS (lanes 4 and 8), Mg2+ --GMPPNP (lanes 5 and 9) or Mg2+ --GDP (lanes 6 and 10) was added to the reaction mixture, as indicated at the top of the panel. Products were analyzed using PAGE. Lane 1, heat-denatured substrate; lane 2, no incubation. Figure 8 | GTP binding to the RECQ5 /QE DNA helicase elevates Vmax without changing the S0.5 (Km) values for DNA and ATP. GTP binding to the RECQ5 /QE DNA helicase elevates Vmax without changing the S0.5 (Km) values for DNA and ATP. The RECQ5 /QE protein (0.05 pmol) and the indicated amounts of 32P-labeled 20mer annealed to M13mp18 DNA [microM (nt)] or Mg2+ --ATP (1:1) were used in the helicase reaction, as described in the Materials and Methods. The helicase reaction was quantitated using a Bas2500, as described earlier. (A) Helicase activity as a function of ATP concentration at a constant Mg2+:ATP ratio of 1.0, in the presence (closed diamonds) and absence (open triangles) of Mg2+ --GMPPNP, as described in the Materials and Methods. The curve was sigmoid, suggesting the cooperativity of ATP-binding sites in RECQ5 /QE . Therefore, the S0.5 value was used for ATP (0.28 mM). (B) Helicase activity of the RECQ5 /QE protein as a function of DNA concentration, in the presence (closed diamonds) and absence (open triangles) of GMPPNP. The Km value for DNA was 0.5 microM. Table 1 | ATPase and GTPase requirements DISCUSSION : RECQ5 /QE has a unique ssDNA-dependent GTPase activity | In this study, we demonstrated that RECQ5 /QE is a DNA-dependent GTPase and ATPase, and an ATP-dependent DNA helicase. Furthermore, GTP binding stimulated the DNA helicase activity. Most helicases favor ATP as their energy source, although in some cases GTP may be utilized in addition to ATP [e.g. HSV-1 helicase-primase , E.coli dnaB and T4 gp41 ]. The RECQ5 /QE protein prefers ATP for the helicase activity, but is more active in hydrolyzing GTP than ATP on ssDNA. No ssDNA-dependent GTPase similar to RECQ5 /QE has been reported so far. Helicase motifs share some homology with GTP-binding consensus sequences . A comparison of the RECQ5 /QE amino acid sequence with other members of the RecQ family and GTPases revealed similarities around the helicase motif I and in the GTPase P-loop [G-1 ] . Interestingly, the E.coli FtsZ protein usually behaves as a GTPase. However, a single amino acid substitution (G to S, producing FtsZ84) results in a change in the activity to an ATPase . This substitution at the phosphoryl-binding site is associated indirectly with purine recognition (adenine and guanine). Therefore, it is possible that the GTPase and ATPase active centers are the same in the RECQ5 /QE protein. This position is well conserved in most RecQ family members (L or V), but varies in Dm RECQ5 /QE (S), Ce E03A3.2 (S), Hs RECQ5(C), Mm RECQ5(C) and Dm RTS (T). It would be interesting to determine whether these RecQ proteins additionally exhibit the GTPase activity. RECQ5 /QE is an ATP-dependent DNA helicase | Recently, the small RECQ5 isoform (54 kDa) was purified and characterized . The ATPase characteristics of this protein are quite similar to those of our large isoform (120 kDa), and its DNA helicase activity may also be comparable (i.e. low processivity and polarity). Significantly, the DNA unwinding reaction catalyzed by the small RECQ5 isoform requires unexpectedly high protein concentrations. On the other hand, our large isoform, RECQ5 /QE, appears to be more active in the helicase reaction. These results are consistent with the results from in vitro translation products. Previously, we demonstrated that the full-length (amino acids 1 --1058) and the N-terminal half (1 --584) of RECQ5 /QE showed DNA helicase activity . The small RECQ5 isoform (1 --473) exhibited very weak DNA helicase activity using in vitro translation products (K. Kawasaki, unpublished results). RECQ5 /QE is an oligomeric helicase | The native molecular mass of the RECQ5 /QE protein was similar480 kDa, and no RECQ5 /QE protein was evident in smaller molecular mass fractions . The DNA helicase family is categorized into at least two classes, specifically, oligomeric ring and monomeric forms (,). Physical data indicate that the RECQ5 /QE DNA helicase belongs to the oligomeric type. The BLM helicase, a member of the RecQ family, exists as a hexamer . Harmon and Kowalczykowski demonstrated that the E.coli RecQ protein has a Hill coefficient of 3.3. The active complex formed upon RecQ helicase binding to DNA substrates is expected to be oligomeric, since multiple ATP-binding sites are utilized by the protein to achieve strand separation. These findings collectively suggest that members of the RecQ family are oligomeric DNA helicases. It should be noted that non-hydrolyzable analogs of helicase substrates assemble subunits into stable hexamers [dTMPPNP for T7 gp4 and ATPgammaS /GTPgammaS for T4 gp 41 ]. However, Superose 6 gel filtration experiments revealed that ATPgammaS did not change the apparent molecular mass of RECQ5 /QE (K. Kawasaki, unpublished results). It remains to be determined whether ATPgammaS has the ability to assemble the RECQ5 /QE protein into its active hexameric form in the presence of DNA . Role of ATP hydrolysis in the RECQ5 /QE helicase activity | We observed that ATP drove the RECQ5 /QE helicase activity, whereas GTP had little effect . However, the RECQ5 /QE protein hydrolyzes GTP on ssDNA more efficiently than ATP . These data suggest that the mechanism of the GTP-dependent DNA binding /release differs from that of the ATP-dependent DNA binding /release. Since the RECQ5 /QE helicase is oligomeric, it is possible that the protein subunits work cooperatively when utilizing ATP, but not GTP. Consequently, the GTPase activity of RECQ5 /QE is not coordinated to DNA binding /release for strand separation. Biochemical analyses of the NS-1 protein of the mouse minute virus and the UL5 protein of Herpes simplex virus revealed that most of the mutants retained significant levels of ATPase activity, while point mutations in the ATP-binding domain severely reduced the helicase activity (,). Accordingly, we hypothesize that the ATP-binding domain is important for the coupling of the ATPase and helicase activities. The ATP-binding sites of RECQ5 /QE function in both the ATP hydrolysis and the coupling of the ATPase and helicase activities. Although GTP hydrolysis occurs at the same or an indistinguishable site as that for ATP hydrolysis, the former reaction does not accomplish efficient strand separation. The non-hydrolyzable GTP analogs, GTPgammaS and GMPPNP, also bind RECQ5 /QE. Therefore, it is possible that GTP analogs induce a specific conformation in RECQ5 /QE through interactions with the ATP-binding site. Possible mechanisms of helicase stimulation by GTP analogs | Despite the partial inhibition of the RECQ5 /QE ATPase activity by GTPgammaS and GMPPNP, the strand separation activity was not inhibited, but rather, was stimulated. This implies that GTP analogs modulate the DNA helicase activity of RECQ5 /QE under conditions where a decreasing supply of energy is available from ATP hydrolysis. Since the S0.5 (Km) values for DNA and ATP remained unchanged in the absence and presence of GTP analogs during the RECQ5 /QE helicase reaction, the increase in Vmax may be explained by an elevated strand separation rate and /or processivity. The helicase reaction consists of multiple steps, specifically, oligomerization, ssDNA binding, encounter with the dsDNA region, ATP hydrolysis, DNA binding /release, coordination between ATP hydrolysis and DNA, polar translocation on DNA, displacement from DNA, and recycling. It is therefore important to determine the specific step of the RECQ5 /QE reaction that is influenced by GTP analogs. There are several possible explanations for the stimulation by GTP analogs. One hypothesis is that the population of active enzyme is increased upon preventing the formation of a non-productive enzyme --DNA complex in the presence of GTP analogs. The helicase activity of RECQ5 /QE is inhibited by ssDNA, presumably because the protein is sequestered by ssDNA. Therefore, the GTP analogs may increase the effective concentration of RECQ5 /QE by blocking the sequestration of the enzyme. However, this possibility is unlikely, since the gel mobility shift experiments demonstrated that GTP analogs do not affect the preference or the strength of the RECQ5 /QE DNA binding (K. Kawasaki, unpublished results). Another potential explanation is that GTP analogs stimulate the formation of an oligomeric helicase. The RECQ5 /QE helicase exhibits an apparent native molecular mass corresponding to a tetrameric protein, while RecQ is a hexameric helicase. However, the native molecular mass of RECQ5 /QE remained unchanged on Superose 6 gel filtration in the presence of GTPgammaS (K. Kawasaki, unpublished results). In addition, a pre-incubation of the RECQ5 /QE helicase with GTP analogs did not cause stimulation of the helicase reaction. We cannot exclude the possibility that the active oligomeric formation is too fast to be detected, and is dependent on DNA. A third explanation for the stimulation of the RECQ5 /QE helicase by GTP analogs is that these compounds directly enhance the translocation rate and /or the processivity of the RECQ5 /QE helicase during strand separation. Non-catalytic nucleotide binding sites have been found in the F1-ATPase protein, which shares structural similarity with hexameric helicases . In this case, it is thought that the non-catalytic sites ensure cooperative catalysis between the catalytic sites. A study by Singleton et al. demonstrated that, in the complex of homo-hexameric T7 gene 4 helicase with ADPNP, only four of the six sites bind nucleotides at any time. GTP analogs may define the non-catalytic sites in the RECQ5 /QE oligomer, which ensures cooperativity between the catalytic sites for efficient strand separation. It is currently unclear how a homo-oligomeric helicase contains non-catalytic (permanently inactive) or empty, but active sites for NTP /NDP binding (,). It remains to be determined whether the binding of GTPgammaS or GMPPNP similarly defines non-catalytic sites in RECQ5 /QE helicase. Note that the above possibilities for the stimulation of RECQ5 /QE helicase by GTP analogs are not exclusive of each other. This study is an initial characterization of the large isoform of the RECQ5 /QE helicase, and suggests a new mechanism of regulation or activation of the helicase activity. The RECQ5 /QE protein is accumulated in early embryos (K. Kawasaki, unpublished results; ). The helicase activity stimulation implies that RECQ5 /QE may be responsible for a quick response to DNA repair or /and processing of stalled replication forks in early embryonic DNA replication. The RECQ5 /QE helicase activity is regulated by cofactors or unknown effectors inducing an active or suppressive state. Figure 9 | Presence of a GTP-binding motif in the RECQ5 /QE protein sequence. Presence of a GTP-binding motif in the RECQ5 /QE protein sequence. (A) Comparison of amino acid sequences around the helicase motif I of RecQ family members. The RecQ helicases are specified on the left. The protein sequence alignment was performed using the Clustal W program . Residues indicated with dark or light shades represent identical or similar amino acids, respectively. Numbers on the right represent positions of amino acids from the first methionine. Dm RECQE, D.melanogaster RECQ5 /QE ; Hs RECQ5, Homo sapiens RecQL5 ; Mm RECQ5, Mus musculus RecQL5 ; Ce E03A3.2, Caenorhabditis elegans E03A3.2 ; Dm RTS, D.melanogaster RecQ4 ; Dm BLM, D.melanogaster BLM ; Hs WRN, H.sapiens Wrn ; Hs BLM, H.sapiens Blm ; Hs RTS, H.sapiens Rts ; Hs RECQ1 /QL, H.sapiens ATPaseQ1 /RecQL ; Sp Rqh1, Schizosaccharomyces pombe Rqh1 /Rad12 ; Sc SGS1, Saccharomyces cerevisiae SGS1 ; Ec RECQ, E.coli RecQ . (B) Comparison of the RECQ5 /QE helicase motif I and the GTPase P-loop. Ec FtsZ, E.coli FtsZ ; GTPaseG-1, consensus sequence motif G-1 present in members of the GTPase superfamily . The underlined residue indicates a single amino acid substitution within FtsZ84. Backmatter: PMID- 12202749 TI - Displacement of DNA-PKcs from DNA ends by the Werner syndrome protein AB - The DNA-dependent protein kinase (DNA-PK) complex, which is composed of a DNA-dependent kinase subunit (DNA-PKcs) and the Ku70 /80 heterodimer, is involved in DNA double-strand break repair by non-homologous end joining (NHEJ). Ku70 /80 interacts with the Werner syndrome protein (WRN) and stimulates WRN exonuclease activity. To investigate a possible function of WRN in NHEJ, we have examined the relationship between DNA-PKcs, Ku and WRN. First, we showed that WRN forms a complex with DNA-PKcs and Ku in solution. Next, we determined whether this complex assembles on DNA ends. Interestingly, the addition of WRN to a Ku:DNA-PKcs:DNA complex results in the displacement of DNA-PKcs from the DNA, indicating that the triple complex WRN:Ku:DNA-PKcs cannot form on DNA ends. The displacement of DNA-PKcs from DNA requires the N- and C-terminal regions of WRN, both of which make direct contact with the Ku70 /80 heterodimer. Moreover, exonuclease assays indicate that DNA-PKcs does not protect DNA from the nucleolytic action of WRN. These results suggest that WRN may influence the mechanism by which DNA ends are processed. Keywords: INTRODUCTION : Werner syndrome (WS) is a premature aging disease (progeria) with features typical of normal aging such as graying and loss of hair, atherosclerosis, osteoporosis, type II diabetes mellitus and vascular disease, as well as an unusually high incidence of tumors ( --). The first signs of this disorder appear after puberty and the full symptoms become apparent in individuals 20 --30 years of age. Myocardial infarction and cancer are the most common causes of death in WS patients, who have a median age of death of similar47 years. Primary fibroblasts from WS individuals show decreased lifespan and divide approximately half the number of times as normal fibroblasts. WS cells show an elevated rate of chromosomal translocations and extensive genomic deletions ( --). WS is caused by mutations in a single gene that encodes a 1432 amino acid protein (Werner syndrome protein, WRN). WRN has a central domain that shows strong homology to the RecQ helicases (,) and the N-terminal region of WRN is highly homologous to the nuclease domain of Escherichia coli DNA polymerase I and ribonuclease D . Helicase and exonuclease activities with a 3'->5' directionality have been demonstrated in vitro using the recombinant protein ( --). Our biochemical results and those from Bohr and colleagues show that WRN interacts with the Ku70 /80 heterodimer (Ku) (,) and that this interaction alters the properties of WRN exonuclease activity . Ku70 /80 is a factor required for the repair of DNA double-strand breaks (DSBs) by non-homologous end joining (NHEJ) ( --). Furthermore, the localization of Ku70 /80 at telomeres has suggested that in addition to its role in DNA repair this factor might participate in the protection of telomeric sequences from nuclease and ligase activities ( --). Thus, depending on the cellular context, Ku70 /80 may facilitate or prevent DNA end joining. DSBs can be caused by a variety of exogenous and endogenous agents and are repaired either by using an intact copy of the broken DNA region as a template (homologous recombination) or by direct rejoining of the broken DNA ends (NHEJ) (,). Both mechanisms operate in eukaryotic cells, however, NHEJ is thought to be the predominant pathway, in particular during the G1 phase of the cell cycle (,). Biochemical and genetic analyses have established that at least five components are involved in the NHEJ pathway: the Ku70 /80 heterodimer, DNA-PKcs, XRCC4 and DNA ligase 4 ( --). The initial step in the NHEJ repair process requires the binding of Ku70 /80 to both ends of a broken DNA molecule ( --,). Then Ku recruits DNA-dependent protein kinase (DNA-PKcs), a serine /threonine protein kinase with homology to phosphatidylinositol kinases that is activated by DNA . DNA-PKcs can phosphorylate itself as well as Ku and other cellular factors such as p53 . However, the physiologically relevant targets of this kinase remain elusive. Once bound to DNA, Ku70 /80 can translocate inward from the DNA end in an ATP-independent manner, thus allowing the sequential binding of multiple Ku70 /80 heterodimers to one DNA molecule. Unlike Ku70 /80, DNA-PKcs is thought to remain bound to the DNA ends (,) and autophosphorylation has been proposed to facilitate the release of DNA-PKcs from the DNA . The final steps in the repair process involve the assembly on the DNA ends of a ligase4 /XRCC4 complex, which is required for the ligation of the two cohesive DNA ends. The events occurring during the transition from a Ku:DNA-PKcs to a ligase4 /XRCC4 complex have not been well characterized and little is known about the cellular factors involved in this process. The observation that both WRN and DNA-PKcs are stably recruited to DNA by interactions with Ku70 /80 raises the question of whether these factors assemble as a mutiprotein complex on DNA ends. In this study, to gain more insight into the relationship between WRN and the NHEJ repair machinery, we have investigated the interaction between WRN, DNA-PKcs and the Ku heterodimer. Our results show that DNA-PKcs is retained with Ku70 /80 on a WRN affinity column, thus demonstrating that a trimeric complex forms in solution. Further analysis indicates that the association between DNA-PKcs and WRN occurs only in the presence of Ku70 /80. In striking contrast, our experiments show that in the presence of DNA ends, the interaction of Ku70 /80 with WRN and DNA-PKcs is mutually exclusive, and equimolar amounts of WRN can displace DNA-PKcs from a Ku70 /80:DNA complex. The removal of DNA-PKcs from DNA-bound Ku70 /80 requires the C-terminal region of WRN, which cooperates with the WRN N-terminal domain in binding to the Ku70 /80 heterodimer. One consequence of this process is that DNA-PKcs cannot prevent the nucleolytic processing of DNA ends by the WRN exonuclease. Thus, the assembly on DNA of alternative complexes may be part of a regulatory process required for the repair of damaged DNA. MATERIALS AND METHODS : Protein purification | DNA-PKcs was purified from 500 mg HeLa cell nuclear extracts. The nuclear extract was prepared as described previously and loaded on a 30 ml fast flow DEAE --Sepharose column in buffer A (20 mM Tris --HCl, pH 7.5, 1.0 mM EDTA, 2 mM dithiothreitol, 10% glycerol and a mixture of protease inhibitors) containing 100 mM KCl. The flow-through from the DEAE --Sepharose (containing DNA-PKcs) was then incubated with a GST --Ku80(500 --732) affinity resin. After extensive washes with buffer A containing 150 mM KCl, the bound Ku70 /80 and DNA-PKcs were eluted from the beads with buffer A containing 400 mM KCl. The eluate was dialyzed against buffer A containing 100 mM KCl and then applied to a native DNA --cellulose column. After washing the column with buffer A containing 100 mM KCl, DNA-PKcs was eluted from the column with a salt gradient in buffer A (from 250 to 350 mM). Under these conditions Ku70 /80 remains bound to the resin. The purity of the purified DNA-PKcs preparation was determined by silver staining. The C-terminal region of Ku80 used in the purification of DNA-PKcs was expressed in E.coli as a GST fusion protein [GST --Ku80(500 --732)]. Escherichia coli BL21 cells were transformed with pGEX-3X-Ku80(500 --732) and logarithmically growing cells were induced with 0.3 mM IPTG for 2.5 h at 30C. The cells were harvested by centrifugation, resuspended in phosphate-buffered saline (PBS) containing 1% Triton X-100 and a mixture of protease inhibitors, sonicated and centrifuged at 15 000 g for 20 min at 4C. The bacterial lysate was incubated with 5 ml of glutathione --Sepharose slurry (Amersham Pharmacia Biotech) for 1 h at 4C. Beads were washed three times in PBS with 1% Triton X-100, followed by washes in buffer A with 150 mM KCl. GST --Ku80 affinity resin was used for the purification of DNA-PKcs. Baculoviruses expressing Flag-tagged wild-type or mutant WRN protein were used to infect Sf9 cells. Whole cells lysates were prepared in lysis buffer (10 mM HEPES, pH 7.5, 100 mM NaCl, 1.5 mM MgCl2, 0.5% Nonidet P-40 and protease inhibitors) and proteins were purified by DEAE --cellulose column and affinity chromatography on anti-Flag resin as described (,). GST --WRN(1 --50) and GST --WRN(1306 --1432) were generated by insertion of the appropriate cDNA sequences into pGEX-2T vector (Pharmacia) and expressed in E.coli. GST fusion proteins were purified on glutathione affinity resin. Histidine-tagged Ku70 and Ku80 were co-expressed in Sf9 cells and the heterodimer was purified on metal affinity resin (Talon, Clontech Biotech) and by DNA --cellulose chromatography . Protein binding assay | Protein binding assays were performed as described (,). Flag --WRN and Flag --HCV polymerase were expressed using a baculovirus expression system. Flag --WRN and Flag --HCV polymerase were first immobilized on anti-Flag resins and then the beads were incubated with nuclear extracts from HeLa cells. Bound proteins were eluted from the resins with BCO buffer (1 M KCl, 10 mM Tris --HCl, pH 7.5, 1 mM EDTA, 5% glycerol, 1 mM dithiothreitol) containing a cocktail of protease inhibitors and visualized by silver staining. Bound proteins were also resolved by 5% SDS --PAGE and transferred to nitrocellulose. Western blotting was performed using rabbit anti-DNA-PKcs antibodies (Santa Cruz Biotechnology). In similar assays, Flag --WRN immobilized on anti-Flag beads was incubated with purified DNA-PKcs or Ku70 /80:DNA-PKcs. After extensive washes, the bound proteins were eluted as described above and analyzed by silver staining and immunoblotting. For GST pull-down assays GST --WRN deletion mutants were first immobilized on glutathione beads and then incubated with 500 microl of cell lysate from Sf9 cells expressing Ku70 and/or Ku80. After extensive washing, bound proteins were eluted with high salt buffer (1.0 M KCl, 10 mM Tris, pH 7.5, 1 mM EDTA, 1 mM dithiothreitol and 10% glycerol) and analyzed by SDS --PAGE and western blotting with anti-Ku70 or anti-Ku80 antibodies. Exonuclease assay | DNA exonuclease activity was measured as described (,). The 30mer oligomer 5'-CGCGCCGATTTCCCGCTAGC AATATTGTGC-3', partially complementary to the 46mer 5'-GCGCGGAAGCTTGGCTGCAGAATATTGCTAGCGG GAAATCGGCGCG-3', was labeled at the 5' end with [32P]ATP and T4 polynucleotide kinase. The 30mer and 46mer oligonucleotides were annealed by boiling and slow cooling to room temperature. Exonuclease reaction mixtures contained 40 mM Tris --HCl, pH 7.5, 4 mM MgCl2, 5 mM dithiothreitol, 1 mM ATP, 0.1 mg /ml bovine serum protein, 40 fmol (100 000 c.p.m.) DNA substrate and 100 fmol Ku70 /80, DNA-PKcs and /or WRN in a final volume of 10 microl. The DNA was incubated with Ku70 /80 and DNA-PKcs for 10 min at room temperature, then WRN was added to the reaction mixture and the incubation was continued for an additional 10 min at room temperature. The reaction was terminated by the addition of 2 microl of a formamide solution. After incubation at 95C for 3 min, the products were resolved on a 14% polyacrylamide --urea gel and visualized by autoradiography. Electrophoretic mobility shift assays | A 40 /46mer DNA substrate was prepared from a 40mer oligomer (5'-CGCGCCGATTTCCCGCTAGCAATATTGTGCAGCCAAGCTT-3', partially complementary to a 46mer oligomer, 5'-GCGCGGAAGCTTGGCTGCAGAATATTGC TAGCGGGAAATCGGCGCG-3'). The DNA duplex (80 fmol, 200 000 c.p.m.) was incubated with 200 fmol Ku70 /80 and wild-type or mutant WRN in 15 microl of buffer (10 mM Tris --HCl, pH 7.5, 80 mM NaCl, 4 mM KCl, 1.5 mM MgCl2, 2 mM EDTA and 10% glycerol) at 25C for 10 min. In the displacement reactions, 200 fmol Ku70 /80 and DNA-PKcs were first incubated with the 40 /46mer DNA substrate for 10 min at room temperature, then 200 fmol wild-type or mutant WRN was added to this reaction and incubated for an additional 5 min at room temperature. The samples were resolved by electrophoresis at 10 V /cm through a 4% polyacrylamide gel at 8C. The gels were dried on Whatman 3MM paper and subjected to autoradiography. In vitro kinase assay | Bacterially expressed GST --p53 was bound to glutathione beads and then incubated with either 200 pmol purified DNA-PKcs, 100 microg HeLa cell nuclear extract or buffer alone (20 mM HEPES, pH 7.5, 100 mM KCl, 2 mM dithiothreitol, 1.5 mM MgCl2, 10% glycerol and 1 mM ATP) in the presence of 20 microCi [32P]ATP and 100 pmol double-stranded oligonucleotide at room temperature for 30 min. The beads were washed four times with buffer containing 20 mM HEPES, pH 7.5, 200 mM KCl, 2 mM dithiothreitol, 1.5 mM MgCl2, 10% glycerol and 0.1% Nonidet P-40 and the bound proteins were released by boiling the beads in SDS --PAGE sample buffer. The products were resolved on 8% polyacrylamide denaturing gels and visualized by autoradiography. Interaction assay on streptavidin-immobilized DNA | A 39 /39mer DNA duplex (100 pmol in 300 microl of H2O) biotinylated on one 5' end was incubated with 30 microl of streptavidin --Sepharose 4B for 1 h at 4C. After extensive washes with buffer A containing 150 mM KCl, the streptavidin-bound DNA was incubated with 1 microg purified Ku70 /80:DNA-PKcs for 1 h at 4C. The unbound proteins were washed off with buffer A containing 150 mM KCl and half of the reaction mixture was incubated with 0.5 microg full-length WRN, while the other half was incubated with 0.5 microg mutant WRN(1 --749). Each reaction mixture was incubated at 4C for 30 min and the beads were then washed with buffer A containing 150 mM KCl. The bound proteins were released by boiling in buffer containing SDS and analyzed by immunoblotting with anti-DNA-PKcs antibody. RESULTS : WRN and DNA-PKcs form a complex with Ku in the absence of DNA | To determine whether WRN binds to DNA-PKcs, a nuclear extract from HeLa cells was incubated with Flag --WRN immobilized on anti-Flag resin and after extensive washes the bound proteins were eluted for subsequent analysis. In parallel reactions, Flag --HCV polymerase immobilized on Flag beads and anti-Flag only beads were used as control columns for the non-specific binding of proteins to the matrix. Silver staining of SDS --PAGE gels revealed that three proteins were selectively bound to the Flag-tagged WRN resin (Fig. A, lane 3) and not to the control resins (Fig. A, lanes 1 and 2). The two proteins, of similar65 and 90 kDa, were previously identified as Ku70 and Ku80. The additional band on the gel was a protein whose high molecular weight was suggestive of DNA-PKcs, a relatively large polypeptide of similar420 kDa. To establish the identity of this protein we performed immunoblot analysis using antibodies against DNA-PKcs. The antibody reacted with the high molecular weight protein eluted from the Flag --WRN resin (Fig. B, lane 4) but not with the eluates from the control resins (Fig. B, lanes 2 and 3), indicating that the protein corresponds to DNA-PKcs. Identical results were obtained when the nuclear extract was treated with DNase I before being loaded on the affinity column (data not shown). Next, to determine whether the association between WRN and DNA-PKcs is direct or requires the Ku heterodimer, Flag-tagged WRN was immobilized on affinity beads and incubated with either purified DNA-PKcs or a mixture of Ku70 /80 and DNA-PKcs. The DNA-PKcs used in this and the following assays was purified from HeLa cells and is functionally active, as indicated by its ability to phosphorylate p53 in the presence of DNA in vitro (Fig. C). After extensive washes, the bound proteins were eluted from the beads and analyzed by silver staining. As shown in Figure D, DNA-PKcs is retained on the Flag --WRN resin only in the presence of Ku70 /80, suggesting that Ku70 /80 is required to facilitate the interaction between WRN and DNA-PKcs. The finding that the binding of DNA-PKcs to WRN requires Ku70 /80 was confirmed in a similar pull-down experiment by immunoblot analysis of the resin-bound material (Fig. E). Both the N- and C-terminal regions of WRN are required to form a stable complex with Ku70 /80:DNA | Our earlier work demonstrated that the Ku70 /80 heterodimer is required for the recruitment of WRN to the DNA substrate . To examine in more detail which domain of WRN is required for the formation of a stable WRN:Ku70 /80 complex on DNA, we analyzed the behavior of full-length WRN and a series of WRN deletion mutants in electrophoretic mobility shift assays (EMSA). The incubation of a 40 /46mer DNA with Ku70 /80 leads to the formation of two complexes, which correspond to one or two Ku heterodimers bound to DNA (Fig. , lane 1) . Upon addition of full-length WRN an additional slower migrating band appeared in the gel (lanes 10 and 11), which we previously identified as the WRN:Ku70 /80:DNA complex . A similar pattern of shifted bands is observed in the presence of the deletion mutant WRNDelta(773 --1306) (lanes 8 and 9), suggesting that the region of WRN between amino acids 773 and 1306 is not required for the recruitment of WRN to DNA by Ku70 /80. On the other hand, the WRN mutants missing either the C-terminus [WRN(1 --50), WRN(1 --388) and WRN(1 --749)] or the N-terminus [WRN(1306 --1432)] failed to form a stable complex with Ku70 /80 on DNA, evident by the absence of a specific shifted complex (lanes 2 --7, 12 and 13). These results suggest that the N- and C-terminal domains of WRN are both required for the formation of a stable complex with Ku on DNA. The C-terminal region of WRN binds to the Ku70 /80 complex, but not to individual Ku subunits | We have previously shown that WRN binds to Ku80 but not to Ku70 and that the interaction between WRN and Ku80 is mediated by the N-terminus of WRN . Because the EMSA indicated that the C-terminal domain of WRN is also required to recruit this factor to DNA, we then asked whether this region of WRN might interact with the Ku70 /80 heterodimer. For this purpose, GST fusion proteins containing the N-terminus [GST --WRN(1 --50)], central domain [GST --WRN(50 --749)] or C-terminus [GST --WRN(1306 --1432)] of WRN (Fig. A) were immobilized on glutathione beads and then incubated with extracts from Sf9 cells expressing individual subunits or the Ku70 /80 heterodimer. After extensive washes, the bound proteins were resolved on a SDS --PAGE gel and analyzed by immunoblotting with antibodies against Ku70 and Ku80. In agreement with our previous data, Ku70 did not bind to any of the WRN mutants (Fig. B, top), while the first 50 amino acids of WRN [GST --WRN(1 --50)] were sufficient to bind to Ku80 (Fig. B, middle, lane 3). The N-terminus of WRN also bound to the Ku70 /80 heterodimer (Fig. B, bottom, lane 3). Interestingly, the C-terminal region of WRN [GST --WRN(1306 --1432)], which did not associate with individual Ku subunits, bound to the Ku70 /80 heterodimer (Fig. B, bottom, lane 5). To substantiate the association between Ku70 /80 and WRN N- and C-terminal regions, WRN mutants missing the N-terminus [WRN(50 --1432)], the C-terminus [WRN(1 --1308)] or both [WRN(50 --1308)] regions were used in protein binding assays. As shown in Figure C, Ku70 /80 bound to WRN(50 --1432) and WRN(1 --1308) but not to WRN(50 --1308), confirming that the N-terminus (amino acids 1 --50) and C-terminus (amino acids 1308 --1432) of WRN make direct contacts with Ku70 /80. Taken together with the results of the EMSA , these data suggest that the stable recruitment of WRN to DNA requires simultaneous interaction of the WRN N- and C-terminal regions with the Ku heterodimer. WRN displaces DNA-PKcs from the Ku70 /80:DNA complex | Studies from several laboratories have indicated that Ku70 /80 recruits DNA-PKcs to DNA ends. This process leads to the activation of DNA-PKcs kinase activity, probably as the result of conformational changes in DNA-PKcs upon binding to the Ku70 /80:DNA complex . Because our results indicate that Ku70 /80 recruits WRN to the ends of DNA we wished to determine whether WRN could bind to a DNA-PKcs:Ku70 /80 complex on DNA. For this purpose, full-length WRN and a series of WRN deletion mutants were tested in gel shift assays in combination with Ku70 /80 and DNA-PKcs. As shown in Figure , neither the wild-type nor any of the WRN mutants used in this experiment formed a stable complex with DNA (Fig. A, lanes 2 --5, Fig. B, lanes 1 --4). As expected, the addition of Ku 70 /80 to DNA leads to the formation of two retarded DNA bands (Fig. A, lane 6, Fig. B, lane 5). A third retarded DNA band is formed upon the addition of WRN or WRNDelta(773 --1306) (Fig. A, lanes 9 --10, Fig. B, lane 9), which corresponds to a Ku70 /80:WRN:DNA complex . DNA-PKcs does not bind to DNA by itself (Fig. A, lane 11, Fig. B, lane 10), but it forms a distinct high molecular weight complex in the presence of Ku70 /80 (Fig. A, lane 12, Fig. B, lane 11). Interestingly, the addition of stoichiometric amounts of wild-type WRN or WRNDelta(773 --1306) to a preformed Ku70 /80:DNA-PKcs complex resulted in the disappearance of the slower migrating band (Ku70 /80:DNA-PKcs:DNA) and the formation of a new complex with the same electrophoretic mobility as the Ku70 /80:WRN:DNA complex (Fig. A, lanes 15 and 16). This process was strictly dependent on the presence of a WRN protein having both Ku-binding domains, since none of the WRN mutants missing either the N-terminus [WRN(50 --1432), Fig. B, lane 12], the C-terminus [WRN(1 --388), Fig. A, lane 13; WRN(1 --749), Fig. A, lane 14; WRN(1 --1308), Fig. B, lane 14] or both domains [WRN(50 --1432), Fig. B, lane 13] displaced DNA-PKcs from the Ku70 /80:DNA complex. These results suggest that the WRN:DNA-PKcs:Ku complex does not form on DNA. Furthermore, they indicate that the Ku interaction domains located in the N- and C-terminal regions of WRN are required to dissociate DNA-PKcs from DNA. To provide further evidence that WRN removes DNA-PKcs from a Ku70 /80:DNA complex, we performed protein binding assays on immobilized DNA templates. Ku70 /80 and DNA-PKcs were loaded onto biotinylated DNA that was immobilized on streptavidin beads and then incubated with either full-length WRN or a mutant WRN missing the C-terminal region [WRN(1 --748)]. The reaction mixture was incubated for 30 min at 4C and the unbound material was removed by extensive washes. Then the bound proteins were released from the beads and analyzed by immunoblotting with DNA-PKcs antibody. As shown in Figure C, the addition of wild-type WRN (lane 1) led to the disappearance of DNA-PKcs from the DNA-bound material, indicating that WRN displaces DNA-PKcs from the Ku70 /80:DNA complex. Moreover, the inability of the mutant WRN(1 --749) (lane 2) to perform this function demonstrates that both Ku-binding domains of WRN are required in the process. DNA-PKcs does not protect DNA from nucleolytic processing by WRN | It has been proposed that the binding of DNA-PKcs to DNA ends might protect DNA from degradation by cellular nucleases. The finding that DNA-PKcs is displaced from the DNA ends by WRN suggests that WRN exonuclease activity may not be influenced by the presence of DNA-PKcs. To test this hypothesis, we performed a WRN exonuclease assay in the presence of purified Ku70 /80 and DNA-PKcs . As shown previously , WRN nucleolytic activity was strongly stimulated by Ku70 /80 (Fig. A, compare lanes 3 and 5). In the absence of Ku70 /80, DNA-PKcs did not influence WRN exonuclease activity (Fig. A, lane 4), either in the presence or absence of ATP (Fig. B, compare lanes 1 and 2). Likewise, the strong stimulation of WRN exonuclease activity by Ku70 /80 is not affected by the addition of DNA-PKcs (Fig. A, lanes 6 --9), also in the presence of ATP (Fig. B, lane 6). Taken together, these results indicate that DNA-PKcs does not protect the DNA termini from nucleolytic degradation by the WRN exonuclease. Figure 1 | WRN interacts with DNA-PKcs in the presence of Ku. WRN interacts with DNA-PKcs in the presence of Ku. (A) WRN-binding proteins from HeLa nuclear extracts. Flag --WRN (lane 3) and Flag --HCV polymerase (HCV-pol, lane 2) were immobilized on anti-Flag beads and incubated with nuclear extracts from HeLa cells. After extensive washes, the bound proteins were eluted from the beads with BCO buffer (1.0 M KCl, 10 mM Tris --HCl, pH 7.5, 1 mM EDTA, 5% glycerol). The eluted proteins were separated on an 8% SDS --polyacrylamide gel and analyzed by silver staining. The arrow indicates high molecular weight protein. (B) Identification of the >350 kDa polypeptide by western blot analysis. The proteins eluted from the anti-Flag resins were resolved by SDS --PAGE and transferred to nitrocellulose. Western blot analysis was performed using antibodies against DNA-PKcs. The arrow indicates DNA-PKcs. (C) Purified DNA-PKcs phosphorylates p53. Bacterially expressed GST --p53 was bound to glutathione beads and then the beads were incubated with buffer alone (lane 1), nuclear extract from HeLa cells (lane 2) or purified DNA-PKcs (lane 3) in the presence of DNA. The beads were then washed and the bound phosphorylated proteins were analyzed by 8% SDS --PAGE followed by autoradiography (lane 1, 200 microl of kinase reaction buffer; lane 2, 100 microg nuclear extract from HeLa cells; lane 3, 200 pmol purified DNA-PKcs). The arrows indicate GST --p53. (D) Ku70 /80 mediates the interaction between WRN and DNA-PKcs. Equal amounts of purified DNA-PKcs (lanes 1 and 3) and Ku70 /80:DNA-PKcs (lanes 2 and 4) were incubated with Flag --WRN immobilized on anti-Flag beads (lanes 3 and 4) and the bound proteins were then eluted and analyzed by SDS --PAGE and silver staining. The arrow indicates DNA-PKcs. (E) Reactions were performed as in (D) and the resin-bound proteins were detected by immunoblotting with anti-DNA-PKcs (top) and anti-WRN (bottom) antibody. Lane 1, Flag --WRN + Ku70 /80 + DNA-PKcs; lane 2, Flag --WRN + DNA-PKcs; lane 3, purified DNA-PKcs (top) or WRN (bottom). Figure 2 | The N- and C-terminal regions of WRN are required for the formation of a stable complex with Ku70 /80 on DNA. The N- and C-terminal regions of WRN are required for the formation of a stable complex with Ku70 /80 on DNA. Radiolabeled 40 /46mer DNA and purified Ku70 /80 and WRN or WRN fragments were first incubated at room temperature for 10 min. The reactions were analyzed by 4% native PAGE and the DNA --protein complexes were visualized by autoradiography. Femtomoles of Ku70 /80 and WRN used in each reaction are indicated at the top of the gel. Lane 1, 200 fmol Ku70 /80; lane 2, 200 fmol Ku70 /80 and 200 fmol GST --WRN(1 --50); lane 3, 200 fmol Ku70 /80 and 300 fmol GST --WRN(1 --50); lane 4, 200 fmol Ku70 /80 and 200 fmol Flag --WRN(1 --388); lane 5, 200 fmol Ku70 /80 and 300 fmol Flag --WRN(1 --388); lane 6, 200 fmol Ku70 /80 and 200 fmol Flag --WRN(1 --749); lane 7, 200 fmol Ku70 /80 and 300 fmol Flag --WRN(1 --749); lane 8, 200 fmol Ku70 /80 and 200 fmol Flag --WRNDelta(773 --1306); lane 9, 200 fmol Ku70 /80 and 300 fmol Flag --WRNDelta(773 --1306); lane 10, 200 fmol Ku70 /80 and 200 fmol Flag --WRN; lane 11, 200 fmol Ku70 /80 and 300 fmol Flag --WRN; lane 12, 200 fmol Ku70 /80 and 200 fmol GST --WRN(1306 --1432); lane 13, 200 fmol Ku70 /80 and 300 fmol GST --WRN(1306 --1432); lane 14, DNA probe only. Figure 3 | The C-terminal region of WRN binds to Ku70 /80. The C-terminal region of WRN binds to Ku70 /80. (A) The Coomassie stained gel shows an aliquot of the GST fusion proteins that were used in the protein interaction studies. Lane 1, GST; lane 2, GST -- WRN(1 --50); lane 3, GST --WRN(50 --749); lane 4, GST --WRN(1306 --1432); lane 5, molecular markers. The arrows indicate the positions of the GST fusion proteins on the stained gel. (B) A sample of 4 microg of each GST fusion protein was immobilized on glutathione beads and incubated with 500 microl of cell lysates from Sf9 cells expressing Ku70 (top), Ku80 (middle) or Ku70 /80 (bottom). After extensive washes, bound proteins were analyzed by western blotting with anti-Ku70 (top), anti-Ku80 (middle) and a mixture of anti-Ku70 and anti-Ku80 (bottom) antibodies. The input lane (lane 1) shows 5% of Ku70 (top), Ku80 (middle) and Ku70 /80 (bottom) used in each pull-down assay. Lane 1, input protein extract; lane 2, GST; lane 3, GST --WRN(1 --50); lane 4, GST --WRN(50 --749); lane 5, GST --WRN(1306 -- 1432); lane 6, purified factor (Ku70, top; Ku80, middle; Ku70 /80, bottom). (C) Flag-tagged WRN(50 --1432) (lane 2), WRN(50 --1308) (lane 3), WRN(1 --1308) (lane 4) and full-length WRN (lane 5) were immobilized on beads and incubated with extracts from Sf9 cells overexpressing Ku70 /80. After extensive washes, the bound proteins were resolved by SDS --PAGE, transferred to nitrocellulose and analyzed with a mixture of anti-Ku70 and anti-Ku80 antibodies (top) and WRN antibody (bottom). Lane 1 represents 5% of the Ku70 /80 extract used in each reaction. Figure 4 | WRN displaces DNA-PKcs from a DNA:Ku70 /80 complex. WRN displaces DNA-PKcs from a DNA:Ku70 /80 complex. (A) Gel retardation assays were conducted with a radiolabeled 40 /46mer DNA substrate and purified Ku70 /80, DNA-PKcs and wild-type or mutant WRN. The DNA substrate was first incubated with 200 fmol Ku70 /80 and DNA-PKcs at room temperature for 10 min, then 200 fmol wild-type or mutant WRN was added to the reactions and the incubation was continued for an additional 5 min at room temperature. The reaction mixture was subjected to 4% native polyacrylamide gel electrophoresis at 8C. Shifted bands were visualized by autoradiography. Lane 1, DNA only; lane 2, WRN(1 --388); lane 3, WRN(1 --749); lane 4, WRNDelta(773 --1306); lane 5, WRN; lane 6, Ku70 /80; lane 7, Ku70 /80 and WRN(1 --388); lane 8, Ku70 /80 and WRN(1 --749); lane 9, Ku70 /80 and WRNDelta(773 --1306); lane 10, Ku70 /80 and WRN; lane 11, DNA-PKcs; lane 12, Ku70 /80 and DNA-PKcs; lane 13, Ku70 /80, DNA-PKcs and WRN(1 --388); lane 14, Ku70 /80, DNA-PKcs and WRN(1 --749); lane 15, Ku70 /80, DNA-PKcs and WRNDelta(773 --1306); lane 16, Ku70 /80, DNA-PKcs and WRN. (B) Gel retardation assays were performed as described in (A). Lane 1, WRN(50 --1432); lane 2, WRN(50 --1308); lane 3, WRN(1 --1308); lane 4, WRN; lane 5, Ku70 /80; lane 6, Ku70 /80 and WRN(50 --1432); lane 7, Ku70 /80 and WRN(50 --1308); lane 8, Ku70 /80 and WRN(1 --1308), lane 9, Ku70 /80 and WRN; lane 10, DNA-PKcs, lane 11, Ku70 /80 and DNA-PKcs; lane 12, Ku70 /80, DNA-PKcs and WRN(50 --1432); lane 13, Ku70 /80, DNA-PKcs and WRN(50 --1308); lane 14, Ku70 /80, DNA-PKcs and WRN(1 --1308); lane 15, Ku70 /80, DNA-PKcs and WRN; lane 16, DNA alone. (C) WRN removes DNA-PKcs from a Ku70 /80:DNA complex. (Left) Silver stained gel showing the proteins used in the assay. Lane 1, WRN; lane 2, WRN(1 --749); lane 3, Ku70/80 + DNA-PKcs. (Right) Ku70 /80 and DNA-PKcs were bound to a 5' end-biotinylated 39 /39 double-stranded oligomer immobilized on streptavidin beads and then incubated with 150 microl of a solution containing WRN (lane 1) or mutant WRN(1 --749) (lane 2). The mixture was incubated at 4C for 30 min, after which the beads were washed extensively and the bound proteins were analyzed by immunoblotting using DNA-PKcs antibody. Lane 3 shows purified DNA-PKcs as a positive control. Figure 5 | DNA-PKcs does not protect DNA from the exonuclease activity of WRN. DNA-PKcs does not protect DNA from the exonuclease activity of WRN. (A) Ku70 /80 (100 fmol) and DNA-PKcs (100 --400 fmol) were incubated with radiolabeled 30 /46mer DNA substrate at room temperature for 10 min. WRN (100 fmol) was added to this reaction and incubated for an additional 10 min. The reaction products were analyzed by 14% polyacrylamide --urea denaturing gel electrophoresis and autoradiography. Femtomoles of Ku70 /80, DNA-PKcs and WRN are indicated at the top of the gel. Lane 1, 100 fmol DNA-PKcs; lane 2, 100 fmol Ku70 /80; lane 3, 100 fmol WRN; lane 4, 100 fmol WRN and DNA-PKcs; lane 5, 100 fmol WRN and Ku70 /80; lane 6, 100 fmol WRN and Ku70 /80:DNA-PKcs; lane 7, 100 fmol WRN and Ku70 /80 and 200 fmol DNA-PKcs; lane 8, 100 fmol WRN and Ku70 /80 and 300 fmol DNA-PKcs; lane 9, 100 fmol WRN and Ku70 /80 and 400 fmol DNA-PKcs; lane 10, probe only. (B) Exonuclease activity of WRN:DNA-PK in the presence or absence of ATP. Reactions were performed as described in (A) and incubated at room temperature for an additional 15 min after the addition of WRN. Femtomoles of Ku70 /80, DNA-PKcs and WRN are indicated at the top of the gel. Lane 1, 100 fmol WRN and DNA-PKcs in the presence of ATP; lane 2, 100 fmol WRN and DNA-PKcs in the absence of ATP; lane 3, 100 fmol WRN and Ku70 /80 in the presence of ATP; lane 4, 100 fmol WRN and Ku70 /80 in the absence of ATP; lane 5, 100 fmol WRN and Ku70 /80:DNA-PKcs in the presence of ATP; lane 6, 100 fmol WRN and Ku70 /80:DNA-PKcs in the absence of ATP; lane 7, probe only. DISCUSSION : In this study we have examined the relationship between WRN and two components of the NHEJ repair pathway, Ku and DNA-PKcs. Our analysis indicates that in the absence of DNA WRN assembles into a triple complex with Ku and DNA-PKcs. The association between WRN and DNA-PKcs requires the presence of Ku, suggesting that the binding of WRN to DNA-PKcs is probably not direct, but rather is mediated by the Ku70 /80 heterodimer. To determine whether DNA influences the molecular interactions between these factors we performed gel shift assays with purified components. The results of these experiments indicate that a WRN:Ku:DNA-PKcs complex does not form on DNA and equimolar amounts of WRN can displace DNA-PKcs from a preformed complex with Ku70 /80 on DNA. By titrating increasing amounts of DNA-PKcs we have observed the formation of distinct WRN:Ku70 /80:DNA and DNA-PKcs:Ku70 /80:DNA complexes in the same reaction mixture, but we have not detected a supershifted band suggestive of a WRN:Ku70 /80:DNA-PKcs:DNA complex (data not shown). Importantly, our data show that the displacement of DNA-PKcs from DNA by WRN requires both the N- and the C-terminal domains of WRN, each of which makes direct contacts with Ku70 /80. The finding that the WRN C-terminus, in addition to the N-terminus, binds to Ku70 /80 is in agreement with the data from another laboratory, which identified the Ku heterodimer as a WRN-interacting factor through an affinity screen that used the C-terminal region of WRN as bait . Our analysis further indicates that while the N- and C-terminal regions of WRN can interact with Ku in solution independently of each other, both domains are required for the formation of a stable WRN:Ku complex on DNA and for displacing DNA-PKcs from the Ku:DNA complex. It is possible that the displacement of DNA-PKcs from the Ku70 /80:DNA complex may be facilitated by specific conformational changes in DNA-PKcs that occur upon recruitment of this factor to the DNA ends and activation of its kinase activity. This work and two other recent studies on the interaction between DNA-PKcs and WRN differ in some of their conclusions (,). One study suggests that the interaction between WRN and DNA-PKcs is direct , while the other indicates, as we have shown in this report, that Ku70 /80 is required for the binding of WRN to DNA-PKcs . Furthermore, both papers suggest that WRN and DNA-PKcs can concurrently form a complex with Ku70 /80 on DNA. However, the functional significance of these molecular interactions is not consistent between the two studies: while the results from one group indicate that DNA-PKcs represses WRN exonuclease activity only in the presence of Ku70 /80 , the other shows that WRN exonuclease activity is inhibited only in the absence of Ku70 /80 . We cannot provide a likely explanation for these differences, but it is possible that the quality of protein preparations and /or minor variation in the experimental procedures may be responsible for these effects. Using different preparations of highly active, Ku70 /80-free DNA-PKcs, we have not detected either a direct interaction between WRN and DNA-PKcs or a significant inhibition of WRN exonuclease activity, even in the presence of excess DNA-PKcs. It is possible that the WRN:DNA-PKcs:Ku70 /80 complex shown in the other reports may be due to the presence of glutaraldehyde in the gel shift assays, since most of these assays were performed with a reaction mixture treated with the cross-linking reagent prior to electrophoresis (,). We have performed gel shift assays in the presence of glutaraldehyede (data not shown), however, the results of these experiments were not highly reproducible and therefore not easily interpretable. Our data suggest one of two models for how WRN could function to maintain genetic stability. First, since one of the known functions of Ku and DNA-PKcs is in the repair of DNA damage by NHEJ, WRN could be directly participating in the NHEJ repair process by releasing DNA-PKcs from the DNA ends prior to the ligation step. Indeed, several studies have suggested that while the Ku70 /80 heterodimer translocates inward, DNA-PKcs remains bound to the DNA ends (,,). The subsequent steps in the NHEJ repair pathway require recruitment of the XRCC4:ligase IV complex, which catalyzes rejoining of the broken DNA ends. It is likely that in order to allow proper DNA ligation, DNA-PKcs needs to be released from the DNA ends. Our results suggest that WRN may be part of this release mechanism during progression of the repair process. As an alternative model, it is possible that by preventing the stable association of DNA-PKcs with DNA-bound Ku, WRN inhibits the functional assembly of the NHEJ repair complex and contributes to the implementation of an alternative pathway. Such a pathway could be triggered by DNA damage caused by genotoxins such as 4 nitroquinoline-1-oxide and camptothecin, two chemical agents to which WRN cells are extremely sensitive, and may require the coordinate action of Ku and WRN but not DNA-PKcs. Thus, WRN and Ku may function in a novel pathway to maintain genomic stability. Backmatter: PMID- 12202765 TI - Drosophila damage-specific DNA-binding protein 1 (D-DDB1) is controlled by the DRE /DREF system AB - We succeeded in cloning the gene, termed d-ddb1, for a Drosophila homolog of the p127 subunit of the human damage-specific DNA-binding protein, thought to recognize (6 --4) photoproducts and related structures. In Drosophila, the gene product (D-DDB1) also appeared to play a role as a repair factor, d-ddb1 knockout Kc cells generated with a RNAi method being sensitive to UV. In addition, UV or methyl methanesulfonate treatment increased d-ddb1 transcripts. However, we found that the gene is controlled by the DRE /DREF system, which is generally responsible for activating the promoters of proliferation-related genes. Moreover, during Drosophila development, the transcription of d-ddb1 changed greatly, with the highest levels in unfertilized eggs, indicating that external injury to DNA is not essential to D-DDB1 function. Interestingly, as with UV irradiation-induced transfer of D-DDB1 to the nucleus from the cytoplasm, during spermatogenesis the protein transiently shifted from one cell compartment to the other. The results indicate that D-DDB1 not only contributes to the DNA repair system, but also has a role in cell proliferation and development. Keywords: INTRODUCTION : The purpose of the present study was to cast light on DNA damage-specific DNA-binding proteins in Drosophila melanogaster. In man a number of such gene products have been identified, including DDB (or XPE), a heterodimeric protein composed of 127 and 48 kDa subunits, which has been shown to be essential for DNA excision repair . DDB reportedly recognizes many types of DNA lesions ( --) and is inducible by treatment with DNA-damaging agents (,,). However, DDB was found not to be required in nucleotide excision repair (NER) reconstitution studies in vitro ( --). The xeroderma pigmentosum group E (XPE) gene product, homologous to DDB, is involved in NER, and the damaged DNA-binding activity of DDB is absent from cells of a subset of XPE patients ( --). Recently, functions of DDB other than directly in DNA repair have been suggested . DNA repair is complex, with multiple overlapping and intersecting pathways with contributions of replication, transcription, meiotic recombination and gene silencing. A large fraction of mutations isolated on the basis of meiotic recombination defects or embryo developmental anomalies are conferred by and predispose to mutagen hypersensitivity. To generate a deeper comprehension of the nature of DDB, we have chosen D.melanogaster as a suitable model animal. In Drosophila many genes associated with DNA metabolism and morphogenesis are well characterized and mutants can be easily prepared. This factor, coupled with a refined system for genetic analysis, provides a valuable research resource. In this report, we document characterization of Drosophila DDB1, the larger subunit of DDB, and details of its relation to repair, cell proliferation and development, including spermatogenesis. Interestingly, the gene was found to be controlled by the DRE /DREF system, which is responsible for activating the promoters of nucleus encoded genes for proliferating cell nuclear antigen (PCNA), the 180 and 73 kDa subunits of DNA polymerase alpha and cyclin A, among others. Our results provide evidence that Drosophila DDB1 acts as a cell proliferation- or development-associated factor as well as a repair factor. MATERIALS AND METHODS : Separation of Drosophila UV-damaged site-binding proteins | Aliquots of 10 g of Drosophila Kc cells were homogenized in 4 vol of buffer containing 10 mM Tris --HCl, pH 8, 1 mM EDTA and 5 mM dithiothreitol, and then 40 ml of ice-cold buffer (containing 50 mM Tris --HCl, pH 8, 10 mM MgCl2, 2 mM dithiothreitol, 25% sucrose and 50% glycerol) and thereafter neutralized, saturated ammonium sulfate solution (10 ml) was added slowly to the homogenized suspension. After gentle stirring for another 30 min on ice, the homogenate was centrifuged at 50 000 r.p.m. (optima L-70k/70Ti, Beckman) for 3 h at 4C. The supernatant was dialyzed against the same buffer (TNMD) as that used for UV cross-linking analysis (containing 20 mM Tris --HCl, pH 7.5, 200 mM NaCl, 5 mM MgCl2, 0.5 mM dithiothreitol and three protein inhibitors, 1 microg /ml pepstatin A, 1 microg /ml leupeptin and 1 mM phenylmethylsulfonyl fluoride). The dialysate was loaded onto a UV-irradiated single-strand DNA --cellulose column (2.5 x 5.0 cm phi) equilibrated with TNMD buffer. After washing the column with 150 ml of 200 mM NaCl in TNMD buffer, elution was performed with a three-step gradient, using 50 ml each of 0.5, 1 and 2 M NaCl in the same buffer. Aliquots of 4 ml of each fraction were tested for UV-damaged site-binding activity. Determination of UV-damaged site-binding polypeptide size by UV cross-linking analysis | UV cross-linking analysis was carried out as described earlier with modifications. Aliquots of 30 ng of oligonucleotide TC31-3' (5'-AAGCTTTATGCCTGCATCATC) and 15 ng of oligonucleotide TC31 (5'-AATTCGAGCTCGTACGATGA CGATGATGCATCATCGGTCATCGTACGATGACGATGATGCAGGCATAAAGCTT) were mixed in 41.5 microl of a solution containing 33 mM Tris acetate, pH 7.9, 10 mM magnesium acetate, 66 mM potassium acetate and 0.5 mM dithiothreitol and incubated for 3 min at 95C, followed by 10 min at 25C. Then the solution was mixed with 8.5 microl of reaction mixture containing 118 microM each dATP, dGTP and dCTP, 1850 KBq [alpha-32P]dATP and 4 U of Escherichia coli DNA polymerase I large fragment. DNA was uniformly labeled with 32P by incubation at 37C for 1 h and then chased for 15 min with 10 microM unlabeled dCTP. The 32P-labeled probe (1.75 ng) was incubated with Kc cell nuclear extract (10 microg protein) or 5 microl of each fraction separated on a UV-irradiated single-strand DNA --cellulose column for 15 min on ice, in 17 microl of buffer containing 20 mM Tris --HCl, pH 7.5, 200 mM NaCl, 5 mM MgCl2, 0.5 mM dithiothreitol, 0.2 microg poly(dI --dC). Uncapped Eppendorf tubes containing reaction mixtures were placed at 8 cm distance from a 254 nm UV transilluminator (model UVGL-58; UVP Inc.) and irradiated on ice for 20 min. UV dose under these conditions was 4.19 kJ /m2. Solutions of CaCl2 and MgCl2 were added to final concentrations of 10 and 100 mM, respectively. Digestion by 2 U DNase I and 9 U exonuclease III was carried out at 30C for 30 min. The reaction was terminated by adding 20 microl of loading buffer containing 100 mM Tris --HCl, pH 6.8, 4% SDS, 0.2% bromophenol blue, 20% (v /v) glycerol and 0.2 M dithiothreitol. The samples were heated and applied to 10% polyacrylamide gels containing 0.1% SDS. After electrophoresis, the gels were stained with Coomassie Brilliant Blue, photographed, dried and autoradiographed. The molecular weights of the protein bands were estimated by comparing their mobilities with those of marker proteins (Bio-Rad). Determination of UV-damaged site-binding polypeptides by electrophoretic mobility shift assay using native gels | The methods used were basically the same as described in an earlier study by Todo et al. . The binding reaction mixture was the same as for the UV cross-linking analysis. The mixture was incubated for 20 min on ice. It was then electrophoresed in a 6% polyacrylamide gel in a buffer containing 6.7 mM Tris --borate pH 7.5, and 1 mM EDTA (TBE buffer) at 50 V for 3 --4 h, after which it was dried and then used to expose Fuji RX medical X-ray film. The gels were also exposed to X-ray films for 6 h. Production of recombinant D-DDB1 | The D-DDB1 fragment (amino acids 592 --1140) was expressed as an N-terminal His-tagged fusion protein using the pET-28b (Invitrogen) vector in E.coli BLR21 (DE3). The recombinant fragments were purified by chromatography through His-bind resin. Purified recombinant D-DDB1 (10 microg) fragment was separated on a 12.5% SDS --polyacrylamide gel and the recombinant fragment was visualized using Coomassie Brilliant Blue staining. Analysis of DNA sequences | DNA sequencing was performed by the dideoxy chain termination method with a sequencing kit and DNA sequencer from Applied Biosystems. DNA sequence analysis was performed using Gentetix Mac v.10 (Software Development Co. Ltd). Production of polyclonal antibody and western blot analysis | The purified recombinant D-DDB1 fragment was used for immunization of a mouse. Polyclonal antibodies reacting with D-DDB1 were affinity purified by D-DDB1 protein- conjugated Sepharose column chromatography. Western blotting analysis was carried out by the method of Towbin et al. . Anti-mouse IgG conjugated with alkaline phosphatase (Promega) was used as the secondary antibody. Color was developed with nitroblue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate as the substrates of alkaline phosphatase. Northern hybridization | Aliquots of 30 microg of total RNA extracted from D.melanogaster bodies at several different stages and from Kc cells, either incubated for 0 --10 h after irradiation with 70 J /m2 UV or for 0 --7 h after addition of 0.1% methyl methanesulfonate (MMS), were resolved on 1.2% formaldehyde --agarose gels and transferred to nylon membranes (Hybond-N+; Amersham). After prehybridization, the filters were probed with 32P-labeled d-ddb1 cDNA (nucleotides 1774 --3423) or with 32P-labeled full-length ribosomal protein 49 (Rp-49) cDNA as a control at 42C for 16 h, followed by washing twice with 2x SSPE + 1% SDS at room temperature for 15 min and twice with 1x SSPE + 0.1% SDS at 50C for 20 min. Blots were exposed to Kodak X-Omat XAR films and quantified with a NIH imaging analyzer. Double-stranded RNA production | DNA fragments 600 bp in length (nucleotides 1 --600), containing the coding sequences for D-DDB1 to be 'knocked out' were amplified using PCR. Each primer contained a T7 RNA polymerase-binding site (GAATTAATACGACTC ACTATAGGGAGA) followed by sequences specific for the targeted genes. The PCR products were purified using a High Pure PCR purification kit (Roche Molecular Biochemicals) and employed as templates with a Megascript T7 transcription kit (Ambion, Austin, TX) to produce double-stranded (ds)RNA. The dsRNA products were ethanol precipitated and resuspended in water and annealed by incubation at 65C for 30 min followed by slow cooling to room temperature. Aliquots of 6 microg were analyzed by 1% agarose gel electrophoresis to ensure that the majority of the dsRNA existed as a single band of similar600 bp. The remainder was stored at --20C. Conditions for RNA interference (RNAi) in Drosophila cell culture | RNAi was carried out by the method of Clemens et al. . Drosophila Kc cells were diluted to a final concentration of 1 x 106 cells /ml in serum-free M3 medium (Sigma) and 1 ml aliquots were plated per well of 6-well cell culture dishes (Corning). dsRNA was added directly to the medium to a final concentration of 2.5 or 5 microg /ml. For dsRNAs of similar600 nt, this corresponds to 1 mg dsRNA. This was followed immediately by vigorous agitation. The cells were incubated for 30 min at room temperature followed by addition of 2 ml of M3 medium containing fetal calf serum. The cells were incubated for an additional 3 days to allow for turnover of the target protein. Immunostaining of Kc cells | Kc cultured cells derived from D.melanogaster embryos were grown at 25C in M3 medium supplemented with 2% fetal calf serum. They were incubated for 1 h with anti-D-DDB1 IgG, diluted 1:50 in phosphate-buffered saline (PBS) containing 1% BSA, then treated for 1 h with Alexa594 anti-mouse IgG (Sigma), diluted to 1:100, as secondary antibody. The cells were also stained with a solution of 20 microg /l DAPI for 5 min, examined under a fluorescence microscope (Olympus BX-50) (x1600) and photographed using Tri X Pan 400 film (Kodak). Monoclonal antibodies | Monoclonal antibodies to DREF, Mab-1 and Mab-4, were raised as described previously . Electrophoretic mobility shift assays and preparation of nuclear extracts | The sequences of double-stranded oligonucleotides containing DREF-binding sites of the PCNA gene promoter (DRE-P) and of their derivatives in the d-ddb1 promoter were as follows. DRE-P, gatccCTGCCTGCTATCGATAGATTCAGGa GGACGGACGATAGCTATCTAAGTCCtctag DRE wt, ACACAACACATATCGATATTCGGTTTTT TGTGTTGTGTATAGCTATAAGCCAAAAA DRE mutD, GGGCAACACAACACAgcTCGATATTCGGTTTTTTC CCCGTTGTGTTGTGTcgAGCTATAAGCCAAAAAAG DRE del3, GGGCAACACAACACATAT---TATTCGGTTTTTTC CCCGTTGTGTTGTGTATA---ATAAGCCAAAAAAG DRElike-wt, AAAGAAACGAAATCGATTTTTCTTGCC TTTCTTTGCTTTAGCTAAAAAGAACGG DRElike-mutD, AAAGAAACGAgcTCGATTTTTCTTGCC TTTCTTTGCTcgAGCTAAAAAGAACGG DRElike-del3, AAAGAAACGAAAT---TTTTTCTTGCC TTTCTTTGCTTTA---AAAAAGAACGG Nucleotides deleted from or inserted into the wild-type sequence are shown by lower-case letters with underlining. 32P-labeled probes (10 000 c.p.m.) were incubated in 16 microl of reaction mixture containing 25 mM HEPES pH 7.9, 100 mM NaCl, 1 mM EDTA, 10% (v /v) glycerol, 0.1% Tween-80, 0.02% 2-mercaptoethanol, 0.5 microg poly(dI --dC) and 0.5 microg salmon sperm DNA on ice for 5 min. When necessary, double-stranded oligonucleotides were added as competitors at this step. Then, Kc cell nuclear extract was introduced and the reaction mixture was incubated for 20 min on ice. In experiments with antibodies, Kc cell nuclear extracts were preincubated with these for 1 h on ice. DNA --protein complexes were electrophoretically resolved on 4% polyacrylamide gels in 100 mM Tris --borate buffer, pH 8.3, 2 mM EDTA containing 2.5% (v /v) glycerol at 25C. The gels were dried and then autoradiographed. DNA transfection and luciferase assay | Drosophila S2 cells (2 x 106 per dish) were grown in 60-mm plastic dishes for 24 h and transfected with 500 ng reporter plasmid DNA and 1 ng pRL-actin5C DNA using CellFectin reagent (Gibco BRL). Cells were harvested 48 h thereafter. The luciferase assay was carried out by means of a Dual-Luciferase Reporter Assay System (Promega). Firefly luciferase activity was normalized to Renilla luciferase activity. To construct the plasmids p5'-510DDBluc(wt), p5'-510DDBwt-mutDluc(wt-mutD), p5'-510DDBwt-del3luc(wt-del3), p5'-510DDBmutD-wtluc(mutD-wt), p5'-510DDBdel3-wtluc(del3-wt), p5'-510DDBmutD-mutDluc (mutD-mutD), p5'-510DDBmutD-del3luc(mutD-del3), p5'-510DDBdel3-mutDluc(del3-mutD), p5'-510DDBdel3-del3luc(del3-del3) and p5'-210DDBluc(-DRE) for the luciferase transient expression assay, a DNA fragment containing the d-ddb1 gene fragment was isolated and inserted into plasmid PGVB. The mutation was introduced using Mutan(R)-Super ExpressTM (Takara). Plasmid pRL-actin5C containing the actin 5C gene promoter was used as an internal control in the luciferase transient expression assay. Fixation and immunocytochemical staining of larval tissues | The imaginal discs were dissected from larvae, then fixed with Tris-buffered saline (TBS) solution (50 mM Tris --HCl, pH 8.3, and 150 mM NaCl) containing 3.5% formaldehyde. Tissues were blocked with TBS solution containing 10% goat serum and 0.15% Triton X-100 (TBT). Incubation with primary antibodies was carried out in the same solution for 16 h at 4C. Samples were washed extensively in TBS containing 0.3% Triton X-100, reblocked with TBT and incubated with alkaline phosphatase-conjugated secondary antibody for western immunoblotting analysis. Stained samples were mounted in a solution containing glycerol. Samples were examined under a microscope (Olympus BX-50) and photographed using Tri X Pan 100 film (Kodak). In situ immunostaining of Drosophila adult bodies | Paraffin-embedded, paraformaldehyde-fixed tissue specimens were used for in situ immunostaining to detect D-DDB1 proteins as described . Samples were examined under a microscope (Olympus BX-50) and photographed using Tri X Pan 100 film (Kodak). Fixation and staining of testes | Adult testes were dissected and fixed according to Pisano et al. . Fixed preparations were washed twice (10 min each) in PBS, 0.1% Tween-20, and once for 5 min in PBS, before incubation with the antibody. The purified recombinant D-RPA30 protein expressed as an N-terminal His-tagged fusion protein using the pET-28b vector (Invitrogen) in E.coli BLR21 (DE3) was employed for immunization of a rabbit. Polyclonal antibodies reacting with D-RPA30 were affinity purified on a D-RPA30 protein-conjugated Sepharose column. Slides were incubated for 1 h with 40 microl of anti D-DDB1 and anti-RPA30 antibodies diluted 1:100 in PBS in a humid chamber at room temperature. They were then washed twice in PBS (5 min each) and incubated for 1 h with the secondary antibody (Alexa488 anti-mouse IgG or Alexa594 anti-rabbit IgG) diluted 1:100 in PBS. After two washes in PBS (5 min each) the immunostained slides were stained with a solution of 20 microg /l DAPI for 5 min, examined under a microscope (Olympus BX-50) and photographed using Tri X Pan 400 film (Kodak). RESULTS : Characterization of UV-damaged DNA-binding proteins in Kc cells in D.melanogaster | To screen for UV-damaged DNA-binding proteins that might play a role in DNA repair-specific events, crude extracts were generated from cultured D.melanogaster Kc cells and first characterized by UV-damaged DNA --cellulose column chromatography. Then they were subjected to a DNA cross-linking assay for UV-damaged DNA-binding protein activity. Figure A shows UV-damaged DNA --cellulose elution patterns of proteins in D.melanogaster Kc cells. The components were separated into at least three peaks in the column at 0.5, 1 and 2 M NaCl. The major fraction eluted at 0.5 M NaCl appeared to be a CPD or (6 --4) photolyase with a mol. wt of similar60 kDa, as reported by Todo et al. (lanes 8 --11), and the 30 kDa protein eluted at 1 M NaCl might be a UV-damage-specific endonuclease 1 (D-DDB P1), as we reported previously (,) (lane 13). Here, we have concentrated our attention on the 126 kDa protein fraction eluted at 2 M NaCl, because it demonstrated the highest affinity for UV-damaged DNA in D.melanogaster. The 126 kDa protein in fractions 37 --46 of the column was collected, purified by SDS --PAGE and used to raise rabbit antiserum against the polypeptide. The antiserum recognized multiple polypeptides, including the 126 kDa protein, in Kc cell extracts. Expression screening of a Kc cell cDNA library using the antiserum resulted in isolation of six clones containing partial DNA sequences with high homology to human DDB1 (p127) . We then carried out a detailed analysis of the 126 kDa protein homologous to human DDB1, provisionally designated Drosophila damage-specific DNA-binding protein 1 (D-DDB1). The nucleotide sequence data have been lodged in the DDBJ /EMBL /GenBank nucleotide sequence databases with accession no. AF132145. The cDNA of d-ddb1, obtained by RT --PCR using specific primers, was characterized as follows. The open reading frame predicted a protein of 1140 amino acid residues, with a molecular weight of 126 038. Overall similarity of the sequence to the human and rat DDB1 sequences was 49.8 and 49.4%, respectively . Figure B shows a western blot of the gel pattern of Figure A using the anti-DDB1 antibody generated against D-DDB1 recombinant fragment (see left panel). The 126 kDa protein indicated to be D-DDB1, and the molecular mass was 126 kDa (Fig. B, lane 1). Fraction 37 in Figure A was tested by SDS --PAGE and confirmed to have another small subunit of similar50 kDa (Drosophila DDB2) (data not shown). Figure C shows the results of a native gel electrophoretic mobility shift assay using fraction 37 (containing D-DDB1) in Figure A. After UV-irradiation, shifted protein bands occurred in the native gel (Fig. C, lane 8). When 32P-labeled unirradiated TC31 was used, the shifted protein bands did not occur (Fig. C, lanes 1 --6). Shifted bands were detected when 32P-labeled irradiated TC31 was used, which were greatly reduced by addition of increasing amounts of unlabeled irradiated TC31 (Fig. C, lanes 11 and 12). Although two bands were observed, the same result was found for human DDB . Influence of UV-irradiation and MMS on expression of the d-ddb1 gene and results of knocking out the gene by RNAi | We measured levels of mRNA for the d-ddb1 gene in Kc cells after UV-irradiation. Since >90% of (6 --4) photoproducts in UV-damaged nuclear DNA could be repaired within 3 h , and since DDB binds to (6 --4) photoproducts, we checked the amount of mRNA every hour after UV-irradiation at 70 J /m2. The amount of mRNA increased after UV-irradiation, reached a maximum within 4 h, and then gradually reduced, indicating the transcripts to be UV-inducible (Fig. A). Transcripts of d-ddb1 were also induced by 0.1% MMS treatment, although they reached a maximum more quickly (2 h) (Fig. B). Figure illustrates the effects of UV-irradiation on Kc cells in which the d-ddb1 gene was knocked out by the RNAi method using 600 bp dsRNA (Fig. A). As shown in Figure B, the transcripts of d-ddb1 were significantly reduced by the RNAi treatment at 2.5 and 5.0 microg /ml compared with same amount of rp49 transcript. The signals between d-ddb1 and rp49 are non-specific bands, because the subjacent band was only detected in Kc cells (see Fig. , lane 14). In Figure C, data for viability of cells every 6 h after UV-irradiation at 70 J /m2 are presented. The cells were clearly more sensitive to UV after gene knockout. To our knowledge, these results represent the first such description, although there is a report that a mutation in human DDB2 makes the cells more UV-sensitive . We also tested the effects of UV-irradiation on the cellular localization of D-DDB1 protein as a function of time by immunostaining with the anti-D-DDB1 antibody. As shown in Figure , after UV irradiation at 70 J /m2, nuclear accumulation of D-DDB1 protein in Kc cells was found in 8% of all the cells examined after 12 h, although no accumulation was observed in unirradiated cells. The protein seemed to be moved into the nucleus from the cytoplasm. Changes in expression of the d-ddb1 gene during Drosophila development and its regulation by the DRE /DREF system | As described briefly in the Introduction, DDB is not only reported to recognize many types of DNA lesions, but might also have a role in other processes. To investigate its relationship to Drosophila development, northern hybridization experiments were performed using the d-ddb1 cDNA as a probe and total RNA extracted from Drosophila at various developmental stages: unfertilized eggs, 0 --2, 2 --4, 4 --8, 8 --12, 12 --16 and 16 --20 h embryos, first, second and third instar larvae, pupae, adult males and females, and Kc cells . A single hybridization band of similar4 kb was detected throughout all stages, the size corresponding to that of the full-length cDNA. The amount of d-ddb1 mRNA changed considerably during development. Although UV-irradiation was not performed, the highest level of mRNA was detected in unfertilized eggs, then early embryonic stages from 0 to 8 h, adults and Kc cells . Low levels were also detected in the 8 --20 h embryos, larvae and pupae . The expres sion pattern was similar to those for the genes for DNA polymerases alpha and epsilon and PCNA (,). Since this suggested a link with cell proliferation, we searched for transcription factor-binding sequences in the d-ddb1 gene and, interestingly, found a DRE site (5'-TATCGATA) and a DRE-like site (5'-AATCGATT, which matched 6 of the 8 bp DRE consensus sequence) in the upper cascade, respectively located at positions --218 and --264 with respect to the transcription initiation site of the d-ddb1 gene (Fig. A). Therefore, we next tested whether gene expression is regulated by the DRE /DREF system. As shown in Figure , the 32P-labeled DRE-P, DRE-wt or DRE like-wt oligonucleotides shown in Figure B were incubated with or without Kc cell nuclear extract and several competitor complex bands were detected. The 32P-labeled DRE-P probe was used as a positive control and when non-labeled DRE-P, DRE-wt and DRE like-wt were added as competitors the complex band was reduced (Fig. A, lanes 2, 3 and 6). Similarly, 32P-labeled DRE-wt or DRE like-wt probes competed with unlabeled DRE-P, DRE-wt and DRE like-wt (Fig. A, lanes 10, 13, 16, 18, 19 and 24). A probe mutated in the DRE site did not compete or only very weakly with DRE-wt, DRE like-wt and DRE-P (Fig. A, lanes 4, 5, 7, 8, 11, 12, 14, 15 and 20 --23). Next, to confirm that the complex band was the DRE /DREF complex, we examined the effects of anti-DREF monoclonal antibodies (Mab-1 and Mab-4) on the binding reaction with Kc cell nuclear extract. Mab-1 inhibited binding of DREF to DRE-P, and Mab-4 supershifted the band with DRE-P (,) (Fig. B, lanes 1 --4). As shown in Figure B, one of the shifted bands with the DRE-wt and DRE like-wt probes was diminished by addition of Mab-1 (lanes 7 and 10) and was supershifted by Mab-4 (lanes 8 and 11). For further confirmation of the stimulatory effects of the two DREF recognition sequences on the promoter, we prepared luciferase expression constructs having the wild-type DRE-wt and DRE like-wt with several mutations in the DRE and DRE-like sequences ligated with the the luciferase gene in a PicaGene basic vector (Fig. A). The plasmids carrying these constructs were then transfected into Drosophila S2 cells and luciferase expression levels were determined. As shown in Figure B, mutation in any of the two DREF-binding sequences caused a reduction in luciferase expression levels. The two DREF-binding sequences appeared to cooperate to provide high promoter activity. For further investigation, an antibody against D-DDB1 was used for western blot analysis with extracts from Drosophila bodies at various developmental stages. As shown in Figure A, the 126 kDa subunit level decreased throughout embryogenesis; D-DDB1 levels were high at early stages when the rate of DNA replication was maximal and decreased as development proceeded (Fig. A). Although no or a very slight signal for D-DDB1 protein was detected in extracts from larval, pupal and adult whole bodies, probably due to dilution of D-DDB1 with large amounts of other proteins, a specific signal was detected from the brain lobes and imaginal discs dissected from third instar larvae, testes and ovaries (data not shown). Figure B shows imaginal discs dissected from third instar larvae and immunostained with antibody against D-DDB1. Anti-D-DDB1 antibody stained the cells of imaginal discs (A --C). No significant staining signal was observed with normal rabbit IgG as the primary antibody (D --F). In Figure C, D-DDB1 protein is evident in ganglion cells in the brain (A, B), flight muscles (C), testis (D) and ovaries (E). Pursuing the roles of D-DDB1 in relation to development, testis immunostaining was conducted because transcripts of d-ddb1 were found to be abundant in adult male flies (Fig. B, lane 12) and D-DDB1 protein was detected in the testis (Fig. C). Interestingly, D-DDB1 appeared to be transiently present in meiotic cells during spermatogenesis . We applied double staining with D-DDB1 and RPA30, because there is a report that DDB mediates NER activity and because RPA30 is known to be an indispensable factor for replication. Triple labeling (D-DDB1 in green, DNA in blue and RPA30 in red) gave, in merged views: green, D-DDB1; blue, DNA; red, RPA30; yellow, D-DDB1 and RPA30. In young spermatocytes D-DDB1 accumulated in the cytoplasm, although RPA30 was present in the nucleus (Fig. A). In mature primary spermatocytes both D-DDB1 and RPA30 accumulated in the nucleus (Fig. B). At the telophase stage of meioses I and II both D-DDB1 and RPA30 were found in the cytoplasm (Fig. C and D). In early spermatids the typical features included a spherical mitochondrial Nebenkern (NK) lacking both D-DDB1 and RPA30 in the NK, but both accumulated in the nucleus (Fig. E). In elongating spermatids D-DDB1 disappeared from the nucleus, although RPA30 was present (Fig. F). In fully elongated spermatids neither D-DDB1 nor RPA30 was detected. Since at this stage chromatin condensation becomes tighter due to protamine, D-DDB1 and RPA might be removed (Fig. G). Thus, during spermatogenesis, D-DDB1 is first located in the cytoplasm, then accumulates in the nucleus of the mature primary spermatocyte, and finally disappears (Fig. H). In the nucleus positive stages (6 --4) photoproducts were not detected using a specific monoclonal antibody. Figure 1 | (A) UV cross-linking analysis of UV-damaged site-binding polypeptides. (A) UV cross-linking analysis of UV-damaged site-binding polypeptides. Manley's whole cell extract from Kc cells (10 microg protein) was incubated for 15 min with a UV-unirradiated 32P-labeled probe (lanes 1 and 2) or UV-irradiated 32P-labeled probe (lanes 3 and 4) in a solution containing 20 mM Tris, pH 7.5, 200 mM NaCl, 5 mM MgCl2, 0.5 mM dithiothreitol, 5% (v /v) glycerol, 2 microg poly (dI --dC). Similarly, similar930 ng protein from Kc cells, separated on a UV-irradiated single-strand DNA --cellulose column, was incubated for 15 min with a UV-irradiated 32P-labeled probe (lanes 5 --21). After UV irradiation, followed by digestion with DNase I and exonuclease III, the reaction products were electrophoresed on a 12.5% polyacrylamide gel containing 0.1% SDS. Migration positions of marker ptoteins are indicated. (B) SDS --PAGE analysis of the purified recombinant D-DDB1 fragment. A part of D-DDB1 protein (amino acids 592 --1140) was expressed in E.coli BLR21 (DE3) as an N-terminal His-tagged fusion protein obtained using pET-28b (Invitrogen) vector and purified for preparation of an antibody for use in western blotting and immunostaining (left). Western blot analysis revealed that the anti-D-DDB1 antibody specifically recognized the 126 kDa D-DDB1 protein from Kc cells (right panel, lane 1). Lanes 2 --17 contained the same fraction separated with the UV-irradiated single-strand DNA --cellulose column indicated in (A). Fraction 37 contained D-DDB1. The results indicated the UV cross-linking band detected in the 2 M eluted fraction [(A) lanes 18 --21] to be D-DDB1. (C) UV-irradiated DNA-specific binding activity was detected by EMSA. UV-irradiated or unirradiated TC31 oligonucleotides were radiolabeled and used as a probe in EMSA using fraction 37, containing the D-DDB. Fraction 37 was added to the reaction mixture (lanes 2 --6 and 8 --12) with increasing amounts of unlabeled unirradiated TC31 (lanes 3, 4, 9 and 10) or inducing amounts of unlabeled irradiated TC31 (lanes 5, 6, 11 and 12). The higher order band corresponds to two independent binding events on the same probe. Figure 2 | (A) Northern hybridization analysis of total RNA from Kc cells. (A) Northern hybridization analysis of total RNA from Kc cells. Cells were incubated for 1 --10 h after irradiation at 70 J /m2, separated on a 1.2% agarose --formaldehyde gel, transferred to a nylon membrane and probed with random primed 32P-labeled d-ddb1 cDNA (1774 --3420 bp). The blot was reprobed for the rp49 message as a loading control. (Lower panel) Amounts of d-ddb1 mRNA as determined using an imaging analyzer and expressed relative to the highest amount of transcripts 4 h after irradiation. (B) Northern hybridization analysis of total RNA from Kc cells. Cells were incubated for 0 --10 h after addition of 0.1% MMS, separated on a 1.2% agarose --formaldehyde gel, transferred to a nylon membrane and probed with a random primed 32P-labeled d-ddb1 cDNA (1774 --3420 bp). The blot was reprobed for the rp49 message as a loading control. (Lower panel) Amounts of d-ddb1 mRNA as determined using an imaging analyzer and expressed relative to the highest amount of transcripts 2 h after addition of MMS. Figure 3 | RNAi in Drosophila Kc cells. RNAi in Drosophila Kc cells. (A) The 600 bp d-ddb1 dsRNAs are indicated. (B) Northern hybridization analysis of total RNA from Kc cells prepared 3 days after addition of 0, 2.5 or 5.0 microg d-ddb1 dsRNAs is illustrated on the right. Total RNA was separated on a 1.2% agarose -- formaldehyde gel, transferred to a nylon membrane and probed with a random primed 32P-labeled d-ddb1 cDNA (1774 --3420 bp). (C) Kc cells were incubated for 0 --24 h with the indicated concentrations of d-ddb1 dsRNAs (1 --600 nt). Cells were prepared after 3 days and tested for survival after UV-irradiation at 70 J /m2. Figure 5 | Northern hybridization analysis of total RNA from D.melanogaster at various developmental stages. Northern hybridization analysis of total RNA from D.melanogaster at various developmental stages. (A) An aliquot of 30 microg total RNA was applied to each lane. Radiolabeled probes were used successively on the same membrane. rp-49 mRNA served as a loading control. (B) Relative amounts of d-ddb1 mRNA (1774 --3420 bp), normalized to rp-49 mRNA and expressed relative to the value for transcripts in unfertilized eggs. Figure 4 | Subcellular localization of D-DDB1 in Drosophila Kc cells after UV-irradiation (x1600). Subcellular localization of D-DDB1 in Drosophila Kc cells after UV-irradiation (x1600). Bar indicates 5 microm. Kc cells were stained with DAPI (left) and antibodies against D-DDB1 (middle). Merged images are shown on the right. (Upper panels) Without UV irradiation; (lower panels) with incubation for 12 h after 70 J /m2 UV-irradiation. Note the nuclear localization of D-DDB1 after UV-irradiation. Figure 6 | (A) Upstream genomic sequence of the d-ddb1 gene. (A) Upstream genomic sequence of the d-ddb1 gene. The bent arrow indicates the transcription initiation site. The boxes enclose the DRE sequence (5'-TATCGATA), located at position --218, and DRE-like sequence (5'-AATCGATT), located at position --264 with respect to the transcription initiation site. The putative TATA box is underlined. (B) Sequences of double-stranded oligonucleotides containing a DRE (DRE-P) in the PCNA promotor and double-stranded oligonucleotides containing DREF-binding sites or their derivatives in the d-ddb1 promoter. Figure 7 | (A) Complex formation between the probe (DRE-P, DRE-wt or DRE like-wt) and the Kc cell nuclear extract. (A) Complex formation between the probe (DRE-P, DRE-wt or DRE like-wt) and the Kc cell nuclear extract. A radiolabeled double-stranded oligonucleotide of DRE-P (lanes 1 --8), DRE-wt (lanes 9 --16) or DRE like-wt (lanes 17 --24) was incubated with the Kc cell nuclear extract (4 microg protein) in the presence or absence of competitor oligonucleotides (DRE-wt, DRE-mutD, DRE-del3, DRElike-wt, DRElike-mutD, DRElike-del3 or DRE-P, 100 ng /lane). (B) DREF bound to oligonucleotides containing the DRE-wt and DRE like-wt sequences. 32P-labeled DRE-P (lanes 1 --4), DRE-wt (lanes 5 --8) or DRE like-wt (lanes 9 --12) were incubated with Kc cell nuclear extract in the presence of anti-DREF monoclonal antibody (Mab-1) (lanes 3, 7 and 10), anti-DREF monoclonal antibody (Mab-4) (lanes 4, 8 and 11), anti-DNA polymerase alpha monoclonal antibody (4-8H) as a negative control (lanes 2, 6 and 12) or in the absence of antibody (lanes 1, 5 and 9). Figure 8 | (A) Constructs of d-ddb1 --luciferase fusion genes [p5'-510DDBluc (wt), p5'-510DDBwt-mutDluc (wt-mutD), p5'-510DDBwt-del3luc (wt-del3), p5'-510DDBmutD-wtluc (mutD-wt), p5'-510DDBdel3-wtluc (del3-wt), p5'-510DDBmutD-mutDluc (mutD-mutD), p5'-510DDBmutDdel3luc (mutD-del3), p5'-510DDBdel3-mutDluc (del3-mutD), p5'-510DDBdel3-del3luc (del3-del3) and p5'-210DDBluc (-DRE)] are shown. (A) Constructs of d-ddb1 --luciferase fusion genes [p5'-510DDBluc (wt), p5'-510DDBwt-mutDluc (wt-mutD), p5'-510DDBwt-del3luc (wt-del3), p5'-510DDBmutD-wtluc (mutD-wt), p5'-510DDBdel3-wtluc (del3-wt), p5'-510DDBmutD-mutDluc (mutD-mutD), p5'-510DDBmutDdel3luc (mutD-del3), p5'-510DDBdel3-mutDluc (del3-mutD), p5'-510DDBdel3-del3luc (del3-del3) and p5'-210DDBluc (-DRE)] are shown. (B) Effects of mutations in DRE and DRE-like sites on d-ddb1 gene promotor activety in Drosophila S2 cells, assessed with 500 ng aliquots of luciferase plasmids harboring wild-type or mutant d-ddb1 promotors (indicated on the left). p5'-510DDBluc (wt), p5'-510DDBmutDluc (mutD), p5'-510DDBdel3luc (del3), p5'-510DDBmutDmutDluc (mutD-mutD), p5'-510DDBmutDdel3luc (mutD-del3) and p5'-210DDBluc (-DRE) were transfected into S2 cells. Cell extracts were prepared to determine the luciferase expression levels. Averaged values obtained from two independent dishes with standard deviations are shown by closed bars as luciferase activity relative to those of p5'-510DDBluc (wt). Figure 9 | (Opposite) (A) D-DDB1 protein levels in D.melanogaster at various developmental stages and in Kc cells. (Opposite) (A) D-DDB1 protein levels in D.melanogaster at various developmental stages and in Kc cells. Aliquots of 25 microg protein from Drosophila bodies at various developmental stages were separated by 10% SDS --PAGE. The anti-D-DDB1 antibody reacted with one polypeptide band at 126 kDa. (B) Immunochemical localization of D-DDB1 in the imaginal discs of third instar larvae. The imaginal discs were dissected from third instar larvae and fixed. Polyclonal anti-D-DDB1 IgG (A --C) and normal rabbit IgG (D and E) were used as primary antibodies as indicated. Alkaline phosphatase-conjugated goat anti-mouse or anti-rabbit IgG was used as the secondary antibody. (A and D) Eye discs; (B and E) leg discs; (C and F) wing discs. (C) Expression of the D-DDB1 protein was immunohistochemically detected in Drosophila tissues at the adult stage. (A and F) Head frontal sections; (B and G) head median sections; (C) flight muscle sagittal section; (D) testis sagittal section; (E and H) ovary frontal and sagittal sections; (F --H) negative control stained with normal rabbit IgG. E, compound eye; GC.I, GC.II, GC.III, ganglion cells of periopticon, epiopticon and opticon; GngCl, ganglion cells (neurocytes); I, periopticon; II, epiopticon; III, opticon; MT, median fiber bundle; Oe, oesophagus; SoeGng, suboesophageal ganglion; Spg, spermatogonium; Spz, sperm. Figure 10 | (Opposite) D-DDB1 organization during D.melanogaster spermatogenesis. (Opposite) D-DDB1 organization during D.melanogaster spermatogenesis. Microscopic imaging of a testis triple labeled with D-DDB1 in green and DNA in blue and RPA30 in red. The merged view (DNA and D-DDB1 and RPA30) is shown: green, D-DDB1; blue, DNA; red, RPA30; yellow, D-DDB1 and RPA30. (A) In young spermatocytes D-DDB1 accumulates in the cytoplasm, although RPA30 is present in the nucleus. (B) In mature primary spermatocytes D-DDB1 and RPA30 are both in the nucleus. (C and D) At the telophase stage of meioses I and II D-DDB1 and RPA30 are both in the cytoplasm. (E) In early spermatids typical features include a spherical mitochondrial Nebenkern (NK) lacking D-DDB1 and RPA30. (F) In elongating spermatids D-DDB1 is outside the nucleus, although RPA30 is present within. (G) In fully elongated spermatids D-DDB1 and RPA30 are undetectable. (H) The merged overview. Bars, 10 microm. DISCUSSION : Damage-specific DNA-binding protein (DDB) has been implicated in numerous aspects of DNA repair ( --). To elucidate precise functions we have focused on DDB in an animal model, using D.melanogaster. We thereby succeeded in isolating four UV-irradiated DNA-specific binding proteins from Kc cells and established one of them to be a Drosophila homolog of the p127 subunit of DDB. We then cloned the d-ddb1 gene. D-DDB1 appeared to be a protein that recognizes UV-damaged DNA in vivo. It has been reported that CREB-binding protein and p300 histone acetyltransferase can interact with the small subunit of DDB (the DDB2 or p48 subunit) , while the human STAGA complex (SPT3 --TAFII31 --GCN5L acetylase) interacts with the large subunit of DDB (the DDB1 or p127 subunit) . DDB might have functions as a repair protein factor in the context of chromatin structure for recruitment of NER factors to DNA damage sites, because it physically interacts with histone acetyltransferase subunits, believed to be important in altering chromatin structure. Transcription of ddb2, but not ddb1, is strongly activated by p53 tumor suppressor protein in human cells . However, the response of DDB to DNA damage remains to be clarified. We investigated D-DDB1, a homolog of DDB1, and found it to be directly related to the Drosophila DNA repair system. This was confirmed in cells knocked out for the d-ddb1 gene by the RNAi method. To our knowledge there has been no report of ddb1-deficient cells, although lack of ddb2 has been studied . D-DDB1 was detected in the cytoplasm of Kc cells at interphase, but not in the nucleus, suggesting that D-DDB1 is ordinarily present in the cytoplasm. Interestingly, after UV-irradiation, D-DDB1 is shifted to the nucleus within 12 h. This phenomenon is in line with recruitment of NER factors to DNA damage sites. It has been reported that DDB2 possesses nuclear localization signals and is required for the nuclear accumulation of DDB1 . D-DDB1 was detected in adult brain (Fig. C). D-DDB1 protein may play an important role in the central nervous system (CNS) and muscles. DDB reportedly recognizes a class of oxygen free radical-induced base lesions . Since O2 consumption in the CNS and muscles is known to be particularly high as compared with other organs, their data suggested that the amount of D-DDB1 protein in the CNS and muscles of Drosophila might correspond to the necessity of NER in the CNS and muscles. Northern hybridization of developmental stages here revealed a pattern of expression very similar to those of Drosophila enzymes involved in nuclear DNA replication (DNA polymerases alpha and epsilon, PCNA, etc.) (,). During embryogenesis d-ddb1 transcript and D-DDB1 protein levels were high in the early stages when rates of DNA replication were maximal and decreased markedly as development proceeded. A clear decrease in d-ddb1 mRNA in the embryos from 0 to 20 h is shown in Figure . This may indicate that the presence of a AUUUA signal within the 3'-untranslated region of d-ddb1 mRNA is associated with rapid mRNA turnover. D-DDB1 protein level also decreases throughout embryogenesis, as does d-ddb1 mRNA level. This may indicate a short half-life of the D-DDB1 protein. Maternal loading of the d-ddb1 transcript indicates an importance related to blastogenesis. Reportedly, it has been shown that DDB2 interacts with the activation domain of E2F1 and stimulates E2F1-activated transcription in human cells . E2F is a transcription factor playing important roles in the regulation of cell proliferation . In mouse liver the nuclear levels of both DDB1 and DDB2 increase in late G1 phase and several hours before the peak of DNA replication . These phenomena may indicate that DDB has an important role in entering S phase. As shown here, DRE and DRE-like sequences are located upstream of the transcription initiation site of the d-ddb1 gene. The DRE /DREF system regulates the expression of several genes encoding key factors involved in nuclear DNA replication and the present gel mobility shift analyses have demonstrated that the DRE and DRE-like sites of the d-ddb1 gene are essential for formation of stable DNA --protein complexes. Furthermore, mutations in any one of these two DREF-binding sites resulted in an extensive reduction in promoter activity. The data suggest that transcription of the d-ddb1 gene is under control of the DRE /DREF system. The results of western blotting analysis of DREF levels during Drosophila development reported by Yamaguchi et al. are similar to those of northern blotting for d-ddb1. We conclude that D-DDB1 must contribute to both DNA repair and replication. Like D-DDB1, another DNA repair protein factor, Rad51, is reportedly induced several-fold in murine cells ectopically expressing E2F1 or E2F2 . It may thus be required during replication to ensure genomic integrity. The dE2F gene has three tandemly arranged DNA replication-related element (DREs) and is one of the most critical target genes of DREF . The d-ddb1 gene has two DRE sites and the fact that its transcription was activated by DREF is clearly of interest in this context. The mechanisms by which DDB stimulates E2F-activated transcription and functions in DNA repair should next be deciphered. Further analysis should provide more precise insights into the molecular roles in different physiological processes. Backmatter: PMID- 12202761 TI - Genome-wide detection of tissue-specific alternative splicing in the human transcriptome AB - We have developed an automated method for discovering tissue-specific regulation of alternative splicing through a genome-wide analysis of expressed sequence tags (ESTs). Using this approach, we have identified 667 tissue-specific alternative splice forms of human genes. We validated our muscle-specific and brain-specific splice forms for known genes. A high fraction (8 /10) were reported to have a matching tissue specificity by independent studies in the published literature. The number of tissue-specific alternative splice forms is highest in brain, while eye_retina, muscle, skin, testis and lymph have the greatest enrichment of tissue-specific splicing. Overall, 10 --30% of human alternatively spliced genes in our data show evidence of tissue-specific splice forms. Seventy-eight percent of our tissue-specific alternative splices appear to be novel discoveries. We present bioinformatics analysis of several tissue-specific splice forms, including automated protein isoform sequence and domain prediction, showing how our data can provide valuable insights into gene function in different tissues. For example, we have discovered a novel kidney-specific alternative splice form of the WNK1 gene, which appears to specifically disrupt its N-terminal kinase domain and may play a role in PHAII hypertension. Our database greatly expands knowledge of tissue-specific alternative splicing and provides a comprehensive dataset for investigating its functional roles and regulation in different human tissues. Keywords: INTRODUCTION : Recently, there has been growing interest in alternative splicing as a mechanism for expanding the repertoire of gene functions. Different combinations of exons can be spliced together to produce different mRNA isoforms of a gene, encoding structurally and functionally different protein products . The discovery from large-scale genomics studies that alternative splicing may occur in a very large fraction of human genes (35 --59%) suggests a major role for alternative splicing in the production of functional complexity in the human genome ( --). This hypothesis implies extensive regulation of alternative splicing. Alternative splicing can display strong specificity to a particular tissue or developmental stage (,), modulating the functional characteristics of protein isoforms in specific tissues . It has also been estimated that similar15% of disease-causing mutations in human genes involve misregulation of alternative splicing and errors in mRNA processing have been associated with cancer and other human diseases ( --). Despite growing interest in how alternative splicing is regulated (, --), relatively little is known about tissue-specific alternative splicing and its regulation, especially when compared with the sheer volume of information known about other mechanisms of functional control such as transcriptional regulation. For example, tissue specificities for only a small number of alternatively spliced genes (about 50) are listed in current alternative splicing databases (,). If alternative splicing plays as large and important a role in gene function regulation as recent genomics studies suggest, much more information is needed. There are many questions that need to be answered . What fraction of alternative splicing is tissue-specific? What proportion of tissue-specific splicing is associated with gross subdivisions of tissues (such as an entire organ like the brain) versus very specific cell types and developmental states? How can we efficiently identify the complete regulatory machinery controlling tissue specificity? What are the regulatory factors that mediate this process and what are the control sites that they recognize? What are the functional consequences of these alternative splicing events? To answer any of these questions, one essential prerequisite is large-scale discovery and characterization of tissue-specific alternative splicing, for example by microarray analysis ( --). This is needed both to provide biologists with information on whether 'their gene' is alternatively spliced in a tissue-specific manner (enabling them to study its functional consequences) and to give splice regulation researchers a big enough dataset to study mechanisms of splice regulation in many genes and tissues. One possible approach is to use genomics datasets such as expressed sequence tags (ESTs) for large-scale analysis of tissue specificity. This poses two challenges. Although the EST database provides some information about tissue source, these data are incomplete and inconsistent. Thus, a consistent tissue classification of this large dataset is needed to enable detection of tissue specificity. Much more importantly, interpretation of these data requires filtering for statistical significance. EST data often have poor coverage (i.e. only a small number of ESTs from a given tissue for a region of interest in a gene) and many sampling artifacts. For example, there can be dramatically different numbers of ESTs from different libraries or tissues, creating sample bias. This could give the erroneous impression that a given splice form is specific to one tissue (simply because many ESTs for this gene have been sequenced from that tissue and few from other tissues). We have sought to address both these problems. In this paper we present an automatic method to detect tissue-specific alternative splicing events using EST and genomic sequences. After constructing a tissue list of 46 human tissues with 2 million human ESTs, we generated a database of novel human alternative splices that is four times larger than our previous report and used Bayesian statistics to compare the relative abundance of every pair of alternative splices in these tissues. Using several statistical criteria for tissue specificity, we have identified 667 tissue-specific alternative splicing relationships and analyzed their distribution in human tissues. We have validated our results by comparison with independent studies. This genome-wide analysis of tissue specificity of alternative splicing will be made available as a part of the Human Alternative Splicing Database , to provide a useful resource to study the tissue-specific functions of transcripts and the association of tissue-specific variants with human diseases. MATERIALS AND METHODS : Data sources | Our analysis is based on three sources of data: human genomic sequence assemblies , human ESTs from the UniGene database and human EST library information. Human genomic assembly sequences (accession no. NT_XXXX) and 'draft' BAC clone sequences (accession nos ACXXXX, ALXXXXX) were downloaded from NCBI ( and ). Human ESTs and library information were downloaded from UniGene . Additional EST library information about human tissue sources was obtained from the NCBI Library Browser, downloaded from . The work described in this paper is based on the January 2002 release of the human genome and UniGene data. Library classification | Tissue source information for approximately 6900 human EST libraries was exhaustively examined to produce a consistent classification of human tissues suitable for tissue specificity calculations. We checked and refined the NCBI Library Browser classification (200 categories) to produce a considerably reduced classification (46 categories). For many libraries with unclear or incomplete tissue information in UniGene, we checked their dbEST entries for extra information about tissue source. Libraries recorded as having the same tissue source (e.g. 'brain') were combined into a single category, including both tumor and normal samples from that tissue. We sought to avoid mixing of multiple tissues during this procedure. If a library could not be clearly assigned to a single tissue (for example, if it was a pool of multiple samples from different tissues, or lacked clear information designating it as a sample from a single tissue), it was excluded from the final set. We excluded a total of 2652 EST libraries by these criteria (see Table ). Our final classification consisted of 4271 EST libraries in just 46 single tissue categories (see Table ). Tissue specificity scoring | Reliable detection of tissue specificity is complicated by poor EST library coverage (in many cases there are few ESTs from a given tissue for a given exon) and sampling bias problems (different tissues may have dramatically different numbers of ESTs sequenced). We therefore needed a statistical measure of tissue specificity that properly accounts for these sources of uncertainty and bias. Specifically, we cannot assume that the observed numbers of ESTs from one tissue exactly reflect the true proportions of different splice forms in that tissue. A larger number of ESTs gives a more confident estimate of those proportions; a smaller number of ESTs causes increased uncertainty. To take this into account rigorously when attempting to compare the proportions of a given splice form between different tissues we used Bayesian statistics, treating the true proportions as hidden variables and calculating confidence from the available observations. Suppose gene G has two mutually exclusive (i.e. alternative) splices S1 and S2. By 'mutually exclusive' we mean two splices that share one splice site but differ at the other splice site and which thus cannot both be present in a single transcript . Throughout this paper we will refer to the set of transcripts containing splice S1 as 'splice form S1'. For our hidden variables, let theta1T represent the hidden frequency of S1 in a specific tissue T and let theta1similar be its frequency in the pool of all other tissues (i.e. all tissues except T, symbolized by similar). Similarly, let theta2T and theta2similar represent the hidden frequencies of S2 in tissue T versus in the pool of all other tissues similar. By definition, these probabilities must be normalized: theta1T + theta2T = 1 theta1similar + theta2similar = 1 For our observations, let N1T and N2T be the total number of ESTs in tissue T observed to have splice S1 or S2, respectively. Similarly, let N1similar and N2similar be the total number of ESTs in the pool of all other tissues similar observed to have splice S1 or S2, respectively. Since our model assumes two mutually exclusive splice forms, the likelihood of the observations should follow a simple binomial distribution. For example, in tissue T We first calculated the confidence that splice S1 is preferred in tissue T as a Bayesian posterior probability: We used P(theta1T) = 1 as an uninformative prior probability. We also computed the posterior probability that splice S1 is preferred in the pool of all other tissues[P(theta1similar > 50%|obs)] in the same way, from the counts N1similar and N2similar. We defined a tissue specificity (TS) score of splice S1 for tissue T as the difference between this posterior probability for tissue T versus the pool of other tissues, TS = 100[P(theta1T > 50%|obs) -- P(theta1similar > 50%|obs)] To assess how stable the TS value is to possible error models, we calculated a robustness value rTS, which measures how much the TS value drops when a single EST observation of splice S1 is removed from tissue T. Specifically, we computed the negative log value of the relative change of the TS value caused by removing that EST, rTS = --log10(DeltaTS /TS), where DeltaTS = |TS(N1T) -- TS(N1T -- 1)|. rTS = 1 means TS drops by 10 --1 = 10% and rTS = 0 means TS drops to 0. Because each EST makes an equal contribution to the TS value, only one resampling step was required to calculate rTS. We also calculated rTSsimilar, the effect of removing one EST observation of splice S2 from the pool of other tissues, defined in the same way as rTS, using DeltaTS = |TS(N2similar) -- TS(N2similar -- 1)|. Criteria for high confidence (HC) tissue specificity were TS > 50, rTS > 0.9, rTSsimilar > 0.9; for low confidence (LC) TS > 0, rTS > 0.5, rTSsimilar > 0.5. A necessary (but insufficient) condition for the HC group was at least three EST observations of S1 in tissue T; for LC at least two EST observations of S1 in tissue T. Validation of tissue-specific splices of known genes | To search for alternative splicing information for a given gene, we performed thorough literature searches using PubMed, OMIM, LocusLink and other databases of alternative splicing. We sought information about sequencing of alternative splice forms and their tissue specificity. Isoform data without a complete reported sequence was not considered sufficient validation. To be counted as a match, an alternative splice identified in our database had to match a specific transcript sequence published in the literature. To be counted as a validated tissue specificity the isoform also had to be independently reported to be specific to the same tissue that we identified. For the validation data shown in Table we validated a sample set consisting of all brain-specific and muscle-specific alternative splices identified by our HC criteria on a previous dataset (UniGene and human genomic sequence data February 2001). All procedures and criteria were identical for different runs. We also used the GeneMine software system to visualize and validate versus the literature all aspects of the genomic mapping of our clusters, exons and introns, splices sites, alternative splicing and the impacts on protein structure and function, by examining all the features in the genomic --EST --mRNA multiple sequence alignments. Bioinformatics analysis of tissue-specific protein isoforms | To assess the effects of alternative splicing on the protein product, we predicted protein isoform sequences for each alternative splice, their protein domain composition and motif analysis. These results will be described in detail elsewhere. ORF prediction was performed using standard methods for identifying the longest open reading frame in each transcript. Protein domain prediction was performed using RPS-BLAST on the protein isoform sequences, against a database of protein domain sequences from SMART and PFAM , using cut-off thresholds of 10 --20 expectation. For all the examples with functional importance shown in this paper we also evaluated the effects of each alternative splice relationship by carefully examining the complete alignment of ESTs to genomic sequence using GeneMine software. Since an alternative splice can change where the coding region starts and ends, we adopted the policy that any alternative splice that alters the protein product will be classified as a 'coding region', regardless of its location relative to the GenBank CDS annotation. Table 1 | Construction of our EST library tissue classification Table 2 | Our EST library classification database of 46 human tissues Table 6 | Validation of brain- and muscle-specific isoforms versus independent literature RESULTS : Tissue classification of human ESTs | Since the cDNA library source of each public EST sequence is recorded, in principle this dataset could provide large-scale detection of tissue-specific alternative splicing, if each cDNA library could be associated with a specific tissue. Unfortunately, these data are from many different contributors and are not annotated in a uniform way. To provide a reliable basis for analyzing tissue specificity, we have carefully classified the approximately 7000 human EST libraries into distinct tissue classes. We began with the NCBI Library Browser classification, which consists of 200 categories covering 6923 libraries. Manually inspecting public information about cDNA libraries available from Unigene, dbEST and GenBank, we combined different categories that were from the same tissue, excluded many categories (e.g. 'head and neck') that did not correspond to a specific tissue and constructed a manually curated library tissue classification database . For this study we combined tumor and normal samples from each tissue source (e.g. 'brain'), although in the future it will be interesting to look for tissue specificities that distinguish these. Our final classification consisted of 46 tissues containing 4271 cDNA libraries and 2.2 million human ESTs . This represents 75% of ESTs in UniGene. This classification is by no means an optimally structured subdivision of the distinct tissues in the human body, but rather is intended to reflect the level of specificity present in the public cDNA library samples themselves. These samples are rarely more specific than an entire organ (e.g. 'brain'). As an example of our tissue classification, Table lists all the libraries classified as 'adipose' tissue. Genome-wide detection of alternative splicing | Using the latest human EST data (UniGene January 2002) and genomic sequence we performed a genome-wide analysis of alternative splicing as previously described . This conservative analysis process takes into account many factors, including mapping of EST consensus sequences to unique genomic locations, validation by intronic splice site sequences and very specific match requirements (two mutually exclusive splices that match exactly at one splice site but diverge at the other splice site) to report an alternative splice. This analysis identified 27 790 alternative splices (see Table ), approximately four times that of our previous analysis . These results will be accessible via our online Human Alternative Splicing Database . Detection of tissue-specific alternative splicing | To identify tissue-specific alternative splicing automatically and with statistical robustness, we developed a tissue specificity (TS) scoring function. This calculation measures the percent confidence that a specific splice S1 is preferred in a given tissue T1 (i.e. that S1 is found in a larger proportion of transcripts there than the alternative splice S2), minus the same confidence calculated for the pool of ESTs from all other tissues (see Materials and Methods for details). TS > 0 means that splice S1 is preferred in tissue T1 more than it is in other tissues. The higher the TS score is, the stronger the evidence of tissue specificity. For example, if the confidence that splice S1 is the major splice form was 70% in brain and 40% in the pool of other tissues, then the TS score would be 70 -- 40 = 30. If splice S1 is preferred in all tissues, it will get a low TS score. If there are insufficient EST counts to be confident about the proportion of S1 in tissue T1 or other tissues, this will also give a low TS score, both by decreasing the certainty that splice S1 is preferred in tissue T1 and increasing the possibility that splice S1 might be preferred in other tissues as well. Table illustrates the use of the TS score to distinguish tissue-specific splicing within ESTs from one gene (GAR22, UniGene cluster Hs.322852). We identified two alternative splicing relationships in this cluster (indicated in Table by their splice IDs, 17571 --17572 and 17577 --17578) and calculated TS scores for these alternative splices in the tissues in which they were observed. Of 11 candidate splice --tissue pairs, four had a positive TS score. However, the number of EST observations supporting these scores was not large. To take this into account, we calculated the 'robustness' of the TS score, which measures how much TS drops when a single EST observation of splice S1 in tissue T1 is removed from the sample. This assesses how stable the TS value is to possible error models (e.g. the library classification may have errors) and is analogous to the 'jack-knife', a common statistical resampling method . Defining the robustness as rTS = --log10(DeltaTS /TS), where DeltaTS = |TS(N1T) -- TS(N1T -- 1)| (see Materials and Methods for details), rTS = 1 means TS drops by 10 --1 = 10% and rTS = 0 means TS drops to 0. This measures the amount of EST evidence supporting the TS score; if N1T >> 1, then the robustness, rTS, can be high. Only one candidate tissue-specific splice in GAR22 (splice ID 17571 in tissue eye_retina) passes the combined criteria of >2-fold tissue specificity (TS > 50) and good robustness (rTS > 0.9, equivalent to a 14 point drop in TS; see Materials and Methods for details). It is striking that the only other observations of this splice form are in brain, which suggests that the putative association of this form with eye_retina is real and that this form may be found exclusively in neuronal tissue. We have computed TS scores for all alternative splices detected in the EST libraries contained in our 46 tissue classification database . We divided those with positive TS scores into two groups: a HC group with TS > 50% and rTS > 0.9, designed to screen out false positives (but causing a high level of false negatives), and a LC group with TS > 0 and rTS > 0.5. We identified 894 tissue-specific relationships for 667 alternative splices in the HC group and a total of 2873 alternative splices showing tissue specificity in the LC group. These data suggest that by our HC criteria 10.4% of alternatively spliced genes in the human genome have tissue specificity discernible in current EST data, and 30.6% by our LC criteria. Given the very gross form of tissue specificity of the EST libraries (typically an entire organ rather than a specific cell type), this probably underestimates the true extent of tissue-specific alternative splicing. Independent validation of our tissue specificity results | We have performed extensive validation analysis of our tissue-specific alternative splice forms. Figure shows one example of brain-specific alternative splicing detected automatically by our procedure in the IRF3 gene (Hs.75254, interferon regulatory factor-3) with a TS score of 88. IRF3 is a member of the IRF family and plays an important role in the virus- and double-stranded RNA-mediated induction of interferon beta (IFNbeta) and RANTES (regulated upon activation normal T cell expressed and secreted) (,). Our automated procedure detected two alternative splice forms: a longer mRNA consisting of eight exons and a short form in which exons I and II are replaced by a new exon IIa. These match isoforms reported in the literature as IRF3 and IRF3a. In the EST data we detected five ESTs in brain, all matching the IRF3a form, and five ESTs elsewhere, of which four match the IRF3 form. According to the literature, both isoforms are expressed in multiple tissues but the ratio IRF3a:IRF3 is dramatically high in brain compared with other human tissues . This example also illustrates the functional interpretability of the large structural changes that alternative splicing often causes. We performed a series of bioinformatics analyses to predict the protein isoform sequences and protein structural domains (see Materials and Methods). Prediction of the protein products identified an ORF in both cases, revealing a replacement AUG start site in exon IIa that encodes 22 amino acids before entering exon III in the same coding frame as in the long IRF3 form. Protein domain prediction using SMART and PFAM showed that the brain-specific splice disrupts the 110 amino acid DNA-binding domain (PFAM id 00605) at the IRF3 N-terminus, by replacing the first 55 amino acids with the 22 amino acids from exon IIa (Fig. B). On the basis of this simple domain analysis, we would predict that in brain the role of IRF3 in regulating gene expression (e.g. activating IFNbeta and RANTES) would be dramatically altered. This is strongly validated by the published literature. IRF3 has been shown to bind to the IFNbeta promoter and up-regulate the transcription of IFNbeta with other enhancers after virus infection (,). The brain-specific splice eliminates the ability of IRF3 to bind to the IFNbeta promoter in vitro and seems to play a protective role in brain, reducing the toxic effect of IFNbeta (,) by suppressing its expression in brain . To assess the overall accuracy of our automated tissue specificity detection method for many genes, we followed this same procedure of testing for independent validation of our results. We used two tissues, brain and muscle, as the test samples for this validation, since many studies of alternative splicing have been done for genes in these tissues . For a random sample of 37 tissue-specific alternative splices identified in the HC group for these tissues, we searched the published literature to see whether tissue specificity was independently reported for those genes. To count a splice form as validated, we required that a complete sequence matching our splice form be found in the literature and be demonstrated to be specific to the tissue reported by our procedure. Previously unknown genes (e.g. Hs.7973) were excluded from this analysis, since no studies of their tissue-specific alternative splicing have been published. We found that 80% (8 /10) of our brain- and muscle-specific splices were validated by the existing literature . In the two cases where the splice forms were not validated by a matching isoform in the literature, published papers reported a different isoform that matched our tissue specificity. For example, in Hs.57435 we detected a brain-specific alternative splice. This splice was not validated by a matching sequence in the literature, but another isoform was reported and shown to be brain-specific. Unfortunately, there are no ESTs from brain that align to this region of the gene, so there was no possibility of our detecting this form. Similarly, our muscle-specific alternative splice for Hs.239069 was not validated by the literature, but another isoform of this gene was reported to be muscle-specific. These coincidences suggest that even the tissue specificities we detected for these genes may not actually be incorrect. There were five additional cases where the splice form we detected was validated by the literature but no tissue specificity studies were reported. Within the limits of available published data, these results indicate that the reliability of our database is likely to be high. We have also used these data to estimate the fraction of our results that are novel discoveries of tissue-specific alternative splicing. These come from two categories: detecting previously unknown tissue specificity for a known alternative splice (in a known gene); detecting tissue specificity for a previously unknown alternative splice (in a known gene or in a novel gene). Combining these categories, 78% (29 /37) of our brain- and muscle-specific alternative splices appear to be novel. It seems likely that our database can be a valuable source of interesting discoveries for biologists who study these genes, as well as researchers who study tissue-specific regulation of splicing. Distribution and enrichment of tissue-specific alternative splicing | We analyzed the distribution of tissue-specific alternative splicing over the 46 human tissues in our classification . We identified tissue-specific alternative splicing in the HC group for 30 of the 46 human tissues. The largest category by total number of tissue-specific splice forms was brain, which represented 18% of all tissue-specific alternative splicing events we observed. The tissues that each accounted for at least 4% of observed tissue-specific forms were brain, eye_retina, lung, liver, pancreas, placenta, ovary, uterus, testis, lymph, muscle and skin. This is consistent with results from a survey of the alternative exons from the published literature, in which brain and neurons were ranked highest among human tissues . Since the number of ESTs sampled is very different for each tissue, we have also normalized the rate of detection of tissue-specific alternative splices within each tissue by its total number of ESTs. We defined the enrichment factor (EF) for a tissue as the proportion of total tissue-specific alternative splicing that it represents divided by the proportion of total ESTs that it represents. Figure compares the enrichment factors for our classified human tissues. Skin, eye_retina and muscle ranked highest (2.4 times more tissue-specific splice forms than average), with lymph (2.0) and testis next (1.8), followed by brain (1.2) and liver (1.1). Other tissues with above-average tissue specificity were bladder, mammary_ gland and pancreas. Overall, this is consistent with previous studies indicating the immune and nervous systems as major loci of alternative splicing (,) and a common focus on regulation of alternative splicing in neuronal tissues . It is also striking that at the high end of the ranking, more sharply defined tissue categories (e.g. retina, muscle, skin, lymph) returned higher yields of tissue specificity detection than broadly defined categories such as brain. This does not necessarily mean that there is more tissue-specific alternative splicing in retina or muscle than in the many tissues in the brain. Instead, it may simply reflect serious limitations in the nature of the library samples (that most of them are grossly defined, combining all the different tissues that compose an entire organ) and of our TS scoring calculation. Bioinformatics analysis of a novel tissue-specific splice form | To demonstrate the value of our database for biological discovery, Figure shows our analysis of a representative example (Hs.184592) in which we detected a novel kidney-specific alternative splicing event (TS score 94). As part of our genome-wide analysis of alternative splicing, we generated predicted protein isoforms and analyzed their domain composition by searching against protein domain databases. This gene encodes a serine /threonine protein kinase, WNK1 (with no K = lysine), which has only recently been described . This gene has 28 exons and encodes a huge protein with a kinase domain near its N-terminus and two coiled-coil conserved domains . Its name refers to the surprising replacement of an active site lysine residue with a cysteine, which leaves kinase activity intact. Our automated procedure identified a novel kidney-specific alternative splicing event between exons IV and V (Fig. B). For the isoform we detected outside kidney our deduced protein sequence is identical to the reported protein sequence . Exons IV, V and VI encode the second half of the kinase domain of this enzyme (Fig. C). In the kidney-specific isoform exon IV is replaced by exon IVa, drastically altering the protein sequence. It contains a 63 nt upstream in-frame stop codon (UGA), as well as a subsequent start codon (AUG) for an ORF that extends in-frame into the rest of the normal protein sequence in exons V onwards. Thus it is likely that exon IVa encodes an alternative 30 amino acid N-terminus of the WNK1 protein, replacing 384 amino acids of the usual WNK1 protein N-terminus. This appears to disrupt the kinase domain and to eliminate WNK1 kinase activity specifically in kidney. Our data show eight ESTs aligning with exon IVa and extending up to 42 nt upstream of the AUG start codon. However, because ESTs are short fragments, in this case they do not extend to a 3' splice site and thus do not show where the beginning of exon IVa might be. However, in the 21 nt between the in-frame stop codon and the start of the first EST alignment there is no consensus 3' splice site (polypyrimidine tract + AG). Thus, even if exon IVa is spliced to upstream exons, the in-frame stop codon would evidently be included and the kinase domain would be removed from the WNK1 protein product. It is also possible that exon IVa represents an alternative promoter site. Although this is a novel discovery, there is experimental data that support it. Previous studies of the expression of WNK1 have reported both an 11 --12 kb band observed ubiquitously in many tissues and a 9.5 --10 kb band expressed at high level in the kidney (,). This truncated transcript is consistent with the kidney-specific exon IVa alternative splice that we have identified, but the reported band has not been further characterized. In 7 --10 ESTs we have also detected alternative splicing in a variety of tissues omitting exons IX, XI and /or XII (and a single EST omitting exon XXVI) in WNK1. Exons IX, XI and XII have been independently reported to be alternatively spliced (,). Bioinformatics analysis of a known splice variant | To illustrate further the value of our database, Figure shows our bioinformatics analysis for CDC42 (cell division cycle 42). Many studies of this gene provide insight into its splicing and isoforms ( --). CDC42 is a member of the Rho family, which is a group of small GTPases. CDC42 plays multiple functional roles in cell regulation . Our automated analysis detected an alternative splice in CDC42, with one splice form showing brain specificity. Our forms matched the two isoforms reported in the literature, CDC42b and CDC42u. In agreement with our EST results, CDC42b has been reported to be expressed exclusively in brain, while CDC42u is expressed in a wide variety of tissues . Bioinformatics analysis of these forms was revealing. ORF prediction for the two splice forms produced almost identical protein sequences. The brain-specific alternative splice replaced exon VII (which encodes the protein C-terminus) with a new exon (VI) as the last exon. Remarkably, exon VI supplied an almost identical C-terminal sequence (Fig. C), of exactly the same length and with 20 of 29 amino acids identical. The divergent nine amino acids include a C-terminal dilysine motif of retrieval receptors that has been shown to be critical for binding to coatomer complex (COP) in the endoplasmic reticulum and to cargo receptors in the Golgi apparatus ( --). CDC42u has this motif at positions 183 --184, but our analysis indicated that CDC42b eliminates this motif, replacing it with glutamine and proline (QP). This suggests that the brain-specific splicing blocks CDC42 binding to COP in brain. This novel hypothesis is consistent with available experimental data. In vitro site-directed mutagenesis of the CDC42u sequence, replacing lysines 183 --184 with serine, eliminated binding of the gamma-COP subunit . Moreover, it has been shown that binding of gamma-COP is necessary for CDC42 to induce malignant transformation , suggesting that this brain-specific splice has functional importance. Figure 1 | Brain-specific alternative splicing of IRF3. Brain-specific alternative splicing of IRF3. (A) Genomic structure of the IRF3 gene. Exons are shown as boxes and colors show alternative exons. Splice a is specific to brain. (B) The two alternative forms of IRF3 mRNA inferred from the expressed sequence data. The protein coding region is indicated by an arrow beneath each form. (C) Schematic representation of the IRF3 protein. The DNA-binding domain, the NES element, the proline-rich region and the C-terminal IRF association domain are indicated. Dashed lines mark the boundaries of the DNA-binding domain. Figure 2 | Tissue distribution of human tissue-specific alternative splicing. Tissue distribution of human tissue-specific alternative splicing. Areas on the pie chart are proportional to the total number of alternative splices with high confidence tissue specificity for a particular tissue. Figure 3 | Enrichment of tissue-specific alternative splicing in 30 human tissues. Enrichment of tissue-specific alternative splicing in 30 human tissues. The y-axis shows the enrichment factor for each tissue, defined as the ratio of the number of tissue-specific alternative splices observed in a tissue divided by the total number of ESTs observed in that tissue, normalized to have an average value of 1 (see text). Figure 4 | Kidney-specific alternative splicing of WNK1. Kidney-specific alternative splicing of WNK1. (A) Genomic structure of WNK1. The genomic segment spanning WNK1 is represented by a horizontal line and exons by numbered vertical lines. Pink vertical lines indicate the alternative exons, IVa, IX, XI, XII and XXVI. (B) Gene structure for exons IV --VIII of the WNK1 gene. Exons are shown as boxes and colors show alternative exons. Splice a is specific to kidney. The putative in-frame stop codon UGA and start codon AUG are indicated. (C) The two alternative forms of WNK1 mRNA inferred from the expressed sequence data and the schematic representation of WNK1 protein sequences. The conserved kinase domain, two coiled-coil (CC) domains and the corresponding protein regions of mRNA forms are indicated. Three amino acids (K233, C250 and D368) that are required for the kinase activity of WNK1 are marked by flags on the WNK1 protein. Figure 5 | Brain-specific alternative splicing of CDC42. Brain-specific alternative splicing of CDC42. (A) Genomic structure of the CDC42 gene. Exons are shown as boxes and colors show alternative exons. Splice a is specific to brain. (B) The two alternative forms of CDC42 mRNA inferred from the expressed sequence data. The protein coding region is indicated by an arrow beneath each form. (C) Alignment between the protein sequences encoded by the alternative exons VI and VII of CDC42. Conserved amino acids are in bold. The dilysine motif is indicated in red and the stop codons by asterisks. Table 3 | Libraries in the tissue category 'adipose' Table 5 | Genome-wide detection of tissue-specific alternative splicing Table 4 | TS calculations for GAR22 (UniGene cluster Hs.322852) DISCUSSION : Our results can be useful to biologists in several ways. First, they provide a validated, automatic method for large-scale discovery of tissue-specific alternative splicing, which can be applied to many EST and other datasets where tissue information is available. Second, we have discovered 667 tissue-specific splice forms in the human transcriptome. Our comparisons with the published literature suggest that up to 78% of our tissue specificity findings are novel. These data can furnish biologists with many new functional insights into well-studied genes (by identifying a novel tissue-specific splice form), which can be of great interest for further experimental study. Our data can also provide interesting functional suggestions for unknown genes, since observation of tissue specificity (combined with other information, such as homology) may itself suggest fruitful directions for research. Moreover, the large scale of alternative splice impact on the protein product (e.g. removal or addition of a domain) often yields interpretable functional implications (e.g. removal of a DNA-binding activity, as in IRF3). Finally, researchers who study regulation of splicing can benefit from this large, searchable database of tissue-specific alternative splicing spanning many distinct tissue types. Benefits and deficiencies of our approach | Our use of readily available expressed sequence data has both advantages and disadvantages. One of the biggest problems with ESTs is their fragmentary character; the difficulty of interpreting individual ESTs (because they are tiny fragments rather than full-length cDNAs) and the high rate of errors in their sequencing and clustering. To address this, we perform a rigorous, comprehensive analysis of the total set of all EST data, which does not assume confidence in single ESTs. Multiple forms of evidence are required. This process depends on making a complete, clearly interpretable match between the genomic sequence of a gene and the total set of ESTs that map to that gene location, carefully considering many details such as intronic splice junction sequences and current knowledge about mechanisms of alternative splicing . Our procedure is conservative (designed to avoid false positives) in the sense that if a given set of ESTs does not fit its rigid model, they are simply excluded from our analysis. This procedure is more likely to give false negative errors (real alternative splices that are missed by the analysis) than false positives (reported alternative splices that are incorrect). An advantage of using ESTs for alternative splice detection is that this furnishes exact sequence information for the novel splice form, in many cases suggesting a clearly interpretable functional effect. In contrast, northern blot, PCR-based and microarray hybridization-based methods do not directly read out the sequence of a novel form as sequencing does and instead would require indirect means to infer a sequence. For example, this can be done by using the genomic sequence to predict what splice forms are possible and then correlate the observed transcript sizes (on a northern blot) or hybridization signals (on a microarray) versus these forms. Thus these experiments are most interpretable when the set of probes has been carefully designed to distinguish the particular set of alternative forms that are expected. EST data, on the other hand, can readily detect an unexpected new form by providing a direct readout of its sequence. This is particularly helpful for interpretation of newly detected forms. Knowing the exact sequence change that an alternative splice produces is the difference between simply having a new 'band on a gel' versus being able to apply the full resources of sequence analysis and available literature to interpreting its likely functional impact. For example, for WNK1, previous studies may have detected our kidney-specific WNK1 isoform (reported only as a 9.5 --10 kb band; ,), but have not published data giving this functional significance. In contrast, detection of a kidney-specific splice form by our automatic procedure immediately suggested an interesting functional impact due to truncation of the highly conserved N-terminal kinase domain. Our approach has many disadvantages. Our results are likely to be far from complete, in that they may not include many genes that do have tissue-specific splice forms. We have designed our approach to reduce the rate of false positives (incorrectly reporting a splice form to be tissue-specific), by accepting a much higher rate of false negatives (failing to report a splice form that actually is tissue-specific). We have tried to set the cut-off score for our HC group high enough to avoid a high rate of false positives, and the independent validation results (80% validation) support this. However, this means that a number of genes with tissue-specific alternative splicing may be missed, either because of inadequate EST data or overly conservative scoring. Our TS scoring function is based on rigorous statistical inference methods, but is far from perfect in dealing with the many possible problems in this data. For example, the TS metric quantifies the evidence that a given splice form is preferred (>50% frequency) in a specific tissue and not preferred in the pool of all other tissues. It is certainly possible that a splice form could be tissue-specific (e.g. found only in brain) and still be a minor splice form in that tissue (i.e. <50% frequency). The TS metric can miss such cases and thus is not a completely satisfactory definition of tissue specificity. This could be corrected by a simple Z-test of the null hypothesis that the proportion of splice types is the same in each tissue. More fundamentally, our procedure depends on pooling (combining many different libraries into one tissue class) to get enough counts for each class so the results will be statistically significant. Pooling can obscure real specificity signals if the libraries that are being pooled as one tissue actually have very different patterns of splice form expression. Our current metric does not address this. However, while our tissue specificity scoring could be improved, we suspect that problems in the very nature of the EST data and libraries are more serious. These problems are as follows. Working with EST data, our method can handle cases where more than two alternative splice forms are detected in a gene, but it cannot distinguish whether they are combined in a coordinated way. Since multiple alternative splicing events are observed as independent events in the EST fragment data, our method treats them as independent events in its scoring, i.e. it calculates TS scores independently for each alternative splicing event (i.e. pair of mutually exclusive splices) in the gene. If multiple alternative splices were combined in a coordinated way (revealed by full-length transcript sequences), our method would still correctly detect their tissue specificity patterns individually. However, it would be up to the user to notice that they all had the same pattern and indeed were all observed in the same transcripts. A database of novel biological discoveries and functional implications | We believe our database of 667 tissue-specific splice forms can be a rich source of discovery for researchers studying human biology and disease. While there is growing interest in alternative splicing and tissue-specific splice forms, there are relatively few large-scale information resources for this field, compared with other areas such as genome annotation /gene discovery (databases such as Ensembl), polymorphism (databases such as dbSNP) or the study of transcriptional regulation (databases such as TRANSFAC). In addition to several databases of alternative splicing from the literature or ESTs ( --,,, --), there are databases of known tissue-specific alternative splicing. The Alternative Exon Database includes 379 human alternative splices, of which 30 exons (in 19 genes) are reported to be specific to brain, muscle or two tissues . As part of a computational analysis of candidate intronic splice regulatory elements, Brudno et al. created a collection of 25 brain-specific alternative spliced exons. However, at this time there is still no single resource where one can go to reliably find all splice form specificities reported in the literature. For example, to seek validation for our brain- and muscle-specific forms we had to perform extensive manual literature searches. Our database can make some contribution to this, since it contains approximately 167 tissue specificities that are likely to be already known. However, its major value is providing previously unknown alternative splice forms that show tissue specificity: approximately 500 with high confidence and 2200 that show evidence of tissue specificity with low confidence. Furthermore, the EST data are continuing to grow. From February 2001 to January 2002, for example, the human EST data grew from 2.4 million to 3 million, a 27% increase over 11 months. Furthermore, ESTs from many other organisms are also being sequenced. Thus, the method presented in this paper can be applied in the future to much more EST data, to greatly expand the database of tissue-specific alternative splice forms. Our database can provide biologists with valuable bioinformatics analyses that suggest hypotheses about function. As we have showed in a number of cases, the large-scale changes in the protein product produced by alternative splicing often make them interpretable enough to suggest exciting ideas that merit further experimental tests. Our database provides biologists with essential information for interpreting functional impact, such as inferred protein isoform sequences and predicted changes in protein domain composition based on conserved domain databases such as SMART and PFAM. We will continue to add useful analyses, such as transmembrane domain prediction , localization signal analysis, etc. Our data can also be useful for studies of the regulation of mRNA splicing. Given the relatively small number of genes and tissues in which the mechanistic details of splice regulation have been studied carefully (for a review see ), it seems likely that a large new database of tissue-specific alternative splicing can be a valuable resource for the field. First of all, most of our tissue-specific splice form data appear to be novel, providing researchers with many new cases of tissue-specific splicing to work on. Second, it spans a large number of genes (454 in the HC group, 1572 in the LC group) with different functions and gene structures, giving researchers a very diverse set to study. Third, it spans a large number of tissues, providing many more examples for tissue specificities that have been previously studied (e.g. neuronal), as well as many examples of tissue specificities that have not been studied in full mechanistic detail before. Finally, our database provides a lot of useful information for mechanistic studies, including the genomic sequence for each gene and the detailed evidence from expressed sequences for each exon -- intron junction and alternative splice. In principle this provides the information needed not only for designing appropriate experimental strategies to study the regulation of these splice forms (e.g. probe sequences, PCR primers, etc.), but also for searching for possible binding sites for known or novel splice regulatory factors. The dataset is large enough (e.g. 213 HC brain- and retina-specific alternative splices) that statistical analysis might be useful for detecting novel tissue specificity motifs. Observation of tissue specificity also adds valuable information to de novo alternative splice discovery databases. An extremely challenging problem for the field is how to validate novel splice forms efficiently. What fraction of the enormous new datasets of alternative splicing discovery (e.g. 27 790 in this paper) represent real biological forms of functional regulation, as opposed to experimental or bioinformatics artifacts (,)? This is not an easy question to answer. High throughput technologies such as microarrays can help address part of this, by providing much more experimental data indicating that these individual forms really are abundant in cells. However, simply showing that a form is present does not prove that it has functional importance for biology. Observation of tissue specificity is one good starting point for answering this harder functional question. First of all, the very fact of observing tissue specificity demonstrates a non-random pattern in the putative alternative splicing data. This makes it much less likely that a given splice form (e.g. detected by an automated procedure such as ours) is simply an experimental or bioinformatics error. Secondly, evidence of biological regulation (i.e. that a form is tissue-specific) can itself be taken as evidence of participation in a functional process. Thus, one way to look at our data is to assert that the most interesting and most reliable alternative splices from our automatic detection procedure (out of the total of 27 790) are simply those for which we found a tissue specificity. These are probably the most fruitful starting point for further experimental studies, both by individual researchers and by high throughput technologies such as microarrays. Medical interest of tissue-specific alternative splicing: WNK1 | For example, there is some evidence that the kidney-specific alternative splicing of WNK1 discovered in this paper could have medical importance. A genome scan of hypertension patients in a family study identified WNK1 as a genetic cause of pseudohypoaldosteronism type II (PHAII) hypertension and found a specific mutation of WNK1 in a large proportion of the patients . The homologous gene WNK4 shares the same key features (conserved kinase domain and two coiled-coil domains), and mutations in WNK4 have also been linked to PHAII hypertension. Consistent with this hypothesis and the chloride-dependent character of PHAII hypertension, the WNK1 and WNK4 proteins have been localized to the distal renal tubules of the kidney, which play a key role in maintaining the body's electrolyte balance . WNK4 appears to be expressed exclusively in kidney . Our discovery of a kidney-specific disruption of the kinase domain of WNK1 by alternative splicing suggests a possible hypothesis about the pathogenesis of PHAII hypertension. Normal WNK1 function (including the kinase) should be expressed only outside the kidney. Within kidney, WNK1 function is restricted or altered by disruption of its kinase domain, whose activity is replaced by that of WNK4. An unusual feature of WNK1 is that the WNK1 mutation observed in PHAII patients is a deletion in the intron between exons I and II, which ordinarily would have no effect on the protein. On the other hand, this could have an important effect on regulation of alternative splicing. Deletion of intronic splice regulatory elements could lead to misregulation or even loss of the tissue-specific splicing of the WNK1 transcript. For example, if PHAII patients produce the normal isoform of WNK1 in kidney, they will express fully functional WNK1 in their kidneys in addition to WNK4. This dosage effect could alter regulatory balances in the kidney and cause hypertension. Backmatter: PMID- 12202775 TI - Human non-synonymous SNPs: server and survey AB - Human single nucleotide polymorphisms (SNPs) represent the most frequent type of human population DNA variation. One of the main goals of SNP research is to understand the genetics of the human phenotype variation and especially the genetic basis of human complex diseases. Non-synonymous coding SNPs (nsSNPs) comprise a group of SNPs that, together with SNPs in regulatory regions, are believed to have the highest impact on phenotype. Here we present a World Wide Web server to predict the effect of an nsSNP on protein structure and function. The prediction method enabled analysis of the publicly available SNP database HGVbase, which gave rise to a dataset of nsSNPs with predicted functionality. The dataset was further used to compare the effect of various structural and functional characteristics of amino acid substitutions responsible for phenotypic display of nsSNPs. We also studied the dependence of selective pressure on the structural and functional properties of proteins. We found that in our dataset the selection pressure against deleterious SNPs depends on the molecular function of the protein, although it is insensitive to several other protein features considered. The strongest selective pressure was detected for proteins involved in transcription regulation. Keywords: INTRODUCTION : A considerable effort is underway to relate human phenotypes to variation at the DNA level. Most human genetic variation is represented by single nucleotide polymorphisms (SNPs) and many of them are believed to cause phenotypic differences between individuals. However, identifying SNPs responsible for specific phenotypes appears to be a problem that is very difficult to solve. The concept of association studies has been proposed as an experimental technique to identify SNPs underlying complex phenotypes, mostly human multifactorial disorders . The question of study design is, however, disputable. Linkage disequilibrium-based whole genome scanning (,) has the advantage of being a completely hypothesis-free approach, though possibly too demanding because of the extraordinary number of markers to be screened. Candidate gene studies (,) try to reduce the number of SNPs to those from genes most likely to constitute the genetic basis of the disease. Although, even in the latter case, especially if large sets of candidate genes are considered, multiple testing of hundreds and even thousands of SNPs makes detection of the association difficult. A possible way to overcome the problem of testing overwhelming numbers of SNPs, especially in the case of candidate gene studies, would be to prioritise SNPs according to their functional significance (,). As a priori biological knowledge can be used to reduce the number of SNPs by focusing on specific genomic regions or gene sets, bioinformatics expertise may help to discriminate between neutral SNPs, which constitute the majority of genetic variation, and SNPs of likely functional importance. Below, we specifically focus on non-synonymous SNPs (nsSNPs), i.e. SNPs located in coding regions and resulting in amino acid variation in the protein products of genes. It has been shown in several recent studies ( --) that the impact of amino acid allelic variants on protein structure and function can be predicted by analysis of multiple sequence alignments and protein 3D structures. As we demonstrated in an earlier work, these predictions correlate with the effect of natural selection seen as an excess of rare alleles (,). Therefore, predictions at the molecular level reveal SNPs affecting actual phenotypes. Here we present: (i) a Web server for annotation of functional nsSNPs ; (ii) a dataset of nsSNPs extracted from a public SNP database, HGVbase ; (iii) an analysis of these data with regard to predicted effect on protein structure and function. Prioritisation of SNPs in the candidate gene approach is not the only suggested use of the PolyPhen (polymorphism phenotyping) server and the collection of nsSNPs. The server could also be useful to reveal the structural basis of disease mutations and explain the molecular cause of a disease. This might help in some cases to identify the causative allelic variant after a disease has been linked to a particular locus. On the other hand, since numerous disease associations published recently could not be confirmed by subsequent independent studies (,), the independent evidence of functionality of a nsSNP could be an additional argument to discriminate true associations from false positives. Analysis of the database of nsSNPs enabled us to test whether certain characteristics of proteins are associated with accumulation of nsSNPs (especially slightly deleterious nsSNPs). MATERIALS AND METHODS : PolyPhen is a World Wide Web server devoted to automated functional annotation of coding nsSNPs. PolyPhen input is the amino acid sequence of a protein or the SWALL database ID or accession number together with sequence position and two amino acid variants characterising the polymorphism. Given the input, PolyPhen starts a fully automated pipeline of several programs described step by step in this section. The pipeline is schematically presented in Figure . The server was used to annotate all SNPs deposited in the HGVbase database and the resulting dataset of annotated SNPs is available at . Identifying nsSNPs in known genes | The necessary first step in the analysis of nsSNPs is to identify whether a given SNP is indeed non-synonymous. For this purpose we map SNPs onto known proteins on the basis of SNP DNA flanking sequences. Flanking genomic sequences of SNPs from HGVbase with length 25 bp each have been translated in all six possible frames and searched for in the proteins in the human proteins subset of the SWALL database . Protein sequences and genomic fragments were pre-processed with the SEG , XNU , RepeatMasker and DUST programs, which are used to filter out areas of low compositional complexity, regions containing internal repeats of short periodicity and known human genomic repeat sequences. ALU subfamily proteins were also excluded from the set. We required that at least one translated flanking sequence should have an exact match with a database protein sequence. If this match was detected, we further required that the second flanking sequence had either an exact match with the protein sequence or matched the protein sequence in all positions until the end of the protein or a conventional exon /intron border is observed. The resulting mapping of a SNP onto a protein sequence is always unique. The above procedure is available as a stand alone World Wide Web-based program snp2prot. The link to this program is provided from the main PolyPhen page. We also provide a link to the SNP annotation tool HNP (Y.Yuan, unpublished results). After processing HGVbase v.12 (983 589 SNP entries), we obtained a set of 20 462 coding SNPs. Of these, 11 152 were non-synonymous, whereas 9310 were synonymous SNPs and do not produce any change in the amino acid sequence. The nsSNPs formed our dataset, which can be downloaded as one text file or searched against with a straightforward World Wide Web-based engine. The search results contain links to the other databases that provide additional information, e.g. chromosomal location of a nsSNP. PolyPhen analysis of nsSNPs | Sequence-based characterisation of the substitution site. The substitution may occur at a specific site, e.g. active or binding, or in a non-globular, e.g. transmembrane, region. A query identifies the protein by its SWALL accession number or ID or by the sequence itself. In the latter case, PolyPhen tries to find the given sequence in the human subset of the SWALL database and use the FT (feature table) section of the corresponding entry. If the sequence cannot be found in the human subset of SWALL, this step is skipped. PolyPhen checks if the amino acid replacement occurs at a site that is annotated in the SWALL database feature table as DISULFID, THIOLEST or THIOETH bond, BINDING, ACT_SITE, LIPID, METAL, SITE or MOD_RES site or as a site located in a TRANSMEM, SIGNAL or PROPEP region. PolyPhen also uses the TMHMM algorithm to predict transmembrane regions, the Coils2 program to predict coiled coil regions and the SignalP program to predict signal peptide regions of the protein sequences. For a substitution in a transmembrane region, PolyPhen uses the PHAT transmembrane-specific matrix score to evaluate possible functional effect of a nsSNP in the transmembrane region. At this step PolyPhen memorises all positions that are annotated in the query protein as BINDING, ACT_SITE, LIPID or METAL. At a later stage, if the search for a homologous protein with known 3D structure is successful, it is checked whether the substitution site is in spatial contact with these critical residues. Profile analysis of homologous sequences. The amino acid replacement may be incompatible with the spectrum of substitutions observed at that position in a family of homologous proteins. PolyPhen identifies homologues of the input sequences via a BLAST search of the NRDB database. The set of aligned sequences with sequence identity to the input sequence in the range 30 --94% (inclusive) is used by the new version of the PSIC (position-specific independent counts) software to calculate the so-called profile matrix . Elements of the matrix (profile scores) are logarithmic ratios of the likelihood of a given amino acid occurring at a particular site to the likelihood of this amino acid occurring at any site (background frequency). PolyPhen computes the absolute value of the difference between profile scores of both allelic variants in the polymorphic position. PolyPhen also shows the number of aligned sequences at the query position; this may be used to assess the reliability of profile score calculations. Mapping of the substitution site to known protein 3-dimensional structures. Mapping of an amino acid replacement to a known 3D structure reveals whether the replacement is likely to destroy the hydrophobic core of a protein, electrostatic interactions, interactions with ligands or other important features of a protein. If the spatial structure of a query protein is unknown, one can use a homologous protein of known structure. PolyPhen carries out a BLAST query of a sequence against a protein structure database [PDB or PQS , see below] and retains all hits that meet the given criteria. For instance, the default sequence identity threshold is set to 50%, since this value guarantees the conservation of basic structural characteristics. Minimal hit length and maximal length of gaps are by default set to 100 and 20, respectively. The position of the substitution is then mapped onto the corresponding positions in all retained hits. By default, a hit with 3D structure is rejected if its amino acid at the position under study differs from the amino acid in the input sequence. Hits are sorted according to the sequence identity or E-value of the sequence alignment with the input protein. Structural parameters used to evaluate the effect of amino acid substitution. Structural analysis performed by PolyPhen is based on the use of several structural parameters, as suggested previously ( --). Importantly, although all parameters are reported in the output, only some of them are used in the final decision rules. PolyPhen uses the DSSP database to obtain the following structural parameters for the mapped amino acid residues: secondary structure (according to the DSSP nomenclature); solvent accessible surface area (absolute value in A2); phi --psi dihedral angles. The following values are also calculated by PolyPhen: normalised accessible surface area [the absolute value divided by the maximal area defined as the 99% quantile of surface area distribution for this particular amino acid type in PDB ]; change in accessible surface propensity (knowledge-based hydrophobic 'potentials') resulting from the substitution; change in residue side chain volume (in A3); region of the phi --psi map (Ramachandran map) derived from the dihedral angles ; normalised B factor (temperature factor) for the residue [following Chasman and Adams ]; loss of a hydrogen bond [following Wang and Moult ] according to the HBplus program . By default, the parameters above are calculated for the first hit only. Contacts with 'critical sites', ligands and other polypeptide chains. The presence of specific spatial contacts of a residue may reveal its role in protein function. PolyPhen checks three types of contacts for a variable amino acid residue. First, contacts with ligands (defined as all heteroatoms excluding water and 'non-biological' crystallographic ligands). Second, interactions between subunits of the protein molecule. Technically these are defined as contacts of a polymorphic residue with residues from other polypeptide chains present in the PDB (PQS) file. For this particular type of interaction, it is more advantageous to use the PQS (Protein Quaternary Structure) database rather than PDB, since PQS entries are supposed to provide a more adequate picture of protein quaternary structure architecture. The third type of contact analysed by PolyPhen is represented by contacts with 'critical' residues, where the latter are derived from the sequence annotation. The suggested default threshold for all contacts to be displayed in the output is 6 A. However, a value of 3 A is used in the decision rule. For evaluation of a contact between two residues or between a residue and a ligand molecule, PolyPhen finds the minimal distance amongst all possible between atoms of two residues. By default, contacts are calculated for all hits with structure. This is essential for cases where several structures correspond to one protein but carry different information about complexes with other macromolecules and ligands (see for example figure in ref. ). Prediction rules. PolyPhen uses empirically derived rules to predict that an nsSNP is damaging, i.e. is supposed to affect protein function, or benign, i.e. most likely lacking any phenotypic effect. The rule is based on the analysis of the ability of various structural parameters and profile scores to discriminate between disease mutations and substitutions between human proteins and closely related mammalian orthologues . We introduced two categories of prediction: nsSNPs possibly damaging protein function / structure and nsSNPs probably damaging protein function /structure. The scheme presented in Table successfully predicts similar82% (similar57% for the more stringent set of rules) of disease-causing mutations annotated in SwissProt database 14 and produces similar8% (similar3% for the more stringent set of rules) false positives given the control set of between-species substitutions. We note that many parameters, though computed by the server, were excluded from the decision rule. Due to correlation with other parameters they did not help to increase sensitivity without significant loss of specificity of predictions. Multiple alignment-based profile scores provided the major contribution to the prediction. Therefore, even in the case of proteins with no homologue with known 3D structure, predictions remain reasonably reliable. Figure 1 | PolyPhen query processing flowchart. PolyPhen query processing flowchart. PolyPhen combines information on sequence features, multiple alignment with homologous proteins and structural parameters and contacts to make a prediction of nsSNP effect on protein function. hs_swall is the abbreviation for the Homo sapiens subset of the SWALL database (also known as SPTR, i.e. SwissProt + TrEMBL). Var1,2, two amino acid variants; ACC /ID, SWALL accession number or ID. Figure 2 | Results of the PolyPhen analysis of the HGVbase database v. Results of the PolyPhen analysis of the HGVbase database v.12. hs_swall denotes the Homo sapiens subset of the SWALL database. snp2prot is an in-house command line tool to map HGVbase SNPs onto sequences of known human proteins. 11 152 nsSNPs were identified. 1591 of them have been predicted as possibly damaging for protein structure and function and an additional 1257 as probably damaging. The number of structure-based predictions is much lower compared with the number of sequence-based predictions because structural information was available in only 1026 cases. Table 1 | Rules used by PolyPhen to predict effect of nsSNPs on protein function and structure RESULTS : Retrieval of nsSNPs | HGVbase v.12 , a comprehensive public database with extensive curation, was chosen as a source of SNP data. The database had 983 589 SNP entries, which represented SNPs from various sources. Importantly, SNPs in the database are classified according to reliability. Namely, SNPs confirmed by independent and solid experimental verification are marked as 'Proven', whereas other SNP candidates are marked as 'Suspected'. Version 12 of the database contained 984 093 entries, 983 589 of these being SNPs, while the rest represent other types of genetic variants. Only 14 986 SNPs, however, appeared in the 'Proven' category. We mapped all available SNPs onto known proteins and found 9310 of them to be synonymous and 11 152 non-synonymous, causing amino acid changes in protein sequences. 1276 of these identified nsSNPs were 'Proven'. Only 1026 nsSNPs were mapped to proteins with at least 50% sequence identity to a protein with known 3D structure. The analysis for the rest of the nsSNPs was performed on the basis of multiple alignment information only. The database of these nsSNPs and their analysis using PolyPhen is available at . PolyPhen analysis was only possible for 9165 (82%) of these nsSNPs, as the remainder have been mapped to proteins with no applicable site annotation and no reasonably close homologous sequences available in the SWALL database for multiple alignment or structural analysis. The results of the PolyPhen analysis are presented in Figure . Structural characterisation of nsSNPs | As has been noted by Wang and Moult , most disease mutations and supposedly deleterious nsSNPs affect protein stability rather than functionality. Various structural parameters have been proposed ( --,) to detect the effects of amino acid substitutions. We selected a group of structural parameters and evaluated their impact through a comparison of disease mutations, nsSNPs and substitutions between human proteins and closely related mammalian orthologues [datasets from Sunyaev et al. ]. We also selected three characteristics responsible for functionality: annotation of the site as BINDING, ACT_SITE, LIPID or METAL (SwissProt feature table terms); proximity to an annotated site; proximity to a co-crystallised ligand. The data presented in Table confirm that functionality parameters have a smaller impact on the molecular origin of disease mutations and deleterious nsSNPs than protein stability characteristics. Among the structural characteristics presented in Table , hydrophobic core stability parameters are the best predictors. Interestingly, for all parameters analysed we observed the same pattern in Table . The fraction of SNPs that affect a structural parameter is always much lower than that of disease-causing mutations. At the same time, it is always higher than the corresponding number of substitutions between species. This observation suggests that all effects associated with these structural parameters are responsible for the accumulation of deleterious alleles in the human genome. Disease-causing mutations are subject to very strong selective pressure and are eliminated from the population very quickly. In contrast, slightly deleterious SNPs detected in panels of healthy individuals are supposedly under lower selective pressure and therefore have a much longer persistence time in the population. As suggested by Table , we did not observe any structural feature responsible solely for strong or solely for weak selection, as all parameters display the same pattern. Although many structural parameters can serve as reasonably reliable predictors of the effect of a substitution, a strong correlation within structural parameters and especially between structural parameters and long-term selective pressure signals seen from multiple sequence alignment made exclusion of many parameters from the combined prediction rule necessary. For the set of nsSNPs predicted to be damaging, based on the combined set of rules that incorporate both multiple alignment and structural information (available for these cases), structural parameters worked as predictors in 40% of cases. However, the prediction cannot be made solely at the sequence level in 22% of cases (28% if the 'probably damaging' category only is considered). Protein structural and functional characteristics and selective constraints | As has been shown by systematic studies on cSNP (coding SNPs) discovery ( --), the distribution of nsSNP density over human genes is highly non-uniform. Apart from differences in the coalescent history of loci, this notable difference in the rate of nsSNPs is likely to be caused by variations in selective pressure against deleterious variants. We expected that the difference in selective pressure might be caused by structural properties because the number of sites important for stability or functionality might depend on the protein structure type. Also, extracellular proteins can be expected to have higher stability compared with intracellular ones and this may affect selective constraints. On the other hand, selective pressure may depend on the impact of the gene on the overall fitness of the organism . In order to test whether the above properties of proteins have an effect on the density of nsSNPs (considered for genes with the same number of synonymous SNPs to correct for various sources of bias), we subdivided genes from our database into groups according to the SCOP and GO classifications. Contrary to our expectations, we did not detect a significant correlation of selective pressure against deleterious nsSNPs with secondary structure class, localisation or biological process. This might be because we grouped genes into very large classes and the effect might be detected if a finer classification were considered. Alternatively, we have to conclude that there is no strong impact of these characteristics on the selective constraints. In contrast, molecular function of the protein showed a statistically significant association with the strength of selective pressure (the P value of the chi2 test was 0.009). The functional class showing the highest selective pressure against deleterious nsSNPs is the class of transcription factors. This class displays the greatest departure from the average level of selective constraints. Enzymes are the class of proteins with the lowest selective pressure. The fraction of nsSNPs predicted as damaging by PolyPhen is also highest for enzymes and lowest for transcription factors. This is expected and shows that low selective constraints allow for accumulation of slightly deleterious SNPs. We hypothesise that this observation can be explained in terms of the molecular basis of dominance . Mutations in enzymes are likely to be recessive because the flux in a metabolic pathway undergoes very minor change in response to a decrease in enzyme activity . In contrast, changes in the activity of transcription factors can have a high impact on the transcription level of the regulated genes. Transcription factors listed in the OMIM database are reported to be dominant genes much more frequently than enzymes. However, we should note that the current SNP databases are probably biased towards 'popular' genes, which could have affected our results. More accurate selective pressure studies will be possible in the future with larger datasets arising from large-scale systematic studies. Table 2 | Structural characteristics of disease mutations, nsSNPs and amino acid substitutions between species DISCUSSION : Server | Ideally, the end point of disease gene identification should be functional analysis of the disease-associated allele and an understanding of the molecular mechanism of causation of the disease phenotype. This functional characterisation can be facilitated by the computational analysis provided by our tool. Unlike fully penetrant mutations causing Mendelian diseases, SNPs involved in complex human phenotypes are not a necessary and sufficient condition defining the phenotype but their effect depends on many other genetic and environmental components. In other words, SNPs may comprise risk factors of having a specific phenotype in the statistical sense. Therefore, the effect of a particular SNP on phenotype might be seen only as a frequency difference between individuals that display the phenotype and unaffected controls. Given the very high rate of false associations recently reported, any independent evidence of the impact of the suspected allelic variant should be valued. Sequence and structure analysis of the suspected amino acid variant can increase the confidence of the finding by revealing the structural background of the disease. The PolyPhen server can be used to evaluate whether the reported /identified association can indeed have a functional meaning and therefore is less likely to represent a false positive due to statistical reasons or reasons of inappropriate study design and population choice. Consequently, even if an association of a genomic locus with a particular phenotype is unambiguously demonstrated, it is not always clear that the identified DNA variant has a causative relationship with the disease and that statistical association is not a result of linkage disequilibrium with the true functional variant . In this case the PolyPhen server can be used to distinguish casual from non-casual relationships between a nsSNP and the phenotype of interest. The database of nsSNPs annotated by PolyPhen provides a source of functionally annotated nsSNPs. The collection might be a useful resource for selection of nsSNPs for candidate gene-based association studies. The question of how to choose the set of SNPs to be screened is critical to the success of a study. The major hurdle in any model of association studies is posed by the large number of these SNPs (,). One side of the problem is the limitations of currently available genotyping technologies, which make studies on large SNP sets in large panels of individuals impractical. The other side, however, is of a purely statistical nature and is therefore independent of the technological progress. Multiple test correction in the case of many thousands of SNPs to be analysed makes the detection of otherwise significant allele frequency differences problematical. Possible allelic and non-allelic heterogeneity, epistatic interactions between alleles, low penetrance of the phenotype and complexity of environmental factors involved make the SNP-based detection of disease genes even more difficult . Without any careful pre-selection of SNPs to be screened, unrealistically large panels of individuals might be required to detect association at a reasonable level of statistical significance. Therefore, computational prediction of functional importance can be considered as one of the reasons to prioritise SNPs while looking for an association. Survey | PolyPhen analysis of the nsSNP database confirmed earlier observations ( --,,) that a significant number of human nsSNPs is represented by slightly deleterious alleles. The fraction of nsSNPs predicted to be damaging in the much larger dataset of 9165 nsSNPs is similar to the earlier result. Most predictions were computed based solely on the multiple alignment information, since structural data are available for only a very small fraction of cases. It is important to note that the number of functional nsSNPs predicted for the whole database is likely to be an overestimate due to pollution of the database by erroneous SNP reports, on the one hand, and possible bias of the database towards disease-related allelic variants on the other. To test the impact of these biases on the overall conclusion of the presence of multiple slightly deleterious SNPs in individual human genomes, we compared fractions of nsSNPs predicted to be damaging (both possibly and probably) for HGVbase entries annotated as 'Proven' and 'Suspected'. Additionally, we compared the prediction rate for 'Proven' nsSNPs originating from systematic studies ( --) with the overall prediction rate. The overall prediction rate for the category 'Suspected' nsSNPs was 31.4%, for the category 'Proven' nsSNPs it was 28.9% and for 'Proven' nsSNPs from systematic studies on healthy individuals ( --) it was 27.6%. This shows that inaccuracy and bias of the database data lead to overprediction of the fraction of deleterious nsSNPs. However, the effect of the prediction rate for nsSNPs compared with the species divergence data on a much higher fraction is seen even from the cleanest possible dataset. Similarly, trends observed in Table are the same for any subset of nsSNP data. Our analysis showed that various effects on protein stability are responsible for accumulation of slightly deleterious nsSNPs in human genes. The selection against these variants is likely to depend on the molecular function of proteins rather than on the type of structure or cellular localisation. This can possibly be explained by the relationship between molecular function and mutation dominance. Transcription factors appear to be the group with the highest selective constraints. With the growth of public SNP data and the improvement in the quality of SNP databases, functional analysis of SNPs can possibly play a role in our understanding of the inheritance of complex human phenotypes. Backmatter: PMID- 12202763 TI - Structural insights by molecular dynamics simulations into differential repair efficiency for ethano-A versus etheno-A adducts by the human alkylpurine-DNA N-glycosylase AB - 1,N6-ethenoadenine adducts (epsilonA) are formed by known environmental carcinogens and found to be removed by human alkylpurine-DNA N-glycosylase (APNG). 1,N6-ethanoadenine (EpsilonA) adducts differ from epsilonA by change of a double bond to a single bond in the 5-member exocyclic ring and are formed by chloroethyl nitrosoureas, which are used in cancer therapy. In this work, using purified recombinant human APNG, we show that EpsilonA is a substrate for the enzyme. However, the excision efficiency of EpsilonA was 65-fold lower than that of epsilonA. Molecular dynamics simulation produced similar structural motifs for epsilonA and EpsilonA when incorporated into a DNA duplex, suggesting that there are no specific conformational features in the DNA duplex which can account for the differences in repair efficiency. However, when EpsilonA was modeled into the APNG active site, based on the APNG/epsilonA-DNA crystallographic coordinates, in structures produced by 2 ns molecular dynamics simulation, we observed weakening in the stacking interaction between EpsilonA and aromatic side chains of the key amino acids in the active site. In contrast, the planar epsilonA is better stacked at the enzyme active site. We propose that the observed destabilization of the EpsilonA adduct at the active site, such as reduced stacking interactions, could account for the biochemically observed weaker recognition of EpsilonA by APNG as compared to epsilonA. Keywords: INTRODUCTION : The saturated exocyclic adduct of adenine, 1,N6-ethanoadenine (EpsilonA) (Fig. A), has been identified as one of the products of the reaction of 1,3-bis(2-chloroethyl)nitrosourea (BCNU) with DNA (,). BCNU belongs to the family of therapeutic nitrosourea compounds used in cancer treatment. The ethano adducts in DNA structurally resemble the exocyclic etheno adducts formed from the reaction of the chemical carcinogen vinyl chloride with DNA (,) or by lipid peroxidation . The etheno adducts, particularly 1,N6-ethenoadenine (epsilonA), have been extensively studied biochemically and structurally . It has been shown that this adduct can be efficiently removed from DNA by rodent or human alkylpurine-DNA N-glycosylase (APNG) (also termed alkyladenine DNA glycosylase, AAG) ( --). The mechanism of epsilonA excision by APNG has been proposed based on the 2.1 A crystal structure of an APNG mutant protein (E125Q) complexed to epsilonA-containing DNA . Crystallization of the protein --substrate complex was made possible by substitution of Glu125 with a glutamine residue, which prevents activation of the active site bound water acting as a nucleophile. The authors showed that flipped-out epsilonA has the ability to stack in a stable position between the aromatic side chains in the enzyme active site . The position of the adduct was also stabilized by a key hydrogen bond between the main chain of His136 and N9 of epsilonA, which offered a unique acceptor lone pair essential for hydrolysis of the C1' --N glycosylic bond. The His136 side chain forms hydrogen bond interactions to Tyr157 and the phosphate group of epsilonA (Fig. B). The ethano adducts differ from etheno adducts by the change of a double bond to a single bond in the 5-member exocyclic ring (Fig. A). In this work we have addressed the issue of whether such a small structural change could affect the recognition and repair efficiency of EpsilonA compared to epsilonA by human APNG. Recent work in this laboratory showed that a small structural change in the adduct structure has an effect on DNA glycosylase activity . Addition of a hydroxymethyl group to the C8 position of 3,N4-ethenocytosine (epsilonC) to form 8-(hydroxymethyl)-3,N4-ethenocytosine (8-HM-epsilonC), a product of the reaction with the mutagen/carcinogen glycidaldehyde, reduced the repair efficiency by Escherichia coli mismatch uracil-DNA glycosylase (Mug) by 2.5-fold as compared with that of the structurally related epsilonC. However, molecular dynamics simulation showed similar alignment and hydrogen bonding patterns for both adduct pairs in the 25mer oligomer duplexes used in the biochemical studies . The lower Mug activity toward 8-HM-epsilonC suggests some degree of steric hindrance to the binding or catalytic activity as a result of the hydroxymethyl group on the etheno ring. In this work the repairability by human APNG of EpsilonA, incorporated into a 25mer DNA duplex, was investigated and compared to the repair efficiency of epsilonA by the same enzyme using a DNA glycosylase assay. EpsilonA was found to be a substrate for the human enzyme, but a much weaker one than epsilonA. The observed difference in rate of excision of epsilonA versus EpsilonA adducts was correlated with the structural data obtained by molecular modeling. The availability of crystal data for the APNG enzyme complexed to epsilonA-containing DNA allowed us to use it as a starting point in our molecular modeling. The observation of structural perturbations caused by replacement of epsilonA by EpsilonA in the enzyme active site might have an effect on the substrate preference of epsilonA over EpsilonA. The complementarity between the substrate and enzyme active site should be one of the factors responsible for the catalytic specificity and efficiency of repair. However, a number of other events, such as initial lesion binding/recognition, ease of rotating the damaged base from the DNA ladder and stabilization of the extrahelical conformation, also contribute to the efficiency of repair for a particular adduct. These factors can be influenced by the conformational features of the adduct-containing duplexes. To evaluate the effects of the EpsilonA adduct on the local and global structural features of the DNA duplex we performed simulation of an EpsilonA-T-containing 25mer DNA duplex. These data were compared to the epsilonA-T- and A-T-containing duplexes. Figure 1 | (A) Chemical structures of the epsilonA and EpsilonA adducts. (A) Chemical structures of the epsilonA and EpsilonA adducts. (B) The APNG active site structure showing the stacking between epsilonA and aromatic side chains of Tyr127, His136 and Tyr157. The black dashed lines show hydrogen bonds between epsilonA N9 and His136 N, Tyr157 O4 and His136 Npi and epsilonA OP1 and His136 Ntau. The picture was generated using the atomic coordinates of the crystallized epsilonA-DNA/APNG complex [PDB ID code 1f4r ]. MATERIALS AND METHODS : Oligonucleotides | Synthesis of the EpsilondA phosphoramidite and its site-specific incorporation into oligonucleotides was described by Maruenda et al. . The epsilondA phosphoramidite was purchased from Glen Research (Sterling, VA). Both derivatives were placed in the sixth position from the 5'-end of a 25mer sequence (X): 5'-CCG CTX GCG GGT ACC GAG CTC GAA T-3'. The unmodified 25mer and complementary strands with T opposite the modified base were purchased from Operon Technologies (Alameda, CA). All the oligomers were purified by HPLC and denaturing PAGE. DNA glycosylase assay | The enzymatic assay used to test APNG-mediated cleavage of EpsilonA or epsilonA from oligonucleotides was carried out essentially as previously described (,). Briefly, 25mer oligonucleotides were 5'-end-labeled with [gamma-32P]ATP (specific activity 6000 Ci/mmol, 1 Ci = 37 GBq; Amersham Pharmacia Biotech) and annealed to a complementary strand in a 1:1.5 molar ratio. The standard reactions (10 microl) contained 2 nM 5'-32P-end-labeled oligomer duplex in 10 mM HEPES --KOH, pH 7.4, 100 mM KCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 0.1 mg/ml acetylated BSA and varying amounts of human APNG protein (a gift from Dr Tim O'Connor, Beckman Research Institute, Duarte, CA) . In these reactions, a 5' AP endonuclease, the major human AP endonuclease (HAP1) (a gift from Dr Ian Hickson, Oxford University, Oxford, UK), was added to cleave the apurinic (AP) site resulting from the excision of EpsilonA or epsilonA by APNG protein. The reactions were stopped by adding equal amounts of F/E solution (90% formamide plus 50 mM EDTA) and then heated at 95 --100C for 3 min. Reactions were then resolved by 12% polyacrylamide --8 M urea denaturing PAGE. For band quantitation, a Bio-Rad FX molecular phosphorimager and Quantity One software (v.4.0.1) were used. Molecular modeling | EpsilonA- and epsilonA-containing 25mer DNA duplexes. A set of force field parameters for epsilonA was previously developed using an ab initio quantum mechanical calculation and procedure described in earlier publications from this laboratory (,). The EpsilonA adduct was built by saturating the C7=C8 double bond in the imidazole ring of epsilonA. Atom-centered charges were calculated with the RESP module of AMBER 6.0 using the partial charges obtained by Hartree --Fock calculation using the 6-311G* basis set in the Spartan 5.0 suite (Wavefunction, Inc., Irvine, CA). Prior to the charge calculations, the conformation of EpsilonA was geometry optimized using the 6-31G* basis set. The epsilonA and EpsilonA adducts were incorporated in the sixth position into the 25mer sequence used in biochemical studies (for sequence see Materials and Methods). The topology and coordinate files for the three DNA duplexes (epsilonA-T-DNA, EpsilonA-T-DNA and A-T-DNA, used as a control) were generated with the xLeap module of AMBER 6.0 . Forty-eight Na+ ions were placed around the phosphate groups to neutralize negative charges, and an aqueous environment was represented by a rectangular water box, which provided no less than 10 A of TIP3P water molecules around the solute. Two nanosecond molecular dynamics simulation runs at 310 K, using particle-mesh Ewald (PME) to treat Coulombic interactions and a 2 fs time step, were generated after the system achieved the correct density and volume . EpsilonA-DNA/APNG and epsilonA-DNA/APNG complexes. In this work the high resolution X-ray crystal structure of the epsilonA-DNA/APNG complex (PDB code 1f4r) served as the starting structure. Hydrogens were added using the xLeap module of AMBER 6.0. To generate the EpsilonA-DNA/APNG complex, epsilonA was replaced by the geometry optimized EpsilonA adduct, using Insight II (Biosym/MSI, San Diego, CA). Two sets of topology and coordinate files for the APNG protein complexed to epsilonA-DNA and EpsilonA-DNA were generated using the xLeap module of AMBER 6.0. A rectangular box of TIP3P water molecules was added, providing at least 10 A of explicit solvent around each DNA/protein complex, yielding 9292 water molecules. The complete system consisted of approximately 31 728 atoms and has the initial dimensions 73.439, 70.948 and 76.668 A in the x, y and z directions, respectively. The initial density of the water around the protein was 0.806 g/cm3. Molecular dynamics simulations were carried out using the SANDER module of AMBER 6.0 with a 2 fs time step. SHAKE was applied to all hydrogen atoms and a 10 A cut-off was used for Lennard --Jones interactions. Constant pressure was maintained with isotropic scaling. All long-range electrostatic interactions were handled using the PME method. In the beginning of the simulations, the water box was subjected to a series of equilibration molecular dynamics runs while holding the DNA/APNG complex fixed, and was similar to the procedure used for the DNA duplexes. The equilibration runs began with 1000 steps of minimization followed by 10 ps of simulation, during which the temperature was slowly raised from 0 to 310 K and kept at this temperature for another 50 ps. During the first 30 ps of simulation the water density and pressure converge to the correct values (1.01 g/cm3 and 1 atm, respectively). This was followed by a second set of 1000 steps of minimization and 3 ps of simulation, which were carried out with the restraints on the solute molecule reduced to 25 kcal/mol. Finally, five rounds of 800 steps of conjugate gradient minimization were performed with the positional restraints reduced by 5.0 kcal/mol in each round. The unrestrained molecular dynamics production runs of 2 ns were initiated after the last round of minimization. The final structures representing the conformational family for the DNA/enzyme complexes produced by molecular dynamics simulation were generated by averaging the molecular dynamics trajectories based on root mean square deviation (RMSD) profiles (from 0.4 to 2 ns). Structural analysis and calculations. The molecular dynamics trajectories were processed using the analytical modules of AMBER 6.0 and visually analyzed with the VMD program . Nucleic acid structural parameters were derived using CURVES 5.1 . Production runs for the 25mer DNA duplexes and DNA/APNG complexes were carried out on 64 processors (16 processors per node) using the IBM SP RS/6000 supercomputer available at the National Energy Research Scientific Computing Center, Lawrence Berkeley National Laboratory. The equilibration runs and trajectory analysis were performed on a Silicon Graphics Origin 200 server interfaced with a dual processor Octane workstation. RESULTS : Biochemical assay | We first tested the excision activity of APNG protein towards EpsilonA since this enzyme excises the closely related adduct epsilonA as well as another ethano adduct, N2,3-ethanoguanine (,,). As shown in Figure (left), APNG protein showed a protein-dependent cleavage of a 32P-end-labeled EpsilonA-containing 25mer oligomer duplex (EpsilonA-T). The cleavage products from EpsilonA- and epsilonA-containing oligonucleotides after 5' AP endonuclease treatment were both 32P-labeled 5mers are shown in Figure (arrows). These are the expected products resulting from the 5' hydrolysis by HAP1 of an AP site at the sixth position. However, the extent of EpsilonA excision by APNG was much lower than that of epsilonA excision, as indicated in Figure (left). Excision of EpsilonA from the 25mer DNA duplex by APNG protein as a function of time is shown in Figure , in which a comparison was made between the rate of excision of EpsilonA and epsilonA under the same assay conditions except that the amount of APNG used was different (3 and 0.15 ng for EpsilonA and epsilonA, respectively). In Figure (right) the scanning results were normalized per ng protein in order to compare the rates of these two activities. It is evident that the excision of EpsilonA occurs much more slowly than that of epsilonA, with an similar65-fold difference. Conformation of the EpsilonA- versus epsilonA-containing duplexes | Two nanosecond molecular dynamics calculations were performed for the two 25mer DNA duplexes used in the biochemical studies (epsilonA-T-DNA and EpsilonA-T-DNA) and a corresponding control A-T-DNA (25mer DNA duplex with an unmodified A at the sixth position). The conformational stability was evaluated by calculating RMSD values of each picosecond relative to the coordinates of the initial energy minimized structures for all three DNA duplexes. RMSD values for all atoms, the five central base pairs and the adduct-containing base pair are shown in Figure A and B for epsilonA-T- and EpsilonA-T-DNA, respectively. Based on RMSD values, both structures reached conformational equilibrium after the first 400 ps and showed a plateau for the rest of the simulation. To monitor integrity of the duplex during the simulation we calculated Watson --Crick hydrogen bond distances and percentage occupancy for all base pairs in the duplexes. All hydrogen bonds, including 5'-TA and 3'-GC base pairs flanking the adduct site, were 98 --100% occupied during the entire simulation (data not shown). Terminal bases were not included in hydrogen bond calculations due to known fraying effects, which were also observed in our simulations. No hydrogen bonding was observed in the epsilonA-T and EpsilonA-T mismatches. Top and side views for the T5X6G7/A46T45C44 motifs, where X = A, epsilonA or EpsilonA, are shown in Figure . In both lesion-containing duplexes, compared to the unmodified duplex, the adduct was displaced towards the major groove, while the opposite T remained stacked between A and C bases. Figure shows average values for the inter- and intra-base pair parameters (Fig. A and B, respectively) of the 5 bp for the epsilonA- and EpsilonA-containing duplexes and corresponding control. Average values were calculated over the simulation trajectory. Presence of the adduct had a similar effect on the conformation of the mismatch and neighboring bases in epsilonA-T-DNA and EpsilonA-T-DNA. A positive shear (SHR) value was observed for both epsilonA-T and EpsilonA-T base pairs, indicating the magnitude of displacement of the adduct towards the major groove . Another two intra-base pair parameters affected by the presence of either epsilonA or EpsilonA in the DNA duplex were buckle and propeller twist. Considerable propeller twist (15 --23), compared with the unmodified DNA (<8) was observed for the T5-A46, epsilonA/EpsilonA-T45 and C8-G43 base pairs in the lesion-containing duplexes (Fig. A). Buckling around the lesion site was also larger in magnitude than for the unmodified DNA. Perturbations in inter-base pair parameters, which probably best describe stacking interactions, were similar in both the epsilonA and EpsilonA duplexes. The most noticeable differences from the unmodified duplex were observed for the tilt (TLT), roll (ROL) and twist (TWS) parameters (Fig. B). The high magnitude of TWS for the T5-A46/epsilonA6-T45 and T5-A46/EpsilonA6-T45 base pair steps (58 and 50, respectively) indicated a larger helical twist at these steps compared with the rest of the adduct-containing and unmodified duplexes. The succeeding steps, epsilonA6-T45/G7-C44 and EpsilonA6-T45/G7-C44, showed much smaller TWS values (2 and 4, respectively), characteristic of untwisting of the DNA at the lesion site. In previous modeling work from our laboratory we reported smaller TWS values at the epsilonA-T base pair in a 15mer DNA duplex. Moreover, the magnitude of TWS was sequence-dependent . The curvature of the DNA was calculated using the CURVES 5.1 algorithm and was not affected by presence of the adduct. To avoid a contribution from the highly flexible DNA ends, the terminal base pairs were not included in the curvature measurements. The values for the EpsilonA-T-DNA and epsilonA-T-DNA duplexes were 14 and 11, respectively. The sugar conformation of EpsilonA falls in the C2'-endo conformation, while epsilonA was closer to the C1'-exo range. Both adducts stack in an anti orientation into the DNA helix. The conformation of the epsilonA-T base pair produced by our modeling approach was similar to the conformation of that lesion reported based on NMR data (,). However, in our modeling we observed a slightly bigger shift of epsilonA towards the major groove than was reported by NMR. The displacement of epsilonA along the x-axis towards the major groove was 2.5 A, while EpsilonA was displaced by 2.0 A, compared to unmodified A in the control duplex. The differences in the conformation of the epsilonA-T pair between NMR and modeling can be attributed to sequence-dependent effects . Effect of the EpsilonA adduct on the APNG active site | The availability of crystal data for human APNG complexed to epsilonA-containing DNA allowed us to use this structure as a starting point in our molecular modeling study in which we addressed the question of substrate preference of this enzyme for epsilonA over EpsilonA. Simple superimposition of EpsilonA over epsilonA did not reveal any conformational effects which EpsilonA might have on the active site of APNG enzyme. First, to validate our modeling protocol, we performed 2 ns simulation of the APNG/epsilonA-DNA complex (PDB ID code 1f4r). The analysis of the overall structure and position of the adduct in the active site showed that the averaged minimized structure produced by molecular dynamics simulation deviates minimally from the crystal coordinates. All averaged RMSD values where <2.0 A, with a value of 0.9 +- 0.06 A for the enzyme active site, 1.7 +- 0.17 A for the enzyme, 1.87 +- 0.31 A for the DNA and 1.97 +- 0.25 A for the all-atom RMSD for the entire structure (Fig. A). The largest RMSD fluctuations observed for the DNA duplex bound to enzyme can be explained by the contribution of more flexible DNA ends. All stacking and key hydrogen bond interactions in the active site remained intact during this simulation. The superimposition of the active sites of the crystal structure of epsilonA-DNA/APNG and the epsilonA-DNA/APNG complex produced by molecular dynamics simulation is shown in Figure . Note that molecular dynamics simulation produced more pronounced plane-to-plane stacking between His136 and the imidazole ring of epsilonA than in the crystal structure. The RMSD values for the EpsilonA-DNA/enzyme complex showed a similar profile to that observed for the epsilonA-DNA/enzyme complex and indicated overall conformational stability for the system when epsilonA was replaced by the EpsilonA adduct (Fig. B). Slightly higher averaged RMSD values were observed for the active site of the enzyme complexed with EpsilonA-DNA than the active site of the epsilonA-DNA complex (1.1 +- 0.1 versus 0.9 +- 0.06 A, respectively). Pertinent observations can be drawn from monitoring the RMSD values for the adduct itself during the course of the simulations. Only corresponding atoms between the structures were compared. The EpsilonA adduct showed significantly higher flexibility and larger deviation from the starting position in our simulation than the epsilonA adduct (Fig. C). The average RMSD value for EpsilonA was 0.36 +- 0.13 A, while for the epsilonA adduct it was 0.2 +- 0.13 A. The main conformational feature observed for the APNG active site complexed with EpsilonA-DNA was displacement and almost 70 rotation of the His136 side chain. This created an edge-to-edge packing interaction with EpsilonA, rather than the much more stable face-to-face stacking observed between the planer epsilonA and His136 in the crystal structure . Face-to-face stacking was also supported during our molecular dynamics simulation of the epsilonA-DNA/APNG complex. The change in the stacking interaction between EpsilonA and His136 resulted in a weakening of two hydrogen bonds: between the side chain of His136 and the 5'-phosphate of EpsilonA (epsilonA OP1 --His136 Ntau) and the side chain of His136 and Tyr157 (Tyr157 O4 --His136 Npi). The evolution of these hydrogen bonds over simulation time is shown in Figure . However, the key hydrogen bond between N9 of EpsilonA and NH of His136 remained intact in the EpsilonA-DNA/APNG complex . Figure 2 | (Left) Protein-dependent cleavage of a 25mer oligonucleotide containing either EpsilonA or epsilonA by human APNG protein. (Left) Protein-dependent cleavage of a 25mer oligonucleotide containing either EpsilonA or epsilonA by human APNG protein. Increasing amounts of APNG protein (0.3 --4.8 ng for EpsilonA and 0.01 --0.15 ng for epsilonA) were incubated with 2 nM 32P-end-labeled oligomer substrates for 30 min at 37C. The AP site produced by DNA glycosylase action was further cleaved by adding HAP1 (5 ng), a 5' AP endonuclease, to the reaction mixture. Note that the use of HAP1 alone had no detectable effect on either EpsilonA- or epsilonA-containing templates. (Right) Time-dependent cleavage of a 25mer oligonucleotide containing either EpsilonA or epsilonA. Oligomer duplexes were reacted with 3 (for EA) or 0.15 ng (for epsilonA) APNG protein for varying times at 37C. The scanning data were normalized as nM oligomer substrate cleaved per ng APNG protein. (Inset) Detailed time-dependent response of EA excision by APNG. Figure 3 | Autoradiogram of gel electrophoresis of 5'-32P-labeled oligonucleotides after reaction with human APNG for varying times (0 --60 min). Autoradiogram of gel electrophoresis of 5'-32P-labeled oligonucleotides after reaction with human APNG for varying times (0 --60 min). The amount of APNG used for these reactions was 3 ng for EpsilonA and 0.15 ng for epsilonA excision. For reaction and gel procedure details see Materials and Methods. The arrows show the position of the 5mer cleavage product. M, a 5mer marker with the same sequence as the expected cleavage product. Figure 4 | Time dependence of RMS deviations of the 25mer DNA duplexes containing epsilonA-T (A) and EpsilonA-T lesions (B). Time dependence of RMS deviations of the 25mer DNA duplexes containing epsilonA-T (A) and EpsilonA-T lesions (B). The data is shown for all atoms (black), the lesion (red) and 5 bp with the lesion in the middle (blue) (C4T5epsilonA/EpsilonA6G7C8/G47A46T45C44G43). Both structures reached conformational equilibrium after the first 400 ps. Figure 5 | Top (left) and major groove (right) views for the 3 bp motifs for the A-T-, epsilonA-T- and EpsilonA-T-containing 25mer DNA duplexes produced by 2 ns molecular dynamics simulations. Top (left) and major groove (right) views for the 3 bp motifs for the A-T-, epsilonA-T- and EpsilonA-T-containing 25mer DNA duplexes produced by 2 ns molecular dynamics simulations. EpsilonA and epsilonA adducts are shown in red and A and T bases are shown in blue. Both epsilonA-T and EpsilonA-T base pairs produced similar structural motifs with the adduct displaced toward the major groove and a non-planner alignment of the bases, as compared to the A-T pair. No hydrogen bonds were observed between the bases in the epsilonA-T and EpsilonA-T pairs. Yellow dashed lines show Watson --Crick hydrogen bonds for the A-T base pair. Figure 6 | (A) Average values for the intra-base pair parameters describing the geometry of base pairing for the 5 bp in the A-T-, epsilonA-T- and EpsilonA-T-containing duplexes. (A) Average values for the intra-base pair parameters describing the geometry of base pairing for the 5 bp in the A-T-, epsilonA-T- and EpsilonA-T-containing duplexes. (B) Average values for the inter-base pair parameters describing the stacking interactions for the 4 bp steps in the A-T-, epsilonA-T- and EpsilonA-T-containing duplexes. The tick marks on the x-axis indicate the base pair step. For example, label C4-T5 corresponds to the C4-G47/T5-A46 base pair step. Figure 7 | Time dependence of RMS deviations (RMSD) of the epsilonA-DNA/APNG (A) and EpsilonA-DNA/APNG complexes (B). Time dependence of RMS deviations (RMSD) of the epsilonA-DNA/APNG (A) and EpsilonA-DNA/APNG complexes (B). Black, complex (DNA + enzyme); blue, enzyme alone; green, DNA alone; red, active site. The conformational families produced by molecular dynamics simulation for the epsilonA-DNA/APNG and EpsilonA-DNA/APNG complexes deviate minimally from the crystal coordinates. High RMSD fluctuations for the DNA duplex (green traces) can be explained by the contribution of the more flexible ends. (C) The RMSD values for the epsilonA (blue) and EpsilonA (gray) adducts. Figure 8 | Superimposition of the epsilonA-DNA/APNG active site from the crystal structure (green) (PDB code 1b4r) over the epsilonA-DNA/APNG active site produced by 2 ns molecular dynamics simulation (red). Superimposition of the epsilonA-DNA/APNG active site from the crystal structure (green) (PDB code 1b4r) over the epsilonA-DNA/APNG active site produced by 2 ns molecular dynamics simulation (red). The RMSD between active site conformations is <0.91 A. The yellow dashed lines indicate the key hydrogen bond between the epsilonA adduct and main chain amide of His136 and two hydrogen bonds which stabilize the position of the His136 side chain. All three hydrogen bonds remained intact during molecular dynamics simulation. Figure 9 | APNG active site structure for the epsilonA-DNA/APNG (blue) and EpsilonA-DNA/APNG (yellow) complexes produced by molecular dynamics simulations. APNG active site structure for the epsilonA-DNA/APNG (blue) and EpsilonA-DNA/APNG (yellow) complexes produced by molecular dynamics simulations. The green dashed lines indicate hydrogen bonds between epsilonA N9 and His136 NH in the APNG/epsilonA-DNA complex and EpsilonA N9 and His136 NH in the APNG/EpsilonA complex. The steric clash between the EpsilonA exocyclic ring and His136 side chain (indicated by red arrow) resulted in reduced stacking interactions (edge-to-edge packing between EpsilonA and His136) and destabilized the position of the ethano adduct in the enzyme active site. Figure 10 | Evolution over time of the three hydrogen bond distances in the APNG binding pocket for the epsilonA- (green) and EpsilonA-containing (blue) DNA/APNG complexes. Evolution over time of the three hydrogen bond distances in the APNG binding pocket for the epsilonA- (green) and EpsilonA-containing (blue) DNA/APNG complexes. The epsilonA N9 --His136 NH and EpsilonA N9 --His136 NH hydrogen bonds remained intact during the entire simulation. DISCUSSION : One of the most important steps in DNA base excision repair (BER) is recognition and excision of the damaged base from the DNA ladder by DNA glycosylases. This step is the key determinant of BER activity against a specific lesion. Recent crystallographic studies revealed details of the base excision mechanism of DNA glycosylases, including human APNG, showing that damaged DNA bases are excised by hydrolysis of the C1' --N glycosylic bond. The result of this reaction is a free DNA base and an abasic sugar residue, which is hydrolyzed by an AP endonuclease, followed by DNA synthesis and ligation, which restores the correct DNA sequence (,). The position of the adduct in the enzyme binding pocket is achieved by rotation of the damaged base out of the DNA in order that it can be inserted into the enzyme active site. Correct alignment of the modified base in the enzyme active site is one of the key steps for successful removal of that base from the DNA. Structural data on the enzyme complexed to adduct-containing DNA provide essential information on the interaction between the substrate and enzyme active site. The use of molecular modeling has allowed refinement of the conformation of DNA/enzyme complexes with adduct structures, which were not used in the X-ray crystallography studies. Additionally, structural data on adduct-containing DNA duplexes should provide valuable information on some initial steps of BER. Pronounced structural perturbation around the lesion might be a signal for the DNA repair enzyme to act on the substrate to prevent binding to the adduct-containing DNA motif. The stacking interaction between the adduct and the flanking bases, together with hydrogen bonding with the opposite base, should influence the ability of the modified base to be flipped out from the DNA duplex into the enzyme binding site. The observed conformational features of the adduct-containing duplexes and enzyme active site bound to the modified base should be carefully examined and compared with the biochemical data, thus providing a possible explanation for differential repair by the particular enzyme. In this work we have demonstrated that human APNG recognizes and excises an EpsilonA adduct in a defined oligonucleotide (Figs and ). Previously this enzyme was also found to act on the epsilonA adduct (,,), a structural analog of EpsilonA, although these two adducts are produced by completely unrelated compounds. Human APNG, as well as homologs in cells from eukaryotic and prokaryotic species, represents a family of enzymes with a wide substrate range (for a review see ). This work showed that the substrate range of APNG is still expanding. Human APNG excises epsilonA from DNA with high efficiency (,). We previously reported that epsilonA is even preferred by APNG over 3-methyladenine, after which the enzyme was originally named. The kinetic comparison made in this work between epsilonA and EpsilonA showed that epsilonA is excised much faster than EpsilonA . Such biochemical data prompted us to explore the structural basis for the observed difference. In this work we employed molecular dynamics simulation to provide structural insights on the EpsilonA- and epsilonA-containing 25mer DNA duplexes and the effect of EpsilonA on the APNG active site conformation. Molecular modeling did not reveal any significant conformational features which can distinguish between the epsilonA and EpsilonA adducts when incorporated opposite T in 25mer DNA duplexes. Both duplexes have similar structural motifs around the lesion sites. Both adducts adopted the anti orientation, were displaced towards the major groove and formed a non-planar, sheared base pair with the opposite T. It has been proposed that sheared base pairs can be a structural feature important for recognition by some DNA glycosylases . No hydrogen bonds were observed between the bases in the epsilonA-T and EpsilonA-T pairs. The sugar pucker of the EpsilonA and epsilonA adducts falls in the C2'-endo/C1'-exo range. The smaller twist values observed for both lesions should contribute to unwinding of the DNA upon binding to the repair enzyme. The unwound DNA around the lesion site allows easy access for the repair enzyme to continue further adduct recognition and discrimination . The overall conformation of the epsilonA-T base pair produced by modeling was in general agreement with NMR data on an epsilonA-T-containing 9mer duplex . Saturation of the imidazole ring in the EpsilonA adduct partially reduced the stacking ability of this molecule, as compared to epsilonA, which favors pi --pi stacking interactions with amino acids in the enzyme active site. The extra, non-planer hydrogens at the C7 and C8 positions of EpsilonA, as compared to epsilonA, contribute an additional van der Waals surface area that makes it more difficult to accommodate the adduct between the conformationally constrained Tyr127 and more flexible His136. The replacement of epsilonA by EpsilonA in the APNG active site resulted in an edge-to-edge packing interaction between His136 and EpsilonA. The conformation produced by molecular modeling shows that in order to accommodate EpsilonA in the enzyme active site the active site required a structural rearrangement involving His136. A comparison of the APNG crystal structure with epsilonA-DNA/APNG and abasic pyr-DNA/APNG complexes showed that the Tyr127, Tyr157 and His136 side chains are in the same orientation, suggesting that the conformation of the APNG active site is predetermined and not influenced by adduct binding . The extra energy required to overcome the steric clash between the aromatic side chain of His136 and the 7,8-dihydro-imidazole ring of EpsilonA should prevent an easy fit of EpsilonA adducts into the APNG active site. However, the mechanism of EpsilonA excision by APNG may be similar to that reported for epsilonA (,). Both epsilonA and EpsilonA have a lone pair acceptor nitrogen (N9), which is unique to the alkylated base. The position of Glu125 is not changed in the presence of EpsilonA and this residue should be able to deprotonate the active site bound water for nucleophilic attack on the C1' sugar carbon of EpsilonA. The hydroxide nucleophile will be stabilized by Arg182, the position of which also remains unchanged in the EpsilonA-DNA/APNG complex as compared to the epsilonA-DNA/APNG complex. Based on the conformations of the epsilonA- and EpsilonA-containing duplexes, it can be suggested that the glycosylase does not distinguish between these adducts based on local DNA distortion. Similar structural motifs for these adducts serve as an initial signal for the enzyme to test the base by forcing Tyr162 into the helix and displacing the modified base into the enzyme active pocket. The enzyme active pocket requires tight interaction between the adduct and the neighboring amino acids and thus is sensitive to the adduct structure and conformation. It was shown that an APNG mutant (H136Q), engineered to eliminate aromatic stacking interactions with epsilonA, has very low repair efficiency (,). Moreover, it has been proposed that base stacking interactions between the damaged bases and the aromatic side chains of amino acids in the active site may provide a basis for recognition and excision by E.coli 3-methyladenine DNA glycosylase II (,), which also excises EpsilonA (B.Hang, A.B.Guliaev and B.Singer, manuscript in preparation). The observed destabilization of the EpsilonA adduct in the active site, such as the weaker stacking interaction of the adduct with the aromatic side chains of His136, is likely to contribute to the lower efficiency of repair and explain why this adduct is a less preferable substrate than epsilonA for human APNG. Backmatter: PMID- 12202772 TI - Mutations altering the cleavage specificity of a homing endonuclease AB - The homing endonuclease I-CreI recognizes and cleaves a particular 22 bp DNA sequence. The crystal structure of I-CreI bound to homing site DNA has previously been determined, leading to a number of predictions about specific protein --DNA contacts. We test these predictions by analyzing a set of endonuclease mutants and a complementary set of homing site mutants. We find evidence that all structurally predicted I-CreI /DNA contacts contribute to DNA recognition and show that these contacts differ greatly in terms of their relative importance. We also describe the isolation of a collection of altered specificity I-CreI derivatives. The in vitro DNA-binding and cleavage properties of two such endonucleases demonstrate that our genetic approach is effective in identifying homing endonucleases that recognize and cleave novel target sequences. Keywords: INTRODUCTION : Homing endonucleases are remarkable in their ability to function autonomously in recognizing and cleaving specific long DNA sequences. These enzymes are encoded by sequences found within self-splicing RNA introns or self-splicing protein introns ('inteins'). Their target sequences ('homing sites') are 14 --40 bp in length and are comprised of DNA sequences from each of the two intron flanks. Intact homing sites are thus present only in the 'intronless' alleles of their corresponding host genes. In nature, homing endonuclease-cleaved DNA can be repaired via homology-based double-strand break repair, resulting in unidirectional transfer of the homing endonuclease-containing intron, along with disruption of the homing site sequence. This mechanism allows for both same species and horizontal intron transfer events (,). Homing endonucleases have been introduced into bacterial, plant, insect and mammalian cells where they have been shown to recognize and cleave their specific targets ( --). Such homing endonuclease-induced DNA double-strand breaks may be lethal, mutagenic, or repaired by homologous recombination. The major limitation of this approach is that it depends on the prior introduction of homing endonuclease target sequences or on the fortuitous existence of host sequences compatible with particular homing endonucleases. The ability to engineer new homing endonucleases that target DNA sequences of interest could enable new gene replacement and inactivation strategies in a wide variety of organisms. Homing endonucleases appear to have evolved independently on multiple occasions, as sequence comparisons reveal four distinct families of enzymes . The homing endonuclease I-CreI is a member of the largest family, the LAGLIDADG endonucleases (,). I-CreI is a 163 amino acid protein which functions as a homodimer, with the conserved LAGLIDADG domain forming the dimerization interface as well as contributing to the catalytic core . I-CreI is specified by sequences in a self-splicing intron in the 23S rRNA gene from the choloroplast genome of Chlamydomonas reinhardtii . The structure of I-CreI bound to homing site DNA has been solved, providing valuable insight into the mechanism by which I-CreI achieves its high DNA target specificity (,). The I-CreI homing site is a semi-palindromic 22 bp sequence, with 7 of 11 bp identical in each half-site . Since I-CreI functions as a homodimer, it is not surprising that the majority of predicted protein --DNA contacts occur at these symmetrical positions . Of the nine amino acids predicted to directly contact DNA, seven are thought to interact with nucleotides at symmetrical positions. Of the seven symmetrical positions in each half-site, six are predicted to interact with one or more I-CreI amino acids. We are interested in better defining the interactions involved in DNA recognition by I-CreI. We have constructed endonuclease mutants in which each of the nine amino acids predicted to directly contact homing site bases and a tenth amino acid predicted to participate in a water-mediated interaction are converted to alanines. We have also constructed a series of homing site mutants where palindromic positions have been systematically altered. Each of our mutants has been examined for function in vivo in a series of Escherichia coli-based assays. Our mutants display a wide range of phenotypes, from nearly wild-type to completely inactive. We have thus determined the relative importance of structurally predicted I-CreI /DNA contacts, defining which contacts are most important for this highly specific protein --DNA interaction. In the course of these studies we determined that our collection of mutant homing sites included a number with decreased affinities for wild-type I-CreI. We have begun to systematically search for I-CreI derivatives that recognize these sites better than does the wild-type enzyme and report here on the isolation of a collection of such altered specificity mutants. Two such mutants have been purified and shown to display altered DNA recognition properties in vitro, demonstrating that our genetic approach is effective in identifying homing endonucleases that recognize and cleave novel target sequences. Figure 1 | Predicted I-CreI /DNA contacts. Predicted I-CreI /DNA contacts. The 22 bp I-CreI homing site is shown, with palindromic base pairs in bold. Cleavage positions are indicated by vertical lines between bases 13 and 14 on the top strand and 9 and 10 on the bottom. I-CreI amino acids are indicated by their single letter abbreviations, with residues from one monomer of the homodimer in bold and those from the other having primes. Solid lines indicate direct hydrogen bond interactions, with double lines indicating two such interactions. Dashed lines indicate water-mediated interactions. The figure is based upon the structure of Jurica et al. . MATERIALS AND METHODS : Bacterial strains and media | The following E.coli K-12 strains were used in this study: CC118, [araD139 Delta(ara,leu)7697 Delta(lac)X74 phoADelta20 galE galK thi rpE rpoB argE(am) recA1], CC136 [F128 (lacIq) /Delta(lac-pro) ara nalA argE(am) thi rpoB] and MC1000 [araD139 Delta(ara,leu)7697 Delta(lac)X74 galE galK thi rpsL]. Standard growth media were used . Where required, they were supplemented with kanamycin (50 microg /ml), ampicillin (200 microg /ml), chloramphenicol (30 microg /ml), tetracycline (10 microg /ml), X-Gal (40 microg /ml) or arabinose (0.2 or 0.04% w /v). Plasmid constructs | Plasmid pBR-O-Xho is a tetracycline-resistant derivative of pBR322 (New England Biolabs, Beverly, MA) used for the construction of F'o-cre alleles. It contains 310 bp of lac operon DNA, beginning at the 3' end of lacI and extending into the 5' end of lacZ, with 25 bp of lacO sequence replaced by a XhoI cleavage site. It was created using overlap extension PCR and cloned into the unique PvuI and EcoRI sites of pBR322. The wild-type and the symmetrically mutated homing site derivative of pBR-O-Xho were created by annealing appropriate complementary 26 base oligonucleotides and ligating into XhoI-cleaved pBR-O-Xho. All mutants were sequence verified by the DNA Sequencing Facility at Rancho Santa Ana Botanic Gardens (Claremont, CA). I-CreI homing site-containing plasmid pKS155 and I-CreI endonuclease-encoding plasmid pA-E have been described previously . I-CreI alanine substitution mutants were generated by overlap extension PCR . The same procedure was used to create silent mutations specifying a unique SalI restriction site at I-CreI codons 17 and 18, and a unique HindIII restriction site at I-CreI codons 40 and 41. A cassette mutagenesis strategy utilizing these two restriction sites was employed to create all non-alanine substitutions at codons 32 and 33. All mutants were sequence verified. Standard recombinant DNA procedures were used for the isolation and restriction analysis of plasmid DNAs . Bacterial manipulations | Reciprocal recombination was used to transfer the wild-type and symmetrically mutated homing sites from pBR-O-Xho to the F' factor F128 of E.coli strain CC136 to create F'o-cre alleles. This was done in two steps, by first selecting tetracycine-resistant co-integrates after a mating with MC1000 and then screening for tetracycine-sensitive F' factors after a mating with CC118. F'o-cre alleles were then transferred back to MC1000 for phenotypic assays. After transfer to F' factors, all mutant homing sites were amplified by PCR and sequence verified. In vivo I-CreI assays | F'cre-kan assays. MC1000 cells containing F'cre-kan were made competent by CaCl2 treatment and transformed by pAE encoding wild-type and mutant derivatives of I-CreI. Trans formants were selected on media containing (i) chloramphenicol only; (ii) chloramphenicol and kanamycin; (iii) chloramphenicol and 0.2% arabinose; (iv) chloramphenicol, kanamycin and 0.2% arabinose. Under these conditions wild-type I-CreI expressed from pAE yielded healthy colonies on media supplemented with chloramphenicol only, tiny irregular colonies on media supplemented with chloramphenicol and 0.2% arabinose, and no colonies on either media containing kanamycin. Mutants that yielded colonies on all four types of media were classified as inactive. A subset of these mutants that displayed no evidence of toxicity on arabinose-containing media were classified as non-toxic. Mutants that failed to form colonies only on media containing chloramphenicol, kanamycin and 0.2% arabinose were classified as partially active. A number of mutants were extremely toxic in that they failed to form the expected numbers of colonies on media supplemented with chloramphenicol and 0.2% arabinose. These mutants were tested for activity on media supplemented with 0.04% arabinose. F'o-cre assays. MC1000 cells containing F'o-cre were made competent by CaCl2 treatment and transformed by pAE encoding wild-type and mutant derivatives of I-CreI. Transformants were selected on media containing chloramphenicol and X-Gal. Photographs of sectored colonies were taken after colonies developed for 24 --48 h at 37C. In vitro I-CreI assays | Binding assays. The I-CreI mutants S32K and Y33C were constructed in pI-CreI using Stratagene's Quikchange site-directed mutagenesis kit. The two mutants and the wild-type protein were overexpressed in E.coli BL21[DE3] and purified as previously described . Gel mobility shift assays were based on retardation of the electrophoretic mobility of 32P kinase-labeled DNA when bound by I-CreI (,). Appropriate 47 base oligonucleotides were annealed and end-labeled with 32P. Endonuclease and 2.5 fmol labeled double-stranded (ds)DNA was incubated for 30 min at room temperature in 20 mM Tris pH 9.0, 10 mM CaCl2, 1 mM DTT, 50 microg /ml non-specific competitor DNA and 3% glycerol (Mg2+ is required for I-CreI cleavage activity; substitution of Ca2+ for Mg2+ permits DNA binding but not cleavage). Samples were electrophoresed on 10% non-denaturing polyacrylamide gels containing 1 mM CaCl2 at 200 V at 4C. Gels were imaged in a Storm Phosphorimager 840 (Molecular Dynamics, Sunnyvale, CA) and the intensity of the free and bound DNA bands were quantified using ImageQuant software (Molecular Dynamics). The Kd values of the I-CreI /DNA complexes were defined as the concentration of I-CreI at which 50% of the DNA was shifted into a complex with slower mobility and represent averages of three independent assays. Cleavage assays. The abilities of mutant I-CreI homing sites (present on pBR-O-Xho) to serve as substrates for purified I-CreI were determined as previously described , with one exception: prior to exposure to I-CreI, pBR-O-Xho derivatives containing mutant homing sites were linearized with NdeI. For site competition assays, pKS155 (containing a wild-type I-CreI homing site) was linearized with XmnI. Assays were performed on 100 ng of each linearized plasmid in 10 microl of 20 mM Tris pH 9.0, 10 mM MgCl2, 1 mM DTT and 50 microg /ml BSA. Minimal enzyme amounts sufficient to achieve complete digestion of each substrate were determined empirically (see Fig. , 1x samples) and used to initiate a series of 2-fold dilutions. Reactions took place for 30 min at 37C and were terminated by placing digestions on ice, followed by addition of loading buffer containing SDS (to 0.5% w /v) and electrophoresis through 1.2% agarose gels in 1x TBE buffer . Figure 5 | In vitro cleavage competition assays. In vitro cleavage competition assays. Wild-type I-CreI and endonuclease mutants C33 and K32 were exposed to linearized plasmids containing wild-type and mutant homing sites. The numbers above each gel indicate relative amounts of endonuclease, with 0 corresponding to no enzyme, 1 corresponding to the minimal enzyme concentration sufficient to completely digest each plasmid and subsequent fractions reflecting serial 2-fold enzyme dilutions. Homing site identities are indicated at the sides of each photo and endonuclease identities below. RESULTS : We have used a series of E.coli-based genetic assays to analyze DNA recognition by I-CreI. Briefly, I-CreI was expressed from an arabinose-inducible promoter on the pACYC184-based plasmid pA-E. Two F' factor borne I-CreI homing site constructs were used as targets. The first target has the homing site located adjacent to a kanamycin resistance gene (the 'F'cre-kan assay', Fig. A). The second target has the homing site located in a lac operon in place of lacO sequences (the 'F'o-cre assay', Fig. B). In the absence of arabinose induction there was sufficient expression of wild-type I-CreI from pA-E to cleave each homing site, as evidenced by loss of kanamycin resistance in the first assay and conversion of cells from lacZ+ to lacZ -- in the second. Increased expression of I-CreI and I-CreI mutants can be toxic to E.coli, as evidenced by the presence of small translucent colonies or by a complete absence of colonies on media containing arabinose. Partially active I-CreI mutants could be identified in each of the two assays. In the F'cre-kan assay such mutants eliminated kanamycin resistance only when expressed at high levels in the presence of arabinose . In the F'o-cre assay partially active mutants gave rise to sectored colonies, with the degree of sectoring indicative of the amount of endonuclease activity (Table and Fig. ). Note that there was a small amount of spontaneous sectoring in control experiments (Fig. , endo --), indicating that the F' lac allele was slightly unstable. As this instability may be heightened by I-CreI-induced toxicity, all F'o-cre assays were carried out in the absence of arabinose induction. Alanine substitutions | According to the I-CreI --DNA co-crystal structure, nine amino acids (Q26, K28, N30, S32, Y33, Q38, Q44, R68 and R70) make potentially important direct contacts with specific homing site bases, and a tenth (T140) participates in water-mediated hydrogen bonding . To determine the relative importance of these 10 residues in DNA recognition, alanine substitutions at these positions were constructed and examined in our E.coli-based assays. The resulting mutants fell into four distinct phenotypic classes . The first two classes of mutants retained the most I-CreI activity . The 'fully active' mutants S32A and T140A behaved exactly like the wild-type in each of the assays: kanamycin resistance was lost in the absence of arabinose induction in the F'cre-kan assay, and cells became completely lacZ -- in the F'o-cre assay. Thus, the S32 and T140 contacts appear least important for homing site recognition. The 'partially active' mutants N30A, Q38A and Q44A displayed intermediate levels of activity in each assay. In the F'cre-kan assay each mutant required arabinose induction in order to eliminate kanamycin resistance. In the F'o-cre assay each gave rise to sectored colonies , indicating that each of these mutants is somewhat active in the absence of induction. Mutants in the remaining two classes failed to demonstrate homing site cleavage in either assay . These mutants were distinguishable based upon their different responses to endonuclease induction. For mutants K28A and R70A, growth in the presence of 0.2% arabinose resulted in relatively healthy-looking colonies. These two mutants were classified as 'non-toxic'. The remaining 'inactive' mutants (Q26A, Y33A and R68A), like the 'fully active' and 'partially active' classes of mutants described above, produced small translucent colonies when grown in the presence of 0.2% arabinose. Surprisingly, the Y33A mutant was even more toxic than wild-type I-CreI in that exposure to 0.2% arabinose resulted in significantly fewer colonies than expected. Due to this extreme toxicity, the Y33A mutant was assayed in the presence of 0.04% arabinose, where it failed to eliminate kanamycin resistance. The toxicity observed upon induction of I-CreI and most of the I-CreI derivatives described here may reflect direct interactions with E.coli chromosomal DNA (see Discussion). Two of the mutants that displayed significant residual activity, T140A and Q44A, are altered at positions predicted to interact with the same homing site bases (positions 8 /15, Fig. ). In the F'o-cre assay the T140A mutant generated all completely white colonies, whereas the Q44A mutant produced highly sectored colonies . It is possible that each of these mutants retains significant activity as a result of the remaining contacts between the other residue, either Q44 or T140, and the homing site bases at these positions. To test this idea, a mutant containing both the T140A and the Q44A mutations was examined . In the F'o-cre assay this mutant produced colonies that were significantly less sectored than those produced by the Q44A mutant , implying that the T140 contact does contribute to site recognition. Position 33 substitutions | The tyrosine at I-CreI position 33 is among the most conserved residues between I-CreI and a group of closely related endonucleases . Having shown that an alanine substitution at this position significantly alters I-CreI activity, we decided to examine the consequences of other amino acid substitutions. A cassette mutagenesis strategy was used to generate each of the other 18 possible amino acid substitutions and the resulting I-CreI derivatives were examined for activity in our assays . One mutant (Y33F) behaved like wild-type I-CreI in each of our assays. This demonstrates that the hydroxyl group present on Y33 is not required for efficient homing site recognition. Three other mutants (Y33C, Y33N and Y33R) displayed intermediate levels of activity in each assay (Table and Fig. for Y33C). Of this latter group, Y33N and Y33R mutants were more toxic than wild-type I-CreI in that they failed to form any colonies on 0.2% arabinose-containing media. However, these mutants did grow on 0.04% arabinose media and eliminated kanamycin resistance in the F'cre-kan assay under these conditions. The remaining substitutions at position 33 each displayed less affinity for the wild-type I-CreI site . Eight mutants (Y33G, Y33I, Y33K, Y33L, Y33P, Y33Q, Y33S and Y33T) failed to demonstrate homing site cleavage in either assay but did retain significant toxicity in the presence of arabinose. Five other mutants (Y33D, Y33E, Y33M, Y33V and Y33W) failed to display either homing site cleavage or toxicity. The final mutant (Y33H) displayed an interesting combination of phenotypes. Like the Y33N and Y33R mutants, Y33H yielded sectored colonies in the F'o-cre assay and failed to yield colonies in the presence of 0.2% arabinose . However, unlike the other two mutants, Y33H retained kanamycin resistance when grown on 0.04% arabinose media. This is the only mutant analyzed with seemingly contradictory behavior in our two assays. It is interesting that even in the absence of arabinose induction Y33H-containing colonies displayed irregular morphologies, indicative of toxicity . It may be that this 'hypertoxic' phenotype is responsible for the sectoring observed in the F'o-cre assay. Symmetrical homing site mutants | Site recognition by I-CreI can be disrupted either by mutations that alter I-CreI (see above) or by mutations that alter the DNA substrate. We have examined a series of mutations altering the seven symmetrical positions in the I-CreI homing site. To ensure that each monomer in an I-CreI homodimer is presented with the same potential contacts, we have restricted the present analysis to symmetrical mutations, where each member of a symmetrical pair is altered in the same fashion. For example, a mutant site with the 5' nucleotide on the top strand having a C->T transition would contain the same mutation on the bottom strand. Thus, for each of the seven symmetrical positions in the I-CreI homing site , three such mutations are possible. Each of these 21 homing site mutations have been made, transferred into our F'o-cre assay system and examined for the ability to serve as substrates for cleavage by wild-type I-CreI. Fifteen of the 21 mutant homing sites were completely resistant to cleavage by I-CreI in our in vivo assays . These 15 mutants correspond to all possible combinations at five of the seven symmetrical homing site positions (Fig. , positions 2 /21, 3 /20, 7 /16, 8 /15 and 9 /14). Thus, five of the seven symmetrical positions in each half-site appear to be essential for efficient site recognition in vivo. Mutants altered at the remaining two positions (Fig. , positions 1 /22 and 4 /19) displayed varying degrees of I-CreI sensitivity. For each position, one of three mutant homing sites acts as an excellent substrate for I-CreI, as evidenced by completely white colonies in the F'o-cre assay . The other two mutant sites at each position display sectored colonies upon exposure to wild-type I-CreI (Fig. , first row). The degree of sectoring differs for these four mutant sites, with site G4 C19 containing the most lacZ -- cells per colony and site G1 C22 the least . Presumably, the relative number of lacZ -- cells per colony is directly proportional to the ability of a mutant site to be cleaved by I-CreI. The fact that mutants altered at positions 1 /22 and 4 /19 are cleaved more efficiently than mutants altered at the other five symmetrical positions fits nicely with the structural data and with our endonuclease mutant data. The outermost position (1 /22) is predicted to interact with S32, which when changed to an alanine behaved exactly like the wild-type (see above). The other position (4 /19) is the lone symmetrical position without any predicted endonuclease contacts . Each of the 21 symmetrical mutant homing sites was also examined for cleavage by I-CreI in vitro (data not shown). Nineteen of the 21 sites displayed partial to complete cleavage. Only two sites (A9 T14 and G9 C14) were completely I-CreI resistant. Thus, 13 of the 15 symmetrical mutant sites that were resistant to I-CreI cleavage in vivo were sensitive in vitro. These results are consistent with previous work, where eight of nine other mutant homing sites selected as being I-CreI-resistant in vivo were sensitive in vitro . S32 and T140 contribute to DNA recognition | The phenotypes of the S32A and T140A mutants were indistinguishable from that of wild-type I-CreI in each of our assays . This could indicate that S32 and T140 are not involved in DNA recognition or that contributions made by these residues are too subtle to be detected by these assays. If the latter is true, then mutant homing sites that are inefficiently cleaved by wild-type I-CreI may display a heightened requirement for the presence of these contacts. To test this prediction, the S32A and T140A mutants were examined for the ability to cleave the four mutant homing sites that resulted in sectored colonies when exposed to wild-type I-CreI. The results of these assays are presented in Figure . The T140A mutant appeared to be significantly less efficient than wild-type I-CreI at cleaving the four mutant homing sites, as evidenced by a greater proportion of lacZ+ cells per colony (Fig. , compare rows one and three). The S32A mutant also displayed a significantly greater proportion of lacZ+ cells per colony than wild-type with the C4 G19 and G4 C19 mutant homing sites. Interestingly, the difference between wild-type I-CreI and S32A was much more subtle with the A1 T22 and G1 C22 mutant homing sites. This makes sense given that S32 is predicted to contact the two outermost positions which are altered in these mutants. Thus, contacts at these positions are predicted to be lost for both the wild-type endonuclease and the S32A mutant. Taken together, these data imply that the S32 and T140 contacts do in fact contribute to homing site recognition. I-CreI derivatives with novel cleavage specificities | Altered specificity derivatives of I-CreI would be expected to involve a combination of existing and new protein --DNA contacts . The I-CreI-resistant mutant homing sites described above provide us with the opportunity to screen for such new contacts. As each of these mutant sites differs from the wild-type homing site in a symmetrical fashion, any new contacts present in one monomer of an I-CreI dimer would be present in the other as well. Since we know which I-CreI amino acid(s) is predicted to interact at a particular homing site position, it is possible to alter specific contact residues and determine whether resulting mutants have increased affinity for corresponding homing site mutants in the F'o-cre assay. The above strategy was used to study the interaction between I-CreI position 33 and homing site bases 2 and 21. Each of the 19 I-CreI derivatives altered at position 33 was examined for activity in the F'o-cre assay against the three mutant homing sites altered at these bases. Mutants with increased affinities were found for two of the three mutant homing sites . The G2 C21 homing site was efficiently cleaved by mutants containing arginine or histidine at position 33, as evidenced by completely white colonies. Similarly, white colonies resulted when the T2 A21 homing site was assayed against an I-CreI derivative containing cysteine at position 33. This latter mutant homing site also resulted in sectored colonies when tested against I-CreI derivatives with leucine, serine or threonine at position 33. The C2 G21 homing site was not efficiently cleaved by any derivative. It is worth noting that each of the I-CreI derivatives that displayed an increased affinity for a mutant homing site also displayed a decreased affinity for the wild-type homing site (Table and Fig. ). We have begun a similar analysis of the interaction between homing site bases 1 and 22 and I-CreI amino acid 32. I-CreI derivatives containing arginine and lysine at position 32 have been created and examined in our assays . Each cleaves the G1 C22 homing site better than does wild-type I-CreI, as evidenced by a greater proportion of lacZ -- cells per colony . When tested against the wild-type homing site and the A1 T22 homing site, each displayed a greater proportion of lacZ+ cells per colony than did wild-type I-CreI, indicating a lower affinity for these two homing sites (see S32K, Figs and ). Interestingly, the T1 A22 homing site is efficiently cleaved by wild-type I-CreI as well as by the S32K and S32R mutants, as all colonies are completely white in each case. Protein was isolated from mutants Y33C and S32K and examined in vitro. To examine DNA cleavage, the endonucleases were exposed to two linear plasmid substrates, one with the wild-type homing site and one with a mutant homing site. Wild-type I-CreI demonstrated a clear preference for the wild-type homing site over the T2 A21 site over a wide range of endonuclease concentrations (Fig. , top left; note the differences in substrate cleavage in lanes 1 /2, 1 /4 and 1 /8, as well as the differences in product appearance in the remaining lanes). The C33 endonuclease displayed a reciprocal pattern, indicating a strong preference for the mutant homing site (Fig. , bottom left). DNA binding studies performed on oligonucleotide substrates confirmed that wild-type I-CreI and the C33 endonuclease each have greater affinities for their respective target sites . Most importantly, the C33 endonuclease displayed a 12-fold lower Kd for the T2 A21 site than wild-type I-CreI, demonstrating a significantly enhanced ability to bind this mutant site. Wild-type I-CreI also demonstrated a clear preference for the wild-type homing site over the G1 C22 site (Fig. , top right). However, this preference was less dramatic than that over the T2 A21 site in that differences were apparent over a narrower range of endonuclease concentrations. The K32 endonuclease displayed no site preference in that both mutant and wild-type substrates were cleaved with similar efficiencies (Fig. , bottom right). Kd measurements confirmed these observations, with wild-type I-CreI displaying a 7-fold lower Kd for the wild-type homing site and the K32 endonuclease displaying virtually identical Kd values for each site . Figure 2 | Escherichia coli-based assays for I-CreI function. Escherichia coli-based assays for I-CreI function. (A) The F'cre-kan assay. Plasmid pA-E contains a chloramphenicol resistance gene and expresses I-CreI from an arabinose-inducible promoter. F'cre-kan contains a wild-type I-CreI homing site adjacent to a kanamycin resistance gene. Introduction of pA-E into F'cre-kan-containing cells results in recipient cells being converted from kanamycin-resistant to kanamycin-sensitive. (B) The F'o-cre assay. F'o-cre contains an I-CreI homing site in place of lacO sequences. Introduction of pA-E into F'o-cre-containing cells results in recipient cells being converted from lacZ+ to lacZ --. Figure 3 | Cleavage of the wild-type I-CreI homing site by partially active I-CreI derivatives. Cleavage of the wild-type I-CreI homing site by partially active I-CreI derivatives. The F'o-cre assay with recipients containing F'o-cre with wild-type I-CreI homing sites. The identities of endonucleases are indicated above each panel, with the 'endo --' control corresponding to pACYC184. Figure 4 | Cleavage of mutant homing sites by various I-CreI derivatives. Cleavage of mutant homing sites by various I-CreI derivatives. The F'o-cre assay with recipients containing F'o-cre with symmetrical mutant homing sites, as indicated above each set of panels. The identities of endonucleases are indicated next to each set of panels. Table 1 | I-CreI mutant phenotypes Table 2 | Symmetrical homing site mutants Table 3 | Altered specificity I-CreI derivatives Table 4 | In vitro binding assays DISCUSSION : Determinants of DNA recognition by I-CreI | A great deal of structural data on homing endonucleases has become available in the past few years (,,, --). Such studies provide important clues as to how these enzymes function to recognize and cleave specific long DNA sequences. However, there are limitations as to what can be concluded from crystallographic studies alone. For example, although the structure of I-CreI bound to homing site DNA reveals which endonuclease amino acids may be sharing hydrogen bonds with homing site bases, the structure reveals little information about the relative importance of these putative hydrogen bonds. Genetic studies can directly address such functional questions. The majority of genetic work on homing endonucleases has involved examining the DNA-binding and cleavage properties of purified mutant proteins ( --). While valuable insights have been obtained from this approach, it may be problematic to extrapolate in vivo DNA recognition behavior from such in vitro studies. For example, I-CreI has been shown to cleave variant homing sites with up to 10 substitutions in vitro, implying that many positions are non-essential (,,). However, we have identified a large collection of 1 and 2 bp homing site substitutions that are I-CreI-resistant in our E.coli-based assay systems (; this work). As many of these mutant sites are efficiently cleaved in vitro, it is clear that our in vivo assays are much more sensitive to relatively small mutational changes. Such in vivo assays more closely mimic the natural situation where homing endonucleases must find and cleave particular DNA sequences in complex genomes. We have previously described the F'cre-kan assay and used it to isolate loss of function endonuclease and homing site mutants . Here we have described a second E.coli-based assay and shown that the results from it correlate quite well with those from the previous assay . The primary advantage of the F'o-cre assay (Fig. B) is that it is capable of identifying both gain and loss of function mutants in a single step. The assay also reveals mutants with intermediate phenotypes in that such mutants yield sectored colonies. We interpret such colonies as resulting from the cleavage of some, though not all, target sites in the initially transformed cell. Subsequent segregation of cleaved from intact F' factors, and later cleavage and segregation events in the course of colony growth, would produce the observed sectored colonies. An essentially identical explanation has been proposed to explain mixed colonies resulting from cleavage of an E.coli chromosomal site by the homing endonuclease I-SceI . To examine DNA recognition by I-CreI we have systematically altered the I-CreI amino acids predicted to make specific DNA contacts, generated as a set of corresponding homing site mutants, and examined the functional consequences of these mutations in our E.coli-based assays. We find evidence that all structurally predicted I-CreI --DNA contacts do contribute to recognition, although they differ greatly in their relative importance. Endonuclease mutants. Two I-CreI amino acids (S32 and T140) appear to be relatively unimportant for DNA recognition. Mutants with alanine substitutions at these positions were virtually indistinguishable from wild-type I-CreI in our assay systems. When three other amino acids (N30, Q38 and Q44) were converted to alanines the resulting mutants retained significant I-CreI function. It is interesting that, with the exception of S32, each of these five residues interacts with a homing site position predicted to contact two amino acids . The Q44A /T140A double mutant, altered at two residues predicted to contact the same base pair, was significantly more defective than was either single mutant, consistent with the notion that each of these amino acids contributes to DNA recognition. Mutants with alanine substitutions at the other five positions displayed less affinity for the wild-type homing site, indicating that each of these native contacts is more important for site recognition. Two mutants in this group alter residues (R68 and R70) predicted from the I-CreI /DNA co-crystal structure to each make two base-specific contacts in each half-site . Two other residues in this group (Q26 and K28) are the only amino acids predicted to make contacts with non-palindromic homing site bases . It is interesting that these base contacts that differ in each DNA half-site are among the most important in the I-CreI /DNA complex. Our analysis of all possible amino acid substitutions at I-CreI position 33 has identified phenylalanine as the only amino acid capable of functioning as well as the native tyrosine when tested against the wild-type homing site. The ability of phenylalanine to substitute for tyrosine is surprising, given that the hydroxyl group of tyrosine was predicted to share either one or two hydrogen bonds with adenine bases present at homing site positions 2 and 21. Three other substitutions at position 33 (cysteine, asparagine and arginine) retained some affinity for the wild-type homing site, while the remaining 15 did not. Homing site mutants. The phenotypes of our homing site mutants are consistent with the above endonuclease mutant data. Of the seven palindromic positions within the homing site, five appear to be essential for efficient site recognition (Fig. , positions 2 /21, 3 /20, 7 /16, 8 /15 and 9 /14). All mutants altered at these positions are resistant to cleavage by wild-type I-CreI in vivo. The remaining two palindromic positions appear to be less essential, as mutants altered at these positions are either completely or partially I-CreI sensitive in vivo. One of these positions (4 /19) is the only symmetrical position without any predicted endonuclease contacts based upon the I-CreI /DNA co-crystal structure . The other position (1 /22) is predicted to interact with S32, shown above to be among the least important of all contact residues examined. The four mutant homing sites that yielded highly sectored colonies upon exposure to wild-type I-CreI provided a sensitive means for examining endonuclease mutants with subtle phenotypes. As both T140A and S32A mutants were less efficient at cleaving these sites than was wild-type I-CreI , we conclude that both T140 and S32 contribute to homing site recognition. For T140, this result is consistent with the behavior of a Q44A /T140A double mutant, where a role for T140 was also implied. Toxicity to E.coli | One caveat of protein genetics is that inactivating mutations may induce gross structural aberrations in protein structure. We do not believe this to be a major problem in these studies since 24 of 31 endonuclease mutants analyzed displayed significant residual activity, toxicity or both . Of the seven mutants that displayed no apparent toxicity, four are predicted to display decreased positive electropotentials on or near their DNA-binding surfaces, either due to loss of a positive charge (K28A and R70A) or addition of a negative one (Y33D and Y33E). Two extremely toxic mutants (Y33H and Y33R) are predicted to add an extra positive charge to the protein DNA-binding surface. As a strong positive electropotential has been postulated to play a role in DNA binding , it could be that a significant portion of I-CreI toxicity to E.coli results from electrostatically driven protein --DNA interactions. There may be a second, sequence-specific component to I-CreI toxicity. The wild-type I-CreI homing site is a 22 bp sequence located in a highly conserved region of the C.reinhardtii chloroplast 23S rRNA gene. The corresponding sequence in the E.coli 23S gene, present seven times per genome, differs by only 3 bp . Of particular interest is an alteration at position 21 , which replaces an adenine with a guanine across from I-CreI amino acid 33. As mutants Y33H and Y33R have each been shown to specifically cleave the symmetrical mutant homing site containing guanines at this position in each half-site, it could be that each of these mutants interacts with the E.coli 23S site better than does wild-type I-CreI. The basis for the extreme toxicity displayed by mutants Y33A and Y33N is less clear, but may involve loss of an unfavorable interaction between the tyrosine at position 33 in wild-type I-CreI and the guanine present at position 21 in the E.coli 23S gene. If I-CreI-mediated cleavage of the E.coli 23S genes is in fact responsible for the observed toxicity, this may serve as a model for homing endonuclease derivatives engineered for use as highly specific antimicrobial agents. Altered specificity I-CreI derivatives | With their ability to function autonomously in recognizing and cleaving specific long DNA sequences, homing endonucleases have been mentioned as ideal reagents for the manipulation of complex genomes (,,). Such manipulations include gene mapping, cloning, and studying the repair of DNA double-strand breaks. Of particular interest are studies in mammalian cells showing that the repair of homing endonuclease-generated double-strand breaks can be mutagenic and can stimulate gene targeting reactions . The ability to alter the target specificities of existing homing endonucleases would expand the utility of such approaches. For example, 'designer' homing endonucleases could be used as gene therapy reagents to specifically target particular common disease-causing alleles. The great deal of structural information available for I-CreI (,,), as well as the ability to genetically manipulate I-CreI and its homing site (; this work), make it an ideal candidate for studies aimed at isolating such altered specificity derivatives. The E.coli-based assays we have developed provide a sensitive means for identifying I-CreI derivatives with altered cleavage specificities. Here, we have taken a systematic approach in examining the interaction between a single I-CreI amino acid (Y33) and its cognate homing site base. I-CreI derivatives with each of the possible 19 amino acids at position 33 were tested against homing sites containing each of the four possible bases symmetrically distributed at target site bases 2 and 21. I-CreI derivatives with increased affinities were identified for two of three possible symmetrical base substitutions at this homing site position . When the adenines present at homing site bases 2 and 21 were converted to guanines, the resulting site was efficiently cleaved only by I-CreI derivatives with histidine or arginine present at amino acid 33. Interestingly, homing site recognition by wild-type I-CreI is predicted to involve two arginine residues (R68 and R70) interacting with guanine bases in each half-site . The presence of the basic residue lysine at position 33 did not result in cleavage of this mutant site. When the adenines present at homing site bases 2 and 21 were converted to thymines, the resulting site was efficiently cleaved only by an I-CreI derivative with cysteine present at amino acid 33. The strong preference of the C33 endonuclease for this mutant homing site was confirmed in vitro (Fig. and Table ). I-CreI derivatives with leucine, serine or threonine present at amino acid 33 each displayed some affinity for this mutant site. As leucine differs greatly from the other three amino acid side chains in its chemical properties, it is hard to imagine each of these I-CreI derivatives making novel contacts with the thymine bases in this mutant homing site. However, all four side chains are of similar size, which may be important in how these I-CreI derivatives interact with this DNA sequence. Preliminary studies on another contact, that between S32 and the outermost homing site bases, have identified two interesting mutants (S32R and S32K). In vivo these mutants interacted better with the G1 C22 site than did wild-type I-CreI (Table and Fig. ). Each also interacted less well with the native homing site than did the wild-type enzyme (Table and Fig. ). In vitro the K32 endonuclease displayed a slightly lower Kd for the G1 C22 site than for the native site . Comparing Kd values for wild-type I-CreI on the two mutant homing sites tested supports the idea that the contacts at bases 1 and 22 are significantly less important for homing site recognition than those at bases 2 and 21 . Theoretically, derivatives with increased affinities for non-native sequences could be altered specificity or relaxed specificity mutants. Altered specificity mutants are those which display an increased affinity for a mutant homing site and a decreased affinity for the native I-CreI homing site. The position 33 substitutions clearly fall into this class, as demonstrated in vitro with the C33 endonuclease. Relaxed specificity mutants would be predicted to display similar affinities for both native and mutant sites. The K32 endonuclease appears to fall into this category. Relaxed specificity derivatives may be ideal for in vitro applications in that they could recognize and cleave sequences of complexity intermediate to those cleaved by native homing endonucleases and type II restriction endonucleases. For three of four mutant homing sites examined, I-CreI derivatives with increased affinities have been identified. If the results of the present work are representative, a systematic search of the remaining I-CreI DNA contacts should reveal a large number of novel contacts. Appropriate amino acid substitutions can then be combined to generate I-CreI derivatives that are specific for DNA sequences containing the appropriate cognate bases. Given the close proximity of many of the I-CreI contact residues to each other, it seems likely that there will be some constraints upon which amino acid substitutions can be used in combination. For example, introducing multiple basic residues in close proximity to each other may inhibit enzyme function. Only by first identifying novel contacts and then testing them in combination will we know the full range of DNA sequences accessible to I-CreI derivatives. Many homing endonucleases of the LAGLIDADG family contain both DNA-binding domains within a single polypeptide (,,,). Such enzymes are free to interact with non-palindromic DNA sequences, as each DNA-binding region can specify a different range of contacts. Having a collection of I-CreI derivatives that recognize and cleave different symmetrical DNA sequences, it should be possible to mix and match 'monomers' with different recognition properties to create 'dimers' that target novel asymmetric DNA sequences. In fact, a chimeric LAGLIDADG endonuclease containing domains from both I-CreI and I-DmoI has already been made and shown to bind and cleave a hybrid target site (B.S.Chevalier, T.Kortemme, M.S.Chadsey, D.Baker, R.J.Monnat Jr and B.L.Stoddard, submitted for publication). We are currently attempting to create analogous active chimeras from divergent I-CreI subunits, using the altered specificity derivatives isolated thus far. Backmatter: PMID- 12202754 TI - DNA double-strand break-induced phosphorylation of Drosophila histone variant H2Av helps prevent radiation-induced apoptosis AB - The response of eukaryotic cells to the formation of a double-strand break (DSB) in chromosomal DNA is highly conserved. One of the earliest responses to DSB formation is phosphorylation of the C-terminal tail of H2A histones located in nucleosomes near the break. Histone variant H2AX and core histone H2A are phosphorylated in mammals and budding yeast, respectively. We demonstrate the DSB-induced phosphorylation of histone variant H2Av in Drosophila melanogaster. H2Av is a member of the H2AZ family of histone variants. Ser137 within an SQ motif located near the C- terminus of H2Av was phosphorylated in response to gamma-irradiation in both tissue culture cells and larvae. Phosphorylation was detected within 1 min of irradiation and detectable after only 0.3 Gy of radiation exposure. Photochemically induced DSBs, but not general oxidative damage or UV-induced nicking of DNA, caused H2Av phosphorylation, suggesting that phosphorylation is DSB specific. Imaginal disc cells from Drosophila expressing a mutant allele of H2Av with its C-terminal tail deleted, and therefore unable to be phosphorylated, were more sensitive to radiation-induced apoptosis than were wildtype controls, suggesting that phosphorylation of H2Av is important for repair of radiation-induced DSBs. These observations suggest that in addition to providing the function of an H2AZ histone, H2Av is also the functional homolog in Drosophila of H2AX. Keywords: INTRODUCTION : A double-strand break (DSB) in chromosomal DNA is a potentially lethal lesion and, if not repaired accurately, can create genetic instabilities and gene mutations that predispose a cell to neoplastic transformation. The response of eukaryotic cells to the formation of a DSB is highly conserved and involves both checkpoint functions that arrest the cell cycle, allowing time for repair to occur, and repair functions that directly fix the break (for a review see ). DSBs can be repaired by homologous recombination (HR) or non- homologous end joining (NHEJ). Both mechanisms are used in higher eukaryotes, with NHEJ being used predominantly in G1 and HR in G2 of the cell cycle (for an example see ; for a review see ). HR is the predominant mechanism of repair used by yeast ( and references therein). One of the earliest known responses to DSB formation is phosphorylation of the C-terminal tails of H2A histones in nucleosomes located in the vicinity of the break (,). Histone variant H2AX is the H2A histone that is phosphorylated in mammals . The amino acid sequence of H2AX is nearly identical to that of H2A except for its C-terminal tail, which has a divergent sequence and 13 additional amino acids (see Fig. ) . H2AX is phosphorylated on Ser139 in an SQ motif located near the C-terminus . The phosphorylated form of H2AX is referred to as gamma-H2AX. H2AX is phosphorylated by the PIK-related protein kinase ATM or ATR, depending on whether the break is introduced by radiolytic damage or replicative stress, respectively (,). Phosphory lation occurs within 20 s of radiation-induced break formation, making it the earliest event known to occur after DNA breakage and kinase activation . While H2AX is uniformly distributed in chromatin, only H2AX located in the vicinity of a DSB becomes phosphorylated, and a single domain of phosphorylation might encompass megabases of DNA . DSBs also induce phosphorylation in budding yeast, but in yeast it is the core histone H2A that is phosphorylated rather than a histone variant . Yeast H2A is phosphorylated within an SQ motif located in its C-terminal tail, analogous to mammalian H2AX . This evolutionary conservation suggests that H2A phosphorylation is a conserved response of eukaryotic cells to DSBs . DSB-induced phosphorylation of H2A histones is important for some mechanisms by which the DSB is repaired. In budding yeast that lack H2A phosphorylation, for example, repair by NHEJ is reduced 2-fold, and knockout mice that lack H2AX have a pleiotropic phenotype that includes increased genomic instability, increased radiation sensitivity, reduced immunoglobulin class switching, and male sterility (,). For some mechanisms of repair, phosphorylation of H2A histones appears to be less important, such as for homology-dependent repair in yeast and V(D)J recombination in mammals (,). So, even though localized phosphorylation of H2A histones has been found at all DSBs thus far studied, irrespective of how the break is formed or the pathway by which it is eventually repaired, H2A phosphorylation may be functionally important for only a subset of repair processes (,, --). Antibodies that specifically bind phosphorylated H2AX in mammals also recognize radiation-induced proteins in Muntiacus muntjac, Xenopus laevis and Drosophila melanogaster . The fruitfly, D.melanogaster, has a single H2A variant, H2Av, in addition to core histone H2A . H2Av is an essential gene and is homologous to the H2AZ family of histone variants, which function in the chromatin-mediated regulation of transcription ( --). In addition to H2Av being the H2AZ ortholog of flies, the C-terminal tail of H2Av also contains an SQ motif similar to that in mammalian H2AX and yeast H2A (see Fig. ; for a broader phylogenetic comparison see also ). The presence of the SQ motif prompted previous investigators to predict that H2Av would be the protein in Drosophila recognized by gamma-H2AX-specific antibodies . We tested this prediction and, in this report, we demonstrate that the C-terminal tail of Drosophila H2Av is phosphorylated in response to radiation-induced DSBs, specifically at Ser137 within the SQ motif. Furthermore, mutant animals that lack H2Av phosphorylation have increased levels of radiation-induced apoptosis of imaginal disc cells, although the increase in apoptosis caused little or no larval lethality. These results suggest a function for H2Av phosphorylation in repair of radiation-induced DSBs. Figure 1 | C-terminal amino acid sequence of H2A histones. C-terminal amino acid sequence of H2A histones. The C-terminal tails of H2Av and H2A1 from D.melanogaster (Dm), H2AX and H2A1 from human (Hs) and H2A1 from S.cerevisiae (Sc) are shown, starting at the conserved lysine pair located at the beginning of the C-terminal tail of these histones. Dots indicate gaps and asterisks indicate stop codons. Amino acid numbers for Dm H2Av are shown above the sequence and the position at which the DeltaCT mutation truncates H2Av is indicated with an arrow. Amino acids of the conserved SQ motif are highlighted and the serine that is phosphorylated in response to a DSB in each histone is aligned below Dm H2Av Ser137. MATERIALS AND METHODS : Antibody synthesis and affinity purification | A polyclonal antiserum that recognizes both phosphorylated and non-phosphorylated H2Av was generated by injecting rabbits with a synthetic peptide homologous to the C-terminal 15 amino acids of H2Av in which Ser137 was phosphorylated. All procedures were approved by the Wadsworth Center IACUC committee. The peptide was conjugated to keyhole limpet hemocyanin via an added N-terminal cysteine before injection (Imject Maleimide Activated KLH Kit; Pierce). This antiserum was used for all experiments except for the experiment illustrated in Figure B, in which phosphorylation- specific antibodies were used. Antibodies that bound only phosphorylated H2Av peptide were isolated by reverse affinity purification. A peptide identical to the C-terminal 15 amino acids of H2Av but lacking a phosphate on Ser137 was coupled to beaded agarose via an added N-terminal cysteine (Sulfolink Kit; Pierce). Serum antibodies that bound non-phosphorylated peptide were then removed by column chromatography and antibodies specific for phosphorylated H2Av were collected in the flow-through. Drosophila strains and tissue culture cells | Information about Drosophila genes and nomenclature may be found at Flybase . Flies were maintained at 23C and 55% relative humidity on cornmeal -- brewer's yeast --glucose medium. Strains P{His2AvDeltaCTXc};l(3)His2Av810 /TM6B, P{His2AvDeltaCTXa};l(3)His2Av810 /TM6B and P{His2AvWTXa};l(3)His2Av810 /TM6B were originally constructed to analyze the essential function of H2Av as the Drosophila H2AZ homolog . The transgene is inserted on the X chromosome in each strain. Tb+ larvae that are P{His2Av};l(3)His2Av810 /l(3)His2Av810 and lack expression of the endogenous His2Av gene and Tb -- larvae that are P{His2Av};l(3)His2Av810 /TM6B and have a wildtype His2Av gene on the TM6B chromosome were collected from each strain as needed for specific experiments. Drosophila S2 tissue culture cells were grown in T-75 flasks at 27C in serum-free medium (Gibco BRL). Histone isolation and western analysis | Histones were isolated from S2 cells or whole third-instar larvae using an acid extraction protocol modified from Thorne et al. . Briefly, cells or larvae were homogenized in 15 mM Tris, pH 7.5, 60 mM KCl, 15 mM NaCl, 3 mM EDTA, 0.1 mM EGTA, 0.15 mM spermine, 0.5 mM spermidine, 0.2% NP-40, 10 mM NaF, 1.5 microg /ml aprotinin, 0.7 microg /ml pepstatin, 0.5 microg /ml leupeptin and 1 mM PMSF and nuclei were isolated from the homogenate by pelleting through sucrose. Histones were then extracted in 0.4 M HCl and precipitated with acetone. Histones from similar107 S2 cells or 40 whole larvae were loaded per sample onto 20 cm SDS --PAGE gels. The separating gels were 18% acrylamide /bisacrylamide (37.5:1) and were run for similar10 h at 13 mA. Western detection was by chemiluminescence (ECL Kit; Pharmacia). Signals were quantitated using a Molecular Dynamics 595 Fluorimager and ImageQuant software. The fraction of H2Av phosphorylated was always calculated as a ratio of phosphorylated H2Av to total H2Av in each sample. Photochemically induced DSBs and other DNA-damaging protocols | DSBs were induced photochemically essentially as described by Limoli and Ward . Briefly, S2 cells were grown in the presence of 25 microM bromodeoxyuridine for 4 days. Hoechst 33285 was then added to a final concentration of 100 microg /ml and the cells incubated for a further 10 min. Cells were then exposed to 49 995 J /m2 UV-A light (365 nm). Histones were isolated from cells as described above. For induction of general oxidative damage, S2 cells were incubated in the presence of 50 microM H2O2 for 30 min. This level of H2O2 exposure caused 20 --30% of the cells to die within 24 h and arrested the cell cycle of all other cells for at least 24 h (unpublished observation). For induction of single-strand breaks in DNA, S2 cells were exposed to 9999 J /m2 UV-C light (254 nm). This level of exposure to UV-C light caused 10 --20% of the cells to die within 24 h and arrested the cell cycle of all other cells for at least 24 h (unpublished observation). A 137Cs radiation source was used to irradiate samples at a dose rate of similar4 Gy /min, as measured by thermoluminescent dosimetry. Radiation-induced apoptosis assay | Tb+ roaming third-instar larvae of genotypes P{His2AvDeltaCTXc};l(3)His2Av810, P{His2AvDeltaCTXa};l(3)His2Av810 and P{His2AvWTXa};l(3)His2Av810 were collected from each TM6B balanced stock (see above). Larvae were exposed to 0, 4 or 8 Gy gamma-radiation at a dose rate of 4 Gy /min and allowed to recover for 4 h at 23C. After recovery, wing imaginal discs were dissected from larvae and stained for 8 min in 1 microg /ml acridine orange in EBR (10 mM HEPES, pH 6.9, 130 mM NaCl, 5 mM KCl, 2 mM CaCl2) and destained for 10 min in EBR. Discs were mounted under a coverslip and allowed to flatten until most cells were spread into a uniform monolayer, which facilitated subsequent analysis and counting. The discs were examined in the fluorescein channel of a fluorescence microscope and the total number of stained nuclei was determined for each disc by manual counting. Apoptotic cells with strongly staining nuclei were first apparent 3 h after irradiation. Recovery times greater than 4 h resulted in a majority of larvae pupating before they could be analyzed. The wing imaginal discs of wildtype and mutant larvae had the same overall size and equivalent cell densities, as determined by DAPI staining (R.L.Glaser, unpublished observation). Whole animal radiation sensitivity assay | Fifty Tb+ roaming third-instar larvae of genotypes P{His2AvDeltaCTXc};l(3)His2Av810, P{His2AvDeltaCTXa};l(3)His2Av810 and P{His2AvWTXa};l(3)His2Av810 were collected from each TM6B balanced stock (see above). Larvae were exposed to 0, 10, 20, 30, 40 or 50 Gy gamma-radiation at a dose rate of 4 Gy /min and then allowed to continue developing at 23C. The fraction of larvae surviving to adulthood was measured. A larva was scored as surviving only if the adult produced was fully mobile. Larvae that produced pharate adults that fell into the medium and died before inflating their wings were scored as not surviving. The fraction of larvae that survived without gamma-irradiation was set to 100% and all other samples were normalized to that value. Figure 2 | Radiation-induced phosphorylation of H2Av on Ser137. Radiation-induced phosphorylation of H2Av on Ser137. (A) Histones were isolated from S2 cells that were untreated (lane 1), were exposed to 50 Gy of gamma-radiation (lane 2) or were exposed to 50 Gy of gamma-radiation followed by treatment with alkaline phosphatase after histone isolation (lane 3). H2Av was visualized by western blotting using unpurified antiserum against the H2Av phosphopeptide. (B) The western blot illustrated in (A) was stripped and reprobed with affinity-purified antibodies that specifically recognize H2Av phosphorylated on Ser137. (C) Histones were isolated from whole larvae of genotype P{His2AvDeltaCT};l(3)His2Av810 /TM6B, which express wildtype H2Av from the TM6B chromosome, untreated (lane 1) or exposed to 40 Gy of gamma-radiation (lane 2). Histones were isolated from larvae of genotype P{His2AvDeltaCTXc};l(3)His2Av810, which express only the DeltaCT allele of H2Av, exposed to 40 Gy of gamma-radiation (lane 3). H2Av was visualized by western blotting using unpurified antiserum against the H2Av phosphopeptide. Antiserum specific for H2A was added to the analysis as a loading control . RESULTS : DSBs in DNA induce H2Av phosphorylation | A polyclonal antiserum directed against a peptide homologous to the C-terminal 15 amino acids of H2Av was generated and used to demonstrate that H2Av is phosphorylated in response to gamma-irradiation. The antiserum recognized a 14.6 kDa nuclear protein in S2 cells, consistent with earlier studies of H2Av (,). A slower migrating protein was also detected when the S2 cells were exposed to gamma-radiation prior to protein isolation (Fig. A). The radiation-induced protein was absent in samples treated with alkaline phosphatase, demonstrating that the protein is generated by radiation-induced phosphorylation (Fig. A). Because the H2Av peptide used to generate the antiserum was phosphorylated on what would be Ser137 of H2Av , phosphorylation-specific antibodies could be isolated using an affinity column containing non- phosphorylated peptide (see Materials and Methods). When such phosphorylation-specific antibodies were used to probe western blots of proteins from control and gamma-irradiated S2 cells, only the radiation-induced phosphorylated protein was bound (Fig. B). This result is consistent with radiation causing the phosphorylation of H2Av on Ser137. The possibility that the radiation-induced protein was not phosphorylated H2Av was considered. Radiation could have induced the phosphorylation of an SQ motif in another protein that is fortuitously recognized by antibodies in the H2Av antiserum. Genetic analysis was used to demonstrate that the phosphorylated protein was, in fact, H2Av. A fly strain was obtained that expresses His2AvDeltaCT, a His2Av gene containing a deletion that removes the C-terminal 15 amino acids of H2Av, including Ser137 . The 15 C-terminal amino acids of H2Av are not required for the essential function of H2Av and a His2AvDeltaCT transgene fully rescues the lethality caused by null mutations in His2Av . A His2AvDeltaCT transgene inserted by P-element transformation onto the X chromosome was crossed into a genotype containing l(3)His2Av810, a null allele of His2Av that contains a 311 bp deletion that removes the second exon of the gene . P{His2AvDeltaCT};l(3)His2Av810 /TM6B larvae, which express wildtype H2Av from the His2Av gene on the TM6B balancer chromosome, and P{His2AvDeltaCT};l(3)His2Av810 /l(3)His2Av810 larvae, which express only H2AvDeltaCT from the His2AvDeltaCT transgene, were both recovered from stock P{His2AvDeltaCT};l(3)His2Av810 /TM6B. Proteins were isolated from third-instar larvae and analyzed by western blotting. Both H2Av and the phosphorylated protein were detected in irradiated P{His2AvDeltaCT};l(3)His2Av810 /TM6B larvae, a result comparable to what was observed in irradiated S2 cells (compare Fig. A and C). In contrast, neither H2AvDeltaCT nor the phosphorylated protein were detected in irradiated P{His2AvDeltaCT};l(3)His2Av810 /l(3)His2Av810 larvae that express only the deleted H2AvDeltaCT protein (Fig. C). The inability to detect H2AvDeltaCT was expected, since the protein is lacking the 15 C-terminal amino acids against which the anti-peptide H2Av antiserum was generated. H2AvDeltaCT is present in these animals, since the His2AvDeltaCT transgene is required to rescue the lethality caused by the l(3)His2Av810 mutation . The inability to detect the phosphorylated protein in irradiated P{His2AvDeltaCT};l(3)His2Av810 /l(3)His2Av810 larvae provides compelling evidence that the radiation-induced protein is H2Av phosphorylated at Ser137 and not an unrelated protein fortuitously recognized by the H2Av antiserum. Phosphorylated H2Av was detected with as little as 0.3 Gy of gamma-radiation, and the fraction of total H2Av phosphorylated increased with increasing radiation exposure up to at least 100 Gy, although phosphorylation was clearly most efficient at low radiation doses (Fig. A). Phosphorylation of H2Av occurred rapidly after DNA breakage. When S2 cells were exposed on ice to 50 Gy and then transferred to medium at 23C, H2Av phosphorylation was detected within 1 min and was maximal after 5 min. The percentage of phosphorylated H2Av remained maximal for at least 5 min and then declined gradually over several hours (Fig. B). The gradual loss of phosphorylated H2Av is likely to be caused by the active dephosphorylation or turnover of H2Av rather than loss of cells due to radiation-induced apoptosis, since apoptotic cell death takes >3 h to occur (see Materials and Methods). Both the dose-response relationship and kinetics of H2Av phosphorylation are very similar to what has been reported for the radiation-induced phosphorylation of H2AX in mammals . Ionizing radiation creates a variety of cellular damage in addition to DNA DSBs, primarily through the generation of reactive oxygen species. It was important, therefore, to establish that H2Av phosphorylation is induced specifically by DSBs and not by other types of damage to DNA or other macromolecules. DSBs were induced photochemically, without the collateral damage caused by ionizing radiation, by incorporating BrdU into the DNA and then exposing the cells to long wavelength UV-A light in the presence of Hoechst dye . Such photochemically created DSBs induced H2Av phosphorylation as efficiently as did gamma-irradiation (Fig. A). In addition, neither general oxidative damage due to H2O2 exposure nor single-strand breaks induced by short wavelength UV-C light caused H2Av phosphorylation (Fig. B and C). These results, in conjunction with those discussed above, establish that H2Av is rapidly phosphorylated on Ser137 in response to DSBs (Figs --). H2Av phosphorylation helps prevent radiation-induced apoptosis | If H2Av phosphorylation has a function in DSB repair, we would predict that a cell lacking H2Av phosphorylation would be more sensitive to radiation and would be more prone to die of radiation-induced apoptosis as a consequence of its reduced ability to repair broken DNA. To test this prediction, we measured the frequency of radiation-induced apoptosis in imaginal disc cells of P{His2AvDeltaCT};l(3)His2Av810 larvae that are unable to phosphorylate H2Av, and in P{His2AvWT};l(3)His2Av810 control larvae that express wildtype H2Av in an otherwise identical genetic background. Apoptotic cells were identified by staining dissected wing discs with acridine orange, which preferentially stains nuclei of apoptotic cells and makes them brightly fluorescent in the fluorescein channel of a fluorescence microscope (see Materials and Methods) . The number of apoptotic cells in each disc was measured by manually counting the number of fluorescing nuclei. Imaginal discs from larvae lacking H2Av phosphorylation had significantly more radiation-induced apoptotic cells than did discs from wildtype larvae . In the absence of gamma-irradiation, discs from both mutant and wildtype larvae had about 300 apoptotic cells (Fig. A). Specifically, discs from P{His2AvWT};l(3)His2Av810 larvae had 331 +- 31 (n = 12), discs from P{His2AvDeltaCTXc};l(3)His2Av810 larvae had 274 +- 40 (n = 12) and discs from P{His2AvDeltaCTXa};l(3)His2Av810 larvae had 286 +- 26 (n = 12). Whatever causes apoptosis in the absence of irradiation appears to be unaffected by H2Av phosphorylation. In contrast, discs from P{His2AvDeltaCTXc};l(3)His2Av810 mutant larvae that are unable to phosphorylate H2Av had 2-fold more apoptotic cells at both 4 Gy (P < 0.001; two-tailed t-test assuming unequal variances) and 8 Gy irradiation (P < 0.0008; Fig. A and B). Discs from P{His2AvDeltaCTXa};l(3)His2Av810 mutant larvae, which have a P{His2AvDeltaCT} transposon inserted at a different location on the X chromosome, had a more modest 1.3-fold increase in apoptotic cells at 4 Gy (P < 0.08) and a 1.6-fold increase at 8 Gy (P < 0.002; Fig. B). Discs from larvae of the same mutant genotypes but derived from homozygous mutant parents lacking wildtype H2Av, as opposed to those having heterozygous parents that were used for the analysis shown in Figure , had the same degree of radiation sensitivity, indicating that maternal wildtype H2Av did not suppress the frequency of apoptotic cells (data not shown). Finally, similar increases in radiation-induced apoptosis were also observed in eye /antennal and leg discs (data not shown). The radiation-sensitive phenotype is likely to be caused solely and specifically by the lack of H2Av phosphorylation at Ser137 and not by any other unanticipated consequence of the deletion of 15 amino acids of the H2Av C-terminus. The C-terminal 15 amino acids of H2Av are not known to encode any functions other than DSB-induced phosphorylation. The amino acids are not required for the essential functions of H2Av ; the C-terminal tail has no homology outside the SQ motif involved in phosphorylation to the C-terminal tails of H2AZ or H2AX histones in other organisms ; and the DeltaCT mutation leaves a C-terminal tail of six amino acids and is therefore unlikely to have a detrimental impact on the structure or function of the globular domain of H2Av, given that the normal C-terminal tail of H2A is only seven amino acids long . Finally, the 2-fold increase in radiation-induced apoptosis of imaginal disc cells did not cause a concomitant increase in lethality. Specifically, the percentage of irradiated roaming stage third-instar larvae to complete metamorphosis and emerge as pharate adults did not differ between larvae that could or could not phosphorylate H2Av, irrespective of radiation dose . This was also true of early third-instar larvae as well as 24 h pupae (data not shown). The 2-fold increase in imaginal-cell apoptosis might have only negligible effects on overall disc development and morphogenesis, or perhaps is a level of cell loss for which the imaginal disc is able to compensate by increased proliferation of viable cells. We also measured the radiation sensitivity of embryos and adults, and again found no difference between animals that could and could not phosphorylate H2Av (data not shown). Figure 3 | Kinetics of H2Av phosphorylation. Kinetics of H2Av phosphorylation. (A) Histones were isolated from S2 cells after exposure to increasing amounts of gamma-radiation. H2Av was visualized by western blotting using unpurified antiserum against the H2Av phosphopeptide and quantitated by fluorimagery. The percentage of total H2Av phosphorylated is graphed as a function of radiation dose. (B) Histones were isolated from S2 cells at various times after exposure to 50 Gy of gamma-radiation. H2Av was visualized by western blotting using unpurified antiserum against the H2Av phosphopeptide and quantitated by fluorimagery. The percentage of total H2Av phosphorylated is graphed as a function of time after radiation exposure. Figure 4 | H2Av phosphorylation is induced specifically by DSBs. H2Av phosphorylation is induced specifically by DSBs. (A) Histones were isolated from S2 cells that were untreated (lane 1), were exposed to UV-A light (lane 2), had incorporated BrdU and been incubated with Hoechst dye (lane 3) or were exposed to UV-A light after incorporation of BrdU and incubation with Hoechst dye (lane 4). H2Av was visualized by western blotting using unpurified antiserum against the H2Av phosphopeptide. (B) Histones were isolated from S2 cells that were untreated (lane 1), incubated with H2O2 (lane 2) or incubated with H2O2 followed by exposure to 100 Gy of gamma-radiation (lane 3). H2Av was visualized by western blotting using unpurified antiserum against the H2Av phosphopeptide. (C) Histones were isolated from S2 cells that were untreated (lane 1), exposed to UV-C light (lane 2) or exposed to UV-C light followed by exposure to 100 Gy of gamma-radiation (lane 3). H2Av was visualized by western blotting using unpurified antiserum against the H2Av phosphopeptide. Figure 5 | Imaginal cells unable to phosphorylate H2Av are radiation sensitive. Imaginal cells unable to phosphorylate H2Av are radiation sensitive. (A) Larvae expressing either wildtype H2Av (WT), which is phosphorylated, or H2Av deleted of its C-terminus (DeltaCT), which is not phosphorylated, were exposed to 0 or 8 Gy of gamma-radiation. Apoptotic cells were identified in dissected imaginal wing discs by staining with acridine orange. Nuclei of apoptotic cells stain brightly and appear as sharp pinpoints of light in the images shown. The DeltaCT discs illustrated are from larvae of genotype P{His2AvDeltaCTXc};l(3)His2Av810. The numbers of apoptotic cells counted were 370 (0 Gy) and 1146 (8 Gy) for the wildtype discs illustrated and 288 (0 Gy) and 2088 (8 Gy) for the DeltaCT discs. (B) The number of radiation-induced apoptotic cells per disc was quantitated by counting the total number of fluorescent nuclei in discs irradiated with either 4 or 8 Gy of gamma-radiation and subtracting the average number of fluorescent nuclei observed in unirradiated discs of the same genotype. Larvae of genotype P{His2AvWT};l(3)His2Av810, which express wildtype H2Av (open column), and larvae of genotypes P{His2AvDeltaCTXc};l(3)His2Av810 and P{His2AvDeltaCTXa};l(3)His2Av810, two independent strains expressing H2Av deleted of its C-terminus (gray columns), were analyzed. Mean and standard errors were calculated from 9 --12 independent measurements for each sample. Figure 6 | Larvae that lack H2Av phosphorylation are not radiation sensitive. Larvae that lack H2Av phosphorylation are not radiation sensitive. Third-instar larvae of genotype P{His2AvWT};l(3)His2Av810 that can phosphorylate H2Av (open boxes) and larvae of genotypes P{His2AvDeltaCTXc};l(3)His2Av810 and P{His2AvDeltaCTXa};l(3)His2Av810 that lack H2Av phosphorylation (filled boxes and filled triangles, respectively) were exposed to increasing amounts of gamma-radiation. The percentage of larvae that survived to adulthood was measured and is graphed as a function of radiation dose. Mean and standard deviations were calculated from the results of at least three independent measurements. DISCUSSION : Drosophila H2Av is unique among histone variants, providing in a single histone both the transcription function of H2AZ variants and the DNA repair function of H2AX variants. Ser137 in the C-terminus of H2Av is phosphorylated in response to DSBs and preventing this phosphorylation causes imaginal disc cells to become radiation sensitive (Figs --). These observations support the conclusion that H2Av phosphorylation contributes to repair of DSBs and that H2Av is the functional homolog in flies of mammalian H2AX. The amino acid sequence of H2Av outside its C-terminal tail clearly identifies H2Av as the H2AZ ortholog in Drosophila. The H2AZ function provided by H2Av is essential for normal fly development and is likely involved in chromatin-mediated regulation of transcription, although a function in transcription has yet to be directly demonstrated in flies ( --). The juxtaposition of both DNA repair and transcription functions in a single variant histone distinguishes H2Av in Drosophila from the situation in both yeast and mammals, in which these functions are provided by separate H2AX and H2AZ histones. The phosphorylation of H2Av in Drosophila is also distinct from DSB-induced phosphorylation in yeast and mammals in that the histone phosphorylated is specifically an H2AZ variant. Mammalian H2AX outside its C-terminal tail is almost identical to H2A, differing at only three of 120 residues, and H2A is the histone phosphorylated in yeast (,). The phosphorylation of H2Av in Drosophila suggests that an H2AZ variant can make the same contribution to DNA repair as does H2A, assuming that DSB-induced phosphorylation in mammals, flies and yeast facilitates repair by a single mechanism. The equivalence of H2AZ and H2A histones with respect to DNA repair was not necessarily expected, given that H2AZ and H2A are not equivalent with respect to their functions in transcription. H2A cannot functionally replace H2AZ in flies or yeast and H2AZ cannot functionally replace H2A, at least in yeast, in which such an experiment can be done (,). Perhaps H2Av in flies, H2AX in mammals and H2A in yeast function equivalently with respect to DNA repair because their C-terminal tails, the substrates for DSB-induced phosphorylation, are located in the same position along the chromatin fiber. The C-terminal tail of all H2A histones extends from the nucleosome near the position where linker DNA enters and exits, and H2A C-termini can directly interact with linker DNA (,). Since the tails are in the same location, it is reasonable to assume that phosphorylation of those tails would have comparable effects. If the location of the C-terminal tail is, in fact, all that matters for DNA repair, this would also imply that any unique functions provided by the globular domains of H2AZ or H2A are unrelated to, and likely unaffected by, the mechanism by which phosphorylation of the C-terminal tail facilitates repair. This is a particularly interesting issue with respect to H2AZ variants, because they may be involved in regulating the equilibrium among differing states of chromatin compaction . Identification of a function for H2Av in DNA repair also provided an explanation for a puzzling aspect of the pattern of localization of H2Av in the Drosophila genome, which was characterized previously. H2Av-containing nucleosomes are found associated with DNA sequences located through the genome, including non-coding satellite sequences within heterochromatin . The association of H2Av with non-coding satellite sequences would be difficult to understand if the only function of H2Av was in transcriptional regulation. Identification of a function for H2Av in DNA repair provides a plausible explanation for why H2Av would be associated with non-coding sequences. A DSB located anywhere in the genome is a potentially lethal lesion, so some amount of H2Av-containing nucleosomes might need to be associated with all sequences of the genome to ensure that repair can be facilitated wherever a DSB might occur. In addition to the localization of H2Av to all regions of the Drosophila genome, the density of H2Av-containing nucleosomes also varies significantly and frequently along the length of each chromosome arm . While variations in density are likely to be a general characteristic of H2AZ variants, this pattern is quite different from the uniform distribution observed for mammalian H2AX and yeast H2A (,). The fact that H2Av functions in both transcription and repair demonstrates that both functions can be compatible with the localization pattern of a single histone; it suggests, more generally, that H2AZ and H2AX variants do not require mutually exclusive locations in the genome to provide their respective functions. The variation in H2Av density also raises the possibility that the efficiency of DSB repair could vary among different regions in the Drosophila genome, if it were the case that the density of H2Av in any given region of chromatin influenced the extent to which DSB-induced phosphorylation facilitated repair. Alternatively, a minimum density of H2Av-containing nucleosomes might be sufficient to provide the function of H2Av in repair and increases in density above that minimum might be necessary for the essential function of H2Av in transcription, but would not influence repair. Overall, the phenotype of Drosophila lacking H2Av phosphorylation is quite modest. Radiation sensitivity in larvae unable to phosphorylate H2Av increased 2-fold when assayed at the level of imaginal cell apoptosis and no increase in sensitivity was observed at the level of larval lethality (Figs and ). If H2Av phosphorylation were an essential step in the DSB repair pathway, radiation-induced larval lethality would be expected to increase substantially. For example, mutation of specific repair proteins, such as DmRAD54, can increase radiation-induced lethality 7-fold, and if a second repair gene is mutated simultaneously the increase in radiation-induced lethality can reach nearly 40-fold . In the light of these observations, the modest phenotype described in this report suggests that H2Av phosphorylation promotes repair efficiency rather than being an essential step in the repair process itself. A similar conclusion is suggested by genetic analyses of DSB-induced phosphorylation of H2A in budding yeast. Prevention of H2A phosphorylation increased the sensitivity of yeast to radiomimetic drugs, but it only reduced NHEJ repair about 2-fold and appeared to have little or no effect on HR . The phenotype of H2AX knockout mice also suggests that H2AX phosphorylation in mammals is probably not essential for most mechanisms of DSB repair, although H2AX phosphorylation in mice does appear to be more important for some types of repair than is DSB-induced phosphorylation for repair in either yeast or flies. Mice lacking H2AX are growth retarded, radiation sensitive and male sterile . In contrast, Drosophila lacking H2Av phosphorylation grow at normal rates, larvae are no more radiation sensitive than the wildtype and both sexes are fertile (Fig. ; R.L.Glaser, unpublished observations). So, genetic analyses of DSB-induced phosphorylation of H2A histones in yeast, flies and mice are consistent with the general conclusion that the function of phosphorylation is to facilitate repair efficiency. The phenotype of H2AX knockout mice further suggests that during evolution phosphorylation of H2AX in mammals acquired greater importance for particular DSB repair processes. Backmatter: PMID- 12202764 TI - Gene expression profiling of the aging mouse cardiac myocytes AB - Heart disease remains the most frequent cause of death in the general population with increasing incidence in the elderly population. The pathologic failure of the aging heart may be related to structural and functional alterations in cardiac muscle cells. However, the molecular mechanisms underlying the aging-related decline in cardiac muscle function are largely unknown. To provide the first analysis of cardiac aging at the level of gene expression, we established and compared cDNA libraries from apparently healthy young and aged mouse ventricular cardiac muscle cells. We report the identification of genes that exhibit aging-related changes of mRNA levels. Aging expression profiles in aged hearts indicate decreased cellular adaptation and protection against stress-induced injury together with the development of contractile dysfunction. The data suggest reduced activity of the mitochondrial electron transport system and reduced levels of cardiac-specific transcription regulators. The cardiomyocyte aging profile of gene expression displays similarities with known heart disorders. Genes whose mRNA levels change with aging in cardiomyocytes might profoundly affect pathological changes in the heart. Keywords: INTRODUCTION : The genetics of aging is one of the biggest challenges to genomic and medical research. The ultimate genomic goals are the understanding of the dynamic network of genes, its effect on the aging process, interaction with disease, and organ specificity . It has been proposed that the genetic component of aging is small , however, studies with centenarians have provided evidence that genes most likely play a prominent role in the ability to achieve older age . Although the heart is the focus of major problems in the aging population and aging-related heart disease is the most frequent cause of death, at present, genetic analysis of cardiac aging is not available. Cellular aging is commonly associated with the instability of the nuclear and mitochondrial genomes and oxidative protein damage because of long-term exposure to reactive oxygen species. Cardiac myocytes, however, are highly specialized high-oxygen-content cells that generally do not divide. Therefore, errors in nuclear DNA duplication and replicative senescence should not be critical in their aging. The cardiac myocytes, however, house a large number of mitochondria that are dividing. Somatic mutations in the mitochondrial genome are documented with aging. These mutations have the potential to influence mitochondrial functions, which in turn may influence functions in the nuclear genome. Mitochondrial electron transport abnormalities because of DNA deletion mutations can result in oxidative cellular protein damage. Oxidative protein damage may directly cause changes in transcription factors, chromatin structure and function. With aging, these changes can result in alterations of nuclear gene expression, mRNA stability or both. Searching for cardiac muscle cellular genes that change expression levels during aging would lead to a better understanding of decreased cardiac performance with aging. We applied a combination of four different methods to screen directly for differences in gene expression levels in cardiac myocytes from young and aged mouse hearts. We found 43 genes that accumulate at different levels with age. The levels of these genes did not change significantly from animal to animal. Our results clearly indicate the presence of cardiac-specific aging gene candidates. We also found that the cardiomyocyte aging profile of gene expression displays similarities with known heart disorders. MATERIALS AND METHODS : Adult mice hearts | Hearts were obtained from C57BL/6 mice with no clinical evidence of heart failure such as labored breathing, ascites, and significant weight gain or weight loss. Adult mouse cardiomyocytes | Left ventricular cardiac myocytes (LVCs) were enzymatically dissociated from the left ventricle according to a previously published protocol with slight modifications. Hearts from young animals were perfused for 20 min and old hearts were perfused for 25 min. Each heart from young animals yielded approximately 2 --3 x 106 cells and from aged animals 0.7 --1.0 x 106 contracting rod-shaped cells. The purity, the shape and the contractility of the myocyte preparations were inspected by light microscopy. Cells were also inspected for sarcolemma structure by confocal microscopy with FITC-labeled rabbit antibodies (Sigma), which are specific to cardiac alpha actinin. The accumulation of autofluorescent material with aging was monitored by confocal microscopy (excitation at 500 nm). Freshly isolated cardiomyocytes typically contained >90% rod-shaped cells and <1% of non-myocytes. Cells were used immediately for preparation of RNA or for microscopy. Young C57BL/6 adult male mice were 4 months old and the aged C57BL/6 adult male mice were 20 months old. Retired C57BL/6 male breeders (Jackson laboratory) were used after reaching the age of 20 months or delivered from the National Institute of Aging at the age of 20 months. Differential gene subtraction | For differential gene expression analysis, we used the PCR-based cDNA subtraction procedure (Clontech). LVCs isolated from male and female hearts were used for the isolation of total RNA as suggested by the manufacturer (Qiagen). cDNA libraries were prepared using the SMART cDNA Synthesis Kit (Clontech) as suggested by the manufacturer. Subtraction hybridization was routinely performed with 600 ng of driver cDNA that was mixed with 20 ng of the tester cDNA. Subtracted fractions were then amplified by selective PCR using the Advantage cDNA PCR Core Kit (Clontech). The subtraction efficiency was determined by virtual northern blot analyses as suggested by the manufacturer. Differential subtractions were independently conducted with six cDNA libraries prepared from hearts from both ages (12 young C57BL/6 male and 12 old C57BL/6 male mice). Cloning and analysis of subtracted products | The subtracted cDNAs were ligated into Zero Blunt TOPOTM cloning vector (Invitrogen) and transferred to DH5alpha Escherichia coli. The positive clones were identified following standard procedures. In brief, the tester and the driver cDNA fractions were 33P radioactively labeled and used for the colony hybridization. The clones that hybridize to the tester probe but not to the driver probe represent differentially expressed mRNAs. These clones were isolated and each cDNA fragment from the positive clones was used as a probe to confirm its differential expression by virtual northern blotting. Confirmed positive clones were sequenced using a 310/317 Genetic Analyzer (Applied Biosystems). Sequence data were compared with the GenBank database using the BLAST program at the National Center for Biotechnology Information (National Institutes of Health, Bethesda, MD). Northern blotting | We routinely used 2 microg of RNA, which was separated on a 0.7% agarose gel, denatured and then blotted onto a nylon membrane (GeneScreen plus; NEN). Hybridization with the radioactively labeled probes was performed according to standard protocols in Quick Hybrid solution (Stratagene). We analyzed the RNA pool made from 10 male and 10 female C57BL/6 mouse heart LVC preparations. Using pairwise, young and respectively aged mice RNA pools, comparisons were made with [33P]dATP radioactively labeled (Prime-It II; Stratagene) individual cDNA clones. The radioactive hybridization signal was visualized after 2 --6 days exposure using the PhosphorImaging system, NIH Image and Adobe PhotoShop software. Differential gene expression analysis by DNA array | An alternative method to analyze gene expressions was performed using DNA-arrayed membranes. We developed a DNA array with immobilized oligonucleotides that recognize the 3' end of the mRNAs of the differentially expressed genes. To ensure gene-specific hybridization for each gene, 20 fmol of two to four different 60 bp oligonucleotides (MWG Biotech) were spotted onto hybridization transfer membranes (GeneScreen plus) with a 1536-pin replicator (V and P Scientific, CA). The selected oligonucleotides satisfied the following criteria: Tm within 85 --95C; absence of secondary structure; and absence of cross-hybridization verified by querying each oligonucleotide against the expressed sequence tag (EST) database. Blank spots without oligonucleotides were included for evaluation of background caused by non-specific interactions of individual probes with the membrane. For the array analysis the cDNA libraries were labeled with [33P]dCTP (Prime-It II). Labeled cDNA was purified from free [33P]dCTP by a QiaQuick PCR column (Qiagen) and heat denatured before hybridization to the membrane array. After washing and exposure, the spot reading on the membranes was performed with the PhosphorImager. Matrix overlay maps spotted oligonucleotides on the array. In a single experiment, two or three identical membranes were hybridized with each of the radioactively labeled cDNA libraries and analyzed. Individual cDNA libraries from eight young and eight old male hearts were labeled and used separately for hybridization. The data were imported into Microsoft Excel spreadsheets. The data from different arrays were normalized using a set of highly and steadily expressed genes, and the probes with hybridization below background noise were excluded from the analysis. The readings of each oligonucleotide probe averaged among replicate arrays hybridized with the same cDNA were considered independent measurements. All such measurements obtained from old animals were pooled to make the 'old' sample, and all from the young made up the 'young' sample. The hypothesis that a gene was differentially expressed in the old as compared with the young hearts was tested by calculating the corresponding P-value using the native Excel t-test option. Gene expression analysis by Affymetrix gene chips | We analyzed total RNA from two sets of six mice independently. cDNA was labeled and hybridized to the gene chip (murine genome array U74Av2) in the Affymetrix facility at Beth Israel Deaconess Medical Center. The data analysis of the micro-arrays was performed using Affymetrix software. Pairwise comparisons between individual mice were made using Excel software as recommended by Affymetrix. Changes higher than 1.2-fold were considered as a difference and they were included in the tables. The data analysis of the micro-arrays was performed in the Affymetrix facility. Each gene in the Affymetrix array was represented by perfectly matched (PM) and mismatched (MM) control oligonucleotides. Fluorescence intensity was read for each oligonucleotide to calculate the average signal intensity (SI) for each gene by subtracting the intensities of the PM from the intensity of the MM control, after discarding the maximum, the minimum and any outliers beyond three standard deviations. The effects of aging were determined by comparing each young (a mixture from six young hearts) with each old (a mixture from six old hearts) cRNA, generating two pairwise comparisons. The data reported in Table S1 of the Supplementary Material, represent the average fold changes obtained through the two pairwise combinations. The calculations were performed by an Affymetrix algorithm. RESULTS : To find cardiac-specific 'aging genes', we first constructed cDNA libraries from total RNA of purified LVCs from C57BL/6 mouse male hearts at age 4 months (young) and at age 20 months (aged). The isolated LVCs contained typically >90% rod-shaped cardiomyocyte cells . Confocal microscopy revealed the expected presence of the autofluorescent cytoplasmic lipofuscin inclusions in the aged cells (Fig. B) (,). Significantly fewer fluorescent inclusions were observed in LVCs from young mice (Fig. A) . We first compared RNA levels from both ages by the cDNA differential subtraction method. Three hundred and sixty individual clones from the subtracted cDNA libraries were sequenced to identify the differentially expressed genes. We found clones that were present in the young cDNA libraries at a significantly higher level than in the aged libraries (Table , D), as well as other clones that are induced in aged LVCs (Table , U). The proportion of mitochondrial RNAs was relatively high (48%) in the libraries from both groups, most likely because of the high copy number of mitochondrial genomes per cardiac muscle cell and the presence of tandem A-run sequences. To further validate the data, we analyzed mRNA expression in young and aged LVCs applying northern blotting. Equal amounts of total RNA from young and aged LVCs were probed for signals from 34 of the differentially expressed genes. Examples of the northern blot assays are displayed in Figure . To confirm results from the cDNA subtraction experiments and to verify the reproducibility of aging-related differential gene expression in individual mouse preparations, we performed expression analysis experiments using an oligonucleotide array. We developed an array for differentially expressed genes. In addition, we included probes for 120 genes previously identified to play roles in the cardiac muscle specialized functions . We performed eight sets of independent hybridization experiments with identical array membranes and radioactively labeled cDNA libraries from individual hearts. We also conducted two sets of analyses with an independent gene array containing 6000 mouse genes and ESTs. The results of the gene array comparative assays are presented in Table S1 of the Supplementary Material. To further validate the data, we analyzed mRNA expression in young and aged LVCs applying northern blotting. Equal amounts of total RNA from young and from aged LVCs were probed for signals from 34 of the differentially expressed genes. Examples of the northern blot assays are shown in Figure . We found 43 RNAs that accumulate at different levels with male mouse strain C57BL/6 age. Their levels did not change significantly from animal to animal . Five of these clones have not been reported previously in the gene bank. Figure 1 | Confocal microscopy of LVCs from healthy young and aged mouse hearts. Confocal microscopy of LVCs from healthy young and aged mouse hearts. Anti-alpha actinin --FITC conjugated antibodies (green) were used to label the alpha actinin distribution in the contractile apparatus. The autofluorescent lipofuscin inclusions (pink and yellow) were visualized with the Texas red filter. Nuclei are labeled with arrows. Pictures were taken at magnification x600 and simultaneously with three of the filters. (A) LVCs from young animals. (B) LVCs from aged animals. Figure 2 | Representative northern blot analyses confirming changes in mRNA levels. Representative northern blot analyses confirming changes in mRNA levels. Total cellular RNA samples (2 microg) from young and aged LVCs were examined by northern blot analysis. Genes to which the probes correspond are identified to the left of the autoradiograms. Y, young LVCs; A, aged LVCs. Probes were radioactively labeled differentially expressed cDNA fragments. The level of mRNA specific for beta actin did not change significantly with aging and was used as an internal control. Figure 3 | DNA arrays were used to confirm differential gene expression profiles. DNA arrays were used to confirm differential gene expression profiles. Identical DNA array membranes were probed with individual 33P-labeled cDNA libraries prepared from (A) young and (B) aged LVCs. Arrays contain oligonucleotide probes that are specific for the 45 differentially expressed clones and for 120 additional genes with known function in the cardiac muscle cells. Bottom panels of (A) and (B) show an enlargement of the boxed portions in the top panels. Matrix overlay maps the individual oligonucleotides for each of the genes. Hybridization position of some of the differentially expressed genes in the magnified region: 1 and 2, cytochrome NADH dehydrogenase subunit 1; 3 and 4, cytochrome c oxidase, subunit 3; 5 and 6, cytochrome b; 33 and 34, cardiac myosin light-chain; 41 and 42, alpha B-crystallin; 65 and 66, Hsp25. Table 1 | mRNAs that decrease or increase in LVCs from aging mouse hearts compared with those from young mouse hearts DISCUSSION : Our results indicate that the induction of an oxidative stress-induced transcriptional response may be a common feature of cellular aging in cardiac and in skeletal muscle of rodents and primates, but the extent to which aging modifies these responses may be cell specific. Our results indicate that the expression of several members of the stress-responsive heat-shock protein (Hsp) family is modulated by aging (Table , Figs and , Table S1 in Supplementary Material). Changes in the oxidative stress-induced transcriptional response may be a common feature of skeletal (,) and cardiac muscle aging. The mRNA for the inducible Hsp70 and Hsp25 decreased in aged LVCs. A similar aging pattern was observed in mouse and monkey skeletal muscle cells (,). The accumulation of Hsp70 in cardiac cells after heat stress is known to protect the myocardium from ischemic injury . We found that the aged LVCs expressed less hsc70 and alpha B-crystallin mRNAs. One of the roles of the constitutively expressed hsc70 is to bind to misfolded polypeptides to deter inappropriate interactions, which may lead to aggregation and loss of function (,). alpha B-crystallin has been shown to protect against hyperthermia, hypertonic stress and various cytotoxic agents, and its ectopic expression in adult rat cardiomyocytes leads to increased protection against ischemic injury . We found that Hsp32 mRNA was present in young LVCs and undetectable in cDNA libraries from aged LVCs. Hsp32, also known as heme oxygenase 1 (HO-1), is inducible in many tissues by various agents, including heme compounds, heavy metals, sulfhydryl reagents and hydrogen peroxide . The heart is rich in cytochromes and is a site of heme synthesis, the substrate for HO-1 . The reduced expression of Hsp32 mRNA might significantly contribute to changes in myocardial mechanisms of cellular adaptation and sensitivity to environment. It is not known exactly how changes in Hsp transcript levels contribute to the decline of cardiac performance with age. The precise mechanism by which heat-shock proteins confer LVC protection remains to be elucidated. This knowledge ultimately would help in developing pharmacological methods to improve the aging heart's ability to respond to environmental and oxidative stresses. We also found age-related changes in RNA levels in mitochondrially encoded transcripts. Constituents of the mitochondrial respiratory chain such as the mitochondrially encoded cytochrome b and cytochrome c oxidase subunit 3 were decreased in aged LVCs. Cytochrome b, a catalytic subunit of complex III, is directly involved in electron transfer. A decrease in both of these enzymatic activities with aging has been previously demonstrated . In mouse and monkey skeletal muscle cells mRNAs levels for some of the nuclear encoded subunits of these enzymes decline with aging (,). Although LVCs house a large number of mitochondria, the ATP consumption in these cells is high and the loss of metabolic activity in a certain number of mitochondria might result in cardiac life span shortening. We observed less mitochondrial creatine kinase (Mi-CK) mRNA in aged LVCs. The Mi-CK enzymatic activity also declines with aging as does the steady state level of the Mi-CK mRNA (data not shown). Mi-CK octamer --substrate complexes have a stabilizing and protective effect against mitochondrial permeability and pore opening . In mouse skeletal muscle, the steady state level of this enzyme increases with aging (,). The cytosolic creatine kinase mRNA and enzymatic activity in LVCs (data not shown), however, were found to increase slightly with aging. Cytosolic creatine kinases play a crucial role in the energetics of Ca2+-homeostasis . It is possible that with aging, cardiac muscle cells switch on compensatory mechanisms that cause elevated activity of the cytosolic creatine kinase. Aging is associated with changes in mRNA level from genes whose proteins play a role in the LVC contractile apparatus. For example, the mRNA level for dystrophin was reduced with aging. In humans, the cardiomyopathies associated with Duchenne muscular dystrophy , Becker muscular dystrophy and X-linked dilated cardiomyopathy are all caused by defects in the dystrophin gene . Since dystrophin-deficient mice have an increased vulnerability to acute pressure overload in vivo the aging-associated reduced dystrophin mRNA level might contribute to a reduced threshold for the development of contractile dysfunction upon exposure to increased levels of mechanical stress . There was aging-specific decreases in mRNA for alpha tropomyosin, alpha skeletal actin and sarcoplasmic reticulum Ca2+-ATPase (SERCA2) transcripts in aged LVCs (Table , Figs and , Table S1 in Supplementary Material). SERCA2 plays a key role in the contraction --relaxation cycle of the myocardium by controlling the intracellular Ca2+ concentration. A reduction in the SERCA2 mRNA level because of transcription repression has been documented in all models of heart failure, including human failing ventricles (,). None of these contractile apparatus protein genes has been associated with aging of mouse and rhesus monkey skeletal muscle (,). The LVC contractile apparatus most likely ages differently from that of skeletal muscle because of its significant functional difference, namely the continuous necessity of cardiac myocytes to produce force and motion. The mechanism of aging modulation of mRNA levels remains to be determined. Aging could at least in part modulate the stability of certain mRNAs by a heat-shock protein --ubiquitin --proteasome pathway through the reduced level of Hsp70 . Aging might either induce or fail to trigger a pathway that leads to repression or activation of the transcription promoters. The levels of transcription factors could also be modulated. The Nkx2.5, GATA4, c-jun and Jun B mRNA levels declined with aging (Table , Figs and , Table S1 in Supplementary Material). Jun homo- and hetero-dimers regulate transcription in response to stimuli of several genes by interacting with the AP-1, phorbol 12-O-tetradecanoate 13-acetate- and cAMP-response promoter elements . Concerted changes in the SERCA2 , cytochrome c , actin isoforms and some of the heat-shock protein mRNA levels may be a direct effect of the reduced levels of c-jun and junB with aging. AP-1 and GATA4 are both involved in the response to the cardiac pressure overload . AP-1 and GATA4 are targets of signaling cascades (,). Differences in response to stress may occur because of aging-related changes in protein phosphorylation or oxidative protein damage of cascade members. GATA4 fine-tunes cardiac myocyte-specific gene expression together with Nkx2.5 (,). The reduced Nkx2.5 mRNA level in aged LVCs coincides with the reduction of connexin-43 and cardiac myosin light-chain mRNAs. Nkx2.5 was previously shown to regulate the expression of both genes (,). The cardiac phenotype of mice expressing the mutant Nkx2.5 are in some ways similar to the aging heart . Both result in progressive cardiac conduction defects that may lead to heart failure. The down regulation of these transcription factors may be important for the transition and establishment of an 'aged mode' of LVC transcription regulation of gene expression. Although we might find some additional expression differences by applying less stringent conditions and testing a larger number of animals, our results suggest that aging alters expression in the ventricular cardiac muscle of the mouse. Aging influences gene expression in different cell types, some of which respond differently than the ventricular cardiac myocytes suggesting that aging gene expression pattern might differ according to the specialized cellular functions. It is not known exactly how the subset of mRNAs is selected for expression changes with LVC aging. It is not known if the LVC aging gene expression profile in cardiac myocytes from C57BL/6 mice displays similarities with the profiles of other mouse strains and of other mammalian species. It is not known if there is a connection between the pathways that lead to accumulation of the autofluorescent lipofuscin inclusions and to the LVC aging gene expression profile. We propose that several cellular genes whose mRNA levels change with LVC aging might profoundly influence the age-related pathologic changes in the heart. SUPPLEMENTARY MATERIAL : Supplementary Material is available at NAR Online. Backmatter: PMID- 12202748 TI - SURVEY AND SUMMARY: Novel domains and orthologues of eukaryotic transcription elongation factors AB - The passage of RNA polymerase II across eukaryotic genes is impeded by the nucleosome, an octamer of histones H2A, H2B, H3 and H4 dimers. More than a dozen factors in the yeast Saccharomyces cerevisiae are known to facilitate transcription elongation through chromatin. In order to better understand the evolution and function of these factors, their sequences have been compared with known protein, EST and DNA sequences. Elongator subcomplex components Elp4p and Elp6p are shown to be homologues of ATPases, yet with substitutions of amino acids critical for ATP hydrolysis, and novel orthologues of Elp5p are detectable in human, and other animal, sequences. The yeast CP complex is shown to contain a likely inactive homologue of M24 family metalloproteases in Spt16p /Cdc68p and a 2-fold repeat in Pob3p, the orthologue of mammalian SSRP1. Archaeal DNA-directed RNA polymerase subunit E" is shown to be the orthologue of eukaryotic Spt4p, and Spt5p and prokaryotic NusG are shown to contain a novel 'NGN' domain. Spt6p is found to contain a domain homologous to the YqgF family of RNases, although this domain may also lack catalytic activity. These findings imply that much of the transcription elongation machinery of eukaryotes has been acquired subsequent to their divergence from prokaryotes. Keywords: INTRODUCTION : Chromatin decompaction is required for efficient RNA polymerase II (RNAP-II)-mediated transcription of eukaryotic protein coding genes . Transcription is divided into an initiation stage, during which transcription factors and RNAP bind to promoter sites and RNA synthesis commences, followed by an elongation stage, during which RNAP traverses along the DNA assembling an RNA transcript. Transcription elongation through chromatin is severely hindered by the nucleosome, a structure containing DNA wrapped around two copies each of histones H2A, H2B, H3 and H4. Disruption of the structural integrity of the nucleosome, by histone acetylation and /or methylation, by DNA unwinding, or by histone translocation, allows passage of the RNAP-II complex along the gene ( --). In the yeast Saccharomyces cerevisiae at least a dozen factors are known to facilitate elongation through chromatin . These are Rad26p, CP (a heterodimeric factor of Cdc68p /Spt16p and Pob3p), Elongator (containing two subcomplexes, each of three subunits), the Spt4p --Spt5p heterodimer and Spt6p. The molecular functions of these 12 differ greatly. Human DSIF, containing orthologues of yeast Spt4p --Spt5p, functionally interacts with other elongation factors as well as physically with the largest subunit of RNAP . Both FACT, the human version of the yeast CP complex, and Spt6p bind histones directly (,) whereas Elongator Elp3p acts as a histone acetyltransferase . An additional Rad26p-associated factor called Def1p /YKL054Cp enables ubiquitin-mediated proteolysis of RNAP . Human orthologues are known for all of the 12 S.cerevisiae elongation factors, with the exceptions of Elp5p and Elp6p, and Def1p /YKL054Cp. Consequently, transcription elongation processes in mammals and yeast are likely to be highly similar. In contrast, only three of the 12 factors, namely Rad26p, Spt5p and Spt6p, have highly sequence-similar homologues in bacteria, and archaea have likely orthologues only of Rad26p, Spt5p and Elp3p. The paucity of candidate orthologues of eukaryotic transcription elongation factors in archaea is curious since they are thought to possess chromatin-like structures . This study sought to determine whether previously undetected homologues, orthologues and domains of S.cerevisiae transcription elongation factors could be detected using in-depth sequence database searches. Its aims included the prediction of molecular function using the homology paradigm, and the identification of candidate orthologues of yeast elongation factors in mammals, bacteria and archaea. Sequence data from diverse sources, including incompletely sequenced genomes and expressed sequence tags, were found to be valuable in identifying previously unforeseen evolutionary relationships. Table 1 | Human, archaeal and bacterial orthologues or homologues (in parentheses) of 12 S.cerevisiae transcription elongation factors MATERIALS AND METHODS : PSI-BLAST, TBLAST-N and BLASTX searches were undertaken at the National Center for Biotechnology Information (NCBI; ) using NCBI databases, including the non-redundant protein sequence database (nr; ftp.ncbi.nlm.nih.gov /blast /db /) currently containing approximately 900 000 sequences. PSI-BLAST searches employed an E-value inclusion threshold of 2 x 10 --3 and composition-dependent statistics , except where stated. The E-value corresponding to an alignment score x is the number of false positive sequences that are expected to be aligned with scores x, or higher, in that search by chance. Additional BLAST searches used the VGE and NCBI unfinished genomes' sites, and organism-specific sites such as dicty.sdsc.edu /annot-blast. html (for Dictyostelium discoideum) and (for Caenorhabditis briggsae). Other searches used the nrdb90 protein sequence database (ftp: / /ftp.ebi.ac.uk /pub /databases /nrdb90 /) for which no pair of sequences has greater than 90% pairwise identity. This database contained 474 487 sequences. Pairwise comparison of sequences was achieved using Blast-2-Sequences . Multiple alignments were initially constructed using Clustal-W and manually edited using Seaview according to the guidelines of Bork and Gibson . Alignments were presented using the CHROMA tool . Hidden Markov model (HMM) searches of protein sequence databases used HMMER2 and an E-value inclusion threshold of 0.1. Domain-based analyses used SMART , Pfam and CDD . Detection of distantly related repeats used Prospero . Conservation of gene order in completely sequenced genomes was investigated using the COG database . Comparison of a conserved alignment block of Def1p orthologues with nrdb90 used MoST, the motif search tool , and parameters I = 80% and E = 0.05. Preliminary sequence data were obtained from the Institute for Genomic Research website at . RESULTS AND DISCUSSION : Archaeal homologues of Spt4p | Initial attempts to identify non-eukaryotic homologues of Spt4p using PSI-BLAST searches of protein sequence databases were unsuccessful. However, Spt4p homologues from incompletely sequenced eukaryotic genomes, such as D.discoideum, were identified from TBLAST-N searches of expressed sequence tag (EST) databases . The conceptual protein sequences were then used to query nr using PSI-BLAST. A database search using, as query, D.discoideum Spt4p, derived from ESTs AU038537 and AU073920, yielded significant similarity (E = 9 x 10 --3) to the DNA-directed RNA polymerase subunit E" (rpoE") from the hyperthermophilic archaeon Methanococcus jannaschii after one iteration. Sequence similarity extends over a N-terminal C4 zinc ribbon and a C-terminal alpha/beta-containing region. Similar rpoE" homologues occur in all other completely sequenced archaea, including a version in Sulfolobus acidocaldarius that is fused to DNA-directed RNA polymerase subunit E . Despite claims of a RNA polymerase architecture common to both archaea and eukarya , previous studies had identified neither archaeal counterparts of eukaryotic Spt4p nor eukaryotic versions of archaeal DNA-directed RNA polymerase subunit E". The identification of these two molecules as homologues resolves these discrepancies and further implicates archaeal subunit E" as a Spt4p-like transcription elongation factor. This might account for the apparent absence of subunit E" in RNAP complexes from Methanobacterium thermoautotrophicum , since it would be expected to bind archaeal NusG (which binds RNAP) rather than binding RNAP directly (see below). A novel NusG N-terminal (NGN) homology domain | The C-terminal regions of NusG and Spt5p contain one, and multiple, KOW motifs, respectively . A PSI-BLAST search with an N-terminal region of yeast Spt5p (amino acids 220 --380, thereby lacking a highly acidic region 1 --219) revealed additional significant similarity to NusG homologues [M.jannaschii NusG (E = 2 x 10 --4) and Pyrococcus horikoshii NusG (E = 6 x 10 --4)] after two rounds. These NGN domains appear to occur in all Spt5p and NusG homologues in archaea, bacteria and eukarya . Thus, Spt5p and NusG contain two distinct regions of homology: an NGN domain and one or more KOW motifs . Database searches using HMMs detected both an NGN domain and a KOW motif in NusG-like bacterial proteins RfaH . This is consistent with the known role of RfaH in regulating bacterial transcription elongation . The newly identified NGN domain in human Spt5p may possess an affinity for Spt4p. Yamaguchi et al. determined the Spt4p-binding region of Spt5p as amino acids 176 --313. This region overlaps both its NGN domain (amino acids 176 --267) and its first KOW motif (amino acids 272 --299). A natural corollary to the finding that an NGN --KOW region of Spt5p binds Spt4p is that their archaeal homologues, rpoE" and NusG, may also associate. An Spt4p /rpoE"-binding role for NGN --KOW cannot be a universally applicable function since bacteria lack apparent Spt4p orthologues. Further investigation of KOW motif sequences resulted in their identification in KIN17, a component of the ultraviolet (UV)-C response , eukaryotic homologues of human T54 (see also Pfam family PF00467), and in eukaryotic ribosomal S4 proteins . Finally, although there has been some disagreement in the literature concerning the number and position of KOW motifs in Spt5p (,), this study resulted in detection of five KOW motifs in fungal Spt5p and six KOW motifs in mammalian Spt5p . The functions and structures of KOW motifs remain enigmatic. From the known structure of the large ribosomal subunit , the KOW sequence motif of L24 occurs in three beta-strands within a larger src homology 3 (SH3) domain fold, which lies at the exit of the polypeptide tunnel. L24 interacts with several RNA domains in the ribosome, which is in agreement with the original proposal of KOW as an RNA-binding motif . Spt6p domain homologues | Little is known about individual functional domains of yeast Spt6p. Maclennan and Shaw identified a src homology 2 (SH2) domain in a C-terminal region whereas, more recently, Doerks et al. described a 'CSZ domain' encompassing most of the N-terminal remainder of Spt6p. The CSZ region encompasses two tandem helix --hairpin --helix (HhH) motifs likely to bind DNA . Although Spt6p homologues in eukaryotes other than fungi contain a readily identifiable ribosomal S1-like RNA-binding domain in a region C-terminal to the CSZ, pairwise alignment of Spt6p homologues demonstrates that the S1 domain is present also in fungi Spt6p (data not shown) . Eukaryotic Spt6p has previously been shown to be homologous, over the CSZ-S1 domains' region, to the Bordetella pertussis Tex (toxin expression) gene product (,). Tex is an essential and ubiquitous factor in bacteria and is hypothesised to regulate transcriptional processes . Iterative database searches revealed that the region of CSZ that is N-terminal to the tandem HhH motifs contains a domain predicted to possess a RNase H fold . For example, a search for conserved domains in Xylella fastidiosa Tex (GenInfo code 11278678) using CDD revealed a possible Pfam-FGGY-like RNase H fold domain (amino acids 330 --391; E = 1 x 10 --3). A PSI-BLAST search using this sequence as the query identified Synechocystis sp. PCC 6803 sll0832 as significantly similar to Tex (E = 3 x 10 --4 in round 2). Sll0832 is a member of the YqgF domain family of RNases that includes the eponymous Escherichia coli Yqgf. A reciprocal search with E.coli YqgF as the query yielded the expected significant similarity (E = 1 x 10 --3) with Bacillus subtilis Tex in four rounds. These findings demonstrate that YqgF-homologous domains occur in bacterial Tex orthologues and eukaryotic Spt6p orthologues within their CSZ regions. Thus, CSZ represents a domain and motif combination ('architecture') that is preserved in Tex /Spt6p homologues rather than being a single large domain. This is the first observation of a protein containing both YqgF and other domain types . Although previously thought to be absent in archaea , YqgF homologous domains are detectable in this kingdom. A PSI-BLAST search with the YqgF domain of Synechocystis sp. (strain PCC6803) sll0832 (amino acids 16 --141) identified M.thermoautotrophicum MTH839 amino acids 23 --129 as homologous in nine search rounds with E = 7 x 10 --4 . YqgF domains in Tex and TexL orthologues are likely to possess a nuclease function. The residues Asp (twice), Glu, and Ser or Thr are absolutely conserved in both Tex orthologues and YqgF-like proteins at positions that are thought to contribute to nuclease activity . The substrate of this Tex nuclease domain is tentatively suggested to be RNA since Tex negatively regulates transcription when overexpressed, and also since it contains a C-terminal RNA-binding (S1) domain . Spt6p paralogues and transcription elongation | It has been suggested that Spt6p is the eukaryotic orthologue of bacterial Tex. Although Spt6p is a Tex homologue, it is not the most sequence-similar Tex homologue in eukaryotes. The hypothetical proteins human FLJ10379, Drosophila melanogaster LD12377p /CG5253 and Caenorhabditis elegans ZK973.1 are significantly more sequence similar to bacterial Tex (similar35% pairwise sequence identity) than are Spt6p homologues (similar25%). Moreover, these three eukaryotic proteins share the same CSZ and S1 domain architecture and predicted catalytic residues in their YqgF domains as bacterial Tex . Thus, it is predicted that bacterial Tex and eukaryotic homologues such as human FLJ10379 are orthologous and may have comparable cellular functions. In this paper, human FLJ10379, D.melanogaster LD12377p /CG5253 and C.elegans ZK973.1 homologues, which from searches of EST databases appear to be widespread in eukaryotes, will be described as TexL (Tex-like) genes. The molecular functions of Tex and TexL orthologues are unknown. In some cases, the repeated co-occurrence of genes in prokaryotic genomes has been used to accurately predict function . Consequently, the genomic contexts of Tex orthologues were investigated using the COG database . Viewing these contexts demonstrated that the GreA transcription elongation factor (COG0782) was the most proximal 5' gene to Tex in four completely sequenced genomes. This is likely to be significant since in E.coli the genes are encoded on the same strand whereas in Vibrio cholerae, Haemophilus influenzae and Pasteurella multocida they are on complementary strands. Bacterial GreA is known to promote efficient RNA polymerase transcription elongation past template-encoded arresting sites . This suggests that bacterial Tex and eukaryotic TexL, in common with GreA, are transcription elongation factors. In addition, Tex was found to be the neighbouring gene to E.coli sprT homologues in four bacterial genomes: Lactococcus lactis (L86677), Streptococcus pyogenes (SPy0581), Bacillus halodurans (BH0532) and B.subtilis (ydcK) (all same strand). These bacterial sequences are homologous to eukaryotic proteins, including human ACRC, since they are found within five PSI-BLAST rounds using the human ACRC sequence (amino acids 411 --691) as query and an E-value inclusion threshold of 0.002. The ACRC gene maps to the Dystonia parkinsonism critical interval in Xq13.1 . It is inferred from their genomic co-occurrence with Tex that SprT and ACRC also function in transcription elongation. Three viral SprT homologues are known, in Mamestra configurata nucleopolyhedrovirus and Leucania separata nucleopolyhedrovirus. These are the only viral homo logues of eukaryotic transcription elongation factors known. Widespread conservation of a HExxH motif and His and Cys residues indicates that SprT homologues are metalloproteases . Homologues of Spt16 /SSRP1 (FACT) | The human orthologue of yeast Spt16p /CDC68p is one subunit of FACT, a heterodimer which is a chromatin-specific transcription elongation factor . A PSI-BLAST database search with yeast Spt16p /CDC68p as query revealed that it is a member of the metallopeptidase family M24: in the first search round, Staphylococcus aureus subsp. aureus N315 Xaa-Pro dipeptidase (GenInfo code 15927110) was found with E = 2 x 10 --5. Interestingly, Spt16p /CDC68p orthologues lack amino acids that are known to be essential for catalysis (data not shown). Thus, Spt16p /CDC68p is predicted to adopt the fold of the peptidase M24 family, but not possess its protease activity. The Spt16p /Cdc68p metalloprotease- homology domain is within an N-terminal region known to affect chromatin structure thereby inhibiting transcription . In the absence of catalytic residues, it might be thought that the molecular function of Spt16p /CDC68p is as a DNA-binding factor. This would be consistent with the observation that a Schizosaccharomyces pombe metallopeptidase M24 family member has been shown to preferentially bind curved DNA . However, as human Spt16 has been reported not to bind unmodified DNA its function still remains to be determined. The sequences of SSRP1 (single-stranded recognition protein 1), the second subunit of FACT, were also investigated for distant homology. Although no previously unknown SSRP homologues were detected, a tandem repeat within all animal SSRP1s was detected . For example, a search for repeats in D.melanogaster SSRP1 using Prospero revealed significant internal sequence similarity (P = 1.6 x 10 --3). This was consistent with the results of PSI-BLAST searches. For example, a search with S.cerevisiae Ynl206c, using an E-value inclusion threshold of 0.002, indicated the presence of a second repeat in Xenopus laevis SSRP1 (DUF87) with E = 3.2. It is not apparent from this analysis what these repeats' function might be. However, it is unlikely to be DNA binding since this function is conveyed by the high-mobility group (HMG) domain present in most of the SSRP1 homologues. An alternative hypothesis is that the repeats in SSRP1 mediate its affinity for Spt16, the other FACT subunit. It is notable that although the isolated HMG domain of SSRP1 binds DNA, it cannot do so in the full-length molecule except when in the presence of Spt16 . Thus, one of the functions of the SSRP1 repeats may be to regulate its multidomain conformational change that is induced by Spt16-binding. Elongator subunits | The histone acetyltransferase complex holo-elongator can be isolated as two subcomplex factors that associate with RNAP-II (,). For the first subcomplex, Elp1p and Elp2p contain WD40 repeats, whilst Elp3p is a histone H3 and H4 acetyltransferase and possible histone demethylase . The functions of the three proteins, Elp4p, Elp5p and Elp6p, in the second subcomplex remain poorly understood. The sequence-based approaches used in this study, however, demonstrate that both Elp4p and Elp6p are inactive homologues of P-loop ATPases /GTPases. Likely orthologues of yeast Elp4p were previously identified in vertebrates and invertebrates . PSI-BLAST searches with these Elp4p orthologues provide evidence that Elp4p are ATPase homologues. For example, a search with the Arabidopsis orthologue (GenInfo code 12321866) reveals significant similarity (E = 4 x 10 --4) in two rounds to the ATPase domain in the X.fastidiosa 9a5c radA-like protein . Saccharomyces cerevisiae Elp6p is not apparently similar to any sequence in the nr, although an orthologue is readily apparent from a search of the Candida albicans unfinished genome (TBLAST-N E = 9 x 10 --18 using ). Detailed searches did reveal a family of orthologues in other eukaryotes, such as human FLJ20211, S.pombe SPBC3H7.10, D.melanogaster diminished discs (DID), D.discoideum ORF (GenInfo code 12007287), and Arabidopsis thaliana F28M11.10, and two lines of evidence indicated that these represent Elp6p orthologues. First, a PSI-BLAST search found significant similarity between the A.thaliana F28M11.10 sequence and C.albicans Elp6p with E = 9 x 10 --3 after three rounds. This search used amino acids 21 --242 of F28M11.10, but corrected to reflect sequence differences manifest between F28M11.10, and ESTs AV552362 and AU226546. The search also employed a nrdb90 database that was supplemented by the C.albicans Elp6p orthologue. Secondly, the family is represented in all free-living eukaryotic genomes sequenced to date except S.cerevisiae. Consequently, Elp6p is the most likely candidate as the S.cerevisiae orthologue of this family. Surprisingly, proposed Elp6p family members are also likely homologues of Elp4p. A PSI-BLAST search with D.melanogaster DID (amino acids 13 --245) and an E-value inclusion threshold of 5 x 10 --3 revealed marginal similarity (E = 6 x 10 --3) in five rounds to a human hypothetical protein (GenInfo code 15214765) that is, in turn, homologous to Elp4p. This search was undertaken using nrdb90 supplemented by Elp6p orthologues' sequences from C.albicans, Lycopersicon esculentum and Ciona intestinalis taken from EST and genome sequencing projects. Construction of a multiple alignment of Elp4p and Elp6p homologues with ATPases demonstrates that the elongator components lack the phosphate-binding P-loop . This implies that these proteins lack ATPase activities. In the absence of ATPase activities the functions of Elp4p and Elp6p orthologues remain to be clarified. Whatever these functions might be, it is possible that archaea possess analogous functions since ATPase homologues that are similar in sequence to Elp4p and Elp6p and have substitutions within their P-loops are apparent for example in Archaeoglobus fulgidus (AF0352, AF0518 and AF1172), Halobacterium sp. NRC-1 (HtlC; GenInfo code 15790668), P.horikoshii (PH1120) and Thermoplasma acidophilum (Ta0084). No orthologues of Elp5p (also known as YHR187w and Iki1p) outside of the fungi are readily apparent from BLAST database searches using composition-dependent statistics . However, a reciprocal PSI-BLAST search with a D.discoideum protein sequence (GenInfo code 19570052) and composition-dependent statistics revealed significant similarity (E = 2 x 10 --4) to S.pombe Elp5p (SPBC18E5.05c) in three rounds. Such searches determined that Elp5p orthologues are present across the eukaryotes, in mammals, Drosophila and C.elegans . Little is known of these proteins except that expression of Rai12, the mouse orthologue gene, is induced by retinoic acid . However, the identification of likely Elp5p orthologues should assist in the investigation of this Elongator subunit in mammals. Homologues of S.cerevisiae Def1p | Rad26p facilitates UV-light-induced DNA damage and appears to protect RNAP-II from degradation during the repair process . In contrast, the association of Def1p with Rad26p in chromatin appears to enable ubiquitination of RNAP-II and leads to its proteolysis by the proteasome . As noted elsewhere , UV-induced RNAP-II ubiquitination and degradation has been observed in fungi and mammals, yet Def1p orthologues have not been detected in standard protein sequence databases. In order to search for candidate Def1p orthologues, the S.cerevisiae Def1p sequence was compared with unfinished genome, EST and protein sequences using the NCBI BLAST web resources. This resulted in the identification of likely orthologues from four fungi: C.albicans (on contig 6 --2503), Aspergillus fumigatus (on fragment 2283), S.pombe (gene SPBC354.10) and Coccidioides immitis (encoded in ESTs BF251037 and BF252062). A MoST search using these sequences identified the N-terminal of two CUE domains in mouse Enhancer-trap-locus-1 (Etl-1; amino acids 271 --300) as being similar to these sequences with E = 4.8 x 10 --2 . CUE domains in Etl-1 were identified using Pfam . Additionally, C.albicans Def1p was the highest scoring sequence, albeit with a non-significant E-value (0.74), in a search of known sequences using the SMART CUE domain HMM. These marginal similarities may not have provided sufficient evidence for the presence of a CUE domain in Def1p orthologues, except for the existence of strong functional similarities between yeast Def1p and Cue1p in the literature. Cue1p is known to recruit the soluble ubiquitin-conjugating enzyme (UBC) Ubc7p to the endoplasmic reticulum (ER) membrane prior to the ubiquitination of products that undergo ER-associated degradation . Def1p coordinates the ubiquitination of RNAP-II, presumably when transcription is stalled at a site of DNA damage . Consequently, similarities in both sequence and function indicate that these two proteins contain a conserved CUE domain. Like the CUE domain in Cue1p, the predicted Def1p CUE domain may recruit UBC E2 to the transcription complex. Interestingly, among mammalian CUE domains, the yeast putative Def1p CUE domains are most similar to those in Etl-1, a member of the SNF2 /SWI2 family of transcriptional regulators. Since yeast Rad26p, the interaction partner of Def1p, is also a member of this family, the domain architecture arising from the conceptual fusion of Def1p and Rad26p is almost equivalent to that of Etl-1. Using the concept of Etl-1 as a 'Rosetta Stone protein' , this suggests that mammalian Etl-1, whose cellular function remains ill determined, may lie in regulating transcription elongation. Conclusions: evolution of eukaryotic transcription elongation factors | These data suggest that of all the modern components of the eukaryotic transcription elongation machinery only NusG /Spt5 and RNAP itself were present in the last common ancestor of the three kingdoms of cellular life, archaea, eubacteria and eukarya. The eukaryotic transcription elongation machinery appears to have appropriated components from other cellular processes such as protein degradation (Spt16p /Cdc68p is a metalloprotease homologue), ATP-dependent chromatin remodelling (Elp4p and Elp6p are ATPase homologues) and nucleic acid hydrolysis (Spt6p contains a YqgF nuclease domain homologue). In each of these three cases appropriation is associated with apparent losses in enzymatic activities, with substitutions of known active site residues. Apart from NusG /Spt5, the only factor that appears to have survived in situ since the common ancestor of eukaryotes and archaea is Spt4 /rpoE", whereas the bacteria and eukaryotes only otherwise share Tex /TexL and possibly sprT. Eukaryotic Spt6 is a specialisation of bacterial Tex with accretions of a single SH2 domain in fungi and a pair of consecutive SH2 domains in animals, and with a loss of YqgF nuclease activity. Eukaryotes have also evolved a transcription elongation apparatus that has no demonstrable homologues in the prokaryotes. This includes subunits of the Paf1 complex ( --) and domains in SSRP1 /Pob3p and Def1p that are only currently found elsewhere in other eukaryotic proteins. Comparison of eukaryotic Spt5 with prokaryotic NusG shows that it too has acquired structural additions. It has accreted many additional domains, in particular multiple KOW motifs , since it diverged from the archaeal and bacterial NusG lineages. This may reflect the numerous physical interactions with the eukaryotic-specific Paf1 and Spt16p /CDC68p /Pob3p /FACT complexes . Figure 1 | Multiple sequence alignment of eukaryotic Spt4p and archaeal DNA-directed rpoE" represented using CHROMA and a 75% consensus. Multiple sequence alignment of eukaryotic Spt4p and archaeal DNA-directed rpoE" represented using CHROMA and a 75% consensus. Cys residues that are likely to bind Zn2+ are shown in white-on-red, whilst conserved Ser /Thr residues that might act as phosphorylation sites are shown as white-on-blue. Secondary structures predicted at expected accuracies of >82% (E, H) or >72% (e, h) are indicated below the alignment (E /e, extended or beta-strand structure; H /h, alpha-helical structure). Numbers in parentheses represent amino acids that have been excised from the alignment. Amino acid numbers (or, in italics, nucleotide numbers for ESTs) and GenInfo numbers are given following the alignment. Species abbreviations: Af, A.fulgidus; At, A.thaliana; Bm, Brugia malayi; Ce, C.elegans; Dd, D.discoideum; Hg, Heterodera glycines (soybean cyst nematode); Hsp, Halobacterium sp. NRC-1; Mj, M.jannaschii; Mm, Mus musculus; Mt, M.thermoautotrophicum; Ph, P.horikoshii; Sc, S.cerevisiae; Sp, S.pombe; Ss, Sulfolobus solfataricus; St, Sulfolobus tokodaii; Ta, T.acidophilum. Figure 2 | Schematic representation of the domain architectures of molecules discussed in this paper, approximately to scale. Schematic representation of the domain architectures of molecules discussed in this paper, approximately to scale. Domain abbreviations: C2HC, uvs-2-like C2HC zinc finger; G, G-patch domain; HhH, helix -- hairpin --helix motifs; K, KOW motif; R, Rpb7-like N-terminal domain; S1, ribosomal protein S1-like RNA-binding domain. Species abbreviations: Bp, B.pertussis; Ec, E.coli; Hs, Homo sapiens; Mj, M.jannaschii; Sa, S.acidocaldarius; Sc, S.cerevisiae. Figure 3 | Multiple sequence alignment of NGN domains represented using CHROMA and a 75% consensus. Multiple sequence alignment of NGN domains represented using CHROMA and a 75% consensus. Secondary structures predicted at expected accuracies of >82% (E, H) or >72% (e, h) are indicated below the alignment (E /e, extended or beta-strand structure; H /h, alpha-helical structure). Numbers in parentheses represent amino acids that have been excised from the alignment. GenInfo and amino acid numbers are given following the alignment. Species abbreviations: Ap, Aeropyrum pernix; Ce, C.elegans; Ec, E.coli; Mj, M.jannaschii; Ps, Pseudomonas syringae pv. maculicola; Sc, S.cerevisiae; Sp, S.pombe; Tt, Thermus thermophilus; Tv, Thermoplasma volcanium. Figure 4 | Multiple sequence alignment of YqgF-homologous domains, including those in Tex and Spt6p orthologues, represented using CHROMA and a 75% consensus. Multiple sequence alignment of YqgF-homologous domains, including those in Tex and Spt6p orthologues, represented using CHROMA and a 75% consensus. Predicted active site residues in YqgF nucleases are shown as white-on-red. Secondary structures predicted at expected accuracies of >82% (E, H) or >72% (e, h) are indicated below the alignment (E /e, extended or beta-strand structure; H /h, alpha-helical structure). Species abbreviations: Aa, Aquifex aeolicus; At, A.thaliana; Bp, B.pertussis; Bs, B.subtilis; Cb, C.briggsae; Ce, C.elegans; Cj, Campylobacter jejuni; Cm, Chlamydia muridarum; Dm, D.melanogaster; Dr, Deinococcus radiodurans; Hi, H.influenzae; Hs, H.sapiens; Mt, Mycobacterium tuberculosis; Mth, M.thermoautotrophicum; Sc, S.cerevisiae; Sp, S.pombe; Ssp1, Synechocystis sp. PCC6803; Ssp2, Synechocystis sp. PCC7002; Tm, Thermotoga maritima; Tp, Treponema pallidum; Xf, X.fastidiosa. GenInfo numbers and amino acid numbers are: Spt6p Sc: 6321552, 750 --858; Spt6p Sp: 11359284, 590 --687; T8F5.22 At: 7487801, 772 --890; F16M19.22 At: 10092249, 506 --625; Supt6h Hs: 1136386 779 --894; Emb5 Cb: 1669619, 747 --862; Spt6 Dm, 7290693, 783 --898; FLJ10379 Hs: 11430242, 194 --293; CG5253 Dm: 7300889, 526 --627; Tex Bp: 2501166, 344 --443; HI0568 Hi: 2501167, 327 --424; Tex Tp: 7445176, 332 --426; sll0832 Ssp1: 7469941, 16 --103; sll0832 Ssp2: 13924476, 2 --89; DR2509 Dr: 11136112, 16 --113; YrrK Bs: 6226496, 1 --103; sll1547 Ssp1: 6226451, 5 --105; Aq_1498 Aa: 6226434, 1 --99; XF2227 Xf: 1135885, 11 --111; TC0456 Cm: 11278851, 8 --110; Rv2554c Mt: 6226486, 22 --123; Cj0635 Cj: 11135891, 1 --97; ORF Tm: 7462233, 12 --100; MTH839 Mth: 7482690, 23 --115. Figure 5 | Multiple sequence alignment of SprT homologues represented using CHROMA and a 75% consensus. Multiple sequence alignment of SprT homologues represented using CHROMA and a 75% consensus. His and Cys residues that are likely to bind divalent metal ions are shown in white-on-red, whilst a likely catalytic Glu residue is shown as white-on-blue. Secondary structures predicted at expected accuracies of >82% (E, H) or >72% (e, h) are indicated below the alignment (E /e, extended or beta-strand structure; H /h, alpha-helical structure). Numbers in parentheses represent amino acids that have been excised from the alignment. GenInfo and amino acid numbers are given following the alignment. Species abbreviations: Bs, B.subtilis; Ce, C.elegans; Dm, D.melanogaster; Hi, H.influenzae; Hs, H.sapiens; Mcn, M.configurata nucleopolyhedrovirus; Sp, S.pombe; Spy, S.pyogenes; TM4, Mycobacteriophage TM4; Vc, V.cholerae. Figure 6 | Multiple sequence alignment of SSRP1 repeats represented using CHROMA and a 75% consensus. Multiple sequence alignment of SSRP1 repeats represented using CHROMA and a 75% consensus. The top two tiers of sequences represent the two repeats in SSRP1 orthologues. Secondary structures predicted at expected accuracies of >82% (E, H) or >72% (e, h) are indicated below the alignment (E /e, extended or beta-strand structure; H /h, alpha-helical structure). Species abbreviations: Ce, C.elegans; Dm, D.melanogaster; Hs, H.sapiens; Sc, S.cerevisiae; Sp, S.pombe; Xl, X.laevis; Zm, Zea mays. Amino acid numbers are given following the alignment and GenInfo numbers are: SSRP1 Hs: 730840; SSRP1 Dm: 12644386; SSRP1 Xl: 4586285; SSRP1 Ce: 1174454; C32F10.5 Ce: 7496859; YML069W Sc: 2497082; SPBC609.05 Sp: 2497082; SSRP1 Zm: 8920409; YNL206C Sc: 732190; SPAC6G9.03c Sp: 2842694. Figure 7 | Multiple sequence alignment of Elp4p orthologues (top five sequences) and Elp6p orthologues (next five sequences) with archaeal probable ATPases (last three sequences) represented using CHROMA and a 75% consensus. Multiple sequence alignment of Elp4p orthologues (top five sequences) and Elp6p orthologues (next five sequences) with archaeal probable ATPases (last three sequences) represented using CHROMA and a 75% consensus. Walker A and B motifs are double underlined beneath the alignment. Numbers in parentheses represent amino acids that have been excised from the alignment. Amino acid limits and GenInfo numbers are shown at the end of the alignment. Species abbreviations: Af, A.fulgidus; At, A.thaliana; Ce, C.elegans; Dm, D.melanogaster; Ec, Encephalitozoon cuniculi; Mj, M.jannaschii; Mm, M.musculus; Hs, H.sapiens; Sc, S.cerevisiae; St, S.tokodaii. Figure 8 | Multiple sequence alignment of the C-terminal regions of Elp5p homologues represented using CHROMA and a 75% consensus. Multiple sequence alignment of the C-terminal regions of Elp5p homologues represented using CHROMA and a 75% consensus. Sequence conservation is detectable throughout these likely orthologues but accurate multiple sequence alignment is problematic. Numbers represent amino acids that have been excised from the alignment. Species abbreviations: At, A.thaliana; Ca, C.albicans; Cb, C.briggsae; Ce, C.elegans; Ci, C.intestinalis; Dd, D.discoideum; Dm, D.melanogaster; Hs, H.sapiens; Sc, S.cerevisiae; Sp, S.pombe; Xl, X.laevis; Zm, Z.mays. GenInfo numbers and amino acid numbers (when known) are: Elp5 /Iki1p Sc: 6321981, 257 --309; At2g18410 At: 15224168, 303 --392; EST Zm: 6721029 (conceptual translation); dd_00450 Dd: 19570052, 201 --253; EST Ci: 16853181 (conceptual translation); SPBC18E5.05c Sp: 19112643, 252 --314; CG2034 Dm: 7292214 210 --262; Rai12 Hs: 20086425, 244 --307; EST Xl: 13166929 (conceptual translation); W09B6.4 Ce: 17536805, 295 --361. Figure 9 | Multiple sequence alignment of CUE domains in Def1p, Enhancer-trap-locus-1 (Etl-1) orthologues and yeast Cue1p, represented using CHROMA and a 75% consensus. Multiple sequence alignment of CUE domains in Def1p, Enhancer-trap-locus-1 (Etl-1) orthologues and yeast Cue1p, represented using CHROMA and a 75% consensus. Species abbreviations: Af, A.fumigatus; Ca, C.albicans; Ci, C.immitis; Dr, Danio rerio; Mm, M.musculus; Sc, S.cerevisiae; Sp, S.pombe; Xl, X.laevis. GenInfo numbers and amino acid numbers (when known) are: Def1p Sc: 6322796, 21 --62; Def1p Sp (SPBC354.10): 19112026, 77 --119; Etl1 Mm: 1082208, 271 --313 and 362 --404; Cue1p Sc: 6323920, 65 --106. Backmatter: PMID- 12202755 TI - Partial reconstitution of human RNase P in HeLa cells between its RNA subunit with an affinity tag and the intact protein components AB - An RNA affinity tag was incorporated into the RNA subunit of human nuclear RNase P. The tagged RNA assembled with the protein components of RNase P inside HeLa cells to generate an active enzyme. Because of the specificity of the RNA tag to streptavidin, the reconstituted complex could be separated from the native enzyme and other ribonucleoproteins (particularly RNase MRP) by streptavidin agarose chromatography and could be recovered by the eluting agent, biotin. A mutant, tagged RNase P RNA, whose P3 domain was partially replaced, could not reconstitute with the proteins to yield an active enzyme. The P3 domain, therefore, is critical for the structure and function of RNase P. Keywords: INTRODUCTION : Ribonuclease P (RNase P) is an essential enzyme required for the processing of the 5' termini of precursor tRNAs in all living organisms . The eubacterial RNase P consists of one RNA subunit and one protein subunit. The RNA subunit retains its catalytic function in a high concentration of salt in vitro in the absence of the protein subunit . However, the protein component facilitates catalysis under physiological salt concentrations and is required for efficient tRNA processing in vivo . The currently characterized eukaryotic (either Saccharomyces cerevisiae or Homo sapiens) nuclear RNase P possesses numerous protein subunits and one RNA subunit (,). Unlike the eubacterial enzyme, the combination of human RNase P RNA transcribed in vitro and its purified recombinant proteins expressed in Escherichia coli when mixed together did not achieve an active enzyme (data not shown). Total, reliably reproduced, reconstitution in vitro of the RNA and protein subunits of the eukaryotic RNase P has not been reported so far. An RNA aptamer (S1) has been developed recently to bind specifically to streptavidin . The RNA subunit of yeast nuclear RNase P was tagged with the streptavidin-binding element and the tagged RNA subunit was shown to be able to reconstitute an active enzyme with the protein subunits in vivo. The enzyme could be isolated from column chromatography by elution with biotin because the S1-binding ability to streptavidin agarose beads is much lower than that of biotin . In this study, we have shown that the RNA affinity tag also works well in a mammalian system, and the aptamer-tagged human nuclear RNase P could be immobilized to streptavidin agarose from crude cell extracts and eluted with biotin under mild conditions. In addition, we have applied this method to the study of an essential protein-binding domain (P3) of the RNase P RNA from HeLa cells. MATERIALS AND METHODS : Materials | Restriction and modification enzymes were obtained from New England Biolabs. Radiochemicals were obtained from Amersham. Oligonucleotides were synthesized by the Keck Facility at Yale University. Cloning and mutagenesis. Four oligonucleotides, oligo 1 (AAAAG GGAGT CGACC GACCA GAATC ATGCA AGTGC GTAAG ATAGT CGCGG GCCGG GGGCG), oligo 2 (TATTA TGTGC GTCTA CATCT AGACT CATAA AAGGC CCCgg cc), oligo 3 (GGGGC CTTTT ATGAG TCTAG ATGTA GACGC ACATA ATACG CCCCC GGCCC GCG), oligo 4 (TATCT TACGC ACTTG CATGA TTCTG GTCGG TCGAC TCCCT TTTgg cc) were annealed and ligated with pUC /H1 cut by ApaI. The bases represented by lower case letters facilitate cloning into the sticky ends of the restriction enzymes. The resulting plasmid was then digested by EcoRI and SalI, and the similar400 bp fragment was ligated to the plasmid pDeltaEGFP which is derived from pEGFP-N1 (Clontech) after removal of the gene encoding EGFP. The final plasmid was named as pDeltaEGFP --H1S1 containing CMV IE promoter, the gene for H1S1 (H1 RNA with S1 tag), kanamycin and neomycin selection markers, and colE1 and SV40 replication origins. To generate the deletion of P3 domain of H1 RNA, two oligonucleotides, P3forward (AGAAT TCATA GGGCG GAGGG AAGCT CATCA GTGGG GATGT CCCTT GGGAA GGT) and P3reverse (AAGGA TCCAA TGGGC GGAGG AGAGT AGT) were used to perform PCR with pDeltaEGFP --H1S1 as template. The PCR fragment was cloned into pDeltaEGFP after being digested by EcoRI to get the plasmid pDeltaEGFP --H1S1DeltaP3 . The coding regions for H1S1 and H1S1DeltaP3 in the plasmids pDeltaEGFP --H1S1 and pDeltaEGFP --H1S1DeltaP3 were confirmed by DNA sequencing with T7 Sequenase (Amersham). Transfection of HeLa cells | Transfection of the plasmids into HeLa S3 cells was done with SuperFect transfection reagent (Qiagen) according to the manufacturer's handbook. The cells were incubated for 24 h after transfection, washed with PBS, harvested and stored at --80C. Purification of S1-tagged human RNase P | The purification was performed at 4C as described with minor modifications. The pellets of cells (similar5 x 107) were resuspended in 0.5 ml of lysis buffer (50 mM HEPES, pH 8.0, 10 mM MgCl2, 100 mM KCl, 1 mM DTT, 10% glycerol) with the Complete protease inhibitor cocktail (Roche). The mixture was homogenized by passing it 15 times through a 25G7 /8 needle fitted into a 1 ml syringe. The lysate was centrifuged at 16 000 g for 30 min. The S16 cell extract was incubated and rotated with 50 microl of streptavidin agarose (Gibco BRL) overnight in a 15 ml tube. The beads were spun down to remove the supernatant and washed five times with 5 ml of lysis buffer, for 15 min each. The beads were then transferred to an Ultrafree-MC centrifugal filter unit, 0.45 microm in size (Millipore), and further washed three times with 400 microl of lysis buffer. Elution was accomplished by incubating with 100 microl of lysis buffer containing 0.5 or 5 mM biotin for 1 h. Northern blot analysis | Aliquots for the streptavidin agarose purification were extracted from column fractions by phenol and precipitated by ethanol. The samples were separated by a 3% agarose gel and then transferred onto a positively charged nylon membrane (Roche). The northern blot was performed with Rapid-hyb buffer (Amersham Pharmacia Biotech), using a 5'-terminally [32P] phosphorylated DNA oligonucleotide complementary to positions 319 --340 of human RNase P RNA or 245 --265 of human RNase MRP RNA. Western blot analysis | Samples were separated in 13% polyacrylamide /0.1% SDS gels, electrotransferred to a nitrocellulose filter and immunoblotted with a 1:100 dilution of the primary antibody serum. As a secondary antibody, a 1:5000 dilution of goat anti-rabbit IgG antibody was used. Blots were washed and antibody --antigen complexes were visualized with ECL-Plus detection reagent (Amersham Pharmacia Biotech), following the manufacturer's instructions. The membrane was stripped of bound antibodies and reprobed with others. Enzyme activity assays | RNase P enzyme activities were assayed in 1x PA buffer (50 mM Tris --HCl, pH 7.5, 10 mM MgCl2 and 100 mM NH4Cl) at 37C. The substrate RNA, E.coli precursor tRNATyr (pTyr), was transcribed in vitro in the presence of [alpha-32P]GTP, purified on a 7 M urea 5% polyacrylamide gel, and used at a final concentration of 100 nM (2000 c.p.m. per reaction). Figure 1 | The secondary structure of human nuclear RNase P from HeLa cells. The secondary structure of human nuclear RNase P from HeLa cells. The P3 domain nucleotides are shown in hollow letters. The restriction cleavage site for ApaI is indicated as an arrow. The S1 aptamer (rectangle) with an extra stem --loop was inserted into the gene for H1 RNA between position +207 and +214 in the original sequence in the construct of pDeltaEGFP --H1S1. The main part of the P3 domain (from +30 to +67) was replaced additionally by a tetraloop AUGU (see circle) in the construct of pDeltaEGFP --H1S1DeltaP3. P12 extends from position 155 to position 240 . RESULTS : Cloning and transfection of the RNA subunit | Helix P12 of H1 RNA, the RNA subunit of human nuclear RNase P, is not conserved in vertebrates . Accordingly, the S1 aptamer (shown in the rectangle in Fig. ) with an extra stem --loop was incorporated between the nucleotides at position +207 and +214 of the native H1 RNA to allow for function of the whole H1 RNA molecule. The first nucleotide cloned is the 5' G at position 208 and the last is the 3' G at position 213 (see Fig. for tag sequence). The S1-tagged H1 (H1S1) gene was placed into a mammalian expression vector pDeltaEGFP which is a truncated version of pEGFP-N1 (see Materials and Methods). The mutant, H1S1DeltaP3, which has a deletion of the main part of the P3 domain of H1 RNA , was also cloned into the pDeltaEGFP vector. After transfection, either with pDeltaEGFP --H1S1 or pDeltaEGFP --H1S1DeltaP3, the HeLa cells had a growth phenotype different from cells without transfection or transfected with pEGFP-N1. The new cells transfected with H1S1 RNA or H1S1DeltaP3 RNA grew quite slowly (doubling time much longer than 24 h, the usual doubling time of these cells) and similar10% of them died and were washed away by PBS washes (see Materials and Methods). The slow growth may be caused by the streptavidin tag in H1 RNA. Purification by streptavidin agarose | The crude extracts (S16) of transfected cells were bound to streptavidin agarose and extensively washed with lysis buffer (Materials and Methods), and then eluted with lysis buffer with 0.5 mM and then 5 mM biotin. The eluates were analyzed for the presence of H1 RNA or MRP RNA by northern blot hybridization as well as for the activity of RNase P. The H1S1 transfected cell extract contained the S1-tagged enzyme. The enzyme bound to streptavidin agarose and was eluted with 5 mM biotin . The one-step purification fold resulted in an similar410-fold enrichment of the RNase P activity, which is close to the 510-fold purification achieved by the classic three-step purification [DEAE Sepharose, glycerol gradient and FPLC monoQ column ]. However, for the mutant H1S1DeltaP3 RNA, the RNA was isolated after chromatography (see below) but no active enzyme was detected after the elution. Subsequent to the streptavidin agarose column, the RNase P with the tagged RNA is still impure because of contaminating proteins that might still bind to H1 RNA or to its protein subunits. The tagged RNA enzyme was then subjected to further chromatography as indicated by the conventional method for RNase P purification . The tagged RNA enzyme purification behavior is shown in Table , which shows a purification of 12 000-fold. The final steps, DEAE Sepharose, FPLC mono Q and Superose 6 and the streptavidin column can all be achieved in 1 week, though the enzyme is not yet pure as judged by SDS gel electrophoresis of the proteins (data not shown). The conventional method, including glycerol gradients, without the streptavidin column, results in a purification factor of 2300 and takes 1 month . Northern blot hybridization | Northern blot analysis of the aliquots from the purification step revealed that both H1S1 and H1S1DeltaP3 were found in their eluates of 5 mM biotin from the agarose beads, respectively. MRP RNA, which normally associates with RNase P in the early steps of its purification, was not present in the eluates. It should be noted that the expression of internal H1 RNA was almost completely inhibited in the cells transfected with pDeltaEGFP --H1S1 (Fig. , lane 3), but not in those transfected with pDeltaEGFP --H1S1DeltaP3. Western blot hybridization | To determine whether there are protein subunits in the samples eluted from streptavidin columns, western blots were done to detect eight of the known protein components (hPop1, Rpp40, Rpp38, Rpp30, Rpp29, Rpp25, Rpp21 and Rpp20) of human nuclear RNase P by their cognate antibodies ( --). As shown in Figure , all the proteins could be detected by their cognate antibodies in the 5 mM biotin eluate of cells transfected with pDeltaEGFP --H1S1, though there were two proteins (Rpp38 and Rpp29) that showed weak signals. This result is anticipated if the tagged RNA is part of an active enzyme. However, none of the proteins could be detected in the corresponding sample from the cells transfected with pDeltaEGFP --H1S1DeltaP3. Figure 2 | RNase P activity assay of fractions from the purification step of streptavidin agarose chromatography. RNase P activity assay of fractions from the purification step of streptavidin agarose chromatography. Reaction conditions are detailed in Materials and Methods. (A) HeLa cells transfected with pDeltaEGFP --H1S1. Lane 1, no enzyme. Lane 2, S16 extract. Lane 3, the supernatant after rotation with streptavidin agarose overnight. Lane 4, the streptavidin column flow through washed with lysis buffer. Lane 5, the eluate (0.5 mM biotin). Lane 6, the eluate (5 mM biotin). Lane 7, the agarose beads assayed after eluates. Lane 8, RNase P purified by conventional methods. (B) The same procedures except extracts of HeLa cells were transfected with pDeltaEGFP --H1S1DeltaP3. Figure 3 | Northern blot analysis of fractions from the purification step of streptavidin agarose (Materials and Methods). Northern blot analysis of fractions from the purification step of streptavidin agarose (Materials and Methods). (A) Northern blot with an oligonucleotide probe against RNase P RNA. The arrowhead indicates H1 RNA and the arrow indicates H1S1 RNA. (B) Northern blot with an oligonucleotide probe against MRP RNA. The left part of each panel is from HeLa cells transfected with pDeltaEGFP --H1S1 whereas the right part is from cells transfected with pDeltaEGFP --H1S1DeltaP3. Lane 1, H1 RNA transcribed in vitro by T7 RNA polymerase. Lane 2, H1S1 RNA transcribed in vitro by T7 RNA polymerase. Lane 3, MRP RNA transcribed in vitro by T7 RNA polymerase. Lane 4, S16 extract of HeLa cells transfected with pDeltaEGFP --H1S1 (left) or pDeltaEGFP --H1S1DeltaP3 (right). Lane 6, flow through washed with lysis buffer. Lane 7, the eluate (0.5 mM biotin). Lane 8, the eluate (5 mM biotin). Lane 9, agarose beads after eluates. The arrow in (B) indicates MRP RNA and the dots indicate fragments of MRP RNA. Figure 4 | Western blot analysis (Materials and Methods) of fractions from the purification step of streptavidin agarose. Western blot analysis (Materials and Methods) of fractions from the purification step of streptavidin agarose. The blots have been probed with antibodies to the various proteins listed between the columns of figures. The left part is from HeLa cells transfected with pDeltaEGFP --H1S1 whereas the right part is from cells transfected with pDeltaEGFP --H1S1DeltaP3. Lane 1, S16 extract. Lane 2, supernatant after rotation with streptavidin agarose overnight. Lane 3, flow through washed with lysis buffer. Lane 4, the eluate (0.5 mM biotin). Lane 5, the eluate (5 mM biotin). Lane 6, agarose beads after eluates. Table 1 | Purification profile of H1S1-tagged human RNase P from HeLa cells transfected with pDeltaEGFP --H1S1 DISCUSSION : We have shown that the S1-tagged H1 RNA could reconstitute with RNase P proteins in vivo to form an active enzyme. The mutant, H1S1DeltaP3, the P3 domain of which was deleted, did not yield any active enzyme. The S1-tagged enzyme could easily be separated from the native enzyme by a one-step streptavidin agarose column with a 410-fold purification. This purification achieves what a three-step conventional method does for human nuclear RNase P . Eight of the protein components could be detected in the eluted sample. Further purification could be conducted by DEAE Sepharose, FPLC monoQ and Superose 6 . At least five protein subunits of human RNase P, namely Rpp29, Rpp30, Rpp38, hPop1 and hPop5, have been shown to be associated with RNase MRP ( --). The overlap of subunit composition and the difficulty in separating easily RNase P from RNase MRP raised the possibility that the two ribonucleoproteins might be in the same complex in vivo. The eluate that contains RNase P from streptavidin agarose did not contain the MRP RNA, indicating that the RNase P and MRP were very likely not associated with each other in one complex. (We assume that lack of MRP RNA is an indication that no functional RNase MRP is in the fractions we show in the figure.) We could not exclude completely the possibility that the S1 tag affects the interaction between RNase P and MRP in vivo, if there is any. The P3 domain is conserved in nuclear RNase P and MRP in eukaryotic cells (,). This domain may play a key role in the entry of RNase P RNA and MRP RNA to the nucleolus in normal rat kidney epithelial cells (,). The P3 domain of yeast nuclear RNase P RNA is an essential protein-binding domain . The P3 domain of human nuclear RNase P RNA interacts with Rpp21, Rpp29, Rpp30 and Rpp38 in vitro by UV cross-linking , suggesting that these proteins are bound to the RNA subunit of the enzyme. The H1S1DeltaP3 RNA was unable to assemble with the protein subunits of RNase P to produce an active enzyme. This result indicated the importance of the P3 domain of RNase P in the assembly and function of nuclear RNase P. There are some limits on the tagged RNA purification method. The prime concern is that a stable HeLa cell strain could not be obtained by transfection with pDeltaEGFP --H1S1. A similar event happened with plasmid pmU6-H1S1 using a mouse U6 promoter and pEGFP --H1S1 (data not shown). This phenomenon might be connected to the fact that H1S1 RNA expression can inhibit native H1 RNA expression and, therefore, must have effects on cell survival. The streptavidin tag method offers further opportunities for studying the roles of particular segments of the RNA in enzyme function, and for purification of RNase P free of RNase MRP. CONCLUSION : A streptavidin tag has been inserted into a segment of H1 RNA, the RNA component of human nuclear RNase P, and has facilitated the purification of an intact enzyme from crude cell extracts. This simple procedure has yielded an enzyme that is free of human RNase MRP and allows further steps in purification. A streptavidin tag of H1 RNA that is missing the deleted P3 helix does not yield RNase P. The method is therefore useful for testing different parts of H1 RNA for function and as a part of a purification procedure. Backmatter: PMID- 12202771 TI - Challenging artificial genetic systems: thymidine analogs with 5-position sulfur functionality AB - Eight different polymerases, chosen from evolutionary families A (Taq, Tfl, HotTub and Tth) and B (Pfu, Pwo, Vent and Deep Vent), were examined for their ability to incorporate 5-position modified 2'-deoxyuridine derivatives that carry a protected thiol group appended via different linkers containing either three or four carbon atoms. This represents the first attempt to incorporate the thiol functionality into DNA via enzymatic synthesis. Each polymerase --substrate combination was evaluated using a hierarchy of increasingly more difficult challenges, starting with incorporation of a single derivative, proceeding to incorporation of two derivatives at adjacent sites and non-adjacent sites, then examining the ability of the polymerase to accept the derivative within the template, and concluding with a challenge involving PCR. The evaluation of thiol-bearing 2'-deoxyuridine derivatives was then extended to consider their chemical stabilities. Stability was found to be less than satisfactory when the thiol functionality has a 'propargylic' relationship to the unsaturation in the linker. The best polymerase --appendage combination used the polymerase from Pyrococcus woesei (Pwo) and the 5'-tBu-SS-CH2-CH2-C=C- linker. This pair supported PCR amplification and therefore should have value in artificial in vitro selection experiments. Indeed, we discovered that Pwo and Pfu preferred the derivative triphosphate over TTP, the natural substrate, in competition studies. These studies confirm an earlier suggestion that membership of an evolutionary family of polymerases is a partial predictor of the ability of the polymerase to accept 5-modified 2'-deoxyuridines. Considerable differences are displayed by different members within a polymerase family, however. This remains curious, as the ability of the polymerase to replicate natural DNA with high fidelity and its propensity to exclude unnatural analogs are presumed to be correlated. Keywords: INTRODUCTION : It has been two decades since the enzymic incorporation of functionalized nucleotides was first suggested to be useful to tag DNA with a functionality that it did not intrinsically carry (,). In the intervening years, enzymic incorporation of modified nucleotides has become the key to automated DNA sequencing . Also, increasing the functionality of nucleic acids has been suggested to be a key to enhancing the power of nucleic acids as catalysts . To implement a functionalized genetic molecule, the DNA alphabet was expanded some time ago from four to 12 letters (,) and modified standard nucleotides have been incorporated into in vitro evolution experiments ( --). The ultimate goal, of course, is a combination of polymerases and unnatural/functionalized DNA-like molecules that support repeated copying and recopying with an efficiency and fidelity sufficient to support Darwinian evolution in vitro, and ultimately in cells. While many polymerases are known to incorporate nucleoside analogs, others have proved to be quite idiosyncratic when challenged with nucleoside analogs , even though all polymerases are believed to be descendants from a single common ancestor . Many polymerases accept pyrimidine species carrying substituents at the 5-position of the nucleobase, at least to some extent. Crystal structures suggest that space is available in the active site for several polymerases to accommodate substituents at this position ( --). However, successful expansion of the genetic alphabet requires incorporation of nucleoside analogs with high efficiency. Neither the crystal structure nor organic chemistry theory has sufficient resolution to predict which polymerase will accept which substituted pyrimidine derivative best. For these reasons, rational empiricism remains an essential part of any effort to generate an artificial genetic system that incorporates 5-position modified pyrimidines the best. Simple tests for nucleotide analog incorporation (such as incorporation of a single nucleotide analog in a primer extension reaction), while useful as a first level, need not, however, be good indicators of the suitability of a derivative for the long-term goal. Likewise, it is clear that simply challenging a polymerase with a particular analog in a PCR reaction is not a strategic way of developing our understanding of polymerases as they interact with unnatural substrates. Rather, a set of graded challenges is needed to evaluate a polymerase --substrate combination for its suitability as part of an expanded genetic information system. This set might begin with simple primer extension experiments challenging the polymerase to incorporate a single nucleotide analog as a triphosphate. In further challenges, however, the polymerase must show its ability to incorporate more than one nucleotide analog, at both adjacent and non-adjacent sites, and to accept the analog in the template. Last, the polymerase might be challenged to amplify, in a PCR reaction, oligonucleotides containing the derivative. Lee et al. recently reported a systematic study of the ability of a single polymerase (Taq) to accept 5-position modified 2'-deoxyuridines carrying four linkers, each with three carbon atoms and bearing an amino group (,). This work prompts us to report our work examining a range of polymerases for their ability to incorporate a series of 2'-deoxyuridine derivatives carrying a thiol functionality appended via different linking groups. Here, we exploit this hierarchy of challenges to evaluate the suitability of a variety of polymerase --analog combinations as part of an expanded alphabet. MATERIALS AND METHODS : Nucleoside synthesis | Full experimental procedures describing the synthesis of the compounds used in this work are available in Supplementary Material. The 2'-deoxyuridine derivatives carrying a 5-position appended functionality were prepared by palladium(0)-catalyzed coupling of the appropriate unit to 5-iodo-2'-deoxyuridine (Scheme ). For example, 3',5'-bis-O-(tert-butyldimethylsilyl)-5-(3-hydroxy-propynyl)-2'-deoxyuridine 3a was prepared in two steps by coupling 5-iodo-2'-deoxyuridine 1 with propargyl alcohol using a Pd(0) catalyst following a procedure adapted from Osborne et al. (; Scheme ). A portion of this material was reduced by catalytic hydrogenation to give 3',5'-bis-O-(tert-butyldimethylsilyl)-5-(3-hydroxypropyl)-2'-deoxyuridine 4a. The free alcohol group was mesylated and then reacted with thiobenzoic acid to give the corresponding thiobenzoate 6a. The silyl groups were removed, the 5'-hydroxyl group was reprotected as its dimethoxytrityl ether (compound 8a), the thiobenzoate was cleaved and the free thiol was converted to the tert-butyl disulfide derivative 9a by reacting with di-t-butyl-1-(t-butylthio)-1,2-hydrazinedicarboxylate. This gave a 2'-deoxyuridine derivative carrying a 5-position thiol attached via a four carbon saturated linker. A corresponding series of transformations, taken from Osborne et al. , generated the disulfide 9b between 5'-O-(4,4'-dimethoxytrityl)-5-(butanethiol) and tert-butyl disulfide. This gave a 2'-deoxyuridine derivative carrying a 5-position thiol attached via a three carbon saturated linker. The corresponding mesylates were also prepared from the propynyl and butynyl intermediates and converted to the corresponding thiols protected as their disulfides with tert-butyl disulfide (10a and 10b). This gave 2'-deoxyuridine derivatives carrying a 5-position thiol attached via three carbon and four carbon acetylenic linkers. The dimethoxytritylated nucleosides in CH2Cl2 were then reacted with N,N-diisopropylethylamine and chloro-N,N-diisopropylamino-beta-cyanoethylphosphine to yield phosphoramidites following standard procedures. Triphosphate synthesis | The 5'-triphosphates of the side chain modified 2'-deoxyuridine derivatives, designated dS1TP, dS2TP, dS3TP and dS4TP , were generated from the corresponding 5'-O-dimethoxytritylated nucleosides 9a, 9b, 10a and 10b by the method of Ludwig and Eckstein . Oligodeoxyribonucleotides | Oligodeoxyribonucleotides containing dS2 were synthesized on a PerSeptive Biosystems Expedite 8900 DNA synthesizer, employing standard beta-cyanoethyl phosphoramidite chemistry. The oligonucleotides were purified by 15% PAGE and characterized by mass spectrometry on a MALDI TOF Voyager mass spectrometer (PerSeptive Biosystems). Standard oligonucleotides were purchased from Integrated DNA Technologies, Inc. as desalted crude products, or synthesized on a PerSeptive Biosystems Expedite 8900 DNA synthesizer and purified by PAGE. Labeling of primer Pr1 at its 5'-terminus with Redivue [gamma-32P]ATP (Amersham) was performed using T4 polynucleotide kinase (New England Biolabs). A list of oligonucleotides used in this work can be found in Table . Polymerases and reaction buffers | Pwo DNA polymerase was from Roche Molecular Systems. The 10x reaction buffer used with Pwo DNA polymerase (PCR buffer with MgSO4 from Roche Molecular Systems) contained 100 mM Tris --HCl (pH 8.85 at 20C), 20 mM MgSO4, 250 mM KCl and 50 mM (NH4)2SO4. Vent DNA polymerase and Deep Vent DNA polymerase were from New England Biolabs. The 10x reaction buffer (ThermoPol reaction buffer from New England Biolabs) contained 200 mM Tris --HCl (pH 8.8 at 25C), 20 mM MgSO4, 100 mM KCl, 100 mM (NH4)2SO4 and 1% Triton X-100. PfuTurboTM DNA polymerase was purchased from Stratagene. It was used with the 10x cloned Pfu reaction buffer from Stratagene [200 mM Tris --HCl (pH 8.8), 20 mM MgSO4, 100 mM KCl, 100 mM (NH4)2SO4, 1% Triton X-100 and 1 mg/ml nuclease-free BSA]. Tfl DNA polymerase and Taq polymerase (in storage buffer A) were from Promega. The 10x reaction buffer used with Tfl DNA polymerase was the Tfl DNA polymerase 10x reaction buffer from Promega [200 mM Tris --acetate (pH 9.0), 20 mM MgCl2, 100 mM (NH4)2SO4, 750 mM KOAc and 0.5% Tween 20]. Taq polymerase was used with the thermophilic DNA polymerase buffer, magnesium-free, from Promega [100 mM Tris --HCl (pH 9.0 at 25C), 20 mM MgCl2, 50 mM KCl and 1% Triton X-100]. The final MgCl2 concentration in the assays with Tfl and Taq polymerases was adjusted to 2 mM. Tth polymerase and HotTubTM polymerase were purchased from Amersham/Pharmacia. The 10x Tth polymerase reaction buffer from Amersham was used in assays with Tth polymerase. The final MgCl2 concentration in the assays was adjusted to 2 mM. In assays with HotTubTM polymerase, the 10x HotTubTM reaction buffer from Amersham [500 mM Tris --HCl (pH 9.0), 20 mM MgCl2 and 200 mM (NH4)2SO4] was used. Primer extension experiments | 5'-32P-labeled primer Pr1 (15 pmol, final assay concentration 750 nM) was annealed to a template sequence (T1 --T7) (18 pmol, final assay concentration 900 nM) in polymerase reaction buffer (1.25x, 16.5 microl) by heating the mixture for 2 min to 95C and subsequently allowing the solution to cool over 2 h to room temperature. The triphosphates (18.8 nmol, final concentration 188 microM each) were added at room temperature, followed by the polymerase (1 U), to a final reaction volume of 20 microl. The reaction mixture was immediately incubated at 72C. After 10 min, the reaction was stopped by addition of EDTA (1.25 equiv. with respect to [Mg2+]). The reaction products were separated by denaturing PAGE (7 M urea, 55 --60C) and visualized and quantified with a MolecularImager (Bio-Rad, Hercules, CA). PCR amplifications | PCR experiments were performed on a 50 microl scale in 0.2 ml thin-walled PCR tubes. Reaction mixtures contained primers Pr2 and Pr3 (50 pmol each, 1 microM), template T8 (2.5 pmol, 50 nM), triphosphates (20 nmol each, 200 microM), the appropriate polymerase reaction buffer (at 1x concentration, see above) and a polymerase (1 U). In 'positive' control reactions, the triphosphates were the standard triphosphates TTP, dATP, dCTP and dGTP (Promega). In the actual test reactions, TTP was substituted with one of the modified triphosphates dS2TP, dS3TP or dS4TP. In 'negative' control reactions, water was added in lieu of TTP. The reaction mixtures were cycled through 15 (Family A polymerases) or 25 (Family B polymerases) cycles of amplification in a Stratagene Robocycler (1 min at 94C, 2 min at 55C and 2 min at 72C). The mixtures were then stored on ice and the reactions were quenched by addition of EDTA (2 equiv. with respect to [Mg2+]). Analysis was performed by 2% agarose gel electrophoresis containing ethidium bromide. The products were visualized under UV light (254 nm). Gel images were recorded and analyzed with GelDoc (Bio-Rad). Scheme 1. Scheme 1. Figure 1 | Structures of the substituted 2'-deoxyuridine derivatives dS1, dS2, dS3 and dS4 modified with a 5-position side chain carrying a thiol group protected as the disulfide with tert-butyl thiol. Structures of the substituted 2'-deoxyuridine derivatives dS1, dS2, dS3 and dS4 modified with a 5-position side chain carrying a thiol group protected as the disulfide with tert-butyl thiol. Table 1 | Primer and templates employed for polymerase screening by primer extension and PCR RESULTS : Four 2'-deoxyuridine derivatives bearing C5-mixed tert-butyl disulfide moieties (dS1, dS2, dS3 and dS4) were examined to learn whether they might be suitable as components of an artificial genetic system that introduces thiol functionality into oligonucleotides . Both dS3 and dS4 are known previously in their 5'-dimethoxytritylated phosphoramidite forms , but not as triphosphates. As such, dS3 and dS4 have been incorporated into oligonucleotides by automated chemical synthesis, yielding thiol functionalized oligonucleotides through post-synthetic reduction of the disulfide moieties. They all share a tert-butyl blocked disulfide attached to the nucleoside through the C5-position of the heterocyclic base. They differed in the length and degree of saturation of the chain appending them to the heterocycle. Two compounds are new to this work. One of them, dS2, carries the protected thiol on a butynyl linker. dS2 differs from dS4 because of its slimmer and more rigid linker. The propynyl tert-butyl disulfide substituted 2'-deoxyuridine, dS1, proved to be insufficiently stable during the standard work-up of the synthetic oligonucleotides to be carried further. During synthesis of oligonucleotides containing dS1, complex HPLC results were observed on crude materials after deprotection. This complexity arose even though trityl release indicated highly successful coupling. Simple TLC spot tests of the nucleoside in ammonia suggested that even after only 1 h at 55C, dS1 nucleoside decomposed. UV spectra indicated that the propynyl linkage had disappeared. In contrast, both dS3 and dS4, both lacking the 'propargylic' thiol, were found to be stable under these conditions. We had noticed previously a lower level of instability with the analog carrying a propargylic amino group (T.R.Battersby, unpublished results). We suspect that this instability increases as the leaving group ability of the functionality increases and recommend that this consideration guide the design of other functionalized 5-position modified 2'-deoxyuridine derivatives. Incorporation of dSxTP opposite dA in a template | The simplest challenge for a polymerase is to place a single nucleotide site specifically into an oligonucleotide opposite a single nucleotide in a template. This is most easily implemented in a 'primer extension' (or 'run-off') experiment. A primer is annealed to a template oligonucleotide containing a single nucleotide that presents the challenge, and then presented to a polymerase. Analysis of the products, generally by gel electrophoresis, determines how the polymerase met the challenge. To study the incorporation of the uridine triphosphate analogs dSxTP, templates T1, T2 and T3 were employed . These oligonucleotides are composed of the four standard 2'-deoxynucleosides T, dA, dC and dG. dS2-containing templates T4 and T6 were utilized to analyze the behavior of polymerases when encountering templates containing 2'-deoxyuridine analogs carrying C5-position butynyl substituents. All five templates anneal to a common primer (Pr1). In the simplest primer extension experiments studying incorporation of functionalized triphosphates dSxTP, all standard nucleoside templates (T1, T2 and T3; Table ) were used. 5'-32P-radiolabeled primer Pr1 was annealed to an excess of template oligonucleotide. The primer --template complex was then incubated with triphosphates dSxTP, dATP, dCTP and dGTP, and a polymerase. An analogous sample in which TTP took the place of the substituted triphosphate dSxTP was prepared simultaneously. This 'positive' control reaction did not involve any modified nucleosides and therefore served as a reference for the performance of the polymerase. In a 'negative' control, water replaced the dSxTP or TTP solution. After a defined incubation time, the reaction was stopped with EDTA and the reaction mixture was analyzed by denaturing PAGE. The 'level one' challenge asks if a polymerase can incorporate a single non-standard nucleotide in a product DNA molecule opposite a single position in a template. To implement this challenge, T1 and Pr1 were used as a template --primer combination. Full-length product is formed by adding 10 nt to Pr1, with a single dSxTP in the product introduced opposite a single A in the template. The reaction products were detected by PAGE. Figure shows a representative set of results in a 'two by two' polymerase --substrate matrix. The two polymerases were Pwo and Vent and the two substrates were the propynyl-linked dS1TP and butynyl-linked dS2TP analogs. Three features of the results are noteworthy. First, the major band observed with both polymerases and both dSxTP substrates corresponded to full-length product. This indicated that both polymerases accepted both substrates under these conditions. Second, a small amount of an 'n -- 1' product was formed in addition to full-length product. As this was seen in the positive control, the n -- 1 product cannot indicate a deficiency in the interaction between the polymerase and the non-dSxTP. Most likely, the n -- 1 product may have arisen because of a small amount of n -- 1 failure sequence in primer Pr1 (Fig. , lane 1). Last, the full-length product incorporating dSxTP migrated on the gel slightly more slowly than the analogous full-length product incorporating T. This is consistent, of course, with the modestly larger 'size' of an oligonucleotide containing dSxTP. In all subsequent experiments, this mobility shift was taken to indicate the presence of dS1 and dS2 in the product strand. Additional confirmation that dS1TP and dS2TP is indeed incorporated, and that the full-length product is not a result of misincorporation, was provided by the negative control. In the absence of dS1TP, dS2TP and TTP, chain elongation stopped (<5% continued elongation) at the position of dA in the template (Fig. , lanes 4 and 8). This indicated that the polymerases could not form full-length product without 2'-deoxyuridine being present as a triphosphate. Together with the mobility shift, this makes compelling the case that full-length product, when it is observed, arises through incorporation of one of these nucleotides opposite dA in the template. As chemical instability suggested that dS1 would never make a suitable analog, it was abandoned at this point. All eight polymerases (Tfl, Tth, HotTub and Taq from Family A and Pwo, Pfu, Vent and Deep Vent from Family B) were then challenged with dS2, dS3 and dS4. To establish a metric for the evaluation of a polymerase --analog pair, the intensity of the band corresponding to full-length product was quantitated using the Molecular ImagerTM software. The quantitative intensity was then compared to the intensity of the analogous band formed when TTP was replaced by dSxTP. When an 'n + 1' band was observed, as is frequently the case with Family A polymerases , this band was included as full-length product. The amount of full-length product formed by the positive control reaction was set at 100%. Experiments then showed that dS2TP, dS3TP and dS4TP were all accepted in lieu of TTP by all eight polymerases tested . Although most of the dS3TP/polymerase and dS4TP/polymerase combinations produced full-length product in amounts >75% of the positive control, larger amounts of abortion products were produced with dS3TP and dS4TP than with dS2TP. Surprisingly, the major abortion products formed by Family B polymerases were only 3 nt shorter than the full-length product and did not correspond to abortion at the position of the critical nucleotide dA, which would form products 5 nt shorter than the full-length product. It can be deduced from the shift of the bands of the abortion products that the functionalized nucleosides were probably incorporated before pausing occurred. This set the stage for a broad survey of the suitability of polymerase --analog pairs that meet the level one challenge. It has been observed (our unpublished results) that the evolutionary family to which a polymerase belongs might be a predictor of its performance with a set of analogs. Therefore, we present data here according to their family (Figs and ). It is worth noting that at this, the lowest level of challenge, all polymerases work with all substrates. This is the level of challenge that is met in a majority of experiments that examine the incorporation of non-standard nucleotide analogs into DNA via template-directed polymerase-catalyzed primer extension. It is clear that performance at this level is, at best, a preliminary indicator of the suitability of the pair to support an artificial genetic system. We illustrated this by attempting to amplify an oligonucleotide using PCR with dS3TP or dS4TP replacing TTP and Vent as the polymerase. PCR produced little amplification under standard conditions (data not shown). This result is hardly surprising. Copying the copies requires a polymerase that can, at the very least, accept a non-standard component in a template. Further, more than one non-standard component will occur in a real, artificial genetic system, possibly adjacent in the sequence. Polymerases must be able to copy these. In general, most polymerases will fail to PCR amplify oligonucleotides containing most nucleotide analogs. Unfortunately, a failed PCR does not provide sufficient information to say what went wrong. Therefore, intermediate level challenges are needed to evaluate and cull polymerase -- analog systems for their suitability to support artificial genetics. At the very least, these will prevent time consuming experiments on polymerase --analog pairs that will not work. At best, these will provide diagnostic information to direct the experimentalist towards those systems that are most likely to work. The next level of challenges: adjacent components and multiple components | Two of these situations were set up for the dSx system. The first challenged the polymerase to incorporate two adjacent analog components into a DNA product. This was achieved using template T2, which has two adjacent A residues. The second challenged the polymerase to incorporate multiple analog components at 'random' positions, both adjacent and non-adjacent, in a DNA product. This was achieved using template T3, which has eight A residues, including three in a row. The data are shown in Figure (for consecutive incorporation) and Figure (for multiple incorporation). Several features of the results are striking. First, the ability to incorporate a single dSx triphosphate was not predictive of the ability of a polymerase to incorporate two dS2 nucleotides in adjacent positions. For example, all Family A polymerases incorporated a single dS3 into a product (Fig. C), but failed to incorporate two consecutive dS3 residues . Second, the data provided clear evidence that membership in a particular evolutionary family of polymerases was a predictor of the performance of a polymerase given a particular challenge. The only polymerase --analog systems that performed with results comparable to those obtained with TTP were samples containing dS2TP and Family B polymerases. A significant amount of a second band, probably the n -- 1 product, was formed in the dS2TP reactions with template T2, but not in the corresponding reactions with TTP. Oddly, this phenomenon did not occur with template T3, which included the sequence of template T2. Family A polymerases generated significant amounts of full-length product containing dS2, but also clearly paused; it seemed as if the incorporation stopped after the dS2TPs had been incorporated (data not shown). It is interesting to note in this context that oligonucleotide T4, which contained two dAs separated by 4 nt, was utilized as a template without pausing by all eight polymerases in the presence of dS2TP as T derivative (see below). Template T3 was designed to identify polymerases that are capable of processing the functionalized triphosphates dS2TP, dS3TP and dS4TP with high efficiency. A total of eight modified triphosphates must be incorporated by a polymerase to form a full-length complement to T3. The decameric region adjacent to the primer-binding region of T3 was identical to T2, which required incorporation of two consecutive modified triphosphates. In addition, T3 posed the challenge of incorporating three successive modified triphosphates, as well as several individual modified triphosphates, in various sequence contexts. Even with the demanding template T3, Pwo, Pfu and Deep Vent polymerases (Family B) generated substantial amounts of full-length product containing dS2, resulting in >75% of the product formed with TTP (Fig. B). The extent of the shift in the full-length product was remarkable. Of all tested polymerases, Pwo polymerase generated the cleanest product and the fewest failure products. Family A polymerases, in contrast, failed to yield any detectable full-length product in the presence of dS2TP as T analog (Fig. B). With the alkanyl substituted nucleotides dS3TP and dS4TP, Family B polymerases were inefficient, while Family A polymerases completely failed to form fully elongated product with template T2 . Neither Family B nor Family A polymerases formed any full-length product with dS3TP and dS4TP and template T3 under the assay conditions (data not shown). Higher level challenges: non-standard components in the template | A second requirement for a non-standard analog to participate in an artificial genetics system is that it be accepted by a polymerase as a component of a template. Only the substituted thymidine derivative dS2TP survived intermediate challenges past simple primer extension reactions. The alkanyl substituted nucleotides dS3TP and dS4TP were not processed efficiently enough by the polymerases tested here. dS1 was eliminated due to its apparent chemical instability. To implement this challenge for dS2, template T4 was prepared containing dS2, and polymerases were challenged to incorporate dA opposite it . Seven of the eight polymerases (the Tfl result is curious given the results in Fig. ) readily incorporated dATP opposite dS2 in template T4 to form full-length product. The longer n + 1 bands are characteristically seen with Family A polymerases . When TTP was employed as complement for dA in the template, the standard oligonucleotide product band migrated with the product band generated by the positive control sample. When dS2TP replaced TTP, full-length product containing two dS2 nucleosides was generated and migrated slower than the product containing T. In the absence of dATP, some full-length product (up to 10% of the amount formed in the presence of dATP) was observed with Family A polymerases. This might be due to contamination with residual [gamma-32P]ATP from the primer labeling procedure (requiring the DNA polymerase to incorporate a ribonucleoside triphosphate) or due to misincorporation. Misincorporation of dATP, dCTP or dGTP opposite dA was found to take place when the standard template T1 was employed (Figs and ). Both observations are consistent with the known low fidelity of the relevant Family A polymerases (similar1 error/1.3 x 105 nt incorporated) . In the negative control samples with Family B polymerases, no full-length product was formed. With Vent (and slightly with Deep Vent), primer elongation continued beyond the first dS2 nucleoside in the template by several nucleotides in the negative control . This suggests that having dS2 in the template may lower fidelity. This hypothesis needs to be examined in greater detail. Further challenges: adjacent non-standard components in the template | The next challenge involved templates containing two adjacent non-standard components. For this purpose, template T6 was used, carrying two adjacent dS2 nucleosides, flanked by dA on each side. Template T7, a standard oligonucleotide of the same length and very similar sequence, was used for the positive control assays. All Family A polymerases formed full-length product in very high yields (Fig. A). However, with all Family B polymerases, significant amounts of shorter products of various lengths were observed (Fig. B). Pwo, Pfu and Deep Vent even yielded <50% of the full-length product formed in the positive control sample. Nevertheless, very little misincorporation was seen in the negative control. When dS2TP was offered with template T6 instead of TTP, dS2TPs must be incorporated nearly across from the dS2 nucleosides in the template, resulting in a duplex with four of the nucleotide analogs in close proximity. Only Tfl and Tth polymerases, both of Family A, were found to be able to generate this duplex with ease (Fig. A). Pwo, Pfu and Deep Vent polymerases paused significantly after encountering the two adjacent dS2 nucleosides in the template (Fig. B). Taq and Deep Vent polymerases generated multiple products shorter than the full-length product, without a single predominant failure product. Lengthening the time of incubation might resolve problems of pausing and slow incorporation, as long as there is little infidelity. Low infidelity in the negative control experiments was characteristic of Family B polymerases throughout these studies. Family A polymerases, in contrast, misincorporated other nucleotides when the complementary triphosphate was not present, especially with adjacent dSx in the template (see Fig. ). We propose that the ratio of incorporation to misincorporation in a 'minus' control may prove to be a more relevant predictor of the suitability of a polymerase -- analog pair in an artificial genetics system. This hypothesis will need further examination. Evaluation using competition experiments | To directly compare the efficiency of polymerase incorporation of dS2TP versus TTP, primer extension experiments were implemented in which dS2TP competes with TTP. Oligonucleotide T1 was chosen as template, as it requires the incorporation of only a single dA complement and therefore enables a simple analysis of the experimental data. The assays were carried out under the conditions used in the polymerase screening experiments. The complex from primer Pr1 (5'-fluorescein labeled) and template T1 was incubated for 10 min at 72C with either Pwo or Vent polymerase, the standard triphosphates, and [alpha-32P]TTP (TTP*). The amount of radioactive product formed in the control reaction was set to 100%. In individual assays, 25, 50 and 75% of the TTP* solution in the control reaction was replaced by an equimolar dS2TP solution, and the decrease in the amount of radioactivity incorporated into full-length product was used as a measure of the ability of dS2TP to compete with TTP* in the elongation reaction. The product was visualized using the 5'-fluorescein label of the primer, ensuring that complete primer elongation had taken place in all samples. If the polymerase made no discrimination between TTP* and dS2TP, T* and dS2 would appear in the full- length product in a ratio identical to the ratio of their triphosphates. Thus, if TTP* and dS2TP were offered in a 1:1 ratio and the polymerase accepted either substrate equally well, the amount of radioactive product would be 50% of the amount of radioactive product formed in the presence of TTP* alone. The amounts of radioactivity actually found in the product bands, relative to the samples without addition of dS2TP, are summarized in Table . Remarkably, for the two polymerases examined, the amounts of radioactivity found in the product bands indicate that both Pwo and Vent polymerase incorporate dS2TP preferentially to TTP opposite dA in template T1. Pwo polymerase discriminates even more strongly in favor of dS2TP than Vent. We asked whether this result might be explained by a miscalculation of the concentration of the dS2TP solution, which was estimated by UV absorbance exploiting the Lambert --Beer Law. The extinction coefficient for dS2TP was taken as that of 5-(1-propynyl)-2'-deoxyuridine (Glen Research; molar epsilon260 = 3.2 x 103). Comparison of extinction coefficients of various alkynyl substituted deoxyuridine derivatives does not show a strong influence of the omega functionality of the alkynyl linker on the extinction coefficient . For example, epsilon260 is reported as 3.5 x 103 M --1 cm --1 for linker --C=C-R, R = CH2CH2OH, while epsilon260 is reported as 3.0 x 103 M --1 cm --1 for linker --C=C-R, R = CH2CH3. This implies that the discrimination evident with Pwo was significantly greater than any plausible misjudgement of the concentration of dS2TP. The discrimination observed with Vent, however, was within these limits. Tables and summarize the results of the primer extension experiments examining the enzymatic incorporation of dS2TP, dS3TP and dS4TP opposite dA nucleosides. In general, the collection of data focuses on Family B, and Pwo polymerase in particular, as a suitable pair for dS2TP. These results provide the rationale for this pair as a candidate to support an artificial genetic system. It should be noted that a further understanding of the competition between these analogs could be gained by comparing the time courses of the primer extension reaction. PCR amplification | Amplifications by PCR are much more exacting than a simple primer extension reaction, with very different demands on polymerases. In PCR, longer templates must be replicated in shorter times than in typical primer extension experiments. PCR amplification also puts higher demands on the fidelity of polymerases, because incorporation mistakes are amplified. Pwo, Pfu, Vent, Deep Vent, Tfl and Taq polymerases were tested for their ability to PCR amplify the 94mer template sequence T8, utilizing dS2TP in place of TTP. The 94mer T8 was used as a template in comparative studies of the polymerases. T8 contained only the four standard nucleotides. dS2 was introduced during the first cycle of PCR. In the following cycles, dS2 was present as both triphosphate and in the template. T8 was designed to contain only very few dS2 nucleotides in the primer-binding regions generated after the first PCR cycle, in order to avoid potential annealing problems. The region between primers and primer-binding regions had a very high dS2 content (30% after the first cycle), to make the test rigorous. A set of three samples was prepared to test each polymerase. All three samples contained template T8 and matching primers (Pr2 and Pr3), the polymerase being tested and the standard triphosphates dATP, dCTP and dGTP. TTP was added to the 'positive' control sample (T sample). The TTP solution was substituted by dS2TP solution in a second sample (S sample). Finally, water replaced the T derivative solution in the 'negative' control sample. In amplifications with dS2TP and Pwo, similar90% amplification product containing dS2 (T8S) was generated compared to the amount of amplification product containing T (T8T) with TTP . The amount of T8S generated by Vent polymerase was similar55% of the amount of T8T generated under the same conditions. For Pfu and Tfl polymerase the ratio T8S:T8T was even less, only similar30 --35%. Taq did not form any detectable T8S product at all. With Pwo and Pfu polymerase PCR was performed over 25 cycles. Both the T and S products were obtained as sharp bands. With Vent, Tfl and Taq amplification was stopped after 15 cycles. When more cycles were performed, non-specific amplification of the template occurred in the T samples, leading to longer products which were visible as a smear on the gel. Figure 2 | Primer extension reactions with 5'-32P-labeled primer Pr1 (p) and template T1 (t), showing the synthesis of full-length product by these polymerases in the presence of dS1TP and dS2TP, but not in their absence. Primer extension reactions with 5'-32P-labeled primer Pr1 (p) and template T1 (t), showing the synthesis of full-length product by these polymerases in the presence of dS1TP and dS2TP, but not in their absence. Polymerases and triphosphates used in each reaction are noted below the lanes. Note the presence of n -- 1 template (lane marked t). Denaturing 15% PAGE (urea, 60C). Figure 3 | Primer extension reactions with 5'-32P-labeled primer Pr1 (p) and template T1 (t), showing the synthesis of full-length product by these polymerases in the presence of TTP and dS2TP, dS3TP and dS4TP, but not in their absence. Primer extension reactions with 5'-32P-labeled primer Pr1 (p) and template T1 (t), showing the synthesis of full-length product by these polymerases in the presence of TTP and dS2TP, dS3TP and dS4TP, but not in their absence. Polymerases and triphosphates used in each reaction are noted below the lanes. Note the presence of n -- 1 template (lane marked t). Denaturing 17% PAGE (urea, 60C). Figure 4 | Primer extension reactions with 5'-32P-labeled primer Pr1 (p) and template T2 (t). Primer extension reactions with 5'-32P-labeled primer Pr1 (p) and template T2 (t). TTP, dS3TP and dS4TP were employed to replace TTP. Polymerases and triphosphates used in each reaction are noted. Denaturing 17% PAGE (urea, 60C). Figure 5 | Primer extension reactions with 5'-32P-labeled primer Pr1 (p) and template T3 (t). Primer extension reactions with 5'-32P-labeled primer Pr1 (p) and template T3 (t). TTP and dS2TP were employed to replace TTP. Polymerases and triphosphates used in each reaction are noted. Denaturing 15% PAGE (urea, 60C). Figure 6 | Primer extension reactions with 5'-32P-labeled primer Pr1 (p) and templates T4 (t) and T5 (marked with *). Primer extension reactions with 5'-32P-labeled primer Pr1 (p) and templates T4 (t) and T5 (marked with *). Polymerases and triphosphates used in each reaction are as noted. The result in lane 3 may be an artifact (see Fig. ). Denaturing 15% PAGE (urea, 60C). Figure 7 | Primer extension reactions with 5'-32P-labeled primer Pr1 (p) and templates T6 (t) and T7 (marked with *). Primer extension reactions with 5'-32P-labeled primer Pr1 (p) and templates T6 (t) and T7 (marked with *). Polymerases and triphosphates used in each reaction are as noted. Denaturing 15% PAGE. Figure 8 | Letters beneath the lanes indicate the triphosphates that were present in the PCR incubation mixture. Letters beneath the lanes indicate the triphosphates that were present in the PCR incubation mixture. PCR products were resolved by 2% agarose gel electrophoresis, stained with ethidium bromide and visualized under UV light (254 nm). The marker, labeled M, is a 25 nt ladder from Life Sciences; the most intense band is at 125 bases. PCR cycling: 94C, 1 min; 56C, 2 min; 72C, 2 min. All amplifications were done for 25 cycles, except for Taq, where 15 cycles were used to avoid 'smears' (presumably a broad product distribution), presumably arising from lower proofreading. The intensity of fluorescence was integrated by the Molecular Imager software. Table 2 | Results of primer extension experiments with direct competition between labeled TTP and unlabeled dS2TP for incorporation opposite dA by Pwo and Vent polymerase Table 3 | Summary of the influence of linker geometry and length on the incorporation of the functionalized nucleoside triphosphates dS2TP, dS3TP and dS4TP into DNA Table 4 | Acceptance of dS2-containing templates by thermostable polymerases DISCUSSION : In the past decade, many laboratories have reported that polymerases incorporate nucleotide analogs, including those that carry appendages that protrude into the major groove. These have proven to be useful tools for the biotinylation of DNA, for DNA sequencing and for the addition of a variety of reporter groups and tags to DNA. These results might suggest that incorporation of major groove functionalized nucleotides into DNA via template-directed polymerization must be easy. In fact, when this is done routinely within the context of a PCR experiment, the outcome is frequently unsatisfactory. This paper makes clear that this unsatisfactory outcome is not necessarily unexpected. In virtually all of these cases, polymerase incorporation was considered to be successful after only the simplest challenge is presented to the polymerase. This does not, of course, detract from the utility of the tools. In DNA sequencing, of course, inefficient incorporation of a chain terminator is desired. Inefficient incorporation of a label can be tolerated in a range of biophysics experiments, where (for example) a fluorescence signal is sought and the signal from only a few labeled molecules is needed. Only when the goal is an artificial genetic system must we be concerned about the highest level of incorporation. There is nothing particularly innovative about the strategy presented here to assess the specifications of a polymerase --analog pair. What is clear, however, is that a strategy of this type is important in developing artificial genetic systems based on them. Many discoveries were made by this strategy. First, we were surprised to see the chemical instability of dS1. This discovery should guide further development of artificial genetic systems. Second, it was a surprise that two polymerases incorporated dS2TP, in some contexts, in preference to TTP. In the Pwo/Pfu sequence(s), however, it is interesting to note that an Ala residue replaces Ser411, which is highly conserved throughout most other Family B polymerases. Although no crystal structure is available for the Pwo/Pfu polymerase, crystal structures of other Family B polymerases allow homology modeling, which in turn suggests that the Ala411 side chain might lie in the position where a major groove substituent might bind. The Ala side chain is, of course, more hydrophobic than the Ser side chain. This difference might therefore explain why the Pwo polymerase so effectively accepts dS2 in both the template and as a triphosphate. Realistically, however, chemical theory is inadequate to enable rational design at this level of resolution, implying that explanations at this level of resolution must be viewed as being conjectural. It was also surprising to see that single incorporation experiments were not necessarily good predictors of the outcome of multiple incorporation experiments. Much of the literature needs to be re-evaluated in the light of this result. Further, we found further evidence that evolutionary classification is a valuable predictor for how a polymerase is likely to handle a non-standard substrate. Polymerases are classified into two evolutionary families according to their amino acid sequence similarity to E.coli DNA polymerase 1 . X-ray crystallography has shown that Family A and Family B polymerases possess analogous tertiary structures ( --). They are, however, quite divergent in their amino acid sequences and it remains an open question as to whether the two folds arose by convergent or divergent evolution. This statement may be general to other classes of nucleotide analogs. For example, it has been suggested that the ability of a polymerase to incorporate the nucleoside analogs disoC and disoG also correlates with membership of a polymerase family . In this context, it is interesting to note that the Family B polymerases examined here all have a strong 3'->5' exonuclease (or 'proofreading') activity, providing replication with higher accuracy than the relatively error-prone Family A representatives, which all lack a 3'->5' exonuclease domain. The inherent proofreading activity of Pwo polymerase results in an increased fidelity of DNA replication that is (remarkably) >10-fold greater than that displayed by Taq polymerase (Roche Molecular Systems, Pwo polymerase product information). It is interesting to ask how this well known property of polymerases might correlate with their ability to accept with fidelity non-standard nucleotide analogs. Finally, the Pwo and Pfu polymerases present a paradox. These enzymes were purchased from Roche Molecular Systems and Stratagene, respectively. The literature records their having the same sequences, despite their different names. Nevertheless, they behave quite differently in their interactions with non-standard nucleotide analogs, and these differences are reproducible. Inquiries to the commercial supply houses failed to obtain any information that resolved this paradox. We cannot say therefore whether the literature has confused the sequences of the genes, or whether the proteins are expressed or processed differentially from the same gene, or whether trace molecule 'effectors' in the two commercial preparations account for their different behaviors. SUPPLEMENTARY MATERIAL : Supplementary Material is available at NAR Online. Figure 1 | Backmatter: PMID- 12202767 TI - Characterisation of site-biased DNA methyltransferases: specificity, affinity and subsite relationships AB - DNA methylation is now seen as a primary signal in the cell for mediating transcriptional repression through chromatin formation. The construction and evaluation of enzymes capable of influencing this process in vivo is therefore of significant interest. We have fused the C5-cytosine DNA methyltransferases, M.HhaI and M.HpaII, which both methylate 4 bp sequences containing a CpG dinucleotide, to a three zinc finger protein recognising a 9 bp DNA sequence. DNA methylation analyses demonstrate specific DNA methylation by both enzymes at target sites comprising adjacent methyltransferase and zinc finger subsites, targeted M.HpaII being the most specific. Binding analysis of the targeted M.HpaII enzyme reveals an 8-fold preference for binding to its target site, compared to binding to a zinc finger site alone, and an 18-fold preference over binding to a methyltransferase site alone, thereby demonstrating enhanced binding by the fusion protein, compared to its component proteins. Both DNA binding and methylation are specific for the target site up to separations of similar40 bp between the zinc finger and methyltransferase subsites. Ex vivo plasmid methylation experiments are also described that demonstrate targeted methylation. These targeted enzymes, however, are shown to be not fully mono-functional, retaining a significant non-targeted activity most evident at elevated protein concentrations. Keywords: INTRODUCTION : Members of the prokaryotic cytosine-5 DNA methyltransferase (C5-Mtase) family of enzymes share a high degree of amino acid sequence homology, most pronounced in 10 highly conserved regions (,). This homology to some extent reflects the shared catalytic function of all these enzymes, that of transferring a methyl group from the cofactor, S-adenosyl-l-methionine (AdoMet), to the 5 position of the target cytosine base in DNA. M.HhaI and M.HpaII are type II C5-Mtases from Haemophilus haemolyticus and Haemophilus parainfluenzae, respectively, and recognise the specific DNA sequences 5'-GCGC-3' and 5'-CCGG-3', respectively, methylating the underlined cytosine in each instance. Crystal structures for one of these enzymes, M.HhaI, as well as for M.HaeIII, complexed with DNA and cofactor have been described (,). Whilst demonstrating subtle differences in DNA --protein contact signatures between the two proteins, both these studies revealed that in each case the target cytosine is flipped out of the DNA helix into the nucleotide-binding pocket of the enzyme, where the methyl transfer reaction can proceed. The functional characteristics of these enzymes, together with a wealth of information relating both amino acid sequence and structure to function, suggests to us that C5-Mtases may have potential as tools both in general research and in the treatment of disease, if their catalytic activity could be delivered to predetermined regions of the genome. The construction and evaluation of a three zinc finger fusion with a prokaryotic methylase was initially reported by Xu and Bestor . While they were able to demonstrate a degree of targeted binding and methylation, they also suggested that the protein was not fully mono-functional, each component protein exhibiting a significant independent activity. However, although members of the C5-Mtase family are highly homologous in terms of primary amino acid sequence, the functional characteristics of these proteins can vary significantly. For example, M.SssI, the methyltransferase used in this early work, contains both a processive methyltransferase activity and a topoisomerase activity , both of which would be expected to modulate gene targeting by a zinc finger protein, to varying extents. Therefore, in an effort to evaluate the properties of other C5-Mtases in a gene-targeted scenario, we have constructed fusions between a three zinc finger DNA-binding protein and the C5-Mtases M.HhaI and M.HpaII. A primary application that we envisage for targeted C5-Mtases is as a tool for the study of the spread of methylation patterns within eukaryotes. While there is evidence that DNA methylation does spread from an initial focus ( --), no clear trends or rules have been established. Recent observations, such as the existence of a novel family of mammalian C5-Mtases and the role of histone deacetylases in mediating methylation-induced packaging of DNA (see below), suggest that a series of complex factors might exist that contribute to DNA methylation spreading in mammalian cells. Previous experimental procedures for studying methylation spread in vivo generally involved in vitro methylation and subsequent stable transfection of plasmid vectors (see for example ), with the integration site and flanking sequences varying between each experiment. This makes interpretation of data extremely difficult, due to large variations in flanking sequence effects. Furthermore, it has recently been demonstrated that the integration event itself can lead to marked alterations in the DNA methylation patterns of cellular genes and DNA segments . Development of the enzymes described in this study will allow an alternative approach to the study of DNA methylation spread in eukaryotes that closely approaches the in vivo state. These enzymes, when expressed sequentially in a clonal cell line whose genome harbours a targetable DNA insert containing multiple C5-Mtase recognition sites, will enable a range of methylation densities to be generated at a defined locus, with a knowledge that both flanking sequences and initial methylation status are identical in each instance. A second potential application of targeted C5-Mtase enzymes stems from observations that mutation of the invariant catalytic cysteine of C5-Mtases to a glycine residue can generate an enzyme that binds with extremely high affinity to its DNA recognition site (see for example ,). These proteins are cytotoxic when expressed in Escherichia coli and the ability to direct such toxic enzymes to specific and unique gene sequences within the mammalian cell may ultimately provide both greater insight into cellular repair processes and lead to a potential strategy for targeted cell death. Lastly, the correlation between DNA methylation of eukaryotic gene promoters and transcriptional repression is widely accepted (for a review see ,) and some of the mechanisms by which this relationship is implemented have been elucidated. These include reduced or abolished binding of transcription factors when their DNA binding site is methylated, as well as proteins that specifically bind to methylated DNA, thus occluding transcription factor binding directly. The observation that methylcytosine-binding proteins accelerate compaction of methylated DNA into a transcriptionally inactive chromatin form prompted studies that have revealed extensive interactions between methylcytosine-binding proteins and histone deacetylase proteins, mediated by a host of adaptor molecules ( --). With these mechanisms in place within the cell, targeted promoter DNA methylation, accomplished using derivatives of the enzymes described in this study, may represent a novel gene therapy approach for the inheritable silencing of genes associated with disease. This would be of particular relevance in the case of viral genomes, which are known to be sensitive to promoter methylation. We now report the construction and evaluation of first generation targeted HhaI and HpaII C5-Mtases. These enzymes are assessed for their ability to specifically methylate targeted versus non-targeted DNA sequences. The most promising of these current enzymes, targeted HpaII, has been studied in more detail, both in terms of binding specificity and affinity for different DNA substrates as well as methyltransferase specificity. The results of experiments designed to assess the cooperative interaction between the zinc finger and methyltransferase components of the fusion enzyme are also reported. The results of ex vivo studies are presented that point the way towards successful function of these enzymes under in vivo conditions in the mammalian cell. MATERIALS AND METHODS : General reagents and chemicals | General reagents were supplied by Sigma unless otherwise stated. Restriction enzymes, Vent DNA polymerase, T4 DNA ligase and T4 polynucleotide kinase were purchased from New England Biolabs. Ultra pure deoxyribonucleotide solutions and poly(dI)poly(dC) were purchased from Pharmacia Biotech. [gamma-32P]ATP (>111 TBq/mmol) was purchased from ICN and [3H-methyl]AdoMet (>1.85 TBq/mmol) from Amersham International. All oligodeoxynucleotides used in this study were synthesised by MWG BioTech (Germany), by the standard CE-phosphoramidite method. Bacterial strains and plasmids | Escherichia coli ER1647 [F --fhuA2 Delta(lacZ)r1 supE44 trp31 mcrA1272::Tn10(Tetr) his-1 rpsL104(Strr) xyl-7 mtl-2 metB1 Delta(mcrC-mrr)102::Tn10(Tetr) recD1014] was used throughout this study and was obtained from New England Biolabs. Bacteria were grown in liquid culture in Luria --Bertani medium (1% w/v Tryptone, 1% w/v NaCl and 0.5% w/v yeast extract) and on LA medium (1% w/v Tryptone, 1% w/v NaCl, 0.5% w/v yeast extract and 1.5% w/v agar) supplemented, if appropriate, with antibiotics as described before . The gene encoding the three zinc finger protein recognising the 9 bp sequence 5'-GCAGAAGCC-3' was a generous gift from Dr Yen Choo (LMB, University of Cambridge, UK). pUCHhaI and pUCHpaII are pUC18-based plasmids harbouring the entire M.HhaI and M.HpaII ORFs and were a kind gift from Dr Geoff Wilson (New England Biolabs). pGEX5X-3, pET22b and pLsyE (pACYC derivative) vectors were purchased from Pharmacia Biotech and Novagen. Construction of targeted and non-targeted C5-Mtases | The gene encoding M.HhaI (Genbank accession no. J02677, bases 437 --1420) was amplified by PCR using pUCHhaI (10 ng) as template and primers complementary to the 5' end (5'-GGCGCGGATCCTCGACTAGTGAATTCATGATTG AAATAAAAGATAAACAGCTCACAGGATTACGCTTT-3') and 3' end (5'-GCGCGGTACCTTACTCGAG ATATGGTTTGAAATTTAATGA-3') of the M.HhaI gene. The M.HhaI PCR product was restricted with BamHI and XhoI (underlined in the above primer sequences) and ligated into these same sites in pET22b to produce pHhaI. The PCR product was also ligated into the same sites in pGEX5X-3 in order to produce a non-targeted, GST --HhaI construct (pGHhaI). Amplification of the three zinc finger peptide described by Choo et al. was performed using primers complementary to the 5' end (5'-GCGCAAGCTTCGCATATGGCAGA AGAGAAGCCTTTTCAGTGTCGAA-3') and 3' end (5'- GCGCGGATCCCTTCTCGCCTGTGTGGGTCTTTAGGT GTCTCTGAAGAGTAGC-3') of the three zinc finger gene. The start methionine codon of the first zinc finger and end lysine codon of finger three are in bold in the respective primer sequence. The PCR product was restricted with NdeI and BamHI (underlined in the primer sequences above) and subcloned into these same sites in pHhaI, immediately 5' to the M.HhaI cDNA, to produce pzfHhaI. Double-stranded oligodeoxynucleotide coding for the flexible (Gly4Ser)3 peptide linker was subcloned between the BamHI and EcoRI restriction sites which separate the two cDNAs in pzfHhaI to produce pzfLHhaI. The resulting peptide sequence between the zinc finger and Mtase proteins was therefore TGEKGS(G4S)3SGEFM. Zinc finger and Mtase residues are underlined. The BamHI and EcoRI restriction sites are incorporated into the sequences coding for GS and EF dipeptides. The zinc finger --linker --M.HhaI gene was excised from pzfLHhaI, as a NdeI --XhoI fragment, and ligated into a pGEX5X-3 vector containing a modified multiple cloning site (P.J.Hurd, unpublished data), cut with the same enzymes to produce pGZfHhaI. M.HpaII (Genbank accession no. L17342, bases 1594 --2670) was amplified by PCR using pUCHpaII (10 ng) as template and the following primers, complementary to the 5' end (5'-GCGCGCGAATTCATGAAAGATGTGTTAG ATGATAACTTGTTA-3') and 3' end (5'-GCGCGCGTC GACGTCATATAAATTTCCTAATTTTTCTAA-3'). After restriction of the PCR product with EcoRI and SalI (sites underlined in the primer sequences above), the fragment was ligated into pGZfHhaI, pre-restricted with EcoRI and XhoI, resulting in direct replacement of the M.HhaI gene with that of M.HpaII (pGZfHpaII). Construction of a vector coding for M.HpaII as a fusion with GST was by direct ligation of the C5-Mtase as a BamHI --SalI fragment into the modified pGEX5X-3 vector, to give pGHpaII. The integrity of constructs was confirmed by sequencing. Expression from all vectors allowed purification of the recombinant protein via glutathione --agarose affinity chromatography. Protein expression and purification | Recombinant plasmids were used to transform E.coli ER1647 cells and single colonies were used to induce high level expression of the targeted enzyme. Induction of protein expression was carried out as follows. A single colony was inoculated into 50 ml of LB medium containing 0.5% glucose, grown overnight at 37C and cells were harvested by centrifugation. The cell pellet was resuspended in 1 ml of LB medium and then used to inoculate 500 ml of LB medium supplemented with 0.5% glucose and 100 microM ZnSO4. Growth was allowed to continue at 37C until the OD600 reached 0.8. Protein expression was induced by the addition of isopropyl-beta,d-thiogalactopyranoside to a final concentration of 1.0 mM and incubation was continued for a further 4 h at 30C. Induced cells were harvested and lysed by sonication on ice; solubilisation of the recombinant enzymes was aided by the addition of Triton X-100 to 1% (v/v). Soluble recombinant enzymes were purified to homogeneity (as judged by 10% SDS --PAGE and Coomassie blue staining; ) by glutathione affinity chromatography using glutathione --Sepharose (Sigma), essentially as described before . All protein concentrations were determined by quantitative analysis of an SDS polyacrylamide gel containing sample protein and a protein standard of known concentration. Imaging was performed using a Las 1000 CCD camera (Fujifilm) and Aida software (Raytest; Isotopenmessgerate, GmbH). Gel retardation analysis | All oligodeoxynucleotides were annealed in 10 mM Tris --HCl, 10 mM MgCl2, 50 mM NaCl, 1 mM dithiothreitol (pH 7.9 at 25C) in a PCR machine. Successful annealing was analysed by 3% agarose gel electrophoresis and ethidium bromide staining. Annealed oligodeoxynucleotide duplexes were labelled at the 5' end with [gamma-32P]ATP using phage T4 polynucleotide kinase. Unincorporated label was removed by gel filtration using Bio-Spin-6 columns (Bio-Rad Ltd). Typically, 0.05 pmol targeted or non-targeted enzyme was incubated in C5-Mtase buffer (20 mM HEPES, pH 7.5, 5 mM dithiothreitol, 20 microM ZnSO4, 5 mM MgCl2, 50 mM KCl, 10% glycerol, 0.1% Nonidet P-40, 0.1 mg/ml BSA) with 0.03 pmol labelled double-stranded oligodeoxynucleotide and 300 ng poly(dI)poly(dC), in a final reaction volume of 30 microl. S-adenosyl-l-homocysteine (AdoHcy) was added to a final concentration of 100 microM. Incubations were carried out for 35 min at room temperature. The reaction mixture was subsequently loaded onto a pre-run, 4% (w/v) non-denaturing polyacrylamide (acrylamide/bisacrylamide 19:1) gel in 1x TBE buffer and electrophoresed for 1 h at 100 V at room temperature. Electrophoresis was carried out as described previously. All protein --DNA complexes were visualised after autoradiography with intensifying screens at --70C overnight. (The high levels of protein relative to probe concentrations that were used in gel retardation assays were dictated by the conditions that yielded significant retardation of the probe, and therefore reflect the percentage activities of these protein preparations after the solubilisation and purification steps described above.) For dissociation constant determination, competitor oligonucleotide was added at time 0, over a final concentration range 15.4 --1078 nM (i.e. DNA always in excess). The reaction products were electrophoresed as described above. Dried down gels were imaged using an FLA-2000 Phosphorimager (Fuji). All experiments were performed at least in triplicate, unless stated otherwise, and used the same purified protein preparation. The proportion of bound competitor DNA was determined by subtraction of the retarded probe signal in the presence of competitor from that of a control shift in the absence of competitor, which was assigned an arbitrary value of 1.0 (i.e. 100%), using the imaging software Aida. This ensured that each gel shift assay could be normalised relative to its own internal control. Data was plotted as amount of competitor bound (normalised against zero competitor level shift intensities) against total amount of competitor added in each binding reaction. Data was fitted to the binding isotherm described by the equation [DNA:P] = ([Ptot][DNA]/[DNA] + Kd using Prizm (v.3.0) software (Graph Pad Inc., San Diego, CA), where [DNA:P] is the complex concentration, [Ptot] is the total protein concentration and [DNA] is the free DNA concentration. Under experimental conditions where the protein concentration is kept constant and the DNA levels are increased, the apparent dissociation constant (Kd) is always estimated as the concentration of free DNA when half the protein has been bound. [As an additional test of the data fit, Kd values derived from the non-linear regression approach described above were also compared with those calculated from data that were transformed and subjected to Scatchard analysis. In all cases, plots were clearly linear, suggesting single binding site kinetics, and Kd values were found to be corroborative (data not shown). Attempts to fit data to double binding site models failed in all cases.] Oligodeoxynucleotide methylation assays | Unless otherwise stated, 1 pmol enzyme was incubated in the relevant C5-Mtase buffer (see previous section) with 3 pmol double-stranded oligodeoxynucleotide, 200 ng poly(dI)poly(dC) and 1.65 microM [3H-methyl]AdoMet in a final reaction volume of 30 microl. Reactions were allowed to proceed for 30 min at 37C before being stopped by the addition of SDS to 1% or AdoHcy (100 microM), followed by heat treatment for 5 min at 72C in formamide gel loading buffer . The reaction mixture was subsequently loaded onto a pre-run 6% (w/v) denaturing (7 M urea) polyacrylamide (acrylamide/bisacrylamide 19:1) gel in 1x TBE buffer and electrophoresed for 1 h at 100 V at room temperature. Following electrophoresis, gels were fixed in 10% methanol, 10% acetic acid and either treated with the flurographic reagent Amplify (Amersham International) or dried down directly. Duplexes that had been methylated were identified by autoradiography with intensifying screens after 5 days at --70C. Titrational methylation assays were performed similarly to above, with protein and oligonucleotide concentrations as indicated in the relevant figures. Time course methylation assays contained 3 pmol duplex oligonucleotide and 20 fmol enzyme. For both these experiments, reactions were stopped by direct addition of formamide loading buffer and subsequent incubation at 85C for 10 min. After electrophoresis, gels were treated and dried down as described previously and exposed overnight to a tritium imaging plate, which was subsequently scanned using an FLA 2000 Phosphorimager and quantified using Aida software. Ex vivo methylation assays | A double-stranded oligonucleotide with 3' terminal single A overhangs, and harbouring adjacent zinc finger (underlined) and HpaII (bold) recognition sites (core sequence 5'-CTCCGGCTTCCATGGAGACGCAGAAGCCCT-3') was ligated into the pCR 2.1-Topo vector (Invitrogen) to generate the vector pZMol. A target site vector was made containing a larger insert but still containing only a single HpaII site. This was accomplished by PCR of pZMol using primers specific for the pCR 2.1-Topo vector at regions close to, but not including, the nearest flanking HpaII sites to the site introduced by oligonucleotide insertion described above (5'-AGCGGGCGCTAGGGCGCTGGCAAGTGT-3' and 5'-GTAGTTGTGTGGAATTGTGAGCGGAT-3'). The PCR product was then ligated into the pCR 2.1-Topo vector, to generate ZMTopo. Ex vivo methylation analysis involved incubation of 0.5 microg ZMTopo vector with increasing concentrations of recombinant purified targeted or non-targeted HpaII enzyme, in HpaII methylase buffer (see above) in the presence of 5 microM AdoMet, in a reaction volume of 10 microl. After either a 30 or 60 min incubation at 37C, the reaction was stopped by heating the samples at 85C for 10 min. The samples were then subjected to digestion by R.HpaII (0.2, 0.5 or 10 U) for 90 min at 37C. Samples were analysed on a 1% agarose gel. RESULTS : Plasmids coding for targeted methyltransferases are fully methylated in vivo | As a preliminary analysis of the targeted methylation properties of zinc finger --methyltransferase fusion enzymes, pGEX-based plasmids harbouring the genes for the GST --3Zf:HhaI or GST --3Zf:HpaII fusion proteins (henceforth referred to as Zf.M.HhaI and Zf.M.HpaII) and which contained multiple HhaI and HpaII restriction sites, were digested with HhaI or HpaII, respectively (Fig. A). Both these plasmids were shown to be fully resistant to cleavage, demonstrating that the Mtases had retained biological activity (lanes 6 and 12). Our expectation had been that the presence of the zinc finger component would reduce the level of wild-type methylation observed by the Mtase component. In order to further examine the nature of the methylation that occurred during bacterial growth, a number of mutant targeted HpaII constructs were made. Firstly, removal of the methionine codon at position 1 in the HpaII gene of the targeted HpaII construct generated a mutant (Mut2) that similarly and predictably produced a fully methylated plasmid (see Fig. B, lane 5 compared to 6). However, additional mutations, in which frameshifts were introduced into the GST --3Zf:HpaII vector, either at the start of the zinc finger gene of Mut2 (Mut3, Zf.HpaII gene out of frame with GST gene) or between the zinc finger and HpaII genes in the wild-type construct (Mut1, HpaII gene out of frame with GST --Zf gene), also gave rise to fully protected plasmids (Fig. B, lane 7 compared to 8 and lane 3 compared to 4). SDS --PAGE analysis of induced protein expression for each of these mutants revealed protein products of the predicted size (see Fig. C). Taken together, these results confirm that while the correct transcription and translational signals are being used in the cell under inducing conditions, internal start codons are also being used in E.coli, which give rise to methyltransferases lacking a fully functional zinc finger targeting component and at levels sufficient to fully methylate plasmid DNA under normal growth conditions. As such, further in vivo analysis of these enzymes in E.coli would have been uninformative under these conditions. In the following series of experiments, gel retardation and methylation analyses were used to characterise the interaction between targeted C5-Mtases (i.e. with a zinc finger moiety) and control, non-targeted C5-Mtase (i.e. no zinc finger moiety) and DNA (oligonucleotides listed in Table ). All experiments were carried out with glutathione S-transferase (GST) fused to the N-terminus of the recombinant enzymes via the zinc finger moiety (see Materials and Methods). The presence of GST has previously been demonstrated to have no effect on the activity of the C5-Mtase MspI, for example ( --), or on the binding of zinc fingers to DNA (,). Targeted C5-Mtases preferentially methylate DNA duplexes containing adjacent zinc finger and C5-Mtase recognition sites | The ability of Zf.M.HhaI to methylate DNA in a targeted manner was investigated by oligodeoxynucleotide methylation and competition --methylation assays. In order to discriminate between targeted and non-targeted methylation in the competition methyltransferase assays described, C5-Mtase site-containing oligodeoxynucleotides were synthesised as 29mers or 30mers (compared to 48mer target site oligodeoxynucleotides). This size difference enhances resolution of the two oligonucleotides during electrophoresis. The results of methylation assays designed to confirm the ability of Zf.M.HhaI to methylate both strands of the target sequence are shown in Figure A. (The designation of 'top' strand for all experiments described subsequently is defined by the sequence orientation of oligonucleotides listed in Table .) The enzyme is clearly able to transfer tritiated methyl groups from [3H-methyl]AdoMet to hemi-methylated cognate DNA duplexes (Fig. A, lanes 2 and 3). Duplexes that contained no C5-Mtase sites (lanes 5 and 6) or duplexes that were fully methylated (lane 4) were not subsequently methylated, confirming that the methylation observed for non-methylated or hemi-methylated duplex target site DNA had occurred only at the HhaI site. A significant level of methylation was observed by Zf.M.HhaI for duplexes that contained a HhaI site only (Fig. A, lane 7). The results of experiments designed to assess the ability of competitor oligonucleotides to modulate the methylation by Zf.M.HhaI of DNA substrates containing a HhaI site are presented in Figure B. Addition of target competitor duplex ZfHhaI (lanes 6 --8) or duplexes that contain the zinc finger recognition site alone (lanes 2 --4) cause a readily observable reduction in non-targeted methylation of duplexes containing only a M.HhaI site (lanes C), accompanied by a steady increase in degree of methylation of the competitor duplex (arrowed on the right hand side of Fig. B). Effective and complete competition is observed even at equimolar levels when the competing duplex contained the zinc finger recognition and C5-Mtase sites, rather than duplexes containing a zinc finger recognition site alone. In a repeat of the experiments described in Figure A for Zf.M.HhaI, the degree of non-targeted methylation of a 30mer oligonucleotide harbouring a single M.HpaII site by Zf.M.HpaII was only just detectable under conditions where the signal acquired by the target site-containing oligodeoxynucleotides was extremely intense (Fig. C, compare lanes 1 --3 with lane 6). Indeed, in three additional repeats of this assay, non-targeted methylation of the HpaII* oligonucleotide by Zf.M.HpaII was not detectable at all (data not shown). Similarly to Zf.M.HhaI, Zf.M.HpaII is able to methylate either strand of the target duplex (Fig. C, lanes 2 and 3). Due to the extremely low levels of non-targeted DNA methylation produced by Zf.M.HpaII, the competition experiments described for Zf.M.HhaI in Figure B could not be performed with this enzyme. However, a series of alternative competition experiments were conducted using the Zf.M.HpaII enzyme, in order to examine the ability of either zinc finger site-containing or non-specific oligodeoxynucleotide sequences to compete with methylation of a target oligonucleotide (ZfHpaII). Results are shown in Figure D. Specific methylation of the target oligodeoxynucleotide is not significantly affected by even a 10-fold molar excess of zinc finger site-containing oligodeoxynucleotide (Fig. D, lane 4). Similar results were obtained for competition by non-specific duplex (lanes 6 --8). These data further confirm that the C5-Mtase component of the targeted enzymes contributes significantly to the specificity of the targeted enzyme as a whole. Time course and titrational methylation assays were performed on Zf.M.HpaII in order to examine the relative catalytic activity at targeted and non-targeted Mtase sites. The results of titrational assays are shown in Figure A. The targeted enzyme is shown to catalyse methylation of the target oligonucleotide to an similar11-fold higher level than a duplex harbouring an M.HpaII site only. The relative catalytic activity of Zf.M.HpaII for a duplex harbouring a zinc finger and adjacent C5-Mtase site is similar to that of non-targeted HpaII for an oligonucleotide harbouring a C5-Mtase site. The results of time course methylation analyses, shown in Figure B, reveal a similar trend, with an 8-fold difference in activity for Zf.M.HpaII methylating target site- versus M.HpaII site-containing oligonucleotides. Only an similar2-fold difference in catalytic activity was observed between targeted and non-targeted HpaII for ZfHpaII DNA. Targeted HpaII C5-Mtase binds specifically and in cis to a recognition site comprising adjacent zinc finger and Mtase subsites | In order to obtain more quantitative binding data describing the interactions of such targeted enzymes with cognate and related sites, kinetic analyses of competition gel shift assays were performed on the Zf.M.HpaII and M.HpaII enzymes, as described in Materials and Methods. The competition approach ensures that within each set of comparative experiments identical levels of protein and probe are used, while the more readily quantifiable competitor DNA level is varied. (For all subsequent band shift assays reported the labelled duplex in each case was ZfHpaII; see Table .) The gel shifts are shown in Figure A. For the self-competition experiment, a full gel is shown. (For all subsequent experiments throughout this work, only the region of the gel showing retarded probe is shown. Each gel figure shown represents one of a triplicate of experiments from which binding data were calculated.) While absolute Kd values per se cannot realistically be determined via oligonucleotide competition approaches, highly reproducible relative Kd values can be determined that are a proportional reflection of the binding affinity of the enzyme for the various DNA substrates. Binding analysis of the gel shifts shown in Figure A are shown in Figure B and reveal that Zf.M.HpaII binds to its target site with an apparent Kd of 4.77 (+- 0.93) x 10 --9 M, but binds to zinc finger or C5-Mtase sites alone with Kd of 38.6 (+- 3.9) and 86.8 (+- 10.8) x 10 --9 M, respectively, similar8- and 18-fold differences in binding affinity, respectively. This result clearly demonstrates specific binding to a sequence comprising adjacent zinc finger and Mtase recognition sites. Under identical conditions, the three zinc finger protein (as a fusion to GST only) binds to its recognition site with an apparent Kd of similar14 x 10 --9 M (data not shown). M.HpaII binds to a M.HpaII site-containing oligonucleotide with a Kd of 35.08 (+- 7.9) x 10 --9 M, significantly weaker than for Zf.M.HpaII binding its target site, but stronger than for Zf.M.HpaII binding just a HpaII site. Thus the targeted HpaII enzyme displays both an enhanced binding affinity compared to its individual components and a reduced affinity for individual subsites associated with each component of the fusion enzyme. Double competition binding experiments for Zf.M.HpaII, using a combination of the competitor duplexes HpaII and Zf, which contain single M.HpaII and zinc finger protein recognition sites, respectively, yielded an apparent Kd of 93.65 (+- 24.6) x 10 --9 M. This is a much larger value than found for binding target site and zinc finger site-containing oligonucleotides alone, despite the opportunity for interaction of each component of the fusion protein with its own recognition subsite, and suggests that both components of the targeted enzyme preferentially bind to linear DNA, with subsites arranged in cis. Competition experiments using an oligonucleotide in which the HpaII site replaced the first 3 bp of the zinc finger subsite (zHpaII, bottom gel in Fig. A) gave an apparent Kd of 169.4 (+- 29.7) x 10 --9 M, again much larger than that found for binding to just a HpaII site-containing oligonucleotide. All these results taken together confirm that the binding affinity obtained for the interaction of the enzyme with the target site sequence is a result of both the C5-Mtase and zinc finger subunits of the targeted enzyme, acting in concert and in cis on the same DNA substrate. Targeted HpaII binding is modulated by the methylation status of the target site | In this experiment, competitor oligonucleotides were synthesised in which either the top, bottom or both top and bottom strand target cytosine positions contained methylcytosine. Representative gel mobility titrations showing the effect of methylation status of the target site on binding of Zf.M.HpaII are shown in Figure A. Kinetic analysis of these data is shown in Figure B. The enzyme binds to the ZM3 oligonucleotide (bottom cytosine methylated) with an apparent Kd of 4.28 (+- 1.2) x 10 --9 M and to the ZM5 oligonucleotide (top strand methylated) with a Kd of 7.6 (+- 0.87) x 10 --9 M, nearly 2-fold weaker. The protein bound to the fully methylated target site ZM5/3 with an apparent Kd of 9.25 (+- 0.53) x 10 --9 M. Taken together, these results indicate a slight preference for the Mtase component of Zf.M.HpaII binding the top strand, presumably as a consequence of motional constraints placed upon the targeted enzyme by the zinc finger --DNA interactions or of the flexibility of the (Gly4Ser)3 linker between the zinc finger and C5-Mtase components of the targeted enzyme. Targeted HpaII C5-Mtase binds and methylates specifically at recognition sites with variable subsite spacing | The ability of the targeted HpaII C5-Mtase enzyme to bind specifically to oligodeoxynucleotides with an increasing base pair separation between the zinc finger and M.HpaII subsites was examined by gel retardation competition and oligonucleotide methyltransferase assays. Figure A shows the gel shift assays resulting from competition with duplexes harbouring zinc finger and M.HpaII subsites separated over the range 2 --44 bp (oligonucleotides Sep2 --Sep44, see Table ). Figure B shows a plot of the binding affinity of the enzyme for each subsite spacing. Also presented on the graph are the results of methylation assays for each oligonucleotide (for details see figure legend). The values have been normalised against the apparent Kd values, so that they can be presented in the same figure. It is immediately apparent that binding and methylation curves correlate extremely well, being nearly mirror images of each other, i.e. as the apparent Kd reduces (tighter binding) the methylation signal increases. A general area of tight binding and high methylation rate can be seen to occur from similar10 bp up to similar40 bp. Two local optima, coincident for both binding and methylation, centred at similar13 bp and at 34 bp separation, may also be present. An extremely sharp transition point can be seen to occur between the 16 and 17 bp subsite separations, where both binding and methylation become weaker. This is a significantly larger change than is seen for the 5 to 6 bp subsite spacing transition and probably reflects a 'stress site', at which flexibility of the fusion protein linker and the related orientational preference of the methylase component directly conflict with the spatial presentation of the target cytosines at the Mtase subsite. Targeted HpaII specifically methylates plasmid DNA harbouring a target sequence ex vivo, against a background of non-targeted HpaII sites | The vector ZMTopo was constructed, harbouring a single targetable site (see Materials and Methods) but containing an additional 23 M.HpaII recognition sites. In the absence of any methylation, the target plasmid should yield a characteristic restriction pattern after digestion by R.HpaII. Specific methylation at the target site would result in an alteration of this restriction pattern, giving rise to a unique 1060 bp fragment. A schematic describing this rationale is shown in Figure A. A titrational ex vivo methylation assay was performed, incubating target site plasmid ZMTopo with increasing amounts of recombinant Zf.M.HpaII, in order to simulate the in vivo scenario of methylating more complex substrates. The DNA was subsequently digested with R.HpaII and the digestion products analysed by agarose gel electrophoresis. The results are shown in Figure B. A band corresponding to the expected size of 1060 bp can be clearly seen, increasing in intensity in those lanes that had been incubated with Zf.M.HpaII (lanes 1 --6), but not in lanes that had been incubated just with M.HpaII (lanes 7 --12). The identification of this band as being the target site-containing fragment was confirmed by further digestion of the DNA with restriction enzymes whose sites only occurred within the target fragment (see Fig. A). As can be seen in Figure C, upon restriction with EcoNI and BamHI, the 1060 bp fragment disappears. The appearance of a unique band from the target site vector in this assay only when incubated with targeted enzyme and corresponding in size to that expected from the result of targeted methylation, and the confirmation of predicted restriction enzyme sites within this fragment, confirms our conclusion that targeted methylation is occurring under these conditions. Figure 1 | Restriction protection analysis of in vivo methylated plasmid DNA. Restriction protection analysis of in vivo methylated plasmid DNA. (A) Plasmids encoding targeted HhaI and HpaII were isolated from E.coli and subjected to restriction by restriction enzymes HhaI and HpaII respectively, as described in Materials and Methods. Lane 1, uncut control pGEX5X-3 vector; lane 2, R.HhaI restricted pGEX5X-3 vector; lane 3, uncut pGHhaI vector; lane 4, R.HhaI restricted pGHhaI vector; lane 5, uncut pGZfHhaI vector; lane 6, R.HhaI restricted pGZfHhaI vector; lane 7, as lane 1; lane 8, R.HpaII restricted pGEX5X-3 vector; lane 9, uncut pGHpaII vector; lane 10, R.HpaII restricted pGHpaII vector; lane 11, uncut pGZfHpaII vector; lane 12, R.HpaII restricted pGZfHpaII vector; lane m, 100 bp marker (NEB); lane kb, 1 kb ladder (NEB). (B) Plasmids encoding mutant targeted HpaII were isolated from E.coli and subjected to restriction by HpaII. Mutants are: Mut1, wild-type vector cut with EcoRI and filled in to generate HpaII out of frame with GST --Zf; Mut2, wild-type vector with first codon (methionine) of HpaII removed; Mut3, Mut2 vector cut with NdeI and filled in to generate Zf-HpaII out of frame with GST (for details see text). Lanes 1 and 2, uncut and R.HpaII cut wild-type targeted HpaII vector, respectively; lanes 3 and 4, uncut and R.HpaII cut Mut1 vector; lanes 5 and 6, uncut and R.HpaII cut Mut2 vector; lanes 7 and 8, uncut and R.HpaII cut Mut3 vector; lane 9, R.HpaII cut pGEX empty vector. (C) SDS --PAGE analysis of protein induction for wild-type and mutant targeted HpaII enzymes. Molecular weights are indicated on the right of the gel. Induced protein products are arrowed. Figure 2 | Methylation analysis of targeted and non-targeted HhaI and HpaII enzymes. Methylation analysis of targeted and non-targeted HhaI and HpaII enzymes. (A) Oligodeoxynucleotide methylation assays were performed to confirm double-strand methylation by Zf.M.HhaI. All reactions contained 1 pmol Zf.M.HhaI protein and 3 pmol duplex DNA as indicated in each lane. Reactions were carried out as described in Materials and Methods. The designations 5M, 3M and 5/3M, in this and other figures, refer to the oligonucleotides being pre-methylated at the target cytosine on either the top, bottom or both DNA strands respectively. (B) Methylation competition assays. All lanes contained 1 pmol Zf.M.HhaI protein and 3 pmol HhaI oligodeoxynucleotide and competitor DNA at the following levels: lane C, no competitor DNA; lanes 2 --4, Zf oligodeoxynucleotide added as competitor at 1-, 5- and 10-fold molar excess, respectively; lanes 6 --8, ZfHhaI oligodeoxynucleotide added similarly at 1-, 5- and 10-fold molar excess. The arrows indicate the different mobilities of the oligonucleotides used. (C) Oligodeoxynucleotide methylation assays were performed to confirm double-strand methylation by Zf.M.HpaII. All lanes contained 1 pmol Zf.M.HpaII protein and 3 pmol duplex DNA as indicated in each lane. (D) Competition methylation analysis. All reactions contained 1 pmol Zf.M.HpaII, 3 pmol ZfHpaII oligonucleotide and competitor DNA at the following levels: lane c, no competitor DNA; lanes 2 --4, as lane c except for the addition of 1-, 5- and 10-fold molar excess of ZfHhaI competitor oligodeoxynucleotide (i.e. effectively zinc finger only site); lanes 6 --8, as lane c but with the addition of 1-, 5- and 10-fold molar excess of non-specific (Non-Sp.) oligodeoxynucleotide. Figure 3 | Titrational and time course analysis of targeted and non-targeted HpaII methylation for different oligonucleotide substrates. Titrational and time course analysis of targeted and non-targeted HpaII methylation for different oligonucleotide substrates. (A) Increasing concentrations of Zf.M.HpaII and M.HpaII enzyme were incubated with ZfHpaII or HpaII oligonucleotides (final concentration 150 nM) over the concentration range shown. The relative levels of methylated oligonucleotide, as evaluated by phosphorimager analysis, were plotted against protein concentration. The intensity of the tritiated oligonucleotides are plotted as PSL values (photo-stimulated luminescence), which are directly proportional to the radioactivity of the samples being measured. The best fit line for Zf.M.HpaII binding to ZfHpaII oligonucleotide is denoted by a solid line. (B) Time course methylation profile for the interaction of targeted and non-targeted HpaII with oligonucleotide substrates. Reactions contained 20 fmol protein and 3 pmol DNA (for details see Materials and Methods). Figure 4 | (A) Gel shift competition assays. (A) Gel shift competition assays. ZfHpaII probe (1.0 nM) was incubated with M.HpaII or Zf.M.HpaII enzymes (1.7 nM) (see Materials and Methods). Lane C, no competitor DNA; subsequent lanes contain competitor DNA at final concentrations of 15.5, 30.8, 77, 154, 308, 616 and 1078 nM. The competitor DNA used is indicated on the right of each gel; the protein assayed in each case is indicated on the left of each gel. For the self-competition experiment, the full retardation gel is shown. For all subsequent experiments, only that portion of the gel containing the retarded probe is shown. (B) Analysis of triplicate binding data described in (A) above. Figure 5 | Gel retardation competition analysis of the interaction of Zf.M.HpaII with pre-methylated target site DNA. Gel retardation competition analysis of the interaction of Zf.M.HpaII with pre-methylated target site DNA. (A) ZfHpaII probe was incubated with Zf.M.HpaII enzyme as in Figure . Lane C, no competitor DNA; subsequent lanes contain competitor DNA at final concentrations of 15.5, 30.8, 77, 154, 308, 616 and 1078 nM. The competitor DNA used is indicated on the right of each gel. (B) Analysis of triplicate binding data described in (A). Oligonucleotides ZM3, ZM5 and ZM5/3 are identical to oligonucelotide ZfHpaII but contain methylcytosine at the target site (i.e. mCCGG) on the top, bottom and both strands, respectively. Figure 6 | Zf.M.HpaII binds specifically to oligodeoxynucleotides with variable 'subsite' spacings. Zf.M.HpaII binds specifically to oligodeoxynucleotides with variable 'subsite' spacings. (A) Gel retardation analysis. Lane C in all cases contains Zf.M.HpaII enzyme and ZfHpaII labelled probe at the same levels as shown in Figure . Competitor DNA was added to the same final concentrations as given in Figure . The base pair separation between subsites for each competitor oligonucleotide is shown to the left of each gel. (B) The apparent Kd values for triplicate competition binding experiments for each subsite spacing were plotted against the subsite spacing. The methylation signal obtained from incubation of each oligonucleotide (30 pmol) with Zf.M.HpaII protein (0.1 pmol) for 30 min in the presence of [3H]AdoMet is also plotted on the same graph. The curve shown is representative of a triplicate set, which all followed the same trend. The methylation intensities have been normalised to fit on the graph. Figure 7 | Ex vivo methylation analysis of Zf.M.HpaII interacting with target site-containing plasmid DNA. Ex vivo methylation analysis of Zf.M.HpaII interacting with target site-containing plasmid DNA. (A) Schematic outline of the key elements of the ZMTopo vector used as substrate in the ex vivo studies described below (see also Materials and Methods). The region of the plasmid harbouring the zinc finger and single flanking HpaII site is shown as a grey box with a lollipop on it. In the scenario where this site has been preferentially methylated (designated by a black lollipop, compared to unmethylated vector HpaII sites, shown in grey), the nearest cleavage by R.HpaII will only occur at flanking HpaII sites a and b, resulting in the generation of a 1060 bp fragment. If this site is not methylated, derivative fragments of similar280 and 780 bp will be produced. Restriction enzyme sites for EcoNI and BamHI are also shown. These sites are unique to DNA fragments harbouring the target sequence. (B) Target site-containing vector ZMTopo was incubated with increasing amounts of Zf.M.HpaII or M.HpaII enzyme for 30 min prior to digestion with R.HpaII. Lanes 1 --6 and 7 --12 represent digestion of vector preincubated with 25, 50, 75, 125, 175 and 225 fmol Zf.M.HpaII and M.HpaII protein, respectively. The expected 1060 bp DNA fragment indicative of targeted methylation is arrowed. Lane U, unrestricted ZMTopo DNA; lane C, unmethylated vector DNA restricted with R.HpaII; lane M, 100 bp ladder (NEB). Key size bands are indicated. (C) All lanes were identical to those described for lane 4 in (B), but with the inclusion of 10 U of the restriction enzyme EcoNI (lane 1) or BamHI (lane 2) or just water (lane c). The 1060 bp DNA fragment present in lane c is arrowed. Table 1 | Oligodeoxynucleotides used during protein --DNA binding and methylation studies DISCUSSION : We have presented data describing the evaluation of fusion proteins comprising a three zinc finger protein covalently linked to either HpaII or HhaI Mtase via a flexible peptide linker. The principle behind this approach was to bias the action of the Mtase component to specific recognition sites than were proximal to the 9 bp DNA sequence recognised by the zinc finger protein. The expectation of this initial approach would be that the intrinsic specificity of each component of the fusion protein should remain unaltered, although one would expect changes in general binding affinities for each component for its subsite, as part of a fusion protein. We have evaluated whether this was indeed the case in detail for the targeted HpaII enzyme, as well as examining orientational and subsite spacing preferences. The initial discovery that plasmids encoding targeted methyltransferases were fully methylated suggested initially that fusion of a Mtase to the zinc finger protein did not modulate either Mtase activity or specificity in E.coli. However, the observation of similar methylation by constructs specifically mutated so as not to express a functional targeted methyltransferase suggested that internal transcription and/or translation occurred from vectors coding for the targeted enzymes in E.coli. Given the repetitive nature of the DNA linker sequence joining the zinc finger and methyltransferase genes, the possibility of both translational and transcriptional stalling present themselves, allowing alternative 're-start' points for both events to re-occur along the template. Due to the nature of methyltransferases, only a small number of non-targeted molecules would need to be produced to fully protect the coding plasmid under typical overnight growth conditions. This property suggests methyltransferases as excellent reporters for constitutive and low level expression/translation studies. However, any evaluation of targeted methylation efficacy in bacteria using this current expression system was not possible, due to the considerations mentioned above. Comparative in vitro analysis of the catalytic activity of purified Zf.M.HhaI and Zf.M.HpaII enzymes revealed that Zf.M.HhaI in fact catalyses a significant level of methylation of duplexes harbouring a M.HhaI site alone (i.e. it is able to function as a discrete methyltransferase at some level, irrespective of its zinc finger component). Binding studies, however, revealed no interaction between targeted HhaI and this duplex (data not shown). Such non-targeted DNA methylation has also been described for the only other report of a targeted C5-Mtase, a fusion between a three zinc finger protein and the 5'-CpG-3'-specific prokaryotic SssI C5-Mtase . Consequently, this enzyme has limited use in in vivo applications. Non-specific methylation associated with the targeted M.SssI enzyme may be related to the observed processive action of this enzyme, proceeding along CpG-containing substrate molecules and methylating one strand of DNA at a time. Such activity would exacerbate the level of non-targeted methylation expected. The HhaI and HpaII C5-Mtases, however, do not display the same processive characteristics as described for M.SssI . The results of methylation competition experiments described for targeted HhaI and HpaII broadly correlate with data from gel shift assays and provide confirmation that both the DNA binding and DNA methylation events are targeted phenomena. The degrees of targeted and non-targeted methylation observed between these two enzymes might reflect differences in their 3-dimensional structures, which would modulate the interaction between the C5-Mtase and zinc finger moieties. It may be significant that the N-terminal regions of each C5-Mtase differ in length before the start of homology block I (,). M.HhaI has 13 amino acids and M.HpaII has 34 amino acids in this region, which are fused directly to the linker sequence between the zinc finger and C5-Mtase components. Contributions made by each region to the overall nature and flexibility of the existing linker region would be expected to modulate the characteristics of each enzyme and represent a potential avenue of exploration for additional studies aimed at further enhancing the specificity of these enzymes. An earlier targeted HhaI prototype enzyme, containing a (Gly)6 linker, was in fact non-functional (data not shown), and confirms the importance of the linker sequence between the components of a fusion enzyme, as demonstrated in other studies (, --). Titrational and time course methylation assays on the targeted HpaII enzyme interacting with target and C5-Mtase site substrate oligonucleotides confirmed methylation specificity and that the enzymatic properties of the C5-Mtase component of the targeted enzyme had not been impaired as a consequence of its fusion protein status. Interestingly, methylation of the shorter (30mer) duplexes by targeted M.HpaII was significantly lower than for the longer DNA substrates (48mers) used in time course and titration assays, and may reflect a requirement for an initial non-specific interaction between Zf.M.HpaII and DNA before methylation proceeds at the specific site. Binding studies on the active targeted HpaII enzyme are necessarily complicated by the fact that the Mtase component of the fusion enzyme has an intrinsic catalytic activity. During the course of a normal binding assay, for example, one would expect large protein conformational changes to occur as well as forced dissociation of the protein from the DNA subsequent to target site methylation. An additional aspect would be the generation of fully methylated and hemi-methylated DNA products that would re-enter the general pool of DNA to be re-bound. A further complication lies in the proposed base flipping mechanism associated with C5-Mtases, which may have an associated DNA 'tracking' component and which might also be expected to distort the DNA. To counter the problems associated with catalysis interfering with accurate measurement of binding affinities, we have performed all binding assays in the presence of the commonly used cofactor analogue AdoHcy, which lacks a methyl donor group. Thus catalytic acceleration of dissociation during the binding assay has been addressed to some degree. Kinetic analysis of Zf.M.HpaII binding to a series of different oligonucleotides representing target sequence permutations revealed that the highest affinity interaction was found for Zf.M.HpaII binding to its full target site, which was 8-fold tighter than to a zinc finger site oligonucleotide and 18-fold tighter than to a Mtase site oligonucleotide. These results confirm that the targeted enzyme displays a binding affinity greater than either of its component parts, although the greater contribution is most likely from the zinc finger moiety and suggests a degree of cooperativity between each component in the binding event. Evidence of cooperativity in binding also exists, from the results of double competition experiments using separate Zf- and Mtase site-containing competitor oligonucleotides. The resultant apparent Kd values are significantly larger than for the protein binding to target site or zinc finger site oligonucleotides alone and are in fact more comparable to that of binding to a methylase site-containing oligonucleotide. This is an unusual result, since one would expect the zinc finger oligonucleotide component of the double titration assay to exert a similar competitive effect as in previous experiments, when it was sole competitor. It can only be inferred that the fusion enzyme cannot interact with subsites located on separate DNA molecules with the same efficiency as each protein would in its wild-type form, or that such interactions are now short lived due to steric problems. For example, crystal structure analysis of the HhaI Mtase and its complex with DNA shows that significant conformational changes occur upon DNA binding , and it is conceivable that such movement in the methyltransferase component of the targeted enzyme would lead to a reduction in zinc finger --DNA interactions in a trans binding scenario. Similarly, competition by the mutated zinc finger site oligonucleotide zHpaII, in which the HpaII site directly encroaches into the zinc finger site to the extent that only 6 of the original 9 bp of the zinc finger site are present, is significantly less effective than by a simple HpaII site-containing oligonucleotide, and again argues for a cooperative (albeit negative in this instance) effect. Similar 'bi-modal' binding systems, such as covalently linked zinc finger multi-arrays or zinc finger --transcription factor fusion enzymes, have also been shown to demonstrate cooperativity in binding between component proteins when similarly analysed (,). Such cooperativity as seen in these and our own study, where the binding affinity for the hybrid target site is significantly greater than the affinity of each protein component for its respective site, is thought to be largely due to an enhanced local concentration of each protein component on the DNA. This is due to 'tethering' of each component near the DNA, as a consequence of binding of the other component to its recognition subsite. Cooperative interactions might also arise due to direct interaction between component proteins of the fusion enzyme, resulting in an enhancement of their DNA binding affinity. Linker-associated interactions may also contribute to the binding event. Binding studies using pre-methylated competitor duplexes showed that the C5-Mtase component of Zf.M.HpaII preferentially interacts with the top strand of the target duplex. This is most likely attributable to the flexibility of the linker peptide joining the zinc finger and C5-Mtase components. The lower binding affinity observed for the fully methylated duplex reflects the fact that a fully methylated Mtase site is no longer a substrate for methylation and thus not an optimal binding site. Significantly, DNA methylation analysis of both the Zf.M.HhaI and Zf.M.HpaII enzymes revealed that these enzymes are capable of transferring methyl groups to both strands of the target sequence at the respective C5-Mtase subsites. In contrast to gel shift competition assays, however, under the conditions of these methylation experiments no obvious strand preference can be inferred. The ability of Zf.M.HpaII to specifically bind duplexes with variable spacing between the zinc finger and C5-Mtase subsites was shown to correlate with the methylation activity of the enzyme for each oligonucleotide. In both instances, maximum activity/affinity occurred over the range 10 --40 bp separation between subsites. Within this region, evidence for the existence of two local optima is observed at 13 and 32 bp. These fluctuations fall at an interval corresponding to whole helical turns of DNA and are not inconsistent with the target cytosines being optimally positioned for interaction with the methylase component of the fusion enzyme at these particular subsite spacings. The ability of the C5-Mtase component to specifically recognise and bind its subsite up to 44 bp away from the anchored zinc finger subunit suggests great potential for this enzyme construct in delivering methylation over a significant length of DNA. The binding affinity at this separation (26 nM) is still stronger than that of the protein binding to just a zinc finger site (38 nM). Interestingly, binding of the enzyme to DNA with a subsite separation of only 2 bp was still significantly tighter than to oligonucleotides with 'infinite' spacing (i.e. the Zf and HpaII oligonucleotides). This suggests, rather surprisingly, that even at such a close subsite separation, where in fact the HpaII and zinc finger sites overlap but are maintained (see Table ), both components of the fusion enzyme are still contributing to the binding event in a cooperative manner. This conclusion is also in accord with results obtained for binding of the protein to the zHpaII oligonucleotide previously discussed. Given that when the linker can be considered either relaxed (at 2 bp subsite separation) or extended (at 44 bp separation) cooperative binding is still observed, it seems unlikely that the linker contributes significantly to the cooperative binding event. Preliminary titrational methylation assays had suggested that the DNA:protein ratio may represent a critical factor for the successful function of these targeted enzymes. This conclusion was confirmed when methylation analysis of a target site plasmid revealed successful targeted methylation at the target site and where the majority of vector DNA was unmethylated. Such targeted methylation occurred at specific ratios of protein:DNA concentration, which we have calculated to be in the range 1:10 --1:50. This observation is also significant, given that in vivo studies on the zinc finger protein activation of target promoters has shown that at high cellular protein concentrations, non-specific binding and transcriptional activation from non-target site-containing promoters was observed . Here we have constructed targeted C5-Mtases that can methylate DNA in a targeted manner. In the absence of a target site, these enzymes are not absolutely mono-functional, although Zf.M.HpaII is close to being so. When the enzymes are presented with both target site- and non-target site-containing DNA, methylation occurs at the target site in preference to any other. The targeted HpaII enzyme displays the most specific binding and methylation attributes and has been shown to be able to deliver targeted methylation over at least a 35 bp region. Methylation analyses using complex substrates, such as target site vectors, more closely approach in vivo conditions and have shown that targeted methylation is achievable at specific ratios of protein to DNA. However, a key problem still exists with these enzymes, which is that the level of normal, i.e. non-targeted, methylation is quite high. In the absence of target sites, the affinity of the targeted enzyme for Mtase sites alone is high enough for significant methylation to occur. Second generation targeted enzymes are therefore currently being evaluated, in which the C5-Mtase moieties harbour mutations that result in reduced DNA binding affinity at the Mtase recognition site and reduced catalytic activity . Such lower activity would be compensated for by the tethering effect of the zinc finger component, which should occur only at the target site. These approaches should significantly reduce, if not eliminate, non-targeted methylation. Additionally, the zinc finger protein used in this study has only moderate binding affinity for its recognition site and targeting could certainly be further enhanced by use of tighter binding zinc finger proteins or larger zinc finger arrays. While there is still room for significant advances to be made in this area, these current and future enzymes represent valuable tools for the study of DNA methylation spread and control in in vivo studies, as well as potential effectors for targeted promoter gene silencing induced through a DNA methylation-mediated cellular response. Backmatter: PMID- 12202759 TI - Synthesis and polymerase incorporation of 5'-amino-2',5'-dideoxy-5'-N-triphosphate nucleotides AB - Owing to the markedly increased reactivity of amino functional groups versus hydroxyls, the 5'-amino-5'-deoxy nucleoside and nucleotide analogs have proven widely useful in biological, pharmaceutical and genomic applications. However, synthetic procedures leading to these analogs have not been fully explored, which may possibly have limited the scope of their utility. Here we describe the synthesis of the 5'-amino-2',5'-dideoxy analogs of adenosine, cytidine, guanosine, inosine and uridine from their respective naturally occurring nucleosides via the reduction of 5'-azido-2',5'-dideoxy intermediates using the Staudinger reaction, and the high yield conversion of these modified nucleosides and 5'-amino-5'-deoxythymidine to the corresponding 5'-N-triphosphates through reaction with trisodium trimetaphosphate in the presence of tris(hydroxymethyl)aminomethane (Tris). We also show that each of these nucleotide analogs can be efficiently incorporated into DNA by the Klenow fragment of Escherichia coli DNA polymerase I when individually substituted for its naturally occurring counterpart. Mild acid treatment of the resulting DNA generates polynucleotide fragments that arise from specific cleavage at each modified nucleotide, providing a sequence ladder for each base. Because the ladders are generated after the extension, the corresponding products may be manipulated by enzymatic and /or purification processes. The potential utility of this extension --cleavage procedure in genomic sequence analysis is discussed. Keywords: INTRODUCTION : Chemically modified nucleotides have been extensively used in the study of many complicated biological systems. In particular, they have proven indispensable in the analysis of protein --nucleic acid interactions, the sequencing of nucleic acids and, more recently, the determination of genotypes. In general, these applications rely on differences in the chemical reactivity or steroelectronic properties of the modified nucleosides as compared to the naturally occurring counterpart. Owing to its markedly increased reactivity, 5'-amino-5'-deoxythymidine (NH2-dT) has been employed in a variety of studies, including model studies on DNA replication , template-directed chemical amplification , mechanistic studies on polymerases and reverse transcriptases and the construction of combinatorial peptide --DNA hybrids and their libraries (,) for antisense applications. Although 5'-amino-5'-deoxythymidine-5'-N-triphosphate (NH2-dTTP) was initially described in the 1970s (,), interest in 5'-amino-2',5'-dideoxynucleotides (NH2-dNTPs) has revived recently due to their potential utility in genomic analysis. Shchepinov et al. have reported a method for the high throughput detection of single nucleotide polymorphisms (SNPs) by MALDI-TOF mass spectrometry that involves the matrix-assisted fragmentation of DNA containing NH2-dT or NH2-dC. Recently, we have developed a novel dinucleotide cleavage method involving polymerase incorporation of a ribonucleotide and a 5'-amino-2',5'-dideoxynucleotide that may be utilized for SNP discovery by MALDI-TOF mass spectrometry (Wolfe,J.L., Wang,B.H., Kawate,T., DeMaria,C. and Stanton,V.P., submitted for publication). In order to take full advantage of these novel technologies and to provide additional tools for model studies that utilize 5'-amino-nucleosides (, --), it is necessary to develop a simple and reliable method for generating 5'-amino-2',5'-dideoxynucleosides (NH2-dNs) and corresponding 5'-N-triphosphate nucleotides (NH2-dNTPs). The present report describes the synthesis and characterization of 5'-amino-2',5'-dideoxy-5'-N-triphosphate nucleotide analogs for adenosine, cytidine, guanosine, thymidine, uridine and inosine. Each of these analogs is readily incorporated into DNA by the Klenow fragment of DNA polymerase I and can partially or completely replace its naturally occurring counterpart. The phosphoramidate bonds formed within the resulting polynucleotides can be specifically cleaved with dilute acetic acid, thereby generating a high quality sequencing ladder for the corresponding base. Furthermore, we demonstrate that when NH2-dTTP is partially incorporated into a 7.2 kb polynucleotide the resulting DNA duplex is a good substrate for restriction endonucleases such as MscI. We propose that these unique properties may render the NH2-dNTPs exceptionally well suited for genomic sequence analysis. MATERIALS AND METHODS : Reagents and solvents were purchased from Sigma-Aldrich (St Louis, MO) or J. T. Baker (Phillipsburg, NJ). The 1H and 31P NMR spectra were recorded on a Bruker DPX-400 NMR spectrometer using either tetramethylsilane (TMS) as an internal standard (1H spectra) or 80% H3PO4 as an external standard (31P spectra) (Dr Jin Hong, 77 Pine Ridge Drive, Ayer, MA). Chemical shifts (delta) are reported in p.p.m. downfield from TMS or H3PO4. Column chromatography was performed with silica gel 60 (230 --400 mesh, Merck 9385) using standard flash methods. Analytical thin-layer chromatography (TLC) was carried out on Merck silica gel 60 F254 pre-coated plates. Oligonucleotides were purchased from Sigma-Genosys (Woodlands, TX). Nucleotide dNTPs were from Amersham Biosciences (Piscataway, NJ). Synthesis of 5'-azido-2',5'-dideoxyadenosine (7) | To a solution of 599 mg (1.6 mmol) of 5'-azido-N6-benzoyl-2',5'-dideoxyadenosine in 4 ml of methanol was added 6 ml of concentrated NH4OH. The resulting mixture was stirred at 58 --60C for 15 h and then concentrated to dryness in vacuo. The residue was dissolved in H2O (60 ml) and extracted with CHCl3 (5 x 5 ml). The aqueous layer was collected and the organic solutions were combined and back-extracted with H2O (2 x 5 ml). Aqueous solutions from both extractions were combined and lyophilized. The residue was fractionated on silica gel (CH2Cl2:MeOH, 49:1 --9:1) to afford 350 mg 7 (80%) as a light yellow foam. 1H NMR (DMSO-d6): delta 2.30 (1H, ddd, J = 3.7, 6.4, 13.3 Hz, 2'a), 2.92 (1H, td, J = 6.7, 13.4 Hz, 2'b), 3.48 (1H, ddd, J = 3.9, 12.9 Hz, 5'a), 3.65 (1H, ddd, J = 7.3, 12.9 Hz, 5'b), 3.96 (1H, m, 4'), 4.41 (1H, m, 3'), 5.48 (1H, d, J = 4.0 Hz, 3'OH), 6.37 (1H, t, J = 6.9 Hz, 1'), 7.29 (2H, s, NH2), 8.14 (1H, s, 2), 8.33 (1H, s, 8). Synthesis of 5'-azido-2',5'-dideoxycytidine (8) | Compound 5'-azido-N4-benzoyl-2',5'-dideoxycytidine (454 mg, 1.27 mmol) was dissolved in a mixture of 5 ml of pyridine and 5 ml of concentrated NH4OH and the resulting solution was stirred at 60C for 8.5 h. The mixture was cooled on ice and subjected to rotary evaporation. The residue was partitioned between H2O (30 ml) and EtOAc (3 ml) and the aqueous layer was washed with Et2O (3 x 3 ml) and EtOAc:Et2O (1:2, 2 x 5 ml). Lyophilization of the aqueous layer gave 316 mg of 8 (98%). 1H NMR (DMSO-d6): delta 2.07 (2H, m, 2'), 3.51 (1H, dd, J = 4.4, 13.1 Hz, 5'a), 3.56 (1H, dd, J = 6.2, 13.1 Hz, 5'b), 3.83 (1H, m, 4'), 4.12 (1H, m, 3'), 5.38 (1H, d, J = 4.3 Hz, OH), 5.73 (1H, d, J = 7.4 Hz, 5), 6.21 (1H, t, J = 6.8 Hz, 1'), 7.16 (1H, brs, NH), 7.20 (1H, brs, NH), 7.60 (1H, d, J = 7.4 Hz, 6). Synthesis of 5'-azido-2',5'-dideoxyguanosine (9) | To a solution of 409 mg (1.13 mmol) 5'-azido-N2-isobutyryl-2',5'-dideoxyguanosine in 4 ml of methanol was added 6 ml of concentrated NH4OH. The mixture was stirred at 58 --60C for 13 h and then concentrated to dryness in vacuo. The residue was dissolved in H2O and extracted with EtOAc. The aqueous layer was collected and lyophilized to afford 340 mg of 9 (100%) as a white powder. 1H NMR (DMSO-d6): delta 2.22 (1H, ddd, J = 3.2, 6.1, 13.2 Hz, 2'a), 2.70 (1H, td, J = 6.3, 13.5 Hz, 2'b, 3.46 (1H, dd, J = 4.2, 13.0 Hz, 5'a), 3.61 (1H, dd, J = 7.2, 13.0 Hz, 5'b), 3.91 (1H, m, 4'), 4.29 (1H, m, 3'), 5.44 (1H, br, OH), 6.14 (1H, t, J = 6.9 Hz, 1'), 6.53 (2H, brs, NH2), 7.90 (1H, s, 8), 9.48 (1H, brs, NH). Synthesis of 5'-amino-2',5'-dideoxyadenosine (12) | A solution of 34 mg (0.12 mmol) of 7 and 100 mg triphenylphosphine (Ph3P; 0.38 mmol) in 1 ml of pyridine was stirred at room temperature for 7 h. Concentrated aqueous NH4OH (0.2 ml) was added and the mixture stirred at room temperature overnight. After additional stirring at 55C for 1 h, 5 ml of H2O was added and the resulting precipitate removed by filtration and washed with H2O. After extraction by EtOAc (5 x 3 ml), the combined filtrate was lyophilized to give 63 mg of crude 12, which was used without further purification. 1H NMR (DMSO-d6): delta 2.23 (1H, ddd, J = 3.3, 6.1, 13.1 Hz, 2'a), 2.80 (3H, m, 2'b and 5'), 3.82 (1H, m, 4'), 4.42 (1H, m, 3'), 5.70 (2H, br, NH2), 6.32 (1H, t, J = 6.6 Hz, 1'), 7.29 (2H, brs, NH2), 8.14 (1H, s, 2), 8.35 (1H, s, 8). Synthesis of 5'-amino-2',5'-dideoxycytidine (13) | A solution of 80 mg of 8 (0.31 mmol) and 245 mg Ph3P (0.37 mmol) in pyridine (2 ml) was stirred at room temperature for 7 h. Concentrated NH4OH (0.2 ml) was added and the mixture was stirred at room temperature for an additional 15 h. Water (5 ml) was added and the resulting precipitate was removed by filtration. The filtrate was extracted with EtOAc and the resulting aqueous layer was lyophilized to afford 104 mg of crude 13, which was used without further purification. 1H NMR(DMSO-d6): delta 1.95 (1H, td, J = 6.6, 13.4 Hz, 2'a), 2.09 (1H, ddd, J = 3.9, 5.7, 13.2 Hz, 2'b), 2.75 (2H, brs, 5'), 3.69 (1H, m, 4'), 4.15 (1H, m, 3'), 5.14 (3H, br, NH2 and OH), 5.73 (1H, d, J = 7.3 Hz, 5), 6.14 (1H, t, J = 6.7 Hz, 1'), 7.31 (2H, br, NH2), 7.72 (1H, d, J = 7.3 Hz, 6). Synthesis of 5'-amino-2',5'-dideoxyguanosine (14) | A mixture of 40 mg (0.14 mmol) of 9 and 111 mg (0.43 mmol) of Ph3P was added to 1 ml of pyridine and stirred at room temperature for 5 h. Concentrated aqueous NH4OH (0.3 ml) was added and the mixture was stirred for 17 h at room temperature followed by 1 h at 55C. Water (6 ml) was added to the mixture and the resulting precipitate was removed by filtration and washed with H2O. The combined filtrate was extracted with EtOAc. The aqueous solution was lyophilized to afford 50 mg of crude 14, which was used without further purification. 1H NMR (DMSO-d6): delta 2.19 (1H, m, 2'a), 2.57 (1H, td, J = 6.6, 13.4 Hz, 2'b), 2.82 (2H, brs, 5'), 3.79 (1H, brs, 4'), 4.37 (1H, brs, 3'), 5.76 (3H, brs, NH2 and OH), 6.11 (1H, td, J = 6.9 Hz, 1'), 6.90 (2H, brs, NH2 and OH), 7.88 (1H, s, 8). Synthesis of 5'-amino-2',5'-dideoxyuridine (15) | A solution of 5'-azido-2',5'-dideoxyuridine (10) (50 mg, 0.20 mmol) and Ph3P (160 mg, 0.61 mmol) in pyridine (1 ml) was stirred at room temperature for 6 h. Concentrated NH4OH (0.2 ml) was added and the mixture was stirred at room temperature for 19.5 h followed by 55C for 2 h. Water (5 ml) was added and the resulting precipitate was removed by filtration. The filtrate was washed with EtOAc and lyophilized to give 59 mg of crude 15, which was used without further purification. 1H NMR (DMSO-d6): delta 2.10 (2H, m, 2'), 2.72 (2H, m, 5'), 3.70 (1H, dd, J = 5.1, 8.5 Hz, 4'), 4.19 (1H, m, 3'), 5.62 (1H, d, J = 8.1 Hz, 5), 6.00 (3H, br, NH2 and OH), 6.13 (1H, t, J = 6.9 Hz, 1'), 7.82 (1H, d, J = 8.1 Hz, 6). Synthesis of 5'-amino-2',5'-dideoxyinosine (16) | Compound 16 was prepared from 5'-azido-2',5'-dideoxyinosine (11) by contract synthesis using hydrogenation catalyzed by Pd (GLSynthesis Inc., Shrewsbury, MA). Compound 11 was made using procedures similar to that described by Yamamoto et al. . Synthesis of 5'-amino-2',5'-dideoxynucleoside-5'-N-triphosphates 1 --6 | A mixture of 250 micromol trisodium trimetaphosphate (TMP) and 50 micromol 12, 13, 14, 15, 16 (described above, the amounts of 12 --15 were based on assuming 100% conversion from 5'-azido-nucleosides) or 17 (obtained from Sigma-Aldrich) was dissolved in 0.5 ml of 0.5 M aqueous Tris (measured pH similar 11). The resulting solution was allowed to incubate at room temperature for 5 --7 days. The progress of the reaction was monitored by reversed phase HPLC. The resulting solutions were used as 100 mM 1, 2, 3, 4, 5 or 6 for enzymatic reactions without purification. Reversed phase HPLC analysis of nucleotides 1 --6 | The reaction mixtures were analyzed on an HPLC instrument (Waters, Marlboro, MA) that consisted of a 515 HPLC pump, a 2700 sample manager, a 996 photodiode array detector and a temperature control module. A Waters Nova-pak C18 (3.9 x 150 mm, 4 microm) column was used with mobile phase buffer A (0.1 M TEAA, 10 mM Tris pH 9) and buffer B (25% MeOH, 75% buffer A) at a flow rate of 1.00 ml /min. The buffer gradient was varied from 98% A:2% B to 100% B to achieve good separation of each nucleoside from its triphosphate: 0 --8 min, 98% A to 70% A; 8 --9 min, 70% A to 0% A; 9 --10 min, 0% A; 10 --11 min, 0% A to 98% A; 11 --15 min, 98% A. The extent of conversion from each starting nucleoside to the corresponding product was determined to be 85.6% for 1, 78.9% for 2, 83.2% for 3, 80.0% for 4, 89.1% for 5 and 91.5% for 6 . The photodiode array detector also provided UV absorption spectra of each compound from which lambdamax for each compound was obtained: lambdamax(1) = 259 nm; lambdamax(2) = 271 nm; lambdamax(3) = 252 nm; lambdamax(4) = 263 nm; lambdamax(5) = 249 nm; lambdamax(6) = 267 nm. 31P NMR analysis of 1 --3, 5 and 6 | The above solutions containing nucleotide triphosphates 1 --3, 5 and 6 and an aqueous solution of a mixture of Tris and TMP were diluted with D2O and analyzed by 31P NMR. The most prominent peak was at --21 p.p.m. and was observed for all samples, corresponding to TMP. A doublet around --1 p.p.m. corresponding to a phosphoramidate (Palpha-N) resonance was observed for samples containing 1, 2, 3, 5 and 6, but not for the mixture of Tris and TMP. Due to the presence of a large excess of TMP and its breakdown products, the resonances corresponding to Pbeta and Pgamma were not assigned. NH2-dATP (1), --0.99 p.p.m. (d, J = 20.8 Hz); NH2-dCTP (2), --0.94 p.p.m. (d, J = 19.5 Hz); NH2-dGTP (3), --1.00 p.p.m. (d, J = 20.8 Hz); NH2-dITP (5), --1.03 p.p.m. (d, J = 20.8 Hz); NH2-dTTP (6), --0.97 p.p.m. (d, J = 20.1 Hz). Preparation of single-stranded DNA template | A biotinylated forward primer 5'-biotin-d(TCGGAGAAACTGGACAGCAC) was paired with reverse primer 5'-d(TTGAGATCCAGCCTCACGAGG) or 5'-d(AACAATGGGAAATTTAGTCTG) to produce biotin-tagged 87 and 480 bp amplification fragments by PCR, using a plasmid containing the human transferrin receptor gene as template (,). Each PCR product was immobilized to streptavidin --agarose beads in 5 mM Tris pH 7.5, 1 M NaCl, 0.5 mM EDTA, 0.05% Tween-20 overnight. After removal of the supernatant, the beads were washed with H2O, then treated twice with 0.3 ml of 0.2 M NaOH for 15 min. After washing with 150 microl of H2O, all three supernatants were combined and cooled on ice before being neutralized to pH 8 with 1 M HCl (similar120 microl). The solution was concentrated on a SpeedVac (Savant, Farmingdale, NY) to similar50 microl and purified using a Quick-Spin G25 (TE) column (Roche Biosciences, Oakland, CA). The collected solutions of non-biotinylated, single-stranded DNA templates (87 or 480 nt) were quantified and stored at --20C. The amount of the 480 nt fragment was estimated by UV analysis using the formula [OD260 /(480 x 10 000)] x 106 = concentration (microM) . The concentration of the 87 nt solution was estimated by capillary electrophoresis analysis by comparing peak sizes with an oligonucleotide of known concentration. Primer extension with complete analog substitution | The primer 5'-d(TCGGAGAAACTGGACAGCAC) was labeled at the 5' end using T4 polynucleotide kinase (New England Biolabs, Beverly, MA) and [gamma-32P]ATP (Perkin-Elmer, Boston, MA) and was purified using a Sephadex G50 column. The radiolabeled primer (similar1.6 pmol) was annealed to the single-stranded 87 nt DNA template (similar2.4 pmol) in 20 mM MgCl2 and 50 mM NaOAc. A typical extension reaction (10 microl) contained similar0.01 microM primer --template duplex, 45 mM Tris pH 9.5, 10 mM DTT, 20 mM MgCl2, 4 mM NH2-dNTP, 0.1 mM each of the other three dNTPs and 5 U Klenow (exo --) polymerase (New England Biolabs). After incubation at 37C for 1 h, 30 microl of TE (10 mM Tris, 1 mM EDTA, pH 8) was added and the mixture was purified using a Sephadex G50 column. Primer extension with partial (statistical) analog substitution | The radiolabeled primer 5'-[32P]dp(TCGGAGAAACTGGACAGCAC) (similar4 pmol) was annealed to the single-stranded 480 nt DNA template (4 pmol) in 20 mM MgCl2 and 50 mM NaOAc. A typical extension reaction (10 microl) contained similar0.02 microM primer --template duplex, 50 mM Tris pH 9.5, 5 mM DTT, 22 mM MgCl2, 5 mM NaOAc, 4 mM NH2-dNTP and 0.016 --0.046 mM of the corresponding dNTP (0.04 mM dATP, 0.026 mM dCTP, 0.016 mM dGTP, 0.046 mM TTP, 0.026 mM dITP or 0.02 mM dUTP), 0.4 mM each of the other three dNTPs and 5 U Klenow (exo --). After incubation at 37C for 1 h, each extension reaction mixture was supplemented with additional d(ACGT)TPs (1 microl of 1 mM) and incubated for another 15 min. The solutions were stored at --20C after dilution with TE pH 8. Acid cleavage of extension products | To each of the Sephadex purified 87 nt extension products (8 microl) was added 2 microl of 1% acetic acid (HOAc), followed by incubation at 40C for 30 min. Deionized water (100 microl) was added to each sample and the diluted solutions were dried in vacuo at room temperature. To each of the unpurified, TE diluted 480 nt extension mixtures (2 microl) was added 1 microl of 10 mM EDTA, 1 microl of 10% acetic acid and 6 microl of dH2O. The resulting mixtures were incubated at 37C for 10 min then diluted with 100 microl of dH2O. The resulting mixtures were dried in a SpeedVac evaporator. Restriction digestion of M13mp18 DNA duplex containing P-N linkers | A 32P-labeled M13 universal primer (5'-GTAAAACGACGGCCAGT, 5 pmol) was annealed to 1 microg single-stranded M13mp18 plasmid (New England Biolabs) in TE, and similar13% of the resulting duplex (2 of 15 microl) was used for each primer extension reaction. The extension reaction was carried out at 37C for 1 h in 50 mM Tris pH 9, 5 mM DTT, 10 mM MgCl2, 0.4 mM d(ACG)TPs, 4 mM NH2-dTTP and 0.08 mM dTTP with 5 U Klenow (exo --). The product was purified using a G50 column and the eluate was concentrated to a final volume of 20 microl. For the cleavage reaction, 4 microl of the duplex was treated with 4% HOAc at 37C for 10 min and analyzed by 5% PAGE. For restriction digestion, 2 microl of the duplex was digested with 3 U of MscI (New England Biolabs) in 50 mM KOAc, 20 mM Tris --OAc pH 7.9, 10 mM Mg(OAc)2 and 1 mM DTT at 37C for 1 h. The product was ethanol precipitated, redissolved in formamide loading buffer and analyzed on a 0.6% TBE --agarose gel. The gel was dried and subjected to autoradiography. Figure 1 | HPLC chromatograms of reaction mixtures containing NH2-dATP, NH2-dCTP, NH2-dGTP, NH2-dUTP, NH2-dITP and NH2-dTTP. HPLC chromatograms of reaction mixtures containing NH2-dATP, NH2-dCTP, NH2-dGTP, NH2-dUTP, NH2-dITP and NH2-dTTP. The conversion yield was determined by comparing the areas of the late eluting peak (nucleotide 5'-N-triphosphate) versus the early eluting peak (nucleoside) in each chromatogram. RESULTS AND DISCUSSION : Syntheses of 5'-amino-2',5'-dideoxynucleosides | Although there are several reports on the preparation of NH2-dU (15) and NH2-dT (17) (, --), the isolation and characterization of other 5'-amino analogs (12 --14 and 16) have not been reported. The Staudinger reaction, which converts azides into primary amines through Ph3P treatment followed by hydrolysis of the iminophosphinimine intermediates, had previously been used to prepare NH2-dT (17) from N3-dT . Using a similar strategy (Scheme ), we synthesized 5'-amino-2',5'-dideoxynucleosides 12 --16 (NH2-dA, NH2-dC, NH2-dG, NH2-dU and NH2-dI) from the corresponding azides 7 --11 (N3-dA, N3-dC, N3-dG, N3-dU and N3-dI). This conversion was virtually quantitative for each nucleoside analog that we examined. Compounds N3-dU (10) and N3-dI (11) were synthesized according to previously described procedures (,), whereas N3-dA (7), N3-dC (8) and N3-dG (9) were prepared from derivatives containing protected exocyclic amino groups followed by an intermediate step of NH4OH aminolysis. We first attempted to carry out the Staudinger reaction directly on the protected form of 5'-azido-N2-isobutyryl-2',5'-dideoxyguanosine, hoping that the NH4OH treatment would also remove the isobutyryl group to yield NH2-dG (14). However, the apparently redundant conversion to N3-dG (9) prior to the Staudinger reaction turned out to be essential. Direct treatment of 5'-azido-N2-isobutyryl-2',5'-dideoxyguanosine with Ph3P in pyridine followed by hydrolysis /deprotection with NH4OH yielded a mixture of products that could not be fully converted to 9, even after prolonged heating in concentrated ammonia. We suspect that the major side-product comes from transamidation of the isobutyryl group; the resulting 2',5'-dideoxyguanosine-5'-isobutyrylamide is much less reactive to NH4OH than the starting material. Preparation of 5'-amino-2',5'-dideoxynucleoside 5'-N-triphosphates | In early reports, the synthesis of NH2-dTTP was accomplished by reacting NH2-dT (17) with TMP (18), with moderate conversion yield . Attempted purification by HPLC fractionation failed to yield pure NH2-dTTP because the isolated NH2-dTTP appeared to slowly degrade until the ratio NH2-dTTP:NH2-dT reached 40:60. In addition, NH2-dTTP was reported to be labile to both acidic and alkaline conditions, yielding either NH2-dT or its 5'-N-diphosphate (NH2-dTDP), respectively, although the identity of the latter was supported only by circumstantial evidence . To our knowledge, there have been no prior reports on the synthesis and characterization of any other NH2-dNTPs. We set out to apply and, we hoped, improve upon this elegant approach in the synthesis of other 5'-N-triphosphates from corresponding 5'-amino-nucleosides. In our attempts to improve the TMP reaction, we reasoned that a more careful regulation of the pH of the reaction mixture might result in higher yields of the desired 5'-N-triphosphates. To accomplish this we included tris(hydroxymethyl)aminomethane (Tris) in the reaction mixture. Tris is a commonly used buffering agent with a pKa of 8.1 and, because it is a sterically hindered primary amine, it should not critically interfere with the desired reaction. An added benefit is that Tris is frequently used as a buffer for polymerization reactions, which should allow the direct use of the TMP reaction mixtures without purification. Six nucleotide analogs (1 --6) were successfully prepared from the corresponding nucleosides (12 --17) using this strategy (Scheme ). Typically, five molar equivalents of TMP (18) and 0.5 M Tris (pH similar 11) were used in each reaction, which was monitored by HPLC until the ratio of NH2-dNTP to NH2-dN reached a plateau. The HPLC chromatograms showed that each NH2-dNTP is well separated from its nucleoside and 80 --90% conversion was achieved in every case. When stored at --20C these products are stable for at least several months. Attempts to purify the 5'-N-triphosphates by HPLC consistently yielded mixtures of NH2-dN and NH2-dNTP, even when alkaline HPLC buffers were used. This observation is in agreement with previous reports on the stability of NH2-dTTP and may imply greater proportions of NH2-dNTP in each reaction mixture, since some degradation might have occurred during the HPLC analysis. Additional characterization of each product by 31P NMR spectroscopy showed the appearance of a single phosphoramidate resonance, which is indicative of the formation of a P-N bond. The absence of additional P-N signals in the spectra confirms the purity of each sample as seen by HPLC analysis. Complete substitution and site-specific cleavage | Based upon the previously reported ability of DNA polymerase I to incorporate NH2-dTTP, we chose to examine the ability of the exonuclease-deficient Klenow fragment of DNA polymerase I to incorporate our newly generated 5'-N-triphosphates. Each of the analogs NH2-dATP (1), NH2-dCTP (2), NH2-dGTP (3) and NH2-dTTP (6) was readily incorporated in polymerization reactions in place of dATP, dCTP, dGTP and dTTP, respectively (Fig. A, lanes 1, 3, 5 and 7). The best extension was achieved when the pH of the reaction mixture was kept at >9, presumably because both the NH2-dNTPs and the DNA containing P-N linkers are more stable under basic conditions . The acid lability of the phosphoramidate linkages was demonstrated by PAGE of acetic acid-treated extension products containing NH2-dA, NH2-dC, NH2-dG or NH2-dT (Fig. A, lanes 2, 4, 6 and 8), in agreement with previous reports on NH2-dT-containing oligonucleotides (,). The site-specific cleavage occurred at every modified base, yielding distinct oligonucleotide bands that represent sequence ladders for each base. We should note that following the completion of this work, a newly reported study demonstrated the ability of Sequenase v.2.0 to incorporate NH2-dTTP and NH2-dCTP (nucleotide synthesis was not described) . In our hands, Sequenase v.2 was not as efficient at incorporating these analogs as Klenow (exo --). Surprisingly, the first cleavage band observed above the primer is {primer+AGp}, not the expected {primer+Ap} (Fig. A, lanes 6 and 2). A closer inspection of lane 2 revealed an extra band just below the 20 nt primer, which is likely to be the 'missing' {primer+Ap} fragment. This fragment runs aberrantly similar0.5 nt below the primer, presumably because the additional 3'-phosphate group alters its charge versus mass ratio. This band is also visible in lane 1 of Figure B, where it resides between the 20 nt primer and a 19 nt band (a side-product of the synthetic 20 nt primer). Thus, in Figure A the sequence ladders are represented by oligonucleotides migrating similar1.5 nt faster than the expected length due to the extra 3'-phosphate. Fragments up to 81 nt long (cleaved at the 82nd base, NH2-dT) were observed, as well as the full-length 87 nt extension product. After prolonged heating at 40C with a higher concentration (5%) of acetic acid, the full-length products were completely degraded and the shorter fragments enriched (Fig. B, lanes 1 --4), indicating more extensive, but still incomplete, cleavage at each modified nucleotide. Whereas the relative extent of cleavage can be roughly estimated as NH2-dC > NH2-dT > NH2-dG ~ NH2-dT, some non-specific cleavage bands were observed and are highlighted with gray boxes in Figure B (lanes 1 and 4). The mobility of these extra bands is less than one base slower than the specific cleavage bands, a mobility difference that was mentioned above as arising from loss or gain of a phosphate group. We speculate that the non-specific cleavage products come from a hydrolysis reaction catalyzed by the harsher acidic conditions, resulting in the formation of oligonucleotides without the 3'-phosphate group (Scheme ). Whatever the cause, the results in Figure B suggest that complete and specific cleavage at P-N linkages in modified DNA is not easily achieved and that the most fruitful applications of phosphoramidate chemistry may be those that require only partial cleavage, such as DNA sequencing. Partial substitution and site-specific cleavage: application in sequencing | Partial substitution of a designated base by its 5'-amino analog may be achieved by using a mixture of NH2-dNTP and dNTP in a polymerase reaction, resulting in a statistical representation of the analog at each matching site in a DNA molecule. This statistical distribution permits the generation of a sequence ladder through site-specific cleavage at the modified nucleotides. The use of partial incorporation of each of the NH2-dNTPs (1 --6) allowed the full-length replication of a 480 nt DNA template . The addition of extra dNTPs at the end of the primer extension reactions was found to minimize incomplete extension products . Direct treatment (without purification) of the polymerization mixtures with acetic acid resulted in sequencing ladders of A (lane 2, from NH2-dA), C (lane 3, from NH2-dC), G (lanes 4 and 6, from NH2-dG and NH2-dI) and T (lanes 5 and 7, from NH2-dT and NH2-dU). The distribution of the band intensities is reasonably uniform and can be adjusted by utilizing different ratios of native and modified nucleotides (Supplementary Material, Fig. A) and /or different cleavage conditions (see Fig. ). Under the conditions shown, NH2-dA, NH2-dC, NH2-dI and NH2-dU provided readable band intensities up to full length. There are several bands in lane 1 (dNTPs) that are less than full length, which indicate stops or positions where the polymerase had difficulty proceeding. These artifacts are common in DNA sequencing reactions and can often impede sequence readout. Fortunately, with the incorporation --cleavage sequencing scheme that is outlined here, such artifacts can be eliminated by simple purification of the full-length extension product prior to the cleavage reaction. To evaluate any length limits of this incorporation --cleavage sequencing method, a 7.2 kb single-stranded M13mp18 plasmid was used as a primer extension template. A 5' radiolabeled primer was annealed to the plasmid and extended in the presence of NH2-dTTP:dTTP 100:1 and d(ACG)TPs using Klenow (exo --) (Fig. A). Upon treatment with dilute acetic acid, the extension product was cleaved to form DNA fragments well dispersed throughout the range similar20 bp to 7.2 kb, indicating successful incorporation of NH2-dT (see Supplementary Material, Fig. B). The successful primer extension to 7.2 kb and the susceptibility of the extension product to restriction endonucleases is demonstrated in Figure . After treatment with the restriction enzyme MscI, which recognizes and cleaves at TGG /CCA and has a unique site within the 7.2 kb M13 sequence, the circular DNA (present as a mixture of supercoiled and non-supercoiled) was linearized and migrated as a 7.2 kb duplex (Fig. B, lanes 1 and 2). Heating of non-digested extension product prior to gel analysis resulted in two major bands near 8 kb, respectively corresponding to a single-stranded 7.2 kb fragment (upper band) and the duplex, probably as a result of incomplete heat denaturation. Heat denaturation of MscI-digested material liberated a single-stranded 1.2 kb fragment, which was the only product observed (Fig. A, and Fig. C, lane 2), as expected. These results indicate that primer extension by statistical incorporation of a NH2-dNTP is very efficient and the length limit for such polymerization may be well above 7.2 kb. Furthermore, despite the presence of P-N linkers in these polynucleotide products, they can be readily digested by restriction enzymes. Based on these observations, we envisioned a long-range sequencing approach (Scheme ) that may be particularly useful for genome sequence assembly, a challenge faced by genome sequencing projects (,). Because the human genome is highly repetitive with >50% of repeated sequences , including multiple Alu repeats (consensus length 280 nt) and LINES (long interspersed DNA sequence elements; up to 7000 nt in length) , whereas the commonly used dideoxy sequencing method has an average read length of <600 nt , it is very difficult to assemble the sequence information thus obtained. Long-range primer extension reactions could be used to generate large duplex DNA containing statistically incorporated NH2-dNTPs, which could be digested with restriction endonucleases into shorter fragments (Scheme ). These fragments could then be labeled with a fluorescent dye or radiolabel, then fractionated by HPLC. Each isolated fragment could be cleaved and sequenced. Digestion with different restriction endonucleases will produce complementary sets of sequencing data, which would facilitate assembly of the template sequence. Scheme 1. (a) (i) Ph3P, pyridine; (ii) NH4OH. Scheme 1. (a) (i) Ph3P, pyridine; (ii) NH4OH. (b) TMP (18), Tris, H2O. Figure 2 | Denaturing PAGE (12%) analysis of polymerization products with dATP, dCTP, dGTP or dTTP completely substituted by the corresponding NH2-dNTPs. Denaturing PAGE (12%) analysis of polymerization products with dATP, dCTP, dGTP or dTTP completely substituted by the corresponding NH2-dNTPs. The sequence of the full-length products are listed with modified bases shown in coded colors: A, red; C, green; G, yellow; T, blue. (A) Primer extension reaction mixtures (E) and corresponding cleavage products (C) elicited by treatment with 0.2% acetic acid at 40C for 30 min. Specific cleavage bands at each NH2-dN are highlighted with dots in coded colors. (B) Cleavage products generated by heating at 40C with 5% acetic acid for 2 h. Non-specific cleavage bands are highlighted with gray boxes. Scheme 2. Scheme 2. Figure 3 | Sequence analysis utilizing statistical incorporation of NH2-dNTPs. Sequence analysis utilizing statistical incorporation of NH2-dNTPs. The cleavage fragments from 480 nt DNA were separated by denaturing PAGE (6%). Figure 4 | (A) Schematic presentation of the primer extension, restriction digestion and heat denaturation procedures and corresponding products. (A) Schematic presentation of the primer extension, restriction digestion and heat denaturation procedures and corresponding products. (B) TBE --agarose gel (0.6%) analysis of the primer extension product containing statistically incorporated NH2-dT:dT using 7249 nt M13 plasmid as template and the corresponding MScI restriction digestion product. (C) Heat-denatured samples of (B) analyzed by 0.6% TBE --agarose gel electrophoresis. Scheme 3. Proposed long-range DNA sequencing method using polymerization, restriction digestion, fragment labeling, chemical cleavage, electrophoresis and sequence assembly. Scheme 3. Proposed long-range DNA sequencing method using polymerization, restriction digestion, fragment labeling, chemical cleavage, electrophoresis and sequence assembly. CONCLUSIONS : We have developed reproducible synthetic procedures to prepare six 5'-amino-2',5'-dideoxynucleotide analogs in high yield and purity, and have characterized each by UV, HPLC, 31P NMR and polymerase incorporation. The availability of these nucleotides will facilitate the development of new technologies that utilize them (; Wolfe,J.L., Wang,B.H., Kawate,T., DeMaria,C. and Stanton,V.P., submitted for publication) and may inspire interests in additional applications. We also demonstrated that 5'-NH2-dNTPs can be efficiently incorporated by Klenow (exo --) polymerase to generate very long polynucleotides containing a modified nucleotide, and that these products can undergo site-specific cleavage at the phosphoramidate bonds under mild acidic conditions to generate sequence ladders. This incorporation -- cleavage-based sequencing approach may provide an experimental means to improve genome sequence assembly. SUPPLEMENTARY MATERIAL : Supplementary Material is available at NAR Online. Backmatter: PMID- 12202757 TI - Polyamine structural effects on the induction and stabilization of liquid crystalline DNA: potential applications to DNA packaging, gene therapy and polyamine therapeutics AB - DNA undergoes condensation, conformational transitions, aggregation and resolubilization in the presence of polyamines, positively charged organic molecules present in all cells. Under carefully controlled environmental conditions, DNA can also transform to a liquid crystalline state in vitro. We undertook the present work to examine the ability of spermidine, N4-methylspermidine, spermine, N1-acetylspermine and a group of tetramine, pentamine and hexamine analogs of spermine to induce and stabilize liquid crystalline DNA. Liquid crystalline textures were identified under a polarizing microscope. In the absence of polyamines, calf thymus DNA assumed a diffused, planar cholesteric phase with entrapped bubbles when incubated on a glass slide at 37C. In the presence of spermidine and spermine, the characteristic fingerprint textures of the cholesteric phase, adopting a hexagonal order, were obtained. The helical pitch was 2.5 microm. The final structures were dendrimeric and crystalline when DNA was treated with spermine homologs and bis(ethyl) derivatives. A cholesteric structure was observed when DNA was treated with a hexamine at 37C. This structure changed to a hexagonal dendrimer with fluidity on prolonged incubation. These data show a structural specificity effect of polyamines on liquid crystalline phase transitions of DNA and suggest a possible physiological function of natural polyamines. Keywords: INTRODUCTION : The natural polyamines, putrescine [H2N(CH2)4NH2], spermidine [H2N(CH2)3NH(CH2)4NH2] and spermine [H2N(CH2)3 NH(CH2)4NH(CH2)3NH2], are ubiquitous cellular components that are involved in a variety of cellular functions . Under physiologic ionic and pH conditions, the polyamines are positively charged and hence negatively charged macromolecules, including DNA and RNA, are their prime targets of interaction (,). The binding of polyamines to DNA results in duplex and triplex DNA stabilization and condensation of dilute solutions to toroids and spheroids, as well as the aggregation and resolubilization of DNA ( --). Toroidal condensates are a highly organized form of DNA and a recent study indicates the organization of DNA in a columnar hexagonal array in toroids ( --). A columnar hexagonal organization has been reported in liquid crystalline DNA, which is generally studied using a polarizing microscope ( --). In dilute solutions (<1 mg /ml), DNA exists in random coils or is randomly oriented and the solution is a classical isotropic liquid. Under polarized light, the DNA solution becomes totally dark. As the DNA is concentrated (>1 mg /ml), the molecules spontaneously undergo unidirectional ordering (the solution starts to become birefringent under polarized light) and transform into liquid crystals of the 'cholesteric' type, which transforms into the 'columnar hexagonal' phase at higher concentrations ( --). Depending upon the concentration of DNA in solution, the condensates assume different degrees of order and packing ( --). The textures of a liquid crystalline phase arise due to the packing or alignment of partially organized units of molecules in space. These alignments are dictated by the nature of the molecules that provoke the liquid crystalline organization as well as the surface properties of the material used as a base to study the liquid crystalline textures. The supercoiled DNA spontaneously organizes into the liquid crystalline phase to minimize the macromolecular excluded volume . This concentration-dependent spontaneous liquid crystal formation is similar to that exhibited by non-electrolyte macromolecules . The highly charged anionic polyelectrolyte nature of DNA, with a persistence length of 50 nm, might affect its liquid crystalline properties due to the counterion layer which determines the effective axial ratio and the excluded volume . Counterion neutralization is essentially required for the induction and stabilization of liquid crystalline DNA. The liquid crystalline transformation of DNA was previously observed with fragmented DNA of similar150 bp length, at concentrations >200 mg /ml ( --). A basic requirement for the exhibition of the liquid crystalline phase in DNA is a critical local concentration (Ci) . Merchant and Rill studied the chain length dependence of Ci and found a dramatic reduction in Ci as the size of the DNA increased. For example, the Ci values for 147 and 8000 bp DNA samples were 135 and 13 mg /ml, respectively. DNA condensation by multivalent ions, including the natural polyamines, results in a significant increase in the local concentration of DNA because this process is achieved by DNA --DNA interactions (,,). In a series of experiments, Livolant and colleagues demonstrated that spermidine and spermine are capable of provoking multiple liquid crystalline forms of fragmented DNA (,,, --). The liquid crystalline organization of DNA in the presence of these endogenous molecules is important to understand the nature and organization of DNA in the cell (,). The nuclear concentration of DNA is very high and of the order of 200 --400 mg /ml . The cellular DNA is in a macromolecular crowded environment, surrounded by proteins and cationic molecules, including the polyamines . The cellular concentration of polyamines is estimated to be in the millimolar range, although the precise distribution of these molecules in the cytoplasm and nucleus is not correctly known (,). Since polyamines associate with DNA by electrostatic interactions, a possible function of polyamines in the cell might involve the organization of DNA, including liquid crystalline DNA. Synthetic polyamines, such as bis(ethyl) derivatives of spermine and its analogs, and oligoamines are under development as chemotherapeutic agents for different forms of cancer (,,). It is not yet clear how the synthetic polyamines interact with DNA and modulate cellular functions and specific gene expression. Therefore, contrasting the effects of natural and synthetic polyamines on liquid crystalline DNA formation might shed new light on the mechanism of action of polyamine-based therapeutic agents. Polyamines and polyamine derivatives, such as polyaminolipids, are under development as DNA delivery vehicles for gene therapy . It has been generally accepted that the first step in the mechanism of action of these agents is the condensation of DNA to nanoparticles that are transported through the cell membrane by mechanisms that are not yet clear ( --). The existence of the liquid crystalline phase of DNA has been shown in complexes of cationic lipids and DNA . However, detailed information on the liquid crystalline textures adopted by DNA under the conditions of gene transfection is lacking at present. Due to the important role played by natural polyamines in DNA packaging in the virus head and chromatin (,) and the emerging use of polyamine analogs and derivatives as gene delivery vehicles and potential drug candidates for chemotherapy, we undertook a detailed investigation of the liquid crystalline phases of DNA in the presence of natural and synthetic polyamines. The liquid crystalline structures formed in the presence of spermidine and spermine were planar cholesteric or hexagonal. In contrast, synthetic polyamines had a tendency to crystallize the DNA, although highly ordered liquid crystalline dendrimeric structures were also formed in the presence of these polycations. MATERIALS AND METHODS : Polyamines and chemicals | Spermidine.3HCl and spermine.4HCl were purchased from Sigma Chemical Co. (St Louis, MO). N4-methylspermidine, 1,11-diamino-4,8-diazaundecane (norspermine or 3-3-3), N1,N11-bis(ethyl)norspermine (BE-3-3-3), N1,N12-bis(ethyl)spermine (BE-3-4-3), 1,10-diamino-4,7-diazadecane (3-2-3), 1,13-diamino-4,10-diazatridecane (3-5-3), 1,14-diamino-4,11-diazatetradecane (3-6-3), 1,15-diamino-4,12-diaza- pentadecane (3-7-3), 1,16-diamino-4,13-diazahexadecane (3-8-3), 1,17-diamino-4,14- diazaheptadecane (3-9-3), 1,15-diamino-4,8,12-triazapentadecane (3-3-3-3), 1,15-bis(ethylamino)-4,8,12-triazapentadecane (BE-3-3-3-3), and 1,21- diamino-4,9,13,18-tetraazahenicosane (3-4-3-4-3) were synthesized by us by procedures described earlier (,). Calf thymus DNA | Calf thymus DNA was purchased from Worthington Biochemical (Freehold, NJ) and dissolved in 10 mM Na cacodylate buffer (10 mM Na cacodylate, pH 7.4, and 0.5 mM EDTA) at a concentration of 25 mM DNA phosphate (8.28 mg /ml). The weight average molecular weight of the DNA was 6 x 106, as determined by multiangle laser light scattering and Zimm plot. The second virial coefficient was 6 +- 1 x 10 --4 mol ml /g2. It had a root mean square radius of 238 +- 3 nm. The observed A260 /A280 ratio of the DNA solution was 1.88, indicating that the DNA was free of protein contamination. The DNA sample was dialyzed extensively against the Na cacodylate buffer. The concentration of calf thymus DNA was determined by measuring the absorbance at 260 nm and using the molar extinction coefficient (epsilon) of 6900 per M /cm. The DNA concentration of 25 mM was selected for the present series of experiments because liquid crystalline phase transitions could be observed with this concentration of high molecular weight DNA in the presence of polyamines. DNA and polyamine solutions were stored at 4C. There was no effect of the time of storing DNA /polyamine solutions before incubation on the nature of liquid crystalline structures adopted by the DNA. The solutions were homogeneous at the start of our experiments. Polarizing microscopy | DNA was precipitated either directly on the glass slide or in an Eppendorf tube and centrifuged to sediment the precipitate for observation . The glass slides were soaked in chromic acid, cleaned with distilled water, rinsed with ethanol and dried for sample preparation. For some experiments, the dried slides were rubbed unidirectionally with a fine cotton cloth to study the effect of grooves on the liquid crystalline phases formed in the presence of polyamine analogs. The DNA precipitate was spread over the glass slides with a coverslip and sealed with a neutral solution of polystyrene and plasticizers in toluene to prevent dehydration of the sample (,). A total of 5 microl final volume was handled when precipitation was performed on glass microscope slides. The preparations were observed under polarizing light or phase contrast light in a Nikon TE-DH100W microscope. In some cases, a lambda plate was inserted between crossed polars to analyze the orientation of the DNA molecules in particular domains. The microscope stage was rotated in a clockwise manner to observe the uniaxial /biaxial nature of the phases and the sign of optical rotation. All the preparations were optically negative (i.e. when the stage was rotated in a clockwise manner, the texture shift or extinction of disinclination lines occurred in a counter-clockwise direction) and showed negative birefringence (textures in black and white instead of a colored pattern). After observing the initial phase appearance at 22C, the preparations were incubated at 37C for extended time periods to observe the phase changes until crystallization or complete darkening (isotropization) occurred. The preparations were examined periodically for phase changes and photographs were taken when the phases became prominent and distinct. The phases and granular boundaries were clear and sharp when the sample was incubated at 37C. A triplicate of each sample was made to ensure the reproducibility of the phase changes. The results were reproducible in three sets of separate experiments. RESULTS : Effects of spermidine and N4-methylspermidine on the structure of calf thymus DNA | Figure shows the phase transitions of calf thymus DNA (25 mM, 8.3 mg /ml) in the presence of spermidine and its derivative N4-methylspermidine. No well-defined structure of the DNA appeared after 2 h incubation at 22C. However, a diffuse planar cholesteric phase, with isotropic bubbles, appeared on incubating the DNA at 37C for 2 h (Fig. A). This phase could flow spontaneously; however, no characteristic fingerprint texture of the cholesteric phase appeared even after incubating this phase for 48 h at 37C. We next examined the effect of spermidine on the precipitation and phase transitions of calf thymus DNA. DNA solution (25 mM) was mixed with 50 mM spermidine on a glass plate and covered with a coverslip. The preparation was sealed completely with polystyrene resin and left for similar15 min at 22C. Polarizing light microscopic observation showed an oily streak texture (not shown) characteristic of the cholesteric phase and the phase flowed spontaneously. The sample was incubated at 37C and the phase changes were monitored at different time points with a polarizing microscope with crossed polars. After 3 h, the oily streak texture coalesced to form a large pitch cholesteric phase (not shown) which later (similar5 h) developed the fingerprint texture (Fig. B) of the cholesteric phase with selective reflection of light in the blue region. The area of the fingerprint texture (Fig. B, arrow) was blue in color and the helical pitch was 2.5 microm. The sample darkened after similar10 h and a flower-shaped columnar hexagonal phase (Fig. C) developed, the core arms of which were isotropic. This could be due to homeotropic (column alignment perpendicular to the glass surface) alignment of the columns. A similar preparation of DNA with N4-methylspermidine (50 mM) showed an oily streak cholesteric texture at 22C (not shown). After incubation for 12 h at 37C, fingerprint textures (pitch length 2.5 microm) developed. This phase transformed to an ordered hexagonal phase, showing bundles of rod-like textures (Fig. D), which remained stable for several days. Effect of spermine and N1-acetylspermine on calf thymus DNA | We next examined the effects of spermine on the liquid crystalline phase transitions of DNA . Mixing of the DNA with 1 mM spermine and incubation at 22C for 15 min produced a planar cholesteric phase with a 3-dimensional network (Fig. A). With 12 h incubation at 37C, a fingerprint texture developed with antiparallel grain boundaries when viewed through the lambda plate under crossed polars (Fig. B). The antiparallel arrangement and the fingerprint texture within the grains might have originated from the cholesteric domains adopting a hexagonal order. Incubation of the sample for 24 h at 37C showed a large pitch cholesteric phase (Fig. C) which darkened at similar48 h, without crystallization. In contrast, DNA treated with 1 mM N1-acetylspermine showed a Schlieren nematic phase after 12 h incubation at 37C (not shown), and this phase transformed to a crystalline phase at 48 h (Fig. D). Thus, the structural organization of calf thymus DNA is different in the presence of spermine and its acetylated derivative. Effect of spermine homologs on the phase transitions of DNA | In the next series of experiments, we examined the effects of several structural homologs of spermine, with the general structure H2N(CH2)3NH(CH2)n=2 --9NH(CH2)3NH2. The structural homologs and analogs of spermine are designated with a number system, indicating the number of methylene groups between the primary /secondary amino groups of spermine. The initial texture observed with calf thymus DNA treated with 3-3-3 for 15 min at 22C was lamellar (myelinic) (Fig. A). After 6 h incubation at 37C, this phase changed to a hexagonal phase with oblique tetragonal symmetry (not shown), which then changed to a flower-shaped homeotropic dendritic texture of hexagonal phase (Fig. B). Incubation of the DNA --polyamine complex for 24 h produced a crystalline phase with shell-like steps and colored arches. An arm of hexagonal ordered phase was also observed (Fig. C). A crystalline phase of DNA was also seen in calf thymus DNA complexed with 3-2-3 after 24 h incubation at 37C (Fig. D). We next examined the effects of higher homologs of spermine on the phase transitions of calf thymus DNA. The initial phase obtained with 3-5-3 and 3-6-3 was large pitch cholesteric (not shown). With 3-5-3, this phase changed to an oblique hexagonal crystalline phase after 24 h incubation at 37C (Fig. A). However, a myelin-like isotropic texture with layered appearance occurred with 3-6-3 after 12 h incubation at 37C. This phase developed a striped appearance (Fig. B) and eventually transformed to a crystalline phase (Fig. C). Among the higher homologs of spermine studied by us, only 3-7-3 (not shown) and 3-9-3 (Fig. D) produced fingerprint textures in calf thymus DNA after incubation for 12 h. All other compounds initially produced a cholesteric phase, which eventually crystallized. Effect of bis(ethyl) substitution of tetravalent polyamines on the phase transitions of DNA | We also examined the effects of two bis(ethyl)spermine analogs on the liquid crystalline phase transitions of DNA. These compounds are gaining considerable attention as chemotherapeutic agents for different forms of cancer (,,). A myelinic cholesteric phase was observed on mixing calf thymus DNA with 1 mM bis(ethyl)spermine (BE-3-4-3) (Fig. A), which transformed to a Schlieren texture (Fig. B) on incubation at 37C for 12 h. A crystalline phase slowly formed from this hexagonal phase and crystallization was complete by similar24 h (Fig. C). A similar pattern of phase changes was evident in calf thymus DNA complexed with bis(ethyl)norspermine (BE-3-3-3); however, the crystal growth showed a stepped lamellar phase and birefringent areas (Fig. D). Effects of a pentamine and its bis(ethyl)-substituted derivative on phase transitions of DNA | In order to test the effect of increasing the number of positive charges on DNA phase transitions, we next examined the effects of two pentamines (3-3-3-3 and BE-3-3-3) on calf thymus DNA. For these experiments, the polyamine concentration was 100 microM. Addition of 3-3-3-3 to calf thymus DNA produced a planar network cholesteric phase (Fig. A). This phase had limited fluidity when compared to the cholesteric phase formed in the presence of triamines and tetramines. On incubating the sample at 37C, the network texture became highly birefringent and a neuron-like dendrite developed after 12 h incubation at 37C due to hexagonal ordering (Fig. B). A crystalline phase developed after 36 h incubation, with both hexagonal dendrites as well as a crystalline phase (not shown). However, a highly ordered columnar hexagonal phase developed on a rubbed glass slide after 12 h incubation at 37C (Fig. C). In the case of BE-3-3-3 --3, a highly birefringent network-like texture developed from the cholesteric phase, which transformed to a neural network-like dendrite (not shown), which later crystallized (Fig. D). Effect of a hexamine on DNA phase transitions | With 3-4-3-4-3, a myelinic cholesteric phase initially appeared (not shown), which transformed to a fingerprint cholesteric texture after 12 h incubation at 37C (Fig. A). The fingerprint texture changed to an oblique hexagonal phase, which transformed to a discotic hexagonal ordered phase at 48 h (Fig. B), and this phase was quite stable for 1 week. Figure 1 | Effects of spermidine and N4-methylspermidine on the liquid crystalline transitions of calf thymus DNA. Effects of spermidine and N4-methylspermidine on the liquid crystalline transitions of calf thymus DNA. (A) Control. Calf thymus DNA solution (25 mM in a buffer containing 10 mM Na cacodylate, pH 7.4, and 0.5 mM EDTA) was incubated on a glass slide at 37C for 2 h (100x). (B) DNA (25 mM) was treated with 50 mM spermidine and incubated on a glass slide for 12 h at 37C (200x). Fingerprint texture, characteristic of the cholesteric phase, is indicated by the arrow. (C) DNA was treated with 50 mM spermidine and incubated on a glass slide for 24 h at 37C (180x). The flower-like texture is a highly ordered liquid crystalline form of DNA. (D) DNA treated with 50 mM N4-methylspermidine and incubated on a glass slide for 24 h at 37C (400x). Crystalline form of DNA is seen. Figure 2 | Effects of spermine and N1-acetylspermine on the liquid crystalline phase transitions of calf thymus DNA. Effects of spermine and N1-acetylspermine on the liquid crystalline phase transitions of calf thymus DNA. (A) DNA (25 mM in Na cacodylate buffer) was mixed with 1 mM spermine and incubated on a glass slide at 22C for 15 min (100x). A planar cholesteric phase with a 3-dimensional network is observed. (B) The glass slide in (A) was incubated for 12 h at 37C and viewed through the lambda plate under crossed polars (200x). Fingerprint texture with antiparallel grain boundaries is found. (C) Large pitch cholesteric phase was observed when the glass slide was further incubated for 24 h at 37C (180x). (D) DNA was mixed with 1 mM N1-acetylspermine and incubated for 48 h at 37C (45x). A crystalline phase is obtained. Figure 3 | Effects of two lower homologs of spermine on the liquid crystalline phase transitions of calf thymus DNA. Effects of two lower homologs of spermine on the liquid crystalline phase transitions of calf thymus DNA. (A) DNA (25 mM in Na cacodylate buffer) was mixed with 1 mM 3-3-3 and incubated for 15 min at 22C on a glass slide (100x). A lamellar (myelinic) phase is obtained. (B) Flower-shaped homeotropic dendritic texture of hexagonal phase is observed on incubating the DNA in (A) for 12 h at 37C (100x). (C) Crystalline phase with shell-like steps and colored arches was found on further incubating the DNA on a glass slide for 24 h at 37C (180x). (D) Crystalline phase of DNA was seen in calf thymus DNA complexed with 3-2-3 after 24 h incubation at 37C (45x). Figure 4 | Effects of higher homologs of spermine on the liquid crystalline phase transitions of calf thymus DNA. Effects of higher homologs of spermine on the liquid crystalline phase transitions of calf thymus DNA. (A) DNA (25 mM in Na cacodylate buffer) was incubated with 1 mM 3-5-3 for 24 h at 37C. A crystalline phase is observed (100x). (B) A myelin-like growth, which developed a striped appearance, is found after incubation of DNA with 1 mM 3-6-3 for 12 h at 37C (360x). (C) A crystalline phase was observed on further incubating the DNA in (B) for 24 h at 37C (90x). (D) Fingerprint textures are obtained on incubating DNA with 1 mM 3-9-3 for 12 h at 37C (200x). Figure 5 | Effects of bis(ethyl)spermine analogs on the liquid crystalline phase transitions of calf thymus DNA. Effects of bis(ethyl)spermine analogs on the liquid crystalline phase transitions of calf thymus DNA. (A) A myelinic cholesteric phase was observed on complexing calf thymus DNA with 1 mM bis(ethyl)spermine (BE-3-4-3) and incubating the complex on a glass slide for 2 h at 37C (200x). (B) A Schlieren texture was observed (BE-3-4-3) on incubation of the DNA in (A) for 12 h at 37C (90x). (C) A crystalline phase slowly formed from the complex in (B) and crystallization was complete by similar24 h at 37C (90x). (D) A similar pattern of phase changes was evident in calf thymus DNA complexed with 1 mM bis(ethyl)norspermine (BE-3-3-3); however, the crystal growth showed a stepped lamellar phase and birefringent areas (400x). Figure 6 | Effects of a pentamine, 3-3-3-3, and its bis(ethyl) analog on the liquid crystalline phase transitions of calf thymus DNA. Effects of a pentamine, 3-3-3-3, and its bis(ethyl) analog on the liquid crystalline phase transitions of calf thymus DNA. (A) Addition of 100 microM 3-3-3-3 to calf thymus DNA (25 mM) produced a planar network cholesteric phase on incubating for 15 min at 22C on a glass slide (100x). (B) On incubating the sample at 37C, the network texture became highly birefringent and a neuron-like dendrite developed after 12 h incubation at 37C (45x). (C) A highly ordered columnar hexagonal phase developed on a rubbed glass slide after DNA complexed with 3-3-3-3 was incubated at 37C for 12 h (200x). (D) DNA was complexed with 100 microM BE-3-3-3-3 and incubated for 12 h on a glass slide at 37C (100x). Figure 7 | Effects of 3-4-3-4-3 on the liquid crystalline phase transitions of calf thymus DNA. Effects of 3-4-3-4-3 on the liquid crystalline phase transitions of calf thymus DNA. (A) DNA (25 mM in Na cacodylate buffer) was incubated with 100 microM 3-4-3-4-3 for 12 h at 37C (200x). (B) A discotic hexagonal ordered phase is observed after incubating the DNA in (A) for 48 h at 37C (180x). This phase was stable for 1 week. DISCUSSION : The results presented in this report show multiple liquid crystalline phase transitions of calf thymus DNA in the presence of natural and synthetic polyamines. In most of the previous studies of liquid crystalline phase transitions of calf thymus DNA, low molecular weight fragments, prepared by either sonication or micrococcal nuclease digestion, were used (, --,). In contrast, we used high molecular weight calf thymus DNA for our studies. The concentration of DNA used in the present study was much lower than that used in previous reports with low molecular weight DNA ( --); however, the liquid crystalline structural transitions are comparable for the spermidine /spermine-induced liquid crystalline DNA (,). Merchant and Rill showed a dramatic decrease in the critical concentration of DNA for liquid crystalline formation as the molecular weight of DNA increased. In addition, the multivalent polyamines can enter into inter and intramolecular interactions between different or the same strands of DNA and increase the local concentrations to levels that are conducive for the liquid crystalline state (,,). A more important finding from this study is that a facile transition of the DNA to the columnar hexagonal phase occurred in the case of the natural polyamines, spermidine and spermine, and the hexamine with a closely related structure, 3-4-3-4-3. In contrast, the initial cholesteric phase of DNA was converted to a crystalline phase in the presence of spermine homologs and alkyl-substituted derivatives. Our results indicate that the overall phase behavior of DNA is complex in the presence of polyamines, with multiple textures exhibiting highly birefringent domains, indicating a supramolecular ordering of DNA molecules with polymorphous behavior. It is interesting to note that the previously reported precholesteric blue phases , which are a transition from the isotropic to the cholesteric phase, are not observed in the present case. This might be attributed to the ability of polyamines to directly order the DNA molecules to the simple twist configuration of the more stable cholestreric phase ( --). Two main phases, cholesteric and columnar hexagonal, are found in our study, either separately or in coexistence, with variations depending on local conditions. This is characterized by highly birefringent domains of oily streaks with finely divided textures with fingerprint patterns, as observed by other investigators on DNA condensed in the presence of multivalent cations (,,,). The DNA molecules are unidirectionally aligned with a lateral hexagonal order. Fan-shaped textures, which might have been formed from the original supple textures, can be seen in the columns. Undulations typical of the hexagonally ordered columnar phase are also noticed. The fluidity and order required for a liquid crystalline state are observed here. The columnar hexagonal phase also showed typical patterns of flower-like /dendrite domains whose homeotropic alignment prevented further analysis. The helical pitch determined in our study is 2.5 microm for fingerprint textures of the cholesteric phase, and this value compares well with the reports in the literature (2 --3 microm) . However, Pelta et al. reported a helical pitch of 22 microm for stripes formed from spermidine and fragmented calf thymus DNA (similar150 bp). This difference might be a consequence of the different lengths of DNA used by Pelta et al. compared to that used in the present study. DNA liquid crystals are viscous solutions in which the molecules are still mobile but are partially ordered at the same time. The mobility of the phases indicates that the mode of binding of polyamines should be along the DNA strands instead of interstrand binding, which would eventually introduce cross linking, leading to an arrest of molecular mobility. This result is consistent with recent Raman spectroscopic investigation showing non-specific interactions between polyamines and DNA . Moreover, the types of phases formed and their interconversions are unique to a particular class of polyamines, showing that the binding should be specific to the structure of polyamines. This was further supported by the variety of crystalline modifications formed by the DNA --polyamine complexes. When the charge density on polyamines increased, a decrease in fluidity was observed. This might be due to non-specific interactions of additional amine functionalities with the neighboring DNA strand. The phase interconversions also slow down when the charge density increases. For example, the phase transition from the cholesteric to columnar phase took only a few hours in the case of spermidine and its N4-methyl derivative, whereas the times taken for transformation were similar12 --44 h in the case of spermine and its N1-acetyl derivative. It is important to note here that the time-dependent changes in liquid crystalline textures of DNA occurred under conditions in which solvent evaporation was prevented by sealing the glass slides with a neutral solution of polystyrene and plasticizers in toluene (,). Therefore, the observed changes are a consequence of the reorganization of polyamines on the DNA strands. Such a mechanism is compatible with the suggestion that polyamine --DNA interaction is a multistep process, involving rapid electrostatic binding, followed by polyamine condensation on DNA and polyamine- mediated cross linking of DNA (,). In the case of diethyl derivatives of polyamines, which are therapeutically important (,,), the sequence of phase transitions was cholesteric to columnar hexagonal to crystalline. It is surprising to note that the spermine --DNA complex did not crystallize, whereas BE-3-4-3 showed crystalline phase formation, although the charge separation in both molecules is the same. This result indicates that the binding preference might be different in these molecules due to the steric hindrance imposed by the bulky ethyl groups. Among spermine and its homologs studied herein, all the tetramines, except spermine, supported growth of the crystalline phase, suggesting the importance of the natural polyamine structure and charge separation in the stabilization of liquid crystalline DNA. DNA crystallization did not occur in the presence of spermidine and spermine even after several days incubation at 37C. This observation gives a clue to the possible physiological role of polyamines in the cell nucleus, where chromatin is condensed to very tight bundles, yet retains the mobility of the double strand within the condensate . The higher polyamine analogs, such as pentamines and hexamines, also induced the liquid crystalline phase of DNA. Even though the fluidity was poor, the phase transitions occurred unambiguously, initially giving the cholesteric phase and then a highly birefringent neuron-like dendrite. The dendrites may be columnar hexagonal internally because a sample prepared on a rubbed glass plate showed a stable and clear columnar hexagonal phase (Fig. C). The rubbed glass experiment also shows the influence of polar surface forces on the stability of liquid crystalline phases. The collapse of high molecular weight DNA to toroidal and spheroidal structures has been reported in the presence of multivalent cations, including spermidine, spermine and Co(NH3)63+ ( --,,,). The organization of DNA in these structures composed of one or only a limited number of DNA molecules has attracted much attention recently because of the technological importance of these 'artificial virus' particles as gene delivery vehicles ( --). A recent report indicates that the columnar hexagonal packing of DNA facilitates the cellular transport of DNA . Most of these transfection agents are composed of multivalent cations, cationic lipids, polyethylenimine, polylysine, polyamines or their derivatives. In a recent study using freeze fracture electron microscopy, Hud and Downing found a hexagonal packing arrangement of DNA in toroids formed from lambda DNA condensed with Co(NH3)63+. The hexagonal packing of DNA has been found in many cases of DNA crystallization ( --); however, the finding of such an arrangement in a toroid composed of two molecules of DNA is very interesting . This result suggests that the hexagonal arrangement is the most efficient form of packing when individual strands of DNA are brought within a distance of 2 --3 nm in the toroids. Our finding of the hexagonal arrangement of DNA by polarizing microscopy further emphasizes the importance of this mode of packing. In summary, multiple liquid crystalline phases of DNA are induced and stabilized in the presence of polyamines. The initial phase is cholesteric in most cases. However, fingerprint textures of the cholesteric phase are found with the natural polyamines, spermidine and spermine, and the hexamine and two higher homologs of spermine (3-7-3 and 3-9-3) only. We observed columnar hexagonal textures in the case of spermine, pentamine and hexamine. There is a structural specificity effect on the facile crystallization of DNA by synthetic polyamines, including the substituted spermidine and spermine. DNA crystallization is not facilitated by natural polyamines under the conditions of our experiment. We could generate liquid crystalline phases of DNA at concentrations that are far less than that necessary for low molecular weight (similar150 bp length) DNA. A possible reason for the facile liquid crystalline phase transitions of high molecular weight DNA might be the ability of polyamines to pull together several DNA molecules by intra and /or intermolecular interactions and thus increase the effective local concentration of DNA. To the best of our knowledge, this is the first investigation of the effects of a series of spermine analogs on the liquid crystalline behavior of DNA. Backmatter: PMID- 12202769 TI - Functional analysis of iceA1, a CATG-recognizing restriction endonuclease gene in Helicobacter pylori AB - iceA1 in Helicobacter pylori is a homolog of nlaIIIR, which encodes the CATG-specific restriction endonuclease NlaIII in Neisseria lactamica. Analysis of iceA1 sequences from 49 H.pylori strains shows that a full-length NlaIII-like ORF is present in 10 strains, including CH4, but in other strains, including strain 60190, the ORFs are truncated due to a variety of mutations. Our goal was to determine whether iceA1 can encode a NlaIII-like endonuclease. Overexpression in Escherichia coli of iceA1 from CH4, but not from 60190, yielded NlaIII-like activity, indicating that the full-length iceA1 is a functional endonuclease gene. Repair of the iceA1 frameshift mutation in strain 60190 and its expression in E.coli yielded functional NlaIII-like activity. We conclude that iceA1 in CH4 is a functional restriction endonuclease gene, while iceA1 in 60190 is not, due to a frameshift mutation, but that its repair restores its restriction endonuclease activity. Keywords: INTRODUCTION : Helicobacter pylori is a gram-negative bacterium that colonizes the stomach of more than half of the world's population, and their presence increases the risk of developing peptic ulcers and gastric adenocarcinomas ( --). In Western populations, H.pylori strains of particular genotypes (such as cagA+,vacA s1m1) are more virulent than other strains ( --). Recently, two unrelated genes, designated iceA1 and iceA2, have been identified, which are present at the same genomic locus among various H.pylori strains . iceA1 expression is up-regulated by contact with epithelial cells and in some (,), but not all, populations the iceA1 genotype is associated with peptic ulcers. However, it is not known whether the gene is functional and whether or how the iceA1 product is involved in H.pylori colonization of the human stomach. DNA analysis indicates that iceA1, but not iceA2, has strong homology to nlaIIIR , which encodes a CATG-specific endonuclease in Neisseria lactamica. A nlaIIIM-like methylase gene, hpyIM , is located immediately downstream of either iceA1 or iceA2. Therefore, the iceA1 --hpyIM gene locus is a homolog of the CATG-specific type II restriction --modification system in N.lactamica (,). Our previous studies indicate that hpyIM is well conserved and encodes a CATG-specific DNA methylase in H.pylori , and that promoters for hpyIM expression vary between iceA1 and iceA2 strains . Whether or not iceA1 encodes a functional endonuclease gene remains to be determined. Studies on H.pylori strains from different geographic origins demonstrated that most iceA1 genes, including those in strains 60190 and 26695 , have frameshifts or nonsense mutations leading to early termination in their nlaIIIR-like ORF (,), whereas iceA1 from a few strains, like CH4, have an intact nlaIIIR-like ORF. To determine whether iceA1 encodes a NlaIII-like endonuclease and whether the polymorphisms in iceA1 affect the function of its product, in this study we examined the function of iceA1 from H.pylori strains CH4 and 60190, which represent full-length and truncated forms of iceA1, respectively. MATERIALS AND METHODS : Nucleotide sequence accession number | The sequence of iceA1 from H.pylori strain CH4 has been submitted to GenBank with the accession no. AF459446. Strains, plasmids, growth conditions and reagents | Helicobacter pylori strains, Escherichia coli strains and plasmids used in this study are listed in Table , and growth conditions of E.coli and H.pylori strains were as described . Restriction enzymes, digestion buffers and substrate DNAs were from New England Biolabs (Beverly, MA) and columns for protein purification were from Biosepra (Marlborough, MA). Oligonucleotides used in this study were synthesized using a Milligen 7500 DNA synthesizer at the Vanderbilt University Cancer Center DNA Core Facility. DNA techniques | Preparation of plasmid and chromosomal DNA, digestion of DNA with restriction endonucleases, automated DNA sequencing and PCR were performed as described . Computer analysis of DNA and amino acid sequences was performed with the GCG program . Construction of plasmids | Plasmids pQXCH4, pQX60T and pQX60A , which carry a full-length nlaIIIR-like ORF from strain CH4, the longest ORF (starting from the TTG site) in strain 60190 iceA1 and an ORF starting from the downstream ATG site in 60190 iceA1, respectively, were constructed as follows. DNA fragments containing the strain CH4 iceA1 ORF were amplified through PCR using IceA-F (with a 5' NdeI site) and IceACH4-R (with a 5' SalI site) as primers and chromosomal DNA from strain CH4 as template. DNA fragments containing the two 60190 iceA1 ORFs starting from the TTG or ATG sites were created by PCR using IceA60T-F (with a 5' NdeI site) or IceA60A-F (with a 5' NdeI site), respectively, with iceA-R (with a 5' SalI site) as primer pairs and 60190 chromosomal DNA as template. After digestion of the DNA fragments with NdeI and SalI, the PCR products were ligated with NdeI /SalI-digested and dephosphorylated vector pSYX22 in the presence of T4 DNA ligase to create pQXCH4, pQX60T and pQX60A. Each ligation mixture was transformed into E.coli strain ER2169 carrying a NlaIII methylase gene (nlaIIIM) in pSYX20-NlaIIIM to protect the E.coli host DNA from cleavage and transformants were selected on LB plates containing ampicillin and kanamycin. Plasmid DNA was prepared from the transformants and the constructs were confirmed by DNA sequencing. Repair of the frameshift in 60190 iceA1 and construction of pQX60R | A site-directed mutagenesis method was used to repair a point deletion mutation in H.pylori strain 60190 iceA1. Two pairs of PCR primers, iceA-F (with a 5' NdeI site) and fix-R (with a 5' PstI site) and fix-F (with a 5' PstI site and an extra G residue in the region corresponding to the site of the frameshift site in iceA1) and iceA-R (with a 5' SalI site) were designed to correct the frameshift of iceA1 located between the upstream (ATG) and the second (TTG) putative start codons. With primer pair iceA-F and fix-R and chromosomal DNA of H.pylori 60190 as template, a 140 bp PCR product was amplified, representing the 5'-most 105 bp region of iceA1, beginning at the first ATG site. With fix-F and iceA-R as primers and the same DNA as template, a 616 bp PCR product was produced, representing the remainder of iceA1, from bp 106 to the end of the gene, including the frameshift site where a point deletion mutation was present. Since primer fix-F has an extra G residue inserted at a position equivalent to bp 123 of 60190 iceA1 , the iceA1 frameshift was repaired in the 616 bp PCR product. After digestion of the 140 bp PCR product with NdeI and PstI, and the 616 bp PCR product with SalI and PstI, the two PCR products were ligated with the NdeI /SalI-digested vector pSYX22 to create pQX60R. The ligation mixture was transformed into E.coli strain ER2169 carrying pSYX20-NlaIIIM and transformants were selected on LB plates containing ampicillin and kanamycin. Plasmid DNA was prepared and sequenced to confirm that the point deletion mutation of iceA1 had been corrected. Expression in E.coli of iceA1 from CH4 and from 60190 | In pQXCH4, pQX60A, pQX60T and pQX60R, each iceA1 gene was inserted downstream of a ribosomal binding site, under the control of the T7 promoter . In E.coli strains such as ER2169, the T7 RNA polymerase gene is under the control of the lac operon. Thus, gene expression was inducible with IPTG in these E.coli strains. To express iceA1, E.coli strain ER2169 /pSYX20-NlaIIIM carrying one of the plasmids was grown in LB medium with ampicillin, kanamycin and IPTG at 37C overnight. Escherichia coli ER2169 /pSYX20-NlaIIIM without an iceA1-containing plasmid was grown under the same conditions as a control. To detect expression of the iceA1 product, the cells from E.coli ER2169 /pSYX20-NlaIIIM with or without each of the iceA1 plasmids were lysed by boiling at 100C. The lysates were resolved by SDS --PAGE (4% stacking gels and 8% separating gels) and gels were stained with Coomassie brillliant blue . Endonuclease activity assay | To determine whether the iceA1 products encode NlaIII-like endonucleases, the cultured cells were disrupted by sonication, followed by centrifugation, as described . The supernatant underwent serial dilutions in NEB buffer 4 and was then used to digest lambda DNA, and the products were resolved on agarose gels. To further purify the iceA1 product, the remainder of the crude extract was applied to a Heparin Hyper D column, which was washed and eluted, as described . The eluted fractions from this column were assayed for endonuclease activity and the active fractions were used in parallel with NlaIII to digest pBR322, pUC19 and phiX174 substrate DNA to determine whether the iceA1 protein has NlaIII-like endonuclease activity. Table 1 | Strains, plasmids and oligonucleotides used in this study RESULTS : Comparison of the DNA sequences of nlaIIIR from N.lactamica and iceA1 from H.pylori strains 60190 and CH4 | DNA analysis demonstrates that iceA1 from H.pylori has similar60% similarity to nlaIIIR, which has a 690 bp ORF and encodes a 230 amino acid protein product, the restriction endonuclease NlaIII . However, the homology at the amino acid level is limited in most of the 49 iceA1 genes studied (,; GenBank accession nos AF239991 --AF239994 and AF001537 --AF001539), because of frameshifts and early terminations in the nlaIIIR-like ORF. Thus far, a full-length nlaIIIR-like ORF has only been observed in 10 (PO31, AU8, AU18, CH4, F38, F72, F43, India227, Alaska209 and Alaska218) of 49 iceA1 strains studied, and similarity between their deduced IceA1 proteins and NlaIII was 52 --57% (,). To determine whether iceA1 encodes a restriction endonuclease, we chose strains CH4 and 60190 for functional analysis. iceA1 from strain CH4 has a full-length nlaIIIR-like 684 bp ORF, encoding a 228 amino acid product, very similar to that of nlaIIIR, whereas the nlaIIIR-like ORF in strain 60190 is not complete and the longest ORF is only 534 bp, encoding a 178 amino acid product. Alignment of sequences with nlaIIIR reveals several mutations in the predicted iceA1 coding region, including a 9 bp in-frame deletion that is present in both CH4 iceA1 and 60190 iceA1, and a single nucleotide G deletion in the run of six G residues present in CH4, which results in a frameshift in the 60190 iceA1 coding region. This frameshift results in the ORF terminating at a TAG beginning at position 193 of 60190 iceA1, and leads to a 64 amino acid ORF representing the N-terminus of a putative restriction endonuclease. The longest predicted coding region from the 60190 iceA1 starts from a TTG site that is 150 bp downstream of the nlaIIIR-equivalent start site ATG, to form a 534 bp ORF. This would encode a 178 amino acid product corresponding to the C-terminus of a putative restriction endonuclease. An ORF starting from an ATG at position 304, an iceA1 translational start site that is conserved in various strains , is present within both CH4 and 60190 iceA1. Expression of the CH4 and 60190 iceA1 products in E.coli | To determine whether iceA1 encodes a functional NlaIII-like endonuclease, we sought to express the potential nlaIIIR-like ORFs in E.coli. The 684 bp (full-length) nlaIIIR-like ORF in iceA1 from strain CH4 was amplified from CH4 chromosomal DNA and cloned into the vector pSYX22 to create pQXCH4. The longest coding region of 534 bp for iceA1 in strain 60190 starts from a TTG site that is 150 bp downstream of the equivalent ATG start site in nlaIIIR . This DNA fragment was amplified and cloned into pSYX22 to create pQX60T. A downstream 381 bp nlaIIIR-like ORF in iceA1 from strain 60190 starting from the ATG site at bp 304 was also cloned into pSYX22 to create pQX60A. The resulting constructs, pQXCH4, pQX60T and pQX60A, were then transformed into E.coli strain ER2169, which carries a functional nlaIIIM gene in pSYX20-NlaIIIM . After IPTG induction, E.coli produced iceA1 protein products of the expected sizes (data not shown). To determine whether the cloned iceA1 products had restriction endonuclease activity, serial dilutions of crude cell extracts from cultured E.coli ER2169 /pSYX20-NlaIIIM, with or without pQXCH4, pQX60T or pQX60A, were used to digest lambda DNA. As expected, the crude extract from E.coli ER2169 /pSYX20-NlaIIIM failed to digest lambda DNA (Fig. A). A fraction of the DNA in these lanes did not migrate from the loading wells, which may be explained by the presence of large amounts of DNA-binding proteins in the crude extract. When the crude extract from E.coli ER2169 /pSYX20-NlaIIIM with pQXCH4 was used for digestion, the substrate DNA was digested into smaller DNA fragments (Fig. A). The lane loaded with the smallest volume (0.1 microl) had more small DNA fragments than the other lanes, which may be due to less interference by DNA-binding proteins. In contrast, crude extracts from E.coli with pQX60T or pQX60A failed to digest the lambda DNA into smaller fragments. In total, these results indicate that endonuclease activities were present in the crude cell extract of E.coli ER2169 /pQXCH4, but not in the control cells or cells with pQX60T or pQX60A. To investigate whether the endonuclease activities are NlaIII-like, the endonuclease from the crude cell extract of E.coli ER2169 with pQXCH4 was purified using a Heparin Hyper D column. The eluted fractions that contained the endonuclease activities were used to digest substrate (pBR322, pUC19 and phiX174) DNAs in parallel with NlaIII. The digestion pattern of DNA by the pQXCH4 iceA1 product is identical to that by NlaIII (Fig. B), which indicates that CH4 iceA1 encodes a NlaIII-like endonuclease. Fixing the frameshift in 60190 iceA1 and examining the endonuclease activity of the repaired iceA1 product | For H.pylori strain 60190, a protein beginning at the TTG site would not include a 50 amino acid region that is equivalent to the N-terminal region of NlaIII, which could explain why no endonuclease activity was detected when the iceA1 protein from strain 60190 was examined. To test whether the frameshift caused by the missing G residue at position 123 was the main problem, it was repaired by inserting an extra G residue into the poly(G) tract and the repaired iceA1 was inserted into the vector pSYX22, to create pQX60R. In the repaired iceA1, in addition to the G insertion, a T->C and an A->G point mutation were also introduced in the repaired region, to create a PstI recognition site (Fig. B). These substitutions are silent, since both TTA and CTG encode leucine as amino acid 36 of the repaired protein. The repaired iceA1 is a 684 bp ORF, encoding a 228 amino acid protein product, with 60% identity with NlaIII. To express the repaired iceA1 in E.coli, pQX60R was then transformed into E.coli strain ER2169 carrying pSYX20-NlaIIIM. After induction by IPTG, a 26 kDa protein was produced, as expected . To determine whether the repaired iceA1 product has endonuclease activity, serial dilutions of crude cell extracts from cultured E.coli strain ER2169 /pSYX20-NlaIIIM, with or without pQX60R, were used to digest lambda DNA. As expected, the crude extract from E.coli ER2169 /pSYX20-NlaIIIM failed to digest lambda DNA (Fig. A). However, the migration of lambda DNA was shifted to a lower position in the lanes with 3 and 1 microl of the crude extract, because of the presence of large amounts of DNA-binding proteins in the crude extract. When the crude extract from E.coli ER2169 /pSYX20-NlaIIIM with pQX60R was used for digestion, the substrate DNA was digested into smaller DNA fragments (Fig. A). Overall, these results indicate that restriction endonuclease activities were present in the crude cell extract of E.coli ER2169 /pQX60R, but not in the control cells. To investigate whether the activities are NlaIII-like, the restriction endonuclease from the crude cell extract of E.coli ER2169 with pQX60R was purified using a Heparin Hyper D column. The eluted fractions that contained the endonuclease activities were used to digest substrate (pBR322, pUC19 and phiX174) DNAs in parallel with NlaIII. The digestion pattern of DNA by the repaired iceA1 product is identical to that by NlaIII (Fig. B), which indicates that iceA1 from strain 60190, after repair of the frameshift, encodes a NlaIII-like endonuclease. Figure 1 | Alignment of the nucleotide sequences of N.lactamica nlaIIIR (GenBank accession no. Alignment of the nucleotide sequences of N.lactamica nlaIIIR (GenBank accession no. U59398) and the iceA1 regions from H.pylori strains 60190 and CH4. Deletion and insertion mutations in the iceA1 regions, relative to the full-length nlaIIIR, are shaded. Dots indicate identical nucleotides. Dashes indicate gaps in the nucleotide sequence. Potential translation start and stop sites are shown in bold. Figure 2 | Endonuclease assay of the products of the iceA1 genes in CH4 and 60190. Endonuclease assay of the products of the iceA1 genes in CH4 and 60190. (A) Assay of crude extracts from E.coli strain ER2169 /pSYX20-NlaIIIM (1) and ER2169 carrying pQX60A (2), pQX60T (3) and pQXCH4 (4). Crude extracts (0.1 --3 microl) were used to digest 1 microg bacteriophage lambda DNA and the digested DNA was electrophoresed on a 1% agarose gel. Arrows indicate the position of undigested lambda DNA in the gels. (B) Assay of the purified CH4 iceA1 endonuclease. Parallel digestion of pUC19, pBR322 and phiX174 DNA was performed with both the NlaIII and the CH4 iceA1 endonucleases. The DNA standards are labled L. Figure 3 | A frameshift is present in H.pylori 60190 iceA1 when compared to NlaIII. A frameshift is present in H.pylori 60190 iceA1 when compared to NlaIII. (A) Schematic representation of the iceA1 --hpyIM region from H.pylori 60190. The locations of three potential translation start sites (ATG, TTG and ATG) and a potential stop site (TAG) in iceA1 are indicated. The DNA sequence and its deduced N-terminal NlaIII-like protein sequence from the first (ATG) to the second translation start site (TTG) in the iceA1 region with the frameshift site also are presented. The ATG and TTG sites are shown in bold. The boxed region indicated by an arrow shows the frameshift site. (B) The 60190 iceA1 sequences near the site of the frameshift, before and after the repair, are presented. The new PstI site created and the polyguanosine tract are underlined. Residues changed by the repair process are shown in bold. Figure 4 | Expression of iceA1 in E.coli strain ER2169 after repair of its frameshift. Expression of iceA1 in E.coli strain ER2169 after repair of its frameshift. 10% SDS --PAGE with Coomassie blue staining of lysates from E.coli strain ER2169 without (WT) or with pQX60R. +, induction with IPTG; --, no induction. The 26 kDa arrow indicates the protein product of the repaired iceA1. Migration (in kDa) of each protein standard is indicated on the left. Figure 5 | Endonuclease assay of the product of the repaired iceA1. Endonuclease assay of the product of the repaired iceA1. (A) Assay of crude extracts from E.coli strain ER2169 (1) and ER2169 carrying pQX60R (2). Crude extracts (0.1 --3 microl) were used to digest 1 microg bacteriophage lambda DNA. The digested DNA was electrophoresed on a 1% agarose gel. Arrows indicate the position of undigested lambda DNA in the gels. (B) Assay of the purified product of repaired iceA1 from strain 60190. Parallel digestion of pUC19, pBR322 and phiX174 DNA was performed with both NlaIII and the repaired iceA1 product. The DNA standards are labeled L. DISCUSSION : This study demonstrates that the full-length iceA1 from strain CH4 encodes a functional restriction endonuclease, which is consistent with our recent observation that NlaIII-like endonuclease activity is present in the cells of this strain . Although iceA1 from 60190 has strong homology to nlaIIIR in N.lactamica, the predicted iceA1 ORF is much smaller than that in nlaIIIR, because of a frameshift mutation . Thus, not surprisingly, the iceA1 ORF from strain 60190 expressed in E.coli lacked NlaIII-like activity, which is also consistent with the absence of NlaIII-like activity in cells of strain 60190 . However, after restoring a full-length (nlaIIIR-like) ORF in iceA1 from strain 60190 by repairing the frameshift, the 'repaired' iceA1 encodes an NlaIII-like endonuclease in E.coli. These results explain why there was no NlaIII-like activity in this H.pylori strain , even though iceA1 is transcribed . Compared to the repaired iceA1 product, the protein transcribed from the predicted start codon (TTG) in iceA1 lacks a 50 amino acid N-terminal region . The lack of endonuclease activity in the native product of iceA1 from strain 60190 indicates that the 50 amino acid N-terminal region is required for NlaIII-like function. One question is whether this type of mutation may be repaired within H.pylori populations to produce a functional NlaIII-like endonuclease. In iceA1 from strains like 60190, only a single nucleotide correction (inserting a G residue) is needed to restore the full-length nlaIIIR-like ORF. The frequency of single transition or transversion mutations in the ribosomal RNA region of H.pylori has been measured in strain 60190 by monitoring the frequency of spectinomycin resistance ( --) and found to be 7.8 x 10 --9 (Q.Xu and M.J.Blaser, unpublished data). If the frequency of spontaneous mutation in the iceA1 region is similar to that in the ribosomal RNA region, the frequency of frameshift repair in 60190 iceA1 would be <7.8 x 10 --9, assuming that production of point insertion mutations is more difficult than transition or transversion mutations (,). However, the frameshift in iceA1 from strain 60190 occurs in a polyguanosine tract (six G residues in strain CH4, five in 60190), suggesting that slipped strand mispairing could occur in this region during DNA replication (,). Thus, the addition of one guanosine residue could occur at a much higher frequency than the background mutation rate ( --). It can be beneficial to bacterial populations to contain genes that undergo phase variation, allowing them to be ready for environmental changes. Examples of such contigency genes in H.pylori have been found in Lewis antigen expression (,), which exhibits phase variation at a 0.2 --0.5% rate in vitro . Analysis of the whole genomic sequences (,,) has revealed the presence of 10 potential phase-variable restriction --modification genes, not including iceA1, in H.pylori. These genes carry either homopolymeric tracts or dinucleotide repeats with their length varying from 5 to 15 (,). Within different H.pylori populations, changes in the lengths of specific homopolymeric tracts have been found in some of the putative phase-variable restriction --modification genes . Contigency restriction --modification genes have also been identified in other bacteria ( --), including Haemophilus influenzae, Neisseria meningitidis and Mycoplasma pulmonis. A recent study demonstrated that a contingency restriction --modification system in M.pulmonis was turned on after the bacteria contacted the lower respiratory tract in infected rats , suggesting that contingency restriction --modification genes may play important roles in host --bacteria interactions. If iceA1 in strain 60190 functions as a contingency gene, H.pylori could adapt to environmental changes, such as contacting epithelial cells in the stomach, or to phage infection, by encoding a NlaIII-like restriction endonuclease. However, whether iceA1 is actually a contigency gene remains to be determined. Another question is how often iceA1 mutations in various strains are located in similar homopolymeric regions. Analysis of iceA1 sequences from 36 H.pylori strains (which are among the 49 strains referred to previously) showed that a total of 14 strains (including 60190) carry iceA1 genes with frameshift mutations in homopolymeric tracts, compared to CH4 iceA1. Among these homopolymeric tracts, 12 are in the same poly(G) tract of five or seven residues, +-1 nt at bp 123 , and two in the poly(A) tract of eight residues near the beginning of the iceA1 ORF, --1 nt at bp 19 . A +1 or --1 nt frameshift in these tracts would repair the frameshift mutations. However, additional mutations are also present in most of these other strains , which makes repair very difficult. Among the 36 strains analyzed, only seven (including CH4) carry a full-length nlaIIIR-like ORF . In the remaining strains, the iceA1 region has various mutations, including insertions, deletions and nonsense mutations. Only 10 (including 60190) have mutations at a single location in iceA1 when compared to nlaIIIR , whereas others have mutations at multiple locations, making them more difficult to repair. For example, corrections of both a deletion and an insertion mutation are needed to restore the full-length nlaIIIR-like ORF in strain J166. Biochemical approaches failed to detect NlaIII-like activity in strain J166 , although the iceA1 region is transcribed . Thus, it appears that there are no special mechanisms, such as ribosomal frameshifting, that could permit translation to bypass these mutations. Furthermore, unless there is a special mechanism that would greatly increase the rate of mutation repair in non-homopolymeric tracts, it seems unlikely that the iceA1 gene serves as an ordinary contingency gene, analogous to the fucosyltransferase genes controlling synthesis of the Lewis antigens ( --). Since H.pylori is naturally competent for the uptake of chromosomal DNA and plasmid DNA, when more than one iceA1 strain is present in the gastric mucosa, horizontal exchange may occur. In such an exchange, an alternative way for H.pylori to repair iceA1 mutations and restore NlaIII-like activity is to acquire DNA from strains with the entire gene or complementary regions by gene conversion. However, restriction --modification systems are highly diversified among various H.pylori strains (, --) and form barriers for transformation of DNA fragments between strains , which would limit the frequency of such repair. Overall, we have shown that iceA1 of CH4 is a functional NlaIII-like endonuclease gene and the iceA1 gene of 60190 can be repaired to code a functional restriction enzyme. In a majority of H.pylori strains, iceA1 appears to be a degenerate gene that was once part of an restriction-modification system. However, expression of truncated iceA1, such as that from strain J166, is up-regulated by contact with epithelial cells . It remains to be determined whether iceA1 plays a role other than encoding a NlaIII-like endonuclease or not. Table 2 | Summary of the frameshift or nonsense mutations in the nlaIIIR-equivalent coding region of iceA1 from 36 H.pylori strains Backmatter: PMID- 12202758 TI - Transcriptome analysis of Escherichia coli using high-density oligonucleotide probe arrays AB - Microarrays traditionally have been used to analyze the expression behavior of large numbers of coding transcripts. Here we present a comprehensive approach for high-throughput transcript discovery in Escherichia coli focused mainly on intergenic regions which, together with analysis of coding transcripts, provides us with a more complete insight into the organism's transcriptome. Using a whole genome array, we detected expression for 4052 coding transcripts and identified 1102 additional transcripts in the intergenic regions of the E.coli genome. Further classification reveals 317 novel transcripts with unknown function. Our results show that, despite sophisticated approaches to genome annotation, many cellular transcripts remain unidentified. Through the experimental identification of all RNAs expressed under a specific condition, we gain a more thorough understanding of all cellular processes. Keywords: INTRODUCTION : Genome sequence information has accumulated at a fast pace in recent years and the generation of whole genome sequences is now commonplace. However, the number of uncompleted genome projects significantly exceeds the number of completely annotated and published sequences ( and ). One of the primary reasons for this gap between sequence generation and public release is the still difficult task of sequence annotation, of interpreting raw sequence data into useful biological information. Most of the genome annotation information is generated using bioinformatics approaches. These in silico methods used for gene prediction in combination with homology searches are applied to the primary genome sequence. However, coding sequences, those portions of the genome that are transcribed and ultimately translated, are not the only elements of the genome which are transcribed into RNA. Transcribed but untranslated regions (UTRs) are common at the 5' and 3' ends of genes since transcription initiation and termination sites generally extend beyond translation start and stop sites. In prokaryotes such as Escherichia coli, operons may be described as consecutive genes, which are transcribed into a single polycistronic mRNA molecule. In the case of these operons, the intergenic regions are transcribed but not translated. In addition, at least 34 untranslated small RNA molecules, which may have regulatory functions, have been reported ( --). Further, there is evidence that current annotation algorithms have limitations which can cause errors in the annotation process ( --). By investigating observed transcripts, which are distinct from previously annotated genes, we are able to identify new potential genes. Several in silico approaches, based largely on primary sequence analysis, have proven successful at identifying many of these transcript elements, including promoter regions (,), transcription termination sites (,), operons and small RNAs ( --). In addition to these transcribed elements, a number of intergenic repeats in E.coli have been computationally identified and documented (,). However, these computational approaches rely on primary sequence analyses and cross-species sequence comparisons. Genome-wide experimental identification of transcripts, such as with microarrays, has been limited primarily to coding sequences. For identifying transcribed intergenic regions, we present an orthogonal approach to in silico primary sequence analysis methods that is based on high density oligonucleotide probe arrays, which interrogate the sense strand of coding sequences and both strands in the intergenic regions of the genome. Using E.coli RNA from cells grown on different media, we have identified over 1100 transcripts corresponding to intergenic regions. We proceeded to classify these transcripts using sequence analysis, expression clustering, sequence homology and information collected from the literature and public databases. MATERIALS AND METHODS : Strain and growth conditions | Escherichia coli strain MG1655 cells were grown in Luria --Bertani broth or on solid medium and used for inoculation of liquid cultures. Cells were grown in 50-ml batch cultures in 250-ml Erlenmeyer flasks at 37C with aeration by rotary shaking (300 r.p.m.). The culture media used were Luria --Bertani (LB) or M9 minimal medium as described elsewhere supplemented with glucose (0.2%) or glycerol (0.2%). Anaerobic growth was performed at 37C in the same flasks fitted with butyl rubber stoppers and the air in the dead space replaced with argon. Growth was monitored at 600 nm on a Hitachi U-2000 spectrophotometer. Cells were harvested in mid log phase, midway between beginning log phase and stationary phase, early stationary phase or deep stationary growth phase (24 h after the culture reached stationary phase) . RNA isolation, cDNA synthesis and target labeling | Total RNA was isolated from the cells using the protocol accompanying the MasterPureTM complete DNA /RNA purification kit from Epicentre Technologies (Madison, WI). Isolated RNA was resuspended in diethyl pyrocarbonate-treated water and quantitated based on the absorption at 260 nm. The cDNA synthesis method has been described previously . Briefly, 10 microg total RNA was reverse transcribed using the Superscript II system for first strand cDNA synthesis from Life Technologies (Rockville, MD). The remaining RNA was removed using 2 U RNase H (Life Technologies) and 1 microg RNase A (Epicentre) for 10 min at 37C in 100 microl total volume. The cDNA was purified using the Qiaquick PCR purification kit from Qiagen (Valencia, CA). Isolated cDNA was quantitated based on the absorption at 260 nm and fragmented using a partial DNase I digest. The fragmented cDNA was 3' end-labeled using terminal transferase (Roche Molecular Biochemicals, Indianapolis, IN) and biotin-N6-ddATP (DuPont /NEN, Boston, MA). The fragmented and end-labeled cDNA was added to the hybridization solution without further purification. Genomic DNA labeling and hybridization | Escherichia coli genomic DNA (5 microg) was fragmented using 0.2 U DNase I (Roche) in one-phor-all buffer (Amersham, Piscataway, NJ), adjusted to a final volume of 20 microl and incubated at 37C for 10 min, followed by inactivation of DNase at 99C for 10 min. The fragmented DNA was subsequently labeled with terminal transferase (Roche) and biotin-N6-ddATP (DuPont /NEN) in accordance with the manufacturers' protocols. Standard hybridization, wash and stain protocols were used (Affymetrix, Santa Clara, CA). Reverse transcription --PCR (RT --PCR) | RNA isolation and cDNA synthesis was performed as described above. The PCR reaction was carried out with 70 ng cDNA as template and 1 microM forward and reverse primers. The reaction was cycled 25 times with a 55C annealing temperature and a 2 --4-min extension time at 72C, depending on the size of the expected product. We used the same RNA in the PCR reaction as a negative control to test for genomic contamination. Array design | A detailed description of the microarray is available . In summary, each array chip contains 295 936 oligonucleotide probes. Half of the probes are designed to be perfect match (PM) probes, which correspond to 25mer oligonucleotides in the E.coli genome, while the other half are designed to be mismatch (MM) probes, which correspond to the same 25mers as the PM probes except that the 13th base pair is complemented. The chip assays every annotated gene with a set of probe pairs and every intergenic region in both orientations with a set of probe pairs. Probe sets generally contain 15 PM and 15 MM probes. Transcript identification | The GeneChip Software analysis program MAS 4.1 and DMT 2.0 (Affymetrix) were used for the analysis of gene expression and expression clustering, respectively. To identify transcripts within intergenic regions, we developed an algorithm for the analysis of the .cel file generated by MAS 4.1. The .cel file contains the probe locations and the individual intensities of the PM and corresponding MM probes on the microarray. In order to identify transcripts, we looked for sets of adjacent probes (two or more probes) in which PM -- MM for each adjacent probe exceeds an expression threshold in both replicates (based on empirical results, we used a difference threshold of 200). We prefer reasonably strict criteria for transcript identification to ensure a high specificity for transcript detection. For each duplicate experiment, we searched for all possible transcripts which met these criteria in all interrogated intergenic regions. In order to correct for possible cross-hybridization effects, labeling inconsistencies or hybridization variations, we combined neighboring transcripts in the same intergenic region into a single transcript if they were separated by a single probe, which failed to meet our expression criteria. We applied this approach to all interrogated intergenic regions genome wide and then proceeded to classify the identified transcripts. Table 1 | The 13 different growth conditions used for the E.coli transcriptome analysis RESULTS : Many approaches that quantify the expression level of a gene based on oligonucleotide array data operate under the assumption that all (or at least most) oligonucleotide probes for a given gene are essentially independent measurements of the same transcript expression ( --). This is a reasonable and convenient assumption for genes, where the existence and the exact position of the transcript are known a priori, but when we search for new transcripts, such as in intergenic regions, we do not have the luxury of this assumption. Rather, we perform RNA expression analysis at a sub-transcript resolution. Initial analysis of the data across all experiments showed a range of hybridization affinities for different probes. We removed 2671 probes in the intergenic regions from the analysis for which there was evidence of significant cross-hybridization or other non-specific hybridization. These probes were determined by hybridizing E.coli genomic DNA labeled directly with terminal transferase to the probe array and removing the probes that failed to meet our difference threshold. The remaining probes were studied by hybridizing biotin-labeled cDNA derived from 13 different growth conditions in duplicate for a total of 26 arrays . For transcript discovery a stringent difference model was developed, which is based on evidence that an average difference model can linearly approximate actual expression levels (,). A probe had to meet the difference requirement in both duplicate experiments before we considered the probe 'expressed' (see Materials and Methods). After identifying a conservative set of potential transcripts in intergenic regions, we then proceeded with their classification based on their genome location as operon elements, 5'-UTRs, 3'-UTRs or as transcripts of unknown function . For additional validation of our classification, we determined the co-regulation of the identified transcripts with their flanking ORFs using the self-organizing map (SOM) algorithm . Transcripts that are co-regulated across many conditions are likely to be from the same transcript . In addition, we performed a homology search against the complete genome sequence of Salmonella typhimurium (the closest fully sequenced relative to E.coli) to identify conserved regions (329). Sequences can be conserved for many different reasons, including coding regions, complex promoters or leader sequences, transcriptional and post-transcriptional regulatory signals, small RNAs, transcriptional terminators and sequences of as yet unknown function. We used the cluster and homology analyses together with annotation programs (,,) and information collected from the literature and public databases to further characterize the transcripts and to classify them as potential new ORFs or RNA transcripts that serve as small regulatory RNAs (sRNA) . Operon elements | A gene can be described as belonging to an operon if it is one of two or more adjacent genes which are transcribed into a single transcript . Similarly, an intergenic region is part of an operon if it is transcribed in the same transcript as both its flanking genes. The ability to identify operons can be very useful in understanding gene function, since genes that are members of the same operon generally code for proteins which have functional roles in the same cellular pathways. While correlated expression of two neighboring genes may be a reasonable indication that the genes are co-transcribed, this correlated expression coupled with evidence of similar intergenic expression between the two genes provides a much stronger signal. Intergenic transcripts are classified as part of an operon if the orientation of the intergenic region matches that of the flanking genes, if both genes are expressed and if the expressed intergenic transcript extends across the entire intergenic region. Using these parameters we identified 289 of these intergenic regions which have been previously documented or predicted as being part of an operon (,; ) (Supplementary Material, Table S1). Based on this comparison the false positive rate for transcript detection was estimated to be <1%. Characterizing the false negative rate proves to be problematic because many of these regions showed little or no expression under our 13 conditions and we are limited to drawing conclusions only from elements which are expressed in our experiment set. In addition to the transcript analysis described above, we also performed an expression cluster analysis using the SOM algorithm to investigate correlation between the expression of an intergenic region and the expression of its neighboring genes. The average difference values for the flanking genes and the intergenic region under all 13 conditions were used to identify co-regulated expression. Of our predicted operons, 71% showed co-regulation in at least two of the three transcripts (flanking genes and intergenic region), while 81% of the documented operons offered this evidence of co-regulation. Figure shows the expression levels for individual probes interrogating the predicted hnr --galU operon. RT --PCR confirmed a single RNA transcript for these two genes and the intergenic region . We confirmed a single RNA transcript using RT --PCR for six additional predicted operons. Of the four intergenic regions which have not previously been documented or predicted as part of an operon but for which we observe operon evidence, we found that two were co-regulated with flanking genes (rpsM /rpmJ and rplN /rpsQ), which code for 30S and 50S ribosomal subunit proteins. Based on our findings and the close functional relationship of the gene products, they are strong candidates for new, previously unidentified operons. 5'-Untranslated region (5'-UTR) | As with the operons described above, experimental evidence for 5' expressed regions can supplement computational approaches by identifying not only transcription start sites for genes, but also multiple start sites when different promoters are employed under different conditions, as well as cis-regulatory sites upstream of known genes. In order for an intergenic transcript to be classified as a 5'-UTR in our analysis, we required the transcript to be in the same orientation as its downstream gene and to be expressed under the same growth conditions. We made the assumption that the transcript must be >=70 nt to encode a 5'-UTR, slightly longer than the expected 50 --60 nt of a promoter, and that the transcript extends close to the downstream gene translational start site, i.e. the transcript must extend to the penultimate or ultimate probe in the probe set of the intergenic region. Figure shows an example for the microarray detection of the transcribed but not translated leader sequence of the ompA mRNA . The PM -- MM probe intensities and the probe locations were used to determine the transcriptional start site, which was found to be close to the predicted promoter location for the ompA gene. We identified a conservative set of 353 transcripts which met our expression criteria for 5'-UTRs (Supplementary Material, Table S2). Of these transcripts, 294 either showed concordant expression with their downstream ORF in all 13 experiments or else showed homology to S.typhimurium with an E value < 0.01 and an over all identity of >65%. Fifteen 5'-UTRs contain conserved regulatory sequences , one of which matches a previously identified small RNA (sraB) and an additional one (crpT) a potential small RNA . However, based on the signal location and co-expression with its downstream gene, our study suggests that crpT is the 5'-untranslated leader of yhfA as originally reported and not an independent transcript as speculated by Carter et al. . An additional 49 other transcripts are classified as potential small ORFs (Supplementary Material, Table S2). 3'-Untranslated region (3'-UTR) | The classification of transcripts as 3'-UTRs is analogous to that of the 5'-UTRs. The intergenic transcript must be in the same orientation as its upstream gene and was required to be expressed under the same growth conditions. In addition, we restricted the transcripts to be at least 70 bp in length and to extend close to the upstream gene predicted translational stop site. With these criteria we identified 122 potential 3'-UTRs, of which 69% are either expressed concordant with their upstream gene in all 13 experiments or have sequence homology to S.typhimurium with an E value < 0.01 and an overall identity of >65% (Supplementary Material, Table S3). Eleven of the 122 transcripts are classified as potential novel small ORFs. Transcripts of unknown function | Finally, we identified 334 transcripts longer than 70 bp that were expressed but which could not be classified as operon elements, 5'-UTRs or 3'-UTRs (Supplementary Material, Table S4). This group of transcripts has a hybridization signal separate from and discontinuous with the signals from neighboring ORFs. Over 200 transcripts in this group showed sequence homology with S.typhimurium or considerable expression levels (more than three times background). Of the 34 reported E.coli sRNA molecules interrogated in intergenic regions on the array ( --), we detected transcripts for 19 on the reported strand (Supplementary Material, Table S5) and an additional five on the opposite strand in this group. Several of the known sRNAs are expressed only under specific growth conditions and we cannot draw conclusions from a negative result. We predict an additional nine transcripts as being good candidates for new small RNAs based on their homology to S.typhimurium, their average transcript intensity and their expression in either late log, early stationary or stationary growth phase. In addition, their sizes range from 50 to 400 nt (Supplementary Material, Table S5). These are conditions under which most of the other sRNA transcripts were identified. Furthermore, we detected transcripts for 31 predicted but not experimentally confirmed ORFs from the Colibri and EcoGene databases ( and ) (Supplementary Material, Table S5). This is the most challenging group of transcripts to characterize and understand. Further sequence analysis to identify promoters, transcriptional terminators and secondary structure is necessary to say how many of these transcripts qualify as small RNA molecules, new ORFs, regulatory transcripts such as antisense RNAs or transcripts of as yet unknown function. Additional protein sequence homology, functional annotation information and experimental validation is also needed to confirm the predicted ORFs. We have intentionally not assigned any new E.coli gene names to our transcripts awaiting further characterization. Figure 1 | The pts operon in E.coli. The pts operon in E.coli. The enlarged pictures of a microarray show the high expression level of the neighboring genes ptsH, ptsI and crr in the pts operon. The bottom picture (from RegulonDB; ) shows the location of the genes within the genome. Figure 2 | Operon detection using oligonucleotide probe intensities. Operon detection using oligonucleotide probe intensities. Individual oligonucleotide probe intensities (PM -- MM) from three conditions are shown to validate the microarray-predicted hnr --galU operon. Intensities for individual probes interrogating hnr, the 200 bp intergenic region and galU are shown. This operon was independently confirmed using RT --PCR . Figure 3 | RT --PCR for seven predicted operons. RT --PCR for seven predicted operons. The DNA bands represent the PCR products for the following operons, including the expected size of the PCR product in parentheses (see also Supplementary Material, Table S1): lane 1, rpsR --rplI (0.9 kb); lane 2, ytfQ --ytfR (2.4 kb); lane 3, yaeR --mesJ (1.4 kb); lane 4, rplA --rplL (1.5 kb); lane 5, ptsH --ptsJ (1.7 kb); lane 6, purA --yjeB (1.8 kb); lane 7, hnr --galU (2 kb); lane St, 1 kb standard DNA ladder. Figure 4 | 5'-UTR detection upstream of ompA. Individual oligonucleotide probe intensities (PM -- MM) from three conditions are shown to validate the microarray-detected 5'-UTR upstream of ompA. 5'-UTR detection upstream of ompA. Individual oligonucleotide probe intensities (PM -- MM) from three conditions are shown to validate the microarray-detected 5'-UTR upstream of ompA. Intensities for individual oligonucleotide probes interrogating ompA, the 356 bp intergenic region and sulA are shown. The arrows above the indicated genes show the direction of transcription. Table 2 | Summary of all detected transcripts and their classification and characterization DISCUSSION : While the molecular biology community has met with great success in developing computational approaches for genome wide transcriptome analysis, experimental evidence to support this analysis tends to be either on a gene-by-gene level or else, in the case of microarrays, only targeted at transcripts that are also translated (,,). A recent study of chromosomes 21 and 22 in human cells identified novel RNA transcripts not detected by sequence analysis . For the analysis of gene expression in an organism and the interpretation of the generated data we will increasingly rely on complete and accurate catalogs of genes, mRNAs, proteins and untranslated but transcribed regions. With the recent advances in oligonucleotide probe array technology and the availability of complete genome sequences we are now in a position to assay the complete transcriptome of many organisms, as opposed to only their coding subset. By assaying the E.coli transcriptome under a range of conditions, we identified multiple non-coding transcript elements, including 5'-UTRs, 3'-UTRs, small RNA molecules and polycistronic elements (operons). In addition, we were able to detect transcripts that escaped gene prediction programs due to non-conforming characteristics. The arrays can be constructed without a priori knowledge of genome annotation and can provide the experimental foundation for a complete transcriptome analysis. By interrogating both strands of a genomic sequence on one array, valuable information on possible antisense gene regulation can be obtained and can provide the basis for a more accurate understanding of gene translation. Our experimental approach to transcriptome analysis in E.coli could be extended to the genomes of other organisms with complete sequence data but incomplete annotation information. Using conservative criteria, we identified a set of 1102 transcripts in the intergenic regions of E.coli which we classified as operon elements, 5'-UTRs, 3'-UTRs, small RNAs, new ORFs or transcripts of unknown function . Using the experimental approach of detecting transcripts under different growth conditions gives us the opportunity to detect >95% of all possible transcripts as judged by the total number of genes detected (data not shown). In fact, we validated most of the reported sRNAs, operons and ORFs by using secondary analysis tools. Our analysis will be most effective if applied in combination with other evidence, such as computational approaches, which predict operons , promoters and ORFs (,,,), sRNAs ( --), transcription termination sites (,), etc. Whenever possible, we attempted to validate our observed intergenic transcripts with independent means of analysis, such as sequence homology, expression clustering or ORF identification programs. The data presented show that a large portion of the E.coli transcriptome remains to be characterized. We also expect that experimental evidence of transcripts will be useful for developing the next generation of gene prediction algorithms, and only with a complete understanding of transcription for both coding regions and intergenic regions can we fully comprehend cellular processes. SUPPLEMENTARY MATERIAL : Supplementary Material is available at NAR Online. Backmatter: PMID- 12202751 TI - Distinct domains in the CArG-box binding factor A destabilize tetraplex forms of the fragile X expanded sequence d(CGG)n AB - Formation of hairpin or tetraplex structures of the FMR1 gene d(CGG)n sequence triggers its expansion, setting off fragile X syndrome. In searching for proteins that destabilize d(CGG)n secondary structures we purified from rat liver quadruplex telomeric DNA binding protein 42 (qTBP42) that disrupts G'2 bimolecular tetraplex d(CGG)n while paradoxically stabilizing the G'2 structure of the telomeric sequence d(TTAGGG)n. Based on peptide sequence homology of qTBP42 and mouse CArG-box binding factor A (CBF-A), we provide direct evidence that recombinant CBF-A protein is physically and immunochemically indistinguishable from qTBP42 and that it too destabilizes G'2 d(CGG)n while stabilizing G'2 d(TTAGGG)n. We inquired whether CBF-A employs the same or different domains to differentially interact with G'2 d(CGG)n and G'2 d(TTAGGG)n. Mutant CBF-A proteins that lack each or combinations of its five conserved motifs: RNP11, RNP12, RNP21, RNP22 and ATP /GTP-binding box were tested for their G'2 d(CGG)n destabilization and G'2 d(TTAGGG)n stabilization activities. We find that either RNP11 or the ATP /GTP motifs are necessary and sufficient for G'2 d(CGG)n destabilization whereas RNP21 suppresses destabilization by either one of these two motifs. Neither RNP11 nor the ATP /GTP motif are required for G'2 d(TTAGGG)n stabilization. Hence, CBF-A employs different domains to destabilize G'2 d(CGG)n or stabilize G'2 d(TTAGGG)n. Keywords: INTRODUCTION : Runs of adjacent guanine residues in DNA are capable of self-association to form four-stranded structures termed DNA tetraplexes or quadruplexes. At the core of these DNA secondary structures are Hoogsteen hydrogen-bonded and cation-coordinated stacked guanine quartets (reviewed in ,). Three major classes of tetrahelical DNA are defined by the stoichiometry and orientation of the DNA strands: G'4 unimolecular tetraplexes, G'2 bimolecular tetraplexes, and G4 four-molecular tetraplexes. These types of tetrahelical DNA are further differentiated into sub-groups by parameters such as the glycosidic torsion angles and molecular geometry of the tetrahelix, the nucleotide sequence of non-guanine spacer stretches and their structure, inclusion of bases other than guanine in tetrad structures and the nature of the coordinating cation (,). Although DNA tetraplexes are readily formed in vitro under physiological-like conditions, their existence in vivo still awaits direct demonstration. However, some indirect lines of evidence suggest that tetrahelical DNA might be present in living cells and contribute to diverse physiological and pathological processes. First, biologically important guanine-rich DNA regions fold into tetraplex structures under physiological-like conditions in vitro. It was argued that in vivo formation of tetraplex structures by such sequences might be necessary for the execution of their proposed biological roles. For instance, transient generation of tetraplex structures by the pairing of guanine runs at intra-chromosomal loci was suggested to mediate meiotic pairing of the homolog chromosome . Likewise, folding of the telomeric G-strand into tetraplex formations was proposed to contribute to the regulation of telomere extension . Also, tetrahelical parallel structures of guanine-rich stretches in regions upstream to genes such as c-myc and insulin were implicated in the regulation of their transcription. Lastly, formation of tetraplex structures by a d(CGG) trinucleotide repeat in the FMR-1 gene was suggested to prompt polymerase pausing and slippage and expansion of the repeat sequence that leads to silencing of FMR-1 and sets off fragile X syndrome . A second argument for the existence of tetraplex DNA structures in vivo is the presence of numerous cellular proteins that interact with tetraplex DNA. Proteins isolated from diverse organisms bind preferentially, and at a relatively high affinity, various types of tetraplex DNA ( --). Other proteins were shown to selectively process tetraplex DNA or to modulate its structure. These are nucleases, identified in fission yeast (,), mouse and human cells , that hydrolyze DNA (,) and RNA next to tetraplex domains. Other proteins alter the equilibrium between single-stranded and tetraplex structures of guanine-rich DNA. The beta-subunit of an Oxytricha telomere-binding protein promotes the formation of a tetraplex structure of telomeric DNA (,). Also, several mammalian proteins tightly bind to tetraplex DNA and increase its stability (,,). Lastly, yeast and human helicases of the RecQ family were shown to preferentially unwind tetraplex structures of diverse guanine-rich sequences ( --). In searching for mammalian proteins that interact with tetraplex DNA we identified in rat hepatocytes a protein, designated quadruplex telomeric DNA binding protein 42 (qTBP42), that bound tightly (Kd = 3.7 --14.6 nM) single-stranded and G'4 unimolecular and G'2 bimolecular tetraplex forms of the telomeric sequence d(TTAGGG)n and a G4 four-molecular quadruplex structure of an immunoglobulin switch region sequence . The association of qTBP42 with tetraplex telomeric DNA structures increased their resistance to heat denaturation and diminished their digestion by micrococcal nuclease . Conversely, without detectably binding to it, qTBP42 efficiently destabilized G'2 tetraplex d(CGG)n disrupting this tetrahelix into its constituent single strands . Amino acid sequences of qTBP42 peptides are fully homologous to segments of the CArG-box binding factor A (CBF-A), a heterogeneous nuclear ribonucleoprotein-related protein originally identified as a muscle-specific transcriptional repressor . More recent data suggest that CBF-A might also be involved in transcriptional and post-transcriptional regulation of the expression of diverse genes ( --). Here we show that mouse recombinant CBF-A is physically and immunologically indistinguishable from qTBP42 and that similarly to qTBP42, CBF-A also contrastingly stabilizes tetraplex telomeric DNA while destabilizing tetraplex d(CGG)n. In undertaking to identify domains in CBF-A that mediate G'2 d(CGG)n destabilization or tetraplex telomeric DNA stabilization, we conducted a systematic study of the activities of truncated and deleted CBF-A mutant proteins. We report the identification of distinct domains in CBF-A that prompt or inhibit the destabilization of G'2 d(CGG)n. These regions are dispensable for the binding and stabilization of tetraplex telomeric DNA. MATERIALS AND METHODS : Oligonucleotides | Deoxyoligonucleotides, 5'-tail TeR2, 5'-d[TAGACATG(TT AGGG)2TTA]-3' and 3'-tail d(CGG)7, 5'-d[(CGG)7CGTG GACTC]-3', were synthesized by Operon Technologies and purified by electrophoresis in 8 M urea, 14% polyacrylamide (acrylamide /bisacrylamide 19:1) denaturing gel . Preparation of tetraplex forms of DNA oligomers | Gel-purified single-stranded 5'-tail TeR2 or 3'-tail d(CGG)7 were 5'-end labeled with 32P in a bacteriophage T4 polynucleotide kinase-catalyzed reaction . Generation of bimolecular tetraplex G'2 structures of the oligomers and their isolation by non-denaturing gel electrophoresis were performed as we described previously . The isolated tetraplex DNA was stored at --20C in TE buffer, 100 mM KCl until used. Due to spontaneous dissociation of the tetraplex structures into their constituent single strands, tetraplex forms constituted 80 --90% and 50 --70%, respectively, of the total gel-purified 5'-tail TeR2 and 3'-tail d(CGG)7 DNA. Bimolecular stoichiometry of the tetraplexes was verified as detailed elsewhere . As previously demonstrated for the Hoogsteen hydrogen-bonded tetraplexes of these guanine-rich tracts (, --), both G'2 5'-tail TeR2 and G'2 3'-tail d(CGG)7 DNA structures resisted methylation by dimethylsulfate. Plasmids | A pGEX-A1 plasmid harboring mouse hnRNP A1 cDNA was the generous gift of Dr Benoit Chabot (Universite de Sherbrooke, Canada). Mouse CBF-A cDNA in a pCDL-SR-CBFA plasmid was kindly contributed by Dr T. Miwa (Osaka University, Japan). Using primers with EcoRI ends, the CBF-A cDNA insert was amplified by polymerase chain reaction (PCR) using Pow DNA polymerase (Roche) and the product cDNA was cloned into pGEX-2T. Generation of deletion, substitution and truncation mutations in CBF-A | Deletion or substitution mutations in the CBF-A cDNA insert were generated according to the Quickchange site-directed mutagenesis protocol (Stratagene) by extending over pGEX-2T DNA oligonucleotide primers that flank the deleted sequence or that contain a desired base substitution, respectively. To produce 5'- or 3'-truncation mutations, oligonucleotide primers positioned at the desired ends of the CBF-A cDNA sequence were extended by PCR using Pow polymerase. The 5'-end primers with EcoRI termini had a d(ATG) methionine-encoding start triplet whereas the EcoRI-ended 3'-primers had a d(TAG) termination triplet. Wild-type or mutant plasmid DNA was electroporated into Escherichia coli XL-1-Blue cells (Eppendorf electroporator 2510), the DNA was purified and mutations were validated by direct nucleotide sequencing of the CBF-A encoding cDNA tract. Expression and purification of wild-type and mutant CBF-A proteins | Plasmid DNA harboring wild-type or mutant CBF-A cDNA was transformed by CaCl2 into E.coli BL21(DE3)pLysS cells. The cells were grown at 37C for 2 --3 h to A600 similar0.6 and expression of glutathione-S-transferase (GST)-fused protein was induced by the addition of 100 microM IPTG for an additional 3 --4 h. The induced cells were harvested, washed once with ice-cold STE buffer (100 mM NaCl, 1 mM EDTA, 10 mM Tris --HCl buffer, pH 8.0) and suspended at 4C in 1:50 original volume of TED buffer (10 mM EDTA, 1 mM DTT, 2.5 mM PMSF, 10 microg /ml aprotinin, 10 microg /ml leupeptin, 50 mM Tris --HCl buffer, pH 8.0). All the subsequent steps were conducted at 4C. The cells were disrupted by ultrasonic disintegration (Heat Systems XL sonicator) and the resulting extract was centrifuged at 27 000 g for 35 min. The supernatant fraction (10 ml) was adsorbed onto a 1.0 ml column of glutathione --agarose (Sigma) and after a wash with 10 ml of buffer T (1 mM EDTA, 50 mM Tris --HCl buffer, pH 8.0), the GST --CBF-A fusion protein was eluted from the column by 5 mM of freshly prepared reduced glutathione in buffer T. Following overnight dialysis of the protein solution against buffer D (0.5 mM DTT, 1 mM EDTA, 20% glycerol, 25 mM Tris --HCl buffer, pH 8.0), the GST protein tag was cleaved by incubating the fusion protein at 16C for 18 h with 1 U bovine thrombin (Pharmacia) per 100 microg of protein. Proteins were resolved by SDS --PAGE to verify complete cleavage of GST and to establish that CBF-A and GST constituted more than 95% of the protein mass in the digest. Purification of qTBP42 | qTBP42 was purified to near homogeneity from extracts of rat hepatocytes by successive steps of column chromatography as previously described . The presence of stabilizing soy bean trypsin inhibitor (STI) protein in the highly purified fractions of qTBP42 prevented determination of its protein concentration and the amount of qTBP42 was thus calibrated by its 5'-tail G'2 TeR2 binding activity . Assay of G'2 3'-tail d(CGG)7 destabilization | Measurement of G'2 3'-tail d(CGG)7 destabilization by wild-type or mutant CBF-A or by qTBP42 was conducted by incubating at 37C for 15 min, specified amounts of the purified proteins in 10 microl reaction mixtures that contained 150 --300 fmol of G'2 5'-32P 3'-tail d(CGG)7, 10 mM KCl and buffer D. The tetraplex destabilization reaction was terminated by adding to the mixtures 1% SDS to a final concentration of 0.3%. Intact and unwound G'2 3'-tail d(CGG)7 were resolved from one another by electrophoresis at 4C and 200 --250 V in a non-denaturing 10% polyacrylamide gel in 0.5x TBE buffer, 10 mM KCl, until the bromophenol blue tracking dye migrated 7 --7.5 cm into the gel. The proportion of unwound G'2 3'-tail d(CGG)7 was quantified by phosphorimaging analysis of the dried gel. Assays for G'2 5'-tail TeR2 DNA binding | G'2 5'-tail TeR2 binding to purified qTBP42 or to wild-type or mutant CBF-A proteins was conducted at 4C for 20 min in 10 microl reaction mixtures that contained specified amounts of the respective protein, 150 --300 fmol of 5'-32P G'2 5'-tail TeR2 DNA, and 10 mM KCl in buffer D. Protein --DNA complexes were resolved from unbound DNA by electrophoresis of the directly loaded reaction mixtures through a mobility shift non-denaturing 9% polyacrylamide gel in 0.5x TBE buffer (1.2 mM EDTA in 0.54 Tris-borate buffer, pH 8.3) containing 10 mM KCl. Electrophoresis at 4C and 200 --250 V was terminated after the bromophenol blue tracking dye migrated 7 --7.5 cm into the gel. The formation of a protein --G'2 5'-tail TeR2 DNA complex was quantified by phosphorimaging of the dried gel. Some CBF-A mutant proteins failed to form complexes with labeled tetraplex TeR2 DNA that were detectable by mobility shift analysis. Immunochemical identification of qTBP42 and CBF-A | Polyclonal antiserum against purified recombinant wild-type CBF-A protein was raised in a rabbit by Sigma (Israel). Anti-CBF-A antibodies were adsorbed to protein A /G-agarose (Santa Cruz) by incubating overnight at 4C and under rotation, 50 microl of anti-CBF-A antiserum with 20 microl of a 50% slurry of protein A /G-agarose thrice pre-washed by buffer D, in a final volume of 1 ml of buffer D. Control mixtures contained 50 microl of either pre-immune serum or buffer D. The protein-coated A /G-agarose was washed three times to remove unabsorbed immunoglobulins and 3.9 microg of either purified CBF-A or an equivalent activity of TBP42 in 100 microl of buffer D were added to the mixtures. Immune adsorption was conducted under rotation at 4C for 90 min, the protein A /G-agarose beads were removed by centrifugation and aliquots of the supernatant fraction were assayed for binding of 5'-32P-labeled G'2 5'-tail TeR2 DNA as described above. Immunochemical identification of GST-fused recombinant CBF-A was conducted by the same immunoprecipitation procedure except that anti-GST antibody (Santa Cruz) was employed in place of the anti-CBF-A antiserum. RESULTS : CBF-A destabilizes G'2 d(CGG)n and binds G'2 telomeric DNA | We reported previously that rat liver qTBP42 protein binds tightly and stabilizes tetraplex forms of the telomeric sequence d(TTAGGG)n while it paradoxically destabilizes bimolecular tetraplex structures of the fragile X syndrome expanded sequence d(CGG)n . Amino acid sequences of qTBP42 peptides are fully homologous to fragments of the hnRNP-related CArG-box binding protein CBF-A . We inquired whether, similarly to qTBP42, CBF-A also interacts contrastingly with tetraplex forms of d(CGG)n and d(TTAGGG)n. Mouse CBF-A cDNA , whose length and nucleotide sequence are identical to rat CBF-A cDNA , was expressed in E.coli cells. The recombinant CBF-A protein was purified and assayed for its ability to destabilize or bind and stabilize G'2 bimolecular tetraplex structures of 3'-tail d(CGG)7 and telomeric 5'-tail TeR2 DNA. The results presented in Figure A show that, similarly to qTBP42 , recombinant mouse CBF-A destabilizes G'2 3'-tail d(CGG)7 and that the extent of destabilization is proportional to the amount of protein added. Destabilization of this tetraplex DNA substrate is undetectable in control reaction mixtures that contain GST protein in place of CBF-A (data not shown). Similarly to qTBP42, CBF-A destabilizes G'2 3'-tail d(CGG)7 at equal rates with or without the presence of ATP or Mg2+ ions (data not shown). Furthermore, purified recombinant CBF-A does not appear to contain bound ATP that contributes to its tetraplex DNA disruption activity. CBF-A pre-treated by apyrase (potato ATPase) or incubated in the presence of the ATP-consuming deoxyglucose-hexokinase system maintains 80 and 113%, respectively, of the G'2 3'-tail d(CGG)7 activity of control protein. Replicate titrations of the CBF-A tetraplex DNA destabilizing activity indicate that resolution of 50% of the tetraplex DNA substrate is attained at a 24 --38-fold molar excess of protein over G'2 3'-tail d(CGG)7. As previously reported for qTBP42 , CBF-A destabilizes to a comparable extent G'2 structures of d(CGG)7, 3'-tail d(CGG)7, 5'-tail d(CGG)7 or 3',5'-tail d(CGG)7 (data not shown). In contrast, with d(CGG)n tetrahelices, incubation of CBF-A with a tetraplex structure of telomeric DNA, G'2 5'-tail TeR2, under conditions of G'2 d(CGG)n destabilization, results in protein --DNA complex formation with no detectable tetraplex destabilization. As seen in Figure B, CBF-A and G'2 5'-tail TeR2 DNA form complexes whose amount is proportional to the amount of protein added. No complex is detected in control mixtures that contain GST protein in place of recombinant CBF-A (data not shown). Notably, larger sized complexes are generated in the presence of excess CBF-A protein. A similar association with G'2 telomeric DNA was reported for qTBP42 and the analysis indicated that the larger-sized complexes have a higher molecular ratio of protein to DNA (,). No binding of blunt-ended or tailed G'2 d(CGG)n substrates by CBF-A was detectable under these incubation conditions (data not shown). Multiple titrations of the complex formation between CBF-A and G'2 5'-tail TeR2 indicate that an 18 --27-fold molar excess of protein over DNA is required to bind 50% of the telomeric tetraplex DNA. A heat-resistant CBF-A stabilizes bound G'2 5'-tail TeR2 DNA | qTBP42 is a heat-stable protein whose association with tetraplex telomeric DNA raises the melting temperature of the bound DNA . The results presented in Figure A show that CBF-A and qTBP42 are similarly heat resistant, both maintaining nearly full DNA-binding activity after being incubated for 10 min at 100C. To examine whether similarly to qTBP42, CBF-A protects tetraplex telomeric DNA against heat denaturation, complexes of G'2 5'-tail TeR2 DNA with each protein were incubated at 51C, the protein residue was removed by SDS and the amounts of remaining tetraplex structure were determined by electrophoresis. As seen in Figure B, relative to free tetraplex DNA or to DNA heated in the presence of excess GST protein, the denaturation rate of DNA bound by CBF-A or qTBP42 is diminished. Since the amount of the added qTBP42 could not be determined directly (see Materials and Methods), the heated mixtures contained different amounts of CBF-A or qTBP42G'2. This difference in protein to DNA ratio is the probable cause for the somewhat greater degree of heat protection of the tetraplex DNA afforded by qTBP42 (Fig. B). Two lines of evidence indicate that CBF-A is directly responsible for the observed thermal stabilization of G'2 5'-tail TeR2 DNA. First, the heat stability of tetraplex telomeric DNA is not increased in the presence of carbonic anhydrase, bovine serum albumin, thrombin protease or ovalbumin added at 10-fold greater molecular excess than CBF-A (data not shown). Secondly, removal of GST-fused CBF-A by immune precipitation with anti-GST specifically obliterates the thermal protection of the DNA tetraplex (data not shown). Anti-CBF-A antiserum recognizes qTBP42 | To confirm synonymy of qTBP42 and CBF-A we inquired whether anti-mouse CBF-A rabbit antibodies bind rat qTBP42. The results presented in Figure demonstrate that whereas pre-immune rabbit serum precipitates neither CBF-A nor qTBP42, both proteins are equally bound by the anti-CBF-A antiserum. The cross-antigenicity of qTBP42 and CBF-A, as well as their sequence homology, similar heat stability and comparable abilities to destabilize G'2 d(CGG)n and to bind and stabilize G'2 TeR2 DNA, establish their identity. It is highly likely, therefore, that rat qTBP42 and mouse CBF-A mediate their differential interaction with tetraplex d(CGG)n and d(TTAGGG)n through the same protein domains. The RNP11 and ATP /GTP-binding motifs of CBF-A are specifically required for tetraplex d(CGG)n destabilization | Mutant CBF-A proteins were expressed in E.coli cells that have truncated N- or C-terminal regions or that contain deletions in one or more of the evolutionary conserved and duplicated RNP1 or RNP2 boxes (,) or in the putative ATP /GTP-binding domain (, --). The purified proteins were assayed for their capacity to destabilize G'2 3'-tail d(CGG)7. Figure shows a typical analysis of G'2 3'-tail d(CGG)7 destabilization by wild-type and representative mutant CBF-A proteins. Whereas G'2 3'-tail d(CGG)7 is fully resolved by wild-type CBF-A, no destabilization is detectable in mixtures that contain equal amounts of mutant proteins with either deleted RNP11 and RNP12 boxes (mutant DeltaR11R12) or truncated C-terminal tract that encompasses the ATP /GTP-binding motif (mutant 1 --260). Notably, recombinant wild-type hnRNP A1 is also unable to unwind G'2 3'-tail d(CGG)7 , indicating that the tetraplex d(CGG)n unwinding capacity of CBF-A is not shared by every hnRNP species. The schemes of all the CBF-A mutants examined and their relative G'2 3'-tail d(CGG)7 destabilization activities are shown in Figure . Our results indicate that the RNP11 box or the ATP /GTP-binding element are required for destabilization of G'2 3'-tail d(CGG)7. Deletion of residues 116 --122 from the RNP11 motif (CBF-A mutant DeltaR11) inactivates G'2 3'-tail d(CGG)7 destabilization, implying that this element is essential for the unwinding activity . This proposition is supported by the similar loss of unwinding activity by CBF-A double mutants DeltaR11R12 and DeltaR11R22 that, in addition to RNP11, lack the RNP12 or RNP22 boxes which are dispensable for tetraplex destabilization (see below). Parallel data indicate that the destabilization activity is also dependent on the presence of an intact ATP /GTP-binding motif. The absence of this element in CBF-A Delta266 --269 or in the truncation mutant 1 --260 results in failure to destabilize G'2 3'-tail d(CGG)7 . Interestingly, truncation that trims a single serine residue at the C-terminal end of the ATP /GTP motif (position 272; mutant 1 --271) causes loss of similar50% of the G'2 3'-tail d(CGG)7 destabilization activity . Lastly, substituting a single tyrosine residue by lysine at position 267 within the ATP /GTP box (T267K mutation) also results in loss of destabilization activity . Put together, these data substantiate the required role of the conserved ATP /GTP box in G'2 3'-tail d(CGG)7 disruption. The evidence summarized in Figure indicates that in contrast to the essential RNP11 and ATP /GTP-binding motifs, the N-terminal sequence of CBF-A and its RNP21, RNP22 and RNP12 boxes are expandable for G'2 d(CGG)n destabilization activity. CBF-A is distinguished from other hnRNP homologs by its unique N-terminal acidic stretch of 75 amino acids (,). As seen in Figure , truncation of an 83 amino acid N-terminal tract (mutant 84 --285) does not affect G'2 3'-tail d(CGG)7 destabilization, indicating the dispensability of this sequence for tetraplex disruption activity. Most of the G'2 3'-tail d(CGG)7 destabilization activity is also maintained by mutant proteins with residues deleted from the RNP21, RNP22 or RNP12 motifs as well as by proteins that contain double mutations; DeltaR21R22 and DeltaR22R12 . These results demonstrate that RNP21, RNP22 and RNP12 are also superfluous for tetraplex d(CGG)n disruption. The RNP21 box represses G'2 d(CGG)n destabilization by the RNP11 or ATP /GTP-binding motifs | Inspection of the activities of some of the compound CBF-A mutant proteins listed in Figure reveals that the RNP21 motif inhibits tetraplex d(CGG)n disruption mediated by the RNP11 element or the ATP /GTP-binding box. Deletion of the RNP11 motif in CBF-A mutants DeltaR11, DeltaR11R22 or DeltaR11R12, results in loss of G'2 3'-tail d(CGG)7 unwinding activity despite the presence of an intact ATP /GTP-binding motif . Concordantly, the failure of CBF-A mutants Delta266 --269 and 1 --260 to destabilize G'2 3'-tail d(CGG)7 shows that tetraplex disruption is inactivated by the absence of an ATP /GTP-binding motif despite the presence of an undamaged RNP11 box. The vigorous destabilization activities of wild-type CBF-A and of the mutant proteins DeltaR22, DeltaR12 and DeltaR2212 indicate, however, that tetraplex disruption is permitted in the presence of both the RNP11 and ATP /GTP-binding motifs . At the same time, data show that removal of the RNP21 motif re-establishes G'2 d(CGG)n activity in mutant proteins that contain either an RNP11 motif but no ATP /GTP-binding element or an ATP /GTP box without RNP11. The nearly full disruption of G'2 3'-tail d(CGG)7 by CBF-A mutants DeltaR21R11, DeltaR21R11R22 or DeltaR21R11R22R12 indicates that deletion of the RNP21 box restores destabilization activity to proteins that contain only an ATP /GTP-binding element without an RNP11 box . Correspondingly, the full disruptive activity of CBF-A mutant protein DeltaR21 indicates that deletion of the RNP21 motif confers tetraplex d(CGG)n destabilization activity to a mutant protein that has an RNP11 box but no ATP /GTP-binding motif . Put together, these results indicate that in the absence of the RNP21 motif, RNP11 or the ATP /GTP-binding box are capable of mediating disruption of G'2 d(CGG)n independently of each other. Since the RNP21 box represses the activity of either one of these elements but is unable to simultaneously inhibit both, the presence of both the RNP11 and ATP /GTP-binding motifs secures destabilization activity despite the existence of an operational RNP21 inhibitory element. A tentative model that satisfies these observations is presented in the Discussion. Binding and stabilization of G'2 TeR2 DNA do not require intact RNP11 motif or ATP /GTP-binding element | Parallel to possessing tetraplex d(CGG)n destabilizing capacity, CBF-A and qTBP42 are distinguished by their ability to bind and stabilize tetraplex telomeric DNA. We examined whether or not the RNP11 motif or the ATP /GTP-binding box that are required for G'2 d(CGG)n disruption, also mediate G'2 5'-tail TeR2 DNA binding and stabilization. The results summarized in Table indicate that tetraplex telomeric DNA is bound to a similar extent by recombinant wild-type CBF-A or by the DeltaR11 and 1 --260 mutant proteins that lack these two domains. Likewise, wild-type CBF-A and the two mutant proteins protect to a similar extent G'2 5'-tail TeR2 against heat denaturation . It appears, therefore, that the RNP11 or the ATP /GTP-binding motifs are not involved in the binding and stabilization of tetraplex telomeric DNA. Figure 1 | CBF-A destabilizes bimolecular tetraplex G'2 3'-tail d(CGG)n and binds bimolecular tetraplex G'2 5'-tail TeR2. CBF-A destabilizes bimolecular tetraplex G'2 3'-tail d(CGG)n and binds bimolecular tetraplex G'2 5'-tail TeR2. (A) Destabilization of G'2 3'-tail d(CGG)7 by CBF-A. Increasing amounts of purified recombinant CBF-A, 5 --2000 ng were incubated at 37C for 15 min with 5'-32P G'2 3'-tail d(CGG)7 under standard tetraplex DNA destabilization assay conditions (see Materials and Methods). Control G'2 3'-tail d(CGG)7 with no added protein was either boiled for 10 min to denature the tetraplex DNA structure or was incubated at 37C under standard tetraplex DNA destabilization conditions to reveal the initial amount of unwound G'2 3'-tail d(CGG)7. The reaction mixtures were cooled to 4C and 1% SDS was added to a final concentration of 0.3% to denature DNA-bound protein. Intact and destabilized G'2 3'-tail d(CGG)7 were resolved by electrophoresis at 4C in 10% polyacrylamide gels in 0.5x TBE buffer, 10 mM KCl. A phosphorimage of the electropherogram is shown. (B) Binding of 5'-tail TeR2 DNA by CBF-A. Increasing amounts of CBF-A, 5 --600 ng, were incubated at 4C for 20 min with 5'-32P G'2 5'-tail TeR2 DNA under standard tetraplex DNA-binding conditions (see Materials and Methods). Control G'2 5'-tail TeR2 DNA without added protein was either boiled for 10 min to denature the tetraplex DNA structure or was incubated at 4C under standard tetraplex DNA-binding conditions to reveal the initial amount of unbound G'2 5'-tail TeR2 DNA. A phosphorimage of a 9% polyacrylamide, 0.5x TBE buffer, 10 mM KCl gel is shown. Figure 2 | CBF-A and qTBP42 are heat-stable proteins that slow down denaturation of associated G'2 5'-tail TeR2 DNA. CBF-A and qTBP42 are heat-stable proteins that slow down denaturation of associated G'2 5'-tail TeR2 DNA. (A) Heat stability of CBF-A and qTBP42. Purified CBF-A at 100 microg /ml or qTBP42 protein to which bovine serum albumin was added to the same concentration, were incubated, each in duplicate, at temperatures ranging from 4 to 100C for 10 min in a final volume of 10 microl of buffer D. The samples were rapidly cooled to 4C and aliquots were assayed under standard conditions for the binding of 90 fmol of 5'-32P G'2 5'-tail TeR2 DNA. Protein --DNA complexes were resolved by electrophoresis at 4C in 8% polyacrylamide gels, 0.5x TBE buffer, 10 mM KCl. Amounts of protein-bound 5'-32P G'2 5'-tail TeR2 DNA were quantified by phosphorimaging. Relative binding of 100% represents association of CBF-A or qTBP42 with 85.5 --89 fmol of tetraplex DNA. Closed circles, qTBP42; open circles, CBF-A. (B) Binding of CBF-A and qTBP42 slows down the denaturation of G'2 5'-tail TeR2 DNA. DNA, 26 fmol of 5'-32P G'2 5'-tail TeR2, was incubated at 4C for 20 min in 10 microl of buffer D, 10 mM KCl reaction mixtures that contained no protein or 167 pmol GST, 13 pmol of CBF-A, or qTBP42 (amount not determined, see Materials and Methods). The mixtures were transferred to 51C for the indicated periods of time, rapidly cooled to 4C and protein was removed from the DNA by adding 1% SDS to a final concentration of 0.3%. Intact and denatured G'2 5'-tail TeR2 DNA were resolved from one another by electrophoresis at 4C in 10% polyacrylamide gels, 0.5x TBE buffer, 10 mM KCl. Amounts of G'2 and single-stranded DNA were quantified by phosphorimaging. A value of 100% G'2 5'-tail TeR2 DNA represents 57% G'2 out of the total 5'-tail TeR2 DNA in samples that were kept at 4C. Closed circles, qTBP42; open circles, CBF-A; open boxes, GST; closed boxes, no protein. Figure 3 | Anti-CBF-A antiserum recognizes TBP42. Anti-CBF-A antiserum recognizes TBP42. Purified CBF-A or qTBP42 proteins were incubated at 4C for 90 min with rabbit anti-CBF-A antibodies adsorbed to protein A /G-agarose (see Materials and Methods). Control mixtures contained pre-immune rabbit serum adsorbed to protein A /G-agarose or protein A /G-agarose with no adsorbed protein. The protein A /G-agarose beads were removed by centrifugation and aliquots of the supernatant fractions were assayed for binding of 5'-32P G'2 5'-tail TeR2 DNA as described in Materials and Methods. Phosphorimages of protein-G'2 5'-tail TeR2 complexes resolved from unbound DNA by electrophoresis through non-denaturing 9% polyacrylamide mobility shift gels are shown. Figure 4 | Destabilization of G'2 3'-tail d(CGG)7. Destabilization of G'2 3'-tail d(CGG)7. Wild-type CBF-A or hnRNP A1 proteins, at 33 pmol each or 165 pmol each of R11R12 or 1 --260 CBF-A mutant proteins were incubated at 37C for 15 min with 0.16 pmol of 5'-32P G'2 3'-tail d(CGG)7 under standard tetraplex DNA destabilization assay conditions. Control samples that did not contain protein were similarly incubated at 37 or 100C to reveal the initial amount of G'2 3'-tail d(CGG)7 and the position of displaced single-stranded DNA, respectively. The tetraplex unwinding reaction was terminated by the addition of 1% SDS to a final concentration of 0.3%, and tetraplex and single-stranded 3'-tail d(CGG)7 were resolved by electrophoresis through a 10% polyacrylamide gel, 0.5x TBE buffer, 10 mM KCl. A phosphorimage of the electropherogram is shown. Figure 5 | CBF-A mutant proteins and their G'2 3'-tail d(CGG)7 destabilization activity. CBF-A mutant proteins and their G'2 3'-tail d(CGG)7 destabilization activity. The conserved motifs marked in the scheme of wild-type CBF-A are: RNP21, MFVGGL; RNP11, SRGFGFIL; RNP22, IFVGGL; RNP12, RRGFVFIT; the putative ATP /GTP-binding domain, GSTNYGKS. The upper numbers in the wild-type protein scheme denote positions of residues at the boundaries of the conserved domains whereas the lower italicized numbers indicate the boundaries of segments deleted from each motif. Schemes show segments truncated or deleted in each mutant protein. The presented relative G'2 3'-tail d(CGG)7 destabilization activity of each mutant protein is an average of the indicated number of independent determinations. Figure 6 | CBF-A mutant T267K fails to unwind G'2 3'-tail d(CGG)7. CBF-A mutant T267K fails to unwind G'2 3'-tail d(CGG)7. Wild-type or T267K variant CBF-A protein mutated in the ATP /GTP-binding motif at 48 pmol each, were incubated at 37C for 15 min with 0.17 pmol of 5'-32P G'2 3'-tail d(CGG)7 under standard tetraplex DNA destabilization assay conditions. Incubation of control samples with no added protein, termination of the destabilization reaction and electrophoretic resolution of intact and destabilized G'2 3'-tail d(CGG)7 were conducted as detailed in the legend to Figure . Quantification of phosphorimages of four independent determinations reveal that wild-type and T267K CBF-A proteins destabilized 53 +- 3 and 4 +- 3%, respectively, of the tetraplex substrate. Table 1 | Deletion of the RNP11 box or the ATP /GTP-binding motif does not affect binding or stabilization of G'2 5'-tail TeR2 DISCUSSION : Tetrahelical structures of the FMR1 gene d(CGG)n tract have been suggested to promote polymerase slippage and expansion of the trinucleotide repeat sequence that silences FMR1 and sets off fragile X syndrome . Disruption of d(CGG)n tetraplexes could possibly lower the probability of expansion of this sequence. Proteins such as the Werner syndrome DNA helicase (,), qTBP42 (,) and uqTBP25 (,) were found to unwind G'2 bimolecular tetraplex forms of d(CGG)n. In this work, we undertook to unravel the protein structure basis for the G'2 d(CGG)n destabilization activity of qTBP42 /CBF-A. The results described here have two major corollaries: First, the hnRNP-related rat protein qTBP42 (,,) and the transcriptional regulator CBF-A ( --) are indistinguishable. Secondly, two conserved domains of CBF-A, the RNP11 box and the putative ATP /GTP-binding fold mediate destabilization of tetraplex d(CGG)n whereas the RNP21 motif acts to suppress this activity. Rat liver qTBP42 is distinguished by its tight binding and stabilization of tetraplex forms of the telomeric sequence d(TTAGGG)n (,,) and by its contrasting capacity to destabilize G'2 tetraplex structures of d(CGG)n . Based on peptide sequence homology of qTBP42 and CBF-A, we inquired whether physical --chemical properties of recombinant mouse CBF-A and its differential interaction with G'2 tetraplex structures of d(TTAGGG)n and d(CGG)n resemble those of qTBP42. Our results show that, similarly to qTBP42, CBF-A also disrupts G'2 bimolecular d(CGG)n tetraplexes while binding G'2 d(TTAGGG)n . Furthermore, CBF-A resembles qTBP42 in being a heat-stable protein that also slows down the heat denaturation of bound tetraplex telomeric DNA . Lastly, the identity of the two proteins is corroborated by the specific precipitation of rat qTBP42 by anti-mouse CBF-A antibodies . We identified domains in CBF-A that mediate G'2 d(CGG)n destabilization by determining this activity in an array of CBF-A mutant proteins. Although CBF-A is distinguished from other hnRNP-related homologous proteins by its unique acidic N-terminal stretch of 75 amino acids (,), results show that this domain is not required for G'2 3'-tail d(CGG)7 destabilization as evidenced by the full unwinding activity of the N-terminal truncated mutant protein 83 --285 . The nearly full G'2 3'-tail d(CGG)7 disruption activity of CBF-A mutant proteins with deletions in the RNP21, RNP22 or RNP12 boxes or combinations thereof, indicates that these elements are also expendable for tetraplex destabilization . In contrast, the RNP11 box and /or the ATP /GTP-binding elements are required for G'2 d(CGG)n disruption. CBF-A mutants that have residues deleted from their RNP11 motif alone or in combination with the dispensable RNP12 or RNP22 boxes lack unwinding activity . Likewise, CBF-A mutant proteins with truncated or deleted ATP /GTP-binding motifs , or with amino acid substitution within this domain (mutant T267K, Fig. ) fail to disrupt G'2 d(CGG)n. Notably, the putative ATP /GTP-binding fold, solely assigned by amino acid sequence homology, has no known function (, --). The participation of this element in tetraplex DNA destabilization is, to the best of our knowledge, its first identified activity. However, since disruption of G'2 d(CGG)n by CBF-A does not require nucleotide triphosphates (see Results), the involvement of this domain in tetraplex disruption is probably unrelated to its presumed binding of ATP or GTP. Surprisingly, we find that the RNP21 motif acts to suppress tetraplex d(CGG)n destabilization. Analysis of multiple mutant proteins indicates that whereas the disruption of G'2 d(CGG)n by intact CBF-A requires the presence of both the RNP11 motif and an ATP /GTP-binding element, removal of the inhibitory RNP21 box allows mutant proteins that lack either RNP11 or ATP /GTP-binding box to conduct tetraplex unwinding by the single remaining element . Which then is the domain in wild-type CBF-A that is active in G'2 d(CGG)n disruption and which is inactivated by the suppressor RNP21 box? Two alternative modes of destabilization of G'2 d(CGG)n by intact CBF-A protein are modeled in Figure . In one modality the RNP21 box blocks RNP11, leaving an uninhibited ATP /GTP-binding motif to conduct tetraplex destabilization (Fig. A). In an alternative mode RNP21 obstructs the ATP /GTP-binding element and allows the unblocked RNP11 to mediate disruption of G'2 d(CGG)n (Fig. B). A third possibility is that the two depicted conformations of CBF-A are in equilibrium and either the RNP11 motif or the ATP /GTP-binding box operates within any individual protein molecule to destabilize G'2 d(CGG)n. However, the known mode of interaction of hnRNP proteins with single-stranded DNA or RNA favors the view that the ATP /GTP-binding box is the primary domain responsible for tetraplex unwinding by wild-type CBF-A. Proteins of the hnRNP super family possess two RNA recognition motifs (RRMs). The crystal structure of hnRNPA1 reveals that RRM1, formed by the coupling of RNP21 to RNP11, and RRM2, consisting of RNP22 and RNP12, are held in close contact in an anti-parallel orientation (,). Several closely related hnRNPs, including CBF-A /qTBP42, bind tightly the single-stranded telomeric repeat d(TTAGGG)n (,, --). The crystal structure of a complex of d(TTAGGG)2 with a dimer of the N-terminal proteolytic fragment of hnRNPA1 (protein UP1) reveals that each of the two DNA strands is in contact with RRM1 of one UP1 monomer and with RRM2 of the other monomer . It appears, therefore, that when they interact with telomeric DNA, RNP21 and RNP11 are coupled together in RRM1. Conceivably, the interaction of CBF-A with G'2 d(CGG)n is also mediated by its RRM1 and RRM2 domains. If this is the case, the RNP11 box of the wild-type CBF-A is coupled within an RRM1 element to the RNP21 suppressor motif that renders it inactive while allowing the uninhibited ATP /GTP box to mediate disruption of tetraplex d(CGG)n (Fig. A). However, with the inhibitory RNP21 element deleted, no RRM1 is formed and the unsuppressed RNP11 takes over the function of unwinding G'2 d(CGG)n when the ATP /GTP-binding motif is absent . Whereas motifs in CBF-A that contribute positively or negatively to the disruption of tetraplex d(CGG)n were pinpointed, CBF-A mutants lacking either RNP11 or the ATP /GTP-binding motif maintain full capacity to bind and stabilize tetraplex telomeric DNA. Hence, specific CBF-A domains that are responsible for the binding and stabilization of G'2 d(TTAGGG)n remain to be identified. Notably, analysis of co-crystals of UP1 and single-strand telomeric DNA reveals multiple contacts between the protein and the d(TTAGGG)2 sequence . Amino acids within and outside the RNP elements interact with bound DNA through hydrogen bonds and charge and van der Waals interactions . If a similar large number of amino acids in CBF-A interact with tetraplex telomeric DNA it is plausible that the binding and stabilization of the tetrahelical telomeric sequence by CBF-A can be impaired only by the removal of amino acids at multiple sites. In the past we speculated that qTBP42 might act in vivo bi-functionally by stabilizing desirable tetrahelices, such as perhaps at telomere ends, and destabilizing unwanted tetrahelices such as G'2 d(CGG)n structures that might be generated during DNA transactions . Interestingly, additional hnRNP-related proteins were also found to interact differentially with tetraplex DNA structures. These are uqTBP25, which binds unimolecular and bimolecular tetraplexes of the telomeric sequence d(TTAGGG)n but destabilizes G'2 d(CGG)n (,) and hnRNP D, which binds tightly the telomeric tract and destabilizes its G --G paired structures . Since hnRNPs play major roles in RNA metabolism, it is highly likely that the observed interactions with tetraplex DNA reflect functions of hnRNPs in stabilizing or disrupting tetraplex structures in specific RNA molecules. Interestingly, recent results indicate that the FMR1 protein binds preferentially tetraplex-forming mRNA molecules, modulating their association with polysomes ( --). A tempting conjecture is that by differentially modulating the stability of tetraplex domains in different mRNA molecules, proteins such as CBF-A, uqTBP25 or hnRNP D might regulate their nuclear -- cytoplasm translocation or control their translation. Figure 7 | Models of two alternative modes of tetraplex d(CGG)n destabilization by CBF-A. Models of two alternative modes of tetraplex d(CGG)n destabilization by CBF-A. CBF-A, schematically presented in the upper diagram with its conserved RNP boxes and ATP /GTP-binding motif, might assume either one of the two illustrated alternative conformations that conduct G'2 d(CGG)n destabilization through either RNP11 or the ATP /GTP-binding motif. (A) G'2 d(CGG)n unwinding mediated by the ATP /GTP-binding motif. The RNP11 box is blocked by the inhibitory RNP21 element (horizontal hatches) whereas the ATP /GTP-binding motif (dashed circle) remains available to conduct G'2 d(CGG)n destabilization. (B) RNP11-mediated G'2 d(CGG)n unwinding. The ATP /GTP-binding motif is blocked by the inhibitory RNP21 element (vertical hatches) whereas the RNP11 box (dashed circle) remains available to conduct G'2 d(CGG)n destabilization. Details of the models are considered in the Discussion. Backmatter: PMID- 12202773 TI - Promiscuous methylation of non-canonical DNA sites by HaeIII methyltransferase AB - The cytosine C5 methyltransferase M.HaeIII recognises and methylates the central cytosine of its canonical site GGCC. Here we report that M.HaeIII can also, with lower efficiency, methylate cytosines located in a wide range of non-canonical sequences. Using bisulphite sequencing we mapped the methyl- cytosine residues in DNA methylated in vitro and in vivo by M.HaeIII. Methyl-cytosine residues were observed in multiple sequence contexts, most commonly, but not exclusively, at star sites (sites differing by a single base from the canonical sequence). The most frequently used star sites had changes at positions 1 and 4, but there is little or no methylation at star sites changed at position 2. The rate of methylation of non-canonical sites can be quite significant: a DNA substrate lacking a canonical site was methylated by M.HaeIII in vitro at a rate only an order of magnitude slower than an otherwise identical substrate containing the canonical site. In vivo methylation of non-canonical sites may therefore be significant and may have provided the starting point for the evolution of restriction --modification systems with novel sequence specificities. Keywords: INTRODUCTION : DNA cytosine C5 methyltransferases are a large group of enzymes whose structure is conserved in bacteria, archaea and eukaryotes. This class of enzymes has a common mechanism of catalysis and 10 highly conserved motifs making up the catalytic site and the cofactor binding site . In eukaryotes, cytosine C5 methyltransferases are responsible for CpG methylation, which is involved in regulation of chromatin structure and gene expression . In bacteria, DNA methylation has roles in restriction --modification (R /M) systems, the control of DNA replication and post-replicative mismatch repair . Bacteria possessing a R /M system express a methylase and cognate restriction endonuclease which recognise the same target site. Methylation of the bacterial chromosome protects it from endonuclease digestion, but unmethylated DNA of infecting bacteriophage is digested by the endonuclease. Type II R /M systems are the most common, with over 200 DNA specificities . The evolution of a R /M system with a novel specificity must necessarily involve the co-evolution of both a methylase and a restriction endonuclease; however, little is known about how this might occur ( --). Both a methyltransferase and a restriction enzyme with novel sequence specificity must evolve, yet a simultaneous change in sequence specificity is not likely. Due to the lethal nature of a novel restriction enzyme without a cognate methyltransferase, the methyltransferase must change its specificity first. Several models for this co-evolution have been suggested, including duplication of the methyltransferase gene and its cognate restriction endonuclease gene (the two genes are frequently in close proximity) followed by changes in sequence specificity. This switch in specificity may start with the enhancement of activity towards a non-canonical site against which the enzyme already has low activity (,), as proposed for other enzymes (,). Star activity of restriction endonucleases, when the enzyme recognises and cuts at a sequence differing at only a single base pair from the canonical site, is a relatively well documented phenomenon (,). In contrast, DNA methylation is generally regarded as being highly sequence specific , with few published observations of non-canonical methylation. However, non-canonical, or aberrant, methylation has been observed with DNA methyltransferases at sites closely related to the normal target sequence (,, --). Here we report an investigation into methylation of non-canonical sites by the cytosine C5 methyltransferase M.HaeIII which is part of a typical type II R /M system in the bacterium Haemophilus influenzae (biogroup aegyptius). M.HaeIII methylates the third base of the sequence GGCC in unmodified or hemimethylated duplex DNA. We have investigated the optimum conditions for methylation by M.HaeIII at canonical and non-canonical sites and examined the sequence preferences of M.HaeIII for non-canonical sites. We observe that M.HaeIII can methylate cytosines in a variety of sequence contexts both in vitro and in vivo. MATERIALS AND METHODS : Oligonucleotides | Haesubplus: 5'-TTCGAGAAGCTGAGGCCGGCGTACC TGGAG-3' Haesubminus: 5'-CTCCAGGTACGCCGGCCTCAGCTTC TCGAA-3' Starsubplus: 5'-TTCGAGAAGCTGAAGCCGGCGTACC TGGAG-3' Starsubminus: 5'-CTCCAGGTACGCCGGCTTCAGCTTC TCGAA-3' GST1Fo: 5'-GAAATAGTTATGATTATGATTATGT TAAGT-3' GST1Ba: 5'-TCTTTCTTATTATTCCACAACACCA TATAC-3' GST2Fo: 5'-TAATTACCCTAAATTATTAAAACAT CCACA-3' GST2Ba: 5'-GTTGTTTTGTTGTTTGGGGTAGTTG GGGAT-3' Pet1Fo: 5'-GGGAGACCATGGTTTTGATTTTTTA TGTTGGATGTATTG-3' Pet1Ba: 5'-CACAACGAGCTCAATAATCAACCCA CTAACACATTACAC-3' Pet2Fo: 5'-GGGAGACCATGGTCTCAATCCTCTA CACCAAACACATC-3' Pet2Ba: 5'-CACAACGAGCTCGATGTGTTGTGTG AGAAGATTGTGTATTGTTG-3' cM1-Clon-F: 5'-ATCGATCGGTACCTTATTACTCTTT CTTGTTGTTCCACAGCGC-3' cM1-Clon-B: 5'-TATCCTGAGCCATGGTGGTTACCCT GGGTTATTG-3' Re-N-Flag-Bc: 5'-GCTGACAAGCTTAATAATTTTGTTT AACTTTAAGAAGGAGATATAGCCATG GACCTACAAGATGACGATGATAAAA ATTTAATTAGTCTTT-3' In vitro methylation with S-adenosyl-l-[methyl-3H]methionine | Complementary oligonucleotides Haesubplus and Haesub minus were annealed by heating equal amounts (0.5 mM each in 10 mM Tris --HCl, pH 8.5) to 90C for 5 min and then cooling slowly to 20C to create a 30 bp template for methylation, termed Haesub (see Fig. A), containing a single canonical M.HaeIII methylation site (GGCC). Similarly, the DNA oligonucleotides Starsubplus and Starsubminus were annealed to create a second 30 bp template for methylation, termed Starsub (see Fig. A), which has no canonical HaeIII sites and differs from Haesub by only a single base pair. Incorporation of methyl-3H into the two DNA substrates by M.HaeIII was assayed. Reactions contained 10 nM M.HaeIII (New England Biolabs), 200 nM S-adenosyl-l-[methyl-3H]methionine (79.8 Ci /mmol; Perkin-Elmer) and 1.25 microM double-stranded DNA in 30 microl of a standard buffer (25 mM Tris --HCl pH 7.5, 25 mM NaCl, 10 mM dithioerythritol, 0.2 mg /ml BSA). In parallel reactions the concentration of NaCl was varied in the range 0 --100 mM, Tris --HCl in the range 5 --100 mM and glycerol in the range 0 --50%, keeping the other two variables constant (see Fig. ). Reactions were performed in triplicate, incubated at 37C and stopped by the addition of 10 microl of 0.3 M sodium acetate and 320 microM unlabelled S-adenosyl-l-methionine. The rate of incorporation of methyl-3H into Haesub was linear for at least 20 min and the rate of incorporation into Starsub was linear for 2 h (data not shown). To measure the initial rate, reactions were quenched after 10 min (Haesub) or after 40 min (Starsub). Thirty-five microlitres of the quenched reaction was spotted onto a Whatman DE81 filter. Filters were washed four times with 2 ml of 0.2 M ammonium hydrogen carbonate, twice with 1 ml of ethanol, air dried and counted using a Beckmann LS6000SC scintillation counter in 3 ml of scintillant (National Diagnostics). The concentration of S-adenosylmethionine in the assays (200 nM) was lower than the Km of M.HaeIII for S-adenosylmethionine (336 nM; H.M.Cohen, unpublished results); however, the use of low concentrations of cofactor with high specific radioactivity was necessary to allow sensitive detection of methyl group incorporation. Determining the methylation pattern of DNA in vitro | The chicken glutathione-S-transferase M1-1 gene was amplified from chicken liver cDNA (Clontech) using primers cM1-Clon-F and cM1-Clon-B, digested with NcoI and KpnI and ligated into pGEM4z (Promega). PCR amplification of this plasmid, using the standard --21M13 forward and M13 reverse sequencing primers, yielded an 832 bp template for methylation. The methylation reaction contained 100 nM M.HaeIII (New England Biolabs), 73 nM PCR fragment (purified using the Wizard PCR purification kit; Promega), and 80 microM S-adenosyl methionine in 50 microl of buffer (25 mM NaCl, 25 mM Tris --HCl pH 8.5, 5 mM DTT, 40% glycerol). The reaction was incubated for 8 h at 37C, with an additional 80 microM S-adenosyl methionine added after 4 h. The DNA was purified (Wizard PCR purification kit), treated with bisulphite and sequenced as described below. An unmethylated sample of the PCR fragment was treated with bisulphite in a parallel reaction. Determination of the methylation pattern of DNA in vivo | The M.HaeIII gene was amplified by PCR from H.influenzae (biogroup aegyptius) (ATCC 11116) using oligonucleotides HaeIII-FoNC and HaeIII-Bc , digested with KpnI and HindIII and ligated into pGEM4Z (Promega). The gene was amplified from this plasmid by PCR using oligonucleotides Re-N-Flag-Bc and LMB2-Nest , digested with NcoI and SacI, ligated into pET30a (Invitrogen) creating pET-MHaeIII, and transformed into Escherichia coli C41(DE3) . The expressed protein has N-terminal His6, S and Flag tags. Cells were grown in 2x YT at 37C to OD 0.6, induced by the addition of 0.3 mM isopropyl-beta-d-thiogalactoside and transferred to 30C for 4 h. M.HaeIII was purified from the cells using NiNTA agarose (Qiagen) according to the manufacturer's instructions. pET-MHaeIII plasmid was also prepared from the induced cells using the Qiagen Plasmid Midi-prep kit. Fifty-six nanograms of plasmid was linearised by digestion with SacI, purified (Qiaquick PCR purification kit; Qiagen), treated with bisulphite and sequenced as described below. pET30a (containing no methyltransferase gene) was prepared from E.coli C41(DE3) after induction, as above, for use in a parallel bisulphite reaction. Bisulphite sequencing | Purified methylated DNA was treated with bisulphite as described . Strand 1 of the bisulphite-converted GST gene was amplified using primers GST1Fo and GST1Rev, strand 2 was amplified with GST2Fo and GST2Rev. PCR products were cloned using the TOPO TA Cloning kit (Invitrogen) and sequenced. Strand 1 of the bisulphite-converted pET plasmid was amplified using Pet1Fo and Pet1Rev and strand 2 was amplified using Pet2Fo and Pet2Rev. PCR products were digested with NcoI and SacI, cloned in pIVEX2.2 (Roche) and sequenced. Figure 1 | In vitro methylation rates of substrate DNA containing canonical and non-canonical sites by M.HaeIII. In vitro methylation rates of substrate DNA containing canonical and non-canonical sites by M.HaeIII. Two 30 bp substrates were used (A), differing by only a single base pair: Haesub contained a single canonical site (GGCC), shown in bold; and Starsub differed at one base pair, shown in bold, and lacked canonical sites. Initial rates of incorporation of methyl-3H into Haesub (continuous line, squares) and Starsub (dashed line, triangles) at various concentrations of glycerol (B), NaCl (C) and Tris --HCl (D). Measurements were taken in triplicate and error bars represent one standard error of the mean. RESULTS : The relative rates of methylation of two substrates, Haesub, which contains a single canonical M.HaeIII methylation site (GGCC), and Starsub, which has no canonical HaeIII sites and differs from Haesub by only a single base pair (Fig. A), were determined under a variety of different conditions in vitro . It is well known that the star activity of restriction enzymes is greatly enhanced under conditions of low ionic strength or high glycerol concentrations; therefore, we investigated whether the star activity of M.HaeIII could be enhanced under similar conditions. Under all conditions tested, the rate of methylation of the non-canonical substrate, Starsub, was significant, ranging between 2.8 and 50.8% of the rate with the canonical substrate, Haesub, under the same conditions . Both the absolute rates and relative rates of methylation of the two substrates were dependent on the reaction conditions. The fastest rate of methylation of Haesub is achieved in 25 mM NaCl and 25 --50 mM Tris --HCl. The reaction is inhibited by glycerol, with an almost 4-fold reduction in rate caused by the addition of glycerol to 50% (Fig. B). Unlike Haesub the rate of methylation of Starsub is increased by the addition of glycerol (Fig. B), and is higher in 100 mM NaCl than at lower salt concentrations (Fig. C). The dependence on Tris --HCl concentration mirrors that of HaeSub (Fig. D) and the optimum pH for methylation of both substrates is between 7.0 and 8.0 (data not shown). Under the optimal conditions for methylation of Haesub the substrate lacking a canonical site, Starsub, is methylated 35.7-fold more slowly. However, the highest rate of non-canonical methylation occurred in 25 mM Tris --HCl, 25 mM NaCl and 50% glycerol, when Haesub is methylated only 1.97-fold faster than Starsub. The pattern of DNA methylation by M.HaeIII was determined both in vitro and in vivo by bisulphite modification and sequencing of methylated DNA . This chemical treatment results in deamination of all the unmethylated cytosines to uracil, which is copied as thymine in the subsequent PCR. C5-methylated cytosines react more slowly and remain unconverted . To assay methylation in vitro, an 832 bp DNA fragment encoding the chicken glutathione-S-transferase M1-1 gene was incubated with purified M.HaeIII in a buffer chosen to maximise the rate of non-canonical DNA methylation (see Fig. ). For methylation of pET-MHaeIII in vivo, the rate of methylation was also enhanced by overexpressing M.HaeIII in E.coli. A total of 3.1 mg of M.HaeIII was purified from 7.8 g of bacteria (data not shown) which corresponds to an approximate intracellular concentration of 9.15 microM enzyme (assuming a density of 1 g /ml and an Mr of the tagged protein of 43 412 Da). After methylation, the DNA fragment encoding glutathione-S-transferase was purified from the in vitro reactions and pET-MHaeIII was purified from the induced E.coli. This DNA was then treated with bisulphite . Sequencing of M.HaeIII-methylated, bisulphite-treated DNA before cloning only revealed methyl cytosine at canonical sites (data not shown), as this method displays an average of all the strands. Cloning the bisulphite-treated DNA allowed the analysis of individual strands (Table and Fig. ). 22 870 bases of in vitro-methylated DNA and 12 730 bases of in vivo-methylated DNA were sequenced. The percentage of canonical HaeIII sites methylated in vitro was 90.3% and the percentage of canonical sites in pET-MHaeIII methylated in vivo was 100%. In addition to these sites, a further 0.9% of all other cytosines in the PCR fragment and 0.6% in pET-MHaeIII remained unconverted. More than half of these unconverted cytosines were M.HaeIII star sites, with unconverted cytosines observed at seven of the nine possible star sites. Additionally, 28 unconverted cytosines were observed in sequence contexts differing by two or three bases from the canonical site; 19 in vitro and nine in vivo. However, unconverted cytosines, in addition to arising from DNA methylation, may also be due to incomplete chemical reaction of the DNA with bisulphite or mutagenesis in the subsequent PCR causing a T to C transition. To test the background level of unconverted cytosines, the unmethylated PCR fragment and unmethylated pET30a were subjected to bisulphite sequencing. A total of 7640 bases of the bisulphite-treated, cloned glutathione-S-transferase gene and 3350 bases of the bisulphite-treated, cloned, pET30a were sequenced. The percentage of cytosines remaining unconverted in these samples was 0.7 and 0.1%, respectively. No bias in the sequence context of unconverted cytosines was observed: of the 64 permutations of NNCN, 51 were never observed unconverted, 12 were observed unconverted only once, and only one site (CGCC) was observed unconverted twice. Methylated DNA, in contrast, has a strongly biased pattern of non-canonical unconverted cytosines that cannot be explained simply by incomplete reaction with bisulphite . Figure 2 | Typical sequences of the methylation template before bisulphite treatment (A), after bisulphite conversion and cloning (B) and after in vitro methylation by M.HaeIII, bisulphite treatment and cloning (C). Typical sequences of the methylation template before bisulphite treatment (A), after bisulphite conversion and cloning (B) and after in vitro methylation by M.HaeIII, bisulphite treatment and cloning (C). Figure 3 | Frequency of occurrence of C5 methyl-cytosine groups in a glutathione-S-transferase gene methylated in vitro by M.HaeIII determined by bisulphite treatment and sequencing of individual clones. Frequency of occurrence of C5 methyl-cytosine groups in a glutathione-S-transferase gene methylated in vitro by M.HaeIII determined by bisulphite treatment and sequencing of individual clones. The sequences of both strands of the gene are shown, with canonical HaeIII sites shown in red. Bars above the sequence indicate unconverted cytosines in the top strand and bars below the unconverted cytosines in the bottom strand. The height of the bars represent the percentage of unconverted cytosines at each position with a scale of 0 --50%. All canonical HaeIII sites were unconverted in >50% of the clones analysed. PCR amplification of the top strand yields a 583 bp product and the bottom strand gives a 480 bp product. Table 1 | The sequence context of C5 methyl cytosines detected in DNA methylated by M.HaeIII in vitro and in vivo DISCUSSION : Traditionally, DNA methyltransferases are viewed as highly sequence specific . However, recent observations, including this study, suggest that methylation of non-canonical sites may be a common feature of DNA methyltransferases. Non-canonical methylation was first reported for the adenine N6 methyltransferase M.EcoRI and has since been reported for two other adenine N6 methyltransferases, M.FokI and M.EcoRV. Twenty-seven oligonucleotides were tested as substrates for methylation by the M.FokI DNA methyltransferase . The enzyme has two domains, each methylating one strand of its asymmetric recognition site, GGATG. Whilst the N-terminal domain only methylated one star site, the C-terminal domain methylated most of the star sites tested and two sites differing by two base pairs from the canonical site. M.EcoRV (canonical site GATATC) is also capable of methylating five different degenerate EcoRV recognition sites, including one site with three altered bases (GATC) (,). Methylation of non-canonical sites has also been reported for three cytosine C5 methyltransferases . Overexpression of M.HpaII (canonical site CCGG), Dcm and EcoRII (CCAGG) methyltransferases in E.coli causes an increase in the C --T mutation rate at the canonical sites for methylation, but also at some non-canonical sites (CAGG and CCGGG). Restriction digestion of the plasmid isolated from E.coli overexpressing M.EcoRII showed that the DNA was partially protected from digestion at the site CCCGGG. Methylation of this star site by M.EcoRII was confirmed using a synthetic oligonucleotide in vitro. In these studies, methylation of a target cytosine or adenine was examined in a limited number of sequence contexts. Here we attempted to explore the full range of sequence contexts for methylation of cytosine residues by M.HaeIII. The pattern of methylation was investigated both in vitro, using purified enzyme, and in vivo in E.coli overexpressing methyltransferase. The pattern of non-canonical DNA methylation was similar in vitro and in vivo. Of the 63 permutations of NNCN (excepting GGCC), 25 were unconverted by bisulphite modification at least once in the clones analysed. Methylation of non-canonical sites was not uniform, either in vitro or in vivo, M.HaeIII methylates star sites more frequently than other sites. Only at two of the nine star sites was methylation not detected either in vitro or in vivo. The enzyme shows less discrimination at the first and fourth base of its recognition sequence, with the central cytosine of AGCC being methylated more frequently than all other non-canonical cytosines, in vitro and in vivo. The star sites GTCC and GCCC were never observed unconverted, and in only one case was the third base of GACC unconverted, indicating a greater stringency for recognition of the second base of the sequence. In addition to star site methylation, DNA methylated in vitro and in vivo contained unconverted cytosines at sites differing from the canonical site by two or three base pairs. However, it is difficult to be sure whether cytosines observed unconverted only once are due to methylation or to the incomplete conversion of cytosines by bisulphite. Although conversion was very efficient (over 99.3% of cytosines converted), some of the cytosines observed in the methylated DNA were almost certainly due to incomplete conversion. However, some sites, notably AGCA, were observed multiple times in methylated DNA but never in unmethylated DNA, and are likely to be the result of methylation by M.HaeIII. This site combines the base alterations of two star sites that are methylated at an appreciable level by M.HaeIII in vitro (AGCC and GGCA). In E.coli overexpressing M.HaeIII, DNA is hypermethylated to the extent that the central cytosine of the sequence AGCC is methylated in 11.9% of cases and methylation is also seen at other star sites . These high levels of methylation suggests that there might still be significant levels of non-canonical DNA methylation in Haemophilus aegyptius, despite the lower concentration of M.HaeIII. The in vitro methylation rates provide further support for this suggestion as the enzyme concentration (10 nM) and DNA concentration (1.25 microM sites) were very close to those predicted in bacteria (50 --500 nM methylase and similar1 --10 microM sites) . Even under the conditions where the enzyme showed highest specificity for its canonical site, the rate of methylation of non-canonical substrate was still only 36.7-fold lower than the rate with the canonical substrate. This corresponds to a 303-fold lower rate of methylation per cytosine, based on two canonical target cytosines in Haesub and 17 non-canonical cytosines in Starsub. The existence of a large number of R /M systems with a wide range of sequence specificities raises the question of how methylases and restriction enzymes were able to co-evolve to recognise new target sites. It has been proposed that many enzymes can catalyse the conversion of molecules which are not their principle substrate with low efficiency, and that new enzymes may evolve by improvements in the enzyme's ability to catalyse the conversion of one of these poor substrates (,,). Promiscuous activities as low as described above (and indeed lower) can provide a selective advantage and thereby a starting point for the evolution of a new activity . In particular, it has been proposed that protein --DNA interactions evolve by relaxing an existing specificity (namely, increasing promiscuity) and then restricting it to the new target; this principle has recently been used to alter the site specificity of Cre recombinase . High levels of star activity by restriction endonucleases would be lethal unless the cognate methylase also had the ability to methylate and protect these star sites. The problems associated with non-canonical methylation are probably less severe. These include the energetic cost and the 10-fold higher mutation rate of C5-methylated cytosines compared with unmethylated cytosines (,), and the different response of prokaryotic DNA-interacting proteins, for example, methyl-directed restriction enzymes, to methylated and unmethylated DNA. In fact the selective disadvantage attributed to a methyltransferase with two distinct recognition sites (the canonical site and a newly arisen specificity) is comparable to that caused by expressing two different methyltransferases, whereas natural populations of bacteria may contain in excess of 10 different R /M systems . Hence, it is likely that a DNA methyltransferase with a new specificity arises before the cognate restriction endonuclease. If DNA methylases possess a significant level of activity at non-canonical sites then this could have served as a starting point for evolutionary divergence towards a new R /M site (,). Type II R /M systems recognise symmetrical DNA targets 2 --8 bp in length so the evolution of a new specificity would usually involve a change of at least two base pairs in the target site. The ability of M.HaeIII to methylate cytosines in a variety of non-canonical sequence contexts indicates that such changes in specificity may not represent a large evolutionary barrier. Backmatter: PMID- 12202750 TI - Predicted structure and phyletic distribution of the RNA-binding protein Hfq AB - Hfq, a bacterial RNA-binding protein, was recently shown to contain the Sm1 motif, a characteristic of Sm and LSm proteins that function in RNA processing events in archaea and eukaryotes. In this report, comparative structural modeling was used to predict a three-dimensional structure of the Hfq core sequence. The predicted structure aligns with most major features of the Methanobacterium thermoautotrophicum LSm protein structure. Conserved residues in Hfq are positioned at the same structural locations responsible for subunit assembly and RNA interaction in Sm proteins. A highly conserved portion of Hfq assumes a structural fold similar to the Sm2 motif of Sm proteins. The evolution of the Hfq protein was explored by conducting a BLAST search of microbial genomes followed by phylogenetic analysis. Approximately half of the 140 complete or nearly complete genomes examined contain at least one gene coding for Hfq. The presence or absence of Hfq closely followed major bacterial clades. It is absent from high-level clades and present in the ancient Thermotogales-Aquificales clade and all proteobacteria except for those that have undergone major reduction in genome size. Residues at three positions in Hfq form signatures for the beta /gamma proteobacteria, alpha proteobacteria and low GC Gram-positive bacteria groups. Keywords: INTRODUCTION : Hfq, also called HF-I, is a 12 kDa heat-stable protein, encoded by the hfq gene at 95 min on the Escherichia coli chromosome map . Originally discovered as a host factor required for bacteriophage Qbeta RNA replication , it was later shown to be associated with ribosomes and, to a lesser extent, with the nucleoid (,). Hfq is a global regulator of E.coli metabolism, and disruption of the hfq gene can cause a pleiotropic phenotype . The broad impact of the protein appears to stem from its role in regulating the stability and /or translation of mRNAs from a number of regulatory genes. One of these mRNAs is the rpoS mRNA that encodes the stationary phase sigma factor sigmas of RNA polymerase (,). Mutational studies suggest that Hfq is involved in the processes that affect the secondary structure near the 5' end of rpoS mRNA alleviating an inhibition of ribosome access to a translation start region . A similar behavior is inferred from studies that indicate Hfq helps open an inhibitory stem --loop structure at the 3' end of Qbeta plus-strand RNA to mediate access of Qbeta replicase . Hfq has been shown to affect the in vivo stability of mRNAs expressed from the ompA, mutS, miaA and hfq genes (,), and to stimulate elongation of the poly(A) tail of the rpsO mRNA . The mechanism of the effect of Hfq on mRNA stability appears to involve its influence on the interaction of non-coding regulatory RNAs with specific mRNAs. It has been shown to affect the binding of regulatory RNAs DsrA, OxyS, RprA and Spot 42 with their target mRNAs (, --). While functional roles for Hfq have been demonstrated, and general models for its mechanism of action proposed , the absence of structural information on Hfq and Hfq --RNA complexes hinders an understanding of its molecular mechanism(s) of action. Recent amino acid sequence analysis of Hfq has shown that the N-terminal portion of Hfq is highly conserved among a number of bacteria and shares a strong similarity with the Sm1 motif of Sm and Sm-like (LSm) proteins found in eukaryotes and archea (,). These results suggest that Hfq is an ancestral Sm protein. Sm proteins are essential components of the small nuclear ribonucleoproteins (snRNPs) that form spliceosomes (,). Sequence comparisons of Sm proteins from a range of species showed that the Sm motif is comprised of two conserved regions, Sm1 and Sm2, separated by a region varying in length and sequence . Biochemical and crystallographic studies (,) have demonstrated that the Sm motif dictates a common folding domain that enables Sm proteins to assemble onto a uridine-rich region of snRNAs and form a ring-like heteroheptamer. Formation of this core structure is essential for the stability and function of the snRNPs . Searches of eukaryotic genome databases have shown that a large number of proteins contain the Sm sequence motif (,). Some of these proteins are similar to the originally characterized splicesomal Sm proteins, and others are referred to as LSm proteins. Analysis of archaeal genomes also revealed the presence of ORFs that encode LSm proteins . Biochemical and crystal studies of three archaeal LSm proteins revealed that they exhibit properties similar to their counterparts in eukaryotes. They bind to RNA with oligo(U) sequences, and assemble a heptameric ring around the RNA (,). A comparison of the monomer subunits in the crystal structures of eukaryotic Sm proteins that form dimers with the monomer subunits of archaeal LSm proteins that form homodimers and heptamers (,) show strong similarities. Each subunit has a short alpha helix followed by five interwoven beta strands separated by short loops. The presence of the Sm1 motif sequence in E.coli Hfq, and the ability of Hfq to form a hexameric ring and bind RNA, support the notion that it is evolutionarily related to the Sm family of proteins. However, the absence of the Sm2 motif in Hfq makes the structural relationship of Hfq with the known structures of Sm proteins uncertain. In this report secondary structure prediction, amino acid solvation properties and three-dimensional (3D) threading algorithms were used to predict a 3D structure for the N-terminal domain of Hfq. The predicted structure fits very well with the major features of the Calpha backbone of Methanobacterium thermoautotrophicum LSm protein. The Sm1 motif sequence in Hfq is structurally aligned with its counterpart in the archaeal LSm protein, and a highly conserved portion of the Hfq sequence assumes a structural fold similar to that of the Sm2 motif of archaeal and eukaryotic Sm proteins. Highly conserved residues of Hfq are also located in the same structural region that is responsible for subunit assembly in the Sm proteins. The strong structural similarity of Hfq and the LSm protein supports the hypothesis that Hfq is an ancestral Sm protein and contributes confidence in its predicted 3D structure. During revision of this manuscript a paper describing the crystal structure of the Staphylococcus aureus Hfq hexameric protein and a complex of this protein with RNA was published by Schumacher et al. . Our 3D model of the E.coli Hfq monomer is in excellent agreement with the monomer of this structure. Comparison of the S.aureus Hfq structure with predicted features of Hfq based on our structure /sequence analysis is presented below. The presence of the Hfq protein in bacteria was explored by BLAST searches against bacterial genomes available in the NCBI databases. Approximately half of the bacterial genomes examined contain an Hfq protein based on strong amino acid sequence similarity, protein sequence length and amino acid conservation pattern. Phyletic distribution of Hfq indicates that it is an ancient protein. We obtained no evidence that Hfq might be a subject of lateral gene transfer and conclude that gene loss played a major role in its evolution. The bacterial species in which Hfq was absent were highly correlated with specific taxonomic or lifestyle trends. MATERIALS AND METHODS : Database searches | Non-redundant database (NCBI) searches were performed by Position-Specific-Iterative (PSI)-BLAST program , using the amino acid sequence of E.coli Hfq (GI 16131994) as the primary query sequence. The inclusion threshold (E value) employed was 0.01. A multiple sequence alignment was constructed by the CLUSTAL W program . Additional BLAST searches were carried out against the Microbial Genomes database at NCBI , using amino acid sequences of Hfq or the consensus Sm motif as the query. Multiple sequence alignment and secondary structure prediction | A multiple sequence alignment of Hfq proteins was constructed by the CLUSTAL W program using the output of the PSI-BLAST searches. The amino acid conservation pattern was determined by calculating consensus using the Perl script by Nigel Brown and Jianmei Lai (available at ). The secondary structure was predicted using the consensus method JPRED2 , which utilizes multiple sequence alignments, along with PSI-BLAST and HMM profiles. Fold recognition and 3D modeling | The 3D-PSSM (Position Specific Scoring Matrix) program was employed to search for proteins with structural similarity to Hfq. This server provides 3D structural information about the backbone of a query protein by scoring the relationship between the residues of a query sequence with the residues of a homologous protein of known structure. The query protein is scanned against a library composed of proteins with known crystal structures and scored for compatibility using several scoring components. These include amino acid sequence profiles built from relatively close homologs, more general profiles containing more remote homologs, matching of secondary structure elements, and matching the propensities of residues to occupy varying levels of solvent accessibility. Known protein structures within the database with significant homology to the query sequence are used to produce closest fit alignments between query sequence and target structures that maximize position specific scores. The top 20 structural alignments to the query sequence are produced, each illustrating regions of similarity and differences. The SWISS-MODEL comparative protein modeling server was employed to generate a 3D model of the E.coli Hfq protein based on the structural alignment of its sequence with the highest scoring template structure determined by 3D-PSSM. In the initial step of modeling, the Hfq query sequence was modified in order to accommodate the four and six residue segments absent from Hfq when compared with the best template structure, the M.thermoautotrophicum LSmalpha protein (see Fig. A and B). Residues were inserted into the Hfq sequence at the minus-labeled segments shown in Figure C and D to produce a query sequence that would match the length of the M.thermoautotrophicum template sequence. This modified Hfq query sequence was submitted to the SWISS-MODEL server, which produced a predicted structure. The query sequence was then changed to the correct Hfq sequence by replacing the inserted residues with gaps and the Hfq sequence and template sequence resubmitted to the server in 'Optimize Mode' after aligning the gaps to the template sequence as indicated by the 3D-PSSM model. Figure 3 | The 3D structure of Hfq generated by SWISS-MODEL program using the same archaeal LSmalpha protein determined to be the best template by 3D PSSM. The 3D structure of Hfq generated by SWISS-MODEL program using the same archaeal LSmalpha protein determined to be the best template by 3D PSSM. (A and B) Front and side views of the predicted Hfq structure as well as the template Sm structure. The backbone features that are constant in both structures are illustrated in blue. Differing structural elements are shown by using a red ribbon for Hfq and a green ribbon to illustrate the LSmalpha protein backbone. The aqua ribbon illustrates the beta4 strand residues SQMVY and beta5 strand residues AISTVV. (C and D) Front and side views of the predicted Hfq model with several potential RNA-interacting residues shown in stick model representation: Lys31 in loop 2, Phe39 and Phe42 in loop 3, Lys56 and His57 in loop 5. RESULTS AND DISCUSSION : Sequence similarity of Hfq and Sm proteins | A search of the non-redundant database (NCBI) using PSI-BLAST was carried out with the E.coli Hfq amino acid sequence as the query sequence. Twenty-five similar (statistically significant) sequences were detected from a range of bacterial species. Multiple alignment of these sequences is shown in Figure . Hfq proteins are highly conserved in their N-terminal halves of the molecules. This conserved domain corresponds to residues 7 --64 in E.coli Hfq. In contrast, the C-termini of Hfq proteins vary greatly among the different species. In some instances it is totally absent, e.g. the 57 amino acid Hfq protein of Bacillus anthracis. This result implies that the C-terminal region might not play a significant role in the major function(s) of Hfq and attention was focused on the N-terminal region. This hypothesis is supported by a recent study showing that the Hfq homolog of Pseudomonas aeruginosa, consisting of 82 amino acids from the N-terminal end, can functionally replace E.coli Hfq for phage Qbeta replication and for rpoS expression . Two conserved motifs are observed in Hfq. The first motif, Sm1, is a counterpart of the Sm1 motif found in archaeal and eukaryotic Sm and LSm proteins . From Figure , it can be seen that the Sm1 sequence is well aligned to residues 20 --52 of the E.coli Hfq sequence. As noted previously , the Sm2 motif of archaeal and eukaryotic proteins does not appear to have a counterpart in Hfq. However, Hfq does have an additional conserved region, YKHA, following the Sm1 motif. The relationship of the YKHA motif of Hfq with the Sm2 motif of Sm proteins was explored by generating a structural model of Hfq. Comparison of predicted structure of Hfq with known structure of the Sm protein | The secondary structure of the consensus Hfq sequence was predicted by JPRED2 . Figure shows that Hfq is a beta-sheet-rich structure with an alpha helix at the N-terminus. All the predicted secondary structure elements fall in the region that is conserved in the multiple alignment. In contrast, no secondary structure elements were predicted in the C-terminus of Hfq. Crystal structures of several Sm proteins show a common fold for the Sm motif. The fold contains an N-terminal helix, followed by five segments of beta strands (, --). Strands beta1, beta2 and beta3 are part of the Sm1 motif, whereas the Sm2 motif corresponds to beta4 and beta5 strands (shown in Fig. ). The topology of the secondary structure elements in an Sm protein is schematically shown in Figure A and B. Strands beta2, beta3 and beta4 are strongly bent to allow the formation of the hydrophobic core. The structural plasticity needed for such a high degree of curvature in the beta1 strand is provided by several strictly conserved glycines that occur near the pivot points (Gly18, Gly23, Gly53 and Gly59 in Sm consensus; Fig. ). The segment linking the beta4 and beta5 lies at the top of the U-shaped trough to close the protein into a beta-barrel-like structure. Hfq has the same predicted secondary structure elements in the Sm1 motif region as does the Sm protein: an alpha helix followed by beta strands. The critical residues that are required for the beta1 strand curvature (Gly29 and Gly34 in Hfq) are identical in all Hfq homologs. Interestingly, another long beta strand (in some predictions, it was two separate beta strands) was predicted in Hfq from Ser51 to Pro65. Although this region could not be aligned to the Sm2 motif in the Sm protein sequence, the length of this region ---referred to as Xm2 in Figure ---closely matches the length of the Sm2 motif. In addition, the Sm2 and Xm2 motifs both have highly conserved residues flanked by hydrophobic residues. There are two highly conserved residues in the middle of the Sm2 motif, Arg --Gly, and four highly conserved residues in the Xm2 motif, Tyr --Lys --His --Ala. The 3D structure of Hfq was predicted using comparative modeling as described in the Materials and Methods. 3D-PSSM was first employed to thread the Hfq sequence as a query against known protein structures in a fold library. The template structure that produced the highest score for the Hfq sequence was the archaeal LSmalpha protein from M.thermoautotrophicum . The structure of the LSmalpha protein is shown in Figure A and B, while Figure C and D shows the best-fit model structure of Hfq with the locations of deletions and insertions from the template structure that produced the highest score. The Hfq model is well aligned to the LSmalpha protein. The major difference is in the beta3 and beta4 region. Four and six amino acids of the template structure are absent from Hfq in the beta3 and beta4 strands respectively. Visually, the Hfq structure suggests that if these 10 residues are simultaneously deleted, the isolated loop 4 may be able to connect the remaining fragments of beta3 and beta4. The missing parts of Hfq, when compared with the LSmalpha protein, fall just within the highly variable region between the Sm1 and Sm2 motifs, which includes loop 4 as well as parts of the beta3 and beta4 strands. This suggests that the amino acid sequences which constitute a minimum Sm fold can be shortened, and may be composed of adjacent Sm1 and Sm2 motifs with no variable linker. The SWISS-MODEL program was then employed to generate a 3D model of the Hfq protein using the archaeal LSmalpha protein as a template, and the information inferred from the 3D-PSSM highest scoring alignment shown in Figure C and D. The optimized structural model of Hfq is shown in Figure where it is compared with the structure of the LSmalpha protein. The features that are constant in both structures are illustrated in blue. A red ribbon designates Hfq and a green ribbon illustrates the LSmalpha protein in the regions where there are differences in their structural features. The beta3 and beta4 strands of Hfq are shortened relative to the LSmalpha protein and connected by loop 4. Loop 4 changes its orientation from up and to the right for the LSmalpha protein to a downward direction for Hfq (Fig. A and B). The aqua-colored ribbon shows the location of the beta4 strand residues SQMVY, and the beta5 strand residues AISTVV. Figure shows the beta4 --loop 5 --beta5 region in greater detail. The residues in the beta5 strand of Hfq, STVVP, appear to occupy similar spatial locations as the corresponding residues of the LSmalpha protein, VLISP. Based on sequence alignment, amino acid characteristics and secondary structure prediction, we anticipated that the highly conserved His --Ala residues of Hfq would occupy the same 3D positions as the highly conserved Arg --Gly of an Sm protein. However, a comparison of the structural models in Figure indicates that the His --Ala pair in Hfq is shifted in their relative location two residues downstream when compared with the Arg --Gly residues in the Sm structure. The predicted structure of Hfq was compared with the Sm protein structure with regard to segments that may be involved in subunit interaction in the formation of multimers. Several studies indicate that Hfq forms a hexamer (,,,,), while Sm proteins form a homo- or hetero-heptamer depending on the number of distinct subunits available in vivo. Archea species form a heptamer composed of seven identical subunits (,). Eukaryotes utilize different polypeptide chains to assemble a hetero-heptamer Sm complex (,,). In both cases, adjacent monomers in the heptamer interact via pairing of the beta4 and beta5' strands (the ' indicates the adjacent subunit). Only the last five residues of the beta4 strand in one Sm protein are involved in pairing with the beta5' strand of the adjacent Sm protein. In the structural model of Hfq the first six residues of the beta4 strand are absent when compared with the Sm protein. However, the five remaining residues that form the beta4 strand and the beginning of loop 5 are located in similar positions to the residues of the LSmalpha protein that participate in quaternary interactions . The residues spanning the beta4 --beta5 strand region of the human Sm protein that are involved in pairing adjacent subunits are LVLLRGSVIVV , while in the archaeal LSm they are TVLIRGQNIVY . In both cases, a group of hydrophobic residues flank a positively charged Arg that is engaged in several hydrogen bonds with main chain and side chain atoms of the adjacent subunit's beta5' strand. In the predicted Hfq structure, the residues spanning the beta4 strand --loop --beta5 strand region are SQMVYKHAISTVV. One again has hydrophobic residues flanking positively charged residues, in this case lysine and histidine. Although, as mentioned above, the His --Ala residues of Hfq are not in the same structural position as the Arg --Gly residues of the Sm protein, the similar nature of the residues in this region suggest that the predicted Hfq structure also supports multimer formation through beta4 --beta5' strand pairing. It is worth noting that the sequence in this part of the Hfq structure, VYKHAIST, is almost completely conserved among Hfq proteins . The recently determined S.aureus Hfq structure shows that beta4 --beta5' strand interface is indeed a key part of intersubunit interactions. In this structure, H bonds occur between the highly conserved Tyr56 in beta4 and Tyr63 in beta5'. In the E.coli Hfq sequence valine occurs at position 63. This is the more dominant amino acid at this location in Hfq proteins and suggests that Tyr56 H-bonds with a different residue in beta5' in E.coli Hfq or this H bond is not essential for this interface. The S.aureus Hfq structure also shows that contacts between alpha helix residues and loop L3 residues of the adjacent subunit and between side chains in beta strands contribute to the dimer interface. The 3D model of Hfq also provides an opportunity to consider its potential sites of interaction with RNA. The RNA determinants important for Sm core assembly appear to be complex. One prerequisite for an RNA to be bound by an Sm protein heptamer is an 'Sm site element', a 7 --10 nt single-stranded segment that has the consensus sequence PuAU3 --6GPu usually flanked by stem --loop structures (,). In vitro analysis with an RNA oligonucleotide consisting of a minimal Sm site element revealed that the 5' adenosine of the element plays a critical role in the heptamer's association, while the uridine bases and the 2' hydroxyl groups collectively provide a binding determinant (,). In human snRNP core, several Sm proteins were shown to interact with the uridine stretch of the Sm site element by UV cross-linking experiments. The most efficient cross-links were observed for the G and B /B' proteins, which are linked to the first and third uridines of the Sm site element respectively . The residues (His37 for B /B', Phe37 for G) involved in contacting the RNA are located at equivalent regions in both proteins, namely in loop L3 of the Sm1 motif. In contrast, crystal structure of the archaeal SmAP protein suggests that residues in other loops (Arg29 in L2, Asp57 in L4 and Glu71 in L5) are more likely to interact with the RNA Sm site element . All four of these loops jut into the inner ring or pore of the doughnut-shaped heptamer . The corresponding regions of Hfq, which by analogy would be expected to be oriented toward the inner ring of the hexamer, also have conserved residues (for the E.coli sequence: Lys31 in loop 2, Phe39 and Phe42 in loop 3, Lys56 and His57 in loop 5). The location of some of these residues is illustrated in Figure . The Phe39 and Phe42 in loop 3 and Lys56 --His57 in loop 5 are almost 100% conserved among different bacterial species examined, implying they have critical roles in structure and function. Hfq has been shown to be an essential participant in facilitating the interaction of some small riboregulator RNAs, such as DsrA and Spot42 , with their target mRNAs. It was proposed that the role of Hfq might be analogous to Rop, in which two phenylalanines intercalate into base pairs and facilitate the pairing of two RNA molecules . If Hfq functions in this way, the highly conserved Phe42 in loop 3 and its nearby Phe39 are candidates for this role (Fig. C and D). Several of the above predictions are verified in the recently published crystal structure of the S.aureus Hfq --RNA complex . In this structure, the backbone of the oligoribonucleotide 5'-AUUUUG-3' was found to form a circular conformation as it bound to an electropositive patch around one face of the pore of the hexameric Hfq. Residues in the Sm1 and Sm2 motifs of adjacent subunits are utilized to build six nucleotide-binding pockets. There are no intramolecular base stacking interactions within the RNA as the bases are splayed out, fitting into the individual binding pockets. Each base is sandwiched between two Tyr42 side chains from adjacent subunits. The presence of Phe42 instead of Tyr42, which occurs for most Hfq sequences, appears to be able to serve the same function for the nucleotide-binding pockets. The highly conserved Lys --His motif located in loop 5 and facing the pore also contacts the RNA. Lys57 (shifted by one amino acid in S.aureus due to an extra residue relative to the E.coli sequence) H-bonds with uracil and His58 makes contacts with the phosphate oxygens of one nucleotide as well as the ribose O2' hydroxyl of the adjacent nucleotide. Phyletic distribution of the Hfq protein | In order to determine the pervasiveness of the Hfq gene and related Sm proteins in bacteria and explore its evolution, a BLAST search was conducted against bacterial genomes available at the NCBI database. Fifty-eight completed and 82 unfinished bacterial genomes were examined. Approxi mately half of the bacterial genomes contain at least one gene that codes for an Hfq protein based on strong amino acid sequence similarity to the E.coli Hfq sequence, sequence length and amino acid conservation pattern. The presence and absence of the Hfq protein in particular species closely follows recently redefined major bacterial clades (,), as shown in Table . Hfq is missing from three high-level bacterial clades: Chlamydia-Spirochaetes, Actinomycetes-Deinococcus- Cyanobacteria and Green sulfur bacteria-Cytophagales. However, it is present in the most deeply branched Thermotogales-Aquificales. The Hfq protein is present in all alpha, beta, gamma and delta proteobacteria except for those that have experienced a massive genome reduction due to their parasitic lifestyle, e.g. Buchnera sp. , Rickettsia prowazekii and Brucella melitensis . This type of phyletic distribution suggests two possible scenarios for the evolution of the Hfq protein. First, Hfq might be an ancient protein, which was lost early in evolution by major clades and retained only by the lineage leading to proteobacteria. Second, the Hfq protein evolved late in evolution during the separation of the proteobacterial clade and was transferred laterally to several species outside the proteobacteria. A phylogenetic tree built from the multiple alignment of the Hfq protein sequences is shown in Figure . The tree has a topology expected from Table . Three major clades of alpha proteobacteria, beta /gamma proteobacteria and low GC Gram-positive bacteria are well defined and supported by bootstrap analysis. A comparison of Hfq sequences from bacteria in the three groups illustrated in Figure is given in Figure . The results reveal three positions, at the borders of loop 2 and loop 3, and in loop 4 of the predicted Hfq structure that have residues characteristic to each group. For the low GC Gram-positive bacteria, the dominant residues corresponding to the E.coli sequence positions 30, 43 and 50 are phenylalanine, tyrosine and lysine respectively. The alpha proteobacteria are dominated at these positions by valine, cysteine, and histidine or glutamine, while the beta /gamma proteobacteria universally have isoleucine, valine and valine at these locations. The positions are highlighted in Figure . These residues may provide some specificity in the interactions of Hfq with RNA or the interactions governing subunit oligomerization. Since the above analysis is based on the assumption that the model structure for the E.coli Hfq is appropriate for other Hfq proteins, it is worth noting that our predicted structure is in excellent agreement with the recently published crystal structure of the S.aureus Hfq . We also note that four independent algorithms which utilize a single amino acid sequence as a query predict secondary structures for S.aureus Hfq and other relatively distant Hfq sequences (e.g. Geobacter sulfureducens Hfq) that closely fit the LSm and consensus Hfq structures (data not shown). Another outcome from the microbial genome search worth noting was that 6 of the 140 eubacterial genomes examined contained two distinct copies of an Hfq protein coding sequence . Duplicated hfq genes are always found within the same clade on the phylogenetic tree as original copies indicating the likelihood of paralogous relationships over lateral gene transfer. Two Hfq sequences are found within a single 193-residue protein of the bacterium Novosphingobium aromaticivorans, which reinforces the notion that Hfq is a subject of relatively frequent gene duplication events. Twenty-three residues separate the two distinct 59 residue Hfq motifs (E values of 3 x 10 --14 and 2 x 10 --10) in the N.aromaticivorans protein. This protein may code for a heterodimeric version of an Hfq structural unit similar to the heterodimers observed for the eukaryotic Sm proteins. Our phylogenetic analysis produced no evidence for lateral transfer of Hfq. This is consistent with the proposal that the bacterial (Hfq) and archaeal /eukaryotic (Sm and LSm) versions of this important RNA-binding protein shared a common ancestor prior to the separation of bacteria and archaea --eukarya. Gene loss appears to be a major driving force in the evolution of Hfq. Figure 1 | Multiple alignment of Hfq proteins from 26 bacterial genomes compared with the LSm protein from M.thermoautotrophicum and a consensus sequence for Sm proteins (shown above the alignment). Multiple alignment of Hfq proteins from 26 bacterial genomes compared with the LSm protein from M.thermoautotrophicum and a consensus sequence for Sm proteins (shown above the alignment). Known secondary structure for the LSm protein and predicted structure for Hfq proteins are shown above the corresponding sequences: H, alpha helix; E, beta strand. The Sm1 and Sm2 motifs of the Sm protein and the Sm1 and Xm2 motifs of Hfq proteins are shown. The 90% consensus shown below the alignment was derived using the following amino acid groupings. Positively charged residues (RKH) are shown as white letters on a red background; polar residues (p, KRHEDQNST) are shown as red letters; turn-like residues (t, ACDEGKNQRST) are green letters; bulky hydrophobic residues (h, ACLIVMHYFW) and the aliphatic subset of these type residues (l, LIVM) have a yellow background; aromatic residues (a, FHWY) are white letters with a purple background; small residues (s, ACDGNPSTV) are blue letters; tiny (u, AGS) are white letters with a blue background. Sequences are denoted by the species abbreviation followed by GI number. Species abbreviations: M.ther, M.thermoautotrophicum; B.halo, Bacillus halodurans; B.subt, Bacillus subtilis; L.inno, Listeria innocua; T.mari, Thermotoga maritima; C.acet, Clostridium acetobutylicum; A.caul, Azorhizobium caulinodans; C.cres, Caulobacter crescentus; M.loti, Mesorhizobium loti; B.meli, Brucella melitensis biovar Abortus; S.meli, Sinorhizobium meliloti; P.mult, Pasteurella multocida; P.prof, Photobacterium profundum; H.infl, Haemophilus influenzae; V.chol, Vibrio cholerae; Y.pest, Yersinia pestis; Y.ente, Yersinia enterocolitica; P.caro, Pectobacterium carotovorum; E.coli, E.coli; S.typh, Salmonella typhimurium; S.flex, Shigella flexneri; P.aeru, P.aeruginosa; X.fast, Xylella fastidiosa; N.meni, Neisseria meningitidis; A.aeol, Aquifex aeolicus; B.anth, Bacillus anthracis; S.aure, S.aureus. Figure 2 | (A and B) Ribbon representations of two views of the crystal structure of an archaeal Sm protein (PDB accession number 1i81) rotated by 90Epsilon. (A and B) Ribbon representations of two views of the crystal structure of an archaeal Sm protein (PDB accession number 1i81) rotated by 90Epsilon. Images are produced by RasMol program . (C and D) 3D line representations of the Hfq structure predicted by the 3D-PSSM web server using the above archaeal Sm protein as template. The views shown in (C) and (D) are the same as in (A) and (B) respectively. The locations of Hfq residues that are inserted or deleted when compared with the template are represented by thin and thick bars respectively, and accompanied by numbers indicating the number of residues involved. Labels B1 --B5 correspond to the beta strands beta1 --beta5; labels L1 --L5 correspond to the loops. Figure 4 | Molecular representation of the beta4 and beta5 strands in Hfq model (A) and LSmalpha protein (B). Molecular representation of the beta4 and beta5 strands in Hfq model (A) and LSmalpha protein (B). Overlapping representation of both is shown in (C). Figure 5 | Unrooted neighbor-joining tree inferred by analysis of Hfq protein sequences. Unrooted neighbor-joining tree inferred by analysis of Hfq protein sequences. Sequences were aligned using CLUSTAL program and all positions with gaps were excluded from the analysis. Bootstrap values of >600 are displayed at deep nodes only. Color code: green, low GC Gram-positive bacteria; red, alpha proteobacteria; purple, beta proteobacteria; blue, gamma proteobacteria; orange, delta proteobacteria. Aquifecales-Thermatogales and unclassified Magnetococcus are shown in black. Figure 6 | Conserved amino acid residues specific to Hfq proteins from major bacterial groups defined by phylogenetic analysis. Conserved amino acid residues specific to Hfq proteins from major bacterial groups defined by phylogenetic analysis. Multiple alignment of Hfq sequences is subdivided according to bacterial groups inferred from the tree shown in Figure . Positions where amino acid conservation are group specific are shown. Table 1 | Presence or absence of Hfq sequence from BLAST search of bacterial genomes Backmatter: PMID- 12202768 TI - Dnmt3L is a transcriptional repressor that recruits histone deacetylase AB - The Dnmt3L protein belongs to the Dnmt3 family of DNA methyltransferases by virtue of its sequence homology in the plant homeodomain (PHD)-like motif. Dnmt3L is essential for the establishment of maternal genomic imprints and, given its lack of key methyltransferase motifs, is more likely to act as a regulator of methylation rather than as an enzyme that methylates DNA. Here, we show that Dnmt3L, like Dnmt3a and Dnmt3b, interacts both in vitro and in vivo with the histone deacetylase HDAC1. Consistent with the binding to a deacetylase, Dnmt3L purifies histone deacetylase activity from nuclear extracts. We find that Dnmt3L can repress transcription and that this repression is dependent on HDAC1 and is relieved by treatment with the HDAC inhibitor trichostatin A. Binding of Dnmt3L to HDAC1 as well as its repressive function require the PHD-like motif. Our results indicate that Dnmt3L plays a role in transcriptional regulation and that recruitment of the HDAC repressive machinery is a shared and conserved feature of the Dnmt3 family. The fact that, despite the absence of a methyltransferase domain, Dnmt3L retains the capacity to contact deacetylase further substantiates the notion that the Dnmts can repress transcription independently of their methylating activities. Keywords: INTRODUCTION : Methylation of cytosines within the CpG dinucleotide is essential for mammalian development and is associated with gene silencing (,). Transcriptional repression by DNA methylation plays a key role in several biological processes such as X-chromosome inactivation, genomic imprinting, the suppression of parasitic DNA sequences and tissue-specific gene expression ( --). It is also becoming increasingly clear that alterations in DNA methylation contribute to tumorigenesis . Methylation of CpG sites is established by the DNA methyltransferases ---the Dnmts. The first isolated DNA methyltransferase, Dnmt1, has a high affinity for hemimethylated DNA substrates and is targeted to replication foci (,). Given these features, the primary function of Dnmt1 is thought to be its ability to function as a 'maintenance' methyltransferase, restoring the pre-existing methylation patterns during DNA replication. It is possible, however, that Dnmt1 also possesses de novo methyltransferase functions . Disruption of Dnmt1 in mice results in embryonic lethality . Besides Dnmt1, another set of DNA methyltransferases has recently been discovered, namely Dnmt3a and Dnmt3b, which are the founders of the Dnmt3 family of methyltransferases . As with Dnmt1, they can be divided at the sequence level into two domains, a C-terminal catalytic domain responsible for the methylation of DNA, and an N-terminal regulatory domain. The N-terminal non-catalytic portions of Dnmt3a and Dnmt3b show no strong sequence similarity to each other apart from a cysteine-rich motif, termed the plant homeodomain (PHD), present in many chromatin-associated proteins . In particular, the cysteine-rich region of Dnmt3a and Dnmt3b is most closely related to the imperfect PHD, or PHD-like, motif found in ATRX, a putative ATPase of the SNF2 family . Mice that lack Dnmt3a or Dnmt3b have been generated and reveal that these proteins are essential for de novo methylation and mammalian development . From this study as well as others , Dnmt3a and Dnmt3b have clearly been established as de novo methyltransferases. However, the possibility that they also act as 'maintenance' methyltransferases cannot be excluded . Recent advances have begun to shed light on the mechanisms by which the Dnmts repress transcription. They can recruit the histone deacetylase (HDAC) repressive machinery, which removes acetyl groups from histones resulting in gene silencing ( --). The N-terminal non-catalytic portion of the Dnmts mediates the association with HDAC and, intriguingly, the Dnmt enzymatic activity was found to be dispensable for transcriptional silencing (,). These observations suggest the attractive possibility that the Dnmts may have additional roles in the cell beyond their ability to methylate CpG dinucleotides. Recently, Dnmt3L, a novel isolated gene, has been placed in the Dnmt3 family by virtue of its strong sequence similarity in the PHD-like fingers (see Fig. A). In contrast to the other Dnmts, Dnmt3L lacks most of the C-terminal catalytic domain and is therefore almost certainly devoid of intrinsic DNA methyltransferase activity. The inactivation of Dnmt3L in mice has been reported recently and it was found that Dnmt3L is required for the establishment of maternal methylation imprints during oogenesis . Given the lack of methyltransferase domain in Dnmt3L, this study suggests that Dnmt3L is likely to act as a regulator of imprint establishment rather than as an enzyme that methylates DNA. In the present work, we have addressed the mechanisms by which Dnmt3L functions. Given that the PHD-like motif of Dnmt3a and Dnmt3b mediates the association with HDAC (,) and that this motif is conserved in Dnmt3L, we examined whether Dnmt3L can function as a transcriptional repressor through the recruitment of histone deacetylase. We now show, both in vitro and in vivo, that Dnmt3L binds the histone deacetylase HDAC1 and associates with histone deacetylase activity. We find that Dnmt3L can repress transcription and that this repression is dependent on HDAC1 and is sensitive to trichostatin A (TSA) treatment. Finally, we show that the association of Dnmt3L with HDAC1 and its repressive activity occur through the PHD-like motif. These data indicate that transcriptional silencing via the HDAC machinery is a shared and conserved feature within the Dnmt3 family. Given the lack of methyltransferase domain in Dnmt3L, this work also further supports the notion that the Dnmts possess functions other than the ability to methylate DNA. Figure 1 | . . Dnmt3L interacts with HDAC1 in vitro. (A) Schematic representation of the Dnmt3 family members. The conserved N-terminal cysteine-rich region, the PHD-like motif, characteristic of the Dnmt3 family is shown. The C-terminal domain is the catalytic methyltransferase domain present in Dnmt3a and Dnmt3b whereas the key catalytic motifs are absent in Dnmt3L. Numbers indicate amino acid residues. (B) Dnmt3L binds the histone deacetylase HDAC1 in GST pull-down experiments. Full-length HDAC1 was IVT and 35S-radiolabeled, incubated with equivalent amounts of GST (lane 2), GST fusion proteins of Dnmt3L 1 --387 (full-length; lane 3), residues 53 --145 (containing the PHD-like motif; lane 4) or Dnmt3L 1 --387 Delta53 --145 (a mutated version that lacks the PHD-like motif; lane 5). GST 1 --387 Delta53 --145 (lane 5) was tested in a separate experiment than the other constructs (lanes 1 --4). Molecular weight in kDa is indicated on the right. The bound IVT HDAC1 is indicated by an arrow on the left. Lane 1, 35S-radiolabeled HDAC1 input (10%). (C) Dnmt3L binds specific regions of HDAC1 in vitro. The upper panel is a schematic representation of the HDAC1 deacetylase with its catalytic domain depicted by a grey box. The indicated GST --HDAC1 fusions were tested in GST pull-down experiments using IVT full-length Dnmt3L (lower panel, lanes 3 --7). Lane 1, 35S-radiolabeled Dnmt3L input (10%). MATERIALS AND METHODS : Constructs | For the GAL4 fusion experiments, we cloned full-length Dnmt3L (residues 1 --387) or the PHD-like motif (residues 53 --145), or Dnmt3L lacking the PHD-like motif (residues 1 --387 Delta53 --145) into the pcDNA3.1 GAL4 vector using PCR. pcDNA3.1GAL4 has the GAL4 DNA-binding domain (DBD) (residues 1 --147) under the control of a CMV promoter. The pGEX plasmids with various domains of HDAC1 were described previously . We cloned the PHD-like motif (residues 53 --145) of Dnmt3L or the full-length (residues 1 --387) or PHD-like domain-deleted Dnmt3L (residues 1 --387 Delta53 --145) into the pGEX4T1 vector (Pharmacia) by PCR using appropriate sets of primers. pcDNA3HDAC1-F, pING 14A-HDAC1, pGEX-Rb (379 --928) and the reporter construct 4XGAL4-TK-Luc have been described previously (,). We verified all constructs by DNA sequencing. pcDNA3.1Dnmt3L was a kind gift of U. Aapola (Tampere, Finland). Glutathione S-transferase (GST) fusion proteins, in vitro translations and pull-down assays | We expressed GST and GST fusion proteins in Escherichia coli TOP 10F (Invitrogen) and proteins from crude bacterial lysates were purified using glutathione --Sepharose 4B (Pharmacia) according to the manufacturer's instructions. In vitro transcription/translation was performed using the TNT system (Promega). GST pull-down experiments were performed essentially as described previously . Cell culture, transfections and luciferase assays | Monolayer cultures of U2OS and 293 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum at 37C in 5% CO2 atmosphere. For transfection assays, 24 h after plating, the exponentially growing cells were transfected with OptiMEM (Life Technologies) and polyethylene imine (Euromedex). All transfections were carried out with the same total amount of DNA (1 microg). Four hours after transfection, the cells were washed once in phosphate buffered saline and recovered with fresh medium. The cells were incubated for an additional 24 h, in either the presence or the absence of TSA (200 nM; Waco Bioproducts). After 24 h of incubation, the cells were lysed in 200 microl of hypotonic IPH buffer . Supernatants were clarified by centrifugation. Luciferase assays were performed with the Promega Luciferase Assay System according to the manufacturer's instructions. Transfection efficiencies were normalized using a cotransfected beta-galactosidase plasmid. Immunoprecipitations and western blot analysis | 293 cells were transiently transfected in culture dishes (10 cm diameter) with 3 microg of each expression vector. Cells were harvested 24 h post-transfection, lysed in 300 microl of IPH lysis buffer at 4C for 30 min and debris removed by centrifugation. Immunoprecipitations and western blotting were then carried out as described previously . Anti-GAL4 (5C1; Santa Cruz) and anti-Flag (M2; Sigma) antibodies were used. GST pull-down and immunoprecipitation of histone deacetylase activity from cell extracts | Equivalent amounts of GST and GST fusion proteins bound to glutathione --Sepharose beads were added to 50 microl of HeLa nuclear extracts (Computer Cell Culture Center, Belgium) in 300 microl of IPH buffer and incubated at 4C for 2 h. The beads were washed four times in IPH buffer and assayed for histone deacetylase activity. We performed HDAC assays essentially as described previously (,) in a volume of 100 microl of IPH buffer containing 250 000 c.p.m. of tritium-labeled acetylated H4 peptide. For HDAC assays from transfected cells, 293 were transiently transfected as described above with an equal amount of expression plasmids (14 microg for a plate of 14 cm diameter). Cell extracts were immmunoprecipitated with 2 microg of anti-GAL4 antibody (5C1; Santa Cruz) and incubated with 50 microl of HeLa nuclear extracts. After six washes with IPH buffer, the immune complexes were tested for histone deacetylase activity. RESULTS : Dnmt3L interacts with the histone deacetylase HDAC1 using its PHD-like motif | The de novo methyltransferases Dnmt3a and Dnmt3b have recently been reported to associate with the histone deacetylase HDAC1 using their conserved PHD-like motif . Given that Dnmt3L shows strong homology with the PHD-like motif of Dnmt3a and Dnmt3b (Fig. A), we set out to establish whether Dnmt3L associates with HDAC1. To test this possibility, we first performed in vitro GST pull-down assays using full-length Dnmt3L (residues 1 --387) fused to GST and incubated with in vitro translated (IVT) 35S-labeled HDAC1. As shown in Figure B, GST-Dnmt3L 1 --387 pulled down radiolabeled HDAC1 (lane 3) whereas GST alone did not (lane 2). As depicted in Figure B (lane 5), a mutated version of full-length Dnmt3L in which its PHD-like motif was deleted (GST-Dnmt3L 1 --387 Delta53 --145) failed to bind with HDAC1. Consistent with this observation, a GST Dnmt3L fusion that contains only the PHD-like motif (GST-Dnmt3L 53 --145) pulled down HDAC1 (Fig. B, lane 4). Thus, the conserved PHD-like motif of Dnmt3L (residues 53 --145) is required for the association with radiolabeled HDAC1. We next wished to determine which part of HDAC1 mediated the observed in vitro association with Dnmt3L. To this end, various fragments of HDAC1 were fused to GST (Fig. C) and incubated with IVT full-length Dnmt3L. Figure C shows that two non-overlapping N-terminal regions of HDAC1 (residues 1 --120 and 150 --242) mediate the binding with Dnmt3L. To further validate the interaction between Dnmt3L and HDAC1, we performed co-immunoprecipitation analyses from mammalian 293 cells co-transfected with GAL4-tagged full-length Dnmt3L (GAL4-Dnmt3L 1 --387) and Flag-tagged full-length HDAC1 (HDAC1-F). The lysed transfected cells were immunoprecipitated with anti-GAL4 antibody and western blotted with anti-Flag antibody. Figure A shows that Dnmt3L interacted with HDAC1 (lane 4) whereas no precipitate was detected after transfection of either HDAC1-F or GAL4-Dnmt3L 1 --387 alone (lanes 2 and 3, respectively). Similar co-immunoprecipitation experiments were performed using GAL4-tagged Dnmt3L 53 --145, which contains the PHD-like motif, and Flag-tagged HDAC1. As shown in Figure A (lane 6), Dnmt3L 53 --145 co-immunoprecipitated specifically with HDAC1-F. This is consistent with the finding presented in Figure B showing that residues 53 --145 mediate the interaction with IVT HDAC1. Taken together, these data indicate that Dnmt3L specifically interacts with HDAC1 both in vitro and in vivo, and that this interaction requires the conserved PHD-like motif of Dnmt3L. Dnmt3L associates with HDAC activity | The association of Dnmt3L with HDAC1 led us to expect that Dnmt3L would be associated with histone deacetylase activity. To test this, we determined whether a full-length GST fusion protein of Dnmt3L (residues 1 --387) could purify HDAC activity from HeLa nuclear extracts. Histone deacetylase activity was assayed by liquid scintillation counting and measured as [3H]-acetate released from an acetylated H4 peptide. As shown in Figure B (lane 2), GST-Dnmt3L 1 --387 precipitates deacetylase activity from HeLa nuclear extracts, whereas GST alone gave only background activity (lane 1). We used GST fused to the retinoblastoma protein (Rb), a protein known to associate with deacetylase activity (,,), as a positive control (Fig. B, lane 5). As the PHD-like domain of Dnmt3L mediates the interaction with HDAC1 (Figs and A), we next asked whether this domain was required for the association with HDAC activity. Figure B (lane 4) indicates that GST-Dnmt3L lacking the PHD-like motif (GST-Dnmt3L 1 --387 Delta53 --145) resulted in only background activity. In agreement with the requirement of the PHD-like finger for associated HDAC activity, GST-Dnmt3L 53 --145 purified significant histone deacetylase activity from HeLa nuclear extracts (Fig. B, lane 3). The association of Dnmt3L with deacetylase activity could also be observed in vivo. We transiently transfected GAL4-Dnmt3L 1 --387 in 293 cells, followed by immunoprecipitation with anti-GAL4 antibody. Immunocomplexes were assayed for histone deacetylase activity. As shown in Figure C (lane 2), GAL4-Dnmt3L 1 --387 immunocomplexes contained significant deacetylase activity compared with precipitates obtained from a control transfection with GAL4 alone (lane 1). GAL4 Rb, which was used as a positive control for the assay, precipitated HDAC activity (Fig. C, lane 5). Consistent with the finding that the PHD-like motif of Dnmt3L is required in vitro to associate with deacetylase activity (Fig. C), we find that exogenous expression of GAL4-Dnmt3L 53 --145 precipitated significant amounts of histone deacetylase activity (Fig. C, lane 3), whereas GAL4-Dnmt3L 1 --387 Delta53 --145 (that is deleted of the PHD-like motif) gave only background activity (Fig. C, lane 4). Dnmt3L actively represses transcription through the PHD-like motif | Dnmt3a and Dnmt3b have been shown to have repressive capacity using their PHD-like domain (,). Since the conserved PHD-like motif is present in Dnmt3L, we wished to determine whether Dnmt3L may have repressive potential and, if so, whether its PHD-like domain is responsible for the repression. To this end, we used GAL4-Dnmt3L 1 --387, which contains full-length Dnmt3L fused to GAL4 DBD 1 --147. Repressor activity was evaluated by transiently transfecting increasing amounts of the GAL4 fusion construct into U2OS osteosarcoma cells together with a reporter plasmid containing GAL4-binding sites upstream of the viral thymidine kinase promoter driving expression of a luciferase reporter gene (4XGAL4-TK-Luc; Fig. A). Fold repression was determined relative to the basal transcriptional activity of the reporter in the presence of GAL4(DBD) alone. Figure C shows that GAL4-Dnmt3L 1 --387 can repress this promoter in a dose-dependent manner. This repression was specific, as a reporter lacking the GAL4 sites was not affected by GAL4-Dnmt3L 1 --387 (data not shown). To address whether the PHD-like domain of Dnmt3L is responsible for the observed repression, we co-transfected the high basal reporter plasmid 4XGAL4-TK-Luc along with GAL4-Dnmt3L 53 --145 containing the PHD-like motif. As presented in Figure C, GAL4-Dnmt3L 53 --145 inhibited luciferase activity in a dose-dependent manner. Dnmt3L-mediated repression is sensitive to TSA and it co-represses transcription together with HDAC1 | Having shown that Dnmt3L can silence transcription using a minimal transferable repressor domain, we next wished to investigate the molecular mechanism governing this repression. HDAC activity has been shown to repress transcription . Since Dnmt3L associates with histone deacetylase activity and binds HDAC1 via its PHD-like motif (see above), we evaluated whether Dnmt3L 53 --145 and HDAC1 may act together to silence transcription. To this end, we performed transient transfections in 293 cells using the reporter plasmid 4XGAL4-TK-Luc. As illustrated in Figure B, expression of a limiting amount of GAL4-Dnmt3L 53 --145 (lane 2) or HDAC1 (lane 3) only slightly repressed transcription. In contrast, co-transfection of GAL4-Dnmt3L 53 --145 along with HDAC1 provided a synergistic repressive effect on transcription (Fig. B, lane 4). We next asked whether the repression mediated by GAL4-Dnmt3L could be relieved by treatment with the specific HDAC inhibitor TSA. As depicted in Figure C, the repressive effect observed with GAL4-Dnmt3L 53 --145 on reporter activity was relieved by the addition of TSA. Collectively, these data indicate that Dnmt3L contains a transcriptional repressor domain, encompassing the PHD-like motif, and requires the histone deacetylase HDAC1 to exert this repressive effect. Figure 2 | . . Dnmt3L co-immunoprecipitates with HDAC1 and purifies HDAC activity in vitro and in vivo using its PHD-like finger. (A) 293 cells were transiently transfected as indicated (+) with 3 microg of either pcDNA3GAL4 Dnmt3L 1 --387 or pcDNA3GAL4 Dnmt3L 53 --145 (encompassing the PHD-like motif), together with 3 microg of pcDNA HDAC1-F (expressing Flag-tagged HDAC1). Whole cell extracts were then precipitated with anti-GAL4 antibody (5C1) and the presence of HDAC1-F in the immunoprecipitates was visualized by western blot analysis using anti-Flag antibody (M2). HDAC1-F is indicated by an arrow on the right. (B) Dnmt3L associates with deacetylase activity in vitro. Equivalent amounts of GST (lane 1) or GST fusion proteins (lanes 2 --5) were incubated with HeLa nuclear extracts. After washing, the complexes were tested for histone deacetylase activity. Histone deacetylase activity is given as radioactivity (c.p.m.) released from an 3H-labeled acetylated histone H4 peptide. GST --Rb fusion protein (lane 5) was used as a positive control for the assay. The results shown are the average of at least two independent experiments with error bars displaying standard deviations. (C) Dnmt3L purifies deacetylase activity from transfected cells. Cells were transiently transfected with equal amounts of the indicated GAL4-tagged constructs. Whole cell extracts were immunoprecipited with anti-GAL4 antibody. After washing, the immune complexes were tested for histone deacetylase activity. Histone deacetylase activity is given as radioactivity (c.p.m.) released from an 3H-labeled acetylated histone H4 peptide. We used Gal4 --Rb fusion protein as a positive control for the assay. Figure 3 | . . Dnmt3L represses transcription through its PHD-like motif when fused to the GAL4 DBD. (A) Schematic representation of the reporter and the effector constructs used. The reporter construct 4XGAL4-TK-Luc contains four copies of GAL4 sites upstream of the thymidine kinase promoter driving expression of the luciferase gene. The effectors used are also indicated and consist of GAL4 DBD alone [GAL4(DBD)], GAL4 DBD fused to either full-length Dnmt3L (residues 1 --387) or Dnmt3L PHD (residues 53 --145). (B and C) U2OS cells were transiently transfected with 250 ng of 4XGAL4-TK-luc reporter with increasing amounts (50 --250 ng) of GAL4-Dnmt3L 1 --387 (B) or GAL4-Dnmt3L 53 --145 (C). Whole cell extracts were used in luciferase assays. The basal activity of the reporter is normalized to a value of 100%. Transfection efficiencies were normalized using beta-galactosidase activity. The results are the average of at least four independent transfections done in duplicate with error bars displaying standard deviations. Figure 4 | . . Dnmt3L requires the deacetylase HDAC1 for transcriptional silencing and represses transcription in a TSA-sensitive manner. (A) Schematic representation of the reporter and the effector constructs used. The reporter construct 4XGAL4-TK-Luc contains four copies of GAL4 sites upstream of the thymidine kinase promoter driving expression of the luciferase gene. Also indicated are effector constructs expressing only the GAL4 DBD [GAL4(DBD)], the GAL4 DBD fused to the PHD of Dnmt3L (residues 53 --145) and Flag-tagged full-length HDAC1 (HDAC1-F). (B) U2OS cells were transiently transfected with 250 ng of the 4XGAL4-TK-luc reporter and, as indicated, with 50 ng of GAL4 Dnmt3L 53 --145 and/or 500 ng of HDAC1-F. Cells were then harvested and assayed for luciferase. The basal activity of the reporter is normalised to a value of 100%. Transfection efficiencies were normalized using beta-galactosidase activity. The results shown are the average of at least two independent experiments with error bars displaying standard deviations. (C) Dnmt3L-mediated repression is sensitive to TSA. U2OS cells were transfected with 250 ng of the 5XGAL4-SV40-CAT reporter and 50 ng of GAL4 Dnmt3L 53 --145. Sixteen hours after transfection, cells were treated (lanes 3 and 4) or not (lanes 1 and 2) with the HDAC inhibitor TSA (200 mM). The basal activity of the reporter is normalised to a value of 100%. Transfection efficiencies were normalized using beta-galactosidase activity. The results shown are the average of at least two independent experiments with error bars displaying standard deviations. DISCUSSION : Here, we report the first biochemical analysis of Dnmt3L, a third member of the Dnmt3 family. We show that Dnmt3L functions as a transcriptional repressor through its ability to associate with the histone deacetylase HDAC1. We find that Dnmt3L interacts both in vitro and in vivo with HDAC1 and, consistent with this, it associates with histone deacetylase activity. In addition, we demonstrate that Dnmt3L actively represses transcription and that this repression requires the recruitment of the histone deacetylase HDAC1 and is sensitive to the HDAC inhibitor TSA. We and others have recently reported a close connection between histone deacetylation and the two Dnmt3 family members Dnmt3a and Dnmt3b (,). Thus, the present study reveals that transcriptional repression through recruitment of histone deacetylation is a conserved and shared feature within the Dnmt3 family. We find that the link between Dnmt3L and deacetylation is mediated by its conserved PHD-like motif. This cysteine-rich region shows high sequence similarity among all three Dnmt3 family members. In Dnmt3a and Dnmt3b, the PHD-like motif also functions as a transcriptional repression domain by recruiting histone deacetylase (,). Thus, together these data suggest that, within the Dnmt3 family, the conservation of the PHD-like fingers at the sequence level reflects a conservation at the functional level. It is worth noting that the PHD-like domain found in the Dnmt3 members most closely resembles a domain found in the ATRX gene that codes for a putative ATP-dependent chromatin-remodeling factor of the SNF2 family . Mutations in the ATRX gene are the cause of ATRX syndrome, an alpha-thalassemia X-linked mental retardation disease . It is interesting to note that the majority of mutations associated with ATRX syndrome are found in its PHD-like fingers , thus highlighting the biological significance of this domain. As mentioned above, Dnmt3a and Dnmt3b recruit HDAC to silence gene expression through their PHD-like motif (,). Recent studies have demonstrated that Dnmt1 also binds to histone deacetylase through its non-catalytic N-terminal portion ( --). Intriguingly, in both instances, the methyltransferase activity of the Dnmts seems to be dispensable for gene silencing. From these observations, a new concept is beginning to emerge concerning the functional roles of the Dnmts in that they are multi-faceted proteins with additional functions other than their capacity to methylate CpG dinucleotides. Dnmt3L lacks the key catalytic motifs characteristic of DNA methyltransferases and is not likely to function as an enzyme that methylates CpG sites. Dnmt3L, however, retains the PHD-like motif characteristic of the Dnmt3L family. Our finding that Dnmt3L can still recruit the HDAC repressive machinery, despite its lack of a methyltransferase domain, further supports the observation that in some instances the HDAC-associated functions of the Dnmts may be working independently of their methyltransferase activity. Recently, Dnmt3L knock-out mice have been generated by two groups (,). They both found that Dnmt3L+/ -- embryos derived from homozygous mutant oocytes show a lack of maternal genomic imprints, causing genes that are normally methylated and maternally repressed to lack methylation marks and to be expressed on both alleles. This indicates that Dnmt3L is required for the establishment of maternal methylation imprints during oogenesis. As stated above, Dnmt3L is most certainly not acting as a DNA methyltransferase and it is therefore likely that it functions as a regulator of methylation at imprinted loci. Interestingly, Hata et al. have recently shown that Dnmt3L interacts with the DNA methyltransferase Dnmt3a as well as Dnmt3b, and that the absence of Dnmt3a, like the lack of Dnmt3L, results in the loss of maternal imprints in mice. Thus, Dnmt3L may act as a co-factor for the Dnmt3a and Dnmt3b methyltransferases. Our finding that Dnmt3L is associated with HDAC activity could also be relevant to its role as a regulator of maternal imprints. It is indeed interesting to note that several imprinted genes that show loss of maternal-specific methylation imposed by Dnmt3L deficiency, such as Snrpn, have been reported to be normally regulated by histone deacetylation . The methylated and non-expressed maternal allele of the imprinted Snrpn gene is found to be underacetylated relative to the unmethylated paternal allele and treatment of cells with the HDAC inhibitor TSA results in transcriptional reactivation of the silent maternal allele . Thus, it is tempting to speculate that the HDAC-associated function of Dnmt3L may contribute to its role as a regulator of methylation at imprinted genes. This would imply that histone deacetylation can influence DNA methylation patterns. This notion is supported by the observation in the fungus Neurospora crassa that HDAC inhibition by TSA can induce DNA demethylation . Taken together, a model could be envisaged in which Dnmt3L would not only act as a co-factor for the Dnmt3a and Dnmt3b methyltransferases but would also recruit HDAC activity to maternally imprinted genes. The methyltransferase and deacetylase activities recruited by Dnmt3L would then act together to potentiate the repressed state. The targeting of the Dnmt3L-containing enzymatic activities to the relevant loci is unlikely to occur through direct binding to DNA, as Dnmt3L does not contain any obvious DNA-binding domain. It is more likely that, as was shown for Dnmt1 and Dnmt3a (,), Dnmt3L is recruited to DNA by an as yet unidentified DNA-binding repressor. In conclusion, our results further emphazise the close connection that exists between the Dnmts and histone deacetylation. All Dnmts identified so far have now been shown to silence gene expression by contacting histone deacetylases. The challenge for the future is to understand how a cell establishes and maintains a transcriptional repressive state through the use of multiple HDAC activities associated with the Dnmts. Backmatter: PMID- 12202753 TI - The switch region on Leishmania major chromosome 1 is not required for mitotic stability or gene expression, but appears to be essential AB - The Leishmania genome project reference strain, Leishmania major Friedlin, is trisomic for chromosome 1. The complete sequence of this chromosome has revealed that the genes are grouped into two large clusters of the polycistronic type, each borne by one DNA strand and located on each side of a 1.6-kb sequence often termed the switch region. Several hypotheses concerning the role of this switch region have been put forward (region of initiation of transcription for both gene clusters, origin of replication or centromeric sequence). In the present study, we have deleted this region on the three copies of chromosome 1 by sequential targeted replacements. The absence of the switch region did not alter the mitotic stability of the three deleted chromosomes. This region therefore does not appear necessary for chromosomal replication or segregation. However, during the third targeting round, which aimed at knocking out the last switch region, a fourth copy of chromosome 1 that retained this region appeared in all clones analysed. This suggests that the persistence of this switch region is necessary for parasite survival. We then showed that the presence/absence of the switch region did not act upon the expression of a resistance marker gene inserted beforehand into the left gene cluster of the same chromosomal molecule. This result suggests that the presence of this 1.6-kb sequence is not necessary for the expression of all genes on chromosome 1. Keywords: INTRODUCTION : Leishmania are protozoal parasites of the family of Trypanosomatids responsible for significant worldwide human morbidity and mortality. The clinical features vary from mild cutaneous lesions to fatal mucocutaneous or visceral involvement depending on the causative species. The genome structure of this organism has now been well explored; 36 heterologous chromosomes, sized from similar0.3 to 3 Mb, are characterised for all Old World species, totalling similar35 Mb (+-2 Mb) for a haploid genome (,). Although Leishmania is considered to be essentially diploid, aneuploidy (mainly in the form of trisomic chromosomes) has been observed in natural isolates: thus, the Leishmania major Friedlin strain (LmF), the reference strain for the genome sequencing project, possesses three homologues of chromosome 1 (,). From the complete sequence of this similar300-kb chromosome (GenBank accession no. AE001274), the three homologues appear to be almost identical in their coding regions; they are essentially distinguished by a size variation of similar30 kb in the right non-coding subtelomeric end between chromosomes 1a and 1b/c (,). The most remarkable finding of the Leishmania and Trypanosoma genome sequencing projects certainly is the gene organisation in these protozoa. Most of the genes are grouped into large (50 --300 kb) clusters of the polycistronic type grouped on the same DNA strand. These clusters can be transcriptionally divergent or convergent depending on the chromosome region . The structure of LmF chromosome 1 is one of the most conspicuous examples of this organisation; two clusters containing 29 and 50 coding DNA sequences (CDSs), each borne by one DNA strand and oriented toward one telomere, are joined by an A/T-rich 1.6-kb sequence, hereafter termed the switch region . Obviously, the particular position of this region as well as its structural features (e.g. A/T richness) have lead to several hypotheses concerning its potential role, whether it be the site of transcription initiation for both clusters (,), a replication origin (,) or even a centromere . Notably, a negative correlation between purine excess and cumulative GC skew with (here termed divergent) coding directions suggested that the switch region might be the replication origin on this chromosome . Here, we report the successive deletions of this region on chromosome 1 homologues and analyse the effects of these targeted replacements upon mitotic stability or transcription. MATERIALS AND METHODS : Cell line cultures and transfection conditions | The L.major Friedlin (LmF) strain (MHOM/IL/81/Friedlin) has been described previously . Promastigote culture and transfection experiments were as described previously , except for the selection of transfectant clones, which was done on RPMI1640 low-melting agarose plates (0.6%). Depending on the resistance gene(s) inserted, transfectants were selected with 30 microg/ml hygromycin B (Gibco Life Technologies), 10 microg/ml phleomycin (Euromedex) and/or 100 microg/ml puromycin (Euromedex). After 10 days, resistant cells were isolated and cultivated for 1 week, and their genomic DNA was analysed by pulsed field gel electrophoresis (PFGE). For the study of chromosomal stability, two clones, termed cHBP1 and cHBP2 (cHBP for cDelta1.6::HYG/Delta1.6::BLE/Delta1.6::PAC/1.6), were cultivated without drug pressure, subcloning was carried out at different time intervals and the subclones cultivated in parallel in the presence or absence of drugs. For the study of transcription, clones cDeltaccp::HYG/Delta1.6::PAC 1, 2, 3 and 4 were selected with puromycin alone. These clones were then maintained in drug-free medium during 32 generations and then transferred cultivated in parallel in the presence or absence of drugs. Cell growth curves were established by daily cell counts on a Thoma haemocytometer after cultures were seeded with 2 x 105 cells ml --1 at day 0. Construction of the transfection vector | The transfection vectors used for this study were derived from a home-constructed vector termed pVV (GenBank accession no. AF315645), the construction of which has been detailed elsewhere . In this study, the HYG , BLE and PAC genes (conferring resistance to hygromycin B, phleomycin and puromycin, respectively) were used as selection markers. The flanking DNAs were from the 5' and 3' untranslated regions of the L.major dihydrofolate reductase-thymidilate synthase (DHFRTS) gene . Two homologous recombining sequences were used for all knock-outs of the switch region, and two others for the insertion of the HYG gene into the CCP2 gene on the left part of chromosome 1: their positions on chromosome 1 (according to the complete sequence of the chromosome, GenBank accession no. AE001274) are 76471 --77383 and 79049 --79558, and 44147 --44578 and 44837 --45253, respectively. These recombining sequences were PCR amplified from LmF total DNA and introduced into both cloning sites of pVV vector. The resulting constructs were termed pVV-1.6/HYG, pVV-1.6/BLE, pVV-1.6/PAC and pVV-CCP2/HYG. DNA preparation and electrophoretic techniques | Chromosome 1 was separated by PFGE as described by Wincker et al. , using a voltage of 7.5 V/cm and pulse times of 35 s for 72 h. Chromosomal DNA preparation and enzymatic restriction analysis were as described previously. DNA probes | Chromosome 1-specific DNA probes P1.6, PCCP2, PORF20, P3 and P5 were PCR amplified from LmF total DNA, cloned into the pGEMT-Easy vector (Promega) and sequenced. Their locations on chromosome 1 are indicated by the position of the primers used for their PCR amplification, referring to the complete sequence of chromosome 1 (GenBank accession no. AE001274): probe P1.6 (77384 --79048) strictly encompasses the 1.6-kb switch region; probes PCCP (44837 --45270), PORF20 (50717 --51024), P3 (79559 --80066) and P5 (171328 -- 172306) correspond to segments of chromosome 1 located 33 and 26 kb to the left side, and immediately and 93 kb to the right side, of the switch region, respectively. Homologues of chromosome 1 were distinguished by probe ST515, specific for the 81-bp minisatellite sequence LiSTIR1 which is present in chromosome 1b/c and absent from the 1a homologue . The HYG, BLE and PAC probes were gel purified after electrophoresis of the corresponding enzyme-restricted pVV vector. The hybridisation conditions were as described by Ravel et al. . Analysis of the recombination events | The replacements of the switch region on chromosome 1 by the drug resistance genes were analysed by double restriction with NheI and SpeI, followed by Southern analysis. NheI was selected for this analysis as its restriction sites encompass a similar9-kb fragment comprising the switch region. The SpeI site had been introduced by the insertion of the linearised transfection vector (see Fig. ). The integration on the same chromosomal molecule of the HYG and PAC genes into CCP2 and the switch region, respectively, was verified by an SpeI restriction. Figure 2 | . . Restriction analysis of the targeted replacements of the 1.6-kb switch region on chromosome 1 homologues. (A) Schematic representation of the NheI restriction fragment encompassing the switch region on the wild-type chromosome 1 (top line) and the mutated homologues (bottom line). In the first, the switch region (1.6) is comprised in an NheI fragment of similar8.7 kb. The insertion of any of the three drug resistance genes (R) used in our study strictly replaces the switch region and introduces an SpeI site. The size of R is different for each of the three resistance genes. Therefore, an NheI --SpeI double restriction should reveal a different sized fragment (X) depending on the specific resistance gene inserted. 5' and 3' are the L.major DHFRTS 5' and 3' untranslated regions used in the vector construct. The scale of the schematic representation is not proportional. (B) An NheI --SpeI double digest of clone cHBP1 DNA was electrophoresed on an 0.7% agarose gel and subsequently Southern blotted and hybridised. Hybridisation with DNA probe P3 [shown in (A)] revealed four different sized fragments corresponding to four homologues of chromosome 1. Probe P1.6 (specific for the switch region) hybridised with an 8.7 kb fragment corresponding to the wild-type locus. The three remaining fragments (3.8, 4 and 4.4 kb in size) each hybridised with one of the drug resistance genes inserted (BLE, PAC and HYG, respectively); their respective size correlated with the expected size after replacement of the switch region by the resistance gene [X in (A)]. RESULTS : Targeted replacement of the switch region by drug resistance genes in the three copies of LmF chromosome 1 | Three homologues of chromosome 1 are present in LmF, sized 285 and 315 kb for chromosome 1a and chromosome 1b/c, respectively (both 1b and 1c therefore being indistinguishable on a PFGE karyotype). Three rounds of transfection were carried out in order to knock out the 1.6-kb switch region in these three homologues. Transfectant clones were studied by PFGE and Southern analysis to verify the successful integration of the resistance genes. After the first transfection round, the HYG gene integrated on chromosome 1b/c in clone cDelta1.6::HYG, termed cH , in replacement of the switch region locus (see below). A second transfection was realised on this clone and the BLE gene was then found integrated on chromosome 1a, with the switch region absent from this chromosome (clone cDelta1.6::HYG/Delta1.6::BLE, termed cHB; Fig. ). Lastly, clone cHB was transfected in order to replace the last copy of the switch region present on the third homologue of chromosome 1 (b/c) by the PAC gene. Twenty-four clones (termed cHBP) expressing the three drug resistances were then isolated and analysed as above. Three interesting points were then found. (i) The switch region was still present in all 24 clones, on chromosome 1b/c. (ii) In 23 clones, the PAC gene integrated at the level of chromosome 1a, but this did not knock out the BLE gene (see for example clone cHBP1 in Figure ). (iii) In only one clone (cHBP2), PAC integrated at the level of chromosome 1b/c, but without knocking out HYG or the switch region in these homologues . Two alternative hypotheses may explain these results: either two resistance genes had integrated the same chromosomal molecule (by a non-homologous recombination event) or the last transfection round had induced the formation of a fourth chromosome 1 homologue. We analysed the targeted replacements in more detail in order to settle this point. Restriction analysis of the switch region replacements | The replacement of the switch region by the different resistance genes should introduce an SpeI site immediately upstream of the resistance gene (Fig. A). The correct replacements were thus verified by NheI --SpeI double digestion of total genomic DNA (see Materials and Methods). In clone cHBP1 (Fig. B), three NheI --SpeI fragments of the expected sizes (3812, 4043 and 4442 bp) were recognised by DNA probes of the three resistance genes, BLE, PAC and HYG, respectively. This result rules out the hypothesis of non-homologous recombination described above, and confirms the insertion of the three markers on three distinct chromosome 1 homologues. In addition, hybridisation with the switch region probe revealed an 8.7 kb fragment corresponding to the NheI fragment on the wild-type chromosome (8676 bp from the complete sequence of chromosome 1) (see Fig. ). These four polymorphic fragments hybridised with probe P3 (part of the CDS immediately to the right of the switch region). Therefore, they correspond on the one hand to the three mutated homologous loci that each integrated a resistance gene and on the other hand to the conserved wild-type locus. We concluded that four chromosome 1 homologues were present in clone cHBP1: two of the size of chromosome 1a (bearing either the BLE or the PAC gene) and two of the size of chromosome 1b/c (bearing either the HYG gene or the switch region). Similar restriction data were obtained in clone cHBP2, suggesting the presence of only one chromosome 1a and three homologues of the size of chromosome 1b/c (data not shown). Mitotic stability of the switch region deletion mutant chromosomes | Clones cHBP1 and cHBP2 were cultivated without any drug pressure for 110 days (i.e. >130 generations in the conditions used here). Subcloning was carried out at different time intervals. The persistence of the resistance genes was directly tested by cultivating the subclones in parallel in the presence or absence of drugs. All the subclones tested conserved the three drug resistances. Moreover, the karyotypes of some of the subclones were verified by PFGE; all the subclones analysed exhibited a karyotype identical to that of the parental clone (not shown). Thus, the three mutant chromosomes bearing the switch region replacements proved mitotically stable in the absence of selective pressure. The presence of the switch region in cis therefore does not appear necessary to the mitotic stability of chromosome 1. Role of the switch region in the transcription of chromosome 1 | We then wished to further investigate the role of the switch region in the transcription of chromosome 1 gene clusters. In order to specifically identify the transcription of one of the homologues, in spite of the low nucleotide divergence between homologues in LmF , we inserted the HYG gene into the left gene cluster of chromosome 1a. The targeted sequence was the CCP2 gene, located at equal distance between the telomere and the switch region. HYG was inserted in the same orientation as CCP2. The selected clone (cDeltaccp::HYG) was then transfected with pVV-1.6/PAC, in order to replace the switch region by the PAC gene in a rightwards orientation (Fig. A). Four transfectants (cDeltaccp:: HYG/Delta1.6 ::PAC 1, 2, 3 and 4) that had integrated PAC in chromosome 1a were isolated. They were selected and cultivated without hygromycin. Southern analysis of the molecular karyotype of these clones (Fig. B) showed the insertion of both resistance genes and the loss of the switch region all on chromosome 1a. SpeI digestion followed by hybridisation with a probe specific for ORF20 (see Fig. A) revealed the expected 34 kb (34 453 bp) fragment (not shown), confirming that both resistance genes were located on the same chromosomal molecule in the four clones. The latter as well as cDeltaccp::HYG were then cultivated in parallel in the presence or absence of hygromycin. All growth curves proved identical (Fig. C), showing that the HYG gene was expressed in a similar manner whether the switch region was present or absent. Figure 1 | . . Targeted replacement of the 1.6-kb switch region by drug resistance genes in the three homologues of chromosome 1 in L.major Friedlin. (Top) The small chromosomes of the wild-type strain (W) and of different transfectant clones (cH, cHB, cHBP1 and cHBP2; see text for details) were separated on an ethidium bromide-stained PFGE gel. (Bottom) The same gel was Southern blotted and sequentially hybridised with DNA probes P1.6, HYG, BLE and PAC. After transfection of clone cHB by the PAC-containing vector, 24 clones were isolated. The 1.6-kb switch region was present at the level of chromosome 1b/c in all clones. In 23 of them (e.g. cHBP1 here), PAC integrated at the level of chromosome 1a, where BLE remained present. In only one clone (cHBP2), PAC integrated at the level of chromosome 1b/c where the switch region and the HYG gene are also present. No hybridisation was observed with all these probes outside of the chromosome 1 size class, in particular in the compression zone (not shown). Moreover, in all clones, chromosome 1-specific DNA probes CCP2 and P5 hybridised at the same level as chromosome 1a and 1b/c in the wild-type strain (not shown), showing that no major rearrangement occurred in these chromosomes. Figure 3 | . . Analysis of clone cDeltaccp::HYG/Delta1.6::PAC 1 bearing a double mutation on chromosome 1a. (A) Schematic map of the left part of chromosome 1a, including the two targeted sequences and the transfection vector sequences inserted; the arrows indicate the transcription sense orientation of the marker resistance genes. The wild-type chromosome 1 has only one SpeI site located at the right end of the chromosome (position 267 268 bp). The double insertion of HYG and PAC introduced two novel SpeI sites. The collinearity of both resistance genes on the chromosome was verified by the hybridisation of probe PORF20 on a similar34-kb SpeI fragment. (B) PFGE separation of the small chromosomes of strain LmF (a), clone cDeltaccp::HYG (b) and clone cDeltaccp::HYG/Delta1.6::PAC 1 (c). The gel was Southern blotted and hybridised sequentially with the P1.6, HYG and PAC probes. (C) Comparison of the growth kinetics of clones cDeltaccp::HYG (a) and cDeltaccp::HYG/Delta1.6::PAC 1 (b) and 2 (c) in hygromycin-containing medium over 7 days (J0 --J7). Identical curves were obtained in drug-free medium (not shown). DISCUSSION : In the present study, we have focused on the as yet unclear role of the 1.6-kb sequence located at the junction between the two large divergent gene clusters of L.major chromosome 1. We performed targeted deletions of this sequence on the three chromosome 1 homologues present in this strain. Nevertheless, we were unable to create null mutants for this switch region, as a fourth wild-type copy of chromosome 1 persisted after the last transfection round, leading to tetrasomy for this chromosome. Other authors have experienced similar difficulties in disrupting essential genes in the Leishmania genome, due to intrachromosomal amplification, creation of extrachromosomal elements , genomic rearrangements of the duplication --transposition type , or duplication of whole chromosomes . We therefore conclude that the persistence of this 1.6-kb sequence is essential for parasite survival. In a second part of the study, we examined possible changes in the expression of the HYG gene inserted into the left gene cluster of chromosome 1a after we deleted the switch region on the same chromosome. The potential role of this region, as well as a number of mechanisms in which it is supposedly involved (replication, transcription, segregation), are discussed below. The first hypothesis tested is whether this sequence is necessary for chromosomal stability, i.e. is a single replication origin or a centromeric sequence . Our results show that its presence in cis is not necessary for chromosomal stability. Indeed, chromosomes which did not bear this sequence were mitotically stable, hence they were able to replicate and segregate normally. Regarding centromeric function, this finding is in agreement with recent studies suggesting that the sequences involved in chromosomal segregation may be located in the subtelomeric part of chromosome 1 . Regarding chromosome replication, although our results do not support the presence of an origin of replication in the switch region, they do not formally rule out this hypothesis. Indeed, in Saccharomyces cerevisiae, the knock-out of a functional autonomous replicating sequence (ARS) can be compensated by the activation of a neighbouring ARS (,). Moreover, a trans-complementation by an ARS present on the other homologue might be responsible for the maintenance of a chromosome fragment without origin . Here, the difficulty encountered in completely knocking out the switch region from the genome might be explained by the second hypothesis. In this case, the only remaining 1.6-kb region (on the fourth homologue) might complement the absence of origin on the three others. More elements will obviously be necessary to settle this point. The second hypothesis that has been proposed about the role of the switch region is that it is a key element in the initiation/regulation of transcription of the two gene clusters . The difficulty in knocking it out of the genome might be due to the fact that all essential genes present on chromosome 1 would remain untranscribed on the mutated homologues. However, our results do not fit with the latter hypothesis. First, the three resistance genes replacing the switch region in the knockouts were expressed. Secondly, when we deleted the switch region on a homologue of chromosome 1 tagged with the HYG gene integrated in the left gene cluster, the growth kinetics of the double mutants showed that the absence of the switch region did not act upon the expression of this collinear gene. The persistence of the switch region might then have another more trivial explanation: its knockout may have disrupted the splice site of the first gene in each of the opposing polycistronic transcription units [similar to exopolyphosphatase and poly(A) export protein genes, respectively]. If either of these genes were essential, one would never generate a switch region null mutant. Finally, one cannot rule out another more speculative hypothesis: that the switch region, which does not appear to contain any open reading frames, might contain essential non-coding RNAs (e.g. the recently identified miRNAs) . On the whole, this work provides direct functional evidence that the main hypotheses about the role of the switch region (whether single origin of replication or region of initiation of the transcription of both chromosome 1 gene clusters) ( --) are not validated by experimental results. The bioinformatic analysis of five switch regions located on different chromosomes did not allow identification of structural elements strictly common to all, with the exception of a high intrinsic DNA curvature . Although their peculiar position between two convergent or divergent gene clusters made them candidates as major regulating/functional sequences, it is not impossible that their role is restricted to the transcription of the sole adjacent genes. The publication of the complete sequence of the LmF genome, predicted for late 2002 , should allow a more systematic structural and functional analysis of these regions, and thus help to define their possible role in chromosome function. Backmatter: PMID- 12202777 TI - Improved quantitative real-time RT --PCR for expression profiling of individual cells AB - The real-time quantitative polymerase chain reaction (rtqPCR) has overcome the limitations of conventional, time-consuming quantitative PCR strategies and is maturing into a routine tool to quantify gene expression levels, following reverse transcription (RT) of mRNA into complementary DNA (cDNA). Expression profiling with single-cell resolution is highly desirable, in particular for complex tissues like the brain that contain a large variety of different cell types in close proximity. The patch-clamp technique allows selective harvesting of single-cell cytoplasm after recording of cellular activity. However, components of the cDNA reaction, in particular the reverse transcriptase itself, significantly inhibit subsequent rtqPCR amplification. Using undiluted single-cell cDNA reaction mix directly as template for rtqPCR, I observed that the amplification kinetics of rtqPCRs were dramatically altered in a non-systematic fashion. Here, I describe a simple and robust precipitation protocol suitable for purification of single-cell cDNA that completely removes inhibitory RT components without detectable loss of cDNA. This improved single-cell real-time RT --PCR protocol provides a powerful tool to quantify differential gene expression of individual cells and thus could complement global microarray-based expression profiling strategies. Keywords: INTRODUCTION : Gene expression is regulated at the level of individual cells, with different cell types or different developmental stages of the same cell expressing distinct sets of genes. Thus, analysis of the gene expression pattern of individual cells is a desirable goal. Laser-based microdissection techniques enable the isolation of identified single cells from fixed tissue (,), while the patch-clamp method enables the harvesting of mRNAs from a single living cell (,). The latter technique is of particular importance as it permits correlation of the functional properties of individual cells with their specific gene expression profile . Reverse transcription of mRNA followed by polymerase chain reaction (RT --PCR) is a reliable method of detecting gene expression. To analyze mRNA expression from single cells without loss of low abundance mRNAs, it is common practice to use all of the total single-cell complementary DNA (cDNA) reaction mixture as template for subsequent PCR . It is of increasing importance to detect not only qualitative but also quantitative differences in gene expression levels, and several approaches have been used to obtain quantitative data from single-cell RT --PCR experiments. These include the addition of known amounts of competitor mRNA (,), serial dilution of single-cell cDNA pools , and fluorescence-based real-time quantitative PCR (rtqPCR) . The latter has the advantage that it is easy to perform, highly reproducible, and, importantly, has the sensitivity to amplify and quantify even a single DNA template molecule ( --). Thus, rtqPCR has become the method of choice for quantitative analysis of gene expression levels ( --). However, I and others have observed that components of the undiluted RT reaction mixture considerably distort the subsequent PCR amplification reaction, presumably by inhibiting Taq-polymerase (,,). As DNA quantification via rtqPCR depends critically on the presence of an unperturbed exponential phase of PCR amplification (,) this inhibition restrains the use of rtqPCR for quantitative single-cell gene expression profiling. To date, RT --rtqPCR analysis of single cells has therefore been limited to either diluted (,), or PCR pre-amplified single-cell cDNA pools. Both these strategies limit sensitivity. Here, I describe a simple single-cell cDNA precipitation protocol that overcomes these inhibition problems and thus enables the full application of rtqPCR in combination with established single-cell RT protocols ( --,,). MATERIALS AND METHODS : Generation of control cDNA and DNA standard | For tissue RT --PCR, RNA was prepared from the midbrains of three 13-day-old C57Bl/6J mice using the Micro-FastTrackTM Kit (Invitrogen). RT was performed with 500 ng of poly(A)+ RNA overnight at 37C in a total reaction volume of 10 microl, containing random hexamer primers (5 microM; Roche), dithiothreitol (DTT, 10 mM; Gibco BRL), the four deoxyribonucleotide triphosphates (dNTPs, 0.5 mM each; Pharmacia), 20 U of ribonuclease inhibitor (Promega) and 100 U of reverse transcriptase (SuperscriptTMII; Gibco BRL). For T4 gene 32 protein experiments, T4gp32 (ChimerX) was diluted to a concentration of 3.2 microg/microl with sterile water. Freshly diluted T4gp32 (1.6 microg) was added to the single-cell reaction mixture directly before the reverse transcriptase was added. Single-cell cDNA was kept at --70C until PCR amplification. cDNA was purified, diluted and quantified using a BioPhotometer (Eppendorf). For the calibration curve, a tyrosine hydroxylase (TH) cDNA fragment (accession no. M69200) was amplified using conventional PCR [forward primer (387 bp), CACCTG GAGTACTTTGTGCG; reverse primer (1525 bp), CCTGTG GGTGGTACCCTATG], purified (Qiaquick Gel Extraction and PCR Purification Kits; Qiagen) and quantified using a BioPhotometer (Eppendorf). The number of DNA molecules was calculated and the DNA was diluted accordingly in serial steps. Harvesting of single-cell mRNA and cDNA synthesis | Single-cell mRNA was harvested from dopaminergic neurons in acute mouse brain slices as previously described . Briefly, for patch-clamp recording and cytoplasm harvest, patch-clamp capillaries (baked overnight at 220C) were filled with 6 microl of RNase-free patch-clamp buffer (140 mM KCl, 5 mM HEPES, 5 mM EGTA, 3 mM MgCl2, pH 7.3; all chemicals were from Sigma-Aldrich). After electrophysiological recording of neuronal activity using the whole-cell configuration, the cytoplasm of the same cell was harvested via the patch-pipette, under visual control, without losing the gigaseal to prevent contamination with extracellular fluid. Subsequently, the pipette contents were expelled into a sterile 0.5 ml reaction tube (Biopure; Eppendorf), containing the RT reaction mixture. RNasin and reverse transcriptase were added immediately, and single-cell cDNA synthesis was performed as described above. Precipitation of single-cell cDNA | All chemicals were molecular biology grade and certificated to be RNase /DNA free. One microgram of glycogen (Ambion), 250 ng of polyC RNA (Amersham Pharmacia), 250 ng of polydC DNA (Amersham Pharmacia) and a 1/10 vol of 2 M sodium acetate pH 4.0 were added to the single-cell cDNA reactions, or to 10 microl of diluted cDNA (generated from brain tissue) in water. cDNA was precipitated overnight with 3.5 vol of ethanol (100%; Sigma) at --20C. After centrifugation for 60 min at 4C (13 000 g) the supernatant was discarded, 100 microl of 75% ethanol was added and after a second centrifugation (13 000 g at 4C) for 15 min, the supernatant was removed. The cDNA pellet was dried in a thermal heating block (Eppendorf) at 45C until all ethanol had evaporated, and then dissolved in 10 microl of sterile water (Eppendorf). In order to completely resolubilize the cDNA, it was incubated for 60 min at 45C prior to quantitative real-time PCR. TaqMan quantitative real-time PCR and data analysis | rtqPCR was performed as recently described using the GeneAmp 5700 instrument (Applied Biosystems). The TaqMan primer/hybridization probe real-time PCR approach uses a fluorescence resonance energy transfer probe as reporter system . Hybridization primer/probe assay specific for real-time PCR detection of TH (accession no. M69200) was optimized according to the recommended criteria using the Abiprism Primer express software (Applied Biosystems) and the 2x TaqMan hybridization-probe Mastermix (Applied Biosystems). The TaqMan Mastermix contains uracil-N-glycosylase (UNG). RT and PCR were performed with the same batches of enzymes, random hexamers and dNTPs for all cells analyzed. All pipetting steps were carried out in a flow hood (HERAEUS), with the same set of pipettes by the same person. Control DNA or purified single-cell cDNA (each in volumes of 10 microl) was used as a template in a 50 microl PCR reaction in 1x TaqMan Mastermix in the presence of 300 nM of the forward primer (TH-F1151, 5'-GAATGGGGAGCTGAAGGCTTA-3'), 300 nM of the reverse primer (TH-R1260, CTGCTGTGTCTGGTCAAAGG), and 125 nM of the specific probe (TH-probe1193, CTATGGAGAGCTCCTGCACTCCCTGTCA). The hybridization probe was 3' labeled with 6-carboxytetramethylrhodamine as quencher and 5'-labeled 6-carboxyfluoroscein as reporter dye. Real-time PCR was performed in a GeneAmp 5700 thermocycler (PCR program: 2 min at 50C, 10 min at 95C, 50 cycles; 15 s at 95C and 1 min at 60C) and analyzed with GeneAmp 5700, Excel and IGOR (Wavemetrics) software. After defining a baseline (normalized background fluorescence of cycles 6 --15) in a linear plot of relative fluorescence (Rf) against PCR cycle number, quantification of the initial template molecules was performed. The PCR cycle, where the increasing relative fluorescence (Rf) crossed a manually set detection threshold (Ct) was defined in a logarithmic plot of Rf values against PCR cycle numbers (Ct at Rf = 0.09 for all analyzed data). Calibration curves were generated in IGOR by plotting Ct values against respective numbers of DNA template molecules, cDNA concentrations or cDNA dilution factors. Overall efficiencies (E) of PCR were calculated from the slopes of the standard curves according to the E = 10( --1/slope) for serial dilution in steps of 10 [log(10) scale] or E = 2( --1/slope) for serial dilution in steps of 2 [log(2) scale]. E = 2 reflects a doubling of DNA in each PCR cycle over all dilution steps. Errors are given as standard deviations (SDs) of the means. Significance was defined according to P-values gained from either two-tailed t-test analysis, or for single-cell DeltaCt comparisons, one-tailed t-test analysis. RESULTS : A TaqMan primer/probe assay was designed and optimized to detect TH, a marker gene expressed in monoaminergic neurons. Figure shows a representative result for an experiment in which different amounts of TH DNA (from 1 up to 106 molecules) were used as templates for rtqPCR. Relative fluorescence levels are plotted against PCR cycle numbers on both a linear scale (Fig. , top) and a semi-logarithmic scale (Fig. , middle). Figure illustrates that the PCR amplification kinetics are similar over the whole range of TH DNA template molecules. This is a precondition for the determination of Ct values and the generation of reliable calibration curves. The standard curve derived by plotting Ct values against DNA template molecule numbers has a calculated slope of --3.31, reflecting an overall PCR efficiency of 2.0 (mean slope for n = 5 experiments: --3.35 +- 0.09, E = 1.99 +- 0.04). This slope indicates a near perfect doubling of amplification products per cycle during the exponential phase of the PCR [the theoretical slope for an ideal PCR amplification is --1 / log(10)2 = --3.32]. When undiluted single-cell cDNA was used as PCR template, the results were dramatically different from those obtained using purified DNA , or cDNA in water (see Fig. A). The rtqPCR amplification kinetics varied from cell to cell in an unpredictable fashion, making the definition of the cycle threshold for detection problematic (Fig. A). These results appear to rule out reliable rtqPCR quantification of cDNA molecules using undiluted single-cell cDNA reaction mixtures as PCR templates. To identify the RT reaction component(s) that distort the rtqPCR amplification kinetics, the inhibitory effect of each of the individual RT reaction components was determined. These included DTT, RT buffer (containing random hexamer primers and dNTPs), the patch-clamp buffer, reverse transcriptase and ribonuclease inhibitor. Real-time PCR was performed with identical amounts of template cDNA (1 fM), generated from brain tissue, in either water or in the presence of an individual RT reaction component. The Ct value for cDNA in water was used as a reference to identify significant inhibition of cDNA-synthesis reaction components. The following Ct values were obtained: Ct = 34.44 +- 0.18 (n = 6) for cDNA in water; Ct = 34.31 +- 0.33 (n = 4) for cDNA in patch-clamp buffer; Ct = 34.44 +- 0.16 (n = 4) for cDNA in 2 U/microl RNasin; Ct = 34.29 +- 0.18 (n = 4) for cDNA in RT buffer; Ct = 35.03 +- 0.23 for cDNA in DTT (n = 4); Ct = 50.00 +- 0.0 (n = 6) for cDNA in reverse transcriptase. A two-paired t-test demonstrated that the TaqMan rtqPCR amplification was significantly compromised only in the presence of 10 U/microl reverse transcriptase (P << 0.000005) and to a much lesser degree in the presence of 10 mM DDT (P = 0.002). Interestingly, the inhibitory effect of the reverse transcriptase was higher in these control experiments than in single-cell RT --PCR experiments , resulting in complete rtqPCR inhibition (Ct = 50 after 50 PCR cycles). It has been reported that adding of the T4 gene 32 protein (T4gp32) to the RT reaction mixture, prior to RT, could remove the inhibitory effects of reverse transcriptase on subsequent PCR amplification (,). However, as illustrated in Figure B, addition of 1.6 microg of T4gp32 prior to single-cell cDNA synthesis did not improve the real-time RT --PCR amplification kinetics. Figure shows serial dilution of single-cell cDNA reaction mixtures. As illustrated, the rtqPCR amplification kinetics were not homogeneous across different cDNA dilutions, and varied both in the same cell (Fig. A, top and middle), and between individual cells (Fig. A, bottom). Thus, analysis of individual single-cell PCR calibration curves did not produce reliable values. The mean slope for the individual linear regressions was --3.45 +- 1.0 (n = 4), the high SD indicates the large scatter in the slopes of individual calibration curves. Figure B depicts the results of a similar set of experiments to those shown in Figure A, except that in this case the single-cell cDNAs were precipitated prior to serial dilution and rtqPCR. It is evident from the individual amplification plots that the kinetics of rtqPCR were independent of cDNA concentration both in the same cell (Fig. B, top and middle) and between different cells (Fig. B, bottom). The mean slope of the calibration curve was --1.03 +- 0.05 (n = 4) for the log(2) scale (ideal slope for serial dilution of 2 = 1.0), corresponding to --3.44 +- 0.16 on a log(10) scale and an efficiency of 1.96 +- 0.06. Importantly, neither the slope of the calibration curve nor the overall PCR efficiency was significantly different when precipitated single-cell cDNA or purified TH DNA was used as PCR template [P = 0.28 (slope) and P = 0.34 (E)]. I next tested whether the purification protocol leads to any loss of low concentration cDNA, which would bias quantification. Sets of serial dilutions of brain cDNA (10 microl each) were either used directly for rtqPCR or were precipitated and redissolved in 10 microl of water prior to rtqPCR. Figure A shows that the amplification plots and calibration curves obtained using these two protocols were almost identical, indicating that precipitation results in no significant loss of template cDNA over the full range of concentrations tested (102 --105 aM). The mean values between the Ct value for non-precipitated and precipitated DNA (DeltaCt) was --0.04 +- 0.15 (for n = 4 dilution steps). The slopes of the calibration curve obtained for non-precipitated or precipitated cDNA were --3.55 +- 0.17 (n = 4, E = 1.92 +- 0.06) and --3.62 +- 0.15 (n = 4, E = 1.89 +- 0.05), and were not significantly different [two-paired t-test: P = 0.54 (slope) and P = 0.52 (E)]. There was also no difference between the calibration curves generated from a single-cell cDNA and from brain tissue cDNA [P = 0.15 (slope) and P = 0.36 (E), respectively]. To evaluate the reproducibility of the precipitation method at the level of the single cell, cDNA reaction mixtures from individual cells were split after RT into two halves, and both halves were separately precipitated and used for comparative rtqPCR. Figure B illustrates the high reproducibility of the single-cell cDNA precipitation protocol. The Ct values of single-cell duplicates were very similar (mean DeltaCt = 0.27 +- 0.16, n = 6) and the mean DeltaCt was not significantly different from those of standard replicates, calculated to have 10 DNA molecules as PCR template (compare Fig. , mean DeltaCt = 0.31 +- 0.13, n = 6, P = 0.32, one-tailed t-test). In contrast, if only one half of the single-cell cDNA reaction mixture was precipitated, and the other was not, the Ct values were very different from cell to cell (mean DeltaCt = 1.73 +- 1.33, n = 6; the high SD indicates the large scatter of DeltaCt values), and DeltaCt values were significantly higher than those of standard replicates (P = 0.01, one-tailed t-test). These results further illustrate that the extent of PCR inhibition is not predictable using unpurified single-cell cDNA reaction mixtures as templates for rtqPCR. In contrast, the precipitated single-cell cDNA behaves like the purified TH DNA template during rtqPCR amplification. Taken together, absolute quantification of single-cell cDNA molecules using real-time PCR is possible, without diluting the single-cell cDNA reaction mixture, providing a quantitative cDNA precipitation step is included in the protocol. Figure 1 | Real-time fluorescent RT --PCR standard curve for TH cDNA quantification. Real-time fluorescent RT --PCR standard curve for TH cDNA quantification. Top, sensitivity of the real-time fluorescent RT --PCR protocol for TH. Relative fluorescence intensities (Rf) were plotted against PCR cycle numbers for five different calculated numbers of TH dsDNA template molecules (ranging from 1 to 1 000 000 molecules, as indicated). Middle, same data as in the top panel are plotted on a logarithmic Rf scale, for better illustration of the exponential phase of the PCR. Slopes of the exponential amplification phases were highly reproducible and independent of template concentration. Bold line (Ct) indicates the fluorescence threshold of amplification detection at Rf = 0.09, manually set within the exponential PCR phase. Bottom, standard curve for TH cDNA quantification derived from data shown in the top and middle panels. Ct value is defined as the cycle number where the relative fluorescence crosses the set threshold of amplification detection. Mean Ct values for different numbers of TH template molecules are plotted against their respective numbers of template molecules on a logarithmic scale. The linear regression fit (r = 0.999) was highly reproducible (mean slope = --3.31 +- 0.09 SD, n = 3) and defined the intercept at PCR cycle = 40.76 for a single TH dsDNA molecule. Figure 4 | No loss of cDNA occurs during the precipitation procedure. No loss of cDNA occurs during the precipitation procedure. (A) rtqPCR of TH with serial dilutions (steps of 10) of cDNA in water, generated from brain tissue, as templates. Red traces, precipitation of cDNA prior to PCR; black traces, no precipitation of cDNA prior to PCR. Linear and logarithmic amplification plots reveal almost identical amplification kinetics (top/middle) over the full range of tested cDNA concentrations. The respective calibration curves (bottom) had very similar slopes and they were similar to the standard curves shown in Figure . (B) Reproducibility of rtqPCR following single-cell cDNA precipitation. Amplification plots for TH and Ct values of cDNA generated from a single cell, split into two reactions prior to precipitation and rtqPCR, are almost identical. Figure 2 | Real-time PCR using undiluted single-cell cDNA as template. Real-time PCR using undiluted single-cell cDNA as template. (A) Amplification plots for TH rtqPCR for three individual cells utilizing undiluted cDNA reactions as templates. Relative fluorescence intensities (Rf) were plotted against PCR cycle numbers on a linear (top) and logarithmic (bottom) scale, illustrating inhomogeneous amplification kinetics between different single cells and compared with the standard (compare Fig. ). (B) Similar experiment as shown in (A). However, single-cell cDNA synthesis was performed in the presence of the T4 gene 32 protein. Figure 3 | Real-time PCR of TH for cDNA derived from single cells. Real-time PCR of TH for cDNA derived from single cells. (A) Top/middle, amplification plots (linear/logarithmic y-axis) of TH rtqPCR with serial dilutions (steps of 2) of cDNA derived from a single cell as template (dilutions: a = 1:2, b = 1:4, c = 1:8, d = 1:16, e = 1:32, Ct at 0.09). Note the diverging slopes of individual amplification kinetics that also differs from those of controls (compare Figs and A) preventing quantification of single-cell TH cDNA molecules. Bottom, linear plot of Ct values against dilution factor of cDNA derived from individual cells, further illustrating the unpredictability of altered amplification kinetics (data points for cell 1 correspond to amplification plots shown in the top and middle panel). (B) Similar experiment as shown in (A), but single-cell cDNA was precipitated and redissolved in water prior to rtqPCR. Top/middle, slopes of the exponential amplification phases (Ct = 0.09) were highly reproducible and independent of template concentration. Bottom, linear plot of Ct values against dilution factor of cDNA derived from individual cells, further illustrating the robustness and reproducibility of results (data points for cell 3 correspond to amplification plots shown in the top and middle panels). DISCUSSION : Quantitative analysis of gene expression at the level of individual cells is of particular relevance in tissues that contain a large variety of different cell types in close proximity, like the brain. Determination of the exact amount of a cDNA related to a specific gene using rtqPCR requires the definition of a cycle threshold for detection (Ct) in the exponential phase of the PCR amplification. For both relative and absolute quantification and comparison of gene expression between different samples, similar PCR amplification kinetics, especially in the exponential phase, are required for all samples analyzed. Relative quantification requires a housekeeping gene that shows constant expression levels between individual samples. The choice of such a gene is problematic at the tissue level and is likely to be even more difficult at the single-cell level. We have previously shown that the classic housekeeping gene beta-actin display large variations in cDNA expression levels between individual dopaminergic neurons . This indicates that beta-actin, at least at the level of single dopaminergic neurons, is not suitable for relative quantification. In order to allow absolute quantification of cDNA molecule numbers, a standard curve is required for the respective gene, which is generated by using defined numbers of DNA molecules as rtqPCR templates (compare Fig. ). In addition, the amplification kinetics and efficiencies of PCR reactions for sample DNA and for standard DNA need to be similar (,,,). In general, it is problematic to extrapolate from the number of cDNA molecules determined, to the absolute number of corresponding mRNA transcripts, as the efficiency of reverse transcriptase depends on the secondary structure, the specific RNA/protein complexation, and the absolute number of the respective individual target mRNA species (,,). I show here that the requirements for reliable rtqPCR quantification of TH cDNA molecules were clearly not met when undiluted single-cell cDNA reaction mixtures were used as templates; rtqPCR amplification kinetics were altered in a non-systematic fashion from cell to cell (compare Fig. ). Serial dilution experiments of single-cell cDNA reaction mixtures revealed that the rtqPCR amplification kinetics were also inhomogeneous across different cDNA dilutions, not only for the same cell, but also between different individual cells (Fig. A). Inhibition of the PCR amplification reaction, and thus altered PCR amplification kinetics, has been found to be a major problem for quantitative PCR (,,). Mathematical procedures to adjust altered PCR amplification kinetics are only valid if PCR kinetics are distorted in a systematic and thus predictable manner (,,,). Thus, mathematical procedures to correct for the different PCR amplification kinetics found for undiluted single-cells cDNA reaction mixtures are not applicable. To find an experimental way to overcome rtqPCR inhibition when undiluted cDNA reaction mixtures are used as templates, I analyzed which component of the cDNA reaction mixture was responsible for PCR inhibition. I found that the presence of the reverse transcriptase itself in the rtqPCR reaction dramatically inhibited PCR amplification. PCR inhibition by reverse transcriptases has previously been described, and reported to be particularly significant at lower concentrations of template DNA (<105 molecules) . It has been reported that the addition of the T4 gene 32 protein prior to RT increases the efficiency of reverse transcriptases and could overcome PCR inhibition . However, I found that addition of T4gp32 prior to single-cell cDNA synthesis did not result in improved amplification kinetics of subsequent rtqPCR (Fig. B). Heat inactivation of the reaction mix after cDNA synthesis, to destroy secondary structure and thus enzymatic activity, has previously been shown not to be preventing the inhibitory effects of reverse transcriptase on PCR . This is in accordance with my results, as all rtqPCR reactions were incubated for 10 min at 95C prior to PCR amplification to inactivate the UNG enzyme of the TaqMan Mastermix. Decreasing the amount of reverse transcriptase to reduce its inhibitory effect on PCR is expected to result in loss, or under-representation, of low abundance mRNAs from single cells because the efficiency of the cDNA synthesis is reduced at low mRNA concentrations . Dilution of the single-cell cDNA reaction (>one-eighth-fold, compare Fig. A) resulted in amplification kinetics that were very similar to those performed with either cDNA in water, or purified DNA as PCR template. However, dilution of single-cell cDNA will inevitably lead to loss of low abundance cDNAs and thus compromise the application of single-cell rtqPCR for the study of low abundance mRNAs. I describe here a simple ethanol precipitation protocol that enables quantitative precipitation and recovery of very low concentrated cDNA, and that removes all rtqPCR-inhibiting components. Ethanol precipitation was carried out in the presence of sodium acetate, in accordance with the original protocol . Precipitation of cDNA was carried out overnight at --20C, as precipitation at --70C has been found to inhibit complete DNA precipitation when ammonium acetate was used as source of monovalent cations (,). The addition of the carrier nucleic acids polyC and polydC allowed quantitative precipitation of low concentrated DNA without affecting subsequent rtqPCR, as illustrated in Figure A. Glycogen was added to increase the reproducibility of cDNA recovery. I did not test linear acrylamide as co-precipitants as there were no problems either with DNA or RNase contamination of glycogen or with re-dissolving of precipitated cDNA. Linear acrylamide has the advantage of being derived from a non-biological source, however, it contributes to absorbance at 260 and 280 nm . Experiments using the same amount of highly diluted cDNA with or without precipitation prior to rtqPCR allowed direct monitoring of any cDNA loss due to the precipitation procedure and demonstrated that this did not happen (Fig. A). PCR amplification plots, PCR efficiencies, and the slopes of calibration curves generated from precipitated single-cell cDNA were very similar to those using purified TH DNA or brain cDNA as templates. Taken together, single-cell cDNA after precipitation fulfills all the requirements for absolute quantification of TH cDNA molecules at the level of individual cells. The protocol enables rtqPCR in combination with established specific RT conditions for single-cell RNA, harvested after electrophysiological characterization ( --,,). With the method described here we have quantified single-cell cDNA expression after electrophysiological characterization for six different genes with high, medium and low expression levels, and identified a high correlation between cDNA transcript numbers and respective functional protein levels in the individual neurons . Furthermore, this approach provides a reliable method for cell-specific validation of tissue-based DNA microarray-generated global gene expression-profiling data (,). Backmatter: PMID- 12202778 TI - Cre recombinase-mediated inversion using lox66 and lox71: method to introduce conditional point mutations into the CREB-binding protein AB - CREB-binding protein (CBP) is a multifunctional cofactor implicated in many intracellular signal transduction pathways. We aimed to investigate the involvement of CBP in the cAMP response element-binding protein (CREB)-mediated pathway. The point mutation Tyr658Ala in the CREB-binding domain (CBD) was shown to abolish the binding activity of CBP to phospho-CREB, the activated form of CREB. By using a mutant Cre/loxP recombination system, this point mutation was aimed to be generated in the mouse genome in a tissue- and time-specific manner. A targeting construct in which CBD exon 5 and inverted exon 5* containing the point mutation flanked by two mutant loxP sites (lox66 and lox71) oriented in a head-to-head position was generated. When Cre recombinase is present, the DNA flanked by the two mutant loxP sites is inverted, forming one loxP and one double mutated loxP site. As the double mutated loxP site shows low affinity for Cre recombinase, the favorable reaction leads to a product where the mutated exon 5* is placed into the position to be correctly transcribed and spliced. Inversion was observed to be complete in both bacteria and mouse embryonic stem cells. Our results indicate that this Cre- mediated inversion method is a valuable tool to introduce point mutations in the mouse genome in a regulatable manner. Keywords: INTRODUCTION : Tissue- and time-restricted genetic modifications in mice can be generated using the Cre/loxP recombination system . Cre is a 38 kDa site-specific DNA recombinase isolated from bacteriophage P1. loxP (locus of crossover in P1) was originally found in the phage P1 genome and is 34 bp in length, consisting of two 13 bp inverted repeats flanking an 8 bp non-palindromic core sequence that determines the polarity of the loxP site . Cre-mediated recombination between two directly repeated loxP sites results in an irreversible excision of the intervening sequence, while recombination between two loxP sites that are positioned in a head-to-head orientation leads to an inversion, which takes place continuously as long as Cre recombinase is present. Recently, various mutant loxP sites were generated and characterized , of which lox66 and lox71 are particularly interesting. After the first Cre-mediated recombination, one wild-type loxP site and one double mutant loxP site are formed; the latter, however, displays a very low affinity for Cre recombinase. Consequently, Cre-mediated recombination in the mutated Cre/loxP system has a favorable forward reaction equilibrium, as depicted in Figure . This mutant Cre recombination system was successfully tested for targeted integration into the mouse genome . However, the efficiency of inversion mediated by this system has not yet been evaluated in the context of the mouse genome. In the present study, we tested whether, in principle, this system could be used to introduce point mutations into CREB-binding protein (CBP) in a tissue- and time-restricted manner using the mutant Cre/loxP recombination system described above. CBP is a transcriptional cofactor implicated in many different intracellular signal transduction pathways . To dissect CBP function specifically in the cAMP response element-binding protein (CREB)-mediated pathway, we aimed to introduce a point mutation (tyrosine to alanine at amino acid residue 658 of mouse CBP, CBPTyr658Ala), which was reported to abolish the binding activity of CBP to phosphorylated and hence activated CREB , presumably leaving intact the activities of the other signaling pathways where CBP plays crucial roles. Figure 1 | Schematic representation of the mutated Cre/loxP system. Schematic representation of the mutated Cre/loxP system. Nucleotide sequences of loxP and its mutated derivatives are listed; mutated sequences are underlined. Arrows indicate non-palindromic core sequence. Cre-mediated recombination between lox66 and lox71 sites generates one wild-type and one double mutant loxP site. Since the double mutant loxP site exhibits much reduced binding affinity for Cre recombinase, Cre- mediated inversion using the mutated Cre/loxP system prefers the forward reaction as indicated. MATERIALS AND METHODS : Plasmids | The mutant loxP sites, lox66 and lox71, were synthesized by PCR with two pairs of partly overlapping primers, Z24 and Z25, and Z26 and Z27, respectively. The following primers were used: Z24, 5'-gcc gaa gct tct cgt gat aac ttc gat tag cat aca tta tac gaa cgg t; Z25, 5'-acc gga att ccc gga cta ccg ttc gta taa tgt atg ct; Z26, 5'-gcc gaa gct tct cgt gta ccg ttc gta tag cat aca t; Z27, 5'-acc gga att ccc gga cat aac ttc gta taa tgt atg cta tac gaa cgg t. HindIII and EcoRI sites (underlined sequences) were introduced concomitantly and used to clone lox66 and lox71 into plasmid vector pBSIIKS( --) (Stratagene), respectively. The correct clones were confirmed by restriction digestion and sequencing. The plasmid CBPTyr658Ala --Neo (Fig. A) was the final targeting construct, in which the region containing exon 5 of the CREB-binding domain (CBD), an inverted exon 5 with the point mutation (tyrosine to alanine) at amino acid residue 658 and the positive selection marker PGK --Neo (phosphoglycerate kinase --neomycin phosphotransferase) were flanked by lox66 and lox71 sites in a head-to-head orientation. The negative selection marker tk (thymidine kinase) was put at the 3' end of the targeting construct. The lox66 site was inserted into CBD intron 4 through a SpeI site. Sequences containing the inverted exon 5*, the PGK --Neo cassette flanked by two frt sites and the lox71 site were inserted into intron 5 through the EcoRV site. DraIII and NdeI sites were used for the margins of the left and right homology arms, respectively (Fig. A). The mutant exon 5* was generated by PCR with primers Z20 (5'-ttt gta ttt tc*g* c*ga ttt tct ctg ct), Z21 (5'-gca gag aaa atc g*c*g* aaa ata caa aaa ga), Z22 (5'-tag ctg tct cca gac act cca gaa) and Z23 (5'-aaa gac act tgc cag cac acc ctt). Mutations were introduced by the nucleotides marked with asterisks, generating a new NruI site. First, two PCRs with lambda phage DNA containing the mouse genomic CBD region of CBP and primers Z22 and Z20, and Z21 and Z23, were performed. Both products were used as templates for a second PCR with primers Z22 and Z23. The product of this PCR was subcloned into plasmid pBSIIKS( --). The correct clone was confirmed by restriction digestion and sequencing. Primers Z33 (5'-ggt cca gct ggc aaa cta aat gat ggg ctc tca) and Z34 (5'-aac tcg agt tga tat caa aga gca cta agc aaa caa gct), containing PvuII, XhoI and EcoRV sites (underlined sequences), were used to shorten the mutant exon 5* clone and in subsequent cloning steps, as longer inverted repeat sequences could not be cloned. In the plasmid pCre --pac , the Cre recombinase gene was cloned into vector pUC9 containing the puromycin resistance gene (Pac) driven by the PGK promoter. Plasmid 705-Cre was used to express Cre recombinase in Escherichia coli. The plasmid pSVaZ11-PGK --Neo contains the PGK --Neo cassette flanked by two frt sites. The plasmid pUSEFUL contains the tk gene. Cre recombinase-mediated inversion in bacteria | Plasmid 705-Cre (containing the chloramphenicol resistance gene) was transformed into competent DH5alpha bacteria, which were in turn made transformation competent using polyethylene glycol. Plasmid CBPTyr658Ala --Neo (containing the ampicillin resistance gene) was then transformed into these competent Cre-DH5alpha bacteria. Colonies were inoculated in 2 ml of LB medium containing ampicillin (100 microg/ml) and chloramphenicol (25 microg/ml), and the cultures were incubated overnight with shaking at 30C. Cre recombinase-mediated inversion was examined by restriction enzyme digestion of isolated DNA using Acc65I and NruI. Embryonic stem (ES) cell cultures | ES cells (clone E14) were cultured in Dulbecco's modified Eagle's medium (DMEM, high glucose, plus sodium pyruvate) (Gibco, Germany) with 15% heat-inactivated fetal calf serum (Roche, Germany), 2 mM l-glutamine (Gibco), 1/12 000 (v/v) leukemia inhibitory factor (Gibco), 100 U/ml penicillin/100 microg/ml streptomycin (Gibco) and 50 microM beta-mercaptoethanol (Gibco), incubated at 37C in 5% CO2. Targeting construct CBPTyr658Ala --Neo was electroporated into ES cells and screened for homologous recombinant clones according to standard procedures. For Cre recombinase-mediated inversion in ES cells, two of the recombinant cell clones were expanded separately on Petri dishes containing mouse embryonic fibroblast feeder cells. Lastly, cells were harvested from 10 cm gelatinized dishes for electroporation. For each recombinant, 1 x 106 ES cells in 800 microl of phosphate-buffered saline were prepared, transferred into the electroporation cuvette (Gene Pulser cuvette, 0.4 cm electrode gap; Bio-Rad, USA) and mixed with 20 microg plasmid pCre --pac. Electroporation was performed at 240 V, 500 microF in a Bio-Rad gene pulser. After electroporation, the cuvette was put onto ice and incubated for 15 min. The cell suspension was then distributed evenly onto three 10 cm feeder dishes, fed with DMEM complete ES cell medium and incubated overnight at 37C in 5% CO2. Cells were then fed daily with fresh complete ES cell medium plus 1 microg/ml puromycin for selection of Cre recombinase-transfected cells for 2 days, then fed daily with medium without puromycin until formation of colonies. Finally, 96 single colonies for each cell clone were picked and expanded on gelatinized 96-well plates for Southern blot analysis. Southern blot analysis | Southern blots with the Neo probe were carried out according to standard procedures. The Neo probe was the 300 bp fragment from a XbaI digestion of plasmid pSVaZ11-PGK --Neo. Figure 2 | Cre recombinase-mediated inversion in bacteria. Cre recombinase-mediated inversion in bacteria. (A) Targeting construct CBPTyr658Ala --Neo was transformed into competent DH5alpha bacteria expressing Cre recombinase (plasmid 705-Cre) and inversion was induced at 30C. (B) Restriction pattern analysis after Acc65I/NruI double digestion of clones before and after transformation revealed complete and irreversible Cre recombinase-mediated inversion. Figure 3 | Cre recombinase-mediated inversion in ES cells. Cre recombinase-mediated inversion in ES cells. (A) ES cell clones containing the targeted CBPTyr658Ala --Neo allele were generated and analyzed by Southern blotting. Relevant restriction sites used for construction of targeting vector are listed: SpeI and EcoRV sites for insertion of lox66 and lox71, respectively; DraIII and NdeI sites as the margins of the 5' and 3' homology arms; Acc65I sites for screening for correct homologous recombinants. The tk cassette was inserted into a SacII site located in the polylinker region of the vector, adjacent to the end of the 3' homology arm. (B) Correctly recombined ES cell clones were electroporated with the Cre recombinase-expressing plasmid pCre --pac. (C) Southern blot analysis using the Neo probe showed a 6.6 kb band for the targeted CBP allele, a 4.6 kb band for the inverted targeted allele of CBP and both bands for an incomplete inversion. Two examples of inversion are shown. As evaluated by densitometric quantification of the autoradiography film, the inversion ratio was 64% (left) and 100% (right). RESULTS : Generation of targeted CBP locus | To achieve the Cre recombinase-mediated mutation CBPTyr658Ala, we generated a targeting construct containing the wild-type exon 5 of the CBD and a mutated exon 5* (Tyr658Ala) in an inverted orientation. These sequences were flanked by lox66 and lox71 sites, which were positioned in a head-to-head orientation. The notion was that after removal of the Neo selection cassette, Cre recombinase-mediated inversion would place the mutated exon 5* into a position where it would be spliced properly, while the inverted wild-type exon 5 would be spliced out. After several cloning steps using mouse genomic DNA of the CBP locus isolated from mouse strain 129, the targeting construct CBPTyr658Ala --Neo was obtained. A schematic representation of the region encoding the CBD region of CBP and several restriction sites used for cloning are shown in Figures A and A. SpeI and EcoRV sites were used for insertion of the lox66 and lox71 sites, respectively. Inverted mutant exon 5* and the PGK --Neo cassette flanked by two frt sites were introduced concomitantly with the lox71 site. DraIII and NdeI sites formed the margins of the 5' and 3' homology arms, respectively. The tk cassette was inserted into a SacII site located in the polylinker region at the end of the 3' homology arm (plasmid data not shown). Integrity of the targeting construct was confirmed by sequencing of all cloning junctions, recombination sites, exon 5 and mutant exon 5*. Cre-mediated inversion in bacteria | We next tested the targeting construct CBPTyr658Ala --Neo in bacteria regarding Cre-mediated inversion through the lox66 and lox71 sites. For this, the targeting construct was transformed into bacteria expressing Cre recombinase. DNA was isolated from five transformed colonies and digested with Acc65I and NruI. All five DNA samples analyzed displayed a digestion pattern with four bands of 1.0, 1.7, 4.2 and 8.0 kb, while the digestion of CBPTyr658Ala --Neo before inversion showed bands of 1.0, 3.8, 4.2 and 5.9 kb (Fig. B). According to the restriction map of the targeting construct and its Cre-mediated inversion product (Fig. A), this result indicates that the sequence flanked by lox66 and lox71 sites was completely inverted and that the sequence after inversion containing one wild-type loxP and one double mutated loxP no longer appeared to be recognized by Cre recombinase (Fig. B). Cre-mediated inversion in ES cells | To investigate the lox66 and lox71 sites regarding Cre-mediated inversion in the mouse genome, targeting construct CBPTyr658Ala --Neo was electroporated into ES cells. Several clones containing the correct homologous recombination event were characterized by Southern blot analysis (Fig. A and data not shown). Two of these clones were then transiently transfected with the plasmid pCre --pac, in order to express Cre recombinase and to induce concomitant inversion. Southern blot analysis of HindIII-digested DNA from these transfected ES cell clones was carried out using the Neo probe (Fig. B). The targeted CBPTyr658Ala --Neo allele gave one band of 6.6 kb. After transient transfection with Cre recombinase, some cell clones showed only one 4.6 kb band for the inversion product (Fig. C), indicating that inversion was complete and apparently irreversible. Other clones that were analyzed by densitometric evaluation of the autoradiography film revealed values of the inverted band (4.6 kb) lower than 100% (e.g. 64% in Fig. C), suggesting that incomplete inversions can also occur. DISCUSSION : Our results showed that the inversion system using lox66 and lox71 sites represents a novel and very valuable tool to introduce conditional point mutations into transcription units of the genome. While the inversion was always complete and irreversible in bacteria, we observed different efficiencies of inversion patterns in ES cells, ranging from 64 to 100%. It could be that inversion is not as efficient in ES cells as in bacteria, in agreement with another report where the evaluation of inversion was performed using a luciferase assay in transient transfections of cell lines. Cre-mediated integration was shown to be 4- to 7-fold more efficient between lox66 and lox71 than between double mutant loxP and wild-type loxP . The variabilities we observed in our experiments using ES cells and inverting a genomic locus might be caused by the different levels of Cre recombinase expression and/or by the different duration of Cre recombinase action in the transiently transfected clones. At any rate, we were able to clearly show that it is possible to get up to 100% inversion using this method. It remains to be investigated how efficient inversion will be in the living mouse, by crossing the CBPTyr658Ala --Neo allele with Cre recombinase-expressing transgenic mice. The method presented here is very appropriate to study CBP function, since the CBP gene is very sensitive to gene dose. Mice lacking only one CBP allele display quite strong phenotypical changes and Rubinstein --Taybi syndrome patients, who have one defective CBP allele, also show many severe developmental defects . Thus, even if the inversion ratio might not reach 100% in all cells, inversion is very likely to be sufficient to introduce phenotypic changes and, thus, we would be able to analyze CBP function in the mouse by the generation of tissue-specific point mutations in CBP. Backmatter: PMID- 12202779 TI - Single nucleotide polymorphism genotyping using short, fluorescently labeled locked nucleic acid (LNA) probes and fluorescence polarization detection AB - Locked nucleic acids (LNAs) are synthetic nucleic acid analogs that bind to complementary target molecules (DNA, RNA or LNA) with very high affinity. At the same time, this binding affinity is decreased substantially when the hybrids thus formed contain even a single mismatched base pair. We have exploited these properties of LNA probes to develop a new method for single nucleotide polymorphism genotyping. In this method, very short (hexamer or heptamer) LNA probes are labeled with either rhodamine or hexachlorofluorescein (HEX), and their hybridization to target DNAs is followed by measuring the fluorescence polarization (FP) of the dyes. The formation of perfectly complementary double-stranded hybrids gives rise to significant FP increases, whereas the presence of single mismatches results in very small or no changes of this parameter. Multiplexing of the assay can be achieved by using differentially labeled wild-type and mutant specific probes in the same solution. The method is homogeneous, and because of the use of extremely short LNA probes, the generation of a universal set of genotyping reagents is possible. Keywords: INTRODUCTION : Numerous methods for single nucleotide polymorphism (SNP) genotyping have been described and have recently been reviewed (,). However, there is still continuing interest in the development of new approaches, especially those that would be suitable for high throughput genotyping experiments. With the completion of the Human Genome Project, it is anticipated that such projects will become of great importance in gene discovery and pharmacogenomics. We have been exploring the possibility of using peptide nucleic acids (PNAs) as probes for such high throughput genotyping applications and have recently reported two fluorescence polarization (FP) based methods using short, fluorescein or rhodamine labeled PNA probes. In the first such method , we used fluorescein labeled PNA probes and detected their hybridization to target DNA molecules by FP in the presence of polylysine. Later, we discovered that the use of another fluorescent dye, rhodamine, does not require the presence of polylysine, because the hybridization of such probes resulted in very significant increases of the FP signal even in its absence . Unfortunately, there were several problems associated with the use of PNA probes. First, some of these were poorly soluble and precipitated upon storage. Second, the length of these probes had to be at least seven bases, because the lower affinity of shorter probes made the hybridization at room temperature unfavorable. Third, guanosine-rich, rhodamine-labeled probes showed very high intrinsic FP values, and these changed very little upon target hybridization. Because of these problems, we decided to evaluate another recently described class of nucleic acid analogs, locked nucleic acids (LNAs) . The main structural feature of these probes is the presence of an additional methylene bridge linking the 2' hydroxyl group of an RNA monomer to the 4' carbon of the ribose ring. The presence of this bridge 'locks' the sugar ring in one fixed conformation (3'-endo), which is the conformation observed for that ring in RNA or DNA hybrids. On the basis of NMR studies of LNA /DNA duplexes, it has been concluded that the main factor contributing to the extraordinary high stability of LNA-containing duplexes is a local change of the phosphate backbone geometry that favors a much higher degree of base stacking . LNA probes have already been used for solid-phase single nucleotide genotyping (,). In these methods, the authors covalently attached LNA probes to the surface of microtiter plates and used them to capture PCR products containing sequences surrounding the Leiden V and the ApoB 100 R3500Q mutation. An ELISA format was used to detect the hybridization event, and the authors reported excellent specificity and sensitivity of the assays. For high-throughput genotyping applications, a homogeneous method requiring as few post-amplification steps as possible would be of great interest. To our knowledge, no hybridization-based homogeneous method for SNP genotyping using LNA probes has been reported to date. In this article we present the first such approach. It is based on our observations that the FP values of a fluorescent dye such as rhodamine or hexachlorofluorescein (HEX) atached to an LNA probe increase significantly upon the hybridization of the probe to a target DNA molecule. Due to the very high affinity of the LNA molecules, very short probes can be used. In this article, we demonstrate that hexamer and heptamer LNA probes hybridize with very high affinity to perfectly complementary targets, and at the same time show an extraordinary specificity, allowing the discrimination of targets that differ by a single base. MATERIALS AND METHODS : LNA probes and other reagents | All LNA probes used in this work were synthesized by Proligo, LLC (Boulder, CO). The sequences of all probes used in this work are shown in Table . All LNA probes were 5' end-labeled with either rhodamine or HEX. The latter dye was attached to the 5' end of the LNA probes using its commercially available phosphoramidite (Glen Research, Sterling, VA). Because of the instability of rhodamine to the concentrated ammonia treatment during the deprotection step, this dye was attached to 5'-amino modified LNA probes after their complete deprotection with concentrated ammonia. The reactive form of the rhodamine was an N-hydroxysuccinimidyl ester, available from Molecular Probes (Eugene, OR). All probes were purified by ion-exchange HPLC to a final purity of at least 90%. Stock solutions were prepared in water and stored frozen between experiments. The concentrations of the labeled probes were determined based on absorbance measurements at the absorbance maxima of the dyes. Synthetic DNA molecules were obtained from Oligos, Etc. (Wilsonville, OR). Hybridization experiments | Hybridization experiments using synthetic DNA molecules were generally carried out in a buffer containing 50 mM HEPES pH 7.5, 100 mM NaCl. Room-temperature hybridizations and melting curve measurements were performed in a fluorescence spectrophotometer (Fluoromax-2 or Fluorolog-3, JY Horiba, Edison, NJ) equipped with polarizers and temperature bath. PCR and SNP genotyping | Thermostable DNA polymerase (Taq) was procured from Qiagen (Valencia, CA) and used in the manufacturer-supplied buffer. PCR primers FS (5' AGT CAA GGA CAC CGA GGA A, phosphorothioated at the first five bases) and R (5' GCT TCT TAC AGT GCT GGA TGT) were synthesized by Oligos Etc. The incorporation of one partially phosphorothioated primer into the double-stranded PCR product allowed the selective enzymatic hydrolysis of the opposite strand by T7 gene 6 exonuclease . The exonuclease was from USB Corporation (Cleveland, OH). Table 1 | Melting temperatures and FP dynamic ranges of 5'-Rho LNAs RESULTS : Detection of LNA /DNA hybridization by FP | We have previously reported the successful use of FP to detect the hybridization of rhodamine-labeled PNA probes to DNA targets . We have now extended these observations to include fluorescently labeled LNA probes. The two fluorescent dyes that we have found to give the best FP responses upon hybridization were rhodamine and HEX. Since LNA molecules have been reported to form exceedingly stable duplexes with complementary target nucleic acids, we decided to test the hybridization behaviour of very short, dye-labeled probes. Thus, a series of dye-labeled hexamer and heptamer probes were obtained from Proligo, LLC. Figure shows the results obtained when fixed concentrations of 50 nM of four different, rhodamine labeled LNA hexamers were titrated with increasing concentrations of synthetic target DNA molecules at room temperature. The targets were either perfectly complementary to the LNA probes, or contained a single nucleotide substitution that upon hybridization would give rise to a single nucleotide mismatch within the LNA /DNA duplex. A 50 nM probe concentration was chosen because this is within the same order of magnitude as the yield of a typical PCR amplification reaction. As expected, LNA probes with higher GC content (Fig. , LNA 601) were saturated at lower DNA target concentrations than probes that were more AT-rich. In order to evaluate the stability of the LNA /DNA duplexes, we performed melting curve measurements by recording the FP as a function of temperature for free LNA probes or mixtures of LNA probes and DNA target. Figure shows the melting behavior of two probes, labeled with rhodamine and HEX, respectively. LNA 715 (5' Rho-TAGTATG, Fig. A) is an AT-rich heptamer which nevertheless possesses sufficient binding affinity for its perfect-match target at 50 nM, combined with perfect discrimination of center T /G mismatch (Fig. A, curve 2). LNA 607H (5' Hex-ACGGAG, Fig. B) is a GC-rich Hex-labeled hexamer which also presents high melting temperature (Tm) and excellent mismatch discrimination (G /T mismatch, Fig. B, curve 2). During the course of our studies, we noticed that although HEX-labeled LNA probes exhibited ample FP dynamic range, their values were generally lower compared with those of rhodamine-LNAs. On the other hand, FP detection of hybridization of LNAs labeled with fluorescein was practically impossible due to the very small (10 --15 mP) changes in polarization (data not shown). Affinity and GC-content of LNA probes | In order to assess the possibility of genotyping a large number of SNPs with extremely short LNA probes, we undertook a Tm study of a series of rhodamine-labeled short (6 --7 nt) LNAs. The sequences of those probes were designed such that one could obtain a comprehensive picture of the affinities of such probes across the whole GC-content range. Table shows the results for 17 different probes. In addition to the Tm values for perfectly matched hybrids, the table lists the FP dynamic ranges obtained during the melting experiments as well as results on mismatch discrimination. The mismatch positions were at or close to the center of the probes, as we expected this to have the most destabilizing effect on the hybrids. A quick survey of the GC-rich sequences (for example, LNA 601, 602, 606, 607 and 701) indicates that they possess enough affinity to be employed for SNP genotyping as hexamers. LNA hexamers with only 2 GCs (LNA 609) had marginal affinity which, however, increased sufficiently once an extra base was added (LNA 713). On the other hand, LNAs with only a single G or C required a total length of an octamer. Further studies would be needed to establish more precise and predictive rules for Tms of LNA /DNA hybrids as a function of the LNA sequence. As seen in Table , the FP dynamic ranges were variable but generally adequate (55 --140 mP), with only few exceptions (e.g. LNA 603, 35 mP). SNP genotyping of PCR products | To demonstrate the possibility of single nucleotide genotyping of PCR products obtained from genomic DNA using rhodamine labeled hexamer LNA probes, we amplified a 107 bp sequence from genomic DNA samples using PCR primers FS and R. This product contains an SNP, a C to T transition, and the genomic DNA samples used in the experiment had been genotyped previously. PCR reactions were performed with four different templates: (i) no-template control (ntc); (ii) homozygous C human genomic DNA sample; (iii) heterozygous C /T human genomic DNA sample; and (iv) homozygous T human genomic DNA sample. One of the PCR primers used in the amplification, FS, contained five phosphorothioate bonds at its 5' end. These rendered the corresponding strand of the double-stranded PCR product resistant to treatment with the T7 gene 6 exonuclease, which was added to the PCR mixture after the amplification to hydrolyze the opposite strand and generate a single stranded product . We had previously established that the LNA probes are completely resistant to this exonuclease, both in the single- and double-stranded (LNA /DNA hybrid) form (data not shown). To generate LNA --DNA melting curves by FP, PCR reaction products were first digested for 10 min with T7 gene 6 exonuclease, and then mixed with 50 nM probe in fluorometer cuvette. Duplex melting was effected by slow temperature increase in a water bath. The two LNA probes exhibited excellent discrimination, displaying typical melting behavior only where there was perfect match between probe and genotyping sample . In addition to performing the genotyping by melting curve analysis, we tested the possibility of SNP detection by real-time FP measurement. Thus, the rhodamine-labeled probes 601 and 602 were added to separate aliquots of the PCR product, and the FP signals of the dye were recorded over 20 min. The result obtained from the analysis with probe LNA 601 is shown in Figure . We observed no changes in the FP signals upon addition of the double-stranded PCR products, in the absence of exonuclease. Then, following the addition of T7 gene 6 exonuclease to a final concentration of 2 U /microl, an increase of the FP was observed when the generated single-stranded PCR products were completely complementary to the LNA probe, but no changes were seen with the PCR product of the opposite genotype, containing a single nucleotide mismatch with the probe used. The heterozygous sample gave an intermediate increase of the FP signal, as expected. Figure 1 | Room temperature hybridizations of 5' rhodamine-labeled LNAs. Room temperature hybridizations of 5' rhodamine-labeled LNAs. LNA probes (50 nM in 50 mM HEPES, pH 7.5, 100 mM NaCl) were mixed thoroughly with the respective amounts of DNA targets and the steady-level FP was recorded in a fluorometer. Shown are the titration curves for (A) LNA 601, 5' Rho-GTCGCC; (B) LNA 602, 5' Rho-GTCACC; (C) LNA 603, 5' Rho-CATGCC; and (D) LNA 604, 5' Rho-TATGCC with their perfect-match (filled squares) and single-base mismatch (empty squares) DNA targets. Figure 2 | Melting curve detection by FP. Melting curve detection by FP. The experiment was performed as described in the Materials and Methods. Shown are (A) LNA 715, 5' Rho-TAGTATG; and (B) LNA 607H, 5' Hex-ACGGAG at 50 nM concentration in the presence of 50 nM perfect-match (curve 1) or mismatch (curve 2) targets. FP correction was performed by subtraction of the free-probe FP at each temperature. Figure 3 | SNP genotyping by FP melting curve measurement. SNP genotyping by FP melting curve measurement. The experiment was performed as described in Results. Shown are the melting curves when PCR reactions were analyzed with (A) LNA 601, 5' Rho-GTCGCC; and (B) LNA 602, 5' Rho-GTCACC. In both (A) and (B), curves denote the following: 1, homozygous perfect-match DNA; 2, heterozygous DNA; and 3, homozygous DNA of opposite genotype. Figure 4 | Real-time SNP genotyping by FP. Real-time SNP genotyping by FP. The experiment was performed as described in Results. Shown are the FP time-courses when PCR reactions were analyzed with LNA 601, 5' Rho-GTCGCC. Only the homozygous (curve 1) and heterozygous (curve 2) DNA samples show an FP increase upon exonuclease digestion. All other traces (curve cluster 3: no-template control, opposite genotype, no exonuclease) remain flat. DISCUSSION : LNAs are a new class of synthetic DNA /RNA analogs. These molecules display a very high affinity for complementary nucleic acid targets, both RNA and DNA. At the same time, their sensitivity for mismatch discrimination makes them uniquely suited for hybridization-based SNP genotyping. These two properties could allow the generation of an essentially complete set of genotyping probes, applicable to the typing of the majority of SNPs. The size of this universal library will be determined by the length of the probes: 1024 for a library of all possible pentamers; 4096 for all possible hexamers; but 16 384 for all possible heptamers. The former two appear reasonable, but the latter would be prohibitively expensive and hard to manage. Ideally, all the probes of such universal genotyping probe libraries will hybridize at or close to room temperature, and provide excellent mismatch discrimination. The relatively low affinities of DNA probes and of all other nucleic acid probes available previously, including PNAs, are not high enough to allow the generation of such libraries of short probes fulfilling these criteria. In our initial evaluation of the hybridization properties of LNA molecules we used probes containing eight to nine bases, but after observing their extraordinary high affinities for the target DNAs, we decided to explore the idea of a library of hexamer LNA probes for SNP typing. As the method for hybridization detection, we decided to explore an approach that we have previously successfully employed to detect the hybridization of PNA probes to DNA targets. This approach uses rhodamine-labeled probes and relies on measuring the FP changes of this dye upon probe hybridization. As shown in the present paper, this approach proved highly suitable for detecting the formation of LNA /DNA hybrid formation in a homogeneous solution. At the same time, the sensitivity of the LNA probes for single nucleotide mismatches allowed the discrimination of perfectly matched and mismatched targets with high efficiency. Application of LNAs in this assay was additionally facilitated by their nuclease stability and general ease of handling. Unlike PNAs, which occasionally tend to aggregate and have limited solubility, we encountered no such issues when working with LNA. From our results, it appears that in theory it may be possible to generate a useful, universal set of genotyping reagents based on LNA probes. Based on our results, all probes containing at least three G or C would be hexamers, and probes containing two or one G or C would be heptamers or octamers. For multiplexing experiments, two identical libraries will be needed, one with all probes labeled with rhodamine and the other with all probes labeled with HEX. Sequences that will be especially problematic to genotype will be the ones that encompass palindromes because self-complementary LNAs are expected to have greater affinity to themselves than to their perfect-match DNA targets. In reality though, at the relatively high current prices of LNA oligomers, the synthesis of such universal reagent set is prohibitively expensive. Backmatter: PMID- 12202780 TI - Multivariate curve resolution: a powerful tool for the analysis of conformational transitions in nucleic acids AB - A successful application is reported of the multivariate curve resolution alternating least-squares method (MCR-ALS) for the analysis of nucleic acid melting and salt-induced transitions. Under conditions where several structures co-exist in a conformational equilibrium, MCR-ALS analysis of the UV and circular dichroism (CD) spectra at different temperatures, ionic strength and oligonucleotide concentration allows for the resolution of concentration profiles and pure spectra of the different species. The methodology is illustrated by the case of the cyclic oligonucleotide d. The melting transition of this molecule at different oligonucleotide concentrations was studied at 0, 2 and 10 mM MgCl2 by UV and CD spectroscopy. In addition, salt titration experiments were carried out at 21.0 and 54.0C. The MCR-ALS analysis indicates that three different conformations of this molecule co-exist in solution. In agreement with previous NMR studies, these conformations were assigned to a monomeric dumbbell-like structure, a dimeric four-stranded conformation and a disordered (random coil) structure. The MCR-ALS methodology allows for a detailed analysis of how this equilibrium is affected by temperature, salt and oligonucleotide concentration. Keywords: INTRODUCTION : The remarkable conformational flexibility of nucleic acids allows for the formation of a great variety of structures besides the canonic B-DNA duplex. Among these structures, those involving more than two strands have received considerable attention in the last decade ( --). There are two main reasons for the intense interest in such multistranded structures. First, there is more than circumstantial evidence for their biological role in replication, transcription and recombination (,). Second, several multi-stranded motifs are used as the basis of some therapeutic approaches. In the antigene strategy, triplex-forming oligonucleotides exhibit sequence-specific recognition of duplex DNA and can be used to modulate gene expression . A comprehensive review has recently appeared discussing the design of agents targeted toward a structure-specific molecular recognition of DNA triplexes or quadruplexes . The G-rich sequence of human telomeric DNA can adopt a G-tetraplex structure in vitro that blocks the telomerase-catalyzed telomer elongation . Since this enzyme is expressed in tumor cells but not in most somatic cells, it is a prime target for cancer therapy. Several small molecules that stabilize G-tetraplexes have shown themselves to be strong telomerase inhibitors and are promising antitumor drugs (,). On the other hand, the disruption of tetraplex forms by small molecules is a potential route to therapeutic intervention against triplet repeat diseases such as the fragile X syndrome . Finally, G-tetraplex-forming oligonucleotide aptamers are potent thrombin or HIV-1 integrase inhibitors. Numerous three- and four-stranded structures have been solved either by X-ray crystallography or by NMR spectroscopy. However, when the aim is only to assess the formation of such structures and to evaluate their stability under certain conditions, low resolution alternatives such as UV absorption spectroscopy or circular dichroism (CD) are the techniques of choice (,). These techniques have the advantage of requiring much smaller amounts of oligonucleotides and allowing easier performance of experimental measures. The drawback is that the interpretation of the experimental results may be difficult if more than two structurally different species are involved in an equilibrium and their occurrence strongly depends on experimental conditions such as ionic strength, type of cation, pH and temperature. A thorough review of energetic and structural considerations for low- and high-order nucleic acid systems can be found in the literature . Certain cyclic oligonucleotides, as shown by X-ray diffraction and NMR spectroscopy, can self-associate to form a new dimeric four-stranded structure, known as the bi-loop motif, containing G:C:G:C and A:T:A:T tetrads ( --). The same tetrads, although differently arranged, have also very recently been described . In the solution NMR study of the cyclic octamer d at low oligonucleotide concentration (<1 mM in water), we perceived the predomination of a dumbbell-like structure. But at higher oligonucleotide concentration, two molecules dimerize to form the quadruplex (,). When the temperature is increased, the two species evolve to a third one, the random coil structure. Preliminary UV and CD studies of this cyclic octamer were undertaken to assess the occurrence of these species under different conditions (water or different salt concentrations, temperatures). The studies were also designed to evaluate, at the dilution conditions used (5 microM), the thermodynamic stability of the structured dumbbell and bi-loop species . The differences perceived between the UV and CD spectra recorded under different conditions made it difficult to ascertain whether the spectrum of a unique species was being observed or that of an equilibrium mixture of two or three species. By way of example, in the UV melting profile in salt medium, an initial decrease in chromicity was followed by a final increase in the UV absorption. The anomalous shape of the curve, with a partially inverted denaturation profile suggested that what was being recorded was the simultaneous melting of the two structured species into the random coil. The problem therefore, we assumed, was to describe how these different equilibria are reached, and also to achieve accurate speciation of the system with regard to temperature, salt and oligonucleotide concentration changes. Traditionally, conformational equilibria changes have been monitored by UV or CD spectroscopies using a single wavelength (univariate data analysis). This approach has clear drawbacks, the most important being the assessment of the number of different conformations present in an experiment when no selective wavelengths are available and there is a lack of appropiate reference spectra. These difficulties can be overcome by multiple wavelength spectroscopic monitoring of these conformational transitions and the application of appropriate multivariate data analysis methods. One such method is the recently proposed multivariate curve resolution method based on alternating least squares or MCR-ALS (, and references therein) and on factor analysis techniques . This paper describes the application of MCR-ALS to the study of the conformational transitions of d. The goal is the assessment and resolution of all the possible conformations, including intermediate states, and the estimation of their pure spectra and concentration profiles when temperature and salt concentration are changed. A certain challenge is involved in checking whether MCR-ALS results could describe the equilibria between the dimeric bi-loop and monomeric dumbbell structures in salt media and their denaturation into random coil structures. Hence, several experimental approaches using different initial conditions were used. First, melting experiments were run at different salt and oligomer concentrations to monitor their influence in conformational changes. Second, salt-induced conformation transitions were studied, carrying out salt titration experiments at two different temperatures. MATERIALS AND METHODS : Sample preparation | Cyclic octamer d was synthesized as reported and purified by standard liquid chromatography. Oligonucleotide sodium salt solutions were prepared in Ultrapure water (Millipore). Appropriate volumes of buffer [200 mM piperazine-N,N'-bis(2-ethanensulfonic acid), PIPES, disodium salt, pH 7.0] and stock salt solutions (200 mM MgCl2 and 2 M NaCl) were added in salt medium experiments. Oligonucleotide concentration (referred to the monomer) ranged from 5 to 70 microM, and was determined by absorbance measurements at 260 nm at room temperature considering a micromolar extinction coefficient of 64.5 OD /(micromolml) in water . Samples for melting studies were heated at 90C for 5 min and allowed to renaturalize, cooling slowly until room temperature. Oligonucleotide samples were kept at 4C until their use. CD and UV absorption experiments | CD and UV molecular absorption spectra (240 --330 nm) were recorded on a Jasco J-720 spectropolarimeter equipped with Neslab RET-110 temperature control unit. Instrumental control, data acquisition and analysis were carried out using personal computers. The work parameters during the registration of the spectra were bandwidth of 1 nm, sensitivity of 10 mdeg, scan speed of 20 nm /min, response of 8 s, step resolution of 1 nm and two averaged spectra. For all measurements, a Hellma quartz cell (path length of 1.0 cm and volume of 750 microl) was used. Melting experiments were carried out measuring UV absorption and CD spectra at 3C increments, using a temperature rate of 20C /h. Each sample was initially stabilized at the starting temperature for 15 min. Three representative melting experiments at 0, 2 and 10 mM MgCl2 concentration (in experiments 1, 2 and 3, respectively) and 100 mM NaCl were selected to study the effect of temperature on equilibria between oligonucleotide conformations. In order to study and remove possible background contributions, melting experiments were also performed on samples without oligonucleotide. Background contributions were especially relevant in melting UV spectra because PIPES buffer UV absorption changes with temperature. Salt titration experiments were carried out measuring UV absorption and CD spectra at each MgCl2 and NaCl concentration until no changes were observed in two consecutive spectra. These experiments were performed at 54C (experiment 4) and 21C (experiment 5) and MgCl2 concentrations ranging from 0 to 7.5 mM and 0 to10 mM, respectively. The first spectrum measured at each titration experiment was the one corresponding to oligonucleotide in buffer solution. The next recorded spectra corresponded to successive additions of stock salt solution (MgCl2 and NaCl). Salt experiments were carried out at 21 and 54C because preliminary results showed that at these two temperatures the system was rather complex and difficult to be resolved when only melting experiments were analyzed (see Results and Discussion). Nucleation equilibria between ordered conformations were studied in melting experiments at 5, 10, 20, 50 and 70 mM oligomer concentrations (in experiments 6, 7, 8, 9 and 10, respectively) in pure water and in salt media (10 mM MgCl2, 100 mM NaCl). Multivariate curve resolution | Experimental CD and UV absorption data were analyzed using the MCR-ALS procedure previously described (,,,). This procedure is used to estimate the concentration profiles and pure spectra for each spectroscopically active conformation present in melting and salt titration experiments. MCR-ALS procedure is based on factor analysis methods and can be used to analyze spectroscopic data obtained in biochemical or biophysical process monitoring. All MCR-ALS calculations were performed using in-house MATLAB (version 6; The Mathworks Inc, Natick, MA) routines (codes are freely available at ). Only a short description of the MCR-ALS procedure is given here. Spectra recorded in one experiment are collected in a table or matrix D (Nr x Nw), Nr rows being the Nr spectra recorded at successive temperature or salt concentration values, and Nw columns being the number of wavelengths measured in every spectrum. Mathematically, the goal of MCR-ALS is the recovery of the concentration profiles matrix C and of the pure spectra matrix ST (the superscript T means the transpose of matrix S). By considering the multiwavelength extension of Lambert-Beer's law (in matrix form), D = C ST + E1 E is the matrix of residual CD signal or absorbance not explained by the model. E should be close to experimental error. The first step of the data analysis procedure is the estimation of the number of spectroscopically distinct conformations present in a particular experiment. This number of conformations (Ns) is estimated by rank analysis or singular value decomposition (SVD) of each data matrix D . In this estimation, it is assumed that singular values associated with experimental noise are significantly lower than those associated with systematic chemical data variance. Therefore, it is assumed that they can easily be distinguished from the plot of their magnitudes. When doubts about the number of conformations occur, MCR-ALS analysis is repeated for the different possibilities and the results finally evaluated in terms of data fitting and of quality of MCR-ALS-resolved profiles. Equation 1 is solved iteratively by an ALS algorithm that calculates concentration C and pure spectra ST matrices optimally fitting experimental data matrix D using the proposed number of Ns conformations. This iterative optimization requires initial estimations either of C or ST, which can be obtained from evolving factor analysis (EFA) or from pure variable detection methods . During the ALS optimization, several constraints were applied including non-negativity for concentration profiles C and for UV absorbance spectra profiles ST (not applied in case of CD spectra profiles), unimodality for concentration profiles C, and closure also for concentration profiles C. See previous works for a more detailed explanation of the ALS iterative optimization procedure (,,,). Convergence is usually achieved in few iterations and the final value of lack of fit is defined as where dij are the absorbance data of spectrum i and wavelength j, and dij* are the MCR-ALS-recalculated data using the specified number of conformations Ns. Concentration profiles C and pure spectra ST resolved for each conformation in the analysis of individual data matrices may differ from the true ones because of possible unresolved underlying factor analysis ambiguities (,). This means that concentration profiles and pure spectra may be only a solution within a band of feasible solutions that are bounded by the constraints applied in the calculation. Some of these ambiguity problems can be more easily solved by means of the simultaneous MCR-ALS analysis of several data matrices obtained under different experimental conditions . When a chemical system is monitored using two spectroscopies (e.g. CD and UV absorption), a row-wise augmented data matrix can be built up from the individual data matrices corresponding to each spectroscopy, Dabs and DCD, respectively. The dimensions of the new row-wise augmented matrix will be (Nr) x (Nwabs + NwCD), and the new linear model used for MCR-ALS analysis is [DCDDabs] = [C][SCDSabs]T + [ECDEabs]3 Solving equation 3 for C and [SCDSabs] helps to resolve the species of the system, especially if their spectra are not well defined in one of the two spectroscopies simultaneously analyzed. For instance, in the study of the evolution of conformational equilibria, CD spectra are richer and with more features than UV absorption spectra. Therefore, the simultaneous analysis of both would allow a better resolution of UV absorption spectra. In some other cases, one of the finally resolved components corresponds in fact to a mixture of two or more conformations due to rotational ambiguity problems or to rank deficiency problems . In all these cases, and also to allow an improved resolution of the system, the simultaneous analysis of different independent experiments is usually very useful and powerful. For instance the model used for MCR-ALS simultaneous analysis of two different melting experiments at different salt concentrations monitored both by CD spectroscopy, giving experimental data matrices D1CD and D2CD, respectively, is described by the equation In this case, better conditions for species resolution will be achieved when the two experiments are analyzed together compared with when each experiment is analyzed individually (see Results). We also used this kind of analysis in the present work for modeling UV absorption changes of background (PIPES salt) at low and high temperatures and at low and high salt concentrations. Finally, a still more involved data analysis approach can be proposed for data acquired where different spectroscopies are applied simultaneously to multiple experiments. The linear model for n experiments studied by CD and UV absorption spectroscopies is described by equation This kind of simultaneous analysis is the more powerful one and allows improvement of the resolution of very complex data structures, combining the profits of both approaches previously described. MCR-ALS analysis of row- and /or column-wise augmented data matrices has been shown to give more reliable solutions, eventually removing totally rotational ambiguities and rank deficiency problems. See previous references for a more thorough discussion of these topics (,). Thermodynamic data analysis | Calculation of thermodynamic data for the equilibria between conformations is traditionally carried out by conversion of experimental absorbances at single wavelength versus temperature profiles into molar oligomer fraction versus temperature curves. However, this conversion is highly dependent on base line drifts and on the assumption that absorbance and CD signal changes at any temperature are directly proportional to the reaction extent at that particular temperature (,). Here, thermodynamic parameters were directly estimated instead from MCR-ALS-resolved concentration profiles. This more powerful multivariate approach eliminates base line problems and takes advantage of the improved resolution of concentration profiles. It allows a better estimation of melting temperatures and of the thermodynamic parameters because of signal averaging improvement and fewer error propagation problems. First, Tm values were estimated from the intersection of MCR-ALS-estimated concentration profiles of the different conformations resolved in melting experiments. Mean error deviation of these Tm values was estimated to be approximately +-0.7C from two replicated independent melting experiments of each system. This mean error deviation value agrees with estimations in previous works . Equilibrium constants at the different temperatures were also directly evaluated from the ratio between concentration values of the two conformations resolved at the same temperature by MCR-ALS. Then, free energy DeltaG values for the corresponding melting equilibrium between the two different conformations were roughly estimated from temperature dependence of equilibrium constants. In Table more details about thermodynamic evaluations are given. Thermodynamic values are given here to corroborate the reliability of the resolved concentration profiles from the proposed MCR-ALS method. Additional work is pursued to validate the proposed method as a general method for the evaluation of thermodynamic parameters Table 1 | Melting and dimerization thermodynamic data RESULTS AND DISCUSSION : Melting experiments in water medium | Thermal denaturation of d was studied first in pure water, i.e., 0 M MgCl2 and 0 M NaCl (experiment 1). CD and UV absorbance spectra recorded in this experiment (Fig. A and B) were arranged in two data matrices [Dm,0CD] and [Dm,0UV], respectively. The superscripts refer to melting experiments and to no salt added (salt concentration is equal to zero) and the subscripts refer to the spectroscopic method used. The dimensions of both matrices were 24 rows (number of different temperatures /number of spectra) and 92 columns (number of wavelengths). CD bands at 21C showed two maxima at 243.0 and 286.5 nm and a minimum at 263.0 nm. When temperature was raised, a gradual spectral variation was observed. The CD spectrum for the denatured conformation at the highest temperature showed the lowest intensity signal and only one maximum at 278 nm. Changes in UV absorption spectra were not so dramatic as those for CD data. Variations in CD and UV absorption could also be observed by recording data ellipticity at 286 nm and UV absorption at 264 nm (univariate analysis, as shown in Fig. C). Melting profiles obtained in this way agreed with the presence of a single transition in the studied temperature range (Tm value similar50C). MCR-ALS was applied to check the validity of the conclusions derived from this single wavelength data analysis. Rank analysis by SVD of the whole CD spectra (data matrix [DCDm,0]) indicated the presence of two components (i.e., Ns = 2). Accordingly, an initial estimation of the concentration profiles was obtained by EFA and later optimized by ALS using the constraints of non-negativity, unimodality and closure for concentration profiles (,,,). A Tm value of 48.2C was obtained from MCR-ALS-resolved concentration profiles, which was in agreement with the value obtained from single wavelength data analysis. The ALS lack of fit calculated for the model with two components was 3.0%, which was considered reasonably good, taking into account the higher noise levels of CD data. MCR-ALS was then applied to analyze UV molecular absorption data [Dm,0UV]. In this case, the resolved profiles were slightly worse than those described above in the analysis of CD data. Resolved concentration profiles did not show a smooth shape like those for CD data, which made it more difficult to estimate an accurate value for Tm. The reason for this difficulty was related to the small UV absorbance variation observed during the melting experiment, the variation being hidden by the experimental noise contribution. When CD and UV absorption data were simultaneously analyzed according to equation 5, the resolution of UV absorption data improved significantly. Figure D --F shows the resolved concentration profiles [Cm,0] and pure CD [STCD] and UV absorption spectra [STUV]. ALS lack of fit for the simultaneous analysis was 2.0%. According to previous results based on NMR spectroscopy (,) carried out in the absence of salt, the two resolved components in Figure D --F can be assigned to the dumbbell and the random coil conformations, respectively (Scheme ). From the resolved concentration profiles, a Tm value of 50.4C was estimated. The two resolved pure CD and UV absorption spectra showed spectral features explaining experimental spectra at low and high temperatures. Melting experiments in salt media | In order to study the influence of salt media on the denaturation equilibria, two melting experiments (experiments 2 and 3) were carried out at 10 and 2 mM MgCl2, respectively, and 100 mM NaCl. CD and UV absorption spectra obtained at 10 mM MgCl2 (experiment 2) and arranged in data matrices [Dm,10CD] and [Dm,10UV], are given in Figure A and B. There were some spectral features that could not be explained satisfactorily using a single wavelength approach. Whereas the CD signal at 289 nm showed a continuous smooth decrease along the whole temperature range studied, UV absorbance at 263 nm decreased its value from 20 to similar60C and, finally, increased slightly from 60 to 90C (Fig. C). Whereas information obtained from CD data seemed to indicate the presence of only two conformations, UV absorption data showed a behavior that cannot be explained by assuming the presence of only two conformations. As in experiment 1, CD and UV absorption data from experiment 2 were analyzed using a row-wise data matrix arrangement, i.e., according to equation 5. Rank analysis of the row-wise matrix showed the presence of two components. Figure D --F shows the results obtained by MCR-ALS analysis. Lack of fit was 1.8%. The Tm value of 55.4C agreed well with the value determined from univariate analysis of CD data. Resolved CD, [STCD] and UV, [STUV], spectra for the component at the highest temperature, 90C, were very similar to the spectra previously obtained in pure water at the same temperature. Therefore, they were assigned to the same random coil conformation. The resolved pure CD spectrum for the component at 20C showed band maxima shifted to 240.5 nm and 289.0 nm and a band minimum shifted to 259.5 nm. These values were not the same as those observed for the major conformation present at 20C in water (experiment 1), which would indicate a different initial conformation at 10 mM MgCl2. According to previous studies, this initial conformation can be related to the bi-loop dimeric structure (Scheme ; ) formed by antiparallel association of two monomer molecules. However, MCR-ALS results for experiment 2 (shown in Fig. D --F) did not explain satisfactorily the univariate melting profile previously observed. Since the absorptivity of the resolved UV absorption spectra for the initial ordered conformation is higher than the absorptivity of the resolved UV spectrum of the random coil conformation, the absorbance at any wavelength should always have decreased smoothly along the whole temperature range. However, the experimental spectra showed a clear increase of absorbance values at temperatures >60C. This fact suggested that MCR-ALS results obtained from the analysis of experiment 2 according to equation 5, although mathematically acceptable (as reflected by the low lack of fit value obtained), could not entirely explain the experimental results. MCR-ALS results for experiment 3 at 2 mM MgCl2 concentration (data not shown) were similar to those obtained for experiment 2 at 10 mM MgCl2 concentration. Rank analysis also showed the presence of only two components. MCR-ALS resolved concentration profiles were similar to those shown in Figure D, i.e., a major component at 20C and another one at 90C. As for experiment 2, resolved spectra for the major component at 90C showed the same spectral features as those observed in water at higher temperatures. Therefore, this component was also related to the random coil conformation. Band positions of CD spectrum for the major component present at 20C were different to those found in pure water, but also different to those found at 10 mM MgCl2. In fact, MCR-ALS-resolved CD spectra for the initial state at 2 mM MgCl2 showed two maxima at 242.0 and 288.0 nm, respectively, and a minimum at 261 nm. Therefore, these maxima and minimum values were located between those obtained in experiments 1 and 2. The CD band positions suggested that the initial state of the oligonucleotide at 2 mM MgCl2 could actually be a mixture of the conformations observed in pure water (dumbbell) and 10 mM MgCl2 (bi-loop). In other words, the MCR-ALS-resolved concentration profile for the first component would correspond to the sum of concentration profiles of, at least, two conformations. This problem is known in chemometrics as rank deficiency and several ways to circumvent it have been proposed . In the present paper, rank deficiency could be solved by the simultaneous analysis of several experiments at different initial conditions (see below). Salt titration experiments at a fixed temperature were carried out and the experimental data analyzed by MCR-ALS. The aim of these experiments was to obtain a better understanding and resolution of the systems under study, taking into consideration previous proposals about the formation of dimeric bi-loop structures that predominate at high concentration salt media (Scheme ) as well as the monomeric dumbbell structures that dominate in pure water media. Salt titration experiments | The conformation found at low temperatures in pure water was also expected to be the main conformation at low temperatures in low salt concentrations. At high MgCl2 salt concentrations and low temperatures, the main conformation was also expected to be similar to that present at 10 mM MgCl2. MCR-ALS analysis of salt titration at 21C (experiment 4, experimental data not shown) indicated the presence of two components. Concentration profiles showed a crossing point at 2 mM MgCl2 concentration, approximately. This value would suggest the presence, at the same initial conditions of melting experiment 3, of a mixture of ordered conformations. MCR-ALS analysis of salt titration at 54.0C (experiment 5, experimental data not shown) gave poorer fitting results. Lack of fit value was now higher, near 7.0%, when analyzed according to equations 1 or 3, and it was possible to resolve (poorly) only two conformations. From previous results, it was apparent that at 54C, and depending on salt concentration, the mixtures of three conformations can be present. A possible way to improve the resolution of the different conformations in this and other experiments is the simultaneous analysis of more than one experiment according to equations 4 and 5. MCR-ALS simultaneous analysis of melting and salt titration experiments | Results of the simultaneous analysis of melting and salt titration experiments monitored using only one of the two spectroscopies (either CD or UV absorption, i.e., the model described in equation 4) are not given here for the sake of brevity. In any case, the results for the simultaneous analysis of CD and UV absorption data sets according to equation 5 were better. The analyzed data matrix consisted of the arrangement of 10 single data matrices corresponding to melting experiments at 0, 10 and 2 mM MgCl2 (experiments 1, 2 and 3) and salt titrations at 54 and 21C (experiments 4 and 5). All of these were monitored by CD and UV molecular absorption. Additionally, background data matrices obtained at the same experimental conditions as the melting experiments but without changes in oligonucleotide concentration (UV absorption changes due to PIPES buffer) were also included in the simultaneous analysis. These background contributions were modeled appropriately with the inclusion of three additional components in the MCR-ALS model: one component is used to take into account temperature background changes in water; the other two components are used to take into account temperature background changes in salt media. Using this approach, lack of fit values and resolution of conformation profiles improved appreciably. Lack of fit for the global analysis was 6.1%, which was considered rather good taking into account the large number of independent experiments simultaneously analyzed and their variety. The results finally achieved are given in Figures and . These results confirmed the presence of three independent oligonucleotide conformations, in agreement with what was proposed in Scheme . SVD analysis of augmented data matrices described by equations 4 and 5 already indicated a rank increase of the system due to the detection of these three conformations. This confirms that using the proposed approach of simultaneous analysis of appropriately designed independent experiments, the elimination of rank deficiency and rotational ambiguity problems is possible, in full agreement with previous results (,,,). In experiment 1 (Fig. A), in pure water, only two conformations were present and assigned respectively to the monomeric dumbbell conformation (Tm = 48.1C) and to the random coil conformations. This result agrees with the one previously obtained in the individual analysis of CD and UV absorption data from experiment 1 . There, the system did not show any factor analysis ambiguity and gave equal results regardless of the model used. In experiment 2, however, three conformations were now detected and resolved (Fig. B). At the lowest temperatures only the conformation previously related to the dimeric bi-loop structure was detected (Tm = 54.8C). But when the temperature was increased, and in addition to the formation of the random coil, a new minor conformation was now resolved with a spectrum matching that previously obtained in pure water at low temperatures (assigned to monomer dumbbell conformation). The presence of this minor conformation would now clearly explain the anomalous behavior observed in absorption melting curves at a single wavelength. Hence, the decrease of absorbance at temperatures <60C would be related to the higher UV molecular absorptivity of the bi-loop in relation to the dumbbell and random coil conformations (Fig. B). At higher temperatures the dumbbell conformation will also disappear to yield the random coil conformation and the total absorbance increases slightly accordingly until the end of the melting process. This result was now achieved because the proposed simultaneous analysis allowed a better resolution of the minor dumbbell conformation (in experiment 2), not detected previously in the individual analysis of this experiment. Notoriously (and also differently to what was obtained when analyzed individually), three conformations were now perfectly resolved for experiment 3 (Fig. C) at 2 mM MgCl2. Dimeric bi-loop and monomeric dumbbell conformations were simultaneously present in this experiment from the beginning of the titration at the lowest temperatures, both decreasing their concentration when temperature was raised because of thermal denaturation. The corresponding concentration profiles of these two conformations could not be distinguished in the individual analysis of experiment 3 due to the so-called rank deficiency problems . Figure D shows the evolution of the three conformations in equilibria, dumbbell, bi-loop and random coil at 54C when salt concentration changes. The resolution of the concentration profiles of these three conformations was not possible when experiment 4 was analyzed individually because the experimental spectra were always a complex mixture of conformations. The resolved concentration profiles in Figure D showed that at 0 M salt and 54C, a mixture of random coil (major conformation) and dumbbell (minor conformation) was present. This agreed with the values of their resolved concentration profiles at the same temperature of 54C in melting experiment 1 (Fig. A). However, salt titration at 54C could not be carried out up to 10 mM MgCl2 because of experimental difficulties and precipitation problems. The trend observed up to 8 mM showed a decrease of the random coil and dumbbell conformations and an increase of bi-loop conformation, which would also be in agreement with the values of concentration profiles in melting experiment 2 at the same temperature of 54C (Fig. B). Additionally, the values of the concentration profiles at 2 mM MgCl2 in Figure D would agree with the values of concentration profiles in experiment 3 at 54C (Fig. C) where random coil is the major conformation. At 21C (Fig. E), no random coil conformation was detected at appreciable concentrations, and only the dumbbell to bi-loop transition was observed. The resolved concentration profile confirmed that at 2 mM MgCl2 and 21C, a mixture of bi-loop (major conformation) and dumbbell (minor conformation) was present at the beginning of experiment 3 (Fig. C). Figure E also confirmed that at 10 mM MgCl2 and 21C the oligonucleotide was only present in its bi-loop form, which was also observed at the beginning of experiment 2 in Figure B. Finally, in Figure , pure CD and UV absorption spectra for the MCR-ALS resolved conformations are shown. Spectral features agreed with those observed experimentally. Hence, CD maxima for bi-loop conformations are located at 286 and 263 nm, for dumbbell at 289 and 260 nm and for random coil at 279 nm. The different intensities of the dumbbell and bi-loop conformation spectra explain the experimental spectra in all the experiments, including melting experiments 2 and 3 at 10 and 2 mM MgCl2 and salt titration experiment 4 at 54C, which could not be satisfactorily explained by single wavelength analysis or by MCR-ALS multiwavelength analysis of their respective individual experiments. Thermodynamic study of melting and of dumbbell to bi-loop dimerization equilibria | As stated in the Materials and Methods section, in order to complete the study of the equilibria between conformations, melting experiments at different oligonucleotide concentrations were performed in pure water and salt media. MCR-ALS results of melting experiments in water showed that Tm values were independent of oligomer concentration . This behavior was consistent with a monomolecular process, such as the intra-molecular unfolding of the dumbbell conformation. The calculated mean value for this equilibrium DeltaG at 25C, --5 +- 1 kJ /mol matches well with the values for previously reported univariate NMR and CD measurements . On the other hand, MCR-ALS results of melting experiments in salt media at several oligomer concentrations showed that Tm values depended on oligomer concentration. This behavior is in agreement with an equilibrium between conformations of different nuclearity, as in dimerization equilibria. The mean value for DeltaG at 25C, --65 +- 10 kJ /mol obtained from that value are in good agreement with results previously found . Figure 1 | Melting of d oligomer in pure water (experiment 1). Melting of d oligomer in pure water (experiment 1). (A) CD data [Dm,0CD], (B) UV molecular absorption once the blank contribution is removed [Dm,0UV], (C) absorbance versus temperature at 264 nm, (D) concentration profiles [Cm,0], (E) pure CD spectra [STCD] and (F) pure UV molecular absorption spectra [STUV]. Blue continuous line, monomeric dumbbell conformation; red dotted line, random coil conformation. Scheme 1. Conformations of d and the equilibria between them when temperature, salt and oligonucleotide concentration are changed [according to Lane et al. Scheme 1. Conformations of d and the equilibria between them when temperature, salt and oligonucleotide concentration are changed [according to Lane et al. and Salisbury et al. ]. Figure 2 | Melting of d oligomer in salt medium (10 mM PIPES pH 7, 10 mM MgCl2 and 100 mM NaCl; experiment 2). Melting of d oligomer in salt medium (10 mM PIPES pH 7, 10 mM MgCl2 and 100 mM NaCl; experiment 2). (A) CD data [Dm,10CD], (B) UV molecular absorption once the blank contribution is removed [Dm,10UV], (C) absorbance versus temperature at 263 nm, (D) concentration profiles [Cm,10], (E) pure CD spectra [STCD] and (F) pure UV molecular absorption spectra [STUV]. Green dashed line, dimeric bi-loop conformation; red dotted line, random coil conformation. Figure 3 | MCR-ALS-resolved concentration profiles for the simultaneous data analysis of melting and salt titration experiments. MCR-ALS-resolved concentration profiles for the simultaneous data analysis of melting and salt titration experiments. (A) Melting at pure water [Cm,0], (B) melting at 10 mM PIPES pH 7, 10 mM MgCl2 and 100 mM NaCl [Cm,10], (C) melting at 10 mM PIPES pH 7, 2 mM MgCl2 and 100 mM NaCl [STUV], (D) salt titration at 54C [Ct,54] and (E) salt titration at 21C [Ct,21]. Blue continuous line, monomeric dumbbell conformation; green dashed line, dimeric bi-loop conformation; red dotted line, random coil conformation. Circles represent experimental points. Figure 4 | MCR-ALS-resolved pure spectra for the three detected conformations in equilibria in melting and salt titration experiments. MCR-ALS-resolved pure spectra for the three detected conformations in equilibria in melting and salt titration experiments. (A) CD pure spectra, (B) UV molecular absorption spectra. Lines and symbols as in Figure . CONCLUSIONS : MCR-ALS methodology was used to study the equilibrium between the different conformations of d. The MCR-ALS procedure is seen to be a powerful tool for the analysis of multiwavelength spectroscopic data in cases where multiple species in equilibrium are present. The application of this method allowed the identification and resolution of three different conformations of d: dimeric 'bi-loop', dumbbell and random coil structures. The pure CD and UV spectra of each of these species were unambiguously resolved, and thermodynamic data describing the different equilibria were obtained. This methodology can be easily applied to other cases where an equilibrium between multiple nucleic acid species prevents a direct analysis of the spectroscopic data. Backmatter: PMID- 12223118 TI - IkB kinase alpha: a link in the chain of the mammary cycle AB - Abstract | The transcription factor NF-kappaB exhibits altered activity in some breast cancers but the relevance of this association has not been established. Cao et al.'s elegant study recently published in Cell reveals a NF-kappaB-dependent signalling pathway responsible for epithelial proliferation in the mouse mammary gland. Could this mechanism, rather than prevention of apoptosis, be responsible for the reported association between NF-kappaB and breast cancer? Could the specificity of NF-kappaB modulators of the IkB kinase complex determine the fate of epithelial cells at different stages of mammary development? Keywords: apoptosis, IkB kinase, mammary, NF-kappaB, proliferation, apoptosis, IkB kinase, mammary, NF-kappaB, proliferation, Contribution and relevance to the field : In vitro studies on NF-kappaB signalling have contributed much to our understanding of the multiple effects of NF-kappaB. However, beyond the largely descriptive studies of NF-kappaB-related phenotypes in animal models, few studies have addressed precisely how NF-kappaB mediates its effects in vivo. The study by Cao and coworkers provides insight into a NF-kappaB-dependent signalling pathway with a specific phenotypic outcome in vivo, placing NF-kappaB as a key regulator of epithelial growth in the developing mammary gland. In the light of previous reports demonstrating a role for NF-kappaB in survival of mammary epithelial cells , the underlying implication of the present study is that the function of NF-kappaB in mammary epithelial cells is context dependent and relies on a level of signalling specificity previously not appreciated. The study involves the use of an inactivatable mutant of IKKalpha, a component of the heterotetrameric IkB kinase (IKK) complex responsible for activation of NF-kappaB by a variety of stimuli, which is thought to convey development-specific signals to NF-kappaB . Interestingly, this mutant line (unlike IKKalpha null mice) was normal in nearly all respects (including NF-kappaB responses to proinflammatory stimuli) albeit for an abnormal mammary development, similar to that of the receptor activator of NF-kappaB (RANK)-deficient mice. The morphology of the IKKalphaaa/aa mammary gland was consistent with the expected timing of NF-kappaB activation in normal mammary glands of gestating mice . This represents an important tissue specificity in IKKalpha-dependent signalling, which may have implications in other NF-kappaB settings and in the treatment of breast cancers. With an upstream IKKalpha pathway identified, Cao et al. turned to probable downstream NF-kappaB targets that could be mediating the IKKalpha-dependent phenotype. A previous report of a similar mammary phenotype for cyclin D1 knockout mice prompted them to look at cyclin D1 expression in their IKKalpha mutant mice. Confirmation of a marked reduction in cyclin D1 expression in these mice led to the 'clinching' experiment: overexpressing a cyclin D1 transgene in IKKalpha mutant mice rescued the IKK phenotype in the mammary gland. This observation confirmed the hypothesis that NF-kappaB regulated the cell cycle during gestation, and firmly established a linear pathway from RANK, via NF-kappaB, to cyclin D1. The observation that NF-kappaB is responsible for proliferation of mammary epithelial cells during pregnancy is in itself not new. Brantley and colleagues had previously demonstrated that NF-kappaB positively regulated mammary epithelial proliferation, ductal branching and maintenance of overall mammary structure . However, Cao et al.'s study is significant not only in its identification of the specific signalling pathway responsible for this NF-kappaB action in the mammary gland, but also in the manner in which it is achieved. The elegant use of crosses between the IKKalpha mutant and other transgenic lines, previously demonstrated to have mammary phenotypes, combined with tissue-restricted expression of these transgenes, thus demonstrates the increasing power of transgenics in the elucidation of complex signalling pathways. Future perspectives : Two intriguing questions arise from the elucidation of this pathway in the mammary gland. How is the specificity for IKKalpha in the mammary gland achieved, and does the function of NF-kappaB change during the course of mammary development? Central to these questions is the identity of the NF-kappaB isoforms regulated throughout the mammary cycle. Could changes in signalling specificity upstream of NF-kappaB result in a shift in the ratio of activated NF-kappaB isoforms with subsequent changes in target gene expression and cellular phenotype? This has not been properly addressed for the complete mammary cycle and warrants further investigation. p52 processing was shown to be affected in IKKalpha mutant mice in the study of Cao et al. The relatively low level of the p52 isoform compared with the p65 and p50 isoforms, and the fact that there is no mammary phenotype in p52-deficient mice , however, suggest that this has little or no role prior to lactation. The relative contributions of IKKalpha and IKKbeta to the signalling complex vary according to the requirements of the cell. IKKbeta plays a predominant role in immune responses, while IKKalpha alone appears to be sufficient for at least some developmental systems . Cao et al. suggest that the relative contribution of IKKalpha and IKKbeta to downstream NF-kappaB activation may occur via alternate upstream kinases, one of which is NIK and the other yet to be defined. This conclusion is supported in an earlier study of human fibroblasts . Perhaps disproportionate activation of components of the IKK complex determines NF-kappaB responses to specific changes in the local environment during mammary development. NF-kappaB DNA binding activity is biphasic during mammary development, being downregulated after pregnancy and during lactation, and reactivated at the onset of involution . RANK expression is negligible in involution (Clarkson, unpublished observations, 2002) and is therefore unlikely to be the predominant NF-kappaB stimulus at this time. Tumour necrosis factor (TNF) and Fas receptor expression is elevated during involution, and Fas-deficient mice exhibit a mammary involution phenotype . Given that IKKalphaaa/aa cells show normal IKK and NF-kappaB activation in response to TNFalpha, it is intriguing to postulate whether TNFalpha or Fas, via IKKbeta rather than IKKalpha, may be responsible for the activation of NF-kappaB in involution. Could NF-kappaB thus have distinct roles in pregnancy (proliferation) and involution (survival)? There is increasing evidence in the mammary gland and other tissues that this could indeed be the case. Observation of NF-kappaB in a regressing mammary gland in vivo and intervention of NF-kappaB in vitro suggested a role for NF-kappaB in mammary epithelial apoptosis . This fits with the paradigm of NF-kappaB action in other cell types, particularly given the concomitant expression of TNF-receptor family members in mammary epithelial cells during involution. Since the first conclusive description of NF-kappaB acting as a survival factor downstream of TNFalpha, literally hundreds of studies have reported the role of NF-kappaB in apoptotic programmes in a variety of contexts. This may have had the effect of skewing perception of the role of NF-kappaB in developing tissues, including the mammary gland. Indeed, our study of NF-kappaB activity in a normal mammary gland concentrated on its role in involution . Studies describing altered NF-kappaB activities in breast cancers refer to apoptosis as a likely causal mechanism. Cao et al.'s timely contribution to the field has readdressed this balance, highlighting the importance of NF-kappaB in cell growth during development and its possible relevance to breast cancers. While evocative, Cao et al.'s suggestion that their observations might explain the association between NF-kappaB and breast cancer belies an inconsistency in the available evidence. There have been several reports describing constitutive activity of NF-kappaB isoforms in human breast cancer cell lines and in primary tumours . It is unclear from these studies, however, whether NF-kappaB activation or suppression contributes to carcinogenesis, which isoforms are aberrantly regulated or, indeed, whether NF-kappaB has a causal role rather than simply being a marker of tumour differentiation. While the study of Brantley and coworkers seemed to implicate NF-kappaB directly in the phenotypic transformation of mammary epithelial cells in virgin mice, the mechanism underlying this activity remained undefined. Direct regulation of cyclin D1 is, however, a promising proposition, as it is overexpressed in 50% of human primary breast cancers . If NF-kappaB is found to be a major contributory factor to human breast cancers then the approach described by Cao et al., to specifically target IKKalpha signalling and to thereby circumvent proinflammatory pathways, would be an attractive proposition. An alternative role for NF-kappaB in the mammary gland has been proposed by Soler and colleagues . Their studies have implicated NF-kappaB in the maintenance of the epithelial barrier during remodelling of the mammary gland. This is relevant to all stages of mammary development. These reports, taken together, support the view that NF-kappaB has different roles to play at different times during mammary development. At this stage, however, we cannot exclude the possibility that NF-kappaB is exclusively linked to cell-cycle regulation during involution. Conclusion : What do Cao et al.'s results mean for the future of NF-kappaB research? An in vivo study of NF-kappaB in mammary involution is clearly required and will no doubt be forthcoming in the near future. Use of conditional transgenics (IKKalpha aa/aa and IKKbetaaa/aa) mutants, activated at or around the onset of involution, would help to establish the relative contribution of each of these signalling components in mammary regression. Identification of the NF-kappaB isoforms throughout the mammary cycle may also help to determine the phenotypic outcome and alternative upstream pathways responsible for apoptosis and/or proliferation, and may shed light on previous observations in breast cancer cells. The present study has raised the possibility that two distinct NF-kappaB signalling pathways exist during mammary gland development. As one of these pathways has been identified, it now remains to be seen whether a similar approach reveals a second pathway later in the developmental cycle. Abbreviations : IKK = IkB kinase; NF-kappaB = nuclear factor-kappa B; RANK = receptor activator of NF-kappaB; TNF = tumour necrosis factor. Backmatter: PMID- 12223119 TI - A new model for ductal carcinoma in situ suggests strategies for treatment AB - Abstract | Human ductal carcinoma in situ (DCIS) of the breast is now diagnosed quite frequently, due largely to the introduction of mammographic screening. It has been shown in a cell culture system that activation of c-erbB-2, but not the epidermal growth factor receptor, results in a DCIS-like phenotype. Since overexpression of c-erbB-2 occurs in 60% of DCIS, this suggests that it could be a target for treatment in this disease. Keywords: ductal carcinoma in situ, growth factor receptor, tyrosine kinase inhibitor, ductal carcinoma in situ, growth factor receptor, tyrosine kinase inhibitor, Introduction : Breast pathologists have for many years been seeking to demonstrate a sequence of morphologically distinguishable cellular changes that define a linear sequence of transformation of normal breast ductal epithelium to the fully malignant state. It is still not entirely clear that this has been achieved, but a general model consists of premalignant changes (a variety of hyperplasias) followed by DCIS with subsequent invasive disease . There are alternative views that these changes are neither linear nor do they always entail each intermediate phenotype . Clinically, however, there are clear distinctions between these stages in terms of the risk to the patient, which affects the choice of treatment. It has been hoped that the description of the underlying molecular changes accompanying these morphologically defined stages, and at least a partial understanding of their significance, would enable pathologists to confirm or refute their models and to allow individual cases to be placed in the sequence. This has to some extent been achieved; many molecular changes at the level of the DNA have been found in DCIS but not in premalignant disease. Somewhat disappointingly, however, little if anything is known definitively about any additional molecular changes that are presumed to result in the transition from DCIS to invasive disease. Research continues into this problem using methods that minimise preconceptions such as genomic analysis and gene expression analysis. The molecular changes seen in DCIS include mutations in the p53 gene , and gene amplification and overexpression at the protein level of the c-erbB-2 receptor tyrosine kinase . There is a higher prevalence of both these events in more undifferentiated DCIS (known as comedo cancer in previous terminologies) , suggesting a functional significance, but this had not been definitively established. Growth factor receptor activation leads to a DCIS-like phenotype : The hypothesis that growth factor receptor activation leads to a DCIS-like phenotype is tested in an article by Muthuswamy et al. . Two groups cooperated in that work, each bringing a different technological contribution. Mina Bissell, the doyen of breast cancer cell biologists, developed three-dimensional cell culture models of normal breast epithelial cells over several years . Joan Brugge explored the mechanisms of growth factor receptor signalling and developed a system in which ectopically expressed receptors can be selectively activated at will by the addition of drugs . The c-erbB-2 protein, which is overexpressed in DCIS, is a member of the epidermal growth factor receptor (EGFR) family. This family consists of the EGFR, c-erbB-2 and two more receptors (c-erbB-3 and c-erbB-4) . Each protein has an extracellular domain, which in some cases recognises an activating ligand, a transmembrane domain and a cytoplasmic domain with protein tyrosine kinase enzyme activity. Ligand binding induces receptor dimerisation, activation of the kinase and phosphorylation of tyrosine residues in the cytoplasmic domain. The phosphorylated protein then recruits and activates a variety of intracellular second messenger systems that induce changes to the cytoskeleton and that can stimulate the rate of cell division. Brugge and colleagues have constructed altered receptors in which the extracellular domain is derived from the nerve growth factor receptor and cytoplasmic domains from either the EGFR or the c-erbB-2 receptor. In addition, however, there is a sequence fused to the C-termini of the constructs derived from the FK506-binding protein. The membrane-permeable bivalent compound AP1510 binds to this region and induces dimerisation, phosphorylation and intracellular signal transduction, allowing selective, acute or chronic receptor activation . In the study reported in Nature Cell Biology, Muthuswamy et al. introduced the chimeric EGFR and c-erbB-2 separately into MCF10A cells (immortal but untransformed human mammary epithelial cells) and established stable lines expressing moderate levels of either receptor. In the absence of AP1510, these grew in three-dimensional cultures, indistinguishable from the untransfected cells, as structures highly reminiscent of mammary epithelial acini. Each acinar structure contained approximately 20 --40 cells with basally located nuclei and a lumen. Staining for cell adhesion molecules showed that the cells were polarised and surrounded by a collagen layer similar to a basement membrane. The chimeric EGFR was able to stimulate growth of the MCF10A cells in monolayer culture, demonstrating that it was functional, but activation by the addition of AP1510 had no effect on cell behaviour or morphology when the cells were grown in three-dimensional cultures. Strikingly, however, activation of the c-erbB-2 protein caused marked changes in the polarised acinar structures. These lost their polarised organisation and grew into structures described as 'consisting of multiple acinar-like units with filled lumina'; in some cases, a hundred times the size of untreated structures. Muthuswamy et al. conclude that the EGFR and c-erbB-2 'have different abilities to affect polarised and growth arrested acini' and that 'acute activation of c-erbB-2 results in the generation of multi-acinar structures that share properties with structures associated with carcinoma in situ'. Significance of the results for future treatment : DCIS is often treated surgically by mastectomy rather than breast-conserving surgery since it is commonly multifocal and sometimes extensive . Failure to completely remove the disease is associated with a low but significant probability of the subsequent development of invasive cancer. It would be highly desirable to identify effective, low-toxicity adjuvant therapy that might justify the use of more limited surgery. The formal demonstration by Muthuswamy et al. that overexpression of c-erbB-2 in normal breast ductal epithelial cells leads to a DCIS-like phenotype provides strong support for studying such treatments in more detail in model systems. In independent work, Bundred and coworkers have developed a xenograft system in which small pieces of human DCIS are implanted subcutaneously in nude mice where the tissue survives and grows. Treatment of the mice with herceptin (a monoclonal antibody to the human c-erbB-2 protein now used for the treatment of some invasive breast cancers) did not have any effect, but treatment with iressa (a small-molecule tyrosine kinase inhibitor that acts principally on the EGFR but that will also, at higher concentrations, inhibit c-erbB-2) did inhibit cancer cell growth. It is also possible that heterodimers between EGFR and c-erbB-2 are important in signalling in this context, and that these will also be inhibited to a degree by the tyrosine kinase inhibitor. The presented results suggest that antibodies may not be the first method of choice for treatment of DCIS of the breast because they are large molecules and may have limited access to the interior of the cancer-filled duct. Orally available small-molecule tyrosine kinase inhibitors are more promising, and several pharmaceutical companies have selective c-erbB-2/EGFR inhibitors in development. Should these studies prove encouraging, there would be a rationale for testing these inhibitors in cases of ductal carcinoma of the breast in clinical trials. Abbreviations : DCIS = ductal carcinoma in situ; EGFR = epidermal growth factor receptor. Backmatter: PMID- 12223120 TI - Familial risks of breast cancer AB - Abstract | A recent analysis by the Collaborative Group on Hormonal Factors in Breast Cancer has provided the most precise quantification to date of the familial risks of breast cancer. The familial relative risks are shown to decrease from more than fivefold in women younger than age 40 years with a first-degree relative aged younger than 40 years at diagnosis, to 1.4-fold in women older than 60 years with a relative diagnosed over age 60 years. These risks increase progressively with the number of affected relatives. The risks associated with an affected mother and an affected sister are similar, and the relative (but not absolute) risks are similar in subgroups defined by other established breast cancer risk factors. These results provide a useful basis for counselling of women with a family history of breast cancer, and they have implications for the genetic basis of the disease. Keywords: BRCA1, BRCA2, familial breast cancer, reproductive factors, BRCA1, BRCA2, familial breast cancer, reproductive factors, Introduction : The increased risk of breast cancer among women with a family history of the disease is one of the oldest established facts about the disease. This familial aggregation has been the inspiration for studies to identify breast cancer susceptibility genes that have borne fruit over the past decade, and has been the basis for defining high-risk groups for intervention studies (e.g. with tamoxifen). Yet despite the fact that questions about family history are asked in almost every epidemiological study of breast cancer, some important questions about the quantitative relationship between family history relationship have not been answered with precision. Among these questions are the magnitude of the risk according to the age of the women and the age of their affected relative(s), the precise effect of numbers and types of affected relatives, and the joint effects of family history and other known risk factors. Since only ~10 --15% of women with breast cancer typically report a family history of breast cancer, individual epidemiological studies have not had the power to answer these questions precisely. The recent analysis by the Collaborative Group on Hormonal Factors in Breast Cancer goes a long way toward resolving some of these uncertainties. This group has brought together data from 52 studies, originally to evaluate the effects of oral contraceptives and hormone replacement therapy. In the current overview, the group examine risks according to family history of breast cancer in a first-degree relative in over 58,000 cases and in nearly 102,000 controls. Summary of main findings : The main results from the overview are straightforward to summarise. The results of the study are mainly expressed in terms of the risk ratio (or relative risk) of breast cancer associated with a family history; that is, the ratio of the incidence rate of breast cancer in relatives of breast cancer cases to the incidence in the relatives of controls. These risk ratios were estimated from the case-control studies in the usual way, and we refer to them as 'familial relative risks'. As anticipated, the familial relative risk of breast cancer declines progressively as both the age at diagnosis of the case and the age at diagnosis of the relative increases. The estimated relative risk is thus 5.7-fold when both the case and the relative are younger than age 40 years, but is only 1.4-fold when both are aged older than 60 years. (An anomalous observation here is that the relative risks decline more clearly with age of the case than with age of the relative; for cases older than age 60 years, there was no apparent trend in risk by age of relative. This anomaly might be explicable by inaccuracies in the reporting of a relative's age at diagnosis.) This pattern of risk was essentially the same whether the affected relative was a mother or a sister. For women aged younger than 50 years, the risk ratio associated with having a relative diagnosed younger than age 50 years was 2.41 (95% confidence interval, 1.86 --3.12) for an affected mother and was 3.18 (95% confidence interval, 2.15 --4.72) for having an affected sister. For women aged older than 50 years, the risk ratio associated with having a relative diagnosed older than 50 years was 1.60 (95% confidence interval, 1.38 --1.84) for an affected mother and was 1.44 (95% confidence interval, 1.19 --1.73) for an affected sister. The second observation was that the risk increased progressively with numbers of affected relatives. The risk ratios were 1.80 with one affected relative, 2.93 with two affected relatives and 3.90 (albeit with wide confidence limits) with three or more affected relatives. Third, the investigators performed detailed analyses to examine the effects of other known breast cancer risk factors in women with and without a family history, and conversely the effect of family history in women in categories defined by other risk factors. They found that the relative risks associated with other risk factors were essentially identical in women with and without a family history. In women with a family history, risk thus reduced with increasing parity, with earlier age at first child, and with earlier age at menopause to a similar relative extent (but, therefore, a larger absolute extent) as in women without a family history. There was no significant evidence of an association of oral contraceptives or hormone replacement therapy in women with a family history, but the confidence limits for these comparisons were extremely wide and the results were consistent with effects similar to the small increased risks seen in the general population. Finally, the investigators examined the effect of tumour spread on familial risk, but found that the familial risks were essentially the same whether or not the tumour was localised to the breast at diagnosis. Limitations of the results : An obvious concern in any studies of familial risk is the accuracy of reporting of cancer diagnoses in relatives. Other studies have indicated that breast cancer is fairly accurately reported, certainly in comparison with other cancer types (for example, ), but there is some inaccuracy, and hence potentially some bias in favour of reporting of cancers by cases. However, the investigators found the familial risks from the cohort studies alone (which are not susceptible to this bias) to be very similar. The main application of these results will be to genetic counselling and identification of high-risk women for screening and intervention studies. There are clearly some important limitations in this respect. Most obviously, the overview did not include data on genotypes at known susceptibility genes (data that would rarely be available in such studies), so these risk estimates would not apply in families where, for example, BRCA1 or BRCA2 mutation testing had taken place. At least for the present, however, genetic testing is mostly restricted to women with a strong family history (e.g. three or more affected relatives) so these empirical estimates will still be of value to the large majority of women. A more subtle problem is that no data on ages of unaffected female relatives were included. Clearly, in practice, the number and ages of unaffected female relatives do affect risk (although to a lesser extent than affected relatives do). Also, the overview had insufficient data to evaluate risks according to whether a daughter was affected, and no data at all on second-degree or more distant relatives (generally poorly recorded in case-control studies ). The latter two issues might be answered more reliably through cohort studies of families, such as those based on the Swedish Population Family Register . The overview did not consider risks according to cancers other than breast cancer in relatives. Other studies, however, have indicated that, with exception of a well established but modest risk of ovarian cancer (probably explicable in terms of the association of both cancers with BRCA1/2 mutations) and an association with childhood sarcoma (perhaps entirely due to TP53 mutations), there is little or no excess risk of breast cancer associated with a family history of other cancers . Implications of the results for breast cancer genetics : Although the aim of the overview was to present empirical risk estimates, these estimates do raise some interesting issues with regard to the genetics of breast cancer. Two studies have estimated that mutations in the BRCA1 and BRCA2 genes only account for approximately 15% of the excess familial risk of the disease , while the contribution of the other known breast cancer susceptibility genes (TP53, PTEN, CHK2 and ATM) is even smaller . The contribution of known genes is higher than this in certain populations where specific BRCA1 or BRCA2 mutations have become common as a result of founder effects. These populations include Iceland, Ashkenazi Jewish populations and parts of Poland. In most Western populations, however, the observed familial risks are largely the result either of other genes or of nongenetic familial risk factors. While the possibility of the latter cannot be definitively ruled out, results from twin studies suggest that the majority of the familial risk is in fact genetic in origin . (Since adjustment for known reproductive and other breast cancer risk factors had essentially no effect on the familial risks, nongenetic contributors to the familial risk, if there are any, must presumably be unrelated to the known risk factors.) The absence of substantial difference in risk by type of affected relative (i.e. affected mother versus affected sister) suggests that the important genes are likely to act dominantly or additively on risk, but not to act recessively (recessive susceptibility genes give rise to higher risks in siblings than in parents or offspring). Some studies using a family-based cohort approach have found higher risks in siblings, and these notably include studies of cases diagnosed at a particularly young age . These results do not necessarily conflict with the overview, since the risk estimates from the overview are imprecise at young ages. Conversely, it may be that the higher risk to siblings in the cohort studies is due at least in part to cohort effects on background incidence rates, and to the artefactually low rate of breast cancer in mothers who are, by definition, parous. The pattern of risk by the number of affected relatives is also revealing. The fact that the risk increases progressively with the number of affected relatives suggests the effect of a fairly large number of genetic risk groups, consistent with, for example, a polygenic model as proposed by Antoniou et al. . The trend in relative risk with age suggests that (like BRCA1 and, to a lesser extent, BRCA2) some or all of the susceptibility genes involved are likely to confer a higher relative risk at young ages. Finally, the similarity of the risk ratios for other risk factors in women with and without a family history suggest that these risk factors act to a similar extent in women at any level of genetic susceptibility. Of course, one cannot necessarily assume that this will hold for carriers of particular susceptibility mutations. Studies of BRCA1 and BRCA2 mutation carriers have shown that early menopause does have the expected protective effect in these groups , but the effects of other risk factors such as parity have not been definitively established . What are the implications for genetic counselling? Perhaps the most important is that the absolute risk of breast cancer in women with just one affected relative is relatively modest, even when the relative is diagnosed at a young age (the authors estimate a cumulative risk of 16% by age 80 years for women with a relative diagnosed younger than age 40 years). The risks associated with having larger numbers of affected relatives are more substantial, and referral to cancer genetics clinics should reflect this. The results of the present study also imply that the effects of reproductive and hormonal risk factors could be usefully incorporated into genetic counselling. Since the effects of family history and these other risk factors on breast cancer risks appear to combine in roughly multiplicative fashion, the absolute effects of risk factors in individuals with a strong family history can be substantial. Backmatter: PMID- 12223121 TI - Progesterone receptors - animal models and cell signaling in breast cancer: Role of steroid receptor coactivators and corepressors of progesterone receptors in breast cancer AB - Abstract | Progesterone, an ovarian steroid hormone, plays a key role in the development and function of the mammary gland, as it also does in the uterus and the ovary. The action of progesterone is mediated through its intracellular cognate receptor, the progesterone receptor (PR), which functions as a transcription factor that regulates gene expression. As with other nuclear receptors, coregulators (coactivators and corepressors) recruited by the liganded or unliganded PR, either to enhance or to suppress transcription activity, modulate the function of the PR. Mutation or aberrant expression of the coregulators might thus affect the normal function of the PR and hence disrupt the normal development of the mammary gland, which may lead to breast cancer. Keywords: breast cancer, coactivator, corepressor, progesterone receptor, breast cancer, coactivator, corepressor, progesterone receptor, Introduction : The progesterone receptor (PR) is a member of the nuclear receptor superfamily, which specifically regulates the expression of target genes in response to the hormonal stimulus. In the absence of progesterone, the PR is sequestered in a nonproductive form associated with heat shock proteins and other cellular chaperones. In this state, the PR is considered unable to influence the rate of transcription of its cognate promoters . On binding with progesterone, the PR undergoes a series of events, including conformational changes, dissociation from heat shock protein complexes, dimerization, phosphorylation, and nuclear translocation, which enables its binding to progesterone-response elements within the regulatory regions of target genes. The binding of PR to the progesterone-response elements is followed by the recruitment of coactivators and the basal transcription machinery, leading to the upregulation of target gene transcription. The PR exists as two isoforms in most rodents and humans, PR-A and PR-B, which are produced from a single gene by translation initiation at two distinct start codons under the control of separate promoters . The difference between PR-A and PR-B is that PR-A is a truncated form of PR-B. In humans, the N-terminal 164 amino acids of PR-B are missing in isoform PR-A. Detailed molecular dissection has identified two distinct activation function domains (AFs) within both PRs: AF-1, which is located in the N-terminal region, is ligand independent; AF-2, which is ligand dependent, is contained in the ligand-binding domain that is located in the C-terminal region. A DNA-binding domain and the hinge region are mapped to the central region of both receptors. Furthermore, a unique activation function domain, AF-3, is contained in the upstream segment of PR-B that is missing in PR-A . Figure 1 | Schematic representation of the progesterone receptor PR-A and PR-B proteins. Schematic representation of the progesterone receptor PR-A and PR-B proteins. The DNA-binding domain (DBD), the ligand-binding domain (LBD) and activation function domains (AFs) are indicated. Progesterone and estrogen are essential regulators of female reproductive activity. Through their cognate receptors, estrogen and progesterone regulate the normal development of the ovary, the uterus and the mammary gland, and play key roles in the tumorigenesis of these tissues. It has been demonstrated by estrogen receptor (ER) and PR knockout mice that estrogen controls the early ductal morphogenesis of the mammary gland, whereas progesterone controls ductal branching and alveolar development of the mammary gland during pregnancy . Although the two forms of the PR have similar structures and are identical in DNA and ligand binding, in vitro studies using a reconstituted progesterone-responsive transcription system in mammalian cells revealed that PR-A and PR-B are not functionally identical. In most cases, PR-B acts as a potent activator of transcription of target genes, whereas PR-A acts as a dominant repressor of transcription of PR-B as well as a few other nuclear receptors . The AF-3 domain in PR-B is partially responsible for the higher transcriptional activity of PR-B relative to PR-A. Moreover, an inhibitory function domain that is located in the N-terminus of both receptor isoforms has been identified . This inhibition function domain can inhibit the activity of AF-1 and AF-2 but not that of AF-3, which explains why PR-B is a potent activator of transcription. The inhibition function domain is functionally independent and is transferable; when placed upstream of the ER, the inhibition function domain can also suppress ER activity. The relative expression of PR-A and PR-B in the target tissues is dependent on species, cellular context, and the physiological and hormonal status. The ratio of PR-A to PR-B in specific tissues or cell types defines the physiological and pharmacological responses to progesterone. In the mammary gland, the ratios of PR-A to PR-B are constant from puberty to pregnancy, although there are species differences . From PR-A knockout mice, it has been demonstrated that PR-B is mainly responsible for the normal proliferative and differentiative responses of the mammary gland to progesterone, because PR-A knockout mice exhibit a similar phenotype to PR knockout mice . Overexpression of PR-A over PR-B in transgenic mice results in extensive epithelial cell hyperplasia, in excessive ductal branching, and in a disorganized basement membrane. All these features are associated with neoplasia . Furthermore, the development of the mammary gland in PR-B overexpressed transgenic mice is also abnormal , indicating that a regulated expression of PR-A and PR-B and the native ratios of the two isoforms are critical to the appropriate responsiveness of the mammary gland to progesterone. Consistent with the findings from animal studies, very low levels of PR-B and a consequently high PR-A:PR-B ratio were found in a significant proportion of human breast cancer samples . Taking these findings together, it can be inferred that imbalance of PR-A versus PR-B may be associated with the development, progression or prognosis of breast cancer. Coactivators are factors that can interact with nuclear receptors in a ligand-dependent manner and enhance their transcriptional activity. Corepressors are factors that interact with nuclear receptors and repress their transcriptional activity. Both types of coregulators are required for efficient modulation of target gene transcription by the PR . Changes in the expression level and pattern of PR coactivators or corepressors, or mutation of their function domains, might therefore affect the transcriptional activity of the PR and hence cause disorders of its target tissues, including the mammary gland. The present review will describe the coactivators and corepressors that are involved in the transcriptional modulation of PRs, with emphasis on their roles in breast cancer development and progression. Progesterone receptor coactivators : The steroid receptor coactivator family | The steroid receptor coactivator (SRC) family is composed of three distinct but structurally and functionally related members: SRC-1 (nuclear receptor coactivator 1), SRC-2 (transcription intermediary factor 2/glucocorticoid receptor-interacting protein 1/nuclear receptor coactivator 2), and SRC-3 (p300/CREB-binding protein [CBP] cointegrator-associated protein/receptor-associated coactivator 3/activator of thyroid and retinoid receptors/amplified in breast cancer 1/thyroid receptor activator molecule 1). SRC-1 was the first identified coactivator for the steroid receptor superfamily, which was cloned and characterized in 1995 . SRC-2 and SRC-3 were then identified thereafter by several laboratories . Sequence analysis of SRC proteins identified a basic helix --loop --helix domain and two Per --Arnt --Sim domains in the amino-terminal region. The basic helix --loop --helix/Per --Arnt --Sim domain is highly conserved among the SRC members, and it serves as a DNA binding and protein dimerization motif in many transcription factors . Following the basic helix --loop --helix/Per --Arnt --Sim domain, there are a centrally located receptor-interacting domain and a C-terminal transcriptional activation domain. Detailed analysis revealed three conserved LXXLL motifs (nuclear receptor box) in the receptor-interacting domain, which appear to contribute to the specificity of coactivator --receptor interaction. Histone acetyltransferase activity was identified in the C-terminal region of SRC members, and there also exist activation domains that can interact with the CBP. All three members of the SRC family interact with the PR and enhance its transcriptional activation in a ligand-dependent manner . Targeted deletion of the SRC-1 gene in mice has indicated that SRC-1 is important for the biological actions of progesterone in mammary gland development since the hormone-induced ductal elongation and alveolar development is greatly impaired in the null mice . In the meantime, the expression of SRC-2 mRNA was elevated in SRC-1 null mice, suggesting that SRC-2 can partially compensate for SRC-1 function . SRC-3 is the most distinct among the three members. It coactivates not only the nuclear receptors, but also other unrelated transcription factors such as those in the cAMP or cytokine pathways . Compared with the widespread expression of SRC-1 and SRC-2, expression of SRC-3 is restricted to the mammary gland and several other tissues . Disruption of the SRC-3 gene in mice causes severe growth and reproductive defects, including the retardation of mammary gland development . Furthermore, amplification and overexpression of SRC-3 were observed in 10% and 64% of human primary breast cancers, respectively . This observation indicates that SRC-3 is not only essential for the normal mammary development, but also plays a role in breast tumorigenesis. E6-associated protein/RPF1 | E6-associated protein (E6-AP) and RPF1, the human homolog of yeast RSP5, are E3 ubiquitin-protein ligases that target proteins for degradation by the ubiquitin pathway. They are also characterized as coactivators of steroid receptors. It has been demonstrated by transient transfection assay that RPF1 and E6-AP could potentiate the ligand-dependent transcriptional activity of the PR, the glucocorticoid receptor, and other nuclear receptors . Furthermore, RPF1 and E6-AP acted synergistically to enhance PR transactivation . Additionally, the coactivation functions of E6-AP and RPF1 are separable from the E3 ubiquitin-protein ligase activity, as ubiquitin ligase-defective E6-AP and RPF1 exhibited normal coactivation function. E6-AP is expressed in many tissues, including the mammary gland. From its ability to coactivate the PR and the ER in a hormone-dependent manner, it was assumed that E6-AP is an essential regulator for the development of normal mammary gland and mammary tumors. The first evidence of a relationship between E6-AP and breast cancers was obtained from the study of a spontaneous mouse mammary tumorigenesis model, which demonstrated that E6-AP was overexpressed in tumors when compared with normal tissues . We recently examined the expression pattern of E6-AP in biopsy samples of human breast cancers, and our results showed that E6-AP expression was decreased in tumors in comparison with the adjacent normal tissues (Gao et al., unpublished data, 2002). Furthermore, we demonstrated that the decreased expression of E6-AP was stage dependent, and that the expression of E6-AP was inversely correlated with that of the ER in breast tumors. Since the ER plays a major role in breast cancer development and the PR is a target of estrogen, the changes of the expression level of E6-AP might interfere with the normal functioning of the ER and the PR. Hence, E6-AP may participate in the formation and progression of breast tumors. Steroid receptor RNA activator | The growing family of nuclear receptor coactivators has recently acquired a unique member, steroid receptor RNA activator (SRA) . Differing from the other coactivators, SRA functions as a RNA transcript instead of as a protein. SRA specifically coactivates the transcriptional activity of steroid receptors, including the PR, the ER, the glucocorticoid receptor, and the androgen receptor. It has been demonstrated that SRA exists in a ribonucleoprotein complex containing SRC-1 and that it mediates transactivation through the AF-1 domain located at the N-terminal region of nuclear receptors, distinguishing it from the other coactivators. SRA is expressed in normal and malignant human mammary tissues . An elevated expression of SRA was found in tumors compared with the adjacent normal region . Although it is currently unknown whether the expression of SRA is correlated with that of the PR or the ER, the increase in the SRA levels in tumor cells may contribute to the altered ER/PR action that is known to occur during breast tumorigenesis. L7/SPA | L7/SPA is a 27 kDa protein containing a basic leucine zipper domain. L7/SPA is an antagonist-specific transcriptional coactivator because it can only potentiate the partial agonist activity of some antagonists, including tamoxifen and RU486, but has no effect on the agonist-mediated transcription . This unique property of L7/SPA suggests that it might play a role in the development of resistance to hormone therapy for breast cancers. CBP/p300 | CBP was initially characterized as a coactivator required for efficient transactivation of cAMP-response element-binding protein, and p300 was first identified as a coactivator of the adenovirus E1A oncoprotein. CBP and p300 share many functional properties: both of them function as coactivators for multiple nuclear receptors as well as p53 and nuclear factor-kappaB , both possess intrinsic histone acetyltransferase activity, and both can recruit histone acetyltransferase and CBP/p300-associated factor . Besides, CBP/p300 interacts with members of the SRC family and synergizes with SRC-1 in the transactivation of the ER and the PR . Other coactivators | In addition to the coactivators already discussed, there are a few other proteins that have been demonstrated to upregulate the transcriptional activity of the PR. Chromatin high-mobility group protein 1, chromatin high-mobility group protein 2, TIP60 (Tat-interacting protein), proline-rich nuclear receptor coregulatory protein 1, proline-rich nuclear receptor coregulatory protein 2, Cdc25B, and GT198 all function as PR coactivators, as demonstrated by transient transfection assays . Cdc25B is prominent among these coactivators in terms of its roles in breast cancer development, because Cdc25B transgenic mice exhibit mammary gland hyperplasia and increased steroid hormone responsiveness . The significance of all these coactivators in vivo needs to be further investigated. Progesterone receptor corepressors : Nuclear receptor corepressor/silencing mediator of retinoid and thyroid receptors | Nuclear receptor corepressor (N-CoR) and silencing mediator of retinoid and thyroid receptor (SMRT) are both corepressors of numerous transcription factors, including steroid hormone receptors. Both N-CoR and SMRT interact with the nuclear receptors through the receptor-interacting domains located in the C-terminal portion of the proteins, while their transcriptional repression domains are mapped to the N-termini . N-CoR and SMRT also associate with HDAC3 in large protein complexes, which is an important pathway for transcriptional repression. Corepressors N-CoR and SMRT interact with the nuclear receptors either in the absence of agonists (in the case of the thyroid receptor and the retinoid acid receptor) or in the presence of antagonists (in the case of steroid receptors) . Since N-CoR and SMRT are common corepressors for transcription factors, slight alteration of their expression level in certain tissues might result in significant transcriptional changes, leading to altered development of the mammary gland, even tumors. BRCA1 | BRCA1 is a breast cancer susceptibility gene, and its inherited mutations are correlated with an increased risk of breast cancer and ovarian cancer . The role of BRCA1 in cancer development is unclear. In addition to its ability to coactivate p53 and to modulate p300/CBP expression, BRCA1 is also a ligand-independent corepressor for the ER, the androgen receptor and the PR . If BRCA1 is mutated, all of these pathways will be more or less impaired. The effect of BRCA1 in cancer development might therefore be multiplex. Other corepressors | Ubiquitin-activating enzyme 3 (Uba3) is the catalytic subunit of the activating enzyme in the ubiquitin-like NEDD8 (neural precursor cell-expressed developmentally downregulated) conjugation (neddylation) pathway. Uba3 was recently demonstrated as a corepressor of the ER, the androgen receptor and the PR in mammalian transfection assays . Uba3 inhibited the transactivation of the ER in a time-dependent manner, and neddylation activity of Uba3 is required for this suppression. This suggests that Uba3 suppresses steroid receptor activity by promoting the termination of receptor-mediated gene transcription rather than by interfering with the initial events. Repressor of tamoxifen transcriptional activity (RTA) has recently been defined as a potent repressor of tamoxifen-mediated ERalpha transcriptional activity as well as an agonist of the ERbeta, the glucocorticoid receptor, and the PR . The interaction of RTA with the nuclear receptors requires the participation of RNA, because mutation of the RNA recognition motif in RTA compromises its ability to repress transcription . The roles of Uba3 and RTA in mammary gland development and tumorigenesis await further study. Summary : As a transcription factor, the PR activates target gene transcription in response to the hormonal stimulus, and its functions are modulated by coactivators and corepressors. Different coregulators exert their actions through different mechanisms, and involvement in the development of normal mammary gland and the formation or progression of tumors has been reported in some coactivators and some corepressors. The coactivators and corepressors of the PR so far identified are not PR specific, since they can also modulate the transactivation of many other nuclear receptors. In addition, no unique coregulators of PR-A or PR-B have been identified. Identification of PR-specific coregulatory proteins, especially PR-A interacting factors or PR-B interacting factors, is an important goal of future study. Abbreviations : AF = activation function domain; CBP = CREB-binding protein; E6-AP = E6-associated protein; ER = estrogen receptor; N-CoR = nuclear receptor corepressor; PR = progesterone receptor; RPF1 = receptor potentiation factor-1; RTA = repressor of tamoxifen transcriptional activity; SMRT = silencing mediator of retinoid and thyroid receptor; SRA = steroid receptor RNA activator; SRC = steroid receptor coactivator; Uba3 = ubiquitin-activating enzyme. Backmatter: PMID- 12223122 TI - Progesterone receptors - animal models and cell signaling in breast cancer: Expression and transcriptional activity of progesterone receptor A and progesterone receptor B in mammalian cells AB - Abstract | Progesterone is an essential regulator of normal female reproductive function. Its effects are mediated by two nuclear progesterone receptor (PR) proteins, PRA and PRB, which are identical except for an additional 164 amino acids at the N-terminal end of PRB. Transcriptional analyses of the two receptor forms have assigned strikingly distinct functional signatures to the two PRs, despite their apparent physical similarity. The basis of these differences is yet to be fully understood. Furthermore, these differences are strongly influenced by the cell type and the promoter used. We review the mammalian transcriptional studies of PRA and PRB, and compare them with what is known about their expression and function in target tissues. Keywords: progesterone receptor, progestin signaling, transcriptional activation, progesterone receptor, progestin signaling, transcriptional activation, Introduction : Progesterone plays a pivotal role in normal female reproduction, in the uterus, the ovary, the mammary gland and the brain . The number of cellular pathways regulated by progesterone reflects the complexity of its physiological role. In normal breast development, progesterone directs the formation of lobular-alveolar structures and also affects differentiation in the breast by modulation of milk protein synthesis . In the human endometrium, progesterone directs glandular differentiation and glycogenesis, as well as stromal proliferation and development of predecidual cells . These effects are mediated through the nuclear progesterone receptor (PR), expressed as two protein forms (PRA and PRB). There is increasing evidence to date that PRA and PRB are functionally different, and that it is the balance between these two forms that may make it possible for progesterone to affect such diverse physiological targets. Much of this evidence is conflicting and model specific, however, and the true differences between the receptor forms in normal tissues are yet to be fully understood. The purpose of the present article is to review the current state of knowledge about the transcriptional differences between PRA and PRB from mammalian in vitro models, and to correlate this with the effects of altered PRA and PRB levels on target tissues. Transcriptional regulation by the PR : The PR is a member of a large family of ligand-activated nuclear transcription regulators, which are characterised by organisation into specific functional domains and are conserved, to differing degrees, between species and family members. The PR is made up of a central DNA binding domain and a carboxyl-terminal ligand-binding domain. In addition, the receptor contains multiple activation function (AF) and inhibitory function elements, which enhance and repress transcriptional activation of the PR by association of these regions with transcriptional coregulators . In most species examined, the PR exists as two distinct forms (PRA and PRB). The expression of human PR is controlled by two promoters, which direct the synthesis of two distinct subgroups of mRNA transcripts encoding the two receptor proteins. The two PR forms are identical except that PRA lacks 164 amino acids contained at the N-terminal end of PRB. The region of the protein that is unique to PRB contains a transcription activation function, AF3 , in addition to AF1 and AF2, which are common to PRA. Newly transcribed cytoplasmic PR is assembled in an inactive multiprotein chaperone complex that dissociates on ligand binding and receptor activation. Progestin binding to the PR causes a conformational change and dimerisation, resulting in association of the progestin-complexed PR dimer with specific coactivators and general transcription factors. The activated complex binds to progestin response elements (PREs) in the promoters of target genes, resulting in modulation of transcription of those genes (reviewed in ). There is now considerable evidence for differences in the transcriptional activities of PRA and PRB from transient cotransfection into a variety of cell lines of PRA and/or PRB and reporter constructs containing progestin-responsive sequences. These constructs range from the simple PRE-tk-CAT (containing one copy of a palindromic PRE) to more complex constructs such as those incorporating the mouse mammary tumor virus long terminal repeat, which contains multiple hormone-responsive elements. PRB exhibits hormone-dependent transactivation in all cell types examined irrespective of the complexity of the response elements, whereas the transcriptional activity of PRA is cell specific and reporter specific. With reporter constructs containing a single palindromic PRE, PRA displays similar transactivation activity to PRB . However, this activity is reduced or inactive when more complex response elements such as the mouse mammary tumor virus long terminal repeat and PRE2TATAtk constructs are used . Interestingly, PRA acts as a transdominant inhibitor of PRB in situations where PRA has little or no transactivational activity . Moreover, PRA can regulate the transcriptional activity of other nuclear receptors such as glucocorticoid, mineralocorticoid, androgen and estrogen , suggesting that PRA may play a central role in regulation of activity of a number of nuclear receptors in addition to PRB. The ability of PRA to act as a transdominant repressor is highly model specific, however, and there is considerable variability between reports. McDonnell and Goldman reported that PRA but not PRB, in the presence of either progesterone or anti-progestins, lessened the ability of estrogen to induce an estrogen-responsive reporter when the two constructs were transfected into CV-1 or HS578T cells, but not into HepG2 cells. PRA had similar anti-estrogenic effects on endogenous estrogen receptor activation of a minimal estrogen-responsive reporter in MCF-7 breast cancer cells in the presence of RU 38486 . When the estrogen-responsive region of the pS2 gene was used as a reporter in MCF-7 cells, however, PRB and not PRA repressed activation of the reporter by estrogen . The mechanisms by which PRA and PRB exert such apparently different transcriptional activities in various cell and promoter systems remain largely unknown, although a number of possible scenarios have been proposed. The physical differences at the N-terminal end of the two receptors are clearly responsible for some transcriptional differences. In addition to the fact that AF3 is unique to PRB, the PRB-specific region has a distinct conformation in solution and is likely to mask an inhibitory domain that is active in the N-terminus of the PRA protein . This could act to enhance the transcriptional activity of PRB, as well as preventing it from acting as an inhibitor of other receptors. The unique AF in PRB may confer a difference in affinities of the two PRs for coregulators. When compared in a phage display assay, the two PR forms bind to distinct subgroups of peptides. This suggests that coactivators may bind differently to the two PRs or that the two receptors bind to different subgroups of coactivators . Motifs contained in AF3, with the same sequence as the NR boxes of coactivators, have been shown to be necessary for the transcriptional activity of the PRB-unique AF and may form contacts between the receptor and a unique set of cofactors, or within the PR dimer itself. Given that the PR acts in combination with multiple other transcription factors to affect transcription, it is possible that variability of the tissue-specific expression of the components of this multiprotein complex may result in different PRA and PRB activities in the same cell. Furthermore, differential cofactor requirements between gene promoters may lead to differences in the transcriptional efficacy of the two PRs on the same promoter . Further evidence is required to support these assertions since, although most cofactors are expressed at limiting levels, they have been demonstrated to be widely expressed throughout cell types. Furthermore, a degree of functional redundancy in PR coactivators seems to exist, with gene knockout studies of SRC-1 delivering only a mild phenotype and the demonstration that TIF2 and SRC-1 are able to activate the PR to a similar extent in transfections . The PR regulates the expression of a diverse population of transcriptional targets , and it would be expected that changes in the relative amounts of PRA and PRB would result in altered target gene expression patterns if the two isoforms are transcriptionally distinct. When patterns of gene regulation were examined in T47D breast cancer cells expressing exclusively PRA or PRB, a remarkably small overlap was seen between the sets of genes regulated by the two receptors, with the subset of genes regulated by PRB far exceeding in number those regulated by PRA . When the relative expression of PRA and PRB was varied in wild-type T47D cells that already express both isoforms, however, the impact on transcription was not dramatic unless PRA was in vast excess over PRB . Furthermore, no evidence was seen of dominant transcriptional inhibition by PRA. These data suggest that coexpression of both isoforms at similar levels, which is common in normal progesterone target cells (see later), is associated with appropriate transcriptional response to progestins and that changes in relative PRA and PRB levels must be quite dramatic before physiological changes in progestin signaling are observed. Expression and function of PRA and PRB in normal and malignant physiology : In general, PRA and PRB are coexpressed in the same target cells in the human , and their relative expression, where it has been examined, is generally close to unity . In some normal physiological circumstances, and in some cell types, there is a predominance of one isoform. PRA is always the predominant isoform in the uterine stroma, and PRB is predominantly expressed in the epithelial glands in the mid-secretory phase of the menstrual cycle . PRA is the predominant isoform in the rodent and is widely expressed in the macaque reproductive system. Predominance of one isoform also occurs in cancers. In breast cancers, the equivalent expression of PRA and PRB seen in normal cells is disrupted early in carcinogenesis, and predominance of one isoform is common . Most endometrial cancers express only one PR isoform, and isoform predominance is associated with higher histological grade . The association of PR isoform predominance with cancer is supported by the demonstration that transgenic mice overexpressing PRA exhibited features in their mammary glands that were abnormal and commonly associated with neoplasia . The expression of PRA and PRB in vivo supports a role for both isoforms in normal physiology, particularly in the human breast. The fact that the equivalent levels of the two proteins seen in the normal breast become disrupted early in breast carcinogenesis, and that predominance of one PR isoform, usually PRA, is seen in cancers, suggests that disrupted progesterone signaling may play a role in development or progression of breast cancer. Other mechanisms of PRA and PRB action : In addition to the ligand-activated transcriptional effects already discussed, which reflect the nuclear activity of this receptor, the PR also regulates transcription via alternative pathways. Ligand-independent activation of the PR can occur and provides evidence for regulation of the PR via membrane-generated signals . The PR also interacts with cytoplasmic signaling pathways to activate c-Src family members , and PR interactions with the mitogen-activated protein kinase pathway are also described. Migliaccio et al. found that PR activation of signaling pathways was independent of the transcriptional activity of the receptor and was indirectly mediated by PRB, but not PRA, through interaction with the estrogen receptor . Other studies have examined the ability of the PR to transiently activate mitogen-activated protein kinase pathways but have generally focused on PRB, which was more transcriptionally active than PRA in most settings (CA Lange, personal communication). Conclusion : The in vitro data on the relative activities of PRA and PRB tend to support the view that PRB is the active PR, whereas PRA is either inactive or acts as an inhibitor of PRB activity. However, this perspective is at odds with the demonstrated coexpression of both isoforms in normal physiology and with the distinct roles ascribed to each protein from knockout studies discussed in the other sections of this review series. An explanation for the discrepancy between the in vivo and the in vitro data resides in the experimental protocols used to examine the question of relative PRA and PRB transcriptional activity. Most of the information has been obtained from transient transfection studies largely in cell lines that are not progestin targets, using exogenously transfected reporter sequences and each isoform in isolation. The relative levels of isoform protein coexpressed under these conditions cannot be known, and is probably highly relevant to interpretation of the data. There is little evidence in vivo that PRA is a dominant inhibitor of PRB. Moreover, when the data in human tissues and null animals are taken together, they suggest that the two isoforms either work cooperatively to mediate progesterone action or suggest that each isoform has distinct physiological roles that are probably cell specific and promoter specific. A combination of cooperative action and distinct activity is probably the best explanation for the complex and divergent pathways of progesterone action in normal and malignant physiology. Abbreviations : AF = activation function; PR = progesterone receptor; PRE = progestin response element. Backmatter: PMID- 12223123 TI - Progesterone receptors - animal models and cell signaling in breast cancer: Progesterone's role in mammary gland development and tumorigenesis as disclosed by experimental mouse genetics AB - Abstract | The progesterone receptor knockout mouse demonstrated progesterone's importance to parity-induced mammary tertiary branching and lobuloalveologenesis. Because early parity provides significant protection against breast cancer whereas prolonged exposure to premenopausal ovarian progesterone (or to postmenopausal supplementations thereof) has been linked to breast cancer risk, this steroid can be considered to exhibit contrasting roles in breast cancer etiology. This review describes the important mouse models that have contributed to our understanding of progesterone's role in mammary gland development and neoplasia. We conclude by emphasising the urgent need to identify the molecular targets of the progesterone receptor, and to determine whether these targets are modulated differently by the progesterone receptor isoforms (A and B) during mammary morphogenesis and tumorigenesis. Keywords: knockout, mammary gland, progesterone, progesterone receptor, tumorigenesis, knockout, mammary gland, progesterone, progesterone receptor, tumorigenesis, Introduction : Epidemiological studies have shown that early onset of menarche, delayed entry into menopause, cycle periodicity, nulliparity, and a late first pregnancy represent individual risk factors for breast cancer. However, early menopause and early first parity decrease this risk . Because progesterone's presence or absence directly influences the establishment of each of these reproductive endocrine states, assessing mammary gland development and tumorigenesis in relation to progesterone exposure is of paramount importance. As for most epithelial cancers, the incidence of human breast cancer is age dependent, implicating a stochastic multistep progression in the development of this disease . Although breast cancer incidence increases with age, the increase is not uniform but achieves its highest rate during the reproductive years of premenopause . The data suggest that ovarian steroidal exposure of the mammary epithelial cell (MEC) during this reproductive period underlies the risk to breast cancer presented by many of the aforementioned reproductive endocrine states . One hypothesis to account for the link between ovarian steroidal exposure and human breast cancer risk is that ovarian steroid-induced MEC proliferation provides a temporal window of opportunity for the progressive acquisition of genetic errors . As a result of these errors, the transformed MEC is predicted to undergo unchecked clonal expansion to a mammary neoplasm. Although estrogen's proliferative effects on mammary gland development and tumorigenesis are well recognized, progesterone's contribution to these processes has been mired in controversy . Much of this controversy has been exacerbated in the past by: first, our inability to mechanistically dissect the individual roles of estrogen and progesterone in vivo; second, progesterone's established inhibitory role on estrogen-induced luminal epithelial proliferation in the uterus, and the extrapolation of this inhibitory role to the mammary gland; and third, the increasing number of conflicting reports concerning the importance of synthetic progestins in mammary tumor progression in the rodent . The acknowledged gaps in our understanding concerning progesterone's involvement in human breast cancer has exposed the urgent need to better understand the role played by progesterone as an endocrine mammogen in normal mammary gland development. In the present review, we provide an overview of how experimental mouse genetics has not only defined progesterone's unique contribution to mammary gland morphogenesis and function, but has also aided in the formulation of new concepts concerning the role of this ovarian steroid in the normal and neoplastic development of this tissue. Postnatal mammary gland development | The mouse as an experimental tool of inquiry | Because many of the fundamental aspects of mammary gland development and tumorigenesis are conserved between rodents and humans , the rodent model (in particular, the rat and the mouse) has historically served as the experimental system of choice for in vivo mammary gland research. With the recent advent of powerful genetic approaches to manipulate the murine genome, the mouse offers an unprecedented level of sophistication with which to query systemic or locally acting hormonal controls on mammogenesis, from the genetic, cellular, and molecular perspectives. As with the human, murine postnatal mammary gland development consists of two distinct allometric growth stages. The first of these stages occurs at the onset of puberty, whereas the second manifests in response to pregnancy . At puberty, in response to systemic estrogen and locally acting growth factors, cap cells of terminal end buds undergo extensive mitosis to drive ductal elongation and dichotomous branching to the limits of the fat pad. On reaching adulthood, the mammary gland is growth quiescent except for incipient side branching and alveolar budding (conspicuous in most strains, but not all), which develop over time as a result of the exposure of the tissue to cycling levels of ovarian steroids. Figure 1 | Progesterone receptor function is required for mammary ductal side-branching and alveologenesis. Progesterone receptor function is required for mammary ductal side-branching and alveologenesis. (a) The salient postnatal stages of mammary gland development. Whole mounts of (b) transplanted progesterone receptor knockout (PRKO) mammary glands and (c) wild-type mammary glands taken from a nulliparous host, and (d) transplanted PRKO mammary glands and (e) wild-type mammary glands taken from a parous host. Scale bar in (b) denotes 500 m and applies to all whole mounts. Adapted from Lydon et al. . In response to the hormones of pregnancy, the epithelial compartment of the mammary gland undergoes proliferation and subsequent differentiation to generate alveoli that progressively fill the interductal spaces during late pregnancy, parturition, and lactation. Following weaning, removal of the suckling stimulus triggers collapse of the lobuloalveolar system through proteinase-mediated and apoptotic-mediated reductive remodeling processes, collectively termed involution. On completion of involution, the reinstatement of the mammary gland to a ductal architecture resembling the mammary phenotype of the prepregnant mouse completes the cycle of development . Importantly, epidemiological and experimental investigations have provided irrefutable evidence that completion of this developmental cycle early in the reproductive life in rat, mouse and human species confers a significant protection against breast cancer in later life . However, this protection is lost with a late first pregnancy. Given that the stage of mammary gland development at the time of carcinogen exposure greatly influences subsequent breast cancer initiation and progression, experimental mouse genetics has recently been applied to delineate progesterone's role in the elaboration of each of these developmental stages. The progesterone receptor knockout mouse | The progesterone receptor | Most of the physiological effects of progesterone are mediated by its intracellular receptor, the progesterone receptor (PR), which is a member of the nuclear receptor superfamily of transcription factors . On binding progesterone ligand, the nuclear receptor ligand complex translocates to the nucleus to induce or to silence the expression of downstream target genes, which in turn manifest the physiological response of the target cell to the hormone, progesterone. In most physiological systems, including the mammary gland, the PR is induced by estrogen, through its cognate nuclear receptor: the estrogen receptor. The PR comprises two isoforms, PR-A and PR-B, which are expressed from the same gene and are structurally identical except for a short amino acid extension contained in the N-terminal region of PR-B. Previous transient transfection experiments demonstrated that the two PR isoforms exhibit distinct transactivational functions, which are dependent on the cell of origin and on the target gene promoter context . Specifically, in cell types in which PR-A is inactive, the PR-B isoform (in the absence of PR-A) is a strong transactivator of several PR-regulated promoters . In a physiological setting, these results suggest that the PR-A and PR-B isoforms may regulate distinct sets of target genes and may exhibit different transactivational capabilities in a given progesterone-responsive target tissue. Moreover, in cell and promoter contexts in which PR-A lacks transactivational activity, the coexpression of PR-A and PR-B demonstrated that the PR-A isoform could act as a dominant repressor of PR-B activity. In a cellular context in which PR-A and PR-B are coexpressed, this observation suggests that PR-A has the ability to attenuate the general progesterone responsiveness of specific PR target genes in specific target tissues in vivo. Characterization of the progesterone receptor knockout mouse | To directly examine the physiological significance of PR function in the murine mammary gland, a progesterone receptor knockout (PRKO) mouse model was generated in which both PR isoforms were simultaneously abrogated through gene targeting approaches . Initial phenotypic characterization of the PRKO mouse revealed that removal of PR function resulted in a spectrum of reproductive abnormalities in the female, which included severe endocrine defects, an intrinsic impairment in ovulation, a dysfunctional uterus, and a loss in mating behaviors . Unlike the estrogen receptor-alpha knockout mouse, in which the absence of estrogen signaling resulted in a block in mammary ductal outgrowth at puberty , the PRKO mouse mammary gland exhibited normal ductal elongation to generate a simple ductal architecture that was morphologically similar to the mammary gland of the young wild-type virgin . Together, the estrogen receptor-alpha knockout and PRKO mouse models highlight the specific importance of ovarian estrogen rather than progesterone in epithelial ductal elongation in the pubescent mammary gland. To address whether the progesterone signal is requisite for the observed increase in ductal side branching and lateral alveolar budding in the cycling nulliparous mouse, as well as for the observed morphological responses to the full spectrum of pregnancy hormones in the parous mouse, PRKO MECs were transplanted into the cleared mammary fat pad of a wild-type host . The whole mammary gland transplantation approach was necessary to circumvent the inherent infertility defect of the PRKO mouse. In the case of the cycling nulliparous host, the absence of ductal side branching and lateral alveolar budding in glands containing transplanted PRKO MECs as compared with transplanted wild-type MECs supports a role for progesterone in ovarian hormone-driven mammary epithelial cell proliferation. As indicated in the Introduction, ovarian steroidal-induced mammary epithelial proliferation has been implicated as a basis for nulliparity-associated breast cancer risk . Indeed, the recent utilization of the PRKO mouse in combination with the chemical carcinogen-induced mammary tumor model demonstrated the critical role that progesterone-induced mammary epithelial proliferation can play in the initiation and progression of carcinogen-induced mammary tumors . As a corollary to the aforementioned PRKO mammary tumor studies, recent investigations have indicated that progesterone may also facilitate chromosomal instability (aneuploidy) in the subsequent stages of mammary tumor progression that follow the loss of p53 function . In the parous host, despite exposure to the hormonal milieu of pregnancy, transplanted PRKO MECs failed to elicit further ductal side-branching and lobuloalveologenesis, as observed with transplanted wild-type MECs . Because early parity provides significant protection against breast cancer in later life, understanding the cellular and molecular mechanisms by which progesterone exerts these pregnancy-associated morphological and functional changes in the mammary gland is now an important priority for contemporary mammary gland research. To address this issue, recent studies have provided convincing evidence that progesterone (with estrogen) exerts a pivotal role in the elaboration of persistent molecular changes (i.e. the activation of p53-signaling pathways) in a subgroup of mammary epithelial cells that may contribute to parity-induced protection against mammary cancer . Indeed, elegant studies recently reported by Wagner et al. suggest that these parity-induced mammary epithelial populations may be identified and isolated in the near future. Future directions | A paracrine mode of action | As in the human mammary gland , immunohistochemical studies on the murine mammary gland have demonstrated that PR expression is restricted to the luminal epithelial cell . Importantly, recent immunofluorescence experiments have revealed that the majority of epithelial cells, which undergo proliferation in response to progesterone, are segregated from, but in close apposition to, PR-positive cells . These observations support a paracrine mode of action in which PR-positive cells (in response to progesterone) express and secrete a paracrine mediator(s) that impacts a neighboring PR-negative cell to proliferate . Interestingly, a paracrine mode of action for progesterone has also been shown to exist in the uterus . Unlike the mammary gland, however, the progesterone-induced paracrine circuit emanates from the uterine stromal compartment to influence luminal epithelial proliferation. Figure 2 | Paracrine mode of action for progesterone-induced branching morphogenesis in the mammary gland. Paracrine mode of action for progesterone-induced branching morphogenesis in the mammary gland. (a) Indirect immunofluorescence analysis of a rat mammary gland shows luminal epithelial cells expressing the progesterone receptor (PR) (red) or undergoing proliferation (green) as measured by 5-bromo-deoxyuridine. The arrow indicates a single proliferating cell (yellow), which also expresses the PR. Scale bar represents 50 m. (b) All nuclei in the field shown in (a), as detected by 6-diamidino-2-phenylindole incorporation. Details of this experiment are described elsewhere . (c) The conceptual idea of a paracrine mode of action for progesterone in mammary branching morphogenesis. 'X' denotes the elusive downstream paracrine mediator(s) of the progesterone (P4) proliferative signal. Adapted from Lydon et al. . The finding that the nonuniform cellular organization pattern for PR-positive cells and proliferating PR-negative cells is conserved between the human mammary gland and the rodent mammary gland has provided strong support for an evolutionary conserved cellular mechanism of action by which PR-positive cells influence neighboring PR-negative cells to participate in ductal morphogenesis. Although the reasons for such an evolutionary conserved nonuniform cellular organization for PR expression is uncertain, derailment of such an important cellular arrangement is predicted to adversely affect normal mammary gland development. Indeed, in the case of many steroid receptor positive human breast tumors, the majority of estrogen and progesterone receptor positive tumor cells also undergo proliferation, clearly contravening the paracrine signaling circuit that exists in the normal gland . Although speculative, the above studies suggest that the genesis of these tumors may require an early switch in steroid-dependent regulation of proliferation from a paracrine to an autocrine mechanism. Additional support for the importance of this cellular organization pattern for PR expression to normal mammary gland development comes from the recent analysis of the adult CCAAT/enhancer binding protein beta knockout mammary gland, which does not respond to the proliferative effects of estrogen and progesterone . In contrast to a nonuniform cellular distribution pattern for PR expression observed in the normal gland, the CCAAT/enhancer binding protein beta knockout mammary gland exhibits a uniform cellular organization pattern for PR-positive cells with the attendant absence of PR-negative cells that proliferate in response to the progesterone paracrine signal. Clearly, to gain a more meaningful mechanistic understanding of progesterone's paracrine mode of action in the mammary gland, identification of the downstream genetic networks that relay the progesterone signal will be essential. Towards this end, recent studies have provided strong evidence that the wnt-4 signaling pathway may act as an important paracrine mediator of the progesterone-initiated proliferative signal in the mammary gland . With the availability of the complete murine genome sequence, of gene discovery approaches, and of genetic strategies to conditionally turn on or turn off potential target genes in the murine mammary gland, we anticipate that progesterone's mammogenic effects will be readily exposed to molecular dissection in the near future. PR-A and PR-B | Transgenic approaches have recently been employed to address whether the PR isoforms (PR-A and PR-B) mediate distinct physiological effects in the mammary gland . Overexpression of PR-A resulted in increased ductal branching and eventual hyperplasia , whereas PR-B overexpression caused reduced ductal elongation and branching . These in vivo investigations strongly support the concept that progesterone's physiological signal is interpreted differently by its two receptor isoforms, and that misexpression of either one of these isoforms can result in aberrant mammary gland development. Using gene targeting approaches, a specific knockout for the PR-A isoform (PRAKO) has recently been generated and characterized . The PRAKO mouse, like the PRKO mouse, exhibited a infertility phenotype. Unlike the PRKO mouse, however, the PRAKO mammary gland developed normally in the absence of PR-A function . Although PR-A is expressed in the mammary gland, these studies suggest that, unlike most other progesterone target tissues, the PR-B isoform may functionally compensate for PR-A in the murine mammary gland. To determine whether PR-B is indispensable for mammary development or whether its actions can be compensated by PR-A, a specific knockout for the PR-B isoform (PRBKO) mouse model has recently been generated and is currently being characterized. Unlike the PRAKO mouse, the PRBKO female is fertile. Initial mammary gland studies, however, indicate that loss of PR-B function results in reduced pregnancy-associated ductal branching compared with wild-type mammary glands (Conneely OM, unpublished data, 2002). These data, although preliminary, suggest that most of the mammogenic effects of progesterone are mediated by the PR-B isoform in the mouse. Conclusions : Experimental mouse genetics has highlighted progesterone's indispensable role in pregnancy-induced morphological and functional changes in the mammary gland. If the aforementioned developmental changes occur early in reproductive life, a significant protection against breast cancer can be achieved. Because prolonged progesterone exposure, either through uninterrupted cyclical ovarian activity or by postmenopausal supplementation, has been linked to breast cancer risk, progesterone can be viewed as an endocrine mammogen with contrasting roles with respect to breast cancer etiology. With this view in mind, our concepts of breast cancer risk in relation to progesterone exposure will remain just that until the downstream genetic pathways through which progesterone affects normal and neoplastic mammary development are identified. With the availability of PRKO mouse models for PR-A, PR-B or both isoforms, in combination with DNA microarray approaches, we believe this goal will soon be realized. Abbreviations : MEC = mammary epithelial cell; PR = progesterone receptor; PRAKO = a specific knockout for the PR-A isoform; PRBKO = a specific knockout for the PR-B isoform; PRKO = progesterone receptor knockout. Backmatter: PMID- 12223124 TI - Progesterone receptors -- animal models and cell signaling in breast cancer: The role of oestrogen and progesterone receptors in human mammary development and tumorigenesis AB - Abstract | A relatively small number of cells in the normal human mammary gland express receptors for oestrogen and progesterone (ER and PR), and there is almost complete dissociation between steroid receptor expression and proliferation. Increased expression of the ER alpha (ERalpha) and loss of the inverse relationship between receptor expression and proliferation occur at the very earliest stages of tumorigenesis, implying that dysregulation of ERalpha expression contributes to breast tumour formation. There is evidence also for alterations in the ratio between the two PR isoforms in premalignant breast lesions. Elucidation of the factors mediating the effects of oestradiol and progesterone on development of the normal breast and of the mechanisms by which expression of the ERalpha and the PR isoforms is controlled could identify new targets for breast cancer prevention and improved prediction of breast cancer risk. Keywords: breast tumours, normal mammary epithelium, oestrogen receptor, progesterone receptor, breast tumours, normal mammary epithelium, oestrogen receptor, progesterone receptor, Introduction : The human mammary epithelium is the tissue from which most breast tumours arise. Understanding how processes such as proliferation and differentiation of the epithelium are controlled by the ovarian steroids oestradiol and progesterone may lead to an increased understanding of the carcinogenic process. The present article reviews some of what is known about the involvement of the receptors for oestradiol and progesterone in the normal mammary gland and in tumorigenesis. Structure of the human mammary gland : The mammary gland is not completely formed at birth, but begins to develop in early puberty when the primitive ductal structures enlarge and branch . Once ovulatory menstrual cycles have begun, branching of the ductal system becomes more complex and lobular structures form at the ends of the terminal ducts to produce terminal ductal lobular units (TDLUs), which become more complex with successive menstrual cycles. During early pregnancy, there is another burst of activity in which the ductal trees expand further and the number of ductules within the TDLUs increases greatly. These ductules differentiate to synthesise and secrete milk in late pregnancy and subsequent lactation. The entire ductal system of the human mammary gland is lined by a continuous layer of luminal epithelial cells that is, in turn, surrounded by a layer of myoepithelial cells. These myoepithelial cells are in direct contact with the basement membrane, and the TDLUs are surrounded by delimiting fibroblasts and embedded in a specialised intralobular stroma. Histological studies have shown that most human breast tumours appear to be derived from TDLUs and have morphological characteristics of luminal epithelial cells (reviewed in ). Moreover, most human breast tumours retain the biochemical features of luminal cells in that they express the appropriate cytokeratins and membrane antigens such as MUC-1 . Human tumours also contain receptors for oestradiol and progesterone that, in the normal breast, are expressed only in the luminal epithelial cell compartment. Luminal epithelial cells must therefore be regarded as the primary targets for malignant transformation and subsequent tumour formation. The process of breast tumorigenesis is thought to result from a 'benign to malignant' progression in which the accumulation of multiple genetic changes allows evolution from normal breast epithelium through benign proliferative lesions to atypical proliferative lesions, and then to carcinoma in situ and frankly invasive tumours. This progression is elegantly reviewed by Allred and colleagues , who report that the lesions associated with the greatest risk of invasive breast cancer are hyperplasia of usual type, atypical ductal hyperplasia, ductal carcinoma in situ (DCIS) and lobular carcinoma in situ. Ovarian steroids, breast development and tumorigenesis : The clinical and epidemiological evidence for an obligate role of oestrogen in human mammary gland development and tumorigenesis is considerable. There is complete failure of breast development in the absence of intact ovarian function, and oestradiol-replacement therapy is necessary to induce breast development . Increased exposure to the fluctuating levels of oestradiol of the menstrual cycle through early menarche, late menopause or a late, first, full-term pregnancy increases breast cancer risk, as does use of exogenous oestrogens in the form of the oral contraceptive pill or hormone replacement therapy . More compellingly, treatment with anti-oestrogens reduces the incidence of breast cancer in high-risk women . The obligate role for oestradiol in mammary gland development and tumour formation has been confirmed in studies on mice where the gene for the ERalpha has been knocked out . The mammary glands in these ERalpha knockout mice comprise rudimentary ducts confined to the nipple area, which cannot be induced to develop further with oestradiol treatment and which are resistant to malignant transformation following transduction with oncogenes. There is far less evidence for a role of progesterone in human breast development. Studies on mouse models in which the PR has been knocked out suggest that, whereas oestradiol stimulates ductal elongation and PR expression, progesterone induces lobuloalveolar development . Generally, it is assumed that progesterone plays a similar role in the human breast and stimulates TDLU formation and expansion during puberty and pregnancy. As far as is known this has never been demonstrated, although this might be because it is almost impossible to study human breast tissue at these stages of development. As far as a role for progesterone in breast tumorigenesis is concerned, there are now some data suggesting that exogenous progestins taken in the form of combined hormone replacement therapy increase the risk of postmenopausal breast cancer to a greater extent than use of oestrogen replacement therapy alone . Effects of oestrogen and progesterone are mediated by the ER and by the PR : Steroid hormones such as oestradiol and progesterone are lipophilic and they enter cells and their nuclei primarily by diffusing through the plasma and nuclear membranes. Once in the nucleus, the steroids encounter proteins known as receptors because they bind their cognate ligands with high affinity and specificity. There are two receptors for oestradiol, the ERalpha and the ERbeta. Both these ERs are members of the steroid/thyroid hormone nuclear receptor superfamily and both may be described as ligand-dependent nuclear transcription factors. The ER proteins have the modular structure that typifies the nuclear receptor superfamily, which includes domains that mediate binding to ligands and to DNA. Although the two ERs are homologous in their DNA-binding and steroid-binding domains, the ERbeta gene is smaller, it has a different chromosomal location and it encodes a shorter protein . The distinctly different but overlapping tissue distribution of the ERbeta compared with the ERalpha suggests that it might mediate some of the non-classical effects of oestrogens and anti-oestrogens. Alternatively, the results of experimental studies suggest that the ERbeta might interact with and negatively modulate the actions of the ERalpha . Progesterone also has two receptors, PRA and PRB. Unlike the ERs, however, these two receptors are transcribed from the same gene via alternative promoter usage. PRB is longer than PRA as it contains an additional 164 amino acids at its N-terminal, but otherwise the two proteins are identical . PRA and PRB also are members of the steroid/thyroid hormone nuclear receptor superfamily, and they function as ligand-dependent nuclear transcription factors. It has been suggested that PRB is the major activator of gene transcription and that PRA is a repressor of PRB activity . However, more recent studies on breast cancer cells engineered to express either PRA or PRB alone or on mice in which the isoforms have been selectively deleted suggest that PRA as well as PRB can activate gene transcription. Moreover, the two isoforms can be differentiated in terms of the profile of genes that they can activate and by the fact that PRB, but not PRA, mediates the effects of progesterone on mouse mammary gland development . ER and PR expression in normal human breast : Most data on ER and PR expression in the normal human breast have been obtained in the course of studies on tissue from adult women who are neither pregnant nor lactating. These studies show that ERalpha is expressed in approximately 15 --30% of luminal epithelial cells and not at all in any of the other cell types within the human breast . Studies on the expression of ERbeta in either normal or malignant human breast epithelium have been hampered by a lack of antibodies that can reliably detect the protein in sections of formalin-fixed, paraffin-embedded tissue. Such antibodies have recently been developed , however, and initial studies indicate that the ERbeta is expressed in most luminal epithelial and myoepithelial cells, as well as being detectable in fibroblasts and other stromal cells within the normal human breast . Unfortunately, this widespread distribution is not very informative as regards the function of the ERbeta in the normal breast. The results of studies on mice in which the ERbeta has been deleted are similarly uninformative as the mammary glands develop normally in these mice and they appear to have no difficulty in nursing their young . These data thus suggest that, despite its more restricted pattern of expression, the ERalpha is the key mediator of oestradiol action in the normal mammary gland and suggest that further studies are required to establish the role of ERbeta. Most of the investigations where immunohistochemistry was used to determine the level and distribution of PR expression in the normal human breast were carried out before reagents capable of distinguishing the two isoforms became available. Nevertheless, these studies showed that, like the ERalpha, the PR was present in 15 --30% of luminal epithelial cells and not elsewhere in the breast . Dual-label immunofluorescent techniques have been used to show that all cells expressing the PR also contain the ERalpha. In contrast, steroid receptor-expressing cells are separate from, but often adjacent to, these labelled with markers of proliferation . This dissociation between steroid receptor expression and proliferation has been confirmed by other groups in both human breast and in rodent mammary glands . The current hypothesis is that oestradiol and/or progesterone controls the proliferative activity of luminal epithelial cells indirectly in a mechanism where the receptor-containing cells act as 'sensors' that secrete positive or negative paracrine and/or juxtacrine growth factors, according to the prevailing oestradiol/progesterone concentrations, to influence the activity of nearby division-competent cells. This would attenuate the sensitivity of the breast epithelium to steroid hormones such that proliferation will occur only when a sufficient concentration of positive growth factors has accumulated. This might be achieved only after prolonged exposure to high levels of steroid and possibly other hormones, as in early pregnancy, and may be a mechanism for preventing excessive proliferative activity at other times. Relationship between the ER, the PR and proliferation in tumorigenesis : Increased ERalpha expression may be one of the very earliest changes occurring in the tumorigenic process. Khan and colleagues have shown increased ERalpha expression in normal epithelium taken from tumour-bearing breasts. In addition, ERalpha expression is higher in the breast tissue of women from a population at high risk of breast cancer compared with that in the tissue of Japanese women who have a relatively low risk of the disease . ERalpha expression is increased at the very earliest stages of ductal hyperplasia and increases still further with increasing atypia, such that most cells in atypical ductal hyperplasias and in DCIS of low and intermediate nuclear grade contain the ERalpha . There are fewer ERalpha-positive cells in DCIS of high nuclear grade, but the expression of markers such as c-erbB-2/HER-2 suggests that these lesions form a different pathway to invasive cancer. As ERalpha expression increases, the inverse relationship between receptor expression and proliferation becomes dysregulated. There are increasing numbers of cells expressing both the ERalpha and the Ki67 proliferation-associated antigen with progression toward malignancy, and this is another early change associated with the process of breast tumorigenesis . Interestingly, a proportion of hyperplasias of usual type also contain proliferating ERalpha-positive cells, and it remains to be seen whether these lesions are the ones that progress to invasive tumours. Approximately 70% of invasive breast carcinomas contain the ERalpha, and preliminary studies indicate that most of these tumours contain ERalpha-positive, proliferating cells . Clearly, patients whose invasive tumours contain the ERalpha are suitable for endocrine therapy, but there is no evidence that dysregulation of the relationship between receptor expression and proliferation has any influence on their response. This is in keeping with the suggestion that dysregulation is an important step in early tumorigenesis but is less important at later stages. There are some data showing that ERbeta expression is downregulated in lesions such as atypical ductal hyperplasia and DCIS when compared with that in normal breast epithelium . The same group has shown that the receptor is inversely correlated with proliferation and that the ratio between the ERalpha and the ERbeta increases with increasing atypia. This is consistent with the suggestion that the ERbeta negatively modulates the effects of the ERalpha . The data with regard to ERbeta expression in invasive tumours and its relationship to prognosis or response to endocrine therapy are somewhat contradictory, with some groups reporting that the presence of this receptor is a good prognostic factor and others reporting the reverse . There are a few studies on PR expression in premalignant and preinvasive lesions, and these few suggest that expression of the PR also increases with increasing atypia . There is some evidence suggesting that the ratio between PRA and PRB is altered during tumorigenesis, such that PRA predominates . How this can be reconciled with the suggestion that PRA acts as a dominant repressor of the action of PRB and other steroid receptors has yet to be determined, but these data suggest that alteration of the PR isoform ratio also has a role in human breast tumorigenesis. Approximately 60% of invasive breast carcinomas express PRA and/or PRB, and PR expression is regarded generally as a marker of intact ERalpha function . Patients whose tumours contain both the ERalpha and the PR have the greatest probability of responding to endocrine therapy and have a better prognosis than those whose tumours do not contain steroid receptors. Whether the PR isoform ratio has any bearing on response to endocrine therapy remains to be determined. Conclusions : There is almost complete dissociation between steroid receptor (ERalpha and PR) expression and proliferation in the normal human mammary epithelium, suggesting that the ovarian steroids oestradiol and progesterone control proliferation and development of the mammary gland indirectly via the secretion of paracrine growth factors. This may be one way of attenuating the sensitivity of the normal mammary epithelium to the effects of the ovarian steroids and of ensuring that significant proliferative activity occurs only when it is needed (i.e. during puberty and pregnancy). Increased ERalpha expression and loss of the inverse relationship between steroid receptor expression and proliferation occurs at the earliest stages of breast tumour development, implying that dysregulation of ERalpha expression is an important step in the tumorigenic process. Clearly, enhanced ERalpha and PR expression would sensitise the premalignant epithelium to the proliferative effects of their cognate ligands, but it remains to be determined whether oestradiol and progesterone continue to drive proliferation by the indirect mechanisms that exist in the normal epithelium or whether an alternative, more direct, pathway has arisen during malignant transformation. Further studies on the mechanisms by which oestradiol and progesterone control the development of the human breast and breast tumours could lead to the identification of new targets for breast cancer prevention, to improved prediction of invasive breast cancer risk and to early detection of breast tumours. Abbreviations : DCIS = ductal carcinoma in situ; ER = oestrogen receptor; PR = progesterone receptor; TDLU = terminal ductal lobular unit. Backmatter: PMID- 12223125 TI - British Cancer Research Meeting, 30 June-3 July 2002, Glasgow AB - Abstract | The 2002 British Cancer Research Meeting was held from 30th June to 3rd July in Glasgow, UK. The meeting was structured to include educational workshops, plenary lectures, symposia, and poster sessions, which brought together scientists and clinicians. Presentations ranged from the impact that modifications to basic chromatin structure can have on diagnosis and targeted gene therapy, to the outcome of novel therapeutics through clinical trials. The emphasis was clear: patient survival is the main priority and treatment of organ-specific cancer must inevitably be replaced by individualised tumour-specific therapy. Keywords: chromatin remodelling, microarrays, targeted therapy, chromatin remodelling, microarrays, targeted therapy, Introduction : Rain and the city of Glasgow are almost synonymous and sadly this was indeed the case for the duration of the British Cancer Research Meeting 2002, held at the Scottish Exhibition and Conference Centre from 30th June to 3rd July. The rain did not dampen the spirits of the almost 600 delegates, however, and a feast of good science was enjoyed. This ranged from plenary sessions and award lectures to proffered paper sessions and posters, covering various aspects of clinically diagnosed and experimental cancer models. Here we present some of the highlights. Meeting report : A number of scientists, both young and established, were honoured at the meeting for their contributions to cancer research. The British Association for Cancer Research (BACR) Tom Connors Award Lecture "Close Encounters of a Molecular Kind" was given by Malcolm Stevens (University of Nottingham, UK). In his lecture, he outlined the need to nurture 'cunning' rather than 'stunning' chemistry as the way forward in drug discovery. The BACR Translational Research Award went to his colleague Tracey Bradshaw for her work on Phortress. In an eloquent presentation, the history of Phortress was outlined, underlining the interplay between chemistry and biology that led to its development and selection as a candidate therapy. Phortress evolved from polyhydroxylated 2-phenylbenzothiazoles designed as potential tyrosine kinase inhibitors. In laboratory studies it has shown very promising results in treating breast and ovarian cancers and it will shortly be undergoing Phase I clinical trials. Marie Boyd (University of Glasgow, UK) received the BACR Young Scientist Award and outlined the exciting prospect of linking conventional radiation-based therapies with gene therapy for treating malignancy. In "Re-designing Cancer Therapy", Sir David Lane (University of Dundee, UK) emphasised the need for patient-specific therapy, a consensus that was reiterated throughout the conference. He also discussed the necessity of looking not only at the gene in question, but also at the whole signalling pathway in which it is involved. We were reminded that 24% of human breast cancers have genetic changes in the p53 tumour suppressor gene and how p53 may be utilised as a target for gene therapy. The role of p53 as a potential therapeutic target was also emphasised by Barbara Foster (Pfizer Global R&D, Groton, CT, USA). Current strategies include the pharmacological rescue of mutant p53 protein, using small chaperone molecules. Both speakers discussed various therapeutic strategies for tumours that have mutations in the p53 gene and left us with the notion that the future of p53 as a therapeutic target was very much alive. The Cancer Research UK Lecture was given by Olli Kallioniemi (National Genome Research Institute, Bethesda, MD, USA), who discussed the application of array technology in cancer research. From the construction of simple RNA or DNA arrays, this field has evolved rapidly in recent years. The emerging challenge now is to uncover the function of the many candidate genes identified in earlier genome arrays by integrating gene copy number and gene expression analysis with functional studies. The development of tissue microarrays, in which it is theoretically possible to represent an entire pathology archive on a single microscope slide ('pathomics'), is one way forward for such high-throughput analysis. Tissue microarrays allow population-level screening and cellular resolution, which are not possible using genome arrays. A novel slant is the use of living gene transfection microarrays . With this technique, cultured cells are overlaid on DNA arrays. Those cells that take up DNA create spots of localised transfection. This exciting new technology has applicability for identifying new drug targets and in discovering gene products which affect cell physiology. Another innovative angle involves harnessing the power of gene transfection microarrays with RNA interference . RNA interference involves the processing of dsRNA to short interfering RNA of about 21 nucleotides, as a form of post-transcriptional gene silencing. This is often manifested in an aberrant phenotype, which can allow gene function to be identified. Quite literally we now have available an array of arrays and, bearing in mind how rapidly this field continues to evolve, the 'laboratory-on-a-chip' may soon be here. The importance of histone modification and chromatin remodelling is becoming increasingly recognised, particularly since many important genes are methylated and epigenetically silenced in tumours. In a whole session dedicated to this subject, the audience was educated on a range of topics, from basic principles of chromatin structure, to the role of histone modifying proteins and methylation in chromatin remodelling. Histone deacetylases and DNA methyl-transferase inhibitors could represent new diagnostic or therapeutic targets. Tony Kouzarides (Wellcome/Cancer Research UK Institute, Cambridge, UK) initiated this session in the BACR Frank Rose Lecture, giving an invigorating and informative insight into the role of "Histone Modifications in Transcriptional Control". Particular emphasis was given to the effect of gene methylation on transcriptional repression, and he gave an illustrative demonstration of the downstream effects of these events. Interestingly, he used the estrogen receptor (ER)-regulated gene pS2 as a model to demonstrate the complex sequence of events regulating gene transcription. The focus of the talk was on lysine and arginine methylation. Using chromatin immunoprecipitations, he showed that methylation of arginine 17 on the histone (H3) tail is required for activation of the pS2 gene. The order of events involves recruitment of histone-modifying proteins, acetylation, and subsequent methylation of the gene before its activation. The presentation was concluded with the possibility of using this detailed information on pS2 to produce better defined targets for therapy and molecular diagnosis. Individual tumours have different patterns of methylation, and Robert Brown (University of Glasgow, UK) also emphasised the potential of identifying gene methylation patterns in specific tumours. Nick La Thangue (University of Glasgow, UK) highlighted the use of histone deacetylase inhibitors as cancer therapeutics. One candidate, PXD101, is a novel histone-deacetylase antagonist that blocks tumour growth by inducing apoptosis and has little toxicity in mice. Advances in cancer therapeutics are continually ongoing, and novel strategies using DNA methyl-transferase inhibitors and histone-deacetylase inhibitors are already showing promise in clinical trials. New predictive markers for breast cancer are always being sought and work presented by Ramsey Cutress (University of Southampton, UK) showed that BAG-1, a multifunctional protein with targets including heat shock proteins and some nuclear hormone receptors, may have potential in predicting response to hormonal therapy. In a study of 138 breast cancers, nuclear expression of BAG-1 was associated with improved survival. BAG-1 interacts directly with ER-alpha and -beta. Brid Ryan (St Vincent's University Hospital, Dublin, Eire) showed that survivin, an inhibitor-of-apoptosis protein, may have potential in prognosis. Survivin is overexpressed in many tumours, including breast, is inversely correlated with expression of ER and progesterone receptor and is positively correlated with tumour grade. Current indicators of response to breast cancer therapy rely on the hormonal status of tumour tissue. Tumours responsive to endocrine hormones are treated with the ER antagonist tamoxifen. The therapeutic efficacy of tamoxifen is compromised by its agonistic activity in some tissue types. Mitch Dowsett (Royal Marsden NHS Trust, London, UK) highlighted that ER-negative (ER-)/progesterone receptor-positive patients also benefit from hormonal therapy and suggested that identification of progesterone receptor in the small subset of ER- tumours (25%) should also be ascertained prior to selecting adjuvant therapy. He described the reduced benefit tamoxifen might have on patients that have ER+ and HER-2+ tumours. Matti Aapro (Clinique de Genolier, Switzerland) addressed a number of consensus guidelines for adjuvant therapy, and both Aapro and Dowsett summarised the findings of the recent ATAC trial in which the aromatase inhibitor anastrozole was found to be superior to tamoxifen or a combination of both drugs in treating postmenopausal women with early breast cancer . Aapro proposed the use of adjuvant drugs such as Paclitaxel, in the early treatment of breast cancer, to improve patient survival. He concluded that one specific drug treatment will not prove successful for all patients and that organ-specific cancer treatment will soon be superseded by tumour-specific therapy. This idea was reinforced by Dowsett, who encouraged the idea that tumours should be seen as individual entities and should be treated specifically. Tumour metastases and drug resistance remains a burden on the current treatment of breast cancer. Both Graham Roche-Nagle (Beaumont Hospital, Dublin, Eire) and Thomas Barry (Mater Hospital, Dublin, Eire) presented data illustrating the role that cyclo-oxygenase-2 (COX-2) may play in these events. Using selective COX-2 inhibition Roche-Nagle demonstrated a significant reduction in the number of tumour metastases in an orthotopic murine model of breast cancer. Microvessel density and levels of serum vascular endothelial-cell growth factor were also reduced by COX-2 inhibition. Drug delivery and targeting are important aspects of cancer therapy that were covered in several presentations. The story of 5-(aziridin-l-yl)-2,4-dinitrobenzamide (CB1954) was presented by Richard Knox (Enact Pharma plc, Salisbury, UK). CB1954, developed in the 1950s, is a proven success in eradicating rat tumours, but unfortunately the same effect was never observed in human cancers. This is due to differences in the bioactivation of CB1954 in rats versus humans: the human form of the enzyme DT-diaphorase (NQ01) is much less efficient at metabolising CB1954, which accounts for its lack of activity against human cancers. New research has identified an enzyme (NQ02) in some human tumours, although not those of the breast. NQ02 is several-fold more effective than NQ01 at metabolising CB1954 and a Phase I trial is about to begin, with the aim of establishing the efficacy of NQ02 as a more active substrate for CB1954. Ruth Duncan (University of Cardiff, UK) presented an overview of tumour-specific drug delivery using polymer --anticancer conjugates. To date, successful receptor-mediated targeting has been limited to the asialoglycoprotein receptor in hepatic carcinomas, but HER2/neu, present in some breast cancers, could represent a new target for polymer-based drug delivery. Gerry Potter (De Montfort University, Leicester, UK) described some of the actions of the human cytochrome P450 CYP1B1, the overexpression of which has been observed in many tumours, including breast. Many cytochrome P450 enzymes are known to metabolise a variety of anticancer drugs; the consequence of cytochrome P450 metabolism is usually detoxification of the drug, although bioactivation occurs in some cases. With this in mind he proposed that CYP1B1 be reclassified as a tumour-suppressor enzyme, through natural prodrug bioactivation. He presented data showing that some natural dietary agents, such as phytoestrogens and resveratrol, are converted into anticancer agents by CYP1B1. Conclusion : As with most international conferences, the gathering of many researchers, scientists and clinicians results in the generation of novel ideas and extensive discussions. The outcome, foremost, drives delegates forward in their areas of cancer research. As the ultimate focus is on improving patient survival using chemopreventative and chemotherapeutic drugs, we are reminded of the demands of the patient. Perhaps the most significant message from this meeting is the necessity and inevitable development of a tumour-specific drug that is tailored to each patient's gene profile, thus meeting their individual requirements. This is keenly awaited. Perhaps the results of research in these areas, including interim results from the Phase I trials mentioned above, will be presented at the next meeting. The next British Cancer Research Meeting will take place at the Bournemouth International Conference Centre on 2 --5 July 2003 . Abbreviations : BACR = British Association for Cancer Research; ER = estrogen receptor. Backmatter: PMID- 12398775 TI - Hypertonic saline resuscitation in sepsis AB - Abstract | The review by Oliveira and colleagues on the subject of hypertonic saline resuscitation in sepsis (included in the present issue) suggests possible benefits for hypertonic saline. There is a firm experimental basis for the actions of hypertonic saline/hyperoncotic solutions in hemorrhagic hypotension, which include expansion of blood volume, improvement in cardiac index, favorable modulation of the immune system, and improvement in survival. These actions are presumed to be of benefit in the treatment of sepsis or septic shock. However, there are few experimental data regarding the use of these solutions, and clinical studies are descriptive. The major impact of early administration of hypertonic solutions may be attenuation of tissue injury, sepsis, and septic shock. Early and aggressive fluid resuscitation with hypertonic solutions to clinical end-points should be investigated in patients with systemic inflammatory response syndrome, sepsis, and septic shock. Keywords: hemorrhage, hyperoncotic, immune function, septic shock, shock, hemorrhage, hyperoncotic, immune function, septic shock, shock, Introduction : The present discussion focuses on the review by Oliveira and coworkers (presented in this issue), and on the use of hypertonic saline resuscitation in sepsis. Early fluid resuscitation of patients with systemic inflammatory response syndrome reduces the incidence of mortality due to septic shock and sepsis . These favorable results have led to consideration of other fluid modalities. In their review, Oliveira and coworkers consider the use of hypertonic saline solutions in the treatment of patients with severe sepsis. Hypertonic saline has been extensively investigated in animal models with regard to its efficacy in treating hemorrhagic hypotension . Numerous clinical studies of patients with traumatic injuries have been initiated with favorable, but not definitive, results . Limited animal studies have been directed at treatment of septic shock using hypertonic solutions, and clinical studies have been exploratory at best. Hannemann and coworkers studied 21 stable patients with septic shock. Patients were administered 2 --4 ml/kg hypertonic saline in hydroxyethyl starch. Following administration there was a small effect on oxygenation by increasing cardiac output and oxygen delivery. There was no control group in this study for comparison. An abstract reported by Oliveira an colleagues described a randomized study of 25 stable patients with sepsis. Administration of 250 ml hypertonic saline in dextran significantly improved cardiac index and pulmonary artery occlusion pressure as compared with administration of an equivalent volume of normal saline. Although favorable, those studies of responses to hypertonic saline solutions in stable patients with sepsis are only descriptive. Furthermore, administration of hypertonic saline solutions during the early stages of septic shock when the patient is not yet hyperdynamic has not been investigated. The major benefit of hypertonic solutions is the rapid expansion of blood volume for a small volume administered . This advantage is sustained by the addition of colloids such as dextran. In their review, Oliveira and colleagues suggest that the major benefit of resuscitation with hypertonic solutions in models of septic shock is the restitution of blood volume and accompanying improvements in cardiovascular and hemodynamic function. Hypertonic solutions have been demonstrated to modulate the immune system . As pointed out by Oliveira and coworkers , hypertonic saline has a pronounced anti-inflammatory effect that could reduce the response to sepsis and attenuate later multiple organ failure. Recently, Shi and coworkers found resuscitation with hypertonic saline to ameliorate the gut and lung injury seen following use of Ringer's lactate. This lack of injury reduced bacterial translocation, and would presumably reduce the incidence of sepsis. Comparisons between various resuscitation solutions on outcome have been controversial. However, the work of Rhee and colleagues in comparing various solutions suggests pronounced differences in immune modulation. The question that must be addressed is whether the presently available solutions are harmful in conditions in which immune function is challenged, and are thus inappropriate to use. Rhee and coworkers showed that hypertonic solutions, especially when combined with dextran, are superior to 'standard of care' solutions. This leads to the conclusion that presently used solutions may be harmful because they compromise immune function and that hypertonic solutions are of benefit because they do not. In their review, Oliveira and coworkers discuss the use of hypertonic solutions for treatment of septic shock; however, they do not refer to the possible prophylactic benefit of early use of these solutions. The etiology of sepsis is based on a sustained period of hypoperfusion of vital organs . Rivers and colleagues demonstrated early resuscitation to improve survival. The use of hypertonic solutions as the initial fluid therapy, irrespective of the cause of hypotension or cardiac insufficiency, may be of benefit in blocking the cascade that leads to septic shock. In a study of trauma patients initially resuscitated with hypertonic saline dextran , no cases of sepsis occurred in the hypertonic saline group as compared with an incidence of 2% in the standard of care group. Although interesting, this difference was not significant. The favorable preservation of tissues and immune function that occurs following early resuscitation of the hypotensive patient with hypertonic solutions bodes well for a favorable outcome in the treatment of sepsis. Oliveira and colleagues point out the probable advantages of using hypertonic solutions in patients with sepsis. There is a rapid and sustained expansion of plasma volume, leading to improvement in cardiovascular function, rectification of microcirculatory blood flow, and favorable modulation of immune function. All of these factors may contribute to improved outcomes in patients with sepsis and septic shock. Beyond the scope of the present commentary is the possibility that early resuscitation with hypertonic solutions could contribute to a reduced incidence of sepsis by modulation of immune responses and reduced tissue injury. These improvements following hypertonic saline resuscitation would have a greater impact on the subsequent course of patients admitted to the intensive care unit. At present, although the experimental data are positive and the rationale for use of hypertonic solutions in the care of patients with sepsis or septic shock is reasonable, definitive clinical studies as to the safety and efficacy in this patient population are still required. Competing interests : CEW is a faculty member of the University of California at Davis, which holds the US patent for hypertonic saline/hyperoncotic solutions. He holds advisory positions in companies developing these solutions and serves as a consultant to other companies. Backmatter: PMID- 12453307 TI - Calmodulin binding to recombinant myosin-1c and myosin-1c IQ peptides AB - Abstract | Background | Bullfrog myosin-1c contains three previously recognized calmodulin-binding IQ domains (IQ1, IQ2, and IQ3) in its neck region; we identified a fourth IQ domain (IQ4), located immediately adjacent to IQ3. How calmodulin binds to these IQ domains is the subject of this report. Results | In the presence of EGTA, calmodulin bound to synthetic peptides corresponding to IQ1, IQ2, and IQ3 with Kd values of 2 --4 muM at normal ionic strength; the interaction with an IQ4 peptide was much weaker. Ca2+ substantially weakened the calmodulin-peptide affinity for all of the IQ peptides except IQ3. To reveal how calmodulin bound to the linearly arranged IQ domains of the myosin-1c neck, we used hydrodynamic measurements to determine the stoichiometry of complexes of calmodulin and myosin-1c. Purified myosin-1c and T701-Myo1c (a myosin-1c fragment with all four IQ domains and the C-terminal tail) each bound 2 --3 calmodulin molecules. At a physiologically relevant temperature (25C) and under low-Ca2+ conditions, T701-Myo1c bound two calmodulins in the absence and three calmodulins in the presence of 5 muM free calmodulin. Ca2+ dissociated nearly all calmodulins from T701-Myo1c at 25C; one calmodulin was retained if 5 muM free calmodulin was present. Conclusions | We inferred from these data that at 25C and normal cellular concentrations of calmodulin, calmodulin is bound to IQ1, IQ2, and IQ3 of myosin-1c when Ca2+ is low. The calmodulin bound to one of these IQ domains, probably IQ2, is only weakly associated. Upon Ca2+ elevation, all calmodulin except that bound to IQ3 should dissociate. Keywords: Background : Myosin-1c (Myo1c), the myosin previously called myosin-Ibeta, myr 2, or MI-110K , is an unconventional myosin isozyme implicated in nuclear transcription , lamellopodia dynamics of motile cells , brush-border dynamics of proximal-tubule cells of the kidney , and adaptation of mechanoelectrical transduction in hair cells, the sensory cells of the inner ear . Myo1c belongs to the myosin-I class, which contains eight members in humans and mice ; the bullfrog genome possesses at least two members . Members of the myosin-I class have a single globular motor domain, followed by a neck region and a relatively short (30 --40 kD) tail domain . This latter domain is highly basic and binds to acidic phospholipids . Like all biochemically characterized unconventional myosins, Myo1c binds calmodulin light chains in its neck region ; this region also interacts with non-calmodulin receptors in hair cells . Figure 1 | Myo1c constructs and IQ-domain sequences Myo1c constructs and IQ-domain sequences A, domain structure of Myo1c, indicating amino acids encompassed by full-length Myo1c and T701-Myo1c constructs. B, SDS-PAGE of 15 pmol each of purified full-length Myo1c and T701-Myo1c constructs co-expressed with calmodulin, and calmodulin alone (CaM; 45 pmol) on a 18% acrylamide gel. Molecular mass markers indicated on left. C, aligned bullfrog Myo1c IQ domains. The IQ motif positions are in bold and the repeat present in IQ3 and IQ4 is underlined. Numbers indicate amino-acid residue positions. Unconventional myosins contain from one to several IQ domains, which are calmodulin-binding motifs that adhere to the general consensus sequence IQX3RGX3R . Calmodulin, which can bind up to four Ca2+ ions, generally binds IQ domains in the Ca2+-free conformation; interaction of Ca2+-bound calmodulin to other proteins occurs through alternative binding motifs . Myo1c contains three readily recognized IQ motifs of 23 amino acids each (Fig. ; refs. [,-]). Purified Myo1c apparently includes 2 --3 calmodulins per Myo1c heavy chain ; calmodulin supplementation can increase the stoichiometry to as many as 4 calmodulins per Myo1c . Unfortunately, the lack of appropriate quantitation standards for the Myo1c heavy chain in those experiments limits the reliability of these values. How Ca2+ and calmodulin regulate Myo1c or indeed any myosin-I is unclear. Although Ca2+ increases ATPase activity of most myosin-I isozymes, in vitro motility is usually blocked under identical conditions . Ca2+ dissociates one or more calmodulins from the myosin-calmodulin complex, which apparently elevates ATPase activity and inhibits motility . In conventional myosin, light chains related to calmodulin appear to be essential for stabilization of the myosin lever arm , a domain that is vital for efficient conversion of chemical energy into mechanical work . Calmodulin probably plays a similar lever-arm stabilizing role for Myo1c; Ca2+-induced calmodulin release would reverse the stabilization and inhibit motility. To better understand the regulation of Myo1c activity by calmodulin, we sought to more accurately determine how calmodulin binds to Myo1c by measuring the Ca2+-dependence of calmodulin binding to individual Myo1c IQ peptides. In addition, to examine the consequences of calmodulin binding to adjacent IQ domains, we measured hydrodynamic properties of recombinant Myo1c-calmodulin complexes, under differing conditions of Ca2+, calmodulin, and temperature. These measurements allowed us to determine the molecular mass and hence stoichiometry of the Myo1c complex. Our results indicate that IQ1, IQ2, and IQ3 have calmodulin bound when the concentration of Ca2+ is low, and that increased Ca2+ induces release of calmodulin from IQ1 and IQ2. Results : Sequence analysis of IQ domains | Examination of the primary sequence of the bullfrog Myo1c neck region reveals an exact repeat of five amino acids located both in the IQ3 region (YRNQP; residues 761 --765) and at residues 786 --790. Alignment of the residues surrounding the repeat revealed reasonable homology with the three known IQ domains, with particular similarity to IQ3, suggesting that this region may be a fourth IQ domain . Although the pair of amino acids (LM; residues 782 and 783) that align with the RG of the IQ consensus motif are not conserved, the first pair of amino acids (IR; residues 777 and 778) that align with the consensus IQ adhere to the consensus better than those of IQ3. Because of the sequence similarity to IQ3 and because this peptide binds calmodulin (albeit weakly; see below), we refer to this domain as IQ4. IQ -- Alexa-calmodulin interaction on plastic plates | To investigate the calmodulin-binding properties of each Myo1c IQ domain, we measured interaction of a fluorescently labeled calmodulin (Alexa-calmodulin) with individual Myo1c IQ peptides that had been conjugated to wells of a plastic plate. We used an IQ peptide from neuromodulin as a positive control; calmodulin binds to this site, with its interaction reduced by high ionic strength . As a negative control, we used a peptide (PVP), corresponding to the 25 amino acids of Myo1c immediately following IQ4. Indeed, Alexa-calmodulin bound to wells derivatized with the neuromodulin IQ peptide and did not bind to the PVP peptide . Figure 2 | Alexa-calmodulin binding to Myo1c IQ peptides on plates Alexa-calmodulin binding to Myo1c IQ peptides on plates A, Alexa-calmodulin (50 nM) was incubated in a 96-well plate covalently bound with Myo1c IQ peptides, a positive-control IQ peptide (NM), or a negative-control peptide (PVP); binding was carried out at room temperature in the presence of low (150 mM) or high (400 mM) KCl and in the presence of 100 muM EGTA. We measured unbound Alexa-calmodulin concentration by fluorescence, then inferred the amount bound. Averaged data from three independent experiments (error bars indicate standard error). B, competition with free IQ peptides. Following incubation of 50 nM Alexa-calmodulin and free IQ peptides in the presence of 100 muM EGTA and 150 mM KCl in wells of an IQ3-derivatized plastic plate, unbound Alexa-calmodulin was measured; error bars indicate standard deviation with n = 6 (single experiment). To correct for quenching of Alexa-calmodulin fluorescence by IQ peptides, each sample was standardized against an identical reaction in an underivatized plate. Data were fit with equation (4) with the Hill coefficient set at 2. Substantial amounts of Alexa-calmodulin bound to wells derivatized with IQ1, IQ2, and IQ3; by contrast, relatively little bound to IQ4-coated wells under these conditions . As has been noted for the neuromodulin IQ domain , increasing the KCl concentration reduced binding to each IQ peptide. Although the data shown in Fig. were obtained at room temperature, we saw a similar rank order of binding -- albeit with lower total Alexa-calmodulin bound -- at 4C (data not shown). To confirm the approximate binding strength reported by this assay, we used free IQ peptides to prevent Alexa-calmodulin binding to an IQ3-derivatized plate. Because the IQ peptides strongly quenched Alexa-calmodulin fluorescence when bound, we corrected fluorescence measurements using an identical assay in an underivatized plate. Although this quenching correction introduced substantial scatter into the data, we found that the apparent affinities for binding of peptides to Alexa-calmodulin followed the order IQ3 > IQ1 ~ IQ2 > IQ4 . IQ -- calmodulin interaction by quenching of Alexa-calmodulin fluorescence | As noted above, Alexa-calmodulin fluorescence was quenched upon binding to an IQ peptide . Because an excess of unlabeled calmodulin was able to reverse 70 --95% of the quench , we inferred that most Alexa-calmodulin bound to the same site as unlabeled calmodulin. We used this fluorescence-intensity quench empirically to measure the affinity of each IQ peptide for Alexa-calmodulin . In some experiments, Ca2+ was held at <30 nM by chelation with 100 muM EGTA; in other experiments, we added 25 muM exogenous CaCl2 in the absence of EGTA. These two concentrations mimic the low- and high-Ca2+ conditions that Myo1c may encounter in hair cells when the transduction channel is closed or open . In the presence of 100 muM EGTA, Kd values followed the order IQ3 < IQ1 ~ IQ2 << IQ4. Although the data were fit somewhat better with a modified Hill equation that with a standard bimolecular-binding isotherm , the physiological significance of Hill coefficients >1 is uncertain, particularly given the 1:1 peptide:calmodulin stoichiometry (see below). Ca2+ had only modest effects on the affinity of the Myo1c IQ peptides for Alexa-calmodulin . Figure 3 | Alexa-calmodulin binding to Myo1c IQ peptides in solution Alexa-calmodulin binding to Myo1c IQ peptides in solution Binding of IQ peptides to Alexa-calmodulin (100 nM) at room temperature with 150 mM KCl and 100 muM EGTA. A, IQ peptides quench Alexa-calmodulin fluorescence; unlabeled calmodulin reverses the quench. IQ peptides (25 muM) and unlabeled calmodulin (75 muM) were added as indicated. Control, Alexa-calmodulin alone; +IQ, Alexa-calmodulin plus IQ peptide; +IQ, +CaM, IQ peptide plus unlabeled calmodulin; IQ only, IQ peptides with no Alexa-calmodulin. B, Alexa-calmodulin (100 nM) was incubated with increasing concentrations of soluble IQ peptides. Total IQ peptide concentration added is plotted versus Alexa-calmodulin fluorescence, reported in arbitrary fluorescence units; error bars indicate standard deviation with n = 4. Data were fit with equation (3) (dashed lines) and equation (4) (solid lines). Table 1 | Interaction of Myo1c IQ peptides with Alexa-calmodulin Despite only minor effects on binding affinity, Ca2+ did influence the calmodulin-peptide complex, as signaled by changes in Alexa-calmodulin fluorescence. Changes in fluorescence intensity during manipulation of a single parameter, like Ca2+ concentration, should report conformational changes in Alexa-calmodulin. For example, the fluorescence intensity of free Alexa-calmodulin in solution was ~15% lower in 25 muM CaCl2 than in 100 muM EGTA (left-hand limits in Fig. ). Because the dye moiety itself is not Ca2+ sensitive , the Ca2+-dependent fluorescent change reflects changes in the dye's surrounding environment, probably signaling the compact-to-open structural change seen when Ca2+ binds to calmodulin . In contrast to the reduction of free Alexa-calmodulin fluorescence by Ca2+, fluorescence of Alexa-calmodulin when saturated by IQ peptides was 1.5- to 2-fold greater in 25 muM CaCl2 than in 100 muM EGTA (Fig. ; Table ). Thus, when Alexa-calmodulin was bound to IQ peptides, Ca2+ induced a conformational change that was substantially different from that seen in the peptide-free state. Figure 4 | Ca2+-dependence of IQ-Alexa-calmodulin interaction Ca2+-dependence of IQ-Alexa-calmodulin interaction Binding of IQ peptides to Alexa-calmodulin (50 nM) in 100 muM EGTA or 25 muM CaCl2 as described for Fig. ; error bars indicate standard deviation with n = 6. Data were fit with a modified Hill equation (equation (4)). IQ -- calmodulin interaction under stoichiometric-titration conditions | To determine the affinities of the Myo1c IQ peptides for unlabeled calmodulin, we used Alexa-calmodulin as a reporter (Alexa-calmodulin : unlabeled calmodulin ratio of 1:100) in our binding studies. This approach assumes that Alexa-calmodulin is functionally equivalent to unlabeled calmodulin. We determined affinities by fitting the IQ-peptide concentration vs. fluorescence quench data with an appropriate model. If the IQ peptides bound only Alexa-calmodulin and not unlabeled calmodulin, the Kd and FIQ/F values of Table would have described the fit to the concentration-quench plots. The line derived from these values did not fit the data , indicating that, as expected, unlabeled calmodulin binds to the Myo1c IQ peptides. Figure 5 | Binding of unlabeled calmodulin to Myo1c IQ peptides Binding of unlabeled calmodulin to Myo1c IQ peptides Fluorescence intensity during titration of a mixture of 50 muM unlabeled calmodulin and 500 nM Alexa calmodulin (used as a reporter) with IQ peptides (total concentration). Thin solid lines: fit with equation (7) assuming m = 1. Dashed lines: fit with equation (7) assuming m = 2 and final fluorescence from m = 1 fit. Dotted lines: results predicted from Table if unlabeled calmodulin does not affect the Alexa-calmodulin -- IQ peptide equilibrium. Thick solid lines (IQ4 only): fit with equation (4). Error bars indicate standard deviation with n = 4. Note difference in the abscissa scale for IQ4. These experimental conditions resembled a stoichiometric titration, where the total concentration of calmodulin was higher than the Kd values for IQ1, IQ2, and IQ3. Under true stoichiometric-titration conditions (fixed concentration of receptor at 100-fold or more than the Kd, varying the ligand concentration up to and beyond the receptor concentration), almost all of the added IQ peptide would bind tightly to calmodulin and linearly decrease the fluorescence; at the point where the IQ-peptide concentration exceeds the calmodulin concentration multiplied by the peptide:calmodulin stoichiometry (m), a plateau in the fluorescence intensity would be reached. Because the relatively weak affinities observed here make such true stoichiometric titration impractical, we used an intermediate concentration of calmodulin (50 muM, ~10-fold larger than Kd) and used equation (7) to describe the equilibrium precisely. This approach allowed us to determine both m and Kd in the same experiment. In the presence of EGTA, the IQ1, IQ2, and IQ3 binding data were much better fit by m = 1 than they were to m = 2, indicating that the binding stoichiometry of peptide to calmodulin was 1:1 . The Kd values determined with equation (7) were very similar to those determined for binding to Alexa-calmodulin alone (Tables and ), confirming that under these conditions, the IQ peptides bind to unlabeled calmodulin and Alexa-calmodulin similarly. Table 2 | Interaction of Myo1c IQ peptides with unlabeled calmodulin In the presence of Ca2+, IQ3 also bound to calmodulin with a stoichiometry of 1:1 . The fits to m = 1 and 2 were equally good for IQ1 and IQ2 in the presence of Ca2+, signifying the inability for this analysis to determine precise binding stoichiometry of IQ1 and IQ2 under these conditions. In addition, these data indicate that the apparent affinities of IQ1 and IQ2 for unlabeled calmodulin were substantially weakened by Ca2+ , unlike results with Alexa-calmodulin alone . Because the assumption that affinities of Alexa-calmodulin and unlabeled calmodulin for IQ peptides are identical was violated for IQ1 and IQ2, the actual affinities of these IQ peptides for unlabeled calmodulin may be even weaker than those reported in Table . By contrast, Ca2+ had only a very modest effect on IQ3 affinity for unlabeled calmodulin. Binding of IQ4 to unlabeled calmodulin was distinct from that of the other IQ peptides. The data with IQ4 were best fit with a Hill equation (equation 4), with a Hill coefficient of greater than 2 (Fig. , thick solid lines), suggesting that binding of two peptides per calmodulin may be required for the fluorescence change. The apparent affinities (~100 muM) were similar to the concentration of calmodulin (50 muM), however, indicating that the apparent affinities did not accurately reflect Kd values. These results with a mixture of unlabeled and Alexa-calmodulin were different from those with Alexa-calmodulin alone, where IQ4 Hill coefficients were close to 1 (data not shown). Nevertheless, these data show that unlabeled calmodulin can bind to IQ4, albeit with weak affinity and uncertain stoichiometry. Hydrodynamic analysis of full-length Myo1c | To determine the stoichiometry and Ca2+-dependent regulation of calmodulin binding to Myo1c with all four IQ motifs, we co-expressed calmodulin and full-length bullfrog Myo1c in insect cells using baculoviruses and subjected the purified Myo1c-calmodulin complexes to hydrodynamic analysis . We carried out velocity sedimentation of Myo1c-calmodulin complexes on 5 --20% sucrose gradients to determine sedimentation coefficients. We measured the Stokes radius of Myo1c-calmodulin complexes using gel filtration on Superdex 200 under temperature and buffer conditions identical to those of the velocity-sedimentation experiments . Although most experiments used 400 mM KCl (which prevented adsorption to the gel-filtration matrix), we obtained identical sedimentation coefficients in the presence of 150 or 250 mM KCl (not shown). Velocity-sedimentation and gel-filtration experiments were carried out at 4C, the temperature used for Myo1c purification, as well as at 25C, a physiologically relevant temperature for a bullfrog. Table 3 | Hydrodynamic analysis of full-length Myo1c To calculate the molecular mass of Myo1c-calmodulin complexes, we applied the modified Svedberg equation, which relates mass to the diffusion constant (calculated here from Stokes radius) and the sedimentation coefficient . The partial specific volume of each protein complex was determined using the amino-acid composition of the constituent proteins (Table ; ref. ). Although the uncertainty in calmodulin stoichiometry leads to ambiguity in this calculation, the calculated partial specific volumes were so close (e.g., 0.734 for one and 0.731 for three calmodulins per Myo1c complex) that the precise value did not significantly affect the final molecular-mass value. Full-length Myo1c bound ~3 calmodulins per Myo1c at 4C in the presence of EGTA or CaCl2 . One of the bound calmodulins was only weakly associated, as elevation of the temperature to 25C induced the release of 1 mole of calmodulin in the presence of EGTA. When Ca2+ was elevated to 25 muM at 25C, however, we could not detect substantial full-length Myo1c in solution after sucrose-gradient centrifugation or gel filtration, suggesting that the protein had aggregated. Hydrodynamic analysis of T701-Myo1c | Because the size of full-length Myo1c (125 kD, including purification and detection tags) is much larger than calmodulin (16.7 kD), we improved our ability to determine stoichiometry from molecular mass by examining a smaller (45 kD) neck-tail recombinant fragment of Myo1c. This construct, T701-Myo1c, contained amino acids 701 --1028 of bullfrog Myo1c, including all four IQ domains, the entire C-terminal tail, and N-terminal purification and epitope tags . T701-Myo1c bound 2.5 moles of calmodulin per mole of heavy chain at 4C in the presence of 100 muM EGTA (Fig. ; Table ). As with full-length Myo1c, elevation of the analysis temperature to 25C induced the release of ~0.7 mole of calmodulin. In contrast to the results seen with full-length Myo1c, elevation of the CaCl2 concentration to 25 muM at 4C also induced the release of ~0.7 mole of calmodulin. The amount of T701-calmodulin complex recovered on sucrose gradients or by gel filtration decreased substantially when the CaCl2 concentration was elevated to 25 muM at 25C, signaling the formation of aggregates, as seen with the full-length complex. Furthermore, the calculated calmodulin stoichiometry of the observed T701-calmodulin complex under these conditions was only ~0.3 mole of calmodulin per mole of Myo1c, reinforcing the suggestion that Ca2+ induced the dissociation of most calmodulins at 25C and that this loss of light chains resulted in aggregation. Figure 6 | Hydrodynamic analysis of T701-Myo1c Hydrodynamic analysis of T701-Myo1c A, measurement of Stokes radius using gel filtration. Globular standards were used to calibrate the column, and retention times were converted to Stokes radius (inset). T701-Myo1c was separated at 25C in the presence of 100 muM EGTA (Stokes radius of 4.8 nm) or 25 muM CaCl2 (3.9 nm). B-D, measurement of sedimentation coefficients using sucrose-gradient centrifugation. Size standards were used to calibrate the gradient, and fraction numbers were converted to sedimentation coefficient (inset, panel B). T701-Myo1c concentration was determined by ELISA. B, T701-Myo1c sedimentation at 4C in the presence of 100 muM EGTA (3.9 S) or 25 muM CaCl2 (3.5 S). C, T701-Myo1c sedimentation in the presence of EGTA at 25C, with (4.6 S) or without (3.8 S) the continuous presence of 5 muM calmodulin (+CaM). D, T701-Myo1c sedimentation in the presence of 25 muM CaCl2 at 25C, with (3.9 S) or without (3.1 S) the continuous presence of 5 muM calmodulin. Table 4 | Hydrodynamic analysis of T701-Myo1c. We could not prevent the release of calmodulin at 25C by saturating T701-Myo1c with excess calmodulin immediately prior to centrifugation (preloading). In EGTA, the sedimentation coefficient of calmodulin-preloaded T701-Myo1c measured at 25C (3.85 +- 0.07 S; n = 2) was nearly identical to that measured without preloading (3.83 S; Table ). Likewise, the sedimentation coefficient of calmodulin-preloaded T701-Myo1c measured at 25C and in 25 muM CaCl2 (2.80 +- 0.42 S; n = 2) was similar to that measured without preloading (3.13 S; Table ). By contrast, we could prevent the temperature-dependent loss of calmodulin by carrying out sedimentation in the continuous presence of 5 muM calmodulin (Fig. ; Table ). Although gel-filtration analysis was impractical with this high calmodulin concentration, we assumed that the Stokes radius of T701-Myo1c in the presence of calmodulin was identical to the value obtained in the absence. Sedimentation at 25C in EGTA gradients supplemented with 5 muM calmodulin resulted in the retention of ~3 calmodulins per T701-Myo1c. In 25 muM CaCl2, supplementation with 5 muM calmodulin resulted in ~1 calmodulin bound per T701-Myo1c. In addition, protein loss due to aggregation was minimal under these conditions. Table 5 | Hydrodynamic analysis of T701-Myo1c in the presence of 5 muM calmodulin Myo1c-calmodulin stoichiometry by gel scanning | To measure Myo1c-calmodulin stoichiometry by an independent method, we separated calmodulin standards and T701-Myo1c by SDS-PAGE . Using densitometry, we quantified the staining intensity of the calmodulin standards to generate a standard curve and determined the amount of calmodulin present in each T701-Myo1c sample. Applying the analysis described in Experimental Procedures and equation (13), we found an average of 2.6 +- 0.2 calmodulins per T701-Myo1c (mean +- standard error) in six experiments, three separate preparations analyzed in duplicate. This value was very close to the value of 2.5 +- 0.1 bound calmodulins determined independently by hydrodynamic analysis . Figure 7 | Myo1c-calmodulin stoichiometry by gel scanning Myo1c-calmodulin stoichiometry by gel scanning A, Coomassie-stained SDS-PAGE gel with two independent T701-Myo1c preparations (left lanes) and dilutions of purified bovine-brain calmodulin (CaM). B, calmodulin standard curve derived by gel scanning. Discussion : Calmodulin interaction with individual Myo1c IQ domains | To examine how calmodulin binds to the Myo1c IQ sites, we developed two binding assays using a commercially available fluorescent calmodulin and individual IQ peptides. In one assay, we covalently attached peptides to plastic plates, then measured the amount of fluorescent calmodulin that remained unbound after incubation with the peptide-derivatized plate. This assay was simple and fast, and allowed us to measure binding under a wide variety of conditions. In our second assay, we exploited the empirical observation that the Alexa-calmodulin fluorescence intensity is quenched by binding of IQ peptides. As in other assays with fluorescently labeled calmodulins (e.g., ref. ), binding of the peptides to Alexa-calmodulin did not perfectly mimic binding to unlabeled calmodulin. For example, Alexa-calmodulin bound IQ peptides more strongly in the presence of Ca2+ than did unlabeled calmodulin. Moreover, excess unlabeled calmodulin could not fully reverse the quenching of Alexa-calmodulin fluorescence induced by IQ peptides, suggesting that IQ peptides could bind to Alexa-calmodulin at two sites, including one where unlabeled calmodulin could not bind. Indeed, binding of IQ peptides to both sites on a single fluorescent calmodulin could account for Hill coefficients of >1 seen in some experiments (e.g., Figs. and ). Nevertheless, these discrepancies should not prevent use of Alexa-calmodulin for measuring interaction with calmodulin's targets, particularly if the interaction with unlabeled calmodulin is compared to the interaction with Alexa-calmodulin. Calmodulin bound to peptides corresponding to each of the four Myo1c IQ domains, although with differing affinity and Ca2+ sensitivity. Affinities for calmodulin binding to IQ1 and IQ2 were relatively modest (Kd values of ~5 muM). As with other IQ domains , Ca2+ weakened the affinity of calmodulin for IQ1 and IQ2 by more than 10-fold. By contrast, calmodulin binding to IQ3 was slightly stronger and was affected much less by Ca2+. Because calmodulin binds strongly to classic IQ domains only in the absence of Ca2+ , its strong binding to IQ3 in the presence of Ca2+ suggests the participation of an additional Ca2+-requiring binding motif. Two common calmodulin-binding motifs, called 1-8-14 and 1-5-10 for the pattern of hydrophobic amino-acid residues, require Ca2+ for calmodulin binding . IQ3 has two nearly perfect 1-5-10 domains that are at +2 net charge instead of the minimum +3 in the consensus . In addition, IQ3 has a 1-8-14 motif with a proline residue at position 14 instead of phenylalanine, isoleucine, leucine, valine, or tryptophan. Because most proteins that bind calmodulin through the 1-8-14 and 1-5-10 motifs do so strongly, the relatively modest affinity of IQ3 for calmodulin in the presence of Ca2+ suggests that calmodulin binds through one of these imperfect motifs located within this IQ domain. To interact with an alternate set of residues, Ca2+-calmodulin must adopt a new conformation. A similar Ca2+-dependent rearrangement was predicted for the complex of calmodulin and the first IQ domain of myosin-1a (brush-border myosin I) . In support of this view, we observed evidence for Ca2+-dependent conformational changes in calmodulin while bound to IQ peptides. When Alexa-calmodulin was bound to Myo1c IQ peptides, its fluorescence was higher in the presence of Ca2+ than in its absence, suggesting that that Ca2+-bound Alexa-calmodulin binds to the IQ peptides in a different conformation than does Ca2+-free Alexa-calmodulin. For example, in the absence of Ca2+, Alexa-calmodulin may bind to IQ peptides in a more compact conformation, quenching fluorescence by burying dye moieties in a less polar environment. Although the Ca2+-induced conformational change could be a property of Alexa-calmodulin rather than calmodulin itself, the Ca2+-dependent changes in affinity of calmodulin for IQ1 and IQ2 and calmodulin's likely shift to a new binding site on IQ3 suggests that the conformational change is probably also a property of authentic calmodulin. Calmodulin also bound to a newly identified domain, IQ4. Because the affinity of calmodulin for IQ4 is very weak, calmodulin should only occupy IQ4 in subcellular locations with a low Ca2+ concentration and a high level of free calmodulin. For example, a small population of Myo1c molecules with calmodulin bound to IQ4 should be present in the stereocilia of inner-ear hair cells, which contain ~35 muM free calmodulin . Although most tissues contain less free calmodulin , concentrations in other individual organelles can reach the millimolar range . On the other hand, the weak affinity of this IQ domain for calmodulin suggests that IQ4 may play another role, such as interacting with another protein. Calmodulin interaction with Myo1c | The binding affinities of calmodulin for the individual IQ peptides do not reflect exactly the affinities of calmodulin for the IQ domains within Myo1c. For example, despite micromolar Kd values for calmodulin-IQ peptide interactions, calmodulin remains bound to Myo1c during long gel-filtration or centrifugation experiments, even at nanomolar Myo1c concentrations . This result suggests that calmodulin binds to some of Myo1c's four tandem IQ domains substantially more strongly than to the individual peptides. For example, other regions of Myo1c could constrain the IQ domains in conformations that are substantially more (or less) favorable for calmodulin binding than the population of conformations adopted by a soluble IQ peptide. Moreover, calmodulin binding to Myo1c could be influenced by interactions with adjacent calmodulin molecules or to the Myo1c head or tail domains. To examine calmodulin binding to IQ domains in Myo1c, we determined the molecular mass (and hence calmodulin:Myo1c stoichiometry) and shape of Myo1c under the appropriate conditions of temperature and Ca2+. Although analytical ultracentrifugation is more commonly used to measure molecular size of protein-protein complexes , we instead used classic hydrodynamic methods of velocity sedimentation on sucrose gradients to obtain sedimentation coefficients and gel filtration to obtain Stokes' radius. One advantage of this approach was that by detecting Myo1c using a sensitive ELISA method, we were able to use very low concentrations of Myo1c. Furthermore, we were able to carry out sedimentation in the presence of a high concentration of calmodulin, a manipulation that prevents Myo1c detection in a standard analytical ultracentrifugation experiment. A disadvantage of this approach was the need for high concentrations of sucrose, which in rare conditions can substantially affect the hydrodynamic properties of a protein ; nevertheless, changes in Myo1c size were observed both in velocity sedimentation (in the presence of sucrose) and in gel filtration (in its absence). Another disadvantage of our classic approach to molecular-mass determination was that the gel filtration and velocity sedimentations were done on different time scales (~1 hour vs. 15 --18 hours). If calmodulin slowly dissociated during the analysis (which in both assays diluted Myo1c well below 1 muM), the degree of dissociation would be larger in the velocity sedimentation experiments than in the gel filtration experiments. Nevertheless, our approach was validated by the demonstration that the number of calmodulins per T701-Myo1c was identical in hydrodynamic and gel-scanning experiments, at least in EGTA at 4C. Because T701-Myo1c mimicked properties of the full-length protein (except under low-temperature, high-Ca2+ conditions), we exploited the neck-tail construct for a more detailed analysis of calmodulin binding. As expected from the large Ca2+-dependent weakening of calmodulin affinity for IQ1 and IQ2 (Fig. ; Table ), Ca2+ decreased the number of calmodulins bound to T701-Myo1c at high ionic strength. When Ca2+ was low at 25C, each T701-Myo1c had about two bound calmodulins, with a third bound if the calmodulin concentration reached 5 muM. At this calmodulin concentration, IQ domains 1, 2, and 3 are likely occupied by calmodulin. When Ca2+ is high at 25C, all but one calmodulin dissociated from T701-Myo1c in the presence of 5 muM free calmodulin. The strong affinity of IQ3 for Ca2+-calmodulin suggests that the remaining calmodulin was bound to this IQ domain. How many calmodulins are bound to Myo1c in the cell at increased Ca2+ concentrations? The elevated ionic strength used for the hydrodynamic analysis probably weakened the affinity of the calmodulin for IQ3 , requiring 5 muM free calmodulin to maintain occupancy of that site. We therefore infer that at a physiologically significant temperature and at a cellular ionic strength, Ca2+ triggers release of calmodulins from IQ1 and IQ2 from T701-Myo1c, leaving only IQ3 occupied. Although these results contrast with those reported for mammalian Myo1c, where only one of three calmodulins is released by Ca2+ , our T701 construct lacks Myo1c's motor domain. It is entirely plausible that even in the presence of Ca2+, calmodulin remains bound to IQ1, albeit in a different conformation and dependent on interactions with the myosin head. Our results therefore suggest that Ca2+ either induces the release of calmodulin from IQ1 or causes it to change its interaction with Myo1c substantially. Of the three calmodulins bound to Myo1c, one of these binds relatively weakly at 25C, even in EGTA. To which IQ domain does this weakly bound light chain bind? Although calmodulin binds to IQs 1 --3 with approximately the same strength in the presence of EGTA, we suggest that the readily released calmodulin is likely to be that bound to IQ2. To bind three calmodulins, IQs 1 --3, each of which are only 23 amino acids long, must be arranged without kinks ; this arrangement may produce unfavorable strain on each of the calmodulin molecules. Release of calmodulin from IQ2 would relieve all of that strain; release from IQs 1 or 3 would not. Strain relief also may accelerate calmodulin release in the presence of Ca2+; because Ca2+ apparently rearranges the three-dimensional interaction of calmodulin with IQ3, binding of an adjacent calmodulin -- on IQ2 -- might be destabilized even more . Despite the loss of calmodulin from T701-Myo1c induced by elevation of the temperature from 4C to 25C, the frictional ratio (a measure of the protein's asymmetry) increased . The neck-tail region of Myo1c thus appears to adopt a compact structure at 4C, becoming more extended at 25C. Less calmodulin may be released at lower temperatures because the Myo1c tail may bind to and stabilize calmodulin's interaction with the Myo1c neck. Implications for Myo1c activity | The Ca2+-dependent change in interaction of calmodulin with IQ1, the IQ domain closest to the motor domain, has important implications for Myo1c mechanochemical function. Although Ca2+ increases Myo1c ATPase activity, the ion completely halts in vitro motility . Ca2+-dependent changes in conformation may prevent amplification of a small converter-domain movement into a large motor step. In the presence of an external force, as is seen by Myo1c during an excitatory mechanical stimulus in a hair cell , Ca2+ (which enters the cell through open transduction channels), should permit Myo1c to go through its ATPase cycle, binding and unbinding from actin, but the altered interaction of calmodulin and IQ1 may prevent force production by the motor. We predict that Ca2+ will decrease the stiffness of a Myo1c-actin interaction, preventing coupling of the energy released by ATP hydrolysis to the swing of the neck . This behavior will assist Myo1c in its role of adaptation in hair cells, where the motor reduces force applied to the hair cell's transduction channel. A limitation of our experiments is the restriction of Myo1c binding to a single type of light chain, calmodulin. Other light chains can interact with IQ domains, including essential light chain isoforms and calmodulin-like protein . Although purified bovine adrenal Myo1c does not appear to have alternative associated light chains , we can not rule out the possibility that other light chains bind in a cellular context. Nevertheless, purified recombinant full-length Myo1c associated with calmodulin light chains exhibited actin-activated ATPase activity and motility in vitro , indicating that calmodulin can function as a Myo1c light chain. That Myo1c does not bind calmodulin tightly is, at first glance, surprising. Weak calmodulin binding may, however, permit access of IQ domains to intracellular Myo1c receptors. Accordingly, we have found that a Myo1c fragment containing only IQs 1 --3, partially complexed with calmodulin, binds avidly to hair-cell receptors; excess calmodulin blocks this interaction, probably by binding to an unoccupied IQ site on the Myo1c fragment . IQ2 is highly conserved between species, leading us to propose that hair-cell receptors interact through this region . Because Myo1c-interacting proteins in hair cells and elsewhere may interact through IQ domains, regulation of calmodulin binding to Myo1c -- for example, by Ca2+ -- likely affects coupling of the motor protein to its cargo. Conclusions : Under low Ca2+ conditions and normal ionic strength, calmodulin binds moderately tightly to three Myo1c IQ domains, IQ1, IQ2, and IQ3. IQ4 will only be occupied when the calmodulin concentration is very high. When linearly arranged in the Myo1c molecule, at least one calmodulin (most likely that bound to IQ2) is bound less tightly, probably due to steric constraints. Upon binding Ca2+, calmodulin bound to IQ2 dissociates; that bound to IQ1 either dissociates or changes its conformation sufficiently that chemomechanical coupling cannot ensue. Methods : Peptide -- calmodulin interaction on plates | Bullfrog Myo1c IQ peptides were synthesized (Genemed Synthesis, South San Francisco, CA) with N-terminal cysteine residues: IQ1 (residues 698 --720), CRKHSIATFLQARWRGYHQRQKFL; IQ2 (721 --743), CHMKHSAVEIQSWWRGTIGRRKAA; IQ3 (744 --766), CKRKWAVDVVRRFIKGFIYRNQPR; and IQ4 (767 --791; native cysteine at residue 767), CTENEYFLDYIRYSFLMTLYRNQPK. Peptide concentrations were measured by determining optical density at 280 nm, using calculated molar extinction coefficients of 7090 (IQ1), 11500 (IQ2), 7090 (IQ3), and 5240 M-1 cm -1 (IQ4). We also synthesized a negative-control peptide ("PVP") corresponding to amino acids 792 --816 of frog Myo1c (SVLDKSWPVPPPSLREASELLREMC; native C816) and a positive control IQ-peptide ("NM") corresponding to amino acids 29 --52 of bovine neuromodulin with an added C-terminal cysteine (KAHKAATKIQASFRGHITRKKLKC) . For measuring interaction of calmodulin with peptides conjugated to plastic plates, we incubated 10 muM peptide in phosphate-buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4, 1.4 mM KH2PO4, pH 7.4) overnight at room temperature in a maleimide-derivatized 96-well plate (Pierce, Rockford, IL). Peptide was present in large excess over free binding sites (25 --50 pmol) on the plates. To remove unconjugated peptides, plates were washed with PBS; unreacted sites were saturated by incubating with 10 mug/ml cysteine for 1 hour. We then incubated the peptide-conjugated plates with 50 nM Alexa Fluor 488 calmodulin (Alexa-calmodulin; Molecular Probes, Eugene, OR) in 100 mul of a solution that contained 150 or 400 mM KCl, 1 mM MgCl2, 100 muM ethylene glycol-bis(beta-aminoethylether)-N,N,N',N'-tetraacetic acid (EGTA) or 25 muM CaCl2, and 15 mM 2-[4-(2-hydroxyethyl)-1-piperazinyl] ethanesulfonic acid (HEPES) at pH 7.5. According to the manufacturer, Alexa-calmodulin had two dye moieties per calmodulin molecule; the modified residues were likely Lys-75 and Lys-94, the most reactive of calmodulin's lysine residues . After incubation for 2 hours at room temperature, we transferred 50 mul of the solution to another 96-well plate and measured fluorescence (excitation 485 nm; emission, 520 nm) using a BMG Labtechnologies Fluorostar 403 microplate fluorometer (Durham, NC). Under the assay conditions, the inner-filter effect (absorption of excitation or emission photons by the sample) was negligible. From this measurement, we calculated the amount of calmodulin bound to the conjugated peptides. In some experiments, we also included 0.1 --100 muM unconjugated IQ peptide; in that case, we carried out duplicate control reactions in underivatized 96-well plates to correct for fluorescence quenching exerted by IQ peptides. Peptide -- calmodulin interaction by fluorescence quench | We used empirically observed changes in the fluorescence intensity of Alexa-calmodulin, large in magnitude, to measure binding of IQ peptides to calmodulin. Peptides and 50 --500 nM Alexa-calmodulin were mixed in 96- or 384-well microtiter plates with 150 mM KCl, 1 mM MgCl2, 100 muM EGTA or 25 muM CaCl2, 0.5 mg/ml bovine serum albumin, and 15 mM HEPES at pH 7.5; in some experiments we added 50 --75 muM bovine-brain calmodulin. Total volume varied from 10 mul (384-well plates) to 100 mul (96-well plates). After 1 --2 hours at room temperature, fluorescence was read directly. When IQ peptides bound to Alexa-calmodulin, the fluorescence intensity was reduced as the quantum yield decreased (fluorescence quenching). We assumed that two fluorescent species were present, Alexa-calmodulin and IQ peptide-bound Alexa-calmodulin, and that the fluorescence intensity (I) was a linear combination of the fluorescence of the two species: I = fCaM ICaM +fCaM-IQ ICaM-IQ (1) where fCaM and fCaM-IQ are the mole fractions of the two components and ICaM and ICaM-IQ are their fluorescence intensities. Because the quantum yield of Alexa-calmodulin is reduced when IQ peptides bind, ICaM-IQ 1. Stoichiometry determination by gel scanning | T701-Myo1c and bovine-brain calmodulin were separated by sodium dodecyl sulfate gel electrophoresis (SDS-PAGE) and stained with Coomassie blue R250. Gels were scanned with a flatbed scanner; calmodulin was quantified using analysis of the resulting images with NIH Image version 1.62. The concentration of the T701-Myo1c heavy chain was determined by measuring absorbance at 280 nm, although the analysis was complicated by the uncertain calmodulin stoichiometry (p). To circumvent this problem, we solved several simultaneous equations for p. The molar extinction coefficient of the T701-Myo1c/calmodulin complex (epsilonT701-CaM) is given by: epsilonT701-CaM = epsilonT701 + pepsilonCaM (10) where epsilonT701 is the extinction coefficient of the T701-Myo1c heavy chain alone (57,990 M-1 cm-1), epsilonCaM is the extinction coefficient of calmodulin (2560 M-1cm-1), and p is the calmodulin:T701 stoichiometry. The concentration of T701-Myo1c heavy chain is given by: where A280 is the absorbance of the complex at 280 nm for a 1 cm pathlength. Finally, p[T701HC] = [CaM] (12) where [CaM] is the calmodulin concentration determined by gel scanning. Solving for p: Other methods | We measured the free Ca2+ concentrations in our solutions using spectrofluorometry with Calcium Green-2 (Molecular Probes). SDS-PAGE was carried out with 18% acrylamide Criterion gels (Bio-Rad Laboratories; Hercules, CA); gels were stained with Coomassie blue R250. Authors' contributions : Author 1 (PGG) conceived of the experimental approach, carried out many of the experiments, supervised the technician who performed the remainder, developed the methods for analysis, analyzed and interpreted the data, and wrote the manuscript. Author 2 (JLC) contributed to the development of the experimental approach, helped analyze and interpret the data, and edited the manuscript. Backmatter: PMID- 12473161 TI - Quality and correlates of medical record documentation in the ambulatory care setting AB - Abstract | Background | Documentation in the medical record facilitates the diagnosis and treatment of patients. Few studies have assessed the quality of outpatient medical record documentation, and to the authors' knowledge, none has conclusively determined the correlates of chart documentation. We therefore undertook the present study to measure the rates of documentation of quality of care measures in an outpatient primary care practice setting that utilizes an electronic medical record. Methods | We reviewed electronic medical records from 834 patients receiving care from 167 physicians (117 internists and 50 pediatricians) at 14 sites of a multi-specialty medical group in Massachusetts. We abstracted information for five measures of medical record documentation quality: smoking history, medications, drug allergies, compliance with screening guidelines, and immunizations. From other sources we determined physicians' specialty, gender, year of medical school graduation, and self-reported time spent teaching and in patient care. Results | Among internists, unadjusted rates of documentation were 96.2% for immunizations, 91.6% for medications, 88% for compliance with screening guidelines, 61.6% for drug allergies, 37.8% for smoking history. Among pediatricians, rates were 100% for immunizations, 84.8% for medications, 90.8% for compliance with screening guidelines, 50.4% for drug allergies, and 20.4% for smoking history. While certain physician and patient characteristics correlated with some measures of documentation quality, documentation varied depending on the measure. For example, female internists were more likely than male internists to document smoking history (odds ratio [OR], 1.90; 95% confidence interval [CI], 1.27 -- 2.83) but were less likely to document drug allergies (OR, 0.51; 95% CI, 0.35 -- 0.75). Conclusions | Medical record documentation varied depending on the measure, with room for improvement in most domains. A variety of characteristics correlated with medical record documentation, but no pattern emerged. Further study could lead to targeted interventions to improve documentation. Keywords: ambulatory care, documentation, medical records, correlation studies, quality of care, ambulatory care, documentation, medical records, correlation studies, quality of care, Background : Documentation in the medical record facilitates diagnosis and treatment, communicates pertinent information to other caregivers to ensure patient safety and reduce medical errors, and serves an important medical-legal function in risk management . Quality of documentation may also reflect the quality of care delivered, although recent studies have suggested that medical record documentation in the outpatient setting tends to underestimate the actual performance of preventive health care services and other indicators of quality care . Electronic medical record (EMR) systems may improve the quality of care delivered as well as the documentation of that care in the outpatient setting, but few studies have examined this issue . Determining the correlates of quality medical record documentation could thus lead to educational programs and other interventions to improve documentation, but few studies have rigorously examined the correlates of quality of chart documentation . We therefore undertook the present study to measure the rates of documentation of quality of care measures in an outpatient primary care practice setting that utilizes an EMR. We also aimed to examine the physician- and patient-level variables that correlated with the quality of medical record documentation. We studied the abstracted outpatient electronic medical records of 834 patients who received care from 117 internists and 50 pediatricians at 14 practice locations in 1998. Methods : Setting | Medical records were abstracted in 1998 at 14 sites of Harvard Vanguard Medical Associates (HVMA, formerly known as Harvard Community Health Plan, and later, the Health Centers Division of Harvard Pilgrim Health Care) in greater Boston, Massachusetts. In 1998, more than 90% of patients at HVMA had capitated health insurance coverage from a single health maintenance organization, Harvard Pilgrim Health Care. At the time of the study, HVMA was transitioning from one electronic medical record system, the COSTAR-based Automated Medical Record System to the current system, EpicCare . EpicCare records constituted the majority of those reviewed for this study. Both systems have designated, coded fields for capturing the quality measures described below, and both systems have free-text fields for clinicians to enter narrative information from patient encounters. Physicians themselves enter the majority of data in EpicCare, either by populating specific fields (e.g., smoking status, including number of cigarettes per day and duration of smoking; alcohol consumption [yes/no and ounces per week]) or by keyboard entry of free text. Depending on the physician's style, most data entry occurs immediately upon completion of the office visit, while some occurs during the visit in the presence of the patient and some occurs at the end of a clinical session. A minority of office notes are dictated; these transcribed notes are entered only as narrative text and do not result in any coding of specific fields. Physicians and patients | At the time of the study, 212 physicians provided primary care at HVMA, of whom 167 had completed their malpractice insurance carrier's biennial survey of clinical and teaching activities; these 167 primary care physicians (117 internists and 50 pediatricians) constitute the study sample. With three exceptions, five patient records per physician were selected at random for review; four records were reviewed for each of two physicians and six records for one physician. A total of 834 patient records were reviewed. Data collection and physician-level variables | Data sets containing medical record review information and self-reported physician teaching and patient care activities were obtained through permission of the Harvard Risk Management Foundation. Medical record reviews were conducted as a component of quality assurance procedures to fulfill accreditation requirements of the National Center for Quality Assurance (NCQA). Trained nurse abstractors from a medical record consulting firm conducted the reviews using an abstraction instrument containing 51 questions modeled after NCQA's 1999 Health Plan Employer Data and Information Set (HEDIS) quality measures . Nurse abstractors reviewed both the free-text and coded-field components of the medical records. Physicians reported the amount of direct patient care they provided in hours per week; they reported the amount of teaching they did as measured by the number of patients per month for whom they supervised care, as opposed to directly provided care. Physicians were classified as teachers if they reported supervising the care of any patients each month or supervising patient care in any months of the year. Specialty, gender and year of medical school graduation were obtained for each physician from the World Wide Web sites of the American Medical Association and the medical licensing boards of Massachusetts, New York, and California . Practice site was dichotomized according to whether or not the site was located within Boston city limits, because we anticipated significant differences in patient populations and physician practices across sites in these locations. Measures of medical record documentation quality | Before examining the data, we selected for analysis five measures from the chart review instrument that we believed were the most important indicators of quality of medical record documentation in ambulatory care. The five measures were medications, allergies, immunizations, smoking history and compliance with age-appropriate screening guidelines. All measures were dichotomous (yes/no) variables that were coded "yes" if the trained abstractors determined that the relevant information was appropriately documented. For medication documentation, we created a dichotomous (yes/no) variable that was scored "yes" if the record contained documentation of current medications either in the encounter text of the complete history and physical examination or in the medication list (i.e., medication field) of the computerized medical record. Allergy documentation was coded in a similar fashion. The other three measures -- immunizations, smoking history and compliance with age-appropriate screening guidelines -- are analogous to items found in the 1999 HEDIS List of Measures under "Effectiveness of Care" . For "compliance with age-appropriate screening guidelines," abstractors referred to Harvard Vanguard's clinical guidelines at the time of the study. These guidelines included colorectal cancer screening for adults age 50 years or older, mammography for women age 50 years or older, and Pap smears for all women. Prostate cancer screening was not included in this measure. For children, age-appropriate screening included sensory screening (vision and hearing), and lead testing. The "compliance with age-appropriate screening guideline" variable was coded "yes" if the abstractor determined that each relevant screening item had been documented. There were no missing data. Chart abstractors coded "not applicable" for one or more measure on less than one percent of the records reviewed. We recoded this small number of "not applicable" values as "no" on the assumption that any items not coded as "yes" were lacking appropriate documentation of the specified measures. Statistical analysis | Characteristics of physicians were compared by specialty using Student's t-test for continuous measures and by Fisher's exact test for categorical variables. Analyses of documentation behavior were stratified by specialty because we anticipated differences between pediatricians and internists due to the nature of their patient populations, methods of practice, and the documentation measures we selected for analysis. No analyses were performed for immunization documentation by pediatricians since all pediatricians documented immunizations. To account for clustering of patients within each physician, we used mixed effects (hierarchical) logistic modeling for documentation outcomes . This technique accounts for the correlation between outcomes measured on patients who share a physician . We determined whether physician teaching status, physician gender, years since medical school, clinical hours per week of direct patient care, practice site, patient age, and patient gender were correlated with better chart documentation for each of the five measures. We hypothesized a priori that physicians' teaching status would be correlated with the quality of documentation. Other physician characteristics were included because we expected that documentation practice would vary by clinician experience and by geography, although we did not have an a priori hypothesis about the direction of these effects. Analyses were performed using the Glimmix Macro in Statistical Analysis System (SAS) software The study was approved by the Human Studies Committee (Institutional Review Board) of Harvard Pilgrim Health Care. Results : Characteristics of physicians and patients | Table shows the characteristics of the 117 internists and 50 pediatricians in the study. About two-thirds of internists and pediatricians practiced outside Boston City limits. While internists reported more clinical hours per week than pediatricians (34 +- 14 vs. 29 +- 9; P = 0.006), years since completing medical school were higher for pediatricians (18 +- 8 vs. 22 +- 9; P = 0.003). Table 1 | Characteristics of 167 HVMA physicians in a study of medical records documentation. The mean (+- SD) age of patients was 9.3 +- 5.5 years in the pediatrics group and 57.5 +- 15.0 in the internist group. Fifty-five percent of all patients were female. Overall rates of medical record documentation quality | Table shows the unadjusted rates of fulfillment of the five measures of medical record documentation quality for internists and for pediatricians. Table 2 | Unadjusted rates of completion of medical record documentation among internists and pediatricians for each of five selected measures.* Correlates of medical record documentation quality | No consistent patterns emerged among the physician or patient characteristics and documentation quality measures in this study (Table 3 -- see ). Female internists were more likely than male internists to document smoking history (odds ratio [OR], 1.90; 95% confidence interval [CI], 1.27 -- 2.83) but were less likely to document drug allergies (OR, 0.51; 95% CI, 0.35 -- 0.75). With each subsequent year since completing medical school, internists were less likely to document drug allergies (OR, 0.97; 95% CI, 0.95 -- 0.99) and immunizations (OR, 0.91; 95% CI, 0.87 --0.96). For internists, increased clinical time was associated with better documentation for smoking history (OR, 1.02; 95% CI, 1.01 -- 1.03) but not for any of the other measures. As compared with internists practicing at sites within Boston city limits, those practicing outside Boston city limits were less likely to document, medications (OR, 0.47; 95% CI, 0.26 -- 0.85), and immunizations (OR, 0.19; 95% CI, 0.07 --0.47) but were more likely to document smoking history (OR, 1.55; 95% CI, 1.08 -- 2.20). Internists were more likely to document drug allergy status (OR, 2.48; 95% CI, 1.65 -- 3.74), medications (OR, 1.49; 95% CI, 1.18 -- 1.87), and compliance with screening guidelines (OR, 1.36; 95% CI, 1.06 -- 1.73) for female patients than for male patients. However, internists were less likely to document smoking history (OR, 0.51; 95% CI, 0.34 -- 0.79) and immunizations (OR, 0.61; 95% CI, 0.47 --0.79) for female patients than for male patients. Internists were less likely to document smoking history (OR, 0.98; 95% CI, 0.96 -- 0.99) but more likely to document immunizations (OR, 1.02; 95% CI, 1.01 -- 1.03) with every yearly increase in patient age. Female pediatricians were less likely than male pediatricians to document drug allergy status (OR, 0.47; 95% CI, 0.28 -- 0.79) but were more likely to document smoking history (OR, 2.78; 95% CI, 0.85 -- 9.11); interpretation of this latter result is limited by imprecision of the effect size as manifested by the wide confidence interval. For pediatricians, increased clinical time was associated with better documentation for compliance with screening guidelines (OR, 1.22; 95% CI, 1.06 -- 1.41) but not for any of the other measures. When compared with pediatricians practicing in the City of Boston, pediatricians practicing outside Boston city limits were more likely to document drug allergies (OR, 3.17; 95% CI, 1.05 -- 9.57) and medications (OR, 7.93; 95% CI, 1.87 -- 33.68) but less likely to document compliance with screening guidelines (OR, 0.08; 95% CI, 0.01 -- 0.58). Pediatricians were more likely to document smoking history with increasing patient age (OR, 1.61; 95% CI, 1.12 -- 2.31). Pediatricians were more likely to document drug allergy status for female patients than for male patients (OR, 1.62; 95% CI, 1.19 -- 2.20) but less likely to document compliance with age-appropriate screening guidelines for female patients (OR, 0.17; 95% CI, 0.08 -- 0.38). Discussion : In this study of the quality of medical record documentation among 167 primary care physicians, we found very high levels of documentation for immunizations and compliance with screening guidelines, but lower levels of documentation for medications, allergies, and smoking status. Although no characteristics consistently correlated with all measures of documentation quality, some characteristics had important relationships to certain measures. For example, female internists were more likely than male internists to document smoking history but were less likely to document drug allergies. Pediatricians and internists excelled in different domains of documentation. For example, all 50 pediatricians fulfilled the documentation measure for immunizations for all 250 patients whose records were sampled. In comparison, internists met documentation standards for immunization for 562 of 584 patients (96%). These rates of immunization documentation are similar to previously reported rates of immunization for patients in this practice group , thereby supporting the validity of this study, and compare favorably with national benchmarks for immunization. Pediatricians and internists also differed in their patterns of documenting smoking status. Documentation of smoking status may be considered a more appropriate measure of medical record documentation quality for internal medicine than for pediatrics, since very young children (e.g., infants and toddlers) are not at risk for smoking cigarettes themselves. However, national guidelines urge pediatricians to address children's exposure to passive smoking during routine pediatric practice, beginning at the prenatal visit . Pediatricians were more likely to document smoking status for their older patients, providing construct validity to this measure. Curiously, however, internists were less likely to document smoking status with increasing age of the patient. These findings may suggest that pediatricians and internists focus their smoking documentation, and perhaps their smoking prevention and cessation efforts, on adolescents and young adults, target populations advocated by national guidelines . Although no prior studies have comprehensively examined the correlates of medical record documentation quality in the primary care setting, one small study measured medical residents' chart-documentation practices . That study of 26 medical records suggested that documentation was more complete in charts of male patients than in female patients, a finding not confirmed in the present study. A recent study reported the medical record documentation practices of resident physicians and faculty members in a university-based internal medicine training program . In that chart abstraction study, which used measures similar to those in the present study, rates of documentation also varied widely, although that study emphasized performance scores for both preventive health services as well as chronic disease management. Others have found that certain educational interventions improved documentation quality in the ambulatory care setting . For example, clinical audit and standardized follow up improved smoking history documentation from 28% to 88% in charts of patients with diabetes . The low pre-intervention rate of smoking documentation in that study is consistent with the rates we observed among internists (38%) and pediatricians (20%) in the present study. This study provided a novel and valuable glimpse into the quality of medical record documentation of primary care physicians using an electronic medical record system in an ambulatory care setting and enabled the assessment of the correlation between physician and patient characteristics and the quality of medical record documentation. Strengths of the study included the sampling of physicians and patients from 14 practice sites, including both pediatrics and internal medicine, in both urban and suburban locations; the systematic medical record review and abstraction process using measures of documentation quality that resonate with widely accepted NCQA measures of quality of care; and the hierarchical (multi-level) statistical modeling techniques used to control for clustering and for potential confounding by variables at the level of the physician and the patient. Several important limitations should be considered in interpreting the results. The first is the lack of evidence of inter-rater reliability among chart abstractors. Although chart abstractors were registered nurses specially trained in medical record review, it is conceivable that different abstractors judged satisfactory documentation of a measure differently. Misclassification of documentation quality measures could have biased the results toward the null, and a true relationship between some of the physician or patient variables and documentation quality may not have been detected by this study. A second limitation is that documentation in the medical record may not accurately reflect care delivered. One study showed that the validity of the medical record for measuring delivery of different health services depended on the service in question. Specifically, medical record review more accurately reflected the administration of immunizations than the provision of smoking cessation advice . A third limitation is the possibility that unmeasured variation in "case-mix" could have accounted for variation in documentation quality. For example, physicians with generally sicker patients might be less likely to document health maintenance items, such as compliance with screening guidelines. Conversely, physicians might be expected to document more carefully the medications, allergies, and immunization status of more chronically ill patients. Finally, the physicians' self report of supervising patient care may not have accurately reflected their true status as teachers. Although social desirability bias may have led those physicians who do any teaching to overestimate their actual teaching time, we believe it is less likely that physicians who do no teaching would actually report some teaching time. Dichotomizing this exposure variable should have helped to minimize the social desirability bias. In summary, medical record documentation varied depending on the measure in question, with room for improving documentation in most domains. Despite the presence of an electronic medical record designed to facilitate documentation, rates of documentation for drug allergy and smoking status fell below desirable levels. We found that no consistent pattern of correlates of medical record documentation quality emerged in this study. Future studies could investigate whether other factors, such as socioeconomic status and insurance carrier of patients, correlate with medical record documentation quality in ambulatory care settings. Clarification of the correlates of medical record documentation quality can enhance and focus educational strategies to improve clinicians' documentation practices. Competing interests : None declared. Authors' contributions : CMS participated in the design of the study, carried out statistical analyses, and drafted the manuscript. KPK participated in the design of the study and supervised the performance of statistical analyses. SRS conceived of the study, participated in its design and coordination, and participated in drafting the manuscript. All authors read and approved the final manuscript. Pre-publication history : The pre-publication history for this paper can be accessed here: Backmatter: PMID- 12509223 TI - Src Kinase becomes preferentially associated with the VEGFR, KDR/Flk-1, following VEGF stimulation of vascular endothelial cells AB - Abstract | Background | The cytoplasmic tyrosine kinase, Src, has been found to play a crucial role in VEGF (vascular endothelial growth factor) -- dependent vascular permeability involved in angiogenesis. The two main VEGFRs present on vascular endothelial cells are KDR/Flk-1 (kinase insert domain-containing receptor/fetal liver kinase-1) and Flt-1 (Fms-like tyrosine kinase-1). However, to date, it has not been determined which VEGF receptor (VEGFR) is involved in binding to and activating Src kinase following VEGF stimulation of the receptors. Results | In this report, we demonstrate that Src preferentially associates with KDR/Flk-1 rather than Flt-1 in human umbilical vein endothelial cells (HUVECs), and that VEGF stimulation resulted in an increase of Src activity associated with activated KDR/Flk-1. These findings were determined through immunoprecipitation-kinase experiments and coimmunoprecipitation studies, and were further confirmed by GST-pull-down assays and Far Western studies. However, Fyn and Yes, unlike Src, were found to associate preferentially with Flt-1. Conclusions | Thus, Src preferentially associates with KDR/Flk-1, rather than with Flt-1, upon VEGF stimulation in endothelial cells. Our findings further highlight the potential significance of upregulated KDR/Flk-1-associated Src activity in the process of angiogenesis, and help to elucidate more clearly the specific roles and mechanisms involving Src family tyrosine kinase in VEGF-stimulated signal transduction events. Keywords: Background : Angiogenesis is a process that is of critical importance to tumorigenesis and tumor metastasis, as well as to the growth and maintenance of normal vasculature. Both the processes of angiogenesis and tumorigenic proliferation are key events essential for the successful growth of vascularized tumors. Indeed, inhibition of neovascularization has been shown to abolish or slow tumor growth in various experimental models . Angiogenesis plays both beneficial and detrimental roles in normal and disease processes. For example, abnormally enhanced angiogenic responses are observed during conditions such as diabetic retinopathy and rheumatoid arthritis . On the other hand, promotion of the angiogenic response has been demonstrated to be beneficial in treating ischemic conditions, such as myocardial ischemia/infarction . Thus, a comprehensive understanding of the molecular mechanisms underlying the angiogenic response may help in preventing and treating many diseases displaying aberrant angiogenesis. VEGF is an angiogenic factor that is very closely associated with the induction and maintenance of neovasculature in human cancers. Although other splice variant forms of VEGF have been discovered to be encoded and expressed, VEGF-A165 remains the main biologically relevant splice form. VEGF-A165 is produced from tumor cells and stromal fibroblast cells , and is able to bind both KDR/Flk-1 and Flt-1 (two main VEGF receptors of the VEGF family expressed by vascular endothelial cells) . VEGF-A165 exhibits two major biological activities. One, is that it promotes increased microvascular hyperpermeability to circulating plasma proteins (one of the steps important in early vasculogenesis) . This leads to extravasation of fibrinogen and other plasma proteins, activation of the clotting system, and deposition of a fibrin gel in the extravascular space. The fibrin gel provides a proangiogenic matrix that favors and supports the migration of fibroblasts and the growth of new blood vessels . Another major biological activity of VEGF-A165 is its ability to stimulate vascular endothelial cell proliferation . The VEGF receptors, KDR/Flk-1 and Flt-1, are receptor tyrosine kinases structurally related to receptors of the Fms/Kit/PDGFR family. Both VEGFRs contain a large insert sequence in their respective intracellular kinase domains and seven immunoglobulin-like domains in their respective NH2-terminal regions. Also, both KDR/Flk-1 and Flt-1 have been shown to be expressed mainly on endothelial cells , and to play important roles in transducing signals upon VEGF stimulation of the endothelium. Previous knockout studies have demonstrated that KDR/Flk-1 (-/-) mice exhibit embryonic lethality (at E8.5-E9.0). They exhibit no development of blood vessels and very low hematopoiesis . Flt-1 (-/-) mice also exhibit embryonic lethality (at E8.5), but as a result of overgrowth of endothelial cells and disorganization of blood vessels . This suggests that at least in the early stages of embryogenesis, KDR/Flk-1 is essential for both proliferation and differentiation of endothelial cells, whereas Flt-1 can potentially act as a negative regulator of cell proliferation in the assembly of the vascular endothelium . These and other observations indicate that KDR/Flk-1 and Flt-1 may have different biological functions, and suggest that they may utilize different downstream signaling pathways. Src is a cytoplasmic protein tyrosine kinase, whose activation and recruitment to perimembranal signaling complexes has important implications for cellular fate. It has well-documented that Src protein levels and Src kinase activity are significantly elevated in human breast cancers , colon cancers , and pancreatic cancers as well. Furthermore, activation of Src by growth factors and growth factor receptors, such as PDGF(R), EGF(R), and HER2/Neu, has been shown to promote cell proliferation . The biological importance of Src in the VEGF angiogenic system was published in a report by Eliceiri and collaborators . Their experiments used a CAM model (chick embryo chorioallantoic membrane model) and a murine subcutaneous angiogenesis model to demonstrate a crucial role for Src in VEGF-dependent vascular permeability. Also, Suarez and Ballmer-Hofer proposed from their study that VEGF-mediated disruption of gap junctional communication in endothelial cells via Src was the cause of local increased endothelial permeability . To date, it has been assumed that Src becomes activated in endothelial cells following binding to an activated VEGF receptor . However, it is not known whether Src binds and is activated by one VEGFR or is able to bind and be activated by both VEGFRs to a substantial degree. Our study reveals that Src preferentially associates with and is subsequently activated by KDR/Flk-1 rather than with Flt-1 in HUVECs (human umbilical vein endothelial cells). Results and Discussion : Coimmunoprecipitation of Src activity with KDR/Flk-1 | A potentially important biological role for Src in VEGF-dependent vascular permeability has been previously reported . However, it is not known if one or both of the main endothelial VEGF receptors (VEGFRs) are specifically able to bind to and activate Src. We compared the ability of the two main types of endothelial VEGFRs (KDR/Flk-1 and Flt-1) to associate with and activate Src upon VEGF treatment, by determining whether Src activity could be coimmunoprecipitated with either receptor . HUVECs were either serum-starved, or serum-starved and then treated with VEGF, or lysed (unstarved, untreated) at subconfluence. Cell extracts of these samples were then immunoprecipitated using alpha-KDR/Flk-1 (N-931), alpha-Flt-1 (H-225), or nonimmune rabbit IgG; and kinase assays were performed using these immunocomplexes. Upon VEGF treatment, both KDR/Flk-1 and Flt-1 become tyrosine-phosphorylated . This is accompanied by a 2.15-fold increase in phosphorylation of a 60 kDa protein (corresponding to the molecular weight of Src) associated with KDR/Flk-1 and a 1.71-fold increase in enolase substrate phosphorylation activity associated with the KDR/Flk-1 immunocomplexes . However, there was very little detectable p60 kDa phosphorylation or enolase substrate phosphorylation activity associated with the Flt-1 immunocomplexes. Yet, Flt-1 was found to be immunoprecipitatable and tyrosine-phosphorylated upon VEGF treatment (Fig. , left panel). Also, nonimmune rabbit IgG did not coprecipitate any significant levels of p60 kDa phosphorylation or enolase substrate phosphorylation. Kinase reactions were also performed on Src optimal peptide substrate (AEEEIYGEFEAKKKK ) with similar results (unpublished observations). Figure 1 | Src Activity coprecipitates with KDR/Flk-1 Src Activity coprecipitates with KDR/Flk-1. (A) HUVECs were either serum-starved (-), or starved and then VEGF-treated (15 ng/mL VEGF for 15 min.) (+), or lysed (unstarved, untreated) at subconfluence (C). Cell extracts of these samples were immunoprecipitated using alpha-Flt-1 (H-225), alpha-KDR/Flk-1 (N-931), or rabbit nonimmune IgG. Kinase assay reactions were subsequently performed using enolase substrate. Samples were resolved on 10% SDS-PAGE gels. The gels were treated with 1 M KOH at 55C for 1 h after electrophoresis to remove background due to serine phosphorylation. The gels were then dried and exposed to autoradiographic film (upper panel). (B) Cell extracts from the indicated treatment conditions were immunoprecipitated using alpha-Flt-1 (H-225) and immunoblotted using alpha-phosphotyrosine (4G10) or alpha-Flt-1 (H-225) (left panel); or the cell extracts were immunoprecipitated using alpha-KDR/Flk-1 (A-3) and immunoblotted using alpha-phosphotyrosine (4G10) or alpha-KDR/Flk-1 (N-931) (right panel). Time-dependent VEGF stimulation of Src association with KDR/Flk-1 | In order to further study the specificity and time-course of Src association with KDR/Flk-1, HUVECs (human umbilical vein endothelial cells) were serum-starved and then treated with VEGF (15 ng/mL) for 0, 15, 30, 60, or 120 min. As seen in Figure , Src association with KDR/Flk-1 peaked around 15 min of VEGF treatment. Concurrently, KDR/Flk-1 phosphorylation initially increased and then decreased, peaking at about 15 minutes . Our results also indicate that the association of activated Src phosphorylated on tyrosine 419 with KDR/Flk-1 peaked at approximately 15 minutes of VEGF treatment. This was detected by alpha-Src phosphotyrosine 419 specific Src antibody that is able to detect an activated form of c-Src . Because of logistical considerations, we did not examine time points between 0 and 15 minutes. Therefore, we cannot exclude the possibility that a substantial level of activation of KDR/Flk-1 and Src did not also occur at these earlier times. We also studied the time course of Src association with Flt-1. However, we did not find detectable Src association with Flt-1 up to 120 minutes after VEGF treatment . At the same time it was evident that Flt-1 was phosphorylated upon VEGF treatment . These results suggested to us that Src preferentially associated with KDR/Flk-1 rather than Flt-1 upon VEGF treatment. Figure 2 | Time-dependent VEGF stimulation of Src association with KDR/Flk-1 Time-dependent VEGF stimulation of Src association with KDR/Flk-1. HUVECs were serum-starved and then treated with VEGF (15 ng/mL) for 0, 15, 30, 60, or 120 min. (A) Cell extracts from the indicated treatment conditions were immunoprecipitated using alpha-Src 2 --17, and immunoblotted using alpha-KDR/Flk-1 (N-931) or alpha-Src 327. (B) Cell extracts from the indicated treatment conditions were immunoprecipitated using alpha-KDR/Flk-1 (N-931), and immunoblotted using alpha-phosphotyrosine (4G10), alpha-phosphotyrosine 419 Src, or alpha-KDR/Flk-1 (A-3). (C) Cell extracts from the indicated treatment conditions were immunoprecipitated using alpha-Src 2 --17, and immunoblotted using alpha-Flt-1 (H-225) or alpha-Src 327. (D) Cell extracts from the indicated treatment conditions were immunoprecipitated using alpha-Flt-1 (H-225), and immunoblotted using alpha-phosphotyrosine (4G10), alpha-Src 327, or alpha-Flt-1 (H-225). Src SH3-SH2 association with KDR/Flk-1 increases upon VEGF stimulation | In order to confirm Src preference for interaction with Flk-1/KDR compared to Flt-1, we tested GST-Src domain fusion constructs for their ability to bind to KDR/Flk-1 in GST-pull-down experiments. The results indicated that VEGF-stimulated KDR/Flk-1 was able to bind strongly to constructs containing both the Src SH3 and SH2 domains . Some basal binding of KDR/Flk-1 to the Src SH3-SH2 and Src U-SH3-SH2 was observed, although the extent of this binding was less than the binding observed after VEGF treatment. The Src unique domain alone was not able to bind to KDR/Flk-1, but it seemed to augment binding when present in a construct containing both the Src SH3 and SH2 domains . Notably, however, upon VEGF stimulation and tyrosine phosphorylation of KDR/Flk-1, the ability of KDR/Flk-1 to bind to Src SH3-SH2 was increased. Figure 3 | GST pull-downs and Far Western blots indicate increased Src SH3-SH2 interaction with KDR/Flk-1 upon VEGF stimulation GST pull-downs and Far Western blots indicate increased Src SH3-SH2 interaction with KDR/Flk-1 upon VEGF stimulation. (A) HUVECs were either serum-starved (0') or starved and then VEGF-treated (15 ng/mL for 15 min) (15'). The indicated GST-Src domain fusion proteins (see below) bound to glutathione-agarose beads were used (500 pmol equimolar amounts each), in pull-down incubations with equal amounts of protein lysate samples (500 mug each). The pull-down complexes were resolved on a 7% acrylamide SDS-PAGE gel and transferred to nitrocellulose. Immunoblotting was then performed using alpha-KDR/Flk-1 (A-3) or alpha-Flt-1 (H-225). GU, GST-Src unique domain fusion protein; GU32, GST-Src unique SH3 SH2 domain fusion protein; G32, GST-Src SH3 SH2 domain fusion protein. (B) Cell extract (pre-pull-down) input levels of phosphotyrosine KDR/Flk-1 and KDR/Flk-1, from HUVECs either serum-starved (0') or starved and then VEGF-treated (15'). (C) Cell extract (pre-pull-down) input levels of phosphotyrosine Flt-1 and Flt-1, from HUVECs either serum-starved (0') or starved and then VEGF-treated (15'). (D) HUVECs were serum-starved (-) or VEGF-treated with VEGF (15 ng/mL for 15 min) (+). The cell extracts were then immunoprecipitated using alpha-KDR/Flk-1 (N-931). The immunocomplexes were resolved on a 7% acrylamide SDS-PAGE gel. Biotinylated GST and biotinylated GST-Src SH3-SH2 fusion protein were used in Far Western blotting. Detection was performed using horseradish peroxidase-conjugated ExtrAvidin, followed by enhanced chemiluminescence detection reagents. Immunoprecipitation-Western blots for levels of phosphotyrosine KDR/Flk-1 and KDR/Flk-1 of the serum-starved (-) or VEGF-treated HUVEC (+) cell extracts were also performed in parallel. In contrast, when we tested the same panel of GST-Src domain fusion proteins for their ability to bind to Flt-1, we did not observe detectable binding of Flt-1 with the GST-Src domain fusion proteins . Levels of Flt-1 were equivalent in serum-starved versus VEGF-treated cells, and Flt-1 exhibited tyrosine phosphorylation upon VEGF treatment . These data revealed that Flt-1 was present and detectable in lysate from HUVECs, but was not able to interact with the major protein binding domains of Src. Thus, this data further supports a binding preference of Src for KDR/Flk-1 following VEGF stimulation. To further confirm that VEGF treatment leads to an increased ability of tyrosine phophorylated KDR/Flk-1 to interact with Src SH3-SH2, we performed Far Western experiments . HUVECs were either serum-starved, or starved and then treated with VEGF. Following immunopreciptiation of cell extracts using alpha-KDR/Flk-1 antibody, biotinylated GST-Src fusion proteins were used in immunoblotting. Figure demonstrates that KDR/Flk-1 exhibits increased interaction with Src SH3-SH2 domains after stimulation of HUVECs with VEGF, and after tyrosine phosphorylation of KDR/Flk-1. Levels of KDR/Flk-1 protein remained similar before and after VEGF treatment, and there was an increase in tyrosine phosphorylation of KDR/Flk-1 after VEGF treatment. Control experiments indicated that KDR/Flk-1 (prior to VEGF treatment and post-VEGF treatment) did not interact with biotinylated-GST alone . At this time, we do not know the exact nature of the binding interactions between Src and KDR/Flk-1. However, we have observed that both the Src SH2 and SH3 domains appear to be required for optimal binding to KDR/Flk-1 (data not shown). It is possible to speculate that VEGF binding and stimulation of KDR/Flk-1 confers the proper conformation to KDR/Flk-1 such that it is able to initially bind the Src SH3 domain via the cytoplasmic KDR/Flk-1 proline-rich regions and/or KDR/Flk-1 phosphotyrosine sites. When Src is able to bind to this accessible KDR/Flk-1 region, the Src protein itself may be further opened up by Src SH2 domain binding to a phosphotyrosine site on KDR/Flk-1, and Src's active site might then be made fully available for further propagation of transducing cellular signals. Thus in this manner, the binding interactions between KDR/Flk-1 and Src SH3 and Src SH2 domain of Src would act to facilitate disruption of intramolecular, inhibitory Src SH3/SH2 interactions. This would ultimately allow Src to open up into an active conformation, thereby permitting autophosphorylation on Src Tyr 419 , and also upregulation of Src activity by dephosphorylation of negative-regulatory C-terminal phosphotyrosine 530 by phosphatases . We are currently conducting further studies to investigate this model for Src-KDR/Flk-1 interaction. Although we did not detect GST-Src unique domain association with KDR/Flk-1, the presence of the Src unique domain together with Src SH3-SH2 did increase the extent of the Src-KDR/Flk-1 interaction. At present, we do not know the basis for this enhanced level of binding. However, it is notable that both GST-Unique-SH3-SH2 and GST-SH3-SH2 Src display enhanced association with VEGF-stimulated, tyrosine-phosphorylated KDR/Flk-1 compared to unstimulated KDR/Flk-1. Our findings suggest that the Src SH3 domain together with the Src SH2 domain is able to interact with KDR/Flk-1. Src family SH3 domains preferentially bind proline-rich motifs with the sequence RPLPXXPX . The optimal Src SH2 binding consensus sequence is PhipYEEI/Phi (where Phi = hydrophobic amino acids, and E = glutamate or other hydrophilic amino acids) . Significantly, several such potential Src-binding sites are present on the C-terminal cytoplasmic domain of KDR/Flk-1. A candidate proline-rich, SH3-binding, consensus sequence exists in KDR/Flk-1 at amino acid region 1192 --1198. Also, several tyrosine phosphorylation sites exist in KDR/Flk-1. Previously, Cunningham et al. have shown that tyrosine residues 801 (juxtamembranal) and 1175 (C-terminal) were phosphorylated on KDR/Flk-1. Tyrosine residues 951 and 996 (in the kinase insert domain), and 1054 and 1059 (in the C-terminal kinase domain) on KDR/Flk-1 have also been demonstrated to be phosphorylated . Also, Takahashi and collaborators have shown that in cultured endothelial cells, tyrosine residues 1175 and 1214 are major phosphorylation sites. In a recent study published by Meyer et al. , VEGF activation of murine Flk-1 was found to stimulate Src activity; however, mutational analysis of tyrosine 1212 (homologous tyrosine in murine Flk-1 to tyrosine 1214 in human KDR) did not impair Src activation, providing an important clue that tyrosine 1212 of murine Flk-1 might not be the major tyrosine involved in binding interactions with Src. Coimmunoprecipitation of Fyn and Yes with Flt-1 | The observed association of Src with KDR/Flk-1 upon VEGF stimulation and the lack of detectable association of Src with Flt-1 upon VEGF stimulation prompted us to examine whether other Src family kinases (such as Fyn and Yes) might be recruited in a similar or different manner upon VEGF stimulation. HUVECs were serum-starved and treated with VEGF 15 ng/mL for 15 min, and the cell extracts were immunoprecipitated with either alpha-Fyn or alpha-Yes, and Western blotted for Flt-1. Unlike Src, Fyn and Yes were found to preferentially associate with Flt-1, following VEGF treatment (Fig. and ). Concurrently, Flt-1 immunoprecipitation and Flt-1 tyrosine phosphorylation was also detectable . In contrast, we did not observe any detectable amounts of Fyn or Yes coimmunoprecipitating with KDR/Flk-1 upon VEGF treatment (Fig. and ). Control experiments indicate that indeed, despite our observation that Fyn and Yes did not seem to coimmunoprecipitate with KDR/Flk-1, KDR/Flk-1 immunoprecipitation and KDR/Flk-1 tyrosine phosphorylation was still detectable . Figure 4 | Coimmunoprecipitation of Fyn with Flt-1, and Yes with Flt-1 Coimmunoprecipitation of Fyn with Flt-1, and Yes with Flt-1.HUVECs were serum-starved (-), or starved and then treated with VEGF (15 ng/mL for 15 min) (+). The cell extracts were then subjected to immunoprecipitation-Western blotting experiments. (A) Cell extracts from the indicated treatment conditions were immunoprecipitated using alpha-Fyn, and immunoblotted using alpha-Flt-1 (H-225), alpha-KDR/Flk-1 (A-3), or alpha-Fyn. (B) Cell extracts from the indicated treatment conditions were immunoprecipitated using alpha-Yes, and immunoblotted using alpha-Flt-1 (H-225), alpha-KDR/Flk-1 (A-3), or alpha-Yes. (C) Cell extracts from the indicated treatment conditions were immunoprecipitated using alpha-Flt-1 (H-225), and immunoblotted using alpha-phosphotyrosine (4G10) or alpha-Flt-1 (H-225). (D) Cell extracts from the indicated treatment conditions were immunoprecipitated using alpha-KDR/Flk-1 (N-931), and immunoblotted using alpha-phosphotyrosine (4G10) or alpha-KDR/Flk-1 (A-3). These results are consistent with previous findings by Waltenberger and collaborators who demonstrated that Fyn and Yes exhibited increased levels of tyrosine phosphorylation upon VEGF stimulation (using stably transfected, VEGFR-overexpressing porcine aortic endothelial cell lines). Interestingly, they observed that, in porcine aortic endothelial cell lines stably overexpressing Flt-1, increases in Fyn and Yes tyrosine phosphorylation occurred in the VEGF-treated cells (2.4 fold and 2.1 fold, respectively). Also, they noted that increases in Fyn and Yes tyrosine phosphorylation did not occur upon VEGF stimulation in the KDR/Flk-1-overexpressing cell line. However, they did not examine Src in their study. Thus, taken together, our data and this previous report indicate that Src preferentially interacts with KDR/Flk-1, whereas Fyn and Yes preferentially interact with Flt-1, following VEGF stimulation. A bit unexpectedly, through the course of our experiments we have found that our detection of Flt-1 in the HUVECs used here is somewhat better than has been reported in many previous publications (Fig. , ). We note that the HUVECs utilized in our study are primary and early passage cells, kindly provided by Dr. K. Patel. Also, they are seeded from harvest initially in human serum rather than fetal calf serum. Therefore, it is possible that they might retain many endothelial specific markers (such as Flt-1) at higher levels than might be present in HUVECs utilized under different conditions. It is also possible that, to some extent, the differences in VEGF-induced signaling between Src compared to Fyn and Yes may be due to the differential localization of the Src family members. For example, perhaps Flt-l-Fyn/Yes complexes localize to a greater extent to caveolae fractions of cells, whereas KDR/Flk-1-Src complexes might be more generally distributed in the plasma membrane. This may allow for appropriate fine-tuning of differential signaling. Notably, there is strong evidence supporting a model that dually acylated (e.g., myristylated and palmitylated) Src family kinases, which include all Src family kinases with the exceptions of Src and Blk, have the potential to be localized to specific plasma membrane subdomains (such as caveolae). On the other hand, Src itself, which is only myristylated, has been reported to not localize in these microcompartments . Undoubtedly, post-translational modifications of proteins would play a very large role in regulating spatial and temporal specificity in signal transduction in these distinct microcompartments. Conclusions : Our experiments have revealed several important features of c-Src involvement in the early steps of VEGF-mediated signaling in HUVECs. Of primary importance was the finding that Src became associated preferentially with KDR/Flk-1, rather than with Flt-1, upon VEGF stimulation and VEGFR activation. This binding of Src to KDR/Flk-1 and subsequent activation of Src was enhanced by tyrosine phosphorylation of KDR/Flk-1. We also observed that, although Src binds preferentially to KDR/Flk-1, two other Src family members, Fyn and Yes, bind preferentially to the Flt-1 receptor upon stimulation of HUVECs with VEGF. These results are consistent with previous results that indicate that Fyn and Yes exhibited increased levels of phosphorylation upon VEGF stimulation in porcine aortic endothelial cells stably overexpressing Flt-1 . This difference in specificities may be a key mode of fine-tuning downstream signals capitalized on by endothelial cells. In this manner, it is possible that endothelial cells could appropriately gear their responses towards induction of vascular permeability and mitogenicity through signals transmitted by KDR/Flk-1 , or vascular remodeling and revascularization through signals transmitted by Flt-1 , to meet the specific need of a tissue depending on local requirements. Methods : Materials | Human recombinant VEGF-A165 and enolase were purchased from Sigma. Polyclonal alpha-KDR/Flk-1 (N-931) raised against amino acids 931 --997 of Flk-1, monoclonal alpha-KDR/Flk-1 (A-3) raised against amino acids 1158 --1345 of the carboxyl terminus of Flk-1, polyclonal alpha-Flt-1 (H-225) raised against amino acids 23 --247 mapping within the extracellular domain of Flt-1 of human origin, alpha-Fyn (FYN3), and alpha-Yes (3) were purchased from Santa Cruz Biotechnology. alpha-Src phosphotyrosine 419 antibody was from Cell Signaling. alpha-Src monoclonal antibodies were purified from hybridomas obtained from Quality Biotech (alpha-Src 2 --17) or from hybridomas provided by Dr. Joan Brugge (alpha-Src 327). alpha-phosphotyrosine (4G10) was provided by Dr. Stephen Robbins. Cell culture | Human umbilical vein endothelial cells (HUVECs) were provided by Dr. Kamala Patel (University of Calgary). These cells were isolated from human umbilical cord veins by collagenase treatment as described previously . HUVECs were maintained in M199 medium (Gibco) containing 10% fetal bovine serum, 2.2 mg/mL sodium bicarbonate, 2.4 mug/mL thymidine, 10 units/mL heparin, 2 mM glutamine, and antibiotics (100 units/mL penicillin, 100 mug/mL streptomycin) at 37C in 5% CO2. Immunoprecipitations | HUVECs were serum-starved for 24 h and then treated with VEGF (15 ng/mL) for 0, 15, 30, 60, or 120 min. Cells were washed two times with PBS on ice. The cells were then scraped in 1 mL of 1% NP-40 lysis buffer (150 mM NaCl, 50 mM Tris pH 7.5, 1% Nonidet P-40, 2 mM EDTA, 50 mug/mL leupeptin, 10 mug/mL aprotinin, 200 muM sodium orthovanadate, 4 mg/mL p-nitrophenyl phosphate). The scraped cells were briefly vortexed, and then incubated on ice for 10 min. Cell extracts were then clarified by centrifugation at 4C, 10 000 x g for 10 min. Cell extract samples for immunoprecipitation were precleared using 10 muL packed Protein A/G (1:1) and mixing on a rotator at 4C 30 min. Centrifugation at 14000 rpm in a cold microcentrifuge for 2 minutes was then conducted and the precleared cell extract supernatant was collected. Immunoprecipitations were performed using 15 mug alpha-Src 2 --17 per 1 mg cell extract, or 3 mug alpha-KDR or 3 mug alpha-Flt-1 per 1 mg extract, and 15 muL packed Protein G agarose. The immunoprecipitations were mixed on a rotator at 4C for 2 h. The immune complexes were washed 3 times with 1% NP-40 lysis buffer. Samples for Westerns were boiled for 5 min in 1X Laemmli's sample buffer, and resolved on 7% or 10% acrylamide SDS-PAGE gels. c-Src Kinase Activity Assays | HUVECs were either serum-starved for 24 h, or starved and then treated with VEGF (15 ng/mL). Cell lysates of these samples were then immunoprecipitated using alpha-KDR/Flk-1, alpha-Flt-1, or rabbit nonimmune IgG (rabbit IgG). The immune complexes were washed 3 times with 0.8 mL of 1 % NP-40 lysis buffer, and once with 0.8 mL Src kinase buffer (20 mM Tris pH 7.2, 5.0 mM MnCl2) and put on ice for 5 min. The pellets were then centrifuged and the supernatant buffer aspirated. Each kinase reaction was started by the addition of 33 muL Src kinase reaction hot mix (consisting of 30 muL of Src kinase buffer, 2 muL of 2.5 mug/mL enolase, and 1 muL (10 muCi) of [gamma-32P]ATP (>3000 Ci/mmol). The reaction was incubated at 30C for 20 min. The reactions were then stopped by the addition of 33 muL of 2X Laemmli's sample buffer and boiled for 5 min. Samples were resolved on 10% SDS-PAGE gels. The gels were treated with 1 M KOH at 55C for 1 h after electrophoresis to remove background due to serine phosphorylation . The gels were dried and exposed to autoradiographic film. Band intensities were quantified by densitometry of the autoradiogram. GST-Src domain fusion constructs and expression | The required Src domain cDNA fragments were generated by restriction digestion from pBA3CS, which contained full-length wild type human c-Src cDNA. pG-U contains the human c-Src unique domain (nt 1 --170, aa 1 --57) obtained by restriction digestion of pBA3CS-human c-Src using NcoI and NotI. pG-U32 contains human c-Src unique, SH3, and SH2 domains (nt 1 --775, aa 1 --258) obtained by restriction digestion using NcoI and EcoNI. pG-32 contains human c-Src SH3 and SH2 domains (nt 170 --775, aa 57 --258) obtained by restriction digestion using NotI and EcoNI. The Src cDNA fragments above were treated with Klenow to blunt their ends before ligation in frame into pGEX-2T. Each pGEX-Src fusion-containing E. coli BL21 clone was cultured in 10 mL 2x YTA broth (16 g/L tryptone, 10 g/L yeast extract, 5 g/L NaCl, 100 mug/mL ampicillin, pH 7.0) overnight at 30C, shaking at 250 rpm. The starter culture was diluted 1:100 into 400 mL of fresh 2xYTA broth and grown at 30C, shaking at 250 rpm, until A600 reached 0.5 --2. The culture was then induced by the addition of isopropyl-beta-D-thiogalactoside to a final concentration of 0.2 mM for 3 h at 30C, shaking at 250 rpm. GST-Src fusion protein harvesting and purification was carried out following pGEX vector supplier's instructions (Amersham Pharmacia Biotech). GST-Src domain fusion protein pull-downs of KDR/Flk-1 | HUVECs were starved in serum-free medium for 24 h and then treated with VEGF (15 ng/mL) for 0 or 15 min. Equimolar amounts of the various GST-Src domain fusion proteins (500 pmol) bound to Glutathione-agarose beads were used in pull-down incubations (at 4C for 3 hours) with equal amounts of cell extracts (500 mug). The fusion protein complexes were pelleted by centrifugation and washed three times with lysis buffer. The washed fusion protein pellets were boiled in 1X Laemmli's sample buffer, resolved on a 7% acrylamide SDS-PAGE gel, and transferred to nitrocellulose. Afterward, immunoblotting was performed and results were developed by using enhanced chemiluminescence (Amersham). Far western blotting | GST-Src domain fusion proteins were biotinylated at room temperature for 3 h with biotinamidocaproate-N-hydroxysuccinimide ester (100 mug/mL, Sigma) at a fusion protein concentration of 2 mg/mL in 0.1 M sodium borate, pH 8.8. Biotinylated protein was purified extensively by washes using PBS, pH 7.5, 2 mM dithiothreitol through a MICROSEP microconcentrator (Filtron Technology, according to manufacturer's instructions). The final biotinylated protein products were used at concentrations of 2 mug/mL in PBS, 1 mM dithiothreitol, in Far Western blotting of nitrocellulose membranes. Biotinylated probes were detected with horseradish peroxidase-conjugated ExtrAvidin (Sigma) at 1:2000 dilution in PBS, 3 % BSA, 0.05% Tween-20. Results were developed by reaction using enhanced chemiluminescence (Amersham). List of abbreviations : Flt-1, Fms-like tyrosine kinase-1; G32, GST-Src SH3 Src SH2 domain fusion protein; GST, glutathione-S-transferase; GU, GST-Src unique domain fusion protein; HUVEC, human umbilical vein endothelial cell; KDR/Flk-1, kinase insert domain-containing receptor/fetal liver kinase-1; pG, pGEX vector; VEGF, vascular endothelial growth factor. Authors' contributions : MC conducted the biochemical and molecular experiments and wrote the first draft of the manuscript. JW participated in making the GST-Src domain fusion constructs. DF conceived of the study, participated in its design, supervised the work of MC and JW, and oversaw subsequent drafting and editing of the manuscript. Backmatter: PMID- 12226484 TI - Inositol Phospholipid Metabolism in Arabidopsis. Characterized and Putative Isoforms of Inositol Phospholipid Kinase and Phosphoinositide-Specific Phospholipase C AB - Phosphoinositides (PIs) constitute a minor fraction of total cellular lipids in all eukaryotic cells. They fulfill many important functions through interaction with a wide range of cellular proteins. Members of distinct inositol lipid kinase families catalyze the synthesis of these phospholipids from phosphatidylinositol. The hydrolysis of PIs involves phosphatases and isoforms of PI-specific phospholipase C. Although our knowledge of the roles played by plant PIs is clearly limited at present, there is no doubt that they are involved in many physiological processes during plant growth and development. In this review, we concentrate on inositol lipid-metabolizing enzymes from the model plant Arabidopsis for which biochemical characterization data are available, namely the inositol lipid kinases and PI-specific phospholipase Cs. The biochemical properties and structure of characterized and genome-predicted isoforms are presented and compared with those of the animal enzymes to show that the plant enzymes have some features clearly unique to this kingdom. Keywords: Introduction : Phosphatidylinositol (PtdIns) is a major phospholipid in eukaryotic cells. Three of the five free hydroxyl groups of PtdIns can be phosphorylated in cells in different combinations. In total, seven phosphorylated derivatives of PtdIns have been detected, one of which, PtdIns 3,4,5-trisphosphate [PtdIns(3,4,5)P3], has not been found in plant cells . These inositol phospholipids are collectively referred to as phosphoinositides (PIs). In animal cells, PIs and their derivatives operate in signal transduction pathways triggered by stimuli as diverse as growth factors, hormones, neurotransmitters, and light . Figure 1 | PI metabolism. PI metabolism. The different steps in the synthesis of PIs and the lipid kinases catalyzing the different reactions are indicated. PtdIns(3,4,5)P3 is present in animal cells but has not been detected in plant tissues, so far. In animal cells, PtdIns(3,4)P2 can be generated from PtdIns4P by a PtdIns 3-kinase or by an as-yet-unidentified PIPkin from PtdIns3P. Plant cells do not contain any homolog of the heterodimeric inositol lipid 3-kinases that are able to phosphorylate PtdIns4P to PtdIns(3,4)P2 and PtdIns(4,5)P2 to PtdIns(3,4,5)P3. PtdIns(4,5)P2 can be synthesized by type I and type II PIPkins from PtdIns4P and PtdIns5P, respectively. On the basis of sequence comparison, plants cells do not possess type II PIPkins. PtdIns5P is present in plants, but an enzyme capable of producing it has not been identified. Historically, the first major insight into the importance of PIs was the discovery that the two PtdIns 4,5-bisphosphate [PtdIns(4,5)P2]-derived second messengers inositol 1,4,5-trisphosphate [Ins(1,4,5)P3] and diacylglycerol promote Ca2+ release from internal stores and activate protein kinase C, respectively (; for review, see ). The production of these two second messengers from PtdIns(4,5)P2 is catalyzed by PI-specific phospholipase C (PI-PLC) isoforms. During the last decade, it has become evident that in addition to serving as precursors to Ins(1,4,5)P3 and diacylglycerol, PIs actively participate in several other cellular processes. They have been shown to regulate the dynamics of the actin cytoskeleton through interaction with actin-binding proteins , and to potentiate the activation of protein kinase C , PI-PLC , and phospholipase D . In addition, PIs phosphorylated at the D3-hydroxy group of the inositol head group are required for specific vesicle trafficking steps and are able to activate the protein kinase Akt/PKB and PI-dependent kinases . A 3-phosphorylated PI, PtdIns 3,5-bisphosphate [PtdIns(3,5)P2], was recently identified and shown to accumulate in yeast cells subjected to hyperosmotic or NaCl stress . In addition, mutations in many of the proteins involved in the PI system cause various diseases or severe defects . The PIs clearly constitute a group of lipids with very important, diverse functions. It is therefore critical that the levels of these phospholipids are tightly regulated. The lipid kinases catalyzing the synthesis of the phosphorylated derivatives of PtdIns, and isoforms of PI-PLC have been extensively studied and characterized; some of them have also been crystallized and their structure determined. In comparison with the wealth of information available for the components, regulation, and function of the animal PI system, and despite the identification of plant homologs of many of these animal components, our knowledge of the plant PI system is, at present, quite limited. However, it has been demonstrated that micro-injected "caged" Ins(1,4,5)P3 can release Ca2+ from internal stores and is also able to trigger stomatal closure . There is also evidence that PIs may participate in the regulation of cytoskeletal structures in plant cells and in polar pollen tube growth . A number of reports also suggest that a wide range of signals, such as light, hormones, and stress, may mediate their effect through PI-dependent processes , although the effects reported were often limited and the identity of the lipid or inositol phosphate species affected not always clearly demonstrated. More recently, hyperosmotic stress was shown to induce an increase in two distinct PtdIns bisphosphate (PtdInsP2) isomers, PtdIns(4,5)P2 in Arabidopsis cells and PtdIns(3,5)P2 in Chlamydomonas moewusii and in some higher plant cells . It is also now clear that, as in other organisms, vesicle trafficking in plant cells is dependent on PIs . A PtdIns 3-kinase was recently found associated with nuclear transcriptionally active sites, suggesting a potential role for this lipid kinase in the regulation of transcription in the nucleus . It is also interesting that chloroplasts contain both PtdIns 3- and 4-kinase activities . Additional evidence that a functional PI system operates in plant cells has been provided by the identification of homologs of the components of the animal PI system, such as PI-metabolizing enzymes , PI-regulated enzymes , and PIs themselves . Of the seven phosphorylated derivatives of PtdIns identified in animal systems, only PtdIns(3,4,5)P3 has never been observed in plant cells (or in yeast). The presence of PtdIns5P in higher plants has only recently been unambiguously demonstrated . Previous reviews on the metabolism and function of PIs in plant cells have provided overviews of some aspects of this field but were published before the availability of a complete plant genome sequence and lack a thorough description of some of the enzymes involved in the metabolism of these lipids . The aim of the present review is to provide a detailed summary of the information available on the enzymes involved in the synthesis and hydrolysis of PIs, i.e. the PI kinases and PI-PLCs. It includes a description of the structure, and of the molecular and biochemical properties of these enzymes, with new information gleaned from sequence comparison and analysis. When possible, in vivo functions are also discussed. PtdIns KINASES : As shown in Figure , PtdIns can be phosphorylated on three of the five free hydroxyl groups of the inositol ring in reactions catalyzed by PtdIns kinases. The enzymes catalyzing the phosphorylation of the hydroxyl groups at positions 3 and 4 have been identified and characterized. There is an animal enzyme capable of generating PtdIns5P from PtdIns in vitro ; however, whether this enzyme is responsible for the synthesis of PtdIns5P in vivo is not known. PtdIns 3-Kinase | Multiple PtdIns 3-kinase isoforms have been identified. On the basis of their structure and substrate specificity, they can be divided into three classes . All PtdIns 3-kinases share three conserved domains: a catalytic domain preceded by a helical domain and a C2 domain (a Ca2+-dependent or -independent lipid-binding domain originally identified in protein kinase C; ; Fig. ). Class I PtdIns 3-kinases are heterodimeric enzymes that can phosphorylate PtdIns, PtdIns 4-phosphate (PtdIns4P), and PtdIns(4,5)P2 in vitro, but preferentially phosphorylate PtdIns(4,5)P2 in vivo. These PtdIns 3-kinases are subdivided into class IA and class IB. The catalytic subunit of class IA enzymes associates with one of three different adaptor proteins, whereas the class IB catalytic subunit (only one has been identified) associates with a p101 adaptor protein. Class II PtdIns 3-kinases preferentially use PtdIns and PtdIns4P as substrates. Class I and II PtdIns 3-kinases possess a ras-binding domain at their N terminus, and class II enzymes are characterized by the presence of a Phox homology (PX) domain and a second C2 domain at their carboxy terminus (see Fig. ). Finally, class III PtdIns 3-kinases include the only PtdIns 3-kinase identified in Brewer's yeast (Saccharomyces cerevisiae), ScVps34p, and all of its orthologs from other organisms. Class III PtdIns 3-kinases phosphorylate only PtdIns and are believed to represent the ancestral form of PtdIns 3-kinase. The Vps34 protein exists as a complex with a Ser/Thr protein kinase both in Brewer's yeast and animals . This protein kinase, encoded by the VPS15 gene in Brewer's yeast, phosphorylates and activates Vps34. A putative ortholog of this protein kinase is also present in Arabidopsis (At4g29380), suggesting that some of the processes regulating protein trafficking may be conserved in animals, yeast, and plants. Figure 2 | Domain structure representation of PtdIns 3-kinases. Domain structure representation of PtdIns 3-kinases. A, The domains conserved in the catalytic subunits of the three PtdIns 3-kinase classes are represented by colored boxes. B, Vps34 proteins from Brewer's yeast (Vps34p; accession no. ), Arabidopsis (AtVps34; accession no. ), and human (HsVps34; accession no. ) were aligned using the ClustalW program from the MacVector package. Identical, conserved, and semiconserved residues are indicated below the alignment by *, :, and ., respectively. // indicates a large gap in the sequence. The only PtdIns 3-kinase identified so far in Arabidopsis, AtVps34, is a homolog of ScVps34p. The VPS34 gene was identified in a screen for Brewer's yeast mutants altered in the sorting of vacuolar proteins , and was shown to encode a PtdIns kinase that phosphorylates PtdIns exclusively at the D-3 position of the inositol ring . Although the substrate specificity of AtVps34 has not been fully determined, it is reasonable to assume that like ScVps34p, it can only phosphorylate PtdIns, not PtdIns4P and PtdIns(4,5)P2. In both Brewer's yeast and fission yeast (Schizosaccharomyces pombe), the VPS34 gene is not essential but is necessary for normal growth . Point mutations in the catalytic domain of Vps34p or deletion of the VPS34 gene lead to dramatic decreases in cellular levels of PtdIns 3-phosphate (PtdIns3P) in Brewer's yeast and severe defects in vacuolar protein sorting . Down-regulation of the expression of the AtVPS34 gene in transgenic Arabidopsis plants resulted in severe inhibition of growth and development , indicating that AtVps34 is required for normal plant development. The Arabidopsis AtVPS34 cDNA was unable to complement a VPS34 yeast deletion mutant. However, a chimeric gene consisting of Brewer's yeast VPS34 in which the region coding for the catalytic domain had been replaced by the corresponding region of the Arabidopsis gene could complement the yeast mutant. This also suggests that regions other than the catalytic domain are required for the function of Vps34p in vivo, and that the regulation mechanisms of Vps34 enzymes differ between species. A distinctive feature of Vps34 proteins versus other PtdIns 3-kinases is the presence of a longer insert between the amino-terminal C2 domain and the helical domains, and of an extended insert (approximately 60 amino acid residues) in the helical domain. The latter insert is not present in AtVps34 (Fig. B). The two most commonly used PtdIns 3-kinase inhibitors are wortmannin (a fungal metabolite) and LY294002 (a synthetic molecule; ). In vitro, their IC50 values are around 5 nm for wortmannin and 1 mum for LY294002 . Surprisingly, ScVps34p is much less sensitive to these compounds (IC50 of 3 mum for wortmannin and 50 mum for LY294002; ), whereas the human Vps34 ortholog exhibits sensitivities similar to those of the most sensitive PtdIns 3-kinases. In addition, wortmannin inhibits other enzymes including PtdIns 4-kinases (e.g. ; ) and PtdIns phosphate (PtdInsP) kinases (PIPkins; ), but with IC50 values 20- to 1,000-fold higher than for PtdIns 3-kinase inhibition. Fifty micromolar LY294002 does not inhibit a range of animal protein kinases . However other enzymes, such as the PtdIns 3-kinase-related protein kinases TOR and DNA-dependent protein kinase, and to a lesser extent casein kinase-2, are sensitive to LY294002 . Casein kinase-2 is not inhibited by wortmannin . One major problem with using wortmannin is that it is rather unstable in solution, and incubation times longer than a couple of hours should be avoided . In summary, in animal systems, wortmannin and LY294002 are, at low concentrations (up to 1 mum for wortmannin and up to 50 mum for LY294002), specific inhibitors of PtdIns 3-kinase; but at higher concentrations, they lose specificity. Wortmannin has been shown to inhibit PtdIns 3-kinase activity in crude tobacco extracts, with calculated IC50 values of 98 nm in vitro and approximately 1 mum in vivo . In this plant system, wortmannin also inhibited the sorting of one type of protein to the vacuole, with an IC50 of approximately 7 mum, a value close to the IC50 of 9 mum for the synthesis of all phospholipids examined, except PtdIns3P . The effects of wortmannin on vacuolar sorting in these studies are difficult to interpret in terms of the involvement of PIs and/or other phospholipids because extremely high concentrations of wortmannin and long incubation times were used, and the specificity of wortmannin has not been tested with plant enzymes. Before using wortmannin and LY294002 to assess the possible involvement of PtdIns3P in physiological processes in plant systems, it must first be established that they specifically inhibit PtdIns 3-kinase and do not significantly affect other enzymes in the same concentration range. Wortmannin was demonstrated to covalently bind to a Lys residue in the ATP-binding site of a class I PtdIns 3-kinase, inducing a conformational change in the catalytic domain of the enzyme . This Lys residue is conserved in all PtdIns 3-kinases. The difference in sensitivity to wortmannin of the different PtdIns 3-kinases is believed to be attributable to differences in amino acid residues surrounding the conserved Lys residue in the ATP-binding site of these isoforms. However, it remains unknown which residues determine wortmannin sensitivity. As the sole PtdIns 3-kinase present in Arabidopsis, AtVps34 is probably responsible for the synthesis of the majority of the PtdIns3P found in plants. Class I and II PtdIns 3-kinases are activated by G-protein-coupled receptors and Tyr kinase receptors, which are represented by a large number of distinct members in animal cells. Plants, on the other hand, possess few putative G-protein-coupled receptors and no receptor Tyr kinases. It is therefore not surprising that class I and II PtdIns 3-kinases are absent from plants. recently showed that the homolog of AtVps34 is associated with active transcription sites in soybean (Glycine max) cell nucleoli. As in yeast, the product of Vps34 activity, PtdIns3P, participates in the trafficking of proteins to the plant vacuole . In that study, Kim et al. also demonstrated that PtdIns3P is present in several compartments, including the Golgi apparatus and vacuole networks. PtdIns3P-dependent trafficking involves proteins with a FYVE domain (named after four proteins that it has been found in: Fab1, YOTB/ZK632.12, Vac1, and EEA1; ), a domain that confers highly specific binding to PtdIns3P . In Arabidopsis, more than 10 proteins contain a FYVE domain (see the SMART database: ), only one of which has been characterized . The Brewer's yeast genome encodes fewer than 10 FYVE domain-containing proteins; the human genome encodes more than 50. Two of the plant FYVE proteins probably represent putative homologs of the PtdIns3P 5-kinase protein Fab1p from Brewer's yeast. Several PX domains, including all of the PX domains present in Brewer's yeast, have been shown to bind PtdIns3P . Arabidopsis contains up to nine putative proteins with a PX domain. Fifteen have been identified in Brewer's yeast, whereas humans have more than 50 (SMART database). This indicates that the functions of PtdIns 3-kinases are much more diverse in animals than in plants and yeast. PtdIns 4-Kinases | Biochemical Characterization | PtdIns 4-kinases catalyze the phosphorylation of PtdIns to PtdIns4P [a lipid believed to be the major precursor of PtdIns(4,5)P2], and therefore represents a potentially crucial point in the regulation of the PI-dependent pathways. Two major types of PtdIns 4-kinase, II and III, differing in size and sensitivity to detergents and adenosine, have been identified in a wide range of tissues and cellular compartments in animals . Type II PtdIns 4-kinase is a membrane-bound, 55-kD enzyme that is readily renaturable after SDS-PAGE. It has comparatively low Km values for PtdIns and ATP, is strongly inhibited by adenosine and Ca2+ , and is insensitive to wortmannin . Two forms of type III PtdIns 4-kinase have been detected in membrane and soluble fractions of animal tissues. One is 110 kD in size, and the other is 200 to 230 kD. Type III enzymes have 3- to 7-fold higher Km values for PtdIns and ATP than the type II enzyme and are not inhibited by adenosine and Ca2+. In addition, type III enzymes are inhibited by wortmannin (at concentrations significantly higher than those inhibiting PtdIns 3-kinases), whereas type II enzymes are insensitive . PtdIns 4-kinase activity is present in many different compartments in plant cells, and has been partially purified from the soluble fraction of the unicellular alga Dunaliella parva and from carrot (Daucus carota) cells . It has also been found in plasma membranes from Catharanthus roseus and spinach (Spinacia oleracea; ). The size of the partially purified plant PtdIns 4-kinases was estimated to be 80 kD for the soluble activity from carrot cells, 500 kD for the activities purified from D. parva and C. roseus, and 65 and 125 kD for the two distinct activities purified from spinach plasma membranes. A soluble 49-kD protein, PIK-A49, was able to stimulate a PtdIns 4-kinase activity solubilized from carrot plasma membranes . PIK-A49 was shown to bind and bundle actin and was identified as an elongation factor 1 . Stimulation of PtdIns 4-kinase was dependent on the phosphorylation status of PIK-A49 . Using antibodies raised against a recombinant PtdIns 4-kinase, were able to immunoaffinity purify an active PtdIns 4-kinase from Arabidopsis microsomes. Wortmannin was recently shown to inhibit the two PtdIns 4-kinase activities partially purified from spinach plasma membranes . Genes, Structure, and Function | cDNAs encoding functional PtdIns 4-kinases have been isolated from animals, Brewer's yeast, and Arabidopsis, using either peptide sequences from purified proteins or homologous cloning using previously available sequences. Genes encoding type III enzymes have been grouped into two distinct subfamilies, alpha and beta, based on sequence and structure similarities . Subfamily beta is represented by proteins of 68 to 122 kD, and subfamily alpha by proteins of 200 to 230 kD (with the exception of human PI4Kalpha, which is a 97-kD protein ). Several domains are conserved in these proteins, including a catalytic domain of about 230 amino acid residues within the C-terminal part of these proteins, and a helical domain (previously referred to as PIKa or LKU domain), the location of which varies among the different isoforms . Members of the alpha subfamily all contain a pleckstrin homology (PH) domain that separates the helical and catalytic domains in these proteins. Another conserved domain, novel homology (NH) is conserved in all of the isoforms belonging to the beta subfamily (Fig. ; ). NH domains were first identified in yeast Pik1p and in a soluble PtdIns 4-kinase from rat . Figure 3 | Domain structure representation of animal, yeast, and plant PtdIns 4-kinases. Domain structure representation of animal, yeast, and plant PtdIns 4-kinases. The various conserved domains are represented by colored boxes. The catalytic domains of type II and III PtdIns 4-kinase share no homology and are therefore shown in different colors. Some plant type II PtdIns 4-kinases contain one or two ubiquitin-like domains. Although PtdIns 4-kinase activity was first detected, characterized, and partially purified from higher plants many years ago, the isolation and functional expression of a full-length plant PtdIns 4-kinase cDNA was first reported in 1999 . The corresponding 126-kD protein from Arabidopsis, AtPI4Kbeta1 (previously AtPI4Kbeta), is similar in size and overall primary structure to the PtdIns 4-kinase beta isoforms from yeast and animals. Another cDNA from Arabidopsis, AtPI4Kalpha1 (previously AtPI4Kalpha), encoding a PtdIns 4-kinase has been described . The deduced amino acid sequence corresponding to this partial cDNA comprised a helical, a PH, and a catalytic domain . The gene corresponding to this partial cDNA has now been identified and deposited in the GenBank/EMBL databases . The predicted protein is 227 kD in size and structurally similar to the alpha-isoforms from animal and yeast . No peptide sequences for any of the partially purified plant PtdIns 4-kinases are available, making it impossible to verify whether any of them correspond to one of the Arabidopsis PtdIns 4-kinase genes. The subcellular localization of these two Arabidopsis type III PtdIns 4-kinases is at present not known, but AtPI4Kalpha1 was shown to be present in a microsomal fraction and an actin-enriched fraction . Table I | List of characterized and predicted PtdIns 4-kinases from Arabidopsis AtPI4Kbeta1 possesses the two conserved domains present in all type III PtdIns 4-kinases, namely the helical and catalytic domains, but lacks a PH domain, features typical of PtdIns 4-kinase alpha isoforms. A conserved domain previously only identified in the yeast PIK1 protein and a rat soluble PtdIns 4-kinase is also conserved in AtPI4Kbeta1 (domain NH in Fig. ). Surprisingly, AtPI4Kbeta1 is the only known PtdIns 4-kinase to possess a unique repetitive motif constituted of 11 repeats of a charged core unit . No function has yet been assigned to the conserved helical, NH, and repetitive motif domains. Because PtdIns 4-kinase activities have been detected in most cellular compartments, it is possible that the domains identified in PtdIns 4-kinases are involved in the targeting of these enzymes to various cellular compartments through interactions with other proteins or lipids. The PH domain of the carrot homolog of AtPI4Kalpha1 was shown, using a lipid-protein overlay, to bind PtdIns4P but not PtdIns in vitro . It, thus, appears likely that the PH domain of type alpha PtdIns 4-kinases is not responsible for binding the lipid substrate. It is consequently possible that substrate binding in PtdIns 4-kinases is controlled by the helical domain, or a domain conserved structurally but not at the sequence level. As recently observed with other PtdIns 4-kinases, AtPI4Kbeta1 is inhibited by wortmannin at concentrations similar to those affecting the yeast and mammalian enzymes ---concentrations that are significantly higher than those required to inhibit PtdIns 3-kinases. The residue corresponding to the Lys residue of PI3Kalpha at which wortmannin binds covalently is conserved in AtPI4Kbeta1 as it is in all other type III PtdIns 4-kinases, including AtPI4Kalpha1. Two PtdIns 4-kinase genes, PIK1 and STT4, are present in the genome of Brewer's yeast, both of which are essential for yeast viability . The PtdIns 4-kinases corresponding to these two genes are responsible for the synthesis of distinct pools of PtdIns4P . Using temperature-conditional mutants, the authors demonstrated that Pik1p is essential for normal secretion, Golgi and vacuole membrane dynamics, and endocytosis, and that Stt4p is required for maintenance of vacuole morphology, cell wall integrity, and actin cytoskeleton organization. Intriguingly, Arabidopsis is, so far, the only organism found to possess two genes encoding type IIIbeta PtdIns 4-kinases . The two corresponding protein sequences share 80% identity, and AtPI4Kbeta2 also possesses the NH domain and the repetitive motif. The genome of Arabidopsis contains large segments of duplicated DNA that can be grouped into 103 blocks . The two AtPI4Kbeta genes do not belong to any of the duplicated blocks present in the genome, but they may have occurred from a duplication event not identified by . Arabidopsis also contains a gene encoding a putative protein, AtPI4Kalpha2, with significant similarity to AtPI4Kalpha1 . However, this putative protein is much smaller than AtPI4Kalpha1 and consists of a catalytic domain and a region corresponding to the 114 first amino acid residues of AtPI4Kalpha1. AtPI4Kalpha2 does not contain any other domain, i.e. the PH and helical domains are lacking. No expressed sequence tag (EST) corresponding to this gene is present in the databases, suggesting that it may not be expressed. The conservation of the structure of PtdIns 4-kinases in the different phyla contrasts markedly with the current data available for PI-PLC. On the basis of protein structure and biochemical properties, all plant PI-PLC isoforms appear to belong to a single family closely related to the delta-type of mammalian PI-PLCs, but lacking a PH domain. The identification of genes encoding type II PtdIns 4-kinases was achieved only recently . The corresponding proteins from rat and human form a novel class of PtdIns 4-kinases that is unrelated to previously identified inositol lipid kinases . Potential homologs of these animal type II PtdIns 4-kinases are present in yeast, fruitfly (Drosophila melanogaster), and Arabidopsis. Surprisingly perhaps, there is only one gene in Brewer's yeast, but search of databases with the entire sequence of human PI4K IIalpha identified eight putative genes (AtPI4Kgamma1 --8) in Arabidopsis . The predicted Arabidopsis proteins show low overall identity with human PI4K IIalpha, but such is also the case for the putative type II PtdIns 4-kinase from Brewer's yeast (28% identical and 10% similar residues). However, identified several candidate kinase motifs in type II PtdIns 4-kinases in more than 10 different putative type II PtdIns 4-kinases. Alignment of the Arabidopsis protein sequences with human PI4K IIalpha and the sequence from Brewer's yeast clearly shows that the kinase motifs identified in the animal proteins, which correspond to residues found in subdomains I, II, VIb, VII of protein kinases , are also present in the Arabidopsis sequences (Fig. A). Although, as noted by , there is no close similarity between type II PtdIns 4-kinases and other PtdIns kinases, a number of residues conserved in the candidate catalytic domain of type II PtdIns 4-kinases are also conserved in other PtdIns kinases and/or PIPkins. The GXXG motif of the candidate subdomain I is found in PIPkins , including all Arabidopsis PIPkins, except AtPIPK11 (see Fig. ). The FK motif of the candidate subdomain II, the DRH motif of the candidate subdomain VIb of the Arabidopsis sequences, and the DXG motif of the candidate subdomain VII are all present in PtdIns 3-kinases and PtdIns 4-kinases . These observations strongly suggest that the Arabidopsis putative type II PtdIns 4-kinases may well be functional homologs of the animal enzymes. However, none of the Arabidopsis genes have yet been shown to encode functional PtdIns 4-kinases. Most of the Arabidopsis putative type II PtdIns 4-kinases are significantly larger than the animal enzymes: Six of them are 61 kD or greater in size , instead of 55 kD for the animal enzymes. This is because of the presence of an N-terminal extension in the plant enzymes. Sequence analysis reveals that six of the putative plant type II PtdIns 4-kinases contain one or two ubiquitin domains (Fig. B). No ubiquitin domain is present in non-plant type II PtdIns 4-kinases. Figure 4 | Comparison of putative Arabidopsis type II PtdIns 4-kinases with the catalytic domain of type II PtdIns 4-kinases from human and Brewer's yeast (A) and with ubiquitin (B). Comparison of putative Arabidopsis type II PtdIns 4-kinases with the catalytic domain of type II PtdIns 4-kinases from human and Brewer's yeast (A) and with ubiquitin (B). Sequences were aligned with the ClustalW program and adjusted by hand. Identical, conserved, and semiconserved residues are indicated below the alignment by *, :, and ., respectively. A, Alignment of predicted catalytic domains. The location of candidate kinase subdomains I, II, VIb, and VII, identified by , are indicated in roman numerals above the alignment. A candidate subdomain VIII is also indicated. The kinase motifs in subdomains I, II, VIb, and VII identified by are boxed. The two P residues in the PXXXXP motif in the candidate subdomain VIII are also boxed. B, Alignment of the N-terminal domains of six of the Arabidopsis putative type II PtdIns 4-kinases with ubiquitin 7 from Arabidopsis (AtUBQ7; accession no. NM_129118). AtUBQ7, AtPI4Kgamma2, AtPI4Kgamma3, and AtPI4Kgamma4 contain two ubiquitin domains, whereas AtPI4Kgamma5, AtPI4Kgamma6, and AtPI4Kgamma7 contain only one. Amino acid residues identical or conserved at identical positions in the two ubiquitin domains of AtUBQ7 and the ubiquitin-like domains of Arabidopsis type II PtdIns 4-kinases are boxed. PtdInsP KINASES : The synthesis of the three PtdInsP2 isomers from PtdInsP isomers is catalyzed by PIPkins (see Fig. ). These enzymes form a family of lipid kinases distinct from PtdIns 3-kinases and PtdIns 4-kinases, and are classified into three subfamilies (type I --III; Table ), depending on their substrate specificity . Type I enzymes are PtdIns4P 5-kinase, and type II enzymes are PtdIns5P 4-kinase . In vitro, type I PIPkins can make PtdIns5P from PtdIns, and PtdIns(3,5)P2 from PtdIns3P , and type II PIPkins can synthesize PtdIns 3,4-bisphosphate [PtdIns(3,4)P2] from PtdIns3P . In yeast, Mss4p is the only type I PIPkin identified and no type II members appear to be present. Last, type III enzymes are PtdIns3P 5-kinases and are represented by the Brewer's yeast FAB1 gene product and its homologs from other species . In vitro, type III PIPkins can make PtdIns5P from PtdIns . Table II | List of characterized and predicted PIPkins from Arabidopsis All three PtdInsP2 isomers are present in plant cells, but their relative abundance varies depending on cell type and/or growth conditions , suggesting that the lipid kinases involved in the synthesis of these lipids and/or the transcription of genes encoding these enzymes are tightly regulated. A plant enzyme capable of phosphorylating PtdIns4P to a PtdInsP2 isomer has long been known to be present in the plasma membrane , and both PtdIns(3,4)P2 and PtdIns(3,5)P2 can be generated from endogenous PIs when plant plasma membranes are incubated with ATP (S.K. Dove and B.K. Drobak, personal communication). In animal cells, PtdIns(3,4)P2 synthesis can proceed via two distinct pathways, involving either 3-phosphorylation of PtdIns4P by heterodimeric inositol lipid kinases or a new PIPkin capable of phosphorylating PtdIns3P on the 4-OH group . As mentioned previously, the only PtdIns 3-kinase present in plants is a homolog of the yeast Vps34p protein, which only phosphorylates PtdIns. This suggests that a distinct enzyme is responsible for the synthesis of PtdIns(3,4)P2 in plant cells. The first gene encoding a functional plant PIPkin was recently identified and characterized in more detail . When expressed as a recombinant protein in Escherichia coli, AtPIP5K1 was able to phosphorylate both PtdIns3P and PtdIns4P to generate PtdIns(3,4)P2 and PtdIns(4,5)P2, respectively . When expressed in insect cells, the enzyme was shown to preferentially stimulate the synthesis of PtdIns(4,5)P2 and, surprisingly, PtdIns(3,4,5)P3 but not PtdIns(3,4)P2 via 5-phosphorylation of endogenous precursors . It is difficult to explain the discrepancy in the lipids generated by the same enzyme in the two heterologous systems. Regulation of AtPIP5K1 may include phosphorylation of one or several residues, because its activity was inhibited upon phosphorylation by cAMP-dependent protein kinase and because a soluble protein kinase activity from Arabidopsis plants was able to phosphorylate recombinant AtPIP5K1 . Which PIs are generated by AtPIP5K1 in vivo is at present not known, but the enzyme must be responsible for the production of part of the PtdIns(4,5)P2 pool in cells where it is present. Furthermore, AtPIP5K1 is believed to play an important function in the response of plants to drought and salt stress because its expression is low in non-stressed cells and induced upon stress . A confirmation of the probable implication of a PtdIns4P 5-kinase in the response of plant cells to osmotic and salt stress was provided by the clear demonstration that the level of PtdIns(4,5)P2 increases dramatically within a few minutes of subjecting Arabidopsis cells to such stress . Type I/II AtPIPkins: Genes, Structure, and Function | The Arabidopsis genome contains a number of genes encoding proteins with significant similarity to AtPIP5K1, all of which have a PIPkin catalytic domain (Table ; Fig. ). These additional putative AtPIPkins have not yet been characterized. Fifteen distinct Arabidopsis genes encode proteins with a PIPkin catalytic domain. Four of these genes encode putative homologs of the Fab1p protein from Brewer's yeast and have therefore been classified as type III PIPkins. The remaining 11 genes code for proteins that, based on sequence comparison alone, cannot be assigned to either type I or type II PIPkins (see below). They can, however, be classified into two distinct subfamilies, designated A and B in Table based on differences in overall structure (see Fig. ). Subfamily A is represented by two members, whereas subfamily B is represented by nine members including AtPIP5K1 . Figure 5 | Domain structure representation of animal, yeast, and plant type I/II PIPkins. Domain structure representation of animal, yeast, and plant type I/II PIPkins. The conserved domains are indicated by boxes of different colors. Two domains, the dimerization and catalytic domains, are conserved in all type I and II PIPkins (Figs. and ; ). The PIPkin catalytic domain is highly conserved in the 11 type I/II AtPIPkins, and like in all other type I and type II PIPkins, this domain contains a highly variable insert. This insert is significantly longer in AtPIPkins than in non-plant enzymes . The function of this insert in PIPkins is not understood. In addition, a significant number of residues are specifically conserved in the catalytic domains of the Arabidopsis proteins . Most of the residues in HsPIPKIIbeta proposed to participate in binding ATP and the lipid substrate are conserved in AtPIPkins . Figure 6 | Alignment of the N-terminal MORN domain of type I/II PIPkins from Arabidopsis. Alignment of the N-terminal MORN domain of type I/II PIPkins from Arabidopsis. The alignment was generated with the ClustalW program and adjusted by hand. Identical, conserved, and semiconserved residues are indicated below the alignment by *, :, and ., respectively. The eight MORN motifs (consensus sequence: Y-Q/E-G-E/Q-T-X-N-G-K-X-H-G-Y-G) are indicated by black lines on top of the alignment. Residues within the MORN consensus sequence that are conserved in only a few of the sequences are boxed. Figure 7 | Comparison of the catalytic domain of type I and II PIPkins from animal, yeast, and Arabidopsis. Comparison of the catalytic domain of type I and II PIPkins from animal, yeast, and Arabidopsis. The PIPkin catalytic domain of the nine Arabidopsis PIPkins, Brewer's yeast Mss4p and human PIPKIbeta and PIPKIIbeta were aligned using the ClustalW program. Arabidopsis subfamily A sequences are in italics. Residues identical in all sequences are indicated by *. The residues conserved or semiconserved are marked with : and ., respectively. Residues identical or conserved in Arabidopsis sequences only are boxed and have gray backgrounds. Residues identical in AtPIPK10 and AtPIPK11 only, and not conserved in other proteins, are boxed. The activation loop is indicated. Residues in HsPIPKIIb proposed to interact with ATP and PtdIns5P are indicated with circle and , respectively . Residues in the activation loop that are conserved among type I or II PIPkins are shaded in dark and light gray, respectively, only when they are also conserved in Arabidopsis sequences. The position of the first alpha-helix of the catalytic domain of human PIPKIIbeta is shown under the sequence alignment, and the location of the end of the N-terminal dimerization domain is indicated by an arrow. The inserts in the catalytic domains are marked by a dotted line on top of the alignment. The GXXG motif conserved in the catalytic domain of protein kinases and type II PtdIns 4-kinases is indicated by a black bar over the alignment. Subfamily A AtPIPkins are composed of the PIPkin catalytic domain and of a shorter dimerization domain in which the first alpha-helix appears to be missing, as compared with other type I and II PIPkins (Figs. and ). In addition, a number of residues are identically conserved in the catalytic domain of the two class A AtPIPkins but not in any other type I or type II PIPkins . Whether these two putative AtPIPkins are expressed is at present not known, because no mRNA or EST sequences have been reported for any of them. Subfamily B AtPIPkins are composed of the conserved PIPkin catalytic domain and a dimerization domain. In addition, they all have a third conserved domain at their N-terminal end (Figs. and ). A comparison of the nine subfamily B AtPIPkins shows that six of them form three pairs of two closely related proteins (Figs. and ). An identical tree is obtained if the catalytic domains only are compared. For each of the three AtPIPkin pairs, the two corresponding genes appear to have occurred from a duplication event because each of them is found in one of the two regions forming one duplicated block of the Arabidopsis genome, whereas the other gene belongs to the other region of the same duplicated block. Two of the PIPkin pairs belong to two duplicated blocks (8 and 62, respectively) that have been assigned to the same age class, i.e. the two blocks were duplicated at the same time, approximately 50 million years ago . The third pair of duplicated AtPIPkin genes belongs to an older age class, which occurred approximately 100 million years ago (block 17). Surprisingly, the two AtPIPkin genes belonging to block 8 show a significantly lower degree of identity (77%) than the two AtPIPkin genes belonging to block 62 (87%). This suggests that distinct physiological functions are fulfilled by the two pairs of AtPIPkins. It is also interesting to note that AtPIP5K6 and AtPIP5K3 are very closely related to the two pairs from block 62 and block 17, respectively . Subfamily B AtPIPkins are characterized by the presence of a conserved domain at their N terminus, preceding the dimerization domain (Figs. and ). This domain is found in no other PIPkin from any organism and consists of eight repeats of a conserved 23-residue long motif. The presence of a repeated motif was noticed in the first PIPkin sequence (AtPIP5K1) obtained from Arabidopsis , but at that time, no other protein with this domain had been identified. The repeated motif was subsequently identified as the membrane occupation and recognition nexus (MORN) motif by the SMART sequence analysis program . Figure 8 | A phylogenic tree of subfamily B AtPIPkins. A phylogenic tree of subfamily B AtPIPkins. Full-length protein sequences were aligned and the tree was constructed using the PAUP software (Phylogenetic Analysis Using Parsimony, version 4.0b4, Sinauer Associates, Sunderland, MA) from 12,000 replicates. An identical tree is obtained if the sequences of the catalytic domains alone are compared. The three pairs of duplicated genes are circled, and the duplicated blocks they belong to are indicated . MORN repeats are present in a relatively small number of proteins from various organisms, fungi being one exception lacking proteins with such a motif. The best-characterized non-PIPkin proteins with such repeats are proteins called junctophilins, which have so far only been described in animal cells. These proteins are components of the junctional complexes present between the plasma membrane and the endoplasmic reticulum . In junctophilins, the eight MORN motifs are arranged in two groups of six and two motifs each . Using green fluorescent protein fusion proteins, it was demonstrated that the MORN motifs are necessary for junctophilin-1 binding to the plasma membrane . The target of the MORN domain in the plasma membrane is thought to be phospholipids rather than proteins . Whether the MORN domain of AtPIPkins plays a role in the subcellular localization of these proteins is not known, but this is a plausible hypothesis. Membrane-binding in human PIPKIIbeta is believed to be provided by a region of the catalytic domain . HsPIPKIIbeta contains a protein kinase ATP-binding core that is believed to be present in all PI kinases . One main difference between the structures of the catalytic domain of HsPIPKIIbeta and PKA is the absence in the former of structures that might hinder their association with the membrane . It is possible that, in AtPIPkins and PIPkins from plants in general, the MORN domain is involved in membrane binding, not the catalytic domain, and that animal type I and II PIPkins have evolved from an ancestral enzymes including a MORN domain. A comparison of the structure of the type I/II AtPIPkin genes is presented in Figure . The two class A AtPIPkin genes have identical structures, consisting of nine exons and eight introns. The structure of class B AtPIPkin genes is conserved within the three groups defined in Figure . In particular, the three genes encoding AtPIP5K4-6 are almost identical in their structure except for one additional intron present in AtPIP5K6. The group consisting of AtPIP5K1-3 also features two genes with identical structures. The third gene in this group lacks two of the introns, and one of its introns is shorter than those of the two other genes. The structure of the three genes composing the third group of AtPIPkins (AtPIP5K7-9) is identical except that the first exon of AtPIP5K8 and AtPIP5K9 is interrupted by an intron in AtPIP5K7, and that AtPIP5K7 contains an additional exon at its 5' end. This exon is apparently absent from all the other type I/II AtPIPkin genes. AtPIP5K1 expression using promoter-GUS fusion and single-cell RT-PCR is strongest in the cells of the procambial tissues of leaves, flowers, and roots . This suggests that AtPIP5K1 may play a role in cell proliferation. Figure 9 | Schematic representation of the genes encoding characterized and putative type I/II PIPkins from Arabidopsis. Schematic representation of the genes encoding characterized and putative type I/II PIPkins from Arabidopsis. The introns and exons for AtPIP5K1, AtPIP5K5, AtPIP5K7, and AtPIP5K8 were determined by comparing the mRNA and gene sequences, whereas for the other genes, they were deduced by comparison with the four known gene structures and examination of the exon-intron splice junctions. The regions coding for the different conserved domains are indicated by double arrows. Dim, Dimerization domain. The molecular basis for the difference in substrate specificity between type I and II PIPkins was recently shown to implicate the so-called activation loop of these enzymes. The activation loop is a short segment, 22 to 27 amino acid residues long, located close to the C-terminal end of the catalytic domain of all PIPkins, including type III (Fig. ; ). When the activation loop of a human type I PIPkin was exchanged with that of a type II PIPkin, the resulting chimera exhibited type II substrate specificity, whereas a chimera consisting of a type II backbone with only the activation loop from a type I enzyme exhibited type I specificity . A number of residues of the activation loop are specifically conserved in all members of each PIPkin type, including type III, and are therefore thought to confer substrate specificity. However, these type-specific residues are not specifically conserved in any PIPkins from Arabidopsis, but rather AtPIPkins possess in their activation loops both type I- and type II-specific residues. The substrate specificity of animal and plant PIPkins has not been determined in vivo. So far, the yeast PIPkin Mss4p is the only PIPkin whose substrate specificity has been studied in vivo, and it was shown that it produces mainly PtdIns(4,5)P2, only minor amounts of PtdIns(3,4)P2, and no PtdIns(3,5)P2 . Fab1 Homologs | Type III PIPkins were recently proposed as a name for the group of enzymes consisting of the Brewer's yeast Fab1p protein and its homologs from other species . These enzymes are PIPkins that only use PtdIns3P as a substrate and produce PtdIns(3,5)P2 . Yeast cells lacking a functional FAB1 gene do not contain any PtdIns(3,5)P2, and a mammalian type I PIPkin gene cannot restore this phenotype . This demonstrates that, in vivo, the enzyme responsible for the synthesis of PtdIns(3,5)P2 is Fab1p, and that the ability of type I PIPkins to produce PtdIns(3,5)P2 from PtdIns3P in vitro does not represent their biologically relevant activity. PIKfyve, a mammalian Fab1, generates PtdIns5P from PtdIns in vitro . However, when expressed in yeast, it did not induce the formation of PtdIns5P, nor did any other type III PIPkin , confirming that in vivo type III PIPkins are so far the only PIPkin able to synthesize PtdIns(3,5)P2 and that they are most likely not responsible for the synthesis of PtdIns5P. In yeast, animal, and plant cells, salt stress stimulates the production of PtdIns(3,5)P2 . The mechanism(s) involved in the regulation of Fab1 enzymes are at present not known. Deletion or mutation of yeast FAB1 results in temperature-sensitive growth and massive enlargement of the vacuole caused by a defective membrane flux . Four Arabidopsis genes encode putative Fab1p homologs, and all four genes are expressed in plant cells, because at least one EST for each has been reported . These four genes do not belong to any of the duplicated blocks defined by , but they may well have arisen from an older duplication event. However, none of the four putative AtFab1 enzymes has yet been characterized in vitro or in vivo. The proteins encoded by AtFab1c and AtFab1d lack a putative FYVE domain. Surprisingly, only one FAB1 gene appears to be present in the human genome as well as in yeast. The structure of most Fab1 proteins is similar and is composed of three conserved domains: an N-terminal PtdIns3P-binding FYVE domain (this domain is apparently absent from AtFab1c and AtFab1d), a central domain with similarities to a conserved motif present in one of the subunits, Cct1p of the chaperonin complex and its homologs, and a C-terminal PIPkin catalytic domain (Fig. ; ). They are much larger than other PIPkins, most of them having a molecular mass of over 200 kD. Figure 10 | Schematic representation of Fab1 proteins. Schematic representation of Fab1 proteins. The conserved domains are represented by blocks of different colors. The intervals between the conserved domains vary among species. Two of the four Arabidopsis genes encoding putative Fab1 proteins do no contain a FYVE domain. PI-PLC : In vitro, PI-PLC isoforms catalyze the hydrolysis of PtdIns, PtdIns4P, and PtdIns(4,5)P2, and are Ca2+-dependent enzymes. Bacteria such as Bacillus cereus and Listeria monocytogenes contain a PtdIns-specific PLC that cannot use PtdIns4P or PtdIns(4,5)P2 as substrates. Biochemical Characterization | PI-PLC activity has been detected in a number of different plant species and tissues. On the basis of in vitro assays, two types of activity were distinguished: One predominantly present in the soluble fraction of plant cells prefers PtdIns as a substrate and requires millimolar Ca2+ concentrations, and a second present in the plasma membrane that prefers PtdIns4P and PtdIns(4,5)P2 and requires much lower (0.1 --10 mum) Ca2+ concentrations . It was also demonstrated early on that PI-PLC activity is present in both soluble and particulate fractions of plant cells. In most cases, the particulate fractions studied were purified plasma membranes. In the instances when substrate specificity was examined, it was found that the PI-PLC activity in plasma membranes showed a net preference for PtdIns4P and PtdIns(4,5)P2, the exception being a plasma membrane fraction from suspension-cultured rice (Oryza sativa) cells , which showed an apparent preference for PtdIns. Although PI-PLC has been purified from several plant tissues, no amino acid sequence has ever been obtained for any of these purified proteins. It is, therefore, still not known whether any of the PI-PLC genes identified so far correspond to any of the activities detected in or partially purified from plant extracts. However, isolated a cDNA encoding a biochemically active PI-PLC from soybean by screening an expression library with an antiserum raised against total proteins from purified plasma membranes and demonstrated that when expressed in transgenic tobacco plants, the fusion protein was present in the plasma membrane but also in the cytosol. Antibodies raised against a peptide specific for one PI-PLC from Arabidopsis recognize a single polypeptide of approximately 66 kD that is significantly enriched in plasma membranes purified from Arabidopsis plants and cannot be detected in the cytosol . Although it is known that plant PI-PLC is activated by Ca2+, its regulation remains unresolved. When tested, guanine nucleotides have not stimulated PI-PLC activity at all in most cases. Even when activity has been stimulated, inositol phosphate production increased only 2- to 3-fold (not 20- to 30-fold as seen in animal systems) and these increases were accompanied by similar decreases in PtdInsP and PtdInsP2 . Clear, marked stimulatory effects of guanine nucleotides, or any other compounds, on plant PI-PLC activity have never been described. Several plant studies have reported that mastoparan (a potent G-protein activator) and/or its analogs induced Ins(1,4,5)P3 formation and suggested that plant PI-PLC may be regulated by a G-protein-dependent pathway. However, it was recently demonstrated that mastoparan and Mas-7 induce the formation of pores in the plasma membrane of human cells . These pores allowed small molecules, such as ethidium bromide, lucifer yellow, but not Evan's blue, to penetrate the cells. Treatment with mastoparan and Mas-7, thus, resulted in a rise in the intracellular Ca2+ concentration. Chlamydomonas spp. was shown to respond to non-permeabilizing concentrations of mastoparan (i.e. concentrations that did not result in cells becoming permeable to Evan's blue) and its analogs by increasing Ins(1,4,5)P3 levels . It was also demonstrated that these non-permeabilizing concentrations did not induce an influx of Evan's blue inside the cells, however, the influx of smaller molecules such as Ca2+ was not examined. Increases in [Ca2+]i upon mastoparan treatment have been observed in a higher plant system . It is therefore possible, as also suggested by , that the Ins(1,4,5)P3 increase and simultaneous PtdIns(4,5)P2 decrease detected in Chlamydomonas spp. upon treatment with mastoparan and its analogs are attributable to a stimulation of PI-PLC activity via an increase in [Ca2+]i, rather than through activation of a G-protein. Further doubt to the possible regulation of plant PI-PLCs by G-proteins is raised by several facts. First, Arabidopsis contains only one gene coding for each of the three subunits forming heterotrimeric G-proteins, whereas in animal cells, several genes for each subunit exist. Second, and as mentioned earlier, only a few genes encoding putative G-protein-coupled receptors are present in Arabidopsis, whereas hundreds have been identified in animals. Third, as described below, the only type of PI-PLC identified in plants is not of the beta- or epsilon-type, the only type of PI-PLC activated by heterotrimeric G-proteins . Genes, Structure, and Function | The first cDNA clones encoding functional plant PI-PLCs were reported in 1995 for Arabidopsis and soybean , 7 years after the first animal PI-PLC clones were described . Plant cDNA clones encoding active PI-PLC isozymes have also been obtained from potato (Solanum tuberosum; ) and wild tobacco (Nicotiana rustica; ). Genes encoding additional putative PI-PLCs have been identified in Arabidopsis, garden pea (Pisum sativum), cowpea (Vigna unguiculata; accession no. ), rape seed (Brassica napus), soybean, and hairy finger-grass (accession no. ). In Arabidopsis, one cDNA clone, AtPLC1 (accession no. ), encoding a protein with demonstrated PI-PLC activity has been partially characterized . It showed a marked preference for PtdIns(4,5)P2 versus PtdIns, with a specific activity near the optimal free Ca2+ concentration of each substrate (approximately 100 times higher with PtdIns(4,5)P2 than with PtdIns). Maximal hydrolysis rates against PtdIns(4,5)P2 were reached at 1 to 50 mum free Ca2+, whereas Ca2+ concentrations above 1 mm were required to observe substantial PtdIns hydrolysis. Similar characteristics have been described for recombinant PI-PLCs from potato and soybean . Arabidopsis contains six additional genes with deduced protein sequences showing high similarity to AtPLC1 (Fig. ; Table ). Figure 11 | Alignment of the seven PI-PLCs from Arabidopsis. Alignment of the seven PI-PLCs from Arabidopsis. The accession numbers for the different proteins are given in Table . The sequences were aligned using the ClustalW program. The four conserved domains are indicated. The X and Y domains together form the catalytic domain of the enzymes. Because no mRNA or EST sequences for AtPLC6 have been obtained, the sequence shown here is the predicted one that fits best with the other sequences. Table III | List of characterized and predicted PI-PLCs from Arabidopsis PI-PLCs are classified in four distinct subfamilies: beta, gamma, delta , and epsilon . Five domains, a PH domain, an EF-hand domain, an X and a Y domain (together forming the catalytic domain), and a lipid-binding C2 domain, are conserved in PI-PLC enzymes and represent the core sequence of all known PI-PLCs from animals and yeast. This core sequence constitutes the structure of delta-isozymes . The beta-, gamma-, and epsilon-isozymes contain specific insertions in the core sequence and are found only in animals. PI-PLC-betas are activated by G-protein-coupled receptors, PI-PLC-gammas by receptor Tyr kinases , and PI-PLC-epsilons by heterotrimeric G-proteins and Ras . Sequence analysis indicates that all plant PI-PLC sequences are structurally identical and composed of four conserved domains, three of which correspond to the X, Y, and C2 domains conserved in non-plant PI-PLCs (Figs. and ). The structure of plant PI-PLCs resembles most closely that of PLC-delta fisoforms. The fourth conserved domain, preceding the X domain in plant PI-PLCs, is less conserved than the rest of the proteins, apart from the linker between the X and Y domains. Secondary structure predictions for this region show a clear resemblance to that of the secondary structure determined for the second loop of the EF-hand domain present in PLC-delta1, as determined by x-ray crystallography . Helices corresponding to the four helices forming the second loop of the EF-hand domain of PLC-delta1 are predicted in all plant PI-PLCs at sequence positions that are also conserved between plant isozymes and PLC-delta1. In PLC-delta1, the second loop of the EF-hand domain is believed to serve a critical structural role, through interaction with the C-terminal C2 domain. Deletion variants of animal PLC-delta1 and PLC-gamma lacking this second loop are inactive . We have recently shown that in AtPLC2, a deletion variant lacking the 36 first amino acids (i.e. including a helix corresponding to helix E3alpha of PLC-delta1), is inactive, whereas a variant lacking the 22 first amino acids is fully active . This strongly suggests that plant PI-PLCs probably contain an EF-hand domain corresponding to the second loop of the EF-hand domain of animal PI-PLCs. Interestingly, all Arabidopsis PI-PLC genes contain an intron immediately after the region encoding the predicted EF-hand domain . The seven Arabidopsis PI-PLC genes are composed of seven to nine exons: AtPLC3 contains only seven, AtPLC1 and AtPLC5 contain eight, and the remaining four genes have nine exons. Despite having the same number of exons and introns, AtPLC1 and AtPLC5 do not have the same structure: They differ by two introns. AtPLC2 and AtPLC4 are characterized by a much longer intron following the first exon. Because no cDNA sequences have been reported for AtPLC6 and AtPLC7, it is not possible to ascertain their structure. Figure 12 | Representation of the modular domain arrangements of PI-PLC delta isozymes from animals and plants. Representation of the modular domain arrangements of PI-PLC delta isozymes from animals and plants. The conserved domains are represented by blocks of different colors. The EF-hand domain of plant PI-PLCs corresponds to the second loop of the EF-hand domain of animal PI-PLCs. The X and Y domains constitute together the catalytic domain of PI-PLCs. Figure 13 | Schematic representation of the seven PI-PLC genes from Arabidopsis. Schematic representation of the seven PI-PLC genes from Arabidopsis. The introns and exons for AtPLC1, AtPLC2, AtPLC3, AtPLC4, and AtPLC5 were determined by comparing the mRNA and gene sequences, whereas for the other three genes, they were deduced by comparison with AtPLC1-5 and examination of the exon-intron splice junctions. In AtPLC6, two putative start codons can be identified. In addition, an insert in the first exon of this gene is present, and indicated by a gray box. The regions coding for the different conserved domains are indicated by double arrows. Plant PI-PLCs do not contain a PH domain. In animal PI-PLC-delta, this domain is required for interaction with the plasma membrane and is involved in the binding of the lipid substrate and in processive catalysis . The C2 domain is also involved in membrane interaction but is not sufficient to position the enzyme in a catalytically active orientation . A thorough investigation of the lipid-binding and cellular localization of the C2 domains from the four mammalian PLC-delta isoforms recently showed that two of them bind specifically and in a Ca2+-dependent manner to phosphatidyl-Ser localized in the plasma membrane . Because plant PI-PLCs lack a PH domain, their interaction with the plasma membrane and PtdIns(4,5)P2 clearly differs from PI-PLC-delta. The C2 domain of plant PI-PLCs may be sufficient for membrane-binding. Apart from their C2 domain, other regions in plant PI-PLCs may be required for plasma membrane association, such as hydrophobic moieties, with or without posttranslational modification of the protein. The function of plant PI-PLCs in vivo has been addressed, but is still obscure ---in particular, the expression and subcellular distribution of the seven Arabidopsis isoforms has not been addressed. Using the aminosteroid compound U-73122, which inhibited the in vitro activity of a PI-PLC from Nicotiana rustica guard cells, it was demonstrated that stomatal guard cell responses to abscisic acid (ABA), aperture regulation, and cytosolic Ca2+ oscillations are inhibited by U-73122 at concentrations that also inhibited the recombinant PI-PLC. This suggests that ABA may activate a PI-PLC leading to cytosolic Ca2+ alterations and ultimately to stomatal closure . Using sense and antisense transgenic Arabidopsis plants, demonstrated that AtPLC1 is necessary for the inhibition of germination by ABA but that overexpression of AtPLC1 did not result in the induction of ABA-regulated genes, demonstrating that AtPLC1 may be involved in secondary ABA responses. It was also demonstrated that a second PI-PLC gene from Arabidopsis, AtPLC2 is not involved in these responses, in agreement with the fact that AtPLC1 is expressed only when plants are treated with ABA or subjected to drought or salt stress, whereas AtPLC2 expression is not affected by ABA . CONCLUDING REMARKS AND PERSPECTIVES : Progress in the understanding of the function of PIs and their derivatives in plant cells has been slow, especially when compared with the animal field. A number of reasons can explain this lag. The identification of PIs from plant tissues has proved to be a challenge in plants. Attempts at purifying enzymes involved PI metabolism have failed to provide protein sequences. It has consequently been difficult to study the regulation of these enzymes, manipulate PI metabolism in vivo, and identify possible targets of these lipids. However, sequencing of the Arabidopsis genome has enabled the isolation and characterization of a number of cDNAs encoding PI-metabolizing enzymes and the identification of many genes coding for putative orthologs of other PI-metabolizing animal enzymes. The enzymes catalyzing the removal of phosphate from PIs and/or inositol (poly) phosphate have not been discussed in the present review, mainly because only inositol (poly) phosphate phosphatases, but no PI phosphatase, from plants have been characterized. The phosphatases catalyzing these reactions form a large family, and can be categorized in four groups according to the position of the phosphoester bond on the inositol group they hydrolyze (i.e. 1-, 3-, 4-, and 5-phosphatases; for review, see ). Some of these phosphatases can hydrolyze inositol (poly) phosphates only, others PIs only, whereas others can use both as substrates. 1-Phosphatases cannot use PIs as a substrate. One Arabidopsis gene (At5g63980) encoding an inositol polyphosphate 1-phosphatase has been identified, characterized, and shown to be involved in ABA and stress responses , and three from tomato have also been characterized . The Arabidopsis enzyme possesses catalytic bifunctional activity, namely a 3'(2'),5'-bisphosphate nucleotidase activity, and an inositol polyphosphate phosphatase activity. In Arabidopsis, five additional genes encode putative inositol polyphosphate phosphatase, two of which are located immediately after At5g63980 on chromosome 5, and another two genes are also found on chromosome 5. 3-Phosphatases can use either inositol polyphosphate or PIs as substrates, but not both. None have been biochemically characterized from plants. However, the Arabidopsis contains a number of putative orthologs of animal 3-phosphatases. Most interesting is the presence of genes encoding putative members of the two families of PI 3-phosphatases, phosphatase and tensin homolog (PTEN) and myotubularin . Mutated forms of PTEN were identified in numerous human tumors. Originally thought to be a Tyr phosphatase, PTEN was demonstrated to have only weak phosphatase activity toward phospho-Tyr-containing peptides. Instead, its physiological substrate is believed to be PtdIns(3,4,5) P3, with other 3-phosphorylated PIs being poorer substrates in vitro. It is, therefore, intriguing that Arabidopsis contains two genes encoding putative orthologs of enzymes whose preferred substrate is a lipid that has never been detected in plant cells. Mutations in the MTM1 gene cause the neonatal disorder X-linked myotubular myopathy, which in severe cases is often fatal . MTM1 and MTM-related genes encode myotubularin isoforms that are highly similar to each other. Similar to PTEN, myotubularin was first thought to be a protein phosphatase because it contains the active site present in members of the protein Tyr phosphatase family of proteins. However, it too has low activity in the presence of artificial protein substrates. It was then observed that the catalytic domain of myotubularin shows significant similarity with the active site of another PI phosphatase, the SacIp protein (see below) from Brewer's yeast. Myotubularin is indeed a PI phosphatase, showing a clear preference for PtdIns3P as substrate . Two genes in Arabidopsis encode putative myotubularin-like proteins. Both genes appear to be expressed because one EST sequence for each gene has been deposited in the GenBank database. Finally, one enzyme, multiple inositol polyphosphate phosphatase, has been demonstrated to hydrolyze various inositol polyphosphates . One Arabidopsis gene encodes a potential ortholog of the animal enzyme. Only two genes encoding 4-phosphatases have identified from animal tissues. They encode phosphatases that can hydrolyze PtdIns(3,4)P2, Ins 1,3,4-trisphosphate, and Ins 3,4-bisphosphate. No putative orthologs of these two proteins are present in Arabidopsis. The largest group of inositol (poly) phosphates and/or PI phosphatase is constituted by enzymes hydrolyzing the 5-phosphate group of these inositol-containing molecules. They all contain a 5-phosphatase catalytic domain and can be divided into type I and type II enzymes. Type I enzymes do not hydrolyze PIs, whereas the larger type II can. Ins(1,4,5)P3 and Ins 1,3,4,5-tetrakisphosphate are the favored substrates of type I enzymes. One cDNA clone from Arabidopsis encoding type I 5-phosphatases with substrate specificities similar to the animal enzymes have been characterized . Ten additional genes encoding putative type I 5-phosphatases are present in Arabidopsis . This multiplicity is intriguing, especially in view of the fact that Brewer's yeast does not appear to possess a 5-phosphatase with type I substrate specificity , although the Inp54p protein is more similar in size to type I than type II 5-phosphatases. This suggests that the metabolism of inositol (poly) phosphates in plant cells is highly complex. Type II 5- phosphatases can hydrolyze PIs, and some enzymes are also able to hydrolyze Ins polyphosphates . They all possess a type II domain, except for Inp54p in which this domain is truncated . A variety of protein modules, e.g. Sac, SH2, and GTPase activating protein domains, are found in individual members. In Brewer's yeast, the four different 5-phosphatases Inp51p, Inp52p, Inp53p, and Inp54p belong to the type II group . No type II 5-phosphatase from plants has been characterized. However, the genome of Arabidopsis encodes four putative type II 5-phosphatases . The predicted proteins do not contain the Sac domain present in some phosphatases, including three of the type II 5-phosphatases from Brewer's yeast. The Sac domain confers phosphatase activity toward PIs, but not Ins polyphosphates . Of all the existing PIs, PtdIns(4,5)P2 and PtdIns(3,4)P2 are not hydrolyzed by Sac-containing enzymes, whereas PtdIns5P has not been tested as a substrate. As just mentioned, the Sac domain is present in some type 5-phosphatases. There are, in addition, proteins with a Sac domain but no 5-phosphatase domain (two such proteins exist in Brewer's yeast; ). None have been characterized in plants, but Arabidopsis contains nine genes encoding putative proteins with a domain highly similar to the Sac domain. From these studies and observations, it is obvious that plant PI-metabolizing enzymes are related to their animal counterparts both in their primary structure and substrate specificity. However, the pathways involved in PI metabolism in plant cells present some unique features, including a higher number of type I and type III PIPkin isoforms than in animal cells, an apparent lack of animal-like type II PIPkins, an absence of class I and II PtdIns 3-kinases and PtdIns(3,4,5)P3, and only one type of PI-PLC. The subcellular distribution of these different enzymes and the PIs has also been only superficially examined, because it is still not simple to obtain some of the subcellular compartments with adequate purity; in particular, there is at present no protocol allowing the isolation of highly purified Golgi and endoplasmic reticulum membranes from plant extracts. A number of major questions regarding plant cell PI metabolism now clearly need to be addressed. Which of the various isoforms of the different inositol lipid kinases are responsible for the synthesis of the different inositol lipid isomers in vivo? How are these kinases and the PI-PLC isoforms regulated? What is the function of the different isoforms of PI-PLC, PIPkins? As pointed out above, Ins (poly) phosphates and/or PI phosphatases may form a family of more than 35 proteins in Arabidopsis. This virtually unexplored and exciting area of plant PI research undoubtedly deserves much attention; the properties and functions of these proteins await characterization. An even more important challenge is to determine the physiological role of PIs in plant cells. One approach could involve the characterization of the numerous predicted proteins from Arabidopsis containing PI-binding domains, such as the PH, PX, FYVE, and epsin N-terminal homology domains . Arabidopsis knock-out mutants should also provide useful information when they result in a clear phenotype. Note Added in Proof | A fifth subfamily of PI-PLC, designated zeta, has been identified in animals. One member from this new subfamily of PI-PLC has been characterized. It is smaller in size, 74 kD, than any other PI-PLC from animals, and lacks an N-terminal PH domain . Backmatter: PMID- 12226485 TI - Cell-Specific Expression of Homospermidine Synthase, the Entry Enzyme of the Pyrrolizidine Alkaloid Pathway in Senecio vernalis, in Comparison with Its Ancestor, Deoxyhypusine Synthase AB - Pyrrolizidine alkaloids (PAs) are constitutive plant defense compounds with a sporadic taxonomic occurrence. The first committed step in PA biosynthesis is catalyzed by homospermidine synthase (HSS). Recent evidence confirmed that HSS evolved by gene duplication from deoxyhypusine synthase (DHS), an enzyme involved in the posttranslational activation of the eukaryotic translation initiation factor 5A. To better understand the evolutionary relationship between these two enzymes, which are involved in completely different biological processes, we studied their tissue-specific expression. RNA-blot analysis, reverse transcriptase-PCR, and immunolocalization techniques demonstrated that DHS is constitutively expressed in shoots and roots of Senecio vernalis (Asteraceae), whereas HSS expression is root specific and restricted to distinct groups of endodermis and neighboring cortex cells located opposite to the phloem. All efforts to detect DHS by immunolocalization failed, but studies with promoter-beta-glucuronidase fusions confirmed a general expression pattern, at least in young seedlings of tobacco (Nicotiana tabacum). The expression pattern for HSS differs completely from its ancestor DHS due to the adaptation of HSS to the specific requirements of PA biosynthesis. Keywords: Introduction : Homospermidine synthase (HSS) catalyzes the first pathway-specific step in the biosynthesis of pyrrolizidine alkaloids (PAs), a group of secondary compounds whose distribution is widely scattered among the angiosperms . PA-producing plants accumulate these compounds constitutively in all plant organs as chemical defenses against herbivores. In annual Senecio spp., the inflorescences are the major sites of PA accumulation . The function of PAs as powerful deterrents and toxins is supported by the impressive adaptations of certain insects that are specialized to feed on PA-containing plants and utilize plant-acquired PAs for their own protection (for review, see ; ; ). The biosynthesis of PAs has been intensively studied, particularly in Senecio spp. (Asteraceae). Here, PAs are synthesized in the roots as N-oxides that are translocated into the shoot through the phloem . Specific carriers are involved in phloem loading and unloading of the polar PA N-oxides because species that do not produce PAs are unable to translocate them via the phloem . In all Senecio spp. studied so far, senecionine N-oxide is the primary product of biosynthesis. The backbone structure is modified by one- or two-step reactions (e.g. hydroxylations, dehydrogenations, epoxydations, O-acetyla-tions, etc.) that are species specific and lead to the unique PA pattern of a given plant population. In the living plant, PAs are spatially mobile but do not show any turnover or degradation . In PA biosynthesis, homospermidine is the first pathway-specific precursor of the necine base moiety of these ester alkaloids . Homospermidine is a rare polyamine, which is not commonly present in plants. It is formed by HSS (spermidine specific; EC ), which catalyzes the transfer of the aminobutyl moiety from spermidine to a putrescine molecule in an NAD+-dependent reaction. In contrast to its substrates putrescine and spermidine, which in Senecio spp. roots show a dynamic turnover, homospermidine does not exhibit any metabolic activity apart from its incorporation into PAs . In Senecio spp. roots, free homospermidine is only detectable in the presence of beta-hydroxyethylhy-drazine, a diamine oxidase inhibitor, which efficiently blocks the subsequent step in PA biosynthesis. If the inhibition is released, PA biosynthesis starts again at the expense of accumulated homospermidine. Figure 1 | Biosynthesis of PAs in the roots of S. vernalis Biosynthesis of PAs in the roots of S. vernalis. Putrescine and spermidine from primary metabolism are used as substrates for HSS to catalyze the formation of homospermidine. This is the first committed step in PA biosynthesis. The resulting homospermidine is exclusively incorporated into the necine base moiety of senecionine N-oxide. During translocation of this parent PA to the aerial parts of the plant, its structure is chemically modified to provide the PA derivatives found in S. vernalis. Recently obtained molecular data about HSS from Senecio vernalis provided conclusive evidence for its close phylogenetic relation to deoxyhypusine synthase (DHS; EC ), an enzyme involved in the posttranslational activation of the eukaryotic initiation factor 5A (eIF5A; ). DHS catalyzes the first of the two enzymatic reactions leading to one of the most specific posttranslational modifications known to produce activated eIF5A. Although inhibition of hypusine formation stops cell growth at the G1/S boundary , the function of eIF5A remains elusive (for review, see ). Recently, eIF5A was localized at the nuclear pore complex , where it is efficiently exported by the transport receptor exportin 4 . It probably functions as a carrier for the export of specific RNAs from the nucleus to the cytosol. In plants, these mRNAs may be required for programmed cell death . DHS and eIF5A seem to be conserved among eukaryotes and archaebacteria . Recently, the same mechanism of activation was confirmed in plants by cloning and functional expression of DHS and eIF5A from tobacco (Nicotiana tabacum; ) and S. vernalis . The recruitment of HSS from DHS has been interpreted as evolution by change of function because HSS was recruited for PA biosynthesis, a totally different function in plant secondary metabolism, and maintained under the selection pressure of herbivory . Mechanistically, HSS and DHS catalyze analog reactions, both transferring an aminobutyl moiety of spermidine to their substrates, i.e. putrescine in the case of HSS and to a specific protein-bound Lys residue in the case of DHS. Interestingly, purified DHS also catalyzes as a side reaction the aminobutylation of putrescine with the same kinetic properties as HSS (; D. Ober, R. Harms, L. Witte, and T. Hartmann, unpublished data). This suggests that HSS could have evolved from DHS by simply losing its ability to bind the eIF5A protein. Because HSS controls the substrate flow into the alkaloid pathway, it occupies a central position in the regulation of PA biosynthesis. HSS and DHS are cytosolic enzymes and the activities of both enzymes seem to be correlated with cell growth . The intention of this study was to elucidate the spatial localization of HSS in roots of S. vernalis and to understand the biochemical aspects of the compartmentation discussed above. Using polyclonal antibodies against HSS, we studied the tissue-specific and subcellular localization of HSS. The identification of groups of distinctive cells close to the phloem as sites of homospermidine formation indicate a highly cell-specific expression of alkaloid biosynthesis with possible symplastic connection to the phloem. In addition, RNA gel-blot analysis and semiquantitative PCR were performed to compare the tissue-specific expression pattern of the closely related HSS and DHS genes in S. vernalis. For comparison, tobacco, a plant that is devoid of PAs, was included in these studies. Due to difficulties in detecting DHS by immunolocalization, promoter-beta-glucuronidase (GUS)/green fluorescent protein (GFP) fusions were constructed to analyze the tissue specificity of the promoter of tobacco DHS. RESULTS : Analysis of hss and dhs Gene Expression in S. vernalis and Tobacco | To compare hss and dhs gene expression in different organs of S. vernalis and tobacco, we employed two strategies. First, northern-blot analysis was performed with complete cDNAs encoding HSS and DHS as probes. Second, because both genes are closely related and the respective enzymes share identities of 83% and 79% on nucleic acid and amino acid level, respectively , we performed semiquantitative reverse transcriptase (RT)-PCR to prove the specificity of the probes in the northern blot. In these experiments primers were applied, which are highly specific for the cDNAs of HSS and DHS, respectively. However, this technique allows only demonstration of the presence or absence of gene expression in a tissue without rigorous quantitative comparison. Figure shows the results of the northern-blot analysis. The hss gene is expressed at high levels in the roots of S. vernalis independently of the root age, but is not expressed in the aerial parts of the plant like the buds, flower heads, leaves, or stems. This result supports the biochemical evidence that roots of Senecio spp. are the exclusive site of PA biosynthesis, as was shown with in vitro root cultures and detached plant organs . Western-blot analysis of six successive 1-cm segments beginning with the root tip showed expression of HSS in all tested root segments except the undifferentiated first centimeter of the root tip. The hypocotyl was devoid of any HSS expression (data not shown). In contrast to the selective expression of the hss gene, the dhs gene was expressed at low levels in all tested organs of S. vernalis, with higher levels in the roots (Fig. B). Semiquantitative RT-PCR confirmed that the dhs gene is expressed in all tested tissues of S. vernalis as well as of tobacco, whereas the hss gene is expressed selectively only in the roots of S. vernalis. The test of whether the expression of the dhs gene correlates with expression of the gene encoding the substrate of DHS, eIF5A, also showed for the eIF5A an expression at constant level in all tested tissues , which is well in accord with the data for eIF5A from maize (Zea mays; ). Also, ovaries of different developmental stages (of young flower bud through developed fruit-containing seeds) showed a constant level of dhs gene expression (data not shown). Figure 2 | Northern-blot analysis and RNA loading control of different tissues of S. vernalis Northern-blot analysis and RNA loading control of different tissues of S. vernalis. RNA of S. vernalis tissues was hybridized with digoxigenin (DIG)-labeled HSS probe (A) and DHS probe (B) of S. vernalis (buds, lane 1; young leaves, lane 2; young stem, lane 3; flower heads, lane 4; old stem, lane 5; young root, lane 6; and old root, lane 7). Figure 3 | Semiquantitative RT-PCR. Semiquantitative RT-PCR. Reverse transcription and PCR were performed with RNA of different tissues of S. vernalis with primers specific for HSS and DHS of S. vernalis (A) and with RNA of different tissues of tobacco with primers specific for DHS of tobacco (B). Figure 4 | Northern-blot analysis and RNA loading control of different plant organs of tobacco with eIF5A probe. Northern-blot analysis and RNA loading control of different plant organs of tobacco with eIF5A probe. RNA of different tissues was hybridized with DIG-labeled probe of eIF5A (roots, lane 1; shoots, lane 2; old leaves [about 20 cm in length], lane 3; young leaf [about 2 cm in length], lane 4; stamina of open flower, lane 5; stamina of closed flower bud, lane 6; petals of open flower, lane 7; petals of closed flower bud, lane 8; ovary of open flower, lane 9; and ovary of closed flower bud, lane 10). Immunolocalization of HSS in Root Sections of S. vernalis | The tissue-specific localization of HSS was studied with roots of field- or greenhouse-grown S. vernalis. Although in vitro-cultured roots proved to be an ideal system for biochemical studies of PA biosynthesis , they turned out to be unsuitable for histological studies because of incomplete tissue differentiation in the central cylinder. The immunostaining of HSS was performed with polyclonal antibodies from rabbits raised against S. vernalis HSS. Before use, the antibodies were purified by affinity chromatography with immobilized recombinant HSS. Figure A shows a fluorescence photograph of FITC-labeled root sections. The xylem vessels of the triarch vascular bundle show intensive yellow autofluorescence, due to lignification. Yellow autofluorescence is also visible in the radial cell walls of the endodermis cells caused by suberin (casparian strip, marked with arrowheads). A strong antibody labeling is detected in endodermis cells and the adjacent parenchyma cells of the root cortex, which are located opposite of the phloem tissue. The phloem is located in between the radially arranged ridges of the xylem, separated from the endodermis by the pericycle, the outer cell layer of the central cylinder. The pericycle cells are devoid of any labeling. Figure 5 | Immunolabeling of HSS in root sections of S. vernalis Immunolabeling of HSS in root sections of S. vernalis. A, Specific labeling using fluorescein isothiocyanate (FITC) detection of specialized cells within the endodermis and the adjacent parenchyma cells of the root cortex with affinity-purified anti-HSS antibody (Co, parenchyma cells of the root cortex; En, endodermis; Ph, phloem; Xy, xylem; and the casparian strips in the radial cell walls of the endodermis are marked by arrowheads). B through E, Specific labeling as in A, but in the presence of purified HSS (B and C) and in the presence of purified DHS (D and E) in a molar ratio of antibody:added protein of 10:1 and 1:3, respectively. F, Immunogold labeling enhanced with silver of a root section with beginning secondary growth (cells introduced by cambium activity are marked by arrows). The HSS-specific label is still detectable. To confirm the specificity of this labeling, various control experiments were performed. Incubation of the section with either pre-immune serum or the secondary FITC-labeled antibody omitting the primary antibody gave no labeling (data not shown). In protein gel-blot analysis, the antibody raised against the HSS showed weak but detectable cross-reactivity with DHS of S. vernalis, even after affinity purification (data not shown). Because DHS was shown to be expressed in roots (Figs. B and A), any influence on HSS immunolocalization by cross-reactivity with DHS needed to be excluded. Therefore, the labeling experiment was repeated by addition of either purified HSS or purified DHS. In case of a specific labeling of HSS, the label should be reduced by pre-incubation of the antibody with added HSS protein, but not with DHS protein. The pre-incubations were performed at room temperature 10 min before the mixture was applied to the root section. Two different protein concentrations were used, resulting in a molar ratio of antibody to added protein of 10:1 and 1:3. Figure , B through E, show that only the addition of soluble HSS reduces (10:1 ratio) and even completely blocks (1:3 ratio) the specific labeling, whereas addition of the DHS protein is ineffective. Figure 6 | Electron micrographs of in situ immunogold-labeled HSS in roots of S. vernalis Electron micrographs of in situ immunogold-labeled HSS in roots of S. vernalis. A, Two cells of the endodermis (En) with the casparian strip (Ca) as incrustation in the radial cell wall. The endodermis cells are accompanied by cells of the pericycle (Pc) and the cortex parenchyma (Co). Gold label is only found in the endodermis and the adjacent cortex cells, but no label is detectable in the pericycle. B, Cell junction of two endodermis cells (En; casparian strip at bottom right) with a cell of the pericycle (Pc). C, Detail of labeled cells that show the label exclusively in the cytoplasm. All detectable organelles are devoid of any label (at the right is an intercellular space, Ic). Localization of HSS by immunogold labeling coupled with silver enhancement under the light microscope shows the same labeling pattern in a root that just started with secondary growth. Between the well-differentiated metaxylem and the phloem, rows of cells parallel to the radii of the axes are formed by the periclinal divisions of the cambium (Fig. F). Intracellular Localization of HSS in S. vernalis | Figure shows electron micrographs of cross sections of S. vernalis roots labeled with 18-nm gold particles after incubation with the affinity-purified antibody against HSS. In the center of the picture, the casparian strip is visible as a dark incrustation in the wall between two endodermis cells. These are joined at left by a cell of the pericycle. The gold particles are localized exclusively in the cytoplasm of the endodermis cells and in the adjoining cells of the cortex. There is no label associated with any cell organelle. The pericycle cells are also devoid of any label. These results clearly support the biochemical observation that HSS is a soluble cytosolic protein . Control incubations with preserum instead of the anti-HSS antibody did not show any label (data not shown). Tissue-Specific Expression of Tobacco dhs-Promoter Fusions | In contrast to HSS, all efforts to localize DHS by immunolabeling have failed so far. Two reasons may be given to explain these negative results. First, the expression level of DHS is comparatively low. We were never able to demonstrate DHS activity in plant extracts, whereas HSS activity is easily detectable in crude root extracts of S. vernalis. Second, the localization of HSS in specific root cells suggests high local enzyme concentrations that should be detected more easily than the suspected more dispersed tissue distribution of DHS. To analyze DHS expression in more detail, we used the promoter-GUS/GFP fusion technique. From the eukaryotic genome projects (i.e. human, Arabidopsis, Drosophila melanogaster, and yeast [Saccharomyces cerevisiae]), it is known that DHS is represented by a single gene copy. In tobacco, a 2.0-kb fragment upstream of the dhs gene was identified that should contain the regulatory elements of the dhs promoter. To analyze the tissue-specific expression of the dhs gene, we used this fragment to construct a promoter GFP/GUS fusion and transformed it into tobacco plants using Agrobacterium tumefaciens. GUS staining was performed with seedlings and young plantlets of the F1 generation of multiple independent transgenic lines. A significant staining was already detectable after 6 h of incubation, and longer incubation times did not alter the staining patterns. Figure 8, A through D, show strong dhs promoter activity in young tobacco seedlings. The radicle and the hypocotyl just emerging from the seed already showed intensive blue coloration (Fig. A). The promoter activity in the hypocotyl remains stable in older seedlings with expanded cotyledons, but only in the lower part, directly above the root-shoot connection (Fig. , B showing a detail of C). Intensive dhs promoter activity is also detectable in the cotyledons and the two primary leaves (Fig. C), whereas the following leaves show only a slight staining, indicating low expression (Fig. D). Roots of young seedlings also exhibit promoter activity, with the exception of the still undifferentiated growing zone of the root tip, which is generally devoid of any blue coloration (Fig. , C and D). If adventitious roots of the F0 generation are stained, some show promoter activity behind the root tips, with the exception of the tip itself in a 0.5- to 1.0-cm-long region, whereas others are devoid of any label (Fig. E). No success was achieved in trying to visualize GFP expression due to autofluorescence in the tissues. Figure 7 | GUS expression in tobacco plants transformed with the dhs-promoter GUS fusion. GUS expression in tobacco plants transformed with the dhs-promoter GUS fusion. A, Young seedling with stained root and hypocotyl. B, Detail of the transition between root and shoot of a seedling shown in C. C, Seedlings with developed cotyledons, one with two primary leaves. D, Young plant with GUS expression in the cotyledons, successive leaves, the hypocotyl, and some of the roots. Older leaves only show weak GUS expression. E, Adventitious roots of an F0 plant, of which only some show GUS expression in a section of approximately 0.5 to 1.0 cm in length directly behind the root tips. DISCUSSION : Plant secondary metabolism is characterized by an immense diversity of chemical structures of restricted taxonomic distribution that are produced with high specificity and under stringent genetic control. Genes encoding enzymes mediating the high specificity of these pathways must have been recruited during evolution from primary metabolism by gene duplication and modified under selection pressure. Their integration into a secondary pathway implies not only the adaptation of enzymatic activities to new functions but also a proper regulatory integration of the gene, including its expression level and tissue specificity. Comparison of hss and dhs Expression Patterns | The expression patterns of the hss and dhs genes revealed that they are differentially regulated. The dhs gene is expressed in all plant organs analyzed, with slightly higher levels occurring in roots, whereas the hss gene is expressed exclusively in roots (Figs. and A). These results suggest that recruitment of the hss gene from dhs was accompanied by changes in the regulation of this gene to accommodate its new and unique role in PA biosynthesis. The root-specific expression of HSS corroborates earlier physiological studies that proved that the biosynthesis of PAs is restricted to the roots in Senecio spp. It is interesting to note that on a whole-root basis in S. vernalis, the mRNA coding for HSS is expressed in much higher concentration than the mRNA coding for DHS. This becomes even more remarkable if we consider that the HSS expression is restricted to certain root cells. Obviously, the regulatory elements are also modulated in regard to the expression level of the gene product. postulate that the increased expression level of an enzyme involved in the biosynthesis of defense compounds in comparison with its ancestor from primary metabolism could increase the fitness of an organism. This is probably the case for the HSS-DHS system because the low turnover number of DHS is well known , although in plants accumulating defense compounds, low enzyme activities may suffice to establish an efficient defense system. Nothing is known about the localization and function of DHS in plants. In fact, tobacco and S. vernalis are the first plants for which the occurrence of DHS and its function in eIF5A activation was confirmed . Recently, isolated the cDNAs for DHS and eIF5A from a senescence-induced cDNA library of tomato (Lycopersicon esculentum). Only few data are available concerning plant eIF5A. succeeded in cloning and sequencing of the first plant eIF5A cDNA from Medicago sativa. Two cDNA sequences for eIF5A were isolated from Nicotinana plumbaginifolia. One of the corresponding genes is constitutively expressed in all tested tissues, whereas the second gene seems to be expressed mainly in photosynthetically active parts of the plant . In maize, eIF5A is also constitutively expressed in all tissues without any correlation of gene expression with cell division activity . Our results suggest that in plants, neither the expression patterns of eif5A nor DHS appear to be correlated with cell growth because root and leaf tissues of various developmental stages do not show noticeable differences in their expression levels (Figs. --). Thus, in plants, the function of activated eIF5A may be different from that described for other eukaryotic systems, where eIF5A appears to be essential for cell proliferation . In addition to the results of , the dhs promoter-GUS fusions we expressed in tobacco also indicate the involvement of activated eIF5A in processes other than plant senescence. An essential role was postulated for eIF5A in RNA export from the nucleus (see above). Such a mechanism may be required in physiological processes such as those involved in senescence or early development of seedlings. Cell-Specific Localization of HSS in Root Sections of S. vernalis | HSS expression was localized to specialized endodermis and the neighboring parenchyma cells just opposite the phloem. Cells of the endodermis are coated with a lipid (suberin)-containing casparian strip in the radial and transverse walls that produce a physical barrier for apoplastic transport of solutes from the cortex to the central cylinder. We assume that these specialized endodermis and adjacent cortex cells are not only the specific sites of HSS expression but are the intrinsic sites of the biosynthesis of senecionine N-oxide, the backbone structure of the Senecio spp. PAs. This view is corroborated by experimental evidence from earlier physiological and biochemical studies. Thus, in Senecio spp., the roots are not only the specific site of HSS expression but also the exclusive site of de novo PA synthesis . The findings of that endogenously formed homospermidine is exclusively incorporated into PAs and is not a substrate for degrading polyamine oxidases indicate that the formation of homospermidine and the subsequent steps of its utilization for alkaloid biosynthesis should share the same cellular compartment. Our data prove that cell organelles are not involved in compartmentation of homospermidine formation. The arrangement of the HSS-expressing cells vis-a-vis the phloem matches with the phloem-specific root-to-shoot translocation of senecionine N-oxide . To enter a sieve tube, newly synthesized senecionine N-oxide has only to cross the single cell layer of the pericycle (see Fig. ). The mechanism of the root-specific phloem loading of alkaloids is not known, but the polar salt-like senecionine N-oxide was found to be only phloem mobile in PA-synthesizing Senecio spp. . The specific phloem loading with senecionine N-oxide may be either apoplastic or symplastic. We know that PAs are spatially mobile . In whole plants, PAs circulate slowly between plant organs and accumulate transiently at the preferential storage sites (i.e. inflorescences and peripheral stem tissue). Even in Senecio spp. root cultures, there is a continuous translocation of PAs between old and new tissues . In plants and cultured roots, there is no significant loss of PAs over prolonged periods (i.e. >4 weeks with plants and >2 weeks with root cultures), indicating that during their spatial mobility, the PAs are well protected from apoplastic degradation. In root cultures, not even traces of PA N-oxides are lost into the medium. Thus, a PA transport via the apoplast seems to be unlikely. The expression of other alkaloid-specific enzymes in particular plant organs, defined cells, or in cell organelles has been described (for review, see ). In this respect, PAs and nicotine as well as the tropane alkaloids, such as the anticholinergic drugs l-hyoscyamine and scopolamine, found in certain species of the Solanaceae, share some striking similarities. Like PAs, they are exclusively synthesized in the roots and translocated into the shoot. However, in contrast to PAs, they are translocated via the transpiration stream in the xylem . Putrescine N-methyltransferase, the first enzyme of scopolamine biosynthesis, and hyoscyamine 6beta-hydroxylase, the last enzyme of scopolamine biosynthesis, are only expressed in the pericycle of young roots of Hyoscyamus niger and Atropa belladonna . The transformation of H. niger and A. belladonna plants with a GUS fusion construct containing the promoter of hyoscyamine 6beta-hydroxylase revealed GUS expression mainly in those cells of the pericycle, which are adjacent to the primary xylem poles . Thus, like PA biosynthesis, tropane alkaloid formation appears to take place spatially adjacent to the translocation path of the alkaloids. However, in contrast to the pericycle-specific expression of these two enzymes of tropane alkaloid biosynthesis, tropinone reductase I, another specific enzyme of this pathway, is expressed in the endodermis and outer cortex. The differential compartmentation of the biosynthetic enzymes indicates that intermediates must be shuttled between the pericycle and the endodermis to complete the pathway leading to the formation of scopolamine . Complex intercellular compartmentation has also been found in the biosynthesis of the monoterpenoid indole alkaloids in Catharanthus roseus. Specific enzymes of this pathway were localized in at least two different cell types, also requiring intercellular translocation of pathway intermediates . The hypothesis of that already compartmentalized reactions were recruited from primary metabolism to participate in alkaloid biosynthesis cannot be the case for PA biosynthesis. HSS exhibits a tissue-specific expression that is clearly different from the expression pattern of DHS, its evolutionary ancestor. The high specificity of expression of enzymes involved in alkaloid biosynthesis may be an important requirement for the establishment of new secondary pathways in evolution. Although we have good evidence that the HSS-expressing cells are the specific sites of PA biosynthesis, a similar intercellular compartmentation of the biosynthetic enzymes, as suggested for tropane and monoterpenoid indole alkaloid biosynthesis, cannot be excluded. Further enzymes involved in PA biosynthesis need to be identified and localized. MATERIALS AND METHODS : Polyclonal Antibody Preparation and Affinity Purification | Polyclonal sera were raised in rabbits against HSS purified from roots of Senecio vernalis Waldst. & Kitaibel and against purified recombinant DHS protein , provided by Eurogentec (Seraing, Belgium) and Bioscience (Gottingen, Germany), respectively. For affinity purification of the sera, purified recombinant HSS and DHS of S. vernalis were coupled to activated Sepharose 4B (Amersham Biosciences, Freiburg, Germany) according to the manufacturer's instructions. These affinity matrices were incubated overnight at room temperature with the respective sera, washed with 0.1 m sodium acetate (pH 4.5) containing 0.5 m NaCl, and eluted with 0.2 m sodium acetate (pH 2.7) containing 0.5 m NaCl. The eluting purified antibodies were rebuffered to phosphate buffered saline (PBS), concentrated, and stored at -20C until further use. Protein Gel-Blot Analysis | Proteins were separated on 12% (w/v) SDS-PAGE gels using a discontinuous buffer system at 200 V of constant voltage. Gels not used for immunoblotting were stained with Coomassie Blue. Protein gels were electroblotted onto polyvinylidene fluoride membrane (Immobilon P, Millipore, Bedford, MA) with a current density of 2.5 mA cm2. The blots were then blocked with Tris-buffered saline supplemented with 0.1% (v/v) Tween 20 (TBS-T) containing 10% (v/v) fetal calf serum (Sigma, St. Louis) for 1 h at room temperature. With the affinity-purified polyclonal antibody (diluted 1:20,000 [v/v] in blocking solution), the blot was incubated for 1 h at room temperature, followed by successive washing steps (each 2 x 5 min) in TBS-T, TBS-T + 0.5 m NaCl, TBS-T + 0.5% (v/v) Triton X-100, and once in TBS-T. After incubation with a goat anti-rabbit secondary antibody conjugated to horseradish peroxidase (diluted 1:5,000 [v/v], Dianova, Hamburg, Germany), the washing steps were repeated before chemoluminescence detection was performed with the ECL Western Blotting System (Amersham Biosciences) and documented on XAR5 x-ray film (Eastman-Kodak, Rochester, NY). RNA Isolation and Gel-Blot Analysis | Tissue samples were collected, frozen in liquid nitrogen, and stored at -80C. Total RNA was extracted with RNeasy Plant Mini Kit (Qiagen, Hilden, Germany). Ten micrograms per sample was separated on a formaldehyde-agarose gel and transferred onto positively charged nylon membranes (Roche Diagnostics, Mannheim, Germany) by capillary blotting. RNA gel blots were hybridized overnight in DIG Easy Hyb buffer or High-SDS hybridization buffer at 42C with HSS and DHS probes and at 39C with the eIF5A probe, all of which were digoxigenin labeled by using the PCR DIG Probe Synthesis Kit (Roche Diagnostics). Chemoluminescent detection was performed with CSPD (Roche Diagnostics) according to the manufacturer's instructions. Exposure times were 2 h for the HSS- and DHS-specific probes and 1 h for the eIF5A-specific probe. Semiquantitative RT-PCR | Per sample, 2 mug of total RNA was used as template for oligo(T) cDNA synthesis with an oligo(dT)17 primer (0.1 mum, 5'-dGTCGACTCGAGA-ATTC(T)17-3', MWG-Biotech, Ebersberg, Germany) using Superscript II RT (Invitrogen, Carlsbad, CA) in a total volume of 50 muL according to the manufacturer's instructions. PCR was performed with specific primers for HSS and DHS, previously used for amplification of the full-length cDNAs from S. vernalis and tobacco (Nicotiana tabacum; ), Taq DNA polymerase (Invitrogen), and the following temperature program: 5 min at 95C initial denaturation, 40 cycles with 95C for 45 s, 57C for 1 min, and 72C for 2 min. Aliquots of the reaction were taken after cycle 26, 28, 30, 32, 34, and 40 and analyzed by agarose gel electrophoresis. Tissue Preparation for Immunohistochemistry | Roots and other plant organs of S. vernalis and tobacco grown in the greenhouse were cut into small segments (approximately 0.5 --1.0 cm) and immediately fixed for 1 h under reduced pressure in 4% (v/v) formaldehyde (freshly prepared from paraformaldehyde) and 0.2% (v/v) glutaraldehyde in sample buffer (0.05 m potassium phosphate buffer, pH 7.2). Afterward, the samples were washed twice for 10 min in sample buffer, dehydrated in a graded ethanol series, and embedded in Technovit 7100 resin (Heraeus-Kulzer, Hanau, Germany) according to the manufacturer's instructions. Sections (3 --5 mum) were cut with a microtome and mounted on glass slides coated with Teflon (Roth, Karlsruhe, Germany). Immunocytochemical Analysis by UV and Light Microscopy | Sections were blocked at room temperature for 30 min with 0.15 m Gly followed by 30 min with PBS supplemented with 10% (w/v) bovine serum albumin and 0.1% (w/v) fish gelatin. After washing with PBS, the sections were incubated with either pre-immune serum (without dilution) or affinity-purified primary antibody (1:500 dilution [v/v]) diluted with 1% (w/v) bovine serum albumin in PBS for 1 h at 37C in a humid chamber. After five 1-min washings with PBS, the sections were incubated for 1 h at room temperature with the secondary goat anti-rabbit antibody. For immunogold labeling, the secondary antibody was coupled to 18-nm gold particles (diluted 1:20 [v/v], Dianova) and for fluorescence detection, coupled with FITC (1:100 [v/v], Sigma). For visualization in a light microscope (Photomicroscope III, Zeiss, Jena, Germany), gold particle-labeled sections were exposed to a silver enhancement reagent according to the manufacturer's instructions (Amersham Biosciences). FITC-labeled sections were excitated by UV light of 450 to 490 nm and recorded using a Photomicroscope III or Axioskop 2 (Zeiss). Immunocytochemical Analysis by Transmission Electron Microscopy | Sections for transmission electron microscopy were fixed and dehydrated as described above, and then were embedded in Unicryl resin (Plano, Wetzlar, Germany) according to the manufacturer's instructions. Sections of 80 nm were cut with an ultramicrotome and mounted onto nickel grids (300 mesh, Plano) coated with Butvar B-98 (Sigma) according to . Blocking and antibody incubation (goat anti-rabbit with 18-nm gold particles as secondary antibody) was performed as described for light microscopy. After post-staining with 2% (w/v) aqueous uranyl acetate for 20 min, the sections were analyzed using a transmission electron microscope (300 EM, Philips, Eindhoven, The Netherlands). Construction of Promoter-GUS Fusions and Plant Transformation | Genomic DNA was isolated from field-grown tobacco leaves using the hexadecyl-trimethyl-ammonium bromide protocol . To identify the promoter region, a gene walking method was used with primers specific to the known cDNA sequence of tobacco DHS . The 2.0-kb fragment of the genomic DNA directly upstream of the open reading frame was amplified by PCR using the Advantage Genomic PCR Kit (BD Biosciences Clontech, Palo Alto, CA) and two specific primers, with the forward primer including a native HindIII restriction site 1,871 pb upstream of the start-ATG of the DHS gene and the reverse primer encompassing a SpeI site directly behind the start-ATG of the open reading frame. The HindIII-/SpeI-digested PCR product was inserted into pCAMBIA1304, replacing the cauliflower mosaic virus 35S promoter directly in front of an mGFP5*/gusA fusion. The hygromycin resistance cassette of this plasmid was replaced by the ampicillin resistance cassette of pCAMBIA2300. This construct was propagated in Escherichia coli XL1-blue and transformed into competent Agrobacterium tumefaciens strain C58C1:pGV2260 to transform tobacco cv SNN as described by . The nucleotide sequences for the promoter region of the dhs gene of tobacco has been submitted to the EMBL Nucleotide Sequence Database (accession no. ). GUS Reporter Analysis | For histochemical GUS analysis, seedlings and young plantlets of the F1 generation of tobacco plants transformed with the dhs promoter-GFP/GUS fusion were immersed in the GUS reaction buffer according to followed by a slight vacuum infiltration for 20 min. Tissues were incubated at 37C for 6 to 18 h and afterward cleared in 70% (v/v) ethanol. Photos were taken using a stereomicroscope (M8, Wild Heerbrugg, Heerbrugg, Switzerland) with an AxioCam HRc (Zeiss). Backmatter: PMID- 12226486 TI - Differential Expression of a Metallothionein Gene during the Presymbiotic versus the Symbiotic Phase of an Arbuscular Mycorrhizal Fungus AB - A full-length cDNA encoding a metallothionein (MT)-like polypeptide, designated GmarMT1, was identified in an expressed sequence tag collection from germinated spores of the arbuscular mycorrhizal fungus Gigaspora margarita (BEG34). The GmarMT1 gene is composed of two exons separated by an 81-bp intron. It codes for a 65-amino acid polypeptide comprising a plant type 1 MT-like N-terminal domain and a C-terminal domain that is most closely related to an as-yet-uncharacterized fungal MT. As revealed by heterologous complementation assays in yeast, GmarMT1 encodes a functional polypeptide capable of conferring increased tolerance against Cd and Cu. The GmarMT1 RNA is expressed in both presymbiotic spores and symbiotic mycelia, even in the absence of metal exposure, but is significantly less abundant in the latter stage. An opposite pattern was observed upon Cu exposure, which up-regulated GmarMT1 expression in symbiotic mycelia but not in germinated spores. Together, these data provide the first evidence, to our knowledge, for the occurrence in an arbuscular mycorrhizal fungus of a structurally novel MT that is modulated in a metal and life cycle stage-dependent manner and may afford protection against heavy metals (and other types of stress) to both partners of the endomycorrhizal symbiosis. Keywords: Introduction : Contaminated soils and waters are a major threat for the environment and for human health. Within the framework of emerging bioremediation technologies, much attention has been given lately to natural biotools capable of removing, containing, or detoxifying various environmental pollutants . Basic and applied research has been mainly directed toward plants and bacteria , whereas the kingdom of fungi, which represents a very versatile microbial entity, has remained largely unexplored. The extraordinary genetic and physiological diversity of fungi ---particularly the richness of species found in soil ---makes them key components of almost all ecosystems, with a great potential for bioremediation purposes. Especially interesting within this kingdom are mycorrhizal fungi, which represent direct links between plants and soil and which are often needed to ensure plant survival in heavily polluted areas. Arbuscular mycorrhizal (AM) fungi (Glomeromycota, ), which are ubiquitous root symbionts, are particularly attractive in this regard. In natural and agricultural environments, they significantly contribute to plant growth not only by improving mineral nutrition , but also by protecting plants against a variety of biotic and abiotic stresses, including heavy metal (HM) stress . Several studies indicate that an increased HM tolerance and other beneficial effects are conferred to the host plant by the AM symbiosis . High levels of mycorrhizal colonization have accordingly been documented in agricultural soils with different kinds of HM contamination . Although AM fungi are presently the object of an increasing attention and many international projects are aimed at exploiting their ecological potentialities for bioremediation purposes, no information about the molecular HM tolerance mechanisms operating in these organisms is currently available. A complex network of transport, chelation, and extracellular and intracellular sequestration processes operates to maintain essential metal (e.g. Cu) homeostasis and to minimize the damage caused by nonessential metals (e.g. Cd; ). A variety of membrane transporters controlling the trafficking of both types of metal ions (e.g. ATP-binding cassette-type transporters and the so-called cation diffusion facilitators) have been identified recently in plants and microorganisms . By comparison, only a fairly limited number of intracellular metal chelators have been identified so far. Citrate and His have been shown to act as major Ni chelators, whereas phytochelatins (PCs) and metallothioneins (MTs) are of primary importance in buffering the intracellular concentration of free thiophilic metal ions, such as Cu, Zn, and Cd (; , and refs. therein). The latter two are Cys-rich polypeptides that chelate metal ions through the formation of tetrahedrally coordinated metal-thiolate clusters. However, at variance with PCs, which are synthesized through the ribosome-independent polymerization of reduced glutathione-derived gamma-glutamyl-Cys units, MTs are gene-encoded . Fungal MTs and PCs have been characterized almost exclusively in yeasts. Brewer's yeast (Saccharomyces cerevisiae) contains a multigene MT (CUP1) family, which is mainly involved in Cu detoxification , and the single-copy MT gene CRS5 , but no canonical PC synthase gene . On the other hand, only one sequence annotated as a putative MT (accession no. ) has been reported so far in fission yeast (Schizosaccharomyces pombe), which, however, produces HM-chelating PC peptides through a plant-like PC-synthase enzyme (SpPCS; ; ). Budding yeast cup1 and fission yeast SpPCS disruptants are both metal hypersensitive, as if distinct HM detoxification strategies were predominantly used by these two organisms. Some fungi, however, are able to synthesize both MTs and PCs. This is the case of Candida glabrata, for example, which produces MTs when exposed to toxic concentrations of Cu but produces mainly PCs in response to a Cd stress . Only a few data are available for filamentous fungi . The occurrence of MTs, PCs, or both, in mycorrhizal fungi is still a matter of debate . There are a few reports on MT-like sequences, obtained in the frame of expressed sequence tag projects. cDNAs coding for putative MT-like polypeptides have been identified in the ectomycorrhizal basidiomycete Pisolithus tinctorius and in the AM fungi Gigaspora rosea and Glomus intraradices (accession no. ; M.J. Harrison, personal communication). However, the metal sequestration capacity of these three predicted polypeptides and, thus, their actual MT-like nature have not yet been determined. Here, we describe the identification and functional characterization of GmarMT1, a MT-encoding gene from the AM fungus Gigaspora margarita. Besides documenting for the first time, to our knowledge, the existence of an MT-based HM detoxification machinery in AM fungi, we also show that the GmarMT1 mRNA is differentially expressed in symbiotic versus presymbiotic life cycle stages of G. margarita. These data raise interesting questions as to the possible role of MTs in metal detoxification and/or more general stress responses in AM fungi. RESULTS : Identification and Sequence Analysis of the GmarMT1 cDNA | A full-length cDNA similar to previously described MTs was identified in an expressed sequence tag collection from G. margarita germinated spores. This cDNA, named GmarMT1, codes for a 6.9-kD polypeptide with a predicted pI of 8.6. As shown in Figure , other diagnostic features of GmarMT1 besides its small size are the presence of 14 cysteines, with only one aromatic residue (Tyr-23), on a total of 65 amino acids. Furthermore, all but two of such Cys residues are part of the characteristic MT motif C-X-C (underlined in Fig. ). Also shown in Figure , is the genomic sequence of GmarMT1, obtained from PCR experiments carried out with oligonucleotide primers designed on the 5'- and 3'-untranslated regions of the GmarMT1 cDNA. As revealed by comparison between genomic and cDNA sequences, a small (81-bp) intron, delimited by consensus splice junctions, is present in the GmarMT1 gene at position 200. Figure 1 | Nucleotide and deduced amino acid sequence of GmarMT1. Nucleotide and deduced amino acid sequence of GmarMT1. The intron sequence is shown in lowercase letters; Cys residues are in bold; C-X-C motifs and GmarMT1-specific primers (MT1/MT2) are underlined. Interestingly, the GmarMT1 polypeptide could not be assigned to any of the six, presently classified subfamilies of fungal MTs . Additional insight into the putative G. margarita MT was, thus, obtained from an extended similarity search using the sequence of the predicted GmarMT1 polypeptide as a query. Quite surprisingly, all but one of the 20 most similar sequences identified by this analysis were type 1 MTs from plants. Besides a partial polypeptide sequence from the related fungus G. rosea , the only other non-plant GmarMT1 homolog within this set of sequences is a predicted (but as-yet-unclassified) MT-like polypeptide from the basidiomycete Agaricus bisporus . As revealed by the multiple sequence alignment reported in Figure , the GmarMT1 polypeptide can be subdivided more specifically into two distinct N- and C-terminal domains, each bearing three C-X-C motifs and separated by an approximately 28-amino acid spacer. The first of such domains best fits the sequence pattern of the corresponding domain of plant type 1 MTs even at amino acid residues other than the cysteines. The C-terminal domain and part of the predicted spacer, instead, most closely resemble the corresponding regions of the Agaricus MT-like polypeptide . Figure 2 | Alignment of GmarMT1 with MT-like polypeptides from other organisms. Alignment of GmarMT1 with MT-like polypeptides from other organisms. The polypeptide sequence of GmarMT1 (boxed) was aligned with the partial sequence of a predicted polypeptide from G. rosea and with seven of the best scoring sequences identified by BLAST analysis: Arabidopsis MT1A (National Center for Biotechnology Information [NCBI] accession no. ; ; ); Arabidopsis MT1C (NCBI accession no. ; ) canola (Brassica napus) MT-like (NCBI accession no. ; ); rice (Oryza sativa) MT1 (NCBI accession no. ; ); barley (Hordeum vulgare) MT1 (NCBI accession no. ; ); Mimulus guttatus MT1 (NCBI accession no. ; ); and A. bisporus MT like (NCBI accession no. ; ). Amino acid residues that are identical in at least seven of the nine sequences are shown on a black background, and residues shared by all three fungal sequences are shaded gray; these two types of conserved residues are indicated with uppercase and lowercase letters, respectively, in the consensus pattern shown below the alignment. Gaps introduced to optimize the alignment are indicated by dots. Functional Complementation Assays | yAP-1 is a transcription factor related to the mammalian AP-1 complex that positively controls various genes involved in HM and, more generally, oxidative stress tolerance in yeast . Although yeast MT genes are not direct targets of such an activator, Deltayap-1 mutants are particularly sensitive to Cd and are, thus, suitable to highlight tolerant phenotypes induced by exogenous cDNAs . To test whether GmarMT1 expression confers an increased tolerance to HMs, we transformed a Deltayap-1 yeast mutant strain with the complete open reading frame of the GmarMT1 cDNA placed under the control of a constitutive yeast promoter provided by the expression vector pFL61 . The empty pFL61 vector and the same vector carrying the MT2a cDNA from Arabidopsis (, ; A. Bolchi and S. Ottonello, unpublished data) were used as negative and positive controls, respectively. The various yeast transformants were streaked onto synthetic dextrose (SD)-agar plates containing a linear 0 to 100 mum gradient of CdSO4. As shown in Figure A, cells carrying the pFL61 vector alone grew only a very short distance into the Cd concentration gradient. In contrast, both positive control (MT2a) and GmarMT1 transformed yeast cells grew up to the highest Cd concentration. No difference between the three yeast transformants (pFL61-GmarMT1, pFL61-MT2a, and pFL61) was observed when similar assays were conducted on CuSO4 or NiSO4 gradients (both ranging from 0 to 3 mm, data not shown). The HM protection capacity of GmarMT1, evidenced by the above experiments, was further verified by similar but functionally more direct complementation assays carried out in a yeast mutant (Deltacup1) that is highly sensitive to HMs (especially Cu) because of the complete disruption of the MT-encoding locus CUP1 . Confirming and extending the results obtained with the Deltayap-1 mutant, an increased tolerance to CuSO4 (0 --200 mum gradient; Fig. B) and to CdSO4 (0 --10 mum gradient; data not shown) was conferred to Deltacup1 cells by the GmarMT1 cDNA. Once again, no differential sensitivity was observed when the three yeast transformants were plated onto 0 to 3 mm NiSO4 gradients (data not shown). Figure 3 | Increased HM tolerance conferred by GmarMT1 in metal-hypersensitive yeast mutants. Increased HM tolerance conferred by GmarMT1 in metal-hypersensitive yeast mutants. Deltayap-1 (A) or Deltacup1 (B) yeast mutants harboring the pFL61-GmarMT1 plasmid (GmarMT1), the positive control plasmid pFL61-Mt2a (AtMt2a), or the empty pFL61 vector (-) were grown on SD-agar (-uracil) plates with the indicated linear gradients of CdSO4 (A) or CuSO4 (B). GmarMT1 mRNA Expression Analysis | GmarMT1 mRNA expression levels were next analyzed by reverse transcriptase (RT)-PCR. Sequence-specific amplification primers designed on the GmarMT1 sequence (MT1/MT2; underlined in Fig. ) were initially tested on total DNA extracted from the host plant white clover (Trifolium repens). The negative results obtained from such control amplifications (data not shown) allowed us to exclude any cross-hybridization with the plant genome. Total RNA extracted from quiescent spores, germinated spores, and mycorrhizal roots was then reverse-transcribed and used for RT-PCR analysis. The amount of cDNA obtained from different fungal samples was first quantified by amplifying a small aliquot with fungus-specific 18S rRNA primers (R1/R2) that did not recognize any sequence in the plant genome (data not shown). Balanced amounts of each cDNA sample were finally amplified with GmarMT1-specific oligonucleotide primers. As shown in Figure , an amplified fragment of the expected size (174 bp) was obtained from all cDNA samples, thus, indicating that the GmarMT1 gene is expressed in all three stages of the G. margarita life cycle. The amount of amplified product, however, was considerably higher (7- to 8-fold) in quiescent and germinated spores than in symbiotic mycelia . Such a difference in GmarMT1 mRNA abundance was reproducibly observed regardless of the extent of AM colonization, which in different experiments varied from 30% to 60% of colonized segments. Figure 4 | RT-PCR analysis of GmarMT1 mRNA levels in presymbiotic and symbiotic life cycle stages of G. margarita RT-PCR analysis of GmarMT1 mRNA levels in presymbiotic and symbiotic life cycle stages of G. margarita. Balanced amounts of cDNA from quiescent spores (lane 1), germinated spores (lane 2), or mycorrhizal roots (lane 3) were amplified with GmarMT1-specific oligonucleotide primers (MT1). 18S rDNA amplicons, obtained from parallel control reactions and loaded in the same order as above, were used as internal standards (18S). The sizes of GmarMT1 and 18S rDNA amplicons are indicated. No cDNA template was added to reaction mixtures run in lane 4; DNA size markers (HaeIII-digested pUC18) were run in lane M. The metal responsiveness of GmarMT1 was then investigated by measuring mRNA levels in germinated spores and symbiotic mycelia exposed for 24 h to increasing concentrations of either CuSO4 or CdSO4. Data reported in Figure a show that Cu did not exert any appreciable effect on GmarMT1 expression in germinated spores, and the same result was obtained after Cd exposure (data not shown). GmarMT1 mRNA levels in symbiotic mycelia were similarly not affected by a Cd treatment (Fig. b). When exposed to Cu, instead, the same symbiotic mycelia exhibited a significant increase in GmarMT1 mRNA abundance (approximately 4-fold) even at the lowest Cu concentration (100 mum; Fig. c). Figure 5 | GmarMT1 mRNA expression levels after metal exposure. GmarMT1 mRNA expression levels after metal exposure. Balanced amounts of total RNA extracted from germinated spores (a) and mycorrhizal roots (b and c), either untreated or exposed to the indicated concentrations of CuSO4 or CdSO4, were reverse-transcribed and amplified with GmarMT1-specific primers (MT1). The 18S rRNA was used as an internal calibration standard for all reactions (18S). The sizes and migration positions of GmarMT1 (MT1) and 18S rDNA (18S) amplicons are indicated. Template RNA was omitted from reaction mixtures shown in lane -; DNA size markers (HaeIII-digested pUC18) were run in lane M. DISCUSSION : MTs are ubiquitous Cys-rich metalloproteins with a characteristically low (or null) aromatic amino acid content . The possible occurrence of MT-like polypeptides in AM fungi was first suggested by electron energy loss spectroscopic analyses, which revealed the presence of polyphosphate granules containing S and N together with Al, Fe, Ti, and B in the fungal cytoplasm . However, neither the potential metal-binding polypeptides contained in such granules nor the genes coding for them were identified thereafter. The results of this work thus provide the first direct evidence for the occurrence of a functional MT gene in the genome of an AM fungus. GmarMT1 Codes for a Functional MT | Standardized protocols for the genetic transformation of Glomalean fungi are not yet available . We, thus, resorted to functional complementation assays in yeast to unambiguously demonstrate that the GmarMT1 gene product can sequester metal ions, thereby conferring in vivo protection against HMs. Two distinct metal hypersensitive mutants (Deltayap-1 and Deltacup1) and three different HMs (Cd, Cu, and Ni) were used for these assays. Only a Cd tolerance phenotype was observed in GmarMT1-transformed, Deltayap-1 mutants. The Deltayap-1 mutant is impaired in reduced glutathione/glutaredoxin and thioredoxin but not in MT synthesis . Therefore, it is likely endowed with a residual tolerance against Cu (the preferred ligand of yeast MTs) that is much too high to allow the detection of an increased Cu tolerance phenotype conferred by a transgene. Such a phenotype, in fact, was observed in an MT-deficient, Deltacup1 mutant , which upon GmarMT1 transformation became more resistant to both Cu and Cd. As furthermore expected for a thionein metal-binding protein, GmarMT1 failed to confer to either mutant an increased resistance against Ni, a metal with an exceedingly low affinity for thiolate groups. GmarMT1 Identifies a Novel Class of Fungal MT-Like Polypeptides | The G. margarita MT displays fairly unique sequence features that distinguish it from known homologs of either fungal or plant origin . Its N-terminal domain is most similar to type 1 MTs from plants , whereas a much closer resemblance with a predicted fungal MT from A. bisporus is observed at the level of its C-terminal domain. The GmarMT1 polypeptide, thus, appears to comprise two evolutionarily distinct domains, one of which is more closely related to plant MTs than to any known fungal MT. Because of this peculiar structural organization and region-specific similarity with an as-yet-unclassified fungal MT, GmarMT1 could not be assigned to any of the presently categorized fungal MT subfamilies . Additional proof of the authentic AM fungal origin of GmarMT1 was provided by the isolation of its corresponding genomic sequence. This also revealed the presence of a short intron separating the region coding for the predicted C-terminal domain (exon 2) from the rest of the coding sequence (exon 1). Besides the N-terminal domain, the first exon of GmarMT1 encodes a putative intervening spacer that comprises the only aromatic residue found in the entire protein. Interestingly, a similar two-exon organization with a centrally located intron and a conserved Tyr residue within the spacer is also found in the genes for Arabidopsis (and other plants) MTs. At variance with GmarMT1, however, Arabidopsis exon 1 only encodes for the N-terminal domain, whereas the spacer and the C-terminal domain are both encoded by exon 2. In many eukaryotes, including Brewer's yeast, MT genes are organized into tandemly repeated multigene clusters . So far, the extremely limited amount of genomic DNA that can be extracted from the spores of G. margarita (an obligate biotroph) has unfortunately precluded a direct examination of MT gene multiplicity in this organism. Differential Expression of the GmarMT1 mRNA | GmarMT1 is expressed in presymbiotic spores and in symbiotic mycelia. Lower expression levels, however, were consistently measured in mycorrhizal roots (sampled at different colonization densities) compared with spores . Considering that MTs are known to respond to a variety of stresses besides HM exposure , one may imagine that such an expression pattern somehow reflects a generalized stress situation occurring within the "free-living" spores of an obligate biotrophic organism. This situation, in which the fungus suffers from both C and N starvation , contrasts with the more favorable conditions experienced inside the roots by the symbiotic mycelium . Interestingly, Glc starvation is known to activate MT (CUP1) genes in Brewer's yeast , and various other metal-unrelated stresses (e.g. wounding, pathogen infection, and senescence) similarly up-regulate MT expression in both plants and fungi . Two features shared by these seemingly diverse stress conditions are an increased production of activated oxygen species and a frequent occurrence of cell death events. Because of their limited life span (and compromised metabolic situation), it is likely that similar events also occur in germinating AM spores. After germination, their hyphae elongate for about 15 to 20 d, after which many morphological modifications occur, including cytoplasmic retraction, production of septa, development of lateral branches, and swollen apices. All of these events correlate with a compromised metabolic situation and the arrest in hyphal growth . Figure 6 | Scheme of the differential expression and HM-induction of GmarMT1 in two phases of the G. margarita Scheme of the differential expression and HM-induction of GmarMT1 in two phases of the G. margarita life cycle. The steps illustrated are germinating spores of G. margarita as seen under the stereomicroscope (presymbiotic phase; bar = 300 mum) and an arbuscule visualized by fluorescence microscopy (symbiotic phase; bar = 5 mum). The GmarMT1 mRNA is selectively up-regulated by Cu ions in the symbiotic mycelium. In keeping with this view, oxidative stress response genes, such as those coding for superoxide dismutase and glutathione S-transferase, were found to be expressed at high levels in germinated spores of G. margarita (; L. Lanfranco, unpublished data). A generalized, and preexisting activation of stress response genes in spores might also explain the lack of metal-induced GmarMT1 mRNA up-regulation in this particular stage of the fungus life cycle. Metal-dependent up-regulation, in fact, was only observed in symbiotic mycelia , as if a basal GmarMT1 expression state, such as the one associated to this more favorable growth condition, were a necessary prerequisite for responding to the stress ensuing from metal exposure. An alternative explanation is that GmarMT1 up-regulation in symbiotic mycelia might be caused indirectly by a secondary stress resulting from metal-induced damage to the host root. Similarly to what has been reported for other unicellular and multicellular fungi relying on MTs to control Cu homeostasis , GmarMT1 responded to Cu, but not to Cd exposure . This selectivity of response, which contrasts with the joint protective effect observed upon constitutive overexpression of GmarMT1 in metal-hypersensitive yeast mutants, likely reflects the Cu specificity of the trans-acting factors regulating GmarMT1 expression in its natural host. It will, thus, be interesting to identify putative metal (and other stress) response elements in the GmarMT1 promoter and to test their role in controlling MT expression by either homologous or heterologous reporter gene experiments. MTs and HM Tolerance in AM Fungi | In both Brewer's yeast and C. glabrata , MTs have been shown to be the chief effectors of Cu tolerance. By contrast, a plant-like PC synthase and a P1-type ATPase seem to play major roles in the detoxification of Cu in fission yeast and C. albicans, respectively. This variety of metal protection strategies, which sometimes coexist in a single organism, raises a question as to the actual role of GmarMT1 in Cu detoxification and to the possible existence in AM fungi of additional HM protection mechanisms, besides MTs. A second, more general question, prompted by the obligate symbiotic and endomycorrhizal nature of AM fungi and by the metal protective effects that are thought to be exerted by plant MTs , is whether mycobiont-derived MTs, such as GmarMT1, can confer an increased HM tolerance to host plants. What is becoming increasingly clear, however, is that mycorrhizal fungi contribute not only to plant nutrition, but also to the uptake and detoxification of various environmental pollutants, including HMs . Understanding the molecular bases of HM tolerance in this symbiotic system will, thus, also aid the selection of the most effective AM isolates and plant-fungus combinations for bioremediation and soil protection purposes. MATERIALS AND METHODS : Biological Materials | Spores of Gigaspora margarita (BEG 34) were collected from pot cultures of mycorrhizal white clover (Trifolium repens) and sterilized with 4% (w/v) chloramine T/0.004% (w/v) streptomycin, plus four rounds of sonication. To induce germination, spores were incubated in water at 24C for 2 weeks. Mycorrhizal roots were obtained from pot cultures of white clover. The percentage of total infected root lengths was evaluated with the grid-line intersect method of . Mycorrhizal plants were collected from pot cultures, and cleaned roots were submerged for 24 h in water solutions containing increasing amounts of CuSO4 (100 and 300 mum) or CdSO4 (50, 150, and 300 mum). Germinated spores were exposed for the same length of time to 500 mum CuSO4 or 300 mum CdSO4 dissolved in water. The G. margarita (germinated spores) cDNA library, with an estimated complexity of 50,000 recombinant clones, was constructed into the lambdaTriplEx II vector using the SMART cDNA library construction kit (CLONTECH Laboratories, Palo Alto, CA). Individual clones were randomly selected and sequenced . PCR Amplifications | Genomic DNA was extracted from spores, mycorrhizal roots, or leaves as described by . PCR reactions were carried out in a final volume of 50 muL containing 10 mm Tris-HCl, pH 8.3, 50 mm KCl, 1.1 mm MgCl2, 0.01% (w/v) gelatin, 200 mum of each dNTP, 1 mum of each primer, 50 to 100 ng of genomic DNA, and 2 units of REDTaq DNA polymerase (Sigma, St. Louis). The PCR program was as follows: 95C for 3 min (1 cycle), 92C for 45 s, 45 s annealing at temperatures indicated below, 72C for 45 s (30 cycles), and 72C for 5 min (1 cycle). The oligonucleotide primers MTG1 (5'-ATCAAATAAGTATATCTCTC) and MTG2 (5'-TTTGAACCCAATATACAACG) were employed for genomic GmarMT1 amplifications at an annealing temperature of 48C. The primers MT1 (5'-TGTGGTTCCGCTTGTCAATG) and MT2 (5'-TTTACAGTTGCCTTTGGTGC) were used to amplify a specific subregion of the GmarMT1 coding sequence at an annealing temperature of 60C. R1 (5'-GAATTTCTACCTTCTGGGGAAC) and R2 (5'-TCGACCATTCAAAAGAATAGCCTG), two oligonucleotide primers specifically recognizing the G. margarita 18S ribosomal gene were used at an annealing temperature of 60C. PCR products were separated on 1.2% (w/v) agarose gels and visualized by ethidium bromide staining. Negative controls for all PCR experiments consisted of reaction mixtures from which template DNA was omitted. Cloning and Sequencing | The PCR product amplified from genomic DNA was extracted and purified from agarose gels using the QIAEX II Gel Extraction Kit (Qiagen USA, Valencia, CA) and directly cloned into the pGEM-T vector (Promega, Madison, WI). XL-2 Blue ultracompetent cells (Stratagene, La Jolla, CA) were transformed and plated onto selective medium following the manufacturer's instructions. Plasmid DNAs were prepared with the Qiagen Plasmid Mini Kit. DNA sequences were determined by MWG Biotech (Ebersberg, Germany) using T7 and Sp6 primers. The genomic sequence of GmarMT1 has been submitted to the DNA data bank of Japan/EMBL/GenBank databases under accession number . DNA sequence analyses were performed with the PC/gene software (IntelliGenetics, Mountain View, CA) and the BLASTX software available through the NCBI. RT-PCR Analyses | RNA was extracted from about 100 spores or 100 mg (fresh weight) of mycorrhizal roots using the SV Total RNA Isolation System kit (Promega). Before RT-PCR experiments, RNA was treated (30 min, 37C) with RNase-free DNase (Amersham, Buckinghamshire, UK), extracted once with phenol:chloroform:isoamylalcohol (25:24:1, v/v), precipitated with 2 m LiCl, and resuspended in 20 muL of sterile diethyl pyrocarbonate-treated, double-distilled water. All of the above RNA samples were routinely checked for DNA contamination by RT-PCR analyses conducted with the 18S rRNA universal primers NS1/NS2 , in the presence or absence of RT. When appropriate, reverse transcription reactions were performed in a final volume of 20 muL containing 50 mm Tris-HCl, pH 8.3, 75 mm KCl, 3 mm MgCl2, 10 mm dithiothreitol, 1 mum dNTPS, 1 unit of RNaseOUT (Invitrogen, Carlsbad, CA), 500 ng of the ribosomal primer NS2 , 1 muL of total RNA, and 200 units of SuperScript II RT (Invitrogen). After 50 min at 42C, a PCR mix was added consisting of 10 mm Tris-HCl, pH 8.3, 50 mm KCl, 1.1 mm MgCl2, 0.01% (w/v) gelatin, 200 mum of each dNTP, 500 ng of the NS1 primer, and 2 units of REDTaq DNA polymerase (Sigma). The same PCR mix, but with both NS primers, was added to parallel samples not subjected to reverse transcription. The PCR program was as follows: 95C for 3 min (1 cycle), 92C for 45 s, 50C for 45 s, 72C for 45 s (30 cycles), and 72C for 5 min (1 cycle). An essentially identical RT-PCR protocol, except for the use of random primers (Invitrogen) and a higher amount (8 muL) of total RNA in the initial reverse transcription step, was used for GmarMT1 mRNA determinations. The amount of reverse-transcribable RNA contained in the different samples was first determined by RT-PCR using oligonucleotide primers (R1/R2; see above) specific for the G. margarita 18S rRNA. Internally balanced amounts of RNA and the sequence-specific primers MT1/MT2 (see above) were then used to amplify the GmarMT1 cDNA. PCR reactions (annealing temperature of 60C) were allowed to proceed for different number of cycles to determine the exponential phase of amplification. The One-Step RT-PCR kit (Qiagen USA) was used for RT-PCR experiments conducted on RNA extracted from HM-treated samples. Reactions were carried out in a final volume of 25 muL containing 5 muL of 5x buffer, 5 muL of Q-solution, 400 mum dNTPs, 0.6 mum of each primer, 1 muL of One-Step RT-PCR enzyme mix, and 0.1 to 2 muL of total RNA. Samples were incubated for 30 min at 50C, followed by a 15 min incubation at 95C. Amplification reactions (92C for 45 s, 60C for 45 s, and 72C for 45 s) were run for a maximum of 26 cycles. RT-PCR products were quantified by densitometric analysis of ethidium bromide-stained bands (Multi-Analyst/PC software, Bio-Rad, Hercules, CA); RT-PCR experiments were conducted in triplicate on three independent samples. Yeast Complementation Assays | The full-length GmarMT1 sequence was first amplified under standard PCR conditions using the NotI site-containing primers MTA (5'-TGACATTGCGGCCGCAAAATGTGCCAAAATTGTA) and MTB (5'-ACTTCGAGCGGCCGCAGAAATTAGCATTTACAGT) at an annealing temperature of 50C. The resulting product was digested with NotI and cloned into the dephosphorylated NotI site of the yeast expression vector pFL61 . The pFL61-GmarMT1 construct, the empty pFL61 vector, and a positive control pFL61 derivative carrying the cDNA for the Arabidopsis MT2a MT (A. Bolchi and S. Ottonello, unpublished data) were then transformed into chemically competent Deltayap-1 (strain WYT; ) or Deltacup1 (strain DTY 113; ) yeast mutants. The various transformants were grown at 30C for 3 d on selective (-uracil) SD-agar medium, before being transferred to SD-agar plates containing linear concentration gradients of Cd, Cu, or Ni . Gradient plate assays, performed at the HM concentrations specified in "Results," were conducted in triplicate. Backmatter: PMID- 12226487 TI - VFL, the Grapevine FLORICAULA/LEAFY Ortholog, Is Expressed in Meristematic Regions Independently of Their Fate AB - The flowering process in grapevine (Vitis vinifera) takes place in buds and extends for two consecutive growing seasons. To understand the genetic and molecular mechanisms underlying this process, we have characterized grapevine bud development, cloned the grapevine FLORICAULA/LEAFY (FLO/LFY) ortholog, VFL, and analyzed its expression patterns during vegetative and reproductive development. Flowering induction takes place during the first season. Upon induction, the shoot apical meristem begins to produce lateral meristems that will give rise to either inflorescences or tendrils. During the second season, after a winter dormancy period, buds reactivate and inflorescence meristems give rise to flower meristems. VFL is expressed in lateral meristems that give rise to inflorescence and flower meristems, consistent with a role in reproductive development. Furthermore, VFL is also detected in other meristematic regions such as the vegetative shoot apical meristem and the lateral meristems that will give rise to tendrils. VFL is also expressed in leaf primordia and in growing leaf margins until later stages of development. Accumulation of VFL transcripts in cell-proliferating regions suggests a role for VFL not only in flower meristem specification, but also in the maintenance of indeterminacy before the differentiation of derivatives of the apical meristem: flowers, leaves, or tendrils. Keywords: Introduction : Genetic and molecular approaches in snapdragon (Antirrhinum majus) and Arabidopsis have allowed the identification of some of the key genes regulating flowering induction and reproductive development . Among them, FLORICAULA (FLO) from snapdragon and its Arabidopsis ortholog, LEAFY (LFY; ; ), seem to have a central role in the specification of flower meristem identity. Inactivation of FLO causes the development of indeterminate shoots in place of flowers , whereas lfy mutants show partial flower-to-shoot conversions . Constitutive expression of LFY is sufficient to promote flower initiation and development from shoot apical and axillary meristems in Arabidopsis and has similar effects in other dicot and monocot species , suggesting a conservation of LFY function across long phylogenetic distances within angiosperms. Despite FLO/LFY sequence conservation among distantly related species , significant differences are emerging in relation to their expression patterns that could indicate the existence of a functional divergence. For instance, although in snapdragon expression of FLO is specific of the reproductive phase (bracts and young floral meristems; ), low levels of expression of FLO/LFY orthologs have been detected in leaf primordia during vegetative growth in Arabidopsis, tobacco (Nicotiana tabacum), Impatiens sp., pea (Pisum sativum), petunia (Petunia hybrida), and tomato (Lycopersicon esculentum; ; ; ; ; ; ; ). Consistent with this evidence, a role in leaf development has been proposed for pea and tomato FLO/LFY orthologs based on the morphological alterations shown in the leaves of loss-of-function mutants. In monocots, expression of FLO/LFY orthologs further deviates from that of their dicots counterparts. In rice (Oryza sativa), the RLF gene is expressed in young panicles but not in mature florets or leaves, thus, making it unlikely its involvement in flower meristem initiation . In Lolium temulentum, LtLFY is expressed later than the SQUA/AP1 ortholog, LtMADS2 (other genes involved in the initiation of floral development; ). In contrast, FLO/LFY precedes SQUA/AP1 expression and LFY is required for AP1 up-regulation . Regulation of flowering in woody perennials shows remarkable differences with respect to herbaceous species, i.e. long juvenile phases, winter bud dormancy and the need of two consecutive growing seasons for flowering. Despite the interest of these processes for the management and improvement of woody species, very little is known about their underlying molecular mechanisms. FLO/LFY orthologs have been cloned and characterized in several woody species such as eucalyptus (Eucalyptus globulus; ), Monterey pine (Pinus radiata; ; ), Populus trichocarpa , and kiwifruit (Actinidia deliciosa; ). However, its specific role in the characteristic features of tree reproductive development is still being elucidated. Furthermore, partial or total FLO/LFY-like sequences have been reported from other basal angiosperms and gymnosperms , although in these cases functional information is not available. We are interested in the reproductive development of grapevine (Vitis vinifera), an important fruit crop in temperate regions that belongs to the family of Vitaceae, a basal family within the eudicots . Grapevine plants grown from seeds go through 2 to 5 years of juvenile phase before they start to flower. In grapevine, flowering requires two growing seasons: flowering is induced in latent buds during the summer but flower initiation and development takes place the following spring . Grapevine, similar to other Vitaceae species, produces either inflorescences or tendrils opposite to leaves. Interestingly, inflorescences and tendrils can substitute for each other depending on the environmental conditions or hormonal treatments , with gibberellins inhibiting inflorescence and promoting tendril development . To understand the genetic and molecular mechanisms underlying the flowering process in grapevine, we have analyzed by scanning electron microscopy (SEM) the development of buds during two growing seasons and have related their development with the temporal and spatial expression patterns of VFL, the grapevine FLO/LFY ortholog gene. Cloning and characterization of VFL indicate that it is a single-copy gene, as in other angiosperm species. In situ hybridization experiments shows that VFL is expressed in lateral meristems irrespective of meristematic fate. These results could suggest that VFL is involved not only in flower initiation and development but also in inflorescence, leaf, and tendril development. RESULTS : Flower Initiation in Grapevine | In grapevine, bud development, from the time of bud initiation until the development of flowers, extends during two growing seasons . This process has been described using SEM in the grapevine var Shiraz . We have used a similar strategy to monitor the development of grapevine var Tempranillo buds during two consecutive seasons. Buds collected at each stage were both analyzed by SEM and used to study VFL expression by in situ hybridization. During the first season, buds are first detectable around March in the axils of current year's young leaves (Fig. A). They are formed by several shoot apical meristems (SAM) protected by bracts. The earliest-formed meristem usually develops as a lateral shoot during this first season. The others remain latent and one of them will give rise to next season's crop of grapes. The whole process described below takes place within this bud (Fig. , A --E). Figure 1 | Stages of development in grapevine. Stages of development in grapevine. A, Newly formed latent bud in the axil of a young leaf. B, Winter bud; phenological stage A according to . C, Swelling bud; phenological stage B. D, Sprouting bud; phenological stage C. E, Phenological stage D. F, Outgrowing shoot; phenological stage E. G, The inflorescences are clearly visible and separated; phenological stage G. H, General view of a growing cane bearing inflorescences with developing flowers at phenological stage H. During the first months of development (March --May; Fig. A), the SAM goes through a vegetative phase producing three to four leaf primordia with distichous phyllotaxis (Fig. A). A pair of scales flanks each leaf primordium. Around May to June, the SAM begins to produce lateral meristems opposite to leaf primordia (Fig. B). Under normal conditions, the first two to three lateral meristems will give rise to inflorescences and the rest to tendrils (Fig. , C and D; , ). The SAM continues to produce two consecutive nodes containing opposed leaf primordia and lateral meristems, which alternate with one node bearing a solitary leaf primordium (Fig. D; see also Figs. H and ). Inflorescence meristems grow rapidly during the summer (June --August). First, a bract is formed at the region farthest from the apex (Fig. D). Then, the inflorescence meristem splits into two meristems (not shown) and each new inflorescence branch meristem begins to produce additional branch meristems subtended by bracts in a spiral phyllotaxy (Fig. C). During this first season, there is no internode elongation, so at the end of summer, the bud encloses a compressed shoot with inflorescence meristems and developing tendrils and leaves (Fig. D). These buds are protected from desiccation and freezing by bracts, scales, and epidermal hairs and enter dormancy around September (Fig. B). Figure 2 | Development of a grapevine var Tempranillo bud and its derivatives as revealed by SEM. Development of a grapevine var Tempranillo bud and its derivatives as revealed by SEM. A through D, First season; E through J, second season. A, Detail of an April latent bud. The vegetative shoot apical meristem is forming leaf primordia flanked by scales (sc) in spiral phyllotaxis. B, Detail of a June bud. The shoot apical meristem has undergone flowering transition and begins to form lateral inflorescence meristem (im) opposite to leaf primordia. C, Inflorescence meristem in August, around the end of the first season. Notice the spiral phyllotaxis of inflorescence branch meristem (ib). One bract (br) subtends each branch. Only some of the ib and br are indicated. D, General view of a July bud showing the derivatives formed by the SAM (sam) during the 1st year. At this stage, the bud encloses developing leaves (lf), inflorescence meristems (im), newly formed leaf primordia (l), and tendril primordia (t). E, Detail of an inflorescence branch in a bud of phenological stage B --C (second season). The inflorescence branch meristem has divided into three to four flower meristems (asterisks). F, Flower meristems derived from an inflorescence branch in a bud of phenological stage C --D. The terminal flower meristem, labeled as 1, is more advanced in development than flanking 2 and 3. G, Developing flower in a bud of phenological stage E. The sepals (sp) grow to enclose the inner part of the flower. H and I, Developing flower at the end of stage E. In I, sepal primordia have been partially removed to show the petal primordia (pt). J, Flower from phenological stage G shoots. The petals have overgrown the calyx (cl). ep, Epidermal hair. All bars represent 50 mum. During the second season, bud growth resumes when the environmental conditions become permissive, around February or March (Fig. C). Upon reactivation, the SAM produces further leaf and tendril primordia. The inflorescence branch meristems form additional inflorescence meristems in a spiral phyllotaxis until mid-March. At the end of March, each racemose inflorescence is formed by many branches that prefigure the conical bunch of grapes. At this stage, each branch meristem divides into a cluster of three to four flower meristems arranged as a dicasium. The terminal flower develops first, then the lateral ones develop, and the basal-most develops the latest (Fig. , E and F). Flower development takes place during April to May (Fig. , D and E), when the bud swells and shoot internodes begin to elongate. Flower meristems sequentially form sepals, petals and stamens, and carpels. Sepal primordia form as a ring of tissue (Fig. G) that grows to enclose the flower bud (Fig. H). Afterward, petal and stamen primordia develop (Fig. I). Petals become fused together by the interlocking of epidermal cells at the petal margins and grow out of the sepal cover (Fig. J). At anthesis, petals cannot separate and they get detached at the base so that the corolla is lifted up and thrown off by the expansion of the filaments of the stamens (not shown, ; ). Cloning of the Grapevine FLO/LFY Ortholog Gene | To analyze how bud development relates to the function of the FLO/LFY grapevine ortholog, we have cloned VFL. The VFL cDNA was isolated by a 3'/5'-RACE strategy. The longest (1,518 bp) cDNA clone (GenBank accession no. ) has a 1,206-bp open reading frame preceded by a 5'-untranslated region of 89 bp. Considerable heterogeneity in length was found in the 3'-untranslated region revealing the use of four different polyadenylation sites at positions 1,393, 1,453, 1,484, and 1,518, respectively. Amplification of the corresponding genomic sequences by PCR showed that the VFL gene contains two introns of 153 and 662 bp (Fig. A) at positions conserved with those described in others species . Genomic DNA-blot hybridization analysis using either a 3' or a 5' probe under high stringency conditions detected a single hybridizing restriction fragment (data not shown). These results suggest the existence of a single VFL gene in the grapevine genome. Figure 3 | The VFL gene and sequence comparison to FLO/LFY-like proteins. The VFL gene and sequence comparison to FLO/LFY-like proteins. A, Genomic organization of VFL (top) and VFL cDNA (bottom). B, The deduced amino acid sequence of VFL was compared with (accession nos. in parentheses): PTLF from Populus balsamifera ; PlaraLFY from Platanus racemosa ; TroLFY from Trochodendron aralioides ; TOFL from tomato ; FLO from snapdragon ; LFY from Arabidopsis ; and ELF1 from eucalyptus (AF34806). Black boxes indicate identical amino acids, shaded boxes similar residues, and dashed lines gaps introduced to optimize the alignment. Sequences were aligned using the ClustalW program. C, Phylogenetic relationship among FLO/LFY-like proteins. The protein sequences shown in B are included, together with: GinLFY from ginkgo (Ginkgo biloba; ); PRFLL from Monterey pine ; NymodLFY from Nymphaea odorata ; LtLFY from L. temulentum ; RFL from rice ; vcLFY from violet cress (Jonopsidium acaule; ); CFL from cucumber (Cucumis sativus; ); UNI from pea ; and NFL1 from tobacco . Bootstrap support values are indicated when over 50. The 1,206-bp open reading frame is predicted to encode a 402-amino acid protein. This predicted protein aligns well with the sequences of other FLO/LFY-like proteins (Fig. B). The VFL protein shares two highly conserved regions with all the FLO/LFY proteins. Furthermore, it shares with other angiosperm FLO/LFY proteins specific regions such as an amino-terminal Pro-rich region (10 Pro residues between positions 14 and 50), a short putative Leu zipper (residues 84, 91, and 98), a basic region formed by a core of Arg and Lys residues from position 180 to 192, and an acidic region (nine aspartic/glutamic residues between positions 197 and 213). The Pro-rich, the acidic, and the basic regions are not so well conserved in gymnosperm FLO/LFY-like proteins. To determine the evolutionary relationships between VFL and other FLO/LFY-like proteins, a phylogenetic tree was constructed using the predicted protein sequences (Fig. C). The general phylogenetic relationships found in this tree are consistent with the currently accepted species phylogenies : The VFL protein is more closely related to other FLO/LFY dicot proteins than to their monocot counterparts or to gymnosperm FLO/LFY-like proteins. Within the dicots, the highest level of identity was found between VFL and FLO/LFY ortholog proteins from P. racemosa (66% identity), T. aralioides (66% identity), P. balsamifera (65%), snapdragon (63%), and several Solanaceae spp. However, the high number of amino acid substitutions precluded a more accurate positioning of grapevine within this group. Expression of VFL during Grapevine Development | To determine the involvement of VFL in grapevine development, we have studied its temporal and spatial patterns of expression in buds and vegetative and reproductive structures. The expression patterns of VFL were first studied by RNA-blot hybridization. RNA levels of VFL were analyzed in axillary buds collected at equivalent branch positions during two consecutive growing seasons (Fig. A). This study revealed that VFL mRNA was already detected in latent buds during the first season (Fig. A, latent buds, lane 3), when inflorescence meristems are being initiated. Expression was barely detectable in dormant buds (Fig. A, stage A, lane 1). During the second season, expression was already detected in advanced stage A (Fig. A, stage A, lanes 2 --3) and increased significantly in stages B and C, when the buds start swelling (Fig. A, stages B and C). At these stages, the inflorescence meristems are dividing to generate new branch meristems and flower meristems. From stage E onward, when the shoot is growing out, VFL expression was analyzed in separate organs (Fig. B). Expression of VFL was mainly detected in flowers of stages E and G (Fig. B). When the branches elongate, during stage H, VFL expression could also be detected in the shoot apex (Fig. B). Figure 4 | Expression of VFL during grapevine development. Expression of VFL during grapevine development. A, VFL expression in latent buds (from June to August) in the first growing season, in winter buds (phenological stage A) and in buds from phenological stages B and C during the spring of the second growing season (see "Materials and Methods" for further details). B, VFL expression in different tissues during cane development (phenological stages E, G, and H). For each lane, 25 mug of total RNA was loaded, blotted, and hybridized with a VFL probe. Filters were also hybridized with 18S rRNA as a quantitative control of loading and blotting. In situ hybridizations were carried out to localize the VFL transcripts during bud development. Expression of VFL was detected in meristematic tissues throughout development. In agreement with the RNA-blot hybridization results, the mRNA levels detected varied during the year and were maximum at the time of flowering induction during the first growing season and at the time of bud reactivation and flower initiation during the second growing season. During the first season, VFL mRNA was detected as early as April at very low levels in the vegetative SAM and at higher levels in leaf primordia (Fig. A). Expression was maintained in the leaf margins until later stages of leaf development (Fig. E). Around May to June, when the SAM begins to produce lateral meristems (Fig. B), VFL levels rose and a strong signal was detected in the dome of those meristems that would become either inflorescences or tendrils (Fig. B). From July to August, VFL was expressed in the inflorescence branch meristems but not in the subtending bract primordia as illustrated in Figure C for second season inflorescences. VFL also accumulated in the apex of developing tendrils (Fig. D). Figure 5 | VFL expression patterns during bud development. VFL expression patterns during bud development. A, Vegetative bud comparable with that in Figure A. VFL accumulates in the inner layers of the SAM and in leaf primordia. B, June bud during floral transition similar to that in Figure B. VFL is expressed in the SAM and in the lateral meristem that will develop an inflorescence. C, Inflorescence of a bud of phenological stage B-C, second season, showing accumulation of VFL in inflorescence branches and in newly formed flower meristems. D, Tendril primordia during the second season. VFL accumulates in the apical region of the tendril that is likely to have meristematic activity. E, Developing leaf. VFL accumulates at the growing tips of the leaf. F, Close up of an inflorescence branch comparable with that in Figure F where flower meristems have just formed. VFL is strongly expressed in newly formed flower meristems and begins to disappear at the regions where sepal primordia form (arrows). G, Flower comparable with that in Figure G. VFL is not detected in sepal primordia but accumulates in the inner part of the flower preferentially in petal primordia. H, Flower comparable with those in Figure , H and I. I, Flower corresponding to a stage slightly earlier than the one shown in Figure J. VFL is detected at very low levels in petals and at higher levels in stamens. VFL mRNA is absent from the bracts throughout development. Nomenclature is the same as in Figure . During the second season, VFL transcripts were already detectable in stage A buds (Fig. B) in the inflorescence branch meristems (Fig. C). However, the expression dramatically increased in the newly formed flower meristems (stages B and C buds; Fig. , C and F). VFL was also expressed in developing floral organs but at lower levels than in inflorescence or flower meristems. When sepal primordia began to grow (Fig. G), VFL expression became restricted to the inner part of the flower meristem that will form petals, stamens, and carpel (Fig. G). Later, VFL was expressed in petal primordia until the petal fuse and, at lower levels, in stamen primordia (Fig. , H and I). DISCUSSION : The basic function of FLO/LFY floral meristem identity genes seems to be conserved in several annual dicot species based on the common aspects of the corresponding mutant phenotypes . However, their role in monocots and basal dicots is not so well established. Moreover, the role of FLO/LFY-like genes in woody perennial species has only started to be established due to the difficulties of the genetic analysis and transgenic approaches in those species. Although the available information suggests that overexpression of LFY is sufficient to promote the conversion of shoots into flowers in species like Populus spp. and Citrus spp. , the role of the endogenous FLO/LFY homologs and their function during meristem development are poorly understood. As a first step to analyze flowering induction and flower initiation in grapevine, we have characterized bud development during this process, cloned the grapevine FLO/LFY ortholog gene and analyzed its expression during vegetative and reproductive development. cDNA cloning and genomic DNA-blot hybridization studies indicate that VFL is a single-copy gene in grapevine as it has been reported for other diploid angiosperms . The number and position of introns found in VFL are identical to those reported for other FLO/LFY-like genes . No alternative splicing sites for the first intron were detected in VFL, in contrast to what has been described for FLO and LFY . Furthermore, analysis of the deduced VFL protein shows the presence of the typical motives found in most FLO/LFY-like proteins . VFL expression is associated to reproductive meristems. During the first season, VFL is expressed in the lateral meristems giving rise to inflorescences and in the developing inflorescence meristems. Winter dormant buds show reduced levels of VFL transcripts, which could be due to either a reduced expression or to a low detectability of the transcripts in these hardy buds. During the second season, the expression levels increase in the proliferating inflorescence meristems generating inflorescence branches, with the highest levels detected in young floral meristems. VFL is also expressed in petal and stamen primordia, and its expression declines as organs expand, as described for other species . In summary, expression of VFL in reproductive meristems and developing floral organs suggests that VFL plays an important role during reproductive development as it has been suggested for most FLO/LFY-like genes studied. Similar temporal patterns of expression spanning two seasons have also been described for the FLO/LFY ortholog in kiwifruit, another woody perennial with winter bud dormancy . However, in the case of kiwifruit, LFY expression levels are higher during the first growing season corresponding to the time of initiation of flower meristems. The expression levels during the second season are lower and coincide with the initiation and development of flower organs. Therefore, in grapevine and kiwifruit, the highest levels of FLO/LFY expression correspond to the time of flower meristem formation (first season in the case of kiwifruit and second season in the case of grapevine), supporting a role for these genes in this process. Expression of VFL is not restricted to reproductive tissues. Low transcription levels are also detectable in the inner cell layers of the vegetative shoot apical meristem and in leaf and tendril primordia. Expression in the vegetative SAM has been reported in violet cress and several Solanaceae species such as tobacco , Impatiens sp. , and tomato , but its role in this meristematic region is still unknown. Leaf primordia and developing leaves also accumulate VFL. Most analyzed FLO/LFY orthologs have been reported to be expressed in developing leaves . In some cases, FLO/LFY expression is restricted to the adaxial side of the developing leaf as in tobacco, eucalyptus, and kiwifruit , but this is not the case for VFL mRNA, which is expressed dorsally and ventrally in leaf primordia. In developing leaves, VFL accumulates at the growing margins. At this region, VFL could be involved in maintaining proliferation and, thus, helping generate the palmate shape of the grapevine leaves. A similar role for FLO/LFY-like genes has been shown in pea where UNIFOLIATA is required to generate the wild-type compound leaves and in tomato where falsiflora mutants have leaves with fewer leaflets than wild-type plants . A unique feature of grapevine is the occurrence of lateral meristems, indistinguishable in morphology and position, that will give rise to inflorescences, tendrils, or even lateral shoots (; this work). Although inflorescence meristems are only seasonally formed as a result of flower induction, tendril meristems are continuously formed together with leaves following the pattern shown in Figure . Poor environmental conditions (e.g. low growing temperatures) can cause lateral meristems to develop as tendrils or even as lateral shoots . In addition, gibberellin and cytokinins have antagonistic effects on the development of these meristems: Gibberellins cause inflorescence meristems to develop as tendrils , and cytokinins can result in the production of inflorescences from tendril meristems. VFL is expressed in all lateral meristems independently of the structure they will form, suggesting that VFL is not sufficient to confer reproductive fate. Thus, the specification of reproductive characteristics may require the observed increase in VFL expression, unknown posttranscriptional regulatory mechanisms or the activity of additional factors. Figure 6 | Bud derivatives in grapevine. Bud derivatives in grapevine. Left, Schematic representation of a latent bud during the first season showing the phyllotaxis of meristems and primordia at this stage. Right, derivatives formed from those meristems and primordia during the second season. Lateral meristems giving rise to inflorescences or tendrils are indistinguishable in morphology and position at the time they are formed. In summary, VFL transcripts accumulate in regions such as meristems and growing primordia that maintain part of its undifferentiated state. Thus, VFL could be involved in maintaining a transient phase of indeterminacy that precedes differentiation of lateral derivatives of the apical meristem: flowers, leaves, or tendrils. A similar role has been previously proposed based on results from other species . This extended expression of FLO/LFY orthologs can be more easily followed in woody perennials such as grapevine than in annual herbaceous plants. Functional studies of the role of VFL in grapevine will be required to demonstrate its function in the development of vegetative and reproductive primordia. MATERIALS AND METHODS : Plant Material | Grapevine (Vitis vinifera L. var Tempranillo) plants were collected in the fields of Instituto Madrileno de Investigaciones Agrarias (Alcala de Henares, Madrid). Young expanding leaves were used for isolation of genomic DNA. RNA-blot and in situ hybridization analyses were performed on plant tissues collected and fixed in different developmental stages during two growing seasons. Parallel samples were also analyzed by SEM. Cloning of VFL | Cloning of VFL was performed using a 3'/5'-RACE strategy , following the instructions of a commercial RACE kit (Roche Diagnostics GmbH, Mannheim, Germany). 3'-RACE was performed with oligo(dT)-primed single-stranded cDNA synthesized from total RNA of developing buds (stage C). Further PCR amplification was performed with anchored primers to the 3'-end and degenerated primers specific to highly conserved regions of FLO/LFY-like genes. These primers were 5'-CAA/GAGGGAGCAT/CCCCTTCATAGTA/GAC-3' for the first amplification and 5'-TACATA/CAACAAGCCGAAA/GATG-3' for the re-amplification. Amplified fragments from two independent experiments were cloned in pGEM-T easy vector (Promega, Madison, WI). Thirty clones corresponding to the 3' region of the gene were sequenced to analyze the sequence diversity at the 3' end. For 5'-RACE, specific single-stranded cDNA was synthesized using a gene-specific primer (5'-AGCAGCAAGGGCGACAAGAGGCTTGTAACA-3'). Then, terminal transferase was used to add a homopolymeric A-tail to the 3' end of the cDNA. Tailed cDNA was amplified with a gene-specific primer (5'-CCAACGTTCTCTCCTCTCTCCTT-3') and an oligo(dT)-anchor primer. The resulting product was re-amplified in a second PCR using a nested, specific primer (5'-GCTCTCCTTAAGGCATTGGAAGCCTCCTCA-3') and an anchor primer. The gene sequence was completed after two successive 5'-RACE experiments, which allowed the isolation of the middle part and the 5' end of the gene, respectively. The complete coding region of the cDNA was obtained by reverse transcriptase-PCR with primers flanking this zone (5'-AGAGATAGAGAGGCAATCAGCAGGATGGAT-3' and 5'-GGGGAATACGTTAAGTTCAGAATGGCAAGT-3'). This cDNA was cloned in pGEM-T easy. To avoid PCR-based errors, six clones, obtained from three independent PCR reactions, were completely sequenced and compared. To analyze the genomic structure of the gene, two overlapping genomic fragments were amplified by PCR on genomic DNA. Amplification products were cloned, and six clones from each fragment were completely sequenced and compared. For sequencing, the Big Dye Terminator Cycle Sequencing kit and a sequencer (Prism 377, ABI, Sunnyvale, CA) were used. Comparisons of sequences of six cDNA clones and six genomic clones revealed a complete identity between them. Only a single-base change that resulted in a synonymous codon (base 221 is G or T) was detected in both the cDNA and the genomic clones. RNA-Blot Hybridization Analyses | For RNA-blot hybridization, plant material was collected at different developmental stages. In the first growing season, young buds in the axils of the leaves (latent buds) were collected every 15 d from June to August (Fig. A, latent buds, lanes 1 --5). For phenological stage A, bud samples were taken in November (dormancy period), February, and March (Fig. A, stage A, lanes 1 --3, respectively). In the second growing season, swelling buds were collected. These samples correspond to buds of early (lanes 1) or advanced (lanes 2) phenological stages B and C (Fig. A). Further in development, when the new cane begins to grow out (phenological stages E, F, G, and H), the different organs of the plant were independently analyzed: shoot apex, young leaves, shoots, tendrils, and flowers. Total RNA extraction was performed following the protocol of . For RNA-blot hybridization analyses, 25 mug of total RNA was loaded per lane on agarose/formaldehyde gels, electrophoretically separated, and transferred to Hybond-N+ membranes. Filters were hybridized with a 32P-radiolabeled VFL 5'-probe, comprising 290 amino acids and the untranslated 5' region. Hybridization was performed overnight at 65C in Church's buffer . Filters were washed several times in 2x SSC and 0.1% (w/v) SDS and once in 0.1x SSC and 0.1% (w/v) SDS for 20 min at 65C. In Situ Hybridization | Digoxigenin labeling of RNA probes, tissue preparation, and hybridization were performed as described by . The template for the VFL digoxigenin-labeled riboprobes was the 1,249-bp fragment, containing the complete coding region, obtained by reverse transcriptase-PCR with the primers indicated above and cloned in pGEM-T easy vector. The hybridized sections were visualized with Nomarski optics in a microscope (DMR, Leica, Wetzlar, Germany). SEM | Samples were collected once every month for 1 year. Samples were dissected in the field and collected in ice-cold fixation buffer (phosphate-buffered saline, 4% [w/v] paraformaldehyde, 0.1% [v/v] Triton, and 0.1% [v/v] Tween 20), vacuum-infiltrated in fresh fixation buffer, and left overnight in the cold room. They were then washed in saline solution and incubated for 24 h in a phosphate-buffered saline solution with 2% (w/v) osmium tetroxide. After washing them once with saline solution, samples were dehydrated through an ethanol series (50%, 70%, 85%, 95%, and 100% [v/v]) and critical-point dried. They were then covered with gold particles and observed under the SEM. Phylogenetic Analyses | To construct the phylogenetic tree, predicted proteins were aligned with ClustalW. Using this original data set, 100 data sets were generated by bootstrap resampling using the SEQBOOT program. Distance matrices were made for each bootstrap data set using the PROTDIST program-Dayhoff PAM matrix algorithm. The distance matrices obtained were used to construct 100 un-rooted trees by the neighbor-joining method using the NEIGHBOR program. A consensus tree was obtained using CONSENSE. SEQBOOT, PROTDIST, NEIGHBOR, and CONSENSE programs belong to the PHYLIP program (Phylogeny Inference Package, version 3.57c, Department of Genetics, University of Washington, Seattle). Backmatter: PMID- 12226488 TI - Effect of Regulated Overexpression of the MADS Domain Factor AGL15 on Flower Senescence and Fruit Maturation AB - We have examined the effect of regulated overexpression of AGL15, a member of the MADS domain family of regulatory factors, on reproductive tissues. Using molecular and physiological markers, we show that constitutive overexpression of AGL15 in Arabidopsis leads to delay and down-regulation of senescence programs in perianth organs and developing fruits and alters the process of seed desiccation. Through genetic crosses, we show that the rate of water loss in the maturing seeds is dictated by the genetic composition and physiological state of the maternal tissue, rather than the embryo. To define the developmental time and/or place when senescence programs are most affected by elevated AGL15 levels, we expressed AGL15 under the control of various promoters. Expression during senescence or in abscission zone cells did not produce delays in floral organ senescence or abscission. Using a glucocorticoid-inducible expression system, we show that an increase in AGL15 levels around the time of flower opening is necessary to delay senescence and increase floral organ longevity. Keywords: Introduction : Members of the MADS domain family of transcriptional regulators participate in a variety of developmental processes in Arabidopsis, including regulation of floral organ identity , inflorescence meristem identity , the transition to flowering , fruit wall differentiation , and lateral root differentiation . All members of this family contain a highly conserved motif of 55 to 60 amino acids, known as the MADS domain, that is essential for the DNA-binding activity of these factors . Many of the family members identified in plants also contain a less-conserved domain called the K domain, which may facilitate protein-protein interactions. The C domain, which constitutes approximately the last one-third of each protein, has a more variable sequence and is likely to contribute to differences in function. Members of the MADS domain family have been shown recently to be important for developmental processes that involve cell separation. For example, SHP1 (SHATTERPROOF1) and SHP2 (SHATTERPROOF2) act redundantly to specify cell identity in the dehiscence zone in Arabidopsis fruits and to promote the lignification of cells adjacent to the dehiscence zone . The MADS domain protein JOINTLESS is necessary to specify pedicel abscission zones in tomato. Abscission zones are absent in the jointless mutant . The Arabidopsis seedstick (agl11) mutant shows defects in the formation of the funiculus, resulting in a persistent connection between the fruit wall and developing seeds . The accumulated information on shp1/shp2, jointless, and stk mutants indicate that these genes are involved in regulating differentiation of abscission or dehiscence zone cells. In a previous study, we showed that constitutive overexpression of the MADS domain protein AGL15 (AGAMOUS like 15) results in delays in abscission of sepals and petals in Arabidopsis . In wild type, AGL15 is expressed in developing embryos, in the vegetative shoot apex, at the base of leaf petioles, and in developing floral organs . Plants overexpressing AGL15 show no structural defects in abscission zones, unlike shp1/shp2, jointless, and stk mutants. If a strong stimulus in the form of exogenous ethylene is added, abscission can occur . In the absence of this stimulus, the plants show a significant delay in floral organ abscission, which is coupled to a delay in floral organ senescence. This combination of features indicates that AGL15's effect on cell separation processes is fundamentally different from that of other MADS domain proteins known to affect abscission or dehiscence. AGL15 preferentially accumulates in young tissues, and our working hypothesis is that endogenous AGL15 operates to maintain or enhance an immature or juvenile or non-senescent state in those tissues. Constitutive overexpression of AGL15 might result in repression of some aspect of programmed senescence in aging floral tissues. Abscission of an organ is often coupled to its senescence, although petals appear to be shed in a turgid state in Arabidopsis . The longevity of perianth organs is dramatically increased when AGL15 is overexpressed . Therefore, plants overexpressing AGL15 present a unique opportunity to explore the control of tissue longevity, the relationship between programmed senescence and abscission, and interactions between senescing organs and adjacent tissues. In the first part of the work reported here, we examined the impact of AGL15 overexpression on senescence programs in floral organs and fruits and explored the relationship between fruit maturation and the process of seed desiccation. We then generated transgenic plants where AGL15 was expressed only in abscission zones or only during senescence to examine the contribution of cell- or program-specific expression to the effect on senescence. Finally, we set up a system where overexpression of AGL15 could be induced by the addition of glucocorticoid. Using this system, we show that AGL15 overexpression must be induced before the onset of senescence and abscission to produce subsequent delays in these processes. RESULTS : The Effect of Overexpression of AGL15 on Perianth Senescence | Arabidopsis plants that constitutively express AGL15 exhibit increased perianth organ longevity , possibly due to some combination of effects on abscission and senescence. To examine the effect of AGL15 overexpression on senescence molecular programs, we used a beta-glucuronidase (GUS) reporter construct to monitor the activity of the SAG12 (Senescence Associated Gene 12) promoter in the sepals and petals. Previous studies showed that expression of SAG12 is correlated with the onset of visible senescence symptoms in both vegetative and reproductive tissues. Plants that were homozygous for an SAG12: GUS reporter construct were crossed to plants that were hemizygous for a 35S:gAGL15 transgene, which drives constitutive expression of AGL15 . Progeny that carried the AGL15 transgene, which represented 50% of the population, could be distinguished from their siblings due to differences in leaf shape and petiole length . In progeny with a wild-type phenotype, SAG12 promoter activity could be detected in stamens, petals, and sepals at 2 d after pollination (DAP) (not shown) and intense GUS staining was visible by 3 DAP (Fig. A). By 4 DAP, the floral organs had generally abscised, and no promoter activity was associated with the remaining tissues of the gynoecium (Fig. B). In progeny showing the AGL15 overexpression phenotype, activation of the SAG12 promoter was delayed in all floral organs (Fig. C). Promoter activity could be detected in the stamens around 4 DAP (not shown). In the sepals and petals however, no sign of strong promoter activity was visible at 4 DAP (Fig. D) or over a 4-week interval after pollination. Faint staining could sometimes be detected at the base of the still-attached perianth organs in older flowers (not shown). We conclude that overexpression of AGL15 affects senescence programs in sepals and petals, leading to down-regulation of the SAG12 gene. Figure 1 | GUS activity in flowers of transgenic plants carrying SAG12:GUS, a senescence reporter construct. GUS activity in flowers of transgenic plants carrying SAG12:GUS, a senescence reporter construct. A and B, GUS activity at 3 (A) and 4 (B) DAP in flowers of plants that do not overexpress AGL15. Perianth organs abscise around 4 DAP in these plants. C and D, Little or no GUS activity can be detected at 3 (C) or 4 (D) DAP in the flowers of plants that carry the 35S:gAGL15 transgene and constitutively overexpress AGL15. Perianth organs are retained for long periods in these plants. The Effect of Overexpression of AGL15 on Fruit Wall Senescence | Overexpression of AGL15 also affects maturation processes in developing fruits. In Arabidopsis, the gynoecium, i.e. fourth whorl floral organs, develops into a silique that undergoes senescence when the seeds become mature. The siliques of plants constitutively overexpressing AGL15 remain green for an extended period of time. To compare the rates of silique maturation in these plants with wild type, we measured changes in chlorophyll content over time. Wild-type siliques contained approximately 1,200 mug chlorophyll g-1 of silique tissue at 9 DAP. The chlorophyll content declined rapidly over the next few days, and was reduced by one-half by 11 to 13 DAP, as shown in Figure A. By 17 DAP, there was approximately 80 mug chlorophyll g-1 of silique tissue. At this stage, the siliques were brownish yellow in color and started to dehisce. Figure 2 | Analysis of fruit maturation in plants that constitutively overexpress AGL15. Analysis of fruit maturation in plants that constitutively overexpress AGL15. A, Chlorophyll content of maturing siliques of wild-type (wt) plants versus plants that constitutively overexpress AGL15 (35S:AGL15). Each data point represents the mean of three experimental replicas. Bars indicate the sd of the mean. B through D, GUS activity in the silique tissues of plants carrying the senescence reporter SAG12:GUS. B, Quantitative assays. Siliques were collected at 17 DAP from plants that do not overexpress AGL15 (gray bar) and at 17, 23, or 25 DAP from plants that overexpress AGL15 (black bars). C and D, Histochemical staining. The siliques on the left in each image were collected at 17 DAP from plants that do not overexpress AGL15. The siliques on the right in each image were collected at 17 (C) or 23 (D) DAP from plants that overexpress AGL15. SAG12 promoter activity was down-regulated and delayed in plants that overexpress AGL15. Siliques from plants that constitutively overexpress AGL15 had a higher chlorophyll content at all stages. As Figure A shows, the maturation process had three distinct phases, marked by different rates of chlorophyll loss. From 9 to 13 DAP, chlorophyll content declined fairly rapidly. From 13 to 15 DAP, chlorophyll content declined very slowly, and then from 15 to 23 DAP, the chlorophyll content declined rapidly again. At 17 DAP, when the siliques of wild-type plants are fully mature, the siliques of plants carrying the AGL15 transgene still contain approximately one-half of the amount of chlorophyll present at 9 DAP. These siliques finally become yellow or brownish yellow around 21 to 23 DAP, when they contained approximately 100 mug chlorophyll g-1 of silique tissue. We conclude that the siliques remain green longer on plants constitutively overexpressing AGL15, partly because they have a higher chlorophyll content initially and partly because chlorophyll loss is slower, particularly during the middle phase of fruit maturation. The effect of AGL15 overexpression on activation of the SAG12 promoter in maturing siliques was also examined. The same population of plants used for the floral organ analysis (35S:gAGL15 segregating in the presence of an SAG12: GUS reporter) was used in this study. In progeny that lacked the 35S:gAGL15 transgene, GUS activity could be detected in tissues of the fruit wall by 16 to 17 DAP (Fig. C). This is quite late in the senescence process and coincides with the final stages of chlorophyll loss in the siliques. In siblings that carried the AGL15 transgene, SAG12 promoter activity could not be detected in the fruits until 21 to 23 DAP (Fig. D), which corresponds to the final stages of chlorophyll loss in these siliques. Thus, based on the chlorophyll loss profiles and SAG12 promoter activity, senescence of fruit tissues appears to be delayed by almost 5 to 6 d when AGL15 is constitutively overexpressed. To determine whether SAG12 promoter activity is down-regulated as well as delayed in the silique tissues, GUS activity was quantified by measuring emitted fluorescence after incubations with 4-methyl umbelliferyl glucuronide. At 17 DAP, the GUS activity in silique tissues of control plants was approximately 10-fold higher than the GUS activity in silique tissues of plants with the AGL15 transgene (Fig. B). The GUS activity increased with age in the silique tissues of the plants constitutively expressing AGL15, but never reached the levels observed in the control plants. The maximal activity, measured at 25 DAP, was approximately 25% of that in control plants at 17 DAP. After 25 DAP, the siliques tended to shatter during harvest, and no further measurements of GUS activity were made. The Effect of Overexpression of AGL15 on Seed Desiccation | What are the consequences of increased silique tissue longevity? One possible effect, because the silique tissues have higher water potentials for a longer period, is an alteration in the decline in water content that marks the later stages of seed development. We compared the water content of seeds collected from self-pollinated wild-type plants and seeds from self-pollinated transgenic plants that contain a single copy of either 35S:AGL15 or 35S:gAGL15 in the hemizygous condition. As shown in Figure A, the initial water content and decline in water content from 9 to 15 DAP were very similar in seeds of wild-type plants (black line) and seeds of plants carrying AGL15 transgenes (blue and red lines). After 15 DAP, the water content of seeds of wild-type plants continued to drop rapidly and reached approximately 5% at 19 DAP. For seeds developing on plants constitutively overexpressing AGL15, the decline in seed water content slowed around 15 DAP. The seed water content was maintained at approximately 30% to 40% from 17 to 23 DAP and then declined rapidly after 23 DAP to reach 6% to 8% at 31 DAP. We conclude that constitutive overexpression of AGL15 leads to alterations in water relations in the developing fruit. Figure 3 | Analysis of seed desiccation in plants that overexpress AGL15. Analysis of seed desiccation in plants that overexpress AGL15. Seeds were excised from maturing siliques at various times (DAP) and seed water content was calculated from measurements of the fresh and dry weight of each sample. A, The water content of seeds of self-pollinated wild-type plants (black line) declines more rapidly than the water content of seeds of self-pollinated transgenic plants that overexpress AGL15. The transgenic plants carried one copy of either of two transgenes (35S:AGL15, blue line; or 35S:gAGL15, red line) in the hemizygous condition. B, The water content of seeds of self-pollinated wild-type plants (black line) was compared with the water content of seeds derived from reciprocal crosses between wild-type plants and plants that overexpress AGL15. The green line shows the water content of seeds derived from crosses where wild-type plants were pollinated with transgenic pollen. The pink line shows the water content of seeds derived from crosses where plants that overexpress AGL15 were pollinated with wild-type pollen. Does expression of the AGL15 transgene in the embryo contribute to this effect? We addressed this question genetically in two ways. First, we pollinated wild-type plants with pollen from transgenic plants carrying a single copy of the 35S:AGL15 transgene in the homozygous condition. The resulting seeds contained embryos with one copy of the AGL15 transgene and wild-type seed coats, and were surrounded by wild-type silique tissues. As shown in Figure B, the decline in water content in the seeds resulting from such crosses (green line) was similar to that in the seeds from self-pollinated wild-type plants (black line). Thus, presence of the AGL15 transgene in embryos is not sufficient to cause a delay in seed desiccation. Second, we pollinated transgenic plants that were hemizygous for the AGL15 transgene with wild-type pollen. This increased the frequency of wild-type embryos in developing fruits from 25% (selfed 35S:AGL15 plant, blue line, Fig. A) to 50% (pink line, Fig. B). However, the rate and pattern of water content decline of the seeds, shown by the pink line in Figure B, did not change. On the basis of these results, we conclude that the genotype of the embryo contributes little to the effect of AGL15 overexpression on seed desiccation. The Effect of Abscission Zone-Specific Expression of AGL15 | In our previous study, we showed that constitutive overexpression of AGL15 leads to delays in perianth organ abscission . Does overexpression of AGL15 in cells outside of the abscission zone contribute to this effect? To address this question, we generated transgenic plants in which overexpression of AGL15 was restricted to the abscission zone cells. To overexpress AGL15 in the abscission zone, we transformed plants with a construct designated Chi: gAGL15, which consisted of the promoter of a bean (Phaseolus vulgaris) chitinase gene fused to sequence encoding AGL15 with 14 additional amino acids at the N terminus. Plants that constitutively overexpress this form of AGL15 show the same phenotype (not shown) as plants that constitutively overexpress unmodified AGL15. The bean chitinase gene promoter shows strong activity in the floral organ abscission zones of Arabidopsis floral organs immediately after bud anthesis . In plants carrying the Chi:gAGL15 construct, AGL15 is overexpressed in the abscission zone cells and targeted to the nuclei, as confirmed with AGL15-specific antibodies (Fig. , A --D). A total of 28 independent transgenic lines were isolated, and a subset of these was followed for two generations. In all cases, floral organs abscised at the same time as in wild type (Fig. , E and F) and earlier than in plants where AGL15 was constitutively overexpressed (Fig. G). To determine whether more subtle changes in tissue integrity had occurred, the petal break strength was measured for successively older flowers along the length of the inflorescence. The petal break strength profiles for individuals from two independent lines, Chi:gAGL15-1 and Chi:gAGL15-2 (each line contains a single Chi:gAGL15 locus in the homozygous condition), were indistinguishable from wild type (Fig. H). We conclude that overexpression of AGL15 in the abscission zone is not sufficient to delay abscission of perianth organs. Figure 4 | Analysis of perianth organ senescence and abscission in transgenic plants that overexpress AGL15 in abscission zones. Analysis of perianth organ senescence and abscission in transgenic plants that overexpress AGL15 in abscission zones. A through D, Immunolocalization of AGL15 in perianth organ abscission zones. No AGL15 accumulation can be detected in cells in the abscission zones in wild-type plants (A and C). AGL15 accumulates at elevated levels in the nuclei of cells in the abscission zones in transgenic plants carrying the Chi:gAGL15 construct (B and D). The location of the abscission zones shown at higher magnification in C and D are indicated by the boxes in A and B. az, Abscission zone; c, carpel; p, petal; r, receptacle; s, sepal. Bars = 20 mum. E through G, Inflorescences of wild-type (E) and transgenic plants carrying either the Chi:gAGL15 transgene (F) or 35S:AGL15 transgene (G). Overexpression of AGL15 in the abscission zones does not produce the visible delay in perianth organ abscission and senescence associated with constitutive overexpression. H, Analysis of the force needed to remove petals in progressively older flowers (in different positions relative to the youngest open flower) in wild-type plants and two lines of transgenic plants carrying the Chi:gAGL15 transgene. Each data point represents the mean of eight to 12 measurements. Bars indicate the sd. The Effect of Senescence-Induced Expression of AGL15 | We have shown that overexpression of AGL15 leads to delays in floral organ senescence and fruit maturation. Does overexpression of AGL15 have the same effect if it occurs after the start of tissue senescence? To address this question, transgenic plants in which AGL15 is expressed under the control of the SAG12 promoter were generated. Previous studies showed that placing expression of bacterial cytokinin synthesis genes under the control of this promoter leads to delayed leaf senescence . We generated a total of 84 independent lines carrying the SAG12:gAGL15 construct, which consists of a fusion between the SAG12 promoter and AGL15 coding sequence. A subset of these lines was followed through two generations. AGL15 mRNA accumulated in senescing tissues in these plants, as confirmed by RNA gel-blot analysis of leaf samples (Fig. A). As Figure A shows, the relative level of mRNA accumulation was comparable with that achieved in younger tissues using the 35S cauliflower mosaic virus (CaMV) promoter. AGL15 mRNA cannot be detected in the younger green leaves of the SAG12:gAGL15 transgenic plants (Fig. A), indicating that accumulation occurs in an age-dependent manner. Figure 5 | Analysis of perianth organ senescence and abscission in wild-type plants and transgenic plants that overexpress AGL15 during senescence. Analysis of perianth organ senescence and abscission in wild-type plants and transgenic plants that overexpress AGL15 during senescence. A, Gel-blot analysis of total RNA samples (5 mug lane-1) isolated from leaves of wild-type or transgenic plants and hybridized with probes specific for AGL15 or 18S rRNA sequences. AGL15 transcripts accumulate in a senescence-dependent manner in transgenic plants carrying SAG12:gAGL15 constructs. AGL15 transcripts can be detected in non-senescing, green leaves (GL) of plants that constitutively overexpress AGL15 (35S:AGL15), but cannot be detected in green leaves of either wild-type plants (wt) or transgenic plants that carry SAG12:AGL15 constructs (SAG12:AGL15). AGL15 transcripts accumulate at higher levels in the yellow, senescing leaves (YL) of plants carrying SAG12:gAGL15 constructs. Hybridization with 18S rRNA probes indicates equal loading of samples in each lane. B and C. The inflorescences of wild-type plants (B) and plants that carry SAG12:gAGL15 transgenes (C) are indistinguishable in terms of floral organ senescence and abscission. Plants that overexpress AGL15 in senescing tissues resembled wild-type plants in most regards. Leaf senescence was not affected and floral organ senescence and abscission were not visibly altered (Fig. , B and C). Fruit maturation appeared to be slightly delayed, however. The siliques turned yellow 2 to 4 d later (relative to the time of pollination) in the transgenic plants than in wild-type plants (not shown). Based on these results, we conclude that, although it is possible to see some effects in fruits, senescence-induced overexpression of AGL15 is not equivalent to constitutive overexpression in terms of enhancing organ longevity. The Effect of Elevation of AGL15 at Different Stages | We used a glucocorticoid-inducible expression system to more clearly define when overexpression of AGL15 exerts its effects. A line carrying a transcriptional regulator activated by the application of the glucocorticoid dexamethasone (35S:GVG), obtained from Dr. Nam-Hai Chua (Rockefeller University, New York), was crossed to a line carrying the regulatory "target" (UAS:AGL15). Plants carrying both constructs were selected in the F3 generation. Overexpression of AGL15 was induced at various times by applying dexamethasone, either by bottom watering (for systemic induction) or by spraying (for local induction). The effectiveness of the induction system was established through gel-blot analysis (Fig. A) of total RNA samples isolated from induced and uninduced plants. AGL15 transcripts accumulated at higher levels in plants carrying the induction system than in wild-type plants, even in the absence of dexamethasone induction. However, the perianth organs of these plants senesced and abscised normally (not shown), indicating that the constitutive expression due to the leakiness of the system is below the threshold needed for overexpression effects. In plants treated with dexamethasone, AGL15 transcripts accumulate at higher levels than in plants carrying 35S:AGL15 constructs (Fig. A). Figure 6 | Effects of glucocorticoid-induced overexpression of AGL15. Effects of glucocorticoid-induced overexpression of AGL15. A, Gel-blot analysis of total RNA samples (5 mug lane-1) isolated from plants in which AGL15 overexpression can be induced by treatment with dexamethasone (+DEX) or controls, hybridized with probes specific for AGL15 or 18S rRNA sequences. Plants carrying the induction system were either not treated (AGL15 [-DEX]) or were treated after bolting (AGL15 [+DEX] ab) or continuously (AGL15 [+DEX] c) with dexamethasone. AGL15 mRNAs were particularly abundant in lines that showed strong phenotypic changes after treatment ([+DEX] st). AGL15 transcripts were less abundant in plants that constitutively overexpress AGL15 (35S:AGL15) and were undetectable in wild-type plants (wt). Hybridization with 18S rRNA probes indicates equal loading of samples in each lane. B and C, Inflorescence (B) and individual flower (C) after induction of AGL15 overexpression by spraying with dexamethasone. D, Inflorescence of a similar plant sprayed with control solution lacking dexamethasone. To determine when AGL15 overexpression is most critical for extended organ longevity, AGL15 was induced systemically by introducing dexamethasone into the solution used for watering at various times after the start of germination. We found that systemic induction during the reproductive phase was sufficient to produce the overexpression effects in floral organs. The plants did not need to be exposed to dexamethasone during the vegetative phase or during the transition to flowering. Spraying the inflorescence with dexamethasone was just as effective as systemic induction. All of the unopened floral buds sprayed with dexamethasone gave rise to flowers showing delayed abscission and senescence of perianth organs (Fig. , B and C). Buds sprayed with solutions lacking dexamethasone showed no effects on either senescence or abscission (Fig. D). Open flowers sprayed with dexamethasone also showed no effects, even though AGL15 transcripts accumulated in these flowers (not shown). We conclude that stage 13 (flower opening) marks a period of transition with regard to the sensitivity of floral tissues and senescence programs to increased AGL15 levels. Induction of AGL15 overexpression can occur as late as stage 11 or 12 but must occur before stage 13 (bud anthesis) to increase floral organ longevity. DISCUSSION : Overexpression of AGL15 Affects Multiple Developmental Processes | Overexpression of AGL15 results in an overall delay in abscission and senescence in reproductive tissues. In addition to the previously reported effects on perianth organ longevity and abscission , we have shown that processes such as fruit maturation, silique dehiscence, and seed desiccation are also affected. Each of these processes involves tissues that undergo senescence or maturation (perianth organs, fruit walls, and seed coats) and cell layers that express programs leading to cell separation (floral organ abscission zones, silique dehiscence zones, and funiculi). In this work, we have used plants overexpressing AGL15 to explore issues related to the regulation of tissue longevity, interactions between organs and separation zones, and the relationship between programmed senescence and abscission. Overexpression of AGL15 leads to both down-regulation and delays in the molecular programs associated with senescence in floral organs and maturing siliques. In wild-type plants, the senescence reporter SAG12:GUS is activated as early as 3 DAP in the floral organs and during the final stages of chlorophyll loss in silique tissues. When AGL15 is overexpressed, SAG12 promoter activation is delayed by up to 3 d in floral organs. SAG12 promoter activation and chlorophyll loss are both delayed in the fruits of plants overexpressing AGL15, and GUS activity is significantly lower at the final stages of fruit maturation than in wild-type plants. We conclude that AGL15 serves as a repressor of senescence when it is constitutively overexpressed, and is likely to act at a point upstream of both the final stages of chlorophyll loss and SAG12 promoter activation. Overexpression of AGL15 results in an extended desiccation phase during seed development. During the maturation phase, seed fresh and dry weight increase rapidly due to synthesis and deposition of storage reserves. During the desiccation phase, both the fresh weight and water content drop as the seeds lose water. The desiccation phase is longer in the seeds produced on plants overexpressing AGL15 largely because of changes in the profile of water loss. Although seed water content usually declines at a steady rate throughout the desiccation phase in wild type, seed water content is maintained around the 30% to 40% level for several days (from 17 --23 DAP) in plants overexpressing AGL15. Although the decline in water content later resumes, seed maturation is delayed. The seeds remain green several days longer and will not germinate readily if they are excised during this period (not shown), indicating that they are not yet fully mature. Although an extension of the maturation period might be expected to enhance reserve accumulation, this does not appear to be the case here. The levels of storage proteins in mature seeds obtained from plants that overexpress AGL15 are not elevated relative to the levels in mature wild-type seeds (K.W. Nichols and D.E. Fernandez, unpublished data). Effects on Fruit Maturation and Seed Desiccation Can Be Attributed to AGL15 Overexpression in Maternal Tissues | Can any part of the effects on fruit maturation and seed desiccation be attributed to signals coming from embryos that carry the AGL15 transgene? We found that introducing the AGL15 transgene into embryos is not sufficient to produce the overexpression effects in wild-type females. In addition, when crosses are performed with transgenic females such that the dosage and/or the frequency of embryos carrying the AGL15 transgene is reduced, no alteration is seen in the pattern of water content decline or rate of fruit senescence relative to self-pollinated transgenic plants. Hence, we conclude that transgene expression in the maternal tissues is necessary, and may be sufficient, to direct the changes seen with AGL15 overexpression. Overexpression of AGL15 affects several features that may contribute to changes in water balance and the delay in seed desiccation. First, delays in fruit wall senescence may result in the maintenance of higher water potentials inside the silique. Seed desiccation is associated with senescence of the pod or capsule in legumes , and it has been suggested that water loss in seeds is primarily due to transpiration of water from the surfaces of surrounding tissues . Both chlorophyll loss and the decline in seed water content transiently slow for a period of several days in plants overexpressing AGL15. When chlorophyll loss resumes, the decline in seed water content does as well. This correlation suggests the two parameters are linked and would be consistent with a cause-and-effect relationship. Alternatively, fruit wall senescence and seed desiccation may be controlled by regulatory elements that are similarly impacted by AGL15 overexpression. Another possible contributing factor in this developmental context would be an effect on seed detachment. The final stages of seed desiccation occur after the vascular supply to the seed is severed through funiculus "abscission" . In plants overexpressing AGL15, the seeds remain green and are connected to the fruit tissues for a longer period. This suggests that the onset of processes that lead to cell separation are delayed in the funiculi, although we have not determined, by direct measurements, when flow through the vascular system ceases. Just as in floral organ abscission zones, AGL15 protein can be immunolocalized at early stages of development in the nuclei of the small, densely cytoplasmic cells that form the detachment zone at the distal end of the funiculus (S. Perry and D. Fernandez, unpublished data). Whether AGL15 plays any direct role in maintenance or specification of these cells has not yet been determined. The Effects of Abscission- and Senescence-Induced Expression of AGL15 | Abscission is frequently associated with senescence of the organ distal to the point of cell separation, and both processes can be initiated in response to the same developmental and environmental factors . In Arabidopsis, the floral organs of Arabidopsis are shed in a turgid state ; however, based on our senescence marker analysis, the senescence process has already initiated in these organs when abscission occurs. Overexpression of AGL15 affects both abscission and senescence. Because extensive cross talk within and between cells is likely in this situation, secondary effects are difficult to distinguish from primary effects. By supplying "extra" AGL15 in particular cells or at specific times, we sought to determine the relationship between effects on abscission, effects on senescence, and other possible primary effects. If AGL15's primary effect is on genes controlling abscission, we might expect that overexpression of AGL15 in abscission zone cells would be sufficient to achieve the floral overexpression phenotype. We found, however, that when AGL15 is supplied ectopically only in the abscission zone, the perianth organs on such plants both senesce and abscise as in wild type. We cannot rule out the possibility that abscission zone-specific expression at an earlier stage, before the chitinase promoter is active, would have the desired effect; however, our results suggest that the effect on abscission is a secondary effect. If AGL15 regulates senescence genes directly, we might expect to see aspects of the overexpression phenotype if AGL15 is supplied at elevated levels during the course of senescence. However, plants that express AGL15 in a senescence-activated manner show no change in the timing of perianth organ senescence and abscission relative to wild type. In fact, in subsequent experiments with glucocorticoid-inducible expression, we found that AGL15 overexpression that occurs anytime after the flowers open is ineffective. Fruit senescence, on the other hand, is somewhat delayed relative to wild type when AGL15 is expressed in a senescence-activated manner, although not to the extent it is in plants that constitutively overexpress AGL15. Thus, although AGL15 may affect some aspects of late senescence programs, it cannot act to halt or reverse the senescence process in either flowers or fruits if it is expressed after senescence is initiated. Overexpression of AGL15 before the Onset of Senescence and Abscission Produces Pronounced Effects | To investigate the contribution the timing of AGL15 overexpression makes to the total picture of overexpression effects, we set up a system that allowed us to inducibly express AGL15 at high levels. Although the system showed some "leakiness" (i.e. AGL15 is expressed at low levels even in the absence of the glucocorticoid inducer), phenotypic changes were associated exclusively with induction and high-level expression. Our studies showed that induction before the transition to flowering or in very young floral buds is not necessary. Induction immediately (1 d) before the flower opens is sufficient to achieve the full overexpression effect on perianth longevity and retention. This corresponds to the period when AGL15 promoter activity normally ceases in wild-type flowers . We suggest that the effects may be due to ectopic expression that adds "extra" AGL15 to a shrinking pool of functional molecules in normal target cells. Induction after the flowers open is ineffective, even though AGL15 transcripts still accumulate in these older flowers after induction (not shown). We conclude that the direct targets of AGL15 action are likely to be in tissues that have largely completed the process of differentiation but are in a presenescent state. Further molecular studies to determine these targets and the primary effects of AGL15 overexpression should focus on this stage and tissues. In summary, overexpression of AGL15 affects the progression of developmental programs, particularly those that involve both senescence and abscission/cell separation, in flowers, fruits, and seeds. Transgenic plants can often provide new insights into important physiological relationships. In this case, we see a particularly striking demonstration of the strong connection between senescence/maturation of the maternal tissues and the process of seed desiccation. Based on our analysis of the effects of regulated AGL15 overexpression, we conclude that programs operating in cells in the presenescent state or just initiating senescence are most sensitive to changes in AGL15 levels. The phenotypic changes likely reflect changes in gene regulation that occur at a stage before obvious signs of abscission or senescence appear. MATERIALS AND METHODS : Plant Material | Arabidopsis ecotype Wassilewskija (Ws) was used in all experiments unless otherwise specified. Plants were grown to maturity in a growth chamber (Enconair Ecological Chamber, Winnipeg, Canada) with a 16-h-light (approximately 125 muEm-2 s-1) and 8-h-dark regime, 22C, and 70% to 80% relative humidity, as described previously . Chlorophyll Content Measurement | Newly opened flowers were hand pollinated and siliques were collected at various DAP. Two siliques from each of two individual plants were collected, weighed, quickly frozen with liquid nitrogen, and stored at -80C until analysis. Each sample was homogenized in 1.2 mL of 80% (v/v) acetone using a pestle in an Eppendorf tube (Eppendorf Scientific, Westbury, NY). After removing cellular debris by centrifugation, the supernatant containing chlorophyll was removed and diluted 1:1 [v/v] with 100% (v/v) acetone. The absorption at 663 and 646 nm was measured for each sample using a spectrophotometer. The chlorophyll content was calculated according to , as follows: Senescence Reporter Gene Analysis | Seeds of a line (Landsberg erecta ecotype) carrying a single insertion of the SAG12:GUS reporter gene construct , which consists of a fusion between the SAG12 promoter and the coding sequence of GUS, were obtained from Susheng Gan (University of Kentucky, Lexington). Plants carrying SAG12:GUS in the homozygous condition were crossed with plants carrying a single insertion of 35S:gAGL15 (CaMV 35S promoter fused to the coding sequence of AGL15 and the first three introns; ) in the hemizygous condition. GUS activities were assayed in the progeny. All of the plants carried the SAG12:GUS construct and one-half of the population showed the phenotypic changes associated with AGL15 overexpression . GUS Activity Assays | For histochemical assays, tissues were incubated with 0.5 mg mL-1 5-bromo-4-chloro-3-indolyl glucuronide as previously described . The tissues were then incubated in 70% (v/v) ethanol until the chlorophyll was completely extracted. For fluorometric assays, siliques at different developmental stages were harvested, weighed, frozen in liquid nitrogen, and stored at -80C. For each data point, two to three siliques were homogenized with 200 muL of extraction buffer (50 mm NaH2PO4, [pH 7.0], 10 mm EDTA, 0.1% [v/v] Triton X-100, 0.1% [w/v] sodium lauryl sarcosine, and 10 mm beta-mercaptoethanol) in an Eppendorf tube using a pestle. Cellular debris was removed by centrifugation. Five microliters of plant extract was added to 1 mL of extraction buffer containing 1 mm 4-methyl umbelliferyl glucuronide, and the reaction was incubated at 37C. At 0, 5, 15, 30, and 60 min, 200 muL was removed, and the fluorometric reaction was carried out as previously described . The emission at 455 nm was measured after excitation at 365 nm in a fluorometer (Hoefer TKO100, San Francisco). Water Content Analysis | For each data point, a total of eight to 12 staged siliques were collected from eight to 12 individual plants. To determine fresh weight, the seeds were removed from the siliques using forceps, pooled, and weighed. To determine dry weight, each seed sample was then baked at 80C for 5 h and weighed again. The water content was calculated as follows: Generation of Transgenic Arabidopsis | To express AGL15 in the floral organ abscission zones, AGL15 was placed under the control of regulatory sequences from a bean (Phaseolus vulgaris) abscission chitinase gene (Chi:gAGL15). A 1.4-kb fragment, representing the chitinase upstream regulatory region plus 24 bp of the coding sequence, was excised from pBD3226 using HindIII and BamHI restriction enzymes and cloned into a pPZP221 transformation vector , which includes the nopaline synthase terminator (NOS). The AGL15 DNA sequence was amplified from 35S:gAGL15 using PCR and was inserted downstream of the chitinase sequence. This created an in-frame fusion that added 14 amino acids to the N terminus of the AGL15 gene product. Plants that constitutively express this version of AGL15 show the AGL15 overexpression phenotype (S.-C. Fang, unpublished data), indicating that the extra amino acids do not interfere with AGL15 function. The AGL15 coding region was also sequenced to confirm that no mutations had been introduced during PCR amplification. The Chi:gAGL15 construct was introduced into Arabidopsis (Ws ecotype) by vacuum infiltration with Agrobacterium tumefaciens strain GV3101, as described by . T1 plants were selected on germination medium (GM) medium supplemented with 100 mug L-1 gentamycin. Plants carrying Chi:gAGL15 insertions in the homozygous condition were identified in the T2 generation and used for immunohistochemistry studies. To express AGL15 during senescence, AGL15 was placed under the control of regulatory sequences from the SAG12 gene . A 2.2-kb fragment of the SAG12 upstream regulatory sequence was excised from pSG506 using EcoRV and NcoI restriction enzymes. After treatment with S1 nuclease, the SAG12 sequence was inserted immediately upstream of the AGL15 coding sequence in a PZP212 transformation vector to create SAG12:gAGL15, which includes a NOS terminator. The SAG12:gAGL15 construct was introduced into Arabidopsis (Ws ecotype) using A. tumefaciens strain GV3101 and the floral dip method, as described by . T1 plants were selected on GM medium supplemented with 50 mug L-1 kanamycin. Immunohistochemistry | Arabidopsis tissues were fixed with 4% (w/v) freshly prepared paraformaldehyde and 0.02% (v/v) Triton X-100 in 50 mm potassium phosphate buffer, pH 7.0, at 4C overnight. Fixed tissues were embedded in paraffin medium (Paraplast Plus, Sigma, St. Louis) and sectioned (7 mum thick) with a steel knife. Immunolocalization was performed using AGL15-specific antiserum, as described previously . Analysis of Petal Break Strength | Petal break strength was measured in flowers of the primary inflorescence at the stage when at least 10 flowers had opened but before global proliferative arrest of the inflorescence meristems. Petal break strength was quantified using a stress transducer as described previously . System for Glucocorticoid-Inducible Expression of AGL15 | The AGL15 coding sequence amplified from 35S:gAGL15 using PCR was cloned into a pPZP212 transformation vector , which includes the NOS terminator. A DNA fragment, UAS-35S, containing six copies of the GAL4 upstream activation site (UAS) and -46 to +9 region of the CaMV 35S promoter, was excised from pTA7001 using HindIII and XhoI restriction enzymes and inserted upstream of the AGL15 sequence. The resulting construct was designated UAS-35S:gAGL15. The UAS-35S:gAGL15 construct was introduced into Arabidopsis (Ws ecotype) by vacuum infiltration with A. tumefaciens strain GV3101, as described by . T1 plants were selected on GM medium supplemented with 50 mug L-1 kanamycin (Sigma). To introduce the glucocorticoid-regulated transcriptional activator GVG, transgenic plants carrying the UAS-35S:gAGL15 construct were crossed to transgenic plants carrying the TA7002:LUC construct (seed obtained from Dr. Nam-Hai Chua, Rockefeller University). TA7002:LUC contains the complete two-component glucocorticoid-inducible system and, therefore, provides the GVG transcriptional factor for induction of AGL15. F1 plants containing both UAS-35S:gAGL15 and TA7002:LUC were selected on GM medium supplemented with 50 mug L-1 kanamycin (Sigma) and 50 mug L-1 hygromycin (Sigma). The induction of AGL15 mRNA in F1 seedlings by glucocorticoid was confirmed by RNA gel-blot analysis (not shown). Plants in the F3 generation were used for further analyses. Glucocorticoid Treatments | Dexamethasone (Sigma), an analog of glucocorticoid, was dissolved in 100% (v/v) ethanol to make a 30 mm stock solution and stored at -20C. For systemic induction, the plants were bottom watered every other day with 10% (v/v) Hoagland solution containing 30 mum dexamethasone and 0.01% (v/v) Silwet L-77 (OSI Specialties, Inc., Sistersville, WV). To treat inflorescence tissues, the pots were covered by plastic wrap and the inflorescences were allowed to grow through holes in the plastic. The exposed inflorescences were sprayed with a solution containing 30 mum dexamethasone and 0.01% (v/v) Silwet L-77 every other day. Control plants were sprayed with a solution containing 0.1% (v/v) ethanol and 0.01% (v/v) Silwet L-77. RNA Gel-Blot Analysis | Total RNA was isolated according to the method described by . Five micrograms of total RNA was denatured and sized-fractionated in a 1% (v/v) formaldehyde-agarose gel following the protocol described in . RNA was then transferred to a nylon membrane (Nytran Supercharge, Schleicher & Schuell, Keene, NH). The DNA template used to make AGL15 probe was generated by PCR amplification from an Arabidopsis AGL15 cDNA clone using primers 5'-CTTGAAGAATCACGCCTC-3' and 5'-TTAGCGGCCGCAGAGAACCTTTGTCTTTTGGCTTC-3'. The amplified DNA product contained 24 bp of the K box region and the full C-terminal region of AGL15. The DNA template used to make the 18S rRNA probe was excised from pRE12 using the EcoRI restriction enzyme. The template DNAs were gel purified and labeled with 32P-labeled dCTP by random priming as described by . Prehybridization, hybridization, and wash conditions were as described previously . The membrane was exposed to a phosphoscreen and the hybridization pattern was visualized by using a Cyclone Storage Phosphor Imaging System (Hewlett-Packard, Palo Alto, CA). The membrane was stripped before the second hybridization by washing twice with boiling 0.1% (w/v) SDS and 0.1x SSC. Distribution of Materials | Upon request, all novel materials described in this publication will be made available in a timely manner for noncommercial research purposes. Backmatter: PMID- 12226489 TI - Molecular Identification of Cytosolic, Patatin-Related Phospholipases A from Arabidopsis with Potential Functions in Plant Signal Transduction AB - Rapid activation of phospholipase A (PLA) by auxin or plant-pathogen interaction suggests a function in signal transduction for this enzyme, but the molecular identification of a cytosolic PLA carrying out this function remains open. We isolated four cDNA sequences from Arabidopsis (ecotype Columbia), AtPLA I, AtPLA IIA, AtPLA IVA, and AtPLA IVC, which are members of the patatin-related PLA gene family in plants and which are homologous to the animal Ca2+-independent PLA2 gene family. Expression was measured by reverse transcriptase-polymerase chain reaction, and AtPLA I transcripts were found preferentially in shoots, AtPLA IIA and AtPLA IVA in roots, and AtPLA IVC in flowers. Transient expression of the four PLA-green fluorescent protein fusion proteins in tobacco (Nicotiana tabacum) leaves showed they were located in the cytosol and not in the vacuoles. Surprisingly, AtPLA::green fluorescent protein was also localized to chloroplasts. The enzymatic activity of the purified recombinant AtPLA IVA toward phosphatidylcholine was dependent on Ca2+, saturated at 0.5 mm, and had a pH optimum of about 7.0. It had both PLA1 and PLA2 specificity. The enzyme showed in vitro highest sensitivity toward the PLA2 inhibitors palmitoyltrifluoromethyl ketone (PACOCF3, Ki approximately 30 nm), arachidonyltrifluoromethyl ketone (AACOCF3, Ki approximately 25 mum), and tetrahydro-3-(1-naphtalenyl)-2H-pyran-2-one (Ki approximately 200 nm) and was also sensitive to other previously used inhibitors 5,8,11,14-eicosatetraynoic acid (Ki approximately 3 mum) and nordihydroguajaretic acid (Ki approximately 15 mum). The influence of these PLA2 inhibitors on elongation in etiolated Arabidopsis seedlings was tested, and tetrahydro-3-(1-naphtalenyl)-2H-pyran-2-one and 5,8,11,14-eicosatetraynoic acid inhibited hypocotyl elongation maximally at concentrations close to their Ki in vitro. Keywords: Introduction : Eucaryotic phospholipases A2 (PLA2s) comprise four major genetically unrelated gene families, which are further subdivided into 11 groups according to their amino acid-similarities and biochemical properties . The first group identified comprised the secreted enzymes (sPLA2) of around 14 kD, which have a function in digestion or as toxins. Plant genes related to sPLA2s have been described . Two types of cytosolic PLA2s were recently identified in animals. The Ca2+-dependent PLAs (cPLA2s), respond to various hormonal stimuli and are activated by phosphorylation by MAP kinase and protein kinase C , and the Ca2+-independent (iPLA2s) have a function in the remodeling of fatty acid composition of phospholipids and can also be activated by stimuli . A third type of cytosolic PLA2, the platelet activating factor acetohydrolases, seems to be a specialized type of enzyme related to the low density lipoprotein-PLA2, which preferentially cleave oxidized fatty acids . Plant enzymes related in sequence to platelet activating-factor acetohydrolases apparently differ in function . The sequences of secreted plant sPLA2, the vacuolar patatins, and the patatin-related PLA are similar to the respective animal PLA2s, but the molecular identification of a cytosolic plant PLA suitable for a function in signal transduction is less clear. The first indication for a function of plant PLA in signal transduction was the rapid activation of PLA activity by auxin . Later, the activation of PLA by pathogens and elicitors was shown . Several attempts to purify soluble plant PLAs have been made, and besides the 14-kD sPLA2 enzymes, other PLA enzymes ranging from 40 to 70 kD were purified from potato (Solanum tuberosum; ), rubber tree (Hevea brasiliensis; ), broad bean (Vicia faba; ), and tobacco (Nicotiana tabacum; ). These were similar by sequence to patatins, but only the rubber tree enzyme was shown to be cytosolic when expressed in the yeast Pichia spp. . When we set out to isolate genes encoding plant PLAs with a potential function in plant signal transduction, our initial attempts failed to isolate enzymes similar to the major PLA2 enzyme involved in animal signal transduction, the cPLA2 . Our focus turned to isolation of plant PLA cDNAs from Arabidopsis, which are related to the animal iPLAs, because the auxin- and elicitor-induced activation of PLA in parsley (Petroselinum crispum) cell cultures is strongly inhibited by the drug HELSS , which is described as a diagnostic inhibitor for iPLAs in contrast to cPLA2s . Here, we report on the isolation of several patatin-related PLA cDNAs, their tissue-specific expression, and the subcellular localization of the respective green fluorescent protein (GFP) hybrid proteins. Data on the inhibitor characteristics of a purified his-tagged PLA enzyme and the effects of these inhibitors on hypocotyl elongation suggest that members of this gene family may have a function in plant signal transduction. RESULTS : Isolation of cDNAs and Tissue-Specific Transcription | By sequence comparisons we could not identify any cPLA2 sequence in the Arabidopsis genome (data not shown), but we identified 10 genes for patatin-like PLA protein sequences, all having the conserved iPLA2-specific sequence motifs. Both the genomic structures and the amino acid comparisons of these sequences suggested that they fall into three groups: group 1 comprising only AtPLA I; group 2 comprising AtPLA IIA, AtPLA IVA, AtPLA IVB, AtPLAVIC, and AtPLA V; and group 3 comprising AtPLA IIB, AtPLA IIIA, AtPLA IIIB, and AtPLA IVD (Fig. , A and B). Genes or putative genes are named with Roman numerals according to the respective chromosomes, and letters, if multiple genes were located on a single chromosome. When AtPLA I was compared with the animal proteins, scores of 24% to 32% identity were obtained , which is reflected in the unrooted phylogenetic tree (Fig. B), so that AtPLA I and group 1 is considered to be evolutionary older than the other two plant gene groups. We assume that group 3 is evolutionary younger than group 2, because the catalytic domains of group 3 is different from those of all other known iPLAs, and because group 2 is closest to the presumably old prokaryotic Anabaena spp. sequence. The fact that group 3 sequences have only one intron may be explained by intron loss, as has been observed in other Arabidopsis gene families , and may be associated with the evolution of small genomes. Figure 1 | Gene structure of the patatin-related AtPLA-family in Arabidopsis and a phylogenetic tree of several patatin-related plant PLA and animal iPLA2 sequences. Gene structure of the patatin-related AtPLA-family in Arabidopsis and a phylogenetic tree of several patatin-related plant PLA and animal iPLA2 sequences. A, Gene structure of the AtPLA family in Arabidopsis. Exons are symbolized by gray boxes, those containing the LRR by black boxes, the catalytic center by white boxes, and introns by black lines. cDNA sequences described in this work are marked by an asterisk. The sequences are deposited in database as follows: AtPLA I (accession no. ), AtPLA IIA (accession no. ), AtPLA IIB (accession no. ), AtPLA IIIA (accession no. .1), AtPLA IIIB (accession no. ), AtPLA IVA (db_xref GI:4006869), AtPLA IVB (db_xref GI:4006870), AtPLA IVC (db_xref GI:4006871), AtPLA IVD (accession no. .1), and AtPLA V (accession no. .1). A 500-bp size standard is indicated. B, Phylogenetic tree produced by the program ClustalW on the amino acid structures of the AtPLA gene-family, a putative bacterial protein from Anabeana sp. (accession no. .1; ), a putative protein from the nematode Caenorhabditis elegans (accession no. .1), a putative protein from the fruitfly (Drosophila melanogaster; accession no. .2), an iPLA2 from human (accession no. JC7284; ), and a patatin class I precursor from potato (accession no. ; ) displayed by the program TREEVIEW . Figure 2 | Alignment of predicted amino acid sequences of isolated PLA cDNAs from Arabidopsis and of selected domains of patatin-related PLAs from other organisms. Alignment of predicted amino acid sequences of isolated PLA cDNAs from Arabidopsis and of selected domains of patatin-related PLAs from other organisms. The conserved residues of LRRs of the consensus sequence LXXLXXLXLXXN/CXXL/IP/RXLXXLXX are highlighted by printing the conserved residues below the AtPLA I sequence. The additional exon found by us and derived from the isolated cDNA of PLA I is boxed. The most highly conserved motifs in the consensus sequence DGGGXRG of the catalytic center and the lipase motif GTSTG are underlined. The PLA amino acid sequences from Arabidopsis AtPLA I, AtPLA IIA, AtPLA IVA, and AtPLA IVC are compared with a putative bacterial protein from Anabeana sp. (accession no. .1), a putative protein from the nematode C. elegans (accession no. .1), a putative protein from the fruitfly (accession no. .2), an iPLA2 from human (accession no. JC7284), and a patatin class I precursor from potato (accession no. ) by using the program vector NTI from InforMax. A consensus sequence is indicated below the sequences. The amino acid sequences derived from cDNAs of AtPLA I, AtPLA IIA, AtPLA IVA, and AtPLA IVC are shown in Figure . We also isolated cDNAs for AtPLA IIB, AtPLA IIIA, AtPLA IIIB, AtPLA IVB, and AtPLA IVD (E. Oppermann, A. Holk, and G.F.E. Scherer, unpublished data) but failed to isolate AtPLA V. The full-length cDNA clone of AtPLA I could only be obtained by reverse transcriptase (RT)-PCR. AtPLA I is not only unique by possessing an iPLA2 domain with the highest similarity to animal iPLA2 sequences (Figs. B and ), but also by having two additional domains, a Leu-rich repeat (LRR) at the N terminus and a C-terminal domain that is not similar to other known sequences. For AtPLA I, we isolated two splice variants. The shorter one did not contain the fifth exon and contained a stop codon before the PLA domain (data not shown), whereas the second, longer cDNA included an uninterrupted reading frame comprising all three domains and an additional exon (in comparison with the annotated sequence in accession no. ), that extended the LRR domain. The other, shorter cDNAs were isolated by screening several libraries (AtPLAIIA, IVA, and IVC) or RT-PCR (AtPLA IIB, IIIA, IIIB, IVB, and IVD). We could find only a few ESTs for the Arabidopsis PLA sequences, so we expected low expression and chose competitive RT-PCR to study their expression in different tissues . The AtPLA I gene was expressed preferentially in shoots and also in flowers and roots, but very little in leaves. Genes AtPLA IIA and AtPLA IVA were expressed in the roots and were much weaker in the flowers, shoots, and leaves. The expression of AtPLA IVC was preferentially observed in flowers but also in roots and was very low in leaves and shoots. Everywhere but in leaves, a somewhat longer splice variant was detected for the AtPLA IVC, but this mRNA was not further investigated. The quite different tissue-specific expression of the isoforms investigated here was reflected in their equally different promoter sequences and was verified by initial experiments with promoter-GUS transformants (not shown). Because the other members of the gene family except AtPLAV could be isolated by RT-PCR (data not shown) this proves that these additional genes AtPLA IIB, AtPLA IIIA, AtPLA IIIB, and AtPLA IVD were also expressed and probably function in Arabidopsis. Sequence AtPLA IIIB was isolated by , and it is expressed most highly in roots, less in flowers and stems, and least in leaves. A mutant resulting from activation tagging of gene AtPLA IIIB exhibited the "sturdy" phenotype . Taken together, these data suggest both tissue preference and functional redundancy for the members of this gene family as has been found for other plant gene families. Figure 3 | Expression of mRNA of the AtPLA I, AtPLA IIA, AtPLA IVA, and AtPLA IVC gene in organs of Arabidopsis. Expression of mRNA of the AtPLA I, AtPLA IIA, AtPLA IVA, and AtPLA IVC gene in organs of Arabidopsis. The total RNA from roots, shoots, leaves, and flowers was analyzed by competitive RT-PCR using gene-specific internal standards for the genes I, IIA, and IVA or by RT-PCR using actin as external standard for the gene IVC; separated on ethidium-bromide gel; and inverted into gray scale for digitization and quantification. Relative amounts of cDNAs (highest value is set to 100%) are shown in A for I, in B for IIA, in C for IVA, and in D for IVC. Properties of Ectopically Expressed Arabidopsis PLA Proteins and Their Cytosolic Localization | Hydropathy plots for various plant patatin-related sequences and the AtPLAs showed that a signal peptide for secretion was present in all the patatins from potato, several from tobacco, and one protein sequence from Cucumis, but not in any of the Arabidopsis sequences (not shown). To demonstrate the suspected cytosolic compartmentation, hybrid PLA-GFP from Arabidopsis proteins were expressed transiently in tobacco leaves . All four expressed sequences, AtPLA I, AtPLA IIA, AtPLA IVA, and AtPLA IVC, clearly were not located to the vacuoles. For AtPLA IIA, AtPLA IVA, and AtPLA IVC, a cytoplasmic localization is suggested, but association with membranes (e.g. plasma membrane or ER) cannot be excluded. As demonstrated by a series of optical sections from palisade parenchyma cells and additional data (not shown), the hybrid protein AtPLA I-GFP colocalized with chloroplasts, although not with all of them in a given cell (Fig. , A --C). Figure 4 | Transient expression of PLA-GFP fusion proteins in tobacco leaf cells as visualized by confocal laser scanning microscopy. Transient expression of PLA-GFP fusion proteins in tobacco leaf cells as visualized by confocal laser scanning microscopy. A through C, Three consecutive optical sections, moving from the epidermal top side of palisade cells toward the spongy parenchyma side containing, expressed PLA-I-GFP fusion protein. Arrows, Localization of PLA-I-GFP fusion protein in an individual chloroplast in consecutive frames. Arrowheads, Localization of PLA-I-GFP fusion protein in the cytosol. D, Single optical section of palisade parenchyma cells expressing the PLA-IIA-GFP fusion protein. E, Summarized image showing the expressed PLA-IVA-GFP-fusion protein calculated from 16 consecutively taken single optical sections (in z-direction) through a large cell. F, Single optical section of an epidermal cell expressing the PLA-IVC-GFP fusion protein. Bar = 10 mum. In previous studies, known PLA2 inhibitors were used to inhibit auxin-dependent growth or auxin stimulation of a PLA2 activity . To be able to compare the effect of inhibitors on an isolated enzyme and on biological responses, it would be desirable to know the enzymological properties of the PLAs and, especially, the reaction of isolated PLA to known PLA2 inhibitors. As a first step toward this goal, the gene AtPLA IVA was expressed as a N-terminally his-tagged protein in Escherichia coli and purified (Fig. A). Enzymatic activity was measured using a fluorescent phosphatidylcholine [1,2-bis-(4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-undecanoyl) - sn-glycero-3-phosphocholine (bis-BODIPY-PC)] labeled at both fatty acids by a fluorochrome that was used in previous experiments in vivo . Hydrolysis by the purified recombinant enzyme yielded fluorescent free fatty acid and fluorescent lysophosphatidylcholine as products (Fig. B). When a different phosphatidylcholine was used, carrying a fluorescent fatty acid only at the C2 atom [2-(4,4-difluoro-5-methyl-4-bora-3a,4a-diaza-s- indacene-3-dodecanoyl)-1-hexadecanoyl-sn-glycero-3- phosphocholine], fluorescent lysophosphatidylcholine and fluorescent free fatty acid originated, indicating that the enzyme had both PLA1 and PLA2 activity. Hydrolysis was linear for 45 min (Fig. C). Although the sequence of AtPLA IVA to the iPLA2 class suggests that its activity would be independent of Ca2+, activity assays showed that it was dependent on Ca2+, being saturated at 0.5 mm Ca2+. The pH optimum of the enzyme was about 7.0 (Fig. , A and B). The enzyme showed highest sensitivity toward the inhibitors palmitoyltrifluoromethyl ketone (PACOCF3; Ki approximately 30 nm) and tetrahydro-3-(1-naphtalenyl)-2H-pyran-2-one (HELSS; Ki approximately 200 nm), and it was also sensitive to other previously used inhibitors such as arachidonyltrifluoromethylcarbon (AACOCF3; Ki approximately 25 mum), 5,8,11,14-eicosatetraynoic acid (ETYA; Ki approximately 3 mum), and nordihydroguajaretic acid (NDGA; Ki approximately 15 mum; Fig. , C --G). Figure 5 | Purification and PLA activity of the recombinant AtPLA IVA protein. Purification and PLA activity of the recombinant AtPLA IVA protein. A, SDS-PAGE of recombinant AtPLA IVA-purification steps. Lane M, Mr size marker; lane 1, E. coli proteins without induction by isopropyl-beta-d-thiogalactosid (IPTG); lane 2, E. coli proteins after induction by 0.5 mm IPTG; lane 3, 12,000g pellet after lysis; lane 4, supernatant of soluble proteins (12,000g centrifugation); lane 5, nickel-nitrilotriacetic acid agarose resin-purified recombinant AtPLA IVA protein. B, Thin-layer chromatogram and comparison of the enzymatic products of PLA digestion of 2-(4,4-difluoro-5-methyl-4-bora-3a,4a-diaza-s-indacene-3-dodecanoyl)-1-hexadecanoyl-sn-glycero-3-phosphocholine (in lanes 1 --3) and bis-BODIPY-PC (lanes 4 --6) by recombinant protein from gene AtPLA IVA. Lane 1, Monolabeled substrate only; lane 2, assay with monolabeled substrate using equivalent amount to lane 3 of control eluate of an empty-vector purification; lane 3, assay with monolabeled substrate using 0.5 mug of IVA purified protein; lane 4, bis-labeled substrate only; lane 5, bis-labeled substrate using an equivalent amount of control eluate from empty-vector purification; lane 6, bis-labeled substrate using 0.5 mug of purified IVA protein in 6 muL; and lane 7, standards for fatty acid and lysophosphatidylcholine. C, PLA activity test. Time-course experiment with recombinant AtPLA IVA and bis-labeled substrate. The amounts of the remaining substrate and the enzymatic products, fatty acid (FA) and lysophosphatidylcholine (LPC), had different apparent molar efficiencies regarding fluorescence emission because of different spot sizes as captured by video photography so that results are expressed on a relative scale. Figure 6 | Catalytic properties of purified recombinant PLA derived from gene AtPLA IVA. Catalytic properties of purified recombinant PLA derived from gene AtPLA IVA. A, Ca2+ dependence of PLA activity; inset, highlighting of low Ca2+ concentrations. B, pH dependence of activity; C, inhibition by HELSS; D, inhibition by NDGA; E, inhibition by AACOCF3; F, inhibition by ETYA; and G, inhibition by PACOCF3. Fluorescent fatty acid was quantified on a relative scale. It was previously suggested that a PLA is involved in auxin signal transduction because it became rapidly activated by auxin and PLA2 inhibitors specifically prevented auxin action . Therefore, we tested the influence of PLA2 inhibitors on auxin-dependent elongation in etiolated Arabidopsis seedlings . Of all inhibitors tested, HELSS and ETYA strongly inhibited hypocotyl elongation at concentrations close to the Ki in vitro. NDGA was only marginally active, and the other inhibitors were ineffective in vivo. This may have been perhaps because of their metabolism in plants or because of other unknown side effects. Figure 7 | Influence of PLA2-inhibitors on auxin-induced hypocotyl elongation growth of Arabidopsis. Influence of PLA2-inhibitors on auxin-induced hypocotyl elongation growth of Arabidopsis. A, Inhibition by HELSS; B, inhibition by ETYA; C, inhibition by NDGA; D, inhibition by PACOCF3; and E, inhibition by AACOCF3. se is shown (25 --40 individuals per sample). DISCUSSION : The goal of this work was to identify which plant PLA or group of plant PLAs could have a function in plant signal transduction. Comparison of the limited number of sequences presented here and about 60 homologous sequences (not shown) revealed that all PLAs from Arabidopsis isolated by us clearly belong to the patatin group, which is related to the animal iPLA2 type. Two sequence features, the N-terminal iLsiDGGGXRGXX (X = aliphatic residues) element and the sequences flanking the catalytic Ser, IXGTSTGGXLXX, are highly conserved in all sequences found in protists, plants, and animals. The catalytic center of the animal cPLA2 contains two motifs similar to those two mentioned for iPLA2s, SGGGXRAX and GXSGS, where both the Arg in the first element and the central Ser in the second element participate in the catalytic mechanism . This similarity in catalytic sites may explain the partially similar sensitivity to inhibitors for both cPLA2s and iPLA2s . Nevertheless, when we searched by BLAST for the presence of cPLA2 sequences in plants using the catalytic centers of cPLA2s as a template, we did not find any, but we found many distantly related fungal enzymes in the databases. Moreover, the Caenorhabditis spp. genome contains six iPLA2s, but no cPLA2. Although it is difficult to exclude the presence of cPLA2 in plants, even after completion of the Arabidopsis genome, it seems unlikely that plants possess this type of enzyme, which is the dominant type of PLA2 signal transduction in higher animals . On the other hand, several additional patatin- and iPLA2-specific motifs such as SAAPtYF, DGGXXANN, and SLGTG were found to be highly conserved in protist, plant, and animal sequences. They may constitute the patatin-type or iPLA2 "signature," which should be named in honor of the first PLA, potato patatin, shown to contain this sequence . Besides functioning as a vacuolar enzyme, plant PLA has been suggested to have a role in auxin and elicitor signal transduction where it hydrolyzes phosphatidylcholine and phosphatidylethanolamine within 1 to 5 min to generate free fatty acid and lysophospholipids as potential second messengers . In this study, we show that four members of the PLA family have cytosolic rather than vacuolar localizations. A vacuolar localization would clearly exclude these patatin-related PLAs from a function in signal transduction so that, in turn, our data support this supposed function. Another patatin-related PLA from rubber tree was previously shown to be a cytosolic enzyme . A surprise was the possibly dual localization of the AtPLA I-GFP fusion protein in the cytosol and in or on chloroplasts. BLAST comparisons indicate that this protein contains a G-protein-binding motif within a LRR motif, and computer analysis (PSORT, version 6.4) of this sequence indicates a certainty of 0.425 for chloroplast membrane localization. Our data obtained by laser confocal microscopy are supported by this sequence analysis, but they need to be corroborated by independent methods in the future. For other investigated patatin-related PLAs the subcellular localization remained as yet unclear . Other known plant PLA enzymes belong to the sPLA2 group, which are secretory proteins in plants and in animals containing a Ca2+-binding pocket necessary for catalysis. Their activity depends on millimolar Ca2+ concentration and on a nonreducing environment. This seems difficult to reconcile with a role in cytosolic signal transduction, although animal sPLA2 participates in the release of arachidonic acid by binding to the plasma membrane . HELSS is not an inhibitor for sPLA2 because the catalytic mechanism does not use Ser but His in the catalytic center . ETYA was not tested as a PLA2 inhibitor before, to the best of our knowledge. Specific inhibitors can define the role of an enzyme in signal transduction by inhibiting a downstream process. Therefore, a series of inhibitors were tested in vitro on the activity of the first enzyme purified in our laboratory (AtPLA IVA) and in vivo on the elongation of hypocotyls, an auxin-dependent process. HELSS and ETYA were strong inhibitors in both sets of experiments. Although AtPLA IVA was expressed at a low level in the shoot, it should be a good biochemical representative of the patatin-iPLA gene family, because it contains all conserved sequence elements, and it is inhibited by HELSS, an inhibitor used to discriminate iPLA2s from cPLA2s . The previously reported enzymatic properties of iPLA2s and especially the inhibitor studies in this paper support a role in plant signal transduction for plant patatin-related PLAs . In the context of signal transduction, it is interesting to consider the role of Ca2+ and PLA. Despite their genetic relationship to the animal iPLA2s, all reported plant PLA or PLA2 assays, including ours, contained Ca2+ at concentrations far above cytosolic concentrations . There is, however, no obvious feature or domain in plant patatin-like PLAs that could indicate such an important difference to animal enzymes, which are Ca2+ independent . The catalytic mechanism of the cPLA2, similarly to the animal iPLA2s, does not depend directly on Ca2+ (as it does for the sPLA2s), but a Ca2+-binding C2-domain binds cPLA2 to the membrane surface, which thereby greatly increases its activity . Either, plant patatin-related PLAs are activated by Ca2+, or by as-yet-unknown mechanisms in the cytoplasm, or the lipid substrate conformation necessary for catalysis may be Ca2+ dependent. For example, it might be necessary that Ca2+ binds to the phosphate group and induces an altered lipid conformation, which may not become induced by the detergents used in vitro. This might lead to an apparent Ca2+-dependent hydrolysis mechanism of an otherwise Ca2+-independent enzyme. The role of possible interactions of Ca2+ and PLA in plant signal transduction warrants further investigations. The substrates reported for this type of plant PLA include phospholipids and galactolipids , which indicates little discrimination between headgroups. Our data also indicate that patatin-related plant PLA has combined PLA1 and PLA2 activity, similarly, as reported for animal iPLA2 . Therefore, if plant patatin-related PLAs are "sloppy" both in selecting head groups and the position of the fatty acid, these enzymatic properties may not be as relevant for their functions compared with their compartmentation. Especially interesting and puzzling is the hydrolysis of galactolipids by patatin-related PLAs, which are lipids specifically found in chloroplasts, which are not known to participate in signal transduction but are the site of jasmonate biosynthesis . It should be emphasized, however, that the same fluorescent phosphatidylcholines employed here as substrates, were also used to show the rapid activation of PLA2 and not of a PLA1 within 5 min in cultured parsley cells by auxin and elicitor so that one or several members of the patatin-related cytosolic PLAs might be activated by auxin. In summary, our data demonstrate that both an Arabidopsis patatin-related PLA and a physiological response are highly sensitive to inhibition by HELSS and ETYA. These data are the first to suggest that a patatin-related PLA with iPLA2-specific sequence motifs may function in plant signal transduction. The apparent absence of cPLA2 from the Arabidopsis genome provides further support for our hypothesis. Because the patatin-related PLA family in Arabidopsis contains a number of isoforms, further confirmation of their exact functions can only arise from investigations of phenotypes of knockout plants and/or plants transformed by single genes. MATERIALS AND METHODS : cDNA Isolation | The different members of the patatin-related PLAs in Arabidopsis were designated by their chromosome number and, if necessary, by additional letters. For isolation of the full-length cDNAs of the PLA genes IIA, IVA, and IVC, we used the gene-specific primers: AtPLA IIA sense (5'-TGC ATG AGG TGA ACG AAT GTC TCG-3') and antisense (5'-TGG TTG ATT TCC ATC TCT CTG CCG-3'); AtPLA IVA sense (5'-TTA TGG AGA ACA AAT CGC CCT CC-3') and antisense (5'-AAC GGC ACA ATG TCT TTA GCC G-3'); and AtPLA IVC sense (5'-GAT GTG TTG GGT CTT TGA AAG TGG-3') and antisense (5'-TTT ACC GTT GAA CTT GGG TCC-3'). The primers were designed for the 5'- and 3'-untranslated region of the respective genes and were used for 35 cycles PCR (each cycle: 94C for 30s, 50C for 30 s, and 72C for 60 s) to amplify the specific cDNA fragment from an Arabidopsis lambda cDNA library (Uni-ZAP XR Library, Stratagene, La Jolla, CA). By stepwise dilution of positive library fractions as described by , the full-length cDNA phage clones from AtPLA IIA, IVA, and IVC were isolated according to the manufacturer's protocol. Plasmids were prepared and the cDNA-insertions were sequenced. The full-length cDNA from AtPLA I with a length of 4.1 kb could not be identified in the lambda cDNA-library Uni-ZAP XR (Stratagene). Total RNA was prepared from 28-d-old Arabidopsis plants by the RNA-purification kit from Qiagen (Hilden, Germany). Three micrograms of RNA was used for first-strand cDNA-synthesis using oligo(dT)12 --18 (Invitrogen, Carlsbad, CA). One hundred nanograms of cDNA was taken for RT-PCR using the primers, sense (5'-GAT CTG AAC GAC GAT CCG ATT C-3') and antisense (5'-GCA GAA ACG AAC AAA ACT TCG-3') to amplify the full-length cDNA of the gene AtPLA I. PCR conditions were 35 cycles of 94C for 15 s, 55C for 30 s, and 72C for 120 s. The amplified cDNA was cloned via adenosine-overhangs produced by the Taq-DNA-polymerase (AGS-Gold, AGS) into the vector pCR2.1 (Invitrogen). RT-PCR Gene Expression Analysis | Total RNAs were prepared from roots, shoots, leaves, and flowers and treated with DNase according to the manufacturer's protocol (Qiagen). The RNAs were subsequently examined with respect to contaminating DNA using PCR. For this purpose, the RNA preparations were used for amplification reactions with each gene-specific PLA primer pair for 40 cycles and analyzed by gel electrophoresis. Initiated by an oligo(dT)12 --18 primer (Invitrogen), first-strand cDNA was synthesized from each 3 mug of total RNA to quantify PLA mRNAs. The cDNAs were amplified with gene-specific primer pairs each derived from different exons and comprising at least one intron sequence in the corresponding genomic DNA-sequence AtPLA I: sense, 5'-ATA TTG GAC GCC AGA CCC TA-3', and antisense, 5'-TGT TTT CTC GTG GTC GCT ATC-3'; AtPLA IIA: sense, 5'-CGT CCT TGA GAG GAT GGT TT-3', and antisense, 5'-TGG ATG GAG AAG AAG CAA GG-3'; AtPLA IVA: sense, 5'-TTC GAC CGG TTT CTC GTT AT-3' and antisense, 5'-TCG TCC GAG AGA ATT TTT GC-3'; and AtPLA IVC: sense, 5'-GGA CCC AAG TTC AAC GGT AA-3', and antisense, 5'-CAC CCT TCA ACG AAT CAT CA-3'. A total of either 33 cycles (AtPLA I) or 35 cycles (AtPLA IIA and AtPLA IVA) were chosen, because then amplification was in the logarithmic range when 100 ng cDNA were used. A total of 100 ng of purified first strand cDNAs was used for subsequent amplification reactions. Conditions for amplification were 94C for 30 s, 55C for 30 s, and 68C for 45 s. The RT-PCR products were 414 bp (AtPLA I), 390 bp (AtPLA IIA), and 432 bp (AtPLA IVA) in length, and their identities were verified by sequencing. Competitor DNA as an internal standard was obtained by the addition of 25 fg (AtPLA I), 1.5 fg (AtPLA IIA), and 4 fg (AtPLA IVA) of the appropriate cloned genomic DNA fragment harboring an intron, which led to the respective larger amplification products (competitor-AtPLA I = 752 bp; competitor-AtPLA IIA = 475 bp; competitor-AtPLA IVA = 650 bp). The PCR products were analyzed by electrophoretic separation on agarose gels and quantified by using the program "scan pack" (Biometra, Gottingen, Germany). The relative amounts of cDNAs were calculated by the formula [amount competitor DNA] x [gray scale value of cDNA]/[gray scale value of competitor DNA], setting the highest value to 100%. Because the gene-specific primer pair for AtPLA IVC resulted in amplification of two splicing-variants, a competitive RT-PCR described above was not possible. To normalize RT-PCR amplification, control reactions were run with the primers sense, 5'-AGG ATA TTC AGC CAC TTG TCT GTG-3', and antisense, 5'-AGA AAC ATT TCC TGT GAA CAA TCG-3', derived from consensus sequence of the Arabidopsis actin genes Act 2 and Act 7 . The amounts of first strand cDNA were 100 ng for AtPLA IVC transcript amplification and 20 ng for actin transcript amplification. PCR was performed through 25 cycles of 94C for 15 s, 56C for 30 s, and 72C for 90 s for actin gene fragments. Transient Hybrid PLA::GFP Expression | The open reading frame of GFP from psmGFP was amplified in a PCR-reaction for 30 cycles (each cycle: 94C for 15 s, 50C for 30 s, and 72C for 60 s) using the sense primer (5'-GCT CTA GAG TCG ACA TGA GTA AAG GAG AAG AAC-3') and the antisense primer (5'-GCG AGC TCG AGC TCT TAT TTG TAT AGT TCA T-3') to introduce a XbaI and SalI recognition site at the N terminus of the open reading frame of the GFP. The amplified PCR product was digested with SacI and XbaI and was ligated into the vector pBI221. The open reading frames of the PLA cDNAs of genes AtPLA I, IIA, IVA, and IVC were amplified by PCR using the sense primers for IA (5'-GCT CTA GAA TGT CTT CTA CAT GTT CTT C-3'), for IIA (5'-AGC TCT AGA ATG CAA ATG GAC AGC CCC-3'), for IVA (5'-AGC TCT AGA ATG GAG AAC AAA TCG CC-3'), and for IVC (5'-AGC TCT AGA ATG GAT ACA GAG AGA-3') containing a XbaI-recognition site and the antisense primers for I (5'-ACG CGT CGA CTA CAC TAG GAA GAT GAC AAG-3'), for IIA (5'-ACG CGT CGA CGA TCC TAA TTG GAG CTT TTG-3'), for IVA (5'-ACG CGT CGA CCT CTT GTG ATT CAT TTG ATG-3'), and for IVC (5'-ACG CGT CGA CAT TAT TAA ACC TTT TGA GAG-3') containing a SalI recognition site. After digestion of the modified pBI221 and the PLA amplification products by XbaI and SalI, the open reading frames of PLA I, IIA, IVA, and IVC were fused into the pBI221 to the N terminus of the GFP. Plasmid preparations of PLA-psmGFP were done as described in the protocol "Plasmid Mini Preparation" (Qiagen) but without using the purification step by columns and then dissolved in sterile deionized water. One microgram of plasmid (each PLA isoform) was bound to wolfram particles (M17, Bio-Rad, Hercules, CA) in the presence of 30 mum CaNO3. PLA-psmGFP-wolfram particles were shot by using a particle gun (PDS100, Bio-Rad) into leaves from Nicotiana benthamiana being 3 to 4 cm in diameter. The transformed leaves were incubated in petri dishes for 72 h in the dark to express the PLA-GFP-fusion proteins and were analyzed by confocal laser microscopy (TCS 4D, Leica Microsystems, Bensheim, Germany). For each construct, three independent repeats were done. Expression and Purification of the Recombinant PLA Proteins | The full-length cDNA from AtPLA IVA was amplified by using the sense 5'-CGG GAT CCG AGA ACA AAT CGC CCT CC-3' containing a BamHI restriction site and the antisense primer 5'-ATT CTG CAG TTA TTT TAT CTC TTG TG-3' containing a PstI restriction site. The PCR product was digested by BamHI and PstI, purified by column (Cycle pure, PEQLAB, Erlangen, Germany), and then cloned into BamHI- and PstI-digested pQE30 plasmid, a his-tagged expression vector (Qiagen). Escherichia coli strain XL 1Blue was used as the host for the transformation. The his-tag was fused to the N terminus of the PLA. E. coli cells with the pQE30-AtPLA IVA plasmid were grown to OD600 = 0.5 in 250 mL of Luria-Bertani broth supplemented with 100 mug mL-1 ampicillin and 25 mug mL-1 tetracycline. IPTG was added to 0.5 mm to induce expression of recombinant proteins, and the culture was continued at 30C for 30 min. After harvesting by centrifugation, the cells were frozen at -20C overnight or in liquid nitrogen. The recombinant protein purification method was slightly modified from that of the QIAexpressionist manual (Qiagen). The bacterial cells were resuspended in 10 mL of ice cold 50 mm Na2HPO4 (pH 7.5), 300 mm NaCl, 10 mm imidazole, 0.1% (v/v) Tween-80, 10% (v/v) glycerol, and 1 mg mL-1 lysozyme, and digested with 10 mug mL-1 DNase I for 1 h on ice. The cells were ruptured by a French press (three cycles, 1,000 psi). After centrifugation for 12 min at 4C (10,000g), the supernatant was incubated on ice with nickel-nitrilotriacetic acid agarose (Qiagen) for 20 min under shaking and washed twice with buffer (50 mm NaH2PO4, pH 8.0, 300 mm NaCl, and 20 mm imidazole) by centrifugation. The recombinant protein was eluted by 50 mm NaH2PO4 pH 8.0, 300 mm NaCl, and 250 mm imidazole. Expression of the transformed E. coli strains as well as purity, Mr, and quantitative amounts of recombinant PLA proteins were analyzed by 10% (w/v) SDS-PAGE according to . Protein amounts were determined by the method of . Assay of PLA Activity | Liposomes were prepared by drying a mix of 186.7 mug of soy PC (Sigma P3644) and 13.3 mug of fluorescence-labeled phospholipid [bis-BODIPY-PC and 2-(4,4-difluoro-5-methyl-4-bora-3a,4a-diaza-s-indacene-3-dodecanoyl)- 1-hexadecanoyl-sn-glycero-3-phosphocholine) in 100 muL of chloroform under a stream of nitrogen, followed by resuspension in 50 mm Tris, pH 8.0, 100 mm KCl, and 0.1% (w/v) sodium deoxycholate at a concentration of 1 mug muL-1 and sonication for 5 min to achieve a clear suspension. For routine enzymatic tests, recombinant protein (0.5 mug in 6 muL) was incubated with 3 muL of liposomes in a total volume of 100 muL of MES-KOH, pH 6.8, and 1 mm CaCl2 at 33C for 30 min. The pH dependence was determined by testing at different pH values of 5.5, 6.0, 6.6, and 7.24 in 50 mm MES buffer; at pH values of 7.3, 7.41, and 7.8 in 25 mm HEPES buffer; and at pH 8.0 and 9.0 in 50 mm Tris buffer. PLA2 inhibitors AACOCF3, NDGA, HELSS, ETYA, and PACOCF3 were added from stock solutions adjusting the solvent concentration to 0.25% (v/v) dimethyl sulfoxide. Reactions were stopped by a slightly modified protocol of by extraction with 2 volumes of the stopping solution (methanol:chloroform = 2:1, v/v), 1 volume 0.1 m KCl was added subsequently, and then incubation at -20C for 20 min followed. After centrifugation for 30 s at 10,000g, the organic phase was evaporated by a stream of nitrogen and redissolved in 10 muL chloroform for thin-layer chromatography on silica gel 60 (Merck, Darmstadt, Germany) in a solvent of chloroform:methanol:water (65:25:4, v/v). Plates were dried and scanned optically by a video camera under UV light for computer-assisted quantification. All assays were repeated at least three times. Hypcotyl Elongation Growth Test | Sterile Arabidopsis seeds were stratified in the dark in deionized water at 4C for 2 d. After transfer into 2 mL of one-half-strength Murashige and Skoog medium, seeds were pretreated with light at 24C and then grown for 48 h under gentle shaking (80 rpm). The PLA2 inhibitors (AACOCF3, NDGA, and HELSS), ETYA, and PACOCF3 were dissolved in ethanol or dimethyl sulfoxide. The solvent concentration was adjusted in all samples to 0.25%, and after additional growth for 18 h at 24C, the length of the hypocotyls were measured from video photographs by using the program dhs-Bild-Datenverarbeitungsprogramm (Leica). Backmatter: PMID- 12226490 TI - The Circadian Clock That Controls Gene Expression in Arabidopsis Is Tissue Specific AB - The expression of CHALCONE SYNTHASE (CHS) expression is an important control step in the biosynthesis of flavonoids, which are major photoprotectants in plants. CHS transcription is regulated by endogenous programs and in response to environmental signals. Luciferase reporter gene fusions showed that the CHS promoter is controlled by the circadian clock both in roots and in aerial organs of transgenic Arabidopsis plants. The period of rhythmic CHS expression differs from the previously described rhythm of chlorophyll a/b-binding protein (CAB) gene expression, indicating that CHS is controlled by a distinct circadian clock. The difference in period is maintained in the wild-type Arabidopsis accessions tested and in the de-etiolated 1 and timing of CAB expression 1 mutants. These clock-affecting mutations alter the rhythms of both CAB and CHS markers, indicating that a similar (if not identical) circadian clock mechanism controls these rhythms. The distinct tissue distribution of CAB and CHS expression suggests that the properties of the circadian clock differ among plant tissues. Several animal organs also exhibit heterogeneous circadian properties in culture but are believed to be synchronized in vivo. The fact that differing periods are manifest in intact plants supports our proposal that spatially separated copies of the plant circadian clock are at most weakly coupled, if not functionally independent. This autonomy has apparently permitted tissue-specific specialization of circadian timing. Keywords: Introduction : Light is a key environmental signal for plants, regulating gene expression and development . Changes in fluence rate and light quality can occur unpredictably and rapidly during the day but have an underlying day-night cycle. Plants have evolved a circadian timing system that allows the anticipation of this predictable rhythm. When plants are deprived of environmental time cues and placed in constant ("free running") environmental conditions, circadian rhythms persist with a period of around 24 h, often for many days . Within the circadian system of the whole organism, the term "circadian oscillator" has been used to denote the parts of the system responsible for rhythm generation. Light-dark signals entrain the oscillator via input phototransduction pathways, synchronizing its phase with the environmental light-dark cycle and also affecting its period. Rhythmic output from the oscillator controls a large number of physiological processes in plants . The abundance of 2% to 6% of RNA transcripts in Arabidopsis plants was scored as circadian-regulated in two recent microarray analyses, for example . The rhythmic expression of chlorophyll a/b-binding protein (CAB or Light-Harvesting Complex [LHCB]) genes has often been used as a marker for circadian regulation in plants (for review, see ), especially using firefly (Photinus pyralis) luciferase (LUC) reporter fusions . CAB genes are strongly expressed in the mesophyll cells of photosynthetically active organs and in epidermal guard cells. However, physiological analysis shows that the plant circadian system comprises many copies of the circadian clock, with at least one clock in all the major plant organs and possibly one in most cells . We use the term "circadian clock" to denote the smallest complete timing unit (comprising an oscillator with light input and output to overt rhythms). CAB rhythms, thus, reveal only a subset of clocks in the plant circadian system. The distributed circadian clocks are thought to function autonomously, because many circadian rhythms persist in isolated tissue explants (for example, ; ; ; ). Within a single plant, we have shown that various organs can maintain rhythmic expression of the same gene set to different phases . Central circadian pacemakers and communicated rhythmic signals, therefore, have little influence on plant rhythms, at least for the gene expression rhythms tested . This is in contrast to their well-documented involvement in mammalian rhythms . The autonomy of plant clocks implies that, if the timing properties of the circadian clocks varied among tissues, those differences would be reflected in each tissue's circadian rhythms. Such specialization in timing might be advantageous . It is unclear how much circadian timing actually varies among plant tissues and what are the molecular causes of such variation. Differences in circadian period between plant rhythms have been reported . The rhythms in question were not only expressed in different cells but also had overtly unrelated mechanisms (leaf movement and CAB gene expression, for example). To investigate the heterogeneity of plant circadian clocks, we required rhythmic markers that can be tested for period under constant conditions, in broader spatial domains than the CAB genes. Here, we characterize the rhythmic expression of a CHS promoter:reporter gene fusion (CHS:LUC). CHS is expressed predominantly in epidermal cells of aerial organs and in roots, in Arabidopsis as in other species . We show that its circadian rhythm has a significantly different period than rhythmic CAB expression, implying that CHS is controlled by a different circadian clock. To test whether the same molecular components are required for circadian control of CHS and CAB, we assayed the two markers in mutant backgrounds that are known to alter CAB rhythms by different mechanisms. The mutations had very similar effects upon both CHS and CAB rhythms, indicating that the circadian clocks share similar molecular components. The heterogeneity among circadian rhythms of plant gene expression is likely attributable to tissue-specific modifiers of a common biochemical oscillator, which is present in many if not all Arabidopsis cells. RESULTS : Expression of CHS:LUC in Roots and Leaves | Fusions of firefly LUC to the promoters of the white mustard (Sinapis alba) chalcone synthase (mCHS) and Arabidopsis chalcone synthase (CHS) genes were transformed into Arabidopsis plants of the Landsberg erecta (Ler) and C24 strains, respectively. Figure B shows that luminescence driven by the CHS promoter was evident in the leaves, hypocotyl, and roots of 12-d-old seedlings. Particularly high expression occurred in the shoot apical region and in young lateral roots. Younger, 5-d-old seedlings showed no CHS expression in the hypocotyl (data not shown), consistent with previous reports . The mCHS promoter had an identical pattern of expression (data not shown). In contrast, CAB:LUC activity was largely confined to green tissues, with a gradient down the hypocotyl and no detectable activity in the roots (Fig. D). Figure 1 | Tissue-specific expression of the CHS promoter. Tissue-specific expression of the CHS promoter. Reflected-light (A and C) and luminescence (B and D) images of CHS:LUC (A and B) and CAB:LUC (C and D) seedlings after 12 d of growth in LD (12, 12) of 150 mumol m-2 s-1. Luminescence video images were processed in false color (see "Materials and Methods"): White and red shades represent the highest photon counts, and darker blues, the lowest. A reflected-light image of the plant is shown to the left of each false color image. E through H, Leaf sections were prepared from plants grown in LD (12, 12) of 150 mumol m-2 s-1 for 7 d and then transferred to 250 mumol m-2 s-1 for 5 d. E, Microtome section of a CHS:beta-glucuronidase (GUS) leaf, stained for GUS activity and counter-stained with ruthenium red. F through H, Thick hand section of a CHS:LUC leaf: bright field (F), luminescence (G), and overlaid (H) images. E and F, Scale bars represent 60 mum. E and H, Arrowheads indicate areas where CHS expression is very weak or absent from the mesophyll. The spatial pattern of CHS promoter activity was examined in greater detail in leaf sections from CHS:LUC plants and plants carrying a CHS-GUS fusion transgene. The CHS:LUC signal was highest in the upper epidermis, but both reporters showed patchy expression also in the palisade mesophyll layer (Fig. , E and G). The epidermal expression of CHS:LUC was about 2-fold higher than the mesophyll expression (in Fig. G, for example, average counts per pixel per minute were 0.36 for epidermis and 0.19 for upper mesophyll). A minor contribution from rhythmic CHS expression in the mesophyll would not be distinguished in our assays. Luminescence from the mesophyll is appreciably reduced by passage through the epidermis , indicating that the CHS:LUC rhythm preferentially reflects epidermal luminescence. CHS-GUS expression was also evident in all cell layers of root cross sections (data not shown). The CHS Expression Rhythm Is Distinct from the CAB Expression Rhythm | The strong CHS:LUC activity in the root might make this a useful marker for circadian rhythms specifically in this organ, where root pressure and ion fluxes are the only previously reported, rhythmic markers . The circadian rhythm of CHS:LUC activity was tested under several fluence rates of constant light, because CHS mRNA levels increase in response to higher fluence rates . Transgenic seedlings were grown for 12 d on vertical agar plates under white light of 60, 150, or 250 mumol m-2 s-1 fluence rate. Seedlings grown and assayed in white light of 60 mumol m-2 s-1 showed low expression of CHS:LUC and mCHS:LUC (data not shown). Luminescence rhythms were measured after the plants were transferred to constant light of either the same fluence rates (Fig. , A and C), or of an increased fluence rate (from 150 to 250 mumol m-2 s-1; Fig. B). CHS:LUC activity showed circadian rhythmicity in aerial organs and roots under all conditions. The peak of activity on the 1st d in constant light occurred before predicted dawn under all conditions (at zeitgeber time [ZT] 20 --22; ZT is defined as the number of hours since lights-on), 6 or 7 h before the peak of CAB expression. Transgenic plants carrying the mCHS:LUC fusion in the Ler background showed very similar luminescence rhythms (Fig. D). Figure 2 | Rhythmic luminescence in roots and aerial organs of Arabidopsis seedlings expressing CHS:LUC. CHS Rhythmic luminescence in roots and aerial organs of Arabidopsis seedlings expressing CHS:LUC. CHS:Luc (A, B, C, and E) or mCHS:LUC (D) expression was assayed by video imaging (see "Materials and Methods") at the times indicated. A, B, D, and E, Seedlings were grown on vertical agar plates, in 7 d of LD (12, 12) at 150 mumol m-2 s-1. Seedlings in A and D were transferred (at 0 h) to continuous light of 150 mumol m-2 s-1; seedlings in B were transferred to continuous light of 250 mumol m-2 s-1; and seedlings in E were transferred first to 250 mumol m-2 s-1 for 12 h and then to constant darkness. Seedlings in C were both grown and tested at 250 mumol m-2 s-1. The data are representative of at least five independent experiments. White box on time axis, Light interval; black box, dark interval. Lv, Luminescence from aerial organs; Rt, luminescence from roots. The lighting conditions affected the amplitude of rhythmic CHS expression. The high-amplitude rhythm in the 1st d of constant light was followed by a reduction (damping) in the amplitude of the CHS:LUC circadian rhythm within two to three cycles. mCHS:LUC was less affected (Fig. D), possibly because of a difference between the C24 and Ler genetic backgrounds. CAB:LUC activity continued to cycle with high amplitude under constant light of all fluence rates. In plants transferred from lower to higher fluence rates of light (Fig. B), CAB:LUC luminescence levels were reduced compared with CHS:LUC. When 12-d-old plants expressing CHS:LUC were transferred to continuous darkness, transcription from the CHS promoter was greatly attenuated (Fig. E). A single, high-amplitude peak at the expected phase was followed by rapid damping of CHS:LUC activity, with a complete loss of rhythmic amplitude by 72h. The CAB and CHS rhythms had different free-running periods under constant light. The lag (phase angle) between the rhythms, therefore, changed progressively during our experiments, most obviously when Ler plants were imaged on the 3rd to 7th d under constant light (Fig. D). The altered period was clear: The first peak of mCHS expression shown (at 68 h) occurred 10 h before the CAB peak (at 78 h), but by 147 h, the mCHS peak occurred with or slightly after that for CAB. Period estimates were derived from luminescence rhythms measured in the aerial tissues, where both CHS:LUC and CAB:LUC are expressed . The period of the CHS rhythm (25.4 h) was significantly longer than the period of the CAB rhythm (23.7 h). A similar period difference was observed between CHS and CAB expression rhythms in the Ler background (as in Fig. D). In contrast, the period of rhythmic CHS expression in the root was not significantly different from that in the leaf, in either genetic background (Fig. ; Table ; data not shown). These results suggested that the circadian system that controlled CHS in aerial organs was distinct from that controlling CAB. Table I | Circadian period of gene expression in the aerial organs of det1 Table II | Circadian period of LUC activity rhythms in toc1 de-etiolated 1 (det1) and timing of CAB expression 1 (toc1) Mutations Shorten the Period of Both CHS and CAB Rhythms | The det1 mutant has a severe short-period phenotype for CAB expression , along with other phenotypes principally related to light signaling . toc1 is a short-period mutant that affects only clock-regulated processes ; TOC1 encodes one of the best candidates for a plant circadian oscillator component . The CHS:LUC construct was crossed into these mutant backgrounds to test whether these genes are involved in the circadian systems that control both the CAB and CHS markers. det1 and toc1 mutant seedlings expressed CHS:LUC rhythmically in both leaves and roots under constant light (Fig. A; data not shown). Light-grown det1 seedlings express the CHS promoter in all leaf cell layers, and express CAB genes inappropriately in roots and in aerial tissue . CAB:LUC activity in det1 roots was barely detectable and was too low for us to measure circadian rhythms (data not shown). The det1 mutation prevents the damping of rhythmic CAB expression in darkness (Fig. B; ). It had little effect on CHS expression, which damped out in the dark similarly in the mutant (Fig. B) and wild type (Fig. E). Consistent with previous data, the toc1 mutation did not affect the mean expression level or damping of CAB and CHS expression rhythms (data not shown; ). Figure 3 | Circadian rhythms of CHS:LUC activity in det1. Circadian rhythms of CHS:LUC activity in det1. Expression was assayed as in Figure . Seedlings were grown in LD (8, 16) at 20 mumol m-2 s-1 to confirm the det1 phenotype. Seedlings were transferred to LD (12, 12) at 150 mumol m-2 s-1, 2 d before the start of the experiment and to constant light of 250 mumol m-2 s-1 at 0 h (A) or to 250 mumol m-2 s-1 at 0 h and constant darkness at 12 h (B). Luminescence data were analyzed from whole seedlings, without separate analysis of roots and aerial tissues. The data are representative of at least four independent experiments. White box on time axis, Light interval; black box, dark interval. Quantitative comparisons showed that the period of CHS expression was longer than the period of CAB expression in both mutant backgrounds (Tables and ), reinforcing the conclusion that separate clocks control these two promoters. However, both rhythmic markers were very similarly affected by the mutations (Tables and ), with much shorter periods in mutant progeny than in wild-type controls. This result indicates that TOC1 and DET1 function similarly in the circadian clocks that regulate CAB and CHS. DISCUSSION : Chalcone synthase is one of the key biosynthetic enzymes controlling anthocyanin formation. These flavonoid pigments function to protect plant cells from UV radiation and from pathogen attack, act as insect repellents, and are involved in plant-microbe and pollen-pistil signaling . We used the noninvasive LUC reporter gene in fusions with the CHS promoter (CHS:LUC), to monitor the dynamic pattern of CHS expression . The rhythmic luminescence of CHS:LUC seedlings showed that the circadian clock regulates CHS expression at the level of transcription . CHS:LUC represents the first noninvasive molecular marker for circadian rhythms in root tissue. Regulation of CHS | The expression of CHS exhibited a very similar circadian rhythm in all tissues . The phase of the circadian rhythm of CHS was earlier than CAB, with a peak occurring in the late subjective night (ZT 20 --22). Chalcone synthase enzyme activity and mRNA abundance have previously been reported to exhibit diurnal and circadian regulation . The reported phase of peak enzyme activity lags slightly behind the peaks of mRNA abundance and of transcriptional activity, suggesting that the circadian system principally regulates CHS expression at the transcriptional level. It may be advantageous for the plant to accumulate photoprotective pigments in advance of the daily photoperiod ; the timing of CHS transcription before dawn is consistent with this notion. Increasing the fluence rate of white light from 150 to 250 mumol m-2 s-1 had little effect on the phase or period of CHS or CAB expression (Fig. ; data not shown). However, mean expression levels were differentially affected by the lighting conditions. Plants entrained at 150 mumol m-2 s-1 but assayed at 250 mumol m-2 s-1 showed reduced levels of CAB:LUC activity compared with CHS:LUC activity (Fig. B), perhaps reflecting a requirement for increased photoprotection and reduced light-harvesting capacity. High-fluence rate light was required to maintain high expression levels of CHS transcription in wild-type plants: CHS:LUC activity was very low in plants assayed at 60 mumol m-2 s-1 (data not shown) or in darkness (Fig. E). The det1 mutation increases CAB expression levels in dark-adapted plants but had little effect on the level of CHS expression (Fig. B; ; ). This presumably reflects a differential involvement of DET1 in the phototransduction pathways that regulate these promoters . Differences in the Circadian Regulation of CHS and CAB | The rhythm of CHS expression was very similar in roots and in aerial organs under constant light, indicating that the clocks controlling CHS in these organs do not differ significantly. The period of CHS expression in aerial tissues was approximately 1.5 h longer than the period of CAB expression. This difference was maintained in two wild-type accessions and in the det1 and toc1 mutants (Tables and ). Distinct circadian clocks, therefore, control the rhythms of CAB and CHS expression, although both are nuclear genes that could in principle respond to the same regulator. The phase of a single rhythm (free calcium concentration) has recently been shown to vary among tissues of transgenic tobacco (Nicotiana tabacum; ). The authors point out that a phase difference could result from tissue-specific clocks or from tissue-specific responses to a single, common clock. Where period differences are observed, the latter interpretation can be ruled out . It is not surprising that the period difference is small. Studies on cyanobacteria indicate that a clock with a period that matches the environmental light-dark cycle provides a competitive advantage . Balancing selection is, therefore, likely to maintain periods in a narrow range around 24 h. Consistent with this notion, we have previously shown that the similar periods of several Arabidopsis accessions are the result of balancing long- and short-period alleles at multiple loci . Period differences under constant conditions are relatively easy to measure. Under light-dark cycles, however, clocks with different periods will entrain to different phases (for example, ; ). A clock with a longer period (such as that controlling CHS) will be set to a later phase, all else being equal. The longer period moves the peak of CHS expression away from midnight and toward dawn. If CAB was controlled by the same, longer-period clock, then the peak of CAB expression would also be later in the day, all else being equal. The delay between the peak of CHS expression and the peak of CAB expression would thus be greater than we observe. Independent clocks allow the temporal sequence of metabolic processes to be fine-tuned: A smaller delay between the peaks of CHS and CAB expression might be one example of this. Such a flexible timing system has potential selective advantages , although these remain to be demonstrated experimentally. Differential Regulation of Circadian Period | The circadian clocks that control CAB and CHS might differ fundamentally in their oscillator mechanism, or they might alternatively be separate copies of a common biochemical mechanism with only minor modifications leading to the period difference . det1 and toc1 mutations are thought to affect the circadian rhythm of CAB expression via the input pathway and the oscillator, respectively . Each mutation shortens the circadian period of CHS in aerial organs in parallel with CAB (Tables and ). This result shows that both circadian clocks share at least the TOC1 and DET1 functions, so the clocks are not radically different. A parsimonious explanation is that the CHS and CAB promoters are controlled by separate copies of the same clock mechanism . The characteristics of circadian timing also vary among rodent organs, outside the brain . Differences in input pathways might be particularly important in that case . The same canonical clock genes seem to be involved in diverse anatomical locations . Figure 4 | Autonomy and specialization of the circadian clock. Autonomy and specialization of the circadian clock. A simplified model of the circadian clock is shown in the mesophyll and in the epidermis, with the output markers and clock-related genes tested in this work. The clock is depicted as containing the same components in each case; TOC1 is thought to function in the oscillator, and DET1 is thought to affect the light input pathway from the photoreceptors. Rhythmic output from the oscillator is shown regulating transcription factors at the CAB or CHS promoter. Two hypothetical tissue-specific factors are shown: modification of the light input pathway and an unknown tissue-specific factor (X). One or both modify the function of an oscillator component, resulting in a longer period in the epidermis. CHS is expressed principally in the epidermal cell layer of wild-type aerial organs, whereas CAB is expressed in the mesophyll layers. It is most likely that their different circadian periods reflect tissue-specific modifications of the clock in epidermal and mesophyll cells, respectively. In support of this conclusion, we have recently shown that the PHYTOCHROME B gene, which is expressed in the epidermis, also has a longer period than CAB (A. Hall, L. Kozma-Bognar, F. Nagy, and A.J. Millar, unpublished data). These data imply that epidermal circadian clocks are not tightly coupled to clocks in the mesophyll of the same leaf. We have previously demonstrated that several autonomous clocks exist within the mesophyll of a single leaf . Taken together, our results suggest that circadian control is local to one or a few cells, not widely coupled within or between tissues, at least for gene expression rhythms in the leaf. The molecular cause of the observed period difference is unclear. Any circadian input pathway might contribute . Photoreceptor genes are differentially expressed in the epidermal and mesophyll cell layers , and photoreceptors are known to alter circadian period in a dose-dependent manner . The light input pathways might thus control circadian period in a tissue-specific fashion. Observed distinctions between circadian rhythms suggest that further specialization of the circadian clock might be present in the cells that sense or respond to photoperiodic signals (for example, ; ) or control rhythmic leaf movement . Rhythmic reporters that peak at various phases in restricted spatial patterns are required to determine how much the circadian clock is modulated for specialized timing functions. Recent microarray experiments suggest many candidate promoters , which can now be tested in detail using LUC fusions. MATERIALS AND METHODS : Plant Material, Media, and Growth Conditions | Experiments were performed with transgenic Arabidopsis seedlings carrying the native firefly (Photinus pyralis) LUC gene under the control of one of two chalcone synthase promoters. Arabidopsis plants of the C24 accession carrying the native Arabidopsis CHS promoter fused to the LUC gene have been described . The Columbia transgenic line carrying a fusion of the same Arabidopsis CHS promoter fragment to GUS was also described previously . Transgenic plants of the Ler accession carried the promoter of white mustard (Sinapis alba) CHS1 from -907 to +26 bp , fused to the LUC gene in the pMON 721 binary vector backbone: Four independently transformed lines were tested with very similar results (data not shown). The CAB:LUC transgene introgressed from C24 into the Ler background has been described . CHS:LUC from the C24 accession was crossed with the det1-1 mutant in the Columbia background (from Dr. Joanne Chory, Salk Institute). det1-1 mutant and wild-type plants were selected from the F2 population on the basis of their morphology; mutant scoring was confirmed in the F3 generation. The line carrying CAB:LUC in the det1-1 background has been described . CHS:LUC in C24 was crossed to the toc1-1 mutant that carries CAB:LUC in the C24 background . Homozygous mutant F2 progeny were selected by scoring short-period CAB:LUC luminescence in leaves; among these, seedlings carrying CHS:LUC were selected by scoring luminescence in roots. The CAB:LUC reporter was removed by segregation in the F3, resulting in greatly reduced luminescence in leaves. The presence of CHS:LUC and absence of CAB:LUC was confirmed by PCR using transgene-specific primers designed from published sequences (data not shown). Seedlings were grown at 22C in 12:12-h light-dark cycles (unless otherwise indicated) on agar medium containing 3% (w/v) Suc under cool-white fluorescent lights of the fluence rates indicated for each figure. Our standard conditions (approximately 60 mumol m-2 s-1) gave low CHS expression (data not shown); higher fluence rates of 150 to 250 mumol m-2 s-1 increased expression from these constructs, so all subsequent experiments were conducted within this range. Localization of CHS Expression | Histochemical localization of GUS (Fig. E) was performed as described . Ruthenium red counterstaining was carried out by standard techniques. Tissue sections were prepared for light microscopy using a Technovit 7100 embedding kit (Kulzer, Wehrheim, Germany). High-resolution imaging of LUC activity was performed essentially as described . The image in Figure , G and H, was captured through a Fluar 20x objective (Zeiss, Jena, Germany) by a 30-min exposure on a back-thinned CCD in a liquid-nitrogen cooled camera (LN/CCD-512-TKB with ST133 controller, Roper Scientific Ltd., Marlow, UK). Camera background has been removed from Figure G. The specific signal in this image ranges from 2 (dark blue) to 24 (white) counts per pixel. Dark current was undetectable and readout noise was 1.8 counts (sd). The lowest luminescence levels have been clipped in Figure H to reveal the tissue outline. Imaging of LUC Bioluminescence Rhythms and Statistical Analysis | After luciferin pretreatment , seedlings were sprayed before each image with 1 mmd-luciferin (Promega, Madison, WI) in 0.01% (w/v) Triton X-100 and left in darkness for 5 min to allow chlorophyll chemiluminescence to decay. Bioluminescence was detected by ultra-low-light cameras (VIM C2400 --47, Hamamatsu, Bridgewater, NJ; and Roper Scientific LN/CCD-512-TKB with ST138 controller), as described . The data from both cameras were similar, although absolute luminescence counts are not directly comparable because of the different image acquisition methods (for example, in Figs. and ). Data are expressed as detected photons per plant per 25-min image, derived from data from groups of 10 to 20 plants. Data in Figures and have been processed with a three-point, boxcar filter. Period estimates from the raw data were produced by the fast Fourier transform-nonlinear least squares method (FFT-NLLS, ). The first 24 h of each time course was excluded from the period estimation to exclude any transient effects of the transfer to continuous light. To derive mean period estimates from the results of seven experiments, the data of Table were analyzed using residual maximum likelihood (REML) in the statistical package Genstat 5 . REML can be thought of as a generalization of analysis of variance to unbalanced designs. Data were weighted for analysis by the reciprocal of the estimated variance of the circadian period for the trace, which was derived from FFT-NLLS . The data were analyzed with each line taken as a fixed effect and experiment and trace within experiment as random effects. The significance of differences between pairs of treatments were assessed using t tests, based on se of the differences derived from REML, rather than the se of individual means presented in Table . The mean periods and se of Table are derived from the same variance-weighting procedure on period estimates from FFT-NLLS for replicated samples in a separate experiment. Waveforms in the data that are unusually close to a cosine wave can give very low estimated ses from FFT-NLLS and, thus, gain disproportionate weight in the variance-weighted means. The analysis was repeated with revised weights that were derived by adding 0.1 to the original, estimated ses to reduce the effects of such rare estimates. The conclusions of the analysis were not altered by this procedure, and the data presented are weighted using the original estimates. Backmatter: PMID- 12226491 TI - A Role for the DOF Transcription Factor BPBF in the Regulation of Gibberellin-Responsive Genes in Barley Aleurone AB - Functional analyses of a number of hydrolase gene promoters, induced by gibberellin (GA) in aleurone cells following germination, have identified a GA-responsive complex as a tripartite element containing a pyrimidine box motif 5'-CCTTTT-3'. We describe here that BPBF, a barley (Hordeum vulgare) transcription factor of the DOF (DNA-Binding with One Finger) class, previously shown to be an activator of reserve protein encoding genes during development, also has a role in the control of hydrolase genes following seed germination. Northern-blot, reverse transcriptase-polymerase chain reaction, and in situ hybridization analyses evidenced that the transcripts of the BPBF-encoding gene (Pbf), besides being present during endosperm development, are also expressed in aleurone cells of germinated seeds where they are induced by GA, an effect counteracted by abscisic acid. Electrophoretic mobility shift assays have shown that the BPBF protein binds specifically to the pyrimidine box motif in vitro within the different sequence contexts that naturally occur in the promoters of genes encoding a cathepsin B-like protease (Al21) and a low-isoelectric point alpha-amylase (Amy2/32b), both induced in the aleurone layers in response to GA. In transient expression experiments, BPBF repressed transcription of the Al21 promoter in GA-treated barley aleurone layers and reverted the GAMYB-mediated activation of this protease promoter. Keywords: Introduction : In cereal seeds, the endosperm is the major organ for storage of the nutritional reserves that will support seedling growth before the onset of photosynthesis. The endosperm is differentiated into two predominant tissue types, the central starchy endosperm and the peripheral aleurone layer . During seed maturation, the starchy endosperm is committed to the synthesis and deposition of carbohydrates and storage proteins. When the seed desiccation phase initiates, it undergoes a degenerative process related to programmed cell death . By contrast, the aleurone cells remain viable in the mature seed, and, following germination, secrete a number of hydrolytic enzymes that mobilize the stored nutrients, mainly starch and proteins . Gene regulation in these aleurone cells is under the control of phytohormones, mainly of the ratio of gibberellins (GA) to abscisic acid (ABA). GAs are synthesized by the embryo and diffuse into the aleurone , where they trigger the transcriptional activation of a number of hydrolase genes such as those encoding several classes of alpha-amylases, proteases, and beta-glucanases . The activating effect of GA is counteracted by ABA . Although a GA receptor has not yet been isolated, many interesting findings have been documented related to the site of perception of GA, elegantly localized to the external face of the plasma membrane , and related to the transduction events that follow after the signal is perceived (for review, see ; ). However, little is known about the actual transcription factors involved in the regulation of the hydrolase encoding genes. Functional analysis of the alpha-amylase promoters from barley (Hordeum vulgare), wheat (Triticum aestivum), and rice (Oryza sativa) have identified a conserved cis-element required for GA induction, termed the GA responsive complex (GARC). Although this GARC may not always be tripartite, most often it includes three sequence motifs, the TAACAAA box or GA responsive element (GARE), the pyrimidine box CCTTTT, and the TATCCAC box . Mutation analysis of the GARC from low and high pI alpha-amylase genes (Amy2/32b and Amy1/pHV19, respectively) have shown that these three boxes are necessary for a full GA response in transiently transformed barley (Hordeum vulgare) aleurone layers . Disruption of either of them resulted in a decreased level of promoter expression in GA-treated aleurone layers, but did not completely abolish the GA induction. Promoter regions closely resembling the GARC have also been described in other genes expressed in the aleurone such as those encoding the (1-3, 1-4)-beta-glucanase isozyme II , and the thiol-proteases cathepsin B-like and EPB1 . In spite of the similarity of most GARC in different hydrolase promoters, significant differences are also noted that may explain the timing and level of expression controlled by them. In this regard, the EPB1 promoter contains a functional GARE, a pyrimidine box, and an upstream element in different orientations than those in alpha-amylase gene promoters. This has been proposed to be the reason why the EPB1 gene shows a lower level of expression than alpha-amylase genes in response to GA . Although the cis-motifs constitutive of the tripartite GARC element have been well defined, the isolation of the corresponding trans-acting factors have lagged far behind. The first such factor to be characterized was the GAMYB protein from barley . GAMYB was induced by GA in aleurone cells and, through binding to the TAACAAA-like sequences, was able to trans-activate a number of hydrolase gene promoters induced in those cells upon germination . A transcription repressor (HRT) has also been described associated with the GARE . However, the trans-acting factors recognizing the pyrimidine and the TATCCAC boxes have not yet been identified. In this context, we found relevant that the pyrimidine box, 5'-CTTTT-3', contains in the anti-sense strand the 5'-AAAAG-3' sequence defined as the core motif required for the binding of transcription factors of the DNA-binding with one finger (DOF) class . Protein BPBF (PBF; gene Pbf) is a DOF transcription factor expressed in the developing endosperm of maize (Zea mays), barley, and wheat where it is involved in the activation of seed storage protein genes . Moreover, reported the isolation of five cDNAs encoding DOF proteins (OsDof1-5) from a rice aleurone library by southwestern screening with a pyrimidine box probe. In transient assays, OsDof3 was a repressor of the GA-induced activity of a type III carboxypeptidase promoter. These data and the fact that the transcription factor Viviparous-1 (VP1) in maize seems to be involved in the control of seed maturation and germination programs led us to explore whether the barley gene Pbf might also have a role in seed germination. In the present study, we have investigated the expression profile of the barley Pbf gene in seeds following germination and we tested the binding ability of BPBF toward the pyrimidine box sequence in promoters of GA-controlled hydrolase genes. More specifically, we show here that the recombinant PBF was able to bind in vitro the pyrimidine box sequence motif in the cathepsin B-type protease Al21 and in the alpha-amylase Amy2/32b promoter contexts. In transient expression experiments, BPBF repressed the GA-induced activity of the Al21 gene promoter and could largely revert its GAMYB-mediated trans-activation. This indicates that Pbf has a functional role as a transcriptional repressor in aleurone cells, beside being a transcription activator of prolamin genes during endosperm development, as previously described by us . RESULTS : The Barley Pbf Gene Is Expressed in the Aleurone Cells following Germination | To explore the possibility of Pbf being expressed in germinating seeds, total RNA was prepared from dissected tissues (aleurone, scutellum, shoots, and roots) of 2-d-imbibed kernels of the barley cv Bomi. The Pbf transcript was detected in the aleurone sample upon northern-blot analysis (Fig. A). It should be noted that the hybridization signal for Pbf was approximately 50 times lower in the aleurone than in the developing endosperm sample run as a control. Our previous results had shown that Pbf was expressed in developing barley endosperms, but its mRNA was not detected in immature embryos or vegetative tissues . To discard the possibility that the Pbf mRNA was present following germination as a result of storage in the dormant aleurone of mRNAs previously synthesized during development, the kinetics of the mRNA accumulation was examined at different times after seed rehydration (Fig. B). The barley Pbf message was not detected in the mature dry seeds, and, upon imbibition, it was first detected in northern blots at 16 h, gradually increasing until 48 h, and decreasing thereafter (96 h). All blots were subsequently hybridized with a probe for a wheat cDNA encoding a cathepsin B-like thiol protease, which is highly expressed in the aleurone layer during germination . The expression pattern of the thiol protease gene closely followed that found for Pbf, which was compatible with its being regulated by Pbf. The blots were finally hybridized with an 18S ribosomal probe as a loading control . Figure 1 | Northern-blot analysis of barley Pbf expression during germination. Northern-blot analysis of barley Pbf expression during germination. A, Ten micrograms of total RNA from different tissues of 2-d germinated kernels of the barley cv Bomi: shoots (S), roots (R), scutellum (Sc), and aleurone (A). For comparison, a sample of 4 mug of total RNA from developing endosperms (Dv/E) at 15 d after pollination (DAP) was included. B, Comparative accumulation of Pbf transcripts in germinating barley aleurones. Total RNA (10 mug) was extracted from aleurones rehydrated during 0, 16, 24, 48, and 96 h. The blot was first hybridized with the barley Pbf-specific probe (BPBF) and subsequently with a probe for the gene Al21 of the cathepsin-B type from wheat (Thiol protease) and/or a 18S rRNA probe as a loading control. To determine the spatial expression of the Pbf mRNA, we did in situ hybridization experiments. This analysis was performed in barley seeds after 2 d of imbibition and in developing kernels at 15 DAP . During development, a clear signal was detected not only in the starchy endosperm, but also in the aleurone layer (Fig. , A and B), which is clearly differentiated at this stage of development. No expression was observed in the pericarp or in the nucellar tissue. When longitudinal sections of 2-d rehydrated grains were hybridized with the barley Pbf antisense probe, a clear signal was observed that was exclusively localized to the aleurone layer (Fig. , D and E). No signal above background was detected when sections of developing or germinating seeds were hybridized to the sense probe (Fig. , C and F). Figure 2 | Spatial expression of the Pbf mRNA in developing and germinating seeds of barley determined by in situ hybridization. Spatial expression of the Pbf mRNA in developing and germinating seeds of barley determined by in situ hybridization. A, B, and E are transversal sections of 15 DAP developing seeds. D through F are longitudinal sections of 2-d germinating seeds. Sections A, B, D, and E were hybridized with a Pbf antisense RNA probe. Hybridization with a Pbf sense RNA probe as negative controls C and F. al, Aleurone; se, starchy endosperm; p, pericarp. Bars = 100 mum in A, C, E, and F; 250 mum in A; and 500 mum in D. The Expression of Barley Pbf in the Aleurone Is Hormonally Regulated | The observed accumulation pattern of the barley Pbf transcript upon seed imbibition suggested that it may be up-regulated by GA. We have used isolated aleurone layers as a system to study the hormonal control of the Pbf gene because they do not synthesize GA, but are able to respond to it. Aleurone layers were isolated from deembryonated grains of Himalaya barley after 4 d of imbibition and were incubated in the presence of 1 mum GA3 over different periods of time up to 24 h. Total RNA was prepared from these samples and the time course of Pbf mRNA accumulation was analyzed using relative reverse transcriptase (RT)-PCR techniques. For comparative purposes, the induction profile of the barley homolog of the wheat gene encoding the cathepsin B-like thiol protease Al21 was also studied. As shown in Figure A, aleurone layers at the beginning of the incubation period (0 time) contained barely detectable levels of the Pbf mRNA. After 4 h of GA treatment, these levels sharply increased, reaching their maximum by 16 h and then declining slightly. As expected, incubation with GA3 induced the thiol protease transcript levels . It is well documented that the GA effect on hydrolase gene transcription in aleurone layers can be reverted by an external concentration of 10 mum ABA. We analyzed the possible interaction between these two hormones in Pbf expression. Figure B shows the levels of the Pbf mRNA, as well as those of the thiol protease run as a control, in aleurone layers incubated for 16 h in the presence of GA3, ABA, and GA3 plus ABA. The treatment with ABA did not activate the expression of either of the genes tested, and was effective in antagonizing the induction of Pbf transcription promoted by GA3. Figure 3 | RT-PCR analysis of hormone response of Pbf expression in rehydrated aleurone layers. RT-PCR analysis of hormone response of Pbf expression in rehydrated aleurone layers. Total RNA was isolated from aleurone layers after different hormone treatments and was reverse transcribed in the presence of random hexamers. The first strand cDNA was then amplified by PCR using gene-specific primers for the Pbf (BPBF) or the cathepsin B-like thiol protease (thiol protease) transcripts. Amplification of a region of the 18S RNA was used as the internal control (18S). The PCR products were fractionated in agarose gels and were visualized by ethidium bromide staining. A, Comparative Pbf expression pattern in aleurone layers incubated in the presence of 1 mum GA3 for 0, 4, 8, 16, or 24 h. B, Expression analysis of Pbf and of the cathepsin B-like thiol protease transcripts in aleurone layers incubated for 4 h with (+) or without (-) 1 mum GA3, 10 mum ABA, or both, as indicated above. BPBF Binds to the Pyrimidine Box Element from Hydrolase Gene Promoters | Because the complementary strand of the pyrimidine box element (5'-CTTTT-3') in GA-induced hydrolase gene promoters was identical to the core sequence (5'-AAAAG-3') recognized by PBF in prolamin gene promoters (P-box: 5'-T/AAAAG-3'; ; ; ), it seemed plausible that the PBF might bind to the pyrimidine box motif. To test this possibility, two pyrimidine box elements placed in different contexts were assayed in vitro in electrophoretic mobility shift assays (EMSA; Fig. ). A glutathione S-transferase (GST)-BPBF fusion protein was expressed and purified from Escherichia coli extracts and was used in the EMSA with appropriate radiolabeled DNA probes. The thiol protease probe was a 29-mer oligonucleotide containing the pyrimidine box motif 5'-AAAAGG-3' (Fig. , D1) from the promoter of the gene Al21 , considered a high-affinity binding site for PBF, according to . Two other pyrimidine boxes of lower affinity (our data not shown and ) were also present in the Al21 promoter (Fig. , D2 and D3). Figure 4 | In vitro binding of the recombinant BPBF protein to the pyrimidine box sequence motif. In vitro binding of the recombinant BPBF protein to the pyrimidine box sequence motif. EMSA performed with affinity purified GST-BPBF protein and oligonucleotides deduced from the thiol protease AL21 (A; D1 motif in Fig. ) and the alpha-amylase Amy2/32b gene promoters (B). The sequences of the corresponding wild-type (WT) and mutant derivative (mt) oligonucleotides, used as probes, are shown at the bottom of each panel. The pyrimidine box motif is underlined and the point mutations are indicated by lowercase letters. Binding reactions were performed with the GST protein as a negative control (C) or with the GST-BPBF protein (+), in the absence (-) or in the presence of probe competitors at the indicated molar excesses (20x, 50x, and 100x). Figure 5 | Schematic structure of effector and reporter constructs. Schematic structure of effector and reporter constructs. A, Schematic representation of the effector constructs. The barley Pbf (BPBF protein) and GAMyb genes were under the control of the cauliflower mosaic virus (CaMV) 35S promoter (p35S) followed by the first intron of the maize AdhI gene (I-ADHI) and, downstream, flanked by the 3'-nopaline synthase (nos) sequences (nos). B, Diagrammatic structures of the reporter constructs used in this study: pCBG13, pCBG13.1, and pCBG13.5 as described in . The white and black boxes indicate, respectively, the location of putative binding sites for BPBF (D1, D2, and D3) and GAMYB (M) transcription factors, along with the cathepsin B-like thiol protease AL21 promoter sequences. C, Sequence of cis-motifs conforming the putative DOF- and MYB-binding sites in the AL21 gene promoter. Numbers at the left indicate the position from the translation start codon. Motifs containing the BPBF-binding sequence core (indicated by uppercase letters) are shown aligned with the consensus sequence for pyrimidine box elements. The alpha-amylase probe was a 31-mer oligonucleotide derived from the promoter sequence of the low-pI alpha-amylase gene Amy2/32b , which contains a canonical pyrimidine box element (5'-CCTTTT-3'). When these probes were incubated independently with the GST-BPBF protein, a shifted complex was observed that was competed out when the probes were incubated with a molar excess of the corresponding unlabeled oligonucleotides. As expected, this binding was not produced when the control GST protein was used in the assay (Fig. , A and B). The specificity of the interaction was also confirmed using variants of these probes. In a previous study, we had shown that the binding of the BPBF protein to the prolamin-box motif in storage protein gene promoters was abolished when the core AAAG sequence was changed to AgAc . The same nucleotide substitutions were introduced at the pyrimidine box sequence in the thiol protease oligonucleotide (AAAAGG was changed to AAgAcG) to generate a mutant thiol protease probe. Even when a single base substitution was introduced, as in the mutated alpha-amylase probe (changing CCTTTT to CCTcTT), the binding of BPBF was abolished. As shown in Figure , none of these mutant versions of the probes were bound by the GST-BPBF protein or able to compete the binding of the corresponding wild-type probes, even at 100 molar excess. BPBF Negatively Regulates the GA-Responsive Expression of a Thiol Protease Promoter in Cobombarded Aleurones | Because promoters containing a pyrimidine box motif are likely to be the target for BPBF in germinating aleurone nuclei, we addressed the question of whether BPBF modulates their transcriptional activity. Transient expression experiments were done by using particle bombardment into aleurone layers of cv Himalaya barley embryoless half-seeds. The pBPBF effector construct expressed the whole barley Pbf cDNA under the control of the CaMV35S promoter followed by the first intron of the maize AdhI gene , and an equivalent construct with the GAMYB cDNA was used as the GAMYB effector; both constructs had the 3'-noncoding region of the nos gene (Fig. A). As reporters, several fragments of the Al21 promoter fused to the beta-glucuronidase (GUS) gene were chosen. The Al21 promoter constructs assayed had been described in an earlier study and consisted of three fragments spanning to positions -480, -381, and -355, respectively, from the translation initiation codon. Sequence analysis of the -480-bp promoter region in the pCBG13 construct showed (Fig. , B and C) that it contained a canonical pyrimidine box motif at position -401 (Fig. , B and C, "D1"), and two additional imperfect boxes at -434 and -235 (Fig. , B and C, "D2 and D3," respectively). One putative GAMYB-binding site was found at position -367 (Fig. , B and C, "M"). In a first set of assays, the pCBG13 reporter construct was bombarded into rehydrated aleurone cells, alone or with the pBPBF effector, at a 1:1 molar ratio. The GAMYB as effector (1:1) was used as a positive control . The bombarded aleurone layers were then incubated in a solution with 1 mum GA3 or without this hormone. As represented in Figure A, the exogenous GA treatment resulted in similar10-fold enhancement of the GUS activity over that found with the GA-untreated controls. Coexpression of BPBF largely reverted this GA-induced expression, reducing levels of GUS activity to similar15% of that found in GA-treated aleurone layers. As expected, GUS activity from pCBG13 was highly induced by coexpression with GAMYB , reaching comparable levels with those found with the GA-treatment alone. Cobombardment with BPBF greatly reduced the GAMYB trans-activation of pCBG13 in the presence and in the absence of exogenous GA3. These data indicate that BPBF functions in rehydrated aleurone cells as a negative regulator of the GA-dependent or GAMYB-inducible GUS activity driven by the Al21 thiol protease promoter. In the absence of GAMYB or when GA incubation was suppressed, GUS activity was very low, and the repression mediated by BPBF, if any, could not be evaluated (Fig. A). Figure 6 | BPBF negative regulation of the transcription of AL21 thiol protease promoter. BPBF negative regulation of the transcription of AL21 thiol protease promoter. Intact aleurone layers were prepared from half-grains and were transfected by particle bombardment with the indicated combinations of reporter and effector plasmids at a 1:1 molar ratio. A, Effect of BPBF on the AL21 promoter activity of the reporter construct pCBG13. + indicates the presence and - indicates the absence of the effector constructs shown at the left (BPBF and/or GAMYB). When indicated, half-grains were incubated in the presence of 1 mum GA3. B, Effect of BPBF on the GA-induced transcription of three different AL21 promoter constructs. Aleurone cells were bombarded with the reporter construct indicated, alone (white bars) or in combination (1:1 molar ratio) with the BPBF effector (dashed bars), and were then incubated with 1 mum GA3. The p35S construct, where the GUS reporter gene was under the control of the CaMV35S promoter, was used as a control. Relative GUS activity is shown, considering expression levels of the pCBG13 construct in the presence of GA3 as 100%. Column height represents the mean value of 12 replicates from three independent experiments. Bars indicate ses. To address the contribution of the various DOF motifs in the AL21 promoter to the BPBF transcriptional repression, the reporter constructs pCBG13.1 and pCBG13.5 (Fig. B), spanning, respectively, to positions -381 and -355 were assayed (Fig. B). The GA-induced GUS expression from pCBG13.1 was only similar50% of that supported by the pCBG13 construct, but still retained its capacity to be repressed in trans by the BPBF transcription factor. Although the percentage of repression obtained relative to the respective background levels was smaller for pCBG13.1 (40% GUS) than for pCBG13 (18% GUS), the actual remaining activity after BPBF cotransformation was identical in both cases, probably indicating that interaction with D3 plays a major role in the repression mediated by BPBF, whereas promoter sequences between -480 and -381, namely the pyrimidine boxes D1 and D2 (Fig. , B and C), are positive elements. The GUS activity driven by the pCBG13.5 construct appears not to be GA inducible or repressible by BPBF. Considering that pCBG13.1 and pCBG13.5 differ only by the deletion of 26 bp of the promoter sequences, which removes the putative GAMYB-binding site (M), but not the D3 sequence likely to be recognized by BPBF, these results suggest that the repressing activity of BPBF in the aleurone cells may work through interaction with GAMYB. In this context, it is pertinent to mention that recent experiments from our laboratory have detected interaction between GAMYB and BPBF in the yeast two-hybrid system . Overexpression of BPBF had no effect over the transcriptional activity of the constitutive promoter CaMV35S (Fig. B), indicating that BPBF does not act as a general negative regulator of transcription in the aleurone layers. DISCUSSION : The two phases of the life cycle of cereal seeds, development and germination, are associated with the expression of different sets of genes, and the ratio of GA to ABA plays an essential role regulating gene expression in both programs. The success of germination depends on the responses of aleurone cells to GA, which activates the expression of genes encoding hydrolytic enzymes, an effect that is counteracted by ABA . Initial analysis of GA-regulated gene promoters identified a cis-element (referred to as GARE) in cereal alpha-amylase promoters that was able to confer GA-regulated expression to truncated 35S promoters in transient assays . However, it became clear that correct GA regulation of gene expression required additional elements that conformed the tripartite GARC . So far, only transcription factors interacting with the GARE have been described are GAMYB, acting as an activator, and HRT, which is a repressor of transcription . However, to our knowledge, no transcription factors interacting with the other two cis-elements of the GARC have been characterized thus far. Here, we describe that barley BPBF (gene Pbf), a transcription factor of the DOF class previously shown to be an activator of reserve protein genes in the developing barley endosperm , functions as a transcriptional repressor upon germination through interaction with the pyrimidine box of the GARC. As shown by northern-blot analysis, RT-PCR, and in situ hybridization analysis, Pbf is expressed in the aleurone cells following germination in such a way that its mRNA is up-regulated by GA. This pattern of expression, together with in vitro binding assays of BPBF to the pyrimidine box in the context of promoters of alpha-amylase and protease-encoding genes is compatible with BPBF being a transcriptional regulator of hydrolase genes induced by GA in the aleurone cells upon germination. In transient expression assays, BPBF behaves as a transcription repressor of the GAMYB-transactivated or GA-inducible GUS reporter activity controlled by the thiol-protease Al21 gene promoter. According to our results, the regulation of the Al21 gene by BPBF may be modeled under the possible scenario that BPBF can compete with another transcriptional activator(s) of the DOF class for binding to the pyrimidine box motifs. This view of BPBF as a competitor of another putative DOF transcription activator(s) is consistent with the observation that pyrimidine box motifs D1 and D2 are positive promoter elements for the response of the Al21 gene to GA (Fig. B; ). It is also supported by similar results reported in the analysis of the Amy 2/32b , Amy 1/6-4 , and EPB-1 genes. From this evidence, the pyrimidine boxes D1 and D2 are expected to be the site for binding of transcriptional activator(s). According to this model, the relative concentrations of BPBF and of such an activator(s), and their relative binding affinities, would direct the overall effect of the pyrimidine box elements upon GA induction. Although the nature of such an activator remains to be determined, it seems probable that it could also belong to the DOF class of transcription factors. In maize leaves, two DOF factors have been described as being involved in the light regulation of certain genes ; modulation being achieved by competition for a common binding site between Dof1 acting as an activator, and Dof2 acting as a repressor . Genetic evidence for two Arabidopsis DOF genes, DAG1 and DAG2, with opposite effects on germination has been recently reported , although their mechanism of action is not completely elucidated . The repressor activity of BPBF might also operate indirectly through the interaction with other transcription factors associated with the AL21 promoter because BPBF contains no obvious repression domain and it is a good transcriptional activator itself in the yeast one-hybrid system and in the developing endosperm cells . Transient expression in planta (this paper) and yeast two-hybrid assays indicate that BPBF interacts with GAMYB. This result opens the possibility that GAMYB function might be regulated by BPBF and, because the Pbf transcripts appear later than those of GAMYB and are simultaneous to downstream GAMYB-regulated genes such as those encoding hydrolytic enzymes, a possible role for BPBF might be to terminate Al21 GA-induced transcription as part of the start of programmed cell death in aleurone cells. Coupled activator and repressor functions have been well documented for VP1 in the developing maize seeds, where it acts as a transcriptional activator of maturation-specific genes and as a repressor of alpha-amylase-encoding genes normally expressed after germination . Similar to VP1 in maize, during barley endosperm development, the PBF would act as an activator of hordein genes and as a transcriptional repressor of the protease gene Al21. Another transcription factor, expressed during seed development and in aleurone cells following germination, is GAMYB. Recent data from our laboratory indicate that GAMYB is a transcription factor involved in seed-storage protein gene regulation , in addition to being a transcriptional activator of hydrolase-encoding genes in aleurone cells following germination . MATERIALS AND METHODS : RNA Isolation and Northern-Blot Analysis | Seeds of barley (Hordeum vulgare cv Bomi) were germinated at 22C in the dark, and were used to collect samples of aleurone, scutellum, and root and shoot tissues. Developing endosperms were prepared from 15 DAP grains. Total RNA was extracted using the method for aleurone samples, or the procedure for the other tissues analyzed. RNA fractionation in formaldehyde-agarose gels, blotting, and hybridization with [32P]-labeled random-primed probes were performed as previously described . The barley Pbf-specific probe was obtained from an 896-bp SmaI-XhoI fragment of the barley Pbf cDNA clone, which spans from position 302 to 1,198 . Expression of a thiol protease gene of the cathepsin B-like class was analyzed using as a probe the 1,065-bp insert in the 2529 wheat (Triticum aestivum) cDNA clone, reported by . As control for loading, the blots were also hybridized with a 18S rDNA probe from wheat. In Situ Hybridization | Barley seeds of the cv Bomi were collected at 15 DAP or after 2 d of germination and were fixed in 50% (v/v) ethanol, 5% (v/v) acetic acid, and 3.7% (w/v) formaldehyde with an occasional vacuum. After dehydration with ethanol and exchange with xylene, tissues were embedded in Paraplast Plus (Sigma, St. Louis). Sections (8 mum thick) were mounted in Poly-l-Lys and were hybridized overnight at 52C with appropriate Pbf Biotin-labeled probes at a final concentration of 100 ng muL-1, according to the mRNA locator-Hyb and nonisotopic labeling kits (Ambion, Austin, TX). For probe preparation, the barley Pbf-specific fragment of 896 bp, used for northern-blot analysis, was subcloned in pBluescript SK+ (Stratagene, La Jolla, CA) and was amplified by PCR with standard M13 reverse and forward primers. About 1 mug of the PCR product was used as template for the synthesis of the Biotin-labeled, in vitro-transcribed RNA from the T7 (antisense probe) or the T3 (sense probe) promoter sequences. Hybridization signals were detected by color development with a Streptavidin alkaline phosphatase conjugate (Ambion), following the manufacturer's instructions. Sections were stained with 0.1% (w/v) calcofluor and were then photographed under UV light on an microscope (BX60; Olympus, Tokyo). Hormone Treatments and RT-PCR Analysis | Cultivar Himalaya barley seeds (1992 harvest, Washington State University, Pullman) were deembryonated and sterilized in 1.7% (w/v) NaOCl for 10 min, treated with 0.01 m HCl for 5 min, and thoroughly washed with distilled water. Half seeds were placed for 48 h at 22C in the dark on filter paper soaked with a buffer containing 20 mm Na succinate, pH 5.0, and 20 mm CaCl2 for 3 d. Aleurone layers were isolated under a dissecting microscope and were incubated in petri plates at 22C in the dark, for various times, with the buffer described above, including no hormone, 1 mum GA3, 10 mum ABA, or a mixture of both hormones. Total RNA for RT-PCR analysis was isolated from aleurone layers by the RNeasy Plant protocol (Qiagen, Valencia, CA). Contaminating genomic DNA in the RNA preparations was then digested by DNase treatment using the DNA-free system (Ambion). First-strand cDNA synthesis was primed with random hexamers and was catalyzed by M-MuLV Reverse Transcriptase according to the manufacturer's recommendations (Amersham Pharmacia Biotech, Piscataway, NJ). PCR amplification of a 460-bp portion of the Pbf cDNA was performed using the following oligonucleotides as primers: The forward BPBF14 was 5'-ACCCTTCGTTCACCTGATGG-3', which spans an intron-exon boundary in the Pbf gene sequence (I. Isabel-Lamoneda, M. Mena, and P. Carbonero, unpublished data); the reverse BPBF15 was 5'-GACCCAAAAGTTCTCAGGGA-3'. A 650-bp portion of the cDNA encoding a cathepsin B-like thiol protease was amplified using the primers CB1, 5'-TCGCGAATTACACTATTGAGC-3', and CB2, 5'-CACCGGTGATGTG-CTTGTA-3'. These primers were designed based on nucleotide sequences highly conserved between the different wheat cathepsin B-like cDNAs . The resulting PCR products were cloned and sequenced, and this showed that we had amplified the desired target messages without detectable heterogeneity. As an internal control, the 18S amplicon was used, performing the PCR reaction with a mixture of 18S primers/competimers (Ambion) at a 2:8 and 3:7 molar ratio for Pbf and the thiol protease analysis, respectively. All amplifications were carried out for 35 cycles with AmpliTaq Gold DNA polymerase (Applied Biosystems, Norwalk, CT). The PCR products were analyzed by agarose gel (2%, w/v) electrophoresis and were visualized by ethidium bromide staining. EMSA | Plasmid expression constructs and procedures for purification of the fusion proteins GST-BPBF and its mutant derivative GST-mtBPBF have been described previously . The oligonucleotide probes, described in Figure , were generated by annealing complementary oligonucleotides designed to create 5' overhangs that were end-filled by treatment with Klenow DNA polymerase in the presence of [alpha-32P]dATP. Each binding reaction contained approximately 5 ng of GST-purified proteins, 1 ng of gel-purified radiolabeled probe (25,000 cpm), and the indicated excess of unlabeled double-stranded oligonucleotide as a competitor in 15 muL of binding buffer [10 mm HEPES, pH 7.9, 50 mm KCl, 10 mm dithiothreitol, 10% (w/v) glycerol, 1 mug d(I)-d(C), and 2 mg mL-1 bovine serum albumin]. After incubation for 30 min at room temperature, the DNA-protein complexes were analyzed by electrophoresis on 7% (w/v) acrylamide gels (29:1), prepared, and run in 0.5x Tris borate-EDTA, pH 8 (10 mm Tris-ClH, 90 mm boric acid, and 2.5 mm EDTA), at 150 V for 3.5 h at 4C. Gels were dried under vacuum and were autoradiographed using film (X-OMAT S; Kodak, Rochester, NY). Transient Expression Assays in Barley Aleurone Cells | Reporter constructs of the wheat cathepsin B-like protease promoter were obtained by linking different promoter fragments of the Al21 gene to the coding sequence for the GUS reporter gene, followed by the 3'-nos terminator. The pCBG13 reporter plasmid contained a -480-bp fragment upstream of the putative translational start site, and the pCBG13.1 and pCBG13.5 plasmids were deletions spanning up to -381 and -355 of the ATG, respectively . Effector constructs were derived from the plasmid p35IN harboring the CaMV35S promoter followed by the first intron of the maize (Zea mays) AdhI gene and the 3'-nos terminator in a pBluescript vector. Plasmids pBPBF and pGAMYB contained the full-length coding regions of the barley Pbf and GAMyb cDNAs , respectively. Particle bombardment was performed with a biolistic Helium gun device (PSD-1000; DuPont, Wilmington, DE), basically as described in a previous study , but with the following modifications: Each shot delivered equal molar amounts (to a maximum of 175 ng) of the tested reporter plasmid on 0.15 mg of gold particles of 1 mum in size; for cotransfection, the reporter was combined with the effector plasmid(s) at 1:1 molar ratio after preliminary studies indicated that the same inhibition rates were obtained at three effector (BPBF) concentrations (0.5:1, 1:1, and 2:1); and rupture discs of 1,500 psi were used and the distance between macrocarrier and sample was set to 6 cm. Cultivar Himalaya barley seeds (1992 harvest) were deembryonated and sterilized as described above. Half seeds were placed for 48 h at 22C in the dark on filter paper soaked with a buffer (pH 5.0) containing 20 mm Na succinate and 20 mm CaCl2 before removing their pericarp and testa layers under a dissecting microscope. Exposed aleurone layers were shot in sets of eight. Four of them were then incubated separately in the buffer described above with no hormone or with 1 mum GA3 for 24 h at 22C in the dark in a petri plate with gentle shaking. GUS expression was determined by histochemical staining for 24 h according to . Blue spots were counted under a dissecting microscope, and the GUS activity was expressed as the mean value of blue spots per half-grain of aleurone. This measure of GUS activity directly correlates with fluorimetrically quantitated GUS activity per milligram of protein with a correlation coefficient of 0.95 (data not shown). Backmatter: PMID- 12226492 TI - A Strobilurin Fungicide Enhances the Resistance of Tobacco against Tobacco Mosaic Virus and Pseudomonas syringae pv tabaci AB - The strobilurin class of fungicides comprises a variety of synthetic plant-protecting compounds with broad-spectrum antifungal activity. In the present study, we demonstrate that a strobilurin fungicide, F 500 (Pyraclostrobin), enhances the resistance of tobacco (Nicotiana tabacum cv Xanthi nc) against infection by either tobacco mosaic virus (TMV) or the wildfire pathogen Pseudomonas syringae pv tabaci. F 500 was also active at enhancing TMV resistance in NahG transgenic tobacco plants unable to accumulate significant amounts of the endogenous inducer of enhanced disease resistance, salicylic acid (SA). This finding suggests that F 500 enhances TMV resistance in tobacco either by acting downstream of SA in the SA signaling mechanism or by functioning independently of SA. The latter assumption is the more likely because in infiltrated leaves, F 500 did not cause the accumulation of SA-inducible pathogenesis-related (PR)-1 proteins that often are used as conventional molecular markers for SA-induced disease resistance. However, accumulation of PR-1 proteins and the associated activation of the PR-1 genes were elicited upon TMV infection of tobacco leaves and both these responses were induced more rapidly in F 500-pretreated plants than in the water-pretreated controls. Taken together, our results suggest that F 500, in addition to exerting direct antifungal activity, may also protect plants by priming them for potentiated activation of subsequently pathogen-induced cellular defense responses. Keywords: Introduction : When plants encounter pathogen attack, they activate diverse cellular defense responses that are aimed at resisting disease. In the case of gene-for-gene resistance, successful pathogen restriction is frequently accompanied by localized cell death of host tissue (hypersensitive response [HR]; ). This localized cell death mostly results in the formation of visible necrotic lesions. After lesion formation, many plants develop an enhanced (acquired) resistance to a broad spectrum of pathogens, both in the tissue surrounding the primary infection site (local acquired resistance [LAR]) and often also in formerly uninoculated organs (systemic acquired resistance [SAR]; ; ; ; ). In NN genotype tobacco (Nicotiana tabacum), both forms of induced resistance become obvious, for example, by a significant reduction in the size of necrotic lesions produced upon secondary infection with tobacco mosaic virus (TMV) compared with lesions that develop after primary TMV attack . Alternatively, both LAR and SAR may be reflected by absence, reduced number, or delayed appearance of necrotic lesions that form either as a part of the HR or during disease development. The identity of the long-distance signal that travels from the site of primary pathogen infection to the uninoculated parts of the plant to enhance disease resistance is still unknown (for review, see ; ). However, in the past decade, it became clear that in tobacco and some other plants, development of LAR and SAR requires the endogenous accumulation of salicylic acid (SA; ; ). When SA accumulation is disrupted, tobacco and Arabidopsis plants display increased susceptibility to pathogen attack and fail to express both LAR and SAR . In many plants, enhanced disease resistance is frequently accompanied by the activation of genes encoding pathogenesis-related (PR) proteins . Because some of these proteins display antimicrobial activity, their accumulation has often been assumed to contribute to acquired disease resistance. In tobacco, acidic PR-1 is the predominant PR protein and accumulates to 1% to 2% of the total leaf protein in infected leaf tissue . The biological activity of the PR-1 protein is still unclear. However, overexpressing the PR-1 gene in transgenic tobacco plants enhanced their resistance against two oomycete pathogens . This result suggested an important role for PR-1 in acquired disease resistance of tobacco. In addition to the accumulation of PR proteins, some enhanced plant disease resistance is associated with a primed state causing potentiated activation of various cellular defense responses, but only once the protected tissue becomes attacked by a pathogen (for recent review, see ). For example, in systemically resistant Arabidopsis plants, pathogen infection caused stronger activation of PR-1 genes when compared with plants without acquired disease resistance . Similarly, in induced transgenic tobacco plants, the pathogen-elicited expression of an Asparagus officinalis PR-10::beta-glucuronidase chimeric gene was stronger than in plants with no induced resistance . The strobilurin class of fungicides comprises a variety of synthetic plant protecting compounds with broad-spectrum antifungal activity and structural similarity to basidiomycete antibiotics . The mode of action of strobilurin fungicides is the inhibition of mitochondrial respiration by binding to the ubihydrochinone oxidation center of the mitochondrial bc1 complex (complex III), thereby blocking electron transfer . Over the past years, there has also been increasing evidence for direct influences of strobilurins on plant physiology . This physiological effect includes the so-called "greening," that is, even in the absence of challenge by pathogen attack, plants treated with strobilurins are intense green and look healthier than plants that have not been treated with strobilurin fungicides . This suggested to us that, in addition to their fungicidal activity, strobilurins might also enhance the capability of plants to ward off pathogens. In the present paper, we demonstrate that a novel synthetic strobilurin fungicide, F 500 (Pyraclostrobin), accelerates the TMV-induced activation of PR-1 genes in tobacco (Nicotiana tabacum cv Xanthi nc) plants, enhances their resistance against TMV and Pseudomonas syringae pv tabaci, and delays responses that are associated with resistance against P. syringae pv tomato DC3000. RESULTS : F 500 Reduces TMV Lesion Size | To investigate whether F 500 enhances the resistance of tobacco cv Xanthi nc (genotype NN) against TMV infection, on two tobacco leaves, one-half was infiltrated with aqueous suspensions of F 500. As a control, the remaining halves of the leaves were infiltrated with water. Twenty-four hours later, both halves of the two treated leaves were infected with TMV (1 mug mL-1) as described in "Materials and Methods." The area of developing lesions was determined 7 d post TMV infection by fluorimaging. As shown in Figure , infiltrating tobacco leaf tissue with F 500 caused a significant (Student's t test, P < 0.01, n = 26, F 500 = 0.5 mm) reduction in TMV lesion size. The reduction was about 50% when the leaf halves had been infiltrated with an aqueous suspension of F 500 at 0.25 mm and was only slightly more pronounced upon infiltrating F 500 at 0.5 mm. It should be noted that in all the experiments performed, we did not detect any effect of the pretreatment with F 500 on the number of developing TMV lesions. Furthermore, TMV lesion-reducing activity could also be observed when pure F 500 instead of the F 500 formulation was used for leaf pretreatment (data not shown). Thus, the TMV-protecting activity of the F 500 formulation is likely due to the F 500 fungicide. Figure 1 | Reduction of TMV lesion size by pretreatment with F 500. Reduction of TMV lesion size by pretreatment with F 500. On two tobacco plants, one-half of a leaf was infiltrated with an aqueous suspension of F 500 at 0.25 or 0.5 mm. The remaining halves of the leaves were infiltrated with water. Twenty-four hours later, both halves of the two leaves were infiltrated with a suspension of TMV (1 mug mL-1) by gently rubbing two layers of cheesecloth soaked with TMV suspension over the carborundum-covered leaf surfaces. The leaves then were washed under a stream of tap water. The size of developing lesions (+-se) was examined with a fluorimager 7 d later as described in "Materials and Methods." F 500 Enhances the Resistance against P. syringae pv tabaci and Delays Resistance Responses to P. syringae pv tomato | To assay the spectrum of pathogens against which F 500 can affect the resistance or diminish disease symptoms in tobacco, we next included two phytopathogenic bacteria in our experiments. By doing so, we found that pure F 500 delayed the appearance of disease symptoms induced by the virulent wildfire pathogen P. syringae pv tabaci (Fig. A). Bacteria-inoculated control tissue areas initially were chlorotic (within 72 h postinfection), then subsequently water soaked and necrotic (Fig. A), whereas the bacterial population in control leaf halves increased at least over the first 24 to 48 h postinfection (Fig. B). In the F 500-pretreated leaf panels, bacterial population size increased to a much lesser extent (Fig. B) and the appearance of disease symptoms induced by P. syringae pv tabaci was delayed (Fig. A). Figure 2 | Enhanced resistance against P. syringae Enhanced resistance against P. syringae pv tabaci and impaired appearance of resistance responses to P. syringae pv tomato DC3000 upon pretreatment of tobacco leaf tissue with F 500. Tobacco leaf halves were infiltrated with 0.5 mm pure F 500 in 1% (v/v) dimethyl sulfoxide (DMSO; +) or with DMSO (1%, v/v) only (-). Twenty-four hours later, individual leaf panels (asterisks) were infiltrated with P. syringae pv tabaci or with P. syringae pv tomato DC3000 at 1 x 105 colony-forming units (cfu) mL-1 (corresponding to 20 x 103 cfu per leaf disc) through small holes punctured with a needle (some of which are marked with an arrow). A, Symptoms were recorded at the indicated time points with a digital camera. B, The increase in bacterial population size (+-se) was determined in discs from infected leaf panels of F 500-pretreated (black symbols) and F 500-non-pretreated (white symbols) leaf halves at the indicated times. hpi, Hours postinfection. When avirulent P. syringae pv tomato DC3000 was used for tobacco leaf infection, there was activation of an HR associated with the onset of necrosis and dehydration of the tissue (Fig. A), along with a slight and transient increase in bacterial growth that was followed by a dramatic reduction in bacterial population size (Fig. B) within 48 h after infection of control leaf areas. Interestingly, the F 500-pretreated and then P. syringae pv tomato DC3000-infected leaf panels, within 48 to 72 h postinfection, remained essentially free of detectable symptoms of necrosis (Fig. A), even upon microscopic examination, though there was multiplication and increased presence of bacteria within the first 20 to 44 h postinfection of this tissue (Fig. B). At later time points (144 --168 h postinfection), tissue dehydration proceeded on control leaf halves, whereas the F 500-pretreated and subsequently P. syringae pv tomato DC3000-infected leaf tissue was subject to some localized chlorosis (Fig. A). Necrotic leaf areas, if present, were confined to the tissue surrounding the holes punctured with a needle to infiltrate bacteria (Fig. A). Similar effects on bacterial population size and symptom development were seen when the F 500 formulation was used instead of pure F 500 for the pretreatment of tobacco leaf tissue (data not shown). F 500 Does Not Affect the Potency of TMV to Infect Tobacco Leaves | The F 500-caused reduction in TMV lesion size can either be due to an enhanced ability of the treated tobacco leaves to ward off TMV attack or may be due to a direct toxic effect of F 500 on TMV. To address the latter question, aliquots of the TMV suspension were incubated, for 1 and 3 d, in the absence or presence of 0.5 mm F 500 in the growth rooms of the plants. Then, tobacco leaf halves were infected with the respective F 500-pretreated or -non-pretreated TMV suspension as described above and the area of the developing lesions determined 5 d later. Figure demonstrates that there was no significant (Student's t test, P = 0.74, n = 6, 1-d treatment of TMV with F 500) difference in the size of necrotic lesions induced by F 500-treated or -non-treated TMV. Therefore, direct inhibition by F 500 of the capacity of TMV to infect tobacco leaves is unlikely. As in the present experiment , there was no prolonged pretreatment with F 500 of the leaf tissue before TMV infection and because F 500 was present only in the soaked cheesecloth used for leaf infection (white bars), lesion-reducing activity of F 500 was not significant in this experiment . Figure 3 | F 500 does not affect the infection potency of TMV. F 500 does not affect the infection potency of TMV. A TMV suspension (1 mug mL-1) was incubated, for 1 and 3 d, in the absence (-) or presence (+) of 0.5 mm F 500 under plant growth conditions. Tobacco leaf halves then were infected with the F 500-pretreated or -non-pretreated TMV suspension. TMV lesion size (+-se) was determined 5 d postinfection. F 500-Induced TMV Resistance Is Independent of SA Accumulation | When enhancing the resistance of tobacco to TMV infection, F 500 could act independently of SA or affect steps upstream or downstream of SA accumulation. If F 500 acts before SA accumulation, it would not enhance TMV resistance in NahG transgenic tobacco plants. Due to the presence of the NahG gene from Pseudomonas putida, which encodes an SA-hydroxylase, these plants are unable to accumulate significant amounts of SA and do not express LAR or SAR in response to pathogen attack . If F 500 acts downstream or independently of the SA accumulation step, NahG plants should respond identically to nontransformed plants. To distinguish between these possibilities, halves of leaves of NahG plants were treated with 0.5 mm F 500 for 1 d before the resistance against TMV was evaluated. As shown in Figure , F 500 enhanced the resistance against TMV infection in NahG transgenic tobacco to the same degree as in nontransformed tobacco cv Xanthi nc plants. Statistical analysis (Student's t test) revealed that the difference in the size of TMV lesions between F 500-pretreated and -non-pretreated NahG tobacco leaves was significant (P < 0.01, n = 16). Thus, the result from this experiment indicates that the F 500-induced TMV resistance of tobacco is independent of SA accumulation. Figure 4 | F 500-induced TMV resistance does not depend on SA accumulation. F 500-induced TMV resistance does not depend on SA accumulation. Leaf halves of nontransformed (tobacco cv Xanthi nc) and transformed NahG (tobacco cv Xanthi nc [NahG]) plants were infiltrated with water (-; control halves) or with 0.5 mm F 500 (+). One day later, the entire leaves were infected with TMV. Lesion size (+-se) was determined 3 d postinfection. F 500 Does Not Directly Induce PR-1 Protein Accumulation | Next, we were interested in whether the F 500-induced resistance of tobacco against TMV is associated with the accumulation of known molecular markers of acquired disease resistance. For this purpose, F 500-treated tobacco leaf tissue was assayed for the accumulation of PR-1 proteins by western-blotting analysis at various time points after treatment. By doing so, we found that, although there was a prominent and time-dependent accumulation of PR-1 proteins in tobacco leaf halves that had been infiltrated with 0.5 mm SA, induction of these proteins was only very low when 0.5 mm F 500 was injected into tobacco leaf halves . Thus, the F 500-induced TMV resistance of tobacco does not depend on a significant, pre-infectional accumulation of the prominent disease resistance marker protein PR-1. Figure 5 | F 500 does not significantly induce the accumulation of PR-1 proteins in tobacco leaves. F 500 does not significantly induce the accumulation of PR-1 proteins in tobacco leaves. One-half of a leaf of a tobacco plant was infiltrated with 0.5 mm F 500, whereas the other one-half was infiltrated with 0.5 mm SA (positive control). At the indicated time points post treatment, leaf tissue was assayed for the accumulation of PR-1 proteins by western-blotting analysis. F 500 Primes Tobacco Leaves for Accelerated PR-1 Induction after TMV Attack | As a next step toward elucidating the mode of action of F 500 in the enhancement of disease resistance in tobacco, we examined whether the pretreatment with F 500 may prime the leaves to better activate cellular defense responses once attacked by TMV. To this end, halves of tobacco leaves were infiltrated with 0.5 mm F 500, whereas control halves were infiltrated with water. After 1 d, both halves of the leaves were infected with TMV. At various time points post-TMV application, leaf tissue was harvested and analyzed for the accumulation of both PR-1 mRNA and protein by RNA gel blot and western-blotting analysis, respectively. Consistent with the results in Figure , there was only faint, if any, accumulation of PR-1 mRNA and protein in leaf halves treated with F 500 only. Interestingly, however, in the F 500-pretreated and then TMV-infected leaf halves, the virus-induced accumulation of both PR-1 mRNA and protein was detectable at least 12 h earlier than in the water-pretreated and then TMV-infected control halves . Thus, the pretreatment with F 500 enables the tobacco leaves to react faster with PR-1 gene expression and PR-1 protein accumulation but only after the protected tissue becomes attacked by TMV . Figure 6 | Pretreatment with F 500 primes tobacco leaf tissue for accelerated accumulation of both PR-1 mRNA (A) and PR-1 protein (B). Pretreatment with F 500 primes tobacco leaf tissue for accelerated accumulation of both PR-1 mRNA (A) and PR-1 protein (B). One-half of a tobacco leaf was infiltrated with water (-), whereas the other one-half was infiltrated with F 500 (0.5 mm; +). After 24 h, the entire leaf was infected with TMV. At the indicated time points, leaf tissue was analyzed for the accumulation of PR-1 transcripts (A) and PR-1 protein (B) by RNA gel-blot and western-blotting analysis, respectively. DISCUSSION : In this study, we demonstrated that tobacco cv Xanthi nc plants treated with the synthetic strobilurin derivative F 500 showed enhanced resistance against infection by either TMV or virulent P. syringae pv tabaci . In contrast, the pretreatment with F 500 delayed induction of the HR and necrosis by P. syringae pv tomato DC3000 (Fig. A) and also led to an increase in bacterial population size (Fig. B), indicating that F 500 interferes with resistance responses of tobacco cv Xanthi nc to avirulent P. syringae pv tomato DC3000. The reason for the paradoxical result with the two bacterial strains remains unclear. One possible explanation for the observed increase in P. syringae pv tomato DC3000 growth in F 500-pretreated leaf panels is the fact that this tissue, in contrast to the untreated control, does not collapse but remains viable, possibly due to F 500 affecting the perception of the avirulent pathogen by the host. This assumption is supported by the highly delayed appearance of visible symptoms on F 500-pretreated and then P. syringae pv tomato DC3000-infected tobacco leaf tissue. Because P. syringae pv tomato DC3000 that has been re-isolated from F 500-pretreated and then infected tobacco leaves was fully avirulent when infiltrated into tobacco leaf tissue (data not shown), we can exclude the possibility that F 500 causes genetic changes in the bacteria that affect avirulence. The lack of correlation between the dose of F 500 applied and the observed reduction in TMV lesion size might be due to the poor solubility of F 500 in water (0.19 mg 100 mL-1 at 20C). Because of this poor solubility, increasing the content of formulated F 500 in the aqueous suspension from a certain amount does not further enhance the concentration of dissolved F 500 and, therefore, may not further reduce the size of TMV lesions . Alternatively, the induced TMV resistance might already be fully triggered at the lower F 500 doses. F 500 had no direct inhibitory effect on TMV' s infection potency and per se did not affect in vitro the multiplication of P. syringae pv tabaci (data not shown). Thus, the enhanced protection against these two pathogens (Figs. , , and ) very likely results from the activation by the F 500 fungicide of an induced disease resistance mechanism in the plant. This conclusion is consistent with an earlier report demonstrating impaired activity of the fungicides Metalaxyl, Cu(OH)2, and Fosethyl in Arabidopsis plants blocked in their disease resistance signal transduction mechanism . Furthermore, induction of SAR in Arabidopsis was reported recently as a mode of action of the plant protecting fungicide Probenazole . F 500 protected tobacco cv Xanthi nc leaf tissue from TMV (Figs. and ) and P. syringae pv tabaci attack , but did not induce TMV resistance in tobacco cv Xanthi and also failed to induce resistance against tobacco etch virus and potato virus Y in tobacco cv Xanthi nc plants (data not shown). Furthermore, F 500 obviously interferes with the resistance of tobacco cv Xanthi nc to P. syringae pv tomato DC3000. Thus, the F 500-induced resistance acts against various, yet not all, pathogens. Similar findings have been made before with induced resistance phenomena, including SAR and rhizobacteria-induced systemic resistance . F 500 was active at the enhancement of TMV resistance in NahG transgenic tobacco plants . In addition, F 500 did not induce the accumulation of SA in wild-type tobacco (data not shown). Therefore, F 500 enhances the TMV resistance of tobacco either by acting downstream of SA in the SA signal transduction network or by functioning independently of SA. The latter possibility is more likely because F 500 did not directly cause significant accumulation of the SA-responsive PR-1 proteins in infiltrated tobacco leaves . The mechanism by which F 500 accelerates the activation of PR-1 genes and enhances the pathogen resistance of tobacco remains unclear, however. In various fungi, strobilurins were reported to stimulate alternative respiration . Interestingly, Carr and coworkers found that SA-mediated enhancement of TMV resistance in tobacco is sensitive to salicylhydroxamic acid, an inhibitor of the alternative oxidase pathway. Thus, it is possible that the F 500-induced increase in TMV resistance and the associated acceleration of PR-1 gene expression in tobacco cv Xanthi nc is mediated, at least in part, by activation of the alternative respiratory pathway as a response to F 500-induced inhibition of mitochondrial respiration. However, alternative oxidation does not confer resistance against bacterial pathogens in tobacco cv Xanthi nc and, thus, the F 500-induced resistance against P. syringae pv tabaci requires another explanation. The TMV-induced activation of PR-1 defense genes and the associated accumulation of the PR-1 proteins occurred much earlier in F 500-pretreated plants than in the water-pretreated controls . Thus, F 500 may enhance disease resistance in tobacco by accelerating the plant's ability for the induction of normal defense responses that occur once the pathogen is sensed by the plant. A similar conclusion has been drawn from earlier experiments demonstrating potentiated activation of various elicitor-induced defense responses in parsley (Petroselinum crispum) culture cells primed with SA or synthetic SA analogs . Reports on stronger induction of pathogen-activated defense responses in intact tobacco and Arabidopsis plants pre-incubated with inducers of enhanced plant disease resistance also supported the conclusion that a primed state for potentiated activation of cellular defense responses might play a crucial role in acquired plant disease resistance. In this context, it is interesting that demonstrated accelerated accumulation of PR-1 transcripts in P. syringae pv tomato DC3000-infected Arabidopsis plants made resistant by pretreatment with the nonprotein amino acid beta-aminobutyric acid. Furthermore, pretreatment of cultured tomato (Lycopersicon esculentum) cells with the wound-generated systemic peptide messenger systemin was reported to enhance the hydrogen peroxide burst subsequently induced by oligogalacturonides or by osmotic stress . MATERIALS AND METHODS : Materials | Pure F 500, as well as a formulation of F 500 (BAS 500 F DI) containing 20% (w/w) active ingredient, were provided by BASF Inc. (Limburgerhof, Germany). The formulation was suspended in water as a stock suspension containing 1 mm F 500. The suspension was left on the bench for exactly 15 min to sediment insoluble materials. The F 500 content in the resulting supernatant was determined by HPLC analysis and found to be 0.5 mm, indicating saturation of the F 500 solution. The supernatant was used, partly after further dilution with water, for infiltration of the tobacco leaf apoplast as described below. Pure F 500 was suspended in 1% (v/v) DMSO at 0.5 mm (final concentration) and infiltrated into leaf tissue also as described below. SA was purchased from Sigma (St. Louis), dissolved in water as a 10 mm stock solution, and adjusted to pH 5.8 with KOH. Pseudomonas syringae pv tomato (strain DC3000) and P. syringae pv tabaci were provided by Brian Staskawicz (University of California, Berkeley) or bought at the German Collection of Microorganisms and Cell Cultures, Inc. (Braunschweig, Germany), respectively. Growth and Treatment of Plants, Determination of TMV Lesion Size, and Evaluation of Symptoms | Nontransgenic and NahG transgenic tobacco plants were grown at 23C in a 16-h-light cycle and used for experimentation at 6 to 8 weeks. One-half of a well-developed leaf of a tobacco plant was infiltrated, through small holes punctured with a needle, by a syringe with either pure F 500 in 1% (v/v) DMSO, an aqueous suspension of F 500 formulation, or with an aqueous solution of SA. The second one-half of the leaf was infiltrated with the respective control solution; that is, either 1% (v/v) DMSO or water. For TMV infection, the entire leaf was inoculated, 24 h later, with a suspension of the U1 strain of TMV (1 mug mL-1) in 50 mm sodium phosphate (pH 7.0) by gently rubbing two layers of cheesecloth soaked with TMV suspension over the carborundum-covered leaf surface. The leaf was then washed under a stream of tap water. After an additional 3 to 7 d, the area (average +- se) of lesions on the leaf halves was determined based on the fluorescence of phenolic compounds in the lesions with a FluorImager 595 (Molecular Dynamics, Inc., Sunnyvale, CA) and ImageQuant software (Molecular Dynamics, Inc.). This computer-based method for determining TMV lesion size was found to be highly reliable and less prone to subjective evaluation errors than the classical determination of lesion size with a ruler. P. syringae pv tomato (strain DC3000) and P. syringae pv tabaci were grown at 30C in King's B media for 1 d. After centrifugation, bacterial cells were washed and resuspended to 1 x 105 cfu mL-1 in 10 mm MgCl2. Bacterial suspensions were infiltrated into the apoplast of tobacco leaf tissue, through small holes punched with a needle, using a plastic syringe. Appearing symptoms were recorded with a digital camera (model C-3030, Olympus, Tokyo) for 1 week. Estimation of Bacterial Populations | Increases in bacterial populations were estimated in two leaf discs (1 cm in diameter) taken from infected leaf areas of two different plants at the indicated time points postinfection (n = 4). The discs were homogenized in 500 muL of sterile water, thoroughly mixed, and serial dilutions of the slurry were plated out on King's B agar. After incubation at 30C for 2 d, colonies were counted and the original population size deducted. Population sizes are given as cfu per leaf disc. Analysis of PR-1 Induction | To determine the accumulation of PR-1 proteins by western-blotting analysis, two leaf discs (1 cm in diameter) of respectively treated tobacco leaf halves were homogenized and fractionated by SDS-PAGE as described . The separated proteins were electrophoretically transferred to a polyvinylidene difluoride membrane. Immunoblot analysis was performed as described with mouse monoclonal antibody 33G1, which specifically recognizes PR-1 proteins . To evaluate the accumulation of PR-1 mRNA by RNA gel-blot analysis, total RNA was isolated from frozen leaf discs using TRI-Reagent (Molecular Research Center, Cincinnati) according to the manufacturer's instructions. Ten micrograms of total RNA was denatured and separated on a 1.2% (w/v) agarose-2.5% (v/v) formaldehyde gel essentially as described . After blotting to a positively charged nylon membrane (Nytran-Plus, Schleicher & Schull, Dassel, Germany) by downstream capillary transfer using 10x SSC (1.5 m sodium chloride and 0.15 m sodium citrate, pH 7.0), RNA was cross-linked to the membrane by UV irradiation. Prehybridization and hybridization were performed at 65C in 0.25 m Na2HPO4, pH 7.2; 1 mm EDTA; 7% (w/v) SDS; and 1% (w/v) bovine serum albumin. Hybridization with a 32P-labeled tobacco PR-1 cDNA probe was for 16 h. After hybridization, the membrane was washed at 65C for 1 h with two changes of the washing solution (40 mm Na2HPO4, pH 7.2; 1 mm EDTA; 5% [w/v] SDS; and 0.5% [w/v] bovine serum albumin). Finally, the blot was exposed to maximum sensitivity x-ray film (Eastman-Kodak, Rochester, NY) at -70C. All experiments shown in this study were performed at least three times with similar results. Distribution of Materials | Upon request, all novel materials described in this publication will be made available in a timely manner for noncommercial research purposes, subject to the requisite permission from any third party owners of all or parts of the material. Obtaining any permissions will be the responsibility of the requestor. Backmatter: PMID- 12226493 TI - An Endoplasmic Reticulum-Bound Ca2+/Mn2+ Pump, ECA1, Supports Plant Growth and Confers Tolerance to Mn2+ Stress AB - Plants can grow in soils containing highly variable amounts of mineral nutrients, like Ca2+ and Mn2+, though the mechanisms of adaptation are poorly understood. Here, we report the first genetic study to determine in vivo functions of a Ca2+ pump in plants. Homozygous mutants of Arabidopsis harboring a T-DNA disruption in ECA1 showed a 4-fold reduction in endoplasmic reticulum-type calcium pump activity. Surprisingly, the phenotype of mutant plants was indistinguishable from wild type when grown on standard nutrient medium containing 1.5 mm Ca2+ and 50 mum Mn2+. However, mutants grew poorly on medium with low Ca2+ (0.2 mm) or high Mn2+ (0.5 mm). On high Mn2+, the mutants failed to elongate their root hairs, suggesting impairment in tip growth processes. Expression of the wild-type gene (CAMV35S::ECA1) reversed these conditional phenotypes. The activity of ECA1 was examined by expression in a yeast (Saccharomyces cerevisiae) mutant, K616, which harbors a deletion of its endogenous calcium pumps. In vitro assays demonstrated that Ca2+, Mn2+, and Zn2+ stimulated formation of a phosphoenzyme intermediate, consistent with the translocation of these ions by the pump. ECA1 provided increased tolerance of yeast mutant to toxic levels of Mn2+ (1 mm) and Zn2+(3 mm), consistent with removal of these ions from the cytoplasm. These results show that despite the potential redundancy of multiple Ca2+ pumps and Ca2+/H+ antiporters in Arabidopsis, pumping of Ca2+ and Mn2+ by ECA1 into the endoplasmic reticulum is required to support plant growth under conditions of Ca2+ deficiency or Mn2+ toxicity. Keywords: Introduction : Plants have a remarkable ability to adapt and grow in soils containing widely different levels of mineral nutrients. To complete a plant's life cycle, macronutrients, like Ca2+ and Mg2+, ideally are required at 100- to 300-fold higher levels than that of micronutrients, such as Mn2+ and Zn2+. Growth is inhibited when tissue content of a nutrient either drops below a critical concentration (deficiency zone) or increases above an adequate level (toxic zone). Nevertheless, plants demonstrate differences in their abilities to tolerate nutrient deficiency or toxicity . For most nutrients, including Ca2+ and Mn2+, the molecular bases underlying these differences are poorly understood. Ca2+ and Mn2+ are essential for distinct functions in plants. The majority of calcium is associated with the middle lamella of cell walls, where it plays roles in support and growth. Inside cells, most Ca2+ is sequestered within organelles, like the vacuole and the endoplasmic reticulum (ER). The transient release of Ca2+ into the cytosol generates calcium waves or oscillations that carry specific information for transduction of hormonal and environmental stimuli . Calcium also mediates many developmental processes, including polarized growth, mitosis, and cytokinesis . Mn2+ is essential for photosynthesis because it is part of the oxygen-evolving complex in photosystem II. This cation is also important for several redox processes and is an activator of enzymes involved in oxidation reduction, decarboxylation, and biosynthesis . At high concentrations, both Ca2+ and Mn2+ are potentially toxic and their cytosolic levels are tightly regulated in the range of 0.1 to 0.2 mum . Both cations enter plant cells down an electrochemical gradient . Cytosolic Ca2+ is maintained at low levels by ATP-driven pumps and Ca2+/H+ antiporters located at membranes, including the plasma membrane, vacuole, and ER . Mn2+ is accumulated mostly in the vacuole and chloroplast ; however, less is known about active transporters of Mn2+. Arabidopsis contains 15 putative Ca-ATPases, as predicted from the completed genome sequence , though the in vivo function of each pump is unknown. ECA1, previously shown to be a Ca2+ pump, is specifically blocked by cyclopiazonic acid (CPA; ), an inhibitor of animal sarcoplasmic/endoplasmic reticulum Ca-ATPase pump. ECA1 is one of four members of the ER-type calcium ATPase subfamily in Arabidopsis . ACA2 localized to the ER represents another subfamily characterized as autoinhibited calcium ATPase. These Ca2+ pumps are inactivated by interactions with an inhibitory region at the N-terminal region and stimulated by calmodulin interaction with the inhibitory region . ACA4 and ACA8 are similar types of calmodulin-regulated Ca-ATPases localized at the vacuolar and at the plasma membrane , respectively. Here, we show that the ER-localized Ca2+ pump, ECA1, has a dual role in both Ca2+ and Mn2+ homeostasis. We provide biochemical evidence that ECA1 provides approximately 70% of the total ER-type calcium pump activity in Arabidopsis. Surprisingly, a plant with a T-DNA disruption (eca1-1) of this major transport system has a wild-type phenotype when grown under standard nutrient conditions. Nevertheless, under conditions of calcium deprivation, growth of mutant plants is impaired, demonstrating that ECA1 provides an important function in calcium nutrition. We further provide biochemical and genetic evidence that ECA1 behaves as an Mn2+ pump, and that it confers tolerance to toxic levels of Mn2+. These studies provide the first genetic evidence for the in vivo function of a calcium pump in plants. RESULTS : ECA1 Can Transport Divalent Cations Other Than Ca2+ | Specific ions that are transported by P-type ATPaseshave been shown to stimulate the formation of a phosphorylated intermediate as part of the reaction cycle . To test which ions are potentially transported by ECA1, microsomes isolated from pECA1-transformed K616 yeast (Saccharomyces cerevisiae) were incubated with [gamma-32P] ATP and 5 mum various cations. Very little phosphoprotein was formed in the absence of any divalent cation (+EGTA). Phosphorylation was enhanced by Ca2+, Mn2+, Zn2+, and perhaps by Ni2+, but not with Cd2+ (Fig. , A --C). The steady-state [32P] phosphoenzyme was decreased rapidly by excess unlabeled ATP, indicating the enzyme turns over rapidly. Ca2+ did not stimulate any phosphoprotein formation in control membranes isolated from vector-transformed yeast . The results are consistent with the idea that ECA1 could transport several divalent cations, including Ca2+, Mn2+, Zn2+, and Ni2+. Figure 1 | ECA1 behaves like a transporter with multi-cation specificity. ECA1 behaves like a transporter with multi-cation specificity. A through C, Ca2+, Mn2+, and Zn2+ stimulated phosphoprotein formation in membranes of ECA1 transformants. [32P]ATP (300 nm) was added to a 2-mL reaction mixture containing 25 mm HEPES-1,3-bis(tris[hydroxymethyl]methylamino) propane (BTP; pH 7.0), 100 mm KCl, and 80 mug of vesicle with 0.5 mm EGTA alone (C, and ) or in the presence of 5 mum divalent cations. The final free concentration of the cations was estimated from total cation added in the presence of 0.5 mm EGTA using the Maxchelator program. Aliquots were sampled, and unlabeled ATP was added at 120 s (arrow) to a final concentration of 1 mm. Black and white symbols represent [32P] phosphoprotein level before and after addition of ATP, respectively. The data are from the average of two experiments. A, Ca2+, circle; Mn2+, ; Cd2+, . B, Zn, circle. C, Ni2+, . D, 45Ca transport into membrane vesicles of ECA1 transformants is blocked by Mn. ATP-dependent 45Ca (approximately 0.6 mum at 0.3 muci mL-1) uptake at 5 min was measured without EGTA in the presence of Mn2+ as indicated. Pump activity shown is the difference in uptake by vesicles of mutants transformed with ECA1 and that of vector alone. Average of two experiments. To confirm that Mn2+ is recognized by the Ca2+-binding site(s) on ECA1, Mn2+ was tested for its ability to inhibit Ca2+ transport in a concentration-dependent manner. Microsomal vesicles were isolated from yeast strains expressing ECA1, and ATP-driven 45Ca2+ uptake was measured in the presence of an estimated Ca2+ concentration of approximately 0.6 mum. Increasing Mn concentration from 10 nm to 100 mum inhibited 45Ca2+ transport into vesicles isolated from yeast expressing ECA1. Mn2+ had little or no effect on Ca2+ binding/uptake in vesicles isolated from yeast harboring the empty vector (not shown). The Mn2+ concentration required to inhibit ECA1-dependent Ca2+ uptake by 50% was estimated as 0.5 mum Mn2+ (Fig. D), suggesting ECA1 had a high affinity for Mn2+ that is similar to that of Ca2+. Evidence that ECA1 can transport Mn2+ and Zn2+ in vivo was obtained by showing that ECA1 restored growth of yeast mutant K616 on media containing high Mn2+ (1 mm) or Zn2+ (3 mm; Fig. ). The K616 strain (pmr1 pmc1 cnb1) is defective in both a Golgi PMR1 and in a vacuolar PMC1 Ca2+ pump , and PMR1 is thought to remove Mn2+ from the cytosol and prevent toxicity . ECA1 can provide the same activity, although a constitutively activated AtACA2-2 cannot . Moreover, ECA1 also reversed the toxic effects of Zn2+. We had shown before that ECA1 is a Ca2+ pump that may also transport Mn2+ . Together, these results indicate that ECA1 behaves as a multi-cation ATPase that transports Ca2+ as well as Mn2+ and Zn2+. Figure 2 | ECA1, but not ACA2, expression restored growth of K616 mutant on Mn2+-supplemented medium. ECA1, but not ACA2, expression restored growth of K616 mutant on Mn2+-supplemented medium. Wild-type (W303) or K616 (pmr1 pmc1 cnb1) yeast cells were transformed with control vector (p426 Gal1) alone. K616 cells were transformed with pECA1 or pACA2-2. Each transformant was diluted with complete synthetic medium (SC)-uracil (URA)/Gal medium to a density at A600 of 1.0, 0.1, 0.01, or 0.001. Then, 10 muL of each dilution was dotted on SC-URA/Gal media (control), or with 1 mm MnCl2 or 3 mm ZnCl2, and incubated for 3 d at 30C. Identification of a Mutant Plant Carrying a T-DNA Disruption in the ECA1 Gene | To investigate the in vivo functions of ECA1, we obtained a mutant plant, eca1-1. A T-DNA insertion associated with the ECA1 gene (eca1-1) was previously detected using a PCR screening strategy applied to a population of T-DNA-transformed Arabidopsis plants . To isolate the individual mutant plant and to determine the exact location of the T-DNA insert, both right and left ECA1/T-DNA borders were PCR amplified and sequenced. Both borders contained sequence from the coding region of ECA1, indicating that the T-DNA insertion was not associated with a major deletion or rearrangement of flanking sequences. Sequence analysis indicated that the insertion went into the middle of the last transmembrane domain (TM10; Fig. A). Figure 3 | Position of the T-DNA insert in eca1-1 mutant. Position of the T-DNA insert in eca1-1 mutant. A, T-DNA inserts in the 10th transmembrane (TM) domain of ECA1 protein. Position 1,025 marks the last amino acid of ECA1 in eca1-1 mutant corresponding to wild-type ECA1. B, PCR analysis showing the absence of a wild-type ECA1 gene in a homozygous eca1-1 mutant plant. Genomic DNA samples from a wild-type plant (Wt) or eca1-1 homozygous plant (Mutant) were used as templates in a PCR reaction using a primer pair that can amplify the entire ECA1 coding sequence (5' + 3'), or a primer pair that amplifies the sequence between the T-DNA left border and the 5' end of ECA1 (5' + TL). A picture of a UV-illuminated ethidium bromide-stained gel is shown. A plant line homozygous for eca1-1 was identified by PCR analysis after two backcrosses to the parental Wassilewskija (WS)-F ecotype. This backcrossing removed a second unlinked T-DNA insert. A single progeny was identified that had no wild-type ECA1. To test for the wild-type gene, a genomic DNA sample was subjected to a PCR amplification reaction using two primers (5' + 3' primer pair) corresponding to the 5' and 3' ends of ECA1. In controls using a wild-type DNA template, the 5' + 3' primer pair amplified an approximately 4.3-kb fragment (Fig. B). However, with a DNA template isolated from a homozygous eca1-1 mutant line, no PCR product was observed, presumably due to the presence of the very large T-DNA insert that prevented the formation of a detectable product. The absence of a wild-type ECA1 PCR product indicated that the T-DNA insertional mutant was homozygous. As a control to show that the DNA sample was a suitable PCR template, a parallel reaction was conducted in which the 3' primer was replaced with TL, a primer based on the left border of the T-DNA insert (i.e. 5' + TL primer pair). In this control, a PCR product was produced showing the expected size (4.3 kb) for a T-DNA insertion near the 3' end of the gene. Growth of the eca1-1 Mutant Is Normal under Standard Conditions But Retarded by Low Ca2+ | Compared with wild-type plants, there is no obvious phenotype in eca1-1 homozygous mutants grown on normal 0.5x Murashige and Skoog medium when Ca2+ and Mn2+ are 1.5 mm and 50 mum, respectively (Fig. A, -Mn). Plants also grew similarly on Gamborg's B5 medium containing 0.5% (w/v) MES at pH 5.7, and in various soils under growth room and greenhouse conditions (not shown). eca1-1 mutant plants appeared to display normal germination, growth rates, morphology, seed set, and gravitropism relative to wild-type plants grown in parallel. They were similar in their fresh weight and their chlorophyll content (Fig. , B and D). Figure 4 | Growth of the eca1-1 mutant is inhibited by high Mn2+ and is rescued by an ECA1 transgene. Growth of the eca1-1 mutant is inhibited by high Mn2+ and is rescued by an ECA1 transgene. Seeds of wild-type plant (ECA1), mutant (eca1-1), and mutants expressing wild-type ECA1 (35S-#6 and 35S-#7) were germinated on a 0.8% (w/v) agarose plate with 0.5x Murashige and Skoog. Five-day-old seedlings were transferred to 0.5x Murashige and Skoog medium alone (control, -Mn) or supplemented with 0.5 mm Mn2+ (+Mn). A, Plant morphology and size 12 d after transfer. B, Fresh weight of 25 plants 10 d after transfer (+-se, n = 4). C, Root hairs of plant 12 d after transfer. D, Chlorophyll content from seedlings 10 d after transfer. Average (+-se) from four extractions. Bar = 1 mm. To test the effect on growth of reduced Ca2+, seeds were germinated on standard 0.5x Murashige and Skoog and 5-d-old seedlings were transferred to a modified Murashige and Skoog medium with reduced Ca2+. After 10 d, mutants and wild-type plants grown in medium containing >=0.5 mm Ca2+ were similar in size although growth in all cases was severely impaired when Ca2+ dropped to 0.1 mm or less (not shown). At 0.2 to 0.4 mm Ca2+, the mutant consistently showed Ca2+ deficiency symptoms not exhibited by wild-type plants. These included small plant size, short roots, small yellowish leaves, and lack of bolts . The differential traits of mutant and control plants on medium with 0.2 mm Ca2+ were consistently observed in three independent experiments. Figure 5 | eca1-1 (-/-) (homozygous) mutant grew poorly on low Ca2+. eca1-1 (-/-) (homozygous) mutant grew poorly on low Ca2+. Seeds were germinated on a 0.8% (w/v) agarose plate with 0.5x Murashige and Skoog salts at pH 5.8. Seedlings (5 d old) of similar size were transferred to medium containing reduced Ca2+ (0.2 mm) and grown for 10 d at 21C. Plants are representative of mutant (eca1-1) and wild type (ECA1) from three independent experiments. Growth of eca1-1 Mutant Is Inhibited by High Mn2+ | When grown on high (0.5 mm) Mn2+, both mutant and wild-type plants had elevated Mn2+ content . However, only mutant plants were severely stunted. This phenotype is easily detected after 4 to 5 d, and becomes dramatic after 10 to 12 d (Fig. A). The fresh weight of eca1-1 mutant after 12 d was reduced by 66% relative to wild-type plants (Fig. B). Both root elongation and leaf expansion were inhibited. Two weeks later, the mutants appeared to stop growing; however, wild-type plants were pale green and continued to increase in size. At 1 mm Mn2+ or higher, both mutant and wild-type plants showed Mn2+ toxicity symptoms, although chlorosis appeared earlier (in 2 --3 d) in the mutants. Table I | Ion content of wild-type and eca1-1 mutant plants Leaves of mutants growing at 0.5 mm Mn2+ were chlorotic and twisted, and young leaves were deformed and stuck together. Wild-type plants growing in media containing 0.5 mm Mn2+ showed a 26% drop in chlorophyll content relative to plants cultivated in control medium (Fig. D). However, the chlorophyll content of eca1-1 mutant grown in 0.5 mm Mn2+ decreased by 74% relative to plants in normal 0.5x Murashige and Skoog. Inhibition of Root Hair Elongation in the eca1-1 Mutant | The morphological change in root growth was striking. Figure C shows that eca1-1 mutant roots grown on normal 0.5x Murashige and Skoog medium had root hairs similar in length to that of wild-type plants. However, the eca1-1 mutant, growing on medium supplemented with 0.5 mm Mn2+, had only stubs or very short root hairs. These results indicated that the ability to initiate root hairs was not blocked; however, tip growth and root hair elongation were inhibited. Reduction of ECA1 Protein and Ca2+ Pumping in the eca1-1 Mutant | Immunostaining showed that ECA1 protein is relatively abundant in the root and the flower, although it is expressed in all major organs of the wild-type plant (Fig. A). The expression of a mutant protein in a homozygous eca1-1 plant line was analyzed using three antibodies raised against the C- and N-terminal ends and middle hydrophilic portion of wild-type ECA1, respectively. All antisera detect a 5- to 10-fold reduced signal in a membrane extract from the homozygous mutant, so only results using anti-ECA1(M) are shown (Fig. B). Because the mutant eca1-1p is expected to have a truncated C-terminal end [hence no anti-ECA1(C)-detectable epitope], the residual signal revealed by anti-ECA1(C) most likely corresponds to a cross-reaction with other related pumps. Anti-ECA1(C) very likely recognizes AtECA4 (Arabidopsis Genome Initiative [AGI] no. At1g07670), which shares 99% protein identity with ECA1 (AGI no. At1g07810), and has only three differences in the C-terminal 27 amino acids used as an antigen to make the anti-ECA1(C) . Figure 6 | Reduction of ECA1 protein and Ca2+ pumping in the eca1-1 (-/-) mutant. Reduction of ECA1 protein and Ca2+ pumping in the eca1-1 (-/-) mutant. A, ECA1-like protein is found in all organs of wild-type plants. Total protein extracts (10-mug samples) from roots (R), leaves (L), flowers (F), and siliques (S) were subjected to SDS-PAGE (8% [w/v] gel), transferred to nitrocellulose, and probed with pre-immune or anti-ECA1(C) serum (1:1,000 [v/v]). Secondary antibody (1:5,000 [v/v]) from donkey anti-rabbit IgG was conjugated with horseradish peroxidase, and activity was detected using enhanced chemiluminescence (Amersham-Pharmacia Biotech, Uppsala). The arrow marks the expected position of ECA1 (116 kD). B, ECA1 protein content in mutant and transgenic plants. Microsome (25 mug of protein) from 1-week-old seedlings of wild-type (ECA1), mutant (eca1-1), and transgenic mutants expressing ECA1 (35S#6 and 35S#7) were separated by SDS-PAGE (10% [w/v] acrylamide), transferred, and probed with rabbit anti-ECA1 (M; 1:1,000 [v/v]). Secondary antibodies (1:5,000 [v/v]) were linked to alkaline phosphatase. Arrow marks expected size (116 kD) of ECA1. C, Ca2+ transport. Assay mixture consisted of 1 mm ATP and 10 mum Ca in the absence of EGTA, and membrane from 1-week-old seedlings of wild type (ECA1) or mutant (eca1-1). Pump activity shown is the difference between uptake at 30 min with and without Mg2+. When added, CPA was 100 nmol mg-1 protein. DeltaCPA, Pump activity sensitive to CPA. Total, Activity without CPA. Average of two experiments. We isolated vesicles from 5-d-old seedlings grown on 0.5x Murashige and Skoog to test Ca2+ pump activity. Total pump activity in the mutant was decreased by 22% relative to that in wild type. However, CPA-sensitive activity was reduced by more than 70% of that in wild type (Fig. C), indicating that activity from ECA1-like pumps was specifically impaired. CPA is a specific and potent blocker of ECA1 activity , but not of ACA2 activity . Similar results were obtained when Ca2+ content inside vesicles was enhanced by oxalate. Thus, a decrease in divalent cation pumping into endolumenal compartments leads to reduced growth and to chlorosis of plants under various nutrient stress conditions. eca1-1 Is Complemented by the Wild-Type ECA1 Transgene | We tested whether these phenotypic changes were caused by loss of ECA1 function alone. The full length ECA1 cDNA under the control of the cauliflower mosaic virus 35S promoter was introduced into the eca1-1 mutant. Five independent transgenic lines expressing 35S::ECA1 were chosen for complementation analyses. All five complementation lines showed wild-type phenotype when grown on high (0.5 mm) Mn2+-supplemented medium, and two examples are shown in Figure A (right). The fresh weight of lines 35S::ECA1-#6 and 35S::ECA1-#7 was 93% and 126% of that of wild type (Fig. B). Chlorophyll content of two transgenic lines, 35S::ECA1-#6 and 35S::ECA1-#7, was 83% and 131% of that in wild type (Fig. D). Furthermore, complementation lines showed abundant and elongated root hairs similar to that observed in wild type (Fig. C). The restoration of a wild-type phenotype was accompanied by an increase in ECA1 protein (Fig. B). Despite the CAMV35S promoter, the ECA1 protein level expressed in transgenic plants was comparable with that in wild-type plants, suggesting potential regulation of translation or protein degradation. Thus, the expression of the wild-type ECA1 gene from a 35S promoter was sufficient to rescue the mutant. DISCUSSION : The mechanisms by which plants regulate the homeostasis of most essential mineral nutrients are not understood. Insufficient levels of a nutrient can lead to deficiency symptoms, whereas excessive levels are toxic . Here, we present: (a) biochemical evidence that ECA1 can transport Mn2+ as well as Ca2+, and (b) genetic evidence that this ER-located pump plays critical roles in growth as well as tolerance to Mn2+ toxicity. T-DNA Insertion of the ECA1 Gene Reduced Ca2+ Pumping in the Mutant | To understand in vivo functions of ECA1 pump, we isolated a homozygous plant line harboring a T-DNA disruption of ECA1 (ECA1::T-DNA-1 = eca1-1). The T-DNA insertion site is predicted to disrupt the last (10th) putative transmembrane domain (see Fig. ). Although a mutant eca1-1 protein may be synthesized, immunoblot analyses indicates that the expression of ECA1-like protein is reduced 5- to 10-fold (Fig. B). It is likely that membrane proteins that fail to fold into a proper conformation are selectively removed from the ER and degraded . Alternatively, the T-DNA insertion may disrupt a transcriptional control or create an unstable mRNA. The residual protein detected by three different antibodies is likely due to the presence of another closely related isoform that shares 97% identity to AtECA4 (AGI no. At1g07670). Regardless of the mechanism, evidence indicates that the T-DNA insertion dramatically reduces both the accumulation of the ECA1-like protein and Ca2+ pump activity into endomembranes . ECA1 Maintains Ion Homeostasis by Distributing Internal Ca2+ and Mn2+ | When grown in the same nutrient medium, the total ion content of wild-type or mutant plants were similar , indicating that net ion uptake was little or not altered. It is interesting that eca1-1 mutant grown in 0.5x Murashige and Skoog medium showed little or no obvious physical differences from wild-type plants. Apparently, other intracellular cation pumps and cation/H+ antiporters could compensate in part for the loss of ECA1 activity when extracellular Ca2+ or Mn2+ was 1.5 mm or 50 mum, respectively. Under these conditions, it is conceivable that cells use a calmodulin-stimulated ACA2 to load Ca2+ into the ER lumen , and CAX2 to remove cytosolic Mn2+ into small vacuoles . Mn2+ Toxicity and Tolerance | However, physical symptoms of Ca2+ deficiency became evident in mutants when external Ca2+ was dropped to 0.2 to 0.4 mm. Furthermore, a 10-fold elevation in Mn2+ accentuated symptoms typical of Ca2+, Mg2+, and Fe2+ deficiencies in the mutant, but less so in the wild-type plants. We have initially focused on the increased sensitivity to Mn2+ toxicity because symptoms of Ca2+ deficiency were less dramatic, and symptoms of Zn2+ deficiency or toxicity were not observed in preliminary studies. When plants were exposed to high levels of Mn2+, both wild-type and mutant plants accumulated Mn2+ to similar, potentially toxic, levels , consistent with the idea that the eca1-1 mutation did not alter uptake of Mn2+ into the plant. Thus, the difference in Mn2+ tolerance between mutant and wild-type plants may be caused by differential internal distribution of ions within tissues or cells. This idea was suggested for two cotton (Gossypium hirsutum) genotypes that differed in their ability to tolerate elevated Mn2+ . Although measurements of ion content in the cytosol and intracellular compartments of eca1-1 mutants are not yet available, we offer several interpretations for observed Mn sensitivity of eca1-1. The average Mn2+ content in plant tissues is low (20 --50 mug g dry weight-1). Mn2+ is accumulated in the vacuole and chloroplast in leaves , whereas cytosolic Mn2+ is estimated as less than 0.2 mum in roots . High external Mn2+ can compete for divalent cation-binding sites, thus reducing uptake of Ca2+, Mg2+, and Fe2+ . In theory, this competition could also occur at sites of intracellular transport. Thus, high cytosolic Mn2+ potentially interferes with proper uptake and sorting of Ca2+, Mg2+, and Fe2+ into intracellular locations. Inadequate Mg2+ and Fe2+ in the chloroplasts could result in chlorosis because they are required for chlorophyll synthesis and as cofactors for many metalloenzymes . Perturbations in cytosolic Ca2+ signaling could also result in poor growth. Interference of Ca2+ sequestration into organelles could potentially change the shape and frequency of Ca2+ oscillations that are so critical to specific signaling events . We propose that ECA1 could promote growth and confer Mn2+ tolerance in several ways. By pumping Mn2+ into the ER, ECA1 could reduce cytosolic Mn2+ to levels that do not interfere with the internal distribution of Mg2+, Fe2+, or Ca2+. This would reduce the induced deficiency of these cations, and also restore, for instance, signal-induced Ca2+ transients normally seen in wild-type plants. Under conditions of Ca2+ deficiency when cytosolic Ca2+ level would be very low, the high-affinity ECA1 may be the only pump capable of loading Ca2+ into the ER lumen for functioning of the secretory system and for stimuli-induced Ca release . Roles of ECA1 | The observation of an eca1-1 phenotype demonstrates that ECA1 has unique functions, despite multiple calcium pumps and cation exchangers in plants. Biochemical studies demonstrate that ECA1 differs from other ACA2-like Ca2+ pumps in its (a) ion specificity, (b) high affinity for cations, and (c) subcellular location . Although ACA2 is present in the ER of some cell types , it does not transport Mn2+ and therefore is unable to replace ECA1. Moreover, ECA1 protein may be more abundant than ECA2-ECA4 because it contributes 70% of Ca2+ pump activity inhibitable by CPA. As a high-affinity divalent cation pump with a KmCa estimated at 0.03 mum , ECA1 would be more effective than other Ca2+ pumps or Ca2+/H+ antiporters when cytosolic [Ca2+] is extremely low. The inhibition of growth by high Mn2+ in eca1-1 Arabidopsis plants and in the yeast pmr1 mutant are remarkably similar, suggesting that disruptions in intracellular Ca2+ and Mn2+ homeostasis affect fundamental processes of cell division and cell expansion. For example, Mn2+ may be important for cell cycle progression in place of Ca2+ as shown in yeast , and for activation of Mn-dependent glycosyltransferases involved in protein processing and cell wall synthesis . Yeast pmr1 mutants lacking a Golgi Ca2+/Mn2+ pump showed defects in glycosylation and protein sorting , and Drosophila melanogaster cells lacking a sarcoplasmic/endoplasmic reticulum-type Ca-ATPase is defective in Notch trafficking . Furthermore, elongating root hairs and pollen tubes depend on a tip-focused Ca2+ gradient that is most likely dependent on spatially localized Ca2+ channels and active transporters. Perturbations of the Ca2+ gradient alter root hair growth . Although the mechanism of this is still unknown, Ca2+ pumps associated with the cortical ER potentially alter the shape or frequency of Ca2+ oscillations that accompany tip growth . This study of the first plant mutant with a disruption of an ER-type Ca2+ pump has provided two new insights. First, despite a 70% reduction in ER-type calcium pump activity, a plant can complete its life cycle under conditions of "optimal" nutrient availability. Second, despite the presence of 14 Ca2+ pumps and 11 CAX1-like antiporters in Arabidopsis, the disruption of ECA1 revealed a critical role of this single pump in plants grown under conditions of low Ca2+ or high Mn2+. Thus, ECA1 clearly functions in mineral nutrition and makes plants more adaptable to soils with variable nutrient conditions. Whether ECA1 also plays a specific role in calcium signaling will require further studies. MATERIALS AND METHODS : Yeast (Saccharomyces cerevisiae) Strain, Plasmid, and Growth Medium | Yeast strains W303-1A (MATa, leu2, his3, ade2, trp1, and ura3) and K616 (MATa pmr1::HIS3 pmc1::TRP1 cnb1::LEU2 and ura3) were used . The entire ECA1 cDNA and the truncated ACA2-2 cDNA from Arabidopsis were constructed into the yeast expression vector p426Gal1 . The K616 mutant was transformed, selected on SC-URA, and grown on medium with 2% (w/v) Gal as described . PCR Analysis of T-DNA Tags in Arabidopsis | Genomic DNA was isolated from individual homozygous eca1-1 mutant and wild-type Arabidopsis WS-F plants according to . The PCR primers sequences are as follows: T-DNA left border (TL), dGATGCACTCGAAATCAGCCAATTTTAGAC; ECA1 primer (5'), dGAGTTTCCGGGAGAATTTGACGAATCTGT; and ECA1 primer (3'), dCCAACGCCGAGGTAAGTAACAACGCTAAT. Genomic DNA samples (40 ng) from a wild-type plant or eca1-1 homozygous plant were used as templates in a PCR reaction using a primer pair that can amplify the entire ECA1 coding sequence (5' + 3'), or a primer pair that amplifies the sequence between the T-DNA left border and the 5' end of ECA1 (5' + TL). PCR reactions were performed using X-Taq and buffers (PanVera Corporation, Madison, WI) for 41 cycles as follows: 94C for 30 s, 65C for 1 min, and 72C for 2 min. Plant Transformation | eca1-1 (-/-) (homozygous) mutant was transformed with the CAMV35S-ECA1 construct via vacuum infiltration . Kanamycin-resistant plants were selected and five independent lines expressing the ECA1 transgene were used for complementation analyses. Growth Conditions of Arabidopsis | The homozygous eca1-1 mutant and transformants were grown side by side with parental wild-type WS-F plants under the same conditions. Seeds were surface sterilized, vernalized at 4C for 48 h, and germinated vertically on 0.8% (w/v) agarose plates with 0.5x Murashige and Skoog medium at pH 5.7 under constant illumination of approximately 100 muE m-2 s-1 at 21C. To test effects of Ca2+ or Mn2+ on growth, 5-d-old seedlings were transferred to agarose plates containing 0.5x Murashige and Skoog (control), or a modified Murashige and Skoog with 0.1 to 1 mm Ca2+ or supplemented with 0.5 to 2.0 mm Mn2+ in the absence of EGTA. Ion Content | Mutant and wild-type seeds were supported on filter paper dipped in 0.5x Murashige and Skoog and germinated vertically in Magenta boxes for 5 d. Seedlings were transferred to either control (0.5x Murashige and Skoog alone) or medium supplemented with 0.5 mm Mn2+ for 2 weeks. Plants (15 --20 g fresh weight) were dried at 70C for 3 to 4 d, and ashed at 480C for 16 h. Ash was digested with 2 mL of concentrated HNO3, and the dry sample was dissolved in 10 mL of 3 n HCl. Ion content was analyzed with a Liberty 150 ICP Emission Spectrometer (Varian, Palo Alto, CA) using cobalt as an internal standard. Chlorophyll Content | Seedlings (approximately 0.5 g) were pulverized in liquid nitrogen and extracted in 2 mL and then diluted with 80% (v/v) ice-cold acetone to 3 mL. The supernatant, collected after centrifuging at 5,000g for 15 min, was examined at 663 and 645 nm, and the chlorophyll concentration was determined using the equation: Isolation of Membrane Vesicles from Yeast and Plants | One-week-old seedlings (8 g) were homogenized in 10 mL of buffer containing 50 mm HEPES-BTP (pH 7.4), 250 mm sorbitol, 6 mm EGTA, 0.5% (w/v) bovine serum albumin, 1 mm dithiothreitol, 1 mm phenylmethylsulfonyl fluoride, 0.1 mm tosylsulfonyl phenylalanyl chloromethyl ketone, 2 mm benzamidine, 1 mug mL-1 leupeptin, and 5 mug mL-1 pepstatin A. The homogenate was centrifuged at 10,000g for 15 min, and the supernatant was either centrifuged at 110,000g for 50 min to get a microsomal pellet for SDS-PAGE, or layered on two 16-mL tubes with 15%/35% (w/w; 5 mL each) Suc step gradient (in buffer: 10 mm HEPES-BTP, pH 7.2; 50 mm KCl; 1 mm dithiothreitol; 1 mm phenylmethylsulfonyl fluoride; 0.1 mm tosylsulfonyl phenylalanyl chloromethyl ketone; and 2 mm benzamidine), and centrifuged at 100,000g for 3 h. Vesicles at the 15%/35% (w/w) Suc interface were collected (3 --4 mL) for calcium transport. Yeast vesicles were isolated as described . Protein concentration was determined with the Bio-Rad reagent (Bio-Rad Laboratories, Hercules, CA). Calcium Transport | Ca2+ uptake into membrane vesicles was measured without EGTA by the filtration method . Transport with plant vesicle was initiated with 1 mm ATP in a reaction mixture (250 muL) containing 250 mm Suc, 25 mm HEPES/BTP (pH 7.0), 10 mm KCl, 0.4 mm NaN3, and 10 mum45CaCl2 (2 muci mL-1, PerkinElmer Life Sciences, Boston) with or without 1 mm MgSO4. To block H+/Ca2+ antiport, mixture included 0.5 mum bafilomycin A1 and 5 mum gramicidin D. CPA was incubated with membranes at 20C for 15 min before the reaction was started. Transport of yeast vesicles was similar except that ATP and MgSO4 is 3 mm, 45CaCl2 is 0.3 muCi mL-1 (estimated to total calcium 0.6 mum), and the incubation time is 5 min at 20C. Antibody Production | Three separate rabbit polyclonal anti-ECA1 were produced. Anti-ECA1(C) number 1,374 was raised against a fusion protein, encoded by the plasmid pECA1-c, containing the last 27 residues of ECA1p fused to the C-terminal end of glutathione S-transferase (GST). Anti-ECA1(N) number 1,705 was also raised against a GST fusion protein, encoded by the plasmid pIN-ECA1-N, containing the first 85 residues of ECA1p. Both GST fusion proteins were constructed using the parent vector pGEX-2T . The fusion proteins were expressed in Escherichia coli DH10alpha, and affinity purified over a glutathione-agarose column (Pharmacia Biotech, Piscataway, NJ). Anti-ECA1(M) number UMY31 was raised against a C-6x His tag fusion protein containing residues Met-393-Gln-605 of ECA1. The fusion protein was constructed using the parent vector pET28b at the NcoI/SalI insertion site. The fusion protein was expressed in BL21 cells, and first purified through Probond nickel-chelating resin according to the manufacturer's protocol (Invitrogen Company Xpress System, Invitrogen, Carlsbad, CA) and purified by SDS-PAGE (approximately 25 kD). The purified fusion protein samples were injected into New Zealand white rabbits with RIBI adjuvant as recommended by the manufacturer (RIBI ImmunoChem Research). Serum of final bleeds recognized a 116-kD protein in immunoblots of microsomal proteins from ECA1 transformants. Distribution of Materials | Upon request, all novel materials described in this publication will be made available in a timely manner for noncommercial research purposes, subject to the requisite permission from any third party owners of all or parts of the material. Obtaining any permission will be the responsibility of the requestor. Backmatter: PMID- 12226494 TI - Movement of Potato Spindle Tuber Viroid Reveals Regulatory Points of Phloem-Mediated RNA Traffic AB - Increasing evidence indicates that the phloem mediates traffic of selective RNAs within a plant. How an RNA enters, moves in, and exits the phloem is poorly understood. Potato spindle tuber viroid (PSTVd) is a pathogenic RNA that does not encode proteins and is not encapsidated, and yet it replicates autonomously and traffics systemically within an infected plant. The viroid RNA genome must interact directly with cellular factors to accomplish these functions and is, therefore, an excellent probe to study mechanisms that regulate RNA traffic. Our analyses of PSTVd traffic in Nicotiana benthamiana yielded evidence that PSTVd movement within sieve tubes does not simply follow mass flow from source to sink organs. Rather, this RNA is transported into selective sink organs. Furthermore, two PSTVd mutants can enter the phloem to spread systemically but cannot exit the phloem in systemic leaves of tobacco (Nicotiana tabacum). A viroid most likely has evolved structural motifs that mimic endogenous plant RNA motifs so that they are recognized by cellular factors for traffic. Thus, analysis of PSTVd traffic functions may provide insights about endogenous mechanisms that control phloem entry, transport, and exit of RNAs. Keywords: Introduction : RNA and protein traffic is a biological function basic to all multicellular organisms. Increasing evidence indicates that selective plant RNAs traffic from cell to cell through plasmodesmata and even from organ to organ through the phloem. Examples include Suc transporter 1 (SUT1) mRNA , over 100 mRNAs from phloem exudates of pumpkin , and a mutant fusion transcript between LeT6, a tomato (Lycopersicon esculentum) homeodomain protein gene, and PYROPHOSPHATE-DEPENDENT PHOSPHOFRUCTOKINASE gene . Gene silencing signals, believed to consist of RNAs, also traffic systemically . These findings raise the prospect that systemic RNA traffic controls various plant developmental and physiological processes, in addition to surveillance and elimination of viral pathogens . The diversity of RNAs transported in the phloem poses questions about how the phloem delivers all cargos to the proper locations. Most important of these are whether a transport cargo has a distinct motif for traffic, whether entering and exiting the phloem involve the same or unique RNA motifs, and what cellular factors recognize and transport an RNA to its final destination . Viroid infection provides a unique experimental system to study phloem-mediated RNA traffic in plants. Viroids are single-stranded, covalently closed circular, and pathogenic RNAs that infect plants . Although they do not encode proteins, viroids can replicate autonomously and traffic systemically throughout their host plants. A viroid genome must apparently interact directly with host components for traffic. We have been using potato spindle tuber viroid (PSTVd) as a model system to study RNA traffic in plants. This viroid replicates in the nucleus. Its systemic movement, therefore, includes nuclear transport , cell-to-cell transport , and phloem transport . In this study, we used several approaches to obtain evidence that (a) a phloem-based mechanism transports PSTVd to sepals but not to the other floral organs of an infected plant, (b) PSTVd replication and phloem exit are mediated by different viroid motifs, and (c) phloem exit of PSTVd is potentiated by specific interactions with host factors. We present our results and discuss their biological implications. RESULTS : PSTVd Trafficked into Selective Floral Organs | Our previous work showed that PSTVd intermediate (PSTVdInt) strain was present in the sepals but not in the petals, stamens, and ovary of developing flowers of mechanically inoculated tomato and Nicotiana benthamiana . We analyzed further the PSTVd infection pattern in mature flowers of mechanically inoculated N. benthamiana by in situ hybridization. As shown in Figure A, the viroid was detected in sepals but not in the other floral organs. Thus, PSTVd consistently infected sepals, but not petals, stamens, and ovary of flowers at different stages. Considering that petals, stamens, and ovary are complete sink organs with functional phloem connections to the rest of the plant body and that CmNACP, CmGAIP, and CmPP16 mRNAs and the green fluorescent protein (GFP; ) can traffic into all floral organs, two possibilities could account for the presence of PSTVd in sepals and its absence in the other floral organs. First, PSTVd was transported into all floral organs, but its replication in the ovary, petals, and stamens was inhibited. Second, PSTVd was transported into sepals but not into the other floral organs. Figure 1 | Selective traffic of PSTVdInt into sepals but not into other floral organs of N. benthamiana Selective traffic of PSTVdInt into sepals but not into other floral organs of N. benthamiana, as detected by in situ hybridization on paraffin sections. A, Transverse view of a mature flower from mechanically inoculated N. benthamiana. The viroid signal (arrows) is detected in sepals (Se) but not in petals (Pe), stamens (St), or the pistil (Pi). B, Longitudinal view of a mature flower from a 35S:PSTVdInt-transgenic N. benthamiana plant. Viroid signal is detected in all floral organs (arrows). Ov, Ovary. C, Absence of hybridization signal in leaf cells of transgenic N. benthamiana expressing the central conserved region of PSTVd under the control of the CaMV 35S promoter. MS, Mesophyll; Ph, phloem. All bars = 40 mum. To distinguish between these possibilities, we tested the ability of PSTVd to replicate in various floral organs of transgenic N. benthamiana that expresses the cDNA of PSTVdInt under the control of the cauliflower mosaic virus (CaMV) 35S promoter . We carried out in situ hybridization using a digoxigenin (DIG)-labeled RNA probe specific for the (-)-strand PSTVd, which is produced only during PSTVd replication . Examination of flowers at various developmental stages from the transgenic plants revealed presence of the (-)-strand PSTVd in sepals, petals, stamens, and ovary (Fig. B). Furthermore, sequence analysis of PSTVd progeny isolated from the transgenic plants revealed wild-type (intermediate) sequence. We emphasize that the hybridization signal in the transgenic plants must be from PSTVd that accumulated as a result of autonomous viroid replication by using the CaMV 35S promoter-generated primary transcripts as the initial templates and not because of aberrant expression of the inserted PSTVd cDNA. The CaMV 35S promoter-generated primary transcripts are not detectable by RNA-RNA hybridization , presumably because the viroid RNA-RNA replication leads to methylation of the viroid cDNA and, therefore, inactivation of further transcription . As shown in Figure C, no hybridization signal was detected from transgenic N. benthamiana expressing only the central conserved region . Our data indicate that PSTVd can replicate in all floral organs in N. benthamiana. Therefore, absence of PSTVd signals in petals, ovary, and stamens in mechanically inoculated plants can be best interpreted as being attributable to restricted traffic of PSTVd into these organs and not to suppression of replication. Two PSTVd Mutants Replicated in But Did Not Exit the Phloem in Tobacco (Nicotiana tabacum) | After mechanical inoculation, PSTVdInt strain infects N. benthamiana and tomato systemically , but it hardly infects tobacco (N. tabacum). The lack of infection in tobacco could be attributable to inability of the viroid to replicate in tobacco cells and/or to move systemically. Tobacco infection would, therefore, provide a complementary experimental system to gain further insights about viroid-host interactions for systemic traffic. We first tested whether PSTVd would replicate in the phloem of tobacco and then exit the phloem to invade nonvascular tissues. To overcome the technical barrier of mechanically inoculating the phloem, we generated transgenic plants expressing the cDNA of PSTVdInt under the control of the companion cell-specific commelina yellow mottle virus (CoYMV) promoter (; Fig. A). Previous studies on transgenic tobacco and N. benthamiana with CaMV 35S promoter-driven expression of PSTVd cDNAs have established the validity of this inoculation approach. As discussed above, the primary transcripts of PSTVd derived from the promoter activity would serve as the templates to initiate autonomous RNA-RNA replication of the viroid . Figure 2 | Restricted traffic of replicating PSTVd out of the phloem in CoYMV:PSTVd-transgenic tobacco but not in transgenic N. benthamiana Restricted traffic of replicating PSTVd out of the phloem in CoYMV:PSTVd-transgenic tobacco but not in transgenic N. benthamiana. PSTVd is detected in the nuclei (purple dots in B --I) by in situ hybridization; H is from a cryosection, and all others are from paraffin sections. A, Companion cell (CC)-specific activity of the CoYMV promoter in transgenic tobacco, as revealed by beta-glucuronidase (GUS) reporter expression . The GUS activity is absent from all other cells including sieve element (SE) and phloem parenchyma (PP). B through E, CoYMV:PSTVdInt-transgenic tobacco. Paradermal (B and C) and transverse (D and E) leaf sections show that PSTVd mutants replicate and remain in the phloem (Ph). The phloem tissue on the left of D is obliquely sectioned. E, A high magnification view of the right portion of D. MS, Mesophyll; Xy, xylem. F and G, CoYMV:PSTVdInt-transgenic N. benthamiana. Paradermal (F) and transverse (G) leaf sections show that PSTVdInt accumulates in phloem (Ph), mesophyll (MS), and epidermis (EP). H, Presence of PSTVdInt U257 -> A in the phloem (Ph) of a systemic leaf of mechanically inoculated tobacco. MS, Mesophyll. I, Presence of PSTVdInt U257 -> A in the phloem (Ph) and mesophyll (MS) of a systemic leaf of mechanically inoculated N. benthamiana. Bar in A = 10 mum. Bars in B through I = 40 mum. We obtained two CoYMV:PSTVdInt-transgenic tobacco lines (4 and 8) that showed viroid accumulation in leaves based on dot-blot analysis (data not shown). We then performed in situ hybridization to determine the cellular localization of the viroid, using DIG-labeled RNA probes specific for the (-)-strand of PSTVd. We have shown previously that PSTVd traffics from the phloem into all other cells in young sink leaves but is restricted to the phloem during sink-to-source transition of a leaf in N. benthamiana and tomato as a result of changes in leaf physiology during development . Therefore, in this and subsequent experiments, we focused on analysis of PSTVd traffic in sink leaves. Our analyses showed that the viroid signal was present only in the phloem of tobacco (Fig. , B --E). In contrast, the viroid was detected in the phloem, mesophyll, and epidermal cells of CoYMV:PSTVdInt-transgenic N. benthamiana (Fig. , F and G). These data indicate that in an appropriate host, companion cell-derived viroid can traffic into surrounding cells for replication. In transgenic tobacco, lack of interactions with a positive host factor or active interactions with a negative host factor may have contributed to the restriction of PSTVd in the phloem. We used reverse transcription (RT)-PCR) to amplify viroid progeny from all transgenic lines for sequencing. We also used PCR to amplify the PSTVd cDNA inserted into the genomes of these plant lines for sequencing. The viroid progeny from the transgenic N. benthamiana maintained the Intermediate sequence. Surprisingly, PSTVd progeny in line 4 of transgenic tobacco contain C259 -> U change and in line 8 contain U257 -> A change. The viroid cDNA sequence inserted into the plant genome was not altered in any lines. The C259 -> U change is identical to the mutation that converts the tomato-strain PSTVd KF440 --2 into the tobacco-infectious strain PSTVd-NT . We designate our two mutants as PSTVdInt U257 -> A and PSTVdInt C259 -> U, respectively. We should point out that PSTVdInt C259 -> U and PSTVd-NT, although both derived from C259 -> U substitution, are not identical because their respective parental strains PSTVdInt and PSTVd KF440 --2 have five nucleotide differences in the pathogenicity domain . To test further whether PSTVdInt U257 -> A and PSTVdInt C259 -> U were confined to the phloem in infected tobacco, we constructed cDNAs of these mutants to carry out two types of experiments. First, we mechanically inoculated in vitro transcripts derived from these cDNAs onto tobacco. In situ hybridization on young leaves of positively infected plants revealed presence of the viroid exclusively in the phloem (Fig. H). The viroid progeny from the infected plants were sequenced to confirm maintenance of the mutant sequences, respectively. Second, we generated transgenic tobacco expressing the cDNAs of these mutants under the control of the CoYMV promoter, respectively. In situ hybridization again showed that the viroid was confined to the phloem (data not shown), and sequence analysis also confirmed maintenance of the mutant sequences. We then tested whether U257 -> A or C259 -> U mutation would impair replication and/or traffic of PSTVd in N. benthamiana. We mechanically inoculated in vitro transcripts of both PSTVd mutants onto this plant. In situ hybridization showed that both PSTVdInt U257 -> A (Fig. I) and PSTVdInt C259 -> U (data not shown) infected all cells of systemic leaves. The mutant sequences were maintained in the infected plants. Thus, both U257 -> A and C259 -> U mutations did not abolish PSTVd replication or systemic movement in N. benthamiana. Altogether, these data provided compelling evidence that PSTVdInt U257 -> A and PSTVdInt C259 -> U replicated in and were confined to the phloem in tobacco. As shown below, this confinement was because of the inability of these mutants to exit the phloem. PSTVdInt C259 -> U and PSTVdInt U257 -> A Replicated in Nonvascular Tissues of Tobacco | Presence of PSTVdInt U257 -> A and PSTVdInt C259 -> U in the phloem and its absence from all other cells in transgenic tobacco raised a number of questions. First, was the phloem restriction attributable to inability of the viroid to traffic out of the phloem into nonvascular tissues such as the mesophyll for replication? Second, did the viroid traffic into nonvascular tissues but not replicate in these tissues? Third, are U257 -> A and C259 -> U changes essential for PSTVd replication in tobacco? To address these questions, we generated transgenic tobacco expressing the cDNA of PSTVdInt under the control of the CaMV 35S promoter. Ten transgenic lines were obtained that showed accumulation of PSTVd based on dot-blot analysis (data not shown). To determine the cellular localization of PSTVd in these transgenic plants, we performed in situ hybridization using a DIG-labeled RNA probe specific for the (-)-strand PSTVd. The viroid was detected in patches of mesophyll and phloem cells (Fig. , A and B), indicating that the viroid can replicate in these tissues. In transgenic N. benthamiana expressing the cDNA of the PSTVdInt under the control of the CaMV 35S promoter, the viroid was also detected in all leaf cells (; Fig. C). Figure 3 | Replication of PSTVd in mesophyll of transgenic tobacco and N. benthamiana Replication of PSTVd in mesophyll of transgenic tobacco and N. benthamiana. As detected by in situ hybridization on paraffin sections, each purple dot represents localization of PSTVd in the nucleus. A and B, Accumulation of PSTVdInt C259 -> U and PSTVdInt U257 -> A in mesophyll (MS) and phloem (Ph) of CaMV 35S: PSTVdInt-transgenic tobacco, respectively. C, Accumulation of PSTVdInt in mesophyll (MS) and phloem (Ph) of CaMV 35S:PSTVdInt-transgenic N. benthamiana. Bars = 40 mum. Sequence analysis revealed that viroid progeny isolated from transgenic N. benthamiana maintained the PSTVdInt sequence. In contrast, viroid progeny from seven of the transgenic tobacco lines contain C259 -> U change and progeny from the other three tobacco lines contain U257 -> A change. These changes are identical to those obtained from the CoYMV:PSTVdInt-transgenic plants. The viroid cDNA sequences inserted into the plant genome in all transgenic plants were not altered. Our data indicate that U257 -> A and C259 -> U substitutions in PSTVdInt are important for replication in tobacco, similar to the conversion of the tomato-strain PSTVd KF440 --2 into the tobacco-infectious strain PSTVd-NT by the C259 -> U substitution . As suggested by , a very low level of PSTVd replication, using primary transcripts derived from the 35S promoter activity as the initial templates, may have occurred in transgenic tobacco. During this replication, mutations occurred randomly and U257 -> A and C259 -> U changes enhanced some aspects of replication and accumulation. It is puzzling why the viroid appeared in patches of cells. Two explanations are possible. First, PSTVdInt U257 -> A or PSTVdInt C259 -> U mutations arose independently in a few cells of the mesophyll and phloem, and these variants replicate and move into surrounding cells (within the phloem and mesophyll, respectively). Second, PSTVdInt U257 -> A or PSTVdInt C259 -> U arose in a few cells in the early stages of organogenesis. Subsequent cell division spread the viroid population into multiple and contiguous cells. These potential mechanisms are not necessarily mutually exclusive. They could have all contributed to the observed accumulation of PSTVdInt U257 -> A or PSTVdInt C259 -> U in patches of cells in the transgenic plants. Because PSTVdInt U257 -> A and PSTVdInt C259 -> U can replicate in tobacco mesophyll and in the phloem, we attribute their phloem limitation in the CoYMV:PSTVdInt-transgenic tobacco and in the mechanically inoculated tobacco to their inability to traffic out of the phloem. Our data, thus, suggest that PSTVd replication and phloem-to-nonvascular tissue traffic are mediated by separate viroid motifs. U257 -> A and C259 -> U Substitutions Did Not Alter PSTVd Structure | Because PSTVd does not encode proteins, all functional information resides directly within the RNA genome itself. To gain insights into the molecular basis of U257 -> A and C259 -> U substitutions for the enhanced replication in tobacco, we analyzed the structure of PSTVdInt U257 -> A and PSTVdInt C259 -> U in comparison with that of the PSTVdInt. Computing with mfold (version 3.1 for 37C; ) or RNAstructure 3.6 revealed no difference in the secondary structure of the three strains . Data suggest that nucleotide changes, and not structural changes, led to the enhanced replication of PSTVd in tobacco. However, alternative approaches (e.g. chemical/enzymatic mapping and x-ray crystallography) are needed to verify independently the computed secondary structure and/or to uncover potential tertiary structures. Figure 4 | Selected portion of computed secondary structure of PSTVdInt, PSTVdInt C259 -> U, and PSTVdInt U257 -> A. Selected portion of computed secondary structure of PSTVdInt, PSTVdInt C259 -> U, and PSTVdInt U257 -> A. Nucleotide substitutions C259 -> U and U257 -> A are indicated. DISCUSSION : It is generally believed that, once in the phloem, long-distance movement of macromolecules including viral pathogens follows mass flow of photo-assimilates from source to sink organs/tissues . Recent studies showed that the 27-kD GFP ectopically produced in the companion cells of mature source leaves is loaded into sieve elements and transported to sink leaves in transgenic Arabidopsis and tobacco . further showed that GFP can traffic all the way into shoot apices and all floral organs. These observations have led to speculations that macromolecular flow from sieve elements to neighboring cells in a sink organ occurs by default without regulation . Our data demonstrate that PSTVdInt can replicate in all floral organs of transgenic N. benthamiana plants, and yet the viroid is present only in sepals but not in petals, stamens, and ovary in mechanically inoculated N. benthamiana. Considering these observations and the finding that some plant mRNAs can traffic into all floral organs , we suggest that a phloem-based mechanism selectively transports PSTVdInt into sepals but not into other floral organs. On the basis of this model, flow of macromolecules within sieve tubes to sink organs does not occur all by default. At present, we cannot exclude the possibility that the viroid is transported into the sieve elements of petals, stamens, and ovary but is not unloaded into surrounding nucleate cells for replication. In this case, the small amount of PSTVdInt in individual sieve elements could have escaped detection by the current in situ hybridization protocol. Whether traffic of PSTVd into various floral organs is regulated during passage within the sieve tubes or during exit from the sieve tubes, our observations support the hypothesis that the phloem has a sophisticated mechanism to deliver macromolecules to specific sink organs. Our studies on PSTVd traffic in leaves of tobacco and N. benthamiana provided compelling evidence that phloem exit of an RNA is highly controlled. Although PSTVdInt U257 -> A and PSTVdInt C259 -> U can replicate in the phloem and mesophyll cells of tobacco leaves, they do not traffic from the phloem to mesophyll. In N. benthamiana, these two mutants and the parental strain PSTVdInt can exit the phloem to invade neighboring tissues. These results lend strong support to the hypothesis that PSTVd contains structural motifs for intercellular movement . Consistent with this hypothesis, in situ hybridization of tomato roots and stems infected by mutant PSTVd-R revealed restriction of the mutant replication in the vascular tissue . PSTVd-R contains UU->AA substitutions at positions 177 and 178 and an additional G residue between positions 176 and 177 . It should be noted, however, that the replication function of mutant PSTVd-R in nonvascular tissues of these organs was not tested . Interestingly, fluorescently labeled in vitro transcripts of PSTVdInt can move from cell to cell when injected into tobacco mesophyll . Furthermore, PSTVdInt U257 -> A or PSTVdInt C259 -> U that was mechanically inoculated onto tobacco leaves apparently moved from epidermis to mesophyll and then into the phloem to spread systemically, but they did not exit the phloem in the systemic leaves. These results suggest that the PSTVd genome contains multiple structural motifs for traffic between different cell or tissue types. In particular, phloem entry and exit appear to be mediated by different motifs. There are examples that phloem entry and exit of a virus also use different mechanisms. showed that nontoxic concentrations of cadmium inhibit systemic infection of turnip vein clearing virus, otherwise an infectious virus, in tobacco. Cadmium has no effect on viral replication, assembly and local movement, but it appears to inhibit virus exit from the vascular tissue . Our data showing that PSTVdInt U257 -> A and PSTVdInt C259 -> U can replicate in the phloem and mesophyll but cannot traffic from the phloem to nonvascular tissues provide evidence that replication and phloem exit are mediated by different viroid motifs. Notably, both U257 -> A and C259 -> U mutations occur in the lower one-half of the same loop within the central conserved region of PSTVd. The analysis of specifically showed that nucleotide C259 is part of the PSTVd loop E sequence, which is conserved in eukaryotic 5S rRNA , 28S rRNA , and viroid RNAs . postulated that C259, as an extrahelical residue in loop E, is involved in interacting with host factors for efficient replication in tobacco. Our isolation of the tobacco-infectious strain PSTVdInt C259 -> U confirms the importance of C259 -> U for PSTVd replication in tobacco. U257 is significantly also part of the loop E (see ). That both U257 -> A and C259 -> U are critical for PSTVd replication but not for phloem exit in tobacco further underscores the involvement of loop E in viroid replication and suggests that the viroid motif(s) for phloem exit is located elsewhere. Work is in progress to identify this traffic motif(s). In search of cellular factors that are involved in phloem traffic of viroids, two recent studies demonstrate that phloem lectin PP2 from cucumber phloem exudate binds viroids and other RNAs . Whether PP2 has a role in viroid traffic is being investigated. In conclusion, our analyses of PSTVd traffic reveal regulatory points for systemic RNA traffic that likely involve distinct interactions between RNA motifs and cellular factors. RNA movement within sieve tubes may not simply follow mass flow from source to sink. It appears that some control mechanisms can sort and direct an RNA to selective sink organs. In addition, phloem exit of an RNA in sink organs is also highly controlled. A viroid most likely has evolved structural motifs that mimic endogenous plant RNA motifs so that they are recognized by cellular factors for traffic. In this regard, a viroid can be considered an exogenous RNA that contains "endogenous" traffic motifs and can, therefore, be used to study the mechanisms that control phloem entry, transport, and exit of plant RNAs. MATERIALS AND METHODS : Plant Material and Growth Conditions | Tobacco (Nicotiana tabacum cv Samsun) and Nicotiana benthamiana were grown in a growth chamber maintained at 27C/22C day/night temperature regimes and 14-/10-h light/dark cycle. Molecular Cloning and Generation of Transgenic Plants | Two types of DNA constructs were generated for plant transformation. First, the cDNA of PSTVdInt flanked by ribozymes was inserted into SmaI site, downstream of the CaMV 35S promoter, of plasmid pRTL2 . The 35S:PSTVdInt construct was then inserted into HindIII site of binary vector pGA482. Second, the cDNA of PSTVdInt was inserted into SmaI site of binary vector pCOI (provided by Dr. Gary Thompson [University of Arkansas, Little Rock]), downstream of the companion cell-specific CoYMV promoter . After sequence verification, these constructs were used to transform Agrobacterium tumefaciens (LBA 4404). A standard A. tumefaciens-mediated leaf-disc transformation method was used to generate transgenic tobacco and N. benthamiana plants. Dot-blot analysis with RNA probes specific for the (+)-strand PSTVd was used for initial screening for transgenic lines that accumulated PSTVd. Protocols for dot-blot analysis were described earlier . Sequencing of PSTVd cDNA and RNA Progeny | Genomic DNA was extracted from PSTVd transgenic plants using DNeasy Plant Mini Kit (Qiagen USA, Valencia, CA). PSTVd was amplified by PCR from genomic DNA, using primers PSTVd-AS (5'-CCTGAAGCGCTCCTCCGAG-3') and PSTVd-S (5'-GATCCCCGGGGAAACCTGG-3') and PfuTurbo DNA polymerase (Stratagene, La Jolla, CA) and the buffer provided by the manufacturer. The PCR cycling profile (28 cycles) was 30 s at 95C, 30 s at 55C, and 45 s at 68C with a final extension step at 68C for 10 min. The PCR products were fractionated on 1.5% (w/v) agarose gel. The band of expected size was cut out of the gel and purified with Qiaquick Gel Extraction Kit (Qiagen USA). Taq DNA polymerase (Promega, Madison, WI) was used to create 3'-A overhangs to the purified PCR products, which were then cloned into pGEM-T vector (Promega). PSTVd cDNAs in the plasmids were sequenced in both directions using the ABI377 DNA sequencer (PerkinElmer Life Sciences, Boston) at the DNA Sequencing Facility at The Ohio State University. To determine the sequences of RNA progeny, total RNA was isolated from the PSTVd-transgenic plants or mechanically inoculated plants using RNeasy Plant Mini Kit (Qiagen USA) following manufacturer's instructions. cDNAs of PSTVd RNA were synthesized by RT-PCR. The first cDNA strand was synthesized using the Thermoscript RT-PCR system (Invitrogen, Carlsbad, CA) and primer PSTVd-AS following the protocols recommended by the manufacturer. To synthesize the second strand of cDNA, aliquots (1/10) of the RT reaction mixture were PCR-amplified with primers PSTVd-AS and PSTVd-S as described. The PCR products were purified, cloned, and sequenced as described. Construction of cDNA Clones of PSTVdInt U257 -> A and PSTVdInt C259 -> U for in Vitro Transcription | To generate cDNA clones of the PSTVd mutants for in vitro transcription, the 294-bp EagI-Eco47 III fragment (positions 145 --359/1 --79) of the mutants were transferred to a double ribozyme expression cassette pRZ6 --2 by replacing the corresponding fragment of the PSTVdInt. This resulted in plasmids pRZ:IntU257 -> A and pRZ:IntC259 -> U. Synthesis of in Vitro Transcripts and Mechanical Inoculation | For production of in vitro transcripts of the PSTVdInt and the two mutants, plasmids pRZ6 --2 , pRZ:IntU257 -> A, and pRZ:IntC259 -> U were linearized with HindIII and used as template for in vitro transcription using the T7 MEGAscript Kit (Ambion, Austin, TX) following manufacturer's protocol. After in vitro transcription, the DNA template was removed by digestion with RNase-free DNase. The RNA transcripts were purified using RNeasy Plant Mini Kit (Qiagen USA) and diluted to a final concentration of 100 ng muL-1 in 20 mm sodium phosphate buffer (pH 7.0). Aliquots of 10 muL were rubbed onto each of carborundum-dusted young leaves of 2-week-old tobacco or N. benthamiana seedlings. Tissue Processing | Samples from transgenic or mechanically inoculated plants were processed to obtain paraffin sections or cryosections for in situ hybridization. Protocols for tissue processing to obtain paraffin sections were essentially as described in . In brief, plant samples were fixed in formaldehyde-acetic acid (10% [v/v] formaldehyde, 50% [v/v] ethanol, and 5% [v/v] acetic acid) at 4C overnight. After dehydration and infiltration, the samples were embedded in paraffin (Electron Microscopy Sciences, Fort Washington, PA). Sections (8 --10 mum) were obtained with a rotary microtome (model 820, American Optical Co., Buffalo, NY). Cryosections were obtained as described in . In brief, samples were fixed in a mixture of 3.7% (w/v) paraformaldehyde, 0.1% (v/v) glutaraldehyde, 0.2% (w/v) picric acid, 50 mm potassium phosphate, and 5 mm EGTA for 2 to 3 h. The fixed samples were infiltrated sequentially with 3:7 (v/v), 5:5 (v/v), and 7:3 (v/v) embedding mixture (two parts of 20% [w/v] Suc and one part of O.C.T. compound; Ted Pella Inc., Redding, CA):potassium phosphate/EGTA buffer. Afterward the samples were infiltrated with pure O.C.T. compound, embedded, and frozen at -20C. The frozen samples were sectioned to 12 mum thickness using a cryostat (HM500, Microm, Walldorf, Germany). In Situ Hybridization | DIG-labeled (+)-strand PSTVd was prepared by in vitro transcription using as template EcoRI-linearized plasmid pST64-B5 . DIG-UTP was purchased from Boehringer Mannheim (Indianapolis, IL), and in vitro transcription was carried out using the Ambion MEGAscript Kit (see above). In situ hybridization was performed as described previously , using reagents purchased from Boehringer Mannheim. In brief, cryosections or dewaxed paraffin sections were prehybridized in a blocking solution and then incubated with DIG-labeled PSTVd probes. Afterward, the sections were washed in 0.2x SSC and then incubated with alkaline phosphatase-conjugated anti-DIG antibodies. Finally, the sections were incubated in the color substrate (nitroblue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate) solution. When color had developed sufficiently, the sections were mounted and examined under an Eclipse 600 microscope (Nikon, Tokyo). Images were captured and processed with a SPOT 2 Slider CCD camera and the associated software (Diagnostics Instruments, Sterling Heights, MI). Distribution of Materials | Upon request, all novel materials described in this publication will be made available in a timely manner for noncommercial research purposes, subject to the requisite permission from any third-party owners of all or parts of the material. Obtaining any permissions will be the responsibility of the requestor. Backmatter: PMID- 12226495 TI - The Bifunctional LKR/SDH Locus of Plants Also Encodes a Highly Active Monofunctional Lysine-Ketoglutarate Reductase Using a Polyadenylation Signal Located within an Intron AB - Both plants and animals catabolize lysine (Lys) via two consecutive enzymes, Lys-ketoglutarate reductase (LKR) and saccharopine dehydrogenase (SDH), which are linked on a single polypeptide encoded by a single LKR/SDH gene. We have previously shown that the Arabidopsis LKR/SDH gene also encodes a monofunctional SDH that is transcribed from an internal promoter. In the present report, we have identified two cDNAs derived from cotton (Gossypium hirsutum) boll abscission zone that encode a novel enzymatic form of Lys catabolism, i.e. a catabolic monofunctional LKR. The monofunctional LKR mRNA is also encoded by the LKR/SDH gene, using two weak polyadenylation sites located within an intron. In situ mRNA hybridization and quantitative reverse transcriptase-polymerase chain reaction analyses also suggest that the cotton monofunctional LKR is relatively abundantly expressed in parenchyma cells of the abscission zone. DNA sequence analysis of the LKR/SDH genes of Arabidopsis, maize (Zea mays), and tomato (Lycopersicon esculentum) suggests that these genes can also encode a monofunctional LKR mRNA by a similar mechanism. To test whether the LKR/SDH and monofunctional LKR enzymes possess different biochemical properties, we used recombinant Arabidopsis LKR/SDH and monofunctional LKR enzymes expressed in yeast (Saccharomyces cerevisiae) cells. The Km of the monofunctional LKR to Lys was nearly 10-fold lower than its counterpart that is linked to SDH. Taken together, our results suggest that the LKR/SDH locus of plants is a super-composite locus that can encode three related but distinct enzymes of Lys catabolism. These three enzymes apparently operate in concert to finely regulate Lys catabolism during plant development. Keywords: Introduction : In plant and animal cells, the essential amino acid Lys is catabolized into acetyl CoA and several molecules of Glu (Fig. ; ; ). The first enzyme in the Lys catabolic pathway, Lys-ketoglutarate reductase (LKR), condenses Lys and alpha-ketoglutarate into saccharopine, which is then converted by the second enzyme saccharopine dehydrogenase (SDH) into alpha-amino adipic semialdehyde and Glu . In animals, Lys catabolism via alpha-amino adipic semialdehyde plays a significant physiological role in providing Glu for nervous signal transmission via Glu receptors . Figure 1 | The Lys catabolism pathway and metabolites derived from it. The Lys catabolism pathway and metabolites derived from it. Broken arrows represent several nonspecified enzymatic reactions. Glu residues are situated inside large boxes. The functional significance of Lys catabolism in plants is still unknown. In plants, the LKR level was shown to be significantly up-regulated in inflorescence tissues and developing seeds and in response to osmotic stress . In addition, LKR activity in tobacco (Nicotiana tabacum) and maize (Zea mays) seeds was shown to be stimulated by excess cellular Lys, via an intracellular signaling cascade involving Ca2+ and protein phosphorylation . In both plants and animals, LKR and SDH are linked on a single, bifunctional polypeptide encoded by a single gene . Yet, the metabolic flux of Lys catabolism does not depend only on the bifunctional LKR/SDH enzyme. We have demonstrated that in Arabidopsis the LKR/SDH locus also encodes a monofunctional SDH enzyme, using an internal promoter . In the present report, we provide evidence showing that the LKR/SDH locus of plants expresses another novel enzymatic form, i.e. a highly active catabolic monofunctional LKR. We propose that this enzyme may provide a transient superefficient flux of Lys catabolism under specific developmental programs. RESULTS : An mRNA Encoding a Novel Catabolic Monofunctional LKR Is Expressed in Cotton (Gossypium hirsutum) Boll Abscission Zone | To obtain a deeper insight into the functional significance of Lys catabolism in plant growth and development, we searched for LKR/SDH related sequences in a number of expressed sequence tag (EST) databases. Of particular interest was the discovery of three LKR/SDH-related ESTs of only approximately 1,800 ESTs derived from a cotton boll abscission zone cDNA library (. edu/projects/cotton/). This relatively high frequency suggested that LKR/SDH gene expression might be significantly up-regulated during abscission. The three cotton ESTs were ordered and sequenced to completion. One of these ESTs encoded a bifunctional LKR/SDH (GhLKR/SDH; EST no. Cabc0005CA08x), which is similar to the LKR/SDH cDNA of Arabidopsis. Yet, the other two ESTs represented two novel catabolic monofunctional LKR mRNAs (EST nos. Cabc0001cE06x and Cabc0007af10x). These two clones were designated GhLKR-1 and GhLKR-2, respectively. As shown schematically in Figure , the cotton GhLKR-1 and GhLKR-2 cDNAs were identical in DNA sequence to the bifunctional GhLKR/SDH cDNA through the coding DNA sequence of the LKR domain and part of the intermediate domain. However, they contained additional 3'-DNA sequences that did not exist in the bifunctional LKR/SDH cDNA. These 3'-DNA sequences included codons for the five C-terminal amino acids (VSIHN) followed by a stop codon, 3'-non-coding sequences (varying in length between GhLKR-1 and GhLKR-2) and a poly(A) tail. The three original cotton cDNAs were all incomplete, lacking part of the 5' region of the LKR open reading frame . Figure 2 | Schematic diagram of the cotton LKR/SDH-related ESTs. Schematic diagram of the cotton LKR/SDH-related ESTs. 1, The Arabidopsis AtLKR/SDH cDNA, used as a reference; 2 through 4, the cotton GhLKR/SDH, GhLKR1, and GhLKR2, cDNAs, respectively. Boxes with identical filling represent the same domain. ATG and TAG represent initiator and terminator codons, respectively. The position of the five deduced new C-terminal amino acids VSIHN of the two cotton monofunctional LKR ESTs is indicated by an arrow at the bottom. The Cotton LKR/SDH and Monofunctional LKR mRNAs Are Encoded by the Same Locus | The identical LKR-related DNA sequences between GhLKR/SDH, GhLKR-1, and GhLKR-2 suggested that the LKR/SDH and monofunctional LKR mRNAs are encoded by the same composite LKR/SDH locus. To study the genetic control of these two mRNAs, cotton DNA was digested with several restrictions enzymes, fractionated by agarose gel electrophoresis, and hybridized in a Southern blot with GhLKR-1 as a probe. As shown in Figure , hybridization with this probe hybridized to mostly two DNA bands. Because the G. hirsutum species of cotton is allotetraploid, it is likely that this species possesses two genes, one for each of its two diploid genomes. To clone the cotton LKR/SDH locus, we screened a cotton bacterial artificial chromosome (BAC) library by hybridization with GhLKR-1 as a probe. A positive BAC plasmid was identified and confirmed to possess the LKR/SDH locus by partial DNA sequence analysis (data not shown). To test whether the LKR/SDH locus inside this BAC plasmid encodes both the bifunctional LKR/SDH and monofunctional LKR mRNAs, it was hybridized with the unique 3' sequence of the GhLKR-1, which is missing in GhLKR/SDH cDNA. The BAC plasmid positively hybridized with this probe too (data not shown). Figure 3 | Southern-blot analysis of LKR/SDH-related sequences in cotton. Southern-blot analysis of LKR/SDH-related sequences in cotton. Cotton genomic DNA was digested with various restriction enzymes as indicated on the top of the panel. The digested DNA was reacted in a Southern blot with a probe derived from the cotton GhLKR-1 cDNA as a probe. The migration of DNA size markers is indicated on the left. The LKR/SDH locus in the BAC clone was incomplete, lacking the promoter and a small part of the 5'-coding region. However, using inverse PCR of cotton genomic DNA, we were able to clone the 5'-coding region and to generate the full open reading frames of LKR/SDH and the two monofunctional LKR cDNAs (GenBank accession nos. , , and , respectively). The Cotton Monofunctional LKR mRNA Is Produced by a Polyadenylation Site Located within a Large Intron in the Linker Region between LKR and SDH Coding Sequences | To test the origin of the unique 3' sequences of GhLKR-1 and GhLKR-2 mRNAs, we subcloned and sequenced the DNA region encoding the intermediate domain between the LKR and SDH domains in the gGhLKR/SDH locus. As shown schematically in Figure A, the 3' unique sequences of GhLKR-1 and GhLKR-2 were located inside a large intron (GenBank accession no. ) located within this intermediate domain region. Furthermore, as shown in Figure B, analysis of this intron revealed at least two putative polyadenylation sites each containing an AT-rich PE plus additional elements upstream and downstream to it, which meet the criteria of plant polyadenylation sites as reported by . Only one of these two putative polyadenylation sites was present in GhLKR-1, whereas both were present in GhLKR-2, slightly upstream of their poly(A) tails. None of these two polyadenylation sites resembled the complete consensus plant polyadenylation site , suggesting that they are rather low-affinity sites. Figure 4 | Localization of the 3'-DNA sequence of the cotton monofunctional LKR cDNAs within an intron of the cotton LKR/SDH locus. Localization of the 3'-DNA sequence of the cotton monofunctional LKR cDNAs within an intron of the cotton LKR/SDH locus. A, Schematic diagram showing the position of the intron within the intermediate domain region. The location of the five deduced C-terminal amino acids of GhLKR1 and GhLKR2 followed by a stop codon (*) at the beginning of this intron (bolded capital letters), the position of the two poly(A) residues of these cDNAs, and the location of the two putative low-affinity polyadenylation sites (black boxes marked 1 and 2) are indicated. ATG and TAG represent initiator and terminator codons, respectively. B, The 5'-DNA sequence within the intron shown in A. The exon/intron junction is indicated by an arrow. The deduced seven C-terminal LEVSIHN amino acids (of which the last five encoded by intron sequences) are indicated in bold letters followed by a TAA stop codon (*). The positions of the two putative polyadenylation elements (PE) are indicated in bolded letters. The positions of sequences upstream and downstream to the PE, which resemble those reported by , are indicated in bold underlined letters. The Monofunctional LKR Is Abundantly Expressed in Parenchyma Cells of the Abscission Zone | To substantiate the presence of a monofunctional LKR in the cotton abscission zone and identify the cells expressing it, we induced abscission zone of cotton leaves, using the ethylene releasing compound ethephon (see "Materials and Methods"). Longitudinal sections of abscission zone at the base of the petiole were then subjected to in situ mRNA hybridization, using monofunctional LKR-specific antisense and sense RNA probes, derived from the 3' intron-located non-coding region of the monofunctional LKR. As shown in Figure , a nonspecific color was observed inside the vacuoles of epidermis cells treated with either the antisense or sense monofunctional LKR probes (Fig. , a --f). This color was also observed in untreated sections (data not shown). Nevertheless, no alkaline phosphatase staining was detected in the cytosol of these cells after hybridization with both the antisense (lanes a, c, and e) and sense (lanes b, d, and f) probes. In contrast, the parenchyma cells beneath the epidermis layers strongly hybridized with the antisense, but not the sense probes. Many of the parenchyma cells are unfortunately highly vacuolated with very little cytoplasm, and it is, hence, very difficult to see the alkaline phosphatase staining in a photograph. However, clear, intense alkaline phosphatase staining is seen in a number of cells where more elaborated cytoplasm is present (Fig. , a, c, and e, staining marked by arrowheads). No such staining is seen in sections treated with the control sense probe (Fig. , b, d, and f). Figure 5 | In situ hybridization pattern of the cotton monofunctional LKR mRNA in cotton leaf abscission zone. In situ hybridization pattern of the cotton monofunctional LKR mRNA in cotton leaf abscission zone. Near median longitudinal sections (8 mum) through the base of the petiole were hybridized with Dig-labeled antisense (a, c, and e) and sense (b, d, and f) probes specific to the monofunctional LKR mRNA (see "Materials and Methods"). c and d are higher magnifications of a and b, respectively. e and f are higher magnification of neighboring cells. EP, Epidermis; NS, nonspecific color inside vacuoles. The location of the monofunctional LKR mRNA in the cytosol, surrounding the large vacuoles, in parenchyma cells is indicated by arrowheads. Similar longitudinal sections as shown in Figure where also probed with antisense and sense RNA probes derived the SDH domain of the cotton LKR/SDH cDNA to localize the LKR/SDH mRNA. The SDH antisense RNA probe, but not the sense probe, labeled the same type of parenchyma cells shown in Figure , although the labeling intensity did not seem as intense as that obtained with the monofunctional LKR probe (data not shown). Expression of the monofunctional LKR in the leaf abscission zone was also studied by a quantitative reverse transcriptase (RT)-PCR, in comparison with normal leaf base (not treated with ethephon) and flower buds as controls (see "Materials and Methods"). Three sets of specific primers were used. One set was specific for the monofunctional LKR, the second specific for the bifunctional LKR/SDH, and the third specific for ubiquitin as an internal control. As shown in the top panel of Figure , the ubiquitin-specific primers amplified DNA bands with comparable intensities from the three different tissues, suggesting that the RT reaction operated at a comparable efficiency on RNAs extracted from these tissues. As shown in the middle and bottom panels of Figure , the relative intensity of the monofunctional LKR band was considerably higher in RT-PCR from abscission zone RNA than from the untreated leaf base and flower buds RNAs. The opposite was observed for the bifunctional LKR/SDH band, which was relatively more intense in RT-PCR from flower buds RNA than abscission zone and untreated leaf base RNAs. The intensity of the monofunctional LKR band derived from the untreated leaf base RNA was notably slightly higher than that derived from the flower buds RNA. This may indicate that leaf bases become committed to form abscission zones long before abscission can be detected morphologically. The quantitative RT-PCR was repeated several times with similar results. Figure 6 | Quantitative RT-PCR analysis of the bifunctional LKR/SDH and monofunctional LKR mRNAs in cotton. Quantitative RT-PCR analysis of the bifunctional LKR/SDH and monofunctional LKR mRNAs in cotton. Top, middle, and bottom, PCR products with primers specific to the control ubiquitin, monofunctional LKR, and LKR/SDH mRNAs, respectively. Lanes a through c, RNA taken from induced leaf abscission zone, noninduced leaf base, and flower buds, respectively. The LKR/SDH Genes of Maize, Arabidopsis, and Tomato (Lycopersicon esculentum) Also Possess Putative Polyadenylation Sites within Introns for Production of a Monofunctional LKR mRNA | In two previous reports , faint mRNA bands of smaller sizes than the LKR/SDH mRNA were detected in northern blots from maize and Arabidopsis plants, which were specifically hybridized with the LKR but not with the SDH domain of the LKR/SDH gene. The origins of these LKR-specific mRNA bands were not elucidated. To explore whether these low abundance mRNAs may have been produced by transcription termination within introns, we analyzed the DNA sequence of introns within the intermediate domain regions of the maize and Arabidopsis LKR/SDH genes (GenBank accession nos. and , respectively). Several low-affinity polyadenylation sites were identified within introns from both genes, which can potentially produce monofunctional LKR mRNAs. Examples of such polyadenylation sites within intron 12 of the maize LKR/SDH gene and intron 11 of the Arabidopsis LKR/SDH gene are illustrated in the supplementary Figure 1 and supplementary Figure 2, respectively (they can be viewed at ). Using a similar DNA sequence analysis, we have also recently identified putative low-affinity polyadenylation sites within introns in the intermediate domain region of the tomato LKR/SDH gene (supplementary Fig. 3; it can be viewed at ). These results suggest that a monofunctional LKR mRNA is produced in many plant species. The Monofunctional LKR Is a More Efficient Enzyme Than Its Counterpart That Is Linked to SDH | To provide some insight into the physiological significance of the new catabolic monofunctional LKR, we wished to compare its biochemical properties with that of its counterpart that is linked to SDH. To address this, we used recombinant Arabidopsis LKR/SDH and monofunctional LKR, fused at their N termini to a tag of six histidines (His tag), which we have previously expressed in yeast (Saccharomyces cerevisiae) cells . We have also previously shown that yeast is a highly reliable system to study the activities of recombinant LKR/SDH enzymes from Arabidopsis . The recombinant LKR and LKR/SDH were purified on a nickel column and analyzed for LKR activity under increasing concentration of the three LKR substrates Lys, alpha-ketoglutarate, and NADPH. Reactions were performed in a 0.1 m phosphate buffer at pH 7.5, which resembles the physiological pH of the cytosol where LKR/SDH and apparently also the monofunctional LKR are localized . As shown in Table , the monofunctional LKR possessed a nearly 10-fold lower Km to Lys than the LKR activity of LKR/SDH. No difference between these two enzymes was observed for the Km of alpha-ketoglutarate and NADPH. The average LKR specific activities of the monofunctional LKR and bifunctional LKR/SDH enzymes under conditions of excess substrate concentrations (Vmax), as determined by units per microgram of protein +- sd, were also comparable being 20.59 +- 1.17 and 15.15 +- 1.11, respectively. Table I | Km of LKR activity of the Arabidopsis LKR/SDH and monofunctional LKR to its three different substrates DISCUSSION : The Cotton LKR/SDH Locus Encodes a Novel Form of a Catabolic Monofunctional LKR, Using Internal Low-Affinity Polyadenylation Sites | By complete sequencing of three ESTs derived from cotton boll abscission zone, we have identified a bifunctional LKR/SDH cDNA and two cDNAs encoding a novel form of a catabolic monofunctional LKR. The coding DNA sequences of the monofunctional LKR cDNAs were identical to each other and to that of the bifunctional LKR/SDH, implying that they are produced from the LKR/SDH locus. This was also confirmed by DNA sequence analysis of the LKR/SDH gene from a BAC clone showing that the monofunctional LKR-specific 3'-DNA sequences were derived from a large intron within the intermediate domain region between the LKR and SDH coding regions. Our Southern-blot analysis also supported this observation, indicating the presence of two LKR/SDH genes in tetraploid cotton, apparently a single gene per diploid genome. Because both of the cotton monofunctional LKR cDNAs ended with a 3'-poly(A) tail, we concluded that their transcription termination is regulated by two polyadenylation sites located inside the large intron within the intermediate domain region of the cotton LKR/SDH locus. Such putative elements were identified within the DNA sequence of the intron . Production of the Cotton Monofunctional LKR mRNA Is Enhanced in the Parenchyma Cells of the Abscission Zone | Using in situ mRNA localization with RNA probes derived from the intron within the intermediate domain of the cotton LKR/SDH gene, we showed that the monofunctional LKR is predominantly expressed in parenchyma cells of the abscission zone. Moreover, our quantitative RT-PCR also suggested that the level of the monofunctional LKR in these cells is relatively higher than in nonabscised leaf bases and flower buds. This was also supported by the relative high frequency of LKR/SH-derived sequences in the cotton abscission zone library (three of only approximately 1, 800 ESTs) and by the fact that two of these three ESTs were derived from monofunctional LKR mRNAs. The mechanism controlling the enhanced production of monofunctional LKR in abscission zones is not known but is likely related to the efficiency of transcription termination and mRNA polyadenylation at the two putative polyadenylation sites within the intron. The sequences of these two polyadenylation signals are notably much more diverged from the consensus plant polyadenylation signals than the polyadenylation sites in the 3' end of the LKR/SDH gene. Thus, it is likely that in many tissues, such as flower buds, transcription termination and mRNA polyadenylation within the intron will be minor resulting in dominant production of the bifunctional LKR/SDH mRNA. Enhanced production of the monofunctional LKR mRNA in the abscission zone may be regulated by a mechanism that enhances the recognition of the two low-affinity polyadenylation sites within the intron. Such a mechanism was previously shown to control the synthesis variant proteins from a number of mammalian genes, with a good example being the production of a soluble and a membrane-bound form of IgM heavy chains from a single locus in human B cells . The switch to the synthesis of the smaller soluble IgM polypeptide occurs by preferred use of a low-affinity polyadenylation site within an intron. This preferred use is associated with enhanced synthesis of CstF64, a member of a protein complex that binds to the polyadenylation site . Whether a similar mechanism also operates in plants still remains to be elucidated. A Monofunctional LKR mRNA Is Likely Also Produced from the LKR/SDH Loci of Other Plant Species | By analysis of the sequences of the intermediate domain regions of the maize, Arabidopsis, and tomato LKR/SDH loci, we identified putative low-affinity polyadenylation signals with introns that can direct the synthesis of monofunctional LKR in all of these plant species. Moreover, transcription termination at these low-affinity polyadenylation sites would have expected to produce transcripts that correspond in sizes to the faint LKR-specific mRNAs bands that were previously detected in maize and Arabidopsis . Transcription termination and polyadenylation in sequences that diverge from the consensus plant polyadenylation sites were previously reported when yeast genes were expressed in plants , supporting the presence of a polyadenylation machinery that can recognize such low-affinity polyadenylation signals. Our results suggest further that such a machinery is quite dominant in abscission zones and is used to produce a novel functional gene product, i.e. a catabolic monofunctional LKR. The Physiological Significance of the Catabolic Monofunctional LKR | Our biochemical studies showed that the Arabidopsis monofunctional LKR possesses a Km of 0.328 mm Lys, whereas its counterpart that is linked to SDH possesses a significantly higher Km of 5.18 mm Lys . We believe that these characteristics are also true for LKR/SDH and monofunctional LKR enzymes of other plant species because the LKR/SDH polypeptides of different plant species possess highly conserved sequences and biochemical properties . Because the level of Lys in the plant cytosol is estimated to be around 1 mm , our results imply that the monofunctional LKR is much more efficient than its counterpart that is linked to SDH in vivo. Taken together, we, thus, hypothesize that plants possess two different fluxes of Lys catabolism. In some tissues, such as inflorescence and developing seeds, the LKR/SDH is apparently the major contributor to the total LKR activity, and the flux of Lys catabolism may therefore be relatively limited because of the inefficient nature of LKR/SDH. The major function of Lys catabolism in such tissues may be to regulate Lys homeostasis. A transient enhanced production of the more efficient monofunctional LKR (and in some plant species also a monofunctional SDH; ) in specific physiological programs, like abscission, may enable a temporary superefficient flux of Lys catabolism into Glu . Glu is an important regulatory metabolite in plants. It serves as a precursor for the stress-related metabolites Pro, gamma-amino butyric acid, and Arg, which is the donor for polyamines and nitric oxide and is also a modulator of Glu (for review, see ). MATERIALS AND METHODS : Materials | Cotton (Gossypium hirsutum) plants were grown in pots under a greenhouse condition (12-h photoperiod at 25C +- 5C). Ethephon (Dropp Ultra soluble concentrate, Agrevo, Cambridge, UK) was kindly provided by Dr. Alon Haberfeld (Hazera Quality Seeds Ltd., Kinyat Gat, Israel). Long-template PCR amplification Taq polymerase, PWO Taq polymerase, dNTPs, RNase-free DNase, and DIG RNA labeling kit were purchased from Roche Diagnostics (Mannheim, Germany). Moloney murine leukemia virus (M-MLV) RT was purchased from Promega (Madison, WI). Super-Therm DNA Polymerase was purchased from JMR Holdings (London). Isolation of the 5' Region of the Cotton LKR/SDH Locus | The 5'-coding region of the cotton LKR/SDH (missing 5' regions in the three cotton ESTs) was cloned by inverse nested PCR as follows: Two micrograms of cotton genomic DNA was cleaved with XbaI for 2 h. After heat inactivating the enzyme at 65C for 20 min, the DNA was allowed to self-ligate overnight. Two microliters of the overnight ligation reaction was used as a template for PCR analysis using the two oligonucleotides Cot-P9-R and Cot-P2-F . Reaction conditions included: annealing temp of 55C, 8-min elongation, 40 cycles, and long-template PCR amplification Taq polymerase. Two microliters of the inverse PCR reaction was used as a template, for a second PCR reaction with the two oligonucleotides Cot-Inv-2-R and Cot-P7-F . Reaction conditions were as for the first PCR reaction, but the PWO Taq polymerase was used. The PCR products were cleaned with phenol/chloroform cut with XbaI and cloned into the XbaI and SmaI sites of the SK plasmid. The subcloned PCR products were sequenced and assembled into the 5' regions of the three cotton LKR/SDH-related ESTs to generate full open reading frames. Table II | DNA sequences of oligonucleotides used in the present study Induction of Leaf Abscission Zone | Blades of fully developed young cotton leaves were excised, and several drops of 20x dilution of the ethephon soluble concentrate were applied to the cutting. Control leaves were kept intact. About 5 d after addition of the ethephon, leaf bases were collected and either immediately frozen in liquid nitrogen and kept at -70C or fixed for in situ analysis as previously described . Primer Use for Specific Amplification of cDNAs | The sequences of the primers used for PCR amplification are illustrated in Table . The cotton ubiquitin mRNA (GenBank accession no. ) was amplified by primers Cot-UBQ-5' and Cot-UBQ-3', generating a 312-bp-long DNA fragment. Specific amplification of the cotton LKR/SDH and monofunctional LKR were performed by two sets of primers, each containing the common forward Cot-P2-F primer derived from the LKR domain of GhLKR/SDH. For specific amplification of the bifunctional LKR/SDH, a second reverse primer derived from the SDH domain (Cot-P24-R) was used; whereas for specific detection of the monofunctional LKR, a second reverse primer (Cot-P5-R), located upstream the polyadenylation site within the intron situated within the intermediate domain was used. These two sets of primers amplified DNAs of 255 and 257 bp, respectively. In Situ mRNA Hybridization | To construct specific sense and antisense probes for in situ detection of the cotton monofunctional LKR, the two primers Cot-P5-R and Cot-P16-F were used for PCR on cotton genomic DNA using PWO Taq polymerase. These primers amplify DNA sequences from the large intron within the intermediate domain of the cotton LKR/SDH gene. The PCR product was cloned into the XbaI and Sam sites of the Bluescript SK vector. To construct specific sense and antisense probes for in situ detection of the cotton bifunctional LKR/SDH, the GhLKR/SDH cDNA was amplified with primers cot-P7-F and cot-P17-R from the SDH domain. The PCR product was digested with EcoRI and SmaI (inside the cot-P17-R primer) and subcloned into the Bluescript SK plasmid. Digoxigenin-labeled sense and antisense RNA probes were obtained by in vitro transcription using the DIG RNA labeling kit. Tissue preparation and in situ hybridization were carried out as described . An antisense probe and the corresponding control sense probe were used in each experiment. Extraction of Nucleic Acids and Quantitative RT-PCR Analysis | Cotton DNA was extracted as previously described . Total RNA from various cotton tissues was isolated as previously described . Total RNA (1 mug) was treated with 1 unit of RNase-free DNase. After enzyme inactivation at 65C 20 min, the RNA was annealed with 10 nmol of oligo(dT) (17-mer) primer. Reverse transcription was performed by adding a mixture of 1x M-MLV buffer, 1 mm dNTPs, 20 units of RNase inhibitor, and 200 units of M-MLV RT in a 20-muL final volume. Reactions were incubated at 42C for 1 h, and then the RT was inactivated at 95C for 5 min. Two microliters of RT reaction was taken into a mixture of 1x PCR buffer, 0.25 mm dNTPs, 5 pmol of primers of the specific genes, 2.5 units of Super-Therm DNA Polymerase in a 50-muL final volume for PCR amplification. PCR was performed for 30 cycles each containing 45 s at 94C, 1 min at 53C, and 1 min at 72C. The PCR products were then separated on a 1% (w/v) agarose gel. Amplification of the different RT reactions with the ubiquitin primers resulted in bands with similar intensities between different tissues, whereas their amplification with the LKR/SDH and monofunctional LKR primers resulted in bands with different intensities between tissues. This suggested that the PCR reactions were still in the linear range of amplification and did not reach the plateau level expected from deletion of the rate-limiting PCR components. To further corroborate the quantitative nature of the results, all first amplification reactions were diluted twice by 10- and 100-fold and subjected to a second amplification as specified above. Amplification of both dilutions resulted in very similar relative intensities of the different bands to that observed by the first amplification of the RT reaction, supporting the linearity of all amplification reactions. Purification of His-Tagged Proteins and Analysis of LKR Activity | Expression in yeast (Saccharomyces cerevisiae) and purification of Arabidopsis bifunctional LKR/SDH and monofunctional LKR polypeptides, fused at their N termini to an epitope tag of six histidines (His tag), has been previously described . LKR activity was performed in an incubation assay containing approximately 0.1 mug of protein from the nearly purified preparations in 0.3 mL of 0.1 m phosphate buffer (pH 7.5), 20 mm Lys, 14 mm alpha-ketoglutarate, and 0.4 mm NADPH as previously described . Experiments elucidating the apparent Km of the three LKR substrates used different concentration of these substrates. The apparent Km values were determined graphically from double-reciprocal plots of activity against the variable substrate concentration. Backmatter: PMID- 12226496 TI - elongated mesocotyl1, a Phytochrome-Deficient Mutant of Maize AB - To begin the functional dissection of light signal transduction pathways of maize (Zea mays), we have identified and characterized the light-sensing mutant elm1 (elongated mesocotyl1). Seedlings homozygous for elm1 are pale green, show pronounced elongation of the mesocotyl, and fail to de-etiolate under red or far-red light. Etiolated elm1 mutants contain no spectrally active phytochrome and do not deplete levels of phytochrome A after red-light treatment. High-performance liquid chromatography analyses show that elm1 mutants are unable to convert biliverdin IXalpha to 3Z-phytochromobilin, preventing synthesis of the phytochrome chromophore. Despite the impairment of the phytochrome photoreceptors, elm1 mutants can be grown to maturity in the field. Mature plants retain aspects of the seedling phenotype and flower earlier than wild-type plants under long days. Thus, the elm1 mutant of maize provides the first direct evidence for phytochrome-mediated modulation of flowering time in this agronomically important species. Keywords: Introduction : The phytochrome family of photoreceptors mediates many of the responses that plants display to changes in their light environment . The basis of phytochrome action is a reversible photoconversion between a red light (R)-absorbing form (Pr) and a far-red light (FR)-absorbing form (Pfr; ). In lower plants, the family is represented by a small number of nuclear genes . However, gene duplication and evolutionary divergence have resulted in the formation of functionally diverse multigene families in flowering plants. In Arabidopsis, the phytochrome family consists of five genes: PHYA, PHYB, PHYC, PHYD, and PHYE , whereas the grasses have three phytochromes: PhyA, PhyB, and PhyC . In maize (Zea mays), an ancestral genomic duplication has increased the total family size to at least six: PhyA1, PhyA2, PhyB1, PhyB2, PhyC1, PhyC2, and possibly PhyC3 . Although loss-of-function phy mutants have been characterized in a broad range of plants, including Arabidopsis (for review, see ), sorghum (Sorghum bicolor; ), barley (Hordeum vulgare; ), and rice (Oryza sativa; ), no phytochrome gene mutants have been characterized in maize. Gene duplication within the maize phytochrome family and the accompanying potential for functional redundancy may have obscured genetic screens for phy mutants. The photoactive holoprotein (phy) consists of a PHY apoprotein (PHY) covalently attached to a linear tetrapyrrole (bilin) chromophore, 3E-phytochromobilin (PPhiB; ). The first committed step in the synthesis of PPhiB is the conversion of heme to biliverdin (BV) IXalpha by the enzyme heme oxygenase . BV IXalpha is then reduced to 3Z-PPhiB by PPhiB synthase and subsequently isomerized to 3E-PPhiB . Of these three activities, genes encoding the first two have now been cloned . The HO1 (HY1) gene encodes heme oxygenase, which is targeted to the plastid . The HY2 gene encodes PPhiB synthase, a ferredoxin-dependent BV reductase, which is also plastid localized . It is not yet known whether the isomerization of 3Z-PPhiB to 3E-PPhiB is enzyme mediated or whether it occurs spontaneously. Although phytochrome apoproteins are encoded by a multigene family, it is likely that all plant apophytochromes bind the same chromophore. Therefore, genetic disruption of linear tetrapyrrole synthesis offers a way to specifically inactivate the entire phytochrome system. There are a number of known mutants in which linear tetrapyrrole synthesis is disrupted. These include the hy1 and hy2 mutants of Arabidopsis , the pew1 (partially etiolated-in-white-light1) and pew2 mutants of Nicotiana plumbaginifolia , the pcd1 (phytochrome chromophore-deficient1) and pcd2 mutants of pea (Pisum sativum; , ), the au (aurea) and yg-2 (yellow-green2) mutants of tomato (Lycopersicon esculentum; ; ), and the se5 (photoperiodic sensitive5) mutant of rice . All these mutants have lesions in either heme oxygenase or PPhiB synthase and show a reduction in light responsiveness. However, a common characteristic of these mutants is that they show a partial recovery of light sensitivity during development , suggesting that other enzymes are present that can partially complement these mutations. Support for such an idea has recently come from , who have shown that heme oxygenase is encoded by a small gene family that may be functionally redundant. Although many molecular characterizations of phytochrome signaling have focused on seedling responses, a number of studies have demonstrated the importance of phytochrome in mature, field-grown plants . In several cases, the mutation of a single phy gene has dramatically changed the mature plant phenotype. Two such examples are the early flowering ma3R line of sorghum and the photoperiod-insensitive BMDR1 line of barley . Overexpression of an oat (Avena sativa) PHYA in tobacco (Nicotiana tabacum) resulted in dramatic morphological changes that increased harvest index . Characterization of the se5 heme-oxygenase mutant of rice has also demonstrated a significant contribution of the phytochrome system to the regulation of flowering time in this crop species . In this study, the isolation and initial characterization of the elongated mesocotyl1 (elm1) mutant of maize is presented. We show that the elm1 mutant has a reduced accumulation of active phytochrome. Under long-day (LD) growth conditions, elm1 mutants flower earlier than near-isogenic wild-type plants, indicating that phytochrome signaling can modulate flowering time in maize. RESULTS : Isolation of the elm1 Mutant | The elm1 mutation was identified in seedling screens of an Ac-mutagenized population. All lines in this population were maintained in a standard W22 inbred, enabling near-isogenic comparisons between any alleles recovered. To identify light-signaling mutants, Ac transpositions were selected from several donor elements located throughout the maize genome. F1 plants were grown and self-pollinated to generate approximately 100 F2 families. Approximately 20 kernels from each ear were then screened in greenhouse sandbenches to identify elongated pale-green seedlings. A similar screen was previously used to identify long-hypocotyl mutants of Arabidopsis and tomato . A single line was identified that segregated pale-green seedlings with elongated mesocotyls as a simple recessive trait. The mutation was designated elm1. Southern-blot analysis has failed to detect linkage of the transposable element Ac to the elm1 locus (data not shown); thus, it is unlikely that the elm1 allele contains an Ac insertion. Mature, field-grown elm1 plants have elongated internodes , pale-green leaves, and display a tendency to lodge (fall down). Under LD growth conditions in the field (14 --16 h of light, Ithaca, NY), elm1 plants flowered approximately 5 d earlier than wild type (Elm1, n = 60, mean = 79.1 d after planting [dap], se = 0.22; elm1, n = 63, mean = 74.4 dap, se = 0.38; Wilcoxon two-sample, non-paired, rank test, U = 3,639.5, P < 0.01). The similarity of elm1 seedling and mature plant phenotypes to previously characterized mutants of Arabidopsis and tomato suggested that the elm1 mutant is impaired in light perception or signal transduction. Figure 1 | Mature plant phenotypes of wild-type (Elm1) and mutant (elm1) plants. Mature plant phenotypes of wild-type (Elm1) and mutant (elm1) plants. Wild-type (left) and elm1 mutant (right) plants grown at summer field site (LD conditions). elm1 mutants are taller, have pale internodes, and flower earlier than near-isogenic siblings. elm1 Mutant Seedlings Show a Disruption of the De-Etiolation Response | The phenotype of elm1 seedlings grown under white light (W) is shown in Figure . Under these conditions, elm1 showed a moderately elongated phenotype , lower levels of chlorophyll (Chl), and an increased Chl a:b ratio . Carotenoid levels were also reduced in elm1 compared with wild-type seedlings . Detailed genetic analyses in Arabidopsis have indicated that phyB is the primary photoreceptor mediating de-etiolation in response to red (R), whereas phyA is the primary photoreceptor mediating responses to far-red (FR) . To further examine the light signal transduction pathway in elm1 mutants, we examined the inhibitory effect of R and FR on mesocotyl elongation. The mesocotyl can be considered functionally analogous to the dicot hypocotyl and, in wild-type seedlings, is greatly elongated in the absence of light stimuli (see Fig. ). Wild-type and elm1 plants were grown in dark (D), R (3 mumol m-2 s-1), or FR (3 mumol m-2 s-1) and mesocotyl lengths measured 10 dap. As shown in Figure , etiolated (D) wild-type and elm1 seedlings showed similar elongation of the mesocotyl. In wild-type plants, elongation was strongly inhibited by both R and FR. In contrast, mesocotyl length was similar in elm1 seedlings under all growth conditions tested . This morphology is indicative of a lack of responsiveness to either R or FR. Figure 2 | Seedling phenotypes of wild-type (Elm1) and mutant (elm1) plants. Seedling phenotypes of wild-type (Elm1) and mutant (elm1) plants. Representative seedlings were photographed after 10 d of growth in D, R (3.0 mumol m-2 s-1), FR (1.2 mumol m-2 s-1), or W (100 mumol m-2 s-1). Arrows indicate boundary between mesocotyl and first internode. Scale bar divisions are in centimeters. Table I | Pigment measurements Figure 3 | De-etiolation responses of wild-type (Elm1) and mutant (elm1) plants. De-etiolation responses of wild-type (Elm1) and mutant (elm1) plants. Mean (+-se) mesocotyl length measurements in wild-type (Elm1) and mutant (elm1) seedlings grown for 10 d in D, R (3.0 mumol m-2 s-1), or FR (2.0 mumol m-2 s-1) growth conditions. Sample size is 15 to 20 seedlings per treatment/genotype. elm1 Seedlings Do Not Contain Spectrally Active Phytochrome Pools | The nonresponsiveness of elm1 seedlings to both R and FR suggests disruption of both phyA- and phyB-mediated responses. To further investigate the activity of phytochrome in elm1 mutants, spectrophotometrically active pools of phytochrome were directly measured in etiolated elm1 and wild-type seedlings. Using in vivo spectrophotometry, the signal from elm1 seedlings was below the level of detection, whereas the signal from etiolated wild-type seedlings was 4.9 +- 0.1 (n = 2) units (1 unit is 1 x 10-3 DeltaDeltaA730 --800 nm). The level of detection is <0.3 units; therefore, elm1 seedlings contain <6% of the spectrally active phytochrome present in wild-type plants. As an additional assay of phytochrome activity, PHYA accumulation was examined in elm1 seedlings. phyA is the most abundant phytochrome in etiolated tissue, but is rapidly degraded upon illumination. Because this degradation requires Pfr formation, the change in phyA stability after a light treatment can be used to assay the degree of photoconversion . As shown in Figure A, wild-type and elm1 seedlings accumulate PHYA in D. After 4 h of R (10 mumol m-2 s-1), phyA pools are rapidly depleted in wild-type but not in elm1 seedlings. Figure B shows that levels of PHYA were approximately 5- to 10-fold higher in elm1 seedlings relative to wild-type plants after a 4-h R treatment. This suggests that although PHYA accumulates in elm1 seedlings, it is not bound to PPhiB. Figure 4 | Immunoblot analysis of PHYA stability in wild-type (Elm1) and mutant (elm1) plants. Immunoblot analysis of PHYA stability in wild-type (Elm1) and mutant (elm1) plants. A, Top, Immunodetection of PHYA apoprotein after SDS-PAGE and western blotting of crude protein extracts from wild-type (Elm1) and mutant (elm1) seedlings grown in dark (D) or given a 4-h R treatment (Red Shift; 10.0 mumol m-2 s-1). Lanes were loaded on an equivalent fresh weight basis and loading confirmed by Coomassie Blue stain (data not shown). Molecular masses (kD) were determined using prestained markers (see "Materials and Methods"). Bottom, Detection of bound PPhiB by zinc-induced fluorescence after gel staining with Zn(OAc)2. Lanes as above. B, Immunodetection of PHYA apoprotein in crude protein extracts of R shift (4 h, 10.0 mumol m-2 s-1) wild-type (Elm1) and mutant elm1 seedlings. Extracts of elm1 seedlings were diluted from 2-fold (1/2) to 100-fold to allow semiquantitative determination of relative PHYA levels. To confirm that elm1 is deficient in holophytochrome, zinc-induced fluorescence was used to visualize the covalently bound chromophore . In wild-type plants, the presence of bound chromophore was detected as a fluorescent band in zinc-stained SDS-PAGE gels that comigrated with PHYA (Fig. A). However, although elm1 plants accumulate high levels of PHYA, no bound chromophore could be detected in extracts of either etiolated or red-shifted seedlings. The absence of chromophore associated with PHYA in etiolated elm1 seedlings suggests that there is either a disruption in holoenzyme assembly or PPhiB synthesis in elm1 mutants. elm1 Mutants Are Unable to Convert BV IXalpha to 3Z-PPhiB | All phytochrome chromophore-deficient mutants characterized to date are blocked in one of two steps: heme to BV IXalpha or BV IXalpha to 3Z-PPhiB . To examine the accumulation of heme oxygenase in elm1, we used an antibody raised to the HO1 (HY1) protein of Arabidopsis . Etiolated elm1 seedlings had an identical level of immunodetectable HO1 protein as wild-type seedlings (data not shown). This result suggests that heme oxygenase is unaffected in elm1 seedlings and that elm1 may be deficient in PPhiB synthase. To directly assay PPhiB synthase activity in elm1, we isolated plastids from dark-grown wild-type and elm1 seedlings, incubated these with BV IXalpha and heme, and analyzed the products by HPLC. Figure shows that incubation of wild-type plastids with BV IXalpha results in the synthesis of two products, identified by their absorbance maxima and co-injection of authentic standards isolated from pea (data not shown), as 3Z-PPhiB (peak 2, trace B) and 3E-PPhiB (peak 3, trace B), respectively. In contrast, incubation of elm1 plastids with BV IXalpha produced no PPhiB peaks (trace C). The major peak under these conditions was confirmed as the substrate BV IXalpha by its absorbance maximum and co-injection of authentic BV IXalpha (data not shown). To confirm that the elm1 mutation specifically affects PPhiB synthase, we also assayed the same etioplast samples for heme oxygenase activity. Wild-type etioplasts were capable of converting heme to both BV IXalpha and 3Z-PPhiB. In contrast, incubation of heme with elm1 etioplasts resulted in a small and reproducible increase in BV IXalpha synthesis, but no synthesis of PPhiB (data not shown). These data strongly suggest that elm1 is specifically deficient in the enzyme PPhiB synthase. Figure 5 | HPLC analysis of BV metabolism by isolated plastids from wild-type (WT) and mutant (elm) seedlings. HPLC analysis of BV metabolism by isolated plastids from wild-type (WT) and mutant (elm) seedlings. HPLC analysis of BV metabolism by isolated plastids from wild-type (WT) and mutant (elm) seedlings. A, Control incubation with BV IXalpha and all reaction components except plastids. B and C, Bilin products obtained after incubation of WT or elm1 mutant plastids with BV IXalpha. Peaks identified as 1 BV IXalpha, 2 3Z-PPhiB, and 3 3E-PPhiB are indicated. Photosynthetic Transcripts Accumulate to Reduced Levels in elm1 Seedlings | To investigate the requirement of phytochrome for photosynthetic development, the levels of transcripts encoding several plastidic proteins were examined in elm1 seedlings. Wild-type and elm1 seedlings were grown in D, under two fluences of R and in W. RNA gel-blot analysis was used to assay the accumulation of transcripts encoded by Cab, RbcS, rbcL, and psbA genes . Cab and RbcS are nuclear transcripts encoding the light-harvesting Chl a/b protein (LHCPII) and the small subunit of Rubisco, respectively. The transcription of Cab and RbcS genes is regulated by both phytochrome and blue light-mediated signaling . rbcL and psbA are chloroplast genes encoding the large subunit of Rubisco and the D1 peptide of photosystem II, respectively. As observed with Cab and RbcS, transcripts encoded by rbcL and psbA accumulate to higher levels after illumination . Under D growth conditions, wild-type and elm1 mutants showed similar low-level accumulation of photosynthetic transcripts. As expected, all photosynthetic transcripts examined in wild-type seedlings accumulated to much higher levels under R relative to D growth conditions. In contrast, there was a relatively modest accumulation of all photosynthetic transcripts examined in elm1 seedlings grown under low- or high-R conditions. Under W, rbcL and psbA transcripts accumulated to similar levels in wild-type and elm1 mutant seedlings. RbcS transcripts accumulated to slightly lower levels in W-grown elm1 relative to wild type, but Cab transcripts were significantly lower in elm1 plants relative to W-grown wild-type seedlings. These data show that the R-mediated nuclear (Cab and RbcS) and plastid (rbcL and psbA) transcript accumulation is impaired in elm1 seedlings. Figure 6 | Northern-blot analysis of photosynthetic transcript accumulation in wild-type and elm1 seedlings. Northern-blot analysis of photosynthetic transcript accumulation in wild-type and elm1 seedlings. Total RNA was extracted from wild-type and elm1 mutant seedlings grown under D, low R (3 mumol m-2 s-1), high R (30 mumol m-2 s-1), or W (100 mumol m-2 s-1) light conditions. Filters were hybridized to gene-specific fragments for the Cab, RbcS, psbA, and rbcL genes as described in "Materials and Methods." Approximately 5 mug of total RNA was loaded per lane. The 26S ribosomal band was visualized by ethidium bromide staining and was used as a loading control. Approximate transcript sizes (kb) are shown on the left. DISCUSSION : The data presented here show that the elm1 mutant of maize is severely deficient in photoreversible phytochrome and responds only weakly to both R and FR irradiation. These results indicate that elm1 lacks multiple phytochromes and is consistent with a deficiency in phytochrome chromophore synthesis or assembly. Mutants disrupted in the synthesis of the phytochrome chromophore have been characterized in a number of species and are disrupted at one of two loci encoding heme oxygenase or PPhiB synthase. To date, no locus has been implicated in the regulation of chromophore synthetic enzymes, post-synthetic chromophore processing or holophytochrome assembly. Measurement of PPhiB synthesis in isolated etioplasts demonstrated that elm1 was unable to synthesize 3Z-PPhiB from BV IXalpha. This result, together with the retention of heme oxygenase protein and activity, suggest that elm1 is specifically deficient in PPhiB synthase. Therefore, the elm1 mutant is similar to the hy2 mutant of Arabidopsis , pcd2 of pea , and the au mutant of tomato , but is not equivalent to any known mutants from monocot species. In Arabidopsis, the HY2 gene has recently been cloned and shown to encode PPhiB synthase . Unfortunately, searches of public maize expressed sequence tag collections have failed to identify putative maize orthologs of HY2. This may be due to the relatively low abundance of HY2-like transcripts in maize or may reflect a high degree of sequence divergence between the maize and Arabidopsis gene sequences. The rapid progress being made in the sequencing of the rice genome could soon help to provide a bridge to the isolation of a maize HY2 ortholog and possibly to the cloning of Elm1 in maize. elm1 mutants are pale green both as seedlings and as mature field-grown plants, a phenotypic trait observed in most chromophore-deficient mutants . Investigation of greening in elm1 seedlings has demonstrated a reduction in the accumulation of transcripts encoding a number of chloroplast components. Under R, Cab, RbcS, rbcL, and psbA transcripts accumulated to reduced levels in elm1 seedlings. Under W, rbcL and psbA transcripts accumulated to similar levels in elm1 and wild-type seedlings, suggesting that R and possibly blue light-signaling systems function redundantly to promote the accumulation of chloroplast-encoded transcripts. In contrast to rbcL and psbA, Cab and RbcS transcripts did not accumulate to similar levels in wild-type and elm1 seedlings grown in W. This result suggests that the blue-light signal transduction pathway is unable to compensate for impaired phytochrome signaling in elm1 mutants. However, it does not exclude the possibility that the reduced levels of Cab and RbcS transcripts reflect an altered physiology directly or indirectly responsible for lower Chl and carotenoid levels in elm1 plants. Although non-plastidic photoreceptor systems may signal directly to chloroplasts, the probable targets of light regulation are nuclear factors required for the accumulation of plastid-encoded transcripts. The identification of nuclear-encoded plastid-localized RNA polymerase and plastid-targeted sigma factors has revealed the importance of transcriptional control in the regulation in plastid gene expression. The accumulation of both nuclear-encoded plastid-localized RNA polymerase and plastid-localized sigma factors is normally light regulated and, therefore, may be disrupted in elm1 plants. Furthermore, genetic analyses have revealed a number of nuclear loci required for the processing of chloroplast mRNAs and highlight the importance of posttranscriptional regulation of transcript abundance . Thus, although we observed dramatic decreases in levels of psbA and rbcL transcripts in elm1 mutants, it is unclear if this reflects a decreased rate of transcription, an increased rate of transcript degradation, or a combination of both. Although it is clear from the above discussion that deficiencies in light signaling are likely to be important in determining the pale-green phenotype of elm1, the analysis of chromophore-deficient mutants from other species has indicated that other factors may play a role . Inconsistencies between the degree of Chl deficiency and the level of functional phytochrome have led to the proposal that feedback inhibition of Chl synthesis results from the block in plastidic heme degradation . Consistent with this hypothesis, dark-grown au and yg-2 mutants have a reduced level of the Chl precursor protochlorophyllide and this phenomenon has also been observed in chromophore-deficient mutants of pea and Arabidopsis . The pale-green phenotype and reduced Chl accumulation in elm1 plants, therefore, may result in part from a similar negative feedback of Chl biosynthesis. Examination of protochlorophyllide levels in D-grown elm1 seedlings should resolve this issue. The elm1 mutant represents the first light-signaling lesion to be characterized in maize and presents evidence that phytochrome influences flowering time in this species. Flowering time in many species is modulated by the relative duration of light and darkness during a daily cycle (the photoperiod; ). In some plants, flowering is promoted by short days (SD), whereas in others, flowering is promoted by LD. Genetic analyses, notably of Arabidopsis (an LD plant), have indicated that both phytochrome and blue light-signaling pathways interact in the perception of photoperiod and in the regulation of flowering. As a generalization, phyA acts to promote flowering , whereas phyB is required to inhibit flowering . The blue light-sensing cryptochromes act to promote flowering, both in a phytochrome-independent manner and by antagonism of phyB signaling . Although accessions of maize grown in the United States are generally considered day neutral, the early flowering phenotype of elm1 shows that flowering is repressed under LD in the W22 inbred. The response is similar to that seen in the se5 mutant of rice (an SD plant), although the magnitude of the effect in rice is greater (under LD wild-type rice flowered after 100.8 +- 0.8 d, se5 flowered after 46.6 +- 0.6 d; ). The expansion of cultivated maize beyond semitropical regions of early domestication has required the selection of day-neutral lines from ancestrally SD stocks. Nevertheless, the elm1 mutant suggests that standard U.S. inbreds retain a weak response to photoperiod. The elm1 mutant, the se5 mutant of rice, the BMDR-1 mutant of barley, and the ma3R line of sorghum collectively show that the establishment of early flowering under nonpermissive photoperiods can be achieved by selection for reduced phytochrome signaling. The utility of loss-of-function alleles in components of the phytochrome pathway has made this an efficient target for past selection and an attractive candidate for future genetic modification. MATERIALS AND METHODS : Plant Material and Growth Conditions | Homozygous elm1 mutants and near-isogenic wild-type seed stocks were maintained in a standard maize (Zea mays) W22 inbred line. Seeds were surface sterilized for 15 min in a 10% (v/v) commercial bleach solution containing 0.1% (v/v) Tween 20, rinsed five times in deionized water, and imbibed overnight with shaking at room temperature. For mesocotyl measurements, seeds were grown in Rootrainers containing vermiculite. On d 10, mesocotyl length was measured to the nearest millimeter using digital calipers. All seedlings were grown under continuous light or in constant darkness at 28C in Percival Scientific (Boone, IA) model E-30LED light chambers with the exception of white-light treatments. LED light modules provided red and far-red light sources with narrow wavebands and peak emissions at 664 and 736 nm, respectively. White-light growth used a combination of incandescent and cool-white fluorescent lighting. Plants were grown for 10 d under light treatments before photography, RNA isolation, and mesocotyl measurements. Measurement of Photosynthetic Pigments | Wild-type and elm1 plants were grown for 2 weeks for 16 h in 180 mumol m-2 s-1 W and 8 h in D at 23C. Three leaf discs were taken from the third leaf of four different plants and Chl and carotenoids were extracted into 80% (v/v) acetone and quantified according to . Protein Gel-Blot Analysis | Seedlings were frozen in liquid nitrogen, ground to a fine powder, and suspended at 2 mL g-1 in extraction buffer (37.5% [v/v] ethylene glycol, 75 mm Tris-HCl [pH 8.3], 7.5 mm Na4EDTA, 15 mm NaS2O5, 0.11% [v/v] polyethylenimine, and 1.5 mm phenylmethylsulfonyl fluoride; ). Extracts were clarified by centrifugation at 3,000g for 30 min at 4C, fractionated by SDS-PAGE (7.5% [w/v] acrylamide gel) and transferred to nitrocellulose membrane (Schleicher & Schull, Keene, NH). Gels were loaded by mass of starting tissue and equal loading confirmed by Coomassie Blue staining. Samples were equivalent to approximately 50 mug of total protein as determined using the Bio-Rad DC protein assay antibody (Bio-Rad, Hercules, CA). PHYA protein was detected using the monoclonal antibody O73D , horseradish peroxidase-conjugated goat-anti-mouse secondary antibody and the Bio-Rad Opti-4CN substrate kit. To detect PPhiB, acrylamide gels were incubated for 2 h in 1 m zinc acetate and visualized under UV light . Molecular masses (kD) were determined using prestained markers (SeeBlue Plus 2, Invitrogen, Carlsbad, CA). Spectrophotometric Assay for Phytochrome | Phytochrome levels in wild-type and elm1 seedlings were assayed by in vivo spectroscopy as described previously . Seedlings were grown for 7 d in D at 25C and the top 1.5 cm of eight seedlings were used for each sample. Assays for PPhiB Synthesis | Maize seeds were sown in wet vermiculite (washed to remove fine particles before use), cold treated for 24h at 4C, and then grown in the dark at 23C for a further 8 to 10 d before etioplast isolation. Maize etioplasts were isolated as described previously and PPhiB synthesis from BV IXalpha was assayed essentially as described before for pea (Pisum sativum; ; ), but with the following modifications; PPhiB assays were performed in 1 mL of reaction buffer, 20 mm TES, 10 mm HEPES-NaOH (pH 7.7) containing 500 mm sorbitol, 1 mm phenylmethylsulfonyl fluoride, 2 mum leupeptin, and 0.5 mm dithiothreitol, and an NADPH regenerating system (1.2 mm NADP+, 10 mm Glc-6-phosphate, and 2.5 units mL-1 Glc-6-phosphate dehydrogenase), 3,000 U mL-1 catalase, 1 mm desferrioxamine, and 5 mm ascorbate. The reaction was initiated by the addition of 10 muL of BV IXalpha (Porphyrin Products Inc., Logan, UT) to give a final substrate concentration of 10 mum. Bilins were recovered and concentrated using a C18 cartridge (SepPak Plus, Waters Corporation, Milford, MA) as described previously . HPLC analysis was performed using an LC-10 system (Shimadzu Corp., Kyoto) running VP-5 software and using an SPD-M10A photodiode array detector. An LC-18 column (5 mum; 250 x 4.6 mm; Supelco UK, Poole, UK) was used with a mobile phase of acetone:ethanol:100 mm formic acid (25:65:10 [v/v]) at an isocratic flow rate of 1 mL min-1 . The photodiode array detector was used to monitor spectra between 300 and 800 nm over 60 min. RNA Gel-Blot Analysis | Seedling tissue was harvested at the base of the coleoptile and flash frozen in liquid nitrogen. Total RNA was extracted from approximately 1 g of tissue as previously described . Approximately 5 mug of total RNA was fractionated on 1.5% (w/v) agarose gels containing 6.8% (v/v) formaldehyde and photographs taken of ethidium bromide-stained gels to visualize ribosomal bands. RNA was transferred to GeneScreen Plus nylon membrane (NEN, Boston, MA) through capillary transfer in 20x SSC. Digoxygenin (DIG)-labeled DNA probes were synthesized using the PCR DIG Probe Synthesis kit (Roche, Indianapolis, IN) according to the manufacturer's recommendations, using T7 and T3 primers. Gene-specific fragments for RbcS (pJL12), rbcL (pJL12), psbA (pSD7), and Cab (LHCP1020) were described previously . Hybridizations were performed using the Roche DIG Easy Hyb solution. Hybridization buffers and conditions were according to the manufacturer's recommendation (DIG Easy Hyb, Roche). In brief, membranes were prehybridized in 15 to 25 mL of DIG Easy Hyb buffer (Roche) for 30 min to 1 h. The prehybridization buffer was removed and 5 mL of hybridization buffer was added to 3 muL of labeled probe. Hybridization was performed overnight at 43C. Membranes were washed twice in 2x SSC and 0.1% (w/v) SDS at room temperature for 5 min and twice in 0.1x SSC and 0.1% (w/v) SDS at 68C for 15 min each. Membranes were then washed in 1x maleic acid buffer (0.1 m maleic acid and 0.15 m NaCl, pH 7.5) for 3 min followed by 1 to 2 h of shaking in blocking solution (10% [w/v] casein, 0.1 m maleic, and 0.15 NaCl) acid before addition of 5 muL of Anti-DIG-AP Fab fragments (Roche). Blots were incubated for 30 min at room temperature with gentle shaking, washed for 15 min twice in 1x Washing buffer (3% [v/v] Tween 20, 0.1 m maleic acid, and 0.15 m NaCl). Blots were incubated with detection buffer (0.1 m Tris and 0.1 m NaCl, pH 9.5) for 3 min and placed in plastic sheet protectors. Excess liquid was removed and 2 mL of CDP-Star solution (20 muL of CDP-Star reagent + 2 mL of detection buffer) were applied directly to membranes. Blots were incubated for 5 min and exposed on an image station 440 CF (Eastman-Kodak, Rochester, NY). Eastman-Kodak 1D 3.5.4 Image Analysis software was used to determine relative transcript abundance. Backmatter: PMID- 12226497 TI - Isolation and Characterization of a Novel Ribosome-Inactivating Protein from Root Cultures of Pokeweed and Its Mechanism of Secretion from Roots AB - Ribosome-inactivating proteins are N-glycosidases that remove a specific adenine from the sarcin/ricin loop of the large rRNA, thus arresting protein synthesis at the translocation step. In the present study, a novel type I ribosome-inactivating protein, termed PAP-H, was purified from Agrobacterium rhizogenes-transformed hairy roots of pokeweed (Phytolacca americana). The protein was purified by anion- and cation-exchange chromatography. PAP-H has a molecular mass of 29.5 kD as detected by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and its isoelectric point was determined to be 7.8. Yeast (Saccharomyces cerevisiae) ribosomes incubated with PAP-H released the 360-nucleotide diagnostic fragment from the 26S rRNA upon aniline treatment, an indication of its ribosome-inactivating activity. Using immunofluorescence microscopy, PAP-H was found to be located in the cell walls of hairy roots and root border cells. PAP-H was determined to be constitutively secreted as part of the root exudates, with its secretion enhanced by a mechanism mediated by ethylene induction. Purified PAP-H did not show in vitro antifungal activity against soil-borne fungi. In contrast, root exudates containing PAP-H as well as additional chitinase, beta-1,3-glucanase, and protease activities did inhibit the growth of soil-borne fungi. We found that PAP-H depurinates fungal ribosomes in vitro and in vivo, suggesting an additive mechanism that enables PAP-H to penetrate fungal cells. Keywords: Introduction : Ribosome-inactivating proteins (RIPs) are widely distributed plant enzymes that inhibit protein synthesis by virtue of their N-glycosidic activity, selectively cleaving an adenine residue from a highly conserved and surface-exposed stem loop structure in the 28S rRNA . This cleavage prevents the binding of the EF-2/GTP complex, with the subsequent arrest of protein synthesis leading to autonomous cell death . RIPs are either enzymatically active single polypeptides (type I) or heterodimers (type II). A type II RIP consists of an A chain, functionally equivalent to a type I RIP, which is attached to a sugar-binding B chain (for review, see ; ; ). Besides RNA N-glycosidase activity, some RIPs have ribonuclease, DNase, DNA glycosylase, and apurinic/apyrimidic lyase activities . In addition, RIPs from Trichosanthes kirilowii cell cultures have been demonstrated to possess chitinase activity . Certain type I RIPs display a variety of antimicrobial activities, including antifungal, antibacterial , and broad-spectrum antiviral effects against different plant and animal viruses , including a human immunodeficiency virus . In addition, pokeweed (Phytolacca americana) antiviral protein (PAP) showed the inhibition of tumor cell growth , and the RIPs from Ricinus communis (ricin) and Saponaria officinalis (saporin) have been shown to possess insecticidal properties against Coleopteran species . Despite extensive enzymatic and antimicrobial characterization, the significance of RIPs for plant biology remains largely unknown. Pokeweed produces a suite of constitutive and induced RIPs in its leaves and seeds. For instance, PAP is a 29-kD constitutive RIP found in pokeweed leaves and localized in the cell wall matrix of leaf mesophyll cells . PAP II is a seasonal 30-kD RIP found in pokeweed leaves harvested in late summer , and PAP-S (29.8 kD) is expressed in seeds . Amino acid comparisons show 80% homology of PAP with PAP-S, and 33% homology of PAP with PAP II. Accordingly, PAP-S cross-reacts with PAP antibodies, but PAP II does not react with PAP antibodies . PAP is thought to play a defense role because it depurinates ribosomes from all organisms tested, and because its expression in transgenic tobacco (Nicotiana tabacum cv Samsan and Nicotiana benthamiana) plants leads to resistance against viral and fungal infection . However, a clear understanding of the role of PAP in pokeweed has not been achieved. To better understand the functional significance of RIPs in pokeweed, we developed a hairy root system for the expression and manipulation of constitutive RIPs. Hairy roots show stable expression of root-specific biosynthetic pathways, and thus have been used as an experimental system to study the biology and biochemistry of underground organs . Plant roots synthesize and store various macromolecules, including storage and defense-related proteins such as chitinase and beta-1,3-glucanase, to cope with pathogenic challenge . Here, we report the identification of a novel RIP, termed PAP-H, located in the cell walls of hairy roots and root border cells of pokeweed. PAP-H was also found to be constitutively secreted as part of the root exudates, and its secretion was enhanced by elicitation with ethylene. Hairy root exudates of pokeweed containing PAP-H and other defense-related proteins showed strong antifungal activity against fungi causing root rot. We have previously reported that in vitro root secretions of secondary metabolites and proteins compare with root secretions under natural settings , suggesting that pokeweed may secrete RIPs into the soil. This paper reveals a new mechanism by which roots are able to secrete RIPs into the rhizosphere to prevent pathogen infection. RESULTS : Development of Pokeweed Hairy Roots | The transformation of pokeweed with Agrobacterium rhizogenes American Type Culture Collection (ATCC) no. 15834 was accomplished. Several hairy root clones were established and selected based on growth and stability, and root cultures were established in 125-mL flasks as indicated in "Materials and Methods" (Fig. A). Pokeweed root cultures showed stable growth and phenotype, producing a substantial biomass yield. As shown in Figure B, pokeweed hairy roots showed a biphasic root growth until d 30, which contained two periods of exponential growth. Maximum tissue accumulation under these conditions was approximately 180 g fresh weight L-1 medium, representing about a 900-fold increase in biomass starting from a single root tip inoculum. Root intracellular (in organ) and extracellular (secreted) proteins that accumulated in the culture medium during the time course were also examined by SDS-PAGE followed by western blotting using a PAP-specific antibody (Fig. C). Cross-reactivities with the PAP antibody were found in both the intracellular and the extracellular protein fractions. In the intracellular protein fraction, PAP antibody cross-reactivity increased during very early stages of growth, and maximum protein accumulation occurred before the end of exponential root growth phase at approximately 20 d. PAP cross-reactivity developed in culture media (extracellular proteins) after d 8, and increased through the time course. Figure 1 | Establishment of pokeweed hairy roots, and time course of root growth and PAP-H accumulation. Establishment of pokeweed hairy roots, and time course of root growth and PAP-H accumulation. A, Developed hairy roots of pokeweed as described in "Materials and Methods." B, Growth curve of fresh and dry weight accumulation over 32 d. C, Western-blot analysis of intra- and extracellular PAP-H accumulation over 32 d. The same amount (0.2 g fresh weight) of hairy roots from each sample was harvested to extract total proteins, and 1 mL of media from each sample was collected to concentrate the root secreted proteins followed by trichloroacetic acid (TCA) precipitation as described in "Materials and Methods." Identification and Purification of RIP from Hairy Roots of Pokeweed | To ascertain whether PAP expressed in established hairy roots is similar to PAP isoforms produced in leaves, seeds, and roots of pokeweed, the protein profiles of these different organs were probed with a PAP antibody by western blotting. Total protein was extracted from 40-d-old hairy root cultures of pokeweed, and from pokeweed leaves, seeds, and roots (Fig. A). Western-blot analysis indicated slight differences in Mr among immunodetected proteins from different extracts. The immunoreactive band observed in hairy roots by western blotting was slightly larger in size than that of PAP (from leaves) but smaller than that of PAP-S (from seeds). It also differed in size from the immunoreactive bands of RIPs detected in storage roots, but was of a similar size to that of an RIP from primary roots. Based on these results, we concluded that RIP is produced in the transformed hairy roots of the pokeweed plant. Interestingly, RIP production was also detected as part of the root exudates secreted both in vitro from hairy root cultures and from whole-plant cultures. Figure 2 | Identification and purification of PAP-H. Identification and purification of PAP-H. A, Profiling of SDS-PAGE and western blot. Total proteins prepared from different organs of the pokeweed plant, and media where the hairy roots and whole pokeweed plant were grown were concentrated and run through 12.5% (w/v) SDS-PAGE. The antibody raised against PAP was used for western blotting. Approximately 25 mug of protein was loaded per lane. B, Cation-exchange chromatography of the total hairy root proteins from pokeweed. Total protein and samples of each fraction were run in 12.5% (w/v) SDS-PAGE, and western blot was performed using the PAP antibody. Approximately 5 mug of total, basic fractions, PAP-S and PAP, and 15 mug of acidic fraction were loaded per lane. To biochemically purify the RIP from hairy roots of pokeweed, total proteins were subjected to ion-exchange chromatography as outlined in "Materials and Methods." Because RIPs are mostly basic proteins, the unretained protein solution from the anion-exchange column was applied to a cation-exchange column (Sep-Pack, Waters, Milford, MA), and basic proteins were eluted with NaCl step gradients from 40 to 600 mm. A 29.5-kD protein corresponding to the RIP detected in hairy roots was resolved in the 80 mm fraction and analyzed by western blotting (Fig. B). The RIP was further resolved using UNO S1 cation-exchange column chromatography (Bio-Rad, Hercules, CA). The RIP was determined to have a molecular mass of 29.5 kD as determined by SDS-PAGE. Characterization and Enzymatic Activity of PAP-H | The N-terminal region of the RIP purified from hairy roots of pokeweed was sequenced and compared with those from PAP isoforms and other RIPs (Fig. A). The data showed that the N-terminal amino acid sequence of the hairy root RIP differed from those of all known PAP isoforms and other RIPs. The N-terminal regions of the hairy root RIP had 61% homology with PAP, and 56% homology with PAP-S. The N-terminal region of the hairy root RIP also showed a significant homology with PAP-alpha , as well as PIP 2 from P. insularis and PD-S2 from P. dioica . Importantly, the data indicated that highly conserved hydrophobic residues reported in the N-terminal region of all other RIPs, such as a Tyr-14 and Phe-17 were found in the N-terminal region of the hairy root RIP. Based on these results, we concluded that the RIP purified from transformed hairy roots of pokeweed is a novel type of PAP, and named it PAP-H. PAP-H was determined to be, unexpectedly, a neutral protein with a pI of 7.8 by isoelectric focusing (IEF)-PAGE (Fig. B), and amino acid composition analysis showed that the amino acid distribution of PAP-H was similar to that of other RIPs (Fig. C). Figure 3 | Characterization of PAP-H. Characterization of PAP-H. A, Comparison of the N-terminal sequences of PAP-H. PAP (GenBank accession no. ); PAP-R ; PAP-S ; PAP II ; PAP-alpha, antiviral protein precursor from pokeweed ; PIP 2, RIP 2 from Phytolacca insularis ; PD-S2, protein synthesis inhibitor from Phytolacca dioica ; ME1 and 2, Mirabilis expansa protein 1 and 2 ; MAP, Mirabilis jalapa antiviral protein ; SAP, saporin from S. officinalis . Shaded boxes represent two amino acids that are absolutely conserved in all RIPs . Bold characters indicate homology regions among RIPs purified from pokeweed plants. B, Determination of pI. SDS-PAGE of the purified PAP-H was performed as described in "Materials and Methods." pH range of the first dimension gel was pH 3 to 10 using Bio-Lyte ampholytes (Bio-Rad). C, Amino acid composition of PAP-H, PAP , PAP-R , PAP-S , PAP II , MAP , and ME 1 and ME 2 . The RNA N-glycosidase activity of PAP-H was tested using yeast (Saccharomyces cerevisiae) ribosomes as a substrate. When depurinated rRNAs are treated with aniline, cleavage occurs at the depurinated site and a small nucleotide fragment is released from the 26S rRNA . Yeast ribosomes were incubated with PAP-H and with RIPs from M. expansa (ME), S. officinalis (saporin), and R. communis (ricin). As shown in Figure , PAP-H dupurinated the rRNAs and released the 367-nucleotide fragment upon treatment with aniline . These results demonstrate the enzymatic activity of PAP-H as an RIP. Figure 4 | Enzymatic activity of PAP-H in vitro. Enzymatic activity of PAP-H in vitro. Ribosomes were isolated from yeast and incubated with PAP-H, ME, saporin, and ricin as described in "Materials and Methods." rRNAs were extracted, treated with aniline, separated on a 4.5% (w/v) urea-polyacrylamide gel, and stained with ethidium bromide. The presence (+) or absence (-) of aniline is denoted. The arrow shows the presence of the diagnostic 367-nucleotide cleavage product of rRNA. PAP-H Is Localized at the Cell Wall Matrix in the Hairy Roots and Root Border Cells of Pokeweed | We examined the location of PAP-H in hairy roots of pokeweed using fluorescent microscopy. As shown in Figure A, thin longitudinal-sectioned hairy roots were treated with an anti-PAP antibody and a fluorescein-labeled secondary antibody as described in "Materials and Methods." Green fluorescence, which indicates PAP-H presence as measured by the fluorescent tag-antibody interaction, was observed in the walls of every cell in the hairy roots. According to the broad density of fluorescence imaging along the entire circumference of the cell wall, PAP-H is most likely embedded in the cell wall matrix rather than bound to the cell wall (Fig. A). Interestingly, we found that PAP-H was also located in the cell walls of root border cells released from the root cap as the root grows (Fig. B). This observation suggested a possible secretion mechanism of PAP-H from the roots into the rhizosphere. Figure 5 | Localization of PAP-H in hairy roots of pokeweed. Localization of PAP-H in hairy roots of pokeweed. The tips of hairy root tissues were cut into thin sections and fixed as described in "Materials and Methods." Sections of cells were then stained with PAP antibody, followed by fluorescein-labeled secondary antibody reaction. Green fluorescence as observed in a fluorescent microscope indicates the presence of PAP-H in the cells. A, Longitudinal section of hairy roots. B, Longitudinal section of hairy roots showing root border cells (the arrows) near the root cap. Phase images were taken by light microscopy. Control root images not incubated with PAP antibody and fluorescein-labeled secondary antibody did not show fluorescent signal, indicating no root endogenous fluorescence (data not shown). Bars = 20 mum. cDNA Cloning of PAP-H | Total RNA was isolated from hairy roots of pokeweed using the RNeasy Plant Mini Kit (Qiagen USA, Valencia, CA; see "Materials and Methods"). To clone the cDNA corresponding to the PAP-H protein, two gene-specific primers (GSPs) were designed for RACE. The sequences of these primers were formulated based upon the partial amino acid sequence of the N terminus region of PAP-H, as well as the codon preferences of several previously cloned PAP-related genes from pokeweed. One outer primer covered amino acids 6 through 14, including the first absolutely conserved hydrophobic residue, Tyr-14 (Y), whereas a second primer was designed from amino acids 9 through 18, overlapping the outer primer and covering the two absolutely conserved hydrophobic residues, Tyr-14 (Y) and Phe-17 (F). A 968-bp fragment obtained by 3'-RACE was cloned into the pCR4Blunt-TOPO and pCR4TA-TOPO (Invitrogen, Carlsbad, CA), and its nucleotide sequence was determined as addressed in "Materials and Methods." A 3'-untranslated region of 51 bp with a polyadenylation site was found downstream from the stop codon. To obtain a 5' sequence of PAP-H, 5'-RACE was performed using two GSPs designed from the fragments produced by 3'-RACE. These primers were designed so that a full-length cDNA clone sequence could be assembled from sequences of overlapping 5'- and 3'-RACE amplification products. A GSP of 27 bp was designed; it corresponded to the DNA sequence located approximately 100 bp downstream from the GSPs used for 3'-RACE. A second GSP of 29 bp was designed, corresponding to the DNA sequence located approximately 200 bp downstream of the 3'-RACE GSPs. 5'-RACE was performed and amplification products were cloned and sequenced as described for the 3'-RACE procedures. After DNA sequencing of the multiple clones for each amplification product produced with the various GSPs, the PAP-H cDNA sequence was assembled using the Jellyfish 1.5 Gene Analysis software package and found to be 1,125 bp in length with an open reading frame of 1,074 bp (Fig. A; GenBank accession no. ). The deduced translated amino acid sequence indicated that PAP-H cDNA has a putative signal peptide sequence of 47 amino acids, and a coding sequence of 292 amino acid residues. The first 18 amino acid residues from the deduced mature PAP-H polypeptide were found to be identical to those obtained by N-terminal sequencing of the purified PAP-H protein (Fig. C). A polyadenylation signal (AATAAA) was found between 10 and 30 bases upstream of the polyadenylation site, as found in most plant and animal mRNA . Using the gene analysis software package mentioned above, the coding region of the PAP-H cDNA was found to be slightly longer than that of PAP (GenBank accession no. ), especially on the 5'-upstream region. When compared with the previously cloned PAP cDNA, PAP-H was found to contain an additional 25 amino acids upstream of the N-terminal sequence. Approximately 50% amino acid sequence homology was found to exist between PAP-H and PAP in the N-terminal extension region. Analysis also indicated that the C-terminal extension region of PAP-H might comprise 28 instead of 29 amino acids as found in PAP, with six amino acids differing between the two translated sequences. The cDNA sequence of the open reading frame region from PAP-H has 59.1% homology with PAP, 69.7% with PAP-S, 64.7% with PAP-alpha, and 52.1% with PAP II (GenBank accession nos. , , , and ). The analysis also shows that the predicted mature protein from PAP-H cDNA has 66.8% identity with PAP, 64.4% with PAP-S, 58.4% with PAP-alpha, and 36.0% with PAP II. Highly conserved hydrophobic residues reported by were found in the deduced amino acid sequences of the PAP-H cDNA (Tyr-14, Phe-17, Arg-22, Tyr-74, Tyr-125, Gly-143, Ala-166, Glu-178, Ala-179, Arg-181, Glu-207, and Trp-210), and the relatively well-conserved region (172AIQMVSEAARFKYI186) thought to be the active site of enzymatic activity of RIPs was also found in PAP-H (Fig. B). Figure 6 | cDNA and deduced amino acid sequences of PAP-H. cDNA and deduced amino acid sequences of PAP-H. A, The amino acid sequence of the mature protein (PAP-H) is in dark bold and the absolutely conserved amino acid residues among RIPs are in dark bold and underlined. The dark bold and italic characters represent amino- and carboxyl-terminal extensions. The polyadenylation sites are indicated in italics and underlined. The cDNA sequence of PAP is aligned along with PAP-H. B, Comparison of the mature proteins between PAP-H and PAP . Ethylene Enhances the Secretion of PAP-H into the Rhizosphere | Based on PAP-H cell wall localization and on the observation that PAP-H was secreted as part of the root exudates (Fig. A), we decided to study the exudation mechanism of PAP-H. Upon closer examination, we observed exudates resembling water drops that were secreted from the root hairs of pokeweed root cultures growing in solid culture media (Fig. A). The exudate was collected using a pipette, filtered to avoid collecting root border cells, and run on an SDS-PAGE gel (Fig. B). The exudate showed slight differences in protein patterns compared with the proteins expressed in hairy roots. To determine if PAP-H was also secreted as a part of the root exudates, the intracellular (in organ) and extracellular (in exudate) proteins of pokeweed hairy roots were probed with the PAP antibody (Fig. B). The antibody strongly cross-reacted with both intra- and extracellular proteins and showed the same band, PAP-H (29.5 kD), which indicated that PAP-H is secreted as part of the root exudates as well as with the root border cells. In accordance, in vitro-grown pokeweed plants growing in liquid media also released PAP-H, presumably through roots (Fig. A). Figure 7 | Identification and ethylene induction of extracellular PAP-H. Identification and ethylene induction of extracellular PAP-H. A, Water drop-shaped exudates secreted from hairy roots of pokeweed grown in solid culture media. B, SDS-PAGE and western blotting of total protein and exudates from hairy roots of pokeweed. The PAP antibody strongly cross-reacted with intra- and extracellular proteins and showed the same band, PAP-H (29.5 kD), indicating that PAP-H was secreted as part of the root exudates. Approximately 20 mug of protein was loaded per lane. C, Western-blot analysis of elicitation experimentation during one time course study. Hairy roots of pokeweed grown for 30 d were transferred into 50 mL of fresh Murashige and Skoog media, and treated with air (top), ethylene (middle), and air + methylcyclopropane (MCP; bottom) through 6 d. The medium from which the hairy roots were grown for 30 d was used as control. One milliliter of media from each sample was collected to concentrate the root secreted proteins followed by TCA precipitation as described in "Materials and Methods," and subsequently the concentrated protein was loaded in each lane. D, Ethylene concentration in the headspace of flasks during the course of each treatment. E, Western-blot analysis of intra- and extracellular PAP-H elicited by ethylene in the time course. One milliliter of media from each sample was collected, concentrated, and loaded as described above. To analyze the biological significance of PAP-H exudation, hairy root cultures of pokeweed were challenged with the stress-related chemical ethylene. As shown in Figure C, an enhanced induction of PAP-H was observed in the root exudates when the hairy roots were treated with ethylene. The secretion of PAP-H was induced shortly after beginning the ethylene treatment, and constantly increased through d 6 of ethylene treatment. On d 6, the secretion of PAP-H was increased up to 8-fold compared with a control, which consisted of roots secreting PAP-H for 30 d without any treatment. Air treatment showed no induction of PAP-H until 3 d after culture transfer, showing PAP-H secretion on d 4. In contrast, treatment with an ethylene inhibitor (MCP) suppressed the natural secretion of PAP-H. This result suggested that MCP blocked the binding of the endogenous ethylene and thus PAP-H secretion. The concentration of ethylene in the headspace of the flask was measured in replicates of all three treatments. Endogenous ethylene produced in the hairy roots treated with air began to increase in d 4, and reached 5 muL L-1 at d 6, a level similar to the concentration supplied in the ethylene treatment (Fig. D). The increase of endogenous ethylene concentration correlates to the increase of PAP-H production starting on d 4, as detected by western blotting. The result shown in Figure D indicated that MCP also blocked the production of endogenous ethylene. Thus, our results clearly show that PAP-H is secreted under natural conditions by a mechanism mediated by ethylene induction. Interestingly, the western-blot analysis of intracellular PAP-H indicates that its concentration inside the hairy roots slightly decreased upon ethylene treatment, which correlates to increasing PAP-H accumulation in the media (Fig. E). These results suggest that ethylene induces secretion of PAP-H rather than the novo production of this RIP. Taken together, these results suggest that ethylene induces secretion of PAP-H into the rhizosphere but does not induce production of PAP-H. Antifungal Activity of PAP-H | To explore the biological significance of PAP-H, we examined the N-glycosidase activity of PAP-H against fungal ribosomes. The depurination experiment described for PAP-H against yeast ribosomes was repeated using ribosomes isolated from Trichoderma reesei. After aniline treatment, a small fragment of about 360 nucleotides was detected by urea-acrylamide gel electrophoresis (Fig. A). This result indicates that PAP-H may be involved in inhibiting fungal growth by actively depurinating the fungal ribosomes. Based on this result, we tested the antifungal activity of PAP-H against an array of fungi (see "Materials and Methods"). However, PAP-H did not show fungal inhibitory activity in plate assays. Subsequently, we tested the total secreted proteins (exudates) from pokeweed hairy roots for fungal inhibition to check any possible antifungal effect with other pathogenesis-related (PR) proteins that may allow PAP-H entrance onto fungal cells. As shown in Figure B, 50 mug of total secreted proteins produced a significant inhibition of the growth of R. solani as well as of T. reesei (data not shown). These data indicate that PAP-H may participate in an active defense mechanism of roots against rhizosphere microbes. Thus, we studied whether fungal inhibition was due to the direct effect of PAP-H on fungal ribosomes. We cocultured R. solani with hairy roots of pokeweed that were grown for 2 weeks before coculture in Murashige and Skoog basal media (Fig. C, a). Fungal hyphae growing in the proximity of the roots were collected after a week of coculture and washed briefly with water to cleanse of medium (solids) attached to hyphae, and ribosomes were then isolated from collected fungal mycelium tissues. Interestingly, a small fragment appeared when the rRNA derived from these ribosomes was run on urea PAGE, indicating rRNA cleavage due to RIP depurination after RNA extraction and aniline treatment (Fig. C, b --d). This result demonstrated that PAP-H entered into the cytosolic region of fungal cells, and depurinated fungal ribosomes. Figure 8 | Antifungal activity of pokeweed hairy root exudates. Antifungal activity of pokeweed hairy root exudates. A, Depurination of T. reesei ribosomes in vitro. Ribosomes were isolated and incubated with PAP-H, ME, saporin, and ricin. The presence (+) or absence (-) of aniline is denoted. The arrow shows the presence of the diagnostic nucleotide cleavage product of rRNA. B, Radial growth inhibition assay of pokeweed hairy root exudates against Rhizoctonia solani. Twenty-five millimolar NaPO4 buffer, pH 7.5 (a), and filter-sterilized 50 mug of total root exudates (b) were applied to the discs, and tested for antifungal activity. C, Enzymatic activity of exudated PAP-H from hairy roots against R. solani. R. solani was cocultured with hairy roots of pokeweed that were grown for 2 weeks before the coculture in Murashige and Skoog basal media (a), and ribosomes of fungal hyphae growing in the proximity of the roots were isolated after a week of coculture. Then, rRNA was isolated from both ribosomes followed by treatment with aniline (b), and compared with ribosomes isolated from normally grown R. solani, incubated with PAP-H and then treated with aniline (c). Fungal hyphae of R. solani growing alone in a petri dish were collected, and their ribosomes were isolated and examined for no natural occurrence of rRNA depurination in fungal cells without prior incubation with PAP-H and/or the root exudates (d). The presence (+) or absence (-) of aniline is denoted. The arrow shows the presence of the diagnostic nucleotide cleavage product of rRNA. D, Enzymatic activities of chitinase and beta-1,3-glucanase in root exudates. For each activity, triplicate samples were assayed at two different dilutions with each assay run in triplicate. Error bars indicate +-sd. E, Determination of proteolytic activity of root exudates. Zymogram gel electrophoresis containing gelatin was performed as described in "Materials and Methods." The gel was then stained with 0.5% (w/v) Coomassie Brilliant Blue R-250. The arrow indicates protease activity in root exudates. Enzymatic Activity Assay of PR Proteins in Root Exudates of Pokeweed | Based on previous results, we hypothesized that root exudates of pokeweed contained other PR proteins that could facilitate the entry of PAP-H into fungal cells by lysing fungal cell walls. Chitinase, beta-1,3-glucanase, and protease enzymatic activities were tested as possible candidates as described in "Materials and Methods." As shown in Figure , D and E, the root exudates demonstrated chitinase and beta-1,3-glucanase (10.18 and 16.82 units mg-1 total protein) activities, as well as protease activities detected by zymogram gel analysis. Correlating the concentration of chitinase and beta-1,3-glucanase in pokeweed root exudates with previously reported enzymatic activity for these enzymes in other plants (approximately 14 units mg-1 for chitinase and 18.5 units mg-1 for beta-1,3-glucanase; ; ; ; ; ), we conclude that the concentration of these PR proteins in the root exudates is sufficient to damage fungal cell walls. These results suggest that these PR proteins in root exudates may generate an additive effect to facilitate entrance of PAP-H into fungal cells. DISCUSSION : We established transformed root clones of pokeweed with A. rhizogenes ATCC number 15834. These transformed roots ("hairy roots") displayed similar morphological characteristics to primary roots, including a well-developed cortex with new tissues developing from an apical meristem. The hairy roots of pokeweed showed stable and fast growth, and produced PAP-H steadily and constitutively. Hairy roots of pokeweed grew rapidly after d 20, and reached the stationary stage in approximately 30 d. The production of intracellular PAP-H increased during the very early stage, and remained constitutive after 15 d. In contrast, extracellular PAP-H accumulated to detectable levels in approximately 8 d, and continuously increased with the growth of roots. These results indicate that extracellular PAP-H accumulation occurs over time along with root growth, whereas the expression of intracellular PAP-H is constant. In this communication, we report the isolation of PAP-H, a new constitutively produced RIP, from A. rhizogenes-transformed hairy roots of pokeweed. The pI of PAP-H was determined to be 7.8 by IEF-PAGE, suggesting that PAP-H is the first neutral and active protein among known RIPs (Fig. B). As shown in Figure C, the lower levels of basic amino acids such as Lys and Arg in PAP-H compared with the levels in PAP and other isoforms may contribute to its lower pI. PAP-H showed N-terminal amino acid sequence similarity with other RIPs found in pokeweed, such as PAP and PAP-S, and to a lesser extent with PAP-II . PAP-H cross-reacts with PAP antibodies, but does not react with the PAP-II antibody (data not shown). Amino acid comparisons show 75.8% homology between PAP and PAP-S, and 33% homology between PAP and PAP II. In accordance, PAP-S cross-reacts with PAP antibodies, but PAP II does not react with PAP antibodies . We suggest that because PAP, PAP-S, and PAP-H are constitutively produced, they share more sequence and antigenic specificity, in contrast to PAP II, in which expression is environmentally regulated. Comparing PAP isoforms detected among different types of roots (Fig. A), PAP-H showed size differences with RIPs found in storage roots. However, PAP-H has an Mr similar to the protein expressed in primary roots, which cross-reacted with the PAP antibody. A. rhizogenes-transformed hairy roots, in general, exhibit morphological and genetic similarity with fibrous primary roots (for review, see ; ). have shown that secondary growth induction leading to storage root formation shifts the protein production and accumulation patterns of RIPs. Therefore, PAP-H may be considered a biochemically identical isoform to the protein found in the primary roots. Our results also showed that storage roots have two immunoreactive bands that indicate a molecular mass (30 kD) close to that of PAP-S isolated from seeds . Furthermore, hairy root and whole-plant root exudates cross-reacted with the PAP antibody, and the immunoreactive band produced by these exudates showed the same Mr as PAP-H, suggesting that PAP-H is released from roots to the rhizosphere as a part of the root exudates (Figs. A and B). The amino acid sequence of the PAP-H cDNA was deduced and found to be identical to the protein sequence corresponding to the N terminus of the purified PAP-H protein (Figs. C and , A and B). Mature PAP-H contains 264 amino acid residues, which is similar to the 262 and 261 amino acid residues found in PAP and PAP-S . Although the numbers of amino acid residues from mature proteins are similar to each other, molecular weights of PAP-H, PAP, and PAP-S clearly showed differences (Fig. A), suggesting that posttranslational modification may occur in pokeweed RIPs. Recently, a number of mutagenesis studies using PAP have revealed certain specific sites important for its enzymatic activities. Transgenic plants expressing nontoxic mutated PAP forms such as PAPx (active site mutant), PAPn (N-terminal mutant), and PAPc (C-terminal mutant) have shown that Glu-176, Gly-75, and 25 C-terminal amino acids are critical amino acid residues involved in ribosome depurination . Comparison of cDNA sequences of PAP-H and PAP shows that PAP-H also contains those amino acid residues involved in ribosome depurination. The deduced amino acid sequences of PAP-H from its cDNA show Glu-178 and Gly-77 corresponding to Glu-176 and Gly-75 from PAP; the latter amino acids are responsible for the N-glycosidase activity of PAP. PAP-H also has high sequence homology with PAP in 25 C terminus amino acid residues. Furthermore, the PAP active site residues (Tyr-72, Tyr-123, and Arg-179) directly participated in the catalytic deadenylation of RNA, and Trp-208 involved in the stabilization of ribosome binding may correspond to Tyr-74, Tyr-125, Arg-181, and Trp-210 from PAP-H. These sequence analyses indicate that PAP-H shares with PAP amino acid residues involved in cytotoxicity, so that the N-glycosidic activity of PAP-H may be similar to that of PAP. Interestingly, the cDNA analysis demonstrates that PAP-H contains 75 amino acid residues corresponding to the N- and C-terminal extensions (47 from N terminus and 28 from C terminus). The sequence comparison of both amino and carboxy extension regions from PAP-H and PAP suggest that these regions from PAP-H may function as cell wall-targeting sequences in hairy roots as was previously reported for PAP . However, the N-terminal extension of PAP-H contains an additional 25 amino acids compared with PAP and PAP-S, and the sequences of these two proteins differ from each other. Therefore, the N terminus sequence of PAP-H may include additional functions such as directing secretion of PAP-H to the rhizosphere under ethylene regulation. We found that PAP-H is located in the cell wall matrix of hairy roots (Fig. A), similar to PAP being localized in the cell wall matrix of leaf mesophyll cells . Because PAP is an exported protein, it may have an endoplasmic reticulum (ER) signal sequence at the N terminus to enter the secretory pathway (for review, see ). As described above, PAP-H contains potentially functional N-terminal extension residues, and shows localization similar to that of PAP. In accordance, PAP-H may share the same processing mechanisms as PAP. Interestingly, PAP-H was also found to be localized in the cell walls of root border cells and released from the root tip as the root grew (Fig. B). Root border cells are considered to be one of the key factors in root-microbe communication because they are programmed to be detached from the root and enter the rhizosphere (for review, see ). Border cells are involved in triggering various responses by producing secondary metabolites, chemo-attractants, repellents, and signals that can lead to agglutination against infesting microbes and root parasites. However, no specific enzyme has been isolated from root border cells. Recently, Hawes and coworkers have suggested the production of a low-pH galactosidase in the root border cells of the pea (Pisum sativum; ). Our data showed that pokeweed root border cells produce and store RIPs. As mentioned above, we also found that hairy roots of pokeweed constitutively secreted water drop-shaped exudates while growing in solid media (Fig. A). Western-blot analysis (Fig. B) of filtered exudates collected from hairy roots indicated that PAP-H was released as part of the root exudates. To our knowledge, PAP-H is the first RIP shown to be secreted in the root exudates, as well as the first shown to be compartmentalized in secreted root border cells. These root-specific functions may act as potential plant defense mechanism against pathogen infection. PAP is regarded as a defense-related protein because it can deadenylate ribosomes from all organisms, and its expression in transgenic plants leads to resistance to viral and fungal infection . Although many elicitors and signals that accompany pathogen recognition and defense responses of plants have been found to induce PR proteins, only limited information has been determined about the induction of RIPs. Some studies have addressed the induction of RIPs caused by stress and stress-related compounds such as jasmonate , osmotic stress or heat shock , and salt shock . reported that methyl jasmonate also rapidly induces and accumulates RIP in leaf tissues of barley (Hordeum vulgare) at the transcriptional level. Ethylene is induced in response to environmental stress, including infection by pathogens (for review, see ). Ethylene has also been shown to be an important factor in plant cell culture systems, affecting the stimulation of secondary metabolites from various tissue and cell cultures . Thus, ethylene is biologically active at a very low concentration ---less than 1 muL L-1 ---and its production is temporarily increased severalfold within 25 to 30 min when tissues are wounded or mechanically perturbed . Exogenous application of ethylene has been shown to induce the transcription of several genes encoding basic-type PR proteins such as class I basic chitinases and class I beta-1,3-glucanase (for review, see ). have reported recently that the ethylene receptor-like protein ETR1 is located in the ER membrane in Arabidopsis. Ethylene receptors such as ETR1 are two-component signaling systems containing a receptor with His kinase activity and a response regulator . In response to ethylene, the receptor is autophosphorylated and a phosphate is then transferred to the response regulator, which mediates downstream responses . Because PAP-H showed strong secretory induction by ethylene treatment and is assumed to be secreted through the ER, the ethylene receptor located in the ER membrane may be involved in activating the secretory mechanism of PAP-H. In conclusion, we feel that ethylene induced by biological stresses, such as pathogen attack, binds to the receptor protein at the ER membrane, initiating a downstream signaling cascade to activate the PAP-H secretory pathway as a defense mechanism. As part of a secretory root defense mechanism, PAP-H creates an antifungal scenario with the aid of other PR proteins secreted from hairy roots of pokeweed. As shown in Figure A, PAP-H has in vitro N-glycosidase activity against fungal ribosomes, which indicates that PAP-H can recognize and depurinate fungal ribosomes; however, purified PAP-H did not show in vitro antifungal activity against R. solani. In contrast, total root exudates of pokeweed showed inhibitory activity against fungi (Fig. B). Upon close examination, the ribosomes isolated from R. solani grown in the presence of hairy roots of pokeweed (Fig. C, a) showed depurination traces (Fig. C, c). We hypothesize that PAP-H depurinates fungal rRNA and inhibits fungal growth by disrupting protein synthesis in fungal cells. The penetration of PAP-H into the fungal cells may be facilitated by PR proteins such as chitinase, beta-1,3-glucanase, and proteases (Fig. , D and E). Both chitinase and beta-1,3-glucanase are widely distributed enzymes in higher plants, and have been hypothesized to function synergistically in plant defense against fungal pathogens (; for review, see ; ). We suggest that PAP-H is constitutively produced as a host defense protein, and secreted into the rhizosphere as a barrier against soil-borne microbe infection. The secretion of PAP-H is regulated in response to endo- and exogenous stresses. Our results indicate that PAP-H may prevent the multiplication of pathogens in the soil, acting in an additive effect with other PR proteins. Another hypothesis is that PAP-H is constitutively secreted to produce organic matter in the soil by depurinating the ribosomes of soil microorganisms. These and other hypotheses are currently being tested to shed light on the function of RIPs in the rhizosphere. MATERIALS AND METHODS : Plant Material | Seeds of pokeweed (Phytolacca americana) were collected in New Brunswick (NJ). Seeds were washed five times with sterile water and germinated on filter papers on a petri dish. The seeds were then transferred to pots and placed in the greenhouse. Establishment of Agrobacterium rhizogenes-Transformed Hairy Roots of Pokeweed | Shoots from pokeweed were collected from greenhouse-grown plants and surface sterilized with 10% (v/v) commercial bleach for 15 min and then washed four times with sterile water. Shoot cultures were placed separately in Magenta GA-7 vessels containing Murashige and Skoog basal medium solidified with 0.3% (w/v) Phytagel (Sigma, St. Louis). Cultures were kept in a light chamber maintained at 24C with a light intensity of 100 mumol m-2 s-1 PAR. To produce hairy root cultures, 1-month-old in vitro plants were infected with a 3-d-old culture of A. rhizogenes (ATCC no. 15834) grown in TY medium (0.8% [w/v] tryptone, 0.5% [w/v] yeast extract, and 0.25% [w/v] NaCl) at 30C . In brief, stems of pokeweed were punctured in several places with A. rhizogenes and then placed in a light chamber. Roots that developed at the infection sites were transferred to petri dishes containing solid Murashige and Skoog medium supplemented with 250 mug mL-1 Claforan (Hoechst-Roussel Pharmaceuticals, Somerville, NJ) and kept in the dark chamber at 24C. After 14 d, 1-cm root tips were subcultured twice to eradicate excess bacteria before transferring them to fresh medium in the absence of antibiotic. Clonal root lines established after serial transfers of root tips to fresh Murashige and Skoog medium were subcultured into 125-mL Erlenmeyer flasks containing 50 mL of liquid Murashige and Skoog medium and placed on a gyratory shaker set at 90 rpm in a dark chamber. Protein Extraction from Hairy Roots | Hairy roots of pokeweed were immersed in liquid N2 and ground to a powder using a mortar and pestle. Ground root tissue was dissolved into two volumes of extraction buffer (25 mm NaPO4 [pH 7.0] with 250 mm NaCl, 10 mm EDTA, 5 mm dithiothreitol, 1 mm phenylmethylsulfonyl fluoride, and 1.5% [w/v] polyvinylpolypyrrolidone), homogenized, and centrifuged for 30 min at 10,000g. The supernatant was brought to 20% (w/v) ammonium sulfate, and centrifuged again for 20 min at 10,000g. The supernatant was dialyzed against 20 mm NaPO4 buffer (pH 7.0) until it was free from sulfate ion. All extraction procedures were conducted at 4C, and the crude extract was stored at 4C until use. Electrophoresis and Western-Blot Analysis | SDS-PAGE gel electrophoresis was performed with 12.5% (w/v) acrylamide discontinuous gels using an electrophoresis cell (Mini-Protein 3 Cell, Bio-Rad) according to manufacturer's instructions. Low-molecular mass protein markers (21.1 --110 kD, Bio-Rad) were run simultaneously for each electrophoresis gel. The gel was stained with Coomassie Brilliant Blue R-250 (EM Science, Gibbstown, NJ). Proteins were electroblotted to Immun-Blot polyvinylidene difluoride membranes (Bio-Rad) using a Mini Trans-Blot Electrophoretic Transfer Cell (Bio-Rad). The blot was then probed with protein A-purified polyclonal rabbit anti-PAP antibody obtained from Dr. Nilgun Tumer (Rutgers University, New Brunswick, NJ), and the membranes were developed using an Opti-4CN Detection Kit (Bio-Rad), following the manufacturer's instructions. An antiserum titer of 1:1,000 (w/v) was used for all experiments. Chromatography and IEF | Ion-exchange separation was performed using Sep-Pack Plus Cartridges (Waters). Waters Accell Plus QMAs were equilibrated with 20 mm NaPO4 buffer (pH 7.0), and the flow-through solution containing unretained (basic) proteins was collected. Basic proteins were applied to Waters Accell Plus CMs equilibrated with 20 mm NaPO4 buffer (pH 7.0) and unbound (acidic) proteins were collected. Basic proteins were eluted with step gradients from 40 to 600 mm of NaCl. The target protein was resolved in the 80 mm fraction. The 80 mm fraction was dialyzed against 20 mm NaPO4 buffer (pH 7.0) and subsequently separated by cation-exchange chromatography using a UNO S1 column (Bio-Rad). Diluted protein solutions and fractions were concentrated by ultrafiltration using a Stirred Ultrafiltration Cell 8050 (Millipore, Bedford, MA). Protein purity and peak size were confirmed by SDS-PAGE stained with Coomassie Brilliant Blue R-250 (EM Science) and Silver Stain Plus (Bio-Rad). Protein concentration was determined by the Bradford (1976) method using a protein assay kit (Bio-Rad). The pI of purified PAP-H was estimated by IEF using a Mini-Protean II 2-D Cell (Bio-Rad) with Bio-Lyte ampholytes (pH range 3 --10; Bio-Rad) following manufacturer's instructions. Second dimension was performed in an SDS-PAGE gel with 12.5% (w/v) acrylamide gel using a Mini-Protein 3 Cell (Bio-Rad), and stained with Silver Stain Plus (Bio-Rad). N-Terminal Sequencing and Amino Acid Analysis | The purified protein was N-terminally sequenced on a Precise Protein Sequencer System (Applied Biosystems, Foster City, CA) at the Macromolecular Resources Facility (Department of Biochemistry, Colorado State University). Amino acid analysis and composition were obtained by the Protein Structure Core Facility at the University of Nebraska (Lincoln). Isolation of Ribosomes and rRNA Depurination Assay | Yeast (Saccharomyces cerevisiae) strain YPH500 was grown in YPD medium (1% [w/v] yeast extract, 2% [w/v] peptone, and 2% [w/v] Glc), and fungi, Trichoderma reesei, and Rhizoctonia solani were grown in potato dextrose media. Yeast was then pelleted by centrifugation and mycelial tissue of T. reesei and R. solani were collected by vacuum filtration. To isolate ribosomes, 10 g of pelleted yeast and fungal hyphae was ground in a mortar with liquid N2, and dissolved in 100 mL of extraction buffer (200 mm KCl, 25 mm MgCl2, 25 mm EGTA, 200 mm Suc, and 25 mm beta-mercaptoethanol in 200 mm Tris-HCl [pH 9.0]). The supernatant, collected by centrifugation at 10,000g for 20 min at 4C, was pipetted onto a Suc cushion (1 m Suc, 20 mm KCL, and 5 mm MgCl2 in 25 mm Tris-HCl [pH 7.6]) in 70 Ti tubes (Beckman Instruments, Fullerton, CA), and centrifuged at 55,000 rpm for 4 h at 4C (L-70 Ultracentrifuge, Beckman Instruments). The pellets were resuspended in 25 mm Tris-HCl buffer (pH 7.6) with 25 mm KCl and 5 mm MgCl2, and stored at -80C. The depurination assay was conducted according to . In brief, ribosomes were resuspended in RIP buffer (167 mm KCl, 100 mm MgCl2, and 100 mm Tris-HCl [pH 7.2]) and incubated with RIPs at 30C for 30 min in a total volume of 100 muL. RNA incubated in the absence of RIPs served as a negative control. After incubation, the RIPs were removed from the mixture by phenol:chloroform extraction and the RNA was divided in half. One-half of the extracted RNA was incubated on ice for 30 min with 1 m aniline acetate (pH 4.5) and precipitated with ethanol. Both aniline-treated and untreated RNAs were subjected to electrophoresis in a 7 m urea/6% (w/v) polyacrylamide gel and stained with ethidium bromide. cDNA Cloning of PAP-H | Total RNA was isolated from hairy roots of pokeweed using the RNeasy Plant Mini Kit (Qiagen USA) according to manufacturer's instructions. cDNA was cloned by the RACE-PCR method using the SMART RACE cDNA Amplification Kit following the manufacturer's instructions (CLONTECH, Palo Alto, CA). To amplify the 3' end of the PAP-H cDNA(s), two GSPs (5'-CCT TCG ATG TTG GAA GTG CAA CCA TTA GC and 5'-GGA AGT GCA ACC ATT AGC AAG TAT ACC ACC) were designed based on the N-terminal amino acid sequence of the purified PAP-H protein and considering codon preferences in previously cloned PAP-related genes from pokeweed. PCR reactions were performed in a GeneAmp System 2400 (Applied Biosystems) as follows: 94C for 1 min; 30 cycles of 94C for 5 s, 68C for 10 s, and 72C for 3 min. The nucleotide sequences of the GSPs used for 5'-RACE (5'-TGG CAA CCA ATA GGA ATC CTG CCT CGG and 5'-TGG CAA CCA ATA GGA ATC CTG CCT CGG) were designed based on the nucleotide sequences derived from the 3'-RACE product. The second PCR was performed as follows: 32 cycles of 94C for 5 s, 60C for 10 s, and 72C for 3 min. Each PCR product was purified using Quantum Prep Gel Slice Kit (Bio-Rad), and cloned into pCR4Blunt-TOPO or pCR4TA-TOPO (Invitrogen) for sequencing. The DNA sequencing was performed using an ABI Prism 377 DNA Sequencer (Applied Biosystems) at the Macromolecular Resources Facility (Department of Biochemistry, Colorado State University). Microscopy | The tips of hairy root tissues were cut into 2-mm segments and fixed for 1 h at 4C with 1.5% (w/v) formaldehyde in Sorenson's phosphate buffer (SPB; 0.03 m sodium phosphate monobasic and 0.12 m sodium phosphate dibasic [pH 7.5]). Samples were placed into gelatin capsules containing 15% (w/v) gelatin solution, and overlaid with one or two drops of the gelatin solution by gentle pipetting. The gelatin capsules containing samples were incubated for 24 h at 4C to be polymerized. After the resins were completely polymerized, gelatin blocks were removed from the capsules, and sections were cut 30 to 50 mum thick in prechilled SPB using a Vibratome (Sorvall MT2-B, Kendro, Newtown, CT). The sections were placed on slides and allowed to adhere for a few seconds. A drop of 50 mm Gly was added and sections were incubated for approximately 10 min. After washing twice with SPB, the sections were incubated with labeling-blocking buffer (LBB; 10% [v/v] goat serum and 1% [v/v] Triton X) for 30 min at 4C, and incubated overnight with diluted PAP primary antibody (1:1,000 [w/v]) in LBB at 4C in dark. Negative controls were incubated with only LBB under the same conditions. After 17 h, the sections were rinsed with LBB four times for 10 min at 4C, and incubated with fluorescein goat anti-rabbit IgG (1:1,000 [v/v] dilution in LBB; Molecular Probes, Eugene, OR) for 4 h in dark at 4C. The sections were washed with LBB again and SPB was used as a final wash. The sections were then mounted with antifade reagent (SlowFade Antifade Kit; Molecular Probes) in glycerol/PBS and sealed, and fluorescent images were taken using a microscope (2000 EXII, JEOL, Tokyo). PAP antibodies were obtained from Dr. Nilgun Tumer (Rutgers University). Preparation of Proteins from Liquid Media | Root cultures grown in liquid media were vacuum filtered with a 0.8-mum cellulose nitrate membrane filter (Whatman, Maidstone, UK), and the media were supplementary filtered to avoid debris using a 0.22-mum Millex-GP syringe-driven filter (Millipore). The samples were precipitated with TCA according to the method of . In brief, to each 1 mL of sample containing approximately 5 to 100 mg of protein, 100 muL of Na-deoxycholate was added and incubated for 10 min at room temperature. Then, 100 muL of 72% (w/v) TCA was dispensed, mixed, and incubated on ice for 15 min and centrifuged for 10 min. The supernatants were immediately removed, and the pellets were washed three times with ice-cold acetone. Pellets were then dissolved in SDS-PAGE sample buffer . Elicitation Experiments and Ethylene Measurement | Hairy roots of pokeweed grown for 30 d were transferred into 50 mL of fresh Murashige and Skoog media, and treated with air, ethylene, and air + MCP (Biotechnologies for Horticulture, Burr Ridge, IL). Air treatment was performed using an air-permeable silicon cap, and ethylene treatment was conducted by feeding 10 muL L-1 ethylene gas into a flask connected to a gas-mixing apparatus at a total flow rate of 15 mL min-1; air was used as the balance gas. These flasks were connected by means of a second tubing to remove effluent gas outside of the incubator. MCP was added as powder at a final concentration of 20 muL L-1. Each treatment was incubated on a gyratory shaker in a dark chamber for a week and 2 mL of each culture media was collected daily for 6 d. Collected samples were stored at 4C and concentrated using TCA precipitation . Ethylene was measured on a model 5840A gas chromatography system (Hewlett-Packard, Palo Alto, CA) equipped with a flame ionization detector. A Porapak N (6 ft, 0.2-mm i.d., stainless steel, Alltech, Nicholasville, KY) column was used at 75C. The injection port and detector temperatures were 90C and 180C, respectively; the mobile gas was helium at 20 mL min-1; sample size was 0.1 mL. Antifungal Assay | Antifungal activity of purified and exudated proteins was determined by a radial growth inhibition assay adapted from the method of . Various fungal plugs were placed in the center of potato dextrose agar plates, and sterile paper discs were placed next to the fungal plugs. Fifty micrograms of each protein, which was sterilized using Ultrafree-MC Sterile (0.22-mum GV Durapore, Millipore), was pipetted onto the discs. The plates were then incubated in the dark at room temperature. Antifungal activity was observed as a crescent-shaped zone of inhibition at the mycelial front. The effect on fungal growth was expressed qualitatively, according to the procedure of . Enzyme Activity Determinations | A colorimetric assay for chitinase and beta-1,3-glucanase activities in root exudates, with chitinase azure and laminarin azure as substrates, was performed as described by . The reaction mixture contained 550 muL of water, 200 muL of substrate (4 mg mL-1 in 0.2 m sodium acetate buffer [pH 5.0]), and 50 muL of root exudates (1 mg mL-1). After incubating the mixtures at 37C for 5 min, 200 muL of 1 n HCl was added, placed on ice for 10 min, and centrifuged for 5 min at 12,000g. The resulting supernatants (900 muL) were measured spectrophotometrically at 550 nm. All assays were performed in triplicate, and blanks were prepared without the addition of root exudates during incubation. Enzymatic activities were calculated according to and expressed as international units per milligram protein. One international unit is defined as the amount of enzyme required to catalyze the formation of 1 nmol of product per minute. Zymogram Ready Gel was used for proteolytic activity. A 10% (v/v) SDS-PAGE gel containing gelatin (Bio-Rad) was used and electrophoresis was performed according to the manufacture's instructions. After electrophoretic separation, the gel was incubated at room temperature for 30 min in 100 mL of 2.5% (v/v) Triton X-100 with agitation, and incubated overnight at 37C in 100 mL of development buffer (50 mm Tris-base, 200 mm NaCl, 5 mm CaCl2 anhydrous, and 0.02% [v/v] Brij-35 [pH 7.5]). The gel was then stained with 0.5% (w/v) Coomassie Brilliant Blue R-250 (EM Science). Backmatter: PMID- 12226498 TI - Engineering Herbicide Metabolism in Tobacco and Arabidopsis with CYP76B1, a Cytochrome P450 Enzyme from Jerusalem Artichoke AB - The Jerusalem artichoke (Helianthus tuberosus) xenobiotic inducible cytochrome P450, CYP76B1, catalyzes rapid oxidative dealkylation of various phenylurea herbicides to yield nonphytotoxic metabolites. We have found that increased herbicide metabolism and tolerance can be achieved by ectopic constitutive expression of CYP76B1 in tobacco (Nicotiana tabacum) and Arabidopsis. Transformation with CYP76B1 conferred on tobacco and Arabidopsis a 20-fold increase in tolerance to linuron, a compound detoxified by a single dealkylation, and a 10-fold increase in tolerance to isoproturon or chlortoluron, which need successive catalytic steps for detoxification. Two constructs for expression of translational fusions of CYP76B1 with P450 reductase were prepared to test if they would yield even greater herbicide tolerance. Plants expressing these constructs had lower herbicide tolerance than CYP76B1 alone, which is apparently a consequence of reduced stability of the fusion proteins. In all cases, increased herbicide tolerance results from more extensive metabolism, as demonstrated with exogenously fed phenylurea. Beside increased herbicide tolerance, expression of CYP76B1 has no other visible phenotype in the transgenic plants. Our data indicate that CYP76B1 can function as a selectable marker for plant transformation, allowing efficient selection in vitro and in soil-grown plants. Plants expressing CYP76B1 may also be a potential tool for phytoremediation of contaminated sites. Keywords: Introduction : Engineering of herbicide tolerance in higher plants can be achieved in many ways: via introduction of an altered target protein that is insensitive to the herbicide, overexpression of a wild-type target, or modification of herbicide transport, compartmentation, or metabolism. Increasing metabolism may be the best strategy because the phytotoxic compound is chemically altered and there is no interference with primary metabolism and no residual herbicide remains in the plant. So far, most crops genetically modified for herbicide metabolism have been transformed with genes isolated from microorganisms ; however, plants themselves offer a wide choice of herbicide-detoxifying enzymes. The introduction of different plant genes or appropriate alterations in expression levels in crop plants could be considered as an accelerated adjunct to classical breeding techniques to engender gene transfer between plants. The genes for herbicide-detoxifying enzymes in higher plants are just starting to be characterized. Efforts are focused on multigene families like those of glutathione S-transferases or glycosyl transferases , and cytochrome P450 monooxygenases . The latter, which is by far the largest family of enzymatic proteins in higher plants (272 P450 genes are found in the diminutive genome of Arabidopsis), offers the widest resource in terms of diversity and possible substrate specificity . In vivo and in vitro experimentation has largely demonstrated the involvement of P450s in the metabolism of all major classes of herbicides and their contribution to herbicide selectivity and weed resistance . Numerous reports confirm that the xenobiotic metabolizing P450s are induced by chemicals such as drugs, metals, ethanol, herbicide safeners, or herbicides themselves . The latter observation led us to isolate and characterize cDNAs of P450s expressed upon plant treatment with xenobiotics , among them CYP76B1, which shows high transcriptional activation after treatment with MnCl2 or drugs such as aminopyrine or phenobarbital in Jerusalem artichoke (Helianthus tuberosus). Expression of CYP76B1 in yeast (Saccharomyces cerevisiae) revealed its ability to metabolize a broad range of exogenous molecules with a catalytic efficiency often comparable with that observed for P450-dependent metabolism of endogenous compounds . Among the xenobiotics tested, CYP76B1 was particularly effective at one or more N-demethylation reactions with phenylurea herbicides . The plant detoxification of these compounds is well understood , and early investigators have established that the demethyl derivative of compounds like diuron and monuron are less phytotoxic than the parent, and that the di-demethyl derivative is essentially nonphytotoxic . Similar results have been obtained with a variety of related compounds, including chlortoluron and isoproturon . The removal of the single N-methyl substituent of linuron results in the nearly complete elimination of phytotoxicity , similar to the effect of two-step detoxification with most other herbicides in this class. Figure 1 | CYP76B1 metabolism of phenylurea. CYP76B1 metabolism of phenylurea. Therefore, CYP76B1 is a good candidate gene to use for engineering herbicide resistance in sensitive crops or for increasing the bioremediation potential of some plants. We have tested the effects of its ectopic constitutive expression on phenylurea tolerance of tobacco (Nicotiana tabacum) and Arabidopsis. To optimize the activity of the recombinant P450 protein and to circumvent possible limiting input of reducing equivalents for CYP76B1, translational fusions of the CYP76B1 cDNA with that of a P450 reductase previously isolated from the same species were constructed. In the present paper, we compare the increased tolerance obtained in the two plant species with the different constructs. A very significant increase in tolerance was observed for several phenylureas, allowing in vitro selection of the transformants and also conferring resistance to formulated herbicide applied by foliar treatment to plants grown in soil. RESULTS : Construction of cDNAs Encoding CYP76B1-P450 Reductase Fusion Proteins | Recent data have suggested that proteins catalyzing the electron transfer to P450 enzymes may become rate limiting in planta. For example, high expression of a cytochrome b5 gene is needed for full activity of the flavonoid 3',5'-hydroxylase activity in the petunia (Petunia hybrida) flower . Furthermore, plants contain different P450 reductases, and a coordinate activation of a cinnamate 4-hydroxylase and a specific P450 reductase gene upon elicitor treatment, infection, or UV light irradiation has been observed in parsley (Petroselinum crispum; ). The availability of an electron donor is particularly likely to be a rate-limiting factor in the case of recombinant overexpression of a P450 terminal oxidase, and the electron drain caused by the recombinant enzyme may interfere with the activity of endogenous proteins with essential functions. For this reason, we constructed chimeric cDNAs coding for fusion proteins between CYP76B1 and a P450 reductase. Natural forms of such fusion proteins were found in bacteria and eukaryotes , and artificial chimera were shown to be catalytically active in yeast and Escherichia coli . Fusions of yeast P450 reductase and mammalian P450 enzymes have been used to engineer herbicide resistance in tobacco and potato (Solanum tuberosum; ; ). The source of the coexpressed reductase influences the stability of recombinant P450s and the efficiency of coupling of the electron transport to substrate oxidation . Thus, we chose to fuse the coding sequence of CYP76B1 to that of a P450 reductase previously isolated from the same plant (Jerusalem artichoke), HTR1 (; GenBank accession no. ). Two types of translational fusions were constructed . One included the full-length reductase sequence attached to the 3' end of CYP76B1 (76B1-HTR1). This fusion comprises the longest open reading frame of HTR1, including the first of several ATGs, codons for a 35-amino acid hydrophilic stretch predicted to form a loop, and those for a hydrophobic helical segment, which is predicted to direct membrane insertion. In the second fusion (76B1-HTR1Delta57), the first 57 codons for the first loop and the membrane-insertion segment were removed to obtain a direct fusion of the flavoprotein domain of the reductase to the P450 hemoprotein . In both cases, the fusion was achieved by PCR modification of the coding sequences so as to introduce a SalI site 3' and 5' of the P450 and reductase sequences, respectively. Figure 2 | Construction of the translational fusions between CYP76B1 and the Jerusalem artichoke P450 reductase HTR1. Construction of the translational fusions between CYP76B1 and the Jerusalem artichoke P450 reductase HTR1. Both sequences were modified by PCR with a proof-reading polymerase using the primers given in text. BamHI and SalI sites were added just at the 5' and 3' ends of the CYP76B1 sequence, respectively, the stop codon being removed. SalI and EcoRI sites were inserted 5' and 3' of HTR1 so as to generate two different translational fusions. A, The HTR1 sequence has four ATG codons (underlined) that may constitute possible translation starts. Arrows indicate the connecting points between the P450 and the reductase. Predicted structures: +, inside loop; -, outside loop; O, outside helix cap; X, central transmembrane helix segment; and I, inside helix cap. B, Our first construct contains the largest reading frame. In the second construct, all ATGs, the first loop, and the transmembrane helix were removed, the two globular domains of the proteins being joined by just a short loop structure. In both cases, the SalI site created a Val-Asp linker predicted to maintain a loop structure between the two proteins. Expression and Activity of the Fusion Proteins in Yeast | The reductase-P450 fusion proteins were first tested in yeast for expression and activity. The different constructs were inserted in a multicopy plasmid under the control of the Gal-inducible promoter GAL10-CYC1. This plasmid was transformed into W303-1B and WHT1, a yeast strain where the endogenous P450 reductase has been replaced with the Jerusalem artichoke P450 reductase HTR1 under the control of GAL10-CYC1. Immunoblot analysis of the microsomal fractions of Gal-induced recombinant yeast with anti-CYP76B1 and anti-Jerusalem artichoke P450 reductase polyclonal antibodies indicated that the fusion proteins were expressed at levels at least equivalent to the wild-type CYP76B1. However, multiple lower Mr bands were evident from both of them, suggesting that they are degradation products and that the fusion proteins are more labile in vivo than the CYP76B1 alone. Staining with anti-P450 reductase antibodies also revealed some reductase protein, not fused to P450, produced in the absence and presence of Gal from the full fusion (76B1-HTR1). This was particularly evident in the W303 --1B blots, where there was no crosshybridization with the endogenous reductase (CYP76B1 lanes), yet there was strong crosshybridization in the 76-HTR lanes. This was unexpected and could be due to the presence of a cryptic promoter in the coding sequence of CYP76B1 or in the 5' segment of HTR1. Figure 3 | Expression in yeast (WHT1) of CYP76B1 and its translational fusions with HTR1. Expression in yeast (WHT1) of CYP76B1 and its translational fusions with HTR1. The immunoblots (20 mug of protein lane-1) were revealed with rabbit polyclonal antibodies directed against His-tagged recombinant CYP76B1 or purified Jerusalem artichoke P450 reductases . -, Microsomes isolated from yeast grown on Glc; +, microsomes from yeast induced for 16 h on Gal; M, Mr markers. To verify if functional enzymes were produced, P450 absorbance, reductase, and monooxygenase activities were also measured. The results of this analysis confirmed the expression of reductase in the absence of induction from the 76B1-HTR1 construct, and, to a much lesser extent, from the 76B1-HTR1Delta57 (ATG-less) chimera. MROD, easy to monitor by fluorometry, was chosen as reporter monooxygenase activity . The full-length fusion 76B1-HTR1 showed no MROD activity. This was probably the result of impaired folding or of the instability of the P450 domain because no CO-binding hemoprotein was detected in the recombinant yeast microsomes, whereas reductase was active for cytochrome c reduction. In a converse manner, the truncated fusion 76B1-HTR1Delta57 was functional in yeast microsomes. Calculation of specific activities indicates that its turnover is similar or is slightly increased compared with that of CYP76B1 separately coexpressed with HTR1 in yeast microsomes. Table I | Expression and activity in yeast of CYP76B1, single or fused to the HTR1 reductase Herbicide Tolerance of Transformed Tobacco | For expression in tobacco, all constructs were cloned into a T-DNA binary vector under the control of a strong constitutive duplicated 35S promoter. T1 (primary) transformants were selected on kanamycin, and the number of integration loci in the host genome was determined by Southern blot. Plants with single insertions were self-pollinated, and subsequent generations were selected on kanamycin and diuron. The herbicide tolerance of the transformants was assayed at all stages of the process using different methods. Phenylurea herbicides inhibit photosynthesis by blocking electron transport in the photosystem II complex . Susceptible plant treatment leads to leaf chlorosis followed by complete necrosis, resulting in a dramatic dose-dependent reduction of growth. The tolerance of the T1 transformants was tested by aging leaf pieces on agar plates containing different concentrations of herbicide under strong light. Leaves of the transformants appeared resistant to bleaching by 0.5 mm linuron (Fig. A). Figure 4 | Increased herbicide tolerance of T1 and T2 CYP76B1 tobacco transformants. Increased herbicide tolerance of T1 and T2 CYP76B1 tobacco transformants. A, Leaf pieces were washed with sterile water and cut under sterile conditions before being aged under strong light for 10 d on Murashige and Skoog medium containing 0.5 mm linuron. B, Selection of the T2 on Murashige and Skoog medium containing 0.9 mum diuron. Growth of wild-type plants is strongly inhibited. Selection of the T2 and T3 generations was performed, and the segregation of the T3 progeny was checked by growing the seeds on sterile plates containing 0.9 mum diuron (Fig. B). Selection of the transformants on diuron or kanamycin was equally effective. T3 homozygous plants were then tested for herbicide tolerance after leaf treatment of soil-grown plants with commercial formulations of the phenylurea herbicides linuron, chlortoluron, and isoproturon. A dramatic reduction of plant growth was observed following treatment with herbicide concentrations around 10 mum (Figs. A and 6). The increase in phenylurea resistance of transgenic plants was evaluated by comparing concentrations leading to 50% (I50) inhibition of shoot growth . The highest increase in herbicide tolerance (around 20-fold) was obtained with linuron for the plants expressing CYP76B1 alone, not fused to any reductase . The CYP76B1-HTR1Delta57 fusion also conferred a significant tolerance (around 10-fold). CYP76B1-HTR1 had a reduced efficiency. The same general trend was observed when tolerance was assayed with chlortoluron or isoproturon, although the overall resistance factor was not as high as with linuron. Linuron was completely detoxified by a single demethylation, whereas chlortoluron and isoproturon needed to be doubly dealkylated to yield a nonphytotoxic metabolite . We propose that the need for two catalytic steps to achieve detoxification with chlortoluron and isoproturon results in the overall lower resistance factor with these compounds. Figure 5 | Comparison of the increase in phenylurea tolerance obtained with the different constructs. Comparison of the increase in phenylurea tolerance obtained with the different constructs. Tests were performed with 1-month-old homozygous T3 plants, leaves being treated twice at a 4-d interval with dilutions of commercial formulations of linuron. A, Tobacco plants 12 d after the first treatment. B, Arabidopsis plants 10 d after first treatment Table II | Compared growth inhibition by phenylureas of wild-type and transgenic tobacco and Arabidopsis Transgene Expression and Herbicide Metabolism in Transformed Tobacco | Data obtained with yeast microsomes indicated that the CYP76B1-HTR1Delta57 fusion had the highest monooxygenase activity when expressed in microsomes also containing wild-type (not fused) plant P450 reductase. This fusion did not provide the highest level of herbicide tolerance in tobacco, therefore, this discrepancy was further investigated. A microsomal fraction was prepared from leaves of the different transformants, and the CYP76B1 and reductase contents were estimated by immunoblot . The staining with antireductase antibodies revealed a much higher expression of the CYP76B1-HTR1Delta57 than of the CYP76B1-HTR1 construct in plant tissues. As in the case of recombinant yeast microsomes, expression of nonfused reductase was also observed. When the immunoblots were stained with anti-CYP76B1 antibodies, only CYP76B1 expressed alone was detected. As the fusion proteins were readily detected in yeast microsomes , a selective epitope masking in plant membranes, although possible, seems rather unlikely. Therefore, it suggests that the levels of expression or stability of the fusions in plants is much lower than that of the P450 expressed alone. The P450 segment of the protein seems particularly prone to misfolding or degradation by proteases. Figure 6 | Shoot growth inhibition by chlortoluron of CYP76B1-transformed and wild-type plants. Shoot growth inhibition by chlortoluron of CYP76B1-transformed and wild-type plants. A, Dose-response: shoot fresh weight (average +- se of three measurements) was determined 9 d after foliar herbicide treatment. B, Appearance of the control and 60 mum-treated plants. Figure 7 | Immunoblot analysis of the transgenic tobacco microsomes. Immunoblot analysis of the transgenic tobacco microsomes. Immunoblots were revealed with the same polyclonal sera as in Figure . Twenty-five micrograms of microsomal protein was loaded in each lane. No reliable measurement of cytochrome c reductase activity was possible in microsomes from light- or dark-grown plants, in vitro or in soil. In the same way, P450 content and CYP76B1-associated demethylase activity could not be measured in plant microsomes, probably due to contamination by chloroplast pigments and redox components, or to plant secondary metabolites such as phenols and alkaloids. To directly evaluate the metabolism of phenylurea herbicides in the transgenic plants, individual leaves of T3 homozygotes were excised and fed with radiolabeled chlortoluron or isoproturon. All transgenic lines showed a higher capacity to metabolize the herbicide than wild-type tobacco . The capacity to produce nonphytotoxic metabolites paralleled herbicide tolerance of the transformants: CYP76B1 > CYP76B1-HTR1Delta57 > CYP76B1-HTR1 > wild type. The products of CYP76B1-dependent metabolism of phenylurea were further converted by other tobacco enzymes into other more polar metabolites (e.g. glucosyl conjugates; see Fig. A). Wild-type plants more actively metabolized isoproturon than chlortoluron (Fig. , B and C). Therefore, the increase in herbicide metabolism of the recombinant plants was smaller for isoproturon than for chlortoluron. Figure 8 | Chlortoluron and isoproturon metabolism in wild-type and recombinant tobacco leaves. Chlortoluron and isoproturon metabolism in wild-type and recombinant tobacco leaves. Excised leaves petioles were fed with 0.84 kBq [phenyl-U-14C]chlortoluron or 1.55 kBq [phenyl-U-14C]isoproturon. After complete absorption, the herbicide solution was replaced with water and leaves were maintained under continuous illumination for 12 h. 14C-labeled parent herbicide and its metabolites were extracted in methanol and were quantified by thin-layer chromatography (TLC). Glc conjugates were identified after treatment of extracts with beta-glucosidase. A, TLC profiles of the chlortoluron metabolites: (1) CTU, (2) N-demethyl CTU (DM-CTU), (3) N-didemethyl CTU (DDM-CTU) + unknown metabolites, (4) unknown metabolites, and (5) glucosides. B, Proportion of residual glucosides and phytotoxic (CTU +DM-CTU) and nonphytotoxic (DDM-CTU + unknown metabolites) aglycones recovered after chlortoluron metabolism. C, Proportion of residual glucosides and phytotoxic (IPU + DM-PU) and nonphytotoxic (DDM-IPU + unknown metabolites) aglycones recovered after isoproturon metabolism. Herbicide Tolerance and Transgene Expression in Arabidopsis | Immunoblot of the microsomes from transgenic Arabidopsis gave essentially the same results as those obtained with tobacco (data not shown). DISCUSSION : Factors Influencing Herbicide Tolerance | To our knowledge, the present study constitutes the first systematic investigation on the level of herbicide tolerance in transgenic plants by the overexpression of a plant P450 enzyme. There are few reports on the engineering of herbicide metabolism with plant genes , but the rather scarce data available do not indicate how large increases in herbicide tolerance have been achieved. CYP76B1, a gene inducible by chemical stress isolated from Jerusalem artichoke, was previously shown to metabolize herbicides of the class of phenylurea with high turnover rates . We show here that its 35S promoter-driven expression confers a 3- to 20-fold increase in resistance to phenylurea on tobacco and Arabidopsis. The level of resistance that is obtained varies with the active compound used for treatment in each plant. In tobacco, for example, a 2- to 3-fold higher increase in tolerance is achieved against linuron compared with chlortoluron, and the increase in chlortoluron resistance is, in turn, 2- to 4-fold higher than the increase in tolerance to isoproturon. This reflects, in part, the affinity and turnover rates for each herbicide, but also depends on the number of catalytic steps needed for complete detoxification and on the initial level of tolerance of the transformed plant. CYP76B1 metabolizes phenylurea exclusively via N-dealkylation. Compounds with two N-methyl substituents (e.g. isoproturon, diuron, and chlortoluron) require two dealkylation steps to become nonphytotoxic. Compounds with N-methoxy and N-methyl substituents, such as linuron and related molecules, lose phytotoxicity with a single demethylation. In addition, their affinity for CYP76B1 was shown to be higher than that of related structures with two methyl substituents (turnover rates are not reported so far; ). This readily explains why CYP76-expressing plants are more tolerant to linuron than to chlortoluron and isoproturon. Wild-type tobacco plants are more tolerant to isoproturon than to chlortoluron and, in good agreement with this observation, metabolize isoproturon much faster than chlortoluron . Catalytic efficiency of chlortoluron metabolism by recombinant CYP76B1 is one order of magnitude higher than that of isoproturon . Thus, it is not surprising if transformed tobacco shows a higher increase in tolerance to chlortoluron than isoproturon. Thus, the increase in herbicide tolerance that will be obtained by ectopic expression of CYP76B1 is expected to vary with the phenylurea molecule. Impact of the P450 Reductase | The highest increase in herbicide resistance was obtained via plant transformation with just CYP76B1. Attempts to further optimize CYP76B1 efficiency in planta, using translational fusions with a homologous P450 reductase to overcome possible rate-limiting electron transfer, proved disappointing. The plant P450 reductase, HTR1, as many other plant P450 reductases, differs from animal and yeast enzymes in that it presents a longer hydrophilic segment with a predicted loop structure at its N terminus. We tested if the presence of this structure as a linker between the two intact proteins results in a functional chimera, with both domains anchored and sufficiently mobile in the membrane. We compared the efficiency of this fusion with another construct deleted of the N terminus of the reductase so as to remove the potentially labile linker between the two globular proteins. The complete fusion, CYP76B1-HTR1, was barely effective. The best fusion protein, CYP76B1-HTR1Delta57, was two to four times less efficient than CYP76B1 alone. There are two possible explanations for the lower activity of the fusion proteins: a low expression or stability of the fusions in planta or a poor intramolecular electron transfer compared with that between single CYP76B1 and the P450 reductases from the host plant. The second alternative is very difficult to verify with plant microsomes, but western-blot analysis of plant membranes and measurements of activities in recombinant yeast largely support the first hypothesis. As in previous reports on yeast or E. coli expression of mammalian or yeast fusion proteins , the deletion of the membrane-anchoring segment of the P450 reductase was required to obtain chimeric protein expressed at significant levels in yeast and plants. Almost no fusion protein was detected in plants transformed with CYP76B1-HTR1 , in agreement with the low metabolism and increased tolerance observed in the recombinant plants. In a converse manner, the higher levels of fusion protein detected in plants transformed with CYP76B1-HTR1Delta57 are also in agreement with the higher levels of metabolism and herbicide tolerance achieved in the transformants. Thus, fusion proteins seem to be prone to protease degradation, and only systematic optimization of the connecting segment between the two proteins could possibly lead to really stable and catalytically effective chimera . A large proportion of the reductase was recovered as single, nonfused protein in yeast and plant. This may indicate activity of a cryptic promoter upstream of the reductase sequence, but also misfolding and fast degradation of the P450 segment of the fused protein. Expression of single reductase in yeast transformed with the fusion constructs in the absence of Gal induction strongly supports the first hypothesis. In support of the second hypothesis, the fusion proteins, which are readily detected by anti-CYP76B1 antibodies in yeast microsomes , are not detected in microsomal fractions isolated from recombinant plants , which suggests a strong alteration in P450 conformation. Instability of the hemoprotein domain compared with the reductase is also apparent upon expression in yeast . Thus, it is possible that combined effects of a cryptic promoter and P450 degradation contribute to the accumulation of isolated reductase protein. An interesting consequence of this high reductase accumulation is that, a least in the case of the CYP76B1-HTR1 transformants, some increased herbicide metabolism might result from reductase-mediated activation of endogenous enzymes. Such a possibility will be further investigated by construction of reductase-overexpressing plants. Potential Applications | Ectopic expression of the CYP76B1 protein has no visible impact on plant growth and development. This indicates that its overexpression does not directly or indirectly lead to major perturbation of an essential pathway or to accumulation of a phytotoxic compound. The P450 reductase isoforms and their constitutive expression in the plants seem to ensure efficient, if not optimal, reduction of the recombinant P450 oxidase. Thus, it seems possible to use the CYP76B1 gene as a positive marker of plant selection. Our experiments indicate that selection is easily achieved in vitro by including herbicide in the growth medium or via leaf treatment of soil-grown plants with commercial phenylurea formulations. The only selection system using a P450 gene described so far is a negative selection system using a bacterial P450 gene that is also effective in vitro and in soil . CYP76B1 has the advantage over other positive selection systems in that it is a plant gene conferring resistance to a whole class of generic herbicides and also leading to their detoxification. It is interesting to note that phenylurea dealkylation by CYP76B1 in planta allows further metabolism, not only conjugation, by the plant enzymes. Thus, CYP76B1 is also a potential tool for phytoremediation of contaminated soils and wastewater, or containment of herbicide leaching into groundwater . MATERIALS AND METHODS : Chemicals | Umbelliferone was purchased from Sigma (St. Quentin-Fallavier, France) and 7-methoxyresorufin was purchased from Molecular Probes (Leiden, The Netherlands). Linuron, chlortoluron, and isoproturon were purchased from Promochem (Molsheim, France). [phenyl-U-14C]Chlortoluron (64 GBq mol-1), [phenyl-U-14C]isoproturon (344 GBq mol-1), and reference metabolites were generous gifts from Syngenta (Basel). The commercial formulations of herbicides used for plant treatments in soil were Afalon 50L (active ingredient: linuron, distributed by Bayer Crop Science, Lyon, France), and Chlorto-Stef GT and Iso-Stef GT (active ingredients: chlortoluron and isoproturon, respectively, distributed by Stefes, Senlis, France). Construction of the WHT1 Yeast Strain | The WHT1 yeast strain was constructed by replacing the gene of the endogenous P450 reductase of W303 --1B with the coding sequence of a P450 reductase from Jerusalem artichoke (Helianthus tuberosus), HTR1, under the control of the Gal-inducible promoter GAL10-CYC1. BamHI and EcoRI sites were generated 5' and 3' of the HTR1 coding sequence, respectively, using the PCR primers N-terminal (5'-cgggatccATGCAACCGGAAACCGTCG-3') and C-terminal (5'-ccggaattcTCACCAAACATCACGGAGGTATC-3'). The Pfu DNA polymerase-amplified fragment was inserted in the integrative plasmid pYeDP110 using these restriction sites. The plasmid was linearized with NotI for integration of HTR1 and GAL10-CYC1 at the locus of the yeast (Saccharomyces cerevisiae) reductase by homologous recombination. Construction of the Plant Expression Vectors | Three vectors were constructed for plant transformation: one with the CYP76B1 (accession no. Y0992) coding sequence alone, and the two others designed to direct expression of translational fusions of CYP76B1 and of the Jerusalem artichoke P450 reductase HTR1 (accession no. ). To optimize the fusion proteins, predictions of secondary structures and trans-membrane segments were performed using the online PSIpred prediction server . Junctions between P450 and reductase were chosen so that the reductase was predicted to start with a loop structure . The first construct, CYP76B1-HTR1, included a predicted membrane-spanning segment near the reductase N terminus. In the second construct, CYP76B1-HTR1Delta57, this segment and the upstream loop were deleted. For the construction of the fusions, the coding region of CYP76B1 was PCR amplified so as to create BamHI (5') and SalI (3') sites with the primers CYP76-2 (5'-cgcggatccATGGATTTTCTTATAATAGTGAGTAC-3') and CYP76-3 (5'-cgcgtcgacGTTCAATGGTATTGGAACAACACAC-3'). CYP76B1 was also inserted alone in the BamHI and SacI sites of the expression vector using CYP76-2 in conjunction with CYP76-4 (5'-cgcgagctcCTAGTTCAATGGTATTGGAACAACAC-3'). The primers used for amplification of HTR1 were designed to create SalI and SmaI sites, respectively, at the 5' and 3' ends of the PCR products: HTR1-1 (5'-cgcgtcgacATGCAACCGGAAACCGTCGAAACG-3'), HTR1-2 (5'-cgccccgggTCACCAAACATCACGGAGGTATC-3'), HTR1-3 (5'-cgcgtcgacAATAGATCGTCCGGTAAGAAGTCCG-3'). HTR1-1 with HTR1-2 amplified the complete coding region, and HTR1-2 with HTR1-3 produced the truncated form lacking the sequence coding for amino acids 1 through 57. The PCR mixtures contained 500 ng of template, 20 pm of both primers, and 0.2 mm dNTPs in a total volume of 50 muL. They were preheated for 2 min at 94C before the addition of 5 units of Pfu DNA polymerase (Stratagene, La Jolla, CA). After 3 min of additional heating at 94C, 30 cycles of amplification were carried out as follows: 1 min of denaturation at 94C, 2 min of annealing at 55C, and a 2-min extension at 72C. The reaction was completed by a 10-min extension at 72C, and an additional 20-min PCR extension step was carried on with Taq DNA polymerase (Roche Molecular Diagnostics, Basel) to add an extra 3'-terminal adenosine to both ends. PCR products were then cloned in pGEM-T plasmid (Promega, Madison, WI). To construct the fusions, after digestion with BamHI and SalI, the CYP76B1 fragment was subcloned into the plasmid pLacZi (Clontech, Palo Alto, CA). The resulting plasmid was digested with SalI and SmaI, and the HTR1 fragments, excised from pGEM by SalI and SmaI digestion, were inserted 3' from CYP76B1. The resulting fusions were digested with BamHI and SmaI and were cloned into pBDX. The single CYP76B1 sequence terminated by a stop codon, excised from pGEM with BamHI and SacI, was also directly subcloned into the corresponding sites of pBDX. The pBDX binary vector, derived from pBD515.3 , contains between the left and right T-DNA borders a kanamycin-resistance gene and a modified multiple cloning site flanked in 5' by a duplicated 35S promoter, and by a nopaline synthase terminator at the 3' end (H. Schaller and P. Crevenat, unpublished data). Yeast Expression | The fusions constructed in pBDX were excised by BamHI and EcoRI digestion, and were ligated in pYeDP60 previously digested with the same restriction enzymes. The CYP76B1 sequence previously cloned in pGEM-T was transferred into pYeDP60 using the BamHI and SacI restriction sites. The different constructs were used to transform the W303 --1B (wild-type) and WHT1 yeast strains. Yeast transformation was performed essentially according to . Yeast growth, induction, and preparation of yeast microsomes were carried out according to . Plant Transformation and Characterization of Transgenic Plants | PBDX and the constructs pBDX-CYP76B1, pBDX-CYP76B1-HTR1, and pBDX-CYP76-HTR1Delta57 were transferred from Escherichia coli XL1 Blue into Agrobacterium tumefaciens LB4404 by triparental mating with the helper plasmid pRK2013 . Five-week-old roots from Arabidopsis C24 were infected with the resulting A. tumefaciens LBA4404 strains, and T1 plants were regenerated as described by . Culture room conditions were 23C with a 16-h light/8-h dark cycle. Tobacco (Nicotiana tabacum var. Xanthi) leaves were transformed via A. tumefaciens as in . In vitro cultures were grown under a 16-h light/8-h dark cycle at 23C. Transformed T1 Arabidopsis and tobacco plants were selected in vitro for kanamycin resistance (on 50 and 100 mg L-1 kanamycin, respectively). The T2 segregation was analyzed in vitro on a selective medium containing kanamycin or 0.9 mum diuron. T3 homozygous transgenic plants were selected among the self-progeny of kanamycin- and diuron-resistant T2 plants on the same media. The segregation of the T3 self-progeny was checked by growing the seeds on a sterile medium containing 0.9 mum diuron. Biochemical Characterization of the Transformed Plants | Microsomes of recombinant plant leaves were prepared essentially according to , except that the ratio of plant material to polyvinylpolypyrrolidone was 1:1 (w/w). Quantification of microsomal protein was carried out using a protein assay (Bio-Rad, Hercules, CA). SDS-PAGE, electroblotting onto nitrocellulose (Hybond C; Amersham Biosciences, Piscataway, NJ), and immunostaining with rabbit polyclonal antibodies raised against recombinant CYP76B1 or purified Jerusalem artichoke reductases were performed as in . Herbicide metabolism and P450 reductase assays have been described in and in , respectively. Herbicide Treatments and Phytotoxicity Evaluation | Foliar treatments of soil-grown plants were performed at the three- to four-leaf stage in the case of tobacco and the six- to eight-rosette leaf stage for Arabidopsis. Plants were sprayed twice with 10 mL of water dilutions of commercial herbicides at 4-d intervals. Plant tolerance was assessed 9 to 12 d after the first treatment by measuring shoots wet weight. Data are means +- se of triplicate experiments. In Vivo Herbicide Metabolism | In vivo metabolism was assayed with excised leaves from the wild-type and transgenic tobacco plants fed with 0.84 kBq [phenyl-U-14C] chlortoluron or 1.55 kBq [phenyl-U-14C] isoproturon dissolved in water, as described by . Herbicide and metabolites were quantified by TLC analysis of methanol extracts. Metabolites were identified by comparison with reference compounds using the solvent system, hexane:chloroform:acetone:ethanol (8:8:4:1, v/v). Metabolites that did not display mobility upon TLC analysis were submitted to hydrolysis with beta-glucosidase, as described by , for quantification of Glc conjugates. Backmatter: PMID- 12226499 TI - The Altered Pattern of Amylose Accumulation in the Endosperm of Low-Amylose Barley Cultivars Is Attributable to a Single Mutant Allele of Granule-Bound Starch Synthase I with a Deletion in the 5'-Non-Coding Region AB - Reasons for the variable amylose content of endosperm starch from waxy cultivars of barley (Hordeum vulgare) were investigated. The mature grains of most such cultivars contain some amylose, although amounts are much lower than in wild-type cultivars. In these low-amylose cultivars, amylose synthesis starts relatively late in grain development. Starch granules in the outer cell layers of the endosperm contain more amylose than those in the center. This distribution corresponds to that of granule-bound starch synthase I (GBSSI), which is more severely reduced in amount in the center of the endosperm than in the outer cell layers, relative to wild-type cultivars. A second GBSSI in the barley plant, GBSSIb, is not detectable in the endosperm and cannot account for amylose synthesis in the low-amylose cultivars. The change in the expression of GBSSI in the endosperm of the low-amylose cultivars appears to be due to a 413-bp deletion of part of the promoter and 5'-untranslated region of the gene. Although these cultivars are of diverse geographical origin, all carry this same deletion, suggesting that the low-amylose cultivars have a common waxy ancestor. Records suggest a probable source in China, first recorded in the 16th century. Two further families of waxy cultivars have no detectable amylose in the endosperm starch. These amylose-free cultivars were selected in the 20th century from chemically mutagenized populations of wild-type barley. In both cases, 1-bp alterations in the GBSSI gene completely eliminate GBSSI activity. Keywords: Introduction : The aim of this work was to investigate the reported variation in the amylose content of starch from the endosperm of waxy mutants of barley (Hordeum vulgare). Amylose is synthesized by granule-bound starch synthase I (GBSSI), an isoform of starch synthase of approximately 60 kD. GBSSI is encoded at the Waxy loci in cereals. In most cereal species, waxy mutants lack any detectable amylose in the starch of the endosperm. The major exception is barley, in which waxy mutant cultivars are reported to have between 0% and 13% amylose in their starch. For most such low-amylose cultivars, endosperm starch is reported to contain between 0.4% and 9% amylose , but starch from a few cultivars has undetectable amylose (barley cv Yon M Kei, ; barley cv CDC Alamo [line SB94794], ). In this paper, the waxy mutants with detectable amylose will be referred to as low-amylose cultivars and those with undetectable amylose will be referred to as amylose-free. In one low-amylose barley line (SW7142-92), the residual amylose has been shown to be concentrated in the outer layer of cells of the endosperm. Starch in the cells in this subaleurone layer stained blue-black with iodine solution, whereas that in the remainder of the endosperm stained red. The amylose contents of starch from tissues dissected from the outer and innermost parts of the grains of this cultivar were 8.6% and 2.2%, respectively . Two possible explanations for the wide variation in amylose content of the starch of barley waxy mutants are suggested by recent studies of GBSSI. First, it has been shown that wheat (Triticum aestivum) possesses two isoforms of GBSSI with different spatial distributions in the plant. In developing wheat grains, one GBSSI isoform accounts for amylose synthesis in the endosperm and a second accounts for much of the amylose synthesis in the pericarp, aleurone, and embryo . The pea (Pisum sativum) plant also has two, differently expressed isoforms of GBSSI . It is likely that in barley, as in wheat, two isoforms of GBSSI are present and expressed in different tissues. However, if in barley the isoform expressed primarily in other parts of the plant was also expressed in the endosperm of barley in addition to the endosperm-specific isoform of GBSSI, then loss of either GBSSI could result in a low-amylose content. Loss of both forms of GBSSI from the endosperm would result in amylose-free starch. Second, several independently derived waxy cultivars of barley have been shown to possess identical deletions in a GBSSI gene expressed in the endosperm ---the only GBSSI gene thus far identified in barley. The deletion overlaps a TATA box, and reverse transcriptase (RT)-PCR failed to reveal any mRNA for GBSSI in endosperm of the mutant cultivars . However, the data do not rule out the possibility that the deletion drastically reduces but does not eliminate expression of the gene. Thus, the presence of amylose in the endosperms of the low-amylose cultivars of barley might be explained by the widespread occurrence of a mutation that reduces but does not entirely eliminate expression of the gene encoding endosperm GBSSI. To discover which, if either, of these explanations is correct, we have examined the occurrence and distribution of amylose and GBSSI protein(s), and the nature and expression of genes encoding GBSSI in the developing grains of low-amylose and amylose-free mutants of barley. RESULTS : The Distribution of Amylose in the Endosperm | Iodine-stained sections of developing grain revealed great variation between waxy cultivars in the proportion of endosperm cells that contained significant amounts of amylose. In the amylose-free barley cv Yon M Kei and cv CDC Alamo, we observed no blue-staining granules at any stage of development (Fig. , G --I). Barley cv Arizona Hulless Waxy, a parent of CDC Alamo, also had no blue-staining granules (Fig. J); therefore, we consider it to be an amylose-free cultivar. In the low-amylose barley cv Iyatomi Mochi, cv Waxy Oderbrucker, cv Waxy Hector, and line SB85750 (the other parent of CDC Alamo; ), no blue-staining granules were present in young endosperm (from grain up to about 20 mg fresh weight), but blue-staining granules appeared in outer cells of the endosperm during the later part of development (Fig. , A --F). In some of the low-amylose cultivars, blue-staining granules in the endosperm were largely confined to cells immediately adjacent to the groove (inside the basal endosperm transfer cell layer; ), whereas in others, blue-staining starch granules were present in cells at the outer edge of the endosperm all around the grain (Fig. , A --D). In most cases, blue-staining granules were not confined to a single layer of cells. There was a gradation from the outer edge of the endosperm of cells with blue staining, through cells in which the peripheral region of the granule stained red and the core stained blue, to cells in which the entire granule stained red. In barley cv Waxy Hector, a low-amylose cultivar reported to have up to 8% amylose in its starch , granules containing some amylose were present from early in development (in grains of less than 20 mg fresh weight). In more mature endosperms of barley cv Waxy Hector, most of the granules stained either completely or partly blue with iodine solution (Fig. F). Figure 1 | A through K, Developing endosperm and starch from waxy barley mutants. A through K, Developing endosperm and starch from waxy barley mutants. For endosperm sections, whole grains were fixed in formaldehyde, embedded in wax, sectioned, and stained with iodine solution. All samples were taken from grain of 50 to 70 mg fresh weight (starting to turn yellow) except for those in E and G, which were from grain of approximately 20 mg fresh weight. Bars represent a distance of 50 mum. A through F, Low-amylose cultivars. G through H, Amylose-free cultivars. K, Wild-type cultivar. P, Pericarp. A, Barley cv Iyatomi Mochi. B, Barley cv Iyatomi Mochi. C, Barley cv Waxy Oderbrucker. D, Barley cv SB85750. E, Barley cv Iyatomi Mochi, young grain. F, Starch extracted from barley cv Waxy Hector. G, Barley cv Yon M Kei. H, Barley cv CDC Alamo. I, Barley cv CDC Alamo, young grain. J, Barley cv Arizona Hulless Waxy. K, Barley cv Shikoku Hadaka. In all of the cultivars, including those with no amylose in the endosperm, starch in the pericarp (Fig. I) and in the embryo (not shown) stained blue with iodine at all of the developmental stages at which it was present. The Presence of GBSSI Protein in the Endosperm | In all of the low-amylose cultivars, the starch contained a protein of approximately 60 kD, immunologically related to the GBSSI present in wild-type barley but present in very much lower concentrations than in wild-type starch (Fig. A). In the amylose-free waxy barley cv Yon M Kei, as reported previously , no 60-kD protein was detectable (Fig. A). However, in the amylose-free barley cv CDC Alamo, the amount of the 60-kD protein was very similar to that in representative wild-type cultivars. Measurements of starch synthase activity associated with granules in developing endosperm of barley cv CDC Alamo suggested that most or all of this GBSSI protein was inactive. The activity was comparable with or lower than that of other waxy cultivars, and only about 10% of that of wild-type cultivars (data not shown). Much or all of this residual activity is likely to be due to isoforms of starch synthase other than GBSSI . Figure 2 | A and B, Presence of GBSSI-like proteins in starch from the developing endosperm of barley. A and B, Presence of GBSSI-like proteins in starch from the developing endosperm of barley. After gelatinization by boiling in SDS-containing buffer, samples of starch from developing endosperms (from grains of approximately 50 mg fresh weight) were subjected to electrophoresis on 7.5% (w/v) SDS-polyacrylamide gels. Gels were either stained with Coomassie Brilliant Blue (right) or blotted onto nitrocellulose (left). Blots were developed with serum containing antibodies against GBSSI of pea embryos at a dilution of 1:2,000 (v/v; A) or 1:750 (v/v; B). A, Gel and blot of starch granule-bound proteins from whole endosperms. The phenotypes of the cultivars with respect to their amylose contents are indicated. WT, Wild type. LA, Low amylose. AF, amylose free. B, Gel and blot of starch from dissected outer layers and inner part of the endosperm of the low-amylose barley cv Iyatomi Mochi, and from whole endosperm of the wild-type barley cv Hector. To study the distribution of the GBSSI-like 60-kD protein in the endosperms of a low-amylose cultivar, the outer layers were dissected away from the inner part of the maturing endosperm of barley cv Iyatomi Mochi. Starch extracted from the outer layers contained considerably more of the 60-kD protein than starch from the inner part (Fig. B). Comparison of GBSSI Isoforms | To discover whether in barley, as in wheat, there is a second form of GBSSI, we searched for a barley expressed sequence tag (EST) similar to the nonendosperm form of GBSSI in wheat (GBSSII; ). A barley EST (accession no. ) was identified and used to clone a cDNA of a second form of barley GBSSI, which we called GBSSIb (submitted to GenBank; accession no. ). The predicted mature GBSSIb protein shares 96.4% identity with wheat GBSSII and 65.3% identity with barley GBSSI. Thus, in barley, as in wheat, there are two forms of GBSSI that are similar in amino acid sequence but differ particularly at the N termini of the mature proteins . To investigate whether GBSSIb is expressed in the outer cell layers of the endosperm of barley, we compared the predicted protein sequences of the two isoforms of GBSSI with protein sequences obtained experimentally from starch from wild-type and waxy barley endosperms. Table I | Comparison of the N-terminal sequences of GBSSI from wheat and barley GBSSI proteins in wild-type and low-amylose cultivars were compared by matrix-assisted laser-desorption ionization (MALDI)-time of flight (TOF) mass spectrometry (MS) and by N-terminal sequencing. MALDI-TOF MS was performed on tryptic digests of GBSSI purified from starch from the wild-type barley cv Hector and the low-amylose barley cv Waxy Hector and Iyatomi Mochi. For barley cv Hector, 13 peptides were identified that accounted for 29% of the amino acids in the mature GBSSI protein. For barley cv Waxy Hector, 16 peptides were identified that accounted for 32% of the amino acids in GBSSI. For barley cv Iyatomi Mochi, 14 peptides were identified that accounted for 32% of the amino acids in GBSSI. For all three samples, the best match of peptide masses obtained was to the amino acid sequence predicted from the barley GBSSI cDNA sequence (accession no. ; ). These results are consistent with the idea that the GBSSI-like protein in the low-amylose barley cv Iyatomi Mochi and cv Waxy Hector is the product of the same gene that encodes the endosperm GBSSI in wild-type barley. Protein sequencing revealed that the N-terminal 12 amino acids of the GBSSI protein from starch from the outer part of the endosperm of the low-amylose barley cv Waxy Oderbrucker matched the sequence of the GBSSI protein from the endosperm of the wild-type barley cv Shikoku Hadaka and cv Vogelsanger Gold . The sequences of these proteins were also very similar to that of the GBSSI expressed in wheat endosperm . They differed considerably from the N-terminal sequences of the nonendosperm form of GBSSI in wheat (GBSSII) and barley (GBSSIb). These data again suggest strongly that the GBSSI in the endosperm of the low-amylose waxy cultivars of barley is the same protein as that in the endosperm of wild-type barley, rather than a different isoform expressed primarily in other parts of the plant. Mutations in the GBSSI Gene of Waxy Barleys | To provide further evidence about the identity of the GBSSI in the endosperm of waxy barley cultivars, we cloned and sequenced the cDNA encoding GBSSI and 1 kb of the promoter region of the GBSSI gene from a wild-type cultivar, barley cv Oderbrucker, and from several low-amylose (Waxy Oderbrucker, Iyatomi Mochi, and SB85750) and amylose-free (Yon M Kei and CDC Alamo) lines and cultivars. The sequence of GBSSI obtained from barley cv Oderbrucker (accession no. ) was almost identical to that of the wild-type barley cv Vogelsanger Gold, published earlier (accession no. ). The 5'-untranslated region (UTR) of these wild-type alleles includes intron 1 , and the region upstream of the 5'-UTR contains a predicted transcription complex-binding site (TATA box) 43 bp upstream of the transcription start site . Figure 3 | Diagrammatic representations of the gene structure of GBSSI. Diagrammatic representations of the gene structure of GBSSI. Upper diagram, The entire GBSSI gene. Blocks represent the 5'-UTR and the exons. Lower diagram, the 5'-UTR and exon 1 expanded to show the 413-bp deletion and 15-bp insertion. The sequences for the promoter and 5'-UTRs of GBSSI have been submitted to GenBank (accession nos.: barley cv Oderbrucker, ; barley cv Waxy Oderbrucker, ; barley cv Iatoma Mochi, ; SB85750, ; barley cv CDC Alamo, ; and barley cv Yon M Kei, ). The GBSSI sequences for all of the low-amylose cultivars were very similar to one another and different from the sequences from wild-type barley in two main respects. First, in the GBSSI alleles from the low-amylose cultivars, there was a 413-bp deletion in the promoter and 5'-UTR including the TATA box, the start of transcription, and part of intron 1 . Second, there was also a 15-bp insertion in exon 1 that does not cause a frame shift but results in the addition of five extra amino acids to the transit peptide of the protein. To discover more about the distribution of this 15-bp insertion among barley cultivars, we sequenced the same region from barley cv Shikoku Hadaka, from which the waxy barley cv Yon M Kei was derived. The 15-bp insertion was present in the GBSSI allele in this cultivar (Table , column 1). Thus, the insert represents allelic variation that has little or no impact upon amylose content: It cannot be responsible for the low-amylose phenotype. We conclude that the reduction in amylose content in low-amylose cultivars is probably due to the 413-bp deletion that is common to all of the cultivars of this type that we have examined. Table II | Comparisons of the cDNA and predicted protein sequences of GBSSI from barley The GBSSI sequence from the amylose-free barley cv CDC Alamo was identical to that of the wild-type barley cv Vogelsanger Gold except for one base substitution (T instead of A) at position 860 in barley cv CDC Alamo. This is predicted to result in the substitution of the aliphatic amino acid (Val) for the acidic amino acid (Asp). Asp is conserved in this position in all of the other GBSSI alleles of barley and in GBSSI from a wide range of other species (data not shown). This single base change in barley cv CDC Alamo is likely to be the cause of the observed production of an inactive GBSSI protein in this cultivar (see above). Thus, it defines an Asp that is essential for GBSSI activity. The sequence of GBSSI from the amylose-free barley cv Yon M Kei also contained a single base substitution compared with wild-type sequences, at position 580 (T instead of C). This is predicted to create a stop codon, and thus is likely to be responsible for the complete lack of GBSSI protein that was observed in this mutant. In addition to the 15-bp insertion in the GBSSI gene already mentioned , there are other sequence differences between the GBSSI alleles for which we have complete sequences. These are summarized in Table . These minor differences are not correlated with amylose content and therefore are unlikely to contribute to the waxy phenotype. However, they represent at least some of the allelic variation in GBSSI that exists within cultivated barleys. On the basis of cDNA sequence comparisons, the GBSSI alleles can be divided into two groups representing two haplotypes. Group 1 contains GBSSI alleles from the wild-type barley cv Vogelsanger Gold and cv Oderbrucker and the amylose-free barley cv CDC Alamo. All of the members of this group lack the 15-bp insertion and vary from the group 2 alleles in 10 other positions (Table , columns 1 --4). Group 2 contains GBSSI alleles from the wild-type barley cv Shikoku Hadaka, the low-amylose lines/cultivars, Iyatomi Mochi, Waxy Oderbrucker, and SB85750, and the amylose-free barley cv Yon M Kei. All of the members of this group have the 15-bp insertion. Comparison of GBSSI Transcripts | The relative amounts of GBSSI transcripts in developing endosperms of wild-type and waxy barleys were compared in two ways. First, semiquantitative RT-PCR (Fig. A) showed that the low-amylose line/cultivars Waxy Oderbrucker, Iyatomi Mochi, and SB85750 all had normal or only slightly reduced levels of GBSSI transcript in endosperms from grains of 30 to 45 mg fresh weight The amylose-free barley cv Yon M Kei (containing a GBSSI allele with an introduced stop codon in the coding region) also had a normal level of transcript. However, the low-amylose barley cv CDC Alamo, which had a normal amount of an inactive form of GBSSI protein, had elevated levels of transcript. Second, the transcript levels in barley cv Oderbrucker and cv Waxy Oderbrucker at two developmental stages were compared by northern analysis (Fig. B). In the older endosperms (from grains of 30 --45 mg fresh weight), the transcripts in these cultivars were of the expected size and were of similar abundance. The GBSSI transcript was also abundant in young endosperms (from grains of 12 --16 mg fresh weight) of barley cv Oderbrucker. However, in young endosperms of barley cv Waxy Oderbrucker, there was very little, if any, transcript. Figure 4 | A and B. A and B. Comparison of GBSSI transcripts. A, Semiquantitative RT-PCR. RNA was extracted from endosperms of grains of 30 to 45 mg fresh weight Tracks are: 1, barley cv Oderbrucker; 2, barley cv Waxy Oderbrucker; 3, barley cv Yon M Kei; 4, barley cv Iyatomi Mochi; 5, SB85750; and 6, barley cv CDC Alamo. Upper, Product generated using primers designed to amplify GBSSI. Lower, Product generated using primers designed to amplify Mub-1 (ubiquitin). B, Northern blots. Tracks are: 1 and 3, barley cv Oderbrucker; and 2 and 4, barley cv Waxy Oderbrucker. RNA was extracted from endosperms of grains of 30 to 45 mg fresh weight was in tracks 1 and 2 and RNA from grains of 14 to 16 mg fresh weight was in tracks 3 and 4. Upper, Products generated using primers designed to amplify GBSSI. The approximate size of the GBSSI transcript is indicated. Lower, Ethidium bromide-stained gels used to prepare the blots shown in the upper panels. Attempts to map the transcription start site for GBSSI in the low-amylose mutants and to search for alternative transcription start sites in the 5'-upstream sequences have been unsuccessful so far. We assume that there is an alternative transcription start site either in the remaining part of intron 1 or further upstream. This results in a longer pre-RNA that is spliced to give a mature RNA of similar size to the mature RNA of the wild type. DISCUSSION : Our results suggest that there is only one GBSSI gene expressed in barley endosperm and that this is responsible for the amylose synthesized in low-amylose waxy cultivars as well as that in wild-type cultivars. MALDI-TOF analysis and amino acid sequence from the GBSSI protein in the endosperm of these waxy cultivars show that this protein is indistinguishable from the GBSSI of wild-type endosperms. The protein is different from the predicted product of a second GBSSI gene, GBSSIb. GBSSIb, like the homologous gene in wheat (GBSSII, ), is probably expressed in parts of the plant other than the endosperm. It is likely to be responsible for the synthesis of the amylose in the pericarp, including that observed in pericarps of the amylose-free cultivars. The GBSSI alleles of all of the low-amylose waxy barleys we examined carry a 413-bp deletion in the promoter and 5'-UTR. This discovery strongly suggests that the alleles in all of these cultivars are derived from a single origin. The cultivars we examined are extremely diverse in phenotype and geographical origin. They include cultivars bred in Japan, Canada, Europe, and the United States, and a wide range of awn, row, pigment, hull, and growth habit characteristics. Nonetheless, records on the origins of the cultivars are consistent with the idea that the waxy characteristic in most or all of these barleys is derived from waxy barleys native to Asia . The Asian waxy barleys are probably all descended from a glutinous (waxy) form of barley recorded in China in the 16th century. Glutinous barleys are believed to have been introduced into Japan from Korea at some time before the 17th century . The European and North American waxy barleys stem from a U.S. Department of Agriculture-Agricultural Research Service (USDA-ARS) breeding program in Aberdeen, Idaho (1940 --1950) in which a Japanese cultivar, barley cv Murasaki mochi, was crossed to a European cultivar, barley cv Oderbrucker (Harold Bockelman, USDA-ARS, personal communication). The resulting barley cv Waxy Oderbrucker was then used in a breeding program in the United States and Canada. We can trace many of the modern waxy cultivars back to barley cv Waxy Oderbrucker. For example, it was a parent of barley cv Waxy Betzes , which was then used as a parent in the breeding program in Saskatchewan from which the line SB85750 was derived. Barley cv Waxy Hector was also derived from either cv Waxy Betzes or cv Waxy Oderbrucker. Our analysis indicates that in the low-amylose barley mutants, the 413-bp deletion alters the spatial and/or temporal expression of GBSSI in the endosperm. The deletion removes the normal transcription complex-binding site (TATA box) and transcription start site. We assume that an alternative, upstream promoter region and transcription start site are used to produce the GBSSI transcript in the mutants. The mutant transcript is produced only late in endosperm development, consistent with the appearance of amylose only later in endosperm development. The fact that amylose and GBSSI protein are found mainly or exclusively in the outer cells of the endosperm indicates that the alternative promoter specifies a different spatial and/or temporal pattern of expression from the normal promoter. Effects on amylose content of mutations in the promoter or 5'-UTR of GBSSI have also been reported in other species. In rice, variations in amylose content between cultivars were shown to be due to differences in the efficiency with which intron 1 in the 5'-UTR was removed from the GBSSI pre-mRNA . Low-amylose (6.7% --16.0% amylose) cultivars of rice had lower levels of mature GBSSI mRNA than high-amylose (20.0% --27.8% amylose) cultivars as well as incompletely processed GBSSI mRNA. In these cultivars, the inefficient processing of the pre-mRNA was due to a single base mutation at the 5' splice site of intron 1. The reduced efficiency of GBSSI pre-mRNA processing also resulted in alternate splicing at multiple sites, some of which had non-consensus splice site sequences. Growth temperature can also affect the efficiency of pre-mRNA processing in low-amylose cultivars of rice. Plants grown at 18C had higher steady-state levels of mature GBSSI mRNA than plants grown at 25 or 32C . At lower temperatures, when splicing was more efficient, the activity of GBSSI in the low-amylose cultivars was higher and more amylose was synthesized . In potato (Solanum tuberosum), GBSSI activity and amylose content are affected by the presence or absence of a 140-bp fragment at a site in the promoter region approximately 0.5 kb upstream of the ATG start codon . Alleles of the gene that contain the 140-bp fragment result in lower GBSSI activity and amylose content than alleles without this fragment. The basis for this effect is not known because the variations in GBSSI activity could not be attributed to large differences in the amounts of either GBSSI RNA or protein. The highly variable amylose content of starch granules from low-amylose cultivars of barley may be an important consideration in assessing the physicochemical properties of starch. Individual endosperms contain granules that stain blue with iodine that probably have near-normal levels of amylose, granules with blue-staining cores that are likely to have a severely reduced amylose content, and granules that stain almost completely red and probably have near-zero amylose contents. This contrasts with the situation in low-amylose starches from other species. For example, in the low-amylose lines of potato created by expression of antisense GBSSI constructs , there was less granule-to-granule variation in amylose content than in the low-amylose barley cultivars. Whether the physicochemical properties of low-amylose starches of comparable bulk amylose content are influenced by the extent of heterogeneity of amylose contents between individual granules remains to be determined. Two of the waxy cultivars with no amylose in the endosperm, barley cv Arizona Hulless Waxy and cv Yon M Kei, were produced by mutagenesis of wild-type cultivars rather than by breeding from low-amylose cultivars traceable to Japan (see below). Barley cv Yon M Kei was generated by mutagenesis of the wild-type barley cv Shikoku Hadaka with sodium azide . Barley cv Arizona Hulless Waxy was generated by mutagenesis of a wild-type line, 76-19-7, with diethyl sulfate (PI 560053; ). The third amylose-free cultivar, barley cv CDC Alamo, was derived by breeding from barley cv Arizona Hulless Waxy, and probably carries the same GBSSI allele as this parent . Barley cv Yon M Kei and cv CDC Alamo do not have the 413-bp deletion seen in the low-amylose cultivars. Instead, they have different mutations in the GBSSI gene that account for the complete absence of GBSSI activity. The endosperm of barley cv Yon M Kei contains no detectable GBSSI protein. There is a single base substitution in this gene that creates a stop codon. This prevents the production of a full-length GBSSI protein. Presumably, the incomplete GBSSI protein is either unstable and is rapidly degraded or it is not capable of binding to the starch granules and, therefore, would not have been detected in our experiments. Barley cv CDC Alamo has wild-type levels of GBSSI protein. The mutant protein is able to bind tightly to starch granules like normal GBSSI, but it is unable to synthesize amylose due to a single base substitution that results in a conserved Asp residue (Asp-217) being replaced by a Val. It is known that enzymes in the glycosyltransferase family of which GBSSI is a member have Asp residues at the catalytic center that participate in the enzymatic reaction . Mutational analysis of several glycosyltransferases, including an isoform of starch synthase, has shown an absolute requirement for certain conserved Asp residues (for example, cellulose synthase, ; chitin synthase 2, ; and starch synthase IIb, ). Substitution of these ---even conservative substitution with similar amino acids ---results in inactive enzymes. However, whether the conserved Asp-217 is at the catalytic center of GBSSI remains to be discovered. MATERIALS AND METHODS : Plant Material and Growth Conditions | Grains of barley (Hordeum vulgare) cultivars were obtained from the John Innes Centre and Crop Development Centre Germplasm collections, from Dr. Tom Blake (Montana State University, Bozeman, barley cv Nubet), and from Dr. Naoyuki Ishikawa (Tochigi Agricultural Experiment Station, Tochigi, Japan, barley cv Shikoku Hadaka 84 and cv Yon M Kei 286). Barley plants were grown in a greenhouse in individual pots at a minimum temperature of 12C, with supplementary lighting in winter. Extraction of Starch | Endosperms, free of pericarp and embryos, were dissected from developing grains of 50 mg fresh weight. For starch extraction from the inner and outer layers of the endosperm, whole endosperms were frozen on dry ice, then "peeled" with a fine razor blade to remove tissue to a depth of very approximately 0.5 mm from the surface. Starch was extracted as described by and stored at -20C. SDS-PAGE and Immunoblotting | Starch samples were washed twice by suspension and centrifugation in 2% (w/v) aqueous SDS, at 30 mg starch mL-1. Washed starch was suspended in gel sample buffer at 50 mg starch mL-1, boiled for 5 min, and allowed to cool. Unlike the wild-type starches, centrifugation of the starches from the waxy lines did not result in a supernatant. Therefore, none of the samples were centrifuged after boiling. Instead, approximately 30 muL of the resulting paste was loaded directly into wells on 7.5% (w/v) SDS-polyacrylamide gels (10 cm long, 1 mm thick). Due to the starch in the samples, which remained in the wells, there was some unavoidable distortion of the protein bands on the gel (see Fig. ). After electrophoresis, gels were stained with Coomassie Brilliant Blue R or electroblotted onto nitrocellulose. Blots were developed with serum containing antibodies against GBSSI of pea (Pisum sativum) embryos at a dilution of 1:2,000 (v/v), as described by . Identification of GBSSI Proteins by MALDI-TOF and by Sequencing | Starch granule-bound proteins were subjected to SDS-PAGE and the separated proteins were stained with Coomassie Brilliant Blue R-250. The major, approximately 60-kD, protein bands were excised and subjected to tryptic digestion according to followed by analysis by MALDI-TOF MS. Mass fragment sizes were used to query the National Center for Biotechnology Information database using the MASCOT search tool . All matched sequences showed mass errors of <75 muL L-1. Protein was blotted from an SDS polyacrylamide gel onto a polyvinylidene difluoride membrane (Immobilon P, Millipore, Bedford, MA) and stained with Coomassie Brilliant Blue R250. The excised band was sequenced directly from the membrane by Edman degradation on a model 494 Procise protein sequencer (PE-Applied Biosystems, Foster City, CA) using the pulsed-liquid mode. Light Microscopy | Freshly harvested tissue was fixed in a formaldehyde solution and dehydrated through a graded ethanol series according to . After transfer to Histoclear (Agar Scientific, Stansted, Essex, UK), tissue was embedded in Paramat wax (BDH, Poole, UK), and sectioned. Wax was removed with Histoclear and sections were transferred through an ethanol series into water and photographed through a light microscope after staining with iodine solution (dilutions of 2- or 5-fold of Lugol's solution; Sigma, Poole, UK). To improve contrast, red/brown and blue colors in photographs were enhanced using Adobe Photoshop software (Adobe Systems, Mountain View, CA). DNA Extraction and the Cloning and Sequencing of Alleles of GBSSI | DNA was extracted from 0.1-g samples of young barley leaves with the DNeasy plant mini-kit (Qiagen, Hilden, Germany). The promoters were amplified using Pfu-Turbo DNA polymerase (Stratagene, La Jolla, CA) with primers designed to GBSSI from barley cv Vogelsanger Gold forward (5'-TATATGACGCACTCCACACCCACACACACA-3') and reverse (5'-CTGTTCCTGAAATCTAAGATCGTTTGCAGA-3'). The blunt-ended PCR products had adenine residues added by incubation with deoxyadenylate triphosphate and Taq polymerase at 72C. The products were then ligated into pGEM-T-easy vector (Promega, Madison, WI) for sequencing. Overlapping sequencing primers were designed at 400-bp intervals. RNA Extraction and DNA Synthesis | RNA was extracted from 0.5 g of endosperms from grains of 30 to 45 mg each or from 21 whole grains of 12 to 16 mg each using Concert RNA reagent (Invitrogen Ltd., Paisley, UK). RNA was treated with DNaseI (Roche, Basel) and cleaned again with phenol:chloroform. cDNA was synthesized at 58C from a reverse primer designed to Vogelsager Gold cDNA (; P2, 5'-TGCTCCATGCACCAGAATGT-3') using Thermoscript RT (Invitrogen Ltd.). The enzyme was denatured at 85C and RNA removed from the duplex by incubation at 37C with RNase H (Roche). RT-PCR | RT-PCR with the cDNA synthesis primer (P2) and a forward primer (P1, 5'-TGCTCTCTCACTGCAGGTAG-3') was done using Pfx polymerase or Platinum Taq DNA polymerase and PCRx reaction buffer (Invitrogen Ltd.). Semiquantitative RT-PCR was done in tandem with primers to ubiquitin (mub1-M60175; 5'-CGGACACCATCGACAACGTCCAG-3' and 5'-GCCA-GTTCTAAGCCTTCTGGTTGTAG-3'). PCR cycles were paused after 15 cycles and 5-muL aliquots removed. Products were separated on 1% (w/v) agarose gel and transferred to Duralon nylon membrane (Stratagene) by capillary transfer. Membranes were hybridized in phosphate buffer (pH 7.4), 10 mm EDTA, and 7% (w/v) SDS with 25 ng of 32P[alpha-dCTP]-labeled GBSSI-cDNA or ubiquitin-cDNA amplified from Oderbrucker, for 4 h at 65C and washed with 0.1x SSC containing 0.1% (w/v) SDS. Cloning and Sequencing of GBSSIb | RNA was extracted from the entire seeds of barley cv Nubet 3 DPA. Ten micrograms of total RNA was used to synthesize cDNA using the Generacer Kit (Invitrogen Ltd.). RACE products amplified with primers designed to EST (5'-TCCTACAACTGGAACAGACTTCCGAGATAA-3' and 5'-ACGGTTCTGCTTTTGTGCTTGCTGCATT-3') were cloned into the TOPO-10 vector (Invitrogen Ltd.) and sequenced. Northern Blotting | Ten micrograms of total RNA was separated on a 1% (w/v) agarose denaturing gel with size standards (Promega). RNA was visualized with ethidium bromide under UV light to ensure equal loading. RNA was transferred to Duralon nylon membrane (Stratagene) by capillary transfer. Hybridization conditions were identical to those used for semiquantitative RT-PCR. Backmatter: PMID- 12226500 TI - Indole Acetic Acid Distribution Coincides with Vascular Differentiation Pattern during Arabidopsis Leaf Ontogeny AB - We used an anti-indole acetic acid (IAA or auxin) monoclonal antibody-based immunocytochemical procedure to monitor IAA level in Arabidopsis tissues. Using immunocytochemistry and the IAA-driven beta-glucuronidase (GUS) activity of Aux/IAA promoter::GUS constructs to detect IAA distribution, we investigated the role of polar auxin transport in vascular differentiation during leaf development in Arabidopsis. We found that shoot apical cells contain high levels of IAA and that IAA decreases as leaf primordia expand. However, seedlings grown in the presence of IAA transport inhibitors showed very low IAA signal in the shoot apical meristem (SAM) and the youngest pair of leaf primordia. Older leaf primordia accumulate IAA in the leaf tip in the presence or absence of IAA transport inhibition. We propose that the IAA in the SAM and the youngest pair of leaf primordia is transported from outside sources, perhaps the cotyledons, which accumulate more IAA in the presence than in the absence of transport inhibition. The temporal and spatial pattern of IAA localization in the shoot apex indicates a change in IAA source during leaf ontogeny that would influence flow direction and, consequently, the direction of vascular differentiation. The IAA production and transport pattern suggested by our results could explain the venation pattern, and the vascular hypertrophy caused by IAA transport inhibition. An outside IAA source for the SAM supports the notion that IAA transport and procambium differentiation dictate phyllotaxy and organogenesis. Keywords: Introduction : In 1880, Darwin stated: "Some influence moves from the tip of an oat coleoptile to the region below the tip where it controls elongation." This moving influence ---later shown to be indole acetic acid (IAA; ; ) ---is the first description of polar auxin transport. Polar auxin transport is ubiquitous among higher plants. Efficient transport of IAA modulates cell shape and differentiation and is necessary for normal organogenesis and vascular patterning . Vascular tissues are conduits for water and nutrients throughout the plant body. They are generated during embryogenesis and organogenesis, expanding along the growth axis of the organ. Vascular development begins with the differentiation of provascular tissue, the procambium , through periclinal cell division, cell elongation, and cell alignment. The procambium of dicotyledonous embryos such as Arabidopsis becomes evident at early heart stage as elongated cells in the center of the embryo distinct from the nearly isodiametric surrounding ground tissue cells . As the embryo matures, the procambial cells differentiate into phloem and xylene elements . Vascular tissues connect the leaves and other parts of the shoot with the roots, enabling efficient long-distance transport between organs. The vascular network is particularly extensive in leaves, with primary, secondary, tertiary (or 1, 2, and 3, respectively), and higher order veins. The veins arise at different times and are arranged in a pattern, referred to as the venation pattern, reflecting the ontogeny and structural organization of the leaf. Arabidopsis leaves are pinnate with a single 1 vein (midvein) from which arise all the 2 veins that rejoin the 1 vein, forming a series of prominent arches . The 3 veins form bridges between 2 veins, whereas quaternary veins extend from 3 veins and end blindly in areoles . The hierarchical differentiation of 1, 2, 3, and higher order veins provides an excellent system to study the mechanism of vascular differentiation and pattern formation . Vascular differentiation is related to auxin flux . Auxin transport appears to be mediated by specific cellular influx and efflux proteins . The directionality of auxin flow is attributed to polar distribution of the efflux carrier molecules in the plant cell membrane . Two models, canalization of auxin flow and reaction-diffusion prepattern, have been proposed to explain the pattern of vascular differentiation . The canalization of signal flow hypothesis is based on a positive feedback mechanism: a proposed gradual restriction of IAA flow from a field to specialized files of cells, resulting in provascular, and later vascular, differentiation . IAA-induced de novo vascular differentiation and the effect of changing IAA flow on vascular pattern support the IAA flow-dependent canalization hypothesis . However, some investigators have argued for the reaction diffusion theory based on observations such as the fragmented vascular strands in some vascular mutants. This theory emphasizes generation of stable patterns autonomously in an initially homogenous field by interacting substances with different diffusion rates . Both theories predict vascular development based on a leaf autonomous signal source . When the polar auxin transport inhibitor 1-N-naphthylphtalamic acid (NPA) is used to block IAA flow, vascular development is impaired . NPA caused central and marginal vascular hypertrophy ---a general increase in the number and size of veins. NPA treatment also interferes with organogenesis, inhibiting both lateral root development and the formation of new leaf primordia . Although auxin transport is implicated in a variety of growth and differentiation processes , little is known about the site of IAA production or its route of transport. It has been generally believed that IAA is produced in the shoot apex and in the tips of older leaves and transported basipetally, but the site of IAA production and its distribution in plant tissue have not been characterized. Methods for detecting IAA in plant tissues are being developed. Constructs containing an IAA-inducible promoter (Aux/IAA) fused to the beta-glucuronidase (GUS) reporter gene can detect IAA in situ . Monoclonal antibodies against IAA have been used to localize IAA in maize (Zea mays; ; ) and peanut (Arachis hypogaea) tissues . In this work, we show that immunocytochemistry with a monoclonal anti-IAA antibody can detect free IAA in Arabidopsis tissues. While studying IAA distribution in growing organs, we found that NPA prevents accumulation of IAA in the shoot apical meristem (SAM) and the youngest pair of leaf primordia, but not in older leaf primordia. The implications of our findings for the direction of IAA flow and vascular differentiation are discussed below. RESULTS : IAA Immunolocalization in Arabidopsis Tissues | The production and isolation of monoclonal antibodies highly specific for IAA provided the means to localize and evaluate in situ IAA levels in maize and peanut tissues . We prefixed Arabidopsis tissue samples with EDAC, which cross-links the carboxyl group of IAA to structural proteins in the plant tissues, creating the epitope recognized by this anti-IAA monoclonal antibody . The prefixed tissues were processed, sectioned, and reacted first with the monoclonal anti-IAA antibody then with the secondary antibody, anti-mouse IgG conjugated with alkaline phosphatase, before the enzymatic reaction was carried out to obtain the color signal. IAA signal was low in pith but high in the epidermal and cortical tissues and vascular bundles of inflorescence stems (Fig. A). We verified the effectiveness of the immunolocalization technique and the specificity of the antibody with several controls. Because the monoclonal antibody was raised against free IAA cross-linked to bovine serum albumin (BSA) through its carboxyl group , we tested unfixed tissues for color reaction. We observed no color comparable with the prefixed stem section (Fig. , A and B), indicating both that prefixation with EDAC is essential for free IAA detection by this antibody and that the antibody does not recognize other epitopes in these tissue sections. No signal was detected when the primary (Fig. C) or secondary antibody (Fig. D) was omitted, indicating that the color reaction is dependent on the presence of these antibodies on the tissue section and demonstrating again that the background color is very low. Figure 1 | IAA immunolocalization in Arabidopsis tissues. IAA immunolocalization in Arabidopsis tissues. A through D, Cross sections of inflorescence stem. A, Stem tissues prefixed with ethyl-3(3-dimethylaminopropyl)carbodiimide hydrochloride (EDAC), embedded in paraffin, sectioned, and reacted with the anti-IAA antibody followed by anti-mouse IgG secondary antibody conjugated with alkaline phosphatase. There is a high level of IAA signal in the epidermal and cortical tissues and around vascular bundles: B through D, controls, showing very low levels of IAA signal; B, no EDAC prefixation; C, no primary antibody; D, no secondary antibody; E through K, longitudinal section of young siliques; E, eight-cell embryo; F, early globular stage embryo, showing high IAA signal in the embryo and endosperm cells (b) and a lower IAA level in the suspensor (a) and ovule cells (c); G, globular stage embryo with the omission of the primary anti-IAA antibody; and H, torpedo stage embryo with the omission of the secondary antibody. These controls show very low levels of the IAA signal: I, heart stage embryo; J, torpedo stage embryo; and K, walking stick stage embryo, showing high levels of IAA in the embryo. The suspensor (a), endosperm (b), and ovule cells (c) are indicated. L and M, GUS activity in embryos of DR5::GUS transgenic plants. L, Heart stage embryo; M, torpedo stage embryo showing high level of IAA in the embryos. Bar = 100 mum in A through D, 10 mum in E and F, 5 mum in G, 20 mum in H through J, 50 mum in K, 20 mum in L, and 50 mum in M. We used the anti-IAA monoclonal antibody to study IAA distribution in growing tissues, embryos, leaf primordial, and SAMs (Figs. and ), where IAA is expected to be high. Immunocytochemistry on longitudinal sections of young siliques with different stages of embryo development is shown in Figure , E, F, and I through K. The signal is high in all embryo cells and lower in the suspensor and the ovule. Lower IAA signal may reflect more disperse cytoplasm in the more vacuolated suspensor and ovule cells. High signal is detected in the endosperm surrounding the suspensor in the early embryonic stages (Fig. , E and F). Figure 2 | IAA immunolocalization in leaf primordia and SAM. IAA immunolocalization in leaf primordia and SAM. A and B, Cross sections of the shoot apex of 4-d-old seedlings were treated as described in the legend to Figure . A, Shoot apex of an untreated seedling, showing high IAA signal in the pair of first node leaf primordial. B, Shoot apex of a seedling grown in the presence of 40 mum NPA, showing little IAA signal in the leaf primordial. C and D, Longitudinal sections of 4-d-old seedlings. C, SAM of an untreated seedling, showing high IAA level. D, SAM of a seedling grown in the presence of 40 mum NPA, showing low IAA level. E through H, Serial cross sections of the shoot apex of 5-d-old seedlings that were treated as described in the legend to Figure . Drawing on the left depicts the sites where the four sections were made through the shoot apex including the first node (1) and second node (2) leaf primordia. Higher IAA levels are detected in the upper than in the lower sections. C, Petiole of the cotyledons. Bar = 20 mum in A and B, 10 mum in C and D, and 25 mum in E through H. Embryonic stages in which procambium can be detected, i.e. heart stage (Fig. I), torpedo stage (Fig. J), and walking stick stage (Fig. K), have high IAA signal evenly distributed in all embryonic cells with no difference in signal level between procambium and ground tissue. No IAA signal was detected when the primary anti-IAA antibody (Fig. G) or the secondary antibody (Fig. H) was omitted, demonstrating again that the immunocytochemical reaction is specific and that background color is very low. The IAA signal pattern was consistent across embryos with only minor variation. All 42 pro-embryos and globular stage embryos examined showed high IAA signal in the embryo and surrounding endosperm and lower signal in the suspensor. Thirty-five of 39 heart stage embryos analyzed had the uniform immunostaining shown in Figure I, whereas four showed high signal only in the distal end of the cotyledons and not uniformly in all the embryonic cells. Seventeen of 18 torpedo stage embryos showed the signal pattern in Figure J. Figure K shows the high IAA signal seen in all 19 walking stick stage embryos examined. To verify the anti-IAA antibody signal, we used the GUS activity of transgenic plants carrying the DR5::GUS construct, composed of the IAA-inducible promoter (Aux/IAA) fused to a GUS reporter gene . Figure , L and M, show GUS activity throughout heart and torpedo stage embryos. The walking stick stage embryo has less GUS activity, which diminishes as the embryo matures and becomes dormant. The overall GUS pattern is consistent with the immunological signals during embryogenesis. For unknown reasons, there are a small percentage, about 10%, of the torpedo stage and walking stick stage embryos that show higher GUS activity in the root cap and the tips of the cotyledons (data not shown). IAA Signal in the Shoot Apex | The immunocytochemistry technique was used to examine IAA distribution in shoot apices of Arabidopsis seedlings. The first node rosette leaves (Fig. A) and SAM (Fig. C) of 4-d-old seedlings had high levels of IAA signal. Figure A shows high levels of IAA signal in a cross section of the first node rosette leaves, and Figure C shows high levels of IAA in a longitudinal section of the SAM. The IAA signals in Figure were representative of 48 leaf primordia and 17 SAMs examined. We also determined the GUS activity of transgenic seedlings carrying the DR5::GUS construct . As shown in Figure A, high levels of GUS activity were detected in the first node leaf and stipule of 4-d-old seedlings. These results are similar to the IAA immunolocalization findings, demonstrating that the IAA inducibility of the DR5::GUS construct is a useful reporter of the endogenous IAA levels in the shoot apex. Figure , A, C, E, and G show the first true leaf at different ages. As the leaf primordium grows, only the distal end of the leaf primordium maintains high GUS activity (Fig. , A, C, E, and G). High GUS activity in the distal end of the leaf primordia was confirmed with the immunological assay for IAA. Figure , E through H, show a series of four cross sections of a 5-d-old shoot apex that has high anti-IAA antibody signal in the sections from the distal end of the first node leaf primordia, particular in the adaxial side of the leaf. Leaf primordia of subsequent nodes have a similar spatial and temporal pattern of GUS activity: high levels in the young leaf primordia that decline with growth (for example, see the second node leaf in Fig. , E, G, and I, and the third node leaf in Fig. I). Figure 3 | IAA distribution and venation pattern in transgenic plants grown in the presence and absence of 40 mum NPA. IAA distribution and venation pattern in transgenic plants grown in the presence and absence of 40 mum NPA. DR5::GUS activity in transgenic plants: A, 4-d-old shoot apex, showing the GUS-positive first true leaf and the stipules (white arrows); B, 4-d-old shoot apex in a seedling grown in the presence of NPA, showing very low levels of IAA; C, 5-d-old shoot apex showing the GUS-positive true leaves and their stipules (white arrows); D, 5-d-old shoot apex in a seedling grown in the presence of NPA, showing some GUS signal in the distal end of the leaf; E, 6-d-old shoot apex showing first node leaf primordium with declining GUS activity and second node leaves (red arrow) and stipules (white arrows) with high level of IAA; F, 6-d-old shoot apex in seedling grown in the presence of NPA, showing more GUS signal in the leaf tip and the emerging marginal veins; G, 8-d-old shoot apex showing the GUS activity in the second node leaves (red arrows) and stipules (white arrows), whereas the signal in the first node true leaves decreased and is concentrated in the leaf tip; H, 8-d-old shoot apex in a seedling grown in the presence of NPA, showing increased GUS signal and the expanding veins along the leaf margin; I, IAA distribution in the subsequent leaf nodes of a 10-d-old seedling, showing high IAA signal in the third node leaves (red arrows) and the stipules (white arrows) and lower signals in the second node leaves; J, second rosette leaves of 10-d-old seedlings grown in the presence of NPA, showing no IAA signal; K and L, GUS activity in the first true leaf of a 10-d-old seedling (K) and a 10-d-old seedling grown in the presence of 40 mum NPA (L); M and N, venation pattern and IAA distribution in the first true leaf of 10-d-old seedling. Seedlings were fixed in 6:1 (v/v) ethanol:acetic acid for 4 h at room temperature and then rinsed and whole mounted as described in "Materials and Methods"; M, venation pattern of the first true leaf, showing 1o, 2o, and 3o veins; N, venation pattern of a first true leaf of seedling grown in the presence of 40 mum NPA, showing the marginal and central hypertrophy; 1, first node leaves; 2, second node leaves; and 3, third node leaves. Bar = 20 mum in A and B, 50 mum in C and D, 100 mum in E and F, 200 mum in G through J, and 400 mum in K through N. Effect of NPA on IAA Distribution during Leaf Ontogeny | To study the effect of inhibition of IAA transport on IAA distribution, we germinated Arabidopsis seedlings in media supplemented with 40 mum NPA (concentration chosen based on studies of ) and sectioned shoot apices for IAA immunolocalization studies. The IAA signal in the leaf primordia (Fig. B) and in the SAM (Fig. D) of 4-d-old seedlings grown in the presence of NPA was very low relative to the untreated control (Fig. , A and C, respectively). (Three of 44 leaf primordia from seedlings grown on NPA showed a slightly higher IAA signal than in Fig. B.) Similarly, DR5::GUS activity was barely detectable in shoot apices of seedlings grown in the presence of 40 mum NPA (Fig. , B and D). However, IAA signal was detectable in the apical area of the first node leaf primordium in 5- to 6-d-old NPA-grown seedlings (Fig. , D and F) and increased in intensity in leaf primordia of 8-d-old (Fig. H) and 10-d-old (Fig. L) seedlings, in the presence of NPA. No GUS activity was detected in the stipules or the SAM (Fig. , B, D, F, H, and J). The second node leaves of NPA-grown seedlings showed the same spatial and temporal distribution of the GUS activity. No IAA signal was detected in newly emerged, 140-mum leaf primordia (Table ; Fig. J) until d 5 to 6, when IAA signal appeared at the leaf tip (data not shown). In contrast, control leaf primordia of the same length showed GUS signal throughout the entire leaf (second node leaf in Fig. E and third node leaf in Fig. I). Table I | Leaf lengths of Arabidopsis seedlings growing with or without NPA Effect of NPA on Organogenesis and Venation Pattern | NPA treatment altered the venation pattern and leaf and root shape. Leaf primordia grown in the presence of NPA (Fig. , D, F, H, and L) have broad leaf blades and broad, short petioles relative to leaves grown in the absence of the IAA transport inhibitor (Fig. , C, E, G, and I). Relative to the control root (Fig. N), the NPA-treated seedling root is shorter, lateral roots are inhibited, and the root tip is wider (Fig. O). The first node leaf primordia of untreated plants triple in length every 2 d. The leaves of NPA-treated seedlings grow much slower . NPA also slows the rate of emergence of leaves from subsequent nodes . Figure 4 | IAA distribution in seedlings containing a different IAA-inducible promoter::GUS and in a DR5::GUS-containing line treated with three different IAA transport inhibitors. IAA distribution in seedlings containing a different IAA-inducible promoter::GUS and in a DR5::GUS-containing line treated with three different IAA transport inhibitors. A through F, GUS activity in 4-d-old transgenic seedlings containing different Aux/IAA promoter::GUS. A, DR5::GUS transgenic line; B, DR5::GUS transgenic line grown in the presence of 40 mum NPA; C, BA::GUS transgenic line; D, BA::GUS transgenic line grown in the presence of 40 mum NPA; E, SAUR-AC1::GUS transgenic line; and F, SAUR-AC1::GUS transgenic line grown in the presence of 40 mum NPA. G through K, GUS activity in a 5-d-old DR5::GUS transgenic line grown with a different IAA transport inhibitor; G, seedlings grown without any inhibitor, showing high level of GUS activity; H, seedlings grown in the presence of 20 mum NPA, showing low level of the GUS activity; I, seedlings grown in the presence of 40 mum NPA; J, seedlings grown in the presence of 20 mum 2-chloro-9-hydroxyfluorene-9-carboxylic acid (HFCA); K, seedlings grown in the presence of 40 mum 2,3,5-triiodobenzoic acid (TIBA); L, 5-d-old cotyledon, showing lower level of the GUS activity than a 5-d-old cotyledon of seedling grown in the presence of NPA (M); and N, 4-d-old root tip and O, 4-d-old root tip of seedling grown in the presence of 40 mum NPA, both showing high level of GUS activity. Bar = 50 mum in A through F, 100 mum in G through K, 200 mum in L and M, and 100 mum in N and O. As reported in , NPA treatment affects the venation pattern. For example, a 5-d-old leaf has a differentiated midvein, but leaves of seedlings grown in the presence of NPA do not (Fig. , F and H). Instead, the first veins form laterally along the margins of the leaf primordia at the distal end of the leaf in 6-d-old seedlings (Fig. F), then expand along the leaf margins to form a broad band of vascular tissue, called the marginal hypertrophy (Fig. H). By 8 d after germination, multiple files of veins, called the central hypertrophy (Fig. N), are formed in the center of the leaf blade, extending into the petiole but not connecting to the vascular system of the hypocotyl. This is consistent with the findings of and . It is worth noting that the area of IAA-driven DR5::GUS activity (Fig. , F and H) coincides with the site of NPA-induced vascular differentiation, especially in the marginal hypertrophy. IAA Distribution in Different promoter::GUS- Containing Seedlings Grown in the Presence and Absence of NPA | Because the expression pattern of DR5::GUS depends on tissue-specific promoter activity as well as the presence of IAA, we tested two other Aux/IAA response promoter::GUS constructs: PSIAA4/5 BA::GUS and SAUR-AC1::GUS . BA is the PSIAA4/5 promoter region that contains two auxin-responsive domains (AuxRD A and AuxRD B). Domain A contains a highly conserved sequence found in various IAA-inducible genes that behaves as a major auxin-responsive element. Domain B functions as an enhancer element. The two domains, which act cooperatively to stimulate transcription , were fused to the GUS reporter gene and introduced into Arabidopsis . The SAUR-AC-1 genes encode auxin-inducible small RNAs. High promoter activity was reported by in vascular tissues of transgenic plants harboring the SAUR-AC-1:GUS construct. The GUS expression patterns in seedling shoot apex of the BA::GUS (Fig. , C and D) and the SAUR-AC-1::GUS (Fig. , E and F) transgenic plants are similar to that of the DR5::GUS plants (Fig. , A and B), with high GUS activities in the shoot apex of 4-d-old seedlings and reduced GUS activities in the shoot apex of seedlings grown in the presence of NPA. Little GUS activity is detected in the SAM, leaf primordial, or stipules of NPA-grown seedlings. The faint GUS activity in the shoot apex of the NPA-grown SAUR-AC-1::GUS transgenic plants (Fig. F) is much lower than in non-treated SAUR-AC-1::GUS transgenic plants (Fig. E). In general, these results confirm the IAA distribution pattern during Arabidopsis leaf ontogeny determined by DR5::GUS activity and by IAA immunocytochemistry. It is presumed that NPA prevents IAA transport to the shoot apex, resulting in the absence of the IAA necessary for inducing GUS activity. IAA Distribution in DR5::GUS Seedlings Grown in the Presence of Three Different IAA Transport Inhibitors | To confirm that NPA affects IAA transport to the shoot apex, we grew DR5::GUS transgenic plants in the presence of two other IAA transport inhibitors, TIBA and HFCA. Seedlings grown in the presence of these two inhibitors had reduced IAA signal in their shoot apices (Fig. , J and K) relative to the untreated control (Fig. G), as did the NPA-grown seedlings (Fig. , H and I). The congruence of results with these three IAA transport inhibitors confirms the conclusion that the reduced IAA signal in the shoot apex is due to lack of auxin transport. Thus, the IAA in the untreated shoot apex comes from an outside source. In an attempt to identify this outside source, we examined GUS activity in the cotyledons and root tip of seedlings grown in the presence of NPA. The signal detected in root tips of NPA-grown seedlings was similar to the control (Fig. , N and O). The cotyledons of seedlings grown in the presence of NPA (Fig. M), TIBA, and HFCA (data not shown) have higher GUS activity than those grown in the absence of inhibitors (Fig. L). This result implicates the cotyledons as a probable source of the shoot apex IAA of germinating seedlings. DISCUSSION : Because polar auxin transport plays a major role in plant development, characterization of the IAA distribution pattern in relation to the processes that are regulated by IAA will contribute to our overall understanding of how organs communicate with each other to coordinate the growth of the whole plant. To elucidate the role of polar auxin transport in vascular differentiation, we investigated the temporal and spatial pattern of IAA localization in the shoot apex of Arabidopsis seedlings. We found that inhibition of auxin transport prevented IAA accumulation in the SAM and the youngest pair of leaves, but not in older leaf primordia. The temporal pattern of IAA localization implies a switch in IAA transport direction that is consistent with the mechanism of vascular differentiation proposed in the canalization of IAA flux hypothesis. IAA Immunocytochemistry | IAA distribution in Arabidopsis was determined using the anti-IAA monoclonal antibody employed to detect IAA in maize and peanut tissues. This monoclonal antibody, raised in mice (Mus musculus) against IAA conjugated to albumin through its carboxyl group , shows maximal cross-reactivity to the methyl ester of IAA in radioimmuno assay and ELISA . The antibody does recognize IAA conjugates , but Arabidopsis has little IAA-Glu conjugate, IAA-Asp conjugate, or IAA-Glc conjugate . We used EDAC to cross-link the exposed carboxyl group of free IAA to the free amino groups of structural proteins in Arabidopsis cells. IAA conjugates lack a free carboxyl and cannot be cross-linked to structural proteins by EDAC, which precludes antibody reaction with any conjugates that might be present in the tissues. The free carboxyl group of Trp can be cross-linked by EDAC, but this monoclonal anti-IAA antibody has a very low cross-reactivity with Trp . Positive controls, including adding exogenous IAA to tissues and radioactive IAA blotting assay to test the EDAC cross-linking of IAA, verified the specificity of the antibody for free IAA . The antibody failed to detect any signal in tissues not prefixed with EDAC (Fig. B), further confirming the specificity of the antibody to cross-linked IAA. Moreover, the immunological signal was absent from shoot apical tissues when IAA transport was inhibited, indicating that the signal is detecting IAA and not other compounds, such as Trp. Finally, the anti-IAA monoclonal antibody and the three different Aux/IAA promoter::GUS constructs produced similar patterns, indicating that immunocytochemistry is a reliable method for identifying IAA in Arabidopsis tissues. IAA immunocytochemistry will be an important tool for detecting IAA in very small tissues such as the SAM, in tissues where the Aux/IAA promoters are not expressed, and in plant species such as monocots in which promoter::reporter constructs are not yet available. IAA Production and Transport in Shoot Apex | Using in situ IAA determination, we have found that Arabidopsis rosette shoot apex, i.e. the SAM and the youngest set of leaf primordia, do not accumulate free IAA if IAA transport is inhibited with NPA. We observed similar depletion of IAA in seedlings treated with two other IAA transport inhibitors, confirming that the reduced IAA in the NPA-treated seedlings results from lack of IAA transport. Thus, IAA is likely transported into the shoot apex, not produced there. However, as leaf primordia grow and mature, IAA is found at the distal end of the leaf regardless of IAA transport inhibition. The presence of IAA in older leaf primordia of NPA-grown seedlings indicates that IAA can be produced in the presence of NPA. Elevated IAA in cotyledons of seedlings grown in the presence of NPA (Fig. , L and M) suggests that IAA is normally transported out of the cotyledons and that they might be the source of the IAA in the first pair of true leaves. IAA accumulation at the distal end of transport-blocked leaves implies IAA production at the leaf tip. IAA produced in the leaf tip must drain out of the leaf . Based on these results, we propose the "IAA flow model" to describe the temporal and spatial pattern of IAA flow in the shoot apex . During leaf development, the IAA source changes from extrinsic to intrinsic, which would change IAA flow direction and affect the pattern of vascular differentiation (see below) and probably cell and organ differentiation as well. The correspondence of the patterns of IAA flux and the leaf vascular formation, e.g. as seen in Figure , E and H, supports and is in good agreement with the leaf venation hypothesis recently proposed by . Figure 5 | "IAA flow model" in the shoot apex of Arabidopsis seedling. "IAA flow model" in the shoot apex of Arabidopsis seedling. The youngest leaf primordia (marked "2" for second node leaves) and SAM do not produce IAA. Rather, IAA is being acropetally transported into this region. IAA is produced at the tip and marginal regions of the older leaf primordia (marked "1" for first node leaves) and basipetally drained, primarily through the midvein. Arrows represent flow direction of IAA. The IAA source for the youngest pair of leaf primordia and SAM may be the cotyledons (C), older leaves, and shoot tissues underneath the SAM. The acropetal flow of the IAA into the leaf primordia can explain the acropetal formation of the midvein. The time of IAA appearance in the distal end of the leaf corresponds to the time of secondary vein differentiation along the leaf margin, consistent with the notion that inability to drain IAA from the leaf tip basipetally through the midvein would cause marginal hypertrophy . Role of IAA Flux on Venation Pattern | There are at least three orders of veins in the Arabidopsis rosette leaf: 1o (or midvein), 2o, and 3o, which form in a hierarchical fashion . The midvein is usually seen growing acropetally from the hypocotyl vasculature into the emerging leaf primordium . The extrinsic IAA source and acropetal flow pattern suggested by our results is consistent with the acropetal differentiation of midvein procambium at the onset of leaf ontogeny. If the 5- to 6-d-old leaf begins to produce IAA at the distal end, IAA could then flow basipetally into the plant. Seedlings germinated on NPA do not form the prominent midvein. Instead, files of veins appear, first along the leaf margins (Fig. , F and H), and afterward toward the leaf base (Fig. N), that are not connected to the hypocotyl vascular system . This is again consistent with the canalization hypothesis: Without the midvein, there is no constant basipetal flow of IAA from the leaf tip. The IAA produced at the leaf tip, the presumed IAA source of 6-d and older leaves, cannot itself be cannalized into specific cell files and cause differentiation of a vascular bundle connected to the plant vascular system. Prior differentiation of the midvein by flow of auxin into the leaf appears to be required for the later flow from the leaf tip to generate the normal venation pattern. The three pairs of secondary veins in the young rosette leaves appear as arches . As the leaf primordium grows laterally, the first pair forms as a loop from the tip of the midvein along the leaf margin that rejoins the midvein at a basal location. The secondary veins may join the midvein again because IAA flows toward the midvein, which drains the IAA out of the leaf. In the absence of a midvein in NPA-treated leaves, IAA flows along the growing leaf margin rather than toward the center of the leaf. Thus, the pair of arches is replaced by a band of vascular tissues along the leaf margin forming the marginal hypertrophy . The mechanism of IAA flows along the leaf margin is unknown, but flow is apparently independent of the IAA transport mechanism that can be disrupted by NPA, TIBA, or HFCA. Role of Localized IAA Production and IAA Transport in Organogenesis | Chemical inhibition of polar auxin transport alters venation pattern and leaf shape. The accumulation of IAA and vasculature in the upper one-half of the leaf may promote lateral leaf growth and inhibit longitudinal growth, resulting in broad leaves with short petioles (Fig. , C, D, G, and H). It has been generally believed that IAA, which plays a major role in regulating tissue differentiation and organogenesis, is produced in the SAM . Recently, a very different concept was proposed ---IAA is transported to, not produced in, the shoot meristem proper. It was further proposed that, because of the acropetally advancing procambial strand from the stem, IAA from an extrinsic source might dictate the position of new leaf primordia, phyllotaxis, and initiate organogenesis . Our results provide the first evidence in support of the concept that IAA flows from the plant to the SAM. Import of IAA into the SAM would explain impairment of organogenesis when IAA flow is blocked. Growth observed in NPA-treated, IAA-depleted shoot apices may be supported by very low levels of IAA or by other forms of auxin such as indole butyric acid . In conclusion, finding both extrinsic and intrinsic sources of IAA in leaf primordia is novel, providing both the first evidence for changing IAA flow pattern and insights into the mechanism of IAA flux-mediated venation pattern. Our results establish a foundation from which to pursue the IAA signaling pathways of phyllotaxis and organogenesis. MATERIALS AND METHODS : Plant Material and Growth Conditions | Arabidopsis ecotype Columbia-0 was used in the study. Transgenic seeds containing the promoter::GUS constructs were kindly provided by: DR5::GUS, Dr. Tom Guilfoyle (University of Missouri, Columbia); PSIAA4/5 BA::GUS, Dr. Anastasios Theologis (The Plant Gene Expression Center, U.S. Department of Agriculture, Albany, CA); and SAUR-AC1::GUS, Dr. Pamela Green (Michigan State University, East Lansing). Seeds were surface sterilized in 70% (v/v) ethanol for 1 min, followed by 10% (v/v) commercial bleach for 15 min, washed three times in sterile distilled water, and plated in 0.4% (w/v) molten agar on top of a solid germinating medium in 9-cm petri dishes. Germinating medium contained 0.5x Murashige and Skoog basal salts , 1.5% (w/v) Suc, and 0.8% (w/v) agar. Plates were sealed with parafilm, incubated at 6C in the dark for 2 d, and then transferred to a growth chamber set at 22C for a 16-h-light (120 --150 mumol m-2 s-1), 8-h-dark cycle (long-day conditions). The time of transfer to growth chamber was considered the starting point of all the experiments. Plants were transplanted into soil 2 weeks later and were grown under long-day conditions until seeds were harvested. For the inflorescence stem tissues, seedlings were germinated as described above and were grown for 2 weeks at 22C in an 8-h-light (120 --150 mumol m-2 s-1), 16-h-dark cycle (short-day conditions), then transferred to soil and grown in the greenhouse under long-day conditions (16-h-light, 8-h-dark cycle). Four- to 10-cm inflorescence stems were collected for the IAA immunocytochemical studies. Auxin transport inhibitors NPA (Chem Service, West Chester, PA), TIBA (Sigma, St. Louis), and HFCA (Sigma) were dissolved in dimethyl sulfoxide (Sigma). The concentration of the dimethyl sulfoxide in the growth media never exceeded 0.1% (v/v). IAA Immunocytochemical Localization | The monoclonal anti-IAA antibody used in the immunolocalization studies was kindly provided to us by Dr. John L. Caruso (Department of Biological Sciences, University of Cincinnati). The antibody was raised against free IAA that was cross-linked to BSA at the carboxyl group in mice (Mus musculus) . Excised tissue samples were immediately prefixed in 3% (w/v) aqueous solution of EDAC (Sigma) and postfixed in FAA (3.7% formaldehyde:50% ethanol:5% glacial acetic acid [v/v]) for 16 h at 4C, dehydrated with a graded ethanol series, embedded in paraffin, and sectioned to 10-mum slices. Sections were affixed onto slides (Probe-On Plus, Fisher Scientific, Pittsburgh). After overnight drying at 42C, sections were deparaffinized with xylene and hydrated in an ethanol-water series. Slides were processed as described in with some modifications: Slides were incubated in a blocking solution containing 10 mm phosphate-buffered saline (PBS; 2.68 mm KCl, 0.15 m Na2HPO4, and 0.086 m KH2PO4), 0.1% (v/v) Tween 20, 1.5% (v/v) Gly, and 5% (w/v) BSA, for 45 min at 22C, then rinsed in a regular salt rinse solution (10 mm PBS, 0.88% [w/v] NaCl, 0.1% [v/v] Tween 20, and 0.8% [w/v] BSA), and washed briefly with 10 mm PBS and 0.8% (w/v) BSA solution to remove the Tween 20. Fifty microliters of 1:200 (w/v) anti-IAA antibody (1 mg mL-1) were placed on each slide, covered with coverslips, and incubated overnight in a humidity chamber at 22C. Two 10-min vigorous washes with high-salt rinse solution, 10 mm PBS, 2.9% (w/v) NaCl, 0.1% (v/v) Tween 20, and 0.1% (v/v) BSA were followed by a 10-min wash with a regular salt rinse and a brief rinse with 10 mm PBS and 0.8% (v/v) BSA. Fifty microliters of 1:100 (w/v) dilution of the 1 mg mL-1 anti-mouse IgG-alkaline phosphatase-conjugate (Promega, Madison, WI) were added to each slide, which was covered with coverslip and incubated for 4 to 6 h in a humidity chamber at 22C. Two 15-min washes with a regular salt rinse were followed by a 15-min wash with water. Two hundred microliters of ready-to-use Western Blue (Promega) were added to each slide; each slide was then covered with a coverslip and incubated in the dark for 15 to 30 min. When blue/purple color was observed, slides were rinsed with water and then dehydrated in a graded water-ethanol series, ethanol-xylene, xylene. Slides were mounted with Permount (Fisher Scientific), dried overnight in a 42C oven, and observed under an Axiophot microscope (Zeiss, Jena, Germany). Photomicrographs were taken by a video camera attached to the microscope and processed with the Scion Image 1.60. The figures were arranged using Adobe Photoshop version 5.5 (Adobe, Mountain View, CA). GUS Activity | To assay GUS activity, dissected samples were incubated with 5-bromo-4-chloro-3-indolyl-beta-d-glucuronide (X-Gluc) solution as described by . Excised samples were vacuum infiltrated in the X-Gluc solution for 10 min at room temperature and then incubated at 37C in the dark for 16 h. Samples were rinsed with 50 mm sodium phosphate buffer, pH 7.0, and then fixed in ethanol:acetic acid (9:1 [v/v]) for 4 h at room temperature. X-Gluc-treated samples were rinsed with 95% (v/v) ethanol and transferred to 70% (v/v) ethanol. Tissue samples were whole mounted on microscope slides in a clearing solution of chloral-hydrate:glycerol:water (8:1:2 [v/v]) as described by . The samples were covered and observed with a Zeiss Axiophot microscope. Photomicrographs were taken as described above. For GUS staining of the zygotic embryos, ovules were dissected from siliques. To minimize wounding the embryo, a hole was punctured to allow the penetration of the X-Gluc solution into each ovule. After 16 h of incubation, embryos were dissected from the broken ovule for fixation, clearing, and photography. 35S::GUS embryos were used as control to ensure proper penetration of the X-Gluc solution through the damaged ovule. Upon request, all novel materials described in this publication will be made available in a timely manner for noncommercial research purposes. Backmatter: PMID- 12226501 TI - Arabidopsis CYP98A3 Mediating Aromatic 3-Hydroxylation. Developmental Regulation of the Gene, and Expression in Yeast AB - The general phenylpropanoid pathways generate a wide array of aromatic secondary metabolites that range from monolignols, which are ubiquitous in all plants, to sinapine, which is confined to crucifer seeds. The biosynthesis of these compounds involves hydroxylated and methoxylated cinnamyl acid, aldehyde, or alcohol intermediates. Of the three enzymes originally proposed to hydroxylate the 4-, 3-, and 5-positions of the aromatic ring, cinnamate 4-hydroxylase (C4H), which converts trans-cinnamic acid to p-coumaric acid, is the best characterized and is also the archetypal plant P450 monooxygenase. Ferulic acid 5-hydroxylase (F5H), a P450 that catalyzes 5-hydroxylation, has also been studied, but the presumptive 3-hydroxylase converting p-coumarate to caffeate has been elusive. We have found that Arabidopsis CYP98A3, also a P450, could hydroxylate p-coumaric acid to caffeic acid in vivo when expressed in yeast (Saccharomyces cerevisiae) cells, albeit very slowly. CYP98A3 transcript was found in Arabidopsis stem and silique, resembling both C4H and F5H in this respect. CYP98A3 showed further resemblance to C4H in being highly active in root, but differed from F5H in this regard. In transgenic Arabidopsis, the promoters of CYP98A3 and C4H showed wound inducibility and a comparable developmental regulation throughout the life cycle, except in seeds, where the CYP98A3 promoter construct was inactive while remaining active in silique walls. Within stem and root tissue, the gene product and the promoter activity of CYP98A3 were most abundant in lignifying cells. Collectively, these studies show involvement of CYP98A3 in the general phenylpropanoid metabolism, and suggest a downstream function for CYP98A3 relative to the broader and upstream role of C4H. Keywords: Introduction : Plants synthesize thousands of secondary metabolites from offshoots of primary metabolism . In most cases the biosynthetic routes are unknown and even in some of the well-studied pathways many aspects remain uncertain. Phenylpropanoid metabolism generates phenolic intermediates and end products that include lignin monomers, flavonoids, isoflavonoids, lignans, tannins, quinones, and sinapate esters . Lignin constitutes approximately 15% to 30% of the dry weight in woody plants, and contributes about 30% of the organic carbon in plant biomass in general . Thus, lignin assembly places a huge demand on phenylpropanoid supply. Lignification is considered a biochemical adaptation to provide mechanical strength and "non-seeping" water transport channels as plants adopted terrestrial habitats. The biosynthetic pathways appear to have been further diversified and recruited to supply metabolites for a variety of other end uses such as attraction of pollinators for promoting sexual propagation, pest deterrence, pathogen resistance, UV radiation protection, and allelopathic exclusion of potentially competing plants . The inherent inter- and intraplant variations in metabolite accumulation and environmentally modulated fluctuations make generalizations of the biosynthetic steps very tentative, but, nonetheless, composite views are useful to address specific aspects. laid out such a foundation for the pathways of general phenylpropanoid metabolism that led to the concept of l-Phe and l-Tyr entering secondary metabolism and undergoing various biochemical transformations. Deamination of l-Phe by phenylalanine ammonia lyase (PAL) is the first committed step in phenylpropanoid synthesis . Cinnamic acid is then hydroxylated at the 4-position of the aromatic ring by cinnamate 4-hydroxylase (C4H) to generate p-coumarate. Subsequently, several independent or sequential transformations occur at the 3 and 5 positions of the ring (hydroxylation and methylation) and at the gamma-carbon in the side chain. Recent studies have revealed that the order in which the ring substitutions and side chain modifications occur are not as clear-cut as previously depicted . The p-coumarate ring has been considered to undergo methoxylation successively at its 3 and 5 positions to generate 4-hydroxy-3,5-dimethoxycinnamic acid (sinapic acid) via the following intermediates in a linear pathway: 3,4-dihydroxycinnamic acid (caffeic acid), 3-methoxy-4-hydroxycinnamic acid (ferulic acid), and 3-methoxy 4,5-dihydroxycinnamic acid (5-hydroxy ferulic acid). Each of these intermediates are activated at their gamma-carbon by 4-coumarate ligase (4CL) to generate corresponding CoA thioesters that in turn are successively reduced to aldehyde and alcohol forms. These linear pathways have been revised to a metabolic grid, and the latter is supported by the following observations: 5-hydroxylation occurs favorably on aldehyde and alcohol forms derived from feruloyl CoA rather than on ferulic acid ; 0-methylation occurs on CoA esters of caffeate and 5-hydroxy ferulate, and on the respective aldehyde and alcohol forms . Figure 1 | Biochemical transformations of the ring and side chain in cinnamic acid. Biochemical transformations of the ring and side chain in cinnamic acid. It is the first phenolic acid in the general phenylpropanoid metabolism arising from deamination of l-Phe by PAL. 4-Hydroxylation of cinnamic acid by C4H generates p-coumaric acid, but further ring transformations can potentially occur when the gamma-carbon is an acid, aldehyde, or ester. In the current depiction of the pathways, 3-hydroxylation occurs on p-coumaric acid or p-coumaroyl CoA, and the later 5-hydroxylation by ferulic acid 5-hydroxylase (F5H) occurs when the gamma-carbon is in aldehyde or alcohol form. A fully methoxylated acid, i.e. 4-hydroxy,3,5-dimethoxycinnamic acid (sinapic acid) is an intermediate in sinapine and other sinapate ester synthesis. The gene encoding 3-hydroxylation has been elusive. Further details appear in the text. Of the three hydroxylases implicated in ring substitutions, C4H was the first to be studied and the best characterized cytochrome P450 monooxygenase (CYP) from plants . It is encoded by a single gene in Arabidopsis and is designated as CYP73A5 . F5H is also a P450 . Contrary to the historical nomenclature, it does not hydroxylate ferulic acid efficiently, but instead shows substrate preference for coniferaldehyde and coniferyl alcohol . Although C4H and F5H add single oxygen to the 4 and 5 positions of the phenolic ring, respectively, the situation with 3-hydroxylation has been unclear. Several types of enzymes have been implicated. Phenolases, also referred to as polyphenol oxidases, tyrosinases, and catechol oxidases, have been suggested . These soluble enzymes differ from membrane-associated P450s in structural and mechanistic characteristics. Their poor substrate specificity and lack of coordinate gene expression with such genes as PAL and C4H have cast doubts on their involvement in the general phenylpropanoid pathway. Furthermore, chemical inhibition of phenolase activity in mung bean (Vigna mungo) seedlings does not prevent formation of caffeic acid derivatives . A p-coumarate-specific hydroxylase responsible for caffeic acid formation has also been reported ; however, there is no evidence for its widespread occurrence, a central characteristic for the general phenylpropanoid-related enzymes. Interestingly, have described a parsley (Petroselinum crispum) p-coumaroyl CoA hydroxylase, thus providing another entry point for 4-hydroxy coumaroyl CoA into the phenylpropanoid grid. An enzyme capable of hydroxylating p-coumaroyl Glc in sweet potato (Ipomoea batatas) has also been purified . There are also reports that P450 enzyme(s) can catalyze the 3-hydroxylation of shikimate and quinate esters of p-coumarate to the corresponding caffeoyl esters . While this paper was in preparation, reported that Arabidopsis CYP98A3 expressed in yeast (Saccharomyces cerevisiae) can catalyze 3-hydroxylation of p-coumaroyl shikimate and p-coumaroyl quinate, but not p-coumaric acid. This finding was surprising because p-coumaroyl shikimate and p-coumaroyl quinate were not generally considered as intermediates in the general phenylpropanoid metabolism. The general phenylpropanoid network includes biosynthesis of seed-borne sinapine in crucifers . Sinapine is an antinutritional factor in canola (Brassica napus), an economically important oilseed crop, and consequently a target for elimination by breeding and biotechnology . Previously, we had characterized F5H genes from canola and showed that their transgenic suppression can result in up to 40% reduction in sinapine . We undertook the work presented here to investigate the elusive coumarate 3-hydroxylase (C3H) in Arabidopsis to gain an additional handle for metabolic engineering of sinapate ester synthesis in canola, a close relative of Arabidopsis. RESULTS : CYP98A3 as a Gene Potentially Involved in the Phenylpropanoid Metabolic Network | As mentioned above, classical biochemical approaches have not led to the identification of a C3H gene. Therefore, we employed in silico and comparative gene expression analyses to identify the putative C3H gene. We set the following criteria for identifying potential candidate(s) for C3H: (a) Its product might have some structural resemblance to other hydroxylases of the phenylpropanoid network, namely C4H and F5H; (b) Expression of the candidate gene(s) should resemble that of C4H and F5H in being active in organs that have a high demand for phenylpropanoids (e.g. stem); (c) The promoter regions of the candidate gene(s) might have some features in common with C4H and F5H, suggesting similar gene regulation; (d) The gene product should be most evident in or very near lignifying cells so as to meet the huge demand for phenylpropanoids; and (e) Expression of the putative gene(s) in yeast might afford in vivo conversion of p-coumarate to caffeate. Initially, we used simple BLASTP searches of the Arabidopsis P450 database (. org) with entire C4H and F5H amino acid sequences as the query. There were approximately 160 P450s in the database when this was done. The sequences identified as being related to both C4H and F5H were sorted out after visual examination of the identity and the scores. Those common to both C4H and F5H were accorded priority for further consideration. Among these, CYP98A3 was the highest scoring sequence, with an identity of 28% to C4H. Interestingly, CYP98A3-like expressed sequence tags had also been found among the cDNAs of xylem tissue from loblolly pine and sweetgum . These observations encouraged us to investigate the Arabidopsis CYP98A3 gene further, and a few lower ranking candidates found in our screening were also included in the subsequent analysis. RNA from stems, roots, flowers, siliques, and leaves that were young, mature, or partially senescent was probed for transcripts of the following: C4H (CYP73A5) and F5H (CYP84), CYP98A3, CYP71A19, CYP71A22, CYP71B2, CYP71B26, CYP71B35, CYP71B36, and CYP703A2 . CYP98A3 was the only one to show expression in stem, in common with C4H and F5H, and it was also the only one to resemble C4H in being highly active in root tissue . However, there were some differences, such as low-level expression of F5H in roots. Figure 2 | Expression patterns of some Arabidopsis CYP genes. Expression patterns of some Arabidopsis CYP genes. PCR-amplified segments of the CYPs were used for probing 15 mug of RNA. After initial analyses, C4H, F5H, and CYP98A3 expression was determined again for quantitation using a phosphor imager. A, The stem RNA signal was set as the reference (100%) for each probe. Expression of C4H, F5H, and CYP98A3 were analyzed successively in the indicated order after removal of the previous probe. B, Collectively, three membranes with identical RNA loading were used; a representative for rRNA loading is shown. The promoters of the phenylpropanoid metabolism genes in many species contain conserved cis sequences that are referred to under various names . MYB transcription factors are considered to regulate certain phenylpropanoid genes via interactions with MACCWAMC (M = A/C; W = A/T) elements present in them . In the Arabidopsis C4H promoter region, have noted sequences resembling the "L box" (two; YCYYACCWACC; Y = C/T), and "P box" (four; YTYYMMCMAMCMMC); some of these overlap with or include the "H box"-like sequence and "Box 3"-like element found by . Some of the above P and L boxes comprise the MYB-binding consensus sequence (MACCWAMC; MYB element). We searched the 2-kb sequence upstream of the CYP98A3 open reading frame (ORF) and, for comparison, the 2-kb sequence upstream of Arabidopsis F5H ORF for perfectly conserved MYB elements. Five were found in CYP98A3 (-143 to -136, -1,116 to -1,109, -1,661 to -1,654, -1,674 to -1,667, and -1,779 to -1,772) and two in F5H (-90 to -83 and -1,126 to -1,119). In addition, A and H box-like sequences that were imperfectly conserved among C4H, F5H and CYP98A3 were also found. These provided additional impetus to characterize CYP98A3. Expression of CYP98A3 in Yeast Affords 3-Hydroxylation of p-Coumaric Acid to Yield Caffeic Acid | We used the P450 expression system of to investigate CYP98A3 for hydroxylase activity. Plasmid pRAM51, a recombinant pYeDP60 with the CYP98A3 ORF, was introduced into the yeast WAT21 strain that produces an Arabidopsis P450 reductase, and the microsomal fractions of the pRAM51 and pYeDP60 strains were assayed for hydroxylation of cinnamic acid, p-coumaric acid, ferulic acid, coniferaldehyde, and coniferyl alcohol at substrate concentrations of 20 mum, 0.1 mm, and 1.0 mm. Even after 1 h of incubation at 30C, no novel UV-absorbing products were detectable by HPLC. Assuming that this might be due to a slow reaction that escaped detection in the microsomal assays, we then assayed the yeast cultures supplemented with the above substrates for in vivo production of novel products. There was a new but very small peak in WAT21 (pRAM51) supplemented with p-coumarate but not with the other substrates. This product was found to be indistinguishable from caffeic acid according to three criteria : (a) retention time in HPLC, (b) UV absorbance spectrum, and (c) analysis of the parent and daughter ions by liquid chromatography (LC)/negative ion electrospray mass spectrometry. The WAT21 (pYeDP60) control cultures did not produce this peak. Thus, the production of caffeic acid could be attributed to CYP98A3 in the recombinant yeast cells. The caffeic acid production analyzed over a period was linear , and the apparent rate of caffeic acid production was 4.3 x 102 pmol h-1 for a unit of 2 x 108 cells and 1.2 x 103 pmol h-1 for 1-mg protein content of the cells. These results showed that CYP98A3 could hydroxylate p-coumaric acid in vivo. Figure 3 | Hydroxylation of p-coumaric acid by yeast WAT21 (pRAM51) cells that produce Arabidopsis CYP98A3. Hydroxylation of p-coumaric acid by yeast WAT21 (pRAM51) cells that produce Arabidopsis CYP98A3. Inset, Immunoblot of 15 mug of protein from the control strain containing the vector (pYeDP60) and the recombinant strain fractionated on SDS-PAGE. The new peak appearing in the CYP98A3+ strain matched an authentic caffeic acid standard. Negative ion electrospray tandem mass spectrometry identified a predominant daughter ion (m/z 135) from the parent ion (m/z 179). Figure 4 | CYP98A3-catalyzed production of caffeic acid in yeast WAT 21 cells. CYP98A3-catalyzed production of caffeic acid in yeast WAT 21 cells. The cells were grown in YPLA medium supplemented with 5 mmp-coumaric acid. Samples withdrawn at indicated times were analyzed for caffeic acid content by HPLC as described in "Materials and Methods." Inset, Caffeic acid standard plot constructed with authentic sample. Comparative Analysis of the Developmental Regulation of CYP98A3 and C4H | The results above and the commonality of CYP98A3 and C4H gene expression in various tissues were intriguing to warrant a comparison of the developmental regulation of the two genes in Arabidopsis. We reasoned that this comparative expression profiling would be informative with regard to the functional relationships of CYP98A3 with C4H. The 5'-upstream promoter sequences of the CYP98A3- and C4H-coding regions retrieved by PCR were fused to the ORF of the Escherichia coli beta-glucuronidase (GUS) gene . At least 20 independent Arabidopsis transgenic lines were produced for each construct and analyzed initially for GUS expression. Five second generation transgenic lines were analyzed for detailed GUS activity in seedlings grown on Murashige and Skoog agar medium and in plants grown on soil . The vascular bundles in stem, petiole, leaf, and silique wall showed evidence of promoter activity in both CYP98A3::GUS and C4H::GUS plants (Fig. , a --c, e, g --i, and k). The GUS staining was always more intense in the C4H::GUS plants than in the CYP98A3::GUS plants. The observed differences between these included an overall GUS staining of the flowers in the C4H::GUS plants (Fig. j), but in the case of the CYP98A3::GUS flowers (Fig. d), only in the vascular tissues of petal, sepal, anther, and stigma. The most contrasting characteristic between these two was the absence of GUS staining in the seeds of CYP98A3 lines, whereas it was very intense in C4H transgenic seeds (Fig. , e, f, k, and l). This difference was evident in transgenic tobacco (Nicotiana tabacum) seeds as well (Fig. , u and v), and the spatial expression of GUS in other parts of transgenic tobacco plants was comparable with that in Arabidopsis (data not shown). In Arabidopsis seedlings, both CYP98A3 and C4H promoters were most active in roots (Fig. , m and q). However, there were some zonal and tissue-specific differences: Unlike the C4H promoter (Fig. t), the CYP98A3 promoter was not active in the apical meristem of roots (Fig. p), and it was active only in stele and endodermis but not in epidermis or cortex, whereas the C4H promoter was active in all cells as shown in the cross sections of the upper and lower regions of the roots. Wounding induced both C4H and CYP98A3 promoters (Fig. , n and r). Thus, the two genes were generally comparable in spatial expression. Figure 5 | Developmental regulation of CYP98A3 and C4H. Developmental regulation of CYP98A3 and C4H. Transgenic plants with promoter::GUS fusion constructs were analyzed for 5-bromo-4-chloro-3-indolyl-beta-d-glucuronide (X-gluc) staining. a, g, c, i, p, and t, Compound microscope; the remainder are from a dissecting microscope. p and t, Differential interference contrast microscopy. The material in a through l and u and v were from soil-grown plants, and the rest from seedlings germinated on Murashige and Skoog agar. The deliberately wounded parts of excised leaves are marked in boxes. Hy, Hypocotyl; Pr, primary root; Lr, lateral root. Cross sections: s, stele; e, endodermis; c, cortex. Bars: 100 mum for p and t; 400 mum for a through l, o, s, u, and v; 1 mm for n and r; and 4 mm for m and q. CYP98A3 Protein Localizes to Lignifying Cells of Arabidopsis Stems and Roots | The common characteristics of CYP98A3 and C4H in the analyses described above suggested that CYP98A3 might also be produced in lignifying cells. This was found to be the case in immunolocalization by polyclonal antibodies raised against a truncated CYP98A3 expressed in recombinant E. coli. Although the pre-immune serum did not bind to lignifying tissue, the antiserum did. The meta- and protoxylem cells undergoing primary lignification in young stem showed a positive immunochemical reaction, and the lignified interfascicular fibers and xylem vessels in older stem showed a stronger reaction . This correlated well with the lignin deposition patterns in Arabidopsis . In root, the stele cells were reactive with the antiserum, providing further evidence that CYP98A3 was associated with lignification. Figure 6 | CYP98A3 localization, and lignification. CYP98A3 localization, and lignification. Immunochemistry with polyclonal antiserum raised against a truncated CYP98A3 produced in E. coli (inset). Paraffin sections of stem (a --d) and root (e) were probed with pre-immune serum (a) or antiserum (b --e). Hand sections of stem (f --h) or root (i) were stained with phloroglucinol for lignin. Bar = 100 mum in a through e and 400 mum in f through i. Arrows indicate tissues that are lignified. vb, Vascular bundles; if, interfascicular fiber; st, stele. DISCUSSION : We have extensively characterized the developmental regulation of CYP98A3 and have shown that it resembles in many respects C4H that encodes the archetypal hydroxylase of phenylpropanoid metabolism. Taken together with the finding that CYP98A3 was most evident in lignifying cells and that it was capable of hydroxylating p-coumarate to caffeate, albeit at a very slow rate in yeast cells, it was attractive to consider CYP98A3 as a phenylpropanoid network enzyme. Historically, all such evidence has been used in assigning the role of a given gene to a specific function in phenylpropanoid biosynthesis, and, as discussed below, recent evidence shows that CYP98A3 is involved in general phenylpropanoid metabolism. CYP98A3, the Elusive 3-Hydroxylase of the General Phenylpropanoid Metabolism | The independent in silico gene discovery approaches employed by and us entailed assumptions of a potential 3-hydroxylase being structurally similar to the previously characterized hydroxylases of phenylpropanoid metabolism. These strategies were productive in the identification of CYP98A3, even with the following caveat that some P450s are enigmatic: Nearly identical P450s (in one example differing by just one amino acid) can mediate reactions on unrelated substrates and, conversely, dissimilar P450s can use a common substrate . A 3-hydroxylase of the latter type, if present, would have escaped detection at the outset. In recombinant yeast cells, CYP98A3 converted p-coumaric acid to caffeic acid, but so slowly that the reaction escaped detection in in vitro assays. The rate of caffeic acid production contrasts highly with that of the production of p-coumaric acid by C4H-expressing recombinant yeast cells; the latter is 1.6 x 106 pmol h-1 for 2 x 108 cells, as calculated from the data of , and, thus, 3,700-fold greater than the caffeic acid production. could not detect p-coumaric acid hydroxylation by yeast-produced CYP98A3 in in vitro assays. While our paper was in revision, have also reported that the CYP98A3 activity with p-coumaric acid in vitro was too low to permit accurate measurement of the catalytic properties, even lower with p-coumaraldehyde, and undetectable with p-coumaryl alcohol. With regard to other potential substrates, p-coumaroyl CoA, p-coumaraldehyde, p-coumaryl alcohol, or 1-O-glucoside or 4-O-glucoside esters of p-coumaric acid are not hydroxylated in vitro . However, CYP98A3 hydroxylates p-coumaroyl shikimate (Kcat/Km = 87 min-1 mum-1; Km = 7 mum) and, at a 4-fold less efficiency, p-coumaroyl quinate (Kcat/Km = 22 min-1 mum-1; Km = 18 mum) to the corresponding caffeoyl esters . The apparent Km for the shikimate ester is similar to that reported for C4H toward cinnamic acid (4 --10 mum; , ) and for F5H toward coniferaldehyde (1 mum) and coniferyl alcohol (3 mum; ). The poor catalytic activity of CYP98A3 in p-coumaric acid hydroxylation, as surmised from our in vivo measurements, is reminiscent of the observations that Arabidopsis F5H and its corresponding enzyme from sweetgum hydroxylate coniferaldehyde at 150- to 1,000-fold greater efficiency in comparison with ferulic acid as the substrate . Although all these observations need to be considered in light of potential metabolic channeling that has been inferred for phenylpropanoid intermediates , they do suggest that shikimate and quinate esters are likely the physiological substrates, as originally conceived by and , and that p-coumaric acid is not. The monooxygenase(s) present in carrot (Daucus carota; ) and parsley may well be encoded by orthologs of the Arabidopsis CYP98A3. How exactly caffeoyl shikimate and quinate esters fit into the general phenylpropanoid metabolism is unclear. Typical with the construction of the metabolic maps of phenylpropanoids, only a composite picture from multiple species is available. The shikimate and quinate esters are not methylated , but caffeoyl CoA is . Because caffeoyl shikimate or quinate do not accumulate in Arabidopsis, these esters might be acted upon by a reversible hydroxycinnamoyl CoA p-hydroxycinnamoyltransferases activity to generate caffeoyl CoA as proposed by . In parsley, caffeoyl CoA can be produced by a coumaroyl CoA 3-hydroxylase . In Arabidopsis, however, this route is either nonexistent or insignificant as deduced from the following observations: p-coumaroyl CoA is not hydroxylated by CYP98A3 ; the Arabidopsis ref8 mutant, with a chemically induced mutation in CYP98A3, accumulates p-coumaroyl esters, instead of sinapoyl esters, in its leaves and synthesizes lignins derived almost exclusively from p-coumaryl alcohol . Thus, the genetic evidence not only confirms CYP98A3 mediating aromatic 3-hydroxylation but also shows that a distinct coumaroyl CoA 3-hydroxylase, if present in Arabidopsis, cannot appreciably substitute for CYP98A3. All of these studies necessitate a revision to the general phenylpropanoid metabolism to include an obligatory esterification at the gamma-position before hydroxylation at the 3 position , and furthermore portend additional revisions in regards to the formation of the 3-methoxy intermediate. Developmental Regulation of CYP98A3. Commonality and Dissimilarity of CYP98A3 and C4H | CYP98A3 promoter activity, like C4H activity, was most abundant in vascular tissue. In addition, the CYP98A3 protein was most evident in lignifying cells in stem and root, which require a large supply of phenylpropanoids. Both promoters were also wound inducible, another characteristic of lignin biosynthesis-related genes . CYP98A3 gene expression was strong in roots (approximately 280% of stem; all comparisons are for a given probe, relative to the signal in stem sample as shown in Fig. ) and resembled C4H (approximately 230%) in this respect. This differs from the results of that show a high level expression of CYP98A3 in stem but much less in roots. In siliques, C4H expression was much more robust (approximately 300%) than in stems; CYP98A3 also showed this trend (approximately 170%). The histochemical analyses of promoter function portrayed the commonality between CYP98A3 and C4H expression in general, and, in addition, showed some specific differences. Although C4H promoter activity was found in all cell layers of the root, CYP98A3 promoter function was localized to the stele and endodermis where the cells were lignified. Because C4H catalyzes the pivotal hydroxylation step that generates p-coumarate, which in turn is a common precursor for the biosynthetic pathways of monolignols, sinapates, and flavonoids, the broad gene expression is explicable on the basis of the upstream function for C4H. Notable in this regard is the in situ localization of chalcone synthase and chalcone isomerase to the epidermal and cortex cells of the primary root where flavonoids accumulate . The absence of CYP98A3 promoter activity in these locations is consistent with a function of CYP98A3 occurring downstream of C4H, but not required for flavonoid biosynthesis. Most of the phenylpropanoid genes that exist as multigene families are differentially expressed. For example, the members within the Arabidopsis PAL, 4CL, and CCR gene families are differentially regulated . For single-copy genes, a higher basal level of gene expression might be necessary in those organs that potentially have a sudden and high demand for their activity. Considering this potential requirement, the deviations in the expression of CYP98A3 and C4H are likely to have some functional implications. The high transcript level of CYP98A3 in roots is intriguing. Given that roots are in intimate contact with other organisms in the biosphere, both CYP98A3 and C4H may also be involved in synthesizing phenylpropanoids pertaining to interactions with these organisms. For example, some phenylpropanoids have antimicrobial activity . Interestingly, F5H expression in roots was relatively low (17%), suggesting that the products further downstream of F5H action might not have such a function. The implications of the differential expression in roots remain to be unraveled, but these observations place C4H and CYP98A3 ---the relatively upstream hydroxylases ---in one category and the downstream F5H in another. The phenylpropanoid pathway in Arabidopsis generates leaf- and seed-associated sinapate esters such as sinapoylmalate and sinapine, respectively. Sinapine synthesis, a distinct hallmark of crucifers, occurs only in seed tissue, and is a subject of our interest . Although the CYP98A3 promoter was active in the silique walls of transgenic Arabidopsis, it was not active in seeds. The ref8 mutant is impaired in sinapine synthesis in Arabidopsis seeds , showing that CYP98A3 in its native context does function in seeds. Together, our observation of the lack of GUS expression in seeds with the promoter construct suggests the presence of a hitherto uncharacterized seed-active cis element(s) in the CYP98A3 gene outside of the 2.2-kb promoter region included in our study. These element(s) might offer a means to differentially regulate this single-copy gene in seeds. MATERIALS AND METHODS : Plant Growth Conditions and Chemicals | Arabidopsis (Columbia) was used throughout. Unless stated otherwise, plants were grown in pots containing RediEarth (Grace Horticultural Products, Ajax, ON, Canada) in a chamber under a 16-h-light/25C and 8-h-dark/22C cycle with a light intensity of 380 mumol m-2 s-1 photosynthetic photon flux density. Cinnamic acid, p-coumaric acid, caffeic acid, ferulic acid, coniferaldehyde, and coniferyl alcohol were from Sigma-Aldrich (Oakville, ON, Canada). DNA and RNA Analysis | Young leaves, mature leaves, and roots were collected from 4-week-old plants. Main inflorescence stems, partly senescent leaves, siliques, and flowers, were collected from 8-week-old plants. Roots were washed with double-deionized water to remove the soil. All tissues were frozen in liquid nitrogen and stored at -80C. All DNA and RNA extractions and analyses were as in . A Molecular Dynamics PhosporImager:SI and ImageQuaNT program (Amersham, Sunnyvale, CA) were used for quantitating the RNA signal intensity according to the supplier's instructions. The CYP probe DNAs were amplified by PCR with the following primers: CYP98A3 (atgtcgtggtttctaatagc and aaggctagccgcgttatgttgt), CYP71A19 (tgtgcttaacaacgctcctt and cttcttgtaagaccggacca), CYP71A22 (gaagaaaagcaacacacctg and gttggtgaggagatggagga), CYP71B2 (cgatcttgctctgtttcttc and gatactagcggtgagggaga), CYP71B26 (tcgacgaataccatctcctc and tgactgcagagcttccttag), CYP71B35 (ggcttctgtcacttatcttc and gtcctgcagaccacactaac), CYP71B36 (ttgtattcttctagccgcct and tcgctcaagttaaccggagt), and CYP703A2 (atgattttcgtgctagcctc and cttgggcttcttttgggcta). GenBank entries are as follows: CYP98A3 sequence, locus At2g40890 and bacterial artificial chromosome ; and F5H sequence, locus At4g36220 and bacterial artificial chromosome . The FINDPATTERNS program in the GCG software suite (Genetics Computer Group, Madison, WI; ) was used for sequence analyses. Reverse Transcriptase-PCR | First-strand cDNA was synthesized using 10 mug of total RNA as described in . PCR was set in a 50-muL reaction volume containing 2.5 muL of the cDNA, 1x Pfu polymerase buffer (Stratagene, La Jolla, CA), 200 mum of each dNTP, and 50 pmol of each of the gene-specific primers (5'-gcggatccg-atgtcgtggtttctaatagcggtgg-3' and 5'-gcgaattca-ttacatatcgtaaggcacgcgt-3') to amplify the CYP98A3 ORF. After initial denaturation of DNA for 2 min, PCR was conducted for 25 cycles with 2.5 units of Pfu DNA polymerase (Stratagene) in a DNA Thermal Cycler (Perkin-Elmer Applied Biosystems, Foster City, CA) at a setting of 94C for 45 s, 56C for 1 min, and 72C for 4 min for each cycle. The PCR product was purified, digested with BamHI and EcoRI (Life Technologies, Rockville, MD), and cloned into pBluescript SK- (Stratagene). The ORF sequence of CYP98A3was confirmed by sequencing and the plasmid termed as pRAM48. Genetic Transformation of Plants | The plant vector pRD420 derivatives containing the promoter from Arabidopsis C4H (CYP73A5) or CYP98A3 as a HindIII-BamHI segment was fused to GUS ORF. These vectors were transformed into Agrobacterium tumefaciens GV3101 (pMP90). The promoters were originally retrieved from Arabidopsis genomic DNA by Pfu Turbo (Stratagene)-mediated PCR. 5'-gcgaagctt-tatttcctgcaaaagatgttataatg-3' and 5'-gcggatcc-gaagttttgcttctatttttattttcgg-3' were used for retrieving 2.2 kb immediately upstream of the CYP98A3 ORF, and 5'-gcaagctt-agaggagaaactgag-3' and 5'-gcggatcc-tatagtttgtgtatccgcaatgatattg-3' for 2.9 kb of the C4H promoter sequence. Arabidopsis plants with 3- to 5-cm stem bolts were transformed using the floral dip method of . Tobacco (Nicotiana tabacum) plants were transformed as described previously . Production of CYP98A3 in Yeast (Saccharomyces cerevisiae), and Biochemical Analyses | The WAT21 strain, pYeDP60, microsome preparation and assays were as in with the indicated modifications. The CYP98A3 ORF from pRAM48 was subcloned between the BamHI and EcoRI sites of pYeDP60 to generate pRAM51. The enzyme assays of microsomal preparations were done in a 500-muL reaction according to the method of : 450 muL of 100 mm sodium phosphate buffer (pH 7.4), containing 1 mm NADP+, 10 mm Glc 6-phosphate, and 1 unit of Glc-6-phosphate dehydrogenase (Sigma-Aldrich), was pre-incubated at 30C in the presence of one of the substrates (cinnamic acid, p-coumaric acid, ferulic acid, coniferaldehyde, or coniferyl alcohol) at three different concentrations of 20 mum, 0.1 mm, or 1.0 mm. The reaction was initiated by the addition of 5 or 50 mug of microsomes from the vector control or pRAM51 cells in 50 muL, allowed to proceed at 30C for 15 or 60 min, and stopped by adding 50 muL of trifluoroacetic acid. The extract supernatant was analyzed by HPLC (Nucleosil C18 AB column, Alltech, Deerfield, IL) using an acetonitrile/phosphoric acid (1.5% [v/v]) gradient of 0% to 25% (v/v) acetonitrile over a 40-min period in a 60F multisolvent delivery system (Waters, Milford, MA) fitted with a Waters 600 controller. A330 was determined with a Waters 996 photodiode array detector. The in vivo assays and the media compositions were as in , except for supplementation with the phenolics used here. A saturated culture of yeast grown at 28C in 2 mL of N3AT medium was subcultured into 100 mL of YPGE medium for 24 h. Fifty milliliters of the culture was centrifuged and resuspended in 100 mL of YPLA medium, and was induced for 6 h with 2% (w/v) Gal. Ten milliliters of the induced culture was added to 100 mL of YPLA containing 2% (w/v) Gal and one of the following phenolics at 0.1, 1.0, or 5 mm: cinnamic acid, p-coumaric acid, ferulic acid, coniferaldehyde, and coniferyl alcohol. One-milliliter samples were then drawn at 0, 16, 24, 40, 48, 64, 72, and 88 h after incubation at 28C, and 500 muL of the supernatant was added to 100 muL of trifluoroacetic acid, and 20 muL of the supernatant analyzed by HPLC as described above. Negative ion electrospray tandem mass spectrometry analysis for caffeic acid was done by the Mass Spectrometry Unit with a Quattro-LC mass spectrometer (Micromass, Manchester, UK) following LC in an HP1100 HPLC (Agilent, Palo Alto, CA). The LC was done with a Genesis C18 reverse phase column (2.1 x 100 mm; Jones Chromatography, Lakewood, CO) and a mobile phase of aqueous (12 mm acetic acid, pH 3.3) solution with acetonitrile increasing from 0% to 25% (v/v) over 40 min. Multiple reaction monitoring was for caffeic acid (m/z of 179) and its most abundant daughter ion (m/z of 135) generated by argon collision. Production of CYP98A3 Antisera, and Immunoblot | A partial CYP98A3 cDNA (bp 309 --1,141 of the ORF) amplified using Pfu and the primers 5'-gaggatcc-tagccgcaacggtcag-3' and 5'-gcaagctt-agcctccgatcttgacatct-3' was ligated into the BamHI-HindIII sites of pRSET B vector (Invitrogen, Carlsbad, CA) to give pRAM45. Escherichia coli BL21(DE3) pLysS strain (Invitrogen) containing pRAM45 or pRSET B was grown overnight at 30C in 2 mL of Luria-Bertani medium, subcultured in 50 mL of Luria-Bertani medium, and grown to A600 of 0.7. Isopropyl beta-d-1-thiogalactopyranoside was added to 0.1 mm, and incubated for 3 h before harvesting and lysing in 6 m urea, pH 8.0. The protein was purified on nickel-nitrilotriacetic acid agarose columns (Qiagen, Valencia, CA) by the manufacturer's protocol. The 34-kD polypeptide, excised from an SDS-PAGE (10% acrylamide, w/v), was used to raise antibodies in rabbits at the facilities of Veterinary and Infectious Disease Organization (Saskatoon, SK, Canada). Western blotting was as described in . Pre-immune serum was used as a control in all immunoblot and immunolocalization experiments. GUS Activity in Situ, Lignin Staining, and Immunohistochemistry | Hand sections or explants were incubated in a solution of X-gluc [1 mm X-gluc in 50 mm phosphate buffer (pH 7.2) containing 0.5 mm K4Fe(CN)6.H2O, 0.5 mm K3Fe(CN)6, 10 mm EDTA, and 0.1% (v/v) Triton X-100] at 37C overnight, destained, and rinsed in 70% (v/v) ethanol. Seedlings were stained for GUS according to . Whole root tips were mounted in 50% (v/v) glycerol. For cross sections, the GUS-stained root tips were fixed, dehydrated, and embedded in paraffin as described in . Counterstaining was done with 0.01% (w/v) Safranin in deionized water for 15 min, and the sections were mounted with Cytoseal (Stephens Scientific, Kalamazoo, MI). Phloroglucinol stains lignin and hand-sectioned stem or root was stained with 1% (w/v) Phloroglucinol (Sigma-Aldrich) in 6 n HCl as described in . Immunolocalization methods for tissue preparation and detection were as in , with CYP98A3 antiserum at 1:500 (v/v) dilution. Backmatter: PMID- 12226502 TI - Regulated Expression of Arabidopsis Phosphate Transporters AB - Phosphorus deficiency is one of the major abiotic stresses affecting plant growth. Plants respond to the persistent deficiency of phosphate (Pi) by coordinating the expression of genes involved in alleviation of the stress. The high-affinity Pi transporters are among the major molecular determinants that are activated during Pi stress. In this study, using three reporter genes (green fluorescent protein, luciferase, and beta-glucuronidase) regulated by two Pi transporter promoters, we have carried out an extensive analysis of transcriptional and spatial regulation of gene expression. Activation of the genes was rapid, repressible, and specific in response to changes in Pi availability. The phytohormones auxin and cytokinin suppressed the expression of the reporter gene driven by the AtPT1 promoter, and that of the native gene, suggesting that hormones may be involved in regulation of some component(s) of Pi starvation response pathway. These studies also provide molecular evidence for a potential role of high-affinity Pi transporters in mobilizing Pi into reproductive organs. The results suggest that members of the Pi transporter family may have similar but nonredundant functions in plants. Keywords: Introduction : Phosphorus is one of the major, but least available, essential nutrients acquired by plants. The evolution of plants in environments where phosphate (Pi) availability is low in the rhizosphere has led to numerous adaptations required for the survival of plants . These include the capacity to enhance Pi uptake during nutrient deficiency . This increase has been correlated with an increased number of high-affinity Pi transporters assembled in the plasma membrane . The high-affinity Pi transporters, encoded by a small family of genes, translocate Pi from the external media containing very low levels of the nutrient into the cytoplasm . Nine genes with a high degree of similarity to high-affinity Pi transporters have been identified in the Arabidopsis genome. Two members of this family have been shown to function as the high-affinity Pi transporters . Most of the cloned Pi transporters are expressed preferentially in roots under Pi starvation, consistent with their role in the nutrient acquisition . In addition, Pi transporters are also involved in in planta translocation of the nutrient. It is generally assumed that low-affinity Pi transporters are responsible for mobilization of Pi within the plant . However, expression of some of the high-affinity Pi transporters in plant parts other than roots, including stem, leaf, tuber, and flowers, points to their potential involvement in internal translocation of Pi . Ion uptake kinetic studies have shown that high-affinity Pi transport activity is inducible during Pi starvation, whereas the low-affinity transport remains constitutive . This observation was further substantiated by the constitutive expression of the low-affinity Pi transporters in Arabidopsis . Furthermore, the transcript levels of high-affinity Pi transporters increased rapidly as a specific response to Pi starvation . There is growing evidence supporting the transcriptional activation of high-affinity Pi transporters during Pi deficiency . An effective approach to study transcriptional regulation is to monitor the activity of reporter genes driven by the specific gene promoter. Expression of reporter genes such as luciferase (LUC), green fluorescent protein (GFP), and beta-glucuronidase (GUS) has been successfully used to understand transcriptional regulation in plants . Although GUS is an ideal reporter for histochemical analysis, the destructive nature of the staining procedure makes it unsuitable for sequential analysis of effects of multiple inducers. Perhaps the most useful aspect of using LUC as a reporter is its noninvasive assay procedure. This unique feature makes it possible to study the effects of multiple factors influencing gene expression by sequentially imposing different stimuli. However, both GUS and LUC reporter systems require the addition of substrates to detect their activity. In contrast, the detection of GFP is noninvasive and does not require any substrate. Illumination of GFP expressing plants with long-wave UV (395 nm) or blue light (475 nm) results in bright-green fluorescence (510 nm). To take advantage of the usefulness of each of these reporter systems, we have generated transgenic plants expressing reporter genes under the regulation of AtPT1 and AtPT2 promoters. Plants exhibit an array of physiological changes, as well as morphological and architectural modifications in roots, in response to altered availability of phosphorus . Some of these responses, particularly changes in root morphology and architecture, are very similar to those resulting from phytohormone treatments. Auxin and ethylene are known to have profound effects on root growth, root hair initiation, and elongation . Ethylene is involved in the lateral root development, root hair initiation, and elongation . Interestingly, Pi deficiency that leads to similar morphological changes also enhances ethylene production in plants . It has been suggested that Pi starvation-induced changes in root morphology may involve both ethylene-dependent and -independent pathways . Application of auxin to Pi-sufficient white lupin (Lupinus albus) resulted in the formation of proteoid roots, a response commonly observed under Pi deficiency . In addition, a decrease in the level of cytokinins during Pi deficiency has been correlated with altered root morphology . Although these findings revealed the involvement of phytohormones in Pi deficiency responses, not much is known about their role in the Pi starvation-induced signaling pathways or in gene expression. Inhibitors of auxin transport and ethylene synthesis have been shown to influence the expression of genes during Pi deficiency in the proteoid roots of white lupin . A recent report showing the suppression of Pi starvation-induced gene expression by cytokinin is quite interesting . The cytokinin effects are presumed to be due to altered long distance signaling during Pi starvation. In this study, we have carried out an extensive analysis of transcriptional regulation and tissue-specific expression of three reporter genes driven by the promoters of two high-affinity Pi transporters of Arabidopsis. The role of hormones such as auxin, ethylene, and cytokinin in the regulation of Pi starvation-induced gene expression is also examined. Results show that the transporters are transcriptionally regulated by Pi in a rapid and reversible manner. This study also showed that auxin and cytokinin specifically suppressed expression of the high-affinity Pi transporter AtPT1. This supports the notion that high-affinity Pi transporters are involved in nutrient remobilization during Pi starvation. RESULTS : Generation of Transgenic Arabidopsis Plants | To understand the transcriptionally regulated tissue specific expression of Pi transporters in Arabidopsis, we have generated transgenic Arabidopsis plants expressing reporter genes under the regulation of the AtPT1 and AtPT2 promoters. About 25 independent transgenic lines from each construct (AtPT1-GUS/GFP, AtPT2-GUS, and AtPT2-LUC) were selected and evaluated for the expression of reporter genes in response to Pi starvation. One transgenic line from each category with representative expression was chosen for detailed analysis of expression. Because AtPT2 gene is specifically induced under Pi starvation, transgenic plants carrying the AtPT2 promoter were subjected to detailed analysis of reporter gene activity. This analysis provided evidence that the reporter genes are under the transcriptional regulation of Pi transporter promoters . Figure 1 | Pi starvation-induced expression of reporter genes. Pi starvation-induced expression of reporter genes. A, Transgenic Arabidopsis expressing the reporter genes GUS (1 and 2) and GFP (3 and 4) under the regulation of AtPT1 promoter are shown. 1 and 3, Expression of reporter genes under Pi sufficiency conditions. 2 and 4, Expression under Pi deficiency conditions. B, Transgenic Arabidopsis expressing the reporter gene GUS (1 and 2) or LUC (5 and 6) under the regulation of AtPT2 promoter are shown. 1 and 5, Absence of expression of GUS and LUC in plants grown under Pi sufficiency. Expression of GUS and LUC under Pi-starved conditions is shown in 2 and 6, respectively. 3 and 4, Pi-sufficient and -deficient plants used for monitoring LUC expression in roots. Pi Deficiency Is the Underlying Factor in Gene Expression | One of the important features of high-affinity Pi transporters is their inducibility by low Pi concentrations . To better understand this response mechanism, a detailed analysis of the effect of different Pi concentrations was carried out. Seven-day-old seedlings of transgenic Arabidopsis lines grown in one-half-strength Murashige and Skoog medium were transferred to Murashige and Skoog medium supplemented with different concentrations of Pi (0 --1.25 mm) for 5 d. LUC activity in plants (AtPT2-LUC) grown in the presence of 1.25 mm Pi was barely detectable (Fig. A). A dramatic increase in the reporter gene activity was noticed with decreasing Pi concentration, which reached nearly 85-fold at 25 mum. Further lowering of Pi concentration did not significantly alter the level of expression of the reporter gene (Fig. A). Analysis of plants expressing GUS under the regulation of AtPT2 promoter also showed a similar trend in the reporter gene activity (Fig. B). The GUS activity was quite obvious in transgenic plants carrying the AtPT1 promoter, even under Pi sufficiency conditions. Transferring these plants to Pi-deficient conditions increased the activity by approximately 2-fold (Fig. A). Figure 2 | Effect of Pi concentration on gene expression. Effect of Pi concentration on gene expression. A, Seven-day-old seedlings of AtPT2-LUC plants were transferred to Murashige and Skoog media containing varying concentrations of Pi (0, 5, 10, 25, 50, 125, 250, 500, and 1,250 mum) for 5 d. Plants were harvested for measuring the LUC activity and RNA isolation. B, Seven-day-old seedlings of AtPT2-GUS plants were grown in the presence of different concentrations of Pi as indicated for 5 d and utilized for GUS reporter expression analysis. C, Northern analysis of total RNA from AtPT2-LUC plants supplemented with different concentrations of Pi. Ten micrograms of total RNA was electrophoretically separated on denaturing formaldehyde agarose gels and blotted on to nylon membranes. The membranes were probed with 32P-labeled AtPT1 and AtPT2 cDNAs. The ethidium bromide-stained gel picture shows uniform loading of RNA samples. Figure 3 | Activity of AtPT1 promoter in response to Pi starvation. Activity of AtPT1 promoter in response to Pi starvation. Seven-day-old AtPT1-GUS/GFP expressing plants were transferred to liquid Murashige and Skoog medium with (1,250 mum) and without (0 mum) Pi for 5 d. Soluble protein was extracted from the Pi-treated plants to determine GUS activity. The enzyme activity was measured using 4-methyumbelliferyl-beta-d GlcUA as the substrate. Error bars show sd. B, Northern-blot analysis of total RNA isolated from Arabidopsis plants grown in the presence (P+) and absence (P-) of Pi for 5 d. The blot was probed with 32P-labeled AtPT1. The ethidium bromide-stained gel shows uniform loading and integrity of RNA. Interestingly, the expression of the native genes (AtPT1 and AtPT2) parallels the expression and activity of the reporter genes. A strong correlation between decreasing Pi concentration and increasing transporter-specific message accumulation was observed (Fig. C). When the Pi in the medium was sufficient (1.25 mm), the expression of Pi transporters either remained at basal level (AtPT1) or was barely detectable (AtPT2). A significant increase in the abundance of these transcripts was observed at 500 mum Pi and it reached a maximum in plants grown in the medium lacking Pi. These results suggest that reporter gene activity could be used as an indicator of transcriptional activation of native gene expression under Pi starvation. Transcriptional Activation of Gene Expression Is an Early and Reversible Response to Pi Availability | The inducible and repressible mode of Pi uptake based on internal Pi requirement is well documented in plants . Accumulation of mRNA and appearance of Pi transporter protein within 12 to 24 h after removal of Pi have been observed in tomato (Lycopersicon esculentum; ; ). In this study, the rapidity of gene induction and repression by Pi was examined. Seven-day-old seedlings (AtPT2-LUC) were transferred to Murashige and Skoog medium with and without Pi. Samples were collected at different time intervals after the initiation of treatment and assayed for the expression of LUC. A small but reproducible increase in LUC activity was observed as early as 12 h after transferring plants to Pi-deficient medium. A 60-fold increase in the LUC activity was recorded after 2 d of starvation (Fig. A). The activity reached the highest level of nearly 100-fold by 4 d of starvation and stabilized thereafter. Northern analysis of total RNA from these plants indicates the accumulation of LUC-specific mRNA as early as 12 h. The transcript levels increased with increasing duration of Pi starvation. This increase in the LUC transcripts appears to serve as a prelude to enzyme activity during Pi starvation (Fig. B). Figure 4 | Temporal expression of Pi transporters. Temporal expression of Pi transporters. A, Rapid induction of AtPT2 promoter-driven expression of LUC during Pi deficiency was monitored in the transgenic plants. Seven-day-old seedlings of AtPT2-LUC plants were transferred to Murashige and Skoog medium without Pi. The seedlings were removed at different times after Pi starvation to measure reporter gene activity and for northern analysis of transcripts. B, Total RNA from the plants harvested at different times of Pi deficiency was separated on denaturing formaldehyde agarose gels and blotted onto nylon membranes. The membranes were probed with 32P-labeled cDNA fragments of AtPT1, AtPT2, and LUC. C, Reversibility of induction of genes was studied by resupplying Pi to Pi-starved plants. The LUC expressing plants (AtPT2-LUC) were starved for Pi for 5 d and then transferred to Murashige and Skoog medium containing sufficient Pi (1,250 mum). Other samples of seedlings were grown continuously under Pi deficiency. Plants were harvested at different time periods (12 h, and 1, 2, 3, and 4 d) after transfer, and analyzed for reporter gene activity and isolation of RNA. DR, Days after replenishment with Pi; DP, days plants continued to be grown in the absence of Pi. D, Northern blot showing the expression of AtPT1, AtPT2, and LUC in plants subjected to Pi replenishment experiments. The nylon membranes containing the RNA were probed with 32P-labeled cDNAs. Expression of native Pi transporters, especially AtPT2, correlated with the expression of LUC and the enzyme activity. These results confirm a link between transcriptional activation of Pi transporters and Pi uptake during the nutrient starvation. Expression of reporter gene was not only inducible by Pi starvation but also repressible by replenishment of the nutrient (Fig. C). A 3-fold reduction in LUC activity was apparent within 24 h of Pi replenishment. The activity progressively decreased to one-sixth of the induced level within 4 d of Pi resupply, whereas the LUC activity remained at a high level in plants that continued to grow in Pi-deficient medium. Northern analysis of total RNA from these plants showed that repression of Pi transporter and the reporter gene is a rapid response to Pi supplementation. In the case of AtPT1, the reversal of induction was gradual, which is consistent with a relatively constitutive nature of its expression (Fig. D). There was a noticeable difference between the relatively rapid decrease in the AtPT2-LUC transcript abundance and the LUC enzyme activity in Pi-replenished plants. Pi Transporters Are Expressed in Specific Tissues under Pi Starvation | One of the effective ways to analyze tissue-specific expression of genes is by localizing the reporter gene activity in transgenic plants. There were distinct differences in the patterns of expression of reporter genes driven by the AtPT1 and AtPT2 promoters in roots of Pi-starved plants. AtPT1 promoter-driven reporter gene activity was lacking in root tips; in contrast, the AtPT2 promoter-driven reporter gene expression was observed in all cells of undifferentiated segments of the root, including the tip (Fig. A). Furthermore, in AtPT1 transgenic plants, GUS and GFP activity was quite high in the newly formed secondary root branches. In the case of AtPT2-GUS plants, the pattern of expression of the reporter gene remained the same irrespective of development of the secondary roots. The expression of GUS in the root hairs in both AtPT1-GUS and AtPT2-GUS plants was quite evident under Pi starvation (Fig. A). The root hairs and emerging secondary roots of AtPT1 transgenic plants showed highest levels of GFP expression. The histochemical analysis further confirmed the lack of expression of the reporter gene driven by AtPT1 promoter in root tips. The tissue sections also revealed that AtPT1 is strongly expressed in epidermal and endodermal layers. The GUS activity was detected in all the cell types in the meristamatic region of AtPT2-GUS roots. In the mature region, the activity of GUS was observed in the epidermis, cortex, and particularly in the stellar region (Fig. B). These studies show that different Pi transporters may have tissue specific role in acquiring and transferring Pi within the roots. A relatively strong expression of reporter genes in epidermal layers provides further evidence for the role of epidermally localized Pi transporters in Pi acquisition. Figure 5 | Pi transporter promoter-mediated expression of reporter genes in Arabidopsis root. Pi transporter promoter-mediated expression of reporter genes in Arabidopsis root. A, AtPT1 promoter-mediated expression of GUS (1 --3) and GFP (4 --6) in Arabidopsis roots and root hairs (3 and 6) are shown. Stronger GUS activity can be observed in newly formed branches, whereas the intensity of GUS staining decreased in the primary root (2). Lack of GUS expression in the tips of primary and lateral roots of AtPT1-GUS plants is clearly depicted (1 and 2). AtPT2 promoter-driven expression of GUS in all parts of the roots, including root tips and root hairs, is shown in 7 through 9. B, Expression of the reporter gene activity was examined in thin sections of roots of transgenic plants. The Pi-deficient plant roots expressing GUS were allowed to develop blue color in the presence of X-Gluc. The root segments were fixed and embedded in Technovit resin and 8-mum sections were cut and placed on a slide. The sections were photographed under a microscope (Olympus Corporation, Lake Success, NY). Top, GUS staining of the transverse sections of roots; bottom, cross-sectional view. 1 through 4, Expression of the AtPT1-GUS in the root tip (1 and 3) and the differentiated region of a root (2 and 4). Similarly, the AtPT2-GUS expression can be seen in the root tip (5 and 7) and the differentiated region (6 and 8). The High-Affinity Transporters May Play a Role in Internal Pi Mobilization | High-affinity Pi transporters have been shown to express preferentially in roots during Pi starvation. However, reporter gene expression studies revealed a very interesting pattern of gene expression in mature plants subjected to Pi starvation. In this analysis, 1-month-old transgenic plants with young flowers and fruits, expressing LUC or GUS under the control of the AtPT2 promoter, were transferred to a hydroponic nutrient solution with or without Pi for 2 weeks. Under Pi deficiency conditions, a strong expression of LUC was observed in flowers and at the junction between silique and peduncle (Fig. A). In addition, some of the senescing leaves of Pi-starved plants also showed the expression of AtPT2 promoter-driven reporter genes (data not shown). These results suggest a possible role for AtPT2 in the mobilization Pi during the nutrient starvation. Figure 6 | A, AtPT2 promoter-mediated expression of reporter genes in flowers and fruits of Pi-starved Arabidopsis. A, AtPT2 promoter-mediated expression of reporter genes in flowers and fruits of Pi-starved Arabidopsis. Expression of GUS (1 and 3) and LUC (2 and 4) in flowers (1 and 2) and fruits (3 and 4) was examined in Pi-starved plants. The reporter gene expression is confined to the silique and stalk junctions in fruits (3 and 4). Transverse (5) and cross-sectional (6 and 7) views of the silique and stalk junction clearly show the GUS expression. Strong expression of reporter gene is observed in outer cell layers (reminiscent of nectaries) and some cells in tracheary elements and vascular tissues. Cross-sectional view of the fruit stalk (8) showing the expression of GUS in some cells of cortical region. B, Heterologus expression of reporter genes in tobacco (Nicotiana tabacum). Expression of LUC (1) and GUS (2) under the regulation of AtPT2 promoter in tobacco roots. Longitudinal (3) and cross-sectional (4) view of the root showing the GUS expression. AtPT2 Promoter Drives the Expression of Reporter Genes in a Heterologous System | Transgenic tobacco plants expressing LUC or GUS under the regulation of the Arabidopsis Pi transporter promoter (AtPT2) were generated by Agrobacterium tumefaciens-mediated leaf disc transformation method. A number of independent transgenic lines were examined for the reporter gene activity. In all the lines examined, reporter gene expression was induced only under Pi starvation and no detectable activity was observed in the presence of Pi. Expression of reporter genes was quite pronounced in the region behind the root cap (Fig. B). Similarly, the tips of newly emerging secondary roots of Pi-starved plants exhibited strong reporter gene activity. In addition, GUS expression was also observed in root hairs of Pi-starved tobacco (data not shown). These data indicate that the conserved cis-elements present in the AtPT2 promoter are sufficient to drive the Pi starvation-mediated expression of reporter genes in a heterologous plant. Auxins and Cytokinins Suppress Expression of the AtPT1 Gene | Because many morphological responses of plants to phytohormones resemble those of Pi deficiency, we examined the effect of auxin, ethylene, or cytokinin on Pi starvation-induced gene expression. Transgenic plants expressing reporter genes under the regulation of AtPT1 (AtPT1-GUS/GFP) and AtPT2 (AtPT2-GUS) promoters were used in these studies. GUS activity under regulation of the AtPT1 promoter was suppressed by auxin and cytokinins (Fig. A). Suppression due to auxin or cytokinin was obvious both under Pi sufficiency and deficiency conditions. Increasing differences in the reporter gene activity were observed with increasing concentrations of Ki. Interestingly, there were no significant differences in the reporter gene activity in AtPT2-GUS plants either in the presence or absence of hormones (Fig. B). The suppression of reporter gene activity by hormones paralleled the suppression of AtPT1 expression (Fig. C). In contrast, expression of AtPT2 and other Pi starvation-induced genes such as RNase2, At4, purple acid phosphatase, and a constitutively expressed tubulin were comparable in hormone-treated and -untreated plants. Furthermore, there was no major difference in the transcript levels of Pi starvation-induced genes, including the transporters in ACC-treated plants. Some of the differences in gene expression observed in this study and an earlier study by may be due to differences in culture conditions and duration of hormone treatments. Figure 7 | Effect of hormones or hormone inhibitors on GUS activity and expression of Pi starvation-responsive genes. Effect of hormones or hormone inhibitors on GUS activity and expression of Pi starvation-responsive genes. Seven-day-old seedlings of AtPT1-GUS/GFP and AtPT2-GUS were transferred to Murashige and Skoog medium with (P+) and without (P-) Pi supplemented with 2,4-dichlorophenoxyacetic acid (2,4-D; 0.5 mum), naphthalene acetic acid (0.5 mum), kinetin (Ki; 0.1, 1.0 and 10 mum), 6-benzyleaminopurine (1.0 mum), 1-aminocyclopropane-1-carboxylic acid (ACC; 50 mum), 2,3,5-triiodobenzoic acid (1.0 mum), naphthalmic acid (1.0 mum), or alpha-2-aminoethoxyvinyl Gly (AVG; 10 mum). Seedlings were harvested after 48 h of treatment for enzyme assay and RNA isolation. Effect of hormones and inhibitors of hormones on AtPT1 (A and D) and AtPT2 (B and E) promoter-driven expression of GUS under Pi-sufficient (white bars) and -deficient (black bars) conditions. Error bars represent sd. Northern analysis of total RNA from AtPT2-GUS plants treated with different hormones and hormone inhibitors (C and F). Fifteen micrograms of total RNA was transferred to nylon membranes. The membranes were hybridized with 32P-labeled AtPT1, AtPT2, At4, PAP, RNase2, and tubulin cDNA fragments. The effect of auxin and ethylene on gene expression was further analyzed by treating plants with auxin transport and ethylene biosynthesis inhibitors. The reporter gene activity of plants treated with inhibitors was similar to that of control plants (Fig. , D and E). There was no obvious change in the transcript levels of Pi starvation-induced genes in the inhibitor-treated or -untreated plants (Fig. F). These studies point to the complex nature of hormone interaction during Pi starvation and highlight the need for further experimentation. DISCUSSION : Pi acquisition is a highly regulated process in plants, and the ability of plants to acquire the nutrient increases during Pi starvation . Interestingly, kinetic analysis of nutrient uptake pointed to an increase in the Vmax of the high-affinity Pi transporters without much change in the Km. Based on these results, it was concluded that during Pi starvation the number of high-affinity Pi transporters increases, whereas the low-affinity Pi transporter's activity remains unchanged . A clear link between increased transcription, translation, and assembly of Pi transporters into the plasma membrane during Pi starvation was obtained using gene-specific probes and antibodies for the transporter . In this study, we report an extensive analysis of transcriptional regulation, and tissue-/organ-specific expression of transporter genes under Pi deficiency. The data obtained with three different reporter genes under the regulation of two high-affinity Pi transporter gene promoters provide compelling evidence for transcriptional regulation of Pi uptake in plants. These promoters represent two highly expressed Pi transporters (AtPT1 and AtPT2) of Arabidopsis. The similarity between the expression of reporter genes and native genes suggests that the promoters used in these studies are likely to contain all the conserved elements. The sequences responsible for root preferential expression and Pi starvation responsiveness appear to be located in the promoters used to drive the expression of these reporter genes. At present, the specific DNA regions associated with these responses are not completely elucidated. A conserved MYB transcription factor (PHR1) binding sequence has been identified in many of the Pi starvation-induced genes . The PHR1 protein interacts with the conserved cis-element present in Pi starvation-induced AtIPS1 gene of Arabidopsis. Interestingly, mutation in PHR1 affects several Pi starvation-induced responses of Arabidopsis . DNA-protein interaction studies have shown that two regions in the AtPT2 promoter specifically interact with nuclear protein factors that are present under Pi sufficiency but either disappear or are modified during Pi deficiency . It is likely that interaction of trans-acting factors with cis-elements located on the promoters is responsible at least in part for root preferential and Pi deficiency-induced gene expression. The reporter gene activity increased as a direct consequence of changes in Pi concentration in the medium. This response is reminiscent of earlier Pi uptake studies carried out in plants that were subjected to Pi starvation . Expression of the AtPT1-reporter gene, which was obvious even under Pi sufficiency condition, increased by 2-fold during Pi deficiency. In contrast, the AtPT2 promoter-driven reporter gene activity was observed only under Pi deficiency conditions. These findings point to the existence of two types of high-affinity Pi transport systems operating under low Pi conditions, one that responds rapidly and specifically to Pi deficiency, and the expression of the other increases during the deficiency. The differential expression of two Pi transporters is a reflection of the ability of plants to adjust uptake rates based on internal Pi demand. Furthermore, induction of the reporter gene activity was not only rapid but also reversible upon resupply of the nutrient to Pi-deficient plants. The reversible induction of transporters responsible for regulated uptake of nutrients is crucial for maintaining cellular ion homeostasis. This is particularly relevant for Pi, where excess uptake results in imbalance of ions such as Zn and leads to physiological disorders . One distinct advantage of using reporter genes is the relative ease of analyzing tissue- and organ-specific expression of genes. Both promoters activated the expression of reporters primarily in roots. Expression of the AtPT2 promoter-driven reporter gene was observed all along the roots, confirming the notion that the entire root system retains the ability to acquire Pi during the nutrient deficiency . The newly emerging secondary roots had noticeably higher levels of AtPT1 promoter-driven reporter gene activity. Both AtPT1 and AtPT2 promoters strongly activated reporter gene expression in root hairs. The elongating root hairs are considered to be an important component of Pi acquisition. Because of their large surface area and ability to access narrow pores in soil, they are geometrically placed in a better position to obtain Pi. It has been shown that nearly 63% of Pi may be acquired by root hairs and their role becomes more important under Pi deficiency conditions . The data presented here provide molecular evidence that under Pi deficiency, root hairs not only increase surface area but also have enhanced Pi uptake ability. The integration of changes in root morphology, biochemistry, and gene expression under Pi deficiency has been studied more extensively in the proteoid roots of white lupin . Under Pi deficiency, these roots increase surface area and act as organs of organic acid biosynthesis and secretion. Furthermore, they also serve as sites for the production of phosphatases and exhibit enhanced Pi uptake. Taken together, these studies suggest that roots and root hairs produced in response to Pi starvation are morphologically and biochemically distinct from those of Pi-sufficient plants. Histochemical analysis of sections of roots further confirmed the distinct patterns of expression of AtPT1 and AtPT2. Although AtPT2-GUS activity was observed in all tissue types in the undifferentiated root tip and elongation zone, strong activity was evident in the vascular tissues of differentiated region. This indicated that high-affinity Pi transporters are not only involved in acquiring Pi from soil but they may also be involved in its transfer to the vascular tissues. In contrast, the AtPT1-GUS activity was clearly absent in the root tips and root caps. Another distinguishing feature is that the level of AtPT1-GUS expression decreased in main branches upon formation of secondary roots. Higher expression of reporter genes in newly emerging roots, root hairs, and epidermis resemble that of in situ transcript localization of Pi transporter in tomato . It is generally assumed that high-affinity Pi transporters are involved in Pi uptake by roots, whereas low-affinity transporters regulate movement of Pi within the plant. This assumption was supported by the preferential expression of a number of cloned high-affinity Pi transporters in roots . However, analysis of reporter gene activity driven by the AtPT2 promoter in Pi-starved mature plants revealed the expression of reporter genes in floral organs. Interestingly, the majority of reporter gene activity was found in the thalamus region of the flower and at the junction of the silique and fruit stalk. Based on these patterns of gene expression, it is tempting to suggest that AtPT2 may be involved in loading of Pi to reproductive organs during Pi deficiency. The reporter gene expression studies also revealed that the AtPT1 promoter specified a different expression pattern compared with the AtPT2 promoter. It will be interesting to analyze the tissue-specific expression and Pi starvation responsiveness of the other seven members of the high-affinity Pi transporter family in Arabidopsis. Expression of reporter genes in roots of Pi-deficient tobacco points to the conservation of the components of Pi starvation-induced signal transduction pathway and response mechanisms across plant species. Because plants are constantly interacting with Pi-limited ecosystems, they may have developed highly conserved Pi starvation sensing (monitoring) and signaling mechanisms. Despite extensive information on gene expression during Pi deficiency, little is known about signal transduction during Pi starvation. Similarities between Pi starvation-induced responses and that of auxin and ethylene has led to the hypothesis that these hormones play a role in Pi starvation . In addition, cytokinins also have been shown to be involved in suppression of Pi starvation-induced gene expression . In this study, auxin, particularly 2,4-D, and cytokinin suppressed the activity of a reporter gene driven by the AtPT1 promoter. In addition, hormones also suppressed accumulation of AtPT1 transcripts. This effect appears to be specific to AtPT1 expression because there was no marked difference in the induction of other Pi starvation-associated genes or the reporter gene activity driven by the AtPT2 promoter. Similarly, ethylene precursor ACC or its biosynthesis inhibitor AVG had no obvious effect on Pi starvation-induced gene expression. These results suggest that some components of the Pi starvation response may be directly affected by auxin and cytokinin and this response may not involve ethylene. There is growing evidence supporting the differential effect of hormones on Pi starvation-induced responses . This has been clearly demonstrated in the proteoid roots of white lupin, wherein auxin transport inhibitors decreased the levels of Pi starvation-induced phosphoenolpyruvate carboxylase and malate dehydrogenase, but not the secretion of acid phosphatase . Similarly, cytokinins have been shown to suppress the expression of some Pi starvation-induced genes, including AtPT1, but not the modifications of root growth . These observations support the notion that both hormone signaling-dependent and -independent pathways are involved in Pi starvation response. The tissue-specific expression of reporter genes in newly emerging secondary roots and root hairs under Pi starvation point to a fine coordination of hormone action and Pi starvation-induced gene expression. The data presented here provide strong evidence for the transcriptional regulation of high-affinity Pi transporters and their potential involvement in internal Pi mobilization. Different members of the high-affinity Pi transporter family are likely to be involved in Pi uptake and mobilization in the plant, along with the low-affinity transporters. These studies also show that morphological and physiological changes occurring during Pi starvation are highly coordinated with the molecular changes. The phytohormones may play a dual role in Pi starvation response by enhancing root biomass and specifically altering the expression of genes. MATERIALS AND METHODS : Generation of Expression Vectors | A 2.1-kb 5'-flanking sequence of the AtPT1 gene representing the promoter region was PCR amplified from Arabidopsis genomic DNA using specific 5' (CGGGATCCGATGGATGAGTCTACGTAC) and 3' (GGGAAGCTTCCTAGAGCTCTATAATC) primers. The amplified fragment was cloned into pGEM5 vector (Promega, Madison, WI) and the authenticity of the PCR product was confirmed by restriction enzyme digestion and DNA sequencing. The promoter fragment was released from pGEM5 by restriction enzyme digestion and transcriptionally fused to GUS/GFP reporter genes in the binary vector pCAMBIA1303 (CAMBIA, Canberra, Australia). Similarly, the AtPT2 promoter (2.3 kb) was also amplified from genomic DNA using the specific 5' (TCCCATTCGTGAATGAAAAC) and 3' (TCTTCTTCTCCTCTGC) primers and cloned into pBSIIKS+. The amplified product contains a 1.1-kb intron located in the untranslated 5' region of the gene. The promoter fragment was released from the pBSIIKS+ and cloned into binary vectors pLPTV-BAR and pGPTV-BAR , resulting in the transcriptional fusion of the promoter to the reporter genes firefly LUC and GUS, respectively. The binary vector pLPTV-BAR was constructed by replacing the HindIII-SacI fragment of pGPTV-BAR representing the coding sequence of GUS with an HindIII-SacI fragment harboring the firefly LUC coding sequence from RD29A-LUC . The binary vectors were transferred to the Agrobacterium tumefaciens (PGV3101) and used for transformation. Plant Material and Transformation | Transgenic Arabidopsis (ecotype Columbia) plants expressing reporter genes were generated by the A. tumefaciens floral dip method as described by . Seeds of transformed plants (T0) were germinated in the Scotts Metro-mix-360 with coir medium (Scotts-Sierra Company, Marysville, OH) or petri plates with sterile agar media to select transgenic plants resistant to herbicide or antibiotics. Homozygous transgenic plants expressing reporter gene were selected in the following generations. These plants were used in all the molecular and biochemical analysis of Pi starvation response. One-month-old transgenic seedlings were also examined for the expression of reporter genes in different parts of the plant. The seedlings grown in the Scotts Metro-mix-360 were removed and roots were carefully washed free of the medium and transferred to modified Hoagland nutrient solution containing 250 mum Pi or no Pi . After 2 weeks of the transfer, different parts of plants were analyzed for the reporter gene expression. Transgenic tobacco (Nicotiana tabacum) plants were generated by A. tumefaciens-mediated leaf disc transformation method. Seeds from transgenic tobacco plants were germinated on the Scotts Metro-mix-360 and seedlings were selected for herbicide resistance. Four-week-old seedlings were transferred to hydroponics solutions containing 250 mum or no Pi for 2 weeks before harvesting for gene expression analysis. Liquid Culture of Arabidopsis Seedlings | Seeds from homozygous Arabidopsis plants expressing reporter genes were treated for a minute in 70% (v/v) ethanol and rinsed twice with sterile water. The seeds were then treated in 50% (v/v) commercial bleach with 0.1% (v/v) Tween 20 for 10 min and rinsed thoroughly with sterile water to remove the residual bleach. Surface sterilized seeds were subject to stratification at 4C for 2 d before dispensing in conical flasks containing sterile Murashige and Skoog liquid medium. The seedlings were grown under a 16-h-light/8-h-dark cycle at 25C with constant shaking (85 rpm). Seedlings grown in liquid culture were used for all biochemical and molecular analysis unless otherwise indicated. All the experiments were repeated for a minimum of three times with at least three replications. Analysis of Temporal- and Pi Concentration-Mediated Expression of Genes | Transgenic Arabidopsis plants expressing GUS or LUC under the regulation of AtPT1 and AtPT2 were used in this study. Seven-day-old seedlings grown in one-half-strength Murashige and Skoog medium were rinsed once with sterile distilled water followed by a rinse with sterile Murashige and Skoog medium without Pi and transferred to sterile Murashige and Skoog medium supplemented with different concentrations of Pi (0, 5, 10, 25, 50, 125, 250, 500, and 1,250 mum). The plants were harvested after 5 d and used for gene expression analysis. To study temporal induction of Pi transporters and the reporter gene in response to Pi starvation, 7-d-old transgenic seedlings (AtPT2-LUC) grown in liquid medium were rinsed once with sterile water, followed by a rinse with sterile Murashige and Skoog medium deficient in Pi, and transferred to fresh medium without Pi. Plants were collected at different time periods (12, 24, 48 h, 3, 4, and 5 d) after the transfer and used for isolating total RNA and measuring the reporter gene expression. Evaluation of the Effect of Hormones on Gene Expression | Seven-day-old transgenic seedlings (AtPT1-GUS/GFP and AtPT2-GUS) grown in one-half-strength sterile Murashige and Skoog liquid medium were rinsed once with sterile distilled water followed by P+ or P- Murashige and Skoog medium and transferred to P+ or P- Murashige and Skoog medium supplemented with auxins 2,4-D or naphthalene acetic acid, the cytokinins Ki or 6-benzyleaminopurine, and the ethylene biosynthesis precursor ACC. Similarly, seedlings were also treated with auxin transport inhibitors 2,3,5-triiodobenzoic acid or naphthalmic acid and the ethylene biosynthesis inhibitor AVG. After 48 h of treatment, seedlings were collected for GUS analysis and RNA isolation. RNA Isolation and Northern Hybridization Analysis | Total RNA was isolated by hot phenol and lithium chloride precipitation method . Ten micrograms of total RNA was electrophoretically separated in a denaturing formaldehyde agarose gel and blotted onto nylon membranes. The nylon filters were hybridized overnight with 32P-labeled DNA probes in a hybridization solution containing 50% (v/v) formamide, 5x Denhardt's solution, 0.1% (w/v) SDS, 6x SSPE, and 100 mug mL-1 denatured salmon sperm DNA at 42C. The membranes were washed three times with 2x SSC/2% (v/v) SDS for 15 min each at 55C before autoradiography. LUC Imaging | Imaging of LUC activity in transgenic Arabidopsis and tobacco plants was done with a high-performance CCD camera (Princeton Instruments, Trenton, NJ). Seedlings or the parts of the plants to be imaged were sprayed uniformly with 100 mm luciferin (dissolved in 0.1% Triton X-100) and kept in the dark for 5 min. Then the seedlings were transferred to a dark chamber equipped with a CCD camera and the images were acquired for 5 min and analyzed with the software "Winview" (Princeton Instruments). Quantitative Analysis of LUC Activity | LUC activity was recorded with a luminometer TD-20/20 (Turner Designs, Sunnyvale, CA) following the manufacturer's instructions. The tissue was ground to a fine power in liquid nitrogen and about 200 mg of the powder was transferred to a microfuge tube containing 500 muL of LUC extraction buffer (100 mm K2HPO4, pH 7.8; 1 mm EDTA; 10 mm dithiothreitol; and 0.25% [v/v] glycerol). The contents were mixed thoroughly on a vortex mixer and centrifuged at 15,000 rpm for 5 min at 4C. Supernatant was transferred to a fresh tube and used as a source of the enzyme. The enzyme activity was determined within 2 h of extraction. The reaction mixture consisted of 100 muL of LUC assay buffer (50 mm HEPES; 20 mm MgCl2; 10 mm ATP in 0.2 m potassium Pi buffer, pH 7.0; and 0.05% [w/v] bovine serum albumin) and 50 muL of enzyme extract. The assay buffer and enzyme extracts were mixed in a polypropylene cuvette and placed inside the luminometer chamber. The reaction was initiated by injecting 50 muL of 1 mm luciferin (Promega). The instrument automatically measures and computes the light produced by the luciferin-LUC reaction. The luminescence values were expressed as relative luminescence units per milligram total protein. Total protein in the crude enzyme extract was determined according to . Quantitative Analysis of GUS Activity | Fluorometric quantification of GUS activity was done as described by . About 200 mg of plant tissue powder was transferred to a microfuge tube containing 500 muL of GUS extraction buffer (50 mm NaHPO4, pH 7.0; 10 mm beta-mercaptoethanol; 10 mm EDTA; 0.1% [v/v] Triton X-100; and 0.1% [w/v] sodium lauryl sarcosyl) and thoroughly mixed. The contents were centrifuged (15,000 rpm) for 5 min at 4C. The supernatant was transferred to a fresh tube and used as the source of enzyme. The reactions were started by mixing 450 muL of prewarmed (37C) GUS assay buffer (2 mm 4-methyumbelliferyl-beta-d GlcUA in GUS extraction buffer) with 50 muL of enzyme extract in a microfuge tube and incubating at 37C for 30 min. The reaction was terminated by transferring 100 muL of the reaction mix to 900 muL of 2% (w/v) Na2CO3. The amount of fluorescent product 4-methylumbelliferyl produced in the reaction was determined using a fluorometer. The GUS activity was expressed as pmoles of 4-methylumbelliferyl formed mg total protein-1 min-1. Histochemical Localization of GUS Expression | Histochemical staining for GUS activity was done according to the protocol described by , with slight modifications. Whole seedlings or parts of plants to be stained were incubated in GUS reaction mix (25 mg of 5-bromo-4-chloro-3-indolyl-beta-d-glucuronide in 50 mL of 100 mm sodium Pi buffer with 0.1% [v/v] Triton X-100) for 6 to 17h. The stained seedlings were transferred to 70% (v/v) alcohol to remove chlorophyll. Microscopic analysis and photographing of GUS expression in seedlings and organs such as flowers and fruits was done with a dissecting scope. Representative samples of GUS stained tissues were taken for histochemical analysis of the reporter gene expression. The tissue fixation and subsequent processing for the sectioning were essentially the same as that described by . Stained tissues were vacuum infiltrated with 4% (v/v) formaldehyde in 50 mm potassium Pi buffer for 20 min, followed by an overnight incubation at 4C in the same medium with gentle shaking. The fixed tissues were thoroughly washed with Pi buffer and dehydrated in alcohol series before embedding in Tecnovit resin (Energy Beam Science, Inc., Agawam, MA). Tissue sections (8 mu) were photographed under a microscope utilizing the differential interference contrast. Analysis of GFP Expression | Expression of GFP was visualized with the Sterio Fluorescent system (Lecia, McHenry, IL) using the standard fluorescein isothiocyanate filter. The optimal exposure time for capturing the images was 30 s. The light source was provided by an HBO-50W high-pressure mercury lamp. The maximum excitation and emission wavelengths for the GFP2 and GFP3 filter set were 510 nm (range of 480 --440 nm) and 525/550 nm (range of 470 --440 nm), respectively. Images were automatically transferred to a computer and analyzed using the SPOT version 3.0.4 software (Apple Event 3.0, Sterling Heights, MI). Backmatter: PMID- 12226503 TI - Arabidopsis UVR8 Regulates Ultraviolet-B Signal Transduction and Tolerance and Contains Sequence Similarity to Human Regulator of Chromatin Condensation 1 AB - To further our understanding of how plants defend against the harmful effects of ultraviolet (UV) light, we characterized an Arabidopsis mutant hypersensitive to UV-B. This mutant, UV resistance locus 8-1 (uvr8-1), contains a single recessive mutation at the bottom of chromosome 5. Fine-scale mapping localized uvr8-1 to a 21-kb locus containing five predicted open reading frames. Sequencing of this entire region revealed that the uvr8-1 allele contains a 15-nucleotide deletion in a gene similar to the human guanine nucleotide exchange factor regulator of chromatin condensation 1. This mutation reduces the UV-B-mediated induction of flavonoids and blocks chalcone synthase mRNA and protein induction. In contrast, uvr8-1 has enhanced induction of PR1 and PR5 proteins in response to UV-B, an indication of increased UV-B injury. These results suggest that UVR8 acts in a UV-B signal transduction pathway leading to induction of flavonoid biosynthesis. Keywords: Introduction : Plants must resist the deleterious effects of UV light because they are dependent on sunlight for photosynthesis and cannot avoid UV light exposure. Although UV is defined as the region of the spectrum from 200 to 400 nm, only the levels of UV-B (280 --320 nm) reaching the earth's surface are increased by the thinning of the stratospheric ozone layer . Thus, studies have focused on UV-B tolerance mechanisms because plants are directly affected by changes in terrestrial UV-B fluence. UV-B is known to cause DNA damage predominantly through cyclobutyl pyrimidine dimer formation and, to a lesser extent, pyrimidine-pyrimidinone (6, 4) photoproducts, both of which form by covalent bonding of adjacent pyrimidines (for review, see ; ; ). Failure to repair these DNA lesions interferes with DNA synthesis and transcription, and can result in heritable mutations (for review, see ). Studies of Arabidopsis have identified a variety of UV-B-hypersensitive mutants deficient in DNA repair (uvr1 , uvr2 [; ], uvr3 [; ], and uvh1 ). For example, photolyases, enzymes that use blue light energy to repair pyrimidine dimers , are critical for plant survival under UV-B in the laboratory . Other light-independent DNA repair mechanisms in plants are currently under study. In addition to directly causing DNA damage, UV-B generates oxidative stress through the formation of reactive oxygen species (ROS; ; ; ; ), which in turn causes enhanced lipid and protein oxidation . Plants counteract this increased ROS by increasing antioxidant enzymes . For example, exposure to UV-B induces guaiacol-peroxidases, ascorbate peroxidases, cytosolic Cu/Zn-superoxide dismutase (SOD), and coniferyl alcohol peroxidases . In addition, a role for ROS in UV-B-mediated plant damage is further evidenced by mutants deficient in ascorbic acid synthesis that are sensitive to UV-B irradiation . Although reactionary defense mechanisms abate the secondary effects of ROS generated by UV-B, plants utilize UV-absorptive secondary metabolites from the phenylpropanoid biosynthetic pathway as sunscreens to avoid UV-B. These compounds, especially the colorless flavonoids and hydroxycinnamic acids , accumulate in plants in response to UV. Several studies using Arabidopsis mutants deficient in flavonoids and hydroxycinnamic acids underscore the importance of chemical sunscreens in protecting against UV radiation (tt4 and tt5 , uvs , fah1 , and uvt1 ). Field studies in soybean (Glycine max) showed that UV-B was specifically required for sunscreen compound induction and this induction leads to a decrease in UV-B-mediated DNA . Here, we report the isolation and characterization of the UV-B-sensitive mutant, UV resistance locus 8-1 (uvr8-1), which defines a new class of UV-resistance gene. Unlike previously reported mutants that are defective in DNA damage repair or sunscreen biosynthetic enzymes, uvr8-1 has altered UV-B signal transduction as shown by a lack of UV-induced accumulation of flavonoids and chalcone synthase (CHS) mRNA and protein. Map-based cloning of uvr8-1 identified a gene with extensive sequence similarity to the human guanine nucleotide exchange factor regulator of chromatin condensation 1 (RCC1). In other eukaryotes, RCC1 functions as a nucleotide exchange factor for the Ran G-protein to regulate diverse biological processes, including RNA processing and nucleocytoplasmic transport . These results suggest that UVR8 plays a role in UV-B-mediated induction of flavonoid biosynthesis and plant defense against UV-B. RESULTS : UV-B Hypersensitivity of uvr8-1 | The uvr8-1 mutant was identified as having increased UV-B sensitivity compared with the progenitor Landsberg erecta (Ler) tt5 chalcone isomerase-deficient mutant line based on increased leaf injury and stunted growth under continuous cool-white fluorescent light chronically supplemented with 0.2 kJ UV-BBE m-2 h-1 UV-B for 10 d (see "Materials and Methods"). It is interesting to note that the parental chalcone isomerase-deficient tt5 line is already quite UV-B hypersensitive under growth chamber conditions due to decreased accumulation of flavonoids and sinapate esters . uvr8-1 was subsequently outcrossed four consecutive times to the wild-type Ler TT5 to generate homozygous Ler uvr8-1 TT5 lines. Homozygous uvr8-1 TT5 lines are indistinguishable from wild-type Ler UVR8 TT5 in the absence of UV-B, as shown in Figure A. However, uvr8-1 shows enhanced UV-B sensitivity in comparison with the wild-type Ler after subjecting 10-d-old plants to 3 d of constant 0.2 kJ UV-BBE m-2 h-1 (Fig. A). This sensitivity is displayed as necrosis of the first true leaves and cotyledons, as well as folding of the youngest leaves (Fig. A). In addition, leaf necrosis progressively worsens during 3 d of recovery under white light minus UV-B (Fig. B). Figure 1 | uvr8-1 is hypersensitive to UV-B in comparison with wild-type Ler. uvr8-1 is hypersensitive to UV-B in comparison with wild-type Ler. Wild-type Ler and uvr8-1 plants were grown in the absence of UV-B for 10 d. They were then treated with 72 h of 0.2 kJ UV-BBE m-2 h-1 (+) and were allowed a 72-h recovery period in white light without UV-B. Identically aged control plants were grown in the absence of UV-B throughout the experiment (-). Plants were photographed immediately after the UV-B treatment or following a 72-h recovery period. A, Plants photographed immediately after a 72-h UV-B treatment. B, Plants photographed after a 72-h recovery period following a 72-h UV-B treatment. uvr8-1 Alters Phenylpropanoid Metabolism | The accumulation of phenylpropanoid-derived metabolites, flavonoids, tannins, and anthocyanins is controlled by developmental (for example, in the seed coat) and environmental (e.g. under high-intensity white light) signals in Arabidopsis and other plants. UV-B also induces anthocyanin pigment accumulation in the hypocotyl of wild-type Arabidopsis seedlings, and this response is nearly abolished in uvr8-1 (J.E. Lim and D.J. Kliebenstein, unpublished data). However, the seeds from these plants have the wild-type brown seed coat coloration, suggesting that uvr8-1 is defective in environmental but not developmental regulation of anthocyanin accumulation. We next tested the hypothesis that uvr8-1 is altered in UV-B-mediated regulation of flavonoid or sinapate ester concentrations. As shown in Figure , HPLC analysis revealed that untreated uvr8-1 plants have normal sunscreen accumulation (Fig. , A and B). As expected, irradiating wild-type Ler TT5 with UV-B leads to increased sinapate esters (peaks 5 and 6) and flavonoids (peaks 7 --10; Fig. , A versus C). In contrast, uvr8-1 accumulates approximately 50% less total flavonoids than wild type following UV-B exposure (Fig. , B versus D). It is surprising that sinapate ester induction is normal in uvr8-1. Thus, uvr8-1 seems to have altered UV-B regulation of flavonoid and anthocyanin metabolism. Figure 2 | uvr8-1 reduces the UV-B-mediated induction of flavonoid accumulation. uvr8-1 reduces the UV-B-mediated induction of flavonoid accumulation. Plants were grown in the absence of UV-B for 14 d and were treated with 0.4 kJ UV-BBE m-2 h-1 for 3 d. Control plants were grown under white light in the absence of UV-B. Leaf tissue was harvested and methanol extracts were fractionated by reverse-phase HPLC. Numbers indicate the sinapate esters (peaks 5 and 6) and flavonoids (peaks 7 --10), as previously identified by . A, Chromatogram of extract from Ler untreated control. B, Chromatogram of extract from uvr8-1 untreated control. C, Chromatogram of extract from Ler treated with UV-B for 3 d. D, Chromatogram of extract from uvr8-1 treated with UV-B for 3 d. uvr8-1 Blocks Induction of CHS mRNA and Protein | CHS is the committing enzyme for flavonoid and anthocyanin biosynthesis (for review, see and ), and it is positively regulated by UV-B . To ask whether the uvr8-1-decreased flavonoid induction is controlled at the level of CHS protein accumulation, we compared CHS protein accumulation in uvr8-1 and wild-type Ler following UV-B treatment. As shown in Figure A, CHS protein continually increased over three consecutive days of UV-B treatment in wild-type Ler. In comparison, uvr8-1 completely blocked the UV-B-mediated induction of CHS protein. The inhibition of CHS induction in uvr8-1 is not caused by a global loss of stress responsive gene expression, as PR-1 and PR-5 proteins are induced more rapidly and to a higher level in uvr8-1 than in wild-type Ler (Fig. B). This suggests that UVR8 is required for transduction of a UV-B response signal. Figure 3 | Induction of CHS, PR1, and PR5 proteins by UV-B. Induction of CHS, PR1, and PR5 proteins by UV-B. Plants were grown in the absence of UV-B for 14 d and were then treated with 0.4 kJ UV-BBE m-2 h-1 for 1, 2, or 3 d. Control plants were grown to the same age in the absence of UV-B. A, Immunoblot with CHS antisera. Control plants showed no detectable CHS protein. B, Immunoblot with PR1- and PR5-specific antisera. To test whether the decreased induction of CHS protein in uvr8-1 is controlled at the mRNA level, we compared CHS mRNA accumulation in uvr8-1 and wild-type Ler following UV-B treatment. As shown in Figure , uvr8-1 and wild-type plants grown in white light without supplementary UV-B had comparable CHS mRNA. However, UV-B-mediated induction of CHS mRNA is nearly blocked in uvr8-1 as compared with wild type. The expression of VTC1 (vitamin C deficient; an ascorbic acid biosynthetic enzyme) and manganese SOD 1 (MSD1) were assayed to examine if uvr8-1 is impaired in antioxidant defense capacity . These genes are expressed at similar levels in uvr8-1 and wild-type Ler before and after UV-B treatment, suggesting that uvr8-1 is not deficient in antioxidant defense . MSD1 was previously shown to not respond to UV-B treatment and, therefore, functions as a control showing the use of equal cDNA amounts in the different reactions (D.J. Kliebenstein, unpublished data). These results further support the hypothesis that UVR8 transduces a UV-B-specific signal. Figure 4 | Analysis of CHS mRNA induction following UV-B treatment. Analysis of CHS mRNA induction following UV-B treatment. Reverse transcriptase (RT)-PCR analysis of CHS mRNA induction by UV-B. Fourteen-day-old plants were grown in the absence of UV-B and were treated with 0.4 kJ UV-BBE m-2 h-1 for 24 h or were left under -UV-B conditions for an additional 24 h. Tissue was then harvested for RNA extraction. Ethidium bromide-stained products are shown from quantitative RT-PCR. For each sample, 1 and 2 mug of total RNA were used for cDNA synthesis prior to PCR amplification. The results are representative of three independent experiments. Map-Based Cloning of UVR8 | To prepare for map-based cloning of UVR8, we initiated genetic analysis of uvr8-1. The mutant was genetically characterized by testing segregation of UV-B sensitivity in the F2 generation of a cross between homozygous uvr8-1 TT5 in the Ler genetic background and wild-type UVR8 TT5 Ler. Analysis of >540 segregating F2 progeny showed a 3:1 ratio of UV-B-resistant:UV-B-sensitive plants (chi2 = 0.241), suggesting that uvr8-1 is a monogenic recessive mutation. An F2 mapping population was created by crossing homozygous uvr8-1 TT5 in the Ler genetic background to Colombia-0 (Col-0). The F2 progeny were scored for UV-B sensitivity and were genotyped (see for information about available markers). As shown in Figure A, uvr8-1 is on chromosome 5, centromere distal to LFY3 . Figure 5 | Chromosome walk to uvr8-1 locus. Chromosome walk to uvr8-1 locus. A, Markers used for localizing uvr8-1 to chromosome 5. Numbers at top represent genetic distance to uvr8-1 in centromeres. B, The four P1 clones (MBK5, MGI19, MVD12, and MHJ24) covering the uvr8-1 locus, with their sizes indicated in parentheses. The physical positions of markers MBK5-1 (A), MBK5C3 (B), MGI19-1 (C), MGI19C7 (D), MGI19C9 (E), MGI19C8 (F), MGI19C6 (G), MHJ24C1 (H), and MHJ24-2 (I) are indicated. C, Map of approximate locations of recombination breakpoints used for fine mapping. D, Smallest region genetically identified to contain uvr8-1 mutation. The five ORFs are putative amino acid transporter (1), RCC1 homolog (2), Ser/Thr protein phosphatase (3), hypothetical protein (4), and histidinol dehydrogenase (5). E, DNA sequence of the uvr8-1 15-bp deletion. Fine-scale genetic mapping of the mutation required the identification of polymorphic markers tightly linked to UVR8. Using publicly available wild-type Col-0 genomic sequences of MBK5, MGI19, and MHJ24 P1 clones (Fig. B and ), we developed nine new markers. These simple sequence length polymorphisms (SSLPs) (MBK5-1, MGI19-1, and MHJ24-2), cleaved-amplified polymorphic sequences (CAPS; MBK5C3, MGI19C7, MGI19C9, and MHJ24C1), and single nucleotide polymorphic (SNP) markers (MGI19C8 and MGI19C6) are documented in Table (Fig. B). Genotyping 1,254 UV-B-sensitive F2 individuals with the flanking markers MBK5-1 (Fig. B, marker A) and MHJ24-2 (Fig. B, marker I) identified 14 recombinants (Fig. C). The 14 recombinants were tested in the F3 progeny to verify the UV-B sensitivity, and the genotype was tested for the other seven markers. The location of the recombination breakpoints indicates that UVR8 is between markers MGI19C7 and MGI19C6 (Fig. C). This 21-kb interval is predicted to contain five open reading frames (ORFs; Fig. D; bac_cosmid?MGI19). Table I | SSLP, CAPS, and SNP markers created in this study uvr8-1 Contains a 15-bp Deletion in an RCC1 Homolog | To identify the molecular lesion responsible for the UV-B sensitivity, the entire 21-kb region containing UVR8 was sequenced from uvr8-1 and wild-type Ler to detect uvr8-1-specific polymorphisms (the Ler genomic sequence is GenBank accession no. , and the experimentally verified cDNA is GenBank accession no. ). The only difference between the uvr8-1 and the wild-type Ler sequence in the entire 21-kb region was a 15-bp deletion in ORF2, renamed UVR8 from here in this manuscript (Fig. , D and E). The predicted UVR8 protein has sequence similarity to the RCC1 family of proteins (35% identity and 50% similarity), which are nuclear-localized guanine nucleotide exchange factors for the small G-protein Ran. In mammals and fungi, RCC1 functions with the Ran G-protein to regulate diverse biological processes, nucleocytoplasmic transport, and the cell cycle (for review, see ). Genomic sequencing has identified Ran homologs in Arabidopsis, suggesting that the Ran G-protein regulatory mechanism may also function in plants. The RCC1 secondary structure contains seven beta sheet blades whose structural integrity is dependent on four absolutely conserved Gly and one invariant cis-Pro . All of these conserved Gly and Pro residues are present in the wild-type UVR8 protein, except one Pro to Asn change. In addition, UVR8 contains the eight amino acids considered essential for RCC1 activity . In contrast, the five-amino acid deletion in uvr8-1 removes one of these absolutely conserved Gly and changes the spacing between two others (Fig. , conserved Gly are in bold print). Figure 6 | Deletion in uvr8-1 removes a Gly conserved among RCC1 homologs. Deletion in uvr8-1 removes a Gly conserved among RCC1 homologs. An alignment of the 15 amino acids surrounding the uvr8-1 deletion. The sequences are UVR8, uvr8-1, RCC1 from human (Homo sapiens; ), and homologs from hamster , Xenopus , fruit fly (Drosophila melanogaster; ), and yeast (Saccharomyces cerevisiae; ). RLD2 is a human protein that contains a domain similar to the entire RCC1 protein . Asterisks mark the uvr8-1 deletion. The numbers designate the distance from the carboxyl terminus. Gly described in the text are in bold. Complementation of uvr8-1 | To confirm that the 15-bp deletion in uvr8-1 is the basis of the UV-B sensitivity, we attempted to rescue the uvr8-1 UV-B phenotype through transformation with plasmid pGPTV-UVR8, which contains a UVR8 genomic fragment in the binary transformation vector pGPTV, as shown in Figure . Wild-type Ler and mutant uvr8-1 were transformed with pGPTV-UVR8 or empty pGPTV, and the transgenic progeny were tested for UV-B tolerance. All 16 uvr8-1 lines containing pGPTV-UVR8 were UV-B tolerant, indicating that UVR8 complements the uvr8-1 mutation and confirming that the 15-bp deletion in uvr8-1 leads to increased UV-B sensitivity. In addition, all six uvr8-1-transgenic lines containing pGPTV were UV-B sensitive, and all Ler transgenics, with pGPTV (five lines) or pGPTV-UVR8 (12 lines), were UV-B resistant, suggesting that transformation did not affect UV-B tolerance . UV-B-irradiated wild-type Ler transformants containing extra copies of UVR8 in the form of pGPTV-UVR8 also displayed an enhanced purple coloration in the leaves and anthocyanin pigment accumulation in the hypocotyl in comparison with transgenics containing the empty pGPTV vector alone. Thus, UVR8 may be a positive regulator for anthocyanin pigment accumulation. In summary, the deletion in uvr8-1 is the basis of the UV-B sensitivity, possibly through the removal of a protein required for the induction of UV-B defense mechanisms. Figure 7 | Generation of the UVR8 complementation construct. Generation of the UVR8 complementation construct. Boxes represent ORFs with gene labels contained within each box. Numbers at the top represent location of restriction enzyme digestion sites in kilobases. A, HindIII fragment containing the putative amino acid transporter and UVR8. B, Final HindIII cassette containing UVR8 obtained after removing the NcII fragment containing the amino acid transporter. C, Restriction map of pGPTV-BAR binary vector. L, Left T-DNA border; R, right T-DNA border. Table II | Primers used in this study DISCUSSION : UVR8 Positively Regulates UV-B Induction of Phenylpropanoid Metabolism | Our results suggest that UVR8 is a positive regulator involved in a UV-B signal transduction pathway. First, the uvr8-1 mutation blocks the UV-B-mediated induction of CHS mRNA and protein, as well as reduces flavonoid and anthocyanin pigment accumulation (Figs. --). Second, the presence of transgenic copies of UVR8 in uvr8-1 rescues anthocyanin production in response to UV-B and enhances the up-regulation of anthocyanin accumulation in wild-type Ler (J.E. Lim and D.J. Kliebenstein, unpublished data). Considering these results, UVR8 appears to be a positive regulator at least in the UV-B signal transduction pathway for CHS. The apparent complete block in CHS mRNA and protein induction only leads to a 50% reduction in flavonoid accumulation. This could result from a UV-B-mediated increase in metabolite flow through the phenylpropanoid pathway that acts to push the production of flavonoids in the absence of increased CHS levels. In contrast to the positive action of UVR8, AtMYB4, a previously identified transcription factor involved in regulating phenylpropanoid metabolism in response to UV-B, is a negative regulator. AtMYB4 represses cinnamate 4-hydroxylase expression, whereas it has minimal impact upon CHS . Upon exposure to UV-B, AtMYB4 transcript levels decrease, allowing increased cinnamate 4-hydroxylase expression and increased sinapate ester accumulation . Furthermore, an AtMYB4 knockout mutation leads to elevated sinapate ester and cinnamate 4-hydroxylase accumulation, but does not alter CHS or flavonoid levels . This suggests that the different phenylpropanoid biosynthetic pathway components are controlled by different regulatory mechanisms involving positive and negative control elements. Specificity of UV-B-Mediated CHS Regulation | A vast amount of work has shown that CHS mRNA accumulation is up-regulated following exposure to UV-B (for review, see ). This UV-B-mediated induction can be attenuated by red light through phyB and can be amplified by blue light independent of cry1 . Thus, UV-B is detected and the signal is transduced through an independent pathway not dependent on the known photoreceptors . Other work has shown that UV-B-dependent signals can be transduced through signal transduction pathways using nitric oxide, salicylic acid, jasmonic acid, ethylene, and/or ROS . Of all of these signals, only nitric oxide plays a role in UV-B-mediated induction of CHS mRNA or protein . Thus, UVR8 is functioning in a very specific UV-B signal transduction pathway that may use nitric oxide. Further evidence for this specificity comes from the observation that uvr8 does not alter tannin accumulation in seeds or anthocyanin accumulation in response to methyl jasmonate treatment (J.E. Lim and D.J. Kliebenstein, unpublished data). Identification of additional UVR8 signal transduction components should provide unique insight into how UV-B regulates CHS. In addition, analysis of CHS regulation in uvr8 under a diverse array of conditions will enable analysis of the UV-B specificity of this pathway. uvr8-1 Is UV-B Sensitive Due to a Deletion in an RCC1 Homolog | Complementation and mapping experiments show that a 15-bp deletion in a gene similar to RCC1 causes uvr8-1's UV-B sensitivity. The similarity between UVR8 and RCC1 suggests that UVR8 may have guanine nucleotide exchange activity . It is interesting that RCC1 mutations in Saccharomyces cerevisiae alter a wide variety of processes, including pre-mRNA processing and transport , mating behavior , initiation of mitosis , and chromatin decondensation . Although RCC1 mutations in fungi and other species are lethal or highly pleiotropic, the uvr8-1 mutation has no discernible effect on growth/development of growth chamber grown plants except in the presence of UV-B. This lack of pleiotropy could be an effect of this specific uvr8-1 mutation or an indication that UVR8 is not an RCC1 ortholog. Evidence for the latter comes from the observation that UVR8 is significantly smaller than RCC1 and does not contain the nuclear localization sequences conserved among the animal and fungal RCC1 proteins. The functional significance of the similarity between UVR8 and RCC1 remains to be determined. uvr8-1 Appears to Be Deficient in Multiple UV-B Defense Mechanisms | uvr8-1 appears to impair UV-B tolerance mechanisms in addition to decreased CHS and flavonoid induction. This conclusion comes from the observation that uvr8-1 was identified as UV-B sensitive in a tt5 background that blocks flavonoid accumulation . Additional support for this hypothesis stems from the observation that the phenotypes associated with uvr8-1 UV-B sensitivity (necrosis and leaf cupping) continue to worsen up to 72 h after UV-B removal. In comparison, UV-B sensitivity in the tt5 mutant does not show a continual progression of UV-B damage after removal of the UV-B . Thus, uvr8-1 may be defective in the induction of mechanisms that detoxify UV-B-mediated damage. However, these mechanisms do not block the accumulation of PR1 and PR5 stress-responsive proteins or alter the regulation of VTC1 and MnSOD antioxidant mRNAs (Figs. and ). Additional work is necessary to fully understand how the impaired UV-B signal transduction in uvr8-1 relates to the displayed UV-B sensitivity. Future Work and Implications | Further studies into the biological role of UVR8 should enhance our understanding of UV-B signal transduction and sensitivity. Identifying the specific mechanism by which uvr8-1 generates increased UV-B sensitivity and isolation of additional uvr8-1 mutations and new mutations that alter different components of the UVR8 UV-B tolerance mechanisms will boost our understanding of how plants resist UV-B. Because G proteins have not been shown to regulate CHS or phenylpropanoid metabolism, biochemical testing of the hypothesis that UVR8 is a guanine nucleotide exchange factor and identification of its substrate will help to elucidate new components in the regulation of secondary metabolism. In addition, investigating where UVR8 is in the UV-B signal transduction pathway and how uvr8-1 affects the synergy between UV-B, UV-A, and white light in regulating CHS will clarify how these signal transduction pathways are coordinated. MATERIALS AND METHODS : Plant Material and Growth Conditions | Wild-type Arabidopsis accessions Ler and Col-0 were used in this study, as were the mutant lines Ler tt5 . All plants were grown under constant light (60 --100 muE m-2 s-1 photosynthetically active radiation) using CW1500 cool-white fluorescent lamps (General Electric, Fairfield, CT) with filters to remove all UV-B, before and after UV-B treatment. Plants used for UV-B induction studies were grown on nutrient agar plates. Plants for UV-B sensitivity assays were grown in Cornell soil-less mix . All experiments were independently replicated at least twice. UV-B Treatment | UV-B treatments were conducted, and UV-B fluences are expressed as previously described . White light (100 muEi m-2 s-1) was supplemented with UV-B from F40 UV-B fluorescent lamps (Phillips, Somerset, NJ) or T12F40 UV-B lamps (UV Resources International, Cleveland). Light was filtered through 3-mm-thick Pyrex glass plates to remove UV-C, wavelengths <280 nm, and to attenuate UV-B levels; control plants were additionally shielded with 0.13-mm-thick Mylar (AIN Plastics, Mt. Vernon, NY) to remove UV-B, wavelengths <310 nm. For plants grown on soil, 0.13-mm-thick cellulose acetate was used instead of Pyrex glass to remove UV-C. UV-A levels were identical between control (-UV-B) and UV-B treatments. For mutant selection, 4-d-old seedlings were irradiated with 0.2 kJ UV-BBE m-2 h-1 (0.7 kJ m-2 h-1 unweighted UV-B) for 10 d. For analysis of uvr8-1, plants were grown for 10 d before being irradiated with 0.2 kJ UV-BBE m-2 h-1 (or 1.4 kJ m-2 h-1 unweighted UV-B) for 3 d. Quantitative RT-PCR Analysis | RNA was isolated from approximately 100 mug of leaf tissue using Trizol reagent (Invitrogen, Grand Island, NY). cDNA synthesis was per the manufacturer's instructions (SuperScript II; Invitrogen). PCR reactions were as described for SSLP reactions using primers 221EF1 and R1013 for the UVR8 cDNA , primers p3-GMP and 3'GMP for VTC1 , primers MnSOD1F and 1R for MSD1 , and primers CHS-1F and CHS-1R for CHS . Titration studies showed the optimal range for each primer pair is 1 or 2 mug of mRNA for cDNA synthesis followed by 3 mug of the diluted cDNA mix in a 25-cycle PCR reaction. Immunoblot Analysis | Protein sample preparation, quantitation, electrophoresis, and immunoblotting were as described by . Twenty micrograms of total leaf protein was loaded per lane, except for 60 mug of total leaf protein per lane for CHS. Proteins were detected using published antiserum concentrations and chemiluminescent detection . uvr8-1 Isolation and Genetic Characterization | Ler tt5 seeds were gamma irradiated with 50 krad, grown on soil, and M1 plants were allowed to generate M2 populations by self-pollination. One M2 plant was identified as more UV-B sensitive than the tt5 parent under continuous UV-B and was allowed to self-fertilize for two generations to obtain M4 seeds. A UV-B-sensitive M4 progeny was outcrossed to Ler TT5 to generate F2 for segregation analysis. UV-B-sensitive F2 progeny with a wild-type testa phenotype were allowed to self-pollinate. Homozygous uvr8-1/uvr8-1; TT5/TT5 lines were selected from the F3 and were outcrossed to Ler three times. Progeny from the fourth Ler TT5outcross (OC4) were used for all biochemical analyses unless otherwise noted. To generate a mapping population, uvr8-1/uvr8-1; TT5/TT5 plants were crossed to wild-type Col-0. F2 seeds were grown on plates without UV-B for 14 d and were then treated with UV-B for 3 d to identify homozygous uvr8-1 plants. All UV-B-sensitive plants were left to recover on plates for 2 weeks before being transferred to soil and allowed to self-cross to generate F3 seed. DNA Extraction, PCR, Restriction Enzyme Digestion, and Gel Electrophoresis Conditions | DNA was extracted as described . All PCR was done in a DNA Thermal Cycler 480 or a GeneAmp PCR System 9600 (PerkinElmer Instruments, Norwalk, CT). SSLP markers were amplified using a standard PCR mixture and program , except that 1.2 mm MgCl2 was used with MHJ24-2. SSLP products were separated on 4% (w/v) agarose and were visualized with ethidium bromide using an EagleEye II (Stratagene, La Jolla, CA). Standard PCR conditions were used for CAPS and SNP markers . CAPS markers were digested with the respective enzyme and were separated on 1.5 or 2.5% (w/v) agarose gels. Mapping Analysis, Sequencing, and Sequence Analysis | SSLP markers were identified by searching published Col-0 genomic sequences for di- or trinucleotide repeats longer than six units. PCR primers were designed to amplify products smaller than 200 bp and were synthesized by the Great American Gene Company or by the Cornell BioResource Center (Cornell University, Ithaca, NY; Table ). These primers were used to screen for polymorphisms between Col-0 and Ler as previously described, and the polymorphic markers are listed in Table . SNP and CAPS makers were identified by designing primers to amplify approximately 2 kb of presumed noncoding regions . These primers were used to amplify Ler genomic DNA, and the resulting PCR products were purified using the QIAquick PCR purification kit (Qiagen, Santa Clarita, CA) and sequenced by the Cornell BioResource Center. The sequence was compiled and compared with the published Col-0 genomic sequence using Sequencer 3.1 (Gene Codes, Ann Arbor, MI) and DNAstar (Madison, WI). Complementation Analysis | The construct scheme is shown in Figure . Escherichia coli containing the P1 clone MGI19 (Arabidopsis Resource Center, Ohio State University, Columbus) was grown overnight, and the P1 clone was isolated using a plasmid maxi kit (Qiagen). MGI19 was digested with HindIII and the fragment, including UVR8, and an amino acid transporter gene (AAT) was subcloned into pBluescript II SK+ phagemid (Fig. , A and B). This plasmid was digested with NcII, and the digested plasmid minus the 4-kb AAT fragment was relegated to produce a HindIII cassette containing only the complete UVR8 gene. This final cassette was then inserted into HindIII-digested pGPTV-BAR (Fig. C; ) to generate pGPTV-UVR8. pGPTV-UVR8 was transformed into uvr8-1 and Ler using Agrobacterium tumefaciens pMP90 strain GB3101 . uvr8-1 and Ler were also transformed with the empty pGPTV-BAR as controls. T1 progeny were screened for BASTA resistance and the BAR gene using PCR . In addition, primers 221EF1 and R1013 were used to screen uvr8-1 transformed with pGPTV-UVR8 to for wild-type UVR8. BASTA-resistant T1 and T2 progeny were tested for UV-B sensitivity as described above. Distribution of Materials | Upon request, all novel materials described in this publication will be made available in a timely manner for noncommercial research purposes, subject to the requisite permission from any third-party owners of all or parts of the material. Obtaining any permissions will be the responsibility of the requestor. Backmatter: PMID- 12226504 TI - Glycerophosphocholine Metabolism in Higher Plant Cells. Evidence of a New Glyceryl-Phosphodiester Phosphodiesterase AB - Glycerophosphocholine (GroPCho) is a diester that accumulates in different physiological processes leading to phospholipid remodeling. However, very little is known about its metabolism in higher plant cells. 31P-Nuclear magnetic resonance spectroscopy and biochemical analyses performed on carrot (Daucus carota) cells fed with GroPCho revealed the existence of an extracellular GroPCho phosphodiesterase. This enzymatic activity splits GroPCho into sn-glycerol-3-phosphate and free choline. In vivo, sn-glycerol-3-phosphate is further hydrolyzed into glycerol and inorganic phosphate by acid phosphatase. We visualized the incorporation and the compartmentation of choline and observed that the major choline pool was phosphorylated and accumulated in the cytosol, whereas a minor fraction was incorporated in the vacuole as free choline. Isolation of plasma membranes, culture medium, and cell wall proteins enabled us to localize this phosphodiesterase activity on the cell wall. We also report the existence of an intracellular glycerophosphodiesterase. This second activity is localized in the vacuole and hydrolyzes GroPCho in a similar fashion to the cell wall phosphodiesterase. Both extra- and intracellular phosphodiesterases are widespread among different plant species and are often enhanced during phosphate deprivation. Finally, competition experiments on the extracellular phosphodiesterase suggested a specificity for glycerophosphodiesters (apparent Km of 50 mum), which distinguishes it from other phosphodiesterases previously described in the literature. Keywords: Introduction : Phospholipids play a key role in the architecture of eukaryote membranes. Membrane lipid composition is under tight regulation that involves both lipid biosynthesis and turnover. Phospholipid turnover may result from the action of acyl-transferases , phospholipases , or lipolytic acyl-hydrolases . Among the catabolic products, acyl groups can be degraded by alpha- or beta-oxidation and used as a respiratory substrates or for triacylglycerol synthesis . The phospholipid head groups are a source of glycerol, phosphate, or polar head moieties that can be reused directly in phospholipid synthesis . In non-plant eukaryotes, phospholipid catabolism often produces glycerophosphodiesters. In yeast (Saccharomyces cerevisiae), glycerophosphodiesters are secreted in the extracellular medium and hydrolyzed at the outer surface of the cell . In animal cells, glycerophosphodiesters, mainly glycerophosphocholine (GroPCho) are synthesized from phospholipids. In HeLa cells, GroPCho secretion was observed . Moreover, GroPCho can accumulate in renal cells where its role as an osmoprotectant has been suggested . GroPCho concentration in renal cells is controlled by its enzymatic degradation rate, involving phosphodiesterase . In plants glycerophosphodiester accumulation was observed in physiological situations involving membrane turnover or degradation. Thus, during seed germination, glycerophosphodiesters, mainly GroPCho, accumulate in young rice (Oryza sativa) shoots and in maize (Zea mays) hypocotyls . Similarly, during the course of Suc starvation in sycamore (Acer pseudoplatanus) cells, reported a marked transient increase of the GroPCho level. This coincided with phosphocholine (P-Cho) accumulation and phospholipid catabolism during autophagy . Very little is known about glycerophosphodiester metabolism in higher plant cells. Because phosphatidylcholine is the major phospholipid in extraplastidial membranes, we decided to examine in detail GroPCho catabolism and started our work after GroPCho incorporation into various plant cells. These experiments led us to characterize a new GroPCho phosphodiesterase (GPC-PDE) at the outer surface of plant cells. An intracellular form of GPC-PDE was also discovered. Its accumulation under different stresses and the possible significance will be discussed. RESULTS : Identification of an Extracellular GPC-PDE Activity | To follow GroPCho catabolism in plant cells, 31P-NMR spectroscopy was performed on carrot (Daucus carota) cells. Figure A illustrates the changes that occurred when the cells were incubated in a nutrient medium containing 1.5 mm GroPCho as the only source of phosphate. A decrease of the major peak (0.1 ppm) was observed within the first hours, corresponding to the consumption of GroPCho from the extracellular medium. This decrease of GroPCho was concomitant with the appearance of another peak at 3.2 ppm, corresponding to P-Cho. After 5 h of incubation with GroPCho, we collected the medium and the cells and analyzed them separately under 31P-NMR spectroscopy (Fig. , B and C). In the medium (Fig. B), we only observed sn-glycerol-3-phosphate (Gro-3-P; peak at 4.4 ppm), inorganic phosphate (Pi; peak at 2.3 ppm), and GroPCho. In contrast, all of the P-Cho was found in the isolated cells (Fig. B). If we take into account the fact that, in our experimental conditions, the perfusion tube analyzed in 31P-NMR contained all of the packed cells (10 g fresh weight) but only a small fraction of the total extracellular medium (5%; see "Materials and Methods"), the amount of P-Cho accumulated corresponded only to a small fraction of the GroPCho hydrolyzed. Figure 1 | Representative 31P-NMR spectra of carrot cells incubated with 1.5 mm GroPCho at pH 6; 10 g (fresh weight) of carrot cells was perfused with 200 mL of a nutrient medium containing GroPCho as the sole source of phosphate. Representative 31P-NMR spectra of carrot cells incubated with 1.5 mm GroPCho at pH 6; 10 g (fresh weight) of carrot cells was perfused with 200 mL of a nutrient medium containing GroPCho as the sole source of phosphate. After 5 h of incubation, medium and cells were collected and analyzed separately (B and C). Glc-6-P, Glc-6-phosphate; cyt-Pi, cytoplasmic Pi; vac-Pi, vacuolar Pi; NTP, nucleotide triphosphate (mainly ATP and UTP). A, In vivo 31P-NMR spectra of carrot cells at the beginning of the incubation (A1, t = 0) and after 3 h of incubation (A2). B, 31P-NMR spectra (expanded scale from -1 to 6 ppm) of the medium collected after the incubation. C, 31P-NMR spectra (expanded scale from -1 to 6 ppm) of the perchloric extract from carrot cells collected after the incubation. To quantify GroPCho hydrolysis more accurately, we incubated carrot cells with 1.5 mm GroPCho in similar conditions and analyzed the incubation medium using 31P-NMR and enzymatic analysis techniques . Accumulation of Pi and Gro-3-P reflected the consumption rate of GroPCho (Fig. A). Accumulation rates of free choline and glycerol (Fig. B) were conversely consistent with GroPCho hydrolysis, confirming that the major proportion of the hydrolysis products accumulated outside of the cells. Under these experimental conditions, the GroPCho hydrolysis rate was estimated at 3.5 mumol h-1 g-1 fresh weight carrot cells. By contrast, the P-Cho accumulation rate deduced from the in vivo 31P-NMR spectra (Fig. A) was approximately 0.3 mumol h-1 g-1, which coincides with the choline incorporation rate (0.4 mumol h-1 g-1) measured on sycamore cells . Figure 2 | Evolution of extracellular choline, sn-glycerol, Gro-3-P, Pi, and GroPCho contents after incubation of carrot cells in a nutrient medium containing 1.5 mm GroPCho as the exclusive source of phosphate. Evolution of extracellular choline, sn-glycerol, Gro-3-P, Pi, and GroPCho contents after incubation of carrot cells in a nutrient medium containing 1.5 mm GroPCho as the exclusive source of phosphate. A, 31P-NMR spectra (expanded scale from -1 to 6 ppm) of the extracellular medium before (A1, t = 0) and after incubation with GroPCho (A2 and A3). B, Extracellular concentrations as a function of time determined using enzymatic analysis techniques (see "Materials and Methods"). GroPCho concentration was deduced from the 31P-NMR spectra. Concentrations are expressed as micromoles per milliliter of the incubation medium. To clarify the GroPCho cleavage site, we performed the same experiment with an alkaline extracellular medium (pH 8.5), which inactivated most of the cell wall acid phosphatase and prevented Gro-3-P hydrolysis. GroPCho was hydrolyzed at a comparable rate (not shown). But, when we compared the extracellular medium composition from the incubation driven at pH 6.0 (Fig. A) and that at pH 8.5 (Fig. B) using 31P-NMR spectroscopy, we observed a major peak at 4.4 ppm, corresponding to Gro-3-P. No P-Cho was observed, conversely indicating that Gro-3-P was the only phosphomonoester released after GroPCho hydrolysis. From these experiments, we conclude that there is a GPC-PDE activity that hydrolyzes GroPCho at the extracellular surface of carrot cells, releasing free choline and Gro-3-P. Under physiological conditions (pH 6.0), Gro-3-P is subsequently converted into glycerol and phosphate by acid phosphatases. Figure 3 | Representative 31P-NMR spectra (expanded scale from -2 to 6 ppm) of the extracellular medium after incubation of carrot cells for 5 h in 1.5 mm GroPCho at pH 6.0 (A) and pH 8.5 (B). Representative 31P-NMR spectra (expanded scale from -2 to 6 ppm) of the extracellular medium after incubation of carrot cells for 5 h in 1.5 mm GroPCho at pH 6.0 (A) and pH 8.5 (B). GroPCho was the sole exogenous source of phosphate, pH was maintained constant (at 6.0 or 8.5) using a pH-stat. After incubation, the medium was removed and analyzed at pH 7.5 to enable comparison between spectra. Finally, it remained unclear whether the intracellular P-Cho pool resulted from the incorporation of exogenous GroPCho and its hydrolysis in the intracellular compartments or from the choline uptake and subsequent phosphorylation in the cytoplasm by choline kinase. To discriminate between these two hypotheses, we incubated carrot cells with 1 mm GroPCho in the presence of 5 mm 3-hemicholinium, an efficient competitor of choline incorporation . After 6 h of incubation, 3-hemicholinium did not affect GroPCho hydrolysis. In the absence of 3-hemicholinium (Fig. B), P-Cho accumulated in the cells, as described previously, but, in the presence of the competitor, no increase of P-Cho was observed, and its concentration remained constant (Fig. , A and C). Comparable results were obtained when GroPCho was replaced by an equimolar (1 mm) mixture of choline and Gro-3-P. These results indicated that the newly produced choline was incorporated into the cells where it became phosphorylated, as recapitulated in Figure . Figure 4 | Representative 31P-NMR spectra (expanded scales from -1 to 5 ppm) of perchloric extracts from carrot cells incubated with 1 mm GroPCho. Representative 31P-NMR spectra (expanded scales from -1 to 5 ppm) of perchloric extracts from carrot cells incubated with 1 mm GroPCho. A, Control cells before incubation (t = 0). B, Carrot cells were incubated for 6 h with 1 mm GroPCho. C, Carrot cells were incubated for 6 h with 1 mm GroPCho and 5 mm 3-hemicholinium, an efficient competitor of choline incorporation. Incorporation and Intracellular Distribution of GroPCho Hydrolysis Products | Because choline was incorporated inside the cells after GroPCho hydrolysis, we followed its intracellular distribution using 31P- and 13C-NMR spectroscopy. Figure illustrates the intracellular distribution of choline, P-Cho, and glycerophosphodiesters after incubation of carrot cells with 2 mm GroPCho for 24 h. 31P-NMR spectra from carrot cells indicated that P-Cho (at 2.2 ppm) became the major phosphorylated compound after the incubation with GroPCho (Fig. A1). Comparison between the intracellular free choline and P-Cho concentrations using 13C-NMR spectroscopy (Fig. A2) revealed that the P-Cho pool (with the three methyl peaks shifted to 54.65, 54.72, and 54.78 ppm) represented two-thirds of the total soluble choline-containing compounds, whereas free choline (with the three methyl peaks shifted to 54.48, 54.55, and 54.6 ppm) represented the remaining one-third. Protoplast preparation with cellulase and pectolyase significantly affected the phosphomonoester contents: The Pi concentration increased, whereas the P-Cho concentration decreased, because it represented only 50% of the soluble choline-containing compounds (Fig. B). The significance of these modifications remains unexplained, however, we should consider that digestion with pectolyase and cellulase and incubation in mannitol may affect cell metabolism in several ways. When we isolated vacuoles from protoplasts (see "Materials and Methods) and analyzed them under 31P- and 13C-NMR spectroscopy, choline was found only as free choline (Fig. C), the low level of P-Cho observed reflecting the amount of cytoplasmic contaminants. When we analyzed a fraction enriched in cytosol (Fig. D), the proportions of P-Cho and free choline revealed by 13C-NMR spectroscopy (Fig. D2) conversely reflected the proportions observed in the protoplasts. Because no choline or P-Cho was found in the organelles precipitated after centrifugation (not shown), we could conclude that cytosolic P-Cho represents the major storage form of soluble-choline-containing compounds, whereas the vacuole represents only a pool of free choline. These results are consistent with in vivo 31P-NMR studies, which localized P-Cho in an alkaline compartment (cytoplasm) on the basis of its chemical shift but are also in agreement with metabolic modeling of choline metabolism, which predicted cytoplasmic P-Cho and vacuolar choline as the major pools of soluble choline-containing compounds . Figure 5 | Representative 31P- and 13C-NMR spectra of perchloric extracts of different carrot subcellular fractions. Representative 31P- and 13C-NMR spectra of perchloric extracts of different carrot subcellular fractions. Carrot cells were incubated with 2 mm GroPCho for 24 h. A, Total cells extract; B, protoplast extract; C, vacuole extract; D, enriched cytosol extract. A1, B1, C1, and D1 (left), 31P-NMR spectra (expanded scale from 0 to 5 ppm); A2, B2, C2, and D2 (right), 13C-NMR spectra (expanded scale from 54.2 to 55.1 ppm). Protoplast preparation affected phosphomonoester concentrations (B1 and B2). Vacuoles were prepared by flotation (see "Materials and Methods"). Contaminants from the cytoplasm did not exceed 5%, as determined with the enzymatic markers Glc-6-PDH and alpha-mannosidase. Enriched cytosol was highly contaminated (up to 40%) with vacuole sap, as shown by the presence of phytate. Glycerophosphodiester (GroPCho, GroPIns, and GroPGro) concentrations observed in total cell extracts were poorly affected by the addition of exogenous GroPCho, as discussed in the text. P-Cho and glycerophosphodiesters were essentially located in the cytosol. The low levels of P-Cho detected in the vacuole (C1) were consistent with the amount of contaminants (5%). Intracellular concentrations of GroPCho (peak at 0.1 ppm), glycerophosphoinositol (GroPIns, peak at 0.2 ppm), and glycerophosphoglycerol (GroPGro, peak at 1.1 ppm) were minimally affected by the addition of exogenous GroPCho. Moreover, analysis by 31P-NMR spectroscopy of glycerophosphodiesters in different cellular subfractions gave us interesting data concerning their intracellular localization. GroPCho, GroPIns, and GroPGro resonance peaks were exclusively found in the cytosol (Fig. D1) and were absent from the vacuoles (Fig. C1). Finally, we should mention the fate of glycerol and phosphate resulting from Gro-3-P hydrolysis, although glycerol and phosphate uptakes have already been described . Glycerol diffused through the plasma membrane, probably through a glycerol facilitating system , and was rapidly metabolized, as revealed by its low accumulation in the extracellular medium during long-term incubation (not shown). In contrast, phosphate accumulated steadily in the medium, its low uptake reflecting the consumption rate during carrot cell culture. Localization of the Extracellular GPC-PDE in the Cell Wall | The results presented above pointed out the existence of a hydrolytic activity splitting GroPCho into choline and Gro-3-P outside the cells. To discriminate between the different extracellular compartments (proteins secreted in the culture medium and proteins bound to the cell wall or to the plasma membrane), we tested GPC-PDE activity in the culture medium, in microsomal fractions obtained by ultracentrifugation, in plasma membrane isolated in a two-phase partitioning system, and in cell wall proteins extracted by CaCl2 treatment (see "Materials and Methods"). No GPC-PDE activity was found in the membrane fractions (microsomes and isolated plasma membrane). After concentration by ultrafiltration (50x), GPC-PDE activity was detected in the medium, however, the recovery rate (1.0%) was too low to explain the activity observed with intact carrot cells . On the contrary, extraction of cell wall proteins by saline treatment yielded significant amounts of GPC-PDE . Digestion of carrot cell wall with 0.25% (w/v) pectolyase and 2% (w/v) cellulase conversely solubilized approximately 90% of the extracellular GPC-PDE. Taken together, these results strongly suggest that extracellular GPC-PDE is bound to the cell wall by ionic interactions. Table I | GPC-PDE activities from different extracellular extracts Finally, we compared the hydrolysis of GroPCho from cell wall extract at pH 6 and 8.5. At pH 8.5, Gro-3-P release represented 75% of the choline produced; in contrast, at pH 6, GroPCho hydrolysis released similar amounts of choline, glycerol, and phosphate (not shown), indicating the combined action of cell wall acid phosphatases and GPC-PDE observed on intact cells. Characterization of an Intracellular GPC-PDE Activity | During the above experiments, we maintained cell integrity to avoid exposing GroPCho to the contents of the intracellular compartments. At 30C, the rate of GroPCho hydrolysis measured on intact cells was approximately 4.9 mumol h-1 g-1. In contrast, when the carrot cells were broken by sonication, the total GPC-PDE activity observed was much higher: 10 mumol h-1 g-1 (Fig. , foreground). This large increase of total GPC-PDE activity revealed the existence of an intracellular GPC-PDE. This result was even more striking when we tested the existence of GPC-PDE in different plant cell cultures. In Arabidopsis, sycamore, and maize cell cultures, intracellular GPC-PDE represented a large majority of the total GPC-PDE measured in broken cells (Fig. , foreground). Figure 6 | GPC-PDE activities in different plant cell cultures. GPC-PDE activities in different plant cell cultures. Activities were measured on intact cells or cells broken by sonication. Cells were grown for 3 d in a culture medium in the presence (+P, foreground) or absence (-P, background) of phosphate 5 mm. Activities are expressed as micromoles of GroPCho hydrolyzed per hour and per gram fresh weight. Mean values and sd were calculated from three independent experiments. Localization of the Intracellular GPC-PDE in Vacuole | To localize more precisely the intracellular GPC-PDE activity, we fractionated carrot protoplasts. When all of the membrane systems were pelleted by ultracentrifugation, GPC-PDE activity remained in the supernatant (not shown). On the contrary, when vacuoles were isolated by flotation (see "Materials and Methods"), we observed a strong enrichment of the GPC-PDE activity expressed per unit weight of protein . This enrichment coincided with enrichment in alpha-mannosidase, a vacuolar marker, and a decrease in Glc-6-phosphate dehydrogenase (Glc-6-PDH), a cytosolic marker . When enzyme activities were expressed per unit weight of fresh cells, the recovery rate of GPC-PDE (13%) was comparable with that obtained with alpha-mannosidase (11%). In contrast, GPC-PDE contained only a minor fraction of the Glc-6-PDH activity (0.54%), reflecting the amount of cytosolic contaminants in the vacuolar fraction. Finally, comparison of vacuolar GPC-PDE activities at different pH also enabled us to distinguish between the phosphodiesterase and the phosphatase activity. GroPCho is similarly hydrolyzed into choline and Gro-3-P (not shown). Table II | Activities of GPC-PDE and different enzyme markers in carrot cells and in vacuoles Induction by Phosphate in Different Plant Cells | recently reported the existence of a cyclic-nucleotide phosphodiesterase induced by phosphate deprivation. Phosphate starvation has been described as a physiological situation that induces several phosphate-rescuing systems like acid phosphatases and phosphate transporters . Thus, we tested the effect of phosphate deprivation on GPC-PDE activities in our different cell culture models (Fig. , background). In carrot, Arabidopsis, and sycamore cells, both extra- and intracellular GPC-PDE activities were strongly stimulated. We conversely investigated the capacity of phosphate-deprived cells to recover from phosphate starvation when they were fed with GroPCho as the exclusive source of phosphorus. When phosphate-starved cells were incubated with 2 mm GroPCho for 24 h, intracellular phosphate and monoester concentrations increased steadily to return to normal levels (data not shown). Substrate Specificity of GPC-PDE Activity | Phosphodiesterases have been described recently in the cell wall or in the extracellular medium . Most were reported to split the artificial substrate bis-p-nitrophenyl phosphate (bis-PNPP). To discriminate between these phosphodiesterases and the extracellular GPC-PDE described in our work, we performed different competition assays on the cell wall protein extract. First, we estimated the Km of GPC-PDE for GroPCho at 50 mum, which denoted a high affinity for its substrate. Addition of 2 mm bis-PNPP, ADP, ADP-Glc, UDP-Glc, and ADP-Rib did not significantly affect GPC-PDE activity in the presence of high (2 mm) or limiting (50 mum) concentrations of GroPCho . Release of p-nitrophenol from bis-PNPP conversely remained constant in the absence or presence of GroPCho (not shown). Figure 7 | GPC-PDE activities in the presence of a second phosphodiester. GPC-PDE activities in the presence of a second phosphodiester. Cell wall proteins were extracted by CaCl2 treatment as described in the text. GPC-PDE was assayed in the presence of a limiting (50 mum) or a high (2 mm) concentration of GroPCho and 2 mm of the second phosphodiester. GPC-PDE activities are expressed as the percentage of GroPCho hydrolysis rate measured at 30C with 50 mum (or 2 mm) of pure GroPCho. We also tested phosphodiesterase activities with other glycerophosphodiesters involved in phospholipid metabolism (glycerophosphoethanolamine [GroPEtn], GroPIns, GroPGro, and glycerophospho-Ser [GroP-Ser]). The cell wall protein extract hydrolyzed GroPEtn, GroPIns, GroPGro, and GroP-Ser at rates comparable with GroPCho (Table , first lane). In the presence of a limiting concentration of GroPCho, GroPEtn inhibited competitively GroPCho hydrolysis, with an apparent Ki of 40 mum. At last, in the presence of a saturating concentration of GroPCho (2 mm), the addition of a second glycerophosphodiester (GroPEtn, GroPIns, and GroP-Ser) did not increase significantly the amount of glycerol and Gro-3-P released (Table , second lane), indicating that the enzymatic activity responsible for GroPCho also hydrolyzes the other glycerophosphodiesters. Table III | Rate of hydrolysis of various glycerophosphodiesters by cell wall GPC-PDE From these experiments, the cell wall GPC-PDE described in this article seems different from other plant phosphodiesterases and specific for glycerophosphodiesters. Vacuole GPC-PDE specificity remains unknown. DISCUSSION : The results presented in this article pointed out the existence of two hydrolytic activities splitting GroPCho into Gro-3-P and choline: The first GPC-PDE was localized on the cell wall, the second in the vacuole sap. In contrast, GroPCho and the other glycerophosphodiesters were found to accumulate steadily in the cytosol. Catabolism of GroPCho by Higher Plant Cells | The catabolism of GroPCho by extracellular GPC-PDE underlines various pathways involved in the uptake of choline and its further distribution within different plant cell compartments, as recapitulated in Figure . The combined actions of GPC-PDE, cell wall phosphatases, and different transport systems permitted the incorporation of all of the GroPCho chemical components: Glycerol diffused through the membranes directly or via glycerol facilitator systems and was rapidly metabolized while phosphate and choline were incorporated by different transport systems . In the cytosol, most of the choline was phosphorylated and steadily accumulated as P-Cho. Choline scavenging in the vacuole was relatively slight compared with the phosphorylation of choline by cytosolic choline kinase, but it pointed out the existence of a transport system between the vacuole and the cytosol. Finally, the cytosolic pool of GroPCho remained unaffected by the presence of exogenous GroPCho. Figure 8 | Catabolism of GroPCho in carrot cells. Catabolism of GroPCho in carrot cells. 1 and 2, Cell wall (1) and vacuole (2) GPC-PDE activities described in the present work; 3, hydrolysis of Gro-3-P by the acid phosphatase; 4, free diffusion of glycerol through the plasma membrane or the tonoplast; 5 and 6, incorporation of phosphate and choline through its transporters; 7, phosphorylation of choline in the cytosol by choline kinase; 8, choline incorporation in the vacuole deduced from the NMR experiment (see Fig. ); 9 and 10 (dashed lines), hypothetical transport of GroPCho through the tonoplast or the plasma membrane; similar transport has been shown through the plasma membrane in non-plant organisms. Originality of Plant GPC-PDE | Various phosphodiesterases have already been reported in the vacuole and in the extracellular compartments . However, none of these authors reported a specific action on glycerophosphodiesters. On the contrary, preliminary results obtained with the competition assays described in this paper suggest that the plant GPC-PDE are specific for glycerophosphodiesters and remain unaffected by previously described phosphodiesterases substrates (bis-PNPP, ADP-Glc, UDP-Glc, or ADP-Rib). Regarding its specificity for glycerophosphodiesters, the plant GPC-PDE seems to be closer to the non-plant glycerophosphodiester phosphodiesterase (GPX-PDE). In other eukaryotes, GPX-PDE activities have been described in yeast and animal cells and were often found to be extracellular. All of these activities hydrolyze glycerophosphodiester into Gro-3-P and the corresponding alcohol (choline, ethanolamine, inositol, or glycerol) and can function in alkaline conditions. Nevertheless, none of these activities have been characterized at the molecular level. Finally, in prokaryotes, a periplasmic GPX-PDE encoded by the glpQ gene was thoroughly described in Escherichia coli . Physiological Significance of Plant GPC-PDE Activity | Plant intracellular GPC-PDE is in the vacuole, a cellular compartment where several hydrolases have been reported, including acid phosphatase and nonspecific phosphodiesterases (for review, see ). In autophagy triggered by carbon starvation, the GroPCho degradation coincided with vesicle formation and fusion with the central vacuole, suggesting an important role for vacuolar hydrolytic enzymes . The vacuole GPC-PDE may also participate in membrane degradation in different physiological situations of programmed cell death: In tracheary element differentiation , tonoplast rupture is considered as a critical event in cell death. In senescence, lipolytic acyl hydrolases, which release glycerophosphodiesters from phospholipids, seem to play an important role in membrane disruption . Thus, the modification of vacuoles during different senescence processes could release different hydrolases from the vacuole sap. GPC-PDE would be one of the enzymes involved in phospholipid catabolism: After phospholipid hydrolysis, the combined actions of GPC-PDE and acid phosphatase would release glycerol, phosphate, and free alcohol. The significance of extracellular GPC-PDE is also enigmatic. Comparison with other organisms could provide some clues. Many extracellular GPX-PDE are enhanced in different starvation conditions. In bacteria, the E. coli glpQ gene belongs to the glycerol operon and is associated to the uptake of glycerol when the carbon supply is very low . In Saccharomyces uvarum , the enhancement of GPX-PDE activities is similarly triggered by inositol or phosphate starvation. In plants, the significance of the extracellular GPC-PDE activity could also be related to starvation conditions. The activation of different phosphatases and phosphodiesterases during phosphate deprivation has already been described in other plant systems . In the case of nucleic acids, the combined actions of ribonuclease, cyclic nucleotide phosphodiesterase, and phosphatase enabled cell growth on a nutrient medium containing nucleic acid as the only source of phosphate . Interestingly, the phospholipid content of the soil represents from 0.5% to 7% of the immobilized phosphate, which is comparable with the nucleic acid concentration . Therefore, GPC-PDE could be one of the enzymes involved in phospholipid degradation, especially when the phosphorus supply is low. In our experiments, GroPCho and other glycerophosphodiesters were supplied exogenously. Nevertheless, the in vivo occurrence of extracellular glycerophosphodiesters remains unclear. During barley (Hordeum vulgare) seed germination, GroPCho accumulation coincides with the accumulation and hydrolysis of starch-bound lysophosphatidylcholine in the endosperm . Moreover, the amount of starch-bound lysophosphatidylcholine released is consistent with the concentration of GroPCho measured in young rice shoots . Therefore, the rapid hydrolysis, which involves the action of different lysophospholipases secreted from the aleurone layers , would release quantitative amounts of GroPCho in the endosperm. In addition to germination, we should mention four hypothetical situations that could trigger glycerophosphodiester accumulation in the extracellular compartment: (a) Although there is no evidence of excretion of glycerophosphodiesters in plants, this mechanism is widespread in other eukaryotes like yeast and animal cells ; (b) the phospholipid turnover of the plasma membrane may also lead to the production of extracellular glycerophosphodiesters because in yeast, it implies the action of extracellular acyl-transferase and phospholipase B ; (c) transport of GroPCho through the xylem may occur because measured high concentrations of P-Cho in the xylem sap; and (d) in physiological situations that affect the plant integrity (wounding, pathogen attack, and senescence), intracellular pools of organic-phosphate are released in the extracellular medium whereas actions of different lipases release phospholipids catabolites. The different phosphatases and phosphodiesterase may allow reuse of various catabolites in the neighborhood of the damaged tissues. MATERIALS AND METHODS : Plant Material | Carrot (Daucus carota) and maize (Zea mays) cell suspension cultures were grown in the liquid medium described by supplemented with 30 g L-1 Suc and 0.1 mg L-1 2,4-dichlorophenoxyacetic acid. Sycamore (Acer pseudoplatanus) cell suspension cultures were grown in a liquid nutrient medium according to the method of . Arabidopsis cell suspension cultures were grown in Gamborg B5 medium supplemented with 15 g L-1 Suc and 0.1 mg L-1 2,4-dichlorophenoxyacetic acid. To establish Pi deprivation, cells were maintained in exponential growth by subculture every 3 d. Pi-starved cells were obtained by removing the culture medium, rinsing the cells three times with water, and adding the same culture medium where Pi was omitted. Chemicals | GroPCho was purchased from Sigma (St. Louis). Because GroPCho was supplied as a CdCl2 complex, we removed the cadmium by elution through an IRC-50 cation exchange column equilibrated with 100 mm potassium acetate, pH 6. Production of [3H]GroPCho | [3H]GroPCho was obtained from [3H]methyl-phosphatidylcholine by mild deacylation in 0.2 m methanolic sodium hydroxide . Five to 50 muCi of labeled phospholipid was incubated for 15 min at 25C with chloroform:methanol:methanolic NaOH 0.2 m (2:3:5, v/v). After centrifugation, the upper methanol-water phase was neutralized by elution through a cation ion-exchange resin (Dowex-50) and addition of NH4OH. In a parallel experiment performed with unlabeled phosphatidylcholine, we checked the purity of GroPCho produced after deacylation using 31P-NMR analysis. NMR Measurements | In Vivo 31P-NMR Measurements | Spectra performed on intact cells were recorded on an NMR spectrometer (AMX 400, wide bore, Bruker, Billerica, MA) equipped with a 25-mm probe tuned at 162 MHz. Acquisition conditions were: 50 radio frequency pulses (70 mus) at 0.6-s intervals; spectral width, 9,800 Hz; 1,500 scans; and Walz-16 1H decoupling sequence (with two levels of decoupling: 2.5 W during acquisition time, 0.5 W during delay). Free induction decays were collected as 4,000 data points zero-filled to 8,000 and processed with 2-Hz exponential line broadening. Spectra are referenced to a solution of 50 mm diphosphonic acid, in 30 mm Tris contained in a 0.8-mm capillary inserted inside the inlet tube along the symmetry axis of the cell sample (see ). The assignment of Pi, phosphate esters, and phosphate diesters was carried out according to and from spectra of perchloric extracts that contained the soluble, low-molecular-weight constituents. Cells (10 g wet weight) were introduced in the 25-mm NMR tube as described by . The perfusion medium (200 mL) consisted of a 10x diluted growth medium where manganese was omitted. The pH of the external medium was adjusted to 6.0 or 8.5 before experiments and recorded using a pH electrode immersed in the reservoir of perfusion medium. Temperature was kept at 27C. After the incubation, perfusion medium and cells were separated by filtration. The cells were rinsed with water, frozen in liquid nitrogen, and conserved at -70C for further analysis. In Vitro 31P- and 13C-NMR Measurements | Perchloric acid extracts were obtained as described in a previous paper . Spectra of neutralized perchloric acid extracts were recorded on the same NMR spectrometer equipped with a 10-mm multinuclear probe tuned at 162 or 100.6 MHz for 31P- or 13C-NMR studies, respectively. The deuterium resonance of 2H2O was used as a lock signal. 31P-NMR acquisition conditions used were: 70 radio-frequency pulses (15 mus) at 3.6-s intervals; spectral width, 8,200 Hz; and Waltz-16 1H decoupling sequence (with two levels of decoupling: 1 W during acquisition time, 0.5 W during delay). Free induction decays were collected as 8,000 data points, zero-filled to 16,000, and processed with 0.2-Hz exponential line broadening. 31P-NMR spectra were referenced to methylene diphosphonic acid, pH 8.9, at 16.38 ppm. 13C-NMR acquisition conditions used were: 90 radio-frequency pulses (19 mus) at 6-s intervals; spectral width, 20,000 Hz; 900 scans; and Waltz-16 1H decoupling sequence (with two levels of decoupling: 2.5 W during acquisition time, 0.5 W during delay). Free induction decays were collected as 16,000 data points, zero-filled to 32,000, and processed with 0.2-Hz exponential line broadening. 13C-NMR spectra were referenced to hexamethyldisiloxane at 2.7 ppm. Spectra of standard solutions of known compounds at pH 7.5 were compared with the spectrum of a perchloric acid extract of sycamore cells. The definitive assignments were made after running a series of spectra obtained by the addition of the authentic compounds to the perchloric acid extracts, according to the methods described in previous publications (for 31P-NMR, see ; for 13C-NMR, see ). Enzymatic Determination of Choline and Glycerol | Choline was determined by the method of . In the presence of choline kinase and ATP, choline was phosphorylated to P-Cho. Successive enzymatic steps (pyruvate formation from ADP and PEP and reduction of pyruvate to lactate) led to an equimolar oxidation of NADH, which was measured photometrically. Glycerol and Gro-3-P were determined by the method of also based on enzymatic analysis. Enzymatic Assays | GPC-PDE activity was assayed in a two-step enzymatic test. At high concentrations of GroPCho (from 1 to 5 mum), active fractions were incubated with GroPCho in 50 mm MES and 5 mm CaCl2, pH 6 buffer. The reaction was stopped with the addition of a few drops of 70% (v/v) perchloric acid. The sample was neutralized with saturating KHCO3. After centrifugation, choline was determined as described above. For competition assays with other glycerophosphodiesters (GroPEtn, GroPIns, GroP-Ser, and GroPGro), glycerol and Gro-3P were also determined after KHCO3 neutralization. At lower concentrations of GroPCho (from 5 to 500 mum), enzymatic assays were performed incubating [3H]GroPCho (1.5 muCi mumol-1) in the reaction buffer (50 mm MES and 5 mm CaCl2, pH 6). Enzymatic reaction was stopped by heating (95C) for 10 min. After drying, the reactive mixture was resuspended in ethanol:water (50:50, v/v) and loaded on a silica-gel plate. [3H]Choline and [3H]GroPCho were separated with thin-layer chromatography in a methanol:0.5% NaCl: NH4OH (100:100:2, v/v) mixture, as described by . Subcellular Fractionation | The clear culture medium was concentrated by ultrafiltration with the Diaflo system (Amicon, Beverly, MA) and a PM-10 membrane. Microsomal fractions were obtained after cell disruption by the French press in 0.5 m Suc, 2 mm EDTA, 10 mm dithiothreitol (DTT), 1 mm phenylmethylsulfonyl fluoride (PMSF), and 100 mm Tris, pH 8. After a first centrifugation (16,000g, 20 min), the supernatant was filtered on a 50-mum nylon netting. Microsomes were pelleted by ultracentrifugation (96,000g, 35 min). Plasma membrane was isolated from the microsomes by phase partition as described by . Cell wall proteins were extracted by CaCl2 treatment using a protocol adapted from . Intact cells were harvested by filtration and carefully rinsed with water. Then, the cells were stirred in 5 volumes (volume/fresh weight) of 0.2 m CaCl2 and Tris 20 mm, pH 7.4, for 5 min at 30C. After removing the cells by filtration, the saline extract was dialyzed against Tris 20 mm, pH 7.4, and concentrated by ultrafiltration on a PM-30 membrane. The absence of intracellular contaminants was checked using Glc-6-PDH as an intracellular marker. Vacuoles were isolated by flotation, using a protocol adapted from . Protoplasts were obtained after cell digestion in 0.6 m mannitol, 2% (w/v) cellulase (Seichin-Kyowa, Osaka), 0.25% (w/v) pectolyase (Seishin), and 25 mm MES, pH 5.5. Protoplasts were filtered on 50-mum nylon netting and pelleted by centrifugation at 800g in a swinging bucket rotor. After rinsing with 0.7 m mannitol and 25 mm Tris, pH 7, protoplasts were resuspended in extraction buffer: 0.5 m Suc, 4% (w/v) Ficoll, 50 mm Tris, pH 8, 2 mm EDTA, 0.2 mm PMSF, and 1 mm DTT. With a syringe, protoplasts were passed through a 20-mum nylon netting, which disrupted the plasma membrane and released intact vacuoles. Protoplast homogenate (about 20 mL) was loaded at the bottom of a centrifuge tube. Above it, 15 mL of flotation buffer (0.5 m Suc, 50 mm Tris, 2 mm EDTA, 0.2 mm PMSF, and 1 mm DTT) was layered. The preparation was centrifuged in a swing-out rotor for 10 min at 500g and for 30 min at 10,000g. After centrifugation, vacuoles formed a thin layer floating on the top of the tube. Supernatant was also collected because it was enriched in cytosol sap though contaminated with broken vacuoles. The purity of vacuoles and cytosol was verified by comparing two enzyme markers : Glc-6-PDH for the cytosol and alpha-mannosidase for the vacuolar sap. Backmatter: PMID- 12226505 TI - Biochemical Characterization of the Arabidopsis Protein Kinase SOS2 That Functions in Salt Tolerance AB - The Arabidopsis Salt Overly Sensitive 2 (SOS2) gene encodes a serine/threonine (Thr) protein kinase that has been shown to be a critical component of the salt stress signaling pathway. SOS2 contains a sucrose-non-fermenting protein kinase 1/AMP-activated protein kinase-like N-terminal catalytic domain with an activation loop and a unique C-terminal regulatory domain with an FISL motif that binds to the calcium sensor Salt Overly Sensitive 3. In this study, we examined some of the biochemical properties of the SOS2 in vitro. To determine its biochemical properties, we expressed and isolated a number of active and inactive SOS2 mutants as glutathione S-transferase fusion proteins in Escherichia coli. Three constitutively active mutants, SOS2T168D, SOS2T168DDeltaF, and SOS2T168DDelta308, were obtained previously, which contain either the Thr-168 to aspartic acid (Asp) mutation in the activation loop or combine the activation loop mutation with removal of the FISL motif or the entire regulatory domain. These active mutants exhibited a preference for Mn2+ relative to Mg2+ and could not use GTP as phosphate donor for either substrate phosphorylation or autophosphorylation. The three enzymes had similar peptide substrate specificity and catalytic efficiency. Salt overly sensitive 3 had little effect on the activity of the activation loop mutant SOS2T168D, either in the presence or absence of calcium. The active mutant SOS2T168DDelta308 could not transphosphorylate an inactive protein (SOS2K40N), which indicates an intramolecular reaction mechanism of SOS2 autophosphorylation. Interestingly, SOS2 could be activated not only by the Thr-168 to Asp mutation but also by a serine-156 or tyrosine-175 to Asp mutation within the activation loop. Our results provide insights into the regulation and biochemical properties of SOS2 and the SOS2 subfamily of protein kinases. Keywords: Introduction : During growth and development of many multicellular organisms, protein kinases function in a variety of signaling pathways critical for cell division, metabolism and response to hormonal, developmental, and environmental signals. The activity of protein kinases can either be stimulatory or inhibitory to downstream targets . Knowledge of how the relevant protein kinases are regulated, therefore, is one key to understanding basic cellular processes involved in growth and development. The SNF1/AMPKs are highly conserved Ser/Thr protein kinases identified in fungi, fruitfly (Drosophila melanogaster), Caenorhabditis elegans, mammals, and plants . Many SNF1-related protein kinase genes (SnRKs) have been isolated in plants, and these SnRK kinases have been classified into three subgroups (SnRK1, SnRK2, and SnRK3) based on sequence similarity . The Arabidopsis Salt Overly Sensitive 2 (SOS2) and Salt Overly Sensitive 3 (SOS3) genes were isolated through positional cloning and were shown to be required for sodium and potassium ion homeostasis and salt tolerance . SOS2 encodes a 446-amino acid Ser/Thr protein kinase with an N-terminal kinase catalytic domain similar to SNF1/AMPK and a novel C-terminal regulatory domain . SOS2 can be classified as a member of the SnRK3 subgroup of SNF1-related protein kinases. SOS3 encodes a myristoylated EF-hand calcium-binding protein that may sense the calcium signal elicited by salt stress . SOS3 physically interacts with SOS2 in the yeast (Saccharomyces cerevisiae) two-hybrid system and in vitro . Both the catalytic and regulatory domains are essential for SOS2 function in salt tolerance . A 21-amino acid sequence in the regulatory domain of SOS2, the FISL motif, has been determined to be necessary and sufficient to bind SOS3 . Salt stress up-regulation of the Salt Overly Sensitive 1 (SOS1) gene encoding a Na+/H+ antiporter is partially under control of the SOS2-SOS3 regulatory pathway . SOS2 and SOS3 are more importantly both required for the posttranslational activation of SOS1 Na+/H+ exchange activity . Recently, we have characterized the functional domains in SOS2 kinase . Recombinant SOS2 protein produced in bacteria exhibits no substrate phosphorylation activity in the absence of SOS3, although it has autophosphorylation activity . In the presence of calcium, SOS3 activates the substrate phosphorylation activity of SOS2 . The substrate phosphorylation activity of SOS2 could also be activated by a Thr-168 to Asp mutation within the activation loop or by removal of the autoinhibitory FISL motif . In this study, we used the constitutively activated SOS2 mutants to characterize the biochemical properties of SOS2. These properties include divalent cation preference, phosphate donor specificity, steady-state substrate kinetics, and the reaction mechanism of autophosphorylation. We also discovered that the substitution of Ser-156 or Tyr-175 within the activation loop with Asp could also activate SOS2. These results help understand the biochemical characteristics and the regulation of SOS2 protein kinase. RESULTS : Expression, Purification, and Kinase Activities of Recombinant SOS2 Mutants | Bacterially expressed SOS2 recombinant protein is inactive by itself in peptide substrate phosphorylation and becomes active in the presence of SOS3 that binds to the autoinhibitory FISL motif of SOS2 . Three SOS3-independent, constitutively active mutants SOS2T168D, SOS2T168DDeltaF, and SOS2T168DDelta308 were produced, which contain the activation loop Thr-168 to Asp mutation or combine the activation loop mutation with removal of the FISL motif or regulatory domain . These were expressed here as glutathione S-transferase (GST)-tagged fusion proteins in E. coli and affinity-purified on glutathione-Sepharose. The eluting proteins were analyzed for purity by SDS-PAGE (data not shown). Each protein migrated as predicted from its molecular mass. The purity of these preparations was estimated to be above 95%, and their identities were confirmed by western analysis (data not shown). These purified kinase samples were used for the remainder of this study. An exogenous peptide, p3 (ALARAASAAALARRR), derived from the recognition sequences of protein kinase C and SNF1/AMPK, was earlier shown to be phosphorylated by SOS2 in the presence of SOS3 . Kinase activity of the purified recombinant proteins was evaluated by measuring phosphorylation activity toward this peptide substrate, without addition of SOS3. A standard kinase assay with 5 mm Mg2+ showed that these SOS2 mutants displayed much higher phosphorylation of the peptide substrate p3 and autophosphorylation (data not shown) than either SOS2 alone or SOS2 in the presence of SOS3 (designated SOS2/SOS3) did. These active SOS2 mutants were, thus, chosen for further biochemical characterization. Figure 1 | Kinase activities of the purified recombinant SOS2T168D, SOS2T168DDeltaF, and SOS2T168DDelta308 fusion proteins. Kinase activities of the purified recombinant SOS2T168D, SOS2T168DDeltaF, and SOS2T168DDelta308 fusion proteins. The SOS2, SOS3, and SOS2 mutant cDNAs were expressed as GST-tagged fusion proteins in E. coli and purified by glutathione-Sepharose affinity chromatography. Peptide phosphorylation activities of SOS2 and SOS2 mutants were measured using p3 as a peptide substrate in the presence of 5 mm Mg2+. Error bars indicate +-sd (n = 3). Phosphorylation of the peptide substrate by all SOS2 mutants was linear during 30 min (data not shown). The autophosphorylation activity of these proteins was detectable in 5 min, the first time point assayed, and completed after 30 min (data not shown). All subsequent kinase assays, therefore, were routinely terminated at 30-min incubations to obtain a good estimate of the initial rate. In addition, we have found that the purified enzyme is highly stable when maintained in a concentrated solution, even at room temperature but rapidly loses activity upon dilution. Enzyme inactivation after dilution could result from alterations in either the tertiary structure of the enzyme or its aggregation state. Our observations are similar to previous reports demonstrating that the aggregation state of a type 1 receptor Tyr kinase catalytic domain significantly affects the rate of catalysis . Divalent Cation Preference | Kinases, like other phosphotransferases, require a divalent cation to coordinate the phosphate groups of the nucleotide triphosphate substrate. These enzymes can also be activated or inactivated by binding of a cation to an additional site of interaction . To determine the divalent cation preferences in vitro of these mutants, we measured peptide substrate phosphorylation in the presence of various concentrations of either Mg2+ or Mn2+ (Fig. , A and B; SOS2T168D data not shown). With all kinases, there was no substrate phosphorylation in the absence of added divalent cation, and all of them showed higher rates with Mn2+ than with Mg2+. Minimal concentrations for any activity were 0.25 mm for Mn2+ and 0.5 mm for Mg2+, and optimal concentrations were 2.5 and 5 mm, respectively. Above 2.5 mm, Mn2+ was inhibiting for all kinases. ATP was held constant at 10 mum in the experiment, and so any concentration of cation above 10 mum is essentially free from bound nucleotide. The difference must, therefore, reflect different affinity of Mn2+ and Mg2+ for binding to a cation site on the enzyme. The apparent Km and Kcat values for ATP with either Mg2+ or Mn2+ were determined at a constant concentration of peptide substrate p3. These experiments were performed at the optimal concentration of free Mg2+ or Mn2+ as seen in Figure . The titrations of ATP with Mg2+ or Mn2+ were conducted in parallel, and the reactions were initiated with the same diluted enzyme mixture to ensure that the results are directly comparable. The results are summarized in Table . For all enzymes, the Kcat/Km was 4- to 5-fold higher for ATP with Mn2+ than for ATP with Mg2+. Figure 2 | Divalent cation dependence of peptide phosphorylation. Divalent cation dependence of peptide phosphorylation. Peptide phosphorylation reactions by the kinases were performed using peptide substrate p3 at various concentrations of Mn2+ (as MnCl2) or Mg2+ (as MgCl2) as indicated. Initial rates were measured and plotted against the Mn2+ or Mg2+ concentrations. A, SOS2T168DDelta308. B, SOS2T168DDeltaF. Table I | Kinetic parameters for ATP metal To determine the metal cation requirement of the SOS2 mutants for autophosphorylation, we used the same concentration series of the two divalent cations in the assay system. Mn2+ provided significant activation by 1 mm with all enzymes (Fig. for SOS2T168DDelta308; SOS2T168D and SOS2T168DDeltaF data not shown). A similar level of activation required 10 mm of Mg2+. It is apparent that autophosphorylation prefers Mn2+ over Mg2+, but there are very different optimal levels than those seen in peptide phosphorylation. In subsequent studies, we used the concentration of Mn2+ (2.5 mm) optimal for substrate phosphorylation and adequate for autophosphorylation. Figure 3 | Dependence of autophosphorylation of SOS2T168DDelta308 on divalent cations. Dependence of autophosphorylation of SOS2T168DDelta308 on divalent cations. Autophosphorylation of SOS2T168DDelta308 in the presence of various concentrations of Mn2+ (as MnCl2) or Mg2+ (as MgCl2), as indicated, was presented as the density of autoradiographic bands. Three independent experiments were performed, and a typical result is shown here. A small number of protein kinases use GTP and ATP as a phosphate donor. These include a receptor-like kinase from Madagascar periwinkle , a human STE20-like Ser/Thr protein kinase , and an Arabidopsis Ser/Thr protein kinase CK2 . To test for GTP use with our enzymes, increasing concentrations of unlabeled GTP were used to compete with [gamma-32P]ATP in the substrate phosphorylation and autophosphorylation assays. Cold GTP did not compete with ATP for all kinases in substrate phosphorylation of p3 or in autophosphorylation (data not shown). To further confirm the specificity of ATP as a phosphate donor, kinase assays were performed using either [gamma-32P]ATP or [gamma-32P]GTP at identical specific activities. None of them could use GTP as a phosphate donor for both autophosphorylation and peptide phosphorylation (data not shown). Steady-State Peptide Substrate Kinetics | Both substrate phosphorylation and autophosphorylation of the three kinases had pH optima between 7.0 and 7.5, and activity was optimal at 30C for all kinases (data not shown). In addition to p3, SOS2 could phosphorylate two other synthetic peptides: p1 (LRRASLG) and p2 (VRKRTLRRL), derived from the recognition sequences of protein kinase C or SNF1/AMPK . To evaluate peptide substrate preference of these kinases, we analyzed the steady-state kinetic parameters toward the three peptides. Apparent Km and Kcat values for p1, p2, and p3 of the three kinases were determined from Eadie-Hofstee plots of V0 versus V0/[S] (data not shown). The Kcat to Km ratios show clearly that all kinases prefer p3 as a substrate to either p1 or p2 . Although p3 is not based on a physiological substrate for SOS2, all these Km values are within the range of those found for plant SNF1-related kinases with peptide substrates that do reflect true physiological substrates. For one example, cauliflower (Brassica oleracea) 3-hydroxy-3-methylglutaryl-CoA reductase kinasehad a Km of 95 mum for the SAMS peptide based on conserved residues of known physiological substrates . Table II | Peptide substrate steady-state kinetic parameters with three peptide substrates Effect of SOS3 on Kinase Activity of SOS2T168D | The regulatory protein SOS3 has been shown to activate SOS2 in a Ca2+-dependent manner by binding to the autoinhibitory FISL motif in the C-terminal domain of SOS2 . To test whether SOS3 still could enhance the activity of the activation loop Thr-168 to Asp mutant SOS2T168D (with no deletions), we compared the phosphorylation of p3 by SOS2T168D with or without SOS3, either in the presence or absence of 0.5 mm calcium. SOS3 had no significant effect on p3 phosphorylation by SOS2T168D either in the presence or absence of calcium . SOS3 also exhibited little effect on autophosphorylation activity, either with or without calcium (data not shown). These observations suggest that the Thr-168 to Asp mutation within the activation loop could release (at least partially) the autoinhibitory effect of the FISL motif on SOS2 kinase activity, thus, making SOS2 independent of the regulatory protein SOS3. In addition, calcium (0.5 mm) was not required for kinase activity of SOS2T168D, although it seemed to slightly activate SOS2T168D either in the presence or absence of SOS3 . At the present time, the significance and potential mechanism of this slight calcium enhancement of SOS2T168D activity is unclear. Figure 4 | Effect of SOS3 on kinase activity of the activation loop mutant SOS2T168D. Effect of SOS3 on kinase activity of the activation loop mutant SOS2T168D. Substrate phosphorylation of SOS2T168D using peptide substrate p3 was measured with SOS3 or without SOS3 in the presence or absence of 0.5 mm Ca2+ (as CaCl2) in the kinase buffer. Error bars indicate +-sd (n = 3). Autophosphorylation Mechanism | In many cases autophosphorylation of a protein kinase has been shown to proceed by an intermolecular mechanism in which the catalytic domains and phosphorylation sites reside on separate molecules. To test by which mechanism SOS2 kinase could autophosphorylate, we determined the dependence of autophosphorylation activity on protein concentration. The autophosphorylation reactions of all these kinases showed first order kinetics (a linear increase in rate with increasing kinase protein) rather than second order (rate increases with the square of kinase concentration; Fig. A for SOS2T168DDelta308; data not shown), which suggests that SOS2 may autophosphorylate by an intramolecular reaction . Figure 5 | Intramolecular autophosphorylation mechanism of SOS2. Intramolecular autophosphorylation mechanism of SOS2. A, A plot of SOS2T168DDelta308 autophosphorylation versus its protein concentration. Actual amounts of SOS2T168DDelta308 varied from 0 to 1.05 mug per 30-muL reaction. A 30-muL reaction mixture contained various concentrations of the purified SOS2T168DDelta308 protein as indicated, 5 muCi of [gamma-32P]ATP, and 2.5 mm Mn2+ (as MnCl2) in kinase assay buffer. Reactions were incubated for 30 min at 30C. After separation on 10% (w/v) SDS-PAGE, the resulting gels were autoradiographed using a phosphor imager. The first order kinetics suggests an intramolecular autophosphorylation mechanism. B, Protein stain. Lanes 1, 2, and 3 represent Coomassie Blue-stained gel corresponding to lanes 1, 2, and 3 in C. C, Autoradiographs of phosphorylation assays of SOS2K40N and SOS2T168DDelta308 proteins. Eight hundred nanograms of either protein as shown in B was incubated alone or together in the presence of 5 muCi of [gamma-32P]ATP in kinase assay buffer as described above, fractionated by 10% (w/v) SDS-PAGE, and exposed to x-ray film. Lane 1, SOS2K40N; lane 2, SOS2K40N and SOS2T168DDelta308; and lane 3, SOS2T168DDelta308. Lack of a labeled band of the same size in lane 2 of B shows that SOS2T168DDelta308 (62 kD) cannot transphosphorylate SOS2K40N (80 kD), again suggesting an intramolecular autophosphorylation mechanism. To provide additional evidence, we tested whether the truncated active protein SOS2T168DDelta308 was able to transphosphorylate the kinase-dead mutant SOS2K40N. This mutant has the residue Lys-40, a conserved amino acid in the catalytic site required for phosphotransfer activity in all protein kinases , changed to Asn through site-directed mutagenesis. The active kinase SOS2T168DDelta308 could be clearly resolved from the full-length protein SOS2K40N by SDS-PAGE (Fig. B). This inactive protein was then co-incubated with SOS2T168DDelta308 in the kinase assay. As expected, SOS2K40N failed to autophosphorylate (Fig. C, lane 1), and SOS2T168DDelta308 had high autophosphorylation activity (Fig. C, lanes 2 and 3). The SOS2K40N was not trans-phosphorylated in the presence of SOS2T168DDelta308 (Fig. C, lane 2), and no autophosphorylation of the inactive SOS2 mutant was detectable even after very long exposures (data not shown). These results are strong evidence for an intramolecular rather than intermolecular reaction mechanism of SOS2 autophosphorylation. Thus, unlike some other Ser/Thr protein kinases , there is no need to postulate oligomerization of the protein as part of the mechanism. Activation by Substitution of Either Ser or Tyr with Asp within the Activation Loop | Many protein kinases are activated by phosphorylation of one or more residues within an activation loop. The introduction of a phosphate results in ionic interactions that are critical to kinase activity . In some protein kinases, such as phosphorylase b kinase and phosphoenolpyruvate carboxylase kinase , the phosphorylation site within the activation loop is replaced by a negatively charged residue. These kinases are indeed constitutively active, and do not require phosphorylation. SOS2 protein is not constitutively active in substrate phosphorylation. In the SOS2 kinase subfamily, the activation loop is located between the conserved DFG and APE residues in the kinase catalytic domain . A comparison of the activation loops of 23 members of the kinase subfamily showed that in addition to a Thr residue, either a Ser or Tyr residue is completely conserved in all members of this subfamily (Fig. ; data not shown). By mutating the conserved Thr to Asp in its activation loop, we earlier created a constitutively active SOS2 kinase . Here, we wanted to see if changing either the conserved Ser or Tyr to Asp could also make these kinases constitutively active. We constructed two activation loop single mutants, designated SOS2S156D and SOS2Y175D by mutating Ser-156 to Asp and Tyr-175 to Asp, respectively via site-directed mutagenesis. Figure 6 | Sequence alignment of the SOS2 activation loop compared with the corresponding region of the sequence of two SOS2-like kinases PKS11 and PKS18. Sequence alignment of the SOS2 activation loop compared with the corresponding region of the sequence of two SOS2-like kinases PKS11 and PKS18. A, Shown is a schematic diagram of the domain structure for SOS2. FISL, FISL motif. B, An alignment of SOS2, PKS11, and PKS18 activation loop. PKS11 and PKS18 are identical to gene products with the following GenBank accession numbers: and , respectively. Amino acids are numbered on the left. Identical residues and conservative replacements are shown with black and gray shading, respectively. The three conserved residues in all members of SOS2 subfamily kinase are marked with asterisks (Ser and Tyr) and a dot (Thr), respectively. Dashed lines represent spaces that were introduced to maximize alignment. We expressed these mutants, and also the wild-type protein as GST fusion proteins in E. coli, and purified them on glutathione-Sepharose (data not shown). SDS-PAGE analysis shows that the expression level of the recombinant mutant proteins was similar to that of the wild type (data not shown). Kinase assays showed that SOS2S156D and SOS2Y175D exhibited 19- and 23-fold higher activity, respectively, in p3 phosphorylation than the wild-type kinase (data not shown). Both activation loop mutant kinases also had higher autophosphorylation activity compared with the wild type (data not shown). DISCUSSION : SOS2 Exhibits Unusual But Not Unique ATP-Metal Substrate Preference | The biochemical characteristics of SOS2 and mechanisms of the regulation of its kinase activity are not fully understood. The current evaluation of the biochemical and kinetic properties of the SOS2 has been greatly facilitated by the availability of recombinant SOS2 mutants that show robust kinase activities. All of these three SOS2 mutants exhibited a strong preference for Mn2+ over Mg2+ for peptide substrate phosphorylation and autophosphorylation . We conducted ATP kinetic studies in the presence of the optimal divalent metal concentration. These studies show that all of these proteins have a higher affinity and catalytic efficiency toward ATP with Mn2+ than with Mg2+ . Activity of a number of other kinases also works better with Mn2+ than with Mg2+ . This divalent metal preference has been thought to reflect involvement of the kinase in a complex for full activation . The intracellular concentration of Mn2+ is in the micromolar range, whereas the Mg2+ concentration is in the millimolar range . The minimal and optimal activation concentrations of Mn2+ for SOS2 mutants in vitro (Figs. and ) suggest that Mn2+ may not play a role in SOS2 activity regulation under physiological conditions. Peptide Substrate Selectivity | SOS2 can be classified as a member of the SnRK3 family of SNF1-related protein kinases . The physiological substrate (s) of SOS2 kinase is still unknown. Testing the three different peptide substrates with these kinases, the Kcat to Km ratios show clearly a preference order of p3 > p2 > p1 . The apparent Km values toward the three peptides for these kinases are similar to those for SNF1/AMPK from yeast, mammalian, and higher plants with Ser-containing peptides as substrates . The preferred peptide substrate p3 (ALARAASAAALARRR) contains within it the following sequence: hydrophobic-X-basic-X (2)-Ser-X (3)-hydrophobic residue. The same motif has been previously established as a minimal recognition motif for the cauliflower AMPK/SNF1 homolog . An alignment of sequences around the phosphorylation sites on 3-hydroxy-3-methylglutaryl-CoA reductases, nitrate reductases, and Suc phosphate synthases from different plant species identified a consensus recognition motif for SnRK1 protein kinases , and this is shared by p3. Therefore, the presence of the hydrophobic and basic residues may be a determinant for the substrate specificity of the SOS2. Comparison of the kinetic data of these SOS2 mutants shows that removing the regulatory domain of SOS2 does not seem to significantly affect cofactor preference, peptide substrate specificity, and catalytic efficiency toward peptide substrates. We have previously reported that salt stress up-regulation of a Na+/H+ antiporter SOS1 gene is controlled by the SOS3-SOS2 regulatory pathway . We have recently shown that SOS3-SOS2 also directly activates SOS1 sodium-proton exchange activity . The C-terminal part of the SOS1 protein contains some putative consensus recognition motifs found in the preferred peptide substrate p3 of the SOS2 mutants. An SOS1-His tagged protein purified from yeast membranes has been recently observed to be phosphorylated in vitro by the SOS2T168DDelta308 . Therefore, SOS1 is likely one physiological substrate of SOS2. Activation of SOS2 by Multisite Phosphorylation within the Activation Loop | Regulation of protein kinases is achieved through many different mechanisms, including protein phosphorylation by other kinase(s) , autophosphorylation , or control by regulatory domains or subunits. A key feature for regulation in many protein kinases is thought to be the phosphorylation of one or more residues within the activation loop of the catalytic subunit . An unphosphorylated activation loop can block access of substrates to the active site, whereas phosphorylation can cause an outward rotation of the activation loop, making substrate accessible to the active site residues for catalysis . Three residues ---Ser-156, Thr-168, and Tyr-175 ---within the SOS2 activation loop are completely conserved among all members of the SOS2 kinase subfamily (Fig. ; data not shown). In addition to activation by converting the conserved Thr-168 to Asp , we have demonstrated here activation by changing the conserved Ser or Tyr to Asp (data not shown). These results are similar to findings with a number of other protein kinases in both plants and animals . It has been thought that phosphorylation of the activation loop shifts the equilibrium toward a conformation that accommodates protein substrate binding, and some data in the literature support this concept . Finally, recombinant SOS2 mutants have been useful because of their relative abundance compared with native SOS2 protein, the purification of which from Arabidopsis has not been possible because of its extreme low abundance (J.-K. Zhu, unpublished data). The recombinant SOS2 mutants are easily purified, and being catalytically active, have permitted biochemical analysis. Although these studies demonstrate the catalytic potential of the proteins, the biologically relevant form of the SOS2 kinase is most likely in the complex it makes with SOS3 and perhaps additional proteins; it will be of interest to see whether there are any significant differences in activity between the complex and the mutated active forms of SOS2. MATERIALS AND METHODS : Site-Directed Mutagenesis | A cDNA containing the complete open reading frame of SOS2 was obtained by reverse transcription-PCR as described by . Substitution of either Ser or Tyr with Asp within the activation loop of SOS2 was introduced using oligonucleotide-directed in vitro mutagenesis. The sequences of mutagenic oligonucleotide primers (MWG-Biotech, High Point, NC) were as follows: 5'-TTTCGGATTTCGGACTCGACGCATTGCCTCAGGAAGGAG-3' (SOS2S156D, forward), 5'-TCCTTCCTGAGGCAATGCGTCGAGTCCGAAATCCGAAACC-3' (SOS2S156D, reverse), 5'-ACATGTGGAACTCCGAACGACGTAGCTCCAGAGGTACTTAG-3' (SOS2Y175D, forward), and5'-AAGTACCTCTGGAGCTACGTCGTTCGGAGTTCCACATGTGG-3' (SOS2Y175D, reverse). In vitro mutagenesis reactions were performed on the plasmid DNA with a 1:1 (v/v) enzyme mix of LA Tag (TaKaRa Shuzo, Ltd., Kyoto) and Pfu Turbo DNA polymerase (Stratagene, La Jolla, CA) using the following PCR cycle: 95C for 30 s, followed by 16 cycles of 95C for 30 s, 60C for 1.0 min, and 72C for 7 min. The PCR products were gel-purified and treated with DpnI to digest the parental double-stranded DNA. The digested PCR products were then transformed into DH5alpha-competent cells. The sequences of mutation and the fidelity of the rest of the DNA in all constructs were confirmed by DNA sequencing. GST-SOS2K40N and GST-SOS3 constructs were produced as described by and , respectively. Expression of Kinase Fusions in E. coli and Protein Purification | All constructs were expressed in bacteria as a C-terminal fusion protein with the bacterial GST under control of the isopropyl beta-d-thiogalactopyranoside-inducible tac promoter. All mutant and wild-type GST fusion constructs were transformed into E. coli BL21 (codon plus) cells (Stratagene). Freshly transformed single colonies were grown overnight at 37C, transferred to fresh 1,000 mL of Luria-Bertani media, and further cultured until the A600 reached approximately 0.8. Recombinant protein expression was induced by 0.6 mm isopropyl beta-d-thiogalactopyranoside for 4 h. The cells were harvested by centrifugation (4,000g, 25 min, 4C), and the pellets were resuspended in a ice-cold bacterial lysis buffer containing 140 mm NaCl, 2.7 mm KCl, 10.1 mm Na2HPO4, 1.8 mm KH2PO4, pH 7.5, 10% (v/v) glycerol, 5 mm dithiothreitol, 2 mug aprotinin mL-1, 2 mug leupeptin mL-1, and 2 mm phenylmethanesulfonyl fluoride. Lysozyme (1 mg mL-1) and Triton X-100 (1%, v/v) were added to the suspension and incubated on ice with gentle shaking for 30 min before sonication. The sonicate was clarified by centrifugation at 15,000g for 30 min at 4C, and the supernatant was recovered. Recombinant proteins were then purified from the bacterial lysates by glutathione-Sepharose (Amersham Pharmacia Biotech) affinity chromatography. Glutathione-Sepharose 4B beads were added to the supernatant, and the mixture was incubated with gentle agitation for 60 min on ice. The Sepharose beads were then sedimented, and the pellets were carefully washed six times with the cell lysis buffer and resuspended in kinase assay buffer. SDS-PAGE (10%, w/v) analysis was used to evaluate the protein composition of each preparation. Gels were stained with Coomassie Brilliant Blue. Kinase Activity Assays | In vitro phosphorylation assays were performed as described previously with some modification. Peptide phosphorylation was measured as the incorporation of radioactivity from [gamma-32P]ATP (Perkin Elmer Life Sciences, Boston) into the peptide substrate. Forty microliters of the reaction mixture contained 20 mm Tris (pH 7.2), 2.5 mm MnCl2 or 5 mm MgCl2, 0.5 mm CaCl2, 10 mum ATP, 5 muCi [gamma-32P]ATP, 150 mum peptide substrate, and 2 mm dithiothreitol. Three peptide substrates used were p1 (LRRASLG; Kemptide, St. Louis), p2 (VRKRTLRRL; Sigma, St. Louis), and p3 (ALARAASAAALARRR, Research Genetics, Huntsville, AL). Enzymatic reactions were initiated by adding 5 muCi of [gamma-32P]ATP, and reaction mixtures were immediately incubated at 30C with gentle shaking. Reactions were terminated after 30 min by adding 1 muL of 0.5 m EDTA, and the GST fusion proteins bound to glutathione-Sepharose beads were pelleted. Fifteen microliters of the supernatant was applied onto P-81 phosphocellulose paper (Whatman, Clifton, NJ) for peptide phosphorylation analysis. The P-81 paper was then washed three times with 1% (v/v) phosphoric acid, and 32P incorporation into the peptide was quantified by phosphor imaging on a STORM 860 PhosphorImager (Molecular Dynamics, Sunnyvale, CA). For autophosphorylation assays, the remaining 25-muL reaction mixture was added with 5 muL of 6x Laemmli sample buffer and then separated by a 10% (w/v) SDS-PAGE gel and autoradiographed. For the analysis of divalent cation requirements, kinase assays were performed in the kinase assay buffers containing 0 to 20 mm of MnCl2 or MgCl2, using 150 mum p3 and 10 mum ATP. For ATP substrate kinetics analysis, 0 to 25 mum of ATP was used while keeping p3 constant (150 mum). Peptide substrate kinetic parameters were determined by varying the concentrations of the peptides (0 --250 mum) at a fixed ATP concentration (10 mum). Data Analysis | Initial rates were determined by measuring the amount of phosphorylated peptide formed in 30 min, because this time period produced adequate amounts of product for both enzymes and was within the linear portion of the reaction progress curve. The kinetic parameters were determined by nonlinear least squares analysis of the averaged initial velocity data fitting to the Henri-Michaelis-Menten equation (Eq. 1). In this equation, V0 is the measured initial velocity; Vmax is the maximum velocity; A is the concentration of ATP-metal; and Km is the apparent Km. The Kcat values were calculated by dividing Vmax by the total enzyme concentration. Three experiments were performed for all kinetic studies, and the average data were fit to the equation. Distribution of Materials | Upon request, all novel materials described in this publication will be made available in a timely manner for noncommercial research purposes. Backmatter: PMID- 12226506 TI - Plasmalemma Abscisic Acid Perception Leads to RAB18 Expression via Phospholipase D Activation in Arabidopsis Suspension Cells AB - Abscisic acid (ABA) plays a key role in the control of stomatal aperture by regulating ion channel activities and water exchanges across the plasma membrane of guard cells. Changes in cytoplasmic calcium content and activation of anion and outward-rectifying K+ channels are among the earliest cellular responses to ABA in guard cells. In Arabidopsis suspension cells, we have demonstrated that outer plasmalemma perception of ABA triggered similar early events. Furthermore, a Ca2+ influx and the activation of anion channels are part of the ABA-signaling pathway leading to the specific expression of RAB18. Here, we determine whether phospholipases are involved in ABA-induced RAB18 expression. Phospholipase C is not implicated in this ABA pathway. Using a transphosphatidylation reaction, we show that ABA plasmalemma perception results in a transient stimulation of phospholipase D (PLD) activity, which is necessary for RAB18 expression. Further experiments showed that PLD activation was unlikely to be regulated by heterotrimeric G proteins. We also observed that ABA-dependent stimulation of PLD was necessary for the activation of plasma anion current. However, when ABA activation of plasma anion channels was inhibited, the ABA-dependent activation of PLD was unchanged. Thus, we conclude that in Arabidopsis suspension cells, ABA stimulation of PLD acts upstream from anion channels in the transduction pathway leading to RAB18 expression. Keywords: Introduction : Abscisic acid (ABA) regulates seed maturation and germination, adaptation of plants to water shortage, cold, and high salinity . Several ABA transduction mutants have been isolated in Arabidopsis in which diverse loci affecting ABA response have been identified (for review, see ; ). Among the best characterized are the ABA-insensitive abi-1 and abi-2 mutations, which affect protein phosphatases ; the abi-3 , abi-4 , and abi-5 , which are mutated in transcription processes; and the ABA-hypersensitive era1 mutant, deleted for the beta-subunit of a farnesyl transferase which acts as a negative regulator . In guard cells, ABA transduction pathways have been extensively analyzed. Stomatal aperture is controlled by ABA through the activation of anion currents, which depolarizes the plasma membrane and promotes the activation of outward-rectifying K+ currents . In aleurone cells, ABA causes a decrease in cytoplasmic Ca2+ content, which is necessary for the inhibition of GA promotion of alpha-amylase activity . In Arabidopsis suspension cells, similar ion channel activation has been observed . However, the precise sequence of events triggered by ABA in these models remains unknown. Recent research has provided new data about the role of phospholipases in plant physiological responses , including the implication of phospholipase C (PLC) in ABA signaling. In Arabidopsis, ABA was shown to induce the expression of the Ca2+-dependent AtPLC1 . In the same species, it was demonstrated that an increase in PLC1 activity was necessary for the induction of the RD29a, KIN2, and RD22 ABA-responsive genes . In Commelina communis guard cells, have shown that , an inhibitor of phosphatidylinositol-PLC (PI-PLC) activity, abolished ABA-induced cytosolic [Ca2+] oscillations and stomatal closure. The involvement of phospholipase D (PLD) in ABA responses is better understood . PLD hydrolyzes constitutive phospholipids to phosphatidic acids (PAs) and free alcohols. A plasma membrane localization of PLD was observed in rice (Oryza sativa; ), tobacco (Nicotiana tabacum; ), and Arabidopsis . Three groups of PLD are distinguished according to the Ca2+ concentration required in in vitro activity tests . PLDs of the first group, named PLDalpha, are active at millimolar levels of Ca2+. The second group is PI-dependent PLD activated by micromolar Ca2+ concentration. The third group is Ca2+-independent PLD. Because of their higher affinity for primary alcohols than for water, Ca2+-dependent PLDs are able to transfer the phosphatidyl group to the alcohols, thus, synthesizing phosphatidyl alcohols. This reaction, associated with radioactive phosphate labeling of phospholipids, allows easy measurement of PLD activity . This method makes it possible to show that PLD activity mediates ABA responses. For example, in broad bean (Vicia faba) guard cells, ABA induced a transient stimulation of PLD . Moreover, PA application was able to mimic ABA action because, in broad bean, PA treatment inhibited inward K+ currents in guard cell protoplasts and regulated stomatal aperture . In barley (Hordeum vulgare) aleurone protoplasts, PA treatment causes a reduction in the cytoplasmic [Ca2+] and inhibits the GA activation of alpha-amylase activity . In animal models, heterotrimeric G proteins relay extracellular signals perceived from cell surface receptors to ion channels or phospholipases . Activation of a G protein-coupled receptor provokes the exchange of GDP for GTP in the guanine nucleotide-binding site of the alpha-subunit, which dissociates from the betagamma-dimer. Subsequent dephosphorylation of the GTP alpha-subunit allows the reassociation of the three G protein subunits. In plants, many biochemical and molecular studies argue in favor of the involvement of G proteins in the transduction of plant hormone signals . For instance, in aleurone protoplasts, the ABA stimulation of PLD is GTP dependent and is inhibited by pertussis toxin, thus, demonstrating an association with G proteins . Heterotrimeric and small G proteins were recently shown to be involved in ABA-triggered stomatal mechanism. However, in Arabidopsis, only one gene, GPA1, coding for an alpha-subunit of G protein, has been found . In Arabidopsis suspension cells, we have previously shown that ABA perception, at the outer side of the plasma membrane, was accompanied by the activation of outward-rectifying K+ channels and led to the expression of RAB18 . We demonstrated also that ABA perception was followed by a Ca2+ influx . Furthermore, the activation of anion channels and the inhibition of inward K+ channels occurred very quickly after ABA perception . Ca2+ influx and anion channel activation belong to the signaling pathway leading to the expression of RAB18, whereas K+ exchanges are independent of this pathway. The activation of specific Ca2+ channels by impermeant ABA prompted us to investigate Ca2+-dependent elements in the cascade of events starting from the outer-plasmalemma perception of ABA and leading to RAB18 expression in Arabidopsis suspension cells. In this report, we analyze the possible role of PLC and PLD. PLC activity was not involved in this signaling pathway, but we demonstrate that ABA stimulates a specific PLD activity that seems unlikely to be connected with heterotrimeric G proteins. This PLD stimulation precedes the activation of anion channels and is required for the expression of RAB18. RESULTS : PLD Activity Is Necessary for RAB18 Expression | In Arabidopsis suspension cells, RAB18 expression was elicited by both free and protein-conjugated ABA. When 1-butanol (1-BuOH), known as a PLD substrate for transphosphatidylation, was co-incubated for 3 h with 10-5m ABA, a dose-dependent inhibition of RAB18 expression was observed (Fig. A). With 0.5% (v/v) 1-BuOH, a total inhibition occurred, whereas 2-BuOH and 3-BuOH, which are not recognized by PLD, did not affect RAB18 expression (Fig. B). Figure 1 | ABA-induced RAB18 gene expression is inhibited by 1-BuOH in Arabidopsis suspension cells. ABA-induced RAB18 gene expression is inhibited by 1-BuOH in Arabidopsis suspension cells. Northern-blot analysis of total RNA (10 mug) extracted from cells. A, Cells were incubated for 3 h with 10-5m ABA and 0.05% to 0.5% (v/v) 1-BuOH. B, Cells were incubated for 3 h with ABA-BSA (10-5m equivalent ABA) and 0.5% (v/v) 1-BuOH, 2-BuOH, or 3-BuOH. Ethidium bromide staining of rRNAs is shown as control. Because the PA produced by PLD activity was required for ABA-induced RAB18 expression, we measured how in vivo PLD activity was modulated by ABA. After an 18-h 33P pulse, 0.5% (v/v) 1-BuOH and 10-5m ABA were simultaneously added to the suspension cells. Lipids were extracted and phosphatidyl-BuOH (P-BuOH) synthesized was separated by thin-layer chromatography (Fig. A). Direct measurement of in vivo PLD activity was expressed as the relative amount of P-BuOH compared with structural phospholipids. Figure B shows that accumulation of P-BuOH is almost linear for 1 h in control cells, whereas it presents a bimodal pattern in ABA-treated cells: 10 min after ABA treatment, PLD increased 20%, reached about 30% at 30 min, and then remained almost unchanged (Fig. B). Therefore, ABA induced a transient increase in PLD activity. Figure C illustrates relative in vivo PLD activity measured from cells treated for 30 min with free or conjugated ABA. ABA-bovine serum albumin (ABA-BSA; 10-5m equivalent ABA) stimulated PLD activity to the same extent as free ABA. Figure 2 | In vivo PLD activity is stimulated by ABA in Arabidopsis suspension cells. In vivo PLD activity is stimulated by ABA in Arabidopsis suspension cells. A, Thin-layer chromatography separation of phospholipids after treatment with ABA-BSA (10-5m equivalent ABA) and 0.5% (v/v) 1-BuOH after an 18-h 33P pulse. B, Time course of in vivo ABA-stimulated PLD activity. , Control; circle, ABA (10-5 M). Bars, sd, n = 3. C, Free (10-5m) or conjugated (10-5m equivalent ABA) ABA-stimulated PLD activity. PLD activity was measured after 30 min of treatment. SPL, Structural phospholipids. Bars, sd, n = 7. PLC Activity Is Not Involved in ABA-Induced RAB18 Expression | Activation of PLD is necessary for ABA-induced RAB18 expression, thus, indicating that PA synthesis is a step of this signaling pathway. Because PA can also be produced by PLC associated with diacylglycerol kinase (DAG kinase) activity, we investigated the involvement of PLC in ABA-specific RAB18 expression. Short incubation of cells with 33P was done to label the ATP pool, thus, making PA labeling only through DAG-kinase activity. Neither free nor conjugated ABA induced significant increase in the measured PA level (Fig. A). Figure B shows that RAB18 expression induced by ABA was not modified by neomycin (50 mum) or by the specific PI-PLC inhibitor (10 mum) and its biologically inactive analog U73343 (10 mum). Furthermore, measurement of inositol 1,4,5-triphosphate (IP3) produced by PI-PLC showed no variation between control and 30-min ABA-treated cells (4.6 pmol IP3 mL-1, sd = 0.14, n = 4; data not shown). These results show that PLC is not implicated in ABA-induced RAB18 expression in Arabidopsis suspension cells. Figure 3 | PLC is not involved in ABA-induced RAB18 expression in Arabidopsis suspension cells. PLC is not involved in ABA-induced RAB18 expression in Arabidopsis suspension cells. A, Relative amount of PA extracted from cells preincubated for 2 h with 33P and treated with ABA (10-5m) or ABA-BSA (10-5m equivalent ABA) for 30 min. Bars, sd, n = 3. B, Northern-blot analysis of total RNA (10 mug) extracted from cells treated for 3 h with ABA-BSA (10-5m equivalent ABA) and neomycin (50 mum), (10 mum), or U73343 (10 mum). Ethidium bromide staining of rRNAs is shown as control. G Proteins Do Not Seem to Be Involved in ABA-Triggered PLD Activation | PLD has been shown to be regulated by heterotrimeric G proteins . Thus, we tested mastoparan, a specific heterotrimeric G protein activator, on RAB18 expression. Mastoparan did not mimic ABA, and when added to ABA, it did not potentialize ABA activation of RAB18 expression . These data are in accordance with those displayed in Figure B. 2-BuOH, a secondary alcohol that enhances G protein activity , did not modify ABA-specific expression of RAB18. From these data, we conclude that heterotrimeric G proteins do not seem to be involved in ABA stimulation of PLD activity required for RAB18 expression. Figure 4 | Heterotrimeric G proteins do not seem to be involved in RAB18 expression in Arabidopsis suspension cells. Heterotrimeric G proteins do not seem to be involved in RAB18 expression in Arabidopsis suspension cells. Northern-blot analysis of total RNA (10 mug) from cells incubated simultaneously with ABA-BSA (10-5m equivalent ABA) and mastoparan (1 and 5 mum) for 3 h. Ethidium bromide staining of rRNAs is shown as control. PLD Precedes Anion Channel in ABA-Signaling Cascade | In a previous study, we showed that ABA outer-plasmalemma perception triggers early activation of plasmalemma anion currents, which is required for RAB18 expression . We tested ABA anion current activation when PLD was inhibited and vice versa to better understand the link between PLD and anion current activation. After pulses from -200 to -40 mV, most of the cells (70%; n = 91) exhibited recordings typical of S-type anion currents, which are enhanced by ABA (Fig. A). Successive application of 0.1% (v/v) 1-BuOH reduced ABA-enhanced currents by 57% +- 15% (mean value +- sd; n = 4) for -200 mV clamping (Fig. A). This reduction was observed in six of the nine cells exhibiting an anion current. By contrast, when cells were first treated with 1-BuOH, ABA could not activate anion currents (Fig. B). When ABA application was followed with 3-BuOH treatment, ABA activation of anion channels was not changed, but further addition of 1-BuOH resulted in a reduction (30% +- 4%) of anion currents (Fig. C). These results indicate that ABA activation of anion currents depends on PA formation by PLD activation. Figure 5 | Shunt of PLD with 1-BuOH suppresses ABA activation of whole cell anion currents in Arabidopsis suspension cells. Shunt of PLD with 1-BuOH suppresses ABA activation of whole cell anion currents in Arabidopsis suspension cells. Whole cell currents were activated by a depolarizing prepulse (+100 mV for 4.5 s), not shown. Then, hyperpolarizing pulses ranging from -200 to -40 mV, in 20-mV steps (A) or from -200 to -120 mV, in 40-mV steps (B and C) were applied for 9.5 s. Representative traces are shown. A, Current recordings from intact cells before (control) and 15 min after application of 10 mum ABA and 6 min after addition of 0.1% (v/v) 1-BuOH. The diameter of the cell was 45 mum. Right, Histogram represents the mean anion current at -200 mV for at least four replicates (bars, sd). B, Currents were recorded 5 min after successive additions, at 5-min intervals, of ABA, 3-BuOH, and 1-BuOH (0.1%, v/v). The diameter of the cell was 42 mum. Right, Histogram represents the mean anion current at -200 mV for at least three replicates (bars, sd). C, Currents were recorded from intact cells (control) and after successive additions of 0.1% (v/v) 1-BuOH (2 min later) and then 10 mum ABA (4 min later). The diameter of the cell was 48 mum. Right, Histogram represents the mean anion current at -200 mV for at least four replicates (bars, sd). The anion channel blocker 5-nitro-2-(3-phenylpropylamino) benzoic acid (NPPB) inhibits ABA activation of anion currents and RAB18 expression in Arabidopsis suspension cells . Addition of NPPB (25 mum) did not significantly affect PLD activity: Only a slight increase was observed in control (110% versus 100%) and, when ABA and NPPB were simultaneously added, the stimulation of PLD activity by free or conjugated ABA was still recorded . Taken together, the data presented demonstrate that the ABA-induced stimulation of PLD activity is upstream the activation of plasmalemma anion currents (Figs. and ). Figure 6 | Inhibition of plasma membrane anion currents does not alter ABA activation of in vivo PLD activity in Arabidopsis suspension cells. Inhibition of plasma membrane anion currents does not alter ABA activation of in vivo PLD activity in Arabidopsis suspension cells. Cells were treated for 30 min with ABA (10-5m) or ABA-BSA (10-5m equivalent ABA) and 25 mum NPPB. Bars, sd, n = 4. DISCUSSION : PLD Activity Is Involved in ABA-Triggered RAB18 Expression in Arabidopsis Suspension Cells | In the present study, we established that treatment of Arabidopsis suspension cells with free or protein-conjugated impermeant ABA triggered a stimulation of PLD activity. Therefore, PLD is an element of the signaling cascade associated with ABA perception at the plasmalemma level. The role of PLD and its PA product in the ABA-induced RAB18 expression was demonstrated because addition of 1-BuOH led to the accumulation of P-BuOH, at the expense of PA formation, and inhibited RAB18 expression (Figs. and ). Such an involvement of PLD has already been described in ABA transduction pathways. For example, in barley aleurone cells, PLD activity mediates ABA inhibition of GA-induced alpha-amylase production . Using transient gene expression in rice protoplasts, PLD was demonstrated to regulate expression of the ABA-inducible wheat (Triticum aestivum) gene Em . In Arabidopsis suspension cells treated with 10-5m ABA, the range of in vivo PLD activation observed (i.e. 30% --40%; Fig. ) is comparable with the 50% increase in PLD activity measured in vitro in the microsomal fraction of barley aleurone protoplasts treated by ABA at the same concentration . In broad bean guard cell protoplasts, a transient increase in PLD activity, within the same range, was obtained after a 30-min application of 10-5m ABA . The activation of PLD by ABA in Arabidopsis suspension cells follows a similar pattern . Nevertheless, our data are in contrast with the observation reported in Arabidopsis T87 culture cells by . They did not detect any PLD activation by exogenous ABA. The reason for this discrepancy remains unknown. Specificity of the PLD Stimulated by ABA | PLDs represent a multiple gene family that is divided into three groups according to Ca2+ requirement . Ca2+-independent PLDs are unable to achieve the transphosphatidylation reaction and cannot belong to the transduction cascade analyzed here because 1-BuOH addition blocks RAB18 expression. PLDbeta and PLDgamma are Ca2+-dependent PLDs that are inhibited by neomycin . We have demonstrated that neomycin did not interrupt the signaling cascade triggered by ABA leading to RAB18 expression (Fig. A). Therefore, this result suggests that the ABA-activated PLD belongs to PLDalpha group. Our observation corroborates other reports in Arabidopsis. In fact, the importance of PLDalpha has been confirmed in Arabidopsis leaves , and the antisense suppression of PLDalpha retarded ABA-promotion of senescence in detached Arabidopsis leaves . ABA-treated cells incubated with 1-BuOH accumulated 30% to 40% more P-BuOH than untreated ABA cells , which is the reflection of specific PAs over-produced by the ABA-stimulated PLD activity. Thus, these PAs are the putative second messengers in the ABA transduction pathway understudied. PAs are considered as second messengers in physiological responses to stresses and pathogen attacks or in symbiosis (for review, see ). In broad bean guard cells, PA was shown to exert similar effects as ABA on the regulation of stomatal aperture and on the inhibition of the activity of K+ inward channels . In barley aleurone protoplasts and layers, similar inhibition of GA-activated alpha-amylase activity could be obtained with the addition of either ABA or PAs . In the same study, the authors highlighted the importance of the PA molecular species for a given biological effect and demonstrated that arachidonoyl-stearoyl PA was the most efficient antagonizer of GA action. In castor bean (Ricinus communis), it was shown that wounding activated a PLD and resulted in accumulation of polyunsaturated fatty acids, DAG, and PAs . The analysis of the acyl composition of PAs in mechanically wounded Arabidopsis leaves has shown that a specific isoform of PLD, using phosphatidylcholine (PC) as substrate, was activated . The importance of acyl groups of phospholipids acting as putative second messengers is well established . Therefore, isolation and characterization of endogenous specific PAs produced by ABA-treated Arabidopsis suspension cells should be undertaken to confirm the role of PAs as the second messenger. PLC Activity Does Not Contribute to ABA-Triggered RAB18 Expression in Arabidopsis Suspension Cells | In Arabidopsis suspension cells, we previously demonstrated that an influx of extracellular Ca2+ through specific plasmalemma Ca2+ channels was required for ABA-induced RAB18 expression . Nevertheless, changes in cytoplasmic [Ca2+] may also be due to the activation of IP3 or cADPR-dependent intracellular Ca2+ channels . Hence, we examined the involvement of PI-PLC, responsible for IP3 production, in ABA signaling in Arabidopsis suspension cells. ABA-induced RAB18 expression was not modified by treatment of the cells with the PI-PLC inhibitors neomycin and (Fig. B). Furthermore, no changes in the phosphorylation rate of PA extracted from cells preincubated for 2 h with 33P and treated with ABA was observed (Fig. A). The system PLC-DAG kinase is, thus, not involved in the ABA-triggered RAB18 expression pathway. As already reported by in similar Arabidopsis suspension cells, we were unable to detect any IP3 formation after ABA treatment (data not shown), indicating that RAB18 expression is not dependent on inner IP3-dependent Ca2+ channels. Taken together, these results show that a common regulation, independent of PLC but dependent on PLD, can be envisaged for RD20 and RAB18, two ABA-inducible genes. However, in Arabidopsis seedlings, it was demonstrated, with transgenic lines expressing an AtPLC1 antisense, that PLC activity was involved for the induction of the RD29a, KIN2, and RD22 ABA-inducible genes. In contrast, the expression of COR47 was unaffected in the same AtPLC1 antisense line . Therefore, the control of the expression of ABA-inducible genes may act through different pathways. It should be noted that in other ABA-dependent responses that involve Ca2+, PI-PLC is not always implicated. For example, in broad bean, a total inhibition of ABA-induced stomatal closure requires both and nicotinamide, an inhibitor of cADPR synthesis whereas, in Commelina communis, treatment with is sufficient to inhibit ABA-induced cytosolic [Ca2+] oscillations and stomatal closure . Thus, it appears that different Ca2+ channels may act in parallel or cooperate in ABA-signaling pathways according to the cell or the organ considered. There Is No Evidence That PLD Involved in ABA-Triggered RAB18 Expression Is Regulated by Heterotrimeric G Proteins | Some reports suggest that G proteins act upstream from PLD in plants . In Chlamydomonas eugametos, deflagellation is associated with PLD activation and can be obtained with G protein activators . The in vitro activity of a PLD recombinant protein, obtained from a cDNA clone isolated from tobacco, is inhibited by a recombinant G protein alpha-subunit expressed in Escherichia coli . In barley aleurone microsomal fraction enriched in plasma membrane, the addition of GDPbetaS rendered PLD activity insensitive to ABA, whereas GTPgammaS stimulated PLD in the absence of ABA . Our data show that mastoparan, which mimics the motif of 7TMS receptors that activates G proteins , was unable to reproduce the effect of ABA on RAB18 expression . Furthermore, we have shown that 2-BuOH did not enhance the effect of ABA on RAB18 expression (Fig. B) Therefore, it is highly improbable that G proteins are involved in the regulation of PLD activated by ABA in this transduction pathway. PLD Activity Precedes Activation of Plasmalemma Anion Currents | The activation of anion currents by ABA was shown to be required for RAB18 expression . Here, we observed that the inhibition of anion currents by NPPB did not reduce the in vivo enhancement of PLD activity by ABA . In contrast, the electrophysiological data clearly showed that addition of 1-BuOH but not of 3-BuOH suppressed the activation of anion currents by ABA . ABA-specific PLD activation must, therefore, occur upstream from anion currents in the signaling pathway leading to RAB18 expression. In conclusion, we demonstrated the role of a specific PLD in the signaling cascade, which begins with the plasmalemma perception of ABA and results in the expression of the RAB18 gene. PLD activation is a very early event in the ABA transduction pathway because it precedes the stimulation of anion currents. The PLD isoform involved remains to be analyzed. Further investigations need to characterize the PAs produced as second messengers and the targets modulated by PA. MATERIALS AND METHODS : Plant Material | Arabidopsis ecotype Columbia cells were obtained by . They were cultured at 24C, under continuous white light (40 mumol m-2 s-1) with an orbital agitation at 130 rpm, in 500-mL Erlenmeyer flasks containing 200 mL of culture medium. A 25-mL aliquot of cell suspension was transferred to a fresh medium every week. All the experiments were conducted on 4-d-old cells. The pH of the culture medium was 6.8. The viability of the cells during the experimental treatment was systematically checked with Trypan blue tests (not shown). RAB18-Responsive Test and Northern-Blot Analysis | A 5-mL suspension was incubated for 3 h under the conditions of culture. ABA-BSA purified conjugate (10-5m equivalent ABA) was added in 50 mm Na2SO4, 50 mm phosphate buffer, pH 6.8. Northern-blot analyses were performed according to the protocol previously described . The 684-bp RAB18 cDNA probe used (GenBank accession no. ) contained the coding sequence (with the exception of the first 100 bp of 5' sequence after the ATG codon) and the 3' noncoding sequence ending with the polyadenylation site of the gene . 18S RNA gene probe was used as control of constitutive gene expression (data not shown). In Vivo PLD Activity | PLD activity was measured as described by . Arabidopsis suspension cells were first incubated in [33P]H3PO4 (53 MBq L-1) for 18 h to label all phospholipids. 1-BuOH was then added to probe PLD activity by transphosphatidylation leading to P-BuOH production. Phospholipids were extracted in 2:1 (v/v) MeOH:CHCl3 solvent mixture (15 mL for 6 mL of suspension cells), and separated on silica gel 60 thin-layer chromatography plates (Merck, Rahway, NJ) with the organic upper phase of 12:2:3:10 (v/v) ethyl acetate:iso-octane:formic acid:water according to . Radioactivity incorporated in structural phospholipids, PA, and P-BuOH spots was measured on a PhosphorImager (Storm, Molecular Dynamics, Sunnyvale, CA). In Vivo PI-PLC Activity | PI-PLC activity was assayed with the Biotrak assay kit (TRK1000, Amersham, Buckinghamshire, UK). Six milliliters of suspension cells was fixed with 1.2 mL of perchloric acid and then centrifuged at 15,000g for 15 min, at 4C. The supernatant was adjusted to pH 7.5 with ice-cold 60 mm HEPES in 1.5 m KOH. The KClO4 precipitated was removed, and IP3 content was determined in the supernatant according to the manufacturer's specifications. Electrophysiology | Cells were immobilized by means of a microfunnel (approximately 30 --80 mum o.d.). Impalements of the cells were carried out with a piezoelectric micromanipulator (PCS-5000, Burleigh Institute, New York) in a continuous-flow chamber (500 muL) made of perpex. ABA and BuOH were diluted in the bathing medium introduced via a polyethylene catheter. The cells were balanced for 24 h before electrophysiological experiments in fresh culture medium (20 mm KNO3, 0.9 mm CaCl2, and 0.45 mm MgSO4, pH 6.8). Microelectrodes were made from borosilicate capillary glass (Clark GC 150F, Clark Electromedical, Pangbourne Reading, UK) pulled on a vertical puller (Narischige PEII, Japan). Their tips were of 0.5 mum diameter, they were filled with 600 mm KCl, and had electrical resistances from 50 to 80 MOmega with the buffer solution. Voltage-clamp measurements of whole cell currents from intact cells were carried out at room temperature (20C --22C) using the technique of the discontinuous single voltage-clamp microelectrode . A specific software (pCLAMP 8) drives the voltage clamp amplifier (Axoclamp 2A, Axon Instruments, Foster City, CA). Voltage and current were digitalized with a PC computer fitted with an acquisition board (Labmaster TL 1, Scientific Solution, Solon, OH). In whole cell current measurements, the membrane potential was held at -40 mV. Anion currents were activated by a depolarizing prepulse (+100 mV for 4.5 s), and then hyperpolarizing pulses ranging from -200 to -120 or to -40 mV for 9.5 s (in 40- and 20-mV steps, respectively) were applied. We systematically checked that cells were correctly clamped by comparing the protocol voltage values with those really imposed. Backmatter: PMID- 12226507 TI - Characterization of a Novel Lipoxygenase-Independent Senescence Mechanism in Alstroemeria peruviana Floral Tissue AB - The role of lipoxygenase (lox) in senescence of Alstroemeria peruviana flowers was investigated using a combination of in vitro assays and chemical profiling of the lipid oxidation products generated. Phospholipids and galactolipids were extensively degraded during senescence in both sepals and petals and the ratio of saturated/unsaturated fatty acids increased. Lox protein levels and enzymatic activity declined markedly after flower opening. Stereochemical analysis of lox products showed that 13-lox was the major activity present in both floral tissues and high levels of 13-keto fatty acids were also synthesized. Lipid hydroperoxides accumulated in sepals, but not in petals, and sepals also had a higher chlorophyll to carotenoid ratio that favors photooxidation of lipids. Loss of membrane semipermeability was coincident for both tissue types and was chronologically separated from lox activity that had declined by over 80% at the onset of electrolyte leakage. Thus, loss of membrane function was not related to lox activity or accumulation of lipid hydroperoxides per se and differs in these respects from other ethylene-insensitive floral tissues representing a novel pattern of flower senescence. Keywords: Introduction : Senescence is a complex, highly regulated process that involves a decline in photosynthesis; dismantling of chloroplasts; degradation of macromolecules such as proteins, nucleic acids, and lipids; loss of chlorophyll; and mobilization of nutrients to developing parts of the plant . In leaves, this process can be reversed; however, in floral tissues it cannot, indicating that a tightly controlled program for cell death exists . The termination of a flower involves at least two, sometimes overlapping, mechanisms. In one, the perianth abscises before the majority of its cells initiate a cell death program . Abscission may occur before or during the mobilization of food reserves to other parts of the plant. Alternatively, the petals may be more persistent, so that cell deterioration and food remobilization occur while the petals are still part of the flower. The overall pattern of floral senescence varies widely between plant genera; therefore, a number of senescence parameters have been used to group plants into somewhat arbitrary categories. One distinction that is often made is the relative response of flowers to ethylene, resulting in the recognition of "ethylene-sensitive" (e.g. Orchidaceae, Campanulaceae, and Cruciferae) and "ethylene-insensitive" (e.g. Iridaceae, Liliaceae, and Amaryllidaceae; ; ) systems. Flowers of Alstroemeria peruviana (Liliaceae) have been reported as relatively "insensitive" to this gaseous hormone . The maintenance of cellular integrity and subcellular compartmentation is integral to cell function. However, during senescence of both ethylene-sensitive and -insensitive flowers, marked changes occur in the biochemical and biophysical properties of the cell membranes. These result from losses of membrane phospholipids, increases in neutral lipids, increases in sterol to phospholipid ratio, and increases in the saturation:unsaturation index of fatty acids . Membrane polyunsaturated fatty acids are prone to oxidation either by enzymatic means (lipoxygenase [lox]) or through autoxidative events (nonenzyme catalyzed). In a number of floral tissues, such as carnation (Dianthus caryophyllus; ), daylily (Hemerocallis hybrid; ), and rose (Rosa hybrid; ), lox activity increases before the onset of electrolyte leakage (a marker of loss of membrane semipermeability). Increase in lipid peroxidation, usually estimated as thiobarbituric acid reactive substances (TBARS), accompanies the increase in lox activity and the products of peroxidation are considered to perturb membrane function, in part, at least, by causing increased membrane rigidification . Examples exist of both ethylene-sensitive lox-mediated peroxidation (e.g. carnation and rose) and ethylene-insensitive lox-mediated peroxidation, e.g. daylily and Gladiolus hybrid . However, daylily flowers have a particularly short life span of around 24 h ; therefore these flowers might be anomalous with respect to other longer lived ethylene-insensitive floral systems. For this purpose, we chose A. peruviana, which is relatively insensitive to ethylene and has a life span of approximately 12 d . To distinguish between enzymatic oxidation and chemical autoxidation of lipids, we analyzed the regio- and stereochemical addition of molecular oxygen to the acyl chain . This level of chemical analysis offers a more complete picture of the nature of lipid peroxidation in plant tissues than can be obtained from TBARS measurements alone. In this study, we characterize the enzymatic activity and protein levels of lox and chemically profile the oxidation products generated in floral tissues during senescence. The data presented illustrate that, in contrast to other ethylene-insensitive systems such as daylily and Gladiolus sp., loss of membrane integrity in A. peruviana is not related to either lox activity or the accumulation of lipid hydroperoxides (LHPOs) and, thus, represents a novel category of floral senescence. RESULTS : Changes in Lipid Composition during Senescence | To determine the catabolic fate of lipids after flower opening and subsequent senescence, total lipid extracts were prepared from seven characterized stages in A. peruviana, designated numerically from stage 0 (S0) to stage 6 (S6) and representing 12 d in time . In brief, S0 and S1 are the stages of floral development when the buds are opening, and by S2, the flowers were fully open with the sepals reflexed. At S3, the top three anthers anthesed and 2 d later at S4, the petals showed initial signs of in-rolling and discoloration (visible senescence), while the bottom three anthers had also anthesed. S5 was defined by the separation of the stigmatic lobes and further signs of petal discoloration and in-rolling. Abscission of the perianth occurred at S6. Analysis of the total lipid content shows that a marked decline in lipid occurs in both sepals and petals after S1, representing a loss of 56% and 68% of total lipid, respectively (Fig. A). Of the major fatty acid constituents degraded, linoleic acid (C18:2 Deltacis 9,12) showed the largest decline, losing almost 40 mumol g dry weight-1 from S1 to S6 (equivalent to 80%). Decreases in palmitic acid (C16:0; 10 mumol g dry weight-1) and 18:3 (linolenic acid, C18:3 all cis Delta9,12,15) of 14 mumol g dry weight-1 were also observed (Fig. B). Small increases in the mass of lauric acid (C12:0) and myristic acid (C14:0) were detected as senescence progressed (from 0.4 +- 0.2 mumol g dry weight-1 at S0 to 2.5 +- 0.3 mumol g dry weight-1 at S6 for both fatty acids). Figure 1 | A, Total fatty acid content of sepals and petals during senescence. A, Total fatty acid content of sepals and petals during senescence. B, Changes in major fatty acid components of petal tissue during senescence. Fatty acids were extracted in a chloroform:methanol-based solvent system, separated by gas chromatography as their methyl esters, and quantified using heptadecanoic acid as an internal standard. Data points represent the mean of n = 3 +- sd. Similar overall changes in the fatty acid constituents of sepals were observed (data not shown). The complex lipid profile was almost identical for both sepals and petals, and the same changes were observed during senescence for both floral organs. For this reason, only the data for petal tissues were shown . In stages S0 through S4, the phospholipids were the major components representing almost 60% of the total lipids, of which phosphatidylcholine was 33%, with phosphatidylethanolamine being the next most abundant lipid over this time period (17%). The chloroplast galactolipids, monogalactosyldiacylglycerol and digalactosyldiacylglycerol, were present in similar amounts and jointly constituted around 22% of the tissues total lipids at S0. Phosphatidylinositol, phosphatidic acid, and phosphatidyl-Ser were also detected in low amounts (<5% of the total lipids combined; data not shown). Neutral lipids were present in young S0 tissues of both floral organs with triacylglycerol (TAG; 4%), diacylglycerol (3%), unesterified fatty acids (UFAs; 4%), and sterol esters (4%) being detected. During the course of development and senescence the phospholipids and galactolipids were extensively degraded and by S6, 87% of phosphatidylcholine, 83% of phosphatidylethanolamine, 95% of monogalactosyldiaglycerol, and 89% of digalactosyldiacylglycerol had been utilized . The levels of neutral lipids, in contrast, remained similar throughout senescence, although some increase in the unesterified fatty acid (UFA) level was detected at S6 . Table I | Changes in the major lipids during petal senescence Lox Activity | Petal lox activity (includes all isoforms of the enzyme present ---at least two activities; see stereochemical profiling of products below) had a pH optimum between 5.5 and 6.0 (Fig. A) and was inhibited, in a dose-dependent manner, by the lox inhibitor esculetin (50% inhibition with 100 mum; Fig. B). Lox(es) showed activity to a range of unsaturated fatty acids, although a preference for 18:2 over 18:3 was observed. The enzyme(s) also readily used gamma-linolenic acid (C18:3 all cis Delta6,9,12) and arachidonic acid (C20:4 all cis Delta5,8,11,14), which are not endogenous fatty acids in the floral tissues (Fig. C). Lox activity declined by 50% from S0 to S2 and continued to decline to S6 (Fig. A). This decline in activity was paralleled by a decrease in the level of lox protein(s) detected in western blots (Fig. A, insert). Figure 2 | Characterization of A. peruviana Characterization of A. peruviana lox activity. A, pH optima. B, Inhibition of activity by esculetin. C, Fatty acid substrate specificity. Lox activity was monitored in 13,000-g supernatants of petal extract by measuring the increase at A234 after the formation of conjugated dienes from fatty acid substrate. Both the pH optima and esculetin inhibition analyses were performed using 18:2 as a substrate. All data points represent n = 9 +- sd. Figure 3 | a, Total lox activity in petal tissue throughout senescence. a, Total lox activity in petal tissue throughout senescence. Lox activity was determined in 13,000-g supernatants by following conjugated diene formation at 234 nm, using 18:2 as a substrate. Data points represent n = 9 +- sd. The insert shows a western blot probed with antibodies raised against a recombinant cucumber (Cucumis sativus) lipid body lox. Results are representative of analyses repeated twice. b, Electrolyte leakage throughout senescence expressed as the percent of total leakage. Total leakage was determined by measuring the conductivity of frozen sepal and petal discs of appropriate age. Data points represent the mean of n = 6 +- sd. Electrolyte Leakage | We postulated that if lox activity were responsible for the initiation of senescence through mechanisms involving free radical damage, then the activity of this enzyme would be correlated to electrolyte leakage. This parameter is a commonly used method for determining the extent of tissue damage indicative of the loss of semipermeability of membranes and/or loss of cellular integrity. Both sepals and petals showed little electrolyte leakage up to S4, although after this point leakage increased in both tissues (Fig. B). Thus, lox activity and electrolyte leakage are chronologically separated, indicating that lox is not a primary initiator of cellular damage in A. peruviana. Stereochemical Profile of Lox-Derived Oxygenated Fatty Acids | Lipoxygenases readily utilize UFA as substrates for oxygenation . To determine the positional specificity of the A. peruviana lox, we froze floral tissues to elevate the UFA pool (released from complex lipids by acyl hydrolases), thereby providing substrate for the endogenous lox enzyme(s) allowing accurate determination of the positional specificity of oxygenation. After freezing, the content of UFA rose from 2.2% to 10.4% in S0 tissues, from 1.4% to 6.5% in S2, and from 2.7% to 6.4% in S3 tissues. At S6, however, substantial levels of UFA are already present in the tissues (11.5%) and freezing did not result in any further increase in their level. Straight phase (SP)-HPLC analyses of lox-derived oxygenated fatty acids (analyzed as hydroxy polyenoic fatty acids) were undertaken and similar results were obtained for both sepal (data not shown) and petal tissue . In the esterified lipid hydroxide fraction, the predominant positional isomers were the 13-oxy derivatives of C18:3 [all cis Delta 9,12,15, 13(S)-hydroperoxy-9(Z), 11(E), 15(Z)-octadecatrienoic acid (13-HOTE)] and C18:2 [cis Delta 9,12, 13(S)-hydroperoxy-9(Z), 11(E)-octadecadienoic acid (13-HODE)] and the enantiomeric form was >94% in the S form, indicating an enzymatic origin (Fig. A). Small amounts of 16-HOTE, 9-HOTE, and 9-HODE were also detected and were racemic, indicating an autoxidative origin. In the UFA hydroxide pool (Fig. B), again the predominant positional isomer at S0, S2 and S4 were 13-HOTE (84% +- 2%) and 13-HODE (74% +- 4%), whereas their corresponding 9-oxy derivatives constituted 16% +- 2% (9-HOTE) and 26% +- 4% (9-HODE), respectively. The 9-oxy derivatives were predominantly of the S type (86% --97%), indicating an enzymatic origin. However, the major lox activity in A. peruviana floral tissues is of the 13 type. At S6, the ratio of 13- to 9-oxy derivatives was similar, although the level of oxygenated fatty acids was significantly lower than at the earlier time points analyzed. Interestingly, relatively high levels of keto fatty acids were detected in both sepals and petals and the predominant isomer detected was 13-keto-9(Z), 11(E)-octadecadienoic acid (Fig. C). Figure 4 | Analysis of oxygenated fatty acid products in petal tissue. Analysis of oxygenated fatty acid products in petal tissue. Products were analyzed by HPLC and separated into three oxylipin classes: A, hydro(pero)xy fatty acids esterified to complex lipids; B, unesterified hydro(pero)xy fatty acids; and C, keto fatty acids. In A and B, the R:S ratio for each stereoisomer ratio is shown above that isomer. Endogenous Oxidative Damage in Freshly Harvested Floral Tissues | In the above experiment, we were interested in elevating the level of UFA in the tissues to determine the stereochemical specificity of lox. However, the determination of the LHPO content of freshly harvested tissues is also necessary to determine whether the levels of these compounds change significantly throughout senescence. Recently, we have developed a rapid and sensitive spectrophotometric method for the detection of LHPOs in plants based upon the hydroperoxide-mediated oxidation of ferrous to ferric ions under acidic conditions and used it to determine LHPO levels in A. peruviana floral tissues. In sepals, the level of LHPOs gradually accumulated with time to a maximum at S5 (Fig. A). Conversely, in petal tissues, the level of LHPOs declined by over 60% from S0 to S6. At times from S2 onwards, the LHPO level in the sepals was between 3 to 4 times higher than in petals. The fatty acid content of both sepals and petals markedly declines throughout senescence (see Fig. A). Expressing the level of oxidized lipid relative to total lipid we observe that the proportion of oxidized lipid markedly increased in the sepals from basal levels of 1% to 3% of total fatty acids, whereas in petals, the level is maintained at <1% at all times (Fig. C). Figure 5 | Assessment of lipid peroxidation in senescing floral tissues. Assessment of lipid peroxidation in senescing floral tissues. A, LHPO content determined by the FOX assay. B, Malondialdehyde (MDA) equivalents as detected by TBARS. C, Percent oxidized fatty acids in tissues, determined from A and data presented in Figure A (n = 9 +- sd). The TBARS assay that estimates the amount of MDA, a secondary end product of polyunsaturated fatty acid oxidation, has often been used as an index of general lipid peroxidation . Determination of TBARS in the tissues again showed that lipid peroxidation is typically 2 to 3 times higher in sepals than petals (Fig. B). In sepals, the TBARS level peaked at S2 and then declined throughout senescence, whereas in petals, the levels remained almost constant between S1 and S5. Clearly, distinctly different patterns of LHPOs are accumulated in sepals and petals; to address why this might be so, we investigated the role that antioxidants might play in this process. Antioxidant Status and Pigment Content of Floral Tissues | Plants have evolved a complex battery of defensive mechanisms to deal with the consequences of oxidative stress that can result in lipid membrane (as well as other macromolecular) perturbation . To avoid making prejudgments regarding the possible nature of the antioxidant contributors to chemoprotection from oxidative stress, we first screened the tissues for their total aqueous soluble antioxidants (TASA). The change in tissue TASA after harvest was monitored using the 2,2-azino-di [3-ethylbe-nzthiazoline sulfonate] (ABTS+) assay . We predicted that the levels of antioxidants would decline in sepals and thereby account for the increase in oxidized fatty acids observed. No clear difference in the level of antioxidants between the two tissue types was evident in stages S0 to S3. However, from S4 to S6, the level of antioxidants actually increased in sepals, yet remained relatively constant in petals (Fig. A). Figure 6 | Antioxidant status and pigment content of senescing floral tissues. Antioxidant status and pigment content of senescing floral tissues. A, TASA determined by the ABTS assay. B, Anthocyanin content. C, Carotenoid content. D, Chlorophyll content throughout senescence. For the TASA measurement n = 4 +- sd, whereas for all pigment measurements, n = 9 +- sd. During bud opening and expansion, the level of anthocyanin in petals was more than double that in the sepals (Fig. B). However, between S4 and S6, the levels were similar and declined in both tissues. Thus, during sepal senescence, the increase in TASA from S4 to S6 (Fig. A) could not be accounted for by anthocyanin, which actually decreased over this time period, suggesting that components other than these are responsible for the elevation in the antioxidant levels. Carotenoids are also considered to be effective free radical scavengers and to play a chemoprotective role against oxidative damage . These components are not soluble in ethanol and, therefore, do not contribute to the antioxidant status in that preparation. Spectrophotometric determination of carotenoids in chloroform extracts showed that the level of these compounds was twice as high in petals compared with sepals (Fig. C). In petals, chlorophyll was degraded from S0, with over 50% degraded by S3. In contrast, chlorophyll degradation was slower in sepals, with 70% still remaining at S5 (Fig. D). Thus, marked differences exist in the content of the major pigments in sepals and petals of A. peruviana. DISCUSSION : Lipid catabolism was extensive in A. peruviana floral tissues, resulting in an over 80% depletion of complex lipids (phospholipids and galactolipids) during senescence. The persistence of the perianth suggests that remobilization of nutrients from the floral tissues may occur. Although complex lipids were degraded, the level of neutral lipids remained fairly constant throughout senescence. In other plant senescing tissues, increases in neutral lipid biosynthesis have been reported and their adverse effects on the biophysical properties of the membranes described . Floral tissues contain diacylglycerol acyltransferase, the enzyme responsible for TAG formation . In stressed leaves, TAG synthesis is activated and may serve to sequester UFAs liberated by galactolipid catabolism . In A. peruviana, TAG production was low throughout flower opening and senescence and an increase in UFA was only seen at the terminal stage of senescence. A similar rise in the content of UFA was observed in carnation petals . UFA arising from membrane lipid metabolism may be removed from the bilayer by blebbing of lipid particles highly enriched in UFA from the membrane surface into the cytosol and may be metabolized by senescence-induced glyoxylate cycle enzymes . Accumulation of UFA results from the action of hydrolytic activity utilizing complex lipid substrates such as phospholipid and the neutral lipids, diacyl- and tri-acylglycerol. Recently, a cDNA clone has been obtained from carnation petals encoding a senescence-induced lipase that utilizes phospholipids . The abundance of the lipase mRNA increases just before petal senescence and the gene is also induced by ethylene. The corresponding senescence-induced lipase gene has been isolated from an Arabidopsis leaf senescence library. Antisensing this gene resulted in delayed leaf senescence, suggesting that it may play a central role in mediating the onset of this process . Fatty acids either present in complex lipids or released as UFA by acyl hydrolase activity are prone to peroxidation by either chemical (autoxidation) or enzymatic (lox) processes . Lipid peroxidation is a process that is often associated with senescence, although whether it is a primary event that initiates many of the downstream symptoms of senescence or is a consequence of this process remains unclear . In broccoli (Brassica oleracea) florets, lipid peroxidation was suggested to be a primary event associated with the onset of senescence . However, we have shown recently that although marked decreases in membrane polyunsaturated fatty acids (PUFAs) occur, as in A. peruviana, no increase in primary LHPO products or secondary TBARS could be detected during senescence initiation in this species . Furthermore, peak lox activity was also chronologically divorced from the rapid phase of PUFA consumption in the tissues. In many studies, lipid peroxidation is generally monitored by measuring some stable secondary end product of PUFA oxidation (e.g. MDA or hydrocarbon gases). The most widely used is the TBARS assay, which measures MDA derived from PUFAs. However, MDA can only be formed from fatty acids with three or more double bonds and because plant tissues often contain high levels of 18:2 (Deltacis,9,12), the TBARS assay may underestimate the actual extent of peroxidation. In A. peruviana floral tissues, 18:2 (Deltacis,9,12) is the predominant acyl constituent, and to estimate more accurately the extent of peroxidation, we employed a new method for detecting the levels of LHPOs . The level of these compounds is more likely to be physiologically relevant than MDAs. LHPOs accumulated in sepal but not in petal tissues, even though they had largely similar fatty acid and complex lipid profiles. This difference was not accounted for by differing patterns of lox enzymatic activity, which were similar in both tissue types and decreased markedly from S0. Perhaps the most surprising result was the difference in content of LHPOs in sepals and petals, while both exhibited similar electrolyte leakage profiles. In both tissues, the basal level of oxidation of the cell membrane was 1% of total fatty acids, similar to that recently calculated for a wide range of plant tissues . In sepals, this level rose to 3% during the course of senescence, whereas it remained constant at <1% in petal tissues. The degradation of LHPO by tissue extracts of A. peruviana petals and sepals was similar and the activity was maintained throughout senescence (data not shown). The observation that lox activity declines and LHPO enzymatic degradation is maintained at a constant level in both sepal and petal tissues indicates that the increase in LHPO in sepals is likely the result of nonenzyme-catalyzed reactions relating to autoxidation as has been shown recently for senescing Arabidopsis leaves . However, regio- as well as stereochemical analysis of lipid oxidation products revealed that 13S-oxy fatty acid isomers were exclusively detected in complex lipids. This is the first time that lox-derived products are found in complex lipids in a tissue other than germinating seeds. Thus, it is tempting to speculate that a lox-dependent degradation of polyunsaturated fatty acids in the esterified lipid fraction may occur in this senescence system as has been suggested for storage lipids in germinating oilseeds . Analysis of the oxygenated UFA, however, revealed a 9-lox that accounted for some 20% of the total lox activity present in both tissues. These 9-lox products were not detected in complex lipids. Whether 13-lox acts directly in situ on complex lipid fatty acids and 9-lox acts on UFA requires further investigation of the substrate specificity of the recombinant enzymes in vitro. In germinating barley (Hordeum vulgare) grains, two isozymes (BLYLOXA and BLYLOXC) were identified, with BLYLOXA characterized as a 9-lox preferentially acting on UFA, whereas BLYLOXC was found to be a 13-lox acting on fatty acids esterified to complex lipids . Interestingly, high levels of keto fatty acids were also present in both A. peruviana floral tissues and most likely arose as secondary products of lox reactions as shown for soybean (Glycine max; ) and pea (Pisum sativum; ). Keto fatty acids are usually minor products of lox reactions (5% --10%) formed under oxygen deprivation, whereas in A. peruviana, they represented about 30% of the oxylipins identified and as such are major products of the lox pathway. Whether these are derived from specific properties of A. peruviana lox requires further study, as does their physiological significance. In non-senescent tissues, lipid autoxidation is low and is controlled by a complex array of antioxidant components within the cells. These include the enzymes of the ascorbic acid cycle and other detoxifying enzymes such as superoxide dismutase and catalase, which remove potentially harmful reactive oxygen species . Pigments such as anthocyanins and carotenoids are also known to be free radical scavengers and limit reactive oxygen species propagation within plant cells . The level of TASA, which includes anthocyanins, increased in sepals coincident with the increase in LHPO levels. Thus, although a role in optical masking of chlorophyll by anthocyanins, thereby reducing the risk of photooxidation in leaf cell senescence, has been suggested , these compounds appear ineffective in preventing oxidative damage to the lipid components in A. peruviana sepal tissues. The level of carotenoids was almost double in the petals and during the course of senescence the level of chlorophyll also declined more markedly in petals. Thus, during the course of senescence, the ratio of chlorophyll to carotenoid increased in sepals and decreased in petals. Light absorption by chlorophyll can initiate the formation of free radicals in sensitized photooxidation . The ratio of chlorophyll to carotenoid would favor this process in sepal tissues and, therefore, could be a contributory factor in generating the higher levels of LHPO seen in sepal tissues. A survey of the different types of floral senescence patterns observed in the plant kingdom studied to date indicate that two categories can be defined on the basis of ethylene sensitivity or insensitivity. In the ethylene-sensitive category, lox activity may play a positive role in promoting senescence through oxidative membrane damage as seen in rose and carnation . In rose petals, a LOX transcript dramatically increased during senescence and its expression was stimulated by ethylene . On the other hand, ethylene-sensitive plants such as Phalaenopsis hyb and Dendrobium hyb (Orchidaceae) have been characterized in which lox does not play any apparent role in this process . In the ethylene-insensitive category, lox promotion of senescence has been proposed in daylily and implicated in Gladiolus sp. . Petals of daylily showed a 5-fold increase in TBARS from the time of opening to the point of electrolyte leakage and lox activity, which increased 2-fold, mirrored the increase in TBARS production. In contrast, in A. peruviana, floral tissues, lox activity had declined by some 80% at electrolyte leakage initiation and the accumulation of oxidized lipids (LHPOs and TBARS) was also not coincident with the loss of cellular ions. These observations indicate an unlikely role for lox in perturbing membrane function in this species and the chronology of lox activity and peroxidation product accumulation is quite different from those species in which an active role of lox in this process has been established. Because loxes preferentially utilize UFA, their activity is dependent on the supply of these substrates released by the action of acyl hydrolases. Thus, in those species in which lox plays an important role in perturbing membrane function, UFA may be more readily available to lox. In species in which lox appears to have a minor role, it could be envisaged that after de-esterification, UFAs are actively channeled by mechanisms such as membrane blebbing to other cellular compartments, like the glyoxysome, for efficient beta-oxidation and thus are unable to partake in lox-mediated reactions. Thus, based on the classification system of ethylene sensitivity/insensitivity and lox mediated or non-lox mediated, A. peruviana represents a distinct pattern of floral senescence that is both ethylene and lox independent. To date, this is the only species that clearly falls into this category. MATERIALS AND METHODS : Plant Material | Alstroemeria peruviana cv Samora plants were grown under glass at 16C +- 1C (day) and 13C +- 1o (night). Floral shoots were cut to about 10 cm, then transferred to distilled water and stored in a Saxcil growth cabinet at 20C, under an irradiance of 150 mumol m2 s-1, with a 16-h photoperiod and a relative humidity of 70%. Only the first flower on the cyme was used in these studies. Extraction of Lipids | Total lipids were rapidly extracted from tissues according to . All procedures were performed in dim light at 4C using chilled solvents (containing 0.01% [w/v] butylated hydroxytoluene) and glassware. Spectrophotometric Determination of LHPOs | LHPOs were determined in chloroform extracts immediately after extraction according to . Standard curves were constructed using linoleate hydroperoxide as standard synthesized by the method of . Lipid Purification and Quantitation | Lipids were purified by thin-layer chromatography on precoated silica gel plates (silica gel 60, Merck, Darmstadt, Germany) solvent systems described elsewhere . Fatty acid methyl esters (FAMEs) were quantified by gas liquid chromatography using heptadecanoic acid as the internal standard and on a 10% SP-2330 100/120 Chromosorb W AW (Supelco, Bellefonte, PA) column at 135C using a 6890 gas liquid chromatograph equipped with a flame ionization detector and a mass selective detector (Agilent Technologies, Stockport, UK). Extraction and Detection of Lox Products by HPLC | Oxidized fatty acids were extracted using a modification of . One gram of frozen floral tissue was added to 30 mL of extraction medium (3:2 [v/v] isohexane:isopropanol with 0.0025% [w/v] butylated hydroxytoluene) and immediately homogenized with an ultra Turax for 30 s under a stream of argon on ice. The extract was centrifuged for 10 min at 4,500g at 4C. The clear upper phase was collected and the pellet extracted three times with 3 mL each of extraction medium. To the combined organic phases, a 6.7% (w/v) solution of potassium sulfate was added to a volume of 47 mL. After vigorous shaking, the upper hexane-rich layer was removed. The upper organic phase containing the oxylipin fatty acid derivatives were dried under nitrogen and redissolved in 2 mL of isohexane:isopropanol (100:5 [v/v]), divided into two parts, and stored under argon at -80C until use. For the analysis of esterified fatty acids, the solvent was removed and 333 muL of a mixture of toluene and methanol (1:1 [v/v]) and 167 muL of 0.5 mm sodium methoxide were added. As internal standards, triheptadecanoate and triricinoleate were added. After incubation of the samples for 20 min, 0.5 mL of 1 m sodium chloride and 50 muL of 37% (v/v) HCl were added and the FAMEs were extracted twice each with 0.75 mL of hexane. The combined organic phases were evaporated to dryness under nitrogen and the corresponding FAMEs were dissolved in 50 muL of methanol:water:acetic acid (85:15:0.1 [v/v]). For the analysis of UFA derivatives, the solvent was removed and the sample was dissolved in 400 muL of methanol. As internal standards, heptadecanoic acid and 15-hydroxyeicosadienoic acid were added. Then, 10 muL of a 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide solution [1 mg of 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide/10 muL of methanol] was added and incubated for 2 h. After adding 200 muL of Tris buffer (0.1 m Tris-HCl, pH 7.5), the FAMEs were extracted twice each with 1 mL of hexane. The combined organic phases were evaporated to dryness under nitrogen and the corresponding FAMEs were redissolved in 50 muL of methanol:water:acetic acid (85:15:0.1 [v/v]). HPLC analysis was carried out on an Agilent 1100 HPLC system coupled to a diode array detector . At first, oxylipins were purified on a reverse phase-HPLC. This was performed on an ET250/2 Nucleosil 120 --5 C18 column (2.1 x 250 mm, 5-mum particle size, Macherey-Nagel, Dueren, Germany) with a methanol:water:acetic acid (85:15:0.1 [v/v]) solvent system at a flow rate of 0.18 mL min-1. SP-HPLC of the hydroxy fatty acids was carried out on a Zorbax Rx-SIL column (Agilent, 150 x 24.1 mm, 5-mum particle size) with a solvent system of n-hexane:2-propanol:acetic acid (100:1:0.1 [v/v]) and a flow rate of 0.1 mL min-1. Chiral phase-HPLC of the hydroxy fatty acids was carried out on a Chiralcel OD-H column (150 x 2.1 mm, 5-mum particle size, Daicel, Merck, Darmstadt, Germany) with a solvent system of n-hexane:2-propanol:acetic acid (100:5:0.1 [v/v]) and a flow rate of 0.1 mL min-1. A234 was monitored. TBARS | MDA was synthesized as described by and the endogenous levels present in the tissues were determined by a modified version of the TBARS assay recently described by . Lox Assays | Total proteins were extracted from approximately 1 g of floral tissue in a pestle and mortar in 2 volumes of 50 mm potassium phosphate, pH 7.0, containing: 1% (w/v) polyvinylpolypyrrolidone, 0.1% (w/v) Triton X-100 (t-octylphenoxypolyethoxyethanol), 0.04% (w/v) potasium bisulfate, and 1 mm dithiothreitol (modified from ) at 4C. The homogenate was centrifuged at 13,000g for 2 min, and the resulting supernatant was passed through a PD-10 column (Amersham, Buckinghamshire, UK), pre-equilibrated with 50 mm potassium phosphate buffer, pH 7.0. The protein eluate was resuspended in glycerol (20% [v/v] final concentration), snap frozen in liquid nitrogen, and stored at -80C until required. Lox activity was measured spectrophotometrically by following the increase in A234 (formation of conjugated dienes) from added fatty acid substrate (10 --100 nmol per assay in ethanol) in a 1-mL reaction volume at 25C with 0.005% (w/v) Triton X-100 (ultrapure, Pierce Chemical, Rockford, IL) in the buffer. Oleic acid (C18:1 cis Delta9) was used as a control fatty acid and the rate minus background (lacking addition of unsaturated fatty acids) was used to correct for activity determination. Lox Detection | Proteins were separated on a 10% (w/v) gel by SDS-PAGE (mini-gel system, Bio-Rad Laboratories, Hercules, CA), transferred to nitrocellulose membranes (Sartorius, Goettingen, Germany), and detected by chemiluminescence (ECL + Plus western blotting, Amersham). Lox antibody raised to recombinant cucumber (Cucumis sativus) oil body lox was used as the primary antibody . Electrolyte Leakage | Ten discs, 7 mm in diameter, were cut from both sepal and petal using a cork borer, avoiding the midrib, and directly weighed to determine their fresh weight. The tissue discs were washed with water for 10 min with constant agitation. The wastewater was then removed and an additional 10 mL of fresh water added. Conductivity was determined at the start (background) and after 2 h of incubation. All measurements were made in triplicate, the experiment was repeated twice, and the results were averaged (n = 6). Measurements were made on a beta800 conductivity meter, with a cell constant of K = 1 (EDT, Dover, UK). Measurements are expressed as the percent of total membrane leakage relative to electrolyte leakage determined for similarly aged prefrozen floral tissue. Total Antioxidant Status Determinations | The change in tissue TASA after harvest was monitored using the ABTS+ assay . Tissues were homogenized in 70% (w/v) ice-cold ethanol and centrifuged at 2,000g for 10 min. Twenty-microliter aliquots were assayed and Trolox (a vitamin E analog) was used to calibrate the assay. The reagents were purchased as a kit (Randox, Crumlin, UK) and assayed according to the supplied methodology. Protein and Pigment Determinations | Protein determinations were made using the standard assay protocol with bicinchoninic acid (Perbio Science UK Ltd, Tattenhall, UK) with bovine serum albumin as a standard. Chlorophyll and carotenoids were determined in the chloroform (lipid) extracts using the equations of and anthocyanins according to . Backmatter: PMID- 12226508 TI - Mitochondrial-Driven Bicarbonate Transport Supports Photosynthesis in a Marine Microalga AB - The CO2-concentrating mechanism (CCM) of the marine eustigmatophycean microalga Nannochloropsis gaditana consists of an active HCO3- transport system and an internal carbonic anhydrase to facilitate accumulation and conversion of HCO3- to CO2 for photosynthetic fixation. Aqueous inlet mass spectrometry revealed that a portion of the CO2 generated within the cells leaked to the medium, resulting in a significant rise in the extracellular CO2 concentration to a level above its chemical equilibrium that was diagnostic for active HCO3- transport. The transient rise in extracellular CO2 occurred in the light and the dark and was resolved from concurrent respiratory CO2 efflux using H13CO3- stable isotope techniques. H13CO3- pump-13CO2 leak activity of the CCM was unaffected by 10 mum 3(3,4-dichlorophenyl)-1,1-dimethylurea, an inhibitor of chloroplast linear electron transport, although photosynthetic O2 evolution was reduced by 90%. However, low concentrations of cyanide, azide, and rotenone along with anoxia significantly reduced or abolished 13CO2 efflux in the dark and light. These results indicate that H13CO3- transport was supported by mitochondrial energy production in contrast to other algae and cyanobacteria in which it is supported by photosynthetic electron transport. This is the first report of a direct role for mitochondria in the energization and functioning of the CCM in a photosynthetic organism. Keywords: Introduction : In many species of cyanobacteria and microalgae, the uptake of CO2 for photosynthesis is mediated by an energy-dependent CO2-concentrating mechanism (CCM). Several key components of the system have been identified in cyanobacteria and include metabolic influx pumps that actively transport and accumulate inorganic carbon (CO2 + HCO3- = Ci), carbonic anhydrase (CA), which catalyzes the conversion of accumulated HCO3- to CO2 near the site of Rubisco, and structurally intact carboxysomes, which house the majority of the cellular complement of Rubisco and CA . Physiological and biochemical studies indicate that there are multiple transport systems for Ci that recognize and use HCO3- and CO2 as substrates . It is widely accepted that Ci transport is light dependent and that photosynthetic electron transport provides the energy required for the active transport of both HCO3- and CO2 at the plasma membrane . In the cyanobacterium Synechococcus sp., there appears to be a division of labor within the photosystems in that HCO3- transport is supported by linear electron flow, whereas cyclic electron transport supplies the energy for CO2 uptake . The CCM is more complex in eukaryotic algae because of the increased number of metabolic compartments. Species diversity in terms of the required components for the CCM and its specific mode of operation in microalgae have been extensively documented . Diversity in the functional elements of the CCM include the participation (or absence thereof) of various forms of perplasmic CA , one or more plasma membrane-localized and/or chloroplast envelope-localized, Ci transport systems , pyrenoid-localized Rubisco , and intracellular CA localized within the pyrenoid and thylakoid membranes and mitochondria . Early experiments using Chlamydomonas reinhardtii indicated that photosynthetic processes provided the energy for the operation of the CCM . The details of energy supply are yet to be fully established, however, inhibition of light-dependent Ci accumulation in isolated chloroplasts of C. reinhardtii by 3(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU) strongly suggested a role for linear electron transport in Ci transport in this organelle . In marine microalgae, studies of the CCM have concentrated largely on examining the Ci species transported during photosynthesis and on the role of CA . Recent studies with the marine cyanobacterium Synechococcus sp. WH7803 and the eustigmatophyte alga Nannochloropsis spp. have, however, identified a new phenomenon associated with the operation of the CCM . In these species, illumination resulted in a significant and sustained rise in the external CO2 concentration as photosynthesis proceeded, rather than the expected draw-down of the external CO2 due to active Ci transport and photosynthesis . In other words, the cells acted as point-source CO2 generators substantially elevating both the internal and external CO2 concentration above their chemical equilibrium, in what might be considered a shot-gun approach to circumventing CO2 limitation. CO2 generation required two elements in Nannochloropsis spp.: an active HCO3- transport system to accumulate intracellular Ci and an intracellular CA to convert HCO3- to CO2. It appeared that a substantial portion of the internal CO2 subsequently leaked to the surroundings, and this leakage accounted for the unexpected rise in the external concentration. Inhibition of HCO3- transport by 4,4'-diisothiocyanatolstilbene-2,2'-disulfonic acid (DIDS) or the inhibition of intracellular CA activity by ethoxyzolamide (EZ) both prevented the rise in CO2 concentration in the surroundings and reduced the rate of photosynthesis . As a consequence, the CCM of Nannochloropsis spp. may be described minimally as HCO3- pump-CO2 leak activity. We have recently discovered that the CO2 generating system in Nannochloropsis gaditana continues to function for up to 20 min in the dark. Like its counterpart in the light, the CO2 generating system was sensitive to DIDS and EZ, suggesting that the same components involved in the HCO3- pump-CO2 leak activity of the CCM participated in CO2 generation in the dark. This observation seriously challenges the notion that the CCM is exclusively energized by the light-dependent photosynthetic electron transport chain. In the present work, we use mass spectrometry and stable isotope techniques to measure fluxes of 12CO2 and 13CO2 in cell suspensions of the marine microalga N. gaditana to distinguish between fluxes arising from respiratory metabolism and the HCO3- pump-CO2 leak activity of the CCM in the dark and light. Flux measurements were made in the absence or presence of inhibitors of mitochondrial respiration and chloroplast linear electron transport to investigate the role of these organelles in providing energy for HCO3- transport. Our results demonstrate that HCO3- uptake in the dark and light is driven by mitochondrial respiration and, thus, identify a novel component of the CCM in this alga. RESULTS : N. gaditana Generates CO2 in the Dark and Light | Illuminated cells of N. gaditana were allowed to reach the CO2 compensation point, and the light was switched off (Fig. a). The 12CO2 concentration in the medium rose to a very high level over the initial 5 min of the experiment and then gradually declined over the next 10 min. Addition of bovine CA to the reaction vessel (Fig. b) during any part of the time course resulted in a rapid diminution of the CO2 signal, indicating that CO2 was present in the medium at levels well above its chemical equilibrium value with HCO3-. The creation and maintenance of this chemical disequilibrium is indicative of the involvement of an energy-dependent step in this process. Such a large rise in CO2 after darkening has been observed thus far only in N. gaditana . The subsequent decline in CO2 concentration would not be anticipated, however, because respiratory metabolism should contribute to a continuous, though slow, increase in the external CO2 concentration. Figure 1 | Measurement of 12CO2 ( --- --- ---) and 13CO2 ( --- --- ---) fluxes in the dark and light. Measurement of 12CO2 ( --- --- ---) and 13CO2 ( --- --- ---) fluxes in the dark and light. a, An illuminated (1 mmol m-2 s-1, photosynthetically active radiation) cell suspension of N. gaditana was allowed to reach the CO2 compensation point, and the light was turned off. Changes in 12CO2 concentration were followed with time, and the light was then turned on. The asterisk indicates the transition from the slow to the fast phase of 12CO2 decline. b, As in a except that bovine CA (40 mug mL-1) was added to the darkened cell suspension during the rise in 12CO2. c, K213CO3 (100 mum) was added to reaction buffer (-cells) to determine the equilibrium 13CO2 concentration at pH 8.0 and 25C. d, As in a except that 100 mum K213CO3 was added 2 min after darkening and both 12CO2 and 13CO2 concentrations were measured over time. The time courses are superimposed for comparison. When the light was turned on (Fig. a), a new transient rise in CO2 concentration was observed that was then followed by a persistent decline as photosynthesis proceeded. Addition of bovine CA during the light phase also resulted in a diminution of the CO2 signal, indicating that the CO2 concentration was above its chemical equilibrium level (data not shown; ). A similar sustained evolution of CO2 in the light during CO2 fixation has been previously reported for N. gaditana and the marine cyanobacterium Synechococcus sp. WH7803 and Synechococcus spp. . In contrast to these results, other eukaryotic algae and cyanobacteria draw-down the CO2 concentration after illumination. Clearly, CO2 fluxes in N. gaditana are complex and cannot be accounted for simply by considering photosynthetic CO2 consumption in the light and respiratory CO2 production in the dark, because in several instances the net flux was opposite in direction to the major metabolic flux of CO2. Light-Enhanced Dark Respiration (LEDR) and HCO3- Pump-CO2 Leak Activity Are Required for Dark CO2 Generation | The initial rise in CO2 concentration in the dark can be attributed to a number of different sources including the release of an internal Ci pool, a photorespiratory postillumination CO2 burst, and mitochondrial respiration. By definition, the Ci pool at the compensation point is small, and in other algal species, it was released to the medium within 1 to 2 min after darkening, resulting in only a minimal rise in the extracellular CO2 concentration. Similarly, the postillumination CO2 burst would contribute only a small portion of the CO2 because the CCM of N. gaditana suppressed the formation of photorespiratory substrates required to initiate the burst . As a consequence, respiratory processes would be expected to be a major source of the CO2 appearing in the medium. N. gaditana is one of a number of plant and algal species that display a LEDR , where the respiration rate immediately after a period of photosynthesis is substantially higher than the steady-state rate. Under our conditions, LEDR measured immediately after darkening (O2 uptake) was on average 2.7-fold higher than the steady-state respiration rate measured 20 min after darkening (e.g. Fig. ). The rate of LEDR gradually declined to the steady-state over a period of 5 to 6 min. Thus, the period of LEDR (O2 uptake) was clearly associated with the postillumination period of maximum CO2 release , indicating that they are linked, probably through mitochondrial respiration. We have recently discovered that the HCO3- transport system in N. gaditana remained active in the dark , and it may, therefore, also contribute to the rise in CO2 through its HCO3- pump-CO2 leak activity. Figure 2 | Measurement of 12CO2 ( --- --- ---) and 16O2 (. Measurement of 12CO2 ( --- --- ---) and 16O2 (. . . . .) fluxes in a cell suspension of N. gaditana in the light (L) and dark (D) and in the absence and presence of the HCO3- transport inhibitor DIDS (500 mum). The experimental procedure was essentially the same as that for Figure . The asterisk indicates the transition from the slow to the fast phase of 12CO2 decline. To follow CO2 fluxes between the medium and the cells independent of respiratory 12CO2 production, we added 100 mum13Ci to the cell suspension 2 min after turning the light off (Fig. , c and d). In the absence of cells, the 13CO2 concentration in the medium rose to the expected equilibrium level and then remained constant (Fig. c). In the presence of cells, the 13CO2 concentration also increased with time (Fig. d) but to a level 5.6-fold the equilibrium value and then declined gradually over time in parallel with the 12CO2 signal. Thus, the 13CO2 signal displayed the same dynamics as the 12CO2 signal, but in this case, the rise in 13CO2 cannot be attributed to LEDR or steady-state respiration because mitochondrial substrates were not enriched with 13C. Participation of H12CO3- in this energy-driven pump-leak activity would also be expected to occur because it is the natural substrate for the transporter. The rise in CO2 concentration may also be due to a cellular acidification of the medium. However, measurement of extracellular pH indicated that this parameter remained constant during the experiments (data not shown). As a consequence, the unusually large rise in 12CO2 concentration in the dark (Fig. a) can be attributed to two superimposed processes. First, LEDR-generated CO2, which was rapidly released to the medium, and some of it was hydrated to form HCO3-. Second, this HCO3- then served as substrate for the HCO3- pump-CO2 leak activity in the dark resulting in a further increase in CO2 efflux. The combined processes of CO2 formation occurred at a rate faster than the uncatalyzed conversion of CO2 to HCO3- in the medium as evidenced by the rapid diminution in CO2 concentration when CA was added (Fig. b). In the light, similar processes occurred that accounted for the initial rise in CO2 upon illumination. The rise was considerably smaller, however, because Rubisco-mediated fixation consumed a portion of the CO2. Both the 12CO2 and 13CO2 signals ultimately declined to the CO2 compensation concentration because the Ci was consumed in photosynthesis (Fig. d). Contribution of HCO3- Transport | To estimate the relative contributions of respiration and the HCO3- pump-CO2 leak activity to the rise in CO2 concentration, experiments were conducted in the absence and presence of the HCO3- transport inhibitor DIDS . The addition of 500 mum DIDS to illuminated cells at the CO2 compensation point significantly reduced the level of CO2 efflux once the cells were darkened. However, DIDS only had a small, negative effect (10%) on the rate of LEDR dark O2 consumption, indicating that the major effect of DIDS was on the HCO3- pump-CO2 leak activity. At the DIDS concentration used, HCO3- transport was inhibited by about 90% . Using the difference in peak heights in the dark as an estimate of activity (e.g. Fig. ), HCO3- pump-CO2 leak activity contributed approximately 40% to 50% to the total rise in CO2. Illumination of the cell suspension resulted in a rapid rise in extracellular CO2 concentration, which was greatly reduced in cells treated with the HCO3- transport inhibitor , indicating that DIDS-sensitive HCO3- transport activity was essential for the rise in CO2 in the light and in the dark. Thus, the rise in 13CO2 concentration above the equilibrium level was diagnostic for HCO3- transport activity in N. gaditana. Oxygen Is Required for HCO3- Pump-CO2 Leak Activity in the Dark | The decline in CO2 concentration in the dark often displayed two distinct phases, an initial slow phase followed by a second and more rapid rate of disappearance (e.g. Figs. a and , *). The accelerated phase of CO2 disappearance was also observed in the 13CO2 signal (Fig. d, *) and its onset corresponded with the approach to anaerobic conditions in the medium, brought about by respiratory O2 consumption . Because the 13CO2 signal solely reflects HCO3- pump-CO2 leak activity, this observation suggests a dependence of a component of the pump-leak activity on O2 availability. To test this hypothesis, we examined the effect of O2 concentration on the magnitude of CO2 efflux . The O2 concentration in the medium was initially set by gassing with a stream of N2, or by the addition of dithionite to achieve zero O2. In the absence of O2, darkening of a cell suspension at the CO2 compensation point resulted in a very small rise in 12CO2, consistent with the suggestion that O2-dependent LEDR was responsible for part of the large rise in CO2 concentration in the dark. The addition of 100 mum13Ci, 2 min after darkening, resulted in a rise in 13CO2 concentration only to the expected equilibrium level (Fig. a). In either case, illumination did not result in a rise in the CO2 signals, as observed in the control (Fig. d, 230 mum O2). As the O2 concentration was increased, the 13CO2 concentration also increased progressively in the dark, although the absolute amount of 13Ci added was the same in each case (Figs. , b --d, and 1, c and d). These data support the concept that the pump-leak activity associated with HCO3- transport required O2. The 12CO2 concentration also rose progressively due to combined mitochondrial respiration and HCO3- pump-CO2 leak activity. The rise in CO2 concentration expected upon illumination was also restored with increasing O2 concentration (Figs. , a --d, and 1c). Figure 3 | Measurement of 12CO2 ( --- --- ---) and 13CO2 ( --- --- ---) fluxes in the dark and light in the presence of 0 (a), 1.5 (b), 5.6 (c), and 130 (d) mum O2. Measurement of 12CO2 ( --- --- ---) and 13CO2 ( --- --- ---) fluxes in the dark and light in the presence of 0 (a), 1.5 (b), 5.6 (c), and 130 (d) mum O2. The plots obtained at 230 mum O2 are shown in Figure d. The time courses are superimposed for comparison. The experimental procedure was essentially the same as that for Figure ; off, light turned off; on, light turned on. Mitochondrial Respiration Energizes HCO3- Transport | The occurrence of HCO3- pump-CO2 leak activity in the dark and its dependence on O2 suggested that mitochondrial respiration may be involved in energizing a component of that system. The most likely candidate is HCO3- transport because CA is a freely reversible enzyme, whereas the uptake of HCO3- would be against its electrochemical potential. To assess mitochondrial involvement, CO2 flux experiments were carried out in the presence of various inhibitors of respiration (Fig. ; Table ). The addition of 250 mum KCN to cell suspensions during the dark CO2 efflux phase mimicked the effect of anoxia and resulted in a rapid decline in both 12CO2 and 13CO2 concentrations (Fig. b). Upon illumination, the efflux of 13CO2 was abolished and that of 12CO2 was markedly reduced. Inclusion of KCN in the cell suspension at the CO2 compensation point in the light resulted in a significant inhibition of 12CO2 efflux after darkening (Fig. c) and a 82% decrease in O2 consumption (data not shown), confirming inhibition of mitochondrial respiration. Inhibition of HCO3- pump-CO2 leak activity was also evident because the addition of 100 mum13Ci resulted in a rise in 13CO2 only to its equilibrium level. The efflux of both 12CO2 and 13CO2 was abolished during a subsequent dark-light transition. Concentrations of KCN as low as 25 mum were also effective in inhibiting CO2 efflux in the dark and in the light , suggesting a requirement for mitochondrial complex IV in HCO3- pump-CO2 leak activity. Consistent with this hypothesis was the observation that NaN3, another inhibitor of complex IV, inhibited 12CO2 and 13CO2 efflux in the dark and light and could also mimic the effect of anoxia (data not shown). Figure 4 | Measurement of 12CO2 ( --- --- ---) and 13CO2 ( --- --- ---) fluxes in the dark and light in the absence (a; control) and presence (b and c) of 250 mum KCN. Measurement of 12CO2 ( --- --- ---) and 13CO2 ( --- --- ---) fluxes in the dark and light in the absence (a; control) and presence (b and c) of 250 mum KCN. KCN was added during the slow decline in CO2 (b) or 2 min before darkening (c). The time courses are superimposed for comparison. The experimental procedure was essentially the same as that for Figure ; off, lights turned off; on, lights turned on. Table I | Effect of inhibitors on 13CO2 efflux In the light, KCN may also inhibit photosynthetic electron transport directly at plastocyanin and lead to the inhibition of CO2 fixation and O2 evolution. At an external Ci concentration (0.1 mm) just sufficient to saturate photosynthesis , 250 mum KCN abolished O2 evolution . However, increasing the external Ci to 50 mm resulted in a restoration of the photosynthetic rate to 75% of the control rate. Although KCN has a clear effect on photosynthesis, these results also indicate that the ability of the photosynthetic electron transport system to supply energy to the Calvin cycle remained largely functional and that it was the availability of Ci that was the limiting factor, consistent with an impairment of the CCM not directly related to chloroplast energy supply. Figure 5 | Effect of 250 mum KCN on the time course of photosynthetic O2 evolution in the presence of various levels of external Ci . Effect of 250 mum KCN on the time course of photosynthetic O2 evolution in the presence of various levels of external Ci . For comparison, O2 evolution in the absence of KCN at 0.1 mm Ci (circle) is also shown. The results obtained with KCN and NaN3 were in marked contrast to those observed with DCMU, an inhibitor of chloroplast linear electron transport . At 10 mum, the rise in 13CO2 was unaffected by the inhibitor in the dark or the light, indicating that the HCO3- pump-CO2 leak activity could be sustained when the energy supply from photosynthetic linear electron transport was restricted. As anticipated, photosynthetic O2 evolution was strongly inhibited (6.4-fold), indicating that DCMU effectively blocked linear electron transport in N. gaditana. At 5 times the DCMU concentration , 13CO2 efflux in the dark occurred at 86% of the control level, but 13CO2 efflux in the light was abolished, indicating that chloroplast energy supply may also contribute to some extent in the pump-leak activity . Figure 6 | Measurement of 12CO2 ( --- --- ---), 13CO2 ( --- --- ---), and 16O2 (. Measurement of 12CO2 ( --- --- ---), 13CO2 ( --- --- ---), and 16O2 (. . . . .) fluxes in the dark and light in the absence and presence of 10 mum DCMU. The time courses are superimposed for comparison. The experimental procedure was essentially the same as that for Figure . DCMU reduced 12CO2 efflux in the dark, but had no effect on 13CO2 efflux or on the steady-state rate of O2 uptake. The effect of DCMU on LEDR-dependent 12CO2 efflux was likely indirect and the result of a decreased supply of oxidizable substrate to the mitochondria from the chloroplasts, leading to less CO2 production and ultimately to reduced availability of HCO3- for the pump-leak activity. This possibility seems reasonable because have shown that LEDR was substantially, but indirectly, reduced by DCMU in C. reinhardtii. The effects of two additional inhibitors of mitochondrial energy metabolism on 13CO2 efflux were also examined . The respiratory electron transport inhibitor rotenone, which blocks complex I, reduced 13CO2 efflux in the dark and the light by 40% and 80%, respectively. The transport inhibitor N-ethylmaleimide, which prevents ATP export from the mitochondria, reduced 13CO2 efflux to 60% of the control level in the dark and abolished it in the light. With time (> 45 min), NEM ultimately reduced dark efflux to zero (data not shown). DISCUSSION : N. gaditana generated CO2 internally in the light and dark, which raised the extracellular CO2 concentration to levels well above its equilibrium with HCO3-. CO2 generation was associated, in part, with HCO3- transport and intracellular CA activity . This HCO3- pump-CO2 leak activity was resolved from respiratory 12CO2 release by using 13Ci (H13CO3-) as the substrate for the HCO3- transporter and monitoring the rise in 13CO2 in the medium.13CO2 concentrations above the measured equilibrium value were diagnostic for HCO3- pump-CO2 leak activity . Mitochondrial Energization of the CCM | The locations of the internal sites of CO2 generation are not known other than that they coincide with sites where CA is located . Net diffusion of CO2 away from these sites would occur randomly, following the chemical potential between the cells and the medium. As a consequence, high levels of internal CO2 would be distributed more or less evenly throughout the cells and would be dependent on HCO3- transport. Minimally, the internal CO2 concentration would be expected to be equal to the highest extracellular CO2 concentration observed. In the light, HCO3- pump-CO2 leak activity would function as a rudimentary CCM increasing the CO2 concentration in the chloroplast (and other compartments), thereby facilitating high rates of CO2 fixation and suppressing photorespiration. N. gaditana lacks CO2 transport capability, extracellular CA, and pyrenoids , features that in other algae play an important role in the efficient and refined operation of the CCM.. It is this unique set of circumstances in N. gaditana that allow us to detect HCO3- pump-CO2 leak activity in this organism by mass spectrometry. HCO3- pump-CO2 leak activity in the light and dark was inhibited by DIDS , by EZ , and by anoxia . The common effect of these very different treatments on reducing CO2 generation indicated that the same biochemical components mediated HCO3- pump-CO2 leak activity in the light and dark. Thus, information obtained from experiments in the dark is relevant in interpreting the results of the experiment in the light and vice versa. The O2 requirement of active HCO3- transport and its occurrence in the dark was inconsistent with light being the sole or primary energy source to drive this process. Inhibitors of respiration such as KCN, NEM, and others reduced or completely inhibited HCO3- pump-CO2 leak activity in the light and dark, consistent with a requirement for mitochondrial energy supply in the process. A primary role for light in energizing HCO3- pump-CO2 leak activity (CCM) was further challenged by the observation that low concentrations of DCMU had little effect on 13CO2 efflux, but significantly reduced the photosynthetic rate. Energy supply from the mitochondrion, therefore, was critical to HCO3- pump-CO2 leak activity, whereas energy from chloroplast linear electron transport was not. We cannot, however, completely exclude a role for light in HCO3- pump-CO2 leak activity because high levels of DCMU blocked 13CO2 efflux in the light, but not in the dark. Thus, complex interactions between chloroplasts and mitochondria may also be associated with the process. In other algae and cyanobacteria, active Ci transport rapidly ceases after darkening, reflecting a requirement for photosynthetically derived energy to fuel Ci accumulation . In N. gaditana, HCO3- pump-CO2 leak activity continues for at least 20 min in the dark before it stops. Thus, regulation of HCO3- transport by chloroplast energy supply seems unlikely. Restoration of dark HCO3- pump-CO2 leak activity can be achieved by briefly (15 --45 s) interrupting the dark phase with white light . Thus, light may activate, but not energize, HCO3- transport. The continuation of HCO3- pump-CO2 leak activity in the dark may, therefore, reflect a rather prolonged deactivation process that is more in tune with the gradual decrease in light during the natural day-night cycle than with the abrupt light-dark transition of our experiments. One hypothesis that can account for the experimental data is that mitochondria-derived ATP is exported to the cytoplasm where it is used to drive, either directly on indirectly, a plasma membrane-localized active HCO3- transport system in the dark and light. The accumulated HCO3- is converted by a thermodynamically reversible CA to CO2 resulting in a high level within the cell and a rise in the external CO2 concentration. In the light, the effect is to concentrate CO2 in the chloroplast and to provide saturating levels of substrate for Rubisco. In the dark, LEDR releases CO2 to the medium, some of which is converted nonenzymatically to HCO3-. This HCO3- serves as the substrate for the HCO3- pump, and the resulting leakage of CO2 is superimposed upon the respiratory efflux. As LEDR diminishes with time, the rate of CO2 production declines, but HCO3- pump-CO2 leak activity and the uncatalyzed conversion of CO2 to HCO3- continue with the net result being a decline in CO2 in the medium. This decline corresponds to the slow decline phase seen in the time course experiments (e.g. Figs. a and ). When respiration depletes the available O2, oxidative ATP and CO2 production stop, causing a cessation in CO2 release to the medium and in HCO3- pump-CO2 leak activity. But conversion of CO2 to HCO3- in the medium continues, causing a further decrease in CO2 concentration that corresponds to the second, more-rapid phase in the CO2 decline curve. At present, we cannot rule out involvement of a chloroplast-localized HCO3- transport system powered by mitochondria-derived ATP. However, either the ATP-binding site of the transporter would need to be oriented toward the cytosol or an import of mitochondria-derived ATP to the chloroplast would have to occur . In the light, it is also possible that a complex shuttle of reductant from the chloroplast to the mitochondria may also contribute to ATP generation in the mitochondrion, which is used to transport HCO3-. This situation could explain the inhibition by high concentrations of DCMU of 13CO2 efflux in the light. The association of active HCO3- transport with mitochondrial energy supply and the operation of the CCM has not been observed before. However, a role for mitochondria in the acclimation of the green algal C. reinhardtii to CO2-limiting growth conditions has been proposed, based on the observations that specific mitochondrial CAs were induced upon transfer to a low-CO2 environment and the relocation of mitochondria from central region of the cells to the periphery . Participation of mitochondria in the CCM in other species may be masked by the presence of external CA, which would prevent the rise in external CO2 concentration, or by active CO2 transport, which would rapidly recycle CO2 that leaked from the cells . New experimental approaches will be needed to detect mitochondrial involvement in the CCM in these species and to determine whether this phenomenon is widespread. The dependence of HCO3- uptake on mitochondria energy supply may well be a primitive characteristic of algae and confined to a few species. The genus Nannochloropsis is a member of the class Eustigmatophyceae within the division Heterkonta and is considered to be one of its most primitive members , based on the lack several photosynthetic pigments found in other members of the division . Intracellular structures of eustigmatophycean algae indicate that they may have arisen by a secondary endosymbiosis where a eukaryotic alga was engulfed by a phagotrophic oomycete. This may explain the features of Nannochloropsis spp. in that the oomycete host may have had a constitutive HCO3- transporter, but is unlikely to have had either an external CA or an active transport of CO2 to facilitate Ci acquisition. MATERIALS AND METHODS : Organism and Growth | The unicellular marine microalga Nannochloropsis gaditana was grown at 20C in artificial seawater (Sigma, St. Louis) supplemented with F/2 medium and bubbled with air (0.035% [v/v] CO2) at a rate of 60 mL min-1 . Cultures were continuously illuminated at a photon flux density of 75 mumol quanta m-2 s-1 provided by a combination of cool-white and Gro-lux fluorescent lamps. Cells were maintained in exponential growth phase by daily dilution. Experimental Conditions | Cells were harvested by centrifugation (2,800g, 10 min), washed twice with and resuspended in a Ci-free medium containing 450 mm NaCl, 40 mm MgCl2, 10 mm KCl, 10 mm Na2SO4, and 5 mm CaCl2, and buffered at pH 8.0 with 25 mm TRIZMA-Base. The medium was depleted of Ci by gassing with N2. The cell suspension (6 mL) was placed in a glass reaction vessel at 25C and 1 mmol quanta m-2 s-1 light (photosynthetically active radiation) and was allowed to fix residual Ci in the medium until the CO2 compensation point was reached. The final cell density was 2 x 108 cells mL-1 and corresponded to 30 mug chlorophyll a mL-1. Mass Spectrometry | Washed cell suspensions (6 mL) were transferred to a glass reaction vessel containing a magnetic stirrer, and the chamber was closed with a plexiglass stopper leaving no head space. Gases and treatment solutions were introduced into the chamber through a capillary bore in the stopper. The reaction chamber was connected to the ion source of a magnetic sector mass spectrometer (model MM 14 --80 SC, VG Gas Analysis, Middlewich, UK) by an inlet covered with a dimethyl silicone membrane that allowed dissolved gases to pass freely but not ions like HCO3- . The illuminated cells were allowed to consume residual Ci before experiments commenced and then were darkened and treated with inhibitors, if necessary. When used, 13Ci was supplied to the cells from a stock solution of K213CO3 (300 mm). Concentrations of 16O2, 12CO2, and 13CO2 (m/z = 32, 44, and 45, respectively) in cell suspensions were measured simultaneously. The mass spectrometer was calibrated for O2 and CO2 as described previously . Rates of O2 evolution or consumption were derived from the slopes of the m/z = 32 signal. In some experiments, net photosynthetic O2 evolution was measured using a temperature-controlled Clarke-type electrode (Hansatech, King's Lynn, UK; ). Backmatter: PMID- 12226509 TI - Early Embryo Development in Fucus distichus Is Auxin Sensitive AB - Auxin and polar auxin transport have been implicated in controlling embryo development in land plants. The goal of these studies was to determine if auxin and auxin transport are also important during the earliest stages of development in embryos of the brown alga Fucus distichus. Indole-3-acetic acid (IAA) was identified in F. distichus embryos and mature tissues by gas chromatography-mass spectroscopy. F. distichus embryos accumulate [3H]IAA and an inhibitor of IAA efflux, naphthylphthalamic acid (NPA), elevates IAA accumulation, suggesting the presence of an auxin efflux protein complex similar to that found in land plants. F. distichus embryos normally develop with a single unbranched rhizoid, but growth on IAA leads to formation of multiple rhizoids and growth on NPA leads to formation of embryos with branched rhizoids, at concentrations that are active in auxin accumulation assays. The effects of IAA and NPA are complete before 6 h after fertilization (AF), which is before rhizoid germination and cell division. The maximal effects of IAA and NPA are between 3.5 and 5 h AF and 4 and 5.5 h AF, respectively. Although, the location of the planes of cell division was significantly altered in NPA- and IAA-treated embryos, these abnormal divisions occurred after abnormal rhizoid initiation and branching was observed. The results of this study suggest that auxin acts in the formation of apical basal patterns in F. distichus embryo development. Keywords: Introduction : The basic body organization of plants and algae is established during embryogenesis, with the apical basal pattern resulting from an initial asymmetric cell division (for review, see ; ). This first cell division leads to apical and basal daughter cells that are precursors to the shoot meristem and root meristem, respectively, in land plants and thallus and rhizoid tissues in algae, respectively. The directional cues that control the polarity of this first cell division are not yet clear in higher plants, although maternal factors have been suggested to orient the embryo axis . In contrast, in embryos of brown algae, environmental gradients, with light being the best characterized, control the polarity of the apical basal pattern . In higher plants and algae, similarities in the mechanism by which the first asymmetric cell division is established are becoming apparent. In Arabidopsis, the product of the GNOM gene is required for stable fixation of the apical basal axis . The GNOM gene encodes a brefeldin A (BFA)-sensitive guanine nucleotide exchange factor for small GTP-binding proteins, which function in membrane trafficking . In Arabidopsis, both the gnom mutation and BFA treatment led to alterations in membrane protein localization, including PIN1, a putative auxin efflux carrier protein . In embryos of brown algae, treatment with BFA leads to alterations in the orientation of the first cell division and to formation of embryos with multiple or branched rhizoid or basal cells (; for review, see ). These results implicate vesicle secretion in early events in formation of cell polarity in land plants and algae embryos. The plant hormone auxin may also play a role in embryo development (for review, see ; ). Several mutants with altered auxin responses have embryo developmental defects, including axr6, bdl (bodenlos), and mp (monopterous; ; ; ). The mutations in the MP and BDL genes alter the division plane of the apical daughter cell and affect both the central and basal cell lineages . Because the mutations in AXR6, BDL, and MP genes give rise to seedlings with no primary root, these genes appear to be required for organizing embryonic root formation, but are not necessary for postembryonic root formation (for review, see ). The nature of the proteins encoded by these genes so far indicates a role for auxin-regulated gene expression in these development processes (for review, see ). Additional studies have suggested that embryo development may be directly tied to the ability to properly transport auxin. Treatment of embryos with inhibitors of indole-3-acetic acid (IAA) efflux led to the development of altered shapes of embryos in carrot (Daucus carota; ; ), Ipomoea batatas , mustard (Sinapis alba; ), wheat (Triticum aestivum; ), and Brassica juncea . The altered embryo structures that result from these treatments resemble the defects found in the gnom mutant, described above and in plants with a mutation in the AtPIN1 gene , which is predicted to encode an auxin efflux carrier . PIN1 is normally asymmetrically localized to one plane of the plasma membrane and this localization is believed to control the directionality of auxin transport (for review, see ). Finally, mutations in the GNOM gene or treatments with BFA also led to embryos with a mislocalization of PIN1 . Therefore, these results suggest interdependence between auxin transport and the targeted vesicle secretion that plays an important role in early events in embryo development. To examine the role of auxin during the formation of the apical basal axis, Fucus distichus embryos provide an excellent system. F. distichus gametes are released into seawater and fertilization and development occur in solution, facilitating the examination of large populations of synchronized embryos (for review, see ). A number of events that occur before the first asymmetric cell division have been identified, and these studies provide an excellent framework for understanding formation of developmental polarity (for review, see ). Because targeted secretion, actin filaments, and ionic gradients have all been implicated in polarity formation in these embryos (for review, see ), it should also be possible to examine the relationship between auxin and these processes in the context of early development. In the present study, the role of auxin and polar auxin transport in the initial stages of F. distichus embryo development was examined. The presence of IAA in F. distichus was verified by gas chromatography (GC)-mass spectroscopy (MS) analysis of extracts from F. distichus tissue. The ability of F. distichus to transport IAA, and the activity of auxin efflux carriers and influx carriers, were examined. Altered embryo developmental patterns were induced by treatment with either auxin or auxin transport inhibitors, with these compounds active in the first 6 h after fertilization (AF), which is before the first cell division. Together, these experiments suggest that auxin plays a role in the formation of the apical basal pattern of these embryos. RESULTS : F. distichus Contains Free IAA at Concentrations Comparable with Higher Plants | Extracts of both zygotes and mature tissues of F. distichus were prepared, IAA was purified, and GC-MS was used to demonstrate the presence of IAA and to quantify the IAA concentration. An extract from the fruiting tips of F. distichus was subjected to GC-MS. A total ion chromatograph (TIC) of a peak eluting from the GC with a retention time of 5.593 to 5.600 min is shown in Figure . The peaks at 189 and 130 mass-to-charge ratio (m/z) result from intact methylated IAA and the predominant fragmentation product, respectively. Additional fragments of the endogenous IAA are an unmarked peak at m/z 77 and a peak at 103. Ions at m/z 136 and 195 are the quinolinium ion and molecular ion from the [13C]-IAA internal standard. This fragmentation pattern of a sample with this retention time on GC is conclusive demonstration of the presence of IAA in these extracts. Figure 1 | IAA is present in F. distichus IAA is present in F. distichus embryos as demonstrated by a TIC of methylated, HPLC-purified samples of free IAA from F. distichus fruiting tips. The ions at m/z 130 and 189 are the quinolinium ion and molecular ion, respectively, of the methylated ester of the endogenous IAA. Additional fragments of the endogenous IAA are an unmarked peak at m/z 77 and a peak at 103. Ions at m/z 136 and 195 are the quinolinium ion and molecular ion from the [13C]-IAA internal standard. Free IAA concentrations measured from extracts of mature fruiting tips and zygotes using GC-selected ion monitoring-MS are reported in Table . The amounts of IAA were determined relative to the [13C]-IAA internal standard, using single ion chromatographs, and are compared with IAA measurements from the literature for other plants. The concentration in fruiting tips was in the range previously reported for tobacco (Nicotiana tabacum) leaves and that reported for Arabidopsis cotyledons . The free IAA concentration in embryos changes during development, with a range between 5 and 31 ng g fresh weight-1 for carrot embryos , and in wheat embryos, a range from 8 to 90 ng g fresh weight-1 . The free IAA concentration obtained for F. distichus zygotes is similar to the lower end of both of these ranges. The calculated IAA concentration may be an underestimation of the concentration, though, because the F. distichus embryos were removed from artificial seawater (ASW) for these analyses and it is possible that some of the ASW was retained with the embryos, thereby causing an overestimation of the weight and an underestimation of the free IAA concentration. Table I | Measurement of free IAA in F. distichus and land plants One concern that has been previously raised about measurement of IAA in algal samples is the possibility that the presence of IAA is due to contaminating microorganisms . This possibility cannot be eliminated in the reported measurements of free IAA in thallus tissues collected from the ocean. This is much less of an issue with F. distichus zygotes, though. During preparation, F. distichus zygotes were filtered to remove thallus tissue and any associated microorganisms and then allowed to settle briefly under conditions that would select against contaminating microbes. In addition, other samples of zygotes were recovered in the presence of antibiotics and IAA amounts were similar to those reported here (data not shown). Naphthylphthalamic Acid (NPA) Causes Accumulation of [3H]IAA in F. distichus Embryos | To determine if F. distichus zygotes have the ability to move IAA across membranes, and whether an IAA efflux carrier controls auxin movement with similarities to those in land plants, auxin accumulation assays were performed. F. distichus zygotes (2 h AF) were incubated for 1 h in the presence of 25 nm [3H]IAA along with a range of NPA concentrations. The [3H]IAA not associated with the zygotes was removed by filtration and the [3H]IAA accumulated within the zygotes was quantified. In land plants, NPA reduces IAA efflux . In F. distichus zygotes, NPA also appears to block efflux, thereby increasing the accumulation of [3H]IAA, as shown in Table . The reduction in IAA accumulation at the highest dose of NPA is often seen in other transport assays using Arabidopsis and zucchini (S.R. Brady and G.K. Muday, unpublished data), although the explanation for these bell-shaped dose response curves are unknown. The magnitude and concentration dependence of the effect by NPA is comparable with that found in tissue segments, tissue culture cells, or membrane vesicles derived from many green plants . Table II | [3H]-IAA accumulation in F. distichus zygotes IAA Does Not Alter Accumulation of [3H]IAA in F. distichus Embryos | To determine if IAA accumulation into F. distichus zygotes is carrier mediated, the ability of unlabeled IAA to prevent [3H]IAA accumulation was tested. If the uptake of IAA is carrier mediated, then increasing concentrations of unlabeled IAA should reduce the amount of radiolabeled IAA accumulation and the total IAA accumulation should be saturable. The amount of total IAA accumulation, using the [3H]IAA as a tracer to estimate the total IAA accumulation, is plotted as a function of unlabeled IAA added and is shown in Figure . There is a linear relationship between the amount of total IAA accumulated and the amount of cold IAA added. In contrast, in corn (Zea mays) coleoptile segments and tobacco cells, auxin concentrations of 100 mum saturate IAA accumulation . This result suggests that movement of IAA into F. distichus zygotes is due to passive diffusion of the protonated and hydrophobic IAA molecule, rather than being mediated by an IAA uptake protein. Figure 2 | IAA accumulation into F. distichus IAA accumulation into F. distichus zygotes is not saturable. Accumulation of [3H]IAA into F. distichus zygotes was measured in the presence of a range of unlabeled IAA concentrations but with a constant 25 nm concentration of [3H]IAA. The total IAA accumulation was calculated using the tracer IAA and is plotted as a function of added cold IAA. IAA and NPA Alter Embryo Development | The effects of IAA and the auxin transport inhibitor NPA on embryo development were examined by culturing embryos in ASW in the presence and absence of 50 mum of these compounds on horizontal slides in the dark, as shown in Figure . In Figure A, a representative embryo with a single unbranched rhizoid that was cultured in untreated ASW is shown. Embryos grown on IAA predominantly developed multiple rhizoids, often with four or more randomly distributed rhizoids, as shown in Figure B. Growth on NPA led to embryos with branched rhizoids (Fig. C). Although most F. distichus rhizoids will ultimately form branches, these normally occur much later in development and in cells considerably further from the thallus cell. If IAA and NPA treatments are performed in the light, the magnitude of these developmental alterations are significantly reduced (data not shown), which is consistent with a previous report in which IAA had no significant effects and an IAA efflux inhibitor had only slight effects on F. distichus embryo development in the light . Because F. distichus embryos normally are released and undergo the earliest stages of development in the light, the physiological relevance of these developmental effects of exogenous IAA and NPA should be considered. There is experimental evidence that in mosses, light suppresses auxin signals. Cryptochrome-disrupted or auxin-treated Physcomitrella patens have increased branching patterns and the cryptochrome disruptants have higher sensitivity to exogenous auxin . Therefore, blue light, acting through cryptochrome, reduces the auxin response much as light reduces the branching effects of auxin on F. distichus embryos reported here. In addition, NPA and IAA do perturb F. distichus development in the light as well, but in a different way. Treatments with IAA and auxin efflux inhibitors will reduce the percentage of embryos that are polarized in response to unilateral light (S. Basu and G.K. Muday, unpublished data). Figure 3 | IAA and NPA led to multiple and branched rhizoids. IAA and NPA led to multiple and branched rhizoids. F. distichus embryos were grown for 48 h in the dark in ASW or in the presence of 50 mum IAA or 50 mum NPA as indicated. The effects of NPA and IAA, and the combination of these two compounds, on embryo development are summarized for three separate experiments using 50 mum of each compound in Table . The reported values are for 48 h of treatment, although the results for 12 and 24 h were very similar (data not shown). IAA caused a greater than 10-fold increase in the number of embryos with multiple rhizoids. NPA also led to a slight 1.5-fold increase in multiple rhizoids, although this increase was not statistically significant (P = 0.17). When both NPA and IAA were added together, the number of embryos with multiple rhizoids was similar to when IAA was added alone. Table III | Effect of IAA and NPA on F. distichus developmental patterns Although IAA did not affect the percentage of embryos with branched rhizoids, NPA treatment resulted in branched rhizoids in greater than 50% of the embryos. This is a statistically significant increase over the DMSO control that leads to an increase of between 2- and 3-fold in the number of embryos with branched rhizoids. The magnitude of the effect is at the lower end of that range in this summary due to the unusually large number of branched rhizoids in the controls, as compared with Figures through . When NPA and IAA were added together, IAA prevented the increased number of branched embryos induced by NPA. Figure 4 | NPA and triiodobenzoic acid (TIBA) cause abnormal embryo development in the dark. NPA and triiodobenzoic acid (TIBA) cause abnormal embryo development in the dark. The percentage of F. distichus embryos with branched and/or multiple rhizoids are reported in response to treatment with BA, NPA, and TIBA at 50 mum after 48 h. The average and se of three separate experiments are reported. Asterisk, P < 0.005 as judged by Student's t test, showing a statistical difference from BA-treated control. To verify that the effects of IAA and NPA were not simply due to the weak acid character of these compounds, additional experiments were performed with benzoic acid (BA) as a control for nonspecific weak acid effects. In the presence of 50 mum BA, 85% of the embryos developed normally as shown in Figure , with a single unbranched rhizoid, which is similar to ASW controls. In the presence of either NPA or another auxin transport inhibitor, TIBA, at the same concentrations, the number of embryos with branched or multiple rhizoids increased. The P value of the Student's t test comparing the frequency of altered embryo development in controls (BA treatments) with the NPA treatments was P = 0.0026 and with the TIBA treatments was P = 0.0033, indicating that the increases in altered developmental patterns due to both treatments were significant. Also, the effect of the solvents in which auxin and auxin transport inhibitors were dissolved were examined. In Table , controls contain similar concentrations of ethanol or DMSO as in the treatments with IAA or NPA, indicating that the solvents lead to significantly fewer embryos with altered rhizoid formation. NPA and IAA Alter F. distichus Embryo Development in a Dose-Dependent Fashion | The numbers of embryos with altered rhizoid formation and branching in response to a range of IAA and NPA concentrations were quantified and results from a representative experiment are reported in Figure . IAA led to a dose-dependent and more than 5-fold increase in the number of rhizoids. The auxin transport inhibitor, NPA, caused a dose-dependent increase in the number of branched rhizoids, with greater than a 3-fold increase. The concentration of IAA at which 50% of the embryos had multiple rhizoids was 31.6 mum and 50% of embryos were branched at 49.5 mum NPA. Again, NPA also increased the embryos with multiple rhizoids, but with complex concentration dependence and with effects on 25% or fewer embryos. The NPA concentrations that alter development were similar to those that affected IAA accumulation. Maximal effects on IAA accumulation were observed at 50 mum NPA, whereas the IC50 for formation of branched rhizoids was 49.5 mum NPA. In addition, the IAA and auxin transport inhibitor concentrations that caused F. distichus embryo alterations are at the high end of the range of concentrations found to alter land plant embryo patterns, in which IAA concentrations in the range of 1 to 40 mum were effective and NPA concentrations of 1 to 40 and 1 to 80 mum NPA were effective . Figure 5 | IAA and NPA led to dose-dependent changes in F. distichus IAA and NPA led to dose-dependent changes in F. distichus embryo development. The percentage of the total number of embryos with multiple or branched rhizoids after culture for 48 h in ASW containing either NPA or IAA is plotted as a function of concentration of added compound. Results from a representative experiment are shown that was repeated three times. IAA and NPA Affect F. distichus Development before the First Cell Division | Because the previous experiments involved growth of embryos in the presence of IAA and NPA for 48 h, it was important to determine more precisely when these compounds were effective. Initial experiments indicated that 12, 24, and 48 h of exposure to IAA and NPA led to similar effects (data not shown). To define the window of time in which NPA and IAA exert their effect, zygotes were treated for 1-h windows during the first 14 h AF. Embryos were placed in ASW and then moved to ASW containing NPA at the indicated times AF and then returned to ASW after a 1-h incubation. The results from this experiment are shown in Figure A and indicate that a treatment for 4 to 5 h AF led to the maximal effect for both NPA and IAA and that treatments with a 1-h duration are sufficient to exert these effects. Figure 6 | Zygotes are sensitive to NPA and IAA treatment early in development. Zygotes are sensitive to NPA and IAA treatment early in development. A, The number of multiple or branched embryos after treatments with 50 mum IAA or NPA, respectively, for 1 h at various times during the first 15 h AF. B, The effect of IAA and NPA treatments for 30 min during the first 6 h AF on multiple and branched rhizoids, respectively. The average and se of three or four separate experiments are reported in A and B, respectively. To more precisely define the time of embryo response to IAA and NPA, embryos were treated with IAA or NPA for 0.5 h between 2 and 6 h AF and the effects on embryo development are reported in Figure B. IAA maximally induces multiple embryos between 3.5 and 5 h AF. The effect of NPA on formation of branched rhizoids is slightly delayed and the maximal effect is between 4 and 5.5 h AF. This difference in timing could be developmentally significant because IAA affects rhizoid initiation, whereas NPA affects the later process of rhizoid branching. The timing of the effects of NPA and IAA are similar to the 4 to 5 h AF required for selection of a light-dependent axis in F. distichus embryos and are well before the 10 to 12 h AF required for embryos to become committed to a developmental polarity and the 24 h needed for germination that were reported previously and experimentally confirmed in this laboratory (data not shown). IAA and NPA Alter Position of Planes of Division in F. distichus Embryos | Because NPA and IAA exert their effect early in development, the possibility that these compounds alter the orientation of division planes was examined. F. distichus embryos were treated with IAA and NPA for durations of 12, 24, or 48 h, beginning 2 h AF, and then stained with fluorescein diacetate (FDA), which facilitates the localization of division planes , as shown in Figure . The control embryos grown in ASW for the 12-h treatment period had germinated, but cell division had not yet occurred (Fig. A). By 24 h, most embryos had undergone one division, resulting in formation of thallus and rhizoid cells (Fig. B), whereas at 48 h, multiple divisions have occurred, resulting in single unbranched rhizoids (Fig. C). Treatment with IAA (Fig. , D --F) resulted in germination of multiple rhizoids before the first cell division, as well as a random orientation of division planes and development of a rhizoid from several of the thallus cell divisions. Most of the embryos treated with NPA also had altered orientation of division planes but with less serious alterations, but again these division plane alterations followed the formation of branched rhizoids (Fig. , G --I). Solvent controls treated with 0.5% (v/v) ethanol and DMSO were similar to those treated with ASW (data not shown). These results indicate that alterations in rhizoid initiation and branching precede cell division. Figure 7 | IAA and NPA alter position of division planes in F. distichus IAA and NPA alter position of division planes in F. distichus embryos. F. distichus embryos were stained with FDA and transverse (parallel to the coverslip) views of division planes in ASW are shown for untreated embryos for 12 (A), 24 (B), and 48 (C) h or with embryos treated with 50 mum IAA for 12 (D), 24 (E) or 48 (F) h, or embryos treated with 50 mum NPA for 12 (G), 24 (H), or 48 (I) h. Size bars = 50 mum. DISCUSSION : The goal of these experiments was to explore the role of auxin and auxin transport in the development of apical basal polarity. F. distichus is an ideal system for these experiments because it is possible to obtain thousands of synchronized embryos in which the ability of the auxin, IAA, and the auxin transport inhibitor, NPA, to alter development can be quantitatively examined. Because these brown algae are relatively distantly related to land plants, it was critical to first determine if these algae contain IAA and whether there is evidence for carrier-mediated influx and efflux of IAA. IAA was detected in both zygotes and mature tissues of F. distichus with an abundance of free IAA that is similar to, but slightly lower than that found in land plants and other green plants . The identification of IAA by GC-MS is consistent with the previous identification of compounds with auxin activity in F. distichus using bioassays . In addition, auxin movements in F. distichus embryos were measured by examination of [3H]IAA accumulation into embryos. IAA accumulation was found to increase upon treatment with NPA. Because NPA acts specifically at the site of auxin efflux , these results are consistent with an auxin efflux carrier with similar pharmacology to those found in land plants. The inability of excess unlabeled IAA to reduce [3H]IAA accumulation into embryos suggests that IAA uptake is not mediated by a protein carrier in F. distichus zygotes, but may rather occur through passive diffusion of a protonated and hydrophobic IAA molecule. Although both uptake and efflux proteins are known to function in land plants, the possibility that one set of carriers is more highly conserved evolutionarily has not been experimentally tested. A number of investigators have surveyed diverse groups within the plant kingdom to test for the presence of auxin metabolism and transport (; ; for review, see ). IAA has been demonstrated convincingly to be present using GC-MS in a number of species including Charophytes, mosses, and liverworts, but with differences in IAA metabolism between plant groups . In addition, there is evidence for the presence of polar auxin transport in bryophytes, mosses, and the green alga (for review, see ). The brown algae, of which F. distichus is a member, are evolutionarily quite distant from the other algae and land plants. The brown algae are part of Kingdom Phaeophyta, which are in the Stramenophiles clade, which is distinct from the clade including Plantae . Therefore, it is somewhat surprising that there are apparently similar mechanisms of auxin efflux used in brown algae and land plants, but this finding does suggest the universal importance of auxin transport in plants and algae. Because F. distichus zygotes contain free IAA from the earliest stages AF and have evidence for IAA efflux carrier-mediated transport, then it was reasonable to examine the role of auxin in F. distichus embryo development. Treatment of embryos with either exogenous IAA or the IAA efflux inhibitor, NPA, led to altered rhizoid formation, but with two different characteristics. Multiple rhizoids were evident upon treatment with IAA, with four or more randomly distributed rhizoids often formed. When both IAA and NPA were added simultaneously, the number of embryos with multiple rhizoids matched that found with IAA alone. This is consistent with a dominant effect of IAA on this process and with external IAA concentrations controlling rhizoid initiation. Similarly, a previous report indicates that culturing of brown algae in the presence of exogenous auxin for 40 d caused increased number of holdfasts, which is the structure that develops from rhizoids . Treatment with NPA led to high numbers of embryos with branched rhizoids, although IAA-treated embryos did not exhibit this phenotype. In fact, treatment with IAA in addition to NPA reduced the number of branched rhizoids to the levels in untreated embryos. This reversal of the NPA effect by IAA is consistent with NPA acting to alter IAA distribution. The effects of NPA and IAA on F. distichus embryo development can only be partially compared with the effects of these compounds reported for embryos of land plants. Although a number of treatments were performed on zygotic embryos, the effects of the treatments were reported on embryos cultured for 1 or 2 weeks on these compounds . The significantly longer exposure to added compounds and the absence of information on the effects on the earliest divisions make it difficult to compare these previous studies with those reported here. Although higher concentrations of auxin and auxin transport inhibitors were required to alter F. distichus embryos, the ability to expose the embryos for less than 1 h and get nearly maximal effects indicates that the embryos are quite sensitive to these compounds. The best comparison of auxin effects on the first cell divisions can be made by examination of Arabidopsis mutants. The bdl (bodenlos) and axr6 mutants have altered auxin sensitivity and show increases in number of embryos with an alteration in the apical basal pattern . In bdl, although 6% of the embryos have a defect normally, this mutation causes that to be elevated to 18% of the embryos . The timing of the effects of IAA and NPA on embryo development was carefully examined. Treatments for 0.5 h indicate that the maximal effects of IAA are between 3.5 and 5 h and for NPA between 4 and 5.5 h AF. It is possible that the developmental significance in the delay in NPA action relative to IAA is due to the effect of IAA on rhizoid initiation and the effect of NPA on later events leading to rhizoid branching. Therefore, it is clear that IAA and NPA exert their effect very early in the developmental sequence. In these single-celled zygotes, IAA movements are detectable across the F. distichus plasma membrane, yet there are clearly no intercellular IAA movements. Therefore, it becomes important to consider the function of IAA efflux at this early developmental stage. First, if polar IAA movement is integrally linked to the polarity of each cell, then it is conceivable that the polarity of IAA efflux is established simultaneously with the development of polarity. Although [3H]IAA accumulation assays cannot provide insight on whether IAA is polarly transported, they do demonstrate the NPA-regulated movement of IAA in these zygotes, consistent with very early auxin transport activity. These results suggest one additional conclusion, which is that the appropriate distribution of IAA is required for establishment of polarity because inhibition of IAA efflux or excess IAA led to alterations in the formation of apical basal pattern. The simplest explanation for this result is that the IAA efflux carrier complex is part of the membrane protein complex that forms at the site of rhizoid initiation to mark this location for growth and that a gradient of IAA across the rhizoid tip is essential to reinforce the signal for rhizoid outgrowth. Consistent with this idea is the evidence that local cell wall loosening precedes rhizoid germination in brown algae . In land plants, cell wall loosening is required for growth and is stimulated by auxin . In addition, the timing of the cell wall loosening in Pelvetia compressa rhizoids, occurring predominantly at 4 to 6 h AF , is later than the period of NPA and IAA sensitivity. The multiple rhizoids that initiate in the presence of excess external IAA are consistent with a delocalization of the rhizoid initiation signal, perhaps mediated by multiple cell wall loosening events. The excess rhizoid branching in the presence of NPA might then arise from a reduction in the strength of the external IAA signal that reinforces the initiation point resulting in a weaker initiation signal and several local rhizoid initiation events. A second important question about these developmental alterations is the role of cell division in this process. In F. distichus embryos, the formation of apical basal polarity precedes the first cell division, in which rhizoid germination is evident by 12 to 14 h, yet division does not occur until approximately 24 h AF. In the case of NPA and IAA, it is clear that the effects of these compounds are on rhizoid formation and that altered cell division is the result, rather than the cause, of these alterations. In support of this conclusion is the evidence that NPA and IAA complete their action in less than 6 h AF , well before the first cell division, and that germination of multiple or branched rhizoids is documented in cells that have not yet undergone division . This conclusion is also supported by , who conclude from their studies with BFA that the cytoskeleton/cell wall complex formed at the site of rhizoid outgrowth directs the rhizoid formation independent of cell division. It is critical to consider the role of IAA in the context of the other signaling events that control embryo development in brown algae. Formation of an F-actin patch has been observed at the site of fertilization of P. compressa embryos and this patch will disassemble and reform at the ultimate site of rhizoid inhibition, if the position of signals that control initiation are shifted . The detection of an actin patch that predicts the site of rhizoid initiation has the best temporal correlation to the window in which NPA and IAA exert their effect, usually being first evident at about 3 h AF in P. compressa . In addition, the actin cytoskeleton has been implicated in the localization of IAA efflux carriers through attachment to one protein that is part of the IAA efflux carrier (; ; for review, see ) and as tracks to deliver vesicles containing auxin transporters (; for review, see ). If elevated IAA concentration on the outside of the zygote is required to reinforce the site of rhizoid initiation, association of efflux carriers with the newly formed actin patch may provide the mechanism to localize the IAA efflux carrier complex, just as it has been suggested to localize channels that move calcium into the embryo and facilitate calcium entry at the rhizoid tip . Finally, consistent with this idea, treatment of land plants with cytochalasin D leads to both a randomization of PIN1, a putative IAA efflux carrier, and to reduction in polar IAA transport . In addition, the connections between auxin and targeted vesicle secretion in the formation of apical basal polarity in embryos should be considered. In both land plants and F. distichus embryos, there is a clear interdependence of these processes and development as well as connections to polar auxin transport. In F. distichus, the F actin patch is believed to mark the position of asymmetric vesicle secretion leading to the deposition of adhesive that facilitates embryo attachment to solid surfaces , as well as providing the vesicles containing membrane components needed for rhizoid tip growth . Treatments with BFA, which blocks vesicle secretion, led to formation of F. distichus rhizoid branches that are strikingly similar to those formed by NPA treatment and, like NPA, BFA leads to loss of gravitropic growth in Arabidopsis . In land plants, mutations in a target of BFA, the GNOM ADP-ribosylation factor-guanine nucleotide exchange factor protein, also led to embryo defects that are very similar to those resulting from IAA efflux inhibitor treatment . The GNOM mutation and BFA treatment also led to randomization of the PIN1 protein, which is a putative IAA efflux carrier . BFA treatments also block IAA efflux from tissue culture cells (for review, see ), suggesting parallel changes in PIN1 localization and transport capacity. These results then demonstrate that inhibition of vesicle transport and IAA efflux cause similar developmental defects and that vesicle transport is necessary for IAA transport. In addition, temporal and spatial control of vesicle movement to specific membranes may be essential for establishment of auxin transport polarity, consistent with the dynamic nature of BFA-dependent PIN1 membrane localization . Yet, it may also be possible that auxin transport is more deeply tied to vesicle secretion. have demonstrated that the IAA efflux inhibitor, TIBA, can perturb the localization of PIN1. Specifically, they report that TIBA will prevent the BFA-induced randomization of PIN1 or the relocalization of PIN1 after BFA removal . Although TIBA does not affect PIN1 localization directly, it interferes with the effect of BFA and, even more significantly, prevents the restoration of PIN1 localization after BFA is removed. Understanding of this aspect of the interaction between IAA efflux and vesicle targeting awaits further experimentation (for review, see ), yet provides an alternative explanation for the similar embryo defects induced by NPA and BFA. In conclusion, this report contains evidence supporting a role for auxin and auxin transport in the early events of F. distichus embryo development. IAA is present in F. distichus embryos at levels similar to those found in land plants, and evidence for an auxin efflux carrier, but not an influx carrier, is reported. Both IAA and auxin transport inhibitors led to alterations in cell division and resulting embryo development when treatments precede the first cell division. Although the function of auxin and auxin gradients in developing embryos of F. distichus is not yet clear, these data suggest that initial events in embryo development are linked to auxin and auxin transport. MATERIALS AND METHODS : Chemicals | [3H]IAA was purchased from Amersham International (Arlington Heights, IL; 25.0 Ci mmol-1). N-1-NPA was purchased from Chemical Services (West Chester, PA). 2,3,5-TIBA and all other chemicals were purchased from Sigma (St. Louis). Fucus distichus Zygote Isolation | Reproductive fronds (receptacles) of sporophytes of F. distichus were collected at South Beach (OR; South Jetty), transported on ice to Winston-Salem (NC), and stored in the dark at 4C for 2 to 3 weeks. The gametes were released into ASW (450 mm NaCl, 10 mm KCl, 9 mm CaCl2, 30 mm MgCl2, 16 mm MgSO4, and 10 mm TES buffer, pH 7.5) and fertilization and development of zygotes were performed at 14C +- 1C in constant illumination with cool-white fluorescent lights at 60 mumol m-2 s-1, as described previously . Fertilization is defined as 15 min after reproductive fronds are placed in the light. Experiments were initiated at 2 h AF, unless noted otherwise. The germination rate for all samples independent of treatments was greater than 85% unless indicated otherwise. Identification of IAA and Quantification of Free IAA Concentrations | Free IAA was purified and quantified using the procedure of in the lab of Jerry D. Cohen (U.S. Department of Agriculture, Beltsville, MD). Between 5 and 15 g fresh weight of F. distichus fruiting tips and 50 to 200 mg of F. distichus zygotes (2 h AF) were used for isolation of free IAA. Tissue was frozen in liquid nitrogen and ground in a mortar and pestle with glass beads and IAA extraction buffer (65% [v/v] isopropanol and 35% [w/v] 0.2 m imidazole buffer, pH 7). [13C]-IAA (20 ng) was used as an internal standard and 50,000 dpm of [3H]IAA was used as a radiotracer. Extraction was for 1 h at 4C and the extract was subjected to centrifugation at 10,000g for 10 min. The supernatant fluid was then analyzed for free IAA. IAA was purified by an amino column, with several organic washes and eluted in methanol with 5% (v/v) acetic acid. After concentration, the sample was purified by HPLC, methylated using ethereal diazomethane, and then analyzed by GC-MS. The GC-MS was used for a TIC of F. distichus fruiting tip extracts for identification of IAA and for selected ion monitoring to quantify the free IAA concentrations in extracts from both F. distichus fruiting tips and zygotes. [3H]IAA Accumulation | Zygotes in ASW at a density of 40,000 zygotes mL-1 were transferred immediately after shedding into Multiscreen HA filtration plates (Millipore, Bedford, MA). The zygotes were incubated in 25 nm [3H]IAA in a total assay volume of 200 muL, in the absence and presence of NPA with IAA at the indicated concentrations in ASW with MES substituted for TES, pH 5.5, at 14C +- 1C in the lighted incubator for 1 h. [3H]IAA accumulation was then determined by filtration to remove the excess [3H]IAA and washing with 200 muL of ASW-MES. The filters were punched out and soaked for 30 min in 250 muL of water and 2.5 mL of scintillation fluid was added. Samples were counted in an LS 6500 scintillation counter (Beckman Instruments, Fullerton, CA) for 2 min. Background was determined by addition of 25 nm [3H]IAA to zygotes immediately before filtration and was subtracted from all the values. Effects of IAA and Auxin Transport Inhibitors on Embryo Development | Aliquots of F. distichus zygotes (2.5 h AF) in ASW at a density of 2,000 zygotes mL-1 were placed in tubes and allowed to settle. ASW was removed and the zygotes were resuspended in ASW containing IAA, NPA, TIBA, or BA at 50 mum or a range of concentrations. Slides were coated with poly-l-Lys and baked at 100C for 15 min and cooled to room temperature. For each treatment, 60 muL of the solution containing zygotes was transferred to a glass slide under normal laboratory lighting conditions, and placed horizontally in a petri dish containing the same solution. The petri dish was placed in the incubator at 14C in the dark for 48 h. Images were obtained with an Axioplan microscope (Zeiss, Jena, Germany) equipped with an Orca cooled CCD camera (Hamamatsu Corporation, Bridgewater, NJ). The percentage of embryos with more than one rhizoid or with branched rhizoids was determined after 48 h under the dissecting scope. Examination of the Temporal Sensitivity to NPA and IAA Treatment | To determine when NPA and IAA exert their maximal effect, zygotes at 2 h AF were treated for 1 h in the dark in ASW containing 50 mum of either IAA or NPA and then slides were transferred to untreated ASW for the remainder of 48 h, after which branched and multiple rhizoids were quantified. At least 200 embryos per treatment were scored and each experiment was replicated at least three times. The average and se of these three experiments are provided. In a second experiment, embryos were placed in ASW and then transferred to ASW containing 50 mum of either NPA or IAA for 30 min. At the end of this 30-min treatment, slides were transferred back into untreated ASW after a brief rinsing with ASW. The numbers of branched embryos in the presence of NPA or embryos with multiple rhizoids were determined and the numbers of altered embryos are reported as a function of the time of the treatment. Effect of IAA and NPA on Orientation of Division Planes in F. distichus Embryos | Microscope slides were coated with poly-l-Lys as described above. At 2 h AF, three aliquots of 100 muL of ASW with F. distichus zygotes at a density of 2,000 zygotes mL-1 were added on the coated portion of the slides, allowed to fix for 1 min, and immediately placed in ASW supplemented with NPA or TIBA at 50 mum under normal laboratory lighting conditions. The slides containing embryos were then placed at 14C in the dark for either 24 h or 48 h. ASW containing DMSO or ethanol was used for controls at the same concentrations as in the treatments (0.5% [v/v]). After the treatments, the embryos were stained with FDA at 1 mug mL-1 according to for 7 min in the dark and fluorescent images were captured by a Zeiss Axioplan microscope equipped with a Hamamatsu Orca cooled CCD camera. Backmatter: PMID- 12226510 TI - Sterol C-24 Methyltransferase Type 1 Controls the Flux of Carbon into Sterol Biosynthesis in Tobacco Seed AB - The first committed step in the conversion of cycloartenol into Delta5 C24-alkyl sterols in plants is catalyzed by an S-adenosyl-methionine-dependent sterol-C24-methyltransferase type 1 (SMT1). We report the consequences of overexpressing SMT1 in tobacco (Nicotiana tabacum), under control of either the constitutive carnation etched ring virus promoter or the seed-specific Brassica napus acyl-carrier protein promoter, on sterol biosynthesis in seed tissue. Overexpression of SMT1 with either promoter increased the amount of total sterols in seed tissue by up to 44%. The sterol composition was also perturbed with levels of sitosterol increased by up to 50% and levels of isofucosterol and campesterol increased by up to 80%, whereas levels of cycloartenol and cholesterol were decreased by up to 53% and 34%, respectively. Concomitant with the enhanced SMT1 activity was an increase in endogenous 3-hydroxy-3-methylglutaryl coenzyme A reductase activity, from which one can speculate that reduced levels of cycloartenol feed back to up-regulate 3-hydroxy-3-methylglutaryl coenzyme A reductase activity and thereby control the carbon flux into sterol biosynthesis. This potential regulatory role of SMT1 in seed sterol biosynthesis is discussed. Keywords: Introduction : Sterols are ubiquitous in plant cells, where they serve crucial functions to control the fluidity and permeability of membranes and as precursors to steroid growth regulators such as brassinosteroids . The sterol pathway is divided into two distinct parts . First, acetate is converted into squalene epoxide via several steps in the mevalonate pathway. Second, in the first committed step of sterol biosynthesis, squalene epoxide is cyclized to give cycloartenol, which is transformed into end product sterols in a series of enzyme catalyzed methylations, demethylations, and desaturations . In contrast to animals, higher plants contain mixtures of 24-alkyl-Delta5-sterols, of which campesterol is the principal 24-methyl sterol, and sitosterol and stigmasterol are the predominant 24-ethyl sterols . It has been proposed that the C-24 methylation of cycloartenol is a major site of regulation in the sterol biosynthetic pathway . Important evidence supporting this view is that transgenic tobacco (Nicotiana tabacum) leaf-overexpressing 3-hydroxy-3-methylglutaryl CoA reductase (HMGR) accumulates high levels of cycloartenol, but only relatively moderate levels of 24-alkyl-Delta5-sterols . Figure 1 | Schematic representation of the sterol pathway post-squalene. Schematic representation of the sterol pathway post-squalene. Solid lines indicate that one enzymatic step is involved in the conversion and dashed lines indicate the involvement of more than one enzyme. The conversion of cycloartenol into 24-methylene cycloartenol is principally catalyzed by an S-adenosyl-l-Met-dependent sterol C-24 methyltransferase type 1 (SMT1). Genes encoding SMT1 enzyme have been cloned from yeast (Saccharomyces cerevisiae) and a range of plants such as Arabidopsis, tobacco, Ricinus communis, soybean (Glycine max), and Oryza sativa . SMT1 preferentially catalyzes the methyl addition to cycloartenol, but it can also catalyze the second methylation step, namely the conversion of 24-methylene lophenol to 24-ethylidene lophenol, albeit at a much lower rate . Conversely, sterol-C24 methyltransferase 2 (SMT2) is mainly responsible for the second methyl addition, but it can also catalyze the first sterol methylation . Therefore, it was not a surprise that an Arabidopsis SMT1 deletion mutant was still able to accumulate alkylated sterols, but at altered levels . The perturbed leaf sterol composition of the SMT1 mutant was associated with several phenotypical abnormalities such as poor growth and fertility, a loss of proper embryo morphogenesis, and sensitivity of the root to calcium . This emphasizes that perturbed sterol composition in vegetative tissue can have severe consequences for plant development. To date, all effort has been devoted to elucidate the function of SMT1 in vegetative plant tissues in relation to sterol composition and plant development . However, little is known of the role of SMT1 in sterol biosynthesis in seed tissue. An effective way to study the function of genes in plants is to generate transgenic plants, which overexpress the gene under investigation. In this report, we investigate the function of SMT1 in sterol biosynthesis in tobacco seed by overexpressing SMT1 under control of either a constitutive or a seed-specific promoter. Herein, we present the metabolic effects that overexpression of SMT1 in tobacco has on SMT enzyme activity, sterol composition, and total sterol content in seed tissue. Furthermore, we also show that the activity of one key enzyme in sterol biosynthesis, HMGR, is up-regulated in seed tissue of plants overexpressing SMT1. RESULTS AND DISCUSSION : Overexpression of SMT1 in Tobacco | The Ntsmt1-1 gene, encoding SMT1, was cloned from a pool of tobacco cDNAs as described in "Materials and Methods." Plant expression vectors were constructed by placing Ntsmt1-1 under control of either the constitutive carnation etched ring virus (CERV; pNH7) or the seed-specific truncated acyl-carrier protein (ACP) promoter (pNH19; Fig. ). The CERV promoter has similar strength and developmental regulation pattern as the more commonly used enhanced cauliflower mosaic virus 35S promoter (M. Harker, N. Holmberg, J.C. Clayton, C.L. Gibbard, A.D. Wallace, S. Rawlins, S.A. Hellyer, A. Lanot, and R. Safford, unpublished data). The ACP promoter is active during the second and third stages of rape seed development, when sterol biosynthesis occurs (; M. Harker and A. Lanot, unpublished data). Figure 2 | Binary plant transformation vector pNH7 and pNH19 holding the CERV-Ntsmt1-1-nopaline synthase (NOS) and ACP-Ntsmt1-1-NOS expression cassettes, respectively. Binary plant transformation vector pNH7 and pNH19 holding the CERV-Ntsmt1-1-nopaline synthase (NOS) and ACP-Ntsmt1-1-NOS expression cassettes, respectively. LB, Left border; RB, right border. Arrows indicate the orientation of the Ntsmt1-1 and the hygromycine resistance genes. The size bar indicates 1,000 bp. Binary vectors pNH7 (CERV-Ntsmt1-1-NOS) and pNH19 (ACPp-Ntsmt1-1-NOS), and the empty vector control, pSJ35, were transformed into tobacco. Thirty individually transformed NH7 and NH19 lines were generated and screened for SMT1 activity in leaf and seed tissue. As demonstrated in Figure , A and B, transgenic NH7 lines exhibited up to 4.3-fold more SMT1 activity in leaf and up to 7.4-fold more activity in seed tissue than the average of control plants. Although the overall trend of increased activity in the population of leaf samples was mirrored in the seed samples, there was only a loose correlation between the SMT1 activity in leaf and seed tissue. The transgenic NH19 lines also demonstrated significant increases in SMT1 activity in seed tissue (Fig. C). Plants NH19:13 and NH19:27 exhibited up to 3.5-fold more SMT1 activity than the average of control plants. Seeds of selected high-expressing plants were germinated on hygromycin and grown on to the subsequent generation (T1). Figure 3 | The SMT1 activity of T0 transgenic and control lines. The SMT1 activity of T0 transgenic and control lines. Transgenic lines are indicated in black staples and control in white staples. The average activities of four independent tobacco SR1 control lines are given and the sds are indicated with error bars. A, The SMT1 activity in expanding leaves transformed with pNH7 (CERV-Ntsmt1-1-nos). B, The SMT1 activity in seeds collected 17 d after anthesis of tobacco transformed with pNH7. C, The SMT1 activity in seeds collected 17 d after anthesis of tobacco transformed with pNH19 (ACP-Ntsmt1-1-nos). Genetic Characterization of SMT1-Expressing Lines | Transgenic NH7 and NH19 lines in the T0 generation, which exhibited high SMT1 activity, were subjected to Southern-blot analysis to deduce the number of transgene integration loci because previous reports have indicated that multiple copies may lead to cosuppression . As shown in Table , the copy number varied in the transgenic lines. Surprisingly, there was not a clear correlation between the insertion of the transgene into a single locus and high SMT1 activity. Lines exhibiting high SMT activity had the transgene inserted into one, two, or four genomic loci (Table ; Fig. ). Table I | No. of transgene insertion loci in SMT1-expressing tobacco lines Segregation of the transgene in T1 lines NH7:27 and NH19:27 was assessed by germinating 25 seeds on hygromycin-containing media. NH7:27 and NH19:27 exhibited Mendelian segregation ratios of 3:1 (chi2 = 0.26, P > 0.9) and 4:0 (chi2 = 0.04, P > 0.99), respectively (data not shown). This observation correlates well with the Southern-blot data, which demonstrated that NH7:27 contains a single transgene insertion locus and that NH19:27 contains several insertion loci . Enhanced Carbon Flux into End Product Sterols in Seed But Not in Leaf Tissue | The sterol level and composition of leaf tissue from five individual T1 NH7 lines was analyzed . In agreement with and , we did not observe any significant changes in the amounts of total sterols in leaf tissue of SMT1-overexpressing plants. Moreover, overexpression of SMT1 in tobacco did not lead to an altered visual phenotype, despite a perturbed sterol composition, which is also in accord with both and . In contrast, transgenic plants overexpressing SMT2 were shown to have an altered campesterol to sitosterol ratio, which resulted in a stunted growth phenotype . Table II | Sterol composition of leaf from T1 tobacco transformed with CERV-Ntsmt1-1 (mug g dry wt-1) The sterol composition of SMT1-overexpressing leaf tissue, however, was altered, with cycloartenol and cholesterol levels dramatically reduced . The level of cycloartenol was reduced from approximately 50 mug g-1 dry weight to non-detectable levels (<2 mug g-1 dry weight) in four of five lines and the level of cholesterol was reduced by up to 5.4-fold as compared with wild-type lines. These decreases in cholesterol and cycloartenol levels are more striking than those observed by when expressing the tobacco SMT1 in tobacco, and by when expressing soybean SMT1 in tobacco. A possible explanation for the more pronounced decreases in cycloartenol and cholesterol in the present study is that higher levels of SMT1 activity were achieved. The increases in SMT1 activity achieved by and were up to 54% and 63% over wild-type levels, respectively. These are moderate effects when compared with the 5-fold increases in SMT1 activity observed in the NH7 lines (Fig. A). Although it should be pointed out that the SMT1 assays were not performed in an identical fashion, , for instance, used the true SMT1 substrate, cycloartenol, when assaying the SMT1 activity, whereas lanosterol was used in the current study and by . However, it is unlikely that the use of lanosterol as substrate would give rise to higher SMT1 activity. In contrast to leaf, seed of transgenic tobacco overexpressing SMT1, under control of either the CERV or ACP promoter, exhibited elevated total sterol levels (Tables and ). The total sterol level was 27% and 44% higher in the NH7:27 and NH19:27 T1 lines, respectively. Interestingly, the relative decrease of cycloartenol (percent of wild type level) was less dramatic in the seed tissue than in the leaf tissue of the NH7 lines (Tables and ). However, it should be pointed out that the absolute amount of cycloartenol was higher in wild-type seed than leaf (approximately 300 versus approximately 50 mug g-1 dry weight) and that the absolute decrease in cycloartenol in the NH7 lines was greater in seed than in leaf (approximately 200 versus approximately 50 mug). Moreover, the cholesterol and cycloartenol levels were reduced in NH7:27 by 25% and 53%, respectively, as compared with wild-type values . Similar decreases in cholesterol and cycloartenol were observed in the seed of NH19:27 (34% and 26%; Table ). The absolute decrease in cycloartenol levels was greater in line NH7:27 than in line NH19:27 (184 versus 73 mug; Tables and ). In addition, the levels of 24-methylene and 24-ethylidene sterols, downstream of cycloartenol, were significantly increased. Another observation was that the increase in 24-methylene lophenol was greater than the increase 24-ethylidene lophenol in the seed of NH7:27 and NH19:27. It has been shown previously that SMT1 can convert 24-methylene lophenol into 24-ethylidene lophenol, albeit at a very low rate . Therefore, one would predict that SMT1 would alter the balance in favor of 24-ethylidene lophenol, but this was shown not to be the case. The slight differences in sterol profile observed between NH7:27 and NH19:27 might be explained by minor differences in temporal regulation afforded by the CERV and the ACP promoters. Table III | Sterol composition of seed from T1 tobacco transformed with CERV-Ntsmt1-1 (pNH7; mug g dry wt-1) Table IV | Sterol composition of seed from T1 tobacco transformed with ACP-Ntsmt1-1 (pNH19; mug/g dry wt-1) The end product sterols, isofucosterol, campesterol, and sitosterol, were all significantly increased in the seeds of NH7:27 and NH19:27 tobacco (Tables and ), but levels of stigmasterol were unaltered in seed as well as leaf tissue (Tables --). The latter finding is easy to explain in leaf tissue because the level of the immediate precursor of stigmasterol, sitosterol, is unaltered. However, the sitosterol level in seed tissue was increased by 30% to 50%, without impacting on the level of stigmasterol, suggesting that the C-22 desaturase that catalyzes the conversion of sitosterol into stigmasterol is a critical slow step in seed tissue (Tables and ; M. Harker, N. Holmberg, J.C. Clayton, C.L. Gibbard, A.D. Wallace, S. Rawlins, S.A. Hellyer, A. Lanot, and R. Safford, unpublished data). The increased levels of end product sterols cannot simply be explained by the conversion of cycloartenol into the concomitant alkylated sterols because cycloartenol is only a minor intermediate (approximately 8% of total sterols in seed). An alternative explanation is that the decreased levels of cycloartenol and/or cholesterol levels feed back to enzymes upstream of SMT1, up-regulating enzyme activities and enhancing carbon flux into the sterol biosynthetic pathway. One potential target enzyme is HMGR, which has been shown to control the flux of carbon into sterol biosynthesis . Therefore, we measured the HMGR activity in the seed of SMT1-overexpressing plants as described below. Increased Endogenous HMGR Activity in Seed with Reduced Levels of Cycloartenol | HMGR activity and transcript levels were measured in seed tissue of two high T1 sterol lines (NH7:27 and NH19:27). As shown in Figure A, lines NH7:27 and NH19:27 have elevated HMGR activity, up to 2.7- and 3.6-fold, respectively. Moreover, these lines exhibit not only elevated Ntsmt1-1 transcript levels, which are due to the transgene expression, but also elevated levels of endogenous hmgr transcripts (Fig. B). The relative differences in the increases of hmgr and Ntsmt1-1 transcript levels between the NH7:27 and NH19:27 lines are most likely due to differences in the promoter strength and temporal expression patterns of the CERV and ACP promoters (Fig. B). Figure 4 | A, Endogenous HMGR activity in the seed of T1 transgenic tobacco lines NH7:27 (CERV-Ntsmt1-1-nos) and NH19:27 (ACP-Ntsmt1-1-nos) collected 14 d after anthesis. A, Endogenous HMGR activity in the seed of T1 transgenic tobacco lines NH7:27 (CERV-Ntsmt1-1-nos) and NH19:27 (ACP-Ntsmt1-1-nos) collected 14 d after anthesis. The average HMGR activity of SR1 tobacco was calculated from the activities of seed samples of four individual plants. The error bar corresponds to the sd. B, Real-time PCR to detect the Ntsmt1-1 and the Nthmgr transcript levels in transgenic seed of NH7:27 and NH19:27. The most obvious conclusion from these experiments is that the overexpression of SMT1 reduces the level of cycloartenol, which in turn feeds back to up-regulate the activity of HMGR. The elevated transcript levels suggest that up-regulation of HMGR activity occurs via transcriptional control, but it cannot be excluded that HMGR is regulated on other levels as well. Concluding Remarks | It has been shown in several studies that the activity of HMGR is important in regulating the carbon flux from acetate to sterols (e.g. ; ). Deregulation of HMGR activity in tobacco via gene overexpression demonstrated that the intermediate cycloartenol overaccumulated (85-fold), whereas the end product sterols (e.g. sitosterol) increased to a much lesser extent (2-fold; ). Therefore, it was proposed that alkylation of cycloartenol is another key regulatory point in sterol biosynthesis . Yet, when SMT1 or SMT2 were overexpressed in tobacco a concomitant increase of total sterol could not be detected in leaf tissue . However, it has been shown that an Arabidopsis smt1 deletion mutant accumulated elevated levels of cholesterol in leaf tissue, which indicates that SMT1 plays an important regulatory role . In accord with , we demonstrated that overexpression of SMT1 reduces the level of cholesterol to undetectable levels in leaf and by about 35% in seed. This raises the interesting possibility to produce vegetables and plant oils with reduced levels of cholesterol. We have demonstrated for the first time, to our knowledge, that overexpression of SMT1 leads to an increase in sterol accumulation in seed, while also confirming the lack of effect on sterol levels in leaf tissue. Hence, we conclude that the carbon fluxes into the sterol pathways in leaf and seed tissue are regulated via different mechanisms. Although overexpressing SMT1 in seed only gave rise to modest absolute decreases in cycloartenol levels, greater significant increases in total sterol levels were observed (Tables and ). Furthermore, we also detected increases in endogenous HMGR activity in the SMT1-overexpressing tobacco . Therefore, we hypothesize that lowering the level of the key intermediate, cycloartenol, feeds back and up-regulates HMGR activity to increase the carbon flux to cycloartenol. However, it cannot be ruled out that lowering the cholesterol level also affects HMGR activity. Furthermore, the elevated hmgr transcript levels suggest that transcriptional regulation is involved. The question of why sterol biosynthesis is regulated via different mechanisms in seed and leaf tissues remains unanswered. However, it has been shown that germinating seed lacks an active de novo sterol pathway, but important changes in sterol composition still occur during germination and seedling development . Therefore, it is plausible that the reduced cycloartenol level, imposed by SMT1 overexpression, enhances the carbon flux as a compensatory mechanism to ensure that the seed has sufficient amounts of sterols for efficient germination and seedling development to occur. However, more research is clearly required to further elucidate the precise mechanism by which HMGR activity is up-regulated when SMT1 is overexpressed. MATERIALS AND METHODS : Strains, Plasmids, Media, and Culture Conditions | Escherichia coli strain DH5alpha (Invitrogen, Carlsbad, CA) was used as the host strain in all cloning procedures. Binary vector pSJ35 was created by filling in the BamHI restriction site of the pGPTV-HYG with the Klenow enzyme . Plasmid pTZ5BS holding the Brassica napus ACP promoter and the ACP5 gene was previously described by . Predigested vector pGEM-T Easy containing poly-T overhangs was obtained from Promega (Madison, WI) and used according to the supplier's recommendations. Bacteria were cultivated in Luria-Bertani medium (10 g L-1 tryptone, 5 g L-1 yeast [Saccharomyces cerevisiae] extract, and 5 g L-1 NaCl) supplemented with ampicillin (100 mug mL-1) on a rotary shaker (210 rpm) at 37C. Plant Material | Tobacco (Nicotiana tabacum) cv SR1 Petite Havana was grown in either Murashige and Skoog medium or a compost:perlite mixture (2:1 [v/v]; ). The temperature in the growth rooms was kept at 22C, and a day/night cycle of 16 h/8 h was used. The light intensity was 40 mumol m-2 s-1. Oligonucleotide Synthesis | All oligonucleotides were synthesized by MWG (Milton Keynes, UK) and they are compiled in Table . Table V | Oligonucleotide primers (given in 5' to 3' direction) Cloning of Ntsmt1-1 | Two-week-old seedlings of tobacco cv SR1 Petit Havana were ground up in a mortar and the total RNA was isolated using the Purescript RNA isolation kit from Flowgen (Shenstone, UK). Total RNA (5 mug) was mixed with primer RoRidT17 (10 pmol) in 11.34 muL of diethyl pyrocarbonate-treated water. The mixture was incubated at 68C for 4 min and thereafter placed in wet ice for 2 min. First strand buffer (1x), dithiothreitol (0.1 mumol), RNAsin (22 units), dNTP (20 nmol), and Superscript (200 units) were added to give a final volume of 20 muL. The mix was incubated at 37C for 60 min. The Ntsmt1-1 cDNA (accession no. ) was amplified by gene-specific primers Ntsmt15 and Ntsmt13 using 35 thermal cycles (30 s at 94C, 30 s at 53C, and 90 s at 72C) and a mix of Taq and Pyrococcus furiosus (10% [v/v]) DNA polymerase. The amplification products were separated on a 1.2% (w/v) agarose gel and the fragment corresponding to the full-length Ntsmt1-1 cDNA was excised and ligated into pGEM-T. Plasmid pGEM-T plasmid harboring the Ntsmt1-1 gene was sequenced using an automatic 373 sequencer (Perkin-Elmer Applied Biosystems, Foster City, CA) and the following primers: pUC/M13 forward, pUC/M13 reverse, Seqsmt5, and Seqsmt3. Sequencing demonstrated that the cloned gene contained two base substitutions (T553 to G and G555 to T) as compared with the published Ntsmt1-1 sequence . It is difficult to assess whether these nucleotide substitutions have arisen from infidelity during the reverse transcription or PCR amplification, or if the discrepancies found are due to natural variation among tobacco varieties. cloned Ntsmt1-1 from a cDNA library of 5-week-old tobacco var Xanthi SH6 calli, derived from leaf protoplasts, and we cloned Ntsmt1-1 from a pool of cDNA derived from 2-week-old SR1 (Petit Havana) seedlings. However, because the identical nucleotide substitutions appeared in five individual clones from two different amplification reactions, we concluded that the most likely explanation is that the discrepancies are due to varietal differences. In addition, the two mutations led to the conservative Leu-167 to Val amino acid substitution. The side chains of these amino acids are nonpolar and very similar, and they only differ by one methyl group. Furthermore, alignment of SMT sequences from tobacco, Arabidopsis (accession no. ), Triticum aestivum (accession no. ), yeast (accession no. NP013706), Neurospora crassa (accession no. ), and Pneumocystis carinii (accession no. ) using ClustalW revealed that the amino acids at that position, which corresponds to 167 in NtSMT1, are not conserved. Therefore, it is highly unlikely that that the Leu-167 to Val substitution will result in altered substrate specificity or binding. Construction of Shuttle Vectors | The 308-bp HindIII/BamHI-digested CERV promoter fragment from pSJ103 and the SacI/EcoRI-digested 275-bp NOS from pSJ35 was inserted into pUC19, giving pUCN. A synthetic DNA linker, assembled by annealing oligonucleotides PCR628 and PCR629, was inserted into the BamHI/SacI-digested pUCV, yielding pUCV2. This linker introduces the following restriction enzyme sites (5' to 3'): NcoI, NheI, and MunI. Another DNA linker was assembled thereafter by annealing PCR645 and PCR646. The sequence of this linker was designed to mimic the consensus Kozak sequence from plants (taa acc ATG G), which has been shown to render efficient translation . This linker was inserted into BamHI/NheI-digested pUCV2, yielding shuttle vector pNH1. A synthetic DNA linker, holding a SalI site, was assembled by annealing SalI and SalII and inserted between the EcoRI and XbaI sites of pNH1, yielding pNH2. Another synthetic DNA linker holding an XmaI site, assembled by annealing oligonucleotides XbaI and XmaII, was inserted between the HindIII and ClaI site of pNH2, giving pNH3. Cloning of Promoters | The CERV promoter was isolated by PCR from infected Dianthus barbatus leaf material using the primers CERV1 and CERV2 under standard conditions. The approximately 380-bp fragment (corresponding to nucleotides 6,737 --7,118 of ) was cloned into the TA cloning vector pT7Blue (Invitrogen) with the promoter in the same orientation as the T7 promoter yielding pSJ103. Sequence analysis showed that the isolated CERV promoter differed from the published sequence at several positions (T6790 to C, C6826 to T, A6872 to G, and T6729 to A). The most significant change was a small deletion in a poly(A+) tract in the 5'-untranslated leader (four As versus nine As in the published sequence). These changes probably represent differences in virus isolates rather than PCR errors, but the latter cannot be ruled out. A truncated 289-bp portion of the ACP promoter and 69 bp of the 5'-untranslated region of the ACP5 gene were amplified from vector pTZ5BS using primers ACP5 and ACP3. ACP5 and ACP3 were designed with 5' overhangs holding restriction enzyme sites for ClaI and NcoI, respectively. The amplified promoter fragment was digested with ClaI and NcoI and inserted into ClaI/NcoI-digested pNH3, yielding pNH12. Construction of Ntsmt1-1 Expression Vectors | The 1,040-bp Ntsmt1-1 gene was amplified from plasmid pGEM-T, harboring Ntsmt1-1, by PCR using standard conditions. To facilitate cloning, the primers used, clsmt5 and clsmt3, were designed with 5' overhangs containing AflIII and EcoRI restriction sites, respectively. A mix of Taq and Pyrococcus furiosus (10% [v/v]) DNA polymerase was used to ensure high amplification fidelity during the amplification. The amplification products were separated on a 1.2% (w/v) agarose gel and the 1,040-bp Ntsmt1-1 gene was excised, purified, digested by AflIII and EcoRI, and inserted into vector pNH3 downstream of the CERV promoter and upstream of the NOS terminator giving pNH6. Plasmid pNH6 was sequenced with primers CERV1S, clsmt5, seqsmt5, seqsmt3, clsmt3, and NosAs to confirm its authenticity. The CERV-Ntsmt1-1-NOS cassette was excised from pNH6 using XmaI and EcoRI and subsequently inserted into XmaI/EcoRI-digested pSJ35, giving pNH7 . Vector pNH7 was sequenced using primers CERV1S and NosAs to confirm that the junction between promoter, Ntsmt1-1, and terminator were intact. The AflIII/EcoRI-digested Ntsmt1-1 amplification product was inserted into NcoI/MunI-digested pNH12, rendering pNH14. Plasmid pNH14 was sequenced with primers 181, clsmt5, seqsmt5, seqsmt3, and NosAs to confirm that the PCR amplification had not introduced any mutations. Vector pNH14 was digested by XmaI and EcoRI to excise the ACPp-Ntsmt1-1-NOS expression cassette. This cassette was subsequently inserted into XmaI/EcoRI-digested pSJ35, resulting in binary expression vector pNH19 . Vector pNH19 was sequenced with primers ACP5, Ntsmt15, and NosAs to confirm that the junction regions between promoter, gene, and terminator were intact. Plant Transformation and Growth Conditions | Binary vectors pNH7, pNH19, and pSJ35 were transformed into Agrobacterium tumefaciens LBA4404 using electroporation as described by . A. tumefaciens-mediated transformation of tobacco SR1 was carried out using the leaf disc method as described previously by . Hygromycin (25 mg L-1)-resistant seedlings were screened by PCR to identify transformants using primer pairs TGN1/TGN2 (NH7) and ACP5/TGN2 (NH19) . Antisense primer TGN2 annealed to the coding sequence of Ntsmt1-1, and sense primers TGN1 and ACP5 annealed to the CERV and ACP promoter, respectively. PCR positive transformants were transferred from tissue culture to growth chambers and potted up into a compost:perlite mix (2:1 [v/v]). Pods were tagged and seeds were collected for SMT and HMGR enzyme activity measurements 14 and 17 d after anthesis. Mature seeds were collected after about 12 weeks. Southern-Blot Analysis | Genomic DNA was isolated from leaf tissue of 4-week-old soil-grown primary transgenics using the cetyl-trimethyl-ammonium bromide method as described by . Approximately 10 mug of genomic DNA was digested with EcoRI and separated on a 0.7% (w/v) agarose gel. The DNA fragments were transferred onto a Hybond N+ nylon membrane using capillary transfer as described by . Southern-blot analysis was performed using a DIG-labeled probe directed against the hygromycin resistance marker gene according to the method described in the DIG System Users Guide (Roche Molecular Biochemicals, Mannheim, Germany). SMT1 Activity Assay | All steps were carried out at 4C unless otherwise stated. Approximately 150 mg of tagged seeds (collected 17 d after anthesis), which had been stored previously at -80C, were homogenized in the ratio of 1:10 (w/v) seed:buffer using an Ultra-turrax at maximum speed. The homogenization buffer consisted of: 0.2 m potassium phosphate, pH 7.5; 0.4 m Suc; 10 mm EDTA; MgCl2; 5 mm glutathione; and 40g L-1 polyvinylpolypyrrolidone. Aliquots of homogenate (400 muL) were centrifuged at 1,200g for 5 min at 4C. The supernatant was removed using a syringe and needle and discarded. The lipid and pellet fractions were extracted together with 400 muL of homogenization buffer containing 10 mm CHAPS to extract SMT activity. The tobacco samples were vortexed and left on ice for 20 min. The samples were again centrifuged at 1,200g for 5 min at 4C and the supernatant removed. This process was then repeated. The two detergent extracts were then mixed 1:1 (v/v) and assayed immediately for SMT enzyme activity. Crude extracts of young developing tobacco leaves were homogenized 1:10 in homogenization buffer (same as above) using an Ultra-turrax at maximum speed. After centrifugation for 5 min at 1,200g, the resulting supernatant was immediately assayed for SMT1 activity. The standard assay system consisted of 50 mumS-adenosyl-l-[methyl-14C] Met (57 mCi mmol-1), 125 mum lanosterol emulsified in Tween-80 (final concentration 0.1% [v/v]), and 80 muL of enzyme extract in a final volume of 100 muL. Control assays contained no enzyme extract. The assays were incubated for 1 h at 30C and terminated by the addition of 100 muL of 12% (w/v) KOH in ethanol. Lanosterol and cholesterol, 15 mug each, were added as carriers. The neutral lipids were extracted with hexane (2 x 600 muL) and the combined eluant was evaporated to dryness under nitrogen. The lipid residue was resuspended in 15 muL of toluene. The toluene containing extract (10 muL) was applied to precoated silica thin-layer chromatography (TLC) plates (silica gel 60 F254, Merck, Rahway, NJ). The TLC plates were developed in dichloromethane. The 4-desmethyl and 4,4-dimethyl sterols were visualized with iodine vapor. The band corresponding to lanosterol (4,4-dimethyl sterol) was scraped off into scintillation vials. Liquid scintillation cocktail (Readysafe, Beckman Instruments, Fullerton, CA) was added and radioactivity was measured with a Beckman LS650 scintillation counter. It should be noted that endogenous substrate, in particular cycloartenol, may also be methylated by SMT1 and hence contribute to the measured SMT1 activity. HMGR Activity Assay | All steps were carried out at 4C unless otherwise stated. Homogenization buffer was 0.2 m potassium phosphate, pH 7.5, containing 0.4 m Suc, 10 mm EDTA, 5 mm MgCl2, 5 mm glutathione, and 4 g/100 muL insoluble polyvinylpyrrolidone. Developing seeds from tagged pods (approximately 150 mg), collected 14 d after anthesis, were homogenized in the ratio 1:10 (w/v) seed:buffer using an Ultra-turrax at maximum speed. After centrifugation for 5 min at 1,200g, the lipid and pellet fractions were extracted together with 400 muL of homogenization buffer containing 2% (w/v) Brij. The samples were vortexed and incubated on ice for 20 min. The samples were again centrifuged at 1,200g for 5 min at 4C. This process was immediately repeated. The two detergent extracts were mixed 1:1 (v/v) and assayed for HMGR enzyme activity. The assay system consisted of 100 mm potassium phosphate, pH 7.5; 3 mm NADPH; 20 mum [14C] HMG CoA (30 nCi); and 20 muL of enzyme extract in a final volume of 26 muL. In control assays, the enzyme extract was omitted. After incubation for 15 min at 30C, reactions were terminated by the addition of 5 muL of 6 m HCl. To this mixture, 5 muL of mevalonate lactone (1 mg muL-1) was added to act as the carrier. The samples were left to lactonize for 15 min at room temperature and clarified by centrifugation at 13,000 rpm for 2 min at room temperature. Aliquots of supernatant (10 muL) were applied to precoated silica TLC plates (silica gel 60 F254, Merck). Unreacted substrate was separated from product by developing plates in diethyl ether:acetone (3:1 [v/v]). Mevalonate lactone was visualized with iodine vapor and scraped off into scintillation vials. Radioactivity was measured as described above. Real-Time PCR | Total RNA was isolated from developing tobacco seed (14 d after anthesis) of NH7:27, NH19:27, and SR1 plants using the RNaqueous kit according to the supplier's instructions (Ambion, Austin, TX). The total RNA was treated with DNase to remove any contamination of genomic DNA and subsequently converted into cDNA using the 3' RACE System from Invitrogen. Taqman primer pairs directed against Ntsmt1-1 (TaqN1 and TaqN2), tobacco tac9 actin (TaqA1 and TaqA2), and tobacco hmgr (TaqNtH1 and TaqNtH2) genes were designed using the Primer Express software (PE-Applied Biosystems). These primer pairs were used together with Sybr Green (PE-Applied Biosystems, Foster City, CA) in Taqman PCR reactions to detect transcript levels of Ntsmt1-1, hmgr, and tac9 in transgenic and control samples. The Ntsmt1-1 and hmgr transcript levels in the transgenic tobacco were calculated in relation to the transcript levels in SR1 tobacco according to the manual supplied by PE-Applied Biosystems. Sterol Analysis | Mature seed and leaf tissue was collected and freeze dried. Freeze-dried tissue (approximately 50 mg) was ground with a pestle and mortar. Dihydrocholesterol (250 muL, 0.2% [w/v]) in chloroform was added to act as the internal standard. Sterols were extracted in 5 mL of 2:1 (v/v) chloroform:methanol for 30 min at 80C. The mixture was allowed to cool and then filtered. The extract was reduced to dryness under nitrogen. Transmethylation of the extract was performed by resuspending the extracts in 1 mL of toluene and 2 mL of sodium methoxide (0.5 m). The mixture was maintained at 80C for 30 min and allowed to cool. Boron trifluoride solution (2 mL, 14% [w/v]) was added and the mixture was heated for a further 10 min at 80C. Diethyl ether (3 mL) was added to the mixture when cool, followed by deionized water (5 mL) with vigorous shaking. The ether layer was removed and a second extraction with ether performed. The ether layers were combined and backwashed with 5 mL of water. The ether phase was dried over anhydrous sodium sulfate overnight. Extracts were resuspended in 300 muL of toluene and the free sterols silylated by the addition of 150 muL of 95:5 (v/v) N,O-bis(trimethylsilyl)acetamide:trimethylchlorosilane followed by incubation at 50C for 10 min. Sterol derivatives were analyzed by gas chromatography using a Perkin-Elmer Autosystem XL GC with flame ionization detection and a BPX5 capillary column (25-m x 0.32-mm i.d., 0.25-mum film thickness, and helium 8 pounds per square inch), ex SGE. The temperature program used a fast rise from 80C to 230C (45C min-1), a slow rise from 230C to 280C (4C min-1), and 280C to 355C (20C min-1), and finally 5 min at 355C. Peak areas were calculated automatically using Turbochrom software (PerkinElmer, San Jose, CA). Sterol structures were identified by mass spectrometry. Backmatter: PMID- 12226511 TI - Two New Loci, PLEIADE and HYADE, Implicate Organ-Specific Regulation of Cytokinesis in Arabidopsis AB - In screens for regulators of root morphogenesis in Arabidopsis we isolated six new recessive mutants with irregular cell expansion. Complementation analyses placed the mutations in two loci, PLEIADE (PLE) and HYADE (HYA). Phenotypic analyses revealed multinucleated cells, cell wall stubs, and synchronized cell divisions in incompletely separated cells that are all characteristics of defective cytokinesis. These defects were pronounced in roots and undetectable in aerial organs. In addition, fertility and germination were not affected by the mutations. Thus, the alleles that we have isolated of PLE and HYA suggest that the genes may encode organ-specific components needed primarily during root development. Analysis of microtubule arrays during cell cycle in ple and hya roots indicates that the presence of several synchronized nuclei influences the position of preprophase band, mitotic spindles, and phragmoplasts. The enhanced and synergistic phenotype of PLE/ple.hya/hya seedlings and double mutants point to a role of PLE and HYA in the same process. These mutants provide tools to elucidate the regulation of nuclear cytoskeletal interactions during cell division and cytokinesis. Keywords: Introduction : The Arabidopsis root has a well-defined and simple structure with a radial pattern of single layers of epidermis, cortex, endodermis, and pericycle tissues, which are produced by rings of meristematic initials located at the tip of the root . In the meristematic zone, cells divide transversely giving rise to organized files along the root axis. Root meristems vary in size depending on environmental conditions, and, thus, the number of mitotic cells actively dividing is flexible . In plants conventional somatic cell divisions and in particular cytokinesis differ in some aspects from other organisms . One major difference is that in contrast to the contractile ring, which is used for partitioning the cytoplasm of most animal cells , plant cells divide the cytoplasm by constructing a new wall inside the cell. Furthermore, plants have developed several different mechanisms to execute cytokinesis. Before prophase of conventional somatic cell divisions, plants develop a special microtubule structure, the preprophase band (PPB) at the plasma membrane. The cortical position of the PPB correlates with the location where the cell plate will fuse with the parental cell walls at the end of cytokinesis. In telophase of conventional cell divisions, another special microtubule structure (the phragmoplast, which has similarity to the midbody of animal cells) forms from spindle remnants, microfilaments, and Golgi-derived vesicles between separated chromosomes . In plants, the Golgi-derived vesicles, which contain cell wall material, fuse to form the cell plate. The cell plate expands outward and finally reaches the parental cell wall at points previously marked by the PPB. However, not all plant cells need the formation of a PPB to divide correctly. For example, no PPB has been detected at the onset of the first mitotic division during pollen-grain development. The mechanism by which the cell plate is guided to the parental cell wall is unknown in divisions with or without the PPB. In Arabidopsis, "nonconventional" syncytial cytokinesis occurs during nuclear endosperm development , during male and female meiosis , and during female gametophyte cellularization . In these cases, nuclear divisions and cytokinesis are uncoupled. This leads to the formation of multinucleated/syncytial cells. Syncytial cytokinesis does not involve the formation of PPBs, and the position of the new cell walls is determined by interacting arrays of microtubules that radiate from the nuclear envelope surfaces. The onset of cellularization occurs simultaneously. In endosperm, the MTs form groups of miniphragmoplasts between a pair of nuclei leading to a patchwork of syncytial-type cell plates . From recently isolated mutants and their affected gene products, a picture has begun to emerge of the mechanism and spatial regulation of plant cytokinesis . Characteristics of all cytokinesis mutants are the presence of incomplete cell walls in enlarged cells. Furthermore, the affected cells are multinucleated and/or contain enlarged polyploid nuclei. The mutants can be categorized according to their cell biological defects and their affected genes according to their similarity to biochemically characterized gene products. For example, the pilz mutant embryos and endosperms lack both interphase and mitotic microtubule arrays, indicating that the PILZ gene products are involved in microtubule array formation in conventional and syncytial-type cytokinesis . Related phenotypes were observed in titan mutants . The molecular analysis revealed that TTN1 encodes a regulatory protein known as tubulin-folding cofactor D and that TTN5 encodes a small GTP-binding protein-related ADP ribosylation factor-like protein (ARL2; ; ). Recent experiments showed that ARL2 interacts with tubulin-folding cofactor D in fission yeast (Schizosaccharomyces pombe) and humans . Another group of cytokinesis mutants is defective in vesicle trafficking and cell plate formation . A cytokinesis-specific syntaxin (KNOLLE) and a SEC1 homolog (KEULE) have been isolated and shown to bind to each other . Furthermore, some cytokinesis-defective mutants can be grouped based on their severe deficiencies in the formation of the primary cell wall. For example the CYT1 gene has been cloned, and it encodes a Man-1-phosphate guanylyltransferase . The cyt1 mutants exhibit a 5-fold decrease in cellulose content and hyper-accumulate callose. The deficiency in N-glycosylation in cyt1 indicates that N-glycosylation is required for cellulose biosynthesis, and improper cell wall synthesis leads to cytokinesis defects. A relationship between improper cell wall synthesis leading to cytokinesis defects fits the phenotype of mutations in the KORRIGAN (KOR) gene, which encodes a putative membrane-bound endo-1,4-beta-glucanase. Whereas weak kor/rsw2 mutants have reduced cellulose and increased pectin content , a strong allele exhibits characteristic cytokinesis defects such as aberrant cell plates, incomplete cell walls, and multinucleated cells . Most cytokinesis-related genes were identified from embryo- or endosperm-defective mutants, indicating that some components are shared between the different modes of plant cytokinesis. However, there is evidence that some components of conventional somatic cytokinesis are developmentally regulated. For example, TSO1 is a gene that appears to be involved in cytokinesis primarily during flower development . In addition, although no cytokinesis occurs from the first zygotic division onward of double mutants between KNOLLE and KEULE, the haploid gametophytes are functional . Moreover, several genes have been identified because of their abnormal male gametophytic meiosis and mitosis , but the molecular basis of these genes is still unknown. In this report, we present the genetic identification of two new Arabidopsis loci, PLE and HYA. Mutations in PLE and HYA result in irregularly expanded root cells. At the cellular level, the ple and hya alleles contain partially formed transverse cell walls and multiple nuclei, characteristics of cytokinesis-defective mutants. During cell division, these multinucleated cells divide synchronously and influence the position of microtubule arrays including the PPBs, the mitotic spindle, and the phragmoplasts. The strong root phenotypes of the ple and hya alleles indicate that the genes may encode components required for organ-specific cytokinesis. We discuss the possible mode of action of the PLE and HYA gene products during cytokinesis. RESULTS : Isolation and Genetic Classification | In genetic screens for root morphogenesis mutants of Arabidopsis, six lines were isolated with irregular root expansion phenotypes . All six mutants segregated as single nuclear recessive mutations (Tables and ). Reciprocal complementation analyses revealed that they belonged to two different genetic loci, PLE and HYA. The PLE locus included two ethyl methanesulfonate (EMS) and one T-DNA mutagenized allele, and the HYA locus comprised three T-DNA mutagenized alleles. The chromosomal locations of the two loci were determined by the use of molecular markers. Both loci are located on the bottom of chromosome V and linked with a map distance of 14.5 centiMorgans (cM) to each other. HYA maps between the simple sequence length polymorphism (SSLP) markers AthS0191 and nga129 and PLE between nga129 and the cleaved amplified polymorphic sequences (CAPS) marker ASBII . Table I | Overview of the ple and hya alleles Table II | Complementation tests of the cytokinesis mutants Table III | Segregation analysis of ple and hya alleles Figure 1 | Genetic map of the PLE and HYA loci. Genetic map of the PLE and HYA loci. Around 300 F2 mutants were used for mapping of each locus with the SSLP markers nga129, AthS0191, and TSBI and the CAPS marker ASBII. The genetic distance in cM was calculated from the recombination frequency after . Both loci map to the bottom of chromosome V. They are linked with a calculated distance of 14.5 cM. PLE and HYA Are Involved in Cytokinesis | Typical of these new mutants is an irregularly expanded root. The phenotype ranges from very thick and extremely short roots to elongated roots with a wavy growth pattern and enhanced lateral root initiation . In ple-2 and ple-3, development of aerial organs appears normal as evidenced by microscopic analyses of shoot and floral meristems (see Fig. ), which produce normal leaves and flowers. Furthermore, no problems with fertility were observed in all ple and hya alleles. Ple-1 plants exhibited a smaller rosette and shorter inflorescence. This suggests that the affected genes may be active predominately in roots. Figure 2 | Phenotype of wild-type (WT) and cytokinesis-defective seedlings on vertical nutrient agar medium. Phenotype of wild-type (WT) and cytokinesis-defective seedlings on vertical nutrient agar medium. Mutant and WT seedlings were photographed 12 and 8 d after germination (DAG), respectively. A, WT Col; B, hya-1; C, hya-2; D, hya-3; E, ple-1; F, ple-3; and G, ple-2. The primary roots of the cytokinesis mutants are significantly shorter, irregularly expanded, exhibit a wavy growth pattern and develop more lateral roots. The shoots were indistinguishable from WT under these conditions. Bars in A and B through G = 1 mm. Wavy root growth patterns have been described for agravitropic mutants. To test whether the newly isolated mutants respond to gravity, seedlings were exposed to an altered gravitropic stimulus by rotating the nutrient plates by 90. All mutant roots grew toward the gravitropic stimulus. Therefore, we conclude that the wavy growth behavior is caused primarily by the irregularly expanded root cells. Transverse sections through the differentiation zone of roots revealed that some of the epidermis and cortex cells are grossly expanded . These spherical cells are visible in whole-mount preparations . Although root cells are malformed and the radial organization of different root tissues is more variable, mutant roots have clearly distinguishable cell layers . Sections revealed that cell wall stubs and multiple nuclei are often present ---a characteristic feature of cytokinesis-defective mutants . All root tissues are affected to different degrees with the epidermis and cortex forming larger cells than the endodermis or stele tissues. Cells containing wall stubs appear to differentiate as witnessed by the presence of plasmodesmata and of suberin in endodermal cells . Thus, the cytokinesis defect does not appear to change the differentiated characteristics of the affected cells. Figure 3 | Root morphology of wild type (WT), ple, and hya. Root morphology of wild type (WT), ple, and hya. Whole-mount preparations of cleared WT (A), hya-3 (B), and ple-2 (C) root tips grown on nutrient agar medium using differential interference contrast microscopy. Median longitudinal and transverse sections of primary root tips of hya-3 (D and I) and ple-1 (E and J) resin embedded and stained with basic fuchsine/toluidine. In E and J, every tissue has aberrant cell wall stubs and multiple nuclei. Fresh transverse sections through the differentiation zone of a primary WT (F), hya-3 (G), and ple-2 (H) roots. The surface areas of hya roots are about 3 times and of ple-1 roots 3.5 to 4 times larger than WT. The three tissue layers epidermis, cortex and endodermis are clearly outlined but radially enlarged. H, In ple, the diarch symmetry of the vascular tissues is disrupted. K through M, TEM of ple-1 root cells. Note the darkly stained uneven nucleoli (arrowheads), the nuclear membrane (open arrows), and the incomplete cell walls. Longitudinal sections through cortex (K and N), endodermis (L), and vascular (M) cells. Although the newly developed cell wall is incomplete, suberin lamellas of the casparian stripe and plasmodesmata are incorporated (L and N, arrows). Bars = 50 mum in A through E, 25 mum in F through J, and 1 mum in K through N. Another feature of cytokinesis-defective mutants is their multinucleated phenotype. Visualization of the nuclei in mutants by YO-PRO and 4,6-diamidino-2-phenylindole (DAPI) staining revealed that the giant root cells contain up to 32 nuclei, indicating that these cells undergo karyokinesis without cytokinesis . This multinucleated phenotype is the reason for naming the genes after the stellar constellations, the PLEIADES (PLE) and HYADES (HYA). Figure 4 | Visualization of nuclei in different organs of wild-type (WT) and cytokinesis-defective mutants. Visualization of nuclei in different organs of wild-type (WT) and cytokinesis-defective mutants. The nuclei were either stained with YO-PRO (A --F) or with DAPI (H and I) and visualized with confocal laser scanning microscopy (CLSM) or a fluorescence microscope, respectively. Optical longitudinal sections of inflorescence and floral meristems of wild type (A and D), ple-1 (B and E), and hya-2 (C and F). The two single-layered L1 and L2 cell files are clearly defined. A through C, Inflorescence meristems are flanked by floral meristems in different stages. D through F, Floral meristems at the stage of early sepal development. The primary and lateral root meristems of ple (K and N) and hya (L and O) accumulate multiple nuclei. Root hairs of ple-1 emerge with altered morphology as bifurcation (I) or as large bulges (H). G, Multiple nuclei phenotype in cortex cells of roots of flowering ple-1 mutants grown in soil. Fewer cells with less morphological defects are detectable in differentiated parts of roots of these plants. S, Sepal; Inf, inflorescence meristem; Fl, floral meristem. Bars = 50 mum in A through F, H, I, M, and O and 25 mum in G and J through L. Longitudinal sections demonstrate that apical cells in the root meristems are less affected. Quiescent center and root cap cells rarely contain more than one nucleus. In addition, there is a phenotypic gradient from the outside (epidermis) with the most severely affected cells toward the inside (stele) with only a few multinucleated cells. Although the polarity of giant multinucleated cells is affected, not all of the multinucleated cells have lost their anisotropic expansion . For example, the epidermal cells are still able to differentiate into root hair cells. But most of the multinucleated root hairs exhibit abnormal bulging, produce several tips, or become bifurcated . Previous analyses of cell expansion mutants indicated that growth conditions can modulate the expansion phenotype . By decreasing the Suc concentration in the nutrient plates and by isolating roots of adult plants from the soil, we found that the multinucleated phenotype was attenuated . The number of nuclei in one cell indicates how many divisions were aberrant. Thus, to accumulate 32 nuclei, five rounds of cell divisions are necessary. This suggests that most of the defective cell divisions may occur after germination. Morphometric analyses revealed that the size of mature embryonic root meristem is similar between wild type and the cytokinesis mutants . However, we also identified multinucleated cells in mature embryos of two alleles, ple-1 and ple-3, of which 20.9% (n = 67) and 48.8% (n = 41) contained multinucleated cells in the embryonic roots, in hypocotyls, or in both . As in adult roots some nuclei seem to be larger, suggesting that DNA endoreduplication or nuclear fusion might occur during mitosis. Because the nuclei of multinucleated cells cluster, we were not able to address these hypotheses by measuring their DNA content. Figure 5 | Embryonic phenotype of ple mutants. Embryonic phenotype of ple mutants. Optical sections through mature embryos of ple-3 (A --D) and ple-1 (F --G) mutants stained with YO-PRO. The overall architecture of the embryos is normal with only a few cells in ple-3 being enlarged and containing multiple or giant nuclei (B and C). In some embryos only minor deviations of the architecture could be detected (arrows in D, F, and G). Bar = 25 mum. To analyze the phenotype during cell division, we crossed the mutants with the cycB1;1:CDB:beta-glucuronidase (GUS) marker line, which labels cells in late G2 and mitosis. We have shown that the subcellular localization of the histochemical GUS staining changes during mitosis and discriminates different mitotic phases . In prophase and telophase, the dye precipitates around the nucleus, whereas in meta- and anaphase, it is distributed in the whole cell. In mid-late anaphase the dye is denser on both sides of the metaphase plate and may point to a spindle localization. After completion of cytokinesis, the dye disappears gradually . The histochemical GUS staining in mutant background revealed that mitosis is synchronized in incompletely separated cells. Their shared cytoplasm probably accounts for this synchronization. Furthermore, the subcellular localization of the dye in the prophase and telophase is similar to wild type. Only in ple alleles was the anaphase-specific spindle localization more frequently detectable than in wild-type roots. This might indicate that cell cycle is slowed down in these mutants. Figure 6 | Histochemical staining of mitotic cells of WT, ple, and hya roots. Histochemical staining of mitotic cells of WT, ple, and hya roots. The CYCB1;1:CDB:GUS marker line stains cells from late G2 to the onset of G1 and exhibits also a subcellular localization during mitosis of WT (A --D), hya-2 (E), and ple-3 (F and G) roots. In pro- and telophase, the staining accumulates around the nucleus (A and B), whereas in meta- and anaphase, it is distributed in the whole cell (B through D) with a slightly denser accumulation on both sides of the metaphase plate in mid-late anaphase (C, D, and G). After completion of cytokinesis, the dye disappears gradually. E through G, Cells with multiple nuclei divide synchronously. Bar = 10 mum. Our phenotypic analyses suggest that PLE and HYA may be primarily active in roots. Because the analyzed alleles may not be null mutants, we are aware that some caution needs to be invoked in the interpretation of their spatial-temporal specificity. Microtubule Organization during the Cell Cycle | During the cell cycle, plant microtubules (MT) undergo a series of conformational changes. In interphase of rapidly expanding cells, cortical microtubules are perpendicular to the axis of cell elongation (Fig. A). In late G2 before cell division, the PPB is formed (Fig. , B and C), and its localization correlates with the position and fusion site of the phragmoplast, cell plate, and new cell wall . To characterize further the role of PLE and HYA at the cellular level, we studied MT organization in wild-type and mutant plants. Figure 7 | Immunolocalization of microtubules in wild-type, ple-1, and hya-3 roots. Immunolocalization of microtubules in wild-type, ple-1, and hya-3 roots. A through C, Optical sections through the epidermis of wild-type roots showing cortical MTs (A; black arrow), phragmoplasts (B and C; white arrow), and the MTs of the PPB (arrowhead). The PPB is always accompanied with strong perinuclear MTs (C). D through G, Optical sections through the epidermis of ple-1 roots showing mitotic spindles (stars), dense accumulation of cytoplasmic MTs (D), cortical (E and F), and the perinuclear MTs of the PPB (G). Note the dark gray nuclei in multinucleated cells (D) and the cell wall stubs where MTs seems to nucleate (G). H through K, Optical sections through the epidermis and cortex of hya-3 roots showing slightly misoriented cortical (H, I, and K), perinuclear MTs and PPBs (I), mitotic spindles (J) of synchronized two nuclei containing cortical cells. K, Shows the misaligned phragmoplasts of a multinucleated epidermis cell. Bar = 10 mum. In ple and hya, the helical arrays of cortical MTs are present but misplaced in giant cells with multiple nuclei (Fig. , E, H, and J). Furthermore, in ple, short, diffuse cytoplasmic MTs were detectable in interphase cells (Fig. , D and G). During late G2, the perinuclear MTs stained strongly in both mutants but were not symmetrically distributed around the nucleus in the ple mutants. Although the PPB is present in cells with single nuclei, it is displaced in hya (Fig. I) and is not discernible in multinucleated ple cells (Fig. G). Thus, cells with numerous nuclei do not exhibit an equatorial arrangement of the PPB. During mitosis, the position of spindles and phragmoplasts is also misplaced in multinucleated cells of both mutants (Fig. , D, F, J, and K). Thus, MT aggregates in multinucleated G2 cells could be interpreted as an attempt to produce PPBs, but signals from several nuclei interfere with correct positioning. These interfering signals could also be the reason why spindles and phragmoplasts are misplaced in mitotic cells. These phenotypes support the hypothesis that signals from nuclei together with signals from the cortex are responsible for positioning the diverse MT arrays in plant cells. Cell Plate Formation in ple and hya | The MT analysis revealed that the phragmoplast is formed in both mutants. One of its proposed functions is to direct Golgi-derived vesicles to the division plane where the cell wall material-containing vesicles fuse to form the cell plate. The cell plate is a transient membrane-bound compartment, which undergoes complex transformations while expanding from the middle out to the division site of the parental cell walls. Callose was identified as the predominant lumenal component of forming cell plates. During cell plate maturation, the callose is replaced by cellulose, xyloglycans, and pectin . To gain insight into the aberrant cell wall formation of ple and hya, we studied callose deposition into the cell plate by histochemical staining with aniline blue. For better orientation, the nuclei were counter-stained with DAPI. Callose-containing cell plates were observed between sister nuclei in root meristems by focusing through the whole cell, because fluorescence of aniline blue and DAPI could rarely be focused in one optical plane (Fig. A). In multinucleated cells of ple and hya, cell plates develop synchronously and with a slight angle to each other (Fig. , B and C). This displacement is comparable with misplaced phragmoplastic MTs in multinucleated cells of fairly normal morphology. In multinucleated root cells with a stronger cell morphological defect, we rarely spotted cell plates in a focal plane. Callose was further observed at the end of cell wall stubs in already expanded multinucleated cells of ple (data not shown). Such plugs of callose can be induced by stress and might be a secondary effect. In summary, cell plates are formed in ple and hya but misplaced, indicating that the phragmoplast is functional. Figure 8 | Callose staining of cell plates in root meristems of wild type ple and hya. Callose staining of cell plates in root meristems of wild type ple and hya. A, An epidermal cell file of wild type with callose in the cell plates. B and C, ple-2 and hya-1, respectively, show callose deposition in the cell plate of a synchronously dividing cells with multiple nuclei. Bars = 25 mum in A through C. Double Mutant Analysis | To determine possible genetic interactions between PLE and HYA, double mutants were generated by crossing homozygous hya-1 and hya-2 with ple-1 and ple-2 mutants. F1 progeny of these crosses were all phenotypically wild type. The F2 seedlings were grouped into five classes: wild type, very weak, weak, ple, and double mutant phenotypes . Among F2 progeny, double mutants were classified as those with extremely short roots (n = 36, root length 1.31 +- 0.5 mm; Fig. C). CLSM analyses revealed that most of the root and more hypocotyl cells of double mutants were multinucleated, indicative of a synergistic enhancement of the cytokinesis defects (Fig. , D and E). Table IV | Double mutant and segregation analysis of the progeny of selfed ple/PLE.hya/HYA plants Figure 9 | Double mutant analysis of ple and hya. For double mutant analyses hya-1 (A) and ple-1 (B) were crossed and the genotype of different phenotypic classes were confirmed with molecular markers. Double mutant analysis of ple and hya. For double mutant analyses hya-1 (A) and ple-1 (B) were crossed and the genotype of different phenotypic classes were confirmed with molecular markers. C, The enhanced phenotype of the hya-1 ple-1 double mutant is exemplified. Optical sections through the root and the hypocotyl of hya-1 ple-1 double mutants show that more cells in both organs contain multiple nuclei. Bars = 1 mm (A --C) and 25 mum (D and E). The genotype of members of these phenotypic classes was confirmed using closely linked markers and segregation analyses of the F2 progeny. The homozygous hya seedlings exhibited the weakest phenotypes, whereas those grouped as weak phenotype were PLE/ple.hya/hya. In contrast to the single mutants, the double mutants were not viable on soil and were not fertile. Furthermore, in the progeny of selfed PLE/ple.hya/hya and ple/ple.HYA/hya plant, a significant number of seeds did not germinate, indicating that they are embryonic lethal . Together with the synergistic phenotype of the PLE/ple.hya/hya seedlings, this suggests that PLE and HYA genetically interact and might be involved in the same process. Table V | Segregation analysis of the progeny of selfed ple/ple.HYA/hya and PLE.ple/hya.hya plants DISCUSSION : Spatio-Temporal Specificity of PLE and HYA Function | We have identified two new loci, PLE and HYA, whose mutant alleles exhibit typical features of cytokinesis-defective mutants. Whereas in ple and hya all root tissues are affected by the mutations, the aerial portions of these mutants have a wild-type appearance and no detectable fertility or germination defects. Homozygous plants of five alleles survive on soil and are indistinguishable from wild type. Only the strongest allele of ple-1 forms a smaller plant. Moreover, ple hya double mutants show synergistic cytokinesis defects and are not viable on soil. Our interpretation is that the root functions of weak alleles are still able to support normal aboveground development, whereas the cytokinesis defects of roots in severe alleles cannot sustain normal plant growth. Plants cannot survive if, in addition to the root, aerial parts are affected, as the hypocotyl in double mutants. Moreover, the synergistic phenotype of double mutants points to a potential genetic interaction between PLE and HYA. The segregation data of homozygous and heterozygous ple and hya alleles indicate that PLE and HYA are not required for gametophytic development. The phenotype of ple and hya alleles is predominantly detectable in primary and secondary roots of seedlings grown on nutrient agar plates. Although most root tissues display cytokinesis defects, no dramatic radial pattern changes are detectable in the root or embryo. The spatial arrangement of initials is the result of early cell divisions, and at the early torpedo stage of embryogenesis, the tissues of both root and hypocotyl are produced . The finding that the maximal number of nuclei within root cells was 32 is in agreement with the approximately four cell divisions that each cell undergoes in postembryonic root meristems of Arabidopsis . This provides support for the hypothesis that most of the defective cell divisions occur after germination and points to a developmental stage-specific function of the PLE and HYA gene products. However, multinucleated cells are detectable at low frequency in mature embryos of strong ple alleles. Our PLE and HYA alleles, alternatively, may not be null mutants, and the phenotypic analyses may lead to an overestimate of their spatio-temporal specificities. Another possibility is that the PLE and HYA genes belong to a gene family with partially redundant functions. Other members of this gene family may act during embryogenesis, in mitotic divisions of somatic cells, in shoot and floral meristems, and during gametophytic cell divisions. Asymmetric Cell Plate Formation in ple and hya | We observed cell wall stubs in the mutants, indicating that the growing cell plate reaches the parental cell wall at late stages of cytokinesis but only on one side of the cell. This implies that as soon as one side of the cell plate reaches the parental wall, it is stabilized, whereas the other side of the plate is degraded. Thus, PLE and HYA gene products may be involved in the spatial-temporal coordination of cell plate growth and/or stabilization during cell wall maturation. Asymmetric cell wall stubs are characteristic for cytokinesis defects either induced by mutations or upon caffeine inhibition of cytokinesis. The possibility that cell plates are formed asymmetrically challenges our current model of plant cytokinesis in which cell plate formation is symmetric and centrifugal. Thus, in addition to the specific function of PLE and HYA, all cytokinesis-defective mutants point to a general novel characteristic of cell plate formation. The idea of asymmetric cell plate formation is supported by a recent paper of using three-dimensional live-cell imaging. The Number of Nuclei Influences Microtubule Structures and Polarity | The cytokinesis defects of ple and hya mutants do not inhibit multiple rounds of nuclear divisions. Thus, these multinucleated cells share some aspects of syncytia where free nuclear divisions occur uncoupled from cytokinesis and cellularization . Studies of MTs during the cellularization process of endosperm syncytia indicate that the formation of phragmoplasts is a primary difference between conventional and nonconventional cytokinesis. It has been shown that in higher plants, the nuclear envelope acts as a microtubule organizing center. Whereas the phragmoplasts of somatic cells derive from anaphase spindle remnants, the syncytial cells build miniphragmoplasts from MTs, which radiate from the microtubule organizing centers of adjacent nuclei . Our study indicates that the presence of multiple nuclei has a dramatic impact on the different MT arrays including short diffuse cytoplasmic, irregular perinuclear, misplaced cortical, PPBs, spindle, and phragmoplast MTs. The diverse aberrant MT arrays become more dramatic with increasing the number of interfering nuclei in the cells of ple and hya mutants. Nothing equivalent to the syncytial-type miniphragmoplasts or radiating nuclear MTs develop, indicating that missing PLE and HYA action does not turn somatic-type into a syncytial-type cytokinesis. The MT phenotypes of ple and hya mutants suggest the possibility that a defect in MT dynamic or positioning may be enough to hinder proper cell wall formation. Once nuclei of daughter cells are incompletely separated, their interfering signals lead to a cellular disaster impeding all events that rely on the coordination between the nucleus, the cytoskeletal elements, and the cell periphery. A morphological process that requires tight control between the nucleus and cell cortex is the positioning and emergence of root hairs. The syncytial epidermis trichoblasts are still able to form root hairs; thus, tip growth is not inhibited. But the position and the number of root hairs on one trichoblast vary. In rare cases, root hairs bifurcate, indicative of defects in tip growth polarity. Similar phenotypes have been observed by disruption the MT cytoskeleton dynamics with taxol or alpha-tubulin reduction . The root hair phenotype might indicate a MT-associated function for PLE and HYA. Thus, in addition to the regulation of cytokinesis, PLE and HYA could help to elucidate the importance of the interplay between the nucleus, cytoskeletal structures, and cell morphogenesis. Threshold Model of PLE and HYA Action | In summary, we propose that a certain threshold activity of the PLE and HYA gene products is needed to stabilize cytokinetic structures. If these stabilizers are not present or functional or if a certain threshold could not be synthesized, cell wall formation cannot be completed . Figure 10 | Schematic representation of cell division pattern within a cell file of wild-type and cytokinesis-defective root meristems. Schematic representation of cell division pattern within a cell file of wild-type and cytokinesis-defective root meristems. The diagram illustrates our interpretation of the cytokinesis defects in roots. The ordered cell files of wild-type roots are generated through transverse divisions in the meristem that has been already established during embryogenesis. After germination, cells start to divide. Our model proposes that the first defect in mutant root meristems may not yet change the normal morphogenesis of the affected daughter cells. But during subsequent rounds of divisions, the syncytial status of daughter cells leads to a cellular disaster. Instead of adding only one new daughter cell into a cell file, the syncytial nuclei divide simultaneously and positional signals between the nuclei and cytoskeletal elements get misinterpreted leading to visible morphological changes. This cellular disaster does not prevent further cell divisions as cells continue their division program up to five times. This model explains the rare embryonic phenotypes of the mutants, because their threshold level may be lower during embryogenesis or under growth conditions that reduce cell division such as growth in soil or on nutrient medium without Suc. This model would also allow another possible interpretation for the organ-specific phenotype, because our alleles may not be null mutants and the residual activity of the proteins reaches a threshold that is sufficient for embryogenesis. It also explains the radial gradient of phenotypic severity from the outside (epidermis) to the inside of the root (stele). In stele cells, the spatial constraints do not allow a dramatic volume increase even if daughter cells are not completely isolated by a new cell wall. On the other hand, these cells have a smaller diameter, thus, a smaller cell plate is needed for a complete separation. This model is also supported by the observation that cell division appears to be slowed down in multinucleated cells (; our histochemical results). Thus, as soon as multiple nuclei are debating cellular decisions, the timing of cytokinesis is disturbed. Isolation of PLE and HYA will complement the increasing number of cytokinesis-specific genes and will provide a more detailed picture of the structural basis of cytokinesis and its regulation. MATERIALS AND METHODS : Plant Material, Growth Condition, and Mutant Screening | Wild-type Arabidopsis accessions (Landsberg erecta [Ler], Columbia [Col], and Wassilewskija [Ws]) were obtained from the Arabidopsis Stock Center (Columbus, Ohio). Seeds were sterilized, seedlings were grown and screened for abnormal root morphology on vertical nutrient agar plates, and plants were propagated on soil according to and . The mutants were isolated because of their wavy growth pattern and irregularly expanded roots among approximately 5,000 EMS and 175,000 T-DNA mutagenized (K. Feldmann, University of Arizona [Tucson] and DuPont [Wilmington, DE]) Arabidopsis M2 and T4 seed pools, respectively. Phenotypic Characterization and Microscopic Analyses | The differential interference contrast microscopy (Nomarski) was performed with cleared preparations. Roots were fixed with methanol:acetic acid (MA = 3:1, v/v) and treated with Herr's solution (chloralhydrate:phenol:lactic acid:xylol:carnation oil = 2:2:2:1:2, v/v). Transverse root sections were obtained as described by . nDNA of MA fixed roots was stained with 0.5 mug mL-1 DAPI in phosphate-buffered saline with Tween 20 (PBT) (130 mm NaCl, 10 mm phosphate buffer, pH 7, and 0.1% [v/v] Tween) for 10 min. Excess stain was removed by several washes with PBT followed by a glycerin series (30% [v/v], 50% [v/v], and 70% [v/v] glycerin in PBT). Seedlings were mounted in Citifluor (Plano, Wetzlar, Germany). Double staining for callose and DNA was performed with MA-fixed plant material. After washes with sterile water, DNA was stained with 2 mum DAPI in water for 5 min and washed with water for 1 min. Callose staining was conducted with 0.5% (w/v) Aniline blue in water for 10 min followed by 2-min washes with 2 mm Na2PO4, pH 7.5. Seedlings were mounted in 0.2% (w/v) Aniline blue, 100 mm Tris-HCl, pH 9, and 50% (v/v) glycerol on microscope slides. Roots were analyzed with a confocal laser scanning microscope (Axiovert, Zeiss, Jena, Germany) under UV with the DAPI filter set (365-nm excitation and 420-nm emission). Pictures were taken with a 167 mt camera (Contax, Japan) using R-100 film (Konica, Japan). The slides were scanned and arranged using Adobe Photoshop (Adobe Systems, Mountain View, CA). Histochemical GUS staining was done as described by . For confocal microscopy, seedlings were fixed in MA and stained with 0.5 mum YO-PRO (Molecular Probes, Eugene, OR) in PBT solution for 10 min. Washing and mounting was similar to the DAPI staining protocol. For YO-PRO staining of mature embryos, dry seeds were fixed in MA at 12 to 48 h, washed twice with water for 10 min, and incubated in 0.5 mum YO-PRO for 2 h. The seeds were kept in water until the testa was dissected under the microscope, and the embryos were mounted in 70% (v/v) glycerin. YO-PRO stains predominately DNA with weaker staining of RNA and the cell wall. Pictures were taken using filters for 488-nm excitation on a confocal laser scanning microscope (MSE 600, Bio-Rad, Hercules, CA) with an argon/krypton laser. Pictures were saved as PIC files translated with the Paint Shop Pro3 (batch conversion of raw file formats, v89a-nominated with 768 width and 512 length and 76-byte hadder size) into TIFF files and manipulated in Adobe Photoshop. The size of the root meristem was defined as by . For resin embedding, root tips were fixed with 2.5% (v/v) glutaraldehyde in 5 mm Na-cacodylate buffer (pH 6.8) for 1.5 h under low vacuum. After six 10-min washes in 5 mm Na-cacodylate buffer (pH 6.8), the roots were post-fixed with 1% (w/v) osmium tetroxide in 5 mm Na-cacodylate buffer (pH 6.8) for 2 h. After several washes, samples were dehydrated in an ethanol series starting with 30% (v/v) for 30 min, 60% (v/v), 75% (v/v), and 90% (v/v) each for 1 h, and 100% overnight. Media exchange started with two additional 100% (v/v) ethanol steps, followed by one 100% (v/v) ethanol:propylene oxide (1:1, v/v) and one pure propylene oxide step, each for 1 h. Media infiltration continued with a 3-h incubation of a propylene oxide:Spurr's resin (3:1, v/v) mixture and an overnight incubation with propylene oxide:Spurr's resin (1:1, v/v). The seedlings were allowed to equilibrate for 24 h in a propylene:Spurr's resin (1:3, v/v) mixture and with pure Spurr's resin two times each for 24 h. Polymerization was performed for 8 h at 70C. Semithin (1- to 2-mum) sections of embedded roots were stained about 30 s to 1 min. with 1% (w/v) aqueous toluidine blue mixed with 2% (w/v) Na2CO3 in 1:1 ratio at 50C to 60C. After washing, staining continued with 0.1% (w/v) aqueous basic fuchsine for about 30 s at the same temperature. Cell walls and polysaccharides stained bright red, nucleoli were dark blue, and the nucleoplasm and the cytoplasm were purple. For transmission electron microscopy (TEM), ultrathin (80- to 90-nm) sections were cut from the Spurr's resin embedded roots and stained under N2 atmosphere with uranyl acetate for 1 h and with lead citrate for 5 min. Sections were visualized using a Zeiss TEM 900. The images were digitized using an image analyzer (Q500MC, Leica Microsystems, Wetzlar, Germany) with Q500MC software and the monochromatic camera system CF8/1 FMC. Immunolocalization of Microtubules | Before fixation, whole seedlings were incubated in microtubule stabilization buffer (MTSB; 50 mm PIPES, 5 mm EGTA, and 5 mm MgSO4, pH 7.0) with 10% (v/v) dimethyl sulfoxide for 15 min at room temperature (RT). This solution was then exchanged with freshly prepared fixative (4% [w/v] p-formaldehyde in MTSB, 5% [v/v] dimethyl sulfoxide, and 0.1% [v/v] glutaraldehyde). After at least 1 h of fixation, the seedlings were washed three times with MTSB for total of 30 min. The roots were transferred to aminopropyltri-ethoxysilan (Sigma, St. Louis)-treated microscope slides. After desiccation, the roots were rehydrated for 10 min with MTSB and 0.1% (v/v) Nonident P40 (Sigma). Slides were carefully treated with 300 muL of 2% (w/v) driselase (Sigma) in MTSB for 30 min to 1 h, followed by washing three times with MTSB and 0.1% (v/v) Triton X-100 for about 30 min at RT. After three additional washing steps with MTSB for about 30 min, roots were blocked with 300 muL of 3% (w/v) bovine serum albumin (BSA; Sigma A-7030) in MTSB in a humid chamber for 1 h at RT, followed by an incubation with 300 muL of a 1:100 (v/v) dilution of the rat monoclonal YOL1/34 (anti-yeast-alpha-tubulin subunit antibody; MCA78S, Serotec, Oxford) in 3% (w/v) BSA/MTSB in the humid chamber overnight at RT. Slides were carefully washed three times with MTSB for 1 h then incubated with 300 muL of a 1:100 (v/v) dilution of the fluorescein-isothiocyanate-labeled secondary anti-rat antibody (Sigma F-1763) in 3% (w/v) BSA/MTSB in a dark, wet chamber for about 4 h. Slides were washed three times with MTSB for 60 min and mounted in Citifluor. The slides were keep at 4C in the dark until microscopic examination. CLSM was performed with a Bio-Rad MSE 600 at 488 nm. Pictures were taken, stored, and manipulated as described above. Genetic Analyses and Mapping | Homozygous plants were used for pair wise crosses between the different cytokinesis mutants to determine complementation groups. The chromosomal locations of the ple and hya genes were determined by measuring the recombination frequency between the mutant genes and microsatellite markers (; modified as ) and CAPS . More than 600 F2 plants from crosses of ple-3, hya-3, hya-1, and hya-2 to Col were used for mapping. For double mutant analysis, homozygous hya-1, hya-2, and ple-1 plants were used to pollinate homozygous ple-1, ple-2, and hya-1 plants, respectively. F1 and F2 progeny of those crosses were analyzed for their root phenotype on vertical agar plates. Among the progeny of F2 plants with [ple/ple.HYA/ple] and [PLE/ple.hya/hya] genotypes we were able to obtain a higher frequency of double mutants which were use to analyze their cellular phenotypes. Genomic DNA for PCR analysis was isolated according to . We designed a new forward primer for the microsatellite nga129 5'-CATAATCGAATCGGACACGAC-3' resulting in 59-bp-smaller PCR fragments for all ecotypes. In addition, the SSLP marker AthS0191 was used . The CAPS marker ASBII detects a HinfI Col/Ws polymorphism . For double mutant analysis, a closer molecular marker was developed from the Trp beta-synthase 1 gene (TSB1; ). The primers TSBI-F 5'-GTGCCGAGGTGATGCTTAGG-3' and the TSBI-R 5'-GTCACCCAATCTCTTATCGCTTCA-3' amplify a 162-bp fragment in Col; the slightly smaller Ws fragment is distinguishable on native 5% (w/v) PAGE. This marker is located between ple and ASBII and has a recombination frequency with PLE of about 5% (w/v). Backmatter: PMID- 12226512 TI - Light-Intensity-Dependent Expression of Lhc Gene Family Encoding Light-Harvesting Chlorophyll-a/b Proteins of Photosystem II in Chlamydomonas reinhardtii AB - Excessive light conditions repressed the levels of mRNAs accumulation of multiple Lhc genes encoding light-harvesting chlorophyll-a/b (LHC) proteins of photosystem (PS)II in the unicellular green alga, Chlamydomonas reinhardtii. The light intensity required for the repression tended to decrease with lowering temperature or CO2 concentration. The responses of six LhcII genes encoding the major LHC (LHCII) proteins and two genes (Lhcb4 and Lhcb5) encoding the minor LHC proteins of PSII (CP29 and CP26) were similar. The results indicate that the expression of these Lhc genes is coordinately repressed when the energy input through the antenna systems exceeds the requirement for CO2 assimilation. The Lhc mRNA level repressed under high-light conditions was partially recovered by adding the electron transport inhibitor 3-(3,4-dichlorophenyl)-1,1-dimethylurea, suggesting that redox signaling via photosynthetic electron carriers is involved in the gene regulation. However, the mRNA level was still considerably lower under high-light than under low-light conditions even in the presence of 3-(3,4-dichlorophenyl)-1,1-dimethylurea. Repression of the Lhc genes by high light was prominent even in the mutants deficient in the reaction center(s) of PSII or both PSI and PSII. The results indicate that two alternative processes are involved in the repression of Lhc genes under high-light conditions, one of which is independent of the photosynthetic reaction centers and electron transport events. Keywords: Introduction : Photosynthesis is regulated at various levels in response to fluctuating light intensity under various ambient temperature and nutrient conditions. The proper responses to the various environmental cues are necessary for photosynthetic plants to use light energy efficiently and to protect themselves from photoinhibitory damage caused by excessive irradiance . Excessive light energy absorbed by chlorophyll is dissipated by non-radiative processes and is properly distributed between two photosystems (PS) by state transition , whereas the energy input is regulated by changes in the size of the light-harvesting antenna systems to modulate the optical cross section. Light-harvesting chlorophyll a/b (LHC)II proteins, which are major components of light-harvesting antennae of PSII in higher plants and green algae, typically change their abundance in response to the intensity of irradiance . Under stress and intense light, enhanced amounts of reactive oxygen species will react with proteins and lipids, not only in chloroplasts but also in the cytosol, and will induce various types of photodamage. Therefore, the quality and quantity control of the LHC protein complex is required to avoid photodamage by alleviating excitation energy pressure. Although the LHC protein complex could be controlled by various mechanisms including pigment synthesis, the repression of the Lhc genes under stressful light conditions must be an important antistress response of plants. However, little is known about the mechanism of how the excessive light intensity is sensed and how the signal is transduced to change gene expression. One proposal is that the redox state of the photosynthetic electron transport carrier(s) between the two PS in green algae monitors the energy balance because such carriers will be over-reduced if the energy input exceeds the requirement for the dark reaction. The abundance of LHCII protein and/or mRNA decreases with the increase of the reduced QA population probed by chlorophyll a fluorescence in Chlorella vulgaris and Dunaliella salina . Expression of the LHCII gene in Dunaliella tertiolecta is enhanced by interrupting electron transfer from QA to QB with 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU), and it is repressed by inhibiting the oxidation of plastoquinol with 2,5-dibromo-3-methyl-6-isopropyl-p-benzoquinone . These findings suggest that the redox state of the plastoquinone (PQ) pool participates in the control of LHCII gene expression. On the other hand, studies of some higher plants have shown that the redox state of the photosynthetic electron carriers is not strictly correlated with LHCII gene expression . The light-harvesting antennae comprise several homologous LHC proteins encoded by a nuclear gene family . Although the exact reasons for the multiplicity of the LHC proteins have not been elucidated, each LHC protein may have distinct functions in light harvesting, such as the optimization of light energy transfer and the dissipation of excessive light. Studies to date on the response of Lhc genes to light intensity have mainly focused on the gene encoding the most abundant LHC (LHCII) protein. Whether each Lhc gene is regulated independently or whether they are all coordinately regulated in response to the light intensity remains unknown. To understand the light-dependent regulation of the entire antenna system, comprehensive studies on the light response of all Lhc genes are required. The unicellular green alga Chlamydomonas reinhardtii has been extensively applied as a model experimental system for studies of photosynthesis. The composition of LHC proteins in this alga has been best characterized in algal species . We characterized the Lhc gene family encoding the LHC proteins of PSII using the C. reinhardtii expressed sequence tag (EST) databases . The results revealed that this alga has at least six genes encoding the major LHC (LHCII) proteins and two genes for the minor LHC proteins (CP29 and CP26). The highly homologous LHCII proteins in C. reinhardtii cannot be assigned to any of the three proposed types in higher plants (Lhcb1-Lhcb3), but they can be classified into four distinct types. Type I is encoded by the three genes: LhcII-1.1, LhcII-1.2, and LhcII-1.3. Types II, III, and IV are encoded by LhcII-2, LhcII-3, and LhcII-4, respectively. Therefore, C. reinhardtii should provide a promising experimental system with which to study regulation of Lhc gene expression under various environmental conditions. The present study uses quantitative reverse transcriptase (RT)-PCR to examine the amounts of the multiple Lhc mRNAs that accumulate in C. reinhardtii cells exposed to various intensities of light at various temperatures and under different CO2 conditions. The Lhc mRNA levels were also examined in the presence of the PSII inhibitor DCMU and in mutants deficient in the PSII or both the PSI and PSII reaction center(s). We discuss the mechanisms involved in the light-intensity-dependent control of Lhc gene expression. RESULTS : Light-Intensity-Dependent Expression of Lhc Genes at Various Temperatures | Figure shows profiles of DNA amplification by PCR in the RT-PCR analysis of LhcII-4 and Lhcb4 encoding LHCII and CP29 protein, respectively. C. reinhardtii cells cultured mixotrophically under continuous light (5 muE m-2 s-1) were dark-adapted for 12 h at 26C and then transferred to various light intensities (0, 50, 100, or 200 muE m-2 s-1) for 6 h at 22C. Total RNA was analyzed by RT-PCR to quantify the mRNA. Relative mRNA abundance revealed by electrophoresis of the RT-PCR products (Fig. A) was more precisely quantified by analyzing the kinetics of the PCR reaction using a real-time PCR assay system (Fig. , B and C). The amounts of LhcII-4 and Lhcb4 mRNAs were significantly larger under low light (50 muE m-2 s-1), as revealed by the appearance of the PCR products at fewer PCR cycles, than in the dark or under high light (200 muE m-2 s-1). In contrast, light intensity barely affected the amount of 18S rRNA (Fig. , A and D), which was used as the internal standard in the quantitative RT-PCR analysis described below. Figure 1 | Quantitative RT-PCR analysis of Lhc mRNAs in C. reinhardtii Quantitative RT-PCR analysis of Lhc mRNAs in C. reinhardtii under various intensities of light. Algal cells were dark-adapted for 12 h and exposed to 0, 50, 100, or 200 muE m-2 s-1 of light for 6 h at 22C. Total RNA prepared from each sample was analyzed by RT-PCR using a primer set for LhcII-4, Lhcb4, and 18S rRNA. A, Electrophoretogram of DNAs amplified by RT-PCR. DNA at the exponential phase of the PCR amplification was resolved by electrophoresis and detected using SYBR Green I. B through D, Kinetics of DNA amplification analyzed by a real-time PCR assay system. LhcII-4 (B), Lhcb4 (C), and 18S rRNA (D) in darkness (circle), low (50 muE m-2 s-1; ) and high light (200 muE m-2 s-1; ). PCR reactions proceeded in the presence of SYBR Green I, and amounts of amplified DNA were monitored by fluorescence intensity. Fluorescence intensity (arbitrary units) is plotted against number of PCR cycles. Figure shows the effects of light intensity on the levels of LhcII-4 and Lhcb4 mRNAs that accumulated at various temperatures. The abundance of Lhc mRNA was normalized with respect to that of 18S rRNA and then presented as values relative to that under dark conditions, where effects of temperature on the amounts of the Lhc mRNAs could be minimized. Figure A shows that the mRNA level of LhcII-4 in the cells exposed to low (50 muE m-2 s-1) and medium (100 muE m-2 s-1) light intensity at 26C was increased 8- to 10-fold compared with that in the cells kept in darkness. However, the level in the cells exposed to high light (200 muE m-2 s-1) only doubled. These findings indicate that the mRNA level is significantly enhanced by illumination at low and medium intensity but the light-dependent enhancement is repressed to be low at high intensity. Low-light exposure at 22C similarly enhanced the mRNA level, but the level under medium light was one-fourth of that under low light, and the mRNA level was not enhanced under high light. In contrast to the findings at 22C and 26C, low light barely enhanced the mRNA level at 18C. The level under medium and high light at 18C was even lower than that in darkness. It is of note that no visible symptom of photodamage was observed and the cells proliferated well under the high-light conditions at 18C. These results indicate that the threshold light intensity required to repress the mRNA accumulation decreased as the temperature lowered. The cells at 18C may perceive the light at 50 muE m-2 s-1 as being more intense than 200 muE m-2 s-1 at 26C. Figure B shows that the mRNA level of Lhcb4 responded to light intensity and temperature in a similar manner to that of LhcII-4, although the mRNA level was slightly higher under 100 muE m-2 s-1 than under 50 muE m-2 s-1 at 26C. Figure 2 | Effects of light intensity and temperature on accumulation of LhcII-4 (A) and Lhcb4 (B) mRNAs. Effects of light intensity and temperature on accumulation of LhcII-4 (A) and Lhcb4 (B) mRNAs. C. reinhardtii cells exposed to light at various intensities for 6 h at 26C (circle), 22C , and 18C were examined by quantitative RT-PCR. Abundance of Lhc mRNA relative to values in darkness was plotted after normalization with respect to that of 18S rRNA. Results from two independent experiments were similar, and mean values are presented. Figure compares the responses of the mRNA levels of four Lhc genes (LhcII-1.1, LhcII-1.3, LhcII-3, and LhcII-4) encoding the major LHC (LHCII) proteins and two (Lhcb4 and Lhcb5) encoding the minor LHC proteins (CP29 and CP26) to light intensities at three different temperatures. The levels of mRNA accumulation are shown relative to those under low light (50 muE m-2 s-1). The responses of the mRNAs of the tested Lhc genes to light intensity and the effects of temperature on their responses were essentially similar. These results suggest that these Lhc genes are coordinately regulated under common signal transduction pathways. The relative mRNA levels for Lhcb4 and Lhcb5 were slightly but reproducibly higher than those for the LhcII genes under high light (200 muE m-2 s-1) at all the tested temperatures. This suggests that the genes of the minor LHC proteins were less sensitive to down-regulation by high light than those of the major LHC proteins. Figure 3 | Comparison of light-dependent changes in levels of mRNAs for multiple Lhc genes at 26C (A), 22C (B), and 18C (C). Comparison of light-dependent changes in levels of mRNAs for multiple Lhc genes at 26C (A), 22C (B), and 18C (C). Quantitative RT-PCR data for LhcII-1.1, LhcII-1.3, LhcII-3, LhcII-4, Lhcb4, and Lhcb5 mRNAs are presented. Abundance of mRNA relative to values under low light (50 muE m-2 s-1) is shown after normalization with respect to that of 18S rRNA. Results from two independent experiments were similar, and mean values are presented. Effects of CO2 on Light-Intensity-Dependent Expression of Lhc Genes | Figure shows the effects of the CO2 concentration on the light-dependent changes in the mRNA levels of the Lhc genes. The cells were photoautotrophically cultured in HS minimal medium under low light (50 muE m-2 s-1) at 26C bubbled with either air (low CO2) or 5% (v/v) CO2 in air (high CO2), then exposed to very high light (1,000 muE m-2 s-1) for 6 h. The mRNA level for all the tested Lhc genes (LhcII-1.1, LhcII-1.3, LhcII-3, LhcII-4, Lhcb4, and Lhcb5) in the cells exposed to very high light under the low CO2 concentration was about 20% of that in the cells incubated under low light. These changes are compatible with the results obtained under high light (200 muE m-2 s-1; Figs. and ), where the cells were dark-adapted before exposure to the various light conditions with no air-bubbling. The mRNA levels under low light were not significantly affected by the CO2 concentration (data not shown). In contrast to low CO2 conditions, exposure to high light under high CO2 conditions elevated the mRNA level to 7- to 30-fold of that under low light. These results indicate that limitation of the supply of carbon source is the primary cause of the decreased Lhc mRNAs levels with high light under our experimental conditions. Figure 4 | Effects of CO2 concentration on light-intensity-dependent changes in Lhc mRNA levels. Effects of CO2 concentration on light-intensity-dependent changes in Lhc mRNA levels. C. reinhardtii cells were cultured under low light (50 muE m-2 s-1) and bubbled with either air (A; Low CO2) or 5% (v/v) CO2 in air (B; High CO2) at 26C, transferred to high light (H; 1,000 muE m-2 s-1) or kept under low light (L), and incubated for 6 h. Quantitative RT-PCR data for LhcII-1.1, LhcII-1.3, LhcII-3, LhcII-4, Lhcb4, and Lhcb5 mRNAs are presented. Abundance of mRNA relative to values under low light is shown after normalization with respect to that of 18S rRNA. Effects of DCMU on Light-Intensity-Dependent Expression of Lhc Genes | The results so far indicated that the Lhc genes are down-regulated when light energy absorbed by the light-harvesting system exceeds the requirement of the photosynthetic dark reactions, which are retarded by a decrease of temperature or a limited CO2 supply. Under the excessive light conditions, the photosynthetic electron carriers subsequent on the acceptor side of PSII are in an over-reduced state. Therefore, the redox state of one of the electron carriers may be used to sense the energy imbalance under excessive light conditions that would down-regulate the Lhc genes. To examine this notion, we examined light-intensity-dependent changes in the Lhc mRNAs levels in the presence of the herbicide, DCMU, which inhibits electron transfer from reduced QA to QB. C. reinhardtii cells were cultured in the dark for 12 h at 26C in the presence or absence of 10 mum DCMU, then transferred to low (50 muE m-2 s-1) or high light (200 muE m-2 s-1) for 6 h at 22C. Photosynthetic oxygen evolution was almost completely inhibited in the presence of 10 mum DCMU (data not shown). Figure shows the results of quantitative RT-PCR analysis of the six Lhc genes described above and LhcII-1.2 and LhcII-2. The level of the LhcII-1.1 mRNA was approximately 50-fold lower under high light than under low light in the absence of DCMU. The presence of DCMU reduced the mRNA level by only 20% in the cells exposed to low light, indicating that the mRNA accumulates under low light even when electron transport at the QB site is interrupted. In contrast, DCMU caused a 10-fold increase in the mRNA level in the cells transferred to high light. These findings are consistent with the view that excessive electron donation by PSII reduces the Lhc mRNA level. We emphasized, however, that the mRNA level enhanced by DCMU under high light was still significantly lower than that under low light. This indicates that the level of LhcII-1.1 mRNA accumulation can respond to high light even when electron delivery from PSII is inhibited by DCMU. The effect of DCMU on the responses of the other Lhc genes to light intensity was similar, although the extent of the responses varied among those genes. With respect to the Lhcb4 and Lhcb5 genes, the mRNA levels were relatively high under high light in the presence and absence of DCMU. The levels under high light reached 50% to 60% of those under low light in the presence of DCMU. Figure also shows the mRNA levels of the CRY1 and psbA genes. The nuclear gene CRY1 encodes the ribosomal protein S14 , and psbA is a chloroplast gene encoding the D1 protein of the reaction center of PSII . In contrast to the levels of the Lhc mRNAs, those of both the CRY1 and psbA mRNAs were relatively higher under high light than under low light. DCMU did not affect the CRY1 mRNA level but enhanced that of the psbA mRNA. The enhancement of the psbA mRNA level by DCMU is compatible with the fact that the oxidized PQ pool can induce psbA expression in chloroplasts and cyanobacteria . As an alternative, this may be ascribed to the inhibitory effect of DCMU on mRNA degradation in chloroplasts , although such effect is not observed for the psbA gene. However, why the mRNA level is enhanced by high light remains unclear, but an increased turnover rate of the D1 protein may be involved. Figure 5 | Effects of DCMU on light-intensity-dependent changes in accumulation of Lhc mRNAs. Effects of DCMU on light-intensity-dependent changes in accumulation of Lhc mRNAs. C. reinhardtii cells dark-adapted for 12 h at 26C were exposed to low (50 muE m-2 s-1) or high (200 muE m-2 s-1) light for 6 h at 22C with or without DCMU. Quantitative RT-PCR data for LhcII-1.1, LhcII-1.2, LhcII-1.3, LhcII-2, LhcII-3, LhcII-4, Lhcb4, and Lhcb5 mRNAs and CRY1 and psbA mRNAs are presented. Abundance of mRNA relative to values under low light without DCMU is shown after normalization with respect to that of 18S rRNA. Results from two independent experiments were similar, and mean values are presented. Light-Intensity-Dependent Expression of Lhc Genes in Mutants Deficient in the Photosynthetic Reaction Center | The low sensitivity of the light response of the Lhc genes to DCMU implies that over-reduction of the photosynthetic electron carriers is not necessarily required to repress the Lhc mRNAs levels under excessive light. The possible contribution of the reaction center function to gene regulation can be studied using mutants deficient in the reaction center. We therefore examined the Lhc mRNA levels in a mutant deficient in the PSII reaction center (ac-u-epsilon) and in another deficient in both the PSII and PSI reaction centers (FUD7-DeltapsaA). Cells were cultured in the dark for 12 h at 26C and then exposed to low (50 muE m-2 s-1) or high (200 muE m-2 s-1) light for 6 h at 22C. The mutant cells did not evolve O2 or accumulate the psbA mRNA, but the LHC abundance seemed to be comparable with that of the wild-type cells judging from their chlorophyll content (data not shown). Figure shows the results of the quantitative RT-PCR analysis of the LhcII-4 and Lhcb4 genes. The level of the LhcII-4 mRNA accumulation under high light was 15% and 10% of that under low light in the ac-u-epsilon (-PSII) and FUD7-DeltapsaA (-PSI and -PSII) mutants, respectively. The level of Lhcb4 mRNA in these mutants also responded to light intensity in a manner similar to that of the wild type, although the extent of changes was somewhat different. The mRNA levels of these two Lhc genes in the mutants were comparable with those of the wild type under low light (data not shown). These results indicate that as in the wild-type strain, LhcII-4 and Lhcb4 mRNA accumulation was repressed by high light in these PS-deficient mutants. On the other hand, the level of CRY1 mRNA was not significantly affected by changes in light intensity in either the mutants or in wild-type cells. Figure 6 | Light-intensity-dependent changes in accumulation of Lhc mRNAs in mutants deficient in the photosynthetic reaction center(s). Light-intensity-dependent changes in accumulation of Lhc mRNAs in mutants deficient in the photosynthetic reaction center(s). Wild-type strain C-9 (WT), a strain deficient in PSII reaction center, ac-u-epsilon (-PSII), and a strain deficient in both PSI and PSII reaction centers, FUD7-DeltapsaA (-PSI and -PSII), were dark-adapted for 12 h at 26C and then exposed to low (50 muE m-2 s-1) or high (200 muE m-2 s-1) light for 6 h at 22C. Quantitative RT-PCR data for LhcII-4, Lhcb4, and CRY1 mRNAs are presented. Abundance of mRNA in each strain relative to values under low light is shown after normalization with respect to that of 18S rRNA. Each value is mean with se of three independent experiments. DISCUSSION : C. reinhardtii has at least six genes corresponding to the major LHC (LHCII) proteins and two corresponding to the minor LHC proteins (CP29 and CP26) according to an analysis of the EST databases . LhcII-1.2, LhcII-2, and Lhcb5 could be expressed at much lower levels than the other LhcII genes if the relative number of EST clones for each gene reflects the differences in expression levels for multiple Lhc genes. The present results showed that levels of all of the tested Lhc mRNAs similarly responded to the light intensity when C. reinhardtii cells were exposed to light for 6 h. The mRNA levels were markedly increased by the exposure to low light, but this induction was repressed under high light. The levels were down-regulated at lower light intensities at lower temperatures or under conditions of a limited CO2 supply, suggesting that the mRNA levels are repressed when the photon energy absorbed by PS exceeds the energy used by the CO2 assimilation reactions. The results showed that all of the tested Lhc genes are coordinately regulated under excessive and low light conditions. Multiple products of these genes are thought to be assembled into a PSII antenna complex. Therefore, the coordinate regulation of the Lhc mRNAs levels may contribute to an efficient control of antenna size in response to changes in some environmental factors. The present results showed that the mRNA levels for the Lhcb4 and Lhcb5 genes encoding the minor LHC proteins (CP29 and CP26) tended to be less responsive to light intensity than those of the major LHC proteins . Results from cells acclimated to various light conditions for a longer time suggest that Lhcb5 is regulated by light in a less strict manner compared with LhcII-1.1 . The major and minor LHC proteins differ in terms of pigment contents and protein-protein interaction . Therefore, the differential expression between these two types of LHC genes may reflect their functional specificity in an antenna complex. It is, therefore, reasonable that the genes for the inner antennae (CP29 and CP26) are somewhat differently regulated than the rest of the peripheral antennae (LHCII). reported that a gene product for CP29 was specifically present in the cDNA library generated from C. reinhardtii cells starved for nutrients. The result may be compatible with the less strict repression of Lhcb4 under excessive light conditions observed in this study. The present study examined the expression of eight genes for LHC proteins of PSII, which have been found based on the analysis of C. reinhardtii EST databases . However, it is of note that genes expressed at very low levels or under specific conditions may not be detected from the EST databases. Several Lhc genes, presumably for LHCII proteins, were recently deposited in GenBank under accession numbers (Mussgnug and Kruse), , , and (Elrad, Niyogi and Grossman). Detailed characteristics of these genes have not been reported. reported that LHCII gene expression under high light is enhanced by inhibiting electron transfer from QA to QB with DCMU but is repressed by inhibiting the oxidation of plastoquinol with 2,5-dibromo-3-methyl-6-isopropyl-p-benzo-quinone in D. tertiolecta. On the basis of these results, they proposed that the redox state of the PQ pool is involved in the light-intensity-dependent expression of Lhc in green algae and that an over-reduced PQ pool is an initial signal for excessive light excitation leading to the down-regulation of gene expression. A similar redox control of gene expression has been proposed for both nuclear and chloroplast genes encoding photosynthetic proteins and for photosynthetic genes in cyanobacteria . Furthermore, a similar redox control mechanism participates in the state I-state II transition operated for the redistribution of absorbed excitation energy between PSI and PSII (; Gal and ). Figure shows that DCMU partially restored the accumulation of Lhc mRNAs, which were repressed under high light. This result could be compatible with the above control mechanism via the redox state of the PQ pool. However, much less Lhc mRNAs accumulated under high light than under low light even in the presence of DCMU, indicating that the regulation mechanism through the redox-state of the pooled PQ cannot simply account for the repression of the Lhc genes under excessive light. The repression of the Lhc mRNAs levels under high light was distinct even in the two mutants deficient in the reaction center(s) shown in Figure . This demonstrated that Lhc mRNAs levels can be down-regulated by high light through a mechanism that is independent of the photosynthetic reaction centers. Therefore, a putative redox-independent mechanism is responsible for repressing the Lhc mRNAs levels under excessive light shown in Figures through and under high light in the presence of DCMU shown in Figure . However, these data cannot completely exclude the possibility that the high-light repression in the mutants occurs via a mechanism that is not directly related to the repression found in the wild-type cells. In general, the PQ pool-dependent redox control mechanism is important for fine-tuning the distribution of excitation energy between the two PS and/or adaptation to physiological light conditions . When the irradiance exceeds the level that triggers photoinhibitory effects, the function of PSII is impaired. The function of PSII can be restored by the repair cycle including selective degradation of the damaged D1 protein and regeneration of the PSII complex with de novo synthesized D1 protein . It has been reported that reactive oxygen species preferentially inhibit the repair process . If the majority of PSII are photodamaged, the control of the Lhc genes through the redox state of a linear electron transport chain will not work well because the damaged PSII cannot reduce the electron transport carriers sufficiently. Under these situations, the redox-independent mechanism for down-regulation of Lhc genes demonstrated in the present study may be important for minimizing excitation pressure to facilitate the repair of the damaged PSII. We considered how C. reinhardtii cells perceive high light to repress the Lhc genes in the presence of DCMU and in the absence of PS. We note that the PS mutants used in this study retained amounts of antennae that were comparable with those of wild-type cells. The LHC antennae could be a potent source of reactive oxygen species in principle. An excited triplet chlorophyll (3Chl*) molecule sensitizes the formation of an excited singlet state of oxygen (1O2*), which is highly reactive, oxidizes many biological components, and may produce other reactive species via superoxide formation . Under optimal light conditions, however, the yield of 3Chl* formation is rather low because of efficient energy transfer to the reaction center and the resulting small amounts of 3Chl* and 1O2* are quenched safely by carotenoids adjacent to the antenna chlorophyll. However, the 3Chl* formation will be enhanced when the energy absorbed by LHC antenna exceeds the requirement for the photosynthetic reactions because the excitation energy can no longer be efficiently transferred to the closed reaction center. Therefore, 1O2* and 3Chl* formation in the LHC antenna should be enhanced even in the presence of DCMU and in the absence of the PS under high light. The amounts of CRY1 mRNA were little affected by high light either in the wild type (+ DCMU) or in the mutants as shown in Figures and . Furthermore, the level of psbA mRNA in the wild-type cells was increased by DCMU, which is compatible with the view that psbA is up-regulated when the PQ pool is oxidized. This indicates that the sensory and signaling process mediated by the redox state of the electron transport chain operates quite normally. These data demonstrate that the direct and nonspecific destruction of mRNA molecules and/or damage to nuclear and chloroplast transcriptional machinery by reactive oxygen species are not the cause of the repressed Lhc mRNAs levels under high light. Taking these into consideration, a reactive oxygen species or its reactant with a cellular component may function as an initial signal of excessive light intensity. Many metabolic systems scavenge reactive oxygen species in chloroplasts and the cytosol. Therefore, one of the metabolic species of such systems may be involved in the repression of Lhc mRNAs levels as a signal. In fact, several lines of evidence indicate that the expression of some nuclear and plastid genes in plants are controlled by reactive oxygen and metabolic species that compose the defense system for oxidative stress . Furthermore, the expression of Lhcb1 is possibly regulated by chlorophyll synthesis precursors , of which levels in chloroplasts may be affected by reactive oxygen species. Further studies are required to elucidate the involvement of these factors in the repression of Lhc mRNAs levels under excessive and stressful light conditions. Up-regulation of Lhc genes accompanying transfer from dark to moderate light is performed not only via photosynthetic reactions but also via a signal transduction mechanism with a specific photoreceptor molecule, such as phytochrome in some plant species . Therefore, the enhanced Lhc mRNA accumulation upon exposure to low light in this study could be at least partly ascribed to up-regulation by this type of photoreceptor system. This view is consistent with the results that the addition of DCMU and the deletion of the PS little affected the low-light accumulation. Although phytochrome has not been reported in C. reinhardtii, a protein with retinal as chromophore (chlamyopsin; ) and a homolog of cryptochrome (CPH1; ) have been identified. proposed that the light-induced expression of a Lhc gene is controlled by a system with a blue-light receptor rather than by a phytochrome in C. reinhardtii, although the photoreceptor has not yet been identified. In Arabidopsis, a blue-light receptor (NPL1) conducts the high-light avoidance response of chloroplasts . Thus, a blue-light receptor might be involved in the high-light repression of the levels of Lhc mRNAs. However, this view is less likely because the light enhancing the Lhc mRNA level turns to repress the level under lower temperatures or less CO2 supply, indicating that the Lhc mRNAs levels do not respond to light intensity itself. We cannot completely exclude the possibility that the Lhc genes are independently controlled under high light by multiple sensory and signaling processes, or in concert with each other. In conclusion, the present results indicate the presence of the mechanism for high-light repression of Lhc genes independent of the function of the both photosynthetic reaction centers, in addition to the process dependent on the redox state of the photosynthetic electron transport chain. The former redox-independent process is assumed to be functional when plants are exposed to excessive and stressful light, which expedites photoinhibitory effects. MATERIALS AND METHODS : Strains and Culture Conditions | The wild-type Chlamydomonas reinhardtii strain C-9 was obtained from the IAM culture collection at the University of Tokyo. The PSII deficient strain ac-u-epsilon was obtained from the Chlamydomonas Genetics Center (Duke University). The PSII/PSI double-deficient mutant FUD7-DeltapsaA generated by deleting the psaA gene in a PSII deficient strain FUD7 was a gift from Dr. K. Redding. C. reinhardtii cells were mixotrophically cultured in Tris-acetate-phosphate medium under continuous dim light (5 muE m-2 s-1) at 26C with constant agitation. The cultures at the mid-logarithmic growth phase (approximately 1 x 106 cells mL-1) were dark-adapted by wrapping the flask with aluminum foil for 12 h, and then 4-mL aliquots of the cultures were transferred to 18-mL test tubes and exposed to light at various intensities (0, 50, 100, or 200 muE m-2 s-1) for 6 h at 18, 22 or 26C in a growth chamber with white fluorescent bulbs. The lower intensities of light were achieved by wrapping the test tubes with layers of gauze or with aluminum foil. DCMU at 10 mm dissolved in dimethyl sulfoxide (DMSO) was added to the cultures to a final concentration of 10 mum when indicated. The same volume of DMSO was added to the control cultures. Because the photosynthesis-deficient strains are relatively sensitive to light, they were cultured under continuous very low light (<1 muE m-2 s-1) before dark adaptation. As an alternative, C. reinhardtii cells were photoautotrophically cultured in HS medium in a 100-mL test tube under continuous low light (50 muE m-2 s-1) at 26C bubbled either with air or with 5% (v/v) CO2 in air. The cultures at the mid-logarithmic phase were exposed to intense light (1,000 muE m-2 s-1) or kept under low light for 6 h. The cell cultures were illuminated with white fluorescent bulbs (up to 200 muE m-2 s-1) or halogen lamps (1,000 muE m-2 s-1). The photosynthetically active irradiance was measured using a quantum sensor (model LI-189, LI-COR, Lincoln, NE). Quantitative RT-PCR | Total RNA was prepared from algal cells using the TRIZOL reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's manual. Using an RNA PCR kit (TaKaRa, Tokyo), single-stranded cDNA was synthesized from 0.5 mug of total RNA with random primers in 20 muL of reaction mixture, and then 1 muL of the cDNA mixture was added as a template in 19 muL of PCR mixture containing each primer (0.2 --0.5 mum) and DNA polymerase master mix with SYBR Green I (SYBR Green PCR Master Mix, PE Biosystems, Foster, CA or LightCycler FastStart DNA Master SYBR Green I, Roche Diagnostics, Mannheim, Germany). After heating at 95C for 10 min, PCR reactions proceeded via 40 cycles of 15 s at 95C, 5 to 15 s at 60C, and 10 to 30 s at 72C. The amount of the amplified DNA was monitored by fluorescence at the end of each cycle using GeneAmp 5700 (PE Biosystems) or LightCycler (Roche Diagnostics). The PCR cycle program and the concentration of Mg2+ and primers in the PCR reaction mixture were optimized for each primer set. To analyze eight LhcII genes by RT-PCR, the common primer 5'-GTTCGGCTTCTTCGTCCAGG-3' or 5'-AGGCTGTCTGGTTCA-AGGC-3', corresponding to the almost identical sequence among these mRNAs, was used in combination with the specific primers 5'-GCTGCTG-CAGAGCTTATACG-3' (LhcII-1.1), 5'-ATCGTCAGCATCATCGTAGG-3' (LhcII-1.2), 5'-AACGACACACGGCACACATC-3' (LhcII-1.3), 5'-CACGCC-ACCAGTCTACACCA-3' (LhcII-2), 5'-CTCACTCAAGTCAAGAAGGC-3' (LhcII-3), and 5'-AACGTTGACAAGTCCCAGCG-3' (LhcII-4). The primers 5'-CCCGGTTGTGTATGATAAGTTGC-3' and 5'-CCTCCGCTCACACAA-ACCA-3' were used for Lhcb4. For Lhcb5, the primers were 5'-GACC-GCAAGCTGTTGCTGCC-3' and 5'-GCTCGTTCTCGCGGTACTTG-3'. The primers for CRY1 were 5'-GGCCCCCAAGGAGGTGGT-3' and 5' G-CGTAGGGCGACGACTCG-3'. The primers for psbA were 5'-GGCCAAGGTTCATTCTCTGA-3' and 5'-CACCGAATACACCAGCAACAC-3', and the 18S rRNA primers were 5'-ACTGCTCTGCTCCACCTTCC-3' and 5'-TATTCAGAGCGTAGGCCTGC-3'. Specific amplification of the targeted DNA was confirmed by electrophoresis and sequencing of the PCR product. Quantification was based on the theory that the cycle threshold value, which is defined as the cycle number required to obtain a fluorescence signal above the background, correlates inversely with the log of the initial template concentration . The relative abundance of the targeted mRNAs from several samples was determined from a standard curve that was constructed from a set of dilution series of single-stranded cDNA from one sample. To standardize the results, the relative abundance of 18S rRNA was also determined and used as the internal standard. Backmatter: PMID- 12226513 TI - Inhibition of Squalene Synthase and Squalene Epoxidase in Tobacco Cells Triggers an Up-Regulation of 3-Hydroxy-3-Methylglutaryl Coenzyme A Reductase AB - To get some insight into the regulatory mechanisms controlling the sterol branch of the mevalonate pathway, tobacco (Nicotiana tabacum cv Bright Yellow-2) cell suspensions were treated with squalestatin-1 and terbinafine, two specific inhibitors of squalene synthase (SQS) and squalene epoxidase, respectively. These two enzymes catalyze the first two steps involved in sterol biosynthesis. In highly dividing cells, SQS was actively expressed concomitantly with 3-hydroxy-3-methylglutaryl coenzyme A reductase and both sterol methyltransferases. At nanomolar concentrations, squalestatin was found to inhibit efficiently sterol biosynthesis as attested by the rapid decrease in SQS activity and [14C]radioactivity from acetate incorporated into sterols. A parallel dose-dependent accumulation of farnesol, the dephosphorylated form of the SQS substrate, was observed without affecting farnesyl diphosphate synthase steady-state mRNA levels. Treatment of tobacco cells with terbinafine is also shown to inhibit sterol synthesis. In addition, this inhibitor induced an impressive accumulation of squalene and a dose-dependent stimulation of the triacylglycerol content and synthesis, suggesting the occurrence of regulatory relationships between sterol and triacylglycerol biosynthetic pathways. We demonstrate that squalene was stored in cytosolic lipid particles, but could be redirected toward sterol synthesis if required. Inhibition of either SQS or squalene epoxidase was found to trigger a severalfold increase in enzyme activity of 3-hydroxy-3-methylglutaryl coenzyme A reductase, giving first evidence for a positive feedback regulation of this key enzyme in response to a selective depletion of endogenous sterols. At the same time, no compensatory responses mediated by SQS were observed, in sharp contrast to the situation in mammalian cells. Keywords: Introduction : In higher plants, two distinct pathways have been shown to operate concomitantly for synthesizing isopentenyl diphosphate, the common precursor for all isoprenoids. Plastid isoprenoids such as carotenoids, mono- and diterpenes, or the prenyl chains of chlorophylls and plastoquinones are formed from 2-C-methyl-d-erythritol 4-phosphate, which itself arises from the initial condensation of pyruvate with glyceraldehyde 3-phosphate (for review, see ; ). In the cytosol, isoprenoids are synthesized via the classical acetate/mevalonate (MVA) pathway, in which 3-hydroxy-3-methylglutaryl coenzyme A (CoA) reductase (HMGR) plays a key role. This enzyme is encoded by a multigene family . In this pathway, farnesyl diphosphate (FPP) occupies a central position from which specific cis- and trans-prenyltransferases dispatch isoprene units to either sterols or non-sterol isoprenoids as represented by sesquiterpenes, ubiquinone, heme a, polyprenols, or prenylated proteins . It has been recently proposed that specific classes of isoprenoids might be produced within distinct metabolic channels or metabolons, probably involving individual HMGR isoforms . Sterols represent the major end products of this multibranched pathway, but what controls the whole pathway is still far from being understood. Such a control might either concern only branch point enzymes or involve coordinated functioning of distinct metabolic channels, each one being regulated independently from one another . Figure 1 | Cytosolic isoprenoid biosynthetic pathway. Cytosolic isoprenoid biosynthetic pathway. Cytosolic isoprenoids are synthesized from acetyl CoA via the intermediate formation of MVA. IPP, Isopentenyl diphosphate; DMAPP, dimethylallyl diphosphate; IPA, isopentenyl adenine; and UQ, ubiquinone. Sq and Tb inhibit SQS and SE, respectively. The first step committed to the sterol branch of the isoprenoid pathway is catalyzed by the squalene synthase (SQS), which mediates the reductive head-to-head condensation of two molecules of FPP to form squalene via presqualene diphosphate . This reaction takes place in membranes of endoplasmic reticulum, as do all subsequent steps involved in sterol biosynthesis. Because of its particular position at the interface between hydrophilic and hydrophobic intermediates, SQS might constitute a major control point for regulating the sterol branch in directing FPP molecules into either sterols or non-sterol isoprenoids in response to changing cellular requirements. The sequence of reactions needed to convert squalene into end products is now well known , and the "state of art" on relevant enzymes and genes has been just reviewed . In contrast to animal and fungal cells, higher plants synthesize a mixture of sterols in which sitosterol, stigmasterol, and 24-methylcholesterol often predominate. These compounds play an essential role as membrane components in regulating the fluidity and permeability of membranes and the activity of membrane-associated proteins , but some sterols or biosynthetic intermediates might also serve as signal molecules during plant growth and development . Despite the critical importance of sterols, mechanisms responsible for sterol homeostasis in higher plants are still largely unknown. Recent reports have pointed out that HMGR could regulate the flux of intermediates toward the sterol branch , but whether or not this enzyme is able to respond to a depletion of sterol end products had not yet been investigated until now. To get some insight into the regulatory mechanisms controlling the sterol biosynthetic pathway, the present study was focused on the role played by the plant SQS. Early work with plant cell suspension cultures has already emphasized its involvement in plant defense reactions. A very fast inhibition of SQS activity was observed after addition of fungal elicitors. The resulting arrest of sterol biosynthesis has been interpreted as a means for the cell to either redirect metabolic intermediates, especially FPP, toward the synthesis of sesquiterpene phytoalexins or simply to leave "house-keeping" metabolism pending better conditions . This work was aimed at probing the effects of a direct inhibition of SQS in tobacco (Nicotiana tabacum cv Bright Yellow-2 [TBY-2]) cell suspension cultures. The recent availability of squalestatins (also called zaragozic acids), which are highly potent and specific inhibitors of SQS , gives the opportunity to investigate whether or not compensatory responses take place in the case of a depletion of only squalene-derived products. In particular, we wanted to check the possibility for HMGR to be a target for positive feedback regulation by endogenous sterols. Our study also comprises an investigation of responses of TBY-2 cells to an inhibition of squalene epoxidase (SE), the next committed enzyme to the sterol pathway. This enzyme catalyzes the stereospecific conversion of squalene to (3S)-2,3-oxidosqualene in the presence of molecular oxygen. It is the first oxygen-requiring step in the sterol biosynthetic pathway and, thus, might constitute a secondary regulatory level. SE can be specifically inhibited by compounds belonging to the class of allylamines . Here, we show that both squalestatin-1 (Sq) and terbinafine (Tb) are able to inhibit efficiently sterol biosynthesis in TBY-2 cells and, thus, to induce a decrease in the sterol cell content. In the case of Tb treatment, this depletion was accompanied by a concomitant accumulation of squalene and by an increase in the content and the rate of synthesis of triacylglycerols (TAG). Squalene was found to accumulate in lipid droplets but could be redirected toward sterol biosynthesis if necessary. Our results demonstrate that inhibition of either SQS or SE triggers an increase in HMGR enzyme activity, giving evidence for a positive feedback regulation of this key enzyme in response to a selective depletion of endogenous sterols. Surprisingly, we observed no compensatory responses mediated by SQS, in sharp contrast to the situation in animal systems. RESULTS : Sterol Enzyme Activities and Free Sterol Levels during a Growth Cycle of TBY-2 Cells | Figure shows a typical growth curve of a TBY-2 cell culture and the corresponding changes in four sterol enzyme activities and in free sterol levels. Microsomal fractions were prepared from cultures collected on the designated days after subculturing and used for both enzymatic assays and sterol determinations. We monitored SQS activity and the activities of HMGR and sterol methyltransferases SMT1 and SMT2 throughout the growth cycle. HMGR, an upstream enzyme, is involved in the synthesis of both sterols and non-sterol isoprenoids. Sterol methyltransferases catalyze two distinct steps downstream of squalene: the S-adenosyl-l-Met-dependent methylation of cycloartenol (SMT1) and of 24-methylenelophenol (SMT2) to give 24-methylenecycloartanol and 24-ethylidenelophenol, respectively. Figure 2 | Sterol biosynthesis in proliferating TBY-2 cells: changes in cell mass (A), sterol enzyme activities (B), and free 4-demethylsterol levels (C) as a function of time after subculturing. Sterol biosynthesis in proliferating TBY-2 cells: changes in cell mass (A), sterol enzyme activities (B), and free 4-demethylsterol levels (C) as a function of time after subculturing. Microsomal membranes were prepared from cultures collected on the designated days after subculturing and used for both measurements of sterol enzyme activities (the values are the means of two replicates and are representative of a standard experiment) and free 4-demethylsterol contents (+-se). SMT1, S-adenosyl Met-cycloartenol methyltransferase; and SMT2, S-adenosyl Met-24-methylenelophenol methyltransferase. One or 2 d after subculturing, the tobacco cell culture entered into a rapid growth phase, as attested by a large increase in fresh cell mass, and reached a stationary phase after 7 d (Fig. A). Activities of the four enzymes exhibited very similar changes over the course of a culture cycle, with a maximum found on d 5 and a sharp decrease in the stationary phase, suggesting a coordinated expression of all the enzymes of the sterol pathway in proliferating TBY-2 cells (Fig. B). In Figure C, changes in the sterol levels over the growth cycle are given. Only free Delta5-sterols, the end products of the pathway, were quantified (as micrograms per milligram of protein). As expected, these compounds mainly accumulated in the stationary phase. The sterol composition of 3-d-old TBY-2 cells is given in Table . Stigmasterol, 24-methylcholesterol and sitosterol were by far the predominant sterols. The other sterols were isofucosterol, 24-methylenecholesterol, cholesterol, and Delta7-cholesterol. Other compounds (about 1.2% of total sterols) represented the usual sterol biosynthetic intermediates: 4alpha-methylsterols (obtusifoliol, cycloeucalenol, 24-methylene-, and 24-ethylidenelophenol) and 4,4-dimethylsterols (cycloartenol and 24-methylenecycloartanol). The relative proportions of the different end products were found not to change over the growth cycle, with the exception of a slight increase (from 1.55 to 1.82) in the ratio of stigmasterol to sitosterol (data not shown). Table I | Free sterol composition of TBY-2 cells Effects of Sq | Cell Growth, Sterol Content, and Biosynthesis | Three-day-old TBY-2 cells were treated with concentrations of Sq ranging from 0 to 0.5 mum for 24 h and incubated with radioactive sodium acetate for 2 h just before cell harvesting. TBY-2 cells were collected by filtration and analyzed for their sterol composition as described in "Materials and Methods." Sterols (as the sum of free and esterified forms) were quantified by gas chromatography (GC) as acetate derivatives. Table shows that Sq severely impaired cell growth as attested by a 50% decrease in the cell mass after a treatment with 0.5 mum. A parallel dose-dependent decrease in the sterol cell content was found. The high sensitivity of TBY-2 cells toward the inhibitor is illustrated by the very fast depletion of the pool of 4,4-dimethylsterols, the early precursors of sterols, which was reduced by 88% after a treatment with only 50 nm Sq (data not shown). The efficiency of Sq as an inhibitor of the sterol pathway was confirmed by measurements of radioactivity from [14C]acetate incorporated into end products. After addition of 50 nm Sq, radioactivity associated with 4-demethylsterols accounted for only 4% compared with that of sterols from control cells . At the same time, a dose-dependent accumulation of a labeled compound was observed. This compound, which was recovered in the band corresponding to 4alpha-methylsterols, was identified as farnesol. It might originate from hydrolysis of radioactive FPP molecules not used for sterol synthesis. Table II | Effect of Sq on cell growth, sterol biosynthesis, and SQS activity SQS Activity | The target of Sq is SQS as illustrated in Table . A dose as low as 50 nm Sq was shown to be sufficient to inhibit almost completely SQS activity in microsomal fractions. Such an inhibition was found to take place in a few hours, with an IC50 value of 5.5 nm, indicating the very potent inhibitory power of Sq. This compound, which partially mimics the structure of FPP and/or presqualene diphosphate, has been described as a competitive inhibitor of SQS . Attempts were made to recover SQS activity in microsomal membranes from Sq-treated cells. However, neither extensive washes of intact cells nor dilution and additional centrifugations of microsomes were successful, suggesting that Sq also acts as a mechanism-based irreversible inactivator of plant SQS . SQS, FPP Synthase (FPPS), and HMGR Expression | To investigate whether SQS transcription could be affected by the inhibition of SQS activity, total RNA was isolated from TBY-2 cells treated with different Sq concentrations. A full-length SQS cDNA from Nicotiana benthamiana , which has 98% identity with the corresponding N. tabacum cDNA, was used as a probe. After hybridization with this probe, transcripts with a size of 1.6 kb could be detected. As shown in Figure B, the levels of SQS transcripts were not found to change significantly, despite an almost complete inhibition of SQS activity. A time-course study (from 0 to 48 h) of effects induced by 0.5 mum Sq also showed no differences between SQS mRNA levels from control and treated cells, whatever the period of time (Fig. B). Because SQS inhibition possibly induces an increased amount of FPP molecules in the cytosol, we checked the effects of Sq treatments on the transcription of FPPS. Northern-blot experiments were performed with a partial FPPS cDNA from TBY-2 cells as a probe. The corresponding mRNA levels, with a size of about 1.7 kb, remained constant whatever the dose of Sq used, suggesting that FPPS transcription was neither affected by SQS inhibition, nor by the excess of farnesol and FPP. Figure 3 | Effects of Sq on HMGR activity (A) and HMGR, SQS, and FPPS mRNA levels (B). Effects of Sq on HMGR activity (A) and HMGR, SQS, and FPPS mRNA levels (B). Three-day-old TBY-2 cells were treated with various concentrations of Sq (from 0 to 500 nm) and used to isolate both microsomal fractions and total RNA. A, HMGR activity was measured in the presence of 30 mumR,S-[3-14C]HMG-CoA and of 30 mug of microsomal protein. Control value (5.5 nmol h-1 mg-1) was set at 1. Enzyme activities were expressed as relative values to the control. Results are from two independent experiments +- sd. B, Total RNA (30 mug) was loaded per lane and transferred to a nylon membrane. Hybridizations were performed with 32P-labeled probes. All of the hybridizations were performed on the same membrane. Relative intensities obtained by PhosphorImage analysis are given in the table after correction for background and normalization relatively to EF1-alpha mRNA content. Control value was set at 1. Figure 4 | Time course of Sq effects on HMGR activity (A) and HMGR and SQS mRNA levels (B). Time course of Sq effects on HMGR activity (A) and HMGR and SQS mRNA levels (B). Three-day-old TBY-2 cells were treated during various times (from 0 to 48 h) by Sq (500 nm) and were used to isolate both microsomal fractions and total RNA. A, HMGR activity was measured as indicated in Figure . For each time, enzyme activities were expressed as relative values to the corresponding control (set at 1). Results are from two independent experiments +- sd. B, Total RNA (30 mug) was loaded per lane and transferred to a nylon membrane. Hybridizations were performed with 32P-labeled probes. All the hybridizations were performed on the same membrane. Relative intensities obtained by PhosphorImage analysis are given in the table after correction for background and normalization relatively to EF1-alpha mRNA content and compared with the corresponding control (set at 1). It is well known that HMGR constitutes the major limiting-step in cholesterol biosynthesis in mammalian cells . The inhibition of SQS and the resulting depletion of squalene-derived products were shown to induce compensatory responses mediated by HMGR . It seemed interesting to us to investigate whether similar responses could also take place in TBY-2 cells. Three-day-old TBY-2 cells were treated for 24 h with Sq concentrations varying from 0 to 0.5 mum and were used to isolate both total RNA and microsomal fractions, for HMGR activity measurements. As shown in Figure A, microsomal fractions from Sq-treated TBY-2 cells were found to exhibit 2- to 4-fold increased HMGR enzyme activities compared with control cells. The highest stimulation rate was observed at 75 nm. Such a stimulation of HMGR activity could be detected as early as 6 h after Sq administration and progressively increased until 24 h (Fig. A). For northern-blot experiments, we used as a probe a 1,400-bp fragment corresponding to the C-terminal part (catalytic site) of a Nicotiana sylvestris HMGR cDNA . HMGR transcripts with a size of about 2.5 kb could be detected. Figure B shows a dose-dependent increase in HMGR mRNA levels from Sq-treated TBY-2 cells. A 5-fold increase was found in cells treated with 500 nm Sq. At this concentration, the highest level of transcripts was observed 24 h after Sq administration (Fig. B). Taken together, these results indicate that the inhibition of SQS by Sq triggered a stimulation of both HMGR steady-state mRNA and enzyme activity, suggesting that the arrest of sterol biosynthesis and the resulting depletion of squalene-derived products exerted a positive feedback regulatory effect on the transcription of HMGR. Surprisingly, this inhibition did not activate SQS transcription. Effect of Tb | In Vivo Sterol Biosynthesis | Three-day-old TBY-2 cells were treated with concentrations of Tb ranging from 0 to 30 mum for 30 h, then collected by filtration and analyzed for their free sterol composition as described in "Materials and Methods." Squalene and free sterols (as acetate derivatives) were quantified by GC. Treatment of TBY-2 cells with Tb was found to induce a dose-dependent accumulation of squalene and a progressive decrease in the content of end products (Fig. A). After treatment with 30 mum Tb, the squalene content amounted to 5 mg g-1 dry weight, whereas it was barely detectable in control cells (about 20 mug g-1 dry weight). At the same time, the remaining Delta5-sterols accounted for only 0.85 mg g-1 dry weight compared with 1.8 mg g-1 for control cells, corresponding approximately to a 50% decrease in the usual free sterol content. Such an accumulation of squalene clearly indicates an inhibition of SE by Tb, leading to an arrest of end product biosynthesis. No change in relative proportions of Delta5-sterols was found (data not shown). It is remarkable to see that TBY-2 cells can accommodate such high squalene intracellular concentrations because no inhibition of cell growth or cell death was observed (data not shown). Figure 5 | Effects of Tb on sterol composition (A) and biosynthesis (B). Effects of Tb on sterol composition (A) and biosynthesis (B). Three-day-old TBY-2 cells were treated with concentrations of Tb ranging from 0 to 30 mum for 24 h and incubated with sodium [1-14C]acetate for 2 h before cell harvesting. Sterols were extracted as indicated in "Materials and Methods." A, Squalene and free sterols (acetate derivatives) were quantified by GC. Sterols amounts are expressed in milligrams per gram dry weight. B, Radioactivities incorporated into sterols and precursors are expressed as 105 cpm g-1 dry weight (dwt). Figure shows a time course of squalene accumulation in TBY-2 cells treated with 3 mum Tb. A significant increase in the squalene content occurred as soon as 6 h after administration of the inhibitor and continued up to 48 h. After 96 h, no additional accumulation of squalene was observed (data not shown). Figure 6 | Time course of squalene accumulation after Tb administration. Time course of squalene accumulation after Tb administration. Three-day-old TBY-2 cells were treated by 30 mum of Tb for different periods of time in parallel with a control. Sterols were extracted as indicated in "Materials and Methods." Squalene and free sterols (acetate derivatives) were quantified by GC. Sterol amounts are expressed in milligrams per gram dry weight (dwt). Further evidence for the SE inhibition was obtained by feeding control and Tb-treated TBY-2 cells with [14C]acetate for 2 h before cell harvesting. In control cells, most of the acetate radioactivity was incorporated into free sterols, whereas in Tb-treated cells, squalene was by far the most labeled compound (Fig. B). After treatment with 3 mum Tb, about 71% of the radioactivity recovered in the sterol branch was already associated with squalene. At this concentration, a time-course study of squalene biosynthesis demonstrated that the highest rate of synthesis was obtained after 24 h of treatment (data not shown). After treatment with 30 mum, 97% of the radioactivity was present in squalene and only 2% in the end products. Cycloartenol Accumulation at Low Tb Concentration | As shown in Figure , treatment of TBY-2 cells with 3 mum Tb was found to induce a significant accumulation of 4,4-dimethylsterols. These compounds were mainly represented by cycloartenol and 24-methylenecycloartanol . GC analysis of the corresponding acetate derivatives indicated a specific increase in the cycloartenol content (data not shown), suggesting that SMT1 involved in the methylation of cycloartenol to give 24-methylenecycloartanol might be down-regulated. To get more insight into this process, TBY-2 cells were treated with 3 mum Tb for periods of time from 0 to 48 h. A 2-h pulse of radioactive acetate was given just before cell harvesting. After lipid extraction, the fraction of 4,4-dimethylsterols was analyzed in more detail. The acetate derivatives were separated on AgNO3-impregnated thin-layer chromatography (TLC) plates. Cycloartenol and 24-methylenecycloartanol were eluted and quantified by GC, and their radioactivity was measured by scintillation counting. As shown in Figure A, the cycloartenol content of treated cells was found to increase progressively in function of the duration of contact with the inhibitor. In contrast, cycloartenol did not accumulate in control cells (Fig. A). As a consequence, the radioactivity associated with cycloartenol declined rapidly in control cells, whereas in treated cells, cycloartenol continued to be actively synthesized for as long as 24 h (Fig. B). Whatever the period of time, only low amounts of radioactivity were detected in 24-methylenecycloartanol. Figure 7 | Time course of content (A) and biosynthesis (B) of 4,4-dimethylsterols after administration of 3 mum Tb. Time course of content (A) and biosynthesis (B) of 4,4-dimethylsterols after administration of 3 mum Tb. Three-day-old TBY-2 cells were treated with 3 mum Tb for different periods of time and incubated with sodium [1-14C]acetate for 2 h before cell harvesting. A, Levels of the three major 4,4-dimethylsterols were quantified by GC (acetate derivatives) and expressed in micrograms per gram dry weight. B, Radioactivities incorporated into cycloartenol and 24-methylenecycloartanol were expressed as 105 cpm g-1 dry weight (dwt). 24-me cycloartanol, 24-Methylenecycloartanol. To investigate whether the methylation of cycloartenol by the SMT1 could be inhibited in TBY-2 cells treated by low Tb concentrations, microsomal fractions were isolated from control and Tb-treated cells and tested for their SMT1 enzyme activities. Similar enzyme activities were found (data not shown), indicating that the SMT1 protein remained active despite the Tb treatment of tobacco cells. We also checked that 30 mum Tb had no direct inhibitory effect on SMT1 activity. Thus, these data indicate that the inhibition of the methylation reaction occurs only in intact treated cells and could result from a secondary regulatory effect. SQS and HMGR Expression | As stated above, treatment of TBY-2 cells with Tb triggered the accumulation of impressive amounts of squalene, the product of the reaction catalyzed by SQS. In control cells, endogenous squalene was barely detectable. Despite such large increases of squalene, SQS was found to exhibit a constant enzyme activity, similar to that of control cells and whatever the duration of the Tb treatment (Fig. B). At the same time, no significant change in the SQS steady-state mRNA levels was observed (Fig. C), indicating no negative regulatory effect on the gene transcription by squalene. Figure 8 | Time course of Tb effects on HMGR (A) and SQS (B) activities and mRNA levels (C). Time course of Tb effects on HMGR (A) and SQS (B) activities and mRNA levels (C). Three-day-old TBY-2 cells were treated for various time periods (from 0 to 48 h) with 30 mum Tb and used to isolate both microsomal fractions and total RNA. A, HMGR activity was measured as indicated in Figure . For each point, enzyme activities were expressed as relative values to the corresponding control (set at 1). Results are from three independent experiments +- sd. B, SQS activity was measured in the presence of [1-3H]FPP and of 10 mug of microsomal protein. For each time, enzyme activities were expressed as relative values to the corresponding control (set at 1). Results are from three independent experiments +- sd. C, Total RNA (30 mug) was loaded per lane and transferred to a nylon membrane. Hybridizations were performed with 32P-labeled probes. All the hybridizations were performed on the same membrane. Relative intensities obtained by PhosphorImage analysis are given in the table after correction for background and normalization relative to EF1-alpha mRNA content and compared with the corresponding control (set at 1). We also investigated effects of Tb treatments on HMGR expression. As shown in Figure A, a 2- to 4-fold increase in the HMGR enzyme activity was observed in TBY-2 cells treated with 30 mum Tb. The stimulation occurred already after 6 h and reached a maximum after 24 h. However, this increase in enzyme activity was not correlated with significant modifications of the corresponding mRNA levels (Fig. C). TAG Synthesis | Besides the accumulation of squalene, Tb treatment of TBY-2 cells also triggered a significant dose-dependent increase in the TAG content, with an 8-fold enhancement of the mean value measured for control cells after administration of 30 mum . This increase directly resulted from a stimulation of their de novo biosynthetic rate, as indicated by the parallel increase in their [14C]radioactivity . A concomitant accumulation of lipid droplets in the cytosol of treated cells was found to take place. Light microscopy observations showed the presence of many orange spheres after staining of cells with Sudan IV, a lipid-specific dye . Very few or none of these lipid droplets were seen in control cells. These results suggest that relationships between sterol and TAG biosynthetic pathways might occur in vivo. Table III | Effects of Tb on TAG content and biosynthesis Figure 9 | Lipid particles in the cytosol of TBY-2 cells treated with Tb. Lipid particles in the cytosol of TBY-2 cells treated with Tb. Three-day-old TBY-2 cells were treated for 24 h with Tb 30 mum and then observed in optical microscopy after staining by Sudan IV, in parallel with a control. Lipid particles appear as spherical orange droplets. Bar = 20 mum. Squalene Intracellular Localization | To address the question of the intracellular localization of the overproduced squalene, TBY-2 cells were treated with 30 mum Tb for 24 h before being used for isolation of a microsomal fraction. After sedimentation at 100,000g, lipid particles appeared as a fluffy lipid layer at the surface of the corresponding supernatant. Both fractions were then analyzed for their squalene content. In control cells, squalene could be not detected in the supernatant. In contrast, most (i.e. higher than 90%) of the squalene from treated cells was recovered in the lipid droplets. The remaining 10% were associated with the microsomal fraction. It seemed to us interesting to investigate whether or not the squalene stored in these lipid particles could be reused as a precursor for sterol biosynthesis. TBY-2 cells were first treated with 30 mum Tb for 24 h and fed with radioactive acetate for 2 h just before cell harvesting. They were then extensively washed and resuspended in a Murashige and Skoog medium containing 0.5 mum Sq to block the synthesis of endogenous squalene. These cells were allowed to grow for 12, 24, and 48 h, respectively, collected by filtration, and analyzed for their content in squalene and in free sterols and their precursors (4,4-dimethyl- and 4alpha-methylsterols). Radioactivity associated with each class of compounds was also measured. A sample of TBY-2 cells treated with Tb for 24 h was taken as a control (t0h). As shown in Table , TBY-2 cells collected after 12 and 24 h exhibited dramatic decreases in both the content and radioactivity of squalene, whereas concomitant and similar increases were found for the sterol end products. A 2-fold increase in the level of 4-demethylsterols occurred after 24 h, despite the inhibition of SQS by Sq, and these compounds contained the most radioactivity initially associated with squalene. In the further 24 h (t48h), their amount was reduced by 50% because no more squalene was available. The data from Table also indicate a transient increase in both the contents and radioactivities of the sterol biosynthetic precursors, giving evidence for a restart of an active sterol biosynthesis from the pool of radioactive squalene, triggered by the Tb removal. Finally, it should be emphasized that such a restoration of the sterol biosynthesis was accompanied by a parallel decrease in the content and radioactivity of TAG, indicating once more likely relationships between both pathways. Taken together, these results clearly demonstrate that squalene, which was previously stored in cytoplasmic lipid particles, could be remobilized for an active sterol biosynthesis. Even if the fate of TAG molecules remains to be investigated, it appears that these lipid particles actually do constitute a pool of available metabolic intermediates. Table IV | Redirection of squalene from lipid particles to the sterol pathway DISCUSSION : In contrast to the situation in animal cells, much less is known regarding regulation of the sterol pathway in plants. In that context, we planned to investigate the potential regulatory role played by the plant SQS for the following reasons: (a) SQS is the first committed enzyme to the sterol branch of the isoprenoid pathway and as such, may play a critical role in directing FPP molecules in either sterol or non-sterol isoprenoids in response to changing cellular requirements ; (b) because sterols are major isoprenoid end products, most part of precursors goes through SQS; (c) SQS has been known to participate in plant defense reactions against pathogens ; and (d) in mammalian cells, SQS is coordinately regulated with HMGR by a sterol feedback mechanism . Moreover, inhibition of SQS would not deprive the cell of important non-sterol compounds such as isoprenylated proteins, dolichol, heme a, or ubiquinone. As a plant material, we used tobacco BY-2 cell suspension cultures. This cell line was originally selected for its very short cell cycle (about 15 h; ). Such a suspension was, therefore, particularly suitable for metabolic studies and labeling experiments. TBY-2 cells were first checked for their ability for sterol synthesis. As shown in Figure , SQS and SMT1 and SMT2 enzyme activities were expressed during the entire cell growth, with a maximum at 5 d after subculturing. HMGR, which is involved in the synthesis of both sterols and non-sterol isoprenoids, exhibited a very similar expression profile, suggesting that all the enzymes of the sterol pathway were coordinately regulated, to sustain an active synthesis of membranes in rapidly dividing cells. To investigate the regulatory response of SQS to depletion of sterols, TBY-2 cells were treated with Sq. This inhibitor belongs to the class of Sqs, a family of fungal metabolites recently discovered. These compounds, which are analogs of FPP and presqualene diphosphate, are potent competitive inhibitors of mammalian SQS . We show here that Sq is also a strong inhibitor of the tobacco SQS. The high efficiency of Sq as an inhibitor of sterol synthesis was revealed by the rapid and dramatic decrease in the radioactivity from [14C]acetate incorporated into sterols after treatment with nanomolar concentrations of Sq, resulting in a decrease in the sterol cell content . At the same time, SQS activity rapidly became barely detectable. Sq was found to inhibit tobacco SQS with an IC50 of 5.5 nm (data not shown), a value similar to that obtained for mammalian SQS . Sq was also found to rapidly impair cell growth of TBY-2 cells . Such an effect on cell growth probably resulted from an inhibition of cell division. When given to synchronized TBY-2 cells, this inhibitor triggers an arrest of the cell cycle specifically in the G1/G0 phase, but without inducing cytotoxicity or cell death . An intriguing question is related to the fate of FPP molecules, which do not contribute to the build-up of sterols. First, the increase in cytosolic FPP resulting from Sq treatments does not appear to regulate negatively FPPS expression because no changes in corresponding steady-state mRNA levels occurred, whatever the Sq concentration . However, FPP might have an effect on MVA kinase, an enzyme upstream in the pathway because it has been shown to be a competitive inhibitor of this enzyme with respect to ATP . As already stated , FPP serves as a substrate for a variety of non-sterol isoprenoids. As a consequence, a redirection of FPP toward such pathways could appear to be likely. We have previously shown that exogenous farnesol could be incorporated into sterols but also into the prenyl side chain of ubiquinone Q10 and proteins from TBY-2 cells . When radioactive farnesol was given to tobacco cells in the presence of 0.5 mum Sq, no increase in the label of ubiquinone and proteins occurred, indicating that no additional FPP molecules were redirected toward these compounds (M.-A. Hartmann, unpublished data). These results are similar to those from obtained with brain cells. Under the same conditions, these authors also demonstrated that Sq had no effect on the synthesis of dolichol-phosphate. Our labeling experiments showed that a significant part of FPP molecules were hydrolyzed in response to Sq treatment, as attested by the dose-dependent accumulation of radioactive farnesol (data not presented). Similar observations were made in the case of Sq-treated mammalian cells . Such a hydrolysis might be catalyzed by a FPP diphosphatase, which could be induced by the stress caused by the Sq treatment . However, it should be pointed out that farnesol is also known to have deleterious effects. In particular, when added to TBY-2 cells at a concentration higher than 20 mum, farnesol induces cell death . As a consequence, its level in the cytosol has to be closely controlled. However, the possibility of a conversion of FPP or farnesol into other metabolites should not be excluded. In sharp contrast to SQS inhibition by Sq, treatment of TBY-2 cells by Tb, an inhibitor of SE, did not affect cell growth. Tb belongs to the class of compounds termed allylamines, which have significant impact as antifungal drugs . Tb is a reversible, noncompetitive inhibitor of SE . We show here that Tb is also active in plant systems, as attested by a dose-dependent decrease in the free sterol content and by a concomitant accumulation of squalene. These data are in agreement with previous results obtained with celery (Apium graveolens) cell suspension cultures or wheat (Triticum aestivum) seedlings . However, under our conditions, no accumulation of Delta5,7-sterols could be observed, and the absence of unusual intermediates like Delta8- or Delta8,14-sterols suggests that Tb had no secondary target in TBY-2 cells . Such high amounts of squalene seemed not to be toxic for the cell. We observed that inhibition of SE by Tb in TBY-2 cells was accompanied by a proliferation of cytosolic lipid droplets , in which squalene accumulated. Our results provide evidence that squalene could be remobilized for an active sterol biosynthesis in response to a depletion of end products . Therefore, lipid particles actually do constitute a pool of available biosynthetic intermediates and not only a metabolically inactive storage compartment, in agreement with recent data in yeast (Saccharomyces cerevisiae) from . Concomitant to the squalene accumulation, a dose-dependent increase in the TAG content and rate of synthesis was observed . These TAG probably also accumulated in lipid droplets because they could not be detected in microsomes. The mechanisms whereby the inhibition of SE induces the TAG synthesis still remain to be elucidated. However, it should be emphasized that a stimulation of TAG synthesis was also observed in leek (Allium porrum) seedlings treated with fenpropimorph, another inhibitor of sterol biosynthesis (M.-A. Hartmann, A.-M. Perret, J.-P. Carde, C. Cassagne, and P. Moreau, unpublished data), suggesting that some regulatory relationships between sterol and TAG biosynthetic pathways might occur in plants. Because Sq and Tb treatments of TBY-2 cells were found to induce significant decreases in the sterol cell content, it was interesting to investigate whether compensatory responses could be mediated by HMGR. In mammals, HMGR is the major rate-limiting enzyme in the cholesterol biosynthetic pathway. This enzyme is encoded by a single gene. It is well established that reductase activity is controlled through multivalent feedback regulation, involving both sterols and non-sterol compounds derived from MVA . In sharp contrast to animal systems, the occurrence of multiple genes encoding HMGR is a general feature of higher plants. Individual genes have been shown to exhibit different expression patterns in response to physiological and environmental stimuli such as light, plant growth regulators, wounding, or pathogen attack . It has been proposed that the different HMGR isoforms might play distinct roles associated with the production of specific isoprenoid compounds . Previous work had given evidence for an involvement of plant HMGR in controlling the flux of intermediates toward sterol biosynthesis . Tobacco plants overexpressing either the HMGR1 gene from Hevea brasiliensis or a truncated HMGR gene from guinea pig synthesize higher amounts of sterols and sterol precursors, which accumulate as steryl esters in cytosolic lipid bodies . However, no change in the free sterol content was observed. Thus, sterol acylation appears as a means for the cell to maintain sterol homeostasis. We show here for the first time, to our knowledge, that tobacco HMGR is able to respond to a selective depletion of endogenous sterols. Decreases in squalene and squalene-derived compounds resulting from treatments of TBY-2 cells with Sq or Tb triggered 2- to 4-fold increases in HMGR enzyme activity (Figs. A, A, and A). In the case of SQS inhibition by Sq, an enhancement of corresponding mRNA levels was observed, giving evidence for an activation of HMGR gene transcription. In rat liver, a similar stimulation of HMGR transcription was previously reported in response to Sq . On the other hand, SE inhibition by Tb was not accompanied by changes in the HMGR transcripts, suggesting that the regulatory response mediated by HMGR could be exerted at a translational or posttranslational (i.e. catalytic efficiency or protein degradation) level. To our knowledge, whereas plant HMGR appears to be mainly transcriptionally controlled , such a feedback regulatory effect occurring at the HMGR protein level in response to a depletion of end products has not yet been reported in plants. Thus, our results show differential regulatory responses induced by both inhibitors. and have presented convincing evidence that the mammalian SQS genes are regulated by their transcription rate in response to exogenous sterols and inhibitors of sterol synthesis. Surprisingly, our results indicate that tobacco SQS exhibits a different behavior. We show that SQS mRNA levels were not altered after treatment of TBY-2 cells with either Sq or Tb. Thus, the expression pattern of SQS appears to be insensitive to both the inhibition of SQS activity and the accumulation of squalene, the product of the reaction. Such an impressive accumulation of squalene, which was induced by the inhibition of SE by Tb, can likely be attributed to the enhancement of HMGR expression, leading to a greater amount of enzyme protein and, thus, to a higher synthesis of early intermediates. At the same time, no change in the SQS enzyme activity occurred, indicating that SQS would not be a limiting step for sterol synthesis in tobacco cells. However, as described above, squalene did not accumulate in endoplasmic reticulum membranes but in lipid droplets. Thus, SQS could not "sense" the excess of squalene. In the same context, no changes in SQS activity were observed when TBY-2 cells were treated with Lab 170250F, an inhibitor of obtusifoliol 14-demethylase , an enzyme downstream of squalene, or with mevinolin, an inhibitor of HMGR (L. Wentzinger and M.-A. Hartmann, unpublished data). Taken together, these data indicate that SQS is regulated differently in TBY-2 cells and in mammalian cells . The next step will be to check whether similar regulatory mechanisms for SQS also take place in intact plants. Genes encoding enzymes making up a specific metabolon must have similar transcriptional networks to coordinate expression of the metabolic unit. In mammals, the cholesterol feedback system is mediated by a family of membrane-bound transcription factors known as sterol regulatory element (SRE)-binding proteins, which recognize a 10-bp sequence (SRE) within the target genes . It has been just reported that such a SRE-binding protein activation mechanism concerns every step of the cholesterol biosynthetic pathway . In the promoters of plant HMGR genes, no consensus sequences similar to these SRE or other cis regulatory elements from animal sterol-regulated genes have been found so far . Moreover, many enzymes involved in plant sterol biosynthesis are encoded by multiple genes . For instance, this is the case for SQS and SE . The biological significance of such a plethora of genes remains to be elucidated. Higher plants have probably evolved specific mechanisms for regulating their complex isoprenoid pathway, mechanisms that remain to be discovered. In this challenging context, we are currently investigating in more detail sterol homeostasis in plant cells. MATERIALS AND METHODS : Chemicals | All chemicals were purchased from Sigma (St Louis). Sodium [1-14C]acetate (54 Ci mol-1), [3-14C]HMG-CoA and S-adenosyl l-[methyl-3H]Met were from Amersham (Buckinghamshire, UK). [1-3H]FPP was from Isotopchim (Ganagobie-Peyruis, France). Sq was obtained from Glaxo (Greenford, Middelsex, UK) and dissolved in 0.1 m Tris-HCl (pH 7.4) to give a 0.1 mm. stock solution. Tb was kindly supplied by Dr. N.S. Ryder (Vienna) and dissolved in dimethyl sulfoxide at a concentration of 50 mm. Final dimethyl sulfoxide concentrations did not exceed 0.05% (v/v). Plant Material | Cell suspension cultures of tobacco (Nicotiana tabacum cv Bright Yellow-2 [TBY-2]) were usually grown in 250-mL Erlenmeyer flasks containing 80 mL of modified Murashige and Skoog medium at 26C in the dark and subcultured weekly as reported . Cells were harvested by filtration. In all cases, sterol inhibitors were added to 3-d-old cell cultures. In Vivo Labeling Experiments | Cells were usually incubated with sodium [1-14C]acetate (5 muCi, 0.2 mm) for 2 h just before cell harvesting. Unincorporated radioactivity was removed by washing cells on the filter. Isolation of Microsomes | Frozen cells were ground in a mortar in the presence of liquid N2. The powder was resuspended in a medium consisting of 0.25 m Suc, 4 mm EDTA, 100 mm potassium fluoride, 40 mm sodium ascorbate, and 0.2% (w/v) bovine serum albumin in 0.1 m Tris-HCl (pH 8.0; 10 mL g-1 fresh wt). After filtration through a nylon blutex, the homogenate was centrifuged at 10,000g for 25 min. The resulting supernatant was centrifuged at 100,000g for 60 min. The microsomal pellet was resuspended in 0.1 m Tris-HCl (pH 7.5) containing 1.5 mm dithioerythritol and 20% (w/v) glycerol and stored at -80C until use. Protein concentrations were determined according to with bovine serum albumin as a standard. Isolation of Lipid Particles | Frozen cells were ground in a mortar in the presence of liquid N2. The powder was homogenized in 0.1 m Tris-HCl (pH 8.0) containing 1 mm EDTA. After filtration through a nylon blutex, the homogenate was centrifuged at 10,000g for 25 min. The resulting supernatant was centrifuged at 100,000g for 60 min. The white fat pad at the top of the tube was collected and lyophilized for lipid analyses. The microsomal fraction was resuspended in 0.1 m Tris-HCl (pH 8.0) and 1 mm EDTA, and centrifuged at 100,000g for 60 min. The microsomal pellet was freeze-dried before lipid analysis. Lipid Analyses | Freeze-dried material was ground and extracted by refluxing twice with dichloromethane:methanol (2:1, v/v) for 2 h. Extracts were combined, dried under reduced pressure, and thoroughly washed at room temperature with hexane to recover squalene, free sterols, and TAG. Sterols were isolated and quantified as previously reported . After recovering from total lipids with hexane, sterols were loaded on TLC plates, which were developed in dichloromethane (two runs). Purified sterols were then eluted and acetylated before being analyzed by GC on a glass capillary column (30 m long, 0.25 mm i.d., coated with DB-1). The temperature program used includes a fast rise from 60C to 230C (30C min-1), then a slow rise from 230C to 280C (2C min-1). A cholesterol standard was added to the samples before analysis. Sterols were identified by GC-mass spectroscopy . Squalene and steryl esters were separated by TLC with cyclohexane:toluene (95:5, v/v) as the solvent. Squalene (RF 0.5) was eluted and quantified by GC. The radioactivity was measured by liquid scintillation spectrometer. TAG were quantified using the colorimetric assay from Sigma (kit 336-10). Assays for Enzyme Activities | HMGR and SQS activities were measured according to and , respectively. SMT activities were measured in the presence of 0.1 m Tris-HCl (pH 7.5) containing 100 mum [3H-methyl]Ado-Met (1 muCi), 0.1% (w/v) Tween 80, 1 mm 2-mercaptoethanol, 50 to 100 mug of microsomal membranes, and 100 mum cycloartenol (SMT1) or 50 mum 24-methylenelophenol (SMT2), in a total volume of 100 muL. Incubations were carried out at 30C for 1 h and stopped by adding 12% (w/v) KOH in ethanol. The neutral lipids were extracted with hexane and loaded on TLC plates. The bands of 4,4-dimethyl sterols (SMT1) or 4alpha-methylsterols (SMT2) were scrapped off the plate and their radioactivities were measured by liquid scintillation counting. Northern Blots | Total RNA was isolated from TBY-2 cells using the guanidine thiocyanate-phenol-chloroform method . It was analyzed (30 mug) by formaldehyde-agarose gel electrophoresis and blotted onto Hybond-N membranes (Amersham). Radiolabeled cDNA probes were prepared by a random priming method . The nylon membranes were hybridized overnight with a 32P-labeled probe (106 cpm mL-1) in a solution containing 5x Denhardt solution, 6x SSC, 0.5% (w/v) SDS, and 5 mg mL-1 denatured salmon sperm DNA, under low stringency conditions (55C). Membranes were washed twice with 2x SSC and 0.1% (w/v) SDS at room temperature, twice with 0.2x SSC and 0.1% (w/v) SDS at 45C for 30 min. Transcript levels were quantified from the blots using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA), after an overnight exposure. The membranes were boiled for 3 min in 0.1% (w/v) SDS to remove the bound probe and to reuse them for other hybridizations. Data were normalized for EF1-alpha mRNA content and corrected for background. The probes, which were used, are the following: (a) HMGR, a 1.4-kb fragment corresponding to the catalytic part of a Nicotiana sylvestris HMGR cDNA ; (b) SQS, a 1.1-kb XhoI/SpeI fragment of a Nicotiana benthamiana SQS cDNA ; (c) EF1-alpha, a 1.7-kb full length of an Arabidopsis EF1-alpha cDNA; and (d) FPPS, a 390-bp cDNA fragment obtained by PCR amplification of a TBY-2 cDNA library with degenerate oligonucleotides: B1 [5'-(T/C) TT(T/C)(T/C) TIGTII(C/T) IGA(T/C) GA(T/C) ATIAATTGA] and E1 [5'-TA(A/G) TC(A/G) TC(T/C) TGIAT(T/C) GA(A/G) AA] as described by . Optical Observations | TBY-2 cells were stained with Sudan IV (70% [v/v] ethanol). Lipid droplets appear as orange spherical granules on light microscopy. Backmatter: PMID- 12226514 TI - Probing in Vivo Metabolism by Stable Isotope Labeling of Storage Lipids and Proteins in Developing Brassica napusEmbryos AB - Developing embryos of Brassica napus accumulate both triacylglycerols and proteins as major storage reserves. To evaluate metabolic fluxes during embryo development, we have established conditions for stable isotope labeling of cultured embryos under steady-state conditions. Sucrose supplied via the endosperm is considered to be the main carbon and energy source for seed metabolism. However, in addition to 220 to 270 mm carbohydrates (sucrose, glucose, and fructose), analysis of endosperm liquid revealed up to 70 mm amino acids as well as 6 to 15 mm malic acid. Therefore, a labeling approach with multiple carbon sources is a precondition to quantitatively reflect fluxes of central carbon metabolism in developing embryos. Mid-cotyledon stage B. napus embryos were dissected from plants and cultured for 15 d on a complex liquid medium containing 13C-labeled carbohydrates. The 13C enrichment of fatty acids and amino acids (after hydrolysis of the seed proteins) was determined by gas chromatography/mass spectrometry. Analysis of 13C isotope isomers of labeled fatty acids and plastid-derived amino acids indicated that direct glycolysis provides at least 90% of precursors of plastid acetyl-coenzyme A (CoA). Unlabeled amino acids, when added to the growth medium, did not reduce incorporation of 13C label into plastid-formed fatty acids, but substantially diluted 13C label in seed protein. Approximately 30% of carbon in seed protein was derived from exogenous amino acids and as a consequence, the use of amino acids as a carbon source may have significant influence on the total carbon and energy balance in seed metabolism. 13C label in the terminal acetate units of C20 and C22 fatty acids that derive from cytosolic acetyl-CoA was also significantly diluted by unlabeled amino acids. We conclude that cytosolic acetyl-CoA has a more complex biogenetic origin than plastidic acetyl-CoA. Malic acid in the growth medium did not dilute 13C label incorporation into fatty acids or proteins and can be ruled out as a source of carbon for the major storage components of B. napus embryos. Keywords: Introduction : Plant oils represent the largest renewable resource of highly reduced carbon chains and there is interest in increasing their production by oilseed crops. Brassica napus (canola, oilseed rape) is a major oil crop and a multitude of literature focuses on the biochemistry and physiology of oil accumulation in developing seeds of B. napus (see ; ; ; ; ). Although the biochemical pathways leading from Suc to oil storage are largely understood, a number of questions remain regarding, for example, the subcellular organization of reactions, and the origin of acetyl-CoA, reducing power, and ATP for fatty acid synthesis. These questions are particularly difficult to address using standard in vitro or organellar biochemical analyses that often result in loss of key activities. In addition, due to several reasons including dilution of isotopes by internal metabolite pools, many classical radioisotope tracer experiments may have led to major misinterpretations of in vivo metabolic fluxes. New stable isotope labeling methods have been developed to dissect a number of aspects of in vivo intermediary metabolism . Most frequently, 13C-labeled substrates are fed to cells followed by NMR or gas chromatography/mass spectrometry (GC/MS) analysis of the products of this metabolism. Because the carbon atoms incorporated into amino acids and fatty acids can be traced back to the structures of a number of central intermediates, different pathways of intermediary metabolism can be said to differently "imprint" the pattern of 13C recorded in end products of metabolism. In addition, under metabolic and isotopic steady-state conditions, fluxes through the central carbon metabolic network can be quantified. So far, the more evolved techniques of metabolic flux analysis with stable isotope labeling have been applied almost exclusively to microorganisms or plant cell cultures, where physiological and growth conditions can be controlled very exactly. If the steady-state condition is fulfilled, the labeling pattern in amino acids and other products gives important information on carbon fluxes in the central carbon metabolism during growth. In principal, these techniques can be extended toward growth of a plant organ on a general 13C-labeled carbon source under conditions similar to those found in planta. Recently, reported initial results in this direction by labeling of maize (Zea mays) kernels. Developing embryos of B. napus take up nutrients from the liquid endosperm, which surrounds them . After the mid-cotyledon stage, rapid accumulation of storage lipid and proteins occurs , resulting in mature seeds with 40% to 50% (w/w) oil and 30% (w/w) protein . During storage product accumulation, Suc at high concentration is the predominant sugar in the seed and has been considered the main carbon source for the embryo . However, in addition to sugars, considerable amounts of amino acids and organic acids have been reported to be present in the endosperm liquid of Phaseolus vulgaris and, as reported in this paper, are also present in the endosperm liquid of B. napus. Therefore, in addition to Suc, the embryo may take up at least a part of carbon in the form of amino acids and/or organic acids. The uptake and influence of amino acids is also suggested by the presence of amino acid transporters in developing embryos of Arabidopsis and by the finding that protein content is correlated with the concentration ratio of amino acids to sugars in the phloem sap between different genotypes of B. napus . In addition, it has been shown for rapeseed plants that with the beginning of seed filling, nitrogen is moved from roots, stem, and leaves into the pods, while the uptake of NO3- into plants comes to an end . This suggests that most or all the nitrogen is imported into the developing embryo in the form of amino acids. During oil and protein accumulation in the developing embryo, both fatty acid and protein biosynthesis have high demands for precursor molecules and cofactors (NADH, NADPH, and ATP). If amino acids are taken up from the liquid endosperm, the demand of precursors and cofactors for protein synthesis will be much lower than if the amino acids must be synthesized de novo from Suc and nitrate. Thus, the carbon economy of the developing embryo is dependent on which sources of carbon, nitrogen, and energy are used. As a consequence, unless all natural carbon sources are considered in the composition of the growth medium for labeling experiments, even very precise measurements of carbon flux ratios in intermediary metabolism would be of reduced significance. Therefore, in this study, we present the analysis of organic constituents of EL and the design of a liquid growth medium that mimics the in planta liquid environment. With this medium, embryos can be grown on different isotopically labeled compounds and accumulate milligram amounts of labeled storage lipids and storage proteins. By supplying uniformly 13C-labeled Glc with either unlabeled amino acids or malate and subsequent measurement of the 13C enrichment in fatty acids and amino acids by GC/MS, we have determined the contribution of the different carbon sources to fatty acid and storage protein biosynthesis. In addition to the measurement of 13C enrichment, the fractional 13C labeling of fatty acids and proteinogenic amino acids can be investigated by GC/MS to provide additional quantitative information on fluxes through alternative pathways of central carbon metabolism in developing B. napus embryos. RESULTS AND DISCUSSION : Analysis of EL for Sugars, Amino Acids, and Organic Acids | To understand which carbon sources are available in planta to developing B. napus embryos, we analyzed the endosperm liquid after dissection of seeds of greenhouse-grown B. napus plants. Sugars | As shown in Table , endosperm liquid of seeds at the beginning of oil accumulation (embryos of 0.1 --0.5 mg fresh weight, mid-cotyledon stage, 20 DAF) contains Glc and Fru as the main sugars and at similar concentrations. As development proceeds, the concentration of Glc and Fru decreases severalfold, whereas that of Suc increases about 10-fold such that for seeds in the late cotyledon stage (>3 mg fresh weight, late cotyledon, 26 DAF), Suc dominates over Glc and Fru. The change in the ratio of hexoses to Suc is similar to that observed earlier by in whole seeds of B. napus as well as by in endosperm liquid. After 26 DAF, the growth of the embryo comes to an end and the embryo takes up most of the volume inside the seed coat. In this stage, the embryos have maximal oil accumulation . Similarly, in Vicia faba, a shift from high hexose to Suc ratio to a high Suc to hexose ratio is believed to govern the developmental process from cell division to cell expansion and accumulation of storage compounds . Table I | Concentration of several organic constituents in the endosperm liquid Amino Acids | TLC of endosperm liquid and staining with ninhydrin revealed that in all stages of embryo development, Gln is the main amino acid constituent of the endosperm liquid. The concentration of l-Gln was determined enzymatically and ranged between 20 (mid-cotyledon stage) and 36 mm at late cotyledon stage . In addition, the concentrations of 16 other proteinogenic amino acids were determined at mid- and late cotyledon stage and the sum totaled 40 and 70 mm, respectively, with Gln, Glu, and Ala as the dominating components . Thus, the amino acids in the endosperm liquid represent substantial possible sources of carbon and reduced nitrogen for seed storage product biosynthesis. In contrast to the endosperm liquid, in phloem sap of B. napus, Gln, Glu, Ser, Thr, and Asp have been found as the major amino acids . Malic Acid | By TLC of endosperm liquid, malic acid was found in all stages of seed development as the principal carboxylic acid. This result was confirmed by GC/MS of a derivatized acidic fraction of endosperm liquid. The concentration of malic acid, as quantified enzymatically, increased during development from 6 to 15 mm . Design of the Culture Medium Composition and Labeling Experiments | Based on the above analysis, we developed a culture medium to mimic the composition of the endosperm liquid and to allow stable isotope labeling. In all experiments, the volume of liquid medium per embryo provided approximately 10-fold excess of carbon and nitrogen sources as related to the expected yield of oil and protein. Glc and Suc were provided at 40 and 80 mm, respectively, to mimic the composition during the most active oil and protein synthesis stage . General Procedure of Labeling Experiments | Embryos were grown on media under day length and low-light conditions simulating in planta growth. Media contained the following carbon sources: sugars (40 mm Glc and 80 mm Suc [S medium]); sugars and amino acids (SA medium); or sugars, amino acids, and malate (SMA medium). For labeling experiments, 13C-labeled Glc (99% 13C enrichment) was mixed with unlabeled Glc and Suc in the molar ratio of 20:20:80, which is a 10:20:160 molar ratio based on hexose units, and results in a 10% isotopic enrichment in hexose units. In an additional experiment, 13C-labeled Suc was supplied (see below). After 15 d of growth, the (fractional) 13C enrichment in fatty acids and amino acids (after protein hydrolysis) was analyzed by GC/MS as outlined by the scheme shown in Figure . Figure 1 | Overview of the analytical techniques used for the measurement of 13C labeling in amino acids and fatty acids (see also "Materials and Methods"). Overview of the analytical techniques used for the measurement of 13C labeling in amino acids and fatty acids (see also "Materials and Methods"). After extraction of labeled embryos, the seed protein was hydrolyzed and the amino acids derivatized to their N,O-t-butyl-dimethylsilyl (TBDMS) derivatives. By GC/MS, the amino acid molecule is represented by the fragment M-57. For most amino acids, additional fragments were measured that represent parts of the amino acid molecule. The abundances of mass isotope isomers (isotopomers) of a measured fragment (m0, m1, m2... mn) were corrected for isotopomer content in the derivatization reagent and for heteroatoms (1H, 13C, 15N, 17O, 18O, 29Si, and 30Si) as well as for natural 13C in the derivatized molecule fragment. Finally, the relative abundance of mass isotopomers (13C1, 13C2, 13C3... 13Cn) was obtained. After transmethylation of seed oil, fatty acid methyl esters were analyzed by GC/MS. The molecular ion and the McLafferty fragment (m/z 74) were measured and the relative abundance of mass isotopomers was obtained as described for the amino acids. Formation of Storage Products in Cultured Embryos Reflects Seed Development in Planta | As shown in Figure , the main accumulation of fatty acids of B. napus embryos in culture occurs during the first 2 weeks of culture (20 --35 DAF). In a labeling experiment, the amount of newly formed labeled fatty acids and protein were more than 10-fold greater than the initial biomass. After 45 d of growth, Suc and Glc were still abundant in the growth medium at similar concentrations as added initially, indicating constant nutrient supply. In some experiments, amino acids and/or malate were included into the media (SA, SM, and SMA media, see "Materials and Methods"). No substantial differences in overall growth of embryos were found between S, SA, SM, and SMA media. In general, the cultured embryos developed similar to in planta embryos. In average, 4 mg fresh weight (2 mg dry weight) and 20% fatty acids/fresh weight was obtained, which is similar to the data given for late cotyledon embryos (26 DAF) grown in planta . In embryos of the rapeseed cv Reston (a high-erucic acid line), the accumulation of seed oil is characterized by a sharp increase in C20:1 and C22:1 . The final fatty acid composition of the embryos grown in our medium is similar to literature values for embryos grown in siliques . The content of soluble protein increased continuously over more than 30 d , which is also the case for B. napus embryos grown in planta . In summary, after growth of B. napus cv Reston embryos on S medium, gain of fresh weight and protein as well as content and composition of fatty acids were similar to embryos developing in planta. We conclude that the development of embryos under these culture conditions mimics development in siliques of intact plants. The major increase in fatty acid content during the 15-d growth period defines the main phase of storage deposition in seed development, during which we assume a condition of metabolic steady state (see "Materials and Methods"). Figure 2 | Gain of lipid (micrograms fatty acid per embryo) and soluble protein (micrograms protein per embryo) of B. napus Gain of lipid (micrograms fatty acid per embryo) and soluble protein (micrograms protein per embryo) of B. napus cv Reston embryos cultured in liquid medium. The embryos were taken into culture about 20 DAF, which is mid-cotyledon stage, and were cultivated up to 30 d. The duration of labeling experiments was 15 d, as indicated. Each data point is the average from three embryos. Table II | Fatty acid composition in embryos of B. napus cv Reston Glc Is Metabolized Preferentially over Suc as Carbon Source for Seed Metabolism | When both Glc and Suc were provided with a 20% 13C enrichment (S medium, [U-13C6]Glc, [U-13C12]Suc), after 15 d growth, the 13C enrichment in both fatty acids and amino acids was found to be very close to 20% (data not shown). This result indicates that fatty acids and proteins have been fully labeled to the same isotope abundance as provided by the 13C carbohydrate in the media. This 13C abundance would not have been reached if preexisting or other carbon sources than Glc and Suc were available. This result also indicates that incorporation of atmospheric CO2 does not contribute substantially to fatty acid or protein synthesis despite the obvious ability for photosynthetic CO2 fixation of developing B. napus embryos (; see below). In preliminary experiments, when embryos were grown on S medium with uniform 13C-labeled Glc (99% 13C enrichment) and unlabeled Suc in a molar ratio of 20:80 (as related to hexose units), the resulting total 13C enrichment in fatty acids and amino acids was 30% to 35%, which is substantially higher than the expected 20%. Although both Glc and Suc clearly have been used as carbon sources in this labeling experiment, the 13C enrichment of approximately 35% in labeled end products indicates a preferential utilization of Glc over Suc. In addition, as described in "Materials and Methods," MS analysis of the 13C label pattern in different fatty acids indicated that a small subpopulation of fatty acid molecules was derived mainly from [U-13C6]Glc, whereas the bulk of fatty acid molecules derived from [U-13C6]Glc diluted as described above. This inhomogeneous distribution of labeled carbon in the metabolic products can pose problems for the interpretation of 13C-labeling patterns . As a result of these considerations, further experiments were performed with a carbohydrate mixture in which, in addition to unlabeled Suc, unlabeled Glc was added (10% [U-13C6]Glc, 10% unlabeled Glc, and 80% unlabeled Suc, as related to hexose units). Under these conditions, analysis of the labeled fatty acids (GC/MS, molecular ion clusters) showed that the inhomogeneity of 13C in the fatty acids was greatly reduced such that valid interpretations based on the isotopic steady-state assumption could be made (see "Materials and Methods"). Glycolytic Breakdown of Glc Dominates in Plastidic Acetyl-CoA Formation | The oxidative pentose phosphate pathway (OPPP) provides an alternative pathway to glycolysis for metabolism of hexose to acetyl-CoA and has been proposed as a source of reductant for fatty acid synthesis in oilseeds . Earlier studies attempted to assess OPPP activity by measurement of the reduction of label from [1-13C]Glc in metabolic products by Glc 6-phosphate dehydrogenase reaction relative to the label from [6-13C]Glc. However, refixation of the 13CO2 released and/or redistribution of label due to exchange reactions make such experiments difficult to interpret . For a first semiquantitative estimate of carbon flux through the OPPP, we used multiple 13C-labeled Glc ([1,2-13C2]Glc and [U-13C6]Glc). When cells metabolize [13C]Glc containing 13C-13C bonds and in the presence of unlabeled Glc, information on metabolic pathways can be gained from the conservation of 13C-13C bonds ("connectivity") in metabolites formed from the Glc . By sole glycolytic breakdown of Fru-6P via triose-P to pyruvate and finally oxidative decarboxylation to acetyl-CoA, the 13C-13C bonds of [U-13C6]Glc or of [1,2-13C2]Glc are 100% retained in pyruvate and in acetyl-CoA. In contrast, during the oxidative pentose phosphate cycle, synthesis of Fru-6P from pentose-P involves the enzyme transfer of two and three carbon units between different intermediate sugar phosphates, which results in the reduction of 13C-13C connectivity, i.e. a characteristic isotopomer pattern in Fru-6P and in its glycolytic products (triose-P, pyruvate, and acetyl-CoA). In addition, the reversibility of transketolase and transaldolase reactions leads to exchange reactions with glycolytic intermediates, producing similar isotopomer patterns. As shown in Figure a, glycolysis of [1,2-13C2]Glc will produce [2,3-13C2]pyruvate. The 13C2 isotopomers will also be retained by the reversible exchange of C1/C2 of Fru-6P by transketolase. However, if Glc enters the OPPP, C-1 of [1,2-13C2]Glc is lost by decarboxylation producing [1-13C]pentose-P. Recycling of Fru-6P from [113C]pentose-P by transketolase and transaldolase reactions will result in different single positional 13C-labeled Fru-6P isotopomers. Figure 3 | Predicted and observed isotopomer pattern in triose-P (TP) and derivatives (pyruvate and acetyl-CoA) after feeding of 13C-labeled Glc. Predicted and observed isotopomer pattern in triose-P (TP) and derivatives (pyruvate and acetyl-CoA) after feeding of 13C-labeled Glc. Developing embryos were labeled with [1,2-13C2]Glc (A) or [U-13C6]Glc (B) in SMA medium (10% enrichment of labeled Glc). Fragments of Ala and C18:1(1-2) were measured by GC/MS, representing pyruvate and acetyl-CoA, respectively. For the sake of clarity, the figure does not show all expected and found isotopomers. Fru-6P is derived from Glc or by reactions of the OPPP. After glycolytic cleavage of Fru-6P (a), dihydroxyacetone phosphate and glyceraldehyde 3-phosphate are assumed to be in isotopic equilibrium (triose-P [TP]). A, By sole glycolysis, the abundance of 13C2 label from [1,2-13C2]Glc is retained in triose-P, whereas the OPPP reduces the abundance of 13C2-labeled triose-P. Nearly 100% of possible 13C2 abundance was found for pyruvate and acetyl-CoA. B, By sole glycolysis, [U-13C6]Glc is transformed to [U-13C3]pyruvate (100%). By involvement of transketolase (TK) and transaldolase (TA) in synthesis of Fru-6P, the abundance of [U-13C3]pyruvate is reduced and mainly [1-13C]- and [2,3-13C2]pyruvate are formed. Because the contribution of the oxidative part of the OPPP is very low (A), the observed labeling pattern from [U-13C6]Glc can be explained by reversible exchange of C1/C2 of Fru-6P by transketolase. Due to highly reversible reactions of the OPPP, an exact determination of fluxes requires the measurement of fractional labeling in intermediates of the OPPP (see e.g. ). PDH, Pyruvate dehydrogenase complex; PP, pentose-phosphate. After labeling with [1,2-13C2]Glc, the distribution of 13C in pyruvate was measured for the fragments Ala(1-3), Ala(2-3), and Phe(1-2), representing pyruvate(1-3), pyruvate(2-3), and pyruvate(1-2), respectively. The fractional 13C label of these three pyruvate fragments allows the determination of the abundances of all eight possible 13C isotopomers in pyruvate, as described by . Our analysis of pyruvate isotopomers revealed that the relative abundance of [2,3-13C2]pyruvate was 89% of the abundance defined by sole glycolysis (Fig. A). Similarly, the abundance of 13C2 isotopomer in acetate units of C18:1 [fragment C18:1(1-2)] was 93% of the maximal possible abundance (Fig. A). Together, these fatty acid and amino acid analyses indicate that approximately 90% of the [1,2-13C2]Glc molecules were transformed to pyruvate without being subjected to oxidative decarboxylation. Although Ala possibly does not represent the plastidic pyruvate pool, Val is derived from plastidic pyruvate . The isotopomer pattern of Val(2-5) was found to be in accordance with the 13C2 abundance found in Ala (data not shown), confirming that approximately 90% of the [1,2-13C2]Glc molecules were transformed to plastidic pyruvate without being subjected to oxidative decarboxylation. In addition, the mass spectrum of His, labeled from [1,2-13C2]Glc, revealed that pentose-P is mainly formed by transketolase, rather than by the oxidative decarboxylation of [1,2-13C2]Glc-6P (data not shown). Additional conclusions on the extent of reversibility of transketolase were derived after labeling with [U-13C6]Glc. This analysis revealed that the transketolase reaction largely reduced the abundance of the [U-13C3]pyruvate isotopomer to 55% of its abundance defined by sole glycolysis (Fig. B). By carbon transfer reactions, mainly [1-13C]pyruvate and [2,3-13C2]pyruvate were produced, which is interpreted as a signature of the transketolase reaction on [U-13C6]Fru-6P (Fig. B). In summary, labeling with both [1,2-13C2]Glc and with [U-13C6]Glc with subsequent analysis of isotopomer abundance in pyruvate and C18:1(1-2) demonstrated that during Glc breakdown, about 90% of 13C-13C connectivity from [1,2-13C2]Glc is retained in pyruvate and acetyl-CoA, indicating low OPPP activity relative to the total glycolytic flux. Furthermore, labeling with [U-13C6]Glc showed that the precursors of pyruvate and acetyl-CoA (Fru-6P and Glc-6P) are subjected to the reactions of the nonoxidative part of the pentose phosphate pathway. Because only the isotopomer pattern of the oxidative part of the OPPP is missing, we consider unlikely the possibility of a "sequestered" OPPP in which [1-13C]pentose-P could be largely produced by oxidative decarboxylation of [1,2-13C2]Glc-6P, but the recycled single 13C-labeled Fru-6P would not enter the glycolytic route leading to fatty acids ---and, therefore, would not contribute to the measured isotopomer pattern . This kind of "sequestered" cyclic OPPP is also not supported by the labeling pattern of His, which is derived from plastidic pentose-P. We conclude that the OPPP has a low contribution to Glc breakdown, and is unlikely to provide the major source of NADPH for fatty acid synthesis in B. napus embryos cultured as described. However, more exact measurements of relative fluxes through glycolysis and OPPP will require labeling experiments specifically designed for the sensitive measurement of small fluxes through OPPP. Also, modeling of the central carbon metabolism, including compartmentation effects and the effect of highly reversible reactions of the nonoxidative part of the OPPP, are necessary to further confirm our findings. Amino Acids Do Not Provide Carbon for Plastidic Fatty Acid Synthesis | From the data given in Table , it can be calculated that amino acids and malate constitute up to 25% of the carbon in endosperm liquid. To assess the contribution of these compounds to protein and fatty acid synthesis, embryo labeling experiments with the basic growth medium containing [U-13C6]Glc were compared with experiments where unlabeled carbon sources (amino acids and/or malate) were also added. In such experiments, a reduction in 13C label in the end product indicates incorporation of carbon from the unlabeled carbon source. As shown in Figure A, the 13C enrichment of fatty acids is not significantly reduced by addition of either unlabeled amino acids or malate as carbon sources. In contrast, as shown in Figure B, 13C levels of several amino acids of proteins were strongly reduced by added unlabeled amino acids. Thus, amino acids provided to the embryos are readily incorporated into proteins but do not serve as a carbon source for de novo fatty acid synthesis. Figure 4 | Incorporation of [U-13C6]Glc into seed oil and storage protein of B. napus Incorporation of [U-13C6]Glc into seed oil and storage protein of B. napus embryos and dilution of label by additional unlabeled carbon sources. Amino acids (Glu, Asp, and Ser) and/or malic acid were added to the growth medium. Isotopic enrichment was measured by GC/MS of fatty acid methyl esters and TBDMS amino acids. A, The fatty acids from seed oil show no isotopic dilution, except for the terminal acetate units of C20:1 and C22:1. B, Several amino acids from seed protein show high isotopic dilution by unlabeled amino acids in the medium. The percentage of carbon contributed from each amino acid (or fragment) to seed protein is given. Due to acidic conditions during hydrolysis of seed protein, Asn and Gln are found as Asp and Glu, respectively. Not measured were Lys, His, Arg, Cys, and Trp. For the calculation of the isotopic dilution in total seed protein, the 13C enrichment of these amino acids was derived from other amino acids according to the closest biosynthetic relations. The seed protein was assumed to consist of 60% (w/w) cruciferin, 20% (w/w) napin, and 20% (w/w) oleosin . For these proteins, the amino acid compositions were taken from and . For five independent experiments, the average 13C label +- se is given. Carbon from Amino Acids Contributes to the Cytosolic Acetyl-CoA Pool | B. napus cv Reston produces 38% C20 and C22 fatty acids in addition to 62% C16 and C18 fatty acids . Whereas de novo fatty acid synthesis of C16 and C18 fatty acids occurs in the plastid from acetyl-CoA, elongation of oleic acid outside the plastids is responsible for the addition of the terminal two or four carbons of C20:1 and C22:1, respectively . Therefore, these very long-chain fatty acids offer an opportunity to compare the labeling of plastid versus cytosolic pools of acetyl-CoA. Label in the terminal acetate unit of fatty acids was examined from the abundant McLafferty fragment (m/z 74, see "Materials and Methods"). For oleic acid (Fig. A) and palmitic acid (data not shown), no dilution of label in the terminal acetate carbon was observed upon addition of unlabeled amino acids to the growth medium (Fig. A). In contrast, 13C labeling of the terminal acetate units of both 20:1 and 22:1 were significantly reduced by added amino acids. The observation that the label of the cytosolic acetyl-CoA pool is significantly diluted by unlabeled amino acids, whereas plastidic acetyl-CoA is not, suggests involvement of different pathways to produce the two pools. Thus, not only is the plastid pool of acetyl-CoA clearly spatially distinct from the cytosolic pool (as previously observed in numerous labeling studies), but these results further demonstrate a different metabolic origin of acetyl-CoA. Source of Cytosolic Acetyl-CoA | Acetyl-CoA apparently does not cross membranes. Based on a lack of cytosolic pyruvate dehydrogenase, direct synthesis of acetyl-CoA from pyruvate in the cytosolic compartment and the formation of cytosolic acetyl-CoA in plants is still not clearly understood . The results in Figure A indicate a reduction of 13C label in cytosolic acetyl-CoA by carbon from unlabeled Asp and Gln. This result also occurred with experiments using only Gln as unlabeled amino acid (data not shown). This dilution of 13C by Asp and Gln points to a possible relation of cytosolic acetyl-CoA to mitochondrial metabolism because both amino acids can be metabolized to tricarboxylic acid (TCA) cycle intermediates. From known metabolic pathways in plants, and from the results in Figure A, the following possible route can be proposed: (a) Unlabeled Glu and Asp are transformed into TCA cycle intermediates 2-ketoglutarate, oxaloacetate, succinate, fumarate, and malate. (b) Mitochondrial pyruvate, imported from the cytosol and derived from sugars, would have maximal 13C enrichment as found in plastid-derived fatty acids and amino acids (Val and Leu). A reduction of 13C enrichment is most likely due to the activity of mitochondrial malic enzyme because malate is in part derived from unlabeled Asp and Glu. (c) By pyruvate dehydrogenase complex, mitochondrial acetyl-CoA is formed and transformed to citrate by citrate synthase. After export of citrate to the cytosol, acetyl-CoA is formed by cytosolic ATP:citrate lyase (EC ; ; ). In particular, due to the Pro-3S specific cleavage of citrate by ATP:citrate lyase, only carbons of mitochondrial pyruvate are the source of cytosolic acetyl-CoA and no carbons from mitochondrial oxaloacetate end up in acetyl-CoA. In summary, given the influx of unlabeled carbon into the TCA cycle via oxaloacetate and 2-oxoglutarate and the export of mitochondrial acetyl-CoA into the cytosol via citrate, the isotopic dilution in mitochondrial pyruvate and in cytosolic acetyl-CoA can best be explained by the action of mitochondrial malic enzyme. As an alternative to the export of acetyl-CoA via citrate, free acetate export from mitochondria, after acetyl-CoA hydrolysis , cannot be ruled out because direct acetate export would lead to the same conclusions on isotope dilution as made with the export of citrate. However, some in vitro biochemical data suggest a lack of cytosolic acetyl-CoA synthase . Is Malate an Intermediate in B. napus Embryo Fatty Acid Biosynthesis? | Unlabeled malic acid did not significantly reduce the incorporation of [U-13C6]Glc into fatty acids or amino acids (Fig. B). Even Asp, which can be formed from malate via malate dehydrogenase and Asp aminotransferase, was not significantly reduced in 13C label (Fig. B, SM medium). This indicates that externally supplied malic acid is not used as a major carbon source during embryo development, although it was found in considerable concentration in the endosperm liquid . This, however, does not exclude the existence of malate metabolism into fatty acids inside the embryo from internally generated pools and in this regard, malate was proposed to be a precursor for plastidic acetyl-CoA in developing rapeseed embryos . Based on in vitro enzyme activity analysis, proposed that cytosolic phosphoenolpyruvate (PEP) is transformed to oxaloacetate (by PEP carboxylase) and reduced to malate (NAD-malate dehydrogenase), which is imported into the plastid and transformed to pyruvate by malic enzyme. We consider this scenario unlikely because the incorporation of [U-13C6]Glc into plastidic fatty acids was not reduced by unlabeled Asp, although carbon from this amino acid was readily incorporated into seed protein and cytosolic acetyl-CoA, most likely via malate as described above. Amino Acid Incorporation into Seed Storage Proteins | On a dry weight basis, mature B. napus embryos contain about 50% oil and 30% protein . Based on the concentrations of amino acids in the endosperm liquid and on studies on nitrogen metabolism in rapeseed plants , we reasoned that the embryos could meet their high demand for amino acids at least in part by uptake from the endosperm liquid rather than by de novo synthesis from sugars and nitrate. As a consequence, in vivo there may be a reduced demand for precursors of amino acids (e.g. oxaloacetate and alpha-ketoglutarate) and cofactors for amino acid biosynthesis as compared with growth where no amino acids are supplied. When embryos were grown on [U-13C6]Glc with unlabeled Gln, Asn, and Ser, 13C enrichment in several seed protein amino acids was highly reduced as compared with the control experiment without amino acids (Fig. B). This reduction in 13C label in particular amino acids indicates incorporation of the unlabeled amino acid or its derivative. Asn and Gln are readily converted to Asp and Glu and can be assumed to deliver amide-bound nitrogen for the synthesis of other amino acids. The incorporation of Gln, Asn, and Ser into different amino acids of the seed protein reflects expected biogenetic relations as follows: Addition of amino acids to the medium reduced 13C label in Ser of seed protein 90% to 95% (Fig. B), indicating that it was nearly completely incorporated from the medium rather than de novo synthesized from carbohydrates. Gly was similarly reduced in 13C label (Fig. B), which can be explained by a synthesis from Ser by Ser transhydroxymethylase . An additional related conclusion can be made. If photorespiration was highly active in developing B. napus embryos, Ser and Gly would be a part of the carbon flux for recycling of phosphoglycerate from phosphoglycolate, produced by the Rubisco oxygenase reaction. Unlabeled Ser would be expected to reduce 13C label in plastidic phosphoglycerate and in derived intermediates and end products as well as the 13C label of plastidic glycolate should end up in Ser. Because the plastid-derived fatty acids showed no dilution of 13C label and Ser was very low in 13C label, these results indicate that a major flux through photorespiration pathways is unlikely. Glu and Pro were also strongly reduced in 13C content (Fig. B) as expected because both Glu and Pro are derived from Gln. Asp, Thr, Ile, and Met were also reduced in 13C label, whereas Thr, Ile, and Met were less reduced in 13C label than Asp (Fig. B). Thr and most of the carbon skeleton of Ile and Met are derived from Asp (via Asp semialdehyde as common precursor). Their biosynthesis is localized in the plastid and the difference in 13C enrichment between Asp and Thr, Ile, and Met (Fig. B) most likely reflects the compartmentation of different Asp pools. Amino acids related to plastidic pyruvate (Val and Leu) or to plastidic PEP [Phe(1-2)] were not reduced in label by the supply of Ser, Gln, and Asn (Fig. B). This is in further agreement with the finding that Ser, Gln, and Asn are not incorporated into the plastid-derived fatty acids. How Much Carbon in Seed Protein Is Derived from Preformed Amino Acids? | To assess the amount of carbon that is contributed from the exogenously supplied unlabeled amino acids to seed protein formation, we calculated the reduction of 13C label (from [U-13C6]Glc) in protein relative to experiments without added amino acids. By considering the percentage of carbon each amino acid contributes to seed protein (see legend of Fig. ) and the isotopic dilution from the unlabeled amino acids, we calculated that the seed proteins reached approximately 70% of the maximum possible label. Thus, 30% of the carbon of the seed protein is derived from the unlabeled amino acids in the growth medium and 70% from de novo biosynthesis from sugars. This substantial contribution of exogenous amino acids to total carbon in seed protein is in contrast to the seed oil, which did not experience dilution in 13C abundance by carbon from amino acids. This statement is valid, although it was found that the terminal acetate units of C20 and C22 fatty acids are diminished in 13C by unlabeled amino acids (Fig. A). Because these terminal acetate units only contribute about 8% of total carbon in seed oil, the contribution of exogenous amino acids to oil synthesis is minimal. The fact that exogenous amino acids are used as a nitrogen and carbon source for protein synthesis, but not as a carbon source for the plastidic fatty acid synthesis, clearly emphasizes that although many intermediates of metabolism can be considered on paper to connect amino acid and fatty acid metabolism, in B. napus embryos, compartmentation or other barriers prevent these connections. This result also emphasizes the independence of protein and oil biosynthesis and may in part explain why decreasing the amount of protein or oil in seeds of Arabidopsis mutants does not generally lead to a balancing increase in the other major storage component. For example, in Arabidopsis abi/aba mutants , storage protein content is strongly reduced but oil content does not change significantly, and in the wri1 mutant , oil is reduced 80% but storage protein content remains similar to wild-type seeds. SUMMARY AND CONCLUSIONS : To better understand the flow of carbon in developing embryos of B. napus during storage product accumulation, we initiated 13C-labeling experiments with embryos growing in culture. Because developing embryos take up nutrients from a liquid environment, there is the opportunity for steady-state stable isotope labeling experiments to closely mimic in planta seed metabolism and the transport of nutrients from the mother plant to the embryo. By proffering nutrients in similar concentrations to those found in the endosperm liquid, labeling experiments on cultured embryos can reveal, therefore, fluxes through central carbon metabolism very close to the in planta conditions and much more reliable conclusions can be made than from studies using extracts, organelles, or pulses of radioactive precursors. The initial studies using this approach have led to the following conclusions: In addition to Glc, Fru, and Suc, endosperm liquid of developing seeds of B. napus contains considerable amounts of amino acids (mainly Gln) and malate as organic constituents . Therefore, a labeling approach with multiple carbon sources is a precondition to quantitatively reflect the fluxes of central carbon metabolism that occur in planta in developing embryos. We found that in addition to hexoses and Suc, amino acids are used to deliver both carbon and nitrogen to the growing embryo. The amino acids of the growth medium are incorporated primarily into storage proteins and did not provide carbon for plastid fatty acid synthesis. Amino acids of the growth medium provide an additional carbon source for cytosolic acetyl-CoA used for fatty acid elongation. The involvement of mitochondrial malic enzyme in the formation of cytosolic acetyl-CoA was postulated as an explanation for incorporation of carbon from Asn and Gln into the terminal carbons of C20:1 and C22:1 fatty acids. In contrast to cytosolic acetyl-CoA, the plastidic fatty acid biosynthesis pathway is clearly independent and fed only by the sugars that are metabolized primarily by the glycolytic pathway and without substantial contribution of the OPPP. This finding underlines the independence of cytosolic amino acid and protein biosynthesis from plastid fatty acid synthesis in B. napus despite the potential exchange of common intermediates between the two pathways. Although developing B. napus embryos are green and have substantial Rubisco and photosynthetic capacity , 13C incorporation from hexose into fatty acids and amino acids showed no evidence of dilution from the fixation of atmospheric CO2. Also, as discussed above, photorespiration is not a substantial part of embryo metabolism. In addition to the above conclusions, our considerations of carbon economy in developing oil seeds raise some new questions. During the formation of fatty acids, it is notable that one-third of the carbon of precursors is released as CO2 when pyruvate is transformed to acetyl-CoA by the pyruvate dehydrogenase complex. Thus, without refixation, a substantial fraction of the carbon entering oilseeds as carbohydrate would be lost. Based on maximal rates of oil synthesis in developing rapeseed embryos , CO2 produced would saturate the embryo cells within 10 min. Because the seed coat surrounding the developing embryo is a major barrier for gas diffusion , the produced CO2 cannot simply escape as fast as it is produced. As a consequence, it may not only be efficient but also essential for the developing oilseed embryo to conduct refixation or export of CO2. B. napus embryos have green chloroplasts with the capability of photosynthetic carbon fixation . Measured rates of oxygen evolution of developing B. napus embryos as well as Rubisco activity are at least theoretically sufficient to refix the CO2 that is produced by maximal oil synthesis. In addition to Rubisco, high activities of PEP carboxylase and malic enzyme are found in Brassica campestris in the developing seeds . The latter enzyme activities are essential to the CO2 concentration mechanism in C4 photosynthesis. By analogy, we speculate that PEP carboxylase in the embryo could refix a part of the CO2 into oxaloacetate and malate, which could be exported to the seed coat or the pod wall, where more light is available for photosynthesis. Thus, the CO2 evolved in oil-accumulating embryos could be refixed by PEP carboxylase into oxaloacetate and malate, which could be exported into seed coat or pod wall. In fact, we have found high malate levels in the liquid endosperm. The use of a complex medium for stable isotope labeling (labeled Glc and unlabeled amino acids) provided additional insights, such as the involvement of amino acids in biosynthesis of cytosolic acetyl-CoA. However, for the formation of acetyl-CoA in plants, a multitude of possible combinations of existing reactions might be proposed. The actual in vivo fluxes may be best found by more detailed "metabolic flux analysis" using in vivo stable isotope labeling approaches. These approaches will involve introduction of 13C-labeled Glc, amino acids, and other precursors labeled in specific positions, followed by detailed isotopomer analysis. Such experiments offer the possibility to further dissect several aspects of central carbon metabolism, including the questions outlined above and the contributions of alternative sources for reducing equivalents [NAD(P) H] for fatty acid biosynthesis in developing seeds. MATERIALS AND METHODS : Chemicals | d-[U-13C6]Glc, d-[1,2-13C2]Glc, and [U-13C12]Suc (99% 13C-enrichment) were purchased from Isotec (Miamisburg, OH). Analysis of the Endosperm Liquid | Collection of Endosperm Liquid | Endosperm liquid was collected from seeds, directly after harvest of siliques, with a thin pipette (1 --3 muL per seed). After centrifugation (10,000g for 1 min), the supernatant was heated to 100C for 1 min and centrifuged again at 10,000g for 5 min. This was done to ensure the absence of any enzymic activities that could interfere with metabolite assays. The resulting supernatant was assayed as described below. Sugars in the Endosperm Liquid | A semiquantitative determination of sugars was performed by TLC of 0.5 to 1 muL of endosperm liquid on cellulose (cellulose MN 300, Machery-Nagel, Dueren, Germany) with t-butanol:ethyl methyl ketone:formic acid:water (40:30:15:15 [v/v]). The sugars and reference standards were visualized by staining with a solution containing 0.2 g of aniline, 0.2 g of diphenylamine, and 1 mL of H3PO4 in 10 mL of ethanol and heating to 80C. Using 0.2 muL of endosperm liquid, the Glc concentration was determined enzymatically with a hexokinase/Glc-6-phosphate dehydrogenase test (Sigma, St. Louis) as described by . With prior addition of invertase to an endosperm sample, the concentration of the sum of Glc and Suc could be determined. Fractionation of Endosperm Liquid into Neutral/Acidic and Basic Fractions | HCl was added to endosperm liquid to a final concentration of 0.1 m. The endosperm liquid was then applied on a short column of DOWEX 50 W kation exchange resin (H+; Sigma, St. Louis). Neutral and acidic substances were eluted with water and amino acids were eluted with 2 n NH4OH. Malate in Endosperm Liquid | One to 5 muL of endosperm liquid was separated by TLC on cellulose (cellulose MN 300, Machery-Nagel) with 1-butanol:acetic acid:water (80:20:20 [v/v]). Acidic spots were visualized with bromphenol blue. An acidic fraction of endosperm liquid was derivatized with ethyl chloroformate as O-ethoxycarbonyl ethyl esters and separated by GC/MS (see below). The content of malic acid in endosperm liquid was measured with malic enzyme (EC , Sigma). To 1 mL of reaction buffer (25 mm HEPES/NaOH, pH 7.5; 10 mm MgCl2; and 2 mm NADP), 2 muL of endosperm liquid and 0.5 units of malic enzyme (Sigma) were added. The increase in A340 was measured for several minutes at room temperature in a spectrophotometer (DU640, Beckman Instruments, Fullerton, CA). One millimolar NADPH formed equals 1 mm malic acid and an increase in A340 of 6.22. Amino Acids in Endosperm Liquid | Endosperm liquid from different developmental stages was separated by TLC on silica gel with n-butanol:acetic acid:water (80:20:20 [v/v]) and with CHCl3:methanol:NH4 (40:40:20 [v/v]). Amino acids were visualized with ninhydrin (0.2% [w/v] in ethanol) and heating to 100C. For quantification of amino acids, a solution of uniformly 13C-labeled amino acids was made by hydrolysis of uniformly 13C-labeled algal crude protein extract (Isotec) in 6 n HCl for 24 h at 110C. After removing HCl, the concentration of labeled amino acids in this standard solution was determined by adding known amounts of unlabeled amino acids, derivatization to the TBDMS derivatives (see below), and analysis of the mass isotopomers of selected fragments by GC/MS (selective ion monitoring). To a small volume (10 --30 muL) of endosperm liquid, an aliquot of the uniformly 13C-labeled amino acid standard was added and the amino acids were purified on an anion-exchange column (see above). After derivatization to the TBDMS derivatives (see below) and analysis of the mass isotopomers of selected fragments by GC/MS (selective ion monitoring), the concentrations of 15 amino acids were determined. Gln, Asn, Trp, Cys, and Arg were not recovered in the uniformly labeled protein hydrolysate and therefore could not be determined. Using 5 --10 muL of endosperm liquid, Glu and Gln were determined enzymatically using glutaminase and l-Gln dehydrogenase as described by . Assuming the same efficiency of derivatization for Gln and Asn, the concentration of Asn was determined by comparison of the respective peak intensities in the total ion chromatogram. For the analysis by GC/MS, amino acid fractions were derivatized to their TBDMS derivatives . Growth Medium for Embryos | Inorganic constituents of all growth media were based on the medium used by for growth of embryos of Capsella bursa-pastoris. A total osmotic pressure of --14 atm (see below) was confirmed to be optimal for growth of Brassica napus embryos. To define the osmotic pressure of the medium, the partial osmotic pressure of PEG in solutions was calculated according to . For the calculation of the partial osmotic pressure caused by Suc and Glc, values were taken from . For all other media components, the osmotic pressure was calculated from the molar concentration (c) by the Van't Hoff equation: P = cRT (R = 8.14 J mol-1 K-1, T = absolute temperature in K). The total osmotic pressure was assumed to be the sum of all partial osmotic pressures. The concentration of K+ is important for embryo growth . If KNO3 was omitted (SA and SMA media), the total concentration of K+ of the S medium was maintained by addition of KCl. The amount of KOH added for titration of pH was also considered. Constituents of the growth medium (pH 5.7) were: KNO3 (19 mm), NH4NO3 (10 mm), CaCl2 (5.99 mm), MgSO4 (1.5 mm), KCl (4.69 mm), KH2PO4(1.25 mm), PEG 4000 (220 g L-1), Na2EDTA (14.9 mg L-1), FeSO4 7H2O (11.1 mg L-1), H3BO3 (12.4 mg L-1), MnSO4 H2O (33.6 mg L-1), ZnSO4 7H2O (21 mg L-1), KI (1.66 mg L-1), Na2MoO4 2H2O (0.5 mg L-1), CuSO4 5H2O (0.05 mg L-1), CoCl2 6H2O (0.05 mg L-1), inositol (100 mg L-1), nicotinic acid (5 mg L-1), pyridoxine HCl (0.5 mg L-1), thiamine HCl (0.5 mg L-1), folic acid (0.5 mg L-1), and biotin (0.05 mg L-1). S Medium | The medium contained Suc (60 mm) and d-Glc (40 mm) as carbon sources. SM Medium | Ten millimolar malic acid was added into S medium. SA Medium | The medium was derived from the S medium: KNO3 and NH4NO4 were omitted. Gln (10 mm), Asn 5 (mm), and Ser 5 (mm) were included. SMA Medium | KNO3 and NH4NO4 were omitted. Ten millimolar malic acid was added and Gln (10 mm), Asn 5 (mm), and Ser 5 (mm) were included. Influence of Osmotic Pressure on Growth | In addition to the nutrient composition, the total osmotic pressure of the growth medium is an important factor influencing embryo development . High osmotic pressure must be maintained to prevent precocious germination . defined a minimum osmotic pressure of -14 atm, which will inhibit precocious germination in 100% of B. napus embryos and enabled continuous growth in embryonic mode. To maintain such a high osmotic pressure in liquid growth media, polyols (without nutritional effect) such as sorbitol or mannitol have been used in liquid culture of B. napus embryos. However, observed an overall better development of microspore-induced embryos if sorbitol was replaced by PEG. To establish optimum water potential for culture, the effect of PEG as osmoticum on the gain of fresh weight and oil content of zygotic B. napus embryos was tested with constant concentrations of Suc and Glc of 80 and 40 mm, respectively. PEG 4000 was included between 14% and 26% (w/v; Fig. ). Growth in culture was initiated with embryos in the early cotyledon stage (20 +- 1 DAF, average 0.3 mg fresh weight). Precocious germination was not observed in any of the tested PEG concentrations. As shown in Figure , with increasing PEG concentration, the final fresh weight decreased over 10-fold from an average 17-mg embryo-1 (14% [w/v] PEG) to 1.5-mg embryo-1 (26% [w/v] PEG). In addition, the relative oil content increased , confirming that the osmotic pressure is a major factor that influences growth and development of the embryo. With 22% (w/v) PEG, an osmotic pressure of -14 atm is maintained in the liquid medium and the embryos grew to about 4 mg fresh weight , which compares well with the 3.7 mg fresh weight for late cotyledon stage (26 DAF) embryos developed in planta . Therefore, for this study the total osmotic pressure was adjusted to -14 atm by adding PEG 4000 to the medium. Figure 5 | Dependence of growth and oil content on osmotic pressure of the culture medium. Dependence of growth and oil content on osmotic pressure of the culture medium. Early cotyledon embryos were cultured for 15 d in liquid medium with different concentration of polyethylene glycol (PEG) 4000. Growth of B. napus Plants | Seeds of B. napus cv Reston were planted in 30-cm plastic pots in a 2:1 (v/v) mixture of peat moss:vermiculite (Therm-O-Rock Inc., New Eagle, PA) and grown under constant conditions (15C nights, 20C days; 16-h day, 600 muE m-2 s-1). Growth of Isolated Embryos | Embryos 20 DAF were defined by their fresh weight of 0.3 to 0.5 mg. Siliques with developing seeds were treated for 5 min with diluted commercial bleach (about 1% [w/v] NaOCl as active ingredient) and rinsed several times in sterile water. Seeds were removed from the siliques and embryos were dissected aseptically and placed immediately into growth medium. Each embryo was grown under aseptic conditions in an Erlenmeyer vessel (250 mL) in 6 mL of filter-sterilized growth medium, sealed with a cotton plug. Under these conditions, the embryo was only 1 to 2 mm under the liquid surface to ensure gas exchange. The vessels were not shaken to avoid mechanical stresses. The vessels were kept in a growth chamber under a temperature of 20C and fluorescent light (50 --100 muE m-2 s-1) in a 16-h-light, 8-h-dark cycle. The silique wall absorbs similar80% of incident light . Therefore, this light level in cultures approximates average light intensities reaching embryos in planta. Embryos were harvested after 15 d of culture for analysis as described below. Analytical Methods | Extraction of Proteins and Lipids | Each embryo was homogenized in a tissue grinder in 2 mL of an ice-cold buffer containing Na-phosphate, pH 7.5 (10 mm), and NaCl (500 mm). Fifty micrograms of triheptadecanoin was added as internal standard for triacylglycerols. The lipids were extracted three times with 1 mL of hexane. After centrifugation (10,000g for 20 min), the hexane phase was evaporated to dryness in a stream of nitrogen. Measurement of protein concentration in the aqueous phase was according to , with bovine serum albumin as the standard. Fatty acid methyl esters were prepared by transmethylation of the extracted lipids in a mixture of 0.7 mL of toluene and 1.3 mL of 10% (w/v) BCl3 in methanol for 1 h at 95C. After subsequent addition of 0.5 mL of water and extraction three times into 1 mL of hexane, fatty acid methyl esters were measured by GC/MS. Protein Hydrolysis and Amino Acid Derivatives | Soluble proteins were precipitated by addition of 10% volume of 50% (w/v) trichloroacetic acid. After 1 h of incubation on ice and centrifugation (10,000g for 10 min), the protein pellet was washed twice with diethylether:ethanol (1:1 [v/v]). Proteins were dissolved in a small amount of water and then hydrolyzed in 6 n HCl for 24 h at 100C. HCl was removed under reduced pressure, amino acids were dried under vacuum, and then derivatized to their TBDMS derivatives . GC/MS Analysis of Amino Acid Derivatives and Fatty Acid Derivatives | All GC/MS analyses were performed using an HP 5890 II gas chromatograph with an HP 5972 mass spectrometer (Hewlett-Packard, Palo Alto, CA). Carrier gas was helium at 1 mL min-1. For fatty acid methyl esters and N,O(S)-ethoxycarbonyl ethyl esters of carboxylic acids, a 30-m x 0.25-mm DB23 column was used (J&W Scientific, Folsom, CA). For TBDMS derivatives of amino acids, a 30-m x 0.25-mm DB1 column was used (J&W Scientific). The GC conditions were as follows for fatty acid methyl esters: The injector was set to 250C, and the detector was set to 300C. The oven temperature programming was: 90C for 3 min, increased to 160C at 30C min-1, increased to 190C at 10C min-1, and 190C for 7 min. Finally, temperature was raised to 240C at 20 min. For N,O(S)-ethoxycarbonyl ethyl esters of carboxylic acids, the injector was set to 250C, and the detector was set to 250C. The oven temperature programming was: The initial temperature was 110C for 2 min, and was increased to 255C at 10C min-1. The final temperature was 255C for 20 min. For TBDMS derivatives of amino acids, the injector was set to 300C, and the detector was set to 300C. The oven temperature programming was: 150C for 14 min, increased to 210C at 5C min-1, and then increased to 295C at 10C min-1. The final temperature was 295C for 10 min. Measurement of Isotopomers in Mass Spectra | Mass isotopomers are molar fractions of m0, m1, m2, etc., according to the number of labeled carbons in the molecule. The sum of all mass isotopomers of the molecules (Sigmami) is 100%. Single ion monitoring was generally used for the measurement of mass isotopomers. The mass spectra of the entire chromatographic peak were integrated to avoid the influence of possible isotope fractionation during GC separation. Background correction was performed with MS spectra directly preceding the chromatographic peak. The isotope ratios were reproducible over the concentration range of at least 2 orders of magnitude. A matrix-based method was applied for correction of natural isotope content in heteroatoms and in derivative residues, as well as in the labeled molecule, as described by . The identity of different fragments of the TBDMS amino acid derivatives was derived from literature . Derivatives of amino acids with natural isotope enrichment were separated by GC/MS and the mass isotopomer distribution of different fragments in the MS spectra was measured and monitored for their deviations from the expected mass isotopomer pattern, if calculated from the element composition of the fragment. Fragments that were in good agreement with the calculated mass isotopomer distribution (<1% deviation of individual peaks) were used for measurements with labeled amino acids. Mass Spectra of Fatty Acids | The molecular ion of fatty acids was measured to determine the average 13C content of fatty acids. A matrix-based method was applied for correction of natural isotope content as described by . The 13C enrichment is the weighted sum of labeled mass isotopomer species (Sigma mi i/n; mi = fractional molar abundance of the mass isotopomer containing i 13C-atoms, n = total number of carbons). McLafferty Ion (m/z 74) in Fatty Acids | The molecular ion of fatty acids represents a polymer of acetate units and information on mass isotopomers in acetate units cannot exactly be derived from this ion. However, in mass spectra of fatty acid methyl ester, the ion m/z 74 is a highly abundant ion, and in saturated fatty acid methyl esters m/z 74 is the base peak. This peak represents a rearrangement reaction of the fatty acid methyl ester (McLafferty rearrangement) involving the transfer of gamma-H and the break of the C-2/C-3 bond yields an ion of the elemental composition C3H6O2, comprising C-1, C-2, and the O-methyl group of the fatty acid methyl ester . The fragment comprises C-1 and C-2 of a fatty acid, i.e. the terminal acetate unit . The identity of the fragment m/z 74 was confirmed using specific deuterium-labeled isomers . Mass 75 has a much higher abundance than expected from natural isotopes of C3H6O2 (m/z 74), due to a second proton transfer to the same fragment resulting in C3H7O2m/z 75 . For saturated fatty acid methyl esters of different chain lengths, this fragment was used to measure incorporation of stable isotopes into the terminal acetate unit . The mass isotopomers m0, m1, and m2 (relative abundance of 12C2, 13C1, and 13C2) of the terminal biosynthetic acetate unit can be derived from the ions C3H6O2+ and C3H7O2+ by measuring the intensity of the masses 74, 75, and 76 of the labeled sample. Correction for natural isotope content as well as for the ion C3H7O2+ (m/z 75) was made using the relative intensities of masses 74 to 76 from an unlabeled reference. To further confirm the suitability of m/z 74 for measurement of stable isotope label, possible superposition by other ions was investigated for different saturated and mono-unsaturated fatty acid methyl esters. Methylation with CD3OD showed that the total intensity of the peaks m/z 74 and 75 is shifted by three mass units, i.e. contains the O-methyl group of the fatty acid methyl esters. By high-resolution MS of methyl oleate, we confirmed again the identity of the following fragments: C3H6O2 (m/z 74) and C3H7O2 (m/z 75). In unlabeled fatty acid methylesters, the ion m/z 76 has a very low abundance (<1% of m/z 74) and it was determined by high-resolution MS to be C4H6. The error contributed by this fragment will be lower than the accuracy of the MS measurements. In methyl oleate, the fragment m/z 73 has 20% intensity of m/z 74 and its elemental composition was determined as C4H9O. Mass isotopomers of this ion will contribute to the masses 74 and 75. Methylation of oleic acid with CD3OD showed that m/z 73 includes the O-methyl group. By assuming that m/z 73 represents the carbon fragment "C3-C2-C1-O-methyl," this contribution can be corrected for, if the labeling pattern of acetate units is first approached from the molecular ion of the spectrum and the isotopomer composition of C4H9O (m/z 73) is calculated. In addition to m/z 73, the possible superposition of other relatively abundant ions with m/z 74 was tested. In methyl oleate, m/z 69 and 70 are nearly as abundant as m/z 74. If the fatty acid is highly 13C labeled, isotopomers of m/z 69 and 70 may shift to mass 74. By high-resolution MS of methyl oleate, m/z 69, 70, and 71 were identified as C5H9, C5H11, and C5H11, respectively. By knowing the composition of the fragments, the possible superposition with m/z 74 by 13C label can be calculated. With an average 13C label lower than 10%, the superposition of fragments m/z 69, 70, and 71 will not be significant. In summary, the abundances of m/z 74, 75, and 76 in mass spectra of labeled fatty acids can be used to accurately determine the 13C labeling of the terminal acetate unit. The masses 74 to 76 must be measured with an unlabeled reference to correct for natural isotopes in C3H6O2+ and for the ion C3H7O2+. Ions that are not derived from the terminal acetate unit and that might interfere with the McLafferty ion m/z 74 are only of significant abundance in unsaturated fatty acids C18:1, C20:1, and C22:1 and their contribution to ion abundance can be corrected for. In the saturated fatty acids, the McLafferty ion m/z 74 is much more dominant and has no significant overlap from m/z 69, 70, 71, and 73. Therefore, to achieve the best accuracy of measurements, labeled unsaturated fatty acid methyl esters were isolated by TLC (AgNO3/silica gel, developed three times with toluene at -20C), reduced (dissolved in methanol; 2 h under oxygen-free H2/Pt catalyst), and measured again by GC/MS. Measurement of Corrected Mass Isotopomers in the Ion Cluster m/z 74 to 76 | After measurement of the ion cluster m/z 74 to 76, a correction method for the contribution of H transfer and natural isotopes to ion abundance was applied, based on correction matrices described by . For different fatty acids, correction matrices were constructed using the relative abundances of the ion clusters m/z 74 to 76 of reference substance with natural isotope abundance. For each fatty acid, the ratio of intensities of m/z 75 and m/z 74 was very constant over a concentration range of fatty acids of 102. However, it was susceptible for changes in tuning of the mass spectrometer. Inhomogeneous Distribution of 13C Label in Fatty Acids and the Steady-State Assumption | Labeling experiments with a complex medium (multiple carbon sources) seems to be at first sight an offense against a dogma for metabolic flux analysis. However, because we are interested in the status of metabolic fluxes in vivo, it seems not to make much sense to grow developing embryos on one carbon source because in vivo multiple carbon sources are involved. To properly interpret the fractional abundance of different isotopomers in 13C-labeled products, an assumption of metabolic and isotopic steady state is required . This means that the labeling pattern in the end products can be interpreted in terms of metabolic fluxes only if the metabolic fluxes and the distribution of labeled molecule species in the metabolic network are constant during the labeling experiment. To justify this assumption, the influx of labeled Glc and unlabeled Suc into the metabolic network must be constant. However, during development, the embryo may change its preferences for carbon sources. This is suggested by the fact that during embryo development of B. napus embryos, the Suc to hexose ratio in the endosperm liquid increases (see Table ; ). With the increase of Suc concentration in the endosperm liquid and the onset of storage product accumulation, induction of Suc synthase is reported in seeds of B. napus and induction of Suc synthase mRNA and a Suc transporter mRNA are observed in seeds of Arabidopsis . Thus, the relative ratio of uptake of [U-13C6]Glc and Suc may change during embryo development. In fact, we observed in mass spectra of all labeled fatty acids that a small fraction of the fatty acid molecules was very highly labeled from [U-13C6]Glc, whereas the bulk of fatty acids were formed from labeled Glc and unlabeled Suc, which were incorporated at a ratio close to that provided (see Fig. ). One explanation for these populations of highly labeled fatty acid molecules is that at the beginning of the labeling experiment, the embryos of mid-cotyledon stage are adapted to hexose as the main carbon source and, thus, the initial fatty acid and amino acid synthesis may mainly consume [U-13C6]Glc. An alternative explanation, which we cannot exclude, is that the different fatty acid labeling patterns arise from fatty acids produced in two different types of cells. Figure 6 | Predicted and measured relative abundance of mass isotopomers of C18:1 methyl ester, after growth on [U-13C6]Glc and unlabeled Suc. Predicted and measured relative abundance of mass isotopomers of C18:1 methyl ester, after growth on [U-13C6]Glc and unlabeled Suc. From the fragment C18:1(1-2), representing the terminal biosynthetic acetate unit, the distribution of mass isotopomers was predicted. The polymerization of nine acetate units to a C18 fatty acid molecule can be described by a binominal distribution (compare with ): (p + (q1 + q2))9. q1, Fraction of (13C1) acetate; q2, fraction of (13C2)-acetate; and p = 1 - q1 - q2, the fraction of (12C2) acetate. The expansion of this binominal distribution leads to the relative abundance of mass isotopomers (m0, m1, m2... m2n) of the fatty acid . After labeling with [U-13C6]Glc (diluted 20:80 with Suc, as related to hexose units), the fractional labeling in C18:1(1-2) was determined to q1 = 0.084, q2 = 0.267, and p = 0.649. This equals an average 13C content of 31%. The predicted distribution (white bars) and the measured molecular ion of C18:1 (black bars) are similar, indicating that the majority of fatty acid molecules was formed from acetate units labeled as measured in the fragment C18:1(1-2). However, a 4% fraction of fatty acid molecules was much higher labeled. Thus, during fatty acid biosynthesis, a small part of fatty acid molecules has been made mainly from labeled Glc, whereas the bulk of fatty acids was made by a homogeneous mixture of about 30% labeled and 60% unlabeled hexose units. Backmatter: PMID- 12226515 TI - Tandemly Duplicated Safener-Induced Glutathione S-Transferase Genes from Triticum tauschii Contribute to Genome- and Organ-Specific Expression in Hexaploid Wheat AB - Glutathione S-transferase (GST) gene expression was examined in several Triticum species, differing in genome constitution and ploidy level, to determine genome contribution to GST expression in cultivated, hexaploid bread wheat (Triticum aestivum). Two tandemly duplicated tau class GST genes (TtGSTU1 and TtGSTU2) were isolated from a single bacterial artificial chromosome clone in a library constructed from the diploid wheat and D genome progenitor to cultivated wheat, Triticum tauschii. The genes are very similar in genomic structure and their encoded proteins are 95% identical. Gene-specific reverse transcriptase-polymerase chain reaction analysis revealed differential transcript accumulation of TtGSTU1 and TtGSTU2 in roots and shoots. Expression of both genes was induced by herbicide safeners, 2,4-dichlorophenoxyacetic acid and abscisic acid, in the shoots of T. tauschii; however, expression of TtGSTU1 was always higher than TtGSTU2. In untreated seedlings, TtGSTU1 was expressed in both shoots and roots, whereas TtGSTU2 expression was only detected in roots. RNA gel-blot analysis of ditelosomic, aneuploid lines that are deficient for 6AS, 6BS, or 6DS chromosome arms of cultivated, hexaploid bread wheat showed differential genome contribution to safener-induced GST expression in shoots compared with roots. The GST genes from the D genome of hexaploid wheat contribute most to safener-induced expression in the shoots, whereas GSTs from the B and D genomes contribute to safener-induced expression in the roots. Keywords: Introduction : Glutathione S-transferases (GSTs) belong to multigene families common to all plants . They are well known for their responses to numerous endogenous and xenobiotic stresses, and glutathione conjugation of toxic electrophilic molecules. The roles of GST proteins in endogenous plant metabolism, as well as their role in stress tolerance, have yet to be clearly defined. GST gene expression is induced after exposure to many stresses, including biotic stresses such as pathogen attack and fungal elicitors, and abiotic stresses such as heat shock, cold, high salt, UV light exposure, heavy metals, and herbicides. Phytohormone treatments such as ethylene, auxins, abscisic acid (ABA), methyl jasmonate, and salicylic acid have also been shown to induce expression of GSTs (for review, see ; ; ). Induction of GST expression by so many diverse stimuli implies that plant GSTs are critical in plant response to stress, either by participating in the signal transduction process and/or detoxifying harmful compounds produced in response to or as a result of a given stress. It is likely that GST gene expression is induced by conditions that lead to oxidative stress . The encoded GST proteins play an important but poorly understood role in plant response to stress, possibly through the central role of antioxidant function. GST enzymatic activity could involve direct glutathione conjugation to toxic electrophilic molecules, or gluta-thione-dependent peroxidase activity, using glutathione as reductant for the detoxification of toxic oxygen species, oxygen radicals, and lipid peroxides formed during or after plant stress . One biochemical function of GST proteins that is well defined is their role in herbicide metabolism in crops. GSTs are the predominant detoxification enzymes in maize (Zea mays) and cereal crops that are responsible for metabolism of triazine herbicides, acetamide herbicides, and certain graminicides, such as fenoxaprop-ethyl in wheat . Herbicide-detoxifying GSTs have been well characterized in maize and soybean (Glycine max; ; , ; ; ; ; ), but have also been identified and partially characterized in wheat . Herbicide safeners are chemical compounds that increase the tolerance of certain grass crops (e.g. maize, grain sorghum [Sorghum bicolor], wheat, rice [Oryza sativa]) to herbicides . Herbicide safeners protect the crop plant by increasing herbicide metabolism and detoxification pathways . The increase in metabolism results from an increase in the activity of herbicide detoxification enzymes, such as GSTs, cytochrome P450-dependent monooxygenases, and glucosyltransferases . Despite the widespread agronomic use of safeners and information about their effects on GST and P-450 enzymatic activity, there is little information on the precise molecular mechanism for safener induction of the genes that encode these herbicide-detoxifying enzymes. A safener-binding protein and its activity have been characterized in maize seedlings, along with its gene expression patterns . An area that is especially lacking in information is the identification and characterization of important regulatory sequences present in the promoters of safener-responsive GST genes, and the transcription factors that bind to these DNA sequences in agronomically important grasses such as maize, rice, and wheat. Molecular analyses of plant GST genomic sequences have focused mainly on dicot species, and have examined GST gene expression in response to pathogen attack, plant hormones, or stress treatments. Promoter regulatory elements that confer increased GST gene expression in response to auxins, ethylene, salicylic acid, hydrogen peroxide, heat shock, heavy metals, and plant pathogens have been identified in several dicot species including potato (Solanum tuberosum), tobacco (Nicotiana tabacum), Arabidopsis, soybean, and carnation (Dianthus caryophyllus; ; ; ; ; ; ; ; ). Safener-responsive expression of GSTs and their cDNA sequences have been reported for maize, wheat, and rice . However, among these safener-responsive genes in monocot crops, only the maize GST-27 promoter has been partially characterized . Our studies have utilized the diploid wheat Triticum tauschii (synonymous with Aegilops tauschii and Aegilops squarrosa) as a model plant and genome to understand regulation of GST expression in grass crops with large and/or polyploid genomes . Previous research focused on a herbicide safener-induced GST isozyme that was purified from T. tauschii using anion-exchange and affinity chromatography, and was biochemically characterized . This safener-inducible GST isozyme can use the chloroacetamide herbicide dimethenamid as a substrate , where its conjugation with reduced glutathione results in metabolic detoxification of the herbicide . In subsequent studies, a corresponding cDNA was isolated from T. tauschii and was used to map the homoeologous GST genes to a chromosome arm in cultivated, hexaploid bread wheat (Triticum aestivum) and to a linkage group in barley (Hordeum vulgare; , ). Here, we report the analysis of genomic sequences for the safener-inducible, tau class GST genes (TtGSTU1 and TtGSTU2) and characterization of their expression profiles in T. tauschii. We also utilized ditelosomic, aneuploid wheat lines and various wheat species that differ in genome constitution and/or ploidy level to determine genome contributions to expression patterns in cultivated, hexaploid bread wheat. Our results provide novel evidence for differential genome contributions to constitutive and inducible gene expression in organs of hexaploid wheat, and differences in expression between closely related gene family members that are organized as a tandem repeat in a large grass genome. RESULTS : Isolation and Sequence Analysis of GST Genes from T. tauschii | Three different GST genes were isolated by screening a T. tauschii large DNA insert genomic library . The genomic library was screened with the safener-inducible GST TSI-1 (renamed TtGSTU1; ) cDNA that had been isolated previously from T. tauschii . A total of four positively hybridizing genomic clones were obtained and further analyzed. DNA gel-blot analysis of the four bacterial artificial chromosome (BAC) clones digested with different restriction enzymes showed that BAC 1 (insert size of 150 kb) and BAC 4 (insert size of 130 kb) contained overlapping genomic fragments, and each BAC clone appeared to contain at least two GST genes. Subsequent experiments focused on analyzing only BAC 1. BAC 1 contained an approximately 14-kb EcoRI/KpnI-hybridizing fragment. An XhoI site within the 14-kb fragment was utilized to separate this fragment into two smaller fragments of 8.5 and 5.3 kb. Each smaller fragment contained one of two tandemly repeated GST genes, TtGSTU2 and TtGSTU1 (accession no. ; Fig. ). These results are consistent with DNA gel-blot analysis (using XbaI as the restriction enzyme) of the cultivated, hexaploid bread wheat genome , where it was hypothesized that the homoeologous GST genes were represented by at least two copies in each of the three wheat genomes. Sequence analysis of the entire 14-kb fragment showed that the coding region and untranslated regions (UTRs) of TtGSTU1 are identical to the TtGSTU1 cDNA . Both of the tandemly duplicated genes contain an intron that interrupts the coding region at the same location, although the length and sequence of the intron varies between the two genes. TtGSTU1 has a single intron of 99 bp, whereas TtGSTU2 has a 319-bp intron , suggesting that these are tau class, or type III, GST genes . Figure 1 | Restriction map of the 14-kb region of the BAC 1 clone from T. tauschii Restriction map of the 14-kb region of the BAC 1 clone from T. tauschii, containing the two tandemly duplicated GST genes. Restriction sites are: B, BamHI; E, EcoRI; H, HindIII; K, KpnI; Xb, XbaI; and Xh, XhoI. LE, Left end (position 0); RE, right end (position 13, 710). DNA gel-blot and sequence analysis of BAC 3 showed that it contained only the TtGSTU1 gene, and was not analyzed further. DNA gel-blot analysis of the BAC 2 clone, digested with several restriction enzymes, showed a single, weakly hybridizing band (data not shown). BAC 2 was found to contain a related, yet divergent, GST-like sequence, which was named TtGSTU3 (accession no. ). The nucleotide sequence of the TtGSTU3 gene's coding region is 76% identical to the corresponding region of the TtGSTU1 gene, but this gene apparently does not contain an intron. Comparison of the deduced amino acid sequences of the three GST genes showed that the TtGSTU1 and TtGSTU2 proteins are 95% identical and 96% similar, whereas the TtGSTU1 and TtGSTU3 proteins are 68% identical and 80% similar (Fig. A). There is a 31-amino acid residue extension at the C terminus of the TtGSTU3-encoded protein, compared with both TtGSTU1 and TtGSTU2 (Fig. A). These three GSTs belong to the tau class of plant GSTs, according to the classification system proposed by and . The encoded proteins contain the triplet of amino acid residues His-Lys-Lys, which are conserved only in the tau class GST sequences, and are located at position 53 to 55 (numbering of amino acid residues is according to the TtGSTU1 sequence throughout the text, unless stated otherwise). The conserved triplet of His-Asn-Gly in the tau class is also conserved in these three GSTs at position 61 to 63 (Fig. A). TtGSTU3 is a unique GST, however, in that it appears to be a tau class GST protein based on its deduced amino acid sequence, but its gene sequence does not contain a single intron, which is characteristic of tau class GST genes. Figure 2 | Comparison of tau class GST proteins from T. tauschii Comparison of tau class GST proteins from T. tauschii. A, Alignment of the deduced amino acid sequences of TtGSTU1, TtGSTU2, and TtGSTU3. Identical amino acids and conservatively substituted residues are shaded with light and dark gray, respectively. The amino acid residue where exon 2 of TtGSTU1 and TtGSTU2 starts is indicated by an asterisk above the sequences. GenBank accession numbers are as follows: TtGSTU1 and TtGSTU2 , and TtGSTU3 . B, Hydropathy profile of the TtGSTU1 protein, as determined by the method of using a window of 10 amino acid residues. The hydropathy profiles of the TtGSTU1 and TtGSTU2 proteins, as determined by the method of , are almost identical. Both proteins contain a strong hydrophilic region around amino acids 110 to 130 (Fig. B, only TtGSTU1 is shown). This hydrophilic region is located where exon 2 of each gene starts (at amino acid 109 in TtGSTU1 and TtGSTU2; Fig. A). The function of this hydrophilic region is not known, but may play an important role in substrate binding and/or substrate specificity. Alternatively, this region may play a role in determining the intracellular localization of the GST proteins. This region in TtGSTU1 and TtGSTU2 is very similar in sequence to the same region of the mouse (Mus musculus) mGSTA4-4 protein (amino acid residues 110 --120), which also contains several charged residues and was shown to be important for a proposed electrostatic interaction with the plasma membrane of mouse liver hepatocytes . Hydropathy analysis of the TtGSTU3 protein showed minor differences relative to TtGSTU1, although slightly less hydrophilic character was noted around amino acids 110 to 130 (Fig. A). Repeat Elements Are Dispersed in the Intergenic Regions of the 14-kb Contiguous Sequence of BAC 1 | Two open reading frames (encoding the proteins TtGSTU1 and TtGSTU2) were found within the completely sequenced 14-kb interval from BAC 1. A BLAST search with this 14-kb interval identified sequences with similarity to several repeat elements from barley, maize, and the diploid wheat Triticum monococcum. The 14-kb interval contains about 2.6 kb of sequence upstream of the TATA box of TtGSTU2 until the EcoRI site (and start of the 14-kb contiguous sequence) is reached. In this region, about 1 kb of sequence (bp 1 --1,039 of accession no. ) was highly homologous to the long terminal repeat (LTR) of several retrotransposons, including the Angela-type retrotransposons from T. monococcum . This 1-kb portion also shares 82% nucleotide identity with a portion of the LTR of the BARE-1 copia-like retroelement from barley . Approximately 8 kb of sequence (bp 3,889 --11, 859 of accession no. ) is located between the TtGSTU2 and TtGSTU1 genes . Within this 8-kb intergenic sequence, there are three regions that show homology to the LTRs of several retrotransposons from cereal species. Sequences from position 5,261 to 5,890 are similar to the LTRs of Angela-type retrotransposons , and also shares about 83% identity with the barley BARE-1 copia-like retroelement over a 170-bp region . Position 8,062 to 8,451 was also similar to the LTRs of Angela-type retrotransposons and also shares 86% identity with the BARE-1 copia-like retroelement in barley over a 130-bp region . Position 8,775 to 9,124 shares about 90% nucleotide identity with LTRs of the BARE-1-like retrotransposons Angela-2, Angela-3, and Angela-4 . An 868-bp sequence flanks the 3' end of the TtGSTU1 mRNA until the KpnI restriction site (and end of the 14-kb contiguous sequence) is reached. A BLAST search with this sequence showed that a region of about 660 bp (toward the 3' end) shares 95% nucleotide identity with intron 10 of the T. tauschii starch synthase I gene , which may indicate the presence of a repeat family that is present throughout different parts of the wheat genome. This region also shares homology with the intron of the maize ACCase gene (accession no. ), which was noted to contain a number of retroelements, such as colonist-1 and colonist-2. 5'-Flanking Regions of TtGSTU1 and TtGSTU2 | In comparison with the 5'-UTR present in the TtGSTU1 cDNA, the start site for transcription initiation was set approximately 90 bp upstream of the MET start codon for both GST genes . The major distinguishing feature of the 5'-UTRs in the two genes is an "AC" dinucleotide simple sequence repeat, present just upstream of the translational start site. The TtGSTU1 genomic sequence contains eight copies of the AC repeat, whereas the TtGSTU2 gene has five AC repeats . An appropriately placed TATA box can be easily recognized 36 bp 5' to the transcription start site in both genes. Comparison of alignments of the nucleotide sequences of the promoters of TtGSTU1 and TtGSTU2 revealed that the two genes are very similar for about 800 nucleotides upstream of the TATA boxes, except for two large gaps that were noted: a 22-bp gap in the TtGSTU1 promoter at position -195, and a large gap in the TtGSTU2 promoter at position -363 . Figure 3 | Comparison of the 5'-flanking sequences of the two tandem GST genes, TtGSTU1 and TtGSTU2, isolated from the BAC 1 clone. Comparison of the 5'-flanking sequences of the two tandem GST genes, TtGSTU1 and TtGSTU2, isolated from the BAC 1 clone. Numbering is determined from the putative transcription start site by comparison of the 5'-UTR sequence of the TtGSTU1 gene with the TtGSTU1 cDNA. Potential transcriptional regulatory elements identified by homology searches are underlined in bold and labeled accordingly. Preliminary analysis of the UTRs and promoter regions using a plant transcription factor homology database identified several potential cis-acting regulatory elements. Sequences similar to the TATA box of many eukaryotic promoters were found at position -36 of both TtGSTU1 and TtGSTU2. The sequence RYACGTGGYR (R = A/G and Y = C/T), which was identified as an ABA-responsive element (ABRE) in Arabidopsis , was found in both TtGSTU1 and TtGSTU2 promoters at -230 and -249, respectively . An ethylene-responsive enhancer element AWTTCAAA (W = A/T) identified in the carnation GSTI gene and a fruit-ripening gene was located in the minus strand of the TtGSTU1 promoter at -525 . The sequences of TGTCTC and CATATG , which were related to auxin-responsive expression, were found in TtGSTU1 at -1,118 and -119 . The auxin-responsive element TGTCTC was also found in the minus strand of TtGSTU2 at -2,078 (not shown). A G box-like sequence containing the palindromic hexamer TAGCTA was found at position -308 in the TtGSTU1 gene and position -351 in the TtGSTU2 gene . This same sequence in the TtGSTU2 gene is actually a palindromic octamer, GTAGCTAC. Expression of GST Genes in T. tauschii following Treatment with Safeners and Hormones | RNA gel-blot analysis showed that constitutive GST expression was detected in roots of T. tauschii, but expression in shoots was low or not detectable . GST expression was highly induced by the safeners cloquintocet-mexyl and fluxofenim in T. tauschii shoots and roots. Induction of GST expression by these safeners in wheat shoots is consistent with previous results . Because potential ABA-, ethylene-, and auxin-responsive regulatory elements were identified (by homology searches) in the promoters of the GST genes , we also examined expression in response to these plant hormones. The synthetic auxin herbicide 2,4-D highly induced GST expression in roots and also induced expression in shoots, although to lower levels than the safeners in shoot tissue. ABA at a relatively high concentration (100 versus 10 mum) slightly increased GST expression in T. tauschii shoots . Ethylene had no effect on GST expression in T. tauschii shoots or roots . The fact that both safeners caused the greatest increase in GST expression, relative to the plant hormones, suggests that safeners may be tapping into a different regulatory pathway for induction of GST expression, or that the signal for induction may be stronger and/or longer lasting for a safener relative to the plant hormones examined. Figure 4 | RNA gel-blot analysis of GST expression in T. tauschii RNA gel-blot analysis of GST expression in T. tauschii after treatment with herbicide safeners and plant hormones. Total RNA (10 mug lane-1) was analyzed, and the blot was probed with the digoxigenin (DIG)-labeled TtGSTU1 cDNA coding region from T. tauschii. Etiolated seedlings were treated for 48 h with aqueous solutions containing 100 mum 2,4-dichlorophenoxyacetic acid (2,4-D), 10 mum or 100 mum ABA, 20 mug mL-1 2-chloroethyl-phosphonic acid for ethylene treatment, and 10 mum of the safeners cloquintocet-mexyl and fluxofenim. All treatment solutions were applied as vermiculite drenches to the seedlings 72 h after transferring pots to room temperature; the seedlings were exposed to the treatments for 48 h, then shoots and roots were harvested (5-d total growth period). Uniform loading was verified by comparing RNA intensities after ethidium bromide staining and by hybridizing the blots with a wheat actin probe (data not shown). Differential Expression of TtGSTU1 and TtGSTU2 in T. tauschii | Due to the high degree of similarity between the two GST genes at the nucleic acid level, we used semiquantitative reverse transcriptase (RT)-PCR to detect individual expression patterns of the two tightly linked GST genes in roots and shoots of T. tauschii. Primers were designed that would selectively amplify only one of the two tandemly duplicated GST genes. Constitutive GST expression in both roots and shoots was contributed mainly by TtGSTU1, and this gene also seems to be the most highly induced and expressed in safener-treated seedlings . TtGSTU1 was also induced by 100 mum ABA and 2,4-D in shoots . TtGSTU2 expression was not detectable in control shoots, but was weakly expressed in control roots. However, TtGSTU2 expression could be detected in control shoots when amplifying for 35 cycles (data not shown), compared with the 25 cycles shown in Figure . TtGSTU2 expression was induced by safeners, 100 mum ABA, and 2,4-D in both roots and shoots, although the level of expression was always less than that of TtGSTU1 . Ethylene had little effect on TtGSTU1 and TtGSTU2 expression in either shoots or roots . Table I | Primers for RT-PCR Figure 5 | Semiquantitative RT-PCR analysis of individual GST gene expression in T. tauschii Semiquantitative RT-PCR analysis of individual GST gene expression in T. tauschii. Expression was analyzed in shoots (A) and roots (B) of T. tauschii. Seedling growth conditions and treatments, and total RNA samples were the same as those used for RNA gel-blot analysis shown in Figure . Total RNA (5 mug) was used to synthesize first strand cDNA, and a fraction (1/20) of the first strand cDNA was used as template for PCR amplification of individual gene transcripts. Ethidium bromide-stained RT-PCR products were separated in 1.2% (w/v) agarose gels and analyzed with 1D image analysis software (Eastman-Kodak, Rochester, NY). The wheat actin gene was used as a constitutively expressed control gene and loading control. We could not detect the expression of TtGSTU3 by RT-PCR in either untreated T. tauschii seedlings or treated seedlings (data not shown), which suggests that TtGSTU3 may be transcriptionally inactive (a pseudogene), or has an expression pattern entirely different from TtGSTU1 and TtGSTU2. This might include expression in other tissues or organs, or possibly under different stress conditions. GST Expression in Cultivated, Hexaploid Bread Wheat and Other Triticum Spp. | In addition to examining GST expression in T. tauschii, we also investigated GST expression in cultivated, hexaploid bread wheat (ABD genomes). Homoeologous GST genes were previously mapped to the short arms of group 6 chromosomes in cultivated, hexaploid bread wheat via Southern hybridization , using the T. tauschii TtGSTU1 cDNA to probe ditelosomic, aneuploid wheat lines of group 6 chromosomes that are missing either the short or long arm of chromosome 6A, 6B, or 6D. RNA gel-blot analyses of these same lines were conducted, again using the TtGSTU1 cDNA as a probe. It is clear that in safener (cloquintocet-mexyl)-treated wheat shoots, removal of the tandem GST genes in the chromosome 6DL line (missing the short arm only) almost completely eliminated GST expression . However, removal of the homoeologous GST genes in the chromosome 6AL and 6BL lines (missing the short arms of 6A and 6B, respectively) had relatively minor effects on GST expression levels, even though the 6A or 6B GST alleles were no longer present. Figure 6 | RNA gel-blot analysis of GST expression in chromosome group 6 ditelosomic, aneuploid wheat lines. RNA gel-blot analysis of GST expression in chromosome group 6 ditelosomic, aneuploid wheat lines. GST expression was analyzed in aneuploid lines derived from cultivated, hexaploid bread wheat cv Chinese Spring, in both roots and shoots. Treatments included the control (unsafened) or 10 mum of the safener cloquintocet-mexyl for 48 h as described in Figure . Total RNA (10 mug lane-1) was analyzed, and the blot was probed with the DIG-labeled TtGSTU1 cDNA coding region from T. tauschii. L, Long arm only (missing the short arm); S, short arm only (missing the long arm); CS, cultivated, hexaploid bread wheat cv Chinese Spring base genotype. GST expression in cultivated, hexaploid bread wheat roots showed a different pattern when individual group 6 chromosome arms were removed . When the short arms of chromosomes 6A and 6D were removed (in lines 6AL and 6DL), there was no detectable difference in the level of GST expression in safener-treated roots. However, when the short arm of chromosome 6B was removed (in the 6BL line), there was a significant decrease in GST gene expression. This suggests that in safener-treated cultivated, hexaploid bread wheat roots, the GST allele(s) on chromosome 6BS is a major contributor to GST expression; in contrast, the GST alleles on chromosome 6DS are most important in safener-treated cultivated, hexaploid bread wheat shoots and the GST alleles on chromosomes 6AS and 6BS appear to be minor contributors in shoots. This difference in genome contribution to GST gene expression in roots and shoots is also noted in Figure , where several Triticum spp. were analyzed for their response to the safener cloquintocet-mexyl in roots and shoots. Triticum spp. were chosen that differ in genome constitution (diploids A only or D only, and tetraploid AB) and were compared with cultivated, hexaploid bread wheat (ABD genomes). GSTs were highly induced in the shoots of cultivated, hexaploid bread wheat (AABBDD) and T. tauschii (DD), although there was also a relatively high constitutive level of GST mRNA in cultivated, hexaploid bread wheat cv Chinese Spring (AABBDD; Fig. ). In contrast, there was no induction of GST expression in the shoots of T. monococcum (AA) and relatively minor induction in Triticum turgidum subsp. durum (AABB) shoots . However, the safener increased GST expression in the roots of all four wheat species, although the lowest induction occurred in T. monococcum roots . These results confirm our previous observation that GST genes in the D genome of wheat are contributing the vast majority of safener-induced expression in shoots; however, in roots, it appears that GST genes in both the B and D genomes are the major contributors to GST expression (Figs. and ). Figure 7 | RNA gel-blot analysis of wheat genome contribution to GST expression. RNA gel-blot analysis of wheat genome contribution to GST expression. Four wheat species were examined, each differing in genome composition and/or ploidy level. Wheat seedlings were either unsafened (control) or treated with 10 mum of the safener cloquintocet-mexyl for 48 h as described in Figure . Total RNA (10 mug per lane) was analyzed, and the blot was probed with the DIG-labeled TtGSTU1 cDNA coding region from T. tauschii. AABBDD, Cultivated, hexaploid bread wheat; AABB, T. turgidum subsp. durum; AA, T. monococcum; DD, T. tauschii. Cultivated, hexaploid bread wheat cv Chinese Spring was used in all of our expression studies because it is the base genotype from which the ditelosomic aneuploids were derived . There was a relatively high level of constitutive GST expression in shoots of cv Chinese Spring, compared with T. turgidum subsp. durum, T. monococcum, and T. tauschii, yet it also showed safener induction (Figs. and ). To further investigate this finding, three other varieties of cultivated, hexaploid bread wheat (CITR 9038, Ernie, and P25R57) were used to compare and contrast with cv Chinese Spring. These four wheat varieties differed in both basal level of GST expression and safener induction in shoots in response to the two safeners cloquintocet-mexyl and fluxofenim . The varieties Chinese Spring and CITR 9038 showed the highest basal levels of GST expression, whereas there was no detectable basal GST expression in P25R57. GST expression was highly induced by both safeners in CITR 9038, which is consistent with the large increases in GST enzyme activity (in response to both safeners) reported previously for this wheat line . The variety Ernie showed a higher level of induction of GST expression after treatment with both safeners relative to the variety P25R57 . These varying ranges of constitutive and safener-induced GST mRNA levels among wheat varieties and genetic lines are consistent with the broad range of constitutive and safener-increased GST activity levels reported previously , and provide further evidence that there is a great amount of genetic diversity present for GST expression levels in wheat germplasm. Figure 8 | Cultivated, hexaploid bread wheat varietal response to safeners and differential GST expression. Cultivated, hexaploid bread wheat varietal response to safeners and differential GST expression. Wheat seedlings were either unsafened (control) or treated with 10 mum of the safener cloquintocet-mexyl or fluxofenim for 48 h as described in Figure . CS, Chinese Spring. Although it was not determined which genome(s) is contributing the constitutive expression observed in cultivated, hexaploid bread wheat cv Chinese Spring shoots (Figs. and ), the variable constitutive and safener-induced expression levels observed in cultivated, hexaploid bread wheat varieties suggest genome complexities indicative of further evolution and diversification of homoeologous GST gene family members after their duplication by polyploidization of the A, B, and D genomes in cultivated, hexaploid bread wheat. The diversification of TtGSTU1 and/or TtGSTU2 expression patterns in both T. tauschii (by a duplication event) and cultivated, hexaploid bread wheat (by polyploidization) is consistent with a relaxation of purifying selection in duplicate genes of recent origin . DISCUSSION : Genomic Organization | Screening a T. tauschii BAC genomic library (average insert size of 119 kb; ) allowed us to isolate the BAC 1 clone (insert size of 150 kb) containing the two tandemly repeated GST genes TtGSTU1 and TtGSTU2, both contained within a 14-kb restriction fragment. Within this 14-kb fragment, there were only two open reading frames, both coding for the two highly homologous tau class GSTs. These two GSTs contain the entire transcription units in two exons interrupted by one intron, with complete conservation of intron position. However, the length and nucleotide sequences of the two genes' introns are different. Similar GST gene clusters have been identified in the carnation genome as well as in the Arabidopsis genome . In contrast to TtGSTU1 and TtGSTU2, the two tightly linked carnation GST genes were obtained from a single lambda phage clone, and contained 10 exons and nine introns , characteristic of zeta class plant GSTs. These two GST genes showed conservation in their intron positions and also the length and nucleotide sequences of the introns . The recent findings from carnation, Arabidopsis, and wheat all indicate that GSTs from the same subclass are grouped on chromosomes as tandem duplications . Comparisons of TtGSTU1 and TtGSTU2 sequences with rice genome sequences from GenBank and the draft recently completed by the Beijing Genomics Institute identified at least two clusters of tandemly duplicated tau class GST genes in rice. One of these clusters contained 20 closely related tau class GST genes on a single BAC clone from rice chromosome 10 (accession no. ). At least 12 of these 20 genes were expressed in leaves, roots, or callus tissue based on their identity with rice GST cDNAs or ESTs. Another tandem duplication of tau class GST genes was found on a rice BAC clone from chromosome 1 (accession no. ). Among these 22 rice tau class GST genes, both TtGSTU1 and TtGSTU2 show the highest similarity to the same OsGSTU4 gene present within the 20-gene cluster on chromosome 10, suggesting that the duplication event leading to TtGSTU1 and TtGSTU2 may have occurred after the divergence of the rice and wheat genomes. The high degree of similarity between TtGSTU1 and TtGSTU2 in both the coding and 5'-flanking regions (except for a few insertions or deletions in their promoters; Fig. ) supports the notion that these genes arose due to a recent duplication event. TtGSTU1 and TtGSTU2 map to wheat chromosome 6 , which is proposed to be syntenic with rice chromosome 2 . Our finding that the rice genes with the highest similarity to TtGSTU1 and TtGSTU2 map to chromosome 10 may suggest a lack of microsynteny for these regions , or that the true rice orthologs of the wheat GST genes may be present in the estimated 8% of the rice genome not covered by the draft sequence from the Beijing Genomics Institute. Among cereal crops, rice appears to have a predominance of tau class GST genes, which was also reported for hexaploid wheat but in contrast with maize, which was noted to have more phi class GST genes . However, these observations in wheat and maize have been based mainly on sequence analysis and immunological characterization of GST enzymes that possess herbicide detoxification activities. A complete sequence analysis of the entire rice, wheat, and maize genomes would be necessary to confirm this preliminary speculation on the relative abundance of specific GST subclasses in cereal crops. BLAST searches showed that the intergenic regions of the 14-kb fragment contained sequences similar to the LTRs of retrotransposons, which is in accord with previous reports that showed a large portion of the wheat genome consists of repetitive DNA elements . Certain types of retrotransposons are preferentially located near the centromeres of the chromosomes of grass species, such as wheat, barley, sorghum, and maize . Mapping the chromosomal location of a homologous GST in barley with the TtGSTU1 cDNA showed that it is located very close to the centromere of chromosome 6HS . Based on these results, we postulate that the precise chromosomal location of TtGSTU1 and TtGSTU2 in T. tauschii and cultivated, hexaploid bread wheat may be the same as in barley (i.e. near the centromere of chromosome 6DS). This would also be consistent with our finding of the retrotransposon-like sequences in the intergenic regions of the BAC 1 clone from T. tauschii. Expression Analyses in T. tauschii | Although TtGSTU1 and TtGSTU2 have similar gene structures and share very high identities at the amino acid and DNA sequence levels, they display different expression patterns in the roots and shoots of T. tauschii seedlings, as well as in response to various chemical inducers. Our RT-PCR results show that TtGSTU1 is constitutively expressed in both roots and shoots of 5-d-old T. tauschii seedlings grown under control (untreated) conditions. Using RNA gel-blot and immunoblot analyses, a similar expression pattern was reported in maize for the phi class GST-29 gene (ZmGSTF1), which was found to be constitutively expressed in a number of maize tissues . Unlike maize GST-29, which showed a minimal increase in transcript levels upon herbicide safener treatment , TtGSTU1 was strongly induced by herbicide safeners in shoots. Using RT-PCR and gene-specific primers, TtGSTU2 transcripts were detected in control roots, but not shoots, of 5-d-old etiolated seedlings. The phi class maize GST-27 gene (ZmGSTF2), like TtGSTU2, was constitutively expressed in roots, and no expression was detected in other tissues . Furthermore, herbicide safener treatments caused dramatic increases in the expression of both TtGSTU1 and TtGSTU2 and the maize GST-27 gene . These genes also showed induction in response to treatments with high levels of 2,4-D in maize leaves and T. tauschii shoots , although the level of induction was always lower than with the safeners in these aerial tissues. Induction of expression in response to ethylene was not detected for the GSTs in T. tauschii or the maize GST-27 gene, except at very high concentrations that also led to phytotoxicity . These similar expression profiles imply that the promoters of these inducible wheat and maize GST genes may contain similar safener-responsive regulatory elements. An alignment of the promoters of TtGSTU1 and TtGSTU2 with the maize GST-27 promoter (reported in patent no. WO 93/01294, ; GenBank accession no. ) did not reveal any conserved regions, and a search with the plant transcription factor homology database only identified several CCAAT box sequences within the first 2 kb of the maize GST-27 promoter (D. Riechers and S. Moose, unpublished data). Because TtGSTU1 and TtGSTU2 are tau class GSTs, and the maize GST-27 is a phi class GST, there may be different safener-responsive regulatory elements in their promoters. Detailed functional analyses of the promoter sequences for these safener-inducible GSTs in maize and wheat may identify important regulatory elements that govern safener-induced expression. In comparison with two wheat GSTs reported in the literature, TtGSTU1 shares 30% amino acid sequence identity with the phi class GstA1 protein and 25% amino acid sequence identity with the zeta class TA-GSTZ1 protein . GstA1 was further characterized for its expression patterns in wheat, detected at both the mRNA and protein levels . GstA1 was induced in response to challenge by pathogens and a cell-free fungal extract, but not by xenobiotics , suggesting a role for this gene in plant defense reactions against pathogen attack. In contrast, both TtGSTU1 and TtGSTU2 were highly induced by safener treatment in T. tauschii, as well as the synthetic auxin 2,4-D (and the phytohormone ABA to a limited extent). These results suggest that the tau class TtGSTU1 and TtGSTU2 proteins have important roles in xenobiotic metabolism in wheat, and may also have significant yet undefined roles in response to plant stresses. Expression Analyses in Cultivated, Hexaploid Bread Wheat and Other Triticum Spp. | In addition to examining GST expression in the diploid wheat T. tauschii, gene expression was investigated in cultivated, hexaploid bread wheat, other diploid and tetraploid wheat species, as well as in ditelosomic aneuploid wheat lines that are missing individual arms of group 6 chromosomes. GST loci were previously mapped to the short arms of chromosomes 6A, 6B, and 6D in cultivated, hexaploid bread wheat, using the TtGSTU1 cDNA as a probe . This allowed us to use the same probe to detect homoeologous GST transcripts in ditelosomic, aneuploid lines of cultivated, hexaploid bread wheat. The results demonstrated that GST expression in safener-treated wheat shoots was mainly contributed by GSTs from the D genome, whereas GSTs from both the B and D genomes contribute to safener-induced GST expression in wheat roots. The GST gene(s) on chromosome 6AS are not expressed to a significant extent in either control or safener-treated roots or shoots of T. monococcum, and also do not appear to be significant contributors to expression in safener-treated cultivated, hexaploid bread wheat seedlings. The most straightforward explanation to describe these expression patterns in wheat shoots is that the tandem GST genes on chromosome 6DS contain safener-responsive element(s) in their promoters or UTRs that are lacking in GST genes on chromosomes 6AS and 6BS. However, our data with ditelosomic, aneuploid wheat lines do not rule out the possibility that a regulatory factor could also be located on chromosome 6DS that controls the response to safeners in wheat shoots. RNA gel-blot analysis of Triticum spp. with different genome constitutions showed that safener treatment dramatically increased GST expression in the shoots of cultivated, hexaploid bread wheat (AABBDD) and T. tauschii (DD), but not in T. monococcum (AA) or T. turgidum subsp. durum (AABB). These results are consistent with the results of expression analyses from the ditelosomic, aneuploid wheat lines of group 6 chromosomes, clearly indicating that GST genes in the D genome of wheat contribute most toward safener-induced GST expression in shoots. When assaying for GST enzyme activity in different Triticum spp. with a herbicide substrate, much higher GST activities were present in safener-treated shoots of both cultivated, hexaploid bread wheat and T. tauschii, relative to T. turgidum subsp. durum or other wheat species that lack the D genome . The results of these studies confirm that the D genome is an important source of GST isozymes that are involved in the safener response in hexaploid and diploid wheats containing the D genome . Expression analyses in safener-treated roots from various Triticum spp. were also consistent with results found with the ditelosomic, aneuploid wheat lines, although the results were different from those observed in shoots. Safener treatment increased GST transcript levels in the roots of species or lines that contain the B or D genomes. Thus, GST genes in the B and D genomes contribute to safener-induced GST expression in wheat roots, whereas GST genes from the D genome contribute most toward expression in safener-treated shoots. This pattern of genome- and organ-specific expression of GSTs implies that the genes' promoters or untranslated sequences may contain different transcriptional regulatory elements that control gene expression in wheat roots versus shoots. Interestingly, comparisons of diploid wheats (T. monococcum and T. tauschii) with cultivated, hexaploid bread wheat indicate that genome contributions to safener-induced GST expression appear to be conserved following the polyploidization of cultivated wheat (Figs. and ). However, differences in GST expression among cultivated, hexaploid bread wheat varieties show that constitutive and safener-induced GST expression is variable within cultivated, hexaploid wheat (Fig. ; ). The recent GST gene duplication event that led to TtGSTU1 and TtGSTU2 may have permitted the diversification of gene expression patterns in T. tauschii . Further duplication of these genes by the poly-ploidization events that led to cultivated, hexaploid bread wheat may have allowed for their continued evolution within cultivated wheat (as noted by varietal differences in Fig. ). Future work will be aimed at characterizing the structure and expression patterns of the TtGSTU1 and TtGSTU2 gene homoeologs in T. monococcum and cultivated, hexaploid bread wheat. Additional functional studies of these homoeologous genes, which differ in their expression response to herbicide safeners, will help identify the regulatory elements and factors that are important for herbicide safener-induced GST gene expression in wheat. MATERIALS AND METHODS : Screening of a Triticum tauschii BAC Library | High-density filters of BAC clones were screened with the coding region of the cDNA encoding the safener-induced TtGSTU1 , previously isolated from T. tauschii. DNA hybridization and washing conditions were as reported by . Plasmid DNA from individual BAC clones was isolated using the alkaline lysis method and DNA insert sizes were estimated by pulse field gel electrophoresis according to . DNA Sequencing and Analysis | DNA fragments containing sequences of interest were subcloned into pBluescript SK+ (Stratagene, La Jolla, CA). Sequencing was conducted on both strands and reactions were performed at the sequencing center at the University of Illinois (Urbana) using the Big Dye kit (Perkin-Elmer Applied Biosystems, Foster City, CA) and an ABI Prism 377 (ABI, Sunnyvale, CA). DNA sequence analyses and amino acid alignments were performed using the AlignX tool of Vector NTI Suite V.6 software (InforMax, Inc., Bethesda, MD). Plant Material | For RNA extraction and analysis, seeds were planted in plastic pots containing vermiculite. Pots were watered to saturation with deionized water, covered with aluminum foil, and subjected to prechilling at 4C for 5 d to increase and synchronize seed germination. Pots were then removed from the cold and incubated at room temperature without light for a total of 5 d. For safener and plant hormone treatments, pots were transferred to room temperature, watered with deionized water, and incubated for 3 d. Then, the pots were watered with 10 mum cloquintocet-mexyl, 10 mum fluxofenim, 100 mum 2,4-D, 10 mum or 100 mum ABA, or 20 mug mL-1 2-chloroethyl-phosphonic acid (Sigma, St. Louis) for ethylene treatments, and incubated for another 48 h at room temperature. Roots and etiolated shoots were harvested separately, frozen in liquid nitrogen, and stored at -80C until RNA extraction. RNA Gel-Blot Analysis | Total RNA was isolated from plant tissues using TRIzol total RNA isolation reagent (Invitrogen Life Technologies, Carlsbad, CA) according to the manufacturer's instructions. The following manufacturer-recommended modification was used: Equal volumes of a high salt solution (1.2 m sodium citrate and 0.8 m NaCl) and isopropanol were added during the RNA precipitation step to decrease polysaccharide contamination and obtain >90% pure RNA (A260/A280 > 1.8). RNA was denatured at 55C in the present of formamide and formaldehyde and separated by electrophoresis in 1.2% (w/v) agarose gels (containing 0.4 m formaldehyde). Equal loading among wells was verified by ethidium bromide staining. RNA was transferred to Nytran N membranes (Schleicher & Schuell, Keene, NH) by capillary blotting in 10x SSC overnight. RNA was fixed to the membrane by cross-linking on an UV Stratalinker (Stratagene). Blots were prehybridized in DIG Easy Hyb buffer (Roche Applied Science, Indianapolis) for 4 h and hybridized to a DIG-labeled TtGSTU1 cDNA coding region probe overnight at 50C. The blots were washed at a final stringency in 0.1x SSC and 0.1% (w/v) SDS at 65C. Blots were developed with CDP-Star chemiluminescent substrate, then exposed to Hyperfilm (Amersham Biosciences, Piscataway, NJ). Semiquantitative RT-PCR | Total RNA (5 g) was annealed to an oligo(dT)12-18 primer (Invitrogen Life Technologies), then first strand cDNAs were synthesized using Superscript II RT (Invitrogen Life Technologies). The genomic sequences for TtGSTU1, TtGSTU2, and TtGSTU3 were used to design primers for subsequent PCR amplification. Gene-specific primers were designed based on the comparisons of DNA and deduced amino acid sequences of the three GST genes. The forward primer was the same for the three GST genes, and was located at the amino acids 28 to 33 of TtGSTU1 and TtGSTU2, or 29 to 34 of TtGSTU3 . The reverse primers were designed to be gene specific, and were located near the C-terminal sequences of the open reading frames . The wheat actin transcript served as an internal, constitutively expressed loading control. The two actin primers used for RT-PCR were designed from conserved sequences in the rice (Oryza sativa) actin gene (GenBank accession no. ). The reaction mixture contained 1 muL of first strand cDNA, 0.2 mm dNTPs, 1.0 mm MgCl2, 0.4 mum each primer, and 1.25 units of Taq polymerase (Invitrogen Life Technologies) in a total volume of 25 muL. PCR cycling conditions were as follows: an initial denaturation step at 95C for 10 min, 25 amplification cycles (95C for 50 s, 65C for 30 s, and 72C for 2.5 min), and a final elongation step at 72C for 10 min. For semiquantitative RT-PCR, linearity for each amplification was confirmed . Specificity of amplification for each GST gene was verified by using plasmid controls containing each individual gene fragment under the same PCR conditions as described above (data not shown). Backmatter: PMID- 12226516 TI - Successive Glycosyltransfer Activity and Enzymatic Characterization of Pectic Polygalacturonate 4-alpha-Galacturonosyltransferase Solubilized from Pollen Tubes of Petunia axillaris Using Pyridylaminated Oligogalacturonates as Substrates AB - Polygalacturonate 4-alpha-galacturonosyltransferase (pectin synthase) was solubilized from pollen tubes of Petunia axillaris and characterized. To accomplish this, an assay method using fluorogenic pyridylaminated-oligogalacturonic acids (PA-OGAs) as acceptor substrates was developed. When the pollen tube enzyme was solubilized with 0.5% (v/v) Triton X-100 and was incubated with PA-OGA and UDP-galacturonic acid (UDP-GalUA), successive transfer activity of more than 10 GalUAs from UDP-GalUA to the nonreducing end of PA-OGA was observed by diethylaminoethyl high-performance liquid chromatography. This activity was time- and enzyme concentration-dependent. The optimum enzyme activity was observed at pH 7.0 and 30C. Among the PA-OGAs investigated, those with a degree of polymerization of more than 10 were preferred as substrates. The crude pollen tube enzyme had an apparent Km value of 13 mum for the PA-OGA with a degree of polymerization 11 and 170 mum for UDP-GalUA. The characteristics of the P. axillaris pollen tube enzyme and the usefulness of fluorogenic PA-OGAs for the assay of this enzyme are discussed. Keywords: Introduction : Pectin, one of the major components of plant cell walls, is mainly composed of homogalacturonan (HGA), rhamnogalacturonan I, and rhamnogalacturonan II . HGA is a homopolymer of alpha-1,4-linked galacturonic acid (GalUA) partially methyl esterified at C-6 of GalUA . Pectic HGA- or polygalacturonic acid (PGA)-synthesizing polygalacturonate 4-alpha-galacturonosyltransferase (PGA-GalUAT; EC ) is a key enzyme for pectin biosynthesis. However, it has not been purified nor has its gene been cloned. There are reports on the activity of the suspended membrane-bound enzyme in mung bean (Phaseolus aureus; ), tomato (Lycopersicon esculentum), turnip (Brassica rapa; ), sycamore (Acer pseudoplatanus; ), tobacco (Nicotiana tabacum; ), and adzuki bean (Vigna angularis; ). Solubilization of the tobacco enzyme with detergents has been attempted; however, when it was solubilized with 40 mm CHAPS, the enzyme did not exhibit successive glycosyltransfer activity (consecutive addition of sugar residues) but transferred only a single GalUA residue . The solubilized tobacco enzyme has been shown to add GalUA from UDP-GalUA onto the nonreducing end of oligogalacturonic acid (OGA; ). To study PGA-GalUAT, we selected the growing pollen tube of Petunia axillaris as the enzyme source. This solanaceous petunia bears relatively large flowers (styles of approximately 5 cm) and can be transformed by an exogenous gene using standard techniques. Its pollen tube grows very rapidly through the transmitting tissue in the style; even under the in vitro system used in this study, its growth rate reached 60 mum h-1. Cell wall synthases appear to be highly expressed in P. axillaris pollen tube tissue. Pectin is localized in an outer layer of the cell wall, and its methylesterification is considered to control the mechanical strength and extensibility necessary for the growth of the tube . In addition to pectin methyltransferase and pectin methylesterase, PGA-GalUAT may be involved in this control. The assay for PGA-GalUAT reported in the literature is based on measurement of the radioactivity of the product incorporating [14C]GalUA from UDP-[14C]GalUA . However, because the degradation of acceptor substrates by glycosidases is not detected, the target enzyme may not be correctly evaluated. Polysaccharide-degrading enzymes are considered to be highly expressed where the corresponding polysaccharide synthase is expressed. Here, we report on a new assay method for PGA-GalUAT. The reducing ends of OGA were modified with 2-aminopyridine, a superior fluorescent-labeling reagent . The method was used to identify and characterize PGA-GalUAT prepared from the microsomal membrane fraction of the P. axillaris pollen tube. The enzyme showed successive glycosyltransfer activity, which is assumed to be a characteristic of polysaccharide synthase. The characteristics of the enzyme and the advantages of our new assay method are also reported and discussed. RESULTS : Pyridylaminated-OGA (PA-OGA) Preparation | OGAs obtained as partial hydrolyzates of PGA were pyridylaminated. PA-OGAs with a degree of polymerization (DP) ranging from 4 to 27 were separated by DEAE anion-exchange HPLC . Each fraction was rechromatographed. The purity of each PA-OGA used for the enzyme assay was checked by ensuring that it gave a single peak on DEAE HPLC and reversed-phase HPLC and that the mass of each PA-OGA corresponded with the value calculated using MALDI-TOF MS (data not shown). The relative fluorescence intensities of the PA-OGAs were almost the same except for that of the PA-OGA of DP 1 , as was also found for pyridylaminated neutral oligosaccharides . The peak area ratios for PA-OGAs with a DP > 2 are thus considered as molar ratios. Figure 1 | DEAE anion-exchange HPLC of PA-OGAs. DEAE anion-exchange HPLC of PA-OGAs. A mixture of PA-OGAs was analyzed by DEAE HPLC. The peak eluted at 35 min was identified as PA-OGA of DP 4 by matrix-assisted laser-desorption ionization time of flight mass spectrometry (MALDI-TOF MS). Arrows and numbers indicate the PA-OGA elution position and DP of the PA-OGA eluted, respectively. Table I | Fluorescence intensity of PA-OGAs Assay Procedure for PGA-GalUAT Using PA-OGAs | The assay for PGA-GalUAT was performed by incubating PA-OGA and UDP-GalUA with the crude enzyme (solubilized with 0.5% [v/v] Triton X-100). The chromatogram for the assay with 15 mum of the PA-OGA of DP 14 and 1 mm UDP-GalUA is shown in Figure A. The elution times of the products corresponded with those of standard PA-OGAs, showing that GalUA was successively transferred to PA-OGA. More than 10 GalUAs were transferred to the PA-OGA of DP 14 in 1 h. PA-OGAs with a DP up to 27 were produced in 2 h (Fig. B). PA-OGAs in the DP range of 15 to 17 initially increased but then decreased because they were also used as acceptor substrates. All PA-OGAs of DP > 18 gradually increased during the 2-h period. No glycosyltransfer activity was detected in the soluble fraction of the pollen tube extract (data not shown). Figure 2 | DEAE HPLC of products obtained with PGA-GalUAT using PA-OGA of DP 14. DEAE HPLC of products obtained with PGA-GalUAT using PA-OGA of DP 14. A, The enzyme reaction was conducted in the presence of 15 mum PA-OGA of DP 14 and 8.7 microunits of the crude enzyme for 0 to 60 min. Arrows and numbers indicate the PA-OGA elution position and DP of the PA-OGA eluted, respectively. B, Amounts of products plotted against incubation time. PA-OGAs of DP 15 , 17 (circle), 19 , 21 , 23 , 25 , and 27 are shown. Characterization of the P. axillaris Pollen Tube Enzyme | The time course of the reaction (Fig. A) shows that incorporation of GalUA increased linearly during the first 60 min. Linearity between the enzyme concentration and GalUA incorporation was observed for 0 to 5 units L-1 of the crude enzyme in a 30-min reaction (Fig. B). Figure 3 | Time- and enzyme concentration-dependence of the enzyme reaction. Time- and enzyme concentration-dependence of the enzyme reaction. A, Time course of GalUA incorporation into PA-OGA of DP 14. B, Relationship between enzyme concentration and incorporation of GalUA into PA-OGA. The enzyme reaction was continued for 30 min. The influence of the PA-OGA size on PGA-GalUAT activity was investigated . The smaller PA-OGAs (DP 5 --10) functioned as acceptors, although the enzyme activity was low, whereas PA-OGAs of DP > 12 worked as better acceptor substrates. The optimum pH for the enzyme activity was around 7 . The optimum temperature of the enzyme was about 30C . The enzyme was active over a wide temperature range and retained about 40% of its maximum activity even at 10C. PGA-GalUAT had no activity without divalent cations and was most activated by the addition of Mn2+ . Mg2+ and Ca2+ ions activated 20% of the activity of the Mn2+-treated enzyme. The optimum Mn2+ concentration for the enzyme activity was 5 to 10 mm . The apparent Km values for the PA-OGAs of DP 7, 11, and 14 were 44, 13, and 11 mum, respectively; for UDP-GalUA, the value was 170 mum. The apparent Vmax values for the PA-OGAs of DP 7, 11, and 14 were 120, 240, and 640 pmol min-1 mg-1 protein; for UDP-GalUA, the value was 480 pmol min-1 mg-1 protein. The enzyme solubilized with other detergents (1% [v/v] digitonin or 40 mm CHAPS) also had an activity of successive transfer of GalUA. The activity for unit weight of protein in a crude enzyme solubilized with these detergents were on a level with 0.5% (v/v) Triton X-100 (data not shown). Figure 4 | Effects of DP of acceptors on PGA-GalUAT activity. Effects of DP of acceptors on PGA-GalUAT activity. The reaction was conducted with 2.6 microunits of the crude enzyme and various PA-OGAs of the following DP: 5 (10.8 mum), 6 (14.2 mum), 7 (13.2 mum), 8 (12.3 mum), 9 (13.4 mum), 10 (13.2 mum), 11 (11.7 mum), 12 (14.2 mum), 13 (13.4 mum), 14 (12.1 mum), and 15 (14.0 mum). The initial transfer activities are shown relative to that of the PA-OGA of DP 14. Figure 5 | Effects of pH on PGA-GalUAT activity. Effects of pH on PGA-GalUAT activity. The crude enzyme (3.3 microunits) was incubated with 7.9 mum PA-OGA of DP 12. The buffers used were 100 mm MES-NaOH ( --circle --), HEPES-NaOH ( -- --), and Tris-HCl ( -- --). The enzyme activity at pH 7.0 was taken as 1.0. Figure 6 | Effect of temperature on PGA-GalUAT activity. Effect of temperature on PGA-GalUAT activity. The crude enzyme (4.3 microunits) and 6.8 mum PA-OGA of DP 15 were incubated for 30 min. The activities are shown relative to that at 30C. Table II | Effects of cations on PGA-GalUAT activity Figure 7 | Effect of Mn2+ concentration on PGA-GalUAT activity. Effect of Mn2+ concentration on PGA-GalUAT activity. The crude enzyme (1.6 microunits) was incubated with 7.9 mum PA-OGA of DP 12. Each reaction mixture contained 0.13 mm EDTA derived from the extraction buffer. The activities relative to 5 mm Mn2+ are shown. Galacturonidase Digestion of PGA-GalUAT Products | The enzyme products, PA-OGAs of DP 12 and 13 prepared from the PA-OGA of DP 11 with PGA-GalUAT (Fig. A), were hydrolyzed with the P. axillaris pollen tube soluble fraction that contained galacturonan alpha-1,4-galacturonidase (EC ) activity (Fig. B). The digest products were eluted at the same retention times as those for the PA-OGAs of DP 11, 10, 9, 8, 7, 6, 5, and 4, substantiating the evidence that GalUA was transferred onto the nonreducing end of PA-OGA through alpha-1,4-linkages. Figure 8 | Galacturonidase digestion of products obtained with PGA-GalUAT. Galacturonidase digestion of products obtained with PGA-GalUAT. A, PA-OGA of DPs 12 and 13 were prepared from the reaction with 2.3 microunits of the crude enzyme, and 26.3 mum PA-OGA of DP 11. B, The enzymatic products (PA-OGAs of DP 12 and 13) shown in A were collected and treated with the soluble fraction containing galacturonidase activity from P. axillaris pollen tubes for 1 and 4 h at 28C. Arrows and numbers indicate the PA-OGA elution position and DP of the PA-OGA eluted, respectively. DISCUSSION : The P. axillaris pollen tube enzyme of the microsomal fraction solubilized with a buffer containing 0.5% (v/v) Triton X-100 has similar enzymatic characteristics to those of the solubilized enzyme obtained from a tobacco-cell suspension culture with respect to optimum pH, requirement of Mn2+, and the Km value for a donor substrate. The P. axillaris enzyme, thus, appears to be in the same family as the solubilized tobacco enzyme. As novel information on this enzyme family, the Km values for the acceptor substrates (PA-OGAs) were determined and were found to be on the order of 10-5m. The enzyme exhibited successive GalUA transfer activity using the assay developed in this study . This is the first report, to our knowledge, of successive glycosyltransfer activity of solubilized PGA-GalUAT, although similar successive glycosyltransfer of other polysaccharide synthases has been observed . Our finding of successive transfer is inconsistent with previous reports that the CHAPS-solubilized tobacco cell enzyme transferred only one GalUA residue to OGA . Because galacturonan alpha-1,4-galacturonidase activity is known to occur in pollen and pollen tube , the synthesis and degradation of pectin in the growing pollen tube must constitute a dynamic equilibrium. No degradation products were detected, that is, only synthase activity was observed under the assay conditions used in this study (Fig. A). However, when the concentration of UDP-GalUA in the reaction mixture was reduced to less than 100 mum, not only PGA-GalUAT activity but also galacturonidase activity against PA-OGA were observed (data not shown). This means that the microsomal fraction of the P. axillaris pollen tube contains both PGA-GalUAT and galacturonidase activity. The galacturonidase activity in the microsomal fraction has already been reported elsewhere . The glycosyltransfer reaction observed in this study did not derive from transglycosylation by alpha-1,4-galacturonidase because no glycosyltransfer was observed in the absence of UDP-GalUA and in the presence of 1 or 5 mm GalUAalpha1 --4GalUA or GalUAalpha1 --4GalUAalpha1 --4GalUA (data not shown). In nature, HGA has some modifications, including methylesterification at C-6 of the GalUA residue, but it has been unclear whether this modification is needed for PGA synthesis. The fact that PGA-GalUAT successively transferred GalUA to PA-OGAs, which do not have any such modification, indicates that this enzyme does not necessarily require this particular HGA modification for its activity. A model for the synthesis of a flip-flop polysaccharide has been proposed based on the idea that every GalUA appears to flip approximately 180 in HGA, giving rise to the notion that PGA is constructed of disaccharide repeating units . The formation of a dimer of GalUA before polymerization in this model conflicts with the successive transfer of every GalUA observed in the present study . The solubilized PGA-GalUAT from the pollen tube of P. axillaris transferred a single GalUA to the nonreducing end of PA-OGA in a non-processive fashion . The assay for PGA-GalUAT using fluorogenic acceptor substrates has several advantages over the assay method using a radiolabeled donor substrate. First, the products are easily identified by DEAE HPLC. Second, elongation of oligosaccharides is observed with higher resolution than the method that relies on radiolabeled nucleotide sugars because a fluorescence detector was connected directly with a HPLC. The process of the enzyme reaction can accordingly be followed more quantitatively. Third, activity for both synthesis and degradation of an acceptor substrate can be observed. In several cases, tissues that highly express a polysaccharide synthase also highly express a polysaccharide-degrading enzyme. The activity for substrate hydrolysis must influence the apparent activity for polysaccharide synthase. Fourth, products present in very small amounts can be detected, because PA-OGA offers highly sensitive fluorescence detection. Less than 10 fmol of PA-OGA as the enzyme product was quantified in this study. Fifth, PA-oligosaccharides are chemically stable and do not require a hot laboratory. In addition to these advantages, this method is applicable to assay for other polysaccharide synthases and analysis of the mechanism for elongation of polysaccharides. For example, PA-cello-oligosaccharides can be an acceptor substrate for cellulose synthase. This method is also applicable to assay for the branching enzymes, which transfer the side chain because PA-oligosaccharides can be separated from one another by HPLC with high resolution . Our success in detecting the successive glycosyltransfer activity of PGA-GalUAT in the crude pollen tube enzyme of P. axillaris solubilized with detergents and using the assay method developed in this study opens the way to the purification of the enzyme and the cloning of its gene. MATERIALS AND METHODS : Materials | PGA was purchased from ICN Biomedicals Inc. (Costa Mesa, CA). 2-Aminopyridine supplied by Wako Pure Chemicals (Osaka) was recrystallized from 1-hexane. Dimethylamine-borane complex and a Wakosil-II 5C30 AR column (4.6 x 250 mm) were obtained from Wako, UDP-GalUA from Sigma (St. Louis), and TSK gel DEAE-5PW (7.5 x 75 mm) from Tosoh (Tokyo). PA-OGA Preparation | PGA (500 mg) dissolved in 50 mL of H2O was titrated to pH 4.2 with NaOH. This solution was autoclaved at 121C for 40 min and then adjusted to pH 2.0 with HCl. The supernatant containing OGA was lyophilized. OGA was pyridylaminated by the method of with slight modifications. In brief, after dissolving lyophilized OGA (50 mg) in 500 muL of 20 mm ammonium acetate buffer, pH 4.5, 500 muL of a coupling reagent (prepared by mixing 552 mg of 2-aminopyridine and 200 muL of acetic acid) was added, and the resultant solution was heated at 90C for 60 min. The Schiff base obtained was reduced with 1,750 muL of a reducing reagent (freshly prepared by mixing 200 mg of dimethylamine-borane complex, 50 muL of water, and 80 muL of acetic acid) at 80C for 35 min. Excess reagents were extracted twice with 2 mL of water-saturated phenol:chloroform (1:1, v/v) and twice with 2 mL of chloroform, after which the aqueous phase was concentrated. Each PA-OGA was separated with TSK gel DEAE-5PW using a linear gradient of ammonium acetate buffer, pH 4.8, from 60 to 800 mm over 60 min at a flow rate of 1.0 mL min-1. Reversed-phase HPLC was performed on a Wakosil-II 5C30 AR column with isocratic elution of 0.1% (v/v) trifluoroacetic acid at a flow rate of 1 mL min-1. The fluorescence intensity of PA-OGAs relative to that of the PA-OGA with a DP of 3 was determined by quantifying the PA-GalUA in the hydrolyzate of each PA-OGA after hydrolysis with 4 m trifluoroacetic acid at 130C for 4 h. PA-OGAs were detected by fluorescence (excitation wavelength, 310 nm; emission wavelength, 380 nm). For MALDI-TOF MS, a PA-OGA sample was cocrystallized in a matrix of 2,5-dihydroxybenzoic acid and analyzed with a Voyager-DE RP biospectrometry workstation (PerSeptive Biosystems, Framingham, MA), using delayed extraction technology and operated in the reflector mode. Preparation of Pollen Tubes | Plants of Petunia axillaris (Lam.) Britton, Sterns & Poggenb. subsp. axillaris (Solanaceae) were grown in a greenhouse, and the pollen was separated from the anthers. Dried pollen was used immediately or kept at -70C. Pollen (2 g) was germinated in a culture medium (8 mg pollen mL-1) containing 0.07% (w/v) Ca(NO3)24H2O, 0.02% (w/v) MgSO47H2O, 0.01% (w/v) KNO3, 0.01% (w/v) H3BO3, 25 mm MES-KOH (pH 6.0), 5% (w/v) Suc, and 20% (w/v) PEG4000 according to . Pollen tubes germinated for 7 h at 25C were collected and separated from ungerminated pollen using a steel sieve (180-mum mesh). Preparation of Crude Enzyme from P. axillaris Pollen Tubes | Pollen tubes were ground with a mortar and a pestle under liquid nitrogen and homogenized with 10 mL of a grinding buffer (50 mm HEPES-NaOH, pH 7.3, containing 50% [v/v] glycerol, 25 mm KCl, 0.25 mm MnCl2, and 0.1% [v/v] 2-mercaptoethanol; ) at 4C for 15 min. The homogenate was centrifuged at 8,500g at 4C for 20 min. The supernatant was centrifuged at 103,000g at 4C for 1 h, and the pellet was washed with the grinding buffer and recentrifuged at 103,000g at 4C for 1 h to yield a membrane pellet. The supernatant was used as a soluble fraction containing alpha-1,4-galacturonidase. The pellet was solubilized in 300 muL of a solubilization buffer (50 mm HEPES --NaOH, pH 6.8, containing 25% [v/v] glycerol, 25 mm KCl, 0.25 mm MnCl2, 2 mm EDTA, and 0.5% [v/v] Triton X-100) with a hand-held pellet mixer for 15 min in a 1.5-mL microfuge tube polyallomer (Beckman Coulter, Fullerton, CA). The supernatant obtained by centrifugation in a Beckman Coulter ultracentrifuge rotor TLA 100.3 at 103,000g at 4C for 1 h was used as the crude enzyme. Assay Procedure for PGA-GalUAT | PGA-GalUAT activity was measured in a reaction mixture (total volume, 30 muL) containing the crude enzyme, 10 muL of a reaction buffer (100 mm HEPES-NaOH, pH 7.3, containing 25 mm KCl, 0.4 m Suc, 0.1% [v/v] bovine serum albumin, and 0.5% [v/v] Triton X-100), 5 mm MnCl2, 1 mm UDP-GalUA, and 1 to 30 mum PA-OGA at 28C for 30 min unless otherwise specified. The reaction was terminated by heating at 100C for 4 min. The reaction mixture was centrifuged, and the supernatant was analyzed by DEAE anion-exchange HPLC with a linear gradient of ammonium acetate buffer, pH 4.8 (60 mm for 3 min, to 130 mm in 2 min, to 280 mm in 5 min, and then to 470 mm in 40 min), at a flow rate of 1.0 mL min-1. The products were detected by fluorescence as described above. One unit of enzyme activity was defined as the amount of enzyme that transferred 1 mumol of GalUA from UDP-GalUA to PA-OGA with a DP of 14 (5 mum) per minute under the conditions described above. The apparent Km and Vmax values of PGA-GalUAT as the crude enzyme for the PA-OGAs of DP 7, DP 11, and DP 14 were determined by assay with various concentrations of the PA-OGA of DP 7 (8 --80.0 mum), DP 11 (3 --59.5 mum), and DP 14 (1.8 --22 mum) in the presence of 1 mm UDP-GalUA and 5.1 (for DP 7), 1.3 (for DP 11), and 0.2 (for DP 14) microunits of the crude enzyme, respectively. The apparent Km and Vmax values for UDP-GalUA were determined with various concentrations of UDP-GalUA (10.8 --1100 mum) in the presence of 212 mum PA-OGA of DP 11 and 1.3 microunits of the crude enzyme. The reaction mixture was incubated at 28C for 20 min. The production of the PA-OGA of DP 12 was quantified under conditions in which the production of the PA-OGA of DP 13 was negligible. Protein Assay | Protein was determined using a BCA protein assay reagent kit (Pierce, Rockford, IL) according to the manufacturer's instructions. Backmatter: PMID- 12226517 TI - Developmentally Regulated Dual-Specificity Kinase from Peanut That Is Induced by Abiotic Stresses AB - Tyrosine (Tyr) phosphorylation represents an important biochemical mechanism to regulate many cellular processes. No Tyr kinase has been cloned so far in plants. Dual-specificity kinases are reported in plants and the function of these kinases remains unknown. A 1.7-kb cDNA that encodes serine/threonine/Tyr (STY) kinase was isolated by screening peanut (Arachis hypogaea) expression library using the anti-phospho-Tyr antibody. The histidine-tagged recombinant kinase histidine-6-STY predominantly autophosphorylated on Tyr and phosphorylated the histone primarily on threonine. Genomic DNA gel-blot analysis revealed that STY kinase is a member of a small multigene family. The transcript of STY kinase is accumulated in the mid-maturation stage of seed development, suggesting a role in the signaling of storage of seed reserves. The STY kinase mRNA expression, as well as kinase activity, markedly increased in response to cold and salt treatments; however, no change in the protein level was observed, suggesting a posttranslational activation mechanism. The activation of the STY kinase is detected after 12 to 48 h of cold and salt treatments, which indicates that the kinase may not participate in the initial response to abiotic stresses, but may play a possible role in the adaptive process to adverse conditions. The transcript levels and kinase activity were unaltered with abscisic acid treatment, suggesting an abscisic acid-independent cold and salt signaling pathway. Here, we report the first identification of a non-MAP kinase cascade dual-specificity kinase involved in abiotic stress and seed development. Keywords: Introduction : Phosphorylation of Ser, Thr, and Tyr residues on target proteins by protein kinases represents an important biochemical mechanism to regulate enzyme activities and many other cellular processes . Although an increasing number of Ser/Thr kinases have been cloned, the existence of protein Tyr kinases is yet to be demonstrated in plants. Nevertheless, protein Tyr phosphorylation has been reported in wheat (Triticum aestivum; ), coconut (Cocos nucifera; ), and pea (Pisum sativum; ). Phosphorylation of cyclin-dependent protein kinases at Tyr-15 is mediated by Wee1 and related kinases . Actin in Mimosa pudica is heavily phosphorylated at the Tyr and the extent of phosphorylation correlates with the bending of petiole . Despite the apparent absence of classical Tyr kinases, dual-specificity kinases that carry out Tyr phosphorylation are reported in plants. Dual-specificity kinases ATN1, APK1, and ADK1 from Arabidopsis and GmPK6 from soybean (Glycine max; ) have been isolated, but the function of these kinases remains unknown. Plant growth is greatly affected by abiotic factors like low temperature and salinity. There is increasing evidence that genes involved in signal transduction are shown to be up-regulated in response to cold acclimation . Furthermore, several of these genes are also regulated by salt stress . Mitogen-activated protein (MAP) kinases and their upstream kinases have been shown to be activated under various stresses . In addition, SNF1 kinases and calcium-dependent protein kinases are shown to be induced by abiotic stresses. Here, we report the identification of a new class of dual-specificity kinase, Ser/Thr/Tyr (STY) kinase, which is distinct from the above-mentioned kinases, involved in abiotic stress response and seed development. STY kinase is a structural mosaic of Ser/Thr and Tyr kinases that autophosphorylates predominantly on Tyr, and phosphorylated the histone primarily on Thr. The isolated kinase is regulated by cold and salt stresses by posttranslational activation mechanism. RESULTS : In Silico Analysis of the STY Kinase cDNA Clone | Sequence analysis of the largest clone obtained after screening of peanut (Arachis hypogaea) expression library with anti-phospho-Tyr antibodies revealed a 1,740-bp gene that encodes a protein with an open reading frame (ORF) of 411 amino acids with a theoretical pI of 8.54 and molecular mass of 46.1 kD (GenBank accession no. ). The ORF is flanked by a 200-bp 5'-untranslated region (UTR) and a 309-bp 3'-UTR with stop codons in all three reading frames upstream of the initiating ATG, and a poly(A+) tail at the end of a 3'-UTR, indicating that the isolated clone represents full-length cDNA. The STY kinase polypeptide contains all 11 subdomains for protein kinases and the kinase catalytic domain spans from 133 to 378 amino acids. In subdomain VIB, STY kinase contains a KPM sequence motif that shows homology to Ser/Thr kinases. In subdomain VIII, RWM sequence motif is more closely related to Tyr kinases. In subdomain XI, the dipeptide CW characteristic of all the Tyr kinases is also found in the STY kinase protein sequence. A BLAST search revealed that the STY kinase showed 75% homology with a putative protein kinase from Arabidopsis. Four other protein kinases, namely Arabidopsis ATN1 (43%), soybean GmPK6 (42%), Dictyostelium discoideum sp1A (37%), and human MLK1 (35%) shared identities greater than 30% with STY kinase over the catalytic domain. Multiple sequence alignment of STY kinase protein sequence with the related sequences is depicted in Figure . The most striking common motifs are located in the core of the catalytic domain and are as follows: the conserved sequences of subdomains V (TEY), VIB (HRDL), VII (DFGVAGI), and VIII (TYRWMAPE). In analogy with the MAP kinases, STY kinase also has the TEY motif in subdomain V. MAP kinases have the same motif between VII and VIII domains. The non-kinase domains do not show homology to any of the sequences in the database. The sequence similarity of the STY kinase revealed homology to Src (Schmidt-Ruppin A-2) Tyr kinases from chicken (Gallus gallus) and human . Five potential glycosylation sites are found throughout the protein sequence (Asn-X-S/T). The hydropathy plot of STY kinase revealed a single transmembrane domain (316 --332). The transmembrane domain has a Leu-/Ile-rich repeat (LRR). There is another LRR between 253 and 267 amino acids. Figure A is the molecular phylogenic estimation of the most parsimonious tree, which revealed that the STY kinase is closely related to a putative protein kinase from Arabidopsis and Dictyostelium discoideum non-receptor Tyr kinase (sp1A). Figure 1 | Sequence alignment of peanut STY kinase catalytic domain sequence and those of the five most closely related sequences in the databases, namely (putative protein kinase) and ATN1 from Arabidopsis, GmPK6 from soybean, DPYK1 from Dictyostelium discoideum, and MLK1 from human (Homo sapiens). Sequence alignment of peanut STY kinase catalytic domain sequence and those of the five most closely related sequences in the databases, namely (putative protein kinase) and ATN1 from Arabidopsis, GmPK6 from soybean, DPYK1 from Dictyostelium discoideum, and MLK1 from human (Homo sapiens). Black boxes indicate positions at which the residues are identical, and gray boxes highlight residues that are similar. The positions of the catalytic subdomains are indicated with roman numerals. Databanks and database accession numbers are shown within brackets. Sequences were aligned using ClustalW and GeneDoc. Figure 2 | A, Phylogenetic tree based on an alignment of peanut STY kinase (STY, accession no. A, Phylogenetic tree based on an alignment of peanut STY kinase (STY, accession no. ) with putative protein kinase from Arabidopsis (accession no. ) and 13 related proteins. GenBank accession numbers are: ATN 1 (Arabidopsis), ; sp1A (D. discoideum), ; EDR 1 (barley [Hordeum vulgare]), ; MAP3K (Arabidopsis), AA7459; TCTR2 (Lycopersicum esculentum), ; CTR1 (Arabidopsis), ; CTR1 (L. esculentum), ; K (Fagus sylvatica), ; PK6 (soybean), ; MLK1 (human), ; MLK2 (human), ; DLK (Rattus norvegicus), JC53399; and DLK (human), NP 00472. Protein sequences were aligned using the ClustalW. Distance trees were calculated using the neighbor-joining method. The lengths of the branches are proportional to the degree of divergence and thus correspond to the statistical significance of the phylogeny between the protein sequences. B, Southern-blot analysis STY kinase. Genomic DNA (20 mug) from immature peanut seed was digested with EcoRI (lane 1), SacI (lane 2), and BamHI (lane 3), separated on 0.8% (w/v) agarose gel, and probed with full-length STY kinase cDNA at high-stringency conditions (hybridization at 65C, and washes at 65C and 0.1x SSC). Molecular organization of the STY kinase gene in peanut was determined by Southern analysis (Fig. B). Peanut genomic DNA was digested with various restriction enzymes and probed with a 1.7-kb full-length STY kinase gene. Digestion of the STY kinase gene with BamHI that cleaves within the probe region resulted in three hybridization signals. The digestions with EcoRI and SacI that do not cut within the probe also produced three hybridization signals. However, additional weak signals were observed, suggesting the presence of a homologous gene(s) in the genome of peanut. A simple hybridization pattern appears to be consistent with the STY kinase being a small multigene family. Dual-Specificity Kinase Activity of STY Kinase | The cDNA corresponding to the ORF was subcloned into pRSET C and expressed in Escherichia coli as an N-terminal fusion protein of oligo-His. The His-6-STY was purified by a nickel affinity column and shown to have a molecular mass of 52 kD, which was 6 kD more than that of theoretical molecular mass. The higher molecular mass of the protein could be due to posttranslational modifications or aberrant mobility of His tag proteins. To determine autophosphorylation kinetics, the STY kinase was incubated with [gamma-32P]ATP in an in vitro kinase assay for various time intervals, and maximum phosphorylation was observed in 20 min (Fig. A). The autophosphorylation activity remained the same even at 60 min (data not shown). The stoichiometry of autophosphorylation was calculated to be 3 +- 0.38 mol of phosphate incorporated per mol of STY kinase, which was obtained from 0.1 mm ATP concentration under standard assay conditions. The reaction was linear with the amount of protein (data not shown). The reaction was dependent on Mg2+; however, no phosphorylation was observed either with Ca2+ or Mn2+ (Fig. B). Phosphoamino acid analysis of autophosphorylated protein indicated that the STY kinase predominantly phosphorylated Tyr (>80%) but less on phospho-Ser and phospho-Thr (Fig. C). This was further confirmed by performing immunoblotting with monoclonal antibodies for all three phosphoamino acids (Fig. D). When histone H1 (type III-S) was used as an exogenous substrate, we detected phosphorylation predominantly in one of its degradation product (15 kD) in addition to the protein (Fig. E). However, recombinant protein did not phosphorylate substrates such as enolase, casein, and aprotinin, suggesting that the STY kinase is not a promiscuous kinase (data not shown). Phosphoamino acid analysis of histone phosphorylation by the STY kinase indicated that the protein phosphorylated the substrate maximally at Thr and less at Tyr. However, phospho-Ser was not detected in the autoradiogram (Fig. F). Figure 3 | A, Time course of autophosphorylation of STY kinase. A, Time course of autophosphorylation of STY kinase. B, Effect of divalent cations (10 mm) on the autophosphorylation of STY kinase. C, Phosphoamino acid analysis of autophosphorylated STY kinase. Recombinant STY kinase was autophosphorylated, resolved on 12% (w/v) SDS-PAGE, and transferred onto a polyvinylidene difluoride membrane. The reaction product was hydrolyzed and separated by silica thin-layer chromatography (TLC) as described in "Materials and Methods." The positions of the origin (ori), phospho-Ser (PS), phospho-Thr (PT), and phospho-Tyr (PY) are indicated along the right side of the TLC. Increasing amounts of hydrolyzed phosphoamino acids were spotted in lanes 1 through 3. D, Autophosphorylated protein was electrophoretically transferred onto a nitrocellulose membrane, and was reacted with the anti-phospho-Ser, anti-phospho-Thr, and anti-phospho-Tyr monoclonal antibodies. E, Five (lane 1) and 10 (lane 2) mug of histone III S was subjected to phosphorylation by STY kinase (750 ng) and the amount of phosphorylated histone was visualized by autoradiography. Molecular mass standards are indicated in the left in kilodaltons. F, Phosphoamino acid analysis of histone-III S phosphorylation by STY kinase. Expression of STY Kinase in Peanut | To study the specificity of the antibodies raised against recombinant protein, the antibodies were affinity purified and used for western-blot analysis. His-6-STY was found to cross-react with the affinity-purified immune serum but not with pre-immune serum and immune serum that had been pre-incubated with His-6-STY (Fig. , A and B). As shown in the Figure C, the monospecific immune serum detected a major protein band of 52 kD from the total protein extracts of immature peanut. The protein was not detected with pre-immune serum and blocked immune serum (Fig. C). As a consequence, immunoblotting of microsomal and soluble proteins of immature peanut with STY kinase specific antibodies detected a 52-kD protein in the cytosol (Fig. D). Immunoprecipitation and immune complex kinase assay of the soluble and membrane proteins of peanut cotyledons resulted in a 52-kD protein present in the cytosol but not in the membranes (Fig. E). The higher molecular mass of the protein could be due to posttranslational modifications. Immune complex kinase assay of the soluble and membrane proteins of peanut with pre-immune serum did not result in any phosphorylated proteins. In addition, immune complex kinase assay of the boiled peanut protein with affinity-purified anti-STY kinase antibodies did not result in labeled proteins (data not shown). Immunoprecipitation of the peanut cytosol followed by immune complex kinase assay with histone revealed that the STY kinase is present as an active kinase in vivo (Fig. F). Immunohistochemical analysis indicated that STY kinase is localized in the cytoplasm of seed, corroborating immunoblotting and immunoprecipitation data. The STY kinase antibodies did not immunodecorate the oil bodies of the seed (Fig. , G and H). Figure 4 | Specificity of the antibodies and Intracellular localization of STY kinase. Specificity of the antibodies and Intracellular localization of STY kinase. A, Recombinant His-6-STY was purified by nickel affinity column, run on 12% (w/v) SDS-PAGE, and stained with Coomassie Brilliant Blue R-250. Molecular mass standards are indicated on the left. B, Each lane of the 12% (w/v) SDS-polyacrylamide gel contains 200 ng of the recombinant His-6-STY protein. The primary antibodies used were affinity-purified immune serum raised against His-6-STY protein (lane 1), pre-immune serum (lane 2), and blocked immune serum (lane 3). C, Each lane of the 12% (w/v) SDS-PAGE contains 50 mug of total protein extracted from peanut immature seed. The primary antibodies used were the same as described in B. The primary antibodies used were affinity-purified immune serum raised against His-6-STY protein (lane 1), pre-immune serum (lane 2), and blocked immune serum (lane 3). D, Fifty micrograms of peanut membranes (lane 1), cytosol (lane 2), and 500 ng of STY kinase (lane 3) were subjected to immunoblotting using anti-STY kinase antibodies. E, Two-hundred fifty micrograms of CHAPS-solubilized membrane (lanes 1 and 2) and soluble (lanes 3 and 4) fractions were immunoprecipitated using pre-immune serum (lanes 1 and 3) and anti-STY kinase antiserum (lanes 2 and 4) as described in "Materials and Methods" and subjected to phosphorylation. Boiled plant protein is immunoprecipitated using anti-STY kinase antibodies (lane 5). The antibodies were used at 1:100 (v/v) dilution. F, Immunoprecipitation and immune complex assay of STY kinase (lane 1) and histone phosphorylation of immune complex (lane 2) of peanut cytosol. Cytosol (250 mug) was subjected to immunoprecipitation with 1:100 (v/v) dilution of the affinity-purified anti-STY kinase antibodies. G and H, Immunohistochemical localization of STY kinase in peanut seed. Transverse section of seed was incubated with 1:500 (v/v) diluted pre-immune serum (G) or with anti-STY kinase antibodies (H). Tissue sections were incubated with peroxidase-coupled secondary antibodies and developed using 3,3'-diaminobenzidine as a chromogenic substrate. Magnification: x63. O, Oil bodies; V, vacuole. Immunoblot analysis was performed to determine the expression of STY kinase protein levels in different organs of peanut plant. STY kinase protein was detected in all organs of the plant with higher levels in the shoot as compared with other organs studied (Fig. A). The developmental expression of the STY kinase gene was examined by RNA gel-blot analysis (Fig. B). A high level of mRNA expression was detected at 20 to 26 d after flowering (DAF) that was approximately a 5-fold increase as compared with that of 14 DAF. The levels of mRNA expression drastically reduced at 36 DAF. The size of transcript was consistent with the sequence data and confirms that the isolated cDNA was full length. Figure 5 | A, Tissue-specific immunoblot analysis of different organs of peanut probed with antibodies raised against the recombinant STY kinase. A, Tissue-specific immunoblot analysis of different organs of peanut probed with antibodies raised against the recombinant STY kinase. The band at approximately 50 kD in the leaf and shoot lane corresponds to the large subunit of Rubisco, to which either the primary or secondary antibodies bound nonspecifically. Molecular mass standards are indicated on the left in kilodaltons. B, Developmental stage RNA gel-blot analysis. Four different seed developmental stages based on DAF were analyzed. For the rRNA panel, size-separated total RNA (20 mug) was stained with ethidium bromide to show that similar amounts of RNA were loaded per lane. The prominent band is 23 S rRNA. C, Relative mRNA levels were densitometrically estimated, and histograms were constructed with the data from Figure B. Error bars represent the sd of three experiments. Transcriptional and Posttranslational Activation of STY Kinase by Low Temperature and High Salt | To analyze the expression of STY kinase under various stress conditions, northern-blot analyses and histone kinase activities were carried out and found that the mRNA level and histone kinase activity of STY kinase increased when peanut seedlings were under cold or high-salinity stress conditions. The magnitude of transcript accumulation and kinase activity was higher with cold treatment (4C) than with NaCl (250 mm) for 24 h (Fig. , A --D). Immunoprecipitation and immune complex kinase assays with cold- and salt-treated seedlings with pre-immune serum did not result in any labeled protein (Fig. D). We did not observe significant change in the expression of STY kinase when seedlings were subjected to various hormonal treatments (indole acetic acid, benzyl amino purine, gibberillic acid, and ABA; data not shown). There was no change in mRNA levels when peanut seedlings were treated with 100 mum ABA (Fig. E). Upon cold treatment, there was a time-dependent up-regulation of mRNA levels (Fig. , A and B). In 48 h, the level of mRNA expression was very high. Cold-treated seedlings were immunoblotted with anti-STY kinase antibodies and a 52-kD protein was specifically decorated (Fig. C). However, there was no change in the steady-state protein levels with cold-treated seedlings. When aliquots of the same seedlings were immunoprecipitated and subjected to histone kinase assay, time-dependent increase in the histone kinase activity was observed (Fig. D). There was no significant change in the mRNA levels and kinase activity with cold treatment at the early time points such as 1 and 6 h (data not shown). Figure 6 | A, Two-day-old seedlings of peanuts were grown in the dark at 25C, and treated either with 250 mm NaCl or with 4C, and harvested after 24 h. A, Two-day-old seedlings of peanuts were grown in the dark at 25C, and treated either with 250 mm NaCl or with 4C, and harvested after 24 h. Untreated samples were used as controls. Total RNAs were isolated from the peanut seedlings and subjected to northern analysis using STY kinase full-length cDNA as the probe. B, Relative mRNA levels were densitometrically estimated, and histograms were constructed with the data from Figure A. Error bars represent the sd of four experiments. C, Two-hundred fifty micrograms of the protein from unstressed, cold-, and salt-stressed seedlings were immunoprecipitated with pre-immune serum (lanes 1 --3) and with affinity-purified anti-STY kinase antiserum (lanes 4 --6). Histone kinase assay was performed with unstressed (lanes 1 and 4), cold- (lanes 2 and 5), and salt- (lanes 3 and 6) stressed seedlings. D, Two-day-old seedlings of peanuts were grown in the dark at 25C, and were treated with 100 mum abscisic acid (ABA) and harvested after the indicated time. Total RNAs were isolated from the peanut seedlings and subjected to northern analysis using STY kinase full-length cDNA as a probe. For the rRNA panel, size-separated total RNA (20 mug) was stained with ethidium bromide to show that similar amounts of RNA were loaded per lane. The prominent bands are rRNAs. Figure 7 | Two-day-old seedlings of peanuts were grown in the dark at 25C, and were transferred to 4C and harvested after the indicated time. Two-day-old seedlings of peanuts were grown in the dark at 25C, and were transferred to 4C and harvested after the indicated time. A, Total RNAs were isolated from the peanut seedlings and subjected to northern analysis using STY kinase full-length cDNA as a probe. For the rRNA panel, size-separated total RNA (20 mug) was stained with ethidium bromide to show that similar amounts of RNA were loaded per lane. The prominent bands are rRNAs. B, Relative mRNA levels were densitometrically estimated, and histograms were constructed with the data from Figure A. Error bars represent the sd of three independent experiments. C, Protein extracts (25 mug) from untreated control (0 min) or seedlings treated with cold (4C) were subjected to immunoblot analysis with STY kinase antibodies at the indicated time periods. D, Protein extracts (250 mug) from untreated control (0 min) or seedlings treated with cold (4C) were immunoprecipitated with 1:100 (v/v) dilution of anti-STY kinase antibodies. Kinase activity of the immune complex was subsequently determined using histone III-S as an exogenous substrate. When peanut seedlings were subjected to 250 mm NaCl, the mRNA levels increased proportionately with time and reached maximum at 48 h (Fig. , A and B). Untreated seedlings and seedlings treated at the early time point (6 h) showed a basal mRNA expression and kinase activity. Salt-treated seedlings were subjected to immunoblot analysis and a 52-kD protein is specifically decorated; however, there was no change in the steady-state protein levels (Fig. C). There was a time-dependent increase in the histone kinase activity with salt-treated seedlings (Fig. D). The other abiotic stresses such as high temperature and dehydration (20% [w/v] polyethylene glycol) treatments did not have much effect on STY kinase mRNA expression and kinase activity (data not shown). These results indicate that the STY kinase is regulated by cold and salt treatments in an ABA-independent manner by posttranslational activation mechanism. Figure 8 | Two-day-old seedlings of peanuts were grown in the dark at 25C, and were treated with 250 mm NaCl at 25C and harvested after the indicated time. Two-day-old seedlings of peanuts were grown in the dark at 25C, and were treated with 250 mm NaCl at 25C and harvested after the indicated time. A, Time course of mRNA expression of salt-treated seedlings. B, Densitometry scan of relative mRNA levels obtained from Figure A. Error bars represent the sd of two independent experiments. C, Immunoblot of salt-treated seedlings with STY kinase antibodies at the indicated time periods. D, Protein extracts (250 mug) from untreated control (0 min) or seedlings treated with cold (4C) were immunoprecipitated with 1:100 (v/v) dilution of anti-STY kinase antibodies. Kinase activity of the immune complex was subsequently determined using histone III-S as an exogenous substrate. DISCUSSION : This study describes molecular characterization of the developmentally regulated and stress-related dual-specificity kinase from peanut. Protein Tyr kinases play a vital role in the regulation of cell growth, differentiation, and development in animal systems . On the other hand, no Tyr kinase has been cloned from plants. The translation product of the isolated cDNA shows homology to Tyr kinases in the database. The autophosphorylated kinase is recognized by anti-p-Tyr antibody, indicating that STY kinase autophosphorylates on the Tyr. Additional evidence for Tyr phosphorylation was from phosphoamino acid analysis. The methods of screening expression libraries with anti-p-Tyr antibodies were successfully employed in mammalian systems , yeast (Saccharomyces cerevisiae; ), and D. discoideum . The significant feature of the STY kinase cDNA is that the gene product has all the sequence attributes of protein Ser/Thr kinases, even though it was isolated by anti-p-Tyr antibody. Earlier attempts to clone plant cDNAs for classical Tyr kinases using anti-p-Tyr antibodies resulted in dual-specificity protein kinases . This raises an intriguing possibility that the dual-specificity kinases alone are responsible for Tyr phosphorylation in plants. Arabidopsis ATN1 and soybean GmPK6 are the dual-specificity kinases that are not studied in detail. Arabidopsis ADK1 has been overexpressed and it is shown to autophosphorylate maximally on Thr than Tyr. Peanut STY kinase is the first plant protein kinase shown to autophosphorylate predominantly on Tyr. We have also demonstrated the dual specificity of the STY kinase, even in substrate phosphorylation with histone. Except for the MAP kinases (MEK1), the physiological significance of other dual-specificity kinases is yet to be determined. The STY kinase sequence has striking features of potential functional significance. There are two LRRs in the STY kinase protein sequence. LRRs are shown to be involved in signal transduction during developmental and environmental cues . The presence of LRRs in the transmembrane domain suggests a possible role of STY kinase in plant defense. The MAP kinase cascade is involved in stress signaling in plants. Besides MAP kinases, other stress-activated plant kinases have been reported wherein the phosphorylation of Ser and Thr, but not Tyr, is responsible for activation of these kinases. These belong to calcium-dependent protein kinases and the SNF1 family of kinases . Peanut STY is a novel kinase that does not belong to any of the category of kinases reported so far and thus forms a new division in the plant kinase family. STY kinase is the first report of a non-MAP kinase cascade dual-specificity kinase involved in abiotic stresses. The activation of the STY kinase is detected after 12 to 48 h of cold and salt treatments, which indicates that the kinase may not participate in the initial response to abiotic stresses, but suggesting a possible role in the adaptive process to adverse conditions. STY kinase is not only induced by abiotic stresses but also responds to developmental cues. The transcript of STY kinase is accumulated in the mid-cotyledonary stage of seed development, where the metabolism of proteins and lipids is highly active. It is possible that peanut STY kinase could also be involved in the signal transduction mechanisms related to storage of metabolites. The regulation of cold-inducible genes (RC12A, RC12B, COR6.6, COR15, COR78, DREB1, RD29A, and KIN1) during Arabidopsis development has been studied in detail, and the results indicated that the expression of these genes is regulated in different tissues both in stressed and unstressed conditions . Low-temperature-responsive genes (BLT101, BLT4.9, and BLT14) have been reported in barley, and these genes are also regulated during development . It is possible that many genes regulated during plant development are also regulated by environmental stresses. It has been reported earlier that some of the salt-inducible genes are also induced by cold stress . Cold and salt stresses increased STY kinase mRNA levels. The activation of STY kinase preceded the increase in mRNA levels, suggesting that the STY kinase gene might be a target of the kinase pathway. This pattern is reminiscent of positive feedback loop that occurs in the activation of several protein kinase cascades. However, an increase of STY kinase transcripts was not correlated with an increase in the protein levels. It could be possible that the transcripts accumulate but are not translated. Alternatively, the translation of the STY kinase transcripts and steady-state protein levels of STY kinase may stay constant. A similar observation was made for alfalfa (Medicago sativa) MAP kinase (MMK4; ). In addition, the transcripts of glycogen synthase kinase-3 were accumulated after wounding in alfalfa and the protein levels remained constant; however, there was an increase in kinase activity . These results indicated that protein amounts were tightly regulated by complex intracellular mechanisms in some protein kinases involved in stress response and adaptation. Stress adaptation in plants is mediated either by ABA-dependent or -independent pathways . ABA has an important role in mediating responses to environmental stresses, including cold and salt stresses. Exogenously applied ABA is also able to induce expression of several cold and salt stress genes . When peanut seedlings were treated with ABA, no changes in the mRNA levels or kinase activity were observed. These experiments indicate that the cold- and salt-induced activation of STY kinase is independent of ABA. The expression of several cold- or salt-induced genes is independent of ABA, indicating the presence of ABA-dependent and -independent pathways for low-temperature and salt responses. Because peanut is a chilling-tolerant species, identification of the STY kinase as a cold stress-responsive gene may enable to genetically engineer the chilling-sensitive plants for chilling tolerance. Also, involvement of STY kinase in seed development offers opportunities to manipulate seed development and maturation. The identification of STY kinase as a stress-responsive and developmentally regulated dual-specificity kinase may provide new insights in unraveling plant signal transduction pathways. MATERIALS AND METHODS : Plant Material and Treatments | Field-grown developing peanut (Arachis hypogaea) cotyledons were harvested at various stages after flowering and used either fresh or frozen at -80C. Toward stress treatments, peanut seeds were surface sterilized and grown in water-saturated filter paper (Whatman, Clifton, NJ) in petri dishes for 2 d. Cold (4C) and salt (250 mm NaCl) treatments were given at various time intervals. Towards hormonal treatment of seedlings, 100 mum indole acetic acid, benzyl amino purine, gibberillic acid, and ABA were used. Construction and Screening of cDNA Library | A seed-specific cDNA library of peanut was constructed in lambda-ZAP II (Stratagene, La Jolla, CA). To produce anti-phospho-Tyr antibodies, rabbit polyclonal antiserum was raised against a polymer of phospho-Tyr, Ala, Gly, and keyhole limpet hemocyanin and affinity purified as described . The expression library was screened using anti-phospho-Tyr antibodies. A total of 2 x 106 plaques were screened. Positive plaques were purified by three additional screenings with the polyclonal antibodies and parallel with a monoclonal anti-p-Tyr antibody (Sigma, St. Louis). Sequencing and Sequence Analysis | Plasmid DNA was isolated from the positive clones by alkaline lysis , and was sequenced on both strands using the Biotech Taq cycle sequencing kit (Amersham-Pharmacia Biotech, Uppsala) on an automated sequencer (Applied Biosystems 377, PE-Applied Biosystems, Foster City, CA). Sequence analysis was performed using BLAST search . Multiple Sequence alignment obtained using the ClustalW were displayed using the program GENEDOC . Molecular phylogenies were computed by PHYLIP (Phylogeny Inference Package, version 3.52, Department of Genetics, University of Washington, Seattle) and tree files generated by the above methods were displayed and printed using TREEVIEW. Genomic Southern and Northern Hybridization Analyses | Towards Southern analysis, high-Mr genomic DNA was isolated from peanut cotyledons . Twenty micrograms of DNA was digested with EcoRI, SacI, and BamHI, fractionated on 0.8% (w/v) agarose gel, and transferred onto a Hybond N+ nylon membrane . For northern analysis, total RNA was extracted from cotyledons at four different stages of development and stress-induced seedlings . Aliquots of total RNA (20 mug) were run on 1% (w/v) agarose gels containing 19% (v/v) formaldehyde, 40 mm MOPS, 10 mm sodium acetate, 1 mm EDTA, and transferred to Hybond N+ membrane (Amersham-Pharmacia Biotech). Southern and northern hybridizations were carried out in aqueous hybridization buffer (phosphate buffer [pH 7.2], 7% [w/v] SDS, and 20 mm EDTA) with a random-primed 32P-labeled STY kinase cDNA as a probe at 65C. Blots were washed in 0.2x SSC and 0.1% (w/v) SDS at 65C. Equal loading of RNA was confirmed by visual comparison of ethidium bromide-stained 18 S and 23 S rRNAs of the agarose gel. Expression and Purification of His-6-STY Kinase Fusion Protein | The cDNA spanning for coding region of STY kinase (lacking 11 amino acids from the N terminus) was subcloned into the His-tagged fusion protein expression vector, pRSET C, at bglII and kpn1 cloning sites. The resultant construct was expressed in Escherichia coli BL-21 (pLysS). The fusion protein was induced with 0.4 mm isopropyl-1-thio-beta-d-galactopyranoside for 4 h. The recombinant protein was induced in large scale and purified by nickel-nitrilotriacetic acid agarose chromatography (Qiagen USA, Valencia, CA). Purified fractions containing the eluted protein were analyzed by 12% (w/v) SDS-PAGE followed by Coomassie Blue staining . In Vitro Kinase Assays | Autophosphorylation assay was performed by incubating 0.5 mug of purified recombinant STY kinase in reaction buffer (50 mm Tris-HCl [pH 7.5] and 10 mm MgCl2) in the presence of 25 mum [gamma-32P]ATP (3,000 dpm pmol-1) at 30C for 20 min. The reaction was stopped by the addition of SDS-PAGE loading buffer, and kinase activity was detected by autoradiography after 12% (w/v) SDS-PAGE. Substrate phosphorylation assay mixture consisted of 750 ng of STY kinase, 5 mug of histone III-S (Sigma), 50 mm Tris-HCl (pH 7.5), and 10 mm MgCl2, and the incubation was carried out at 30C for 20 min. Phosphorylated products were separated by 12% (w/v) SDS-PAGE and the labeled proteins were detected by autoradiography. Phosphoamino Acid Analysis | The purified STY kinase was labeled in vitro with [gamma-32P]ATP as described above and electroblotted onto a polyvinylidine difluoride membrane. After autoradiography, radioactive band of interest was excised and hydrolyzed in 200 muL of 6 m HCl for 2 h at 110C. The hydrolyzate was dried in a Speed-Vac concentrator and resuspended in 20 muL of water containing 1 mg mL-1 of each of the phosphoamino acid markers such as phospho-Ser, phospho-Thr, and phospho-Tyr (Sigma). Two microliters of the hydrolyzate was analyzed by ascending silica TLC (Merck, Rahway, NJ) using a solvent system containing a mixture of ethanol and ammonia (3.5:1.6 [v/v]; ). The position of phosphoamino acid markers was detected by ninhydrin staining of the silica-TLC plate (0.25% [w/v] ninhydrin in acetone). The plate was then exposed for autoradiography to locate the position of the 32P-labeled amino acids. Immunoblotting | Two New Zealand white male rabbits were subcutaneously immunized at 2-week intervals with 0.5 to 1.0 mg of fusion protein. The polyclonal antiserum was purified by protein A-Sepharose column chromatography. Subcellular fractionation of peanut cotyledon was performed and protein concentrations were determined by the Bradford method using bovine serum albumin as the standard. For immunodetection of STY kinase proteins, the soluble and membrane fractions were isolated from immature peanuts and separated on 12% (w/v) SDS-polyacrylamide gels. The proteins were transferred onto a nitrocellulose membrane and the blots were then probed with anti-STY kinase antibodies (1:2,000 dilution [v/v]) for 60 min. Blots were washed in phosphate-buffered saline with 0.05% Tween-20, and incubated for 60 min with peroxidase-coupled secondary antibodies. Proteins were detected by using 3,3'-diaminobenzidine as a chromogenic substrate. Immunoprecipitation and Immune Complex Kinase Assay | The soluble and membrane proteins (200 mug) from immature peanuts were precleared with 25 muL of protein A-agarose (Bangalore Genei, Bangalore, India) beads for 1 h. The agarose beads were removed by centrifugation and 2 mug of affinity-purified antibody was added to the extracts and incubated with agitation for 6 h at 4C. After brief centrifugation, the precipitate was washed three times with buffer (250 mm Tris-HCl [pH 7.5] and 10 mm MgCl2). The immunoprecipitated proteins were phosphorylated as described before. Immunohistochemistry and Microscopy | Peanut tissues were fixed in formaldehyde:acetic acid:70% (v/v) ethanol (5:5:90 [v/v]), dehydrated in ethanol followed by infiltration with paraffin in n-butanol, and microtome sectioned (0.5 mum). The sections were mounted on glass slides, deparaffinized with xylene, rehydrated, and incubated with 0.6% (v/v) hydrogen peroxide in methanol. The sections were then incubated in a blocking solution containing 5% (v/v) goat serum and 1% (w/v) bovine serum albumin in phosphate-buffered saline for 60 min at room temperature. The slides were placed in a humidified chamber and sections were covered with 2 mug mL-1 of affinity-purified anti-STY kinase antibodies diluted in phosphate-buffered saline with 1% (w/v) bovine serum albumin for 1 h. Control for specificity included 2 mug mL-1 rabbit pre-immune serum in place of primary antibodies. The sections were then incubated with 0.05% (w/v) 3,3'-diaminobenzidine in 0.1% (v/v) hydrogen peroxide to produce a brown reaction product. Images were captured using a Zeiss-Axioscope microscope (Zeiss, Jena, Germany). GenBank Accession Number | The GenBank accession number for the nucleotide sequence reported in the article is . Backmatter: PMID- 12226518 TI - N-Acylethanolamines Are Metabolized by Lipoxygenase and Amidohydrolase in Competing Pathways during Cottonseed Imbibition AB - Saturated and unsaturated N-acylethanolamines (NAEs) occur in desiccated seeds primarily as 16C and 18C species with N-palmitoylethanolamine and N-linoleoylethanolamine (NAE 18:2) being most abundant. Here, we examined the metabolic fate of NAEs in vitro and in vivo in imbibed cotton (Gossypium hirsutum) seeds. When synthetic [1-14C]N-palmitoylethanolamine was used as a substrate, free fatty acids (FFA) were produced by extracts of imbibed cottonseeds. When synthetic [1-14C]NAE 18:2 was used as a substrate, FFA and an additional lipid product(s) were formed. On the basis of polarity, we presumed that the unidentified lipid was a product of the lipoxygenase (LOX) pathway and that inclusion of the characteristic LOX inhibitors nordihydroguaiaretic acid and eicosatetraynoic acid reduced its formation in vitro and in vivo. The conversion of NAE 18:2 in imbibed cottonseed extracts to 12-oxo-13-hydroxy-N-(9Z)-octadecanoylethanolamine was confirmed by gas chromatography-mass spectrometry, indicating the presence of 13-LOX and 13-allene oxide synthase, which metabolized NAE 18:2. Cell fractionation studies showed that the NAE amidohydrolase, responsible for FFA production, was associated mostly with microsomes, whereas LOX, responsible for NAE 18:2-oxylipin production, was distributed in cytosol-enriched fractions and microsomes. The highest activity toward NAE by amidohydrolase was observed 4 to 8 h after imbibition and by LOX 8 h after imbibition. Our results collectively indicate that two pathways exist for NAE metabolism during seed imbibition: one to hydrolyze NAEs in a manner similar to the inactivation of endocannabinoid mediators in animal systems and the other to form novel NAE-derived oxylipins. The rapid depletion of NAEs by these pathways continues to point to a role for NAE metabolites in seed germination. Keywords: Introduction : In mammalian cells, N-acylethanolamines (NAEs) have varied physiological roles. N-Arachidonylethanolamine (anandamide), a type of NAE in mammalian brain tissue, is an endogenous ligand for the cannabinoid receptor and modulates neurotransmission. Anandamide also can activate vanilloid (capsaicin) receptors and function as an endogenous analgesic , and appears to be involved in neuroprotection . In other animal tissues, NAEs have been implicated in immunomodulation , synchronization of embryo development , and induction of apoptosis . These endogenous bioactive molecules termed "endocannabinoids" are hydrolyzed by fatty acid amidohydrolase (AHase) to terminate their signaling functions. In plants NAEs are present in substantial amounts in desiccated cotton (Gossypium hirsutum) seeds (1.6 mug-1 g fresh weight), and their levels decline after a few hours of imbibition . Individual NAEs were identified predominantly as 16C and 18C species with N-palmitoylethanolamine (NAE 16:0) and N-linoleoylethanolamine (NAE 18:2) being the most abundant. NAEs in both plant and animal cells are derived from N-acylphosphatidylethanolamines (NAPEs), a minor membrane lipid constituent of cellular membranes . NAEs are produced by the action of a phospholipase D (PLD). In plants NAEs were produced in cell suspensions and leaves in minutes after pathogen elicitor perception, raising the possibility that these molecules function in plant defense signaling. Exogenous NAE 14:0 at submicromolar concentrations was sufficient to activate Phe ammonia-lyase (PAL) expression in cell suspensions and leaves of tobacco (Nicotiana tabacum; ). The occurrence of NAEs in seeds with substantially different structural properties than those found in elicitor treated leaves and their rapid depletion during seed imbibition suggest that these lipids may have a role in the regulation of seed germination. Recent studies have shown that NAE 18:2 and NAE 18:3 could be converted into hydroperoxy NAE by purified soybean (Glycine max) lipoxygenase-1 (LOX; , ). The hydroperoxides of NAE 18:2 and NAE 18:3 could subsequently be converted by alfalfa (Medicago sativa) hydroperoxide (HPO) lyase and flax (Linum usitatissimum) seed allene oxide synthase (AOS) into novel oxylipins . N-Arachidonoylethanolamine (anandamide), the mammalian neurotransmitter, could be converted into hydroperoxy NAE by purified 5-LOX from barley (Hordeum vulgare) and tomato (Lycopersicon esculentum; ). As an alternative, NAE AHase, also designated fatty acid amide hydrolase (FAAH; ), acts upon NAE to produce free fatty acid (FFA) and ethanolamine, and in mammalian cells this pathway is responsible for inactivation of endocannabinoid lipid mediators . Thus, there are two possible enzymatic pathways (LOX and NAE AHase) that might be responsible for the observed decline in NAEs during seed imbibition. To begin to understand the role of NAEs in seeds, we investigated the metabolic fate of NAEs in cottonseeds upon imbibition, germination, and during postgerminative growth, a period previously noted to be active in NAPE/NAE metabolism . Our results indicate that there are two pathways capable of metabolizing NAEs in seeds: a LOX-mediated pathway selective for unsaturated NAEs (e.g. NAE 18:2) and a NAE AHase activity, which uses both unsaturated and saturated NAEs. Both enzymatic pathways were most active in imbibed seeds consistent with depletion of NAEs in vivo and at a time period just preceding or coincident with radicle emergence, suggesting NAE metabolism may play a role in the regulation of seed germination. These results will provide the basis for future studies aimed at understanding the functional role of NAE metabolism in seed germination and seedling growth. RESULTS : Identification of NAE Metabolites and Subcellular Distribution of the Enzymes | Several types of NAEs are prevalent in the desiccated seeds of plants . For example, the total NAE content in desiccated cottonseeds was approximately 1,600 ng g-1 fresh weight, of which NAE 18:2 was approximately 940 ng g-1 fresh weight and NAE 16:0 was about 380 ng g-1 fresh weight. The levels of these NAEs declined sharply within few hours of seed imbibition to 200 ng for NAE 18:2 and to about 160 ng for NAE 16:0. Here, these two most abundant NAEs in desiccated seeds, NAE 16:0 and NAE 18:2, were used to evaluate the metabolic fate of NAE in imbibing seeds. When [1-14C]16:0 NAE was used as substrate, there was only one product (comigrating with palmitic acid, Rf = 0.48) formed by extracts of imbibed cottonseeds (Fig. A). When [1-14C]18:2 NAE was used as a substrate there were two apparent products formed (Fig. B): one comigrating with linoleic acid (Rf = 0.46) and the other more polar, near the origin (Rf = 0.12), was tentatively identified as NAE oxylipin. These data suggested that endogenous NAEs in desiccated seeds were metabolized by two pathways: one producing FFAs and other likely producing NAE-derived oxylipins. Figure 1 | Representative chromatograms of radioactive lipids separated on thin-layer chromatography (TLC) plates. Representative chromatograms of radioactive lipids separated on thin-layer chromatography (TLC) plates. For the assay, 100 mum [14C]NAE with 20,000 dpm in 50 mm MES buffer (pH 6.5) was used. Reactions were initiated by adding 400 muL of microsomes and were incubated at 30C for 1 h with shaking. Lipids are extracted as described in "Materials and Methods," and distribution of radioactivity on TLC plates was evaluated by radiometric scanning (Bioscan system 200 image scanner). When synthetic NAE 16:0 was used as a substrate, FFA was produced (A) and when synthetic NAE 18:2 was used as a substrate, FFA and an additional lipid product(s) were formed (B). The subcellular distribution of NAE-LOX (NAE oxylipin formation) and NAE AHase (FFA formation) differed in cell fractions prepared from 4-h imbibed cottonseeds (Tables and ). The presumed NAE-oxylipin formation was distributed both in membrane and cytosol-enriched fractions, whereas the enzyme responsible for FFA formation was localized almost exclusively to microsomes (Tables and ). AHase activity toward NAE 18:2 was higher than AHase activity toward NAE 16:0, whereas formation of NAE oxylipin from NAE 18:2 was considerably higher than the corresponding AHase activity. These relative activities toward the different NAEs are consistent with the more rapid consumption of NAE 18:2 in vivo during seed imbibition. Table I | Subcellular distribution of NAE-LOX (NAE oxylipin formation) and NAE amidohydrolase (FFA formation) activities in 4-h imbibed cottonseed cell fractions utilizing NAE 16:0 as substrate Table II | Subcellular distribution of NAE-LOX (NAE oxylipin formation) and NAE amidohydrolase (FFA formation) activities in 4-h imbibed cottonseed cell fractions utilizing NAE 18:2 as substrate To test whether the oxylipins were formed by the LOX pathway, the influence of two widely used LOX inhibitors on their formation was determined . Both 5,8,11,14-eicosatetraynoic acid (ETYA) and nordihydroguaiaretic acid (NDGA) reduced NAE-oxylipin formation in a concentration-dependent manner. NDGA appeared to be a more potent inhibitor of NAE 18:2-LOX than ETYA, particularly at higher concentrations. As an alternative, NAE 18:2-dependent lipid peroxide formation was estimated spectrophotometrically . Consistent with the above results, inclusion of both inhibitors reduced the formation of NAE 18:2 lipid hydroperoxide. The small amount of lipid peroxide detected in the absence of enzyme (control-enzyme) was likely attributable to the spontaneous oxidation of NAE 18:2 during assay reactions, because no lipid peroxide was detected when NAE 18:2 was omitted from reactions (not shown). These data indicate that the polar product in the incubation is formed by the LOX pathway. Figure 2 | The effects of LOX inhibitors on the metabolism of NAE 18:2 in vitro. The effects of LOX inhibitors on the metabolism of NAE 18:2 in vitro. The amount of NAE-oxylipin was determined by incubating (1 h) synthetic NAE 18:2 with a 150,000gmax (60 min) supernatant of imbibed cottonseeds. Total lipids were extracted from the reaction mixture and were separated by TLC (hexane:ethyl acetate:methanol, 60:40:5; v/v). Identification and quantification of radiolabeled lipids were performed by radiometric scanning. ETYA is a dual-specific inhibitor, affecting both LOX and cyclooxygenases, and is irreversible . NDGA is a classical inhibitor of different LOXs . There was almost complete inhibition of oxylipin production at 400 mum NDGA and 50% inhibition at 400 mum ETYA. The data points are means and sd of three replicates of one experiment. Figure 3 | Determination of LOX activity was performed with a lipid hydroperoxide assay kit. Determination of LOX activity was performed with a lipid hydroperoxide assay kit. Control, 80 nmol of NAE without enzyme showed the natural hydroperoxidation, which is 0.200 nmol h-1; control with enzyme (3.33 mg protein per assay), the total activity was 0.947 nmol h-1; NDGA (100 mum) with enzyme showed the effect of a classical LOX inhibitor. This inhibitor inhibited the effect of natural peroxidation as well which was also observed in other radiolabeled experiments; and ETYA (100 mum) with enzyme showed the expected effect of LOX inhibitor. There was 50% inhibition, which was also observed in radiolabeled NAE 18:2 substrate metabolism experiments. Experiments without synthetic substrate was also carried out to confirm the absence of lipid hydroperoxide in the cell extract itself. Also, an experiment without EDTA was carried out to investigate any possible role of EDTA. All of those experiments were negative. Determination of LOX activity was performed with a commercially available lipid hydroperoxide (LPO) assay kit (catalog no. 705002, Cayman Chemical). For each assay, 80 nmol of NAE 18:2 was used as substrate and incubated with crude extract for 1 h at 30C with shaking (110 rpm). The lipid peroxides that were formed were extracted from the samples into chloroform and quantified by measuring A500 compared with the standard lipid hydroperoxide (13-hydroperoxy octadecadienoic acid). The data points are means and sd of three replicates of one experiment. To elucidate the structure of the polar compound, gas chromatography-mass spectrometry (GC/MS) analysis was performed . Selective ion monitoring at m/z 116 (diagnostic of ethanolamine containing lipids) revealed the presence of two oxygenated NAE 18:2 metabolites in incubations of cottonseed extracts incubated with NAE 18:2, with retention times of 18.22 and 18.29 min, respectively (Fig. A). These compounds were identified as trimethylsilylated, reduced alpha-ketols (diastereomers) 12-oxo-13-hydroxy-N-(9Z)-octadecenoylethanolamine by their electron impact mass spectrum (Fig. C). Predictable fragmentation ions including the molecular ion [M+] (m/z 573) were clearly identifiable, and spectra were comparable with those recorded in previous studies . Only alpha-ketols originating from 13-hydroperoxy NAE (18:2) were detected, indicating that imbibed cottonseeds contained both a 13-LOX and 13-AOS that metabolized NAE 18:2. Of interest also was the identification of the alpha-ketols 12-oxo-13-hydroxy-(9Z)-octadecenoic acid in the total ion chromatogram at retention times 14.85 and 14.94 (Fig. B, diastereomers), which can be explained by the subsequent actions of NAE AHase, LOX, and AOS. These results most importantly demonstrate unequivocally that NAE 18:2 was metabolized by 13-LOX (and 13-AOS) in extracts of imbibed cottonseeds, raising the possibility that a new class of oxylipins may be involved in seed germination. Figure 4 | A, Single-ion chromatogram at m/z 116 (characteristic fragment for the ethanolamine group) of NAE metabolites after incubating with 150,000g supernatant of cottonseeds. A, Single-ion chromatogram at m/z 116 (characteristic fragment for the ethanolamine group) of NAE metabolites after incubating with 150,000g supernatant of cottonseeds. The peaks with RT = 18.22 and 18.29 are alpha-ketols of 13-hydroperoxy NAE (two diastereomers) = 12-oxo-13-hydroxy-N-(9Z)-octadecenoylethanolamine. B, Total ion chromatogram of derivatized lipid products after incubating NAE 18:2 with the 150,000g supernatant of imbibed cottonseeds. Several peaks were identified by MS: 12.28 min, C18:2 linoleic acid; 12.33 min, C18:1 oleic acid; 12.55 min, C18:0 stearic acid; 13.90 min, 13-HPOD (13-hydroperoxy octadecadienoic acid); 14.85 and 14.94 min, alpha-ketols of 13-HPOD; 18.22 and 18.29 min, alpha-ketols of 13-hydroperoxy NAE = 12-oxo-13-hydroxy-N-(9Z)-octadecenoylethanolamine. C, Electron impact mass spectrum of fully reduced TMS-ethers of 12-oxo-13-hydroxy-N-(9Z)-octadecenoylethanolamine, the compound eluting with retention time of 18.22 min (A and B). Inclusion of FAAH inhibitors to study sensitivity of the cottonseed FFA producing enzyme showed a concentration-dependent effect . There was approximately a 40% inhibition of FFA production with 10 mm phenylmethylsulfonyl fluoride, a potent inhibitor of FAAH activity consistent with the catalytic mechanism of a Ser hydrolase . In contrast, there was only 10% inhibition by 10 mum arachidonyl trifluoromethyl ketone (ATMK), an analog of anandamide. ATMK is a potent inhibitor of mammalian anandamide hydrolysis showing complete inhibition at 7.5 mum . The difference in sensitivity to ATMK may indicate a different property of cottonseed NAE AHase or may simply be a reflection of the lack of arachidonyl fatty acid derivatives in higher plant tissues. Table III | The effects of FAAH inhibitors on metabolism of NAE 18:2 in vitro Developmental Changes in NAE Metabolism | The capacity for NAE oxylipin formation by cytosol-enriched fractions increased during seed imbibition to the highest levels (7 nmol h-1 mg-1 protein) by 8 h after commencing imbibition . Activity remained at this level throughout the first 24 h of postgerminative growth. Under these conditions, cottonseeds germinate at 12 to 18 h after commencing imbibition, and lipid mobilization (marked by glyoxylate cycle enzymes) is most active in 24- to 48-h-old seedlings (for summary, see ). As before, inclusion of NDGA helped to confirm that the activity was attributable to LOX-like enzyme. Figure 5 | Time course of NAE 18:2 metabolism in cytosol-enriched fractions isolated at various times of seed imbibition (up to 4 h), germination (at about 12 h), and postgerminative growth. Time course of NAE 18:2 metabolism in cytosol-enriched fractions isolated at various times of seed imbibition (up to 4 h), germination (at about 12 h), and postgerminative growth. The amount of oxylipin production was determined by incubating 100 mum synthetic NAE 18:2 (20,000 dpm) with 400 muL of the supernatant (150,000gmax supernatant of 10,000gmax supernatant) in a final volume of 800 muL with shaking for 1 h at 30C. Lipids were extracted as described in "Materials and Methods." Identification and quantification of radiolabeled NAE-lipids were performed by radiometric scanning. The maximum specific activity of "LOX" was at 8 h (just before seed germination). When 100 mum NDGA was used, there was almost 50% inhibition of oxylipin production. It indicates that the product of NAE 18:2 was most likely attributable to enzymatic action of LOX. The data points are means and sd of three replicates of one representative experiment repeated for three times. The developmental change in cytosolic NAE oxylipin formation was somewhat different from that associated with microsomes . Microsomal NAE oxylipin formation, like that associated with the cytosol-enriched fractions, increased during imbibition to its highest levels by 8 h after commencing imbibition; however, the membrane-associated activity dropped substantially by 16 h and was undetectable by 24 h (not shown). The developmental change in membrane-associated NAE oxylipin formation paralleled that of NAE AHase . In the case of microsomes, NDGA had a profound effect on NAE oxylipin formation (indicative of a LOX-mediated pathway) but had only a modest effect on AHase activity. Overall, these results indicate that metabolism of NAEs is most active during seed imbibition, just before seed germination, and well before the period of lipid mobilization for postgerminative seedling growth. This conclusion is consistent with the time period of NAE depletion in vivo . Figure 6 | Time course of NAE 18:2 metabolism in microsomes isolated at various stages of cottonseed imbibition, germination, and postgerminative seedling growth. Time course of NAE 18:2 metabolism in microsomes isolated at various stages of cottonseed imbibition, germination, and postgerminative seedling growth. Both AHase and LOX activities were detected in microsomes (see also Tables and ). Both activities increased before and decreased after seed germination. NDGA reduced substantial oxylipin formation, whereas there was minimal effect on FFA production in vitro. The highest specific activities of LOX and NAE AHase were at 8 h and 4 to 8 h, respectively, after commencing imbibition (just before seed germination, 12 --18 h). The data points are means and sd of three replicates within a given experiment and are representative of three experiments. Enzymatic Properties of AHase and NAE Oxylipin Formation | We compared the enzymatic properties of the NAE AHase pathway and NAE-LOX-mediated pathway in cottonseed microsomes to estimate the relative capacity of each pathway to contribute to the metabolism of the predominant seed NAEs (Fig. ; Table ). Both pathways exhibited typical Michaelis-Menten kinetics when initial velocity measurements were made at increasing NAE concentrations . For the NAE AHase, the apparent Km values were similar for NAE 16:0 and NAE 18:2 (Km values for NAE 16:0 and NAE 18:2 are 83 mum and 74 mum, respectively), although the apparent Vmax values estimated for NAE 18:2 were nearly twice that as for NAE 16:0 (Vmax values for NAE 16:0 and NAE 18:2 are 1.6 and 3.0 nmol h-1 mg-1 protein, respectively). NAE oxylipin formation from NAE 18:2 also exhibited typical saturation kinetics at increasing NAE concentrations (Fig. B). The apparent Km value (70 mum) was similar to that estimated for the NAE AHase, indicating similar substrate affinities for both pathways. On the other hand, the apparent Vmax (12 nmol h-1 mg-1 protein) was four times that of the NAE AHase-mediated pathway, indicating a greater capacity for NAE oxylipin formation than for NAE hydrolysis. Although comparisons of kinetic parameters cannot be interpreted to indicate relative metabolic flux, these data do indicate that the capacity for NAE consumption by these two pathways estimated in vitro exceeds that required for NAE depletion in vivo. The degradation rates for NAE 18:2 and NAE 16:0 were calculated to be 15 and 4.5 ng h-1 seed-1, respectively, in vivo ; whereas the rates were 160 and 80 ng h-1 seed-1 for NAE 18:2 and NAE 16:0, respectively, in vitro through NAE AHase pathway, based on data presented in Table . The conversion rate for NAE 18:2 through NAE-LOX pathway was 20 mug h-1 seed-1 in vitro , indicating this pathway was relatively more capable of contributing to the metabolism of polyunsaturated NAE (18:2) than was AHase. Figure 7 | Concentration-dependent formation of FFA from NAE substrate in cottonseed microsomes under initial velocity conditions. Concentration-dependent formation of FFA from NAE substrate in cottonseed microsomes under initial velocity conditions. The apparent Km and Vmax for NAE 16:0 were estimated to be 83 mum and 1.6 nmol h-1 mg-1 protein, respectively. The apparent Km and Vmax for NAE 18:2 were estimated to be 74 mum and 3.0 nmol h-1 mg-1 protein, respectively. Concentration-dependent formation of oxylipin from NAE 18:2 incubating the same cell fraction showed the apparent Km and Vmax to be 70 mum and 12 nmol h-1 mg-1 protein, respectively. In A, NAE 16:0 and NAE 18:2 were the substrates, and in B, NAE 18:2 was the substrate. Lines represent nonlinear regression fits of the data using Michaelis-Menten equation (Prism software, v3.0, GraphPad Software, San Diego). Kinetic parameters were estimated from regression analyses. Although the microsomal NAE AHase appears to have similar affinities for saturated and polyunsaturated NAE species, the maximum rate of product formation from NAE 18:2 is about twice that for NAE 16:0. Similarly in B, the enzyme has similar affinity of 70 mum with four times greater production rate of conversion of NAE 18:2 into NAE-LOX product. Data points are averages of triplicate samples within a representative experiment. Table IV | Summary of kinetic parameters of NAE utilizing enzymes in cottonseed microsomes NAE Metabolism in Vivo | NAE was metabolized in vivo by both the NAE-LOX pathway and NAE AHase pathway during seed imbibition . Radiotracer experiments with imbibing seeds showed that [1-14C]18:2 NAE was converted to NAE oxylipin and FFA in a time-dependent manner. In the presence of the LOX inhibitor, NDGA there was reduction of NAE-oxylipin formation, which was particularly evident after 4 h. Production of FFA from NAE 18:2 also was reduced in the presence of the LOX inhibitor, which may suggest these two pathways are somewhat interdependent. Application of NAE 18:2 or NDGA under these conditions did not influence seed germination. The metabolic results in vivo are consistent with those obtained in vitro and confirm that NAE 18:2 is metabolized by LOX and AHase pathways, which together likely account for the in vivo depletion of NAE 18:2 during seed imbibition . Figure 8 | Metabolism of [14C]NAE 18:2 in vivo by imbibing cottonseeds. Metabolism of [14C]NAE 18:2 in vivo by imbibing cottonseeds. Seed coats were removed from imbibing (4 h) seeds, which were then incubated 30 min with (or without, dimethyl sulfoxide-only control) LOX inhibitor (5 muL of 16 mm NDGA per seed) before application of radiolabeled NAE 18:2 (0.1 muCi seed-1, 2.04 mCi mmol-1). Imbibed seeds were incubated on moist filter paper in covered petri dishes for additional 1, 2, and 4 h in the dark. Lipids were extracted and separated by TLC as described in "Materials and Methods," and distribution of radioactivity was evaluated by radiometric scanning (Bioscan system 200 image scanner). The data points are means and sd of four replicates within a single experiment. Additional experiments showed identical trends, although the efficiency of incorporation of radiolabel varied somewhat from experiment to experiment. Both oxylipin and FFA production increased with time. The conversion of NAE 18:2 to NAE-oxylipin was reduced by application of a classical LOX inhibitor, NDGA (see Fig. ). NDGA had some effect on production of FFA, as well. For clarity, the amount of radioactivity in NAE 18:2 is not included but represented nearly all of the remaining proportion of radioactive lipid. DISCUSSION : The recent identification and quantification of NAEs in desiccated seeds and their disappearance after 4 h of imbibition raised the question as to the metabolic fate of these compounds. Preliminary evidence suggested that these NAEs could be hydrolyzed in imbibed seeds by an NAE AHase activity , similar to the FAAH found in some animal systems. However, here, a detailed evaluation now indicates a more complicated scheme for NAE metabolism in seeds than originally anticipated . NAEs with saturated fatty acid constituents, like the endogenous NAE 16:0, are indeed hydrolyzed by an AHase activity, whereas polyunsaturated NAEs (NAE 18:2) appear to be metabolized by two pathways, the NAE AHase pathway and a LOX-mediated pathway. The AHase pathway leads to the formation of FFAs, which could be reincorporated into NAPE, the precursor for NAEs (see Fig. ). In fact, NAPE biosynthesis was shown to be increased during seed imbibition, germination, and early postgerminative growth as judged by increases in lipid levels and enzyme activity. Figure 9 | Proposed scheme for the metabolism of NAEs in seeds and seedlings. Proposed scheme for the metabolism of NAEs in seeds and seedlings. NAPEs are hydrolyzed by PLD to yield saturated and unsaturated species of NAE. FAAH hydrolyzes NAEs to FFAs and ethanolamine. As an alternative, some polyunsaturated NAEs are metabolized by 13-LOX and 13-AOS to yield NAE oxylipins (13-hydroperoxy NAE, 13-hydroperoxy octadecadienoylethanolamine and 12-oxo-13-hydroxy NAE, 12-oxo-13-hydroxy octadecenoylethanolamine). FFAs formed from hydrolysis of NAE or glycerophospholipids can be incorporated directly into the N-position of NAPE by NAPE synthase. This overall cycle could be used for signal transduction (formation of NAE lipid mediators) or to scavenge FFAs (for membrane protection) depending upon cellular demands. The "X" in the glycerophospholipid molecule represents the head group of any phospholipid class (e.g. Ser, choline, ethanolamine, etc.). Metabolism of NAEs during seed imbibition by a membrane-associated AHase activity is reminiscent of the mechanism for NAE activation in animal systems. The FAAH has been cloned from a number of mammalian tissues and shown to encode an enzyme with amidase and esterase activities of broad substrate specificity including several fatty acid amides and acylglycerols . More extensive studies of substrate specificity for the purified recombinant rat enzyme recently showed that it was capable of hydrolyzing a wide array of unsaturated, and to a less extent saturated, fatty acid primary amides . As the chain length of fatty acid (saturated) constituents decreased, the rate of hydrolysis increase . Mammalian FAAH inhibitors had a relatively modest effect on cottonseed NAE AHase enzyme(s) . This may indicate that plant NAE AHases are diverged from the mammalian counterparts to reflect a specificity for NAEs abundant in plant tissues. However, recently reported a new enzyme from lung tissues, designated NAE 16:0 hydrolase, that was much less sensitive to inhibition of FAAH, suggesting that animal systems may contain more than one NAE AHase. Additional work will be required at the biochemical and molecular levels to more fully understand the nature of NAE AHase activity(ies) in plant systems, and to define what specific role this pathway plays in seed germination. Recent evidence demonstrated that purified plant LOX, AOS, and HPO lyase could metabolize synthetic NAEs to generate novel oxylipins , raising the possibility that plants might catalyze these reactions in vivo. Here, we provide several lines of evidence that support this concept and highlight seed imbibition and germination as a period intensely involved in formation of these metabolites. The time course of NAE 18:2 metabolism in cytosol-enriched fractions isolated at various times of seed imbibition, germination, and postgerminative growth showed the maximum specific activity of LOX (7 nmol h-1 mg-1 protein) at 8 h after imbibition, which is just before seed germination . Microsomes isolated at the same developmental stages similarly showed the same time period for the highest specific activity of LOX (8 nmol h-1 mg-1 protein) and NAE AHase (2 nmol h-1 mg-1 protein; Fig. ) consistent with results in vivo . The fate of NAE-derived oxylipins is unclear at this point, but it is tempting to speculate these NAE oxylipins may have a role of their own during seed germination. For example, recently proposed that 13-LOX-mediated pathway is associated with a "priming" function in oilseeds (Brassica napus) for postgerminative triacylglycerol mobilization. It should be emphasized, however, that here the timing of the most intensive NAE metabolism of cottonseed NAEs preceded radicle emergence and lipid mobilization (Figs. and ; also see ), so these NAE-derived oxylipins may play other roles perhaps as lipid mediators involved in the regulation of seed germination. In any case, the role of NAE metabolism in imbibing seeds is not specific to oilseeds because a similar depletion of seed NAEs was observed in non-oilseeds (e.g. pea [Pisum sativum]; ). MATERIALS AND METHODS : Chemicals | [1-14C]Palmitic acid (53 mCi-1 mmol in ethanol) and [1-14C]linoleic acid (53 mCi mmol-1 in ethanol) were from PerkinElmer Life Sciences (Boston). Chemical Synthesis of NAEs | Specific NAE types were synthesized from respective radiolabeled FFA by first producing the fatty acyl chloride . The FFA was dissolved in dichloromethane and then mixed with dimethylformamide (1 mol equivalent) and oxalyl chloride (1.2 mol equivalent). The fatty acyl chloride was mixed with a 10-fold excess of ethanolamine to convert the acyl chloride to the corresponding NAE. Products were extracted in dichloromethane and purified by TLC. Yield and purity of NAEs were estimated by radiometric scanning (System 200 image scanner, Bioscan, Washington, DC). The yield was routinely 65% to 70% (starting from FFA), and purity after TLC was >99%. Radiospecific activity was calculated from the original 14C-labeled FFA and adjusted accordingly with non-radiolabeled synthetic NAE produced by the same method. Plant Material | Cotton (Gossypium hirsutum L. Stoneville 7A glandless) seeds were provided by Dr. R.B. Turley (Cotton Physiology and Genetics Laboratory, U.S. Department of Agriculture-Agricultural Research Service, Stoneville, MS). For all of the experiments, seeds were surface-sterilized with 20% (v/v) commercial bleach (sodium hypochlorite) solution for 5 min. Seeds were rinsed several times and imbibed in distilled water (in the dark) for 4 h at 30C with aeration. For time-course experiments, imbibed seeds were placed in filter paper scrolls as previously described and germinated and grown in the dark (30C). Preparation of Cellular Fractions | Cell fractions were prepared by differential centrifugation as described with some modifications. In brief, seeds imbibed for 4 h were chopped with a steel blade on ice in homogenization medium containing 100 mm potassium-phosphate (pH 7.2), 10 mm KCl, 1 mm EDTA, 1 mm EGTA, and 400 mm Suc. The homogenates were filtered through four layers of cheesecloth and centrifuged at 650gmax (4C) for 10 min in a centrifuge (RC 5C, SS 34 rotor, Sorvall, Newton, CT). The 650gmax supernatant was centrifuged at 10,000gmax (4C) for 30 min in the same centrifuge. The resulting supernatant was centrifuged at 150,000gmax (4C) for 60 min in Sorvall Discovery 90 model ultracentrifuge centrifuge (Ti45 rotor, Beckman Coulter, Fullerton, CA). Microsomes (150,000gmax pellet) were resuspended in homogenization medium (0.3 mL per original gram fresh weight). Protein concentration was estimated according to using bovine serum albumin as the standard. Lipid Extractions and Analysis | For enzymatic assays in vitro, 100 mum (20,000 dpm) of [14C]NAE substrate (combined with nonradioactive NAE and radiolabeled on the carbonyl carbon) was suspended by sonication in 50 mm MES buffer (pH 6.5). Reactions were initiated by adding 400-muL cell fractions. Reactions were terminated, and lipids were extracted into chloroform according modified to eliminate endogenous PLD activity . In brief, reactions were stopped by adding 2 mL of hot 2-propanol (70C) to 800 muL of the aqueous assay reaction mixture and heated at 70C for 30 min. One milliliter of chloroform was added to the mixtures, and lipids were extracted at 4C overnight. One milliliter of chloroform and 2 mL of KCl (1 m) were added to induce phase separation. The aqueous layer was aspirated off, and the organic layer was washed twice with 2 mL of 1 m KCl and once with deionized water (MilliQ UF plus). The organic phase was collected and dried under nitrogen. Lipid classes were separated by TLC (hexane:ethyl acetate:methanol, 60:40:5; v/v). Identification and quantification of radiolabeled lipids were performed by radiometric scanning (Bioscan system 200 image scanner) and comigration with known standards. NAE Metabolism in Vivo | For radiolabeling experiments in vivo, seed coats were removed from imbibed (4 h) seeds, which were then preincubated for 30 min with (or without, dimethyl sulfoxide-only control) LOX inhibitor (5 muL of 16 mm NDGA per seed) before application of radiolabeled NAE 18:2. Synthetic NAE 18:2 (0.1 muCi, 2.04 mCi-1 mmol) was applied in a small volume to each seed. Imbibed seeds were incubated for various time periods in the dark on moistened filter paper in 100- x 15-mm covered petri dishes. Radiolabeled lipids were extracted and analyzed as described above. Spectrophotometric Assay of Lipid Hydroperoxide Formation | Determination of LOX activity also was performed with a commercially available lipid hydroperoxide (LPO) assay kit (catalog no. 705002, Cayman Chemical, Ann Arbor, MI). For each assay, 80 nmol of NAE 18:2 was used as substrate and incubated with crude extract for 1 h at 30C with shaking (110 rpm). The lipid peroxides that were formed were extracted from the samples into chloroform and quantified by measuring A500 compared with the standard lipid hydroperoxide (13-hydroperoxy octadecadienoic acid). Lipid peroxide formation in imbibed cottonseed extracts was protein, temperature, and NAE 18:2 dependent. Identification of NAE Oxylipins by GC/MS | NAE-derived oxylipins were evaluated by GC/MS as previously described , except lipids were reduced with NaBH4 instead of NaBD4. In brief, imbibed cottonseed cell fractions, prepared as described above, were diluted 1:1 in 50 mm potassium phosphate buffer, pH 6, and incubated with 100 mum NAE 18:2 for 2 h at room temperature. Lipid products were extracted, reduced with NaBH4, methylated, and trimethylsilylated before identification by GC/MS using the conditions described previously . Backmatter: PMID- 12226519 TI - Loss of Nuclear Gene Expression during the Phytochrome A-Mediated Far-Red Block of Greening Response AB - We have examined the expression of the HEMA1 gene, which encodes the key chlorophyll synthesis enzyme glutamyl-tRNA reductase, during the phytochrome A-mediated far-red light (FR) block of greening response in Arabidopsis. Our results demonstrate that the FR block of greening comprises two separate responses: a white light (WL) intensity-independent response that requires 3 d of FR and is associated with a loss of expression of the nuclear genes HEMA1 and Lhcb following the transfer to WL (transcriptionally coupled response) and a WL intensity-dependent response that is induced by 1 d of FR and is transcriptionally uncoupled. Both responses required phytochrome A. The transcriptionally uncoupled response correlated with a deregulation of tetrapyrrole synthesis and potential photooxidative damage and was inhibited by cytokinin. The transcriptionally coupled FR response was additive with the loss of expression following Norflurazon-induced photobleaching and was absent in the presence of sucrose or after lower fluence rate (1 mumol m-2 s-1) FR treatments. Both pathways leading to the loss of nuclear gene expression were inhibited by overexpression of NADPH:protochlorophyllide oxidoreductase, indicating a role for plastid signaling in the FR-mediated pathway. The significance of identifying a distinct phytochrome A-mediated plastid signaling pathway is discussed. Keywords: Introduction : The coordinated synthesis of chlorophylls and chlorophyll-binding proteins is critically important during emergence of the etiolated seedling into light, and is essential for the normal development of functional chloroplasts. The key regulatory step in chlorophyll synthesis is the formation of 5-aminolevulinic acid (ALA), which is rate limiting for the tetrapyrrole pathway . As a consequence, inhibition of ALA synthesis through inhibitors or anti-sense approaches has a dramatic effect on chloroplast development, resulting in plants that are highly susceptible to photobleaching. In a converse manner, excess ALA also affects chloroplast development and leads to an overaccumulation of porphyrins, which can lead to severe photooxidative damage . Uncoupling tetrapyrrole synthesis from chloroplast development may also be achieved by irradiating etiolated seedlings with far-red (FR) light . Growth of Arabidopsis seedlings under continuous FR (FRc) can induce partial photomorphogenesis with reduced hypocotyl elongation growth and cotyledon expansion. These responses are characteristic of the FR-high-irradiance response (FR-HIR; ), which is a specific function of phytochrome A . FRc irradiation can also activate many of the processes required for chloroplast development. These include the induction of nuclear genes encoding chlorophyll a/b-binding proteins and other photosynthetic proteins and the transcription of chloroplast genes and replication of plastid DNA . However, because photoconversion of protochlorophyllide (Pchlide) to chlorophyllide (Chlide) by the enzyme NADPH:Pchlide oxidoreductase (POR) is inefficient under FR wavelengths, chlorophyll synthesis is severely limited under these conditions. Thus, phytochrome A-mediated induction of selected facets of plastid development may occur in the absence of conditions that allow the synthesis of corresponding levels of chlorophyll. It has previously been shown that seedlings of Arabidopsis and tomato (Lycopersicon esculentum) grown under prolonged FR not only fail to accumulate chlorophyll, but also are unable to green when subsequently transferred to white light (WL; ; ). This phenomenon, known as the FR block of greening response, has been characterized at the molecular level as a depletion of PORA (and partially of PORB) and the concomitant loss of the ordered membrane system of the prolamellar body . Each of these effects displays characteristics of an FR-HIR and is dependent on phytochrome A activity . Further evidence that the loss of POR is crucial to the FR block of greening comes from a transgenic approach in which overexpression of PORA was able to maintain WL-mediated greening in FR-pretreated seedlings . It is likely that PORA has two important roles in maintaining chlorophyll synthesis in WL. Not only is it required for light-dependent chlorophyll synthesis, but it has a role in buffering against photooxidative damage that occurs during the illumination of etiolated, Pchlide-containing tissues . This second role has been largely supported by subsequent studies with PORA- and PORB-overexpressing lines . However, it is interesting that POR overexpression could not fully counteract the FR block, and FR-pretreated transgenic seedlings still contained approximately two-thirds of the chlorophyll levels of dark-treated controls . This suggests that although the level of POR is a critical determinant of this response, other FR-mediated processes may also be important. Prolonged FR irradiation has been shown to cause the formation of aberrant plastids in developing cotyledons, and the irreversible nature of the FR block of greening response has been attributed to the ensuing degradation of plastid components suggested by the formation of vesicles in the stroma . One possibility is that this FR-mediated effect on plastid integrity leads to the loss of a plastid signal required for normal expression of photosynthetically related genes. Such a signal has been previously identified through the action of photobleaching herbicides and is required for the expression of numerous photosynthetically related nuclear genes , including Lhcb in Arabidopsis . This signal is also closely involved with phytochrome signaling pathways and can be thought of as gating phytochrome responses in a similar way to that proposed for circadian control of light-induced gene expression . We have examined the hypothesis that plastid signaling is involved in the FR block of greening response by analyzing the expression profile of HEMA1. This gene, which encodes glutamyl-tRNA reductase, the first committed enzyme of tetrapyrrole synthesis, is strongly expressed in photosynthetic tissues and is regulated by light, including an FR-HIR, and a plastid signal . However, the light-regulated expression pattern of HEMA1 is not identical to other light-regulated genes, as HEMA1 does not exhibit a very low-fluence response and utilizes light signaling pathways that partly diverge from those for light regulation of Lhcb (A.C. McCormac and M.J. Terry, unpublished data). Given the importance of tetrapyrroles in the FR block of greening response and the increasingly significant relationship between tetrapyrroles and plastid signaling , the expression of HEMA1 under these conditions is likely to be highly informative for our understanding of plastid development. RESULTS : The FR Block of Greening Response Is Associated with an Inhibition of Promoter Responsiveness to WL | Figure A shows that, as seen previously , 3 d of FRc (10 mumol m-2 s-1) irradiation of etiolated seedlings fully inhibited their ability to green under subsequent WL, whereas dark-grown seedlings at the same developmental age could green normally. This effect was independent of WL intensity (Fig. A; ) and was also found to be associated with a loss of HEMA1 and Lhcb expression (3 d FRc; Fig. B). Seedlings that had received a pretreatment of just 1 d of FRc suffered only a small, but consistently observed, loss of greening capacity that was enhanced by high-intensity WL (Fig. A). These seedlings were still able to accumulate high levels of HEMA1 and Lhcb mRNA (1 d FRc; Fig. B). Under these conditions, FRc stimulated subsequent WL induction of HEMA1 expression by up to 2-fold compared with seedlings transferred directly from darkness to WL (Fig. B). This stimulation was most evident under low-intensity (8 mumol m-2 s-1) WL, but was absent under high-intensity (250 mumol m-2 s-1) WL (data not shown). These results were confirmed using transgenic Arabidopsis seedlings expressing a HEMA1 promoter:gusA construct (Fig. C). Furthermore, the 1-d FRc-mediated stimulation of WL-induced HEMA1 expression and the 3-d FRc-mediated inhibition of HEMA1 expression following transfer to WL were absent in a phyA genetic background, indicating that these are both phytochrome A-mediated responses (Fig. C). Therefore, the FR block of greening response is characterized not only as a progressive loss of greening ability, but also by a marked change in the transcriptional response to WL. Figure 1 | FRc preirradiation leads to a reduction in greening and nuclear gene expression following transfer to WL. FRc preirradiation leads to a reduction in greening and nuclear gene expression following transfer to WL. A, Greening of cotyledons under different WL fluence rates after growth in the dark or FRc. Seedlings were germinated without Suc in the dark for 2 d. They were subsequently grown for a further period (as indicated) under FRc (white symbols) or were maintained in the dark (black symbols) for the equivalent period. Seedlings were then transferred to a WL source of 250 mumol m-2 s-1 (circles) or 8 mumol m-2 s-1 (triangles). B, HEMA1 and Lhcb mRNA accumulation under WL (130 mumol m-2 s-1) following a 1- or 3-d FRc pretreatment. Seedlings were grown without Suc. The RNA gel blot was sequentially hybridized with probes for HEMA1, Lhcb, and 18S transcripts. C, Effect of an FR pretreatment on the subsequent response to WL of a HEMA1 promoter:beta-glucuronidase A (gusA) transgene. Transgenic WT or phyA Arabidopsis seedlings were germinated for 1 d in the dark followed by a 1- or 3-d FRc treatment (or equivalent darkness). Seedlings were then transferred to WL (8 mumol m-2 s-1) for 3 d prior to measurement of GUS activity. Data shown are the mean +- se (n = 3). The FR Block of Greening Response Develops without Inhibition of HEMA1 Transcription in the Presence of Suc | A full FR block of greening response was avoided if seedlings were grown on a medium supplemented with 3% (w/v) Suc (Fig. A; ). However, even in the presence of Suc, FR pretreatments still elicited a partial impairment of the greening response as compared with that in dark-grown seedlings (Fig. A). This partial block of greening was not seen following FR irradiation of phyA lines (data not shown). In contrast to the FR block of greening response in the absence of Suc (Fig. A), the partial FR block in wild-type (WT) lines was shown to be dependent on WL intensity (Fig. A). In addition, for seedlings grown on 3% (w/v) Suc, an FR pretreatment produced no discernible inhibition of HEMA1 or Lhcb transcript accumulation following transfer to 130 mumol m-2 s-1 WL (Fig. B), even with a 75% inhibition of greening capacity. The accumulation of GUS activity driven by the HEMA1 promoter also showed a normal response to WL following a 3-d FRc preirradiation and was even elevated compared with the response of dark-grown seedlings in low-intensity WL (data not shown). Thus, under these conditions, the effect of an FR preirradiation was to separate HEMA1 expression from chlorophyll accumulation. Figure 2 | Suc inhibits the loss of nuclear gene expression in FRc-grown seedlings transferred to WL. Suc inhibits the loss of nuclear gene expression in FRc-grown seedlings transferred to WL. A, Greening of cotyledons under different WL fluence rates after growth in darkness or FRc. Seedlings were germinated on 3% (w/v) Suc in the dark for 2 d. They were subsequently grown for a further period (as indicated) under FRc (white symbols) or were maintained in darkness (black symbols) for the equivalent period. Seedlings were then transferred to a WL source of 250 mumol m-2 s-1 (circles) or 8 mumol m-2 s-1 (triangles). B, HEMA1 and Lhcb mRNA accumulation under WL (130 mumol m-2 s-1) following a 1- or 3-d FRc pretreatment. Seedlings were grown on 3% (w/v) Suc. The RNA gel blot was sequentially hybridized with probes for HEMA1, Lhcb, and 18S transcripts. The response characteristics on Suc are similar to those seen in the absence of Suc, but following a short period (1 d) of FR irradiation (Fig. , A and B) and they enable us to define two separate responses leading to an FR block of greening. The first response requires only 1 d of FRc and results in a WL intensity-dependent, incomplete loss of greening ability that is not associated with a loss of HEMA1 and Lhcb expression. We subsequently refer to this as the transcriptionally uncoupled response. The second response requires a longer period of FRc (3 d), is independent of WL intensity, and leads to a complete loss of greening ability (<10%) and the inhibition of HEMA1 and Lhcb expression. This WL intensity-independent response is specifically inhibited by Suc and is now referred to as the transcriptionally coupled response. Different Transcriptional Responses under FRc Define the Two FR Block of Greening Responses under WL | To further define the characteristics of the two response pathways leading to the FR block of greening, we examined the expression of HEMA1 during FRc. Figure A shows a time course of HEMA1 and Lhcb expression for a 4-d period of FR irradiation. A peak in HEMA1 and Lhcb mRNA levels occurred at around 1 d from the start of the FR treatment and declined thereafter, returning almost to dark levels by d 4 (Fig. , A and B). This type of short-lived expression profile is comparable with that reported previously for a wide range of plastid-associated nuclear genes under FRc , but is in contrast to the sustained transcriptional response seen under Rc (Fig. C). The post-peak phase of FR-mediated HEMA1 expression (i.e. >=72 h FRc; Fig. A) temporally coincides with the inability to reinitiate transcription when exposed to WL (3 d FRc; Fig. B). Because Suc can prevent the loss of HEMA1 expression in WL after a prolonged FRc treatment (Fig. B), we tested whether it could also block the loss of FR-induced HEMA1 expression. Figure D shows that the loss of HEMA1 expression under FRc was prevented by the presence of 3% (w/v) Suc with maximal expression levels (detected after 3 d of FRc) remaining for >=7d. However, the presence of 3% (w/v) Suc also substantially reduced the accumulation of HEMA1 (and Lhcb; data not shown) mRNA seen within the first 24 h of FRc (Fig. , compare A and D). This is consistent with previous reports of the effect of Suc on phytochrome A signaling . Figure 3 | The response of the HEMA1 promoter to light is subject to developmental aging, FR fluence rate, and Suc. The response of the HEMA1 promoter to light is subject to developmental aging, FR fluence rate, and Suc. A, Transcript levels of HEMA1 and Lhcb under prolonged (16 --96 h) FR irradiation. Seedlings were grown without Suc and were germinated for 2 d in the dark before transfer (at time = 0) to FRc. B, Densitometry scans of band intensities as shown in (A). C, HEMA1 transcript levels under prolonged (24 --96 h) R irradiation. Seedlings were grown without Suc and were germinated for 2 d in the dark before transfer (at time = 0) to Rc. D, HEMA1 mRNA levels in seedlings grown under prolonged (1 --9 d) FR irradiation in the presence of 3% (w/v) Suc. Seedlings were germinated for 3 d in the dark prior to transfer (at t = 0) to FRc. E, HEMA1 mRNA levels in seedlings grown under 1 or 10 mumol m-2 s-1 FRc. F, HEMA1 mRNA levels under WL in seedlings grown previously for 3 d in the dark or under 1 or 10 mumol m-2 s-1 FRc. G, Effect of aging on the subsequent response of a HEMA1 promoter:gusA construct to WL. Seedlings were grown with or without 3% (w/v) Suc in the dark and were subsequently transferred to WL for 3 d prior to measurement of GUS activity. Data shown are the mean +- se (n = 5). Phytochrome A-mediated FR responses are also typically affected by the rate of FR delivery . Therefore, we examined this relationship for the transcriptional response to FR and the FR block of greening response by comparing the effect of irradiating seedlings (grown without Suc) with FRc at 1 mumol m-2 s-1 (Fig. E) instead of 10 mumol m-2 s-1 (Fig. , A and E). Under the lower fluence rate, the levels of HEMA1 mRNA showed the same pattern of transient up-regulation, but the rate of transcriptional increase and magnitude of the peak were reduced (Fig. E). A 3-d pretreatment under 1 mumol m-2 s-1 FRc failed to induce the FR-mediated block of the WL transcriptional response (Fig. F). Instead, the FR-irradiated seedlings displayed the characteristics of the transcriptionally uncoupled FR block of greening, showing strong nuclear transcription (Fig. F) and a WL intensity-dependent loss of greening (data not shown). In each case, phyA mutants demonstrated that there was an absolute requirement for phytochrome A for the response to FR (data not shown). To test whether the decline in mRNA accumulation under FRc (Fig. B; 24 --96 h) could be explained as a general loss of light responsiveness rather than the specific loss of phytochrome A-mediated signaling, we examined the effect of prolonging the period in darkness on the ability of seedlings to induce HEMA1 in response to WL. There was a progressive loss of HEMA1 light responsiveness to WL (Fig. G) and also FR (data not shown). This effect was also seen in phyA mutants, which demonstrates that this is a time-dependent, but phytochrome A-independent, response (data not shown). In addition, a time-dependent depletion of greening in dark-grown seedlings was seen in the absence of Suc (Fig. A). When 3% (w/v) Suc was included in the growth medium, the time-dependent loss of promoter responsiveness to WL (Fig. G) and FR (data not shown) was prevented, suggesting that starvation may be a key factor in this process. Thus, it appears that in darkness or FR light (in the absence of exogenous sugars), nuclear genes progress into a nonresponsive state. By contrast, under R (Fig. C) or WL, the transcriptional response escapes this process. Figure 4 | Relationship between HEMA1 expression, ALA synthesis, and greening ability in FRc-grown seedlings. Relationship between HEMA1 expression, ALA synthesis, and greening ability in FRc-grown seedlings. A, ALA synthesis rates in seedlings of WT and the phyA mutant grown in the dark (black columns) or under 1 d of FRc (hatched columns), with or without 3% (w/v) Suc (as indicated). The inset graph shows the corresponding levels of Pchlide accumulated in the FR-irradiated WT and phyA seedlings. Data shown are the mean +- se (n = 3). B, Relationship between relative HEMA1 mRNA levels (as measured by densitometry scans of RNA gel blots) and corresponding rates of ALA synthesis in WT and phyA (inset) seedlings. Seedlings were allowed to germinate in the presence of 3% (w/v) Suc for 2 d in the dark and were transferred to FRc or maintained in the dark for an additional period of 1 or 3 d. Each datum point shows the individual value under either irradiation condition for these two incubation times for the Landsberg erecta (Ler) and Colombia (Col) backgrounds. C, Relationship between greening capacity measured after transfer to WL and ALA synthesis rates prior to transfer in dark- and FRc-grown WT and phyA (circle) seedlings. Seedlings were grown as for B. Data shown are the mean +- se (n = 3). These results demonstrate that a transcriptionally coupled block of greening response can proceed in the absence of light, although more slowly than under high-intensity FRc (see also Fig. A) and therefore appears to have no absolute requirement for phytochrome A. In contrast, the transcriptionally uncoupled effect on greening has a strict requirement for phytochrome A at the seedling stage (see also Fig. A). In addition, the two FR block of greening responses, as observed in WL, can be further distinguished by the different patterns of transcriptional response during the preceding period of FR irradiation. Thus, for FR irradiation to accelerate the development of the transcriptionally coupled block of greening response, a maximal FR-HIR must be achieved. By contrast, a submaximal FR-HIR can trigger the transcriptionally uncoupled block of greening. The Transcriptionally Uncoupled Response Is Associated with an Increase in ALA Synthesis | The FR block of greening has previously been ascribed, at least in part, to the photoexcitation of excess Pchlide in the absence of adequate POR buffering , and we hypothesized that the WL-dependent, transcriptionally uncoupled response was likely to be related to these processes. ALA synthesis is the rate-limiting step in Pchlide accumulation , and GluTR activity and HEMA1 expression are believed to be key determinants of this activity (see ). Because FRc resulted in an increase in HEMA1 expression (Fig. , A and B), we tested whether HEMA1 induction could also contribute to the transcriptionally uncoupled response by increasing ALA synthesis and, therefore, the pool of photosensitive Pchlide. Figure shows that irradiation with FRc significantly elevated ALA synthesis in WT seedlings (Fig. A) and this was correlated (r = 0.82) with HEMA1 mRNA levels under a range of different conditions (Fig. B). On 3% (w/v) Suc, the FRc induction of ALA synthesis was also seen, consistent with the observation previously that although Suc inhibits the transcriptionally coupled response, the transcriptionally uncoupled response is retained. In phyA seedlings, the FR-mediated increase in ALA synthesis was significantly reduced on 0% and 3% (w/v) Suc (Fig. A). The greater increase in ALA synthesis in WT seedlings grown under FRc also correlated with a higher level of Pchlide in WT versus phyA seedlings in FRc (Fig. A, inset). In contrast to the WT response, phyA seedlings showed no correlation between ALA synthesis rates and HEMA1 mRNA levels (Fig. B, inset), indicating that the FR-mediated increase in ALA synthesis in phyA was not attributable to transcriptional regulation of HEMA1. This strongly suggests a role for posttranscriptional effects on GluTR expression and/or activity in regulating ALA synthesis under these conditions. We next investigated the relationship between ALA synthesis rates and subsequent greening capacity. Figure C shows this data for WT and phyA lines grown under conditions in which only the transcriptionally uncoupled response is occurring (i.e. 1 --3 d of FRc with Suc or 1 d of FRc without Suc). There is a significant (r = 0.88) inverse relationship between ALA synthesis rates and the ability to green following transfer to WL. This suggests that increased ALA synthesis contributes to the transcriptionally uncoupled response. Loss of Nuclear Gene Expression following FRc- and Norflurazon (NF)-Induced Photobleaching Is Additive | The loss of nuclear gene expression in the transcriptionally coupled FR block of greening response is reminiscent of the loss of HEMA1 expression following plastid photobleaching . Therefore, we investigated the interaction between these two responses. Figure A shows that the carotenoid biosynthesis inhibitor NF inhibits HEMA1 expression under WL following growth in darkness on 0% or 3% (w/v) Suc. This response is the same in WT and phyA seedlings (Fig. , A and B). NF does not affect the HEMA1 response of etiolated seedlings while under FR (data not shown). However, when NF and a 3-d FR pretreatment were combined on 0% (w/v) Suc, the inhibitory effect on the subsequent WL responsiveness of the HEMA1 promoter exceeded that of either of the individual treatments (Fig. A). It should be noted that immediately prior to transfer to WL, the FR-irradiated seedlings in each case had higher levels of HEMA1 expression than dark controls and so had the potential to show a higher level of transcriptional inhibition under WL. However, direct observation of HEMA1 mRNA levels by northern blotting demonstrates that in the presence of NF, a substantial level of HEMA1 mRNA remains (data not shown). Therefore, the magnitude of the NF-mediated reduction of GUS activity was not limited following darkness, and only the combination of FR and NF produced the minimum promoter activity (Fig. A). The additive transcriptional effect of NF and an FR pretreatment was dependent on phytochrome A (Fig. B). In contrast, the transcriptionally uncoupled response, as seen in WT seedlings grown on 3% (w/v) Suc , did not enhance the NF-mediated inhibition of the HEMA1 WL response (Fig. A). Therefore, these results show that the loss of nuclear gene expression during the FRc-mediated transcriptionally coupled response and following NF-induced photobleaching is additive. This may result from two inputs into the same plastid signaling pathway or through two independent pathways. Figure 5 | POR overexpression rescues the loss of HEMA1 expression in WL following growth in FRc or NF. POR overexpression rescues the loss of HEMA1 expression in WL following growth in FRc or NF. A, Effect of a combined FR preirradiation and NF treatment on the WL response of a HEMA1 promoter:gusA reporter gene. WT seedlings were grown with or without 3% (w/v) Suc for 3 d in FRc or darkness prior to transfer to WL (130 mumol m-2 s-1) for an additional 3 d. NF was included in the media at 5 mum and resulted in the cotyledons appearing completely white under all treatments. Values were normalized to the respective dark control levels (=100) within each experiment. Data shown are the mean +- se (n = 5). B, HEMA1 promoter:gusA expression in phyA seedlings under the same conditions as shown in A. C, The effect of POR overexpression on HEMA1 expression. Seedlings of PORA- or PORB-overexpressors (PAO-3 and PBO-1) were germinated in the dark without Suc and were irradiated for 0 to 3 d under FRc. HEMA1 mRNA accumulation was measured following transfer to WL (130 mumol m-2 s-1). D, The effect of 5 mum NF on the accumulation of HEMA1 mRNA in WL-irradiated seedlings of WT and transgenic lines overexpressing PORA (PAO-3) or PORB (PBO-1). Seedlings were grown in the absence of Suc for 3 d in the dark prior to transfer to WL for an additional 3 d. POR Overexpression Leads to a Maintenance of Nuclear Gene Expression following FRc- and NF-Induced Photobleaching | It has previously been shown that overexpression of POR can partially rescue the FR block of greening response, and this has been attributed to an inhibition of WL intensity-dependent photooxidative damage . We wanted to examine whether POR overexpression could also affect the transcriptionally coupled response and, therefore, we subjected transgenic Arabidopsis lines overexpressing PORA (PAO-3) or PORB (PBO-1; ) to conditions resulting in the loss of HEMA1 in WT seedlings (i.e. 2 --3 d of FRc on 0% [w/v] Suc). Upon transfer to WL, FR-treated PAO-3 and PBO-1 seedlings retained full expression of HEMA1 (Fig. C). Thus, POR overexpression inhibited the transcriptionally coupled FR block of greening response. The plastid localization of POR in the overexpressing lines supports the idea that loss of HEMA1 expression following FRc is the consequence of a signal emanating from the plastid. We also tested the effect of POR overexpression on the loss of plastid signaling following NF treatment. The cotyledons of NF-treated seedlings of WT and both POR-overexpressing lines appeared totally white. However, as seen for the WL transcriptional response following FRc, the POR-overexpressing seedlings accumulated significant levels of HEMA1 mRNA (Fig. D), indicating that positive plastid signaling was still functional despite photobleaching of the cotyledons. Therefore, POR overexpression results in a phenotype that is apparently similar to that of the genomes uncoupled mutants in which nuclear gene expression is maintained in the absence of functional plastids . However, in this case, it is not yet known whether POR overexpression results in a reduced level of plastid damage or has a direct affect on plastid signaling. The Transcriptionally Uncoupled But Not the Transcriptionally Coupled FR Block of Greening Response Is Inhibited by Cytokinin | We have described the FR block of greening as comprising two separate responses with respect to their different sensitivities to WL intensity and Suc and their effects on nuclear gene expression. Furthermore, there is a time-dependent transcription effect that is superimposed on these processes. Thus, the FR block of greening is a complex process encompassing a phytochrome A-dependent FR-HIR and phytochrome-independent effects. To try to isolate these components further, we investigated the effects of cytokinin treatment, as this hormone has previously been shown to influence greening capacity (e.g. ). WT and phyA seedlings grown for 5 d in the dark in the absence of Suc showed a time-dependent loss of light responsiveness (see Fig. ) and, therefore, showed a characteristically poor transcriptional response of the HEMA1 promoter to a WL treatment (Fig. A) and a partially reduced greening response (Fig. C). Application of the cytokinin 6-benzylaminopurine (BAP) inhibited both of these time-dependent processes (Fig. , A and C). In contrast, the inhibition of promoter responsiveness and cotyledon greening seen in WT seedlings subject to FRc (the transcriptionally coupled response) was not prevented by cytokinin (Fig. , A and C). Analysis of HEMA1 expression in darkness or FR light showed that cytokinin increased HEMA1 expression by an equal increment in both conditions (Fig. E) and, therefore, the FR response relative to darkness was unchanged. Figure 6 | The effect of cytokinin on the FR block greening response and HEMA1 expression. The effect of cytokinin on the FR block greening response and HEMA1 expression. A and B, The effect of cytokinin on WL-mediated induction of HEMA1 expression. WT and phyA seedlings were germinated in the dark for 2 d, transferred to FRc (or maintained in the dark) for 3 d, and then transferred to WL for an additional 3 d prior to measurement of GUS activity. Seedlings were grown in the absence (A) or presence (B) of 3% (w/v) Suc, and data shown are the mean +- se (n = 3). C and D, The effect of cytokinin on greening following transfer to WL. Seedlings were grown as shown in A and B and greening was measured in the absence (C) or presence (D) of 3% (w/v) Suc. Data shown are the mean +- se (n = 3). E and F, The effect of cytokinin on HEMA1 expression in the dark and FRc. WT and phyA seedlings were germinated in the dark for 2 d and were transferred to FRc (or maintained in the dark) for 3 d prior to measurement of GUS activity. Seedlings were grown in the absence (E) or presence (F) of 3% (w/v) Suc, and data shown are the mean +- se (n = 3). When seedlings were grown on 3% (w/v) Suc, an FR pretreatment had little affect on the subsequent HEMA1 response to WL (Fig. B), as seen previously , and cytokinin did not further influence this (Fig. B). For seedlings grown in darkness on 3% (w/v) Suc, the greening capacity was very slightly reduced by the presence of cytokinin (Fig. D). However, cytokinin could almost completely prevent the transcriptionally uncoupled block of greening response following FR irradiation of WT seedlings (Fig. D). Again, cytokinin stimulated HEMA1 expression in dark-grown seedlings (Fig. F), but in contrast to the effect on 0% (w/v) Suc (Fig. E), cytokinin inhibited the FR-mediated increase in HEMA1 expression on 3% (w/v) Suc (Fig. F). These results demonstrate that exogenous cytokinin can rescue the transcriptionally uncoupled response as it develops in the presence of Suc, but that it does not block the transcriptionally coupled response. However, there is also an effect of cytokinin on the time-dependent, FR-independent inhibition of greening. Therefore, the response to cytokinin defines a developmental separation of the two FR block of greening responses and can also isolate the FR-mediated transcriptionally coupled response from FR-independent events. DISCUSSION : Two Distinct Responses Leading to an FR Block of Greening | Here, we demonstrate that the FR block of greening comprises two distinct FR-dependent responses. As shown in Figure , the two responses can be distinguished by a number of different parameters. Following the onset of FRc irradiation, the first response detected (within 1 d of FRc) is a WL intensity-dependent incomplete loss of greening ability with a retention of WL induction of HEMA1 and Lhcb expression. We have shown that this response, which we have termed the transcriptionally uncoupled response, is prevented by the phyA mutation and by cytokinin treatment. In addition, it has previously been demonstrated that overexpression of POR can inhibit this response . Following longer periods of FRc irradiation (3 d of FRc), a WL intensity-independent response is also observed . This response is characterized by a complete loss of greening ability and the inhibition of HEMA1 and Lhcb expression following transfer to WL, and is inhibited by Suc and POR overexpression. We have termed this response the transcriptionally coupled response and it is also absent in the phyA mutant, consistent with the FR fluence rate dependence of this process. We have also identified a time-dependent effect on greening and the transcriptional light response that proceeds in the absence of FR. Figure 7 | Model for the progression of the FR block of greening response. Model for the progression of the FR block of greening response. FR block of greening is comprised of two separate phytochrome A-dependent responses that can be separated based on their dependence on WL intensity and their effects on greening and nuclear gene expression. The transcriptionally uncoupled response is WL intensity dependent, results in a partial loss of greening following transfer to WL, and is inhibited by cytokinin, the phyA mutation, and POR overexpression. The transcriptionally coupled response is WL intensity independent, and results in a complete loss of greening ability and an inhibition of HEMA1 and Lhcb gene expression in WL. This response is inhibited by Suc, the phyA mutation, and POR overexpression and requires high fluence rate FRc. 1. The Transcriptionally Uncoupled Response Results from a Perturbation of the PORA:Pchlide Ratio Leading to Increased Photosensitivity to WL | Previous studies on the FR block of greening response have demonstrated that there is an FR-HIR-mediated depletion of PORA in WT seedlings and that this leads to an overaccumulation of nonphotoactive Pchlide . This Pchlide species cannot be reduced to Chlide upon illumination and therefore acts as a potent sensitizer for photooxidative damage to plastids . The combination of photooxidative damage and reduced availability of POR for chlorophyll synthesis leads to a loss of greening ability in WL. Consistent with this interpretation, overexpression of POR has been shown to protect against the FR block of greening by decreasing the amount of nonphotoactive Pchlide present in seedlings prior to the shift to WL . Because the effects of photooxidative damage are dependent on WL intensity, we believe that the transcriptionally uncoupled response characterized in this study corresponds to an increase in photosensitivity through inhibition of POR synthesis as shown previously . One aspect of the block of greening response is that there is an FR-dependent increase in total Pchlide (Fig. ; , ). This would be expected to compound any depletion in POR. Here, we show that the increase in Pchlide results from an FR-HIR-mediated up-regulation of HEMA1 expression and ALA synthesis . The correlation between increased ALA synthesis and loss of greening ability under conditions resulting in the transcriptionally uncoupled FR block of greening response supports the hypothesis that the transcriptionally uncoupled response is a consequence of deregulation of the tetrapyrrole pathway . The transcriptionally uncoupled response was inhibited by the presence of cytokinin. The effect of cytokinin in inhibiting the greening response of FR-treated WT seedlings grown on 3% (w/v) Suc (Fig. D) corresponded to a decrease in the FR-mediated induction of HEMA1 expression (Fig. F). Cytokinin has also been shown to increase PORA levels , and these combined effects could well be sufficient to maintain adequate buffering of Pchlide when exposed to WL. It should also be noted that in addition to blocking the transcriptionally coupled response, Suc slightly delayed the transcriptionally uncoupled response (Fig. A). Again, this corresponded to a delay in the induction of HEMA1 (Fig. D) and in the down-regulation of POR expression . These results are all consistent with the PORA:Pchlide ratio being the determining factor in the transcriptionally uncoupled FR block of greening response. With regard to the physiological significance of these results, the observation here that a 1-d FR pretreatment (or longer irradiations with low-intensity FRc) could enhance the subsequent transcriptional response to low-intensity WL (Fig. C) has some precedence in reports that an FRc pretreatment caused an amplification of the low fluence growth response to a subsequent RL treatment . Stimulation of ALA synthesis in WL through an FR pretreatment has also been noted before (e.g. ). Therefore, it appears that an FR-HIR serves a similar function as a R light pulse (followed by a short darkness incubation period) in priming the autotrophic transition in the emerging seedling, in this case, for example, under dense canopy shade. The Transcriptionally Coupled Response Results in the Loss of Nuclear Gene Expression through an FR-HIR Effect on Plastid Gating of Phytochrome Signaling | The transcriptionally coupled response was characterized by the almost complete loss of HEMA1 and Lhcb expression in WL following transfer from prolonged (>2 d) FR treatments . Because this effect is also seen using the HEMA1 promoter:gusA lines, it clearly results from a loss of promoter-driven transcriptional activity as opposed to transcript stability. Suc , but not cytokinin , inhibited the development of the transcriptionally coupled response. We propose that the loss of nuclear gene expression results from an FR-HIR-mediated inhibition of a plastid signal that gates HEMA1 and Lhcb gene expression. This proposal is based on the similarity of this response to the characteristic loss of nuclear gene expression following NF-induced photobleaching and the ability of POR overexpression to rescue HEMA1 expression . Because POR is localized to plastids in these overexpressors , it is reasonable that its ability to rescue HEMA1 expression results from alterations in plastid function. It is not clear whether the deterioration of the plastid signal in the transcriptionally coupled response occurs directly during the FR irradiation period or subsequently under WL as the result of sensitization by the FR pretreatment. However, our current hypothesis is that plastid signaling is lost during the period of FRc irradiation. This hypothesis is based on the observations that the complete loss of greening ability and nuclear gene expression was independent of WL intensity , but was dependent on the FR fluence rate (Fig. F). Furthermore, there was an additive interaction of an FR pretreatment with NF-induced photobleaching , which suggests that plastid signaling is not completely abolished by either treatment alone. Although it is not possible to say whether these two treatments use the same or independent signaling pathways, these results do suggest that at least some element of the transcriptional block was irreversibly determined during the FR preirradiation period. This is in contrast to the effect of NF-induced photobleaching, which occurs exclusively during the WL irradiation period. A number of events may be occurring during the period of FRc that could lead to the loss of plastid signaling. Previous studies on the effects of FR irradiation on plastid structure have revealed that this treatment leads to a deficiency in the ordered membrane system of the prolamellar body . This can be largely attributed to the dramatic decline in PORA , which is the major protein component of the prolamellar body . In addition, the development of vesicles, which may be related to degradation of plastid components, has also been observed . These ultrastructural changes all occur during the FR treatment, prior to transfer to WL, and seedlings treated with 2 d of darkness following the FR pretreatment do not recover greening ability , indicating that FRc leads to an irreversible arrest of plastid development. It is interesting that Suc, which inhibits the WL-independent loss of nuclear gene expression, also inhibits vesicle formation . It is plausible that this FR-mediated deterioration of plastid structure would compromise plastid/nuclear signaling and is consistent with the observation here of a specific FR-HIR-mediated impairment of nuclear gene expression. Because POR overexpression rescues the observable effects of FRc on plastid ultrastructure , the demonstration that POR overexpression also rescues HEMA1 expression supports the idea that FRc-mediated changes in plastid ultrastructure lead to a loss of plastid signaling. In this context, it may be that cytokinin can mediate plastid repair (sufficient to permit greening) in the event of low-level damage such as within the transcriptionally uncoupled response, but cannot overcome this more severe damage associated with the transcriptionally coupled response. Loss of nuclear gene expression can also be seen following transfer to WL after longer growth periods (>4 d) in darkness . In this experiment, Arabidopsis seedlings showed a progressive loss of the HEMA1 promoter response to light and this was independent of FR and phytochrome. Lhcb and RBCS mRNA levels have previously been reported to be subject to light-independent developmental mechanisms , but it is equally possible that this is simply the result of starvation. The inhibition of this response by 3% (w/v) Suc supports such an idea. The decline in the transcriptional response of seedlings while still under FRc displayed a similar time course to the FR-independent loss of WL induction, and this response was also prevented by the presence of 3% (w/v) Suc . However, the time-dependent loss of transcriptional activity is distinct from that seen following the complete FR block of greening response. First, the use of the GUS reporter showed that an FR pretreatment resulted in a significantly greater level of HEMA1 promoter inhibition under subsequent WL than could be accounted for by an extended dark period (Fig. C). Second, the time-dependent loss of greening ability and promoter responsiveness, as seen in dark-grown seedlings in the absence of Suc, was alleviated by cytokinin, whereas the FR-dependent loss of greening and HEMA1 expression was not (Fig. , A and C). The most likely explanation is that in the absence of Suc or cytokinin, these time- or starvation-dependent effects proceed within FR-irradiated seedlings in parallel with FR-specific responses. In contrast, when seedlings are transferred to R or WL, they become dissociated from such a time-dependent loss of transcriptional activity (Fig. C) through the photosynthetic supply of sugars or direct chloroplast signaling. There is clearly a close link between the transcriptionally uncoupled response proposed to act through a perturbation of the PORA:Pchlide ratio and the transcriptionally coupled response in which it is proposed that irreversible changes in plastid ultrastructure result in a loss of nuclear gene expression. The important role of POR in determining plastid structure and the ability of POR overexpression to rescue both responses, at least in part (Fig. ; ), supports such a hypothesis and suggests that the effects caused by the transcriptionally uncoupled response may have an input into the transcriptionally coupled response . Given the temporal separation of the two FR-mediated responses, it is possible to think of them proceeding sequentially following the onset of FRc, with the transcriptionally uncoupled response leading to the transcriptionally coupled response. However, as cytokinin inhibits the transcriptionally uncoupled response while having no effect on the transcriptionally coupled response, there is clearly no absolute requirement for the transcriptionally uncoupled response for the transcriptionally coupled response to proceed. Implications for Plastid-Nuclear Signaling | In this study, we have identified a plastid-signaling pathway that is affected by prolonged periods of FRc irradiation and is at least partially independent from the pathway inhibited by NF-induced photobleaching. Previous studies have also concluded that more than one plastid-signaling pathway must exist . There are a number of similarities between the FRc dependence of plastid signaling described here and previous observations. For example, treatment of barley (Hordeum vulgare) seedlings with the carotenoid inhibitor Amitrole (but not NF) results in the loss of Lhcb and RbcS expression in low-intensity, nonphotodamaging WL . In a similar manner, inhibition of etioplast development resulting from the reduced accumulation of Pchlide and POR in the phytochrome chromophore-deficient aurea mutant of tomato also leads to reduced Lhcb expression in the darkness . This can also be considered as a WL intensity-independent loss of plastid signaling. The possibility that changes in plastid development through FRc, inhibition of chromophore biosynthesis, and Amitrole treatment all lead to inhibition of the same plastid signaling pathway is intriguing and will require further experiments. One common feature of these conditions is that they all perturb tetrapyrrole biosynthesis . This observation, together with our finding that POR overexpression can sustain this signaling response, supports the growing evidence that tetrapyrroles play a major role in mediating plastid signaling . However, the exact role of tetrapyrroles in these pathways is still unknown and further work is clearly needed to elucidate this. The identification of a new pathway leading to a phytochrome A-dependent loss of plastid signaling will provide a new experimental system for addressing questions on the mechanisms involved in plastid-nuclear signaling. MATERIALS AND METHODS : Plant Material | Transgenic lines expressing a HEMA1promoter:gusA reporter construct were the kind gift of Andrea Fischer and Prof. Dieter Soll (Yale University, New Haven, CT) and have been described previously . Two homozygous lines were used in these experiments, each containing the full-length (-2,435/+252) HEMA1 promoter region fused upstream of the GUS coding sequence. The Arabidopsis WT ecotypes Col and Ler and the phyA (Col and Ler) mutant used in this study were kindly provided by Drs. Haruko Okamoto and Xing-Wang Deng (Yale University). Surface-sterilized seeds were plated onto a 1% (w/v) agar medium containing Murashige and Skoog salts and were supplemented with 0% or 3% (w/v) Suc as indicated in "Results." Plates were placed at 4C in darkness for 2 d prior to receiving a 30-min WL irradiation to synchronize germination. Unless indicated otherwise, seeds were allowed to germinate in darkness at 23C for 2 d (0% [w/v] Suc) or 3 d (3% [w/v] Suc) prior to transfer to the FR light source. Where indicated, NF (kindly provided by Prof. John Gray, University of Cambridge, Cambridge, UK) was added to the medium at a level of 5 mum. For treatments with the cytokinin BAP, seeds were first germinated on sterile filters placed over Murashige and Skoog medium for 1 d in darkness and were then transferred to plates containing 10 mg L-1 (44 mum) BAP. Light Sources | Broad-band WL was provided by white fluorescent tubes (400 --700 nm = 130 mumol m-2 s-1 unless indicated otherwise). This fluence rate was equivalent to high-intensity WL (250 mumol m-2 s-1; Figs. and ) for the greening response. Narrow waveband sources were provided by light-emitting diode displays in environmental control chambers (Percival Scientific, Boone, IA). R light corresponded to a peak at 669 nm (25-nm bandwidth at 50% of peak magnitude) with a fluence rate of 80 mumol m-2 s-1. FR from the light-emitting diodes had a peak at 739 nm (25-nm bandwidth at 50% of peak magnitude) and was passed through two filters (nos. 116 and 172; Lee Filters, Andover, UK) to remove lambdas <700 nm to give a final fluence rate of 10 mumol m-2 s-1. RNA Gel-Blot Analysis | RNA gel-blot analysis was performed using total RNA (30 mug lane-1) exactly as described previously . The HEMA1 probe used was a 3'-cDNA fragment that is gene specific . The Lhcb probe (kindly provided by Dr. Joanne Chory, The Salk Institute, La Jolla, CA) contained the majority of the coding region of the Lhcb1*2 gene and is predicted to crosshybridize with other members of the Lhcb gene family . GUS Fluorometric Analysis | Quantitative assays of GUS activity in seedling cotyledons were conducted exactly as described previously . ALA Biosynthesis Assay | Seedlings were incubated, under the appropriate light conditions, in 100 mm sodium phosphate buffer (pH 7.0), with or without 40 mm levulinic acid (Sigma-Aldrich, Poole, UK), at 23C for 7 h. Seedlings were then blotted dry, frozen in liquid nitrogen, and stored at -80C. The frozen seedlings were ground in 20 mm sodium phosphate buffer (pH 7.0), incubated on ice for 20 min, and centrifuged at 8,500g in a benchtop microfuge for 5 min. The supernatant was incubated with ethylacetoacetate at 100C followed by the addition of modified Ehrlich's reagent . Absorbance was read at 526, 553, and 600 nm and the concentration of ALA was calculated using a molar absorption coefficient of 7.45 x 104m-1 cm-1. Pigment Extraction | Pchlide was extracted based on the method of . Etiolated Arabidopsis seedlings were homogenized in acetone:0.1 m NH4OH (90:10, v/v). The extract was centrifuged at 8,500g for 2 min and the pellet was re-extracted as above. The supernatants were combined and mixed with an equal volume of hexane. The aqueous and hexane fractions were collected separately, and relative fluorescence emission spectra were recorded using a fluorescence spectrophotometer (F-2000; Hitachi, Tokyo) with an excitation wavelength of 440 nm. Greening is shown as the percentage of seedlings with visibly green cotyledons. Comparison with direct measurements of chlorophyll extractions showed that such an assessment was linearly correlated to total chlorophyll within a given genetic background. Backmatter: PMID- 12226520 TI - Short-Term Boron Deprivation Inhibits Endocytosis of Cell Wall Pectins in Meristematic Cells of Maize and Wheat Root Apices AB - By using immunofluorescence microscopy, we observed rapidly altered distribution patterns of cell wall pectins in meristematic cells of maize (Zea mays) and wheat (Triticum aestivum) root apices. This response was shown for homogalacturonan pectins characterized by a low level (up to 40%) of methylesterification and for rhamnogalacturonan II pectins cross-linked by a borate diol diester. Under boron deprivation, abundance of these pectins rapidly increased in cell walls, whereas their internalization was inhibited, as evidenced by a reduced and even blocked accumulation of these cell wall pectins within brefeldin A-induced compartments. In contrast, root cells of species sensitive to the boron deprivation, like zucchini (Cucurbita pepo) and alfalfa (Medicago sativa), do not internalize cell wall pectins into brefeldin A compartments and do not show accumulation of pectins in their cell walls under boron deprivation. For maize and wheat root apices, we favor an apoplastic target for the primary action of boron deprivation, which signals deeper into the cell via endocytosis-mediated pectin signaling along putative cell wall-plasma membrane-cytoskeleton continuum. Keywords: Introduction : Boron is an essential microelement for plant cells. Boron deficiency causes defects in assembly and mechanical properties of cell walls, in structural and functional integrity of the plasma membrane, and in several metabolic and physiological processes (for reviews, see ; ; ; ; ). Nevertheless, it is still not clear which of these responses are of primary and which of secondary nature. Recent research has revealed several rapid reactions to boron removal (within 5 --20 min), including altered cell wall physics, with a transitory decrease of the elasticity modulus epsilon followed by a secondary rehardening, and a reduction of inducible plasma membrane-bound reductase activity (for review, see ). Although rhamnogalacturonan II (RGII) pectins were identified as the major boron-binding fraction in plant cell walls , it is unclear how these rapid structural alterations at the cell wall-plasma membrane interface are related to the RGII-borate complex. In addition to boron cross-linked RGII pectins (e.g. ), the pectin network of calcium cross-linked de-esterified homogalacturonan pectins (e.g. ) is also important for the regulation of mechanical properties of cell walls. In tip-growing pollen tube apices, for instance, pectin epitopes with a relatively high level of methylesterification are abundant and responsible for the loosened nature of cell walls . In muro de-esterification of homogalacturonan pectins in the subapical region of pollen tubes makes them proned for cross-linking by calcium, which increases mechanical strength of cell wall . Intriguingly, boron deprivation caused an immediate halt of the tip growth, and apices of pollen tubes became enriched with JIM5-reactive de-esterified pectins . This implies that boron may affect the extensibility of cell walls not only via cross-linking of RGII pectins but also by affecting the distribution of de-esterified homogalacturonan pectins cross-linked with calcium . Increasing evidence supports the hypothesis that the cytoskeleton-plasma membrane-extracellular matrix continuum represents an essential structural assembly directing the growth and morphogenesis of higher plants . Structural alterations of the pectic matrix, attributable to the removal of boron, are expected to interfere directly and/or indirectly with the cell wall-plasma membrane-cytoskeleton continuum. The extracellular part (apoplast) of this structural continuum has been suggested to represent the essential site of aluminum toxicity . Using anti-actin and anti-tubulin monoclonal antibodies, we revealed increased levels of actin and tubulin proteins upon short-term boron deprivation in roots of hydroponically grown Arabidopsis . This response is also associated with changes of cytoskeletal polymerization patterns in cells of maize (Zea mays) root apices . A possible explanation of these findings could be that these rapid responses of the root cytoskeleton are related to changes in cell wall structure mediated via alterations of cell wall pectins. The aim of this study was to investigate the rapid responses of cell wall pectins to boron deprivation from the emerging perspective of putative cytoskeleton-plasma membrane-extracellular matrix continuum of plant cells. RESULTS : Cell Wall Pectins Become More Abundant in Maize Roots under Boron Deprivation | JIM5 antibody recognizes partially (up to 40%) esterified homogalacturonan pectins . These were localized in cell walls of all root cells with the exception of epidermis and root cap. RGII antibody, which recognizes RGII cross-linked by a borate diol diester , was reactive in cell walls throughout root apices in all plant species under investigation. In the maize root apex, JIM5-reactive pectins accumulated in cell walls facing intracellular spaces, and only a weak signal was associated with cross-walls (Fig. A). On the other hand, RGII-borate pectins were distributed in all cell walls (Fig. B) throughout the maize root apex. Importantly, both RGII- and JIM5-reactive pectins were excluded from the cytoplasm (see also ), which corresponds well with the fact that their epitopes are formed in muro. Deprivation of maize root apices of boron for 1 h was associated with an enhancement of cell wall signal for both of these pectin epitopes (Fig. , C and D). Intriguingly, JIM5-reactive pectins accumulated abundantly also at cross-walls of boron-deprived root apices (Fig. C), which were almost devoid of these pectins in control root apices (Fig. A). Figure 1 | Localization of cell wall pectins in control and boron-deprived maize root apices. Localization of cell wall pectins in control and boron-deprived maize root apices. Immunofluorescence microscopy revealed that boron deprivation (1 h) induced alterations of cell wall pectins, reactive to JIM5 (A and C) and to RGII (B and D) antibodies, in cells of maize root apices. A and B, Control images taken from the cortex document that the signal is preferentially associated with cell walls. Especially longitudinal cell walls are strongly labeled with JIM5 antibody (A), whereas all cell walls are labeled with RGII antibody (B). Distribution patterns of de-esterified homogalacturonan pectins (responsive to JIM5 antibody) change under boron deprivation (1 h). C, Besides slightly stronger signal throughout root apices, cross walls of boron-deprived cells are also strongly labeled with JIM5 antibody. Similarly, the RGII-reactive signal is enhanced in cells deprived of boron (D) when compared with the control (B). Bars = 25 mum in A and C, 22 mum in B, and 15 mum in D. Boron Deprivation Inhibits Internalization of Cell Wall Pectins in Maize and Wheat (Triticum aestivum) Roots | Brefeldin A (BFA) is very useful drug for visualization of internalizing and recycling molecules because it blocks exocytosis but still allows endocytotic internalization of extracellular molecules such as RGII- and JIM5-reactive cell wall pectins . The internalization of JIM5-reactive cell wall pectins occurred mainly in cortical cells of the meristem (Fig. , A and B). On the other hand, stele cells formed only few and smaller BFA compartments (Fig. A). Note that stele cells above quiescent center are devoid (Fig. A, star), similar to the epidermis and root cap cells (Fig. A, asterisk), of any signal in response to JIM5 antibody. Figure 2 | Effects of BFA on localization of cell wall pectins in maize root apices. Effects of BFA on localization of cell wall pectins in maize root apices. Effects of BFA on distributions of JIM5- and RGII-reactive pectins in cells of maize root apices. A, JIM5-reactive pectins are absent from stele cells above quiescent center (star) and epidermis/root cap cells (asterisk). A and B, Cortical cells internalize JIM5-reactive pectins as documented by large and clear BFA compartments. C, The internalization of JIM5-reactive pectins was reduced as a result of 5 h of boron deprivation. D, Active internalization of RGII-reactive pectins in cortical and particularly in stele cells, as evidenced by intracellular aggregates in the presence of BFA. E, Note that especially phloem cells are active in internalization of RGII-reactive pectins. F, Internalization of RGII-reactive pectins was completely blocked upon 5 h of boron deprivation. Bars = 35 mum in A, 16 mum in B and C, and 25 mum in D through F. In contrast, RGII-borate pectins were internalized in all meristematic cells exposed to BFA (Fig. D). Dramatic accumulation of RGII-borate pectins within BFA compartments was observed especially in phloem cells (Fig. E). The internalization of RGII-borate pectins in maize was completely blocked after 5 h of boron deprivation (Fig. F). In contrast to RGII pectins, internalization of JIM5 pectins in maize was not blocked completely under boron deprivation, although its magnitude was highly reduced (Fig. , C versus B). In cells of wheat root apices, internalization of both JIM5-reactive (Fig. , A and B) and RGII-reactive (Fig. , C and D) pectins was blocked completely under boron deprivation (Fig. , A --D). Figure 3 | Localization of cell wall pectins in control and boron-deprived wheat root apices. Localization of cell wall pectins in control and boron-deprived wheat root apices. Distributions of JIM5-reactive (A and B) and RGII-reactive (C and D) pectins in BFA-treated root apices of wheat. Note that boron deprivation (B and D) blocks completely cytoplasmic internalization of these cell wall pectins, which accumulate abundantly within prominent BFA compartments in the presence of boron (A and C). Bars = 5 mum in A, 15 mum in B, 12 mum in C, and 22 mum in D. Cell Wall Pectins Do Not Internalize in Cells of Zucchini (Cucurbita pepo) and Alfalfa (Medicago sativa) Roots | Intriguingly, this rapid internalization of cell wall pectins reactive to JIM5 (Fig. , A, B, E, and F) and RGII (Fig. , C, D, G, and H) antibodies was not detectable in BFA-treated alfalfa (Fig. , A --D) and zucchini (Fig. , E --H) root apices. Some pectin-positive BFA compartments were occasionally found in epidermis cells of both alfalfa and zucchini root apices (data not shown). Importantly, we did not detect strongly increased amounts of these pectins in their control cell walls (Fig. , A, C, E, and G) when boron was deprived (Fig. , B, D, F, and H), although a slight increase was visible in alfalfa root apices (Fig. , B and D). Figure 4 | Localization of cell wall pectins in alfalfa and zucchini root apices. Localization of cell wall pectins in alfalfa and zucchini root apices. Distributions of JIM5-reactive (A, B, E, and F) and RGII-reactive (C, D, G, and H) pectins in BFA-treated root apices of alfalfa (A --D) and zucchini (E --H) root apices. Note that there are no BFA compartments formed in these dicot species. In contrast to maize and wheat, boron deprivation (B, D, F, and H) does not prominently change the abundance of these pectins in cell walls of alfalfa and zucchini root apices. Bars = 10 mum in A and B, 15 mum in C and D, 20 mum in E and F, 15 mum in G, and 10 mum in H. DISCUSSION : Eukaryotic cells perform continuous recycling of the plasma membrane proteins and extracellular matrix molecules from the cell surface back to the cytoplasm (for plant cells, see ; ). Our recent study provides experimental evidence that cell wall pectins are internalized after in muro de-esterification and cross-linking with calcium and boron . These almost exclusive cell wall pectin epitopes, reactive to JIM5 and RGII antibodies, accumulate abundantly within intracellular BFA-induced compartments, which are obviously formed through aggregation of trans-Golgi networks and putative plant endosomes . Here, we report that intracellular BFA compartments accumulate these cell wall pectins in meristematic cells of maize and wheat, but not of zucchini and alfalfa, root apices. Intriguingly, boron deprivation inhibits endocytosis of cell wall pectins. These findings implicate differences between pectin turnover in cell walls of graminaceous monocots and dicots. Interestingly in this respect, root apices of gramineae are relatively insensitive to boron deprivation. Internalization of these cell wall pectins must have profound effects on wall structure and porosity. For instance, the RGII-borate dimer, being the antigen of RGII antibody , has been convincingly shown to play a crucial role in both plant growth and development . Comparative analysis of mur1 and mur2 mutants of Arabidopsis implies that pectins, instead of xyloglucan, are the most important partner of cellulose microfibrils in rendering strong mechanical properties of plant cell walls . Tip-growing pollen tubes, which have high-pectin content in their apical cell walls, might be expected to rely on cross-linkings of pectins with boron and calcium . In accordance with this concept, tip-growing pollen tubes burst at their apices under boron deficiency (e.g. ). Interestingly, tip-growing root hairs show a high capacity to bind fluorescein isothiocyanate (FITC)-phenyl boronic acid, indicative of high boron-binding capabilities, at their expanding apices . In extending root hair apices, initiation of Rhizobium sp. bacteria versus legume plants interactions starts with formation of infection threads growing inwardly into root hairs. Growth of infection threads through root tissues bears a histological similarity to pollen tubes growing through pistil tissues. Interestingly, growth of infection threads is arrested under boron deficiency . Later in the Rhizobium sp. host plant interactions, boron obviously plays an important role in mediating surface interactions between Rhizobium sp. bacteria and root cortex cells that lead to endocytosis-like internalization of bacteria by host cells during the nodule development . demonstrated a dramatic increase of homogalacturonan cell wall pectins, reactive to JIM5 antibody , at tips of pollen tubes suffering from boron deficiency. A similar response was demonstrated in pollen tubes exposed to Yariv reagent , a compound that cross-links plasma membrane-associated arabinogalactan proteins as evidenced by altering lateral diffusion of plasma membrane proteins . These findings would indicate that, as with the absence of boron, the exposure to Yariv reagent also inhibits internalization of JIM5-reactive cell wall pectins. What are further physiological consequences of reduced or even completely blocked internalization of cell wall pectins in boron-deprived root apices? Boron obviously affects size of pores within cell walls . Moreover, it is known that a high-plasma membrane tension, which may rapidly build up under boron deficiency because of the initial weakness of the cell wall challenged with a high turgor pressure , generally inhibits endocytosis . This lead us to speculate that cell wall boron might regulate internalization of RGII- and JIM5-reactive acidic pectins through interference with receptor-mediated endocytosis. One characteristic hallmark of receptor-mediated endocytosis in eukaryotic cells is that hypertonic media inhibit effectively this process by blocking clathrin-coated pit formation . Further hallmarks of receptor-mediated endocytosis are its dependence on temperatures above 4C (for plants, see ; ) and on an intact F-actin cytoskeleton . In our accompanying paper , we showed that endocytosis of cell wall pectins is blocked upon cold treatment and after depolymerization of F-actin. Importantly, pectin-derived elicitors were reported to be taken up into plant cells via receptor-mediated endocytosis . In fact, endocytosis of cell wall pectins might be part of pectin-mediated signaling between the plasma membrane and nucleus . Interestingly in this respect, boron was suggested to function at "membrane rafts" , which are known to act as platforms for signaling and endocytosis in cells of multicellular eukaryotic organisms (; ; for plants, see ). All of this gives us a new fresh look on still enigmatic roles of boron in plant cells. MATERIALS AND METHODS : Plant Growth | Maize (Zea mays), wheat (Triticum aestivum), alfalfa (Medicago sativa), and zucchini (Cucurbita pepo) seeds were soaked in distilled water for 8 h with aeration and germinated in moistened rolls of filter paper for 2 d in dark at 24C (maize) or 28C (zucchini). Uniform seedlings with straight primary roots were transferred to solution culture, which was maintained in a phytotrone at 24C under 14 h of light (400 mumol m-2 s-1) and 70% humidity. The nutrient solution contained 2,000 mum Ca(NO3)2, 3,000 mum KNO3, 1,000 mum KH2PO3, 500 mum Mg(NO3)2, 100 mum NaCl, 1,000 mum MgSO4, 44.8 mum FeEDDHA, 18.2 mum MnSO4, 3.1 mum CuCl2, 6.1 mum ZnSO4, 0.2 mum (NH4)6Mo7O24, 0.016 mum CoCl2, 0.017 mum NiCl2, and 40 mum H3BO3. All the nutrient stock solutions were prepared with Milli-Q ultrapure water that was deprived of boron with B-specific ion-exchange resin Amberlite IRA-743 (Sigma, St. Louis). The pH of the nutrient solution was adjusted to 6.0 by addition of 1 n NaOH. Seedlings were allowed to grow for 2 d in a one-tenth-strength nutrient solution with 15 min of aeration every 45 min and then in a full-strength nutrition solution for a further 2 d. Boron Deprivation Treatment | After transplanting, seedlings were transferred to a fresh nutrient solution with (+B) or without boron (-B). In the -B treatment, an ample amount of the borate-specific exchange resin Amberlite IRA 743 was also placed within the nutrient container, and the nutrient solution was constantly stirred to adsorb virtually all possible B contaminations. The roots were rinsed in a 5-L -B nutrient solution for a few seconds and then allowed to grow in -B solution for 10, 30, and 60 min and 5 h. When the internalization of cell wall pectins was studied, seedlings were grown further for 2 h in the +-B nutrition solution containing 100 mum BFA. Indirect Fluorescence Microscopy | Indirect fluorescence microscopy was performed essentially according to with a few modifications. Apical segments of primary root from +B and -B-treated seedlings were vacuum infiltrated for 10 min with 3.7% (w/v) formaldehyde made up in stabilizing buffer (SB; 50 mm PIPES, 5 mm EGTA, and 5 mm MgSO4, pH 6.9) and then fixed at room temperature for 1 h. After a brief rinse in SB, they were dehydrated in a graded ethanol series diluted with phosphate-buffered saline (PBS; pH 7.3) and embedded in Steedman's wax (for further details, see ). Eight-micrometer-thick longitudinal sections were prepared from the embedded materials, and the most median sections were allowed to expand on a small drop of distilled water onto slides coated with glycerol-albumen (Serva, Heidelberg). After drying at room temperature overnight, the mounted sections were dewaxed in ethanol, rehydrated in a PBS-diluted ethanol series, and then left in SB for 30 min before being rinsed with methanol at -20C for 20 min. The sections were transferred to SB for 30 min and incubated with the following primary antibodies for 1 h at room temperature: the monoclonal anti-acidic pectin antibody (JIM5) raised against rat and the polyclonal anti-RGII-boron-complex antibody raised against rabbit . All the primary antibodies were diluted in PBS supplemented with 0.1% (w/v) bovine serum albumin. After a rinse in SB, the sections were incubated with FITC-conjugated anti-mouse IgG (Sigma) and diluted 1:100 (w/v) in PBS containing 0.1% (w/v) bovine serum albumin. The labeled sections, after being rinsed with PBS and further stained with 0.01% (w/v) toluidine blue to diminish autofluorescence from root tissue, were mounted using an antifade mountant containing p-phenylenediamine . Fluorescence was detected with an Axiovert 405M inverted microscope (Zeiss, Oberkochen, Germany) equipped with epifluorescence and standard FITC exciter and barrier filters (BP 450 --490, LP520). Photos were taken on T-Max films rated at 400 ASA (Eastman Kodak, Rochester, NY). Distribution of Materials | Upon request, all novel materials described in this publication will be made available in a timely manner for noncommercial research purposes, subject to the requisite permission from any third-party owners of all or parts of the material. Obtaining any permissions will be the responsibility of the requestor. Backmatter: PMID- 12226521 TI - F-Actin-Dependent Endocytosis of Cell Wall Pectins in Meristematic Root Cells. Insights from Brefeldin A-Induced Compartments AB - Brefeldin A (BFA) inhibits exocytosis but allows endocytosis, making it a valuable agent to identify molecules that recycle at cell peripheries. In plants, formation of large intracellular compartments in response to BFA treatment is a unique feature of some, but not all, cells. Here, we have analyzed assembly and distribution of BFA compartments in development- and tissue-specific contexts of growing maize (Zea mays) root apices. Surprisingly, these unique compartments formed only in meristematic cells of the root body. On the other hand, BFA compartments were absent from secretory cells of root cap periphery, metaxylem cells, and most elongating cells, all of which are active in exocytosis. We report that cell wall pectin epitopes counting rhamnogalacturonan II dimers cross-linked by borate diol diester, partially esterified (up to 40%) homogalacturonan pectins, and (1->4)-beta-d-galactan side chains of rhamnogalacturonan I were internalized into BFA compartments. In contrast, Golgi-derived secretory (esterified up to 80%) homogalacturonan pectins localized to the cytoplasm in control cells and did not accumulate within characteristic BFA compartments. Latrunculin B-mediated depolymerization of F-actin inhibited internalization and accumulation of cell wall pectins within intracellular BFA compartments. Importantly, cold treatment and protoplasting prevented internalization of wall pectins into root cells upon BFA treatment. These observations suggest that cell wall pectins of meristematic maize root cells undergo rapid endocytosis in an F-actin-dependent manner. Keywords: Introduction : Eukaryotic cells perform endomembrane flow accomplished by vesicles shuttling among endoplasmic reticulum (ER), Golgi apparatus (GA), the plasma membrane (PM), and endosomes (for plants see ; ). These compartments and pathways of endomembrane flow are highly conserved in unicellular yeast, higher plants, and animals (for plant cells, see ; ). A major breakthrough in our current understanding of this complex endomembrane flow was provided by rediscovery of the fungal metabolite brefeldin A (BFA; ). BFA action prevents vesicle formation in the exocytosis pathway by stabilizing abortive complexes between conserved ADP ribosylation factor 1 (ARF1) and the Sec7 domain of its guanine nucleotide exchange factor during the assembly of coat protein complexes of budding vesicles (for plants see, ; ). Because of this action, BFA inhibits anterograde vesicular pathways while allowing endocytosis and some retrograde pathways to proceed further . Moreover, BFA inhibits the endosome to vacuole transport in budding yeast . The introduction of BFA to investigate the cell biology of endomembrane flow in plant cells occurred some years later , but most of the major findings concerning the effects of this drug in animal and yeast systems have been confirmed for plant cells. For example, low concentrations of BFA (<10 mug mL-1) effectively inhibit secretion attributable to blockage of the ER to GA step , whereas higher BFA levels (>50 mug mL-1) are needed to inhibit the GA to PM transport step and to induce vesiculation of GA stacks . This is accompanied by redistribution of GA proteins into ER . Similar to other eukaryotic systems, endocytosis remains intact in BFA-treated plant cells . Before the formation of GA-ER hybrid organelle, the trans-most GA cisterna is lost and apparently contributes to the formation of perinuclear vesicular bodies . Such compact areas of densely packed heterologous vesicles were named "BFA compartments" , and they represent the most dramatic morphological response of plant cells to high BFA levels. Importantly, BFA compartments form before GA disintegration , and ER elements do not participate in the formation of BFA compartments . These observations strongly suggest that there must be an additional membranous source that feeds into these compartments of plant cells. Our data suggest that this source is the PM. The nature of BFA compartments remains controversial also because several other studies failed to report such compartments, even in plant cells that have their GA totally disassembled in response to BFA treatment . In our study, we have addressed these issues using an embedding-sectioning technique based on Steedman's wax , which enables reliable development- and tissue-specific localization of diverse intracellular (for tubulin and actin, see , ) and cell wall (; this study) antigens in the context of intact root apices. Taking the advantage of BFA and pectin antibodies reactive to cell wall pectin epitopes , we report that cell wall pectins represent the first complex macromolecules that are shown to be internalized into the cytoplasm of meristematic plant cells. RESULTS : Development- and Tissue-Specific Distributions of BFA Compartments | Rhamnogalacturonan II (RGII) antibody recognizes cell wall rhamnogalacturonan pectins cross-linked by a borate diol diester formed within cell walls in muro . We have taken advantage of this antibody to probe whether cell wall pectins accumulate within intracellular BFA compartments. We report that the RGII antibody labeled predominantly cell walls in control root apices (Fig. A). Importantly, in BFA-treated roots, RGII antibody recognized prominent BFA compartments in all meristematic root cells (Fig. , B --D). In contrast, RGII-positive BFA compartments did not form in cells having active exocytosis like post-mitotic metaxylem elements (Fig. C) and secretory root cap cells (Fig. D). Interestingly, BFA compartments ultimately achieved perinuclear positions and maintained the longest distance from each other at the opposite sides of centrally positioned nuclei matching cellular polarity axes (Fig. , E and F). In the elongation region, only epidermal cells embarking on root hair formation formed BFA compartments (Fig. F), whereas all other rapidly elongating cells did not form BFA compartments (Fig. G). Figure 1 | Development- and tissue-specific distributions of RGII-borate pectins in cells of control (A) and BFA-treated (B --F) maize root apices. Development- and tissue-specific distributions of RGII-borate pectins in cells of control (A) and BFA-treated (B --F) maize root apices. A, RGII-borate pectins localize preferentially to cell walls of all cells of the root apical meristem. B, In BFA-treated roots, all meristematic cells accumulate RGII-borate pectins within BFA compartments. The only exceptions to this feature are metaxylem elements (asterisk in C) and secretory cells of the root cap periphery (asterisk in D). Prominent BFA compartments are found in epidermis cells in the meristem (E) and in the apical part of the elongation region (F). Note the switch in positioning of BFA compartments in post-mitotic epidermis cells in E and F. In contrast, all other elongating root cells are devoid of BFA compartments; for cortical cells see G. The basical-apical root axis of each cell (in this and all other figures) runs from the top to bottom of the page. Stars indicate nuclei. Bar = 10 mum in A and G; 46 mum in B through D; 12 mum in E; and 15 mum in F. Epitope-Specific Accumulation of Cell Wall Pectins within BFA Compartments | We have tested further cell wall pectin epitopes for their localization in control and BFA-treated root apices. Partially esterified (up to 40%) homogalacturonan pectins are recognized by monoclonal JIM5 antibody (; for an overview of antibodies used, see Table ). In control root apices, JIM5-reactive pectins located exclusively to cell walls of cortex cells, whereas the signal in stele cells was rather faint (Fig. A). Root cap and epidermal cells did not react with JIM5 antibody (data not shown). This situation changed after exposure to BFA, where JIM5-reactive pectins accumulated within BFA compartments in all JIM5-positive cells of the apical meristem (Fig. B). At variance with meristematic cells, BFA-treated elongating cells retained JIM5-reactive pectins within their cell walls (Fig. C). LM5 antibody recognizes (1->4)-beta-d-galactan side chains of rhamnogalacturonan I (RGI; ; see Table ) and labeled predominantly cell walls in control cells (Fig. D). In contrast, labeling by this antibody was almost exclusively confined to large BFA compartments in BFA-treated cells (Fig. E). Importantly, cell wall labeling diminished considerably after accumulation of LM5-reactive pectins within intracellular BFA compartments (Fig. , D and E). In contrast, non-blockwise de-esterified homogalacturonan pectins reactive to LM7 monoclonal antibody , which also labeled exclusively cell walls in control root apices (Fig. F), did not accumulate within BFA compartments (Fig. , F and G). Table I | Summary of antibodies used and localization of their antigens Figure 2 | Distributions of JIM5- (A --C), LM5- (D and E), and LM7-reactive (F and G) pectins in control (A, D, and F) and BFA-treated (B, C, E, and G) root apices. Distributions of JIM5- (A --C), LM5- (D and E), and LM7-reactive (F and G) pectins in control (A, D, and F) and BFA-treated (B, C, E, and G) root apices. B and C, In BFA-treated root apices, JIM5-reactive pectins accumulate within BFA compartments in all meristematic cells (B) but not in elongating cells (C). D and E, LM5-reactive pectins redistribute almost completely from cell walls (D) into BFA-induced compartments (E). F and G, In contrast, LM7-reactive pectins do not accumulate within BFA compartments and remain in cell walls also in BFA-treated cells. Bar = 20 mum in A and B; 40 mum in C; and 11 mum in D through G. Actin Filaments Are Essential for Internalization of Cell Wall Pectins | Root cells devoid of F-actin because of their exposure to latrunculin B failed to internalize JIM5-reactive cell wall pectins and to accumulate them within BFA compartments (Fig. , A and B). In contrast, oryzalin-treated cells devoid of microtubules formed normal or sometimes even slightly larger JIM5-reactive BFA compartments (Fig. D) than those found in control root cells (Fig. C). Figure 3 | Effects of latrunculin B (A and B) and oryzalin (D) on accumulation of JIM5-reactive pectins within BFA compartments. Effects of latrunculin B (A and B) and oryzalin (D) on accumulation of JIM5-reactive pectins within BFA compartments. A and B, Cells devoid of F-actin in latrunculin B pretreated root apices do not internalize JIM5-reactive pectins both in cortex (A) and stele (B). C and D, Cells of oryzalin pretreated roots (D) form even slightly larger JIM5-positive BFA compartments than cells treated only with BFA (C). Bar = 25 mum. Distributions of GA-Derived Exocytotic Pectins in Control and BFA-Treated Cells | JIM7-reactive (up to 80% esterified homogalacturonan) pectins localized to intracellular spots (Fig. A) representing presumably GAs where homogalacturonan pectins are synthesized in their esterified form . In control root cells, pectins reactive to JIM7 antibody did not associate abundantly with cell walls. This characteristic distribution pattern changed only slightly in cells of BFA-treated roots when accumulation of JIM7-reactive spots into a small number of larger aggregates was scored occasionally (Fig. B). Importantly, typical compact, roundish, and large BFA compartments were never detected with the JIM7 antibody suggesting that GA is not the major source of membranous structures accumulating within BFA compartments. Figure 4 | Distribution of JIM7-reactive pectins (A and B), GA-derived AGPs (C and D), ER-based HDEL proteins (E), and PM-associated recycling proteins (F through H) in control (A) and BFA-treated (B through H) cells. Distribution of JIM7-reactive pectins (A and B), GA-derived AGPs (C and D), ER-based HDEL proteins (E), and PM-associated recycling proteins (F through H) in control (A) and BFA-treated (B through H) cells. In control cells, JIM7 antibody recognizes numerous spots distributed throughout the cytoplasm (A), and this pattern does not change dramatically in BFA-treated cells (B). C and D, Secretory AGPs reactive to MAC207 (C) and LM2 (D) antibodies and ER-based HDEL proteins (E) do not accumulate within BFA compartments of BFA-treated root cells. F through H, In contrast, both PM-H+-ATPase (F and G) and PIN1 auxin efflux carrier (H) accumulate abundantly within BFA compartments. Stars indicate nuclei. Bar = 20 mum in A through E; 11 mum in F and H; and 23 mum in G. GA and ER Are Not Major Contributors to Formation of BFA Compartments | To further explore possible contributions of GA and ER elements to the formation of BFA compartments, we used two arabinogalactan protein (AGP) antibodies (MAC207 and LM2) that label the endomembrane system of maize (Zea mays) root cells . Final stages of AGP synthesis occur in GA, after which these complex molecules are secreted into the extracellular space. As a visual marker for ER elements, we used the HDEL antibody (; Table ). Importantly, these GA- and ER-related molecules did not accumulate within BFA compartments of maize root cells (Fig. , C --E). These findings confirm that both the GA and ER do not contribute substantially to the formation of BFA compartments in maize root cells (for similar data, see ). Recycling PM Proteins Accumulate within BFA Compartments | To look for other molecules accumulating within BFA-induced compartments, we have probed subcellular distributions of PM-associated proteins that are expected to perform recycling. Consistent with our expectations, antibodies raised against both the PM H+-ATPase and PIN1 auxin-efflux carrier labeled the PM in control cells (data not shown) and BFA compartments in treated cells (Fig. , F --H). The PM H+-ATPase (Fig. , F and G) and the PIN1 auxin efflux carrier (Fig. H) accumulated within BFA compartments after 2 h exposure to BFA (for similar results in Arabidopsis, see ). Depolymerization of F-actin with latrunculin B inhibited accumulation of these recycling PM proteins within BFA compartments (data not shown). ARF1, But Not cis-Golgi Marker, Accumulates within BFA Compartments | In addition, we have analyzed distributions of two Golgi-associated molecules in BFA-treated maize roots. ARF1, a small GTPase of the Ras family , is the actual target of BFA action in eukaryotic cells localized also to TGN (for maize, see ). In cells of maize root meristem, ARF1 localized diffusely throughout the cytoplasm but also to the PM. Upon BFA treatment, ARF1 accumulated within BFA compartments and got depleted from the cytoplasm and the PM (Fig. , A and B). In contrast, cis-Golgi marker 58-K protein did not accumulate in BFA compartments (Fig. , C and D). Figure 5 | Distribution of ARF1 (A and B) and 58K (C and D) GA proteins in control (A and C) and BFA-treated (B and D) cells. Distribution of ARF1 (A and B) and 58K (C and D) GA proteins in control (A and C) and BFA-treated (B and D) cells. Although ARF1 accumulates prominently within BFA compartments (B), 58K antibody labels spots distributed throughout the cytoplasm that correspond to GA (C), and this pattern does not change dramatically in BFA-treated cells (D). Stars indicate nuclei. Bar = 8 mum in A; 12 mum in B; and 17 mum in C and D. Protoplasting and Cold Treatment Prevent Accumulation of Cell Wall Pectins within BFA Compartments | To provide further experimental evidence for endocytosis of cell wall pectins, we exposed wall-less protoplasts of meristematic maize root cells to BFA. As predicted, we did not score any intracellular accumulation of RGII-borate pectins and LM5-reactive cell wall pectins in BFA-treated protoplasts (Fig. , A --F). In the case of JIM5-reactive pectins, we occasionally found small intracellular aggregates distributed throughout the protoplasts that, however, never coalesced into large BFA compartments (data not shown). Figure 6 | Protoplasts of maize root apex cells do not accumulate cell wall pectins in BFA compartments. Protoplasts of maize root apex cells do not accumulate cell wall pectins in BFA compartments. JIM5-reactive pectins (A and B), LM5-reactive pectins (C and D), and RGII-borate pectins (E and F) show the same distribution pattern both in control (A, C, and E) and BFA-treated (B, D, and F) protoplasts. Bar = 20 mum in A through E; and 34 mum in F. Additional evidence in favor of the active endocytotic internalization of cell wall pectins was provided using cold treatment of maize roots. A hallmark of clathrin-supported endocytosis is its sensitivity to low temperature . BFA treatment at low temperature prevented intracellular internalization of cell wall pectins in all root meristem cells (Fig. , A --D). Low temperature similarly blocked accumulation of PM H+-ATPase and PIN1 auxin-efflux carrier within BFA compartments (data not shown). Figure 7 | Cold treatment (B and D) prevents internalization of cell wall pectins into BFA compartments (for BFA treatment at room temperature, see A and C). Cold treatment (B and D) prevents internalization of cell wall pectins into BFA compartments (for BFA treatment at room temperature, see A and C). A and B, RGII-borate pectins; C and D, JIM5 pectins. Stars indicate nuclei. Bar = 8 mum in A; 10 mum in B; 12 mum in C and D. DISCUSSION : We report here that BFA compartments do not form in all root apex cells. Intriguingly, active secretory cells, like elongating root body cells and root cap periphery cells, do not form characteristic BFA compartments. The only exceptions to this rule are epidermal cells embarking on root hair formation. In contrast, all meristematic root cells formed BFA compartments, which ultimately achieved perinuclear positions. Surprisingly, cell wall pectins accumulated in these large compartments. In addition to cell wall pectins, PM-associated proteins undergoing internalization-mediated recycling also accumulated within BFA compartments. Therefore, the most plausible explanation for our present data is that membranous structures accumulating in characteristic BFA compartments are predominantly of endocytotic/recycling origin. Two major implications arise from our present data. First, it is important to be aware that not all intracellular pectins belong to the GA-derived secretory pathway. Some are clearly transported along endocytotic pathways heading for either recycling or degradation. Second, it is apparent that internalization of cell wall pectins must play a key role in the dynamic turnover of pectins in dividing cells of higher plants. Importantly, we document internalization of those pectin molecules which can be cross-linked with boron (RGII pectins) and calcium (RGII-borate pectins and de-esterified homogalacturonan pectins). BFA Compartments as Endosome-trans-Golgi Network (TGN) Hybrid Organelle | Our data show that BFA compartments accumulate internalized macromolecules via retrograde endocytotic pathway. However, TGN also appears to be involved in the formation of BFA compartments. The most trans-cisterna of plant GA, corresponding to TGN, is rapidly lost via sloughing in response to BFA, and presumably participates in the formation of BFA compartments (see Fig. 12 in ). In accordance with this notion, JIM84 antibody recognizes a complex carbohydrate epitope generated late in the GA pathway that does not relocate back into ER during BFA treatment. Instead, the JIM84 antigen ends up within BFA compartments together with the actual BFA target ARF1 . In contrast to TGN, cis- and median-Golgi cisternae merge with ER as reported for BFA-treated tobacco (Nicotiana tabacum) cv Bright Yellow-2 cells . Strong support for this concept, namely that the trans-most (TGN) cisterna of plant GA aggregates together with putative endosomes to form large BFA compartments, is obtained from studies using targeting of mammalian alpha-2,6-sialyltransferase into plant GA . This protein localizes exclusively to the trans-most (TGN) cisterna of untreated transgenic Arabidopsis root cells but accumulates within BFA compartments of BFA-treated Arabidopsis meristematic root tip cells . Accumulation of alpha-2,6-sialyltransferase within BFA compartments of root cells is apparently dependent on their meristematic nature because post-mitotic leaf cells redistribute the same trans-Golgi marker rather into ER . This is in a full agreement with our present finding that characteristic BFA compartments form only in meristematic cells. Furthermore, the dynamin-like protein ADL6 localizes to trans-Golgi and to BFA compartments in Arabidopsis root cells . In contrast, antibody raised against a cis-Golgi marker 58K protein , which recognizes plant GA , does not label BFA compartments of maize root cells but is presumably associated with the Golgi-ER hybrid organelle described by . Cell Wall Pectins, But Not GA and ER Molecules, Accumulate within BFA Compartments | We have compared distributions of JIM7- and JIM5-reactive homogalacturonan pectins in cells of control and BFA-treated root apices. In accordance with data from tobacco pollen tubes , BFA compartments accumulate large amounts of low-esterified (up to 40%, JIM5) but not high-esterified (up to 80%, JIM7) homogalacturonan pectins. Importantly, besides JIM5-reactive pectins, RGII dimers cross-linked by a borate diol diester and (1->4)-beta-d-galactan side chains of RGI are further pectin epitopes which accumulate within BFA compartments. As boron-cross-linked RGII pectins of cell walls are critical not only for cell wall integrity but, in due course, also for cell growth and overall plant form , their internalization might be expected to have profound impacts on growth and development of plants. Intriguingly in this respect, JIM5-reactive pectins , LM5-reactive pectins , and RGII pectins cross-linked by a borate diol diester all localize preferentially at the innermost part of cell walls adjacent to the PM and undergo internalization. In contrast, non-blockwise de-esterified homogalacturonan pectins of cell walls, reactive to LM7 antibody, do not localize close to the PM and are not internalized into the cytoplasm of dividing cells (this study). Importantly, those cell wall pectin epitopes that are internalized become depleted from walls of BFA-treated root apices. To demonstrate that endocytosis is involved in internalization of these cell wall pectins, we showed that disintegration of F-actin inhibits internalization of cell wall pectins. Further evidence that cell wall pectins are internalized via endocytosis was provided by performing the BFA treatment at 4C when active processes like endocytosis are blocked . At this temperature, no accumulation of wall pectins within BFA compartments was observed. Finally, protoplasting of meristematic root cells prevented formation of pectin-enriched BFA compartments. In contrast to extracellular wall pectins, GA-derived JIM7-reactive pectins did not accumulate within typical compact BFA compartments but instead localized into smaller irregular aggregates corresponding presumably to the pleiomorphic GA-ER hybrid organelle pervading the whole cytoplasm . Importantly in this respect, ER-based proteins did not accumulate in BFA compartments (see also ). Antibodies (MAC207 and LM2) raised against cell wall- and PM-associated epitopes of AGPs, secreted via the exocytotic pathway , similarly did not label BFA compartments of root cells (at least not after 2 h of BFA treatment). These observations argue against the concept that GA-derived vesicles represent the major constituents of BFA compartments. F-Actin, But Not Microtubules, Is Essential for Cell Wall Pectin Internalization | An intact F-actin cytoskeleton is required for endocytosis (; for plant cells, see ; ). Depolymerization of F-actin with latrunculin B interfered with endocytotic internalization of PM/cell wall-associated molecules. JIM5-reactive cell wall pectins and PM-associated proteins, such as PM-H+ ATPase and auxin efflux carrier PIN1, failed to be internalized in the absence of F-actin. These results support the notion that accumulation of internalized pectins within BFA compartments results from the unbalanced recycling of vesicles in BFA-treated cells. In contrast to F-actin, depolymerization of microtubules did not inhibit endocytosis of cell wall pectins. Indeed, the opposite appeared to be the case in that BFA compartments seemed larger in the absence of cortical MTs. This observation might be explained by the dense arrays of cortical microtubules in plant cells (for maize root cells see ) sterically interfering with the assembly of the endocytic protein complexes at the PM. For instance, the distance between neighboring cortical microtubules is much smaller than the size of coated vesicles of plant cells . Thus, a PM devoid of a dense cortical microtubule array might be expected to perform more internalization events. Recycling PM Proteins Accumulate within BFA Compartments | To further substantiate the idea that endocytosis-driven internalization/recycling contributes significantly to the formation of BFA compartments, we have taken advantage of two well-defined antibodies against PM proteins known, or expected, to perform recycling at the PM. Importantly, PIN1 auxin efflux carrier accumulates rapidly within BFA compartments in cells of Arabidopsis embryos . More recent studies on PIN1 and PIN3 demonstrated that BFA-induced accumulation of PIN1/PIN3 within BFA compartments results from unbalanced endocytosis of a steady-state pool of these molecules that rapidly recycle between the PM and endosomal compartment . The identity of this endosome compartment awaits further experimental studies but might be expected to be an early and/or recycling endosome. Here, we have shown that both PM H+-ATPase and PIN1 auxin efflux carriers accumulate within BFA compartments of maize root cells. As recycling PM proteins accumulate within BFA compartments, it might be expected that the internalized cell wall pectins also accomplish recycling. One plausible scenario would be that that boron- and calcium-cross-linked cell wall pectins lose their cross-linkages within endosomes and then return back to cell walls. CONCLUSIONS : In the present study, we have revealed an unexpected new aspect of pectin behavior in plant cells: internalization of cell wall pectins. Clearly, not all intracellular pectins are on their exocytic pathway, but a portion of these molecules are located within compartments of the endocytic pathway. Our data show that boron-cross-linked RGII pectins are internalized. Calcium-cross-linked homogalacturonan pectins (reactive to 2F4 antibody) also undergo an internalization process in meristematic cells of carrot (Daucus carota) root apices (Francoise Liners, personal communication). Because calcium- and boron-cross-linked cell wall pectins are critical for both mechanical strength and porosity of plant cell walls , their internalization might prove to be of critical importance for growth, polarity, and morphogenesis. In plants, dynamic redistribution of pectins between cell walls and the cytoplasm might also have an impact on pectin-based signaling . Interestingly in this context, a pectin-derived elicitor was reported to bind to the PM and to be taken up into plant cells via receptor-mediated endocytosis . Moreover, besides growth, polarity, and morphogenesis, also cell-to-cell transport via plasmodesmata , pollen tube guidance , and sensitivity of cells toward aluminum toxicity are critical pectin-dependent processes. In light of our findings, it will be important to characterize those mechanisms that drive internalization of cell wall pectins and to unravel how this process impinges on mechanical properties of cell walls and, ultimately, on signaling, cell growth, polarity, and morphogenesis in plants. MATERIALS AND METHODS : Plant Material and Inhibitor Treatments | Maize (Zea mays L. cv Careca S230) grains were soaked for 6 h and germinated in well-moistened rolls of filter paper for 4 d in darkness at 20C. For cold treatment, seedlings were kept at 4C for 6 h. Young seedlings with straight primary roots, 50 to 70 mm long, were selected for inhibitor treatments and subsequent immunolabeling studies. Unless stated otherwise, all chemicals were obtained from Sigma Chemicals (St. Louis). For pharmacological experiments, root apices were submerged into appropriate solutions at room temperature. For BFA treatment, we used a 10-2m stock solution (made in dimethyl sulfoxide) further diluted in distilled water to achieve an effective working solution of 10-4m (see also , ) before submergence of root apices for 2 and 6 h. Latrunculin B was used at 10-5m for 3 h, oryzalin at 10-5m for 3 h, and colchicine at 10-3m for 3 h. Indirect Immunofluorescence Microscopy | Excised apical root segments (7 mm in length), encompassing the major growth zones, were fixed in 3.7% (w/v) formaldehyde prepared in stabilizing buffer (SB; 50 mm PIPES, 5 mm MgSO4, and 5 mm EGTA, pH 6.9) for 1 h at room temperature. After rinsing in SB, the root apices were dehydrated in a graded ethanol series diluted with phosphate-buffered saline (PBS). They were embedded in low-melting-point Steedman's wax and processed for immunofluorescence (for details, see ). After a 10-min rinse with absolute methanol at -20C, the sections were transferred to SB containing 1% (w/v) BSA for 30 min at room temperature. They were then incubated with the following primary antibodies: anti-Golgi 58K monoclonal antibody (Sigma G2404) diluted 1:50 (w/v), JIM5 and JIM7 monoclonal antibodies diluted 1:20 (w/v), LM5 monoclonal antibody diluted 1:20 (w/v), LM7 monoclonal antibody diluted 1:10 (w/v), RGII polyclonal antibody diluted 1:100 (w/v), LM2 monoclonal antibody diluted 1:20 (w/v), MAC207 monoclonal antibody diluted 1:20 (w/v), PM H+-ATPase monoclonal antibody diluted 1:100 (w/v), ARF1 polyclonal antibody diluted 1:100 (w/v), and PIN1 polyclonal antibody raised against auxin efflux carrier of maize diluted 1:100 (w/v). All primary antibodies were diluted in PBS supplemented with 1% (w/v) BSA, and sections were incubated in primary antibody for 1 h at room temperature. After rinsing in SB, the sections were incubated for 1 h at room temperature with fluorescein isothiocyanate (FITC)-conjugated anti-mouse IgGs (58K and PM H+-ATPase), with anti-rat IgGs (JIM5, JIM7, LM2, LM5, LM7, and MAC207), or with anti-rabbit IgGs (RGII, ARF1, and PIN1), diluted 1:100 (w/v; mouse and rabbit antibodies) or 1:20 (w/v; rat antibodies) in PBS containing 1% (w/v) BSA. A further rinse in PBS (10 min) preceded a 10-min treatment with 0.01% (w/v) toluidine blue (made in PBS), which diminished autofluorescence of root tissues. The sections were mounted using an anti-fade medium containing p-phenylenediamine . Sections were examined with an Axiovert 405M inverted microscope (Zeiss, Oberkochen, Germany) equipped with epifluorescence and standard FITC excitation and barrier filters (BP 450 --490, LP 520). Photographs were taken on T-Max film (Eastman Kodak, Rochester, NY) rated at 400 ASA. Protoplast Preparation and Immunofluorescence | The 50- to 70-mm-long root apices were selected for protoplasts preparation. Root caps were removed, and 2-mm-long segments were used for further protoplast preparation using the method described by . The segments were transferred to the solution containing 1 mm CaCl2, 0.5% (w/v) polyvinylpyrrolidone, 0.5% (w/v) BSA, 0.8% (w/v) cellulase, 0.1% (w/v) pectolyase, 8 mm MES-KOH to pH 5.5, and 0.6 m sorbitol. They were incubated at 65 rpm for 60 min at 30C. The same solution, but without pectolyase, was subsequently added, and another incubation followed for 90 min. The suspension was then filtered through a nylon mesh, and protoplasts were washed three times with washing solution (1 mm CaCl2, 5 mm MES/Tris, pH 5.5, and 0.6 m sorbitol). Washing solution was replaced with modified growth medium containing 10-4m BFA, and protoplasts were treated for 2 h. For control, washing solution was replaced with growth medium without BFA. Immunolabelling of protoplasts was done according to . Protoplasts were fixed with 4.5% (w/v) paraformaldehyde and 0.5% (v/v) glutaraldehyde for 60 min in PBS containing 0.6 m sorbitol. After washing, they were permeabilized with 0.2% (v/v) Triton X-100 for 30 min. The samples were then incubated with the following primary antibodies diluted in PBS and supplemented with 0.2% (w/v) BSA, JIM5, LM5, and RGII, each at 1:100 (w/v) dilution. After washing, the protoplasts were incubated for 1 h at room temperature with FITC-conjugated anti-rat (JIM5, LM5) and anti-rabbit IgGs (RGII), diluted 1:100 (w/v) in PBS containing 0.2% (w/v) BSA. The protoplasts were mounted using an anti-fade mounting medium containing p-phenylenediamine . The images were taken using a confocal microscope (TCS 4D, Leica, Heidelberg). Backmatter: PMID- 12226522 TI - Arabinoxylan Biosynthesis in Wheat. Characterization of Arabinosyltransferase Activity in Golgi Membranes AB - Arabinoxylan arabinosyltransferase (AX-AraT) activity was investigated using microsomes and Golgi vesicles isolated from wheat (Triticum aestivum) seedlings. Incubation of microsomes with UDP-[14C]-beta-l-arabinopyranose resulted in incorporation of radioactivity into two different products, although most of the radioactivity was present in xylose (Xyl), indicating a high degree of UDP-arabinose (Ara) epimerization. In isolated Golgi vesicles, the epimerization was negligible, and incubation with UDP-[14C]Ara resulted in formation of a product that could be solubilized with proteinase K. In contrast, when Golgi vesicles were incubated with UDP-[14C]Ara in the presence of unlabeled UDP-Xyl, the product obtained could be solubilized with xylanase, whereas proteinase K had no effect. Thus, the AX-AraT is dependent on the synthesis of unsubstituted xylan acting as acceptor. Further analysis of the radiolabeled product formed in the presence of unlabeled UDP-Xyl revealed that it had an apparent molecular mass of approximately 500 kD. Furthermore, the total incorporation of [14C]Ara was dependent on the time of incubation and the amount of Golgi protein used. AX-AraT activity had a pH optimum at 6, and required the presence of divalent cations, Mn2+ being the most efficient. In the absence of UDP-Xyl, a single arabinosylated protein with an apparent molecular mass of 40 kD was radiolabeled. The [14C]Ara labeling became reversible by adding unlabeled UDP-Xyl to the reaction medium. The possible role of this protein in arabinoxylan biosynthesis is discussed. Keywords: Introduction : Plant cells are surrounded by an extracellular matrix known as the cell wall, which plays an important role in development, defense against pathogen attack, and mechanical resistance. The cell wall is composed mainly of polysaccharides, which can be divided in cellulose, hemicellulose, and pectin. The synthesis of these polymers takes place in different subcellular compartments. Cellulose and callose are made at the plasma membrane, whereas other hemicelluloses and pectin are believed to be synthesized in the Golgi apparatus . Xylans are common polysaccharides in plant cell walls, particularly in secondary walls where they are deposited as the major noncellulosic polysaccharide. Xylans consist of a backbone of (1->4)-linked beta-xylosyl residues. About 10% of the Xyl residues typically have single residues of 4-O-methyl-GlcUA and/or GlcUA attached, usually through alpha-(1->2) linkages. The xylosyl residues may also be substituted with short side chains containing l-Ara, Gal, and Xyl, and they may be acetylated at C-2 and/or C-3. In type II walls, which are present in grasses and some related plants, xylans are also the major noncellulosic polysaccharides in the primary walls. The xylans in type II walls have abundant alpha-l-arabinofuranosyl side chains attached through (1->3) and (1->2) linkages and have only a small amount of glucuronosyl side chains . This type of xylan is known as arabinoxylan and may have a role in the cross-linking of cellulose microfibrils and may thereby regulate cell expansion and strengthen the wall . In the endosperms of grasses, arabinoxylans are also abundant and their properties are important for the functionality of flour and the nutritional value of animal feed . Heteropolysaccharide biosynthesis can be divided into four steps: chain or backbone initiation, elongation, side chain addition, and termination and extracellular deposition . Our understanding of these different steps in biosynthesis is still very incomplete. The main enzymes responsible for heteropolysaccharide biosynthesis are glycosyltransferases, but only very few genes for these have been identified. Notable exceptions include the genes encoding a galactomannan galactosyltransferase from legume seeds and a xyloglucan fucosyltransfease . Both of these enzymes are responsible for adding single, terminal residues to a polysaccharide backbone. The enzymes responsible for synthesizing the backbone of xylans are not known. The backbone-synthesizing enzymes may belong to the cellulose synthase-like proteins, but this assumption may be false as it is now known that callose synthase does not resemble cellulose synthase . The biosynthesis of (1->4)-linked beta-xylosyl backbones in xylans is catalyzed by beta-1,4-xylosyltransferase. This enzyme has been investigated in a number of plants by different groups (e.g. ; ; ; ), and we have recently reported the characterization of beta-1,4-xylosyltransferase activity from microsomal membranes isolated from wheat (Triticum aestivum) seedlings . The addition of side chains to xylans has been less investigated and little is known about the way in which the different glycosyltransferases interact to form the complete polysaccharide. A study of glucuronosyltransferase has shown an interaction with xylosyltransferase . The incorporation of arabinosyl groups into (arabino) xylans remains to be explored. Although l-Ara is a common monosaccharide in plant polysaccharides and glycoproteins, there are very few reports of the arabinosyltransferases involved in polysaccharide synthesis, and no l-arabinosyltransferase has yet been identified in any organism. The difficulty and expense in obtaining the UDP-beta-l-arabinopyranose substrate are probably major reasons for the relatively few investigations of arabinosyltransferases. Arabinosyltransferase has been investigated in French bean (Phaseolus vulgaris), but the main polysaccharide product was arabinan rather than xylan . Also, these studies have shown that arabinosylated protein was sometimes formed . The partial purification of arabinan arabinosyltransferase from Golgi membranes isolated from French bean has been reported, but final identification of the enzyme was not achieved . This work is the first to report the presence of arabinoxylan arabinosyltransferase (AX-AraT) in microsomal and Golgi membranes isolated from wheat seedlings. In addition, we present the characterization of the enzyme activity and its product. Furthermore, we demonstrate the presence of an arabinosylated protein and we discuss the possibility that this protein could participate in arabinoxylan biosynthesis. RESULTS : Arabinosyltransferase Activity in Microsomes | We have previously investigated the presence of beta-1,4-xylosyltransferase activity in microsomal membranes isolated from wheat seedlings. The microsomes efficiently incorporated radioactivity from UDP-[14C]Xyl into xylan . In the present work, we have used similar conditions to investigate the presence of arabinosyltransferase by incubating microsomal membranes with UDP-[14C]Ara. Particulate enzyme preparations obtained from wheat seedlings catalyzed the synthesis of a polymeric product from UDP-[14C]Ara. The product was characterized by digestion with well-defined, monospecific enzymes and by gel-filtration chromatography . The intact product eluted in two peaks (I and II) with molecular masses of similar500 and similar15 kD, respectively (Fig. A). After digestion of the radiolabeled product with endo-xylanase A, a major peak eluted with an elution time corresponding to Xyl, xylobiose, and xylotriose (Fig. B). Treatment with arabinofuranosidase, an enzyme that hydrolyzes terminal alpha-l-arabinofuranosyl residues, resulted in a partial digestion of peak I, whereas peak II remained in the original position (Fig. C), indicating that only peak I contains alpha-l-arabinofuranosyl residues. Treatment with proteinase K did not affect the elution of the products, indicating that radiolabeled protein was not present (Fig. D). To confirm the composition of the two peaks eluted from the Superose 12 column, the corresponding fractions were subjected to complete acid hydrolysis in 2 m trifluoroacetic acid, and the hydrolysates were analyzed by thin-layer chromatography (TLC). Peak I contained 22% of [14C]Ara and 78% of [14C]Xyl, whereas peak II contained only [14C]Xyl (data not shown). From these data, it was concluded that microsomal membranes would be difficult to use for characterization of arabinosyltransferase because 4-epimerization of UDP-Ara by the membrane preparations was very high. Therefore, we decided to prepare Golgi vesicles for the further investigation of arabinosyltransferase. Figure 1 | Gel-filtration chromatography of solubilized 14C-labeled product. Gel-filtration chromatography of solubilized 14C-labeled product. 14C-Labeled product was generated under standardized conditions using microsomes corresponding to 180 mug of protein and 22,000 dpm (1 mum) of UDP-[14C]Ara. The product was then solubilized using buffer (A), endo-xylanase A (B), arabinofuranosidase (C), and proteinase K (D) as described in "Materials and Methods." The solubilized material was then separated over a Superose 12 column, and the radioactivity was determined in collected fractions. The elution of monosaccharide and Dextran standards is indicated. Arabinosyltransferase Activity in Golgi Vesicles | Solubilization of the 14C-Labeled Product Formed from UDP-[14C]Ara | Preparations enriched in Golgi membranes were obtained by centrifugation on Suc density gradients. The identity of the fractions from the gradients was confirmed using antibodies against marker proteins . The Golgi-enriched membranes synthesized a radiolabeled product, which was hydrolyzed in 2 m trifluoroacetic acid and analyzed by TLC. The radioactive product was composed mainly of Ara, whereas only traces (<5%) of Xyl were present. Thus, epimerization of UDP-Ara by Golgi membranes was very low compared with microsomal membranes. The nature of the 14C-Ara-containing product was determined by treatment with well-defined enzymes (Fig. A). Incubation of the radioactive product with pure endo-xylanase solubilized no more of the radioactive product than treatment with buffer alone, i.e. 15% to 20%. In a similar manner, treatment with xylanase or arabinofuranosidase alone or in combination had no significant effect. In contrast, proteinase K solubilized 73% of the radioactive product. Thus, we conclude that most of the 14C-Ara was incorporated into protein, whereas no detectable radioactivity was incorporated into arabinoxylan. The inability of arabinofuranosidase to solubilize the product indicates that Ara was not linked to the protein as alpha-arabinofuranoside residues. Figure 2 | Identification of Golgi-enriched fraction. Identification of Golgi-enriched fraction. Microsomal membranes were separated by Suc gradient centrifugation as described in "Materials and Methods." The vesicles at the upper interphase (between 0.25 and 1.1 m Suc; lanes 1 and 3) and lower interphase (between 1.1 and 1.3 m Suc; lanes 2 and 4) were collected. Vesicles corresponding to 1 mug of protein were separated by SDS-PAGE and were analyzed by immunoblotting using antibodies against a Golgi marker (RGP, lanes 1 and 2) and an endoplasmic reticulum marker (calnexin/calreticulin, lanes 3 and 4). Figure 3 | Solubilization of 14C-labeled product. Solubilization of 14C-labeled product. 14C-Labeled product was generated under standard conditions using Golgi membranes corresponding to 10 mug of protein and 22,000 dpm (1 mum) of UDP-[14C]Ara. The incubation was carried out in the absence (A) or the presence (B) of 1 mm of unlabeled UDP-Xyl. The product was solubilized using buffer, a combination of xylanase and arabinofuranosidase (Araf), or proteinase K (PK). After incubation, ethanol was added to precipitate undigested material. The suspension was then centrifuged to remove nonsolubilized material, and the radioactivity in the supernatant was determined. The data represent the average of duplicated samples from two to four separate experiments. The total radiolabel in the products varied between 300 and 500 dpm in the different experiments. Analysis of the 14C-Labeled Product Obtained by Addition of Unlabeled UDP-Xyl in the Reaction Mixture | To determine if xylan backbone synthesis is a requirement for incorporation of 14C-Ara, we analyzed the radioactive product formed from UDP-[14C]Ara in the presence of unlabeled UDP-Xyl (Fig. B). In the presence of 1 mm of cold UDP-Xyl during the reaction, a very different product was obtained. Proteinase K had no significant effect, whereas xylanase A and a combination of xylanase A and arabinofuranosidase solubilized 69% and 76% of the radioactive product, respectively (Fig. B). Arabinofuranosidase alone solubilized 51% of the radioactivity. Similar results were obtained in the presence of 0.5 mm of unlabeled UDP-Xyl (data not shown). Thus, the addition of unlabeled UDP-Xyl directed the flow of 14C-Ara to a product that was sensitive to digestion with xylanase while eliminating the formation of a product sensitive to digestion with proteinase K. Characterization of AX-AraT Activity | From the above results, we concluded that the presence of unlabeled UDP-Xyl results in a radiolabeled product consisting primarily of arabinoxylan. Therefore, we have optimized the AX-AraT assay in presence of unlabeled UDP-Xyl . The time and protein concentration dependence were determined by stopping the reaction mixture, washing the product, and measuring the total radioactivity incorporated by scintillation counting. The total incorporation of 14C-Ara into radiolabeled product was dependent on the amount of protein used and the reaction time (Fig. , A and B) as expected for an enzymatic reaction. The incorporation reached a maximum when Golgi vesicles with a protein content of similar30 mug were used (Fig. B) and after 40 min of incubation (Fig. A). The total incorporation was proportional to the amount of UDP-[14C]Ara used up to 5 mum (data not shown). Due to the limited amounts of substrate available, it was unfortunately not feasible to determine the apparent Km. The effect of divalent cations and pH was determined by stopping the reaction mixture, washing the products, and incubating them with a combination of xylanase and arabinofuranosidase. After incubation with the enzymes, insoluble and higher molecular mass materials were reprecipitated by the addition of ethanol and were then pelleted by centrifugation. In all cases, most of the radioactivity was present in the supernatant, indicating that the major product was arabinoxylan. AX-AraT activity was enhanced by the addition of divalent cations, but the addition of divalent cations was not an absolute requirement (Fig. C). At a concentration of 10 mm Mn2+ or Mg2+, the AX-AraT activity increased approximately 4- and 2-fold, respectively. AX-AraT was active in the investigated pH range of 5.0 to 7.6, with peak activity at pH 6 (Fig. D). The remaining experiments reported below were carried out under the optimal conditions determined from the experiments reported in Figure (see "Materials and Methods"). Figure 4 | Characterization of 14C-incorporation and arabinosyltransferase activity. Characterization of 14C-incorporation and arabinosyltransferase activity. 14C-Labeled product was generated under standard conditions as described in "Materials and Methods. " Golgi membranes corresponding to 10 mug of protein (except in B), 0.5 mm of UDP-Xyl, and 22,000 dpm (1 mum) of UDP-[14C]Ara were used. A, The incubation time was varied. B, The amount of Golgi vesicles (protein) used was varied. C, The concentrations of Mn 2+ (circle) or Mg 2+ were varied. D, The pH was varied using MES buffer (circle) and phosphate buffer . Dpm in A and B represents total incorporation of 14C-Ara into ethanol-insoluble product, and dpm in C and D represents radioactivity measured in the supernatant after digestion with a combination of xylanase A and arabinofuranosidase. The data represent the average of duplicate or triplicate samples. Similar results were obtained in two separate experiments. Analysis of the 14C-Labeled Product by Gel-Filtration Chromatography | The radiolabeled product resulting from incubation of Golgi vesicles with UDP-[14C]Ara in the presence of 0.5 mm of unlabeled UDP-Xyl was further analyzed by gel-filtration chromatography . The intact product eluted in a peak corresponding to a molecular mass of similar500 kD (Fig. A). The peak with an elution time similar to monosaccharides and small oligosaccharides represents residual unincorporated substrate (Fig. A). Treatment with xylanase A or a combination of xylanase A and arabinofuranosidase resulted in the disappearance of the high molecular mass peak and the appearance of a much smaller product that eluted in the included volume similar to monosaccharides and small oligosaccharides (Fig. , B and C). When the radioactive product was treated with proteinase K, the pattern obtained was the same as for the intact product (Fig. D). Figure 5 | Gel-filtration chromatography of solubilized 14C-labeled product. Gel-filtration chromatography of solubilized 14C-labeled product. 14C-Labeled product was generated in the presence of 0.5 mm of unlabeled UDP-Xyl using Golgi membranes corresponding to 30 mug of protein and 95,000 dpm (5 mum) of UDP-[14C]Ara. The incubation time was 60 min. The recovered product was solubilized using buffer (A), xylanase A (B), a combination of xylanase A and arabinofuranosidase (C), or proteinase K (D). The solubilized material was then separated over a Superose 12 column, and the radioactivity was determined in collected fractions. Reversibility of 14C-Arabinosylation of Protein | Polypeptides that become labeled upon incubation with UDP-[14C]Glc have been reported in membrane and soluble enzyme preparations from different species, including pea (Pisum sativum; , ), Arabidopsis, maize (Zea mays), and tobacco (Nicotiana tabacum; ). These polypeptides have been suggested to facilitate the channeling of UDP-activated sugars from the cytoplasm through Golgi membranes to lumenal sites, where they can be used as substrates for glycosyltransferases to synthesize products such as xyloglucan . In the present work, it was demonstrated that when Golgi vesicles were incubated in absence of cold UDP-Xyl, 14C-Ara was incorporated into protein. To determine the reversibility of 14C-Ara labeling onto the protein, we carried out the following experiment. Golgi vesicles were first incubated with UDP-[14C]Ara for 30 min to produce the labeled protein, and aliquots were removed as controls to be analyzed. Following this incubation, cold UDP-Xyl was added to the reaction, and incubation was continued for an additional 30 min. After this second incubation, the reaction was stopped and the samples were analyzed by enzymatic treatment. After the initial incubation with only UDP-[14C]Ara, 62% of the radioactivity (318 +- 13 dpm) could be solubilized with proteinase K, whereas the combination of xylanase A and arabinofuranosidase did not solubilize more radioactivity than the buffer control (116 +- 8 dpm). After the addition of unlabeled UDP-Xyl and incubation for an additional 30 min, 78% of the product (407 +- 23 dpm) could be solubilized with a combination of xylanase A and arabinofuranosidase, whereas treatment with proteinase K released no more radioactivity than treatment with buffer (130 +- 12 dpm). Thus, the protein was reversibly labeled and the removal of label took place simultaneously with the transfer of label to de novo synthesized xylan. Similar results were obtained with several different preparations of Golgi vesicles. Presence of Radiolabeled Protein | To further investigate the presence of arabinosylated protein, we carried out the same experiment described above in which samples were removed after the first incubation with only UDP-[14C]Ara and after the second incubation with the addition of unlabeled UDP-Xyl. The samples were subjected to SDS-PAGE and were exposed to PhosphorImager screens . A single labeled band of approximately 40 kD was found in samples removed after the first incubation containing only UDP-[14C]Ara (Fig. , lane 1), whereas no band was detected in samples removed after the second incubation with addition of unlabeled UDP-Xyl (Fig. , lane 2). Figure 6 | Reversibility of protein arabinosylation. Reversibility of protein arabinosylation. 14C-Labeled product was generated under optimized conditions using Golgi membranes corresponding to 30 mug of protein and 66,000 dpm (3 mum) of UDP-[14C]Ara. The reaction products were formed after initial incubation of 30 min with UDP-[14C]Ara (lane 1) and after subsequent incubation of 30 min in the presence of 0.5 mm unlabeled UDP-Xyl (lane 2). The products were precipitated with ethanol and were separated by SDS-PAGE. The gel was exposed to PhosphorImager screens. DISCUSSION : We have previously used microsomal membranes from wheat seedlings to investigate xylosyltransferase involved in arabinoxylan biosynthesis . The microsomal preparation would be expected to contain UDP-Xyl 4-epimerase, and this enzyme is responsible for the production of UDP-l-[14C]Ara from UDP-d-[14C]Xyl. However, the radioactive product obtained in this earlier work consisted mainly of Xyl, whereas only traces of Ara were found, indicating a low 4-epimerization of UDP-Xyl . Attempting to detect the presence of arabinosyltransferase activity, we have incubated microsomal membranes with UDP-[14C]Ara. Two radioactive products were separated by gel-filtration chromatography (Fig. A). In contrast to the earlier results, one of the products consisted of 78% of Xyl, whereas the other consisted exclusively of Xyl, indicating a high 4-epimerase activity. The equilibrium of the epimerase reaction favors UDP-Xyl , but the very pronounced difference in the degree of epimerization when microsomal membranes were incubated with UDP-[14C]Ara or UDP[14C]Xyl was nevertheless surprising. Other previous studies of xylan synthesis have also been hampered by the presence of UDP-Xyl-4-epimerase . Because hemicelluloses and pectin are synthesized in the Golgi apparatus and the interference of 4-epimerase makes the properties of the enzyme difficult to study in microsomal preparations, we decided to use Golgi-enriched vesicles for the investigation of arabinosyltransferase. Golgi vesicles were incubated with UDP-[14C]Ara and the resultant radioactive product consisted mainly of [14C]Ara, whereas only traces of [14C]Xyl were found, indicating an insignificant 4-epimerase activity. Similar results were found for particulate enzyme preparations obtained from bean that were incubated with UDP-l-[l-3H]Ara . In bean, most radioactivity was incorporated into pectin, and radioactive Xyl only accounted for 6% to 8% of the radioactivity. In the present investigation, characterization of the radioactive product revealed that most of the labeling was incorporated into protein and only traces of [14C]Ara were incorporated into xylan (Fig. A). In the presence of unlabeled UDP-Xyl, most of the radioactive product was arabinoxylan, whereas no radiolabeled protein was detected (Fig. B). Thus, nascent xylan acted as an acceptor for the further incorporation of [14C]Ara from UDP-[14C]Ara. Xylan xylosyltransferase is presumably a processive enzyme and its synergy with the arabinosyltransferase suggests that the two enzymes form a complex with the nonreducing end of the growing polysaccharide. Similar interactions between backbone-synthesizing enzymes and decorating enzymes have been reported for the xylan glucuronosyltransferase . Sustained incorporation of [14C]GlcUA from UDP-d-[14C]GlcUA into glucuronoxylan only occurred in the presence of UDP-d-Xyl, indicating that xylan was being synthesized and acted as an acceptor for the further incorporation of GlcUA. A similar interaction is also known between the galactosyltransferase and mannosyltransferase involved in galactomannan biosynthesis . Because the presence of UDP-Xyl results in a product consisting mainly of arabinoxylan, the AX-AraT activity was further characterized in the presence of nonradioactive UDP-Xyl during the incubation. AX-AraT activity was dependent on time and protein concentration. The maximum yield of product occurs within 40 min, and the enzyme was active over a broad pH range, with optimum at pH 6.0. Optimum pH values between 6 and 6.5 were reported for arabinosyltransferase from mung bean (Vigna radiata) shoots involved in arabinan synthesis . Xylan xylosyltransferase from wheat is also active over a wide range, but has peak activity at pH 6.8 . The AX-AraT activity was enhanced by addition of Mg2+ or Mn2+ as has been found for arabinosyltransferases involved in arabinan biosynthesis and in general for other glycosyltransferases. The xylan xylosyltransferase from wheat was also enhanced in the presence of Mg2+ or Mn2+, although the effect was less pronounced . In the absence of added UDP-Xyl, only protein was radiolabeled with [14C]Ara. A single labeled protein migrated on SDS-PAGE with an apparent molecular mass of 40 kD . However, upon subsequent incubation with UDP-Xyl, the label disappeared. The addition of UDP-Glc similarly led to a disappearance of the radiolabel. These properties are similar to what has been described for reversibly glycosylated proteins (RGPs), which are soluble proteins found in association with the Golgi membranes. During studies of polysaccharide synthesis in pea Golgi membranes, identified a 41-kD protein doublet that they suggested was involved in polysaccharide synthesis. The authors showed that this protein could be glycosylated by radiolabeled UDP-Glc, but that this labeling could be reversibly competed with unlabeled UDP-Glc, UDP-Xyl, and UDP-Gal, the sugars that make up xyloglucan, but not by other nucleotide sugars. The effect of UDP-Ara was not reported in these previous investigations. The 41-kD protein was named Pisum sativum reversibly glycosylated polypeptide-1 [PsRGP1]; ), and antibodies raised against PsRGP1 showed that it is soluble and localized to the Golgi compartment . have isolated and characterized a cDNA clone encoding the Arabidopsis homolog of PsRGP1, named AtRGP1. Sequence comparisons with previously defined UDP-Glc-binding sites showed that AtRGP1 contains a similar motif, which may be involved in the binding of UDP-sugars. A single amino acid, Arg-158, was found to be labeled with [14C]Glc, in accordance with the single glycosylation of PsRGP1 . Several authors have speculated that RGPs may in some way be involved in polysaccharide biosynthesis, as protein primers, as intermediates involved in transport, or as real glycosyltransferases . There is no evidence for a primer function of RGP other than an analogy to protein-primed starch and glycogen synthesis . However, the ability of RGPs to be reversibly glycosylated, their exposure to the cytoplasm in which nucleotide sugars are found, and their association with Golgi membranes support the notion that RGPs could act as carriers of UDP-sugars from the cytoplasm to the Golgi apparatus . The existence of RGP in dicots and monocots , but apparently not in other organisms, suggests a plant-specific function. Most authors have assumed that the RGPs contained glycosidic bonds, as this would agree with the behavior of the protein on SDS-PAGE. However, a recent report provides evidence that the glycosylated RGP is an unreactive glycoprotein formed relatively slowly by glycosyl transfer from a rapidly formed UDP-sugar-binding polypeptide . The authors suggest that it is the evanescent-bound sugar nucleotide that is capable of acting as a sugar donor and not the final stable glycoprotein. Arabinoxylan formation is enhanced by addition of unlabeled UDP-Xyl in the reaction medium. We cannot conclude through our experiments that [14C]Ara is transferred from the labeled protein onto arabinoxylan. The alternative that UDP-Xyl is replacing the Ara on the protein in a reaction unrelated to xylan biosynthesis cannot be excluded. A chase experiment with excess unlabeled UDP-Ara would be required to make this conclusion, but unfortunately, the substrate was not available. It is interesting that [14C]ferulic acid has also been shown to be transiently incorporated into a 40-kD protein in wheat (N. Obel and H.V. Scheller, unpublished data). In this protein, ferulic acid appeared to be bound to C5 of an arabinofuranosyl residue. If the two 40-kD proteins are identical, the linkage of ferulic acid eliminates the possibility that UDP-Ara is bound as an intact nucleotide sugar as found for the protein studied by . This would make it more likely that the 40-kD protein in wheat is directly involved in arabinoxylan biosynthesis. We are currently investigating the identity of the labeled proteins. MATERIALS AND METHODS : Chemicals, Reagents, and Enzymes | UDP-l-[14C]Ara with specific activity of 9.9 GBq mm-1 was prepared as described in . Wheat (Triticum aestivum cv Cadenza) arabinoxylan and arabinofuranosidase from Aspergillus niger were purchased from Megazyme International (Bray, Ireland). Homogeneous endo-xylanase A from A. niger was a gift from Drs. Troels Gravesen and Susan Madrid (Danisco Biotechnology, Copenhagen). The xylanase had no detectable arabinanase, xyloglucanase, or arabinofuranosidase activity, and the arabinofuranosidase had no detectable xylanase activity. Proteinase K was from Boehringer Mannheim (Mannheim, Germany) and had no detectable hydrolytic activity with arabinoxylan. Dextran molecular mass standards were purchased from Fluka (Buchs, Switzerland). Plant Material | Wheat seedlings were grown in trays of vermiculite in controlled environment chambers at 20C with 150 mumol photons m-2 s-1 and a 16-h photoperiod. Four-day-old seedlings were used for preparation of microsomes and Golgi vesicles. Preparation of Microsomes | The entire preparation of microsomes took place in a cold room (similar4C). Shoots and coleoptiles were harvested with a razor blade and were ground with a mortar and pestle in a buffer (1 mL g-1 of plant material) of 50 mm potassium phosphate, pH 7.2, 10% (w/v) polyvinylpolypyrrolidone, 1 mm dithiothreitol (DTT), 1 mm MgCl2, and 0.4 m Suc. The suspension was filtered through a nylon cloth (30-mum mesh), and was centrifuged for 15 min at 3,000g to remove debris. The resulting supernatant was centrifuged at 48,000g for 1 h to pellet the microsomes, which were resuspended in homogenization buffer without polyvinylpolypyrrolidone at a ratio of approximately 30 muL of buffer g-1 fresh weight of plant tissue. Total protein was determined according to with bovine serum albumin as a standard. Preparation of Golgi Vesicles | The method to obtain Golgi-derived vesicles was based on the procedure of with minor modifications. Shoots and coleoptiles (8 --12 g) were homogenized by hand with razor blades in a buffer (1 mL g-1 fresh weight) of 0.5 m Suc, 0.1 m potassium phosphate, pH 7.2, 5 mm MgCl2, and 1 mm DTT. After the tissue was finely chopped, it was homogenized for 2 min in a mortar. This procedure was carried out at 0C. The homogenate was filtered through nylon cloth (30-mum mesh) and was centrifuged at 1,000g for 2 min. The supernatant was loaded onto a 4-mL 1.3 m Suc cushion and was centrifuged at 100,000g for 90 min. The upper phase was removed without disturbing the interphase fraction. A discontinuous gradient was then formed by overlaying the solution with 5 mL of 1.1 m Suc and 4 mL of 0.25 m Suc. The Suc solutions were prepared in a buffer containing 0.1 m potassium phosphate, pH 7.2, 5 mm MgCl2, and 1 mm DTT. The gradients were centrifuged for 90 min at 100,000g. The interphase at 0.25/1.1 Suc was collected and stored at -80C until use. The identity and purity of the fractions was analyzed by immunoblotting using antibodies against representative marker proteins. Polyclonal antibodies against RGP from pea (Pisum sativum; ) and calnexin/calreticulin from barley (Hordeum vulgare; ) were kind gifts from Dr. Kanwarpal S. Dhugga (Pioneer Hi-Bred International, Des Moines, IA) and Dr. David J. Simpson (Carlsberg Laboratory, Copenhagen), respectively. The immunoblots were visualized using secondary antibodies conjugated with horseradish peroxidase (Dako, Glostrup, Denmark) and a chemiluminescence detection kit (Amersham Pharmacia Biotech, Buckinghamshire, UK). Standard Assay for Arabinosyltransferase | The standard assay was used unless otherwise indicated. The reaction mixture (final volume of 40 muL) consisted of 10 muL of reaction buffer (200 mm potassium phosphate, pH 7.2, 0.8 m Suc, 40 mm MgCl2, and 4 mm DTT), 7 mm MnCl2, 22,000 dpm of UDP-[14C]Ara (similar1 mum), and 20 muL of the microsomal membranes (containing approximately 180 mug of protein) or 20 muL of Golgi vesicles (containing approximately 10 --15 mug of protein). The reaction mixture was incubated for 60 min at 30C, and the reaction was then terminated by boiling for 5 min, cooled on ice, and 500 muL of chloroform:methanol (3:2, v/v) was added. The sample was mixed on a vortex mixer, and the precipitate was collected by centrifugation at 10,000g for 10 min. The pellet was resuspended in 500 muL of aqueous 65% (v/v) ethanol, and 90 mug of wheat arabinoxylan was added as carrier. The sample was centrifuged again and the supernatant was removed. The resultant pellet was washed with 65% (v/v) ethanol until the washes were free of the radioactivity. The final pellet was suspended in 200 muL of water, boiled for 5 min, and counted in a liquid-scintillation counter after addition of 2 mL of scintillation fluid (Ecoscint, National Diagnostic, Manville, NJ), or was treated with different enzymes as described below. Optimization of AX-AraT Assay | Time course and protein concentration dependence for enzyme activity was determined by using the standard assay. After reaction, the product was washed, and the total radioactivity in the final pellets was determined by liquid-scintillation counting after the addition of 2 mL of scintillation fluid. Effect of divalent cations and optimum pH for the enzyme activity was investigated by using the standard assay in which the buffer reaction did not contain any divalent ions and by adding different concentrations (0 --30 mm) of MnCl2 or MgCl2 or by using the following buffers to obtain the required pH: MES, pH 5.0 to 6.7, and potassium phosphate, pH 6.2 to 7.6. After the reaction, the recovered product was washed and incubated with a combination of xylanase A and arabinofuranosidase at 30C for 3 h (see below). After stopping the reaction by boiling, the products were reprecipitated by adding ethanol (70% [v/v] final concentration) and arabinoxylan (160 mug) as a carrier and they were pelleted by centrifugation at 10,000g for 10 min. The supernatants containing the solubilized material and the pellets were subjected to liquid scintillation counting. Optimized Assay for AX-AraT | The incorporation reaction (final volume of 60 muL) consisted of 15 muL of reaction buffer (200 mm MES, pH 6, 0.8 m Suc, and 4 mm DTT), 5 mm MnCl2, 4.5 mum of UDP-[14C]Ara (95,000 dpm), 0.5 mm UDP-Xyl, and Golgi vesicles corresponding to approximately 15 to 20 mug of protein. After the incubation at 30C for 60 min, the sample was treated as described above for the standard assay. Enzymatic Digestion of the 14C-Labeled Product | In general, the final pellet obtained was first suspended in 120 muL of water and boiled for 3 min. After cooling down, buffer and enzymes were added. The digests/solubilizations were incubated for 3 h at 30C. Insoluble and higher molecular mass materials were reprecipitated by adding ethanol (70% [v/v] final concentration) and arabinoxylan (160 mug) as a carrier, they were pelleted by centrifugation at 10,000g for 10 min, and the supernatants were counted in a liquid-scintillation counter or subjected to gel-filtration chromatography as described below. Enzymes were used in the following concentrations and buffer systems: endo-xylanase A (0.002 U, 1 unit releases 1 mumol of reducing arabinoxylan oligosaccharide min-1) and arabinofuranosidase (0.01 U, 1 unit releases 1 mumol of Ara from arabinoxylan min-1) were incubated in 50 mm sodium acetate (pH 5.2). Proteinase K (0.2 mg) was incubated in water. Gel-Filtration Chromatography of 14C-Labeled Product | The samples were treated with different enzymes as described above. Supernatants obtained after reprecipitation by the addition of ethanol and centrifugation were spin-filtered and analyzed by gel-filtration chromatography on a Superose 12 column (30 cm long, 1 cm i.d.; Amersham Pharmacia Biotech). The column was equilibrated in 50 mm ammonium formate (pH 5.0). The sample was applied and eluted with the same buffer at a flow rate of 0.4 mL min-1. Fractions (0.8 mL) were collected and the radioactivity was determined by liquid-scintillation counting. The Dextran standards were monitored in the eluate using a refractive-index detector (model 131; Gilson, Middleton, WI). Total Acid Hydrolysis | The 14C-labeled product recovered was resuspended in water (250 muL) and was incubated for 120C for 1 h after the addition of 44 muL of trifluoroacetic acid (13.5 m). The treated product was dried, resuspended in water, and separated on TLC plates (Silica Gel 60 F254; Merck, Darmstadt, Germany) in ethyl acetate:acetic acid:methanol:water (12:3:3:2, v/v) for 2 h. Unlabeled standards were detected by dipping the TLC plates into a solution of sulfuric acid:ethanol (1:9, v/v) and heating the plate until the sugars charred. Radiolabeled products were detected by cutting the sheet corresponding to the sample into 2.5-mm strips for scintillation counting. To perform total acid hydrolysis of fractions collected from the Superose 12 column, the fractions were combined, dried, resuspended in water, and treated as described above. Presence of Radiolabeled Protein | Optimized assay was used to generate six reaction samples, which were incubated for 30 min. After this first incubation, three samples were stopped by boiling and were combined. The remaining three reaction tubes were incubated for an additional 30 min after the addition of 0.5 mm cold UDP-Xyl after which the samples were stopped and combined. The pooled samples were dried and resuspended in 30 muL of SDS sample buffer and were subjected to SDS-PAGE in 8% to 25% (w/v) High-Tris gradient gels. The gels were dried and exposed to storage phosphor screens for approximately 3 d. The exposed screens were analyzed with a PhosphorImager (model 425F; Molecular Dynamics, Sunnyvale, CA). Backmatter: PMID- 12226523 TI - Patterns of Expression and Normalized Levels of the Five Arabidopsis Phytochromes AB - Using monoclonal antibodies specific for each apoprotein and full-length purified apoprotein standards, the levels of the five Arabidopsis phytochromes and their patterns of expression in seedlings and mature plants and under different light conditions have been characterized. Phytochrome levels are normalized to the DNA content of the various tissue extracts to approximate normalization to the number of cells in the tissue. One phytochrome, phytochrome A, is highly light labile. The other four phytochromes are much more light stable, although among these, phytochromes B and C are reduced 4- to 5-fold in red- or white-light-grown seedlings compared with dark-grown seedlings. The total amount of extractable phytochrome is 23-fold lower in light-grown than dark-grown tissues, and the percent ratios of the five phytochromes, A:B:C:D:E, are measured as 85:10:2:1.5:1.5 in etiolated seedlings and 5:40:15:15:25 in seedlings grown in continuous white light. The four light-stable phytochromes are present at nearly unchanging levels throughout the course of development of mature rosette and reproductive-stage plants and are present in leaves, stems, roots, and flowers. Phytochrome protein expression patterns over the course of seed germination and under diurnal and circadian light cycles are also characterized. Little cycling in response to photoperiod is observed, and this very low amplitude cycling of some phytochrome proteins is out of phase with previously reported cycling of PHY mRNA levels. These studies indicate that, with the exception of phytochrome A, the family of phytochrome photoreceptors in Arabidopsis constitutes a quite stable and very broadly distributed array of sensory molecules. Keywords: Introduction : The phytochromes are a family of soluble chromoproteins that function in the absorption of red (R) and far-red (FR) light and the transduction of intracellular signals during light-regulated plant development. In early studies, physiological and biochemical experiments indicated that higher plants contain at least two different forms of phytochrome . These were frequently referred to as type I or "light-labile" phytochrome, which predominated in etiolated tissue, and type II or "light-stable" phytochrome, which predominated in light-grown tissue (for review, see ; ). Identification of PHY gene families, first in Arabidopsis and subsequently in many other plant species, and analysis of their phytochrome protein products have generally confirmed this view , although most plants contain more than two forms and the levels and light stabilities of all of the phytochromes from a single species have not been analyzed up to this time. Complete or partial sequences of PHY genes from a large number of flowering plants have been isolated. Among these, the most intensively studied are those of Arabidopsis, the complete PHY gene complement of which consists of five members, PHYTOCHROME A through PHYTOCHROME E (PHYA --PHYE; ; ). Tomato (Lycopersicon esculentum), the only other eudicot from which all of the phytochrome coding sequences have been isolated, also contains five PHY genes . These are similar to the Arabidopsis genes and, with the exception of the pairs of recently derived phyB-like genes Arabidopsis B/D and tomato B1/B2, are thought to be orthologs of the Arabidopsis genes . PHY genes sequenced from a variety of higher plants, including both monocots and dicots, all show extensive similarity in sequence and overall structure to the Arabidopsis sequences. Phylogenetic analysis indicates that a duplication event in early seed plant evolution gave rise to two initial PHY gene lineages, PHYA/C and PHYB/D/E, which underwent later duplications in angiosperms to generate the divergent PHYA, PHYC, PHYE, and PHYB/D lineages . Members of these lineages are found in most plants examined, with the exception of PHYE genes, which have not been found in sampled monocots and may have evolved only in dicots after the monocot/dicot divergence or may have been lost in an early monocot progenitor . Hence, available evidence strongly supports the view that the five Arabidopsis PHY genes and their protein products are representative of phytochrome diversity found across a broad range of angiosperm plants. The phenotypic consequences of mutations in the Arabidopsis PHY genes and in PHY genes from other species have been used to assess the respective roles of these molecules in regulation of R/FR photomorphogenic responses. A single Arabidopsis gene, PHYA, encodes a light-labile type I phytochrome, which has been shown through mutant analysis to function in very low fluence and FR high irradiance responses . Mutations in PHYA genes from tomato, pea (Pisum sativum), and rice (Oryza sativa) result in similar defects in these general classes of phytochrome responses . showed that the Arabidopsis PHYB and PHYC genes encode proteins that are less abundant than phyA in etiolated tissue and appear to be light stable. These, therefore, resemble type II phytochromes. The abundance and light stability of the phyD and phyE phytochromes have not been analyzed, but these proteins are most closely related in sequence to phyB. Arabidopsis phyB, phyD, and phyE mutants have been isolated and used to demonstrate that all three function in low fluence, R/FR-reversible responses and in shade avoidance responses to altered R to FR ratio . Again, mutations in several PHYB-related genes from other plant species affect low fluence, R/FR-reversible responses and shade avoidance responses in a similar way . Mutations in the Arabidopsis PHYC gene or in PHYC homologs from other species have not yet been described. Before the complexity of the phytochrome family was known, the expression pattern and distribution of phytochrome were characterized in several different plant species using assays based upon spectroscopy, biological activity, and immunolocalization (for review, see ). These experiments were limited to addressing only the type I or phyA form of phytochrome or were nonselective among the various forms. In general, phytochrome was found in most tissues tested spectroscopically or immunologically, although significant variation in its localization was seen in different species. Variation was also observed in the phytochrome content of different cell types in tissues from a given plant, and phytochrome was often observed to be most abundant in young, rapidly expanding cells . Following the description of the five Arabidopsis PHY genes, preliminary characterization of the distributions of the PHY mRNAs in isolated plant organs indicated that these transcripts are relatively constitutive throughout the mature plant and throughout development . A quantitative analysis of the steady-state levels and distributions of all of the five tomato PHY mRNAs was performed and was again found fairly constitutive expression in seedlings . In these studies, induction of all five transcripts was observed over the course of seed germination, 2- to 3-fold diurnal cycling of the PHYA, PHYB1, and PHYB2 mRNAs was seen, and some variation in expression was seen in mature tomato organs, although all five transcripts were detectable in all tissues examined. Promoter-reporter gene fusion constructs for the Arabidopsis PHYA, PHYB, PHYD, and PHYE and tobacco PHYA and PHYB transcription regulatory regions have also been constructed and their activities monitored in transgenic plants . In these experiments, examples of well-defined organ and tissue specificities for promoter activity were observed, however all of the PHY promoters tested were expressed in a large number of different cell types and in most plant organs. Variation between Arabidopsis and tobacco for their respective PHYA and PHYB promoter expression patterns was also described. All of these approaches to characterization of PHY gene transcription patterns indicate that, although some regulation of PHY gene promoter activity occurs with respect to different cell types and stages of development, phytochromes are likely to be broadly distributed in a given plant and most, perhaps all, plant cells are R/FR responsive. Earlier efforts at determining the in vivo levels of the phytochrome apoproteins have been carried out principally using oat or pea seedlings and ELISA, radioimmunoassay, or immunoblot assays for phytochrome . However, in none of these cases was it possible to distinguish completely among the different forms of phytochrome present in the extracts because a comprehensive set of phy-specific antibodies was not available. The most comprehensive analyses were performed by , , ) and used monoclonal antibodies (MAb) specific to each of three phytochrome forms from oat. Levels of these phytochromes were determined in unimbibed oat seeds and in dissected 3-d-old seedlings, and it was concluded that one form, 124-kD phytochrome, was predominant in dark-grown seedlings, whereas the other two, the 123- and 125-kD forms, were more abundant in the light and that there was little evidence for marked differences in the spatial distribution of the three forms. Several lines of evidence indicate that the level of photoreversible phytochrome in plants is a critical determinant of the sensitivity of the plant to light and the strength of the physiological or developmental response to a given light environment. Overexpression of phyA or phyB phytochromes in transgenic plants has striking effects on growth and development in response to light , even at the level of only a doubling of PHYB gene dosage . The PHYA and PHYB wild-type alleles conversely are incompletely dominant to the phyA and phyB null mutations , indicating again that a 2-fold change in photoreceptor level is physiologically significant. Arabidopsis has been used extensively in the analysis of phytochrome function and phytochrome mutants and MAb that specifically detect each of the five Arabidopsis phytochrome apoproteins have been described . We have used these reagents here to investigate the light stabilities, distributions, and levels of this family of photoreceptors in wild-type plants throughout the growth cycle, under different light conditions, and over diurnal and circadian light cycles. RESULTS : Normalization of Protein Gel Loads to DNA Content of Extracts | Tissues of seedlings grown under different light conditions, of mature plants at various ages, and from different plant organs vary in water content, cell size, and protein content. This raises the question of what basis should be used to compare the levels of the phytochromes present in such different tissues. The amount of phytochrome per gram fresh weight, per seedling or organ, or per milligram of total protein fails to account for the variability in the parameters described above. One useful comparison that would take into account several of these variables is an estimate of the average amount of each phytochrome per cell. As an approximation of this, in our analysis, the amount of extracted plant protein loaded on gels has been normalized to the DNA content of each extract. Frozen tissues were ground in liquid nitrogen, and separate aliquots were extracted and assayed for total protein content, using the standard phytochrome extraction buffer, or for total DNA content. The results of these protein and DNA assays are shown in Table . The protein to DNA ratios range from 65:1 to 138:1 and, for a given tissue, are highly reproducible. Although there is a 7-fold difference in the amounts of protein and DNA per gram fresh weight of tissue when comparing, for example, dark-grown seedlings to flowers, the protein to DNA ratios among all the samples vary only 2-fold. Lanes of gels throughout this paper were loaded as indicated in the last column of Table where a dark-grown seedling is set as the standard of a 1x load. Therefore, with the exception of the seed germination experiments, where extracts were loaded on a per seed basis (see below), the immunoblot signals on a given immunoblot in the figures in this paper can be interpreted as indicating the level of that phytochrome protein present in a similar number of DNA or genome equivalents from the various tissue samples tested. Table I | Protein to DNA ratios in Arabidopsis seedlings, mature plants, and plant organs Light Stabilities of Phytochrome Proteins in Seedlings | described the isolation and characterization of MAb against all five of the Arabidopsis phytochromes. With the exception of the recognition by anti-phyD MAb 2C1 of a low abundance non-phytochrome cross-reacting protein (NPXR) in dark-grown tissue that is described in that work, all available evidence indicates that the anti-phytochrome MAb lines are phytochrome specific. Moreover, it was shown that MAb 2C1 can be used to identify and quantify phyD in dark-grown plant extracts if a 20% (w/v) (NH4)2SO4 fraction rather than a 25% fraction is analyzed . To characterize the light stabilities of the Arabidopsis phytochromes, immunoblots of (NH4)2SO4-precipitated fractions of 7-d-old seedlings that were grown in the dark, grown in the dark but irradiated with 3, 12, or 24 h of continuous R light before harvest, or grown in continuous white light (WL) were probed with each of the five anti-phytochrome MAbs. These blots are shown in Figure . For the phyD blot, a 20% (w/v) (NH4)2SO4 fraction was prepared to eliminate NPXR from the extract. The seedling extracts were loaded on gels as indicated in Table , and for each antibody, dilution curves of the dark-grown extracts indicate the response of the assay to relative antigen levels. Figure 1 | Immunoblot analysis of the light stabilities of the five Arabidopsis phytochromes. Immunoblot analysis of the light stabilities of the five Arabidopsis phytochromes. Seedlings were grown for 7 d in continuous darkness (D) or WL, or in darkness and then placed under R for 3 h (R3), 12 h (R12), or 24 h (R24) before harvest at 7 d. Extracts were prepared and precipitated with 25% (w/v) (NH4)2SO4 for phyA, B, C, and E or with 20% (w/v) (NH4)2SO4 for phyD. An amount of protein equivalent to 300 mug of total extractable protein was loaded in the D, R3, R12, and R24 lanes and an amount equivalent to 390 mug in the WL lane. For each phytochrome, dilution curves of the darkness extract were included to estimate the -fold difference in the level of that apoprotein under the different light conditions. MAbs used were: phyA, 073d; phyB, B6B3; phyC, C11 and C13; phyD, 2C1; and phyE, 7B3. In Figure , the phyA protein exhibits the strong R-dependent down-regulation that has previously been described for light-labile or type I phytochrome, resulting in a 50- to 100-fold reduction in level by 12 h in R. The phyB and phyC proteins also show significant R-light-induced down-regulation, although the reduction in protein level is more gradual and much less pronounced compared with the phyA down-regulation . Compared with dark-grown seedlings, levels of phyB and phyC are 4- to 5-fold reduced in seedlings grown either for 6 d in the dark followed by 24 h of R light or for 7 d in continuous WL. The levels of phyD and phyE are not strongly changed upon extended exposure to R or WL, showing at most a 2-fold reduction compared with the dark-grown seedlings . Thus, phyD and phyE most closely resemble truly light-stable phytochromes. Taken together, the varied responses of the five phytochrome protein levels to R and WL indicate that there is a gradation of light stabilities in this receptor family. Normalized Phytochrome Levels in Arabidopsis Dark-Grown and Light-Grown Seedlings | It is important to note that, in Figure , exposure times for detection of the chemiluminescence signal varied from blot to blot, and although the chemiluminescence signals in the figure accurately reflect the relative levels of each phytochrome protein under the set of conditions being tested, they cannot be used to compare the levels of the different phytochrome proteins with each other. Therefore, the absolute levels of the five phytochromes were determined by comparison with standard curves of purified proteins. Several protocols for preparation of protein extracts for use in immunoblot analysis of phytochrome have been described. We compared four procedures using Arabidopsis seedling tissues: extraction of fresh tissue into hot SDS buffer , (NH4)2SO4-precipitated extraction , the EZ protocol , and grinding and direct extraction of frozen tissue . Extraction of fresh tissue into hot SDS buffer yielded low-quality immunoblots with high backgrounds, and we observed low-phytochrome recovery with the EZ protocol such that the calculated protein levels in the various tissues were approximately 2-fold lower using this method compared with the others. The (NH4)2SO4-precipitated extraction protocol and direct extraction of frozen tissue yielded the most consistent results, and because the (NH4)2SO4-precipitation introduces an additional correction for the fraction of total protein precipitated, the very simple direct extraction method was adopted for most of the quantitative experiments. To construct standard curves of purified phytochrome apoproteins, the five Arabidopsis PHY cDNA sequences were expressed from the pET3c vector in Escherichia coli, and the approximately 120-kD phytochrome products were purified from gels by electroelution. Figure A shows a silver-stained gel of the five purified proteins. For phytochromes B, C, and E, direct extracts of seedlings grown for 7 d in darkness, for 6 d in darkness followed by 24 h in R light (R24), or for 7 d in continuous WL, were analyzed by immunoblotting with standard curves of each of the purified apoproteins. Figure B shows representative blots, which were loaded with 50 mug of the dark-grown or R24 extracts or 65 mug of the WL extract to normalize to DNA content . For the highly light-labile phyA, a R irradiation of only 12 h was used (R12), and separate blots were prepared for the dark and light extracts because of the large difference in the level of phyA in these samples (Fig. C). Measurement of the level of phyD in dark-grown tissue is complicated by cross-reactivity of the anti-phyD 2C1 MAb to a protein (NPXR) found in dark-grown extracts . For this reason, the three extracts for phyD quantitation were precipitated with 20% (w/v) (NH4)2SO4, which does not precipitate NPXR protein , and these samples were blotted and probed as shown in Figure D. Figure 2 | Quantitation of phytochrome levels in seedling extracts with purified apoprotein standards. Quantitation of phytochrome levels in seedling extracts with purified apoprotein standards. A, Silver-stained 6% (w/v) SDS-PAGE gel of 50 ng of each of the E. coli-expressed purified phytochrome apoprotein standards and a commercial preparation of 50 ng of purified beta-galactosidase. B, phyB, phyC, and phyE immunoblots of direct extracts of seedlings grown for 7 d in continuous darkness (D) or WL, or for 6 d in the dark followed by 24 h in R (R24) with dilution curves of the purified apoproteins. Fifty micrograms of D and R24 extract protein and 65 mug of WL extract were loaded, and blots were probed with the MAbs described in Figure . C, Two separate phyA immunoblots, one with 10 mug of darkness direct extract and one with 130 mug of the R12 or WL direct extracts and appropriate dilution curves of the purified phyA apoprotein, were probed with MAb 073d. D, To eliminate cross-reactivity of the anti-phyD MAb 2C1 with a non-phytochrome protein, 20% (w/v) (NH4)2SO4 precipitated fractions were used instead of direct extracts on the phyD blot. An amount of precipitated protein equivalent to 300 mug of darkness or R24 extract or 390 mug of WL extract was loaded. Three independent immunoblots for each of the phytochromes, similar to those shown in Figure , B through D, were scanned and quantitated by densitometry. Figure shows curves generated from representative blots and Table gives the average values determined for nanograms of the five phytochromes per 100 mug of total protein extracted from dark-grown and R-irradiated seedlings and per 130 mug of total protein from WL-grown seedlings (see Table for normalization to DNA content). The choice of expressing the phytochrome levels using nanograms per 100 mug of total extracted protein from dark-grown seedlings as a base, and all other values relative to this, is arbitrary. It allows phytochrome levels to be compared with each other and, using corrections for the weight, protein content, or estimated number of cells per seedling, can be converted to other units (see "Discussion"). For phyD blots, the amounts of (NH4)2SO4-precipitated protein loaded on gels were the equivalent of 300 mug of total protein from dark-grown seedlings and 390 mug from R- or WL-grown seedlings . To confirm the comparable nature of the two different types of samples, in a separate experiment, the phyD level in direct extracts of light-grown tissues, where the cross-reacting NPXR protein is absent , was measured as 3 ng 100 mug-1 total protein (data not shown). Figure 3 | Quantitation of phytochrome immunoblots by densitometry. Quantitation of phytochrome immunoblots by densitometry. Chemiluminescence signals on films of blots similar to those shown in Figure were scanned and analyzed by densitometry. Densitometry units in integrated OD units are plotted versus the amount of purified apoprotein loaded per lane of the gel. The values obtained for darkness (D), R24 (phyB, C, D, and E), or R12 (phyA), and WL extracts are indicated by arrow heads. Table II | Nanograms of phytochrome apoprotein per DNA equivalent In dark-grown seedlings, where all five phytochromes are at their highest levels (Figs. and ), the normalized amounts of the phytochromes per 100 mug total extracted protein are measured as: 355 ng of phyA, 39.7 ng of phyB, 8.1 ng of phyC, 5.4 ng of phyD, and 6.1 ng of phyE. Hence, 85% of phytochrome in the etiolated seedling is phyA, 10% is phyB, and 5% is phyC, D, and E combined. In continuous WL-grown seedlings, the measured levels per 130 mug of total extracted protein are: 0.7 ng of phyA, 7.3 of ng phyB, 2.2 ng of phyC, 2.9 ng of phyD, and 4.8 ng of phyE. This represents a large change in photoreceptor content compared with the results from etiolated seedlings, with approximately 5% phyA, 40% phyB, 25% phyE, and 15% each of phyC and phyD. With reference to overall light stability, these values are consistent with the results in Figure obtained with (NH4)2SO4-precipitated extracts. Phytochrome A, as expected, is a highly light-labile or type I phytochrome, whereas among the type II forms, phyB and phyC have intermediate light stability compared with the most light-stable phyD and phyE forms. Phytochrome Levels Are Similar in Commonly Used Arabidopsis Ecotypes | Because various Arabidopsis ecotypes exhibit differences in light-regulated morphological and developmental characteristics and one ecotype, Wassilewskija (Ws), lacks phyD , the levels of the phytochromes in four commonly used ecotypes were examined. Immunoblots of (NH4)2SO4-precipitated extracts of dark-grown 7-d-old seedlings of the Columbia (Col), Landsberg erecta (Ler), Nossen (No-0), and Ws ecotypes probed with the phytochrome-specific MAbs are shown in Figure A. Little variation in the expression levels of these proteins is observed among the lines aside from the lack of phyD in the Ws ecotype. All other experiments reported here were performed on the Col ecotype. Figure 4 | Analysis of phytochrome levels in diverse Arabidopsis ecotypes, in mature rosette plants, and in plant organs. Analysis of phytochrome levels in diverse Arabidopsis ecotypes, in mature rosette plants, and in plant organs. A, Seedlings of Arabidopsis ecotypes Col, Ler, No-0, and Ws were grown in darkness for 7 d, proteins were extracted and precipitated with 25% (w/v) (NH4)2SO4 (for phyA, B, C, and E) or 20% (w/v) (NH4)2SO4 (for phyD), and an amount of precipitated protein equivalent to 300 mug of total extractable protein was fractionated on 6% (w/v) SDS-PAGE gels and blotted. Blots were probed with the MAbs described in Figure . B, Seven-day-old seedlings grown in the dark (D) or under continuous WL (WL7) and rosette-stage plants grown under continuous WL for 14 (WL14) or 21 d (WL21) were extracted. Extracts were precipitated with 25% (w/v) (NH4)2SO4 (for phyA, B, C, and E) or 20% (w/v) (NH4)2SO4 (for phyD), and amounts of precipitated protein equivalent to 300 mug (D) or 390 mug (light-grown) of extractable protein were fractionated and immunoblotted as in A above. C, Leaves, inflorescence stems, and flowers were harvested from soil-grown plants, and roots were harvested from liquid grown plantlets. Proteins were extracted and precipitated with 25% (w/v) (NH4)2SO4. Amounts of precipitated protein equivalent to 240 mug (roots), 390 mug (leaves), 450 mug (stems), or 210 mug (flowers) of total extractable protein were fractionated and immunoblotted as in A above. Phytochrome Levels in Mature Vegetative Tissues and in Plant Organs | To determine whether phytochrome levels change over the transition from seedling to mature plant, the levels of the five phytochrome proteins in 14- and 21-d-old continuous WL-grown Arabidopsis plants were compared with those in 7-d-old dark- and light-grown seedlings. Figure B shows that the phyA protein remains very low, below the detection limit on this blot, throughout this time. The levels of phyB, phyC, phyD, and phyE are not strikingly altered in 14- and 21-d-old plants compared with 7-d-old light-grown seedlings. At 21 d, when the plants were beginning to bolt, all four of these receptor proteins are present at equivalent or, at most, 2-fold reduced levels compared with 7-d-old light-grown seedlings, on a per DNA content basis (see Table ). In Figure C, the relative levels of the five phytochrome proteins in four mature plant organs are compared. Extracts of roots, harvested from 14-d-old liquid-grown plants, and rosette leaves, flowering bolt stems, and whole flowers from 28-d-old soil-grown plants were assayed on immunoblots. In each case, the amount of plant extract loaded on the gel was normalized to the DNA content of the extract, as indicated in Table . After a long exposure to film, the phyA protein is detected weakly in roots and, at very low levels, in flowers. The phyB, phyC, phyD, and phyE proteins are observed in all four plant organs, in each case most abundantly in flowers. Phytochrome Levels during Germination and Early Seedling Development | To determine which phytochromes are present during seed germination and early seedling development, the levels of the five Arabidopsis phytochrome proteins over a time course from imbibed seeds through 5 d growth in the dark and in the light were measured as shown in Figure . Dry seeds were sterilized (approximately 20 min), then imbibed for 1 h shaking in water in the dark. The 0-h time points in Figure are extracts prepared at the end of this imbibition period. The time course was performed in liquid growth medium, in the dark or in continuous light, to facilitate collection of tissue samples. A parallel experiment done on solid growth medium gave similar results (T. Clack and R.A. Sharrock, unpublished data). The first four lanes of all of the blots shown in Figure contain extract from 20 seeds per lane except the light-grown blot probed for phyA, which contains extract from 50 seedlings per lane to increase the detection sensitivity. The fifth lanes in each blot, containing the 120-h samples, were not loaded on a per seed basis but on a DNA equivalent basis, as in Figure , because, by 120 h of development, the quantity of protein in 20 seedlings had increased to the point where it caused the lanes to distort. The stained gel at the bottom of Figure B illustrates the protein loads in each lane of the blots shown in Figure B. Although not shown here, these blots were compared with standard curves of purified phytochrome apoproteins, as in Figures and . The levels of the antigens determined in this way are given below the lanes in the blots in Figure . The level of phyD was not assayed in the dark in these experiments because of the detection of the cross-reacting NPXR protein by MAb 2C1 in dark-seedling direct extracts. Figure 5 | Phytochrome levels in germinating seeds and young seedlings. Phytochrome levels in germinating seeds and young seedlings. A, Seeds were germinated in the dark for 0, 4, 24, or 48 h, and the amounts of protein extracted from 20 seeds were loaded in wells of a 6% (w/v) SDS-PAGE gel. One batch of seeds was allowed to grow in the dark for 5 d, the seedlings were extracted, and 10 mug of total protein was loaded in the fifth lane. Seeds were also germinated in continuous WL for 0, 4, 24, or 48 h, and the amount of protein extracted from 50 seeds was loaded on a gel. In the fifth lane of this gel, 130 mug of protein from 5-d-old light-grown seedlings was loaded. After electrophoresis, the gels were blotted and probed with the anti-phyA MAb 073d. B, Seeds were germinated in the dark or in continuous WL for 0, 4, 24, or 48 h, and the amounts of protein extracted from 20 seeds were loaded on 6% (w/v) SDS-PAGE gels. Batches of seeds allowed to grow for 5 d in darkness or continuous light were extracted and 50 or 65 mug of total protein from these seedlings was loaded in the fifth lane of the respective gels. Duplicate identical gels were run, blotted, and probed with MAbs for phyB (B6B3), phyC (C11 and C13), phyD (2C1), and phyE (7B3). Dilution curves of purified apoprotein standards were run next to the lanes shown here and the levels of the phytochromes were measured by scanning densitometry. These levels are given below each lane (ud, undetectable). A representative blot containing the dark and light protein samples and stained with copper stain is shown at the bottom of B to illustrate the relative protein loads. A blot of the dark extracts was not probed for phyD level because of the presence of a cross-reactive protein to that antibody in the dark. Figure A shows that phyA is undetectable in this assay in seeds at the end of the sterilization and 1-h imbibition period (0-h lanes). The seeds used in this experiment were harvested from plants grown in continuous light, and whether phyA accumulates in seeds that are allowed to develop in darkness or under a short photoperiod was not investigated. In the dark, the level of phyA becomes detectable by 4 h after imbibition and rises dramatically over the 5-d time course to reach a level of 230 ng 100 mug-1 extracted protein at d 5. In the light, phyA is detectable by 24 h and reaches a level of 0.5 ng 130 mug-1 extracted protein by d 5 (Fig. A). These results are consistent with those in Table for 7-d-old seedlings, assuming the level of phyA in the dark continues to rise as the seedlings get older. Phytochromes B, D, and E are all present at significant levels (1 --3 ng seed-1) in freshly imbibed seeds, whereas phyC is 5- to 10-fold lower (Fig. B). On a per seed basis, phyB and phyC increase slightly over the first 48 h of germination and seedling development in the dark but are relatively constant in continuous light. The levels of phyD and phyE do not change over early seedling development. By d 5, the measured levels of all of the phytochromes in Figure are consistent with those measured in 7-d-old seedling in Table . Diurnal and Circadian Cycling of Phytochromes | Studies using phy null mutants have demonstrated that phyA, phyB, phyD, and phyE have roles in sensing the input light signals to the circadian clock . In addition, pronounced circadian cycling of the activities of the PHYA, PHYB, PHYD, and PHYE promoters has been observed . To determine whether the levels of the Arabidopsis phyA, B, C, and E apoproteins cycle in a diurnal fashion, seedlings were grown for 5 d under a 12-h WL/12-h dark photoperiod (LD) and harvested at 3-h intervals throughout the 5th d. Protein extracts were prepared, precipitated with 25% (w/v) (NH4)2SO4, and analyzed by immunoblotting. The levels of phyD were not followed in these experiments because of the complication of detection of the NPXR protein by the 2C1 MAb. Figure A shows that over a single LD cycle, clear diurnal cycling of the levels of phyA and phyC is evident. Low amplitude cycling of phyB is also indicated, whereas the level of phyE remains constant. The data in Figure are representative of two independent repetitions of this experiment, both of which showed qualitatively similar responses of the four phytochrome proteins. The potential circadian regulation of cycling of the phytochromes was also analyzed. Figure B shows steady-state levels of phyA, B, C, and E over a 48-h time course in seedlings that were grown for 5 d under a 12 WL/12 dark photoperiod and transferred to continuous WL. In no case was strong circadian cycling of a phytochrome observed. Again, two independent experiments were performed and data from one experiment are presented in Figure B. In both experiments, cycling of phyA at very low protein levels, close to the detection limit of the assay, and cycling of phyC with a peak in the subjective dark and progressive damping over the 48-h time course were seen (Fig. B). In these experiments, phyB and phyE did not show significant circadian cycling. Figure 6 | Diurnal and circadian cycling of phytochrome protein levels. Diurnal and circadian cycling of phytochrome protein levels. A, Seedlings were grown for 5 d under a 12-h WL/12-h dark photoperiod and harvested at 3-h intervals over the course of the 5th d. Proteins were extracted and precipitated with 25% (w/v) (NH4)2SO4, and amounts of precipitated protein equivalent to 400 mug of extractable protein were fractionated and immunoblotted. Blots were probed with MAbs for phyA (073d), phyB (B6B3), phyC (C11 and C13), and phyE (7B3). X-ray films of chemiluminescence signals were scanned and quantitated by densitometry, and these data are presented graphically in the lower panel. B, Seedlings were grown for 5 d under a 12-h WL/12-h dark photoperiod and transferred to continuous WL. Seedlings were harvested at 3-h intervals, proteins were extracted and precipitated with 25% (w/v) (NH4)2SO4, and these proteins were fractionated, blotted, probed with antibodies, and quantitated as in A. DISCUSSION : Light is a critical environmental effector of plant growth and development, and it readily penetrates aboveground tissues, so there are few barriers to its detection in cells . Therefore, the localization and the physical abundance of a given photoreceptor are likely to be important determinants of its biological activity. In support of this, the expression level of the phyA and phyB phytochromes has been shown to have a quantitative correlation with light responses, both in overexpression assays and in heterozygotes of null mutations . This suggests that the level of phyA or phyB Pfr formed in the plant is a limiting factor in signaling in these photosensing pathways and that the levels of the phytochrome photoreceptors themselves are critical determinants of photoresponsiveness. We have used type-specific MAb to the five Arabidopsis phytochromes to assess the levels of these molecules in tissue extracts. This analysis does not address the spectral or biological activity of the phytochrome proteins detected, only the abundance of the apoproteins. However, given that overexpressed phytochrome apoprotein in Arabidopsis is converted to correspondingly high levels of spectrally active photoreceptor , it is likely that the phytochromobilin chromophore is not limiting in most plant cells and, therefore, that the estimates derived here represent biologically active phytochrome levels. The Assay | Quantitative analysis of protein levels in plant tissues by immunoblotting is influenced by several considerations including the specificity of the antibodies used, the purity of the proteins used to generate standard curves, and the efficiency of the extraction method. The specificity of the MAb used here has been extensively demonstrated . The apoprotein preparations used as standards were highly overexpressed in E. coli cultures and cut out and eluted from denaturing gels. Compared with other methods of preparation we tried, such as affinity chromatography approaches, these samples consistently showed less contamination with extraneous proteins in silver-stained gels. Nevertheless, the concentrations of these phytochrome standards, as determined by protein assay, doubtless reflect small amounts of contaminating protein in each sample. These considerations are likely to affect the quantitation of the phytochromes in relatively small and consistent ways such that the absolute values of nanograms of phytochrome per microgram of protein derived here are very good estimates of their actual levels and the levels of the five forms relative to each other are quite accurate. We compared several different protocols for preparation of tissue extracts, settling on the method that gave the most consistent and highest recoveries. However, the extraction buffer used in this method does not contain detergent, making it possible that some phytochrome pellets with cellular debris in the centrifugation step of the protocol. Moreover, it has been shown that phyA and phyB are translocated from the cytoplasm to the nucleus in the presence of light , raising the additional possibility that the extractable amounts of some of the phytochromes could be reduced in light-grown compared with dark-grown tissue as a result of association with pelletable nuclear material (see phyB and phyC in Fig. ). To address this, experiments were performed in which the pelleted debris from extractions of dark- and light-grown seedlings were taken up in SDS buffer, boiled, and analyzed on immunoblots. Negligible amounts of any of the phytochromes were recovered from the pellets from either tissue (T. Clack and R.A. Sharrock, unpublished data). In addition, experiments in which dark-grown and light-grown seedlings were directly extracted into hot SDS buffer showed similar light-associated reductions of phyB and phyC to those presented here (T. Clack and R.A. Sharrock, unpublished data). Hence, although it remains possible that some phytochrome is nonextractable using current protocols and that this could be differential between dark and light growth conditions, we have no evidence that this has significantly influenced our results. The amount of each protein extract loaded on the gels used for the immunoblots performed here was normalized to the DNA content of the extract. In doing this, we attempted to normalize the very different tissues assayed to genome equivalents and, in a crude way, to cell numbers. This approach could be compromised by extensive endoreduplication of the nuclear genome or by large changes in the numbers of plastid genomes per cell under different light conditions or at different stages of plant growth. Currently available evidence indicates that, although these parameters vary somewhat in different cell types and at various stages of plant growth, the differences are not large. For instance, using quantitative PCR to follow replication of nuclear, mitochondrial, and chloroplast DNA during leaf growth, showed that these DNA molecules exhibit highly parallel rates of replication during growth of cotyledons and fifth true leaves in planta, so that the ratio of the three genomes remains constant throughout leaf development. Endoreduplication of the nuclear genome is observed in many cell types, especially those undergoing differentiation and expansion , and an extra round of endoreduplication often occurs in hypocotyl cells of dark-grown seedlings compared with light-grown seedlings . This differential cellular polyploidization is phytochrome regulated but affects only some of the cells and leads to only a small change in the overall DNA to cell ratio of this tissue . In our experiments, protein to DNA ratios of seedlings and isolated organs vary only 2-fold between the lowest, flowers, and the highest, stems (see Table ), whereas the protein content per gram fresh weight varies 7-fold. Phytochrome Contents of Dark-Grown and Light-Grown Seedlings | Given the relative constancy of cellular DNA contents, we believe that normalizing to DNA content of extracts is a meaningful way of comparing values from different tissues or growth conditions. Moreover, by correcting the data for other measured or estimated parameters, these results can be converted to other units. The data in the first column of Table are taken from Table and are expressed as nanograms of phytochrome per 100 mug of extracted protein from dark-grown seedlings and nanograms of phytochrome per 130 mug from light-grown (WL) seedlings. We recover 3.7 mug protein/dark-grown seedling and 9.6 mug protein/WL seedling, so the values in the first column also represent nanograms of phytochrome per 27 dark-grown seedlings and nanograms of phytochrome per 13.5 WL seedlings, and the amount of phytochrome/seedling can be directly calculated . We determined the wet weights of both dark-grown and WL seedlings as 0.8 mg seedling-1 and, therefore, the amount of phytochrome per gram fresh weight can also be directly calculated . Table III | Conversion of nanograms of phytochrome perDNA equivalent in 7-d-old seedlings to other units To estimate the content per cell of each of the phytochromes, four assumptions have been made: (a) that the five phytochromes in the plant are each evenly distributed among all the cells, (b) that the density of dark-grown and WL tissues can be approximated as that of water (1 g cm-3), (c) that an average cell of a light-grown Arabidopsis plant can be roughly estimated as having the dimensions 20 x 20 x 15 mum, giving a volume of 6,000 mum3 , and (d) that the DNA per gram fresh weight values in Table are acceptable indicators of relative cell numbers. All of these assumptions clearly represent simplifications and first approximations but, with that in mind, they yield a useful preliminary picture of phytochrome distribution. Using a density of 1 g cm-3, the 0.8 mg weights of dark-grown and WL seedlings can be converted to volumes of 0.8 mm3 seedling-1. The DNA per gram fresh weight values in Table indicate that there are 2.3 times as much DNA, and, therefore, 2.3 times as many cells per gram fresh weight in WL cells as dark-grown cells. Because dark-grown and WL seedlings weigh the same and are therefore assumed to have similar volumes, this indicates that there are 2.3 times as many cells per seedling in WL versus dark-grown tissue and, conversely, that dark-grown cells are 2.3 times the size of WL cells. Hence, WL cells are estimated to be 6,000 mum3 and dark-grown cells 14,000 mum3, and seedlings with volumes of 0.8 mm3 are estimated to contain, respectively, 1.3 x 105 cells in WL and 6 x 104 cells in dark-grown seedlings. These values allow the calculation of picograms of phytochrome per cell and, with an average dimeric molecular mass of 250,000 D, number of molecules per cell for each of the phytochromes . These calculated values indicate that phytochromes are present at low to moderate levels in cells, levels similar to those of hormone and growth factor receptor proteins in animals. Levels of the Five Phytochromes in Seeds, Seedlings, and Mature Plant Organs | Our results indicate that the original distinction between light-labile type I phytochromes and light-stable type II phytochromes is valid and useful in its broad outlines. We have shown that: (a) phyA is a uniquely light-labile type I form, which is clearly the most strongly regulated in its expression level; (b) among the four light-stable type II forms, phyB and phyC are significantly reduced upon extended exposure to R light, whereas phyD and phyE are nearly unaffected; (c) the four type II phytochromes are present quite constitutively, being detectable in all samples from imbibed seeds to inflorescence stems and flowers, although significant alteration in some phytochrome levels, such as those of phyB and phyC during early seedling growth, does occur; and (d) in most of the light-grown tissues examined here, phyB is the most abundant of the four type II phytochromes, consistent with its predominant role in the most obvious light-regulated traits. These observations do not markedly change our concept of the phytochrome family. However, they provide a set of quantitative reference values from which physiological, genetic, and molecular experiments on phytochrome function can be interpreted. The varied responses of the five phytochrome protein levels to light indicate that there is a gradation of light stability in this receptor family, which along with the spectral attributes of the different forms, such as the extent and rate of dark reversion , may have an important influence on their respective activities. It has long been recognized that the rapid and strong down-regulation of phyA by light may reflect a role for this phytochrome as an antenna for very low fluence light signals. That phyB and phyC are down-regulated to a much lower extent than phyA but in similar fashion to each other indicates that alteration of receptor stability by light may be a general mechanism for controlling the activities of individual phytochromes. It is notable that the level of phyC is reduced severalfold in phyB null mutants, both in dark- and light-grown seedlings . These observations and those described here suggest that phyB and phyC may be subject to coordinated turnover. Using null mutants, phytochromes A, B, and E have been shown to mediate the induction of seed germination by R and FR irradiation and to strongly influence several early seedling growth responses . Roles for the other two phytochromes in germination have not been described and, although phyD mutants are defective in early seedling hypocotyl elongation and cotyledon opening responses , little has been described relating to the roles of phyC or phyE in early seedling growth. In earlier studies, phyA was not detectable on immunoblots of freshly sterilized Arabidopsis seeds or after a 24-h imbibition but was detectable in 2-d-old dark-grown seedlings, whereas phyB was present in seeds at a level similar to that in etiolated seedlings . This induction of phyA synthesis correlated with the development of photoresponsiveness of seed germination in a phyB mutant line . We also observe that phyA is not detectable in seeds, however it is faintly detectable after 4 h of dark growth, is clearly present after 24 h, and continues to accumulate as the seedling grows. In addition, we find that phyB, D, and E are present at significant levels in freshly imbibed seeds whereas phyC is 5- to 10-fold lower. These phytochromes are either maintained at those levels (phyD and phyE) or accumulate slowly (phyB and phyC) over the course of seedling development, irrespective of whether the seedlings develop in the dark or in the light. The level of phyE is higher in seeds relative to the other phytochromes than at any other time we assayed, perhaps consistent with its novel role in R/FR light control of germination . Circadian Regulation of Phytochrome Levels | Diurnal and circadian cycling of the activities of the PHYA, PHYB, PHYC, PHYD, and PHYE promoters, at levels of from 2- to 8-fold, has been observed using both PHY-LUC reporter genes and analysis of the endogenous mRNA levels . In those experiments, the PHYC promoter showed a low amplitude rhythm under diurnal LD conditions compared with the others and weak cycling under continuous light or continuous darkness circadian conditions. For all of these promoters, the peak of LUC activity and PHY mRNA abundance was seen during the light phase under light-dark cycles. This contrasts with our findings in that we observe significant diurnal cycling of the phyA and phyC apoproteins and weak cycling of phyB but no significant oscillation, even in continuous light, of phyE. Because of cross-reactivity of our anti-phyD MAb, we did not investigate the cycling of phyD. Moreover, the peaks of the oscillations in the phyA, B, and C apoproteins occur in the dark phase, when the mRNA levels are seen to be low . This suggests that, although there are fairly robust responses of PHY promoter activities to light cycles, oscillations in the PHY mRNAs are not rapidly converted to corresponding oscillations in protein levels. The diurnal cycling of phyA, B, and C may more closely correlate with light effects on their stabilities than on regulation of their biosynthetic capacity. The identification of the F-box containing family of ZTL/ADO/LKP/FKF proteins as components of the light input pathway to the clock and the physical interaction of ZTL/ADO1 with phyB reinforce this possibility. Under LL conditions, which test for circadian regulation, we observe very weak cycling of phyA and phyC, indicating that the mechanisms controlling the levels of these receptors are responsive to the circadian clock. Comparison with Previous Studies | A pool of anti-phyB MAbs that are now known to cross-react with phyD and the same anti-phyC MAbs used here were used previously to evaluate the light stability of these two proteins . At that time, it was concluded that phyB and phyC were light stable in comparing dark-grown tissue levels to tissue that had been exposed to 24 h of R light. However, in contrast to this, a later study showed that phyB was reduced 2-fold by 2 h after a 2-min R-light pulse . Consistent with this later work, the results shown in Figures through here indicate that phyB and phyC are significantly light-labile compared with highly stable phyD and phyE. With reference to this, presented evidence that the PHYB and PHYC mRNA levels are unchanged as a fraction of total RNA in dark-grown compared with WL-grown seedlings, suggesting that the down-regulation of the phyB and phyC proteins likely occurs posttranslationally. The levels of the phyB, D, and E proteins in seeds, seedlings, isolated Arabidopsis organs, and over the time course of plant development determined here correlate well with the respective promoter activities of these genes measured in seedlings using RNA blots or promoter-beta-glucuronidase fusion transgenes . In contrast, PHYA-beta-glucuronidase promoter activity and the level of PHYA mRNA are higher in the light, relative to the other PHY genes, than would be expected for the low abundance of the phyA protein. This indicates that the phyA level may be controlled posttranslationally in light-grown plants by continuous light-induced proteolytic turnover of this phytochrome form. Previous efforts at quantifying the different phytochromes in plant tissues have most frequently been performed in peas or oats and are, in general, consistent with our current findings. used immunoblotting and ELISA techniques with antibodies selective for phytochrome I (phyA) and phytochrome II (phytochrome purified from light-grown tissue) and standard curves of immunopurified phytochromes I and II to determine the levels of these antigens in dissected pea axes. They found approximately 7 ng of each phytochrome per axis in unimbibed seed and observed increases to 200 and 50 ng axis-1, respectively, for phytochrome I and II after 12 h of imbibition of axes in the dark. The levels of three immunologically distinguishable phytochrome apoproteins in the embryo-containing portions of unimbibed oat seeds and in the shoot, scutellum, and root of 3-d-old oat seedlings were determined more recently by comparing immunoblot signals from tissue extracts to standard curves of each of the partially purified phytochromes . The three oat phytochromes were called the 123-, 124-, and 125-kD forms, with the 124-kD form corresponding to phyA. In the dissected embryo portions of oat seeds, the 123-, 124-, and 125-kD phytochromes were measured at 6.1, 1.6, and 1.4 ng embryo-1, respectively . In etiolated seedlings, the 123-, 124-, and 125-kD proteins were measured at 3.2, 83, and 1.5 ng per 100 mug total extracted protein and in light-grown seedlings at 2.2, <0.5, and 0.5 ng 100 mug-1 total protein . The values determined in these studies were clearly derived from different plant species and from different tissues than those presented here. In most cases, the measured quantities of phytochrome were also expressed in units that are not directly comparable with the units used here. Nevertheless, the values are in the same ranges as those we observe, the light lability of oat phyA is of a similar order to that of Arabidopsis, and one of the "light stable" oat phytochromes, the 125-kD form, exhibits a significant reduction in the light relative to the dark, as we describe for Arabidopsis phyB and phyC. CONCLUSION : The opportunity in Arabidopsis to monitor the levels of all of the members of the phytochrome family is unprecedented, and a comprehensive description of the distributions of these photoreceptors in this model plant will aid in analysis of their functions and mechanisms. Our current findings support an overall picture of a broadly distributed photoreceptor array for R- to FR-light signals in seeds and mature plant organs of Arabidopsis and the functioning of two general types of phytochrome, a single light-labile type I phyA form and multiple light-stable type II forms. The quantitative results presented here are consistent with previous work in both the distantly related dicot pea and the monocot oat. Hence, it is likely that these observations are, in their general implications, applicable across plant genera. MATERIALS AND METHODS : Plant Materials and Growth Conditions | The Arabidopsis Col, Ler, No-0, and Ws ecotypes were obtained from previously described sources . For seedling tissues, seeds were surface-sterilized for 30 min in 15% (v/v) bleach/0.2% (w/v) SDS, rinsed at least five times with sterile water, and plated on sterile filter papers overlaying Murashige and Skoog basal agar medium (Sigma, St. Louis) containing 2% (w/v) Suc and 0.5 g L-1 MES in 100- x 25-mm petri dishes. Plates were kept in the dark at 4C for 2 to 3 d, treated for 2 h with WL to induce germination, and then placed either in complete darkness or under the described light condition. Light sources for seedling growth were WL from a bank of 40-W cool-white fluorescent bulbs (10 W m-2) or R light from fluorescent bulbs (20WT12/2364, Sylvania, Danvers, MA) filtered through plastic (Roscolene no. 823, Musson Theatrical, Santa Clara, CA; 2.0 W m-2). Radiant output was determined using an IL1400A radiometer equipped with an SEL033/F/W detector (International Light, Newburyport, MA). Rosette and reproductive stage plants were grown on soil under continuous fluorescent light at 20C in a growth chamber (Conviron, Pembina, ND). Roots were obtained from sterile plantlets grown shaking slowly in liquid Murashige and Skoog medium with 2% (w/v) Suc and 0.5 g L-1 MES under fluorescent lights for 2 weeks. For circadian studies, seeds were sterilized and plated as above and grown in a light-sealed growth chamber at 20C under a 12-h fluorescent light (35 W m-2)/12-h dark photoperiod. Mature plants were grown on soil under continuous WL at 20C. Extraction Conditions and Measurement of Protein to DNA Ratios | The following protocol was adapted from that of . Seedlings, mature plants, or organs were harvested and immediately frozen in liquid nitrogen. For each sample, 300 mg of tissue was ground to a fine powder under liquid nitrogen in a mortar and pestle, and aliquots of approximately 100 mg were transferred to two tubes and carefully weighed. To one tube, 1.8 mL of protein extraction buffer (50 mm Tris-HCl, pH 8.5, 5 mm EDTA, 75 mm (NH4)2SO4, and 25% [v/v] ethylene glycol) was added, and the tube was vortexed for 2 min and then centrifuged at 12,000g for 10 min. The protein concentrations of 20-, 40-, 60-, 80-, and 100-mul aliquots of the supernatant were determined using bovine serum albumin as a standard. To the other tube of ground tissue, 0.6 mL of DNA extraction buffer (200 mm Tris-HCl, pH 7.6, 5 mm EDTA, 250 mm NaCl, and 0.5% [w/v] SDS) was added, the tube was vortexed for 2 min, 0.6 mL of phenol/chloroform/isoamyl alcohol 50:49:1(PIC) was added, and the tube was vortexed for an additional 10 min. The samples were centrifuged (12,000g for 5 min) at room temperature, and the aqueous phase was removed. The interface and PIC fraction from this first extraction was re-extracted with an additional 0.6 mL of DNA extraction buffer and the two aqueous phases were mixed. Two volumes of ethanol were added, and nucleic acids were allowed to precipitate overnight at -20C. The precipitate was collected by centrifugation at 20,000g for 30 min, excess ethanol was aspirated, and the pellet was air-dried briefly. The pellet was dissolved in 0.5 mL of TCS (10 mm Tris, pH 7.5, 05 mm CaCl2, and 1% [w/v] SDS) and transferred to a clean microtube. Proteinase K was added to 40 mug mL-1, and the samples were incubated at 50C for 30 min. Samples were extracted with PIC, the aqueous phase was collected, and nucleic acids were precipitated with ethanol. The precipitate was collected as described above and dissolved in 100 mul of water. The DNA concentrations of 10-, 20-, and 30-mul aliquots of the sample were measured by staining with Hoechst dye 33258 and determining fluorescence with a Hoefer TKO fluorometer, using calf thymus DNA as a standard. Protein Extraction and Electrophoresis, Immunoblotting, and Quantification of Immunoblots | For ammonium sulfate-precipitated protein fractions (Figs. , , and ), extracts were prepared and precipitated with 25% or 20% (w/v) (NH4)2SO4, as described . For direct extracts (Figs. , , and ), 0.25 g of frozen tissue was briefly ground in liquid nitrogen in a mortar and pestle, 0.5 mL of 2x protein extraction buffer plus protease inhibitors (100 mm Tris-HCl pH 8.5, 10 mm EDTA, 150 mm (NH4)2SO4, 50% [v/v] ethylene glycol, 2 mug mL-1 aprotinin, 1 mug mL-1 leupeptin, 1 mug mL-1 pepstatin, 2 mm phenylmethylsulfonyl fluoride, 10 mm iodoacetamide, and 5 mug mL-1 NaHSO3) was added, and the samples were ground for 2 min in the mortar and centrifuged for 5 min at 12,000g in a microfuge at 4C. A sample of the supernatant was removed, and the protein concentration was determined . The remaining supernatant was mixed with an equal volume of 2x SDS-PAGE sample buffer and frozen in liquid nitrogen. Protein extracts were fractionated on SDS-polyacrylamide gels, blotted to nitrocellulose, and probed with MAb as described . Phytochromes were detected using the Supersignal West chemiluminescence kit (Pierce, Rockford, IL), exposed x-ray films were scanned on a Fluor-S Multiimager (Bio-Rad, Hercules, CA), and stored image files were analyzed using the Quantity One software package (Bio-Rad). MAbs were anti-phyA 073d, anti-phyB B6B3, anti-phyC C11 and C13, anti-phyD 2C1, and anti-phyE 7B3 . Preparation of Purified Phytochrome Apoprotein Standards | The five Arabidopsis phytochrome apoproteins were expressed in Escherichia coli BL21 (DE3) from full-length cDNA sequences cloned in the pET3c vector as described previously . Cultures of E. coli carrying the pET3c-phyA, B, C, D, or E constructs were induced with isopropylthio-beta-galactoside for 4 h, and the cells were collected by centrifugation and suspended in 50 mm Tris-HCl, pH 8.2, 2 mm EDTA. For each cell suspension, lysozyme was added to 0.1 mg mL-1 and Triton X-100 to 0.1% (v/v), and the suspension was incubated for 15 min at 30C. The suspension was chilled on ice, sonicated, and centrifuged for 15 min at 12,000 rpm. The pellet containing inclusion bodies was washed three times with 0.5% (v/v) Triton X-100, 1 mm EDTA, dissolved in 50 mm Tris-HCl pH 8 m urea, and centrifuged at 20,000 rpm for 15 min to pellet any debris. This preparation was fractionated on a preparative 6% (w/v) SDS gel, proteins were visualized with GelCode E-zinc stain (Pierce, Rockford, IL), and the full-length phytochrome apoprotein band was excised with a razor blade. The excised phytochromes were electroeluted into 25 mm Tris-HCl, pH 8, 192 mm Gly, and 0.01% (w/v) SDS. Protein concentration was determined using a detergent-compatible protein assay (Bio-Rad). For immunoblots, the indicated amounts of the purified phytochrome apoproteins were fractionated on SDS gels in the presence of approximately 10 mug of carrier HELA cell protein. This carrier protein was prepared from cells grown as a monolayer in Dulbecco's modified eagle medium containing 10% (v/v) fetal calf serum (Invitrogen, Carlsbad, CA). Confluent plates of HELA cells were washed with PBS and trypsinized, and the cells collected. The cell pellet was suspended in SDS sample buffer, heated at 95C for 5 min, and centrifuged at 12,000g for 10 min. These carrier protein preparations were shown to contain no proteins that cross-react with the anti-phy MAbs and were included to aid in the loading of the small amounts of the phytochrome apoprotein standards onto the gels and their electrotransfer to the nitrocellulose membrane. Backmatter: PMID- 12226524 TI - Characterization of a Strong Dominant phytochrome A Mutation Unique to Phytochrome A Signal Propagation AB - Here, we report the isolation and characterization of a strong dominant-negative phytochrome A (phyA) mutation (phyA-300D) in Arabidopsis. This mutation carries a single amino acid substitution at residue 631, from valine to methionine (V631M), in the core region within the C-terminal half of PHYA. This PHYA core region contains two protein-interactive motifs, PAS1 and PAS2. Val-631 is located within the PAS1 motif. The phyA-V631M mutant protein is photochemically active and accumulates to a level similar to wild type in dark-grown seedlings. Overexpression of PHYA-V631M in a wild-type background results in a dominant-negative interference with endogenous wild-type phyA, whereas PHYA-V631M in a phyA null mutant background is inactive. To investigate the specificity of this mutation within the phytochrome family, the corresponding amino acid substitution (V664M) was created in the PHYTOCHROME B (PHYB) polypeptide. We found that the phyB-V664M mutant protein is physiologically active in phyB mutant and causes no interfering effect in a wild-type background. Together, our results reveal a unique feature in phyA signal propagation through the C-terminal core region. Keywords: Introduction : One of the most important environmental factors affecting plant growth and survival is light. Plants employ arrays of photoreceptors to detect and respond to a broad spectrum of light . One of the best-studied groups of photoreceptors is the red (R)/far red (FR) light-absorbing phytochromes . In Arabidopsis, there is a family of five phytochromes whose apoproteins are encoded by a multigene family (PHYA --PHYE; ; ). Among the five phytochromes, the best characterized is the FR-absorbing phytochrome A (phyA). Like all phytochromes, phyA is capable of a photoreversible conformational change between the Pr and the Pfr forms in response to light. Whereas the Pfr form is rapidly degraded, the Pr form of phyA is stable in the cytosol. Upon exposure to R light, the Pr form is converted to the Pfr form, whereas exposure to FR light converts the Pfr form back to the Pr form. This photoconversion correlates with the absorption spectra of these forms, where Pr absorbs maximally at 666 nm and Pfr absorbs maximally at 730 nm . It is currently hypothesized that both the Pfr and the photocycled Pr forms of phyA are the active species and possess kinase activity . Upon light perception, phyA and phytochrome B (phyB) migrate into the nucleus, where they presumably trigger downstream events . Since the discovery of the phytochrome system, the search has been on to identify the signal transduction mechanism through which light perception is coupled to changes in gene expression and cell physiology that control growth and development. Use of both biochemical and genetic means have identified candidate downstream-signaling components of the phytochrome pathway. Microinjection into hypocotyl cells of the phytochrome-deficient aurea mutant of tomato (Lycopersicon esculentum) has been used to biochemically assay the activities of various pharmacological agents and putative signaling intermediates . Such studies suggest that heterotrimeric G proteins Ca2+-calmodulin and cyclic-guanosine-5' monophosphate may mediate phytochrome-induced responses. A reverse genetic approach has recently implicated the involvement of the only G-alpha protein in Arabidopsis in light inhibition of hypocotyl elongation . Both yeast (Saccharomyces cerevisiae) two-hybrid protein-protein interaction screens and mutational approaches have resulted in identification of over a dozen putative downstream components and interacting partners of phytochromes (for review, see ; ). Those putative signaling components include both cytoplasmic and nuclear factors, some of which are transcription factors directly acting on light-responsive promoters. For phyA, the majority of the signaling components identified so far are nuclear localized. This is consistent with the view that light-activated phyA rapidly migrates into nucleus and triggers most of the signaling events there. A plethora of studies have revealed that phytochrome molecules contain multiple functional domains (for review, see ; ). The N-terminal half contains the chromophore attachment site and is responsible for the observed spectrum specificity of phyA and phyB. The C-terminal half seems to be responsible for dimerization and downstream signal transmission . A region in the C-terminal half, designated as the "core region" or "Quail box," was revealed to be critically important for phytochrome signal propagation . This core region in phyA contains two PAS motifs, presumed to be involved in protein-protein interaction. This region was shown to be important for interaction with the reported phytochrome-interactive partners . Although it was reported that the C-terminal half of phyA and phyB could be interchangeable for some R and FR responses , some aspect of distinction in the phyA and phyB signal propagation through their C-terminal core regions to downstream targets might be expected because there are a number of unique downstream components specific to either phyA or phyB. Here, we report that in a mutant screen for an etiolated phenotype under continuous FR light (FRc), we have recovered a strong dominant-negative mutation of PHYA. This dominant interfering phenotype observed in the heterozygous state is even stronger than those previously reported for transgenic seedlings overexpressing phyA dominant-negative mutant forms . Characterization of this new phyA mutant revealed an insight specific to phyA-signaling propagation through its C-terminal domain. RESULTS : Isolation of phyA-300D, a Strong Dominant-Negative Allele of PHYA | In an effort to isolate mutants impaired specifically in phyA signaling, long hypocotyl mutant seedlings specific for FRc were screened from several available T-DNA insertion mutagenized collections. Seven elongated hypocotyl mutants were identified . Genetic complementation tests revealed that most of the mutations defined new alleles of known genes; however, one new locus, designated FHY4, was defined by a single-mutant allele (fhy4-1). Table I | Summary of the mutant screen One of the mutants, now designated phyA-300D, exhibited a 3:1 mutant:wild-type phenotype segregation in a backcross to wild type . This result suggests that the phyA-300D mutation behaves as a dominant trait. This was further supported by the fact that during our allelism test between phyA-300D and various known phyA-signaling mutants, the F1 progenies always displayed long hypocotyl phenotypes, contrasting to those of a normal recessive phyA allele (Table ; data not shown). In the F2 populations, it was revealed that phyA-300D complements all other mutations except phyA, indicating that this may be a new allele of PHYA. As shown in Figure A, seedlings heterozygous for the phyA-300D allele possess long hypocotyls under FRc, similar to that of the homozygous seedlings. However, phyA-300D seedlings grown under other light wavelengths displayed normal de-etiolated phenotypes, suggesting that the phyA-300D phenotype is FRc specific (Fig. B). Table II | Complementation analysis of phyA mutants Figure 1 | Phenotype characterization of the dominant phyA-300D mutant. Phenotype characterization of the dominant phyA-300D mutant. A, Three-day-old FRc-grown phyA-300D homozygote (-/-), heterozygote (-/+), and wild-type (WT) seedlings. Average hypocotyl lengths (in millimeters) are shown above each seedling (n = 25) with sd shown in parentheses. B, Hypocotyl length of wild-type (WT) and phyA-300D seedlings grown in continuous white (W) and blue (B) light and Rc (R) and FRc (FR), or in darkness (D) for 6 d. Each column represents a mean value of the hypocotyl length (in millimeters) taken from 25 seedlings with the sd indicated by an error bar. To further verify this dominant mutant as a phyA allele, mapping analysis of the phyA-300D mutation was carried out (see "Materials and Methods"). As expected, phyA-300D was mapped to chromosome 1 near the PHYA locus (data not shown). phyA-300D Is Caused by a Single Amino Acid Change at Residue 631 | To determine the exact nature of the phyA-300D mutation, overlapping regions covering the PHYA gene from the mutant background were PCR amplified and sequenced. This sequence analysis revealed two nucleotide changes, both from guanine (G) to adenine (A), affecting two adjacent codons of the PHYA open reading frame . The first G to A mutation is silent, altering the codon 630 from ACG to ACA, both encoding the same amino acid, Thr. The second G to A nucleotide change is a missense mutation, which alters the codon at position 631 from GTG (encoding for Val) to ATG (encoding for Met). Thus, the overall consequence in phyA-300D is a Val to Met change at amino acid 631 (V631M) in the PHYA polypeptide (Fig. C). This Val-631 is located within the first of the two PAS motifs, within the core region or Quail box of the phytochrome C-terminal half . The core region of phyA has been demonstrated to be critical in phytochrome signal propagation to downstream targets . However, all those reported mutations had been described as recessive in nature . Figure 2 | The structural features of phyA and the molecular lesion of the phyA-300D mutation. The structural features of phyA and the molecular lesion of the phyA-300D mutation. A, Diagram of the Arabidopsis PHYA protein. The chromophore-binding site (C-323), core region, and its two PAS motifs are indicated. B, The conserved Val and its surrounding amino acid residues among four Arabidopsis phytochromes. Note that the Val is invariant among all known phyA species from higher plants, although it can be substituted by Ala in all other phytochromes among different higher plant species. The single-letter codes of amino acids are shown in bold. C, Nucleotide sequence comparison of the PHYA gene in phyA-205, phyA-300D, and wild type (WT) around the Val-631 region. Amino acids (in bold single-letter code) are shown above their codons. In both mutant alleles, there is a single identical amino acid residue mutation (V631M) in the translated proteins. phyA-205 Is an Independent Allele of PHYA with the Same Amino Acid Alteration as phyA-300D | It is intriguing that a V631M mutation of PHYA was previously reported for another allele, phyA-205 . However, at that time phyA-205 was described as a recessive mutation with an intermediate phenotype. We further examined the segregation property of the phyA-205 backcross with wild type and our results clearly indicated that it is a strong dominant mutation . Furthermore, phyA-205 failed to complement phyA-300D in both the F1 and F2 populations , as expected from their allelism. As shown in Figure A, seedlings heterozygous for the phyA-205 mutation showed long hypocotyls under FRc similar to that of homozygous mutant seedlings (Fig. A). Further sequence analysis of the phyA-205 mutant confirmed that it contains the reported single G-to-A mutation in codon 631 (; Fig. C). Thus, there are two independent mutations, phyA-300D and phyA-205, both resulting in the same V631M mutation in the PHYA protein (Fig. C). This result is consistent with the conclusion that the V631M mutation caused a dominant-negative phenotype for phyA. Figure 3 | The phyA-205 has a dominant-negative phenotype and is not dependent on Suc concentration. The phyA-205 has a dominant-negative phenotype and is not dependent on Suc concentration. Average hypocotyl lengths (in millimeters) are shown above each seedling (n = 25) with the sd shown in parentheses. A, Phenotypic comparison of 3-d-old FR light-grown phyA-205 homozygous (-/-), heterozygous (-/+), and wild-type (WT) seedlings on 0.3% (w/v) Suc medium. B, Phenotypic comparison of 3-d-old FR light-grown phyA-205 homozygous (-/-), heterozygous (-/+), and wild-type (WT) seedlings on 2% (w/v) Suc medium. Because the original characterization of phyA-205 used growth media (GM) supplemented with 2% (w/v) Suc and our studies used 0.3% (w/v) Suc, we examined the effect of Suc concentration on hypocotyl length in the phyA-205 mutant. As shown in Figure , the high concentration of Suc indeed inhibited hypocotyl elongation under FRc. However, this inhibition was observed in both phyA-205 homozygous and heterozygous seedlings. At either concentration of Suc, the heterozygous and homozygous mutant seedlings have similar hypocotyl lengths under the same FRc growth condition. phyA-300D homozygous and heterozygous mutant seedlings similarly also displayed this Suc inhibition of hypocotyl elongation (data not shown). This effect of Suc was apparently not unique to the dominant mutations, because both phyA-1 and phyA-301 (recessive mutations) displayed a similar decreased hypocotyl length under the higher Suc concentration. The PHYA-V631M Is Sufficient to Confer a Dominant-Negative Interference to phyA-Mediated FRc Inhibition of Hypocotyl Elongation | To test whether a V631M mutation in PHYA is sufficient to confer a FR-specific dominant-negative phenotype, we introduced a 35S promoter-driven PHYA-V631M and wild-type PHYA into wild-type Arabidopsis and a null mutant (phyA-101; Fig. A). When grown in FRc, transgenic seedlings expressing PHYA-V631M (from the phyA-300D allele) exhibited an etiolated phenotype (Fig. B). When grown complete darkness, there is no observable difference between transgenic and wild-type seedlings (Fig. C). However, overexpression of PHYA-V631M in a phyA null mutant did not rescue its mutant phenotype (Fig. D), further confirming the loss-of-function nature of the PHYA-V631M protein. Thus, our result implies that not only did phyA-V631M lose its ability to respond to FR light, but that the presence of PHYA-V631M is sufficient to interfere with endogenous wild-type phyA function in a dominant-negative fashion. Figure 4 | Expression of PHYA-V631M mutant protein is sufficient to cause a dominant interfering effect on endogenous phyA signaling. Expression of PHYA-V631M mutant protein is sufficient to cause a dominant interfering effect on endogenous phyA signaling. A, Diagrams of the transformation cassettes. The wild-type PHYA (WT) and PHYA-V631M full-length coding regions were driven by the cauliflower mosaic virus 35S promoter. B, Morphological comparison of 3-d-old FR light-grown seedlings of wild-type PHYA overexpressor (PHYA OE), wild-type (WT), and two mutant PHYA-V631M overexpressor (PHYA-300D OE) lines. PHYA protein immunoblots are shown in the bottom panel. C, Morphological comparison of 3-d-old dark-grown seedlings with PHYA-V631M overexpressor transgene in wild-type background (left) and the wild-type control (right). D, Morphology of 3-d-old FR light-grown seedling with PHYA-V631M overexpressor transgene in a phyA null mutant background. Scale bars in B through D = 1 mm. Interestingly, the degree of dominant-negative interference of the transgenic plants depends on mutant PHYA-V631M protein dose. As shown in Figure B, a high-expression level of mutant PHYA-V631M resulted in seedlings with an extremely elongated hypocotyl and closed cotyledons, resembling that of the phyA null mutants, whereas a lower expression level resulted in an intermediate phenotype, a long hypocotyl, and open but small cotyledons (Fig. B). As expected, expression of wild-type PHYA resulted in seedlings with short hypocotyls and open cotyledons (Fig. B), similar to the reported phyA overexpressor phenotype . The phyA-V631M Abundance Is Not Reduced in Mutant Seedlings Grown in Darkness But Is Slightly Higher under FR Light | As a first step to determine a possible molecular basis of the phyA-V631M mutation effect, the phyA protein levels of the phyA-300D and wild-type seedlings grown in darkness and in FRc were examined using immunoblot analysis. As shown in Figure , the level of phyA protein in dark-grown phyA-300D seedlings is as abundant as in wild-type seedlings. Interestingly, phyA-300D grown under FRc accumulated to about 2-fold higher levels of phyA protein than that of wild type. This result rules out the possibility that the phyA-300D phenotype is a result of lower PHYA levels under FR light. It should be noted that the dominant mutant phenotype of phyA-300D could be a result of increased levels of the mutant PHYA-V631M protein. However, this hypothesis would be argued against by the observation that the phyA-105 mutant, which has a single amino acid change, accumulated to similar high levels of the PHYA-V631M mutant protein in FRc, and still exhibited completely recessive properties . Figure 5 | Characterization of wild-type and PHYA-V631M stability under R and FR light. Characterization of wild-type and PHYA-V631M stability under R and FR light. A, phyA levels in 3-d-old dark (D) and FR light-grown wild-type Columbia (WT) and phyA-300D, and in 3-d-old dark-grown seedlings exposed to Rc for 1 or 4 h (hr; top). The immunoblots were probed with a phyA-specific polyclonal antibody. An equal amount of total protein was loaded in each lane, as verified by CSN5 immunoblot (bottom). B, phyA levels in 3-d-old dark-grown (D) wild-type Columbia (WT), and phyA-300D seedlings and in 3-d-old dark-grown seedlings exposed to FRc light for 1, 4, or 24 h (hr). An equal amount of total protein was loaded in each lane, as verified by immunoblot using anti-tubulin antibodies (bottom). phyA-V631M Possesses Normal Light Triggered Degradation Kinetics and Photochemical Properties | It is known that phyA accumulates in darkness as the Pr form and is rapidly degraded upon exposure to light, because of the susceptible nature of the phyA Pfr form to proteasome-mediated protein degradation. Therefore, we examined the degradation kinetics of the phyA-V631M mutant protein compared with wild-type phyA protein upon exposure to continuous R light (Rc) or FRc. Dark-grown mutant and wild-type seedlings were exposed to Rc for 0, 1, and 4 h and harvested for immunoblot analysis. The degradation kinetics shown in Figure A indicates that there is no significant difference between the phyA-V631M mutant and wild-type phyA proteins. As expected, FRc treatment of dark-grown mutant or wild-type seedlings does not cause observable degradation of the phyA proteins (Fig. B). This result suggests that the mutant phyA-V631M protein was able to photoconvert from Pr to Pfr upon exposure to Rc, and only the Pfr form of phyA-V631M was labile and subjected to rapid degradation. To test whether phyA-V631M protein retains typical phyA R-FR reversibility, the differential absorbance spectra for phyA-300D and wild-type seedlings were examined. To this end, phyA-300D, phyA-101 (a phyA null allele), and wild-type seedlings were grown in darkness for 3 d and extracts were prepared under green safelight as described in "Materials and Methods." This assay revealed that phyA-V631M has an essentially identical spectral property as wild type (data not shown). The PHYA-V631M Protein Retains Dimerization Capability | The Val residue at position 631 is located in the first PAS motif of PHYA that overlaps with the phyA dimerization region. Thus, we examined the ability of PHYA-V631M to form dimers in vivo. To this end, crude extracts from homozygous mutant and wild-type seedlings grown in darkness or FRc were subjected to native gel electrophoresis followed by immunoblot analysis . The migration and amount of PHYA-V631M dimer from phyA-300D seedlings under native gel conditions are comparable with that of wild-type plants, indicating that the PHYA-V631M can effectively dimerize in vivo. Figure 6 | Native gel analysis of phyA dimerization in FR and dark-grown wild-type and mutant seedlings. Native gel analysis of phyA dimerization in FR and dark-grown wild-type and mutant seedlings. Extracts were separated on a 4% to 20% (w/v) gradient non-denaturing gel and probed with a polyclonal PHYA antibody. The phyA dimer is marked by a triangle on the right. Equal loading was ensured by visualizing a protein band staining intensity (marked by control). The positions of two mass markers are indicated on the left side. The phyA-101 is a null mutant control. A, FR light-grown seedlings (5 d); B, dark-grown seedlings (5 d). Overexpression of PHYB-V664M in Wild-Type Plants Failed to Interfere with phyB Signaling | The PHYA and PHYB polypeptides are highly homologous . The region surrounding the Val-631 of PHYA and the corresponding region in PHYB is highly conserved (Fig. B). For PHYB, its Val residue at position 664 is equivalent to the Val at position 631 of PHYA. Thus, it is of interest to determine whether a corresponding PHYB-V664M mutant protein would confer a similar dominant-negative effect on phyB function. For this reason, we specifically created a corresponding PHYB-V664M mutant transgene under 35S promoter control. The transgene was transformed into wild-type Arabidopsis and a phyB null mutant (phyB-101; Fig. A). Homozygous transgenic lines were obtained from each construct, and their hypocotyl elongation under Rc was examined. Contrasting to PHYA-V631M, the PHYB-V664M mutant protein is clearly functional and active in both the phyB null mutant and wild-type seedlings in the lines tested. As shown in Figure B, PHYB-V664M can rescue the phyB null phenotype when overexpressed in the phyB-101 mutant background. Furthermore, transgenic seedlings overexpressing the PHYB-V664M in a wild-type background exhibited a hyper-photomorphogenic phenotype with shorter hypocotyls than those of wild type (Fig. C), similar to seedlings overexpressing wild-type PHYB (Fig. A). Several PHYB-V664M transgenic lines that showed longer hypocotyl under Rc turned out to be cosuppression lines and had an undetectable level of phyB protein. This result indicates that PHYB-V664M behaves differently from PHYA-V631M. Our results imply that the mutant PHYB-V664M protein neither reduces its ability to respond to Rc, nor is capable of interfering with wild-type phyB signaling. Thus, the role of Val-631 in phyA defined in this work is specific to phyA and not applicable to phyB. Figure 7 | The PHYB-V664M protein is physiologically active and does not confer an interfering effect on phyB signaling. The PHYB-V664M protein is physiologically active and does not confer an interfering effect on phyB signaling. The 3-d-old seedlings are shown on top, and their corresponding PHYB protein immunoblots are shown on the bottom. A, Three-day-old Rc light-grown wild type (WT), PHYB overexpressor (PHYB OE) in wild-type background, and the phyB-101 null mutant. B, Three-day-old Rc light-grown seedlings of three independent PHYB-V664M overexpressor lines (L1, L2, and L3 of PHYB-V644M OE) in the phyB-101 null mutant background. C, Three-day-old Rc light-grown seedlings of two different PHYB V644M overexpressor (PHYB-V644M OE) lines in a wild-type background. L'1 is an overexpressor line for PHYB-V664M, and L'2 is a cosuppression line. DISCUSSION : PHYA-V631M Has a Strong Dominant-Negative Interfering Activity | Several lines of evidence presented in this report definitively show that PHYA-V631M is capable of conferring a strong dominant-negative interfering effect on normal phyA signaling. First, two independent alleles with the same V631M amino acid change exhibited a strong long hypocotyl phenotype under FR in the heterozygous state. The hypocotyl length of heterozygous mutant seedlings is nearly identical to that of their homozygous mutant seedlings. The only distinction between the heterozygous and homozygous seedlings is the slightly more advanced cotyledon development in the heterozygous mutants. The heterozygous seedlings exhibit opening and slight enlargement of the cotyledons after extended growth under FRc. Second, introduction of the PHYA-V631M transgene into wild-type Arabidopsis confirmed that PHYA-V631M is sufficient to confer the dominant-negative interfering effect. Our results imply that not only does the mutant PHYA-V631M protein reduce its ability to respond to FRc, it is also able to interfere with the ability of endogenous phyA to mediate FRc responses in a dominant-negative fashion. The PHYA-V631M is clearly photochemically active but not functional in mediating FRc responses, because its homozygous seedlings exhibited a phenotype essentially indistinguishable from that of the phyA null mutations. Furthermore, overexpression of PHYA-V631M in a phyA null mutant background also failed to show any phenotypic rescue (Fig. D). Although a large number of phyA mutations within the two PAS motifs have been described, all of them were reported as recessive . In fact, only one report described a slightly weak semidominant effect of phyA mutations under low-FR photon fluence conditions and with no sugar in the GM . In this case, heterozygous PHYA/phyA seedlings for two alleles, at least one is a null allele, exhibited slightly longer hypocotyls (3 versus 1 mm in wild type) under FRc but were still significantly shorter than those of phyA homozygous mutants (12 --13 mm) under the same growth condition. However, this weak semidominant effect of phyA mutations is largely growth-condition dependent and was not observed by other researchers using different growth conditions . We also failed to observe a semidominant effect of the same null phyA-1 allele under our growth conditions, whereas the dominant effect observed in PHYA-V631M is very strong and consistent under our experimental conditions. On the other hand, a dominant interfering effect has been reported for transgenic seedlings overexpressing several heterologous phyA species. For example, overexpression of three distinct oat (Avena sativa) phyA mutant forms (major deletions in very N- or C-terminal half) in Arabidopsis resulted in a strong dominant-negative interfering effect over endogenous Arabidopsis phyA function . Overexpression of an amino-terminal deletion of rice (Oryza sativa) phyA in tobacco seedlings also resulted in a dominant interfering effect . Last, overexpression of an intact fern phyA in Arabidopsis also caused a weak dominant interfering effect . A general explanation for those dominant interfering effects is that those mutant phyA molecules (oat or rice) or the too-diverged fern phyA may not be able to mediate a productive interaction with phyA-signaling partners, and thus, titrate out the available phyA-signaling partners. Nevertheless, comparing the degrees of the dominant-negative interfering phenotypes in those overexpressing lines, the phyA-V631M heterozygous mutant seedlings exhibited possibly the strongest interfering phenotype. Furthermore, this strong dominant-negative interfering phenotype in phyA-300D and phyA-205 is caused by a single amino acid substitution. The Role of the PAS-Like Domain in phyA- Signaling Activity | Molecular and biochemical characterization of phyA indicates that the apoprotein folds into two major structural domains: a globular NH2-terminal domain cradling the covalently attached chromophore in a hydrophobic pocket, and a more extended C-terminal domain, with a short flexible hinge region connecting the two . Evidence also indicates that the determinants of the photosensory specificity of phyA and phyB to FRc and Rc reside in the NH2-terminal domains. In addition, the proximal end of the C-terminal domain, designated the core region or Quail box of the polypeptide, contains determinants necessary for regulatory activity or for the effective communication of perceived light signals to the cellular transduction circuitry. It is generally assumed that the phytochrome molecules, rather than directly binding to target gene promoters, relay information via signaling intermediates or directly to DNA-binding transcription factors . The V631M mutation in phyA-300D and phyA-205 falls within the "core region" of the PHYA protein (Fig. A). This structural domain has shown a propensity for light-signaling mutations . The severity of the homozygous phyA-300D mutant phenotype reveals a critical role of the Val residue within this core region for phyA function. This region of the PHYA protein has been shown to be a "hot spot" for missense mutations resulting in regulatory mutants that produce PHYA protein, but cannot transduce the light signal to elicit the photomorphogenic response. However, phyA-V631M is the first case where a strong dominant-negative interfering phenotype has been described, whereas other mutations in this region have been reported as recessive . This core region of PHYA contains two PAS-like motifs and may be involved in phytochrome signal transduction by contacting downstream partners (Fig. A; ). Previous studies have shown that this core region of phytochrome interacts directly with phytochrome partners including NDPK2 and PIF3 , supporting the conclusion that the PAS motifs or the core region defines protein-protein interactive surfaces. One possible explanation of the negative interference of the PHYA-V631M on phyA signaling is that the phyA-V631M mutant protein interacts too strongly with downstream positively acting components of the pathway. These stronger than wild-type interactions with downstream components for PHYA-V631M would compromise the capability of the activated target protein to initiate downstream events. In heterozygous plants, the PHYA-V631M protein would titrate out the downstream phyA target because of the higher affinity between PHYA-V631M and the target, thus, effectively preventing wild-type phyA from interacting with the target and blocking the proper phyA-mediated event. However, this is certainly not the only explanation, and there are other alternative possibilities. One alternative is that PHYA-V631M could have a stronger affinity to form a functionally inactive heterodimer with the wild-type PHYA. In heterozygous plants, this higher affinity to form nonfunctional PHYA-V631M::PHYA heterodimers could prevent the formation of active wild-type phyA homodimers. As a result, the heterozygous plants would fail to respond to FRc. A third possibility is that the PHYA-V631M homodimer and its heterodimer with wild-type PHYA might have an elevated Pfr to Pf form dark reversion rate, which could potentially account for the interfering phenotype. An important task of future work will be to determine which mode of action is used in plants. The phyA-V631M Dominant-Negative Effect Is Not Applicable to phyB | Although there is a high-sequence homology between the Arabidopsis PHYA and PHYB apoproteins surrounding PHYA amino acid Val-631 (Fig. A), we demonstrated that the corresponding Val to Met mutation (V664M) in Arabidopsis phyB does not trigger a similar dominant-negative effect on phyB signaling. The PHYB-V664M mutant protein was active in both phyB null mutant and wild-type plants when introduced into the proper backgrounds. These results indicate that Val-631 is critical for PHYA-mediated FR signaling but not applicable to PHYB-mediated signaling. This specificity to phyA is also consistent with the observation that all known phyA proteins, including those from both monocots and dicots, have the invariant Val at the corresponding position . However, the Val residue is not absolutely conserved in other higher plant phytochromes, and in some cases, the corresponding position has changed to Ala . This observation is consistent with the conclusion that the role of Val-631 in phyA as defined by the V631M mutation is unique to phyA, and not applicable to phyB. Thus, it defines a key distinction in the manner through which phyA and phyB transduce light signal via their C-terminal core regions. MATERIALS AND METHODS : Plant Materials, Mutant Screen, and Mapping | Arabidopsis mutants phyA-1, phyA-101, phyA-205, and phyB-101 (Landsberg erecta ecotype) were described previously . Several available T-DNA insertion mutagenized collections were used, including those from the Arabidopsis Stock Center and Dr. Steven Dellaporta (Yale University; ). Either T2 or T3 seeds were screened for mutant seedlings exhibiting a complete (long hypocotyl and closed cotyledons) or partial etiolated (long hypocotyl and open cotyledons) phenotype after 3 d of growth under FRc. Seed sterilization and plating were essentially the same as previously described except the GM was supplemented with 0.3% (w/v) Suc. To improve germination rate, the seeds on the plates were exposed to 24 h of white light after vernalization for 2 d at 4C in the dark. The light conditions used were: Rc (100 mumol m-2 s-1), continuous blue light (50 mumol m-2 s-1), continuous white light (200 mumol m-2 s-1), and FRc (175 mumol m-2 s-1). Growth chambers (E-30 LED 2/3, Percival Scientific, Perry, IA) were maintained at 22C constant temperature. For mapping, mutant phyA-300D (Columbia ecotype) was crossed to wild-type Landsberg erecta ecotype. F2 progenies were grown on 0.3% (w/v) GM plates and screened under FR for a short hypocotyl, whereas F3 progeny were grown on 0.3% (w/v) GM plates and screened under FR light for a long hypocotyl. DNA from individual homozygous mutant seedlings was prepared and used for PCR-based simple sequence length polymorphism or cleaved-amplified polymorphic sequence mapping . Native Protein Gel and Immunoblot Analyses | Five-day-old FR or dark-grown seedlings were collected under green safelight for protein analysis. Seedlings were frozen in liquid nitrogen and ground with a mortar and pestle in 100 muL of grinding buffer (10% [w/v] glycerol, 400 mm Suc, 50 mm Tris, pH 8.0, and 1 mm phenylmethylsulfonyl fluoride). For native gel analysis, 20 mug of each sample was run on a 4% to 20% (w/v) non-denaturing gradient gel (Tris-Gly, pH 8.0). After electrophoresis, proteins were transferred to Immobilon-P membrane (Millipore, Bedford, MA) for immunoblot analysis. For the PHYA protein level, protein was extracted from 3-d-old dark-grown seedlings and from dark-grown seedlings subjected to Rc or FRc for the specified period of time. Protein concentration was determined by Bradford assay (Bio-Rad, Hercules, CA), loaded equally to SDS-PAGE, and transferred to nylon membrane. The protein-blot analysis was performed according to procedures described . Type-selected antibodies for phyA and phyB were used. Phytochrome Difference Spectrum | Arabidopsis protein extracts were prepared from 6-d-old dark-grown seedlings as described . The extract was dispensed into a spectrophotometer ultra-microcuvette and placed in a FR-LED chamber (200 mumol m-2 s-1) for 360 s to induce photoconversion of phytochrome to the Pr form. After this exposure, the cuvette was placed into the chamber of a spectrophotometer (DU-70, Beckman Coulter, Fullerton, CA), and the baseline was taken. The cuvette was then exposed to R light for 180 s. R light was generated by using a light projector (Ektagraphic III AMT, Eastman Kodak, Rochester, NY) with a Kodak extra bright lamp module. The light passed though a Plexiglas filter (CBS Red 650, Westlake Plastics Co., Lenni, PA) and was reflected into the chamber of the spectrophotometer using a mirror. After R light exposure, an absorbance versus wavelength plot was generated, with readings taken every nanometer from 500 to 800 nm. DNA Sequencing of the PHYA Gene | Genomic DNA was isolated from phyA-300D and phyA-205 mutants. A series of primers was synthesized based on the published PHYA sequence , and PCR was used to amplify the segments of the PHYA gene from the phyA-300D and phyA-205 genomic DNA. PCR products were purified using a PCR purification kit (Qiagen USA, Valencia, CA) and sequenced. Construction of Full-Length PHYA and PHYB Overexpressing Plasmids and Arabidopsis Transformation | Total RNA was extracted from 6-d-old dark-grown seedlings of either wild type or the phyA-300D mutant using Trizol reagent (Invitrogen, Carlsbad, CA). cDNA was synthesized using a reverse transcriptase kit (Advantage RT for PCR kit, CLONTECH Laboratories, Palo Alto, CA). The entire coding region of PHYA was divided into two fragments for PCR amplification. The first fragment used a primer, 5'-GCG TCG ACA TGT CAG GCT CTA GGC CGA C-3', which covers the translation start codon of PHYA with an incorporated SalI cleavage site, and the second primer was 5'-AAT TTT GAG ATC ATT TAG CTT CG-3'. The second fragment used a primer covering the stop codon of PHYA with an incorporated SmaI cleavage site 5'-CGC CCG GGC TAC TTG TTT GCT GCA GCG-3', and the second primer was 5'-GAA TAC CAC ATG GAT TCA ACG-3'. The two PCR fragments were cloned into the pCR 2.1 TOPO cloning vector (Invitrogen). A unique internal enzyme site (Bst 1107) and a XhoI site within the TOPO vector were used to ligate the two PCR fragments to construct the full-length coding region for either the mutant or wild-type PHYA gene. Using the incorporated SalI site and a SacI site within the TOPO cloning vector, the full-length PHYA gene was ligated into the binary vector pPZPY122 , and the sequence was confirmed. The wild-type full-coding region of PHYB was generated using a similar RT-PCR method as for PHYA above except that it used one primer set. The forward primer (5'-GCC CCG GGA TGG TTT CCG GAG TCG GG-3') was designed over the translation start codon of PHYB with an incorporated SmaI cleavage site. The reverse primer (5'-CGT CTA GAA CTA ATA TGG CAT CAT CAG CAT C-3') was designed over the stop codon of PHYB with an incorporated XbaI site. The generated PCR product was cloned into the pCR 2.1 TOPO cloning vector (Invitrogen) and sequenced. This wild-type PHYB cDNA clone was used as the template for generating the mutant PHYB construct via PCR that specifically incorporates an A-to-G nucleotide on codon 664 to convert Val to Met. As a result, a full-length PHYB-V664M mutant coding region was produced. Using the incorporated SmaI site and a KpnI site within the TOPO cloning vector, the full-length phyB genes were ligated into the binary vector pPZPY122 (using the filled-in XbaI and KpnI sites within the vector; ). All PHYA and PHYB binary constructs were electroporated into the Agrobacterium tumefaciens strain GV3101 (MP90) and transformed into the proper Arabidopsis strains using the floral dip method . Backmatter: PMID- 12226525 TI - Characterization of an Acyltransferase Capable of Synthesizing Benzylbenzoate and Other Volatile Esters in Flowers and Damaged Leaves of Clarkia breweri AB - A cDNA encoding a protein with 456 amino acids whose sequence shows considerable similarity to plant acyltransferases was identified among 750 Clarkia breweri flower expressed sequence tags. The cDNA was expressed in Escherichia coli, and the protein produced was shown to encode the enzyme benzoyl-coenzyme A (CoA):benzyl alcohol benzoyl transferase (BEBT). BEBT catalyzes the formation of benzylbenzoate, a minor constituent of the C. breweri floral aroma, but it also has activity with a number of other alcohols and acyl CoAs. The BEBT gene is expressed in different parts of the flowers with maximal RNA transcript levels in the stigma, and no expression was observed in the leaves under normal conditions. However, BEBT expression was induced in damaged leaves, reaching a maximum 6 h after damage occurred. We also show here that a closely related tobacco (Nicotiana tabacum) gene previously shown to be induced in leaves after being challenged by phytopathogenic bacteria also has BEBT activity, whereas the most similar protein to BEBT in the Arabidopsis proteome does not use benzoyl CoA as a substrate and instead can use acetyl CoA to catalyze the formation of cis-3-hexen-1-yl acetate, a green-leaf volatile. Keywords: Introduction : Floral scents are often rich in volatile esters. In Clarkia breweri ([Gray] Greene; Onagraceae), an annual plant native to California, for example, benzylacetate constitutes 20% to 40% (w/w) of the total scent output (depending on the particular C. breweri line), whereas the two other esters present in its aroma, benzylbenzoate and methylsalicylate, each contribute about 5% to the total volatile output . Many plants also emit volatile esters from leaves damaged by herbivores. The most commonly reported volatile esters in this class are those derived from the octadecanoid fatty acids, such as cis-3-hexen-1-yl acetate , and methylsalicylate, mostly likely derived from the phenylpropanoid pathway or the shikimate pathway , is also common . Total volatiles (including, but not limited to, esters) emitted from injured leaves have been shown to function as orientation cues for predatory wasps and mites that feed upon herbivorous insects . In addition, some of these volatiles may serve as antimicrobial or antifungal agents to prevent further spread of disease . We have previously reported the characterization of the enzyme acetyl CoA:benzyl alcohol acetyl transferase (BEAT), which is responsible for the production of the floral volatile benzylacetate in C. breweri flowers . The concurrent characterization of BEAT and several other structurally similar enzymes has led to the recognition of a novel class of evolutionarily related acyltransferases (EC 2.3.1.x) commonly referred to as the BAHD family of acyltransferases . There are approximately 60 BAHD gene family members in the model organism Arabidopsis, although to date, the substrates and products of the enzymes encoded by these genes have not been determined. Members of the BAHD family have been identified in other plant species through expressed sequence tag (EST) database construction and analyses, and in a few cases, the biochemical function has been determined as well. For example, benzoyl-CoA:anthranilate N-benzo-yltransferase (HCBT) from carnation (Dianthus caryophyllus) is expressed during infection with Fusarium oxysporum or Phytophthora parasitica and produces several different benzoylated and coumaroylated anthranilide phytoalexin derivatives . Other BAHD proteins that are also known to be involved in plant defense are TAT and DBAT from Taxus cuspidata, which catalyze the acetylation and benzoylation, respectively, of taxol precursors , MAT and DAT from Catharanthus roseus, which catalyze the acetylation of precursors of the alkaloid vindoline , and SALAT, which is involved in morphine biosynthesis . BAHD enzymes involved in caffeoylation and malonylation of anthocyanin pigments have also been reported as well as those involved in volatile production in strawberry (Fragaria spp.) fruit . It should be noted that the enzymes responsible for the synthesis of methyl esters such as methylsalicylate and methyljasmonate belong to a different family of enzymes . Here, we report the characterization of the gene and enzyme for the biosynthesis of the floral volatile benzylbenzoate in C. breweri. The gene for benzoyl-CoA:benzyl alcohol benzoyl transferase (BEBT) is expressed in flowers, but it is also expressed in leaf tissue after damage. The structure of BEBT indicates that it belongs to the BAHD family of acyltransferases, and its wide substrate specificity may allow it to catalyze the formation of other esters as well. RESULTS : Benzylbenzoate-Forming Activity in Flowers | We have previously shown that flowers of C. breweri emit benzylbenzoate . The emission of this volatile is highest from the stigma, with petals contributing most of the rest . Total floral emission peaks 36 h postanthesis, with a second minor peak 72 h postanthesis. To look at the synthesis of benzylbenzoate, we developed an assay (see "Materials and Methods") to detect the enzymatic activity of BEBT, the hypothetical enzyme that would catalyze the formation of benzylbenzoate from benzyl alcohol and benzoyl CoA . Crude extracts of stigma and petal tissues from flowers of different ages were assayed. BEBT activity was detected in both petals and stigma. In the stigma, it was high already in unopened flowers, and it peaked on d 1 of anthesis . On d 2, when the stigma becomes receptive, there was a precipitous drop in BEBT activity of about 3-fold, and activity remained low on d 3 and 4. In petals, BEBT activity was much lower than in stigma but it gradually increased until d 4, at which time BEBT levels were similar to those in the stigma . Figure 1 | The reaction catalyzed by BEBT. The reaction catalyzed by BEBT. Figure 2 | BEBT enzymatic activity in different floral tissues during the lifespan of the flower. BEBT enzymatic activity in different floral tissues during the lifespan of the flower. Petal and stigma tissues were collected from C. breweri flowers daily starting 2 d before flower opening (d -2) and ending on d 4 postanthesis. For each data point, tissues from three different plants were combined for each assay, at least three independent assays were conducted, and the mean was obtained. circle, Petals; , stigma. pkat, Picomoles of product per second. Isolation of a cDNA Clone Encoding BEBT | We have recently undertaken the sequencing of approximately 750 EST cDNAs from a C. breweri flower tissue cDNA library. Because the acylation reaction performed by BEBT is similar in principle to the acylation reactions performed by BAHD-type enzymes, we searched the database with BAHD-type protein sequences, including C. breweri BEAT and putative BAHD proteins from Arabidopsis, using the BLAST2 program . The search identified one cDNA with homology to BAHD acyltransferase sequences. Because this cDNA was incomplete on both the 5' and 3' ends, we performed 5'- and 3'-RACE experiments to obtain a full-length clone. The complete cDNA, which we tentatively designated BEBT, has an open reading frame encoding a protein with 456 amino acid residues and a calculated molecular mass of 50.6 kD. Protein sequence comparisons indicate that BEBT is most similar (72% identity) to a protein encoded by the tobacco (Nicotiana tabacum) hypersensitive response cDNA HSR201 and to two BAHD-like Arabidopsis proteins (protein ID nos. CAC01898.1 and AAF01587.1) that are 57% and 54% identical, respectively . In addition, proteins encoded by a cDNA from cantaloupe (Cucumis melo) and a gene from rice (Oryza sativa) identified during the sequencing of the rice genome are also very similar to BEBT, being 53% and 63% identical, respectively, to BEBT. The proteins encoded by these genes all share the salient features of the BAHD protein family, most notably the HXXXD motif in the center of the protein believed to be involved in catalysis, and the DFGWG motif of unknown function near the C terminus . Figure 3 | Sequence comparisons of BEBT and BEAT from C. breweri Sequence comparisons of BEBT and BEAT from C. breweri, HSR201 from tobacco, and the two most similar proteins to BEBT from Arabidopsis. The BEBT gene accession number is , and the gene accession number of HSR201 is . The two Arabidopsis acyltransferase proteins in this figure are identified by their protein accession numbers and correspond to genes from two BAC clones (accession nos. and , respectively). A, Protein sequence alignments of all five proteins using the ClustalX program. Amino acids shaded in black represent identical matches; gray shaded boxes represent conservative changes. The HXXXD motif is indicated with arrowheads; the DFGWG motif is indicated with asterisks. B, Maximum parsimony tree based on protein sequence alignments from A using the program PAUP*. Enzymatic Characterization of C. breweri BEBT Expressed in Escherichia coli | To determine whether the isolated C. breweri cDNA indeed encoded an enzyme with BEBT activity, we subcloned the complete open reading frame of BEBT into the expression vector pET11a, transformed E. coli cells with the recombinant vector, and induced the expression of BEBT with IPTG. Because E. coli strain BL21(DE3) pLysS, the standard strain for expressing heterologous genes from pET vectors, has a low level of acyltransferase activity derived from the chloramphenicol O-acetyltransferase gene, we used E. coli strain B834 instead. Strain B834 cells have no activity with either acetyl CoA or benzoyl CoA regardless of whether they carry a pET expression vector and regardless of whether they have been treated with IPTG (data not shown). The spent media of BEBT-expressing E. coli cultures were extracted with pentane and analyzed by gas chromatography-mass spectroscopy (GC-MS), revealing a wide variety of acylated esters carrying acyl moieties ranging from C2-C6 . In contrast, spent media of induced cultures carrying a pET11a plasmid with no insert, used as a control, contained none of these esters. Because the ester composition in the spent media of BEAT-expressing E. coli cultures is >98% benzylacetate , it indicates that BEBT, unlike BEAT, may accept several acyl-CoA substrates in addition to acetyl-CoA. Although the products that formed in E. coli when BEBT was expressed clearly depended both on the affinity of the BEBT enzyme to potential substrates and the availability of these substrates in E. coli and although the substrate availability and concentration in the plant cell are likely to be very different from the situation in E. coli, nonetheless these preliminary results demonstrated that BEBT possessed ester-forming activity. Table I | Volatile esters extracted from the spent media of E. coli cells expressing BEBT To characterize the biochemical properties of BEBT in vitro, we purified the complete, non-fusion BEBT enzyme from the crude E. coli extract by anion-exchange chromatography on DEAE, followed by another anion-exchange column, Mono Q, and finally by size-exclusion chromatography on a Q-Superose column , as described in "Materials and Methods" (we had originally tried to introduce a His tag at either the N- or C-terminal ends of the protein, however both constructs yielded a protein with a dramatically lower activity). After the Q-Superose purification step, BEBT was essentially separated from all but one other protein with a molecular mass of 22.6 kD, which constituted about 40% of the total protein concentration. Although the latter protein was present in the Q-Superose fraction with the highest BEBT activity, its concentration in this and adjacent fractions did not correlate well with the levels of BEBT activities in these fractions, whereas BEBT protein concentrations and activity levels did (data not shown). It is noteworthy that BEBT migrated on SDS-PAGE as a protein with a molecular mass of 55 kD, whereas its calculated molecular mass is 50.6 kD. BEAT protein similarly has a calculated molecular mass of 48.2 kD, but migrates on a gel as a protein with a molecular mass of 58 kD . Figure 4 | Purification of C. breweri Purification of C. breweri BEBT produced in E. coli. Lane 1, Mr markers. Lane 2, Q-Superose fraction with the highest levels of BEBT-specific activity. In initial assays of BEBT activity, we used a variety of alcohols with benzoyl CoA and acetyl CoA. With acetyl CoA, benzyl alcohol was the preferred substrate, but other alcohols could also be used (data not shown). However, the Km value for acetyl CoA, 818 mum , strongly suggests that acetyl CoA is not commonly used by BEBT as the acyl donor in vivo. With benzoyl CoA, for which a Km value of 20.5 mum was determined, BEBT could use benzyl alcohol (with a Km value of 46.8) and a variety of other alcohols, some of them with higher rates (at a substrate concentration of 1 mm), particularly octanol and cinnamyl alcohol . BEBT had a Km value for cinnamyl alcohol of 97.8 mum . We also examined the activity of BEBT with other acyl CoA substrates by performing competition experiments of these nonradioactive CoA esters with radioactive acetyl CoA. Benzoyl CoA and cinnamoyl CoA were better competitors than either butanoyl CoA or hexanoyl CoA. The Km value for cinnamoyl CoA was determined to be 464 mum . Coumaroyl CoA, on the other hand, did not act as a competitor and could not serve as a substrate. On the basis of the these results and the measured Km values, we concluded that benzoyl CoA and benzyl alcohol are likely to be in vivo substrates of BEBT. It is possible that other alcohols similar to benzyl alcohol (such as octanol) and other acyl CoA esters similar to benzoyl CoA (such as cinnamoyl CoA) may also be acted upon by BEBT, but at present, we are not aware of other possible ester products in C. breweri besides benzylbenzoate that could result from catalysis by BEBT. Table II | Kinetic parameters of BEBT Table III | Relative activity of BEBT with a variety of substrates The pH optima of BEBT was determined to be pH 7.7, whereas only 39% and 69% of maximal activity were observed at pH of 6.5 and of 9.0, respectively. The enzyme was 100% stable for 30 min at 30C and 80% stable for 30 min at 37C. After incubation at 50C for 30 min, the enzyme was completely inactivated. The effects of monovalent and divalent cations were also tested. None were found to be stimulatory, however Mg2+, Ca2+, Co2+, Zn2+, and Cu2+ had a strong inhibitory effect (50% --100% inhibition at a final concentration of 5 mm). The apparent molecular mass of the protein as determined by size-exclusion chromatography was approximately 47 to 49.5 kD. Because the predicted molecular mass of the BEBT protein is 50.6 kD, we concluded that BEBT is a monomeric enzyme. Organ Specificity and Developmental Changes of BEBT Expression in C. breweri | RNA gel-blot analysis with d 1 postanthesis flowers and leaf tissue was carried out to examine the tissue(s) in which the BEBT gene is expressed . The highest steady-state levels of BEBT transcripts were observed in the stigma. Stamens, sepals, and petals had approximately the same amount of transcripts, although it was approximately 30% of that found in the stigma. Transcript levels in style tissue were 10% of those found in the stigma, whereas healthy leaves had little discernible expression of BEBT . Figure 5 | RNA gel-blot analysis of the relative abundance of BEBT mRNA transcripts in the stigma, stamen, style, sepals, petals, and leaves of C. breweri RNA gel-blot analysis of the relative abundance of BEBT mRNA transcripts in the stigma, stamen, style, sepals, petals, and leaves of C. breweri plants. Tissues were harvested from mature plants and floral tissue samples were taken from flowers on d 1 of anthesis. Lanes were loaded with 4 mug of total RNA. After probing with the BEBT probe and quantitation of the results, each blot was rehybridized with an 18S rDNA probe to normalize samples. A sample blot is shown above. The graphical representation below represents an average of three independent experiments. The levels of BEBT transcripts were examined over the life span of the flower by comparing the stigma, the organ with the highest levels of BEBT transcripts, to the petals, the largest organ (by mass) of the flower. The stigmata of flower buds 2 d before anthesis contained the highest levels of BEBT transcripts, and these levels subsequently dropped but remained within 40% to 60% of their initial value by d 3 to 4 postanthesis (Fig. A). In contrast, BEBT mRNA transcript levels from petal tissue steadily increased throughout the lifespan of the flower, reaching a 5-fold increase on d 4 postanthesis as compared with d -2 (Fig. B). However, BEBT mRNA levels in petals at any given day were always severalfold lower than those in the stigma. Figure 6 | Expression of BEBT in petal and stigma tissues of C. breweri Expression of BEBT in petal and stigma tissues of C. breweri flowers during floral development. A, RNA gel-blot analysis of the relative abundance of BEBT mRNA in stigma tissue. Stigma tissue was collected daily from flowers starting 2 d before flower opening and continuing until d 4 postanthesis. B, RNA gel-blot analysis of the relative abundance of BEBT mRNA in petal tissue. Petal tissue was collected at the same times as the tissues in A. For all experiments, each lane was loaded with 4 mug of total RNA. After hybridization with the BEBT probe and quantitation of the results, blots were stripped and reprobed with an 18S rDNA probe to normalize samples. In each panel, a sample blot is shown above, and the graphical representation below represents an average of two independent experiments. Quantitation of BEBT Protein Levels in the Stigma by Immunoblotting | BEBT protein levels were quantified in the stigma by SDS-PAGE followed by immunoblots of stigma samples together with samples of known amounts of BEBT purified from E. coli (Fig. , top). In stigma extracts from all stages of development, with the exception of d -2 stigma, the antibodies made against purified C. breweri BEBT produced in E. coli (see "Materials and Methods") recognized a single protein that migrated on the gel to the same position as did BEBT produced in E. coli. In d -2 stigma, the anti-BEBT antibodies recognized an unfocused protein band migrating slightly more slowly than E. coli-produced BEBT (Fig. , top). The reasons for the aberrant behavior of BEBT in d -2 stigma in SDS-PAGE could not be determined. The quantitative results of the immunoblots (Fig. , bottom) indicated that BEBT protein levels followed the measured levels of BEBT enzymatic activity , dropping equally rapidly on d 2 postanthesis, the day when the stigma lobes open up and the stigma becomes receptive, compared with d 1 postanthesis. Figure 7 | Variation of the levels of BEBT protein over the lifespan of the stigma. Variation of the levels of BEBT protein over the lifespan of the stigma. Stigma tissue was collected daily from flowers starting 2 d before flower opening, and continuing until d 4 postanthesis. Samples were run on SDS-PAGE, and the gels were blotted onto filters. The filters were first probed with anti-BEBT antibodies, followed by incubation with secondary antibodies conjugated to alkaline phosphatase. Bands were visualized by chemiluminescence. A sample blot is shown above, and the graphical representation below represents an average of two independent experiments. Characterizaton of the Activities of the Proteins Encoded by Tobacco HSR201 cDNA and by Another BEBT-Related cDNA from Arabidopsis | The sequence of the protein encoded by the HSR201 cDNA from tobacco is the most similar to BEBT of the protein sequences currently found in the databases . It has previously been shown that HSR201 is expressed in damaged leaves of tobacco after infection with the microorganism Pseudomonas solanacearum . To analyze the function of the protein encoded by HSR201, we isolated total RNA from tobacco leaves 6 h after damaging them with a razor blade, and cloned the open reading frame of HSR201 into the pCR T7 TOPO-CT expression vector, using oligonucleotides based on the published sequence. The HSR201 protein obtained in the E. coli expression system had BEBT activity with Km values for benzoyl CoA and benzyl alcohol of 35 and 19 mum, respectively, similar to the corresponding Km values of BEBT. The HSR201 protein showed no activity with acetyl CoA at concentrations <1 mm. We also examined the activities of the proteins encoded by the two Arabidopsis genes whose proteins show the highest similarity to BEBT. cDNAs encoding CAC01898.1 and AAF01587.1 were expressed in E. coli. No activity could be detected for protein CAC01898.1 with a variety of alcohols and either acetyl CoA or benzoyl CoA. Protein AAF01587.1 had no detectable activity with benzoyl CoA at concentrations <1 mm but exhibited strong activity with acetyl CoA and cis-3-hexen-1-ol and related alcohols (e.g. hexanol and octanol). AAF01587.1, which we have designated CHAT (for acetyl CoA:cis-3-hexen-1-ol acetyl transferase) had a Km value of 10.5 mum for acetyl CoA and 165 mum for cis-3-hexen-1-ol. BEBT Expression in Wounded Leaves | Because HSR201 proved to encode BEBT and its expression was known to be induced in leaves upon wounding, we examined the expression of C. breweri BEBT in wounded leaves. Total RNA was extracted from damaged leaves at intervals of 0, 1, 2, 4, 6, 8, 12, and 24 h after wounding, and the RNA samples were analyzed by gel blotting and probing with a BEBT probe . BEBT mRNA levels rose dramatically from 2 to 6 h after wounding, with peak expression evident after 6 h, decreasing afterward. It should be noted that BEBT transcript levels in wounded leaves were still roughly one-half of that found in the stigma of d 1 flowers. Figure 8 | RNA gel-blot analysis of the relative expression of BEBT in wounded and non-wounded leaf tissues of C. breweri RNA gel-blot analysis of the relative expression of BEBT in wounded and non-wounded leaf tissues of C. breweri. Mature leaves were harvested either before (untreated) or after mechanical wounding. Leaves were wounded by making two parallel incisions approximately 8 mm long to each side of the midvein with a sterile razor blade, and tissue was collected at 0, 1, 2, 4, 6, 8, 12, and 24 h after wounding. RNA from stigma tissue (included for comparison) was collected from d 1 flowers. Total RNA (4 mug) was loaded onto each lane of the gel. After probing with the BEBT probe and quantitation of the results, the blots were then stripped and reprobed with an 18S rDNA probe to normalize samples. A sample blot is shown above, and the graphical representation below represents an average of two independent experiments. DISCUSSION : BEBT, a BAHD-Type Acyltransferase, Is Responsible for Benzylbenzoate Synthesis in C. breweri Flowers | The BAHD family of acyl transferases is a recently discovered group of enzymes . The function and substrate specificity of only a few representatives, which include enzymes that catalyze the formation of both volatile and nonvolatile esters, have so far been determined. Even so, it is clear that a hallmark of these enzymes is the relatively wide range of alcohols that they can react with. On the other hand, the range of acyl-CoA substrates has not yet been extensively tested, due mostly to the lack of availability of these compounds. However, at least in one case, that of HCBT from carnation, it was shown that the enzyme has a somewhat lower Km value for cinammoyl CoA than for benzoyl CoA, even though benzoyl CoA is the substrate used in vivo . The products detected in the spent media of E. coli cells expressing BEBT show that BEBT can use several medium-chain acyl CoA substrates and a variety of mostly primary alcohols in addition to benzoyl CoA and benzyl alcohol. However, what products are formed in E. coli depend on the presence and concentrations of the respective substrates vis-a-vis the affinity of BEBT for those substrates. The possibility that the in planta substrates of BEBT are not found in E. coli is highly likely. The Km values of BEBT for the substrates benzyl alcohol and benzoyl CoA ---46.8 and 20.5 mum, respectively ---are in a range that is similar to those of BEAT and several other BAHD acyltransferases for their substrates, and they strongly argue that BEBT is responsible for the synthesis of benzylbenzoate, at least in flowers. For comparison, the enzyme benzoyl-CoA:taxane 2alpha-O-benzoyltransferase from Taxus spp. was reported to have a Km value of 300 mum for benzoyl CoA . The enzyme HCBT, another BAHD benzoyltransferase, has a Km value of 50 mum for benzoyl CoA, similar to that of BEBT . There is currently no data on the concentrations of benzoyl CoA in C. breweri or any other species known to make benzoate esters, although benzoic acid:CoA ligase has recently been identified and partially purified from C. breweri . On the other hand, the Km values of BEBT for acetyl-CoA and cinnamoyl CoA ---818 mum and 464 mum, respectively ---suggest that these acyl CoAs are not commonly used by BEBT. In addition, the ratio of Kcat/Km (benzoyl CoA) to Kcat/Km (acetyl CoA) shows that BEBT preferentially uses benzoyl CoA to acetyl CoA by a factor of 5.8. Additional support for the role of the protein encoded by the BEBT gene in benzylbenzoate production in flowers comes from the general correlation between BEBT mRNA levels and BEBT enzyme activity in the different floral parts (Figs. and ) and from the correlation between BEBT enzymatic activity in the stigma and the amount of BEBT protein as determined by immunoblotting with antibodies made against the protein encoded by the BEBT gene (Figs. and ). Although BEBT transcript levels and BEBT enzymatic activity levels in the stigma are always higher than those in the petals, nonetheless the levels of BEBT enzymatic activity in the stigma show an interesting pattern of change relative to the BEBT mRNA levels. BEBT activity levels show a moderate buildup until d 1 of anthesis followed by quick decline, even though BEBT mRNA levels do not show such a corresponding decline. The buildup in BEBT activity levels until d 1 of anthesis suggests that the turn-over of the BEBT protein is slow (as is the case for other scent enzymes , and therefore BEBT protein continues to accumulate even though BEBT steady-state transcript levels are relatively stable. The precipitous decline in BEBT enzymatic levels (per gram fresh weight) on d 2 of anthesis is clearly attributable to a decrease in total BEBT protein , and this decline is not caused by a similar decline in the steady-state levels of BEBT mRNA (Fig. A). Day 2 of anthesis is the day in which the stigma opens up and becomes receptive, a process that involves rapid cell expansion and consequently a reduction in protein content per cell . However, even when BEBT activity levels and BEBT protein levels are plotted after normalization to protein content (data not shown), a substantial drop of 50% in both is still observed from d 1 to 2. This reduction in BEBT activity, therefore, must involve a mechanism of specific protein degradation and may be related to the yet-undetermined function of benzylbenzoate in the stigma. Acyltransferase Activities in Wounded Leaves | It is well established that fatty acid-derived aldehyde, alcohol, and ester "green-leaf" volatiles are emitted upon injury . Our results showing the induction of BEBT in wounded leaves and the induction of Arabidopsis CHAT by wounding (J.C. D'Auria, unpublished data) point to the possible role of such acetyltransferases in the defense response of leaves to injury. Although CHAT may be involved in the biosynthesis of cis-3-hexenyl acetate, it is likely that BEBT activity in the leaf is responsible mostly for the production of benzylbenzoate. Other species are also known to produce this compound in leaves. For example, leaves of rice produce benzylbenzoate in response to oviposition by the insect whitebacked planthopper (Sogatella furcifera), and it was shown that benzylbenzoate has ovicidal properties . Our results showing that the tobacco hypersensitive-response cDNA HSR201 encodes BEBT suggest that tobacco leaves may also synthesize benzylbenzoate or a related compound under stress. Other esters that are known to be synthesized in damaged leaves include rho-coumaroyl- and feruloyltyramine in tomato (Lycopersicon esculentum; ) and 5-caffeoylquinic (chlorogenic) acid and its derivatives in lettuce (Lactuca sativa; ). It is worth noting that RNA-blot analysis indicated that the BEBT-related gene BEAT, which encodes an enzyme that catalyzes the formation of benzyl acetate, is not generally expressed in C. breweri leaves nor is it induced upon damage (N. Dudareva and E. Pichersky, unpublished data). Evolution of BEBT and CHAT | Although BEBT and HSR201 have strong affinity to benzoyl CoA and low affinity to acetyl CoA, they are closely related to CHAT, which has a much stronger affinity to acetyl CoA than to benzoyl CoA. A similar situation was reported for two closely related acyltransferases from Taxus spp., one that uses acetyl CoA and the other benzoyl CoA . This suggests that the specificity for the acyl CoA moiety can evolve relatively easily in the BAHD family of enzymes. MATERIALS AND METHODS : Plant Material and Growth Conditions | Details of the construction of true-breeding Clarkia breweri stocks and growth conditions are as described . Arabidopsis ecotype Columbia plants were used of for cloning CHAT. Tobacco (Nicotiana tabacum) seeds were germinated on Murashige and Skoog medium plates and grown for 2 weeks. Seedlings were then transplanted to soil (Sunshine mix no. 2, Sun Gro Horticulture, Bellevue, WA) and grown under 12-h light/12-h dark light conditions in a growth chamber. Reagents | All solvents and reagents were molecular biology grade or reagent grade and were obtained from Sigma (St. Louis), Aldrich (Milwaukee), or Fisher Scientific (Loughborough, Leicestershire, UK). Radiolabeled [acety-1-14C]CoA was purchased from ICN (Irvine, CA). [7-14C]Benzoyl-CoA was enzymatically synthesized from [7-14C]benzoic acid (16.6 mCi mmol-1; Sigma) as previously described . Enzyme Extraction and Assay | Crude protein extracts were prepared by macerating tissue parts in a microcentrifuge tube with a Teflon-coated homogenizer in the presence of ice-cold buffer containing 50 mm Bis-Tris-HCl, pH 7.0, 5 mm dithioerythritol (DTE), 1% (w/v) polyvinylpyrrolidone-40, and 10% (v/v) glycerol. A ratio of extraction buffer:tissue fresh weight of 10:1 (v/w) was used. The slurry was then centrifuged for 10 min at 4C, and the supernatant was transferred to a new tube. For each time point, flower tissue and leaves from three different plants were combined, and assays were repeated at least twice. Assay samples were prepared by adding the following to a 0.6-mL microcentrifuge tube: 5.0 muL of crude extract, 10.0 muL of 5x assay buffer (250 mm Bis-Tris Propane, pH 7.7), 1.0 muL of 50 mm alcohol substrate dissolved in dimethyl sulfoxide, 1.0 muL of [7-14C]benzoyl-CoA, and 33 muL of water to bring the assay volume to 50 muL. Acetyltransferase assays for both BEBT and CHAT were performed as above with the substitution of 0.25 muL of [acetyl-1-14C]CoA (50 mCi mmol-1) for the radiolabeled benzoyl-CoA. The assays were carried out at room temperature for 30 min. The radioactively labeled acylated product was extracted twice with 180 muL of hexane, and the total organic phase was transferred to a scintillation vial with 2 mL of nonaqueous scintillation fluid (Econo-Safe, Research Products International, Mount Prospect, IL) and counted in a liquid-scintillation counter (model LS6500, Beckman Coulter, Fullerton, CA). The raw data (counts per minute) were converted to picokatals based on the specific activity of the radiolabeled substrate and the known efficiency of the scintillation counter used. Assays in which no alcohol substrate was added were performed to test background thioesterase activity. In addition, boiled enzyme extracts were substituted for intact enzyme to test nonenzymatic breakdown of the CoA thioester via reaction with reducing agents in the enzyme extract. The identity of the products were verified by GC-MS as described previously using authentic standards . BEBT cDNA Identification | A total of 735 cDNAs were randomly and automatically isolated and sequenced similar to those previously reported for a basil (Ocimum basilicum) EST database . The C. breweri flower EST database can be found at: (login, clarkia; password, breweri). The BEBT cDNA was identified by BLAST searches with other BAHD sequences. Cloning of BEBT and Its Homologs from Arabidopsis and Tobacco | The BEBT cDNA clone identified in the C. breweri EST database was incomplete. RACE was used to obtain the missing sequence from either end . Two specific primers were designed (forward primer, 5'-CCCATTCGACGACCTGGCTCACCGG-3'; backward primer, 5'-CCCAGCTCGTGTCACGTCTGAAACCAAG-3') and used in 5'- and 3'-RACE, respectively. First-strand cDNAs for the RACE experiments were made from total RNA of 1-d-old C. breweri flowers. On the basis of the sequences obtained, the full-length cDNA was amplified from first-strand cDNA using the forward primer 5'-AATGGCGCATGATCAATCTCTCT-3' and the backward primer 5'-TCTCATCAAAGGGAAGACTGTGT-3'. The resulting 1,375-nucleotide (nt) DNA fragment was subcloned into a TA cloning vector (pCRT7/CT TOPO-TA, Invitrogen, Carlsbad, CA). The sequences of independent clones were determined in their entirety on both strands. For further expression analysis, the open reading frame of BEBT was subcloned into the pET-T7(11a) expression vector (Novagen, Madison, WI). For the gene residing on BAC clone (CHAT), a full-length cDNA was obtained by RT-PCR using total RNA extracted from Arabidopsis flower tissue using the forward primer 5'-AATGGACCATCAAGTGTC-TCTGC-3' and the backward primer 5'-TCATCCTTTAGACACATT-TAGCACTCC-3'. The resulting 1,366-nt DNA fragment was cloned as described above for BEBT. All attempts to clone the gene residing on the BAC clone via RT-PCR using total RNA from various plant tissues were unsuccessful. Because this gene has only one intron, the putative open reading frame was amplified from genomic DNA, and the introns were spliced out as described by by using the following two pairs of primers (front pair, forward primer 5'-AATGTCCGGGTCACTCACGTT and backward primer 5'-GCATTTCAAGCGCGTGACCTGCATGAGCAT-3'; back pair, forward primer 5'-CCTTTGATGCTCATGCAGGTCACGCGCTTGAAATGC-3' and backward primer 5'-TTACAGAGAAGACATGATCAA-3'). The resulting 1,387-nt DNA fragment was again cloned as described above. The full-length cDNA clone HSR201 from tobacco was cloned by RT-PCR using first-strand cDNA obtained from total RNA extracted from mechanically wounded tobacco leaves 6 h after wounding. Wounds were made by making two lateral incisions to each side of the midvein with a sterile razor blade. The primers used in the RT-PCR were the forward primer 5'-AATGGATTCAAAGCAATCATCAGA-3' and the backward primer 5'-TCAAAGGGCAGGTCTGATAATG-3' and resulted in a DNA fragment that was 1,384 nt in length. This fragment was also cloned into a TA cloning vector (pCRT7/CT TOPO-TA, Invitrogen) for expression analysis. Sequence Alignment and Phylogenetic Analysis | Amino acid sequence alignments were generated using the ClustalX computer program . Aligned sequences were analyzed using the PAUP* program (Phylogenetic Analysis Using Parsimony, Sinaur Associates, Sunderland, MA), with amino acids treated as unordered characters, and resulted in a single shortest tree shown in a PAUP* phylogram format. A neighbor-joining tree was also generated. It showed the same branches as that in the maximum-parsimony tree. Protein Production in Escherichia coli and BEBT Purification | BEBT expression in E. coli was induced as previously described with the following minor modifications: All expression constructs in this study were transformed into the E. coli cell line B834. E. coli cells were grown to an OD600 of 0.4 and then induced with 1 mm IPTG. After harvesting, the cells were resuspended in lysis buffer containing 50 mm Tris-HCl (pH 8.0), 10 mm NaCl, 1 mm EDTA, 5 mm DTE, 1 mm phenylmethylsulfonyl fluoride, and 10% (v/v) glycerol, lysed by sonification (3x for 30 s), and centrifuged at 10,000g for 15 min. Soluble BEBT protein was first loaded onto DEAE anion-exchange column (10 mL of DE53, Whatman, Clifton, NJ) pre-equilibrated with buffer A (50 mm Bis-Tris, pH 7.5, 10% [v/v] glycerol, and 5 mm DTE) at a flow rate of 1 mL min-1 on an FPLC apparatus (Amersham Pharmacia Biotech AB, Uppsala). After washing off unbound materials from the column with buffer A, BEBT was eluted with a linear gradient of 0 to 0.5 mm KCl in buffer A. BEBT activity eluted in the 0.25 to 0.3 mm range. Fractions with the highest BEBT activity (10 mL) were pooled and diluted with buffer A to a final volume of 40 mL and loaded onto a Mono-Q FPLC column (0.5-cm diameter x 6.0 cm; Amersham Pharmacia Biotech AB). BEBT was eluted with a 100 to 500 mm KCl linear gradient with 3-mL fractions being collected. BEBT activity eluted with 130 to 140 mm KCl. Peak fractions from Mono-Q were examined by SDS-PAGE gel electrophoresis followed by Coomassie Brilliant Blue or silver staining of the gel. BEBT protein concentration was determined as previously described . Preparation of BEBT Antibodies and Protein Gel Blots | C. breweri BEBT protein produced in E. coli was purified from inclusion bodies as previously described . The final step in the purification consisted of electrophoresis on SDS-PAGE gel, staining with Coomassie Blue, and excision of the gel segment where the separated BEBT was located. Antibodies were prepared by Cocalico Biologicals (Reamstown, PA) by injecting macerated gel fragments containing the purified BEBT and following the company's protocol. Protein gel blots and quantitation of results were performed as described in . Extraction of Volatile Esters from the Spent Media of E. coli and GC-MS Analysis | Extraction of volatile esters from the spent media of E. coli expressing BEBT or containing pET11a vector without any insert were performed and analyzed by GC-MS as reported by with the substitution of pentane for hexane as the solvent. Molecular Mass Estimation | Peak BEBT activity from the Mono-Q fractions was pooled, concentrated down to 200 muL, and run on a gel filtration column (Q-Superose, Amersham Pharmacia Biotech AB) precalibrated with molecular mass markers as previously described . Determination of Kinetic Parameters for BEBT and CHAT | In all kinetics studies, appropriate enzyme concentration and incubation times were chosen so that the reaction velocity was linear during the incubation time period. To determine the Km value for each substrate, one substrate concentration was fixed at a saturated level and the concentration of the other substrate to be measured was varied. Lineweaver-Burk, Eadie-Hofstee, and Hanes plots were constructed and an average of their Km values were used to report the apparent Km value. Temperature Effect on BEBT Activity | BEBT protein from the Mono-Q fraction was incubated at temperatures ranging from 4C to 65C for 30 min and then chilled on ice. Samples incubated at each temperature were then used for the enzyme assays. At least four independent assays were performed for each point, and an average was taken. pH Optimum of BEBT Stability | The optimum pH for BEBT activity was determined using Bis-Tris propane as the buffering system ranging from 6.5 to 9.0. Results presented are an average of at least four independent assays. Effectors | Enzyme assays were performed with one of the following cations present in the assay buffer at a final concentration of 5 mm: K+, Na+, Ca2+, Cu2+, Fe2+, Mg2+, Mn2+, Zn2+, and Co2+. Results presented are an average of at least four independent experiments. RNA Extraction and Northern-Blot Analysis | Total RNA was isolated from both C. breweri and Arabidopsis tissues using the RNeasy plant mini kit (Qiagen), using buffer RLC for C. breweri and RLT for Arabidopsis. For the leaf-wounding treatments of C. breweri leaves, two parallel incisions of 8 to 10 mm were made on either side of the midvein using a sterile razor blade. Leaf tissue was collected at 0, 1, 2, 4, 6, 8, 12, and 24 h after the initial incision. In all cases, 100 mg of tissue was used, and the protocol described in the kit was followed. To reduce the amounts of polysaccharides extracted from C. breweri leaves using the RNeasy kit, 20 muL of 50% high Mr PEG (15,000 --20,000) was added to the microcentrifuge tube after the RLC buffer was added. This mixture of tissue, buffer, and PEG was incubated for 10 min at room temperature. After this, all steps were performed as directed in the kit manual. Total RNA (4 mug) from C. breweri tissues was resolved on 1% (w/v) agarose-formaldehyde gels, blotted to Hybond-N+ nylon membranes, hybridized, and stripped as described by and as per the instructions in the Hybond-N+ manual (Amersham Pharmacia Biotech AB). Probes were synthesized using the rediprime II kit (Amersham Pharmacia Biotech AB) from a PCR-amplified fragment of BEBT using the full-length primers used previously for cloning. Hybridization signals were counted in a phosphor imager (Bio-Rad), stripped, and reprobed with a 18S rRNA probe to normalize mRNA levels and control for discrepancies in concentration readings given by spectrophotometer. Backmatter: PMID- 12226526 TI - Cloning and Functional Characterization of a beta-Pinene Synthase from Artemisia annua That Shows a Circadian Pattern of Expression AB - Artemisia annua plants produce a broad range of volatile compounds, including monoterpenes, which contribute to the characteristic fragrance of this medicinal species. A cDNA clone, QH6, contained an open reading frame encoding a 582-amino acid protein that showed high sequence identity to plant monoterpene synthases. The prokaryotically expressed QH6 fusion protein converted geranyl diphosphate to (-)-beta-pinene and (-)-alpha-pinene in a 94:6 ratio. QH6 was predominantly expressed in juvenile leaves 2 weeks postsprouting. QH6 transcript levels were transiently reduced following mechanical wounding or fungal elicitor treatment, suggesting that this gene is not directly involved in defense reaction induced by either of these treatments. Under a photoperiod of 12 h/12 h (light/dark), the abundance of QH6 transcripts fluctuated in a diurnal pattern that ebbed around 3 h before daybreak (9th h in the dark phase) and peaked after 9 h in light (9th h in the light phase). The contents of (-)-beta-pinene in juvenile leaves and in emitted volatiles also varied in a diurnal rhythm, correlating strongly with mRNA accumulation. When A. annua was entrained by constant light or constant dark conditions, QH6 transcript accumulation continued to fluctuate with circadian rhythms. Under constant light, advanced cycles of fluctuation of QH6 transcript levels were observed, and under constant dark, the cycle was delayed. However, the original diurnal pattern could be regained when the plants were returned to the normal light/dark (12 h/12 h) photoperiod. This is the first report that monoterpene biosynthesis is transcriptionally regulated in a circadian pattern. Keywords: Introduction : In plants, many aspects of circadian and photoperiodic behaviors are regulated by an endogenous clock that helps plants to anticipate and adapt to the daily and seasonal fluctuations in light and temperature . Functional genomic studies of Arabidopsis show many genes to be controlled by the clock, and their functions oscillate rhythmically . It is well documented that genes involved in primary metabolic processes such as photosynthesis are circadianly regulated . For example, in crassulacean acid metabolism plants, the gene for phosphoenolpyruvate carboxylase kinase is regulated by a circadian oscillator to enhance enzyme activity in the dark . In other branches of metabolism, attention has been paid to the circadian pattern of plant hormones such as abscisic acid , or photoprotective pigments, including phenylpropanoids . However, reports on the biosynthetic rhythm of terpenes, the biggest group of plant secondary metabolites, has been limited to in planta chemical analysis and emission profiles of flowers, grasses, shrubs, and trees . In nature, plants are the main source of volatile organic compounds (VOCs) emitted to the atmosphere, among which some terpenoids and benzenoic compounds are released in diurnal or nocturnal patterns . beta-Pinene is one of the most widely detected VOCs , and in some plants, alpha-pinene, beta-pinene, and sabinene may account for more than 80% of the monoterpenes emitted . In Quercus ilex grown under natural conditions, terpene emission was found to peak at noon, with the quantity shifting with light intensity and temperature . The emission of alpha- and beta-pinene in laboratory tests peaked in as short as 30 min after the light was switched on . Because this plant has no special storage tissue (such as the oil glands of peppermint, Mentha piperita) for terpenoids produced, the emission pattern implies fluctuations in the activity of specific enzymes, which could be controlled at any step from transcription to posttranslational modification, or in the availability of substrate. The first committed reactions in the formation of the various classes of terpenes are catalyzed by terpene synthases, which convert acyclic diphosphate substrates (geranyl diphosphate [GPP], farnesyl diphosphate [FPP], geranylgeranyl diphosphate [GGPP], etc.) into acyclic or cyclic terpenes (compare with ). These terpene products may be further modified by other enzymes, such as P450 monooxygenases, leading to various terpenoid derivatives . Monoterpenes, which are formed from GPP by monoterpene synthases, are widely present in plants , and some of them account for the flower scent of plants (e.g. ). In different developmental stages of peppermint, coincident temporal change in enzyme activities, enzyme protein levels, and steady-state transcript abundances of monoterpene synthases was demonstrated, indicating that most of the monoterpene synthases in this plant are regulated at the level of gene expression . Artemisia annua is an annual herb widely distributed in Asia, Europe, and North America. Different from most of the other plants in the Asteraceae family, the pollination of A. annua is mediated by wind or insects . The plant is rich in terpenoids . Among these are the sesquiterpene lactone artemisinin, which is one of the most efficient drugs against Plasmodium species involved in malaria , and monoterpenes camphor, 1, 8-cineole, alpha-pinene, and beta-pinene, etc. , which give the plant a sweet scent. A key enzyme for artemisinin synthesis, amorpha-4, 11-diene synthase, has recently been purified, and its cDNA was cloned . At least two monoterpene synthases (QH1 and QH5) were proved to be inducible at the transcriptional level by mechanical wounding. In vitro assay showed that both of them converted GPP into (3R)-linalool, which, however, could not be detected from the plant extract . Here, we report a new A. annua monoterpene synthase cDNA, QH6, which encodes a (-)-beta-pinene synthase. This is the first reported terpene synthase of which transcripts and enzymatic products were shown to fluctuate in a circadian pattern. RESULTS : Cloning and Functional Analysis of QH6 | The isolated clone, QH6, contained a cDNA insert of 1,863 bp with an open reading frame encoding a 582-amino acid protein, of which the calculated Mr was 67 kD and the calculated pI was 5.7. R286W287W288 and D337D338X339X340D341 motifs, which are conserved in terpene synthases , were also present in QH6 . As in other monoterpene synthases, a transit peptide sequence upstream to the absolutely conserved R48R49 was found that may function in targeting the protein into plastids . Sequence analysis showed that the deduced protein was closest to monoterpene synthases from other angiosperms (Tpsb subfamily; ; Fig. ). QH6 showed 57.1% and 57.2% sequence identities, respectively, with another two monoterpene synthases, QH1 and QH5, from A. annua , but had only 32.5% with a beta-pinene synthase , and 31.6% with a (-)-limonene/(-)-alpha-pinene synthase from the gymnosperm grand fir (Abies grandis). Figure 1 | Nucleotide and deduced amino acid sequences of QH6. Nucleotide and deduced amino acid sequences of QH6. The conserved motifs are in boldface. Positions and directions of primers used are underlined and labeled. The HindIII site, before which the fragment is used as template for probe labeling in northern blotting, is in italic (AAGCTT, position 566 --571). Figure 2 | Phylogenetic analysis of monoterpene synthases of which the enzymatic functions reported. Phylogenetic analysis of monoterpene synthases of which the enzymatic functions reported. The neighbor-joining method was used for the protein sequences. Bootstrap values were labeled besides the branches, and the scale bar indicates 20% sequence divergence. The recombinant plasmid pET32b/QH6, in which the transit peptide sequence had been truncated, resulted in successful expression of QH6 fusion protein in Escherichia coli after isopropyl beta-d-thiogalactoside induction. With GPP as substrate, the fusion protein catalyzed the formation of two products in the ratio of 94:6, as revealed by two peaks in gas chromatography (GC). Mass spectra (MS) showed these to be beta-pinene (the major product) and alpha-pinene (the minor product; Fig. ). By coinjection with authentic standards for GC, both products were found to be (-)-enantiomers . These results were supported by 1H nuclear magnetic resonance (NMR) spectra of the products. The signals from the major product were 4.625 (br s, 1H), 4.556 (br s, 1H), 2.539 (m, 1H), 2.455 (t, 5.4 Hz, 1H), 2.316 (dtd, 9.9 Hz, 5.8 Hz, 1.2 Hz, 1H), 2.248 (m, 1H), 1.975 (m, 1H), 1.84 (m, 1H), 1.82 (m, 1H), 1.420 (d, 9.8 Hz, 1H), 1.236 (s, 3H), and 0.720 (s, 3H), and those from the minor product were 5.185 (t of sextet, 3 Hz, 1.5 Hz, 1H), 2.162 (d of sextet, 17.5 Hz, 2.5 Hz, 1H), 1.931 (td, 5.6 Hz, 1.5 Hz, 1H), 1.658 (br q, approximately 2 Hz, 3H), 1.151 (d, 8.5 Hz, 1H), and 0.835 (s, 3H). These signals are within experimental error of literature values . These in vitro results indicate that QH6 encodes a beta-pinene synthase, which produces alpha-pinene as a minor product. The 94:6 ratio of beta-:alpha-pinene in in vitro assay was similar to that measured from A. annua leaf extract (92:8), but higher than that from stems (89:11) or inflorescences (79:21). Figure 3 | Identification of the major enzymatic product as beta-pinene, and the minor as alpha-pinene. Identification of the major enzymatic product as beta-pinene, and the minor as alpha-pinene. MS are of the major product (A), authentic beta-pinene (B), minor product (C), and authentic alpha-pinene (D). Figure 4 | Enantiomer-selective GC identification of alpha- and beta-pinene produced by the recombinant enzyme are (-)-enantiomers. Enantiomer-selective GC identification of alpha- and beta-pinene produced by the recombinant enzyme are (-)-enantiomers. Left, Identification of the major product; right, identification of the minor product. A, A mixture of (+)- and (-)-authentic standards; B, enzymatic products; and C, coinjection of enzymatic products with authentic (-)-enantiomer. QH6 Transcripts in Different Tissues | Reverse transcriptase-PCR detected a high level of QH6 transcripts in juvenile leaves (2 weeks postsprouting and not fully expanded), and a lower level in stem, mature leaves (fully expanded), and inflorescences (data not shown). Northern-blot analysis also showed that QH6 transcripts were mainly present in juvenile leaves; much lower, but detectable levels were found in mature leaves, inflorescences, and stems . Figure 5 | Transcript abundance of QH6 in different organs of A. annua Transcript abundance of QH6 in different organs of A. annua A, Northern blot; B, ethidium bromide-stained gel for monitoring equal amount of total RNA loaded per lane. Each of 20 mug of total RNA was isolated from roots (lane 1), stems (lane 2), inflorescences (lane 3), mature leaves (lane 4), and juvenile leaves (lane 5). Twenty micrograms of total RNA was loaded per lane. Circadian Pattern of QH6 Gene Expression | Many plant monoterpene synthases, including QH1 and QH5 of A. annua , show an increase in steady-state mRNA level due to mechanical wounding . However, for the (-)-beta-pinene synthase QH6, the wounding-treatment of juvenile leaves caused a suppression, rather than an induction, of the steady-state level of transcripts. During the first 2 d postwounding, the QH6 transcripts in juvenile leaves were almost undetectable; at the 3rd d, the mRNA started to accumulate and reached a level comparable with the original at the 5th d postwounding (Fig. A). Similar to mechanical wounding, treatment with the fungal elicitor also suppressed the accumulation of QH6 transcripts in juvenile leaves, although the effect was more significant and longer lasting (Fig. B). Without these treatments, QH6 transcripts in the leaves collected in the same light phase did not show significant fluctuation, as can be seen in Figure A. For mature leaves in which the QH6 transcript abundance was much lower, no induction was detected by either of the treatments (data not shown). The treatment of juvenile leaves brought no systemic induction to the neighboring leaves, regardless of their age (data not shown). Figure 6 | Effects of mechanical wounding (approximately 40% damage of upper leaf surface; A) and Verticillium dahliae elicitor treatment (1 mug of Suc equivalent mL-1; B) on the steady-state level of QH6 transcripts. Effects of mechanical wounding (approximately 40% damage of upper leaf surface; A) and Verticillium dahliae elicitor treatment (1 mug of Suc equivalent mL-1; B) on the steady-state level of QH6 transcripts. In each panel, an ethidium bromide-stained gel was used for monitoring equal amount of total RNA loaded. RNA samples were isolated from juvenile leaves, which were before (+) and at 1 to 5 d (lanes 1 --5) posttreatment. Figure 7 | Circadian pattern of transcriptional regulation of QH6. Circadian pattern of transcriptional regulation of QH6. A, Fluctuation of steady-state level of QH6 transcripts with day-night shifting (d 1 --3); B, when plants were entrained by LL photoperiod, the oscillation pattern was in advanced cycles (d 4 --8); and C, entrained by DD photoperiod, the oscillation was in delayed cycles (d 4 --8). When plants were released from the entrainments, the original diurnal pattern of QH6 transcript fluctuation was resumed (d 9 and 10 in B and C). For each of the northern blots, the signal strength of QH6 was normalized against that of ubiquitin extension protein (UEP; internal control), and the relative percentage was plotted. We then investigated if other environmental factors could affect QH6 expression. Examination of the diurnal pattern of QH6 expression by northern-blot analysis showed that the steady-state mRNA level fluctuated with the day-night rhythm. With a 12 h/12 h (light/dark, LD) photoperiod, the QH6 mRNA level was generally higher in the day (L) than in the night (D). However, the mRNA level started to increase around 9th h in the dark phase (D9), approximately 3 h prior to the light phase. This accumulation reached its peak level in 12 h, i.e. at the 9th h in light (L9; Fig. A). A key issue is whether plants of A. annua could still maintain a circadian pattern of QH6 transcript accumulation when entrained in constant light (LL) or constant dark (DD; Fig. ). After the plants were moved to a LL regimen for 3 d (d 6 of Fig. ), the fluctuation was still clearly detectable, with the peak level appearing at D9, about 12 h ahead that of the original cycle at L9. The rhythm was further accelerated, as in d 8, two peaks appeared at around D9 and L12, respectively (Fig. B). When plants were entrained in DD, the transcripts accumulated much slower, and the level did not reach the peak until phase D3 in d 4. At the 3rd d under DD (d 6), the peak was delayed by about 15 h, and it was delayed by 12 h on d 8. In addition, the oscillation magnitude was reduced after 5 d of DD treatment (Fig. C). Therefore, the fluctuation cycles were advanced under LL and were delayed under DD conditions. When plants were returned from LL to LD, the peak of QH6 transcript abundance stayed at phase L12 in the 1st d, and then went back to its original phase L9 on the day after (Fig. B). For plants from DD, the fluctuation also returned to its original phase on the 2nd d under LD (Fig. C). Diurnal Pattern of (-)-beta-Pinene in Planta Content and Emission | Under LD conditions, contents of beta-pinene in the VOCs collected and the leaf extract fluctuated, and peaked at L9 , the time phase at which QH6 transcripts reached maximum. This indicates a close correlation between (-)-beta-pinene formation and QH6 gene expression, suggesting that (-)-beta-pinene synthase activity is mainly regulated at the transcriptional level. In the volatiles collected, a much lower amount of (-)-beta-pinene was emitted within 6 h before daybreak (D9 and D12), when the mRNA level was just past its lowest level. At the same time, the in planta content of (-)-beta-pinene was still clearly detectable in leaves . Figure 8 | Diurnal fluctuation of beta-pinene content. Diurnal fluctuation of beta-pinene content. A, beta-Pinene in total VOCs collected in each 3-h time phase (nanograms per plant); B, beta-pinene in volatiles extracted from juvenile leaves (nanograms per gram of fresh weight). Time phases in which samples were collected are labeled at the bottom. DISCUSSION : The Enzymatic Products of (-)-beta-Pinene Synthase, QH6 | Although beta-pinene is one of the most prevalent components of VOCs , no enzyme has been reported so far that produces only beta-pinene. All pinene synthases produce multiple products in vitro with GPP as substrate. For examples, the pinene synthase from grand fir produces (-)-alpha- and (-)-beta-pinene, at a ratio of 42:58 , and the (-)-pinene synthase from sage (Salvia officinalis) produces an assortment of five monoterpenes: (-)-camphene, (-)-alpha-pinene, (-)-beta-pinene, myrcene, and (-)-limonene . However, the enzyme cloned from A. annua revealed in vitro a great preference for (-)-beta-pinene, which represented 94% of the product, much higher than that produced by its counterpart in grand fir or sage. Furthermore, the beta-/alpha-ratios detected from the in vitro assay and from different plant tissues were different, and this may suggest the presence of another pinene synthase that produces more alpha-pinene in this plant. Spatial and Temporal Characteristics of (-)-beta-Pinene Biosynthesis | QH6 transcripts are organ specific, being mainly present in green tissues of leaves and stems. These organs bear glandular trichomes, which could theoretically serve for storing volatiles. QH6 transcripts and the (-)-beta-pinene itself (data not shown) cannot be detected from roots in this work. In a developmental manner, juvenile leaves, which are no older than 2 weeks postsprouting, have the highest steady-state level of QH6 mRNA. This is similar to the findings from peppermint in which monoterpenes were found to be synthesized only in 12- to 20-d-old leaves that were not yet fully expanded . This rapid decline during development is supported by , who found that in leaves of 4-month-old A. annua plants, no beta-pinene could be detected, whereas alpha-pinene represented as much as 26.7% of the total volatiles extracted. It is interesting to note that the QH6 mRNA level and in planta (-)-beta-pinene content displayed a circadian pattern of variation, and their peak levels even appeared in the same time phase (L9). This suggests that the biosynthesis of (-)-beta-pinene in A. annua is largely controlled at the transcriptional level. Such a regulatory pattern is similar to that of monoterpene biosynthesis in peppermint glandular trichomes , which showed coincidental temporal changes in steady-state transcript abundance and enzyme activities. Furthermore, the similar fluctuation of (-)-beta-pinene emission suggests that this monoterpene is probably not stored in glandular trichomes for long periods of time, but is released directly after synthesis from glandular or nonglandular tissues. Although alpha- and beta-pinene were reported to take part in direct plant defense against herbivores and pathogens, allelopathy, and plant pollination , little is known about the biological or ecological function of (-)-beta-pinene in A. annua. In contrast to QH1 and QH5, which could be induced by wounding, QH6 transcription was suppressed by mechanical wounding or fungal elicitation. It seems unlikely that (-)-beta-pinene synthase is involved in the chemical defense of A. annua. One possible explanation for this suppression is that those up-regulated "defensive enzymes" such as QH1 and QH5 would have more substrate (GPP) available, if other, possibly nondefensive, monoterpene synthases were down-regulated. A similar result was obtained with ponderosa pine (Pinus ponderosa) in which it was shown that an increase in limonene content was accompanied by the decreased levels of alpha-pinene, beta-pinene, 3-carene, and myrcene . Circadian Rhythm of Transcriptional Regulation of (-)-beta-Pinene Synthase | Under LD photoperiod, the fluctuation of QH6 transcripts showed an obviously diurnal pattern, which peaked shortly after noon (L9), although in our experiment, the light intensity did not change in the entire light phase. Because the transcripts started to accumulate well before entering the light phase (at D9), QH6 transcription seems not directly triggered by light, but controlled by internal oscillation signals. This is consistent with our findings that the circadian pattern persisted when the plants were transferred by LL or DD photoperiod, although the circadian oscillation of QH6 transcripts became accelerated (LL) or decelerated (DD). It has been proposed that a circadian rhythm is determined by three factors: a free-running period, which is determined by the nature of the gene in certain plants, temperature compensation of the free-running period, and the entrainment of the oscillator through photic and nonphotic time cues (zeitgeber; ). The (possible) only explanation for the varied oscillation frequency of QH6 transcript abundance reported herein is the zeitgeber from LL or DD entrainment signals of A. annua. It should be mentioned that (-)-beta-pinene is not the only example of a plant volatile that can be synthesized and emitted in a diurnal mode. Floral volatiles from rose (Rosa hybrida L. cv Honesty; ) and snapdragon (Antirrhinum majus; ), as well as damage-induced volatiles of cotton (Gossypium hirsutum; ), are emitted in a clear circadian rhythm. and pointed that monoterpene might improve thermotolerance of photosynthesis in leaves and all isoprenoids are thought to protect membrane from denaturation. It is possible that a peak amount of pinene, as found in this study for A. annua, is synthesized and released for protective purpose when light intensity and temperature are reaching their maximal values under nature conditions. Isopentenyl diphosphate (IPP) is the biosynthetic precursor of all the terpenoids. In Arabidopsis, transcripts of DXS, the first enzyme in the pathway to IPP biosynthesis in plastids , vary in a diurnal pattern that peaks at 4 h in the light phase . DXR generates IPP precursors and was thought to catalyze the critical step in the entire pathway . With A. annua, we found that both QH6, one of the monoterpene synthases, and DXR, a key enzyme of the plastid isoprenoid pathway, are up-regulated in the later light phase (S. Lu and X.-Y. Chen, unpublished data) when the biological clock in plant is set to allocating assimilated carbohydrates to different pathways . At this point, it is still unclear whether the QH6 transcription is directly regulated by an intrinsic circadian clock or by some intermediates in the pathway downstream of DXR, e.g. IPP or GPP. It remains to be learned what signal regulates DXR and QH6 transcription. MATERIALS AND METHODS : Plant Material | Seeds of Artemisia annua were collected from Guilin, Guangxi Province of China and were planted in the green house at 25C under a light intensity of 150 mumol m-2 s-1 with 12 h/12 h L/D regime. Leaves at 2 weeks postsprouting that were not yet fully expanded were designated as "juvenile" and those at 6 to 8 weeks postsprouting were designated as "mature." For stems, only the green tissue was sampled. The inflorescences used in this work did not include their axes and stipules. For transcriptional analyses, juvenile leaves were used, unless otherwise specifically noted. Plant Treatments | For mechanical wounding, the upper leaf surfaces of plants growing under LD were rubbed with fine sandpaper to approximately 40% surface damage. Elicitation treatment was performed by spreading Verticillium dahliae elicitor preparation at 1 mug of Suc equivalent mL-1 on the upper surfaces of leaves. Leaves were collected before, and at 1 to 5 d after each treatment, at time phase L9. For circadian analysis, a group of cultivated plants were entrained under LD. One group was kept under LD as control, and the rest were divided for two treatments. One was removed to LL, and the other to DD. After 5 d, all of these treated plants were returned to the original LD cycles. Juvenile leaves were collected from 3 d prior to different photoperiod treatments to 1 d after returning to LD, at 3-h intervals. In each of the triplicate experiments, the plants were destroyed once a leaf was sampled. Isolation of cDNA | PCR was conducted on the cDNA library of A. annua with the degenerate primers 9701 and 9315 . A number of cDNA fragments of terpene synthase candidates were obtained. From the sequence of one of the PCR amplicons, specific forward primer P2 (5'-GCTTCTTATCATTCAGTAGAGG-3') and reverse primer M35 (5'-TCGCAAACTCAAGCACC-3') were designed for screening the library by a PCR-mediated method, as previously described . The final isolated clone was named QH6 and the insert was sequenced. Prokaryotic Expression | A forward primer QH6Trun (5'-ATGAATTCCATGAACAGAAGATCAGCTA-3'), containing an EcoRI cloning site (underlined), was designed for truncating the transit peptide sequence just before the nucleotide sequence of R48R49 motif (boldface). With this and the universal reverse primer T7 (5'-GTAATACGACTCACTATAGGGC-3') and QH6 as template, the open reading frame was modified by PCR with Pfu DNA polymerase. The product was digested by EcoRI and XhoI and was ligated into pET-32b. The resulting plasmid, named pET32b/QH6, was confirmed by sequencing and was transferred into Escherichia coli BL21(DE3). The transformed E. coli cells were cultured overnight at 37C in Luria-Bertani (LB) medium supplemented with 100 mug mL-1 ampicillin. A 500-muL aliquot of the saturated culture was used to inoculate 50 mL of fresh LB medium containing 100 mug mL-1 ampicillin. After adding 1 mm isopropyl beta-d-thiogalactoside, the induced culture was allowed to grow at 22C overnight. The cells were pelted and resuspended in 10 mL of cold assay buffer and were disrupted by sonication. After centrifugation, the supernatant was then transferred to a glass tube and the insoluble fraction was resuspended in 10 mL of the same assay buffer. Enzyme Assay | To assay the terpenoid synthase activity of the bacterial extract, the supernatant was adjusted to a final concentration of 10 mm MgCl2, 20 mum MnCl2, and 30 mum GPP or FPP. Each reaction mixture was then covered with 2 mL of pentane to trap volatile products, sealed, and incubated at 30C for 3 h. The reaction mixture was extracted with pentane (3 x 2 mL). The remaining aqueous phase was subsequently extracted with diethyl ether (3 x 2 mL). The pentane extracts were combined and passed through a silica gel column (3 cm x 5 mm i.d.) covered with MgSO4 (1 cm) to yield the terpene hydrocarbon fraction. The ether extract was passed through the same silica column to provide an oxygenated terpene fraction. The pentane extract was concentrated to 500 muL under a nitrogen stream. The solvent in the ether extract was evaporated similarly to near dryness and was redissolved in 500 muL of ethyl acetate. Both samples were analyzed by GC and GC-MS. Product from E. coli BL21 (DE3) expressing empty pET32b vector was included as a negative control. To obtain sufficient product for NMR analysis, a large-scale preparation was conducted by inoculating 750 muL of an overnight culture into 750 mL of LB medium containing 100 mug mL-1 ampicillin. The culture was prepared similarly to the small-scale assay, except that the supernatant of the sonicate was adjusted to a final concentration of 10 mm MgCl2, 20 mum MnCl2, and 200 mum GPP. Cyclopentane (30 mL) was used instead of pentane to minimize the solvent interference in subsequent NMR analysis. After incubation at 30C overnight, the reaction mixture was extracted with cyclopentane (3 x 100 mL). The combined organic extracts were passed through a silica gel column (8 x 2 cm i.d.) overlaid with MgSO4 (2 cm) to yield the terpene hydrocarbon fraction. A 2-muL aliquot was analyzed by GC and the total amount of product was estimated to be approximately 1 mg based on the peak area-to-weight ratio calculated from the (-)-beta-pinene standard. The solvent was evaporated under a stream of nitrogen to near dryness for NMR analysis. GC, GC-MS, and NMR Analyses | To isolate A. annua essential oil, stems, inflorescences, and mature leaves were ground in liquid nitrogen to fine powder and were then extracted with hexane twice. The organic extracts were combined and were passed through a column of silica gel 60 (EM Science, Gibbstown, NJ) overlaid with MgSO4, and were concentrated under a nitrogen stream. For emitted volatiles, a collecting device according to with about 10 L of headspace and an air flow of 0.2 L min-1 was used, and the plants were grown under normal cultivation conditions as mentioned above. The released chemicals were trapped by 20 mL of hexane, which was finally condensed under nitrogen stream to 200 muL for GC analysis. The authentic (-)-beta-pinene was diluted with hexane in series and was then used for quantifying (-)-beta-pinene contents in samples. GC analyses were performed on a 6890 system (Hewlett-Packard, Palo Alto, CA) equipped with a Rtx-5 capillary column (30-m x 0.25-mm i.d., 0.10 mum df; Restek, Bellefonte, PA). Separation conditions were the following: injection port 200C, flame ionization detector 250C, a split ratio of 39:1 for plant extracts and enzymatic products, and splitless for volatile collections, and helium flow at 20 cm s-1 (0.6 mL min-1). The temperature program was 50C for 5 min, an increase to 250C (10C min-1), and 250C for 5 min. GC-MS was performed on a 5890A instrument (Hewlett-Packard) equipped with a DB-5 ms column (60-m x 0.25-mm i.d., 0.10 mum df; J&W Scientific, Folsom, CA). Separation was achieved with splitless injection at 200C, helium flow at 30 cm s-1 (1 mL min-1), and the identical temperature program as above. Mass spectra (m/z 35 --500) were obtained on a ZAB-HF reverse-geometry double-focusing instrument at 70 eV with an electron-impact ion source (200C). The accelerating voltage was 8 kV and the resolution was 1,000 (10% valley). Enantioselective GC was performed on a GC-9A instrument (Shimadzu, Tokyo) with a CycloSil-B capillary column containing a modified beta-cyclodextrin stationary phase (30-m x 0.25-mm i.d., 0.25 mum df; J&W Scientific). The separation of chiral components was achieved with injection port and flame ionization detector at 200C, helium flow at 29 cm s-1 (0.8 mL min-1), a 36:1 split ratio, and the following temperature program: 50C for 1 min and a 5C min-1 ramp to 200C, where the temperature was held for 10 min. Authentic (+)-alpha-pinene, (-)-alpha-pinene, (+)-beta-pinene, and (-)-beta-pinene from Sigma-Aldrich (St. Louis) were used as standards. The QH6 product was also coinjected with authentic samples to determine the pinene stereochemistries. 1H NMR spectra were obtained on a AMX500 spectrometer (500.1 MHz for 1H; Bruker, Billerica, MA) at 25C in CDCl3 solution (approximately 10 mm) and were referenced to internal tetramethylsilane. Transcriptional Analysis | RNA was extracted according to . A total of 2 mug of RNA was used as template for reverse transcription by using a Reverse Transcription System (Promega, Madison, WI) with oligo (dT)15 as the primer. As an internal control, transcripts of UEP were detected by a forward primer UbiF2 (5'-CTTGGGGGAAGACGGGC-3') and a reverse primer UbiR2 (5'-GCCAAGATTCAGGACAAGGAAGG-3'). Primers P2 and M35 were used to analyze QH6 transcripts. For northern-blot analysis, a total of 20 mug of RNA was loaded per lane, and an ethidium bromide-stained gel or a duplicated blot probed for UEP transcripts was used to monitor the amount of RNA loaded. A fragment containing the first 566 bp of QH6 (released by EcoRI and HindIII digestion of pBK-CMV/QH6), which showed a relatively lower sequence identity with QH1 or QH5 , was 32P labeled by the Prime-a-Gene System (Promega) to detect QH6 transcripts on the membrane. The blots were hybridized, washed as described , and exposed to x-ray film at -70C. The images were collected with GeneGenius gel documentation system and were then digitalized with GeneTools software (Syngene, Cambridge, UK). Corresponding data from UEP transcripts were used for normalization. Backmatter: PMID- 12226527 TI - Nitric Oxide Negatively Modulates Wound Signaling in Tomato Plants AB - Synthesis of proteinase inhibitor I protein in response to wounding in leaves of excised tomato (Lycopersicon esculentum) plants was inhibited by NO donors sodium nitroprusside and S-nitroso-N-acetyl-penicillamine. The inhibition was reversed by supplying the plants with the NO scavenger 2-(4-carboxiphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide. NO also blocked the hydrogen peroxide (H2O2) production and proteinase inhibitor synthesis that was induced by systemin, oligouronides, and jasmonic acid (JA). However, H2O2 generated by glucose oxidase and glucose was not blocked by NO, nor was H2O2-induced proteinase inhibitor synthesis. Although the expression of proteinase inhibitor genes in response to JA was inhibited by NO, the expression of wound signaling-associated genes was not. The inhibition of wound-inducible H2O2 generation and proteinase inhibitor gene expression by NO was not due to an increase in salicylic acid, which is known to inhibit the octadecanoid pathway. Instead, NO appears to be interacting directly with the signaling pathway downstream from JA synthesis, upstream of H2O2 synthesis. The results suggest that NO may have a role in down-regulating the expression of wound-inducible defense genes during pathogenesis. Keywords: Introduction : Nitric oxide (NO) and reactive oxygen species (ROS) have important roles in the activation of defense responses against pathogen attacks (for review, see ; ; ). The addition of sodium nitroprusside (SNP), an NO donor, can cause cell death to soybean (Glycine max) suspension cultures at millimolar concentrations, when ROS are present . The molecular mechanism for NO synthesis or action is currently unknown. NO can activate or inhibit certain heme-containing enzymes , and it can stimulate plant defense responses through a cGMP-dependent signaling cascade involving, at least in some cases, the generation of cADPR and the activation of mitogen-activated protein kinases . The generation of ROS in response to pathogen and herbivore attacks has been well documented (for review, see ; ; ; ). One ROS species, hydrogen peroxide (H2O2), appears to be a key signaling molecule as well as a defensive chemical that physically damages attacking organisms . Tomato (Lycopersicon esculentum) plants are known to accumulate relatively high levels of H2O2 in response to wounding and elicitors, without apparent toxicity to the plants . Although NO has been associated with ROS and the activation of defense responses against pathogens, its possible role in wound signaling has not been reported. Herein, we show that supplying young excised tomato plants with NO donors strongly inhibited the expression of wound-inducible proteinase inhibitors, but did not inhibit the activation of octadecanoid pathway genes. The inhibitory action of NO did not involve the synthesis of salicylic acid (SA), a potent inhibitor of wound-inducible defense gene signaling . RESULTS : The effect of NO on the synthesis of proteinase inhibitor I (Inh I) in leaves of young tomato plants in response to wounding was initially studied using two NO donors, S-nitroso-N-acetyl-penicillamine (SNAP) and SNP, which are known to elevate levels of NO when supplied to plants . When young tomato plants were supplied through their cut stems with solutions of SNAP or SNP and wounded 1 h later, the synthesis and accumulation of Inh I in response to wounding was severely diminished compared with control plants that were not supplied with NO donors (Fig. A). The inhibition of the wound response by both SNAP and SNP was reversed by the simultaneous addition of the NO scavenger 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (CPTIO; Fig. A). Wounding tomato leaves does not cause an increase in NO , and CPTIO and the NO synthase inhibitors S,S'-1,3-phenylene-bis(1,2-ethanediyl)-bis-isothiourea and Nomega-nitro-l-Arg had no effect on the wound response (data not shown). SNP inhibited the wound-induced accumulation of proteinase Inh I in a concentration-dependent manner. The concentration of SNP required for half-maximal inhibition was about 0.1 to 0.2 mm (Fig. B). SNP similarly inhibited accumulation of proteinase inhibitor I induced in tomato plants by the elicitors systemin, jasmonic acid (JA), and oligogalacturonides (OGA) . Treatment of plants with SNAP under the same conditions as SNP resulted in similar effects (data not shown). Figure 1 | NO inhibits the wound-induced accumulation of proteinase Inh I in tomato leaves. NO inhibits the wound-induced accumulation of proteinase Inh I in tomato leaves. A, Effect of NO donors on the accumulation of Inh I. B, Concentration dependence of SNP inhibition of proteinase Inh I accumulation induced by wounding. Fourteen-day-old plants having two expanded leaves and a small apical leaf were excised at the base of the stems and supplied through the stems with a solution of phosphate buffer alone, pH 6.5, for 1 h (control) or a buffer containing 1.0 mm SNAP, 1.0 mm SNP, 1.0 mm SNAP + 1.0 mm CPTIO, or 1.0 mm SNP + 1.0 mm CPTIO for 1 h. All experiments were carried out under light (300 muEm-2 s-1). Plants, except controls, were wounded twice at the middle of each expanded leaf, perpendicular to the main petiole and were incubated in water under light as described in "Materials and Methods." Proteinase Inh I levels in leaf extracts were assayed immunologically in leaf juice 24 h later. Data are means +- sd; n = 6. Figure 2 | Effect of SNP on NO accumulation in leaves of young excised tomato plants. Effect of SNP on NO accumulation in leaves of young excised tomato plants. The plants were supplied with phosphate buffer alone (control) or 1.0 mm SNP in buffer for 1 h. Leaf extracts of wounded and unwounded plants were assayed for NO accumulation at the times indicated. Data are means +- sd; n = 4. Figure 3 | Inhibition of elicitor-induced accumulation of proteinase Inh I by SNP. Inhibition of elicitor-induced accumulation of proteinase Inh I by SNP. Excised tomato plants were supplied through the stem with phosphate buffer alone (control) or with 1.0 mm SNP for 1 h and were transferred for 30 min to 25 nm systemin, 250 mug mL-1 OGA, and 100 mum JA in phosphate buffer as described in "Materials and Methods." After each treatment, plants were incubated in water for 24 h and were then immunologically assayed for proteinase Inh I content in leaf juice. Data are means +- sd; n = 6. In leaves of young tomato plants, the genes that code for components of the octadecanoid signaling pathway are activated within 0.5 to 1 h after wounding . This is in contrast with proteinase inhibitor genes that are activated about 4 to 12 h after wounding . Gel-blot analyses were carried out to determine whether SNP inhibited the early (0.5 --1 h) and/or late (4 --12 h) genes induced by wounding, systemin, or JA. The early-associated signaling pathway mRNAs included prosystemin, lipoxygenase, allene oxide synthase, and polygalacturonase catalytic subunit. The late-associated mRNAs included Inh I, Inh II, cathepsin D inhibitor, and metallocarboxypeptidase inhibitor . The levels of mRNAs coding for the signaling pathway-related proteins induced by wounding, systemin, or JA were not inhibited by SNP, whereas levels of mRNAs encoding the defensive genes were all strongly reduced when SNP was present . Therefore, SNP was not blocking the activation of wound-inducible signaling pathway genes, but was inhibiting the pathway downstream from JA. Figure 4 | Effects of SNP on the expression of genes induced by wounding, systemin, and JA. Effects of SNP on the expression of genes induced by wounding, systemin, and JA. Young excised tomato plants were supplied with phosphate buffer alone (control) or 1.0 mm SNP for 1 h. Plants, except controls, were wounded, transferred to water, and assayed by RNA gel blotting after 2 h for allene oxide synthase, lipoxygenase, prosystemin, and polygalacturonase catalytic subunit (and after 8 h for proteinase Inh), proteinase Inh II, cathepsin D inhibitor, and metallocarboxypeptidase inhibitor. Equal amounts of RNA were loaded as confirmed by probing with an ubiquitin cDNA. Tomato leaves had previously been shown to maximally accumulate H2O2 between 4 and 6 h following wounding, decreasing thereafter . Here, we report the direct quantification of wound-inducible H2O2 that accumulated in wounded and unwounded leaves of young tomato plants 6 h after they had been supplied with SNP. SNP reduced the accumulation of H2O2 to less than 50% of levels caused by excision alone (unwounded control), and by wounding, systemin OGA, and JA . Figure 5 | Effect of SNP on wound- and elicitor-induced accumulation of H2O2. Effect of SNP on wound- and elicitor-induced accumulation of H2O2. Young tomato plants were treated as described in Figure . H2O2 concentration was measured 6 h after elicitor treatment as described in "Materials and Methods." Previous research has shown that H2O2 can act as a second messenger for the expression of the late-associated defensive genes during the wound response . An H2O2-generating system composed of Glc oxidase plus Glc was employed to generate enough H2O2 to cause the induction and accumulation of defensive proteinase inhibitor proteins in excised tomato plants , and was used to investigate whether NO can inhibit H2O2-mediated synthesis of Inh I. Young excised tomato plants were supplied with SNP for 1 h and then with Glc and Glc oxidase. The results shown in Figure indicate that SNP did not block the synthesis of Inh I induced by H2O2, suggesting that the site of inhibition of late genes was at a step or steps between JA and H2O2 generation. Figure 6 | Inhibition of H2O2-mediated accumulation of proteinase Inh I by SNP. Inhibition of H2O2-mediated accumulation of proteinase Inh I by SNP. Excised tomato plants were supplied through the stem with phosphate buffer alone (control) or 1.0 mm SNP for 1 h and were transferred for 1 h to a buffer containing 50 mum Glc plus 2.5 units mL-1 Glc oxidase. Plants were then incubated in water for 24 h and were immunologically assayed for proteinase Inh I content in leaf juice. Data are means +- sd; n = 6. Because NO had previously been reported to induce the synthesis of SA in tobacco (Nicotiana tabacum) leaves , the possibility that NO may be inducing the accumulation of SA was considered as a possible cause of the inhibitory action of NO on wound signaling. To evaluate whether NO is able to inhibit wound-defense gene expression in the absence of SA, NahG transgenic tomato plants overexpressing the bacterial salicylate hydroxylase enzyme, which removes endogenous SA by converting it to catechol , were assayed to see if SNP had the same inhibitory effect on the wound inducibility of Inh I protein in the wild-type and the transgenic NahG plants. NO had the same inhibitory effect in the NahG plants as in wild-type plants . In addition, direct quantification of SA in leaves of unwounded and wounded wild-type tomato plants that were pretreated with SNP showed no differences compared with SA found in leaves of untreated plants within 8 h after SNP treatment (data not shown). It is within this time period that wound signaling takes place , indicating that SA was likely not to have a significant effect on wound signaling. Figure 7 | NO-mediated inhibition of wound-inducible accumulation of proteinase Inh I in SA-deficient (NahG) transgenic tomato plants. NO-mediated inhibition of wound-inducible accumulation of proteinase Inh I in SA-deficient (NahG) transgenic tomato plants. Plants were treated and assayed as described in Figure . Data are means +- sd; n = 6. DISCUSSION : Pathogen-induced production of H2O2 and NO in plant cells has been shown to regulate the hypersensitive response and cell death . H2O2 is also generated in response to mechanical wounding, and acts as a second messenger that regulates the expression of wound-inducible defense-associated genes . Whether NO has a role in the wound-inducible signaling pathway has not been assessed. Therefore, a possible role for NO in the wound-inducible signaling pathway was investigated in tomato plants. Supplying young excised tomato plants with the NO generators SNP and SNAP before wounding caused a nearly complete inhibition of the induction of synthesis of proteinase Inh I, one of several wound-inducible proteinase inhibitor proteins in tomato leaves (Fig. A). The inhibition of wound-inducible Inh I by SNP was concentration dependent (Fig. B) and could be reversed by supplying the NO scavenger CPTIO (Fig. A). When wounded and unwounded excised tomato plants were supplied with 1 mm SNP, NO levels from 0.7 to 1.0 mm were detected in both set of plants within 0.5 h. The levels remained essentially constant for the next 1.5 h, and by 4 h, the levels had declined to 0.5 mum . However, in the absence of the NO donors, NO levels remained unchanged in the leaves of wounded and unwounded plants, indicating that NO synthesis was not regulated by the wounding. NO inhibited Inh I synthesis in young, excised tomato plants induced by systemin, OGA, and JA , indicating that the inhibition of the signaling pathway likely occurred downstream from JA synthesis. Previous reports had demonstrated that wounding, systemin and JA all induced the expression of signal pathway genes leading to JA synthesis within 0.5 to 2 h after wounding, whereas defensive proteinase inhibitor genes were synthesized much later, from 4 to 12 h after wounding . Further studies indicated that JA was not only activating the early signal pathway genes, but was activating the expression of a polygalacturonase gene and the production of H2O2 , with H2O2 acting as a second messenger for proteinase inhibitor gene expression . Therefore, the effects of NO on the expression of early and late genes induced by wounding, systemin, and JA were examined. NO did not inhibit signal pathway gene expression, but strongly inhibited the expression of several proteinase inhibitor genes . Supplying the excised tomato plants with an H2O2-generating system of Glc oxidase plus Glc is known to induce the expression of the proteinase inhibitor genes, but not the early-inducible signal pathway-associated genes . In this study, the levels of H2O2 detected in the leaves of plants supplied with this H2O2-generating system were not affected by the presence of SNP (data not shown), and the H2O2-induced synthesis of proteinase Inh I was reduced by only 12% . H2O2 levels generated in the plants by Glc/Glc oxidase were approximately 13 mum, which is near the levels induced by wounding in the wild-type tomato plants . NO also inhibited the accumulation of H2O2, which occurs in planta in response to wounding or treatment with chemical elicitors, and JA . Together, these results suggest that NO is inhibiting signaling downstream from JA, but before the steps that generate H2O2. The wound signaling pathway in tomato plants was shown previously to be inhibited by SA , indicating that "crosstalk" between the pathogen-inducible defense signaling pathway and the herbivore (wound)-inducible defense signaling pathways can be mediated by SA . However, SNP inhibited the wound-induced accumulation of proteinase inh I in SA-deficient tomato plants , and SA levels remained unaffected in the wild-type plants during the first 8 h after wounding (data not shown). Thus, the effects of NO were likely not due to the induction of synthesis of SA. During plant defense against pathogens, NO potentiates the hypersensitive cell death in soybean cell cultures , and inhibition of NO synthesis compromises hypersensitive disease resistance in Arabidopsis and tobacco plants . Moreover, NO at 0.5 to 1.0 mm mediates plant defense gene activation, triggering the expression of pathogenesis-related proteins and Phe ammonia lyase, and the synthesis of protective natural products . However, in this study of the tomato wound-defense response, NO at 1.0 mm inhibited the activation of antiherbivory proteinase inhibitor genes, which is mediated by the accumulation of nonlethal levels of about 10 mum H2O2 . In this latter system, NO might be acting as an antioxidant agent , protecting the plant cells and tissues from ROS damage. It is interesting that after 24 h, no symptoms of necrosis or hypersensitive cell death were observed in the leaves of plants treated with SNP and exposed at the same time to the H2O2-generating system of Glc plus Glc oxidase. Therefore, in addition to H2O2 and NO, SA may be required for the onset of pathogen-induced programmed cell death in tomato (for review, see ; ; ). Although there is no evidence to support a specific mechanism of inhibition by NO, NO appears to be specifically inhibiting a step between JA synthesis and H2O2 production . Although NO does not seem to be a component of the wound defense response in tomato plants, it might act antagonistically to inhibit the expression of antiherbivory defense genes during the plant defense response against pathogens. In this regard, it has been reported that NO can antagonize ethylene biosynthesis and action in plants (Leshem and Pinchasov, 2000). Because ethylene is required for the wound-defense response in tomato plants , NO may be interfering with the participation of ethylene in the wound signaling pathway. However, at present, the specific molecular target for NO inhibition of wound signaling remains to be identified. Figure 8 | Differential regulation of signal pathway genes and defensive genes in leaves of tomato plants in response to wounding and pathogens. Differential regulation of signal pathway genes and defensive genes in leaves of tomato plants in response to wounding and pathogens. In this model, wounding or systemin produces JA, which up-regulates the signal pathway genes (early genes). NO, produced after pathogen attacks, inhibits only the wound-inducible (late) genes. MATERIALS AND METHODS : Plant Material and Bioassays | Two-week-old tomato (Lycopersicon esculentum cv Castlemart and Moneymaker) plants and the transgenic tomato line cv Moneymaker containing the salicylate hydroxylase (nahG) gene were grown from seeds in growth chambers having 18-h days of 300 muEm-2 s-1 of light at 28C and 6-h nights at 18C. The plants had two expanding leaves and a small developing apical leaf when used for experimentation. SNAP and CPTIO were from Molecular Probes (Eugene, OR). All other reagents were from Sigma. To investigate the effects of NO on the induction of proteinase Inh I synthesis and accumulation in response to wounding and chemical elicitors, 14-d-old tomato plants were excised at the base of the stem and were supplied for 1 h with 10-3m potassium phosphate buffer, pH 6.5, or buffer containing NO-related compounds. Thereafter, the plants were wounded across the main vein of each terminal leaflet or were incubated for another 0.5 to 1.0 h in buffer solutions alone or a buffer solution containing systemin (25 nm), OGA (250 mug mL-1), JA (100 mum), or Glc (50 mum) plus Glc oxidase (2.5 units mL-1), as previously described . The plants were transferred to glass vials containing water, placed within closed Plexiglas boxes, and incubated for 24 h in light (300 mEm-2 s-1) at 28C. Levels of wound- and elicitor-inducible proteinase Inh I protein was quantified in juice expressed from the leaves by use of radial immunodiffusion assays , or the Inh I and II mRNAs were analyzed by gel-blot analyses. RNA Gel-Blot Analyses | Leaves of treated and control tomato plants were removed and immersed in liquid nitrogen, ground to a fine powder, and stored at -80C to isolate total RNA. Total RNA was extracted, fractionated by electrophoresis in 1.4% (w/v) agarose-formaldehyde gels, blotted onto nylon membranes, and hybridized with radioactive 32P-dCTP-labeled probes as described previously . An 18S ribosomal RNA gene probe was used as a loading control. Membranes were washed once with 2x SSPE for 20 min at room temperature, two to three times with 2x SSPE and 1% (w/v) SDS for 15 to 30 min at 65C, and then exposed for 15 to 32 h to x-ray film or to a PhosphorImager (Bio-Rad, Hercules, CA). Quantification of NO | Leaves of young excised tomato plants that had been pretreated with water or SNP for 1 h and then wounded or not wounded were assayed for NO concentration. In brief, 200 mg of frozen leaves of young tomato plants were grounded and homogenized in 1 mL of cooled buffer (0.1 m sodium acetate, 1 m NaCl, and 1% [w/v] ascorbic acid, pH 6.0). The homogenates were centrifuged at 10,000g for 20 min at 4C and the supernatants were passed through 0.8- x 4-cm columns in 1-X8 resin (Bio-Rad). NO was quantified in cleared extracts spectrophotometrically measuring the conversion of oxyhemoglobin to methemoglobin . Quantification of H2O2 | The quantification of H2O2 in extracts from tomato leaves was according to . Leaves were frozen and ground to a powder under liquid nitrogen and stored at -80C. Leaf powder (500 mg) was extracted with 1 mL of 0.2 m HClO 4, incubated on ice for 5 min, and pelleted by centrifugation at 10,000g for 10 min at 4C. The supernatant was neutralized to pH 7.0 to 8.0 with 0.2 m NH 4OH, pH 9.5, and was briefly centrifuged at 3,000g for 2 min to sediment the insoluble material. The extracts were passed through 0.8- x 4-cm columns of AG 1X-8 resin (ionic-form chloride; Bio-Rad) and were eluted with double-distilled water . The quantification of H2O2 in the cleared extracts was carried out using an Amplex Red Hydrogen Peroxide Assay kit (Molecular Probes), following the manufacturer's recommendations. In brief, 50 to 100 muL of extract was mixed with an equal volume of a solution containing 1 U mL-1 horseradish peroxidase in 50 mm sodium phosphate buffer, pH 7.4, and was incubated for 1 h at room temperature. Fluorescence was measured with a fluorescence microplate reader (Perkin-Elmer, Beaconsfield, Buckinghamshire, UK) using excitation at 560 +- 5 nm and fluorescence detection at 590 +- 5 nm. The concentration of H2O2 in each sample was calculated using a standard curve obtained with known concentrations of pure H2O2. Quantification of SA | Leaves of young excised tomato plants that had been pretreated with water or SNP for 1 h and then wounded or not wounded were assayed for SA. SA was extracted from leaves with methanol and was quantified by using HPLC as described by . Backmatter: PMID- 12226528 TI - Resistance of Cultivated Tomato to Cell Content-Feeding Herbivores Is Regulated by the Octadecanoid-Signaling Pathway AB - The octadecanoid signaling pathway has been shown to play an important role in plant defense against various chewing insects and some pathogenic fungi. Here, we examined the interaction of a cell-content feeding arachnid herbivore, the two-spotted spider mite (Tetranychus urticae Koch), with cultivated tomato (Lycopersicon esculentum) and an isogenic mutant line (defenseless-1 [def-1]) that is deficient in the biosynthesis of the octadecanoid pathway-derived signal, jasmonic acid (JA). Spider mite feeding and fecundity on def-1 plants was significantly greater than on wild-type plants. Decreased resistance of def-1 plants was correlated with reduced JA accumulation and expression of defensive proteinase inhibitor (PI) genes, which were induced in mite-damaged wild-type leaves. Treatment of def-1 plants with methyl-JA restored resistance to spider mite feeding and reduced the fecundity of female mites. Plants expressing a 35S::prosystemin transgene that constitutively activates the octadecanoid pathway in a Def-1-dependent manner were highly resistant to attack by spider mites and western flower thrips (Frankliniella occidentalis), another cell-content feeder of economic importance. These findings indicate that activation of the octadecanoid signaling pathway promotes resistance of tomato to a broad spectrum of herbivores. The techniques of amplified fragment length polymorphism (AFLP) and bulk segregant analysis were used to map the Def-1 gene to a region on the long arm of chromosome 3 that is genetically separable from the map position of known JA biosynthetic genes. Tight linkage of Def-1 to a T-DNA insertion harboring the maize (Zea mays) Dissociation transposable element suggests a strategy for directed transposon tagging of the gene. Keywords: Introduction : Plant resistance to arthropod herbivores is often mediated by phytochemicals that negatively affect the feeding, growth, or reproduction of the attacking pest . Although many defensive compounds have been identified from diverse plant species, relatively little is known about the underlying genetic mechanisms that control their biosynthesis in response to developmental and environmental cues. Lycopersicon spp. provide an attractive model system to address this question. Cultivated tomato (Lycopersicon esculentum) is a natural host to over 100 arthropod herbivores that feed on roots, leaves, or fruit . Included among the major pests of tomato are adult and larval stages of Coleoptera (beetles), Lepidoptera (moths), Diptera (flies), Thysanoptera (thrips), Heteroptera (true bugs), Homoptera (aphids and whiteflies), and Acari (spider mites). Natural resistance of tomato to many herbivores is attributed to both constitutive and inducible defensive phytochemicals . Among the most thoroughly studied inducible defenses in tomato are proteinase inhibitor (PI) proteins that inhibit digestive enzymes in the gut of some insect herbivores . Wound-induced expression of PI genes is controlled by the jasmonate family of signaling molecules that includes jasmonic acid (JA), its methyl ester (MeJA), and their metabolic C18 precursor, 12-oxo-phytodienoic acid . Jasmonates are synthesized from linolenic acid via the octadecanoid pathway . In tomato leaves, jasmonate biosynthesis is positively regulated by wounding and by leaf-derived wound signals such as systemin . Genetic analysis indicates that systemin and its precursor protein, prosystemin, are upstream components of a signaling cascade that involves both the synthesis and perception of jasmonates . A tomato mutant (defenseless-1 [def-1]) that is deficient in wound- and systemin-induced JA accumulation and expression of downstream target genes was shown to be more susceptible to attack by Manduca sexta larvae, indicating that the octadecanoid pathway is essential for defense against chewing insects . Theses findings have been extended to field studies showing that exogenous jasmonate promotes resistance of tomato plants to insects in agricultural settings . In contrast to the firmly established role of octadecanoid signaling in plant defense against chewing insects, much less is known about how this pathway affects the interaction of plants with herbivores that use a piercing/sucking mode of feeding . The two-spotted spider mite (Tetranychus urticae Koch) represents one such economically important pest of a wide range of plants including many fruit, vegetable, grain, and ornamental crops, and it is perhaps the most serious pest in greenhouses around the world . The two-spotted spider mite uses specialized stylets to puncture epidermal cells of the host tissue; subsequent withdrawal of cellular contents leads to collapse of the underlying mesophyll tissue and formation of a chlorotic lesion at the site of feeding. Severe infestations usually result in complete desiccation and death of the affected tissue. Resistance of some wild tomato species to two-spotted spider mite has been shown to involve trichome-based physical (i.e. entrapment) and chemical (i.e. toxicity) mechanisms (e.g. ). However, the role of octadecanoid signaling in shaping the interaction between tomato and two-spotted spider mite, or other cell-content feeders, has not been thoroughly explored. To address this question, we studied the performance of two-spotted spider mite on near-isogenic lines of tomato in which this signaling pathway is either attenuated by def-1, or genetically enhanced by overexpression of prosystemin. Our findings, together the results of previous studies, indicate that octadecanoid signaling plays a critical role in regulating defense responses of tomato to a broad spectrum of herbivore pests. As a step toward understanding the molecular function of Def-1 in octadecanoid defense signaling, we mapped the Def-1 locus to the distal end of the long arm of chromosome 3. RESULTS : Octadecanoid-Mediated Resistance of Tomato to Cell-Content Feeding Herbivores | We used the JA-deficient def-1 mutant to investigate the role of the octadecanoid pathway in resistance to the two-spotted spider mite. Two-leaf-stage (15-d-old) wild-type (WT) and def-1 plants were infested with adult female mites that had been reared on bean plants, a preferred host of the two-spotted spider mite. Mites initiated feeding on both host genotypes within 1 d of challenge as evidenced by the appearance of small (approximately 0.25 mm2) chlorotic lesions at the feeding site. Estimation of leaf damage during a time course of infestation indicated that def-1 plants were significantly more susceptible than WT to mite feeding (Fig. A). This effect was accompanied by a significant increase in the number of mite eggs found on def-1 leaves compared with WT leaves (Fig. B). These results indicate that the Def-1 gene plays an important role in reducing the quality of tomato leaves as a food source and oviposition substrate for two-spotted spider mite. Figure 1 | Performance of two-spotted spider mite on wild-type and def-1 plants. Performance of two-spotted spider mite on wild-type and def-1 plants. Five different sets of 15-d-old wild-type (black bar) and def-1 (white bar) plants were challenged with adult female mites (10 mites per plant). Leaf damage (A) and egg counts (B) were determined at various times thereafter, using one set of plants to evaluate each time point. Ten plants of each genotype were used for each time point except the 10-d point, where eight plants per genotype were used. Values represent the mean and sd. Two-way ANOVA was used to evaluate the statistical significance of differences in leaf damage and egg count at each time point. Single asterisks denote a significant difference at P < 0.05. Double asterisks denote a significant difference at P < 0.0001. The differential performance of two-spotted spider mite on WT and def-1 plants suggested that the octadecanoid pathway regulates the production of defensive compounds in leaves of WT plants. To test this hypothesis, the level of Ser PI-II, a well-characterized marker of octadecanoid signaling in tomato , was measured in untreated and mite-infested plants (Fig. A). WT plants subjected to mite feeding for 10 d accumulated high levels of PI-II relative to untreated control plants. Within this population of WT plants, a positive correlation (r2 = 0.49) was observed between the level of leaf damage and PI-II accumulation. PI-II levels in mite-infested def-1 plants were only slightly greater than the detection limit of the assay (approximately 15 mug mL-1), even though mutant plants received approximately 3.9-fold greater damage than WT. To further test the hypothesis that the octadecanoid pathway regulates induced defense responses to spider mites, levels of endogenous JA were measured in control and infested leaves of WT and def-1 plants. Mite feeding resulted in a 2.6-fold increase (P < 0.05) in JA accumulation in WT plants, whereas JA levels of def-1 plants were unaffected by herbivory (Fig. B). These findings demonstrate that two-spotted spider mite feeding strongly induces PI-II accumulation in tomato leaves and that this response involves activation of the octadecanoid pathway. Figure 2 | The octadecanoid signaling pathway regulates the synthesis of PI-II in response to spider mite feeding. The octadecanoid signaling pathway regulates the synthesis of PI-II in response to spider mite feeding. A, Fifteen-day-old wild-type (WT) and def-1 plants were either not treated (black bars) or were infested (white bars) with 20 adult female mites per plant (10 mites per leaf). PI-II levels were measured in both leaves of individual plants 10 d after infestation. Values indicate the mean and sd of each treatment group (n = 14). Different lowercase letters denote a significant difference at P < 0.01 (Student's t test). B, Sixteen-day-old WT and def-1 plants were infested with spider mites as described in "Materials and Methods." Two days after challenge, leaflets showing visible signs of damage were harvested for analysis of JA levels (white bars). JA was also quantified in leaflets of noninfested plants (black bars). Values indicate the mean and sd of three independent experiments. Different lowercase letters denote a significant difference at P < 0.05 (Student's t test). RNA-blot hybridization was used to determine whether spider mite-induced accumulation of PI-II protein results from activation of the PI-II gene. In accordance, plants were infested with spider mites and RNA was prepared from leaf tissue 1, 2, or 3 d thereafter . PI-II mRNA levels in WT plants were strongly up-regulated (>50-fold relative to untreated controls) 1 d after challenge, and remained high at the 2- and 3-d time points. The expression pattern of two other PI genes, PI-I and cathepsin D inhibitor (CDI), was essentially identical to that of PI-II. Consistent with the data on PI-II protein accumulation , PI transcript levels in def-1 plants were less than 10% of that in WT. These findings indicate that two-spotted spider mite feeding activates the octadecanoid pathway leading to the coordinate expression of several defense-related genes, and that Def-1 plays an essential role in this induced response. Feeding of spider mites on WT plants resulted in a modest and gradual increase in the accumulation of LoxD and AOS1, two wound-inducible transcripts that encode the octadecanoid pathway enzymes lipoxygenase and allene oxide synthase, respectively (Fig. ; ; ). Interestingly, the mite-induced expression pattern of LoxD and AOS1 in def-1 plants was nearly identical to that observed in WT. This finding is consistent with previous studies suggesting that tomato uses genetically distinct signaling pathways for the regulation of different classes of wound-responsive genes . Figure 3 | Accumulation of defense-related mRNAs in response to spider mite feeding. Accumulation of defense-related mRNAs in response to spider mite feeding. Wild-type and def-1 plants (15-d-old) were challenged with 20 adult female mites as described in the legend to Figure . At various times (1, 2, or 3 d) thereafter, leaf tissue from 10 plants was harvested from control (0 d) and mite-infested leaves for RNA isolation. RNA-blot hybridization was performed using 32P-labeled cDNAs for Ser PI-I and PI-II, cathepsin D inhibitor (CDI), lipoxygenaseD (LoxD), and allene oxide synthase (AOS1). Blots were also hybridized to a probe for translation initiation factor eIF4A as a loading control. To test the hypothesis that increased performance of two-spotted spider mite on def-1 plants results from a deficiency in JA-induced defense responses, experiments were conducted to determine whether exogenous MeJA could restore resistance to the mutant. The results showed that spider mites caused significantly less damage on MeJA-treated def-1 plants compared with control def-1 plants (Fig. A). Moreover, the fecundity of female mites on MeJA-treated def-1 plants was significantly reduced relative to controls (Fig. B). These findings indicate that jasmonate is necessary and sufficient to restore defense of the mutant against two-spotted spider mite. MeJA treatment also reduced spider mite feeding and fecundity on WT plants, as recently reported by . This observation is consistent with the notion that applied MeJA induces the synthesis of defensive compounds that are normally produced in response to herbivory. Control experiments demonstrated that MeJA vapor, at concentrations 50-fold higher than those used for experiments with tomato, had no significant effect on the mortality or fecundity of mites reared on excised bean leaves (data not shown). Thus, direct toxicity of MeJA vapor to two-spotted spider mite is not likely responsible for the observed effects. Figure 4 | Exogenous MeJA protects def-1 plants from spider mite attack. Exogenous MeJA protects def-1 plants from spider mite attack. Wild-type and def-1 plants were treated for 24 h in a closed container in which MeJA was applied to cotton wicks (white bars). As a control (black bars), an equivalent volume of ethanol was applied to wicks in a box containing a separate set of plants. Treated plants were incubated an additional 24 h in the absence of MeJA before mite challenge. All plants were challenged with 20 adult female mites (10 mites per leaf). Ten days after challenge, leaf damage (A) and egg counts (B) were determined for each of the four treatment groups. Values indicate the mean and sd of each treatment group (n = 8). PI-II measurements showed that the MeJA treatment was equally effective in both genotypes (data not shown). Asterisks denote a significant difference (P < 0.05, Student's t test) in leaf damage or egg count in comparisons between control and MeJA-treated plants of the same genotype. To provide additional evidence for a role of the octadecanoid pathway in resistance to two-spotted spider mite, we examined the performance of mites on a transgenic line of tomato that overexpresses prosystemin from the cauliflower mosaic virus 35S promoter. Previous studies showed that plants expressing this transgene (called 35S::prosys) constitutively express PI and other defensive genes in the absence of wounding and that 35S::prosys-mediated signaling requires octadecanoid biosynthesis and perception . We observed that 35S::prosys plants were much more resistant to mite damage than either def-1 or WT plants (Fig. A). A substantial reduction in mite fecundity on the transgenic line was also evident (Fig. B). Given the significance of these effects, it was of interest to examine the interaction of various tomato genotoypes with another cell-content feeding herbivore, western flower thrips (Frankliniella occidentalis). As was the case for spider mites, thrips feeding resulted in the accumulation of high levels of PI-II in WT but not def-1 plants (Fig. A). The high constitutive levels of PI-II in undamaged 35S::prosys plants were further increased in response to thrips damage, similar to the previously reported effects of mechanical wounding on these plants . Thrips larvae inflicted a comparable amount of damage to WT and def-1 plants during the feeding trial (Fig. B). By contrast, 35S::prosys plants were highly resistant to damage. These findings indicate that thrips feeding induces octadecanoid-mediated host responses and that constitutive activation of the signaling pathway by overexpression of prosystemin enhances resistance to multiple cell-content feeding herbivores. Figure 5 | Activation of the octadecanoid pathway by 35S::prosystemin confers enhanced resistance to spider mites. Activation of the octadecanoid pathway by 35S::prosystemin confers enhanced resistance to spider mites. Fifteen-day-old wild-type (WT), def-1, and 35S::prosys (Ps) plants were challenged with 10 adult female mites on the terminal leaflet of each of two expanded leaves. Eight days after challenge, leaf damage (A) and egg counts (B) were determined for each treatment group. Data represent the mean and sd of 12 plants for each genotype. Different lowercase letters denote a significant difference at P < 0.01. Figure 6 | Octadecanoid signaling mediates defense responses of tomato to thrips. Octadecanoid signaling mediates defense responses of tomato to thrips. Fifteen-day-old wild-type (WT), def-1, and 35S::prosys (Ps) plants were challenged with thrips larvae on the terminal leaflet of each of two expanded leaves (five larvae per leaf). A, Five days after challenge, PI-II protein levels were measured in leaves of infested (white bars) and control untreated (black bars) plants. B, Leaf damage to each host genotype was determined 5 d after challenge. Different lowercase letters denote a significant difference at P < 0.005. Data represent the mean and sd of 14 plants for each genotype. Mapping of the Def-1 Gene | The essential role of Def-1 in induced responses to herbivory prompted us to initiate mapping of this gene as a first step toward understanding its molecular function. The wound response phenotype of def-1 homozygotes can be scored reliably using an immunodiffusion assay to measure wound-induced accumulation of PI-II in two-leaf-stage plants . Phenotypic analysis of an F2 population (168 plants) produced from self-pollination of a Def-1/def-1 heterozygote showed that the proportion of wound-responsive (W+) to wound-nonresponsive (W-) progeny was 123:45, in good agreement with the ratio predicted for a single recessive mutation (chi2 = 0.26; P > 0.5). To generate an interspecific mapping population, L. esculentum (def-1/def-1) was crossed as a pistillate parent to the wild tomato species Lycopersicon pennellii (Def-1/Def-1). All resulting F1 hybrids (def-1/Def-1; n = 12) were W+, indicating that def-1 is recessive in the L. pennellii background. A segregating backcross (BC1) population was generated from a second cross between a F1 plant (staminate parent) and a def-1/def-1 homozygote. Analysis of 509 BC1 progeny showed that 144 plants were W-, whereas the remaining plants were W+. This ratio deviated significantly from the expected value of 1:1 (chi2 = 96; P < 0.001) and likely reflects reduced transmission of the def-1 allele through the pollen (C. Li and G.A. Howe, unpublished data). Bulk segregant analysis was used in combination with AFLP to identify markers that are linked to Def-1. Genomic DNA from 10 W+ and 10 W- BC1 plants was pooled to construct a W+ bulk (B+) and W- bulk (B-), respectively. Among 64 primer combinations used to screen the bulks for AFLPs, two combinations (E-ACA/M-CTG and E-AGC/M-CTC) generated a polymorphic band that was present in both the W+ parent (Def-1/def-1) and the B+, but absent in the W- parent (def-1/def-1) and the B- (Fig. A). DNA bands corresponding to the two polymorphisms, designated EM-1 and EM-2, were cloned into a plasmid vector. Genomic DNA hybridization experiments showed that EM-1 and EM-2 probes detected single- or low-copy sequences in the genome and easily scorable RFLPs in genomic DNA digested with HindIII and HaeIII, respectively (Fig. B, lanes 1 --3). Linkage of EM-1 and EM-2 to Def-1 was confirmed using the 10 W+ and 10 W- BC1 individuals that composed the two bulks. This experiment showed that all W+ plants were heterozygous for both markers, whereas all W- plants were homozygous for the L. esculentum RFLP pattern (data not shown). The absence of recombinants in this population of 20 BC1 plants demonstrates that EM-1 and EM-2 are linked to the Def-1 locus. Figure 7 | Identification and chromosomal localization of AFLP markers linked to the Def-1 gene. Identification and chromosomal localization of AFLP markers linked to the Def-1 gene. AFLP analysis for Def-1-linked markers was performed on bulked segregants from a BC1 segregating population. Four genomic template DNAs were used for each primer combination: P-, W- parent (L. esculentum, def-1/def-1); P+, W+ parent (L. pennellii, Def-1/Def-1); B- and B+, bulks composed of 10 W- and 10 W+ progeny, respectively, from a BC1 mapping population (see text for details). A, Arrows indicate the position of the EM-1 (top) and EM-2 (bottom) AFLP markers identified using two different primer combinations. Only the portion of the autoradiographed AFLP gel is shown. B, Conversion of EM-1 and -2 to RFLP markers and mapping of the markers to IL3-5. Cloned AFLP markers were labeled with 32P and hybridized to genomic DNA digested with HindIII (EM-1, top) or HaeIII (EM-2, bottom). Genomic DNA was obtained from plants with the following genotypes: lane 1, F1 hybrid (Def-1/def-1) between L. esculentum (def-1/def-1) and L. pennellii (LA716); lane 2, L. esculentum (def-1/def-1) parent; lane 3, L. pennellii (Def-1/Def-1) parent; and lane 4, IL (LA3490) harboring the IL3-5 segment of L. pennellii DNA. The chromosomal location of EM-1 and EM-2 was determined using a set of 50 introgression lines (ILs) harboring defined segments of L. pennellii DNA in an otherwise L. esculentum background . DNA from each IL was screened for the presence of the EM-1 and EM-2 RFLPs. One line, LA3490, displayed the L. pennellii RFLP pattern for both markers (Fig. B, lane 4). The introgressed region of DNA contained in LA3490 is located on the end of the long arm of chromosome 3 and is referred to here as IL3-5. This region of chromosome 3 includes the RFLP marker TG152 and all other markers (e.g. TG214 and TG244) distal to the centromere . Southern hybridization experiments using IL3-5-specific RFLP markers confirmed the identity of LA3490 (data not shown). To confirm and refine the position of Def-1 on IL3-5, RFLP markers TG152, TG214, and TG244 were tested for linkage to Def-1 using 305 plants (144 W- and 161 W+) from the above-mentioned BC1 mapping population. The results showed that Def-1 is linked to TG152, TG214, and TG244 at distances corresponding to approximately 17, 9, and 6 centimorgans, respectively . These findings are consistent with the established genetic map for chromosome 3 and position Def-1 distal to TG244. Figure 8 | Genetic map of the long arm of tomato chromosome 3. Genetic map of the long arm of tomato chromosome 3. The map is oriented with the centromere and telomere at the top and bottom of the figure, respectively. Recombination distances between markers and Def-1 are based on segregation analysis of a BC1 population of 305 individuals. Molecular markers are indicated on the right. Map distances (in centimorgans) indicated on the left were calculated as the proportion of individuals in the mapping population having a recombination event between the indicated markers. Markers EM-1 and LoxD co-segregated in all BC1 plants. The T-DNA insertion present in line ET570 is located approximately 2 centimorgans from Def-1, but its position relative to other markers was not unambiguously determined. The tomato line ET570 carries a T-DNA insertion approximately 3 centimorgans distal to TG244 . The T-DNA in ET570 harbors the maize (Zea mays) Dissociation (Ds) transposon, which was introduced by Agrobacterium tumefaciens-mediated transformation for the purpose of transposon-tagging experiments . We performed a test cross to determine the relative distance between Def-1 and the T-DNA insertion. ET570 (Def-1/Def-1 Ds/Ds) was crossed to a def-1/def-1 homozygote to produce an F1 plant (Def-1/def-1 +/Ds) that was subsequently backcrossed to a def-1/def-1 homozygote. The resulting progeny were analyzed for their wound response phenotype and the presence of Ds-containing T-DNA. Among 146 plants tested, three recombinants that were either W+Ds- or W-Ds+ were recovered. This finding confirms the location of Def-1 on the distal end of chromosome 3 and positions the gene approximately 2 centimorgans from the Ds-containing T-DNA harbored by ET570 . Previous studies indicated that the def-1 lesion affects a step in JA biosynthesis (; Fig. B). JA biosynthetic enzymes encoded by genes that map to IL3-5 would thus represent candidates for Def-1. Among the JA biosynthetic genes identified in tomato are those encoding two plastidic lipoxygenases (LoxC and LoxD; ), two plastidic allene oxide synthases (AOS1; ; AOS2; ), plastidic allene oxide cyclase (AOC; ), and 12-oxo-phytodienoic acid reductase (OPR3; ). The location of AOC on chromosome 2 indicates that this gene is not a candidate for Def-1. To test the hypothesis that LoxC, LoxD, AOS1, AOS2, or OPR3 corresponds to Def-1, these cDNAs were converted to RFLP markers and mapped using the ILs described above. The results showed that LoxD maps to IL3-5, whereas LoxC, AOS1, AOS2, and OPR3 map to different chromosomes (data not shown). To further test the genetic relationship between LoxD and Def-1, the position of LoxD was refined using the aforementioned BC1 mapping population. Among 305 meiotic events analyzed, 15 recombinants between Def-1 and LoxD were detected. Thus, the LoxD and Def-1 loci appear to be genetically distinct. DISCUSSION : In this study, we examined the role of the octadecanoid pathway in resistance of cultivated tomato to the arachnid herbivore two-spotted spider mite. The rapid life cycle, ease of rearing, broad host range, and economic importance of two-spotted spider mite make it well suited for the study of plant-herbivore interactions. The two-spotted spider mite has proven particularly valuable for studying herbivore-induced plant volatiles and their role in influencing tritrophic interactions . Increasing evidence indicates that JA and related signaling molecules play an important role in regulating volatile-mediated plant defenses against two-spotted spider mite . By contrast, relatively little is known about the role of jasmonates in regulating the synthesis of phytochemicals that have a direct effect on the two-spotted spider mite. The availability of isogenic lines of tomato that are either down-regulated (i.e. def-1) or up-regulated (i.e. 35S::prosys) in the octadecanoid pathway provide valuable tools to address this question. Several lines of evidence indicate that induced defense of tomato against two-spotted spider mite is regulated by the octadecanoid pathway. First, infestation of WT plants with spider mites induced the expression of several JA-responsive, defense-related genes (i.e. PI-I, PI-II, CDI). Second, def-1 plants were deficient in PI expression in response to two-spotted spider mite feeding. This phenotype was tightly correlated with both increased susceptibility of the mutant to mite damage and increased mite fecundity. Third, spider mite feeding was accompanied by increased JA accumulation in WT but not def-1 plants. The increase in JA levels in mite-infested WT plants was notably less than that observed in mechanically wounded tomato leaves (e.g. ; ). This may reflect differences in the type of damage caused by mite feeding (i.e. piercing/sucking) and mechanical wounding (i.e. leaf crushing). Fourth, pretreatment of def-1 plants with exogenous MeJA before challenge resulted in a significant decrease in two-spotted spider mite performance. This finding is consistent with recent reports that exogenous jasmonate promotes host plant resistance to spider mites . We also observed that activation of the octadecanoid signaling pathway by overexpression of prosystemin significantly reduced the performance of both spider mites and thrips. To our knowledge, this finding represents the first report of genetically engineered resistance to cell-content feeding herbivores. Transgene-mediated activation of octadecanoid signaling may have important implications for the generation of broad-spectrum pest resistance in agricultural crop plants. It will be interesting to determine whether 35S::prosys plants are resistant to other classes of herbivores, such as phloem-feeding insects that induce both JA-dependent and -independent defense responses . Our results support the idea that the octadecanoid pathway regulates the synthesis of one or more foliar compounds that have a negative effect on cell-content feeders. Additional studies are needed to identify these compounds and to determine whether the mechanism of resistance involves antibiosis (i.e. toxicity), antixenosis (non-preference), or other factors. As recently noted by , reduced egg production by spider mites grown on JA-treated tomato plants suggests a nutritional rather than toxic mechanism of resistance. It is possible that JA-regulated proteins such as PIs and polyphenol oxidases implicated as anti-feedants against Lepidopteran insects are also effective against the two-spotted spider mite. As an alternative, the observed effects may be attributed to secondary metabolites whose biosynthesis is regulated by JA . Resistance of wild tomato species to two-spotted spider mite is associated with defensive phytochemicals (e.g. methyl ketones, sesquiterpenenes, and acyl sugars) that kill, repel, or entrap the herbivore . However, these compounds are typically found in secretions of glandular trichomes and generally do not accumulate to high levels in cultivated tomato. We determined the chromosomal location of Def-1 as a first step toward molecular characterization of the gene. The mapping studies relied on the ability to score wound-inducible PI-II accumulation in a segregating population generated from an interspecific cross between def-1 and L. pennellii and involved three basic steps. First, bulk segregant analysis was used to identify AFLP markers linked to Def-1. Second, ILs harboring defined segments of L. pennellii DNA were used to map AFLP markers to a specific region on chromosome 3. Finally, the map position was refined using RFLP markers on the tomato linkage map. The location of Def-1 was confirmed in experiments showing linkage of the gene to a T-DNA insertion previously mapped to the distal end of chromosome 3 . To our knowledge, genes affecting resistance to herbivores have not been mapped previously to this region of the tomato genome. The inability of def-1 plants to accumulate normal levels of JA in response to herbivory (Fig. B), mechanical wounding, and systemin suggests that Def-1 may encode an enzyme involved in JA biosynthesis. We tested this hypothesis by determining whether genes encoding known or putative JA biosynthetic enzymes co-map with Def-1. The finding that LoxC, LoxD, AOS1, AOS2, and OPR3 do not co-map with Def-1 would appear to exclude them as candidate genes. Mapping of a tomato AOC-encoding cDNA to chromosome 2 likewise excludes it as a candidate gene. Given the existence of several isoforms for JA biosynthetic enzymes in tomato, it should be emphasized that these findings do not rule out the possibility that Def-1 corresponds to a JA biosynthetic gene that has not yet been cloned. It is also possible that Def-1 functions indirectly to regulate the activity of a JA biosynthetic enzyme or the transport or stability of an octadecanoid intermediate, but virtually nothing is known about these processes in tomato or other plants. Map-based cloning methods have been successfully used to isolate tomato genes whose biochemical function was not known . However, we have found that the telomeric location of Def-1 hinders the identification of tightly linked markers that flank the target gene (C. Li and G.A. Howe, unpublished data), indicating that this may not be the optimal approach for isolation of Def-1. Transposon tagging may provide a useful alternative. The maize Ac/Ds transposon system has been introduced into the tomato genome by Agrobacterium tumefaciens-mediated transformation, and numerous transgenic lines have been developed in which Ds elements are integrated at defined locations throughout the genome . The tendency of Ds to transpose from its donor site to linked acceptor sites has facilitated targeted tagging of genes whose position is known (e.g. ). The Ds element in line ET570 was previously used for the nontargeted tagging and isolation of the feebly gene, which is located approximately 9 centimorgans from the Ds insertion site . The close proximity of Def-1 to this same Ds element suggests that a targeted transposon tagging experiment may be useful for identification of Def-1. MATERIALS AND METHODS : Plant Material and Herbivore Rearing | Tomato (Lycopersicon esculentum Mill cv Castlemart) seedlings were grown in Jiffy peat pots (Hummert International, Earth City, MO) in a growth chamber maintained under 17 h of light (200 muE m-2 s-1) at 28C and 7 h of dark at 18C. Seed for def-1 was collected from a def-1/def-1 homozygous line that was backcrossed four times using tomato cv Castlemart as the recurrent parent. Seed for the 35S::prosystemin transgenic plants was collected from a 35S::prosys/35S::prosys homozygous line that was backcrossed five times using tomato cv Castlemart as the recurrent parent. Seed for L. pennellii (LA716) and the set of ILs was obtained from the Tomato Genetics Resource Center (University of California at Davis). Two-spotted spider mite (Tetranychus urticae Koch) was obtained from a colony maintained in the Pesticide Research Center greenhouses at Michigan State University. Mites were reared on lima bean (Phaseolus lunatus cv Fordhook) plants grown in course vermiculite and maintained under 18 h of light per day. Bean plants were typically 2 to 5 weeks of age, and contained an average of 50 adult spider mites per leaf. Western flower thrips (Frankliniella occidentalis) were reared on marigolds (cv Golden Boy) in the Pesticide Research Center greenhouses at Michigan State University. Plant Treatments | Adult female spider mites were transferred, using a small soft-bristled paintbrush, to the adaxial surface of the terminal leaflets of 14- to 15-d-old tomato plants. Plants of this age contained two fully expanded leaves and an emerging third leaf. Care was taken to avoid wounding of plants during the transfer procedure. Feeding by adult female mites at a single site resulted in the appearance of a chlorotic lesion, the average size of which was estimated to be 0.25 mm2. The extent of leaf damage resulting from feeding was estimated by counting the number of lesions with the aid of a dissecting microscope. In cases where the area of a single site of damage exceeded 0.25 mm2, the total area of the damaged site was estimated as a multiple of 0.25 mm2. A dissecting microscope was used to count eggs on both the upper and lower leaf surface. Adult and larval stages of thrips were obtained by gently shaking infested marigold flowers onto white paper. Larvae were used for all experiments because, unlike adults, they are fairly immobile. Larvae were transferred individually to tomato leaves as described above. Infested plants were confined individually to Magenta boxes covered with thrips-proof gauze. Leaf damage was estimated by creating a pictorial grid representation of each leaf. During examination of damaged leaves under a dissecting microscope, areas representing damage were marked on the pictorial grid. For experiments involving MeJA treatment, plants were transferred to a lucite container (8 L) and exposed to MeJA composed of approximately 20% (v/v) (+-)-7-iso-MeJA (product no. 399E, Bedoukian Research, Danbury, CT). All experiments were performed using 1 muL of MeJA, diluted into 100 muL of ethanol, applied to several cotton wicks distributed evenly throughout the box. Twenty-four hours after MeJA treatment, the cotton wicks were removed, and plants were acclimated to ambient humidity for an additional 24 h before mite infestation. Control plants were incubated in a separate container in which ethanol was applied to cotton wicks. Measurement of JA | Spider mite-infested lima bean leaves were cut into slices and placed (abaxial side-down) onto the upper surface of leaves of 16-d-old WT and def-1 plants. Two days later, tomato leaflets (5 g fresh weight) showing comparable levels of damage were harvested and frozen in liquid nitrogen. JA was extracted and quantified using gas chromatography-mass spectrometry as previously described . Nucleic Acid Gel-Blot Analysis | Total RNA was isolated and analyzed by blot hybridization as previously described , except that Hybond-N Plus membranes (Amersham Biosciences, Sunnyvale, CA) were used in place of nitrocellulose. Gels were run in duplicate, with one set stained with ethidium bromide to check for equal loading of the samples and intactness of the RNA. DNA probes were obtained and labeled as described by . Hybridization signals were visualized by autoradiography using XAR-5 film (Eastman Kodak, Rochester, NY) and, when appropriate, quantified using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA). Tomato expressed sequence tag (EST) clone cLEC9C14 was used as a probe for AOS1 . Hybridization signals were normalized to the hybridization signal obtained using a cDNA probe (EST clone cLED1D24) for translation initiation factor eIF4A. Tomato DNA was isolated from fresh leaves as described by . Southern-blot hybridization experiments were performed according to . DNA probes were prepared using a T7 Quickprime Kit (Pharmacia Biotech, Piscataway, NJ). Mapping Experiments | The wound response of plants from segregating populations was scored as follows. Sixteen-day-old plants were wounded with a hemostat across the midvein of all leaflets of the basipetal leaf. This procedure was repeated 3 h later on the same leaflets at a position proximal to the petiole. Wounded plants were incubated under standard growth conditions for 48 h, after which the level of PI-II in the wounded (local) and upper unwounded leaf (systemic) was determined using a radial immunodiffusion assay . BC1 progeny showing no systemic response and less than 10% of a wild-type local response were classified as wound-nonresponsive (W-) def-1/def-1 homozygotes. Progeny in which the local and systemic response was comparable with WT were classified as wound-responsive (W+) Def-1/def-1 heterozygotes. AFLP was performed using an AFLP kit from Invitrogen (Carlsbad, CA). AFLP products were labeled with [gamma-33P]ATP, separated on a 6% (w/v) polyacrylamide gel, and visualized by autoradiography. Bands corresponding to putative AFLPs were excised from the gel by marking the position of the polymorphic band on the dried polyacrylamide gel after alignment with the autoradiogram. Excised DNA fragments were eluted in 400 muL of high-salt buffer (20% [v/v] ethanol, 1 m LiCl, and 10 mm Tris-HCl, pH 7.5) for 2 h at 65C and then precipitated with ethanol. One quarter of the resuspended DNA was re-amplified with the same primer combination that was used for the selective amplification. Re-amplified PCR products were cloned into the pGEM-T easy vector (Promega, Madison, WI) according to the manufacturer's instructions. Cloned DNA fragments corresponding to EM-1 and EM-2 were 183 and 249 bp in length, respectively. To convert these markers to RFLP markers, survey blots containing restriction enzyme-digested (DraI, EcoRI, EcoRV, HaeIII, HindIII, and XbaI) genomic DNA from parental lines was probed with the radiolabeled AFLP fragments. The relative position of EM-1 and EM-2 on chromosome 3 was determined by mapping the markers using the subset of BC1 plants containing recombination events between TG244 and Def-1. A probe for detection of the Ds-containing T-DNA in line ET570 was obtained by PCR amplification with primers act5b and 35s, as previously described . Backmatter: PMID- 12226529 TI - Regulation of Transcript Levels of the Arabidopsis Cytochrome P450 Genes Involved in Brassinosteroid Biosynthesis AB - Cytochrome P450 enzymes of the closely related CYP90 and CYP85 families catalyze essential oxidative reactions in the biosynthesis of brassinosteroid (BR) hormones. Arabidopsis CYP90B1/DWF4 and CYP90A1/CPD are responsible for respective C-22 and C-23 hydroxylation of the steroid side chain and CYP85A1 catalyzes C-6 oxidation of 6-deoxo intermediates, whereas the functions of CYP90C1/ROT3, CYP90D1, and CYP85A2 are still unknown. Semiquantitative reverse transcriptase-polymerase chain reaction analyses show that transcript levels of CYP85 and CYP90 genes are down-regulated by brassinolide, the end product of the BR biosynthesis pathway. Feedback control of the CYP90C1, CYP90D1, and CYP85A2 genes by brassinolide suggests that the corresponding enzymes might also participate in BR synthesis. CYP85 and CYP90 mRNAs show strong and transient accumulation during the 1st week of seedling development, as well as characteristic organ-specific distribution. Transcripts of CYP90A1 and CYP85A2 are preferentially represented in shoots and CYP90C1, CYP90D1, and CYP85A1 mRNAs are more abundant in roots, whereas CYP90B1 is ubiquitously expressed. Remarkably, the spatial pattern of CYP90A1 expression is maintained in the BR-insensitive cbb2 mutant, indicating the independence of organ-specific and BR-dependent regulation. Quantitative gas chromatography-mass spectrometry analysis of endogenous BRs in shoots and roots of Arabidopsis, pea (Pisum sativum), and tomato (Lycopersicon esculentum) reveal similar partitioning patterns of BR intermediates in these species. Inverse correlation between CYP90A1/CPD transcript levels and the amounts of the CYP90A1 substrate 6-deoxocathasterone in shoots and roots suggests that transcriptional regulation plays an important role in controlling BR biosynthesis. Keywords: Introduction : Brassinosteroids (BRs) are plant steroid hormones that influence a wide range of important developmental processes, including germination, cell elongation, differentiation of vascular elements, photomorphogenesis, and pollen fertility . The maintenance of optimal local BR concentrations by coordinated biosynthetic and inactivation mechanisms, in combination with the differential responsiveness of target cells and tissues, enables the proper regulation of these physiological functions during plant development. The pathways of BR biosynthesis have been elucidated by a series of detailed biochemical studies. Brassinolide (BL), the biologically most active BR, is synthesized from campesterol via either early or late C-6 oxidation routes . The BR biosynthesis pathways are conserved between Catharanthus roseus, Arabidopsis, pea (Pisum sativum), tomato (Lycopersicon esculentum), and rice (Oryza sativa), although a limitation of early C-6 oxidation has been observed in some of these species . Figure 1 | The pathway of BR biosynthesis. The pathway of BR biosynthesis. Black arrows represent conversion steps with confirmed or assumed involvement of cytochrome P450 monooxygenases. Identified Arabidopsis P450 enzymes of the pathway are indicated. Numbering of the carbon positions oxidized in BRs is given at the structural formula of campesterol. Molecular genetic analysis of BR-deficient mutants has identified several BR biosynthesis genes that, with the exception of DET2 (DEETIOLATED2; ) were all found to encode cytochrome P450 monooxygenases (for review, see ). Arabidopsis DWF4 (DWARF4), a protein classified as CYP90B1 according to the international cytochrome P450 nomenclature , was shown by BR intermediate feeding to catalyze C-22 hydroxylation of the steroid side chain . Likewise, rescue of the Arabidopsis cpd (constitutive photomorphogenesis and dwarfism) mutant revealed that CPD/CYP90A1, another member of the CYP90 family, functions as C-23 steroid hydroxylase . Rescue of the dwarf phenotypes of cpd and dwf4 mutants by BR intermediates indicates that CYP90A1 and CYP90B1 are responsible for the C-23 and C-22 side chain hydroxylation reactions in both the early and late C-6 oxidation pathways of BR biosynthesis. Mutation of the Arabidopsis ROT3 (ROTUNDIFOLIA3) gene, encoding CYP90C1, results in defective cell elongation and reduced leaf expansion. Because of the apparent lack of phenotypic rescue with externally supplied BRs, the role of CYP90C1 in BR biosynthesis remained unclear . Similarly, due to the lack of mutants, no function has been assigned for CYP90D1, the fourth Arabidopsis gene of the CYP90 family. C-6 oxidation of BR intermediates is catalyzed by an enzyme of the CYP85 family , as was demonstrated in vitro with yeast (Saccharomyces cerevisiae)-expressed CYP85A1 of both tomato (DWARF) and Arabidopsis . In these assays, CYP85A1 oxidized only the late biosynthetic intermediates 6-deoxotea-sterone, 3-dehydro-6-deoxoteasterone, 6-deoxotypha-sterol, and 6-deoxocastasterone, but did not catalyze the C-6 oxidation of campestanol, a substrate of CYP90B1 . The function of CYP85A2, the second member of the CYP85 family in Arabidopsis, is so far unclear. Two further oxidative reactions, namely C-2 hydroxylation and the formation of BL by Bayer-Villiger lactonization of the steroid B ring, are also thought to be catalyzed by yet unidentified cytochrome P450 enzymes . Recently, have detected steroid C-2 hydroxylase activity of DDWF1 (dark-induced DWF-like protein 1), a pea P450 designated CYP92A6. Because the Arabidopsis genome does not encode any member of the CYP92 family, in this plant, the C-2 hydroxylation reaction is probably performed by a different cytochrome P450 enzyme. The regulatory mechanisms of BR homeostasis are little understood. , ) observed the accumulation of BL and its precursors, as well as up-regulation of the DWF4 and CPD transcripts, in the BR-insensitive bri1 mutant of Arabidopsis, suggesting a role for BRI1 in the regulation of BR biosynthesis. Furthermore, BL treatment of Arabidopsis seedlings markedly decreased the steady-state level of CPD mRNA, and this transcriptional response was shown to require de novo protein synthesis . These results suggest that BR synthesis is controlled by an elaborate feedback regulation, one that shows analogy to the negative control of GA biosynthesis genes by GAs . The cellular concentration of active BRs is also influenced by the catabolism of BL and/or its precursors. In feeding experiments, the activation-tagged Arabidopsis bas1-D mutant overexpressing BAS1/CYP72B1 was found to accumulate biologically inactive 26-hydroxybrassinolide . The dwarf phenotypes of bas1-D and chibi2, another activation-tagged Arabidopsis mutant with high CYP72C1 level , are very similar to those of the BR-deficient mutants. As compared with the wild type, the BR-deficient and -insensitive Arabidopsis mutants contain diminished BAS1 transcript levels, indicating that BL may induce expression of the corresponding catabolic enzyme . So far, only limited information is available about the temporal and spatial control of the genes responsible for BR biosynthesis. Strong CPD expression was detected during the 1st week of seedling development, and in transgenic plants, a CPD promoter-driven GUS reporter fusion showed activity in cotyledons, leaves, and floral organs, but not in roots . Similar GUS histochemical (G.J. Bishop, unpublished data) and in situ hybridization assays revealed that tomato DWARF promoter activity is localized mainly in the apical and root meristem regions, whereas ROT3 is expressed in all organs and cell types of Arabidopsis seedlings . Thus, further studies are required to elucidate how and to what extent differential expression of particular BR biosynthesis genes affects active hormone levels and intermediate partitioning during plant development. In this paper, we report that genes of the closely related CYP85 and CYP90 cytochrome P450 families implicated in BR biosynthesis are coordinately regulated by BL. Feedback control of the genes encoding ROT3/CYP90C1 and CYP90D1 suggests that these enzymes may also be involved in BR synthesis. Although all CYP85 and CYP90 genes are strongly expressed during the 1st week of seedling development, their transcripts have characteristically different accumulation patterns in the shoots and roots of seedlings and fully developed plants. The expression level of the CPD/CYP90A1 gene shows correlation with the spatial partitioning of the CYP90A1 substrate 6-deoxocatha-sterone, suggesting that transcriptional control of the CYP85 and CYP90 genes can contribute to the regulation of BR biosynthesis. RESULTS : Cytochrome P450 Monooxygenases in BR Biosynthesis Are Evolutionarily Related | All cytochrome P450 enzymes of Arabidopsis with known function in BR biosynthesis belong to either the CYP85 or CYP90 families. Protein sequence comparison based on BLAST homology analysis revealed that these two P450 families are closely related, sharing approximately 35% amino acid sequence identity. We found that the CYP85 and CYP90 proteins also show high levels (about 30%) of sequence identity with ent-kaurenoic acid oxidases, members of the CYP88 P450 family involved in GA biosynthesis . In contrast, the two Arabidopsis CYP72 hydroxylases responsible for BR inactivation are only distantly related to the P450s of BR biosynthesis, featuring less than 20% sequence identity with any member of the CYP90 and 85 families. A phylogenetic tree generated by the ClustalW multiple alignment program shows the close relationship between BR-biosynthetic and CYP88 P450s, as compared with CYP72 proteins, in Arabidopsis (Fig. A). Figure 2 | Structural relationship between selected Arabidopsis cytochrome P450 proteins and their genes. Structural relationship between selected Arabidopsis cytochrome P450 proteins and their genes. A, Unrooted cladogram based on the primary structure of P450 families involved in BR biosynthesis (CYP85 and CYP90), BR catabolism (CYP72), and GA biosynthesis (CYP88). Amino acid identity values, as compared with CPD/CYP90A1, are given in brackets. B, Exon/intron structure of the genes encoding CYP85A1 , CYP85A2 , CPD/CYP90A1 , DWF4/CYP90B1 , ROT3/CYP90C1 , CYP90D1 , CYP88A3 , CYP88A4 , BAS1/CYP72B1 , and CHIBI2/CYP72C1 . Exon sizes are given in bp. The analysis of exon-intron organization of the same P450 genes uncovered similar relationships (Fig. B). Each intron of the CYP85, CYP90 and CYP88 genes was found at one of eight conserved positions, whereas CYP72B1 and CYP72C1 showed a different exon-intron pattern. The close relationship indicated by both protein and gene structure analyses, together with the similarity of enzymatic functions, suggest that during their evolution, the CYP85 and CYP90 families diverged after their specialization to steroid substrates. Coordinated Feedback Regulation of CYP85 and CYP90 Genes | It was demonstrated previously that transcription of the CPD gene is down-regulated by BL, the end product of BR biosynthesis . Therefore, we were interested in determining whether the transcript levels of other CYP90 or CYP85 transcripts are similarly regulated by this phytohormone. Because of the low abundance of these P450 mRNAs, in these experiments, the steady-state transcript levels were monitored by more sensitive semiquantitative RT-PCR, rather than northern hybridization. BL treatment reduced the amount of CPD/CYP90A1, DWF4/CYP90B1, ROT3/CYP90C1, CYP90D1, CYP85A1, and CYP85A2 transcripts to approximately 10% or less of the level detected in untreated control seedlings (Fig. A). These data show that in Arabidopsis all CYP85 and CYP90 gene activities are controlled by BR-dependent feedback regulation. Figure 3 | Effect of BL on the steady-state mRNA levels of BR-biosynthetic P450s. Effect of BL on the steady-state mRNA levels of BR-biosynthetic P450s. Reverse transcriptase (RT)-PCR products obtained from total RNA of 7-d-old seedlings incubated for 4 h in the presence (BL) or absence (Ctr) of 100 nm BL. A, Wild type; B, BR-deficient cpd and cbb3 mutants; C, BR-insensitive cbb2 mutant. UBQ10 was used as internal control. To confirm our results, we also analyzed the relative amounts of CYP85 and CYP90 transcripts in BL-treated and untreated Arabidopsis mutants impaired in BR biosynthesis or perception. Compared with wild-type plants, the mRNA levels of CYP85 and all four CYP90 genes were significantly higher, indicating derepressed expression, in the BR-deficient cpd and cbb3 mutants (cbb3 being allelic to cpd; ). The amounts of CYP85 and CYP90 transcripts were reduced in these mutants upon external application of BL, but remained somewhat higher than in BL-treated wild-type plants (Fig. B). In the BR-insensitive cbb2 mutant , however, BL had no effect on the expression of CPD (Fig. C) or any other BR-responsive CYP85 or CYP90 genes (data not shown). This result indicates that BR-mediated feedback regulation of the CYP85 and CYP90 genes is dependent on the function of the BRI1 Leu-rich repeat receptor kinase , which has been inactivated in the cbb2 mutant. Regulation of CYP85 and CYP90 mRNA Levels during Germination and Seedling Development | To gain better insight into the regulation of BR-biosynthetic P450 genes during the early stages of plant development, we determined the relative amounts of CYP85 and CYP90 transcripts by RT-PCR in seedlings and young plants throughout the first 8 d after imbibition and after 2 weeks of development . At the earliest, transcripts of the CYP85A2 and ROT3 genes were already detectable from the 1st d of germination. Each CYP85 and CYP90 mRNA reached a peak level of abundance during the 1st week of seedling development but, with the exception of CPD, their levels declined to about 10% or less of the maximum values by the end of this period. Subsequently, between d 8 and 14 of the time course, only little or no change was detectable in the transcript levels. Although individual CYP85 and CYP90 genes featured different temporal expression profiles, their transient induction during the 1st week after germination suggests the requirement of BR biosynthesis enzymes during the early stages of seedling development. Figure 4 | Changes in transcript levels of CYP85 and CYP90 genes during germination and seedling development. Changes in transcript levels of CYP85 and CYP90 genes during germination and seedling development. RT-PCR products prepared from total RNA of developing wild-type seedlings and young plants (1 through 8 and 14 d after imbibition). Quantitative data are plotted as percentage of the highest value measured during the experimental period. Differential Regulation of CYP85 and CYP90 Transcript Levels in Shoots and Roots | Previously, we reported that CPD expression is much stronger in the aerial parts than the roots of Arabidopsis . To obtain information on the organ-specific regulation of CYP85 and CYP90 genes, their transcript levels were compared by RT-PCR in roots and shoots (representing combined cotyledon and hypocotyl tissues) of 7-d-old seedlings. CPD and CYP85A2 were found preferentially expressed in cotyledons and hypocotyls, whereas the expression of CYP85A1, ROT3, and CYP90D1 was stronger in roots (Fig. A). Intriguingly, the highly homologous CYP85A1 and CYP85A2 genes exhibited different spatial expression patterns, whereas the closely related ROT3 and CYP90D1 displayed similar ones. We also used RT-PCR assays to determine the expression of the DET2 and DIM1 genes that encode non-P450-type enzymes acting upstream of CYP85 and CYP90 monooxygenases in BR synthesis. In contrast to most CYP85 and CYP90 messages, the DIM1 and DET2 transcripts were equally abundant in the shoots and roots of the seedlings. Figure 5 | Differential accumulation of BR-biosynthetic P450 mRNAs in shoots and roots. Differential accumulation of BR-biosynthetic P450 mRNAs in shoots and roots. A, Transcript levels in shoots (S) and roots (R) of wild-type seedlings. B, CPD/CYP90A1 transcript levels in shoots (S) and roots (R) of BR-insensitive cbb2 seedlings. C, CPD/CYP90A1 transcript levels in roots (R) of wild-type seedlings incubated for 4 h in the presence (BL) or absence (Ctr) of 100 nm BL. RT-PCR products were obtained from total RNA of 7-d-old seedlings. UBQ10 was used as internal control. The shoot to root ratios of mRNA levels in 7- and 20-d-old Arabidopsis plants are shown in Table . These values indicate a preferential accumulation of CYP85A2 and CPD transcripts in the shoots of 7-d-old seedlings. The data obtained from 20-d-old plants reflect the same distribution pattern as those of the young seedlings but, probably due to the lower activity of CYP85 and CYP90 genes in older plants, less pronounced organ-specific differences in CYP85A1, CYP85A2, and CPD expression were detected. Table I | Shoot to root ratio of transcript levels of genes involved in BR biosynthesis To clarify whether BR regulation is required for organ-specific accumulation, we also assayed levels of the CPD transcript in shoots and roots of BR-insensitive cbb2 seedlings. As it is shown in Figure B, the difference between the amounts of the CPD transcript in shoots and roots was found similar to that observed in the wild type (Fig. A). We also found that the low CPD transcript level in the roots of wild-type seedlings could be further decreased by BL treatment (Fig. C). These data indicate that the organ-specific control of CPD expression acts independently from the hormonal feedback regulation. Endogenous BR Levels in Shoots and Roots of Arabidopsis, Pea, and Tomato | Differential expression of CYP85 and CYP90 genes in the roots and aerial parts of the plant might influence BR biosynthesis and BR levels. To see if this was the case, we determined the amounts of endogenous BRs in the shoots and roots of Arabidopsis using quantitative gas chromatography (GC)-mass spectrometry analysis. In addition, similar analyses were performed on roots and shoots of pea and tomato to ascertain the conservation of organ-specific BR distribution. The data revealed a differential accumulation of BR biosynthesis intermediates in the aerial and underground parts of these plants . Despite the varying levels of particular BR forms in the three species, the pattern of their organ-specific partitioning was found to be very similar. The early intermediates 6-deoxocathasterone, 6-deoxoteasterone, 3-dehydro-6-deoxoteasterone, and 6-deoxotyphasterol were preferentially represented in the roots, whereas 6-deoxocastasterone and castasterone, synthesized later in the pathway, were more abundant in the shoots. The intermediates of the early C-6 oxidation pathway were at or below the detection level in all samples, whereas BL could only be observed in the Arabidopsis samples and the roots of pea. In Arabidopsis, the amount of 6-deoxocathasterone was more than 2-fold higher in the roots, which were shown to contain low CPD transcript level (Table ; Fig. A). Because CPD/CYP90A1 catalyzes the conversion of 6-deoxocathasterone to 6-deoxoteasterone, the accumulation of its substrate in the root indicates a low conversion rate in this organ and, hence, a good correlation between transcript abundance and the actual enzyme activity. Table II | Endogenous BR levels in shoots and roots of Arabidopsis, pea, and tomato DISCUSSION : Ensuring optimal physiological levels of active BRs requires a sensitive regulation of their biosynthesis. A recent comparative analysis of endogenous BR levels in Arabidopsis, pea, and tomato suggested similar control mechanisms in these species and indicated that C-22 and C-23 hydroxylation and C-6 oxidation are likely rate-limiting reactions of the pathway . All of these oxidative steps are catalyzed by cytochrome P450-type enzymes of the CYP90 and CYP85 families; therefore, the levels of these enzymes are expected to influence the efficiency of BR synthesis. Previously, we reported that the activity of Arabidopsis CPD, encoding the C-23-hydroxylase CYP90A1, is negatively controlled by BRs at the transcriptional level . Our present data show that DWF4, coding for the C-22 steroid side chain hydroxylase CYP90B1 acting immediately upstream of CPD, is also down-regulated by BL. In the pathway of GA synthesis, similar transcriptional feedback mechanisms have been identified and shown to modulate the expression of GA 20-oxidase and GA 3beta-hydroxylase genes . We found that in Arabidopsis, in addition to CPD and DWF4, the remaining CYP90 and CYP85 genes are also subject to feedback regulation by BL. Considering the highly similar primary structure of CYP85 and CYP90 proteins and that all of them with identified enzymatic functions participate in BR biosynthesis , the BR-repressible expression of the CYP85A2, ROT3, and CYP90D1 genes strongly suggests a role for their P450 products in BR biosynthesis. CYP85A2, sharing 82% amino acid sequence identity with CYP85A1, may represent a second Arabidopsis enzyme with steroid C-6 oxidase activity. At least a partial redundancy of this function in Arabidopsis is suggested by the lack of BR-deficient dwarf mutants defective in CYP85A1. Because P450 monooxygenases of the BR pathway are known to accept multiple substrates , these two enzymes may also differ in their substrate preferences. In vivo feeding experiments using radiolabeled precursors revealed an early and a late C-6 oxidation of campestanol and 6-deoxocasta-sterone, respectively , whereas less efficient conversion of 6-deoxotyphasterol to typhasterol was also demonstrated . In yeast expression systems, CYP85A1 of both tomato (DWARF) and Arabidopsis was shown to oxidize 6-deoxocastasterone. The Arabidopsis enzyme also utilized the upstream intermediates 6-deoxoteas-terone, 3-dehydro-6-deoxoteasterone, and 6-deoxoty-phasterol, but not campestanol and 6-deoxocathaste-rone . If the lack of campestanol conversion was not due to its limited uptake by the yeast cells, this early C-6 oxidation would require the action of another enzyme for which CYP85A2 is a likely candidate. The possible role(s) of the closely related (53% amino acid identity) ROT3/CYP90C1 and CYP90D1 proteins in BR synthesis is unclear. The rot3 mutant phenotype is much weaker than those of the other BR biosynthesis mutants, which may indicate that CYP90C1 has overlapping function with another enzyme, possibly CYP90D1. In Arabidopsis, there are two potentially P450-mediated reactions in BR synthesis for which the genes have not yet been identified, namely the C-2 hydroxylation reaction and the Bayer-Villiger lactonization step converting castasterone to BL . Therefore, it seems conceivable that CYP90C1 and/or CYP90D1 might participate in one of these enzymatic reactions. Transcript levels of the CYP85 and CYP90 genes were found to change in a wide range, from about 10% of the wild-type amount in BL-treated plants to 5 times the wild-type value in BR-deficient mutants. Thus, under normal developmental conditions, BR biosynthesis can be efficiently controlled through feedback regulation of these genes because their expression is partially repressed at physiological BR concentrations. The similarity of BR response suggests that the activity of CYP85 and CYP90 genes might be controlled by the same transcriptional regulators that modulate CPD expression . The BR response of CPD, and probably all other feedback-controlled CYP85 and CYP90 genes, requires an intact BR perception mechanism. Down-regulation of CPD was abolished in mutants deficient in the BRI1 BR receptor function, just as in bin2, another BR-insensitive mutant . Differential organ-specific expression of the CYP85 and CYP90 genes may provide another means of controlling BR biosynthesis. This regulation appears to be independent of BR action because: (a) BR insensitivity does not interfere with shoot-specific accumulation of the CPD transcript, (b) low root levels of this mRNA further decrease upon BL treatment, and (c) CYP85 and CYP90 genes, displaying similar steroid responses, show different preferences for shoot- and root-specific expression. Transcripts of CYP85A1 were detected primarily in the roots, whereas those of CYP85A2 accumulated preferentially in the shoots. The differential organ specificity can be seen as further indication for the different functions of the two Arabidopsis CYP85 enzymes. In a recent microarray-based transcript analysis of 142 Arabidopsis cytochrome P450 genes, have shown that both CYP90A1 and CYP85A1 are preferentially expressed in the aerial portion of 30-d-old plants. In the case of the CYP85A1 transcript, the difference between the shoot versus root ratio detected by these authors and our organ specificity data is likely caused by hybridization of the array probe with both the CYP85A1 and CYP85A2 mRNAs. These transcripts share 82% sequence homology, which is above the claimed 70% distinction limit of these microarray assays . In contrast to genes of the CYP85 and CYP90 families, DIM1 and DET2, encoding enzymes acting farther upstream in the BR pathway were found to be ubiquitously expressed. In addition to the differences in transcript levels, we have also detected differential distribution of BR biosynthesis intermediates between the aerial and underground organs. We found that in Arabidopsis 6-deoxotyphasterol and earlier precursors were more abundant in the roots, whereas the level of 6-deoxocastasterone and castasterone was higher in the shoots. A similar pattern of organ-specific intermediate accumulation could be observed in pea and tomato, indicating that analogous mechanisms may regulate BR distribution in these plants. In accordance with these findings, C-27 BRs (i.e. 28-norca-stasterone and its precursors) have also been shown recently to be differentially partitioned between the shoots and roots of tomato . The potential significance of higher early intermediate levels in roots and the accumulation of 6-deoxo-castasterone and castasterone in shoots is unclear, but worthy of further investigation. The distribution of CYP85 and CYP90 transcripts suggests that roots actively participate in BR synthesis. Because in several plant species root development is inhibited at sub-nanomolar BR concentrations , suppression of metabolic flow to biologically active BR forms might help to maintain the low hormone level in this organ. With the sensitivity of our GC-selected ion monitoring analysis, BL could only be detected in Arabidopsis and pea, but not in tomato, where castasterone is thought to be the only active BR . In Arabidopsis roots, the low level of CPD expression was found to coincide with the accumulation of 6-deoxocathasterone, the substrate of CPD/CYP90A1. This seems to indicate a role for transcriptional regulation in determining the abundance and activity of CYP90A1, and perhaps also other P450 enzymes of the BR pathway. Considering the importance of BRs in regulating early developmental functions, high-level expression of the BR-biosynthetic CYP85 and CYP90 genes in germinating seeds and young seedlings implies that, in addition to BL accumulation in the seeds , efficient de novo synthesis might be required for ensuring the optimal hormone concentration. Although in the whole plant, the activity of these genes declines after the seedling stage, strong expression may be maintained in differentiating regions, as it was shown in the case of CPD . Transcriptional activity of the genes involved in BR metabolism is controlled by multiple physiological factors. High level of active hormone results in the repression of biosynthetic P450 genes and the induction of BAS1 responsible for BR catabolism . In addition to their feedback regulation, the activities of CYP85 and CYP90 genes are also subject to organ-specific and developmental control. Furthermore, a recent DNA microarray analysis revealed that the expression of several genes required for the synthesis of early sterol intermediates in the BR pathway are down-regulated by light . Therefore, it is reasonable to believe that these transcriptional mechanisms are crucial for adjusting the optimal levels and maintaining the homeostasis of active BRs. MATERIALS AND METHODS : Plant Material and Growth Conditions | In vitro cultures of wild-type Arabidopsis (ecotype Columbia-0), the BR-deficient cpd , cbb3, and BR-insensitive cbb2 mutants were grown from surface-sterilized seeds on Murashige and Skoog medium supplemented with 0.5% (w/v) Suc and 0.2% (w/v) Phytagel (Sigma, St. Louis) at 22C, under 14-h-light/10-h-dark cycles. BR treatments were carried out in the same Murashige and Skoog liquid medium supplemented with 100 nm BL (CIDtech Research Inc., Missisauga, ON, Canada), whereas hormone-free control samples received only the ethanol carried in with the BL stock solution (0.01% [v/v]). The Arabidopsis plants used for organ-specific mRNA and BR analyses were grown under similar conditions in Gamborg's B5 liquid medium with continuous shaking at 50 rpm. Seeds of pea (Pisum sativum L. cv Torsdag) and tomato (Lycopersicon esculentum Mill. cv Sekaiichi) were sown in vermiculite and the germinated plants grown in the greenhouse under natural light (13-h day, 11-h night). Five-day-old pea seedlings and 22-d-old tomato plants were then grown hydroponically in the same greenhouse using liquid medium for further 10 and 14 d, respectively. Analysis of Steady-State Transcript Levels | Steady-state mRNA levels were analyzed by semiquantitative and quantitative RT-PCR assays according to with minor modifications. Total RNA was isolated from 1 g of fresh plant material using TRI Reagent (Sigma). After treatment with RNase-free DNaseI, cDNA was prepared from 5 mug of RNA with Ready-To-Go T-Primed First-Strand Kit (Pharmacia Biotech, Piscataway, NJ). One-tenth of the cDNA obtained was PCR amplified within the linear range of accuracy by specific primers spanning 250- to 300-bp regions near the 3' ends of the translated sequences. One percent of the RT-PCR products was labeled with [alpha-32P]dCTP using a single detection primer that was three nucleotides longer in the 3' direction than one of the amplification primers. Signal intensities were detected by autoradiography after size separation on a 2% (w/v) agarose gel and quantitatively evaluated using a PhosphorImager 445 SI (Molecular Dynamics Inc., Sunnyvale, CA). The constitutively expressed UBQ10 mRNA was used as internal control. The cDNA-specific PCR primers used are given in Table . The number of amplification cycles was 15 for UBQ10, 20 for CPD, DIM1 and DET2, and 25 for CYP85A1, CYP85A2, ROT3, and CYP90D1. Table III | Primers used in the RT-PCR analyses Quantitative Determination of Endogenous BR Levels | Twenty-day-old Arabidopsis, 15-d-old pea (n = 192), and 36-d-old tomato (n = 147) plants were separated into shoots (130, 289, and 241 g fresh weight, respectively) and roots (115, 293, and 78 g fresh weight, respectively). BR extraction and analysis were carried out as has been described by . In brief, methanol extracts of these tissues were subjected to solvent partitioning and purified by LH-20 chromatography and then reversed phase HPLC. Before LH-20 chromatography, charcoal chromatography was applied to the Arabidopsis shoot extracts; silica gel chromatography was applied to the extracts of pea shoots, pea roots, and tomato roots; and both silica gel and charcoal chromatography were applied to the tomato shoot extract. Quantitative analyses of BRs were conducted by GC-mass spectrometry/selected ion monitoring, using a JMS AX 505W instrument (JEOL, Tokyo). Backmatter: PMID- 11997340 TI - Discovery of Regulatory Elements by a Computational Method for Phylogenetic Footprinting AB - Phylogenetic footprinting is a method for the discovery of regulatory elements in a set of orthologous regulatory regions from multiple species. It does so by identifying the best conserved motifs in those orthologous regions. We describe a computer algorithm designed specifically for this purpose, making use of the phylogenetic relationships among the sequences under study to make more accurate predictions. The program is guaranteed to report all sets of motifs with the lowest parsimony scores, calculated with respect to the phylogenetic tree relating the input species. We report the results of this algorithm on several data sets of interest. A large number of known functional binding sites are identified by our method, but we also find several highly conserved motifs for which no function is yet known. Keywords: Introduction : One of the great challenges currently facing biologists is to understand the varied and complex mechanisms that regulate gene expression. We focus on one important aspect of this challenge, the identification of binding sites for the factors involved in such regulation. A number of computer algorithms have been proposed for the discovery of novel regulatory elements in nucleotide sequences. Most of these try to deduce the regulatory elements by considering the regulatory regions of several (putatively) coregulated genes from a single genome. Such algorithms search for overrepresented motifs in this collection of regulatory regions, these motifs being good candidates for regulatory elements. Examples of this approach include , , , , and . We adopt an orthogonal approach of deducing regulatory elements by considering orthologous regulatory regions of a single gene from several species. This approach is called "phylogenetic footprinting" . The simple premise underlying phylogenetic footprinting is that selective pressure causes functional elements to evolve at a slower rate than that of nonfunctional sequences. This means that unusually well conserved sites among a set of orthologous regulatory regions are excellent candidates for functional regulatory elements. This approach has proved successful for the discovery of regulatory elements for many genes, including epsilon-globin , gamma-globin , rbcL , cystic fibrosis transmembrane conductance regulator , tumor necrosis factor-alpha , and interleukin (IL)-4, IL-13, and IL-5 . See the review by for more details. The same idea of using comparative analysis to identify conserved elements, but among only two or three species (particularly human and mouse), has recently become popular . The major advantage of phylogenetic footprinting over the single genome, multigene approach mentioned earlier is that the latter requires a reliable method for assembling the requisite collection of coregulated genes. In contrast, phylogenetic footprinting is capable of identifying regulatory elements specific even to a single gene, as long as they are sufficiently conserved across many of the species considered. Genome projects are quickly producing sequences from a wide variety of organisms, so the data necessary for phylogenetic footprinting are becoming increasingly available. The standard method that has been used for phylogenetic footprinting is to construct a global multiple alignment of the orthologous regulatory sequences and then to identify conserved regions in the alignment. A tool such as CLUSTALW is appropriate for this purpose, as it can take advantage of knowledge of the phylogeny relating the species. To see why this approach to phylogenetic footprinting does not always work, consider typical lengths of the sequences involved. Regulatory elements tend to be quite short (5 to 20 nucleotides long) relative to the entire regulatory region in which we search for them (a 1000-bp promoter region would be typical). Given these relative lengths, if the species are somewhat diverged, it is likely that the noise of the diverged nonfunctional background will overcome the short conserved signal. The result is that the alignment may not align the short regulatory elements together. In that case, the regulatory elements would not appear to belong to conserved regions and would go undetected. Thus, when the entire regulatory regions considered are moderately to highly diverged, global multiple alignment is likely to miss significant signals. made similar observations in conjunction with their comparative analysis of several Saccharomyces species. They discovered that if the species are too closely related, the sequence alignment is obvious but uninformative, because the functional elements are not sufficiently better conserved than the surrounding nonfunctional sequence. On the other hand, if the species are too distantly related, it is difficult or impossible to find an accurate alignment (for discussion of these issues, see ). Rather than relying on multiple alignment, a more successful recent approach to phylogenetic footprinting is to use one of the existing motif discovery programs ---such as MEME , Projection , Consensus , AlignAce , or ANN-Spec ---or the segment-based multiple alignment program DIALIGN . , for instance, reported some successes using AlignAce when global multiple alignment tools failed. Another example of this approach is the work of , who used a Gibbs sampler to perform phylogenetic footprinting in bacterial sequences. Such general motif discovery algorithms were designed for a different purpose, however, and have their own drawback: None take into account the phylogenetic relationship of the given sequences; that is, these methods assume the input sequences to be independent. This can be problematic, for example, in data sets containing a mixture of some closely related species and some distant ones. If the phylogeny underlying the data is ignored, similar sequences from the set of closely related species will have an unduly high weight in the choice of motifs reported. Even if these methods were modified to weight the input sequences unequally, this would still not capture the information in an arbitrary phylogenetic tree. The method we present does capture this information. In this paper, we describe an algorithmic method designed specifically for phylogenetic footprinting in multiple species. Because it is tailored to this purpose, it avoids the drawbacks described above of both multiple alignment and general motif discovery algorithms. Given a set of unaligned orthologous sequences, our approach identifies all DNA motifs that appear to have evolved unusually slowly compared with the surrounding sequence. More precisely, given n orthologous input sequences and the phylogenetic tree T relating them, our algorithm is guaranteed to produce every set of k-mers, one from each input sequence, that have parsimony score at most d with respect to T, where k and d are parameters that can be specified by the user. As orthologous sequences from more and more species are included in the input, the distinction between conserved motif and diverged background generally becomes clearer. However, when including many orthologous sequences, particularly distantly related ones, there is increased chance that some of them may have lost or completely altered some regulatory elements over the course of evolution. For example, a species may not need the regulatory mechanism in which some regulatory element was involved, in which case selective pressure would no longer operate. As an example of an altered regulatory element, LexA has an entirely different binding motif in gram-positive bacteria than in gram-negative bacteria . For these reasons, we developed a variant of our phylogenetic footprinting algorithm that identifies motifs that occur in many, but not necessarily all, of the input sequences. This variant requires some way of comparing the levels of conservation among motifs that occur in different subsets of the input species and with different parsimony scores. For example, should one prefer a motif that occurs in all the species with parsimony score 2 or a motif that occurs in most of the species but with parsimony score 0? To address this, for each parsimony score s, we allow the user to set a minimum threshold on the fraction of the phylogeny that must be spanned by any reported motif with score s. For example, the user can ask to see all motifs with parsimony score 0 that span at least 200 Myrs of the phylogenetic tree (i.e., the sum of all branch lengths induced by the leaves containing the motif is at least 200 Myrs), plus those with score 1 that span at least 350 Myrs, plus those with score 2 that span at least 500 Myrs. Thresholds are to be set in such a way that the motifs reported are conserved at a statistically significant level (see Methods). The focus of the present paper is an explanation of what the new algorithms do and a discussion of the results obtained on several interesting data sets available in the public databases. Although a high-level description of how the algorithm works is given in Methods, the details of the algorithm, and in particular several algorithmic optimizations that render it practical on realistic problems, are beyond the scope of the present paper. These algorithmic details are described in a companion paper devoted to that purpose . The algorithms are implemented in a program called FootPrinter that has been used to obtain the results presented here. FootPrinter is available at . RESULTS : In this section, we report the highly conserved motifs found by FootPrinter in nine sets of orthologous or paralogous sequences. We identified many previously known regulatory elements, as well as many highly conserved motifs with unknown function. The data sets considered in this study were chosen according to two main criteria: (1) the availability of several orthologous promoter sequences in GenBank and (2) the availability of information about the regulation of the genes considered (to validate our results). Some sets of orthologous sequences come from the ACUTS database , which lists a number of genes for which regulatory regions have been sequenced in several vertebrates. Other data sets were built by the authors directly from GenBank. The sequences, accession numbers, phylogenetic trees, and detailed results from FootPrinter can be found at . The phylogenetic relationships among the sequences considered were derived from and , unless mentioned otherwise (see Discussion). All results are summarized in Table , but there is more detail available at the web site mentioned above. Table 1 | Motifs Found by Phylogenetic Footprinting Metallothionein Gene Family | The metallothionein gene family is particularly well suited to show the merits of our approach, as a large number of promoter sequences are available from a wide variety of species, the phylogenetic relationships among these sequence have been studied, and a large number of regulatory elements have been experimentally determined in several species. Notice that although we described phylogenetic footprinting as applied to orthologous sequences, the approach applies equally well to paralogous sequences, in which two sequences diverged because of duplication rather than speciation, as long as the gene family tree is known. The primary function of proteins in the metallothionein family is to bind to heavy metal ions and to mediate cellular detoxification of metals. They have also been shown to act as antioxidant agents, protecting DNA from free radicals (for review on the function of metallothionein, see ). The metallothionein gene family appears to have evolved through a series of gene tandem duplications and losses. Most mammals have four major isoforms (MT-I, -II, -III, and -IV). Humans actually have 13 copies of the MT-I gene. Some nonmammals (Caenorhabditis elegans, sea urchin, icefish, and trout) also have several copies of the metallothionein gene, but the duplication events that led to this situation most likely took place quite recently, in such a way that, for example, the MT-I C. elegans gene is not more closely related to the MT-I mouse gene than to the MT-II mouse gene. The phylogenetic relationships among the various members of the gene family have been studied by , and the phylogenetic tree used here (see Fig. ) is derived from theirs. Figure 1 | FootPrinter identified 12 highly conserved motifs in the metallothionein gene family. FootPrinter identified 12 highly conserved motifs in the metallothionein gene family. Each input sequence is the 590 bp located upstream of the start codon. For more details on each motif, see Table . The phylogenetic tree is derived from ; branch lengths not to scale). Numbers along branches indicate when each motif was created (unboxed) or lost (boxed), ignoring any less conserved occurrences of the motif not reported by FootPrinter. Genes from the metallothionein family are known to be regulated by a number of transcription factors. The most important of them, MTF-1, required for basal expression, binds to cis-acting elements known as metal response elements (MREs). A metallothionein promoter typically contains several MREs. In addition to MREs, the mouse MT-I promoter contains one or more GC boxes bound by Sp1, and major late transcription factor/antioxident response element (MLTF/ARE) binding site. The human MT-II promoter contains three basal level enhancer (BLE) elements known to bind transcription factors from the AP-2 family, and a glucocorticoid-responsive element (GRE), bound by the glucocorticoid receptor. Some of these bindings sites have also been identified in other metallothionein promoters. Most binding sites known in any of the species we consider occur within 300 bp of the start codon (for more on the regulation of genes in the metallothionein family, see ). We ran FootPrinter on the 590 bp of sequence located upstream of the start codon of each of the metallothionein genes listed in Figure . The 5'-UTR is usually short (between 50 and 100 bp for species for which the transcription start site is known), and was included in the sequences considered. We searched for conserved elements of the lengths 7, 8, 9, and 10, in consecutive runs, each time adjusting the parameters to ensure that the motifs reported are well conserved at a statistically significant level (see Methods). Because the family contains both orthologous and paralogous genes, the ability of FootPrinter to allow for losses of regulatory elements is particularly crucial. Indeed, duplicated copies of a gene may evolve to have slightly different functions, and it is likely that the same holds for their promoter regions. Our analysis identified 12 motifs, plotted in Figure and listed in Table . Motifs labeled 3, 4, 5, and 10 all correspond to different variants of MREs. That is, they are all experimentally verified binding sites of the same transcription factor MTF-1 in at least one of the sequences. The most common motif, labeled 3, corresponds to MREs located on the reverse strand, whereas the others are on the forward strand. Motif 3 is present in all isoforms of all deuterostomes (echinoderms and vertebrates) studied. It is often present in multiple copies, with up to four copies in human MT-IV and rat and hamster MT-II. Note that other promoter regions in Figure may also contain more copies of this motif than shown, but not sufficiently conserved to be reported (see Discussion). Motif 5 was only found in mammalian MT-I and MT-II, and motif 10 was only found in mammalian MT-II, whereas motif 4 appears to have been lost on the branch leading to mammalian MT-I, -II, and -IV. In all, FootPrinter identified five of the seven MREs documented in the TRANSFAC database for human MT-II, but only two of the six known in rat MT-I (see Discussion for an explanation of why these were missed). The other motif known to have regulatory function is motif 1, a TATA-box, which was found in all isoforms of all mammals, except in MT-I of nonrodents. Had we insisted that the motif be present in all genes of the MT-I/MT-II mammal phyla, its parsimony score would have increased greatly, as it is so poorly conserved in nonrodent MT-I, to a point at which it would no longer have been significant. The other seven conserved motifs identified are not documented in TRANSFAC . None of them were found in the mammalian MT-I and MT-II families, which may partly explain our lack of knowledge about them, as genes in these families have received much more attention than the other isoforms. Motifs 7, 8, and 9, found only in frog and in the MT-III gene family, may be of particular interest. (Motif 2 also occurs in this set of sequences, but its order with respect to that of the other motifs is not as well conserved.) Another interesting point is that most of the motifs found are present in more than one isoform family, which indicates that we gained accuracy by considering the gene family as a whole, instead of running FootPrinter separately on each of the four mammal isoforms. Insulin Gene Family | The upstream sequences of insulin genes are currently known only for three primates, two rodents (with two gene copies in each), and pig. This set of species is much less diverged than the metallothionein gene family. Because of this, we searched only for motifs with 0 mutations (for motif length 8) or 1 mutation (for motif lengths 9 and 10) at most, as motifs with more mutations would be likely to happen by chance in nonfunctional sequences. Our search identified four conserved motifs, all of them corresponding to known binding sites for that gene. (See Table for this gene family and those that follow in the remainder of this section.) Several other binding sites were missed by FootPrinter because they contained a few too many mutations (see Discussion). Had we had a few more diverged species (or many other mammalian sequences), we may have been able to identify these sites as well. Another way to counteract the lack of diversity is to search for longer motifs, as we would then be able to accept motifs with more mutations. However, in this case, searches for motifs of the lengths 12 and 15 did not yield any new motifs. IL-3 | The data set for IL-3 is quite similar to that of insulin, as only six mammalian sequences are available. We were thus limited to reporting only motifs with at most one mutation. Two of the five motifs identified are known binding sites. In fact, motif 3 actually contains binding sites for a number of factors. C-myc Promoter | The upstream sequences of c-myc are known for only 7 species, but these contain members from diverse animal phyla (fishes, batrachians, birds, and mammals). This allowed us to do a much more sensitive motif search. Four of our nine predictions are known binding sites. Notice that some of these were found only in mammals and chicken, others in all terrestrial animals, and one in all species considered, which again illustrates the necessity to search for conservation in subsets of the species. All these known binding sites are located in a 120-bp promoter region known to be very rich in binding sites. We also identified motifs with unknown function but which are as well conserved as those corresponding to known sites. Most of these novel motifs are located well upstream of the known regulatory-rich region, in an area where very few binding sites were previously known. C-myc Second Intron | Although regulatory elements are often located in the 5'-promoter region, some genes also have regulatory elements in their introns, and this is believed to be the case for c-myc. The second intron of c-myc varies in size between 971 bp (in chicken) and 1376 bp (in human). This intron was shown by to contain an antisense promoter that results in the transcription of the reverse strand of part of the c-myc gene. Our analysis indeed identified 10 highly conserved motifs. Many of these are located just downstream of the transcription start site for the reverse strand (located at position 832 in rat, corresponding to position 860 in human). All but one of these are novel motifs. C-fos Promoter | Our set of sequences for c-fos covers a set of species that is similar to that for c-myc. The results are also quite similar. Five of our eight predictions are known binding sites, four of them concentrated in an 80-bp area located about 500 bp upstream of the start codon. C-fos First Intron | The first intron of the c-fos gene is known to contain a long sequence that acts as a transcription terminator by blocking elongation . The presence of Ca2+ has been shown to prevent this premature termination of transcription. The mechanism that underlies this interesting regulation mode is still unclear. FootPrinter very clearly identified that long sequence, as it is almost perfectly conserved in all tetrapoda considered. However, we also observed that this 103-bp-long sequence is broken into three segments. The first segment is conserved in all species considered, except fugu. The second one was found only in tetrapoda. The third segment is conserved in all species and contains a cAMP response element in humans . This division into three segments indicates that each segment has a different function. Notice that in the absence of the two fish species, we would not have seen this division and would not have learned anything new about the region. Growth Hormone 1 | The growth hormone data set is our second largest; it contains sequences from nine fishes, one bird, and six mammals. Again, the subsets of species containing the motifs found vary greatly, from mammals only or fishes only to 15 of the 16 species. Five of the eight motifs identified are known binding sites in either rat or human. DISCUSSION : Known Versus Predicted Binding Sites | In the Results section, we showed that FootPrinter identifies a large number of binding sites with a function that has been established experimentally. However, there are also many known binding sites that were not found by our approach. It is illuminating to analyze why FootPrinter did not detect those motifs. Of course, in the end, the reason must be that these motifs were not sufficiently well conserved to be reported, but a more detailed study is instructive. The known binding sites missed by FootPrinter can be divided into five categories. First, some binding sites appear to have no significant matches in most other species. For example, the thyroid hormone receptor T3R binding site upstream of the rat growth hormone 1, and the Pur-1 binding site upstream of the rat insulin gene are both conserved only in rodents. There is very little hope of detecting these sites by phylogenetic footprinting, unless a large number of closely related species (in this case, rodents) are available. Second, some binding sites show very good conservation, but only over a region that is shorter than the ones we looked for. This is the case for the GC-box of metallothionein, the sequence GGGGCGG of which is perfectly conserved in the four MT-II sequences and has only one mutation in the MT-III sequences. The substring GGGG is actually conserved in almost all mammalian isoforms. We may have been able to detect these kinds of motifs, had we searched for motifs of that length. (We did search for motifs of length 7, but the GC-box was not reported because it did not span a large enough part of the tree.) However, such short motifs are often likely to occur simply by chance in nonfunctional sequences. We could have allowed such short motifs, but our results would have been more likely to contain false positives. Third, a small number of binding sites appear to be relatively well conserved but have had insertions or deletions (although it is not clear if the sequences with insertions or deletions are still functional). FootPrinter can allow for insertions and deletions in the motifs found, but we chose not to use this option, as it is believed that insertions and deletions are rare in binding sites. Allowing for insertions and deletions would thus have produced a few more true positives, but most likely at the price of many more false positives. Fourth, some motifs are quite well conserved, but they barely fail to meet our statistical significance thresholds. This is the case for the CREB binding sites and CT-I regulatory element of insulin, both of which have parsimony score 3 over a motif of length 8. Again, allowing for that many mutations would have produced a number of false positives. However, if sequences from more organisms had been available, these two motifs might have been detected without increasing the false positive rate. Fifth, some transcription factors bind as dimers, in which case the binding site may consist of two conserved regions, separated by a few variable nucleotides. For example, in metallothionein MT-II, transcription factors from the AP-2 dimer family are known to bind the BLE element. Visual inspection reveals that the pattern TGACnnnnnGCGG (where n is a variable nucleotide) is perfectly conserved in all four MT-II genes. Because of the variable internal sequence, FootPrinter did not discover this motif. However, a future version of the program will allow one to search for motifs containing a variable sequence in the middle, where mutations should not be counted. More generally, it is well known that some transcription factors can tolerate more than one type of nucleotide at a given position of the binding site. For example, the MRE binding sites of metallothionein can be described by the consensus string CTCTGCRCNCSGCCC, in which bold characters are absolutely required for metal response, R is A or G, and S is C or G . In this case, one would want to assign a smaller penalty to purine-purine transitions at position 7 of the motif than to other substitutions. The current implementation of FootPrinter assigns equal cost to each type of substitution, but we are investigating ways for the program to learn different mutation cost matrices for each position in the sequence. When reading the results presented in Figure or Table , the reader should be aware that there may be more occurrences of regulatory elements than shown, if they are not sufficiently well conserved. Comparison to Other Computational Methods | A number of existing computational techniques have been used or could be used to identify conserved motifs in orthologous sequences, although none has been designed precisely for that purpose (see Introduction). By far the tool most commonly used for phylogenetic footprinting is CLUSTALW , a tree-based global multiple alignment program. We also consider DIALIGN , a segment-based multiple alignment program, and MEME , a motif-finding technique based on expectation maximization. The output of both CLUSTALW and DIALIGN is a global multiple alignment of the input sequences. Given a correct multiple alignment, one can easily identify conserved motifs, for example, by computing the parsimony score of each column of the alignment and outputting motifs with low overall parsimony score. One could also allow for motif losses and compare the score of a motif to its evolutionary span, as we propose in this paper. (Note that neither CLUSTALW nor DIALIGN currently uses either of these approaches.) However, correctly aligning a set of diverged sequences is a difficult task. For example, CLUSTALW produces very good alignments for closely related sequences (e.g., those from the insulin family and from IL-3, which all come from mammals), but most often incorrectly aligns more highly diverged sequences, thus failing to show the conservation of some motifs. DIALIGN produces better alignments for the purpose of phylogenetic footprinting, because it starts by identifying short conserved regions and then incorporates them into a multiple alignment. In fact, for most data sets, DIALIGN correctly aligned most of the conserved motifs found by FootPrinter (and vice versa: Most conserved regions present in the alignment of DIALIGN were reported by FootPrinter). However, in the metallothionein data set, several conserved sites were misaligned by DIALIGN. In general, we believe that for large data sets containing weakly conserved motifs, or motifs present in a small subset of the input sequences, the advantage of FootPrinter over DIALIGN will become clearer. MEME is a motif-finding program that searches for motifs with high information content, but makes no use of phylogenetic information. Moreover, MEME does not consider the position at which motifs are found in each sequence, so that the motifs reported may occur in a different order in each input sequence (see Methods). Nonetheless, the majority of the motifs reported in this paper are also found by MEME. This is probably because these motifs are very highly conserved, which makes them relatively easy to identify. A notable exception is again the large metallothionein gene family, for which MEME fails to find many of the motifs that occur in small subsets of the input sequences. From the point of view of running time, DIALIGN is about 10 times slower than FootPrinter on large data sets, with motif lengths and scores as in Table , whereas CLUSTALW and MEME run roughly as fast as FootPrinter. A more quantitative analysis of the accuracy of each method on biological data such as that considered here is problematic, as there is no definitive classification of false positives and false negatives. We are currently conducting such comparative experiments on simulated data. Please refer to for more details on how the methods compare on biological sequences. Phylogenetic Information | Throughout this paper, we assume that we are given the correct phylogenetic relationship among the sequences under study. It is the use of this phylogenetic information that allows FootPrinter to accurately identify regions of interest. The phylogenetic tree should represent the evolutionary history of the sequences considered, which may be different from that of the species they come from, because of lateral gene transfers. In vertebrates, such events appear relatively rare, and we thus used the species tree as an estimate of the sequence tree. When such a trusted tree is unavailable, one could infer the phylogenetic tree directly from the sequences considered or from their neighboring coding regions. (This is what Binz and Kagi [1997] did in the case of the metallothionein gene family, and this is what we did for the insulin gene family.) In cases in which the correct topology of the phylogenetic tree remains unclear, an unresolved multifurcating tree can be used. The correctness of the parsimony scores computed obviously depends on that of the tree. Using a completely incorrect tree may greatly affect the accuracy of FootPrinter. However, using a tree with a small number of topological errors should still yield better results not than using a tree at all. Improving Accuracy | The predictions of FootPrinter could be made more accurate by injecting more prior knowledge as to what interesting solutions ought to look like. For example, the order and orientation in which regulatory elements occur in a sequence should be the same in all species, unless large-scale genome rearrangements occurred. Using this information may allow us to reject spurious motifs with order that is not consistent across species. Regulatory elements often occur several times in the same promoter (e.g., some metallothionein promoters contain up to 15 imperfect copies of MREs). Incorporating this type of information may allow us to detect regulatory element that are not sufficiently conserved to be reported by FootPrinter but that occur in several copies in each input sequence, thus boosting the statistical significance of the motif. Finally, if one had some idea about the transcription factors potentially regulating a given gene, one may want to allow motifs that look like potential binding sites for those factors to have slightly larger parsimony scores. METHODS : Algorithm | For the sake of clarity, we present here the simplest (but least efficient) version of the algorithm of FootPrinter, and we also assume that the only mutations allowed are point substitutions. The interested reader can find the extension to handle more general mutations, and the details of optimizations that make the algorithm truly practical, in a companion paper . The basic method is a dynamic programming algorithm similar to one presented by for the computation of the parsimony score of a fixed set of aligned sequences (whereas what we seek is the most parsimonious choice of k-mer from each of the input sequences). The inputs to the algorithm are n homologous sequences S1, S2, ..., Sn; the phylogenetic tree T relating them; the length k of the motifs sought; and the maximum parsimony score d allowed. The algorithm proceeds from the leaves of T to its root. At each node u of T, it computes a table Wu containing 4k entries, one for each possible k-mer. For each such k-mer s, let Wu[s] be the best parsimony score that can be achieved for the subtree of T rooted at u, if the ancestral sequence at u was forced to be s. Let the set of children of u be denoted C(u); let h(s,t) be the number of positions at which k-mers s and t differ; and let Sigma = {A,C,G,T}. The table Wu is computed according to the following recurrence: A straightforward implementation of this recurrence computes all W tables in time O(nk(42k + l)), where l is the average length of the input sequences S1, S2, ..., Sn. The main term nk 42k in this expression comes from the fact that, for each of the O(n) edges (u,v) of T, for each of the 4k possible values of s labeling u, and for each of the 4k values of t labeling v, the recurrence calls for the computation of h(s,t). If r is the root of T, each entry of Wr that is at most d gives rise to one or more solutions to be reported. For each such entry, the corresponding k-mers of the n input sequences can be recovered by retracing the recurrence from the root back to the leaves. By maintaining appropriate pointers that reflect the computation of the W tables, the set of solutions can be recovered in time linear in its size. In nonrepetitive biological sequences, the number of solutions is usually small (when d is small), and the time to enumerate them is negligible compared to the time to compute the W tables. The 42k factor in the complexity of the algorithm as described makes it impractical to use for most interesting values of k. In the companion paper , we show how various algorithmic optimizations can reduce the running time to O(nk min (l(3k)d/2, 4k + l)), which makes it quite practical for the type of data sets given in Results. Notice that the running time is proportional to nl, which is the total length of all the input sequences. This means that the performance of the algorithm's scales, as well as the number of species or length of regulatory region provided, is increased. Although the running time is exponential in either d/2 or k (depending on which of l(3k)d/2 or 4k + l is the lesser), in practice both of these parameters are quite small: Typical values in our experiments were k = 10 and d = 3. Using a desktop workstation, a typical run of the algorithm on a data set of n = 10 sequences of length l = 700 each might take 30 seconds if only substitutions are allowed or a few minutes if insertions and deletions are allowed as well. Handling Motif Losses | Here we discuss the generalization of the phylogenetic footprinting algorithm to identify motifs that may be missing (or highly mutated) in some of the input sequences. For this problem to make sense, there must be a way to compare two solutions containing motifs from different subsets of species. To do so, consider the total amount of evolution (measured, e.g., in millions of years) that the motif has survived. Motifs that have resisted a large amount of evolution are more likely to be interesting than those that span a short time. To be able to estimate the amount of evolution spanned by a set of species, the algorithm must be given not only the phylogenetic tree T that relates the species, but also the length of each of its branches. We estimated branch lengths by computing pairwise alignment scores for the input sequences and using the Fitch-Margoliash algorithm from the PHYLIP package to find the branch lengths that make the tree distances match the pairwise distances as closely as possible. Estimating branch lengths is a notoriously difficult problem, and our estimates may be inaccurate. However, our experience indicates that the quality of the results obtained by the method does not depend very strongly on the accuracy of these estimates. The algorithm identifies motifs that have small parsimony score but span a large part of T. More precisely, the algorithm solves the following problem. In addition to the inputs provided to the basic phylogenetic footprinting algorithm (described above), the user also provides thresholds delta0, delta1, ..., deltad. The problem is to find all sets of k-mers, one from each of the leaves i1, i2, ..., im, where {i1, i2, ..., im} is any subset of the n leaves of T, such that the parsimony score P of this set of k-mers on the subtree induced by the leaves i1, i2, ..., im is at most d, and such that the subtree induced by the leaves i1, i2, ..., im has total branch length at least deltaP. For example, the user can ask to see all motifs with parsimony score 0 that span at least 200 Myrs of T (i.e., delta0 = 200 Myrs), plus those with score 1 that span at least 350 Myrs (i.e., delta1 = 350 Myrs), plus those with score 2 that span at least 500 Myrs (i.e., delta2 = 500 Myrs). The algorithm that solves this generalized problem is very similar in spirit to the dynamic programming algorithm described above. It is a few times slower than the algorithm that does not allow losses, but it produces much more accurate results. This is the algorithm that was used to identify the motifs reported in Table . The interested reader can find further details of the algorithm in . Other Useful Parameters | FootPrinter has a number of options that help to find more actual binding sites, although leaving out spurious hits. We briefly discuss some of them here. First, notice that in our formulation of the phylogenetic footprinting problem, the position at which a motif is found in each sequence is ignored. This is a typical feature of local alignment methods, to which our approach belongs. However, in some circumstances, it is desirable to penalize motifs with positions in the set of homologous sequences that vary too much and are thus unlikely to be instances of a single conserved binding site. There is a natural way to incorporate the notion of position into a parsimony score: We simply augment the definition of motif (which until now was just a k-mer) with a number that indicates the approximate position of the k-mer in the sequence. For this study, we usually divided each input sequence into 10 equal-sized regions and assigned a cost of one mutation for a motif to move to an adjacent region. In fact, to avoid inaccuracies when a motif occurs near a region boundary, we view each motif as also occurring in the two adjacent regions. This approach of dividing the sequences into regions only makes sense if we believe that corresponding regions of each input sequence are approximately homologous. For upstream sequences, this may be a reasonable assumption. However, for introns we did not use this option, as the variation in intron size makes it unclear which portions are homologous. In our original definition of the phylogenetic footprinting problem allowing for regulatory element losses, there is no cost associated with losing a regulatory element, except that the motif spans a smaller part of the tree. This sometimes leads to undesirable situations, in which FootPrinter finds a motif that seems well conserved in two very distantly related species X and Y (thus spanning a large part of the tree) but that appears to have been lost independently in all phyla branching between X and Y. These multiple independent losses are quite unlikely in evolution, and one would like to penalize motifs that have been lost along too many branches. Once again this fits very nicely into our parsimony framework. We do so by assigning a cost to losing a motif along a given branch. For the results reported in this paper, we equate this cost to one substitution. However, one could also assign different loss costs along different branches, so that losses along long branches cost less than those along short branches. In the case of gene families, one may want to assign a smaller cost to motifs lost on branches that follow duplication events, as regulatory elements may be likely to be lost at these times. Finally, it is often useful to restrict the number of mutations along any given branch of the phylogenetic tree. For example, in cases in which a motif is very well conserved in some subset of the sequences, this avoids finding spurious poorly conserved instances of the motif in sequences that actually do not contain the true binding site. Limiting the number of mutations per branch also has a very positive effect on the running time of the algorithm. Statistical Significance | Any set of sequences contains some best conserved motif, but that does not mean that this motif was actually under selective pressure. To make sure that the motifs reported have a mutation rate significantly less than that of the surrounding nonfunctional sequence, we generated a set of random sequences with approximately the same evolutionary history as the input sequences. This set of sequences was generated by simulating evolution over the given phylogenetic tree with the inferred branch lengths. These simulated sequences thus mimic the real input sequences, except that the mutation rate is the same at all sites, and thus we should not find any unusually well conserved motifs in them. In this paper, a motif M with parsimony score s over a tree of size deltas was reported only if the probability of finding such a motif (or one better conserved) in simulated sequences is <5% (for more details on measuring statistical significance, see ). WEB SITE REFERENCES : ; Sequences, accession numbers, phylogenetic trees, and detailed results from FootPrinter. ; FootPrinter, the computer algorithm described in this paper. ; The tree of life web project. ; A database on transcription factors and their DNA binding sites. ; University of Zurich web site. Backmatter: PMID- 11997345 TI - Large-Scale Protein Annotation through Gene Ontology AB - Recent progress in genomic sequencing, computational biology, and ontology development has presented an opportunity to investigate biological systems from a unique perspective, that is, examining genomes and transcriptomes through the multiple and hierarchical structure of Gene Ontology (GO). We report here our development of GO Engine, a computational platform for GO annotation, and analysis of the resultant GO annotations of human proteins. Protein annotation was centered on sequence homology with GO-annotated proteins and protein domain analysis. Text information analysis and a multiparameter cellular localization predictive tool were also used to increase the annotation accuracy, and to predict novel annotations. The majority of proteins corresponding to full-length mRNA in GenBank, and the majority of proteins in the NR database (nonredundant database of proteins) were annotated with one or more GO nodes in each of the three GO categories. The annotations of GenBank and SWISS-PROT proteins are available to the public at the GO Consortium web site. Keywords: Introduction : Biomedical research over the last century has made tremendous progress in our understanding of biology and medicine. The recent genomic sequencing of human, mouse, and other organisms, and high-throughput studies, such as those based on microarray technology, have been yielding massive amounts of data. However, the knowledge accumulated so far is mainly fragmented. Full utilization of this data and its integration with existing knowledge can be facilitated by a systematic representation of knowledge, that is, the development of ontology. Ontology is the formalized specification of knowledge in a certain subject. Great potential exists for ontology-based literature retrieval in biomedical research , ontology-based database integration in drug discovery, and ontology-facilitated biomedical research. Recently, the Gene Ontology (GO) Consortium has developed a systematic and standardized nomenclature for annotating genes in various organisms. Using three main ontologies ---molecular function, biological process, and cellular component ---a significant number of genes in yeast, Drosophila, mouse, and other model organisms have been annotated, either manually or automatically . Association between ontology nodes and proteins, namely, protein annotation through gene ontology, is an integral application of ontology and has many practical uses. For example, designing of microarray probes would be greatly facilitated by a comprehensive understanding of all the genes involved. A microarray aimed to examine a particular process, such as apoptosis, would optimally have probes against all the genes significantly and directly involved in apoptosis. These genes can be chosen using GO annotations. To efficiently annotate proteins, we have developed a software platform, the GO Engine, which combines rigorous sequence homology comparison with text information analysis. During evolution, many new genes arose through mutation, duplication, and recombination of the ancestral genes. When one species evolved into another, the majority of orthologs retained very high levels of homology. The high sequence similarity between orthologs forms one of the foundations of the GO Engine. Text information related to individual genes or proteins is immersed in the vast ocean of biomedical literature. Manual review of the literature to annotate proteins presents a daunting task. Several recent papers described the development of various methods for the automatic extraction of text information . However, the direct applications of these approaches in GO annotation have been minimal. We used simple correlation of text information with specific GO nodes in the training data to predict GO association for unannotated proteins. The GO Engine combines homology information, a unique protein-clustering procedure, and text information analysis to create the best possible annotations, as represented schematically in Figure . Figure 1 | A schematic representation of automatic GO annotation. A schematic representation of automatic GO annotation. Solid, dotted, and dot and dash intervened arrows indicate flow of GO information, sequence information, and text information, respectively. The availability of GO annotations for a significant number of proteins from different organisms presents an opportunity to examine the cellular localization, molecular function, and involvement in a biological process of each of these proteins through the multiple and hierarchical structure of Gene Ontology. We report here our brief analysis of human proteins using the GO Engine annotation system. RESULTS AND DISCUSSION : Protein Database, Input GO Annotation, and Homology Analysis | As a first step in the annotation process, we collected proteins from different sources to build a database of proteins, some of which have been annotated by the members of the GO Consortium. This database is considered to contain the majority of prototype proteins and served as the main driver of the GO Engine. The rationale for using such a database is that there is a finite number of nonmutated proteins in existence, corresponding to a finite number of cognitively distinguishable functions, locations, or biological processes. If the majority of proteins in this database are accurately annotated with GO nodes, then the annotation for any query protein can be deduced from the annotation of a known protein in this database, which is homologous to the query protein. The National Center for Biotechnology Information (NCBI) nonredundant protein database contains proteins from a diverse array of sources, and thus it served as the major source for our protein database. Proteins from the Saccharomyces genome database (SGD) and the Drosophila genome database (Flybase) were added. The database used in this study comprises 670,130 proteins. Initial GO annotations of proteins were obtained from several sources. Members of the GO Consortium have annotated a substantial number of proteins. Their annotations were collected and mapped to proteins in our protein database. In addition, various conversion tables that link Enzyme Commission number, InterPro protein motifs, and SWISS-PROT keywords to GO nodes, which are available from the Gene Ontology Consortium web site, are used to annotate additional proteins in the protein database. The combined GO annotations of proteins served as the training data for the text information analysis and also served as input GO annotation for the GO Engine. The current annotation process exploits the transitive nature of protein homology. This homology transitivity has been used previously , and the merits of this approach have been debated. We found that, with additional input data, such as information derived from protein-domain features, text information analysis, and cellular localization prediction, this homology transitivity can be used as the main engine for predicting GO annotations of unknown proteins. Rigorous and detailed homology comparisons among these 670,130 proteins were performed to delineate the degree of homology between protein pairs by using BLASTP in BLAST with default parameters . Table 1A lists the distribution of the BLASTP results. Overall, 78.5 million pairs of proteins were found to have E scores lower than 10 --2. To accurately calculate the sequence similarity, we performed global alignment for each pair of homologous proteins identified with the BLAST program, using the Needleman-Wunsch algorithm . Table 1B shows the distribution of protein pairs in terms of the identity percentage between them. The majority (68.5%) of protein pairs have identity percentages in the range of 10% --50%. Table 1A | Distribution of the Homology Levels among Pairs of Proteins in Our Protein Database Table 1B | Statistics of Identity Levels of Homologous Protein Pairs Identified by Blastp in Table 1A Textmining and Prediction of Cellular Localization | Many earlier GenBank records and all SWISS-PROT records contain text information, which generally describes the functions of gene products. In addition, one or more reference articles were sometimes identified in the respective field of the GenBank and SWISS-PROT records. The reference articles relevant to the proteins in our database were obtained from the MEDLINE database in the National Library of Medicine, National Institutes of Health. Almost all of them have titles, abstracts, and MeSH terms. In total, 115,527 unique proteins from our protein database were linked to 86,599 MEDLINE records. A few of them lack contents in abstracts or medical subject headings (MeSH) terms. Among those proteins, 61,032 were linked with a single paper. Forty-six MEDLINE records have over 100 protein correspondences. Such records tend to be those reporting on high-throughput cDNA sequencing studies. We applied a simple computational linguistics technique to analyze the textual information from titles, abstracts, MeSH terms, and definition lines of gene records. Text contained in the sequence-related papers and definition lines in sequence records were extracted. The extraction process involves elimination of negative sentences, word stemming, and generation of predictive words. Table lists some general statistics of text information from available sequence databases. A simple, yet predictive, probabilistic model was then applied to create possible GO annotations based on the associated text information. Definition lines of sequence records, MeSH term annotations, titles, and abstracts from the sequence-related papers were modeled separately. Table 2 | Statistics on Textual Information Analyzed by GO Engine For the text analysis, the frequency of association of a specific term with a specific GO node in the training data was examined. Parameters such as boundaries of the frequency of MeSH terms and other words were optimized through the training process, using self-validation and cross-validation methods. Logarithm of odds (LOD) scores, defined as the logarithm of the ratio between the association frequency of any term --GO pair and the calculated frequency of the random combination of this pair, were used to indicate the relatedness of certain terms with certain GO nodes. These LOD scores were found to be correlative with the accuracy of GO prediction, as shown in Figure . Text information from titles of MEDLINE records appears to have more predictive power, in particular at lower LOD scores, than does text information from other categories . This probably reflects the fact that the title tends to summarize the gist of an article in a straightforward manner. MeSH terms havesimilar predictive capabilities as the abstracts, possibly because the MeSH terms are derived from the abstracts, and thus have similar information content. Figure 2 | The calculated logarithm of odds (LOD) scores in textual information analysis correlated well with the accuracy of GO predictions. The calculated logarithm of odds (LOD) scores in textual information analysis correlated well with the accuracy of GO predictions. The result is based on self-validation studies. LOD score is calculated as defined in Methods. Only predictions made with LOD scores above 2 were evaluated here and used in the GO Engine. Any LOD scores above 99 are collapsed to 99. GO prediction for any particular protein is considered accurate if the predicted GO node is the same as one of the input GO nodes for this protein or the predicted GO node is a parent or a child of one of the input GO nodes. Based on text information, a significant number of proteins were predicted to be associated with one or more GO nodes. Table lists the number of proteins with predicted GO nodes from four types of text information in the three categories of GO. These predicted GO annotations were incorporated in GO Engine to increase the accuracy of homology-based GO annotation and to generate de novo annotations. To further enhance the accuracy and coverage of GO Engine, we used a computational platform for predicting cellular localization, ProLoc (A. Novik et al., in prep.), to predict the cellular localization of individual proteins based on their inherent features such as specific localization signatures, protein domains, amino acid composition, isoelectric point (pI), and protein length. Only protein sequences that begin with methionine underwent ProLoc analysis. Thus, 88,997 of 93,110 proteins in SWISS-PROT version 39 were analyzed, and 78,111 proteins have one to three GO predictions in the cellular component category through ProLoc. Table 3 | Statistics on the Number of Proteins with One or More GO Predictions from MeSH Terms, Titles, and Abstracts of Articles Referenced in the GenBank and SWISS-PROT Records and from Definition Lines of Protein Records GO Annotation | To use the homology transitivity between proteins from different species, we developed a progressive single-linkage clustering process. GO Engine clustered the proteins through single linkage; that is, a protein belongs to a cluster if this protein has sequence homology above a certain threshold with one member of the cluster. The threshold progressed from high homology levels to lower ones, with some defined granularity. The protein clustering and GO annotation were performed at each granularized homology level. The granularity resolution is 1% for global alignment identity; that is, for example, clustering was first performed at 98%, then at 97%, and so on. The granularity is 10-fold for the E score of a BLASTP homology pair; for example, clustering was performed at 10 --50, then at 10 --49, and so on. To show clustering efficiency and homology transitivity, we examined all homology pairs clustered with at least 90% identity. There were a total of 57,004 clusters containing 263,259 protein members in this level. Among these clusters, 23,321 clusters contained at least three protein members. Figure shows a histogram of identity percentage between proteins with the 90% identity clusters with at least three protein members. The lowest homology pairs had an identity of 46% while being clustered at 90% or higher identity levels. Figure 3 | The number of protein pairs at different identity levels in clusters identified at greater than 89% identity level through single-linkage clustering. The number of protein pairs at different identity levels in clusters identified at greater than 89% identity level through single-linkage clustering. The result indicates that single-linkage clustering efficiently grouped proteins into clusters. Clusters containing proteins with GO annotations were analyzed along with the GO prediction from the text information, and best annotations for individual proteins of the cluster were selected through an error weight calculation. Table 4A shows the number of the input and output GO annotations through GO Engine. Over 85% of proteins were annotated with one or more GO nodes in each of three GO categories. GO annotations for the majority of proteins with complete coding sequence in GenBank 122 and the majority of SWISS-PROT version 39 were deposited in the web site. Table 4B lists the breakdown of the number of proteins annotated at different homology levels. The results indicate that GO annotations were achieved throughout the whole spectrum of homology level. Table 4A | Statistics on the Number of Proteins in Each of the Three GO Categories with Original GO Input (from GO Consortium, Enzyme Conversion, SWISS-PROT Keyword Mapping, InterPro Mapping) and the Number of Proteins in GO Engine Output Table 4B | Statistics on the Number of Proteins Annotated at Different Homology Ranges in Each of the Three GO Categories The accuracy of these annotations by GO Engine was assessed through cross-validation. One-fifth of input GO annotations were withheld during the GO annotation process, and the resultant annotations were compared with these withheld GO nodes. For each protein, the GO node with the lowest error score was examined. Table 4C lists the coverage and accuracy of such a representative test. The coverage ranges from 96% to 99% and the reproducibility is between 65% and 80%. The lower reproducibility of GO annotation in the "cellular component" category, as compared with that in the other two GO categories, is consistent with the notion that a short amino acid segment of a particular protein such as a signal peptide and a nuclear localization signal affects the cellular localization. The presence or absence of short amino acid segments cannot be completely captured through sequence similarity comparisons. Detailed analysis of the validation data indicates that the accuracy of the annotation correlates with the homology level during the annotation (data not shown). Manual validation of GO Engine annotations was performed on a total of over 500 annotations, and about 85% --93% of annotations were found to be correct. The higher percentage of accuracy in the manual examination compared with the automatic cross-validation may result from the incomplete input GO annotations. An additional analysis was performed using manually curated GO annotation from European Bioinformatics Institute (EBI), which was recently available and not yet incorporated in GO Engine. This validation evaluated the overall accuracy of automatic approaches, including various mappings (InterPro2GO, SWISS-PROT keyword2GO, and Enzyme2GO) and GO Engine. In EBI gene association data deposited in the GO consortium web site as of March 6, 2002, there were 9666 human proteins manually annotated with 32,590 GO nodes (with evidence of codes other than 'IEA'). Among these 9666 human proteins, 5413 have accession numbers present in SWISS-PROT version 39. In GO Engine annotation, there were 5839 human proteins with 27,115 GO annotations from SWISS-PROT version 39 (see Compugen file in the GO Consortium web site). The 5359 overlapping proteins between EBI manual annotations with 18,537 unique GO nodes and GO Engine annotations with 27,115 GO nodes were further used for comparison. Among these 5359 proteins, 3603, with 19,477 GO Engine annotations, had no direct annotations from GO mappings in all three GO categories, indicating the majority of GO annotations in this cohort were from GO Engine. Between the EBI set of data and the GO Engine set of data, 14,695 annotations were exactly matched. Among the rest of the unmatched EBI GO assignments, 47 had no GO Engine assignments in the corresponding GO category, 1329 of them were the parents, 411 of them were the children of one of the GO Engine assignments, and 2055 of them were incompatible with any of the GO Engine assignments for corresponding proteins. A separate analysis was performed to examine the compatibility of GO Engine prediction with the EBI manual curation. GO annotations with the lowest error scores in GO Engine above the homology level of an E score of 10 --10 in each of three GO categories for the 5359 overlapping proteins were compared with EBI manual annotation. Among a total of 15,416 GO annotations, 6815 matched exactly with EBI annotation and 1558 were parents of one of the EBI annotations. Such annotations are correct, yet not specific enough. Six hundred thirty six GO Engine annotations were children of one of the EBI annotations. Three thousand two hundred forty three GO Engine annotations are in different paths of GO hierarchy from the corresponding EBI annotations. They are likely to be incorrect, although some of them may indicate novel or rare protein functionalities. In addition, 3164 GO Engine annotations had no corresponding GO annotations from the EBI data set, and thus these annotations may provide some potentially correct annotations. These results suggested that automatic approaches, mainly through GO Engine, could capture the majority of the GO annotations achievable through manual curation, and provide reasonable ground for future curation and experimental verification. Table 4C | Results of One of the Cross-Validation Tests of GO Engine Annotation Human Protein GO Annotations and Analysis | The availability of a large number of GO-annotated proteins from other organisms presents an opportunity to investigate GO features of human proteins in detail. For this purpose, we obtained the Ensembl version 1.0.0 , and annotated proteins through InterPro scanning , InterPro-to-GO node conversion, and GO Engine. Of 27,333 proteins corresponding to 16,913 contigs in version 1.0.0 Ensembl, 23,036 had been annotated in one or more GO categories: cellular component GO for 15,466 contigs, molecular function GO for 15,271 contigs, and biological process GO for 14,939 contigs. Figure indicates the number of proteins and contigs in each major GO node of the three GO categories that contain more than 300 proteins. Figure 4 | Histograms show the number of proteins and contigs from Ensembl version 1.0.0 in the major nodes in three GO categories: cellular component (A), molecular function (B), and biological process (C). Histograms show the number of proteins and contigs from Ensembl version 1.0.0 in the major nodes in three GO categories: cellular component (A), molecular function (B), and biological process (C). The number of any particular node represents the sum of the number of proteins annotated with this node and that with all children nodes. The sum of all numbers may exceed the total number of proteins or the total numbers of contigs because the annotations of some nodes, which are the children of several higher nodes, are counted multiple times. The genomic localizations of the majority of Ensembl contigs have been identified in the Ensembl data release. We investigated the distribution of chromosome localization of annotated proteins across the GO hierarchy. Each major GO node in the three GO categories with at least 50 contigs assigned to that GO or its children nodes was analyzed. The actual number of contigs localized in each chromosome was compared with the expected number of contigs for this GO node and its children nodes, and significant differences between these two distributions were identified using a chi-square test. Table lists chromosomal distribution of a few selected GO nodes in biological process, cellular component, and molecular function categories, including the actual number of contigs annotated with the particular GO node and its children nodes, the chi-square test score, and the associated P values. In general agreement with previous reports , current annotation identified 1443 transcriptional factors in molecular function GO and 1771 proteins involved in 'transcriptional regulation' in biological process GO. Transcription-related proteins are densely coded in chromosome 19. Although the biological significance of this distribution remains to be determined, it is conceivable that such a clustering of transcriptional factors in specific chromosomal segments, and the more localized clustering in local genomic regions, may be related to particular chromosome structure and, more importantly, this clustering might serve as a regulatory mechanism for the coordinate activation of such genes and the rapid accumulation of transcriptional activity during a period of hyperactivity of gene expression, such as in early development after fertilization, or proliferative activation following a quiescent state. As would be expected, proteins involved in spermatogenesis are clustered on the Y chromosome. Further examination of some of the chromosomal distribution patterns revealed the existence of gene clusters for specific protein families (proteins annotated with a specific GO node or its children GO nodes). For example, we found that in chromosomes 12 and 17, the genes encoding intermediate filament proteins are clustered, and in chromosome 17, at least 10 of those genes also reside in a short stretch of genomic DNA (data not shown). Such observations were not surprising, because protein families such as certain globin family , protocadherin , and Hox genes among many others, are clustered on a segment of genomic region. The annotation of proteins and the structured hierarchy of GO Engine may allow sophisticated and systematic analyses of human proteins. For instance, a broader definition of 'gene cluster' can be used to refer to genes having similar functions, or being involved in a defined process, or being localized in a defined cellular component in a contiguous genomic segment. Such gene clusters may correlate well with specific chromosomal structure, or specific evolutional events. With systematic GO annotation, these clusters can be computationally identified and investigated. Table 5 | Number of Contigs Annotated with Selected GO Nodes and Their Children Nodes in the Three GO Categories, and Chi-Square Test Results METHODS : Data Collection | Gene Ontology and gene association files were obtained from the Gene Ontology Consortium , InterPro from and the enzyme database from . The following databases and versions were used: GenBank release 122.0; SWISS-PROT release 39.0 ; Enzyme database release 26.0; InterPro database as of April 6, 2001; NCBI LocusLink data as of March 6, 2001; MEDLINE databases as of April 6, 2001; and the following files from the Gene Ontology Consortium: gene_association.fb (version 1.26, 2001/02/19), gene_association.mgi (version 1.19, 2001/03/01), gene_association.sgd (version 1.251,2001/03/13), gene_association.pombase (version 1.2, 2000/07/22), ec2go (version 1.2, 2000/10/23), and swp2go (version 1.4, 2000/11/15). Fifty-eight thousand one hundred eighteen SWISS-PROT proteins have been assigned with at least one GO node by the following sources: 15,534 proteins were assigned with at least a functional GO node by conversion of enzyme nomenclature (EC) to GO nodes. Mouse Genome Informatics (MGI) has assigned 5984 SWISS-PROT proteins with GO nodes . Thirty-one thousand eight hundred sixty-nine SWISS-PROT proteins were assigned at least one GO node using SWISS-PROT keyword correspondence and 33,048 SWISS-PROT proteins were assigned at least one GO node by InterPro scanning. The nonredundant protein database is constructed from GenPep file from NCBI, along with proteins collected from Saccharomyces Genome Database (SGD) and Flybase, with a total of 670,130 proteins. Sequence Similarity Analysis | A two-stage strategy was used to build a detailed homology map between all proteins in our protein database. In the first stage, all protein pairs with an E score lower than 0.01 using BLASTP with default parameters were cataloged. In the second stage, all of these homologous protein pairs were aligned through the Needlman-Wunsch algorithm with a global alignment to obtain the percentage of identical amino acids between the two proteins. BLOSUM62 was used as the substitution matrix. The percentage of identity is defined as the number of amino acids aligned with nonnegative scores divided by the number of amino acids in both aligned and unaligned length of two proteins in the global alignment. This two-stage homology searching was the most computation-intensive part of GO annotation. Text Information Extraction | Both GenBank and European Molecular Biology Laboratory (EMBL) databases contain references to the bibliographic information. NCBI staff also add specific MeSH terms to MEDLINE records. In addition, almost all of these records contain abstracts. Efforts were made to obtain the correlations between the presence of specific MeSH terms, or specific English words, in the referred papers and GO assignments in the training data. The correlations were then used to predict GO nodes for unassigned genes. Noncharacters in titles and abstracts and in the definition line of gene records were eliminated and words were stemmed through the Lingua::stem module from . Because of the standardized and curated nature of MeSH terms, MeSH terms were not parsed or stemmed. The frequency of each word in all the available text information was calculated. Words that occur at least five times over the whole text information space are retained for further studies. This cutoff threshold is used to eliminate rare words, wrong spellings, and sometimes even the base-pair sequence present in either the definition lines or abstracts. In addition, an upper limit of word frequency (common words such as 'and,' 'gene,' and 'protein' have very high frequencies) and a lower limit of word frequency are defined through a repeated training process and manual review. The words within the upper and the lower limits are considered as predictive. Because the correlation between the GO nodes and specific words is in a positive nature, negative sentences with words such as 'not' and its variants, such as 'unlikely' or 'unresponsive,' were excluded from consideration. Proteins with GO annotation from other sources such as GO Consortium, InterPro scanning, or keyword mappings were used as training data to obtain the correlation between specific words with specific GO nodes. The following formula was used: S = log(P(m,g)/P(m)P(g)), where S is the LOD score for the word m --GO g combination, P(m,g) is the frequency of the term m and GO node g co-occurrence among all word and GO combinations, P(m) is the frequency of occurrence of the term m among all word occurrences, and P(g) is the frequency of occurrence of GO node g among all GO occurrences. To predict GO nodes for any specific protein that is linked to one to a few dozen words, we calculate and sort the sums of LOD scores from all of these words for each possible GO node, and we use them for further GO annotation. Multiple MeSH terms --GO correlations were tested and were found to be no more informative than the single MeSH term --GO correlation, and therefore they were not used. GO Assignment --GO Engine | GO Engine uses the existing GO annotations as inputs. A substantial number of proteins have been annotated by different groups in the GO Consortium. Their association files and LocusLink GO association were obtained. Additional protein --GO associations were built by using translation files between Enzyme nomenclature and GO nodes, between InterPro entries and GO nodes, and between SWISS-PROT keywords and GO nodes. All of these translation files were available in . Progressive single-linkage clustering was used to assign GO node to proteins. The assignment started with clustering proteins at the highest homology ---99% identity from the global alignment. In any cluster with GO-assigned proteins, other proteins are assigned GO nodes based on both the cluster GO nodes and the GO predictions based on text information analysis and ProLoc for any individual proteins and for other proteins in the cluster. For any cluster with a single GO node, all members of this cluster are assigned this GO node. For any cluster with multiple GO inputs, an error weight scheme was applied to determine the final GO nodes. The error weight scheme works as follows. For each prediction method, the error score matrix is obtained from the validation studies. For example, during text information analysis of titles, a LOD score of 29 has 90% accurate GO prediction in the cellular component category from validation studies (see Fig. ); then any GO cellular component prediction made at a LOD score of 29 has an error score of 0.10. During GO Engine annotation, in any cluster, for any input GO node, the product of all error scores associated with this GO node is the final error score of that GO node. The final GO outputs are sorted according to low final error score for each GO node. The error score of any GO assignment is inherited throughout the GO Engine process. Precedence was also given to an individual protein with its own GO textual information predictions and then the combined GO textual information predictions for the whole cluster. A specified number of GO nodes (for example, from 1 to 5) were predicted for each protein in the cluster along with the calculated error score. After the assignment of a cluster, the homology linkage with the cluster was broken. The newly assigned protein with GO nodes and associated error scores was then clustered with other proteins at a lower homology level, and proteins in the new cluster could then be assigned with GO inputs from the original GO input from other proteins, if any, and the input GO nodes from these just-annotated proteins, using the same error weight scheme. The clustering and GO annotation reiterated from highest homology 99% to 35% (with the granularity of 1%) and shifted to the BLASTP-based homology scheme (E score with the granularity of E-1). At all homology levels, any clusters with only GO inputs from text information analysis were not analyzed, and the clusters were not broken until at least one protein member had original GO annotation, or had GO annotation through the GO Engine. By then, GO Engine assigned GO nodes to each member protein of the cluster, and the cluster was then broken. In the final step, any cluster containing proteins with only one or more textual information GO predictions were analyzed, and proteins within the cluster were assigned GO nodes accordingly. ProLoc predictions were treated the same as the prediction from textual information analysis with its own error score consideration. Statistical Analysis and Quality Assurance | The chi-square test was used for investigating the GO distribution across chromosomes. After each production, a manual check was performed for at least 100 GO assignments. The manual check involved homology searches, literature review, and expert evaluation. WEB SITE REFERENCES : ; Peri modules. ; Source for InterPro files. ; Ensembl project. ; The enzyme database. ; Gene Ontology Consortium web site. ; MGI web site. ; MEDLINE. Backmatter: PMID- 11997348 TI - A Fine Physical Map of the Rice Chromosome 4 AB - As part of an international effort to completely sequence the rice genome, we have produced a fine bacterial artificial chromosome (BAC)-based physical map of the Oryza sativa japonica Nipponbare chromosome 4 through an integration of 114 sequenced BAC clones from a taxonomically related subspecies O. sativa indica Guangluai 4 and 182 RFLP and 407 expressed sequence tag (EST) markers with the fingerprinted data of the Nipponbare genome. The map consists of 11 contigs with a total length of 34.5 Mb covering 94% of the estimated chromosome size (36.8 Mb). BAC clones corresponding to telomeres, as well as to the centromere position, were determined by BAC-pachytene chromosome fluorescence in situ hybridization (FISH). This gave rise to an estimated length ratio of 5.13 for the long arm and 2.9 for the short arm (on the basis of the physical map), which indicates that the short arm is a highly condensed one. The FISH analysis and physical mapping also showed that the short arm and the pericentromeric region of the long arm are rich in heterochromatin, which occupied 45% of the chromosome, indicating that this chromosome is likely very difficult to sequence. To our knowledge, this map provides the first example of a rapid and reliable physical mapping on the basis of the integration of the data from two taxonomically related subspecies. [The following individuals and institutions kindly provided reagents, samples, or unpublished information as indicated in the paper: S. McCouch, T. Sasaki, and Monsanto.] Keywords: Introduction : Rice is the staple food crop for more than one-half of the world's population and has been used as the major model system in cereal genome research because of its small genome size and high synteny to other monocots . For a model plant, rice also has an efficient transformation system , a high-density genetic linkage map , and a comprehensive rice transcript map with 6591 expressed sequence tag (EST) sites . The rice genome, estimated to be similar430 Mb , is the smallest one among major cereals, but it is by far the largest plant genome to be sequenced. Cultivated rice (Oryza sativa) is classified into two subspecies, indica and japonica , which include most of the rice cultivars grown in the world. An international consortium, the International Rice Genome Sequencing Project (IRGSP) has adopted a "clone by clone" strategy to completely and accurately sequence the entire genome of the rice O. sativa ssp. japonica cultivar Nipponbare . This strategy requires a comprehensive physical map of the entire japonica Nipponbare genome. Approaches based on BAC fingerprinting, iterative hybridization, and sequence tag connector (STC, which is to use BAC-end sequences for connecting BAC clones by sequence identity) have been successfully applied in the physical mapping of large complex genomes, including human and Arabidopsis thaliana . In addition to sequencing the genome of the O. sativa ssp. indica cultivar Guangluai 4, we are sequencing the Nipponbare chromosome 4 as part of the international effort. Here we present the construction of a sequence-ready BAC physical map of the chromosome 4 of O. sativa ssp. japonica Nipponbare through an integrated approach using a large set of sequence scaffolds from a taxonomically relative subspecies O. sativa indica variety Guangluai 4 as a starting point. This approach, based on the assumption that the indica and japonica genomes are collinear, included the experimental hybridization and STC approaches in parallel with the sequencing project and can serve as a model system with publicly available genome sequence data from one species to construct the genome physical map of a taxonomically related species. RESULTS AND DISCUSSION : Identification of the Chromosome 4 Seed BACs from the Rice Subspecies Indica Guangluai 4 | We initiated a genome-sequencing project of the O. sativa indica Guangluai 4 in 1993. Two BAC libraries of the Guangluai 4 were constructed with restriction enzymes HindIII or BamHI designated as OSIGBAC-H and OSIGBAC-B, respectively. OSIGBAC-H clones were fingerprinted and assembled into BAC contigs to cover the entire genome . Because of a low genome coverage of the OSIGBAC-H library (<4X) and limited unique markers available for BAC anchoring, the actual genome coverage of the contigs was less than estimated. To construct a more complete physical map of the Guangluai 4 chromosome 4, we took a total of 182 genetic restriction fragment length polymorphism (RFLP) markers of the Nipponbare chromosome 4, provided by RGP , as probes to hybridize the Guangluai 4 BAC libraries. Collectively, 566 BACs were identified by hybridization with 108 RFLP markers and 50 contigs assembled. The anchored BAC contigs were formed as a regional physical map of the Guangluai 4 chromosome 4. We also generated 13,000 BAC-end sequences from the OSIGBAC-H library for the STC approach to extend the anchored BAC contigs. The relationship within the contigs was further confirmed by restriction enzyme digestion and PCR analysis. We selected 84 BACs identified by at least two adjacent genetic markers, according to the genetic linkage map , as seed BACs for complete sequencing. Overall, 114 Guangluai 4 BACs from the chromosome 4 with an average overlap of 21.9% have been sequenced. The total length of the sequenced Guangluai 4 BAC clones is similar8.4 Mb. All of the sequenced Guangluai 4 BACs and the corresponding markers are listed in a table (at ). Construction of a Fine Chromosome 4 Physical Map of the Rice Subspecies japonica Nipponbare | During the course of this study, the Clemson University Genomics Institute (CUGI) made the fingerprint data of two Nipponbare BAC libraries (20X coverage) publicly available (by FTP at ) , as well as assembled contigs by a fingerprinting analysis and 127,459 BAC-end sequences ( ) . Integration of the fingerprinted contig map with the genetic map and the Monsanto working draft data resulted in a comprehensive physical map of the entire rice genome with an estimated genome coverage of 89% . The chromosome 4 map had a coverage of 87.7% with 23 gaps . As part of the IRGSP effort to sequence the rice genome, we have made an effort to construct a sequence-ready Nipponbare chromosome 4 physical map. Because O. sativa indica and japonica are closely related subspecies, a synteny is expected to be very high . Therefore, we could use genetically anchored and fully sequenced indica BACs to search the Nipponbare BAC-end sequences for locating BAC contigs on its chromosome 4. To implement this strategy, we first assessed the sequence synteny of the two subspecies. A sequenced OSIGBAC-H0102C09, anchored by marker R2502, was used to search the Nipponbare BAC-end databases. A BAC clone OSJNBa0068L06 was subsequently identified on the basis of its end sequence identity to OSIGBAC-H0102C09 and then completely sequenced. We performed sequence alignment analysis between them. The result showed that they are almost collinear and identical except for four insertions or deletions (indels) of three miniature inverted transposable elements (MITEs) and one retrotransposon (Han, Q. Feng, Y.J. Zhang, Q. Zhao, et al., in prep.). The sequence identity between their collinear regions is about 99%. We also compared other chromosomal regions and found that the major differences between the two subspecies were caused by retrotransposons and transposons (Han, Q. Feng, Y.J. Zhang, Q. Zhao, et al., in prep.). These repetitive sequences could be identified and marked from unique sequences, which indicates that the large segments of sequences that contain continued unique sequences from one of the subspecies could be used confidently to locate the sequences of collinear regions from another subspecies. This enabled us to use the known indica sequence scaffolds to rapidly and accurately construct a fine physical map of the Nipponbare chromosome 4 through the integration of the fingerprinted data from the CUGI. Conversely, the japonica sequence scaffolds could also used for anchoring the BACs from the Guangluai 4 libraries by searching their BAC-end sequences. This approach has at least three advantages. First, the large number of the Nipponbare BAC contigs will be anchored rapidly on the chromosome 4 by use of the fully sequenced indica chromosome 4 BACs. Second, the integration of large segments of indica sequences with the Nipponbare BAC ends and fingerprinted BAC contigs will produce a minimum tiling path of the BACs for complete sequencing of the chro mosome 4. Third, the localization of identified BAC contigs will be unambiguous. To further cover the chromosome, a total of 182 genetic and 407 EST markers of the chromosome 4 from Nipponbare was also used for the physical mapping by an in silico anchoring . Anchored contigs were extended by chromosome walking through a re peated colony hybridization, as well as the in silico approach. Thus far, we have integrated 114 sequenced seed BACs of Guangluai 4 with the Nipponbare fingerprinted BAC-contig data, the BAC-end sequences, and the genome physical map to construct the chromosome 4 physical map ready for BAC clone selection for its complete sequencing. Taken together, 27 Nipponbare BAC contigs from the CUGI database were anchored in silico. Twenty contigs were initially anchored by the Guangluai 4 sequence scaffolds. Comparison between the Guangluai 4 sequence scaffolds and the Nipponbare BAC ends enabled us to quickly make a large minimum-tiling path of BACs from the large sets of candidate clones. Locations of the selected Nipponbare BACs were unambiguously confirmed by combining the information of the fingerprinted contigs and the sequence identity between the Nipponbare BAC ends and the Guangluai 4 BAC sequence scaffolds . The sequenced RFLP and EST markers of Nipponbare were also directly searched against the Nipponbare BAC-end database and 23 contigs identified by 25 RFLP and 93 EST markers with five contigs not previously known on the chromosome 4. The five contigs were anchored by the RFLP markers with three of them also anchored by the EST markers. Two more contigs (contigs 273 and 403) were identified through contig extension by chromosome walking. The 27 contigs were eventually merged into 11 (designated as CTG4 --1 to CTG4 --11) after the identification of several bridging clones. The location and length of the contigs and gaps are shown in Figure and Table together with the original CUGI's contigs. The detailed map of CTG4 --1 as an example contig from the whole-chromosome map is shown in Figure . It shows the integrated data used for anchoring and joining the CUGI's fingerprinted contigs in this merged larger contig. The contigs were merged with the following three approaches . First, the contigs were connected by contig-end sequence walk. Therefore, five bridging clones were identified, allowing a merge of seven contigs. Second, two Nipponbare contigs were merged by use of Guangluai 4 BACs. CTG4 --8 originally consisted of three contigs (418, 88, and 89); the gap between contigs 418 and 88 was covered by a fully sequenced OSIGBAC-B0812A04 based on the overlaps with the sequenced Nipponbare BAC clones OSJNBa0072F16 and OSJNBa0060P14. Initially, we were unable to identify a Nipponbare BAC clone to bridge the two clones because no BAC-end sequence hits were detected in the CUGI and Monsanto BAC-end databases. Third, 11 gaps between the CUGI's contigs were joined after the sequencing analysis. Contigs 88 and 89 were finally merged after the completion of the two adjacent BAC sequences from both contigs (see the map of CTG4 --8, ). The bridging BACs are described in Table . They were not previously assembled in the contigs because of either poor quality of the fingerprinting data or little overlap with the connected clones. The relationship between the fully sequenced BACs of the Guangluai 4 and the anchored Nipponbare contigs can be accessed at . In addition, the sequence scaffolds from the two subspecies were used for extending contigs by searching against the BAC-end sequences from either the Guangluai 4 or Nipponbare variety by an iterative approach. In summary, the total length of the 11 BAC contigs was estimated to be similar34.5 Mb (Fig. ; Table ). Therefore, on the basis of an estimated 36.8 Mb of the rice chromosome 4 , the physical map constructed here has a 94% chromosomal coverage and is the most covered chromosome 4 map obtained so far. Figure 1 | Comparison between the genetic and physical maps of the rice chromosome 4. Comparison between the genetic and physical maps of the rice chromosome 4. (A) A genetic linkage map. Genetic markers (GM), genetic distance (cM), and centromere (in yellow) are indicated. (B) A physical map. The portions of the 11 BAC contigs covering 34.5 Mb of the chromosome are in orange and the gaps in plain. The physical distance is indicated (Mb). The maps also show a discrepancy of two markers R2502 (3.1 cM) and R2373 (0 cM) between the genetic and physical maps revealed by DNA sequencing and fluorescence in situ hybridization (FISH) analysis (cross lines). Table 1 | Status of the Rice Chromosome 4 Map Figure 2 | The BAC contig map of CTG4-1 near the short arm telomere together with its collinear Guangluai 4 region. The BAC contig map of CTG4-1 near the short arm telomere together with its collinear Guangluai 4 region. (A) Markers associated with the map. Fifteen EST (in green) and 11 RFLP (in blue) markers used to anchor contigs are shown together with their genetic positions (numbers in cM). (B) Anchored Guangluai 4 BAC contigs by the markers. Fully sequenced BACs (red lines), BACs identified using the Nipponbare sequence scaffold against the Guangluai 4 BAC ends (orange lines), and other clones (black lines) in the contigs are indicated. (C) Fifteen tiled and completed sequenced Nipponbare BACs. Three bridging clones OSJNBa0094015 and OSJNBb0004G23 for joining CUGI contigs 297 and 78 and OSJNBa0054H10 for CUGI contigs 78 and 79 are shown. (D) Display of three original fingerprinted CUGI contigs 297, 78, and 79. CTG4-1 is composed of 295 fingerprinted CUGI BACs and three extended bridge clones. Contig 79 was anchored by marker Y1065L by in silico anchoring. Contigs 297 and 78 were anchored by the Guangluai 4 sequence scaffolds as indicated in syntenic positions. Confirmation of the Chromosome 4 Contigs by Fluorescence In Situ Hybridization (FISH) and Cytogenetic Analysis | To validate the contig relationship of the physical map, we conducted cytogenetic analysis with BAC clones within the assigned contigs. Eight CUGI's contigs (297, 78, and 79 covered by CTG4 --1; 404 by CTG4 --4; 83 and 421 by CTG4 --5; 257 by CTG4 --6; and 59 by CTG4 --11) were initially confirmed on their chromosome locations by FISH of meiotic pachytene chromosomes . The rest of six merged contigs (CTG4 --2, 3, 7, 8, 9, and 10) were eventually identified (their locations on the chromosome 4) by FISH analysis (data not shown). OSJNBb0085F13 from CTG4 --1 and OSJNBb0020J19 from CTG4 --11 were mapped to distal regions close to the telomeres of the short and long arms, respectively . To further determine its telemeric location, two additional BACs within CTG4 --1 were subject to pachytene FISH analysis . The orientations of the BAC clones showed that the order of the genetic markers R2373 (0 cM) and R2502 (3.1 cM) was misidentified on the genetic map. Their correct order on the physical map was further confirmed by sequence analysis of the marker-containing BACs . The CUGI's contig 83 was about 2 Mb long, representing the largest contig identified by but anchored to 19.9 --20.9 cM by DNA sequences of only three RFLP (E679S, E31045S, and E2466S) and two EST ( and ) markers through in silico anchoring. No sequenced indicaBACs were available at this region either. To verify this location, three BACs (OSJNBa0019J05, OSJNBb0068N06, and OSJNBb0056F09) from CTG83 were used for FISH analysis. The results showed that this contig was located near the pericentromeric region of the chromosome 4, and its location was therefore confirmed independently. All of the FISH results can be accessed at . The centromere of the chromosome 4 was physically mapped with a rice centromere-specific satellite DNA pRCS2 as a probe . DAPI staining of the chromosome 4 at the pachytene stage showed that the entire short arm and the pericentromeric region of the long arm are rich in heterochromatin . This region represented about 30% of the chromosome when examined visually . By calculating 50 chromosome 4 pachytene spreads, we measured the length ratio of the long arm versus the short arm to be 5.13 (data not shown) , but this ratio was measured to be 2.9 according to the physical map (Fig. B). Taken together, the results showed that the heterochromatic region of the chromosome 4 appears highly condensed at the pachytene stage and could be larger than that estimated by DAPI staining. It likely occupied 45% of the chromosome 4 as calculated by the physical map. Figure 3 | BAC-FISH mapping of a pachytene chromosome 4. BAC-FISH mapping of a pachytene chromosome 4. FISH of OSJNBb0085F13 (red signal) and OSJNBb0020J19 (green signal) revealed that they are near the telomeres of the short and long arms, respectively. The centromere of the chromosome 4 was detected by a rice centromere-specific satellite DNA pRCS2 as described previously and indicated in yellow. The short arm and the pericentromeric region of the long arm are deeply stained by DAPI and therefore represent the heterochromatin-rich regions, which occupied almost one-third of the chromosome when visually measured. Figure 4 | FISH analysis of CTG4-1. FISH analysis of CTG4-1. Three BAC clones, OSJNBa0068L06 (red signal), OSJNBb0085F13 (green signal), and OSJNBa0042N22 (pink signal), within CTG4-1 were chosen for FISH analysis to confirm the contig localization and orientation. RFLP markers R2502 (3.1 cM) and R2373 (0 cM) matched OSJNBa0068L06 and OSJNBb0004G23, respectively. Because OSJNBb0004G23 and OSJNBb0085F13 are overlapped, the order of the two markers was confirmed on the physical map by FISH analysis. Therefore, their genetic order is reversed on the physical map. Relationship between Genetic Recombination Frequency and Physical Distance on Chromosome 4 | To reveal the relationship between the genetic and physical distances, we calculated the genetic distance (cM) per megabase and found a significant variation along the chromosome . For example, CUGI's CTG83 covered by CTG4 --5 has been estimated to be 2 Mb long in physical distance, but its genetic distance is only 1 cM with a ratio of 0.5 cM/Mb, whereas a ratio of 5.4 cM per Mb is found for CUGI's CTG257 covered by CTG4 --6. The average ratio of the euchromatic region was estimated to be 4.8 cM/Mb and that of the heterochromatic region to be 2.3 cM/Mb . A substantial reduction of recombination was found to occur in the regions of the short arm and the pericentromeric region in agreement with the DAPI staining result . Figure 5 | A relationship between the genetic recombination frequency and the physical distance of the rice chromosome 4. A relationship between the genetic recombination frequency and the physical distance of the rice chromosome 4. X axis shows the physical distance in cM along the chromosome 4. Y axis indicates the ratio of the genetic distance versus the physical distance (cM/Mb). The fine BAC physical map of the Nipponbare chromosome 4 can be accessed at along with the integrated genome map of the CUGI at . The BACs and BAC contigs of indica that have been sequenced or identified with the sequenced ends are shown with the map. In summary, we rapidly constructed a fine physical map of rice O. sativa japonica Nipponbare chromosome 4 through the data integration. Eleven BAC-based contigs covered 94% of the entire chromosome, including euchromatin and heterochromatin. The centromere was covered by the contig CTG4 --5. Therefore, the centromere region was cloned and subsequently sequenced. The total length of the remaining gaps was estimated to be 2.5 Mb. We also identified that the chromosome 4 is a highly heterochromatic one by a detailed cytogenetic analysis. Therefore, a combined approach including iterative STC and hybridization and cytogenetic methods are required for its physical mapping. Because of highly repetitive sequences in chromosome 4, sequence scaffolds from related subspecies are very useful to efficiently identify the collinear clones by searching the clone-end sequences. Until now, 222 BAC clones representing 80% of the Nipponbare chromosome 4 had been completely sequenced and deposited into the EMBL database (accession nos. can be accessed at ). The comparison of the available RFLP and EST marker sequences with the sequenced clones fully supported that the physical map constructed in this study is an accurate one. To fill as many as possible remaining gaps in the Nipponbare map, we are currently trying to identify the clones from additional genomic libraries such as P1-derived artificial chromosome (PAC) libraries constructed by the RGP. In addition, other approaches including single chromosome isolation with optical tweezers are also being tested. To complete the Guangluai 4 chromosome 4 gap filling, two deep-coverage BAC and PAC genomic libraries have been constructed. The two maps are being constantly refined and can be accessed at . Because of its high-resolution representation of the rice O. sativa japonica chromosome 4, the map is of high value for studies devoted to comparative genome analysis between the two subspecies, as well as their genome organization and function . Our comparative physical mapping will also help understand hybrid heterosis between indica and japonica subspecies. In conclusion, the strategy described will be effective for constructing a physical map of the entire genomes between closely related subspecies or species in other plants, as well as in mammals. Although some of the genome sizes of the closely related species vary greatly, the gene order and content and other single copy DNA sequences are to be very conserved along the chromosomes. The genomic sequence scaffolds from one species are therefore to be efficiently used for constructing a physical map of another genome of closely related species. Similarly, this approach can also be applied in building up super contigs with draft sequences of one variety of rice according to a complete genome sequence or genome sequence scaffold from another variety of rice. It is obvious that once the reference genome sequence is completed, the genome research on closely related species will be much more effective. To our knowledge, this is the first report on the comparative physical mapping of the two major cultivated rice subspecies chromosomes. METHODS : Genetic and EST Markers Used for Colony Hybridization and In Silico Anchoring | A total of 182 RFLP markers of Nipponbare cloned in pBSSK or pT7T3 --18U or pCRII were obtained from the Rice Genome Program (RGP) in Japan . Another 12 RFLP markers (RZ602, RZ656, RG788, RG476, CDO539X, RG161, RG169, RZ590, RZ250, RZ819, RG620, and CDO36) were provided by Susan McCouch at Cornell University. A total of 407 EST markers of Nipponbare were provided by RGP . All of the genetic and EST markers known on chromosome 4 were searched against the database of BAC-end sequences to anchor the contigs. In Silico mapping was performed in BLASTN, searching with a stringent cutoff of >95% identity against the databases of BAC-end sequences, marker sequences, and BAC-draft sequences by use of proper sequences. BAC Library Construction | An additional BAC library (OSIGBAC-B), constructed in the laboratory of the National Center for Gene Research, was constructed from BamHI partially digested genomic DNA from O. sativassp. indica cultivar Guangluai 4 seedling tissue and cloned into the pBeloBAC11 vector as described previously . Labeling and Hybridization | Radioactive and ECL (Random Prime Labeling and Detection System, version II, Amersham plc) labeling of the RFLP markers, PCR fragments, BAC-end sequences, and plasmids, as well as hybridization procedures, have been described previously and followed according to the protocols of the ECL labeling kit. Fluorescence In Situ Hybridization | Preparation of rice pachytene chromosome spreads on slides and the FISH procedures were performed according to . WEBSITE REFERENCES : ; Rice Genome Research Program (RGP). ; Clemson University Genomics Institute (CUGI) ; National Center for Gene Research. ; The Institute for Genome Research. Backmatter: PMID- 11997341 TI - Transcriptional Regulation of the Stem Cell Leukemia Gene (SCL) --- Comparative Analysis of Five Vertebrate SCL Loci AB - The stem cell leukemia (SCL) gene encodes a bHLH transcription factor with a pivotal role in hematopoiesis and vasculogenesis and a pattern of expression that is highly conserved between mammals and zebrafish. Here we report the isolation and characterization of the zebrafish SCL locus together with the identification of three neighboring genes, IER5, MAP17, and MUPP1. This region spans 68 kb and comprises the longest zebrafish genomic sequence currently available for comparison with mammalian, chicken, and pufferfish sequences. Our data show conserved synteny between zebrafish and mammalian SCL and MAP17 loci, thus suggesting the likely genomic domain necessary for the conserved pattern of SCL expression. Long-range comparative sequence analysis/phylogenetic footprinting was used to identify noncoding conserved sequences representing candidate transcriptional regulatory elements. The SCL promoter/enhancer, exon 1, and the poly(A) region were highly conserved, but no homology to other known mouse SCL enhancers was detected in the zebrafish sequence. A combined homology/structure analysis of the poly(A) region predicted consistent structural features, suggesting a conserved functional role in mRNA regulation. Analysis of the SCL promoter/enhancer revealed five motifs, which were conserved from zebrafish to mammals, and each of which is essential for the appropriate pattern or level of SCL transcription. [The following individuals kindly provided reagents, samples, or unpublished information as indicated in the paper: N. Tanese.] Keywords: Introduction : The SCL gene (also known as Tal1) encodes a basic helix-loop-helix (bHLH) transcription factor with a critical role in hematopoiesis and vasculogenesis. It was identified by virtue of its disruption in T-cell acute leukemia and rearrangements of the SCL locus are perhaps the most frequent molecular pathology associated with this tumor . Targeted mutation of the SCL gene has shown that it is essential for the development of all hemopoietic lineages , and also for normal yolk sac angiogenesis . Ectopic SCL expression in zebrafish embryos specifies hemangioblast development from early mesoderm, with a consequent excessive production of blood and endothelial progenitors, and can also partially rescue endothelial and hemopoietic phenotypes of the cloche mutant . A crucial role for SCL in haemopoietic and endothelial development has also been revealed in an analysis of the in vitro differentiation potential of SCL-/- embryonic stem cells . SCL is normally expressed in hemopoietic cells, endothelium, and within specific regions of the central nervous system (CNS). This pattern of expression is highly conserved throughout vertebrates from mammals to teleost fish . Murine and human SCL expression is tightly regulated and involves two lineage-specific promoters . In addition, a detailed analysis of the chromatin structure of the mouse SCL locus identified a number of DNaseI hypersensitive sites associated with enhancer or silencer activity . More recently, studies using transgenic mice have identified five separate enhancers, which direct reporter gene expression in vivo to endothelium, midbrain, hindbrain/spinal cord, or hemopoietic progenitor cells, all subdomains of the normal SCL expression pattern . We have recently cloned and sequenced the SCL locus from human, mouse, chicken, and pufferfish . Comparative sequence analysis of the human and mouse loci showed that all known regulatory regions are highly conserved, and revealed a number of additional conserved noncoding regions that represent candidate gene regulatory elements . Inclusion of chicken sequences into multiple sequence alignments allowed us to prioritize some of these regions for functional studies, but failed to detect chicken homologs for other known enhancers . The pufferfish SCL locus exhibits considerable genomic compression, and a 10.5-kb region containing the SCL gene and extending to the immediate flanking genes was sufficient to produce appropriate expression in zebrafish embryos . These results suggest that all of the regulatory elements necessary for conserved embryonic expression are present in this construct. However, this approach would not detect regulatory elements necessary for expression in adult tissues and is also limited by the paucity of information on the pattern of SCL expression in pufferfish. Zebrafish represent a powerful model organism for studies of vertebrate development , and zebrafish studies have provided considerable insight into hematopoiesis . Plans for a zebrafish genome sequencing project are well advanced , but little information is available on the utility of zebrafish/mammalian genomic sequence comparisons (for examples of conserved enhancer sequences, see ; ). The pattern of SCL expression is highly conserved between mammals and zebrafish and we therefore reasoned that comparison of the zebrafish SCL locus with the other four available vertebrate SCL loci would be likely to illuminate the transcriptional regulation of the SCL gene. In this work, we describe the characterization of the zebrafish SCL locus and its comparison with the SCL loci from man, mouse, chicken, and pufferfish. RESULTS : Structure of the Zebrafish SCL Gene | As large-scale gene duplications have been shown in zebrafish , it was important to determine that only one SCL gene was present within the zebrafish genome. To address this, genomic DNA from six individual and pooled fish was subjected to Southern analysis. Digestion with four restriction enzymes each resulted in a single band hybridizing to a zebrafish SCL cDNA probe, and a fifth enzyme identified a restriction fragment length polymorphism (data not shown). These results strongly suggest the presence of a single SCL locus within the zebrafish genome, and are consistent with previously reported mapping studies . We next screened a Zebrafish genomic PAC library and identified three positive clones. Southern analysis showed that the SCL gene was centrally situated in one clone (data not shown). This clone was completely sequenced and annotated as described previously . The insert size was 84,489 bp with the start and stop codons of the SCL gene at 41,051 and 47,373 bp, respectively. The boundaries of SCL-coding exons were obtained by aligning the zebrafish cDNA to the genomic sequence. The position of noncoding exons was determined by RT --PCR using total RNA derived from whole adult zebrafish. A forward primer was designed from a putative first exon immediately downstream of a sequence highly homologous to mouse promoter 1a, and a reverse primer was designed from a sequence within the coding region. A single product of 350 bp was generated and sequenced. Sequence analysis revealed a single 5' untranslated exon immediately downstream of the promoter 1a homology region (data not shown). Therefore, as in pufferfish, no exon homologous to mammalian exon 1b was identified. The size of the zebrafish SCL gene from transcriptional start site to poly(A) site was 9.2 kb, compared with 16.5 kb for human, 15.5 kb for mouse, 8.4 kb for chicken, and 4.8 kb for pufferfish . By comparison, the sizes for the vertebrate SCL RNA sequences are 4.7 kb for human, 4.3 kb for mouse, 3.3 kb for chicken, 2.7 kb for pufferfish, and 2.8 kb for zebrafish, with the different lengths being due to variations in the length of the 3' UTR. A total of 20.7% of our zebrafish sequence was masked by RepeatMasker compared with only 5.6% for the pufferfish sequence. Less abundant fish repetitive elements may yet be identified. Nevertheless, our data suggest that a mere lack of repetitive DNA may not account for the full compression of the pufferfish genome. Figure 1 | Structure of the human (H), mouse (M), chicken (C), pufferfish (P), and zebrafish (Z) SCL loci. Structure of the human (H), mouse (M), chicken (C), pufferfish (P), and zebrafish (Z) SCL loci. All loci are drawn to the same scale. Boxes represent exons and arrows indicate gene orientation. The gray shading illustrates the variation in size of the five vertebrate SCL loci. Numbering refers to distances in kilobases. Only the 3' ends of the human, mouse, and chicken SIL genes are shown. Gene Content of the Zebrafish SCL PAC Clone | Three new genes flanking zebrafish SCL were identified using exon predictions, BLAST database searches, and where required, RT --PCR to confirm exon/intron structure. At the 5' extent of the SCL PAC, a single exon was predicted (195 --1004 bp) with high homology to human and mouse immediate early response gene IER5, which are also intronless (; Figs. and ). The second gene identified lay 3' to SCL (58,174 --61,906 bp; Figs. and ) and showed homology to human and mouse MAP17 . MAP17 is known to lie downstream of SCL in the human, mouse, and chicken genomes . A putative zebrafish MAP17 cDNA sequence was assembled from the predicted exons and confirmed by RT --PCR, followed by direct sequencing of RT --PCR products. At the 3' extent of the zebrafish contig (68,676 --82,972 bp), we identified the last four exons of a gene highly homologous to the human multi-PDZ protein MUPP1 (; Figs. ,). Conserved gene order between zebrafish and human/mouse therefore does not extend beyond the MAP17 genes (Fig. D). Figure 2 | Newly identified genes flanking zebrafish SCL. Newly identified genes flanking zebrafish SCL. (A) Alignment of predicted zebrafish IER5 protein (Dr_IER5) to the corresponding human (Hs_IER5) and mouse (Mm_IER5) proteins. (B) Alignment of zebrafish MAP17 protein (Dr_MAP17) to the respective human (Hs_MAP17) and mouse (Mm_MAP17) proteins. (C) Alignment of predicted zebrafish MUPP1 protein (Dr_MUPP1) to the respective rat (Rn_MUPP1) and human (Hs_MUPP1) proteins. (D) Diagram indicating the region of conserved gene order in the mouse (M) and zebrafish (Z) SCL loci. Within the zebrafish SCL PAC clone, we failed to identify any homology to either the SIL gene, which is upstream of human, mouse, and chicken SCL , or the PDZ and RING finger genes upstream of pufferfish SCL . This suggests that the region of conserved gene order ends between SCL and the immediate upstream gene. However, the interval between IER5 and SCL is relatively large (similar39 kb), and we therefore carried out a detailed analysis of this region (see Methods), yet found no evidence of additional genes. The expression pattern of SCL is highly conserved from zebrafish to mammals . Our synteny data therefore suggest that the murine genomic domain, which contains all regulatory elements necessary for the conserved pattern of SCL expression, lies within a region extending from the end of SIL to the beginning of CYP4x1 (Fig. D). Comparative Sequence Analysis Identified Three Regions of Noncoding Homology | The availability of five orthologous SCL genomic sequences covering a wide evolutionary range allowed us to explore the potential usefulness of different vertebrate genome sequences for comparative analysis/phylogenetic footprinting. By use of currently available programs (VISTA, BLAST, PipMaker, DiAlign, and Synplot) , we performed pairwise alignments using the mouse sequence as a reference sequence (Fig. A). This analysis demonstrated that mouse/human alignments showed high sequence similarity for all coding exons and all known regulatory regions (Fig. A). Mouse/chicken alignments identified similarity with all coding exons and also five known regulatory regions (promoters 1a and 1b, +1, +3, +23), but failed to identify significant similarity to three other known murine regulatory regions (-4, +18, +19; see Fig. A). Pairwise mouse/pufferfish and mouse/zebrafish alignments failed to identify some coding exons (exon 4 in pufferfish and exon 5 in zebrafish) and found similarity for only two of the eight known regulatory regions in pufferfish (-4 region and promoter 1a), and no significant similarity for any known regulatory region in the zebrafish sequence. Because mouse/zebrafish synteny comprises the SCL and MAP17 genes, our analysis could potentially have revealed MAP17-conserved noncoding sequences. However, as for the SCL locus, no such sequences were identified from mouse/zebrafish comparisons. Moreover, these comparisons failed to identify any of the MAP17 coding exons, even though these were clearly identifiable using the BLASTX program. Our results therefore suggest that the zebrafish genome sequence may be of limited use for the identification of mammalian gene regulatory regions using simple pairwise comparisons. Figure 3 | Comparative sequence analysis of vertebrate SCL loci. Comparative sequence analysis of vertebrate SCL loci. (A) VISTA global alignment plots displaying similarity between the mouse locus (used as the reference sequence) and the human (H), chicken (C), pufferfish (P), and zebrafish (Z) loci. Gray boxes indicate the positions of known mouse regulatory regions. Black and white boxes depict exons with the coding part shaded red. (B) VISTA global alignment plots displaying sequence similarity of chicken SCL locus (used as the reference sequence) and pufferfish (P) and zebrafish (Z) SCL loci. (C) VISTA global alignment plot displaying sequence similarity between the pufferfish locus (used as the reference sequence) and the zebrafish (Z) SCL locus. VISTA plots depicted in A to C were calculated using a 50-bp window and displayed using a 25% lower cut-off limit. Pairwise comparisons of pufferfish and zebrafish with chicken sequences revealed higher similarity than the comparisons with the mouse sequences described above (Fig. B). Only one coding exon was not identified (exon 4 in pufferfish), and high sequence similarity was detected for the promoter 1a region in all three species. This sequence similarity extended to a region near the 3' end of the first noncoding exon. Moreover, pairwise comparisons of the pufferfish with the zebrafish genomic sequence clearly identified all coding sequences. Furthermore, high similarity was seen for the promoter 1a and exon 1 regions as well as a region just upstream of the poly(A) site (Fig. C). This analysis showed that pairwise comparisons of two teleost genomic sequences can identify noncoding conserved sequences missed by mammalian/teleost comparisons. Moreover, multiple sequence alignments of the promoter 1a, exon1, and poly(A) sequences from all five species showed that all three were, in fact, highly conserved between all species. This suggests that whereas current pairwise global alignment programs may struggle to identify conserved noncoding sequences between fish and mammalian sequences, inclusion of additional genomic sequences can be instrumental in identifying meaningful regions of homology. The Polyadenylation Sequence Lies Within a Highly Conserved Region | A 200-bp stretch of sequence surrounding the AAUAAA polyadenylation sequence showed conservation across all five species (Fig. A). The GU- or U-rich sequence usually found downstream of the AAUAAA poly(A) signal could not be identified. In genes lacking this sequence, efficient polyadenylation has been shown to be dependent on an AU-rich region upstream of the poly(A) signal . As can be seen from the alignment in Figure A, all five SCL sequences share an AU-rich motif upstream of the poly(A) signal. Figure 4 | : Conserved noncoding sequences from vertebrate SCL loci. : Conserved noncoding sequences from vertebrate SCL loci. (A) Alignment of the conserved sequences upstream of the poly(A) signal. The two predicted stem-loop structures, the AU-rich region and the poly(A) signal are indicated. (B) Alignment of the conserved sequences in exon 1. The conserved YY1 site is indicated. (C) Alignment of the conserved SCL promoter/enhancer sequences. The three motifs studied previously (SKN1 and two GATA sites) and the two newly identified conserved sequences (CS1 and CS2) are indicated. The vertical arrow indicates the transcriptional start site. The conserved 3' UTR sequences are likely to reflect conservation of a secondary structure necessary for interaction with RNA-binding proteins. However, secondary structure predictions for an individual RNA sequence are often unreliable, as potential structures may have similar minimum energy (DeltaG) values. However, prediction accuracy can be greatly improved by taking comparative sequence data into consideration . Therefore, we used a set of algorithms that incorporates evolutionary history into the process of predicting conserved structures , and thereby allows identification of conserved regions with highly significant secondary structure predictions. By use of this approach, a 50-bp stretch (position 32 to 86, Fig. A) was predicted to form two stem-loop structures in all five sequences (see Figs. and ), suggesting that this structure is important for some conserved aspect of SCL RNA regulation. Interestingly, the first loop is made up entirely of U residues and the second stem loop contains a bulge structure composed of a perfect palindrome (GGUACC). Both of these motifs have been shown to interact with RNA-binding proteins . Figure 5 | Predicted conserved stem-loop structures for the human (H), mouse (M), chicken (C), pufferfish (P), and zebrafish (Z) poly(A) regions. Predicted conserved stem-loop structures for the human (H), mouse (M), chicken (C), pufferfish (P), and zebrafish (Z) poly(A) regions. (A) Stem loop 1 from Fig. . (B) Stem loop 2 from Fig. . Sequence Conservation in Exon1 | Within the first noncoding exon, 58/74 nucleotides (78%) were conserved between pufferfish and zebrafish, and within this sequence, two blocks of 19 and 15 nucleotides were identical (Fig. B). Furthermore, an 18-bp stretch of sequence was conserved across all five species and contained a consensus-binding motif for the transcription fac-tor YY1 (5'-AANATGGC-3') . YY1 is a ubiquitously expressed zinc finger transcription factor belonging to the GL1-Kruppel family members , suggesting that this region may act as a transcriptional enhancer on the nearby promoter 1a. However, because the sequence conservation is within a transcribed region (i.e., the 5' UTR), our data do not exclude the possibility that the observed sequence conservation reflects the recognition sequence for an RNA rather than a DNA-binding protein, and thus indicates a possible conserved post-transcriptional function for this region. Sequence Conservation Immediately Upstream of Exon 1a | A 170-bp region upstream of SCL exon 1 in zebrafish (38,869 bp --39,024 bp) was highly homologous to regions upstream of the first exons of human, mouse, chicken, and pufferfish SCL genes (Fig. C). This region has promoter activity in hemopoietic cell lines and also contains a midbrain enhancer . A high level of human/mouse homology is seen throughout this region, but with inclusion of pufferfish, zebrafish, and chicken sequences, five blocks were defined that were conserved across all five species. Three of the conserved blocks, two GATA sites, and a putative SKN1 site have been shown to be important for promoter activity in hematopoietic cell lines or for activity of the midbrain enhancer in transgenic mice . Functionally significant regulatory elements have therefore been conserved within this SCL promoter/enhancer from fish to man and can be identified by phylogenetic footprinting. The remaining two conserved sequences (CS1 and CS2) were unrecognized previously and represent candidate binding sites for additional transcription factors. Functional Analysis of the CS1 and CS2 Motifs in Transgenic Mice | The functional significance of CS1 and CS2 was investigated in transgenic mice. Each site was mutated in the context of the --10E3/lacZ/3'enh reporter construct, which contains mouse genomic sequences corresponding to both SCL promoters (1a and 1b) together with four SCL enhancers responsible for directing expression to endothelium, midbrain, hindbrain/spinal cord, and hematopoietic progenitors . Following oocyte microinjection, embryos were examined at day 11.5 post coitum (E11.5), at a time when endogenous SCL is expressed in yolk sac hemopoietic progenitors, fetal liver, brain, and endothelium . The expression pattern of the --10E3/lacZ/3'enh construct was consistent with the known expression patterns of the --10E3/lacZ transgene and transgenes containing the 3' enhancer . Strong expression of beta-galactosidase was evident in the midbrain and endothelium in 2/2 embryos and sections showed lacZ expression in endothelial cells (endocardium, dorsal aorta, intersomitic vessels, capillary networks, and vitelline vessels) as well as in a subset of fetal liver and circulating blood cells (see arrowhead, Fig. B). Mutation of either the CS1 or CS2 site did not alter the pattern of expression. Following mutation of the CS1 site, 2/2 transgenic embryos showed a wild-type pattern of expression in brain, yolk sac, embryonic endothelium, and blood cells (Fig. B). Following mutation of the CS2 site, 3/5 transgenic embryos displayed the same pattern of expression. In the two remaining embryos, staining was seen in brain and fetal liver, but was not evident in endothelium. The lack of detectable endothelial staining in the latter two embryos may reflect position effects that differentially influence the endothelial elements more than the brain regulatory elements. These data therefore show that neither CS1 nor CS2 are necessary for the staining pattern observed with the --10E3/lacZ/3'enh cassette at E11.5. However, our results do not exclude a critical role for these motifs either at other stages of development or in regulating the level of SCL transcription. Figure 6 | Transgenic analysis of the CS1 and CS2 motifs. Transgenic analysis of the CS1 and CS2 motifs. (A) Diagram of the mouse SCL locus with the constructs used for transgenic analysis shown underneath. Vertical arrows indicate previously mapped DNaseI hypersensitive sites. Colored boxes specify enhancer regions mapped previously in transgenic mice [E = endothelial, M = midbrain, (H) hindbrain/spinal cord, (E/B) endothelial/blood]. (B) beta-galactosidase activity of constructs shown in A in E11.5 transgenic mouse embryos. A whole-mount view of an E11.5 embryo together with sections of fetal liver (top) and dorsal aorta (bottom) are shown for each construct. Arrowheads in sections indicate lacZ-positive hematopoietic cells. (B) Brain; (H) heart; (L) fetal liver. CS1 and CS2 Are Both Necessary for Full SCL Promoter Activity in Erythroid Cells | To investigate whether CS1 or CS2 are important for the level of SCL transcription, we analyzed the effect of corresponding mutations on the activity of SCL promoter 1a in a transient transfection assay. Mutation at CS1 or CS2 reduced promoter activity to 67% and 63%, respectively in the erythroid cell line, J2E (Fig. A). This degree of reduction is comparable with that seen with a mutation in the --69GATA site . Our results indicate that both sites contribute to the level of SCL transcription in erythroid cells. Figure 7 | Functional analysis of the CS1 and CS2 motifs in the J2E erythroid cell line. Functional analysis of the CS1 and CS2 motifs in the J2E erythroid cell line. (A) Relative luciferase activity of a wild-type mouse reporter construct containing 2 kb upstream of exon 1a (-2E1aLuc) compared with a CS1 mutant (-2E1aLucCS1), CS2 mutant (-2E1aLucCS2), and promoterless control construct (pGL2Basic). Results represent the mean of four independent experiments (a minimum of 10 luciferase values for each construct) using two different DNA preparations. (B) Gel-shift analysis identifies two specific complexes (arrowheads) binding to the CS1 motif --- see text for details. (Comp) Competitor oligonucleotides; (NE) nuclear extract; (NFI) antibody to NFI; (NS) nonspecific control antibody . (C) Gel-shift analysis using a rapid protocol for nuclear extract preparation identifies a high molecular weight complex (arrowhead) binding to the CS2 site. (Comp) Competitor oligonucleotides; (NE) nuclear extract. Details of oligos used in B and C can be found in the Methods section. Band shift analysis was performed to identify protein complexes binding at CS1 and CS2 in J2E cells (Fig. B,C). An oligonucleotide was used that spanned the region of promoter 1a containing both CS1 and CS2. Two specific complexes (arrowheads, Fig. B) bound the wild-type oligonucleotide (see Fig. B, lane 2). Binding of both complexes was competed by an excess of unlabeled wild-type (see Fig. B, lane 3) or CS2 mutant (see Fig. B, lane 4) competitor oligonucleotides but not by an oligonucleotide with a mutation at CS1 (see Fig. B, lane 5), an oligonucleotide with mutations at both sites (see Fig. B, lane 6), or an oligonucleotide of unrelated sequence (see Fig. B, lane 10). These results show that both complexes bound specifically to the CS1 site and that binding of both was prevented by the mutation of this motif. CS1 contains an NFI (Nuclear Factor I) consensus half-site and a variant C/EBP site. However, the complex could neither be competed by an oligonucleotide containing a NFI consensus site known to bind all NFI family members (see Fig. B, lane 8, mutant NFI oligo control, see Fig. B, lane 9), nor supershifted using an NFI antiserum (see Fig. B, lane 11). Similarly, no competition was seen using an oligonucleotide carrying the CAAT consensus sequence (see Fig. B, lane 7) known to bind C/EBP alpha, beta, and delta . These results therefore suggest that the complexes binding to CS1 do not contain NFI or C/EBP family members. CS2 is a 7-nucleotide sequence precisely conserved across all five species (Fig. C). The TAAT sequence represents the core-binding motif of many homeobox transcription factors , and is used preferentially by members of the Ultrabithorax, Antennapedia, Sex combs reduced, and Deformed families of proteins , together with their mammalian homologs . However, the extended CS2 sequence (5'-ATAATGC-3') has not been reported to bind any specific vertebrate homeobox protein. No complexes could be seen binding to CS2 using standard nuclear extracts and an oligonucleotide containing both CS1 and CS2 sites. An oligonucleotide was therefore designed with CS2 centrally positioned, and nuclear extracts were prepared using a modified rapid protocol (see Methods) to minimize protein degradation. Using this approach, a specific low mobility complex was detected that was competed by wild-type but not CS2 mutant oligos (see Fig. C, lanes 2 --4). Given the large number of possible candidate proteins that might bind to the TAAT core of CS2, further studies will be needed to identify the nature of the protein(s) present in this complex. DISCUSSION : A Genomic Domain Necessary for the Conserved Pattern of SCL Transcription | Chromosomal rearrangements occuring during evolution may shed light on the location of transcriptional elements necessary to produce conserved patterns of gene expression. We have previously characterized and sequenced the SCL genomic loci from the pufferfish, fugu rubripes and have found that the genes immediately flanking pufferfish SCL were unrelated to those known to flank both avian and mammalian SCL genes. These results imply that SCL regulatory elements might be confined to the region between the upstream and downstream flanking genes, a region of 65 kb in human and 8.5 kb in pufferfish. Consistent with this concept, a 10.4-kb fragment of pufferfish genomic DNA, containing the SCL gene and extending to the 5' and 3' flanking genes, directed appropriate expression to hematopoietic and neural tissue in transgenic zebrafish embryos . However, three caveats need to be borne in mind when considering the pufferfish synteny data. Firstly, expression of the pufferfish transgene was assessed at a single time point in zebrafish embryos. The transgene may therefore lack elements necessary for expression in adult fish or at other developmental timepoints. Secondly, compression of the pufferfish genome may have been accompanied by a simplification of gene regulatory mechanisms , which, if true, would limit the ability of pufferfish/mammalian sequence comparisons to shed light on mammalian gene regulation. Thirdly, the pattern of SCL expression in pufferfish has been assessed by RT --PCR and not by in situ hybridization. It is therefore possible that the expression pattern of pufferfish SCL may differ from other vertebrates, possibly as a consequence of the compact pufferfish genome. In contrast, the zebrafish genome is much less compressed (1800 Mbp compared with 400 Mbp), and the pattern of SCL expression as assessed by in situ techniques is highly conserved between zebrafish and mammals, both during embryogenesis and, where it has been assessed, also in adults. In both zebrafish and mouse, SCL is expressed in specific cells within the anterior horns of the spinal cord, in the ventral midbrain, in the hindbrain, in progenitors of blood and endothelium, and in erythroid cells . The results presented here show the existence of a region of synteny conserved between zebrafish and mouse extending from the 3' end of the gene immediately upstream of SCL to the beginning of the gene immediately downstream of MAP17. The comparative synteny data presented here, therefore, raise the possibility that additional SCL regulatory elements may be found between the beginning of the MAP17 gene and the start of the gene immediately downstream, a region known to contain several peaks of human/mouse sequence homology , but which has not yet been assessed for the presence of SCL regulatory elements. However, it is important to bear in mind that comparative mapping studies of zebrafish and human genes have shown that gene order within syntenic regions may be inverted or transposed . It therefore remains possible that as yet unsequenced regions neighboring the SCL locus may exhibit conserved synteny with the SCL gene, and could conceivably harbor long-range regulatory elements. Comparative Sequence Analysis of Zebrafish and Mammalian Loci | The zebrafish SCL gene (from upstream to downstream flanking genes) is 59 kb, mainly as a result of the large interval between Dr-IER5 and zebrafish SCL. The zebrafish SCL gene is therefore similar in size to its mouse or human homologs (similar60 and 65 kb, repectively) and seven times larger than the pufferfish gene. Our data show that the mammalian and zebrafish SCL loci all have MAP17 downstream of SCL and that this is therefore likely to represent the ancestral pattern of gene order. Conservation of gene order between mammals and zebrafish does not extend downstream of the MAP17 gene. In contrast, the pufferfish SCL locus does not share immediate 5' or 3' flanking genes with those present at zebrafish or mammalian SCL loci. Moreover, none of the genes present either side of zebrafish SCL flank the pufferfish SCL gene. These results are consistent with the suggestion that gene shuffling may have been especially prevalent during the speciation of bony fish . Most studies comparing mammalian and fish sequences have focused on the pufferfish because of its compressed genome. Conservation of regulatory elements between pufferfish and mammalian homologs has been found in a number of genes. In some, but not all cases, the regulatory elements are functionally equivalent. For example, comparative analysis of pufferfish and mouse hox gene loci identified conserved enhancers and the pufferfish enhancers functioned appropriately in transgenic mice . In contrast, a comparison of the pattern of activity in transgenic mice exhibited by conserved pufferfish and mouse Wnt1 regulatory elements revealed consistent differences . There are also several examples of genes in which sequence comparisons between pufferfish and mammalian homologs have not revealed any noncoding sequence homology, despite the genes having highly conserved expression patterns and functionally homologous regulatory elements . In this situation, it seems likely that current computational methods are not sensitive enough to detect individual conserved transcription factor binding sites in a background of extensive sequence divergence . There has been relatively little comparative analysis of zebrafish regulatory elements. Here we show that comparisons of the zebrafish SCL locus with mammalian SCL loci did not identify zebrafish elements homologous with the majority of known murine enhancers (e.g., enhancers corresponding to -4/3, +3, +17, +18, +19 DNase1 hypersensitive sites). In contrast, all known mouse regulatory regions were highly conserved in human, whereas significant similarity for half of them could be identified in chicken. We suspect that once the transcription factors binding to these enhancers are identified, it will prove possible to detect corresponding conserved motifs within the zebrafish sequence. In any case, our results did reveal three regions of noncoding sequence conservation, two of which are of particular interest. First, the poly(A) region contained a stretch of conserved sequence homology. Conserved stretches of 3' UTR sequence can be identified in 36% or 17% of orthologous genes in pair-wise comparisons between mammals and birds or mammals and fish, respectively . These regions have been implicated in the regulation of mRNA stability and translation . The conserved sequences in the SCL 3' UTR predict the formation of two stem-loop structures, and contain motifs implicated in interactions with RNA-binding proteins . Further studies will be needed to ascertain whether these sequences play a role in the post-transcriptional regulation of SCL expression that has been reported previously . The second conserved region of particular interest was located immediately upstream of the first SCL exon. This region functions as a promoter in hematopoietic cell lines and as a powerful midbrain enhancer . Five short blocks of homology were identified which were conserved between all five vertebrates studied, and each of which corresponded to a transcription factor binding motif. Two of these motifs are important for promoter activity in hematopoietic cells and for the pattern of SCL expression in the brain . The remaining three motifs are necessary for the level of SCL promoter activity in hematopoietic cells (this study; ). These results are consistent with the concept that within modular promoters and enhancers, only a minority of transcription factor motifs influence the pattern of gene expression, whereas the majority will only be detected by quantitative assays . Zebrafish has emerged as one of the most powerful experimental models to study vertebrate development and organogenesis. This has culminated recently in the announcement of plans to sequence its entire genome . We have described previously the genomic loci of human, mouse, chicken, and pufferfish SCL , and are therefore in a unique position to place zebrafish genomic sequence data into a wider comparative context. The results presented here represent the largest segment of zebrafish genomic DNA so far subjected to comparative sequence analysis. Our analysis suggests that comparative analysis of zebrafish and mammalian genomic sequences may be of limited value for the identification of functionally significant noncoding sequences. However, comparison of zebrafish with pufferfish sequences revealed conserved noncoding sequences missed by the zebrafish/mammalian comparisons, emphasizing the potential benefits of having multiple fish genome sequences. Moreover, the zebrafish genome sequence will greatly accelerate the characterization of genes controlling vertebrate organogenesis identified in large-scale mutation screens. METHODS : Isolation and Sequencing of Zebrafish SCL Locus | A gridded zebrafish genomic PAC library (library no. 706) from the RZPD (German Human Genome Project) was screened using a zebrafish SCL cDNA probe. The three positive clones (BUMSP706C0952Q3, BUMSP706B2050Q3, and BUMSP706I22129Q3) were further analyzed by Southern analysis following standard protocols. The insert of clone BUMSP706I22129Q3 was fully sequenced as described . Sequence Analysis | Interactive annotation of genomic sequences was performed within ACeDB as described . Long-range sequence comparisons were performed using DiAlign version 1 , dotter , and PipMaker . Small-scale DNA alignments were performed and displayed as described . RNA secondary structure predictions were performed using the programs described in . Repeat content was analyzed using RepeatMasker. In an attempt to determine whether any exons predicted 5' of zebrafish SCL were, in fact, part of a gene, RT --PCR was performed using total RNA from whole adult zebrafish with multiple primers. Where possible, primers were designed from exons that were predicted by more than one program. Whenever RT --PCR products were obtained, these were sequenced and shown to result from mispriming of one or both of the primers used, and thus contained no spliced exons (data not shown). Transgenic Analysis | Mutations into the CS1 and CS2 motifs were introduced as described . The CS1 mutation converted the wild-type GCCAAAT sequence to GCTAGCT, whereas the CS2 mutation converted ATAATG to ATGGTG. The transgenic reporter constructs contained a 139-kb BglII/XcmI fragment of the mouse SCL locus ranging from 10 kb upstream of the promoter to exon 3 upstream of a lacZ reporter gene, which was followed by a 5.5-kb BglII fragment containing the 3' enhancer . F0 transgenic mouse embryos were prepared and analyzed as described . Reporter Assays and EMSA Analysis | J2E cells were grown as described . Transient transfections and luciferase assays were performed as described previously . For the rapid nuclear extract preparation, 2 x 106 cells were collected by centrifugation, washed in PBS, and resuspended in 400 muL of buffer A (10 mM HEPES at pH 7.9, 10 mM KCl, 0.1 mM EGTA, 1 mM DTT, 0.5 mM PMSF), and incubated at 4C for 15 min. After adding 25 muL of 10% NP40, the sample was mixed and briefly centrifuged. The pellet was resuspended in 25 muL of buffer C (10 mM HEPES at pH 7.9, 0.4 M NaCl, 1 mM EDTA, 1 mM DTT, 1 mM PMSF), and incubated for 30 min followed by a 5-min centrifugation. A total of 1 muL of the resultant supernatant was used per track for EMSA analysis. Standard nuclear extracts were prepared and EMSAs performed as described . Oligonucleotides used in Figure B are as follows: WT, (GAAATTGCCAAATTAAAAT GAATCATTTGGCCCATAATGGCCGA); CS1, (GAAATTGC TAGCTTAAAATGAATCATTTGGCCCATAATGGCCGA); CS2, (GAAATTGCCAAATTAAAATGAATCATTTGGCCC AT GGTGGCCGA); DM, (GAAATTGCTAGCTTAAAATGAAT CATTTGGCCCATGGTGGCCGA); CAAT, (TGCAGATTGCG CAATCTGCA); NFIc (AGGTCTGGCTTTGGGCCAAGAGC CGC); mNFIc, (AGGTCTCGCTTTGGGCCAAGAGCCGC); NR, (AGGCGGCTCCTTATCTCGGC). Oligonucleotides used in Figure C were as follows: WT, (ATCATTTGGCCCATAATG GCCGAGGCGCTTATCGGGGGCG); CS2m, (ATCATTTGGC CCATGGTGGCCGAGGCGCTTATCGGGGGCG). WEB SITE REFERENCES : ; A resource for masking repetitive DNA in genomic sequences. Backmatter: PMID- 11997349 TI - RePS: A Sequence Assembler That Masks Exact Repeats Identified from the Shotgun Data AB - We describe a sequence assembler, RePS (repeat-masked Phrap with scaffolding), that explicitly identifies exact 20mer repeats from the shotgun data and removes them prior to the assembly. The established software Phrap is used to compute meaningful error probabilities for each base. Clone-end-pairing information is used to construct scaffolds that order and orient the contigs. We show with real data for human and rice that reasonable assemblies are possible even at coverages of only 4x to 6x, despite having up to 42.2% in exact repeats. [The following individuals kindly provided reagents, samples, or unpublished information as indicated in the paper: P. Green and A.F. Smit.] Keywords: Introduction : All large-scale genome-sequencing projects to date have used a shotgun strategy in which some target region is oversampled by a random collection of sequence reads of typical 500 bp length. There is a wide variation in the size of the target region. The International Human Genome Sequencing Consortium targeted bacterial artificial chromosome (BAC) clones of size similar150-kb. Celera, however, targeted the entire 3-Gb human genome . Regardless of the size of the target region, the primary difficulty for assembling a shotgun data set is the frequent appearance of repeated motifs. The difficulty is affected by how many repeats there are, how large they are, how similar they are, and how they cluster. All these characteristics are organism specific. The objective is to put the reads together in the correct order and orientation, despite the repeat-induced ambiguities. Software used by the IHGSC included Staden , Phrap (P. Green, unpubl.), and GigAssembler . Phrap pioneered the concept of using base-level error probabilities to help distinguish nearly identical but distinct repeats from identical repeats that differ because of a sequencing error. This was effective because many of the troublesome repeats were derived from transposon insertions that diverged over evolutionary time and therefore were nearly identical but distinct. Distinction of transposon repeats was not scalable to larger data sets because the explosion in the number of putative overlaps consumed an intolerable amount of computer time. Even so, Phrap's ability to compute a meaningful error probability for each base has been instrumental in the IHGSC's efforts to establish a data quality standard of 1 error per 10,000 bases. The Celera assembler tamed the overlap explosion problem by masking all known repeat classes. To further reduce the number of false joins, it estimated the likelihood that an overlap was unique before joining any two sequences together . This procedure resulted in a fragmentary assembly, but because Celera sequenced both ends of the shotgun-library clones, masked repeats could be inserted back into the assembly, guided by the clone-end-pairing information, as long as both ends were not masked. The clone-end-pairing information allowed them to bridge across many of the remaining gaps, due either to repeat masking or to missing sequence . Because their clone-insert sizes were so tightly controlled (e.g., 2 kb with +-10% variance), they could also estimate the sizes of the bridged gaps. METHODS : We have combined all the hard-earned lessons of the past into a single software package, RePS (repeat-masked Phrap with scaffolding). Rather than reinvent the wheel, we used Phrap to handle the detailed sequence assembly, preserving its ability to compute a meaningful base-level error probability. As the critical pre-Phrap process, we explicitly identify exactly repeated 20mers and mask them out (i.e., remove them from consideration by Phrap). This eliminates the overlap explosion problem. At the same time, it also minimizes the likelihood of making a false join. As a post-Phrap process, we analyze the clone-end-pairing information to fill gaps due to repeat masking and construct scaffolds across any other gaps. In essence, we borrowed concepts used by the Celera assembler to create a scalable version of Phrap for whole-genome shotgun assembly. RePS is available by contacting the authors at . There are advantages to explicitly identifying exact 20mer repeats without regard to their underlying biological context. Mathematically defined repeats (MDRs) are more useful than biologically defined repeats. From an algorithmic perspective, it does not matter if the sequence is a transposon, a microsatellite, or a gene duplication. If it is repeated, it will hinder the assembly process. To the extent that the repeat databases are incomplete, it should be more reliable to identify MDRs from the shotgun data set itself. This will be an increasingly important issue as the large sequencing centers move to less well-characterized genomes. Our software was tested on a data set of genuine sequence reads, taken from an 11.9-Mb region of human chromosome 3, which was finished to the standards set by the IHGSC with a BAC-by-BAC approach. Because the key analysis steps were effectively decoupled, we could explore tradeoffs between different aspects of data quality and how they might be affected by different experimental parameters. The results were used to guide our assembly of the 430-Mb rice shotgun sequence , for which RePS was originally created. [Note: As we were preparing this paper, a similar algorithm was submitted for publication and has been published . A comparison would have been interesting, but it was not practical because the two programs were optimized for different multiprocessor supercomputers and not readily ported.] We adopt a purely mathematical definition of repeats. Any 20mer that is exactly duplicated in the target region is a repeat. The Nmer unit cannot be too small because, at some point, every Nmer is repeated. The number of different 20mers is 420 ~ 1012, which is larger than any genome that we might reasonably try to assemble. However, it does not help to make the Nmer unit much larger than the minimum detectable overlap, which is 14 to 26 bp, based on our Phrap minmatch and minscore settings. Figure depicts the two primary components, repeat-masked Phrap and clone-end-pairing analysis. The latter is divided into repeat-gap closure and scaffold construction. All of the sequence joins are made with Phrap. Repeat masking is used only to prevent Phrap from making a false join. Clone-end pairs are used only to tell Phrap when an otherwise ambiguous join can safely be made. By letting Phrap handle all the details at the base level, we preserve its ability to compute a meaningful base-level error probability. Figure 1 | The RePS algorithm. The RePS algorithm. Any 20mer that appears in the shotgun data set more often than a threshold depth is likely to be an exact repeat and is therefore masked out. Some sequence reads end up fully masked, but most have enough unique sequence in them to be used by Phrap. Repeat gaps are those for which the gap sequence is in the reads, but masked out by our procedure. LW gaps are those for which the gap sequence is not in the reads for statistical reasons (i.e., Lander-Waterman). Clone-end-pairing information is employed to help close the smaller repeat gaps. Large repeat gaps cannot be closed in this manner. Neither can LW gaps. But as long as the clone-insert sizes are larger than the remaining gaps, there is a reasonable probability that we can build scaffolds to bridge over the gaps and order and orient the contigs. Repeat-Masked Phrap | Let C be the coverage, or the number of times, that a genome is represented in the shotgun data set. The number of times that any 20mer appears in the shotgun data set is its depth D. Consider a 20mer with copy number N across the genome. It should have an average depth of N x C. For masking purposes, we define repeats as 20mers with a depth that is greater than some preset threshold. There are tradeoffs between false-positives (unique 20mers incorrectly called repeated) and false-negatives (repeated 20mers incorrectly called unique). False-positives result in excessive masking and smaller contigs. False-negatives are difficult to avoid for low-copy repeats, but the potential for misassemblies is not as serious as it might appear. Expected overlaps are 500 bp divided by coverage, or 125 bp at 4x coverage. To result in a misassembly, the low-copy repeat must be exactly duplicated across the entire 125 bp, which is unlikely but not impossible. In the end, one must test the algorithm on genomes with substantial repeat fractions, like the human and rice genomes, to assess the severity of this problem. Masking the 20mer repeats serves two purposes simultaneously. The first is that it liberates Phrap from having to decide among an exponential number of possible joins, and so the algorithm runs much faster. The second is that it prevents Phrap from making ambiguous joins with a high probability of being incorrect. On the other hand, Phrap does not assemble the sequence in a masked region, let alone compute an error probability. To recover this information, we use a local reassembly. After the initial Phrap assembly, all repeats are unmasked and every contig is Phrap-ed again. Using a 100-bp sliding window, we search for discrepancies. Wherever we find them, we extract all sequence reads that fall within the window, Phrap them again, and replace the 100 bp of contig sequence with this local reassembly. In essence, the initial assembly puts the reads into more or less the right place, while the local reassembly recovers any masked sequences and establishes a Phrap quality for each base. From an implementation perspective, the existing version of Phrap is constructed for a single-processor environment. To make it work in our multiprocessor environment, we first do a pairwise comparison of all the reads using BLAST to cluster any reads that have even a remote chance of being joined. The clusters can then be distributed among as many processors as desired and assembled independently using the single-processor version of Phrap. Clone-End-Pairing Analysis | The clone-end-pairing analysis examines the names of the sequence reads that are already assembled into a Phrap contig, considers the sizes of the clone inserts, and on that combined basis, existing contigs are validated and gaps between contigs are closed. There are two different types of gaps, repeat gaps and LW gaps, which are depicted in Figure . In a repeat gap, the sequence is in the shotgun data, but it has been masked out. If the gap is small, the existing contigs may already overlap, and all we need is clone-end evidence that they can be joined. Even if the gap is bridged by fully masked sequence reads, it can be filled in if the opposite clone ends are not fully masked. In contrast, for an LW gap, the required sequence is not even in the shotgun data, due to sampling statistics . In practice, LW gaps are usually smaller than a read. Regardless of the nature of the remaining gaps, as long as they are smaller than our clone-insert sizes, there is a good chance that they can be scaffolded across to order and orient the contigs, even if the gap sequences remain undefined. Although the widespread use of capillary sequencers has reduced the frequency of mislabeled clone-end pairs to well under 1%, it is more prudent to always make decisions based on at least two sets of clone-end pairs. We adhered to this rule for scaffold construction; but we relaxed this rule for repeat-gap closure because we used Phrap to validate the sequence overlap before the join is made. Specifically, we extracted 500 bp of sequence from the two flanking contigs and fed the constituent reads to Phrap, along with the unmasked reads from the gap, as identified by the clone-end pairs. Glossary | Shotgun Assembly | Shotgun library. A collection of clones that over-sample the target genome. Clone-end pair. Sequence reads derived from both ends of a shotgun-library clone. Clone-insert size. The size of the clone-insert from which a clone-end pair is taken. Contig. The result of joining an overlapping collection of sequence reads. Scaffold. The result of connecting non-overlapping contigs by using clone-end pairs. LW-gap. A gap in the assembly resulting from Lander-Waterman statistics. Repeat Analysis | Coverage. The number of times a genome is represented by the shotgun data. Copy number. The number of times a sequence occurs in the genome. Depth. The number of times a sequence appears in the shotgun data set. For example, if a transposon has copy number N, and the shotgun data set has coverage C, the transposon will appear at an average depth of NxC. 20-mer repeat. Any 20-mer with a depth D beyond a preset threshold. Repeat-masked Phrap. A shotgun-assembler based on Phrap, for which all the 20-mer repeats are first eliminated from consideration, before determining the extent of overlap between different sequence reads. Repeat-gap. A gap in the assembly that is attributed to the repeat masking. Quality Measures | N50 size. As applied to contigs or scaffolds, that size above which 50% of the assembled sequence can be found. Single-base error rate. The number of small-scale discrepancies per unit length, from a comparison with the reference sequence. Small-scale means smaller than a typical 500-bp sequence read, and usually just a few bases. Contig mis-assembly. An error in how the sequence reads are assembled. By definition, it involves segments larger than a 500-bp sequence read. Comparisons with the reference sequence might reveal missing segments, segments in the wrong orientation, or segments in the wrong order. Scaffold mis-assembly. An error in how the non-overlapping contigs are linked together. Comparisons with the reference sequence might reveal interleaving scaffolds, or contigs in the wrong orientation or order. Mis-assembly rate. The number of erroneous contigs (or scaffolds) divided by the total number of contigs (or scaffolds). RESULTS : We selected an 11.9-Mb region of human chromosome 3, from 3p24.3 to 3p26.1, which was sequenced at the Beijing/Hangzhou Genome Center as part of our contribution to the Human Genome Project. The region was covered by an 87 BAC-clone tiling path, finished to an error rate of <10-4. Each BAC was shotgun sequenced by subcloning into plasmids. Two simulated data sets, at 4x and 4x + 2x coverage, were created by uniformly thinning the plasmid subclone sequences to the desired coverage, thereby preserving any cloning biases. Clone ends were simulated by pairing subclones separated by the desired clone-insert distance in the finished sequence. Because the BACs were covered by 11x of plasmid subclones, and we allowed for a +-10% variance in clone-insert sizes, there were relatively few practical constraints. Clone-end pairs were also deliberately mislabeled, to a worst-case frequency of 1%. A proportionate number of contaminant reads (chimeras, clone deletions, and rearrangements) were included in these simulated data sets to be as realistic as possible. The 4.2x data set for the 430-Mb rice genome is discussed in another paper . To validate this sequence assembly, we used finished BAC sequences from a related cultivar of Oryza sativa ssp. indica. The whole-genome shotgun data came from the 93 --11 cultivar, whereas the BAC sequences came from the Guang-Lu-Ai cultivar, with the GenBank accession nos. , , , , , , , and . The cumulative length of all these finished BAC sequences was 0.89 Mb. Contig-Assembly Accuracy | As an initial test of our algorithm for identifying 20mer repeats, we compared the predicted to the actual probability-of-detection. The key parameter is the threshold depth, which we chose to make the false-positive rate nearly 0.1%. For shotgun coverages of 2x, 4x, and 6x, the threshold depths were 7, 11, and 14, respectively. Our concern was whether or not there are cloning biases that are repeat sensitive. As we show in Figure , actual performance was in agreement with expectations based on Poisson statistics, which means that the cloning biases are not a major problem. However, perfect repeat detection is only possible in the limit of infinite coverage. Figure 2 | Probability of detection for repeats of given copy number at a shotgun coverage of 2x, 4x, and 6x. Probability of detection for repeats of given copy number at a shotgun coverage of 2x, 4x, and 6x. The threshold depths are 7, 11, and 14, respectively. The solid lines are theoretical predictions, which assume Poisson statistics, and symbols of the same color refer to actual performance on the human data sets. Ideally, the probability is 0 at copy number 1, and 1 at copy numbers larger than 1. We cannot detect every low-copy repeat, but the number of Alu and Line1 transposons in this target region is 4749 and 1797, respectively, so we should be able to detect virtually every transposon. In the human 4x data set, 15.9% of the finished BAC sequences and 17.2% of the shotgun sequences are masked by 20mer repeats identified from the shotgun data. These numbers are comparable to the 19.4% and 15.4% of the finished BAC sequences that are attributable to repeats of copy numbers at least 2 and 3. RepeatMasker (A.F. Smit and P. Green, unpubl.), however, identifies 43.7% of the finished BAC sequences as being of transposon origin. The difference is that we mask 20mers that are exactly duplicated in the target region; RepeatMasker identifies anything exhibiting similarity to a known transposon sequence. In contrast, for the rice 4.2x data set, 32.1% of the finished BAC sequences and 42.2% of the shotgun sequences are masked. This larger fraction of masked sequences reflects the fact that rice transposons are of more recent origin , less diverged from their ancestral sequences, and more likely to lead to exactly duplicated 20mers. To a lesser extent, it also reflects the fact that our rice data set comes from a larger target region. Table assesses the performance of RePS to unmasked Phrap. The metrics include the number of contigs, the size of the contigs, the single-base error rate, and the contig or scaffold misassembly rates. The number of contigs is compared to the Lander-Waterman expectation. However, it is often the case that, even when there are a large number of small contigs, the bulk of the assembled sequence may be contained in a small number of large contigs. A simple mean or median would obscure this possibility. We therefore characterize the assembly using N50 sizes, defined to be that number above which 50% of the contig or scaffold sequences can be found. Repeat-masked Phrap does produce more (and smaller) contigs than unmasked Phrap, but this is all right, as it is only the first of many steps, and it is more important to avoid making mistakes early on than it is to build large contigs from the outset. Table 1 | Software Performance Comparisons against the reference sequence reveal two kinds of problems: single-base errors and misassemblies. The single-base error rate is the quantity that is estimated by Phrap. It is measured by counting the number of base discrepancies per unit length in a BLAST-alignment segment . Separate error rates are quoted for unique and repeated sequence. In the human 4x data set, the measured rates are 0.066% (0.063%) for unmasked Phrap and 0.077% (0.076%) for repeat-masked Phrap. These differences are negligible. The estimated error rates from Phrap are somewhat higher, but not by much, and our experience with Phrap is that it tends to overestimate the error rate at low coverages. We add that most of the sequencing errors are at the ends of the contigs. If we restrict the BAC comparisons to contigs >3 kb and trim 500 bp off both ends, the error rates become 0.042% (0.025%) for repeat-masked Phrap. In rice, most of these BAC discrepancies were actually polymorphisms, not sequencing errors, as different rice cultivars were used. This is also reflected by the large differences between the Phrap estimates and the BAC discrepancies. By our definition, contig misassemblies involve segments larger than the typical 500-bp sequence read. They are revealed by a BLAST-alignment with missing segments, segments in the wrong orientation, or segments in the wrong order. These reflect each of the specific problems in Figure . We define the contig misassembly rate as the ratio of bad contigs to total contigs. There is a tradeoff between the contig size and misassembly rate, as shown in Figure , but for low-copy repeats, the tradeoff is minor. Only in the limit of no repeat masking would the misassembly rates increase dramatically, say by a factor of 11, in the human 4x data set. This reflects the fact that transposon copy numbers run into the thousands, whereas, gene duplications rarely go past 10 in copy number . Although some of the misassemblies made by unmasked Phrap are in those repeat-masked regions that are never assembled into a contig by repeat-masked Phrap, a huge majority, 79.5%, are not. Figure 3 | Common contig misassemblies, and how they may be detected during scaffold construction. Common contig misassemblies, and how they may be detected during scaffold construction. There are three common problems: missing segment as a result of two repeats in the same direction (a), segment orientation error due to inverted repeats in opposite direction (b), and segment ordering mixed up as a result of three repeats all in the same direction (c). A consistent scaffold is a unidirectional path that puts the contigs in a definite order. It will not turn around and suggest that different contigs should be put in the same place. However, this is precisely what happens if the scaffold is forced to use a misassembled contig. A different scaffolding problem is depicted in d. When the clone-insert sizes are too large, the scaffolds start skipping contigs, leading to an interleaving morass of scaffolds with no obvious relation between overlapping scaffolds. Figure 4 | Tradeoff between contig size and accuracy of assembly. Tradeoff between contig size and accuracy of assembly. This analysis is based on the human 4x data set, using only repeat-masked Phrap without the clone-end-pairing analysis. Increasing the threshold depth results in less of the sequence being masked, so the N50 contig sizes increase. For low-copy repeats, the resultant increase in misassembly rates is minor. The asymptotic contig size and misassembly rate, in the limit of no repeat masking, is somewhat larger than implied by this figure because transposon copy numbers run into the thousands, and this is well off the scale of the figure. One could argue that by not counting the number of misassemblies in the contig, we are underestimating the severity of the problem in the largest contigs, which are likely to have more than one misassembly. In principle, this is certainly true, but in practice, it is a problem only if the distance between misassemblies is small compared to the contig size, which it is not. As supporting evidence, we did the BAC comparisons using a 1-kb window and searched for breakpoints at which the window could only be matched to disjoint loci. In the human 4x, human 4x + 2x, and rice 4.2x data sets, such breakpoints were found on average once for every 649 kb, 825 kb, and 425 kb, respectively, after repeat-masked Phrap. In no case did we find more than two breakpoints per contig. Therefore, as long as the contigs are relatively small, there is little practical difference between these two definitions of contig misassembly rates. After repeat-masked Phrap, we are left with many gaps that are due to the repeat masking. If these gaps are small enough, the clone-end-pairing information can be used to tell Phrap how to close them. How aggressively repeat gaps must be closed is debatable, especially in the grass genomes, like rice, where most of these repeats are attributable to nested retrotransposons in the intergenic regions between genes . The resultant improvements in N50 contig size after repeat-gap closure are larger in rice than in human because rice has a higher repeat-masked fraction. In every instance, the contig misassembly rates increase significantly, from 0.51% to 1.1%, in the human 4x data set. Comparisons against the BAC sequences reveal that the contig order and orientation are correct and that it is the sequences inside the repeat-gaps that are being misassembled. One outstanding issue is that we do not know how well our software will work on outbred organisms in which there are large polymorphic differences between homologous chromosomes. Perhaps if we fix the sequencing errors in advance , it might be easier to resolve the homologs. However, this is not a problem for data from human BACs or inbred rice strains. Scaffold-Assembly Accuracy | After repeat-gap closure, the contigs are as big as they will ever be, and all we can do is build scaffolds to order and orient the contigs. The remaining gaps are either repeat gaps that are too large to be closed or LW gaps, which are typically smaller than a 500-bp sequence read. Scaffold construction is therefore mostly dependent on how the 20mer repeats cluster in the target genome. In practice, one has to generate a certain amount of whole-genome shotgun data just to compute the 20mer repeats. It is this scenario that we are simulating in our 4x + 2x human data set, in which the clone-insert size is 2 kb in the first 4x and 15 kb in the last 2x. The repeat-cluster size distributions of Figure give us an idea of what we have to do to close the gaps. Notably, when the clone-insert sizes are too large, the scaffolds skip over adjacent contigs, resulting in the interleaving problem of Figure d. This problem can be minimized if the scaffolding starts with the smaller clone inserts and slowly works up to the larger clone inserts. Nevertheless, there are limits, and they become apparent for clone-insert sizes that exceed the contig sizes prior to scaffold construction. In the human 4x + 2x data set, clone-insert sizes >15 kb result in total scaffold sizes that exceed the target genome, as shown in Figure , and indicative of serious interleaving scaffold problems. Figure 5 | Repeat-cluster size characteristics. Repeat-cluster size characteristics. Clusters are defined by placing the 20mer repeats, determined from the shotgun data, onto 11.9 Mb and 0.89 Mb of finished human and rice bacterial artificial chromosome sequence, respectively. Any 20mers separated by <26 bp of unique sequence are merged together, and it is the sizes of these merged clusters that are plotted. In the distribution function for human, the peak near 300 bp is due to Alu transposons. In rice, the distribution is scaled up to reflect the entire rice genome. The cumulants show that a significant fraction of rice repeats lie in kilobase-sized clusters. Another way to demonstrate this fact is to highlight 20mer repeats by blue histogram bars proportional to copy number in typical human and rice segments. Figure 6 | Optimization of clone-insert size. Optimization of clone-insert size. This analysis is based on the human 4x + 2x data set. The first 4x of data has a 2-kb clone-insert size, and the final 2x has the clone-insert sizes indicated here. Scaffold size is the sum of the contigs, plus the estimated gaps between these contigs. For this figure, we plot only the scaffolds with two or more contigs, sum over all the scaffolds, and then divide by 11.9 Mb. Similarly, we plot only contigs that are part of such a scaffold, sum over all the scaffolds, and again divide by 11.9 Mb. As the clone-insert sizes increase, more of the target region is subsumed. After 15 kb, however, the total contig size stops growing, but the total scaffold size does not. This can be explained by the interleaving scaffold problem of Fig. d. In the process of building the scaffolds, we can detect some fraction of the contig misassemblies. A consistent scaffold describes a unidirectional path that puts the contigs in the correct order and orientation. However, as shown in Figure , when the scaffold encounters a misassembled contig, the clone-end analysis tells the path to turn around and put different contigs in the same place. Clearly, this cannot be the correct answer, but the fact that the path misbehaves at a specific contig can identify misassembled contigs and these can be left out of the scaffolds. The scaffold misassembly rate is defined as the ratio of bad scaffolds to total scaffolds. In addition to contigs with the wrong orientation or order, interleaving problems are counted as bad. In fact, we had only one bad scaffold in the entire human 4x + 2x data set, and it was an interleaving problem. The benefits of scaffolding are worthwhile, nevertheless, because the resultant N50 scaffold size is 14 times larger than the initial N50 contig size. Even larger scaffolds would be possible if we linked contigs joined by only a single clone-end pair. DISCUSSION : One could ask if there is an advantage to explicitly identifying the 20mer repeats. Notwithstanding the possibility of fine-tuning the Nmer length, the concept is similar to Celera's estimation of the probability that any overlap is unique . Both are reliant on Poisson sampling statistics and hence we would expect their abilities to detect misassemblies to be similar, given equivalent data sets. On the other hand, some of the arguments for or against whole-genome shotgun were based on the precise nature of the repeats. By being more explicit about the repeats that matter, not the biological repeats, but the mathematical repeats, one can begin to put the arguments on a concrete foundation. For example, how large are the repeat clusters? What is the typical copy number for a transposon and a gene duplication? Where are the high copy number repeats with respect to the genes? All the answers are highly organism specific, and only by being explicit about the repeats can we design the experiment to suit the organism being sequenced. As sequencing moves on to nonhuman genomes, with more limited funding, the continuing high costs of sequencing will place a premium on strategies that can generate useful information at the earliest stages of the project. Even at rough-draft coverages of 4x to 6x, in which sequence assemblies are necessarily more fragmentary, the resultant data can be useful . Frankly, single-base error rates are largely dependent on coverage, and the only real challenge is to build ever larger contigs and scaffolds with as few mistakes as possible. Our scaffolding strategy does leave the larger repeat clusters unassembled, but whether or not this matters depends on the organism being sequenced. In rice, we achieved an estimated 92.% functional coverage (i.e., genes and immediate regulatory sequences), despite leaving a large fraction of the repeats unassembled . Therefore, the approach embodied by RePS is appropriate when sequencing through every last repeat is not a high priority. WEB SITE REFERENCES: : . RepeatMasker screen for intersperesed repeats and low complexity DNA. .Phred/Prap/Consed; Shotgun sequence assembly. Backmatter: PMID- 11997346 TI - Integration of Cot Analysis, DNA Cloning, and High-Throughput Sequencing Facilitates Genome Characterization and Gene Discovery AB - Cot-based sequence discovery represents a powerful means by which both low-copy and repetitive sequences can be selectively and efficiently fractionated, cloned, and characterized. Based upon the results of a Cot analysis, hydroxyapatite chromatography was used to fractionate sorghum (Sorghum bicolor) genomic DNA into highly repetitive (HR), moderately repetitive (MR), and single/low-copy (SL) sequence components that were consequently cloned to produce HRCot, MRCot, and SLCot genomic libraries. Filter hybridization (blotting) and sequence analysis both show that the HRCot library is enriched in sequences traditionally found in high-copy number (e.g., retroelements, rDNA, centromeric repeats), the SLCot library is enriched in low-copy sequences (e.g., genes and "nonrepetitive ESTs"), and the MRCot library contains sequences of moderate redundancy. The Cot analysis suggests that the sorghum genome is approximately 700 Mb (in agreement with previous estimates) and that HR, MR, and SL components comprise 15%, 41%, and 24% of sorghum DNA, respectively. Unlike previously described techniques to sequence the low-copy components of genomes, sequencing of Cot components is independent of expression and methylation patterns that vary widely among DNA elements, developmental stages, and taxa. High-throughput sequencing of Cot clones may be a means of "capturing" the sequence complexity of eukaryotic genomes at unprecedented efficiency. [Online supplementary material is available at . The sequence data described in this paper have been submitted to the GenBank under accession nos. -. Reagents, samples, and unpublished information freely provided by H. Ma and J. Messing.] Keywords: Introduction : When a solution of denatured genomic DNA is placed in an environment conducive to renaturation, the rate at which a particular sequence reassociates is proportional to the number of times it is found in the genome. This principle forms the basis of DNA renaturation kinetics (also called Cot analysis), a technique by which the redundant nature of eukaryotic genomes was first demonstrated (; see Fig. A in the online supplement to this article for a review of Cot analysis, ). In a typical renaturation kinetics study, samples of sheared genomic DNA are heat-denatured and allowed to reassociate to different Cot values [Cot value = the product of nucleotide concentration in moles per liter (C0 or Co), reassociation time in seconds (t), and, if applicable, a factor based upon the cation concentration of the buffer; for review, see ]. For each sample, renatured DNA is separated from single-stranded DNA using hydroxyapatite (HAP) chromatography, and the percentage of the sample that has not reassociated (%ssDNA) is determined. The logarithm of a sample's Cot value is plotted against its corresponding %ssDNA to yield a Cot point, and a graph of Cot points ranging from little or no reassociation until reassociation approaches completion is called a Cot curve . Mathematical analysis of a Cot curve permits estimation of genome size, the proportion of the genome contained in the single-copy and repetitive DNA components, and the kinetic complexity of each component. Interspecific comparison of Cot data has provided considerable insight into the structure and evolution of eukaryotic genomes (e.g., ; ; ; ; ; ). With the advent of molecular cloning techniques, most genome researchers abandoned Cot analysis. However, the principles of nucleic acid hybridization developed through Cot research form the basis of many molecular biology techniques, and information generated in Cot studies remains central to current knowledge of genome structure (for review, see ). Repetitive DNA has proven a particularly difficult problem in the investigation of eukaryotic genomes, especially in plants where large genomes are common. In some plants (e.g., maize, ), recent retroelement amplification has made BAC end sequencing and assembly of shotgun-sequenced clones almost impossible. Consequently, substantial efforts are taken to isolate DNA regions that do not contain repetitive sequences. One means of obtaining single/low-copy sequences is to prepare cDNA libraries. However, the representation of genes in a given cDNA library is only indicative of gene expression in the source tissue(s), and gene copy number is not accurately reflected in cDNA libraries even if "normalization" techniques (e.g., ; ; ; ) are employed. Repetitive DNA is often more highly methylated than low-copy DNA, and consequently some researchers have used methylation-sensitive restriction enzymes (e.g., ) or bacterial host strains that preferentially restrict methylated DNA to produce genomic libraries enriched in low-copy (ostensibly genic) DNA. However, cloning strategies involving the preferential exclusion of hypermethylated DNA may result in the loss of important/interesting genes because the pattern and significance of DNA methylation can differ markedly between species (e.g., ), genes within an organism (e.g., ; ), developmental stages (for review, see ), and different regions of the same gene (e.g., ; ). Clearly, alternative strategies for isolating and sequencing the unique elements of genomes are needed. The results of a Cot analysis provide the information needed to isolate the major kinetic components of a genome in a manner independent of sequence expression and/or methylation . However, to our knowledge DNA fractionated via Cot/HAP techniques has not previously been used in the construction of genomic libraries. Here we describe the production and characterization of genomic libraries derived from the three major kinetic components of sorghum (Sorghum bicolor) DNA. Sorghum was chosen for this study because (to our knowledge) it has not been the subject of a Cot analysis, it has a 4000 --6000 year history of cultivation , it is one of the most agronomically important plant species in the world , and its relatively small genome is a valuable "window" into the low-copy sequence diversity of closely related, large-genome crops such as maize and sugarcane (see ). Our results suggest that cloning of isolated kinetic components is a useful and powerful means to clone genomic sequences based upon their relative iteration and to efficiently discover new DNA sequences in a manner independent of expression and/or methylation patterns. The combination of Cot-based cloning and high-throughput sequencing of Cot libraries [Cot-based cloning and sequencing (CBCS)] represents a means by which the sequence complexity of large genomes can be "captured" at a fraction of the cost of shotgun sequencing. RESULTS : Melting Temperatures and GC Content of Sorghum DNA | Melting curves were generated for sheared sorghum DNA in 0.03, 0.12, and 0.5 M sodium phosphate buffer (SPB), and melting temperatures (Tm) for DNA in each buffer were determined using first-derivative analysis. The melting temperatures for sorghum DNA in 0.03, 0.12, and 0.5 M SPB are 75.1, 84.1, and 93.1C, respectively. For DNA dissolved in buffers with a monovalent cation concentration (Mmvc) between 0.01 and 0.2 M, the GC content of the DNA can be calculated using the formula %GC = 2.44 (Tm - 81.5 - 16.6 log Mmvc) . Consequently, the sorghum DNA samples in 0.03 M SPB (Na+ = 0.045 M) and 0.12 M SPB (Na+ = 0.18 M) result in %GC estimates of 38.9% and 36.5%, respectively. The average of these two values is 37.7%. Cot Analysis | A Cot curve for Sorghum bicolor was prepared according to and analyzed using the computer program of . The analysis providing the lowest RMS (root mean square deviation) and goodness of fit values (0.02554 and 0.02712, respectively) is a three-component fit with no constrained variables. In all Cot analyses, a certain fraction of the DNA forms duplexes even at Cot values approaching zero. Such early renaturation is thought to be due to base pairing between complementary sequences on the same DNA molecule (i.e., foldback DNA; ). As shown , approximately 16% of sorghum DNA had reassociated by the earliest Cot point (10-5 M*sec) and consequently was not included in the curve. No detectable reassociation was observed until a Cot value of about 0.02 M*sec. Figure 1 | Sorghum cot analysis. Sorghum cot analysis. (A) Complete Cot curve, data analysis, and component isolation. A least-squares curve (thick black line) was fitted through the data points (open circles) using the computer program of . The curve consists of highly repetitive (HR), moderately repetitive (MR), and single/low-copy (SL) components characterized by fast, intermediate, and slow reassociation, respectively. For each component, the following values have been placed to the right of the component's general location: Fraction = the proportion of the genome found in that component, Kinet. Comp. (kinetic complexity) = the length in nucleotide pairs of the longest nonrepeating sequence calculated from the Cot data, k = the observed reassociation rate in M-1*s-1, and Cot1/2 = the value on the abscissa of the complete Cot curve at which half the DNA in the component has reassociated. Black diamonds mark the positions on the complete Cot curve of the Cot1/2 values for HR, MR, and SL components. For a Cot component, 80% of the sequences in that component will renature in the "two Cot decade region" (TCDR) flanking the component's Cot1/2 value (brackets centered at Cot1/2 markers; ). We utilized this principle in isolating HR, MR, and SL Cot components for Cot library construction. In brief, all double-stranded DNA within a component's TCDR was isolated except for areas that overlap the TCDRs of other components (regions marked by upward vertical dashes, diagonal stripes, and crosshatching delimit areas of the curve used in HRCot, MRCot, and SLCot library construction, respectively). Note that the area isolated for use in constructing the SLCot library extends a short way past the right end of the TCDR for the SL component; this is presumably not a problem as any double-stranded DNA in the region to the right of the SL component TCDR is likely to be single-copy. (B) The predicted individual renaturation profiles of the HR component, MR component, and SL component are shown. Four percent of the sorghum DNA did not reassociate by the highest Cot value (20,000 M*sec). DNA that does not reassociate by such a high Cot value is thought to be damaged and incapable of binding to HAP (e.g., ). The sorghum Cot curve consists of a fast, an intermediate, and a slow reassociating component. The complete Cot curve, renaturation profiles of the three Cot components, and the reassociation rate (k), Cot1/2 value, kinetic complexity, and genome fraction of each component are presented in Figure . In diploid organisms, the slowest reassociating component of a Cot curve generally represents single-copy DNA sequences. In such cases, genome size can be estimated by comparing the k value of the slow reassociating component to E. coli's rate constant (k = 0.22 M-1*sec-1) and DNA content . The genome of E. coli (strain K12, substrain MG1655) is 4,639,221 bp . Assuming that the sorghum slow reassociating component (k = 0.001474 M-1*sec-1) is composed of single-copy DNA, the estimated 1C genome size of sorghum would be G = (4,639,221 bp x 0.22 M-1*sec-1) / 0.001474 M-1*sec-1 = 6.92 x 108 bp or 692 Mbp. While this value is slightly lower than the reported values based on Feulgen densitometry (753 --837 Mbp, ) and flow cytometry (748 --772 Mbp, ), there is only a 7.5% --17% difference between the Cot-based genome size and these previous estimates. Consequently, it is likely that the slow reassociating component is primarily single-copy DNA, and thus we refer to it as the single/low-copy (SL) component. Assuming that the SL component has a repetition frequency of 1, the average repetition frequency of the DNA in the other components can be estimated by dividing their k values by the k value of the SL component . The predicted repetition frequencies of sequences in the fast reassociating component and in the intermediate reassociating component are 7.8864/0.001474 (5350.3) and 0.1062/ 0.001474 (72.1), respectively. In light of their relative repetitiveness, the fast and intermediate reassociating components are hereafter referred to as the highly repetitive (HR) and moderately repetitive (MR) components. Library Construction, Blot Analysis, and Sequencing | HRCot, MRCot, and SLCot libraries were generated from isolated Cot components (see Methods for details on component isolation and cloning). The relative iteration of the insert DNA in the three Cot libraries was examined by comparing the intensity with which Cot clone probes hybridized to replica Southern blots of sorghum genomic DNA. The average intensity of hybridization to blots incubated with radiolabeled HRCot sequences was 43,067 cpm (+- 6248) while the average values for the MRCot and SLCot blots were 3783 cpm (+- 1419) and 1377 cpm (+- 253), respectively. We sequenced a total of 384 HRCot, 480 MRCot, and 576 SLCot clones of which 253, 409, and 499 (respectively) met our sequence quality criteria (i.e., Ph/Pr value >16 over 300 continuous bases, high-quality insert sequence >=50 bp). The (253 + 409 + 499 = 1161) "quality clones" were BLASTed against the GenBank Nr (nonredundant), GenBank EST, and SUCEST Sugarcane EST databases. For each quality clone, only bit scores (S') of 55.44 or greater were deemed significant and used in characterizing the clone. Unlike E values (E) commonly used to compare the quality of hits, bit scores provide a means of comparing the significance of database hits independent of database and query size (see for details). For a database of 3.5 billion nucleotides (slightly larger than the effective size of the GenBank Nr and EST databases at the time of sequence analysis) and an effective query length of 159 nt, a bit score of 55.44 is roughly equivalent to an E value of 1 x 10-5. For a given quality clone, the term "primary hit" was used to indicate the database sequence (if any) showing the highest significant homology to that clone. Categorization of Cot Clones | Each Cot clone was placed into a single descriptive category ("BLAST category") based upon the scheme shown in Figure . Because of the difficulty associated with evaluating EST hits (see Limitations of the EST Data and Table A in the online supplement to this article, ), GenBank Nr database hits were given priority in the classification scheme with EST hits used only to categorize clones without significant Nr hits or with Nr hits to genomic sequences of unknown character. Results of category assignment and a list of characterized gene and repeat sequences recognized by various Cot clones are given in Table . An overview of the data is shown in Figure . Figure 2 | Classification of Cot clones based on sequence analysis. Classification of Cot clones based on sequence analysis. For comparative purposes, each of the sequenced Cot clones meeting the minimum sequence quality requirements (see Methods section) was assigned to a single descriptive "BLAST category" based upon its significant database hits according to the classification scheme shown. Table 1 | BLAST-Based Categorization of HRCot, MRCot, and SLCot Clonesu Figure 3 | Sequence composition of different Cot libraries. Sequence composition of different Cot libraries. Black bars represent HRCot clones, white bars represent MRCot clones, and diagonally striped bars represent SLCot clones. The BLAST group "Dispersed repeat sequences" is composed of the Retroelement, MITE, and Other dispersed repeat BLAST categories (see Fig. and Table ). All three libraries possessed more clones in the no significant hitBLAST category than any other. Roughly 70% of the SLCot clones showed no significant database hits, whereas about 50% of the MRCot clones and 35% of the HRCot clones fell into this category. HRCot hits were primarily to plant repetitive DNA sequences including retrotransposons and other dispersed repeat sequences, rDNA sequences, and sorghum centromeric repeat sequences. The relative percentage of clones showing homology to repetitive ESTs was considerably higher for the HRCot library (19.4%) than the other two libraries (6.6% for MRCot, 2.2% for SLCot). None of the HRCot clones produced a significant hit to a characterized gene sequence, and the percentage of unique EST clones in the HRCot library was much lower than corresponding values for the MRCot and SLCot libraries (Table , Fig. ). Among the three Cot libraries, the SLCot library showed the highest percentage of hits to characterized gene sequences and unique ESTs. No centromeric sequences were detected, and only 6.2% of the SLCot clones fell into any of the repeat sequence categories. The MRCot library showed intermediate levels of repeat sequences and unique sequences. With regard to low-copy sequences, the percentage of MRCot sequences in the unique EST category was roughly the mean of the corresponding values for the HRCot and SLCot libraries . Of the characterized gene sequences detected in the Cot libraries, 25% were found in the MRCot library (the remaining 75% were in the SLCot library). Although the HRCot library had the greatest fraction of clones with homology to known repeats , some repeat sequences (presumably of moderate iteration) were more abundant in the MRCot library. For example, clones with homology to the retroelement Leviathan were three times more common in the MRCot library than the HRCot library. Likewise, sequences with homology to retrotransposon genes/pseudogenes in the GenBank Nr database were limited to MRCot clones . Of particular note, 10% of the MRCot sequences correspond to chloroplast DNA, presumably a contaminant in the nuclear DNA isolation process . However, chloroplast sequences were detected in less than one percent of the SLCot clones and none of the HRCot clones. While chloroplast DNA was not a desired end product of Cot library construction, the observation that chloroplast sequences are almost exclusively limited to MRCot clones neatly illustrates the "two Cot decade" principle used in the isolation of individual Cot components; that is, 80% of the copies of a given DNA sequence are contained within a span of two Cot decades (Fig. ; ). Based on the Cot curve, the MR component constitutes 41% of the genome, but if a tenth of this component is actually chloroplast DNA, the percentage of the genome found in the MR component may be closer to 37%. Of the 1161 Cot clone sequences used in sequence analysis, only one clone showed a significant primary hit to a mitochondrial DNA sequence. This clone (SLCot4G05) appears to contain a portion of the Sorghum bicolor F0-F1 ATPase alpha subunit gene (GenBank ). Retrosor-6 | One of the largest continuous sorghum DNA sequences in the GenBank Nr database is a 126 kb BAC clone containing the 22 kD kafirin cluster (GenBank ; V. Llaca, A. Lou, and J. Messing, unpubl.). A total of 15.2% of the HRCot clones, 2.4% of the MRCot clones, and 1.0% of the SLCot clones showed primary hits to this BAC. Interestingly, 34 of 39 (87.1%) of the HRCot and 7 of 7 (100%) of the MRCot primary hits to the kafirin cluster BAC are localized within a 7377 bp sequence found only once in the BAC (bases 127,895 --135,271). None of the SLCot hits to the kafirin cluster BAC recognize the 7377 bp sequence. Although the 7377 bp sequence represents only 4.5% of the bases in the kafirin cluster BAC, it accounts for 13.4% of all primary HRCot hits, making it the most frequently recognized S. bicolor sequence. In their annotation of the kafirin cluster BAC, (GenBank ) V. Llaca, A. Lou, and J. Messing have deemed the 7377 bp sequence a "retroelement". Although they have named five other sorghum retroelements (Retrosor-1, Retrosor-2, Retrosor-3, Retrosor-4, and Retrosor-5), they did not name the 7377 bp retroelement sequence. Our study of the sequence likewise suggests that it is a retroelement (see Fig. A), and with the support of J. Messing (pers. comm.), we have named the sequence Retrosor-6. Retrosor-6 possesses no large open reading frames (ORFs) although nucleotide-protein BLAST (blastx) results indicate that it shares limited homology to an ORF1 polyprotein (S' = 43.9 bits) of the gypsy-type retroelement Athila and a putative Arabidopsis pol protein (S' = 42.4 bits). The apparent absence of gag and env genes and the limited homology to known pol sequences suggest that the copy of the retroelement found in the kafirin cluster is no longer capable of autonomous replication. Figure 4 | Retrosor-6. Retrosor-6. (A) The structure of Retrosor-6 and the distribution of the 41 sorghum Cot clones with primary homology to Retrosor-6. Retroelement features of Retrosor-6 include (alpha) duplicated target site sequences flanking both ends of the sequence, (beta) long terminal repeats (LTRs) (bases 1 --1279 and 6098 --7377) with the canonical LTR start/end nucleotides 5'-TG...CA-3', (chi) a primer binding site complementary to the plant tRNA for asparagine (bases 1286 --1301), (delta) an internal sequence region with homology to ORF1 of the Arabidopsis gypsy-type retroelement Athila, and (epsilon) a polypurine tract (bases 6083 --6095) (see for review of retroelement structure). A scale showing distance in base pairs has been positioned underneath the diagram of the retroelement. For each Cot clone recognizing Retrosor-6, a thin black line has been placed above the retroelement marking the relative position(s) and length of the sequence shared by that clone and Retrosor-6. Because the LTRs have almost identical sequences (99.5% sequence identity), all of the clones with homology to one LTR have a similar/identical degree of homology to the other LTR. For these clones, lines have been positioned above both LTRs. (B) Hybridization of a Retrosor-6 probe (diamond-headed arrow in A) to a Southern blot. The labels at the head of each lane indicate the source of DNA in that lane and the restriction enzyme with which the DNA was digested. Specifically, b = S. bicolor, p = S. propinquum, E1 = EcoRI, H3 = HindIII, E5 = EcoRV, and X1 = XbaI. The two species show essentially identical hybridization patterns and intensities. (C) S. bicolor grid probed with a sequence from the Retrosor-6 LTR (chevron-headed arrow in A). (D) A grid identical to that in 'C' probed with a Retrosor-6 partial internal sequence (triangular-head arrow in A). The hybridization patterns observed for grids probed with the internal Retrosor-6 sequence are virtually identical to those produced by the LTR sequence probe for both S. bicolor (C and D) and S. propinquum (data not shown). (E) A section of the S. propinquum BAC grid probed with part of the LTR sequence of Retrosor-6. The number of copies of Retrosor-6 in the S. propinquum and S. bicolor genomes were estimated as described in Table B (available in the online supplement to this article, ) and the Methods section. The region on the grid used as "background" is enclosed within a circle. Examples of clones showing relatively intense hybridization signals are marked by arrows (triangular-heads) whereas clones with relatively weak but interpretable hybridization signals are marked by arrowheads (chevrons). To examine the abundance and dispersal pattern of Retrosor-6 in the genome of S. bicolor and to check for its presence in the wild species S. propinquum, a Cot clone containing 190 bp of the Retrosor-6 sequence was radiolabeled and used to probe a Southern blot containing restriction-digested S. bicolor and S. propinquum DNA. As shown in Figure B, the Retrosor-6 hybridization pattern for both sorghum species is essentially the same, consisting of a few dark bands within a smear of hybridization signal. While most of the Retrosor-6 retroelement shows numerous Cot clone hits (Fig. A), the region between bases 2000 and 4000 has only two hits. To explore whether this region has diverged more rapidly than other parts of Retrosor-6, high-density BAC grids from both sorghum species were probed with a Cot clone containing part of the Retrosor-6 LTR sequence (e.g., Fig. C), and duplicate copies of the grids were probed with a sequence from the 2 --4 kb region of the retroelement (Fig. D). When the autoradiograms for the LTR region and the 2 --4 kb region were digitally aligned and compared, only minimal differences could be detected in the hybridization patterns for a particular species (e.g., Fig. C and D). To estimate the copy number of Retrosor-6 in the genomes of S. bicolor and S. propinquum, the BAC grids probed with the Retrosor-6 LTR sequence were analyzed using a densitometer (see Fig. E). The grid densitometry results suggest that there are approximately 6275 copies of Retrosor-6 in the S. bicolor genome and 6748 copies in the S. propinquum genome (see Table B in the online supplement to this article, ). Assuming an average size for the retroelement of 7377 bp, Retrosor-6 accounts for approximately 6.0% and 6.3% of genomic DNA in S. bicolor and S. propinquum, respectively. Of note, two of the randomly selected HRCot clones hybridized to Southern blots (see Library Construction, Blot Analysis, and Sequencing above) were later shown to contain portions of Retrosor-6. One clone (HRCot2G11) containing part of the Retrosor-6 LTR produced the highest level of hybridization of any of the randomly selected Cot clones with a specific activity of 10,000 cpm. The second clone (HRCot3B01), carrying part of the internal sequence of Retrosor-6, resulted in a hybridization intensity of 5000 cpm, that is, half that of the clone containing the LTR sequence. Molecular Genetic Markers, BAC End Sequences, and Cot Clones | Cot clones were BLASTed against approximately 1500 molecular markers (see the section Molecular Markers in the online supplement to this article, ) from a high-density sorghum molecular map based on RFLP segregation in the progeny of a cross between S. bicolor and S. propinquum . Fourteen Cot clones contained inserts with significant homology (S' >=76.28) to a total of nine markers on the molecular map (see Table C in the online supplement). The Cot clone sequences also were compared to 116 sorghum BAC end sequences (H. Ma, J. Bowers, and A. Paterson, unpubl.). None of the BAC ends showed significant homology to SLCot clones. However, 12 BAC ends possessed significant homology to HRCot clones, six recognized MRCot clones, and two recognized both HRCot and MRCot clones. In total, 20 of the 116 BAC ends (17%) exhibited significant homology to at least one of the 1161 sorghum Cot clone sequences. Assuming that the Cot libraries are representative of the sequence complexities of the components from which they were prepared, a 15% probability that any randomly selected sorghum genomic sequence will share significant sequence identity with one or more of the 1161 quality Cot clones (see the section Probability of Significant BAC End/Cot Clone Homology in the online supplement) would be predicted. The observed percentage of BAC ends with homology to the Cot clones (17%) and predicted percentage (15%) are not significantly different (see Test of Significance of a Binomial Proportion in the online supplement), suggesting that the Cot libraries are reflective of their respective Cot components. DISCUSSION : Although renaturation kinetics has long been used to characterize genomes , to our knowledge the present study is the first report in which Cot components isolated from genomic DNA have been cloned and sequenced. Our results indicate that (1) the Cot libraries differ with regard to sequence iteration and composition in a predictable manner, (2) most of the highly repetitive DNA in the sorghum genome is found within the sequenced HRCot quality clones, (3) a previously unnamed sorghum retroelement is a major component (and perhaps the most abundant sequence) in both the S. bicolor and S. propinquum genomes, (4) Cot clones can be used to augment the information content of both molecular and physical maps, and (5) sequencing of clones from Cot libraries may represent a means by which the diversity of sequences found in a genome can be efficiently "captured". Effectiveness of Cot-Based Cloning | The three Cot libraries differ in relative sequence iteration and composition in a manner reflecting the nature of the components from which they were derived; that is, construction of repetition-based DNA libraries using Cot techniques is effective. When Southern blots of sorghum genomic DNA were probed with randomly selected, radiolabeled Cot clone inserts, those blots hybridized with HRCot sequences exhibited a mean labeling intensity (cpm) >10 times that of MRCot-probed blots and >30 times that of SLCot-probed blots. Detailed sequence analysis of 250 --500 Cot clone inserts from each of the three Cot libraries revealed that the HRCot library is rich in sequences traditionally found in high-copy numbers (e.g., retrotransposons, rDNA, centromeric repeats), the SLCot library is enriched in sequences with homology to characterized genes and unique ESTs, and the MRCot library possesses its own subset of repeat sequences as well as exhibiting some overlap with the HRCot and SLCot libraries (Table , Fig. ). Additionally, the observed percentage (17%) of random sorghum BAC end sequences recognizing one or more of the 1161 sorghum Cot clone sequences was found to be statistically indistinguishable from the percentage expected if the Cot libraries are representative of their respective Cot components (15%). Because the methods employed do not rely upon differential expression and/or methylation of sequences, Cot-based cloning provides a means by which any genomic sequence (including genes expressed at low levels or during short developmental timeframes) can be isolated and cloned based upon its relative iteration. Retrosor-6 | The previously discovered sequence exhibiting primary homology to the greatest number of Cot clones is a 7377 bp retroelement found in the sorghum kafirin cluster sequence (GenBank ; V. Llaca, A. Lou, and J. Messing, unpubl.). This retroelement (now called Retrosor-6) is highly reiterated in both S. bicolor and S. propinquum (Fig. B). These two species possess the same chromosome number and can be crossed (e.g., ), but the resulting progeny exhibit the aberrant segregation and partial sterility that typifies interspecific hybrids. Two lines of evidence suggest that most copies of Retrosor-6 in both sorghum species are similar to the copy of the retroelement found in the kafirin gene cluster. First, S. bicolor and S. propinquum BAC grids probed with part of the Retrosor-6 LTR showed hybridization patterns nearly identical to those observed for duplicate blots probed with part of the internal region of the retroelement (Fig. C,D). Second, a Southern blot probed with a portion of the Retrosor-6 LTR exhibited hybridization signal about twice that of a duplicate blot probed with an internal sequence of similar length ---an observation suggesting that there are roughly two copies of the LTR for each copy of the internal sequence. Based on the assumption that most copies of Retrosor-6 are similar to the kafirin cluster copy of the retroelement, densitometric analysis of BAC grids indicates that Retrosor-6 accounts for approximately 6% of the DNA in both sorghum species (see Table B in the online supplement for this article, ). Because S. bicolor and S. propinquum have similar genome sizes and possess roughly the same number of copies of Retrosor-6, the retroelement may have been introduced into a common ancestor of the two species rather than into the species separately. However, on the assumption that Retrosor-6 provides no selective advantage to the genome and hence can undergo mutation without influencing fitness, the preponderance of apparently intact copies of Retrosor-6 and the relatively high level of shared sequence identity between the LTRs of the kafirin cluster copy of Retrosor-6 (615/618 bp matches, S' = 1171 bits) suggest that the retroelement may be fairly new to the Sorghum genus. Cot Clones and the Sorghum Molecular Map | Cot clone sequences were compared with the sequences of markers on the sorghum molecular map . Three of the nine molecular markers recognized by Cot clones appear to be rDNA sequences. The three "rDNA molecular markers" are found at essentially the same locus on S. bicolor linkage group C (see Table C in the online supplement to this article, ). The 18S-5.8S-26S rDNA locus has been localized by fluorescence in situ hybridization to the longest S. bicolor mitotic metaphase chromosome . Likewise, we recently demonstrated that the longest S. bicolor pachytene chromosome is the nucleolus organizer chromosome . Consequently, it appears that S. bicolor mitotic metaphase chromosome 1, meiotic chromosome 1, and linkage group C are the same entity, the first instance in which a sorghum linkage group has been assigned to a cytologically distinguishable chromosome. Potential Bias in the Sorghum Cot Libraries | E. coli possesses three endonuclease systems that preferentially restrict methylated DNA; McrA, McrBC, and Mrr. These restriction systems do not cleave DNA that has been methylated by the bacterium's endogenous methylase systems . In preparing the sorghum Cot libraries, we used the Promega pGEM-T Easy cloning kit and the accompanying host strain JM109. While JM109 lacks functional McrA and Mrr restriction systems (it is mcrA-, mrr-), it does possess a functional McrBC protein (mcrBC+). The McrBC protein cleaves DNA sequences with the following configuration: 5'-PumCN40 --80PumC-3' . Consequently, it is possible that certain methylated (presumably highly repetitive) sequences from sorghum are underrepresented in one or more of the Cot libraries due to preferential restriction by the McrBC system. However, it is likely that the relatively small size of the Cot clone inserts (similar100 --400 bp) and the relatively large size of McrBC recognition sites (>=44 bp) substantially decreased possible effects of McrBC during cloning. The limited effect of the McrBC genotype on sorghum Cot library construction is suggested by the observation that the highest proportion of HRCot clones showing significant hits to the GenBank Nr database contain sequences that are frequently methylated in plants, that is, retrotransposons and centromeric sequences (Table , Fig. ). Regardless, to construct a Cot library that is truly representative of a particular Cot component, one should use a host strain with a genotype that is insensitive to DNA methylation patterns. Continued Use of Sorghum Cot Libraries | Now that the feasibility of Cot-based cloning has been demonstrated, we have begun to use the sorghum Cot libraries as a means to augment the information content of the rapidly-growing S. bicolor and S. propinquum physical maps . For example, Cot clones with homology to Retrosor-6 are being used to determine the genetic and physical distribution of this element by evaluating colocalization of Retrosor-6 and genetically mapped RFLPs on S. bicolor and S. propinquum BACs. This basic principle will likely be used to physically map other repeat sequences. Cot clone insert sequences with homology to characterized plant genes (see Table ) will be used to find sorghum homologs/orthologs in BAC clones and position these sequences on the physical maps. Comparison of BAC ends with sorghum Cot clone sequences provides a means to identify BAC ends that contain repetitive DNA sequences. As described in the Results, 17% of sorghum BAC ends (n = 116) show homology to HRCot/MRCot sequences. These BAC ends likely contain repetitive elements and thus may be of limited use in contig assembly. Experimental Modifications and Applications | While the goals of the present study were to investigate the feasibility/usefulness of cloning isolated Cot components and further characterize the sorghum genome, Cot clones could be employed in other ways. Additionally, many of the experimental parameters utilized in this project could be altered to meet different research needs. For example:(1) = In our research, only double-stranded DNA resulting from reassociation was used in preparing Cot libraries. However, HAP-fractionated single-stranded DNA can be used in Cot-based cloning as well. In this regard, we have taken single-stranded Cot DNA, generated complementary strands via the random primer method , and used TA-cloning techniques to produce ssDNA-derived Cot clones (D. Peterson, A. Nagel, S. Wessler, and A. Paterson, unpubl.). The use of ssDNA fractions in cloning would be advantageous in instances where the quantity of genomic DNA is limited. Additionally, fewer base pair mismatches would be expected if primer extension techniques rather than strand renaturation were used to generate duplexes for cloning and sequencing. (2) = Foldback sequences could be cloned to produce a "foldback Cot" (FBCot) library. Although most foldback DNA is probably repetitive in nature , some foldback sequences may be single/low-copy DNA; likewise the foldback fraction may contain some sequences not represented in the HR, MR, and/or SL components. Consequently, FBCot libraries may be a source of useful sequence information. We have used random primer/TA-cloning techniques to produce FBCot clones for S. bicolor, although these clones have not yet been sequenced (D. Peterson, A. Nagel, S. Wessler, and A. Paterson, unpubl.). (3) = If the DNA fragments in a component are of a length optimal for automated sequencing (about 500 --1000 bp), the fragments can be cloned using standard techniques. If the DNA fragments in an isolated component are relatively short (e.g., 200 bp as in the present research), prior to cloning the fragments can be joined together using DNA linkers with highly recognizable sequences under reaction conditions that result in concatemers with mean lengths in the optimal size range for sequencing. The generation and cloning of concatemers as described above is similar to the SAGE (serial analysis of gene expression) technique . (4) = By using renaturation kinetics to further purify/characterize isolated Cot components into subcomponents (i.e., minicot analysis; see ; ; ), the resolution of Cot analysis (and subsequently Cot-based cloning) could be increased. (5) = In species where methylation is known to be associated with repetitive DNA (e.g., ), cloning of isolated SL sequences into mcrBC+/mcrA+/mrr+bacterial strains may further decrease contamination of the resulting library with repeat sequences. (6) = EST/cDNA and genomic libraries could be screened with isolated Cot fractions to identify populations of clones containing probable unique and/or repetitive sequences. (7) = The possibility of affordably automating (and thus standardizing) many of the Cot analysis/HAP fractionation procedures is well within modern capabilities. Capture of Sequence Complexity Using Cot-Based Cloning and Sequencing (CBCS) | While analysis of complete genome sequences is the ultimate means by which the genomes of different species can be compared, genome sequencing may not be an affordable, realistic, and/or desirable option for species with large, highly repetitive genomes. An alternative to genome sequencing is the "capture" (isolation, cloning, and sequencing) of an organism's sequence complexity ; that is, the combined length in nucleotide pairs of the different DNA sequences that comprise a genome . Because most prokaryotic genomes are relatively devoid of repetition, the sequence complexity of a bacterial genome is roughly the same as its genome size . In contrast, the sequence complexity of a eukaryotic genome is the combined length of all of its single-copy DNA sequences plus one copy of each repeat sequence (e.g., a genome composed of 100,000 copies of sequence A, 9000 copies of sequence B, 3400 copies of sequence C, two copies of sequence D, and one copy each of sequences E --Z would have a sequence complexity of A + B + C + D + E + F...+ Z bp). Cot analysis provides an accurate means of estimating the sequence complexity of kinetic components . To distinguish between the exact sequence complexity of a component/genome (presumably only determinable by complete component/genome sequencing) and an estimate of its sequence complexity based on a Cot analysis, the term "kinetic complexity" is used to identify the latter . However, this convention does not mean that kinetic complexity values do not accurately reflect sequence complexity ---as an analogy, the exact genome size of an organism cannot really be determined except by complete genome sequencing, although genome size can be accurately estimated using Feulgen densitometry, flow cytometry, Cot analysis, and other methodologies. Because each repeat sequence is counted only once in determination of a genome's sequence complexity, the contribution of repeat sequences to sequence complexity is generally quite small. In contrast, single/low-copy sequences account for the vast majority of a genome's sequence complexity (e.g., 98% of the combined kinetic complexity of the sorghum HR, MR, and SL Cot components is found in the SL component; Fig. ). The use of bacterial strains sensitive to DNA methylation has been proposed as a means to capture the low-copy sequences that comprise most of a genome's sequence complexity (i.e., "methyl filtration", ). However, methyl filtration and similar approaches such as PstI cloning are based on the assumption that hypermethylated sequences represent DNA that is nongenic whereas hypomethylated sequences represent low-copy DNA. In most (if not all) instances, using cloning/sequencing techniques centered on differential sequence methylation will result in the loss of many important and interesting genes: (1) it is common knowledge that methylation is one of the primary means by which genes are regulated, and that the methylation status of genes (or portions of genes) differs markedly between tissues and/or developmental stages , (2) some genes are normally active when hypermethylated and may not function if they are demethylated , (3) in some genes methylation at one site enhances transcription whereas methylation at another site reduces transcription , and (4) some species normally possess hypermethylated genes and hypomethylated repeat sequences . We propose "Cot-based cloning and sequencing" (CBCS) as a means to capture the sequence complexity of a genome in a manner independent of methylation. In CBCS, isolated kinetic components are cloned to produce Cot libraries, and clones from each library are sequenced using high-throughput methods. To obtain comparable sequence complexity coverage for different Cot components, Cot clones from each Cot library are sequenced in proportion to the kinetic complexity of the component from which they were derived. The usefulness of CBCS is best demonstrated when compared with shotgun sequencing (the sequencing of randomly selected clones from a genomic library), the primary means by which genomes are currently sequenced. In shotgun sequencing, the number of different clones (n) that need to be sequenced in order to have 99% confidence that all genomic elements have been sequenced at least once (i.e., that the sequence complexity of the genome has been captured) can be calculated using the following formula: where Z = mean insert size in bp and G = 1C genome size in bp . For a standard sorghum (1C = 760 Mb) genomic library containing 600 bp inserts, 99% confidence can be obtained by sequencing 5.8 x 106 randomly selected genomic clones. In Cot library construction, genomic DNA is separated into kinetic/sequence complexity-based components prior to cloning and sequencing. Consequently, for a Cot library the probability of sequencing 99% of the DNA elements in the component used to construct that library is a function of the component's kinetic complexity (gamma) rather than genome size: Assuming one had sorghum HRCot, MRCot, and SLCot libraries with 600 bp inserts, 99% confidence could be obtained for (1) the HR component by sequencing 142 HRCot clones, (2) the MR component by sequencing 3.1 x 104 MRCot clones, and (3) the SL component by sequencing 1.3 x 106 SLCot clones. The total number of Cot clones that would need to be sequenced to have 99% confidence that all HR, MR, and SL component elements had been sequenced would be (142 + 3.1x104 + 1.3x106 =) 1.33x106 clones. However, the HR, MR, and SL components comprise only 80% of sorghum DNA; the remaining 20% is divided between foldback DNA (16%) and damaged (unannealable) sequences (4%). With regard to the latter sequence category, it represents a small proportion of the genome and presumably contains no sequences that are not found in one or more of the other fractions. Assuming that damage is a random event that would affect all portions of the genome in a manner proportional to their relative fractions, less than one-quarter of the 4% unannealable DNA (i.e., <7.6 million base pairs or 1.0% of the entire genome) would be high sequence complexity (single/low-copy) DNA. Consequently, the unannealable DNA can essentially be ignored. The same is not true of the foldback fraction (see Experimental Modifications and Applications above). To be fairly secure of retrieving the useful sequence information from the foldback fraction, it can be assigned a "kinetic complexity" equal to the number of base pairs it contains. In sorghum, the foldback fraction contains (0.16 x 760 Mbp =) 1.2 x 108 bp of DNA, and thus using Equation and a mean insert size of 600 bp, sequencing of 9.2 x 105 "foldback Cot" (FBCot) clones would give 99% confidence that all sequences in the foldback fraction had been sequenced at least once. Consequently, using CBCS and the highly conservative assumption that the foldback fraction is largely single-copy DNA, the sequence complexity of the entire sorghum genome could be captured (similar99% confidence) by sequencing a total of (1.33 x 106 + 9.2 x 105 =) 2.3 x 106 Cot clones. Undoubtedly sequencing of 2.3 million clones would be a significant undertaking. However, capturing the sequence complexity of the sorghum genome using the shotgun approach would require sequencing of (5.8 x 106 / 2.3 x 106 =) 2.5 times as many clones. The relative advantage of CBCS over shotgun sequencing is even more pronounced for species possessing genomes with higher proportions of repetitive DNA ---for some plants and animal species, CBCS allows genome sequence complexity capture using less than one-tenth the number of clones that would be required using shotgun sequencing (D. Peterson, S. Wessler, and A. Paterson, in prep.). In all cases, the minimum number of clones needed to attain a specific level of sequence complexity coverage can be calculated in advance of initiating sequencing. Sorghum has a relatively high percentage of foldback DNA compared to many species for which Cot analyses have been performed. Several strategies employed prior to FBCot sequencing (e.g., screening high-density grids of the FBCot library with randomly selected FBCot clones) could be used to identify highly redundant FBCot sequences and subsequently reduce the number of FBCot clones that would need to be sequenced to attain a desired level of sequence complexity coverage. CBCS may not provide information on small variations in individual members of repetitive DNA families; such information is important in the disambiguation and assembly of complete genomic sequences. This limitation of CBCS might be remedied by coupling it with various techniques designed to detect small variations in related sequences (i.e., DNA resequencing techniques; see ; ; ; ). Regardless, the ability to capture the sequence complexity of a higher organism with far less investment than is required by shotgun sequencing may greatly accelerate the timetable for genome-wide study of many of the world's biota. METHODS : Plant Material | Sorghum bicolor (L.) Moench (breeding line BTx623) DNA was used for Cot analysis, Cot library construction, and as a source of DNA in blotting experiments. For comparative purposes, Southern blots and colony blots containing DNA from Sorghum propinquum Kunth, a noncultivated sorghum species crossed with BTx623 to make a detailed genetic map (; A. Paterson and J. Bowers, in prep.) were probed with BTx623 DNA probes (see below). Melting Curves and Cot Analysis | DNA isolation, preparation, and melting analyses were performed as described . Cot analysis was performed according to except that 0.5 M SPB was used to elute double-stranded DNA from HAP columns rather than 0.48 M SPB. A least squares analysis of the Cot data was performed using the computer program of . Cloning of Cot Components | Highly repetitive (HR), moderately repetitive (MR), and single/low-copy (SL) DNA components of the Cot curve were prepared for cloning as outlined in Figure . The sections of the Cot curve used for cloning (i.e., roughly the two Cot decade regions flanking the Cot1/2 value of each component) are shown in Figure . DNA sample concentrations were determined using KOH-denaturation and spectrophotometry as described by . Figure 5 | Overview of the steps involved in cloning HR, MR, and SL Cot components. Overview of the steps involved in cloning HR, MR, and SL Cot components. DNA was denatured by heating samples in boiling water for 5 --10 min. For samples in a particular sodium phosphate buffer (SPB), renaturation was allowed to occur at the criterion (Tm - 25C) unless noted otherwise (see for details). Single-stranded DNA (ssDNA) and double-stranded (dsDNA) were separated using hydroxyapatite (HAP) chromatography. To attain the equivalent of a specific Cot value when starting with an isolated Cot fraction, the desired Cot value should be multiplied by the fraction of the genome remaining single-stranded at the Cot value of the starting material (see ). This principle was employed once in the isolation of the MR component (*) and once in the SL component isolation : * From the Cot curve, 0.67 of the genome is single-stranded at a Cot of 0.94161. To achieve renaturation equivalent to Cot 67.8 with whole genomic DNA, the Cot >0.94161 DNA was renatured to a Cot of (67.8 x 0.67 =) 45.4. From the Cot curve, 0.28 of the genome is single-stranded at a Cot of 94.161. To achieve renaturation equivalent to Cot 10,000 with whole genomic DNA, the Cot >94.161 DNA was renatured to a Cot of (10,000 x 0.28 =) 2800. Isolated Cot components were digested with mung bean nuclease (Promega) to remove single-stranded DNA overhangs (see manufacturer's instructions), and the resulting blunt-ended molecules were cloned into E. coli (JM109) using the Promega pGEM-T Easy cloning kit (cat. no. A1380). The HRCot, MRCot, and SLCot libraries were plated onto selective media, and positive clones were transferred via sterile toothpicks into freezing medium in 96-well microtiter plates. In total, four plates of HRCot, five plates of MRCot, and six plates of SLCot clones were obtained. Cot libraries were replicated using a hand-held 96-pin replicator and stored at -80C (see for details). Each clone was named based upon the library, plate, row, and column in which it was found (e.g., HRCot3A10 = HRCot library, plate 3, row A, column 10). Sequencing | Plasmids were isolated from Cot clones using an alkaline lysis method with modifications made for the 96-well plate format . Cycle sequencing reactions were performed using the BigDye Terminator Cycle Sequencing Kit Version 2 (Applied Biosystems, Foster City, CA) and an MJ Research (Watertown, PA) PTC-100 thermocycler. Finished cycle sequencing reactions were filtered through Sephadex filter plates directly into Perkin-Elmer MicroAmp Optical 96-well reaction plates. Sequencing was performed using an ABI 3700 automated DNA Analyzer. ABI sequencer trace data was evaluated using the programs PHRED, CROSSMATCH, and PHRAP (see for additional information). Only clones with a Ph/Pr value >16 over 300 continuous base pairs and insert sequences >=50 bp in length were used in sequence analyses. Sequence Analysis | The sequence of each Cot clone was compared to sequences in the GenBank Nr and EST databases , and the SUCEST Sugarcane EST database using standard BLAST (blastn) protocols . Based on the nature of the hits (if any), each Cot clone insert sequence was placed into a single descriptive "BLAST category" according to the scheme shown in Figure . The Retrosor-6 sequence (bases 127,895 --135,271 of GenBank ) was compared to data in the GenBank Nr database using standard blastn (nucleotide query -- nucleotide database) and blastx (nucleotide query -- protein database) programs . Southern Blots | Southern blots containing S. bicolor and S. propinquum DNA were prepared and probed as described by . For simple determination of hybridization intensity, 15 clones from each Cot library were randomly selected as sources of probes. Clone inserts were preferentially amplified by PCR and labeled with 32P-dCTP using nick translation. Each blot was hybridized with 1.8 ng/mL (= 20 muCi/mL) of radiolabeled probe DNA in hybridization buffer for 16 h at 65C. Excess solution was drained from blots, and blots were given three successive 20 min washes (65C) in 0.25x SSPE (aqueous 0.75 M NaCl, 50 mM NaH2P04H20, 6.3 mM EDTA, pH 7.4) containing 0.25% SDS (1.0 L per wash with agitation). Membranes were blotted dry with paper towels and wrapped in plastic wrap. A Geiger-Muller counter was used to measure the relative amount of hybridization (cpm) of each probe to its corresponding blot. After sequence analysis, one of the S. bicolor/S. propinquum Southern blots was probed with radiolabeled insert from a Cot clone with substantial sequence identity to Retrosor-6 (HRCot3E04). Hybridization conditions were identical to those described above. An autoradiogram of the blot was obtained using standard protocols. Colony Blotting | High-density grids containing 18,432 double-spotted clones were prepared from the S. bicolor BAC library BTx623 (D. Begum, unpubl.) and the S. propinquum library SP/YRL as described by . For each BAC library, two identical BAC grids (i.e., two grids containing the same clones in the same order) were selected for analysis. One S. bicolor grid (SB1) and one S. propinquum grid (SP1) were each probed with part of the long terminal repeat (LTR) sequence (clone MRCot2B04) of Retrosor-6, whereas the duplicate filters (SB2 and SP2) were probed with a sequence found in the central region of Retrosor-6 (clone HRCot3C12) . Autoradiogram images were digitally captured using an Alpha Innotech (San Leandro, CA) AlphaImager 2200 image capture/analysis system. The two SB images were aligned, superimposed, and compared using Adobe Photoshop 6.0. SP images were likewise compared and analyzed. To estimate the Retrosor-6 copy number in the genomes of S. bicolor and S. propinquum, the AlphaImager Spot Densitometry application (AlphaImager 2200 v. 5.1) was used to analyze one section (i.e., one-sixth) of BAC grid SB1 and one section of grid SP1 (see Fig. E). For a section, a region within the section containing no visible probe hybridization was selected and set as "background." The "Integrated Density Value" [IDV = Sigma (each pixel value - background)] for the entire section was then determined. Because BAC clones are double-spotted on grids, the IDV of the section was divided by two to yield the "Section IDV." Using a circular sampling tool with a fixed diameter slightly smaller than a clone, IDV readings were taken for 50 different clones ranging from the lowest detectable hybridization signal to the highest hybridization intensity (Fig. E). Clones were selected from all areas of a grid section. The mean density value of the five clones with the lowest IDVs (LowIDV) and the mean value of the five clones with the highest IDVs (HighIDV) were determined. For both S. bicolor and S. propinquum, comparison of the LowIDV and HighIDV indicate an approximately fourfold difference in clone hybridization intensity. It was assumed that the LowIDV represents clones with one copy of Retrosor-6, and therefore inferred that the HighIDV represents clones with four copies of Retrosor-6. To determine the mean number of clones per section, the SectionIDV was divided by the LowIDV. The resulting value was used to estimate the Retrosor-6 copy number per genome and the percentage of the genome composed of Retrosor-6 DNA (see Table B in the online supplement to this article, ). Comparison of Cot Clones With Sorghum Molecular Markers and BAC End Sequences | Cot clone sequences were compared to roughly 1500 molecular markers (see the section Molecular Markers in the online supplement for GenBank accession numbers) on the sorghum molecular genetic map using standard BLAST (blastn) procedures . The chromosomal positions of Cot clones containing sequences with high sequence similarity to molecular genetic markers (S' >=76.28) are shown in Table C of the online supplement. BAC end sequences (n = 116) obtained from H.-M. Ma were BLASTed against the GenBank dbGSS database (which contains the sorghum Cot clone sequences). Significant hits to Cot clones (S' >=76.28) were noted. WEBSITE REFERENCES : ; SUCEST: The Sugarcane EST Project. ; Genome Research website. ; National Center for Biotechnology Information (home of GenBank). ; The Phred/Phrap/Consed System home page. Backmatter: PMID- 11997343 TI - Evidence Suggesting That a Fifth of Annotated Caenorhabditis elegans Genes May Be Pseudogenes AB - Only a minority of the genes, identified in the Caenorhabditis elegans genome sequence data by computer analysis, have been characterized experimentally. We attempted to determine the expression patterns for a random sample of the annotated genes using reporter gene fusions. A low success rate was obtained for evolutionarily recently duplicated genes. Analysis of the data suggests that this is not due to conditional or low-level expression. The remaining explanation is that most of the annotated genes in the recently duplicated category are pseudogenes, a proportion corresponding to 20% of all of the annotated C. elegans genes. Further support for this surprisingly high figure was sought by comparing sequences for families of recently duplicated C. elegans genes. Although only a preliminary analysis, clear evidence for a gene having been recently inactivated by genetic drift was found for many genes in the recently duplicated category. At least 4% of the annotated C. elegans genes can be recognized as pseudogenes simply from closer inspection of the sequence data. Lessons learned in identifying pseudogenes in C. elegans could be of value in the annotation of the genomes of other species where, although there may be fewer pseudogenes, they may be harder to detect. [Online supplementary material available at .] Keywords: Introduction : The complete genome sequence with its annotation, for the nematode worm Caenorhabditis elegans , is a considerable resource with which to investigate biology. The most recent estimates predict, on the basis of sequence data, that this worm's genome contains 18,959 protein-coding genes , although only similar4000 of these have been genetically or biochemically characterized, despite the intense study of this experimentally highly tractable system. A primary aim of C. elegans research is to understand how the genome, via the developmental program, generates the animal, but, as yet, there is little functional knowledge for the vast majority of genes predicted in the genome. This laboratory has been determining gene expression patterns for the annotated C. elegans genes using reporter gene fusion technology as one approach through which to explore genome function . Although there are caveats in using this technology (e.g., the need for caution in assuming that an expression pattern observed is an accurate reflection of the expression of the endogenous gene), this approach has the advantage that an expression pattern is linked absolutely to an annotated gene in the genome. In our current strategy, the annotated C. elegans genes are effectively assayed at random, sampling the genome annotation. While analyzing our data, we have noticed that genes duplicated relatively recently in C. elegans' evolution are much less likely to drive reporter gene expression. One interpretation implies that many of the predicted genes, possibly a fifth of the genes in the annotated genome, are nonfunctional pseudogenes. RESULTS : We have examined the expression of 364 of the annotated C. elegans genes using our current reporter gene fusion approach. After shotgun cloning of 5 --7-kb genomic DNA restriction fragments into lacZ or GFP reporter gene expression vectors , the fusion junction was sequenced for randomly selected clones. Plasmids with a C. elegans gene to reporter gene translational fusion that would be appropriate for expression analysis, according to the genome annotation in ACeDB/WormBase, were thereby identified. The point of fusion for any particular gene was random and could therefore be at any position within the predicted protein-coding region. The translational reading frame was corrected when fusions were to an exon, in the appropriate orientation, but in the wrong reading frame. Expression of the reporter was examined in situ in worm strains generated by transformation with the identified plasmids. The expression pattern data generated are presented on our laboratory web site and in the C. elegans database WormBase/ACeDB . Of the 364 effectively randomly selected annotated genes examined, 186 (51%) failed to drive reporter gene expression to observable levels. There are a number of potential reasons for lack of reporter gene expression. The endogenous gene may be expressed at very low levels or only under specific environmental conditions. Transgene expression in the germ line can be suppressed . Approximately one quarter of C. elegans genes are organized into polycistronic units , and no attempt was made to avoid such operons in this study, because operons are difficult to predict on the basis of sequence data alone. A reporter gene fusion to a gene that is downstream in an operon may not contain distant upstream promoter elements, and may therefore fail to show expression. Finally, the structure of many of the annotated genes is based primarily on predictions by the computer program Genefinder. It is thought that 5.45% of all exons may be mispredicted in a way that a correct translational fusion would not be formed within the reporter gene fusions assayed , and splitting of one gene into two predicted genes can mean an assayed fragment would not contain the necessary promoter elements. While analyzing our data, we noted a remarkable correlation that addresses this issue of why a C. elegans gene's promoter region may fail to drive reporter expression. The proteins encoded by the 364 genes analyzed for reporter gene expression were classified as unique, duplicated, or conserved on the basis of BLAST comparisons with the C. elegans, Drosophila melanogaster, Saccharomyces cerevisiae, and human genomes. Either BLAST scores were extracted from the Proteome database or generated using the BLASTP algorithm applied to the NCBI nonredundant database . Genes were considered homologous if a BLAST score E value was less than or equal to 10-6. By this criterion, genes that have no homolog were classified as unique. For the genes that do have homologs, those for which there is only a C. elegans homolog or the C. elegans homolog is a markedly better match than the best non-C. elegans homolog were classified as duplicated. [The criterion used was that the -log10 (E value of the C. elegans homolog) was more than twice the -log10 (E value of the best other organism homolog)]. The remaining genes, which have close homologs in other species, were simply classified as conserved. Duplicated genes were far less likely to drive reporter gene expression than genes in the unique or conserved categories. Whereas 62% (36 of 58) of the unique genes, and 64% (104 of 162) of the conserved genes gave reporter gene expression, only 26% (38 of 144) of the duplicated genes did so (Fig. A). Figure 1 | Percentage of genes that gave reporter gene expression. Percentage of genes that gave reporter gene expression. Genes are classified as duplicated, unique, or conserved as explained in the text. (A) All genes examined. (B) Genes examined divided between those with ESTs (open bars), and those without (shaded bars). A list of the genes examined, divided into these various categories, is available as supplementary data. A number of possible explanations can be identified for why a smaller proportion of the genes, which have undergone relatively recent duplication within C. elegans evolution, give reporter gene expression as compared with nonduplicated genes. Duplicated genes may be more likely to be expressed to lower levels or only upon environmental induction. Alternatively, the annotation for a large proportion of the duplicated genes may be incorrect, either with regard to intron/exon structure or because they are, in fact, pseudogenes. Low or conditional expression could, at best, only partially explain our observations, according to the following considerations. Genes expressed to very low levels or only under specific environmental conditions are less likely to have associated ESTs/cDNAs. According to ACeDB, 207 of the 364 genes examined have identifiable ESTs/cDNAs (with at least 95% nucleotide identity between the cDNA and gDNA sequence), and the proportion is lower for the duplicated genes: 60% (35 of 58) of the unique genes, 37% (53 of 144) of the duplicated genes, and 73% (119 of 162) of the conserved genes have ESTs. Nevertheless, possession of an EST/cDNA only makes it very slightly more likely that a unique or conserved gene will give reporter gene expression (Fig. B). Therefore, level of expression does not appear to correlate with ability to drive reporter gene expression for genes in these categories. Furthermore, whereas possession of an EST/cDNA increases the likelihood of obtaining reporter gene expression for a duplicated gene, the probability of obtaining expression still does not reach that for the other gene categories. The proportion of genes with ESTs that are able to drive reporter gene expression is 63% for unique genes (22 of 35), 40% for duplicated genes (21 of 53), and 65% for conserved genes (77 of 119) (Fig. B). An explanation for why duplicated genes with ESTs are less likely to give reporter gene expression than unique or conserved genes with or without ESTs and why duplicated genes without an EST are even less likely to give reporter gene expression is still wanting. The remaining explanation for our observations is that a significant proportion of the duplicated genes are really pseudogenes or have an incorrect intron/exon structure prediction, both being errors in the C. elegans genome annotation. Gene structure predictions might be expected to be least reliable for the unique genes, rather than the duplicated genes, because they lack homology that can be used to guide exon identification. The key difference between the duplicated genes and the unique or conserved genes is that duplicated genes have an identifiable paralog that must have arisen through a genetic duplication that occurred since the evolutionary split with D. melanogaster. Only rarely will a duplicated gene acquire a new essential function and become fixed within a population. For most duplicated genes, one copy is expected to accumulate mutations and become a nonfunctional pseudogene . Depending upon the order in which mutations accumulate over evolutionary time, a pseudogene may still be transcribed and, therefore, even annotated genes with ESTs may be pseudogenes, as implied in our data. Pseudogenes will continue to drift until they are either deleted or become unrecognizable as a genetic copy. The number of pseudogenes in a genome will depend on the relative rates of gene duplication and pseudogene loss. If the low frequency of reporter gene expression for the duplicated genes was mainly due to the presence of pseudogenes, then 20% of the annotated genes in the C. elegans genome would be pseudogenes. The basis of this claim is as follows. Given that 64% of the unique and conserved genes gave reporter gene expression, then, for 38 genes in the duplicated category to give reporter gene expression as observed, 59 of the 144 duplicated genes would need to be real genes, leaving 85 as pseudogenes. No conclusions can be drawn as to the number of pseudogenes in the unique or conserved categories, but for the purpose of this calculation, there are assumed to be none. If 85 of the 364 genes examined are pseudogenes, then >4000 of the 18,959 annotated genes in the C. elegans genome would be predicted to be pseudogenes. A similar, but less reliable extrapolation, based only on the results for genes with ESTs, suggests that approximately one-quarter of these pseudogenes are transcribed. Of course, some predicted genes that have given reporter gene expression could also be pseudogenes. Such a large number of pseudogenes in the C. elegans genome is not inconsistent with the observations of other investigators. Biochemical or genetic evidence concerning function has been generated for only a minority of the 18,959 predicted C. elegans genes . Extensive analysis of clones from cDNA libraries have identified ESTs for just 10,000 genes , and microarray analysis has been able to detect transcripts for only 56% --59% of C. elegans genes . Transcripts could be specifically amplified by PCR, from a cDNA library, for 84% of predicted C. elegans genes , a percentage consistent with our estimates of the numbers of transcribed and nontranscribed pseudogenes in the C. elegans genome. There are already 543 predicted pseudogenes identified in ACeDB/Wormbase. In bioinformatic analysis of specific large gene families in C. elegans, such as chemoreceptor genes , an even higher percentage of pseudogenes has been identified. Bioinformatic analysis found 2168 genomic sequences, which do not overlap with annotated pseudogenes or genes, but nevertheless have homology to known or predicted C. elegans exons, and the presence of stop codons or frameshift mutations suggest that these are pseudogenic . Finally, on the basis of a number of close paralogs, it has been proposed that C. elegans has a very high rate of gene duplication, generating 383 duplicated genes every million years, as compared with 31 and 52 for D. melanogaster and S. cerevisiae, respectively . However, we cannot totally rule out the possibility that another peculiar and unrecognized property of the recently duplicated genes, such as use of more distant promoter elements or distinct splicing mechanisms, is causing the differential rates of success in our reporter gene fusion experiments. Proving that a gene unit is totally nonfunctional, and is therefore definitely a pseudogene, is impossible. Nevertheless, a search of the sequence data was undertaken for obvious evidence that might suggest that some of the annotated genes we had assayed were likely to be nonfunctional. It was anticipated that stop codons, translational frameshifts, or deletions in otherwise conserved protein-coding regions may have been avoided in the gene structure predictions. The sequences of the 74 annotated genes in the duplicated category that had no EST, and which failed to give reporter gene expression, were examined. Three-quarters of these annotated genes would need to be pseudogenes if our interpretations are correct. The predicted amino acid sequence was used in a BLAST search of C. elegans WormPep, and the closest homologs were aligned using CLUSTALX . These alignments and the alignments presented within ACeDB by BLIXEM were simply visually inspected, because only clear examples of pseudogenes were sought. Annotated genes would need to show extensive sequence identity if faults were to be apparent in this preliminary analysis, and for 26, the homology wasn't good enough to draw any firm conclusions. For two annotated genes, the point of reporter gene fusion lay in potentially mispredicted coding regions, nonhomologous regions that are probably in introns. For 33 annotated genes, the coding region did appear to be intact. However, the integrity of the promoter region would not be assessed in this analysis and for three of these (F14H8.4, K02E2.3, F14H8.4), either another gene or repetitive DNA was located very close upstream of the initiation codon, suggesting that the promoter region may not be intact. Despite the cursory nature of this analysis, potential faults were identified for 13 of these 74 annotated genes . Table 1 | Annotated Genes that Appear to be Pseudogenes From Inspection of the Sequence, Listed According to the Fault Found For nine of the annotated genes examined (e.g., F10D2.8; Fig. ), coding region for well-conserved amino acid residues appears to have been deleted. Gene structure predictions across these incongruities often either incorporate nonhomologous, presumably intronic DNA in the coding region, and/or designate homologous, presumably coding DNA as intron to maintain the integrity of the coding region. One annotated gene, F19G12.2, appears to consist of two pseudogenes joined together. The downstream pseudogene is homologous to the ribonuclease-diphosphate reductase encoding gene, C03C10.3, and has a coding-region deletion. The remaining upstream unit is then a single predicted exon, a rare structure for a real C. elegans gene, with multiple deleted homologs scattered around the genome. Figure 2 | CLUSTALX alignment of the predicted amino acid sequences for F10D2.2, F10D2.5, F10D2.6, F10D2.7, F10D2.8, and F09G2.6. CLUSTALX alignment of the predicted amino acid sequences for F10D2.2, F10D2.5, F10D2.6, F10D2.7, F10D2.8, and F09G2.6. Each of these genes appears to encode UDP-glucuronosyl transferase. The amino acids in the gap in the alignment for F10D2.8 (positions 120 --174) are encoded by an extension of the predicted exon 2 into nonhomologous, presumably intronic DNA. No vestiges of the missing coding region can be seen in this 175-bp second intron. The deletion of this conserved region suggests that F10D2.8 is a pseudogene. Apparent deletions or insertions for the other sequences in the alignment can be accounted for by errors in gene structure prediction, suggesting that the other annotated genes in this cluster have intact coding regions. Extra, nonhomologous amino acids predicted at the amino terminus for F10D2.2 and F09G2.6 have been removed for this alignment. Whereas coding region deletions seem unlikely to be a consequence of simple errors in the sequence data, apparent frameshifting alterations or stop codons could be. Apparent frameshifting alterations were found in four of the annotated genes examined (e.g., B0281.4; Fig. ), two being associated with coding-region deletions. Three of the annotated genes examined had stop codons within well-conserved protein-coding regions with one of these, F56D6.1, also showing a coding-region deletion and a frameshifting alteration. Figure 3 | An apparent frameshift in the annotated gene B0281.4. An apparent frameshift in the annotated gene B0281.4. This image was generated from a BLIXEM window of ACeDB. The top of the figure shows the predicted gene structure for B0281.4, and each horizontal line represents a homology. The ruler is marked in base pairs. The region within the box is expanded to give the lower part of the figure. The theoretical translation across this window is given in the three reading frames, (+1), (+2), and (+3). The amino acid sequence encoded at the end of the predicted first exon, ALVETDIPVSYFLLTHHG, lies in the second reading frame, whereas the amino acid sequence encoded at the start of the predicted second exon, ESEN, lies in the third reading frame. The homologous sequence for the other two genes in this tandem cluster, B0281.5 and B0281.6, is retained from the BLIXEM window, whereas additional homologous sequences have been removed for clarity. The homology extends from the second exon into the predicted first intron in the third reading frame (underlined). The homology continues with the sequences AEKDDS for B0281.5 and AEDGDS for B0281.6 (these sequences have been added to the BLIXEM window), but this homology has shifted to the first reading frame (underlined). A splice to a position inside the predicted first exon would then remove nonhomologous residues of the predicted B0281.4 protein, yielding a full coding-region match between B0281.4, B0281.5, and B0281.6, and other homologs. The stop codon found for E02C12.7 is an interesting example. E02C12.7 is one predicted gene in a cluster of tandem duplications. E02C12.6, E02C12.7, E02C12.8, E02C12.9, E02C12.10, E02C12.11, and E02C12.12 show strong homology with each other and with an unlinked predicted gene, F56A4.5. Realigning the predicted protein sequences revealed that E02C12.8 forms a gene unit with E02C12.7, as does E02C12.12 with E02C12.11, and the end of E02C12.10 with E02C12.9. The protein-coding regions are interrupted by stop codons at a different position in each case, and these separate the previously predicted gene units. Whereas the start of E02C12.10 appears to be an intact gene unit, like E02C12.6, the middle of E02C12.10 is another truncated copy. Figure 4 | A stop codon in the annotated gene E02C12.7. A stop codon in the annotated gene E02C12.7. (A) A representation, derived from ACeDB, of the tandem gene cluster containing E02C12.6, E02C12.7, E02C12.8, E02C12.9, E02C12.10, E02C12.11, and E02C12.12, which shows homology to a putative choline kinase. The grey arrows indicate the extent of the annotated genes, with the gene structure predictions depicted at top. The black arrows represent the gene units after realignment of the coding sequences as described in the text. The ruler is marked in 500 base pair units. (B) A BLIXEM window from ACeDB with E02C12.8 and E02C12.7 depicted in the top half of the window. Each horizontal line is a homology and the region within the box, covering the gap between the two annotated genes, is expanded to give the lower part of the figure. The theoretical translation across this window is given in the three reading frames, (+1), (+2), and (+3), although only the third reading frame is relevant here. The homologous sequence for three other genes in this tandem cluster, E02C12.6, E02C12.10, and E02C12.12 is retained from the BLIXEM window, whereas other homologous sequences have been removed for clarity. The sequence homology extends upstream from the predicted start of E02C12.7 (MIIDFVPNIQ...) into the predicted intergenic region. A small intron would then link from the homology of E02C12.8 to that of E02C12.7, matching the coding regions of E02C12.6, E02C12.10, E02C12.12, and others, seamlessly. This perfect alignment only fails because of the stop codon (arrow) in the sequence VYCLK*FDNE, which led to the prediction of two gene units. In fact, E02C12.8 and E02C12.7 appear to form a single pseudogene. These apparently faulty genes could still be functional units having acquired novel function or mode of expression since their duplication, they could appear faulty because of errors in the sequence data, or they are fully defective genes, that is, pseudogenes. The final explanation seems the most likely. Although apparent faults were found in only 13 of the 74 annotated genes examined, a smaller proportion than expected to be pseudogenes from our interpretation of the reporter gene fusion results, the mode of analysis means that this must be an underestimate. Effectively, only annotated genes with close homologs in the C. elegans genome were assessed, and it may be easier to generate potential gene structures for more diverged gene units with lower sequence homology (26 of the 74). Furthermore, the examination was only cursory. Nevertheless, there is obvious sequence evidence suggesting that at least 13 of the 364 annotated genes randomly assayed using reporter gene technology are faulty and, by implication, that at least 4% of the annotated genes in the C. elegans genome are pseudogenes. DISCUSSION : Pseudogenes may be difficult to distinguish from functional genes by sequence analysis alone or even when combined with experimental analysis. The predominant fate of duplicated genes will be to accumulate mutations that render them nonfunctional pseudogenes. Premature stop codons and frameshifting mutations are the most obvious defining characteristics of a pseudogene, but gene structure prediction programs may find alternative splicing patterns around such obstacles, particularly if there is no good homology with a functionally well-characterized gene or EST data to act as a guide. Genes that have been disabled by damaged splice sites or promoters will be even harder to recognize as pseudogenes, and such genes may linger before genetic drift makes them clearly pseudogenes from inspection of the sequence alone. Although it might have been anticipated that the integrity of the protein-coding region of a recently duplicated gene may be more sensitive to genetic drift than the promoter, our results suggest that this may not be the case. The conservation of protein-coding regions beyond that of introns for these recently duplicated genes suggests that these genes were initially functional and subject to evolutionary selection before they became inactivated by genetic drift. These findings raise many questions about the evolution of the C. elegans genome and, more generally, molecular evolution. We suggest that many of the considerable number of recently duplicated genes in the C. elegans genome, being in fact pseudogenes, explains the low rate of reporter gene expression among recently duplicated genes. This implies that the C. elegans genome contains substantially fewer real genes than current annotation suggests, and that as many as a fifth of the predicted genes are pseudogenes. Other sequenced animal genomes may contain fewer pseudogenes, which could have made this problem easier to detect in C. elegans. Nevertheless, this problem may be present , but harder to deal with in other species in which gene structure is even more difficult to predict, and experience gained with C. elegans may guide this aspect of genome annotation. METHODS : Reporter Gene Fusion Construction and Analysis | Generation of the reporter gene fusions involved standard molecular biology procedures as described previously . Genomic DNA fragments were derived from the standard wild-type strain, Bristol N2. The vectors were modified from pPD21.28 (lacZ), pPD95.67 (gfp), or pPD95.70 (gfp) by insertion of a 31-bp frameshifting cassette between the multiple cloning site and the reporter gene. This cassette allows the reading frame to be corrected simply by digestion with either AscI or NotI, depending on the shift needed and recircularization. Expression of the reporter gene was examined in wild-type N2 C. elegans that had been transformed by microinjection, using the dominant marker gene rol-6(su1006) to identify the transgenics. Details of the genomic DNA fragments assayed and expression patterns obtained are presented on our web site and have been submitted to ACeDB/WormBase. All assayed plasmids are available on request. Bioinformatic Analysis | All 74 annotated genes in the duplicated category, which had failed to drive reporter gene expression, were analyzed. The predicted protein sequence was extracted from ACeDB/WormBase and used in a BLASTP search of WormPep. The 5 to 10 closest homologs of each, the precise number depending on the distribution of E values obtained, were aligned using CLUSTALX (version 1.81; ). Obvious defects in what were otherwise excellent alignments were sought by direct inspection. A gene would be investigated further if the predicted protein product lacked several consecutive amino acid residues that were highly conserved across the protein family. The gene structure prediction in the vicinity of the potential defect was examined using BLIXEM in a local version of ACeDB. Translations in all three reading frames and homologys identified in ACeDB, both presented in the BLIXEM window, were searched for the missing protein-coding region. Frequently, part of the missing coding region could be found, but had been omitted from the gene structure prediction because a smaller coding-region deletion, a translational reading frame-shift, or a stop codon prevented their inclusion in any potentially functional gene structure. WEB SITE REFERENCES : ; The Hope laboratory web pages with descriptions of reporter gene expression patterns. ; Access to the Fire laboratory web pages with descriptions of the reporter gene plasmid vectors. ; the NCBI (National Centre for Biotechnology Information) home page. ; The C. elegansBLAST server at The Sanger Institute. ; The C. elegans database WormBase. Backmatter: PMID- 11997350 TI - rVista for Comparative Sequence-Based Discovery of Functional Transcription Factor Binding Sites AB - Identifying transcriptional regulatory elements represents a significant challenge in annotating the genomes of higher vertebrates. We have developed a computational tool, rVISTA, for high-throughput discovery of cis-regulatory elements that combines clustering of predicted transcription factor binding sites (TFBSs) and the analysis of interspecies sequence conservation to maximize the identification of functional sites. To assess the ability of rVISTA to discover true positive TFBSs while minimizing the prediction of false positives, we analyzed the distribution of several TFBSs across 1 Mb of the well-annotated cytokine gene cluster (Hs5q31; Mm11). Because a large number of AP-1, NFAT, and GATA-3 sites have been experimentally identified in this interval, we focused our analysis on the distribution of all binding sites specific for these transcription factors. The exploitation of the orthologous human --mouse dataset resulted in the elimination of >95% of the similar58,000 binding sites predicted on analysis of the human sequence alone, whereas it identified 88% of the experimentally verified binding sites in this region. Keywords: Introduction : A major challenge of the postgenome-sequencing era is decoding the regulatory networks underlining gene expression. In eukaryotes, modulation of gene expression is achieved through the complex interaction of regulatory proteins (trans-factors) with specific DNA regions (cis-acting regulatory sequences). Intensive efforts over several decades have identified numerous regulatory proteins, transcription factors (TF), whose sequence-specific DNA binding activity is central to transcriptional regulation. Traditionally, DNA binding specificity of many TFs has been experimentally determined primarily with in vitro techniques such as DNase I footprinting and electromobility shift assay (EMSA) . Recently, alternative techniques such as expression DNA microarrays, in silico oligonucleotide binding, and phylogenetic footprinting have been adopted to identify DNA targets for TFs . Unfortunately, despite the fact that the binding sites of many TF have been experimentally defined, most TFs bind to short (6 --12 base pairs [bp]), degenerate sequence motifs that occur very frequently in the human genome. The binding specificities of these factors can be summarized as position weight matrices (PWM) that are compiled in various databases such as the TRANSFAC database . Pattern-recognition programs such as MATCH or MatInspector use these libraries of TF-PWMs to identify significant matches in DNA sequences. A major confounding factor in the use of PWMs to identify transcription factor binding sites (TFBSs) is that only a very small fraction of predicted binding sites are functionally significant. Accordingly, the use of PWMs has proved to be a poor resource for sequence-based discovery of biologically relevant regulatory elements . In complex organisms, gene expression results from the cooperative action of many different proteins exerting different effects in time and space. Multiple TFs are simultaneously required to cooperatively activate and modulate eukaryotic gene expression . One potential avenue for improving the discovery of functional regulatory elements is to identify multiple TFBSs that are specifically clustered together . This strategy has been successfully implemented in the analysis of regulatory regions involved in muscle and liver-specific gene expression . An additional powerful strategy that has been shown to counter the large numbers of false positives derived from the analysis of sequences from a single organism is the use of multispecies comparative sequence alignments or phylogenetic footprinting . Several recent studies have shown that noncoding regulatory sequences tend to be evolutionarily conserved and support the use of comparative genomics as an extremely effective tool for the discovery of biologically active gene regulatory elements . The computational algorithms developed to perform comparative sequence analysis are based either on local alignments (BLAST ; PIPMaker ) or on global alignments (AVID; VISTA ), both of which have proved very efficient in detecting regions of high DNA conservation. To facilitate the efficient and accurate identification of regulatory sequences in large genomic intervals from complex organisms, we have developed a computational tool, Regulatory VISTA (rVISTA: ), that enriches for evolutionarily conserved TFBSs. rVISTA uses orthologous sequence analysis and clustering to overcome some of the limitations associated with TFBS predictions of sequences derived from a single organism. Here we introduce the rVISTA program and illustrate its ability to identify functional TFBSs as it dramatically reduces the total number of AP-1, NFAT, and GATA-3 sites predicted in a similar1-Mb genomic interval of the well-annotated cytokine gene cluster (Hs5q31; Mm11) . RESULTS : Computational Design of the rVISTA Program | To take advantage of combining sequence motif recognition and multiple sequence alignment of orthologous regions in an unbiased manner, rVISTA analysis proceeds in four major steps: (1) identification of TFBS matches in the individual sequences, (2) identification of globally aligned noncoding TFBSs, (3) calculation of local conservation extending upstream and downstream from each orthologous TFBS, and (4) visualization of individual or clustered noncoding TFBSs . The program uses available PWMs in the TRANSFAC database and independently locates all TFBS matches in each sequence with the MATCH program. A global alignment generated by the AVID program and the corresponding sequence annotations are used to identify aligned TFBS matches in noncoding genomic intervals. Figure 1 | rVISTA data flow. rVISTA data flow. The user submits a global alignment file (generated by the AVID program) and optional annotation files for the two orthologous sequences. The imported TRANSFAC matrix library and the MATCH program are consequently used to identify all transcription factor binding site (TFBS) matches in each individual sequence and to generate a file with all TFBS matches in the reference sequence (used as baseline for visualization). Next, the global alignment and the sequence annotations provided are used to identify all aligned TFBSs present in the noncoding DNA (in the absence of annotation, the program will identify all aligned sites across the entire alignment). A second file is generated containing aligned noncoding TFBSs. DNA sequence conservation is determined by the hula-hoop module, which identifies TFBSs surrounded by conserved sequences and generates a data table with detailed statistics. The final data processing step includes a user-interactive visualization module. The user customizes the data by choosing which TF sites to visualize (we are giving an example for choosing GATA-3 sites), what TRANSFAC parameters to use for all TF matches (rVISTA default 0.75/0.8), and by selectively clustering individual or combinatorial sites. The user can customize the clustering of any of the three data sets (all matches in the reference sequence are depicted as blue tick marks, aligned TFBS matches are in red, and conserved TFBS matches are in green). An aligned TFBS represents a region in the global alignment that corresponds to identical TFBS matches in each orthologous sequence. Orthologous regions correspond to similar DNA sequences from different species that arose from a common ancestral gene during speciation and are likely to be involved in similar biological functions. Because the global alignment forces two closely related sequences to generate the best possible pairwise alignment by introducing gaps, an aligned TFBS site can be present in a region of poor DNA conservation that is below 80% ID. To identify TFBSs present in regions of high DNA conservation, the "hula hoop" component of the algorithm calculates DNA conservation for each aligned TFBS as percent identity (% ID) over a dynamically shifting window of 21 bp that centers on a nucleotide inside the TFBS with the maximum % ID. This process identifies TFBSs located at the edges of highly conserved sequences that would falsely fall below the established conservation criteria threshold if the DNA conservation was determined by a static DNA window perfectly centered on the TFBS alignment. By use of the same principle, rVISTA calculates the maximum DNA conservation over larger DNA segments (up to 201 bp) facilitating the identification of sites present in larger, highly conserved regions. The rVISTA algorithm generates two types of outputs: (1) a static data table with detailed statistics for all aligned TFBSs and (2) a dynamic web-interactive module that allows the user to customize the data for unfiltered, aligned, or conserved TFBS sites and graphically visualize them as colored tick marks. Visualized conserved binding sites fit the criteria of >=80% ID over a 21-bp region. Combinatorial Analysis of TFBSs with Multiclustering | Detailed molecular analyses addressing the architecture of complex regulatory regions in higher eukaryotes have established that the majority of transcriptional control elements such as enhancers and repressors represent a conglomerate of multiple TFBSs that act in concordance to directly modulate the expression patterns of the linked genes . In addition, it has been observed that regulatory elements involved in similar physiological functions, such as the enhancement of liver-specific genes , are associated with distinct patterns of coordinate TF binding. These regulatory regions are frequently present in clusters of two or more repeated sites for the same TF or in combinatorial clusters of two or more adjacent sites belonging to unique regulatory proteins that act together to modulate gene expression . To analyze combinations of multiple TFBSs and identify TF binding patterns that control gene expression in novel sequences, rVISTA calculates the distance between all neighboring TFBSs and allows the user to perform customized clustering of individual or multiple unique transcription factors. One clustering module allows the user to selectively cluster two or more sites of the same TF present in regions of user-defined lengths (Fig A, B), facilitating the identification of evolutionarily conserved elements that harbor multiple clusters of various unrelated TFBSs. A second clustering module allows the user to identify groups of multiple TFBSs present in DNA segments of user-specified length (Fig C). Figure 2 | Visualizing rVISTA cluster analysis for a 25-kb region across the GM-CSF and IL-3 genomic interval. Visualizing rVISTA cluster analysis for a 25-kb region across the GM-CSF and IL-3 genomic interval. (A) Ikaros-2 TFBS clusters (two sites over 60-bp region). Ikaros-2 matches fitting the clustering criteria for the human sequence alone are depicted in blue, aligned clusters in red, and conserved clusters in green. (B) Multiclustering of individual sites can be performed by independently choosing the clustering criteria for each TF. AP-1 (blue), NFAT (red), and GATA-3 (green) clusters (two sites over 100 bp) of conserved TFBSs are illustrated. (C) Combinatorial clustering of TFBS. By use of the clustering criteria of 1 NFAT and 1 AP-1 across a 60-bp DNA fragment, the rVISTA program identifies all the AP-1 (blue) and NFAT (red) paired and displays them as tickmarks. This clustering module can be applied to the three data sets allowing the visualization of clusters in the reference sequence, among the aligned sites, and the conserved sites. Collection of Experimental Data and Validation of rVISTA | To evaluate the biological significance of TFBS data generated by the rVISTA algorithm, we analyzed similar1 Mb of a well-annotated cytokine gene cluster (Hs5q31; Mm11) (IL-3; IL-4; IL-5; IL-13; IRF-1; GM-CSF) plus the intensively characterized cytokine 2 (IL-2) promoter region (Hs4q26; Mm3) . Cytokines are of particular biomedical importance because they augment the growth and differentiation of T helper cell subsets and have been directly implicated in having a major role in determining susceptibilities to asthma phenotypes and inflammatory disorders . As such, much interest has focused on the regulatory mechanisms by which naive helper CD4+ T cells establish their cytokine repertoires, events that are predominantly regulated at the transcriptional level. Because of the vast interest in understanding the regulation of cytokines, we focused our analysis on transcription factors known to transcriptionally activate these genes. One of the best known examples of cooperative binding is the NFAT : AP-1 TF complex that has been described for genes involved in various immune responses. NFAT and AP-1 synergistically form stable complexes with DNA sequences that contain composite elements of adjacent NFAT and AP-1 TFBSs to induce the expression of genes . We have compiled a representative collection of AP-1 and NFAT experimentally defined TFBSs from the published data on this similar1-Mb interval and used it to examine the ability of rVISTA to identify true TFBSs. By analyzing similar925 kb noncoding human sequence independent of the mouse sequence, the MATCH program predicted 23,457 AP-1 and 14,900 NFAT sites with the PWMs available in the TRANSFAC database for these transcription factors (parameters: 0.75/0.8). A comparable number of sites were independently predicted for the orthologous mouse sequence. Among the large number of predicted AP-1 and NFAT sites for the human sequence were also included 17 of the 19 functional AP-1 sites and 19 of the 21 functional NFAT sites (Fig. A). The omitted AP-1 and NFAT functional sites failed to meet the TRANSFAC default parameters. Table 1 | Localization of Functionally Characterized Binding Sites to Conserved Blocks Figure 3 | rVISTA analysis algorithm identifies experimentally characterized TFBS. rVISTA analysis algorithm identifies experimentally characterized TFBS. (A) Two functionally characterized NFAT/AP-1 clusters indicated by black vertical arrows ([two sites/60 bp] [Table : E7 and E8]) are identified by rVISTA and are the only two clusters of conserved TFBSs present in the IL-5 promoter. The VISTA alignment highlights exons in blue, UTRs in yellow, and conserved noncoding in red. (B) A GATA-3 pair in the IL-5 promoter indicated by black vertical arrow is highly conserved and represents the only functional GATA-3 cluster ([2 GATA/60 bp] ) in the proximal promoter (500 bp upstream of the 5'UTR) of this cytokine. Subjecting the orthologous human and mouse sequences to rVISTA analysis reduced the total number of predicted AP-1 and NFAT sites by >95%, identifying 1114 conserved AP-1 and 734 conserved NFAT sites. rVISTA also identified 16 of the 19 AP-1 and 19 of the 21 functionally characterized NFAT sites. Whereas only 4.5% of the total NFAT and AP-1 predicted sites for the human sequence were conserved in the orthologous mouse sequence, in sharp contrast, 88% of the experimentally defined TFBSs were present in highly conserved DNA blocks. This data establishes a strong correlation between the presence of TFBSs in regions of high DNA conservation and biological function . However, only a small percentage of the total identified conserved sites correspond to functional sites that have been experimentally verified. Some of the other conserved TFBSs may also be functional but remain to be experimentally confirmed. Table 2 | Enrichment of Functional TFBS Using Clustering and DNA Conservation Cytokine Promoter Analysis to Assess rVISTA Predictions | In addition to AP-1 and NFAT, the GATA-3 TF has also been implicated in the transcriptional control of the large number of Th2-specific cytokines present in this interval (IL-4, IL-13, IL-5, GMCSF, IL-3) . GATA-3's direct involvement in gene activation has been extensively shown for the IL-4 and IL-5 promoters and has been postulated for the activation (or repression) of all the cytokine genes present in this interval . On the basis of GATA-3's predicted binding to upstream regions of cytokine genes, we hypothesized that there should be an increased distribution of GATA-3 sites across the six cytokine promoters compared with the promoters of the 16 non-Th1/Th2 expressing genes in this region. To test this hypothesis, we determined the GATA-3 site distribution for the 2-kb promoter region of all 22 annotated genes in this interval. Because of the highly degenerate nature of the GATA binding profile that is recognized by all members of the GATA-family , TRANFAC predicted an average of 50 GATA-3 sites per promoter that were evenly distributed across both cytokine and noncytokine gene promoters. In contrast, the rVISTA analysis dramatically reduced the total number of GATA-3 sites per promoter and, most importantly, resulted in an increased representation of GATA-3 sites in cytokine promoters (Fig. A). On average, rVISTA detected eight conserved GATA-3 sites per cytokine promoter while yielding only two conserved GATA-3 sites per noncytokine promoter. In addition, the experimentally characterized GATA-3 sites in both the IL-4 and IL-5 promoters were among the highly conserved sites identified by rVISTA (Fig. B). Figure 4 | Distribution of conserved GATA-3 binding sites across the 22 promoter regions (2 kb upstream of 5'UTR) of all annotated genes from the 1-Mb cytokine gene cluster (Hs5q31; Mm11). Distribution of conserved GATA-3 binding sites across the 22 promoter regions (2 kb upstream of 5'UTR) of all annotated genes from the 1-Mb cytokine gene cluster (Hs5q31; Mm11). Cytokine genes are labeled by arrows, gray bars indicate observed GATA-3 sites, and open bars represent predicted GATA-3 sites as a result of random distribution. Random distribution was estimated on the basis of the frequency of GATA-3 sites across the 1-Mb human sequence and the DNA conservation of each promoter. (A), conserved individual GATA-3 sites. (B), conserved GATA-3 present in clusters (two or more conserved sites enclosed in a 60-bp DNA fragment). Because functional GATA-3 sites are present in pairs , we next analyzed the distribution of GATA-3 sites clustered (two or more sites present within 60-bp regions). By clustering the conserved GATA-3 sites, we observed a further enrichment of GATA-3 sites in the cytokine promoters. In each cytokine promoter there were an average of six GATA-3 clustered sites, whereas no such clustered sites were noted in the promoters of noncytokine genes. These GATA-3 clustered sites, although not yet experimentally verified, were exclusively found in the promoters of genes predicted to be GATA-3 responsive. rVISTA's ability to recognize what are likely true TFBSs in the promoters of cytokine genes supports the hypothesis that GATA-3 plays an important role in the regulation of all the cytokine genes present on human 5q31 (Fig. B). DISCUSSION : Annotating the noncoding portion of the human genome remains among the greatest challenges of the post-sequencing era. Clues for identifying sequences involved in the complex regulatory networks of eukaryotic genes are provided by the presence of TFBS motifs, the clustering of such binding site motifs, and the conservation of these sites between species. rVISTA takes advantage of all these established strategies to enhance the detection of functional transcriptional regulatory sequences controlling gene expression through its ability to identify evolutionarily conserved and clustered TFBSs. By performing an unbiased analysis of the distribution of NFAT and AP-1 binding sites across similar1 Mb of human/mouse orthologous region, we were able to show that although rVISTA reduces more than 95% of the predicted TFBSs derived from the sequence analysis of a single organism, it still recognizes 88% of the biologically characterized AP-1 and NFAT in this region. The PWM compiled from experimentally determined TFBSs available in the TRANSFAC database pose a major limitation in the rVISTA analysis, because the computational approach described relies on the available DNA binding profiles of known transcription factors . Of the total 19 AP-1 and 21 NFAT experimentally described sites, 17 AP-1 and 19 NFAT sites had TRANSFAC values greater than 0.75/ 0.8, two AP-1 and one NFAT site had values of 0.7/0.7, and one NFAT site had a value of 0.6/0.7 . Of the 36 experimentally defined AP-1 and NFAT sites recognized by the PWMs available in the TRANSFAC database (with the 0.75/0.8 parameters), only one aligned AP-1 site (71%) was below our established conservation threshold (>=80%) and failed to be identified by the rVISTA program. Our data indicates that the rVISTA program dramatically eliminates a large number of false-positive TFBSs while it enriches for functional TFBSs. Although the identification of conserved TFBSs on a small genomic interval can be achieved by phylogenetic footprinting , a great strength of the rVISTA algorithm is its ability to efficiently analyze large genomic intervals and potentially whole genomes. The clustering modules and the user-defined customization of visualized sites makes this a further useful tool for the investigation of TFBSs. Through the use of a global alignment, rVISTA takes into account the linear structure of sequence conservation across a large DNA segment. By allowing small gaps and DNA shifts in the aligned TFBSs, we are maximizing the identification of functional TFBSs that have diverged slightly yet are present in highly conserved regions; similar25 --35% of all aligned and similar15 --20% of all conserved TFBSs identified have one gap in their alignment (data not shown). In addition, similar25% of the aligned and similar18% of the conserved NFAT and AP-1 sites have shifts (1 --6 bp) across their alignments (data not shown). The presence of gaps in the alignments of experimentally characterized TFBSs further supports the use of a global alignment for rVISTA analysis and the need for loose parameters for the identification of aligned sites, as well as stringent percent identity criteria for detecting highly conserved TFBSs. Properties related to protein --protein interaction and chromatin structure, as well as clusters of multiple unique sites that have been reshuffled in one of the human or the mouse genome and have lost their positional linearity, are not addressed. Also, clustering does not take into account the spacing between sites but rather counts the number of adjacent sites of a given TF spanning DNA segments of specified length. Although TFBS clustering has been suggested for identifying regulatory sequence, no data to date has proved the effectiveness of this approach . Our clustering analysis results indicate that this approach has the potential to efficiently prioritize functionally relevant noncoding sequences. rVISTA represents the only publicly available program that allows the user to identify customized clusters of multiple TFBSs in large genomic intervals. Our analysis of the AP-1 and NFAT TFBS in the cytokine gene cluster illustrates the effectiveness of the rVISTA algorithm in eliminating many false positives while retaining the majority of experimentally verified sites. In our analysis of GATA-3 sites in the putative promoters of the 23 genes from human 5q31 we were able to prioritize, exclusively on the basis of sequence analysis, a limited number of GATA-3 sites with a high likelihood of being functional that can be used for further biological investigation. With the increasing availability of sequence data for multiple organisms, rVISTA's ability to use comparative data and clustering options in a user-friendly manner makes it particularly suited to assist investigators focused on biologically defined genomic intervals, as well as those interested in performing whole genome analyses to identify functional TFBSs and regulatory elements. METHODS : rVISTA is implemented as a publicly available web-based tool that requires a sequence alignment file and optional gene annotation files as user input. The rVISTA analysis tool consists of four major modules: (1) motif recognition, (2) identification of aligned TFBSs, (3) conservation analysis, and (4) visualization of TFBSs. The system units are implemented with the C++ computer language equipped with Web user-interactive interface written in Perl. For the conservation analysis rVISTA uses an alignment file in the AVID format obtained using the AVID program from the AVID or VISTA servers. The following methodological scheme was implemented as a core for the rVISTA tool. Initially, the user chooses a set of TFs and the PWM parameters to be used. Next, rVISTA extracts all TFBS coordinates independently in the two orthologous sequences before the analysis of the alignment. The locally installed TRANFAC 5.2 library and the MATCH program from Biobase, Inc. are used at this step. Subsequently, the global alignment is scanned for pairs of neighboring human and mouse TFBSs that are aligned and match identically in both sequences. An aligned TFBS is allowed to have a maximum 6-bp shift (majority of TF matrices have core sequences of 4 --6 bp) in the alignment of the TFBS core and a single gap present across the entire local alignment of the TFBS. The conservation analysis module contains one major unit, the hula hoop, which is designed to analyze the local DNA conservation of each aligned TFBS to eliminate aligned sites present in regions of weak DNA conservation. A fixed-size DNA window (21 bp) is being shifted through all the positions of an aligned TFBS, whereas the entire sequence spanning the TFBS is permanently enclosed by the shifting DNA window. The percent identity is calculated at every base pair across the aligned TFBS and extending 10 nucleotides upstream and downstream from it, similar to a hula hoop. The position with the highest percent identity is used to assign the conservation level of that particular TFBS. This process allows the identification of the maximum percent identity for the local alignment of a conserved TFBS. The program calculates % ID for each binding site with dynamically shifting windows of up to 200 bp. These data are provided in a table format and allow the identification of TFBSs present in large regions of high conservation. The visualization module is a web-based tool that post-processes the rVISTA output. One unit of the program eliminates redundancy. Overlapping TFBS matches (within 3 bp from each other) belonging to the same family of regulatory proteins are considered to be an identical match. A second unit of the program measures the distance between adjacent matches belonging to the same TF family and allows the user to selectively cluster TFBSs into groups of x number of sites over y base pair length. The clustering parameters are user-defined and are assigned independently for every family of TF. Any combination of unfiltered, aligned, or conserved TFBSs with customized clustering for the selected set of TFs are interactively visualized as a 'tick-plot' track overlaid on the conservation VISTA-type track and the gene annotation track. All conserved binding sites displayed fit the criteria of >=80% over a 21-bp alignment. Evaluation of True TF Enrichment with GATA Sites in Promoters | To quantitatively measure the enrichment of GATA predictions for functional sites, we performed a statistical simulation for the expected number of conserved sites and compared it with the observed number of conserved sites (>=80% ID; 21 bp) present in promoter regions (2 kb upstream of the 5'UTR). Redundant GATA sites (defined to be overlapping sites) were excluded before the analysis. GATA site clustering was performed for two or more neighboring GATA sites present over a <=60-bp region. The upper bound for the expected number of conserved GATA sites in a promoter under consideration, i, was calculated as follows: (d = 41 bp; d is the average distance between two nonredundant GATA sites in the human sequence; 1/d is the probability of a given nucleotide to be at the starting position for a GATA site; and li is the number of nucleotides inside a promoter region conserved >=80%.) To obtain a more accurate value for the number of expected conserved and aligned GATA sites, we also considered the fact that a human GATA site present in a region of high DNA conservation (>=80%) will not always have an aligned match in the mouse sequence. Any given conserved GATA site could either have no corresponding aligned binding site in the second sequence or the binding site alignment could exceed the gap and shift requirements. We approached this problem by introducing a scaling parameter sigma. (sigma is the probability of a conserved site to be aligned and is approximated to be a constant for all the promoter regions.) The estimation for the sigma value was calculated on the basis of the number of conserved sites that were also aligned in noncytokine promoters. We estimated the expected distribution of conserved GATA sites across the promoter sequences as follows: Thus, we were able to compare the distribution of observed conserved and aligned GATA sites in cytokine promoters with the estimated distribution of such sites extrapolated from noncytokine promoters (Fig. A). Similarly, we estimated the number of expected conserved, aligned, and clustered GATA sites. The length of the conserved and clustered segment of each promoter was obtained by checking all the possible paired coordinates in the conserved regions of all promoters and obtaining the ratio of sites closer than 60 bp but greater than 4 bp apart from each other. The sigma values for conserved and for conserved and clustered sites were found to be 0.23 and 0.19, respectively. Close sigma values for these two types of sites (conserved; conserved and clustered) indicate that the probability of a conserved site to be aligned is independent from the probability of the same site to be clustered (Fig. B). WEB SITE REFERENCES : ; AVID program. ; rVISTA program. ; TRANSFAC database. ; MATCH program, Biobase, Inc. Backmatter: PMID- 11997347 TI - Protein Coding Palindromes Are a Unique but Recurrent Feature in Rickettsia AB - Rickettsia are unique in inserting in-frame a number of palindromic sequences within protein coding regions. In this study, we extensively analyzed repeated sequences in the genome of Rickettsia conorii and examined their locations in regard to coding versus noncoding regions. We identified 656 interspersed repeated sequences classified into 10 distinct families. Of the 10 families, three palindromic sequence families showed clear cases of insertions into open reading frames (ORFs). The location of those in-frame insertions appears to be always compatible with the encoded protein three-dimensional (3-D) fold and function. We provide evidence for a progressive loss of the palindromic property over time after the insertions. This comprehensive study of Rickettsia repeats confirms and extends our previous observations and further indicates a significant role of selfish DNAs in the creation and modification of proteins. Keywords: Introduction : Interspersed repeated DNA sequences are usually confined in the intergenic regions of bacterial genomes. However, Rickettsia appears to be a unique exception in this respect. In our previous work, we identified evolutionarily related sequences of 50 amino acid residues dispersed in protein-coding regions of Rickettsia conorii and other Rickettsia . The peptide segments showed no sequence similarity to known protein domains. On the other hand, the corresponding nucleotide sequences (similar150 bases) showed imperfect palindromic (self-complementary) properties that resemble other bacterial intergenic repeats like IRU and RSA . The repeats were designated as Rickettsia palindromic elements (RPEs). On the basis of the predicted locations of the inserts in the three-dimensional (3-D) folds of proteins, and on the observed transcript sizes, it is most likely that the repeat-derived peptide is expressed as part of the proteins encoded by those open reading frames (ORFs) . Thus the RPE appears to have a unique capability to spread over the coding, as well as the noncoding, regions of the bacterial genome. The completion of the genome sequence of R. conorii revealed a high density of repeated sequences in the genome . In this study, we systematically analyzed the repeat locations in regard to coding versus noncoding regions. Three different palindromic sequence families showed clear cases of insertions into ORFs. In addition, several palindromic sequences were identified within RNA coding genes. We also found that the palindromic property of the repeats has a tendency to be dimmed over time after their insertions in the genome. RESULTS : We identified 656 interspersed repeated DNA sequences in the genome of R. conorii. On the basis of sequence similarity, the repeated sequences were classified into 10 distinct families . There is no significant sequence similarity between the various repeat families. Their copy numbers range from 5 to 223. Nucleotide sequence alignments of the repeated sequences are shown in Figure . A coloring scheme is used to help visualize the predicted RNA secondary structures in the alignments. Of the 10 families, eight showed palindromic sequences with consensus sizes from 95 to 149 bases. Stable RNA secondary structures were predicted for most of the sequences in those eight families. The predicted secondary structures showed hairpin-like forms or variants with additional branched stems. However, the precise base-pairing pattern in the structures varied across, as well as within, the repeat families. The eight palindromic sequence families were named RPE-1 to RPE-8. The previously reported 44 RPEs were classified into the RPE-1 family, for which 11 additional copies have been identified in this study. The two remaining families are composed of shorter repeats (25 bases and 27 bases) showing no stable predicted secondary structure. They were designated as Rickettsia repeat-1 (RR-1) and repeat-2 (RR-2). Table 1 | Interspersed Palindromic Repeats and Short Repeats in Rickettsia conorii Figure 1 | Nucleotide sequence alignments of the 10 repeat families identified in Rickettsia conorii. Nucleotide sequence alignments of the 10 repeat families identified in Rickettsia conorii. The coloring scheme is equivalent to the bracket representation of RNA secondary structure; the bases predicted to make base pairing are represented by the same color (instead of opening and closing brackets). The consensus sequences were derived from the aligned positions in which a single base is observed in at least 70% of the sequences (lower cases for 70% --90%; upper cases for >=90% conservation). ClustalX alignment quality scores are shown at the bottom of the alignment. Exhaustive BLAST searches against sequence databases revealed that all of the 10 families except the RPE-6 are specific to Rickettsia species or limited to R. conorii . The RPE-6 repeat contains two directly oriented RS3 core motifs (ATTCCC-N8-GGGAAT) frequently found in Neisseria genomes . All the repeat families are relatively GC-rich (similar40% of GC) compared with the average GC content of the entire genome (32%), with the exception of the AT-rich RPE-8 (22%). Size variation within family is large for six families (RPE-3 to RPE-8). More than 50% of the identified repeats are "partial" repeats for those RPEs. Size variation is relatively small for the other families (RPE-1 and RPE-2; RR-1 and RR-2), which are mostly composed of "full-length" copies (see Methods). The analysis of the whole 656 repeat locations revealed a large number of insertions within ORFs for seven RPEs (RPE-1 to RPE-7) and a single occurrence for RR-1 . Those cases include full-length repeat insertions within "annotated" ORFs (ORFs with functions predicted by homology), as well as partial repeat insertions within ORFans (ORFs lacking similarity in other organisms). None of the insertions interrupts the reading frame of the host ORFs. Table shows the 38 ORFs harboring the full-length repeats. Of the seven palindromic families found within ORFs, three families (RPE-1 to RPE-3) showed a number of full-length insertions into annotated ORFs. Most of those ORFs appear to be important for R. conorii, because they constitute parts of biological pathways or molecular complexes involving many other genes that are present in the R. conorii genome . However, there is no apparent functional relationship between the predicted functions of those altered ORFs. Table 2 | Full Size Repeat Insertions within the Rickettsia conorii Genes By use of BLAST, we examined the occurrence of the homologous repeats in the Rickettsia prowazekii genome . Interestingly, three copies of the RPE-3 were found within the ORFs of R. prowazekii: RP037 (putative O-sialoglycoprotein endopeptidase) and two ORFs of unknown functions (RP012 and RP707). Each of the three ORFs has a clear ortholog in R. conorii, which lacks the repeat insert. We previously reported nine cases of the RPE-1 within R. prowazekii ORFs . In addition, a copy of an RS3-like element has been found in the N-terminal of the alpha subunit of DNA polymerase III (DnaE) in Rickettsia felis . Thus, the insertions of palindromic sequences within ORFs seem to be a widespread phenomenon in different Rickettsia species. Multiple alignments of the peptide sequences derived from the RPE-1 to RPE-3 are shown in Figure . Each of the three RPEs shows distinct peptide sequences with unbiased amino acid compositions. The peptide sequences are well aligned within the family and correspond to the same reading frame. A remarkable feature of those RPEs is the capability of occupying any site, even in the middle, along the primary sequences of the ORFs . For instance, an RPE-1 sequence is located in the middle part of the R. conorii MesJ protein. The predicted protein secondary structure is a central alpha-helix for the RPE-1 and the RPE-3. In contrast, extended conformations (beta-strands) were predicted at both extremities of the peptide sequences derived from the RPE-2. Those three peptide families may thus show different 3-D folds. Figure 2 | Multiple sequence alignments for the amino acid sequences translated from the palindromic sequences (RPE-1, RPE-2, and RPE-3). Multiple sequence alignments for the amino acid sequences translated from the palindromic sequences (RPE-1, RPE-2, and RPE-3). Amino acid residues are colored as follows: blue for F, W, and Y; yellow for C; orange for A, G, P, S, and T; green for I, L, M, and V; red for D, E, H, K, and R; purple for N and Q. The letters 'H' and 'E' in the first line of each alignment represent the predicted alpha-helices and beta-strands, respectively. ClustalX alignment quality scores are shown at the bottom of the alignment. We then examined the insertion sites of the RPE-derived peptides with 3-D structure data for the homologs of the host proteins. Seven protein structures for RPE-1 insertions , two for RPE-2 and two for RPE-3 , were available for this analysis. In all cases, the insertion site corresponded to the solvent-exposed area of the proteins: mainly loops (nine cases) and occasionally short helices (one case) or beta strands (one case). Furthermore, none of the predicted insertions appeared to hinder known catalytic sites or protein/cofactor binding sites. Four cases corresponding to the ORFs with RPE-2 and RPE-3 insertions are shown in Figure . Figure 3 | Predicted insertion sites of the RPE-2 (a, b) and RPE-3 (c, d). Predicted insertion sites of the RPE-2 (a, b) and RPE-3 (c, d). Green segments labeled by the green arrows in the reference structures indicate the insertion sites. (a) Bovine mitochondrial ATP synthase. gamma chain is shown in red. (b) Streptococcus pneumoniae rRNA methyltransferase used as a reference for the R. conorii dimethyladonosine transferase. (c) E. coli UDP-N-acetylmuramoylalanine --D-glutamate ligase. (d) Pig prolyl oligopeptidase used as a reference for the R. conorii protease II. Seven cases corresponding to the predicted insertions sites for the RPE-1 have been previously reported . The images are generated with MolScript . We identified several RPEs within RNA genes. The tmRNA coding genes (ssrA) of R. conorii and R. prowazekii harbor palindromic sequences of different families (an RPE-1 for R. conorii and an RPE-5 for R. prowazekii). tmRNA is an RNA molecule present in all known bacterial genomes. Its function is to rescue the ribosome stalled on an mRNA . tmRNA is composed of a tRNA-like domain and an mRNA-like domain (Fig. a). The structure of the tmRNA genes from alpha-proteobacteria , together with the locations of the RPE-1 and RPE-5, are shown in Figure , b and c. The two insertions of the palindromic sequences within the tmRNA genes were both located right after the CCA-(3') bases of the acceptor arm, where alanine is added by alanyl-tRNA synthetase. We examined the transcription status from the R. conorii tmRNA gene by reverse transcriptase-polymerase chain reaction (RT-PCR). The RT-PCR product showed the expected size with the repeat insert, demonstrating transcription of RPE-1 with the rest of the genes. Because there is no detectable sequence similarity between the RPE-1 and the RPE-5, those two insertions must have occurred independently in the different lineages of Rickettsia. Another case found in R. conorii is an RPE-1 within a ribozyme gene, rnpB, which encodes M1 RNA of the ribonuclease P. The insertion site corresponded to the P12 helix of the RNA secondary structure model of M1 RNA (Fig. d) . The P12 helix shows a highly variable sequence, and the helix is unlikely to involve functionally important tertiary interactions in vivo . Figure 4 | RPEs found within the RNA genes. RPEs found within the RNA genes. (a) The standard form of the tmRNA genes observed in most bacterial species. (b) The permutated form of the tmRNA gene observed in alpha-proteobacteria . (c) Locations of the RPE-1 and RPE-5 inserted in the tmRNA genes of Rickettsia. Approximate locations of the primers are indicated by triangles. (d) The secondary structure model of the M1 RNA and the insertion site of the RPE-5 in R. conorii. The palindromic property (hairpin-like secondary structure) of the RPEs is probably required for repeat insertion as suspected for IRU and RSA . Alternatively, the hairpin structures might have an important function for Rickettsia. In the former case, the hairpin-like structure might lose its utility after insertions and could disappear over time. In the latter case, nucleotide secondary structures should be conserved despite sequence changes (as in ribosomal RNAs). To investigate the significance of the palindromic property of the RPEs, we computed the minimum free energy for every sequence of the RPE-1 to RPE-8 and obtained the relevant Z-score (see Methods). If we take Z-score > 2 (P-value < .0423) as a threshold, the energy values for 10/45 sequences (22%) of the RPE-1 failed to be significant. Such energy values below the threshold were also observed for the RPE-2 (4/7; 57%), the RPE-5 (6/20; 23%), and the RPE-8 (3/15; 20%). This result indicates that the palindromic properties of some repeats are unlikely to be constrained after their insertions. In Figure , the Z-score of the RPE-1 is plotted against the sequence divergence D, the average sequence difference against the other sequences of the RPE-1. The pair of most similar sequences, which might correspond to most recent inserts, showed very high Z-scores (Z = 8.30 and 8.24). The two sequences were identified in the truB gene and in RC0071. The 144 bases sequences are 94.4% identical with each other. There is also a global tendency for the better conserved sequences to show higher stabilities of the RNA secondary structures (the correlation coefficient is R = -0.74; P < .005). This decay of palindromic property indicates the absence of structural constraints on the repeats after their insertions. The lack of significant differences in the structural stability between the coding and the noncoding repeats also argues against a specific role of the palindromic structures at the transcription and translation levels. Figure 5 | The minimum free energy of the predicted RNA secondary structures of the RPE-1 plotted against sequence divergence. The minimum free energy of the predicted RNA secondary structures of the RPE-1 plotted against sequence divergence. The energy values are scaled into Z-score. Circles correspond to the repeats in coding regions. Triangles correspond to the repeats in noncoding regions. The arrows point to the two most similar sequences. DISCUSSION : In this study, we identified 10 families of repeated sequences in the genome of R. conorii and examined their locations in regard to coding versus intergenic regions. Three palindromic sequence families (RPE-1, RPE-2, and RPE-3) showed clear cases of insertions within predicted coding regions, and eight families in total (RPE-1 to RPE-7 plus RR-1) showed insertions within coding regions including ORFans. Therefore, the surprising mechanism of repeat insertion within protein coding regions initially described for RPE-1 applies to many other repeat families in Rickettsia. The analysis of the locations and the sequences of the repeat-derived peptides (RPE-1, RPE-2, and RPE-3) reinforced our previous observations . First, they are inserted into ORFs with only one reading frame of six possibilities, as indicated by the aligned sequences of the repeat-derived peptides . Second, there are no clear functional links between the ORFs harboring the repeats. Third, the insertion sites of the repeats vary along the primary sequence of the ORFs but always appear compatible with the preexisting protein folds. This study revealed two additional aspects. First, the predicted protein secondary structures for the three RPEs (RPE-1 to RPE-3) correspond to two different conformations. alpha-Helices were predicted for the RPE-1 and RPE-3, whereas beta-strands were predicted for the RPE-2. Thus, the two regular conformations in protein structure (alpha-helix and beta-strand) could occur from repeat insertions. However, the possibility is not ruled out that those peptides are "neutral" and might adapt variable conformations in response to the surrounding structural environment at the insertions sites. Circular dichroism spectroscopy failed to show any property of the regular conformation for synthetic peptides (similar50 amino acids) corresponding to the RPE-1 (C. Abergel, unpublished data). Four repeat-containing proteins have been expressed in Escherichia coli (V. Monchois et al., in prep.), and experiments are in progress to determine the structural properties of the repeat-derived peptides within these proteins. Second, we showed that some copies of the RPEs do not exhibit a significant palindromic structure. Because the pair of most similar RPE-1 sequences correspond to highly stable hairpins, it is plausible that this secondary structure is a feature of the original copies that are mobile within the genome. The structural property might then be lost after the insertion regardless of the site of the genome (coding or noncoding), as the initial repeat continued to diverge in both sequence and structure. However, the possibility is not ruled out that some of the repeats have been recruited for host cellular functions or recombination , as already suggested for other bacterial repeats. Proteins contain structurally flexible regions, usually corresponding to surface loops. Such loops are known to be tolerant of insertions of individual amino acids or peptides. For instance, insertions of peptides between 7 and 17 residues into a loop of the chymotrypsin inhibitor-2 (64 amino acids) have little effect on the stability and the folding rate . This physical flexibility parallels the evolutionary flexibility of protein sequences. Available sequence and structure data indicate a high preference of insertions and deletions within loops . However, most (99%) of the accepted insertions and deletions are shorter than 10 amino acid residues. In contrast, the palindromic repeats described in this study could contribute to insertions up to similar50 residues. Repeat insertion can freely occur within the 20% of R. conorii genome corresponding to the noncoding regions. If one accepts that surface loops account for a quarter of every protein sequence , another 20% of the genome (from the coding moiety) is available for additional repeat insertions. RPEs appear to invade both of the two genomic regions. The mechanism by which the bacterial palindromic sequences of the size of RPEs spread within genomes is not known . However, the coincidence of the two insertion sites of the palindromic sequences in the tmRNA genes (ssrA) of Rickettsia is intriguing. The upstream sequences of those insertion sites are highly similar to bacteriophage attachment site (att) . The homologous sites of the other tmRNA genes and tRNA genes have been known to harbor bacteriophages, retron phages, and pathogenicity islands in other bacteria . Such retron phages and pathogenicity islands are supposed to be integrated by use of the integrases of phages. Some of the RPEs might have used a similar mechanism. We have proposed that RPEs are selfish DNA elements that can break the barrier of genetic material between coding and noncoding sequences . Recursive insertions of such selfish DNAs might provide the initial genetic material for rather "neutral" protein segments, which could then later evolve to create new functions . The genomes of Rickettsia show by far the highest number of occurrence of such insertions, even if a few instances have been reported in other bacteria. For instance, a partial copy of RSA was found in the C-terminal of a hypothetical ORF of 99 amino acids in Salmonella typhimurium . Recently another case has been reported in Sinorhizobium meliloti. The DNA helicase II (UvrD) of this legume symbiont has a 47 amino acid residues insert encoded by a palindromic DNA sequence (motif C) . The ongoing accumulation of more bacterial genome sequences should lead to better appreciation of the importance of the phenomenon of repeat insertions in the origin and evolution of proteins. METHODS : The genomic sequence and annotation data for R. conorii are available at RicBase and NCBI GenBank (; accession no. ). Other complete genomes including those used in Table were obtained from KEGG . Database searches were performed with the NCBI BLAST package against the complete genomes as well as the NCBI nonredundant sequence database. Repeated DNA sequences of R. conorii were initially identified on the basis of the self-comparison of the genomic DNA by BLASTN (E-value < 10-4). The BLAST result was then analyzed to delineate the left and right edges of the repeated sequences with the repeat identification program Mocca . Some trivial repeats such as tRNAs or paralogous ORFs were removed from the dataset a posteriori. A complete list of the repeats described in this paper is available at RicBase. The "full-length" repeats were defined as the sequences with lengths within 70% to 100% of the longest repeat of the family. The remaining shorter sequences were defined as "partial" repeats. The nucleotide sequence alignments and the consensus sequences in Figure were constructed from the full-length repeats with T-Coffee and ClustalX . Sequence divergence D used in Figure was defined as the average sequence difference against the other sequences of the family. The definition of D does not take into account nucleotide secondary structures. In the computation of D, the positions with gaps in the pairwise alignments were omitted. The minimum free energy and the corresponding RNA secondary structures were computed with the Vienna package at . The minimum free energy value was then converted to Z-score. To compute Z-score, every sequence was randomly shuffled 30 times, from which the mean and the standard deviation values were computed. We used an approximation by the extreme value distribution to obtain the relevant P value. Protein secondary structures were predicted with PHDsec at . The following protein structure data were obtained from the Protein Data Bank : Bovine mitochondrial ATP synthase F1 domain ; Streptococcus pneumoniae rRNA methyltransferase ; E. coli UDP-N-acetylmuramoylalanine --D-glutamate ligase ; and Pig prolyl oligopeptidase . The rRNA methyltransferase was used as a reference for R. conorii dimethyladenosine transferase (KsgA); they belong to the rRNA adenine N-6-methyltransferase family. The prolyl oligopeptidase was used as a reference structure for R. conorii protease II (PtrB); they both belong to the prolyl oligopeptidase family. The presence of the RPE-1 in the transcript from the tmRNA gene of R. conorii was assessed by RT-PCR by use of a primer pair, P1 (5'-TAA TTT AGA ATA GAG GTT GCG GAC T-3') and P2 (5'-CGT TTG CGT TTC TTT GTT TT-3'), designed to be specific to the target gene. The expected size of the RT-PCR product was 311 bp including the RPE-1 (146 bp). For RNA extraction, a suspension of fresh R. conoriistrain Malish (seven) was adjusted to 108/mL, and bacteria were separated from cells with a sucrose gradient. RNA extraction from bacteria was then performed with the RN easy Midi kit (Qiagen, Hilden, Germany) as recommended by the manufacturer. RT-PCR was performed on the resulting RNA with the One-step RT-PCR kit (Qiagen) following the manufacturer's instructions. Reverse transcription and amplification were performed in PTC-200 thermocyclers (MJ Research, Watertown, USA) with 40 PCR cycles and an annealing temperature of 50C. RT-PCR products were run in 1% agarose gels, stained with ethidium bromide, and revealed on a UV box. Following the RT-PCR assay, we performed a PCR assay with the same primers on the RNA extract with the Elongase polymerase (Life Technologies, Cergy Pontoise, France). The PCR assay was negative, thus verifying the absence of contaminating DNA in the RNA. The RT-PCR product was sequenced with the same primers with the d-Rhodamine terminator cycle-sequencing ready reaction kit (PE Applied Biosystems, Les Ulis, France) and an ABI-PRISM 3100 automated DNA sequencer (PE Applied Biosystems), as recommended by the manufacturer. An RT-PCR product of 311 bp was obtained from the R. conorii RNA. The sequence of the RT-PCR product was 100% identical to the genomic sequence of the R. conorii tmRNA gene containing the RPE-1. WEB SITE REFERENCES : ; PHDsec. ; RicBase. ; MolScript. ; NCBI GenBank. ; Protein Data Bank. ; Vienna package. Backmatter: PMID- 11997344 TI - Analyses of the Extent of Shared Synteny and Conserved Gene Orders between the Genome of Fugu rubripes and Human 20q AB - Cosmid and BAC contig maps have been constructed across two Fugu genomic regions containing the orthologs of human genes mapping to human chromosome 20q. Contig gene contents have been assessed by sample sequencing and comparative database analyses. Contigs are centered around two Fugu topoisomerase1 (top1) genes that were initially identified by sequence similarity to human TOP1 (20q12). Two other genes (SNAI1 and KRML) mapping to human chromosome 20 are also duplicated in Fugu. The two contigs have been mapped to separate Fugu chromosomes. Our data indicate that these linkage groups result from the duplication of an ancestral chromosome segment containing at least 40 genes that now map to the long arm of human chromosome 20. Although there is considerable conservation of synteny, gene orders are not well conserved between Fugu and human, with only very short sections of two to three adjacent genes being maintained in both organisms. Comparative analyses have allowed this duplication event to be dated before the separation of Fuguand zebrafish. Our data (which are best explained by regional duplication, followed by substantial gene loss) support the hypothesis that there have been a large number of gene and regional duplications (and corresponding gene loss) in the fish lineage, possibly resulting from a single whole genome duplication event. [Reagents, samples, and unpublished information freely provided by D. Barnes and I.D. Hickson.] Keywords: INTRODUCTION : The pufferfish (Fugu rubripes) is now an established tool for comparative gene and genome analyses. Fuguis only a distant relative to human and has one of the smallest vertebrate genomes, but herein lies its success as a comparative analysis tool (for review, see ). The genomic features that are well conserved between the two species are the ones that are fundamental to vertebrate existence ---the genes and functional noncoding elements. Because of its compact size (400 Mb), and relative scarcity of repeat sequences, gene hunting and sequencing in Fugu is relatively simple . Whereas the pufferfish has been used for comparative gene and short-range regional comparisons, the regions studied to date have generally been over distances covered by single BACs and cosmids. Although the pufferfish genome is similar7.5x smaller than the 3000-Mb human genome , estimates of gene numbers for Fugu are similar to those for human . The extent of genomic data available for the pufferfishes and human is, however, vastly different. At the time of writing this letter, both Fugu, via a sequencing consortium , and the freshwater pufferfish Tetraodon nigroviridis are the subject of genomic shotgun sequencing projects, with similar2x genomic coverage currently available for each. Both, however, lack detailed physical and genetic maps. For human, both physical and genetic mapping data are extensive, and the complete sequence, although unfinished in places, is all but complete. The only other fish species that has been so well studied is the zebrafish (Danio rerio). Work on zebrafish has focused on the examination of genetic mutants and pathways (for review, see ; ) and on the creation of expressed sequence tag (EST) resources and radiation hybrid and genetic linkage maps . Although the zebrafish is also being sequenced, this effort has yet to make real headway, and the types of data available for Fugu and zebrafish are therefore quite different. The short sections of assembled pufferfish sequence that are available have not been mapped to individual linkage groups. In zebrafish, long-range linkage data lacks complementary short-range analyses of gene content and orders. The types of genomic and evolutionary analyses that are possible between both fishes and human are therefore still relatively limited. The analysis presented here is an attempt to compare gene content and gene orders for regions of both fish genomes that share synteny with one region of the human genome, the long arm of chromosome 20. Comparisons of gene content and orders in different species are helping us dissect the physical processes that have allowed the development of complex vertebrate genomes. Analyses of vertebrate gene content with species similar to those present at the base of the vertebrate lineage have, for example, indicated that there were two rounds of whole genome duplication in the early evolution of vertebrates. In particular, studies of hox gene complexes have shown that although there is one hox gene cluster in the cephalochordate amphioxus, there are four paralogous copies in human and mouse (for review, see ). Opinions do, however, differ, and recent analyses of the timing of the expansion of vertebrate gene families have also been used to argue against two rounds of simultaneous (whole genome) duplication (e.g., see ). Within the vertebrate lineage, analyses of different genomes can also highlight similarities and differences in genome structures. Recent analyses from zebrafish, for example, have indicated that an additional whole genome duplication may have occurred in zebrafish evolution. In zebrafish, seven hox complexes have been identified , and recent analyses of gene families have indicated that many other gene families also have further additional copies in the fish lineage . Recent analyses in zebrafish have also identified conserved synteny groups between the zebrafish and human genomes . Instead of one orthologous chromosomal segment being present for each human chromosomal region, however, two paralogous chromosome segments have been commonly found in the zebrafish. In many cases, these paralogous segments each contain one copy of a single copy human gene. Knowledge of this kind is of particular importance to researchers using model organisms to dissect gene functions and structures. Functions may be shared or divided between duplicated genes, or duplicates may have adopted new roles within the organism. As a whole, the zebrafish data indicates that an additional genome duplication may have occurred in the fish lineage, but it is currently unclear whether this duplication occurred before the separation of Fugu and zebrafish. There is currently little Fugu data on gene duplication or paralogous linkage groups, although an analysis of the Fugu hox clusters has indicated that at least one has been duplicated in the Fugulineage . In this study, we have generated some of the longest sections of Fugu contig data available to date and used these to conduct comparative analyses between Fugu and human as well as between Fugu and zebrafish. In doing so, we have found both duplicated genes and paralogous genomic segments. The two Fugu regions studied were initially selected for their shared synteny with human 20q11.2 --12, a region commonly deleted in human myeloid malignancies. When a human 20q gene probe for topoisomerase 1 (TOP1) was used, however, two Fugu genes were identified. We have focused on creating gene contigs and examining the extent of Fugu and human shared synteny around these Fugu genes. Both contigs form Fugu paralogous linkage groups, each sharing synteny with human 20q. The contigs, which span 65 identified Fugu genes in total, have allowed cross-species comparisons of shared synteny, gene content (including copy numbers), and gene orders. These two regions (containing duplicated Fugu genes) are also comparable to a pair of paralogous regions in the zebrafish, which indicates they were duplicated before the separation of the two fish lineages. They provide evidence that the two genomes have undergone similar evolutionary processes and suggest that the Fugu genome may also have been duplicated. RESULTS : Two Fugu Contigs Sharing Synteny with Human 20q | We have used BAC and cosmid clones to construct two Fugucontigs across regions sharing synteny with human 20q. Clones from both Fugu contigs have been sequence scanned to allow the examination of gene content . Our experience has shown that this is generally an effective method of determining the vast majority of genes present . In general, the cosmid and BAC coverage is good, although in some places only one clone was available for the sequence scanning process. Genes have been named according to the orthologous human gene unless they have been previously named in Fugu. Where official names for the human genes did not exist, these genes are referred to by the corresponding Unigene cluster. Gene orders have been determined by use of a combination of complete sequencing, sequence scanning, and sequence tagged site (STS) mapping techniques. In most cases, the sequence scanning and STS mapping techniques have allowed the definitive ordering of the identified genes, whereas the depth of the sequence scanning data has allowed the identification of the probable human orthologs of the Fugu genes present. In some cases, however, a Fugu gene has been found to have close sequence similarity to more than one human gene or a number of gene family members. In these cases, it has been difficult to determine the true human ortholog and its corresponding map location. Figure 1 | Tiling paths for scanned BACs in the two Fugu contigs. Tiling paths for scanned BACs in the two Fugu contigs. An additional 14 cosmids have been scanned to generate data for Fugu contig 1 and a further 13 cosmids for Fugu contig 2. In total, 64 BACs and 77 BACs were used in the STS mapping of the genes on contig 1, whereas 43 BACs and 25 cosmids were used to create the contig 2 gene map. The contigs have been physically placed onto Fugu chromosomes by FISH. The larger contig, covering similar400 kb of genomic sequence, is composed of 64 BACs and 77 cosmids of which 14 BACs (of which two were coligated) and 15 cosmids have been sequence scanned (Fig. A). In total, 42 genes have been identified, of which 20 have orthologs on Hsa20q (Fig. A). In addition to the section of this contig, which shares synteny with human 20q, this contig contains a region containing seven genes with orthologs on human 3p21 and a short section of genes with putative orthologs predominantly on human chromosomes X and 7. A number of the Fugu genes in this latter segment of DNA are members of gene families with high sequence similarity. One of the genes, for example, is a member of the methyl CpG binding domain (MBD) family. This has made it difficult to determine orthologous human genes, especially when there appears to be no unique human genome segment sharing synteny with this region of Fugu DNA. In Figure A, therefore, alternative human orthologs are shown for a number of genes. Elsewhere in both contigs, assigning probable orthologs has been less problematic, but when a human ortholog could not be assigned with a high degree of confidence, this is denoted with a question mark following the gene name in Figure . Figure 2 | Gene content on the two Fugu contigs. Gene content on the two Fugu contigs. Genes are named according to the nomenclature for the orthologous human gene. Human mapping locations are shown in italics after the gene name. For genes mapping to Hsa20, locations are shown as positions on bacterial clones in the Hsa20 contig map. In contig 1 there is a clear break in shared synteny with human 20q, whereas in contig 2 synteny with 20q is conserved along the entire contig length. If a gene is duplicated in Fugu, or the Fugu gene has already been named elsewhere, the corresponding Fugu gene name is shown in italics. The relative positions of the duplicated Fugu genes are indicated by arrows. The smaller of the two contigs spans similar250 kb and contains 43 BACs and 25 cosmids, of which 13 cosmids and 13 BACs (Fig. B) have been sequence scanned. Despite the presence of two genes with probable mapping locations elsewhere in the human genome, synteny with human 20q is well conserved across the entire contig (which contains 22 genes) (Fig. B). One end of this contig is flanked by an uncharacterized retroelement. Verifying the contig assembly across this element has proved problematic, but our additional data (not presented here) indicate that synteny with human 20q is not conserved after this point. At the other end of the contig, all of the genes present have orthologs on Hsa20q, and it is likely that shared synteny with human 20q would continue if the contig were extended further in this direction. Given the size of the contigs, and the numbers of genes present, this equates to a gene every 10 kb. Gene distribution on the two contigs appears relatively uniform, with approximately five genes detected on each individual cosmid and no large gene voids. If this figure were extended to the whole Fugu genome, this would give a total of similar40,000 genes in the 400-Mb sequence. This figure is comparable to current estimates of gene content in the human genome and is slightly lower than some of the previous (65,000 --70,000) estimates for Fugu . Duplicated Genes | Sequence analyses of two of the duplicated genes, the Fugu top1 (Fugu top1alpha and Fugu top1beta) and snail genes (Fugu sna1 and Fugu sna2) , have already shown that these genes were duplicated in the fish lineage. In the case of the snail genes, this duplication can be placed before the separation of the Fugu and zebrafish lineages. The lack of full-length top1 gene data from other fish species has meant that we have been unable to date the top1 duplication, but degenerate PCR analysis has also shown the presence of two top1 gene fragments in zebrafish . The third duplicated gene found on both Fugu contigs is a maf gene. These genes are members of the basic region/leucine zipper (b-Zip) superfamily and are transcription factors associated with the regulation of cell differentiation . Members of the maf superfamily can be further divided into the small and large maf proteins. In human, one of the large maf genes, KRML, a member of the mafb gene family, has been mapped to human 20q and has been placed on PAC dJ644L1 similar400 kb away from human TOP1 and PLCG1 in the draft human sequence. The ortholog of this gene in zebrafish (krml1/val) has been mapped to linkage group 23 on the Oregon gene map, again close to plcg1 . In zebrafish, a second homolog, krml2, has also been identified. This gene has been mapped to zebrafish linkage group 11 . The coding regions of the two maf genes have been sequenced in Fugu, and phylogenetic analyses (not shown) show that the KRML gene on the Fugu top1alpha contig is the probable ortholog of zebrafish val and that these are both members of the mafb gene family. The second Fugu KRML is similar to zebrafish krml2, but neither of these genes have high sequence similarity to other members of the mafb gene family. Chromosomal Locations of the Contigs | To determine whether the two paralogous Fugu regions are linked or separated in the genome, BACs from each contig were mapped onto Fugu metaphase chromosomes. BACs B299O09 (from the contig containing Fugu top1 alpha, Fig. A) and B294L23 (from the contig containing Fugu top1beta, Fig. B) were found to map to different Fugu chromosomes (Fig. A). Figure 3 | (A) FISH mapping of nonsyntenic BACs B294L23 (contig 1) and B299O09 (contig 2) to Fugu metaphase chromosomes. (A) FISH mapping of nonsyntenic BACs B294L23 (contig 1) and B299O09 (contig 2) to Fugu metaphase chromosomes. The two Fugu contigs are therefore not linked in the Fugugenome. (B) FISH mapping of syntenic BACs B234K08 and B294L23 (both contig 1) to Fugu metaphase chromosomes. These BACs, which span the breakpoint in shared synteny between contig 1 and human 20q, show no evidence of chimerism. The second Fugu contig has at least two apparent breakpoints in shared synteny with the human genome. To check the integrity of this contig, BAC clones were selected for FISH mapping to Fugumetaphase chromosomes. BAC B234K08 (Fig. A), which spans the genes KRML to TFE3, and B294L23 (Fig. A), which spans much of the region sharing synteny with 3p21, were both confirmed as being syntenic and nonchimeric (Fig. B). Comparisons with Human 20q | On both contigs, within the segments sharing synteny with human 20q, there are a small number of genes that appear to map elsewhere in human. Given the amount of human sequence data now available for 20q, it is thought unlikely that more similar genes will be found on this chromosome. It is likely that these genes have been translocated onto this chromosome segment in the Fugu lineage or, conversely, have been moved from this region in human. The three duplicated genes with orthologs on human 20q are colinear on the two Fugu genomic segments and in the same order (krml-top1-sna) as they are found in human. The order of these genes does not therefore seem to have been disrupted by inversions in either lineage. The orders of many of the other 20q genes are different in the Fugu and human lineages , indicating that other gene orders have been disrupted by localized inversions. The dramatic differences in gene orders between the two species may be clearly seen by comparing gene orders between the two Fugu contigs and smaller sections of human 20q. A comparison of the region spanning 20q11.22 --20q12, for example , shows that genes present on this section of human 20q are split between the two contigs in Fugu. This indicates a pattern of gene duplication in the Fugu lineage followed (in the majority of cases) by loss of one of the two gene copies. There does not appear to be any pattern associated with the loss of the second gene copy, as the retained genes are distributed quite evenly between the two Fugu regions. Figure 4 | Fugu/Human 20q comparative gene maps. Fugu/Human 20q comparative gene maps. The positions of the human 20q orthologs of genes identified on the two Fugu contigs are shown on an ideogram of human chromosome 20. Mapping locations for the human genes span the whole of human 20q. Genes are named according to the nomenclature for the orthologous human gene. If a gene is duplicated in Fugu, the corresponding Fugugene name is shown in italics. The positions of the Fuguand human orthologs are linked by black lines. Genes that have been identified as being present in two copies in Fugu are linked to the mapping position of the corresponding human gene by dashed lines. Although Fugu contig 1 has been inverted in this diagram to reduce the number of crossover events, the number of changes in gene orders between the Fugu contigs and human 20q remains high. This indicates that gene orders in the two lineages have been disrupted by a number of localized rearrangements. Figure 5 | Fugu/human 20q11.2 --12 comparative gene map. Fugu/human 20q11.2 --12 comparative gene map. This ideogram provides a more detailed comparison between the two Fugu contigs and a 7-Mb section of human 20q (a section that contains the human orthologs of two of the genes, KRML and TOP1, known to be duplicated in Fugu). The orthologs of similar25% of the genes mapping to this region are represented on one or another of the two Fugu contigs. The distribution of the genes between the two Fugu contigs, however, appears to be random, indicating a pattern of duplication in the Fugu lineage followed by the loss (in the majority of cases) of one of the additional gene copies. When comparing the Fugu and human gene maps, it is clear that many genes mapping to human 20q are not represented on the two assembled Fugu contigs. This is characteristic of the two contigs, which contain only a subsection of the genes now known to map to human 20q. It is, however, apparent that the human orthologs of identified Fugu genes tend to be clustered in short blocks separated by longer tracts of genes for which no Fugu orthologs have been identified. This indicates that the ancestral chromosome segment that gave rise to these two Fugu contigs lacked the additional genes now seen on human 20q. It is not clear whether these genes were, however, present in the same ancestral linkage group. If this were the case, it is likely that additional genes with orthologs on human 20q could be found by further extension of these regions in Fugu. Comparisons with Zebrafish | To determine whether these linkage groups are conserved in both Fugu and zebrafish, we have tried to identify the mapping locations of the zebrafish orthologs of the identified Fugu genes. In some cases this data was already published. In others, in silico searches of EST mapping and cluster databases could be used to find mapping locations for orthologous zebrafish genes. This information is summarized in Table . Table 1 | Conserved Linkage Groups in Fugu and Zebrafish The ten genes from the larger Fugu contig, which have also been mapped in zebrafish (including genes such as gata1 and rcp , which do not have orthologs mapping to human 20q), all map to zebrafish linkage group 11 . We were unable to find zebrafish mapping locations for any of the genes from this contig that have orthologs on human 3p21. The published mapping data for zebrafish linkage group 11 do, however, indicate that this linkage group also shares synteny with Hsa3p21 . The seven genes from the smaller Fugu contig that have also been mapped in zebrafish all map to zebrafish linkage group 23. Our data therefore indicate that (at least for these regions of the genome) there have been relatively few interchromosomal translocations in the 150 million years since the divergence of Fugu and zebrafish. Detailed comparisons of the Fugu and zebrafish gene orders have not been possible because of the fact that the zebrafish genes have been separately localized onto a number of different mapping panels. DISCUSSION : These mapping data have allowed the comparison of a large section human 20q and two chromosome segments sharing conserved synteny in Fugu. Large-scale mapping data are generally not yet available in Fugu, and previous studies have focused on comparatively mapping short regions of the human genome. Despite this, there are other Fugu data that indicate additional paralogous linkage groups containing duplicated genes do exist in Fugu. Studies of genes mapping to human 9q34, for example, have shown that the Fugu orthologs of 9q34 genes are also divided between two different Fugu chromosomes, and that at least one of the 9q34 genes studied (DNM1) is duplicated in Fugu . In addition to the duplicated genes that we have identified as part of this project, two further genes (from the contig region sharing synteny with 3p21) are also duplicated in Fugu. In a study of the plasminogen-related growth factor (PRGF) genes in Fugu, identified two paralogs,prgfr2 and prgfr3, which both mapped to regions sharing synteny with 3p21. Phylogenetic analysis indicated that the duplication was specific to the fish lineage. One of these two genes, prgfr3, is now known to be present on the larger of the two contigs discussed here and is positioned next to a calmodulin-binding protein, a probable homolog of human CAMK1 (also from Hsa3p21). identified a second CAMK1 homolog neighboring, prgfr2. Thus the region sharing synteny with 3p21 is also known to be duplicated. Although the mapping location in Fugu for these genes is unknown, our data would indicate that prgfr2 and the neighboring calmodulin-binding protein will map to the same Fugu linkage group as the smaller of the two Fugu contigs presented here. These data show that Fugu, like zebrafish, contains pairs of linkage groups that share synteny with single chromosome segments in human. These contain the semiorthologs (both equally related but not necessarily equally similar) to single copy human genes (for definitions of orthology and paralogy, see ). The presence of duplicated genes on these paralogous linkage groups, in addition to their presence on different chromosomes, indicates that they were initially generated by an ancestral duplication, rather than a translocation event, followed by subsequent gene loss. The patterns of gene distribution, as well as gene loss and retention, between the paralogous linkage groups that we examined in Fugu and zebrafish indicate that most gene loss following the duplication event occurred before the ancestral separation of these two fish species similar150 million years ago (for a review of the evolution of bony fishes, see ). We have not tried to detect duplicates of any of the other 20q genes that we have identified; therefore, we are unable to say whether or not any of these are also present in duplicate in Fugu. Hence it has not been possible to estimate the retention rates for duplicated genes. These patterns of duplication of genomic segments and the partial retention of additional gene copies are similar to those seen in the zebrafish genome. It therefore seems likely that Fugu and zebrafish have undergone similar duplication events. If this is the case, we would expect to find increasing numbers of Fugulinkage group pairs that share synteny with single human chromosomes. This is not at odds with the Fugu data generated to date. In general, Fugu gene searches have been focused around single Fugu genes of interest. Single gene probes have been used to select genomic libraries for the target gene, and the examination of synteny has been a secondary rather than a primary focus. Unless the gene probe used were for a duplicated Fugu gene, paralogous Fugu linkage groups would not be identified. METHODS : Identification of Cosmid and BAC Clones | A Fugu genomic cosmid library available from the UK HGMP Resource Centre was probed with a human chromosome 20 TOP1 gene probe (A. Bench, Department of Haematology, University of Cambridge). This identified two sets of nonoverlapping cosmids each containing a Fugu top1 gene. Contigs were extended with cosmids from the same library; subsequently, BAC clones (Incyte Genomics) were also added. Sample Sequencing | Gene content on the cosmids and BACs was assessed by sequence scanning. Approximately 50 random sequences were generated from each cosmid and similar150 sequences from each BAC or PAC were generated with the methods of . Complete Sequencing and Analyses of Duplicated Genes | The complete sequencing of the duplicated snail and top1 genes has been detailed in and . Comparative analyses of the Fugu maf sequences were facilitated by assembling the two sets of genomic maf fragments in GAP4. Gaps in sequence coverage were closed by the design of primers at fragment ends and the use of PCR across gaps to produce templates for dye terminator sequencing. STS Mapping of Individual BACs and Cosmids | Primers for STS mapping were primarily designed from Fugusequences showing sequence similarity to known human genes. Additional STSs were designed from sequences showing no high similarity to known genes in regions of low gene density. Sequence templates were selected for high quality (as shown by a low percentage of ambiguous bases); where possible, primers were designed for regions where sequence data were confirmed by other sequenced subclones. Primers were designed with PRIME (Wisconsin Package, Genetics Computer Group Ltd.). STS primers were used on individual BACs and cosmids in the two Fugu contigs. Data were ordered manually and displayed with ExCelsior (UK HGMP Resource Centre). This allowed the ordering of both genes and genomic clones. Preparation of BAC Pools for STS Mapping | A Fugu genomic BAC library (InCyte Genomics), stored in 111 384-well microtiter plates, was used to prepare DNA pools for use as STS mapping templates. DNA was prepared from a single 384-well microtiter plate by inoculating four separate cultures with individual sets of 96 clones with a 96-pin replicating tool. Cultures were grown overnight and DNA from each 96-clone pool extracted with a standard alkaline lysis technique. All four DNA pools from a 384-well plate were then recombined to produce a single DNA pool representing the genomic content of a single 384-well microtiter plate. Ninety six of the 384-well plate DNA pools, representing plates 193 to 271 (79 plates) and plates 287 to 303 (17 plates), of the library were stored in 96-well microtiter plate format and used as templates for STS mapping. BAC Fingerprinting | Cosmids and BACs were mini-prepped in 96-well plate format with an alkaline lysis technique. Clones were digested with HindIII enzyme and the fragments separated on agarose gels. Gels were stained with SyBr Gold (Cambridge Bioscience) and scanned on a Typhoon 8600 phosphoimager (Molecular Dynamics). Gel images were tracked and bands called with Image (Sanger Centre). Clone fingerprints were then assembled into contigs with FPC . Comparative Sequence Analyses | Genomic sequences were used in BLASTv.2 sequence similarity searches of the SWISS-PROT , TREMBL , and EMBL databases. Search results were filtered through MSP-Crunch and viewed via the Fugu web pages at the UK HGMP Resource Centre. Sequence similarities to database entries were assessed on an individual basis, taking account of the sequence quality, organism, and sequence type matched, and the quality and length of the match (including 'Score' and 'Expect' values). The numbers of different matches to the same database entry were also considered. In some cases, in which BLAST matches were to ESTs, these were used to search the human UniGene database to allow assignment of matches to specific human UniGene clusters. Cases in which ESTs in the same UniGene cluster had been mapped also allowed human chromosomal locations to be found, even for uncharacterized human genes. Localizations for well-characterized human genes were also obtained by use of gene names in searches of GeneMap and LocusLink . BLAST similarity to human genomic PAC clones sequenced as part of the whole genome sequencing effort allowed more precise chromosomal localizations to be established for specific genes. Where BLAST similarity was to a number of similar human genes, probable human orthologs were identified through examination of the extent of shared synteny around the Fugu and human genes. These searches were repeated at frequent intervals to allow for comparisons with new database entries. These new searches became increasingly informative, allowing easier identification of probable human orthologs as human whole genome sequence data were submitted to the databases in increasing quantities. Final locations for human chromosome 20 genes (including locations on human bacterial clones) were found with Ensembl . Human gene names used are those found in LocusLink. Genes that have no official name have been identified according to the corresponding Unigene EST cluster. FuguBLAST EST data were also used to identify zebrafish ESTs with high sequence similarity to Fugusubclones. The names of the ESTs identified were used to search the zebrafish EST mapping data of the WashU-Zebrafish Genome Resources Project . This allowed the identification of zebrafish EST clusters for the corresponding zebrafish genes. Cases in which radiation hybrid mapping data were available for one or more of the clustered ESTs also allowed the probable mapping location of the zebrafish gene to be found. If an identified zebrafish EST was not in a mapped cluster but data were available for its paired read from the other end of the cDNA, this mapping location was used. Tissue Culture, Chromosome Preparation, and Banding | Metaphase chromosomes of Fugu rubripes were prepared from a fibroblast line maintained in Ham's F-12 medium supplemented with 15% fetal calf serum, 1 mM L-glutamine, and antibiotics. Cells were grown at 27C in a 5% CO2 incubator. For chromosome preparation, Colcemid (final concentration, 0.1 mug/mL medium) was added to the culture medium at least 4 h before harvest to arrest cells in metaphase. Cells were removed from the flask by gentle trypsinization. The cell pellet was resuspended in a hypotonic solution consisting of 50 mM KCl for 20 min. Cells were fixed overnight in a 3 : 1 mixture of methanol : acetic acid. Slides were prepared with the conventional 'drop-splash' technique. DNA Probes and FISH | DNA probes were labeled with either biotin-16-dUTP or digoxigenin-11-dUTP by nick translation. Standard FISH protocols were followed. Briefly, the slides were treated with 100 mug/mL RNase A in 2x SSC, pH 7.0 at 37C for 30 min and with 0.01% pepsin in 10 mM HCl at 37C for 10 min. After refixing the preparations for 10 min in 1x PBS, 50 mM MgCl2, 1% formaldehyde, they were dehydrated in an ethanol series (70%, 80%, 100%). Slides were denatured for 1 min at 90C in 70% formamide, 2x SSC, pH 7.0 and again dehydrated in an alcohol series. For hybridization of one slide, 400 ng of biotinylated and/or digoxigenated probe DNA were coprecipitated with 50 --100 mug sheared Fugu genomic DNA (as competitor for single-copy probes) and 10 --20 mug sheared human placental DNA (as carrier) and redissolved in 50% formamide, 10% dextran sulphate, 2x SSC. The hybridization mixture was denatured for 10 min at 80C. Preannealing of repetitive sequences DNA was performed for 30 min at 37C. Then the hybridization mixture was applied to each slide and sealed under a coverslip. The slides were hybridized for at least 3 d in a moist chamber at 37C. The slides were then washed three times for 5 min in 50% formamide, 2x SSC at 42C and once for 5 min in 0.1x SSC, pH 7.0 at 60C, before being blocked with 4x SSC, 3% BSA, 0.1% Tween 20 at 37C for 30 min. Probes were detected with fluorescein isothiocyanate (FITC)-conjugated avidin (Vector Laboratories) and Cy3-conjugated antidigoxin antibody (Dianova). Chromosomes and cell nuclei were counterstained with 1 mug/mL 4',6-diamidino-2-phenylindole (DAPI) in 2x SSC for 1 min. The slides were mounted in 90% glycerol, 0.1 M Tris-HCl, pH 8.0, and 2.3% 1,4-diazobicyclo-2,2,2-octane. Digital Imaging Microscopy | Images were taken with a Zeiss epifluorescence microscope equipped with a thermoelectrically cooled charge-coupled device camera (Photometrics CH250) controlled by an Apple Macintosh computer. Vysis imaging software was used to capture gray scale images and to superimpose the source images into a color image. WEB SITE REFERENCES : ; Ensembl home page. ; Fugu Genomics Projects. ; UK HGMP Resource Centre Fugu Group home page. ; WashU-Zebrafish Genome Resources Project home page. ; Genoscope Tetraodon nigroviridis home page. ; NCBI home page. ; the Zebrafish Information Network home page. Backmatter: PMID- 11997342 TI - Determination of Redundancy and Systems Properties of the Metabolic Network of Helicobacter pylori Using Genome-Scale Extreme Pathway Analysis AB - The capabilities of genome-scale metabolic networks can be described through the determination of a set of systemically independent and unique flux maps called extreme pathways. The first study of genome-scale extreme pathways for the simultaneous formation of all nonessential amino acids or ribonucleotides in Helicobacter pylori is presented. Three key results were obtained. First, the extreme pathways for the production of individual amino acids in H. pylori showed far fewer internal states per external state than previously found in Haemophilus influenzae, indicating a more rigid metabolic network. Second, the degree of pathway redundancy in H. pylori was essentially the same for the production of individual amino acids and linked amino acid sets, but was approximately twice that of the production of the ribonucleotides. Third, the metabolic network of H. pylori was unable to achieve extensive conversion of amino acids consumed to the set of either nonessential amino acids or ribonucleotides and thus diverted a large portion of its nitrogen to ammonia production, a potentially important result for pH regulation in its acidic habitat. Genome-scale extreme pathways elucidate emergent system-wide properties. Extreme pathway analysis is emerging as a potentially important method to analyze the link between the metabolic genotype and its phenotypes. Keywords: Introduction : Access to the complete genome of an organism provides a basis for studying cellular processes as a whole. Organism-level metabolic modeling is a "genome-enabled" science; without a sequenced genome, organism-level modeling is essentially impossible. The sequencing of entire genomes has given us a "parts list" for a cell, and now the challenge is to integrate those parts to understand the mechanisms and organization by which cells use these parts to achieve their phenotypic expressions. Extensive cataloging of biological components has enabled the reconstruction of genome-scale models of cellular metabolism . Metabolism involves the production of mass, energy, and redox requirements for all cellular functions, and thus provides the driving force for cellular activity. As one of the most thoroughly studied aspects of cellular function, it affords the best opportunity for the development of methodologies to characterize and analyze systems-level cellular properties of genome-scale models. A metabolic network consists of the group of reactions and transport processes associated with the production and depletion of cellular metabolites. Using genomic, biochemical, and physiological data, the metabolic pathways and transporters known to exist in an organism can be modeled as a network with a specific environment (Fig. A). Exchange fluxes that cross system boundaries are defined as input and output fluxes. The stoichiometric matrix concisely represents this information in mathematical form, with the rows corresponding to all the metabolites in the system and the columns representing all of the known biochemical reactions and transporters in the system. Figure 1 | A sample biochemical reaction network (A) and the convex representation of its metabolic possibilities (B). A sample biochemical reaction network (A) and the convex representation of its metabolic possibilities (B). The extreme pathways (EPs) in B correlate with those in A, serving to represent the relationship between extreme pathways of a biochemical reaction network and the characterization of the possible phenotypes. Metabolic pathways are commonly used as a means of simplifying the description and analysis of metabolic networks. Pathways have different definitions in the literature. Most commonly, however, metabolic pathways are defined as a linked set of biochemical reactions, in which the product of one reaction is the reactant of the subsequent reaction in the chain . With this definition, glycolysis, the Krebs cycle, and the pentose phosphate pathway are referred to as metabolic pathways. Segmenting a network into pathways is generally performed based on historical discovery, conceptual understanding, or certain heuristics. Extreme pathways, studied herein, are mathematically defined metabolic pathways calculated from a reconstructed metabolic network. Extreme pathways differ from other pathway definitions in that they not only are sets of connected reactions but also describe the functions of an entire metabolic network. Extreme pathways are to heuristically defined pathways what traffic patterns are to roads: Although knowledge of a road network can assist in evaluating how to get from one location to another, a knowledge of traffic patterns (dependent on weather, time, and road conditions) can more completely assess the capability of traveling . Heuristically defined pathways are linked reaction sets (roads), whereas extreme pathways describe the fluxes of molecules undergoing each reaction (flow of traffic along the roads). Extreme pathways are flux maps through a biochemical network that characterize the functioning of the network (Fig. A); in other words, extreme pathways account for the relative magnitude of the number of molecules undergoing each reaction in the cell. Extreme pathways have the following important characteristics: (1) they are generated for steady-state condition with no metabolite build-up allowed within the system, enabling the analysis of time-invariant properties related to the structure of the metabolic network; (2) they are a unique and minimal set of flux maps that can describe the network and are determined directly from the stoichiometric matrix; (3) extreme pathways, although contiguous, can have multiple inputs and outputs; and (4) the extreme pathways circumscribe all potential steady-state flux maps through the network. Extreme pathways can be represented graphically as the edges of a convex cone (Fig. B). Because all valid solutions are non-negative linear combinations of the extreme pathways (property 4), the convex cone circumscribes all potential metabolic phenotypes (i.e., all valid steady-state solutions of the network must lie within the cone). Thus, the extreme pathways characterize the extreme functions of the network; the network cannot attain any steady-state yields greater than that of the maximum-yield extreme pathways. Extreme pathway analysis of metabolic networks presents a mathematically rigorous approach to systems-level analyses. At a systems-level, emergent properties develop that can be determined by looking only at large-scale interactions rather than at individual components. Many other methods have been used to study cellular metabolism, including flux balance analysis , metabolic control analysis , and various types of pathway analysis . A detailed review of the development of extreme pathway analysis has been published elsewhere . In this study, we analyzed the extreme pathway structure of the H. pylori reconstructed metabolic network. This pathogen inhabits the gastric lining of nearly one half of the world's population , with a disproportionately high occurrence of infection in developing countries . It has received increasing interest for its role in various gastric-associated diseases, such as gastritis, peptic ulcers, and gastric cancer . The genome sequence of H. pylori was recently published for strains 26695 and J99 , enabling the reconstruction of its metabolic network and subsequent analysis. The in silico model used in this study is based on the genome sequence of H. pylori strain 26695 (; C.H. Schilling, M.W. Covert, I. Famili, G.M. Church, J.S. Edwards, and B.O. Palsson, in review.). Genome-scale extreme pathways were calculated for H. pylori using a previously described algorithm . Previous work has also been performed in the analysis of the extreme pathways of the metabolism of H. influenzae for the production of individual amino acids . Herein, we present a genome-scale analysis of the H. pylori metabolic network. With this organism, we were able to analyze the production not only of a single biomass compound (as was performed in the H. influenzae study) but also of significant subsets of biomass constituents, approaching a more complete model of how a cell produces all of its biomass constituents simultaneously. We studied the production of the set of non-essential amino acids, the set of ribonucleotides, the production of individual amino acids under various conditions, and the effect of urea in amino acid production. These studies resulted in large numerical data sets that have been analyzed to provide physiologically important characterizations. Definitions : A succinct definition of important terms that are used throughout the text is provided for clarity. Core allowable inputs included alanine, arginine, adenine, phosphate, sulfate, oxygen, histidine, isoleucine, leucine, methionine, phenylalanine, valine, and thiamin. Allowable outputs included ammonia, carbon dioxide, and the target product (outputted amino acids, ribonucleotides). Figure illustrates those inputs and outputs that were used by the model. Figure 2 | Diagrammatic representation of the in silico strain of Helicobacter pylori and its used exchange fluxes. Diagrammatic representation of the in silico strain of Helicobacter pylori and its used exchange fluxes. Other inputs from minimal medium (data not shown) were allowed, but only the inputs taken up by the system in this study are indicated. Carbon sinks include acetate, succinate, formate, and lactate. Target outputs include asparagine, aspartic acid, cysteine, glutamine, glutamic acid, glycine, lysine, proline, serine, threonine, tryptophan, tyrosine, the set of nonessential amino acids, and the set of nucleotides. Case 1 is the core inputs and outputs with succinate as an additional allowed output; case 2, core inputs and outputs with succinate and acetate as additional allowed outputs; case 3, core inputs and outputs with succinate, acetate, and formate as additional allowed outputs; and case 4, core inputs and outputs with succinate, acetate, formate, and lactate as additional allowed outputs. Exchange flux is the flux of a metabolite that crosses system boundaries. An input represents a substrate taken up by the cell, whereas an output can be thought of either as a molecule that is secreted or as a metabolite that is produced. An external state refers to the set of values for the exchange fluxes of a given extreme pathway. A flux map refers to the magnitudes of the fluxes through every reaction in a specific metabolic network, and the internal state is the flux map corresponding to an extreme pathway without the exchange fluxes. Pathway redundancy is a network property characterized by the degree to which multiple extreme pathways correspond to an identical external state (i.e., the number of internal states per unique external state), and the spectrum of yields is a rank-ordered representation of the yields for all pathways associated with the synthesis of the target product in this study. Nitrogen fate is the final destination of the input nitrogen of the system, that is, whether it ends up in the target product or as secreted ammonia. Nonessential amino acids are the amino acids that can be produced metabolically in H. pylori (essential amino acids are constituents of the minimal medium). Target product is the output of interest, constrained to be positive, and yield is the moles of the target amino acid produced per mole of carbon that enters the system. In this study, carbon enters in the form of alanine (three carbon atoms) and arginine (six carbon atoms). RESULTS : The extreme pathway structure of the reconstructed metabolic network of H. pylori was analyzed. The extreme pathways were calculated for the production of (1) the individual nonessential amino acids and ribonucleotides, (2) the equimolar set of nonessential amino acids, (3) the set of nonessential amino acids in physiologic Escherichia coli ratios (caused by the lack of data for amino acid composition in H. pylori), and (4) the equimolar set of ribonucleotides. These analyses resulted in large data sets, necessitating the development of an analytical framework to study them. The developed approaches elucidated important emergent properties of the defined metabolic network of H. pylori. Minimal Medium and Outputs Considered | The in silico strain of H. pylori was allowed inputs that constituted the minimal medium for growth (; ; ; C.H. Schilling, M.W. Covert, I. Famili, G.M. Church, J.S. Edwards, and B.O. Palsson, in review). Urea and adenine were also allowable inputs for some cases. The H. pylori model was allowed up to seven outputs: the target product (i.e., amino acid, ribonucleotide), carbon dioxide, acetate, succinate, formate, lactate, and ammonia. Acetate, succinate, formate, and lactate have been previously determined as the significant carbon byproducts of amino acid metabolism in H. pylori . This combination of outputs allowed the in silico model to produce all of the amino acids. Figure diagrams the exchange flux constraints implemented in this analysis. Substrates used by the model included alanine, arginine, adenine, oxygen, phosphate, and sulfate. Input fluxes that were allowed as part of the minimal medium, but which were never used for nonessential amino acid or ribonucleotide production, were histidine, isoleucine, methionine, phenylalanine, valine, and thiamin. These compounds are necessary for the cell and are taken up for growth, but do not participate in the biosynthesis of nonessential amino acids or ribonucleotides under tested conditions. Various cases with different allowed inputs or outputs were used in generating the extreme pathways for H. pylori. Each case is described in Definitions. It has been established that adenine is not a necessary component of the medium in which H. pylori grows . For this study, however, it was included in the minimal medium for the synthesis of the ribonucleotides. The model was not allowed sufficient byproduct secretion to synthesize adenine de novo. Adenine was subsequently accounted for in the redundancy, nitrogen fate, and yield calculations for ribonucleotide synthesis. Carbon-Nitrogen Substrates | It has been shown experimentally that in the presence of glucose and all amino acids, H. pylori completely consumed alanine, arginine, asparagine, aspartate, glutamine, glutamate, proline, and serine . Of this set, only alanine and arginine are components of the previously determined minimal medium. Our genome-scale extreme pathway analysis is in agreement with these data because the extreme pathways for amino acid and ribonucleotide production only used alanine and arginine from the available carbon and nitrogen sources. An in silico analysis was performed to ascertain the degree of dependence on alanine and arginine as substrates for production of amino acids. Table lists the number of extreme pathways for amino acid production that do not use either alanine or arginine as inputs. Only a few of these extreme pathways do not require alanine from the minimal medium. A notable exception is the extreme pathways associated with the production of proline. Production of proline always required the presence of arginine, but 11 of these pathways did not require alanine. On average, 25.3% of all of the extreme pathways for the production of the individual amino acids did not use arginine as an input. However, only 0.1% of the extreme pathways for individual amino acid production did not use alanine as an input. Thus, the extreme pathways necessary to characterize the metabolic network have a far greater dependence on alanine than arginine. Table 1 | The Number of Extreme Pathways Associated with the Production of the Indicated Amino Acid that Do Not Use the Indicated Carbon and Nitrogen Substrate Biomass Subsets and Nitrogen Fates | The metabolic network of H. pylori allowed the calculation of the extreme pathways for the production of multiple subsets of various constituents of the biomass. We computed the extreme pathways for the production of the set of nonessential amino acids in equimolar ratios. This equimolar set was also analyzed with urea as an allowable input of the metabolic network. The extreme pathways were also calculated for a different set of fixed ratios. The amino acid composition of H. pylori protein was not available, so we used the amino acid composition of E. coli protein as an approximation. The relative amount of amino acids in H. pylori protein was represented to one significant figure (asn-2, asp-2, cys-1, gln-2, glu-2, gly-6, lys-3, pro-2, ser-2, thr-2, trp-1, tyr-1); more precise ratios were computationally unfeasible. Although the ratios used were not ideal, they did help elucidate the extent to which the extreme pathways changed as the ratio of amino acid demand changed. We also analyzed extreme pathways involved in the production of the ribonucleotide set in equimolar ratios. Equimolar Amino Acid Production by H. pylori | Because the metabolic network was unable to produce glycine under the conditions represented in cases 1 and 2 , the equimolar nonessential amino acid set had a flux of zero for these cases. However, with the addition of formate and lactate as allowable outputs in cases 3 and 4, respectively, the network was capable of producing the complete equimolar set of nonessential amino acids . For the production of the individual amino acids in case 3, there were 6916 extreme pathways required to characterize the solution space. For the production of the equimolar amino acid set under otherwise identical conditions, there were 4991 extreme pathways needed to characterize the solution space. Thus, 1925 fewer extreme pathways were needed to characterize the solution space for the production of the set of amino acids than were needed to characterize the production of all of the amino acids independently. Under the system constraints in case 4, 6032 extreme pathways characterized the network for the production of an equimolar amino acid set, which was 2274 fewer pathways than those for the independent production of the amino acids under similar conditions. Thus, setting the amino acid composition that must be produced can be thought of as a type of network requirement that affects the number of extreme pathways needed to characterize the solution space. Table 2 | Extreme Pathway Characteristics Associated with the Production of the Indicated Amino Acids Table 3 | Extreme Pathway Characteristics for the Production of Subsets of Biomass and the Individual Nucleotides The yield plots for the equimolar production of the amino acids are presented for cases 3 and 4 . A large number of the extreme pathways were found within a narrow range of yield values. For case 3, similar66% of the extreme pathways were found between the dashed lines, representing only 17% of the range of capable yield values. For case 4, similar57% of the extreme pathways were found within the dashed lines, also representing only 17% of the possible yield values. Although the yield values within this range represent distinct extreme pathways, these plots indicate the relatively narrow range of yield values that may be experimentally indistinguishable. Figure 3 | Plot of carbon yield values for the production of equimolar amino acid sets. Plot of carbon yield values for the production of equimolar amino acid sets. Cases 3 and 4 are presented. The pathway values between the two cases do not correlate with each other; both data sets are independently rank ordered in terms of increasing yield. The pathways with yield values between the two dashed lines (spanning 17% of the range of yield values) correspond to 66% and 57% of the total number of pathways for cases 3 and 4, respectively. In the production of the equimolar set of nonessential amino acids, nitrogen has two potential systemic outputs: the amino acids and ammonia. Figure (without urea input) shows how nitrogen is directed in the extreme pathways that have been calculated for equimolar nonessential amino acid production in case 3. The percentage of nitrogen directed to these outputs is presented as percentages of total nitrogen input into the system. Here, 88% of the extreme pathways require at least 80% of the total input nitrogen for ammonia production. Urea was added as an allowable input to see what effect it would have on the distribution of nitrogen. There were 1477 more extreme pathways with this additional input (Fig. , with urea input). However, the number of pathways with >80% of the nitrogen diverted to ammonia increased only slightly, from 88% to 89%, with the addition of urea as an input. The minimum and maximum values for both conditions remained unchanged. Figure 4 | Nitrogen fate plots for equimolar amino acid production. Nitrogen fate plots for equimolar amino acid production. Case 3 exchange flux constraints, with and without urea, are shown. H. pylori Amino Acid Production in Physiologic E. coli Ratios | The relative ratios of the individual amino acids in the produced set affects the extreme pathways that are generated. Because data is not available for the amino acid composition of H. pylori, the amino acid composition of E. coli was used as an estimate. We computed the extreme pathways for the set of nonessential amino acids in E. coli ratios (Fig. illustrates yield values and nitrogen fates). As shown, the shape of these curves resembles that for the equimolar production of the amino acids, although the number of generated extreme pathways changed. The range of the yield values was different for the production of the set of amino acids in the two situations (equimolar and E. coli ratios) because there were many more amino acids produced in the set with E. coli --based ratios (i.e., six glycine molecules per set instead of one per set). Figure 5 | Carbon yield (A) and nitrogen fate (B) plots for the production of the nonessential amino acid set in Escherichia coli ratios. Carbon yield (A) and nitrogen fate (B) plots for the production of the nonessential amino acid set in Escherichia coli ratios. Nucleotide Production | The nitrogen fates of the extreme pathways for the production of an equimolar ratio of ribonucleotides are shown in Figure . The shape of this curve was similar to that for the production of the amino acid set, although with a less pronounced shallow slope region. The percentage of nitrogen that could be incorporated into ribonucleotides was 42%. Figure 6 | Carbon yield (A) and nitrogen fate (B) plots for the production of an equimolar set of the ribonucleotides in case 4. Carbon yield (A) and nitrogen fate (B) plots for the production of an equimolar set of the ribonucleotides in case 4. The extreme pathways for the equimolar set of deoxyribonucleotides were not determined because of computational unfeasibility. In attempting to calculate the extreme pathways for synthesis of the deoxyribonucleotides, the network was unable to produce deoxythymidine triphosphate (dTTP) under the conditions that have been outlined above. The stoichiometric matrix was investigated, and further in silico experiments were performed to ascertain the reason. It was subsequently determined that the model was unable to produce dTTP without simultaneously producing glycine. Extreme Pathway Redundancy | The degree of pathway redundancy was evaluated for the individual production of the nonessential amino acids in H. pylori . There was an average of two internal states per unique external state. Previous work has been performed on the pathway redundancy of H. influenzae . In contrast with H. influenzae, there was relatively little variation in the degree of pathway redundancy among the individual amino acids in H. pylori. In H. pylori, the maximum value for the extreme pathway redundancy was 3.5 for the production of aspartic acid in case 1; the minimum value was 1.3 for the production of tryptophan in case 4. This result is strikingly different from the range of values in H. influenzae; in one tested condition for H. influenzae, the minimum value was 10 and the maximum value was 236 for the production of glutamine and tryptophan, respectively. The numbers of extreme pathways per unique external state for the nonessential amino acid sets were found to be of the same order of magnitude as that of the average values for the production of the individual amino acids . The pathway redundancy was approximately two for all the evaluated cases. In our previous study with H. influenzae, the addition of succinate as another carbon output caused the average pathway redundancy to increase from 37 to 52 internal states per unique external state. However, similar increases were not seen in H. pylori with the addition of carbon outputs (acetate, case 2; formate, case 3; and lactate, case 4). The redundancy calculations for the production of ribonucleotides in H. pylori indicated a reduced degree of redundancy . For the production of ATP, CTP, GTP, and UTP, there was an average of 1.2 extreme pathways per unique external state. For the production of the set of ribonucleotides in equivalent molar ratios, the ratio of extreme pathways per unique external state was 1.1. This indicates less redundancy in the defined metabolic network of H. pylori for the production of these metabolites than for the amino acids. The average degree of pathway redundancy for the production of the individual amino acids in H. pylori was an order of magnitude less than that in H. influenzae. The degree of redundancy for the two organisms was evaluated under similar conditions . The outputs included in the comparison were acetate, succinate, and the individual nonessential amino acids common to both organisms. These amino acids included: asparagine, aspartic acid, glutamine, lysine, proline, serine, threonine, and tyrosine. Even with similarly sized organisms, there was a vastly different degree of redundancy . There was an average of two extreme pathways with equivalent external states in H. pylori. The same outputs for H. influenzae resulted in an average of 46 extreme pathways per external state. Although the differences in the minimal medium requirements for each of the organisms make direct comparisons difficult, this difference in pathway redundancy does indicate a much more rigid metabolic network architecture for H. pylori than for H. influenzae. Figure 7 | The exchange flux constraints used for the comparison of the two organisms H. influenzae The exchange flux constraints used for the comparison of the two organisms H. influenzae and H. pylori. Note that more inputs were allowed into both systems, but the indicated inputs were the only ones used by these in silico strains. The organisms had identical output fluxes, with the exception of ammonia as a necessary output for H. pylori. The individual amino acids included in the comparison were asparagine, aspartic acid, glutamine, lysine, proline, serine, threonine, and tyrosine. Table 4 | Network Redundancy Comparison Between the Two Indicated in silico Organisms DISCUSSION : This study presents the first genome-scale extreme pathway analysis for a reconstructed genome-scale metabolic network of H. pylori, and it details physiological, emergent, systems-level properties of this organism. This in silico study resulted in the establishment of important systemic properties of the reconstructed network. We found that (1) the metabolic network associated with the production of individual amino acids in H. pylori is redundant, but to a much lesser degree than that of H. influenzae; (2) the pathway redundancy for the simultaneous production of multiple amino acids was on the same order of magnitude as the redundancy in the production of the individual amino acids, but the redundancy in ribonucleotide production was nearly half of that seen in the amino acids; and (3) the network was unable to achieve a high level of conversion of input nitrogen to the set of either amino acids or nucleotides, and the inclusion of urea as an input did not significantly affect the percentage of input nitrogen secreted as ammonia. The degree of extreme pathway redundancy for the individual amino acid production was much lower in H. pylori than previously calculated in H. influenzae . There were approximately two internal states per unique external state for the production of the amino acids in the H. pylori network, whereas H. influenzae was found to have an average of 46 extreme pathways per external state under similar conditions (minimal medium with acetate and succinate as carbon outputs). This result indicates that the reconstructed H. pylori network is much less metabolically redundant and therefore less robust than the H. influenzae network. In its very specific environmental niche (the human stomach), H. pylori are surrounded by a wealth of amino acids and other precursors that it needs for growth. The ratio of internal to external states in the H. pylori network for amino acid production was found to be close to two despite the addition of more allowable outputs. This result was also somewhat surprising because the pathway redundancy changed dramatically with the inclusion of succinate as an additional carbon sink in our previous study on H. influenzae. In this previous study, the degree of redundancy also varied widely for the production of each different amino acid, in contrast to the data on H. pylori presented herein. This lack of variance again shows a much less flexible and versatile metabolism, an interesting fact in view of the similar size and gene number of H. pylori and H. influenzae. An implication of these in silico results is that the metabolic network of H. pylori may be very specifically tailored to its environment. The extreme pathway analysis of the amino acid and ribonucleotide sets showed similar degrees of pathway redundancy in H. pylori . For the production of the nonessential amino acid sets, there were approximately two internal states per external state, similar to that seen in the production of the individual amino acids. The yield curves for the production of the amino acid and nucleotide sets in H. pylori all showed a similar shape (an inverted S-curve). Thus, there was a region on each of the carbon yield and nitrogen fate plots that had a shallow slope, where many extreme pathways corresponded to a relatively small range of the respective yields. These regions could potentially hold significant physiological meaning. Specifically, the metabolic network has an extremely large number of different ways to produce a relatively small band of yield values. It has been hypothesized that these regions correspond to flux values relevant to primary objectives of the organism , such as reproduction and growth. Further investigation into the genome-scale functioning of metabolic networks is needed to determine if this in silico --generated hypothesis is valid. Redundancy in the metabolism of ribonucleotides was reduced in comparison to that of the amino acids. For the production of the individual ribonucleotides, there were 1.1 internal states per external state. For the set of ribonucleotides, there were 1.2 internal states per external state. This degree of redundancy indicates a much more rigid metabolic network than that for amino acid production. Full integration of input nitrogen into either the amino acids or the ribonucleotides was not possible in the reconstructed H. pylori network. A maximum of 40% of input nitrogen was used in equimolar amino acid synthesis; alternatively, a maximum of 42% of the input nitrogen could be incorporated into ribonucleotide synthesis. In the in silico H. pylori model, the excess nitrogen was necessarily diverted to ammonia production. Because of computational intractability, we were unable to calculate the extreme pathways for the full complement of biomass constituents. Therefore, as has been previously indicated , it is possible that nitrogen taken up from the amino acids is primarily incorporated into biomass when all constituents are taken into account, and that ammonia production for pH control comes predominantly from the breakdown of urea. However, these initial extreme pathway calculations, in which uncoupled protein production can use a maximum of only 50% of the input nitrogen for all cases tested, indicate that nitrogen input from arginine and alanine is potentially used to produce ammonia to some extent in vivo. Also, although 40% utilization of nitrogen in equimolar amino acid production and 42% nitrogen utilization in ribonucleotide synthesis were maximum values, the bulk of the pathways located in the plateau regions achieved much smaller nitrogen utilization. Surprisingly, the addition of urea as an input did not alter the shape of the curve or the percentage of pathways that use small amounts of nitrogen. This lack of change in the nitrogen fate plots indicates that the network was incapable of directing a greater percentage of input nitrogen to ammonia production with the addition of urea. Because ribonucleotide synthesis and amino acid synthesis directed at least 50% of the inputted nitrogen to ammonia production, it is therefore possible that the production of nonessential amino acids with ribonucleotides will maintain a similar nitrogen fate. These two metabolite groups represent a substantial amount of the biomass, and consequently, this characteristic may hold true for the complete growth of H. pylori. The results obtained were based on a reconstructed genome-scale network using all the currently available information for H. pylori. However, until we have 100% open reading frame assignments, no metabolic network reconstruction is complete. In silico models, such as the one used herein, can be valuable tools in helping to guide research. For example, it was found that dTTP could not be produced without simultaneous production of glycine in the defined metabolic network. If the production levels of these two metabolites were measured in vivo and found not to correlate directly, it would be strong evidence that there remains at least one unknown reaction in the network that involves either or both of these metabolites. Such critical comparison between data-driven model predictions and experimental results is thus a valuable tool for our quest to obtain comprehensive in silico representations of living cells. In summary, this study presents the first calculation of genome-scale extreme pathways for the production of large subsets of biomass components (i.e., the full amino acid set, the nucleotide set). It should be emphasized that extreme pathways are completely and uniquely determined by the structure of the metabolic network. They differ in this description from many other pathway definitions that are often arbitrary or intuitive groupings of contiguous reactions in the network. Extreme pathway analysis is a form of genome-enabled science that can be used to determine the capabilities of a reconstructed metabolic network. Thus, it is useful to compare the in silico analysis of extreme pathways with experimental data to determine the extent to which the known metabolic map can be used to reconstruct observed phenotypic behaviors. Future studies will focus on enabling the calculation of the extreme pathways for the complete biomass of H. pylori and for larger genomes. As the complexity of life increases dramatically with the addition of a relatively small amount of genes , apparently so do the numbers of extreme pathways that characterize a metabolic system, allowing cells the ability to fine-tune its metabolic phenotype. Extreme pathways allow us to analyze the fundamental structure of metabolic networks to evaluate the metabolic functions of a cell from a pathway perspective. METHODS : Formulation of the H. pylori Metabolic Genotype | The metabolic network for H. pylori was reconstructed from reviews of biochemical literature and genome databanks following previously used methods . The complete listing of the reactions included in this H. pylori model can be found at . Isozymes were not included in the in silico model used for this study. The selection between isozymes had no effect on the extreme pathway calculations made in this analysis. For any given pathway, the respective flux would be equivalent independent of the catalyzing enzyme per details in the aforementioned extreme pathway analysis algorithm . Extreme Pathways and Convex Analysis | The m x n stoichiometric matrix, S, of this reconstructed network includes all network metabolites (m rows) and all corresponding transport processes and metabolic reactions (n columns). The flux represents the amount of mass moving through the associated reaction. An exchange flux corresponds to a flux across the system boundary. An internal flux corresponds to a reaction within the system. A metabolic network can be constrained by implementing simple thermodynamic principles regarding the irreversibility of reactions. If a reaction is reversible, it is decomposed into two reactions (for the forward and backward directions). Thus defined, all internal fluxes are constrained to be nonnegative. By decoupling the reversible reactions, we have where v is a vector of all the fluxes in the metabolic network. Another set of constraints on this system involves the conservation of mass, which in a steady state is described by the following flux balance equation: A linear basis for this system enables a characterization of the metabolic capabilities of an organism. Each solution to this equation would represent a steady-state set of flux values for the metabolic network. These correspond to a particular cell phenotype. Any solution to equation 2 can be reached with a combination of the linear basis vectors for the S matrix. However, this combination can be negative or positive, violating the thermodynamics of the chemical reactions. To overcome this difficulty, developed an algorithm to evaluate a set of basis vectors, called extreme pathways, using convex analysis. With a convex basis, every point in the space defined by equation 2 can be written as a nonnegative combination of the extreme pathways of the system (Equation ). C represents the convex cone encompassing all possible solutions of the metabolic network; v represents the vector of fluxes for each reaction in the network; and alphai and pi are the weights and the extreme pathways of the network, respectively. Every steady-state solution of the metabolic stoichiometric network can then be represented as a nonnegative combination of the extreme pathways. These extreme pathways are a unique and mathematically rigorous characterization of the solution space for the metabolic network. WEB SITE REFERENCE : ; Complete listing of the reactions included in this H. pylori model. Backmatter: PMID- 11997336 TI - A Complete Sequence of the T. tengcongensis Genome AB - Thermoanaerobacter tengcongensis is a rod-shaped, gram-negative, anaerobic eubacterium that was isolated from a freshwater hot spring in Tengchong, China. Using a whole-genome-shotgun method, we sequenced its 2,689,445-bp genome from an isolate, MB4T (Genbank accession no. ). The genome encodes 2588 predicted coding sequences (CDS). Among them, 1764 (68.2%) are classified according to homology to other documented proteins, and the rest, 824 CDS (31.8%), are functionally unknown. One of the interesting features of the T. tengcongensis genome is that 86.7% of its genes are encoded on the leading strand of DNA replication. Based on protein sequence similarity, the T. tengcongensis genome is most similar to that of Bacillus halodurans, a mesophilic eubacterium, among all fully sequenced prokaryotic genomes up to date. Computational analysis on genes involved in basic metabolic pathways supports the experimental discovery that T. tengcongensis metabolizes sugars as principal energy and carbon source and utilizes thiosulfate and element sulfur, but not sulfate, as electron acceptors. T. tengcongensis, as a gram-negative rod by empirical definitions (such as staining), shares many genes that are characteristics of gram-positive bacteria whereas it is missing molecular components unique to gram-negative bacteria. A strong correlation between the G + C content of tDNA and rDNA genes and the optimal growth temperature is found among the sequenced thermophiles. It is concluded that thermophiles are a biologically and phylogenetically divergent group of prokaryotes that have converged to sustain extreme environmental conditions over evolutionary timescale. [Supplemental material is available online at .] Keywords: Introduction : Thermoanaerobacter tengcongensis, isolated from a hot spring in Tengchong, Yunnan, China, is a rod-shaped, gram-negative (by empirical definitions) bacterium that grows anaerobically under extreme environment. It propagates at temperatures ranging from 50 to 80C (optimally at 75) and at pH values ranging between 5.5 and 9 (optimally from 7 to 7.5). It shares several key genomic and physiological features common to the genus Thermoanaerobacter, such as a relatively low genomic G + C content (<40%), reduction of thiosulfate/sulfur to hydrogen sulfide, and fermentation of glucose to acetate/ethanol . T. tengcongensis, however, has several important phenotypic properties that contradict its membership to the genus. Some of the examples include the absence of spore production, negative gram-staining result, lack of motility under cultural conditions, and exclusive metabolic pathways (such as deficiencies in lactate production and xylan utilization; ; ). To obtain a global view of genes possessed by the organism and to resolve some of the controversies at molecular levels, as well as to understand the biology of thermophilic prokaryotes in general through comparative genomics, we set out to sequence the T. tengcongensis genome. Using a whole-genome-shotgun method, we acquired sequence data at high-genome coverage (9.87x) and assembled the complete sequence of the T. tengcongensis genome of a laboratory strain, MB4T (Genbank accession no. ; see also ). Computational analyses of the high-quality genomic sequence not only confirmed many of the early experimental observations, but also uncovered the heterogeneous nature of thermophilic prokaryotic genomes. The T. tengcongensis genome sequence should provide vital information for understanding cellular and molecular mechanisms that are employed by microorganisms under extreme environments. RESULTS AND DISCUSSION : General Features | T. tengcongensis has a single, circular chromosome of 2,689,445 bp (base pairs) in length (Fig. a,b; Table ). Second only to the Sulfolobus solfataricus genome, it is one of the largest genomes of thermophiles sequenced to date (; ; ; ; , ; ; ; ; ; Heilig, unpubl.). The genomic sequence has an average G + C content of 37.6%, similar to those of other members of the genus Thermoanaerobacter (Table ; Supplemental Table A [available at ]). The genome has 4 rRNA gene clusters (12 rRNA genes) and each cluster encompasses a single copy of 5S, 16S, and 23S RNA genes. The G + C content of the rRNA genes or rDNAs varies from 58.2% to 60.3%. There are 55 tRNA genes scattered over the genome in 28 loci (1 --8 tRNAs in each locus). The G + C content of tDNAs has a broader distribution than that of rDNAs, from 52.6% to 69.3%. The characteristically high G + C content of rDNA and tDNA genes found in T. tengcongensis appears common to all thermophiles (discussed in detail later; also see Supplemental Table A). The elevated G + C content of rDNAs and tDNAs as a function of genomic G + C content increase is also evident in most of the mesophiles, albeit less pronounced (Supplemental Table B). Figure 1 | (a) Circular representation of the Thermoanaerobacter tengcongensis genome. (a) Circular representation of the Thermoanaerobacter tengcongensis genome. Circles display (from the outside): (1) Physical map scaled in megabases from base 1, the start of the putative replication origin. (2) Coding sequences transcribed in the clockwise direction. (3) Coding sequences transcribed in the counterclockwise direction. (4) G + C percent content (in a 10-kb window and 1-kb incremental shift); values >37.6% (average) are in red and smaller in blue. (5) GC skew (G-C/G + C, in a 10-kb window and 1-kb incremental shift); values greater than zero are in magenta and smaller in green. (6) Repeated sequences; short 30-bp repeats are in red and other types in blue. (7) tRNA genes. (8) rRNA genes. Genes displayed in 2 and 3 are color-coded according to different functional categories: translation/ribosome structure/biogenesis, pink; transcription, olive drab; DNA replication/recombination/repair, forest green; cell division/chromosome partitioning, light blue; posttranslational modification/protein turnover/chaperones, purple; cell envelope biogenesis/outer membrane, red; cell motility/secretion, plum; inorganic ion transport/metabolism, dark sea green; signal transduction mechanisms, medium purple; energy production/conversion, dark olive green; carbohydrate transport/metabolism, gold; amino acid transport/metabolism, yellow; nucleotide transport/metabolism, orange; coenzyme metabolism, tan; lipid metabolism, salmon; secondary metabolites biosynthesis/transport/catabolism, light green; general function prediction only, dark blue; conserved hypothetical, medium blue; hypothetical, black; unclassified, light blue; pseudogenes, gray. (b) Linear representation of the T. tengcongensis genome. Genes are color-coded according to different functional categories as described above for a , with above character-string representing gene names or IDs. Arrows indicate the direction of transcription. Genes with authentic frameshift and point mutations are indicated with X. Paralogous gene families are indicated by family ID in a box above the predicted genes. Numbers next to GES (Goldman-Engleman-Steitz) represent the number of membrane-spanning domains predicted by Goldman-Engleman-Steitz scale calculated by TMHMM. Proteins with five or more GES are indicated. The 305 copies of the 30-bp short repeat, clustered in two regions, are indicated with the greater-than symbol. RNA genes, including those of rRNA, tRNA, and other RNA genes, signal peptides and long repeats are also indicated. Numbers on the tRNA symbols represent the number of tRNAs in the cluster. Table 1 | General Features of the Thermoanaerobacter tengcongensis Genome Repetitive Sequences | The T. tengcongensis genome has a significant fraction (9.1%) of repetitive sequences that include simple repeats of a few dozen base pairs in length as a limited number of clusters to complex ones, such as transposase coding (Tables , ). In this study, all repeats were categorized by the means of a suffix tree algorithm , coupled with intensive manual alignment and visual inspection. Table 2 | Selected List of Repetitive Elements in the Thermoanaerobacter tengcongensis Genome The most characteristic repeat family of the T. tengcongensis genome consists of 305 copies of a unique 30-bp AT-rich repeat, TSR001 (Fig. b). They are further grouped into two subfamilies, TSR001a and TSR001b. The two subfamilies differ from each other only by a single substitution at position 18, an adenosine (67 copies) in TSR001a and a guanine (238 copies) in TSR001b, respectively. Sixty-five copies of TSR001a are clustered between 2,326,770 bp and 2,331,141 bp and all units are oriented in the same direction. The two remaining copies are arrayed together with a single cluster of TSR001b (238 copies) from 2,537,291 bp to 2,555,096 bp. The repeat units are not attenuated directly but interrupted by nonrepetitive sequence spacers, most of which are 34 to 41 bp in length. However, three of the spacers are longer than 100 bp (2,329,533 --2,329,637 bp, 2,538,340 --2,538,450 bp, and 2,550,689 --2,550,793 bp) and another one is 1632 bp (2,540,469 --2,538,790 bp) in length, which encodes a transposase (TTE2646). Repeats of similar types are found in other thermophiles, from both archaea and eubacteria. Most of them are distinct, short (20 to similar60 bp), relatively abundant, and organized in a single cluster or multiple clusters . The function of such repeats is yet to be defined and they might play important roles in chromosome anchorage and segregation in these thermophilic organisms. Thirty-seven families of protein-coding repetitive sequences longer than 300 bp were also categorized. Most of them are related to transposases (10 families; 54 copies) and ABC transporters (6 families; 13 copies). Others are unknown or hypothetical (11 families; 62 copies). The largest repeated sequence, TLR028 (3565-bp in length), is composed of two different transposases flanking a hypothetical protein. The most abundant one, a 1596-bp repeat (TLR008), consisting of a single hypothetical gene and a 200-bp noncoding region, occurs 21 times over the entire genome. Origin of Replication | Of a half dozen methods for determining origins and termini of DNA replication, including asymmetric distribution of oligomers , GC-skew (G-C/G + C; ), accumulated GC-skew , and orientation of coding sequences (CDS), all worked satisfactorily in determining the origin of replication for T. tengcongensis. Figure depicts results from some of the analyses. The predicted origin is defined between ribosomal protein L34 (TTE2802) and dnaA (TTE0001) genes, which is dictated by the asymmetry of the nucleotide composition between the leading and the lagging strands. The first base of an octamer repeat (TTTTTCTT)1423, 307-bp upstream of dnaA, is assigned as base-pair number one, whereas the terminus is about halfway into the genome, similar1345-kbp from the origin (Fig. a). Figure 2 | The replication origin of the Thermoanaerobacter tengcongensis. The replication origin of the Thermoanaerobacter tengcongensis. GC skew [(G-C)/(G + C)] was calculated with a nonoverlapping sliding window of 10 kb for a single strand over the length (upper horizontal line). Cumulative GC skew was plotted from position 1 of the genome (upper solid line). Cumulative gene direction (upper dotted line) was plotted from position 1 of the genome sequence, showing that the majority of genes transcribe along the same direction following the replication forks. In the skewed oligomer (TTTTTCTT)1423 part (lower), vertical lines above the center represent the location of this octamer on one DNA strand, and lines below the center indicate the positions on the complementary strand. The transition in GC and oligomer skews, maxima of the curves at the middle of the genome sequence, is identified as the putative terminus of replication. T. tengcongensis has the most biased gene distribution on the leading strand, in the same direction as genome replication, among all sequenced prokaryotic genomes known to date (Fig. a). Of the genes, 86.7% (41.9% and 44.8% from the two replication forks) are transcribed along the leading strand from the two halves of the genome divided by the replication origin. The lagging strand encodes only 13.3% (7% and 6.3% from the two replication forks). The biases in gene orientation have been observed in many other bacteria , but only three of them exceed 80% of the total encoded genes. The extreme case is not seen in prokaryotes but in a eukaryotic organism, Leishmania major, in which the leading strands of all chromosomes encode all the genes . Further analysis and experimentation are of essence to address what is the driving force that instigates such extreme gene distributions. Coding Sequences | Identified were 2588 predicted CDS, covering 87.1% of the genome (Table ; for functional classifications, see Table 3 and Supplemental Table C). Genes for stable RNAs populates 0.9% of the genome. The average length of the CDS is 905 bp, slightly longer than that of a mesophile, Bacillus halodurans (880 bp; ). Of the CDS, 72.9% start with ATG, 13.2% with TTG, and 13.9% with GTG. Such a distribution is similar to that of the B. halodurans genome, of which 78% of the CDS begin with ATG, 10% with TTG, and 12% with GTG. There are 1764 CDS (68.2%) that are homologous to known proteins or protein domains/motifs in public databases; thus, their biological functions are putatively assigned. Identified were 301 CDS (11.6%) in other sequenced prokaryotic genomes as conserved protein sequences of unknown function; 523 CDS (20.2%) have no homologous counterparts in all public databases. When protein similarity was scored in a genome-wide fashion, 54.4 % of T. tengcongensis genes have extensive similarity (BLASTP; 1e-10) to those of B. halodurans. Their overall genome similarity ranks the highest among all the sequenced genomes, regardless if they are thermophiles or mesophiles . Figure 3 | Relative distance of the Thermoanaerobacter tengcongensis genome with those of other 47 completely sequenced genomes, measured by a collective similarity score of the 2588 predicted coding sequences (CDS). Relative distance of the Thermoanaerobacter tengcongensis genome with those of other 47 completely sequenced genomes, measured by a collective similarity score of the 2588 predicted coding sequences (CDS). All the sequences were retrieved from NCBI databases. A tally was kept of which genome produces the significant similarity with the BLASTP program above an expected value of 1e-10. The number of T. tengcongensis CDS matched to those of each genome is tabulated. Bacillus halodurans has the highest value of 54.4%, indicating its highest similarity to T. tengcongensis. Replication, Recombination, and DNA Repair | Genes for the primary replication machinery, the DNA polymerase III complex in T. tengcongensis, are similar to those of well-characterized components in Escherichia coli, which is composed of alpha-subunit (dnaE, TTE1818), beta-subunit (dnaN, TTE0002), gamma-tau-subunit (dnaX, TTE0039), and delta-subunit (holA, TTE0942). In addition, a polC-like gene encoding an alternative DNA polymerase III alpha-subunit was also identified in T. tengcongensis (TTE1398). The presence of two alpha-subunits is not exceptional for T. tengcongensis: this function of polC gene has been reported in Bacillus subtilis . Both dnaE and polC genes are found in several fully sequenced bacterial genomes of the Bacillus/Clostridium group. The thermophilic Thermotoga maritima also harbors these two genes. Although the essential DNA polymerase I homolog is present in T. tengcongensis (TTE0874), DNA polymerase II ---recently being shown to be involved in replication-related DNA damage repair in E. coli but not being essential ---is absent. Many other essential DNA replication-related genes are readily determined by sequence homology. For instance, topoisomerases I/II (topA, TTE1449; gyrA, TTE0011; and gyrB, TTE0010), single-stranded DNA- binding protein (Ssb), DNA helicase (dnaB, TTE2774), and primase (dnaG, TTE1757) are all readily defined by sequence homology. Homologs of recombination and DNA repair-related genes, such as recA/B/D/F/G/N/O/R (TTE1374, TTE0264, TTE0489, TTE0004, TTE1492, TTE1302, TTE0976, and TTE0041, respectively), and >20 genes that are involved in postreplicational mismatch/excision, ultraviolet-induced damage and transcription-coupled DNA repairs, including the mutT/mutS gene families, uvrA/B/C (TTE1970, TTE1971, and TTE1966, respectively) gene cluster, and the uvrD (TTE0604) gene, were found in T.tengcongensis. Although none of the methylation-related dam/dcm homologs was found, suggesting that the genome DNA has no dam/dcm methyl modification, the T. tengcongensis genome possesses seven putative endonuclease genes and a type-I restriction-modification system that is composed of four genes in a single operon. The functions of these putative genes are currently being evaluated. Transcription and Translation | Three RNA polymerase core-enzyme genes (rpoA, TTE2263; rpoB, TTE2301; and rpoC, TTE2300), which encode subunits alpha, beta and beta', and another gene that encodes polymerase subunit Omega (rpoZ, TTE1510) are all documented. Seventeen sigma factors belonging to four groups that constitute the holoenzyme of the RNA polymerase complex are found. The first group contains four of the rpoD (sigma70)-like genes, believed to have housekeeping functions. rpoN (sigma54)-like gene stands alone. The third group, the largest of all, is composed of seven rpoE (sigma24) homologs of the Extracytoplasmic function (ECF) subfamily, whose function is postulated as stress-related, and they perhaps are responsive to the high-temperature environment . The last five fliA-like genes as a group (sigmafliA) are alternative sigma factors. Additional transcription-related factors, such as the elongation factor (greA), the rho factor, the termination factors (nugA, nugB), and three antitermination factors (nugG-like genes) are all unambiguously recognized. Among these documented genes, greA, nugB, and rho have homologs only in Eubacteria. T. tengcongensis has >50 transcriptional regulators acting as activators or repressors involved in many physiological and metabolic pathways. There are similar15 response regulators also related to transcriptional regulation. Twelve of them are two-component response regulators , characterized by a CheY-like receiver domain and an HTH (helix-turn-helix) DNA-binding domain. Two of them are serine phosphatases (encoded by rsbU) with orthologs found in B. subtilis and T. maritime. The last one is a ppGpp synthetase/hydrolase (TTE1195) whose product is believed to be the effector involved in bacterial stringent response . All translation-related genes are highly conserved as seen in other prokaryotes, and shared by both Eubacteria and Archea. Twenty-three genes that encode 20 essential tRNA synthetases are predicted. Two copies (TTE1394 and TTE2299) of an archaeal gene that encodes the ribosomal subunit RpL8A protein are identified in the T. tengcongensis genome. This gene has been found in two other related eubacterial genomes, a thermophile, Thermotoga maritina, and a mesophile, B. halodurans. Many gene products involved in posttranslational processes are also inevitable, including those heat-shock proteins (such as GroES, GroEL, DnaJ/K, and HslU) and chaperones (such as Hsp33 and Hsp20, ATPases associated with various cellular acts and peptidase). A homolog of cold-shock protein, CspC, and a protein that has a regulatory function in transcription and stationary phase survival, SurE, is also present in the genome. Respiratory Pathways | T. tengcongensis gains energy anaerobically by sulfur respiration and uses thiosulfate or element sulfur as electron receptors because its growth increases in the presence of thiosulfate or sulfur but not in the presence of sulfate . Such an observation seems to contradict a common feature observed in most sulfur-respiratory prokaryotes, a heterogeneous group of microorganisms that have the ability to use sulfate as a terminal electron acceptor , including both eubacteria and archaea. What has happened to the sulfate pathway in the T. tengcongensisgenome? First, neither the genes related to sulfate transport systems, nor the key genes involved in the sulfate reduction (such as sulfate adenylate transferase, 3'-phosphoadenosine 5'-phosphosulfate sulfotransferase and adenylylsulfate kinase) are present. Secondly, in the reduction process, thiosulfate is generally reduced to sulfite and further to sulfide. Thiosulfate reductase and sulfite reductase, which play crucial roles in these steps, are not found in the T. tengcongensis genome. Instead, a rhodanese-related sulfurtransferase (TTE1148), which employs thiosulfate as electron acceptor in the presence of cyanide ion , is identified. Because sulfite is not an end product of sulfur metabolism and cannot be reduced to sulfide, it might be recycled back to thiosulfate through a thiosulfate-synthesis pathway in T. tengcongensis as it has been described in Desulfovibrio vulgaris . In D. vulgaris, a trithionate reductase system consisting of two proteins was identified. One is bisulfite reductase, which reduces bisulfite to trithionate, and the other putative protein is designated as TR-1. Both enzymes are required to reduce trithionate to thiosulfate. If this is also the case in T. tengcongensis, it is expected to find flavodoxin (TTE0566, TTE0694, TTE1329, and TTE1531) and cytochrome c3 (TTE1025), which are essential to this pathway. Indeed, the two genes are present in T. tengcongensis. Moreover, two putative ancient conserved regions (ACR) (TTE0085 and TTE0087, stress proteins believed to be involved in the bacillary response to adverse conditions and in non-replicating persistence) related to intracellular sulfur reduction and oxidation also exist in the genome. Although most of the sequenced bacterial genomes have rhodanese-related sulfurtransferases, the two ACR genes are detectable only in a few other bacterial genomes, including Methanobacterium thermoautotrophicum , T. maritime , E. coli , Pseudomonas aeruginosa , and Vibrio cholerae . M. thermoautotrophicum is a methanogen that utilizes CO2 as the electron acceptor , and T. maritima is a thermophile that has an ability to gain energy through a fermentation pathway in the presence of Fe (III) and utilizes sulfur as electron acceptor but does not consequently produce any ATP . No rhodanese-related sulfurtransferase has been recognized in the T. maritima genome either. P. aeruginosa and V. cholerae are oxygenic-respiration bacteria. E. coli has both aerobic and anaerobic respiratory pathways, and the pathway involving formate oxidation and nitrate reduction constitutes a major anaerobic respiratory pathway in E. coli , which is completely absent in T. tengcongensis. Metabolisms | As an anaerobic and heterotrophic eubacterium, T. tengcongensis utilizes both monosaccharides and polysaccharides as carbon sources and yields H2, CO2, and acetate as its major metabolic end products . Among the complex sugars, it is capable of metabolizing starch but not cellulose or xylan. It is known that thiosulfate reducers, such as T. brockiiand T. thermohydrosul furicus, as well as several other thermoanaerobacteria, consume a variety of sugars, including polymeric sugars . However, only a few sulfate-reducers are known to grow on sugars, including Archaeoglobus fulgidus, D. nigrificans, D. geothermicum, D. simplex, D. termitidis, andD. fructosovorans . A. fulgidus is the only one among the group capable of utilizing polymeric sugars. T. tengcongensis has a complete set of genes constituting the glycolysis and the pentose phosphate pathways. It, however, has a few key metabolic enzymes yet to be found for other related pathways. One of the examples is fructose-1,6-biphosphatase, a key enzyme in the gluconeogenesis pathway. Such a depletion is not extraordinary, as similar cases are encountered in all other sequenced thermophiles and certain nonthermophilic bacteria, such as B. subtilis , Deinococcus radiodurans , and Xylella fastidiosa . Another example is the absence of 2-keto-3-deoxy-6-phosphogluconate aldolase in the Entner-Doudoroff pathway. The metabolism of pyruvate reflects the microaerophilic nature ofT. tengcongensis. Neither the aerobic pyruvate dehydrogenase (COG0567; ) nor the strictly anaerobic pyruvate formate lyase (COG1882) is present in T. tengcongensis. Similar to the cases of Helicobacter pylori and Campylobacter jejuni , T. tengcongensis has 12 genes (TTE0445, TTE0960, TTE0961, TTE1209, TTE1210, TTE1211, TTE1340, TTE1341, TTE1342, TTE2193, TTE2194, and TTE2198) related to the pyruvate:ferredoxin oxidoreductases and 2-oxoacid:ferredoxin oxidoreductases. The conversion of pyruvate to acetyl coenzyme A (acetyl CoA) is performed by the pyruvate ferrodoxin oxidoreductase (POR; ; ), a four-subunit enzyme described in H. pylori and other hyperthermophilic organisms . Acetyl CoA is converted to acetate and this process is catalyzed by four enzymes, phosphate acetyltransferase (TTE1482, TTE2195, and TTE2204), acetate kinase (TTE1481), NADH:flavin oxidoreductase (TTE0012, TTE0988, TTE2131, and TTE2625), and Acyl-CoA dehydrogenase (TTE0545; ). These four enzymes are identified in T. tengcongensis. Anaerobic acetogenic bacteria with acetate as their primary reduced end product are capable of utilizing H2 and CO2 to produce acetyl CoA in an autotrophic biosynthetic scheme known as the Wood-Ljungdahl pathway (or the acetyl-CoA pathway). This pathway, catalyzed by enzymes of carbon monoxide dehydrogenase (CODH), formyltetrahydrofolate synthetase, and acetyl-CoA synthetase, synthesizes acetyl CoA from two molecules of CO2 . The key enzymes for the acetyl-CoA pathway, such as a CODH subunit CooS (TTE1708) and a formyltetrahydrofolate synthetase (TTE2391), are identified in T. tengcongensis. The existence of this pathway might reflect the acetogenic aspect of T. tengcongensis. The same pathway was described in A. fulgidus, a thermophilic, anaerobic sulfate-reducing archaeon that grows chemolithoautotrophically on H2 and CO2 with sulfate or thiosulfate as electron acceptor and grows chemoorganoheterotrophically with sulfate and lactate, as well as other carbohydrates . Many chemolithoautotrophic sulfate-reducing prokaryotes, such as those of the genus Desulfobacterium, are acetogenic bacteria , whereas no acetogenic features have been clearly reported so far about the thermophilic anaerobic thiosulfate-reducing Thermoanaerobacter bacteria, including T. tengcongensis. The tricarboxylic acid cycle (TCA) is also incomplete in T. tengcongensis and only half of the relevant clusters of Orthologous groups (COG), 8 out of 16, are present. The absence of the TCA-cycle enzymatic components have only been seen in other anaerobic bacteria, such as Pyrococcus horikoshii , Methanococcus jannaschii , and A. fulgidus . These three bacteria have only 3, 9, and 7 of the COGs, respectively. T. tengcongensis has a complete collection of genes involved in most of the amino acid biosynthetic pathways for threonine, valine, leucine, histidine, phenylalanine/tyrosine, tryptophan, arginine, and methionine. However, it lacks a few key genes such as threonine dehydratase for isoleucine biosynthesis and ornithine cyclodeaminase for proline biosynthesis. For nucleotide metabolism, it also has a complete set of genes for purine biosynthesis, purine salvage, and pyrimidine biosynthesis pathways, but an enzyme, ribonucleotide reductase beta-subunit for either pyrimidine salvage or thymidylate biosynthesis, appears absent. Similarly, the genes involving in coenzyme metabolism, such as ubiquinone and thiamine biosynthesis, are also incomplete. It is in fact quite common in other sequenced bacteria genomes that one or more genes in certain metabolic pathways are unidentifiable as gene identification and classification are based solely on sequence homology. Transporters | Coping with a heated aquatic environment, T. tengcongensis evolves to have a complex ion transport system and a large number of functionally defined transporter genes, crucial for acquiring essential substrates. It encodes ion transporters, not only for monovalent cations, such as K+/Na+, but also for divalent cations, such as Mn2+, Zn2+, and Ca2+. It also encodes transporters for both Fe2+ and Fe3+, as well as for other heavy-metal cations, such as cobalt and nickel, often serving as components of coenzymes. In addition, four undefined cation-transporting ATPases and three anion ion transporter genes for formate/nitrite, phosphate, and nitrate/sulfonate/taurine/bicarbonate are identified. Most of these genes are clustered in the genome, and the majority is composed of ABC-type transporters that require ATP as energy source, such as seven nickel-chelating ABC-type transporters that are involved in the uptake of di- or oligopeptide. Furthermore, 15 genes encoding permeases, members of the major facilitator superfamily, are found scattered over the genome. Finally, as the growth of T. tengcongensis takes place on many carbohydrate substrates , the operons for related substrate transport, including maltose, lactose, galactose, and spermidine/putrescine, are all readily identifiable. Cell Structure | Genes contributing to the cellular structure of T. tengcongensis are quite complex, especially those related to flagellar formation and gram staining. Despite the fact that flagella were not found in the cultured cells , T. tengcongensis does appear to be well equipped with all essential genes for flagellar biogenesis and with nearly all the genes for the chemotaxis signaling pathways. However, it remains puzzling why T. tengcongensis does not assemble functional flagellar under the culture conditions. Bacteria sense a wide range of environmental cues, including nutrients, toxins, and compounds that alter electron transport, pH, temperature, and even Earth's magnetic field . Histidine protein kinase (CheA, TTE1039 and TTE1417) plays a central role in bacterial chemotaxis signaling. Autophosphorylated CheA passes its phosphoryl group onto CheY (TTE0136, TTE0288, TTE1038, TTE1063, TTE1101, TTE1203, TTE1302, and TTE1428), and phosphoryl CheY (CheY-P) then acts on the flagellar motor/switch complex, FliG/FliM/FliN (TTE1441 and TTE1430). Consequently, the complex switches on and controls the flagellar movement. Two auxiliary proteins, CheW (TTE0700, TTE1034, TTE1136, and TTE1416) and CheZ, and two receptor modification enzymes, methylesterase (CheB, TTE1035 and TTE1418) and methyltransferase (CheR, TTE1037 and TTE1135), manipulate the fluctuation of phosphoryl groups within this central pathway . All genes in the chemotaxis signaling pathways except CheZ are unambiguously found in the T. tengcongensis genome. CheZ, a protein known to accelerate dephosphorylation of the responsive regulator phosphoryl CheY, has only been found in a few nonthermophilic eubacteria, such as E. coli , P. aeruginosa , and V. cholerae , and it neither affects the flagellar motors directly nor sequesters the CheY . The presence of these "silent" components involved in flagellar structure and movement in T. tengcongensis suggests a possibility that they might be activated only under certain environmental conditions or they used to be active not long before the present day. Another controversy is that T. tengcongensis, as a gram-negative rod by staining, shares many genes that are characteristic of gram-positive bacteria but lacks some characteristics of gram-negative bacteria. First, sporulation is generally one of the important features for certain gram-positive and rod-shaped bacteria . There are, surprisingly, 23 CDS, which are related to sporulation, in the T. tengcongensis genome. Even with such a remarkable number, only next to the genus Bacillus, which has an additional CDS of polysaccharide biosynthesis protein F (COG 1861) involved in spore-coat formation , no spore formation has been observed in T. tengcongensis culture. None of the other prokaryotes sequenced to date have more than 15 CDS implicated in sporulation. Secondly, gram-negative organisms have lipopolysaccharides (LPS), which gram-positive lacks. In the gram-negative organisms, lipopolysaccharides not only offer structural rigidity, but also affect surface permeability, charges, and hydrophobicity. Consequently, they alter the way bacteria interact with the environment. Biosynthesis of O-antigen polysaccharides takes place in multiple steps involved in synthesis of sugar precursors in the cytoplasm, formation and polymerization of the repeating units, and export to the cell surface . The T. tengcongensis genome, though having a few CDS related to lipopolysaccharide biosynthesis (TTE0652 and TTE0199), does not possess three of the key genes: the one related to lipopolysaccharide biosynthesis (LPS:glycosyltransferase, COG1442), and the two related to lipopolysaccharide transport (i.e., a periplasmic protein involved in polysaccharide export, COG1596) and an ATPase component of ABC-type polysaccharide/polyol phosphate transport system, COG1134. At least one of these three CDS is present in most of the gram-negative prokaryotes, such as P. aeruginosa, V. choleraeserotype, Neisseria meningitidis, X. fastidiosa, and E. coli. Thermophiles of archaea and eubacteria are not exceptional, such as A. fulgidus, Aquifex aeolicus, and T. maritima. Of the sequenced gram-positive bacteria, only the genus Bacillus contains two of the key proteins. Thirdly, none of the four CDS involved in lipid A synthesis are found in the T. tengcongensis genome, although they are well documented in most of the gram-negative prokaryotes, including a thermophilic eubacterium, A. aeolicus. Finally, CDS for porins unique to gram-negative bacteria also appear absent in T. tengcongensis. Less complicated but relevant examples, in which a decision was made for gram staining, do exist. For instance, T. wiegelii, a thermophilic, spore-forming and rod-shaped bacterium in the same genus of T. tengcongensis, is in fact gram-negative by the gram-staining protocol . Members of the genus Mycobacteria, believed to be phylogenetically closer to T. tengcongensis, are also recalcitrant to gram staining under standard conditions. Similar cases are encountered when staining other sulfur/sulfate-reducing species, such as the bacteria of the genus Desulfotomaculum. Although stained as gram-negative, they have many features related to the gram-positive organisms, such as that they form endospores and can be grouped according to their 16S rDNA sequences with the genus Clostridium. Some of them are indeed thermophilic acetogens . Sporomusa sphaeroides represents another similar case . It is clear that sensitivity to gram staining is a delicate feature of the bacterial world and the staining results are not readily explained at molecular levels. Features Associated with Thermophily | Only 15 CDS predicted in T. tengcongensis appear unique to thermophiles, which are found in various thermophlic genomes but not shared by all of them. Only a single copy of reverse gyrase (TTE1745) seems common to most, if not all, thermophiles. Other genes include CODH maturation factor (TTE1709), MinD superfamily P-loop ATPases (TTE1891 and TTE1892), metal-dependent hydrolase of the beta-lactamase superfamily II (TTE1889), predicted methyltransferase (TTE1898), uncharacterized Fe-S center proteins (TTE0177), uncharacterized Fe-S protein PflX (TTE1779), and conserved hypothetical proteins (TTE0285, TTE1224, TTE1505, TTE2664, TTE2667, TTE2636, and TTE2662). It is unlikely that thermophiles would have unique cellular machinery to make themselves capable of living in the extreme environment; rather it could be a result of an evolutionary process leading to the changes at many levels of their biochemical makeup (i.e., proteins and RNAs) and physiology . A strong correlation is observed between G + C contents of tDNA/rDNA clusters and the optimal growth temperatures (OGT) in all 12 sequenced thermophiles . Similar finding has been reported recently in thermophilic archaea . No correlation has been observed between G + C contents of the overall genomic average and OGTs in these thermophiles. In hyperthermophilic archaea, the chromosomes exist as relaxed to positively supercoiled in vivo due to the action of the enzyme, reverse gyrase, and this peculiarity is believed relevant to the stabilization of DNA double-helix against heat-denaturation . In mesophiles, a correlation between G + C contents of rDNA/tDNA and the genome average becomes noticeable . When G + C contents of all the sequenced mesophiles are analyzed, the linear regression coefficients are R = 0.88 for rDNA and R = 0.8 for tDNA, respectively. Nevertheless, especially in the case of mesophiles, G + C content changes not only affect the stability of functional RNAs but also have potential effects on amino acid composition of proteins. However, the interpretation of the underlying mechanism is expected to be statistical and multifaceted . Figure 4 | Correlation of G + C contents and optimum growth temperatures (OGT) of thermophilic bacteria. Correlation of G + C contents and optimum growth temperatures (OGT) of thermophilic bacteria. G + C contents of genomes (solid squares), rDNAs (solid circles), and tDNAs (solid triangles) of 12 thermophilic archaea and eubacteria are plotted against the corresponding OGT. G + C contents of tDNAs and rDNAs show significant correlation with OGTs (linear regression coefficients R = 0.9 and R = 0.92, respectively), but no significant correlation is observed between genomic G + C contents and OGT (R = 0.09). Figure 5 | Correlation of G + C contents between the genome average and rDNA/tDNA clusters from 36 mesophiles. Correlation of G + C contents between the genome average and rDNA/tDNA clusters from 36 mesophiles. G + C contents of tDNA and rDNA (underlined) show significant correlation with genome G + C contents (linear regression coefficients R = 0.88 and R = 0.8, respectively). Numbers in the figure stand for the sequenced prokaryotes: 1, Uure; 2, Buch; 3, Mpul; 4, Bbur; 5, Rpxx; 6, Cjej; 7, Cace; 8, Mgen; 9, SaurN; 10, Llact; 11, Hinf; 12, Spyo; 13, Hpyl; 14, Spneu; 15, Mpneu; 16, Pmul; 17, Cpneu; 18, Ctra; 19, Bsub; 20, Bhal; 21, Vcho; 22, Synecho; 23, Ecoli_O157; 24, Ecoli; 25, Nmen; 26, Xfas; 27, Tpal; 28, Mlep; 29, Atum; 30, Smel; 31, Mlot; 32, Mtub; 33, Paer; 34, Drad; 35, Ccre; and 36, Hbsp. The addition of the T. tengcongensis genome sequence to the growing list of sequenced microbes provides a pivotal view on the genome biology of thermophilic prokaryotes. However, to understand how thermophiles adapt themselves to the ever-changing environment over evolutionary timescale is still an ongoing effort. Systematic computational analysis and experimental verification of complex cellular and molecular mechanisms are essential for understanding the conservation and diversification of bacterial genomes regarding to their many specialized lifestyles. Valuable hypotheses and insights from such endeavors will be applied to medical research and the developing biotech industry. METHODS : Sequence Assembly and Quality Control | Genomic DNA libraries were made in pUC18 carrying insert sizes from 1.5 to 10 kb. The genomic DNA was isolated from a laboratory strain of T. tengcongensis, MB4T. To avoid cloning bias and to achieve optimal genome coverage, DNA inserts were prepared in two different ways, physical shearing (sonication) and enzyme digestion (Sau3AI). There were 75,971 successful sequence reads (>50 bp at Phred value Q20; ; ) generated, which gave rise to an overall genome coverage 9.87x, of which 2084 were from large insert libraries (similar10 kb) and sequenced from both ends. Phred/Phrap/Consed software package was used for quality assessment and sequence assembly. The initial assembly yielded 273 contigs. The number of gaps was effectively reduced to 46 by two basic steps. One is to resequence the low-quality reads flanking the contig ends. The other is to carry out intensive primer walking, based on the sequence information from the initial contig assembly and by using plasmid clones that extend outwards from the contigs as PCR templates. The remaining gaps were closed by a random primer-walking strategy against each contig ends. Some of the larger gaps were closed by long-range PCRs (Advantage Genomic PCR Kit, CLONTECH). In the latter cases, genomic DNA was used as template for PCR amplifications. All gap-closing clones and PCR products were sequenced from both directions to ensure high-sequence quality. The low-quality regions, often a few dozen base pairs, were improved by PCR-based methods. The overall sequence quality of the genome was further improved by insisting the following: (1) three independent, high-quality reads as minimal coverage, (2) sequence coverage accountable from both strands, and (3) Phred quality value >Q40 for each given base. Collectively, an additional 4089 finishing reactions were added to the final assembly at the finishing stage. Based on the final consensus quality scores generated by Phrap, we estimated an overall error rate of 0.86 in 10,000 bases for the final gap-free genome assembly. Physical Map Verification | The complete sequence assembly was verified based on restriction digests of genomic DNA with a panel of three restriction enzymes. DNA fragments were resolved on 1% agarose gels in a pulse-field electrophoresis system (Bio-Rad) at 4 volts/cm in 0.5x TBE buffer for 23 h at 14C. Lambda DNA concatemer was used as molecular-weight markers. All major fragments resolved by the electrophoresis system were unambiguously identified, including fragments for Sfi I (790, 760, 530, 279, 270, and 58 kb), Asc I (1398, 594, 498, 104, 40, 30, and 23 kb), and SgrA I (504, 447, 354, 327, 291, 158, 153, 145, 109, 56, 40, 37, 28, 17, 12, 4, 2, and 1 kb). The result was in complete agreement with the predicted physical map based on the fully assembled genome sequence to the extent that the restriction fragments were resolvable within the dynamic range of the electrophoretic system. Sequence Annotation | The first set of potential CDS were established with GLIMMER 2.0 trained with a set of CDS larger than 500 bp from the genomic sequence and with ORPHEUS at their default settings. Both predicted CDS and putative intergenic sequences were subjected to further manual inspections. Exhaustive BLAST searches with an incremental stringency against NCBI nonredundant protein database were performed to determine homology. Translational start codons were identified based on protein homology, proximity to ribosome-binding site, relative positions to predicted signal peptide, and putative promoter sequences. Rho-independent transcription terminators were identified based on TransTerm in nonprotein coding regions. A few methodological criteria were followed to resolve problematic cases. For instance, when two translation starts were identified, the first was always chosen to yield a larger predicted protein. When frameshifts and point mutations were discovered from two adjacent CDS, they were classified as inactive or pseudogene after careful inspections of the raw sequence data. When significant overlaps of two predicted CDS were encountered, those showing similarity to known genes or protein motifs/domains were preferentially taken, and the longer one was always the choice unless a biological argument favored the shorter. CDS <150 bp, which lack detectable similarity to known protein motifs/domains and distinguishable promoter/termination regions, were also excluded from the annotated CDS. The results were assembled together with manual refinements with Artemis sequence viewer . Each gene or CDS was then assigned with a unique numeric identifier prefixed with "TTE". The first CDS from the origin of replication, the putative dnaA gene, was assigned as TTE0001, and each subsequent CDS was numbered consecutively in a clockwise direction. To find putative orthologs in other completed genome sequences, CDS from the genomes were identified based on the COG database and classified accordingly . Protein motifs and domains of all CDS were documented based on intensive searches against publicly available databases and by using their application tools, including Pfam, PRINTS, PROSITE, ProDom, and SMART. The results were summarized with InterPro . Transfer RNAs, together with tRNA-like and mRNA-like sequences (such as 10Sa RNA or SsrA; see also ), and RNase P genes were predicted with tRNAscan-SE . The program was trained with a prokaryotic dataset and by using suggested procedures at tmRDB and RNase P databases . Signal peptides, transmembrane domains, putative membrane proteins, and ABC transporters were defined with TMHMM and SIGNALP-2.0 after intensive trainings with a dataset of gram-negative bacteria. Sequence data for comparative analyses were obtained from NCBI databases . When there was more than one strain sequenced for a given species, only one was chosen arbitrarily for the comparative study. Forty-seven fully sequenced genomes were used in the analyses. Their full names and abbreviations (in parentheses) are as follows: Agrobacterium tumefaciens (Atum), Aeropyrum pernix (Aero), Aquifex aeolicus (Aquae), Archaeoglobus fulgidus (Aful),Bacillus halodurans (Bhal), Bacillus subtilis (Bsub), Borrelia burgdorferi (Bbur), Buchnera sp APS (Buch), Campylobacter jejuni (Cjej), Caulobacter crescentus (Ccre), Chlamydia trachomatis (Ctra), Chlamydophila pneumoniae CWL029 (Cpneu), Clostridium acetobutylicum (Cace), Deinococcus radiodurans (Drad), Escherichia coli K12 (Ecoli), Escherichia coli O157:H7 EDL933 (Ecoli_O157), Haemophilus influenzae (Hinf), Halobacterium sp. NRC-1 (Hbsp), Helicobacter pylori 26695 (Hpyl), Helicobacter pylori J99 (Hpyl99),Lactococcus lactis (Llact), Mesorhizobium loti (Mlot), Methanobacterium thermoautotrophicum (Mthe),Methanococcus jannaschii (Mjan), Mycobacterium leprae(Mlep), Mycobacterium tuberculosis H37Rv (Mtub), Mycoplasma genitalium (Mgen), Mycoplasma pneumoniae(Mpneu), Mycoplasma pulmonis (Mpul), Neisseria meningitidis MC58 (Nmen), Neisseria meningitidis Z2491 (NmenA), Pasteurella multocida (Pmul), Pseudomonas aeruginosa (Paer), Pyrococcus abyssi (Pabyssi), Pyrococcus horikoshii (Pyro), Rickettsia prowazekii (Rpxx), Sinorhizobium meliloti (Smel), Staphylococcus aureus N315 (SaurN), Streptococcus pneumoniae(Spneu), Streptococcus pyogenes (Spyo), Sulfolobus solfataricus (Ssol), Synechocystis PCC6803 (Synecho),Thermoplasma acidophilum (Tacid), Thermoplasma volcanium(Tvol), Thermotoga maritima (Tmar), Treponema pallidum (Tpal), Ureaplasma urealyticum (Uure),Vibrio cholerae (Vcho), and Xylella fastidiosa(Xfas). To handle recursive-input sequences with efficiency, several custom-designed, perl-based scripts were also developed. The raw data were imported into an Oracle relational database. The user interface for this database was a series of web pages that allow frequent access to the databases. WEB SITE REFERENCE : ; Beijing Genomics Institute's T. tengcongensis genome project web site. ; Tmrna information web site. Backmatter: PMID- 11997337 TI - Identification of a Novel cis-Regulatory Element Involved in the Heat Shock Response in Caenorhabditis elegans Using Microarray Gene Expression and Computational Methods AB - We report here the identification of a previously unknown transcription regulatory element for heat shock (HS) genes in Caenorhabditis elegans. We monitored the expression pattern of 11,917 genes from C. elegans to determine the genes that were up-regulated on HS. Twenty eight genes were observed to be consistently up-regulated in several different repetitions of the experiments. We analyzed the upstream regions of these genes using computational DNA pattern recognition methods. Two potential cis-regulatory motifs were identified in this way. One of these motifs (TTCTAGAA) was the DNA binding motif for the heat shock factor (HSF), whereas the other (GGGTGTC) was previously unreported in the literature. We determined the significance of these motifs for the HS genes using different statistical tests and parameters. Comparative sequence analysis of orthologous HS genes from C. elegans and Caenorhabditis briggsae indicated that the identified DNA regulatory motifs are conserved across related species. The role of the identified DNA sites in regulation of HS genes was tested by in vitro mutagenesis of a green fluorescent protein (GFP) reporter transgene driven by the C. elegans hsp-16 --2 promoter. DNA sites corresponding to both motifs are shown to play a significant role in up-regulation of the hsp-16 --2 gene on HS. This is one of the rare instances in which a novel regulatory element, identified using computational methods, is shown to be biologically active. The contributions of individual sites toward induction of transcription on HS are nonadditive, which indicates interaction and cross-talk between the sites, possibly through the transcription factors (TFs) binding to these sites. [The following individuals kindly provided reagents, samples, or unpublished information as indicated in the paper: L. Hillier.] Keywords: Introduction : All living cells display a rapid molecular response to adverse environmental conditions, a phenomenon broadly termed as the heat shock (HS) response . The HS response is characterized by increased expression of a set of proteins, the heat shock proteins (Hsps), which have been conserved in evolution . The Hsps function as molecular chaperones in regulating cellular homeostasis and promoting survival . In eukaryotes, the enhanced HS gene expression has been shown to be regulated by the heat shock transcription factors (HSFs), which acquire DNA binding activity in response to various kinds of stress . A single HSF gene has been isolated from yeast Saccharomyces cerevisiae and Drosophila melanogaster . Several members of the HSF family have been shown to exist in vertebrates and plants (HSF1 --4) in which different HSFs are indicated to respond to various forms of stress. HSF1 in vertebrates is orthologous to HSF in the yeast and Drosophila and has been indicated as the HSF that mediates heat stress-induced expression of Hsps. HSF, in response to HS, binds to the DNA sites, commonly referred to as the heat shock elements (HSEs) characterized as multiples of the motif 5'-nGAAn- 3' . A few general transcription factors (TFs) are implicated to interact with the HSF for regulation of Hsp expression on HS. The HSF has been observed to interact with other general DNA binding factors, such as the TBP (TATA-box binding protein) and GAGA-factor , which aids the binding of the HSF to the Hsp promoters. Both the TBP and GAGA-factor occupy the promoter regions before induction by HS and are thus positioned to facilitate HSF recruitment. There has also been evidence that HSF1 can interact with the STAT-1 (Signal Transducer and Activator of Transcription-1) TF to induce the expression of HS genes in human peripheral blood cells treated with the cytokine interferon-gamma (IFN-gamma) . However, apart from the HSEs, no other cis-elements are known to be specifically responsible for induction of the HS genes on heat stress. We have been interested in studying the transcription regulatory mechanism in HS response. Gene expression patterns in C. eleganswere determined, before and after HS, using DNA microarrays containing probes for approximately 12,000 genes from the C. elegans genome. We followed a general scheme for identification and computational validation of potential transcription regulatory elements responsible for heat stress induction . Upstream promoter regions of the genes that were consistently up-regulated 1 and 4 hr after HS were analyzed using DNA pattern recognition programs. Two DNA motifs were found, the HSE (consensus: TTCTAGAA), which appears to represent the binding sites for the HSF, and HSAS (for heat shock associated site) (consensus: GGGTGTC), a previously unknown candidate regulatory motif. Statistical analyses and cross-species sequence comparison indicated that these motifs are significantly overrepresented in the promoter regions of the Hsps and are conserved across closely related species. Figure 1 | Schema describing the steps for identification and validation of transcription regulatory elements from coregulated genes. Schema describing the steps for identification and validation of transcription regulatory elements from coregulated genes. We determined the biological significance of the DNA motifs in regulation of HS genes using green fluorescent protein (GFP) technology. Two HSE and one HSAS site were predicted in the promoter region of hsp-16 --2. To monitor hsp-16 --2 promoter-dependent gene expression, a reporter construct was used that contained the hsp-16 --2 promoter fused to a GFP coding sequence. Transgenic C. elegans animals containing this construct showed strong GFP induction on HS. When DNA sites corresponding to the two motifs were mutated, the promoter was no longer inducible by heat stress. Therefore, in addition to the HSE we identified a novel DNA element that plays a significant role in the transcriptional regulation of HS genes. It was observed that mutation of multiple DNA sites was required to eliminate heat stress-induced expression from the hsp-16 --2 promoter, and the extent of expression induced by individual sites was non-additive. This indicates that an interaction between sites (possibly mediated through the TFs that bind to these sites) may be important for efficient transcription regulation of the HS genes under the HS condition. RESULTS : Identification of HS Up-Regulated Genes in C. elegans | Five microarray hybridization experiments were performed with independently prepared mRNA: In two experiments, mRNAs were taken from the animals that were harvested 1 hr after the HS treatment, and in three experiments, animals were heat shocked for 2 hr, allowed to recover for 2 hr, then harvested. Genes that were induced in at least four of the five experiments and were overexpressed by an average factor of two or more over the five experiments compared with the normal non-HS worms were identified . Table 1 | Genes Up-Regulated on Heat Stress DNA Motifs Identified from Promoters of HS Genes | Because some amount of noise is frequently observed in DNA microarray experiments, we considered only those genes up-regulated by an average factor of four or more for the purpose of transcription regulatory element identification. Genes F44E5.5 and F44E5.4 share a common upstream region of 450 nucleotides (nt), hence to avoid redundancy, only the upstream of F44E5.5 was considered. Two DNA motifs were identified by the application of DNA pattern recognition programs, Consensus and ANN-Spec , on regions upstream (-500 to -1) of the up-regulated genes . Given the relatively closely spaced gene distribution in C. elegans, the selected upstream regions are likely to contain relevant promoter elements. However, it is possible that the selected regions may exclude relevant motifs in genes with large promoters, long 5'UTRs, or membership in operons. We have used the translation start site (the -1 position) to select the candidate promoter regions because it is unambiguous, and because transcriptional start sites have not been determined for the large majority of C. elegans genes. One of the motifs (with a consensus sequence of TTCTAGAA) is the HSE, a well-known DNA binding site for the HSF; the other (with a consensus sequence of GGGTGTC) is called HSAS (for heat shock associated site). A thorough search of the Transfac database and published literature indicated that the HSAS motif is novel and does not correspond to any known TF binding sites. Figure 2 | DNA motifs identified from upstream regions of heat shock (HS) genes. DNA motifs identified from upstream regions of heat shock (HS) genes. Motifs identified by pattern recognition programs from the upstream regions (-500 ---1) of HS genes up-regulated genes. Information content (I.C.) in bits (log2) and sequence logos for the DNA motifs are given. DNA logos were generated according to . Determination of DNA Binding Probability and Cutoff Scores | A "site" corresponding to a particular motif in a sequence is simply a high-scoring subsequence that is obtained by the Patser program using the appropriate motif weight matrix as an input (see Methods). Weight matrices for the motifs were determined using the Consensus and ANN-Spec programs or were obtained from the Transfac database. From a consideration of the thermodynamics of protein --DNA interactions and the statistics of the scores , we expect that the score should be proportional to the free energy of binding. Therefore, at equilibrium, the probability of the protein binding to a site with a score, s, is simply: The exact proportionality factor depends on a number of things, including the availability of binding sites within the genome and the concentration of the TF in the nucleus, but because we only use it to rank different potential binding sites, we can ignore it. We also know that there are commonly multiple binding sites in the promoter region for a regulatory TF, so we calculate the probability that it will bind at any of those sites, referred to as the pp-value, as: where m denotes the DNA binding motif for the TF. This treatment is likely oversimplified, given the known cooperative binding of TFs to promoter elements. Nevertheless, more complicated models have not proven more effective for the analysis presented here, and this simplified approach has produced meaningful results (see below). For a given set of N sequences, the geometric mean of the pp-values is given by: For a motif, m, the appropriate cutoff score for eliminating low scoring subsequences is calculated as follows. The candidate promoter regions (-500 to -1) of the 13 genes that were up-regulated by fourfold or more on heat stress and 3000 random genes were obtained from the C. elegans genome. Considering both strands of the DNA sequences, the sites scoring above a particular arbitrary threshold were determined. The geometric mean of the pp-values for the motifs were obtained for the two sequence sets. The arbitrary threshold value was gradually increased from zero to a certain high positive value, and the cutoff that maximized the difference of the log of geometric means of the pp-values from the two sets (DLGM = log HS --log Rand) was chosen to be the appropriate cutoff value for the motif. For both HSE and HSAS, the cutoff values could be efficiently determined. For each of these motifs the DLGMs peaked at a certain threshold value before decreasing at higher thresholds. At low cutoffs, low scoring sites, which are present in substantial amounts in all sequences, are not eliminated. This results in a substantial number of sites being considered in the calculation of the pp-values, resulting in a small difference in the DLGM between the HS inducible and random promoters. As the threshold value is increased, the low scoring sites are eliminated leaving only the high scoring sites for calculation of the pp-values. For the HS regulatory motifs, the high scoring sites are expected to be more prevalent in the promoters of HS-inducible genes compared with random promoters, hence with increasing cutoffs, the DLGM value increases. As the cutoff is increased further, the high scoring sites are now ignored, eliminating those sites from pp-value calculation. Therefore for HSE and HSAS, the DLGM values decrease at high thresholds (at a very high cutoff value, where all sites are ignored, DLGM is zero). For several other motifs, the DLGMs remained low throughout the range tested and did not show a distinct maximum . Table 2 | Statistical Parameters for DNA Motifs For both HSE and HSAS, we calculated the average number of sites per sequence scoring above the respective cutoff values, and the geometric mean of the pp-values for the -500 to -1 regions for genes that are up-regulated by fourfold or twofold, and a set of 3000 genes picked at random from the C. elegans genome (Table A). For the purpose of comparison, the same parameters for four other unrelated patterns are shown: MSE (consensus: CCCGCGGGAGCCCG), a muscle-specific transcription regulatory element ; GATA (consensus: ACTGATAA), a potential intestine specific regulatory motif ; and two other DNA motifs, skn-1 and ces-2, taken from the Transfac database . skn-1 represents the DNA binding site (consensus: TAATGTCATCCA) for the C. elegans skn-1 protein, which is a TF required for the correct specification of certain blastomere fates in early C. elegans embryos , and ces-2 represents the DNA binding site (consensus: ATTACGTAAT) for C. elegans protein ces-2, a TF that controls the cell death fate of individual cell types in programmed cell death . For C. elegans, skn-1 and ces-2are the only two regulatory motifs for which weight matrices are available in the Transfac database. The ratios of the average number of sites per sequence for HS up-regulated genes to the random genes and the DLGMs are given in Table B. It can be seen that the DLGMs for HSE and HSAS are significantly higher compared with the other four unrelated motifs. Nonparametric Mann-Whitney Analysis of Identified Motifs for HS Genes | We took the upstream regions (-500 to -1) of all 19,804 genes from the C. elegans genome, determined the DNA sites for a motif, m, above the cutoff using the Patser program, and calculated the pp-value for each of the sequences (equation 2). A combined pp-value for multiple motifs, M, can also be calculated for the upstream sequence of each gene in the C. elegans genome. For lack of more specific information regarding the mode of TF binding and interaction, we assumed that for up-regulation of genes on heat stress (1) relevant TFs (corresponding to the motifs being considered) need to bind to the upstream sequence, and (2) if there are multiple sites scoring above the cutoff for a particular motif, any one of those binding sites may be occupied by the corresponding TF. For a particular upstream sequence, the combined pp-value for multiple motifs is calculated by taking a product of individual pp-values (from equation 2) for the motifs: All (19,804) upstream sequences were sorted according to the decreasing log of the pp-value, ln(Pmseq) (equation 2), for individual motifs or combined pp-value, ln(Pseq) (equation 4), for multiple motifs. Among the most commonly used biostatistical procedures is the comparison of two sample sets to infer whether differences exist between the two populations sampled. We have used the one-tailed Mann-Whitney nonparametric testing method to see whether the HS genes are placed significantly higher on the list of all genes sorted by the pp-values. We calculated the Mann-Whitney statistic for testing the null hypothesis, H0: Genes in a given set are placed no higher on the list of all genes sorted by the pp-value, compared with a random set of genes. The alternative hypothesis, HA was: Genes in a given set are placed higher on the list of all genes sorted by the pp-value, compared with random genes. For different lists generated by sorting the pp-values (as above), the Mann-Whitney statistic, U, was calculated for the HS genes up-regulated by a factor of four- or twofold. The U statistic can be used to determine z scores, which might be used to determine the probability for the null hypothesis. However, in our case the patterns were discovered using the same set of sequences that we are testing, and we do not have an independent test set. We use the z scores merely to measure the extent to which the identified motifs can help to distinguish the HS responsive genes from other genes and, in particular, to see if the HSAS motif increases the specificity of that observed for HSE alone. Not only are the z scores for the two identified motifs much larger than for other TF patterns, but only with the HSE --HSAS combination does the z score increase over the HSE z score alone. Table 3 | Mann-Whitney Statistic for Nonparametric, One-Tailed Test of the Null Hypothesis, H0: Genes in a Given Set are Placed No Higher on the List of all Genes Sorted by the pp-value, Compared with a Random Set of Genes Conservation of Regulatory Sites across Related Species | Each of the HS up-regulated protein sequences was searched for potential orthologs in the C. briggsae sequences using the gapped-BLAST method. For a given C. elegans protein, the C. briggsae protein, with the lowest (50 GFP+ nuclei were observed. The two non-GFP animals observed in the pPD122.18 and H1 mutated construct injections likely result from rare events when the reporter construct was not incorporated in the Rol marker-containing array. Transgenic animals were scored as "any GFP" if at least one GFP positive nucleus could be observed. Footnotes: a, GFP+ animals from these injections contained an average of 5.5 (predominantly neuronal) GFP positive nuclei; b, one GFP positive animal was observed with 10 GFP+ head neurons; c, one head neuron in one animal was GFP positive. Figure 5 | Effect of HSAS element on reporter transgene expression. Effect of HSAS element on reporter transgene expression. First generation (F1) transgenic animals containing a GFP reporter construct with either both HSE2 and HSAS mutated (HSE1 only) or HSE2 mutated (HSE1 + HSAS) were heat shocked at 35C, returned to 20C for 18 hr, then imaged for GFP expression. This digitally merged differential interference contrast/epifluorescence image shows the requirement for the HSAS element for strong transgene expression when the HSE is mutated. Size bar = 50 mum. DISCUSSION : Identification of a New Transcription Regulatory Element | We have identified a set of genes reproducibly induced by HS in C. elegans using DNA microarray hybridization. These initial experiments involved a limited set of hybridizations and a single developmental stage and used microarrays containing probes for similar of the predicted C. elegans genes. Therefore, it is likely that we have only identified a subset of all the C. elegans genes induced by HS. Nevertheless, these studies have enabled us to identify a novel HS-responsive promoter element by computational DNA pattern recognition methods followed by statistical analysis. The role of this element (HSAS), along with the other well-known element (HSE), has been shown for one of the Hsps in vivo by mutational analysis using GFP reporter constructs introduced into transgenic animals. This is one of only a few instances in which a completely novel cis-regulatory site, identified solely by computational DNA pattern recognition methods, has been supported by experimental evidence (e.g., ; ; and ). Traditionally, time-con-suming experimental methods such as systematic sequence deletions and mutations have been used to identify cis-regulatory regions and sites responsible for regulation of a particular gene. We note that the elements we have identified, being individually neither necessary nor sufficient for full reporter expression, would be difficult to identify using standard molecular techniques. With the advent of techniques like SAGE or DNA microarrays cohorts of coregulated genes can be easily identified. Because the genes that show similar expression profiles are assumed to have similar transcriptional mechanisms governing their expression, DNA pattern recognition methods should be a very useful way of identifying the cis-elements governing the expression of a set of coregulated genes . Statistical Significance of the Identified Motifs | One important concern regarding DNA pattern recognition methods for regulatory element identification is the significance and specificity of individual motifs. Because the ANN-Spec program takes into consideration a background sequence set and discriminates against commonly occurring motifs in the background, it is designed to yield only those DNA patterns that are specific toward the training set . However, statistical validation of the motifs identified by the pattern recognition programs is useful for selecting the most promising candidate motifs for further experimental verification. We find that several parameters like the pp-values and the Mann-Whitney z scores (Tables and ) are useful measures of the specificity of an identified DNA motif. It is expected that if a set of genes are regulated by a common cis-regulatory motif, then the upstream promoter regions of those genes will either contain high scoring site for that motif or contain multiple sites (clusters of sites) or both. In either case, the pp-values (equation 2) should be significantly higher for the coregulated genes compared with other genes. With HSE or HSAS, the mean pp-values or the Mann-Whitney z scores are substantially higher for the HS genes. Cross-Species Conservation of Regulatory Elements | One method that is very useful for identifying conserved DNA motifs is cross-species comparative sequence analysis . We identified the C. briggsae orthologs of several C.elegans Hsp genes using BLAST searches. On the basis of BLAST alignments, we determined that genes CB024O8.9 and CB024O8.10 were orthologous to genes F44E5.5 and F44E5.4. The DNA sites corresponding to HSE and HSAS are remarkably similar in the upstream regions of these orthologous genes. Analysis of the upstream region of T27E4.2 and its likely C. briggsae ortholog G39L17.4 also revealed conservation of HSE and HSAS sites (data not shown). As more sequences become available from C. briggsae and other related organisms, comparative sequence analysis and phylogenetic footprinting will become a powerful tool for identification of DNA regulatory motifs and adding confidence to the motifs identified by DNA pattern recognition methods. Interestingly, the human small HS genes HSPB2 and CRYAB, like the C. elegans HS gene pairs T27E4.2/T27E4.8 and F44E5.4/F44E5.5, are also closely linked and transcribed divergently, sharing a putative promoter region of <1 kb . This promoter region contains four possible HSAS sites, in addition to two classic HSE sites. It is therefore possible that the HSAS element also plays a role in HS-dependent gene expression in other non-nematode species. Identification of HSE Motif | Our identification of a motif corresponding to the classic HSE was not unexpected, as C. elegans contains a single gene (Y53C10A.3) that encodes a likely ortholog of HSF. Double-stranded RNA inhibition (RNAi) of Y53C10A.3 expression reduces HS induction of an hsp-16 --2/GFP reporter transgene (C.D. Link, unpubl.), supporting the view that the identified HSE motif does function in a typical HSF-dependent HS induction mechanism. Mode of TF-DNA Interaction in the Hsp Promoters | A promoter element is organized in a hierarchical manner: Individual binding sites are organized in specific arrays to form 'promoter modules', which are substructures of the functional promoter, and the complete promoter element is composed of specifically organized promoter modules . Hence, individual binding site detection, although important, is not sufficient for understanding the regulatory mechanism and elucidation of complete promoter function . The following discussions illustrate our efforts toward determining the relationship between the HSE and HSAS sites with the goal of understanding the regulatory mechanism of the Hsps. In addition to the mutations of HSE and HSAS sites, we did three preliminary experiments in which we implanted either a single HSE, two closely paired HSEs, or a single HSAS in a "virgin" promoter, devoid of any known heat-inducible elements. These experiments did not attempt to replicate the spacing of these elements. In no case did we observe induced expression from the promoter on HS. This result indicates that (1) both HSE and HSAS may be required, (2) specific distances between the two or more HSEs or HSASs might be important for heat induced expression, or (3) additional sites remain to be identified. We note that the comparison of C. elegans and C. briggsae orthologous HS promoters identified a well-conserved stretch of nucleotides (AGAGACGCAGA) upstream of the HSAS that might represent such a site (Figs. D and A). However, this candidate site is not generally found among the HS-inducible genes identified in our gene expression analysis, perhaps because it has regulatory functions specifically in highly induced, divergently transcribed HS gene pairs such as F44E5.4/F44E5.5 and T27E4.2/T27E4.9. It is also possible that we may have missed additional sites by using rather stringent cutoffs that eliminated some low scoring sites that could be biologically functional. Weaker sites may also be placed at optimal distances from each other so that multiple TFs can bind the respective DNA sites and maintain intermolecular (TF --TF) interaction at the same time. This is observed in cooperative DNA binding by the TFs, in which the binding of a TF to its DNA site may be weak (which can happen when a DNA binding site is low scoring, i.e., does not conform well with the consensus), but cooperative binding with another TF, which binds a nearby DNA site, may be strong enough for stable TF --DNA complex formation. It is interesting to note that single-site mutants do not affect the HS induction of hsp-16 --2; however, double- and triple-site mutants have a dramatic effect on the expression of the gene (Fig. B). The extents of transcription induced by the individual sites are nonadditive (Fig. B), indicating an interaction between the sites, probably mediated through protein --protein interactions between TFs binding to those sites. If this is the case, it can be imagined that the TF binding sites would be located at certain distances where optimal TF --TF interactions can occur. By use of DNA sites corresponding to HSE and HSAS motifs, we tried to build a consistent model for the organization of TF binding sites in the promoter regions of the Hsps, which would distinguish these genes from all other genes in the genome. We studied the strength, frequency of occurrence, and distances between the TF binding sites. However, we failed to obtain such a model. It is clear that many genes with high rankings for HSE, HSAS, or HSE + HSAS sites do not appear to be HS inducible in our gene expression analysis. Furthermore, our hsp-16 --2 reporter mutagenesis studies indicate that individual HSE or HSAS elements are neither necessary nor sufficient for full HS inducibility. This indicates that, although we have identified a new HS regulatory element, our understanding of the regulatory mechanism governing the HS response is still incomplete. We note that the most strongly HS-inducible genes identified in our studies are arranged as divergently transcribed gene pairs (e.g., Fig. A); the contribution of this gene arrangement to HS regulation is unknown. The HSAS site identified in the hsp-16 --2 promoter overlaps a 13-bp imperfect repeat previously indicated to be capable of forming a hairpin structure . It is also possible that other TFs are involved in the transcriptional pathway involving HSE and HSAS, which needs to be elucidated (some TF sites may be in further upstream regions that remain to be analyzed) or that the HSE and HSAS sites are organized in subtle patterns that remain difficult to identify computationally at this point. Involvement of an alternative transcription regulatory pathway, which uses a different set of TFs, is also a distinct possibility. We intend to address these complex issues of TF --DNA interaction further with computational and experimental means in the future. METHODS : Sequences and Gene Annotations | All C. elegans sequences and their annotations were obtained from the WormBase web-site . C. briggsae, a closely related nematode to C. elegans, is currently being sequenced at the Washington University Genome Sequencing Center, St. Louis, Missouri. We obtained the DNA and protein sequences for C. briggsae from the Washington University Genome Sequencing Center and L. Hillier (pers. comm.). cDNA Microarray Experiments and Identification of C. elegans HS Genes | The microarray data were compiled from five independent HS experiments. These experiments were originally designed to investigate whether age-1 mutant animals, which have increased intrinsic thermotolerance , have an altered gene expression response after HS when compared with wild-type animals. Age-synchronous (4-d-old) populations of wild-type or age-1(hx546) animals were harvested as young adults, then split, and half of the populations were heat shocked at 35C. and half of the animals were maintained at 20C to generate a control population. Two HS regimes were used: 'immediate response,' in which animals were harvested immediately after 1 hr of HS (two experiments: one wild-type and one age-1 population), and "recovery response," in which animals were heat shocked for 2 hr, then allowed to recover for 2 h4 (20C) before harvesting (three experiments: two wild-type and one age-1 population). Poly A+ RNA was prepared from these populations and reverse transcribed into Cy3- or Cy5-labeled cDNA; then HS and control cDNAs were cohybridized to glass-slide DNA microarrays containing probes for 11,917 known or predicted C. elegans genes, as previously described . Relative HS-dependent expression changes for each gene was calculated from the ratios of Cy3 and Cy5 hybridization signals. No significant difference in HS-dependent gene expression was observed between age-1 and wild-type animals in this dataset. We therefore compiled the data from the five experiments to generate a list of genes that showed reproducible HS induction independent of genotype or HS regime . Identification of DNA Motifs | Two DNA pattern recognition programs, ANN-Spec and Consensus , were used to identify significant DNA patterns from the promoter regions (-500 to -1 relative to the translation start) of the HS genes up-regulated by an average factor of four or more over the five cDNA experiments . Consensus and ANN-Spec are local multiple sequence alignment programs that run on a given set of sequences (training set) to identify conserved motifs commonly present in those sequences. Both the programs use weight matrix-based models to represent ungapped DNA sequence motifs. Because the cis-regulatory sites in a set of similarly regulated sequences are expected to be conserved to a certain extent, the conserved motifs identified by these programs represent potential regulatory elements. Consensus | The Consensus program uses a greedy algorithm and searches for a matrix with a low probability of occurring by chance or, equivalently, having a high information content. Version 6.c of Consensus was used and the top scoring result was reported. Different pattern lengths were tested, and both strands of the DNA were searched for motifs because TFs can bind to either strand. The patterns with high information content and the lowest expected frequency were considered. ANN-Spec | ANN-Spec uses a simple artificial neural network and Gibbs sampling method to define DNA binding site patterns. The program searches for the parameters of a simple perception network (weight matrix) that maximize the specificity for protein (TF) binding to a positive sequence set (or training set) compared with a background sequence set. The use of background sequences allows the method to find patterns with greater discriminatory capability when compared with the original version of the Gibbs sampling method . Binding sites in the positive data set are found with the resulting weight matrix and these sites are then used to define a local multiple sequence alignment. ANN-Spec Version 1.0 was used. A background sequence set of upstream regions from 3000 randomly picked genes was used. Different motif lengths were tried and both strands of the DNA were searched for motifs. Because of the nondeterministic nature of the algorithm, multiple training runs are performed (100), with each run iterating 2000 times. The results were sorted by their best attained objective function values. Weight matrices corresponding to the ten highest scoring runs were compared and if more than five of these top scoring ten runs give a motif with one consistent pattern consensus, that pattern is considered significant. Calculation of "Site" Scores and Searching for "Sites" in Sequences | A position weight matrix (PWM) has previously been found to be a good model for describing protein binding sites in DNA . An l-long DNA binding site pattern may be described by a 4xl weight matrix, with four weights (for four DNA nucleotides) per pattern position. Let us assume each weight in the matrix is the binding energy contribution of each nucleotide at a particular pattern position. With the additional assumption that protein --DNA contacts at individual residue positions in the binding site are independent of each other , the total binding energy (or score) for a TF molecule to a particular site is given by: where, omega denotes the PWM weights, x denotes the inputs from the site (DNA bases at different positions), k ranges over the l positions of the site, and b ranges over all four DNA bases. The Patser program (G.Z. Hertz and G.D. Stormo, unpubl.) allows one to score the words of a given sequence against a weight matrix. Once the weight matrices for regulatory motifs are obtained by Consensus, ANN-Spec, or from the Transfac database the matrices can be used as input for Patser to identify high scoring subsequences (or "sites") in given sequences. Patser also calculates the p value (or probability) of observing a particular score or higher at a particular sequence position . Nonparametric Analysis with Mann-Whitney Statistics | Nonparametric or distribution-free tests may be applied in any situation in which actual measurements are not used, but instead the ranks of the measurements are used. The data may be ranked either from highest to lowest or from lowest to highest values. In our case, we have the pp-values for all C. elegans genes, based on the DNA binding site motifs, arranged in decreasing order. We use the nonparametric analog of the two sampled t test, commonly known as the Mann-Whitney test . We take the sorted list of pp-values calculated using either individual motifs or combinations of motifs. We then consider two sets of ranks of the HS up-regulated genes and that of the random genes. Because the labeling of the two samples as 1 and 2 is arbitrary, the Mann-Whitney statistic can be calculated as one of two ways: where, n1 and n2 are the two sample sizes and R1 and R2 are the summation of the ranks in the two samples. Usually, U and U' are different, and we take only the larger of the two values. It is known that the distribution of the Mann-Whitney statistic approaches normal distribution for large samples with a mean, muu, of n1n2/2. In our case the random sample size is large (3000), so we can use the above approximation. We observe that usually several ranks in the samples are tied; in such cases the standard error is given by: where, Sigma T = ti3 --ti, ti is the number of ties in a group of tied values, and N = n1 + n2. Therefore, if a U is calculated from data where either n1 or n2 is large, the significance of U can be determined by computing the test statistic: Recalling that the t distribution with infinite degrees of freedom is identical to the normal distribution, the critical value of Z is equal to the critical value of tinfinity. Construction of Reporter Plasmids and DNA "Site" Mutation | The vector pPD122.18 (Fire lab 1999 expression vector kit, see ) was used for the construction of the mutated promoter constructs. This plasmid contains the entire promoter element for the hsp-16 --2/16 --41 gene pair from C. elegans oriented so that it drives the expression of a GFP coding sequence with four nuclear localization signals from the hsp-16 --2 side. (Use of a nuclear-targeted GFP construct simplified quantitation of GFP induction, as it allowed simple counting of GFP+ nuclei.) The promoter of the hsp-16 --2 gene contained two HSE and one HSAS sites (Fig. A). The two HSE sites had identical sequences corresponding to the HSE consensus, TTCTAGAA, whereas the HSAS site sequence was GGGTCTC. The DNA sites of interest were mutated in the promoter region using the Stratagene QuickChange protocol , which allows high efficiency mutagenesis. All these sites were altered by substituting noncomplementary bases at all positions (Fig. A), and the sequences of the altered sites were confirmed by DNA sequencing. Microinjection of Vectors and Identification of HS Expression Pattern Using GFP | Injection mixtures were prepared with 100 ng/muL pPD122.18 (or promoter mutation derivative) and 100ng/muL pRF4 (dominant rol-6 marker). Approximately 20 N2 (wild-type) animals were injected for each construct; typically half of the injected animals segregated F1 roller animals. Pooled F1 progeny were propagated at 16C for 5 d after the injection, heat shocked for 2 hr at 35C, and then returned to 16C. Rol F1 animals were recovered and mounted on slides for assaying GFP response 14 --16 hr after the end of the HS. Animals were scored using a 40x objective on an Axioskop epifluorescence microscope. pPD122.18 F1 animals were observed carefully for GFP expression before and after the HS treatment; GFP expression was found to be completely HS dependent. WEB SITE REFERENCES : ; FireLabWeb. ; Washington University Genome Sequencing Center. ; Transfac database. ; VISTA. ; Stormo lab web site. Access to DNA motif finding program. ; WormBase. Backmatter: PMID- 11997338 TI - Genes in a Refined Smith-Magenis Syndrome Critical Deletion Interval on Chromosome 17p11.2 and the Syntenic Region of the Mouse AB - Smith-Magenis syndrome (SMS) is a multiple congenital anomaly/mental retardation syndrome associated with behavioral abnormalities and sleep disturbance. Most patients have the same similar4 Mb interstitial genomic deletion within chromosome 17p11.2. To investigate the molecular bases of the SMS phenotype, we constructed BAC/PAC contigs covering the SMS common deletion interval and its syntenic region on mouse chromosome 11. Comparative genome analysis reveals the absence of all three similar200-kb SMS-REP low-copy repeats in the mouse and indicates that the evolution of SMS-REPs was accompanied by transposition of adjacent genes. Physical and genetic map comparisons in humans reveal reduced recombination in both sexes. Moreover, by examining the deleted regions in SMS patients with unusual-sized deletions, we refined the minimal Smith-Magenis critical region (SMCR) to an similar1.1-Mb genomic interval that is syntenic to an similar1.0-Mb region in the mouse. Genes within the SMCR and its mouse syntenic region were identified by homology searches and by gene prediction programs, and their gene structures and expression profiles were characterized. In addition to 12 genes previously mapped, we identified 8 new genes and 10 predicted genes in the SMCR. In the mouse syntenic region of the human SMCR, 16 genes and 6 predicted genes were identified. The SMCR is highly conserved between humans and mice, including 19 genes with the same gene order and orientation. Our findings will facilitate both the identification of gene(s) responsible for the SMS phenotype and the engineering of an SMS mouse model. Keywords: Introduction : Smith-Magenis syndrome (SMS) is a contiguous gene-deletion syndrome , in which a distinct and consistent phenotype is associated with deletion of a portion of chromosome band 17p11.2 . The clinical features of SMS patients include mental retardation, delayed speech and motor development, behavior problems, sleep disturbance, minor craniofacial abnormalities, short stature, and brachydactyly . Less common features include otolaryngological abnormalities, hearing impairment, opthalmological abnormalities, and renal and cardiac abnormalities . The incidence of SMS is similar1:25,000 births, which is likely underestimated given the often subtle clinical features, particularly early in life . A heterozygous interstitial deletion of similar4 Mb in chromosome 17p11.2 was identified in >90% of SMS patients, whereas the remaining patients have rare smaller- or larger-sized deletions . To narrow the critical interval responsible for the SMS phenotype, STS content mapping was performed on somatic cell hybrids refining the deleted chromosome from 62 SMS patients. A common deletion region was defined between markers D17S58 and cDI17-498 . Ten patients with a deletion distinct from the common deletion region were identified, and molecular analyses of these patients delineated an SMS critical interval between D17S29 and cCI17-638 . Three copies of a low-copy repeat (SMS-REPs) were identified within the common deletion region, and homologous recombination and unequal crossing over between the flanking SMS-REPs were shown to comprise the mechanism responsible for the genomic deletion . The predicted reciprocal duplication of the SMS common deletion region dup(17)(p11.2p11.2), which was identified recently, causes a subtle clinical syndrome . The first gene identified within the SMS common deletion region, snU3, encodes a small nuclear RNA U3 . Since then, more than 15 genes have been mapped. In addition, a large number of ESTs with no homology to known genes were identified . However, a potential role for any of these genes in the SMS phenotype through haploinsufficiency effects remains unclear. The shaker-2 (sh2) mouse represents a mouse model for human deafness (DFNB3), resulting from mutation of MYO15A . A physical map of the sh2 region revealed that 11 genes mapping within the SMS common deletion region have murine homologs in the sh2 region . The gene order was not completely conserved secondary to two independent genomic inversions. However, the order of the seven genes within the critical region was apparently conserved between humans and mice. To identify potential causative gene(s) of the SMS phenotype and to facilitate the construction of an SMS mouse model, we constructed a large insert clone contig of the SMS common deletion region and its syntenic region in the mouse. Here we report the comparative genomic analysis between humans and mice, comparison of the human genetic and physical maps, delineation of the SMS critical region (SMCR), and the identification and characterization of genes in both human SMCR and the mouse syntenic region. RESULTS : Construction of the BAC/PAC Contig of the SMS Common Deletion and Its Syntenic Region in the Mouse | We constructed a BAC/PAC contig covering the SMS common deletion region with a minimal tilting path of 30 clones . Three low-copy repeat gene clusters were identified previously inside and flanking the SMS deletion region . Clones specific to an individual SMS-REP were identified based on cis-morphisms, sequence differences among repeats on the same chromosome . From the DNA sequence of these clones, we deduced that the size of the common deletion region including the three SMS-REPs is similar3.7 Mb and the sizes of the SMS-REPs are between 176 and 256 kb. The middle SMS-REP is inverted with respect to the direct orientation of the proximal and distal copies . Figure 1 | The Smith-Magenis critical region (SMCR) refined by breakpoint studies of patients with deletions in 17p11.2. The Smith-Magenis critical region (SMCR) refined by breakpoint studies of patients with deletions in 17p11.2. The SMS common deletion region falls between D17S959 and D17S1857, including SMS-REPs. Above are shown genetic markers and the cytogenetic bands on 17p. TEL represents telomeric orientation, and CEN represents centromeric orientation. The minimum BAC/PAC tiling path of the SMS common deletion region is shown toward the top of the figure, with STS-content markers represented by dots and BAC/PAC clones represented by horizontal bars. Clones without a prefix are BACs from RPCI-11; those with prefix P are PAC clones; those with prefix C are CTD clones. BAC end sequences were used as markers with R representing the BAC end sequence derived from the Sp6 primer, and F representing the sequence from the T7 primer. Above the BAC contig are listed the individual genes and genetic markers from this region. Below, patients are identified by number. The deleted region is indicated by dashed lines, whereas bold lines refer to genomic sequences retained. The distal breakpoints of patients 357 and 765 are outside the SMS common deleted region. The refined similar1.1-Mb SMCR (double-edged bold arrow) and the similar210 kb (hatched box) inside the SMCR, but not deleted in patient 765, are indicated. We also constructed a complete BAC contig covering the syntenic mouse region . The DNA sequence from these BACs indicated that the SMS-REPs are not present, consistent with previous hybridization-based observations . The 28 genes that mapped within or around the common deletion region of SMS have homologs on the syntenic region in mouse chromosome 11. Comparisons between humans and mice confirmed previous observations that gene order is conserved within the genomic interval flanked by the distal and middle SMS-REPs . In addition, we found that the order of genes ULK2 and AKAP10 is also conserved. The gene order between COPS3 and AKAP10 is conserved with the exception of an inversion between PRPSAP2 and ALDH3A1. In contrast, the gene TACI is located centromeric to distal SMS-REP, whereas KCNJ12 is centromeric to proximal SMS-REP. In the mouse, these genes are adjacent and located in the middle of the syntenic SMS region. PMP22, the gene responsible for Charcot-Marie-Tooth disease type 1A (CMT1A) and hereditary neuropathy with liability to pressure palsies (HNPP; ), mapped telomeric to distal SMS-REP, but it is located in an inverted position with respect to the syntenic SMS region in mice. Thus, during evolution disruptions of linkage conservation apparently occurred in the genomic region around the SMS-REP low-copy repeats. Figure 2 | Comparison of the gene order in the human SMS common deletion region and its mouse syntenic region. Comparison of the gene order in the human SMS common deletion region and its mouse syntenic region. Genes within the human SMS region on 17p11.2 are shown above. Open boxes represent the SMS-REPs. Below is shown the minimum BAC tiling path of the mouse syntenic region of the SMS common deletion interval and its flanking region. Each BAC clone is represented by a horizontal bar with STS-content markers represented by dots. Blocks of genes that show linkage conservation (i.e., identical gene order) in humans and mice are boxed and connected via gray shading. Comparison between the Physical and Genetic Maps | We previously identified a striking difference in recombination rates between the sexes in the CMT1A genomic region , with reduced recombination noted for the male meiotic map. De novo CMT1A duplication occurs 10 times more frequently in male gametogenesis than in female germ cells . We hypothesized that reduced recombination may increase unequal crossing over resulting in an increased propensity to generate unequal reciprocal recombination products . We extended the genetic/physical map correlation over an similar8.0-Mb region including both the CMT1A and SMS regions . Reduced recombination is observed in both genders for most of the SMS region. Interestingly, no parent-of-origin frequency differences are observed for de novo SMS deletion . Figure 3 | Comparison between genetic and physical maps. Comparison between genetic and physical maps. The STR genetic markers from Marshfield are aligned to the sequence-based physical map. The marker order is as following: D17S1871, D17S959, D17S805, D17S1794, D17S620, D17S740, D17S2196, D17S1857, D17S953, D17S1843, D17S793, D17S918, D17S921, D17S1856, D17S947, and D17S1803 (markers within the SMS region are underlined). The three SMS-REPs are indicated by gray bars, the two CMT1A-REPs by hatched bars. Reduced recombination in both sexes was observed for most of the SMS region. Refining the Smith-Magenis Syndrome Critical Region (SMCR) | Somatic cell hybrid analysis revealed that the breakpoints in most patients with SMS were located within the distal and proximal SMS-REPs (; data not shown). We used an SMS junction fragment identified by PFGE analysis to distinguish SMS patients with the common deletion from those with an unusual-sized deletion. The breakpoints for these latter SMS patients were analyzed by FISH of patient lymphoblasts using BAC/PAC DNA probes; the results are summarized in Figure . Smaller-sized deletions within 17p11.2 were identified in six patients who have typical behavioral and physical features consistent with SMS . The deleted region for two SMS patients, 1190 and 1456, was mapped between the distal and middle SMS-REPs . Furthermore, the distal deletion breakpoints for SMS patients 1615, 1774, and 1354 were mapped proximal to COPS3, a gene located within the genomic interval between the distal and middle SMS-REPs . We therefore conclude that the genomic interval between COPS3 and the middle SMS-REP is the critical region for the major SMS features of mental retardation, craniofacial and behavioral abnormalities, and sleep disturbance. We propose that the dosage-sensitive genes responsible for the SMS phenotype are located inside this newly defined SMCR. Table 1 | Clinical Features of Patients with Atypical Deletions within 17p11.2 Molecular analyses of three patients, 540, 357, and 765, have been reported . By our clinical analysis, patients 765 and 357 do not have SMS because they do not show self-destructive behavior, sleep disorder, or characteristic SMS facies . Mapping of the deletion breakpoints in these patients by PCR analysis of hybrid DNA revealed that the deleted region for patient 765 is located distal to DRG2. Thus, the genomic region responsible for the SMS phenotype may be only similar210 kb in size . However, this conclusion is based on the absence of an SMS phenotype in one patient (765) only. Sequences of the SMCR and Its Syntenic Region in the Mouse | Ten BAC/PAC clones cover the entire SMCR . The location of individual clones was confirmed by FISH on lymphoblast chromosomes derived from SMS patients with common deletions. The genomic sequence assembly using public (NCBI, ) and private (Celera, ) genome databases revealed that the size of entire SMCR from the putative distal end of the middle SMS-REP to the promoter of COPS3 is similar1.1 Mb. RepeatMasker identified interspersed repeats that account for 42.19% of the SMCR. The repetitive elements include 20.54% Alu sequences and 6.19% LINE1 sequences, similar to that of chromosome 22, but different from chromosome 21, which contains 9.48% Alu and 15.51% LINE1 sequences, and the CMT1A/HNPP region in 17p12, that contains 9.97% Alu and 13.43% LINE1 elements . Seven BAC clones cover the mouse region syntenic to SMCR (Fig. ; NCBI). Genome sequences from the private mouse database (Celera) were initially used to assemble some unordered pieces in the public database. However, subsequent analysis of the public sequence database revealed more robust sequence (fewer small gaps) than the Celera database. The mouse genome sequence is similar1.0 Mb with interspersed repeats accounting for only 27.6%. This similar15% decrease in repetitive sequences in the mouse genome as compared to the human has also been observed in other genomic regions . Figure 4 | Transcript map of the SMCR and its mouse syntenic region. Transcript map of the SMCR and its mouse syntenic region. The BAC and PAC clones are identified by name and accession number. (A) Clones without a prefix are BACs from RPCI-11 (human) or from RPCI-23 (mouse); those with prefix P are PAC clones; those with prefix C are CTD clones. The hatched line represents the SMCR, and the open boxes represent the flanking sequences of the SMCR: COPS3 and the middle SMS-REP. The position 0 on the size scale is assigned to the nucleotide proximal to the putative promoter of the COPS3 gene. Genes inside SMCR are represented by black boxes, predicted genes by blue boxes, and pseudogenes by green boxes. The predicted genesUbl and Pabplp are located between Flj20308 and Llglh, and both are present as a gene cluster. To distinguish from Pabplp, gray boxes were used to represent the Ubl. Human genes transcribed from centromere to telomere are located above the central line, and genes transcribed from telomere to centromere are below the line. The mouse genes are drawn in the reciprocal manner (i.e., genes transcribed from telomere to centromere are above the central line). (B) The structure of the similar110-kb gene clusters between Flj20308 and Llghlthat is boxed in A. Seven copies of Ubl and four copies of Pabplp are intermixed. Solid lines connecting exons represent the splicing pattern. All copies are potentially transcribed in the same direction (arrows). The Ubl gene appears to have four copies with spliced variants. Transcript Map of Human SMCR and Its Syntenic Region in the Mouse | Potential genes inside the SMCR and the mouse syntenic region were identified through a combination of sequence similarity searches and sequence analysis using gene prediction programs. The potential genes were categorized into three groups: (I) genes, (II) predicted genes, and (III) pseudogenes, using definitions that we employed previously . Within the similar1.1-Mb SMCR, 30 genes were identified, including 20 genes (Group I), 10 predicted genes (Group II), and 3 pseudogenes (Group III; Fig. ; Table ). Thus, the gene density (1 gene per 37 kb) is much higher than the average calculated for the complete human genome (1 gene per 90 kb). The human SMCR genes are unevenly distributed; 23 genes are in the similar730-kb interval between the middle SMS-REP and SMCR2, and only 7 genes are in the remaining similar400-kb segment . Table 2 | Putative Genes Identified within the SMCR We identified 16 genes (Group I), 6 predicted genes (Group II), and 3 pseudogenes (Group III) in the mouse region syntenic to SMCR (Fig. ; Table ); 8 fewer than in the human. Similar to the gene distribution in the human SMCR, the gene density in the region between Shmt1 and Rai1 is higher than in the remaining portion. Comparison of human and mouse sequences indicated that 19 genes in the human SMCR have orthologs in the mouse syntenic region, with conservation of both the order and the orientation, and the same numbers of pseudogenes are present . Homology is higher across the exons, but extends to some introns and intergenic regions. Table 3 | Putative Genes Identified within the Mouse Syntenic Region of the SMCR Figure 5 | Comparison of the order of the putative genes in the SMCR and the mouse syntenic region. Comparison of the order of the putative genes in the SMCR and the mouse syntenic region. The open boxes represent the flanking sequences of the SMCR: COPS3 and the middle SMS-REP. Genes are in black, and predicted genes are in blue. Genes in the SMCR are connected with their mouse homologs by lines. The region that contains two mouse gene clusters that are not inside the SMCR is boxed. The location of the SMCR in the common deletion region is indicated at the top. Differences in the gene content between the human SMCR and the mouse syntenic region were also observed. RPL13 and 10 predicted genes within the human SMCR are not present within the mouse syntenic region. In the SMCR no genes were identified in the similar20-kb genomic sequence between LLGL1 and FLJ20308 ; as defined by RepeatMasker, this region consists of 56% repetitive sequences. In contrast, in the mouse, an similar110-kb genomic region separates Flj20308 and Llglh, two genes, Ubl and Pabplp, were predicted within this interval . Genes Ubl and Pabplp are homologous to ubiquitin and poly(A)-binding protein, respectively. The human ubiquitin gene subfamily consists of primarily processed pseudogenes . The poly(A)-binding protein is a conserved protein that binds to the 3' poly(A) tail on mRNAs in eukaryotic cells . Both genes are present as intermixed gene repeating units with seven copies for Ubl and four copies for Pabplp (Fig. B). Because three copies of each Ubl and Pabplp could be translated without a premature stop codon, we placed both genes in Group II. In addition, another predicted mouse gene, 1700013G23Rik, is also not located within the human SMCR . Genes in the SMCR (Group I) | We identified 20 genes within the human SMCR . Of the 12 genes mapped in 17p11.2, the genomic structures have been described previously for the following 9: NT5M, PEMT, SREBF1, RAI1, DRG2, MYO15A, FLII, TOP3A, and SHMT1. Here we describe the genomic structures of the other 3 genes: LLGL1, TOM1L2, and ATP12. In addition, two known genes: RASD1 and RPL13, and six unknown genes: FLJ10193, DKFZp586M1120, MGC3048, FLJ20308, SMCR7, and SMCR8, are newly mapped to this region . Figure 6 | Genomic structure of the genes in the SMCR (A) and in the mouse syntenic region (B). Genomic structure of the genes in the SMCR (A) and in the mouse syntenic region (B). Exons of DKFZp586M1120 are represented by gray boxes; exons of other genes are represented by black boxes. Alternative splicing is observed for RAI1, ATP12, and SMCR7. Dashed lines connecting exons represent the alternative splicing. Exon 10 of DKFZp586M1120 is located inside exon 6 of the ATP12 splice variant. Arrows indicate the orientation of transcription. The introns 2 and 3 of KIAA1820 are not to scale. Previously, only seven genes in mice: Pemt, Srebp1, Drg2, Myo15, Llglh, Fliih, and Top3a, were mapped within the SMS syntenic region. Of these, the genomic structure of Drg2 is not known. We describe the gene structure of Rasd1, Rai1, and Drg2, and assemble six unknown mouse genes homologous to human NT5M, FLJ10193, TOM1L2, DKFZp586M1120, MGC3048, and FLJ20308 . FLJ10193 | The human and mouse FLJ10193 have the same two-exon gene structure , but the proteins that they encode have only 64% homology, which is the lowest homology among all genes identified in the SMCR. FLJ10193 is expressed ubiquitously in human and mouse tissues and has a predicted proline-rich region (Fig. ; Tables and ). Figure 7 | Northern blotting of the genes in the SMCR (A) and the genes in its mouse syntenic region (B). Northern blotting of the genes in the SMCR (A) and the genes in its mouse syntenic region (B). Tissues are indicated at the top of each lane, and the size markers in kilobases (kb) are the numbers beside the blots. (He) heart; (Br) brain; (Pl) placenta; (Lu) lung; (Li) liver; (Sk) skeletal muscle; (Ki) kidney; (Pa) pancreas; (Te) testis. For mRNA quantity, beta-actin was probed as a control. Ubiquitous expression was observed for FLJ10193, MGC3048, FLJ20308, and SMCR7. Tissue-specific various-sized transcripts were observed for TOM1L2 and SMCR8. DKFZp586M1120 is predominantly expressed in the kidney, whereas 4930449E07Rik is in the testis. Table 4 | RT-PCR Analysis of the Genes within the SMCR Table 5 | RT-PCR Analysis of the Genes within the Mouse Syntenic Region of the SMCR RASD1 | Mouse Rasd1 is a member of the RAS superfamily, which was induced rapidly by dexamethasone in AtT-20 cells . RASD1 and Rasd1 have the same genomic structure of two exons separated by a <1-kb intron . RAI1 | The genomic structure of RAI1 was described recently . We found that there are two additional transcripts: KIAA1820 and DKFZp434A139 in an similar130-kb region, which share part of their coding regions with RAI1 (; Fig. A). A polyglutamine was observed in RAI1 and KIAA1820 but not in DKFZp434A139. Mouse Rai1 contains two exons corresponding to the exons 2 and 3 of transcript KIAA1820 (Fig. B). Mouse transcript , the homolog of DKFZp586M1120, was also identified (Fig. B). TOM1L2 | We assembled TOM1L2 and Tom1l2, the human and mouse homologs of chicken Tom1B, respectively, by multiple EST alignments. Human TOM1L2 shares a similar genomic structure with its mouse homolog . VHS and GAT domains, which are involved in vesicular trafficking , are present in both the human and mouse hypothetical proteins. A major 6.0-kb and a minor 2.4-kb transcript were identified in all human tissues examined, with a significantly higher level in the heart and skeletal muscle (Fig. A). However, when mouse tissues were analyzed, seven transcripts were observed, and the alternative splicing patterns displayed tissue-specific variations (Fig. B). ATP12 | ATP12 is a human homolog of a yeast nuclear gene required for the assembly of the mitochondrial F1-ATPase . Alternative splicing is observed for exon 3, exon 6, and the last exon (Fig. A). The 3'-UTR of ATP12 overlaps with exon 10 of DKFZp586M1120, which encodes the 3'-UTR (Fig. A). However, the corresponding mouse genes are not overlapping. DKFZp586M1120 | DKFZp586M1120 encodes a putative 225-amino-acid protein , in which five leucine-rich repeats (LRR) were identified. LRR-containing proteins often are involved in protein --protein interactions and cellular adhesion . The putative protein shares 61% homology with PPP1R7, the human homolog of yeast sds22, a mitotic regulator of protein phosphatase-1. A 1.8-kb transcript was observed predominantly in human kidneys and mouse testes . MGC3048 | A putative 217-amino-acid protein encoded by human MGC3048 has the same size as that encoded by its mouse homolog. The two proteins share significant homology with 93% identity. PipMaker analysis revealed conservation of sequence not only in all exons but also throughout the first intron, which may potentially contain regulatory elements for MGC3048 (data not shown). Ubiquitous expression was observed in both human and mouse tissues . FLJ20308 | FLJ20308 contains 4 exons encoding a putative 378-amino-acid transmembrane protein . This gene is well conserved in the mouse with conservation in all exons, introns, and the promoter region (data not shown). Northern blotting identified two transcripts in both human and mouse tissues . The expression is observed in all tissues, with relatively higher expression in heart and skeletal muscle. LLGL1 | Lethal(2) giant larvae is a Drosophila tumor-suppressor gene. Both human and mouse homologs have been described . The overlapping of the 3' end of LLGL1 with the 3' end of FLII is similar to that observed for ATP12 and DKFZp586M1120, but in contrast the LLGL1 and FLII overlap is found in both humans and mice . LLGL1 spans 14.5 kb with 22 exons and is transcribed toward the centromere (Fig. A). SMCR7 | SMCR7 was represented by a cluster of more than 20 ESTs, and two splicing variants with different 5'-UTRs were identified (Fig. A). The putative 353-amino-acid transmembrane protein shows no homology to proteins with defined functions. PipMaker analysis revealed sequence conservation throughout the coding region, indicating the presence of the mouse homolog in the syntenic region. We have not obtained the full-length mouse gene, but several mouse ESTs have been identified. Two transcripts were detected by Northern blotting on human tissues with a similar expression pattern, a major 2.8-kb transcript and a minor 3.4-kb transcript, confirming alternative splicing for SMCR7 (Fig. A). Both transcripts were expressed ubiquitously with a relatively higher expression level in heart and skeletal muscle. SMCR8 | A novel gene, SMCR8, was identified between SHMT1 and TOP3A, with its putative promoter only 286 bp proximal to the promoter of TOP3A. SMCR8 contains two exons encoding a putative 787-amino-acid protein that encompasses an N-terminal domain of the LBP/BPI/CETP family involved in lipid binding (Fig. ; ). More than 75% homology between human and mouse sequences in its coding region was identified. SMCR8 is expressed in all tissues examined (Tables and ), and several transcripts were observed in both human and mouse tissues . Predicted Genes in the SMCR (Group II) | Ten predicted genes were identified within the human SMCR . The existence of these genes is indicated by the presence of multiple ESTs, poly(A), or intron structure, as well as expression in RT-PCR analysis. Each of the predicted genes has no homologs identified in the syntenic region of the mouse. Seven genes are located inside introns of other genes. This potentially explains why these genes have not been identified by gene prediction programs . RT-PCR showed that each predicted gene is expressed in almost all adult and fetal tissues examined except SMCR2, which was expressed in the brain, fetal brain, spinal cord, and trachea (Table ; data not shown). SMCR9 contains two putative PDZ domains, a protein --protein interaction domain likely involved in protein clustering and scaffolding . Pseudogenes (Group III) | Three pseudogenes were identified within the SMCR, including EVPL and two adjacent ribosome protein genes, RPL17 and RPL7A. EVPL is located adjacent to the middle SMS-REP and distal to SHMT1, a region with no conservation to the mouse syntenic region as shown by PipMaker. Also, three pseudogenes, Rps4, Gapd, and Protein tyrosine phosphatase IF2, were identified in the mouse syntenic region. DISCUSSION : Chromosome 17p11.2 is an unstable genomic region harboring several low-copy repeats associated with genomic disorders . Our comparative analysis between the genomic sequence of the SMS region and the syntenic region on mouse chromosome 11 revealed information regarding genomic architecture. Moreover, we refined the critical region responsible for SMS to a minimum interval. New genes were identified within this refined SMCR as well as its mouse syntenic region. Physical and Genetic Map Comparisons | We previously hypothesized that the reduced male recombination frequency at the CMT1A locus may increase the unequal crossing over. This was proposed to result from an extended region of allelic chromosomes without synapse formation to provide an anchor and prevent chromosomal slipping. Indeed, the SMS region shows reduced recombination in both sexes, as might be expected because there is no parent-of-origin preference for the de novo deletion. In the SMS case, the reduced recombination may also reflect interference owing to proximity to the centromere. Nevertheless, it will be interesting to determine if reduced recombination is a general feature of genomic regions that undergo nonallelic homologous recombination. Genomic Architecture Revealed through Comparative Genomics | The percentage of low-copy repeats (LCRs), also termed segmental duplications, in the human genome is greater than in other sequenced genomes, such as the fly and worm . Three low-copy repeats SMS-REPs on human chromosome 17p11.2 mediate DNA rearrangements (deletions and duplications) associated with genomic disorders, by nonallelic homologous recombination . Sequence analyses show that all three SMS-REPs within the SMS common deletion are not present in the mouse syntenic region . Apparently, except for a chromosome inversion of the region between the middle and proximal SMS-REP syntenic region in mouse, the gene order between SMS-REPs is conserved . Interestingly, transposition occurred for the TACI and KCNJ12 genes adjacent to the SMS-REPs. This rearrangement of gene order might have occurred during the evolution of the SMS-REPs, indicating that segmental duplications might transpose surrounding genes. Another genome architectural feature revealed by our human/mouse comparative genome analysis is a region containing two intermixed gene repeat units (Fig. B). Four of the Ubl copies retain a putative intron. The four Pabplp copies are interspersed among the Ubl copies. How this complex array evolved is not immediately obvious, but this entire segment is absent in the human genome. Gene Density and Evolution | The gene density of the human SMCR is higher than the estimated average in the human genome. Within an similar1.1-Mb interval, 19 genes are also present in the same order and orientation in the mouse chromosome . Highly conserved orthologous regions can contain a high gene density. The number, order, and orientation of all 17 genes in a gene-rich cluster at human 12p13 are conserved between humans and mice . Conservation between these two species was also shown for the distal 700 kb of the Cat eye syndrome (tetrasomy of 22q11.2) critical region and for human chromosome 11p15.3, both of which are gene-rich regions . According to the mosaic model of genomic evolution, different portions of the genome evolve at different rates . Thus, gene-rich regions in the genome might have evolved much more slowly than the gene-poor regions. None of the 10 predicted genes in the SMCR was found in mice. Most of these predicted genes match multiple homologous ESTs. Their existence was further confirmed by their expression in RT-PCR analysis. These genes might have important functions for silencing gene expression as an antisense RNA . However, the fact that seven predicted genes are located inside introns of other genes indicates that these predicted genes could represent differentially spliced exons or untranslated exons. Therefore, the absence of these predicted genes in mice could be due to less alternative splicing. In humans, an alternative-spliced last exon of ATP12 overlaps DKFZp586M1120; however, no alternative splicing was observed for mouse Atp12, and the two corresponding mouse homologs are separated from each other. A Newly Defined SMS Critical Interval | Analyses of the deleted intervals of the SMS patients with smaller-sized deletions enabled refinement of the SMS critical region . Six SMS patients ---540, 1190, 1456, 1354, 1615, and 1774 ---were determined to harbor unusual deletions in 17p11.2. The same deletions were identified for patients 1190 and 1456; but some SMS features such as brachycephaly; short, broad hands; and self-destructive and self-hugging behavior are present in 1190 but not in 1456. Slightly different phenotypes were also observed for patients 1615 and 1774 with the same deletions. Thus, genetic background and/or stochastic factors may contribute to penetrance of the phenotype. However, because the major SMS features including craniofacial, skeletal, and neuro/behavioral phenotypes were present in two or more patients , we determined that the smallest overlap among these patients is the SMCR and that the genes that contribute to the craniofacial, skeletal, and neuro/behavioral features are likely located within this genomic region. Two patients, 357 and 765, with an unusual deletion in 17p11.2 have atypical clinical features. The deletions for both 357 and 765 include part of the region between the distal SMS-REP and the middle SMS-REP, with the proximal breakpoints located inside the SMCR. Further study of the breakpoint of patient 765 using a somatic hybrid cell line indicated that it is located inside RP11-258F1, similar210 kb away from the middle SMS-REP. Because patient 765 does not manifest SMS, it is possible that genes responsible for SMS are located within this similar210 kb . Candidate Genes for the SMS Phenotype | To delineate genes that contribute to the SMS features, we identified and characterized genes within the SMCR. SMS patients manifest abnormalities in multiple tissues/organs including: neural (100% mental retardation, 100% self-hugging, 75% peripheral neuropathy, and 69% sleep disorder), eyes (68% iris abnormalities), ears (81% hearing impairment), hearts (29%), kidneys (28%), and skeletal (93% midface hypoplasia and 85% brachydactyly) . The SMS-causing genes are therefore likely expressed in multiple tissues. With the exception of PEMT and MYO15A, all genes mapped within the SMCR are widely expressed in multiple tissues. The genes that are responsible for the SMS phenotype are located within the SMCR and are probably dosage-sensitive. One way to evaluate whether a gene manifests haploinsufficiency effects is by the evaluation of animal models, for example, mice. Until now, targeted disruption of five mouse genes: Pemt, Srebp1, Myo15, Top3a, and Fliih, within SMCR has been reported. The normal heterozygous mutant mice indicated that these genes are not haploinsufficient (; ; ; ; H. Campbell, pers. comm.) and therefore are less likely to contribute to the SMS phenotypes. However, dosage sensitivity for a given gene in humans does not necessarily correlate with haploinsufficiency in mice. For example, haploinsufficiency of GATA3 is the underlying mechanism for the hypoparathyroidism, sensorineural deafness, and renal anomalies (HDR) syndrome. Mice heterozygous for the Gata3 mutation appear to be normal . We can also exclude MYO15A because the heterozygous individual carrying this recessive deafness gene does not show SMS . One potential SMS candidate gene is LLGL1, the human homolog of lethal giant larvae ( Lgl), a tumor-suppressor gene in Drosophila. LLGL1 is expressed ubiquitously, with the most abundant expression in the brain and testis . Moreover, Lgl in Drosophila is essential for asymmetric cortical localization of basal determinants in mitotic neuroblasts, and is required for neural fate decisions . Another SMS candidate is RAI1. Mouse Rai1 has neuron-specific expression in the brain and is induced by retinoic acid, which is involved in craniofacial development . Therefore, haploinsufficiency of RAI1 may account for both neuro/behavioral abnormalities and facial abnormalities in SMS. Two GTP-binding proteins were identified within the SMCR: DRG2 and RASD1. DRG2 is a developmentally regulated protein, and is closely related to DRG1 . RASD1 is a brain-enriched G protein that is strongly and rapidly induced during treatment with dexamethasone . Both the LRR in DKFZp586M1120 and the PDZ domain in SMCR9 are associated with protein --protein interaction. Genes involved in signal transduction or protein --protein interactions have the potential to show haploinsufficiency effects, because half dosage may affect the balance of protein complexes, and then interfere with a series of related events required for normal development. The three genes FLJ20308, SMCR7, and SMCR8 located within the similar210-kb region also remain promising SMS candidates, although we have no clues as to their functions. Targeted mutation of these genes will help unravel their potential roles in SMS. One or More SMS-Causing Genes? | It is unclear whether haploinsufficiency of either one single gene or several contiguous genes causes the characteristic features of SMS as classically proposed . The causative genes for several microdeletion syndromes have been identified. In some syndromes, haploinsufficiency of a single gene is responsible for the entire phenotype. The Rubinstein-Taybi syndrome (RTS), associated with microdeletion of 16p13.3, consists of facial abnormalities, broad thumbs, broad big toes, and mental retardation. Loss of one functional copy of a transcriptional coactivator cyclic AMP response-element-binding protein, (CREB)-binding protein (CBP), underlies all the developmental abnormalities in RTS . This was clearly documented by identifying patients with CBP null alleles caused by frameshift or nonsense mutations. Another example of a monogenic microdeletion syndrome is Alagille syndrome, sometimes associated with the deletion of 20p12. The causative gene for all the features of this syndrome is JAG1, which encodes a ligand for Notch1 . Because deletions of genomic regions inside 17p11.2 are identified in almost all SMS patients, it is possible that haploinsufficiency of more than one gene underlies the SMS phenotype. In some contiguous gene-deletion syndromes, mutation in one gene accounts for only a portion of the phenotype . For example, deletion of the elastin gene, ELN, leads to vascular stenoses, one of the clinical features of Williams syndrome (WS), a developmental disorder caused by a deletion of band 7q11.23 . The specific gene(s) for other features of WS including growth retardation, hypercalcemia, renal anomalies, and mental retardation are still unknown . Miller-Dieker syndrome (MDS) is a multiple malformation syndrome characterized by lissencephaly and facial abnormalities . LIS1, the gene inside the MDS critical region, encodes a subunit of the brain platelet-activating factor acetylhydrolase (PAFAH), and is responsible only for brain malformation . The additional features such as facial abnormalities are thought to be caused by other unknown genes. Therefore, to identify the genes involved in manifesting phenotypes of a contiguous gene deletion syndrome, an important step is to determine the finished sequence of the genomic region corresponding to the deleted interval, and identify all the genes within that region. Conclusions | We have refined the critical SMS interval to an similar1.1-Mb genomic region and performed gene identification and characterization for this newly defined critical region and its syntenic region in the mouse. Our data provide insights into genome architecture and evolution, and new genomic information for comparative analysis between humans and mice, indicate potential SMS candidate genes, facilitate the identification of the haploinsufficient genes involved in this syndrome, and provide information necessary for engineering a mouse model of SMS. METHODS : Construction of BAC/PAC Contig | CTD and PAC clones were selected from the NCBI database based on their sequence homology with the STS markers mapped inside the SMS common deletion region. The BAC clones were isolated by radioactive filter hybridization of human RPCI-11 and mouse RPCI-23 BAC libraries (BACPAC Resources) using probes to STS markers. BAC end sequences were determined either from the database or by direct sequencing. The extent of the genomic clones was determined by STS content mapping. Gaps were filled by multipoint walking based on end sequences of several selected BACs. Up to 20 oligonucleotide overgo probes were included in one hybridization, and positive clones were identified by PCR using primers derived from the parental clone ends. The location of each BAC in the SMCR was further confirmed by FISH on cells from SMS patient lymphoblasts. Human BACs for the minimum tiling paths were submitted to the Whitehead Institute for Biomedical Research at the Massachusetts Institute of Technology. Mouse BACs were submitted to the Human Genome Sequencing Center at Baylor College of Medicine or the Sanger Center for sequencing. Refining the SMCR | The SMCR was refined by FISH on lymphoblast cell lines derived from SMS patients with unusual smaller deletions. BAC/PAC DNA was prepared using the PSI Clone BAC DNA kit (Princeton Separations, Inc.) or the Plasmid Midi Kit (QIAGEN) according to manufacturers' instructions. The DNA probe (1 mug) was labeled by nick-translation using biotin (Life Technologies --GIBCO BRL) or digoxigenin (Boehringer Mannheim). Biotin was detected with FITC-avidin DCS (Vector labs), and digoxigenin was detected with rhodamine-anti-digoxigenin antibodies (Sigma). Chromosomes were counterstained with DAPI diluted in Vectashield antifade (Vector Labs). Cells were viewed under a Zeiss Axioskop fluorescence microscope. Informatics | Sequencher 3.1 software (Gene Codes) was used for sequence alignment, DNA translation, and annotation. Human and mouse interspersed repeat sequences were detected and masked using RepeatMasker . DNA sequences were separated into similar50-kb segments and analyzed using the NIX analysis ; an integrated Web-based multiple DNA analysis bioinformatics tool including GRAIL, Fex, Hexon, MZEF, Genemark, GeneFinder, FGENES, Polyah, RepeatMasker, tRNAscan, and BLAST, that searches many databases. Potential genes were further analyzed individually using FGENES for gene structure and ORF Finder for translation and ORFs. The putative proteins were analyzed using Pfam 6.6 , InterPRO for domains, and TMpred for transmembrane regions. DNA and protein sequence similarity was analyzed with BLAST against the nr, EST, and htgs databases using the default parameters. Human noncontinuous DNA sequences proximal to COPS3 and distal to the middle SMS-REP, corresponding to the SMCR, were repeat-masked and compared with mouse noncontinuous DNA sequences of five BACs (a gap between RP23-82E8 and RP23-181C17 was filled using Celera's genome sequences) using the PipMaker program .PipMaker computes alignments of similar regions in two DNA sequences, and the resulting alignments are summarized with a percent identity plot (PIP; ). RT-PCR Analyses and Northern Blotting | Gene expression profiles in human and mouse tissues were analyzed by Northern blotting and/or RT-PCR (Tables and ). RT-PCR analyses were performed using the first-strand cDNA from various adult and fetal tissues (Clontech). Primers were designed using Primer3 . Hotstart DNA polymerase (QIAGEN) was used to reduce the amplification of nonspecific PCR products; PCR conditions consisted of 95C for 15 min, 1 cycle; 95C for 30 sec, 60C for 30 sec, 72C for 1 min, 32 cycles; and a final extension cycle at 72C. There was no amplification of genomic untranscribed sequences consistent with the absence of genomic DNA contamination in both human and mouse cDNAs. For Northern blotting, probes were designed to DNA sequences of the 3'-UTR of each gene. Radioactive hybridization was performed on multiple tissue blots following the manufacturer's instructions (Clontech). WEB SITE REFERENCES : ; PipMaker to compare two or more noncontinuous DNA sequences. ; RepeatMasker to detect and mask human and mouse interspersed repeat sequences. ; FGENES for gene prediction. ; Pfam 6.6 to analyze proteins. ; Celera private database. ; TMpred for transmembrane regions. ; InterPRO for domains. ; Primer3 to design primers. ; NIX for DNA sequence analysis. ; NCBI public database. ; BAC ends database. Backmatter: PMID- 11997339 TI - Structure and Evolution of the Smith-Magenis Syndrome Repeat Gene Clusters, SMS-REPs AB - An similar4-Mb genomic segment on chromosome 17p11.2, commonly deleted in patients with the Smith-Magenis syndrome (SMS) and duplicated in patients with dup(17)(p11.2p11.2) syndrome, is flanked by large, complex low-copy repeats (LCRs), termed proximal and distal SMS-REP. A third copy, the middle SMS-REP, is located between them. SMS-REPs are believed to mediate nonallelic homologous recombination, resulting in both SMS deletions and reciprocal duplications. To delineate the genomic structure and evolutionary origin of SMS-REPs, we constructed a bacterial artificial chromosome/P1 artifical chromosome contig spanning the entire SMS region, including the SMS-REPs, determined its genomic sequence, and used fluorescence in situ hybridization to study the evolution of SMS-REP in several primate species. Our analysis shows that both the proximal SMS-REP (similar256 kb) and the distal copy (similar176 kb) are located in the same orientation and derived from a progenitor copy, whereas the middle SMS-REP (similar241 kb) is inverted and appears to have been derived from the proximal copy. The SMS-REP LCRs are highly homologous (>98%) and contain at least 14 genes/pseudogenes each. SMS-REPs are not present in mice and were duplicated after the divergence of New World monkeys from pre-monkeys similar40 --65 million years ago. Our findings potentially explain why the vast majority of SMS deletions and dup(17)(p11.2p11.2) occur at proximal and distal SMS-REPs and further support previous observations that higher-order genomic architecture involving LCRs arose recently during primate speciation and may predispose the human genome to both meiotic and mitotic rearrangements. Keywords: Introduction : Smith-Magenis syndrome (SMS) is caused in >90% of cases by a common deletion of an similar4-Mb gene-rich genomic segment in 17p11.2. Recently, the same chromosome region has been shown to be duplicated in patients with the reciprocal chromosome duplication, dup(17)(p11.2p11.2) . Physical mapping studies have shown that the SMS-common-deletion interval is flanked by large (similar200 kb), highly homologous, low-copy repeat (LCR) gene clusters termed SMS-REPs . The proximal and distal SMS-REPs likely act as substrates for nonallelic homologous recombination (NAHR; also known as unequal crossing-over), resulting in both common deletions and reciprocal duplications of the same chromosome segment. A third copy, the middle SMS-REP, has been mapped in 17p11.2 between the proximal and distal SMS-REPs . The chromosome 17p11.2 genomic region is involved in several other rearrangements. Isodicentric chromosomes idic(17)(p11), with breakpoints mapping within or just adjacent to the SMS critical region, have been identified in various hematological malignancies, including chronic myeloid leukemia, and in solid tumors, such as childhood primitive neuroectodermal tumors . Chromosome amplifications within 17p11.2 have been described in patients with osteosarcoma and leiomyosarcoma . In medulloblastomas, an aberrant hypermethylation of the major breakpoint cluster region in 17p11.2 has been proposed to be an additional genomic feature responsible for the chromosomal fragility . These data, together with identified somatic mosaicism for SMS deletions , indicate that the presence of unique genome architecture features, including highly homologous SMS-REPs, makes the chromosome 17p11.2 a highly unstable region in the human genome, prone to both meiotic and mitotic rearrangements. To delineate the genomic structure and evolution of the SMS-REPs, we constructed and sequenced a complete bacterial artificial chromosome/P1 artifical chromosome (BAC/PAC) contig. Based on genomic sequence analysis, we elucidated the size and the orientation of each SMS-REP, the extent of homology among the SMS-REPs, genes within the SMS-REPs, and here provide a model for the evolution of SMS-REPs. Sequence-based structural analysis of the SMS-REPs is essential to determine the molecular mechanism of chromosome rearrangements in SMS, as well as other genomic disorders . These studies add to a growing body of evidence that implicate genome architecture in DNA rearrangements responsible for genomic disorders . RESULTS : Aligning BAC Genomic Clones to Specific SMS-REPs by cis-Morphism Analysis and DNA Fingerprinting | The identification of genomic clones specific to an individual SMS-REP was possible due to a limited number of sequence differences among them, which we called cis-morphisms. Cis-morphisms refer to sequence variation among nonallelic, highly similar sequence copies on the same chromosome. This term has been proposed to distinguish from polymorphisms resulting from variation between sequences on different chromosome homologs (alleles; ). We identified SMS-REP-specific cis-morphisms using Southern blotting with genomic DNA and BAC clones as targets and compared the hybridization band patterns with those of established SMS-REP-specific yeast artificial chromosome (YAC) clones (912D7 and 567A2 for distal SMS-REP; 907E8 and 126H9 for proximal SMS-REP; 951G11 for middle SMS-REP) that were anchored with flanking unique sequences . We documented previously that SMS-REPs contain at least four genes or pseudogenes (CLP, TRE, SRP, and KER, a keratin gene cluster) by sequence skimming of SMS-REP-specific cosmid clones . By utilizing probes from these genes in Southern analysis of genomic DNA and YAC clones anchored by flanking unique sequences, we have detected cis-morphisms that serve as unique identifiers to each specific SMS-REP. Multiple restriction enzyme digestions of large-insert genomic clones were utilized and enabled several cis-morphisms to be identified (Fig. ; Table ). Using CLP, TRE, and SRP sequences as probes on a BAC genomic library we identified >50 SMS-REP-like BAC clones. The CLP probe was a 1.1-kb HindIII fragment of the cDNA clone 41G7A , and TRE and SRP probes were amplified by 3' TRE and SRP primers, respectively. Cross-hybridizing fragments were then analyzed for SMS-REP specific cis-morphisms to uniquely identify to which SMS-REP (proximal, middle, or distal) each BAC clone mapped. Independently, all SMS-REP cross-hybridizing BAC clones were subjected to DNA-fingerprinting analysis. The restriction patterns of the majority of these clones were highly similar, although minor differences allowed the BACs to be grouped into contigs. Fingerprint analysis was combined with hybridization analyses to select minimal clone sets for genomic sequencing. Figure 1 | An example of a cis-morphism among three SMS-REPs. An example of a cis-morphism among three SMS-REPs. Southern blotting with CLP cDNA probe on HindIII-digested large-insert genomic clones revealed an 8.1-kb band for the distal SMS-REP (lanes 3 and 4 for yeast artificial chromosomes [YACs]; lanes 7 and 8 for bacterial artificial chromosomes [BACs]) and 5.1 kb and 0.9 kb for the proximal SMS-REP (lanes 5 and 6 for YACs; lanes 9 and 10 for BACs). Hybridization to total genomic DNA (gDNA in lanes 1 and 2) reveals additional cross-hybridizing bands, which map to different genomic locations. Table 1 | Cis-Morphisms Between SMS-REPs and the Related Structures. (Sizes in kb) Generation of the BAC Contigs Spanning the SMS-REPs | We collected BAC clones identified by hybridization screening of the RPCI-11 library and by computational methods using BLAST searches. Fifty-eight clones were assigned to individual SMS-REPs by cis-morphisms and DNA fingerprinting, and we constructed BAC contigs spanning the three individual SMS-REPs . A minimal-tiling path of assigned BAC genomic clones, selected using the combined results of hybridization and fingerprinting analyses, was chosen for sequence analysis. Figure 2 | Bacterial artificial chromosome/P1 artificial chromosome (BAC/PAC) contig map spanning three SMS-REPs. Bacterial artificial chromosome/P1 artificial chromosome (BAC/PAC) contig map spanning three SMS-REPs. Thick bold lines represent minimal-tiling-path, large-insert clones utilized for genomic finished sequence of SMS-REPs. Clones are designated with their clone name and their GenBank accession number. Markers in boxes represent SMS-REP-flanking sequences used to determine SMS-REP orientations. Final orientation was determined by the construction of a complete BAC/PAC contig spanning the common deletion . The mapping of the genomic BAC clones to the SMS-REP region in 17p11.2 was further confirmed by fluorescence in situ hybridization (FISH). Metaphase FISH analysis using the SMS-REP BAC clones displayed hybridization to the 17p11.2 region, whereas interphase FISH analysis showed three hybridization signals for each chromosome (one per chromatid for each of the three SMS-REP copies in this G2 interphase nucleus; Fig. ). Although FISH analysis can identify SMS-REP cross-hybridizing genomic sequences and distinguish three copies in 17p11.2, it cannot determine from which specific individual SMS-REP the signal arises. Therefore, cell lines from patients with selected 17p11.2 deletions were examined by FISH to further confirm BAC localization to the SMS-common-deletion region. Figure 3 | Human metaphase chromosome 17 and G2 interphase nucleus after fluorescence in situ hybridization (FISH) with SMS-REP-specific bacterial artificial chromosome (BAC) RP11 --158M20. Human metaphase chromosome 17 and G2 interphase nucleus after fluorescence in situ hybridization (FISH) with SMS-REP-specific bacterial artificial chromosome (BAC) RP11 --158M20. To the left is a human-chromosome-17 ideogram with the positions of hybridization signals (green) shown next to specific cytological bands. White horizontal arrows show specific locations of FISH signals with BACs containing the genomic segments listed above the arrows. To the right of the figure is a G2 interphase FISH analysis (the G2 phase of the nucleus was determined without an internal control probe). Note the three copies of SMS-REP (arrows in interphase nucleus) and SMS-REP-like sequences. Analysis of the Structures and the Sequences of Three SMS-REPs in 17p11.2 | Sequencing of the genomic clones in the minimal-tiling paths was performed at the Whitehead Institute/MIT Center for Genome Research. Sequences of two additional PAC clones, RP1 --48J14/AL353996 and RP1 --37N07/AL353997, were downloaded from the NCBI web site . We thus analyzed genomic sequence spanning each individual SMS-REP. Based on the sequence information, a structural map of the SMS-REPs was constructed and analyzed for size and orientation of each SMS-REP, the extent of homology among the SMS-REPs, and gene content within the SMS-REPs. Figure 4 | Sequence-based genomic structure of the SMS-REPs. Sequence-based genomic structure of the SMS-REPs. There are four regions of sequence identity >95% between the proximal and the distal SMS-REPs (A, B, C, and D). The A (red), B (black), and C (yellow) sequence blocks have >98% identity between distal and proximal REPs; the D regions (green) show >95% identity. Blue represents the regions of homology between proximal and middle SMS-REPs. The proximal copy is the largest and is localized in the same orientations as the distal copy. The middle SMS-REP shows almost the same sequence and structure as the proximal copy except for two terminal deletions, an UPF3A gene interstitial deletion and a small (similar2 kb) insertional duplication. However, it is inverted with respect to proximal and distal SMS-REPs. SMS-REP-specific CLP, TRE, and SRP cis-morphisms were confirmed by DNA sequencing. Fourteen genes/pseudogenes were found and are summarized in Table . The two additional KER copies in distal SMS-REP represent repeated fragments of the KER pseudogenes, the accession numbers of which are given in Table . Crosshatched areas (NOS2A in the proximal and KER in the distal) denote two genes spanning the high homology and nonhomology area between the distal and proximal, which suggest a three-step event for the hypothetical model of the evolution of the SMS-REPs (see text). At the bottom, the chromosome 17 distribution of fragments of SMS-REP, which constitutes a chromosome 17 low-copy repeat, LCR17, is shown. The above data were obtained through BLAST analysis of sequence database. The proximal SMS-REP is the largest, similar256 kb (probably 255,808 bp; np 50,751 in RP11 --121A13/AC008088 and np 4,010 in RP11 --434D2/AC015818), and contains 14 genes/pseudogenes. Based on flanking markers P836L9 --5' and R193 , the distal --proximal orientation of proximal SMS-REP is from SRP to CLP (Figs. , ). The distal SMS-REP is 176,482 bp (np 10,612 --187,093 in RP11 --219A15/AC022596). Distal SMS-REP-flanking markers 92B11-R and 315G18-R showed that the distal SMS-REP is in the same SRP --CLP orientation. Therefore, proximal and distal SMS-REP represent direct repeats within 17p11.2. The markers SHMT and 484D23-R flanking middle SMS-REP (241 kb; probably 240,806 bp, np 42,370 in RP11 --448D22/AC105100, and np 53,872 in RP1 --37N07) showed that it is inverted with respect to proximal and distal SMS-REPs. There are four regions of high homology between the proximal and the distal SMS-REPs (A, B, C, and D regions in Fig. ). The sum of these high-sequence-similarity regions is similar170 kb (169,905 bp), and the homology is greater than 98%, with the exception of the D region (>95%). The largest conserved segment (region A in Fig. ) is similar126 kb in size. Two large sequence blocks (between A and B, and C and D) in the proximal SMS-REP are absent in the distal SMS-REP. Two smaller blocks, flanking areas of the B region in the distal SMS-REP, are absent in the proximal SMS-REP. At least 14 genes/pseudogenes were identified in SMS-REPs and are summarized in Table and Figure . Two potential genes in the SMS-REPs include KIAA0565 and FLJ11800. The remaining genes appear to represent pseudogenes or pseudogene-like structures (UPF3A, LGALS9, SRP68, TL-132, NOS2A, MIP, KER cluster, USP6 [also known as TRE], CLP, and several expressed sequence tags: BG108241, AIG53154, AW977532, AL138090, AI913336, BF223454, AW469705, AL136790, AA405442). Importantly, sequence analysis confirmed the presence of all four genes/pseudogenes (SRP, KER, TRE, and CLP) previously identified within SMS-REPs . Interestingly, many (7 of 14) of the genes/pseudogenes also map to the long arm of chromosome 17 . Some parts of the gene sequences are repeated (e.g., a part of TL-132 sequence; np 4173 --4326 of AJ012755 is duplicated and inverted). As predicted by previous hybridization studies, no CLP sequence was found in the middle SMS-REP, and the distal SMS-REP is devoid of two genomic fragments containing LGALS9, MIP, and USP6 . In addition, BLAST analyses identified that an similar2.3-kb UPF3A gene, present in proximal and distal SMS-REPs, is absent in the middle SMS-REP. Interestingly, adjacent (similar400 bp) to this middle SMS-REP-specific deletion, a 1838-bp insertional duplication from chromosome 15q26.1 was identified . BLAST analysis revealed that the overall similarity among SMS-REP-homologous segments is 98% identical, indicating that SMS-REPs can likely act as substrates for NAHR, resulting in both deletions and reciprocal duplications of the same chromosome segment. Table 2 | Genes/Pseudogenes Found in the SMS-REPs Fragments of SMS-REPs Are Also Located Outside the SMS Region | DNA-fingerprinting analysis of CLP cross-hybridizing BAC genomic clones indicated several distinct contigs, two of which were mapped to proximal and distal SMS-REPs by cis-morphism analysis. To map other CLP genes/pseudogenes in the human genome, we used the clones from the unique BAC contigs that did not map to proximal or distal SMS-REPs as probes in FISH mapping studies. The majority of these BACs mapped to both chromosome 17p11.2 and 16qter, indicating the presence of a CLP gene on chromosome 16q24. FISH analysis using SMS-REP BAC clones (distal SMS-REP RP11 --219A15/AC022596, middle SMS-REP RP11 --158M20/AC023401, and proximal SMS-REP RP11 --434D2/AC015818) as probes revealed strong hybridization signals on metaphase chromosome 17p11.2 and three strong signals on the interphase chromosomes. However, SMS-REP-specific BACs also showed weaker hybridization signals in interphase analysis and metaphase spreads; these map to chromosomes 17p13.1, 17p12, 17q11.2, 17q12, 17q21.2, and 17q23.2 . Computational analysis also supported the existence of other SMS-REP-like sequences on chromosome 17. In concordance with FISH results, BLAST analysis revealed that similar11 --30-kb fragments of SMS-REPs are localized on 17p13.1 (similar28 kb; RP11 --500A7/AC008058; RP11 --333E1/AC087500; CTD-2013F15/AC074339), 17p12 (similar11 kb; RP11 --640I15/AC005324), 17q11.2 (similar30 kb; RP11 --218M11/AC011840; RP11 --271K11/AC005562), 17q12 (similar11 kb; RP11 --698P18/AC067923; RP1191J4/AC003976), 17q21.2 (similar25 kb; RP11 --156A24/AC019349), and 17q23.2 (similar28 kb; RP11 --767P9/AC023352; RP11 --178C3/AC005702). A >19-kb fragment of average >90% homology to a portion of SMS-REP, including the USP6 (TRE) gene, is located within the 17q11.2 BAC clone RP11 --271K11/AC005562. This BAC clone also contains similar85 kb of the neurofibromatosis type 1 LCR, proximal NF1-REP. In addition, we found that the distal NF1-REP on 17q24 (RP11 --147L13; AC005332; ) maps directly adjacent to the breakpoint of the translocation t(1;17) associated with Russell-Silver syndrome (RSS; ). Complex Genome Architecture in Proximal 17p | Analysis of the DNA sequence within proximal 17p, including the SMS common deletion and CMT1A chromosome regions, revealed the presence of several genomic segments of sequence similarity >=95% to SMS-REP and/or clones that directly flank SMS-REP . Two LCRs, similar90-kb LCR17pC and similar150-kb LCR17pD (also known as, proximal SMS-REP flanking sequences [PSFS]) were found to flank proximal SMS-REP and to be part of at least 410-kb LCR17pA, flanking proximal CMT1A-REP. LCR17pA has also been proposed to be responsible for the evolutionary chromosome translocation 4;19 in gorilla . In addition, an at least 210-kb segment LCR17pB within overlapping BAC clones RP11 --448D22 and CTD-2145A24, proximally adjacent to the middle SMS-REP, was found to have 98% homology to the other portion of LCR17pA . Interestingly, the breakpoints of an unusual chromosome deletion in two SMS patients have been identified to map within these LCR17p LCRs (our unpubl. data). The recognition that well-characterized chromosome 17 genomic duplications (SMS, CMT1A, and NF1) are further interspersed with previously unidentified homologous repeated sequences and that these additional LCRs are distributed elsewhere on this and other chromosomes underlines the overall complexity of the chromosome 17 genomic architecture. Figure 5 | Schematic representation of low-copy repeats (LCRs; >10 kb) in proximal 17p, including SMS and CMT1A chromosome regions. Schematic representation of low-copy repeats (LCRs; >10 kb) in proximal 17p, including SMS and CMT1A chromosome regions. Colored boxes represent positions of highly similar sequence of LCR structures. Note that the at least 410-kb repeat, LCR17pA, flanking the proximal CMT1A-REP has three partial copies: LCR17pB, proximally adjacent to the middle SMS-REP, and LCR17pC and LCR17pD (also described as PSFS; ) flanking the proximal SMS-REP. Evolutionary Studies | Studies of LCRs involved in several genomic disorders suggest a recent evolutionary origin during primate speciation. In fact, the CMT1A-REP LCR in 17p12 is present in human and chimpanzee but not in gorilla, orangutan, and other Old and New World monkeys . We studied SMS-REP evolution during primate speciation by FISH analysis using human SMS-REP-specific BAC clone probes on different primate cell lines. One BAC from each SMS-REP (distal SMS-REP RP11 --219A15/AC022596, middle SMS-REP RP11 --158M20/AC023401, and proximal SMS-REP RP11 --434D2/AC015818) was used in independent FISH experiments. As expected, the same result was obtained in each of the three separate experiments that used the different SMS-REP probes. The characteristic fluorescence signal pattern after FISH with SMS-REP-containing BAC clones observed on human metaphase and interphase chromosomes 17 was visible on syntenic metaphase chromosomes of all great apes and Old World monkeys analyzed. The presence of three SMS-REP copies was confirmed by the analysis of the respective interphase nuclei (Fig. A). However, presumably due to reduced genomic sequence homology and thus lower hybridization efficiency on the metaphase and interphase chromosomes of New World squirrel monkey, the fluorescent signal after FISH with SMS-REP BAC was of significantly weaker intensity. Therefore it was difficult to determine unequivocally whether all three SMS-REP copies are present in this species. Figure 6 | FISH analyses for evolutionary studies. FISH analyses for evolutionary studies. (A) Metaphase and interphase cells of chimpanzee, gorilla, orangutan, gibbon, baboon, and rhesus after hybridization with SMS-REP bacterial artificial chromosome (BAC) RP11 --158M20. (B) The interphase nuclei of New World squirrel monkey after cohybridization of differentially labeled SMS-REP-specific BAC (RP11 --158M20) and a P1 artificial chromosome clone that directly flanks SMS-REP (RP1 --178F10). Note the single yellow foci, indicating the <100-kb physical vicinity of the clones. Similar results were obtained using five other clones directly flanking individual SMS-REPs, showing the presence of three copies of SMS-REPs. To control for potential misinterpretation of absence of FISH signal secondary to decreased hybridization related to primary sequence divergence, we applied dual-color interphase FISH co-hybridizing SMS-REP BACs with differentially labeled human BAC clones that directly flank SMS-REP. In each of six possible combinations (two flanking unique BAC probes for each three SMS-REPs), nuclei from 50 cells were analyzed. We observed a yellow hybridization signal due to the overlapping of SMS-REP signal (green) and flanking clone signal (red; representative co-FISH in Fig. B). These data indicate the presence of all three repeats in squirrel monkey. Therefore, the age of the three SMS-REPs is assessed to be at least similar40 million years ago (Mya) ---the time of divergence of New World monkeys from pre-monkeys . To further delineate the origin of the SMS-REPs, we investigated evolutionarily older species, including pre-monkey lemur. Only weak unique SMS-region-specific sequence signals were visible. Based on the structure of the individual SMS-REPs and evolutionary investigations using primate cell lines, we propose that an SMS-REP progenitor copy was triplicated after the divergence of New World monkeys and pre-monkeys between similar40 --65 Mya. DISCUSSION : Chromosome 17 Low-Copy Repeats | Chromosome 17 contains several LCRs. It harbors two Charcot-Marie-Tooth disease type 1A CMT1A-REPs (17p12; ; ), three SMS-REPs (17p11.2; ), and three neurofibromatosis type 1 NF1-REPs (17q11.2; ). These LCRs can serve as substrates for both intra- or interchromosomal NAHR, thus making the respective genomic region unstable and prone to chromosome rearrangements responsible for genomic diseases (for reviews, see ; ; ; ; ). Ongoing work is identifying and characterizing additional LCRs, although their contribution to genomic disease is currently unknown. The proximal SMS-REP spans similar256 kb and is in the same orientation as the distal SMS-REP, which is shorter (similar176 kb) and devoid of some small repeat fragments. The middle SMS-REP (similar241 kb) is inverted with respect to proximal and distal SMS-REPs. This architecture potentially explains why the common SMS deletions occur between proximal and distal SMS-REPs. The deletions and duplications resulting from NAHR require direct repeats as recombination substrates, whereas inverted repeat recombination substrates may lead to inversions . We have identified recently five SMS patients with smaller-sized deletions who have rearrangements in which their proximal breakpoints map within or adjacent to the middle SMS-REP and the distal breakpoints within or adjacent to the distal SMS-REP (; our unpubl. observations). A candidate genomic structure, which could be potentially responsible for this smaller deletion, is the inverted repeated sequences of 639 bp, including a part of the TL-132 gene (np 4173 --4326 of AJ012755). This segment in the distal SMS-REP has the same orientation as that repeated in the middle copy (np 81890 --82528 and 69851 --70489 in RP11 --219A15/AC022596 vs. np 135928 --136566 and 147947 --148585 in RP11 --158M20/AC023401), thus yielding a direct repeat. We also suggest that one of the parents of these 5 SMS patients may be a carrier of a paracentric inversion of the respective segments within 17p11.2, with one chromosome breakpoint involving middle SMS-REP. Similar chromosome polymorphisms involving large genomic regions mediated by olfactory receptor gene cluster LCR on chromosome 8p have been identified recently in 26% and 27% of populations of European and Japanese descent, respectively , in the parents of patients with Williams-Beuren (WBS) syndrome (33%), and in some atypical WBS patients (27%; ). Evolution of Low-Copy Repeats | Previous hybridization studies suggested the absence of SMS-REP sequences in rodents . Here we provide evidence based on FISH analysis of primate cell lines that SMS-REP genomic duplication occurred after the divergence of New World monkeys and pre-monkey lemurs. Evolutionary studies of several different LCRs have shown that they arose recently, apparently during primate speciation . A high sequence similarity but <100% identity between LCRs appears to be responsible for genomic disorders and suggests that one copy of the LCR is the progenitor and recent genomic duplication gave rise to its additional copies . The two copies of CMT1A-REP have been shown to be absent from rodent species . Both CMT1A-REP copies were found in chimpanzee and human, whereas only the distal copy has been identified in gorilla, indicating that it must have been duplicated after the gorilla had diverged from a common human/chimpanzee ancestor . Sequence analysis of CMT1A-REPs revealed that the distal copy is the original and is part of the COX10 gene. The proximal CMT1A-REP represents a duplicated copy of COX10 exon 6 and surrounding intronic sequences . Recently, it was shown that the CMT1A-REP insertional/duplicational event divided an ancestral gene, AGIP, into two genes: HREP and CDRT1. These findings illustrate the generation of new genes by DNA rearrangement during mammalian genome evolution . Interestingly, sequences flanking proximal CMT1A-REP have been found to be duplicated on gorilla chromosome 19 at the breakpoints of the evolutionary reciprocal translocation t(4;19) in gorilla, equivalent to human t(5;17) . The repeats flanking the Williams syndrome chromosome region were identified in chimpanzee, gorilla, orangutan, and gibbon but were absent in mice, indicating that the duplication event occurred after divergence of rodents and humans (similar80 Mya) but before the diversification of hominoids (similar15 --20 Mya; ). LCR-sequence comparisons show that block B-mid (common telomeric breakpoint) is probably the ancestral copy. Interestingly, repeat-like sequences were found on human chromosome 7p22 and 7q22 very close to the sites where the breakpoints of evolutionary inversions occurred. It was proposed, therefore, that these segmental duplications had been generated by the evolutionary inversions ; however, it is possible these duplicated segments stimulated inversions. The LCR15s, implicated in the pathogenesis of Prader-Willi and Angelman syndromes, were shown to be present in nonhuman primates, including New World monkeys, and only one copy in lemur, suggesting that the LCR duplication may have occurred similar50 Mya . The repeat block LCR15-BP3A was proposed to be an ancestral copy. Analysis of NF1-REPs and LCR22s suggests that their duplication may predate the divergence of the great apes similar8 --9 Mya and New World monkeys similar40 Mya , respectively. A Model for the Evolution of the SMS-REPs | Based on the structural information of three SMS-REPs, we propose a model to explain the evolution of the SMS-REPs. Evidence supporting this model includes the fact that two genes span the border of high- and low-homology areas between the proximal and the distal SMS-REPs, the NOS2A gene in the proximal, and the KER pseudogene (M22927) in the distal SMS-REP (crosshatched area in Fig. ). A long sequence of the NOS2A gene is located across the border of the A region and its flanking area (A --B in Fig. ) in the proximal SMS-REP, and a part of the gene in the A --B region is absent in the distal SMS-REP. Also the KER pseudogene (M22927) spans the C region and its flanking area (B --C in Fig. ) in the distal SMS-REP, and a part of the gene in the B --C region is absent in the proximal SMS-REP. Therefore, only parts of the NOS2A and the KER pseudogene remain in the distal and the proximal SMS-REP, respectively. To explain these phenomena, we adopted the following three-step-event hypothesis . Our data indicate that in the first step, similar40 --65 Mya, a progenitor SMS-REP must have arisen in an ancient chromosome. Its structure was almost the same as the proximal SMS-REP of today but included flanking sequences on each side of the B region, as is the case of the distal SMS-REP. The distal SMS-REP resulted from the deletions of two large areas between the A and B regions and the C and D regions. Secondly, deletion of both flanking sequences of the B region in the progenitor resulted in the proximal SMS-REP. Finally, two terminal deletions (2 kb and similar14 kb) and one interstitial deletion (similar2 kb) associated with interchromosomal insertional duplication (similar2 kb), along with an inversion, generated the middle SMS-REP . Figure 7 | Hypothetical model for evolution of SMS-REPs. Hypothetical model for evolution of SMS-REPs. Our data indicate that in the first step, a progenitor SMS-REP must have arisen in an ancient chromosome. Its structure was almost the same as the proximal SMS-REP at the present time, but included the B region flanking sequences on both sides similar to the distal SMS-REP. The distal SMS-REP resulted from the deletions of two large areas between the A and B, and C and D regions. Secondly, deletion of both flanking sequences of the B region in the progenitor resulted in the proximal SMS-REP. Finally, two terminal deletions and one interstial deletion involving the UPF3A gene accompanied by interchromosomal insertional duplication together with an inversion generated the middle SMS-REP. Based on the structure of the identified LCRs flanking the proximal SMS-REP and proximal CMT1A-REP and adjacent to the middle SMS-REP, we suggest that before the origin of SMS-REPs, an at least 410 kb chromosome segment flanking the proximal CMT1A-REP (LRC17pA) must have undergone partial segmental duplication in an ancestral chromosome syntenic to the proximal HSA 17p, thus further supporting that segmental duplications have accompanied karyotype evolution in primates. Conclusion | Our molecular analysis of the SMS-REP LCRs reveals a complex structure consisting of at least 14 genes/pseudogenes. There is remarkable preservation of sequence homology, >160 kb of similar98% identity, providing a sizable substrate for homologous recombination. Portions of the SMS-REP are present on 17q, constituting a chromosome 17 low-copy repeat (LCR17) similar to what has been identified on chromosome 22 , whereas some copies of its gene constituents appear to map to other chromosomes. Like other LCRs that mediate DNA rearrangements responsible for genomic disorders, SMS-REP appears to have evolved during primate speciation. Further analyses of proximal chromosome 17p suggest even more complex genome architecture. Why and how large homology segments are preserved in the human genome is not immediately obvious. Nevertheless, in-depth analyses of such LCRs is essential for further determining the higher order architecture of the human genome and to predict regions of susceptibility for rearrangements leading to chromosome imbalance. METHODS : PCR | PCR was performed with extracted BAC DNA using each of the following primer sets: CLP (TCTGTAAACTGTCTGAGTGCAGAG/CGTCTGCACC-ACACAATCAAAAGG), 3'TRE (ACAGGTAGCACAATCTACTAA/TTCTGTGTTT-ACTTGTATGAGG), SRP (GGAAGCACTTGCTGTCATCC/GCCCAGGCCAAATGG-CCCTGG), KER (CTCTGCTCTGACCCTCTA/AGCCCTGATCCTTGGGGTCCAG), SHMT (TGGACGCACATTTGTCCTAC/CAGGGACCTGCAGAACTGAC), R193(GGCAGCTCAGGGTGAGCTCTTC/TATTGGCCTTAAATGCATCTCA), RP11 --92B11 reverse end (GGCTGAATGTTTTCCCACAT/AAGGAGATGAAAGGCAGCAA), RP11 --315G18 reverse end (CTCCACCGAAAAGCCTACAG/TGCCCTGGAGTTACA-AGATG), TOP3 (TGGTTGC TGTTAGCAGAGGA/CCTTGCATTACACCGTCCTT), RP11 --484D23 reverse end (AAGTCTCTGGAGCTCTCATTCA/CCCAGGCACACTA-AACCATT), RP5 --836L9 5' end (ACCTCAGAGGCTACCTCACG/CCAAAGACAG CTATCCACTGC), RP11 --121A13 forward end (GGTTGTCTGGGCTTGGTAGA/TG AGTGCCAGCTAAGTGCTC), and D17S959 (TCAGATTGAACTCTCGGTAT/GCTG AC ACAGGCAATG). PCR reactions were performed on a final volume of 25 muL, using 0.7 U of Ampli-Taq polymerase in 200 muM each of dNTPs; 1x GeneAmp PCR buffer (10 mM Tris-HCl at pH8.3 and 25C; 50 mM KCL; 1.5 mM MgCl2; 0.001% [w/v] gelatin) and 10 pmoles of each primer. PCR conditions were 95C for 5 min; 35 cycles of 94C for 30 sec, 55C for 30 sec, and 72C for 1 min, followed by 72C for 7 min. The products were visualized on 2% agarose gels. BAC Library Screening | We screened segments 1 and 2 (11.8x) of RPCI-11 BAC library (BACPAC Resource Center, Children's Hospital Oakland Research Institute; ) using CLP, SRP, and SHMT gene probes. CLP probes were cDNA probe (1.1 kb HindIII fragment of the cDNA clone 41G7A), overgo oligonucleotide CLP1 (CGGCGATGGCCACCAAGA/GAAGCCTCTTTGTCGATCTTGGTG), and overgo CLP2 (CCAAAGCGACAAGATCGT/CGGGCCTCTCTCCCTAACGATCTT); SRP probe was the PCR products amplified by SRP primer sets; SHMT probe was overgo (GAAGACTACACAGGGCCTTCCTCA/AAGCGACAGGCTTAAGTGAGGAAG). Hybridizations were performed according to the previously described methods BAC end sequences of the assigned clones were obtained from TIGR . We also identified several candidate clones from BLAST searches using the BAC end sequences. We checked the extent of the assigned clones by PCR for STSs and BAC ends. BAC Sequencing and Sequence Analysis | BAC and PAC clone sequencing was performed after generation of m13 and/or plasmid subclone libraries. Individual clones were either sequenced to fourfold coverage (draft quality) and then further sequenced to greater than eightfold coverage or sequenced immediately to greater than eightfold coverage. After greater than eightfold coverage and assembly, directed finishing techniques were applied to determine the sequences of the remaining gaps. All clones were sequenced and assembled independent of each other and then analyzed to confirm their placement in the tiling paths and determine their degree of overlap with neighboring clones. Specific sequencing and finishing procedures were performed as summarized . Regional assemblies for each SMS-REP were constructed following BLAST searches against the high-throughput and the nonredundant sequence database and assembled using the Sequencher software (Gene Codes). Genes and transcripts were identified by a database search of the genomic sequences using the BLAST nonredundant database . Fingerprint Analysis | BAC clones identified following library screens with CLP and SMS-REP markers were fingerprinted and analyzed as described by . Purified BAC DNAs were digested with HindIII, electrophoresed on 1% agarose gels, and stained with SYBR Green I (Roche Diagnostics); then gel images were recorded using a FluorImager SI (Molecular Dynamics). Relative mobility files of observed bands were created using Image ; then fingerprint data were stored and analyzed using FPC . Clones were grouped into fingerprint contigs as described by , with every internal clone requiring having all of its fragments shared with overlapping clones. Southern Blotting | Southern hybridization was performed as described by . The target DNAs (human genomic DNA, SMS-REP specific YAC and BAC clones) were cut with various restriction enzymes (HindIII, XbaI, PstI, PvuII, RsaI, TaqI, BglII, XhoI, BamHI, EcoRI KpnI, MspI, and DraI) and hybridized with CLP cDNA probe and the PCR products for TRE and SRP. Fluorescence In Situ Hybridization | FISH was performed on metaphase and interphase cells of transformed peripheral blood lymphoblasts and skin fibroblasts according to a modified procedure of . Briefly, 1 mug of isolated DNA of BAC/PAC clones: RP11 --98L14 (proximally adjacent to the proximal SMS-REP), RP11 --434D2 (proximal SMS-REP), RP5 --836L9 (distally adjacent to the proximal SMS-REP), RP11 --28B23 (proximally to the middle SMS-REP), RP11 --158M20 (middle SMS-REP), RP1 --178F10 (distal to the middle SMS-REP), RP11 --416I2 (proximal to the distal SMS-REP), and RP11 --219A15 (distal SMS-REP), and RP11 --209J20 (distally adjacent to the distal SMS-REP) was labeled by nick-translation reaction using biotin- (Life Technologies-GibcoBRL) or digoxigenin- (Boehringer Mannheim) labeled nucleotides. Biotin was detected with FITC-avidin DCS (fluoresces green; Vector Labs) and digoxigenin was detected with rhodamine-anti-digoxigenin antibodies (fluoresces red; Sigma). Chromosomes were counterstained with DAPI diluted in Vectashield antifade (Vector Labs). Cells were viewed under a Zeiss Axioskop fluorescence microscope equipped with appropriate filter combinations. Monochromatic images were captured and pseudocolored using MacProbe 4.2.2/Power Macintosh G4 system (Perceptive Scientific Instruments, Inc.). Cell Lines | The human and primate lymphoblastoid cell lines and fibroblasts were grown and harvested using standard methods. The nonhuman-primate-immortalized Epstein-Barr virus stimulated cell lines of lowland gorilla (Gorilla gorilla, CRL 1854), orangutan (Pongo pygmaeus), gibbon (Hylobates lar, TIB-201), New World monkey baboon (Papio hamadryas), Rhesus monkey (Macaca mulatta), and New World squirrel monkey (Saimiri sciureus) were purchased from the American Type Culture Collection (ATCC; ). African green monkey (Cercopithecus aethiops) fibroblasts were also obtained from ATCC. The lemur (Varecia variegates rubber) fibroblasts were obtained from the Center for Reproduction of Endangered Species, Zoological Society of San Diego, and the pygmy chimp (Pan paniscus) lymphoblast sample was kindly provided by Dr. D. Nelson from Baylor College of Medicine. WEB SITE REFERENCES : ; The American Type Culture Collection (ATCC) web site. ; NCBI web site. ; Web site for Image software. ; The Wellcome Trust Sanger Institute. ; TIGR. ; Center for Reproduction of Endangered Species, Zoological Society of San Diego, San Diego, CA. ; Children's Hospital Oakland -- BAC-PAC Resources. Backmatter: PMID- 12359838 TI - Memory Consolidation for the Discrimination of Frequency-Modulated Tones in Mongolian Gerbils Is Sensitive to Protein-Synthesis Inhibitors Applied to the Auditory Cortex AB - Differential conditioning of Mongolian gerbils to linearly frequency-modulated tones (FM) has recently received experimental attention. In the study of the role of cerebral protein synthesis for FM discrimination memory, gerbils received post-training bilateral injections of anisomycin into the auditory cortex under light halothane anesthesia. Compared with saline-treated controls, anisomycin-treated gerbils showed a discrimination decrement during the subsequent three days of training. They markedly improved their performance within training sessions, but started each session at low levels. When repeatedly trained gerbils received post-session injections of anisomycin, discrimination performance during subsequent sessions was similar to the pre-injection performance, indicating that retention, retrieval, reconsolidation, and expression of the established reaction were not affected. However, the improvement of a partially established discrimination reaction was impaired after this treatment. Intracortical injections of emetine confirmed this finding. Neither drug affected FM discrimination learning when given several days before the initial training. Our results suggest that protein-synthesis inhibitors applied to the auditory cortex of gerbils during the post-acquisition phase interfered with learning and memory-related aspects of FM processing. The resulting deficit was evident for a number of post-injection training days. This effect was probably due to impaired consolidation, i.e., processes required for long-term stabilization or retrieval of the memory trace while leaving short-term memory intact. Keywords: Introduction : Long-term plastic changes in the brain, including those supporting memory formation, are assumed to depend on permanent functional alterations in neural cells that require reprogramming of gene expression . Experimental evidence derived from a variety of species and learning paradigms is indicative for consolidation processes that are relevant for stabilization and/or retrieval of long-term memory . These processes are sensitive to disruption by interfering events and agents, such as electroconvulsive shock, inhibition of protein synthesis, or brain injury. Inhibition of cerebral protein synthesis at the level of transcription or translation was shown to disrupt processes supporting long-term memory formation in animals including insects (e.g., ; ), mollusks (e.g., ; ; ; ; ), birds (e.g., ; ; ), and rodents (e.g., ; ; ; ; ). Long-term memory formation consists of multiple phases. Characteristically, the formation of long-term memory is sensitive to protein-synthesis inhibitors applied within narrow time windows around or after acquisition. (For review, see ; ; .) However, some findings have suggested that the same manipulations that cause memory loss after initial learning may also lead to loss of that memory after its reactivation or retrieval. (For review, see .) Thus, a recent study implies that inhibition of protein synthesis can interfere with reactivated as well as new memories . It has been suggested that, when retrieved, memories may reenter a labile state and require reconsolidation processes in order to be maintained. The auditory cortex of mammals is believed to be the substrate for particular aspects of auditory stimulus processing, mediating certain forms of auditory performance as well as playing a role in auditory learning (e.g., ; ; ; ; ; ; ; ; ; ; ). Previous studies on the Mongolian gerbil (Meriones unguiculatus) have revealed physiological correlates of learning-related changes in the auditory cortex of this small desert rodent. Thus learning-induced alterations of neuronal activity patterns have been demonstrated through the use of metabolic labeling techniques . Moreover, single- and multi-unit electrophysiology revealed that the receptive fields of single units in the gerbil auditory cortex show learning-induced plasticity . Lesion studies have suggested that the gerbil auditory cortex plays a critical role in discriminating the directions of modulation of linearly frequency-modulated tones (FM) . The lesion effects on FM discrimination were compatible with both the hypothesis of a purely sensory deficit in FM processing and the hypothesis of an involvement of the auditory cortex in mechanisms of learning and memory. The following experiments were performed to define the role of cerebral protein synthesis in differential conditioning of gerbils to FM. Our findings suggest that post-training bilateral injections of the protein-synthesis inhibitors anisomycin (ANI) or emetine (EME) into the auditory cortex impair processes required for formation and/or retrieval of long-term, but not short-term, memory of the FM discrimination reaction. RESULTS : Gerbils were trained daily to discriminate between rising and falling tones. Learning behavior and performance were studied in a GO/NO-GO task aiming at avoidance of a mild foot shock by crossing the hurdle in a two-way shuttle box. The discrimination performance was quantified by the measure D, i.e., the difference between the relative frequencies of hurdle crossings in response to the reinforced and unreinforced stimuli. Protein-synthesis inhibitors were bilaterally applied to the auditory cortex, and the effect of treatment in relation to control-injected animals was tested in subsequent training sessions performed every 24 h. Three sets of experiments were performed. In the initial experiment, gerbils received injections of ANI after completion of the first training session. In a second set of experiments, gerbils were repeatedly trained to attain either a fully or a partially established discrimination reaction before post-session injections of ANI or EME were performed. In a third set of experiments, drugs were administered several days before the initial training. Three gerbils were used to validate our procedure of positioning the injection tracks using the 2-deoxyglucose autoradiography. Brains of injected animals were horizontally cut (40 mum) and the sections were exposed to autoradiographic analysis as detailed elsewhere . Each injection yielded an injection track visible in the horizontal sections. Metric analysis confirmed the proper placement of the injection sites in the primary (AI), anterior (AAF), and posterior (DP/VP) fields of the auditory cortex (see ; ). A representative case is shown in Figure . Figure 1 | Localization of injection sites using 2-deoxyglucose autoradiography in a representative case. Localization of injection sites using 2-deoxyglucose autoradiography in a representative case. Five autoradiographs of horizontal sections through the left brain hemisphere of a gerbil having received 1 mul injections into the auditory cortical primary (AI), anterior (AAF), and posterior (DP/VP) fields are shown. For each section its dorsoventral position is indicated as referenced to the dorsal tip of the corpus striatum (cf. ). The injection tracks for each of the auditory cortical fields are marked by arrows connected with solid lines. The dashed line marks the rostrocaudal position of the anterior tip of the hippocampus along which sections were aligned and which provides a suitable anatomical landmark for referencing the position of the individual cortical fields and the corresponding injection tracks. Other abbreviations: c.str. corpus striatum; cer. cerebellum; i.c. colliculus inferior; hip. hippocampus. Effects of ANI Injected After Initial Training | In an examination of the sensitivity of a newly acquired FM discrimination reaction to protein-synthesis inhibitors, gerbils received injections of saline or of a 113 mM solution of ANI into the auditory cortex twice: immediately and 2 h after completion of the initial training session. The effect of treatment was tested in subsequent training sessions performed on three consecutive days. Figure A compares the discrimination rates D of ANI-treated gerbils with those of the control-injected animals. An analysis of variance (ANOVA) over sessions across treatment groups revealed significant main effects of treatment (F(1,19) = 5.81; P < 0.05) and session (F(3,57) = 8.51; P < 0.001) and no significant treatment x session interaction (F(3,57) = 1.49). The significant effect of session in conjunction with the lack of a significant treatment x session interaction implies that both treatment groups were able to improve their discrimination rates over sessions. The significant effect of treatment indicates a decrement in the discrimination performance of gerbils that received injections of 113 mM ANI when compared with saline-treated gerbils. Figure 2 | Post-training injections of protein-synthesis inhibitor anisomycin (ANI) impaired frequency-modulated tone (FM) discrimination learning. Post-training injections of protein-synthesis inhibitor anisomycin (ANI) impaired frequency-modulated tone (FM) discrimination learning. Gerbils were trained on the FM discrimination task every 24 h. Saline (open symbols, n = 10) or 113 mM ANI (filled symbols, n = 11) were bilaterally injected twice, i.e., immediately after and 2 h after completion of the first training session. The discrimination rates D estimated (a) across all four training sessions or (b) across the initial 24 (block A) and final 36 (block B) trials of all four training sessions are shown. (c) Discrimination rates D estimated across blocks A and B are expressed as averages of sessions 2 to 4 [(*) significantly different from the values in block B of ANI-treated gerbils and in blocks A and B of saline-treated contols (t-test)]. (d) The inter-session differences DeltaD, i.e., the differences between D recorded in block A of a given session and in block B of the preceding session, are expressed as averages of all three inter-session intervals [(#) significant difference between treatment groups (t-test)]. All data points represent group means +- S.E.M. To further analyze the performance decrement monitored after injections of 113 mM ANI, the discrimination rates D were calculated for the initial 24 trials (referred to as block A) and the remaining 36 (referred to as block B) of each training session. ANOVA over the post-injection training sessions across treatment groups and blocks revealed a significant main effect of treatment (F(1,19) = 11.36; P < 0.005) but not of block (F(1,19) = 2.64). The treatment x block interaction reached nearly the level of statistical significance (F(1,19) = 4.00; P = 0.055). These findings indicate again a significant difference between the discrimination rates of ANI-treated gerbils and controls, whereas the discrimination rates monitored in blocks A and B did not significantly differ from each other. The finding that the interaction between treatment and block nearly reached the level of statistical significance might point to an uneven expression of the effect of treatment in blocks A and B. Indeed, this is reflected in the patterns of performance shown in Figure B. ANI-treated gerbils markedly improved their discrimination scores within the post-injection training sessions, but showed low levels of performance in block A, or at the beginning, of each of the sessions. Saline-treated controls did not show this type of behavior. Planned comparisons revealed significant differences between treatment groups in block A of sessions 3 (t(19) = 2.34, P < 0.05) and 4 (t(19) = 2.22, P < 0.05) but not in block B of sessions 3 (t(19) = 0.27, P > 0.7) and 4 (t(19) = 1.11, P > 0.2). The same pattern of results is reflected in Figures C and D. Figure C shows the discrimination rates averaged over the post-injection training sessions. The mean discrimination rate of ANI-treated gerbils monitored in block A was significantly lower than those monitored in block B of the same group (t(10) = 2.60, P < 0.05) and in blocks A (t(19) = 4.03, P < 0.001) and B (t(19) = 3.67, P < 0.002) of controls. In contrast, the mean discrimination rate of ANI-treated gerbils monitored in block B did not significantly differ from those monitored in blocks A (t(19) = 1.23, P > 0.2) and B (t(19) = 0.96, P > 0.3) of controls. Moreover, the mean discrimination rate of controls monitored in block A did not significantly differ from the level in block B of the same group (t(19) = 0.26, P = 0.8). This latter finding suggests that, in contrast to the ANI-treated group, the major increase in the discrimination performance of controls occurred between block B of a given session and block A of the following session, rather than within the training sessions. Indeed, the mean inter-session differences shown in Figure D differed significantly between groups (t(19) = 2.50, P < 0.025). They exhibited positive values for controls but negative values for ANI-treated gerbils, indicating that, on average, the discrimination rates of controls increased whereas those of ANI-treated gerbils decreased between the final part of a given session and the initial part of the following one. To look for side effects of ANI on mechanisms that may interfere with the learning performance, general parameters (such as footshock intensities and avoidance times, i.e., the times required to enter the goal compartment in response to the CS+), were recorded during the training sessions. These parameters that reflect the sensitivities of sensory and motor systems did not differ between treatment groups . To assess the state of arousal and activity, the inter-trial crossings and the total numbers of conditioned reactions, i.e., hurdle crossings in response to CS+ and CS- presentations, were counted. Additionally, as the discrimination deficit of ANI-treated gerbils was most pronounced during the first part of a given training session, the hurdle crossings performed during 3 min of habituation immediately preceding each training session were monitored. Compared with saline-treated controls, injections of ANI caused no significant changes in any of these parameters . Thus, neither the drug per se nor an inhibition of cerebral protein synthesis caused alterations in the state of arousal and activity or deficits in sensory transduction and motor systems that could account for the learning deficits. Moreover, compared to the initial day of training, ANI-treated gerbils as well as saline-treated gerbils showed decreasing exploratory activities and increasing numbers of conditioned reactions over days . These changes might reflect an increasing familiarity with the training chamber and the experimental procedure, probably indicating that, irrespective of treatment, these gerbils were able to remember the experimental situation. Table 1 | Data recorded during 3 min of habituation to the shuttle box and during the sessions of FM discrimination training of gerbils injected with saline (n = 10) or 113 mM ANI (n = 11) after initial training Taken together, these findings indicate that gerbils injected with 113 mM ANI after the initial session of FM discrimination training showed a decrement in discrimination performance during subsequent training performed on three successive days. ANI-treated gerbils improved their discrimination rates within the sessions but started each of the sessions at low levels. In contrast, saline-treated controls improved their discrimination performance between the sessions. Additional parameters suggest that ANI did not cause effects on the state of arousal and activity or on sensory and motor systems that could account for the decrement in discrimination performance. To control for effects of ANI apart from those caused by protein-synthesis inhibition, a 0.04 mM solution of ANI was bilaterally infused into the auditory cortex of gerbils twice: immediately after and 2 h after completion of the initial training session. This dose was shown to cause only a non-significant inhibition of protein synthesis following injection into rat hippocampus . In contrast to injections of 113 mM ANI, administrations of 0.04 mM ANI to the gerbil auditory cortex did not affect FM discrimination learning during the subsequent three days of training (D of session 1 to session 4, 0.04 mM ANI: 5.50 +- 1.26, 12.17 +- 1.54, 14.00 +- 2.27, 15.17 +- 1.58; D of session 1 to session 4, saline: 0.50 +- 1.28, 10.67 +- 2.88, 10.50 +- 1.61, 15.50 +- 2.92; means +- s.e.m., n = 6 per group). ANOVA revealed no significant main effect of treatment (F(1,10) = 3.03), a significant main effect of session (F(3,30) = 15.69; P < 0.001), and no significant interaction between both (F < 1). Effects of Protein-Synthesis Inhibitors Injected After Repeated Training | For the assessment of potential effects of ANI on the behavioral expression and on processes involved in retention, retrieval, and reconsolidation of an established FM discrimination reaction, gerbils were trained every 24 h to reach an asymptotic level of discrimination performance. Surgery was performed after session 9, and saline or 113 mM ANI were injected after session 10. Figure A compares the discrimination rates of sessions 8 to 13 between treatment groups. ANOVA revealed neither significant main effects of treatment (F < 1) and session (F(5,20) = 1.14) nor a significant interaction between both (F < 1). Thus, post-session injections of ANI into the auditory cortex of well-trained gerbils did not affect the performance of the established FM discrimination reaction during subsequent days of training. The higher variations in the discrimination rates monitored during session 10 probably reflect an effect of surgery. Figure 3 | Protein-synthesis inhibitors impaired the improvement but not the expression of a partially established frequency-modulated tone (FM) discrimination reaction. Protein-synthesis inhibitors impaired the improvement but not the expression of a partially established frequency-modulated tone (FM) discrimination reaction. Gerbils were trained on the FM discrimination task every 24 h. Bilateral injections of saline or drugs were performed twice, immediately and 2 h after completion of the sessions indicated by dashed lines. (a) Gerbils trained up to the asymptotic level of FM discrimination performance were injected with saline (open circles, n = 3) or 113 mM protein-synthesis inhibitor anisomycin (ANI) (filled circles, n = 3). (b) Gerbils with a partially established FM discrimination reaction were injected with saline (open circles, n = 6) or 66 mM ANI (filled circles, n = 6). (c) Gerbils with a partially established FM discrimination reaction were injected with saline (open circles, n = 9) or 15 mM protein-synthesis inhibitor emetine (EME) (filled rectangles,n = 7). Panels a to c represent data obtained from different sets of animals. All data points represent group means +- S.E.M. Next, the effect of ANI on processes involved in the addition of new information to existing stores was assessed. For this purpose, gerbils were trained on the FM discrimination task daily for a total of 11 d (Fig. B). Surgery was performed after session 6 and saline or a 66 mM solution of ANI was administered after completion of session 7, i.e., when the discrimination rate, on average, exceeded the level of 30%. At this stage of training the discrimination reaction was only partially established and therefore still improvable by additional training. ANOVA comparing D over sessions 7 to 11 across treatments revealed significant main effects of treatment (F(1,10) = 6.66, P < 0.05) and session (F(4,40) = 3.94, P < 0.01), and no significant treatment x session interaction (F(4,40) = 1.24). Thus ANI-treated gerbils as well as controls were able to improve their discrimination performance over sessions. This improvement, however, was delayed in ANI-treated gerbils when compared with saline-treated controls. Planned examinations revealed significant group differences in D in sessions 8 (t(10) = 2.45, P < 0.05) and 9 (t(10) = 2.74, P < 0.025), but no longer in session 10 (t(10) = 1.86, P > 0.05). On day 11, the discrimination scores of ANI-treated gerbils nearly reached the level of controls (t(10) = 1.34, P > 0.2). When compared with the pre-injection level, administration of ANI did not impair the performance of the established discrimination reaction. To confirm that the effect of ANI on FM discrimination learning was caused by protein-synthesis inhibition, another inhibitor of translation, EME, was bilaterally applied to the auditory cortex of repeatedly trained gerbils after completion of session 4, i.e., when the discrimination reaction was partially established and the discrimination rate, on average, exceeded the level of 30%. ANOVA comparing D over sessions 4 to 7 across treatment groups revealed no significant main effect of treatment (F < 1), a significant main effect of session (F(3,42) = 2.96, P < 0.05), and a significant treatment x session interaction (F(3,42) = 3.04, P < 0.05). The interaction plot shown in Figure C indicates that saline-treated but not EME-treated gerbils were able to improve their performance over sessions. Accordingly, ANOVA of D within each of the treatment groups revealed a main effect of session within the vehicle-treated group (F(3,24) = 5.24, P < 0.01) but not within the EME-treated group (F < 1). Taken together, post-session administrations of two unrelated protein-synthesis inhibitors to the auditory cortex of repeatedly trained gerbils impaired the improvement of a partially established FM discrimination reaction as measured during subsequent training sessions in saline-treated controls. When compared with the pre-injection level, the performance of the task was normal, implying that neither sensory and motor systems required for performance of the discriminative behavior nor mechanisms involved in retention, retrieval, and reconsolidation of the established reaction were affected. Effects of Protein-Synthesis Inhibitors Injected Several Days Before Initial Training | To assess unspecific long-term effects of the protein-synthesis inhibitors that might interfere with FM discrimination learning, injections were performed several days before the first training session. When 113 mM ANI was applied to the auditory cortex two days before the initial training, performance of the FM discrimination reaction was not impaired in comparison to saline-treated controls (Fig. A), suggesting that acquisition, retention, and recall of the task were not affected. ANOVA comparing D over sessions across treatment groups revealed no main effect of treatment (F < 1), a significant main effect of session (F(2,18) = 6.27; P < 0.01), and no treatment x session interaction (F < 1). Similarly, injections of EME three days before the initial training had no significant effects on FM discrimination learning (Fig. B). ANOVA comparing D over sessions across treatments revealed no main effect of treatment (F < 1), a main effect of session (F(1,6) = 6.92, P < 0.05), and no treatment x session interaction (F < 1). These findings indicate that neither the drugs per se nor an inhibition of cerebral protein synthesis exerts long-term toxic effects that could account for the discrimination deficit observed when the inhibitors were given during the post-acquisition phase. Figure 4 | Administration of protein-synthesis inhibitors several days before initial training did not affect frequency-modulated tone (FM) discrimination learning. Administration of protein-synthesis inhibitors several days before initial training did not affect frequency-modulated tone (FM) discrimination learning. Gerbils were trained on the FM discrimination task every 24 h. (a) Saline (open circles, n = 5) or 113 mM protein-synthesis inhibitor anisomycin (ANI) (filled circles, n = 6) were bilaterally injected twice 48 h and 46 h before the initial training. (b) Saline (open circles, n = 3) or 15 mM protein-synthesis inhibitor emetine (EME) (filled rectangles, n = 5) were bilaterally injected twice 72 h and 70 h before the initial training. All data points represent group means +- S.E.M. DISCUSSION : Using a pharmacological approach, we addressed the question of the functional relevance of cerebral protein synthesis in the Mongolian gerbil for learning the discrimination of the directions of modulation of FM. In relation to saline-treated controls, bilateral injections of the protein-synthesis inhibitor ANI into the primary auditory cortex and its adjacent fields after FM discrimination training led to discrimination deficits measured in daily training sessions performed for up to three successive days. Evaluation of blocks of trials revealed that ANI-treated gerbils markedly improved their discrimination rates within the training sessions but started each session again at low performance levels. In contrast, saline-treated gerbils showed on average higher discrimination rates at the beginning of a given training session than at the end of the preceding one. When gerbils were repeatedly trained to reach a certain level of discrimination performance before post-session injections of ANI, the improvement, but not the expression, of the partially established discrimination reaction was impaired during subsequent training. Intracortical injections of EME, another inhibitor of translation, confirmed this finding. Neither ANI nor EME affected FM discrimination learning when administered several days before the initial training. A comprehensive set of behavioral data was collected to control for unspecific effects of pharmacological treatment. Firstly, as recorded on subsequent training days, gerbils injected with saline after the initial training showed FM discrimination rates very similar to those monitored in untreated gerbils (data not shown), indicating that treatments and procedures required for drug delivery, i.e., surgery, injection, and anesthesia, caused no impairments in FM discrimination learning and performance. These findings support a previous report demonstrating that, in rodents, halothane anesthesia causes no retrograde amnesia and an anterograde amnesia only when applied within 2 h before a learning experiment . Secondly, previous studies on Mongolian gerbils have shown that either bilateral or right auditory cortex lesions impair the discrimination of FM . In the present study, a discrimination deficit was evident only after bilateral drug administrations. Post-training unilateral injections of 113 mM ANI into the right auditory cortex twice, immediately and 2 h after initial training, did not impair FM discrimination learning as measured on subsequent training days (data not shown), suggesting that drug treatment caused no functional damage to this brain structure. Thirdly, FM discrimination learning was normal when ANI or EME was applied several days before initial training and, finally, post-session injections of the inhibitors to repeatedly trained animals did not affect performance of an established FM discrimination reaction on subsequent days . Together these findings imply that neither the procedures of surgery and injection, nor the drugs per se, caused long-term effects on mechanisms involved in acquisition, retention, retrieval, and expression of the discrimination reaction that could account for the amnesic effect. Moreover, the results indicate that, as previously shown for other species and learning paradigms , administration of the protein-synthesis inhibitors in a narrow temporal relation to the learning experiment is also required for their amnesic actions on FM discrimination learning in the gerbil to occur. ANI was shown previously to interfere with the activity of certain protein kinases even at doses that did not inhibit protein synthesis . Therefore, the effect of ANI on FM discrimination learning observed in the present study could be attributable to nonselective actions on the physiology of cortical neurons. However, the discrimination deficit was monitored after intracortical injections of 113 mM or 66 mM solutions of ANI, whereas administrations of a 0.04 mM solution of ANI were ineffective. In rat hippocampus, application of 1 mul of a 10 mM solution of ANI has been shown to cause more than 80% inhibition of protein synthesis, whereas a 0.04 mM solution had only a marginal effect . Therefore, the amnesic action of ANI monitored in the present study was dose-dependent and occurred with dosages that, as deduced from literature data, should induce more than 80% inhibition of protein synthesis in the target region. Moreover, administration of another protein-synthesis inhibitor, EME, impaired FM discrimination learning in a manner similar to that of ANI. The two drugs have the same primary action, i.e., inhibition of protein synthesis, but a different spectrum of side effects. The finding that they both act on FM discrimination learning increases the confidence that the observed effect is attributable to their common primary action. Changes in an animal's behavior measured in re-learning sessions some time after initial learning may reflect effects on many processes, including performance, acquisition, consolidation, retention, and retrieval of memory . To look for side effects of ANI on mechanisms that may interfere with the learning performance, general parameters (such as exploratory activities, inter-trial crossings, avoidance times, and footshock intensities) were monitored during the behavioral experiments . None of these parameters was significantly changed in comparison to saline-treated gerbils, indicating that the drug did not cause alterations in the state of arousal and activity or in the sensitivities of sensory and motor systems that could account for the learning deficits. Gerbils that received injections of ANI after the first training session showed a marked improvement in the discrimination rates between the initial and the final blocks of trials of subsequent training sessions (Fig. B,C). Consequently, the discrimination rates of this group monitored during the final block of trials of the post-injection training sessions did not significantly differ from the level of controls, indicating that ANI-treated gerbils were still able to perform the sensory processing required for discriminating FM as well as the sensorimotor integration and behavior necessary to demonstrate the discrimination reaction. However, these same animals failed to perfom well during the initial trials of the post-injection training sessions. The resulting mean discrimination rate in block A of ANI-treated gerbils was near zero and differed significantly from the level in controls, implying that much of the previously learned information was lost during the inter-session intervals in which the animals were not exposed to the maze. Therefore, these findings suggest that, in the gerbil, mechanisms that support learning and memory-related aspects of FM processing are sensitive to protein-synthesis inhibitors applied to the auditory cortex shortly after training. The results can be explained by postulating that inhibition of protein synthesis in the gerbil auditory cortex during the post-acquisition phase causes a deficit in the consolidation of FM discrimination memory, that is, in a progressive stabilization of the newly acquired memory trace and in processes that ensure its retrieval. In this sense, the increase in performance observed within each of the post-injection training sessions might be indicative either of a recovery during the course of the session from a retrieval failure, or of the unimpaired capabilities of acquiring the discrimination reaction and of forming a short-term memory for it, thus of re-learning the discrimination reaction on a daily basis. Because neither ANI nor EME impaired the expression of a previously established FM discrimination reaction , the amnesic action of the protein-synthesis inhibitors cannot be explained as an effect on the general abilities to retain and retrieve the information. Moreover, in contrast to some suggestions in relation to other species, brain structures, and learning paradigms , the present findings, consistent with other studies , do not lend support to a general view that memory retrieval activates reconsolidation processes that are susceptible to protein-synthesis inhibitor --induced amnesia. However, the inhibitors exerted a decrement in the improvement of a partially established FM discrimination reaction. As this decrement was comparable in extent to that observed after original learning, initial acquisition of the discrimination reaction and the addition of information to pre-existing stores may recruit common mechanisms of consolidation. Gerbils treated with ANI showed some improvement over training days (Fig. A,B) and were able to reach discrimination rates similar to those of control-injected animals at four days post-injection (Fig. B), indicating that long-term memory retention and/or retrieval were not completely and irreversibly blocked. However, compared with saline-treated controls, a discrimination deficit was evident in drug-treated gerbils for several training days. The exact reasons for this finding are currently not clear. Effects of ANI topically applied to distinct brain regions on learning-related behavior during two or more successive days have recently been observed in some species and paradigms, for example, in a classical conditioning of rabbits and in the extinction of a fear-motivated response of rats , but not in others, such as the extinction of a conditioned taste aversion in rats . This implies that the long-term effect of ANI application may be specific to the task, species, and/or brain structure under investigation. The present results led to the assumption that, during the post-acquisition phase, some type of protein-synthesis --dependent trace is produced that either sensitizes relevant local nodes (e.g., neurons, synapses) for future use or modifies them for later processes of consolidation required for stabilization and/or retrieval of long-term memory. The finding that the discrimination scores of controls observed during the first block of trials of a re-training session were increased over those in the final block of the preceding session (Fig. B,D) would support this assumption. Experimental evidence suggests the importance of such mechanisms for plasticity-related processes in the nervous system of invertebrates and vertebrates . Whether similar mechanisms contribute to the long-lasting effect of protein-synthesis inhibition observed in the present study is subject of future investigations. Interestingly, as published during the reviewing of this manuscript, learning-induced tuning plasticity in the auditory cortex of the guinea pig may exhibit long-term consolidation, attaining asymptote at 3 d . In summary, the present results indicate that application of protein-synthesis inhibitors to the auditory cortex of gerbils during the post-acquisition phase interfered with long-term FM discrimination memory for a number of subsequent training days, while leaving short-term memory intact. Control experiments imply that this effect was likely due to impaired memory consolidation. Future studies are required to test hypotheses that account for these observations. MATERIALS AND METHODS : Animals | Seventy-seven adult male Mongolian gerbils (Meriones unguiculatus) were used for the experiments. The animals were housed in groups of five and given free access to standard laboratory chow and tap water on a 12-h light/dark cycle (light on at 6 a.m.). All experiments were approved by the animal care committee of Sachsen-Anhalt (No. 53b-42502/2 --111/150; IfN MD) in accordance with the NIH Guide for the Care & Use of Laboratory Animals (NIH, 1996). Surgical Procedures and Intracortical Injections | The procedures of surgery and intracortical injection have previously been described in detail . In brief, gerbils were deeply anesthetized (0.4 mg ketamine and 0.3 mg xylacine per 100 g body weight, i.p.). The cranial skin was disinfected and incised and three holes of about 1 mm in diameter were bilaterally drilled into the skull at locations covering the primary, anterior, and posterior fields of the auditory cortex (for coordinates, see ). After approximately 20 h of recovery, intracortical injections were performed under light halothane anesthesia using glass micropipettes mounted on a nanoinjector (WPI). Per target region, 1 mul of either anisomycin (Sigma; 0.04 mM = 0.01 mug/mul, 66 mM = 18 mug/mul, or 113 mM = 30 mug/mul), or emetine (Sigma; 15 mM = 8 mug/mul), or vehicle (0.9 % saline) were injected over a period of 4 min. Injections were repeated after an interval of 2 h. All injections were administered to all three sites every time an injection was performed. All solutions used for intracortical injections were adjusted to pH 7. The dosages used in the present study were deduced from literature data showing that ANI and EME efficiently inhibit protein synthesis in various species and brain structures (for references, see Introduction) and that a level of 80% protein-synthesis inhibition is critical for amnesic effects to occur . Specifically, the highest dose (113 mM) of ANI was chosen based on previously published studies showing 80% suppression of protein synthesis in rat cerebral cortex and hippocampus after injections of 100 mug ANI per hemisphere . During the course of our experiments, data were published showing that even lower doses of ANI are sufficient to induce >80% suppression of protein synthesis in rat hippocampus . Therefore, in subsequent experiments the dose of ANI was reduced to 66 mM as indicated. To control for amnesic effects of ANI apart from those caused by protein-synthesis inhibition, the lowest dose (0.04 mM) was used. This dose was shown to induce only a nonsignificant inhibition of protein synthesis following injection into rat hippocampus . The dose of EME was chosen based on previous studies on chicks showing a highly localized inhibition of protein synthesis after intracranial injections . Behavioral Experiments | Gerbils were trained in daily sessions to discriminate the directions of modulation of linearly frequency-modulated tones as described earlier . In brief, behavioral experiments were performed in a shuttle box (E10 --15, Coulbourn Instruments) consisting of two compartments separated by a 3 cm-high hurdle. Before each of the training sessions, gerbils were allowed to habituate for 3 min to the experimental situation in the shuttle box without acoustical stimulation and footshock. During training, conditioned stimuli consisting of FM of 1 --2 kHz ascending frequency (CS+) or 2 --1 kHz descending frequency (CS-) were presented in a pseudorandomized order. Animals were required to perform a hurdle crossing when presented with CS+ and to suppress this response when presented with CS-. Hurdle crossings occurring within 6 sec of the onset of CS+ were regarded as correct conditioned responses (CR+). If, during CS+ presentation, a response did not occur within 6 sec, the gerbil was punished by footshock of 200 --500 muA through the grid floor. The footshock lasted for up to 4 sec and could be terminated by the gerbil by crossing the hurdle. Footshock intensities required to cause the gerbils to cross the hurdle were individually adjusted, starting with 250 muA, during the first footshock presentations of each training session. Hurdle crossings occurring within 6 sec of the onset of CS- were regarded as false alarms (CR-) and punished by a footshock of 300 muA for 1 sec. A training session consisted of 60 trials, that is, 30 presentations of each CS+ and CS-. The inter-trial interval was pseudorandomly varied (mean = 15 sec, range = 12 --18 sec). Analysis of Behavioral Data | Relative frequencies of CR+ and of CR- were calculated as the number of CR+ expressed as percent of trials with CS+ presentations, and the number of CR- expressed as percent of trials with CS- presentations, respectively. The discrimination performance was quantified by the measure D[%], i.e., the difference between the relative frequencies of CR+ and CR-. D was estimated for each animal and session. Where indicated, D was additionally calculated for blocks representing the initial (block A) and the final (block B) parts of a session. To ensure equal numbers of CS+ and CS- presentations within each of the blocks, block A consisted of 24 trials and block B consisted of 36 trials. To assess drug effects on arousal and activity, the numbers of hurdle crossings during the habituation period preceding each training session as well as the inter-trial activity, i.e., the numbers of hurdle crossings occurring during the inter-trial intervals, were monitored. To assess drug effects on sensory systems and motor coordination, footshock intensities required to cause the gerbils to cross the hurdle and avoidance times, i.e., the times required to change the compartment during CR+, were recorded within the training sessions. All data are expressed as group means +- S.E.M. Statistical evaluation was performed using a repeated measures ANOVA (StatView 5.0.1, SAS). Planned examinations, i.e., comparisons of individual data points, were performed using Student's two-tailed t-test for paired or unpaired comparisons where appropriate. P values of <0.05 were considered as statistically significant. Backmatter: PMID- 12359839 TI - Eyeblink Classical Conditioning and Interpositus Nucleus Activity Are Disrupted in Adult Rats Exposed to Ethanol as Neonates AB - Neonatal exposure to ethanol in rats, during the period of brain development comparable to that of the human third trimester, produces significant, dose-dependent cell loss in the cerebellum and deficits in coordinated motor performance. These rats are also impaired in eyeblink conditioning as weanlings and as adults. The current study examined single-unit neural activity in the interpositus nucleus of the cerebellum in adults following neonatal binge ethanol exposure. Group Ethanol received alcohol doses of 5.25 g/kg/day on postnatal days 4 --9. Group Sham Intubated underwent acute intragastric intubation on postnatal days 4 --9 but did not receive any infusions. Group Unintubated Control (from separate litters) did not receive any intubations. When rats were 3 --7 mo old, pairs of extracellular microelectrodes were implanted in the region of the interpositus nucleus. Beginning 1 wk later, the rats were given either 100 paired or 190 unpaired trials per day for 10 d followed by 4 d of 100 conditioned stimulus (CS)-alone trials per day. As in our previous study, conditioned response acquisition in Group Ethanol rats was impaired. In addition, by session 5 of paired acquisition, Group Sham Intubated and Group Unintubated Control showed significant increases in interpositus nucleus activity, relative to baseline, in the CS --unconditioned stimulus interval. In contrast, Group Ethanol failed to show significant changes in interpositus nucleus activity until later in training. These results indicate that the disruption in eyeblink conditioning after early exposure to ethanol is reflected in alterations in interpositus nucleus activity. Keywords: Introduction : Early exposure to alcohol in humans is known to produce a variety of behavioral and neural abnormalities (for reviews, see ; ; ). In some cases of heavy prenatal exposure to alcohol, fetal alcohol syndrome (FAS) can be diagnosed by the presence of the constellation of facial dysmorphology, growth deficiency, and central nervous system damage in the affected child. The development of animal models of FAS has been essential in understanding, at a detailed level, some of the behavioral and neural abnormalities associated with early exposure to alcohol . This research has revealed that cortical, hippocampal, basal gangliar, and cerebellar areas are particularly vulnerable to the teratogenic effects of alcohol, and that the vulnerability of specific brain structures changes over stages of development. Much recent research has concentrated on cerebellar abnormalities that occur in conjunction with early exposure to alcohol. The use of a neonatal rat model of binge alcohol exposure during the period of brain development comparable to that of the human third trimester has been critical to this research. This model involves exposing neonatal rats to high doses of alcohol over a short period of time via either a chronic or an acute intragastric infusion procedure. Either of these procedures allows control of the dose, pattern, and resulting profile of blood alcohol concentrations produced (BACs) while administering alcohol during a stage of brain development that is vulnerable to environmental insults . These are critical factors when cross-species comparisons of the effects of alcohol are made. Research using this model has shown that BACs in excess of 200 mg/dL produce significant loss of Purkinje cells in neonatal rats that increases with increasing BACs . More recent studies have confirmed and extended these findings using unbiased stereological procedures, in which estimates of the total number of cells in a structure are not affected by changes in cell density . Cell loss is greatest during the early stages of Purkinje cell dendritic outgrowth and synaptogenesis (which occurs in rats on postnatal days [PDs] 4 --9). Less Purkinje cell loss occurs with binge exposure during neurogenesis (which occurs in rats on prenatal days 13 --16) , or during later differentiation . PD days 4 --6 appear to be the period of greatest vulnerability, with a dose-related loss of Purkinje cells that can reach 50% --60% of control values . Early binge exposure to ethanol also produces loss of cells in the inferior olive , a brainstem area that is the source of climbing fiber projections to Purkinje cells. We have recently shown that early exposure to ethanol produces loss of cerebellar deep nuclear cells (J.T. Green, T. Tran, J.E. Steinmetz, and C.R. Goodlett, in prep.), the target cells of Purkinje cells and the final relay for the majority of cerebellar output to other brain regions. Rats exposed to binge levels of ethanol as neonates show disrupted performance in motor tasks thought to require an intact cerebellum . However, the precise nature of cerebellar engagement by these tasks is not well defined, making correlations between cell loss and behavioral performance somewhat difficult to interpret. Furthermore, it is difficult to separate learning from performance deficits with these tasks. A task that is particularly well suited to addressing these issues is eyeblink classical conditioning. In eyeblink classical conditioning, a conditioned stimulus (CS), which initially elicits no response, precedes an unconditioned stimulus (US), which elicits a reflexive, or unconditioned response in the form of an eyeblink. Learning is shown by the occurrence of eyeblinks after the CS but before the US (conditioned responses; CRs). An extensive body of research in rabbits, using a variety of approaches, has characterized the brainstem-cerebellar substrates of eyeblink conditioning (for review, see ). The auditory or visual CS is thought to project to cerebellar cortex and deep nuclei via mossy fibers arising from lateral pontine nuclei. The somatosensory US is hypothesized to project to cerebellar cortex and deep nuclei via climbing fibers arising from the dorsal accessory olive. One of the deep nuclei, the interpositus nucleus ipsilateral to the eye that receives the US, appears to be the site of plasticity for learning the CR. Overlying cerebellar cortex seems to play at least a modulatory role in learning eyeblink conditioning, perhaps through control of the amplitude and timing of CRs. These substrates appear to be the same in the rat . In addition to knowledge of the neural circuitry engaged by eyeblink conditioning, control procedures, such as unpaired presentations of the CS and US, allow separation of performance factors from learning. Knowledge of the precise cerebellar substrates of eyeblink conditioning makes it particularly useful for assessing the effects of early exposure to alcohol on the developing cerebellum (for review, see ). Studies have shown that rats exposed to a high dose of alcohol on PDs 4 --9 are impaired in eyeblink conditioning both as weanlings (23 --24 d old) and as adults (>3 mo old) . Given that the number of deep nuclear neurons is highly correlated with the rate of acquisition of eyeblink conditioning ( J.T. Green, T. Tran, J.E. Steinmetz, and C.R. Goodlett, in prep.), deficits in eyeblink conditioning in rats exposed to binge levels of ethanol as neonates may be directly caused by neuronal depletion in the deep nuclei, particularly in the interpositus nucleus. The purpose of the present study was to assess activity of interpositus nucleus neurons during eyeblink conditioning in adult rats after neonatal exposure to binge levels of ethanol. Although we have confirmed that deep nuclear neurons are lost after neonatal exposure to binge levels of ethanol, we do not know if those that remain show normal activity during eyeblink conditioning. The current study shows that adult interpositus nucleus neurons show a delay in the development conditioning-related activity after neonatal exposure to binge levels of ethanol. RESULTS : Histology | Electrode tracks and marking lesions were used to confirm electrode placement. Forty-seven rats (29 underwent paired conditioning; 18 underwent unpaired conditioning) had electrode placements in the region of the left interpositus nucleus, distributed throughout the dorsal --ventral extent. Group totals were as follows: Group Paired Ethanol (EtOH; n = 9); Group Paired Sham Intubated (SI; n = 11); Group Paired Unintubated Control (UC; n = 9); Group Unpaired EtOH (n = 5); Group Unpaired SI (n = 6); and Group Unpaired UC (n = 7). An additional 19 rats had placements outside of the interpositus nucleus, most commonly in either white matter just dorsal to the interpositus nucleus or in brainstem areas somewhat ventral to the interpositus nucleus, such as the dorsal cochlear nucleus or the lateral vestibular nucleus. Only data from the 47 rats with placements in the interpositus nucleus are described. Note that because of technical difficulties, data from extinction sessions for one rat from Group EtOH and three rats from Group UC were not available for analysis. Figure shows a typical electrode placement, along with a sample of spontaneous neural activity recorded from this location during adaptation. Figure 1 | An example of a microelectrode placement in the ipsilateral interpositus nucleus and corresponding spontaneous unit activity from this site. An example of a microelectrode placement in the ipsilateral interpositus nucleus and corresponding spontaneous unit activity from this site. Blood Alcohol Concentrations | Blood samples were available for 11 of 14 EtOH-treated animals with a confirmed interpositus nucleus electrode placement. The mean BAC of these 11 rats on PD 6 was 344 (+-36 SEM) mg/dL. Spike Separation | Mean threshold for detecting spikes across all treatment groups was 0.39 mV. There were no differences between treatment groups in threshold for detecting spikes. For paired groups, a total of 19 --26 units were separated per group per session. For unpaired groups, a total of 10 --18 units were separated per group per session. Mean number of units separated per animal per session was approximately two across all treatment and training groups. Given that our previous study established that rats from Group EtOH have about 50% fewer deep nuclear neurons than do control groups (J.T. Green, T. Tran, J.E. Steinmetz, and C.R. Goodlett, in prep.), one might expect that we would have recorded from fewer neurons in this group. However, assuming uniform loss of cells in the interpositus nucleus (which is our working assumption), loss of neurons in the interpositus nucleus in rats from Group EtOH would be expected to result in depressed background activity as well (i.e., the signal-to-noise ratio would be the same in Group EtOH as in the control groups), allowing discrimination of similar numbers of units in Group EtOH and control groups, even though fewer units exist in Group EtOH. In addition, we limited spike separation to no more than the three best units per rat in any one session. It is possible that more than three units would be separable in control groups, whereas three was approaching the limit in Group EtOH. Adaptation | During adaptation, all rats received 20 presentations of periocular stimulation followed by 20 presentations of the tone CS. Eyeblink responses to the 20 presentations of the tone were analyzed using a one-way analysis of variance (ANOVA). Adaptation data were available for 46 of 47 rats. There was no difference between treatment groups in either the percentage of eyeblink responses, F(2,43) = 1.72, P > 0.05 or in the amplitude of eyeblink responses, F(2,43) = 1.71, P > 0.05. Average percentage of spontaneous eyeblink responses to the tone (during the 81 --350 msec after the tone in which we measured CRs during acquisition) on these initial exposures was 33.7%. Baseline firing rates (spikes per second) for 350 msec before US onset on the 20 US-alone trials and for 350 msec before CS onset on the 20 CS-alone trials were calculated. A one-way ANOVA revealed no difference between treatments for either baseline firing rate before periocular stimulation onset, F(2,89) = 2.37, P > 0.05 or baseline firing rate before CS onset, F(2,93) = 0.99, P > 0.05. Mean firing rates before periocular stimulation onset were 35.9 (+-6.6 SEM) spikes/sec for Group EtOH, 28.4 (+-6.0 SEM) spikes/sec for Group SI, and 47.4 (+-6.4 SEM) spikes/sec for Group UC. Mean firing rates before CS onset were 46.7 (+-7.7 SEM) spikes/sec for Group EtOH, 38.9 (+-7.2 SEM) spikes/sec for Group SI, and 53.5 (+-7.5 SEM) spikes/sec for Group UC. See Figure for an example of unit activity during adaptation. Paired Groups | Acquisition: Behavior | Data from the 29 rats that underwent paired eyeblink conditioning were analyzed. Group EtOH attained a lower asymptote of learning than did control groups, whether measured by percentage of CRs or CR amplitude, that was comparable to our previous study . Groups SI and UC did not attain quite as high an asymptote as in our previous study. These results were confirmed by statistical analyses. Learning for the rats that underwent paired eye-blink conditioning was analyzed with a 3 (treatment: EtOH, SI, UC) x 10 (acquisition session) repeated-measures ANOVA with percentage of CRs as the dependent measure. This analysis revealed a significant effect of session, F(9, 234) = 7.44, P < 0.001. Although the percentage of CRs was consistently lower in Group EtOH compared with control groups, the treatment effect did not quite attain statistical significance (P = .064). A second ANOVA with CR amplitude as the dependent measure revealed a significant effect of session, F(9, 234) = 7.55, P < 0.001, and a significant effect of treatment, F(2,26) = 3.40, P < 0.05. Post hoc Tukey's Honestly Significant Difference (HSD) tests revealed a significant difference between Group EtOH and Group UC, with Group SI performing intermediate to, but not significantly different from either of them . Figure 2 | Learning in treatment groups that received paired eyeblink conditioning as a function of acquisition and extinction session. Learning in treatment groups that received paired eyeblink conditioning as a function of acquisition and extinction session. (A) Percentage of conditioned eyeblink responses. (B) Amplitude of eyeblink responses. The lack of a statistically significant difference between Group EtOH and Group SI was surprising, given the results of our previous study . The current results can be attributed to two rats in Group SI who conditioned exceptionally poorly. Without these two rats, there were clear and consistent significant differences between Group EtOH and Group SI in both percentage of CRs and CR amplitude. Comparisons were conducted between paired and unpaired training for each treatment group to assess further the issue of whether Group SI performed differently from Group EtOH and similarly to Group UC. For each treatment group, a 2 (training: paired, unpaired) x 10 (session) ANOVA with percentage of CRs as the dependent measure was conducted. For Group EtOH, this analysis revealed no significant effects. In contrast, for Group SI, this analysis revealed a significant effect of session, F(9,135) = 3.32, P < 0.01, whereas the effect of training just missed attaining significance, F(1,15) = 3.68, P = 0.07. For Group UC, this analysis revealed a significant training x session interaction, F(9, 126) = 1.89, P = 0.05. Post hoc one-way ANOVAs comparing training within each session revealed that rats in Group UC receiving paired training showed a significantly higher percentage of CRs in sessions 4, 6, 7, 9, and 10 (P's = 0.006 to 0.05). A similar set of analyses was conducted with CR amplitude as the dependent measure. For Group EtOH, this analysis revealed a significant training x session interaction effect, F(9,108) = 2.19, P < 0.05. Post hoc one-way ANOVAs comparing training within each session revealed that Group EtOH rats differed significantly in CR amplitude only in session 1, in which rats that received unpaired training actually outperformed those that received paired training. For Group SI, this analysis also revealed a significant training x session interaction effect, F(9,135) = 1.92, P = 0.05. In contrast to Group EtOH, however, post hoc ANOVAs revealed a significantly higher CR amplitude in Group SI rats that underwent paired training compared with those that underwent unpaired training in sessions 4 --7, 9, and 10. Finally, for Group UC, this analysis revealed significant effects of training, F(1,14) = 5.68, P < 0.05 and session, F(9,126) = 3.54, P < 0.01. Group UC rats that underwent paired training showed a significantly higher CR amplitude than did Group UC rats that underwent unpaired training across all sessions. Acquisition: Unit Activity | Baseline firing rates (spikes per second) were calculated across all paired trials of a session for the 350-msec period before the tone CS onset. Separate one-way ANOVAs for sessions 1, 5, and 10 comparing treatment groups, with baseline firing rate as the dependent measure, revealed no difference in baseline firing rate. Collapsed across sessions, mean baseline firing rate was 38.1 (+-3.2 SEM) spikes/sec for Group EtOH, 47.0 (+-3.1 SEM) spikes/sec for Group SI, and 52.1 (+-3.2 SEM) spikes/sec for Group UC. Figure shows an example of behavioral and unit activity on a CR trial recorded in session 5 of acquisition in a rat from Group EtOH (Fig. A) and a rat from Group UC (Fig. B). Figure 3 | Rectified and integrated eyelid EMG activity (top trace), interpositus nucleus activity (middle trace), and tick marks representing action potentials of a discriminated unit (bottom trace) on a CR trial during session 5 of paired eyeblink conditioning in a rat from (A) Group EtOH and (B) Group UC. Rectified and integrated eyelid EMG activity (top trace), interpositus nucleus activity (middle trace), and tick marks representing action potentials of a discriminated unit (bottom trace) on a CR trial during session 5 of paired eyeblink conditioning in a rat from (A) Group EtOH and (B) Group UC. The analysis of changes in firing from baseline to the CS --US interval can be summarized as follows: (1) when a CR was executed, activity increased during the CS --US interval; in the middle of training, Groups SI and UC showed greater increases between 80 and 280 msec after CS onset than did Group EtOH; and (2) when no CR was executed, changes in activity from baseline to the CS --US interval were minimal and did not differ between treatment groups. To examine changes in firing from baseline to the CS --US interval, we analyzed standard scores of neural activity separately for five 70-msec periods during the CS --US interval for acquisition sessions 1, 5, and 10 with a 3 (treatment: EtOH, SI, UC) x 5 (period) repeated-measures ANOVA with treatment as a between-subjects factors and period as a within-subjects factor. Analyses were conducted separately for CR and non-CR trials. On CR trials, there were no differences between treatment groups at the beginning of acquisition (session 1) or at the end of acquisition (session 10). In contrast, a significant period x treatment interaction was observed in session 5, F (8,256) = 2.26, P = 0.024. Examination of the interaction effect using one-way ANOVAs comparing treatment groups in each period revealed differences between treatment groups in all five periods, P's = 0.03 to 0.003. Post hoc Tukey's HSD tests showed that Group EtOH showed significantly less interpositus nucleus activation on CR trials compared with both Group SI and Group UC in periods 2 --4 and less activation compared with Group UC in period 1 and Group SI in period 5 (Fig. A). Similar analyses of non-CR trials revealed only a significant effect of period in session 10, F (4,224) = 3.88, P < 0.01 caused by a significant linear decrease in activation across periods that was similar in all treatment groups (Fig. B). Figure 4 | Standard scores for sessions 1, 5, and 10 of paired eyeblink conditioning as a function of the 70-msec period during the CS --US interval for each treatment group on (A) paired trials with a CR; (B) paired trials without a CR. Standard scores for sessions 1, 5, and 10 of paired eyeblink conditioning as a function of the 70-msec period during the CS --US interval for each treatment group on (A) paired trials with a CR; (B) paired trials without a CR. A standard score of approximately 2.0 represents a significant increase in activity above baseline. Representative examples of mean behavioral and summed unit activity across CR trials in session 5 of paired acquisition for a rat from each treatment group are shown in Figures --. Figure 5 | Mean rectified and integrated eyelid EMG activity (top trace) and summed peristimulus time histograms for two separate units recorded from a Group Ethanol rat on trials with a CR during session 5 of paired eyeblink conditioning. Mean rectified and integrated eyelid EMG activity (top trace) and summed peristimulus time histograms for two separate units recorded from a Group Ethanol rat on trials with a CR during session 5 of paired eyeblink conditioning. Figure 6 | Mean rectified and integrated eyelid EMG activity (top trace) and summed peristimulus time histograms for two separate units recorded from a Group Sham Intubated rat on trials with a CR during session 5 of paired eyeblink conditioning. Mean rectified and integrated eyelid EMG activity (top trace) and summed peristimulus time histograms for two separate units recorded from a Group Sham Intubated rat on trials with a CR during session 5 of paired eyeblink conditioning. Figure 7 | Mean rectified and integrated eyelid EMG activity (top trace) and summed peristimulus time histograms for two separate units recorded from a Group Unintubated Control rat on trials with a CR during session 5 of paired eyeblink conditioning. Mean rectified and integrated eyelid EMG activity (top trace) and summed peristimulus time histograms for two separate units recorded from a Group Unintubated Control rat on trials with a CR during session 5 of paired eyeblink conditioning. The analyses of standard scores were supported by calculation of the proportion of neurons showing significant activation in periods 4 and/or 5 on CR trials but not on non-CR trials during the CS --US interval. These neurons are the strongest candidates for conditioning-related units. Fewer of these potential conditioning-related neurons showed significant activation in sessions 5 and 10 in Group EtOH compared with control groups (Fig. A). Figure 8 | Proportion of neurons showing significant activation in the (A) last 140 msec of the CS --US interval on CR trials but not on non-CR trials, as a function of a paired acquisition session and (B) the equivalent interval on CS-alone trials with an eyeblink but not on trials without an eyeblink, as a function of an unpaired training session. Proportion of neurons showing significant activation in the (A) last 140 msec of the CS --US interval on CR trials but not on non-CR trials, as a function of a paired acquisition session and (B) the equivalent interval on CS-alone trials with an eyeblink but not on trials without an eyeblink, as a function of an unpaired training session. Acquisition: Relationship Between Behavior and Unit Activity | Cross-correlations between eyelid electromyographic (EMG) activity during eyeblink conditioning and interpositus nucleus activity for each paired acquisition session were calculated across all paired trials that were not discarded because of excessive eyelid activity in the pre-CS period. Separate one-way ANOVAs for sessions 1, 5, and 10 comparing treatment groups were calculated using the maximum correlation coefficient as the dependent measure. This analysis revealed a significant treatment effect in session 5, F(2,63) = 5.77, P < 0.01. Post hoc Tukey's HSD tests showed that Group EtOH had a significantly lower cross-correlation coefficient in session 5 compared with both Group SI (P = 0.006) and Group UC (P = 0.030) but Groups SI and UC did not differ. A similar set of analyses conducted using the lead/lag latency of the cross-correlation coefficient (i.e., the latency by which interpositus nucleus activity preceded or followed eyelid EMG activity) as the dependent measure revealed a difference between treatment groups in session 1, F(2,63) = 3.59, P = 0.033. Post hoc Tukey's HSD tests showed that Group EtOH had a significantly shorter amount of time between behavior onset and unit onset compared with Group UC (P = 0.027) in session 1. Figure shows a scatterplot of cross-correlations from sessions 1 and 5 of paired acquisition. Figure 9 | Scatterplot of cross-correlation coefficients and lead/lag latencies during paired eyeblink conditioning in (A) session 1 and (B) session 5. Scatterplot of cross-correlation coefficients and lead/lag latencies during paired eyeblink conditioning in (A) session 1 and (B) session 5. Extinction: Behavior | A 3 (treatment: EtOH, SI, UC) x 4 (extinction session) ANOVA with percentage of CRs as the dependent measure revealed a significant interaction effect, F(6,60) = 2.38, P = 0.039. Further analysis, using separate one-way ANOVAs comparing treatment groups in each extinction session separately, revealed no significant difference between treatment groups in any individual extinction session, (P's = 0.16 to 0.82). Results were similar with CR amplitude as the dependent measure (see Fig. ). Extinction: Unit Activity | To examine changes in firing from baseline to the post-CS interval, standard scores of neural activity for five 70-msec periods after CS onset were analyzed separately for extinction session 1 and 4 with a 3 (treatment: EtOH, SI, UC) x 5 (period) repeated-measures ANOVA with treatment as a between-subjects factors and period as a within-subjects factor. Analyses were conducted separately for CR and non-CR trials. On CR trials, there were no differences between treatment groups at the beginning (session 1) or end (session 4) of extinction, but there were significant effects of period for both sessions, P's < 0.01. Standard scores increased linearly across post-CS periods on trials with a CR and this pattern was similar across treatment groups. On trials without a CR, there were no significant effects in session 1 but, in session 4, there was a significant effect of period, F(4,176) = 4.97, P = 0.001 and treatment, F(2,44) = 4.02, P = 0.025. In the last session of extinction, standard scores increased linearly across post-CS periods on trials without a CR and were larger in Group UC compared with Group EtOH. Unpaired Groups | Training: Behavior | Data from the 18 rats that underwent unpaired training were analyzed. Responses to CS-alone presentations were analyzed with a 3 (treatment: EtOH, SI, UC) x 10 (training session) ANOVA with percentage of eyeblinks as the dependent measure. This analysis revealed no significant effects. A similar analysis with eyeblink amplitude as the dependent measure also revealed no significant effects. These results indicated that nonassociative responses to the tone were no different in the EtOH rats compared with controls. Furthermore, responding to the tone remained stable across sessions. Average percentage of eyeblinks to the tone across groups and sessions was 33.9%, which was almost identical to the average spontaneous percentage of eyeblinks to the tone during adaptation (see Fig. ). Although not statistically significant, we did note that tone-evoked responses were relatively high in the EtOH rats compared with other rats on sessions 1 to 3. The higher level of responding disappeared, however, after the fourth session. Figure 10 | Eyeblink responses in treatment groups that received unpaired training, as a function of training and extinction session. Eyeblink responses in treatment groups that received unpaired training, as a function of training and extinction session. (A) Percentage of eyeblink responses. (B) Amplitude of eyeblink responses. Training: Unit Activity | To examine changes in firing from baseline to the post-tone interval for the unpaired groups, standard scores of neural activity for five 70-msec periods during the post tone interval were analyzed separately for training sessions 1, 5, and 10 with a 3 (treatment: EtOH, SI, UC) x 5 (period) repeated-measures ANOVA with treatment as a between-subjects factors and period as a within-subjects factor. Analyses were conducted separately for trials with an eyeblink response and trials without an eyeblink response. For trials with an eyeblink response, this analysis revealed significant period x treatment interaction effects for session 1, F(8,172) = 5.37, P < 0.001, session 5, F(8,160) = 2.20, P = 0.030, and session 10, F(8,164) = 3.01, P < 0.01. For session 1, examination of the interaction effect using one-way ANOVAs comparing treatment groups in each period revealed significant differences between treatment groups for periods 2 --5, P's < 0.02. Post-hoc Tukey's HSD tests showed that Group EtOH showed a significantly higher standard score compared to Group UC in period 2 (P < 0.01), and compared to both Group SI and Group UC in periods 3 --5 (P's < 0.02). For sessions 5 and 10, examination of the interaction effect revealed no difference between treatment groups in any period, although the F-values varied widely and approached significance for period 1 of session 10 (Fig. A). Figure 11 | Standard scores for sessions 1, 5, and 10 of unpaired eyeblink conditioning as a function of the 70-msec period during the post-CS interval for each treatment group on (A) CS-alone trials with an eyeblink response and (B) CS-alone trials without an eyeblink response. Standard scores for sessions 1, 5, and 10 of unpaired eyeblink conditioning as a function of the 70-msec period during the post-CS interval for each treatment group on (A) CS-alone trials with an eyeblink response and (B) CS-alone trials without an eyeblink response. A standard score of approximately 2.0 represents a significant increase in activity above baseline. The same analyses were performed for tone presentations with no eyeblink response. These analyses revealed no significant effects for session 1 or 5. For session 10, a significant treatment effect was revealed, F(2,41) = 7.54, P < 0.01. Post-hoc Tukey's HSD tests showed that Group EtOH had significantly less change in interpositus nucleus activity from the pre-tone to the post-tone interval compared to Group SI (P = 0.001). (Fig. B). The results of the standard score analysis were supported by the proportion of neurons that showed significant activation in periods 4 and/or 5 on trials with a response but not on trials without a response (see Fig. B). DISCUSSION : Learning and Interpositus Nucleus Activity in Ethanol-Exposed Rats | As in our previous study , rats exposed to binge levels of ethanol as neonates were impaired in acquisition of 350-msec delay eyeblink conditioning as adults. Deficits did not appear quite as large in the current study because the control animals did not perform quite as well as previously. In particular, two rats in Group SI performed exceptionally poorly compared with the other nine rats in Group SI in the current study and the eight rats in Group SI in our previous study. Comparisons between rats that received paired training versus those that received unpaired training in each treatment group showed that rats in Group EtOH showed no differences in performance (except for session 1) whether they received paired or unpaired training, whereas rats in Group SI and rats in Group UC that underwent paired training almost always differed from their counterparts that underwent unpaired training in sessions 4 to 10. Overall, in the current study, as in our previous study, there appear to be learning deficits in adult rats exposed to binge levels of ethanol as neonates in that they always learned eyeblink conditioning more poorly than did unintubated controls. Rats that undergo sham intubation usually perform somewhat more poorly than unintubated controls but better than ethanol-exposed rats, indicating that the stress of the intubation procedure may have some long-term effects on eyeblink conditioning. However, the effects of early exposure to binge levels of ethanol on eyeblink conditioning in adult rats are clearly a greater factor than is the stress of the intubation procedure in producing deficits in adult eyeblink conditioning. The learning deficits in Group EtOH rats do not appear to be due to a simple inability to perform an eyeblink, as indicated by (1) similar numbers of spontaneous blinks in ethanol-exposed and control rats ; (2) similar number of startle responses to the tone in ethanol-exposed and control rats ; and (3) similar eyelid EMG activity to periorbital stimulation in ethanol-exposed and control rats (J.T. Green and J.E. Steinmetz, unpubl.). Also, in agreement with our previous study, there were no significant behavioral differences between treatment groups during unpaired training, although there was some initial heightened responding in Group EtOH. In contrast to our previous study, no differences were noted between groups during extinction training after either paired or unpaired training. In addition to behavioral deficits in the ability to acquire CRs in eyeblink conditioning, adult rats that had been exposed to binge levels of ethanol as neonates showed patterns of interpositus nucleus activity that differed from both sham intubated and unintubated controls. Overall baseline firing rates were similar across treatment groups and were approximately 28 --54 Hz, which accords well with previous studies of extracellular deep nuclear activity in cats and rabbits . However, during paired acquisition, there was a slower-than-normal engagement of interpositus nucleus neurons during eyeblink conditioning in Group EtOH. Although both Group SI and Group UC showed similar and significant unit activation in the CS --US interval by session 5 on trials when a CR was executed, Group EtOH did not show this pattern until the end of acquisition. This delayed pattern of interpositus nucleus activation in Group EtOH indicates that these rats may be capable of eventually exhibiting the level of CRs shown by control groups and further indicates that CS and US input pathways to the interpositus nucleus must be at least somewhat intact. The most likely explanation of the delayed activation of the interpositus nucleus in Group EtOH during eyeblink conditioning is that these rats have fewer interpositus nucleus neurons available to become plastic, although abnormalities in cerebellar cortex, inferior olive, and pontine nuclei cannot be ruled out. It is important to note, however, that these results indicate that the interpositus nucleus remains, functionally as well as structurally, partially intact after exposure to binge levels of ethanol during the brain growth spurt. A cross-correlation analysis between behavior and unit activity across trials further strengthened the conclusion that CR-related interpositus nucleus activity in Group EtOH was slow to develop during eyeblink classical conditioning. Cross-correlations were lowest in Group EtOH in session 5 of paired acquisition, indicating less connection between behavior and interpositus nucleus activity. Analysis of the proportion of units showing significant activation around the time that CRs were executed indicates that this difference in activation patterns was due to fewer units showing significant activation in Group EtOH, rather than to an overall depression of unit activity. This accords well with the fact that Group EtOH has fewer cells in the deep nuclei (J.T. Green, T. Tran, J.E. Steinmetz, and C.R. Goodlett, in prep.) and supports our conclusion that there are still plastic units in the interpositus nucleus of adult rats exposed to binge levels of ethanol as neonates, but they have been reduced in number. Activation of interpositus nucleus units in ethanol-exposed rats during the initial sessions of unpaired training was observed when eyeblinks were executed. This heightened activation disappeared later in training, but there was still evidence of some activation of interpositus nucleus units on trials with an eyeblink during unpaired training in all treatment groups. Although the interpositus nucleus does not show activation in the absence of CRs in rabbits , it is important to note that rabbits rarely blink at all during unpaired training. In contrast, rats (much like humans), appear to blink spontaneously quite often. Given the activation of deep nuclear neurons during eyeblinks , it is not surprising that we observed some activation during unpaired training. However, in contrast to paired acquisition, unit activation during unpaired training was generally lower and did not increase across sessions. Thus, these units may have been coding the eyeblink itself, but did not appear to be showing learning-related plasticity. Possible Mechanisms for Differences in Interpositus Nucleus Activity Between Ethanol-Exposed and Control Rats | There are several possible reasons (which are not mutually exclusive) for why we observed delayed learning-related activity in the interpositus nucleus of rats exposed to binge levels of ethanol as neonates: (1) loss of interpositus nucleus cells; (2) loss of Purkinje cells; and (3) loss of inferior olivary cells. One possible explanation for our results is that rats exposed to binge levels of ethanol as neonates had fewer interpositus nucleus cells available to become plastic. Recently, we have shown that adult rats have fewer deep nuclear neurons after early exposure to binge levels of ethanol (J.T. Green, T. Tran, J.E. Steinmetz, and C.R. Goodlett, in prep.). In the current study, ethanol-exposed rats did not develop strong conditioning-related activity in the interpositus nucleus during eyeblink conditioning until late in acquisition. If eyeblink conditioning normally engages a population of interpositus nucleus neurons and some of this population is missing, conditioning-related activity of individual neurons of the population may be reduced. This reduced conditioning-related activity may not be enough to consistently generate a CR. The fact that we saw some conditioning-related activity in ethanol-exposed rats agrees with the finding that these rats do learn during eyeblink conditioning, but not as well as controls. Thus, ethanol-exposed rats have some interpositus nu- cleus neurons remaining that can become plastic, but the number is significantly reduced compared with control rats. We were able to find similar numbers of units across treatment groups but we were unable to find similar numbers of conditioning-related units across treatment groups. A second possible explanation for our results is that rats exposed to binge levels of ethanol as neonates had fewer Purkinje cells available and that this directly affected interpositus nucleus activity. Numerous studies have shown that early exposure to binge levels of ethanol reduces Purkinje cell numbers in a dose-dependent manner, with severe loss at high binge doses . The interpositus nucleus receives inhibitory afferents from Purkinje cells. It is possible that there were not enough Purkinje cells to properly shape interpositus nucleus activity for generation of CRs in ethanol-exposed rats. Damage to cerebellar cortex before training, in the absence of damage to the interpositus nucleus, is known to slow the rate of eyeblink conditioning in rabbits with permanent or reversible lesions , rats with fewer cortical cells because of neonatal treatment with an antimitotic agent , and mutant mice that lose all of their Purkinje cells as preweanlings . Although it is possible that damage to the interpositus nucleus in ethanol-exposed animals is secondary to loss of Purkinje cells in causing deficits in eyeblink conditioning, it seems unlikely that damage to the critical substrate would not play a key role in slower eyeblink conditioning. Fewer Purkinje cells is likely to play a role secondary to loss of cells in the interpositus nucleus in the deficits we observed. It may be the case that additional interpositus nucleus units would become plastic if a full complement of Purkinje cells were available, perhaps through control of plasticity by Purkinje cells at mossy fiber-to-interpositus nuclear cell synapses . A third possible explanation for our results involves loss of inferior olivary input to both Purkinje cells and interpositus nucleus cells. Rats exposed to binge levels of ethanol as neonates have been shown to have fewer inferior olivary cells compared with controls . Furthermore, this loss appears to have consequences on Purkinje cell activity. Spontaneous complex spiking in Purkinje cells (which is caused by inferior olivary input) is significantly diminished in anesthetized rats that were exposed to ethanol as neonates compared with controls . In contrast, simple spiking remained normal, indicating that the mossy/parallel fiber input to Purkinje cells (and therefore pontine nuclear input) may remain normal in ethanol-exposed rats. Thus, rats exposed to binge levels of ethanol as neonates may show less learning-related activity in the interpositus nucleus during eyeblink conditioning because of loss of direct inferior olivary input to the interpositus nucleus and indirect inferior olivary input to the interpositus nucleus via Purkinje cells. This inferior olivary input is critical for learning eyeblink conditioning, because it conveys US-related activity to the cerebellum. Damage to the inferior olive, before training, in the absence of other damage, prevents eyeblink conditioning in rabbits . Given that ethanol-exposed rats do learn eyeblink conditioning, although not as well as controls, it is likely that at least some of the critical inferior olivary cells that process the US remain and that they can relay that information to both Purkinje cells and to the interpositus nucleus. However, this information may be less than complete and could potentially be a major cause of reduced conditioning-related activity in the interpositus nucleus in adult rats exposed to binge levels of ethanol as neonates. Conclusions | The current study showed that rats exposed to binge levels of ethanol as neonates are impaired in learning eyeblink conditioning. In addition, CR-related neural activity in the critical site of plasticity for eyeblink conditioning, the ipsilateral interpositus nucleus of the cerebellum, does not develop as quickly in these rats compared with controls, indicating that the interpositus nucleus may be only partially functional in ethanol-exposed rats. Future studies will examine neural activity during eyeblink conditioning in other areas of the eyeblink conditioning circuitry of ethanol-exposed rats, including Purkinje cells, inferior olivary cells, and lateral pontine nuclear cells. Collectively, these studies should provide a detailed description, at the neural level, of the impact of early exposure to ethanol on the developing cerebellum in the form of the long-term consequences on cerebellar-dependent learning. MATERIALS AND METHODS : Subjects | Long-Evans male and female breeders, housed in the Indiana University --Purdue University at Indianapolis vivarium, were placed together on designated evenings at approximately 1700 h and allowed to mate overnight. Gestational day (GD) 0 was defined as the first morning that sperm was detected on a vaginal smear after an overnight mating. GD 22 was designated as PD 0, and >80% of births occurred on GD 22. Rats born on GD 21 were considered PD 0 the following day and rats born on GD 23 were considered PD 1 on that day. Litters were culled to eight pups, four males and four females when possible, on PD 1. Pups were identified by a code using a permanent tattoo on one or more paws, produced by subcutaneous injection of a small amount of nontoxic black ink (performed on PD 3, before group assignment). A total of 66 rats (31 males; 35 females) derived from 25 litters were included in this study. Other rats from these litters were used in other studies. Neonatal Treatment Groups | On GD 26 (PD 4), pups of a given litter were randomly assigned to one of three treatment groups. Two males and two females were assigned to Group EtOH and an additional two males and two females were assigned to Group SI. These eight pups, all given intubation treatments, comprised a litter that remained with their birth dam. Separate litters, also culled to eight pups, comprised Group UC. Pups in Group EtOH and Group SI were intubated three times a day from PD 4 to PD 9, with the intubations separated by 2 h (typically, 10:00, 12:00, and 14:00 h). The first two daily intubations for Group EtOH delivered a solution of 11.9% (v/v) ethanol in milk formula in a volume of 0.02778 mL/g body weight ; the third intubation consisted of milk formula only. Group SI pups were also intubated three times daily, but no formula was ever infused. This choice of not matching infusion calories in Group SI was derived from our previous observation that giving extra calories to nonintoxicated controls resulted in growth curves that were accelerated relative to Group UC . Pups remained with the dam until weaning on PD 21. At that time they were marked with an ear punch for litter identification. Between 60 and 80 d of age, rats were transported in their home cages by vehicle to the Indiana University --Bloomington campus (less than a 90-min drive), and housed thereafter in the Indiana University --Bloomington Department of Psychology vivarium. Determination of Blood Alcohol Concentrations | Blood samples were collected from a tail clip 2 h after the second ethanol intubation on PD 6. Blood was sampled from all Group EtOH and Group SI pups, but not from Group UC pups; only the Group EtOH samples were kept for analysis. BACs were determined using the Analox GL5 Analyzer (Analox Instruments). Twenty microliters of tail blood was collected into a heparinized capillary tube and dispensed into a microcentrifuge tube. Samples were centrifuged, and plasma was separated and either analyzed immediately or frozen at -20C for later analysis. The Analox Analyzer used an oxygen-sensitive electrode to measure the rate of oxygen consumption resulting from oxidation of ethanol in the sample (by the ethanol oxidase reagent provided by Analox). The Analox Analyzer was calibrated (within the range of linearity for expected sample concentrations) with a known alcohol standard before each use. The alcohol concentration of each sample was calculated by comparison to the contemporary standard. Reliability of the Analox Analyzer was periodically checked by testing multiple samples of a known concentration of ethanol; variation typically was within similar3% of the target concentration. Surgery | When rats were at least 90 d old, they were prepared for eyeblink conditioning. Rats were anesthetized using intramuscular (IM) injections of an anesthetic cocktail (2.0 ml/kg) consisting of physiological saline, ketamine (37.0 mg/ml), xylazine (1.85 mg/ml), and acepromazine (0.37 mg/ml). Anesthesia was maintained with additional IM injections of a second anesthetic cocktail (0.15 ml) consisting of ketamine (5.0 mg) and xylazine (10.0 mg). Each animal was surgically prepared with differential EMG recording wires, bipolar periocular stimulation US wires, and a pair of insulated microelectrodes. The EMG wires for recording activity of the external muscles of the eyelid, the orbicularis oculi, were constructed of two strands of ultra-thin (0.003") Teflon-coated stainless steel wire passed subdermally to penetrate the skin of the upper eyelid of the left eye. Bipolar stimulation wires were constructed from the same Teflon-coated wire and were positioned subdermally immediately caudal to, and dorsocaudal to, the left eye. Microelectrodes were constructed from stainless steel rods that were etched to a fine tip and insulated with Epoxylite. Tip impedance was brought to 1.5 --4.0 MOmega (at 1 kHz) by passing a small amount of current between the electrode tip and a counter electrode . A pair of microelectrodes was stereotaxically placed in the region of the left interpositus nucleus (AP: -2.3 relative to the interaural line; ML: +2.3; DV: -5.7). A ground wire was connected to two stainless steel skull screws. All of the wires were attached to a 10-pin Augat-style connector and fixed into a cap of dental cement. The wound was salved with antibiotic ointment, and the animals were given at least 1 wk to recover before the start of the training procedures. Apparatus | Eyeblink conditioning took place in an operant box within a sound-attenuating chamber. A fan provided background noise of approximately 65 --70 dB sound pressure level (SPL). Stimulus delivery was controlled by an IBM PC-compatible computer running custom software . Recording of behavioral and neural activity was controlled by a computer interfaced with a Micro 1401 data acquisition unit and running Spike2 software (CED). The 10-pin Augat socket cemented to the rat's skull carried two leads from the upper left eyelid for recording EMG activity, two leads for delivering periocular stimulation, and two leads from the cerebellar microelectrodes. The rat was plugged into a 10-channel commutator, which carried leads to and from peripheral equipment. The EMG signal was amplified 1000x, bandpass filtered at 100 --1000 Hz, and rectified and integrated before being passed to a computer. The neural signal was passed through a JFET configured as a source follower, amplified 10,000x, bandpass filtered at 500 --5000 Hz and passed to the same computer. A 450-msec, 2800-Hz, 90-dB SPL tone delivered from an overhead speaker within the chamber served as the CS. A 100-msec periocular stimulation (60 Hz, 5-msec pulse width) was delivered through two of the commutator leads and served as the US. The amount of current delivered (mean of 1.7 mA) was adjusted for each subject to elicit a clear eyeblink that was sometimes accompanied by a slight head turn. Post hoc analyses revealed that there were no differences between treatment groups in the intensity of the periocular stimulation delivered as a US for training. Conditioning Procedures | Mean age at testing across all rats was 154 days (range = 96 to 217 days) and there was no difference between treatment groups in age at testing. Because we were interested in the effects on the adult cerebellum of PD exposure to ethanol, we did not try to limit age at testing to a narrower range. Experimenters were blind to the neonatal treatment of the animal. Before training, all rats underwent one session of exposure to the training stimuli. During this session, 20 presentations of a single pulse of periocular stimulation (0.1-msec pulse width) were delivered followed by 20 presentations of the tone CS. On the day following exposure to stimuli, rats began either paired or explicitly unpaired training. In a typical trial of paired training, the onset of the tone CS preceded the onset of the 100-msec coterminating periocular stimulation US by 350 msec. However, on every tenth trial, a CS-alone trial was delivered, which allowed inspection of any potential long latency responses without contamination because of the presence of the US. Rats given paired training received 100 trials per day for 10 days. Trials were separated by 20 --40 sec. Following the last day of paired training, 4 d of 100 CS-alone trials per day were given to examine extinction of the learned response. For rats that underwent explicitly unpaired training, each trial consisted of the presentation of either the tone CS or the periocular stimulation US in a pseudorandom order. Rats given explicitly unpaired training received 100 presentations of the tone CS intermixed with 90 presentations of the periocular stimulation US per day for 10 days. Trials were separated by 10 --20 sec. Following the last day of unpaired training, 4 d of 100 CS-alone trials per day were given to compare with similar trials given to rats that had undergone paired training. Histology | Following the last day of training, rats were killed with sodium pentobarbital (120 mg/kg). A small dc electrolytic lesion (100 muA, 20 sec) was made by passing current through the electrodes. Rats were perfused with 0.9% saline followed by 10% formalin, and the brain was extracted and stored in a 10% sucrose/30% formalin solution for at least 1 wk. Before sectioning, brains were embedded in albumin --gelatin and frozen. Using a sliding microtome, frozen coronal sections were taken at 80 mum. The tissue was stained with cresyl violet (for cell bodies) and Prussian blue (for iron deposits left by the marking lesions) and coverslipped with Permount. Behavioral Data Analysis | Eyelid EMG activity was sampled for 1050 msec on each trial. Each trial was divided into three periods: (1) a pre-CS period, 350 msec before CS onset; (2) a CS --US period, 350 msec between CS onset and US onset; and (3) a post-US period, 350 msec after US onset. Trials in which an eyeblink occurred during the pre-CS period that was >50% of maximum eyelid closure were labeled as trials with excessive spontaneous movement and discarded. Percentage of discarded trials across all rats was 5.5 --14.2%, depending on the session. There were no significant differences between treatment groups in percentage of discarded trials during either paired or unpaired training. An eyeblink within 80 msec after CS onset that was greater than the mean activity during the pre-CS period + 5 SDs was scored as a reflexive startle response to the tone. An eyeblink 81 --350 msec after CS onset that was greater than the mean activity during the pre-CS period + 5 SDs was scored as a CR. Across all trials, percentage of CRs and amplitude of eyeblinks during the CS --US period served as the dependent measures of learning. Unit Data Analysis | Neural activity was sampled for 1050 msec on each trial. Offline separation of individual units was done with the threshold discrimination and template-matching algorithms of the Spike2 system (CED). For spike separation, the maximum percentage amplitude change for a match was 20%, and the percentage of a spike that had to lie within a template for a match was 70%. Threshold for spike detection was set to two times the noise threshold. Using this system, 1 --3 units per session could be clearly separated from the multiple-unit signal. Following spike separation, behavioral and unit data were binned (bin size = 2.832 msec), and analyzed using custom software . Data were separated according to trial type (CS only, US only, or paired) and behavior type (CR, non-CR), and analyzed independently. For analysis of increases and decreases in neural activity during trials, the 350-msec period between CS onset and US onset was divided into five periods of 70 msec each. For each trial, five difference scores were calculated by subtracting the mean activity for the entire pre-CS period from the mean activity for each of the five post-CS periods. For a session, five CS-period standard scores were formed by dividing the mean of the corresponding difference score by the standard error of the corresponding difference score. Criterion for a significant increase or decrease in activity in a particular CS period was 1.96 across sessions. Average unit firing frequency during the pre-CS and the CS ---US interval was also calculated. Behavior-Unit Relationship Analysis | Cross-correlations were conducted for each session between mean eyelid EMG activity and summed unit activity across all good trials of a session. For each cross-correlation, CS period neural data was lined up with pre-CS period behavioral data, a correlation coefficient was calculated, and neural data was shifted one bin. This process was repeated until pre-CS period neural data was lined up with CS period behavioral data. In the current study, there were 120 bins in the CS period, so that cross-correlations were calculated from a lag of -120 to a lag of +120. The lag of the largest correlation coefficient, when multiplied by the bin size, indexed the time by which neural activity preceded or followed behavior. Backmatter: PMID- 12359840 TI - Age-Related Impairment in the 250-Millisecond Delay Eyeblink Classical Conditioning Procedure in C57BL/6 Mice AB - In this study we tested 4-, 9-, 12-, and 18-month-old C57BL/6 mice in the 250-msec delay eyeblink classical conditioning procedure to study age-related changes in a form of associative learning. The short life expectancy of mice, complete knowledge about the mouse genome, and the availability of transgenic and knock-out mouse models of age-related impairments make the mouse an excellent species for expanding knowledge on the neurobiologically and behaviorally well-characterized eyeblink classical conditioning paradigm. Based on previous research with delay eyeblink conditioning in rabbits and humans, we predicted that mice would be impaired on this cerebellar-dependent associative learning task in middle-age, at similar9 months. To fully examine age differences in behavior in mice, we used a battery of additional behavioral measures with which to compare young and older mice. These behaviors included the acoustic startle response, prepulse inhibition, rotorod, and the Morris water maze. Mice began to show impairment in cerebellar-dependent tasks such as rotorod and eyeblink conditioning at 9 to 12 months of age. Performance in hippocampally dependent tasks was not impaired in any group, including 18-month-old mice. These results in mice support results in other species, indicating that cerebellar-dependent tasks show age-related deficits earlier in adulthood than do hippocampally dependent tasks. Keywords: Introduction : The eyeblink classical conditioning paradigm is a powerful tool for studying learning and memory because it is well-characterized on a neurobiological and behavioral basis. In addition, the sensitivity, reliability, and generalization across species that eyeblink conditioning has shown is remarkable. There are more published studies on rabbits and humans tested in eyeblink classical conditioning than on any other form of Pavlovian conditioning. Furthermore, dramatic parallels exist between the effects of normal aging in rabbits and humans in eyeblink conditioning. Given the short life expectancy of mice, the fact that the mouse genome is mapped, and the availability of transgenic and knock-out mouse models of neurodegenerative diseases related to aging, the mouse is an excellent species for expanding knowledge on age effects in eyeblink classical conditioning. The delay eyeblink classical conditioning procedure is frequently used in conditioning studies. In the delay procedure, a neutral stimulus such as a tone conditioned stimulus (CS) is presented shortly before a blink-eliciting unconditioned stimulus (US), such as a puff of air directed to the eye. The two stimuli then briefly overlap and coterminate. After many pairings, the organism associates the stimuli and learns to blink to the tone CS in anticipation of the airpuff US. This behavior is characterized as a conditioned response (CR). The neural circuitry underlying formation of CRs has been almost entirely delineated, and striking parallels have been observed between several mammalian species (; ; ,). The essential site responsible for acquisition and retention resides in the cerebellum, ipsilateral to the conditioned eye . The cerebellum receives converging CS and US information, and develops learning-related neural activity during conditioning. Lesions of cerebellar cortex disrupt learning . Although not essential for delay eyeblink conditioning, the hippocampus can modulate the rate of learning in delay conditioning, increasing or slowing the rate of learning . Age differences in the cerebellum and hippocampus with normal aging are associated with deficits in eyeblink conditioning. Age-related impairments in eyeblink conditioning are documented in humans , rabbits , cats , rats , and mice . Additional serious impairment in eyeblink conditioning, beyond that observed in normal aging, has been observed in older adults with Alzheimer's disease (AD) . Furthermore, the 400-msec delay procedure may detect AD at a preclinical phase . Thus, the eyeblink conditioning paradigm is a fundamental tool for studying age- and neurodegeneration-related changes in the brain. Nonhuman mammalian models of eyeblink conditioning, with life expectancies shorter than the 76.9-year human life expectancy, provide a means to focus and experiment with processes of aging. The rabbit model has been used extensively. However, rabbit life expectancy is 8 years, far exceeding that of rodents. With the rapid evolution of mouse genetics, mouse models have gained increased attention in the neurobiology of aging . The short lifespan of mice and the extensive knowledge base on mouse genetics are among the qualities that make this species an excellent animal model for investigation of learning, memory, and aging. Techniques for testing rats, including neonatal rats, have been recently created , along with techniques to test eyeblink classical conditioning in normal, mutant, and transgenic mouse strains . Thus, the mouse model has recently emerged as a promising direction for future study. If aging mice are to be tested in the eyeblink conditioning paradigm, it is imperative that age parallels also be drawn between mice and other species. Using reproductive capacity as a biomarker, mice aged 2 months can reproduce and are roughly comparable to 4- to 6-month-old rabbits and adolescent humans. Likewise, an 8-month-old mouse begins to show decline in reproductive capacity and is comparable to an 18-month-old rabbit or a middle-aged human. On the basis of results in rabbits and humans, it is likely then that age-related impairments in eyeblink conditioning will begin in mice shortly after 8 mo. In the first published study of eyeblink conditioning in aging C57BL/6 mice, reported age-related impairment in the 252-msec delay procedure in 20-month-old mice. The result of impairment only in very old mice is not completely parallel with studies in rabbits and humans. Age differences are observed in rabbits in the 750-msec delay eyeblink conditioning procedure by 24 mo ---the age when reproductive capacity begins to decline (for summary, see ). The longer 750-msec CS-US interval is significantly more difficult for both young and older rabbits. Age differences in the delay procedure at the shorter and less difficult CS-US interval, 250 msec, occur at a later age in rabbits ---36 months . Based on life expectancy, a 20-month-old mouse is comparable to an 80-month-old rabbit. However, we observed age-related deficits in the 250-msec delay procedure in 36-month-old rabbits. Extrapolating rabbit life expectancy to mouse life expectancy, data from our laboratory indicate that mice should show age-related impairment in the 250-msec delay eyeblink conditioning procedure around the age of 9 mo. To examine fully age-related impairments in mice, we have developed a test battery to compare young and older mice across multiple behaviors. These behaviors include the acoustic startle response, prepulse inhibition, rotorod, and the Morris water maze. Acoustic startle response is a test that has been used previously to examine hearing abilities in mice . Several studies have pointed out that middle-aged C57BL/6 mice begin to show presbycusis, and that this age-related hearing loss can be profound . The acoustic startle response and prepulse inhibition are measures that we use to ensure that auditory acuity is not a confound in our eyeblink classical conditioning measures of learning and memory that use an auditory CS. Prepulse inhibition is also a measure that provides insight into mechanisms of sensorimotor gating, alertness, information processing, and attention . These processes are affected in normal aging and AD, and prepulse inhibition in aging C57BL/6 mice shows decline around 18 months of age . Rotorod is a motor learning and coordination task for which the cerebellum is essential. Purkinje cell loss with age reduces cerebellar volume. Because deficits in the cerebellar-dependent eyeblink conditioning paradigm begin to appear in middle age, and because these deficits are associated with a reduction in cerebellar volume, it is useful to compare performance decline in these cerebellar-dependent tasks. The Morris water maze is a hippocampally dependent spatial learning task. The hippocampus is affected profoundly in AD, and it is necessary to examine the integrity of the hippocampus in normal aging. We predicted that age differences in the cerebellar-dependent eyeblink conditioning and rotorod tasks would occur at an earlier age than do age differences in the hippocampally dependent Morris water maze task. The prediction for age differences in acoustic startle and prepulse inhibition performance was that they should occur at an intermediate age between the age at which eyeblink conditioning and rotorod deficits appear and the age at which Morris water maze deficits appear. Taken together, these results with eyeblink conditioning and other behavioral tasks will help to integrate eyeblink classical conditioning data into the extensive body of research literature on aging mice. RESULTS : Age Differences in Eyeblink Classical Conditioning | Paired CS and US data were first collected for 10 days from 9- and 12-month-old groups. These preliminary data were hand-scored by a trained technician who observed the mouse in the conditioning chamber and scored CRs when an eyeblink response to the tone CS occurred before a head twitch response to the shock US. Because the responses were scored by direct visual observation in the short 250-msec CS-US interval, response latencies were not assessed. It was not possible to correct the responses scored as CRs with short-latency responses, and it was also difficult to eliminate all of the bad trials that included eyeblinks before CS onset in the hand-scored analysis. A 2 (age group) x 10 (training session) repeated- measures analysis of variance (ANOVA) determined that age differences in percentage of CRs were not significant (F[1,12] = 3.56, P = 0.083). There was a significant effect of acquisition of CRs over trials (F[9,108] = 8.04, P < 0.0001; Fig. ). Figure 1 | Percentage of conditioned responses (CRs) in mice aged 9 (n = 7) and 12 (n = 7) months over ten 90-paired trial sessions in the 250-msec delay eyeblink classical conditioning procedure as assessed by hand-scoring of CRs. Percentage of conditioned responses (CRs) in mice aged 9 (n = 7) and 12 (n = 7) months over ten 90-paired trial sessions in the 250-msec delay eyeblink classical conditioning procedure as assessed by hand-scoring of CRs. Hand-scored data indicate that mice aged 9 and 12 months acquire CRs at a similar rate. CRs are hand-scored by observing blinking to a 10-kHz tone conditioned stimulus (CS) just before a head twitch response to the 0.5-mA shock unconditioned stimulus (US). Paired CS and US data were then collected for 10 d from 4- and 18-month-old groups, and for 5 d from a different group of 12-month-old mice. Data in the explicitly unpaired condition in which the CS and US are presented independently were also collected from a 4-month-old group for 5 d. These data were scored by a computer as described in the Materials and Methods section. To determine if our hand-scoring method was comparable to the computer-scoring methods, we compared five sessions of acquisition data in 12-month-old mice tested with the hand- or computer-scoring methods. A 2 (scoring method) x 5 (training session) repeated-measures ANOVA revealed no significant scoring method, training session, or interaction effect, indicating that our hand-scoring and computer-scoring methods were roughly comparable for assessing percentage of CRs. Rapid acquisition occurred in 4-, 12-, and 18-month-old mice over five training sessions, although the young mice acquired CRs more rapidly than did the older mice. Using the dependent measure of percentage of CRs, a 3 (age group) x 5 (training session) repeated-measures ANOVA was conducted. This analysis revealed a significant main effect for training session (F[4,84] = 11.37, P < 0.001), and a significant interaction between age group and training session (F[8,84] = 2.94, P < 0.01; Fig. ). The age group effect was not significant. Using the dependent measure of CR onset latency, a 3 (age group) x 5 (training session) repeated-measures ANOVA was conducted. This analysis revealed a significant interaction between age group and training session (F[8,84] = 2.23, P < 0.05), which occurred because CR onset latency declined over sessions in the 12-month-old group but remained relatively stable in the 4- and 18-month-old groups. The main effect for training session approached significance (P = 0.07), and the age group effect was not significant. Figure 2 | Percentage of conditioned responses (CRs) in mice aged 4 (n = 8), 12 (n = 8), and 18 (n = 8) months over five 90-paired trial sessions in the 250-msec delay eyeblink classical conditioning procedure as assessed by computer scoring of CRs. Percentage of conditioned responses (CRs) in mice aged 4 (n = 8), 12 (n = 8), and 18 (n = 8) months over five 90-paired trial sessions in the 250-msec delay eyeblink classical conditioning procedure as assessed by computer scoring of CRs. CRs are scored by the computer if electromyography activity in the orbicularis oculi muscle exceeds 5 standard deviations above baseline activity between 80 and 250 msec after the CS onset. Using the dependent measure of percentage of short-latency responses (responses occurring between 0 and 80 msec after CS onset that have traditionally been called alpha or startle responses), a 3 (age group) x 5 (training session) repeated-measures ANOVA was conducted. This analysis revealed a significant main effect for training session (F[4,84] = 6.13, P < 0.0001), as well as a significant interaction between age group and training session (F[8,84] = 2.66, P < 0.05). There was not a significant age group effect. Whereas the 4-month-old group showed no difference in short-latency responses across training sessions, both 12- and 18-month-old groups produced significantly more short-latency responses on training session 2 than on training session 1. The number of short-latency responses produced by these groups did not increase significantly beyond training session 2. A 3 (age group) x 5 (training session) repeated-measures ANOVA was conducted on the number of excluded bad trials. There were no significant age group or training session effects, and no interaction between age group and training session was observed. Except in rare circumstances, the number of excluded trials generally remained <10 (10%), with most training sessions excluding <5% of the trials. For the 4- and 18-month-old mice on which there were data for 10 training sessions, a 2 (age group) x 10 (training session) repeated-measures ANOVA revealed a significant training session effect (F[9,126] = 2.98, P < 0.01) and a significant interaction between age group and training session (F[9,126] = 2.31, P < 0.05; Fig. ). The age group effect was not significant. Using the dependent measure of CR onset latency, a 2 (age group) x 10 (training session) repeated-measures ANOVA was conducted. This analysis revealed no significant main effects and no interaction between age group and training session. Nevertheless, mice in both age groups decreased their CR onset latencies numerically from the first to the last training session. CR onset latency for mice in training session 1 was 121.1 and 119.8 msec for young and old mice, respectively. CR onset latency in training session 10 was 115.1 and 111.7 msec for young and old mice, respectively. Using the dependent measure of percentage of short-latency responses, a 2 (age group) x 10 (training session) repeated-measures ANOVA was conducted. This analysis revealed a significant main effect for training session (F[9,126] = 2.67, P < 0.01), as well as a significant interaction between age group and training session (F[9,126] = 2.28, P < 0.05). The age group effect was not significant. Whereas the 4-month-old group showed no difference in short-latency responses across training sessions, 18-month-old mice produced significantly more short-latency responses on training sessions 4, 7, 8, and 9 than on training session 1. A 2 (age group) x 10 (training session) repeated-measures ANOVA was conducted on the number of excluded bad trials. Neither the main effects nor the interaction was significant. Figure 3 | Percentage of conditioned responses (CRs) in mice aged 4(n = 8) and 18 (n = 8) months over ten 90-paired trial sessions in the 250-msec delay eyeblink classical conditioning procedure as assessed by computer scoring of CRs. Percentage of conditioned responses (CRs) in mice aged 4(n = 8) and 18 (n = 8) months over ten 90-paired trial sessions in the 250-msec delay eyeblink classical conditioning procedure as assessed by computer scoring of CRs. Data from young and old mice indicate that young mice show ceiling performance on the first day, whereas old mice do not perform at a comparable level until the second day of training. Because young mice produced such a high percentage of CRs in the first training session, the data were analyzed block by block (groups of nine paired trials/block) over the first training session in a 3 (age group) x 10 (block of nine paired trials) repeated-measures ANOVA. There was a significant effect of age (F[2,21] = 21.51, P < 0.0001). A post hoc analysis using the Tukey honestly significant difference test (HSD) indicated that 4-month-old mice significantly outperformed both 12- and 18-month-old mice on the first day of acquisition (Fig. A). There was also a significant effect of training block (F[9,189] = 2.27, P < 0.02) and a significant age group by training session interaction (F[18,189] = 2.70, P < 0.0001). Given this significant interaction, we performed a one-way repeated-measures analysis for the 4-month-old group, comparing percentage of CRs over the 10 training blocks. There was a significant simple main effect of block within the group of 4-month-olds (F[9,63] = 4.19, P < 0.0001), indicating that learning occurred in this group throughout the first 100 trials of training. Block-by block analysis for training session 2 indicated that there were no group differences on this subsequent training day (Fig. B). Percentage of CRs collected in 4-month-old mice in the paired and explicitly unpaired conditions over five training sessions were compared using a 2 (condition) x 5 (training session) repeated-measures ANOVA. Group differences were revealed (F[1,14] = 21.37, P < 0.0001), indicating percentage of CRs in the paired CS-US condition were significantly greater than were responses in a CR period in the unpaired condition . Figure 4 | (A) Percentage of conditioned responses (CRs) on session 1 in mice aged 4 (n = 8), 12 (n = 8), and 18 (n = 8) months over 10 nine-paired trial blocks in the 250-msec delay eyeblink classical conditioning procedure as assessed by computer scoring of CRs. (A) Percentage of conditioned responses (CRs) on session 1 in mice aged 4 (n = 8), 12 (n = 8), and 18 (n = 8) months over 10 nine-paired trial blocks in the 250-msec delay eyeblink classical conditioning procedure as assessed by computer scoring of CRs. Block-by-block CR acquisition on the first day of training reveals that older mice acquire CRs considerably more slowly than do young mice. (B) Percentage of CRs on session 2 in the same mice over 10 nine-paired trial blocks in the 250-msec delay eyeblink classical conditioning procedure as assessed by computer scoring of CRs. The second day of acquisition shows that age-related differences in percent CR do not extend beyond the first training session. Figure 5 | Percentage of responses, as assessed by computer scoring, to a 10-kHz tone stimulus in mice aged 4 months (n = 8) who received paired conditioned stimulus (CS)-unconditioned stimulus (US) training in the 250-msec delay eyeblink classical conditioning procedure, and 4 months (n = 8) who received explicitly unpaired presentations of the CS and US. Percentage of responses, as assessed by computer scoring, to a 10-kHz tone stimulus in mice aged 4 months (n = 8) who received paired conditioned stimulus (CS)-unconditioned stimulus (US) training in the 250-msec delay eyeblink classical conditioning procedure, and 4 months (n = 8) who received explicitly unpaired presentations of the CS and US. Eyeblink data from 4-month-old mice shows that mice in the paired CS-US procedure produce significantly more responses than do mice in the explicitly unpaired procedure. Responses during the CR period decreased from day 1 to day 5 in the explicitly unpaired procedure. Age Differences in Rotorod | Two dependent measures, taken from the rotorod task that are often examined as indicators of motor coordination and motor learning are walk time and latency to fall. Walk time is a measure of the time that the mouse is actively walking on the rotating beam. Latency to fall is a measure of the amount time the mouse is able to avoid falling. Because mice may have a tendency to grip the rotorod and passively rotate rather than actively locomote, these two dependent measures may differ. Paired samples t tests were conducted between the daily mean for walk time and latency to fall at each rotation speed. The absence of significant differences indicated that mice spent their time on the rotorod actively walking rather than passively rotating. Because no difference was revealed between walk time and latency to fall, the latter measure was used in the present analyses to assess motor abilities. Figure depicts the rotorod performance of all age groups at both 15 and 25 revolutions per minute (rpm). A 3 (age group) x 3 (training session) repeated-measures ANOVA, analyzing the 15-rpm data set, revealed a significant main effect for age group (F[2,80] = 8.17, P < 0.01) and for training session (F[2,160] = 182.18, P < 0.001), as well as a significant interaction between age group and training (F[4,160] = 5.01, P < 0.01). Examining the main effect of training session within each age group, post hoc tests using the Bonferroni adjustment for multiple comparisons indicated that learning occurred in all groups. The 18-month-old group showed a significant difference in latency to fall between the training sessions, showing longer latencies each session. The 4- and 12-month-old groups both showed significant motor learning between the first and second training sessions, and no difference between sessions 2 and 3. This result most likely indicates ceiling performance in the 15-rpm rotorod task by session 2 in young and middle-aged mice. A Tukey HSD post hoc test determined that the 4-month-old group outperformed the 12-month-old group in session 1, as well as the 18-month-old group in sessions 1 and 2 of the 15-rpm rotorod task. There were no differences between 12- and 18-month-old 15-rpm rotorod performance in any training session . Figure 6 | Rotorod performance of mice aged 4 (n = 28), 12 (n = 27), and 18 (n = 28) months, at 15 or 25 revolutions per minute (rpm), over three test days. Rotorod performance of mice aged 4 (n = 28), 12 (n = 27), and 18 (n = 28) months, at 15 or 25 revolutions per minute (rpm), over three test days. Learning in the cerebellar-dependent rotorod task is impaired in older mice at both 15 and 25 rpm. The 15-rpm procedure is sensitive to deficits found in 18-month-old mice. These mice do not perform similarly to 4 month olds until the third day of training. The 25-rpm procedure is sensitive to deficits found in both 12- and 18-month-old mice. Learning, nonetheless, is expressed by all age groups in both procedures. A 3 (group) x 3 (training session) repeated-measures ANOVA, analyzing the 25-rpm data set, also revealed a significant main effect for age group (F[2,80] = 21.30, P < 0.001) and training session (F[2,160] = 92.81, P < 0.001), as well as a significant interaction between age group and training session (F[4,160] = 2.57, P < 0.05). Examining the main effect of training session within each age group, simple effects tests revealed the degree to which repeated training was effective in improving 25-rpm rotorod performance in each age group. Significant differences were revealed in the 4-month (F[2,79] = 16.78, P < 0.001), 12-month (F[2,79] = 40.23, P < 0.001), and 18-month-old groups (F[2,79] = 23.92, P < 0.001). Post hoc tests using the Bonferroni adjustment for multiple comparisons found that learning occurred in all groups. The 12- and 18-month-old groups showed a significant difference in latency to fall between the training sessions, showing longer latencies each session. The 4-month-old group showed significant motor learning between the first and second test session, and no difference between sessions 2 and 3. This most likely indicates ceiling performance in the 25-rpm rotorod task by session 2 in young mice. A Tukey HSD post hoc test determined that the 4-month-old group outperformed the 12-month-old group on sessions 1 and 2, as well as the 18-month-old group on all training sessions of the 25-rpm rotorod task. There were no differences between 12- and 18-month-old 25-rpm rotorod performance on any training session . A Pearson product-moment correlation was performed between percentage of CRs on the first training day for the 46 mice in the age range of 4 to 18 months that were tested with the 250-msec eyeblink classical conditioning procedure and rotorod. The latency to remain on the rotorod at 25 rpm was positively correlated with percentage of CRs (r = 0.56, P = 0.004). Percentage of CRs was not correlated at significant level with dependent measures from the acoustic startle, prepulse inhibition, or Morris water maze assessments. Age Differences in Acoustic Startle | The intensity of a startle response is measured according to the amount of stabilimeter deflection that a mouse's body movement causes when it is presented with an auditory stimulus. The deflection is transduced into a congruent electrical pulse, which is read into the computer in millivolts. Response intensity is expressed as Vmax, the measure of maximum voltage transduced when movement occurs in the startle chamber. It is possible that the weight of an animal may contribute to the intensity of the response, and therefore, the effect of weight on response intensity must first be considered. Correlations between weight and Vmax were conducted within the three age groups to determine to what extent an animal's weight has contributed to the response intensity evoked by the startle stimuli. None of the correlations approached significance, indicating that weight had little effect on Vmax. A 3 (age group) x 3 (dB level) ANOVA indicated a significant main effect for age group (F[2,76] = 35.42; P < 0.001) and decibel level (F[1,76] = 196.58; P < 0.001), and a significant interaction between age group and startle decibel level (F[2,79] = 27.41; P < 0.001). Post hoc comparisons using the Tukey HSD test revealed differences in responding between groups at all dB levels. The only nonsignificant difference was between the 4- and 12-month-old startle responses to a 110-dB stimulus . Within subjects post hoc analyses using the Bonferroni test revealed that mice in the 4-month-old group showed a significant difference in Vmax between the decibel levels, showing larger responses to louder sounds. In all other age groups, however, response intensity did not necessarily increase as stimulus intensity increased. Although there were large startle responses shown by the 12-month-old group, no significant differences in Vmax were detected between the decibel levels. The 18-month-old group showed no difference in startle between 95 and 110 dB, but a significant increase in Vmax was detected at 125 dB. This indicates a deficit in responding to lower decibel levels in older mice. Figure 7 | Maximum startle response to a 95-, 110-, or 125-dB stimulus in mice aged 4 (n = 24), 12 (n = 27), and 18 (n = 28) months. Maximum startle response to a 95-, 110-, or 125-dB stimulus in mice aged 4 (n = 24), 12 (n = 27), and 18 (n = 28) months. Young mice produce larger startle responses to louder stimuli. This is not necessarily true for older mice. Although 12-month-old mice show strong reactions to the three dB levels, there is no difference between these reactions. Mice aged 18 months produce small startle responses to 95- and 110-dB stimuli, but a significant increase in startle is not observed until the loudest stimulus is presented. The low Vmax values produced by 18-month old mice at 95 and 110 dB may indicate that the mice were not hearing the stimuli well. To examine the possibility that older mice did not hear the stimulus at 95 and 110 dB, paired samples t tests were used to examine differences in Vmax between no-stimulus trials (which record baseline chamber activity without presenting stimuli) and the lower decibel levels. Significant differences were observed between 95 dB and no-stimulus trials (t[27] = 3.47, P < 0.01) and between 110 dB and no-stimulus trials (t[27] = 3.57, P < 0.01). In each case, the stimulus trials evoked a significantly higher Vmax than the no-stimulus trials, indicating that 18-month-old mice heard the stimuli and produced startle responses. Age Differences in Prepulse Inhibition | Percentage prepulse inhibition (PPI) was calculated as follows: [100 -(Prepulse + Startle/Startle Alone) x 100], where "Prepulse + Startle" is the average response in the prepulse trials, and "Startle Alone" is the average response in trials without prepulse stimuli . A 3 (age group) x 3 (prepulse decibel level) ANOVA indicated a significant age group effect (F[2,76] = 25.40; P < 0.001) and a prepulse dB level effect (F[2,152] = 59.53; P < 0.001), as well as an interaction between age group and prepulse dB level (F[4,152] = 15.92; P < 0.01; Fig. ). Post hoc comparisons using the Tukey HSD test examined differences in percent inhibition between age groups. These analyses revealed significant differences between both the 4- and 12-month-old groups and the 18-month-old group for the 80-dB prepulse stimulus and the 85-dB prepulse stimulus. The 90-dB prepulse stimulus produced significant differences between the 12-month-old and the 18-month-old groups only. There were no differences between the 4- and 12-month-old groups at any decibel level of the prepulse stimulus. Figure 8 | Percent inhibition of a startle response as a result of an 80-, 85-, or 90-dB prepulse stimulus being presented shortly before a 125-dB startle stimulus in mice aged 4 (n = 24), 12 (n = 27), and 18 (n = 28) months. Percent inhibition of a startle response as a result of an 80-, 85-, or 90-dB prepulse stimulus being presented shortly before a 125-dB startle stimulus in mice aged 4 (n = 24), 12 (n = 27), and 18 (n = 28) months. A high percent PPI indicates that the prepulse inhibited the startle response. Mice aged 4 and 12 months perform at a similar rate, but 18-month-old mice are impaired. Because startle was not inhibited with an 80-dB prepulse stimulus, the hearing threshold for 18-month-old mice is most likely between 80 and 85 dB. Impairment in attention and sensory gating may be to blame for low inhibition at 85 and 90 dB. Examining the main effect of prepulse intensity within each age group, simple effects tests revealed the degree to which the prepulse decibel level is differentially effective at inhibiting the startle response in each age group. Significant differences were observed in the 12- (F[2,75] = 4.87; P < 0.05) and 18-month-old (F[2,75] = 74.89; P < 0.001) groups. Post hoc tests using the Bonferroni adjustment for multiple comparisons revealed significant differences between all prepulse intensities except 85 and 90 dB in 12-month-old mice, and between all prepulse intensities in 18-month-old mice. Although 4-month-old mice showed inhibition of the startle response across all prepulse intensities, there were no significant differences. The change in PPI across prepulse intensities did approach significance, however, and a general trend was that louder prepulse decibel levels produced greater inhibition. This trend was observed across all age groups. The lack of any response inhibition at the 80-dB level in the 18-month-old mice likely indicates that the oldest mice were unable to hear the prepulse stimulus. Age Differences in Morris Water Task | Hidden PlatformTraining | During acquisition, or hidden platform training, mice were repeatedly introduced to a pool filled with aversively cool water and were required to learn the location of a hidden escape platform using stationary contextual cues. The dependent measure of learning was latency to escape from the water by climbing onto the platform. Figure depicts the mean latencies to escape for each day of hidden platform training. A 3 (age group) x 3 (training session) repeated-measures ANOVA was conducted. A significant main effect of training session was discovered for latency to escape (F[2,142] = 64.97; P < 0.001). There was no main effect for age and no interaction between age group and training session. This indicates that all age groups learned at equal rates, producing shorter latencies to escape over the 3 days of acquisition. Examining the main effect of training session within each age group, significant differences were observed in the 4-(F[2,70] = 8.29; P < 0.01), 12-(F[2,70] = 24.62; P < 0.001), and 18-month-old groups (F[2,70] = 33.99; P < 0.001). Post hoc tests using the Bonferroni adjustment for multiple comparisons found that the 4-month-old group showed significant differences between sessions 1 and 3, the 12-month-old group showed significant differences between all training sessions, and the 18-month-old group showed differences between session 1 and sessions 2 and 3. Figure 9 | Latency to escape from the Morris water maze during sessions 1 through 3 (spatial learning) and sessions 5 and 6 (cued training) in mice aged 4 (n = 21), 12 (n = 26), and 18 (n = 28) months. Latency to escape from the Morris water maze during sessions 1 through 3 (spatial learning) and sessions 5 and 6 (cued training) in mice aged 4 (n = 21), 12 (n = 26), and 18 (n = 28) months. TS indicates training session. Young mice outperform older mice on session 1 of spatial learning, and group differences disappear by session 2. A one-way ANOVA examining cued training indicated significant differences on session 5 (F[2,74] = 4.48, P < 0.05), with the 18-month-old group outperforming the 4-month-old group. This is most likely a result based on the observation that young mice tend to show thigmotaxis, an inclination to circle the pool in search of an external escape rather than an internal one. Probe Trials | After hidden platform training, the escape platform is removed from the pool. Probe trials are used to measure place learning retention. One measure taken from probe trials is the number of times the mouse crosses over the former platform area. This measure is taken from each quadrant, and it is expected that mice will cross over the area in the trained quadrant more often than the same area in the other three untrained quadrants. All three age groups crossed the former platform area in the trained quadrant significantly more times than they crossed the same area in the other three quadrants (4-month: F[3,70] = 7.41, P < 0.0001; 12-month: F[3,70] = 10.41, P < 0.0001; 18-month: F[3,70] = 3.20, P < 0.05), with the exception that 18-month-old mice did not differ between the trained quadrant and one adjacent untrained quadrant (Fig. A). Figure 10 | (A) Morris water maze platform crossings in trained and untrained quadrants during session 4 (probe trial) in mice aged 4 (n = 21), 12 (n = 26), and 18 (n = 28) months. (A) Morris water maze platform crossings in trained and untrained quadrants during session 4 (probe trial) in mice aged 4 (n = 21), 12 (n = 26), and 18 (n = 28) months. In the Morris water maze probe trials, all mice crossed the area in the platform trained quadrant more times than they crossed the same area in untrained adjacent and opposite quadrants. (B) Morris water maze time spent swimming in trained and untrained quadrants during session 4 (probe trial) in mice aged 4 (n = 21), 12 (n = 26), and 18 (n = 28) months. Mice generally spent more time swimming in the trained platform quadrant than they did swimming in the untrained adjacent and opposite quadrants. Another measure of retention is the time spent swimming in each quadrant, regardless of whether or not mice pass over the platform area. Figure B depicts the time spent swimming in each quadrant by the three age groups. Generally, the age group means indicated that mice preferred the former platform quadrant. A 3 (age group) x 4 (quadrant) ANOVA revealed that all three age groups spent significantly more time swimming in the quadrant that formerly contained the escape platform than at least one other quadrant (4-month: F(3,70) = 8.89, P < 0.0001; 12-month: F(3,70) = 4.56, P < 0.01; 18-month: F(3,70) = 4.07, P < 0.05). Visible Platform Training | In this condition, a flag was attached to the platform to mark its location below the surface of the water. A 3 (age group) x 2 (training session) repeated-measures ANOVA comparing latency to escape between groups revealed a significant training session effect (F[1,72] = 22.94, P < 0.0001) and a significant interaction between age group and training session (F[2,72] = 3.91, P < 0.05). The age group effect was not significant. A Bonferroni post hoc test determined that mice aged 4 and 12 months old performed significantly better from session 1 to session 2 of cued training, whereas 18-month-old swimming performance did not change between the two sessions. Sex Differences in Eyeblink Conditioning, Rotorod, Acoustic Startle, Prepulse Inhibition, and Morris Water Maze | Early in the development of a mouse model of eyeblink classical conditioning in our laboratory, we heard from one experienced colleague that female mice conditioned more poorly than did male mice. To reduce costs, we aged only male mice, and our older age groups have few female mice. The only older female mice we tested at the age of 12 months were older breeders, and their data were hand-scored. A 2 (sex) x 5 (training sessions) repeated-measures ANOVA indicated that there was not a significant sex difference or sex by training session interaction between the three males and four females tested with this method. The 4-month-old computer-scored group had eight males and eight females, but they were dispersed through the paired and explicitly unpaired conditions. A 2 (sex) x 10 (training sessions) repeated-measures ANOVA on the three males and five females tested in the paired condition indicated that there was not a significant sex effect or sex by training session interaction in percentage of CRs in this age group. The same result occurred in the analysis of the five males and three females tested in the explicitly unpaired condition. Low power may have contributed to our inability to detect sex differences in eyeblink conditioning. Sex differences in 15-rpm rotorod performance were analyzed for all groups. Using the dependent variable of latency to fall, a 3 (age group) x 2 (sex) ANOVA was conducted on data from all three training sessions. The main effect for sex on training session 1 approached significance (F[1,77] = 3.76, P = 0.056), with both the 4- and 12-month-old females outperforming the males. The main effect for sex on training sessions 2 and 3 was not significant, and the interaction between age group and sex was not significant on any training session. Sex differences in 25-rpm rotorod performance were analyzed for all groups. Using the dependent variable of latency to fall, a 3 (age group) x 2 (sex) ANOVA was conducted on data from all three training sessions. The main effect for sex and the interaction between age group and sex was not significant on any training session. Sex differences in acoustic startle response were analyzed for all groups. A 3 (age group) x 2 (sex) ANOVA was conducted on the dependent measure of Vmax for all startle decibel levels. The main effect of sex and the interaction between age group and sex was not significant at any startle decibel level. Sex differences in prepulse inhibition were analyzed for all groups. A 3 (age group) x 2 (sex) ANOVA was conducted on percent prepulse inhibition for all startle decibel levels. The main effect of sex and the interaction between age group and sex was not significant at any decibel level. Sex differences in Morris water maze performance were analyzed for all groups. Using the dependent measure of latency to escape, a 3 (age group) x 2 (sex) ANOVA was conducted on training sessions 1 through 3 (hidden platform training) and sessions 5 and 6 (visible platform training). A significant main effect for sex was observed on training session 1 (F[1,69] = 4.40, P < 0.05), with 12-month-old females outperforming the males. The main effect for sex on training sessions 2 and 3 and sessions 5 and 6 were not significant, and the interaction between age group and sex was not significant on any training session. Sex differences in session 4 probe trial platform crossings in trained and untrained quadrants were analyzed using a 3 (age group) x 2 (sex) ANOVA. There were no significant main effects for sex and no significant interactions between age group and sex. Sex differences in time spent swimming in trained and untrained quadrants were analyzed using a 3 (age group) x 2 (sex) ANOVA. There were no significant main effects for sex and no significant interactions between age group and sex. DISCUSSION : In this study, several behavioral tests, including 250-msec delay eyeblink classical conditioning, were used to examine age differences between 4-, 9-, 12-, and 18-month-old C57BL/6 mice. The hypothesis that mice >9 months old (middle age) would be impaired in cerebellar-dependent tasks such as 250-msec delay eyeblink classical conditioning and rotorod was supported. A second hypothesis, that mice <24 months would not be impaired in the hippocampally dependent Morris water maze, was also supported. These results parallel results from other species tested on delay eyeblink classical conditioning and also on hippocampally dependent behaviors and show that aging in behaviors dependent on the cerebellum occurs earlier than aging in behaviors dependent on the hippocampus. Results also indicate that the mouse is a useful model for investigating age-related changes in learning and memory. Contributions from the Model System of Eyeblink Classical Conditioning | Although the research reported here is purely behavioral, we are confident that the mouse model of eyeblink classical conditioning will have utility in elucidating neurobiological aging processes underlying behavioral aging changes. Richard Thompson's model of the neural circuitry involved in CS-US association and development of CRs in eyeblink conditioning identifies the cerebellum ipsilateral to the conditioned eye as the essential structure . One working hypothesis regarding age-related impairment in eyeblink conditioning in mammals, including humans is that age-related changes in the cerebellum impair eyeblink conditioning beginning in middle age. Testing four age groups of mice with the 250-msec delay eyeblink classical conditioning procedure, we have shown that mice >=9months show impaired performance in cerebellar-dependent tasks when compared to 4-month-old mice. In the eyeblink conditioning procedure, young 4-month-old mice rapidly developed CRs to the 85-dB 10-kHz tone, whereas older mice aged 9, 12, and 18 months took longer to develop a comparable level of conditioning. It is possible that young mice developed CRs rapidly owing to their excessive sensitivity to a loud high-frequency stimulus. We used a 10-kHz tone because this frequency is in the optimal hearing range of C57BL/6 mice, including older mice . Although hearing loss has been documented in aging mice of this strain, our results from the acoustic startle and prepulse inhibition tests indicate that C57BL/6 mice aged 12 and 18 months heard and responded to an 85-dB auditory stimulus. Eighteen-month-old mice showed startle responses to a 95-dB white noise stimulus, and both 85- and 90-dB prepulse stimuli inhibited startle responses in these older mice. However, an 80-dB prepulse stimulus had no inhibiting effect on the startle responses of 18-month old mice. This result may indicate that the broadband hearing threshold of mice aged 18 months lies somewhere between 80 and 85 dB. Both 4- and 12-month-old groups showed no signs of hearing impairment. Therefore, it is unlikely that differences in eyeblink conditioning performance are simply the result of impaired auditory acuity ---although older mice probably hear the CS less clearly than do younger mice. This conclusion that hearing cannot account for the full magnitude of the age differences in conditioning is apparent in the case of 12-month-old mice that showed no signs of hearing impairment but had impaired eyeblink conditioning. It is also the case that rotorod, a task that has no auditory requirement is impaired in 12- and 18-month-old mice. The discrepancies between our results and the results of may be associated with age differences in auditory acuity in C57BL/6 mice. We used a 10-kHz tone to maximize the CS audibility for older mice. Kishimoto and associates used a 1-kHz tone CS. Both groups tested C57Bl/6 mice in the 250-msec delay procedure using a CS intensity of 85-dB. Results of the two studies are at variance in regard to two aspects: (1) An impairment in conditioning was found in 9- and 12-month-old mice in the present study, whereas Kishimoto et al. found no impairment in 10-month-old mice; and (2) both studies detected reduced eyeblink classical conditioning in old mice, but the 18-month-old mice used in the present study showed a high plateau level of conditioning (similar80% CRs after the second day of training), whereas the 20-month-old mice tested in Kishimoto et al.'s study did not exceed a level of similar30% CRs, even after 7 days of training. Perhaps 10-month-old mice in the Kishimoto et al. study heard the 1-kHz tone CS well, whereas 20-month-old mice did not hear it as well. Data presented in Figure of that study indicate that 4- and 10-month-old mice startled equally (>60% of the time) to the 1-kHz CS, but 20-month-old mice startled less to that CS (40%). We used a 10-kHz tone CS because studies of C57BL/6J mice, including 23-month-old mice, show that their hearing thresholds are lowest for frequencies between 8 and 16 kHz . By the age of 18 months, a 10-kHz CS is clearly more audible than a 1-kHz tone to C57Bl/6 mice. Indeed, our 18-month-old mice were able to attain 80% CRs in training session 2 in our study. With a 1-kHz tone CS, 20-month-old mice in the Kishimoto study produced only 30% CRs after seven sessions. The deficit in conditioning observed in both 12- and 18-month-old mice in the present study is not likely to reflect only changes in hearing acuity. An ANCOVA yielded a significant age effect on percentage of CRs for training session 1 when prepulse inhibition performance was held constant (P = 0.03). Note that the prepulse stimulus used in the present study was a white-noise stimulus and thus does not necessarily reflect the salience of the 10-kHz tone used for eyeblink conditioning. We acknowledge that the influence of age-related changes in hearing acuity on eyeblink conditioning in aging mice requires further investigation. Nonassociative factors, including hearing acuity, account for some of the age differences in our eyeblink conditioning data. However, there are some age-related associative learning effects present when the nonassociative effects are partialled out. Furthermore, in the case of rotorod performance, there are clear age effects independent of hearing acuity. An explanation for the observed age differences in these two cerebellar-essential tasks is age-related changes in the cerebellum, the structure essential for eyeblink classical conditioning and rotorod performance. With regard to eyeblink conditioning, this explanation has received support in other aging mammals such as rabbits (,; ) and humans , with observations of a high correlation between the integrity of Purkinje cells and cerebellar volume and performance on eyeblink conditioning. Undertaking a comparative study of aging using the brains of 47 species, including mice, observed age-related decline in the number of Purkinje cells. Age-related changes in the cerebellum may account for the age-related deficits in conditioning and rotorod performance in mice. Motor systems, including the cerebellum, are essential in tasks such as the rotorod, a task that shows impairment in transgenic mice with abnormalities of motor neurons and cerebellar Purkinje cells . Transgenic mice that loose all cerebellar Purkinje cells within the first two months of life perform significantly worse in eyeblink conditioning than their do wild-type litter mates . Purkinje cell loss or cerebellar cortical aspirations do not prevent the acquisition of CRs. Rather loss of the cerebellar cortical machinery normally engaged in acquisition of CRs slows the rate of acquisition . Challenges Posed by the Mouse Model of Eyeblink Classical Conditioning | There is significant potential in the mouse model of aging and eyeblink conditioning to contribute to an understanding of neurobiological and behavioral mechanisms of learning, memory, and aging. However, the mouse model of eyeblink conditioning and aging is challenging to implement. Technical Problems: Activity Level, Size, and Electrode Continuity | The high activity level of mice precludes restraint tactics that are effective in rabbits and makes it necessary to implant stimulating and recording electrodes into the obicularis oculi muscles so that mice can move freely during the conditioning procedure. The small size of mice makes such surgeries difficult. Once the surgery is complete and the mouse has recovered, implanted electrodes are not 100% effective or headstages do not remain attached for the entire 10-day duration of the conditioning sessions. After 2 years of improving on our surgical skills and techniques, we have achieved a successful surgical-electrode efficacy of 75% and headstage maintenance rate of nearly 100%. This means that we still lose some data on valuable mice we have aged for 8 to 18 months. Measurement Problems: Associative Learning Versus Learned Fear Responses | Mice responded at a short latency (less than 80 ms) to the 10-kHz tone CS, making it difficult to score eyeblink conditioning records because the duration of what was scored as a startle response based on results from other species frequently extended to the period when a CR could be scored. There were age differences in the number and period in training of the onset of short-latency responses, with young mice being most likely to produce short-latency responses immediately in the first training session and older mice showing short-latency responses beginning in the second training session. We have spent hundreds of hours hand- and computer-scoring mouse eyeblink conditioning records to establish valid procedures and criteria, and we still do not believe that we have resolved all the issues. We have determined that white noise, with its multiple frequencies, may be a more optimal CS than a 10-kHz tone because it evokes less startle in young C57BL/6 mice and is still audible to the older mice. The magnitude and duration of short-latency responses in young mice make it difficult to discriminate between associative learning and emotional responding especially in the short 250-msec period between CS and US. Theorists have proposed that conditioning with aversive stimuli results in at least two processes of learned behaviors: an initial emotional response and, later, an associative response . It has also been argued that processes of learned behavior develop in parallel rather than sequentially . In either case, the critical brain substrates for the two processes are separate. Observations of the two processes in eyeblink classical conditioning studies indicate that conditioned emotional (fear) responses involve pathways mediating pain as well as limbic forebrain structures including the amygdala, whereas learned discrete responses involve the cerebellum and associated brainstem circuitry . Our data in mice indicate that age differences may exist in both processes. Analyses of short-latency responses showed that they were significantly greater in young mice in the first training session and were hardly apparent in older mice at that point. In the second session, these responses increased significantly in 12- and 18-month-old mice. This pattern of an appearance of short-latency responses after 100 training trials is not characteristic of alpha or startle responses that typically habituate within the first training session in rats, rabbits, and humans. Our analysis of these early responses indicated that they were maintained throughout training. In the case of 4- and 18-month-old mice, the early latency responses were maintained for 10 sessions after they appeared in the first session in young mice and after the second session in older mice. We may be mistakenly calling these responses alpha or startle responses. They may be associative responses in mice that respond more rapidly than the other species that have been conditioned. The optimal CS-US interval for mice has yet to be determined empirically for mice, and there is no convention among laboratories for the latency window for startle responses or onset latency for a valid CR in mice. Data from the present study indicate that these parameters will be shorter for mice than they are for rats, rabbits, and humans. Given the high rate of responding, including the high production of short-latency responses in young mice in the first training session, we nevertheless assert that there is significant associative learning in these animals that exceeds associative learning in mice in the 12- and 18-month-old groups. First, although the 4-month-olds do show a high level of responding already on the first day, breaking up the first session into 10 blocks of 10 trials show a progressive increase in percentage of CRs. In the block-by-block analysis shown in Figure , there is a significant simple main effect of block in the first training session within the group of 4-month-old mice (P < 0.0001), indicating that learning occurred throughout the first 100 trials of training. Second, in contrast to the 4-month-old mice that appear to reach plateau level within the first 100 trials, both 12- and 18-month-old mice significantly improve eyeblink conditioning performance on the second day of training (Tukey post hoc tests, P = 0.034 and 0.023, respectively), indicating a somewhat slowed acquisition of CRs in older mice. This increase in CRs from the first to the second day is paralleled by a significant increase in short-latency (<=80 msec) responses in both 12- and 18-month-old mice (P = 0.035 and 0.008, respectively), making a clear attribution of the increase to purely associative learning processes at the traditional CR onset latency difficult. Third, the latency to remain on the rotorod at 25 rpm of training correlates significantly with percentage of CRs on the first day of training. The cerebellum is critical for associative learning as assessed in delay eyeblink conditioning, and the cerebellum is also essential for rotorod performance. The emotional form of learning represented by the startle response uses nociceptive pathways and limbic forebrain structures (for review, see ). Thus, the observed correlation between performance on the two cerebellar tasks supports the validity of the interpretation that associative learning is assessed by our measures of the CR. With all the challenges posed by the aging mouse model of eyeblink conditioning, we continue to pursue this model because of its obvious value. In mice, it is possible to assess a number of behaviors that are much more challenging to assess in rabbits. Sensory gating and attention assessed with prepulse inhibition uses neural circuitry in the forebrain, including the nucleus accumbens and associated hippocampus, prefrontal cortex, and amygdala. One of the most widely used tests of rodent learning and memory, especially in research relevant to aging and AD, is the Morris swim task, for which the hippocampus is essential in C57BL/6J mice . For the investigation of age-related degenerative diseases such as AD, the availability of transgenic mice that over- or underexpress acetylcholinesterase, of transgenic mice that lack specific nicotinic acetylcholine receptor subtypes, and the existence of techniques to switch on and off the transgenic effects are attractive features of mouse models. A hallmark of AD is the presence of neuritic plaques composed of the beta-amyloid peptide (Abeta1 --42) and neurofibrillary tangles composed of tau protein. A number of transgenic mouse models of these neuropathologies now exist. These transgenic mouse models of AD are the closest animal model of the disease that develops only in humans. These unique features of mouse models make the challenges of implementing eyeblink classical conditioning in aging mice appear potentially worthwhile. MATERIALS AND METHODS : Overview | All mice, originally from Jackson Laboratory (Bar Harbor, Maine), were bred and raised in a colony room in the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC)-accredited Central Animal Facility at Albert Einstein Medical Center. The colony room was temperature-and humidity-controlled and ventilated using a dedicated system. Room lighting was timed for a 12-h/12-h light/dark schedule. Mice were grouped by gender at weaning, with <=12 per cage, and housed individually beginning at least 48 h before testing. Housing consisted of a polycarbonate microisolator filtered-top cage. All mice had ad libitum access to sterile food (PMI autoclavable rodent lab diet RHI5010) and water. Research was given approval by Albert Einstein Medical Center's Institutional Animal Care and Use Committee (IACUC). Each mouse was weighed at the beginning of the study. Previous experimenters have found that a fixed sequence of tasks is preferable to randomized test order, as randomization of task order will likely introduce extraneous variability (,; ; ). For this reason, tests were administered in the fixed order in which they are presented in Table . Testing took place Monday through Friday over a five-week period. Table 1 | Sequence of Behavioral Testing, Surgical, and Recovery Procedures for All Mice Eyeblink Classical Conditioning | Subjects consisted of 46 C57BL/6 mice aged to 4 (n = 16, eight males and eight females), 9 (n = 7, seven males and zero females), 12 (n = 15, 11 males and four females), or 18 (n = 8, eight males and zero females) months at the start of testing. These mice were tested in the 250-msec delay procedure and, in the case of half of the 4-month-old mice, in the explicitly unpaired condition. Additional mice for which data on rotorod, acoustic startle reflex, prepulse inhibition, and Morris water maze are reported were tested using long-delay or trace classical conditioning procedures that are not reported here. Animal husbandry and housing is as indicated in the Materials and Methods overview. Surgery | Apparatus | For anesthesia, a nonrebreathing isoflurane administration system was constructed according to the specifications of . The system consisted of a standard isoflurane vaporizer and an oxygen delivery system to deliver the isoflurane. An extension tube delivered anesthesia directly to the mouse. A scavenging tube attached to the evacuation side of the nonbreathing apparatus collected waste anesthetic gasses into an activated charcoal filter (f/airt/m). A chamber was used to induce anesthesia, and two surgical platforms were used so that multiple surgeries could be performed simultaneously. General Procedures | Surgery took place on training days 11 through 13 . Mice were transported on a covered cart to the surgery room. Surgery was conducted under a standard ventilation hood. Anesthesia was induced with O2 + 3% isoflurane at a flow rate of 1 L/min. The mouse was introduced to the induction chamber and was in a surgical plane of anesthesia within 1 min. The isoflurane was then reduced to 2.5% as the mouse was placed on a surgical platform and fitted with a nose cone for anesthesia maintenance throughout the procedure. Ophthalmic ointment was applied to each eye to prevent drying, and mice were covered with gauze strips to maintain normal thermoregulation. Surgical Procedures | Four Teflon-coated stainless steel wires (0.003 in bare, 0.0045 in coated; A-M Systems, Inc.), soldered to a four-pin male header (Jameco Electronics), were implanted intramuscularly in the orbicularis oculi of the left upper eyelid. Wires were stripped of Teflon and carefully placed such that only the muscle-embedded wire was bare. To ensure that the wires did not move or recede back into the periorbital cavity, they were glued to the skull. The two wires most rostral were used to record differential electromyography (EMG) activity, and the two most caudal were used to deliver the eyeblink-eliciting stimulus. When all wires were placed, the four-pin header (headstage) was cemented to the skull and the incision was closed. Recovery | After surgery, mice were given Baytril antibiotic (85 mg/kg subcutaneously) to prevent infection, and Buprenex anesthetic (0.075 mg/kg subcutaneously) for analgesia. Mice usually awoke within 15 min after completion of the procedure, and motor coordination appeared normal within a few hours. Recovery from surgery took place on test days 14 and 15 and the weekend following. Eyeblink Conditioning | Apparatus | The conditioned eyeblink training apparatus consisted of four sound- and light- attenuating chambers (Med Associates, Inc.). Each chamber contained a clear polycarbonate mouse cage, a ventilation fan, and a wall-mounted speaker. A shielded four-conductor wire entered the ceiling of the chamber and was used to deliver a blink-eliciting stimulus to the orbicularis oculi and to record EMG activity. EMG activity was passed through a 300- to 5000-Hz filter and amplified x10 K. The signal was then integrated and digitized before being read into an IBM-compatible system described by for processing. Data were collected in RAM and saved to a hard drive for offline analyses. General Procedures | Eyeblink conditioning took place on test days 16 through 25 . Each training session was controlled by a program written in C++ language and run on an IMB-compatible 386 computer. The intertrial interval was random, ranging from 15 to 30 sec at 1-sec intervals. Mice were tested in groups of four. Each session lasted similar1 h. Testing Procedures | Each mouse was placed in a clean cage within a chamber, and the four-conductor wire was fastened to its headstage. Mice were allowed to move freely around the cage during testing. The ventilation fan remained on and maintained a 70-dB background noise. There were 100 trials presented in blocks of 10. Each block consisted of nine paired trials and one CS-only test trial. Paired trials presented a 350-msec, 85-dB, 10-kHz tone CS, followed 250 msec after its onset by a 100-msec 0.5-mA shock US. It was determined by observation that a 0.5-mA stimulus was sufficient to cause a blink/head jerk in all mice. The CS and US co-terminated. Each session was computer-scored using a Microsoft Excel macro, which analyzed each trial individually for responses. Whenever EMG activity in the orbicularis oculi, recorded in 3-msec bins, exceeded 5 SDs above baseline, a response was considered to have occurred. If a response took place in the first 100 msec before the CS onset, the trial was excluded. For each session, several variables were observed. A startle or short-latency response was scored if the response occurred in the first 80 msec after the CS onset. A CR was scored if a response occurred after the 80-msec startle period and before the 250-msec US onset. On trials in which a short-latency response occurred, a CR could also be scored if it exceeded baseline by 5 SDs and occurred between 81 and 250 msec after CS onset. A UR was scored if no response occurred before the US onset. Rotorod | Subjects | Subjects consisted of 83 C57BL/6 mice aged to 4 (n = 28, 13 males and 15 females), 12 (n = 27, 16 males and 11 females), or 18 (n = 28, 27 males and one female) months at the start of testing. Animal husbandry and housing is as indicated in the Materials and Methods overview. Apparatus | A four-lane motorized rotorod (San Diego Instruments) was used for testing locomotor coordination in mice. Each rod was 3 cm in diameter and 11 cm long, and maintained at 46 cm above the foam-covered base. An electronically controlled motor maintained the rod speed with the speed indicated on the front panel of the control chassis (with a resolution of 0.1 sec). Seven closely spaced photo beams detected the falling of the subject in each lane. The time to fall was shown on the timer display of each lane. General Procedures | Rotorod testing took place on test days 1 through 3 . Each training session consisted of eight trials. Mice were trained at speeds of 15 and 25 rpm (counter-balanced for each age group), with four trials for each speed. There was a 5-min intertrial interval and a 30-min rest period between speed changes. Testing Procedures | On each day, mice were placed on the rotating rod in the orientation opposite to the direction of rotation. Thus, to avoid a fall, subjects were required to locomote forward in a coordinated manner. Trials were discontinued if the subject did not fall within 80 sec. Latency before falling and time spent walking was calculated for each mouse at both rotation speeds. Acoustic Startle and Prepulse Inhibition | Subjects | Subjects consisted of 79 C57BL/6 mice aged to 4 (n = 24, 10 males and 14 females), 12 (n = 27, 16 males and 11 females), or 18 (n = 28, 27 males and one female) months at the start of testing. Animal husbandry and housing is as indicated in the Materials and Methods overview. Apparatus | A two-channel SR-Lab System (San Diego Instruments) was used to test mice for reflexive startle responses to acoustic stimuli. The system included two 35 x 33 x 38.5-cm sound-attenuating chambers that were ventilated and illuminated. The chambers contain a stabilimeter affixed to a clear Plexiglas cylinder (16 x 8.75 cm) mounted to a Plexiglas frame (12.5 x 20.5 x 0.6 cm). The cylinder and frame are elevated 2.75 cm above a 30 x 30 x 4 cm Plexiglas base by four screws stationed under each corner of the stabilimeter frame. A 6-cm speaker, placed 27 cm above the cylinder, delivered acoustic stimuli. Startle responses were transduced by a perizoelectric accelerometer mounted beneath the stabilimeter frame. Output signals were digitized, rectified, and recorded as consecutive 1-msec readings on a Gateway Pentium II computer with San Diego Instruments Windows-compatible Startle Reflex software. Intensity of acoustic stimuli was verified by placing an audiometer (Radio Shack) in the Plexiglas cylinder, with the chamber door closed, and monitoring decibel levels through a viewing lens while running a test session. General Procedures | Acoustic startle response and PPI testing took place on test day 4 . The chamber light and ventilation fan remained on throughout the session. A 75-dB white noise was presented through the overhead speaker to provide continuous diffuse background noise. Each session was controlled by San Diego Instruments software developed for the Microsoft Windows 98 platform. There were a total of 80 trials. The intertrial interval was random, ranging from 10 to 20 sec at 1-sec intervals, and all trials were presented in a pseudo-randomized order such that no two of the same trial-type were contiguous. Mice were tested in groups of two and each session lasted one-half hour. Testing Procedures | Mice were placed in the cylinder for a 5-min acclimation period. Testing began immediately after acclimation. Startle trials consisted of a 40-msec burst of white noise at one of three startle intensity levels (95, 110, or 125 dB). Acoustic startle consisted of a total of 15 trials with an inter-trial interval (ITI) of 15 +- 5 sec. Each startle intensity level was presented five times. Prepulse inhibition testing immediately followed acoustic startle testing. Prepulse trials were 160 msec in length and consisted of a 40-msec presentation of the 125-dB startle stimulus 100 msec after a 20-msec precursor stimulus of white noise at one of three intensities (80, 85, or 90 dB). These prepulse intensities are 5, 10, and 15 dB above the background noise and are each presented 10 times. In addition, 10 no-stimulus trials were presented to record baseline chamber activity, and 10 startle trials (125 dB) were also presented. There were a total of 50 trials presented during the prepulse inhibition testing session. Immediately after the session, 15 more startle trials were presented at the three startle intensities. Four variables are recorded during the test session: Peak amplitude of the startle response (Vmax), average startle response over the 100-msec recording period, the amount of movement at the start of the trial (baseline activity), and the time required to reach peak startle response (latency). With the exception of no-stimulus trials, recording of chamber activity only took place while the startle stimulus was being presented. Vmax was used as the primary measure of startle response. Morris Water Task | Subjects | Subjects consisted of 75 C57BL/6 mice aged to 4 (n = 21, 11 males and 10 females), 12 (n = 26, 15 males and 11 females), or 18 (n = 28, 27 males and one female) months at the start of testing. Animal husbandry and housing is as indicated in the Materials and Methods overview. Apparatus | The training apparatus was a circular pool, 100 cm diameter and 60 cm deep. The pool was located in a laboratory room containing camera and computer equipment, a screen, tables, and other furniture. The interior of the pool was painted white. The water temperature was maintained between 20 and 26C, and the depth was 16 cm. White nontoxic Crayola paint was used to make the water opaque. The hidden platform was a square white tile platform (11cm2) with a surface that was 1 cm below the surface of the water. The visible platform was the same except a white flag (10 x 7 cm) was suspended 15 cm above the platform by a wooden stick. General Procedures | The Morris water task took place on test days 5 through 10 . Around the test room were multiple cues such as furniture and graphic prints placed on the walls. Computer and camera equipment used to record the session were also visible to the mouse. Each trial was started by placing the animal in the water at the edge of the pool in a quadrant either opposite or adjacent to the quadrant containing the platform. The start locations were varied among the three quadrants not containing the platform; with three different start locations being used in each block of four trials. The platform remained in the same location on every trial during the hidden platform task and varied across the four quadrants in the visible platform task. Each trial lasted 60 sec or until the subject located the platform. Subjects that did not find the platform were guided to it and given a latency score of 60 sec. All subjects received a 15-sec rest period on the escape platform between trials. Between blocks of four trials, the mice were placed in individually heated plastic holding cages with paper towels in the bottom to absorb any moisture for at least 30 min. Hidden Platform Training | The mice were started from one of the three designated start locations, which were counterbalanced. Each subject was given three blocks of four trials each (12 trials per day per mouse) for three consecutive training sessions. Mice were given 60 sec to locate the platform. Whether the platform was located or not, each mouse was required to spend 15 sec on the platform between trials. Mice were returned to the holding cage between blocks. On the fourth training session, the subjects were given a probe trial, in which the platform was removed from the pool. After swimming for 60 sec, the mouse was removed from the pool and returned to its holding cage. The training trials and the probe trial were videotaped and recorded using the SMART (Spontaneous Motor Activity Recording and Tracking) program manufactured by Panlab. Several variables were measured during testing. The time to reach the platform (latency to escape) was recorded for each trial and averaged for each block of four trials. The probe trial was analyzed to measure the amount of time spent in each quadrant and the number of crossing made over the platform location in the trained quadrant and the equivalent area in the untrained quadrants. Visible Platform Training | On the fifth and sixth training sessions, all mice were given the visible platform task in which the training was the same as the hidden platform version except the location of the platform and the start position were varied across trials. The latency to escape was recorded. Backmatter: PMID- 12359836 TI - Rapid Adaptation to Auditory-Visual Spatial Disparity AB - The so-called ventriloquism aftereffect is a remarkable example of rapid adaptative changes in spatial localization caused by visual stimuli. After exposure to a consistent spatial disparity of auditory and visual stimuli, localization of sound sources is systematically shifted to correct for the deviation of the sound from visual positions during the previous adaptation period. In the present study, this aftereffect was induced by presenting, within 17 min, 1800 repetitive noise or pure-tone bursts in combination with synchronized, and 20 disparate flashing light spots, in total darkness. Post-adaptive sound localization, measured by a method of manual pointing, was significantly shifted 2.4 (noise), 3.1 (1 kHz tones), or 5.8 (4 kHz tones) compared with the pre-adaptation condition. There was no transfer across frequencies; that is, shifts in localization were insignificant when the frequencies used for adaptation and the post-adaptation localization test were different. It is hypothesized that these aftereffects may rely on shifts in neural representations of auditory space with respect to those of visual space, induced by intersensory spatial disparity, and may thus reflect a phenomenon of neural short-term plasticity. Keywords: Introduction : Since the pioneering studies of , ) and a multitude of psychophysical investigations has dealt with the issue of human cross-modal adaptation (for review, see ; ; ). One of the various adaptive phenomena found in those experiments is the so-called ventriloquism aftereffect. In this perceptual effect, localization of sound sources is systematically shifted after a period of exposure to a consistent auditory-visual spatial disparity. Until now, only few attempts have been made to demonstrate the ventriloquism aftereffect. Adaptation with an auditory-visual disparity implemented by prism lenses has been shown to induce systematic errors in sound localization that were in the same direction as the lateral displacement of vision during the preceding adaptation period . This auditory shift was generally smaller in amplitude than the angle of cross-modal disparity, but could be increased by instructions requiring participants to attend to the visual stimuli during adaptation. The effect occurred regardless of whether participants moved or could see parts of their body. , ) have shown that synchronization of auditory and visual input is a decisive factor, whereas the complexity of the stimulus situation is not relevant for producing the aftereffect. In this respect, the ventriloquism aftereffect differs from the ventriloquism effect (denoting bias of auditory localization by simultaneously presented, spatially disparate, visual stimuli or even perceptual fusion of the two events), which can be additionally increased by cognitive factors, such as the compellingness of a common cause of auditory and visual events (for review, see ). Moreover, it should be noted that the ventriloquism aftereffect also emerges when participants are aware of the spatial discordance of the stimuli and may, consequently, be independent of perceptual fusion with adaptation. The ventriloquism aftereffect was also induced when cross-modal disparity during the adaptation period was implemented by manipulation of auditory spatial cues in such a way that sound directions were laterally displaced (by use of a pseudophone), whereas vision remained uninfluenced. Post-adaptive shifts of auditory localization found in those experiments were opposite to the direction of sound displacement . Most importantly, in a recent study by the ventriloquism aftereffect was shown to occur after mere presentation of synchronized, spatially disparate repetitive sound bursts and flashing light spots in total darkness for a few minutes. Thus, taken together, it seems that it is primarily the temporal correlation of the spatially deviating auditory and visual repetitive events that induces rapid development of a relative post-adaptive shift of representations of auditory and visual space. The origin of this psychophysical effect is still unclear. However, there is evidence from experiments on owls, reared with prismatic spectacles, that both midbrain representations of auditory space and sound localization behaviour are shifted in the direction of visual displacement . These processes may rely on neural mechanisms on the basis of simple Hebbian learning rules . Even though these animal studies refer to long-lasting reorganizations, partially restricted to specific sensitive periods in development (; ; ; for review, see ), the concept that vision calibrates sound localization may also apply to potentially rapidly induced neural plasticity in the human cortex. proposed that such a form of rapid plasticity may be the substrate of the ventriloquism aftereffect. This hypothesis was based primarily on the author's finding that the ventriloquism aftereffect was frequency specific as follows: when pure tones were used as acoustic stimuli during adaptation, the aftereffect fails to appear when sound localization was later tested with tones that differred by two octaves in frequency from the adapting stimuli. This frequency specificity found at the perceptual level could be related to the widths of frequency-tuning functions, measured in single neurons of the monkey primary auditory cortex during perfomance of a sound-localization task. However, conclusions were based on psychophysical experiments with a small number of participants and restricted stimulus conditions. Thus, because of the high theoretical significance of this issue, it was essential to replicate those findings with a larger sample of participants and greater variety of stimulus parameters. Moreover, a number of important questions still remained left open, which the present investigation aimed to address. For this purpose, an experimental design was chosen that was generally similar to that used by , but with some significant differences from that study. First, the present study investigated post-adaptive shifts not only in auditory, but also in visual localization. Second, post-adaptive localization was measured by a task of manual pointing (for review, see ), whereas both a task of head pointing and a task of disparity detection were used by . Third, the present task required participants to attend to the visual stimuli during adaptation, whereas participants in study attended to auditory stimuli. There were four main experiments. In Experiment 1, pulses of broad-band noise were presented 20 to the left or right of synchronized light flashes during adaptation. Pre- and post-adaptive localization performances were compared (Figs. and ). Experiment 2 was a control experiment in which auditory and visual stimuli were presented in spatial alignment during adaptation. Finally, in Experiments 3 and 4, the acoustic stimuli were tone pulses, with either the same frequency in all experimental conditions, or with the frequency during the adaptation period differring from the pre- and post-adaptation conditions. Assuming that neural circuits, composed of sharply frequency-tuned auditory units, are involved in the ventriloquism aftereffect, the expectation was that this aftereffect would be demonstrable only when sound localization is tested with the same acoustic stimuli as were presented during adaptation. Figure 1 | Stimulus positions in Experiment 1. Stimulus positions in Experiment 1. In the adaptation condition (A), sound stimuli were always 20 from visual stimuli. Positions of active loudspeakers and light-emitting diods (LEDs) were varied between trials. Immediately after completion of the adaptation condition, auditory (B) and visual localization (C) were tested by use of a pointing method. Trials with auditory and visual stimuli were presented in an alternating sequence. Figure 2 | Temporal sequence of stimuli and responses in adaptation and localization conditions. Temporal sequence of stimuli and responses in adaptation and localization conditions. (A) In adaptation trials, the spatially disparate sound and light pulses were synchronized. Participants pressed a key as soon as the brightness of the visual stimulus was reduced. (B) In localization trials, participants pointed with a hand pointer toward the azimuthal position of the sound or light source. Participants pressed a key to indicate the final adjustment of the pointer. RESULTS : Experiment 1 | Experiment 1 was designed to measure shifts in auditory and visual localization induced by adaptation to auditory-visual spatial disparity of 20 (Figs. and ). Figure shows individual data for two representative participants. In both, pointing responses to acoustic targets were more to the left after adaptation to +20 disparity (auditory stimuli to the right of the visual stimuli) than those obtained after adaptation to -20 disparity (auditory stimuli to the left of the visual stimuli; paired t-test, P <0.0001). In addition, there were opposite trends for visual localization; responses measured after adaptation to +20 disparity were to the right of those found after adaptation to -20 disparity (P <0.025). In one of the two participants, visual and auditory shifts were similar in magnitude (Fig. A). In the other participant, the visual shift was close to zero, although still significant (Fig. B). Figure 3 | Shifts in auditory and visual localization in two participants (A,B) after adaptation to 20 auditory-visual disparity (Experiment 1). Shifts in auditory and visual localization in two participants (A,B) after adaptation to 20 auditory-visual disparity (Experiment 1). Final pointer positions obtained after adaptation with the auditory stimuli presented to the right of the visual stimuli (+20 disparity) are plotted against the respective positions measured after adaptation with the auditory stimuli presented to the left of the visual stimuli (-20 disparity). Data for auditory (circle) and visual localization were fit by regression lines. The shift of the regression lines for auditory localization in both participants to the right of the diagonal (broken line) indicates that pointing responses to acoustic targets were more to the left after adaptation to +20 disparity than those obtained after adaptation to -20 disparity. A slight opposite trend was found for visual localization. Negative azimuths indicate stimuli to the left, and positive azimuths stimuli to the right. Figure shows mean pointing responses for all participants, plotted as a function of stimulus azimuth. In all experimental conditons, data could be fit by regression lines (r2 >0.989; P <0.0001). In the pre-adaptation condition (Fig. A,D), performance of the participants was similar to that known from previous studies on auditory and visual localization, with characteristic overestimations of both auditory and visual eccentricities (for review, see ; ). In the two post-adaptation conditions (-20 and +20 disparity), displacements of the regression lines calculated from the data suggested that sound localization shifted to the right of the pre-adaptation values after adaptation with -20 disparity (Fig. B) and to the left after adaptation with +20 disparity (Fig. C). For visual localization, adaptation effects seemed to be reversed, although a clear shift was found only after adaptation with +20 disparity (Fig. E,F). Figure 4 | Mean azimuthal angles (+- SE) of the pointing responses for all participants, plotted as a function of stimulus azimuth (Experiment 1). Mean azimuthal angles (+- SE) of the pointing responses for all participants, plotted as a function of stimulus azimuth (Experiment 1). Localization of auditory (A --C) and visual targets (D --F) is shown prior to adaptation (A,E), after adaptation with auditory stimuli to the left of the visual stimuli (-20; B,E), and after adaptation with auditory stimuli to the right of the visual stimuli (+20; C,F). Data were fit by regression lines. Parameters of the resulting function y = ax + b and coefficients of determination for each fit are as given in the panels. Negative angles are to the left, positive to the right. Mean post-adaptive shifts averaged over all stimulus azimuths are shown in Figure . As already suggested by the analyses described above, there were significant systematic shifts in auditory localization to the right (mean + 3.36, SE +- 0.93; paired t-test, P = 0.003) and left (mean -1.41, SE +- 0.64; P = 0.046) after adaptation to auditory-visual disparities of -20 and +20, respectively. Also, a significant shift in visual localization (mean + 1.71, SE +- 0.52; P = 0.006) was induced by adaptation to +20 disparity, whereas only an insignificant shift (mean -0.21, SE +- 0.61; P >0.05) occurred after adaptation to -20 disparity. Even though the amplitudes of auditory and visual shifts appeared to be greater to the right than to left, these differences were not significant (P >0.1). Differences between shifts induced by adaptation to -20 and +20 disparities were significant both in the auditory (mean 4.77; P <0.0001) and in the visual modality (mean 1.92; P = 0.0009). Figure 5 | Mean normalized shifts (+- SE) in auditory (hatched bars) and visual localization (open bars), averaged over all stimulus positions, after adaptation to auditory-visual spatial disparity (Experiment 1; same data as in Fig. Mean normalized shifts (+- SE) in auditory (hatched bars) and visual localization (open bars), averaged over all stimulus positions, after adaptation to auditory-visual spatial disparity (Experiment 1; same data as in Fig. ). Negative shifts are to the left, positive to the right. Asterisks next to bars indicate significant differences of pre- and post-adaptive localization; and asterisks next to brackets indicate significant differences of the respective auditory or visual shifts measured after adaptation to -20 and +20 disparity. Experiment 2 | Experiment 2 was performed to provide a control for Experiment 1. In particular, Experiment 2 examined whether presentation of visual stimuli in one hemifield may have any influence on post-adaptive auditory and visual localization. The main rationale was that possible effects of eccentric gaze direction and spatial attention could not be excluded in Experiment 1. For this purpose, conditions in Experiment 2 were as in Experiment 1, with the single exception that auditory stimuli were presented in spatial alignment with the visual stimuli. That is, during adaptation, paired sound and light pulses were presented together either to the left or right. Unexpectedly, small, but significant, effects were obtained even under these control conditions. After stimulus presentation in the right hemifield, both auditory and visual stimuli were localized more to the left than after stimulus presentation to the left hemifield (paired t-test; auditory: mean difference 2.22, P = 0.037; visual: mean difference 2.10, P = 0.043; Fig. A). The most pronounced absolute shift was found for visual localization after stimulus presentation on the left side (mean shift +2.28, P = 0.014). Figure 6 | Mean normalized shifts (+- SE) in auditory and visual localization after presentation of auditory and visual stimuli that were spatially aligned (Experiment 2). Mean normalized shifts (+- SE) in auditory and visual localization after presentation of auditory and visual stimuli that were spatially aligned (Experiment 2). (A) Shifts measured after stimulus presentation to the right (0 R) and to the left (0 L) of the participant's median plane. (B) Shifts in the same participants as shown in A, obtained after adaptation to auditory-visual spatial disparity in Experiment 1. (C) Differences between normalized shifts measured in Experiments 1 (20 disparity) and 2 (0 disparity) with identical positions of visual stimuli. Conventions are as in Fig. . (Hatched bars) Auditory; (open bars) visual. The results obtained in Experiment 1 from the same seven participants are shown in Figure B for comparison. As in the complete sample of participants (see above), this subgroup exhibited significant post-adaptive shifts in auditory localization (-20 disparity: mean +2.59, SE +- 0.91, P = 0.030; +20 disparity: -2.77, SE +- 0.67, P = 0.006), with a clear difference of 5.36 between shifts obtained in the two adaptation conditions (P <0.0001). Also, shifts in visual localization were in the same direction as those found for the whole group of participants (-20 disparity: mean -0.66, SE +- 0.88; +20 disparity: +0.53, SE +- 0.53), even though this effect did not reach statistical significance (P >0.1). To cancel any effects associated with the asymmetric presentation of the visual stimuli, the differences between normalized shifts measured in Experiments 1 and 2 for identical positions of visual stimuli were calculated. As shown in Figure C, the post-adaptive shifts in auditory localization were even stronger after this correction (-20 disparity: 3.07, SE +- 0.81, P = 0.009; +20 disparity: -4.50, SE +- 1.05, P = 0.005). However, no significant effects were apparent for visual localization (P >0.1). Variation of Spatial Disparity | In addition to the Experiments 1 and 2 that used either 20 disparity, or spatial alignment of auditory and visual stimuli during adaptation, some further measurements were made with adaptation to smaller disparities of 4 and 8. Figure shows the mean normalized auditory and visual shifts after adaptation as a function of spatial disparity for one exemplary participant. There was a significant correlation, indicating an increase of auditory shifts with increasing spatial disparity (Spearman rank correlation coefficient rS = 0.11, P = 0.036). No correlation was found for shifts in visual localization (rS = 0.08, P >0.1). Figure 7 | Mean normalized shift (+- SE) in auditory (circle) and visual localization for one participant, plotted as a function of auditory-visual spatial disparity during adaptation. Mean normalized shift (+- SE) in auditory (circle) and visual localization for one participant, plotted as a function of auditory-visual spatial disparity during adaptation. Results obtained after adaptation with auditory stimuli to the left and right of the visual stimuli are combined. Positive shifts indicate that localization is shifted toward the direction in which the auditory stimuli were presented during adaptation, and negative shifts indicate a shift to that side in which the visual stimuli have been presented during adaptation. Experiment 3 | Whereas in the experiments described above, broad-band noise was used as the standard auditory stimulus, 1-kHz pure tones were used in the localization conditions of Experiment 3. In the adaptation conditions, the auditory stimulus was composed of either 1- or 4-kHz tone pulses that were presented 20 from the visual stimuli. The rationale for this experimental design was to investigate whether or not post-adaptive shifts in auditory localization, such as shown in Experiment 1, also occur when the frequency of the auditory stimuli to be localized differs from that of the stimuli presented during adaptation before. As shown in Figure A, when the stimulus frequencies in the adaptation and localization conditions were identical, the post-adaptive shift in auditory localization was similar to that found in Experiment 1, but larger (cf. Figs. and B). After adaptation to -20 auditory-visual disparity, tones were significantly localized more to the right compared with adaptation to +20 disparity (mean difference 6.17; paired t-test, P = 0.003). The largest auditory shift was found after adaptation to -20 disparity (+4.13, SE +- 1.65; P = 0.046). Only insignificant differences between shifts after adaptation to -20 and +20 disparity were obtained for visual localization (mean difference 0.24; P >0.1). Figure 8 | Mean normalized shifts (+- SE) in localization of 1-kHz-tone and light pulses after adaptation to auditory-visual spatial disparity with either 1-kHz (A) or 4-kHz tone pulses (B). Mean normalized shifts (+- SE) in localization of 1-kHz-tone and light pulses after adaptation to auditory-visual spatial disparity with either 1-kHz (A) or 4-kHz tone pulses (B). Conventions are as in Fig. . In contrast, no significant effects at all were found when stimulus frequencies in adaptation and localization trials were different (Fig. B). After adaptation to 4-kHz tones, localization of 1-kHz stimuli seemed to shift in the opposite direction from that found after adaptation with 1-kHz tones, with an insignificant mean difference of 2.21 between shifts after adaptation to -20 and +20 disparity (P >0.1). Thus, these results generally resemble those obtained in Experiment 2 when auditory and visual stimuli were spatially aligned during the adaptation period (Fig. A). Experiment 4 | Pure-tone frequencies in Experiment 4 were reversed with respect to Experiment 3. That is, localization of 4-kHz tone pulses was tested after adaptation to 20 auditory-visual disparity with either 4- or 1-kHz tone pulses. As in Experiment 3, significant shifts were measured when auditory localization was tested after adaptation with auditory stimuli of the same frequency, but not when adaptation and test frequencies were different. In the first case (Fig. A), the magnitude of the post-adaptive shifts in auditory localization was +5.02 (SE +- 1.24; paired t-test, P = 0.007) for -20 disparity and -6.57 (SE +- 1.90; P = 0.013) for +20 disparity (mean difference 11.59; P = 0.001); and in the latter case (Fig. B), the mean difference of the auditory shifts after adaptation to -20 and +20 disparity was only 0.40 (P >0.1). Also, no significant effects of adaptation on visual localization were found (mean difference for -20 and +20 disparity 0.8, P >0.1; Fig. A,B). Figure 9 | Mean normalized shifts (+- SE) in localization of 4-kHz-tone and light pulses after adaptation to auditory-visual spatial disparity with either 4-kHz (A) or 1-kHz tone pulses (B). Mean normalized shifts (+- SE) in localization of 4-kHz-tone and light pulses after adaptation to auditory-visual spatial disparity with either 4-kHz (A) or 1-kHz tone pulses (B). Conventions are as in Fig. . DISCUSSION : The present results clearly confirm earlier work on the ventriloquism aftereffect by showing a post-adaptive shift in sound localization after presentation of synchronized and spatially disparate auditory and visual stimuli . The direction of this shift tends to correct for the deviation of the sound azimuth from visual positions, thus suggesting a shift in the internal representation of auditory space toward that of visual space. Most importantly, the present study suggests that the mere presentation of synchronized sound bursts and light flashes with consistent angular disparity in absolute darkness is sufficient to induce this aftereffect. Moreover, Experiments 3 and 4 provided strong evidence for the conclusion of that there is no transfer of the aftereffect across frequencies. As will be discussed below in detail, this finding suggests that the aftereffect may be related to rapidly induced coordinate transformations of brain representations of auditory space, formed by neurons that are tuned to specific frequencies. Even though a very small, but still significant, aftereffect with pointing to visual targets was found in Experiment 1 (20 disparity), its occurrence in Experiment 2 (no disparity) suggested that it was not the adaptation to auditory-visual disparity that shifted visual localization . Rather, a different kind of adaptation aftereffect may also have been involved that was induced by the eccentricity of the visual stimuli, which resulted in a slight post-adaptive shift in the representation of visual space opposite to the side of stimulus presentation. The origin of this unexpected effect, which may interfere with the genuine aftereffect of cross-modal disparity, has to be clarified by future experiments. The magnitude of the present post-adaptive shift in sound localization was generally within about one-third of the adapting stimulus disparity, and, thus, roughly resembled that reported by most earlier studies . However, in apparent contrast to those investigations, as well as to the present one, found post-adaptive displacements of apparent sound locations of approximately the same magnitude as that of the adapting cross-modal disparity (8) when localization was measured by a task of head pointing. On the other hand, when in study sound localization was measured by a task that only required detection of spatial disparity, the aftereffect was incomplete, as it was in the present study. Thus, it seems likely that the method of head pointing used by may have involved additional factors that either increased or were superimposed on the genuine ventriloquism aftereffect. In contrast, in the present experiments, involvement of head-motor or head-proprioceptive factors were excluded by fixating the participants' head position. Because the method used here differed in many details from that used by , further discussion of potential reasons for the obvious quantitative discrepancy in the results would be too speculative. In any case, it is clear that it was not the difference in the angle of adaptive auditory-visual disparity that was relevant (see Fig. ). Also, additional experiments with similar30 min exposure to the adaptative stimuli (data not shown) did not give any suggestion of substantial increases of the magnitude of the aftereffect. The present findings may be related to a rapid shift in representations of acoustic space in the human brain, induced by synchronized visual input. This hypothetical shift is presumably restricted to those auditory neuronal circuits that are selective for the frequency of the adaptive sound stimuli. On the other hand, neurons that are tuned to frequencies other than that used for adaption may have remained unaffected by the adaptive stimuli. Thus, the present data may predict the existence of visual input, including spatio-temporal information, to spatially selective and sharply frequency-tuned auditory neurons. Which areas in the human brain are compatible with these predictions? At first glance, one may assume that bimodal neurons, exhibiting spatio-temporal auditory-visual interactions, could be involved in the ventriloquism aftereffect. This type of neuron has been found in many subcortical and cortical areas, in the superior colliculus as well as in frontal, temporal, insular, parietal, and occipital cortex (e.g., ; ; ; ; ; ; ; ); and neuroimaging studies have indicated potential correlates of these interactions in the human cortex (for review, see ). However, sharp frequency tuning (narrower than two octaves), as was predicted by the present Experiments 3 and 4, seems to occur very rarely (e.g., ; ; ). Instead, bimodal neurons are known to generally prefer nonrepetitive, complex noise bursts (; for review, see ; ). This makes it doubtful that this type of neuron actually subserves the present perceptual aftereffect, which appears to be even stronger with pure-tone adapting stimuli. As proposed by , unimodal neurons in the primary auditory cortex could also be involved in the ventriloquism aftereffect. In fact, these cells perfectly match the present predictions of sharp frequency tuning and spatial selectivity . Moreover, there is some evidence from neuroimaging studies that visual information could modulate activity in the human auditory cortex . At present, the possibility that synchronized visual input specifically modulates spatial properties of neurons in auditory cortex is, however, pure speculation. A further potential substrate of the ventriloquism aftereffect may be the inferior colliculus, which is a center of integration of input channels and distribution of output lines, lying in the middle of bottom-up and top-down processing within the auditory pathway . In the inferior colliculus, auditory information is integrated to create topographically organized maps of both frequency and space, formed by neurons exhibiting relatively sharp-frequency tuning and spatial selectivity (; , ; ,; for review, see ). In the barn owl, the topographical representation of auditory space in the external subnucleus of the inferior colliculus homolog has been shown to be calibrated by a visual instructional signal, originating in the visual map of the optic tectum, which arises from topographic projections from the retina . Investigations in mammals have been far less extensive, but mechanisms similar to those in the owl may occur . However, it should be emphasized that these animal studies observed long-lasting neural reorganizations that developed over weeks, whereas the present perceptual adaptation effect emerges within minutes. Thus far, neurophysiological data that could be related to such a rapid plasticity of the processing of sound location in the auditory system are not available, other than a quite remarkable exception. Spatially selective responses to auditory stimuli in the primate inferior colliculus, superior colliculus, and parietal cortex have been shown to be modulated by changes in eye position, suggesting that an eye-position signal (of unknown origin) is used to transform the (originally head centered) auditory spatial cues into eye-centered coordinates with changes in gaze direction . These findings provide evidence that an extremely rapid plasticity of spatial representations within the auditory system also exists. Possibly, short-term (order: <1 sec), intermediate (order: minutes), and long-term plasticities (order: weeks) are based on neural circuits that integrate auditory, visual, and eye-position information, to maintain a stable alignment of auditory and visual spatial representations. The present, intermediate aftereffect could reflect a first, preliminary stage of longer-lasting adaptive processes that are of particular importance during development, when head and pinnae grow and the relation of acoustic directional cues (interaural time and level differences, spectral cues) to the headcentric spatial coordinates changes. In the adult, the functional significance may lie in the maintenance of perceptual space constancy by permanent visual calibration of sound localization. MATERIALS AND METHODS : Participants | A total of 14 volunteers, 10 female and 4 male, with a mean age of 27.8 (range 21 --42) participated in this study. None had any known hearing deficiencies. The vision of all participants was either normal or corrected to normal by glasses or contact lenses. All were naive with respect to the purpose of these experiments. All participants completed the main Experiment 1, and a subgroup of seven participants performed the remaining experiments. All individuals of the latter subgroup first participated in Experiment 1. The sequence of Experiments 2 --4 was balanced across subjects. Initial practice trials on each of the different tasks were given prior to the beginning of the experimental sessions. Apparatus | The participant sat on a chair in an absolutely dark, sound-proof, and anechoic room (5.4 x 4.4 x 2.1 m; ). The participant's head was fixed in a straight-ahead position by a custom-made restraint that consisted of stabilizing rests for the chin, forehead, and occiput. For auditory stimulation, a horizontal array of broad-band loudspeakers (Visaton SC5.9; 5 x 9 cm) was mounted along the arc of a circle (radius 1.5 m) at eye level and centered around the midpoint of the participant's interaural distance. This apparatus consists of 91 loudspeakers, covering an azimuthal range of 180. However, only 11 loudspeakers were used in the present experiments, 1 loudspeaker was straight ahead of the participant, 5 were on the left, and 5 were on the right with constant angular separation of 4. For visual stimulation, one white light-emitting diode (LED; 3 mm, ca. 5 mcd) was attached to the lower edge of the chassis of each of the loudspeakers. The luminance of the LEDs was so low that the participant could not see any details of the experimental apparatus when visual stimuli were presented. A hand pointer was mounted in front of the participant. The swivel pointer consisted of a metal rod that the participant could rotate in the horizontal plane. A key was mounted on the upper side of the rod. The azimuthal angle of the pointer was recorded by a potentiometer as soon as the key was pressed (for further details, see ). Negative angles of pointing are to the left, and positive values are to the right. Procedure for Experiment 1 | Experiment 1 consisted of two sessions, conducted on different days. Each session was subdivided into five blocks that alternately used localization (blocks 1, 3, and 5) and adaptation conditions (blocks 2 and 4; Figs. and ). In block 1 of session 1, the participant was instructed to point with the unseen hand pointer toward either auditory or visual targets (localization condition; Fig. B,C). Block 1 was composed of 98 trials (48 trials with presentation of auditory targets and 48 trials with presentation of visual targets). Trials with presentation of auditory and visual stimuli alternated. The azimuthal positions of auditory and visual targets were varied between trials following a fixed, quasi-random order from -12 to the left to +12 to the right of straight ahead. The auditory stimulus was a sequence of twelve identical sound pulses of band-pass-filtered frozen noise that were presented at a constant rate of 2 s-1 from one of the loudspeakers (cutoff frequencies 0.5 and 8 kHz; sound pressure level 60 dB re 20 muPa; duration 100 msec; rise/fall time 20 msec; Fig. B). The visual stimulus consisted of a sequence of 12 light pulses (duration 100 msec), presented at a constant rate of 2 s-1 from one of the LEDs. Participants were instructed to direct the pointer as accurately as possible toward the perceived auditory or visual stimulus azimuth. As soon as the pointer adjustment was finished, the participant pressed the key on the pointer. At the moment the key was pressed, the stimulus disappeared. One second later, the next trial began. In the cases in which the key was pressed after presentation of the last sound or light pulse, the trial was repeated automatically at the end of the block. The duration of the block was usually similar10 min. After the block was completed, the participant was allowed to rest for similar5 min. In Block 2, auditory and visual stimuli were presented simultaneously with constant spatial disparity (adaptation condition; Fig. A). In each of a total of 50 trials, auditory stimuli were 20 to the left of the visual stimuli; that is, the auditory azimuth was varied between trials following a quasi-random order from -20 to -4 to the left of straight ahead, and visual azimuth was varied from 0 to +16 to the right of straight ahead. The auditory-visual bimodal stimulus consisted of 36 synchronous noise and light pulses (duration of the pulse train 17.6 sec; all other parameters were as in block 1; Fig. A). After 16, 20, 24, 28, or 32 pulses, the luminance of the visual stimulus was reduced by 40%. The moment of the change in luminance varied between trials following a quasi-random order. The participant was instructed to press the same key as used in block 1 as soon as the brightness changed, but was not allowed to move the pointer. Thus, even though eye position was not measured, one may assume that participants fixated the visual stimulus during the adaptation period. This procedure was used only to keep constant fixation and spatial attention to the visual stimulus component, and results of this task were not analyzed further. The duration of each trial was 20 sec. The complete block, consisting of 1800 pairs of sound and light pulses in all, thus lasted about 17 min. No instruction was given with respect to the auditory stimuli. When questioned after completion of the experiment, most participants reported that they were not aware of the spatial disparity between the auditory and visual stimuli. Block 3 was identical with block 1 and began immediately after completion of block 2. Block 4 was presented after a rest of about 5 min. Conditions in block 4 were as in block 2, with the single difference that auditory stimuli were now 20 to the right of the visual stimuli. Auditory azimuth was varied between trials following a quasi-random order from +4 to +20 and visual azimuth was varied from -16 to 0. The final block 5, following immediately after completion of block 4, was identical to blocks 1 and 3. Session 2 only differed from session 1 in that the sequence of the blocks was reversed. That is, in block 2, auditory stimuli were presented 20 to the right of the visual stimuli, whereas in block 4, auditory stimuli were 20 to the left of the visual stimuli. The sequence of sessions 1 and 2 was balanced across participants. Additional measurements were made with some of the participants, using smaller auditory-visual disparities in the adaptation condition (blocks 2 and 4). The position of the visual stimuli was as in the main experiment, but auditory stimuli were either 4 or 8 to the left or right of the visual stimuli. Procedure for Experiment 2 | The main conditions of Experiment 2 were as in Experiment 1. However, auditory and visual stimuli in the adaptation trials (blocks 2 and 4) were always presented in spatial alignment, with the visual stimuli being in the same positions as in Experiment 1. That is, in block 2 of session 1, the azimuthal position of the auditory-visual stimulus pair was varied from 0 to +16 to the right of straight ahead, and in block 4 from -16 to the left to 0. In session 2, the sequence was reversed. Procedure for Experiment 3 | Experiment 3 differed from Experiment 1 by the use of pure tones as the auditory stimuli. In the localization condition (blocks 1, 3, and 5 of sessions 1 and 2), the frequency of the sound stimulus was always 1 kHz (all other parameters were as in Experiment 1). In the adaptation condition with block 4 of session 1 and block 2 of session 2, the frequency was also 1 kHz, whereas in block 2 of session 1 and block 4 of session 2 it was 4 kHz. Thus, localization of 1-kHz tones was tested after adaptation to disparity of visual stimuli and auditory stimuli that were either identical or deviating in frequency by two octaves. Procedure for Experiment 4 | The main conditions in Experiment 4 were as in Experiment 3. However, localization of 4-kHz tones was now tested after adaptation to auditory stimuli of either 4- or 1-kHz frequency. That is, the frequency of the auditory stimulus in the localization conditions, as well as in the adaptation conditions of block 4 of session 1 and block 2 of session 2, was always 4 kHz, and in block 2 of session 1 and block 4 of session 2 it was 4 kHz. Data Analysis | For the main analyses, final pointer positions measured in block 1 of each session were taken as the preadaptation reference values. The differences between these reference values and the post-adaptation measurements of blocks 3 and 5 were calculated. Data for related post-adaptation conditions obtained in different sessions were pooled. Also, data for all stimulus azimuths were collapsed, because relations of pointing responses and target positions were shown to be approximately linear (cf. Figs. and ). Statistical comparisons were made between the resulting mean normalized shifts in azimuthal localization measured for different adaptation conditions. Negative shifts in localization are to the left, positive shifts to the right. Negative spatial disparities indicate that auditory stimuli were presented to the left of visual stimuli during the adaptation period; positive disparities indicate that auditory stimuli were to the right of visual stimuli. Backmatter: PMID- 12359842 TI - Memories in Drosophila Heat-box Learning AB - Learning and memory processes of operant conditioning in the heat-box are analyzed. In a search for conditioning parameters leading to high retention scores, intermittent training is shown to give better results than those of continuous training. Immediate retention tests contain two memory components, a spatial preference for one side of the chamber and a "stay-where-you-are-effect." Intermittent training strengthens the latter. In the second part, memory dynamics is investigated. Flies are trained in one chamber and tested in a second one after a brief reminder training. With this direct transfer, memory scores reflect an associative learning process in the first chamber. To investigate memory retention after extended time periods, indirect transfer experiments are performed. The fly is transferred to a different environment between training and test phases. With this procedure, an aftereffect of the training can still be observed 2 h later. Surprisingly, exposure to the chamber without conditioning also leads to a memory effect in the indirect transfer experiment. This exposure effect reveals a dispositional change that facilitates operant learning during the reminder training. The various memory effects are independent of the mushroom bodies. Keywords: Introduction : Behavioral plasticity is a key to the study of central brain function. It provides access not only to various forms of memory but also to cognitive functions such as attention, context influences , and configural learning . In the fly Drosophila melanogaster, several paradigms of associative learning and memory have been developed in which different behavioral processes are modified . Here we investigate heat-box learning (Fig. ; ), which had been developed for large-scale mutant screening and is one of the simplest and most efficient paradigms. Figure 1 | Schematic diagram of heat-box. Schematic diagram of heat-box. For details see text. Conditioning in the heat-box is an operant process in which flies develop a spatial preference for one side of an experimental chamber. Single flies, walking freely back and forth in a narrow alley in complete darkness, are conditioned to avoid one half of the length of the alley by being heated instantaneously on entering that half. The temporal scheme of heating and cooling simulates a spatial temperature gradient in the chamber for the fly. The training is followed by a test period without any heat. During the whole experiment, the position of the fly in the chamber is monitored, and the fraction of time the flies spent on the "unpunished" side is calculated. Besides temperature, the fly can use only tactile information and path integration for orientation (ideothetic orientation, i.e., the accumulation of the internal representations of the fly's turns and steps; ). In the heat-box, well-known memory mutants like dunce, rutabaga (rut), and amnesiac show reduced performance in the test . Flies mutant for rut were used by to map the structures in the central nervous system requiring normal rut adenylate cyclase for heat-box learning . Candidate structures in the antennal lobes, median bundle, and ventral ganglion were identified. Neither the mushroom bodies nor the central complex requires normal rut expression. Mushroom body --less flies perform as well as normal ones in heat-box learning . A critical step in developing the heat-box paradigm had been to show that performance in the test indeed shows memory. One problem arises from the fact that the test directly follows the conditioning process. At the end of the training, most flies avoid being heated and are therefore found on the unpunished side. With the test starting directly after the training, all these flies contribute positively to the retention score. However, using such a procedure, it is not possible to distinguish between an aftereffect of heat-avoidance and a conditioned preference for the previously unpunished side of the chamber. To avoid this problem had started the evaluation of the position traces for the retention test after the first midline crossing of the fly. This evaluation, however, underestimates the spatial preference component of the memory trace. A further problem addressed in previous studies was potential odor marks: While being heated, flies might deposit odorants and later during the test avoid these. To investigate this possibility, flies were transferred from one chamber to another between training and test. As flies turned out to lose track of the unpunished side during transfer, a 10-sec reminder training was introduced to reestablish after the transfer the polarity of the new chamber with respect to hot and cold. From these experiments, it was concluded that flies indeed learned a spatial preference for the unpunished side . However, an alternative explanation still remains. Instead of remembering from the first training that its position in the chamber can influence the heat, the fly might be conditioned by the experimental situation of the training period (darkness, isolation, etc.) to learn faster during reminder training. Here we explore the memory processes in the heat-box in more detail. In the first part, we compare various training procedures and try to separate the two components of the retention score. We show that the increase in memory with more training is not caused by the associative memory component but by the fraction of flies that stay on the unpunished side after the last encounter with heat ("stay-where-you-are" effect). In the second part, we measure how long the memory persists. To avoid the effect of extinction between training and retention test, we transfer the flies to a different environment (food vial) for that period. The procedure reveals a third memory component which represents conditions of the training other than the heat/position contingency. RESULTS : Part I: Training Procedures and Memory Components | Improved Memory After Intermittent Training | The standard 4-min of training in the heat-box leads to a final avoidance of performance index (PI) = 0.60 +- 0.04 and a 3-min memory score of PI = 0.35 +- 0.03 (Fig. A; see also ). If the training is extended to 12 min, a final avoidance of PI = 0.85 +- 0.02 and a retention score of PI = 0.56 +- 0.02 is obtained (Fig. B). As spaced training in other learning paradigms has been shown to generate a more robust memory , we investigated whether splitting the training session into several cycles of training and intermittent test phases might further increase performance. In Figure C, training consists of six 2-min periods separated by 1-min test phases. Flies of Figure , B and C, were taken from the same batches. Trained intermittently, they show higher PIs during the training and test phases than with continuous training (U test: Z = 4.34, P < 0.001 for final 3-min memory score). Also, memory decay is slightly slower after intermittent training (data not shown). Figure 2 | Continuous (A, B) versus intermittent (C) training in CantonS flies. Continuous (A, B) versus intermittent (C) training in CantonS flies. Performance index (PI) includes 30-sec pretest (black bars), training (densely hatched bars, 2 min each), intermittent test phases (only in C, empty bars, each 1 min), and final test (broadly hatched bars, 3 min). Error bars are SEMs; n indicates number of flies; *P < 0.05, **P < 0.01, ***P < 0.001, as in all subsequent figures. Influence of Cycle Number and Duration of Training | To optimize the retention score, we performed a parametric study of increasing cycle number (from three to five) for short cycles of 2-min training and 30-sec test, as well as for cycles of 4-min training and 1-min test. After training, all six experimental groups received a 3-min retention test. Figure shows that the duration of cycles influences test performance. In all groups, from three to five cycles, long cycles lead to a significantly higher test performance than do short cycles (U tests: three cycles: Z = 4.13, P < 0.001; four cycles: Z = 2.31, P < 0.05; five cycles: Z = 3.02, P < 0.01). In contrast, the number of long cycles does not significantly influence test performance (ANOVA: H = 1.97, P = NS [not significant]). Comparing test performance of short cycles, four cycles lead to a higher test performance than do three cycles (ANOVA: H = 7.18, P < 0.05; U test, four versus three cycles: Z = 2.68, P < 0.01), whereas five cycles give no significantly better result than that of four or three cycles. Our data show that long training/test cycles lead to better test performance than do short ones. Whether this difference is owing to the total training time or to an intertrial-interval effect remains open. In any event, based on the above experiments, we use four 4-min cycles with 1-min intervals for the experiments in Part II below, as they seem to yield asymptotic values. Figure 3 | Different training regimes. Different training regimes. Flies were trained either five, four, or three times with either short cycles of 2-min training/30-sec intermittent test (empty bars), or long cycles of 4-min training/1-min test (hatched bars). Figure shows performance indices of the final 3-min tests for all six groups. Separation of Two Memory Components | In the heat-box, the fly can avoid the punished side because itself can switch off the heat. If the fly has no control of the heat, it does not develop a side preference. This obvious effect can be visualized in a "yoked control" experiment (Fig. A). Flies are treated with seven training/test cycles (2-min of training and 1-min test). One group is able to control heat by its position in the chamber (experimental), whereas in the other group, each fly experiences a temporal sequence of hot and cold generated by one of the flies in the first group, but has no influence on the temperature (yoked control). Experimental flies reach a performance index of PI = 0.84 +- 0.06 in the test after 12 min of training, whereas yoked flies have no positive avoidance or memory scores. Interestingly, flies that have the possibility to control the heat reduce their locomotor activity more than do flies that have no influence on the temperature (Fig. B). Already in the second training/test cycle, the performance value of experimental flies is significantly reduced compared with that of yoked flies (U tests, second training period: Z = 3.10, P < 0.01; second test period: Z = 2.95, P < 0.01). Figure C shows this difference. The standard training procedure with experimental and yoked control groups gives a similar result (data not shown). Figure 4 | Yoked control experiment. Yoked control experiment. Experimental flies have the possibility to control heat punishment. Each fly of the yoked control group gets the same heat regime as a particular fly of the experimental group, independently of its behavior. (A) Performance index (PI) of the experimental group (experimental) with intermittent training versus the yoked control group (yoked control). PI includes pretest (pre, black bars; 30 sec), training (tr, hatched bars; each 2 min), and test phases (te, empty bars; each 1 min). (B) Locomotor activity of experimental and yoked flies (same experiment as in A). (C) Difference in locomotor activity between experimental and yoked control group. Several explanations can account for the additional decrement in locomotor activity in experimental over yoked control animals. One possibility is that experimental flies use activity reduction to avoid the heat. They might learn that with heat off, slow/no walking is a successful strategy (contributing to the stay-where-you-are effect). Another explanation takes the temporal patterns of spontaneous locomotor activity into account . Flies have their individual schedule of activity and rest periods. Activity bouts and pauses are not synchronized between flies. During training, the flies in the experimental group can follow their endogenous temporal pattern with minimal adjustments, whereas in the yoked flies, the heat pulses during rest periods may induce additional activity bouts. As mentioned in the Introduction, the retention test in the present paradigm immediately follows the training phase and is therefore not a pure measure of the fly's preference for one or the other half of the chamber. It includes an aftereffect of heat avoidance at the end of the training period that leaves most of the flies on the unpunished side. The contribution of this effect is difficult to assess directly. A lower estimate of the true spatial memory component can be obtained by starting the memory test for each fly only after the first midline crossing (Fig. ; ). This evaluation excludes flies that after training stay on the unpunished side for the whole test period (stay-where-you-are). Figure 5 | One-minute test scores starting immediately after training (composite; hatched bars) versus conservative estimates (pure; empty bars) after different training regimes (4 min, 12-min continuous, and 12-min intermittent training; same data as Fig. One-minute test scores starting immediately after training (composite; hatched bars) versus conservative estimates (pure; empty bars) after different training regimes (4 min, 12-min continuous, and 12-min intermittent training; same data as Fig. ). Note different numbers of flies in composite and pure scores. After continuous 12-min training, this low estimate during the first minute of the evaluated test phase is PI = 0.37 +- 0.06; after intermittent training, it is PI = 0.41 +- 0.08 . This small difference indicates that most of the retention increment of the intermittent training over the continuous training is caused by an increasing fraction of flies spending the whole test period on the formerly unpunished side. Taking into account that the early part of the test is discarded, we conclude that the spatial choice component accounts for at least half of the total retention score. Part II: Memory Dynamics | Associative Memory After Transfer | Flies were trained intermittently with four cycles of 4 min, removed from their chamber, and immediately transferred to a new chamber where they received a 30-sec reminder training, during which they had to experience heat at least once to be included in the ensemble average (experimental design, Fig. a). In the subsequent 6-min test without heat, they showed a small but significant retention score (Wilcoxon, P < 0.01), as observed before under slightly different conditions . They were compared with control groups of naive flies and to flies that had been kept in the chamber for 20 min just like the first group but without training, at constant low temperature (Fig. ; experimental design, Fig. a --c). Neither the naive group nor the exposed group showed a significantly positive PI in the test. This result shows that after a short training of 30 sec, the stay-where-you-are effect is minimal. In all transfer experiments, we therefore disregarded the stay-where-you-are effect and directly used the retention scores for further evaluation. Figure 6 | Experimental schedules. Experimental schedules. (a --c) Direct transfer: experimental group (a); exposed group, no training (b); naive control (c). (d --g) Indirect transfer with retention period in food vial; experimental group (d); exposed group, no training (e); handling control (flies have only short chamber experience of 1 to 2 sec), no training (f); and transfer experiment with single flies in small plastic vials (g). Figure indicates pretest (pre), training (tr), test (te), and reminder training (rem). Figure 7 | Direct transfer between two chambers. Direct transfer between two chambers. Flies are either trained (trained; densely hatched bar) or just kept in the first chamber for the corresponding time without heat (exposed; broadly hatched bar). After transfer, all flies receive a short training of 30 sec, and finally, their retention is tested for 6 min. Control animals (naive; empty bar) undergo only the reminder training and the final test. Figure shows performance indices (PIs) of the final 6-min tests. To test whether retention scores after transfer are the result of an operant associative learning process or are caused by a motivational change, a yoked control experiment was performed. One group of flies was able to control during training the temperature by its position in the chamber (experimental group), whereas the other group received the heat punishment independently of its behavior (yoked control group). The flies were subsequently transferred to a new chamber in which all of them received a 30-sec reminder training and were finally tested. Yoked control flies had a significantly decreased test performance compared with that of experimental flies (U test: Z = 2.76, P < 0.01; Fig. ). Their retention score was statistically indistinguishable from zero (Wilcoxon, P = NS). Thus, we conclude that positive performance values of flies after transfer are the result of an associative learning process in the heat-box. Figure 8 | Yoked control experiment with direct transfer. Yoked control experiment with direct transfer. Flies of the experimental group can control heat punishment during intermittent training, whereas flies of the yoked control group can not. Immediately after training, flies are transferred to another heat chamber, where they received a 30-sec reminder training and a 6-min test. Both groups can control heat punishment during the reminder training. Figure shows performance indices (PIs) of final 6-min retention tests for the experimental and yoked control groups. Two-hour Memory | How stable is the memory trace? How long does the fly retain the association between its position in the chamber and temperature? To measure the temporal dynamics of retention without extinction training in the time interval between acquisition and test, flies must be kept in a different environment during that period. Flies were transferred after training to a food vial for various intervals (either 1 to 3 min, 30 min, 2 h, or 4 h) and then back into a chamber for reminder training and test (Fig. , squares; experimental design, Fig. d). They showed PIs significantly different from zero for retention intervals of up to 2 h (Wilcoxon: 1 to 3 min, P < 0.01; 30 min, P < 0.001; 2 h, P < 0.05; and 4 h,: P = NS). Figure 9 | Indirect transfer. Indirect transfer. Flies are either trained (squares) or just exposed to the chamber without heat (triangles). Between conditioning and test, flies are first transferred to a food vial for the indicated time and then back to a new chamber where all flies undergo a short training of 30 sec and a final 6-min retention test. Control animals (naive; filled circle) undergo only the 30-sec training and the final test. Each group includes about 200 flies. Figure shows performance indices (PIs) of the final 6-min test phases. Control flies were kept in the chamber without any heat for the same amount of time before the double transfer (20 min; experimental design, Fig. e). Surprisingly, they also showed a significantly positive PI for the 1- to 3-min retention interval, similar to that of the trained flies (Fig. , triangles). The mere exposure to the chamber improves acquisition during the reminder training. This effect lasts only briefly, though. Already for the 30 min retention interval, the test PIs in merely "exposed" flies were significantly lower than those in trained flies (U test: Z = 2.70, P < 0.01) and at 2 h were not significantly different from zero. In naive flies, as shown before , the reminder training in itself did not lead to PIs significantly different from zero (Wilcoxon-matched pairs test, P = NS). Hence, with the double transfer, a further type of aftereffect is observed: a contextual memory relating to characteristics of the situation in the chamber (exposure effect) rather than to the heat/position contingency. It should be noted that without the short intermission in the food vial, this exposure effect is not observed. In the experiment of Figure , for the 1 to 3 min interval retention scores of trained flies are not larger than those of merely exposed flies. Because trained flies are necessarily also exposed, one can ask whether their memory reflects the heat/position contingency or only the situation in the chamber as for the merely exposed animals. To answer this question, again a yoked control experiment was performed that deviated from the yoked experiment in Figure only in that the flies were kept in a food vial for 1 min between conditioning procedure and reminder training. Test performance of the yoked control group was significantly lower than that of the experimental group (U test: Z = 2.01, P < 0.05) and statistically not different from zero (Wilcoxon, P = NS; data not shown), indicating that the experimental flies remember an association between punishment (heat) and behavior from the operant conditioning procedure. The exposure effect seems to be suppressed by the heat in trained and yoked control animals, at least for the 1- to 3-min retention interval. If this applies also for the 2-h retention interval, the 2-h memory of the trained group can also be regarded as a memory of the heat/position contingency. In the following experiments, we address the issue of what is learned during exposure to the chamber without heat. No Contribution of Handling | In the transfer experiments above, each fly is sucked into and blown out of the aspirator three times: at the transfer from the home vial to the chamber, from the chamber to the food vial, and from the food vial to the new chamber. To investigate whether this handling might contribute to the exposure effect, we reduced the period in the chamber to a few seconds (handling control; experimental design, Fig. f). Afterward, flies were treated just like animals of the trained and exposed groups. They stayed in the food vials for 1 min, were transferred back to the new chambers and after the reminder training were tested for 6 min with heat off. Only flies of the training and exposed group showed significantly positive PIs in the final test (Wilcoxon: trained group, P < 0.01; exposed group, P < 0.01; Fig. ). Flies that had received the full handling but had spent only a few seconds in the chamber showed no significant retention (Wilcoxon: handling control, P = NS). Both, trained (ANOVA: H = 7.15, P < 0.05; U test: trained versus handling, Z = 2.35, P < 0.05) and exposed groups ( U test: exposed versus handling, Z = 2.00, P < 0.05) had a significantly higher test performance than that of the handling control. Apparently, handling per se does not contribute to the exposure effect. It is the experience of the 20-min period in the chamber that enhances the effectiveness of the reminder training in building up a memory. Figure 10 | Handling does not cause the exposure effect. Handling does not cause the exposure effect. Before the transfer to the food vial, flies are kept in the chamber for only a few seconds (handling group) but receive the same handling as those in the trained and exposed groups. After 1 min in the food vial, flies are transferred to the chamber where they undergo a 30-sec reminder training and a 6-min memory test. Only final memory scores are shown. Isolation and Chamber Characteristics Contribute to the Exposure Effect | The experiment of Figure indicated that during exposure, the flies learned characteristic features of the chamber, enabling them afterward to acquire the heat/position contingency more readily during the reminder training. This is not the only interpretation, though. With their first transfer to the heat-box, they are separated from their home vials and their sibling flies for the first time in their life. We therefore asked whether the flies during the exposure to the chamber just learned to cope with isolation in a strange environment rather than memorized specific properties of the geometry and material of the chamber. Before the transfer, we kept flies one by one for a 20-min time period in transparent small plastic vials (phi 22.0/63 mm; experimental design, Fig. g). A group of flies exposed to the heat-box before the transfer and a group of naive flies, both from the same culture vials as the experimental animals, served as controls. After the exposure, all groups received the same treatment, in that they were transferred to a food vial, after 1 min were transferred back to a chamber, received reminder training, and were finally tested . Figure 11 | Chamber-specific and chamber-independent components of the exposure effect. Chamber-specific and chamber-independent components of the exposure effect. Flies of the experimental group are exposed to plastic vials (vial) before the transfer. Control groups include flies that are exposed to the heat chamber (exposed) or naive flies (naive). All flies have a 1-min rest period in the food vial before being transferred to the chamber to undergo a 30-sec training and a 6-min memory test. Figure shows the performance indices (PIs) of the memory tests. Flies kept in plastic vials showed significantly smaller PIs in the test than those of the exposed group (ANOVA: H = 29.85, P < 0.001; U test: exposed versus vial, Z = 2.85, P < 0.01), indicating that the flies learned characteristics of the situation in the chamber. Additionally, however, flies kept in vials showed significantly larger PIs than naive flies (U test: vial versus naive, Z = 2.29, P < 0.05), arguing that chamber-independent aspects of the exposure such as isolation may facilitate acquisition during the reminder training. Length of Chamber Is Not Critical | We next investigated whether chamber length was a critical parameter learned during exposure. Flies were kept in chambers of either full, half, or quarter length by using stoppers that filled part of the chambers. After transfer into a food vial for 1 min and then back into the chambers, flies were tested in full-size chambers. If chamber length was learned, we expected a decrement in the test scores of flies exposed to smaller-sized chambers. As Figure shows, this was not observed: There was no significant difference in test performance between the three groups (ANOVA: H = 0.49, P = NS), all of which showed positive PIs significantly different from zero (Wilcoxon: one chamberP < 0.001; one-half chamber, P < 0.01; one-fourth chamber, P < 0.01). We conclude that chamber length is not a critical feature of the memory in the exposure effect. Figure 12 | No influence of chamber length on exposure effect. No influence of chamber length on exposure effect. Flies are exposed to chambers of different length (full, half, or quarter length). After exposure, flies are transferred to the food vial for 1 min and subsequently to chambers of normal size. Figure shows the performance indices (PIs) during the 6-min memory test after the second transfer and 30-sec training. Mushroom Bodies Are Not Required for Training and Exposure Effects | As a first step toward identifying the neural substrate of the training and exposure effects, we investigated whether mushroom body --less flies still showed any of these types of memory. Heat-box learning with the standard procedure is independent of the mushroom bodies , but the transfer experiment used to document the training and exposure effects involves severe context changes (chamber/food vial/chamber) to which in a different learning experiment, flies without mushroom bodies were shown to be more sensitive than were normal control animals . Flies treated as first-instar larvae by hydroxyurea (HU), and flies treated the same but omitting HU (HU controls) were either trained or merely exposed to the chamber. After being transferred from the chambers to a food vial for 1 min, they were transferred back to the chambers for reminder training and test. Brain sections of tested HU flies gave 90% of animals with total loss of postembryonic mushroom bodies. In <10% one tiny mushroom body was left. Neither for the trained group nor for the exposed group were significant differences between HU and HU control flies observed (U tests, P = NS; Fig. ). All groups gave a positive 6-min retention score (Wilcoxon: trained group HU, P < 0.05; trained group HU control, P < 0.001; exposed group HU, P < 0.05; exposed group HU control, P < 0.01). The mushroom bodies that are not necessary for heat-box learning are also dispensable for the associative and nonassociative memories in the transfer experiments. Figure 13 | No requirement of the mushroom bodies for training and exposure effect after indirect transfer. No requirement of the mushroom bodies for training and exposure effect after indirect transfer. Hydroxyurea (HU)-treated and control flies are compared in the indirect transfer experiment for training and exposure for 1- to 3-min retention interval. Figure shows the 6-min memory tests after transfer and reminder training. DISCUSSION : Conditioning in the heat-box can be very effective. After a training of 20 min, flies stay on the previously heat-associated side for only similar10% of the time. This conditioned avoidance is about as strong as that in odor discrimination learning . At closer inspection, however, the two values are not really comparable, as in the heat-box avoidance in the test is only partly caused by the fly's preference for certain locations in the chamber. As the position of the flies cannot be "randomized" between training and test, the high PI of heat avoidance at the end of the training period is carried over into the retention test. One may account for this effect in the data evaluation procedure if one delays the retention test until the fly has crossed the midline. This eliminates flies showing no further midline crossing. In most cases, these flies show extremely low walking activity, perhaps because of a particularly strong conditioned spatial preference or mere heat avoidance. The delayed retention score contains only the spatial preference component but for several reasons underestimates it. This becomes obvious if we consider that flies have the tendency to patrol the length of the chamber. Therefore, even a virtual fly equipped with only the spatial preference memory would have a high probability to be on the previously punished side after the first midline crossing, owing to heat avoidance during training. This bias would diminish the delayed retention score. The low activity effect shows not only in the locomotion data of Figure , in which trained flies are compared with their yoked controls and training leads to a stronger reduction of locomotor activity than heating per se, but also in the number of flies evaluated in Figure . For instance, after the intermittent training only 39 of the 145 flies could be included in the conservative estimate because only those flies crossed the midline within the first two minutes. As expected, a training interrupted by rest periods is more effective than a continuous conditioning phase. In many organisms and learning situations, spaced training regimes with very different temporal patterns are known to improve memory . In Drosophila extended memory spans after spaced training have been documented for odor and visual pattern discrimination learning . For operant conditioning, we could show that intermittent training mainly strengthens the stay-where-you-are effect. This is unexpected, as with intermittent presentation of the reinforcer, the stay-where-you-are strategy should be more difficult to learn, whereas the conditioned side preference should become more robust against extinction training, as is indeed observed for the composite retention score (data not shown). The heterogeneous composition of the memory score must be taken into account in mapping experiments , as well as future genetic and pharmacological analysis. In the transfer experiments, the avoidance at the end of the primary training is irrelevant for the final retention score after the transfer. Because of the high symmetry of the chamber, the fly has no cue as to its position in relation to the potentially heated side after the transfer. Only the 30-sec reminder training that immediately precedes the test and provides the hot/cold polarity may still affect it. The control experiments with naive flies, however, show that the stay-where-you-are effect from the reminder training is negligible. Moreover, after the transfer and the reminder training, locomotor activity is high for all groups (data not shown). We therefore assume the memory scores in the transfer experiments to represent primarily the conditioned side preference. In one of the control experiments with a 1- to 3-min retention interval in the food vial, a new memory phenomenon is discovered, indicating that a reminder training of 30 sec can be sufficient to induce a subsequent retention score provided that the fly is in the right disposition. If in the first phase (what would be the training phase), the fly is kept in the chamber without the heating regime, the transfer back to a food vial and to a group of other flies between training and test is necessary to establish this dispositional state. A direct transfer from the exposure chamber to the test chamber does not. In other words, after the first transfer from the regular food vial and group situation to the narrow dark chamber, the naive fly is not in the right disposition to build up a memory of the spatial distribution of heating periods during the following half minute. If, however, the same transfer occurs a second time, the fly is ready to attend to the contingency between the heat pulses and its own position in the chamber. Thus, the memory of the first transfer and exposure to the chamber disposes the fly favorably for the learning task after the second transfer. It is a well-known phenomenon that preexposure to the training context without reinforcement can facilitate subsequent acquisition . Here, this is only part of the story. The transfer from the group of flies in the food vial to the chamber and the time in the chamber seem both to be relevant because omitting the rest phase in the food vial (direct transfer, Fig. ) and shortening the first stay in the chamber (handling, Fig. ) both abolish the effect. The length of the chamber is not critical , whereas a plastic vial instead of a chamber does not fully serve as an adequate preexposure , perhaps because it is not dark. To fully understand what the fly is learning in the first phase to master the 30-sec learning task in the second phase will require more detailed investigations. With the transfer experiments and yoked controls, we have finally shown beyond doubt that the heat-box records an associative memory. The fly can remember, even 2 h later , that its position in the chamber controls temperature. Acquisition of this memory is an operant process. The fly's discovery that its behavior can modify temperature leads to a lasting modification of the fly's behavior. How the fly modifies its behavior to take advantage of its conditioned side preference remains to be found out. The fly may try to stay close to the "cold" end of the alley, and it may avoid long straight walks or even any locomotion. In any case, the side preference persists independently of the fly's actual position in the chamber. It must therefore be based on a "percept" or "cognitive map" of the chamber, simple as this representation may be. The map may consist of nothing but two antiparallel vectors for the safe and dangerous directions, which the fly maintains irrespective of its own changing position and orientation. To relate heat-box memories to the brain and to other forms of memory in Drosophila, mushroom body --less flies and their controls were included in this study. Flies store memories of odors in their mushroom bodies . In many other forms of learning, flies without mushroom bodies perform perfectly well. These include visual pattern recognition, color discrimination learning, motor learning, conditioned courtship suppression in day light, and learning in the heat-box (for summary, see ). We reinvestigated this problem here because had discovered that the mushroom bodies render visual memories less sensitive to context changes. Our transfer procedure necessitates a context change: the transfer from the chamber to the food vial and back. As it turns out, heat-box memories are sufficiently robust to sustain these context changes, even in mushroom body --less flies. Apparently, different neural circuits underlie the robustness of memories in the visual and ideothetic domains. This result, however, should not surprise. "Context" is a broad concept. Everything besides the conditioned and unconditioned stimuli and the behavior in question might be regarded as the context. In visual pattern recognition at the flight simulator, the part of the context that changes is the quality of illumination. All other aspects of the fly's precarious situation remain the same. In the present transfer experiments, the situation of the fly dramatically changes from ample space, fresh food, light, and company to isolation, confinement, and darkness. The difference of these two types of context change could hardly be more profound. Nevertheless, one has to abandon the idea that the mushroom bodies might support a general mechanism protecting against all kinds of context changes in memory processes. MATERIALS AND METHODS : Flies | The D. melanogaster CantonS (CS) wild-type strain was used in all experiments. Flies were reared on standard cornmeal/molasses medium in a 16-h-light/8-h-dark cycle at 60% humidity and 25C. Adults of both sexes were studied (similar50%) at 2 to 7 d after eclosion. To create animals lacking mushroom bodies, first instar larvae were treated with the cytostatic drug hydroxyurea which leads to ablation of the mushroom body neuroblasts and hence to adult flies lacking mushroom bodies . After behavioral experiments, a sample of 101 out of 518 HU treated flies was controlled for loss of mushroom bodies by paraffin sectioning of brains. Among the 101 flies, 98 completely lacked the mushroom bodies, whereas three had lost one calyx with one tiny calyx left. Experimental Setup | The conditioning apparatus was built in the workshops of the Biocenter and is a modified version of the one used by . It consists of an array of 15 chambers (26 x 4 x 2 mm) operated in parallel, each with Peltier elements on top and bottom allowing for fast heating and cooling . The Peltier elements cover the whole length of the chamber. A control circuit and a thermo sensor keep the chamber at a defined temperature. Glass side walls enable transmission and detection of infrared light from a LED source (invisible to the flies). The fly casts a shadow on a bar code reader (light gate array in Fig. ) on the opposite side of the chamber. The position signal of the bar code reader is sent to the computer with a frequency of 10 Hz. Experiments are performed in complete darkness. Chambers are cleaned with a pipe cleaner every day before experiments. Measurements are performed on at least three days to avoid effects of daily variability. The different groups in one graph are measured strictly in parallel. Standard Experiment | The standard experiment consists of three phases: pretest, training and test. One half of the chamber is defined as the punished and the other as the unpunished side. These designations are altered for every experiment to reduce systematic effects of side use and of potential asymmetries of the apparatus. During the 30-sec pretest, the fly can explore the chamber at a constant temperature of 20C; this provides a measure of experience-independent spatial preference. During the subsequent 4-min training period, the whole chamber is heated to 40C whenever the fly enters the punished side and is cooled down to 20C when it enters the unpunished side. For analysis, the training and test phases are binned into 1- or 2-min blocks and a performance index (PI) is calculated for each block as detailed below. During training, this index provides a combined indicator of heat avoidance and learning or only heat avoidance. In the following 3-min test period, the chamber is constantly at 20C. The performance index (PI) is calculated as the difference between the time the fly spent in the unpunished versus punished half of the chamber divided by the total time. Thus, the PI can range from -1 to 1, with a PI of 0 indicating no side preference. To yield a measure of general activity, the sum of position changes per time is calculated. Transfer Experiment | The temporal sequence of events in the transfer experiments is explained in Figure . Direct transfer (experimental design: Fig. a): During the training period flies are subjected to four cycles of 4 min training and 1 min test. Afterward they are removed from the chamber by gently aspirating them into a pipette tip and immediately transferring them into another chamber where they are again trained for 30 sec (reminder training). During the reminder training the same side is defined as punished side as in the first training period. Subsequently, animals undergo a 6-min memory test. (This procedure differs from that of Wustmann et al. [1996]. They trained the animals for 3 min. After the transfer they applied a reminder training of 10 sec and tested memory only for 1 min.) The control conditions for the transfer experiment are outlined in Figure , b and c. Flies of the control groups are either exposed to the chamber for 20 min without any heat before the transfer (exposed group) or are taken directly from the food vial before "reminder" training (naive group). Both groups undergo a 6 min memory test after the reminder training. Indirect transfer (experimental design: Fig. d --g): Flies are removed from the chambers after training and transferred into a regular food vial (experimental design: Fig. d; phi 36.0/83 mm). All flies of a given experiment are stored together in a vial until they are, one by one, transferred back into the chambers. After returning flies into the chamber all steps are identical to the direct transfer. Control groups are flies that either had been exposed to the chamber for 20 min without any heat before the indirect transfer (exposed group, Fig. e) or naive flies which had neither received training nor exposure (naive group). Both control groups were then trained for 30 sec and tested for 6 min. Analysis of Data | To exclude animals which do not show substantial motor activity or do not experience punishment, the following criteria are established: flies have to walk at least one chamber length and get at least two heat exposures. For transfer experiments, the following additional criteria apply: After the transfer, flies have to walk one chamber length and have to experience at least one heat period to be included in the data set. As tests for normal distribution of performance indices yield varying results, non-paramentrical tests are used for statistical evaluation. Two independent groups are compared by Mann-Whitney U tests. For comparison of three and more groups Kruskal-Wallis ANOVA tests are used. Wilcoxon tests are applied to compare single performance indices to zero. Backmatter: PMID- 12359841 TI - Side-Specificity of Olfactory Learning in the Honeybee: US Input Side AB - In honeybees, Apis mellifera L., the proboscis extension reflex (PER) can be conditioned by associating an odor stimulus (CS) with a sucrose reward (US). As the neural structures involved in the detection and integration of CS and US are bilaterally symmetrical in the bee brain, we ask what respective role each brain side plays in the conditioning process. More specifically, the US normally used in conditioning experiments is the compound stimulation of the antennae (which triggers the PER) and of the proboscis (where bees lick the sucrose solution). Anatomically, the brain receives unilateral US input through each antenna, but bilateral input from the proboscis. By controlling each US component, we show that an antenna --US produces unilateral sensitization, whereas a proboscis --US or a compound --US induces bilateral sensitization. Bees can learn a unilateral odor CS with all three USs, but when a proboscis --US is used, new learning is inhibited on the contralateral side, owing to a possible US-preexposure effect. Furthermore, we show that the antenna --US induces both unilateral and bilateral reinforcement processes, whereas the proboscis --US produces only bilateral effects. Based on these data, we propose a functional model of the role of each brain side in processing lateralized CSs and USs in olfactory learning in honeybees. Keywords: Introduction : Associative learning requires the anatomical and functional convergence of the signal to be learned with the reinforcing signal. If the input sides of the two signals are lateralized, one might ask whether the associative connections are established independently on both sides or whether sufficient cross-talk exists between the two sides of the nervous system to lead to a memory trace independent of the respective input side. In particular cases, for example, visual imprinting and aversive learning in the chick, access to the formation and retrieval of the memory trace might be limited to only one input side, although the sensory systems are bilaterally symmetrical . In such a case, the memory exists or is accessible only in one side of the brain . Side-specificity of cognitive functions in the human brain is also indicative of asymmetrical representations of the respective memory traces , and, again, the memory can be established and retrieved only via the appropriate lateralized sensory input. In honeybees, Apis mellifera L., evidence exists for both side-specific and bilateral learning phenomena. Concerning the CS in both free-flying and harnessed bees, it was found that the olfactory input side is strongly connected to the respective odor, forming a spatial --olfactory compound as a learning signal . Bees can thus learn side-specific olfactory information. However, a unilaterally learned association was found to also be retrievable from the other brain side after a retention period . Moreover, interfering with the early process of memory consolidation after a single unilateral olfactory conditioning trial by local cooling revealed that only the ipsilateral antennal lobe, the primary olfactory neuropil, contributes to memory formation, but that both sides of the mushroom bodies, the second-order olfactory and higher-order multisensory integration center, are involved in memory formation . These observations are interpreted as indicating a distributed memory trace across the midline of the higher-order neuropils, even after unilateral conditioning. Other studies indicate that both brain sides are needed for solving particular learning tasks. For instance, suggested that reduced learning of odor component A in a mixture of two odorants (AB) after learning B (the phenomenon of "blocking") requires CS input on both sides (for a detailed discussion on the existence of the blocking phenomenon in bees, see ; ). Consistently, we found recently (B. Komischke, J.C. Sandoz, H. Lachnit, and M. Giurfa, in prep.) that bees can solve the nonelemental discrimination task A+ B+ AB- (negative patterning discrimination) only with olfactory input from both sides. Therefore, associative olfactory learning in the honeybee relies both on unilateral and bilateral CS processes. As for the US, until now very little work has sought to understand the influence of the US input site on learning. The US normally provided in conditioning experiments with honeybees (and present in natural foraging situations) is a compound of sucrose stimulation, first at the level of the antenna (which triggers the proboscis extension reflex), and then at the proboscis (the bee then licks some of the sucrose solution). already indicated that each US component alone can act as a reinforcer. However, the influence of each US component on the side-specificity of olfactory learning is utterly unknown. One can hypothesize that the antenna --US involves either a unilateral stimulation (when given to one antenna) or a bilateral stimulation (when given to both antennae) and that the proboscis --US provides a bilateral stimulation. How are these either unilateral or bilateral US components processed, and what is their respective role in the establishment of associative memory traces? We addressed these questions by using a series of behavioral procedures in which we controlled and varied the site of US input. In addition, we compared the ability of bees to be conditioned with an antenna --US or a proboscis --US alone, as compared with the usual compound --US. RESULTS : A general overview of the experimental protocols is provided in Table . On each experimental day, identical numbers of bees received a given set of stimulations on the left as on the right side. Table 1 | Experiment Protocols: Schematic Design of the Experiments, Detailing the CS and US Presentations Performed in Each Experimental Group Experiment 1: Effect of US Input Site on Sensitization | In this experiment, we evaluated the influence of the different US components on sensitization. In particular, we asked whether the antennal components of the US produce unilateral or bilateral sensitization. Protocol | Between 24 and 12 min before the sensitization procedure began, bees were subjected to one pretest with clove oil odor (see Materials and Methods) on each side. Bees that responded spontaneously to the odor on either side were discarded (17.6%, n = 142). Three groups of animals were sensitized with (1) an antenna --US, (2) a proboscis --US, or (3) a compound --US, respectively. Bees were then tested twice with the clove oil odor on one side, first after 30 sec and again after 12 min. According to previous studies , the highest responsiveness to the odor is expected after 30 sec (sensitization), whereas after 12 min, sensitization effects should have disappeared. In cases 1 and 3, responses were compared between the side ipsilateral to the US stimulation (one subgroup) and the contralateral side (second subgroup). In case 2, where no lateral US was given, the responses were compared between the left (one subgroup) and right (second subgroup) sides. Results | Figure presents the percentages of responses to the clove oil odor, 30 sec or 12 min after a sensitizing stimulus was applied as an antenna --US, a proboscis --US, or a compound --US. Thirty seconds after the US, sensitization responses to clove oil were high, between 26% and 45% in all groups, except in one case: bees that received an antenna --US showed sensitization only when the odor was ipsilateral to the US (36%) and not when presented on the contralateral side (6%). This difference was significant (G = 5.3, P < 0.05, 1 df). No statistical differences appeared between sides for groups sensitized using a proboscis --US or a compound --US. After 12 min, responses in all groups were <15%, showing that the effect of sensitization had by then disappeared. Thus, an antenna --US provides unilateral sensitization, restricted to the side of US application, whereas proboscis --US and compound --US induce bilateral sensitization. Figure 1 | Responses of bees in a sensitization experiment with the different US components: antenna --US, proboscis --US, and compound --US. Responses of bees in a sensitization experiment with the different US components: antenna --US, proboscis --US, and compound --US. Sensitization is observed as an increased response to an odor shortly (30 sec) but not later (12 min) after US stimulation. Antenna --US led to unilateral sensitization, whereas proboscis --US and compound --US showed no significant differences between sides. (*) P < 0.05, G-test. Experiment 2: Effect of US Input Site on Conditioning | We investigated the influence of the US input site on conditioning with a unilateral CS. More precisely, we asked whether bees can learn a CS when associated with any US components, and what effect such learning has on the opposite brain side. Protocol | The experiment consisted of two acquisition phases of four conditioning trials (9-min intertrial intervals). The CS was directed to one antenna during Phase 1, and to the opposite antenna during Phase 2. As in Experiment 1, three groups of bees were trained according to the properties of the US: antenna --US, proboscis --US, or compound --US. Antennal components of the US (in the case of antenna --US or compound --US groups) were always presented ipsilaterally to the CS, that is, on one given side in Phase 1, and on the opposite side in Phase 2. Results | During the first phase of the procedure, bees from all three groups showed acquisition of conditioned responses to the unilateral CS, from 8% --20% at the first trial to 42% --45% at the fourth trial . No significant difference appeared among groups in this phase. When the CS changed sides (trial 5 ---note that the antennal US also changed sides, but was given after the response in trial 5 was recorded), responses were low on the opposite side (between 11% and 26%). This indicates that in such a procedure with short intertrial intervals, a transfer between sides has not yet occurred, confirming a previous study . During the second acquisition phase, performances differed between groups: whereas bees from antenna --US and compound --US groups showed an increase in conditioned responses (from 11 --26 to 41% --45%), performances of the proboscis --US group remained low (<31%). This difference was significant (log-linear analysis, partial and marginal association chi2 > 8.0, P < 0.05, 2 df). Two-by-two comparisons showed that the compound --US and antenna --US groups performed similarly in the second phase (partial and marginal associations chi2 < 0.6, NS, 1 df), but both groups responded significantly more than the proboscis --US group (partial and marginal associations chi2 > 4.08, P < 0.05, 1 df). Thus, bees' learning performances were similar with all US components, and when the CS was switched to the opposite side, bees conditioned with a proboscis --US showed a resistance to acquisition as compared with bees conditioned with an antenna --US or a compound --US. Figure 2 | Bees' performances in a conditioning procedure with a unilateral CS and the different US components: antenna --US, proboscis --US, and compound --US. Bees' performances in a conditioning procedure with a unilateral CS and the different US components: antenna --US, proboscis --US, and compound --US. In the first phase, bees from all groups learned the CS --US association. In the second phase, after the CS changed sides, bees receiving a proboscis --US showed a significant blocking of acquisition. (NS) Nonsignificant; (*) P < 0.05, log-linear analysis. Different letters indicate significantly different response levels (pairwise log-linear analyses). Experiment 3: Associative Dependency of the Proboscis --US Effect | This experiment was designed to explain the resistance to acquisition obtained for the proboscis --US group in the second phase of Experiment 2: Was the observed effect caused by an associative process, that is, by an influence of the CS --US association on the contralateral brain side? Or was it caused by an effect of the proboscis --US alone on the opposite brain side? In the latter case, we hypothesize that the reinforcing function of the proboscis --US on the side where no CS is given could produce a US-preexposure effect through the development of a context --US association. Protocol | The experiment consisted of two experimental phases of four trials (9-min intertrial intervals). The CS was directed to one antenna during Phase 1, and to the opposite antenna during Phase 2. The US was always applied at the level of the proboscis. In Phase 1, bees were subjected to one of three treatments: (1) placement, as a control situation, in which the bee was placed at each trial in the experimental situation but received neither CS nor US; (2) proboscis --US-only presentations; and (3) conditioning trials (CS --proboscis --US presentations). In Phase 2, all bees were subjected to conditioning trials with a proboscis --US. The third group was thus equivalent to the proboscis --US group of Experiment 2. Results | Figure presents the acquisition performances of bees in Phase 2. Although performances of the placement group clearly increased during the procedure (from 10% to 50%), those of both the US --proboscis-only group and the conditioning group showed resistance to acquisition (responses from 7% --10% to 27% --28%). Responses of the different groups of bees were significantly heterogeneous (log-linear analysis, partial and marginal association chi2 > 6.8, P < 0.05, 2 df). Two-by-two comparisons further showed that bees that received only placement trials in the first phase responded significantly more than bees subjected to conditioning trials (partial and marginal association chi2 > 7.0, P < 0.01, 1 df). Responses of bees subjected to US-only trials in the first phase of the experiment showed a near-significant difference from those of bees from the placement groups (partial and marginal association chi2 > 3.5, P < 0.061, 1 df). These results indicate that the presentation of a proboscis --US, with or without a lateral CS, induces a resistance to the acquisition of this CS on the opposite side. Figure 3 | Bees' performances in the second phase of a conditioning procedure with a proboscis --US. Bees' performances in the second phase of a conditioning procedure with a proboscis --US. In the first phase, bees received either placement trials, US-only trials, or CS --US associations. In the second phase, all bees received CS --US associations. Whereas bees receiving placement trials learned efficiently in the second phase, both groups of bees receiving either proboscis --US, or CS --proboscis --US associations showed reduced acquisition. This supports the argument for a nonassociative dependency of the proboscis --US effect found in Experiment 2. (*) P < 0.05, log-linear analysis. Different letters indicate significantly different response levels (pairwise log-linear analyses). Experiment 4: CS-Specificity of Antenna --US Conditioning | We asked whether conditioning with an antenna --US produces an associative memory trace as a compound --US does, that is, a memory trace that is specific to the CS. Critical for such a conclusion is a differential conditioning experiment, where a CS+ is associated to the US, and a CS- is not. If animals significantly differentiate the two CSs (responding to the CS+ but not to the CS-), then performance has an associative basis. Protocol | Honeybees were subjected to a differential conditioning procedure, wherein two odors were presented alternately, one being rewarded (CS+) and the other not (CS-). At least 30 min before conditioning began, one antenna was cut off at the scapus. On each experimental day, half the bees were missing the left antenna and half the right antenna. One group of bees ("US same side") received the two CSs and the US on the intact antenna. A second group ("CS cut side") received the CSs on the side where the antenna was missing, and the US on the intact side. Such a control group allowed us to test the influence of the nonassociative component of the US and the quality of our apparatus: If this group showed responses to the odors during conditioning, it would be either caused by sensitization to the US (in which case, they would respond indifferently to CS+ and CS-), or by odor contaminations from one side of the apparatus to the other (in which case, they would respond more to the CS+ than to the CS-). In both groups, the conditioning procedure consisted of 6 CS+ trials and 6 CS- trials, with intertrial intervals of 9 min. On CS+ trials, bees received the CS+ in association with an antenna --US, and on CS- trials, they were stimulated with the CS- odor alone. The procedure began with a CS- trial, and CS+ (+) and CS- (-) were pseudorandomized in a - + + - + - - + order. Results | Bees subjected to a differential conditioning procedure with the CSs presented on the cut antenna (Fig. , "CS cut side") showed only very few responses throughout the procedure (<11%). In contrast, bees conditioned with the CS and the antenna --US on the intact side ("US same side") learned to respond to the CS+ and not to respond to the CS-, showing, respectively, 59% and 7% responses at the end of training. The log-linear analysis shows a highly significant difference between groups (partial and marginal association chi2 > 85.2, P < 0.001, 1 df). In the "US same side" group, bees responded significantly more to the CS+ than to the CS- (partial and marginal association chi2 > 30.1, P < 0.001, 1 df), whereas in the "CS cut side" group they did not (partial and marginal association chi2 < 1.29, NS, 1 df). Thus, an antenna --US is capable of producing a specific CS --US association like the compound --US used in other studies. Nonassociative effects (like sensitization) alone cannot explain the differentiation performance of bees conditioned with an antenna --US. Furthermore, the results of the control group prove that our experimental setup guarantees side-specific odor stimulation. Figure 4 | Performances of bees with a cut antenna in a differential conditioning procedure with unilateral CS+ and CS- and an antenna --US on the same side ("US same side"). Performances of bees with a cut antenna in a differential conditioning procedure with unilateral CS+ and CS- and an antenna --US on the same side ("US same side"). As a control, another group of bees received the CS on the side without an antenna ("CS cut side"). Whereas the "US same side" group learned to differentiate CS+ and CS- with an antenna --US, no amount of differentiation was observed in the "CS cut side" group. (NS) Nonsignificant; (***) P < 0.001, log-linear analysis. Experiment 5: Side-Specificity of Antenna --US Input | In this experiment we aimed to understand the properties of the antenna --US; in particular, we asked whether the reinforcing effect of an antenna --US (found in Experiment 4) is limited to one brain side, like sensitization (Experiment 1), or if contralateral reinforcement occurs. Protocol | The experiment consisted of two acquisition phases of four conditioning trials each (9-min intertrial intervals). The CS was directed to one antenna during Phase 1, and to the opposite antenna during Phase 2. Three groups of bees were trained according to the input side of the antenna --US: the US was thus given (1) ipsilaterally to the CS ("US same side"), (2) contralaterally to the CS ("US opposite side"), or (3) bilaterally ("US both sides"). Note that the protocol applied in group 1 was the same as in the antenna --US group of Experiment 2, and that in group 2, the US was on a different side with respect to the CS both in Phase 1 and in Phase 2. Results | The performances of bees trained in an absolute conditioning procedure with different antenna --USs is presented in Figure . In the first phase, bees from all three groups showed low acquisition of conditioned responses to the unilateral CS, from 0% --3% at the first trial to 15% --25% at the fourth trial. No significant difference appeared among groups in this phase (log-linear analysis, partial and marginal association chi2 < 0.85, NS, 2 df). In Phase 2, when the CS (and US) has changed sides, responses differed between groups: whereas bees from the "US same side" group showed a strong increase of conditioned responses (9% --41%), performances of the "US both sides" and of the "US opposite side" groups changed little (remaining below 22%). This difference is significant (log-linear analysis, partial and marginal association chi2 > 8.6, P < 0.013, 2 df). Two-by-two comparisons show that the same side group responded significantly more than any of the other groups (partial and marginal association chi2 > 4.3, P < 0.035, 1 df). Bees could be conditioned with all kinds of antenna --USs (although with relatively low performance: cf. the antenna --US group of Experiment 1), and when the side of CS input changed, only the group receiving CS and US on the same side learned efficiently. Figure 5 | Performances of bees in a conditioning procedure with a unilateral CS, and an antenna --US on the same side as the CS, on the opposite side, or on both sides. Performances of bees in a conditioning procedure with a unilateral CS, and an antenna --US on the same side as the CS, on the opposite side, or on both sides. In the first phase, all groups showed limited performances. In the second phase, after the CS had changed sides, bees receiving the US ipsilaterally to the CS showed better acquisition than the two other groups. (NS) Nonsignificant; (*) P < 0.05, log-linear analysis. Different letters indicate significantly different response levels (pairwise log-linear analyses). Experiment 6: CS-Specificity of Opposite Antenna --US | Because some conditioning performance was observed for all three kinds of antenna --USs in Experiment 5, this experiment was designed to test if such performance actually has an associative basis, that is, is it caused by CS --US associations? As before (Experiment 4), we applied a differential conditioning experiment, in which we checked whether animals significantly differentiate CS+ and CS-, that is, whether they establish an associative memory trace with the antenna --US. Protocol | All bees were subjected to a differential conditioning procedure, with one-sided CS+ and CS-, the US being provided either (1) ipsilaterally to the CSs ("US same side"), (2) contralaterally to them ("US opposite side"), or (3) bilaterally ("US both sides"). In all three groups, the conditioning procedure consisted of six CS+ trials and six CS- trials with intertrial intervals of 9 min. On CS+ trials, bees received the CS+ in association with an antenna --US, and on CS- trials, they were stimulated with the CS- odor alone. The procedure began with a CS- trial, and CS+ (+) and CS- (-) were pseudorandomized in a - + + - + - - + order. Results | The performances of bees trained in a differential conditioning procedure with different antenna --USs are presented in Figure . Bees trained with the US on the same side as the CS, or with the US on both sides showed fast learning. Responses to the CS+ increased from 6% at the first trial to 50% --58% at the end of training, whereas responses to the CS- decreased to <3%. In the group receiving the US contralaterally to the CS, acquisition was slower, starting after the fourth CS --US association (recorded at the fifth) trial, and reaching 34% at the end of training. The log-linear analysis shows a significant difference among groups (log-linear analysis, partial and marginal associations chi2 > 24.7, P < 0.001, 2 df). In each group, responses to the CS+ were significantly higher than responses to the CS- ("US same side" and "US both sides": partial and marginal association chi2 > 41.2, P << 0.001, 1 df; "US opposite side": partial and marginal association chi2 > 14.2, P < 0.001, 1 df). Responses to the CS+ are significantly different among groups (partial and marginal associations chi2 > 24.6, P < 0.001, 2 df), and two-by-two comparisons show that the "US opposite side" group responded significantly less than each of the other groups (partial and marginal associations chi2 > 22.4, P < 0.001, 1 df), with no difference between them (partial and marginal associations chi2 < 2.1, NS, 1 df). Bees thus established CS-specific memories with all types of antenna --USs, although learning took place more slowly with contralateral CS and US. Figure 6 | Bees' performances in a differential conditioning procedure with a unilateral CS and different types of antenna --US: on the same side as the CS, on the opposite side, or on both sides. Bees' performances in a differential conditioning procedure with a unilateral CS and different types of antenna --US: on the same side as the CS, on the opposite side, or on both sides. All groups of bees learned to significantly differentiate the CS+ and the CS-. (***) P < 0.001; log-linear analysis. However, when the antenna --US was provided contralaterally to the CSs, learning took place more slowly. Different letters indicate significantly different levels of response to the CS+ (pairwise log-linear analyses). No difference appeared among responses to the CS-. DISCUSSION : The present study indicates that (1) An antenna --US produces unilateral sensitization, whereas proboscis --USs and compound --USs produce bilateral sensitization. (2) All three types of US support similar acquisition with a unilateral CS. However, after rewarding with a proboscis --US, new learning appears to be blocked on the opposite side. (3) This proboscis --US effect also appears after US-only trials, which indicates a nonassociative basis. (4) Conditioning with an antenna --US leads to an associative memory trace, because bees perform well in a differential conditioning procedure. (5) Bees also show acquisition when a unilateral CS and the antenna --US are presented on opposite input sides. (6) Such performances also rely on an associative process, indicating that the antenna --US supports both a unilateral and a bilateral reinforcing function. Nonassociative US Processes | In Experiment 1, we show that an antenna --US produces side-specific sensitization. In contrast, a US stimulation to the proboscis produces bilateral sensitization. A compound --US presentation (antenna and then proboscis) shows a tendency for more sensitization on the side that received the antennal part of the US, although this difference was not significant. This indicates that the compound --US induces two processes, a unilateral antenna --US process, and a bilateral proboscis --US process. The fact that sensitization after an antenna --US is lateralized is particularly interesting, because another nonassociative learning phenomenon, habituation, was already found to be limited to the stimulated side . After habituation, however, a strong stimulation on the opposite side dishabituates the reflex, indicating some cross-talk of US pathways between brain sides with respect to nonassociative plasticity. As we discuss below, our data provide further evidence of bilateral antenna --US processes. Associative US Processes | In an associative conditioning procedure, all three kinds of US (antenna --US, proboscis --US, and compound --US) induced an increase in responses to the CS with repeated forward-pairing trials (Experiment 2; see also ). We know from extensive previous work that the memory trace produced by a compound --US is associative in nature, that is, responses to the CS are caused by the learning of the CS --US association, and not to nonassociative effects like sensitization . Critical for such a conclusion are experiments like pseudoconditioning (where US and CS are presented explicitly unpaired) or differential conditioning (where a CS+ is associated to the US, and a CS- is not). In the latter case, associative learning can be inferred when animals significantly differentiate the two CSs. The fact that the antenna --US is equally capable of inducing such associative phenomena (Experiment 4) may appear contraintuitive because bees do not get any real appetitive reinforcement (meaning energy as food uptake) in an antenna --US. However, gustatory (sucrose) receptors are present at the level of the antennae (Sensilla chaetica, see below). Therefore, sucrose stimulation, rather than food uptake, activates the reward pathway in the bee brain, and both sites of sucrose stimulation (antennae, proboscis) appear to activate the reward system. We subjected bees to conditioning procedures with an antenna --US either ipsilateral to the CS, contralateral to it, or bilateral. A performance increase was observed in all cases, which was also shown to be associative, because bees conditioned with all three antenna --US types responded specifically to the CS+ (Experiment 6). This suggests that, in addition to side-specific nonassociative learning (habituation and sensitization, see above), an antenna --US can also act as a reinforcer on the opposite brain side. Contralateral reinforcement appears to be weaker than ipsilateral reinforcement because slower acquisition was observed for contralateral CSs and USs in Experiment 6, and more effective acquisition was found for ipsilateral reinforcement in the second phase of Experiment 5. Proboscis --US Effect | Although bees learn an olfactory CS --US association with any US component (Experiment 2, Phase 1), switching the CS to the opposite side (Phase 2) showed a resistance to acquisition after reinforcement with a proboscis --US. To evaluate whether this was caused by an effect of the CS --US association on the contralateral side, or to the proboscis --US stimulation alone, we repeated this experiment and compared the bees' performances to those of animals receiving placement trials (control) or US --proboscis-only trials (Experiment 3). We found that the retardation of acquisition in the second phase was actually due to US processes, because the proboscis --US-only group showed the same reduced performances as the CS --US group. The proboscis --US effect is reminiscent of the US-preexposure effect that was shown in bees in an aversive conditioning task and suggested for PER conditioning . During US preexposure, an animal is repeatedly stimulated with the US alone and may form an association between context stimuli and the US, which would in turn induce resistance to CS --US acquisition. Based on the morphology of the identified reward pathway in bees (the VUMmx1, see below), we believe that the reinforcement produced by proboscis --US stimulations is a bilateral process. Therefore, in the first phase of Experiment 2, bees received a CS on one brain side associated with the proboscis --US and simultaneously no CS on the other side, associated with the proboscis --US. This could have produced a CS --US association on one side and a context --US association (including no CS) on the other. When the CS was then switched to this side, this might lead to a typical US-preexposure effect, retarding the CS --US acquisition. Interestingly, the same effect was found for the antenna --US. In Experiment 5, after changing the CS side (Phase 2), acquisition was much higher for bees receiving CS and US on the same side than for bees receiving the US on the opposite side or on both sides. When conditioning bees with the US on the side opposite to the CS, two associations are probably built at the same time. On the CS side, a CS --US association that takes time to develop (see Experiment 6) and relies on bilateral processes is built. On the other side, where no CS was given, a context --US association (including no CS) could develop, which would, as postulated for the US --proboscis, retard later acquisition of the CS --US association on this side. A logical consequence of this hypothesis is that when bees receive a bilateral antenna --US, the same effect should appear. The reduced acquisition of the both-sides group in the second phase of Experiment 5 confirms this view. Properties of the Compound --US | An intriguing result was observed in Experiment 2. Although a proboscis --US induced US-preexposure effects when the CS was switched to the opposite side, this did not occur with a compound --US. This indicates that the compound --US provides more (and other) reward information than just the addition of antenna --US and proboscis --US stimulations. From a biological point of view, one might argue that the antenna --US has a signaling function for the reinforcement, whereas the proboscis --US assesses the overall quality of the reinforcement. Indeed, previous work showed that when applying a compound stimulation in a tactile learning paradigm, the quality of the proboscis --US but not of the antenna --US is crucial in determining the level of acquisition of the CS --US association (; R. Scheiner, pers. comm.). For instance, conditioning with a compound --US consisting of fixed proboscis --US (30% sucrose solution) but of various antenna --USs (water, 1.6% and 30% sucrose solutions, and even a salt solution), led to good conditioning performance in all cases. In contrast, when a 30% sucrose solution was proposed to the antenna, bees learned only when the sucrose concentration given to the proboscis confirmed what was given to the antenna: when given a low sucrose concentration (1.6%), water or a salt solution to the proboscis led to very low acquisition. According to these data and ours, the antenna detects the presence of water and sucrose, which triggers, depending on the motivational state of the bee (and its response threshold; ), the extension of the proboscis, and builds a first CS --US association. Additionally, information about US quality, gathered at the proboscis, could serve to modify the strength of this association during the consolidation process, or alternately to create an additional CS --US association. Further work should attempt to elucidate this question. Bilateral Transfer | In two experiments (Experiments 2 and 5), bees were subjected to two conditioning phases, the side of CS input changing from one phase to the other. After this change of CS input, a drop in responses was observed, indicating that the CS --US association was not yet retrievable on the other brain side. This was the case for all US components: compound --US, antenna --US, and proboscis --US. This result complements those from a previous study, where we found that a CS --US association built on one side could be retrieved on the contralateral side . Interestingly, the main difference between the two studies resides in the amount of time after which transfer was tested. In the present experiments, transfer was tested in Phase 2 after a 9-min intertrial interval, whereas in the previous study, tests for bilateral transfer were carried out 3 h and 24 h after conditioning. Indeed, we think that time is a critical factor for the occurrence of bilateral transfer, as we found recently that retrieval on the contralateral side increases from 10 min to 3 h after conditioning (J.C. Sandoz and R. Menzel, in prep.). This indicates that the first phases of the odor CS --US association are unilateral, but later phases are bilaterally accessible. Neural Substrates of the US | The neural substrates of both the CS and the US pathway in PER conditioning are rather well-known . A key finding was the discovery of a single neuron, the VUMmx1 (ventral unpaired median neuron of the maxillary neuromere 1), which converges bilaterally with the CS pathway, at the level of the antennal lobes, the mushroom bodies, and the lateral protocerebral lobes . This neuron was shown to mediate the reinforcement in PER conditioning, because the forward (but not backward) pairing of an odor CS with an artificial depolarization of VUMmx1 produces an associative memory trace. Because VUMmx1 responds both to sucrose stimulations to the antenna and to the proboscis, sucrose receptors on both the antennae and on the proboscis must connect to VUMmx1. However, the exact projection patterns of gustatory receptors are as yet poorly understood. On the antenna, sensilla trichodea type D (i.e., sensilla chaetica; ; ) are thought to be involved in gustatory perception (; ; ,). Two possibilities concerning their projections have been suggested: most of the sensory neurons of the antenna lead directly to the antennal lobe (antennal tracts T1 --T4, mostly olfactory; ; ; ), but a considerable portion of the sensory neurons bypass the antennal lobe altogether, and project to the dorsal lobe (tract T5), to the caudal part of the protocerebrum (tract T6I), and to the suboesophageal ganglion (SOG; tract T6II; ). Therefore, US information perceived at the antennal level could potentially be directed to either (or both) the antennal lobe and the SOG. Side-specific nonassociative processes like habituation and sensitization, as well as side-specific associative processes (antenna --US), could take place at the level of the antennal lobe, under direct influence of projections from the gustatory receptors of the antenna. Contralateral reinforcement of the antenna --US (Experiments 5 and 6) is probably assured by a projection of antennal gustatory receptors to the SOG, where they would feed either directly or via local interneurons, to the VUMmx1. In this way, bilateral reinforcing effects could also take place, albeit with a different time course, as seen in Experiment 6. At the level of the mouthparts, gustatory receptors (sensilla chaetica; ) are placed on the glossa, galeae, and labial palps, but also in the hypopharynx and epipharynx . They project to the SOG, in the same area where the antennal T6II fibers project . It is reasonable to think that they synapse there either directly, or via local interneurons, onto the VUMmx1 neuron, thus inducing bilateral reinforcement. However, the connections of antennal and proboscis gustatory receptors with reinforcing circuits (probably the VUMmx1 neuron) we propose are still hypothetical, and particular effort should be invested in anatomical studies that would help to determine precisely the neural processes of the US pathway. Functional Model of Side-Specific Processes | Based on our results and on our knowledge and predictions concerning the CS and US pathways, we developed a model summarizing side-specific and bilateral processes in olfactory PER conditioning . This model consists of three functional subcompartments, related to Learning, Memory, and Retrieval, respectively. In the Learning phase, the animal receives a set of stimulations with a CS and/or a US, on one or both sides. Each side of the brain houses primary association centers where the CS/US coincidence is detected and associations are built. These associations, held in Short-Term Memory (STM) storage, lead through a time-, and event-dependent process to different forms of Memory ---Mid-Term Memory (MTM) and Long-Term Memory (LTM) phases. When a learned CS is presented to the bee on either side, the Retrieval process starts, and the stored information about the learned CS is retrieved from this side. This process controls the behavior of bees via the proboscis extension motor program. As shown by anatomical studies , the CS input is unilateral on each side, olfactory receptor neurons on the antenna projecting to the glomeruli of the antennal lobe on the ipsilateral side only, thus providing the bee with unilateral odor information on each side (CS-A1). The antenna --US, in parallel to triggering the extension of the proboscis (data not shown), induces first a unilateral reinforcing process. Coincidence detection in the association center on this side (Ass-1) builds a unilateral memory trace (STM1), which has a strong nonassociative component (is highly dependent on sensitization; , ). CS-specific memory consolidates through time and other conditioning trials into associative unilateral MTM1 and LTM1. As was found in previous work , the learned association is retrievable from the opposite side after some time (3 --24 h). Therefore, some of the content of LTM is transferred from side to side. As found in the present study, the proboscis --US induces bilateral reinforcing processes (conveyed by the VUMmx1 neuron), which build unilateral or bilateral memories, depending on the unilateral or bilateral presentation of the CS. Importantly, when the CS is provided on only one side, a olfactory memory is formed on this side, but on the opposite side, a context --US association (including no CS) is formed, which retards the formation of a subsequent CS --US association on this side (US-preexposure effect, Experiments 2 and 6). In parallel to its unilateral reinforcing function, the antenna --US induces bilateral reinforcement (through activation of VUMmx1), because contralateral presentation of CS and US (CS-A1 and US-A2 or CS-A2 and US-A1; Experiments 5 and 6) also leads to significant learning. In this case, a context --US association is built on the US side, which will retard acquisition if the CS is switched to this side (Experiment 5). Lastly, the side itself (as a context stimulus) can also, in association with the presented CS, enter into association with the US. As was found in a previous study , bees can learn a discrimination task of the type A+ B- / B+ A-, where each of the two odors is rewarded on one side, but not on the opposite side. In this case, bees would rely on higher-order cognitive capacities, learning a kind of configuration between an odor and the side on which it was rewarded. During retrieval, the side would thus play a crucial role in retrieving either an excitatory memory on one side or an inhibitory memory on the other side. Figure 7 | Functional model of the side-specificity of olfactory learning in the honeybee. Functional model of the side-specificity of olfactory learning in the honeybee. CS and US stimulations are represented according to the input site: CS-A1 and US-A1 refer to stimulations to the antenna of side 1, and CS-A2 and US-A2 to stimulations to the antenna of side 2. US-P corresponds to the proboscis --US component. On each side, CS and US processes project to an association center (Ass-1 and Ass-2), where the associative pairing effects take place. CS ---US associations as well as context --US associations are built, according to the salience (presence or absence) of a CS on this side. The content of this association as well as transient sensitization effects control unilateral Short-Term Memory phases (STM1 and STM2). Through time and/or additional training, memories are consolidated into Mid-Term Memory (MTM1 and MTM2) and Long-Term Memory phases (LTM1 and LTM2). LTM phases are characterized by the use of both brain sides and bilateral transfer effects . When tested with the CS on one side, the retrieval process starts, and information stored in the memory stores of this side is used to control the animal's responses (MP) Proboscis motor program; (CR) conditioned response. Conclusion | Our work has addressed the side-specificity of the US pathway, by conditioning bees with unilateral and bilateral antenna --US, with proboscis --US, and with compound --US. This and a previous study have allowed us to build a model of side-specific olfactory conditioning in the honeybee. Although both the CS and the US pathways are relatively well known in the bee, our model is highly speculative with respect to the neural substrates involved. Nevertheless, based on the few anatomical studies carried out so far, as well as on our behavioral results, we can make a number of predictions concerning US processes that should be tested in the future. In particular, special attention should be paid to elucidating the projections of sucrose gustatory receptors (antenna, proboscis, and tarsae of the front legs) in the brain of the bee, as well as their convergence with the VUMmx1 neuron. MATERIALS AND METHODS : Honeybees | Worker bees were collected from an indoor flight room with constant temperature and humidity conditions (24C, 60% relative humidity). They were immobilized by brief cooling and fixed in metal harnesses with tape. The head capsule of each bee was fixed to the tape with a drop of low-temperature melting wax to prevent movement. In Experiments 1 --3, the bees were fed to satiation with a 40% sucrose solution and kept overnight. Conditioning experiments were performed the next morning. In Experiments 4 --6, the bees were kept for 2 h before the experiments began. Stimulation Apparatus | The olfactory space of bees was delimited using two different methods, depending on the type of experiment. In Experiments 1 --3, a removable wall device was used. To allow control over the individual components of the US, the device consisted of a removable vertical wall, which could be placed between the antennae of each bee before each trial. The separation wall was made of Plexiglas (1 mm thick) and had a small perforation at the level of the proboscis to make it easily accessible from both sides. This allowed us to give a sucrose stimulation to the proboscis from either side of the apparatus. Odor presentations were precisely directed to each antenna using glass pipettes. Two pipettes, one on each side of the wall, were fixed in front of the bee, with their outlet placed at a distance of 1 cm from the antenna. Each pipette could be connected to a 20-mL syringe containing the odor source. During one-sided odor stimulations, an odor syringe was activated, and a thin odor flow was given to one antenna. An exhaust placed 10 cm behind the bee ensured that the flows on each side of the wall were laminar and that all released odors were extracted from the experimental room. Bees were placed into the apparatus one at a time, the wall was installed between the antennae, and each pipette was precisely directed to one antenna. The bee was left for 2 min to adapt to the experimental conditions. In Experiments 4 --6, a new device was developed that allowed applying side-specific odor stimulations under computer control (it was necessary when more than one odor was used, as in Experiments 4 and 6). As in a previous study , each individual bee had a thin plastic wall on its head, which separated the two sides of its olfactory space. These walls were secured on the head and waxed in place in such a way that all remaining spaces were closed and the proboscis was placed on one side (left side for half the bees, right side for the other half). Therefore, although only the bee's antennae received a US in the experiments, the proboscis could move freely. During conditioning, each bee was placed individually in front of the bilateral odor-supplying device. Identical airstreams were provided on the two sides, and were directed to each antenna (on each side of the plastic wall) via Teflon tubing (5 mm diameter, 5 cm length). Before reaching the bee, each air stream was directed through one of three channels, regulated by three valves controlled by the experimenter via a computer. Each channel contained a cartridge made from the cut end of a 1-mL syringe. The first channel, where the air stream flowed continuously when no odor stimulation was given, contained a cartridge holding a piece of filter paper. The second and third channels each contained an odor source, that is, a piece of filter paper soaked with odor substance. Behind the head of the bee, an exhaust removed the released odorants from the experimental room. Stimuli | Clove oil (from a local pharmacy, Berlin-Dahlem, Germany) was used as the CS throughout, because it provides intermediate performance levels, which allow the researcher to carry out a greater number of conditioning trials. In two experiments where differential conditioning procedures were performed (Experiments 4 and 6), orange flower oil (local pharmacy, Berlin-Dahlem, Germany) was used as the second odor. The odor source was a 1-cm2 piece of filter paper soaked with 5 muL of pure substance, which was inserted into either a 20-mL syringe (Experiments 1 --3) or into an empty cartridge (Experiments 4 --6). A 40% (w/w) sucrose solution was used as the US. Stimulations | Rewarded (CS --US or CS+) Trials | The CS odor was presented to the bee for 4 sec on one side. Then, 3 sec after onset of the CS, US stimulation began. Three possible USs could be given:1. = Antenna --US. The US was given to only one antenna by briefly touching it with the sucrose solution (until a PER appeared). 2. = Proboscis --US. The US was given directly to the proboscis. The bee could lick sucrose solution for 3 sec. 3. = Compound --US. The US was first given to one antenna, thereby eliciting the proboscis extension reflex, then to the proboscis. The bee could then lick sucrose solution for 3 sec. For antenna --US and compound --US, the antennal US could be given either on the same side as the CS (referred to as "ipsilateral to the CS") or on the other side ("contralateral to the CS"). Sensitizing (US-Only) Trials | The US was given without any CS stimulation. As above, three possible USs could be given (antenna --US, proboscis --US, or compound --US). Unrewarded (CS-) Trials | The CS- was presented on one side, without any US stimulation. Statistics | In Experiment 1, differences in the percentages of animals responding to the odor stimuli between the ipsilateral and contralateral sides were evaluated in each group by a G-test with 1 df. In Experiments 2 --6, comparisons among groups of performances were made using a log-linear analysis of the frequencies of bees responding at each trial. Interactions of the design variables (conditioning trials, treatment, and/or CS+ vs. CS-) with the development of odor-evoked PER were considered significant only if both partial and marginal association chi2s were significant (P < 0.05). When significant, two-by-two comparisons were made between groups with log-linear analyses, and were considered significant under the same conditions as mentioned above. Backmatter: PMID- 12359833 TI - Recognition Memory for Single Items and for Associations Is Similarly Impaired Following Damage to the Hippocampal Region AB - The formation of new associations between items is critical for establishing episodic memories. It has been suggested that the hippocampus is essential for creating such associations but is not involved, or is much less involved, in memory for single items. In Experiment 1, we tested controls and amnesic patients with bilateral lesions thought to be limited primarily to the hippocampal region in both single-item and associative recognition memory tasks. In the single-item task, a conventional recognition memory task was administered in which participants studied either houses or faces and were tested for their ability to recognize the individual items. In the associative task, participants studied paired pictures of houses and faces with instructions that encouraged associating the two stimuli, and were tested for their ability to recognize the specific pairings that were presented at study. Like the controls, the amnesic patients performed more poorly on the associative task. Relative to the controls, the amnesic patients were impaired to a similar extent on the single-item and associative tasks. In Experiment 2, the performance of the amnesic patients was improved by increasing the number of presentations of the study lists (eight presentations instead of one). On both the single-item and associative tests, the performance of the amnesic patients after eight presentations was now identical to the performance of the controls who had been given only one presentation of the study list. Thus, the associative condition was not disproportionally difficult for the amnesic patients. These results are consistent with the idea that the hippocampus is similarly involved in single-item and associative memory. Keywords: Introduction : Declarative memory (memory for facts and events) relies upon structures in the medial temporal lobes, including the hippocampal region (the hippocampus proper, the dentate gyrus, the subiculum) and the adjacent structures that lie along the parahippocampal gyrus (the entorhinal, perirhinal, and parahippocampal cortices). Currently, the precise contribution that these structures make to declarative memory is not well understood. There have been several efforts to distinguish the role of the hippocampal region from the role of the adjacent structures. Many of these share a common thread, proposing that the hippocampus is particularly involved in declarative memory tasks that require the formation and use of associations between the separate components of presented material (e.g., ; ; ; ; ; ; ). In strong versions of this view, the hippocampus is proposed to be essential for overtly associative tasks such as recall or paired-associate learning, and it also supports the recollective component of recognition memory. In contrast, the capacity for single-item declarative memory tasks (including familiarity-based recognition) is supported by adjacent structures in the parahippocampal gyrus (e.g., the perirhinal cortex). An alternate view is that the hippocampus and parahippocampal gyrus are both broadly important for declarative memory (e.g., ; ; ; Stark and Squire, in press). According to this view, all tasks of declarative memory, including familiarity-based recognition, depend on forming associations, and both the hippocampus and the adjacent cortex are important for such tasks. Accordingly, a simple distinction between single-item and associative memory does not capture the division of labor between the hippocampus and adjacent structures in the medial temporal lobe. RESULTS : Experiment 1 | In Experiment 1, we examined the status of single-item and associative recognition memory in four amnesic patients with damage thought to be limited primarily to the hippocampal region. In the single-item condition, participants studied 10 pictures of either faces or houses. After a brief delay, a yes/no recognition memory test was given. Associative memory was also tested with a yes/no recognition memory task. In the associative condition, participants studied 10 pairs of houses and faces with instructions that encouraged forming an association between the items of each pair. At test, participants were shown house-face pairs and asked to determine whether each pair was an intact repetition of a studied pair or whether it was the recombination of a house and a face that had been studied as part of different pairs. If the hippocampal region provides a specific ability to explicitly associate components in declarative memory, damage to the hippocampal region should impair performance more severely on the associative house-face task than on the single-item task. Figure (left panel) shows the results. The controls scored 91% correct (correct "yes" responses plus correct "no" responses) in the single-item condition (hit rate = 90%; correct rejection rate = 92%) and 74% correct in the associative condition (hit rate = 77%; correct rejection rate = 72%). Amnesic patients with damage to the hippocampal region scored 75% correct in the single-item condition (hit rate = 76%; correct rejection rate = 74%) and 59% correct in the associative condition (hit rate = 64%; correct rejection rate = 55%). A repeated-measures analysis of variance (ANOVA) revealed an effect of group (F(1,12) = 7.6, P<.02), indicating that the amnesic patients were impaired overall relative to the controls, and an effect of test condition (F(1,12) = 21.0, P<.005), indicating that performance was poorer in the associative condition than in the single-item condition. However, there was no interaction between group and test condition (F(1,12) = 0.03), indicating that the relative difficulty of the associative condition was similar for the two groups. Figure 1 | Percent correct scores for control volunteers (CON; n = 10) and for amnesic patients with damage to the hippocampal region (H; n = 4) on a recognition memory test for pictures of houses or faces presented in isolation (single-item; white bars) and on an associative recognition test involving house-face pairs (associative; gray bars). Percent correct scores for control volunteers (CON; n = 10) and for amnesic patients with damage to the hippocampal region (H; n = 4) on a recognition memory test for pictures of houses or faces presented in isolation (single-item; white bars) and on an associative recognition test involving house-face pairs (associative; gray bars). In Experiment 1 (left), control volunteers and amnesic patients studied each item (single items or face-house pairs) only once (1x). In Experiment 2 (right), the amnesic patients studied each item eight times (8x). Error bars indicate the standard error of the mean, and the dashed lines indicate chance performance. Nevertheless, whereas performance of the controls was reliably above chance in both the single-item (t(9) = 31.0, P<.001) and associative (t(9) = 5.6, P<.001) conditions, the performance of the amnesic patients was not. In neither the single-item condition (t(3) = 2.9, P = .06) nor the associative condition (t(3) = 1.8, P = .18) was the performance of the patients reliably above chance. Experiment 2 | In Experiment 1, amnesic patients with damage to the hippocampal region were impaired in both the single-item and the associative conditions, and the relative difficulty of the associative condition was similar for amnesic patients and controls. Yet, because the performance of the amnesic patients was not reliably above chance, it is possible that the patients might have been disproportionally impaired in the associative condition and that a particularly severe impairment in this condition was obscured by a floor effect. In Experiment 2, we attempted to elevate the performance of the amnesic patients by increasing the amount of their exposure to the study items (for a discussion of various methods of equating amnesic and control performance, see ). Figure (right panel) shows the results, averaged across the two test sessions. The increase in study exposure (from 1 to 8 presentations) significantly improved the performance of the amnesic patients (67.2% to 83.2% overall; t(3) = 4.2, P<.05). The patients scored 92% correct in the single-item condition (hit rate = 92%; correct rejection rate = 93%) and 74% correct in the associative condition (hit rate = 83%; correct rejection rate = 65%). Their scores for both the single-item condition (t(3) = 8.3, P<.005) and the associative condition (t(3) = 3.7, P<.05) were reliably above chance levels. Further, eight repetitions of the study items effectively equated the score of the amnesic patients with the scores of the controls in Experiment 1, who had seen the study items only once (single-item condition: 92% correct for the patients, 91% correct for the controls; associative condition: 74% correct for the patients, 74% correct for the controls). Accordingly, an ANOVA comparing the performance of these two groups found no effect of group (F(1,12) = 0.01) and no interaction of group and test condition (F(1,12) = 0.06). The effect of test condition (F(1,12) = 28.1, P<.001) confirmed that the associative condition was more difficult than the single-item condition. DISCUSSION : In two experiments, we examined the status of single-item memory (memory for individual houses or faces) and associative recognition memory (memory for house-face pairs) in amnesic patients with damage thought to be limited primarily to the hippocampal region. In this way, we tested the hypothesis that the hippocampal region is particularly involved in the formation and use of associations, but is not critical (or is less critical) for single-item recognition memory. If the hippocampal region were more involved in associative than in single-item memory, one would have predicted an impairment limited to, or more severe in, the overtly associative house-face task. In Experiment 1, amnesic patients and controls performed more poorly on the associative task than on the single-item task. In addition, the amnesic patients were impaired overall relative to the controls. However, there was no evidence of a differential impairment for the associative task. In Experiment 2, the performance of the amnesic patients was improved by presenting each study list eight times. With these additional presentations of the study list, the performance of the amnesic patients with damage to the hippocampal region matched the performance of the controls from Experiment 1 on both the single-item and associative memory tasks. In summary, in neither experiment did we find evidence for a differential impairment on the associative memory task. Thus, these results are inconsistent with the view that the hippocampal region is any more involved in memory for associations than single-item memory. These results complement our report (Stark and Squire, in press) of similarly impaired memory for single items and memory for conjunctions in patients with damage to the hippocampal region. The present results extend our previous results in two ways. First, in the previous study, a continuous recognition memory task was used in which participants were asked to decide whether each item in a continuing stream of items had been encountered previously. In the task (based on one developed by ), items could be entirely novel, novel with one previously encountered (repeated) component, novel but with both components repeated, or a true repetition. We observed similar levels of impairment for all three groups of novel items and found no evidence of differential impairment for the explicitly associative recombined stimuli. In the present study, we extended this finding to the case of a traditional recognition memory task in which separate study phases and test phases were presented. Second, while three of the patients tested in the prior work had damage thought to be limited primarily to the hippocampal region, two others had damage that extended into adjacent medial temporal lobe structures. Although the performance of these two groups of patients did not differ overall, in the present study we were able to examine four patients with damage thought to be limited primarily to the hippocampal region. There have been a number of findings that have been taken as support for the view that the hippocampal region is particularly involved in associative forms of memory. For example, two neuroimaging studies reported greater hippocampal activity during an associative memory encoding task than during a nonassociative memory encoding task . Two studies reported similar results during memory retrieval . Yet, in each of these cases, the associative versus nonassociative contrast revealed activity in both the hippocampal region and the parahippocampal gyrus (also see , for an example of similar levels of hippocampal activity during associative and nonassociative recognition tasks). Therefore, although these data support a role for the medial temporal lobe in associative memory, they do not differentiate between the hippocampus and parahippocampal gyrus. The observation that neurons within the hippocampus often respond maximally to conjunctions of features (for reviews, see ; ; ) has also been taken to support the hypothesis that the hippocampal region is especially important for associative memory. Yet, such conjunctive codes are present as well in the parahippocampal gyrus . In addition, what appear to be single-feature codes are also present in the hippocampus . When these findings are considered in the light of the neuroanatomical evidence for associational connections not only within the hippocampus, but also within the entorhinal, perirhinal, and parahippocampal cortices (for review, see ), support for a distinction between the hippocampal region and the parahippocampal gyrus with respect to associative memory is weakened. From these and other studies (see Stark and Squire, in press, for additional discussion), we suggest that although there is likely functional specialization within the medial temporal lobe memory system, attempts to differentiate between the hippocampal region and the adjacent cortex based on simple dichotomies such as associative versus nonassociative memory are unlikely to be successful. That the hippocampal region is important for associative, recollective, episodic, conjunctive, and relational memory is quite clear. However, ascribing the capacity for associative (or recollective, episodic, conjunctive, or relational) memory only to the hippocampal region and the capacity for nonassociative (or familiarity, semantic, single-item, or nonrelational) memory only to structures in the parahippocampal gyrus does not account for the available neuropsychological, neuroimaging, neuroanatomical, or electrophysiological data. We suggest, as have others (e.g., ; ; Norman and O'Reilly, in press), that whatever the different roles of individual structures in the medial temporal lobe may be, the division of labor among these structures is not absolute. The present data suggest that the hippocampal region is equally important for single-item recognition memory and for recognition memory tests that overtly assess memory for associations between two stimuli. MATERIALS AND METHODS : Experiment 1 | Participants | Four amnesic patients (A.B., G.W., L.J., and M.J.) with damage thought to be limited primarily to the hippocampal region (CA fields, dentate gyrus, and subiculum) participated in the study . A.B. became amnesic in 1976 after an anoxic episode following cardiopulmonary arrest. G.W. became amnesic in September 2001, following an overdose of heroin and associated respiratory distress. L.J. became amnesic during a 6-mo period that began in 1998 with no known precipitating event. Her memory impairment has remained stable since that time. M.J. had a 10-y history of cardiovascular disease. On June 6, 1996, he awoke from a night's sleep complaining of memory difficulties. His impairment has remained stable since that time. Table 1 | Characteristics of Amnesic Patients For three of the patients (G.W., L.J, and M.J.), magnetic resonance imaging (MRI) has identified the damage within the medial temporal lobe by comparison with three to four age- and gender-matched healthy controls (for G.W. and M.J., volumes of hippocampal region and parahippocampal gyrus relative to intracranial volume; for L.J., areal measurements, see ). All three patients have reduced hippocampal area bilaterally (reductions of 45%, 46%, and 10%, respectively) and substantially less damage to the parahippocampal gyrus (reductions of 15%, 6%, and 3%, respectively). Patient A.B. wears a pacemaker and is ineligible to participate in MRI studies. In CT scans obtained in 2001, temporal lobe volume appeared normal, and the temporal horns were symmetric and normal in size. The basal ganglia and thalamus also appeared normal. The only focal lesions detected were small bilateral foci in the white matter lateral to the head of the caudate nucleus, which appeared to be old lacunar infarctions. Neurological exam indicated well-circumscribed amnesia. These findings, together with reports that histologically confirmed damage limited to the hippocampal formation can occur after anoxia , suggest that A.B.'s memory impairment is likely due to damage within the hippocampal region. Ten healthy control volunteers (five men and five women) were also tested. The controls were matched to the patients with respect to age (mean = 62.8 years; range = 38 --75; patient mean = 58.0), education (mean = 16.2 years; range = 12 --20; patient mean = 15.0), and WAIS-III subtest scores for Information (23.3; patients = 20.5) and Vocabulary (57.3; patients = 52.3). Because one of the patients (G.W.) was substantially younger than the other three, two of the control volunteers were specifically matched to G.W. Materials | The materials consisted of 80 color pictures of houses and 80 color portrait-style pictures of faces, presented on a computer screen . A total of 40 houses and 40 faces were randomly assigned to the single-item condition. The remaining 40 houses and 40 faces were randomly paired and assigned to the associative condition. Figure 2 | Examples of a house and a face stimulus. Examples of a house and a face stimulus. In the single-item condition, either a house or a face was presented in the middle of the screen. In the associative condition, the house and the face were presented side-by-side with instructions that encouraged associating the two stimuli. Procedure | Each participant completed four single-item and four associative study/test sequences. Single-item and associative tests were alternated, counterbalancing the initial test across participants. For each single-item test, participants were shown either 10 color pictures of houses or 10 color pictures of faces. Houses were used for two of the single-item tests, and faces were used for the other two (participants alternated between houses and faces, with the stimulus type for the initial single-item condition counterbalanced across participants). Participants were asked to make a judgment about each item within 4 sec. In the case of single-item stimuli, participants were asked to judge whether the house was built before or after 1960. In the case of faces, participants were asked to judge whether the person was a "cat person" or a "dog person". After 4 sec, the screen was cleared, and participants were asked to press the space bar to begin the next trial. After all 10 study trials were presented, a 3-min delay was imposed before the test phase began. To test memory for the items presented in the single-item tests, 20 stimuli (houses or faces) were presented. Half of the stimuli were studied targets and half were novel foils. On each trial, participants were asked to make a yes/no recognition memory judgment using keys labeled "yes" and "no". Test trials were self-paced. In each test of the associative condition, participants first studied a set of 10 house-face pairs. On each trial, a picture of a house and a picture of a face were presented on the computer screen side-by-side, and participants were asked to judge whether that person might live in that house. Participants were given four seconds to indicate their decision using keyboard keys labeled "yes" and "no". After four seconds, the screen was cleared, and participants were asked to press the space bar to begin the next trial. After all 10 study trials were presented, a 3-min delay was imposed, and 10 house-face pairs were presented. Participants were asked to judge (by pressing keys labeled "yes" and "no") to indicate whether the house and face had been studied together previously or whether they had been studied separately (half of the pairs were intact pairs and half were recombined). Test trials were self-paced. Experiment 2 | Participants | The same four amnesic patients participated. Materials | Two lists of 90 houses and 90 faces were constructed, similar to those used in Experiment 1. The two lists were used in separate test sessions. In each list, 60 houses and 60 faces were assigned to the single-item condition. The remaining 30 houses and 30 faces were randomly paired and assigned to the associative condition. Procedure | The procedure was identical to that of Experiment 1 with three exceptions. First, each patient was tested in two sessions, separated by at least 6 mo. Second, in contrast to the single 4-sec presentation of the study list in Experiment 1, each study list was presented eight times in succession for 6 sec each prior to the test. Third, within each session, patients were given four associative study/test sequences, as in Experiment 1, but only three single-item sequences (vs. two of each type in Experiment 1). In each session, all three single-item tests involved either houses or faces, and the single-item stimulus type in the second session was the one that had not been tested in the first session (stimulus type in the first session was counterbalanced across participants). Backmatter: PMID- 12359834 TI - Overexpression of hAPPswe Impairs Rewarded Alternation and Contextual Fear Conditioning in a Transgenic Mouse Model of Alzheimer's Disease AB - One of the hallmarks of the pathology in Alzheimer's disease is the deposition of amyloid plaques throughout the brain, especially within the hippocampus and amygdala. Transgenic mice that overexpress the Swedish mutation of human amyloid precursor protein (hAPPswe; Tg2576) show age-dependent memory deficits in hippocampus-dependent learning tasks. However, the performance of aged Tg2576 mice in amygdala-dependent learning tasks has not been thoroughly assessed. We trained young (2 --4 mo) and old (16 --18 mo) Tg2576 and wild-type mice in a T-maze alternation task (hippocampus-dependent) and a Pavlovian fear-conditioning task (amygdala- and hippocampus-dependent). As previously reported, old Tg2576 mice showed impaired acquisition of rewarded alternation; none of these mice reached the criterion of at least five out of six correct responses over three consecutive days. In contrast, old Tg2576 mice showed normal levels of conditional freezing to an auditory conditional stimulus (CS) and acquired a contextual discrimination normally. However, when the salience of the fear-conditioning context was decreased, old (12 --14 mo) Tg2576 mice were impaired at acquiring fear to the conditioning context, but not to the tone CS. Histological examination of a subset of the mice verified the existence of amyloid plaques in the cortex, hippocampus, and amygdala of old, but not young, Tg2576 mice. Hence, learning and memory deficits in old Tg2576 mice are limited to hippocampus-dependent tasks, despite widespread amyloid deposition in cortex, hippocampus, and amygdala. Keywords: Introduction : The development of transgenic (Tg) mice that overexpress the Swedish mutation of human amyloid precursor protein (hAPPswe; Tg2576; ) has provided a springboard for research into the cellular and molecular bases and behavioral sequelae of Alzheimer's disease (AD). Of the many neuropathologies evident in AD, amyloid plaques derived from deposited amyloid beta-peptide (Abeta) and neurofibrillary tangles have received a great deal of attention. The former may be responsible for initiating a biochemical cascade that leads to cognitive decline in AD patients . Amyloid plaques in hAPPswe mouse brain, like those seen in human AD patients, lead to activated microglia , oxidative stress , and alterations in neurotransmitter systems ; all are potential mechanisms by which amyloid deposition disrupts normal neural function. Distribution of amyloid plaques is not uniform throughout the brains of Tg2576 mice ; rather, amyloid deposition is densely concentrated in the cortex, hippocampus, and amygdala. This is similar to the nature and pattern of the neuropathology seen in human AD patients. The anatomy of amyloid deposition in Tg2576 mice indicates that these mice might be deficient in forms of learning that depend on the hippocampus and amygdala, for example . Consistent with this hypothesis, learning and memory deficits have been reported in several hippocampus-dependent learning paradigms in Tg2576 mice. In most cases, these deficits are age-dependent and are only shown in senescent mice . Although aged Tg2576 mice are impaired at learning several tasks that depend on the hippocampus, the performance of these mice on tasks requiring an intact amygdala has not been thoroughly established . To address this issue, we trained wild-type and Tg2576 mice in a Pavlovian fear-conditioning paradigm that is both hippocampus- and amygdala-dependent . We hypothesized that old Tg2576 mice would be deficient compared with old wild-type and young mice in learning to fear a tone conditional stimulus (CS) and the context in which that CS was paired with an aversive footshock unconditional stimulus (US). We also expected that old Tg2576 mice would be deficient at learning a contextual discrimination, a more difficult task that requires both the hippocampus and amygdala. RESULTS : Histology | The photomicrograph in Figure shows representative Abeta plaque pathology in the hippocampus (Fig. A) and amygdala (Fig. B) of an old/APP+ mouse brain stained with Bielchowsky silver. Numerous plaques were found throughout these brain structures, compared with respective areas of an old/APP- mouse brain (Figs. C,D). Young mice, regardless of genotype, were used as negative controls and displayed no Abeta deposition (data not shown). Figure 1 | Amyloid deposition in Bielchowsky silver-stained sections of old mouse brains. Amyloid deposition in Bielchowsky silver-stained sections of old mouse brains. Plaque deposition (indicated by arrows) is distributed throughout the brains of old/APP+ mice, but is concentrated in the cortex, hippocampus (A), and amygdala (C), as compared with the same regions of the brains of old/APP- mice (B and D, respectively). Young/APP- and young/APP+ mice (data not shown) do not show any amyloid plaque deposition (magnification, 40x). Behavior | All of the mice used in these experiments, regardless of age and genotype, displayed a healthy appearance and normal behavior . During food deprivation, some mice became ill, which was evident by changes in body temperature (mice were cold to the touch and shaking), posture (crouching), and activity (very few movements, which were slow and labored). None of these behaviors were evident when the mice were allowed ad libitum food access. Mice that became ill were given a solution of sucrose in water and placed on a heating pad, and in the most serious cases, given a small subcutaneous injection of lactated Ringers solution. Experiment 1: Rewarded Alternation in the T-Maze | It has previously been shown that Tg2576 mice that overexpress hAPPswe are deficient at several hippocampus-dependent tasks . In the first experiment, we attempted to replicate those studies that have shown a deficit among old Tg2576 mice in rewarded alternation in a T-maze task. After a 4-d habituation period on the T-maze, mice were trained to alternate arm choices to receive a sucrose pellet reward. Training consisted of 6 trials per day for 14 d. Acquisition of rewarded alternation was retarded in old/APP+ mice. In fact, none of the mice in the old/APP+ group reached the behavioral criterion of at least 5 out of 6 correct responses over three consecutive training days . A Kruskall --Wallis test confirmed a significant difference in the rate at which the groups reached the criterion for learning the task (H = 12.2; p < 0.01). Post hoc comparisons indicated that old/APP+ mice were significantly impaired relative to young and old/APP- mice, as well as young/APP+ mice, which did not differ from one another (Mann-Whitney U, p < 0.05). Latencies to choose an arm of the T-maze were collapsed across forced and free trials into 2-d blocks. Running speeds were similar across groups, as there were no significant group differences in latency to make an arm choice (F(3,43) < 1; Fig. ) trials. The results of this experiment support previous findings of age-related deficits in Tg2576 mice on a rewarded T-maze alternation task . That the young/APP+ group was comprised entirely of female mice could suggest a sex effect between young/APP+ and old/APP+ mice. However, there were no sex differences in learning rates either within any of the other groups or across all animals (p > 0.05). Figure 2 | Old/Tg mice are deficient at learning a rewarded T-maze alternation task (Experiment 1). Old/Tg mice are deficient at learning a rewarded T-maze alternation task (Experiment 1). Mean +- SEM number of days for each group to reach the criterion for learning of at least 5 out of 6 correct responses on 3 consecutive days. There were no differences among young/APP- (solid open bar), young/APP+ (hatched open bar), and old/APP- (solid filled bar) groups in days required to reach criterion, or in the percentage of animals in each group that reached criterion. None of the old/APP+ mice (hatched filled bar) had reached criterion by the 14-d point, at which the experiment was stopped. Figure 3 | Arm choice latencies are similar across young and old APP+ and APP- mice (Experiment 1). Arm choice latencies are similar across young and old APP+ and APP- mice (Experiment 1). Mean +- SEM latency (in seconds) to make an arm choice across the 14 training days (collapsed into 2-d blocks). There were no group differences in arm choice latencies, indicating no effects of age or transgene on gross motor abilities. Experiment 2: Pavlovian Fear Conditioning | Mice that overexpress hAPPswe show elevated levels of Abeta40 and Abeta42 along with Abeta deposits in the cortex, hippocampus, and amygdala . Although several studies have reported a deficiency among old Tg2576 mice in learning hippocampus-dependent tasks, it has yet to be shown if these same mice have difficulty in amygdala-dependent tasks. To this end, we trained the mice used in Experiment 1 in a Pavlovian fear-conditioning task, which has been shown to be amygdala-dependent . The hippocampus also plays a role in fear conditioning to the places where aversive events occur, so-called contextual fear conditioning . Thus, we hypothesized that old Tg2576 mice, because of the deposition of amyloid plaques in the structures critical for learning Pavlovian conditional associations, would be deficient compared with old/APP- and young mice in conditional fear to both a discrete tone CS and to the context in which the CS was paired with an aversive footshock. For Pavlovian fear conditioning, mice received 3 or 5 tone --footshock trials in a novel conditioning chamber. Conditional freezing to the conditioning context and to the tone CS are shown in Figures , A and B, respectively. Conditional freezing was not different among the groups receiving 3 or 5 CS-US pairings, and these data were collapsed. Inspection of Figure reveals that old/APP+ mice were not deficient in the acquisition of Pavlovian fear conditioning to either auditory or contextual CSs. In fact, old mice showed more conditional freezing than young mice. There was a significant main effect of age (F(1,37) = 7.5; p < 0.01) in levels of freezing to the conditioning context during retrieval testing (Fig. A), and Fisher's PLSD confirmed that old mice froze more during the retrieval test than young mice (p < 0.05). Greater freezing among the old mice during context fear retention testing was probably nonassociative, insofar as old mice also showed greater levels of freezing prior to footshock on the conditioning day (F(3,37) = 19.8; p < 0.0001; data not shown). There were no group differences in retrieval of fear to the tone CS (F(3,37) < 1; Fig. B), and no age x genotype interactions in any of these effects (F < 1). Contrary to our hypotheses, old Tg2576 mice were unimpaired in acquiring auditory and contextual fear conditioning when compared with their littermate controls. Figure 4 | Amyloid deposition does not disrupt learning of conditional fear (Experiment 2). Amyloid deposition does not disrupt learning of conditional fear (Experiment 2). (A) Context test. Mean +- SEM percentage freezing across the 8-min context retrieval test in context A. Contrary to expectations, old/APP+ mice (filled squares) showed no deficits in conditional fear to the context in which fear conditioning had taken place 24 h prior as compared with young/APP- (open circles), young/APP+ (open squares), and old/APP- (filled circles) mice. (B) Tone test. Mean +- SEM percentage of freezing across the 10-min tone retrieval test. The continuous tone CS (horizontal bar) was turned on 2 min after the mice were placed in the testing chambers. Contrary to expectations, old/APP+ mice showed no deficits in conditional fear to the tone CS when tested 48 h after fear conditioning took place as compared with young/APP-, young/APP+, and old/APP- mice. Experiment 3: Decreased Context Salience and Pavlovian Fear Conditioning | Our fear-conditioning experiment failed to elucidate an effect of hAPPswe overexpression in old mice on conditional fear to either the training context or the tone CS. However, one recent report has shown impaired contextual fear learning in Tg2576 mice . One potential source for this discrepancy is that the conditioning context used by Dineley and colleagues supported much less conditional fear than the context we used in Experiment 2. Moreover, auditory fear was much higher than contextual fear in the experiments reported by Dineley and colleagues. This indicates that their tone CS was more salient than the contextual CS. The difference in salience in these cues allowed the tone CS to overshadow the context CS , resulting in a weaker context --US memory that may have been more easily disrupted by hippocampal pathology . We addressed this issue by fear-conditioning naive Tg2576 and wild-type mice in a context that produced less conditional and unconditional freezing than the training context in Experiment 2. We hypothesized that old Tg2576 mice would show impaired contextual fear conditioning when the context is overshadowed by the tone CS. Mice were fear conditioned as before, except that the conditioning context was modified to reduce its salience. Conditional freezing to the conditioning context and to the tone CS are shown in Figure , A and B, respectively. A one-way ANOVA revealed a significant effect of genotype on retrieval of fear to the training context (F(1,35) = 13; p = 0.001). Post hoc comparisons (p <= 0.05) indicated that APP+ mice froze less to the conditioning context than APP- mice (Fig. A). On the second testing day, there were no group differences in levels of freezing to the tone CS (F < 1; Fig. B). As hypothesized, APP+ mice showed impaired contextual conditioning compared with wild-type controls when the salience of the conditioning context was reduced. This deficit was not due to impaired performance of the freezing response, insofar as freezing to the auditory CS was normal. Figure 5 | Amyloid deposition disrupts fear conditioning to the training context but not to the tone CS when context salience is decreased (Experiment 3). Amyloid deposition disrupts fear conditioning to the training context but not to the tone CS when context salience is decreased (Experiment 3). (A) Context test. Mean +- SEM percentage freezing across the 8-min context retrieval test in context A. Old APP+ mice (open circles) show less fear to the context in which fear conditioning had taken place 24 h prior than old/APP- mice (filled circles). (B) Tone test. Mean +- SEM percentage of freezing across the 10-min tone retrieval test. The continuous tone CS (horizontal bar) was turned on 2 min after the mice were placed in the testing chambers. Old/APP+ mice showed no deficits in conditional fear to the tone CS when tested 48 h after fear conditioning took place as compared with old/APP- mice. Experiment 4: Pavlovian Fear Conditioning ---Context Discrimination | Recent reports indicate that contextual fear conditioning can be acquired by rats with hippocampal lesions under some conditions . Therefore, we also trained mice in a contextual discrimination, which has proved to be more hippocampal-dependent than standard contextual fear conditioning . We hypothesized that the old Tg2576 mice would show poor contextual discrimination. In the contextual discrimination task, mice were placed in two distinct contexts each day; an unsignaled footshock was presented in one of these two contexts (no auditory cues were used). Context discrimination across the nine training days is shown in Figure . All mice acquired a significant contextual discrimination. This was confirmed by a significant context x day interaction in the ANOVA (F(8,224) = 9.972; p < 0.0001). There was not a significant effect of age or genotype or an interaction of these variables, indicating intact contextual discrimination in APP+ mice. To more easily compare discrimination across the groups, a difference score was calculated by subtracting freezing in the no-shock context from that in the shock context . An analysis of these data revealed a significant interaction of age x training block (F(2,56) = 3.9; p < 0.05); Fisher's PLSD performed on the last 3-d block suggested that old mice were deficient at acquiring the discrimination (p < 0.05). However, there was no difference between old/APP+ and old/APP- mice on this measure. The old mice appeared to generalize fear across the contexts, which minimized the difference score between freezing in the "shock" and "safe" contexts. Because neither group of old mice reliably discriminated between the contexts, a floor effect could account for the failure to detect a deficit in the old/APP+ mice relative to the old/APP- mice. Nevertheless, it appears that hAPPswe overexpression does not impair the acquisition of a contextual discrimination. Figure 6 | Old mice show less context discrimination than young mice (Experiment 4). Old mice show less context discrimination than young mice (Experiment 4). Mean +- SEM percentage freezing averaged across the 3-min preshock period in the shock (open/filled symbols) and safe (gray symbols) contexts on each of the nine training days. Whereas all mice conditioned to the shock context at the same rate and to the same extent, this fear was generalized more to the safe context among old animals (B,D) than young animals (A,C). Figure 7 | Amyloid deposition does not affect learning on a contextual discrimination task (Experiment 4). Amyloid deposition does not affect learning on a contextual discrimination task (Experiment 4). Mean +- SEM difference scores (percentage freezing in shock context - percentage freezing in safe context) across the three 3-d blocks of context discrimination trials. Although old/APP+ mice (filled squares) were not different from old/APP- mice (filled circles), all old mice (filled symbols) were impaired at learning the discrimination as compared with the young mice (open symbols), which did not differ from one another. These results indicate that the deficiency among old mice in discriminating contexts is caused by overgeneralization of fear across contexts. DISCUSSION : In the present experiments, we aimed to further characterize the nature of the behavioral deficits in transgenic mice that overexpress hAPPswe. To this end, we tested Tg2576 mice on learning and memory tasks that require the hippocampus, amygdala, or both structures. Consistent with other reports, we found that Tg2576 mice showed a pronounced age-dependent deficit in the acquisition of rewarded alternation on a T-maze. This deficit was not caused by any obvious performance deficit in old/APP+ mice, as these mice showed choice latencies and motivations for reward that were similar to those of animals in the other groups. Interestingly, old/APP+ mice performed close to chance on choice trials ---they did not perseverate responses. This indicates that their poor performance on the task was due to a deficit in spatial working memory and not due to a predisposition to return to previously visited places or repeat motor responses. The deficit in rewarded alternation was extreme ---it was apparent even with a minimal delay between sample and choice trials. Consistent with their deficits in spatial alternation, old Tg2576 mice also showed deficits in contextual fear conditioning under some conditions. Although these mice acquired context fear when the context was salient (Experiment 1) or the only signal for shock (Experiment 4), they were deficient in the acquisition of background contextual fear . That is, contextual fear-conditioning deficits in old Tg2576 mice were apparent when a discrete tone CS overshadowed the contextual CS (Experiment 3). Importantly, Tg2576 mice were no different from wild-type controls in acquiring conditional fear to a tone CS. Collectively, this pattern of results indicates that pathology in the hippocampus, not the amygdala, is responsible for the contextual fear-conditioning deficits in Tg2576 mice. These results replicate and extend a recent report that also shows contextual fear-conditioning deficits in old Tg2576 mice . In this study, both 5- and 9-month-old Tg2576 mice showed impairments in contextual fear conditioning, and these impairments were overcome by additional training in the 5-month-old group but not in the 9-month-old group. Collectively, these results are consistent with the effects of hippocampal lesions on contextual fear conditioning in rats , and indicate that amyloid plaque pathology in the hippocampus may mimic the effects of lesions on contextual fear conditioning in mice. Interestingly, contextual fear-conditioning deficits have only been shown in Tg2576 mice of intermediate age (5 --14 mo). It is conceivable that, rather than being mediated by conditioning context salience, these deficits are age-dependent; that is, they are only apparent in middle-aged, but not young (2 --4 mo) or old (16 --18 mo) mice. The pattern of behavioral deficits we have observed in mice may be related to the progression of pathology in AD patients. That is, in humans, it has been shown that the neuropathological effects of AD are first observed in the perforant path-dentate granule cell synapses . Additionally, AD pathology specifically disconnects the hippocampus with its cholinergic afferents and efferent structures . Because the hippocampus shows pathology before other brain structures, it is conceivable that hippocampal-dependent learning is more sensitive to disruption in AD than forms of learning that are mediated by other brain regions, regardless of the eventual amyloid deposition levels in these other brain regions. Whether alterations in synaptic plasticity , excitotoxicity , or some other mechanism underlie the deficits in hippocampal learning is an issue that remains to be resolved. Of course, AD is also associated with marked neuronal loss in the amygdala . In AD, damage to the amygdala and its associated cortex increases the severity of memory deficits attributable to hippocampal damage , and normal enhancement of memory for emotional compared with neutral stimuli is disrupted (; cf. ). Impairment of emotional event memory in AD patients is related to the intensity of amygdalar damage . Additionally, AD patients report a decreased affective response to pain as well as a decreased ability to detect fear in others' faces . Furthermore, fear conditioning to discrete conditional stimuli is impaired in AD patients . These findings predict a deficit in emotional learning in old Tg2576 mice, which also show amygdaloid plaques, although these mice do not display the marked neuronal loss and neurofibrillary tangles characteristic of human AD . Nonetheless, although the old/APP+ mice in our study had severe deficits in T-maze learning, they acquired auditory fear and some forms of contextual fear normally. This indicates that amygdala pathology in old Tg2576 mice is not sufficient to yield deficits in fear conditioning. Indeed, amygdala damage must be complete to observe fear-conditioning deficits . Although old Tg2576 mice show impaired contextual fear conditioning under some conditions (Experiment 3), they were capable of learning a contextual discrimination (Experiment 4) and freezing in a salient context (Experiment 2). In view of the parallel deficits in spatial alternation in these animals, our results reiterate others' suggestions that spatial working memory and contextual learning and memory may be mediated by different neural systems. have reported that spatial learning and contextual conditioning are differentially sensitive to hippocampal lesions. Indeed, fear conditioning is much less reliant on cortical inputs to the hippocampus than spatial learning on the T-maze . Our results provide another dissociation between spatial learning (rewarded alternation) and contextual fear conditioning. In summary, the results of the present study confirm the learning and memory deficits that have previously been shown in transgenic mice overexpressing hAPPswe . Deficits in spatial working memory in hAPPswe mice are consistent with the memory deficits reported in other rodent models of AD . Additionally, the deficits we have found in contextual fear conditioning mirror those reported in a recent study . However, we have also found that old Tg2576 do not have a global deficit in contextual fear conditioning; they are only impaired when the context is overshadowed by a discrete CS. Surprisingly, we found no deficits in Tg2576 mice in auditory fear conditioning, despite the presence of dense amyloid plaque deposition in the amygdalae of aged Tg2576 mice. Therefore, further studies are necessary to fully understand the behavioral effects of amyloid neuropathology on learning and memory processes. Whereas tasks dependent on a functional hippocampus seem to be especially vulnerable to amyloid deposition, amygdala-dependent learning appears less sensitive. Further understanding of the behavioral effects of Alzheimer's disease and the neuropathology associated with it will lead us to more effective treatments for this devastating disease. MATERIALS AND METHODS : Experiment 1: Rewarded Alternation in the T-Maze | Subjects | The subjects used in Experiments 1 and 2 were 57 C57B6/SJL mice. Breeding and histological methods were performed as previously described . Young (2 --4 mo) and old (16 --18 mo) transgenic (Tg2576; APP K670N/M671L;) and wild-type (nontransgenic littermates) mice were used in this study, yielding the following groups (M = male; F = female; APP+ = Tg2576; APP- = wild type): young/APP- (n = 14; 3 M, 11 F), young/APP+ (n = 14; 0 M, 14 F), old/APP- (n = 16; 6 M, 10 F), and old/APP+ (n = 13; 9 M, 4 F). After arrival, the mice were individually housed in opaque plastic (Nalgene) cages on a 14/10-h light/dark cycle (lights on at 7:00 a.m.) and allowed ad libitum access to food and water for 2 wk prior to any behavioral testing. The experiments were performed in two replications, with different sets of mice for each replication, in a counterbalanced order. Mice received both T-maze training (Experiment 1) and fear conditioning (Experiment 2), although not all subjects completed both tasks. For the first replication, mice received T-maze training prior to fear conditioning; in the second replication, the order of the training procedures was reversed. Approximately 2 wk elapsed between experiments within each replication. Because several mice died before completing both experiments, the number of subjects in each group for each experiment is different. Behavioral Apparatus | The mice were trained on a rewarded alternation task using a conventional T-maze. The maze was constructed of clear Plexiglas walls (20 cm tall) and a black Plexiglas floor. The arms and stem of the T were each 45 cm long by 10 cm wide. Clear Plexiglas guillotine doors were placed at the entrance to each of the arms. A third guillotine door was placed 10 cm from the base of the T and defined the start box. At the end of each of the goal arms was a small well into which food rewards (~10-mg sucrose pellets; Noyes, Inc.) were placed. The maze was mounted on a cart (~90 cm tall) that was located in the center of the mouse colony room. Extra-maze cues included the rack of mouse cages, a door, sink, and computer, the experimenter, and several posters on the walls. Procedure | Before T-maze testing began, the mice were placed on a food-deprivation schedule, in which food access was gradually reduced from ~3.5 g/d to ~2.5 g/d over 3 d. Both young and old Tg2576 mice show a hypersensitive hypothalamic --pituitary --adrenal response to stressors such as restraint and food deprivation; in the case of the latter, this response often leads to severe hypoglycemia and mortality . Of the original 57 mice, 48 completed the T-maze training, in the following groups: young/APP- (n = 12; 3 M, 9 F), young/APP+ (n = 13; 0 M, 13 F), old/APP- (n = 13; 4 M, 9 F), and old/APP+ (n = 10; 8 M, 2 F). The mice were given 4 d to habituate to the maze, in which they were allowed to explore the apparatus for 5 min/d or until reward pellets scattered throughout the maze were all consumed. On each of the 14 training days, mice were run in squads of 3 --4 and given 6 pairs of training trials. The first trial of each pair was a forced trial, in which one of the goal arm guillotine doors was closed and the mouse was constrained to selecting the opposite arm; the order of forced choices was pseudorandomly selected so that each mouse was forced to traverse each arm three times per day, but on no more than two trials in a row were they forced down the same arm. The mouse was returned to the start box 15 --20 sec after consuming the reward. On the second, or free-choice trial, both goal arm doors were opened, but only the arm opposite the one selected in the forced trial was baited. Latency to make an arm choice from the time the start box guillotine door was opened was also recorded during the free and forced trials. The criterion for a mouse having learned the rewarded alternation task was 3 consecutive days of at least 5 correct responses out of the 6 free trials . For example, a mouse that correctly alternated on 5 out of 6 free trials on training days 5, 6, and 7 was given a value of 7. Training on the T-maze lasted for 14 d. After the completion of the T-maze learning task, the mice were returned to ad libitum food and water availability for at least 2 wk prior to any further experimental testing or sacrifice. Data Analysis | Mice that did not reach criterion before training was terminated were assigned a score of 14 for days to reach criterion. Nonparametric statistics were used to analyze these data. A Kruskall --Wallis test was performed on the average number of days the mice in each group took to reach criterion performance. After a significant H value was found, post hoc comparisons in the form of Mann --Whitney U were performed. Also reported is the percentage of mice from each group that reached the criterion level of performance. An ANOVA was performed on the latencies to make an arm choice after the start of each trial over the course of training for both forced and free trials. All data are presented as means +- SEMs. Experiment 2: Pavlovian Fear Conditioning | Subjects | Of the 57 mice at the beginning of the experiment, a total of 49 survived to the end of the fear-conditioning procedure, in the following groups: old/APP+ (n = 9; 8 M, 1 F), young/APP+ (n = 14; 0 M, 14 F), young/APP- (n = 13; 3 M, 10 F), and old/APP- (n = 13; 4 M, 9 F). The mice were housed as described in Experiment 1 and were maintained on ad libitum food and water access throughout this experiment. Behavioral Apparatus | Fear conditioning took place in four identical observation chambers (30 x 24 x 21 cm; MED-Associates, Inc.). The chambers were constructed from aluminum (sidewalls) and Plexiglas (rear wall, ceiling, and hinged front door) and were situated in sound-attenuating cabinets located in a brightly lit and isolated room. The floor of each chamber consisted of 36 stainless steel rods (3 mm in diameter) spaced 8 mm apart (center-to-center). Rods were wired to a shock source and solid-state grid scrambler (MED-Associates, Inc.) for the delivery of footshock USs. A speaker mounted outside a grating in one wall of the chamber was used for the delivery of acoustic CSs. A 15-W house light was mounted on the opposite wall. The chambers were cleaned with a 5% ammonium hydroxide solution, and stainless steel pans containing a thin film of the same solution were placed underneath the grid floors to provide a distinct odor before the mice were placed inside. Ventilation fans in each cabinet supplied background noise (65 dB; A scale). This chamber configuration is subsequently referred to as context A. Procedure | Mice were conditioned and then tested for fear to the conditioning context and to the tone CS on the two subsequent days. For fear conditioning, mice were transported in squads of 2 --4 (counterbalanced for age and genotype) and placed in the conditioning chambers; the chamber position was counterbalanced for each squad. The mice received 3 or 5 tone (10 sec, 5 kHz, 85 dB) --footshock (0.5 sec, 0.5 mA) trials (74-sec intertrial interval) beginning 128 sec after being placed in the chambers. Sixty-four seconds after the final shock, the mice were returned to their home cages. Twenty-four hours after the conditioning session, the mice were returned to the conditioning chambers in squads of 2 --4 animals and tested for fear to the context in which they had received the tone --footshock trials. This test for fear to the context lasted 512 sec, after which the mice were returned to their home cages. Twenty-four hours after the context test, the mice were tested for fear to the tone CS in context B. This context consisted of the same chambers used for context A; however, the room lights and chamber houselights were turned off (three 40-W red lights provided illumination). In addition, the doors on the sound-attenuating cabinets were closed, the ventilation fans were turned off, and the chambers were cleaned with a 1% acetic acid solution. To provide a distinct odor, stainless steel pans containing a thin film of acetic acid were placed underneath the grid floors before the mice were placed inside. In this context, the mice received a single continuous tone (512 sec, 5 kHz, 85 dB) beginning 128 sec after being placed in the chambers. Fear to the context and to the tone CS was assessed by measuring freezing behavior . The output from video cameras mounted above each chamber was fed into a video processor (Robot), and the mice were videotaped throughout each of the sessions. Freezing behavior, defined as the absence of all movement except for that necessitated by breathing, was scored using a time-sampling procedure by an experimenter who was blind to the ages and genotypes of the mice. Measurements were made every 8 sec for each mouse, yielding 64 observations during the 512-sec test. Freezing was quantified by computing the percentage of observations in which the mouse had been scored as freezing during the test. Approximately 2 wk passed between fear-conditioning testing and any further behavioral testing or sacrifice. Histology | Approximately 2 wk after the completion of all behavioral testing, the mice were killed by cervical dislocation, and their brains were removed and hemisected saggitally. The right half of each brain was fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS), pH 7.4, at 4C and processed for paraffin embedding. The paraformaldehyde-fixed hemibrains were sectioned coronally at 5 --8 mum thickness, and sections were processed for hematoxylin and eosin, Bielchowsky silver, and Congo Red staining. Data Analysis | For each session, the freezing data were transformed to a percentage of total observations, a probability estimate that is amenable to analysis with parametric statistics. These probability estimates of freezing were analyzed using ANOVA. Post hoc comparisons in the form of Fisher's PLSD tests were performed after a significant omnibus F ratio. All data are represented as means +- SEMs. Experiment 3: Decreased Context Salience And Pavlovian Fear Conditioning | Subjects | The subjects were 37 naive mice obtained and housed as described in Experiment 1. At the beginning of training, all mice were 12 --14 months old. The two groups were: APP+ (n = 15; 7 M, 8 F) and APP- (n = 22; 8 M, 14 F). Throughout this experiment, the mice were maintained on an ad libitum regimen of food and water availability. Behavioral Apparatus | The apparatus used in this experiment was identical to that used in Experiment 2, except that the 1% acetic acid solution was used to clean and provide a distinct odor in context A and a 1% ammonium hydroxide solution was used in context B. Procedure | The mice were fear conditioned as described in Experiment 2. All mice received 3 tone --shock pairings in context A. Testing for fear to the conditioning context and to the tone CS was performed on the two subsequent days as described in Experiment 2. Histology | After behavioral testing, all mice were retained for further testing. Therefore, we have no histological data for these mice. Data Analysis | Freezing behavior was measured and analyzed as described in Experiment 2. Experiment 4: Pavlovian Fear Conditioning ---Context Discrimination | Subjects | The subjects were 32 naive mice obtained and housed as described in Experiment 1. At the beginning of training, young mice were 2.5 months old and the old mice were 17 months old. The four groups were: young/APP- (n = 8; 4 M, 4 F), young/APP+ (n = 8; 4 M, 4 F), old/APP- (n = 9; 5 M, 4 F), and old/APP+ (n = 7; 5 M, 2 F). Throughout this experiment, the mice were maintained on an ad libitum regimen of food and water availability. Behavioral Apparatus | The apparatus used in this experiment was identical to that used in Experiment 2. In Experiment 2, we observed high levels of freezing among old animals in context A prior to any fear conditioning, which could possibly have resulted from some nonassociative fear to this context (in context B there was relatively little freezing before the tone CS was presented). To mitigate this effect, and to equate the salience of the two contexts, the configurations of contexts A and B were adjusted. After these changes, context A consisted of the conditioning chambers with grid floors placed in sound-attenuating cabinets. The chambers were lit by a 15-W houselight, and ventilation fans provided background noise. Instead of ammonium hydroxide, a 70% ethanol solution was used to clean the cages and provide the distinct odor. In context B, the same conditioning chambers were used. However, the red lights were no longer used, and the houselights in the chambers were turned on. The ventilation fans and room lights remained off, and the 1% acetic acid solution was used to clean the chambers and provide a distinct odor. Procedure | On each of 9 d, the mice were brought in squads of 4 (counterbalanced for age and genotype) and spent 256 sec in both context A and context B; the order of context exposure was counterbalanced. In context A, half of the squads received a single footshock (1 sec, 0.75 mA) 192 sec after being placed in the chambers, whereas the other half received the shock in context B. In the "safe" (no shock) context, the mice were simply exposed to the context for 256 sec. Approximately 4 h elapsed between placement in the two contexts each day. Histology | Histological verification of amyloid plaques was performed as described in Experiment 2. Data Analysis | Freezing behavior was measured and analyzed as described in Experiment 2. In this experiment, percentage of time spent freezing in the safe context was subtracted from percentage of time spent freezing in the shock context, yielding a difference score that was used as an index of discrimination between the two contexts. Difference scores were collapsed into 3-d blocks, and an ANOVA was performed on the blocked data. Post hoc comparisons in the form of Fisher's PLSD were performed after a significant omnibus F ratio was observed. Backmatter: PMID- 12359835 TI - Plasticity of the Human Auditory Cortex Induced by Discrimination Learning of Non-Native, Mora-Timed Contrasts of the Japanese Language AB - In this magnetoencephalographic (MEG) study, we examined with high temporal resolution the traces of learning in the speech-dominant left-hemispheric auditory cortex as a function of newly trained mora-timing. In Japanese, the "mora" is a temporal unit that divides words into almost isochronous segments (e.g., na-ka-mu-ra and to-o-kyo-o each comprises four mora). Changes in the brain responses of a group of German and Japanese subjects to differences in the mora structure of Japanese words were compared. German subjects performed a discrimination training in 10 sessions of 1.5 h each day. They learned to discriminate Japanese pairs of words (in a consonant, anni --ani; and a vowel, kiyo --kyo, condition), where the second word was shortened by one mora in eight steps of 15 msec each. A significant increase in learning performance, as reflected by behavioral measures, was observed, accompanied by a significant increase of the amplitude of the Mismatch Negativity Field (MMF). The German subjects' hit rate for detecting durational deviants increased by up to 35%. Reaction times and MMF latencies decreased significantly across training sessions. Japanese subjects showed a more sensitive MMF to smaller differences. Thus, even in young adults, perceptual learning of non-native mora-timing occurs rapidly and deeply. The enhanced behavioral and neurophysiological sensitivity found after training indicates a strong relationship between learning and (plastic) changes in the cortical substrate. Keywords: Introduction : Our brain is a highly adaptive learning device, reorganizing itself in accordance with environmental constraints. Nevertheless, adult people experience difficulties when learning a language that has a completely new and unfamiliar phonemic structure. Language perception is altered by the linguistic experience we make throughout our life. We perceive and produce speech through the filter of our native language. Whereas neonates are able to discriminate most phonetic categories in the first months of life , they develop prototypical phoneme representations of their native language in the first year of life . Exposure to a specific language in the first half year of life directs infants' phonetic perception toward this language and forms prototypes for each category. A prototype functions like a perceptual magnet for all members of that category and attracts all variations of a phoneme into "native" perception . Once established, these phoneme categories form a stable basis for consecutive speech perception and speech production . As a consequence, adults have difficulties in distinguishing non-native phonemic contrasts. For instance, Finnish and Estonian adults perceive their native phoneme prototypes more sensitively than non-native phonemes . Japanese listeners have difficulties in distinguishing /l/ and /r/ and perceive both phonetic categories as one . Nevertheless, after extended training, adult Japanese listeners improved their perceptual identification as well as their production performance of English /r/ and /l/ . In another experiment, the behavioral training of just-perceptible differences in speech stimuli resulted in a significant change in the duration and magnitude of the cortical potentials . It is generally assumed that representational maps in the brain are subject to plastic changes subsequent to altering sensory input. Intensive frequency-discrimination training enhances the area of representation of the trained frequency range in the primary auditory cortex of owl monkeys . In humans, changes in the perceptual acuity of frequency discrimination correlate with enhancements in the neuromagnetic responses to these frequencies . Moreover, intensive phonetic experience of speech sound discrimination is apt to alter the neural activity that underlies coding of these events in infants , as well as in adults . showed that learning a new language is accompanied by the development of new cortical representations for unknown phonemes of the new language. Hungarians, who learned to speak Finnish fluently in adulthood, developed cortical memory representations for Finnish phoneme categories that do not exist in their native language. Phonemes are the smallest units of speech that affect meaning. A small difference in the phonemic structure of a word entirely changes its meaning. Each language has its own "music," that is, languages differ in their rhythm and stress patterns. Language has a certain rhythm or timing because phonological units (i.e., syllables or segments like moras) are organized into rhythmic sequences. Whereas English is generally considered as a stress-timed language and French as a syllable-timed language, Japanese and even Finnish are often cited as mora-timed languages . In Japanese, the "mora" is a temporal unit that divides words into almost isochronous segments (e.g., na-ka-mu-ra and to-o-kyo-o each comprises 4 mora). For Japanese and Finnish, the relative duration of a vowel or consonant is a decisive feature for differentiating among meaningful words. Nevertheless, in stress-timed languages, for instance in English, the difference between heed and hid, bead and bid, or wooed and wood is sometimes only encoded by the length of the vowel . In the German language, vowel duration can also be decisive for the meaning of a word (as in biete and bitte, Miete and Mitte), but is additionally accompanied by a different accentuation or consonant duration that facilitates the perception of the difference. However, in languages such as Finnish, Estonian, Hungarian, and Japanese, every short vowel has a corresponding long counterpart. Similarly, the duration of some consonants has both a short and long variant. These durational contrasts are used to encode different meanings. For instance, in Finnish, all variants of these words, tule --tuule --tuulle --tuulee --tulee --tullee --tulle --tuullee, are to be distinguished accurately by their duration quantity because they encode different meanings. Mora-timing might also be relevant in Finnish . Enhanced processing of speech sounds, especially for duration differences as compared with equivalent changes in tones, was found for Finnish speakers . The Hungarian language also has many long/short differences to be distinguished (e.g., fulel --fullel, sor --so:r, megy --meggy). Recently, the moraic/nonmoraic distinction was postulated even in natural spoken Danish . Japanese is a mora-timed language par excellence ; its temporal units consist mostly of consonant --vowel (CV or CVV) syllables, single vowels, or the nasal /n/. Each unit takes about the same length of time . Long syllables (ko-o) may consist of two moras, but short syllables (ko) only consist of one. These units are expressed in the two phonetic alphabets hiragana and katakana with maximally 111 possible Japanese CV (ka, ba, ra) or CyV monomoraic (kyo, byo, ryo) syllables, five vowels, and the consonant /n/, and form the entire set of elementary phonetic units of Japanese. Thus, the rhythm of Japanese is controlled by a regular temporal sequence, mostly isochronous. The Japanese language has a restricted number of moras, which results in a large number of homonyms. These homonyms are often differentiated only by the duration of a mora when the context is not available. For instance, the sound sequence /hoshu/ exists in four variants: with a long /o/, with a long /u/, with both a long /o/ and a long /u/, or with both a short /o/ and a short /u/. The words hoshu, hooshu, hoshuu, and hooshuu have completely different meanings: hoshu means "a catcher" or "conservative"; hooshu means "a gunner" or "artilleryman"; hoshuu means "mend, repair"; and hooshuu means "remuneration, reward." The same is true for a large number of other Japanese words. Some of these distinctions as in ojiisan (grandfather) and ojisan (uncle) or obaasan (grandmother) and obasan (aunt) can lead to embarrassing misunderstandings, if the duration of the vowel is not considered precisely. studied the role of mora-timing in the recognition of spoken Japanese words. They showed that Japanese listeners are sensitive to the moraic structure of speech and that it is easier for them to respond to moras than to phonemes. even concluded that Japanese listeners parse the speech signal into units larger than phonemes, namely, moras, because they mainly use the prevailing moraic consonant --vowel (CV) unit in current speech. The mora-unspecific VC cues are only used in natural VCV words where the consonant is the onset of a CV mora. This has led to several questions: (1) What are the effects of long-term experience with one's native language, when this language affords accurate temporal discrimination? (2) What kind of changes can be produced by intensive auditory discrimination training of non-native durational differences, and what are the related neurophysiological correlates? (3) What is the difference between native and non-native perception and discrimination learning of durational differences within one mora? To answer these questions, we compared the perception and processing of Japanese words in German and Japanese subjects. The duration quantity in bi- and trimoraic words was manipulated in 8 steps over the length of one mora. German subjects (test group) were trained to discriminate these differences. The control group was composed of Japanese native speakers. Accurate perception of non-native phonemic categories can be quantified by means of an automatic, preattentive "change detector" component of the Auditory Evoked Response (AER), called the Mismatch Negativity (MMN) and its magnetic counterpart, the Mismatch Field (MMF; ; ). Phonemic memory traces, revealed by the MMN/MMF, have shown language specificity by showing that there is an enhancement of MMN/MMF for native but not for non-native phoneme categories . During the processing of speech sounds, the amplitude of the MMN is higher in the left than the right hemisphere . Recently, MMN was found to be larger for words than for pseudowords . There is increasing evidence that speech and language in right-handed subjects are processed and produced preferentially by the left cerebral hemisphere . Also, a left hemispheric dominance for speech has been shown for a majority of left-handed subjects . For these reasons, the present study's recordings were carried out over the left temporal cortex. There were three objectives for the present magnetoencephalographic (MEG) study. The first objective was to assess if learning and experience induce cortical plasticity in the "phonemotopic" maps of the brain. The second objective was to determine if intensive discrimination training of very small differences in non-native speech affects the discrimination acuity. The third objective was to measure the affect of discrimination training on the brain's automatic change-detection process as reflected by the MMF or other neuromagnetic components. A group of German subjects, unfamiliar with Japanese, were trained to discriminate eight durational variants of four Japanese pairs of words (anni --ani, itte --ite, kiyo --kyo, and kiyou --kyou). The length of a consonant or a vowel was varied in eight small durational steps. Subjects performed 10 training sessions. To test the behavioral discrimination performance, each training session was accompanied by discrimination tests without feedback. Before the first and after the last training session, MEG recordings using three of the eight duration variants were performed. A group of Japanese subjects received two consecutive training sessions and were tested in the same manner as the German subjects. This was done to test whether the cortical networks that encode speech duration in Japanese native speakers were also subject to short-term plasticity changes, or whether these well-established representations in the cortical maps are not subject to further reorganization processes. RESULTS : Behavioral Results | Reaction Time | Reaction time (RT) was recorded in all training sessions and referred to the onset of the second stimulus of each stimulus pair. Within German test subjects, training resulted in a clear decrease in RT. After longer training, subjects responded faster to pairs of standard stimuli, as well as to pairs of standard-deviant stimuli (Fig. a). The logarithmic function characterizing the decrease in RTs over 10 training sessions showed smaller values for the standard as compared with the deviant stimuli. The corresponding equation for the results of the standards was y = -0.0366 ln(x) + 0.7667, with R2 = 0.9274, and for the results of the deviants was y = -0.027 ln(x) + 0.7746, with R2 = 0.8623, where R2 represents the variance in RT attributable to the variance in training days. These equations show that there is a smaller decrease for deviants than standards. German subjects' RTs to deviant stimuli were, on average, 21.4 msec (SEM = 0.54 msec) longer than their RTs to standards (Fig. a). From the first to the last training session, a decrease of 97 msec for standard stimuli and of 76 msec for deviant stimuli was obtained. These decreases for both standards and deviants were significant (P = 0.0026 for standards and P = 0.0038 for deviants), as shown by a paired two-tailed Student's t-test. The RTs of the Japanese subjects also decreased significantly from the first to the second training session for each stimulus condition (P = 0.0221 for standards and P = 0.0595 for deviants in a paired two-tailed Student's t-test) by 20.86 msec (SEM = 9.75), on average (Fig. b). The results from Japanese subjects were similar to those from German subjects, showing that RTs for deviant stimuli were longer than for standard stimuli, in the first (by 29.65 msec, SEM = 6.45), and in the second training session (by 40.45 msec, SEM = 8.75). Statistical comparison (unpaired 2-tailed t-test) of RTs in the first and second training sessions between Japanese and German subjects revealed no significant differences. Both groups started with similar RTs, but German subjects showed a markedly greater decrease from the first to the second training session (46 msec for standards and 37 msec for deviants) in comparison with Japanese subjects (26 msec for standards and 15 msec for deviants). Figure 1 | (a) Grand averaged RTs and SEM across German subjects for each training session. (a) Grand averaged RTs and SEM across German subjects for each training session. A constant decrease approximated by a logarithmic trend (thick line for deviants, thin line for standards) occurs from training day 1 to 10. (b) Grand averaged RTs and SEM of Japanese subjects for each stimulus condition. Categorization | In natural speech, humans are able to realize the variation in the duration of a speech signal and make categorical judgments. Perception of speech occurs categorically rather than as a matter of continuous, monotonic variation . In a separate test, 10 subjects classified eight variants of four stimuli in a two-alternative-forced-choice task as being "long" or "short." The purpose of this test was to determine whether subjects form phonetic categories in a long versus short manner or whether they recognize the differences in duration in a monotonic increasing or decreasing way. The results of the categorization test are illustrated in Figure . A set of 120 stimuli of each condition of stimuli was presented in a semirandomized order with an equal number of each difference variation. Subjects perceived the long versus short version of the stimuli in a categorical, nonmonotonic manner. Most subjects recognized the variants 1 --3 as long and 6 --8 as short. During the variants 4 --5, there was a clear change in the recognition pattern, pointing to a categorical perception at this duration. Nevertheless, some subjects (S4, S6, S9) varied these category boundaries to more extreme positions. They seemed to construct wider categories than the other subjects. Figure 2 | Category boundaries for one exceptional subject (S4) with a varying duration concept, a subject (S8) with a nonmonotonic category boundary between the long and short variants of stimuli, and the averaged responses. Category boundaries for one exceptional subject (S4) with a varying duration concept, a subject (S8) with a nonmonotonic category boundary between the long and short variants of stimuli, and the averaged responses. A clear category boundary is observed between differences 4 and 5. Discrimination Performance | Improvements in the training occurred rapidly (already apparent in the first training session) and evolved continuously until reaching an asymptotic saturation level. This is shown by the hit rate, or true positive rate, in Figure . Before training, German subjects discriminated the deviant stimuli differing in consonant duration (anni, itte) at a hit rate of 43% --45% in tests without feedback. They improved their performance by up to 35%, reaching a level of similar80% at the end of training. For deviant stimuli differing in vowel duration (kiyo, kiyou), the starting discrimination performance began at 55% --63% and improved by 20% --27%, reaching a level of similar80% at the end of training. Initial discrimination was better for vowel duration than for consonant duration, and reached a higher end level after training. Therefore, it was easier for German subjects to detect differences in vowel duration than in consonant duration. The correct recognition of pairs of standard stimuli was very high from the first training session, and reached a saturation level by the second or third training session. In training sessions with feedback, improvement of hit-rate performance grew continuously by similar15% for each stimulus, indicating that training performance was independent of stimulus type, but initial performance differed because of stimulus difficulty. Japanese subjects started to discriminate deviants at a higher performance level than German subjects (average hit rate = 61%). After only two training sessions, Japanese subjects' hit-rate performance improved by up to 7% for consonant duration (anni, itte) but remained almost unchanged for vowel duration (kiyo, kiyou). Figure 3 | (a) Hit rates of German subjects for deviants and standards in tests without feedback and in (b) 10 training sessions with feedback. (a) Hit rates of German subjects for deviants and standards in tests without feedback and in (b) 10 training sessions with feedback. (c) Hit rates of Japanese subjects in 2 tests (each condition left) and trainings (each condition right). Psychometric functions of the hit-rate performance for German subjects during training are illustrated in Figure . The largest increase in the percentage of correctly recognized deviants occurred in the variants 2 --4 (Deltat = 30 --60 msec). Small increases were found for variants that were very similar to the standards (variant 1 being the smallest difference, Deltat = 15 msec). In variants 5 --7 (Deltat = 75 --105 msec), which differed greatly from the standards (variant 7 being the largest difference, Deltat = 105 msec), the best performance was reached early and was maintained until the end of training. The middle difference variant 4 (Deltat = 60 msec), where the phoneme boundary is expected to be, showed improvement throughout the training. This improvement was 30% --45% for /anni/, /itte/, and /kiyou/; whereas for /kiyo/, which is less difficult to discriminate, the improvement was <10%. These findings indicate that the categorical boundary, determined by a hit rate of 50%, was at the beginning of the training for consonants (/anni/, /itte/) at the difference variant 4, and for vowels (/kiyo/, /kiyou/) at the difference variant 3. Over the course of the training this boundary shifted for consonants by about 2 (Deltat = 30 msec) and for vowels by about 1.5 (Deltat = 22 msec) difference degrees nearer to the standard. This shift is depicted in Figure . The fitted psychometric curves develop from a rather flat, logistic function before training to a higher and steeper function at the end of training. The areas between the two fitting curves map the "learning gain" throughout training, that is, the hit rates of the small difference degrees significantly increased after training. These results are paralleled by the performance obtained in the training sessions. The steep decrease in difference magnitude for the first three training sessions was followed by a slow, but constant improvement until the end of the training. Figure 4 | Psychometric functions of German and Japanese subjects for each deviant stimulus in the tests before (thin line) and after (thick line) training. Psychometric functions of German and Japanese subjects for each deviant stimulus in the tests before (thin line) and after (thick line) training. Before training, the deviant variant 4 exceeds the random level (50%). After training in each condition, the deviant variant 2 is recognized as different. German subjects improved maximally in the difference variants 2 --4 from the first test before training to the last test after training. Japanese subjects already recognized the deviant variant 2 in almost each condition above random level. Evoked Neuromagnetic Responses | Amplitude Differences | To assess differences in perceptual abilities and the effects of training, the results from a group of native Japanese speakers were compared with those from a non-native German group. The assumption was that language-dependent memory traces exist and will be activated in the native but not in the non-native group. Effects of training in the non-native group should reveal the "trainability" of these memory traces. This hypothesis is based on the notion of language-dependent memory traces that are activated when the stimuli are perceived as native speech and not activated when the stimuli are perceived as non-native, "phonetic" material. Because the mismatch response is considered to be an automatic, preattentive response to stimulus change, it should be sensitive to the ability to discriminate fine perceptual differences. For evaluation of the mismatch fields, the averaged responses to the standard stimuli were subtracted from averaged responses to deviant stimuli. A source analysis based on a moving equivalent current dipole (ECD) in a spherical model fitted to the head surface was performed separately for the derived (deviant minus standard) waveforms as well as for the standard and deviant responses for each subject and condition. The origin of the head-based coordinate system was set at the midpoint of the mediolateral axis (y-axis). The Cartesian coordinates (x, y, z) of the source location, (the posteroanterior x-axis connecting the inion and nasion, positive toward nasion; the mediolateral y-axis connecting the two center points of the entrances to the acoustic meatuses of the left and the right ear, positive toward the left ear; and the superoinferior z-axis projecting perpendicular to the x --y plane, positive toward the vertex), the orientation of the ECD (phix, phiy, phiz), as well as the global field power or root mean square (RMS) amplitudes and the source strength (Q) were calculated. When the goodness of fit between calculated and measured waveforms was >90%, these parameters were used to calculate individual source waveforms for each stimulus condition under the assumption that the source space projection (SSP) on the supratemporal plane explains most of the field variance. Thus, the MMFs to the deviant responses established with the method described above appear to originate from distinct neural representations in the supratemporal plane (see Table ). Table 1 | Source Coordinates of Responses to Standard and Deviant Stimuli Because of the anatomical differences between subjects and restrictions of the method, these source coordinates have to be judged carefully. The location and orientation of the generators of responses to deviant stimuli (MMF) were calculated and compared with those of the standard stimuli. The standard stimuli were located mainly in secondary auditory areas (i.e., planum temporale). In contrast, the main sources of the MMF were found to be located more medial and more anterior to the sources of the standard stimuli. With respect to the x, y, and z coordinates, Student's t-tests with a corrected alpha level of 0.008 revealed no significant differences for pre- versus posttraining localizations or for Japanese versus German subjects. However, Japanese subjects had their sources consistently more anterior, more lateral, and more inferior than did German subjects. On average, the MMF sources were found consistently more anterior, more medial, and more inferior than the corresponding sources of the standards. The MMF for /anni/ is detected in the time window from 370 to 450 msec and for /kiyo/ from 170 to 250 msec after the stimulus onset. The MMF is also detected for both /anni/ and /kiyo/ in the time window from 150 to 230 msec after the "point of uniqueness" of the stimuli, which is the time point within the deviant stimulus from where the response is unequivocally identified as different from that of the standard stimulus (see Fig. ). Figure 5 | MMF amplitudes of one exemplary German subject. MMF amplitudes of one exemplary German subject. A remarkable increase of the MMF amplitudes is seen after training (second and fourth rows) compared with the amplitudes before training (first and third rows). The global field power (or RMS) of the rare deviants as compared with that of the frequent standard stimuli is shown in Figure for all conditions and groups. Before and after training, Japanese subjects showed a "narrower" time window of the derived MMF amplitudes in the first and second MEG sessions as compared with the time window for the German subjects. The responses to duration deviants in /kiyo/ increased in German subjects after training, but showed a flatter, more scattered MMF in Japanese subjects as compared with German subjects. After only two training sessions, the MMFs of Japanese subjects were clearer and concentrated in a narrower time band. Figure 6 | (Left) Grand averaged responses of German subjects before and after 10 d of training in three difference degrees of consonant and vowel duration. (Left) Grand averaged responses of German subjects before and after 10 d of training in three difference degrees of consonant and vowel duration. (Right) Japanese subjects before and after two training sessions. German subjects elicited higher responses before training than Japanese subjects and showed a remarkable improvement throughout training. In Japanese subjects the responses did not change much in the course of the training, but showed a narrowed MMF response compared with the German subjects. Thick lines denote the deviant, thin lines the standard response. The time domain marked by the two thin vertical lines denotes the area of the MMF. The MMF was originally defined as a component resulting from the difference between responses to (frequent) standard stimuli and (rare) deviant stimuli. For statistical analysis, the peak RMS values of the subtraction waveforms were calculated within a latency window of 20 msec. A repeated-measure analysis of variance (ANOVA) was separately calculated for each group of subjects (German vs. Japanese) with the factors "training" (before/after), "phonetic category" (anni/kiyo) and "difference degree" (8, 6, 4) as the repeated within-subjects factors. For German subjects, who performed the extended training, a significant overall effect of training (F(1,9) = 15.772, P = 0.0032) was observed. The interaction of training and phonetic category (anni/kiyo) also reached significance (F(1,9) = 8.982, P = 0.015). The main effect for the phonetic category (anni/kiyo) was highly significant (F(1,9) = 34.35, P = 0.0002), and, typically for MMN studies, the difference degrees (8, 6, 4) evoked highly differential responses (F(2,9) = 71.11, P < 0.0001). Within-group comparisons of the training effect (before/after) were significant for anni8 (F(1,9) = 8.42, P = 0.017), for kiyo8 (F(1,9) = 10.25, P = 0.01), kiyo6 (F(1,9) = 13.43, P = 0.005), and kiyo4 (F(1,9) = 14.83, P = 0.003). In conditions anni6 and anni4 the differences, though not significant, also showed an overall increase. For Japanese subjects, only the difference degrees (8, 6, 4) reached significance (F(2,9) = 11.00, P = 0.0008). No significant training effects were observed. Comparisons between groups (German vs. Japanese) showed significant differences for anni8 (F(1,18) = 5.99, P = 0.02), for kiyo8 (F(1,18) = 6.82, P = 0.017), for kiyo6 (F(1,18) = 7.76, P = 0.012), and for kiyo4 (F(1,18) = 5.64, P = 0.028). Thus, in summary, German subjects showed higher MMF amplitudes after training than before training, whereas Japanese subjects (in four of six cases) started at a significantly higher MMF level and maintained this level throughout the training. The grand-averaged global field power of the MMF peak amplitudes in the left part of Figure illustrates these effects. Figure 7 | (Left) Grand average of the Global Field Power of the MMF (mean, SEM) of German (circle) and Japanese (square) subjects before and after training for the consonant condition in /anni/ (left) and the vowel condition in /kiyo/ (right). (Left) Grand average of the Global Field Power of the MMF (mean, SEM) of German (circle) and Japanese (square) subjects before and after training for the consonant condition in /anni/ (left) and the vowel condition in /kiyo/ (right). Significant increases occur after training in conditions anni8, kiyo8, kiyo6, and kiyo4. In all (but kiyo8) conditions, the MMF of Japanese subjects started at larger values than that of German subjects. (Right) Mean (+-SEM) latency differences of mismatch responses for German subjects (circle) and Japanese subjects (square) grand averaged over all subjects. Latency decreases in all conditions for German subjects, except for the kiyo4 condition. Before training, Japanese subjects have mismatch responses at earlier latencies than German subjects in all conditions (except kiyo4) and maintain this without significant changes. The N1m component (at similar100 msec) disappeared almost completely in most subjects. This disappearance resulted from a smaller recovery time for the N1m because of the very small onset-to-onset ISI (900 +- 100 msec) and the remarkable length of stimuli (400 --505 msec for /anni/ and 250 --355 msec for /kiyo/). An N1m-off at similar100 msec after the end of the stimulus also limited the refractoriness of the following N1m and, instead, a very clear P1m (at similar60 msec) and P2m (at similar160 msec) were elicited. The P1m and P2m for Japanese subjects in their responses to /anni/ were higher than the corresponding components for German subjects. In the responses to /kiyo/, the P1m and P2m also reached higher peak amplitudes than did the N1m. However, in the responses to larger differences (kiyo8 and kiyo6), these peaks merged with the MMF response. For the responses to the small-difference kiyo4, the P1m --N1m --P2m complex was again, as in the anni conditions, more pronounced. Because the duration of the deviant stimulus was longer, the latency of MMF was later (at similar250 msec) and separated from the P1m --N1m --P2m complex. Latency Differences | The same procedure for analyzing amplitudes was also used for analyzing latencies of the MMF (see Fig. ). For German subjects, within-group comparisons of the training effect (before/after) were significant for anni8 (F(1,9) = 20.85, P = 0.001) and for kiyo8 (F(1,9) = 18.18, P = 0.002). In all other conditions (except kiyo4) there were small decrements in the latency of the MMF after training. For Japanese subjects, there were small decreases in latency for anni8, kiyo8, and kiyo6. Japanese subjects had shorter latency mismatch responses in each condition (except kiyo4) as compared with German subjects. These differences were very pronounced for anni but reached significance only in condition anni8 (F(1,18) = 4.82, P = 0.04). DISCUSSION : In several studies , enhanced neurophysiological responses were found native compared to non-native speakers. These studies indicated that at a neurophysiological level, the mismatch negativity is differentially altered by lifelong linguistic experience with a particular language. On this basis, it was suggested that a language-specific phonemic code has a separate neural representation in sensory memory. In this study we showed the same result for mora contrasts. Japanese people perceive [l] and [r] as allophones of the same sound, whereas Western people hardly place them in the same phonemic category . In contrast, Japanese people have to differentiate, in their native language, durations between similar words and to extract where this difference actually occurs. For instance, they have to categorize instantaneously whether a perceived word consists of one mora or two (e.g., /o/ or /o:/) because the number of moras determines the word's meaning. Long-lasting experience with phonemes of slightly different durations is thought to facilitate perception of mora-timing. Therefore, we suggest, that the neuronal networks responsible for this ability are functionally well organized into representational maps, even possibly in phonemotopic maps. The first question of this experiment was whether intensive daily training of non-native speech features, which leads to a remarkable shift of the phonetic boundary toward smaller differences, has the power to trigger reorganization within these cortical maps. The second question was whether native speakers show a different behavioral and neuromagnetic response to these stimuli than do non-native speakers. Behavioral Results | The behavioral results showed clear learning effects across all stimuli. In the first test, the mean RT was at a similar level for both Japanese and German subjects. German subjects showed a markedly greater decrease in RT from the first to the second training than Japanese subjects. Both groups showed a smaller decrement of RTs for deviants than for standards. It is generally assumed that decreases in RT reflect the facilitation and automation of task performance, implying that additional processing is required for detecting a deviant. The discrimination performance, as reflected by the hit rates obtained in tests without feedback, indicated that after a relatively short training time, subjects learn to discriminate phonemes differing only in small durational differences. The German subjects reached an improvement in performance up to 52% for the smallest duration difference after the last training as compared with their performance before the first training. The performance of Japanese subjects started in the same range as that of German subjects, but showed smaller improvements and a higher rate of errors than that of German subjects in the course of two training sessions. Apparently, the meaning of the stimulus interfered with the phonetic processing of the stimuli in Japanese subjects. Japanese subjects had to categorize the stimuli based on their phonemic and semantic content; whereas German subjects categorized phonetically different, but meaningless non-words. Their decision was based on a phonetic category boundary effect. The categorization test showed that German subjects formed long/short categories out of the 8 variants of each stimulus. The training sessions alleviated this process and resulted in a more automated and accelerated processing that shifted the phonetic boundary toward the smaller variants . Japanese subjects made their decisions based on phonetic and further (semantic) processing. This requires time-consuming and extensive reprocessing . Neurophysiological Changes | With regard to neurophysiological changes, training had an enhancing effect on the MMF, which reflects the brain's sensitivity to small differences in auditory input. Even small differences are perceived easier and faster and result in a higher activation of the underlying neuronal networks. MMF responses for Japanese subjects were more "narrow" than for German subjects, which may be an indicator that fewer subcomponents were required for the mismatch response. Moreover, most of the Japanese subjects had MMF responses of slightly higher amplitudes and earlier latencies than did German subjects. In most subjects, the amplitude of the N1m component was diminished. However, the very clear appearance of the surrounding P1m and P2m components raised the question, whether the neuronal networks underlying these responses were also subject to plastic changes caused by the training. For /anni/, the P1m and P2m of Japanese subjects were higher than German subjects. Whereas there is no remarkable change between the two measurements in Japanese subjects, German subjects developed an explicit N1m component throughout training. This N1m component merged in the waveform of the global field power with the P1m and P2m components. In conclusion, speech-related plasticity of cortical structures seems to occur as a consequence of discrimination learning. This implies that the cortical representations of phoneme features can be remodeled dynamically through intensive training, even in adulthood. In this study, the observed effects of amplitude increment and latency decrement are similar to the results of frequency-discrimination learning reported previously . Intensive training improved the discrimination performance of the subjects and changed the neural representation of speech sounds as reflected by the automatic change-detection response (i.e., MMF). Assuming that a German listener can learn to perceive the duration quantity of phonemes as a matter of different phonemic categories, the shift of the category boundary to smaller differences ---achieved in the training ---is a clear behavioral indicator of the changes found in the neuromagnetic measures. The main components (P1m, N1m, P2m) of the standard stimuli showed locations in secondary areas of the auditory cortex (parts of Heschl's gyrus and planum temporale), which are consistent with previous reports (cf. ). The main sources of the MMF were found in the supratemporal plane (see Table ). This result is in accordance with previous assumptions that the cerebral generators of the MMN/MMF have one source in the auditory cortex and one in the frontal cortex. The auditory component of the MMN/MMF is assumed to get its main contribution from the auditory memory trace. It is also assumed that this auditory component reflects the preperceptual deviance from the auditory memory trace. The other component located in the frontal lobe might be related to involuntary switching of attention to a stimulus change . Results showed small, insignificant differences between conditions. However, the major generator in the temporal lobe could represent a more distributed network of neurons. Owing to the restricted area of our measuring unit, additional sources (e.g., in the frontal lobe) cannot be excluded. Furthermore, contributions of subcortical sources from the hippocampus and thalamus have to be reconsidered in further studies. Based on the results of this study, we propose a more continuous view between perception and memory that allows for interactions between memory and sensory systems in the processing of small perceptual differences. If this processing does not differ between native and non-native speech perception, we can exclude a contribution of experience or memory to perceptual organization. However, there were significant differences between these two groups and we should, therefore, conclude that perceptual organization is affected by experience or memory. Because perceptual skills appear to be learned gradually and specifically, human subjects cannot generalize a perceptual discrimination skill to solve similar problems with different attributes. Typically, subjects start off with poor discrimination and improve substantially with training. But when tested with another stimulus with a new attribute, their performance starts again from the baseline. In other words, learning does not transfer from the first to the second attribute. Further studies have to clarify whether a well-tuned neuronal network in a native speaker is able to reach better performance than an untrained network. The total number of cells responding to this specific feature could decrease as cells specialize during the training more and more for the trained feature. In the present study, semantic information was absent in the non-native group and present in the native Japanese group. The native group showed high automaticity of lexical processing, pointing out that fewer brain resources were required. The non-native group uses an explicit, more laborious processing system, whereas the native group processes the mora differences in a "flatter," more implicit procedural fashion. Deficits in learning foreign languages as well as temporal processing deficits can be assessed with methods similar to the procedure used in these two experiments. A method is needed wherein subjects' discrimination abilities are tested with small difference stages without feedback and trained with a staircase method of diminishing differences. Temporal-perception impairments are often associated with an inability to discriminate small temporal lags in speech. The training procedure presented here could help to improve their temporal resolution. MATERIALS AND METHODS : Subjects | Ten German subjects (aged between 23 and 33 yr, mean age: 25.6 yr) and 10 Japanese subjects (aged between 21 and 40 yr, mean age: 26.5 yr) participated in this study. Nine of the subjects in each group were right-handed, and one in each group was left-handed. Each of the right-handed subjects showed a laterality score higher than 85% on the Edinburgh Handedness Questionnaire . All German subjects and most Japanese subjects were students at the University of Munster with no history of neurological, otological, or speech disorders. None of the German subjects had any knowledge of a mora-timed or duration-sensitive second language. Japanese subjects grew up in Japan and had Japanese as their native language. In agreement with the Declaration of Helsinki, informed consent (approved by the Ethics committee of the University of Munster) was obtained from each subject after the nature of the experimental procedure had been fully explained to them. Stimuli | The Japanese word stimuli were based on naturally spoken Japanese words and were generated with 8 degrees of duration for each critical vowel or consonant within one mora (cf. Fig. ). These stimuli were produced at the Department of Speech and Cognitive Sciences at Tokyo University using the STRAIGHT algorithm Figure 8 | (Left) The duration of one mora in /anni/ and /itte/ is varied in 8 steps of 15 msec. (Left) The duration of one mora in /anni/ and /itte/ is varied in 8 steps of 15 msec. (Right) The duration of one mora in /kiyo/ and /kiyou/ is varied in 8 steps of 15 msec. In MEG measurements, the differences 8, 6, and 4 of /anni/ and /kiyo/ were tested. During the behavioral discrimination training, all 8 differences were trained and tested. For the consonants (/anni/ --/ani/), the continuum was resolved into 8 different durations. The stimulus duration decreased from 505 msec to 400 msec, with the consonant /n/ decreasing in 15-msec steps. In the /itte/ --/ite/ continuum, the silent period between the onset and offset of /t/ also decreased in 15-msec steps, resulting in 8 duration variants of /itte/ from 505 msec to 400 msec. Additionally, the duration of the vowel /i/ was manipulated in 15-msec steps. The continuum for /kiyo/ --/kyo/ ranged from 355 msec to 200 msec and for /kiyou/ --/kyou/ from 530 msec to 425 msec. All stimuli were low-pass-filtered at 4.5 kHz to eliminate high-frequency noise portions. The first and last 20 msec of the stimulus were smoothed with a cosine-shaped envelope to avoid sharp beginnings and endings. The Japanese stimuli were chosen such that the first mora was intonated high (H) and the following moras low (HLL -> HL). There was one mora difference between the long and the short variant of the stimulus. In geminates (/an-ni/, /it-te/, and /kiyo-o/), the doubled phonemes were counted as two moras . Because Japanese, in contrast to Western languages, has intonation shifting between a high and low level within the word and no stress on single syllables, the fundamental frequency for the high-intonated beginning vowel portion of all stimuli was chosen at similar300 Hz, decreasing to the low-intonated ending vowel at similar150 Hz. The starting (high) fundamental frequency, the formants, and the amplitude envelope were kept constant for all 8 stimuli in each condition. The stimuli were sampled at 20 kHz with 16-bit resolution and were delivered during training at 70 dBnHL through AKG (K240-DF) studio headphones with circumaural shells. For Japanese listeners the first (long) and the last (short) variant of each word formed separate perceptual categories, which corresponded to two meaningful words. Obviously, this was not the case for German listeners, who could not associate any meaning to the long or short variant. Table gives an overview of the stimuli, their intonation, and their meaning. Table 2 | Phonetic Structure and the Most Salient Meaning of the Stimuli All difference variants were used in the behavioral training and test sessions. In the MEG measurements, three difference variants of the word pairs /anni/ --/ani/ and /kiyo/ --/kyo/ were selected, where the duration of either a consonant or a vowel was varied. The long version of the stimuli served as the standard stimulus and differed from the large deviant (anni8/kiyo8) by 105 msec, from the middle deviant (anni6/kiyo6) by 75 msec, and from the small deviant (anni4/kiyo4) by 45 msec. The duration of /nn/ (long) in /anni/ ranged from 150 msec to 310 msec and of /n/ (short) from 150 msec to 205 msec. The duration of /i/ in kiyo ranged from 75 msec to 180 msec (long) and from 70 msec to 75 msec (short). Depending on these time relations, the MMF was expected to occur at similar100 --150 msec after the "point-of-uniqueness" of the deviant stimulus (i.e., the point in the stimulus from where the deviant unequivocally differs from the standard). Discrimination Training | Ten German subjects were trained on 10 consecutive workdays for similar1.5 h each day. A self-adjusting staircase method with two-alternative forced-choice judgments and visual feedback was used for training. Stimulus pairs were distributed in a pseudorandom order with an interstimulus interval (ISI) of 1 sec and an interblock interval (IBI) of 2.5 sec. A stimulus pair consisted of either two standard stimuli or of a standard and a deviant stimulus. For training purposes, the probability of occurrence for a deviant was set at P = 0.28. The feedback interval was 0.2 sec (cf. Fig. ). When the subject pressed the right mouse button, correctly classifying a pair of standard and deviant stimuli as "different," a green square was generated on the computer monitor, signaling a "hit" or "true positive." When the subject pressed the left mouse button falsely, classifying a standard and deviant pair as the "same," a red square was generated on the screen, signaling a "miss" or "false negative." Similarly, when a pair of two standard stimuli was correctly classified as the "same," a "correct rejection" or "true negative" was signaled by a green square. In the case that it was incorrectly classified as "different," a red square indicating a "false alarm" or "false positive" was shown as feedback . Figure 9 | Stimulation diagram. Stimulation diagram. Pairs of stimuli with an interstimulus interval (ISI) of 1 sec and an interblock interval (IBI) between pairs of 2.5 sec were distributed in a pseudorandom order with a probability of occurrence of a deviant, P = 0.28. The feedback interval (FBI = 0.2 sec) is followed by a short prestimulus interval (PSI = 0.3 sec) before the next stimulus pair. Subjects press a button if the pair of stimuli is perceived as "same" or a different button if the pair of stimuli is perceived as "different." Hits and correct rejections are fed back as a green color flash, whereas misses and false alarms are signaled as a red color flash. Standard stimuli were the long versions of the stimuli /anni/, /itte/, /kiyo/, and /kiyou/, whereas the deviant stimuli were one of the seven shorter versions . The training started at the highest durational difference (105 msec) and continued with progressively smaller differences, corresponding to discrimination performance, until the smallest durational difference relative to the standard stimulus (15 msec) was reached. After a correct response to a deviant, the magnitude of the deviance was reduced in the next pair of stimuli by one step. After a false response to a deviant, the deviance magnitude was increased by one step. Each training performance was cross-validated by a discrimination test without feedback. This test consisted of 50% pairs of standard --standard stimuli of /anni/, /itte/, /kiyo/, or /kiyou/ and 50% pairs of shorter, standard --deviant stimuli with an equal amount of the seven difference variants. Thus, within a total of 210 pairs of stimuli, there were 105 pairs of standard stimuli and 105 pairs of deviant stimuli, with 15 stimuli from each difference variant. The test stimuli were delivered in a pseudorandomized order, and no visual feedback was provided. In these tests, subjects had to recognize the difference strictly by comparison with the standard. The timing of the stimulus presentation was the same as in the training session. Each test session began with a standard stimulus for orientation. After each training session, subjects were informed about their best performance in the session. MEG Recordings | Magnetoencephalographic recordings were performed before and after training sessions with a 37-channel neuromagnetometer (Magnes, Biomagnetic Technologies) in an acoustically and magnetically shielded room. In each measurement session, three difference variants of each stimulus pair (/anni/ --/ani/, /kiyo/ --/kyo/), differing in length by 105 msec (anni8, kiyo8), 75 msec (anni6, kiyo6), and 45 msec (anni4, kiyo4) from standard, were presented. Each variant was presented three times in blocks of 300 sec to ensure test --retest reliability. The three identical blocks of one variant were then averaged. The resulting number of stimulus-related epochs was similar1000 for each condition. After exclusion of artifacts, one measurement session comprised, on average, 146 stimulus-related epochs to deviants and 698 epochs to standards. The MEG session for /anni/ lasted, with pauses, similar1 h and was followed by a short break, where the subjects walked around and relaxed. After this session, the MEG session for /kiyo/ was performed with the same timing. The probability of occurrence for deviants was P = 0.15%, with a randomized stimulus-onset-asynchrony (SOA) of 900 +- 100 msec, and a minimum of 3 and a maximum of 9 standards before a deviant. The intensity of each stimulus was 70 dBSL above the sensation threshold (SL), individually measured for each subject. Continuous MEG recordings were low-pass filtered at 0.1 --100 Hz and sampled at a rate of 297.6 Hz. Stimulation was delivered monaurally to the contralateral side of the recorded hemisphere. The sensor array, with gradiometer-type coils (circular concave array with diameter of 144 mm and spherical radius of 122 mm), was centered above the T3 position of the international 10 --20 system for EEG electrode placement for the left hemisphere and centered above T4 for the right hemisphere. A sensor-positioning system was used to determine the sensor location relative to the head and to indicate whether head movements occurred during the measurement. The subjects rested in a supine position supported by a vacuum cast to ensure a stable body and head position. They were instructed to stay relaxed and awake during MEG measurements by watching an animated video of their choice. This was expected to keep their state of arousal constant and attention continuously oriented toward the visual distraction. Compliance was verified by video monitoring. Any stimulus-related epochs contaminated by eye blink or head movement artifacts (usually no more than 20% of the data) were rejected. Stimulus-related epochs of 700 msec with prestimulus intervals of 100 msec were selectively averaged, baseline-corrected, and low-pass filtered from 0 to 20 Hz with a second-order Bessel filter (no phase shift). Backmatter: PMID- 12359837 TI - Patterns of Interference in Sequence Learning and Prism Adaptation Inconsistent With the Consolidation Hypothesis AB - The studies reported here used an interference paradigm to determine whether a long-term consolidation process (i.e., one lasting from several hours to days) occurs in the learning of two implicit motor skills, learning of a movement sequence and learning of a visuo-motor mapping. Subjects learned one skill and were tested on that skill 48 h later. Between the learning session and test session, some subjects trained on a second skill. The amount of time between the learning of the two skills varied for different subjects. In both the learning of a movement sequence and the learning of a visuo-motor mapping, we found that remote memories were susceptible to interference, but the passage of time did not afford protection from interference. These results are inconsistent with the long-term consolidation of these motor skills. A possible difference between these tasks and those that do show long-term consolidation is that the present tasks are not dynamic motor skills. Keywords: Introduction : The notion that memories initially exist in a labile state and with time become stable was first proposed in 1900 by Muller and Pilzecker (cited in ). These researchers found that memory for a list of words was disrupted if new information was learned less than 1 min later, but not if learned 10 min later. They concluded that the memories had undergone a stabilization process, which they termed consolidation. In the current study, we will use the term consolidation to refer to the stabilization of memory (i.e., a decrease in susceptibility to forgetting via decay or interference) in the absence of further practice of that memory. There are two putative phases of consolidation processes, short-term consolidation processes, which operate over a period of seconds to hours post-training and are mediated by local cellular mechanisms resulting in long-term potentiation (LTP), and long-term consolidation processes, which operate over a period of hours to months post-training and are mediated by neural reorganization (for review, see ; ). The focus of the current study is on long-term consolidation processes. Much of the research on long-term consolidation processes has focused on explicit memory tasks, tasks in which memory is demonstrated through conscious recollection of previously learned information . Although less consolidation research has focused on implicit memory tasks, tasks in which memory is revealed through performance and does not require conscious recollection of learned information, recent interest in the long-term consolidation of motor learning is one exception . The experiments described here were designed to assess whether a long-term consolidation process is ubiquitous for different components of motor skill learning. Although there is much evidence that a long-term process of consolidation is evident in performance on explicit memory tasks (e.g., ; ), it is not immediately apparent that a long-term consolidation process would be evident in the performance of implicit memory tasks. A critical difference between performance on explicit and implicit memory tasks is that performance on explicit tasks relies on the integrity of the medial temporal lobe (i.e., the declarative memory system), whereas performance on implicit tasks does not (; see , for an exception). Performance on implicit tasks depends on the integrity of various anatomical structures outside of the medial temporal lobe (i.e., nondeclarative memory). One result of this difference is that patients with hippocampal damage are impaired on tests of explicit memory (e.g., ), but their performance is relatively spared on tests of motor skill learning . We will focus on the memory supporting motor skill learning as a particular type of nondeclarative memory, although the tasks discussed here are not meant to be representative of all forms of nondeclarative memory. There is behavioral evidence of long-term motor skill consolidation in a force-field task requiring learning of dynamic transformations, that is, the learning of new muscle forces . In this task, subjects must compensate for a perturbing force while moving a cursor to a target location on a computer screen. In one demonstration of temporally graded retroactive interference (RI) in the learning of the force-field task, all subjects practiced on one pattern of forces (force A). Some subjects practiced on a second, opposing pattern of forces (force B) at varying amounts of time after initial learning of the first force. All subjects were then tested on force A 24 h after the initial learning of A. Subjects who trained on force B either 5 min or 1 h after training on force A had impaired retention of A (i.e., greater RI) compared with groups that never trained on force B or that trained on force B 4 h after force A training. This pattern of temporally graded RI, in which only those subjects learning force B soon after having learned force A show an impairment in their memory for A, was interpreted as evidence that a long-term process of consolidation occurs for motor skill. Furthermore, it has been suggested that a reorganization of neural connectivity may be responsible for the consolidation seen in the force-field task . Recent work does indicate that the primary motor cortex (M1) may be responsible for the consolidation of dynamic information. found that administering repetitive transcranial magnetic stimulation (rTMS), which interferes with synaptic activity, to M1 immediately following the learning of a finger movement task-disrupted retention of performance. Administering the rTMS to M1 6 h after completing practice on the task did not, however, disrupt retention of performance. These results provide a possible neuropyschological mechanism for the changes in RI witnessed in the force-field learning task . One problem with attempting to generalize the results that a process of consolidation exists for implicit motor skill learning beyond the force-field learning task is that different brain structures are responsible for the learning and execution of different components of motor skill (for review, see ; ). In particular, learning new patterns of muscles forces relies on M1 , and their representation also relies on pools of interneurons in the spinal cord (see for a review of recent evidence). On the other hand, learning and performing a sequence of actions relies on the basal ganglia and supplementary motor area ; whereas learning kinematic transformations involving the integration of perceptual and motor information relies on the posterior parietal cortex . Therefore, demonstrating consolidation in the learning of dynamic information, which is dependent on M1, is not necessarily informative about the existence of consolidation in other motor skills that do not rely upon the same brain structure. The set of experiments reported here explored the existence of a long-term consolidation process in two different nondynamic components of motor skill whose anatomical basis is well understood, the learning of a movement sequence and the learning of a new visuo-motor mapping. We used an interference design to assess consolidation. Subjects learned a skill at one point in time (skill A) and their retention of that skill was tested 48 h later. Between the learning session and the test session, some subjects learned a second skill that was incompatible with the first (skill B). The appropriate response for skill B was as different as possible from that for skill A, given very similar retrieval cues. The amount of time between the learning of the first and second skills varied for different subjects. EXPERIMENT 1: RESULTS AND DISCUSSION : To assess whether consolidation occurs in the learning of a movement sequence, we chose the serial response time (SRT) task. In this task, subjects see circles appear one at a time in one of four boxes arranged horizontally on the computer screen. The task of the subjects is to press a key on the computer keyboard that directly corresponds to the spatial location of the circle as quickly as possible. Unbeknownst to the subject, the circles appear in a repeating sequence of spatial locations. Subjects become faster at responding to the repeating sequence of circles. Additionally, if the experimenter later introduces randomly determined circle locations, the subjects' reaction time (RT) increases. These effects occur even if the subjects do not become aware that the circles are repeating in a sequence . As such, learning in this task can be implicit. SRT | As learning in the SRT task is heavily dependent on making the correct series of spatial responses , six subjects who failed to get at least 90% correct on the SRT task during training were excluded from the analysis, leaving 24 subjects in the control condition, 24 in the 5-min, 24 in the 1-h, 21 in the 5-h, and 23 in the 24-h. For the remaining subjects, accuracy in the SRT task was uniformly high (95% --98% correct for each block in each condition). There were no reliable effects in analyses of the accuracy scores (all Fs <1.5, Ps >0.20). This result is unsurprising given that the subjects were instructed to respond as quickly and as accurately as possible. The remaining SRT analyses reported here are based on RT as recorded in msec. RTs were summarized by taking the median of each group of 12 trials and then finding the mean of the 8 medians for each block, yielding a single summary RT per trial block for each subject. RT across blocks for each of the conditions is depicted in Figure . Figure 1 | Experiment 1 RT during training and testing sessions for each condition. Experiment 1 RT during training and testing sessions for each condition. (A) Control; (B) 5-min; (C) 1-h; (D) 5-h; (E) 24-h. Training on Sequence A; training on Sequence B; testing on Sequence A. (Rs) Random blocks; (Ss) sequenced blocks. Error bars, +/- 1 SE. The measures of learning were taken when subjects performed the SRT task alone by subtracting performance in the S block from the average of that of the two R blocks. Learning of A | Although it is possible to assess learning during training on sequence A (with the concurrent memory task), any improvements during the sequenced training blocks may be attributable to one of three sources, learning of the movement sequence, general improvements on the button-pushing task, or learning to better coordinate simultaneous performance of both tasks. Additionally, the concurrent memory task may have influenced the expression of learning. Therefore, learning was assessed at transfer (without the concurrent memory task) only. A learning score for sequence A was created by subtracting the response time in the final sequenced block from the average of that in the final two random blocks. (Similar measures were created for the learning of B and testing of A.) All groups showed equivalent learning of sequence A, regardless of when they learned sequence B (see Table ). An ANOVA with time-learned sequence B (control, 5 min, 1 h, 5 h, or 24 h) as a variable failed to reach significance, F<1. The learning scores for sequence A differed reliably from zero, t(115)=8.66, P<0.0001. Subjects responded faster to the sequenced than to the random blocks. Table 1 | Means and SE of the Learning Scores for Skill A Even though there was not a between-group difference in sequence A learning, it is possible that subjects' learning of A was related to their ability to learn a second sequence, as well as to their ability to retain the initial learning. As proactive interference (PI) and RI depend on prior learning, those who learn more have a greater opportunity for RI and PI ---at the extreme, no learning at all should leave minimal opportunity for RI or PI. To assess this possibility, we used subjects' sequence A learning scores to predict the interference measures (described below) in separate simple linear regressions. The extent of sequence A learning was a reliable predictor of both PI and RI (see Table for regression statistics). Subjects who showed greater learning of sequence A also experienced greater PI and greater RI. As learning of sequence A was related to subsequent interference measures, all reported ANOVAs for the RT results are analyses of covariance (ANCOVA), with the learning score for sequence A as the covariate. These ANCOVA results for the PI and RI analyses did not differ qualitatively from those results obtained when an ANOVA was used. Proactive Interference | The nature of PI bears on the interpretation of the RI results. Namely, if a particular group shows no RI, but extreme PI, it may be that this group never learned the second skill well enough to interfere with their memory for the first skill. For example, if RI was temporally graded in the direction predicted by a consolidation account (i.e., the greatest amount of RI occurred when the two skills were learned close together in time), but PI was temporally graded in the opposite direction (i.e., the least amount of PI occurred when two skills were learned close together in time), then the temporal gradient of the RI may only indicate that some subjects never learned the second skill. The PI measure was the difference between each subject's score when learning skills A and B. Positive scores on this measure indicate positive transfer; negative scores indicate PI. As would be expected from a visual inspection of the covariate-adjusted means of the PI scores (see Fig. A), there was PI in the learning of sequence B; however, there was no indication of a temporal gradient to the PI. Overall, the measure of PI was reliably less than zero (M = -8.37, SE=3.8), indicative of interference, t(91) = -2.282, P=0.0249 An ANCOVA on the PI scores with time learned sequence B as a variable yielded no significant effects, F<1. Two planned contrasts were performed to assess the temporal gradient of the PI scores, a planned linear contrast, and a planned comparison contrasting the 5 and 24 hr groups with the other groups learning sequence B. Both failed to reach significance, Fs<1. Even when the second sequence was learned 24 h after having learned the first, there was PI from having learned a different sequence previously. This pattern of proactive interference does not pose a problem for the interpretation of the RI results. Figure 2 | Covariate-adjusted means of the (A) PI scores and (B) the RI scores for the RT measure in Experiment 1. Covariate-adjusted means of the (A) PI scores and (B) the RI scores for the RT measure in Experiment 1. PI scores were obtained by subtracting the Learning of A score from that for the Learning of B. RI scores were obtained by subtracting the Learning of A score from that for the Test of A. Negative values on these measures indicate interference. Positive values indicate positive transfer or retention. Assessing Consolidation: Retroactive Interference Measure | The RI results are of critical importance in addressing the question of whether there is consolidation in sequence learning. If the memory for a movement sequence existed in a labile state for some period of time and slowly became stabilized (i.e., it was consolidated), then a person learning sequence B soon after having learned A should have an impaired memory for A. In particular, one would expect a temporally graded RI. That is, the memory impairment for A decreases as the temporal interval between the learning of the two skills increases. The RI measure is the difference between performance when subjects train on sequence A and when they test on A. Positive scores on this measure indicate an improvement in the memory for skill A. Negative scores indicate RI. Statistical analyses confirmed the visual impression created by the covariate-adjusted means of the RI scores (see Fig. B); there was RI in the memory for sequence A. An ANCOVA on the RI scores with time learned sequence B (control, 5 min, 1 h, 5 h, or 24 h) as a variable failed to yield an overall effect, F<1.5, P<0.25. However, a planned comparison revealed that the groups that learned sequence B had greater RI than did the control group, F(1, 106) = 7.995, P=0.0056, MSE=335.662. This result indicates that all groups that learned a second sequence exhibited RI in their memory for the first sequence. However, the extent of RI did not depend on the amount of time that passed between the learning of the two sequences. To test for a temporal gradient in the RI, we performed two planned comparisons, a linear contrast among those groups that learned a second skill and a planned comparison contrasting the groups that learned skill B within 5 h of having learned skill A, with those groups that learned skill B 5 or more hours after having learned skill A. Neither contrast was reliable, Fs<1. Given the lack of temporal gradient in the RI scores, there is no evidence of consolidation in the learning of a movement sequence. Explicit Memory Measures | Confidence Ratings | A repeated measures ANOVA with time-learned sequence B (control, 5 min, 1 h, 5 h, or 24 h) as a between-subjects variable and session (learn A, learn B, or test A) as a within-subjects variable yielded a reliable effect of session, F(2, 174) = 4.09, P=0.0184, MSE=1.236. Helmhert comparisons revealed that confidence ratings were greater for the second and third training sessions (M=4.22, SE=0.16 and M=4.20, SE=0.15, respectively) than for the first training session (M=3.92, SE=0.14), F(1, 174) = 6.909, P=0.0093. Subjects may have become more suspicious about the nature of the task after they had participated in at least one session. The effect of time-learned sequence B failed to reach significance, as did the interaction between time-learned sequence B and session, Fs<1.5, Ps>0.25. These results indicate that there were no differences between the groups in their confidence in having seen a sequence. Free Recall | Each subject's free recall score reflected the number of positions in the sequence correctly recalled. A position was considered correct when it was included within a correctly recalled segment consisting of a minimum of three consecutive positions, but these recalled segments themselves need not be consecutive. For example, if a subject saw 314324123142 and recalled 123143, the score would be 6, because both 123 and 143 occurred in the sequence. The time at which sequence B was learned had no influence on subjects' recall of sequences A and B. Overall, however, subjects had greater recall of sequence A (M=5.45, SE=0.19) than of sequence B (M=4.10, SE=0.22), indicating PI in the explicit memory for sequence B. A repeated measures ANOVA with time learned sequence B (control, 5 min, 1 h, 5 h, or 24 h) as a between-subjects variable and learning session (learn A, learn B, or test A) as a within-subjects variable yielded a significant effect of learning session, F(2, 174) = 11.48, P=0.0001, MSE=4.344. Helmhert comparisons revealed that recall of sequence A was greater than that of sequence B, F(1, 174) = 22.701, P=0.0001. Recall of A (learning or test) did not vary with the learning session, F<1. The effect of time-learned sequence B was not reliable, nor was the interaction, Fs<1.3, Ps>0.28. Previous reports of explicit knowledge in SRT with similar amounts of sequence training have found that mean guessing performance is 4.6 . The interference analyses on RT were rerun separately for those subjects who recalled five or more positions of sequence A, that is, more positions than would be expected by chance (n=79) and those subjects who recalled four or less positions (n=36). Splitting the analyses on the basis of recall did not change the qualitative nature of the results. There was ungraded PI and RI in both high and low recall groups. Summary | In Experiment 1, interpolated sequence learning hurt the memory for a previously learned sequence, regardless of the amount of time that had passed between the original and interpolated learning. This pattern of RI is inconsistent with the notion that implicit sequence knowledge consolidates. Why might there be consolidation in the learning of new muscle dynamics, but not in the learning of a new sequence of movements? One possibility is that the learning of a new sequence of movements and the learning of new muscles dynamics have different anatomic bases and, therefore, different mechanisms of long-term storage. A second possibility is that the sequence-learning task used in Experiment 1 is behaviorally different from the force-field learning task used to assess the learning of dynamic information. In the force field learning task, prior to executing any movements, all of the perceptual cues available to the subject are the same when learning force A and force B. This is not the case in the SRT task. The two sequences in Experiment 1 were chosen to be as different as possible from one another. Therefore, the perceptual cues for learning sequences A and B were not identical. Additionally, in the SRT task, a person does not have to interact with the environment before knowing the correct movement to make. Whereas in the force-field learning task, there is nothing in the environment that dictates the correct movement until the subject begins to interact with the robot arm. In the force-field learning task, the subject must gradually learn the correct movements on the basis of error feedback. In this way, the force-field learning task is an adaptation paradigm. Lastly, learning one sequence of movements is not necessarily incompatible with learning a second sequence of movements. In the force-field learning task, however, learning one force for interacting with the robot arm is incompatible with learning a different force for interacting with that robot arm. In Experiment 2, we investigated this second possibility, that the failure to find consolidation in the learning of a movement sequence was due to the different behavioral requirements of that task, by testing whether there was temporally graded RI in a task that was behaviorally similar to the force-field learning task, yet which has a unique anatomic basis ---prism adaptation. The prism adaptation paradigm involves a kinematic transformation in which subjects must learn a new mapping between their vision and proprioception. It requires the subject to interact with the environment before knowing the correct movement to make. Likewise, the cues for the appropriate response are identical when learning two incompatible prism displacements. EXPERIMENT 2: RESULTS AND DISCUSSION : Experiment 2 investigated consolidation in the learning of a new mapping between the perception and action systems (i.e., a kinematic transformation). We chose the prism adaptation paradigm to investigate the learning of a new visuo-motor mapping. In studies of prism adaptation, subjects don prism goggles that displace their vision laterally. During prism exposure, subjects are trained to point at targets with visual feedback. Before and after prism exposure, subjects' target-pointing accuracy is measured with normal vision. During the prism exposure period, the target-pointing performance of subjects is initially inaccurate, but with training, they learn to point accurately. A comparison of subjects' target-pointing performance with normal vision before and after the prism exposure period provides a measure of aftereffects. After prism exposure, subjects usually point inaccurately in the direction opposite of the prism displacement, that is, a negative aftereffect. Performance was measured as the horizontal displacement from the target in inches (rounded to the nearest quarter inch). For throws made while wearing, or after wearing the leftward displacing goggles, the sign on all error was reversed so that the data could be averaged with that for the rightward displacing goggles. Positive values on this measure indicate error in the direction of the displacement; negative values indicate error in the direction opposite the displacement. Exposure Performance | Exposure performance was assessed on trials in which the subjects wore the prism goggles and was summarized by averaging the error across the trials within a block separately for each of the target locations. This was done for each of the three sessions (learning of A, learning of B, and testing of A). Exposure error across sessions appears in Figure . Figure 3 | Experiment 2 exposure performance across blocks during the training and testing sessions for each condition. Experiment 2 exposure performance across blocks during the training and testing sessions for each condition. (A) Control; (B) 5-min; (C) 5-h; (D) 24-h. Training on Sequence A; training on Sequence B; testing on Sequence A. Error bars, +/- 1 SE. Learning of A | All groups of subjects learned the initial prism displacement equally well. Additionally, subjects were equally good at learning to throw with both right and left displacing prisms. In a repeated measures, ANOVA with trial block (one, two, or three) and target location (left, center, or right) as within-subjects variables and time learned displacement B (control, 5 min, 5 h, or 24 h) and initial prism displacement (leftward or rightward) as between subjects variables, there were no reliable effects of time-learned displacement B or of initial prism displacement, Fs<1. None of the effects involving target location were significant, Fs<1.5, P>0.25, which indicates that no target location was more difficult to learn than another. The only within-subjects variable to reach significance was that of trial block, F(2, 192) = 656.072, P=0.0001, MSE=3.062. As would be expected, subjects' error decreased across blocks of trials during their initial training on displacement A. Because error performance did not reliably vary with target location or initial prism displacement, these variables were excluded from further analyses. Accuracy in the last block of trials while learning displacement A (see Table ) reliably predicted PI scores (see Table for regression statistics). Subjects with greater learning of displacement A had greater PI. Accuracy in the learning of A did not predict RI scores (see Table ). Therefore, accuracy in the last block of trials of learning displacement A was used as a covariate in the PI analyses only. As in Experiment 1, ANCOVA results for the PI analyses did not qualitatively differ from those results obtained when an ANOVA was used on the PI scores. Table 2 | Regression Results for Predicting Proactive and Retroactive Interference Scores from Initial Learning of Skill A Proactive Interference | PI scores were created in the same manner as in Experiment 1. Covariate-adjusted means for the PI scores appear in Figure A. As is apparent in the figure, overall, there was PI in the learning of displacement B. Overall, the PI scores were reliably less than zero (M = -1.2, SE = 0.26), t(71) = -4.644, P<0.0001. However, the extent of PI did not depend on when subjects learned displacement B. In a repeated-measures ANCOVA with trial block (one, two, or three) as a within-subjects variable and time-learned displacement B (5 min, 5 h, or 24 h) as a between-subjects variable, there were no reliable effects, Fs<1.5, Ps>0.25. The planned contrast comparing the 5-min group with the 5- and 24-h group failed to reach significance, as did the linear contrast, Fs<1. All groups showed a similar impairment in their learning of displacement B regardless of the amount of time that had passed between the learning of the two displacements. Figure 4 | (A) Covariate-adjusted means for the PI scores on the measure of exposure performance. (A) Covariate-adjusted means for the PI scores on the measure of exposure performance. PI scores were obtained by subtracting subjects' Learning of B score from their Learning of A score. (B) Raw means for the RI scores on the measure of exposure performance. Error bars, +/- 1 SE. RI scores were obtained by subtracting subjects' Test of A score from their Learning of A score. Negative values on both the PI and RI measures indicate interference; positive values indicate either positive transfer or retention. Assessing Consolidation: Retroactive Interference Measure | Once again, the RI measure is of critical importance to the question of whether there was consolidation in the exposure performance of prism adaptation. RI scores were created in the same manner as in Experiment 1. Means for the RI scores appear Figure B. As is apparent in the figure, overall, there was RI; the groups that learned displacement B were impaired in their memory for displacement A relative to the control group. An ANOVA on the RI scores with trial block (one, two, or three) as a within-subjects variable and time-learned displacement B (control, 5 min, 5 h, or 24 h) as a between-subjects variable yielded a reliable effect of time learned displacement B, F(3, 100) = 6.799, P=0.0003, MSE=4.725.The planned comparison between the control group and the groups that learned a second displacement revealed that the control group had greater retention of displacement A, F(1,100) = 9.533, P=0.0026. In the above ANOVA, there was a reliable effect of trial block, F(2, 200) = 7.721, P=0.0006, MSE=1.719, as well as a reliable interaction between time-learned displacement B and trial block, F(6, 200) = 7.548, P=0.0001. Helmhert comparisons revealed that the interaction was due to greater facilitation in the memory of displacement A for the control group in trial block one (M=2.159, SE=0.296 for block one, and M = -0.104, SE=0.196 for blocks two and three), F(2, 100) = 15.51, P=0.0001, whereas, for the remaining groups, there was no difference between trial block 1 and trial blocks 2 and 3, Fs<1. As is apparent in Figure B, there was a temporal gradient to the RI scores. This gradient, however, was in the direction opposite from that which would be expected due to consolidation. The planned linear contrast revealed that impairment in the memory for displacement A increased as the amount of time between the learning of displacements A and B increased, F(1, 69) = 9.782, P=0.0026, MSE=5.175. There is, therefore, no evidence of consolidation in subjects' exposure performance during prism adaptation. Aftereffects Performance | Exposure performance is thought to reflect the combination of an implicit adaptation between the perceptual and motor systems and the use of explicit strategies that allow for accurate pointing performance. For example, subjects may learn that when wearing the prism goggles, which displace their vision 12 to the right, that accurate performance can be obtained by consciously selecting a spatial target for pointing that is 12 to the left of where the target appears visually. Aftereffects performance, however, is thought to reflect only the implicit adaptation between the perception and action systems. As the subjects are not wearing the prism-displacing goggles during tests of aftereffects, it would not be reasonable for them to consciously draw on a strategy to correct the effects of a displacement. As such, the results central to the question of whether there is consolidation in the implicit learning of a new visuo-motor mapping are those for the aftereffects performance. Aftereffects were assessed in trials during which subjects wore nondisplacing goggles. Only the first aftereffects trial in each block was used, as subjects may begin applying strategies to combat the aftereffects on subsequent throws within that block. To rid the aftereffects scores of any directional bias, albeit small (overall M=0.14, SE=0.14), that the subjects may have started out with prior to prism training, the average of each subject's baseline performance was subtracted from his or her error on each aftereffects trial. Positive aftereffects scores reflect error in the direction of the prism displacement; negative scores reflect error in the direction opposite the prism displacement. Once a subject takes the prism goggles off, error in the opposite direction of the displacement reflects the extent of the implicit adaptation between the perception and action systems. Therefore, aftereffects performance differs from exposure performance in that greater negative values are indicative of greater learning. Subjects' aftereffects performance across blocks for each of the training session is depicted in Figure . Figure 5 | Experiment 2 aftereffects performance across blocks during the training and testing sessions for each condition: (A) Control; (B) 5-min; (C) 5-hr; (D) 24-hr. Experiment 2 aftereffects performance across blocks during the training and testing sessions for each condition: (A) Control; (B) 5-min; (C) 5-hr; (D) 24-hr. Training on Sequence A; training on Sequence B; testing on Sequence A. Negative values are indicative of a greater adaptation between the perception and action systems. Error bars, +/- 1 SE. Learning of A | All groups of subjects acquired similar aftereffects during the learning of displacement A (see Table ) and these aftereffects were reliably less than zero, t(103) = -16.373, P<0.0001, indicating an adaptation between the perceptual and motor systems. In a repeated-measures ANOVA with time-learned displacement B (control, 5 min, 5 h, 24 h) as a between-subjects variable and trial block (one, two, or three) as a within-subjects variable, there were no reliable effects, Fs<1.5, Ps>1.8. The lack of an effect of trial block suggests that the aftereffects were acquired fairly quickly ---within the first block of trials. The aftereffects for the last block of training on displacement A reliably predicted both PI and RI (see Table for regression statistics). Those subjects who had greater adaptation during the learning of displacement A also experienced greater PI, as well as greater RI. Therefore, the aftereffects score on the last block during training on displacement A was used as a covariate in the subsequent analyses. These ANCOVA results for the PI and RI analyses did not differ qualitatively from those results obtained when an ANOVA was used. Proactive Interference | PI scores were created in the same manner as in Experiment 1 and appear in Figure A. As is apparent in the figure, there was PI when learning displacement B 5 min after having learned displacement A, but this PI disappeared once at least 5 h passed between the learning of the two displacements. In a repeated-measures ANCOVA with condition (5 min, 5 h, 24 h) as a between-subjects variable and trial block (one, two, three) as a within-subjects variable there was only a marginal effect of condition, F(2, 68) = 2.345, P= 0.1035, MSE=34.98. Whereas the linear contrast among those groups learning displacement B was not reliable, F(1, 68) = 2.3.85, P=0.1271, the planned comparison contrasting the 5-min group with the 5- and 24-h groups was reliable, F(1, 68) = 4.396, P=0.0397. Figure 6 | (A) Covariate-adjusted means for the PI scores on the measure of aftereffects performance. (A) Covariate-adjusted means for the PI scores on the measure of aftereffects performance. PI scores were obtained by subtracting subjects' Learning of B score from their Learning of A score. (B) Means for the RI scores of aftereffects performance. Error bars, +/- 1 SE. RI scores were obtained by subtracting subjects' Test of A score from their Learning of A score. Negative values on both the PI and RI measures indicate interference; positive values indicate either positive transfer or retention. In the above analysis, there was also a reliable effect of trial block, F(2, 136) = 8.034, P=0.0005, MSE=10.888. Post hoc analyses on the effect of trial block revealed that across all conditions, the PI scores were greater in blocks one and two (M = -0.812, SE=0.547, M = -0.888, SE=0.566, respectively) than those in block three (M = -0.479, SE=0.573). The interaction between condition and trial block did not reach significance, F<1. As one might expect, the extent of PI decreased with extensive practice on the second displacement; this decrease was present in all groups that learned displacement B. Assessing Consolidation: Retroactive Interference | RI scores were created in the same manner as in Experiment 1 and appear in Figure B. Overall, those groups learning displacement B experienced RI in their memory of the aftereffects for displacement A. In a repeated-measures ANCOVA with time-learned displacement B (control, 5 min, 5 h, 24 h) as a between-subjects variable and trial block as a within-subjects variable (one, two, three), there was a significant effect of time-learned displacement B, F(3, 99) = 3.803, P=0.0126, MSE=22.775. A planned comparison revealed that the control group retained the aftereffects for displacement A, whereas those groups that learned a second displacement experienced an impairment, F(1, 99) = 7.308, P=0.0081. Although the RI scores in Figure B appear to be temporally graded in the direction opposite that predicted by a consolidation account, this effect was only marginally significant. In Fisher's PLSD post hoc analyses, the RI scores of the 5-min and 5-h groups did not differ from each other, F<1, but the 24-h group had marginally greater RI than did the 5-min and 5-h groups, F(1, 99) = 3.737, P=0.0561. In the above ANCOVA, there was also an effect of trial block, F(2, 198)=6.714, P=0.0015, MSE=11.548. Post-hoc analyses on the effect of trial block revealed that the RI was greater in block three (M = -0.528, SE = 0.444) than in blocks one or two (M = -0.258, SE = 0.409, M = -0.272, SE = 0.404, respectively), F(1, 198) = 11.066, P = 0.001, MSE = 11.479. This result suggests that the interference produced by learning displacement B did not dissipate with training as one might expect. The interaction between trial block and time-learned displacement B did not reach significance, F <1. Summary | When subjects learned a new mapping between their vision and their proprioception (i.e., a kinematic transformation) they experienced RI in their exposure performance, but the gradient to this interference was in the direction opposite that predicted by a consolidation account. The RI for the aftereffects when learning two displacements was qualitatively similar to that for the exposure performance. Whereas all groups that learned a second displacement were impaired in their memory for the aftereffects of the first displacement, there was a trend for the RI in the aftereffects to be temporally graded in the direction opposite that predicted by a consolidation account (although this trend was not reliable). Nevertheless, the pattern of RI in the aftereffects performance is inconsistent with the notion that memory for the implicit perceptual-motor adaptation consolidates. GENERAL DISCUSSION : The current set of experiments tested the ubiquity of a long-term consolidation process for different components of motor-skill learning. A consolidation account predicts temporally graded RI in the learning of two skills, such that RI would decrease as the amount of time that passed between the learning of the two skills increased. Results of the current experiments are inconsistent with the notion that the learning of nondynamic motor information undergoes a long-term consolidation process. Experiment 1 examined the existence of a long-term consolidation process in the learning of a sequence of movements. In this experiment, the learning of one movement sequence proactively interfered with the learning of a second movement sequence, and this PI was not temporally graded. Most notably, in Experiment 1, learning a second movement sequence caused an impairment in the memory for the first movement sequence (i.e., RI). This interference, however, was not temporally graded as would be predicted by a consolidation account. Experiment 2 examined the existence of a long-term consolidation process in the learning of new mappings between the perception and action systems. In this experiment, subjects experienced PI in both their exposure performance (which is a combination of conscious strategies and of the implicit adaptation between the perceptual and motor systems) and in their aftereffects performance (which is thought to be a pure measure of the implicit adaptation between the perceptual and motor systems). The extent of this PI was temporally graded in the aftereffects performance, but not in the exposure performance. When attempting to learn a new displacement only 5 min after having trained on an opposing displacement, the adaptation between the perceptual and motor systems was impaired. However, whether 5 or 24 h passed between the attempts to learn the two displacements, the adaptation when learning the second displacement was not impaired. These results are congruent with results from an attempt to train monkeys to reach under two opposing visual displacements (via prism spectacles). When training sessions were separated by only 5 min, the monkeys were impaired at learning the second displacement. However, when the training sessions were separated by 24 h, the monkeys were not impaired at adapting to the second displacement . Consistent with other demonstrations of RI in the learning of kinematic transformations , in Experiment 2, learning of one prism displacement hurt the memory for a previously learned displacement (i.e., there was RI) in both the exposure and the aftereffects performance. In the exposure performance, this RI was temporally graded, but in the direction opposite that predicted by a consolidation account. As the amount of time between the learning of the two prism displacements increased, the amount of RI also increased. Subjects were influenced by what they had done most recently. This pattern of RI, although inconsistent with a consolidation account, is consistent with a number of accounts of RI in paired-associate (i.e., explicit) memory. In one account, the retrieval-induced forgetting account, retrieving any particular target item involves a temporary, reversible, and active suppression of competing memories . Having recently retrieved a competing response (e.g., displacement B) from memory hurts the retention of a previously learned response (e.g., displacement A) more so than if a longer period of time has passed between retrieving the competing response and the retention test for the original response . A second alternative is the unlearning account , which posits that learning a second response to the same cue (learning displacement B while wearing goggles) will weaken the association between the first cue and response (that between displacement A and the goggles). Unlearning is analogous to the extinction of conditioned responses in conditioning paradigms, in that it predicts spontaneous recovery of the original association when memory is tested at longer delays after the interpolated learning. A third alternative is that the more recently the competing memory is learned, the stronger it is and, therefore, the more likely it is to interfere with what has been learned previously. Experiment 2 was not designed to distinguish between these explanations. It is therefore possible that any one may account for our results. Note that all of these explanations rely on the notion that the same cue becomes associated with two different responses that directly compete with one another. We would only expect to see this pattern of reversed temporally graded RI in a task in which the responses do directly compete with one another. This is exactly the condition of the experimental situation in the prism adaptation task of Experiment 2, but not the SRT task of Experiment 1. We do not see this reverse pattern of temporally graded RI in Experiment 1. Furthermore, a similar reversed pattern of temporally graded RI has been observed in the force-field learning task, in which responses do directly compete with one another . In this study, recent learning of a new force interfered retroactively with the memory for a previously learned force, even though 24 h had passed since the old force was learned (i.e., it should have been consolidated). It appears that even if a memory has consolidated, recently retrieving a different response for the same cue can interfere with the retrieval of this consolidated memory. One might expect the RI scores of the exposure performance in our Experiment 2 to be temporally graded in the direction predicted by a consolidation account, because it is known that explicit knowledge consolidates. In this instance, however, any explicit knowledge subjects acquired for improving performance on one displacement could potentially be used for improving performance on the second displacement. For example, a subject could decide to throw to a visual target in the opposite direction of his or her error. This type of strategy would not generate RI. The results of the current set of experiments appear to be inconsistent with demonstrations of RI that are thought to reflect consolidation (e.g., ; ). These studies have shown that learning a second kinematic transformation immediately after having learned a first transformation interferes with the memory for that first transformation. We argue, however, that these studies are not convincing demonstrations of a long-term consolidation process, as they have not tested whether the RI dissipates as predicted by a consolidation account. We demonstrated RI in the learning of two skills in the current set of experiments while failing to find evidence for a long-term consolidation process. Why have there been sufficient demonstrations of long-term consolidation in the learning of dynamic information (e.g., ) but not in the learning of a movement sequence or of a kinematic transformation? One possible reason for the discrepancy between our results and those with dynamic tasks is differences in long-term memory mechanisms due to the differences in the underlying anatomical bases of the tasks. Consolidation of motor skill has been shown for dynamic transformations that may rely largely on M1, but the evidence of a long-term consolidation process for other motor areas such as the basal ganglia, SMA, and posterior parietal cortex is less consistent. have found that parietal cortex lesions produce a retrograde amnesia that is not temporally graded, inconsistent with the notion that the parietal cortex is involved in the consolidation of a spatial discrimination problem in rats. In the striatum, there is plenty of evidence for short-term consolidation processes, but less consistent evidence for long-term consolidation processes. Various treatments have been found to mediate a striatal-dependent memory when injected into the striatum of rats immediately following training (e.g., ; ; ), indicating the presence of a short-term consolidation process. There is, however, conflicting evidence as to whether a long-term consolidation process that continues beyond 1.5 h exists in the striatum. For example, scopolamine injected into the anterior striatum produced deficits in the retention of a passive avoidance task at 2- and 8-min post-training delays, but not at a 15-min delay , consistent with a short-term consolidation process. Furthermore, immediate inactivation of the NMDA receptors of the ventral striatum post-training produced retention deficits, but inactivation 120 min post-training did not , again consistent with a short-term consolidation process. Whereas another study found that post-training injections of tetrodotoxin in the whole striatum produced deficits in a passive avoidance task at 15 min and 1.5 h, but not at 6 h , consistent with a longer-term consolidation process. Additionally, there is evidence that long-term memory mechanisms in the basal ganglia may rely on more immediate changes in the tuning of neural connectivity that occur with practice rather than a slower long-term consolidation process that occurs in the absence of practice. For example, have found that the basal ganglia are important in the on-line modification, and continual tuning of birdsong with practice and feedback and are not responsible for the maintenance of birdsong in the absence of practice with feedback. Whether or not the striatum is involved in consolidation also appears to depend on how well the task is learned or how strong the learning experience is. Functional inactivation studies of the striatum have found that positively rewarded tasks that are well mastered fail to show temporally graded retrograde amnesia , as do passive avoidance tasks that use a very strong footshock . These findings imply that overlearned striatal-dependent tasks may not rely on the striatum for consolidation. It may be that subjects in our sequence-learning experiment had overlearned the SRT task, and therefore, we failed to see a pattern of temporally graded retroactive interference consistent with a long-term consolidation account. The goal of the current set of experiments was to assess whether a long-term consolidation process is ubiquitous for different components of motor skill learning -- in particular, components of motor skill learning not involving dynamic information. Although it is possible that a short-term consolidation process (i.e., <5 min) may have been at work in the current tasks, in both the learning of a movement sequence and in the learning of a new perceptual-motor adaptation, we found that remote memories were not less susceptible to interference from new learning as predicted by a consolidation account. These results are inconsistent with the notion that a process of long-term consolidation is ubiquitous for different components of motor skill learning. MATERIALS AND METHODS : Experiment 1 | Subjects | A total of 122 University of Virginia students (43 male, mean age = 20.5 years) participated in the study either as partial fulfillment of a course requirement or for payment of $15 --$20. A total of 25 students participated in the control condition, 26 in the 5-min group, 25 in the 1-h group, 22 in the 5-h group, and 24 in the 24-h group. Design | All subjects trained on sequence A and were tested on their memory for that sequence 48 h later. The time at which training on the incompatible sequence (sequence B) took place was varied. Subjects trained on sequence B either 5 min, 1 h, 5 h, or 24 h after training on sequence A. A fifth group, the control group, did not train on sequence B. Stimuli and Apparatus | Stimuli appeared on a video monitor and subjects made responses on a computer keyboard, both controlled by a Macintosh G3 computer. Four boxes (2-cm square, center-to-center distance 5 cm), arranged horizontally on the screen, appeared continuously throughout the SRT task. On each trial, a black filled circle (.3 cm diameter) appeared in the center of one of the boxes. Procedure | SRT Task: Subjects rested the index and middle fingers of each hand on the z, c, b, and m keys of the computer keyboard. Subjects responded to the circle by pressing the key that corresponded to the circle's location. The keys corresponded to the circle locations such that z, the leftmost key, corresponded to the leftmost box and likewise for the remaining locations. Once the subject made the correct response, the circle disappeared and after 250 msec, appeared in a different location. Upon an incorrect response, a tone sounded for 120 msec, and the circle remained on screen until the subject made the correct response. In the first experimental session, subjects trained on sequence A in the SRT task. For each subject, sequence A was selected randomly from a corpus of 563 12-unit sequences that met the following criteria: a stimulus could not repeat itself (e.g., 1332); each stimulus appeared an equal number of times in the sequence; the sequence could not contain runs of four units (e.g., 1234) or trills of four units (e.g., 2424). For each subject, sequence B was selected randomly from the corpus of sequences with the constraint that it did not share any triplets (i.e., did not share three stimulus positions in a row) with that subject's sequence A. Sequenced blocks were created by appending the stimulus sequence to itself 8 times (for a total of 96 trials). Random blocks were created by pseudorandomly selecting eight different sequences from the corpus and appending them, with the stipulation that the entire random block met the same criteria detailed above. Sequence Training: Training on sequence A and sequence B proceeded identically. To minimize the acquisition of explicit sequence knowledge during training, subjects performed a concurrent task. Prior to each block of SRT trials, a series of seven letters appeared on the screen for 5 sec. Subjects were instructed to commit the letters to memory and retain them throughout the block of SRT trials. Subjects were asked to report back the letter string at the end of the block. Subjects performed six blocks of SRT with the concurrent task. The first of these was a random block, followed by four sequenced blocks and another random block. Immediately following, the subjects performed three blocks of SRT without the concurrent task, one random, one sequenced, one random. It has been shown that a concurrent memory task may disrupt performance in SRT, but has only a small effect, if any, on learning . The measure of learning was taken during these last three blocks when the subjects were not performing the concurrent task. Subjects were not informed of the repeating sequence in the SRT task, but were told to respond on each trial as quickly and as accurately as possible. Subjects were told that the memory task and the SRT task were equally important. A 48-h test of sequence A: Forty-eight hours following training on sequence A, subjects were tested on their retention of that sequence. Subjects performed three blocks of SRT without a concurrent task: random, sequenced, random. Measures of explicit knowledge: To assess whether subjects became explicitly aware of the sequence, after the 48-h test of sequence A, we asked subjects to rate their confidence in having seen a sequence during each of the three sessions. All subjects were told that they may or may not have been in a group that saw a repeating sequence of stimuli during initial training. Subjects rated their confidence regarding which condition they were in on a scale from 1 (confident in random group) to 7 (confident in sequenced group). Subjects first made a confidence rating for the session they had just completed and then made similar confidence ratings for their second and first training sessions. Subjects rated their confidence on the basis of how they felt while performing the SRT task in the relevant session. A free recall task was administered to assess explicit knowledge of the sequence. Subjects were told that during the last training session the stimuli had sometimes appeared in a 12-unit repeating sequence and were asked to recall as much of the sequence as possible. On a sheet of paper containing 12 rows and 4 columns, with each row corresponding to a unit of the sequence and each column corresponding to a screen location, subjects indicated where a stimulus had appeared by putting an X in the corresponding column. Subjects then made similar recall judgments for their second and initial training session. Experiment 2 | Subjects | A total of 104 University of Virginia students (42 male, mean age = 18.8 years) participated in the experiment either in partial fulfillment of a course requirement or for a payment of $15 --$20. A total of 32 subjects participated in the control condition, and 24 subjects participated in each of the other conditions. All subjects were right-handed and had normal or corrected-to-normal vision. Design | The design of the current experiment was identical to that of Experiment 1, with the exception that there were only four levels of the between-subjects variable, the time at which the incompatible prism displacement, displacement B, was learned. The 1-h group was excluded in the current study. Stimuli and Apparatus | Subjects threw balls of clay (1-in diameter) at three targets. Targets were three black strips of tape (1.25 in x 7 in), placed vertically on a wall 8-in apart and 72 in from where the subject stood. The center target corresponded to the subject's body midline. Subjects' vision was displaced 12.4 to the left or to the right with prism goggles. The prism goggles were created by adhering 3M Press-On Optics fresnel prism lenses of 20 diopters to clear plastic safety goggles. Procedure | Learning of Displacement A: Subjects were tested individually. They made underhanded throws with their right hand. Each subject made 15 practice throws to the center target without wearing goggles. During these practice throws, no measurements were taken. For all subsequent throws, the experimenter handed the subject a clay ball and called out the target they should throw to on that trial. The clay balls were of various colors and when thrown against the wall, they left a small, colored grease mark. After each throw, the experimenter measured and recorded the horizontal distance (in inches) between the center of this grease mark and the closest edge of the target. After every five throws, the experimenter erased the grease marks. After practice, subjects made three throws to the center target while wearing clear goggles (no prism). These first three measured throws were used as a baseline to assess aftereffects. Subjects then put on the prism goggles and performed three blocks of exposure trials. One 12-trial exposure block consisted of 4 sets of trials. Each set consisted of one throw to each of the three targets. Within each set, the order of the targets was pseudorandomly determined. After each block of exposure trials, subjects took off the prism goggles and made three throws to the center target while wearing clear, nondisplacing goggles, allowing for assessment of aftereffects across the learning period. Learning of Displacement B and Testing on A | Training on the second displacement (displacement B) and the 48-h test of displacement A proceeded identically to training on displacement A, with the exception that subjects did not repeat the initial practice throws. During training on displacement B, those subjects who had trained with the rightward displacing goggles during initial training now trained with the leftward displacing goggles and vice versa. At the 48-h test, all subjects trained with the same set of goggles they initially trained with. Backmatter: PMID- 0 TI - AB - Keywords: null: null Backmatter: PMID- 12915560 TI - Comparative Study of Regulation of RTA-Responsive Genes in Kaposi's Sarcoma-Associated Herpesvirus/Human Herpesvirus 8 AB - Replication and transcription activator (RTA) (also referred to as ORF50), an immediate-early gene product of Kaposi's sarcoma-associated herpesvirus (KSHV)/(human herpesvirus 8), plays a critical role in balancing the viral life cycle between latency and lytic replication. RTA has been shown to act as a strong transcription activator for several downstream genes of KSHV. Direct binding of RTA to DNA is thought to be one of the important mechanisms for transactivation of target genes, while indirect mechanisms are also implicated in RTA transactivation of certain selected genes. This study demonstrated direct binding of the DNA-binding domain of RTA (Rdbd) to a Kaposin (Kpsn) promoter sequence, which is highly homologous to the RTA-responsive element (RRE) of the PAN promoter. We undertook a comparative study of the RREs of PAN RNA, ORF57, vIL-6, and Kpsn to understand how RTA regulates gene expression during lytic replication. Comparing RNA abundance and transcription initiation rates of these RTA target genes in virus-infected cells suggested that the transcription initiation rate of the promoters is a major determinant of viral gene expression, rather than stability of the transcripts. RTA-mediated transactivation of reporters containing each RRE showed that their promoter strengths in a transient-transfection system were comparable to their transcription rates during reactivation. Moreover, our electrophoretic mobility shift assays of each RRE demonstrated that the highly purified Rdbd protein directly bound to the RREs. Based on these results, we conclude that direct binding of RTA to these target sequences contributes to their gene expression to various extents during the lytic life cycle of KSHV. Keywords: Introduction : Kaposi's sarcoma-associated herpesvirus (KSHV), also known as human herpesvirus 8 (HHV-8), is an etiological agent of all clinical forms of Kaposi's sarcoma (KS), including AIDS-associated KS, classic KS, endemic forms of KS, and renal transplant-related KS . KSHV is also associated with two AIDS-related lymphoproliferative diseases, primary effusion lymphoma (PEL) and multicentric Castleman's disease . Like other herpesviruses, KSHV exhibits two distinct phases of its life cycle, latency and lytic replication. Although latent infection by KSHV has been proposed to be critical for tumorigenesis, it is clear that lytic replication is also important in KSHV pathogenesis. Most tumor cells in KS and PEL are latently infected with KSHV; however, the virus undergoes lytic replication in a small number of cells in these tumor lesions . Upon reactivation of KSHV from latency to lytic replication, the infected cells express viral chemokines and proinflammatory cytokines, such as viral macrophage inflammatory proteins and viral interleukin-6 (vIL-6) . These lytic gene products may play a key role in disease progression in an autocrine or paracrine fashion . In addition, antibodies against lytic proteins are significantly elevated in the sera of patients developing KS . Treatment with ganciclovir, a drug that inhibits lytic replication of herpesviruses, significantly reduced the incidence of KS development in AIDS patients at high risk for KS . Therefore, it is important to study the molecular mechanisms of reactivation for a better understanding of KSHV pathogenesis. KSHV replication and transcription activator (RTA) (also referred to as ORF50) is an immediate-early gene product encoded primarily by open reading frame 50 (ORF50), which is well conserved among all gammaherpesviruses . Ectopic expression of RTA is sufficient to disrupt latency and activate lytic replication to completion in latently KSHV-infected PEL cell lines . A dominant-negative mutant of RTA with a 160-amino-acid deletion at the C terminus inhibited the ability of RTA to reactivate KSHV from latency . Thus, the expression of KSHV RTA is sufficient and necessary for viral reactivation, indicating a central role of RTA in the switch of the viral life cycle from latency to lytic replication. In Epstein-Barr virus (EBV) (a gamma-1 herpesvirus), ZEBRA (also referred to as BZLF1, Zta, or Z) plays a key role in disruption of viral latency in synergy with RTA . The RTA homologues from gammaherpesviruses function as a transcriptional activator, and RTA autostimulates its own expression . The Oct-1, Sp1, and Sp3 cognate sites appear to be involved in RTA auto-activation . KSHV RTA also interacts with cellular factors, such as the cyclic AMP-responsive element binding protein-binding protein (CBP), histone deacetylase-1, RBP-Jkappa, and MGC2663 (named KSHV RTA binding protein) . EBV RTA has been shown to interact with CBP and the retinoblastoma protein . KSHV RTA transactivates the expression of several downstream genes, including polyadenylated nuclear (PAN) RNA, Kaposin (K12), ORF57, K-bZIP (K8, the ZEBRA homologue of KSHV), K5, K9, ORF6 (single-stranded DNA binding protein), ORF59 (DNA polymerase-associated processivity factor), TK, K14 (vOX-2), viral G-protein coupled-receptor, and viral interleukin-6 (vIL-6) . EBV RTA alone can activate a subset of lytic promoters that do not require ZEBRA, and vice versa . In other cases, RTA activates downstream genes in synergy with ZEBRA. However, the molecular mechanism by which RTA activates target genes during lytic replication has not yet been elucidated. To date, both direct and indirect mechanisms have been suggested for RTA transactivation . We previously showed a very strong binding affinity of RTA to the RTA-responsive element (RRE) of the PAN promoter (pPAN RRE; 5'-GCTTCCAAAAATGGGTGGCTAACCTGTCCAAAATATGGGAAC-3') . A sequence search of the KSHV genome revealed a sequence remarkably similar to that of pPAN RRE in the Kaposin promoter. This sequence of the Kaposin (Kpsn) promoter (5'-CCCGGGAAATGGGTGGCTAACCCCTACATAAGCAGTTTG-3') contains a 25-bp sequences (shown in italics) with 16-bp and additional 5-bp matches with pPAN RRE (shown in bold type). Other research groups have shown that the RTA-responsive sequence of ORF57 (pORF57 RRE; 5'-GCAAGTGTAACAATAATGTTCCCACGGCCC-3') confers direct binding affinity for RTA . A highly homologous sequence of the pORF57 RRE was also found in the K-bZIP promoters (5'-TATTTGTGAAACAATAATGATTAAAGGGG-3') and has been shown to have a binding affinity to RTA . However, a recent study by Liang et al. showed that a cellular transcriptional factor, RBP-Jkappa, plays an important role in recruitment of RTA to the ORF57 promoter , suggesting a complicated mode of gene regulation by RTA. The RREs of the ORF57 and KbZIP promoters do not appear to have any significant sequence homology with those of the PAN and Kpsn promoters. Another RRE was recently identified which has been shown to regulate vIL-6 gene expression (pvIL-6RRE; 5'-GTGGTTCTAAGTCGCACGTTAGAAACCCCGCCCCCTGGTGCTCACTTT-3') and to confer binding affinity for RTA . Again, there is no obvious homology in the RRE of the vIL-6 promoter to other known RREs. It will be intriguing to investigate how RTA can regu-late target gene expression during lytic replication, utilizing diverse sets of responsive sequences. We undertook a comparative study of a group of viral early genes that are strongly activated by RTA. We measured steady-state RNA levels and transcription initiation rates in virus-infected cells, promoter activities of reporter plasmids in a transient-transfection system, and in vitro RTA-DNA binding affinities. The results from these analyses suggest that direct DNA binding of RTA make a significant contribution to the expression of these target genes during lytic replication. MATERIALS AND METHODS : Cell culture. | All cells were cultured at 37C in the presence of 5% CO2. 293T (a human embryonic kidney cell line transformed with the E1 region of adenovirus and the simian virus 40 [SV40] T antigen) and 293 cells were cultured in Dulbecco's modified Eagle's medium (Cellgro) containing 10% fetal bovine serum (FBS) and antibiotics (50 U of penicillin/ml and 50 mug of streptomycin/ml). The KS-1 cell line was derived from primary effusion lymphoma (PEL) patient at the University of California at Los Angeles (UCLA), which is the same type of lymphoma as the BC-3 cell line, infected with KSHV. KS-1 cells were maintained in RPMI 1640 medium supplemented with 10% FBS and antibiotics. BJAB or DG75, a human B-cell line not infected by any herpesvirus, was grown in RPMI 1640 containing 10% FBS and antibiotics. Plasmid construction. | To make the reporter construct pE4T/Luc, the adenovirus E4 minimal TATA box region was PCR amplified from plasmid Z6E4Tluc (kindly provided by M. Carey, UCLA), using a pair of primers, E4T-F/BglII (5'-tctggaagatctgactctagaggatcc-3') and E4T-R/HindIII (5'-atacccaagcttacaccactcgacacg-3'), and Pfuturbo DNA polymerase (Stratagene, La Jolla, Calif.). In all sequences, the underlined nucleotides represent restriction enzyme sites. The PCR product was cut with BglII and HindIII and cloned into the pGL3-basic plasmid (Promega, Madison, Wis.) which had been cut with the same enzymes. The luciferase reporter construct pE4T/RREs contains a single copy of the RREs in the promoters of PAN RNA, Kaposin (Kpsn), ORF57, and vIL-6 in a forward orientation. Each RRE sequence is shown in Fig. . We used the double-stranded oligonucleotide pan1 (5'-cgggatccGCTTCCAAAAATGGGTGGCTAACCTGTCCAAAATATGGGAACAgatcttcg-3') containing a region from nucleotide (nt) -78 to -37 in the PAN promoter sequence (nt 28588 to 28629 in the KSHV genome) to construct pE4T/PAN. The uppercase letters indicate viral sequences. The double-stranded oligonucleotide pan1 was cut with BamHI and BglII and cloned into pE4T/Luc, which had been cut with BglII. The double-stranded oligonucleotide Kpsn (5'-cgggatccCCCGGGAAATGGGTGGCTAACCCCTACATAAGCAGTTTGagatcttcg-3'; nt 118863 to 118825) was used to generate constructs for pE4T/Kpsn, K-ORF572 (5'-cgggatccGCAAGTGTAACAATAATGTTCCCACGGCCCagatcttcg-3'; nt 81900 to 81929) for pE4T/ORF57, and K2p25 (5'-cgggatccGTGGTTCTAAGTCGCACGTTAGAAACCCCGCCCCCTGGTGCTCACTTTagatcttcg-3'; nt 18337 to 18290) for pE4T/vIL-6. In addition, each RRE was also cloned into the pGL3 promoter plasmid (Promega) containing the SV40 promoter in a similar manner, resulting in pSV40/PAN, pSV40/Kpsn, pSV40/ORF57, and pSV40/vIL-6 in a reverse orient-ation. The copy numbers and the orientations of the inserts were confirmed by sequencing analysis. For nuclear run-on assays, the sequences of PAN RNA, Kaposin, and ORF57 were PCR amplified from BC-1 genomic DNA, using a set of primers specific for each gene. pan10/KpnI and pan2/EcoRI were used for pPAN, containing the region from nt 28659 to 29803 of the KSHV genome, KaposinF (5'-ccggaattcACGGAGGACGGATCTCTTGG-3') and KaposinR (5'-cgcggatcctctagaGAACATGTGGCACACGCTGC-3') for pKpsn (nt 118741 to 117454), and kORF57F (5'-gccgagaattcATGGTACAAGCAATGATAGAC-3') and kORF57R (5'-cgcggatcctctagaGCCATGGGGTTTGGCAATCCTTAAG-3') for pORF57 (nt 82069 to 83564). The PCR product for PAN was cloned into the pcDNA3 plasmid. A TA cloning kit (Invitrogen) was used to clone pKpsn and pORF57 in the pCRII vector provided by the manufacturer. Slot blot analysis of RNA. | Total RNA was extracted from KS-1 cells uninduced or induced with either chemicals or RTA transfection, using TriReagent (Molecular Research Center, Cincinnati, Ohio) according to the manufacturer's instructions. Five micrograms of RNA was denatured with 7.4% formaldehyde and 6x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate) for 15 min at 65C and applied to nylon membranes using the slot blot apparatus. Various amounts of DNA containing the gene of interest without any vector sequences were also denatured in 0.2 N NaOH and then neutralized with an equal volume of 10 M ammonium acetate. The denatured DNA mixtures were dotted onto the same nylon membrane to serve as a standard. DNA fragments only containing a gene of interest were labeled, using a random priming method. The membrane was prehybridized and then hybridized with labeled probes, as previously described . The membranes were exposed to a phosphoimage screen, and quantitative analysis of signals was performed using a STORM imaging system (Molecular Dynamics, Sunnyvale, Calif.). Nuclear run-on assays. | Nuclei from 4 x 107 KS-1 cells untreated or treated with 3 mM sodium butyrate for 18 h were isolated as previously described . Briefly, cells were washed twice with cold 1x phosphate-buffered saline (PBS) and resuspended in 8 ml of cold lysis buffer (10 mM Tris-HCl [pH 7.4], 10 mM NaCl, 3 mM MgCl2). The cells were then centrifuged, resuspended in 4 ml of cold lysis buffer with 0.5% NP-40, and incubated on ice for 20 min. The lysates were passed 10 times through a 25G5/8 needle and incubated on ice for another 10 min. The nuclei pellet was resuspended in 210 mul of nuclear storage buffer (50 mM Tris-HCl [pH 8.0], 40% glycerol, 5 mM MgCl2, 0.1 mM EDTA) and stored at -80C. Nuclear run-on assays were performed as previously described . Each reaction consisted of 210 mul of nuclei and 60 mul of 5x nuclear run-on buffer (5x consisting of 25 mM Tris-HCl [pH 8.0], 12.5 mM MgCl2, 750 mM KCl, 1.25 mM ATP, GTP, and CTP). [alpha-32P]-UTP (30 mul; 3,000 Ci/mM) was then added, and the nuclear suspension was incubated at 30C for 30 min, after which DNaseI (345 U; GibCO-BRL, Grand Island, N.Y.) in 10 mM CaCl2 was added. After 5 min at 30C, the reaction was stopped with 1x SET (1% sodium dodecyl sulfate [SDS], 5 mM EDTA, 10 mM Tris-HCl [pH 7.4]), and proteinase K was added to a concentration of 200 mug/ml. After 45 min of incubation at 37C, the solution was extracted with an equal volume of a mixture of phenol and chloroform, and the interphase was again extracted with 100 mul of 1x SET. Ammonium acetate (10 M) was added to the combined aqueous phases (original plus reextraction) to a final concentration of 2.3 M, an equal volume of isopropanol was added, and nucleic acid was precipitated (-70C for 15 min). The precipitate was centrifuged, and the pellet was resuspended in 100 mul of TE (10 mM Tris-HCl [pH 8.0], 1 mM EDTA). Unincorporated nucleotides were removed using a G-25 spin column. NaOH was added to the elute at a final concentration 0.2 N and incubated for 10 min on ice. HEPES was added to a concentration of 0.24 M, and then 2.5 vol of ethanol was added to precipitate labeled RNA overnight at -20C. The pellet was resuspended in hybridization buffer (10 mM TES [pH 7.4], 0.2% SDS, 10 mM EDTA, 0.3 M NaCl, 1x Denhardt's solution, yeast tRNA [250 mug/ml]). Nylon membranes containing denatured plasmid DNAs were prepared using a slot blot apparatus (GibCO-BRL). Plasmids used for the templates were pPAN (a 1.1-kb fragment of the PAN RNA sequence in pcDNA3), pKpsn (a 1.3-kb fragment of the Kpsn ORF and GC-rich upstream direct repeats in pCRII), pORF57 (a 1.5-kb fragment of the ORF57 coding sequence in pCRII), pvIL-6 (a 0.6-kb fragment of vIL-6 ORF in pcDNA3.1), and pGAPDH (glyceraldehyde-3-phosphate dehydrogenase [GAPDH]; a 0.6-kb fragment of the 3'-end of GAPDH in pCRII). After overnight prehybridization, the membranes were hybridized with the nuclear run-on products (107 cpm) for 36 h at 65C. The radioactivity of labeled products was measured using a scintillation counter, LS1800 (Beckman Coulter, Fullerton, Calif.). After hybridization, the membranes were washed twice with 0.1% SDS and 2x SSC for 30 min and then washed twice with 0.1% SDS and 0.1x SSC for 30 min at 65C. The membranes were exposed to a phosphoimage screen, and quantitative analysis of signals was performed using a STORM imaging system (Molecular Dynamics). Transfections. | For the luciferase reporter assays, 2.5 x 105 293T cells were transfected with 100 ng of pcDNA3 or pcDNA3/RTA and 20 ng of a reporter plasmid into 12-well plates, using a calcium-phosphate transfection method. Each transfection for reporter assays was performed in duplicate and contained 10 ng of pRLSV40 (Promega) as a control for transfection efficiencies, as well as 270 ng of pcDNA3. At 24 h posttransfection, cells were washed with 1x PBS and subjected to reporter assays. BJAB or KS-1 cells (5 x 105) were transfected with 250 ng of pcDNA3 or pcDNA3/RTA and 50 ng of a reporter plasmid into 24-well plates, using the Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer's instructions. In addition, each transfection contained 10 ng of pRLSV40 and 700 ng of pcDNA3. At 24 h posttransfection, cells were harvested, washed with 1x PBS, and subjected to reporter assays. Dual luciferase assays. | The dual luciferase reporter assay system (Promega) was used to test promoter activity. Transfected 293T cells in 12-well plates were washed with 1x PBS and incubated with 250 mul of 1x passive lysis buffer provided by the manufacturer. Transfected BJAB or KS-1 cells were resuspended in 100 mul of 1x passive lysis buffer after being washed with 1x PBS. Lysates were frozen, thawed once, and centrifuged at top speed in a microcentrifuge for 5 min. Lysates were assayed using a Luminometer Lmax (Molecular Devices, Sunnyvale, Calif.). The reporter assays were carried out according to the manufacturer's protocol for the dual luciferase reporter assay system (Promega). The Renilla luciferase activities from the pRLSV40 plasmid were not affected in cotransfection of RTA in the cell lines tested, including 293, 293T, BJAB, and KS-1 cells, and were therefore, used as an internal control to normalize transfection efficiency in different transfection samples. EMSAs. | The recombinant Rdbd protein (amino acids 1 to 320) tagged with a FLAG peptide at the N terminus and six histidine residues at the C terminus was expressed in bacteria and purified as previously described . A set of double-stranded oligonucleotides, including pan1 (pPAN RRE), MJmulti (a mutant version of pPAN RRE), Kpsn (pKpsn RRE), Kpsn/TG (a mutant version of pKpsn RRE), K-ORF572 (pORF57 RRE), and K2p25 (pvIL-6 RRE) were used for electrophoretic mobility shift assays (EMSAs), (see Fig. and for sequences). Two pairs of unrelated sequences, NS1 (5'-cgagatcggggtgaggcatgggggatcccg-3') and NS2 (5'-cgggatccgagatcgaagtgaggcatgggggatgagatcttcg-3'), were synthesized and serve as negative controls for EMSA. All double-stranded oligonucleotides were end labeled with [gamma-32P]ATP followed by fill-in reaction, and EMSAs were performed as previously described . For competition assays, various amounts of cold competitors (indicated in Fig. ) were incubated with the purified Rdbd protein 30 min prior to the addition of end-labeled probes. Supershift assays were performed using a monoclonal antibody (2.2 mug) against a FLAG peptide (Sigma) and polyclonal rabbit sera against RTA (1 mul). Normal rabbit sera (1 mul) as well as polyclonal rabbit sera against an irrelevant protein (1 mul) were also used as a negative control for supershift assays. RESULTS : Comparison of RTA binding affinities for PAN and Kaposin promoters. RTA has been shown to act as a strong transcription activator for several downstream genes of KSHV, including PAN RNA, ORF57, K-bZIP, Kaposin, vIL-6, K14, viral G-protein coupled-receptor, and thymidine kinase . We previously used a highly purified RTA binding domain protein (Rdbd) to demonstrate a very strong binding affinity of RTA for the RRE in the PAN promoter (pPAN RRE; 5'-GCTTCCAAAAATGGGTGGCTAACCTGTCCAAAATATGGGAAC-3'). Through sequence analysis of the KSHV genome, it was noted that the Kaposin promoter bears remarkable homology with the pPAN RRE sequence . The sequence of the Kaposin (Kpsn) promoter (5'-CCCGGGAAATGGGTGGCTAACCCCTACATAAGCAGTTTG-3') contains a 25-bp homologous sequence of pPAN RRE (shown in italics) with a 16-bp consecutive sequence and 5-bp additional exact matches (shown in bold type). We tested whether this sequence in the Kaposin promoter confers RTA binding affinity. End-labeled probes, PAN* and Kpsn* (asterisks indicate labeled oligonucleotides), were incubated with 20 or 50 ng of Rdbd and yielded a specific complex with Rdbd, suggesting that RTA binds to the Kaposin promoter (Fig. , lane 5). Rdbd binding affinity for Kpsn* was fourfold lower than that for PAN* . Throughout the paper, we calculated percent binding of an RRE as a ratio of the bound probe to the total (bound plus unbound) probe using a software program (ImageQuant version 1.1) to quantitate signals from EMSA gels. Our previous detailed analysis of pPAN RRE showed that mutations at the 3' end (nt -46 to -54) of the shared sequence with Kpsn significantly reduced Rdbd binding affinities in vitro compared to the wild type (wt). To test whether Kpsn* mimicked the pPAN RRE mutants, which have different nucleotides at the positions critical for Rdbd binding, we introduced a 2-bp mutation into Kpsn (CC->TG). Now, Kpsn/TG shares 27 bp matches, including the 19-bp consecutive sequence and 5 mismatches at nt -52, -49, and -46 to -44, of pPAN RRE . Once percent binding of each probe, PAN*, Kpsn*, or Kpsn/TG*, was calculated, we set PAN binding as 100% to show Kpsn* and Kpsn/TG* binding of RTA in a scale relative to that for PAN*. EMSA results with 32P-labeled Kpsn/TG* showed that an Rdbd binding affinity for Kpsn/TG* is higher than that for Kpsn* but still lower than that for PAN* (Fig. , lanes 4 to 6, and 1C). This difference between Kpsn* and Kpsn/TG* is statistically significant (P < 0.05, t test) and consistent with the results from pPAN RRE mutagenesis studies of members of our group . MJmulti*, containing a 5-bp mutation in the PAN RRE , showed no significant Rdbd binding, indicating the importance of these nucleotides in RTA binding (Fig. , lane 2). Although it contains four out of these five important nucleotides, the same as in pPAN RRE, Kpsn/TG showed approximately 40% relative binding of pPAN RRE, suggesting that other mismatches at nt -52, -49, and -46 might also contribute to reduced binding of pKpsn RRE. A single-base pair mutation at nt -52 (MJ52S: C->A) showed 60% relative binding of wt pPAN RRE. Two-base pair mutations at nt -49 and -48 (MJ49: AA->CC) or at nt -47 and -46 (MJ47: AT->CG) showed 40 to 50% of wt binding. Hence, Kpsn/TG behaves similarly to these pPAN RRE mutants in terms of reduced Rdbd binding, suggesting that the Kpsn sequence may be a derivative of a very strong RRE of the PAN promoter. Transcript abundance and transcription initiation rates of RTA-responsive viral genes during lytic replication. | Others have reported that RTA expressed using a baculovirus expression system binds to the RREs in the ORF57 promoter (pORF57 RRE) . In addition, a highly homologous sequence of pORF57 RRE was noted in the KbZIP promoter and was shown to confer RTA binding activity . There is no apparent homology between pORF57 RRE and pPAN RRE . The RRE from the vIL-6 promoter (pvIL-6 RRE) was identified and shown to bind to bacterially expressed RTA protein . This sequence is unique among other known RREs, without any apparent homology . Thus, we undertook a comparative study of these RTA target genes to understand how RTA activates diverse sets of responsive elements. First, we sought to determine relative expression levels of these genes in virus-infected cells. We assessed the abundance of RTA-responsive gene transcripts in KS-1 cells during lytic replication, using slot blot analysis. Total RNAs were extracted from KS-1 and DG75 (a human B-cell line without viral infection) cells untreated or treated for 18 h with 3 mM sodium butyrate, a known strong activator of KSHV lytic replication. Total RNA (5 mug) immobilized on a membrane was hybridized with the probe specific to each sequence of the RTA-responsive genes tested. The membrane also included DNA of known amounts for each gene to serve as a standard. Gene expression levels in KS-1 cells from the hybridized membranes were calculated and compared to the signals obtained from the standards . No cross-hybridization to RNA was detected from DG-75 cells (data not shown). There were relatively high levels of Kpsn and vIL-6 transcripts in uninduced cells. The amount of PAN RNA was estimated to be 3.9 ng out of 5 mug of total RNA from induced KS-1 cells. Assuming the portion of poly(A) RNA in total RNA in eukaryotic cells is 1 to 4% and the percentage of PEL cells undergoing lytic replication upon chemical induction is 10 to 20%, PAN RNA was calculated to constitute 10 to 80% of total polyadenylated RNA, consistent with our previous observation . Although expressions of other genes were somewhat comparable in induced KS-1 cells, besides PAN RNA being by far the most abundant, relative expression levels were in order of Kpsn, ORF57, and vIL-6. The expression level of a gene is determined by two factors, frequency of transcription initiation and RNA stability. Therefore, we further examined the transcription rates of the RTA-responsive viral genes, PAN RNA, Kpsn, ORF57, and vIL-6, using nuclear run-on assays. Nuclear run-on assays provide a measure of the frequency of transcription initiation that contributes to regulated gene expression and is largely independent of RNA stability. The numbers of nascent transcripts on the gene are thought to be proportional to the numbers of paused RNA polymerase complexes on the DNA template, indicating the frequency of transcription initiation. Nuclei from KS-1 cells untreated or treated with 3 mM sodium butyrate for 18 h were isolated and incubated with ATP, CTP, GTP, and radiolabeled UTP. The majority of radiolabeled nascent transcripts from uninduced and induced KS-1 cells ran equivalent to the size of 200 to 300 nt, suggesting that the nuclear run-on reactions were efficient (data not shown). Radiolabeled newly synthesized RNAs from uninduced and induced KS-1 nuclei were hybridized with a membrane containing immobilized DNA templates of RTA-responsive viral genes, as well as GAPDH as a control. DNA templates included genomic sequences encompassing the ORF of each gene in a vector, as described in Materials and Methods, except PAN RNA and Kpsn. A template for PAN contains 1.1 kb of the PAN RNA genomic sequence, and that for Kpsn contains upstream sequences of GC direct repeats in addition to the Kpsn ORF (K12) . Results from the nuclear run-on assays are shown in Fig. . For quantitative analysis of data, the signal for GAPDH in uninduced KS-1 cells was set as 1, and those for other genes were quantitated relative to the value of GAPDH. Compared with those from uninduced KS-1 cells, high levels of radiolabeled nascent transcripts of these genes were detected in induced KS-1 cells, which reflects the dramatic increase in transcription rates of these genes upon induction of lytic replication . Similar to results obtained from slot blot analysis, it was noted that there was limited but detectable radiolabeled viral lytic transcription from uninduced KS-1 cells (Fig. , left panel). In induced KS-1 cells, the transcription initiation rates of tested viral genes were consistent with their RNA abundance, suggesting that expression of these RTA target genes is mainly regulated at the transcriptional level, rather than at the posttranscription levels (Fig. , right panel). For direct comparison of the effect of RTA on endogenous RTA-responsive gene expression, we also performed transfection of RTA in KSHV-infected cells and obtained similar results. Either pcDNA3 vector alone or pcDNA3/RTA was transfected into KS-1 cells. Total RNA was harvested at 40 h posttransfection and subjected to the slot blot analysis . Upon RTA expression, viral transcripts in RTA-transfected cells were strongly induced and correlated with their transcription initiation rates. We also obtained similar results when we used a different KSHV-infected cell line, BCBL-1 (data not shown). Slot blot analysis of RNAs from KS-1 and BCBL-1 cells treated with 3 mM sodium butyrate for 40 h also showed comparable trends (data not shown). All together, these data suggest that RTA regulates its target gene expression primarily at the level of transcription of the promoters. The transcription activities of the RRE promoters mediated by RTA in reporter assays. | The RREs have been identified for all four promoters and shown to be critical for promoter activation by RTA . We investigated whether promoter strengths of RTA-responsive genes would be a key factor for their gene expression. Since the distances between a TATA box and an RRE and putative binding sites for cellular proteins varied in their own promoter contexts, we tested the strength of the RREs in a uniform background. All the RRE sequences are listed in Fig. , with numbers indicating genomic locations. These RREs were fused to a heterologous promoter, using a luciferase reporter construct, pE4T/Luc, containing the adenovirus E4 TATA box. A single copy of each RRE was cloned into the same locus of the reporter plasmid, pE4T/Luc, and termed pE4T/RREs . Each reporter construct (20 ng) was transfected into 293T cells with 100 ng of a vector, pcDNA3, or an RTA expression plasmid, pcDNA3/RTA. Plasmid pRLSV40 (10 ng), containing the coding sequence for Renilla luciferase under the control of a constitutively active SV40 enhancer/promoter, was included in each transfection and served as an internal control for transfection efficiency. At 24 h posttransfection, cell lysates were assayed for both firefly and Renilla luciferase activity. Firefly luciferase activity was normalized with the corresponding Renilla luciferase activity in each transfection. The normalized levels of luciferase activity of pcDNA3-transfected cells were subtracted from that of pcDNA3/RTA-transfected cells and then used as levels of the promoter activity. pE4T/PAN manifested the strongest promoter activity among the reporter constructs tested. Its fold activation by RTA was up to 3,300-fold, equivalent to results using our previous reporter construct containing a minimal PAN promoter, pLUC (-69) . The relative promoter activity was calculated as a ratio of the promoter activity of reporter constructs to that of pE4T/PAN . The transfections were performed at a suboptimal condition for RTA to avoid saturating the promoters. The promoter activities of the pE4T/RRE reporters in the presence of RTA had an order similar to that of transcription rates during reactivation of KSHV-infected cells. We then explored the possibility that B-cell-specific or viral factors are involved in activating these elements by testing the reporter constructs in BJAB, a human B-cell line with no viral infection, and in KS-1, a PEL cell line latently infected with KSHV. Reporter constructs (50 ng) were transfected into BJAB cells in the absence or presence of pcDNA3/RTA (250 ng). The ratio between the amount of reporter constructs and that of activator transfected into BJAB cells was the same as in 293T cells. pRLSV40 (10 ng) was cotransfected to serve as an internal control. KS-1 cells were also transfected in the same condition as BJAB cells. Promoter activity of pKpsn RRE in comparison of that of pPAN RRE was less strong in BJAB cells than in 293T or KS-1 cells . Differences among promoter activities of other RREs were less prominent in KS-1 cells than in 293T cells, suggesting that there might be other factors involved in activation of RREs from the Kpsn, ORF57, and vIL-6 promoters . However, the relative promoter strengths of pE4T/RREs in BJAB and KS-1 cells were similar to those in 293T cells, indicating that RTA is the main contributing factor to the promoter activity. We compared relative promoter strengths in a different promoter context by cloning a copy of each RRE into another heterologous promoter, such as SV40 promoter . Results from transient transfections in 293 cells were consistent with previous results showing RRE promoter strengths, although levels of the promoter activity were lower in 293 cells than in 293T cells, mainly due to high basal activity of the reporters . After all, these results suggest that the RREs may be a primary contributor in regulating gene expression, regardless of B-cell-specific or viral factors and types of heterologous promoters. RTA binds viral RREs with various affinities. | Although each of four RREs has been shown to confer RTA binding using different protein expression systems by different groups, a recent study showing that RTA might be recruited by a cellular transcriptional factor, RBP-Jkappa, to the ORF57 promoter, suggested a complicated mode of gene regulation by RTA. In addition, results from the mutagenesis study of pORF57 RRE indicated that there might be a sequence requirement for this promoter activation by RTA, other than a simple RBP-Jkappa binding site , thus serving as a composite site. Since both direct and indirect mechanisms were implicated in RTA transactivation of a target gene, the significance of DNA direct binding of RTA has yet to be revisited in terms of regulation of RTA-responsive gene expression. Therefore, we set out to test whether RTA can bind to these diverse sequences under the same conditions. To compare relative binding affinities of RTA to these viral RREs, increasing amounts of Rdbd (0, 50, 150, and 500 ng) were incubated with the labeled oligonucleotides, including PAN*, Kpsn*, ORF57*, and vIL-6*. These labeled probes bound to the Rdbd protein in a dose-dependent manner, but the binding affinities for Rdbd varied, depending on the RRE sequence . For example, PAN* yielded nearly saturated binding to 150 ng of Rdbd (Fig. , lane 3), while vIL-6* showed only limited binding to 500 ng of Rdbd (lane 16). The arrows indicated RTA-containing complexes, all of which were abolished by anti-RTA antibody and further supershifted by anti-FLAG antibody (data not shown). Distinct conformation of Rdbd proteins that migrated with slightly different mobility in the complexes (e.g., isomers) or different stoichiometry of the RTA protein in the complexes might account for the presence of these multiple complexes . Based on these results, the strength of Rdbd binding affinities is in the following order: PAN*, Kpsn*, ORF57*, and vIL-6*. Relative Rdbd binding affinities were further confirmed by cross-competition assays . To make direct comparisons of Rdbd binding between the two homologous groups of RREs, the labeled oligonucleotides, PAN* and ORF57*, were incubated with 10 ng or 1 mug of Rdbd, respectively, in the absence or presence of cold competitors. After several preliminary tests, various preselected amounts of the unlabeled oligonucleotides, PAN and ORF57, were used to compete with PAN* binding to Rdbd and competed efficiently . It was noted that the amounts of the unlabeled oligonucleotide, ORF57, that were required to cross-compete PAN* binding at similar levels were significantly higher than those of the unlabeled oligonucleotide, PAN, consistent with the previous observation that PAN has a higher binding affinity to Rdbd than ORF57. Binding of Rdbd to the labeled probe, ORF57*, was also competed more efficiently by the unlabeled PAN oligonucleotide than by the unlabeled oligonucleotides containing ORF57. These results confirmed that Rdbd has a higher binding affinity for pPAN RRE than for pORF57 RRE. More comprehensive comparisons of relative Rdbd binding were made using PAN* with a panel of cold competitors corresponding to each RRE. A set of unlabeled oligonucleotides was incubated with PAN* in the presence of 10 ng of Rdbd . The amounts of cold competitors required to reduce complex formation at similar levels varied and correlated with the relative strengths of Rdbd binding affinities to these RREs. The weaker the binding affinity, the more a competitor was required. We also determined relative binding affinities, using the labeled oligonucleotides with weaker binding affinities than PAN*, to test whether the strength of the RRE used as a probe would influence relative cross-competitions with other RREs. Competition assays for the labeled oligonucleotide, Kpsn*, in the presence of 50 ng of Rdbd were performed as described for Fig. . The competition data for Kpsn* correlated with the results from competition assays using PAN* to determine relative binding affinities . The specificities of the complexes were also shown by supershift analysis with anti-FLAG antibody (Fig. and C, lane 10). Likewise, the Rdbd-DNA complexes with other RREs, including ORF57* and vIL-6*, were supershifted by anti-FLAG antibody, indicating the specificity of these complexes (data not shown). An unlabeled oligonucleotide, vIL-6, when used in large amounts, appeared to compete more efficiently for Rdbd binding to PAN* and Kpsn* than ORF57 (Fig. and C, lanes 7 and 9). However, our own previous study showed that Rdbd binding affinity of pORF57 RRE was comparable to that of pvIL-6 RRE overall as determined by cross-competition assays against each other . Binding of Rdbd to these elements was sequence specific, since excess of nonspecific competitors (NS1 and NS2) with unrelated sequences failed to compete this binding (Fig. , lanes 4 to 7). In addition, polyclonal rabbit sera against RTA inhibited formation of the protein-DNA complex while anti-FLAG antibody supershifted the complex, indicating the specificity of the complex (Fig. , lanes 9 and 10) . As negative controls for supershift assays, normal rabbit sera and polyclonal rabbit sera against an irrelevant protein were used and neither altered nor supershifted the formation of the complex (Fig. , lanes 11 and 12). Taken together, these results indicate that RTA binds to the RREs in a sequence-specific manner with various binding affinities, which might contribute to transactivation of RTA target genes. DISCUSSION : In our efforts to understand how RTA, a single viral transactivator, concomitantly activates many target genes to various expression levels during lytic replication, we compared four RTA-responsive viral early genes containing RREs in their promoter regions. We undertook several independent approaches to measure steady-state RNA levels, transcription rates, and promoter strengths to examine how RTA regulates target gene expressions. We found that the strength of RREs in the promoters may be one of the major contributors in expression of the four target genes tested during lytic replication. Our data also suggest that DNA direct binding of RTA to the RREs with different affinities at least partially account for their various levels of gene expression. This leads us to hypothesize that varying binding affinities serves as one of the mechanisms for the virus to express multiple genes at various levels, using a single transcription activator. This is intriguing, given the difference in the molecular switch system of reactivation between EBV (gamma1) and KSHV (gamma2); while KSHV seems to make a main use of RTA, another human gamma herpesvirus, EBV utilizes two viral transactivators, ZEBRA and RTA, in regulation of viral gene expression. Delineating a mechanism underlying RTA regulation of KSHV viral genes may highlight similarities and differences in viral gene regulation of two similar herpesviruses. The roles of cellular factors involved in regulating RTA-target gene expression are not fully understood at present. Our results indicating that relative strengths among the RTA-responsive promoters varied, depending on the assays used, suggest that in conjunction with RTA, there are likely to be cellular or other viral factors mediating the activation of these viral genes in vivo. Since each assay is designed to examine a different aspect of gene regulation, it would not be surprising to find variations among these assays. For instance, we detected dramatic differences among the relative affinities of Rdbd binding to the pPAN RRE and other known RREs. To achieve a similar level of competition, 10- to 100-fold more of the unlabeled oligonucleotide ORF57 than the unlabeled oligonucleotide PAN was required . Given that the Kd of Rdbd for the pPAN RRE was determined to be approximately 8 x 10-9 M , the Kd of Rdbd for other RREs are in a 10-8 to 10-6 M range, which is considered to be a reasonable binding affinity for a transcription activator. However, we considered these results to determine relative affinities of Rdbd binding in a qualitative fashion rather than in a quantitative one, because cold competitors may not compete in a linear manner. In contrast, in terms of promoter strengths, differences in reporter analysis results between pPAN RRE and pORF57 RRE were less than twofold, which is much less than that by EMSAs (Fig. and D), suggesting that not only RTA binding but also other cellular or viral factors may play a role in these gene expressions regulated by RTA. For example, a recent study showed that an RTA-interacting cellular protein, RBP-Jkappa, recognizes a part of pORF57 RRE and plays a critical role in RTA transactivation of pORF57, while RBP-Jkappa does not seem to work on pPAN RRE in the same manner . As a viral protein, K-bZIP, the positional homologue of EBV/ZEBRA, is shown to repress RTA transactivation of the ORF57 promoter through RTA-KbZIP protein-protein interactions but not of the PAN promoter . In addition to promoter-specific factors, a cellular or viral factor common to target genes may play an important role in RTA transactivation. The defined RREs tested in this study were 30 to 50 bp in length, which could accommodate a binding site of other factors in conjunction with RTA, although interaction of RTA with other factors that do not directly bind to DNA is also a possible mechanism. Alternatively, a cellular protein might facilitate recruitment of RTA and other coactivators to its responsive regions without direct interaction with RTA . Recently we have found that a non-histone DNA architectural protein, HMG-1, can further increase KSHV RTA binding to all of these target sites (M. J. Song, S. Hwang, W. Wang, J. Round, B. Wang, R. C. Johnson, M. Carey, and R. Sun, unpublished results). Importantly, enhancement of RTA binding by HMG-1 was greater to low-affinity binding sites than to high-affinity ones, which, at least in part, would account for smaller differences seen in reporter assays in vivo than in binding assays in vitro. While binding assays and reporter analyses utilized artificial systems, nuclear run-on assays were performed in vivo in a natural genomic context. Transcription rates of PAN and ORF57 showed more than a sixfold difference in nuclear run-on assays . The distances between each RRE and the TATA box in its own promoter, the efficiency of promoter demethylation, and the local concentrations of activators during lytic replication could account for larger differences in this assay than in reporter analyses. The Kpsn locus was transcriptionally more active than the ORF57 locus during lytic replication, but fold activation in the transcription rate of ORF57 was higher than that in Kpsn compared with their expressions during latency . This might be due to a complicated transcriptional regulation pattern of the Kpsn locus, depending on the viral life cycle, as discussed later. Nevertheless, despite the quantitative variations, the results showing the relative strengths of the promoters, as determined from all assays, are consistent and strongly support our hypothesis that DNA binding of Rdbd makes significant contributions to expression levels of PAN RNA, Kspn, ORF57, and vIL-6. We used slot blot analysis to quantitate steady-state RNA levels. Compared to nuclear run-on assays, this assay enabled us to determine whether any posttranscriptional regulation may contribute to RNA abundance. Since the ORF57 product was shown to enhance PAN RNA expression as a posttranscriptional activator, the expression level of PAN RNA could be further increased . Members of our group and others previously performed Northern analysis, demonstrating highly specific hybridization patterns for all the genes tested . No splicing event was found to be involved in expression of PAN RNA and vIL-6 . Northern analysis with an ORF57-specific probe and a Kpsn-specific probe gave rise to a single size of the transcript, although ORF57 and Kspn appear to undergo splicing in induced PEL cells ; therefore, it would not affect our interpretation of the results. Thus far, relative expression levels of these genes have not been well addressed. Although global analyses of transcription in KSHV-infected cell lines using an array technique have provided excellent insights on expression kinetics of each gene over time , comparison of relative abundance among different genes would be difficult, possibly due to various hybridization efficiencies of different genes. For example, depending on detection methods employed to determine gene expression levels, i.e., array versus Northern analysis, relative abundance of one gene seems to peak at different time points, suggesting technical differences in different methods . Therefore, our study of four different genes set out to compare relative abundance of RTA-responsive genes using various amounts of DNA standard for each gene. Results from both slot blot analysis and nuclear run-on assays showed high levels of Kpsn and vIL-6 expression in uninduced KS-1 cells . This may reflect a low level of lytic replication in a small subset of KS-1 cells, resulting from spontaneous reactivation. However, relative to PAN and ORF57 in induced KS-1 cells, we observed high transcription rates of Kpsn and vIL-6 in uninduced KS-1 cells, consistent with our slot blot analysis data, suggesting that there might be ongoing transcription of Kpsn and vIL-6 during latency . Kpsn was first identified as the most abundant latent transcript in KS tumor samples, but transcription at this locus was further activated during lytic replication . Kpsn was also expressed in some of the PEL cell lines in a similar pattern . However, it is not well understood whether Kpsn utilizes the same transcription unit during both latency and lytic replication. Our study is mainly focused on the relevant Kpsn transcription unit proximal to Kpsn ORF that is responsive to RTA . Interestingly, a recent study reported an unusual spliced Kpsn transcript in PEL tumor and identified a novel Kpsn promoter (nt 124,242 to 123,842), located farther upstream of the known Kpsn transcription unit . A reporter analysis of the novel Kpsn promoter (-400 + 317Luc) showed a constitutive promoter activity similar150-fold over that of the vector alone in a B-cell line. The result suggests that this region might be active during latency, since its activation does not require any other stimulation, such as by 12-O-tetradecanoylphorbol 13-acetate or RTA. Overall, the kpsn gene locus shows very complicated patterns of transcription as well as translation, depending on tumor types and cell lines tested , and remains to be further elucidated. Constitutive expression of vIL-6 in PEL cells was reported in both primary tumor samples and cell lines, and its expression increased during lytic replication . Thus, these observations are consistent with our interpretation that Kpsn and vIL-6 might be latently transcribed. In conjunction with the fact that PAN RNA is the most abundant lytic transcript of KSHV, the transcription initiation rate of PAN RNA was the highest of four viral genes tested during lytic replication, suggesting that very strong promoter activity may be the main contributor to the abundance of PAN RNA. This is also consistent with interpretation of our previous Northern analysis result showing that the luciferase transcript, driven by the minimal PAN promoter, was expressed at a level comparable to that of PAN RNA driven by the same size of PAN promoter in a transient-transfection system . In addition, the relative abundance of Kpsn, ORF57, and vIL-6 gene transcripts ranks similarly to their relative transcription rates. Therefore, the transcription initiation rates, rather than RNA stability, are likely to be a key factor in determining the transcript abundance of these RTA-responsive genes. Although the RTA-binding sequences tested here are quite diverse, pPAN RRE is strikingly homologous to pKpsn RRE. With variances in the 5' and 3' flanking sequences of these two RREs, the PAN and Kpsn genes showed the highest RTA-binding affinities, promoter strengths, abundance, and transcription rates. Given that PAN RNA and Kpsn genes are unique in KSHV among gammaherpesviruses, it is interesting that the virus utilizes similar regulatory elements to efficiently control expression of these two unique genes during lytic replication. An independent study also reported importance of the conserved regulatory sequence between pRAN RRE and pKpsn RRE . Combined deletion of 5' and 3' flanking sequences in pKpsn RRE failed to show RTA binding and reduced the promoter activity, suggesting that these flanking sequences also contribute to RTA binding as well as transactivation of pKpsn RRE. In conclusion, our comparative study of RTA-responsive gene expression revealed that direct binding of RTA to the RREs makes a significant contribution to activation of the promoters and gene expression during lytic replication. These results led us to hypothesize that when the virus undergoes lytic replication, RTA, a single viral transactivator, activates the different levels of viral gene expression by modulating its binding affinities to the diverse responsive elements. Although viral gene expression is regulated in much more complicated ways than just RTA direct binding alone, our results affirmatively suggest that modulating DNA binding affinity of RTA to various target sites is at least one of the mechanisms by which KSHV RTA regulates viral gene expression during lytic replication. FIG. 1. : RTA DNA-binding domain protein (Rdbd) binds a Kpsn promoter sequence which is highly homologous to the RRE of the PAN promoter. RTA DNA-binding domain protein (Rdbd) binds a Kpsn promoter sequence which is highly homologous to the RRE of the PAN promoter. (A) Comparison of the pPAN RRE and the homologous Kpsn promoter sequence. The pPAN RRE (PAN) shares significant homology to the Kpsn promoter sequence, and matched sequences are aligned. MJmulti is a mutant version of pPAN RRE with 5 nt mismatched , as indicated with arrows. For Kpsn/TG, a mutation of Kpsn (CC->TG) was introduced to further liken Kpsn to pPAN RRE. (B) EMSA of PAN*, MJmulti*, Kpsn*, and Kpsn/TG*. End-labeled probes were incubated with 0, 20, or 50 ng of Rdbd, as indicated. Rdbd was expressed in bacteria with a FLAG peptide at the N terminus and 6x histidine residues at the C terminus and purified using a Ni+-nitrilotriacetic acid column. (C) Quantitative analysis of Rdbd binding. Rdbd binding affinities to Kpsn* and Kpsn/TG* were calculated relative to those for PAN*. The values represent averages of relative binding affinities from three independent EMSAs, with the standard deviation shown. A diamond symbol indicates statistically significant difference in Rdbd binding between Kpsn* and Kpsn/TG* (P < 0.05, t test). FIG. 2. : Transcript abundance and transcription rates of RTA-responsive genes in KS-1 cells during lytic replication. Transcript abundance and transcription rates of RTA-responsive genes in KS-1 cells during lytic replication. (A) Abundance of RTA-responsive genes during lytic replication. Total RNA was extracted from KS-1 cells untreated (-) or treated (+) with 3 mM sodium butyrate at 18 h postinduction and denatured with formaldehyde. In addition to 5 mug of RNA from KS-1 cells, different amounts of denatured DNA corresponding to the gene of interest were dotted onto nylon membranes and served as the standard (STD). The membranes were subjected to hybridization with gene-specific probes, including PAN RNA, Kpsn, ORF57, and vIL-6. GAPDH served as a control. (B) Nuclear run-on assays of KS-1 cells untreated (-) or treated (+) with 3 mM sodium butyrate. Nuclei were isolated from uninduced and induced KS-1 cells at 18 h postinduction and incubated with NTPs and [alpha-32P]UTP. Nylon membranes containing 5 mug of indicated plasmids, PAN, Kpsn, ORF57, vIL-6, and GAPDH (see Materials and Methods), were hybridized with nuclear run-on products. Equal amounts of run-on products (107 cpm) were added to each membrane. Representative images of the hybridized membranes are shown. (C) Quantitative analysis of nuclear run-on assays in uninduced (left panel) and induced (right panel) KS-1 cells. The intensity of a signal represents the amount of labeled nascent transcripts, which reflects the transcription initiation rate of the gene. The signal for GAPDH in uninduced KS-1 cells was set as 1, and the signals for other genes were analyzed relative to the value of GAPDH. FIG. 3. : Expression of RTA-responsive genes in KS-1 cells transfected with RTA. Expression of RTA-responsive genes in KS-1 cells transfected with RTA. Total RNA was extracted from KS-1 cells transfected with pcDNA3 vector alone (Vec) or with pcDNA3/RTA (RTA) at 40 h posttransfection and denatured with formaldehyde. In addition to 5 mug of RNA from transfected KS-1 cells, different amounts of denatured DNA corresponding to the gene of interest were dotted onto nylon membranes and served as the standard (STD). The membranes were subjected to hybridization with gene-specific probes, as described in the legend to FIG. . FIG. 4. : Promoter strengths of reporters containing each RRE under a heterologous promoter, the E4 TATA box. Promoter strengths of reporters containing each RRE under a heterologous promoter, the E4 TATA box. (A) Schematic representation of a reporter construct. One copy of each RRE was cloned into a reporter construct containing the adenovirus E4 minimal TATA box. Reporter constructs were tested in 293T (B), BJAB (C), and KS-1 (D) cells, using a transient-transfection system. At 24 h posttransfection, transfected cells were harvested and subjected to dual luciferase assays. Promoter activities were calculated from levels of luciferase activity of reporters transfected with pcDNA3 or pcDNA3/RTA, in addition to a control vector, pRLSV40, levels of which constitutively expresses Renilla luciferase. Relative to that of pE4T/PAN, the promoter activities of the other pE4T/RREs were calculated. The values represent averages of at least three independent transfections, with the standard deviation shown. FIG. 5. : Promoter strengths of reporters containing each RRE under another heterologous promoter, the SV40 promoter. Promoter strengths of reporters containing each RRE under another heterologous promoter, the SV40 promoter. (A) Schematic representation of a reporter construct. One copy of each RRE was cloned into a reporter construct containing the SV40 promoter. (B) Reporter constructs were transiently transfected into 293 cells, and levels of luciferase activity were assayed as described in the legend to FIG. . FIG. 6. : Rdbd binds with various affinities to the RREs of PAN, Kpsn, ORF57, and vIL-6 promoters. Rdbd binds with various affinities to the RREs of PAN, Kpsn, ORF57, and vIL-6 promoters. (A) Sequences of the RREs from the promoters of PAN, Kpsn, ORF57, and vIL-6 of KSHV. Numbers indicate the locations of each RRE in the KSHV genome. (B) Dose-dependent binding of Rdbd to the RREs. Each probe contained an RRE with common flanking sequences at both ends. Increasing amounts of Rdbd (0, 50, 150, and 500 ng) were incubated with end-labeled probes, PAN* (lanes 1 to 4), Kpsn* (lanes 5 to 8), ORF57* (lanes 9 to 12), and vIL-6* (lanes 13 to 16), respectively. Arrows indicate Rdbd-specific binding. FIG. 7. : Cross-competition assays confirm relative affinities of Rdbd binding to the RREs. Cross-competition assays confirm relative affinities of Rdbd binding to the RREs. (A) Cross-competition assays between PAN and ORF57 RREs. Rdbd (10 ng) was incubated without (lane 2) or with excess unlabeled oligonucleotide, PAN (5- and 50-fold; lanes 3 and 4), or ORF57 (50-, 500-, and 1,000-fold; lanes 5 to 7), for competition assays prior to addition of PAN*. Indicated amounts of unlabeled oligonucleotide, PAN (0.1-, 0.5-, and 5-fold; lanes 9 to 11) or ORF57 (fivefold; lane 12), were mixed with 1 mug of Rdbd, followed by addition of ORF57*. Note that the amounts of the unlabeled oligonucleotide PAN used to compete for Rdbd binding were far less than those of the unlabeled oligonucleotide ORF57. (B) Cross-competition assays of PAN* (10 ng of Rdbd) were performed with a set of unlabeled oligonucleotides (PAN, Kpsn, ORF57, and vIL-6). As indicated, different amounts of cold competitors were used to produce similar levels of competition. Unlabeled oligonucleotides were preincubated with Rdbd, and an end-labeled probe, PAN* or Kpsn*, was added (lanes 2 to 9). A monoclonal antibody against the FLAG peptide was used to supershift protein-DNA complexes containing Rdbd (lanes 10). (C) Cross-competition assay of Kpsn* (50 ng of Rdbd) was performed as described for panel B. (D) Specificity of Rdbd binding to the RRE. The labeled PAN* probe were incubated in excess of unlabeled oligonucleotides NS1 and NS2, both of which contain nonspecific sequences, as negative controls. Coincubation of PAN* with unlabeled oligonucleotides containing its own sequence (PAN) resulted in efficient competition of the complex formation (lanes 2 and 3), while NS1 and NS2 did not show any significant competition (lanes 4 to 7). Supershift assays were performed with anti-FLAG antibody and polyclonal rabbit sera against RTA (lanes 9 and 10). Normal rabbit sera and polyclonal rabbit sera against irrelevant protein (lanes 11 and 12) were also included as negative controls. Backmatter: PMID- 12915534 TI - Adenovirus Type 11 Uses CD46 as a Cellular Receptor AB - The 51 human adenovirus serotypes are divided into six species (A to F). Many adenoviruses use the coxsackie-adenovirus receptor (CAR) for attachment to host cells in vitro. Species B adenoviruses do not compete with CAR-binding serotypes for binding to host cells, and it has been suggested that species B adenoviruses use a receptor other than CAR. Species B adenoviruses mainly cause disease in the respiratory tract, the eyes, and in the urinary tract. Here we demonstrate that adenovirus type 11 (Ad11; of species B) binds to Chinese hamster ovary (CHO) cells transfected with CD46 (membrane cofactor protein)-cDNA at least 10 times more strongly than to CHO cells transfected with cDNAs encoding CAR or CD55 (decay accelerating factor). Nonpermissive CHO cells were rendered permissive to Ad11 infection upon transfection with CD46-cDNA. Soluble Ad11 fiber knob but not Ad7 or Ad5 knob inhibited binding of Ad11 virions to CD46-transfected cells, and anti-CD46 antibodies inhibited both binding of and infection by Ad11. From these results we conclude that CD46 is a cellular receptor for Ad11. Keywords: Introduction : The adenovirus family contains 51 human serotypes, which are divided into six species designated A through F . Most human adenoviruses cause disease in the respiratory tract, kidney, eyes, intestine, or lymphoid tissue . The antigenic determinants on the outer parts of the capsid are mainly contributed by three polypeptides: the hexon, the penton base, and the fiber. The hexon protein is the major constituent of the particle, and each of the 12 vertices carries a pentameric penton base protein bearing one trimeric fiber protein . The knob domain of the viral fiber protein has been shown to mediate attachment to host cells in vitro, whereas the penton base facilitates the subsequent internalization step . Adenovirus type 2 (Ad2) and Ad5 (both of species C) have been the best characterized with respect to host cell interactions and both have been demonstrated to utilize the coxsackie-adenovirus receptor (CAR) for binding and infection of host cells in vitro . In addition, adenovirus members from species A, D, E, and F have been demonstrated to interact with soluble CAR in vitro . The CAR-binding site is conserved on CAR-binding fibers and has been localized to the AB-loop in the knob domain . CAR belongs to the immunoglobulin superfamily and has been shown to immunoprecipitate, together with both ZO-1 and beta-catenin, which are components of tight and adhesion junctions, respectively . Moreover, CAR is capable of forming intercellular homodimers and appears to be an important component in contacts between epithelial cells . CAR is also expressed basolaterally in well-differentiated airway epithelium but not apically . Thus, CAR is not easily accessible for adenoviruses, and it has been suggested that transient breaks in the epithelium may be required in order for adenoviruses to infect host cells via CAR . Recently, a novel function for adenovirus-CAR interactions was described. Before lysis of the infected cells, an excess of virus fibers is produced, and these are secreted basolaterally. Since the affinity of CAR-fiber interactions is greater than that of of CAR-CAR interactions, fibers disrupt intercellular CAR dimers, increase intercellular space, and allow complete particles to filter from the basolateral side and escape apically . Consequently, a main function of fiber-CAR interactions may be to facilitate prelytic viral escape. Species D contains 32 different serotypes, but most of these serotypes are rarely isolated from humans. However, three serotypes (Ad8, Ad19, and Ad37) cause a severe ocular disease designated epidemic keratoconjunctivitis . We have recently shown that these three viruses use sialic acid as cellular receptors instead of CAR and that, in these serotypes, the affinity of the fiber for sialic acid is mediated by a charge-dependent interaction between the unusually positively charged fiber knobs of these viruses (pI 9 to 9.1) and sialic acid (pKa 2.6) . We have also shown that the sialic acid receptor sacharrides are linked to membrane proteins rather than to lipids, but the identity of the sialylated protein(s) is still unknown. Recently, and contrary to our own findings, E. Wu et al. suggested that Ad37 attaches to cells via a 50-kDa membrane protein independently of sialic acid and that this interaction is dependent on divalent cations . In contrast to the results of these workers, we found that the attachment of Ad37 to Chang C cells, and thus infection, is entirely dependent on sialic acid but independent of divalent cations . Species B adenoviruses have been subdivided into B:1 (Ad3, Ad7, Ad16, Ad21, and Ad50) and B:2 (Ad11, Ad14, Ad34, and Ad35) . Within species B, Ad3 and Ad7 are among the most commonly isolated viruses . Several species B adenoviruses cause respiratory and ocular infections. In addition, species B:2 viruses, other than Ad14, cause infections of the kidney and/or urinary tract . Species B adenoviruses do not compete with species C adenoviruses for binding to CAR and do not bind to CAR in vitro . We reported recently that two different species B:2 adenoviruses (Ad11 and Ad35) could block the binding of two different B:1 viruses (Ad3 and Ad7) to host cells but that the B:1 viruses were unable to block binding of the B:2 viruses . Moreover, the binding of B:1, but not B:2, viruses was dependent on divalent cations and sensitive to prior trypsination of host cells. Based on these findings, we suggested that adenoviruses from species B:1 and B:2 interact with a cellular receptor designated sBAR (species B adenovirus receptor). In addition, the species B:2 adenoviruses Ad11 and Ad35 were suggested to interact with an additional receptor designated sB2AR (species B:2 adenovirus receptor) . Adenoviruses in general have many receptor-binding characteristics in common with picornaviruses, including the ability to use CAR, sialic acid, VCAM-1, heparan sulfate, MHC-1, and alphav integrins as receptors or coreceptors . Several picornaviruses are also capable of using decay accelerating factor (DAF; CD55) as a cellular receptor . CD55 and CD46 (membrane cofactor protein [MCP]) are two closely related members of the complement system. CD55 regulates the activity of C3 convertases, whereas CD55 acts as a cofactor for factor 1-mediated cleavage of C3b and C4b protein . CD55 has been suggested to associate directly with CAR , and even if CD55 is not used as an adenovirus receptor, monoclonal anti-CD55 antibodies partially inhibit Ad2 binding to HeLa cells , indicating that many adenoviruses attach to host cells in close proximity to CD55. Both CD55 and CD46 are expressed on most cells throughout the body. However, CD46 is expressed on monkey erythrocytes but not on human erythrocytes, whereas CD55 is expressed on both human and monkey erythrocytes . This is of interest, since it has been shown that species B adenoviruses agglutinate erythrocytes from monkeys but not from humans . With this in mind, we set out to determine whether CD55 and/or CD46 could serve as cellular receptors for human adenoviruses. MATERIALS AND METHODS : Cells, viruses, antibodies, and recombinant fiber knobs. (i) Cells. | Human respiratory epithelial A549 cells were grown in Dulbecco modified Eagle medium (DMEM; Sigma Chemical Co., St. Louis, Mo.), containing 10% fetal calf serum (FCS), HEPES, and penicillin-streptomycin (all from Sigma). Chinese hamster ovary (CHO) cells, including CHO cells expressing human CAR, DAF (CD55), or MCP (CD46) isoforms BC1 or BC2 have been described elsewhere . (ii) Viruses. | Ad5 (strain Ad75), Ad7 (Gomen), Ad11 (Slobitski), and Ad37 (1477) virions were propagated in A549 cells with or without isotope labeling and were purified as described elsewhere . The specific activities for each serotype were 1.5 x 10-5 cpm/particle (Ad5), 1.4 x 10-5 (Ad7), 8.1 x 10-6 (Ad11), and 9.4 x 10-5 (Ad37). The correct identity of each virus strain was confirmed by restriction enzyme digestion of purified virus genomes (data not shown). (iii) Antibodies. | Mouse monoclonal antibodies (clones GB24 and TRA-2-10) and rabbit polyclonal antibodies to human CD46 have been described elsewhere . Mouse monoclonal antibodies (clone E4.3) against human CD46 were purchased from Ancell, Bayport, Minn. Monoclonal anti-CD55 (IF7) and anti-CAR (RmcB) antibodies were kindly provided by J. M. Bergelson, Children's Hospital of Philadelphia, Philadelphia, Pa. Rabbit polyclonal anti-Ad5, anti-Ad7, anti-Ad11, and anti-Ad37 antisera were prepared as described elsewhere . (iv) Recombinant fiber knobs. | DNA isolations and manipulations were performed by standard techniques. DNA fragments encoding the fiber knob of Ad11p (fk11p) were amplified by PCR with the following primers, preserving the last shaft motif and the trimerization signal , as described earlier : fk11p forward (5'-GTA CTG GCC ATG GGA CTT ACA TTC AAT TC03') and fk11p reverse (5'-GC TAC GCA GGA TCC TCA GTC GTC TTC TCT GAT G-3'). The PCR products were cloned into a pET-3d expression vector (Novagen, Madison, Wis.) containing an N-terminal His6 tag motif by using NcoI-BamHI restriction sites. The constructs were confirmed by nucleotide sequence analysis. Proteins were expressed in E. coli (strain BL21) and purified by using Ni-nitrilotriacetic acid (NTA) agarose according to the instructions of the manufacturer (Qiagen, Hilden, Germany). Virion-binding assays. | Unless otherwise stated, the binding assays were performed at least twice with duplicate samples in a final volume of 100 mul/sample by using host cells harvested with phosphate-buffered saline (PBS) containing 0.05% (wt/vol) EDTA and allowed to recover for 1 h in DMEM containing 10% FCS, with constant agitation. 35S-labeled virions were used in all binding experiments. Binding assays. | A total of 2 x 105 cells/well in 96-well microplates were incubated with 104 virions/cell in binding buffer (BB; DMEM, penicillin-streptomycin, HEPES, and 1% bovine serum albumin) and then incubated on ice for 1 h. Nonbound virions were removed by washing, and the cell-associated radioactivity was measured by using a Wallac 1406 scintillation counter. Inhibition of virion binding by antibodies. | A total of 105 cells/well in 96-well plates were incubated on ice with or without (control) 5 mul of GB24 (2.5 mg/ml), TRA-2-10 (2.5 mg/ml), IF7 (cell culture supernatant), RmcB (ascites), rabbit anti-CD46 serum, or serum from two different nonimmunized rabbits. After a further 30 min, 5 x 103 virions/cell were added, and incubation was continued on ice. The cells were then washed, and the cell-associated radioactivity was measured as described above. Effect of divalent cations on virion binding. | A total of 2 x 105 nonrecovered CHO or CHO-BC1 cells/well in 96-well plates were incubated first with 5 mul of rabbit anti-CD46 serum on ice for 1 h in Tris-buffered saline with or without 5 mM CaCl2, MnCl2, or MgCl2 and then incubated with 104 virions/cell in the same buffer. Two hours later, the cells were washed with incubation medium, and the cell-associated radioactivity was measured as described above. Effect of prior trypsination of the cell surface on virion binding. | Adherent CHO-BC1 cells in 75-cm2 culture bottles were harvested on ice with PBS-EDTA or with 200 mug of trypsin (Invitrogen, Paisley, United Kingdom)/ml. Cells harvested with PBS-EDTA were recovered as described above. One hour later, Pefabloc C (a serine protease inhibitor; Roche, Basel, Switerland) was added to a final concentration of 2 mM, and 10 ml of DMEM containing 5% FCS was added to the trypsinized cells. The cells were then washed, and 105 cells/well in 96-well plates were incubated on ice in BB containing 3.6 x 104 Ad7 or Ad11 virions/cell. Inhibition of virion binding by soluble fiber knobs. | A total of 5 x 104 cells/well in 96-well plates were incubated on ice in BB (volume 50 mul) with or without 0.5 mug of recombinant, soluble fiber knobs of Ad11, Ad7, or Ad5. One hour later, 5 x 104 Ad11 virions/cell were added, and the mixtures were incubated for 1 h. Nonbound virions were removed by washing, and the cell-associated radioactivity was measured as described above. Flow cytometry. | Expression levels of CD46 on A549, CHO, CHO-BC1 and CHO-BC2 cells were investigated by using mouse monoclonal (E4.3) anti-human CD46 antibodies (1 mg/ml) diluted 1:200 in PBS supplemented with 10% FCS and 0.01% NaN3 (staining buffer). A total of 105 cells were incubated with the diluted primary antibody in a volume of 100 mul for 45 min on ice and thereafter washed once in 200 mul of staining buffer. For detection, fluorescein isothiocyanate-conjugated rabbit anti-mouse Fab fragments were used (Dako, Glostrup, Denmark) that were diluted 1:20 in staining buffer. After incubation for 45 min on ice, the cells were washed as described above and analyzed in a FACScan (Becton Dickinson) flow cytometer. The data were analyzed by using CellQuest software (Becton Dickinson). Fluorescent focus assay. | A total of 1.5 x 105 adherent cells/well in 24-well plates were incubated with 90,000 virus particles/cell on ice to synchronize infection. One hour later, nonbound virions were removed by washing. In some cases, 18 mul of rabbit anti-CD46 serum or serum from nonimmunized rabbits was incubated with the cells on ice in a final volume of 300 mul before virions were added. After incubation for 44 h at 37C, the cells were fixed in methanol and stained with polyclonal rabbit serum raised against Ad11 virions, followed by fluorescein isothiocyanate-conjugated swine anti-rabbit immunoglobulin G antibodies (Dako). Infected cells were then identified in a fluorescence microscope (x100 magnification; Xiovert 25; Carl Zeiss, Jena, Germany) as described previously . RESULTS : Expression of human CD46 on CHO cells enhances binding of Ad7 and Ad11 virions. | In a first experiment, we investigated the ability of several 35S-labeled adenovirus serotypes to bind to CHO cells transfected with cDNAs encoding human CAR, DAF, or two isoforms of CD46 designated BC1 and BC2. As expected, Ad5 virions attached to CHO-CAR cells 20 times more efficiently than to the other transfectants . Ad37 virions attached to all transfectants with similar efficiency , which is in agreement with the apparent ability of this virus to use sialic acid as a cellular receptor. Ad7 virions attached to CHO-BC1 cells three to four times more efficiently than to the other cells, including the CHO-BC2 cells . Ad11 virions attached at least 10 times more efficiently to CHO-BC1 cells and 4 times more efficiently to CHO-BC2 cells than to CHO-CAR cells and CHO-DAF cells . The only differences between the BC1 and BC2 isoforms reside in their cytoplasmic domains. Thus, the extracellular domains are identical and could not account for differences in virus binding. Instead, the differences in the binding of Ad7 or Ad11 virions to CHO-BC1 and CHO-BC2 cells is in line with the higher expression of the BC1 isoform on CHO cells compared to the expression of the BC2 isoform . These experiments indicated that Ad11 and, to some extent, Ad7 interact with cell surface CD46 proteins. Anti-CD46 antibodies inhibit binding of Ad7 and Ad11 virions to CHO-BC1 cells. | In a second experiment, we preincubated A549 cells or CHO-BC1 cells with antibodies to CD46, CAR, or CD55 and investigated the effect on virus binding. Monoclonal antibodies to CAR (RmcB) or CD55 (IF7) had no effect on the binding of Ad7 and Ad11 . Serum from nonimmunized rabbits was also ineffective. However, rabbit anti-CD46 serum efficiently inhibited the binding of both Ad7 and Ad11 to CHO-BC1 cells to background levels, indicating that these antibodies blocked all specific binding sites. Moreover, the anti-CD46 monoclonal antibody GB24 also inhibited binding of Ad7 and Ad11 to CHO-BC1 cells to background levels (Ad7) or close to background levels (Ad11), whereas the ability of anti-CD46 monoclonal antibody TRA-2-10 to inhibit binding was considerably less. GB24 recognizes an epitope located in the SCR3/4 region, and TRA-2-10 recognizes an epitope in the SCR1 region , which suggests that the virus-binding site is located near the cell membrane rather than being in the membrane-distal domain of CD46. Both GB24 and rabbit polyclonal anti-CD46 inhibited the binding of Ad11 to A549 cells but did not completely block binding. Also, none of the anti-CD46 antibodies inhibited the binding of Ad7 to A549 cells, indicating that A549 cells exhibit additional receptor sites for both Ad7 and Ad11. Thus, these results indicated that our previous hypothesis about two different cellular receptors for species B adenoviruses is correct and that CD46 may be equivalent to the previously suggested species B:2 adenovirus receptor (sB2AR). Expression of CD46 is required for efficient infection of CHO cells by Ad11 virions. | To investigate whether CD46 was involved in infection, and not only attachment, we allowed virions to attach and infect CHO, CHO-BC1, and CHO-BC2 cells . Ad11-infected CHO-BC1 and CHO-BC2 cells more efficiently than it infected CHO cells, indicating that the presence of CD46 facilitates both attachment and subsequent infection by Ad11. On the other hand, Ad7 did not infect any of the transfectants, including the CHO-BC1 cells, suggesting that even if the presence of CD46 facilitates attachment of Ad7, it is not sufficient to support infection. The inhibitory effect of the rabbit anti-CD46 serum was also confirmed in an infectivity assay. Anti-CD46 serum efficiently inhibited Ad11 infection of BC1 cells, whereas sera from two nonimmunized rabbits did not affect infection . However, rabbit anti-CD46 serum did not inhibit Ad11 infection of A549 cells, indicating that these cells express additional receptors beside CD46, which are able to support infection of Ad11 (data not shown). Ca2+ and Mn2+ ions are required for the efficient binding of Ad7 to CHO-BC1 cells. | In an earlier publication, we demonstrated that species B:1 viruses Ad3 and Ad7 require Ca2+ ions for binding to host cells, whereas species B:2 viruses attach to host cells almost independently of Ca2+ and Mg2+ ions . To test whether divalent cations are involved in the binding of species B adenoviruses to CD46-expressing cells, we harvested CHO-BC1 and CHO-BC2 cells in PBS-EDTA, thus removing all divalent cations, and then incubated the cells with 35S-labeled Ad7 or Ad11 virions in the presence of Ca2+, Mn2+, or Mg2+ ions. The binding of Ad7 virions to CHO-BC1 cells was strongly enhanced in the presence of Mn2+ ions and moderately enhanced in the presence of Ca2+ and Mg2+ ions, whereas the binding of Ad11 virions to CHO-BC1 cells was doubled in the presence of Mn2+ and unchanged in the presence of Ca2+ or Mg2+ ions . Rabbit anti-CD46 serum efficiently inhibited the binding of both Ad7 and Ad11 to CHO-BC1 cells in the presence of divalent ions. Thus, binding of Ad7 to CD46-expressing CHO cells requires Ca2+ or Mn2+ ions, whereas Ad11 virions are able to interact with CD46-expressing cells in the absence of divalent cations. Furthermore, divalent ions enhanced the binding of Ad7 to CD46-negative CHO cells with the same efficiency as to CD46-positive cells, indicating that the divalent-ion dependent binding of Ad7 was independent of the presence of CD46. Prior trypsination of CHO-BC1 cells inhibits the binding of Ad7 virions but enhances the binding of Ad11 virions. | It has been shown previously that prior trypsination of A549 cells inhibits the binding of Ad3 and Ad7 virions and enhances the binding of Ad11 and Ad35 virions . To investigate whether adenovirus binding to CD46 is affected by cellular trypsination, we pretreated CHO-BC1 cells with trypsin and investigated the binding of Ad7 and Ad11 virions to these cells. In agreement with previous results (using A549 cells), we found that the binding of Ad11 to CHO-BC1 cells was increased by 41%, which supports our hypothesis that CD46 is sB2AR . In the case of Ad7, the binding to CHO-BC1 cells was reduced to background levels, indicating that all specific binding sites were abolished. Soluble fiber knobs inhibit the binding of Ad11 virions to CHO-BC1 cells. | Most adenoviruses, including species B adenoviruses, attach to host cells mainly through an interaction mediated by the knob domain of the viral fiber protein . To investigate whether species B adenoviruses use the fiber knob for attachment to A549 and CHO-BC1 cells, we preincubated these cells with soluble recombinant Ad11 fiber knobs and then added homologous 35S-labeled virions. As expected, soluble knobs inhibited the binding of homologous virions to A549 and CHO-BC1 cells by 82 and 93%, respectively . Thus, soluble Ad11 knobs efficiently inhibited the binding of homologous virions to both A549 and CHO-BC1 cells, indicating that the knob domain of the viral fiber protein mediates Ad11 interactions with CD46. Fiber knobs from Ad7 or Ad5 did not compete with Ad11 virions for binding to CHO-BC1 cells, which is in agreement with our conclusion that CD46 is utilized as a cellular receptor by Ad11 but not by Ad7 or other serotypes. DISCUSSION : In humans, a cluster of genes designated the RCA (for regulator of complement activation) gene cluster is located on the long arm of chromosome 1 (1q32) . Although the proteins within the RCA family vary in size, they have significant similarities in primary structure. Some of the primary sequence is organized in tandem structural units termed short consensus repeats (SCRs), which are present in multiple copies in the protein . Thirteen different SCR-containing proteins in the complement system have been identified . At least three of these13 (CD55, CD46, and complement receptor 1 [CR1]) are associated with host cell membranes. Both CD55 and CD46 contain 4 SCRs, whereas CR1 contains between 23 and 44 SCRs, depending on the isoform. CD55 is linked to the cell membrane via a glycosyl phosphatidylinositol anchor, whereas CD46 and CR1 contain a transmembrane domain. Despite the different forms of membrane linkage in CD46 and CD55, the overall amino acid identity is 45%. In terms of receptor usage, adenoviruses share many features with picornaviruses. Since many picornaviruses use CD55 as a cellular receptor, we hypothesized that members of the adenovirus family may use CD55 or its close relative CD46 as a cellular receptor. CD55 is expressed widely throughout the body but in relatively low quantity on natural killer cells. CD46 serves as a cellular receptor for certain strains of measles virus and human herpesvirus 6 and is expressed in many different tissues, but the expression appears to be higher in the kidney and in the airway and conjunctival epithelium . Thus, the expression of CD46 correlates well with the tropism of most species B:2 adenoviruses (i.e., Ad11, Ad34, and Ad35). Here we have demonstrated that Ad11 interacts with human CD46-expressing CHO cells but not with equivalent cells expressing CD55 or CAR instead and that anti-CD46 antibodies efficiently inhibit the binding of Ad11 to CD46-expressing CHO cells, indicating that the presence of CD46 on the surface of host cells provides Ad11 with novel virus-binding sites. Also, the presence of CD46 converted nonpermissive CHO cells to be permissive for Ad11 infection, and this infection could be blocked with anti-CD46 antibodies. Furthermore, soluble Ad11 fiber knobs efficiently inhibited the binding of Ad11 to CD46-positive cells, indicating that Ad11, like most other adenoviruses, uses the knob domain of the viral fiber protein for interactions with host cell receptors. Taken together, these results demonstrate that CD46 may serve as a cellular receptor for Ad11. We have previously suggested that all species B adenoviruses (including Ad7 and Ad11) use the cellular receptor designated sBAR and that species B:2 adenoviruses may use an additional receptor designated sB2AR. Here we found a clear difference in the interactions of Ad7 (species B:1) and Ad11 (species B:2) with CD46. Ad7 bound to CD46-expressing cells less efficiently than Ad11 and was also unable to infect CD46-expressing CHO cells. Also, anti-CD46 antibodies inhibited the binding of Ad11 to A549 cells but not the binding of Ad7 to the same cells. Moreover, the binding of Ad11 virions to CD46-expressing cells was not dependent on divalent cations and was elevated upon pretreatment of the cells with trypsin, whereas the binding of Ad7 was almost entirely dependent on divalent cations, and the binding was reduced after trypsination. Finally, Ad7 knobs did not compete with Ad11 virions for binding to CD46-expressing CHO cells. Thus, these experiments confirm our previous suggestion about two cellular receptors for species B adenoviruses. The finding that rabbit anti-CD46 serum inhibits the binding of Ad7 to CHO-BC1 cells stimulated with divalent ions suggests that the cellular receptor for Ad7 is in close proximity to CD46 rather than being CD46 itself. The finding that Ad7 binds to CD46-negative CHO cells with the same efficiency as to CD46-positive cells supports this suggestion. Since Ad11 and Ad35 (also species B:2) compete for the same receptor site and Ad34 (the third of four species B:2 adenoviruses) exhibits high amino acid homology with Ad11 in the fiber knob (99.5%) , we propose that the previously designated species B:2 adenovirus receptor (sB2AR) is equivalent to CD46. The identity of sBAR remains to be determined, and the role of CD46 as a cellular receptor for Ad7 and other species B:1 adenoviruses (i.e., Ad3, Ad16, Ad21, and Ad51) requires further evaluation, but our results suggest two possible mechanisms for Ad7 binding to host cells. (i) Ad7 uses a different isoform of CD46 as a cellular receptor, and the isoforms used in the present study contain a partial region of the binding site, which is sufficient for weak binding but not for subsequent infection. (ii) CD46 associates with an unknown receptor in the cytoplasm of CHO cells and brings this receptor to the cell surface, where it is exposed for Ad7. CD46 is known to associate directly or indirectly with several proteins at the cell surface, such as the alpha2beta1-, alpha3beta1-, and alpha6beta1-integrins , which are cell surface proteins that need divalent cations for heterodimerization and interactions with various ligands. Most CHO cell lines express few or no integrins on the cell surface (-). beta Integrins, for example (including beta1 [130 kDa]), are expressed in these cells but are usually not brought to the cell surface due to the absence of alpha integrins. It has been shown previously that Ad3 (which belong to the same subspecies as Ad7) virions bind to a cell surface protein with a mass of 130 kDa in a divalent-cation-dependent manner . However, transfection of CHO cells with cDNAs encoding human integrins alpha2, alpha2beta1, alpha3, alpha5, or alphaV did not enhance the binding of Ad7 nor Ad11 (data not shown), indicating that none of these integrins corresponded to sBAR. Thus, the role of CD46 during Ad7 binding to host cells is still unclear, and we cannot exclude the possibility that sBAR is nonrelated and nonassociated with CD46. Adenoviruses are frequently used as vectors in gene therapy trials, but with limited success. In most cases, Ad5 is used as a vector backbone. The main disadvantages with Ad5-based vectors appear to be low transduction of target cells and high preexisting immunity . In several cases, species B:2 adenoviruses have been shown to attach to host cells and deliver genes much more efficiently than other adenoviruses, including Ad5 . Thus, we believe that the identification of CD46 as a cellular receptor for these adenoviruses widens the understanding of adenovirus tropism and may also contribute to development of more effective adenovirus-based gene therapy vectors. FIG. 1. : Ad7 and Ad11 virions exhibit increased binding to CHO-CD46 cells. Ad7 and Ad11 virions exhibit increased binding to CHO-CD46 cells. The ability of Ad5 (species C) (A), Ad37 (species D) (B), Ad7 (species B:1) (C), and Ad11 (species B:2) (D) to bind to CHO cells transfected with cDNAs encoding human CAR, CD55 (DAF), and CD46 (BC1 or BC2 isoforms) was investigated as described in Materials and Methods. FIG. 2. : CD46 expression on A549 cells and CHO cell transfectants. CD46 expression on A549 cells and CHO cell transfectants. The expression of CD46 on A549 cells, CHO cells, and CHO cells transfected with cDNAs encoding the BC1 and BC2 isoforms of CD46 was investigated by using flow cytometry with the E4.3 monoclonal anti-CD46 antibodies as described in Materials and Methods. Black and white fields represent cells incubated in the absence and presence of anti-CD46 antibodies, respectively. FIG. 3. : Anti-CD46 antibodies inhibit binding of species B adenoviruses to CHO-BC1 and A549 cells. Anti-CD46 antibodies inhibit binding of species B adenoviruses to CHO-BC1 and A549 cells. The binding of Ad7 (A and C) and Ad11 (B and D) virions to CHO-BC1 (A and B) and A549 (C and D) preincubated with or without (CTRL) antibodies was performed as described in Materials and Methods. GB24 and TRA-2-10, monoclonal antibodies directed against the SCRs 3/4 and 1, respectively; Poly, rabbit anti-human CD46 polyclonal antiserum; RmcB and IF7, monoclonal antibodies directed against human CAR and CD55, respectively; rabbit-1 and rabbit-2, serum from two nonimmunized rabbits. FIG. 4. : CD46 promotes infection of Ad11 in CHO cells. CD46 promotes infection of Ad11 in CHO cells. (A) Ad11 virions were allowed to infect CHO cells or CHO cells transfected with cDNAs encoding the BC1 or BC2 isoforms of human CD46, as described in Materials and Methods. (B) CHO-BC1 cells were preincubated with rabbit anti-CD46-serum before infection with Ad11 virions. At 44 h after infection, the cells were fixed, stained, and visualized by using a fluorescent microscope as described in Materials and Methods. FIG. 5. : Ca2+ or Mn2+ ions promote binding of Ad7 virions to CHO and CHO-BC1 cells. Ca2+ or Mn2+ ions promote binding of Ad7 virions to CHO and CHO-BC1 cells. The ability of 35S-labeled virions to bind to EDTA-pretreated CHO-BC1 (Ad11 [A] and Ad7 [B]) or CHO (Ad7 [C]) cells in the presence of Ca2+, Mn2+, or Mg2+ and either with or without rabbit anti-CD46 serum, was investigated as described in Materials and Methods. FIG. 6. : Trypsination of CHO-BC1 cells inhibits the binding of Ad7 but enhances the binding of Ad11. Trypsination of CHO-BC1 cells inhibits the binding of Ad7 but enhances the binding of Ad11. The ability of Ad7 and Ad11 virions to bind to CHO-BC1 cells either pretreated with trypsin or left untreated was investigated as described in Materials and Methods. FIG. 7. : Ad11 fiber knobs block the binding of Ad11 virions to A549 and CHO-BC1 cells. Ad11 fiber knobs block the binding of Ad11 virions to A549 and CHO-BC1 cells. Binding of Ad11 virions to A549 preincubated with homologous fiber knobs or binding to CHO-BC1 cells preincubated with homologous fiber knobs or fiber knobs from Ad5 or Ad7 was investigated as described in Materials and Methods. Backmatter: PMID- 12915540 TI - Characterization of the RNA-Binding Domains in the Replicase Proteins of Tomato Bushy Stunt Virus AB - Tomato bushy stunt virus (TBSV), a tombusvirus with a nonsegmented, plus-stranded RNA genome, codes for two essential replicase proteins. The sequence of one of the replicase proteins, namely p33, overlaps with the N-terminal domain of p92, which contains the signature motifs of RNA-dependent RNA polymerases (RdRps) in its nonoverlapping C-terminal portion. In this work, we demonstrate that both replicase proteins bind to RNA in vitro based on gel mobility shift and surface plasmon resonance measurements. We also show evidence that the binding of p33 to single-stranded RNA (ssRNA) is stronger than binding to double-stranded RNA (dsRNA), ssDNA, or dsDNA in vitro. Competition experiments with ssRNA revealed that p33 binds to a TBSV-derived sequence with higher affinity than to other nonviral ssRNA sequences. Additional studies revealed that p33 could bind to RNA in a cooperative manner. Using deletion derivatives of the Escherichia coli-expressed recombinant proteins in gel mobility shift and Northwestern assays, we demonstrate that p33 and the overlapping domain of p92, based on its sequence identity with p33, contain an arginine- and proline-rich RNA-binding motif (termed RPR, which has the sequence RPRRRP). This motif is highly conserved among tombusviruses and related carmoviruses, and it is similar to the arginine-rich motif present in the Tat trans-activator protein of human immunodeficiency virus type 1. We also find that the nonoverlapping C-terminal domain of p92 contains additional RNA-binding regions. Interestingly, the location of one of the RNA-binding domains in p92 is similar to the RNA-binding domain of the NS5B RdRp protein of hepatitis C virus. Keywords: Introduction : Viral-coded replicase proteins are essential for replication of tombusviruses, similar to other positive-strand RNA viruses . Tomato bushy stunt virus (TBSV), the prototypical tombusvirus, codes for p33 and p92 replicase proteins (reviewed in reference ), which are essential for TBSV replication . Based on the presence of the signature motifs for RNA-dependent RNA polymerase (RdRp) within the C-terminal portion of p92, the predicted function of the p92 is to synthesize viral RNA progenies . The function(s) of p33 in TBSV replication is currently unknown. An interesting feature of tombusviruses is that the larger replicase protein is expressed from the genomic RNA via a ribosomal readthrough mechanism of the p33 termination codon . Due to the expression strategy, the N-terminal portion of p92 overlaps with p33. Both p33 and p92 have been proposed to be part of the TBSV replication complexes . Accordingly, Western blot analysis of the partially purified TBSV RdRp preparation revealed the presence of both p33 and p92 in the transcriptionally active RdRp fractions (J. Pogany and P. D. Nagy, unpublished data). Both replicase proteins are localized to membranous structures in infected cells ---the putative sites of tombusvirus replication . The membrane localization domains in the replicase proteins are present within the N-terminal overlapping domain . Based on their essential roles in tombusvirus RNA replication, it is possible that both tombusvirus replicase proteins can bind to the viral RNA in infected cells. For example, the viral genomic RNA must be recruited to the viral replicase complex after translation . In addition, the viral replicase complex, which likely contains both viral- and host-coded proteins, must synthesize a complementary (minus-strand) RNA on the genomic (+) RNA. This is followed by robust plus-strand RNA synthesis utilizing the minus-strand intermediates . Tombusviruses also synthesize subgenomic RNAs for the expression of their 3'-proximal genes . In addition, tombusviruses are frequently associated with defective interfering (DI) RNAs that are generated from the genomic RNA by multiple deletions (reviewed in reference ). Synthesis (i.e., replication) of all these RNAs requires both replicase proteins in vivo . Accordingly, a recently developed in vitro assay based on partially purified RdRp from tombusvirus-infected cells was used to demonstrate that the TBSV RdRp could synthesize cRNA on both plus- and minus-stranded TBSV templates . Further studies confirmed that the tombusvirus RdRp (obtained for either TBSV or the closely related cucumber necrosis virus) could bind to TBSV RNA in vitro and recognize cis-acting sequences, such as the genomic and complementary promoters and a replication enhancer , which leads to RNA transcription in vitro. To gain insights into the functions of the tombusvirus replicase proteins, we tested their abilities to bind RNA in vitro, and we report the results in this paper. We demonstrated that both p33 and p92 of TBSV could bind to TBSV-derived RNA sequences in vitro. Gel mobility shift assays performed with a series of truncated recombinant p33 purified from Escherichia coli revealed that an arginine- and proline-rich motif (termed RPR motif), which is conserved among tombusviruses, is critical for efficient RNA binding. The corresponding region in the overlapping domain of p92 may also bind to RNA, although this has not been confirmed. The lack of confirmation is because we also found that p92 had additional RNA-binding domains within its nonoverlapping C-terminal region; therefore, mutations within the RPR motif in p92 did not abolish RNA binding. Overall, these experiments supply direct evidence that both TBSV replicase proteins bind to viral RNA, and this feature might be important for the functions of these proteins during tombusvirus infection. MATERIALS AND METHODS : Construction of expression plasmids. | The full-length TBSV cDNA clone (T-100; generous gift of Andy White ) was used to amplify the p33 open reading frame (ORF) using primers 3 and 4 . The PCR product was kinased, digested with XmnI, and cloned into a protein expression vector pMAL-c2X (NEB) to generate expression construct p33 (Fig. and Table ). Construct p33 contains the in-frame fusion of the maltose-binding protein (MBP) and the p33 ORF. To express the p92 ORF, we used a mutant TBSV clone (pHS-175; supplied by H. Scholthof ), which had the amber (UAG) termination codon replaced by a tyrosine codon at the end of the p33 ORF. The mutated p92 gene was amplified via PCR using primers 3 and 5 and template pHS-175. The obtained PCR product was cloned into pMAL-c2x to generate construct p92 (Fig. and Table ) as described for construct p33 above. A similar strategy was used to generate construct p92C by PCR using T-100 template and primers 6 and 11 , except that we applied XbaI for digestion of the PCR product. All other deletion constructs of p33 and p92 (see Fig. and ) were generated by the method described for p33, with the exception that the obtained PCR products were digested with EcoRI and XbaI for cloning . The primers used to generate the expression plasmids are listed in Table . The expression construct N was generated by digestion of construct p33 with EcoNI, followed by religation. Constructs R1, R2, and R4 were generated by digesting the p92C clone with HindIII, BstBI-HindIII, and EcoNI-HindIII, respectively, followed by religation. Constructs R15 to R22 were generated using alanine-serine scanning mutagenesis via two separate PCRs for each construct carried out with the R6 template. One PCR product represented sequences that coded for the N-terminal portion of R6, while the other PCR product coded for the C-terminal portion of the R6 protein. The obtained PCR products (the primer pairs used for PCR are shown in Table for R15 to R22) were digested with NheI, and the appropriate PCR products representing the modified N- and C-terminal portions of R6 were ligated together, and then reamplified by PCR. The PCR products were then digested with EcoRI and XbaI before being cloned into pMAL-c2X. Purification of p33 and p92 proteins and their derivatives from E. coli. | The expression and purification of the recombinant TBSV proteins were carried out as described earlier for the TCV p88 replicase protein . Briefly, individual expression plasmids (see above) were transformed into Epicurion BL21-CodonPlus (DE3)-RIL (Stratagene). The overnight cultures from the transformed bacterial cells were diluted to 1:100 in rich growth medium (10 g of tryptone, 5 g of yeast extract, 5 g of NaCl) containing 0.2% glucose and 100 mug of ampicillin per ml and grown at 37C until the optical density reached 0.6 to 0.8. Protein expression was then induced at 14C with 0.3 mM IPTG (isopropyl-beta-d-thiogalactopyranoside) for 8 to 10 h. The induced cells were harvested at 4,000 x g at 4C for 10 min, resuspended in ice-cold column buffer (10 mM Tris-HCl [pH 7.4], 1 mM EDTA, 25 mM NaCl, 10 mM beta-mercaptoethanol), sonicated on ice to disrupt the cells, and centrifuged at 9, 000 x g for 30 min at 4C. The supernatant was added to equilibrated amylose resin column (New England Biosciences), washed thoroughly with 20 volumes of column buffer, and eluted with 10 mM maltose in column buffer. All protein purification steps were carried out in a cold room. The purified recombinant proteins were analyzed in sodium dodecyl sulfate-10% polyacrylamide gel electrophoresis (SDS-PAGE) for their purity. The cleavage of MBP/p33 protein was carried out using factor Xa protease (New England Biosciences) as recommended by the manufacturer (1 mug of factor Xa was applied for 50 mug of MBP/p33). To measure the amount of purified recombinant proteins, we used the Bio-Rad protein assay, which is based on the Bradford method. Preparation of RNA templates. | The DNA template representing the 82-bp region III of DI-72 RNA of TBSV was generated by PCR using DI-72XP as a template . The primers used were 253 (5'-TTGGAAATTCTCCTTAGCGAGTAAGACAGACTC-3') and 23 (5'-GTAATACGACTCACTATAGGGACCCAACAAGAGTAACCTG-3'). The PCR template also included the T7 promoter to facilitate synthesis of RNA probes. Both labeled and unlabeled RNAs for gel mobility shift and competition experiments were prepared in vitro using T7 RNA polymerase, as described previously . The labeled RNA probes were obtained using [32P]UTP in the T7 transcription reaction , followed by removal of free nucleotides using P-30 micro Bio-Spin columns (Bio-Rad). Template DNA was removed by DNase I, followed by purification of the RNA transcript with phenol-chloroform extraction and 95% ethanol precipitation. The pellet was washed with 70% ethanol to remove residual salts. The RNA transcripts were quantified by UV spectrophotometry (Beckman), followed by either 1% agarose or 5% PAGE . For competition experiments (see Fig. ), we used unlabeled minus-stranded region III of DI-72 RNA. The dsRNA competitor was generated via annealing of the positive and negative strands of region III RNA. The dsDNA competitor was the PCR-amplified region III DNA fragment, while the ssDNA competitor was an artificially synthesized 51-bp oligo DNA (5'-CCC-AGACCCTCCAGCCAAAGGGTAAATGGGAAAGCCCCCCGTCCGAGGAGG-3'). RNA constructs AU and GC were obtained from Chi-Ping Cheng . Gel mobility shift assay. | Approximately 1 muM protein was incubated with 5 ng of radioactively labeled minus-stranded region III RNA probe (see above) in a binding buffer (50 mM Tris-HCl [pH 8.2], 10 mM MgCl2, 1 mM EDTA, 10% glycerol, 200 ng of yeast tRNA [Sigma], and 2 U of RNase inhibitor [Ambion]) at 25C for 30 min . After the binding reaction, the samples were analyzed by either 4% nondenaturing PAGE performed at 200 V or 1% agarose gel electrophoresis run at 100 V in Tris-borate-EDTA buffer in a cold room . The gels were dried, exposed and analyzed in a phosphorimager, and quantified using ImageQuant version 1.2 (Amersham). For competition experiments, unlabeled competitors (applied in molar excess as indicated in Fig. and ) were added simultaneously with the labeled RNA probe to the binding reaction. The experimental binding curves were statistically fit to data using a Microsoft Excel spreadsheet. Northwestern assay. | Approximately equal amounts (similar2 mug) of recombinant proteins were run in an SDS-10% PAGE and then transferred to polyvinylidene difluoride (PVDF) membranes . The membranes were renatured at room temperature in a renaturation buffer (10 mM Tris-Hcl [pH 7.5], 1 mM EDTA, 50 mM NaCl, 0.1% Triton X-100, and 1x Denhardt's reagent ) with three changes of buffer for 20 min each. The membranes were probed with 32P-labeled RNA (minus-stranded region III; see above) for 1 h, washed three times with the renaturation buffer, air dried, and analyzed using a phosphorimager. Biosensor analysis. | The surface plasmon resonance (SPR) experiments were carried out using BIACORE X (Biacore, Inc., Piscataway, N.J.) at 25C as recommended by the manufacturer. Briefly, the running buffer (10 mM HEPES [pH 7.4], 150 mM NaCl, 3 mM EDTA, 0.05% surfactant P-20) was filtered and degassed every time before use. An SA sensor chip (Biacore) was used to immobilize a 20-mer RNA from the 3' end of gTBSV RNA with a biotin label at the 5' end (5'-UGUAACGUCUUUACGUCGGG-3'; Dharmacon, Inc.). The surface of the chip was first preconditioned with three 1-min pulses of 50 mM NaOH in 1 M NaCl. Flow cell 1 (Fc1) was used to immobilize 440 resonance units (RU) of RNA, while flow cell 2 (Fc2) was kept as a control surface to account for nonspecific binding and bulk refractive index changes upon injection of protein samples as suggested by the manufacturer. The recombinant protein samples were diluted with running buffer to 1 muM final concentration before injection. The interactions between RNA and the recombinant proteins were analyzed in real time through a sensogram , in which the RUs were plotted as a function of time. One RU is equivalent to a change in adsorbed mass of 1 pg/mm2 of the sensor surface (BIA Applications Handbook; Biacore). All the data shown in Fig. were corrected based on data obtained from the control (Fc2). RESULTS : Expression, purification, and RNA binding by recombinant p33 and p92 replicase proteins. | In order to obtain sufficient amounts of soluble TBSV p33 and p92 proteins for biochemical studies, we overexpressed them in E. coli as maltose-binding fusion proteins. This expression strategy allowed for affinity-based purification of the recombinant p33 and p92 proteins, as shown in Fig. . Standard gel mobility shift experiments with a 32P-labeled RNA probe, representing the 82-nt TBSV replication enhancer, termed region III(-) (; this probe was used in all the experiments unless stated otherwise), demonstrated that both the recombinant p33 and p92 could bind to RNA efficiently . In contrast, comparable amounts of bovine serum albumin and the full-length MBP alone (expressed and purified under the same conditions as the recombinant p33 and p92 proteins) did not bind to the RNA probe efficiently , thus ruling out the possibility that MBP or any contaminating proteins from E. coli contribute to RNA-binding activity. Interestingly, the N-terminally truncated version of p92 that included the entire unique sequence of p92 (termed p92C; Fig. ) also bound to RNA efficiently (Fig. , lane p92C). This observation suggests that p92 contains a minimum of two RNA-binding regions ---one in the overlapping region and another in the unique region (see below). Since p33 and p92 have a common RNA-binding region within the overlapping sequence, we used mostly the recombinant p33 in the experiments below (unless stated otherwise). To test whether the presence of the N-terminal MBP fusion might affect the ability of p33 to bind to RNA and/or affect the migration of the RNA-protein complex in the gel, we compared a purified recombinant p33 preparation, which had the MBP cleaved off by protease factor Xa (sample p33 [cleaved], Fig. ), with the uncleaved recombinant MBP/p33 fusion protein preparation (sample MBP/p33, Fig. ) in a gel mobility shift assay (Fig. and F). We found that the overall efficiencies of RNA binding by the recombinant p33 and MBP/p33 fusion protein were similar (Fig. and F), thus suggesting that the MBP fusion does not alter the ability of p33 ---and likely p92 ---to bind to RNA. Also, the mobility of the RNA-protein complex was the same for the two preparations (Fig. and F). Therefore, we used the maltose-binding fusion proteins in the following experiments. To corroborate the results obtained from gel mobility shift analysis, we carried out SPR measurements with a Biacore biosensor as described in Materials and Methods. Briefly, the SPR provides data about real-time protein-RNA interactions by measuring the change in refractive index that takes place between the immobilized RNA and the protein that is being passed in an aqueous buffer over the surface of the chip. The refractive index of the medium changes near the chip surface due to change in mass resulting from RNA-protein interaction . For this study, we fixed a 20-nt-long 5'-biotinylated TBSV RNA, which includes the minimal genomic (i.e., minus-strand initiation) promoter to the surface of a streptavidin-coated chip. The recombinant p33, p92, p92C, and MBP proteins diluted with running buffer were passed separately over the immobilized RNA . These experiments confirmed that p33 , p92 , and p92C could bind efficiently and stably to the RNA, while MBP could not. Detailed kinetic measurements on p33/p92 and RNA interactions will be published elsewhere. Preferential binding of recombinant p33 to single-stranded RNA. | To test whether the recombinant p33 could bind only to single-stranded RNA or to other nucleic acids as well, we used various nucleic acids in template competition experiments, which were evaluated by using gel mobility shift assay. Briefly, the same amounts of 32P-labeled region III(-) RNA probe and purified recombinant p33 were used in the presence of increasing amounts of unlabeled competitors, such as single-stranded RNA (ssRNA), double-stranded RNA (dsRNA), ssDNA, and dsDNA . All these sequences were derived from the same region of the TBSV genome (minus- or double-stranded region III). These experiments demonstrated that the ssRNA was the best competitor in binding to p33, while ssDNA competed moderately well . In contrast, dsRNA and dsDNA templates were poor competitors under the experimental conditions used. Overall, the data suggest that p33 is a single-stranded nucleic acid-binding protein with the highest preference toward ssRNA. We also tested if the recombinant p33 could preferentially bind to a TBSV-derived ssRNA sequence by using four different but comparably sized ssRNAs in competition experiments as shown in Fig. . One competitor was region III(-) [named DI-RIII(-); Fig. ], while the other three RNAs were nonviral. One of these competitor ssRNAs consisted of an AU-rich sequence (named AU; Fig. ) and another consisted of a GC-rich (named GC; Fig. ) artificial sequence , while the fourth RNA was tRNA. Comparison of the abilities of these RNAs to compete with the 32P-labeled ssRNA probe in binding to p33 based on the gel mobility shift experiments revealed that the region III(-) was by far the best competitor among the ssRNAs (Fig. and C). The artificial AU-rich RNA was a moderately good competitor, while the artificial GC-rich RNA and the tRNA were poor competitors (Fig. and C). These experiments suggest that p33 preferably binds to the TBSV-derived sequence and, at a lower efficiency, to an AU-rich sequence. Note that the artificial AU and GC competitor RNAs have double- versus single-stranded regions that are comparable in length to those present in DI-RIII(-) . Thus, increased binding by p33 to DI-RIII(-) is likely due to the favorable sequence and structure of region III(-). Cooperative binding of p33 to ssRNA. | In order to characterize the binding behavior of p33 to TBSV RNA, we incubated progressively increasing amounts of recombinant p33 proteins in the presence of a 32P-labeled probe followed by gel mobility shift assay. In the presence of a small amount of p33 , the RNA probe was found mostly in unbound form, while increasing the amount of p33 in the binding reaction resulted in rapid transition of the RNA probe to bound form (samples on right side in Fig. ). The absence of intermediately shifted bands, resulting from limited binding of the probe by p33 between the completely bound and the free probe in the gel , suggests that most of the RNAs are either coated with p33 or are not bound to p33 at all ---depending on the amount of protein present in the RNA-binding reactions. The Hill coefficient of 1.8 for RNA-binding by p33 also supports the supposition that p33 binds to RNA in a cooperative manner . Interestingly, when we used a truncated p33 containing only a 60-amino-acid segment of p33 that includes the RNA-binding domain (construct C10, Fig. ), then we still observed cooperative RNA binding by this truncated p33, as is clearly evident from the rapid transition from the totally free to completely bound state of probe RNA with a marginal rise in protein concentration . The ability of the truncated p33 to bind cooperatively to RNA was surprising (because of the small size of the protein); therefore, we also used the three-membrane sandwich method as described by Pata et al. to confirm the above finding. Briefly, the first polysulfone membrane, which has low affinity to both RNA and protein, was expected to retain only large RNA-protein complexes . The second nitrocellulose membrane can bind to small protein-RNA complexes (i.e., those that have not been retarded on the first membrane), while the third, positively charged membrane retains all unbound RNA. Increasing amounts of C10 protein were added to the same amount of RNA probe, as described in Fig. . Aliquots of the binding reactions were filtered through the three-membrane sandwich, followed by detection of the retarded 32P-label on each membrane . These experiments demonstrated that the RNA probe was retarded as part of a large complex in the presence of elevated amounts of C10 proteins (see membrane 1, Fig. ), while the formation of small RNA-protein complex was relatively inefficient (Fig. , membrane 2). Overall, the shown data are most consistent with the model of cooperative RNA binding, which predicts that most of the RNA is present in a coated form (seen as a large complex) when a sufficient amount of the truncated p33 is present in the binding reaction. The goal of future experiments will be to analyze the structure of the p33-RNA complex. Mapping the RNA-binding site in p33. | In order to identify the RNA-binding domain in p33 and in the corresponding region of p92, we made a series of 13 overlapping constructs with various truncations within the p33 gene . The truncated p33 proteins were overexpressed in E. coli, followed by affinity purification of the proteins as MBP fusions (see above) and SDS-PAGE analysis . The ability of the obtained truncated p33 proteins to bind to the 32P-labeled probe was tested in gel mobility shift experiments . These experiments revealed that the C-terminal segment of p33 contains an RNA-binding domain (Fig. , lane C1), while the N-terminal segment does not bind to RNA under the in vitro conditions (Fig. , lane N). Further testing of C-terminally nested segments of p33 indicated that the central portion of the C-terminal region harbors the RNA-binding domain (Fig. , compare lanes C2 through C4 versus C5 and C6). This was further supported by the observation that deletions of the C-terminal 20 to 56 amino acids in p33 did not affect RNA-binding (Fig. , lanes C7 and C8). Deletion of 103 amino acids from the C terminus, however, abolished RNA binding (Fig. , lane C9). Deletions starting from both the N and C termini confirmed that the shortest p33 derivative that still bound to RNA was 30 amino acids long and covered the central portion of the C-terminal region (Fig. , lane C11). Deletion of an 8-amino-acid portion of construct C11 ---which has the sequence TGRPRRRP ---completely abolished RNA binding (Fig. , lane C12). Indeed, all the p33 derivatives that carried the above arginine- and proline-rich motif bound to the RNA probe efficiently, while those lacking this domain did not bind to the RNA probe . Based on these data, we propose that the arginine- and proline-rich motif, which we call the RPR-motif, might be the primary RNA-binding site in p33 and p92. To confirm that the RNA binding by the truncated p33 derivatives is an inherent feature of these proteins and not due to the presence of a contaminating protein from E. coli, we analyzed the RNA-binding ability of a selected group of short-deletion mutants of p33 in a Northwestern assay . Briefly, the full-length p33 and four of the purified recombinant p33 derivatives, C9, C10, C11, and C12 , were subjected to SDS-PAGE, followed by blotting to a PVDF membrane. This was followed by probing the membrane with a 32P-labeled RNA probe as described in Materials and Methods. Importantly, we could detect the protein-RNA complex by autoradiography in the portion of the membranes that contained proteins p33 (WT), C10, and C11, but we could not detect RNA-protein complex in the case of C9 and C12 (Fig. , panel Northwestern). Therefore, the Northwestern analysis confirmed the results obtained in the above gel mobility shift experiments, those being that the RPR motif is the core region in RNA binding in p33. Mapping additional RNA-binding sites in p92. | In order to identify the RNA-binding domains present in the unique segment of p92 (Fig. , p92C), we used a series of deletion derivatives of p92C that were overexpressed and purified from E. coli. Surprisingly, many of the expressed truncated p92C proteins were unstable in E. coli, thereby preventing us from generating large enough numbers of p92C derivatives that could have been useful to pinpoint the RNA-binding sites precisely (data not shown). To this end, we were able to obtain 14 truncated p92C proteins in large enough amounts suitable for biochemical assays . Nested truncations of 82 to 133 amino acids from the C terminus of p92C resulted in reduced RNA binding, thus suggesting that this portion of p92C contributes to RNA binding, although the residual RNA binding by these proteins was still significant (Fig. , lanes R1 and R2). Deletion of 434 amino acids from the C-terminal end of p92C abolished RNA binding (Fig. , lane R4). The p92C derivatives carrying segments from the central portion of the protein were found to bind to RNA (Fig. , lanes R5 through R11). In contrast, p92C derivatives, which lacked the above central segment and the C-terminal segment, did not bind efficiently to RNA (Fig. , lanes R12 and R13). Interestingly, protein R14, which contained the 303-amino-acid C-terminal segment of p92C, bound to RNA ---though with reduced efficiency when compared to p92C (Fig. , lane R14). Overall, analysis of RNA binding by the above deletion series of p92C revealed that two segments of p92C are involved in RNA binding: the central segment (represented by R6 and R7) and the very C-terminal segment (similar131 amino acids). These two segments can bind to the RNA independently of each other, although each shows somewhat reduced efficiency when compared to p92C. To further delineate the similar108-amino-acid-long central RNA-binding site in p92C, we separately introduced five clusters of five to seven alanine and serine mutations into construct R6 , as shown schematically in Fig. . We targeted short regions that contained clusters of positively charged amino acids, such as arginine, lysine, and histidine (Fig. , lanes R15 through R19). Surprisingly, all these mutated proteins retained their abilities to bind to RNA (Fig. , lanes R15 through R19), suggesting that either (i) the selected amino acids are not involved in RNA binding or (ii) several amino acids, possibly located at different parts of the central segment of the p92C protein, are brought together by protein folding to form an RNA-binding groove ---as is the case for other viral RdRps . Therefore, it is possible that the cluster mutagenesis approach was not effective since it did not modify all the important amino acids at once. Therefore, we decided to introduce 27- to 45-amino-acid deletions into construct R6 as shown schematically for constructs R20 through R22 . These deletion derivatives bound to RNA poorly (especially protein R22 ---but compare lanes R20 and R21 of Fig. ). Overall, this analysis suggests that an RNA-binding region might be present within the 45-amino-acid-long segment of the central region in p92. Since we had unexpected difficulties in obtaining many other mutants (due to protein stability problems during expression in E. coli; see above), we could not further map the actual amino acids involved in RNA binding for the central segment in p92C using this approach. DISCUSSION : Binding of the replicase proteins to the viral RNA is predicted to be important during many steps of the viral infectious cycle, including the recruitment of the viral RNA to the site of replication, recognition of cis-acting elements during replication, cRNA synthesis, etc. . To gain insights into the replication process of tombusviruses, we have in this paper characterized the ability of p33 and p92 replicase proteins to bind to RNA. Although we found that binding of p33 to ssRNA was the strongest, ssDNA was also bound by p33 with moderate efficiency . In contrast, binding to dsRNA or dsDNA was poor, suggesting that p33 is a single-stranded nucleic acid binding protein. Interestingly, the ability of p33 replicase protein to bind single-stranded nucleic acids is similar to other plant viral proteins, such as movement proteins and coat proteins (, , -, , , , , , , ). In spite of its ability to bind to viral as well as nonviral RNAs, p33 showed preference for the TBSV-derived sequence, which was the best substrate among four different similar-sized ssRNA templates . An artificial AU-rich template was also bound by p33 moderately well. In contrast, a GC-rich RNA and tRNA bound poorly to p33 . Overall, the observed selectivity of p33 in RNA binding may not be enough for p33 (and possibly p92) to bind to only TBSV-related RNAs in infected cells. It is possible that other factors, or a combination of factors, are needed to achieve such levels of selectivity in template use. Based on p33-RNA binding experiments, we propose that that both p33 and p92 (data not shown) bind RNA in a cooperative manner. Results show typical "all or none" behavior characteristic of cooperative binding . The RNA bound by the recombinant p33, which was either fused to MBP or was cleaved off the MBP domain , stayed in the well, probably due to the large size of the complex. In contrast, the p33 or p33 fused to MBP did migrate into the gel in the absence of RNA, as shown in Fig. . In addition, a 60-amino-acid-long truncated p33 (Fig. , protein C10) also bound to RNA in a cooperative manner . Indeed, detection of large complexes between the truncated p33 and the RNA probe in the three-membrane sandwich assay is also consistent with cooperative RNA binding by this truncated p33. The proposed ability of p33 to bind RNA in a cooperative manner suggests that, after the initial binding by the first p33 to the RNA, subsequent binding of additional p33 molecules to the same RNA is greatly facilitated not only by the stabilizing effect coming from binding to the RNA but also by p33-p33 (protein-protein) interactions. This may lead to complete coating of the RNA with p33 molecules. In support of this model, we found p33-p33 interactions in vitro (K. S. Rajendran and P. D. Nagy, unpublished data). Binding to RNA in a cooperative manner can be advantageous for both the viral RNA and the replicase proteins, since this may increase the stability of RNA-protein complexes inside the infected cells. Therefore, it is not surprising that many viral proteins, including 2D of poliovirus, NS5B RdRp of hepatitis C virus (HCV) , viral coat proteins, and plant viral movement proteins (, , -, , , , , , , ), were found to bind viral RNAs in a cooperative manner. The functional significance of cooperative binding by p33 and p92 is currently not known. It is possible that p33 can coat the viral ssRNAs in infected cells, which may be beneficial during template recruitment and/or replication. The p33-coated viral RNAs may be more resistant to nucleases and less accessible to host-mediated gene silencing than are free viral RNAs . Since cooperative binding depends on the amounts of replicase proteins and viral RNAs present in the cells, it is likely that this feature may be important for the function of p33, which is 20-fold more abundant than the p92 replicase protein in the infected cells . The ability to bind cooperatively to RNA, possibly in combination with p33, may also be important for the function of p92 during replication. For example, cooperative binding between p92 and p33 may facilitate recruitment of the less abundant p92 proteins to the viral RNA in infected cells. Alternatively, binding of p92 to the viral RNA templates in a cooperative manner may enhance its RdRp activity, as it has been shown for the poliovirus 2D (RdRp) protein that binds RNA cooperatively . In addition, it has been shown that the poliovirus 2D protein forms large complexes that might be involved in virus replication . Deletion analysis revealed that p33 has one high-affinity RNA-binding site in its C-terminal region. This RNA-binding region includes the RPR motif, which has been shown to be essential for RNA binding (Fig. , compare proteins C11 and C12). Mutagenesis of the RPR motif in p33 has also revealed that this domain is essential for the replication of a tombusvirus and an associated DI RNA in vivo . Mutations within the RPR motif have affected subgenomic RNA synthesis as well . These observations demonstrate that the RPR motif plays a central role in viral RNA synthesis and metabolism. The RPR motif is highly conserved among the replicase proteins of tombusviruses and the related turnip crinkle virus (TCV) . Accordingly, a recombinant TCV p88 protein, which is the RdRp protein similar to p92 of TBSV, contains an RPR motif in its N-terminal portion and has been shown to bind RNA efficiently in vitro . Moreover, the RPR motif in TBSV is similar to the previously proposed ARM motif (i.e., arginine-rich motif), which is present in several viral- and host RNA-binding proteins , including the trans-activator protein (Tat) of human immunodeficiency virus type 1 . The structure of the ARM motif, however, is rather different in several RNA-binding proteins. Therefore, it is not known if the ARM motif functions similarly in different proteins . Protein analysis software predicts that the RPR motif in TBSV p33 constitutes a hydrophilic pocket, and it is exposed to solvent . This structural prediction is compatible with the proposed function of the RPR motif in RNA binding. The secondary structure analysis predicts that the RPR motif itself has mainly turns , suggesting that this motif may take stable conformation upon binding to RNA. Indeed, the ARM motif in the human immunodeficiency virus type 1 Tat protein constitutes flexible turns that can adopt two different conformations upon binding to two different viral RNAs . In addition to its role in p33 functions, the RPR motif is likely functional in p92 as well, since its N-terminal sequence (the so-called "overlapping domain") is identical to p33. We could not test this in vitro, however, since p92 has additional RNA-binding sites ---one in the central part (named the RBR2 region, located in the vicinity of the RdRp signature motifs; Fig. ) and another in the C-terminal segment (named RBR3; Fig. ), which can facilitate binding to RNA in the absence of the RPR motif (Fig. , construct p92C). Nevertheless, separate mutagenesis of the RPR motif in p33 and p92, followed by testing virus replication in the protoplast, revealed that the RPR motif in p92 is essential for tombusvirus replication . Therefore, based on (i) sequence identity between p33 and the overlapping domain of p92 and (ii) the in vivo requirement of the wild-type RPR motif in p92 for tombusvirus replication, we propose that the RPR motif in p92 is a functional RNA-binding site. However, the actual function(s) of the RPR motif in p92 will need further studies. Sequence comparison of the central RBR2 RNA-binding domain within the unique p92C segment of TBSV with other tombusviruses and related viruses revealed that this 45-amino-acid region contains several highly conserved amino acids, including motif F . Interestingly, an RNA-binding region, which includes the motif F, has also been defined for the NS5B RdRp protein of HCV , as shown in Fig. . This region has been shown for HCV RdRp to include some of the basic amino acids that line up the RNA-binding groove and bind to RNA template during viral replication . In addition, the highly conserved arginine and isoleucine or leucine in motif F have been found to bind to ribonucleoside triphosphates (rNTPs) . These observations suggest that this 45-amino-acid-long region in TBSV p92 possesses the amino acids responsible for binding to both rNTPs as well as to the template RNA. Based on its similarity in location to the RNA-binding region in the HCV NS5B, the function of the RBR2 RNA-binding region in p92 of TBSV may be to "channel" the RNA template toward the active site in the RdRp. The significance and possible function(s) of the C-terminal RBR3 RNA-binding region in TBSV p92 is currently under further investigation. In summary, our in vitro studies defined the RNA-binding regions in p33 and p92 replicase proteins of TBSV, which likely play a major role(s) in tombusvirus RNA replication and metabolism. These results will open the way for future studies on the functions of these proteins in particular and on tombusvirus replication in general. FIG. 1. : RNA binding by the recombinant p33 and p92 replicase proteins of TBSV in vitro. RNA binding by the recombinant p33 and p92 replicase proteins of TBSV in vitro. (A) Schematic representation of the expression strategy of the replicase genes in TBSV. The plus-strand genomic RNA is used in the infected cells to produce the replicase proteins p33 and p92 (wavy lines above the boxes represent the individual replicase proteins expressed from the TBSV RNA). p92 is translated via ribosomal readthrough of the stop codon at the end of the p33 ORF. An artificial deletion derivative of p92, termed p92C, was also generated to produce the unique, nonoverlapping portion of p92 protein, which contains the signature motifs of RNA-dependent RNA polymerases. The other genes shown in black boxes are expressed from subgenomic RNAs . (B) SDS-PAGE analysis of purified recombinant TBSV replicase proteins from E. coli. The p33, p92 (the stop codon at the end of p33 was modified to a tyrosine codon to ensure the production of full-length p92; reference ), and the truncated p92C genes were cloned into pMAL-c2X and expressed as C-terminal fusion proteins with the MBP. The fusion proteins were purified using amylose resin affinity chromatography and analyzed in an SDS-10% polyacrylamide gel. Lane MW shows molecular weight markers (in thousands) on the left, while the other lanes contain samples purified from E. coli: lane 2, MBP/lacZ fusion protein; lane 3, MBP/p33; lane 4, MBP/p92C; lane 5, MBP/p92. (C) A gel mobility shift assay showing interactions between the recombinant TBSV replicase proteins and TBSV RNA. The 82-nt 32P-labeled minus-stranded region III RNA was separately incubated with bovine serum albumin and one of the recombinant proteins (1 muM) as shown, in a binding buffer at 25C for 30 min and then analyzed in 4% nondenaturing polyacrylamide gel. The unbound, free RNA probe and the shifted (bound) RNA-protein complexes are marked on the right. (D) Mobility of the recombinant MBP/p33 and the p33 (after cleavage with factor Xa) in the absence of RNA probe in a 1% agarose gel. The electrophoresis was performed under the same conditions as in panels E and F. The proteins were stained with Coomassie brilliant blue. (E and F) Comparison of the RNA-binding abilities of two recombinant p33 preparations, which were either fused with MBP or cleaved off the MBP. The gel mobility shift assays were performed as in panel C, except that increasing amounts of MBP/p33 (0.03, 0.06, 0.13, 0.27, 0.65, 1.3, and 2.6 muM protein per lane) or recombinant p33 (cleaved) (0.03, 0.06, 0.13, 0.25, 0.50, 1.0, and 2.0 muM total protein per lane) were applied. The samples were analyzed using 1% agarose gel electrophoresis run at 100 V in a cold room. Note that the faint band located between the fully shifted (top) and free (bottom) RNA bands (marked with *) in panel F was not consistently detectable when we repeated these experiments. FIG. 2. : SPR analysis of interactions between the TBSV RNA and the recombinant replicase proteins. SPR analysis of interactions between the TBSV RNA and the recombinant replicase proteins. SPR analysis was carried out using BIACORE X (Biacore), as described in Materials and Methods. A 5'-biotinylated 20-nt RNA derived from the 3' end of TBSV RNA (the minimal TBSV genomic promoter ) was immobilized (440 RU) onto a streptavidin-coated sensor chip. The purified recombinant proteins (1 muM) in a binding buffer were passed over the RNA-coated surface of the chip, and the change in mass due to interaction between RNA and protein that altered the refractive index of the medium was recorded in real time in a sensogram. The time allowed for association and dissociation phases was 180 and 200 to 260 s, respectively, in each protein-RNA interaction assay. Interactions between the target RNA and p33, p92C, p92 (all three were tested as MBP fusion proteins), and MBP are shown in panels A, B, C, and D, respectively. The sensogram data were corrected for nonspecific binding based on data obtained using a control surface (shown as the baseline), which was free of RNA. Note that the RU were not normalized; therefore, they should not be compared directly for proteins of different molecular mass. The small positive response in the MBP test (in panel D) is likely due to loose, nonspecific binding, which was subsequently washed out with the buffer ---as is evident from the curve reaching the baseline rapidly during the washing step. FIG. 3. : Preferential binding of the recombinant p33 to ssRNA. Preferential binding of the recombinant p33 to ssRNA. (A) A 1 muM concentration of purified recombinant p33 was incubated with 32P-labeled ssRNA probe, representing the 82-nt minus-stranded region III in the absence or presence of increasing amounts (in 4-, 16-, or 64-fold excess) of unlabeled competitors (as shown at the top of the figure). The experiments were repeated twice. (B) Graphical representation of data obtained in panel A. The extent of competition was quantified as the percentage of displaced labeled RNA probe from the bound complex (indicated on the right in panel A) using a phosphorimager and ImageQuant software (version 1.2). FIG. 4. : Testing binding preference of the recombinant p33 to RNA. Testing binding preference of the recombinant p33 to RNA. (A) Predicted secondary structures of the competitor RNAs, including the 82-nt minus-stranded region III of TBSV [DI-RIII(-)] and the similarly sized artificial AU and GC templates . Note that the double- versus single-stranded regions are comparable in length in these competitor RNAs. (B) Unlabeled competitor RNAs at increasing amounts (in 10-, 100-, or 200-fold excess) were added to the mixture containing the labeled probe (82-nt minus-stranded region III; Fig. ) and 1 muM purified recombinant p33 and the bound complexes were analyzed in gel mobility shift assay. The tRNA was from yeast. (C) Graphical representation of data obtained in panel B. The quantification of the experiment was done as described in the legend to Fig. . The experiments were repeated twice. FIG.5. : Cooperative RNA binding by the full-length and truncated recombinant p33. Cooperative RNA binding by the full-length and truncated recombinant p33. (A) Increasing molar concentrations of p33 were incubated with 32P-labeled probe [region III(-); Fig. ], and the RNA-protein complex was resolved in a gel mobility shift assay. The middle panel shows the semi-log plot of the percentage of RNA bound versus molar concentration of p33 determined by using a phosphorimager. The bottom panel shows a Hill plot of the experimental data obtained from the binding assay (top of panel A). To deduce the Hill coefficient, we used the points that correspond to the middle of the curve in the semi-log graph (as cooperativeness is negligible at the extremes). (B) The gel mobility shift assay was performed with a truncated recombinant p33, termed C10 , which contained only a 60-amino-acid-long region, including the RNA binding site in the recombinant p33. The middle and bottom graphs were prepared as described for panel A above. FIG. 6. : Three-membrane sandwich experiments to demonstrate cooperative RNA binding by a truncated recombinant p33. Three-membrane sandwich experiments to demonstrate cooperative RNA binding by a truncated recombinant p33. The purified recombinant C10 protein (Fig. ; the applied amount is shown on the left side) was incubated with 32P-labeled region III(-) RNA probe for 30 min at 25C, followed by filtering the reaction products through a three-membrane sandwich in a slot blot apparatus. The membranes were then washed twice with (300 mul) binding buffer, air dried, and analyzed using a phosphorimager. The membranes were used in the following order: polysulfone on the top (column 1), nitrocellulose in the middle (column 2), and Hybond N+ on the bottom (column 3). Note that the unbound label at the highest protein concentration likely represents labeled ribonucleotides (which, in spite of a purification step, are still present in the RNA probe as a carryover from the labeling reaction) that were unbound were to protein C10. FIG.7. : Mapping the RNA-binding domain in the recombinant p33. Mapping the RNA-binding domain in the recombinant p33. (A) A schematic representation of the deletion derivatives of p33. The names of the constructs and the positions of the amino acids present in the truncated proteins are shown on the right. These truncated p33 proteins were expressed in E. coli as fusions to MBP (indicated schematically by a dotted box). The shaded boxes indicate the portions of p33 protein that were present in given expression constructs. The horizontal lines represent the deletions. (B) SDS-PAGE analysis of the purified recombinant proteins in a 10% polyacrylamide gel stained with Coomassie brilliant blue. The lane MW refers to molecular mass markers (in kilodaltons). (C) RNA binding activities of the truncated p33 proteins. The labeled RNA probe and the gel mobility shift assay were as described in the legend to Fig. . Equimolar concentrations (2 muM) of proteins were used for the gel shift assay. (D) Northwestern analysis of selected truncated p33 proteins. The purified recombinant proteins (similar2 mug) were run in SDS-10% PAGE as shown in the left panel, transferred to a PVDF membrane, and then probed with a 32P-labeled probe [region III(-); Fig. ]. The positions in the Northwestern blot, which represent a particular recombinant protein, are marked with asterisks. FIG. 8. : Mapping the RNA-binding domains within the unique portion of the p92 protein, termed p92C. Mapping the RNA-binding domains within the unique portion of the p92 protein, termed p92C. (A) Schematic representation of the deletion derivatives of p92C. The names of the constructs and the positions of the amino acids present in the truncated proteins are shown on theright. These truncated p92C proteins were expressed in E. coli as fusions to MBP (indicated schematically by a dotted box). The shaded boxes indicate the portions of the p92C protein that were present in given expression constructs. The horizontal lines represent the deletions. (B) Schematic representation of clustered alanine-serine scanning and deletion mutagenesis of a segment in p92C to map the RNA-binding site. The alanine-serine scanning mutations were targeted at five different groups of basic amino acid clusters as shown. Expression constructs R20 to R22 were made by deleting two or more of the basic amino acid clusters as indicated by the straight lines. (C) RNA-binding activities of the truncated recombinant p92C proteins were analyzed in a standard gel mobility shift assay (refer to the legend of Fig. for details). Equimolar concentrations of proteins (similar2 muM) were used in the binding assay in the presence of a 32P-labeled RNA probe [region III(-); Fig. ]. The samples containing particular recombinant proteins are indicated above the lanes, using the same numbering as in panels A and B above. FIG. 9. : Primary and secondary structure analysis of the RNA-binding region in p33. Primary and secondary structure analysis of the RNA-binding region in p33. (A) SEQWeb version 1.1 from GCG Wisconsin package was used to predict the secondary structure, surface probability, and hydrophilicity of the RNA-binding region in p33. The arginine- and proline-rich motif involved in RNA binding is boxed. (B) Sequence alignment of the arginine- and proline-rich motifs present in the p33-like replicase proteins of different tombusviruses and the related proteins in carmoviruses. The conserved motif is boxed. The following abbreviations were used: CNV, cucumber necrosis virus; AMCV, Artichoke mottled crinkle virus; CIRV, Carnation Italian ringspot virus; CymRSV, Cymbidium ringspot virus; CRSV, Carnation ringspot virus (a dianthovirus); TCV, turnip crinkle virus (a carmovirus); MNeSV, maize necrotic streak virus; and HIV-1, human immunodeficiency virus type 1. (C) Alignment of the arginine- and proline-rich motif in TBSV p33 and arginine-rich motif in HIV-1 Tat, which is involved in RNA binding . The residues important for RNA binding in the Tat protein are underlined, while the arginine-rich sequence is boxed. FIG. 10. : Locations of the three RNA-binding regions in TBSV p92. Locations of the three RNA-binding regions in TBSV p92. (A) Schematic representation of the three mapped RNA-binding regions ---RPR, RBR2, and RBR3 ---in p92. The N-terminal portion of p92, which overlaps with p33, is indicated as a dark box, while the unique C-terminal segment is shown as a light box. The signature motifs of RdRps are marked with letters A through F . The names of the RNA-binding regions are shown above the filled boxes. (B) Sequence alignment of the RBR2 region in TBSV and in other replicase proteins of tombusviruses and related viruses as well as the RNA-binding region in NS5B of HCV. The RBR2 region in TBSV is indicated by a solid line above the sequence, while the RNA-binding sequence in NS5B of HCV is marked with a dotted line beneath the sequence. The conserved motif F is boxed. The consensus (cons) sequence is also shown at the bottom. The abbreviations used for viruses are the same as in the legend to Fig. , except that we also included Pothos latent virus (PoLV; Tombusviridae) and HCV. TABLE 1 : List of primers used for PCR to generate expression constructs TABLE 2 : List of primers and templates used for PCR to generate expression constructs Backmatter: PMID- 12915537 TI - Production of Human Papillomavirus Type 16 Virus-Like Particles in Transgenic Plants AB - Cervical cancer is linked to infection with human papillomaviruses (HPV) and is the third most common cancer among women worldwide. There is a strong demand for the development of an HPV preventive vaccine. Transgenic plants expressing the HPV major capsid protein L1 could be a system to produce virus-like particles for prophylactic vaccination or could even be used as edible vaccines to induce an L1-specific prophylactic immune response. Here, we describe the generation of transgenic tobacco and potato plants carrying the HPV type 16 major structural gene L1 under the control of the cauliflower mosaic virus 35S promoter. All attempts to express either the original, unmodified L1 gene or an L1 gene with a codon usage optimized for expression in plants failed. Surprisingly, small amounts of the protein were detected using an L1 gene optimized for expression in human cells. However, Northern blot analysis revealed that most of the L1 transcripts were degraded. Introduction of the translational enhancer Omega derived from the tobacco mosaic virus strongly increased transcript stability and resulted in accumulation of L1 protein to approximately 0.5 to 0.2% of total soluble protein in transgenic tobacco and potato plants, respectively. The plant-derived L1 protein displayed conformation-specific epitopes and assembled into virus-like particles. Furthermore, we did not find any indications of protein modification of the L1 protein produced in plants. Plant-derived L1 was as immunogenic as L1 expressed in baculovirus-infected insect cells. Feeding of tubers from transgenic potatoes to mice induced an anti-L1 antibody response in 3 out of 24 mice, although this response was only transient in two of the mice. Our data, however, indicate that an anti-L1 response was primed in about half of the 24 animals. Keywords: Introduction : The human papillomaviruses (HPV) comprise a heterogeneous group of more than 80 epitheliotropic genotypes. The clear link between infection with "high-risk" HPV types and the development of malignant diseases was established in numerous epidemiological and molecular studies (for reviews, see references , , and ). Worldwide, HPV are responsible for approximately half a million new cervical cancer cases every year, as HPV DNA of high-risk genotypes can be found in more than 95% of these cancers. About half of these cases are associated with infections with HPV type 16 (HPV-16) . Despite improved abilities to detect and cure premalignant lesions, a high percentage of patients still develops persistent or metastatic disease for which no effective therapy is available. Therefore, there is urgent need to develop a prophylactic subviral vaccine for preventing infection with HPV and thereby, most likely, the development of cervical cancer . The fact that for a long time it was not possible to produce papillomaviruses under cell culture conditions has hindered the development of a capsid-directed vaccine, e.g., inactivated virions. About a decade ago, it was demonstrated that expression of the papillomavirus major capsid gene L1 alone, or together with the minor capsid protein L2, is sufficient for the generation of virus-like particles (VLPs). Subsequently, VLP production was shown for virtually all experimental systems, such as mammalian cells, baculovirus-infected insect cells, yeasts, and Escherichia coli as well as in cell-free systems . VLPs mimic in some aspects the infection with virions and induce virus-neutralizing antibodies. Thus, VLPs became the most attractive candidate for developing a prophylactic vaccine against HPV infections. Papillomavirus VLPs are very efficient in inducing both humoral and cellular immune responses. VLP-induced antibodies provided protective immunity against challenge with infectious viruses in animal models . In addition to the induction of a prophylactic immune response, papillomavirus VLPs bind to and lead to the activation of dendritic cells in vitro . The potential of VLPs as prophylactic vaccines is currently being assessed in a number of different clinical trials. In fact, a recent report demonstrated that administration of HPV-16 VLPs significantly lowers infection with HPV-16 and leads to a reduction HPV-16-associated disease . Cervical cytology screening has resulted in a significant decline in mortality from cervical cancer in highly developed countries, but cervical cancer remains one of the main causes of cancer-related death in most of the less well developed countries. However, costly production and distribution of current VLP vaccines, e.g., by the use of recombinant baculoviruses, will prevent their widespread application in less developed countries. Many biopharmaceuticals are traditionally produced using a variety of transgenic systems including mammalian cells or insect tissue cultures; however, in these expression systems it is usually difficult and costly to control GMP production. Transgenic plants have been evaluated as an alternative expression system for vaccines and other therapeutic proteins and have several economic and qualitative benefits. Expression of vaccines in plants eliminates the risk of contamination with animal pathogens, provides a heat-stable environment, and would, if administered as an edible vaccine, avoid injection-related hazards. Apart from these advantages, plants can be grown on a large scale, and existing cultivation, harvest, and storage facilities can be utilized. Within the last years a number of vaccines have been successfully expressed in plants and orally delivered to animals to determine their immunogenicity, such as the Norwalk virus capsid protein , the hepatitis B surface antigen , and the heat-labile nontoxic subunit B of the E. coli enterotoxin . Recently, it has been shown that oral application of HPV-16 VLPs leads to the induction of capsid-specific antibodies in mice, indicating that oral vaccination against papillomaviruses is a serious option . Here, we describe the development of a plant expression system to produce the HPV-16 L1 structural protein and demonstrate its functionality for vaccination purposes. MATERIALS AND METHODS : Plant material and growth conditions. | Tobacco plants (Nicotiana tabacum cv. Samsun NN) and potato plants (Solanum tuberosum cv. Solara) were grown in tissue culture under regimens of 16 h of light, 8 h of darkness on Murashige Skoog medium containing 2% (wt/vol) sucrose. After initial characterization of in vitro-grown plants, tobacco and potato plants were transferred to the greenhouse. Plants were cultivated in soil with 16 h of supplemental light (ca. 250 mumol quanta photosynthetic active radiation m-2 s-1) followed by 8 h darkness. The temperature regimen followed a day-night cycle with 25 and 20C for tobacco and 22 and 18C for potato plants. Plasmid construction. | To produce L1 transgenic plants, the unmodified and the codon-optimized HPV-16 L1 genes (L1ori, L1p, and L1h) were excised from the cloning vector pBK-CMV (Stratagene) using the SalI/KpnI restriction sites and inserted into a Bin-19-derived binary vector containing the constitutive Cauliflower mosaic virus (CaMV) 35S promoter and the octopine synthase polyadenylation signal. L1p and L1h display a codon composition adapted to the codon usage database shown in table format for S. tuberosum and Homo sapiens in reference , respectively. Both sequences are available from the EMBL database under the accession numbers and . In the OD-L1h construct, the TMV-U1 translational enhancer element (Omega) was fused upstream of the L1 initiation codon . To allow the insertion of the enhancer element, a novel NcoI restriction site was introduced into the L1 gene, overlapping with the ATG initiation codon, changing the second amino acid of L1 into an alanine (S2->A). Plant transformation. | The binary constructs were transformed into Agrobacterium tumefaciens strain CV58C1, which carries the virulence plasmid pGV2260 . Transformation of tobacco and potato plants using Agrobacterium-mediated gene transfer was carried out according to Rosahl et al. and Rocha-Sosa et al., respectively . RNA isolation and Northern blot analysis. | Isolation of total RNA was performed as described by Logemann et al. . Aliquots (20 to 30 mug) of total RNA were separated on 1.5% formaldehyde-containing agarose gels and blotted onto nylon membranes (GeneScreen; NEN, Boston, Mass.) by capillary blotting overnight. The membranes were prehybridized and hybridized at 65C. Radioactive labeling of respective cDNA fragments was performed with [alpha-32P]dCTP using the High Prime kit (Boehringer Mannheim, Germany). After stringent washing, radioactive membranes were exposed to X-ray films (Kodak) overnight at -70C. Western blot analysis. | Leaf disks (0.28 cm2) or tuber slices (ca. 50 to 100 mg, fresh weight) were homogenized in extraction buffer containing 50 mM Tris-HCl (pH 6.8), 5 mM MgCl2, 1 mM EDTA, 1 mM EGTA, 5 mM beta-mercaptoethanol, 0.2 mM Pefabloc proteinase inhibitor, and 15% glycerol. Samples were centrifuged for 5 min at 4C, and protein content of the supernatant was determined according to the method in reference . Sodium dodecyl sulfate (SDS)-containing sample buffer was added, and after heat-denaturation 10 to 30 mug protein was separated on SDS-12.5% (vol/vol) polyacrylamide gels. Proteins were transferred onto nitrocellulose membranes (Hybond C; Amersham Pharmacia Biotech, Braunschweig, Germany), blocked for 1 h in 5% skim milk-TBST (20 mM Tris, 500 mM NaCl, 0.1% [vol/vol] Tween 20) and incubated for a minimum of 2 h with rabbit antiserum raised against insect cell-derived VLPs . Immunodetection was performed using the ECL system (Amersham Pharmacia Biotech, Braunschweig, Germany) according to the manufacturer's instruction. Enzyme-linked immunosorbent assay (ELISA). | For the detection of HPV-16 L1-specific antibodies, microtiter plates were coated overnight with 50 mul of phosphate-buffered saline (PBS) containing 35 mug of VLPs derived from insect cells per ml. After blocking of the plates (in 5% skim milk in PBS for 1 h at 37C), 1:100 to 1:12,800 dilutions of mouse sera were added and incubated for 1 h at 37C. To determine nonspecific binding, the same dilutions of the antisera were tested on plates coated with PBS only. After washing, peroxidase-conjugated goat anti-mouse antibodies (Sigma) were added at a 1:4,000 dilution. After 1 h at 37C plates were washed and stained with ABTS (2,2'-azino-bis-3-ethylbenzthiazolin-6-sulfonic acid) substrate solution (1 mg/ml, containing 0.015% H2O2). Extinction at 405 nm was measured after 20 min in a Titertek automated plate reader. To allow relative quantification of VLPs in CsCl fractions, we used an antigen capture ELISA. Microtiter plates were coated overnight with a 1:500 dilution of protein A-purified mouse monoclonal antibody (50 mul/well) specific for HPV-16 VLPs (final concentration of 2 mug per ml, in PBS) . Plates were blocked with 5% skim milk-PBS for 1 h. After adding 50 mul of CsCl fractions using 1:5 to 1:300 dilution (in 5% skim milk-PBS) plates were kept for 1 h at 37C. After three washings with PBS-0.05% Tween 20, 50 mul of a polyclonal rabbit antiserum (1:3,000 dilution in PBS) raised against HPV-16 VLPs was added, and plates were incubated at 37C for 1 h. Plates were washed again and further incubated with 50 mul of a goat-anti-rabbit peroxidase conjugate (1:5,000 in milk-PBS; Sigma). After a final washing step, ABTS substrate was added and allowed to develop for 30 min. Extinction was measured at 405 nm in a Dynatech automated plate reader. Purification of VLPs. | TN-High Five cells were grown to a density of 1 x 106 to 2 x 106 cells per ml of Ex-Cell 405 serum-free medium (JRH Biosciences). About 2 x 108 cells were pelleted at 1,000 x g for 15 min, resuspended in 20 ml of medium, and infected with recombinant baculoviruses at an MOI of 2 to 5 for 1 h at room temperature. Cells were further incubated in shaker bottles for 3 days and then harvested by centrifugation. Cell pellets were resuspended in extraction buffer. All following steps were performed at 4C. After sonication for 1 min at 60 W, the extract was centrifuged at 10,000 rpm in a Sorvall SS34 rotor. The pellet was resuspended in 10 ml of extraction buffer, sonicated again, and centrifuged. The combined supernatants were layered onto a two-step gradient consisting of 8 ml of 40% sucrose on top of 8 ml of CsCl solution and centrifuged in a Sorvall AH629 swinging bucket rotor for 2 h at 27,000 rpm (10C). The interphase between CsCl and sucrose and the complete layer of CsCl were collected and transferred to 13.4-ml Quickseal tubes (Beckman). Volume was adjusted by adding extraction buffer, and tubes were centrifuged overnight at 50,000 rpm at 20C in a Beckman 70 Ti rotor. One-milliliter fractions were collected, and subsequently 2.5 mul of each fraction was separated on an SDS-10% polyacrylamide gel and L1 protein was detected by Western blotting. VLPs from plant cells were extracted using a modified protocol. Leaf material was frozen in liquid nitrogen and pulverized using a mortar and pestle; potato tubers were sliced and mechanically homogenized. Five grams of plant material (in 20 ml of buffer additionally containing 1% polyvinyl pyrrolidone to adsorb the phenolic compounds) was then subjected to further extraction using a French press. Extracts were cleared by centrifugation. L1-containing supernatant was sedimented through a sucrose cushion and further purified by CsCl density centrifugation. Analysis of L1 assembly by sucrose sedimentation. | To separate different assembly forms of L1, crude extracts from tobacco leaves and potato tubers were loaded onto linear (5 to 50%, wt/vol) sucrose gradients. Plant material (2.8 g) was sliced and homogenized in liquid nitrogen using a mortar and a pestle. The material was resuspended in 1 to 2 ml of ice-cold extraction buffer and sonicated four times for 30 s at a low setting. The extract was cleared by centrifugation at 15,000 x g for 15 min. Six hundred microliters, representing approximately one-eighth of the total extract, was loaded on top of a 5-to-50% linear sucrose gradient and centrifuged at 160,000 x g for 3 h in a Beckman SW41Ti rotor. Twenty fractions were collected and analyzed by antigen-capture ELISA and Western blotting for the presence of L1 protein. Electron microscopy. | Twenty-microliter samples of the fractions collected from the CsCl gradient were dialyzed against 10 mM HEPES (pH 7.5) for 45 min on floating filter pads (0.02-mum pore size; Millipore). Carbon-coated copper grids (200-mesh size; EM Sciences) were treated with 20 mul of poly-l-lysine (1 mg/ml; Sigma) for 2 min. The samples were placed onto the grid and stained with 30 mul of a 2% uranyl acetate solution for 2 min. After air-drying of the grids, specimens were inspected with a Zeiss EM 900 electron microscope at 80 kV . Immunization of mice. | BALB/c mice were immunized subcutaneously (s.c.) with 40 ng of CsCl-purified VLPs derived either from plants or from insect cells. The first immunization was performed using complete Freund's adjuvants. After 4 and 8 weeks, respectively, two booster immunizations were carried out using incomplete Freund's adjuvants. Two weeks after the third immunization mice were sacrificed and serum was collected. For oral immunization, the five different groups of mice were fed four times with 5 g of sliced, raw potato tubers per mouse on days 1, 14, 32, and 46. Prior to each feeding, the mice were starved for 24 h. Each feeding lasted 20 h, and the weight of uneaten tubers was determined thereafter (ranging between 0 and 12%). Blood samples were collected on days 0, 12, 30, 44, and 58. On day 79 all mice received a subimmunogenic dose of insect cell-derived VLPs s.c. without adjuvants. Final bleeds were collected on day 79. In one experimental and one control group the potato tubers in each of the feedings were spiked with 10 mug of CpG DNA per 5 g of tubers in the from of plasmid DNA (with no HPV-specific sequences), in one additional group the transgenic potatoes were spiked with 10 mug of cholera toxin B peptide (Sigma) per 5 g of tubers. RESULTS : Optimization of HPV-16 L1 for expression in plants. | Our aim was to evaluate transgenic plants as a tool to express the major structural protein L1 of HPV-16 in the form of VLPs suitable for prophylactic vaccination purposes. To this end, transgenic plants were engineered using the Agrobacterium-mediated transformation system followed by selection of transgenic plants for kanamycin-resistance. Initially, transgenic potato plants were generated expressing the unmodified HPV-16 L1 gene (L1ori) under the control of the CaMV 35S promoter. However, the L1 protein was not detectable in 10 to 30 mug of plant protein extracts in Western blots using an L1-specific antibody. This indicates that the expression level was below the detection limit, which is about 5 ng of L1-protein (purified from insect cells). In order to achieve L1 expression in plants, codon composition of the L1 gene was adapted according to codon usage tabulated in reference for S. tuberosum (16 L1p; EMBL database AJ313181). Earlier, we reported that this gene was efficiently expressed in mammalian cells via transient-transfection assays . A total of 55 different transgenic tobacco plants were obtained, in which a CaMV 35S promoter was driving the expression of 16 L1p. Accumulation of L1p in these lines was analyzed by Western blotting. However, as before no L1-specific signal was detected in extracts made from these transgenic lines, whereas the specific 55-kDa band appeared for the L1 protein purified from insect cell which was loaded as a control (data not shown). To confirm the presence of L1p-specific transcripts, RNA of the transgenic lines was isolated and analyzed by Northern blotting. No distinct L1p-specific signal was obtained for any of the transgenic lines investigated. However, in some cases putative degradation products of the L1 messenger RNAs became apparent. Hybridizing the same membrane with a probe for the small subunit of Rubisco (rbcS) revealed similar transcript abundance in each lane (data not shown). These data indicate that L1p mRNA might be unstable in the transgenic plants. Recently, our laboratory described a humanized HPV-16 L1 gene (L1h; EMBL database AJ313179) that produces at least 100- and 10,000-fold the amount of L1 protein in mammalian cells when compared to the L1p and the L1ori gene, respectively . Since L1 expression was not achieved in transgenic plants using either the L1ori or the L1p construct, the L1h gene was introduced into tobacco plants . A total of 75 transgenic tobacco plants were regenerated and screened for L1 protein accumulation by immunodetection. The L1 protein was recognized by the antibody in some transgenic lines as shown in Fig. , although an L1-specific signal was observed only after extended exposure leading to a high background and the amount of protein was less than 0.01% of total soluble protein. Northern blot analysis performed with selected lines showed that the bulk of L1-transcript, although clearly detectable, was degraded in the L1h-transgenic tobacco plants . Integrity of the isolated mRNA was again confirmed by hybridization of the membrane with the rbcS probe . One reason for the low levels of L1 protein obtained in L1h-transgenic plants could be that the L1 transcripts had been poorly recognized by the cellular translation machinery, and thus the transcripts were degraded. To improve protein translation of the L1h transcripts, the translational enhancer Omega of the tobacco mosaic virus (U1) was cloned upstream of the L1 initiation codon (OD-L1h) . Due to the cloning strategy employed, the second amino acid of L1 was altered from a serine into an alanine. It is unlikely that this mutation influences the ability of L1 to assemble into capsomeres or VLPs since a number of different papillomavirus types (e.g., HPV-1, -18, -30, and -45 and bovine papillomavirus types 1 and 2) do contain an alanine at position 2 of their L1 sequence. Using the OD-L1h expression construct, 88 transgenic tobacco lines were obtained. In addition, about 80 transgenic lines harboring the L1p gene fused to the TMV translational enhancer were regenerated (OD-L1p). Transgenic lines were again screened by Western blotting. While no L1 protein was detectable in the OD-L1p lines (data not shown), significant amounts of L1 protein accumulated in several of the OD-L1h expressing plants. Ten highly expressing lines were selected and transferred to the greenhouse. These tobacco lines did not show any phenotypic alterations. As shown in Fig. , L1h mRNA is highly abundant in these lines. Apparently, protein accumulation was also improved by introduction of TMV overdrive Omega sequence as indicated by strong signals in Western blots . The percentage of L1 protein as fraction of total soluble protein was estimated to be approximately 0.5%. This calculation is based on comparison with the amount of insect cell-derived purified L1 protein (80 ng) which was loaded on the gel as a control . In order to demonstrate that expression of the optimized OD-L1h construct is not restricted to tobacco plants and to produce L1 in a plant species more suitable for feeding experiments, the construct was also introduced into potato plants. Immunodetection revealed the presence of the L1 protein in potato tubers in 15 out of 48 transgenic lines in similar quantities as in leaves of the transgenic tobacco plants. We estimated that L1 accounts for about 0.2% of total soluble tuber protein . In general, the size of the L1 protein extracted from either tobacco or potato was indistinguishable from that of the L1 purified from insect cells, indicating that no significant modification such as glycosylation had occurred. Purification of HPV-16 L1 VLPs from transgenic plants. | After L1 expression was observed in transgenic plants, we wished to determine whether the L1 protein assembles into viral pentameres (capsomeres) and forms VLPs. Both states of virion assembly can be distinguished from the L1 monomer by the presence of conformation-specific epitopes, their sedimentation coefficient in sucrose gradients, and their buoyant density. Plant tissues were extracted using a modified protocol which was originally established for the purification of HPV VLPs expressed in insect cells . Cleared plant extracts were sedimented through a sucrose cushion followed by a CsCl equilibrium density centrifugation as described in Materials and Methods. The L1 protein was detected using an antigen-capture ELISA which is specific for conformational epitopes presented on the surface of viral capsomeres and capsids. Most of the L1-antigen concentrated at a density of 1.32 g/ml in the CsCl gradients . These results show that conformation-specific epitopes were displayed. Therefore plant-derived L1 apparently assembled into viral capsomeres and most likely formed VLPs. Peak fractions of CsCl gradients were analyzed by electron microscopy to determine whether VLPs were indeed formed in OD-L1h potato plants. Thereby, we were able to visualize VLPs in fractions with a buoyant density of 1.32 g/ml . To determine the amount of L1 protein that is in an assembled form, crude extracts of tobacco leaves and potato tubers were fractionated by sucrose density gradient centrifugation. This analysis allows quantitative separation of L1 monomers, pentamers, and higher assembly forms such as VLPs. The extracts were loaded onto linear sucrose gradients (5 to 50%) and centrifuged at 160,000 x g for 3.5 h where after 20 fractions were collected (fraction 1 bottom of the tube) and L1 protein was detected by ELISA. Under these conditions, insect cell-derived VLPs are found in fractions 3 to 6 (approximately 120S) . Treatment of VLPs leads to disassembly into capsomeres with a lower sedimentation coefficient, which are concentrated in fractions 16 to 18 (data not shown). No signal was obtained when extracts of tubers from nontransgenic plants were analyzed . When extracts of tobacco leaves were analyzed, two peaks of the L1 protein were observed , a smaller peak in fractions 16 and 17, consistent with the sedimentation pattern of capsomeres and a larger peak in the fractions 9 and 10. Using empty VLPs of adeno-associated viruses as sedimentation markers (60S) we estimated the sedimentation coefficient of the L1 assembly forms present in fractions 9 and 10 to be approximately 60 to 70S. When extracts of L1-expressing potato tubers were analyzed, we again found a major peak of L1 at 70S. Extracts of some transgenic lines gave rise to the capsomere peak while extracts from other lines did not and it remains to be determined, whether this observation is specific for the respective transgenic lines. Only small amounts of L1 protein were detected in fractions corresponding to intact VLPs. From the amount of insect cell-derived VLPs loaded onto the gradient, we projected that the peak fractions represent approximately 3 to 5 mug of L1 protein from approximately 0,35 g of leave or tuber material. Immunogenicity of plant-derived HPV-16 L1 VLPs. | To analyze whether the plant that produced L1 is able to induce a humoral immune response, groups of six mice were either immunized with a low dose of VLPs purified from plants (40 ng of VLPs per mouse) or with the same amount of insect cell-derived VLPs. Generation of capsid-specific anti-L1 antibodies was measured by ELISA using insect cell-derived VLPs as antigen. Results show that VLPs formed in plants were as immunogenic as VLPs produced in insect cells . While sera of all mice immunized with plant-derived VLPs responded in the ELISA, only five out of the six mice immunized with insect cell-derived VLPs gave a positive signal . No differences in titers were observed comparing the two groups of mice. Immunogenicity of transgenic plants after oral ingestion. | To determine whether L1 transgenic plants can be used as edible vaccine, L1-positive potato tubers were fed to C57/BL6 mice. Thirty mice falling into five different groups were fed four times for 20 h at an interval of approximately 15 days with tubers from transgenic and nontransgenic potatoes . Groups 1 to 3 (mice 1 to 8, 9 to 17, and 18 to 24) received L1 tubers from transgenic plants; groups 4 and 5 (mice 25 to 27 and 28 to 30) received tubers from nontransgenic plants. In three groups, cholera toxin B-peptide (group 2) and CpG plasmid DNA (groups 3 and 5) were added onto the tubers as adjuvants in all the feedings. Sera from mice were collected before the first and 2 to 3 weeks after each feeding (see Materials and Methods) and tested in an ELISA for the presence of L1-specific antibodies . Results indicated that all sera from mice of the control groups (groups 4 and 5) remained negative throughout the feeding. After the third feeding, three mice of the groups (one in group 2 and two mice in group 3) fed with transgenic material showed an L1-specific response. After the fourth feeding, this response was no longer detectable in two of the three mice while it was increased in one mouse of group 3. Sera of four mice of group 1 also showed an elevated reaction in the ELISA, but this was rather weak and not significantly different from the control group. Since only weak and mostly transiently anti-L1 antibody responses were detected after oral vaccination in all but three sera, we wished to determine whether there was a priming of a humoral response by the feeding of L1-containing tubers. Therefore, all remaining 27 mice were injected with a subimmunogenic dose of purified VLPs (20 ng). In a separate experiment, we determined that a minimum of approximately 800 ng of purified VLPs in a single injection was required to mount a significant anti-L1 response in mice when injections were administered s.c. Two weeks after receiving the subimmunogenic dose, all mice were sacrificed and sera were tested for the presence of L1-specific antibodies by ELISA . About half of the mice (11 of 21) that had been fed with tubers from transgenic plants but none of the control mice developed L1-specific antibodies after receiving the subimmunogenic dose, indicating that oral ingestion of L1-positive tubers had primed a humoral response against the L1 protein. DISCUSSION : We developed transgenic tobacco and potato plants expressing the HPV-16 major structural protein L1 by changing the L1 gene codon usage and increasing its transcript stability. Infections with HPV are causatively related to the development of cervical cancer. For this reason, efforts are under way to develop prophylactic vaccines that prevent primary infection with the virus. Prophylactic vaccination against HPV could prevent the development of almost half a million of cancer cases every year . Prime candidate for the production of prophylactic vaccine is the papillomavirus major structural protein L1 in the form of VLPs or viral capsomeres (pentamers) . Both assembly forms have been demonstrated to induce long lasting protective immune responses in animal models . Currently, VLP-based vaccines are under clinical evaluation, and a licensed product can be expected in the coming years. High costs for production and distribution, however, will not immediately allow widespread application of a papillomavirus vaccine that is based on recombinant VLPs. One of the alternatives might be the production of VLPs or capsomeres using L1-transgenic plants. These plants could be either used as bioreactors for VLP production or as edible vaccines. Two reasons allow one to argue that production of an edible plant-derived papillomavirus vaccine is feasible. First, a number of different antigens was already produced in plants and their oral delivery led to the induction of a protective immune response (summarized in reference ). Secondly, it has been demonstrated that papillomavirus capsid antigens are immunogenic when delivered orally , most likely because they are recognized by the immune system due to their virus-like structure. While the majority of antigens in the intestine are ignored by the immune system, M cells located in the Peyer's patches selectively take up particulate antigens such as virus particles and transport them from the intestine to underlying follicles. The activation of a mucosal immune response leads to the production of secretory immunoglobulin A (IgA) antibodies but also to serum IgG antibodies. A putative prerequisite for the induction of a mucosal immune response against viral capsid antigens by edible vaccines, however, is that the proteins are presented to the M cells in a particulate state. This requires that the capsid proteins be able to assemble into VLPs or capsomeres within the transgenic plants and that upon ingestion and passage through the low pH of the stomach, these particles remain intact. In general, it can be expected that the sturdy plant cell walls provide a protective environment for the expressed antigens. In addition, it has been recently demonstrated that the papillomavirus capsid antigens remain immunogenic after exposure to low pH through oral delivery . To demonstrate that L1 assembles into VLPs in the transgenic plants generated in this study, VLPs were purified and enriched by sedimentation and equilibrium centrifugation. The plant-derived L1 antigen accumulated at a density of 1.32 g/ml, which is consistent with the presence of VLPs or capsomeres. Electron microscopy confirmed the presence of VLPs in the CsCl fractions, although the yield was relatively low. Sucrose sedimentation analysis of soluble proteins derived from crude extracts of leaves and tubers revealed that a significant portion of the L1 antigen is in the form of viral capsomeres but also of higher-order structures with a lower sedimentation coefficient compared to intact VLPs and which are not easily detectable by electron microscopy. The reason for preferential formation of these structures compared to VLPs needs to be further investigated. From our observation in a mammalian expression system, we believe that relatively high concentrations of L1 protein are required for formation of VLP structures. Under such conditions, the same L1 open reading frame used to generate the transgenic plants leads to the formation of capsomeres, intact VLPs as well as smaller quantities of the 60 to 70S assembly forms, which might represent assembly intermediates or assembly by-products (unpublished observations). Subcellular localization analysis studies in a transient viral plant expression system suggest that the L1 protein fails to enter the nucleus but it remains to be investigated whether this is also the case for the transgenic lines. This would be consistent with our observation that in mammalian cells the 60 to 70S intermediates are predominantly localized within the cytoplasm (unpublished observations). Proof of concept for the presence of L1-specific conformational epitopes, however, was provided by immunization of mice with low doses of plant-derived VLPs revealing that these were immunogenic. Feeding L1-expressing tubers to mice led to the induction of a weak but detectable immune response in 3 out of 24 mice; in 2 of these mice the response was only transient. Boosting of immunity in the mice with a subimmunogenic dose resulted in measurable anti-L1 titers in almost half of the animals. These data indicate that oral ingestion of transgenic plant material can, in principle, induce a capsid-directed and possible protective immune response. Because of the low response rate in the experimental groups there is no indication if the oral adjuvants (CpG DNA or cholera toxin) we used in two of the experimental groups are able to improve the immune response. One of the major limitations of using plants as an expression system for oral vaccines and other therapeutic proteins are low yields which might be not sufficient to confer protection (reviewed in references and ). Therefore, to achieve sufficient protein levels in transgenic plants optimization of the respective expression system is necessary. After the initial attempt to express the original L1 gene in plants failed, codon usage of the L1 gene was changed according to that of plant cells. However, this did not also lead to L1 expression in transgenic plants. Since the L1 transcript could not be detected in Northern blots, low transcript stability was assumed to be the main reason for the failure to achieve L1 expression in plants. Interestingly, the expression rate of the plant-adapted L1 sequence was increased by a factor of 100 when expressed in mammalian cells , indicating that the high turnover of the L1p mRNA was apparently a plant-specific effect. Surprisingly, an L1 gene optimized for expression in mammalian cells (L1h) led to accumulation of L1 protein in transgenic tobacco plants, although most of the respective mRNA was found to be degraded. The successful expression of L1h in plants was unexpected because this gene carries codons rarely used in plant cells and has a high GC content, which is very untypical for plant genes. Inefficient translation of the L1-specific transcript might be responsible for subsequent RNA degradation. This is supported by previous reports showing that expression of L1 underlies a tight posttranscriptional control, including low mRNA stability, nuclear transport, and poor translational efficiency probably due to rare codon usage . Presumbly, both, L1 transcript stability and recognition by the translational machinery was strongly improved after insertion of a 5'-leader sequence of TMV, referred to as Omega (OD-L1h), and resulted in accumulation of L1 up to 0.5 and 0.2% of total soluble protein in transgenic tobacco and potato plants, respectively. The Omega sequence was described earlier to increase efficiency of translation in eukaryotic and prokaryotic cells ; however, an impact on mRNA half-life was not observed . Various viral 5' untranslated regions were tested in other plant expression systems to enhance expression levels. While introduction of leader sequences had no effect on expression levels of hepatitis B virus surface antigen in potato tubers , the translational enhancer of the tobacco etch virus allowed significant higher accumulation of Norwalk virus capsid protein in transgenic tobacco plants . Expression levels of Norwalk virus capsid protein could be increased to 0.37% of total soluble protein in potato tubers, which is comparable to the amount of L1 protein accumulated in our plants. For the Norwalk virus capsid protein this amount proved to be too low for large-scale oral administration . Generally, levels of heterologous proteins produced in plants have mostly been less than 1% of total soluble protein (reviewed in references , , and ). One approach to achieve higher yields is targeting of the proteins to the apoplastic space or the endoplasmic reticulum. Another promising tool to produce larger amounts of proteins is to make use of transient expression systems such as viral vectors or infiltration of plants with Agrobacterium. Very recently, Franconi et al. described the expression of HPV16 E7 protein in Nicotiana benthamiana using a potato virus X-derived vector . Recently, we tested transient expression of the OD-L1h construct by infiltration of tobacco plants with Agrobacterium, and this resulted in >=10-fold-higher accumulation of L1 protein (unpublished results). Highest levels of protein accumulation have been achieved in transplastomic plants. In contrast to the conventional nuclear transformation which is widely used to engineer transgenic plants, plastids can be transformed allowing the production of proteins within this subcellular compartment. Chloroplasts have a highly polyploid genome and offer an ideal compartment for protein overexpression . Exceptionally high protein production (47% of total soluble protein) has recently been reported for the BT toxin , and expression of vaccines in chloroplasts might make plant-based production commercially feasible. Here, we demonstrated as a first step that it is possible to produce HPV-16 L1-transgenic plants expressing the HPV-16 L1 protein in a form appropriate for immunization purposes. Although neither tobacco plants nor potato plants are ideal plants for oral application, they are widely employed for research and proof-of-concept studies, since they are easy to transform and to work with. However, recombinant cholera toxin B produced in potato tubers was stable upon cooking and preserved its biological activity . Another advantage of using potato tubers is that they are biologically active during storage, and expression of the transgene can be induced postharvest and/or in nonsprouting tubers, which provides a safe production system and prevents environmental exposure. Our observations from feeding transgenic tubers to mice indicate that the oral immunogenicity, although detectable, clearly needs further improvement; such improvement will be achieved by further enhancement of transgene expression. FIG. 1. : Expression analysis of L1h-expressing tobacco plants. Expression analysis of L1h-expressing tobacco plants. (a) Total RNA was isolated from leaves of tissue-cultured plantlets of 12 different transgenic lines (lanes 1 to 12) and wild-type (wt) controls (lanes 13 and 14). Fifteen micrograms of total RNA was loaded per lane and hybridized with L1h and Rbcs cDNA, respectively. (b) Immunodetection of L1 protein accumulation. Protein extracts were prepared from leaves of tissue-cultured plants. Eleven transgenic lines (lanes 2 to 11) and wild-type plants (lane 1 and 14) were analyzed by Western blotting using a rabbit-derived polyclonal anti-L1 antiserum (dilution 1:5,000) and a goat-derived anti-rabbit IgG antibody conjugated to horseradish peroxidase (1:20,000). Equal amounts of protein were loaded to each lane. L1h protein migrated at approximately 55 kDa. (c) Expression-construct L1h used to generate the lines shown in (a) and (b). L1h is driven by the CaMV 35S promoter. Transcription is terminated at the octopine synthase polyadenylation signal (OCS). FIG. 2. : Expression analysis of OD-L1h-transgenic tobacco plants. Expression analysis of OD-L1h-transgenic tobacco plants. (a) Total RNA was isolated from greenhouse-grown source leaves of 10 different transgenic lines (lines 7, 35, 41, 44, 46, 50, 65, 71, 84, and 86) and wild-type (wt) plants (lanes 11 and 12). Thirty micrograms of RNA was loaded per lane and hybridized with L1-h and Rbcs cDNA, respectively. (b) Immunoblot analysis of L1 protein accumulation in OD-L1h expressing tobacco plants. Protein extracts were prepared from mature source leaves of 10 different transgenic lines (7, 35, 41, 44, 46, 50, 65, 71, 84, and 86) and wild-type plants (lane 7 and 13). Equal amounts of leaf protein (25 mug) were loaded onto each lane. Western blotting was performed using a rabbit-raised polyclonal anti-L1 antiserum (dilution 1:5,000) and a goat-raised anti-rabbit IgG antibody conjugated to horseradish peroxidase (1:20,000). As a control and in order to assess percentage of L1 accumulation, 80 ng of L1 protein purified from insect cells was loaded (lane 1). L1h protein migrated at approximately 55 kDa. Protein migrating at ca. 65 kDa is assumed to be a cross-reacting plant protein. (c) Expression-construct OD-L1h used to generate the lines shown in panels a and b. FIG. 3. : Accumulation of L1 protein in potato tubers of OD-L1h-expressing potato plants. Accumulation of L1 protein in potato tubers of OD-L1h-expressing potato plants. Protein extracts were prepared from freshly harvested potato tubers of eight selected transgenic lines (lanes 2 to 8, corresponding to lines 7, 14, 27, 28, 30, 33, 34, and 37) and a wild-type (wt) control (lane 10). Thirty micrograms of soluble tuber protein were loaded onto each lane. As a control, 40 ng of L1 protein purified from insect cells were loaded (lane 1). Western blotting was performed as described in the legend to Fig. . FIG. 4. : Purification of VLPs from transgenic potato plants. Purification of VLPs from transgenic potato plants. (a) Detection of L1 antigen by antigen-capture ELISA in fractions of CsCl gradient. The arrow indicates the fraction in which VLPs were detected. (b) Electron microscopy of potato plant-derived VLPs. Fractions of CsCl gradients with a density of approximately 1.32 g/ml were analyzed by negative staining. Bar = 200 nm. (c) Sucrose sedimentation analysis of L1 derived from transgenic tobacco leaves. Soluble proteins from tobacco leaf (upper right) and potato tuber (lower left and right) extracts were fractionated by sucrose density gradient centrifugation (fraction 1 corresponds to tube bottom). As control, purified VLPs from insect cells were loaded onto the gradients (upper left). As a 60 S sedimentation marker, empty VLPs of adeno-associated virus were used . As negative controls, extracts from tubers of nontransgenic plants were separated (upper left, open squares). Closed circles indicate the refractive index of the respective fraction. FIG. 5. : VLPs derived from transgenic plants are immunogenic. VLPs derived from transgenic plants are immunogenic. A total of six mice were immunized twice at a 4-week intervals with 40 ng of VLPs purified from plants (top) or from baculovirus-infected insect cells (bottom), respectively. Sera from the immunized mice were then tested in an ELISA against insect cell-derived VLPs. FIG. 6. : Oral vaccination of mice with transgenic potato tubers. Oral vaccination of mice with transgenic potato tubers. Five groups of mice were fed four times within 46 days with 5 g of L1-transgenic (groups 1 to 3, mice 1 to 8, 9 to 16, 17 to 24, respectively) potato tubers. As control (groups 4 and 5, mice 25 to 27 and 28 to 30), six mice were fed with nontransgenic tubers. Groups 2 and 5 received tubers spiked with CpG plasmid DNA; mice of group 2 received the tubers with cholera toxin B-peptide as adjuvants. Serum samples were collected before the first and after each of the four feedings and tested in an ELISA for the presence of L1-specific IgG and IgM antibodies. Bars represent standard deviation from duplicate wells. (Note that serum of mouse 3 of group 1 was no longer available after the third feeding). FIG. 7. : Priming of anti-L1 antibody responses in mice after oral uptake of transgenic plant material. Priming of anti-L1 antibody responses in mice after oral uptake of transgenic plant material. To detect priming of a humoral immune response by ingestion of L1-transgenic potato tubers, all mice received a boost with a subimmunogenic dose of purified VLPs injected s.c. Sera were collected 21 days after the boost and tested in an ELISA for the presence of anti-L1-specific antibodies (IgG and IgM). A cutoff line was arbitrarily set according to the reactivity in the two control groups, groups 4 and 5 (mice 25 to 30). Sera of 12 of the remaining 21 mice from groups 1 to 3 that were fed with L1-transgenic tubers showed a reaction in the ELISA above this cutoff. Bars represent standard deviation from duplicate wells. (Note that sera from mice 1, 2, and 12 were not available for the assay.) Backmatter: PMID- 12915563 TI - Monkey Rotavirus Binding to alpha2beta1 Integrin Requires the alpha2 I Domain and Is Facilitated by the Homologous beta1 Subunit AB - Rotaviruses utilize integrins during virus-cell interactions that lead to infection. Cell binding and infection by simian rotavirus SA11 were inhibited by antibodies (Abs) to the inserted (I) domain of the alpha2 integrin subunit. To determine directly which integrins or other proteins bind rotaviruses, cell surface proteins precipitated by rotaviruses were compared with those precipitated by anti-alpha2beta1 Abs. Two proteins precipitated by SA11 and rhesus rotavirus RRV from MA104 and Caco-2 cells migrated indistinguishably from alpha2beta1 integrin, and SA11 precipitated beta1 from alpha2beta1-transfected CHO cells. These viruses specifically precipitated two MA104 cell proteins only, but an additional 160- to 165-kDa protein was precipitated by SA11 from Caco-2 cells. The role of the alpha2 I domain in rotavirus binding, infection, and growth was examined using CHO cell lines expressing wild-type or mutated human alpha2 or alpha2beta1. Infectious SA11 and RRV, but not human rotavirus Wa, specifically bound CHO cell-expressed human alpha2beta1 and, to a lesser extent, human alpha2 combined with hamster beta1. Binding was inhibited by anti-alpha2 I domain monoclonal Abs (MAbs), but not by non-I domain MAbs to alpha2, and required the presence of the alpha2 I domain. Amino acid residues 151, 221, and 254 in the metal ion-dependent adhesion site of the alpha2 I domain that are necessary for type I collagen binding to alpha2beta1 were not essential for rotavirus binding. Rotavirus-alpha2beta1 binding led to increased virus infection and RRV growth. SA11 and RRV require the alpha2 I domain for binding to alpha2beta1, and their binding to this integrin is distinguishable from that of collagen. Keywords: Introduction : Virus attachment and entry into host cells are multistep processes that influence cellular tropism and can involve sequential recognition of multiple receptors and coreceptors. Rotaviruses, a genus within the Reoviridae family, cause severe gastroenteritis following infection of intestinal enterocytes. The virus spike protein, VP4, which is a major determinant of tropism and receptor binding , is proteolytically cleaved by trypsin into VP5* and VP8*, which increases the virus infectivity and internalization rate . Several glycoconjugates have been implicated in rotavirus attachment . Although a minority of animal rotaviruses, including simian strains SA11 and RRV, can utilize terminal sialic acids (SA) as receptors , SA are not essential for infectivity . SA-using porcine rotaviruses OSU and CRW-8 appear to use ganglioside- and glycolipid-based receptors, respectively . RRV binds sialosides with low affinity via a galectin-like region in VP8* . In searching for rotavirus receptors, Coulson et al. found that VP4 and rotavirus outer capsid protein VP7 contain sequences corresponding to integrin recognition sites . Integrins are alpha/beta heterodimeric, transmembrane glycoproteins important for cell surface adhesion and signaling. The RDGE sequence in VP4 at amino acids (aa) 307 to 310 corresponds to the putative alpha2beta1 integrin recognition sequence DGE(A) in type I collagen . VP7 contains the alphaxbeta2 integrin ligand sequence, GPR , and several potential alpha4beta1 integrin ligand sites . Monoclonal antibodies (MAbs) to alpha2beta1 and alphaxbeta2, and peptides containing these integrin ligand sequences, inhibited SA11 and human rotavirus RV-5 infection of MA104 and Caco-2 cells, which were shown to express alpha2beta1 and alphaxbeta2 integrins, by 30 to 90% . As Caco-2 cells model small intestinal epithelial cells, this suggested that SA11 could use alpha2beta1 for infection of intestinal cells. Surface expression of alpha2beta1 correlated with susceptibility of MA104, Caco-2, RD, K562, and COS-7 cells to SA11 infection . SA11 showed increased levels of binding and growth in alpha2beta1- and alpha4beta1-transfected K562 cells, which were specifically blocked by anti-alpha2 and anti-alpha4 MAbs, respectively. From these data, it was concluded that alpha2beta1 and alpha4beta1 can act as SA11 receptors . It has been proposed that rotavirus-cell binding can involve initial carbohydrate recognition followed by integrin interaction . More recently, the neuraminidase-resistant RRV mutant nar3 was shown to bind alpha2beta1 . Another integrin, alphavbeta3, has been shown to promote infection by RRV, nar3, and human rotavirus Wa. Rotavirus binding to alphavbeta3 was not detected . It has been confirmed that infection by SA11 and RRV is inhibited by anti-alpha2 MAbs and DGE-containing peptides . The infectivity of several other rotaviruses, including Wa, was also inhibited by anti-alpha2 MAbs . However, binding of RRV to alpha2beta1 could not be detected in two studies , and evidence that SA11, Wa, and other rotavirus strains bind to alpha2beta1 was not found by one of these groups . A summary of the previously published studies of anti-integrin MAb blockade of rotavirus-cell binding and infection is presented in Table . To directly investigate the cell surface proteins bound by rotaviruses, we developed a novel virus immunoprecipitation technique designed to maintain integrin ligand-binding function, which included divalent cations, maintenance of the alpha/beta subunit association, and preservation of disulfide bonds. Using this method, we show here that SA11 precipitates surface beta1 from alpha2beta1-transfected Chinese hamster ovary (CHO) cells, SA11 and RRV precipitate two surface proteins with the characteristics of alpha2 and beta1 from MA104 cells, and SA11 precipitates these two proteins and a third protein of 160 to 165 kDa from Caco-2 cells. As SA11 binding and infectivity were inhibited by anti-alpha2 MAbs which also block alpha2beta1 binding to collagen and as infectivity was inhibited with the collagen-derived peptide sequence DGE , it was proposed that SA11 might bind alpha2beta1 similarly to collagen . Within alpha2beta1, the upper surface of the inserted (I) domain of alpha2, including the metal ion-dependent adhesion site (MIDAS), contains all the components required for collagen binding . Within the alpha2 MIDAS, aa residues 151, 221, and 254 are essential for alpha2beta1 binding to collagen . Echovirus type 1 binds to alpha2beta1, its cellular receptor, via a region of the alpha2 I domain partially overlapping that bound by a collagen-derived, triple-helical peptide . However, in contrast to collagen, the MIDAS is not involved in echovirus binding to alpha2beta1 . It has been reported that the infectivity of RRV and other rotaviruses was increased in CHO cells following transfection with human alpha2 or alpha2beta1 but that the level of virus binding was not altered . From this study, it was concluded that an anti-alpha2 MAb inhibited the increased infectivity but had no effect on virus binding to the alpha2- or alpha2beta1-transfected CHO cells. To analyze the roles of human alpha2, beta1, the alpha2 I domain, and the alpha2 MIDAS in rotavirus recognition of alpha2beta1, we utilized integrin-transfected CHO cell lines which had been used previously to map the human alpha2beta1 regions involved in collagen binding . We found that infectious SA11 and RRV specifically bound CHO cell-expressed human alpha2beta1 integrin and, to a lesser extent, human alpha2 combined with hamster beta1. This binding was inhibited by anti-alpha2 I domain MAbs but not by non-I domain MAbs to alpha2. It required the presence of the alpha2 I domain but not MIDAS residues 151, 221, and 254. Our experiments showed that the infectivity of SA11 and RRV and the replication of RRV were increased following virus binding to alpha2beta1 integrin on CHO cells. MATERIALS AND METHODS : Cell lines, viruses, and antibodies. | MA104 and Caco-2 cells were propagated as described before . The derivation of CHO K1 cell transfectants expressing human alpha2 (CHO alpha2), the human alpha2 and beta1 integrin subunits (CHO alpha2beta1), the human alpha2 subunit with alanine-swapping mutations (CHO alpha2 D151A, CHO alpha2 D254A, and CHO alpha2 T221A), the human beta1 and alpha2 subunits with an alpha2 I domain deletion (CHO alpha2delIbeta1), and the PBJ-1 empty vector (CHO K1 PBJ-1) has been described previously (, , , -). Cultivation of CHO cell lines was performed as for MA104 cells, except that G418 sulfate (Gibco BRL, Grand Island, N.Y.) at 0.1 to 1 mg/ml was included in transfectant growth medium. Integrin expression on CHO cell lines was monitored regularly by flow cytometry as described previously . Rotaviruses SA11, RRV, and Wa were grown in MA104 cells . For purification by glycerol gradient ultracentrifugation, virus was grown in cell cultures on beads . Rabbit antisera to SA11 (anti-SA11) and RRV (anti-RRV) were produced as described before and showed fluorescent focus reduction neutralization assay titers of 1:100,000 and 1:80,000, respectively, against homologous virus. According to the results of enzyme immunoassays (EIA) performed as described previously , preimmune rabbit sera (control Ab) showed reciprocal titers of <100 against SA11. Hybridoma cells producing the SA11-neutralizing, anti-VP4 MAb 2G4 were provided by M. Estes and J. Burns, Baylor College of Medicine, Houston, Tex. By fluorescent focus reduction neutralization assay, the titer of MAb 2G4 ascitic fluid against SA11 was 1:40,000. The derivation of antirotavirus MAbs RVA (to VP6), 60 (to a VP7 nonneutralization epitope), and RV-3:5 (to a VP7 neutralization epitope) has been described previously . MAbs P4C10, QE2.E5, and P5D2 to human beta1 were provided by D. Leavesley and P. Simmons, Royal Adelaide Hospital, Adelaide, South Australia, Australia (P4C10, QE2.E5, and P5D2), and E. Wayner, Fred Hutchinson Cancer Research Center, Seattle, Wash. (P4C10 and P5D2). Anti-human alpha2 MAbs AK7 , which (according to flow cytometry results) binds MA104 cell surface-expressed alpha2beta1 integrin and blocks SA11 infection of MA104 cells by 50 to 60% at 12 to 40 mug/ml , and P1E6 were purchased from Pharmingen, San Diego, Calif., and Invitrogen, Burwood, Victoria, Australia, respectively. Anti-human alpha2 MAb 12F1 and anti-beta1 MAb B3B11 were purchased from Chemicon, Temecula, Calif. Purified anti-human alpha2 MAbs HAS3 and HAS4 were provided by F. Watt, Imperial Cancer Research Fund, London, England. The epitopes recognized by the anti-integrin MAbs used are listed in Table . The pan-class I-reactive MAb W6/32 was provided by A. Brooks, Department of Microbiology and Immunology, The University of Melbourne, Melbourne, Victoria, Australia. Negative control MAbs MOPC21 (purchased from Cappel, ICN Pharmaceuticals Inc., Aurora, Ohio) and RV5:2 (specific to RV-5 rotavirus VP4) were matched with test MAbs for isotype and protein concentration. Binding of MAbs to cells was characterized by flow cytometry and expressed as the relative linear median fluorescence intensity (MFI), as described previously . Immunoprecipitation of integrin subunits using virus or anti-integrin antibodies. | Surface proteins on confluent cell monolayers (5 x 107 cells) were biotinylated as described previously , and then cells were lysed in 50 mM Tris-HCl buffer (pH 7.5) containing 5 mM iodoacetamide (Sigma, St. Louis, Mo.), 1 mM phenylmethylsulfonyl fluoride (Sigma), 1 mM CaCl2, 1 mM MgCl2, and 0.5% (vol/vol) Triton X-100 (Sigma) for 30 min at 4C. Lysates were clarified by centrifugation at 10,000 x g for 10 min at 4C, and lysate protein concentrations were determined using a detergent-compatible protein assay kit (Bio-Rad, Hercules, Calif.). Purified virus infectivity was activated with porcine trypsin type IX-S (Sigma) (10 mug/ml) for 10 min at 37C. Activated virus (5 mug; 107 to 108 infectious virions) was mixed with cell lysate (70 mug of protein in 500 mul of phosphate-buffered saline [PBS]) overnight at 4C. Anti-rotavirus Ab was added to a final dilution of 1:100 for 1 h at 4C, and Ab complexes were collected onto protein A-Sepharose beads (Amersham Pharmacia Biotech, Uppsala, Sweden) (2 mg in 100 mul) which had been blocked for 1 h at 4C in PBS containing 5% (wt/vol) skim milk powder (Diploma; Bonlac Foods, Altona, Australia). Beads were collected by centrifugation at 10,000 x g for 10 s and washed extensively with PBS. For MAb precipitations, cell extract (70 mug of protein) was reacted with 5 to 10 mug of MAb for 1 to 16 h at 4C. For precipitation of alpha2beta1 from MA104 cell lysates with alpha2-specific MAbs, MAb-cell protein complexes were collected with protein A-Sepharose beads that had been coated with rabbit anti-mouse immunoglobulins (Ig) (Dako, Germany) (0.2 mg anti-mouse Ig, mainly IgG, per 20 mg of beads) overnight at 4C and washed extensively with PBS. For all other MAb precipitations, Ab complexes were collected onto blocked protein A-Sepharose beads as described above for virus complexes. Precipitated proteins were separated from beads by boiling for 5 min in Laemmli's sample buffer containing 5% (vol/vol) beta-mercaptoethanol. Proteins and protein molecular weight standards (Novex, San Diego, Calif.) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis in a 7.5% (wt/vol) acrylamide resolving gel at 125 V. Proteins were electrophoretically transferred to a polyvinylidene difluoride membrane (Immobilon-P; Millipore, Bedford, Mass.) for 1 h at 100 V. After blocking in PBS containing 0.1% (vol/vol) Tween 20 (PBS-T) and 5% (wt/vol) skim milk powder for 1 h at 37C or overnight at 4C, the membrane was washed extensively with PBS-T. Biotinylated proteins were detected by incubation with streptavidin-conjugated horseradish peroxidase (StAv-HRP) (Silenus, Melbourne, Victoria, Australia), optimally diluted in PBS-T, for 1 h at 20C. Following washing as described above, bound StAv-HRP was detected by enhanced chemiluminescence (ECL) with ECL Hyperfilm (Amersham Pharmacia Biotech) as specified by the manufacturer, the positions of molecular weight markers were marked, and films were scanned using Adobe Photoshop software. Virus binding, infectivity, and growth assays. | Assays of infectious rotavirus binding to CHO K1 cell lines were performed using confluent monolayers in 24-well trays (Nunclon, Roskilde, Denmark) as previously described . Harvested virus titers were determined by indirect immunofluorescent staining of confluent MA104 cell monolayers inoculated with serial dilutions of virus and were expressed as the number of fluorescent cell-forming units (FCFU) per well . For MAb blockade of rotavirus binding, washed cell monolayers were incubated with MAbs (200 mul/well) at 37C for 1.5 h and then cooled to 4C prior to MAb removal and virus addition for the binding assay. Rotavirus infectivity in CHO cell lines was quantitated using confluent monolayers in 96-well trays (Nunclon), as described previously . In some experiments, virus was not trypsin activated to maintain cellular integrity. The number of virus-infected CHO cells was determined (using MAb RVA at a 1:2,000 dilution ) by indirect immunofluorescent staining. Rotavirus infection blockade (using anti-integrin MAbs) in MA104 cells was performed using the fluorescent focus reduction assay described previously , and growth curves were obtained using confluent MA104 or CHO cells in 24-well trays as described previously . Assay for MAb AK7 binding to the alpha2 integrin I domain. | Soluble glutathione S-transferase-alpha2 I domain (GST-alpha2I) and wild-type GST (wt-GST) were produced and purified as described previously , and their binding to MAb AK7 was determined using an EIA adapted from earlier studies . Microtiter plates (Immunosorb; Nunclon) were coated with GST-alpha2I or wt-GST diluted in PBS, at 0 to 500 mug/ml, for 2 h at 37C. MAbs (10 mug/ml) diluted in PBS containing 2.5% skim milk powder and 0.05% (vol/vol) Tween 20 (Sigma) were reacted overnight at 4C. Bound MAb was detected using HRP-conjugated anti-mouse Ig (Silenus). Statistical analysis. | The unpaired, two-tailed Student t test or two-way analysis of variance was used to assess the statistical significance of differences in virus-cell binding and infectivity. Significance was set at the 95% level. Unless otherwise stated, error bars represent the 95% confidence interval and the data presented in each figure were derived from at least three experiments. RESULTS : Rotaviruses specifically precipitate MA104 and Caco-2 cell surface proteins that migrate indistinguishably from those immunoprecipitated by MAbs to alpha2beta1. | The proteins precipitated by purified SA11 from lysates of surface-biotinylated MA104 cells and collected using anti-SA11 are shown in Fig. . Precipitated proteins ranged in molecular mass from approximately 160 to 30 kDa. Apart from a nonspecifically detected protein migrating at 50 kDa that was present in all precipitations, no proteins were precipitated by control Ab in the presence of virus or by anti-SA11 in the absence of virus. To confirm that the proteins visualized were biotinylated and derived from the cell surface, proteins precipitated by SA11 and collected using anti-SA11 from lysates of mock-biotinylated MA104 cells were compared with those obtained from cell surface-biotinylated lysates . The two proteins of 140 and 160 kDa (present in virus precipitates of biotinylated MA104 cell lysates) were the only proteins not visualized from virus precipitates of mock-biotinylated lysates. Thus, these were the only biotin-accessible proteins on the MA104 cell surface that were specifically precipitated by SA11. The nonbiotinylated proteins of 120 kDa or less proved to be viral in origin, as they migrated in gels indistinguishably from the VP1, VP2, VP5*, VP6, and VP7 proteins solubilized from purified SA11 rotavirus . They were precipitated by anti-SA11 only in the presence of virus and were detected in the presence or absence of MA104 cell proteins. Sufficient viral protein, but not 140- and 160-kDa proteins, was precipitated to be visible on Coomassie blue-stained gels, suggesting that ECL was detecting the large amounts of viral proteins present in the precipitations nonspecifically. This nonspecific reaction was not eliminated by StAv-HRP concentration optimization or alternative blocking reagents (data not shown). Streptavidin can bind to peptides of sequences unrelated to biotin . As shown in Fig. , RRV bound to 140- and 160-kDa proteins was precipitated by anti-SA11 from MA104 cell lysates. As SA11 and RRV share VP4 and VP7 serotype specificity, these antisera can be used interchangeably. Anti-SA11 was used rather than anti-RRV, as the anti-SA11 serum showed a higher neutralization and EIA titer to both SA11 and RRV than did the anti-RRV serum. Consistent with the lower titer of anti-RRV against SA11, SA11 that bound to the 140-kDa but not the 160-kDa protein was precipitated by anti-RRV from MA104 cell lysates . Anti-VP4 MAb 2G4 precipitated virus complexed with both the 140- and 160-kDa MA104 cell proteins similarly to anti-SA11 and more readily detected the 140-kDa protein . MAb 2G4 did not precipitate the 140- and 160-kDa MA104 cell proteins in the absence of virus. Neither protein was detected using MAbs RVA (VP6), 60 (VP7), RV-3:5 (VP7), or MOPC21 (negative control). The 160- and 140-kDa proteins precipitated by SA11 are comparable in size with the human alpha2 (approximately 165 kDa) and beta1 (approximately 130 kDa) integrin subunits and so might correspond to alpha2beta1. We therefore analyzed the abilities of two anti-alpha2 MAbs to recognize alpha2beta1 on MA104 cells and inhibit virus infection of these cells. MAb P1E6 bound to MA104 cell surface-expressed alpha2beta1 similarly to MAb AK7 according to flow cytometry, as the histograms obtained with the two MAbs used at the same concentrations were indistinguishable (data not shown). P1E6 and AK7 showed similar ratios of the MFI with anti-integrin MAb to the MFI with control MAb +- range (7.23 +- 0.48 and 6.41 +- 0.57, respectively) (data not shown). MAb P1E6 also blocked SA11 infection of MA104 cells by a mean range of 39% +- 7.1% at 40 mug/ml (P = 0.04). As illustrated in Fig. with data obtained with MAb P1E6, both anti-alpha2 MAbs AK7 and P1E6 immunoprecipitated two proteins of the expected molecular mass for the alpha2 and beta1 integrin subunits from cell surface-biotinylated MA104 cell extracts which migrated indistinguishably from the 160- and 140-kDa proteins precipitated by SA11. Although anti-beta1 MAbs P4C10, P5D2, and QE2.E5 also immunoprecipitated both the alpha2 and beta1 integrin subunits, these anti-beta1 MAbs immunoprecipitated the beta1 integrin subunit to a greater extent than the alpha2 subunit. This is shown in Fig. for MAb QE2.E5. The alpha2 band precipitated by QE2.E5 is not visible in Fig. , as exposure of the film sufficient to visualize alpha2 greatly overexposed the beta1 band. This is expected, as beta1 partners several alpha subunits, whereas alpha2 partners only beta1. Control MAbs did not precipitate these proteins. Anti-beta1 MAb B3B11 recognized a single protein of approximately 140 kDa, according to Western blotting of whole MA104 cell lysate directly applied to the gel (data not shown). As each integrin subunit-specific MAb precipitated both subunits from cell lysates, a significant proportion of the alpha2 and beta1 subunits in the lysates retained their noncovalent association, which is required for efficient ligand interaction . Although the intensity of the 140-kDa band tended to be higher than that of the 160-kDa band in SA11 precipitations, the 160- and 140-kDa protein bands precipitated by anti-alpha2 MAbs were approximately equal in intensity. This suggests that the alpha2 and beta1 proteins were labeled to a similar extent with biotin and that SA11 may have been precipitating more 140- than 160-kDa protein. However, it is also possible that these differences resulted from the nonquantitative nature of the immunoprecipitation and protein visualization techniques. Overall, these data suggest that the 140- and 160-kDa MA104 cell surface proteins bound by SA11 rotavirus are the alpha2 and beta1 integrin subunits. Possibly, SA11 precipitates free beta1 integrin, as well as alpha2beta1, from MA104 cell lysates. As shown in Fig. , SA11 precipitated a similar range of proteins from Caco-2 cell lysates as from MA104 cell lysates. The exceptions were that a 130-kDa Caco-2 protein was precipitated, rather than a 140-kDa protein, and two Caco-2 cell proteins of approximately 165 and 160 kDa were resolved in several experiments. No SA11-bound proteins were precipitated by control Ab (Fig. and B). The origin of the Caco-2 cell proteins visualized was determined by comparison of proteins precipitated by SA11 from mock- and surface-biotinylated cell lysates and in the absence or presence of cell lysate . As the proteins of approximately 120 kDa or less precipitated from biotinylated lysates were also present in precipitates from mock-biotinylated Caco-2 cell lysates and in the absence of lysate, only the three proteins at approximately 165, 160, and 130 kDa precipitated by SA11 originated from the Caco-2 cell surface. These SA11-precipitated proteins migrated indistinguishably from the two Caco-2 cell proteins precipitated by anti-beta1 MAb P5D2 and anti-alpha2 MAb P1E6 . The 165- and 160-kDa proteins precipitated with virus could not be resolved sufficiently to determine which protein comigrated with the 160-kDa alpha2 integrin band precipitated by MAb P1E6. Control MAbs RV-5:2 and MOPC21 did not precipitate alpha2beta1. The protein precipitated by RV-5:2 that migrated near beta1 did not comigrate with beta1 in gels with increased protein separation (data not shown). Thus, the 130- and 160- or 165-kDa Caco-2 cell surface proteins bound by SA11 appear to include alpha2 and beta1. Differences in integrin glycosylation between MA104 and Caco-2 cells might explain the difference in sizes estimated for the beta1 subunit in MA104 (140 kDa) and Caco-2 (130 kDa) cells . In contrast to the results of MA104 cell studies, SA11 precipitated similar amounts of the 160- or 165- and 130-kDa proteins as judged by band intensity. No surface-biotinylated Caco-2 cell proteins were visualized in precipitations with purified Wa. As the maximum ratio of infectivity (FCFU) to micrograms of protein obtainable with Wa (6.8 x 105) was 10- to 67-fold less than that obtained with SA11, it appears that the amount of infectious Wa obtainable was insufficient for detection of Wa binding to lysate proteins. Under standardized assay conditions which included using identical MAbs at the same concentrations, anti-alpha2 and anti-beta1 MAbs precipitated more alpha2beta1 integrin from Caco-2 cells than from MA104 cells, as judged by the intensity of staining of the protein bands. This is consistent with the finding of a higher level of alpha2beta1 expression on Caco-2 cells than on MA104 cells as detected by flow cytometry . The amount of alpha2beta1 precipitated by the MAbs also would have been affected by their specificity for human rather than monkey integrins. N-terminal sequence determination and Western blotting with the anti-beta1 MAb B3B11 that detected beta1 in whole MA104 cell lysate of the 160- and 140-kDa proteins precipitated by SA11 from cell lysates were unsuccessful due to insufficient amounts of protein for analysis. Mapping of MAb AK7 to the alpha2 I domain and flow cytometry characterization of integrin-transfected CHO cells. | As shown in Fig. , AK7, the most effective alpha2-specific MAb tested for blockade of SA11 cell binding and infectivity (, ; P. Halasz and B. S. Coulson, unpublished data), bound to GST-alpha2I as determined by EIA, confirming previous data showing that AK7 maps to the alpha2 I domain . Flow cytometry analysis using MAb HAS4 (which maps to the N-terminal region of alpha2 outside the I domain [, ]) and MAb AK7 showed that CHO K1 cells lacked human alpha2, CHO alpha2beta1 cells expressed both I domain and non-I domain epitopes, and CHO alpha2delIbeta1 cells lacked the I domain but expressed non-I domain epitopes similarly to CHO alpha2beta1 cells . MAb HAS3 maps to the same alpha2 region as MAb HAS4 and showed the same results as MAb HAS4 (data not shown). CHO alpha2, CHO alpha2 D151A, CHO alpha2 T221A, and CHO alpha2 D254A cells all expressed the alpha2 I domain AK7 epitope . Human beta1 (combined with human alpha2) was present on CHO alpha2beta1 and CHO alpha2delIbeta1 cells but was absent from CHO K1 and CHO alpha2 cells, as expected . We calculated from the flow cytometry histograms in Fig. that approximately 40% of the CHO alpha2beta1 cells expressed human alpha2 dimerized with hamster beta1. These histograms show that approximately 55% of the CHO alpha2beta1 cells expressed human beta1, as detected using MAb QE2.E5, which is specific for human beta1 and does not recognize hamster beta1 . Thus, the remaining 45% of these cells expressed hamster beta1 only. As detected using MAbs AK7 and HAS4 and determined from the data presented in Fig. , 86% of CHO alpha2beta1 cells expressed human alpha2. So, on average, approximately 90% of the CHO alpha2beta1 cells that expressed hamster beta1 also expressed human alpha2 and, overall, approximately 40% (calculated as 90% of 55%) of all the CHO alpha2beta1 cells expressed human alpha2 dimerized with hamster beta1. As the N and C termini of the I domain are closely associated in the three-dimensional structure and the I domain is inserted as an additional region sitting above the alpha subunit in a subset of integrins , deletion of the I domain is expected not to affect the function of the remaining regions of alpha2 or beta1, and this appears to be the case from our studies. SA11 rotavirus precipitates a CHO alpha2beta1 cell surface protein that migrates indistinguishably from beta1 and is not precipitated by SA11 from CHO K1, CHO alpha2, CHO alpha2delIbeta1, or CHO alpha2 T221A cell lysates. | Anti-alpha2 MAbs immunoprecipitated two proteins from CHO alpha2 and CHO alpha2beta1 cell lysates (AK7 and P1E6) and from CHO alpha2 Tau221Alpha lysates (P1E6) which migrated at the expected molecular masses for alpha2beta1 integrin (approximately 120 and 140 kDa) . No proteins were immunoprecipitated from CHO K1 lysates and CHO alpha2delIbeta1 cell lysates (data not shown) by these MAbs. According to the results of Western blotting of CHO alpha2beta1 and CHO alpha2delIbeta1 cell lysates, anti-beta1 MAb B3B11 recognized a single protein, of approximately 140 kDa that was not detected in CHO K1 or CHO alpha2 lysates (data not shown). SA11 weakly precipitated two proteins, of approximately 140 and 120 kDa (Fig. and C), from CHO alpha2beta1 cell lysates which were not precipitated from CHO K1, CHO alpha2, and CHO alpha2 Tau221Alpha lysates or CHO alpha2delIbeta1 lysates (data not shown). As shown in Fig. , the 140-kDa protein precipitated by SA11 from CHO alpha2beta1 cells migrated indistinguishably from the beta1 integrin subunit immunoprecipitated in conjunction with the alpha2 subunit by MAb AK7. As CHO alpha2 and CHO alpha2beta1 cells differ only in the presence of human beta1, this shows that SA11 precipitated the beta1 integrin from CHO alpha2beta1 cell lysates. The inability of SA11 to precipitate alpha2 or beta1 from CHO alpha2delIbeta1 cells suggests that the beta1 precipitation from CHO alpha2beta1 cell lysates was dependent on the presence of the alpha2 I domain. The identity of the 120-kDa protein precipitated by virus from CHO alpha2beta1 cell lysates (Fig. and C) is unclear. It may have been associated with beta1 or may have been viral in origin. As shown in Fig. , the beta1 band precipitated by anti-alpha2 MAbs P1E6 and AK7 appears more intense than the alpha2 band. This suggests that in contrast to results with MA104 cells, the beta1 may be more heavily biotinylated than the alpha2 in CHO cells. This could explain the detection of SA11 precipitation of beta1, but not alpha2, from CHO alpha2beta1 cell lysates. Binding of SA11 and RRV to alpha2beta1 and infectivity of SA11 in MA104 cells are blocked by an anti-alpha2 I domain MAb but not by MAbs directed to a region of alpha2 outside the I domain. | At a multiplicity of infection (MOI) of 5 (calculated from MA104 infectious cell titers) and in the absence of any Ab treatment, titers of SA11 and RRV bound to CHO alpha2beta1 and CHO alpha2 cells were significantly higher than those of SA11 and RRV bound to CHO K1 or CHO K1 PBJ-1 cells (P < 0.001) (Fig. and ). CHO alpha2beta1 and CHO alpha2 cells bound 1.6- to 1.7-fold- and 1.3-fold-higher titers of SA11 and RRV, respectively, than did CHO K1 or CHO K1 PBJ-1 cells. CHO alpha2beta1 cells bound significantly higher virus titers than did CHO alpha2 cells (P < 0.01), suggesting that the beta1 subunit plays a role in facilitating virus binding to alpha2beta1 integrin. As shown in Fig. , the increase in RRV binding to CHO alpha2beta1 and CHO alpha2 cells was abolished by cellular treatment with MAb AK7 but not with MAbs HAS4 and MOPC21. Similarly, the increased SA11 binding to CHO alpha2beta1 and CHO alpha2 cells was reduced to approximately 110% of binding to CHO K1 cells by treatment with MAb AK7 but not by treatment with MAbs HAS4 and MOPC21. MAb HAS3 had the same effect as MAb HAS4 on SA11 and RRV binding to MA104 cells (data not shown). Thus, SA11 and RRV bind human alpha2beta1 expressed on CHO K1 cells to a higher level than to that of human alpha2 dimerized with hamster beta1 and this binding involves the alpha2 I domain. At the maximum MOI achievable (an MOI of 1), CHO K1, CHO alpha2, and CHO alpha2beta1 cells bound Wa at similarly low titers (data not shown). Thus, no binding of Wa to human alpha2beta1 or human alpha2 dimerized with hamster beta1 was detected using integrin-transfected CHO cells. MAbs HAS4 and HAS3 (data not shown) had no effect on SA11 infectivity in MA104 cells. In contrast, MAb AK7 inhibited SA11 infectivity in a dose-dependent fashion, as reported previously . According to the results of flow cytometry, the major histocompatibility complex (MHC) class I-specific MAb W6/32, which has been used previously as a negative control for anti-alpha2 MAb blockade of echovirus 1 cell binding , bound 99.5% of MA104 cells with an MFI of 20.1 +- 0.5, but this binding did not affect SA11 infectivity, further demonstrating the specificity of the inhibition by MAb AK7. W6/32 also had no effect on the infectivity of RRV and Wa (data not shown). Thus, the interaction of SA11 with monkey alpha2beta1 on MA104 cells is specific for alpha2 and involves the alpha2 I domain. A comparison of the abilities of anti-alpha2 MAbs to inhibit rotavirus, echovirus 1, and collagen binding is presented in Table . Echovirus and collagen both bind the alpha2 I domain . As the patterns of MAb inhibition are very similar between echovirus, collagen, and rotavirus and only anti-I domain MAbs block the interaction of these three ligands with alpha2, these MAb data strongly suggest that rotavirus also binds the alpha2 I domain. The binding of SA11 and RRV rotaviruses to alpha2beta1 is eliminated by deletion of the alpha2 I domain but not by MIDAS point mutations. | As shown in Fig. , titers of SA11 and RRV bound to CHO alpha2delIbeta1 cells were significantly (1.6- to 1.7-fold) lower than those of SA11 and RRV bound to CHO alpha2beta1 cells (P < 0.001), whereas titers of SA11 and RRV bound to CHO alpha2delIbeta1, CHO K1, and CHO K1 PBJ-1 cell lines were indistinguishable (P > 0.05). Thus, deletion of the alpha2 I domain eliminated virus binding to alpha2beta1. CHO alpha2beta1 cells bound levels of SA11 similar to those of permissive MA104 cells (P > 0.05). The CHO alpha2, CHO alpha2 D151A, CHO alpha2 D254A, and CHO alpha2 T221A cell lines all bound similar levels of SA11 and RRV (P > 0.05), so binding of these viruses to human alpha2 dimerized with hamster beta1 is not affected by mutagenesis of alpha2 I domain amino acid residues in the MIDAS region (D151, T221, and D254) that are critical for the ligation of collagen . The infectivity of SA11 and RRV mediated via alpha2beta1 is eliminated by deletion of the alpha2 I domain. | Rotavirus infectivity was measured as the number of cells infected after a single round of virus replication at 16 h after infection. Consistent with previous studies , CHO K1 cells were highly resistant to infection with SA11 and RRV. As shown in Fig. , at a MOI of 20 (calculated using MA104 cells), SA11 infected only 1 to 20 of 50,000 (<=0.04%) cells and RRV infected only 40 to 70 (<=0.1%) cells. Trypsin-activated virus infectivity levels were higher in CHO alpha2 cells than in CHO K1 cells. SA11 titers in CHO alpha2 cells were 90-fold higher (P = 0.03) and RRV titers were 1.5-fold higher (P < 0.05) than those in CHO K1 cells. Greater increases in infectivity of trypsin-activated virus were measured in CHO alpha2beta1 cells than in CHO K1 cells (190-fold for SA11 [P = 0.02] and 2.8-fold for RRV [P < 0.001]). Although SA11 and RRV infected significantly more CHO alpha2beta1 cells than CHO K1 cells, the infected cells represented only 0.1 to 0.4% of the cells expressing alpha2beta1. For each CHO cell line, the relative levels of virus infectivity closely mirrored the bound virus titers. SA11 and RRV that were not trypsin activated infected fewer CHO alpha2beta1 cells than did fully activated virus , but levels of trypsin-activated and nonactivated virus bound relative to CHO K1 cells were similar. Nonactivated SA11 infected 1.9-fold more CHO alpha2beta1 cells (P < 0.05) than CHO K1 cells, and nonactivated RRV infected 2.2-fold more CHO alpha2beta1 cells (P < 0.001) than CHO K1 cells. CHO alpha2delIbeta1 cells were highly sensitive to trypsin and lost adhesion readily, apparently due to their loss of the ability to bind extracellular matrix by using the alpha2 I domain, so they were infected with nonactivated viruses only. Consistent with the relative abilities of SA11 and RRV to bind CHO alpha2delIbeta1 and CHO K1 cells, nonactivated SA11 and RRV each infected similar numbers of these cells (P > 0.05), showing that the infectivity of nonactivated SA11 and RRV mediated by alpha2beta1 depends on the presence of the alpha2 I domain. No Wa-infected CHO K1, CHO alpha2, or CHO alpha2beta1 cells were detected using virus at a MOI of 1 (data not shown). The growth of RRV, but not SA11 or Wa, in CHO K1 cells is enhanced by transfection with alpha2beta1. | As shown in Fig. , RRV growth was barely detectable in CHO K1 cells, with a 1.6-fold increase in titer at 24 h postinfection (p.i.) and a reduced titer at 48 h p.i. RRV titers in CHO K1 PBJ-1 and CHO alpha2 cells continuously declined. Thus, the process of transfection and/or the requirement for G418 selection for maintenance of transfectants rendered the cells completely resistant to production of infectious RRV and this resistance was not overcome by alpha2 transfection. Others have shown that G418 exposure of CHO cells carrying empty vector selected for endogenous Bcl-2 expression and that Bcl-2 expression limited reovirus growth , suggesting that G418 selection might have been responsible for limiting RRV growth. However, CHO alpha2beta1 cells supported a sixfold increase in RRV titer at 24 h p.i., showing that alpha2beta1 transfection allowed completion of the RRV replicative cycle and production of infectious RRV. By 48 h p.i., the increase was reduced to twofold, suggesting that only a single cycle of RRV replication had occurred. In all CHO cell lines tested, net yields of SA11 and Wa (data not shown) declined. A summary of all the experiments described in this paper, their outcomes, the MAbs used, and the MAb epitopes is presented in Table . DISCUSSION : In the present study, rotaviruses SA11 and RRV were shown to precipitate surface proteins migrating identically with alpha2beta1 from highly permissive monkey kidney and colonic epithelial cell lines and to precipitate surface beta1 from CHO cells transfected with human alpha2beta1. On transfected CHO cells, these rotaviruses bound to human alpha2beta1 and human alpha2 combined with hamster beta1. Virus binding to human alpha2beta1 required the presence of the alpha2 I domain and was facilitated by the presence of the homologous beta1 subunit. Virus binding to human alpha2beta1 and to human alpha2 combined with hamster beta1 was inhibited by anti-I domain MAbs but not by anti-alpha2 MAbs mapping outside the I domain. This is strong evidence that SA11 and RRV bind alpha2beta1 and that this binding requires the alpha2 I domain. In confirmation of these studies, Graham et al. have shown that that the alpha2 ligand peptide DGEA and anti-alpha2 MAb AK7 specifically inhibit SA11, RRV, and Wa binding to MA104 and alpha2-transfected K562 cells and that as determined by EIA, Escherichia coli-expressed, purified alpha2 I domain binds to RRV VP5*-GST . In an earlier study, it was reported that levels of RRV binding to CHO cells transfected with human alpha2 or alpha2beta1 were similar to those of RRV binding to a range of other cells, including CHO K1 and MA104 cells, and that anti-alpha2 MAb AK7 did not affect RRV binding to CHO cells transfected with human alpha2 or alpha2beta1 . It was concluded that, with the exception of the nar strain, rotaviruses do not bind to cells via alpha2beta1 . Our ability to demonstrate SA11 and RRV binding to alpha2beta1 on CHO cells may have been due to methodological differences between the binding assays used in the cited negative study and that used in our work. Although Ciarlet et al. tried several different binding assay protocols , they unfortunately did not use the method which Hewish et al. had found earlier to be successful in the detection of SA11 binding to alpha2beta1 on K562 cells and which had been used by two groups in the mapping of rotavirus cell attachment to VP4. The modifications made by Ciarlet et al. to this method included measurement of virus binding to trypsinized cells in suspension. Removal of cells from the extracellular matrix produces changed expression and activation of cell adhesion molecules, including integrins , alterations in activation of signal transduction pathways, and anoikis , which could affect rotavirus binding. Exposed sites on suspended cells that are not normally accessible on cells in monolayers could lead to increased levels of nonspecific virus binding. Ciarlet et al. demonstrated rotavirus infectivity blockade by anti-alpha2 MAb AK7 in monolayers of integrin-transfected CHO cells , which has been confirmed in this study, so the rationale for measuring virus binding to cells in suspension is unclear. Using flow cytometry, the specificity of the virus-cell binding assay used here was demonstrated with three MAbs that bound alpha2 on CHO alpha2 and CHO alpha2beta1 cells. The anti-alpha2 I domain MAb AK7 inhibited SA11 and RRV binding to alpha2beta1 on these cells, whereas virus binding was not affected by two MAbs which mapped to alpha2 outside the I domain. Binding assay specificity was also shown by the increase in titers of virus bound to CHO alpha2 and CHO alpha2beta1 cells, but not to CHO alpha2delIbeta1 cells, over those of controls. The inability of anti-class I MAb W6/32 to inhibit virus infectivity, even though this MAb bound 100% of MA104 cells as determined by flow cytometry, further demonstrates the specificity of this binding assay. It has been shown elsewhere that MAb W6/32 does not affect RRV and SA11 binding to MA104 cells, whereas MAb AK7 at 20 mug/ml inhibits binding of these viruses by 30 and 35%, respectively . Echovirus 1 binding to alpha2beta1 was reduced by mutations at aa 199 to 201, 212, 214, and 216 in the alpha2 I domain and required N289 . In contrast, type I collagen binding required D151, T221, and T254, which provide coordinating side chains for the metal ion in the MIDAS region , and collagen binding was reduced by mutations in the alpha2 I domain at N154, N190, D219, E256, H258, D259, D292, and E299 . As SA11 and RRV rotavirus binding to alpha2 I domain was not affected by alanine mutagenesis of D151, T221, and T254, these MIDAS residues do not appear to be critical for virus binding and rotavirus binding to the alpha2 I domain is distinguishable from that of type I collagen. More detailed mapping of the regions and residues involved in rotavirus binding to the alpha2 I domain would be needed to determine whether rotaviruses and echovirus 1 have similar binding requirements. Interestingly, Graham et al. have shown elsewhere that binding of SA11 to alpha2beta1 on K562 cells is increased following alpha2beta1 activation . In contrast, echovirus 1 binding to alpha2beta1 on K562 cells is not affected by alpha2beta1 activation . This suggests that alpha2beta1 binding by rotaviruses, echoviruses, and collagen involves distinct alpha2beta1 residues. The 140-kDa protein precipitated by SA11 from CHO alpha2beta1 cells was identified as the beta1 integrin subunit by its comigration with the beta1 coprecipitated (using anti-alpha2 MAbs) with alpha2 and by the inability of virus to precipitate this protein from parental CHO cells, CHO cells transfected with alpha2 alone, or CHO cells transfected with alpha2beta1 lacking the alpha2 I domain. This supports the conclusion that the 130- and 140-kDa proteins, migrating indistinguishably from beta1, which were precipitated from Caco-2 and MA104 cells by SA11 and/or RRV, respectively, are likely to be beta1. The beta1 subunit displays differing degrees of mobility in different cell types due to differing degrees of glycosylation . The detection of SA11 precipitation of beta1 from CHO alpha2beta1, but not CHO alpha2, cells is consistent with the levels of binding of infectious SA11 and RRV to alpha2beta1 being highest on CHO alpha2beta1 cells. The precipitation of beta1 but not alpha2 from CHO alpha2beta1 cells was unexpected. As the beta1 subunit appeared to be more heavily biotinylated than the alpha2 subunit in lysates of these cells, it is possible that precipitated alpha2 was not detected due to a lack of assay sensitivity. Alternatively, SA11 may be capable of binding beta1 epitopes directly. The precipitation of an apparently larger amount of beta1 than alpha2 from MA104 cell lysates by SA11 is consistent with the latter possibility. An additional explanation is that these differences in virus-precipitated protein band intensity result from interexperimental variation rather than true biological differences. MAb 2G4 binds to the putative fusion region of VP5* , which consists of 90 aa in the linear sequence from the DGE tripeptide sequence implicated in alpha2beta1 binding. As MAb 2G4 precipitated SA11 complexed with proteins indistinguishable from alpha2beta1, some 2G4 recognition sites were available for binding after the complexes were formed. Previous studies of cellular proteins bound by rotaviruses have used virus overlay protein blot assays (VOPBA), in which cellular proteins were reduced and dissociated before being reacted with virus . MA104 cell proteins of 66 to 97 kDa bound by SA11 in VOPBA and octyl glucoside-extracted MA104 cell proteins, including heat shock cognate protein 70 , which blocked rotavirus infection , were not precipitated in our studies. Two proteins were recognized by SA11 in VOPBA through a galactose moiety and so may not have bound virus with sufficient affinity for detection by precipitation. SA11 and RRV did not precipitate alphaxbeta2 or alphavbeta3 from MA104 cell lysates, as the only cellular proteins specifically precipitated by virus migrated indistinguishably from alpha2beta1 at 160 and 140 kDa. The alphax, beta2, alphav, and beta3 integrin subunits migrate at approximately 150, 95, 150, and 110 kDa . Consistent with this, RRV, Wa, and nar3 rotaviruses do not bind beta3 on these cells . However, the unidentified Caco-2 cell surface protein of approximately 160 to 165 kDa precipitated by SA11 might be alphax or alphav. It is also possible that this 160- to 165-kDa protein was associated with alpha2beta1 rather than precipitated by virus directly. It is significant that proteins with the characteristics of alpha2beta1 were precipitated by SA11 from Caco-2 cells, as these cells share some characteristics with human intestinal enterocytes. In microarray analysis of gene expression 16 h after RRV infection of Caco-2 cells, alpha2 and beta1 expression were increased 3.1-fold and 2.5-fold, respectively, and it was suggested that infection might upregulate this receptor on infected or nearby uninfected cells . CHO cell transfection with alpha2 or alpha2beta1 increased cell binding by SA11 and RRV to similar degrees, and both viruses used alpha2beta1 to infect the cells. However, RRV infected larger numbers of cells and only RRV infection produced detectable infectious virus (in CHO alpha2beta1 cells only) in growth assays. Also, only a small fraction of the SA11 and RRV bound to alpha2beta1 entered CHO K1 cells, as measured by the number of infected cells at 16 h after infection. This low level of infectivity was shown previously for RRV . These observations suggest that these cells have a block in rotavirus entry after attachment via alpha2beta1 (and an additional barrier to production or release of infectious virions following viral protein synthesis) and that these blocks affected SA11 to a greater extent than RRV. The existence of a block in rotavirus cell entry is supported by the report that liposome-mediated infection of CHO K1 cells with double-layered rotavirus particles, which are noninfectious when added exogenously to cells and lack VP4 or VP7, resulted in levels of virus infection similar to those seen with MA104 cells . It would be interesting to determine whether CHO K1 cells expressing alpha2beta1, alphaxbeta2, and alphavbeta3 allow more efficient entry of SA11 and RRV, as alphaxbeta2 and alphavbeta3 have a role in rotavirus entry, and when a mixture of alpha2beta1 and alphaxbeta2 ligand peptides or combinations of MAbs to alpha2beta1, alphaxbeta2, and alphavbeta3 are used, blockade of rotavirus infectivity in permissive cells is additive. It is noteworthy that the elevated binding of SA11 and RRV to alpha2beta1 in the presence of homologous, rather than heterologous, beta1 led to increased SA11 and RRV infectivity and increased RRV growth. The presence of human beta1 may have stabilized alpha2beta1 in a higher activation state. It has been shown that activation of human beta1 on K562 cells expressing alpha2beta1 results in increased levels of infectious SA11 binding to alpha2beta1 . A rotavirus preference for homologous integrin could help explain species-specific restrictions in virus tropism. In contrast to the results seen with SA11 and RRV, Wa neither bound alpha2beta1 on CHO cells nor infected CHO cells, a result in line with those of previous studies . However, anti-alpha2 MAb blockade of Wa infectivity in MA104 cells has been demonstrated . Also, Graham et al. have shown Wa binding to alpha2beta1 on K562 cells, MAb AK7 blockade of Wa binding to these cells, and AK7 blockade of Wa binding to MA104 cells . Thus, an accessory molecule or additional receptor required for Wa, but not SA11 or RRV, binding to alpha2beta1 appears to be lacking from CHO cells but is present on K562 and MA104 cells. This provides a basis for further studies to determine the full requirements for human rotavirus cell attachment. Rotavirus binding to alpha2beta1, and interaction with alphaxbeta2 and alphavbeta3 during cell entry, is consistent with the proposed mechanisms of rotavirus cell entry. Integrin activation following ligand binding leads to formation of focal adhesions in which integrins are clustered and cytoskeletal and signaling proteins are recruited to the integrin cytoplasmic domains . Activated integrins can facilitate endocytosis through direct interaction with protein kinase Calpha and caveolin . Although the pathway of rotavirus cell entry is not defined, there is evidence for both endocytosis and direct membrane penetration . Adenovirus recognition of alphavbeta3 via the penton base leads to both clathrin-dependent endocytosis and macropinocytosis of virus , whereas echovirus 1 binding to alpha2beta1 results in virus entry through caveolae . Thus, even though both these viruses use integrins which can be associated with caveolin , they enter cells through different pathways. Therefore, rotavirus cell entry pathway(s) cannot be predicted from the viral integrin usage. It will be important to determine whether rotavirus is internalized in a complex with integrin following binding or whether the virus-integrin interaction is transient and serves to increase the proximity of the VP5* fusion region to the cell membrane. Induction of signaling cascades by rotavirus-integrin interactions could facilitate cell entry and might also enhance virus replication by modifying cell functions such as survival and differentiation. The studies reported here have helped establish the role of alpha2beta1 in rotavirus attachment and entry. On CHO cells, SA11 and RRV bind to alpha2beta1, resulting in a low level of cellular infection. This binding requires the alpha2 I domain and is facilitated by the presence of homologous beta1. As SA11 and RRV do not require the same MIDAS residues as type I collagen for binding to the alpha2 I domain, human rotavirus binding to alpha2beta1 also may be distinguishable from that of collagen. The development of therapeutic agents capable of selectively inhibiting rotavirus binding to alpha2beta1 might therefore be feasible. It will be important to determine whether rotavirus-neutralizing antibodies (induced by natural infection or candidate rotavirus vaccines) act by inhibition of virus binding to alpha2beta1. FIG.1. : Comparison of MA104 cell surface proteins precipitated by SA11 and RRV with those precipitated by MAbs to alpha2beta1. Comparison of MA104 cell surface proteins precipitated by SA11 and RRV with those precipitated by MAbs to alpha2beta1. (A) SA11 precipitated a range of proteins. Cell lysate proteins were reacted with SA11 (+) or PBS (-) and then collected with anti-SA11 or control Ab. The data represent those obtained in twenty experiments. (B) Of the proteins precipitated by SA11, only the 160- and 140-kDa proteins were biotinylated and derived from the cell surface. SA11-precipitated, mock-biotinylated (-Bn) or biotinylated (+Bn) cell lysate was reacted with anti-SA11 or control Ab. Mock-biotinylated lysates were prepared identically to biotinylated lysates, with the omission of N-hydroxysulfosuccinimide-biotin. (C) The nonbiotinylated proteins in SA11 precipitates were viral in origin. Precipitated proteins were separated in duplicate gels run simultaneously. The biotinylated proteins in one gel (ECL) were detected with StAv-HRP and ECL after transfer to a polyvinylidene difluoride membrane (see Materials and Methods). Proteins in the other gel (Coomassie) were stained at 20C with 0.1% (wt/vol) Coomassie brilliant blue (Bio-Rad) in 20% (vol/vol) methanol and 20% (vol/vol) acetic acid for 3 h and destained overnight in 40% (vol/vol) methanol and 10% (vol/vol) acetic acid. Cell lysate proteins (+) or PBS (-) were reacted with SA11 and then collected with anti-SA11 and corun in comparison with 2 mug of reduced, purified SA11 proteins (virions). (D) RRV also precipitated two MA104 cell surface proteins of 160 and 140 kDa. Cell lysate proteins (+) or PBS (-) were reacted with RRV (+) or PBS (-) and then collected with anti-SA11, control Ab, or PBS. (E) Rabbit anti-RRV serum and anti-VP4 MAb 2G4 also precipitated the 160- and 140k-Da proteins following SA11 incubation with cell lysate. Lysate proteins precipitated by SA11 (+) or PBS (-) were collected using anti-RRV, 2G4 MAb, or MOPC21 (MOPC) MAb. These data represent those obtained in two experiments. (F) Comparison of the 160- and 140-kDa proteins precipitated by SA11 with those immunoprecipitated by MAbs to alpha2beta1. Lysate proteins were reacted with SA11 and then precipitated with anti-SA11 or control Ab as described previously. These were corun with lysate proteins reactive with anti-beta1 MAb QE2.E5, anti-alpha2 MAb P1E6, or control MAb MOPC21. Complexes in the latter two reactions were collected onto anti-mouse Ig-coated beads (see Materials and Methods). The molecular masses of the major protein bands, calculated from the positions of the molecular mass standards, are indicated in kilodaltons on the right sides of the panels. The identified rotaviral proteins are designated using the "VP" nomenclature, and their molecular masses in kilodaltons are shown in parentheses on the right sides of the panels. FIG. 2. : Comparison of Caco-2 cell surface proteins precipitated by SA11 with those precipitated by MAbs to alpha2beta1. Comparison of Caco-2 cell surface proteins precipitated by SA11 with those precipitated by MAbs to alpha2beta1. (A) SA11 precipitated a range of proteins. Proteins in cell lysates were reacted with SA11 (+) or PBS (-) and then collected with anti-SA11 or control Ab. (B) Of the proteins precipitated by SA11, only the 165-, 160-, and 130-kDa proteins were biotinylated and derived from the cell surface. SA11-precipitated, mock-biotinylated (-Bn), or biotinylated (+Bn) cell lysate was reacted with anti-SA11 or control Ab. (C) The nonbiotinylated proteins in SA11 precipitates are viral in origin. Cell lysate proteins (+) or PBS (-) were reacted with SA11 and then collected with anti-SA11. The 130-kDa band was not detected in this particular experiment. (D) Comparison of the 165-, 160-, and 130-kDa proteins precipitated by SA11 with proteins immunoprecipitated by MAbs to alpha2beta1. Lysate proteins were reacted with SA11 and then precipitated with anti-SA11 as described previously. These were corun with lysate proteins reactive with anti-beta1 MAb P5D2, control MAb RV-5:2, anti-alpha2 MAb P1E6, or control MAb MOPC21. The estimated protein sizes and identities are indicated as described in the legend to Fig. . FIG. 3. : Mapping of MAb AK7 to the alpha2 I domain and flow cytometry analysis of alpha2beta1 integrin expression on CHO K1 cell lines. Mapping of MAb AK7 to the alpha2 I domain and flow cytometry analysis of alpha2beta1 integrin expression on CHO K1 cell lines. (A) MAb AK7 binds by EIA to GST-alpha2I but not to wt-GST control. (B, C, and D) Detection by flow cytometry of surface expression of human alpha2 and beta1 on CHO cell lines used in this study. (B) The I domain deletion from CHO alpha2delIbeta1 cells and its presence on CHO alpha2beta1 cells were confirmed using MAb AK7, and the presence of the human alpha2 subunit on CHO alpha2beta1 and CHO alpha2delIbeta1 cells was demonstrated using MAb HAS4. (C) The presence of the alpha2 subunit, including the I domain, on CHO alpha2, CHO alpha2 D151A, CHO alpha2 T221A, and CHO alpha2 D254A cells was confirmed using MAb AK7. (D) Human beta1 expressed on CHO alpha2beta1 and CHO alpha2delIbeta1 cells, but not on CHO alpha2 cells, was detected using MAb QE2.E5. FIG. 4. : SA11 immunoprecipitated beta1 and a 120-kDa protein from CHO alpha2beta1 cells. SA11 immunoprecipitated beta1 and a 120-kDa protein from CHO alpha2beta1 cells. (A) The precipitation of cell surface alpha2 and beta1 by anti-alpha2 MAbs AK7 and P1E6 from CHO alpha2, CHO alpha2 T221A, and CHO alpha2beta1 cells, but not from CHO K1 cells, is shown. (B) SA11 precipitated 140- and 120-kDa surface proteins from CHO alpha2beta1 cells but not from CHO alpha2, CHO alpha2 T221A, and CHO alpha2beta1 cells. (C) The 140-kDa protein was identified as beta1 by comigration with the beta1 that was coprecipitated with alpha2 by MAb AK7. The estimated protein sizes and identities are indicated as described in the legend to Fig. . FIG. 5. : Inhibition by anti-I domain MAb AK7, but not by non-I domain MAb HAS4, of infectious SA11 (upper panel) and RRV (lower panel) binding to human alpha2beta1 and to human alpha2 combined with hamster beta1 on CHO cells (A) and of SA11 infection of MA104 cells (B). Inhibition by anti-I domain MAb AK7, but not by non-I domain MAb HAS4, of infectious SA11 (upper panel) and RRV (lower panel) binding to human alpha2beta1 and to human alpha2 combined with hamster beta1 on CHO cells (A) and of SA11 infection of MA104 cells (B). MOI = 5. MAb W6/32 is directed to MHC class I. MOPC21 is an isotype-matched negative control MAb for AK7, W6/32, and HAS4. Bars represent 95% confidence intervals (A) or standard deviations (B). FIG. 6. : Titers of SA11 (upper panel) and RRV (lower panel) bound to CHO K1 cells and their transfectants show that the alpha2 I domain is essential for virus binding to alpha2beta1 and that point mutations in the MIDAS do not affect virus binding. Titers of SA11 (upper panel) and RRV (lower panel) bound to CHO K1 cells and their transfectants show that the alpha2 I domain is essential for virus binding to alpha2beta1 and that point mutations in the MIDAS do not affect virus binding. MOI = 5. FIG. 7. : The increased infectivity of SA11 (left panels) and RRV (right panels) in CHO alpha2beta1 cells compared to that seen with CHO K1 cells was abolished by deletion of the alpha2 I domain (A), and the growth of RRV, but not SA11, was increased in CHO alpha2beta1 cells compared to that seen with CHO K1 and CHO alpha2 cells (B). The increased infectivity of SA11 (left panels) and RRV (right panels) in CHO alpha2beta1 cells compared to that seen with CHO K1 cells was abolished by deletion of the alpha2 I domain (A), and the growth of RRV, but not SA11, was increased in CHO alpha2beta1 cells compared to that seen with CHO K1 and CHO alpha2 cells (B). (A) Panels labeled "Trypsin-activated virus" show data for virus grown in the presence of trypsin; panels labeled "Non-activated virus" show data for virus grown in the absence of trypsin. MOIs of 20 (A) and 5 (B) were used. TABLE 1 : Summary of previously published studies of anti-integrin MAb blockade of rotavirus-cell binding and infection TABLE 2 : Ability of anti-alpha2 MAbs to inhibit cellular interactions of rotaviruses, echovirus 1, and collagen TABLE 3 : Outline of experiments performed and results obtained in this study Backmatter: PMID- 12915572 TI - CCAAT/Enhancer-Binding Protein-alpha Is Induced during the Early Stages of Kaposi's Sarcoma-Associated Herpesvirus (KSHV) Lytic Cycle Reactivation and Together with the KSHV Replication and Transcription Activator (RTA) Cooperatively Stimulates the Viral RTA, MTA, and PAN Promoters AB - During the immediate-early (IE) phase of reactivation from latency, the Kaposi's sarcoma-associated herpesvirus (KSHV) replication and transcription activator protein (RTA) (or ORF50) is thought to be the most critical trigger that upregulates expression of many downstream viral lytic cycle genes, including the delayed-early (DE) gene encoding the replication-associated protein (RAP) (or K8). RAP physically interacts with and stabilizes the cellular transcription factor CCAAT/enhancer-binding protein-alpha (C/EBPalpha), leading to upregulated expression of the cellular C/EBPalpha and p21CIP-1 proteins followed by G0/G1 cell cycle arrest. Furthermore, RTA also interacts with C/EBPalpha, and both RAP and RTA cooperate with C/EBPalpha to activate the RAP promoter through binding to a strong proximal C/EBP binding site that also serves as an RTA-responsive element (RRE). Here we show that C/EBPalpha also activates the IE RTA promoter in transient-cotransfection reporter gene assays and that addition of either RTA or RAP enhances the effect. Electrophoretic mobility shift assay and deletion analysis revealed three C/EBP binding sites that mediate cooperative transactivation of the RTA promoter by C/EBPalpha and RTA. Furthermore, chromatin immunoprecipitation assay results showed that the endogenous C/EBPalpha, RTA, and RAP proteins all associate with RTA promoter sequences in tetradecanoyl phorbol acetate-induced primary effusion lymphoma (PEL) cells. Induction of endogenous KSHV RTA mRNA in PEL cells by exogenously introduced C/EBPalpha was confirmed by reverse transcription-PCR analysis and by double-label indirect immunofluorescence assays. Reciprocally, expression of exogenous RTA also led to an increase of endogenous C/EBPalpha expression that could be detected by Western immunoblot assays even in KSHV-negative DG75 cells. Cotransfected RTA also increased positive C/EBPalpha autoregulation of the C/EBPalpha promoter in transient-cotransfection reporter gene assays. Finally, C/EBPalpha proved to strongly activate the promoters of two other KSHV DE genes encoding PAN (polyadenylated nuclear) RNA and MTA (ORF57), which was again mediated by C/EBP binding sites that also contribute to RTA activation. Overall, these results support a model in which the cellular transcription factor C/EBPalpha and RTA:C/EBPalpha interactions play important roles both upstream and downstream of the two major KSHV regulatory proteins RTA and RAP during the early stages of lytic cycle reactivation. Keywords: Introduction : Latent infection with Kaposi's sarcoma-associated herpesvirus (KSHV) is detected in the spindle cells of all forms of KS , as well as in plasmablast-like cells of AIDS-associated multicentric Castleman's disease and primary effusion lymphoma (PEL) . Genome analysis revealed that KSHV belongs to the gamma-2 herpesvirus subfamily and is related to Epstein-Barr virus (EBV) . Based on the kinetics of gene expression after viral reactivation, KSHV-encoded genes are classified into four broad categories: latent, immediate-early (IE), delayed-early (DE), and late. Several PEL cell lines carrying multicopy KSHV episomes have been established , and virus derived from them can convert dermal microvascular endothelial cells (DMVECs) into KS-like spindle cells . Only latent-state KSHV proteins, including LANA1, v-CYC-D, v-FLIP, and K15 (or LAMP), are expressed in the majority of PEL cells, KS spindle cells, and infected DMVECs . However, treatment with tetradecanoyl phorbol acetate (TPA) or sodium butyrate can disrupt the latency of KSHV infection, especially in PEL cells, and induce lytic viral replication in a subset of the cells . Induction of several transcripts, including those from the major RTA (or ORF50), ORF45, and ORF-K4.2 promoters as well as a minor RAP (or K8) promoter of KSHV have been detected within 4 h after lytic cycle induction in the presence of cycloheximide, and therefore these genes have been defined as the likely IE regulatory genes of KSHV . The KSHV-encoded RTA protein is generally considered to function as the principal molecular switch converting infected cells from a latent state into the lytic cycle by specific transcriptional activation of many downstream KSHV promoters . The 120-kDa RTA protein (691 amino acids [aa]) is a homologue of the EBV DNA-binding transactivator RTA (or BRLF1), which both play crucial roles as the triggers of complete lytic reactivation from latency. KSHV RTA transactivates various downstream KSHV lytic cycle promoters, including those for RAP (a positional homologue of EBV ZTA), KSHV MTA (or ORF57 a homologue of EBV MTA), vIL6, vMIP (or vCCL), K12 (or Kaposin), vOX2/GPCR, and PAN (polyadenylated nuclear) (or T1.1) RNA (, -, , , , ). In at least some target promoters (e.g., PAN, K12, and vIL6), RTA clearly functions by direct DNA binding to specific type II RTA-responsive elements (RREs) . Although Lukac et al. reported that baculovirus-expressed RTA also binds to the MTA and RAP RRE motifs in electrophoretic mobility shift assay (EMSA) experiments, we observed that in vitro-translated RTA does not bind directly to the type I RAP RRE . Instead, RTA activates RAP expression through an interaction with CCAAT/enhancer-binding protein-alpha (C/EBPalpha), which itself directly binds to the RRE site and activates the RAP promoter . Furthermore, a strong physical interaction was demonstrated between C/EBPalpha and RTA that requires basic residues in the RTA DNA-binding domain (DBD) between aa 151 and 273. RTA has also been reported to activate target gene expression at another type I RRE within the MTA promoter via a functional interaction with the cellular CBF1 (or RBP-Jkappa) DNA binding factor , a target of the Notch signaling pathway that also plays an important role in EBNA2-mediated functions during EBV latency . Previous studies have suggested that KSHV RTA positively autoregulates its own promoter in reporter gene cotransfection assays in 293 cells . In addition, a consensus motif at position -220/-213 in the RTA promoter was shown to bind to the OCT1 protein and was interpreted to be important for RTA-mediated autoregulation in 293L cells . However, because of our evidence that KSHV RTA physically interacts with C/EBPalpha and the predicted presence of C/EBP sites in the RTA promoter, we wished to evaluate other potential mechanisms for controlling expression of the RTA lytic cycle trigger protein. C/EBPalpha was the first identified member of the leucine zipper family of transcription factors, which includes c-JUN, c-FOS, ATF, CREB, and EBV ZTA . Members of this family bind to specific DNA elements as homodimers or heterodimers through the conserved bZIP domain located at the C terminus and modulate gene expression through an N-terminal activation domain. Numerous studies have shown that C/EBPalpha promotes differentiation and inhibits cell proliferation through multiple molecular pathways leading to G1 cell cycle arrest, including elevating the level of p21CIP-1/WAF-1 protein by both directly binding to and activating the p21CIP-1 promoter and stabilizing p21 through a protein-protein interaction . C/EBPalpha also inhibits E2F transcription of genes, including c-MYC, which is involved in DNA synthesis and mitosis through direct interaction with retinoblastoma (Rb) and Rb-like proteins , as well as directly inhibiting CDK2 and CDK4 . The 35-kDa KSHV-encoded RAP protein (237 aa) is also a bZIP-like protein that targets to nuclear PML oncogenic domains and seems to play an important role in lytic cycle viral DNA replication . Unlike its positional homologue ZTA, the major lytic cycle trigger protein encoded by EBV, RAP alone does not bind to any known downstream KSHV target gene promoters, nor does it act alone as a trigger for KSHV lytic cycle induction . However, both RAP and ZTA display similar properties in upregulating levels of the cellular C/EBPalpha and p21CIP-1 proteins, resulting in G0/G1 cell cycle arrest during progression of the KSHV or EBV lytic cycles, respectively . The mechanism of C/EBPalpha induction was found to involve strong protein-protein interactions between C/EBPalpha and RAP or C/EBPalpha and ZTA, leading to both stabilization of C/EBPalpha protein and cooperative transcriptional upregulation of the cellular C/EBPalpha and p21 promoters . C/EBPalpha also proved to reciprocally activate viral RAP and ZTA gene expression by directly binding to strong promoter proximal C/EBP binding sites (, ; F. Y. Wu et al., unpublished data). In the present study, we have asked whether C/EBPalpha might also play a role in activating KSHV IE gene expression and have used EMSA and mutation analysis to identify three C/EBPalpha binding sites within the proximal 914-bp RTA promoter. Cotransfection of a series of deletion mutant or mutated RTA-LUC target reporter genes in several cell types was used to evaluate both direct and cooperative responsiveness to C/EBPalpha, RAP, and RTA. A chromatin immunoprecipitation (ChIP) assay was employed to investigate whether C/EBPalpha, RAP, and RTA associate in vivo with the RTA promoter in KSHV lytically infected PEL cells. We then also tested the ability of C/EBPalpha to upregulate endogenous RTA in PEL cells and possible cooperative effects of RTA plus C/EBPalpha in regulating the cellular C/EBPalpha promoter. The kinetics of induction of C/EBPalpha mRNA and protein levels in TPA-induced PEL cells were also examined. Finally, we searched for C/EBP binding sites and investigated C/EBPalpha responsiveness, C/EBPalpha plus RTA cooperativity, and in vivo ChIP associations in two other downstream KSHV DE promoters that are associated with the PAN and MTA genes and have been shown to respond strongly to RTA. MATERIALS AND METHODS : Cells and plasmids. | HeLa and 293T cells were grown in Dulbecco's modified Eagle's medium (Invitrogen, Carlsbad, Calif.) containing 10% fetal bovine serum in a humidified 5% CO2 incubator at 37C. KSHV-positive human PEL cells BCBL1 cells, and JSC1 cells, as well as the KSHV-negative DG75 B-lymphoblast cell line, were grown in RPMI 1640 medium (Invitrogen) containing 10% fetal bovine serum. For KSHV lytic cycle induction, TPA was added to the medium at a final concentration of 20 ng/ml. Plasmids pSEW-C01, pSEW-C02, pJX15, and pCTC581a expressing full-length human C/EBPalpha, Flag-tagged C/EBPalpha, KSHV RTA, and Flag-tagged KSHV RAP effector genes driven by CMV or SV2 promoter-enhancers were described previously . Plasmid pSEW-R03 encodes the RTA(1-377) DBD segment; plasmid pSEW-R23 encodes RTA(1-691Delta151-167) with an in-frame internal deletion from position 151 to 167 . Reporter plasmid pGL3-FL(-914) containing the intact RTA(-914/+34)-LUC reporter gene was kindly provided by Koichi Yamanishi . Deletion derivatives of the full-length RTA-LUC reporter gene were generated by PCR using RTA(-914/+34)-LUC as the template to generate RTA(-587/+20)-LUC (in pSEW-RP1), RTA(-554/+20)-LUC (pSEW-RP2), RTA(-261/+20)-LUC (pSEW-RP3), RTA(-241/+20)-LUC (pSEW-RP4), and RTA(-212/+20)-LUC (pSEW-RP5) or by PCR-based mutagenesis using pSEW-RP4 or pSEW-RP1 as the templates to generate mutants RTA(-241/+20 pm-48/-43)-LUC (pSEW-RP6) or RTA(-587/+20 pm-48/-43)-LUC (pSEW-RP7), in which GCAATG at position -48 to -43 has been changed to GATATC (throughout this work, mutated nucleotides are indicated in boldface type). A plasmid encoding the human C/EBPalpha(-500/+10)-LUC reporter gene (pSEW-CP1) contains the proximal C/EBPalpha promoter region inserted into a minimal promoter in the pGL3-Basic background, as described previously . The PAN(-210/+15)-LUC reporter gene contains the wild-type PAN promoter inserted into the pGL3-Basic background (in plasmid pSEW-PP1) and point mutant derivatives PAN(-190/+15)-LUC (pSEW-PP2), PAN(-190/+15pm2A)-LUC (pSEW-PP3), PAN(-190/+15pm2B)-LUC (pSEW-PP4), PAN(-190/+15pm4A)-LUC (pSEW-PP5), and PAN(-190/+15pm4B)-LUC (pSEW-PP6) (see Fig. ) Point mutant reporter gene PAN(-190/+15RREm)-LUC (in plasmid pSEW-PP7) contains five point mutations within the RRE region at position -61 to -52, (GCTAACCTGT to TCTACCCGTG). A set of MTA promoter reporter plasmids expressing MTA(-160/+10)-LUC, MTA(-100/+10)-LUC, and MTA(-60/+10)-LUC in pSEW-MP1, pSEW-MP2, and pSEW-MP3, respectively, were generated by PCR amplification from BCBL1 DNA. Point mutant derivative pSEW-MP4 expressing MTA(-100/+10 pm -94/-90)-LUC in which ACAAT at position -94 to -90 has been changed to TCTAG was generated by a PCR-based method using pSEW-MP2 as the template. DNA transfection and luciferase assay. | Transfection of HeLa cells was performed with 5 x 105 cells/sample in six-well plates using Lipofectamine (Invitrogen) according to the Invitrogen protocol. BCBL1 and DG75 B-lymphoblast cell lines (107 cells/sample) were transfected using the electroporation method described previously . Transfected cells were harvested at 48 h posttransfection for luciferase assays as described by Wang et al. . All samples also received added CMV-betaGAL expression plasmid DNA as an internal control for transfection variability, and the betaGAL activity was found not to be significantly affected by cotransfection with either C/EBPalpha, RTA, or RAP or combinations of them. Extraction of mRNA and RT-PCR. | BCBL1 cells were transfected by electroporation, and mRNA was extracted at 40 h posttransfection using a GenElute Direct mRNA Miniprep kit (Sigma) according to the product instructions as described previously . The purified mRNA samples (1 mug) were incubated with 1 U of DNase I (Invitrogen) and 10 mul of DNase I reaction buffer for 15 min at 20C. Then, DNase I was inactivated by adding 1 mul of 25 mM EDTA to the mixture and heating for 10 min at 65C. Reverse transcription (RT) was performed using avian myeloblastosis virus reverse transcriptase (Promega) as previously described , and the synthesized cDNA samples were used as templates for PCRs using RTA coding region primers LGH4289 (5'-GAACTACTCGAGATGGTCACTGGACTGTCCTATCC-3') and LGH4290 (5'-TGACGAAGCTTCAGGTACCAGGTGTCGTGGTCG-3'). The cDNA PCR products (915 bp) were analyzed on a 1.2% agarose gel. IFA. | Indirect immunofluorescence assay (IFA) was performed 24 h after BCBL1 cells were transfected or induced by TPA using the methanol fixation method. Primary antibodies included rabbit antipeptide antiserum against KSHV RTA , mouse anti-Flag monoclonal antibody (MAb) (Sigma, St. Louis, Mo.), and goat (Santa Cruz, Santa Cruz, Calif.) or rabbit anti-C/EBPalpha polyclonal antibody (PAb). Secondary donkey- or goat-derived fluorescein isothiocyanate (FITC)- or rhodamine-conjugated anti-rabbit or anti-mouse or anti-goat immunoglobulin G (Jackson Pharmaceuticals, West Grove, Pa.) was used to detect the primary antibodies. Mounting solution with DAPI (4',6'-diamidino-2-phenylindole) (Vector Shield) was used to visualize cellular DNA. EMSA. | C/EBPalpha protein samples used for EMSA were in vitro translated using the TNT Quick Coupled Transcription/Translation system (Promega) according to the manufacturer's procedures, and pBBV-C/EBPalpha plasmid DNA was used as the template. Annealed double-strand oligonucleotides were radiolabeled with [alpha-32P]dCTP by Klenow DNA polymerase. For EMSA, approximately 50,000 cpm of the 32P-labeled probe and 2 mul of in vitro-translated C/EBPalpha protein were incubated for 30 min at 20C in a binding system containing 10 mM HEPES (pH 7.5), 50 mM KCl, 1 mM EDTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 1% Triton X-100, 5% glycerol, and 2 mug of poly(dI-dC). For supershift experiments, 0.5 mul of C/EBPalpha rabbit antiserum or RTA rabbit antiserum against an RTA epitope between aa 527 and 539 was added to the mixture after 30 min and allowed to incubate for another 30 min before gel loading. Samples were separated on a 4.5% polyacrylamide gel in 1x HEE buffer (10 mM HEPES [pH 7.5], 1 mM EDTA, 0.5 mM EGTA) at 150 V at 4C as described previously . The gels were dried and subjected to autoradiography with Kodak X-ray film. All oligonucleotides used were purchased from Qiagen Operon. The following RTA promoter probes were used: LGH4341 (5'-GATCTGGCGCTACAGTGGGTGATTTCTTCTACCACGG-3') and LGH4342 (5'-GATCCCGTGGTAGAAGAAATCACCCACTGTAGCGCCA-3') annealed to form probe RP-1; LGH4343 (5'-GATCTCATACATTGGTGGCACCCACAGGCCTGTTCCA-3') and LGH4344 (5'-GATCTGGAACAGGCCTGTGGGTGCCACCAATGTATGA-3') annealed to form probe RP-2; LGH4345 (5'-GATCGTGGTACCGAATGCCACAATCTGTGCCCTCCAG-3') and LGH4346 (5'-GATCCTGGAGGGCACAGATTGTGGCATTCGGTACCAC-3') annealed to form probe RP-3; LGH4347 (5'-GATCACAATTTTCATCTCCAATACCCGGAATTGGGAT-3') and LGH4348 (5'-GATCATCCCAATTCCGGGTATTGGAGATGAAAATTGT-3') annealed to form probe RP-4; LGH4349 (5'-GATCCCAGAAACCAGTAGCTGGGTGGCAATGACACGT-3') and LGH4350 (5'-GATCACGTGTCATTGCCACCCAGCTACTGGTTTCTGG-3') annealed to form probe RP-5; LGH4363 (5'-GATCCCAGAAACCAGTAGCTGGGTGGATATCACACGT-3') and LGH4364 (5'-GATCACGTGTGATATCCACCCAGCTACTGGTTTCTGG-3') annealed to form probe RP-5M. The entire PAN promoter insert from PAN(-210/+15)-LUC was isolated by cleavage with SacI/HindIII and end labeled to obtain probe PAN-225. The PAN-1 probe was generated by annealing oligonucleotides LGH4971 (5'-GATCATTAATGAAAGTTTATTAATGTTCATCCGT-3') and LGH4972 (5'-GATCACGGATGAACATTAATAAACTTTCATTAAT-3'); probes PAN-2, PAN-2A, and PAN-2B were generated by annealing LGH4973 and LGH4974, LGH4990 and LGH4991, and LGH4992 and LGH4993, respectively, which contain an additional GATC at the 5' ends (see Fig. ); probe PAN-3 was generated by annealing LGH4975 (5'-GATCATGGAGTTTTCTTATGGATTATTAAGGGTCAGCTTGAAGG-3') and LGH4976 (5'-GATCCCTTCAAGCTGACCCTTAATAATCCATAAGAAAACTCCAT-3'); probes PAN-4, PAN-4A, and PAN-4B were generated by annealing LGH4977 and LGH4978, LGH4994 and LGH4995, and LGH4996 and LGH4997, respectively, which contain an additional GATC at the 5' ends (see Fig. ). Probe PAN-RRE (41 bp) was described previously . Probe MTA-1 was generated by annealing LGH4980 (5'-GATCCTTCATTCCATTAGGGTGAGCGAAGTCACGGTAACACTTATGA-3') and LGH4981 (5'-GATCTCATAAGTGTTACCGTGACTTCGCTCACCCTAATGGAATGAAG-3'); probe MTA-2 was generated by annealing LGH4982 (5'-GATCGTCAGTGTTTTGCCAGCAAGTGTAACAATAATGTTCCCACGGC-3') and LGH4983 (5'-GATCGCCGTGGGAACATTATTGTTACACTTGCTGGCAAAACACTGAC-3'); probe MTA-3 was generated by annealing LGH4984 (5'-GATCCCATTTTTCGTTTGTGGTACCTGTGGGACTGGCCAGTTAATCC-3') and LGH4985 (5'-GATCGGATTAACTGGCCAGTCCCACTGGTACCACAAACGAAAAATGG-3'); probe MTA-2M was generated by annealing LGH5236 (5'-GATCGTCAGTGTTTTGCCAGCAAGTGTATCTAGAATGTTCCCACGGC-3') and LGH5237 (5'-GATCGCCGTGGGAACATTCTAGATACACTTGCTGGCAAAACACTGAC-3'). ChIP. | JSC1 or BCBL1 PEL cells were treated with TPA before harvesting at different time points for ChIP assays as described elsewhere using rabbit PAbs against C/EBPalpha, KSHV RTA, or KSHV RAP or mouse MAb against EBV ZTA or mouse MAb against CHOP10 as controls. Primers LGH4354 (5'-GAACTACTCGAGCTGTGCCCTCCAGCTCTCAC-3') and LGH4355 (5'-GGACGTAAGCTTACAGTATTCTCACAACAGAC-3'), specific for a 261-bp region in the KSHV RTA promoter from -241 to +20; primers LGH4973 (described for EMSA) and LGH4998 (5'-TGATCTAAGCTTCTGGGCAGTCCCAGTGCT-3'), specific for a 205-bp region in the KSHV PAN promoter from -190 to + 15; and primers LGH4980 (described for EMSA) and LGH4987 (5'-GATCGCGGGCTATTTTGGGAACCTGGCAGCCAGGTTATATAGTG-3'), specific for a 168-bp region in the KSHV MTA promoter from -160 to +8, were used for PCR detection. Primers LGH4930 (5'-TTCGCCTGTTAGACGAAGC-3') and LGH4929 (5'-GATTCGCAAGCTTCAGTCTCGGAAGTAATTACG-3'), specific for RTA coding region aa 591 to 691, were used as negative control to detect a nonpromoter region. The PCR products were analyzed on a 2% agarose gel. Extraction of total RNA and Northern blot assay. | BCBL1 cells were induced with TPA, and total RNA was isolated at different time points using Trizol reagent (GIBCO BRL) according to the manufacturer's protocol. RNA samples (30 mug) were separated on a 1.2% formaldehyde agarose gel in 1x morpholinopropanesulfonic acid (MOPS) running buffer and transferred to a Nytran nylon membrane (Schleicher & Schuell). Restriction enzyme-released fragments containing C/EBPalpha (aa 251 to 358), KSHV RTA (aa 244 to 548) or KSHV RAP (aa 1 to 237), as well as a PCR-generated fragment containing a 286-bp GAPDH (glyceraldehyde-3-phosphate dehydrogenase) cDNA region, were labeled using the Prime-a-Gene labeling system (Promega) in a 50-mul reaction mixture containing 1x labeling buffer, 20 muM concentrations of dATP, dGTP, and dTTP, 25 ng of denatured DNA template, 20 mug of bovine serum albumin, 333 nM [alpha-32P]dCTP(3,000 Ci/mmol), and 5 U of Klenow DNA polymerase. After end labeling at 20C for 60 min, the probes were purified with Sephadex G-50, denatured, and incubated with the membrane in a hybridization mixture (5 ml) containing 50% formamide, 5x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate), 1x Denhardt's solution, 20 mM NaPO4 (pH 6.5), 10% dextran sulfate, and calf thymus DNA (100 mug/ml). Sequential hybridizations were carried out overnight at 42C, and the membrane was washed twice in 2x SSC-0.1% sodium dodecyl sulfate (SDS) at room temperature and then twice in 0.5x SSC-0.1% SDS at 50C before autoradiography and reuse. For an RNA loading control, the membrane was stripped again and reprobed with GAPDH cDNA. Protein isolation and Western immunoblot assay. | Approximately 107 BCBL1 cells or 106 293T cells were harvested at different time points after TPA induction or at 48 h posttransfection, respectively; washed one time with phosphate-buffered saline; gently resuspended in 0.2 ml of ice-cold immunoprecipitation buffer (50 mM Tris-Cl [pH 7.9], 50 mM NaCl, 0.1 mM EDTA, 1% glycerol, 0.2% NP-40, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride); and sonicated for 30 s. The samples were then centrifuged at 15,000 x g for 5 min at 4C, and the supernatants were collected. Western blot analysis was performed after SDS-polyacrylamide gel electrophoresis, using 10 mul of each protein sample. The proteins were detected by immunoblotting with rabbit polyclonal antiserum against C/EBPalpha, KSHV RTA, or KSHV RAP or mouse MAb against Flag as described previously . RESULTS : C/EBPalpha activates the RTA IE promoter in cooperation with KSHV RTA and RAP. | Previous reports indicated that cotransfected RTA (ORF50) autostimulates expression from a target RTA promoter by 40-fold in 293T cells using a large RTA(-3000/+20)-LUC reporter gene or by 6-fold in 293L cells using a shorter RTA(-914/+34)-LUC reporter gene . In the latter study, OCT1 binding to the RTA promoter was suggested to be important for RTA autoactivation, although OCT1 had little effect on RTA expression when transfected alone. We suspected that C/EBPalpha might also be involved in the efficient activation of RTA gene expression during the KSHV lytic cycle. Exogenously introduced C/EBPalpha activates expression of the KSHV DE gene RAP (K8) protein in PELs and cooperates with both RTA and the RAP protein itself to stimulate the RAP promoter in transient-cotransfection assays . Those studies also revealed that the C/EBPalpha and RTA proteins physically interact with each other (as well as with RAP), but whether these two transcription factors regulate each other's expression at the transcriptional level was not known. Initially, we transfected the RTA(-914/+34)-LUC target reporter gene into HeLa cells and tested the effects of cotransfected C/EBPalpha, RTA, and RAP effector genes on RTA promoter expression. Only 2.3-fold autoactivation by RTA alone was observed. However, C/EBPalpha alone gave a 15-fold stimulation of RTA-LUC activity, and addition of RTA and C/EBPalpha together resulted in further enhancement up to 31-fold . Although RAP had no effect on RTA promoter activity when transfected alone (1.1-fold), cotransfection of RAP and RTA together resulted in up to 14-fold activation in the absence of C/EBPalpha. When cotransfected with C/EBPalpha, RAP also elevated the level of transactivation by a factor of 2 up to 30-fold. To confirm these observations about transactivation of the RTA promoter in the physiologically more relevant B cells, the cotransfection experiments were also carried out in the KSHV- and EBV-negative DG75 lymphoblast cell line by electroporation . C/EBPalpha alone induced RTA-LUC expression up to 8.1-fold (10 mug), but again, although RTA only autoactivated its promoter by 2.1-fold (10 mug), it more than additively increased the activation by C/EBPalpha, from 6.5-fold up to 14-fold at the 5-mug input DNA level. Addition of RAP also enhanced both C/EBPalpha and RTA activation levels nearly twofold. Finally, when C/EBPalpha, RTA, and RAP were all added together to the transfection (5 mug each), the strongest level of activation was obtained (18-fold). Therefore, C/EBPalpha proved to be the most significant activator of the RTA promoter discovered so far, and just as for the RAP promoter, KSHV RTA and RAP cooperated with C/EBPalpha to achieve the highest level of activation. Transfection by C/EBPalpha increases the level of RTA mRNA in PEL cells. | The cotransfection data above strongly suggest that C/EBPalpha is capable of upregulating KSHV RTA expression at the transcriptional level. To ask whether this effect can be demonstrated to occur in vivo in KSHV-infected PEL cells, we performed RT-PCR in C/EBPalpha-transfected BCBL1 cells. The mRNA samples were isolated and treated with DNase I to eliminate genomic DNA contamination before RT was carried out. After cDNA synthesis, two primers specific for the RTA gene were used for PCR amplification, giving an expected product of 915 bp In empty vector-transfected control cells, the level of RTA mRNA detected was very low (Fig. , lane 2); however, TPA treatment for 24 h resulted in the activation of KSHV lytic replication in 15% of the cells (as detected by IFA for RAP) and led to a 12-fold increase in the level of RTA mRNA (Fig. , lane 3). In BCBL1 cells electroporated with C/EBPalpha expression plasmid DNA, the RTA mRNA level increased by 6.7-fold compared to empty vector-transfected cells, whereas there was no effect on RTA expression in RAP-transfected cells (Fig. , lanes 4 and 5). In BCBL1 cells cotransfected with both C/EBPalpha and RAP, there was a small additive increase in the expression of RTA mRNA (7.8-fold) compared to cells transfected with C/EBPalpha alone (Fig. , lane 6). Authentication of the expected size of the RT-PCR product was provided using RNA from BCBL1 cells transfected with an RTA cDNA expression plasmid (Fig. , lane 7); however, any effect of RTA on autoactivation here was unable to be detected because both endogenous RTA expression and the transfected RTA plasmid give RT-PCR products of the same size. Introduction of C/EBPalpha induces endogenous RTA protein expression in PEL cells. | To confirm that RTA protein expression can be induced by C/EBPalpha, double-label IFA experiments were performed with CMV-Flag-C/EBPalpha transfected latently infected BCBL1 cells. Expression of endogenous RTA protein was absent in empty vector-transfected control cells (Fig. to C); however, in C/EBPalpha-transfected cells, 37% of the cells that expressed exogenous Flag-tagged C/EBPalpha were also now positive for RTA (Fig. to F). In comparison, transfection with SV2-Flag-RAP did not induce RTA protein expression in any of the RAP-positive cells as detected by anti-Flag MAb and anti-RTA PAb (Fig. G to I). Interestingly, in BCBL1 cells doubly transfected with CMV-C/EBPalpha and SV2-Flag-RAP, 51% of RAP-positive cells became positive for RTA protein expression. Additional IFA experiments confirmed that more than 95% of cells expressing either one of the transfected effector plasmids in the double-transfection experiments were positive for expression of both proteins (data not shown). Therefore, expression of exogenous C/EBPalpha alone can activate RTA expression from a latent state in some cells and RAP can evidently augment C/EBPalpha-mediated transactivation of the RTA promoter, although it has no effect alone. Identification of three C/EBP binding sites in the RTA promoter. | DNA sequence analysis revealed five potential C/EBP binding sites within the RTA promoter region between positions -914 and +20 . Complementary oligonucleotide pairs (37 nucleotides long) containing each of these putative sites (designated as RP-1 to RP-5) were synthesized, annealed, and 32P-labeled as probes for EMSA to test their affinity for C/EBPalpha protein. Strong EMSA bands were observed with probes RP-1 and RP-3 when in vitro-translated C/EBPalpha protein was added (Fig. , lanes 2 and 8), and the addition of antibody against C/EBPalpha resulted in the formation of typical slower-migrating supershifted bands (Fig. , lanes 3 and 9). Low-affinity binding to the C/EBPalpha protein was also observed with probe RP-5 (Fig. , lanes 14 and 15), but neither probe RP-2 nor RP-4 showed any affinity for C/EBPalpha, and a triple-point mutation introduced into RP-5 M destroyed its ability to bind C/EBPalpha (Fig. , lanes 17 and 18). The three putative C/EBP sites that proved to interact with C/EBPalpha in vitro were designated C/EBP-I (within probe RP-1), C/EBP-II (within probe RP-3), and C/EBP-III (within probe RP-5). To examine the contribution of these sites to the transactivation of the RTA promoter by C/EBPalpha, we constructed a series of deletion mutants with 5' boundaries at positions -587, -554, -261, -241, and -212 that sequentially removed the C/EBP-I and C/EBP-II sites from the full-length RTA(-914/+34)-LUC reporter gene, as illustrated in Fig. . In addition, the same mutated version of C/EBP-III that was generated in RP-5M was introduced into both RTA(-587/+20)-LUC and RTA(-241/+20)-LUC. The responsiveness of each of these truncated RTA reporter genes to cotransfection with increasing doses of C/EBPalpha effector plasmid DNA was measured, and the results are summarized in the histogram shown in Fig. . Basal activity across the series did not change by more than 50% and is not included in the data. Only the mutant version of RTA(-241/+20)-LUC without any of the three identified C/EBP sites remaining showed a fourfold reduction, with all of the others giving less than twofold reduction in responsiveness to C/EBPalpha alone. Therefore, evidently both the C/EBP-II and C/EBP-III sites alone are sufficient for direct C/EBPalpha activation. We also examined the role of the C/EBP binding sites in cooperative activation by C/EBPalpha plus RTA and RAP. Deletion of C/EBP-I in RTA(-554/+20)-LUC led to partial impairment of enhanced C/EBPalpha activation by both RTA and RAP, whereas when C/EBP-II was also deleted in RTA(-241/+20)-LUC, cooperativity between RTA and RAP with C/EBPalpha was nearly abolished . However, although mutation of C/EBP-III in RTA(-241/+20 pm-48/-43)-LUC significantly impaired direct C/EBPalpha activation, the cooperative enhancement by RTA and RAP was not affected. Therefore, either one or both of the C/EBP-I and -II sites that bind strongly to C/EBPalpha protein in EMSA appear to be required for the cooperative activation by C/EBPalpha, RTA, and RAP, with C/EBP-II alone being sufficient and C/EBP-III not contributing. In parallel experiments, neither deletion nor point mutation of the OCT1 site (-220/-213) described by others had any significant effect on basal activity or on C/EBPalpha or RTA transactivation in our studies in HeLa or DG75 cells (data not shown). A recent study by Saveliev et al. suggests that the KSHV RTA promoter region may initiate a second set of IE transcripts in a bidirectional fashion from near position -100 and that in addition a minor IE version of RAP mRNA may initiate from position -250 in the KSHV RAP promoter. If so, the C/EBP-III site described here would still represent an upstream regulator for both the spliced RTA (ORF50) and ORF48/19 IE transcripts, but C/EBP-I and -II would be downstream and possible negative regulators of the ORF48/29 leftward spliced IE transcript. Similarly, the overlapping C/EBP-RRE motif in the RAP promoter would be downstream of the minor IE RAP transcription start site and would likely have negative regulatory effects compared to the positive effects observed on the predominant DE version of the RAP promoter that we worked with previously . C/EBPalpha, RTA, and RAP all associate with the RTA promoter in vivo by ChIP assay. | ChIP assays provide a powerful method to investigate the association between proteins and specific DNA regions in vivo. For example, we showed previously that addition of either in vitro-translated intact RTA or RAP proteins to C/EBPalpha in EMSA experiments interferes with the C/EBPalpha:DNA-shifted bands but fails to produce a supershifted band . That result suggested but did not prove that RTA and RAP can piggyback bind onto the DNA-bound C/EBPalpha protein. However, the use of chromatin immunoprecipitation assays with PEL cell extracts confirmed that both RTA and RAP do indeed associate with the RAP promoter in a C/EBPalpha-dependent fashion in vivo, despite their inability to bind to it directly in vitro . To conduct a similar experiment with the RTA promoter, we incubated a cross-linked cell extract obtained from KSHV-infected JSC1 cells after 40 h of TPA treatment with antibodies against C/EBPalpha, RTA, or RAP followed by immunoprecipitation. The presence of RTA-promoter DNA fragments within the immunoprecipitated samples was detected by PCR amplification using specific primers encompassing a 261-bp region in the RTA promoter . The recovery of PCR products of this size on an agarose gel revealed that all three antibodies were able to precipitate complexes containing RTA-promoter DNA fragments (Fig. , upper panel, lanes 2 to 4). In contrast, antibody against EBV ZTA or a negative control using no antibody failed to recover any RTA-promoter DNA (lanes 5 and 6). Importantly, these same immunoprecipitates did not display any preferential association with nonpromoter DNA as judged by parallel PCR analysis with RTA-coding region primers (Fig. , lower panel). Both KSHV RTA and RAP have been shown to physically interact with C/EBPalpha by immunoprecipitation and in vitro glutathione S-transferase binding assays , but neither bound directly to any of the identified C/EBP site probes from the RTA promoter in EMSA experiments (data not shown). Therefore, to investigate whether the association of RTA and RAP with the RTA promoter depends on indirect interaction with C/EBPalpha bound to C/EBP binding sites, we next precleared the cell extract with C/EBPalpha antibody before performing the ChIP assay on immunoprecipitates using antibodies against RTA or RAP. After removing C/EBPalpha, neither of these two antibodies was able to precipitate any RTA promoter region target DNA (Fig. , lanes 8 and 9). However, C/EBPalpha antibody was still able to recover RTA promoter DNA after preclearing with antibody against RTA or RAP (Fig. , lanes 10 and 11). This suggests a model involving the formation of transcriptional transactivating complexes on the RTA promoter in which C/EBPalpha binds directly to its target sites, whereas KSHV RTA and RAP only associate more indirectly with this promoter mediated through piggyback protein-protein interactions with C/EBPalpha. Expression of C/EBPalpha is induced during the early phase of KSHV reactivation. | The KSHV IE gene RTA is likely to be the first viral gene expressed after primary infection as well as during reactivation from latency. We have shown above that exogenous C/EBPalpha can activate RTA expression both in latently infected PEL cells and in cotransfection reporter gene assays; however, whether C/EBPalpha is a factor that naturally triggers RTA early expression or vice versa remains to be determined. To investigate the kinetics of C/EBPalpha gene expression compared with that of RTA and RAP during KSHV reactivation in PEL cells, we performed a Northern blot experiment using total RNA samples extracted from BCBL1 cells treated with TPA for 0, 3, 6, 24, 33, or 48 h. A 32P-labeled C/EBPalpha DNA probe was used to detect the 2.5-kb C/EBPalpha mRNA. Expression of C/EBPalpha mRNA was limited in uninduced BCBL1 cells (TPA 0 h) but was elevated by 3 h after TPA induction, when the 3.2-kb RTA mRNA was also first observed, and it appeared to reach a peak at 24 h somewhat earlier than RTA . In contrast, although a low level of the 1.1-kb RAP mRNA was induced at 3 h, the major stimulation occurred significantly later and peaked between 33 and 48 h. To evaluate the time course of C/EBPalpha induction at the protein level, a Western blot experiment was carried out using BCBL1 lysates from cultures treated with TPA for 0, 0.5, 1, 2, 3, and 12 h . An increase in both C/EBPalpha (42 and 30-kDa) and RTA (120-kDa) protein levels was first observed at 3 h, whereas the RAP (35-kDa) protein could not be detected until 12 h. Therefore, evidence at both the RNA and protein levels indicated that expression of C/EBPalpha is elevated at the very early stages of KSHV reactivation in PEL cells in parallel with RTA and even before RAP is synthesized. To address how many and which PEL cells were induced to express endogenous C/EBPalpha and KSHV RTA proteins, double-label IFA experiments were carried out in BCBL1 cells after 24 h in the presence or absence of TPA treatment. Neither protein was detectable in more than 1% of the cells before induction, but both became positive in approximately 15% of the cells after induction. Importantly, 79% of the RTA-positive cells at 24 h after TPA induction were also positive for C/EBPalpha, and 82% of the C/EBPalpha-positive cells were also positive for RTA . Again, we emphasize that TPA treatment did not induce detectable C/EBPalpha protein levels in the majority of PEL cells that did not also proceed into the lytic cycle. The observed correlation in expression of both proteins in the same cells during the lytic cycle further supports our in vitro findings that C/EBPalpha can upregulate the RTA promoter and raised the question of whether RTA might in turn upregulate the C/EBPalpha promoter. The expression of C/EBPalpha is not induced by TPA in KSHV-negative DG75 cells, as judged by both IFA and Western immunoblotting . Therefore, virus-encoded factors seem likely be responsible for even the earliest induction of C/EBPalpha expression in TPA-treated BCBL1 cells. Because RTA and C/EBPalpha mRNA and proteins were both expressed as early as 3 h in TPA-induced BCBL1 cells, well before RAP was detected, we examined the effect of exogenous RTA expression on endogenous C/EBPalpha protein levels in uninfected DG75 lymphoblast cells. Western immunoblotting performed to detect the 42-kDa (and 30-kDa) forms of C/EBPalpha showed that (despite only a 20% transfection efficiency) their overall level of expression in the culture was indeed elevated severalfold by addition of the vector DNA expressing RTA (120 kDa) alone, and in fact reached the same level as that attained after addition of the C/EBPalpha expression vector alone . RTA stimulates activation of the C/EBPalpha promoter cooperatively with C/EBPalpha. | To investigate the mechanism by which RTA induces C/EBPalpha expression, we first tested the effect of cotransfected RTA on a target C/EBPalpha-LUC reporter gene in HeLa cells. RTA alone activated the C/EBPalpha promoter by only 2.7- (0.5 mug) to 3.2-fold (1 mug), compared to C/EBPalpha alone, which had an 18-fold effect . However, when RTA was cotransfected together with C/EBPalpha, the combined effect was elevated more than additively to more than 53-fold. In contrast, the RTA(Delta151-167) mutant, which is unable to interact with the C/EBPalpha protein , had little activity on its own and failed to enhance the activation by C/EBPalpha. Similarly, in DG75 lymphoblast cells, cotransfection with RTA, but not with RTA(Delta151-167), again enhanced C/EBPalpha activation from fourfold to nearly eightfold . These results suggest that RTA can in turn stimulate the C/EBPalpha promoter cooperatively with C/EBPalpha, presumably because of an increase in transcriptional activity by DNA-bound RTA:C/EBPalpha complexes. To address the possibility that the cooperative enhancement of C/EBPalpha transactivation by RTA might in addition involve an effective increased stabilization of the level of C/EBPalpha protein, we compared the amounts of each protein present in the single or combined transfected extracts . Because cotransfection with RTA had no effect on the high constitutive levels of CMV-betaGAL expression in similar experiments (data not shown), we anticipated that there would be no transcriptional effects on the CMV-C/EBPalpha target promoter used here. Nevertheless, the results clearly revealed that cotransfection with wild-type RTA(1-691), but not with mutant RTA(1-691Delta151-167), led to a threefold increase in Flag-C/EBPalpha (42-kDa) protein levels over those obtained with the C/EBPalpha expression plasmid alone. Therefore, the cooperativity of RTA:C/EBPalpha complexes, like that of RAP:C/EBPalpha complexes, may involve a stabilization effect on the relatively short half-lived C/EBPalpha protein, as well as a transcriptional component. C/EBPalpha transactivates the KSHV PAN promoter by specifically binding to two C/EBP sites. | RTA is also known to transactivate expression of the KSHV DE gene encoding PAN RNA by binding directly to a type II RRE in the proximal region of the PAN promoter . To investigate whether C/EBPalpha may also play a role in regulating the PAN promoter, the effects of cotransfected C/EBPalpha and RTA, either alone or in combination, were tested on a target PAN(-210/+15)-LUC reporter gene in HeLa cells. C/EBPalpha alone proved to activate the intact wild-type PAN target promoter up to 32-fold, whereas RTA alone produced up to 128-fold activation . Moreover, the combination of C/EBPalpha and RTA together led to a further dramatic increase in the activation level up to 500-fold. In comparison, the RTA(Delta151-167) deletion mutant, which has lost both its direct DNA binding activity and the ability to interact with C/EBPalpha , neither activated the PAN promoter when transfected alone nor cooperated with C/EBPalpha when cotransfected. In EMSA experiments, the in vitro-translated C/EBPalpha protein and either the intact RTA(1-691) protein or the RTA(1-377) DBD protein segment were all able to bind independently to a 225-bp PAN DNA probe encompassing the RRE . Evidently, in this case, both C/EBPalpha and RTA probably activate the PAN promoter through direct binding to their respective target recognition sites. However, additional more complex interactions remained plausible to explain the functional cooperativity demonstrated above. Therefore, to map the number and position of C/EBP target sites within the PAN promoter, we examined the C/EBPalpha binding ability of a series of overlapping 34- to 44-bp EMSA probes referred to as PAN-1 to -4 that span the whole 180-bp PAN promoter segment upstream of the TATA box. Two of these probes, PAN-2 and PAN-4, were able to form specific shifted complexes with the C/EBPalpha protein, but probes PAN-1, PAN-3, and the PAN-RRE probe did not do so . To further define the location of these C/EBP sites, we introduced multiple adjacent point mutations into several motifs that resembled either C/EBP or other known recognition motifs within the PAN-2 and PAN-4 probes, as indicated in Fig. . The EMSA results indicated that both the PAN-2A and PAN-4A probes lost the ability to bind to C/EBPalpha, whereas the mutations that inactivated putative SRE or SP1 sites within the PAN-2B and PAN-4B probes did not significantly affect C/EBPalpha binding . Therefore, the C/EBP-I site maps around position -186 to -171 and the C/EBP-II site maps around position -102 to -88. As expected, the in vitro-translated RTA(1-377) DBD protein fragment bound only to the PAN-RRE probe from position -77 to -37, which encompasses the previously defined type II RRE, but not to any of the other upstream PAN probes. Quantitation of the relative strengths of C/EBP-I and C/EBP-II showed that the C/EBP-I site had an affinity for C/EBPalpha two- to threefold higher than that for the C/EBP-II site (data not shown). When the mutated versions of the PAN-LUC reporter gene were used in cotransfection experiments , the C/EBP-I mutated PAN(-190/+15pm2A)-LUC gene showed an 8-fold decrease in C/EBPalpha responsiveness and 2.5-fold reduced RTA responsiveness relative to the wild-type version (panel 2), whereas the SRE mutated PAN(-190/+15pm2B)-LUC gene was unaffected (panel 3). However, the C/EBP-II mutated PAN(-190/+15pm4A)-LUC showed only a 30% decrease in both C/EBPalpha and RTA responsiveness (panel 4), and all three mutant probes retained similar twofold levels of RTA plus C/EBPalpha cooperativity . In contrast, although the SRE and SP1 point mutations in PAN(-190/+15pm4B)-LUC did not affect C/EBPalpha transactivation or RTA augmentation (panel 5), there was a fivefold decrease in direct RTA-responsiveness, presumably because the first base pair within the defined type II PAN-RRE motif was also altered . To evaluate the response to C/EBPalpha in the total absence of the direct RTA binding motif, we also performed the cotransfection experiments using a mutant reporter gene PAN(-190/+15RREm)-LUC, which contains the same multiple point mutations that completely abolished both RTA binding and RTA responsiveness of the PAN promoter as reported by Song et al. . Indeed, only 5-fold RTA-activation remained, but normal levels of C/EBPalpha activation (52-fold) were observed (panel 6). Importantly, this RRE-knocked-out version of the PAN target promoter still retained some C/EBPalpha plus RTA cooperativity (giving 126-fold activation). These results indicate that, although RTA alone is sufficient to activate the PAN promoter by binding directly to the proximal type II RRE motif, C/EBPalpha also activates the PAN promoter primarily through the C/EBP-I site. Furthermore, an additional level of cooperative transactivation also occurs that is mediated by RTA:C/EBPalpha complexes binding to one or both of the C/EBP-I and C/EBP-II sites independently of the RRE itself. The KSHV MTA promoter is also transactivated by C/EBPalpha through a C/EBP site adjacent to the RRE. | Both the KSHV MTA and RAP (K8) promoters have previously been reported to contain partially homologous versions of a different type of RRE referred to as a palindromic type I RRE . However, we showed previously that C/EBPalpha transactivates the KSHV RAP promoter through a strong C/EBP binding site at position -73 to-67 that overlaps with the identified type I RRE motif . Furthermore, RTA did not bind directly to the RAP type I RRE at all, and instead the activation was mediated by interaction with C/EBPalpha bound to this C/EBP site. Liang et al. have also recently shown that RTA binds indirectly to the MTA type I RRE via interaction with the cellular CBF1 (or RBP-Jkappa) protein. Nevertheless, we suspected that the MTA promoter might also use the same indirect mechanism for RTA regulation as does the RAP promoter, because a similar potential C/EBP-like motif at position -95/-87 also forms part of the core of the defined type I RRE sequence within the MTA promoter . To examine the C/EBPalpha binding activity of this site and to search for other potential C/EBP sites within the MTA promoter, we performed EMSA experiments using three 43-bp probes spanning the proximal 130-bp upstream MTA promoter region. Probe MTA-2 from position -116 to -74 encompassing the putative C/EBP site within the RRE proved to bind strongly to C/EBPalpha, whereas probe MTA-1 showed only relatively weak binding affinity for C/EBPalpha and probe MTA-3 failed to bind . Importantly, none of these probes, including that containing the RRE, bound directly to either the in vitro-translated intact RTA protein or to the RTA(1-377) DBD fragment . To further test the functions of these C/EBP sites, we carried out cotransfection experiments using a set of MTA-LUC reporter mutants with 5' truncations at positions -160, -100, and -60 . Reporter plasmid MTA(-160/+10)-LUC contains both of the C/EBP-I and C/EBP-II sites detected by EMSA, whereas MTA(-100/+10)-LUC retains only the strong C/EBP-II site within the RRE, and MTA(-60/+10)-LUC lacks both C/EBP sites as well as the overlapping type I RRE motif. C/EBPalpha alone proved to activate both MTA(-160/+10)-LUC and MTA(-100/+10)-LUC by 160-fold, whereas RTA alone gave 430-fold up-regulation. In both cases, C/EBPalpha and RTA together cooperated to achieve an enhanced activation of up to 1,700-fold . In contrast, the C/EBPalpha responsiveness of MTA(-60/+10)-LUC was reduced 5-fold and the RTA responsiveness was virtually abolished (down 100-fold). To further define the location of the strong C/EBP site within the EMSA probe MTA-2, we introduced multiple adjacent point mutations into the putative C/EBP-II site AACAATAAT from -95 to -87. The EMSA results indicated that probe MTA-2M containing the mutated C/EBP-II site had lost the ability to bind to C/EBPalpha . Cotransfection experiments using MTA(-100/+10pm-94/-90)-LUC, which contains the same mutated C/EBP-II site as in the probe MTA-2M, gave a threefold decrease in direct C/EBPalpha activation. Furthermore, it retained threefold-augmented C/EBPalpha plus RTA responsiveness, but it gave only a 30% decrease in direct RTA responsiveness . The surprisingly high residual responsiveness (30- to 50-fold) of both MTA(-60/+10)-LUC and MTA(-100/+10pm-94/-90)-LUC to C/EBPalpha alone implies that there may still be another more proximal as yet unidentified C/EBP site present within the MTA promoter (perhaps at position -40). Overall, the type I RRE overlapping with the C/EBP-II site in the MTA promoter shows some similar properties to that in the RAP promoter by being partially dependent upon RTA forming piggyback complexes with C/EBPalpha. However, unlike the RRE motif in the RAP promoter, the presumed interaction of RTA with CBF1 within the adjacent segment of the MTA RRE clearly produces a much greater effect. Association of C/EBPalpha, RTA, and RAP proteins with both IE and DE KSHV promoters during early lytic reactivation. | The above-mentioned experiments demonstrated that C/EBPalpha plays an important role in transactivating the KSHV RTA, PAN, and MTA promoters and that its own expression is elevated by RTA as early as 3 h after TPA treatment in BCBL1 cells. To investigate when C/EBPalpha, RTA, and RAP can first be detected associating with these three viral promoters during the early stages of lytic reactivation in infected cells, we carried out a time course ChIP assay experiment using BCBL1 cell lysates at 0, 4, 8, 12, and 24 h after TPA treatment. Antibodies against RTA, RAP, C/EBPalpha, or CHOP10 (negative control) were used to immunoprecipitate the associated DNA fragments . In the PCR amplification, DNA regions representing all three promoters (RTA, PAN, and MTA) were detected as early as 8 h after TPA treatment in all three (RTA, RAP, and C/EBPalpha) immunoprecipitates, but not in those obtained with CHOP10 antibody (Fig. to C), whereas no protein-bound DNA fragments were detected in the uninduced BCBL1 cell extracts (0 h) or in cell lysates induced by TPA for only 4 h (data not shown). At 24 h after TPA induction, increased amounts of promoter DNA were precipitated by the RTA, RAP, and C/EBPalpha antibodies, presumably reflecting increased expression of all three proteins, and the 12-h samples gave intermediate effects (data not shown). Interestingly, for the RTA and MTA promoters, C/EBPalpha was able to precipitate higher levels of promoter DNA than did RTA and RAP at both the 8- and 24-h time points (Fig. and C), whereas for the PAN promoter, RTA had the strongest association with the promoter DNA . This is consistent with the in vitro evidence that RTA strongly and directly binds to PAN-RRE but only indirectly associates with the RTA and MTA promoters through the piggyback interactions with C/EBPalpha or CBF1. Comparison of four different types of RTA responsiveness. | The relative responsiveness to both C/EBP and RTA transactivation varies greatly among the five promoters that we have studied extensively both here and previously. Table compares and summarizes the average level of enhancement of LUC reporter gene activity from the cellular C/EBPalpha and KSHV RTA, RAP, PAN, and MTA promoters under parallel conditions in HeLa cells. Interestingly, the properties of the IE RTA promoter here most closely resemble those of the C/EBPalpha promoter, whereas the three DE promoters are more similar to one another, especially with regard to direct RTA response levels. All five promoters contain one or more relatively high- affinity C/EBP binding sites, and we suggest that perhaps all C/EBP sites of at least moderate affinity act as weak RREs by mediating severalfold cooperative effects through C/EBPalpha:RTA complex formation. Both the C/EBPalpha and RTA promoters appear to have only this type of RRE. In contrast, the three DE viral promoters each contain in addition a single more powerful and specific RRE. However, our interpretation of the present evidence indicates that all three DE RREs operate by distinctly different mechanisms. In the case of the PAN promoter, the mechanism involves classical direct RTA binding to a type II RRE recognition motif. This may also be true for the K12 and vIL6 RREs , although there is no nucleotide sequence homology between the identified vIL6 motif and the other two type II RTA binding sites. For MTA and RAP, the partially homologous type I RREs both encompass C/EBP binding sites and both fail to bind directly to RTA. However, based on recent studies from Lukac et al. and Liang et al. , it now seems clear that a CBF1 binding motif lies directly adjacent to the C/EBP site and mediates indirect piggyback RTA binding to the MTA RRE. In contrast, for the RAP RRE, although the C/EBP site is essential for RTA responsiveness , this particular site apparently represents a special highly responsive C/EBP motif, presumably because of some other unknown adjacent feature (but it does not resemble a CBF1 site). We suggest that the indirectly responding MTA and RAP sites be referred to as type IA and type IB RREs, respectively. Finally, there remains the question as to why both the C/EBPalpha and RTA responses of the MTA promoter are so much higher than those of the other promoters tested. One possible explanation may reflect the relatively low basal activity of the MTA promoter, which is nearly 20-fold lower than that of the PAN promoter and might be attributable to the presence of the strong binding site for the CBF1 repressor. DISCUSSION : Early lytic cycle gene expression by both KSHV and EBV is driven primarily by three virus-encoded nuclear regulatory proteins ---namely, in KSHV by RTA (ORF50), MTA (ORF57), and RAP (K8) or in EBV by RTA (BRLF1), MTA (BMLF1), and ZTA (BZLF1), respectively. These proteins from the two subclasses of human gammaherpesviruses have diverged enormously, with only the two RTA DNA-binding transcriptional transactivators retaining as much as 15% overall amino acid identity, whereas the two MTA posttranscriptional RNA shuttle proteins retain only small scattered motifs in common, and the two bZIP family proteins resemble one another only at the structural level and in their splicing patterns. Therefore, there are many questions about the overall degree of similarity or differences in their functional activities. Genetic knockout experiments in EBV have clearly shown that there are considerable overlap and redundancy in the functional roles of RTA and ZTA , although little in the way of any common mechanisms has been described. The ability of both constitutively expressed EBV RTA and ZTA to individually trigger full lytic cycle expression in latently EBV-infected B cells or epithelial cells is highly illustrative of this redundancy . Nevertheless, for KSHV this property has only been ascribed to RTA but not to RAP . For EBV RTA-mediated transactivation, almost as many different mechanisms have been proposed as there are identified target promoters, including direct binding to specific DNA-recognition elements, and indirect mechanisms involving interactions with CBP, USF, Rb/E2F, SP1/3, or activation of JN kinase and mitogen-activated protein kinase. Similarly, for the KSHV version, although the target DNA recognition motif is different, both a direct promoter DNA-binding mechanism and several indirect mechanisms involving CBP, OCT1, SP1, AP1, and CBF1 (RBP-Jkappa) have all been proposed by others . We have been focusing instead on another indirect cellular regulatory pathway involving C/EBPalpha, which unlike many of the others, is evidently targeted by both KSHV and EBV and also includes a physical interaction between C/EBPalpha and RAP or ZTA, as well as with KSHV RTA . For RAP and ZTA, these C/EBPalpha interactions lead to p21-mediated G1 cell cycle arrest in both viruses and also contribute to complex positive autoregulatory loops that apparently serve to strongly and coordinately upregulate levels of the RAP, ZTA, C/EBPalpha, and p21 proteins during the early lytic cycle (; Wu et al., unpublished data). The mechanisms involved include both presumed piggyback targeting of RAP or ZTA to identified C/EBP binding sites in all four promoters and stabilization of the normally short-half-lived C/EBPalpha and p21 proteins by RAP and ZTA . More recently, we found that interactions between KSHV RTA and C/EBPalpha also contribute to the autoregulatory loops acting on the RAP promoter . In addition to RTA, the cellular transcription factor C/EBPalpha is itself a potent activator of the DE class RAP promoter , acting through direct binding to an upstream C/EBP site present in the RAP promoter, which overlaps with and appears to constitute part of the previously identified type I RRE. Direct interactions between RTA and C/EBPalpha as well as between RTA and RAP occur in lytically infected cells as well as in in vitro assays . Presumed piggyback DNA binding by both RTA:C/EBPalpha and RAP:C/EBPalpha complexes mediated through the C/EBP binding site in the RRE evidently results in high-level cooperative activation of the RAP promoter in transient-cotransfection reporter gene assays, as well as in KSHV lytically infected cells. The results described here now provide further evidence for a critical and important role of C/EBPalpha in KSHV lytic cycle regulation during both earlier and later events than just those associated with the DE RAP promoter. In particular, we have demonstrated here that C/EBPalpha on its own can positively regulate the KSHV IE class RTA promoter, as well as both the DE class PAN and MTA promoters, and that it contributes cooperatively to RTA autoregulation as well as to RTA mediated-enhancement of at least these two other downstream DE viral promoters, even where other direct and indirect mechanisms have already been described. Aspects of our present results also hint that, like both RAP and ZTA, the interaction with RTA might lead to stabilization of C/EBPalpha, and they also raise as yet untested questions about whether RTA alone might be capable of independently inducing cell cycle arrest and p21 levels. RTA mRNA belongs to the IE class of transcripts whose expression is induced very early (even in the absence of new protein synthesis) upon primary infection or after reactivation from latency. However, the mechanism that initially triggers RTA promoter expression during the earliest stages of KSHV lytic reactivation is unknown. In cell culture, TPA and butyrate mimic this process, and RTA was originally reported to positively autoregulate its own promoter in cooperation with the cellular OCT1 protein . However, we have shown here that C/EBPalpha plays an even more important role than previously appreciated during the early stages of KSHV lytic reactivation. Not only does C/EBPalpha mediate G0/G1 host cell cycle arrest, but it also directly activates the expression of early viral transcriptional and posttranscriptional nuclear regulatory proteins (RTA and MTA), as well as that of a protein involved in initiating viral DNA lytic replication (RAP). We also demonstrated that RTA cooperates with C/EBPalpha to enhance activation of both the RTA and C/EBPalpha promoters, that RTA can induce endogenous C/EBPalpha proteins in uninfected cells, and that exogenously introduced C/EBPalpha can induce RTA protein expression in latently infected PEL cells. We have previously suggested that the ability of C/EBPalpha and RAP to cooperatively transactivate both the cellular C/EBPalpha and viral RAP promoters provides a mutually self-reinforcing loop to boost the levels of both proteins at early times in KSHV-infected cells . However, as shown here by Northern and Western blotting, the increase of C/EBPalpha mRNA and protein levels in latently infected PEL cells began to occur within 3 h after TPA treatment, whereas the appearance of RAP mRNA and protein did not occur until 12 h. Therefore, the presence of a much earlier trigger of the C/EBPalpha promoter (other than RAP) seems logical. Furthermore, in the DG75 KSHV-negative control cells, we know that TPA treatment itself does not induce endogenous C/EBPalpha expression in B cells, although the introduction of RTA alone does. Obviously, that model must now be modified to argue that C/EBPalpha and RTA also produce a self-reinforcing loop that reciprocally activates both promoters and protein levels at an even earlier stage and that subsequent RAP-mediated induction of C/EBPalpha expression occurs predominantly at later stages and functions to further enhance and maintain the levels of both C/EBPalpha and p21 proteins to promote cell cycle arrest . In fact, the highest levels of C/EBPalpha mRNA and protein were indeed observed between 12 and 24 h, a period when RAP protein is abundantly expressed. However, there is still an interesting dilemma. Despite the fact that C/EBPalpha can activate the RTA promoter and that either KSHV RAP or RTA alone strongly induces C/EBPalpha expression in PEL cells, RAP alone (unlike RTA) is unable to induce RTA, vIL6, or other markers of the lytic cycle when introduced into latently infected PEL cells . It is probable that the promoters for lytic cycle viral genes, especially RTA, are transcriptionally silenced or repressed during latency and that even the high levels of C/EBPalpha protein induced by the exogenously introduced RAP alone is not sufficient to reverse the latent repression of RTA. In fact, the introduction of exogenous C/EBPalpha into latent BCBL1 cells was much more effective in inducing endogenous RAP expression (82%) than endogenous RTA expression (37%), suggesting that C/EBPalpha itself probably cannot fully overcome the latency-associated repression of the RTA promoter. Similarly, C/EBPalpha and RTA together boosted the level of cells coexpressing RAP to 98% but that of RTA as shown here to only 51%. Presumably, this differential effect is caused by the presence of negative repressor elements within the KSHV RTA promoter similar to those in the EBV ZTA promoter , which may only be relieved by lytic cycle-inducing agents such as the protein kinase C agonist TPA, histone acetylation by butyrate, or perhaps B-cell receptor signaling. Alternatively, maybe only RTA but not RAP can access or displace chromatin appropriately, or the unknown relative contributions of C/EBPalpha protein stabilization compared to transcriptional effects in the PELs may also influence these results. In summary, our model suggests that C/EBPalpha plays a central and highly complex role in KSHV lytic cycle reactivation in a process that involves multiple steps. The initial lytic switch at the RTA promoter is likely to be caused by a cellular activation or differentiation signal that is mimicked by chemical agents and/or cellular stress signals (such as TPA or butyrate). This trigger probably involves removal of cellular repressors from the RTA promoter that may in turn have been elevated or sequestered by latency-associated viral proteins. Once derepression is achieved, initial low levels of RTA may in turn cooperate with already existing low levels of C/EBPalpha to upregulate endogenous C/EBPalpha protein expression. After heightened C/EBPalpha protein levels are achieved, these could cooperate with RTA to further upregulate the RTA IE promoter. Both proteins then cooperate to activate the downstream DE class RAP promoter, and as levels of RAP build up, a second reinforcing loop between C/EBPalpha and RAP is established, which also involves stabilization of the C/EBPalpha (and p21) proteins, including their relocalization into PML oncogenic domains, and all leading to cell cycle arrest. Independently of C/EBPalpha, there are also more powerful direct effects of RTA or interactions of RTA with the cellular factor CBF1 that act on certain DE promoters such as PAN and MTA, and presumably also those for the core viral replication proteins, thus pushing the cells further into the lytic cycle and stimulating viral DNA replication and the formation of DNA replication compartments (RC). The concepts presented here also raise new questions that have yet to be addressed experimentally about (i) whether RTA (like RAP) can independently stabilize the usually short half-lived C/EBPalpha protein, followed by inducing p21 expression and G1 cell cycle arrest; (ii) whether C/EBPalpha is itself one of the normal cellular triggers of RTA expression during KSHV reactivation in PELs and DMVECs, and perhaps any differentiation or cell cycle arrest events that may activate C/EBPalpha expression can themselves lead directly to the triggering of KSHV reactivation; and (iii) whether in addition antagonists of C/EBPalpha, perhaps including C/EBPbeta, C/EBPdelta, and CHOP10, may also contribute toward establishment, maintenance, or reactivation from latency. FIG. 1. : C/EBPalpha activates RTA-LUC reporter gene expression in cotransfected cells. C/EBPalpha activates RTA-LUC reporter gene expression in cotransfected cells. (A) HeLa cells were transfected with 0.2 mug of plasmid DNA encoding RTA(-914/+34)-LUC and the indicated amounts of effector plasmid DNA expressing CMV-C/EBPalpha, CMV-RTA, or SV2-RAP. The total amount of effector plasmid DNA used in each transfection was normalized to 1.5 mug by adding empty CMV promoter vector (pcDNA3.1) DNA. Activation of LUC activity was calculated relative to the basal level control obtained using 1.5 mug of empty CMV promoter vector DNA as the effector. (B) DG75 B-lymphoblast cells were transfected by electroporation with 2 mug of the RTA-LUC target plasmid DNA and the indicated amounts of effector plasmid DNA. The total amount of all plasmid DNA used in each transfection sample was normalized to 15 mug by adding empty CMV promoter vector DNA. Error bars, standard deviations. FIG. 2. : Induction of endogenous RTA mRNA by transfected C/EBPalpha in PEL cells. Induction of endogenous RTA mRNA by transfected C/EBPalpha in PEL cells. (A) Schematic diagram showing the structure of the RTA gene and the positions of the two primers used for RT-PCR amplification (915 bp). (B) RT-PCR analysis showing that the level of RTA mRNA was elevated in C/EBPalpha-transfected BCBL1 cells is depicted in the upper panel. Control RT-PCR analysis for cellular GAPDH mRNA levels showing equal loading of all lanes is depicted in the lower panel. Lane 1, DNA size marker; lane 2, negative control RNA for BCBL1 cells transfected with empty pcDNA3.1 vector DNA only; lane 3, positive control RNA for untransfected BCBL1 cells treated with TPA for 24 h; lane 4, CMV-C/EBPalpha-transfected BCBL1 cell RNA; lane 5, SV2-RAP-transfected BCBL1 cell RNA; lane 6, CMV-C/EBPalpha- and SV2-RAP-cotransfected BCBL1 cell RNA; lane 7, positive control RNA from BCBL1 cells transfected with CMV-RTA plasmid DNA, used as a size marker for the PCR amplification of RTA cDNA. FIG. 3. : Exogenously introduced C/EBPalpha triggers endogenous RTA protein expression in PEL cells. Exogenously introduced C/EBPalpha triggers endogenous RTA protein expression in PEL cells. (A to C) Control double-stain IFA showing lack of background expression of either Flag epitope or the endogenous RTA protein in BCBL1 cells transfected with empty CMV-Flag vector DNA as detected with anti-Flag MAb (A) (rhodamine [red]) or anti-RTA PAb (B) (FITC [green]). (C) DAPI nuclear staining (blue) showing all cells in the same field. (D to F) Expression of Flag-C/EBPalpha protein (D) (red) in transfected BCBL1 cells induces expression of endogenous RTA protein in the same cells as detected with anti-RTA PAb (E) (green). (F) Merged image. (G to I) Expression of Flag-RAP protein in transfected BCBL1 cells (G) (red) fails to induce endogenous RTA protein expression (H) (green). (I) DAPI (blue) showing the whole cell population. (J to L) Expression of Flag-RAP protein (J) (red) in C/EBPalpha plus Flag-RAP cotransfected BCBL1 cells and induction of RTA (K) (green) in the same cells. (L) Merged image. Note that spontaneous background RTA expression occurs in approximately 1% of untreated BCBL1 cells. FIG. 4. : Deletion analysis of the RTA (ORF50) promoter and identification of C/EBP binding sites. Deletion analysis of the RTA (ORF50) promoter and identification of C/EBP binding sites. (A) Schematic diagram of the RTA promoter region between positions -914 and +20 relative to the mRNA start site at genomic coordinate 71560, including showing the locations of the three C/EBP binding sites defined here. Five oligonucleotide probes encompassing potential C/EBP motifs predicted by sequence analysis are designated RP-1 to RP-5. The OCT site, which has been reported to be important for RTA autoactivation, and the TATA box are also indicated. The promoter region segments retained in a set of RTA-LUC deletion derivative plasmids that sequentially remove motifs RP-1 to RP-5 are designated according to their 5' boundaries and are represented by solid lines. The X indicates that the C/EBP-III site (RP-5) was destroyed by changing GCAATG (positions 71512 to 71517) to GATATC. (B) EMSA experiment showing the C/EBPalpha binding ability of each putative C/EBP motif in the RTA promoter. Annealed 34-bp double-stranded oligonucleotide probes containing motifs RP-1 to RP-5, as well as the mutated probe RP-5M, were 32P end labeled and tested for the ability to bind to in vitro-translated C/EBPalpha protein (lanes 2, 5, 8, 11, 14, and 17). A sample of the unprogrammed reticulocyte lysate was used as control for nonspecific shifted bands (lanes 1, 4, 7, 10, 13, and 16). Rabbit polyclonal antiserum against C/EBPalpha was used to generate specific supershifted bands (lanes 3, 6, 9, 12, 15, and 18). Abbreviations: S, C/EBPalpha-specific shifted bands; SS, antibody supershifts; NS, nonspecific bands. (C) Responsiveness of seven different RTA-LUC deletion mutant derivatives to C/EBPalpha transactivation in cotransfected HeLa cells. Transfections were carried out using 200 ng of target reporter promoter DNA and 0-, 50-, 100-, or 500-ng input doses of C/EBPalpha effector expression plasmid DNA as described in the legend to Fig. . Error bars, standard deviations. (D) Cooperative activation of seven different RTA-LUC deletion derivatives by C/EBPalpha, RTA, and RAP in HeLa cells. Transfection was carried out using 200 ng of target reporter plasmid DNA, 250 ng of C/EBPalpha expression plasmid DNA and 500 ng of either RTA or RAP expression plasmid DNA alone or a combination of both as indicated. Error bars, standard deviations. FIG. 5. : ChIP assay showing that the C/EBPalpha, RTA, and RAP proteins all associate with RTA promoter DNA in vivo. ChIP assay showing that the C/EBPalpha, RTA, and RAP proteins all associate with RTA promoter DNA in vivo. (A) Schematic diagram of the 261-bp region of the RTA promoter that was targeted for PCR amplification. (B) Results of ChIP assays carried out with samples from JSC1 cells after TPA treatment for 40 h are shown in the upper panel. Lane 1, DNA size marker; lanes 2 to 6, ChIP assay PCR products from immunoprecipitates obtained using antibodies against either RTA, C/EBPalpha, RAP, or EBV ZTA or with no antibody; lane 7, size control for 261-bp PCR amplification product from the RTA(-914/+34)-LUC reporter plasmid; lanes 8 and 9, ChIP assay PCR products obtained from immunoprecipitates using antibodies against RTA or RAP with cell extracts that had been precleared with C/EBPalpha antibody; lanes 10 and 11, ChIP assay PCR products obtained from immunoprecipitates using C/EBPalpha antibody and cell extracts that had been precleared with antibodies against RTA or RAP. Control PCR amplification of RTA coding region sequences (300 bp) from the same ChIP immunoprecipitates are shown in the lower panel. These results demonstrate that the binding is specific for promoter region sequences. FIG. 6. : Enhanced expression of C/EBPalpha mRNA and protein at very early stages during the KSHV lytic cycle. Enhanced expression of C/EBPalpha mRNA and protein at very early stages during the KSHV lytic cycle. (A) Kinetics of C/EBPalpha and KSHV early gene mRNA synthesis during TPA induction in PEL cells. Northern blotting was carried out using total RNA extracted from BCBL1 cells at different time points after TPA induction (0, 3, 6, 24, 33, and 48 h, as indicated). Upper panels show membranes that were sequentially hybridized with denatured 32P-labeled DNA fragment probes detecting mRNA for C/EBPalpha (2.5-kb), RTA (3.2-kb), or RAP (1.1-kb). The lower panel shows a probe for GAPDH (1.3-kb) that was used as loading control for the different samples. Positions of an RNA size marker ladder are indicated at the left of each panel. (B) Examination of C/EBPalpha protein levels at early times during the KSHV lytic cycle in PEL cells. Protein samples were isolated at 0, 0.5, 1, 2, 3, and 12 h after TPA induction from BCBL1 cells and at 0, 12, and 24 h from DG75 cells as indicated. The upper three panels show Western immunoblots that were performed using rabbit antibodies against C/EBPalpha, RTA, or RAP. In the lower panel, antibody against beta-actin was used as a loading control. The positions of protein size markers are indicated at the left of each panel. FIG. 7. : Evidence that RTA expression enhances C/EBPalpha protein levels. Evidence that RTA expression enhances C/EBPalpha protein levels. (A) Expression of endogenous C/EBPalpha protein and that of RTA proteins are both induced in the same subpopulation of BCBL1 cells after TPA treatment. Double-stain IFA shows the absence of C/EBPalpha and RTA expression in uninduced BCBL1 cells (a to c), compared to high-level expression of both endogenous C/EBPalpha and RTA in BCBL1 cells after treatment with TPA for 24 h (d to f). Goat anti-C/EBPalpha PAb (a and d) (rhodamine [red]) and rabbit anti-RTA PAb (b and e) (FITC [green]) wereused for IFA detection. (c and f) Merged images. (B) Immunoblot detection of increased endogenous C/EBPalpha protein (42- and 30-kDa forms) in KSHV-negative B cells transfected with RTA. DG75 cells were transfected with either empty vector DNA (lane 1), two different input doses of CMV-RTA (lanes 2 and 3), or the CMV-C/EBPalpha (lane 4) expression plasmids as indicated. Western blotting was performed using antibodies against C/EBPalpha (upper panel), RTA (middle panel), or beta-actin (lower panel). (C) Immunoblot detection of increased cotransfected C/EBPalpha protein in the presence of RTA. 293T cells were cotransfected with CMV-Flag-C/EBPalpha together with either the empty CMV promoter vector DNA (lane 1), wild-type CMV-RTA(1-691) expression plasmid (lane 2), or mutant CMV-RTA(1-691Delta151-167) expression plasmid (lane 3). Western immunoblotting was performed using Flag MAb (to detect exogenous C/EBPalpha) (upper panel), RTA PAb (middle panel), or beta-actin MAb (lower panel). Positions of protein size markers are indicated to the left of each panel. FIG. 8. : RTA activates the C/EBPalpha promoter cooperatively with C/EBPalpha in transient cotransfection assays. RTA activates the C/EBPalpha promoter cooperatively with C/EBPalpha in transient cotransfection assays. (A) HeLa cells were transfected with 0.2 mug of C/EBPalpha-LUC reporter gene plasmid and the indicated amounts of effector plasmids expressing either C/EBPalpha, wild-type RTA(1-691) or mutant RTA(Delta151-167), or both together. The total amount of effector plasmid DNA used in each transfection was normalized to 1.2 mug by adding empty CMV promoter vector DNA. Activation of LUC activity was calculated based on the control transfection using 1.2 mug of empty CMV vector DNA, representing the basal level. (B) DG75 B-lymphoblast cells were electroporated with 2 mug of C/EBPalpha-LUC target plasmid and the indicated amounts of effector plasmids. The total amount of effector plasmids used in each transfection was normalized to 15 mug by adding empty vector DNA. Error bars, standard deviations. FIG. 9. : C/EBPalpha binds to and transactivates the KSHV PAN promoter. C/EBPalpha binds to and transactivates the KSHV PAN promoter. (A) HeLa cells were cotransfected with 0.1 mug of PAN(-210/+15)-LUC reporter plasmid DNA and indicated amounts of effector plasmids expressing C/EBPalpha, wild-type RTA(1-691), or mutant RTA(Delta151-167). Error bars, standard deviations. (B) EMSA with probe PAN-225 showing that in vitro-translated C/EBPalpha, intact RTA, and the RTA DBD protein segment all bind independently to the PAN(-210/+15) promoter region. Abbreviations: S, shifted bands; SS, antibody supershifts; NS, nonspecific bands. FIG. 10. : Identification and functionality of two C/EBP binding sites within the PAN promoter. Identification and functionality of two C/EBP binding sites within the PAN promoter. (A) Schematic diagram of the PAN promoter region from position -210 to +15 relative to the defined transcription start site at genomic coordinate 28667, including the previously reported RRE, a putative SP1 site, three potential SRE sites, and the two C/EBP sites identified here. Promoter regions contained in probes PAN-1 to -4 and PAN-RRE and in wild-type and mutant PAN-LUC reporter gene derivatives are represented by solid lines. Mutations (X) are indicated. (B) Nucleotide sequences of six wild-type and mutant PAN EMSA probes and the PAN-LUC reporter genes shown above. Motifs that may serve as C/EBP and other consensus binding sites are delineated by arrows and boxes. Mutated nucleotides are indicated in boldface type. (C) EMSA to test the binding affinity of six different 34- to 44-bp PAN promoter probes to in vitro-translated C/EBPalpha, C/EBPalpha plus specific antibody, or the RTA(1-377) DBD protein fragment. Abbreviations: S, shifted bands; SS, antibody supershifts; NS, nonspecific bands. (D) EMSA comparing C/EBPalpha binding affinity of wild-type and mutant forms of the PAN-2 and PAN-4 probes. (E) Results of transient-expression assays with mutated PAN(-190/+15)-LUC target reporter genes introduced into HeLa cells to test their responsiveness to cotransfected C/EBPalpha or RTA or both together. Transfections were carried out as described above using 0.1 mug of reporter plasmid DNA and the indicated amounts of effector DNA. Error bars, standard deviations. FIG. 11. : C/EBPalpha transactivates the KSHV MTA promoter by binding to specific C/EBP sites. C/EBPalpha transactivates the KSHV MTA promoter by binding to specific C/EBP sites. (A) Schematic diagram of the MTA promoter region from positions -160 to +10 relative to the defined transcription start site at genomic coordinate 82003, including the previously reported RTA response element (RRE), CBF1 site, a putative AP1 site and the C/EBP sites identified here. Positions of the probes used for EMSA as well as the sequence for probe MTA-2 are indicated. Promoter regions contained in different derivatives are represented by solid lines. The mutated C/EBP site in MTA(-100/+10 pm-94/-90)-LUC is indicated (X). (B) EMSA to test the binding affinity of three different MTA promoter probes to either unprogrammed reticulocyte lysate (lanes 1, 5, 10, and 14) or in vitro-translated C/EBPalpha (lanes 2, 6, 11, and 15), C/EBPalpha plus specific antibody (lanes 3, 7, 12, and 16), full-length RTA(1-691) protein (lanes 9 and 17) or the RTA(1-377) DBD protein segment (lanes 4, 8, and 13). Abbreviations: S, shifted bands; SS, antibody supershifts; NS, nonspecific bands. (C) Results of transient reporter gene expression assays with deleted or mutated MTA-LUC target reporter genes introduced into HeLa cells to test their responsiveness to cotransfected C/EBPalpha (open bars), wild-type RTA(1-691) (black bars), or both together (shaded bars). Transfections were carried out using 0.2 mug of reporter plasmid DNA and the indicated amounts of effector DNA. Error bars, standard deviations. FIG. 12. : Time course ChIP assays showing the association of C/EBPalpha, KSHV RTA, and RAP proteins with target viral promoters during early KSHV lytic reactivation. Time course ChIP assays showing the association of C/EBPalpha, KSHV RTA, and RAP proteins with target viral promoters during early KSHV lytic reactivation. At different time points after TPA induction, aliquots of the BCBL1 cell lysates were precipitated with antibodies against KSHV RTA, RAP, C/EBPalpha, or CHOP10. Associations between proteins and target promoters were detected by PCR using primers specific for promoter regions of KSHV RTA (A), PAN (B), and MTA (C). Lane 1, DNA size marker; lanes 2 to 5, ChIP assay PCR products from either uninduced input BCBL1 cell lysates (0 h, lanes 6 to 9) or those at 8 h (lanes 10 to 13) and 24 h (lanes 14 to 17) after TPA induction; lanes 6, 10, and 14, PCR products from immunoprecipitates obtained using antibodies against RTA; lanes 7, 11, and 15, PCR products from immunoprecipitates obtained using antibodies against RAP; lanes 8, 12, and 16, PCR products from immunoprecipitates obtained using antibodies against C/EBPalpha; lanes 9, 13, and 17, PCR products from immunoprecipitates obtained using antibodies against CHOP10; lane 18, size control PCR amplification products from the corresponding promoter reporter plasmids including RTA(-914/+34)-LUC, PAN(-210/+15)-LUC and MTA(-160/+10)-LUC reporter plasmids. TABLE 1 : Different types of RTA responsiveness Backmatter: PMID- 12915569 TI - Structural and Phylogenetic Analysis of Adenovirus Hexons by Use of High-Resolution X-Ray Crystallographic, Molecular Modeling, and Sequence-Based Methods AB - A major impediment to the use of adenovirus as a gene therapy vector and for vaccine applications is the host immune response to adenovirus hexon ---the major protein component of the icosahedral capsid. A solution may lie in novel vectors with modified or chimeric hexons designed to evade the immune response. To facilitate this approach, we have distinguished the portion of hexon that all serotypes have in common from the hypervariable regions that are responsible for capsid diversity and type-specific immunogenicity. The common hexon core ---conserved because it forms the viral capsid ---sets boundaries to the regions where modifications can be made to produce nonnative hexons. The core has been defined from the large and diverse set of known hexon sequences by an accurate alignment based on the newly refined crystal structures of human adenovirus types 2 (Ad2) and Ad5 hexon. Comparison of the two hexon models, which are the most accurate so far, reveals that over 90% of the residues in each have three-dimensional positions that closely match. Structures for more distant hexons were predicted by building molecular models of human Ad4, chimpanzee adenovirus (AdC68), and fowl adenovirus 1 (FAV1 or CELO). The five structures were then used to guide the alignment of the 40 full-length (>900 residues) hexon sequences in public databases. Distance- and parsimony-based phylogenetic trees are consistent and reveal evolutionary relationships between adenovirus types that parallel those of their animal hosts. The combination of crystallography, molecular modeling, and phylogenetic analysis defines a conserved molecular core that can serve as the armature for the directed design of novel hexons. Keywords: Introduction : The human adenovirus family has at least 49 distinct serotypes, which fall into six subgroups. These subgroups, A through F, are associated with a variety of pathologies and tissue tropisms . The subgroup C human adenoviruses, including types 2 (Ad2) and 5, have been widely used to develop vectors for gene therapy applications . Because of prior exposure, many patients have preexisting immunity to these common adenovirus types that limits their use . When introduced, they stimulate B cells to produce neutralizing antibodies that block vector uptake. Nevertheless, infection-induced humoral immunity to a particular serotype does not provide cross-immunity to serotypes in other subgroups . The main targets of type-specific antibodies are the adenovirus structural proteins hexon and fiber . As antifiber antibody neutralizes virus infectivity only in vitro and not in vivo , the neutralizing antibody response relevant to vector applications is directed to hexon. Hexon is the largest and most abundant of the structural proteins in the icosahedral adenovirus capsid . The Ad2 virion has a particle mass of at least 150 x 106 Da , with hexon accounting for more than 83% of the capsid protein. The 240 copies of the homotrimeric hexon molecule are distributed symmetrically, with 12 copies forming each of the 20 capsid facets . With a length of 967 residues, the Ad2 hexon polypeptide chain is the longest of the known hexon sequences (109,064 Da [including the acetylated N terminus]). The other two major capsid proteins, the penton base and fiber, form the penton complex at each virion vertex. Two distinct viral attachment and cellular uptake events are required for infectivity, with each involving a separate site on the penton. Initially, a globular knob at the distal end of the fiber attaches to the host cell primary receptor. This is the coxsackie virus and adenovirus receptor (CAR) for Ad subgroups A, C, D, E, and F . For cells that lack CAR, the receptor is the major histocompatibility complex class I alpha2 subunit (for subgroup C) or sialic acid for Ad37 (subgroup D) . After binding, receptor-mediated endocytosis results from an interaction between an Arg-Gly-Asp sequence on the penton base and its alphavbeta integrin target . Several different approaches have been taken to engineer the capsid of Ad5- or Ad2-based vectors to enhance infectivity and redirect native virus tropism. By exchanging fiber genes, vector uptake has been redirected to a non-CAR-mediated pathway . Alternatively, retargeting can be achieved by attaching bispecific molecules. These include bifunctional peptides and chemically conjugated bispecific antibodies. The antibodies have been attached through an epitope incorporated into penton base or by specific recognition of the fiber knob domain . These approaches show great promise for altering normal infection patterns and allowing specific targeting, but they do not overcome the problem of preexisting immunity. One approach to circumventing preexisting host immunity is to mask the surface antigens with protective polymers such as polyethylene glycol , polylactic glycolic acid microspheres , or lipids . Another approach is to derive vectors from different adenovirus serotypes . A recent example uses a nonhuman adenovirus isolated from chimpanzee (AdC68) . Detailed restriction endonuclease mapping and subsequent full-sequence analysis have shown that AdC68 is closely related to Ad4. However, cross-reactivity of type-specific antisera between AdC68 and Ad4 is absent or diminished . Currently, efforts are under way to engineer the AdC68 vector for use as a vaccine carrier . Another way to accomplish immunological distinction is to modify or replace the type-specific epitopes that are most likely located within seven hypervariable regions (HVRs) in the hexon sequence . The HVRs occur in loops at the top of the molecule that lie on the exterior of the virion and cover nearly its entire surface . Genetic engineering approaches creating chimeric viruses have replaced the entire Ad5 hexon, or just the hexon FG1 loop, with that from Ad2 , replaced all four Ad5 (subgroup C) hexon loop domains with those of Ad12 (subgroup A) , and packaged the Ad5 genome into the capsids of Ad3 (subgroup B), Ad4 (subgroup E), or Ad9 (subgroup D) . Nevertheless, attempts to make a chimeric vector by replacing the Ad5 hexon with that of Ad7 (subgroup B) failed . The authors of that previous work suggested that the Ad7 hexon may be incompatible with the Ad5 100,000 folding protein, which is involved with hexon assembly and transport to the nucleus, or with other capsid proteins. The alteration of hexon epitopes need not be limited to switching serotypes. In fact, an 8-amino-acid sequence from the major antigenic site in the VP1 capsid protein of poliovirus type 3 has been introduced into two regions of the Ad2 hexon . These results show that the hexon structure is amenable to specific alterations but that these must be made with care, as some changes can have a profound effect on viral assembly and infectivity. In this study, new high-resolution crystallographic refinements of the Ad2 and Ad5 hexon structures have been performed to take advantage of newer protocols and to resolve the observed differences between the previous Ad5 (Protein Data Bank [PDB] identifier, 1RUX) and Ad2 (PDB identifier, 1HDX) hexon models. To extend the diversity of the structural data, models for human Ad4, a chimpanzee adenovirus (AdC68), and an avian adenovirus (FAV1 or CELO) were built by molecular modeling. The five structures were then used in conjunction with sequence-based methods to define the common core of the hexon molecule that is required for capsid assembly. The resultant hexon armature describes the parts of the molecule that should not be violated when designing novel adenovirus-based vectors for efficacious human gene therapy or vaccine delivery. MATERIALS AND METHODS : Crystallographic refinement. (i) Ad2 hexon. | A new crystallographic refinement of the Ad2 hexon structure was performed, based on a previous model for Ad5 hexon at 2.5-A resolution (PDB identifier code, 1RUX) . Ad2 hexon crystals are isomorphous with those grown from the homologous Ad5 hexon (86% amino acid identity) . The methods for virus growth, hexon purification and crystallization , and X-ray data collection and processing have been described previously. The Ad2 hexon diffraction data were collected at room temperature from a single crystal (dimensions, 0.6 x 0.5 x 0.5 mm) grown in 0.5 M sodium citrate buffer at pH 3.2. The Ad2 space group is P213 (a = b = c = 150.5 A, alpha = beta = gamma = 90), with one subunit of the hexon trimer in the asymmetric unit. The data include 48,199 of the 57,118 unique reflections in the resolution range from 10 to 2.2 A (i.e., 84% complete). The overall agreement between related reflections, Rsymm, was 9.8%: Rsymm = Sigmah [Ih - average(Ih)]/Sigmah average(Ih), where Ih is the intensity of reflection h. The starting model for Ad2 hexon refinement was the crystallographic model for Ad5 hexon (PDB identifier, 1RUX). This has seven segments (residues 5 to 135, 165 to 186, 193 to 250, 258 to 269, 278 to 430, 437 to 442, and 445 to 946) with intervening gaps where the polypeptide chain is disordered. The segments were renumbered to match the Ad2 hexon sequence (residues 5 to 135, 172 to 193, 204 to 261, 269 to 280, 290 to 442, 454 to 459, and 461 to 962). Of the 884 residues in the starting Ad2 model, 65 residues that differ in sequence from Ad5 were truncated at the beta-carbon and modeled as Ala. To serve as a control for potential model bias, the remaining 110 residues with large aromatic side chains (i.e., Phe, Tyr, and Trp) were also modeled as Ala, and all water molecules were removed. All atomic temperature factors (B factors) were initialized to 15 A2, a value typical of atomic positional disorder in macromolecular structures at room temperature. Initial refinement was performed as previously described for Ad5 using TNT modified to include the MLF1 maximum-likelihood target function, TNT-ML . Crystallographic R factors, which provide an estimate of model error, were monitored during refinement. The R factor is the sum of the absolute differences between experimentally observed (Fo) and calculated (Fc) amplitudes over all observed diffraction reflections (R = Sigma |Fo - Fc|/Sigma |Fo|). The working R (Rwork) is calculated with the reflections used for refinement, and an unbiased or free R (Rfree) is calculated with a subset of the data that has not been included in the refinement. Refinement, commencing with a rigid body, was performed using all data except for 5% used to monitor Rfree . The model was rebuilt between refinement rounds using program O . To take advantage of the greater convergence potential of torsion angle dynamics, refinement was continued using the Crystallography and NMR System (CNS) software version 0.9a with the protocol based on the supplied "refine.inp" procedure. For each round of CNS refinement, diffraction data from 6.0- to 2.2-A resolution were used for initial B factor and bulk solvent corrections. Solvent molecules were harmonically restrained with a constant of 10. The model was first subjected to 200 steps of conjugate gradient minimization, and then torsion angle dynamics with slow-cooled annealing were performed from 5,000 to 300 K with 100-K temperature drops per cycle. Finally, another 200 steps of conjugate gradient minimization were applied, followed by 20 steps of restrained individual B-factor minimization. The model was rebuilt between refinement rounds as before, and water molecules were located with the waterpeaks script in the CCP4 program package (Collaborative Computational Project, 1994). The coordinates and structure factors have been deposited in the Protein Data Bank (identifier 1P2Z). (ii) Ad5 hexon. | The same CNS methods that improved the Ad2 hexon model (above) were also applied to the previously published TNT-ML-refined 2.5-A resolution Ad5 hexon structure (PDB identifier, 1RUX) . The Ad5 data set used was identical to that for 1RUX. The space group is P213 (a = b = c = 150.7 A, alpha = beta = gamma = 90), with one subunit of the Ad5 hexon trimer in the asymmetric unit. The diffraction data include 36,549 of 39,573 unique reflections (92% complete from 20- to 2.5-A resolution) and have an Rsymm of 8.1%. The coordinates and structure factors have been deposited in the Protein Data Bank (identifier 1P30). Molecular modeling. | The methods used to model the Ad4 and AdC68 hexons were described previously and used to model the CELO hexon. Briefly, models of the homologous Ad4, AdC68, and CELO hexons were initially produced using the Swiss-PdbViewer protein modeling environment . Sequence alignments were used to guide the threading of model sequences onto the known molecular structures. These initial molecular models were then improved manually by moving gaps to exposed variable regions and optimizing side chain packing. Side chains or stretches of main chain that were missing from the crystal structures because of disorder were selected from libraries of side chain rotamers or known loop structures. The conformation of each model was further refined by energy minimization using the molecular mechanics program CHARMM . Graphical images prepared within the Swiss-PdbViewer program were exported and rendered with the Persistence of Vision Ray Tracer program (POV-Ray 2000; version 3.1g). Structure- and sequence-based alignments. | The X-ray and homology model structures were each aligned with the Ad5 hexon model with transformation matrices calculated with the program LSQMAN and applied with the program O . Secondary structure elements were assigned with PROMOTIF . A high-quality reference alignment was prepared by aligning the sequences corresponding with the five structurally aligned models. Forty unique full-length (>900 residues) hexon sequences were retrieved from a search of GenBank with the Ad5 hexon sequence using BLASTP (sequence accession numbers are listed below). The 35 unique and unmodeled full-length hexon sequences were then aligned with the Clustal_X program using the reference alignment as a profile. The multiple-alignment parameters used in this heuristic approach included the following: Gonnet 250 series matrices, sequence divergence delay of 30%, gap opening 10, gap extension 0.2, residue-specific and hydrophilic penalties included, gap separation distance of 4, and no end gap separation. Hexon sequence labels and GenBank accession numbers (in parentheses) are as follows: ad12_A , ad3_B , ad7_B , ad7a_B , ad16_B , ad2_C , ad5_C , ad21_B , ad34_B , ad35_B , unk , ad48_D , ad4_E , ad41_E , ad40_F , bovine3 , bovine4 , bovine5 , bovine6 , bovine7 , bovine8 , bovine10 , canine1 , canine2 , chimp68 or adC68 , duck1 , equine1 , equine2 , fowl1 or CELO , fowl8 , fowl10 , frog1 , goat1 , murine1 , odocoileus , ovine7 , ovine287 , porcine3 , porcine5 , and turkey3 . Phylogenetic analyses. | Phylogenetic analyses of the aligned hexon sequences were performed with PHYLIP (version 3.5.7c) . The following program protocols were used: seqboot to calculate 100 bootstrap trials, PROTPARS and CONSENSE for parsimony analysis, and a combination of PROTDIST using the Dayhoff PAM matrix , FITCH, and CONSENSE for distance-based analysis. The phylogenetic trees were displayed with TreeView 1.6.1 . RESULTS : Crystallographic refinement. | Initial rigid body refinement of the Ad2 starting model reduced the Rwork and Rfree values from 38.3 to 33.2 and from 38.1 to 33.8%, respectively. The large aromatic side chains that were intentionally removed from the starting model as a control were immediately apparent in the resultant electron density. After four rounds of model building and TNT-ML refinement, the model contained 883 of 967 residues (91%), 296 water molecules, with an Rwork value of 19.0% and an Rfree value of 23.7%. Subsequent CNS refinement converged after five rounds to yield 887 of 967 Ad2 hexon residues (93%), five citrate molecules, 430 water molecules, with an Rwork value of 17.8% and an Rfree value of 22.1% at 2.2-A resolution. For Ad5 hexon, the CNS protocol reduced the initial Rwork and Rfree values from 21.5 to 16.9 and 24.3 to 21.0%, respectively. The refinement converged after four more rounds to yield 891 of 951 Ad5 hexon residues (94%), 304 water molecules, with an Rwork value of 15.8% and an Rfree value of 20.7% at 2.5-A resolution. The presence of citrate molecules reflects the composition of the crystallization buffer. The statistics for the final Ad2 and Ad5 hexon models are shown in Table . Structure alignment. | While the fold of the newly refined Ad2 hexon model is generally similar to that of an earlier model , the alterations bring the new Ad2 model into close agreement with the Ad5 model. An alignment of the current two models with the program LSQMAN shows an excellent match for 874 residues in the Ad2 (887 residues) and Ad5 (891 residues) models, with a root mean square (RMS) distance between Calpha atoms of 0.3 A. Of these 874 matching residues, 812 have the same side chains, whereas the remaining 62 residues differ. The matching residue positions for the two X-ray structures are as follows: for Ad5, positions 5 to 135, 165 to 186, 192 to 211, 213 to 250, 258 to 268, 279 to 305, 308 to 430, and 445 to 946; for Ad2, positions 5 to 135, 172 to 193, 203 to 222, 224 to 261, 269 to 279, 291 to 317, 320 to 442, and 461 to 962. As the hexon homology models for the AdC68, Ad4, and CELO hexons were based upon the Ad2 and Ad5 crystal structures, they are also closely related to one another. Hexon regions. | For simplicity, all hexon models will be discussed together in this section, although only two are based on crystallographic data. Each hexon structure can be divided into eight regions (Fig. ; Table ). The list of secondary structural elements shows a few that have been newly identified in the two X-ray structures (underlined) and some that are predicted in the homology models (in parentheses). The N-terminal region, 55 residues long in the Hominidae (human and chimpanzee) hexons and 60 for CELO, forms an extended loop on the basal surface of the trimer. It is important for stability, as it extends from one subunit, under a second, to touch the third. The additional five residues in CELO precede the alpha0 helix (Fig. ; Table ) and would be exposed at the edge of the base where they could be involved in intermolecular contacts in the viral capsid. Some flexibility is present at the N terminus, as the first four residues are disordered in both crystal structures. The NT-alpha0 helix secondary structural element is newly observed in the current Ad2 and Ad5 crystal structures (residues 5 to 12). The only significant difference between the two models in the N-terminal region is a different side chain conformation for His 52. The N-terminal region is followed by the first of the two viral jellyrolls, V1 (Fig. ; Table ), with its eight strands labeled B to I. The size of V1 is essentially constant, with 198 residues in human hexons and 197 residues in CELO. The DE1 loop is an extended structure extending from strands D and E of V1 to form the largest domain in the Ad2 and Ad5 hexons (220 and 208 residues, respectively). In all other known hexon sequences, DE1 is smaller. DE1 is the most variable of the hexon regions and is flexible; the X-ray models have four gaps due to low electron density . Ad5 hexon has two new very short strands in DE1 at residues 191 to 194 (beta5a) and 236 to 237 (beta6a). The FG1 loop, emerging between the F and G strands of V1, is smaller than the DE1 loop and relatively constant in size compared to DE1, varying from 159 residues (AdC68) to 166 residues (CELO). Two new secondary structural elements, FG1-beta4a and FG1-alpha3a, are observed in both X-ray structures (Fig. ; Table ). The newly observed FG1-beta4a strand hydrogen bonds with the previously observed FG1-beta4 strand to form a second short antiparallel pair close to that formed by FG1-beta3 and FG1-beta5. The viral jellyroll connector domain (VC) is located in the base of the molecule and serves, with V1-alpha1 and loop DE2, to separate the first (V1) and second (V2) viral jellyrolls and to clamp them together. VC is formed by stretches from the middle and end of the molecule (i.e., residues 638 to 660 and 938 to 967 in Ad2), but its size is almost invariant ---53 residues for the human hexons and 52 residues for CELO. This conservation is particularly significant, as any alteration in VC size would alter the separation of V1 and V2 and so affect the hexon diameter (see Fig. ). With 720 copies of hexon present in the capsid, any small change in the separation of V1 and V2 would be greatly magnified and have a significant effect on the overall size of the virion. VC extends to the C terminus, with the last five residues disordered in both X-ray structures. The V2 domain is the smaller of the two viral jellyrolls, with 151 residues for all Hominidae hexons and 153 for CELO . V2 also has loop extensions between the DE and FG strands, but they are much smaller than their counterparts in V1. Their structural similarities indicate that the jellyrolls arose by gene duplication . The double jellyroll is a key feature that distinguishes viruses such as adenovirus and bacteriophage PRD1 from viruses with single-jellyroll capsid proteins like the picornaviruses . Trimeric double-jellyroll coat proteins have pseudohexagonal shapes that make them well suited to accurate assembly of large viral capsids . The DE2 loop is the smallest of the four hexon loops, and its length is remarkably conserved at 30 residues in all the hexon structures. This conservation probably arises because DE2 plays the same role as VC in separating and stabilizing V1 and V2 ---but at the upper half of the jellyrolls. The path of the Ad2 residues ---706 Leu->707 Gly->708 Ser ---differs from that of the corresponding residues in Ad5 (residues 690 to 692), even though the residues are the same. The flexibility at this point explains how the frog and turkey hexons can accommodate a Trp residue in place of the Gly that is present in the other 38 hexon sequences. The Ad2 Gly 707 residue lies on the basal surface, so the Trp in frog and turkey may be involved in hexon-hexon interactions in the capsid. The FG2 loop, which extends between the F and G strands of V2, is much larger than the DE2 loop. FG2 has 96 residues in the Hominidae hexon models and 115 in CELO. In contrast to the other upper loops, DE1 and FG1, FG2 has no disordered residues in the X-ray structures. It is also much smaller and lower in height. FG2 forms an intriguing solvent-accessible pocket underneath the tower that accommodates hydrophobic residues of quite different sizes in different hexons. As they are normally buried, such residues are usually conserved. The pocket explains the puzzling variability in these residues, which lie within a generally conserved region in the sequence, but the feature has no obvious function. In Ad2, the pocket is occupied by Val 861, with Phe 845 in Ad5, and Ala, Tyr, or Trp in other hexons. Sequence alignment. | The structural superposition of the hexon crystallographic and homology models aligned their five sequences and produced a profile to guide the alignment of the remaining 35 sequences. The resultant sequence alignment is colored to reveal conserved chemical properties using default Clustal_X definitions. Sequence diversity is indicated by the Clustal_X quality scores, which are shown as a histogram and as a colored molecule . The conservation pattern reflects hexon's function as an architectural building block. Its pseudohexagonal shape enables hexons to closely pack and form a closed spherical capsid. The inside of the molecule's central cavity and its upper surface both vary, particularly in the tower regions at the top, while the inside and lower surfaces are conserved. Thus, the regions that make up the base of the molecule, NT (77% identity and 83% similarity), V1 (76% identity and 84% similarity), VC (76% identity and 81% similarity), V2 (79% identity and 87% similarity), and DE2 (81% identity and 86% similarity) have highly conserved sequences (Fig. and ). In contrast, the regions forming the top of the molecule, DE1 (37% identity and 46% similarity), FG1 (62% identity and 71% similarity), and FG2 (64% identity and 72% similarity) are less highly conserved (Fig. and ). The most divergent of the 1,064 positions in the hexon are the 208 (20%) with Clustal_X quality scores in the bottom 10th percentile. These unconserved residues are distributed throughout the hexon structure (colored red in Fig. ), but all are located on the molecular surface. Many of these residues fall at the top of the molecule in the HVRs defined by Crawford-Miksza and Schnurr . The new alignment made it possible to assign the HVRs quantitatively as stretches of more than five residues with quality scores in the bottom 10th percentile (with at most one residue allowed to have a higher score). This procedure defines the HVR boundaries more precisely and yields nine regions: HVR1 (residues 146 to 181), HVR2 (199 to 221), HVR3 (264 to 270), HVR4 (288 to 293), HVR5 (307 to 330), HVR6 (359 to 366), HVR7 (482 to 489), HVR8 (503 to 513), and HVR9 (900 to 911). Phylogenetic analysis. | The results of the phylogenetic analyses of the 40 aligned full-length hexon sequences are summarized in Fig. . Parsimony analysis builds an evolutionary tree depicting the order requiring the least number of residue alterations to change each sequence to another. Sequences that group together are more likely to have evolved together. The parsimony analysis for the hexon shows that the Hominidae hexons clearly segregate from all others and that the human hexons form groups consistent with their serotypes. The hexon of unknown origin (unk) is likely to be in subgroup B. That Ad16 (subgroup B) seems more similar to Ad4 (group E) than to others in its subgroup illustrates the limitations of an analysis based solely on sequence. Phylogenetic distance analysis estimates the divergence time between each pair of sequences. The smaller the distance, the more likely it is that the pair has evolved from a common ancestor. The results of the distance matrix analysis are shown as trees . The chimpanzee hexon (AdC68) is clearly grouped with the human serotypes and is most similar to Ad4 and Ad16. It is clear from the data that the adenoviruses share a common origin. For the human, equine, porcine, bovine, canine, ovine, and fowl hosts, the different hexons largely group together. As noted previously , there are at least two distinct groups of bovine hexons. Bovine3 and -10 hexons group close to human hexons, whereas bovine4, -5, -6, -7, and -8 hexons lie elsewhere. The phylogenetic distances presented here are shorter than earlier measurements , as the current calculation includes all positions in the full-length hexon sequences. It is curious that turkey and duck hexons are less closely related to the fowl hexon group than to the frog hexon and the main bovine hexon group, respectively. While predictions of the origins of the adenoviruses should become clearer as more sequences are examined, at present the hexon data suggest that they form at least three and possibly four distinct groups that are evolving independently. It is clear from Fig. that the Hominidae adenoviruses evolved from a common ancestor, with the closest separate group including adenoviruses infecting horses, pigs, cows, dogs, and mice. More distant is the group infecting ducks, cows, deer, goats, and sheep. The turkey and frog adenoviruses appear to be members of a third group, with the fowl adenoviruses as the most distant group. These conclusions are strengthened by the fact that the results of the distance-based analyses are consistent with those of the parsimony analysis , as the two methods rely on different assumptions. Although the phylogenetic tree shown in Fig. is not based on extensive data, it indicates that the evolution of the animal adenoviruses parallels that of their hosts. For example, adenoviruses infecting Hominidae (humans, chimpanzees) or Artiodactyla (cloven-hooved animals, such as cows, deer, goats, and sheep) cluster together. It further suggests that horizontal transfer is not common, despite the intimate contact between humans and animals such as chickens. The extent to which adenovirus transfer can occur between closely related hosts, such as chimpanzees and humans, is still unclear. DISCUSSION : HVRs and epitopes. | To make reproducible assignments for the HVRs, we introduced a quantitative approach. Seven HVRs in the hexon sequence were described by Crawford-Miksza and Schnurr . The present analysis of the positions of the most divergent residues for 40 full-length hexon sequences reveals nine HVRs. In general, the positions of the newly defined HVRs correlate with those defined previously, but the greater number of sequences and the use of hexon structural data to guide their alignment has allowed a more precise definition of these regions. We previously showed that the HVRs lie in the extended loops at the top of the molecule . This is in agreement with the present analysis, in which the DE1 loop contains HVR1 to -6 and the FG1 loop contains HVR7 to -9. The alignment positions of the current (previous) HVR assignments are as follows: HVR1, 146 to 181 (144 to 197); HVR2, 199 to 221 (206 to 216); HVR3, 264 to 270 (235 to 242); HVR4, 288 to 293 (282 to 296); HVR5, 307 to 330 (302 to 333); HVR6, 358 to 366 (355 to 375); HVR7, 482 to 489, HVR8, 503 to 513, and HVR9, 518 to 522 (HVR7, 489 to 525). The criteria used for the new assignments are more stringent than those used earlier and define the HVRs more tightly. In consequence, they focus on "hot spots" where sequence modifications are likely to be possible without structural disruption. The new assignments for HVR1, -4, -5, and -6 are subsets of those defined earlier. Similarly, it is now clear that the earlier HVR7 should be defined more precisely as HVR7, -8, and -9. The one exception to the general pattern of agreement with the previous assignments is HVR3. Its original definition places it in a well-conserved segment in the current sequence alignment . In the absence of evidence to physically map virus-specific neutralizing epitopes, it is reasonable to assume that they lie within one or more of the HVRs. The HVRs lie within the loops at the top of the hexon molecule that form the virion surface, and so it is not unexpected that the most significant differences between the hexon sequences are at these locations. Recently, Moffat et al. showed that rabbit and mouse anti-Ad5, antibovine (bovine3), and antiporcine (porcine3) sera do not cross-react. This finding shows that their neutralizing epitopes differ significantly and is consistent with the idea that the epitopes must lie in positions of sequence variation ---particularly as bovine3 and porcine3 are quite closely related . In an early effort to engineer the adenovirus, Crompton et al. inserted an 8-amino-acid sequence from the major antigenic site in the poliovirus type 3 VP1 capsid protein into two regions of Ad2 hexon. Antisera specific for the poliovirus sequence efficiently neutralized the modified adenovirus, and antisera raised against the modified adenovirus recognized the VP1 capsid protein of poliovirus type 3. Substitution into residues 284 to 291 of the DE1 loop (corresponding to alignment positions 309 to 330 in HVR5; Fig. ) resulted in a virus with wild-type growth characteristics, whereas substitution into residues 442 to 449 of the FG1 loop for Ad2 hexon (alignment positions 499 to 506) yielded viruses that grew poorly. This result was difficult to understand earlier, as the FG1 loop insertion site is within the region previously assigned to HVR7 . In contrast, the insertion extends beyond the range now assigned to HVR8. This suggests that the modification lay outside the most malleable region of hexon and so altered parts of its structure involved in virion formation. It is very encouraging that the more precisely defined HVRs presented here explain the earlier results. The explanation suggests that the new definitions will also be more accurate in predicting which hexon sites are amenable to future modification. The topology of the hexon monomer , with two axially oriented jellyrolls and long loop extensions, suggests a flexible structure capable of supporting a wide variety of sequence lengths. However, in the native trimeric form , the jellyrolls are closely packed and the long loops that emerge between jellyroll strands are tightly wrapped around one another to form compact tower domains. In addition, the hexon base interacts closely with other protein partners in the virion. Due to these severe structural constraints, hexon sequences longer than that for Ad2 are likely to arise from insertions into one or more of the HVRs. Crystallography. | In our earlier crystallographic analysis of the Ad5 hexon , we applied new crystallographic methods, including maximum-likelihood refinement and the modeling of bulk solvent, that were not available at the time that the Ad2 hexon structure was first determined . The results suggested that a reevaluation of Ad2 would be rewarding . The present study shows that this was not only worthwhile but that the Ad5 structure could be improved still further. Initially, the methods used in the Ad5 hexon structure determination improved the Ad2 model so it was of similar quality to that for Ad5 (PDB identifier, 1RUX). Subsequently, still further improvements were made in crystallographic refinement procedures. The program CNS can now combine simulated-annealing and torsion angle dynamics with refinement against the maximum-likelihood target function developed for TNT-ML . Application of the latest methods has now produced excellent models for both the Ad2 and Ad5 hexons. This study shows that determining crystal structures for closely related molecules can be rewarding. Only by determining the Ad5 hexon structure was the necessity to revisit the Ad2 structure made apparent. Likewise, this process suggested further improvements for Ad5. The quality of the current models is such that no further improvements are anticipated with the available diffraction data. Hexon stabilization. | While monitoring the quality and stability of recombinant adenovirus preparations by anion exchange high performance liquid chromatography, Blanche et al. noted that the retention time of the hexon peak increased with storage time. They showed that the shift corresponded to an increase in isoaspartate residue formation and thus resolved the long-standing puzzle of why chromatograms exhibit a trailing shoulder in the hexon peak . To reduce the rate of Asn deamidation, Blanche et al. changed all four accessible Asn residues engaged in Asn-Gly pairs in the Ad5 hexon to Leu (Asn 243, Asn 254, and Asn 436) or Ala (Asn 275). These modifications reduced the microheterogeneity in the recombinant adenovirus without adversely affecting virus infectivity. In the Ad5 hexon X-ray model, only Asn 243 in loop DE1 has interpretable density, while the other three asparagines are located in disordered ---or flexible ---segments of loops DE1 and FG1. The other 39 full-length hexon sequences do not have Asn residues at these positions, but all have other Asn-Gly pairs. For example, the chimpanzee AdC68 hexon has Asn's forming Asn-Gly pairs at positions 414, 486, and 533 in loop FG1. The homology model indicates that each of these is surface accessible and so would be subject to isoaspartate formation. For optimum stability, appropriate modifications should be made to all such Asn-Gly pairs when designing adenovirus vectors for clinical use. Adenovirus-based vector design. | Adenoviruses are maintained episomally in infected cells and do not integrate with the host genome. Consequently, the transgene is lost as cells infected with a recombinant adenovirus vector turn over so that repeated administrations of vector are necessary. Problems then arise with preexisting or acquired antiadenoviral immunity. A conservative approach to overcoming this obstacle is to design and use adenovirus-based vectors with modified hexons. By far the most common serotype of adenovirus-based vectors for human gene therapy is Ad5. A recent study by Youil et al. lists several reasons why this is so. First, Ad5 is the most extensively studied serotype. Vectors based on another serotype would require additional studies to determine safety and tropism profiles before human use. Second, sequencing and annotation would be required before new replication-incompetent adenovirus vectors with deleted early genes could be designed. Third, alternative serotypes have an uncharacterized potential for targeting different cell types, due to variability in the cellular receptors and binding strategies used for attachment. Fourth, serotypes outside human subgroup C are likely to require the generation of new complementing cell lines to support their propagation. As described above, the insertion of an 8-amino-acid peptide corresponding to a poliovirus epitope into the Ad2 hexon by Crompton et al. shows that the hexon molecule is amenable to careful modification. The most comprehensive hexon modification made to an Ad5-based vector was to replace the entire hexon with that of another serotype, Ad2 . Expression levels were monitored after intravenous injections of an Ad5-based vector containing the gene for liver chloramphenicol acetyltransferase. Expression was reduced only 100-fold in animals preimmunized with an Ad2 hexon-bearing Ad5 chimera, as opposed to 10,000-fold when an Ad5 vector was used for preimmunization. Similar results were observed when the FG1 loop of Ad5 was replaced with that of Ad2 . Attempts to create an Ad7 hexon-bearing Ad5 chimera were unsuccessful, which indicated that some residues cannot be readily substituted. In another study, Roy et al. created a viable, albeit lower-titer, Ad12-Ad5 chimera by replacing all four Ad5 hexon loops (i.e., DE1, FG1, DE2, and FG2) with those of Ad12 (subgroup A). More recently, Youil et al. have expanded these studies by creating chimeric recombinant Ad5 constructs in which the hexon gene was replaced with that from several serotypes in subgroups A through E. Of the 18 constructs created, only those with the Ad1, Ad2, Ad6, and Ad12 hexon genes formed viable viruses. Ad1, Ad2, and Ad6 (like Ad5) are all in subgroup C. Only one non-subgroup-C hexon (i.e., Ad12, subgroup A) was able to substitute for Ad5 hexon. Viable vectors were not formed by a second subgroup A hexon (Ad18), or by four from subgroup B (Ad7, -11, -16, and -35), eight from subgroup D (Ad9, -10, -13, -15, -17, -19, -27, and -37), and one from subgroup E (Ad4). These authors echo the suggestion by Gall et al. that the failure to produce viable virions is most likely due to structural incompatibilities between the alternate hexon proteins and the 100-kDa scaffolding protein, which is required for assembly, or with the other capsid proteins (i.e., penton, fiber, and protein IX) . Summary. | We have applied crystallography, molecular modeling, and phylogenetic techniques to determine the conserved core of the hexon molecule and to locate sites that are most likely amenable to modification. The approach used here shows how excellent results can be achieved by combining both experimental and theoretical structural data with a heuristic alignment algorithm. The resultant sequence alignment uses all currently available data and produces phylogenies that are consistent with experimental groupings of adenovirus serotypes. The current alignment includes a wider range of hexon sequences than examined previously and allows a more precise definition of the HVRs. The new definitions now explain experimental results that were published previously. They show that the hexon molecule has a conserved core that forms the capsid and dictates its size and so should not be altered. The nine newly defined HVRs on the molecular surface are primary candidates for hexon modification. FIG. 1. : Hexon structure. Hexon structure. Ribbon representation of the Ad2 hexon subunit. The view is perpendicular to the molecular threefold axis from the inside of the molecule. The top of the molecule, which contains the loops (DE1, FG1, and FG2), forms the outer surface of the viral capsid. The hexon base contains a small loop (DE2) and two eight-stranded "viral" jellyrolls (V1 and V2), which are separated by the connector, VC. The eight jellyroll beta strands are labeled B to I. The N-terminal loop, NT, lies underneath the base. The secondary structural elements are labeled, as are the positions of chain segments that are not observed in the crystallographic model (<...>). The figure was made with MOLSCRIPT . FIG. 2. : Hexon sequence alignment. Hexon sequence alignment. Multiple-sequence alignment of 40 hexon sequences calculated with Clustal_X . An alignment profile based on the structural alignment of two hexon crystal structures and three homology models was used to guide the multiple-sequence alignment. The default Clustal_X color scheme indicates significant features. All G (orange) and P (yellow) are colored. Frequent occurrences of a property at a sequence position are colored: hydrophobic, blue; hydrophobic tendency, light blue; basic, red; acidic, purple; hydrophilic, green; unconserved, white. Conserved positions are marked as complete (*), strong (:), or weak (.). A histogram indicates the conservation level, and the eight molecular regions are labeled. Previously assigned HVRs (HVR1 to -9) are boxed in yellow, and the new assignments (HVR1 to -9) are boxed in red. A Clustal_X alignment file is available for download at . FIG. 3. : Hexon trimer colored by sequence diversity. Hexon trimer colored by sequence diversity. Layers of the trimeric hexon molecule displayed as ribbons. The two columns show hexon slices viewed along the molecular threefold axis. The arrows indicate the direction of view from either the virion exterior top to bottom (left) or from the virion interior bottom to top (right). The structures are colored by sequence diversity from blue (conserved) to yellow (less conserved) to red (not conserved). The jellyrolls V1 and V2 and their alpha1 helices are labeled. Note that the size of the hexon's pseudohexameric base is dictated by structural elements, such as V1-alpha1 and V2-alpha1, which separate the jellyrolls. A PDB file of the hexon trimer with temperature factors (B factors) representing sequence diversity is available for download at . FIG. 4. : Phylogenetic analysis of adenovirus hexon. Phylogenetic analysis of adenovirus hexon. Phylogenetic trees of 40 full-length hexon sequences calculated with the Phylip program package . (A) An unrooted tree generated by parsimony analysis with PROTPARS and CONSENSE. (B) An unrooted tree generated by distance matrix analysis with PROTDIST (Dayhoff's PAM 001 scoring matrix), FITCH (global rearrangements option), and CONSENSE. Branch lengths are proportional to the number of substitutions, and the scale bar represents 10 mutations per 100 sequence positions. Bootstrap values for the 100 trials are indicated for each branch. (C) A close-up view of the Hominidae adenovirus hexon branches. TABLE 1 : Hexon refinement statistics TABLE 2 : Hexon regions TABLE 3 : Hexon secondary structure Backmatter: PMID- 12915574 TI - Antiviral Activity of Limitin against Encephalomyocarditis Virus, Herpes Simplex Virus, and Mouse Hepatitis Virus: Diverse Requirements by Limitin and Alpha Interferon for Interferon Regulatory Factor 1 AB - Limitin has sequence homology with alpha interferon (IFN-alpha) and IFN-beta and utilizes the IFN-alpha/beta receptor. However, it has no influence on the proliferation of normal myeloid and erythroid progenitors. In this study, we show that limitin has antiviral activity in vitro as well as in vivo. Limitin inhibited not only cytopathic effects in encephalomyocarditis virus- or herpes simplex virus (HSV) type 1-infected L929 cells, but also plaque formation in mouse hepatitis virus (MHV) type 2-infected DBT cells. In addition, administration of limitin to mice suppressed MHV-induced hepatitis and HSV-induced death. The antiviral activity may be mediated in part by 2',5'-oligoadenylate synthetase, RNA-dependent protein kinase, and Mx protein, which inhibit viral replication or degrade viral components, because limitin induced their mRNA expression and enzyme activity. While limitin has antiviral activity as strong as that of IFN-alpha in vitro (the concentration that provided 50% inhibition of cytopathic effect is similar30 pg/ml), IFN regulatory factor 1 (IRF-1) dependencies for induction of an antiviral state were different for limitin and IFN-alpha. In IRF-1-deficient fibroblasts, a higher concentration of limitin than of IFN-alpha was required for the induction of antiviral activity and the transcription of proteins from IFN-stimulated response element. The unique signals and the fewer properties of myelosuppression suggest that a human homolog of limitin may be used as a new antiviral drug. Keywords: Introduction : Interferons (IFNs) are divided into two major groups of biologically active proteins. IFN-alpha, IFN-beta, IFN-omega, and IFN-tau, which belong to the IFN-alpha/beta family, have structural homology to each other and share a common receptor . IFN-gamma is a product of activated T lymphocytes and natural killer cells . Although only small amounts of IFNs are produced under steady-state conditions, these can be markedly increased by virus infections or by exposure to double-stranded nucleic acids . IFNs are highly pleiotropic cytokines with potent immunoregulatory and antiproliferative properties in addition to their antiviral activities . IFNs regulate immune responses, such as major histocompatibility antigen expression, cytokine production, and T-lymphocyte activation . IFNs suppress the proliferation of many normal and transformed cells through prolongation of the G1 phase of the cell cycle and reduce the rate of entry into the S phase, as well as inducing apoptosis . The antiviral activity of IFNs results in part from induction of the enzymes 2',5'-oligoadenylate synthetase (2',5'-OAS) and RNA-dependent protein kinase (PKR) and of Mx protein, which inhibit viral replication or degrade viral components . These broad biological properties have suggested many clinical applications in cancer and virologic diseases . Limitin is an IFN-like cytokine that we recently cloned on the basis of its ability to arrest the growth of or kill a myelomonocytic leukemia cell line . The initial cDNA clone carried a 997-bp insert (EMBL-GenBank-DDBJ accession number ) whose deduced protein is composed of 182 amino acid residues. It contains a signal peptide of 21 amino acids at the N-terminal end and one N-linked glycosylation site at amino acid residue 68 but lacks an internal transmembrane domain. In addition, limitin protein was detected in the supernatants of transfectants with an expression plasmid of limitin. Thus, limitin seems to be a secreted glycoprotein. Analysis with the translated protein sequence reveals that limitin has similar30% identity with IFN-alpha, IFN-beta, and IFN-omega. Computer modeling of limitin structure shows an IFN-like globular structure composed of five long alpha-helices and one short helix in the middle of the loop connecting helices B and C, with possible disulfide bonds between residues 52 and 157 and between residues 80 and 130 . In addition, receptor-binding assays and experiments using IFN-alpha/beta receptor-deficient cells have indicated that limitin binds to and utilizes the IFN-alpha/betareceptor . Thus, limitin seems to be a new member of the IFN-alpha/beta family. As has been reported for previously known IFNs, limitin inhibits not only B lymphopoiesis but also proliferation of a variety of lympho-hematopoietic cell lines . Limitin enhances major histocompatibility antigen expression on antigen-presenting cells, as well as the killer activity of cytotoxic T lymphocytes . However, limitin is unique among IFNs because it has no influence on the proliferation of normal myeloid and erythroid progenitors . Although limitin transduces signals via the IFN-alpha/beta receptor, its signals could be different from those of IFN-alpha and IFN-beta. In this study, we show that limitin has antivial activity, a major biological function of IFNs, and we clarify the difference between signals induced by limitin and IFN-alpha. MATERIALS AND METHODS : Cells and mice. | A murine myeloid leukemia cell line (WEHI3), a murine fibroblast cell line (L929), and a murine brain tumor cell line (DBT) were maintained in Dulbecco's modified Eagle's medium (Nikken, Kyoto, Japan) supplemented with 10% fetal calf serum (Flow, Aurora, Ohio). Cultured fibroblasts were generated from murine lungs and maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. IFN regulatory factor 1 (IRF-1)-deficient mice were kindly provided by T. Taniguchi (Tokyo University, Tokyo, Japan). C57BL/6 mice were purchased from Japan Clea (Tokyo, Japan) and used as control wild-type mice for IRF-1-deficient mice. BALB/c mice were purchased from Nippon SLC (Shizuoka, Japan). All mice were maintained under specific-pathogen-free conditions at the Institute for Experimental Animals, Osaka University. The mice were similar6 to 10 weeks of age at the time of use. Viruses and cytokines. | Encephalomyocarditis virus (EMCV) and L929 cells were kindly provided by M. Kohase (National Institute of Infectious Disease, Tokyo, Japan). EMCV was propagated in L929 cell cultures. Human herpes simplex virus type 1 (HSV-1) was isolated from the vesicle of a patient and propagated in L929 cell cultures. Mouse hepatitis virus type 2 (MHV-2) and DBT cells were kindly provided by F. Taguchi (National Institute of Neuroscience, Tokyo, Japan). MHV was propagated in DBT cell cultures . EMCV and HSV were titrated by the cytopathic effect (CPE) dye uptake method, and MHV was titrated by plaque assay. Recombinant murine IFN-alpha was purchased from GIBCO BRL (Grand Island, N.Y.). The production and purification of recombinant limitin and anti-limitin polyclonal antibody were described in a previous article . The fusion proteins limitin-immunoglobulin (Ig) and CD44-Ig were prepared as described previously . Gel filtration chromatography. | Analysis of purified recombinant limitin was performed by gel filtration chromatography using a Superdex 75 PC 3.2/30 column (Amersham Biosciences, Piscataway, N.J.). A mobile phase consisted of 20 mM Tris HCl, pH 8.0, and 150 mM sodium chloride. Eluted protein was detected by UV light at 280 nm. An LMW calibration kit (Amersham Biosciences) was used as a standard protein marker. [3H]thymidine incorporation assay. | WEHI3 cells were incubated in flat-bottom 96-well microplates (Corning Costar, Tokyo, Japan) for 48 h. The cells were pulsed with 0.5 muCi of [3H]thymidine (Amersham, Tokyo, Japan)/well for the last 4 h of culture and then harvested onto glass filters (Wallac, Turku, Finland) with a semiautomatic cell harvester (Pharmacia), and incorporated radioactivity was measured with a liquid scintillation counter. CPE dye uptake method. | Antiviral activity was measured by the CPE dye uptake method with minor modifications . Briefly, confluent monolayers of L929 cells were exposed to serial dilutions of limitin or IFN-alpha overnight, challenged with either EMCV or HSV until 75 to 95% CPE was evident, and stained with naphthol black solution (0.1% naphthol blue-black, 0.1% acetic bicarbonate, 9% acetic acid). The bound dye in the wells of the microtiter plates was eluted with 100 mM NaOH, and the optical density at 650 nm (OD650) was measured colorimetrically using a spectrophotometer (Emax precision microplate reader; Molecular Devices). Plaque assay. | DBT cell monolayer cultures on six-well dishes were washed with phosphate-buffered saline (PBS) and inoculated with MHV for 1 h. The cultures were overlaid with minimal essential medium containing 1% Bacto Agar (Difco, Detroit, Mich.) and 10% tryptose phosphate broth. After 48 h of incubation, each well was stained by a second overlay containing 1:10,000 neutral red for 12 h, and then the plaques were counted . Northern blot analysis. | Total RNAs were isolated using a TRIZOL reagent (Invitrogen, Carlsbad, Calif.). The RNAs (15 mug/lane) were electrophoresed through a formaldehyde-agarose gel and transferred onto a nylon membrane (Amersham, Arlington Heights, Ill.). The cDNA fragments were labeled with [32P]dCTP using a randomly primed DNA-labeling kit (Boehringer Mannheim, Indianapolis, Ind.) and hybridized to the membrane. The blots were then washed and autoradiographed. The cDNA fragment of Mx1 was kindly provided by O. Haller (Freiburg University, Freiburg, Germany), and that of IRF-1 was a kind gift of Y. Kanakura (Osaka University, Osaka, Japan). The cDNA fragments of PKR and 2',5'-OAS were amplified by PCR with the following primers: sense (5'-TTAGGTGGATTTGGTCAA-3') (895 to 912) and antisense (5'-TCGGTCCTTGGGTTTCTCTG-3') (1600 to 1617) for PKR and sense (5'-GTCTCCAAGGTGGTGAAGGG-3') (201 to 220) and antisense (5'-TACTCAAGTTGTGTCGGGTC-3') (781 to 800) for 2',5'-OAS. 2',5'-OAS activity. | Measurement of 2',5'-OAS activity was performed using a 2',5'-oligoadenyl-5'-triphosphate (2-5A) radioimmunoassay kit (Eiken Chemical Co., Tokyo, Japan) according to the manufacturer's instructions and with some modifications described elsewhere . Briefly, L929 cells (105) were seeded in 24-well plates and stimulated with limitin for the appropriate time. The cells were then suspended in 1 ml of lysis buffer consisting of 10 mM HEPES (pH 7.5), 50 mM KCl, 3 mM Mg(CHOO)2, 0.3 mM EDTA, 1 mM 2-mercaptoethanol, 20% glycerol, and 0.5% NP-40 and centrifuged at 10,000 x g for 5 min at 4C. Then, 50 mul of supernatant medium and cell lysates was mixed with 50 mul of poly(rI) poly(rC) agarose beads and then incubated for 10 min at room temperature. After centrifugation at 400 x g for 5 min, the pellets were resuspended with 500 mul of ATP solution and incubated at 37C for 1 h. To the mixture was added 100 mul of 125I-labeled 2-5A solution to quantify the 2-5A produced by enzymatic reaction of 2',5'-OAS and ATP. The mixture was incubated at 37C for 1 h and centrifuged at 800 x g for 30 min. After the beads were washed, their radioactivity was measured with a well-type gamma counter (WALLAC WIZARD 1470). The amounts of 2-5A synthesized by 2',5'-OAS were calculated from the standard curve. Infection. | Mice were infected intraperitoneally with an appropriate dose of virus and then treated intraperitoneally with limitin or PBS. The procedures were in compliance with the guidelines set by the institute for care and use of laboratory animals. GPT enzyme activity. | The concentrations of glutamate pyruvate transaminase (GPT) in the sera of MHV-infected mice were measured using a Rikitech GPT IFCC assay kit (Roche Diagnostics, Tokyo, Japan). Histopathological analysis. | Histopathological examinations were carried out using paraffin-embedded sections of livers from mice infected with MHV. The sections were deparaffinized and stained with hematoxylin-eosin. Luciferase assay. | Luciferase assays were performed using the Dual-Luciferase Reporter System (Promega, Madison, Wis.), in which transfection efficiency was monitored by cotransfected pRL-SV40 (Promega), an expression vector of Renilla luciferase. The cultured cells were transfected with 500 ng of pCIS-ISRE (Stratagene, La Jolla, Calif.), together with 50 ng of pRL-SV40, by the Lipofectamine transfection method (Invitrogen). The transfected cells were lysed in buffer supplied by the manufacturer, followed by measurement of the firefly and Renilla luciferase activities on a luminometer (LB96P; Berthold Japan, Tokyo, Japan). The relative firefly luciferase activities were calculated by normalizing transfection efficiencies according to the Renilla luciferase activities. RESULTS : Recombinant limitin protein. | We made recombinant limitin protein from a stably limitin-producing Chinese hamster ovarian cancer cell line as previously described . When the purity of recombinant protein was evaluated by gel filtration chromatography, only one large peak whose molecular mass was similar20 kDa was seen . The activity of recombinant limitin protein was also confirmed by the previously clarified limitin function . Limitin suppressed the proliferation of WEHI3 cells in a dose-dependent manner, and the activity of limitin was stronger than that of the fusion protein limitin-Ig at the same concentration . Therefore, our recombinant protein, whose purity is >98%, has higher specific activity than limitin-Ig. In this study, we used this lot of recombinant limitin protein to analyze the antiviral activity of limitin. Limitin has antiviral activity in vitro. | When L929 cells were infected with EMCV, CPE was detected within 24 h. As shown in Fig. , CPE after EMCV infection was inhibited by limitin in a dose-dependent manner. The antiviral activity of limitin was evident at 15 pg/ml and maximal at 100 pg/ml, and 30 pg of limitin/ml provided 50% inhibition of CPE, which was similar to that of the control IFN-alpha. The inhibition of CPE by limitin was completely cancelled by the addition of anti-limitin polyclonal antibody . In contrast, the anti-limitin antibody could not cancel the inhibition of CPE by IFN-alpha. These facts indicate that the anti-limitin antibody is specific to limitin and that limitin truly induces an antiviral state in L929 cells. Although CPE of L929 cells was observed 72 h after HSV infection, limitin exhibited dose dependency for antiviral activity similar to that of IFN-alpha, and the dose of both limitin and IFN-alpha that provided 50% inhibition of CPE against HSV was also similar30 pg/ml . Limitin, as well as IFN-alpha, also inhibited plaque formation by MHV in DBT cells . Therefore, limitin displays antiviral activity in vitro against three kinds of viruses: EMCV, HSV, and MHV. The concentration (30 pg/ml) of limitin and IFN-alpha that provided a 50% inhibition of CPE in L929 cells (with EMCV as the challenge virus) was defined as 1 experimental unit (EU)/ml in the subsequent experiments . In vivo antiviral activity of limitin. | To investigate the antiviral activity of limitin in vivo, limitin was injected intraperitoneally every day into MHV-infected mice . The mice injected with PBS, but not those injected with limitin, started to die within 5 days after MHV infection (data not shown). We evaluated the GPT concentrations in serum with the GPT IFCC assay and analyzed hepatitis histologically 3 days after MHV infection. In mice injected with PBS, the GPT concentrations in serum were markedly elevated , and their liver sections revealed massive infiltration of lymphocytes, massive necrosis of hepatocytes, and intrahepatic hemorrhage . In mice injected with limitin, neither elevation of the GPT concentration in serum nor the hepatitis effects in liver sections were observed. When BALB/c mice were infected with HSV , they started to die within 7 days after infection, and similar50% of the mice in the PBS-treated group died during our observation period . In the limitin-treated group, all mice survived throughout the period. HSV was then inoculated into athymic nude mice . Although limitin could not rescue the HSV-infected mice, administration of limitin apparently prolonged their survival compared with that with PBS injection . Therefore, limitin injection produces antiviral status against MHV, as well as HSV, in mice. Induction of antivirus-related genes. | PKR activates eIF-2alpha, which inhibits mRNA translation . 2',5'-OAS activates RNase L, leading to RNA degradation, by producing 2-5A . Mx1 is a GTPase that affects the intracellular transport of viral particles . Whole-cell lysates of L929 cells stimulated with limitin contained approximately twice the 2',5'-OAS enzyme activity of controls (0.24 +- 0.05 pmol/dl/cell/h with PBS versus 0.55 +- 0.04 pmol/dl/cell/h with limitin at 4 h). In addition, the enhanced enzyme activity was also detected in the supernatants of limitin-stimulated L929 cells (data not shown). Message of 2',5'-OAS was also induced after limitin stimulation, and it was enhanced till 16 h after the stimulation . Although PKR mRNA consists of three transcripts of 6.0, 4.0, and 2.5 kbp , only 6-kbp mRNA was induced by limitin and IFN-alpha in L929 cells. Only weak message of Mx protein was induced within 8 h after limitin stimulation (data not shown), because L929 cells were derived from C3H/He mice, which belong to the phenotypically Mx- strain . When L929 cells were stimulated with limitin, messages of 2',5'-OAS and PKR were induced as strongly as with IFN-alpha . Therefore, limitin induces 2',5'-OAS, PKR, and Mx proteins, which mainly mediate an antiviral state via IFNs . Different requirements for IRF-1 by limitin and IFN-alpha for induction of antiviral activity. | Activation of signal transducer and activator of transcription 1 (STAT1) is essential for the antiviral activity of IFNs . A heterotrimer composed of STAT1, STAT2, and IRF-9 binds to the interferon-stimulated response element (ISRE) that exists in the promoters of 2',5'-OAS, PKR, and Mx1 genes . The STAT1 homodimer is another mediator, and it transduces IRF-1 , which recognizes consensus promoter sequences in part overlapping the ISRE . Our Northern blotting revealed that limitin enhanced IRF-1 gene expression within 2 h in L929 cells (data not shown), and the message was induced more strongly than by IFN-alpha 4 h after stimulation . To further examine the contribution of the IRF-1-dependent pathway, we performed a CPE dye uptake assay using IRF-1-deficient fibroblasts. As shown in Fig. , the antiviral activity of limitin was impaired more severely than that of IFN-alpha in IRF-1-deficient fibroblasts (the doses that provided 50% inhibition of CPE were 3 EU/ml with limitin versus 3 EU/ml with IFN-alpha in wild-type fibroblasts and 7 to 8 EU/ml with limitin versus 3 to 4 EU/ml with IFN-alpha in IRF-1-deficient fibroblasts). The reduced antiviral activity of limitin was observed in fibroblasts established from six independent IRF-1-deficient mice . The relatively higher requirement for IRF-1 for the limitin-induced antiviral state was also confirmed by ISRE-luciferase reporter assays. IRF-1-deficient fibroblasts stimulated by limitin produced less luciferase protein from the ISRE sequence than those stimulated by IFN-alpha, while similar luciferase production was observed in wild-type fibroblasts . Therefore, IRF-1 is more critical for the limitin-mediated antiviral state than for the IFN-alpha-mediated antiviral state. DISCUSSION : A novel IFN-like cytokine, limitin, has in vitro and in vivo antiviral activity in part via induction of PKR, Mx protein, and 2',5'-OAS, which inhibit viral replication and degrade viral components. Although limitin has antiviral activity as strong as that of IFN-alpha, IRF-1 dependencies for induction of an antiviral state were different in limitin and IFN-alpha. These findings support the notion that a human homolog of limitin may be applicable as a new therapeutic drug. Mice lacking the IFN-alpha/beta receptor are abnormally sensitive to certain kinds of viruses, suggesting a critical role for IFN-alpha/beta cytokines in this type of immune defense . As described here, limitin has antiviral activity like those of IFN-alpha and IFN-beta. Limitin inhibited not only CPE in EMCV- or HSV-infected L929 cells, but also plaque formation in MHV-infected DBT cells, indicating that limitin displayed antiviral activity against at least three types of viruses: EMCV (RNA nonenveloped), HSV-1 (DNA enveloped), and MHV-2 (RNA enveloped) . We also showed that administration of limitin to mice suppressed MHV-induced hepatitis and HSV-induced death. These facts suggest that limitin may be applicable clinically as an antiviral drug. Interestingly, the antiviral effects of limitin were higher in BALB/c mice than in athymic nude mice. Thus, in addition to direct effects of antiviral activity, limitin seems to have indirect antiviral activity, such as induction of IFNs, enhancement of apoptosis of virus-infected cells, activation of natural killer cells, and induction of nitric oxide, which eliminate invading viruses in vivo. Viruses are intracellular pathogens, and IFNs are thought to constitute the first line of defense in host resistance to virus infections before immune mechanisms come into play . It is noteworthy that the antiviral effects of IFNs and limitin in vivo may be different. IFNs seem to inhibit the expansion and/or extension of virus infections, because IFN-alpha and IFN-beta are induced by double-stranded nucleic acids that are frequently produced as a result of viral replication . Previous data from immunohistological staining showing that limitin is produced constitutively in healthy mice may suggest that limitin mainly inhibits invasion of the body by viruses. This possibility may also be supported by the fact that the main limitin-producing cells are T lymphocytes and bronchial epithelial cells, a major entrance for viruses into the body. IFNs can act directly at various stages of the virus infectious cycle, including penetration, transcription, translation, and budding from the cell surface . Some IFN-inducible proteins, such as PKR and 2',5'-OAS, inhibit translation of viral mRNA, and 2',5'-OAS also leads to RNA degradation . Another IFN-inducible protein, Mx protein, inhibits virus replication . PKR, Mx protein, and 2',5'-OAS may participate in the limitin-induced antiviral state, because limitin can enhance these proteins in fibroblasts. The available information suggests that there are two major pathways for induction of gene expression of PKR, Mx protein, and 2',5'-OAS. In general, ligation of IFN-alpha and IFN-beta to the IFN-alpha/beta receptor causes phosphorylation of two Janus kinases, Jak1 and Tyk2, and the activated Jak1 then phosphorylates STAT1 and STAT2 . The phosphorylated Stat1-Stat2 complexes combine with IRF-9 and migrate to the nucleus, where they bind to DNA (the ISRE sequence) and initiate the transcription of PKR, Mx1, and 2',5'-OAS . Another important pathway is mediated by the Stat1-Stat1 homodimer, which binds to IFN-gamma-activated sequence, resulting in transcription of IRF-1 . IRF-1 recognizes consensus promoter sequences in part overlapping the ISRE . In addition, at least nine IRF members have been identified , and some interact with each other, resulting in modification of their transcriptional activities. Although signals from the IFN-alpha/beta receptor are complex, the important roles of Stat1, IRF-1, and IRF-9 in IFN-induced antiviral activity have been clarified by experiments using knockout mice . Our data clearly indicate that signals for induction of an antiviral state are different for limitin and IFN-alpha, while limitin has antiviral activity as strong as that of IFN-alpha. Transcriptional activity from ISRE sequence, as well as the antiviral activity of limitin, was impaired more severely than that of IFN-alpha in IRF-1-deficient fibroblasts. Thus, the IRF-1-dependent signaling pathway is more critical for the limitin-mediated antiviral state than for the IFN-alpha-mediated antiviral state. Although direct proof is lacking, the phosphorylation rates of Stat1 and Stat2 or the association rates of Stat1-Stat1 and Stat1-Stat2-IRF9 might be different for limitin and IFN-alpha. Alternatively, limitin might induce more or fewer signals that modify the activities of the IRF family than IFN-alpha. This is the first report showing that signals are different for limitin and IFN-alpha. While limitin transduces signals via the IFN-alpha/beta receptor , there are some precedents for the unique signals of limitin. Differences between IFN-alpha and IFN-beta affinities for the IFN-alpha/beta receptor have been described, as well as between subtypes of IFN-alpha. The beta-R1 gene is induced by IFN-beta but not by IFN-alpha . Association of a 95- to 100-kDa tyrosine-phosphorylated protein with the IFN-alpha/beta receptor was found in an IFN-beta- but not an IFN-alpha-treated myeloma cell line . On the other hand, recent advances in the structural and functional analysis of IFN-alpha/beta cytokines allow us to discuss the interactions between IFNs and the IFN-alpha/beta receptor. The three-dimensional crystal structures of murine IFN-alpha and IFN-beta revealed that they have a globular structure with five alpha-helices (A to E) and four loops . Experiments using monoclonal antibodies against IFNs and site-directed mutagenesis revealed the functional importance of the N-terminal half of the loop AB and the loop DE, together with the nearest segments of the helices D and E. The sequence identity between limitin and IFNs is still similar60%, even in the functionally important portions. The diversity of signals among IFN-alpha/beta cytokines, including limitin, might be explained by differences in their interactions with the IFN-alpha/beta receptor or by differences in their sites that bind to the IFN-alpha/beta receptor. IFN-alpha and IFN-beta are used for the treatment of patients with AIDS-related Kaposi's sarcoma, as well as chronic hepatitis B and C . IFN-alpha-treated patients have a significantly higher rate of HBe antigen seroconversion and suppression of hepatitis B virus DNA. IFN-alpha decreases the hepatitis C virus load and reduces the risk for development of hepatocellular carcinoma . In spite of wide acceptance of IFNs as therapeutic drugs, they produce some adverse reactions. Some patients chronically treated with IFNs experience weight loss, bone marrow suppression, alopecia, thyroid dysfunction, and depression and other psychiatric disorders . These severe problems sometimes require dose reduction or discontinuation of treatment. Another clinical problem is that several virus-derived proteins enable viruses themselves to evade the antiviral activities of IFN-alpha and IFN-beta. The conventional IFN monotherapy is effective in only a small fraction of patients with chronic hepatitis C. Furthermore, the hepatitis C virus titer often rebounds after cessation of IFN treatment. To increase the sustained response rate, increasing IFN dosage, prolonging the duration of IFN treatment, and/or combination therapy with other antiviral drugs, such as ribavirin, is often required, although adverse reactions to IFNs also increase. Limitin has some important and unique features described here and elsewhere. Limitin has fewer myelosuppressive properties than IFN-alpha, indicating that high-dose- and long-duration therapy with limitin may be done more safely than with IFN-alpha and IFN-beta. Limitin used different signals for induction of antiviral activity than IFN-alpha and IFN-beta, suggesting that limitin might be effective for IFN-alpha-resistant patients. A human homolog of limitin or an engineered cytokine with useful features of the limitin structure could be superior to IFN-alpha and IFN-beta in clinical utility. FIG. 1. : Characterization of recombinant limitin. Characterization of recombinant limitin. (A) Gel filtration profile of limitin on a high-performance liquid chromatography Superdex 75 PC 3.2/30 column precalibrated with the standard protein markers. Recombinant limitin (1.5 mug) was applied, and the flow rate was 75 mul/min. The arrows indicate the elution peaks of the marker proteins (albumin, 67.0 kDa [1]; ovalbumin, 43.0 kDa [2]; chymotrypsinogen A, 25.0 kDa [3]; and RNase A, 13.7 kDa [4]). (B) WEHI3 cells (2.5 x 103) were seeded in each well and stimulated with the indicated dose of recombinant limitin (solid circles), limitin-Ig (open circles), and CD44-Ig (open squares) for 48 h. The results represent the means +- standard deviations of triplicate cultures. Similar results were obtained in three independent experiments. FIG. 2. : Reduction of susceptibility of L929 cells to EMCV by limitin and IFN-alpha. Reduction of susceptibility of L929 cells to EMCV by limitin and IFN-alpha. (A) L929 cell monolayers on 96-well plates were pretreated with the indicated concentrations of limitin (solid circles) or IFN-alpha (open circles) overnight and challenged with EMCV (5 x 102 PFU/well). After 24 h of incubation, the cells protected from EMCV virulence were evaluated by the CPE dye uptake assay. The percent inhibition of CPE was calculated as 100 x [OD650 (each sample) - OD650 (virus control)/OD650 (cell control) - OD650 (virus control)]. The results represent the means +- standard deviations (SD) of triplicate cultures. Similar results were obtained in three independent experiments. (B) Limitin (1,000 pg/ml) or IFN-alpha (1,000 pg/ml) was incubated with the indicated antibodies (10 mug/ml) for 30 min on ice. L929 cells were cultured in the presence of the mixtures overnight and then challenged with EMCV (5 x 102 PFU/well). After 24 h of incubation, the cells protected from EMCV virulence were evaluated by the CPE dye uptake assay. The results represent the means +- SD of triplicate cultures. Similar results were obtained in three independent experiments. FIG. 3. : Antiviral effects of limitin and IFN-alpha against HSV and MHV. Antiviral effects of limitin and IFN-alpha against HSV and MHV. (A) L929 cell monolayers on 96-well plates were pretreated with the indicated concentrations of limitin (solid circles) or IFN-alpha (open circles) overnight and challenged with HSV (50 PFU/well). After 72 h of incubation, the cells protected from EMCV virulence were evaluated by the CPE dye uptake assay. The results represent the means +- standard deviations (SD) of triplicate cultures. Similar results were obtained in three independent experiments. (B) DBT cells were seeded in a six-well dish, and then the subconfluent monolayer was treated with 30 EU of limitin/ml or 30 EU of IFN-alpha/ml for 24 h. The pretreated DBT cells were infected with MHV (103 PFU/well) for 1 h, followed by a plaque assay. The results represent the means +- SD of triplicate cultures. Similar results were obtained in three independent experiments. FIG. 4. : Repression of MHV infection in vivo by limitin. Repression of MHV infection in vivo by limitin. (A) Ten BALB/c mice were divided into two groups. One group was injected with 3 x 104 EU of limitin/day. The other group was treated with PBS as controls. The mice were treated intraperitoneally with PBS or limitin on days -1, 0, 1, and 2 relative to MHV exposure (103 PFU/mouse). After 72 h of virus inoculation, the mice were anesthetized with pentobarbital and bled to measure the activity of GPT in serum. Each dot represents the GPT concentration in the indicated mouse. Similar results were obtained in two independent experiments. (B) Two mice from each group were anatomized, and their livers were investigated histologically. Magnification, x100. FIG. 5. : Prolongation of survival of HSV-challenged mice by limitin. Prolongation of survival of HSV-challenged mice by limitin. (A) Twenty-four BALB/c mice were separated into two groups and challenged with HSV (5 x 104 PFU/mouse). The mice were treated with 3 x 104 EU of limitin (solid circle) or PBS (open circles)/mouse on days -1, 0, 1, 2, 3, 4, and 5 relative to HSV exposure. The duration of observation for survival was 14 days. Similar results were obtained in two independent experiments. (B) Eighteen athymic nude mice were separated into two groups and challenged with HSV (104 PFU/mouse). The mice were treated with 3 x 104 EU of limitin (solid circles) or PBS (open circles)/mouse on days -1, 0, 1, 2, 3, 4, and 5 relative to HSV exposure. The duration of observation for survival was 14 days. Similar results were obtained in two independent experiments. Statistical analysis was performed by the Kaplan-Meier method, and the P value was calculated by the generalized Wilcoxon test. FIG. 6. : Induction of antiviral molecules by limitin. Induction of antiviral molecules by limitin. (A) Whole-cell lysates of L929 cells stimulated with limitin (30 EU/ml) for 4 h were prepared and subjected to measurement of 2',5'-OAS activity using a 125I radioimmunoassay kit. The results were represented as 2-5A production by each sample within 1 h. The assay range of this method is 10 to 810 pmol/dl. The results represent the means +- standard deviations of triplicate cultures. Similar results were obtained in three independent experiments. (B) Total RNAs of L929 cells stimulated with the indicated doses of limitin or IFN-alpha for 4 or 16 h were prepared and subjected to Northern blot analysis and probed with the indicated materials. beta-Actin was used as an internal control. Similar results were obtained in three independent experiments. FIG. 7. : Antiviral activities of limitin and IFN-alpha in fibroblasts established from wild-type and IRF-1-deficient mice. Antiviral activities of limitin and IFN-alpha in fibroblasts established from wild-type and IRF-1-deficient mice. (A) Fibroblast monolayers on 96-well plates were pretreated with the indicated concentrations of limitin (solid circles) or IFN-alpha (open circles) overnight and challenged with EMCV (5 x 102 PFU/well). After 24 h of incubation, the cells protected from EMCV virulence were evaluated by the CPE dye uptake assay. The results represent the means +- standard deviations (SD) of triplicate cultures. Similar results were obtained in three independent experiments. (B) Fibroblast monolayers on 96-well plates were pretreated with limitin (10 EU/ml) or IFN-alpha (10 EU/ml) overnight and challenged with EMCV (5 x 102 PFU/well). After 24 h of incubation, the cells protected from EMCV virulence were evaluated by the CPE dye uptake assay. The results represent the means +- SD of the results obtained from six sets of fibroblasts independently established from wild-type and IRF-1-deficient mice. FIG. 8. : ISRE-dependent luciferase activity after stimulation of limitin and IFN-alpha. ISRE-dependent luciferase activity after stimulation of limitin and IFN-alpha. The pCIS-ISRE plasmid was transfected into fibroblasts derived from wild-type or IRF-1-deficient mice. The transfectants were stimulated with the indicated concentrations of limitin or IFN-alpha for 6 h. The cells were lysed and subjected to measurement of luciferase activity. The data are expressed as the fold increase over the average for controls and represent the means +- standard deviations of triplicate assays. Similar results were obtained with four sets of fibroblasts independently established from wild-type and IRF-1-deficient mice. N.S., not significant. Backmatter: PMID- 12915535 TI - Association of the Herpes Simplex Virus Type 1 Us11 Gene Product with the Cellular Kinesin Light-Chain-Related Protein PAT1 Results in the Redistribution of Both Polypeptides AB - The herpes simplex virus type 1 (HSV-1) Us11 gene encodes a multifunctional double-stranded RNA (dsRNA)-binding protein that is expressed late in infection and packaged into the tegument layer of the virus particle. As a tegument component, Us11 associates with nascent capsids after its synthesis late in the infectious cycle and is delivered into newly infected cells at times prior to the expression of viral genes. Us11 is also an abundant late protein that regulates translation through its association with host components and contains overlapping nucleolar retention and nuclear export signals, allowing its accumulation in both nucleoli and the cytosol. Thus, at various times during the viral life cycle and in different intracellular compartments, Us11 has the potential to execute discrete tasks. The analysis of these functions, however, is complicated by the fact that Us11 is not essential for viral replication in cultured cells. To discover new host targets for the Us11 protein, we searched for cellular proteins that interact with Us11 and have identified PAT1 as a Us11-binding protein according to multiple, independent experimental criteria. PAT1 binds microtubules, participates in amyloid precursor protein trafficking, and has homology to the kinesin light chain (KLC) in its carboxyl terminus. The carboxyl-terminal dsRNA-binding domain of Us11, which also contains the nucleolar retention and nuclear export signals, binds PAT1, whereas 149 residues derived from the KLC homology region of PAT1 are important for binding to Us11. Both PAT1 and Us11 colocalize within a perinuclear area in transiently transfected and HSV-1-infected cells. The 149 amino acids derived from the KLC homology region are required for colocalization of the two polypeptides. Furthermore, although PAT1 normally accumulates in the nuclear compartment, Us11 expression results in the exclusion of PAT1 from the nucleus and its accumulation in the perinuclear space. Similarly, Us11 does not accumulate in the nucleoli of infected cells that overexpress PAT1. These results establish that Us11 and PAT1 can associate, resulting in an altered subcellular distribution of both polypeptides. The association between PAT1, a cellular trafficking protein with homology to KLC, and Us11, along with a recent report demonstrating an interaction between Us11 and the ubiquitous kinesin heavy chain (R. J. Diefenbach et al., J. Virol. 76:3282-3291, 2002), suggests that these associations may be important for the intracellular movement of viral components. Keywords: Introduction : Juxtaposed between the nucleocapsid and the envelope lies the tegument, an electron-dense, structured region within herpesvirus particles (reviewed in references and ). Included among the RNA molecules and at least 15 polypeptides that populate this zone in HSV-1 virions is the product of the Us11 gene. As a tegument component, Us11 is delivered into the cytosol after fusion of the viral envelope with the host cell plasma membrane . Once inside the host cell, viral tegument components are free, in certain cases, to perform a diverse array of tasks prior to the onset of viral gene expression. Some tegument components associate with strategic host proteins, thereby modifying the cellular environment such that it supports more efficient viral replication. For example, to facilitate the loading of cellular ribosomes onto viral mRNAs, the HSV-1 Vhs nuclease selectively destabilizes mRNAs through an association with the translation initiation factor eIF4H . Similarly, VP16, a potent transcriptional transactivator, assembles into a complex, along with the cellular transcription factors HCF-1 and Oct-1, to stimulate transcription from viral immediate-early promoters . Although introduced into the cytoplasm extremely early in the infectious cycle, Us11, like many tegument proteins, is synthesized late in the viral life cycle. In addition, since Us11 accumulates both in nucleoli and in the cytosol, it has the potential to perform discrete actions at different times in various cellular compartments . Deciphering the functions executed by Us11 has proved difficult, since genetic analysis has revealed that the Us11 gene product is not required for productive viral growth in cultured cells . A role for Us11 in the regulation of viral protein synthesis emerged from the genetic analysis of HSV-1 gamma34.5 mutants. In a variety of cultured cell lines, gamma34.5 mutants grow poorly due to the premature cessation of translation prior to the completion of the viral life cycle . In the absence of the gamma34.5 gene product, the activated cellular PKR kinase phosphorylates the translation initiation factor eIF2alpha, inhibiting the initiation of protein synthesis . Extragenic suppressor mutants selected for their restored ability to grow and synthesize proteins all overexpressed Us11 as an immediate-early protein . Further analysis of Us11 mutants revealed that Us11 is also required for maximal rates of translation when expressed in its natural context late in infection (M. Mulvey and I. Mohr, unpublished data). Inhibition of PKR activation by the carboxyl-terminal 68 residues of Us11 is responsible for the observed effects on protein synthesis . This same segment of Us11 binds double-stranded RNA (dsRNA) through a novel RNA-binding motif , interacts with PKR and the PKR activator PACT , associates with 60S ribosomes and polysomes , contains an overlapping nucleolar retention and nuclear export signal that allows both nucleolar and cytoplasmic accumulation , and can potentially bind to a variety of structured cellular and viral RNAs found in infected cells . Although both the amino- and carboxyl-terminal portions of Us11 can be packaged into virions, the amino-terminal 40 amino acids have been reported to possess an activity similar to that of transactivator proteins encoded by complex retroviruses . It is entirely possible that the catalogue enumerated above, although extensive, does not represent a comprehensive picture of Us11's activities. Our understanding of the role(s) played by Us11 molecules contained within the tegument, both extremely early and late in the infectious program, is also far from complete. Indeed, Us11 protein packaged in the tegument and subsequently discharged into the cytoplasm of a newly infected host cell is not sufficient to overcome the block to protein synthesis observed in cells infected with a gamma34.5 mutant virus, even at very large multiplicities of infection . Thus, although ca. 600 to 1,000 molecules of Us11 per virion enter the cytosol, some of which associate with polyribosomes and enter the nucleolus , this population of Us11 molecules might engage in other actions that have yet to be documented. To explore the possibility that Us11 might indeed contain other intrinsic activities, we searched for cellular proteins that associate with the Us11 polypeptide, since the interaction between viral and host proteins is important for many events in the viral life cycle. Here, we identify the cellular PAT1 polypeptide as a Us11 binding-protein by using multiple, independent assays. PAT1 binds microtubules, is involved in the intracellular trafficking of amyloid precursor protein (APP), and contains a region homologous to kinesin light chain (KLC) . This interaction requires the carboxyl-terminal dsRNA-binding domain of Us11 and a region of PAT1 that contains homology to the kinesin light chain. In addition, PAT1 and Us11 colocalize when ectopically expressed in transiently transfected cells and in HSV-1-infected cells. The association between a viral tegument protein and a cellular protein with the potential to interact with a molecular motor may be important for the intracellular movement of viral components. MATERIALS AND METHODS : Cell culture. | 293T and MDCK cells were propogated in Dulbecco modified Eagle medium supplemented with 10% fetal bovine serum (FBS; Gibco), 2 mM l-glutamine, 50 U of penicillin/ml, and 50 mug of streptomycin/ml. Cos-1 cells were grown in identical media except 10% calf serum was used in place of FBS. The inducible MDCK585 and MDCK411 lines were grown in minimal essential medium supplemented with 1x minimal essential medium nonessential amino acids, 10% FBS, 2 mM l-glutamine, 3 mug of blasticidin/ml, 0.3 mg of zeocin/ml, 50 U of penicillin/ml, and 50 mug of streptomycin/ml as described previously . To induce the production of PAT1 and PAT1Delta412, 1 mug of tetracycline (Invitrogen)/ml and 2 mM sodium butyrate were added to the culture media for 18 h. Construction of recombinant HSV-1 expressing a GFP-Us11 fusion protein. | The HSV-1 Patton strain was used throughout the present study. Nucleotide coordinates refer to the published strain 17 sequence (GenBank accession no. ). Derivatives of the plasmid pSXZY contain the Us10 -12 region, along with flanking sequences (nucleotides 143481 to 147040) and served to target the Us11 alleles to the homologous region in the viral genome. To replace Us11 with an open reading frame (ORF) that fuses enhanced green fluorescent protein (EGFP) coding sequences in frame to the first ATG of the Us11 gene, pSXZY was digested with XhoI, the 5' overhangs were filled in with the Klenow fragment, and the plasmid was subsequently cleaved with BspEI. The plasmid pEGFP-Us11 contains the entire EGFP ORF fused to the first ATG of the Us11 gene and produces an EGFP-Us11 fusion protein under the control of the human cytomegalovirus promoter. After cleavage of pEGFP-Us11 with AgeI, the 5' overhangs were filled in with the Klenow fragment, and the plasmid was digested with BspEI, releasing an 898-bp fragment that contains the EGFP-Us11 fusion gene. This fragment was ligated into XhoI (filled)- and BspEI-digested pSXZY to create pSXZY-GFP-Us11. After cleavage with PvuII to release the HSV-1 sequences from the plasmid backbone, 10 mug of digested plasmid DNA was cotransfected, along with purified GFPDeltaUs11 viral DNA, into Vero cells. EGFP-positive plaques were subjected to two rounds of purification. The physical structure of the viral recombinants was confirmed by Southern analysis, and labeling of viral proteins was performed as described previously . Proteins. | GSTDelta1-87, GSTDelta88-155, and GST-RXP have been described . In vitro-translated PAT1 was generated with the TNT quick-coupled reticulolysate system according to the manufacturer's instructions (Promega). Plasmids. | The plasmids expressing PAT1 or PAT1Delta412 in pCDNA4TO have been described . To generate the amino-terminal fragments Delta281, Delta352, and Delta436, stop codons were introduced into the PAT1 coding sequence by PCR. PAT1 cDNA was excised from the pcDNA4TO vector with KpnI and XhoI and inserted into pBluescript SK(+) that had been digested with the same enzymes in order to produce a plasmid capable of synthesizing PAT1 RNA that could be translated in vitro. Yeast two-hybrid screen. | The yeast two-hybrid assay was performed as described according to the MATCHMAKER two-hybrid user manual. The yeast strain EGY188 harboring the reporter genes LacZ and LEU2, under the control of the upstream LexA DNA-binding domain, was used in the assay. For pEG-202 (a bait vector containing the LexA DNA-binding domain) constructs, the entire ORF (155 residues) of HSV-Us11 were subcloned in frame into pEG-202. The bait-containing vector, pEG-202, and the human brain MATCHMAKER LexA cDNA library in the pJG4-5 plasmid were sequentially transformed into the yeast strain EGY188 and grown on minimal synthetic medium lacking tryptophan, histidine, leucine, and uracil to select potential interactors. The X-Gal (5-bromo-4-chloro-3-indolyl-beta-d-galactopyranoside) indicator dye was included to screen transformants for lacZ expression. Approximately 3 x 106 clones were screened against the Us11 bait. Plasmid DNA isolated from potential positive clones was transformed into Escherichia coli KC8 cells. Bacterial colonies selected on M9 minimal medium lacking tryptophan contained the library plasmid, which was subsequently isolated, and its insert was sequenced. To map the interaction domains, deletion mutants of PAT1 were cloned into pJG4-5, and Us11Delta1-87 and Us11Delta88-161 were cloned into pEG-202. The pJG4-5 and pEG-202 constructs were cotransformed into yeast strain by using a small-scale LiAc/ssDNA/PEG transformation protocol. The transformants were plated onto appropriate minimum synthetic dropout media and were tested for beta-galactosidase activity. Preparation of protein extracts. | 293T cells seeded in six-well dishes were transfected with either wild-type (WT) or mutant PAT1 expression plasmids by using Lipofectamine 2000 (Invitrogen) according to the manufacturer's specifications. After 24 to 48 h, the cells were washed twice with 50 mM HEPES-KOH (pH 7.4)-150 mM NaCl-1 mM EDTA-5% glycerol and subsequently lysed in the same buffer supplemented with 0.5%% Triton X-100 and 200 muM phenylmethylsulfonyl fluoride. After 10 min on ice, the extract was clarified by centrifugation at 14,000 x g for 10 min at 4C. The supernatant was used as the source of recombinant PAT1 protein. In vitro binding assays. | A total 100 mul of a 10% glutathione-agarose slurry was equilibrated in binding buffer (50 mM HEPES-KOH [pH 7.4], 150 mM NaCl, 1 mM EDTA, 5% glycerol, 0.5% Triton X-100) and mixed with 1 mug of purified glutathione S-transferase (GST) fusion protein for 1 h on ice with periodic agitation. The beads were subsequently collected by centrifugation and suspended in 1 ml of binding buffer plus 3% BSA (fraction V). After the mixture was rocked for 1 h at room temperature, 3 mul of 35S-labeled PAT1 that was previously treated with micrococcal nuclease or various quantities of an extract prepared from 293T cells was added. Binding reactions were incubated for 30 min at 30C and manually agitated every 5 min. The beads were subsequently collected by centrifugation and washed three times with binding buffer, and the associated proteins were fractionated by electrophoresis. Whereas labeled PAT1 was visualized after exposure of fluorophore-impregnated gels to film, samples containing FLAG-tagged PAT1 derivatives were electrophoretically transferred to a polyvinylidene difluoride (PVDF) membrane and probed with an anti-FLAG monoclonal antibody (Sigma). Antigen-antibody complexes were visualized by using horseradish peroxidase-conjugated secondary antibodies and enhanced chemiluminescence (ECL; Amersham). For far-Western analysis, 5-mug samples of GST, GST-US11 Delta1-87, and GST-Us11 Delta88-155 were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred to a PVDF membrane. Renaturation of the immobilized polypeptides was accomplished by washing the membrane once for 10 min in PBB (50 mM Tris-HCl [pH 8.0], 150 mM NaCl, 10% glycerol, 1 mM dithiothreitol, 100 muM phenylmethylsulfonyl fluoride), followed by an overnight incubation in PBB at 4C. Nonspecific binding sites were next blocked in PBB plus 3% nonfat dry milk for 4 h at 4C. The membrane was probed with in vitro-translated 35S-labeled PAT1 or luciferase for 24 h at 4C. After five washes at room temperature with PBB plus 0.1% NP-40 (10 min for each wash), the membrane was air dried and exposed to either a phosphorimager screen or film. Immunofluorescence. | Cells were seeded on coverslips and processed for immunofluorescence as described previously . Briefly, Cos-1 or MDCK cells were seeded onto glass coverslips for 24 h. After transfection of the appropriate expression plasmids with Lipofectamine 2000 (Invitrogen), the cells were fixed with 3.7% paraformaldehyde, quenched with 100 mM NH4Cl, and permeabilized with 0.1% Triton X-100. Samples were next incubated with anti-FLAG M2 primary antibody (Sigma) for 30 min at 37C, followed by the addition of a Texas red-conjugated secondary antibody (Vector Laboratories). Nuclei were stained with Hoechst stain. Visualization, quantitative analysis, and photography were performed by using a Zeiss Axioplan 2 fluorescence microscope equipped with a filter for detecting rhodamine and fluoroscein. For confocal studies, Cos-1 cells were seeded onto 35-mm dishes (MatTek Corp.) and, when the cells achieved 70% confluence, they were transfected with plasmids expressing WT PAT1 or PAT1Delta412. At 48 h posttransfection, the cells were infected with GFP-Us11R at an MOI of 5. The samples were processed at various times postinfection as described above and observed by using a Zeiss Axiovert 100 M confocal microscope. RESULTS : PAT1 is a Us11 interacting protein. | To identify Us11 interacting proteins, we performed a yeast two-hybrid screen. The entire Us11 ORF was fused to the LexA DNA-binding domain and cotransformed with a LacZ reporter plasmid into the EGY188 yeast strain. The yeast cells were subsequently transformed with a human brain cDNA library fused to a synthetic activation domain. Proteins that associate with Us11 will activate transcription of the lacZ gene via the fusion to the synthetic activation domain. Three million clones were screened, and four clones were isolated that interact with the full-length Us11 protein (M. Mulvey and I. Mohr, unpublished data). One of these clones fused amino acids 41 to 585 of PAT1 to the synthetic activation domain and was selected for further study. PAT1 (named for protein interacting with APP tail-1) was originally identified in a yeast two-hybrid screen as a protein that binds the basolateral sorting sequence contained in the APP . APP is a proteolytically processed transmembrane protein that is involved in the pathogenesis of Alzheimer's disease . PAT1 binds microtubules and is involved in the intracellular trafficking of APP . Recent data indicate that PAT1 may also participate in intracellular signaling mediated by the cytoplasmic tail of APP . The central region of the 585-residue PAT1 protein contains a 35-amino-acid segment predicted to adopt the structure of a coiled coil, and the carboxyl-terminal portion includes a region with homology to kinesin light chain . PAT1 specifically associates with 68 residues within the carboxyl terminus of Us11. | Many of the activities ascribed to the Us11 protein reside within the carboxyl-terminal 68 residues . A repetitive motif composed of the amino acids Arg-X-Pro comprises a novel dsRNA-binding domain and is thought to fold into a type II polyprolyl alpha-helix . This segment of Us11 also associates with ribosomes , interacts with PKR and PACT , inhibits PKR activation , and contains interwoven nucleolar retention and nuclear export signals, allowing nucleolar accumulation, along with subsequent translocation to the cytosol . To discern which segment of Us11 associates with PAT1, the ability of PAT1 to complex with a portion of the Us11 protein fused to GST was evaluated. Full-length PAT1 was translated in vitro and incubated with either purified GST, GST fused to the carboxyl-terminal 68 amino acids of Us11 (GST-Us11Delta1-87), or GST fused to the amino-terminal 87 amino acids of Us11 (GST-Us11Delta88-155). After the isolation of GST-containing proteins on glutathione-agarose beads, associated proteins were fractionated by SDS-PAGE and visualized by fluorography. Figure demonstrates that the full-length PAT1 protein associates with a GST fusion protein that contains the carboxyl-terminal 68 amino acids of Us11 and not with GST-Us11Delta88-155 or GST. Labeled PAT1 also bound GST-Us11Delta1-87 in a far-Western assay. GST, GST-Us11Delta1-87, and GST-Us11Delta88-161 were subjected to SDS-PAGE, transferred to nitrocellulose, renatured, and probed with labeled PAT1. Only the immoblized GST-Us11Delta1-87 protein bound the PAT1 polypeptide . A labeled luciferase probe did not associate with any of the proteins immobilized on the filter (not shown). Together, these in vitro assays provide multiple, independent physical lines of evidence that confirm the PAT1-Us11 association discovered in the yeast two-hybrid screen and further delineate the PAT1 binding site on the Us11 protein. In addition, PAT1 does not interact with GST-RXP , a GST fusion protein containing 36 amino acids from the Epstein-Barr virus SM polypeptide. GST-RXP contains eight iterations of the RXP repeat sequence, binds RNA, associates with PKR, and inhibits PKR activation . Thus, although a different protein containing RXP repeats can execute a number of Us11-related functions, only the RXP segment derived from Us11 associates with PAT1. PAT1 residues 412 to 585 containing the kinesin light-chain homology region are required to interact with the C-terminal domain of Us11. | To determine the region(s) of the PAT1 protein important for binding to Us11, a series of PAT1 deletion mutants was constructed, and the resulting polypeptides were expressed both in yeast cells, as fusions to the synthetic activation domain, and in mammalian cells . PAT1 fusions expressed in yeast were examined for their ability to associate with LexA-Us11 fusion proteins and to activate transcription of a beta-galactosidase reporter gene. Only the full-length, WT PAT1 protein and a 40-amino-acid amino-terminal deletion were able to associate with Us11 or Us11Delta1-87. Amino-terminal PAT1 fragments ranging in size from 281 to 436 amino acids, although expressed at similar levels (not shown), were unable to associate with Us11 and activate beta-galactosidase transcription (Table and Fig. ). Extracts prepared from mammalian cells expressing these FLAG-tagged PAT1 amino-terminal fragments were incubated with equal amounts of purified GST, GST-Us11Delta1-87, or GST-Us11Delta88-155. Glutathione-agarose-bound complexes were collected, fractionated in SDS-polyacrylamide gels, transferred to nitrocellulose, and incubated with an anti-FLAG antibody. Although similar amounts of FLAG-tagged PAT1 were present in the various extracts, only the full-length PAT1 protein was able to associate with GST-Us11Delta1-87 (Fig. to D). Thus, sequences in the carboxyl-terminal 149 amino acids of PAT1 are important for binding to Us11. Us11 expression alters the subcellular distribution of PAT1 in transfected cells. | To discern whether the physical association of PAT1 and Us11 could affect the localization of the two polypeptides in intact mammalian cells, the subcellular distribution of the proteins, individually and in tandem, was evaluated by fluorescence microscopy. A fusion of Us11 protein coding sequences to EGFP was used to visualize Us11, whereas PAT1 tagged with a FLAG epitope was identified by indirect immunofluoresence. In contrast to the pattern displayed by EGFP after transfection into MDCK cells, EGFP-Us11 accumulated most intensely in the nucleolus and diffusely stained the nucleoplasm . EGFP-Us11 was also present in the cytosol, albeit in much smaller amounts . PAT1, however, was predominantly concentrated in the nucleus and exhibited nucleolar sparing. Numerous small, discrete PAT1 foci were observed in the cytosol as well . A mutant derivative of PAT1 that lacks residues 412 to 585 (Delta412) accumulated diffusely in the cytosol and exhibited areas of more intense punctate staining in the perinuclear region . These data, concerning the intracellular localization of each protein expressed individually, agree with previously published reports and serve to validate our assay. Experiments examining the localization of PAT1 and Us11 expressed together in the same cell were conducted in MDCK cell lines stably expressing either Flag-tagged WT PAT1 (MDCK585) or PAT1Delta412 (MDCK411) from an inducible promoter and in Cos-1 cells cotransfected with EGFP-Us11 and PAT1 derivatives. In MDCK585 cells, EGFP-Us11 accumulated predominantly in the perinuclear region of the cytosol, as did the WT PAT1 protein. Indeed, the two proteins colocalized in this area to a large extent in ca. 44% of cells that expressed both proteins (Fig. and C). Interestingly, very little, if any, EGFP-Us11 is seen in the nucleolus, and the majority of the PAT1 protein appears to be excluded from the nucleus, its normal location . In addition to colocalizing in cells expressing both polypeptides, coexpression of PAT1 and Us11 appears to mutually alter the subcellular distribution observed when each protein was expressed individually . Importantly, this analysis was performed by directly observing a number of individual cells, unlike biochemical methodologies that yield results from entire populations. Therefore, the approach described here is subject to variations inherent within the overall population, which might reflect the asynchronous nature of the culture or genetic heterogeneity resulting from its continuous passage. It is well established that all of the cells within a population, especially those of established immortalized cell lines, do not respond identically to a variety of signals that regulate gene expression . Likewise, in a population of cells that coexpress two polypeptides demonstrated to associate by biochemical techniques, it should come as no surprise that these proteins might not colocalize in 100% of the expressing cells. To demonstrate that the observed frequency of colocalization depends upon PAT1, the distribution of Us11 was evaluated in cells expressing a mutant PAT1 derivative that does not interact with Us11. Significantly, the PAT1 mutant derivative expressed in MDCK411 cells did not colocalize with EGFP-Us11 and displayed the punctate cytoplasmic distribution observed previously . PAT1Delta412 and Us11 colocalized in only 2.4 to 3.2% of cells expressing both polypeptides . The nucleolar accumulation of Us11 in MDCK411 cells was restored as well . A similar picture emerged when the proteins were expressed transiently in Cos-1 cells. PAT1 again appeared to be excluded from the nuclear compartment and extensively colocalized with Us11 in the cytoplasmic and perinuclear area . Thus, Us11 expression can alter the subcellular distribution of PAT1 such that it is excluded from the nucleus; furthermore, this redistribution of WT PAT1 requires 149 amino acids in its carboxyl terminus that shares homology to kinesin light chain. Intriguingly, whereas Us11 is barely if at all detectable in the nucleoli of MDCK cells that express WT PAT1, the accumulation of Us11 in nucleoli of Cos-1 cells that express WT PAT1 is readily discernible . Although the exact reason(s) for this different behavior is not immediately obvious, it certainly could reflect species differences (canine versus monkey), cell state differences (nontransformed versus transformed), or the levels of the expressed proteins achieved in the different cells. In any event, whereas WT PAT1 appears to be capable of altering the subcellular distribution of Us11 in both MDCK and Cos-1 cells, it does not always prevent the nucleolar accumulation or retention of Us11 in transiently transfected Cos-1 cells. Us11 and PAT1 colocalize in HSV-1 infected cells. | Although our studies indicate that Us11 can colocalize with PAT1 and alter its subcellular distribution in the absence of any other HSV-1 components, they have not addressed whether this can occur in infected cells. To evaluate the distribution of Us11 and PAT1 in infected cells, we first constructed a recombinant HSV-1 that fused EGFP coding sequences to the 5' end of the Us11 gene (GFP-Us11R, Fig. ). This allowed us to localize Us11 expression in infected cells by monitoring EGFP fluorescence. Beginning with a Us11-null virus (DeltaUs11 virus) in which the Us11 gene was replaced by the GFP ORF (Fig. and ), we subsequently repaired the mutant Us11 allele with a GFP-Us11 fusion gene . Southern analysis of recombinant viral genomes demonstrated that Us11 coding sequences had been replaced with DNA encoding an EGFP-Us11 fusion protein . Aside from producing an EGFP-Us11 fusion protein displaying an apparent molecular mass of 50 kDa , the recombinant GFP-Us11R was otherwise indistinguishable from its WT counterpart in terms of its ability to direct viral protein synthesis in multiple cultured cell lines (Fig. and M. Mulvey, unpublished results). EGFP-Us11 expression was first detected in nucleoli as early as 6 h postinfection and was clearly visible mainly in nucleoli and the surrounding nucleoplasm by 8 h postinfection, a finding consistent with its expression as a true late gene. Smaller quantities of Us11 were also detected in the cytosol . To assess the effects of viral Us11 expression on PAT1 localization, Cos-1 cells were transfected with vectors expressing either WT PAT1 or PAT1Delta412 and subsequently infected with GFP-Us11R 24 h later. At various times postinfection, cells were fixed and processed for indirect immunofluoresence to detect PAT1. Us11 localization was monitored by the intrinsic fluorescence of EGFP-Us11, and confocal images were obtained. In 59% of cells expressing WT PAT and Us11, PAT1 was excluded from the nucleus and colocalized (yellow signal) with the majority of the EGFP-Us11 polypeptide in a perinuclear zone and a cytosolic patch . Only a small fraction of Us11 (green signal) did not colocalize with WT PAT1 in the merged panel . Although some PAT1 did not appear to colocalize with Us11 in the merged image (red signal), a large proportion of PAT1 clearly colocalized with Us11, as indicated by the yellow perinuclear area . Perhaps the association of both Us11 and PAT1 with other polypeptides accounts for the fraction of both proteins that does not colocalize. Indeed, both Us11 and PAT1 have been demonstrated to interact with other cellular proteins . Finally, Us11 did not accumulate in the nucleoli of PAT1-expressing cells . In marked contrast to cells that express WT PAT1, Us11 amassed in the nucleoli of infected cells that expressed PAT1Delta412 and colocalized with the mutant PAT1 protein in only 6.25% of cells coexpressing the two polypeptides . Thus, the subcellular distribution of PAT1 was altered in infected cells in a manner that required its carboxyl-terminal 149 residues. Moreover, Us11 can colocalize with PAT1 in HSV-1 infected cells and failed to accumulate in the nucleoli of infected cells expressing the WT PAT1 polypeptide. To determine whether Us11 was required for the redistribution of PAT1 in HSV-1 infected cells, Cos-1 cells transfected with a plasmid expressing WT PAT1 were infected with either GFP-Us11R or DeltaUs11 recombinant viruses. At 10 h postinfection, the cells were fixed and processed for immunofluoresence to detect PAT1. All of the cells examined were infected as evidenced by their EGFP fluorescence (data not shown). Figure demonstrates that PAT1 remains nuclear in cells infected with DeltaUs11, as evidenced by visible nucleolar sparing and the coincidence of Texas red and DAPI (4',6'-diamidino-2-phenylindole) stains. However, in cells infected with a virus that produces Us11, PAT1 is predominantly excluded from nuclei and concentrates in a perinuclear region. Thus, the redistribution of PAT1 from the nucleus to the cytosol in HSV-1-infected cells depends exclusively upon the Us11 gene product. DISCUSSION : Us11 is a multifunctional dsRNA-binding protein produced late in the HSV-1 productive growth cycle that is required for maximal rates of late viral protein synthesis (; Mulvey and Mohr, unpublished data). The carboxyl-terminal 68 amino acids binds a variety of structured RNAs ; associates with 60S ribosomal subunits and polysomes ; contains overlapping nucleolar retention and nuclear export signals, allowing nucleolar and cytoplasmic accumulation ; binds the cellular PKR kinase , as well as the PACT kinase activator ; and can prevent activation of PKR . In addition, Us11 is a component of the viral tegument, packaged into the viral particle between the envelope and the nucleocapsid . Since genetic analysis of Us11 mutants is hampered by the fact that it is not essential for replication in any cultured cell lines examined to date , we chose to search for host proteins that associate with Us11 as a means of identifying novel functions that Us11 might influence. We now report that the carboxyl-terminal segment of Us11 interacts with the cellular PAT1 polypeptide, a microtubule-binding protein that participates in the trafficking of APP . A region of PAT1 that contains homology to kinesin light chain is required for binding to Us11 and for the perinuclear colocalization of the two proteins in transfected and HSV-1-infected cells. The association between PAT1, a microtubule-binding protein involved in APP trafficking, and Us11, an RNA-binding tegument polypeptide, results in the redistribution of both polypeptides. Furthermore, this association could be important for the intracellular movement of a variety of viral components such as ribonucleoproteins or nucleocapsids with associated tegument proteins in infected cells. While incoming HSV-1 capsids are propelled along microtubules from the cell periphery to the nucleus by dynein and dynactin , a recent study suggests that tegument proteins may be important for antereograde transport of unenveloped capsids in human neurons . In support of this hypothesis, Us11 was observed to redistribute from neuronal cell bodies into dendrites late in the infectious cycle and to physically associate with the ubiquitous kinesin heavy chain (muKHC). However, an interaction between kinesin light chain and Us11 was not detected in that study . Since discrete KLC isoforms appear to be important for associating with different cargo molecules , perhaps Us11 targets the KLC homology region of PAT1 for this purpose. This may reflect a natural function for PAT1 or represent another example of the virus redirecting host components for its own purposes. Additionally, different neurotrophic alphaherpesviruses may have devised alternative approaches to mobilize various virion components or subassemblies. Unlike the product of the Us9 gene which is conserved among alphaherpesviruses and important for the transit of glycoproteins into axons , Us11 is only found in HSV-1, HSV-2, and simian B virus. Whether other members of the alphaherpesvirus family have targeted PAT1 or another KLC-like molecule remains to be explored. Finally, if the PAT1-Us11 complex is indeed capable of intracellular trafficking, it is possible that unenveloped capsids are not the only cargo transported. As an RNA-binding protein, Us11 can interact with a variety of structured RNA molecules and could distribute them to discrete sites in the cytosol. Movement of ribonucleoprotein components through an association with molecular motor molecules has been demonstrated in Drosophila melanogaster, and this in turn leads to localized translation . Since HSV is a neurotrophic virus, it is certainly worth noting that such localized translation of mRNAs has been described in neurons (reviewed in references and ), and dsRNA-binding proteins have been documented to play a role in the delivery of mRNAs to discrete neuronal subcellular destinations . Besides potentially transporting RNAs in infected cells, the localization of Us11 to discrete sites of translation may be important to prevent ribosome-bound PKR from becoming activated. Our studies demonstrate that PAT1 and Us11 can colocalize in a perinuclear region in epithelial cells. In this regard, it is intriguing that overexpression of PAT1 seems to prevent the accumulation of Us11 in the nucleoli of infected cells. After their translation on ribosomes, the association of the two proteins might preclude each from reaching their respective compartments. Alternatively, their localization could be the result of a dynamic process in which overexpressed PAT1 binds to Us11 in the nucleoplasm, interfering with nucleolar retention while promoting nuclear export, resulting in the accumulation of both proteins in the perinuclear space. In support of this latter model, the impairment of host protein synthesis after infection with HSV-1 makes it likely that Us11 might need to first enter the nucleus in order to encounter the PAT1 protein in infected cells. Furthermore, the segment of Us11 that associates with PAT1 contains the overlapping nucleolar retention and nuclear export signals . Perhaps PAT1 targets Us11 to a region where unenveloped capsids acquire tegument components in the cytosol. One proposed pathway for HSV-1 egress suggests enveloped capsids between the inner and outer leaflets of the nuclear membrane fuse with the outer nuclear membrane, releasing unenveloped capsids into the perinuclear region. These capsids transit through the cytosol and are thought to acquire tegument components by budding into a post-Golgi compartment prior to exiting the cell (; reviewed in reference ). Us11, in turn, might similarly foster the incorporation of the cellular PAT1 protein into the virion. Despite its many activities, Us11 does not appear to contribute to viral neurovirulence in any of the existing animal models examined to date and remains nonessential for replication in cultured cells . It is certainly possible that some of the functions carried out by Us11 are important enough that more than one viral gene product might compensate for the loss of Us11. Different arrays of cellular functions expressed in different cell types could contribute to this as well. Indeed, the gamma34.5 and Us11 polypeptides both regulate eIF2alpha phosphorylation in infected cells, albeit through different mechanisms . In a similar vein, it is conceivable that other viral proteins are able to compensate for the absence of Us11 in the tegument. For example, the VP5 capsid protein, along with the VP22 tegument protein, could potentially interact directly with muKHC as well . That other virion components would possess multiple functions would not be surprising. In addition to transactivating immediate-early promoters after its delivery into infected cells, at late times postinfection VP16 binds vhs, preventing indiscriminate mRNA destruction that could lead to the inhibition of protein synthesis . Furthermore, VP16 also plays an important essential role in viral egress distinct from either of these activities . Likewise, through its association with VP16, vhs inhibits the VP16-mediated activation of immediate-early genes late in infection . It is likely that Us11, like other tegument proteins, participates in a spectrum of activities, some of which remain to be characterized. The inability of tegument-derived Us11 to surmount the translational block imposed by host defenses in cells infected with a virus deficient in gamma34.5 functions , despite the fact that immediate-early expression of Us11 can effectively neutralize the equivalent cellular response , is consistent with this hypothesis. Although the analysis of Us11 functions has proved complicated, perhaps insight into additional novel activities can be gleaned through continued investigation of the host components, such as PAT1 and muKHC, with which it associates. FIG. 1. : Summary of structural and functional motifs found in PAT1 and Us11. Summary of structural and functional motifs found in PAT1 and Us11. (A) Schematic representation of sequence motifs present in the WT PAT1 protein and the mutants used in the present study. The full-length PAT1 protein is composed of 585 amino acid residues, binds to microtubules, and is involved in the intracellular trafficking of APP. The protein contains a segment predicted to adopt the structure of a coiled coil (amino acids 243 to 278) and a region that shares homology with kinesin light chain (amino acids 352 to 538). Mutant PAT1 proteins used in the present study and the regions they contain are also illustrated. (B) Illustration of functional domains in the HSV-1 Us11 protein. The Us11 protein encoded by HSV-1 (Patton strain) is 155 amino acids long. The carboxyl-terminal 68-amino-acid domain contains a novel dsRNA-binding motif composed primarily of 21 repeats of the amino acid triplet RXP. The activities associated with the amino and carboxyl-terminal domains and a representation of the GST fusion proteins are shown. FIG. 2. : PAT1 specifically interacts with the carboxyl-terminal 68 amino acids of Us11. PAT1 specifically interacts with the carboxyl-terminal 68 amino acids of Us11. (A) 35S-labeled PAT1 was incubated with purified GST fusion proteins containing the C-terminal domain of Us11 (Us11Delta1-87), the N-terminal domain of Us11 (Us11Delta88-155), or GST. After the complexes were isolated on glutathione-agarose beads, the proteins were separated by SDS-PAGE, and the fixed dried gel was exposed to Kodak film. The major translated product (input) corresponds to full-length PAT1. PAT1 specifically forms a complex with the C-terminal 87 amino acids of Us11. (B) GST fusion proteins were incubated with lysates prepared from 293T cells transfected with plasmids expressing FLAG-tagged, full-length PAT1. GST-containing complexes were collected on glutathione agarose beads, fractionated by electrophoresis in SDS-polyacrylamide gels, and transferred to a PVDF membrane. The membrane was probed with anti-FLAG antibody, followed by a horseradish peroxidase- conjugated secondary antibody, and PAT1 was visualized via enhanced chemiluminescence. (C) GST and the two GST-Us11 fusion proteins (GST-Us11Delta1-87 and GST-Delta88-155) were subjected to electrophoresis in SDS-polyacrylamide gels, blotted onto nitrocellulose membranes, renatured, and then probed with in vitro-translated [35S]methionine-labeled PAT1. The filter was washed and subsequently exposed in a phosphorimager cassette. FIG. 3. : PAT1 residues 412 to 585 containing the kinesin light-chain homology region are required to interact with the C-terminal domain of Us11. PAT1 residues 412 to 585 containing the kinesin light-chain homology region are required to interact with the C-terminal domain of Us11. (A) Full-length PAT1, PAT1 Delta281, Delta352, Delta412, Delta436, and Delta1-40 were expressed in yeast as activation domain fusions. WT Us11, Us11Delta1-87, and Us11Delta88-155 were expressed in yeast fused to the LexA DNA-binding domain. Interaction between Us11 and PAT1 was detected by the ability to form blue colonies on indicator plates. PAT1Delta412, which lacks amino acids 412 to 585 containing the kinesin light-chain repeats, does not interact with the C-terminal domain of Us11 and appears as a representative example. (B) Cell-free lysates were prepared from Cos cells transfected with plasmids expressing either FLAG-tagged WT PAT1 or PAT1 Delta281, Delta352, Delta412, or Delta436 (input). Equal amounts of purified GST, GST-Us11Delta1-87, or GST-US11Delta88-155 were incubated in the extracts. After the GST proteins were collected on glutathione-agarose beads, the complexes were then washed and fractionated by SDS-PAGE. The proteins were transferred to a solid support, reacted with antibodies, and visualized as described in the legend to Fig. . FIG. 4. : Subcellular localization of GFP Us11, PAT1, or PAT1Delta412 after expression of each protein individually in transfected cells. Subcellular localization of GFP Us11, PAT1, or PAT1Delta412 after expression of each protein individually in transfected cells. MDCK cells were fixed and permeabilized 24 to 48 h after transfection with plasmids expressing either GFP, GFP-Us11, FLAG-tagged PAT1, or FLAG-tagged PAT1Delta412. They were subsequently examined for intrinsic GFP fluoresence or processed for immunofluorescence to detect PAT1. FIG. 5. : Redistribution of PAT1 and colocalization of PAT1 and Us11 require 173 C-terminal residues of PAT1 that contain homology to KLC. Redistribution of PAT1 and colocalization of PAT1 and Us11 require 173 C-terminal residues of PAT1 that contain homology to KLC. MDCK585 (A) or MDCK411 (B) cells were transfected with 7 mug of pEGFP-Us11. Cos-1 cells were cotransfected with 7 mug of pEGFP-Us11 and an additional plasmid (7 mug) expressing either FLAG-tagged WT PAT1 (A) or PAT1Delta412 (B) polypeptides. At 24 to 48 h after transfection, the cells were processed for immunofluoresence to detect PAT1. Nuclei were stained with Hoechst stain. The distribution of GFP-Us11 was monitored by its intrinsic fluorescence. (C) Cells expressing both GFP-Us11 and either PAT1 or PAT1Delta412 were counted (open bars), and the numbers of cells in which Us11 and PAT1 colocalized were recorded (solid bars). The percentage of cells that expressed both GFP-Us11 and PAT1 (WT or Delta412) in which colocalization was observed appears on top of the bar graph for each experiment. Whereas on average, 44% of cells expressing both GFP-Us11 and PAT1 demonstrated colocalization of both proteins, only 5% of cells that expressed GFP-Us11 and PAT1Delta412 exhibited colocalization of the polypeptides. FIG. 6. : Structure of DeltaUs11 and GFP-Us11 recombinant viruses. Structure of DeltaUs11 and GFP-Us11 recombinant viruses. (A) The HSV-1 genome is depicted with the unique long (UL), unique short (Us), and repetitive segments (rectangles) delineated. (B) Expanded view of the WT Us-TRs junction region is shown along with pertinent restriction enzyme cleavage sites. The Us10, Us11, and Us12 ORFs appear as boxes. cis-Acting promoter elements are indicated by stars, and each promoter is normally associated with either the Us10, Us11, or Us12 ORF (denoted by the numbers 10, 11, or 12 at the lower right of each star). Arrows located above each ORF represent mRNA transcripts produced from each promoter. All three transcripts are polyadenylated at a common site downstream from the 3' end of the Us10 ORF. The noncontiguous segment of the Us12 transcript denotes an RNA splicing event. (C) An expanded view of the Us-TRs junction in the DeltaUs11 recombinant virus at which the Us11 ORF has been deleted and replaced with the GFP ORF, followed by a heterologous polyadenylation signal. As a consequence of this mutation, the Us10 ORF is disrupted (the breakpoint is denoted by a broken line at the right end of the box), the Us10 promoter embedded in the Us11 ORF is deleted, and an mRNA capable of encoding the Us10 polypeptide is not produced. (D) An expanded view of the Us-TRs junction in the GFP-Us11R recombinant virus that expresses a GFP-Us11 fusion protein. The GFP ORF has been fused to the 5' end of the Us11 ORF such that a GFP-Us11 fusion protein is produced. Note that the Us10 ORF and promoter are intact and a Us10 mRNA is produced. FIG. 7. : Construction and characterization of a recombinant HSV-1 expressing an EGFP Us11 fusion protein. Construction and characterization of a recombinant HSV-1 expressing an EGFP Us11 fusion protein. (A) Southern analysis of the Us-TRs junction fragments. Viral (GFP-Us11R, DeltaUs11, and WT) or plasmid (pDeltaUs11) DNA was digested to completion with EcoNI, fractionated by electrophoresis in a 1% agarose gel, and transferred onto a nylon membrane. After the blot was probed with a 32P-labeled NcoI-PflMI DNA fragment derived from the Us-TRs region shown in Fig. , the washed membrane was exposed to X-ray film. The plasmid pDeltaUs11 was used to construct the recombinant virus DeltaUs11. EcoNI-digested viral DNA from two different isolates of DeltaUs11 appears as a standard. The genetic structure that surrounds the Us-TRs junction in both the plasmid (pDeltaUs11) and the DeltaUs11 viruses is depicted in Fig. . DeltaUs11 was the parental virus from which the GFP-Us11R recombinants were constructed. Two plaque-purified isolates that contain a GFP-Us11 fusion gene as depicted in Fig. are shown. (B) Pattern of late protein synthesis in cells infected with a recombinant virus that expresses a GFP-Us11 fusion protein. Vero cells were infected with WT HSV-1, the DeltaUs11 mutant, or the GFP-Us11R virus. At 16.5 h postinfection, the cultures were labeled with 35S-labeled amino acids for 1 h, the samples were solubilized in SDS-PAGE loading buffer, and the proteins were fractionated in SDS-polyacrylamide gels. The fixed, dried gel was subsequently exposed to X-ray film. Molecular mass standards (in kilodaltons) appear to the left of the panel. (C) The GFP-Us11R recombinant virus expresses a GFP-Us11 fusion protein. Total protein was isolated from cells infected with either WT, DeltaUs11, or GFP-Us11R HSV-1. After fractionation by SDS-PAGE, the polypeptides were electrophoretically transferred to a membrane and incubated with either anti-Us11 or anti-GFP antibodies. Proteins were detected by chemiluminescence after incubation with horseradish peroxidase-conjugated secondary antibodies. The relative migrations of the GFP-Us11 fusion protein, GFP, and Us11 appear to the left of the image. Molecular mass markers (in kilodaltons) are shown on the right. FIG. 8. : Subcellular localization of Us11 in cells infected with a recombinant HSV-1 that expresses an EGFP Us11 fusion protein. Subcellular localization of Us11 in cells infected with a recombinant HSV-1 that expresses an EGFP Us11 fusion protein. Cos-1 cells infected with vEGFP Us11 were fixed at either 6 or 8 h postinfection, and the subcellular distribution of Us11 was monitored by the intrinsic fluorescence of GFP Us11. FIG. 9. : Redistribution and colocalization of PAT1 and Us11 HSV-1-infected cells. Redistribution and colocalization of PAT1 and Us11 HSV-1-infected cells. (A) Cos-1 cells transfected with plasmids expressing either FLAG-tagged WT PAT1 or PAT1Delta412 were infected with a recombinant HSV-1, which encodes an EGFP Us11 fusion protein, at 24 h posttransfection. Infected cells were fixed at 10 h postinfection and processed for immunofluoresence to detect PAT1 proteins. Localization of Us11 was monitored by the intrinsic fluorescence of EGFP-Us11. Images were captured on a confocal microscope. N, infected cell nuclei. (B) Infected cells expressing both EGFP-Us11 and either FLAG-tagged WT PAT1 or PAT1Delta412 were counted (open bars) and the number of cells in which PAT1 and Us11 colocalized were recorded (solid bars). The percentages of infected cells that expressed both EGFP-Us11 and PAT1 (WT or Delta412) in which colocalization was observed appear on top of the bar graph for each experiment. PAT1 and Us11 colocalized in ca. 59% of infected cells that expressed both proteins, whereas PAT1Delta412 and Us11 only colocalized in 6.25% of infected cells that expressed both proteins. FIG. 10. : Redistribution of PAT1 in HSV-1-infected cells requires the Us11 gene product. Redistribution of PAT1 in HSV-1-infected cells requires the Us11 gene product. Cos-1 cells transfected with a plasmid expressing FLAG-tagged WT PAT1 were infected with either GFPUs11R or DeltaUs11 at 24 h posttransfection. Cells were fixed and processed for immunofluoresence to detect PAT1 at 10 h postinfection. Nuclei were stained with DAPI. TABLE 1 : Defining regions of PAT1 and Us11 required for interaction in the yeast two-hybrid system Backmatter: PMID- 12915576 TI - Existence of Transdominant and Potentiating Mutants of UL9, the Herpes Simplex Virus Type 1 Origin-Binding Protein, Suggests that Levels of UL9 Protein May Be Regulated during Infection AB - UL9 is a multifunctional protein required for herpes simplex virus type 1 (HSV-1) replication in vivo. UL9 is a member of the superfamily II helicases and exhibits helicase and origin-binding activities. We have previously shown that mutations in the conserved helicase motifs of UL9 can have either a transdominant or potentiating effect on the plaque-forming ability of infectious DNA from wild-type virus (A. J. Malik and S. K. Weller, J. Virol. 70:7859-7866, 1996). In this paper, the mechanisms of transdominance and potentiation are explored. We show that the motif V mutant protein containing a G to A substitution at residue 354 is unstable when expressed by transfection and is either processed to a 38-kDa N-terminal fragment or degraded completely. The overexpression of the MV mutant protein is able to influence the steady-state protein levels of wild-type UL9 and to override the inhibitory effects of wild-type UL9. Potentiation correlates with the ability of the UL9 variants containing the G354A mutation to be processed or degraded to the 38-kDa form. We propose that the MV mutant protein is able to interact with full-length UL9 and that this interaction results in a decrease in the steady-state levels of UL9, which in turn leads to enhanced viral infection. Furthermore, we demonstrate that inhibition of HSV-1 infection can be obtained by overexpression of full-length UL9, the C-terminal third of the protein containing the origin-binding domain, or the N-terminal two-thirds of UL9 containing the conserved helicase motifs and the putative dimerization domain. Our results suggest that transdominance can be mediated by overexpression, origin-binding activity, and dimerization, whereas potentiation is most likely caused by the ability of the UL9 MV mutant to influence the steady-state levels of wild-type UL9. Taken together, the results presented in this paper suggest that the regulation of steady-state levels of UL9 may play an important role in controlling viral infection. Keywords: Introduction : The UL9 gene is required for herpes simplex virus type 1 (HSV-1) replication in vivo . The UL9 protein is a dimer in solution and exhibits helicase, ATPase, and origin-binding activities . UL9 is believed to play a key role in the initiation of HSV-1 replication by binding the HSV-1 origin of replication via its C-terminal domain and unwinding it in the presence of ATP and ICP8, the HSV-1 single-stranded DNA binding protein. It is likely that UL9 plays an important role in the assembly of the viral replisome through its interactions with other viral replication proteins . UL9 is a member of the superfamily II helicases . The conserved helicase motifs that are characteristic of this superfamily are positioned within the N-terminal domain of the protein . Genetic studies have previously shown that conserved residues within the helicase motifs are essential for HSV-1 replication in vivo; most engineered motif mutants fail to complement the growth of hr94, a UL9 null virus . Furthermore, biochemical analysis showed a correlation between the failure to complement hr94 and the lack of helicase activity , indicating that helicase activity is essential for UL9 function. Interestingly, a truncated form of UL9 originating from a unique transcript within the UL9 open reading frame designated UL8.5 or OBPC has been observed . OBPC encompasses the 480 C-terminal amino acids of UL9. It is able to bind the origin of replication and localizes to the nucleus, but its significance for the biology of the HSV-1 is not well understood. Several lines of evidence indicate that overexpression of UL9 can inhibit HSV-1 infection. We previously showed that cell lines containing a low copy number of the wild-type UL9 gene could efficiently complement hr94. whereas cell lines harboring a high copy number exhibited lower levels of complementation . In addition, cell lines harboring a high copy number of the UL9 gene were found to inhibit wild-type HSV-1 infection . Furthermore, the cotransfection of wild-type infectious DNA with an excess of plasmid encoding wild-type UL9 reduced the number of plaques observed compared to transfection of wild-type infectious DNA alone . The inhibitory effect of wild-type UL9 overexpression is mediated at least in part by the origin-specific DNA binding function of UL9, harbored in the C-terminal domain (UL9 CTD). In a plaque reduction assay, UL9 CTD severely reduces the efficiency of plaque formation and is thus considered transdominant (dominant negative). The OB mutation which disrupts the origin-binding activity of UL9 reverses the inhibitory effect of wild-type UL9 as well as the transdominant effect of UL9 CTD . The inhibitory properties of the overexpressed wild-type UL9 are consistent with a model in which HSV-1 DNA replication occurs in two steps or stages . According to this model, early in infection HSV-1 replication initiates by a UL9-dependent process at one or more origins of replication (stage I). Later in infection, replication proceeds in an origin-independent manner (stage II). We have proposed that, if UL9 remains bound to the origin of replication late in infection, it may inhibit the transition between stage I and stage II . Consistent with this model, studies with temperature-sensitive UL9 mutants indicate that UL9 is essential for the early stages of HSV-1 replication and appears to be dispensable for the later ones . According to this scenario, we speculate that it may be necessary to downregulate the levels of UL9 protein during stage II in order for infection to proceed. The characterization of transdominant (dominant negative) mutants is a widely employed genetic approach to gain insight into protein function . Mutations in UL9 helicase motifs I, Ia, II, and VI were previously shown to be transdominant in a plaque reduction assay . UL9 mutant proteins bearing mutations in helicase motifs I, Ia, II, and VI are able to dimerize and bind the origin of replication but are defective for helicase activity . We proposed that the transdominance exhibited by these mutants is a result of inhibition of HSV-1 replication at stage I, due to lack of helicase activity, and inhibition of the transition between stage I and stage II, due to overexpression of protein capable of binding the origin, as proposed above for wild-type UL9. In contrast to the transdominant mutations in helicase motifs I, Ia, II, and VI, a mutation in motif V (G354A) was found to confer a potentiating phenotype. When cells were cotransfected with this mutant and wild-type infectious DNA, the plaque number increased compared to transfection with wild-type infectious DNA alone . Interestingly, no stable mutant protein could be detected in the transient transfection-Western blot assay . Intrigued by the ability of an apparently unstable protein to potentiate HSV-1 infection, we sought to further understand the mechanism of potentiation. In this paper, we show that the ability of the UL9 G234A (motif V) mutation to potentiate HSV-1 infection inversely correlates with protein stability. The overexpression of UL9-MV is able to reduce the steady-state wild-type UL9 protein levels and override its inhibitory effects. Furthermore, we report that a fragment consisting of the N-terminal 535 amino acids of UL9 is also transdominant. Based on the genetic and biochemical properties of UL9 helicase motif mutants, we propose a model for the mechanism of transdominance and potentiation. MATERIALS AND METHODS : Reagents and materials. | All restriction enzymes were purchased from New England Biolabs. The pFastBac vector was purchased from Gibco-BRL; the pCDNA-1 and pCDNA-3 vectors were from Invitrogen. Supplemented Grace's medium and penicillin-streptomycin were purchased from Gibco-BRL; fetal calf serum was from Gemini Bioproducts Inc. Viruses and cells. | HSV-1 strain KOS was used as the wild-type virus in all assays. hr94, a UL9 lacZ insertion mutant, was described previously . Vero cells (American Type Culture Collection) and 2B-11, a Vero-derived cell line stably transfected with wild-type UL9 , were maintained in Dulbecco's modified Eagle's medium as described previously . Spodoptera frugiperda Sf21 insect cells were maintained in serum-free medium containing 0.1 mg of streptomycin per ml and 100 units of penicillin per ml. Escherichia coli DH5alpha cells were used for plasmid amplification. DH10Bac competent cells (Gibco-BRL) were used for bacmid packaging and propagation. Antibodies. | Several anti-UL9 antibodies were used in this study: R250 (a kind gift from M. Challberg, National Institutes of Health) is a polyclonal antibody that recognizes the C-terminal 10 amino acids of UL9; 17B is a monoclonal antibody that recognizes the N-terminal 35 amino acids of UL9 ; and RH7 (a kind gift from D. Tenney, Bristol Meyers Squib) was raised against a glutathione S-transferase fusion with the UL9 C-terminal domain (residues 535 to 851). The polyclonal anti-UL6 antibody was described previously . Anti-AU1 tag antibody was purchased from Babco (Berkeley, Calif.), and polyclonal anti-ICP8 antibody was a kind gift from W. Ruyechan, State University of New York, Buffalo, N.Y. . Fluorescein isothiocyanate-conjugated anti-mouse or anti-rabbit immunoglobulin and Texas red-conjugated anti-mouse or anti-rabbit immunoglobulin antibodies were used as secondary antibodies for immunofluorescence analysis. Commercially available antitubulin antibody (Sigma) was used in Western blot analysis to monitor sample loading. Anti-His tag India probe (Pierce) was used to detect His-tagged UL9 fusions. Plasmids. | The pFastBac transfer vectors were used to generate recombinant baculoviruses bearing the C-terminal 317 amino acids of UL9 (pFastBac-UL9-CT) and the N-terminal 535 amino acids of UL9 (pFastBac-UL9-NT). The BamHI- and EcoRI-digested PCR product containing the C terminus (UL9-CT) was inserted in the pFastBac plasmid to generate pFastBac-UL9-CT. The PCR product was amplified by stepdown PCR with pFastBac-UL9-WT-expressing plasmid as a template and the following oligonucleotides: BamHI-UL9Delta1-534 (5'-GCCGAGGTGGATCCATGGATCCCGAGGCGTCGCTGCCGGCCCA-3') and 3'PCL-pFastBac (5'-GATTATGATCCTCTAGTACTTCTC-3'). The pFastBac-UL9-NT plasmid was generated by cloning a BamHI- and EcoRI-digested PCR product containing UL9-NT into the pFastBac plasmid. The UL9-NT PCR fragment was amplified by stepdown PCR with the pFastBac-UL9-WT plasmid as a template and the following oligonucleotides: UL9Delta535-851-EcoRI (5'-GCCGAGGTGAATTCTTACTACCGGCACCGCAGCTCCCGTAGATCG-3') and 5'PCL-pFastBac (5'-GATTATTCATACCGTCCCACCATC-3'). The pCDNA3-UL9-WT plasmid, expressing wild-type UL9 under the control of a cytomegalovirus promoter, was generated by inserting a BamHI-EcoRI fragment containing the wild-type UL9 gene from pFastBac-UL9-WT into the pCDNA3 vector (Invitrogen). The pCDNA3-UL9-MI plasmid, carrying the K87A mutation in helicase motif I, was generated by inserting a BamHI-EcoRI fragment from pCDNA1-UL9-MI into the pCDNA3 vector. The pCDNA3-UL9-MV plasmid was generated by replacing the wild-type UL9 sequence in pCDNA3-UL9-WT with an AscI-EcoRI fragment containing the G354A mutation from p6UL9-MV . The pCDNA3-UL9-MI/MV double mutant was generated by subcloning an AscI-EcoRI fragment containing the G354A mutation from p6UL9-MV into pCDNA3-UL9-MI, replacing the corresponding wild-type fragment. pCDNA3-UL9 NTD and pCDNA-UL9-CTD were generated by inserting the BamHI-EcoRI fragment from pFastBac-UL9 NTD and pFastBac-UL9-CTD, respectively, into the pCDNA3 vector. The pCDNA3-UL9Delta322-851, pCDNA3-UL9Delta355-851-MV, and pCDNA3-UL9Delta450-851 plasmids were generated by inserting the AscI- and EcoRI-digested PCR fragments containing the corresponding truncations into pCDNA3-UL9-WT, replacing the wild-type fragment. All PCR fragments were generated by stepdown PCR with the pFastBac-UL9-WT plasmid as the template and OligoI (5'-GCGGAGTCGGGAGATCCTCTGGGG-3') as the upstream primer. The pCDNA3-UL9Delta450-851-MV truncation was generated with the same procedure except that the p6UL9-MV plasmid was used as a template in the PCR. The specific downstream oligonucleotides used to generate the above truncations were UL9Delta322-851 (5'-GCCGAGGTGAATTCTCATTAGTCCGTAAACTGACGGCAGAACCGGGCCAC-3'), UL9Delta355-851-MV (5'-GCCGAGGTGAATTCTCATTAGGCCACCGTTACGACCGTCGTGTATATAACCACGCG-3'), and UL9Delta450-851 (5'-GCCGAGGTGAATTCTCATTAGCATGCCGACGCGTCACAGCGCCCCTTGAACC-3'). The pUL9Eco plasmid, cont-aining UL9 under the control of its natural promoter, was generously provided by P. Schaffer (University of Pennsylvania School of Medicine). J. Nellissery kindly provided the pCDNA3-UL6 plasmid. The p6UL9-119b plasmid, containing the UL9 gene under the control of the ICP6 promoter, was described previously . The pCDNA3-UL9-NT-AU1 plasmid was generated by subcloning an AscI- and NotI-digested PCR product into pCDNA3-UL9-WT, replacing the corresponding wild-type fragment. The PCR product was generated by stepdown PCR with the pFastBac-UL9-WT plasmid as a template, OligoI (5'-GCGGAGTCGGGAGATCCTCTGGGG-3') as the upstream primer, and UL9-NT-AU1 (5'-GCCGGCTTCTGCGGCCGCTCACTATATATAGCGATACGTATCATCCCGGCACCGCAGCTCCCGTATG-3') as the downstream primer. The nucleotide sequence encoding the AU1 tag is underlined. The cloning procedure resulted in the addition of the AU1 tag (DTYRYI) at the C terminus of UL9-NT. The pCDNA3-AU1-UL9 plasmid was created by inserting a BamHI-EcoRI fragment containing the wild-type UL9 gene into the pCDNA3-AU1 vector, resulting in the addition of an AU1 tag at the N terminus of UL9. The pCDNA3-AU1 vector was generated by ligation of an AU1-containing linker in the polycloning site via the HindIII and BamHI sites. The AU1-containing linker was created by annealing the BM-HB-AU1-T (5'-AGCTTACCATGGATACGTATCGCTACATAG-3') and BM-HB-AU1-B (5'-GATCCTATGTAGCGATACGTATCCATGGTA-3') oligonucleotides. The nucleotide sequence encoding the AU1 tag is underlined. Generation of recombinant baculoviruses expressing UL9 variants. | Recombinant baculoviruses (Autographa californica nuclear polyhedrosis baculoviruses) expressing wild-type or mutant UL9 variants were generated with the pFastBac (Gibco-BRL) system according to the manufacturer's instructions. Cell lysates. | Two types of mammalian cell lysates were used in this study. Regular cell lysates were obtained from one 60-mm2 plate which was washed with phosphate-buffered saline containing a Sigma protease inhibitor cocktail, harvested, and resuspended in sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer, resulting in a total of 300 mul of lysate. The lysates were boiled and analyzed by Western blot analysis. Concentrated cell lysates were prepared from one 60-mm2 plate, which was harvested and resuspended in SDS-PAGE loading buffer, resulting in a total of 60 mul of five-fold-concentrated lysate. The samples were cup-sonicated twice (Sonicator Ultra Processor XL; Misonix, Inc.; power setting 9 for 1 min), boiled, and analyzed. Transient transfection-complementation assay. | The transient transfection-complementation assay was performed as described previously . Each experiment was repeated three times, and the mean complementation index was calculated. Plaque reduction assay. | The plaque reduction assay was performed as described previously . In brief, Vero cells in 60-mm2 plates at 50% confluence were cotransfected with wild-type HSV-1 infectious DNA (strain KOS) and the plasmid of interest (at a ratio of 1:10) with Lipofectamine Plus (Gibco), according to the manufacturer's recommendations. The plates were overlaid with methylcellulose and incubated until plaques were observed. Plaques were stained with crystal violet and counted. Each experiment was performed three times, and the mean number of plaques and the standard deviation were calculated. Immunofluorescent microscopy. | Vero cells were transfected with 2 mug of the plasmid of interest with Lipofectamine Plus (Gibco) and processed for immunofluorescence 18 h posttransfection. The transfected cells were fixed, permeabilized, and stained as described previously except that all procedures were performed at room temperature and phosphate-buffered saline was supplemented with 0.1 mM CaCl2 and 1 mM MgCl2 to reduce cell detachment. The primary antibodies were used at the following dilutions: 17B, undiluted; RH7, 1:100 dilution; anti-ICP8, 1:400 dilution; anti-AU1, 1:500 dilution. The secondary antibodies were used at a dilution of 1:200. Images were taken with an Olympus BX60 microscope (at 100x magnification) equipped with a Hamamatsu Orca digital camera (model C4742-95) and Open Lab software (Improvisions). Images were processed and arranged with Adobe Photoshop 6.0 and Adobe Illustrator 7.0 software. RESULTS : The UL9 helicase motif mutants were classified as transdominant or potentiating based on a plaque reduction assay in which cells were cotransfected with wild-type infectious DNA and an excess of UL9-encoding plasmid. The number of plaques observed in cells transfected with infectious DNA only was normalized to 100 . Overexpression of wild-type UL9 slightly reduced the number of plaques in this assay (60 compared to 100); therefore, wild-type UL9 is considered inhibitory. Mutants whose overexpression resulted in significant reduction of the plaque number were classified as transdominant. For instance, in cells transfected with motif I, Ia, II, and VI mutants, the number of plaques dropped to 10 to 15. In contrast, mutants whose overexpression led to increased plaque number were classified as potentiating. In cells transfected with the motif V mutant, the plaque number increased to 180 . It was shown previously that the motif V mutant protein was undetectable in mammalian cells transiently transfected with a mammalian expression vector under the control of the ICP6 promoter . No protein could be detected in immunoblots of cell lysates from UL9-MV-transfected cells with any of the available anti-UL9 antibodies (17B, raised against the extreme N terminus; R250, recognizing the extreme C terminus; or RH7, raised against the whole C-terminal domain) . In addition, a mutant carrying the motif V mutation and the OB mutation (a 4-amino-acid insertion after residue 591 in the C-terminal domain of UL9 that abolishes the origin-specific DNA function) was even more potentiating than the motif V mutant alone; the number of plaques in this case was increased to 220 . These results indicated that the motif V mutant was capable of influencing the outcome of a viral infection and thus must be expressed, at least transiently. N-terminal fragment(s) but not full-length UL9-MV protein can be detected in heterologous expression systems. | In order to determine if the motif V mutant protein could be detected in heterologous expression systems, UL9-MV was expressed in insect cells with recombinant baculoviruses. Cell lysates of insect cells infected with the recombinant virus bearing wild-type UL9 or the UL9-MV mutant were subjected to Western blot analysis with antibodies which recognize the extreme N terminus (17B) or the extreme C terminus (R250) of the UL9 protein. Both antibodies detected full-length wild-type UL9 protein (Fig. , lanes 2 and 5, respectively). Consistent with our previous findings in mammalian cells , no full-length UL9-MV protein was observed with 17B or R250 antibodies (Fig. , lanes 3 and 6, respectively). Surprisingly, the 17B antibody was able to detect two fragments of approximately 38 kDa (marked with an arrowhead) and 30 kDa (marked with two arrowheads). No fragments were detected with the anti-C terminus antibody (R250) (Fig. , lane 6). The relative amount of the 30-kDa fragment varied between different sample preparations, suggesting that it may be a degradation product of the 38-kDa fragment. We next asked whether the appearance of the N-terminal fragments in insect cells but not in mammalian cells reflects different expression levels and/or stability in these two types of cells. In order to determine if our previous failure to detect UL9-MV was due to expression levels, we cloned the UL9-MV mutant in a mammalian expression vector (pCDNA3; Invitrogen) under the control of the cytomegalovirus immediate-early promoter, a strong constitutive promoter. Vero cells were transfected with the pCDNA3-UL9-MV plasmid, and the cell lysates were assayed for the presence of UL9-MV protein or UL9-MV fragments. As previously seen with vectors with the ICP6 promoter , no protein species were detected with any of the anti-UL9 antibodies (data not shown), but when concentrated lysates of cells transfected with pCDNA3-UL9-MV were prepared as described in Materials and Methods, the 17B antibody detected an N-terminal fragment with an apparent mobility of approximately 38 kDa (Fig. , lane 4). No specific bands were detected with the R250 antibody (data not shown). Thus, we have demonstrated that in both insect and mammalian expression systems, no full-length UL9-MV protein is detectable; instead, an N-terminal fragment with similar if not identical mobility was observed in both cell types. Interestingly, a fragment with the same mobility was observed when UL9-MV was cloned with an N-terminal His tag and expressed in E. coli (Fig. , lane 6). The apparent molecular mass of the N-terminal UL9-MV fragment was determined to be approximately 38 kDa, as judged by Coomassie-stained SDS-PAGE gels with Benchmark (Invitrogen) or broad range (New England Biolab) molecular weight protein standards. Thus, it appears that a similar 38-kDa band is generated by expression of the motif V mutant in all three expression systems. We propose that the G354A mutation results in a conformational change(s) in the mutant protein leading to the exposure of a normally buried protease cleavage site(s). The difficulties that we encountered in detecting the 38-kDa fragment at first may indicate that some of the mutant MV protein is degraded completely. It is not clear whether the observed 38-kDa fragment is a result of cleavage by a protease present in all three cell types or whether this fragment represents the minimal stably folded fragment and is not the result of the initial cleavage event. Potentiation is not due to absence of the transdominant origin-binding domain of UL9. | The UL9 C-terminal domain (UL9 CTD), encompassing residues 536 to 851, is transdominant in the context of the plaque reduction assay and in coinfections of wild-type HSV-1 and recombinant HSV-1 expressing the UL9 CTD . The transdominant effect of the UL9 CTD is abolished if a mutation disrupting the origin-binding function is introduced . Since the C terminus of UL9 is clearly transdominant and since no stable C-terminal fragments were detected when the potentiating mutant UL9-MV was expressed, we wanted to rule out the formal possibility that the absence of the C-terminal domain is responsible for potentiation. We hypothesized that the potentiating phenotype may simply be a result of the absence of the transdominant C-terminal portion of UL9. If this hypothesis is correct, it would be expected that the N-terminal portion of UL9 (UL9 NTD), lacking the transdominant origin-binding domain, would be potentiating. In order to test this hypothesis, a construct expressing the UL9 NTD was generated and tested in the plaque reduction assay. Figure shows that the UL9 NTD was transdominant rather than potentiating. Interestingly, the UL9 NTD protein was not detectable by Western blot analysis with the 17B antibody when used to transfect mammalian cells, even if concentrated cell lysates were prepared , suggesting that the UL9 NTD is unstable. This is consistent with previous observations from the baculovirus expression system . It should be noted that when much larger amounts of plasmid DNA were used for transfection, the UL9 NTD could be detected but only at very low levels (see below). In summary, the transdominant phenotype of UL9 NTD rules out the possibility that potentiation is a result of the absence of the UL9 CTD. Level of transdominance of UL9 NTD correlates with its stability and/or steady-state protein levels in the cell. | Since the UL9 NTD is unstable by itself while full-length UL9 is stable, we reasoned that the addition of a tag at the C terminus of the UL9 NTD might improve its stability. An AU1 tag was added at the C terminus of the UL9 NTD after residue 534, resulting in UL9 NTD-AU1. When mammalian cells were transfected with this construct, the tagged version of the N terminus (UL9 NTD-AU1) was readily detected in Western blots with the 17B antibody as well as with the anti-AU1 antibody (data not shown). When UL9 NTD-AU1 was tested in the plaque reduction assay, it was found to be even more transdominant than the UL9 NTD . Transfection experiments with increasing amounts of plasmid DNA demonstrated significant differences in the steady-state levels of UL9 NTD and UL9 NTD-AU1. Whereas UL9 NTD-AU1 was readily observed when 2 mug of plasmid DNA was used for transfection, UL9 NTD was hardly detectable when 7.5 mug of plasmid DNA was used . Although it is possible that the lower steady-state protein levels of UL9 NTD are due to lower expression levels, we consider it unlikely because both pUL9-NTD-AU1 and pUL9-UL9-NTD were expressed from the same promoter. We therefore favor a scenario in which the low steady-state levels of the UL9 NTD are due to inherent instability of the N-terminal fragment of UL9. Thus, the AU1 tag appears to increase the stability of the N-terminal domain of UL9. In summary, we have demonstrated that the potentiation phenotype of UL9-MV is not due to the absence of the C-terminal origin-binding domain of UL9. Furthermore, the UL9 NTD by itself is transdominant, and its level of transdominance correlates with stability and/or steady-state protein levels in the cell. The potentiation phenotype of UL9-MV does not depend on the helicase activity of UL9. | In order to test whether the helicase activity of UL9 is essential for the potentiating phenotype of UL9-MV, a double mutant, UL9-MI/MV, carrying the K87A mutation in helicase motif I (Walker box A) was engineered. The K87A mutation abolishes the ATPase and helicase activities of UL9 . The UL9-MI mutant was previously shown to be transdominant in the context of the plaque reduction assay . The double mutant UL9-MI/MV was found to be potentiating, resulting in approximately the same level of stimulation of viral plaque formation as UL9-MV by itself (data not shown). Thus, the addition of the MI mutation does not affect the potentiation phenotype of UL9-MV, indicating that potentiation does not rely on the helicase activity of UL9. Presence of the MV (G354A) mutation by itself is not sufficient for potentiation. | As described above, UL9-MV potentiation is not simply a result of the absence of the transdominant CTD of UL9 and is independent of the helicase activity of the protein. Therefore, the potentiation is most likely a property of the G354A mutation itself. In order to investigate this hypothesis, several UL9 truncations were engineered : UL9-Delta322-851, which ends 22 amino acids upstream of the G354A mutation; UL9-Delta355-851-MV, which ends immediately after the MV mutation G354A and corresponds to a protein fragment with predicted molecular weight close to the experimentally determined size of the UL9-MV fragment observed in mammalian, insect, and bacterial cells (Fig. and Fig. ); UL9-Delta450-851, which ends shortly after the last helicase motif of UL9 and contains wild-type sequence; and UL9-Delta450-851-MV, which ends shortly after the last helicase motif and contains the MV mutation (G354A). All the mutants except UL9-Delta450-851-MV were found to express proteins of the predicted sizes . Interestingly, the UL9-Delta450-851-MV truncation produced a 38-kDa fragment comparable to the fragment seen with full-length UL9-MV . In the plaque reduction assay, all truncations except UL9-Delta450-851-MV were found to be transdominant . UL9-Delta450-851-MV was potentiating, and the level of potentiation was comparable to that seen with UL9-MV itself . From these experiments, it appears that there is a trend correlating the potentiation phenotype in the plaque reduction assay with the ability of UL9 variants to be processed or degraded to the 38-kDa N-terminal fragment. Our results show that the presence of the MV mutation by itself is not sufficient to confer potentiation, as seen from the transdominant phenotype of the UL9-Delta355-851-MV truncation. If the G354A mutation was sufficient for potentiation, then this truncation would be expected to be potentiating and not transdominant. It appears that, in addition to the G354A mutation itself, a region C-terminal to the G354A mutation (residues 354 to 450) is required for potentiation. Interestingly, neither the G354A mutation nor the region encompassing amino acids 355 to 450 was sufficient to confer potentiation when present alone. The G354A mutation may introduce a conformational change which leads to destabilization of the protein. The 355 to 450 region may be involved in protein-protein interactions within the UL9 dimer. Overexpression of UL9-MV influences wild-type UL9 protein levels and is able to override the inhibitory effect of UL9. | Since the overexpression of wild-type UL9 is inhibitory for HSV-1 infection, the potentiation phenotype of UL9-MV may be a result of removal of UL9 during the course of infection, thereby preventing wild-type UL9 from exercising its inhibitory effect. In order to test this hypothesis, wild-type UL9 and UL9-MV were cotransfected in a ratio of 1:15, and the steady-state levels of full-length UL9 were followed by Western blot. Figure shows that the overexpression of UL9-MV resulted in lower steady-state levels of wild-type UL9 (compare lanes 2 and 3). In order to determine whether destabilization would also occur in the context of infection, Vero cells were transfected with wild-type UL9 and UL9-MV (ratio, 1:15) and 24 h later were superinfected with the UL9 null virus hr94. In both experiments, shown in Fig. , lanes 1 to 4 (transfection), and Fig. , lanes 5 to 8 (transfection and hr94 superinfection), the level of full-length UL9 was lowered approximately two- to threefold, as estimated from control Western blots in which various amounts of UL9 protein were analyzed for comparison . The possibility that the decrease in levels of full-length UL9 protein was due to transfection of large amounts of plasmid (a total of 16 mug per 60-mm2 plate) was ruled out by transfecting the cells with the same amount of empty plasmid in combination with wild-type UL9 (Fig. , lanes 3 and 7). In order to determine whether the overexpression of UL9-MV resulted in the decrease in levels of an unrelated protein, the same experiment was performed but the levels UL6, an HSV-1 protein essential for cleavage and packaging, were monitored (Fig. , lanes 9 to 12). The overexpression of UL9-MV did not lower UL6 steady-state protein levels. In summary, we found that overexpression of UL9-MV reproducibly lowered the levels of wild-type UL9 in the context of both transfection and infection. To ask whether the decrease in steady-state levels of UL9 also correlated with the potentiation phenotype, the same experiment was performed in the context of the plaque reduction assay . In this case, Vero cells were cotransfected with wild-type HSV-1 infectious DNA and plasmids encoding wild-type UL9 and UL9-MV in a ratio of 1:10:150. As shown in Fig. , wild-type UL9 had an inhibitory effect on plaque formation, but the overexpression of UL9-MV overrode the inhibition. Interestingly, cotransfection of wild-type UL9 and UL9-MV with wild-type HSV-1 infectious DNA resulted in even greater plaque numbers compared to the same experiment in which wild-type UL9 was omitted (Fig. , compare the third and second bars). This further stimulation in the presence of wild-type UL9 may be due to differences in the timing of expression of UL9 from plasmids compared to expression from the viral genome. It may reflect the earlier activation of a degradation pathway which acts to lower the steady-state levels of UL9 (see below). In any case, these experiments indicate that the potentiation phenotype of the UL9-MV mutant is inversely related to the protein levels of wild-type UL9. Indirect evidence that UL9 NTD and UL9-MV are able to interact with full-length UL9. | Our finding that UL9-MV lowers the steady-state protein levels of wild-type UL9 but not the steady-state levels of UL6 implies that UL9-MV may be capable of protein-protein interactions with wild-type UL9. In addition, it is possible that the UL9 NTD may exert its transdominant phenotype via heterodimerization with wild-type UL9. If this is correct, it would be expected that the UL9 NTD fragment would be able to form dimers. In order to test these hypotheses, UL9 NTD and UL9-MV proteins were expressed in insect cells with recombinant baculoviruses, partially purified, and subjected to gel filtration to evaluate their oligomerization state. This biochemical approach was unsuccessful due to problems with solubility and protein aggregation. We therefore took advantage of the knowledge of the position of the nuclear localization signal (NLS) of UL9 to demonstrate protein-protein interactions. The UL9 NLS is localized within the last 107 residues of the protein . Thus, the UL9 NTD would be expected to be cytoplasmic when transfected alone. If the UL9 NTD is able to interact with full-length UL9, it would be expected to enter the nucleus when cotransfected with wild-type UL9. Our experiments show that when transfected alone, UL9 NTD-AU1 was cytoplasmic (Fig. , panels a, b, and c), but it was predominantly nuclear when cotransfected with wild-type UL9 (Fig. , panels g, h, and i). Furthermore, the UL9 NTD and wild-type UL9 staining colocalized in the nucleus (Fig. , panel i). In summary, since the UL9 NTD is able to enter the nucleus only when cotransfected with wild-type UL9 and not when transfected alone, we conclude that the UL9 NTD is able to heterodimerize with wild-type UL9 and utilize its NLS. A similar experiment was performed for UL9-MV . When cells were transfected with UL9-MV alone, nuclear and cytoplasmic staining was observed with the 17B antibody, which recognizes the N terminus of UL9 (Fig. , panel a). No staining was observed with the RH7 antibody, raised against the C-terminal domain of the protein (Fig. , panel b). When cells were cotransfected with UL9-MV and wild-type UL9, the majority of UL9-MV was seen in the nucleus, colocalizing with wild-type UL9 (Fig. , panels g, h, and i). In this experiment, we were able to distinguish between the UL9-MV protein and wild-type UL9 by taking advantage of the observation that the addition of the AU1 epitope to the N terminus of wild-type UL9 masked the N-terminal epitope recognized by monoclonal antibody 17B (Fig. , panel d). Thus, using a tagged version of wild-type UL9 allowed us to differentiate between the N terminus of UL9-MV, detected by the 17B antibody, and the N terminus of the AU1-tagged wild-type UL9, detected by the anti-AU1 antibody but not by 17B. Control double staining with anti-AU1 and RH7 antibodies showed that AU1-tagged wild-type UL9 was nuclear (Fig. , panels j and k) and that the staining patterns overlapped (Fig. , panel l). UL9-MV was both nuclear and cytoplasmic when transfected alone and predominantly nuclear when cotransfected with wild-type UL9. The fact that the localization of UL9-MV to the nucleus is improved in the presence of full-length wild-type UL9 is consistent with an interaction between them. The partial nuclear localization of the UL9-MV protein in the absence of wild-type UL9 may be due to the transient expression of full-length mutant protein which can enter the nucleus prior to being processed to the 38-kDa form. Alternatively, the UL9-MV N-terminal fragment may enter the nucleus by diffusion, but we consider this explanation unlikely due to its size. Inhibitory effect of wild-type UL9 depends on level of UL9 protein. | Previous experiments have suggested that the inhibitory effect of wild-type UL9 is dose dependent , but a direct correlation between the actual protein levels of UL9 and the level of inhibition has not been demonstrated. The transfection of increasing amounts wild-type UL9 plasmid in the plaque reduction assay results in a proportional inhibition of HSV-1 infection , but in these experiments the UL9 protein levels have not been examined. In order to show a direct connection between the UL9 protein level and its inhibitory effect, the UL9 gene was placed under the control of three promoters with different strengths, and the ability of each construct to support the growth of hr94, a lacZ insertion UL9 mutant, was investigated in the transient transfection-complementation assay. The UL9 natural promoter is a weak promoter, and the ICP6 promoter is a relatively strong promoter which requires activation by ICP0 or VP16 in order to be transcribed. On the other hand, the cytomegalovirus promoter is a strong constitutive promoter. Figure shows that there was an inverse correlation between the steady-state level of UL9 protein in the cell and the complementation index for wild-type UL9. The inhibitory effects of the wild-type UL9 protein are thus more pronounced when larger amounts of protein are present. DISCUSSION : UL9 is believed to play a key role in the initiation of viral DNA replication and the assembly of the viral replication machinery at the origin of replication, but many questions about its precise mode of action have not been answered. In this paper, we demonstrate that (i) inhibition of HSV-1 infection can be obtained by overexpression of full-length UL9, the C-terminal third of the protein containing the origin-binding domain, and the N-terminal two-thirds of UL9 containing the putative dimerization domain and the helicase motifs; (ii) the level of inhibition of HSV-1 infection by overexpression of wild-type UL9 correlates directly with the steady-state protein levels of the UL9 protein itself; and (iii) the MV mutant protein is unstable when expressed by transfection and is apparently processed to a 38-kDa N-terminal fragment or degraded completely. Potentiation correlates with the ability of the UL9 variants containing the G354A mutation to be processed or degraded to the 38-kDa form. Finally, overexpression of the MV mutant protein was able to influence the steady-state protein level of wild-type UL9 and to override the inhibitory effects of wild-type UL9. Based on the data presented here and in previous studies, we propose that the MV mutant protein is able to interact with full-length UL9 and that this interaction results in a decrease in the steady-state level of UL9, which in turn leads to enhanced viral infection. Model for transdominance and potentiation of UL9 helicase motif mutants. | The behavior of the UL9 helicase motif mutants in the plaque reduction assay and their biochemical properties can be explained, at least in part, by a two-stage model for the progression of HSV DNA replication previously proposed by us and others . The two-stage model is depicted schematically in Fig. . The model posits that UL9 binds the origin of replication via its C-terminal domain (depicted with black circles). We propose that the binding of UL9 at an origin of replication results in origin unwinding and recruitment of the rest of the replication machinery. There is no direct evidence to indicate whether this initial UL9-dependent step occurs via a theta mechanism or by some other mechanism. Later in infection (stage II), we propose that replication proceeds in an UL9-independent manner to form the head-to-tail concatemers which are the observed products of viral DNA replication. Consistent with this model, studies with temperature-sensitive UL9 mutants indicate that UL9 is essential for the early stages of HSV-1 replication and appears to be dispensable for the later ones . No direct evidence exists to answer the question of whether the later stages of replication occur via a recombination-driven and/or rolling-circle mechanism in infected cells. In insect cells infected with recombinant baculoviruses expressing the HSV replication proteins, origin-containing plasmids appeared to replicate by a rolling-circle mechanism, and this mode of replication was inhibited by the presence of UL9 . On the other hand, replicating DNA isolated from mammalian cells has a very complex structure, consistent with a recombination-mediated mechanism (reviewed in references and ). Whether stage II proceeds by rolling-circle or recombination-driven replication or both in HSV-infected cells, we propose that during the transition between stage I and stage II, UL9 is removed from the origin of replication so that it will not pose a physical barrier for the progressing replication fork. If an excess of UL9 is present, we predict that the origin clearance will be delayed or inhibited, resulting in lower overall viral titers (plaque numbers). Figure depicts a model in which overexpression of wild-type UL9 inhibits HSV-1 replication by inhibiting the progression from stage I to stage II. The inhibition appears to be mediated, at least in part, by the origin-binding function of UL9, since mutations which disrupt origin binding relieve the inhibitory effect of wild-type UL9 overexpression . Furthermore, the UL9 CTD is transdominant in the context of the plaque reduction assay . It is likely that the transdominance of the UL9 C-terminal domain is due to inhibitory effects at both stages of replication. The UL9 CTD is known to bind the origin of replication as efficiently as full-length UL9; therefore, this domain would be expected to compete for origin binding with wild-type UL9 originating from the HSV-1 infectious DNA. Expression levels of UL9 CTD exceeded that of the full-length wild-type UL9 in this experiment; therefore, UL9 CTD binding to the origin would be expected to result in a nonproductive complex at the origin due to the absence of the helicase domain, which is responsible for origin unwinding. Thus, HSV-1 replication will be inhibited at stage I. Nevertheless, some of the HSV-1 genomes will be bound by wild-type UL9 originating from the cotransfected infectious DNA, and these will result in some stage I intermediates. Formation of stage II intermediates, however, will also be inhibited because of the inhibitory effects of UL9 remaining bound at the origin . The OBPC protein described by Baradaran et al. would also be expected to be inhibitory for HSV-1 infection because it contains the UL9 origin-binding domain . In this paper, we report that the N terminus of UL9 can also exert a transdominant effect on HSV-1 infection . It is likely that the UL9 NTD inhibits HSV-1 replication at stage I by virtue of its ability to heterodimerize with full-length wild-type UL9. This is consistent with previous reports mapping the dimerization domain of UL9 to the N terminus and the results presented in Fig. . Although the N terminus would be expected to retain helicase activity , the transdominant phenotype of the UL9 NTD indicates that the heterodimer of wild-type UL9 and UL9 NTD is probably not capable of unwinding the origin of replication. It is possible that the requirement for UL9 to bind the origin as a dimer is related to the ability of UL9 to bind to the origin of replication in a cooperative manner . In any case, it appears that the transdominance of the N-terminal domain of UL9 occurs by a different mechanism from that used by the C-terminal domain. The implications for mapping the dimerization domain of UL9 will be discussed below. The transdominant helicase motif I (K87A), Ia (S110T, R112A, and R113A/F115A), II (E175A), and VI (R387K) mutants were found to be able to dimerize and bind oriS as efficiently as wild-type UL9 but did not exhibit helicase activity . Thus, as a result of the overexpression of mutant protein in the context of the plaque reduction assay, the majority of the UL9 present in the cell would be expected to consist of dimers of mutant proteins. Since the mutant homodimers bind the origin of replication as well as the wild type does, the majority of the origins of replication will be bound by helicase-negative mutant homodimers. Therefore, stage I is likely to be inhibited due to the absence of helicase activity . In addition, the progression from stage I to stage II replication would be inhibited by the mutant UL9 molecules remaining bound at the origin of replication by a mechanism similar to the inhibition by overexpression of wild-type UL9 described above . Although we consider it unlikely, we cannot rule out the possibility that the inhibitory effect of wild-type UL9 and the transdominant effect of the helicase motif mutants and the UL9 NTD also involves the ability of these forms of UL9 to interact with other viral and/or cellular factors needed for HSV-1 replication. Further experiments will be necessary to determine whether these potential interactions contribute to the transdominance and inhibition. The potentiating phenotype of the UL9-MV mutant is particularly intriguing. Potentiation is not simply a result of removing the origin-binding domain of UL9 and is independent of the helicase activity of UL9. In three different expression systems, a 38-kDa N-terminal fragment was observed, indicating that full-length mutant protein is unstable and processed to a smaller form; it is also likely that a portion of UL9 is degraded completely, given the difficulties we have experienced even detecting the 38-kDa fragment. It is possible that two degradation pathways exist, resulting in either complete degradation or the formation of the 38-kDa fragment, or that the 38-kDa fragment is a "semistable" intermediate of the complete degradation pathway. Although further investigation will be needed to distinguish between these possibilities, it is noteworthy that the levels of a cellular precursor of NF-kappaB appear to be controlled by two different proteolytic pathways: limited processing to a 50-kDa protein and complete degradation . Despite the fact that no full-length UL9-MV protein was observed in any expression system tested, it is possible that full-length UL9-MV may be transiently expressed and be capable of protein-protein interactions with wild-type UL9. This hypothesis is supported by the observation that the double mutant UL9-MV-OB, containing both the G354A and an origin-binding mutation, was more potentiating than the UL9-MV mutant alone . Since the OB mutation is an insertion after residue 591 in the C-terminal domain of UL9, in order this for mutation to have an effect, full-length protein must be synthesized at least transiently. Interestingly, we found that overexpression of UL9-MV lowered the wild-type UL9 protein level and overrode the inhibitory effect of wild-type UL9 in the plaque reduction assay. In addition, the partial ability of the motif V mutant protein to enter the nucleus suggests that the full-length version of this mutant protein may be transiently expressed, since the NLS is located at the very C terminus. These data, taken together, suggest that the MV mutant protein is transiently expressed as a full-length protein and is capable of interacting with wild-type UL9 and influencing both UL9 levels and the progression of the infection. Based on the results presented in this paper, it is tempting to speculate that the expression levels of UL9 are regulated in the course of infection. It is possible that UL9 functions early in infection and that later in infection it is targeted for degradation. Thus, the regulation of steady-state levels of UL9 may play an important role in controlling viral infection. Posttranslational regulation of protein levels has been shown to play a major role in various cellular processes. Targeting specific proteins for degradation, often through an ubiquitin-mediated proteosomal pathway, is a commonly employed regulatory mechanism in cell cycle progression, apoptosis, and other biological processes . For instance, the specific degradation of the human cdc6 protein, a protein essential for DNA replication, was recently reported. During apoptosis induced by various stimuli, cdc6 is cleaved by caspase 3, preventing wounded cells from replicating . Interestingly, UL9 has been shown to be phosphorylated during infection, although the sites of phosphorylation have not been identified and the biological significance of the phosphorylation has not been elucidated . Furthermore, the sequence analysis of UL9 revealed the presence of an internal PEST sequence at a position near the G354A mutation. PEST sequences are proline-, serine-, and threonine-rich motifs which have been shown to target proteins for ubiquitination and degradation, often in a phosphorylation-dependent fashion . Taken together, these data suggest a possible mechanism for the regulation of UL9 through phosphorylation at or near the PEST sequence. It is possible that during infection, the steady-state levels of UL9 are controlled through a ubiquitin-mediated pathway. The G354A mutation may act to alter the conformation of the UL9 protein by exposing the PEST sequence and enhancing the degradation of itself and of wild-type UL9, with which it can interact. This scenario, although speculative, is consistent with the available data. Further studies will be necessary to determine whether the putative PEST sequence and phosphorylation play a role in the regulation of UL9 levels during infection. Implications of UL9 transdominance for UL9 dimerization. | Since wild-type UL9 is a dimer in solution and the UL9 CTD is a monomer , it has been assumed that the region responsible for dimerization is localized within the N-terminal domain of UL9 (residues 1 to 535). We attempted to map the dimerization region of UL9 by combining deletion analysis with biochemical methods, but our efforts were hampered by the protein insolubility of the tested deletions (data not shown). In this paper, we showed by immunofluorescent microscopy that the UL9 NTD transfected alone is cytoplasmic. In contrast, when cotransfected with wild-type UL9, it is imported into the nucleus. Presumably, the UL9 NTD heterodimerizes with wild-type UL9 and is imported into the nucleus by utilizing the NLS located at the C terminus of wild-type UL9. This result confirms the notion that the dimerization determinant(s) is localized in the N terminus of UL9. If our model that the transdominance phenotype of the UL9 NTD is mediated by dimerization is correct, then the plaque reduction assay might be used to map the region(s) of UL9 that is involved in dimerization. Our results show that the shortest N-terminal transdominant fragment is the UL9-Delta322-851 truncation, suggesting that at least one region involved in dimerization may lie within the first 322 amino acids of UL9. Further experiments will be needed to test this prediction. In summary, in this paper, we have shown for the first time a direct correlation between the steady-state protein level of UL9 and the level of inhibition of HSV-1 infection. Transdominance appears to be mediated by overexpression, origin-binding activity, and dimerization. whereas potentiation is apparently caused by the ability of the potentiating UL9-MV mutant to influence the steady-state levels of wild-type UL9. Our results suggest that UL9 protein levels are an important factor in the progression of HSV-1 infection and perhaps subject to regulation. FIG. 1. : N-terminal fragment but not full-length UL9-MV protein is detected in insect, mammalian, and bacterial cells expressing UL9-MV. N-terminal fragment but not full-length UL9-MV protein is detected in insect, mammalian, and bacterial cells expressing UL9-MV. (A) Western blot analysis of cell lysates of insect cells infected with baculovirus expressing wild-type UL9 or UL9-MV with the 17B antibody, recognizing the N-terminal 35 residues of UL9 (lanes 1 to 3), or the R250 antibody, recognizing the far C terminus of UL9 (lanes 4 to 6). Full-length UL9 is marked with an asterisk, and the N-terminal UL9 fragments are marked with single and double arrowheads. The molecular size markers (10-kDa protein ladder from Gibco) are depicted on the right. (B) Western blot analysis of cell lysates of insect (lanes 1 and 2), mammalian (lanes 3 and 4), and bacterial (lanes 5 and 6) cells expressing the UL9-MV mutant. The 17B anti-UL9 antibody was used for detection in lanes 1 to 4. The anti-His tag India probe was used for detection in lanes 5 and 6. Broad Range (New England Biolabs) prestained molecular size markers are depicted on the right. FIG. 2. : UL9 N-terminal domain is transdominant. UL9 N-terminal domain is transdominant. (A) The plaque reduction assay with pCDNA3-UL9-NT and control plasmids was performed as described in Materials and Methods. The UL9-MI-expressing plasmid is shown as a representative of a transdominant mutant, and the UL9-MV-expressing plasmid was used as a representative of a potentiating mutant. EP, empty plasmid; NP no plasmid. The error bars reflect the standard deviation calculated from at least three independent experiments. (B) The 17B antibody, recognizing the N-terminal 35 amino acids of UL9, was used to detect UL9 variants in concentrated lysates from cells transfected with the plasmids used in the plaque reduction assay, described for panel A. Broad Range prestained protein markers are depicted on the right. FIG. 3. : Level of transdominance of UL9-NT depends on its stability and/or level of expression. Level of transdominance of UL9-NT depends on its stability and/or level of expression. (A) A plaque reduction assay with the indicated UL9 NTD variants was performed as described in Materials and Methods. The UL9-MV-expressing plasmid was used as a representative for a potentiating mutant. EP, empty plasmid; NP no plasmid. The error bars reflect the standard deviation calculated from at least three independent experiments. (B) The 17B antibody, recognizing the N-terminal 35 amino acids of UL9, was used to detect UL9 variants in concentrated lysates from cells transfected with the plasmids used in the plaque reduction assay, described forA. Broad Range prestained protein markers are depicted on the right. (C) Titration of the amount of plasmid DNA used for transfection is shown. The 17B antibody detected a protein species with an approximate molecular mass of 60 kDa, which correlates with the expected size of the UL9 NTD. FIG. 4. : Presence of motif V (G354A) mutation by itself is not sufficient to confer the potentiating phenotype. Presence of motif V (G354A) mutation by itself is not sufficient to confer the potentiating phenotype. (A) Diagram of UL9 truncations. The N-terminal domain of UL9 is depicted as an open box. The C-terminal domain of UL9 is depicted as a gray box. The UL9 helicase motifs are depicted as black boxes. An asterisk points the position of the G354A mutation. The black line below the UL9 gene depicts the UL9-MV N-terminal fragment seen by Western blot analysis. (B) Plaque reduction assays with the indicated UL9 truncations were performed as described in Materials and Methods. The UL9-MV plasmid was used as a representative of a potentiating mutant. EP, empty plasmid; NP no plasmid. The error bars reflect the standard deviation calculated from at least three independent experiments. (C) The 17B antibody was used to detect UL9 truncations in concentrated lysates from Vero cells transfected with the plasmids used in the plaque reduction assay, described for panel A. Broad Range prestained protein markers are depicted on the right. FIG. 5. : Overexpression of UL9-MV affects the steady-state protein levels of UL9 and is able to override the inhibitory effect of wild-type UL9 in the plaque reduction assay. Overexpression of UL9-MV affects the steady-state protein levels of UL9 and is able to override the inhibitory effect of wild-type UL9 in the plaque reduction assay. (A) Western blot analysis of lysates from cells transfected with pCDNA3-UL9-WT, pCDNA3-UL9-MV, or pCDNA3 empty plasmid (EP). The 17B anti-UL9 antibody was used for detection in lanes 1 to 8. The polyclonal anti-UL6 antibody was used for detection in lanes 9 to 12. The full-length UL9 is depicted with an asterisk (*). The UL9-MV fragment is depicted with two asterisks (**). UL6 is depicted with an arrowhead. A faster-migrating band cross-reacting with the anti-UL6 antibody is evident in lanes 9 to 12. Broad Range prestained molecular size markers are depicted on the right. (B) Titration of the UL9 amount in cell lysates with the 17B antibody. (C) A plaque reduction assay with the indicated plasmids was performed as described in Materials and Methods. The mean normalized plaque number and standard deviation were calculated from at least three independent experiments. FIG. 6. : The N-terminal domain of UL9 is able to dimerize. The N-terminal domain of UL9 is able to dimerize. Vero cells were transfected with UL9-NT-AU1 (panels a, b, and c), wild-type UL9 (panels d, e, and f), or both (panels g, h, and i) and stained with anti-AU1 and RH7 antibodies. Green (fluorescein) represents staining with anti-AU1 antibody. Red (Texas Red) represents staining with RH7 antibody. The arrowheads mark untransfected cells in the same field, which can be used as a reference for background staining. Panels c, f, and i are merged images of panels a and b, d and e, and g and h, respectively. FIG. 7. : Cotransfection of UL9-MV with wild-type UL9 results in enhanced nuclear localization. Cotransfection of UL9-MV with wild-type UL9 results in enhanced nuclear localization. Vero cells were transfected with UL9-MV (a, b, and c), AU1-wild-type UL9 (d, e, and f), or both (g, h, and i) and stained with 17B (anti-UL9-NT) and RH7 (anti-UL9-CT) antibodies. Green (fluorescein) represents staining with 17B antibody. Red (Texas Red) represents staining with RH7 antibody. The arrowheads mark untransfected cells in the same field, which can be used as a reference for background staining. Panels c, f, and i are merged images of panels a and b, d and e, and g and h, respectively. Panels j and k contain images of Vero cells transfected with AU1-WT-UL9 and stained with anti-AU1 and RH7 antibodies. Green (fluorescein) represents staining with anti-AU1 antibody. Red (Texas Red) represents staining with RH7 antibody. FIG. 8. : Inhibitory effect of wild-type UL9 overexpression correlates with UL9 protein levels. Inhibitory effect of wild-type UL9 overexpression correlates with UL9 protein levels. (A) Transient transfection-complementation assay was performed as described in Materials and Methods. The bars indicate the mean complementation index. The error bars reflect the standard deviation calculated from three independent experiments. (B) Western blot analysis of lysates from cells transfected with the plasmids used in the complementation assay described in the legend to panel A. The R250 antibody was used to detect UL9, and polyclonal antitubulin antibody was used as a reference for sample loading. It has been estimated that lane 2 contains two- to threefold less UL9 than lane 3. On the other hand, lane 3 contains 1.5- to 2-fold less UL9 than lane 4. FIG. 9. : Model for the mechanism of potentiation-transdominance of the UL9 helicase motif mutants. Model for the mechanism of potentiation-transdominance of the UL9 helicase motif mutants. The two-stage model for HSV-1 replication is depicted in the center of the figure. The behavior of wild-type UL9, the C- and N-terminal domains, and the transdominant motifs I, Ia, II, and VI is shown above the two-stage model. The cotransfection of these UL9 variants with wild-type infectious DNA results in decreased plaque number, as shown by the downward-facing arrow. Parentheses signify the proposed stage of inhibition of the viral infection. The behavior of the potentiating mutant MV is shown below the model with upward-facing arrows to show both the proposed stage of potentiation and the overall increase in plaque number. Abbreviations: I, inhibitory; TD, transdominant; P, potentiating. See text for details. Backmatter: PMID- 12915565 TI - Pre-S1 Antigen-Dependent Infection of Tupaia Hepatocyte Cultures with Human Hepatitis B Virus AB - The susceptibility of the tree shrew Tupaia belangeri to human hepatitis B virus (HBV) has been demonstrated both in vivo and in vitro. In this study, we show that purified HBV infects primary T. belangeri hepatocyte cultures in a very specific manner, as detected by HBV covalently closed circular DNA, mRNA, HBV e antigen, and HBsAg production. A monoclonal antibody (MAb), MA18/7, directed against the pre-S1 domain of the large HBs protein, which has been shown to neutralize infectivity of HBV for primary human hepatocytes, also blocked infection of primary Tupaia hepatocytes. MAbs against the pre-S2 domain of HBs inhibited infection only partially, whereas an S MAb and polyvalent anti-HBs antibodies neutralized infection completely. Thus, both pre-S1 and S antigens are necessary for infection in the tupaia. Using subviral particles, >70% of primary Tupaia hepatocytes are capable of specific binding of pre-S1-rich HBsAg, showing localization in distinct membrane areas. The data show that the early steps of HBV infection in Tupaia hepatocyte cultures are comparable to those in the human system. Keywords: Introduction : Persistent infection with hepatitis B virus (HBV), a small enveloped DNA virus, is a frequent cause of liver cirrhosis and hepatocellular carcinoma. With >350 million chronically infected people, HBV is a major health problem worldwide. HBV is a member of the Hepadnaviridae, a familiy of hepatotropic and highly species-specific viruses which are able to infect in culture only well-differentiated primary hepatocytes of their specific hosts . Human and related hepadnaviruses from higher primates (Hominoidae) are considered to form one virus species, HBV (ICTVdB: The Universal Virus Database of the International Committee on Taxonomy of Viruses [.]). Further, hepadnavirus species are found in woolly monkeys (Platyrrhini) and in woodchucks and squirrels (both Rodentia) and form the genus Orthohepadnavirus (.). Since HBV naturally infects only humans and higher primates, many efforts have been made to establish a small-animal model for the study of HBV infection (reviewed in reference ). Some mouse-related models, like the uPa-mouse or trimera-mouse , use transplanted human hepatocytes, but they are difficult to work with. Experimental hepatitis delta virus and HBV infection (in 1995 and 1996, respectively) of a small animal, the tupaia , have been reported. These mammals, also called tree shrews, are found in tropical forests of Southeast Asia. In contrast to previous assumptions , they are not primordial primates but form their own order, Scandentia, in the family Tupaiidae . Tupaias can be infected experimentally with HBV in vivo, but the infection is self-limiting and does not lead to a chronic carrier state . Furthermore, a natural tupaia hepatitis B virus has not yet been reported. Primary Tupaia hepatocyte cultures can be infected with HBV and with woolly monkey hepatitis B virus, but not with woodchuck hepatitis virus . This finding confirmed the susceptibility of primary Tupaia hepatocyte cultures to primate hepadnaviruses, but the specificity of virus attachment and entry has not yet been demonstrated. It is known that after a nonspecific uptake of hepadnavirus genomes, later steps of the viral life cycle are not rigidly host restricted, as is shown by transgenic HBV-producing mice . The orthohepadnaviruses contain three coterminal surface proteins (large [LHBs], medium [MHBs], and small [SHBs] HBs) with the three domains pre-S1, pre-S2, and S. Attachment of HBV to human hepatocytes is mediated by the pre-S1 domain and is blocked by a monoclonal antibody (MAb) against pre-S1 (Ma18/7) . Furthermore, antibodies against S protect against infection, whereas the pre-S2 domain seems to be nonessential for attachment . In the study presented here, we established optimized primary hepatocyte cultures from tupaia livers and developed quantitative real-time PCR techniques for detecting HBV DNA transport to the nucleus and viral gene expression. We found that uptake and gene expression of HBV may be specifically blocked by antibodies against those protein sequences which have been found to be essential for infection of human hepatocytes. The results show that primary Tupaia hepatocyte cultures are suitable for studying early steps in the life cycle of HBV, for assay of its infectivity, and for assays of neutralizing antibodies. MATERIALS AND METHODS : HBV-containing plasma. | HBV and HBsAg was isolated from the plasma of two chronic HBV carriers. One carrier (ID1) had 4.3 x 109 HBV genomes/ml, genotype D, and 100 mug of HBsAg subtype ayw2/ml and was positive for HBV e antigen (HBeAg). The second carrier (ID259) had 2 x 109 HBV genomes/ml, genotype D, and 10 mug of HBsAg/ml subtype ayw2 and was positive for HBeAg. HBV antibodies. | Monoclonal antibody (MAb) MA18/7 was generated using purified HBV particles for immunization . MA18/7 detects an epitope (DPXF) in the pre-S1 amino acids 20 to 23 (31 to 34 in genotype A). Other MAbs were generated by immunization with purified HBsAg particles and were characterized as described previously . Polyvalent anti-HBs serum with a high proportion of antibodies against the common "a" determination was generated by immunization as follows. A sheep was injected subcutaneously with 200 mug of highly purified native HBs protein filaments (genotype D) in complete Freund's adjuvant. Booster injections with 200 mug of highly purified native HBs protein filaments of different genotypes in incomplete Freund's adjuvant were given after 3 (genotype A) and 6 (genotype C) weeks. After 9 weeks, a mixture of all three genotypes (200 mug) was injected. Blood was collected 10 days after the last booster injection (Eurogenetec, Searing, Belgium) and tested for reactivity to HBsAg by Laurell electrophoresis and Western blotting . Immunoglobulin G (IgG) was purified from that antiserum by precipitation with 17% (wt/wt) Na2SO4. Isolation and purification of HBV virions and subviral particles from plasma of HBV-infected patients. | HBV and subviral particles were purified as follows. Eighteen milliliters of human plasma was ultracentrifuged through a discontinuous sucrose density gradient (15, 25, 35, 45, and 60% [wt/wt]) in TNE buffer (20 mM Tris-HCl [pH 7.4], 140 mM NaCl, 1 mM EDTA) for 15 h at 25,000 rpm with a SW 28.38 rotor (Beckman, Munich, Germany) . Virus-containing fractions at 40 to 45% sucrose were identified by quantitative real-time PCR (LightCycler system, Roche, Mannheim, Germany) using primers and hybridization probes against the HBV X region as described previously . The assay was calibrated using the Eurohep reference plasma , which has been converted to a World Health Organization international standard sample . Sucrose was removed by using Vivaspin 6 concentrators (Vivascience, Sartorius, Germany), and the concentrated virus in TNE buffer (0.2 ml from 19 ml of plasma) was snap-frozen in liquid N2 and stored at -70C until it was used. The fractions containing 35 to 40% sucrose were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and silver stained for detection of HBs proteins in subviral HBsAg particles. Fractions were pooled, adjusted with solid CsCl to 1.30 g/ml, and layered within a CsCl density gradient ranging from 1.16 to 1.35 g/ml in TNE buffer for 36 h at 25,000 rpm in an SW 28.38 rotor (Beckman). HBs protein-containing fractions at 1.18 to 1.22 g/ml were dialyzed and purified by sedimentation into a discontinuous sucrose density gradient (5, 10, 15, 20, 25, and 60% [wt/wt]) in TNE buffer for 15 h at 25,000 rpm with an SW 28.38 rotor. The LHBs protein-rich fractions containing HBsAg filaments were separated from the HBs spheres, pooled, and concentrated in a Vivaspin-20 ultrafiltration device (Vivascience). The concentration of purified HBsAg was estimated from the optical density at 280 nm (OD280), assuming a value of 4.3 for 1 mg of HBsAg/ml . The purity of the HBsAg filament preparation was controlled by gel electrophoresis and silver staining of the gel as described previously . Isolation and culture of primary Tupaia hepatocytes. | Tupaia belangeri (the Asian tree shrew) was bred in the animal facility of the Institute of Anatomy and Cell Biology at the Justus Liebig University of Giessen. Primary hepatocytes were isolated by the modified two-step collagenase method first described by Seglen . In brief, the livers of 2- to 4-year-old tupaias (8 to 10 g [wet weight]) were perfused via the portal vein with Hanks solution (Invitrogen, Karlsruhe, Germany) containing 5 mM EGTA at a flow rate of 10 ml/min for 7 min, followed by perfusion with Dulbecco's modified Eagle's medium (Invitrogen) containing 0.05% collagenase (type II; Sigma, Deisenhofen, Germany) for at least 15 min at 37C. The cell suspension was rinsed through a layer of nylon filter (pore size, 250 mum) and thereafter through a 75-mum-pore-size nylon filter, yielding isolated single or double hepatocytes. After the hepatocytes were pelleted three times at 30 x g for 5 min at 4C, the cells were resuspended in plating medium (Williams E medium [Invitrogen] supplemented with 5 mug of insulin/ml, 5 mug of transferrin/ml, 5 ng of sodium selenite/ml, 100 mug of gentamicin/ml, and 0.25 mug of amphotericin B/ml with 5% fetal calf serum) and poured on collagen-coated coverslips (collagen I from rat tail; BD Bioscience, Heidelberg, Germany) in 24-well plates (103 hepatocytes per well) or on collagen-coated cell culture dishes (105 hepatocytes per 3-cm-diameter dish). After 2 to 4 h at 37C, the plating medium with unattached cells was removed and the cells were incubated with maintenance medium (Williams E medium supplemented with 0.1% bovine serum albumin [crystallized; Sigma], 5 mug of insulin/ml, 5 mug of transferrin/ml, 5 ng of sodium selenite/ml, 50 muM hydrocortisone, 0.1 muM dexamethasone, 2% dimethyl sulfoxide, 100 mug of gentamicin/ml, and 0.25 mug of amphotericin B/ml) until infection. Infection of primary Tupaia hepatocyte cultures. | Primary Tupaia hepatocyte cultures were inoculated 3 days after being plated with purified virus at appropriate concentrations in infection medium (Williams E medium supplemented with 0.1% bovine serum albumin [crystallized; Sigma], 2% dimethyl sulfoxide, 100 mug of gentamicin/ml, and 0.25 mug of amphotericin B/ml) for 12 h at 37C. The cells were extensively washed with infection medium and further cultivated in maintenance medium. The medium was changed every 3 days and measured for the appearance of secreted HBV antigens. After 12 days, the cells were washed with phosphate-buffered saline (PBS) and lysed with lysis buffer (100 mM NaCl, 50 mM Tris, 0.1% Triton X-100, 5 mM MgCl2, pH 8.0) for 10 min at 4C. Nuclei were separated from the cytoplasmic fraction by centrifugation at 270 x g for 2 min at 4C. The cytosolic and nucleic fractions were snap-frozen in liquid nitrogen and stored at -70C until they were used. Assay for HBV-specific proteins. | HBeAg was determined semiquantitatively by a commercially available enzyme-linked immunosorbent assay (ELISA) (AxSym; Abbott Laboratories, Weisbaden, Germany). HBs was measured by an in-house sandwich ELISA as described previously . In brief, 96-well plates (Maxi-Sorb; Nunc, Hereford, United Kingdom) were coated with 1 mug of MAb C20/2 per ml against HBsAg a determinant in PBS for 2 h at 37C. After being extensively washed (three times with PBS-0.1% Tween 20 and two times with PBS) and subsequently blocked with 10% fetal calf serum in PBS for 1 h at 37C, 100 mul of undiluted cell culture supernatant from infected hepatocyte cultures was incubated for 12 h at 4C. After removal of the supernatant, the plate was again washed with PBS and PBS-0.1% Tween 20, and peroxidase-conjugated anti-HBs from the Enzygnost HBsAg ELISA kit (Dade Behring, Marburg, Germany) was added to each well and incubated for 1 h at 37C. After the plate was washed as described above, an o-phenylenediamine-H2O2 substrate (tablets from Abbott Laboratories) was added for 15 min at room temperature, and the amount of colored product was determined by measuring the OD492. The concentration of HBsAg was calculated in comparison to a dilution series of a well-defined HBV plasma containing 100 to 0.01 ng of HBsAg, according to previous calibrations . Quantification of HBV cccDNA in HBV-infected cells. | Episomal DNA was isolated from nuclei of infected Tupaia hepatocytes by Hirt extraction as described previously . Eluted HBV DNA was further purified using a DNA extraction kit (Highpure; Roche). The purified DNA was used for PCR (LightCycler real-time system with Fast Start PCR kit; Roche) with the primer set HBV C1 sense (nucleotides 1580 to 1596; 5'TGCACTTCGCTTCACCT) and HBV C1 antisense (nucleotides 2314 to 2298; 5'AGGGGCATTTGGTGGTC) for detection of HBV DNA without nick and gap (covalently closed circular DNA [cccDNA]) under the following conditions: denaturation (1 time), 95C for 10 min; amplification (45 times), 95C for 10 s, 57C for 20 s, and 72C for 32 s; slope of temperature changes, 20C/s. The resulting amplificate (735 bp) was subjected to 2% agarose gel electrophoresis and visualized by ethidium bromide staining. Purification of poly(A)+ RNA and quantitative one-tube HBV RT-PCR. | Poly(A)+ RNA from cytoplasmic fractions of infected cells was isolated by using oligo(dT)-coated paramagnetic beads (Dynal, Oslo, Norway) according to the manufacturer's guidelines. The HBV mRNA in this extract was quantified using the one-tube real-time reverse transcription (RT)-PCR kit for LightCycler (Roche) according to the manufacturer's guidelines with primers and hybridization probes for the X region as described previously . For generation of a reference sample HBV RNA, the X region was transcribed from plasmid pCX using the T7 in vitro transcription kit (Roche) according to the manufacturer's guidelines. The resulting RNA was incubated with 10 U of RNase-free DNase I (Roche)/mul for 1 h at 37C and further purified by phenol-chloroform extraction. The RNA copy number was determined by measuring the OD260. The HBV RNA served as a standard in quantitative HBV RT-PCR using serial dilutions. Binding and uptake studies. | Primary Tupaia hepatocytes (1 x 103 to 5 x 103) were plated on collagen-coated glass coverslips in plating medium as described above. After 2 to 4 h at 37C, the plating medium with unattached cells was removed and the cells were further incubated with maintenance medium until they were used. For binding experiments, cells were washed twice with Williams E medium and once with binding medium (Williams E medium supplemented with 0.1% crystallized bovine serum albumin sigma). Purified pre-S1-rich HBs filaments were diluted in binding medium to yield a final concentration of 2 mug/ml. For inhibition of binding, the subviral particles were preincubated with an excess of purified polyvalent anti-HBs IgG (from sheep) for 1 h at 37C. The cells were incubated for various times at various temperatures and were washed several times with ice-cold Williams E medium. The cells were fixed with 4% paraformaldehyde for 30 min at 4C and permeabilized with 0.1% Triton X-100 in PBS for 30 min at room temperature. HBsAg staining was done with the APAAP staining kit (DAKO, Hamburg, Germany) according to the manufacturer's guidelines using an anti-HBs specific MAb at 1:400 dilution (Novocastra, Newcastle, United Kingdom). The slides were counterstained with Meyer's hemalaun (Merck) and mounted in Glycergel (DAKO). RESULTS : Isolation and cultivation of primary Tupaia hepatocytes. | The two-step collagenase perfusion of tupaia livers routinely yielded pure hepatocyte cultures with a viability of >98% hepatocytes (in the trypan blue assay) and <1% contamination with nonhepatocytes . By plating cells on collagen-coated dishes or coverslips with a well-defined maintenance medium, hepatocytes are maintained in a differentiated state, with the appearance of bile canaliculi after 2 days (Fig. , day 3). These primary Tupaia hepatocyte cultures have been maintained for >1 month in a highly differentiated state. Infection of primary Tupaia hepatocytes with purified HBV. | None of the previously known cell culture systems for hepadnaviruses produces more output than input virus, and a cytopathic effect has never been detectable. Instead, "infectivity" is considered the ability of hepadnaviruses to specifically enter cells and have their nicked and gapped DNA transported into the nucleus, where it is converted into cccDNA . The formation of cccDNA is a prerequisite for entering the replication cycle of HBV and is generally considered to be the earliest marker of HBV infection of the hepatocyte . To detect the very low cccDNA content in infected hepatocytes, we used a special PCR protocol that discriminates specifically between the nick-and-gap-containing HBV genome of the infecting virus and the cccDNA in the nucleus. Reliable discrimination between cccDNA and relaxed circular DNA (rcDNA) is only possible when a special temperature protocol is used for PCR in quartz capillaries with very rapid temperature changes (Fig. , left). This PCR detects 1,000 copies of cloned dimeric HBV DNA within a plasmid but not 105 genome equivalents (ge) of HBV DNA from human plasma-derived virions (Fig. , right). For infection experiments, we usually incubated 2 x 105 to 5 x 105 primary Tupaia hepatocytes in 3-cm-diameter dishes with defined numbers of purified plasma-derived HBV particles determined as HBV ge. To determine the minimal amount of purified HBV necessary to detect infection in our system, we titrated the input virus; 105 hepatocytes were infected with ratios of 1,000, 100, 10, 1, and 0.1 ge/hepatocyte and were subsequently cultivated as described in Materials and Methods. Using cccDNA-specific PCR, after 12 days we could detect HBV DNA without nick and gap in nuclear extracts of cultures incubated with HBV input virus down to a ratio of 10 ge/hepatocyte. When the viral input was increased from 10 to 100 ge/hepatocyte, the amount of cccDNA also increased (Fig. , center). Further increasing the viral input (to 1,000 ge/hepatocyte) increased the cccDNA content only marginally. The HBV mRNA in the cytosolic fraction was isolated on day 12 postinfection (p.i.) by oligo(dT) purification and determined by real-time RT-PCR (with primers from the X region) as described above. Due to the relatively large abundance of HBV mRNA in the cytoplasm of the same cells on day 12 p.i., a 1-ge/hepatocyte ratio of input virus was sufficient to detect an HBV infection in primary Tupaia hepatocyte cultures. The amount of cytosolic HBV mRNA increased proportionally with increased HBV input . Kinetics of infection. | By changing the culture medium every 3 days, we determined the onset of HBsAg and HBeAg production after infection. The soluble HBeAg is not associated with virus and is therefore not present in the purified input virus . Thus, HBeAg is a marker for de novo HBV protein expression. With a ratio of 100 HBV ge/hepatocyte, newly synthesized HBeAg could be detected after 6 days and reached a plateau at day 9 to day 12. Increasing the viral input to 1,000 HBV ge/hepatocyte did not lead to increased HBeAg secretion on day 12 (data not shown), while reducing the input from 100 to 10 HBV ge/hepatocyte also decreased the HBeAg output 10-fold, and HBeAg production could first be detected only on day 9. At 1 ge/hepatocyte or below, no significant ELISA signal for HBeAg could be detected. HBsAg production followed a pattern similar to that of HBeAg production. Although present in the input, HBsAg could not be detected on day 3 . This excluded the possibility that nonspecific binding to the cells and subsequent release of virions into the medium without de novo synthesis would generate an HBsAg signal. With a ratio of 100 HBV ge/hepatocyte, newly synthesized HBsAg could be detected 6 days p.i. and reached a plateau at day 9 to day 12. As detected for HBeAg, a further increase of the input to 1,000 HBV ge/hepatocyte did not lead to increased HBsAg secretion (data not shown), while reducing the input from 100 to 10 HBV ge/hepatocyte also decreased the HBsAg output 10-fold, with HBsAg secretion first detectable on day 9. A virus titer of 1 ge/hepatocyte still led to detectable production of HBsAg on day 12. A viral inoculum of 0.1 ge/hepatocyte never generated a detectable HBsAg or HBeAg signal (data not shown). Thus, 1 HBV ge/hepatocyte is necessary to establish a detectable HBV infection in the type of primary Tupaia hepatocyte culture used in this study, while 100 ge/hepatocyte led to maximal HBsAg and HBeAg secretion. Assay of neutralizing antibodies against HBV. | To determine whether the uptake of HBV by primary Tupaia hepatocytes required specific attachment sites on HBV, we preincubated the virus with various antibodies against HBV surface proteins. We used a large stoichiometric excess (>10-fold) of different MAbs directed against the pre-S1, pre-S2, or S domain of HBV surface proteins . Preincubation of virus with MAb MA18/7, which binds to amino acids 20 to 23 of the pre-S1 domain (DPAF), completely prevented infection, because HBeAg and HBsAg production were not detectable in culture supernatants as they were in control infections . Analysis of HBV cccDNA content and HBV mRNA in these cells on day 12 p.i. confirmed that preincubation with MAb MA18/7 also prevented the formation of cccDNA and HBV mRNA . Complete inhibition of infection could also be achieved by preincubation with polyclonal anti-HBs antibodies produced in sheep against purified plasma-derived HBsAg of different HBV genotypes (Fig. and C). Among the two MAbs directed against the S domain, only MAb C20/2, which recognizes a conformational epitope, was also capable of inhibiting infection, as shown by negative results for HBeAg and HBsAg (Fig. and C) and HBV cccDNA and mRNA . MAb 1-9C1, which binds to a linear genotype D-specific S epitope, was not able to completely inhibit infection (Fig. and C). MAbs against the pre-S2 domain showed variable neutralization capacity. MAb 2-12F2 (an IgM antibody) did not inhibit infection efficiently, since we could clearly detect HBV cccDNA 12 days p.i. The HBV mRNA signal was only 74% less than in the control infection. HBeAg and HBsAg secretion were also reduced by 74 and 82%, respectively. The other IgG pre-S2 MAbs neutralized more efficiently but were not able to inhibit infection completely. MAbs recognizing the amino-terminal region of the pre-S2 domain (residues 5 to 17) showed from 96 to 99% inhibition of HBsAg secretion (MAbs S26 and 2-11B1 ), except for MAb Q19/10, which preferentially binds to the N-glycosylated pre-S2 domain of MHBs. In the case of MAb S26, the cccDNA contents in the nuclei of infected cells dropped below the detection limit , but an HBV mRNA signal was still detectable and was reduced only 10-fold, indicating active replication of HBV . Binding and uptake of HBsAg. | To study the binding and uptake of HBV by primary Tupaia hepatocytes, we used highly purified pre-S1-containing HBsAg filament-rich particles from the same patients' plasmas from which we isolated viral fractions for infection experiments. After incubating cell cultures with HBsAg for 1 h at 37C, followed by several washes, >70% of the hepatocytes had retained HBsAg and reacted with an HBsAg-specific immune stain . This effect could be completely abrogated by preincubating HBsAg with a polyclonal anti-HBs antibody . Interestingly, HBsAg was not evenly distributed on the membranes of hepatocytes but appeared to be located in distinct areas of the cell surface (Fig. and C). These structures, most likely actin bundles containing filopodia or microvilli, seem to contain large concentrations of HBV receptors. Some hepatocytes with no visible signs of filopodia or microvilli did not show HBsAg on the cell surface but rather a perinuclear accumulation of HBsAg. This pattern may represent active transport from the plasma membrane to cellular subcompartments. After uptake and further cultivation for 2 h, the signal disappears (data not shown). A similar pattern has been observed using primary human hepatocytes . HBsAg synthesis in infected hepatocytes. | For detection of de novo HBsAg in infected hepatocytes, we infected primary Tupaia hepatocytes with 100 ge/hepatocyte on collagen-coated coverslips. After 12 days, most cells were clearly HBsAg-negative, but up to 5% of the hepatocytes showed strong cytoplasmic staining for HBsAg . Maximum infectivity (20% of cells clearly HBsAg positive) was achieved by increasing the viral input to 10,000 HBV ge/cell . This effect could be inhibited completely by preincubating the input virus with polyclonal anti-HBs antibody . The majority of infected cells showed clear cytoplasmic and occasionally very strong perinuclear staining (Fig. and D), while a small population of cells showed only marginal staining for HBsAg expression, even at the highest viral input . The localized staining most likely reflects HBsAg associated with the endoplasmic reticulum. A similar pattern of HBsAg expression has also been detected using a human hepatoma cell line (HepG2) transfected with HBV DNA (data not shown). DISCUSSION : In this study, primary Tupaia hepatocyte cultures were infected with HBV. There have been previous reports of successful in vitro infection of primary Tupaia hepatocytes with HBV or hepatitis delta virus , but details of the viral attachment and neutralization by specific antibodies have not yet been described. Furthermore, the methods used to detect infection were relatively insensitive semiquantitative hybridization techniques for HBV cccDNA and HBV mRNA . We have developed highly sensitive PCR protocols for cccDNA and mRNA, and furthermore, we used de novo HBeAg or HBsAg production as a marker of successful infection. More sensitive detection of infection allows the use of fewer hepatocytes, which are available only in limited amounts. Similar to what Kock et al. reported , we also achieved our best results with purified HBV particles instead of whole serum. As expected, HBeAg and HBsAg production increased with increasing concentration of the input virus. A ratio of 1 ge/hepatocyte was the minimum for establishing a detectable infection. The highest levels of secreted viral antigens resulted from a ratio of 100 ge/hepatocyte, while at 1,000 ge/hepatocyte, no increase in HBsAg and HBeAg beyond that obtained with the 100-ge/hepatocyte input was detected. HBV cccDNA levels showed the strongest increase at 10- to 100-ge/hepatocyte viral input. On the other hand, the total HBV mRNA content steadily increased with 1 to 1,000 input virus ge/hepatocyte. The reason for this difference among HBV cccDNA, mRNA, and protein production is at present unclear. The specificity of the viral attachment and uptake leading to infection was shown by the use of several MAbs against the HBs proteins. LHBs is preferentially localized on HBV particles . Its pre-S1 domain plays an important role during attachment to a liver receptor(s) . Furthermore, antibodies raised against a pre-S1 peptide (amino acids 21 to 47) neutralized HBV infection of chimpanzees . The pre-S1 antibody used in this study (MA18/7) neutralized HBV infection of primary Tupaia hepatocytes by binding to a pre-S1 epitope that overlaps with the species specificity-determining region of HBV using primary human hepatocytes . These data support the conclusion that HBV uses homologous pre-S1 receptor systems on human and Tupaia hepatocytes for the infection process. Although the MHBs protein is not essential for infection , peptides with the N-terminal half of pre-S2 were found to induce protective immunity and neutralizing antibodies . Furthermore, a MAb recognizing amino acids 13 to 24 of pre-S2 has been reported to neutralize infection of primary human hepatocytes . Indeed, in our study, MAbs against this region (2-11B1 and S26) also showed strong but incomplete neutralizing potential, and the cccDNA contents of the inoculated cultures dropped below the detection limit. Unfortunately, Ryu et al. used detection of cccDNA by Southern blotting as a readout, which is certainly much less sensitive than the techniques used in this study. In our system, we could show that mRNA was still detectable, indicating active gene expression and resulting in HBeAg and HBsAg secretion, although reduced by a factor of 30 to 100 in comparison to the control infection. Thus, none of the pre-S2-specific antibodies had the potential to completely neutralize infection like the pre-S1-binding MAb MA18/7, the conformation-dependent S antibody C20/2, or the polyclonal anti-HBs serum. Although MAb Q19/10 binds strongly to the N-terminal region of pre-S2, it had the lowest neutralization potential of all IgG class MAbs binding to this domain. Binding of MAb Q19/10 to pre-S2 amino acids 1 to 6 is glycan dependent, and removal of this glycan by peptide N-glycosidase F results in complete loss of reactivity . While the pre-S2 domain is present in both LHBs and MHBs, only the pre-S2 domain of MHBs is glycosylated with a complex diantennary glycan at Asn 4 . In LHBs, the pre-S2 domain remains unglycosylated at Asn 4 due to its different topology at the endoplasmic reticulum . Therefore, MAb Q19/10 binds only to the N-glycosylated pre-S2 region of MHBs but not to LHBs. Possibly, the stronger neutralizing potential of some pre-S2-specific antibodies used in this study resulted from binding to the pre-S2 domain of LHBs. This possibility is in agreement with the nonessentiality of MHBs for infectivity . Although no role for pre-S2 protein sequences has been convincingly proven for the attachment and entry of HBV, binding of antibodies to the amino acid sequence 5 to 17 may impair attachment or entry. It is noteworthy that this region of pre-S2 binds modified human serum albumin , which has been postulated to mediate indirect binding to hepatocytes . Our data show that the system is feasible for the assay of neutralizing anti-HBV antibodies. Until now, very few neutralizing studies of anti-HBV antibodies have been done because they required either chimpanzees or primary human hepatocytes . However, the increasing appearance of escape mutants in the S gene after vaccination with SHBs antigen or treatment with HBV immune globulin highlights the need for feasible neutralization assays. In this respect, the predominant role of pre-S1 antibodies in neutralization of HBV infectivity should lead to more attention in strategies for control and prevention of HBV. Currently, SHBs is the only component of the most widely used vaccines. One licensed vaccine on the market also contains MHBs , although the immunogenicity of the pre-S2 sequence of this vaccine appeared weak in humans . There is at present no vaccine on the market which contains the pre-S1 neutralizing epitopes, but two vaccines have already undergone successful clinical evaluation. Whether these vaccines induce protective levels of anti-pre-S1 antibodies is unknown. Infection of primary Tupaia hepatocytes seems to be dependent only upon LHBs and SHBs, as in the human system . Unexpectedly, by the use of subviral particles, >70% of the primary Tupaia hepatocytes were capable of specific binding to pre-S1-rich HBsAg. Interestingly, the bound HBsAg was not evenly distributed but localized to distinct membrane areas of hepatocytes, often associated with special membrane structures like lamellipodi or microvilli. Since the sinusoidal side of hepatocytes in the liver is also heavily covered with microvilli , these structures may be enriched in HBV receptors. Furthermore, many hepatocytes showed signs of HBsAg uptake and transport to perinuclear rather than to perimembranous sites. The staining strongly decreased within 2 h after uptake, which may be due to destruction of the HBsAg conformational epitope by reduction in an endosomal or lysosomal compartment and nonreactivity with anti-HBs. A similar uptake phenomenon has been observed with primary human hepatocytes and fluorescently labeled subviral particles from the same patient using confocal microscopy imaging . Until now, fluorescence microscopy has unfortunately not been applicable in our studies of Tupaia hepatocytes due to strong autofluorescence. Thus, the pathway of HBsAg uptake to Tupaia hepatocytes is unknown and needs to be elucidated. Besides the strong binding and uptake of HBsAg, only a portion of primary Tupaia hepatocytes could be infected to the point that they were actively producing HBsAg at levels detectable by immunocytochemistry. A viral input of 100 ge/hepatocyte resulted in 5% infected hepatocytes. To increase the number of infected hepatocytes to 20%, a 100-times-higher viral input (10,000 ge/hepatocyte) was needed. Only 1,000 to 5,000 out of 105 cells could be infected by an inoculum of 107 particles. In cultures of 105 cells inoculated with 105 particles, obviously very few highly producing cells were sufficient to generate detectable signals for HBV mRNA and HBsAg. This suggests that in our system only one in several thousand particles is infectious. This contrasts with reports that in human plasma, one 50% infectious HBV dose for chimpanzees corresponds to 10 to 100 HBV particles . Unfortunately, infection of Tupaia hepatocytes with HBV-containing plasma requires polyethyleneglycol (PEG) as a nonspecific membrane fusion inducer and does not generate a more efficient infection (reference and our unpublished data). Other results suggest that even with primary human hepatocytes of good quality, only 1 out of 10,000 unpurified recombinant HBV particles results in productive infection (T. Kurschner and H. Schaller, personal communication). With a newly established redifferentiated hepatoma cell line (HepaRG), Gripon et al. reported HBV infection of up to 10% of the cells . It must be noted that these HBV infection studies were done with unpurified recombinant virions and in the presence of the fusion-promoting agent PEG. Although PEG does not induce nonspecific HBV uptake per se, the absence of PEG during the infection process of HepaRG cells reduced HBV mRNA levels to 3% of those for a control infection with PEG . Thus, PEG may favor nonphysiological routes of attachment or entry which may not occur in vivo. In summary, we have shown that the HBV infection of primary Tupaia hepatocytes is specific and comparable to that of primary human hepatocytes. The system has some advantages over other cell culture systems: (i) primary Tupaia hepatocytes are more readily available and show a more constant susceptibility than primary human hepatocytes, which are very often of poor quality; (ii) in contrast to the HepaRG cell line, they do not require the presence of fusion promoters like PEG; (iii) results from in vitro HBV infections using primary Tupaia hepatocytes can be verified in vivo by infecting tupaias with HBV. FIG. 1. : Morphologies of primary Tupaia hepatocyte cultures. Morphologies of primary Tupaia hepatocyte cultures. Shown are phase-contrast micrographs of uninfected primary Tupaia hepatocytes at various time points after two-step collagenase isolation. (A) After perfusion; (B) day 3; (C) day 6; (D) day 9. Hepatocytes preserve their shape and functionality during the time required for infection. Note the frequent presence of two nuclei and the bright borders between the cells (bile canaliculi), typical features of well-differentiated hepatocytes (magnification, x100). FIG. 2. : Detection of replicative intermediates in primary Tupaia hepatocytes infected with increasing amounts of purified virus. Detection of replicative intermediates in primary Tupaia hepatocytes infected with increasing amounts of purified virus. (A) schematic diagram of the early events in HBV infection of hepatocytes. Formation of cccDNA from the incoming HBV rcDNA form with its nick-and-gap structure is essential for RNA transcription and all further steps in establishing an HBV infection. (B) Agarose gel electrophoresis of PCR products specific for HBV double-stranded DNA without nick and gap. A specific amplificate at 735 bp (arrow) could be synthesized from cloned dimeric HBV DNA but not from purified virion rcDNA, with its nick-and-gap structure, from human plasma. HBV cccDNA was detected in nuclear Hirt extracts of HBV-infected primary Tupaia hepatocytes 12 days p.i. or in the positive control cell line HepG2.2.15, but not from the parental HepG2 cell line. Lane M, marker; lane , negative control. (C) HBV mRNA quantification of cytoplasmic extracts from primary Tupaia hepatocytes 12 days after infection with increasing amounts of purified HBV (real-time RT-PCR of the X-gene region). The detection limit was 104 copies per culture. ge, genomic equivalent. FIG. 3. : HBeAg (A) and HBsAg (B) secretion of primary Tupaia hepatocytes infected with increasing amounts of purified HBV. HBeAg (A) and HBsAg (B) secretion of primary Tupaia hepatocytes infected with increasing amounts of purified HBV. The medium was changed every 3 days. Each point represents HBeAg or HBsAg in the supernatant newly synthesized within 3 days. , 100 ge/cell; , 10 ge/cell; , 1 ge/cell. The dotted lines indicate the cutoffs for HBeAg (multiples of cutoff signal) and for HBsAg (0.1 ng/ml). S/CO, sample to cutoff signal. FIG. 4. : Neutralizing epitopes of antibodies against HBV surface proteins. Neutralizing epitopes of antibodies against HBV surface proteins. Schematic diagram of MAb epitopes of LHBs, MHBs, and SHBs (genotype D). NG, N-glycan in the pre-S2 domain of MHBs; OG, O-linked glycan. R122 refers to the subtype determinant y in the antigenic region of SHBs. FIG. 5. : Neutralization test using HBeAg or HBsAg secretion as readout. Neutralization test using HBeAg or HBsAg secretion as readout. (A) Kinetic analysis of HBeAg production of infected primary Tupaia hepatocytes after preincubation of input virus with anti-pre-S1 MAb MA18/7 or with an irrelevant anti-mouse MAb (the dotted line indicates the cutoff). (B) HBeAg production of infected primary Tupaia hepatocytes 12 days p.i. after preincubation of input virus with different MAbs against HBV surface proteins. (C) HBsAg production of infected primary Tupaia hepatocytes 12 days p.i. after preincubation of input virus with different MAbs against HBV surface proteins (the cutoff for HBsAg was 0.1 ng/ml). FIG. 6. : Inhibition of HBV cccDNA formation and transcription by pretreatment of HBV with neutralizing MAbs. Inhibition of HBV cccDNA formation and transcription by pretreatment of HBV with neutralizing MAbs. (A) Agarose gel electrophoresis of PCR products specific for HBV DNA double stranded in the nick-gap region as described in the legend to Fig. . The purified virus inoculum was preincubated before infection with MAbs Ma18/7 (pre-S1), 2-12F2 (pre-S2), S26 (pre-S2), and C20/2 (S) and an irrelevant anti-mouse antibody (anti-mouse). Lane M, marker; lane , negative control. (B) HBV mRNA quantification of cytoplasmic extracts from HBV-infected primary Tupaia hepatocytes 12 days p.i. by real-time RT-PCR. The purified input virus was preincubated before infection with MAbs as for panel A. FIG. 7. : Binding and uptake of subviral HBsAg particles. Binding and uptake of subviral HBsAg particles. Shown is immune staining of Tupaia hepatocytes after incubation with purified LHBs-rich HBV subviral particles for 1 h at 37C with a MAb against the S domain (red) and counterstaining with Meyer's hemalaun (blue). (A, C, and D) HBsAG staining; (C) inhibition by polyclonal anti-HBs. (A and B) Binding of HBsAg particles to primary Tupaia hepatocytes is detected (A) (arrows) (magnification, x100), but not after preincubation of the inoculum with polyclonal anti-HBs antibodies (B) (magnification, x100). (C) At higher magnification, HBs particles (arrows) are detected in distinct areas of the plasma membrane (magnification, x400). (D) Some hepatocytes show a pattern suggesting uptake of HBsAg (magnification, x630). FIG. 8. : (A, C, and D) HBsAg immune staining of infected Tupaia hepatocytes. (A, C, and D) HBsAg immune staining of infected Tupaia hepatocytes. Tupaia hepatocytes were infected with purified HBV (100 ge/cell) as described in Materials and Methods and immune stained for newly produced HBsAg 12 days p.i. (A) (magnification, x200). Infection is inhibited by preincubation of virus inoculum with polyclonal anti-HBs antibody (B) (magnification x200). Increased viral input (10,000 ge/hepatocyte) resulted in infection of up to 20% of hepatocytes (C) (magnification, x200). HBsAg is distributed within the cytoplasm (D) (magnification, x630) with typical perinuclear accumulation (arrow). Backmatter: PMID- 12915562 TI - The Mason-Pfizer Monkey Virus PPPY and PSAP Motifs Both Contribute to Virus Release AB - Late (L) domains are required for the efficient release of several groups of enveloped viruses. Three amino acid motifs have been shown to provide L-domain function, namely, PPXY, PT/SAP, or YPDL. The retrovirus Mason-Pfizer monkey virus (MPMV) carries closely spaced PPPY and PSAP motifs. Mutation of the PPPY motif results in a complete loss of virus release. Here, we show that the PSAP motif acts as an additional L domain and promotes the efficient release of MPMV but requires an intact PPPY motif to perform its function. Examination of HeLaP4 cells expressing PSAP mutant virus by electron microscopy revealed mostly late budding structures and chains of viruses accumulating at the cell surface with little free virus. In the case of the PPPY mutant virus, budding appeared to be mostly arrested at an earlier stage before induction of membrane curvature. The cellular protein TSG101, which interacts with the human immunodeficiency virus type 1 (HIV-1) PTAP L domain, was packaged into MPMV in a PSAP-dependent manner. Since TSG101 is crucial for HIV-1 release, this result suggests that the Gag-TSG101 interaction is responsible for the virus release function of the MPMV PSAP motif. Nedd4, which has been shown to interact with viral PPPY motifs, was also detected in MPMV particles, albeit at much lower levels. Consistent with a role of VPS4A in the budding of both PPPY and PTAP motif-containing viruses, the overexpression of ATPase-defective GFP-VPS4A fusion proteins blocked both wild-type and PSAP mutant virus release. Keywords: Introduction : Retrovirus assembly and release are driven by the viral Gag polyprotein, which contains all determinants necessary for the formation of virus-like particles . Within Gag, an interaction domain mediates Gag-Gag association, while a membrane binding domain directs Gag to the plasma membrane. Based on the sequence of Gag assembly and membrane association, two major pathways of virus assembly can be distinguished: betaretroviruses, such as Mason-Pfizer monkey virus (MPMV) and mouse mammary tumor virus (MMTV), assemble immature procapsids in the cytoplasm (B/D-type assembly). These procapsids are subsequently transported to the plasma membrane. Other retroviruses, e.g., the lentivirus human immunodeficiency virus (HIV), form electron-dense assembly complexes after targeting Gag to the plasma membrane (C-type assembly). Irrespective of the mode of Gag assembly, however, virus bud formation is generally initiated at the plasma membrane. The efficient separation of the bud from the cell membrane has been shown to be an active process which requires so-called late (L) domains . During or after budding, the virus undergoes maturation, i.e., the viral protease (PR) cleaves Gag into the mature matrix (MA), capsid (CA), and nucleocapsid (NC) proteins as well as other products, depending on the virus. This process leads to the condensation of the viral core and is essential for virus infectivity. To date, three different sequence motifs have been shown to provide L-domain function: PT/SAP, PPXY, and YPDL. While L domains have first been described in retroviral Gag proteins , they are also present in the M and matrix proteins of rhabdoviruses and filoviruses, respectively . Most retroviruses except lentiviruses contain a PPXY motif, in some cases together with a PT/SAP motif. In contrast, most lentiviruses harbor only the PT/SAP motif. The lentivirus equine infectious anemia virus is the only virus that has been shown to use the YPDL motif as an L domain . The L-domain sequence is generally very conserved, and even subtle mutations result in the retention of virus buds at the plasma membrane (see, e.g., reference ). The budding structures arrested at the cell surface stay immature in morphology. Host proteins have been identified that directly bind to the different L-domain motifs, and it is believed that the recruitment of these proteins to the site of virus budding ultimately results in the membrane fission event required for virus detachment. The exact set of factors needed, however, and the mechanism of their action are not understood. TSG101, which functions in the vacuolar protein sorting (Vps) pathway , binds to the PT/SAP motif of HIV-1 and HIV-2 . Similarly to mutations in the PT/SAP motif, depletion of TSG101 from virus-producing cells causes the accumulation of virus buds tethered to the cell surface . PPXY motifs are potential binding sites for WW domains , and the WW domain containing Nedd-4 family E3 ubiquitin ligases has been shown to bind to retroviral PPXY L domains . A function of ubiquitin ligases in retrovirus budding is supported by the observations that a small percentage of Gag proteins within purified HIV and murine leukemia virus (MLV) is monoubiquitinated and that large amounts of free ubiquitin are present inside retroviruses . In addition, the intracellular depletion of free ubiquitin by treatment with proteasome inhibitors has been shown to block release of HIV-1, HIV-2, and Rous sarcoma virus (RSV) at a very late stage . In the case of RSV, this block could be partially overcome by fusing ubiquitin to the C terminus of Gag . It is presently not known if the machinery needed to separate viral and cellular membranes is identical or at least overlapping for viruses containing different L domains. The observation that the overexpression of ATPase-defective VPS4A proteins blocked release not only of HIV-1, carrying a PTAP motif, but also of MLV , which functionally depends on a PPPY motif , suggested that these two pathways may utilize the same downstream effectors. Like TSG101, VPS4 is a component of the cellular Vps pathway, normally involved in formation of multivesicular bodies (MVB) . A potential role of Nedd-4-like ubiquitin ligases in the MVB pathway and whether the PPXY and PT/SAP motifs act in a redundant or synergistic way in those viruses where both are found are unclear. Different L-domain motifs have been shown to be at least partially exchangeable between retroviruses . Some viruses (e.g., HIV-1) carry only one known L-domain motif, while several retroviruses (e.g., MPMV and human T-cell leukemia virus type 1 [HTLV-1]) as well as Ebola virus and vesicular stomatitis virus contain closely spaced or overlapping PPPY and PSAP motifs. In the case of MPMV, PPPY and PSAP motifs are found in the pp24/16 domain between the MA and CA regions of Gag. Yasuda et al. showed that the PPPY motif is essential for virus release . Thus, the MPMV PSAP motif alone is not sufficient to mediate virus release. Here, we report that the MPMV PSAP motif acts as an additional L domain and promotes MPMV release in a PPPY motif-dependent manner. Viruses that contain a mutated PSAP motif were defective at a late stage of release, since late budding structures and chained procapsids accumulated at the cell surface. Consistent with a function of TSG101 in MPMV release, we found endogenous TSG101 to be enriched in purified MPMV in a PSAP-dependent manner. MATERIALS AND METHODS : MPMV expression constructs. | The pp24/16 PPPY and PSAP motifs were mutated in the context of pSHRM15 (provided by E. Hunter ). All mutations were introduced by standard PCR mutagenesis, and amplified fragments were sequenced. A new MPMV expression plasmid (pMPMV) was constructed to achieve high levels of virus production after transient transfection. To this end, the MPMV proviral sequence (wild-type [wt] or mutant) was excised from the respective pSHRM15 plasmids and cloned into the EcoRI site of pcDNA3.1/Zeo(-) (Invitrogen). The region between the pcDNA3.1/Zeo(-) and MPMV transcription start sites was deleted by PCR. Detailed cloning procedures are available on request. The resulting constructs carrying the PPPY, PSAP, and double mutations were designated pMPMV/PY(-), pMPMV/PP(-), and pMPMV/2x(-), respectively. pSHRM15 derivatives were termed analogously. Cell lines and transfections. | HeLaP4 and 293T cells were maintained in Dulbecco's modified Eagle medium (DMEM) containing 10% fetal calf serum and antibiotics. For labeling experiments and analysis by electron microscopy (EM), HeLaP4 cells were transfected with FuGENE 6 (Roche) as instructed by the manufacturer. All other transfections were carried out by the calcium phosphate precipitation method. Metabolic labeling. | For pulse-chase experiments, 4 x 106 HeLaP4 cells were transfected with 20 mug of the indicated plasmid. Twenty-four hours after transfection, cells were detached by using phosphate-buffered saline (PBS) containing 15 mM EDTA, washed, and starved for 20 min in methionine-free medium (ICN) containing 2% fetal calf serum. Then, cells were pulse-labeled for 30 min with [35S]methionine (2 mCi/ml; SJ-5050, Amersham). The pulse period was ended by adjusting the medium to 1 mM unlabeled methionine (Sigma), followed by removal of the labeling medium. Aliquots were chased in DMEM containing 0.4 mM methionine for the indicated period of time. For overnight labeling, 4 x 105 cells were transfected with 4 mug of plasmid. Thirty hours posttransfection, cells were rinsed with PBS and incubated with [35S]methionine (0.1 mCi/ml, diluted as above; SJ-5050, Amersham) for 12 h. Subsequent to pulse-chase or overnight labeling, cells were washed with PBS and lysed in RIPA buffer (50 mM Tris [pH 8.0], 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS]). Cell lysates were cleared by centrifugation (13,000 x g for 15 min) and preblocked by using a mixture of three preimmune sera and protein A Sepharose (Amersham) (1 h at 4C). In the case of the pulse-chase experiments, virus-containing media were directly lysed by addition of 5x RIPA buffer. After overnight labeling, media were passaged through 450-nm-pore-size filters and virus was sedimented through sucrose cushions (20% [wt/vol] sucrose in PBS) by ultracentrifugation (90 min at 192,000 x g) and lysed in RIPA buffer. All samples were immunoprecipitated by using anti-MPMV CA antiserum. The immunoprecipitates were resolved by SDS-polyacrylamide gel electrophoresis (PAGE). Gels were fixed in 30% methanol-10% acetic acid and dried. Labeled viral proteins were quantified by using a Bio-Rad Personal FX Phosphorimager and Bio-Rad Quantity One software. Virus purification. | Virus was prepared from 293T cells transiently transfected with pMPMV or pMPMV/PP(-). Twelve hours after transfection, cells were rinsed with PBS, and the medium was replaced. Virus-containing medium was harvested 36 h after transfection and filtered through a 450-nm-pore-size filter. Virus was pelleted through sucrose cushions by ultracentrifugation (60 min at 53,000 x g), resuspended in 0.8 ml of PBS, and loaded onto a 6 to 18% Optiprep gradient as described previously . The visible virus band was isolated and diluted with PBS. Virus was recovered by ultracentrifugation (30 min at 164,000 x g). Purity of the preparation was monitored by silver staining and evaluated based on the loss of albumin. For external digestion of viral particles with trypsin, sucrose pellets were resuspended in 0.1 mg of trypsin/ml (ICN) in PBS and incubated at 30C for 30 min. When virus was lysed with 0.1% Triton X-100 before trypsinization, a complete loss of the viral CA band was observed by Western blotting. Western blotting and antibodies. | Western blotting was performed by using standard procedures. The following primary antibodies were used: rabbit anti-MPMV CA (provided by T. Ruml) (1:1,000); mouse anti-TSG101-4A10 from GeneTex (1:100); rabbit anti-Nedd-4 from Becton Dickinson Biosciences (1:5,000); rabbit anti-green fluorescent protein (GFP) that was raised against bacterially expressed GFP (1:1,000); and rabbit anti-14-3-3gamma from Santa Cruz (1:1,000). EM. | HeLaP4 cells (1 x 106) were transfected with 8 mug of wt or mutant pMPMV. At 48 h after transfection, cells were gently rinsed with PBS and fixed with ice cold 2.5% glutaraldehyde in 0.05 M sodium cacodylate (pH 7.2) for 20 min on ice. Fixed cells were scraped from the plates, collected by low-speed centrifugation (200 x g for 10 min at 4C), successively stained with 2% osmium tretroxide and 0.5% uranyl acetate, and processed for ultrathin sectioning. Micrographs were taken with a Zeiss EM-10 electron microscope at 80 kV. The magnification indicator was routinely controlled by the use of a grating replica. RESULTS : The PPPY and PSAP motifs both contribute to MPMV release. | To facilitate the biochemical analysis of MPMV release, a novel MPMV expression plasmid was constructed. To this end, the MPMV transcription start site was placed under the control of a CMV promoter such that the resulting transcripts are identical to those derived from the LTR promoter. The new plasmid (pMPMV) and the parental construct (pSHRM15) were transfected into HeLaP4 cells, and cell extracts and virus released into the growth medium were analyzed by Western blotting 24 h after transfection. The same pattern of Gag and CA proteins was detected in both cases . Transfection of pMPMV, however, reproducibly yielded significantly higher levels of Gag protein expression and virus release than transfection of pSHRM15 (Fig. , compare lanes 1 and 3 to lanes 2 and 4). In agreement with previously published results, intracellular processing of Gag was observed for both constructs . The extent of intracellular processing did not correlate with the relative level of Gag expression. Within its pp24/16 domain, MPMV Gag carries a PSAP motif four amino acids downstream of a PPPY motif . To evaluate the relative contribution of the two motifs to MPMV release, we individually mutated the PPPY motif to PGAA and mutated the PSAP motif to AGAP. The mutations were introduced into plasmids pMPMV and pSHRM15. In addition, we constructed plasmids containing both mutations . The amino acid substitutions were chosen based on published data and should result in complete inactivation of the respective motif. To determine the efficiency of wt and mutant virus release, HeLaP4 cells were transfected with the indicated plasmids and, 36 h after transfection, metabolically labeled with [35S]methionine for 12 h. After the labeling period, cells were lysed. Virus was collected by centrifugation through sucrose cushions. Cell extracts and virus samples were subjected to immunoprecipitation with an antiserum against the MPMV CA protein followed by SDS-PAGE and phosphorimage analysis. The result of a representative experiment is shown in Fig. . Transfection of wt pMPMV yielded the typical pattern of cell-associated Gag and CA proteins and caused the release of virus into the growth medium (Fig. , lane 1). It should be noted that all virus samples contained only completely processed CA and no precursor or intermediate processing products, independent of the processing defect observed in transfected cells (data not shown). Mutation of the PPPY motif resulted in a pronounced impairment of intracellular Gag processing and a defect in particle release (Fig. , lane 2). Gag and a protein with slightly lower mobility than Gag (see below) accumulated in PY(-)-transfected cells, while the amount of processed CA protein was reduced. Analysis of the virus fraction (Fig. , lower panel) revealed that the PPPY mutant displayed a strong defect in virus release. This result is in agreement with the report by Yasuda et al., who also observed a complete loss of virus release as well as a defect in Gag processing for a PPPY mutant . Analysis of PSAP mutant-transfected cells showed a slight impairment of processing and a significant reduction in virus release (Fig. , lane 3). Thus, the PSAP motif cannot substitute for the PPPY motif, but it acts as an additional L domain and contributes to the efficient release of MPMV. The phenotype of PSAP inactivation was less pronounced than that of the PPPY mutation. Consistent with the absolute dependence of MPMV release on the intact PPPY motif, the PPPY and PSAP motif double mutant behaved like the PPPY mutant (Fig. , lane 4). It has been suggested that the magnitude of the defect in virus release caused by L-domain mutations may correlate with the level of Gag expression . The described experiments were performed with a CMV-driven high-level expression plasmid transfected by lipofection. To exclude the possibility that the intermediate phenotype of the PSAP mutant was a result of high levels of expression, we repeated the experiment by using transfection of pSHRM15 into HeLaP4 cells by calcium phosphate precipitation (Fig. , lanes 1 to 3). As observed for transfection of the pMPMV plasmid, the PPPY mutant Gag protein remained largely unprocessed and exhibited a block to virus release (Fig. , lane 2). Intracellular processing of Gag and virus release were also significantly reduced in the case of the PSAP mutant but were less impaired than for the PPPY mutant (Fig. , compare lanes 1 to 3). Accordingly, the intermediate phenotype of the PSAP mutant is not due to overexpression. To quantify the relative efficiencies of wt and mutant virus release from pMPMV-transfected cells, the virus-associated CA signal was related to the sum of the signals of the intracellular and virus-associated CA-containing proteins. Only the three major CA-containing bands (namely, Gag, the protein migrating slightly above Gag, and processed CA) detected in cell extracts were quantified. The obtained numbers were normalized for the methionine content of the respective proteins. The protein migrating above Gag was assumed to carry the same number of methionines as Gag. These quantifications indicated that virus release is reduced by approximately fivefold in the case of the PPPY and double mutants, while PSAP inactivation resulted in a reduction of virus release by roughly 30% . Rescue of L-domain mutant Gag proteins. | It has been reported that RSV L-domain mutant Gag protein can be rescued into virus-like particles by coexpression of Gag protein carrying an intact L domain . These results suggest that only a minor fraction of Gag molecules of a budding virion needs to carry an L domain. To test the release efficiency of mixed populations of MPMV Gag proteins, we cotransfected equal amounts of mutant and wt plasmids and analyzed and quantified CA-reactive proteins in cell and virus extracts 24 h post transfection (Fig. , lanes 5 to 7 and Fig. ). Cotransfection of the wt plasmid with either PPPY or double mutant pMPMV resulted in a 40% reduction of virus release compared to wt virus (Fig. , compare lanes 1, 5, and 7; Fig. ), indicating that only a partial rescue occurred. A similar analysis for the PSAP motif mutant was not conclusive, because differences in virus release were too small (Fig. , compare lanes 1 and 6; Fig. ). The experiments shown in the previous section suggested that both motifs contribute to MPMV release. To test whether the PPPY and PSAP motifs can complement each other in trans or need to be present on the same polyprotein, we performed cotransfection experiments. Transfecting equal amounts of pMPMV plasmids carrying either the PPPY or the PSAP mutation yielded higher levels of virus release than transfection of the same amount of either variant alone (Fig. compare lanes 2, 3 and 8; Fig. ). A partial rescue of virus release was also observed by Western blot analysis of virus produced from HeLaP4 cells cotransfected with PPPY and PSAP mutant pSHRM15 . Thus, both motifs can complement each other in trans under low- and high-expression conditions. Gag lacking the PPPY motif trans-dominant-negatively inhibits wt virus release. | Since virus release is not completely rescued by expressing equal amounts of PPPY mutant and wt Gag polyproteins , we determined whether this mutant has a dominant-negative influence on wt virus release. wt and PPPY mutant pSHRM15 derivatives were cotransfected at a ratio of 1:1 and 1:2. In all cases, the total amount of transfected DNA was adjusted to 15 mug by using empty vector. As noted above , cotransfection of equal amounts of PPPY mutant and wt proviral plasmids (5 mug each) did not yield more virus production than transfection of 5 mug of wt plasmid alone (Fig. , compare lanes 2 and 5). When wt and mutant plasmids were cotransfected at a ratio of 1:2, virus release was reduced to nearly undetectable levels, even though higher levels of Gag protein were observed in cell extracts. Cotransfection of wt and PPPY mutant plasmids also did not restore intracellular Gag processing, since only Gag and the protein slightly above Gag were detected in cell lysates (Fig. , lanes 5 and 6). The dominant-negative effect of the PPPY mutant Gag protein was also evident when high levels of expression were used (Fig. , lanes 7 to 9) and is therefore independent of the Gag expression level. In addition, we confirmed our finding by quantitative Western blotting (data not shown). Taken together, these results suggest that a molar excess of PPPY mutant Gag polyproteins over wt Gag has a dominant-negative influence on virus production. We also tested whether cotransfection of the MPMV PSAP mutant or of a HIV-1 L-domain mutant with the respective wt plasmid inhibited virus release. In both cases, no dominant-negative effect on virus release was observed (data not shown). Analysis of wt and mutant Gag processing kinetics. | As shown above, intracellular processing of the PPPY and double-mutant Gag polyproteins was significantly impaired . To analyze the processing kinetics of wt and mutant Gag proteins and to determine whether the protein migrating slightly above Gag represents a posttranslational modification of Gag, we performed pulse-chase experiments. For this, we transfected HeLaP4 cells with wt or mutant pMPMV constructs. Twenty-four hours after transfection, cells were pulse-labeled with [35S]methionine for 30 min and chased for the times indicated. Cell and virus extracts were immunoprecipitated with anti-CA antibodies. wt Gag processing and release kinetics are shown in Fig. . Directly after the pulse, the Gag polyprotein was the predominant band detected (lane 1), and processing to CA became clearly detectable after a 1-h chase (lane 3). Most of the intracellular Gag was processed to CA after 8 h. Release of virus particles containing processed CA was detected after 1 h and increased over the entire chase period. A comparison of wt and mutant Gag processing kinetics is shown in Fig. . Similar amounts of Gag polyproteins were observed in all cases after the pulse period, and the protein migrating above Gag was not detected in these samples (Fig. , lanes 1 to 4). Processing kinetics of the PSAP variant were largely similar to the wt polyprotein (Fig. , compare lanes 5 and 7 and lanes 9 and 11). A significant processing defect was observed for the PY(-) variant and for the double mutant, with both exhibiting reduced levels of CA after 2-h and 8-h chase periods (lanes 6 and 8 and lanes 10 and 12). Furthermore, the protein migrating slightly slower than Gag (marked by an arrow in lanes 6, 8, 10, and 12) became detectable in cells transfected with the PPPY or double mutant after a chase period of 2 h and remained present after 8 h. Therefore, this product corresponds to a posttranslationally modified version of Gag. We hypothesize that it represents Gag lacking the proline-rich C-terminal p4 domain . Cleavage of p4 by the viral PR may result in altered migration of the protein in SDS-PAGE. CA-containing cleavage products migrating slower than their precursors have been described for other retroviruses . Alternatively, the product migrating just above Gag could result from a posttranslational modification of Gag other than processing. EM analysis of virus assembly and release. | Deletion of either the PPPY or the PSAP motif impaired MPMV release, but the effect of the PSAP mutation was much less pronounced, and virus production was still observed in this case . Analysis of particle formation by EM is a more direct way to determine virus release at the morphological level. We therefore performed thin-section EM analysis of HeLaP4 cells transfected with wt, PPPY mutant, PSAP mutant, or double-mutant pMPMV. In all cases, the typical sites of intracellular assembly of immature procapsids were observed (shown representatively for cells transfected with wt pMPMV in Fig. ). Furthermore, thin sections of cells transfected with the wt plasmid contained many procapsids budding from the plasma membrane as well as numerous mature and immature virions . In contrast, no mature virions were detected in thin sections of cells transfected with the PPPY mutant MPMV . Instead, many immature procapsids accumulated underneath the plasma membrane and appeared to be arrested at an early stage of bud formation . Typical late budding structures connected to the plasma membrane by a thin membrane stalk as well as chained budding structures containing more than one procapsid were also present, indicating that the early budding arrest was overcome by some procapsids . Interestingly, in several cells, procapsids were found to decorate intracellular vesicles . Again, the majority of these procapsids failed to initiate bud formation, and no bulging of the membrane was observed. Thin-section EM analysis of cells transfected with PSAP mutant pMPMV revealed a different phenotype. In this case, we observed predominantly late immature budding structures, many of which were tethered to each other . Some of these budding chains containing multiple procapsids were apparently released from the cell (Fig. , upper right). Less extracellular particles were observed than in the case of wt pMPMV, but some mature virions containing a condensed core were detected (e.g., Fig. , middle). For a more quantitative determination of the late defect of the PSAP variant, the relative abundance of different budding structures and free mature and immature particles was counted for wt and PSAP mutant pMPMV-transfected cells . This analysis confirmed that much less mature virus was observed in the case of the PSAP variant, and the amount of total cell-free virus detected accounted for only 16% of all procapsids, compared to 60% for the wt constructs. On the other hand, the relative abundance of budding structures and virus chains was increased in the case of the PSAP variant. Only 7% of wt procapsids were found in virus chains attached to the plasma membrane, while 26% of PSAP mutant procapsids were found in such chains. Nearly one-half of the chains containing multiple procapsids were apparently released from the cell, i.e., no connection to the plasma membrane was visible. The relative abundance of immature procapsids was also significantly higher for the PSAP variant, indicating that maturation is impaired in this case. Taken together, these results showed that the PSAP mutant is indeed defective at the stage of virus release. Interestingly, the budding arrest in the case of the PSAP variant appeared to be at a later stage than in the PPPY variant, with more advanced budding structures extending further away from the plane of the plasma membrane. As in the biochemical experiments described above, the double mutant was indistinguishable from the PPPY mutant at the morphological level as well. TSG101 and Nedd-4 are incorporated into MPMV particles. | The interaction of the HIV-1 PT/SAP motif with TSG101 has been shown to be essential for HIV-1 release , and the binding of Nedd-4-like proteins to the PPPY motif is presumed to be necessary for the function of this late motif. Accordingly, the cellular binding partners should be recruited to the budding site and may be incorporated into virions. The overexpression of a segment of TSG101 by transfection led to incorporation of this protein into HIV-1 particles . To test whether endogenous cellular TSG101 or Nedd-4 are incorporated into MPMV virions, we analyzed highly purified particle preparations. To this end, 293T cells were transfected with either wt or a PSAP mutant pMPMV. Media were gathered 36 h after transfection and filtered, and virus was centrifuged through cushions of 20% sucrose. Virus samples were further purified by velocity gradient centrifugation as described previously . The visible virus band was collected and diluted, and virus was recovered by ultracentrifugation. Purity of samples was evaluated by silver staining. Judging from the silver-stained gel, the general protein content and the degree of processing were very similar for wt and PSAP mutant virus . The viral Gag-derived proteins represented the major constituents of the virion. To examine the presence of TSG101 in purified MPMV, similar amounts of wt and PSAP mutant virus (the same ratio used in Fig. , lanes 1 and 3) were probed for incorporation of endogenous TSG101 by Western blotting. TSG101 was readily detected in wt virus preparations, while the amount of TSG101 was significantly reduced in the PSAP mutant virus preparation (Fig. , lanes 3 and 4), and no Tsg101 was detected in parallel preparations from the medium of mock-transfected cells (Fig. , lanes 3 and 4). The data in Fig. slightly exaggerate the amount of TSG101 present in PSAP mutant virus preparations, because more PSAP mutant virus than wt virus was loaded (compare Fig. , lanes 1 and 3 or Fig. , lanes 3 and 4). Quite clearly, however, TSG101 was present in PSAP mutant virus preparations, albeit at a reduced level. To determine whether other cytosolic proteins were nonspecifically incorporated into MPMV particles, we probed for the protein 14-3-3gamma, which localizes mostly to the cytosol (data not shown). While the 14-3-3 protein was highly abundant in cells, it was completely absent from virus preparations . Thus, TSG101 is clearly enriched in particle preparations compared to cell extracts, while other cytosolic proteins are largely excluded. The incorporation of TSG101 into MPMV was confirmed by analysis of virions that had been treated with trypsin, which degrades proteins outside of the virion, but did not reduce the amount of TSG101 in the preparation (data not shown). The PPPY motif binding protein Nedd-4 was readily detected in cell lysates (Fig. , lanes 1 and 2), but a protein migrating at the position of intact Nedd-4 was detected only in small amounts in wt and PSAP mutant virus preparations (Fig. , lanes 3 and 4). The stronger signal in lane 4 than in lane 3 probably reflects the higher amount of virus loaded for the PSAP mutant (compare Fig. , lanes 1 and 3). Besides intact Nedd-4, several other reactive proteins were detected with Nedd-4 antiserum in virus preparations. Some of these proteins may correspond to cross-reaction with (overloaded) viral proteins or have also been detected in cell lysates of untransfected cells (not shown), but this does not account for the protein with an apparent molecular mass of 50 kDa (Fig. , arrow), which was not present in parallel preparations from mock-transfected cells (Fig. , arrow). The presence of Nedd-4 and the smaller immunoreactive products inside MPMV virions was also confirmed by analysis of trypsin-treated virus (data not shown). VPS4A is required for wt and PSAP mutant MPMV release. | The cellular AAA type ATPase VPS4A, which is a component of the Vps pathway involved in MVB sorting , has been shown to be required for the release of both HIV-1 and MLV. We therefore wanted to test whether VPS4A is also required for MPMV budding and whether the residual budding observed for the MPMV PSAP mutant depends on VPS4A function as well. For this, we cotransfected 293T cells with wt or PSAP mutant pMPMV and with either empty vector, wt GFP-VPS4A, or dominant-negative GFP-VPS4A mutants, which are defective for ATP binding (VPS4AK173Q) or ATP hydrolysis (VPS4AE228Q) . As a control, we also transfected PPPY mutant pMPMV. As observed for HeLa cells, 293T cells transfected with PPPY mutant pMPMV exhibited no virus release (Fig. , lanes 1 and 2). In contrast, there was no significant phenotype for PSAP mutant MPMV in 293T cells (Fig. , lanes 1 and 3), indicating that the requirement for the PSAP motif depends on the cell type used. Upon cotransfection of wt GFP-VPS4A with pMPMV or the L-domain mutant plasmids, less virus was released than from cells cotransfected with empty vector (Fig. , compare lanes 4 to 6 and lanes 1 to 3). A similar finding was reported by Garrus et al., who observed a reduction of virus release upon cotransfection of wt GFP-VPS4A and HIV-1 . Strikingly, cotransfection of either VPS4A mutant construct reduced virus release for both wt pMPMV and the PSAP variant to levels as low as that observed for the PPPY mutant construct (Fig. , lanes 7 to 14). Furthermore, in the presence of ATPase-defective VPS4A variants, wt and PSAP mutant Gag exhibited a defect in processing similar to that of PPPY mutant Gag proteins. This result indicates that VPS4A is essential for MPMV release and acts downstream of both the PSAP and the PPPY motif. DISCUSSION : Here, we present a detailed analysis of the relative contributions of the closely spaced MPMV PPPY and PSAP motifs to MPMV particle production. In agreement with previous reports , a block to viral release was observed for MPMV lacking the PPPY motif. Procapsids accumulated underneath the plasma membrane or on intracellular vesicles, displaying an apparent defect mostly in the early stages of budding. Surprisingly, release of particles from cells cotransfected with equal amounts of wt and PPPY mutant constructs was inefficient as well, and a twofold excess of PPPY mutant Gag protein over wt Gag completely abolished virus release in a trans-dominant-negative way. Thus, at least one-half of the MPMV Gag molecules needs to carry an intact PPPY motif to allow efficient virus release. This is in contrast to findings which showed that one-fifth of HIV-1 Gag molecules carrying a PTAP L domain were sufficient to allow efficient particle release . Conceivably, there is a certain threshold for factor recruitment which may be different depending on the virus and/or the host cell. While the MPMV PSAP motif cannot replace a functional PPPY motif (; this study), our results clearly showed that the PSAP motif functions as a second L domain and promotes virus release in the context of an intact PPPY motif. The phenotype of the PSAP mutation was less pronounced than that of the PPPY mutation, since virus release was reduced by only a factor of two when PSAP was changed. The release defect was much more evident on thin-section EM analysis, however, which revealed a late budding arrest, with most procapsids in immature budding structures or budding chains containing multiple procapsids. These budding chains were apparently released from transfected cells, thus contributing to the extracellular virus detected. The phenotype of a construct carrying both the PPPY and PSAP mutations was indistinguishable from that of the PPPY mutant. Thus, in the context of MPMV Gag, PPPY and PSAP motifs are not redundant. This is different from results observed in substitution experiments, where the PPPY L domains of RSV and MLV were artificially exchanged for a HIV-1-derived PTAP motif. In these studies, the first 12 or 18 residues of HIV-1 p6 were used to replace the respective PPPY L-domain peptide, thereby reconstituting particle release . Like MPMV, HTLV-1, vesicular stomatitis virus, and Ebola virus contain closely spaced or overlapping L-domain sequences. The Ebola virus VP40 protein harbors the peptide P7TAPPEY13. Martin-Serrano et al. showed that substitution of P7, which disrupts the PTAP but not the PPXY motif, severely reduced the release of VP40 . This result indicated that the Ebola virus PTAP motif contributes to virus release as well, similarly to the result observed for MPMV in this study. One may speculate that the combined function of two different L-domain motifs is also required for efficient release of the other viruses that contain more than one L domain. Furthermore, the Ebola virus VP40 L-domain peptide has been shown to functionally replace the HIV-1 PSAP motif of the HIV-1 Gag polyprotein . Like in the context of VP40, the P7 mutation, which leaves the PPPY motif intact, resulted in a complete loss of HIV-like particle release . Similarly, Strack et al. replaced the entire p6 domain of HIV-1 Gag with a VP40-derived peptide . Consistent with the observation by Martin-Serrano et al., substituting Y13 induced no phenotype. When the same Ebola virus-derived peptide was used in the context of a minimal Gag protein, which lacked the globular domain of MA and the N-terminal part of CA and in which the NC-sp2 region was replaced by a leucine zipper, the PT/SAP motif did not suffice to promote particle release. Instead, as was observed in this study for MPMV, a PPPY motif was required to achieve release of particles. Strack et al. suggested that in the context of HIV-1, the PT/SAP motif cooperates with NC-p1 or the remainder of p6 . Thus, it seems likely that whether or not a PT/SAP motif can function as an L domain depends on the context of the Gag protein. The EM results obtained in this study suggest that the defects of the PPPY and the PSAP variants of MPMV are both at the stage of virus release but are not identical. Most PPPY variant budding structures appeared to be much closer to the plane of the plasma membrane than was observed for the PSAP variant, indicating an earlier budding defect for the PPPY variant. Some late budding structures connected to the plasma membrane or to each other by only a thin membrane stalk were observed for the PPPY variant as well, but this was not common. In most cases, procapsids completely failed to initiate budding and did not induce membrane curvature at all. Le Blanc et al. reported a similar finding for HTLV-1, indicating that the intact PPPY motif may be required at a relatively early stage for the initiation of budding . If the PPPY motif acts upstream of the PSAP motif, this could explain why an intact PPPY motif is needed for the action of PSAP as an L domain (see also below). In several cells, we observed intracellular vesicles decorated with PPPY mutant viral procapsids that did not initiate budding. Conceivably, some wt MPMV procapsids may also be transported to intracellular vesicles, but this may be normally overlooked, because it constitutes a minor pathway which becomes visible only under conditions of budding arrest. It is interesting in this regard that budding of HIV-1 into intracellular major histocompatibility complex class II vesicles has recently been reported, and this may be an important route of release at least in some cell types . Alternatively, these decorated vesicles may be the result of a partial defect in viral assembly even earlier than budding, i.e., during procapsid trafficking, or may be due to the expression of large amounts of viral Gag proteins. TSG101 has been shown to bind to the HIV-1 PTAP L domain, and this interaction is essential for virus release . We find readily detectable amounts of endogenous TSG101 present within MPMV particles, suggesting that the protein is recruited to the budding site and incorporated via its interaction with the PSAP motif. This incorporation is likely to be specific, because Tsg101 was not detected in preparations from mock-transfected cells, there was no incorporation of a highly expressed cytosolic protein into particles, and the virion levels of TSG101 were strongly reduced in the case of the PSAP variant. The finding that TSG101 incorporation is not abolished in particles lacking the PSAP motif raises the question whether TSG101 packaging is also specific in this case. Furthermore, it is unclear if the amount of TSG101 still detected accounts for the intermediate phenotype caused by PSAP mutation. Based on published data, TSG101 should not be able to bind to the inactivated PSAP motif . Since TSG101 shows weak binding to ubiquitin alone , it could be recruited by binding to ubiquitinated Gag protein. The fact that an L-domain phenotype was observed, even though some TSG101 was present in virus preparations, may indicate that a relatively high level of TSG101 is required at the budding site. Besides TSG101, we also detected the E3 ubiquitin ligase Nedd-4 in MPMV particles, albeit at low levels. Nedd-4 incorporation was unaltered by mutation of the PSAP motif, as expected. Recently, a Nedd-4-like protein named Bul-1 has been reported to interact with MPMV Gag in a PPPY motif-dependent way, and Bul-1 overexpression was shown to stimulate MPMV release . Accordingly, Nedd-4 itself may be only weakly recruited by the MPMV Gag polyprotein, explaining its low concentration inside viral particles. It should be noted, however, that the major Nedd-4 immunoreactive protein in viral preparations migrated with an apparent mobility of ca. 50 kDa, which suggests that significantly more Nedd-4 may be incorporated into viral particles and cleaved by the viral PR. The 50-kDa protein was not detectable in cytoplasmic extracts, and the signal could not be assigned to any of the major viral proteins. It seems unlikely that a cellular protein is incorporated into virions at levels high enough to be detectable by unspecific binding of the antibody, and the possibility of PR-mediated cleavage of Nedd-4 is currently under investigation. As observed for the release of HIV-1 and MLV , MPMV particle production was completely abolished by overexpression of dominant-negative variants of the cellular ATPase VPS4A as well. Virus release was also blocked in the case of the PSAP variant of MPMV, indicating that VPS4A promotes virus release independently of the PT/SAP motif and that the two L-domain motifs target the same pathway. This supports and extends previous findings for MLV, which also contains an essential PPPY motif but carries an additional PSAP motif in its MA domain, which may also contribute to virus release. In summary, our results indicate that both L-domain motifs of MPMV contribute to virus release, with the PPPY motif being essential and a PPPY mutation causing budding arrest at an earlier stage than a PSAP mutation. In the case of the VPS machinery-dependent sorting of endosomal proteins into MVBs, initial monoubiquitination of cargo and adapter proteins is followed by recruitment of the TSG101-containing ESCRT-1 complex and further complexes of the Vps pathway . By analogy, one may speculate that PPPY-dependent recruitment of Nedd-4 (or a Nedd-4 like protein) serves to monoubiquitinate Gag or a Gag-associated protein with subsequent binding of TSG101, both through the PSAP motif within Gag and through the ubiquitin moiety. Since TSG101 was incorporated into virions even in the absence of a PSAP motif, this motif in the case of MPMV may mainly serve to recruit sufficient amounts of TSG101 for efficient virus release, e.g., in cell types where TSG101 concentrations are limiting. This hypothesis predicts that TSG101 would also be required for the release of an MPMV variant lacking the PSAP motif, and this is under investigation. In the case of MLV, however, virus release was not severely altered following depletion of cellular TSG101 , indicating that the PPPY motif can function independently of TSG101 at least in this case. FIG. 1. : Comparison of MPMV expression from pMPMV and pSHRM15. Comparison of MPMV expression from pMPMV and pSHRM15. HeLaP4 cells were transfected with equal amounts of pMPMV (lanes 1 and 3) or pSHRM15 (lanes 2 and 4), and levels of intracellular (lanes 1 and 2) and virus-associated CA-reactive proteins (lanes 3 and 4) were analyzed by Western blotting with anti-CA antiserum. Molecular mass standards in kilodaltons are shown on the right; MPMV Gag and CA are identified on the left. FIG. 2. : (a) Schematic representation of MPMV Gag, showing the L-domain sequence within pp24 and the mutations introduced. (a) Schematic representation of MPMV Gag, showing the L-domain sequence within pp24 and the mutations introduced. (b) wt and L-domain mutant MPMV release. Cells were transfected with the indicated pMPMV constructs and metabolically labeled with [35S]methionine for 12 h. Cell-associated (upper panel) and virus-associated (lower panel) CA protein was immunoprecipitated with anti-CA antiserum, resolved by SDS-PAGE, and subjected to phosphorimage analysis. Note that no precursor or intermediate processing product was observed in the virus samples. (c) Quantification of wt and mutant release efficiencies. CA-containing bands were quantified. The release efficiency of wt and mutant virus as well as that of mixed virus populations was calculated by dividing the signal from virus-associated CA protein by the sum of the signals from cell- and virus-associated CA-containing proteins. The obtained numbers were normalized for wt release efficiency. n = 4 for columns 1, 2, 3, and 8; n = 2 for all other columns. (d) Steady-state levels of wt and mutant virus release. Cells were cotransfected with the indicated pSHRM15 plasmids, and cell- and virus-associated viral proteins (upper and lower panel, respectively) were analyzed by Western blotting with an anti-CA antiserum. Again, no precursor or intermediate processing product was observed in the virus samples. FIG. 3. : Effect of PPPY mutant Gag protein on wt virus release. Effect of PPPY mutant Gag protein on wt virus release. HeLaP4 cells were cotransfected with the indicated pSHRM15 (lanes 1 to 6) or pMPMV (lanes 7 to 9) wt and PY(-) plasmids. In each case, the amount of total DNA was adjusted to 15 mug. Steady-state levels of cell- and virus-associated CA protein (upper and lower panel, respectively) were analyzed by Western blotting with anti-CA antiserum. FIG. 4. : Kinetic analysis of wt and mutant Gag processing. Kinetic analysis of wt and mutant Gag processing. (a) Kinetics of wt Gag processing and virus release. HeLaP4 cells were transfected with pMPMV and pulse-labeled with [35S]methionine for 30 min. Cells were chased in DMEM for the indicated time. Cell- and virus-associated CA-containing proteins (lanes 1 to 6 and 7 to 12, respectively) were immunoprecipitated with anti-CA antiserum, resolved by SDS-PAGE, and visualized by phosphorimage analysis. (b) wt and L-domain mutant processing kinetics. HeLaP4 cells were transfected with the indicated pMPMV constructs, labeled as described above, and chased for 0, 2, or 8 h. At the 2-h time point, a Gag-related band (arrow) that migrates slower than Gag was detected in cells expressing PPPY mutant and double mutant Gag. FIG. 5. : Thin-section EM analysis of HeLaP4 cells transfected with pMPMV (wt). Thin-section EM analysis of HeLaP4 cells transfected with pMPMV (wt). Bars, 500 nm (a) and 200 nm (b). FIG. 6. : Thin-section EM analysis of HeLaP4 cells transfected with PPPY mutant pMPMV. Thin-section EM analysis of HeLaP4 cells transfected with PPPY mutant pMPMV. Bars, 1 mum (a) and 100 nm (b). FIG. 7. : Thin-section EM analysis of HeLaP4 cells transfected with PSAP mutant pMPMV. Thin-section EM analysis of HeLaP4 cells transfected with PSAP mutant pMPMV. Bars, 500 nm (a) and 200 nm (b to d). FIG. 8. : Incorporation of cellular proteins into MPMV. Incorporation of cellular proteins into MPMV. (a) wt and PSAP mutant virus was purified by centrifugation through sucrose cushions followed by sedimentation in an Optiprep velocity gradient. wt (lanes 1 and 2) and PSAP mutant (lane 3) virus protein content was analyzed by silver staining. (b) Comparable amounts of lysates from wt and mutant virus-expressing cells (lanes 1 and 2) as well as highly purified virus (the same amounts as in panel a, lanes 1 and 3) were probed for the cellular proteins TSG101 (upper panel), 14-3-3gamma (middle panel), and Nedd-4 (lower panel) by Western blotting. In the lower panel, the major Nedd-4-related bands migrating at 120 and 110 kDa are indicated. The arrow marks a band reacting with Nedd-4 antiserum which does not comigrate with one of the major viral proteins detected in the Ponceau S stain (not shown). (c) To analyze whether TSG101 and Nedd-4 are released from 293T cells in the absence of MPMV expression, we probed sucrose pellets from culture media of cells transfected with pMPMV or EGFP expression vectors (lanes 3 and 4) for TSG101 (upper panel) and Nedd-4 (lower panel). The arrow marks the same product as in panel b. FIG. 9. : Effect of coexpression of wt or ATPase-defective GFP-VPS4A fusion proteins on MPMV release. Effect of coexpression of wt or ATPase-defective GFP-VPS4A fusion proteins on MPMV release. 293T cells were cotransfected with the indicated pMPMV plasmids. In each case, the amount of total DNA was adjusted to 4 mug. Steady-state levels of cell- and virus-associated CA protein (upper and lower panel, respectively) were analyzed by Western blotting with anti-CA antiserum, and the expression of the VPS4A constructs was confirmed by Western blotting with an anti-GFP antiserum (middle panel). TABLE 1 : Relative abundance of wt and PSAP mutant budding structures Backmatter: PMID- 12915530 TI - A Wild-Type Porcine Encephalomyocarditis Virus Containing a Short Poly(C) Tract Is Pathogenic to Mice, Pigs, and Cynomolgus Macaques AB - Previous studies using wild-type Encephalomyocarditis virus (EMCV) and Mengo virus, which have long poly(C) tracts (61 to 146 C's) at the 5' nontranslated region of the genome, and variants of these viruses genetically engineered to truncate or substitute the poly(C) tracts have produced conflicting data on the role of the poly(C) tract in the virulence of these viruses. Analysis of the nucleotide sequence of an EMCV strain isolated from an aborted swine fetus (EMCV 30/87) revealed that the virus had a poly(C) tract that was 7- to 10-fold shorter than the poly(C) tracts of other EMCV strains and 4-fold shorter than that of Mengo virus. Subsequently, we investigated the virulence and pathogenesis of this naturally occurring short-poly(C)-tract-containing virus in rodents, pigs, and nonhuman primates. Infection of C57BL/6 mice, pigs, and cynomolgus macaques resulted in similar EMCV 30/87 pathogenesis, with the heart and brain as the primary sites of infections in all three animals, but with different disease phenotypes. Sixteen percent of EMCV 30/87-infected pigs developed acute fatal cardiac failure, whereas the rest of the pigs were overtly asymptomatic for as long as 90 days postinfection (p.i.), despite extensive myocardial and central nervous system (CNS) pathological changes. In contrast, mice infected with >=4 PFU of EMCV 30/87 developed acute encephalitis that resulted in the death of all animals (n = 25) between days 2 and 7 p.i. EMCV 30/87-infected macaques remained overtly asymptomatic for 45 days, despite extensive myocardial and CNS pathological changes and viral persistence in more than 50% of the animals. The short poly(C) tract in EMCV 30/87 (CUC5UC8) was comparable to that of strain 2887A/91 (C10UCUC3UC10), another recent porcine isolate. Keywords: Introduction : Encephalomyocarditis virus (EMCV) is a picornavirus belonging to the Cardiovirus genus that infects many animal species including pigs , rodents , cattle , elephants , raccoons , marsupials , baboons, macaques, chimpanzees, and humans . Rats and mice are the natural hosts of the virus, but pigs are the most commonly and severely infected domestic animals . EMCV strains have been isolated from primates, pigs, and rodents . EMCV Rueckert (EMCV R/45), the prototype strain, was isolated from a 5-year-old chimpanzee that suffered acute fatal myocarditis in 1945 , whereas EMC-M/58 virus was isolated from a naturally infected domestic pig suffering from severe myocarditis . A number of EMC-M/58 variants, including EMC-D/58, EMC-B/58, PV2/58, and PV21/58, were generated in laboratories and used to study the pathogenesis of viral diabetes . Mengo virus, another cardiovirus that shares the same serotype as EMCV, was isolated from the cerebrospinal fluid of a paralyzed rhesus macaque in 1948 . EMCV 30/87 and 2887A/91 are more recent EMCV isolates obtained from aborted swine fetuses in the United States and Belgium, respectively, following natural EMCV outbreaks in domestic pigs . Nucleotide sequencing of most EMCV strains and Mengo virus has identified a poly(C) tract, consisting of 61 to 146 cytosine residues (C61 to C146), occasionally interrupted by 1 to 3 uridine residues, located at the 5' nontranslated region (NTR) of the positive-sense RNA genome . Studies to determine the role of the poly(C) tract in virus replication, virulence, and host range have produced conflicting findings. For instance, in earlier studies, truncation of the poly(C) tract of Mengo virus M/48 (from C44UC10 to C8- or C13UC10) attenuated the virus in mice, resulting in lower virus titers in the brain and development of milder meningoencephalitis . However, later, more-extensive experiments using a number of inbred strains of mice demonstrated that both short- and long-poly(C)-tract variants of Mengo virus M/48 were highly virulent in newborn mice of any strain . Studies using both EMCV R/45 and EMCV PV2/58 demonstrated no correlation between the length of the poly(C) tract and virulence. Truncation of the poly(C) tract in EMCV R/45 from C115UC3UC10 to C4, C9, or C20 produced only slight decreases in virulence among 12-week-old SJL mice . Strain PV2/58 (containing 118 C's) was cloned, and three viruses with different lengths of the poly(C) tract were generated; rPV2/dT had the poly(C) tract replaced by a poly(U) tract, rPV2/C20 had the poly(C) tract replaced with 20 C's, and rPV2/ran had the poly(C) tract replaced with a random sequence of 238 nucleotides . Infection of mice with either wild-type PV2/58, rPV2/dT, rPV2/C20, or rPV2/ran resulted in insignificantly higher virus loads and degrees of pathological lesions in mice infected with wild-type PV2/58 than in mice infected with recombinant viruses, suggesting that the poly(C) tract was not essential for virulence in mice . Another interesting aspect of EMCV pathogenesis is the ability of these viruses to cause interspecies infections . EMCV outbreaks in zoos in Australia and the United States have involved multiple animal species including lemurs, squirrels, macaques, mandrills, chimpanzees, hippopotami, kangaroos, and possibly humans . The few documented cases of EMCV infection in humans have been associated with fever, neck stiffness, lethargy, delirium, headaches, or vomiting . In Germany, strains of the virus have been isolated from children suffering from meningitis and encephalitis, but no causal relationship between EMCV and the symptoms has been demonstrated . There is renewed interest in pig-to-human zoonotic viruses because of advances in xenotransplantation as a means of overcoming the acute shortage of transplantation tissues and organs for humans. Porcine cells, tissues, and organs are the primary animal tissues being considered for human transplantation because of the similarities in anatomical and physiological features between humans and pigs, the availability of the species, and the relative ease of breeding pigs . It has recently been demonstrated that infection of 5-week-old pigs with EMCV 30/87 resulted in the deaths of 16% of the pigs 2 to 3 days after infection from acute myocarditis, characterized by extensive lysis of sarcoplasm, cellular degeneration, and mineralization in the myocardium . The rest of the infected pigs (84%) did not develop clinical illness but showed virus persistence for as long as 90 days in cardiomyocytes and central nervous system (CNS) cells, which was associated with extensive apoptosis and viral antigen expression. More importantly, EMCV 30/87 productively infected primary human cardiomyocytes, resulting in production of 100 to 1,000 PFU of infectious virus per cell within 6 h. In the present study, we demonstrate from the nucleotide sequence that EMCV 30/87 has the shortest poly(C) tract of any naturally occurring EMCV strain, and we determine its virulence and pathogenesis in rodents, pigs, and nonhuman primates. MATERIALS AND METHODS : Viruses. | An EMCV strain isolated from naturally infected pigs in Minnesota in 1987 (EMCV 30/87) was used in all our experiments . The virus was amplified by infection of HeLa cells, which were maintained in RPMI 1640 medium supplemented with 5% fetal bovine serum, l-glutamine, antibiotics, and antimycotics. The nucleotide sequences of EMCV 30/87 were compared to the sequences of Mengo virus M/48 and EMCV strains R/45, EMC-B/58, EMC-D/58, PV2/58, PV21/58, and 2887A/91. The year of the parent isolate has been added to the designation of each of these strains, and the same year is given to laboratory variants of a particular isolate (e.g., EMC-B/58, EMC-D/58, PV2/58, PV21/58). Nucleotide sequencing of EMCV 30/87. | EMCV 30/87 RNA from virus-infected HeLa cells was extracted by using a guanidinium thiocyanate-phenol-chloroform reagent (TRIzol; Invitrogen, San Diego, Calif.). The RNA, derived both from a mixed population of low-passage-number EMCV 30 and from a plaque-purified clone, was reverse transcribed using oligo(dT) primers with Superscript II reverse transcriptase (Invitrogen) at 42C for 50 min. Initial primers for amplification of 800 bp at the 5' untranslated end of the EMCV 30/87 sequence were designed by using EMCV strain R/45 (GenBank accession number ). Subsequent primers designed from EMCV 30/87 sequences were used to complete the rest of the genome. After reverse transcription, PCR was performed using 10 mul of the cDNA. For PCR, the cDNA was denatured at 94C for 5 min, followed by 30 cycles of denaturation (at 94C for 1 min), annealing (at 60C for 1 min), and elongation (at 72C for 1 min). The PCR products were gel purified (Qiagen, Valencia, Calif.) and either cloned into the PCR2.1 plasmid vector (Invitrogen) or directly sequenced. The subcloned EMCV 30/87 gene segments were sequenced with T7 forward primers and M13 reverse primers. Three independent clones of each segment, and more than five clones for the poly (C) tract, were sequenced. For the 5' and 3' ends of EMCV 30/87, the GeneRace kit (Invitrogen) was used as described previously . Nucleotide sequence analysis. | Nucleotide sequence editing and prediction of amino acid sequences were performed with DNAStar (Madison, Wis.) software. Alignments were performed by the CLUSTALW method by using the following previously sequenced EMCV strains: Mengo virus M/48 (GenBank accession number ), EMCV R/45 (accession number ), EMC-B/58 (accession number ), EMC-D/58 (accession number ), PV2/58 (accession number ), PV21/58 (accession number ), and 2887A/91 (accession number ). Nucleotide and predicted amino acid sequences were compared for each of the 12 genes and proteins (leader, VP1, VP2, VP3, VP4, 2A, 2B, 2C, 3A, 3B, 3C, and 3D) and also for the long 5' and short 3' NTRs. To determine relationships among the EMCV strains, phylogenetic analysis of the predicted amino acid sequences for the translated region was performed with PAUP (phylogenetic analysis using parsimony) software utilizing both parsimony and neighbor-joining analyses. The two analyses resulted in duplicate phylogenetic relationships and were each evaluated with 2,000 bootstrap replicates. Infection of mice. | Four- to 6-week-old C57BL/6 mice were purchased from Jackson Laboratory and housed in the University of Minnesota biosafety level-2 mouse facilities at the St. Paul campus. Handling of animals, including feeding and euthanasia, was in conformity with the National Institutes of Health and University of Minnesota institutional animal care guidelines. Mice were intraperitoneally inoculated with 104, 100, or 2 PFU of EMCV 30/87 in a 1-ml volume and were monitored for clinical signs of EMCV disease. Terminally sick mice were euthanatized, whereas mice that did not develop clinical disease (mice inoculated with 2 PFU) were sacrificed at day 14, 21, or 45 postinfection (p.i.). Infection of pigs. | Thirty-seven 5-week-old pigs were obtained from an EMCV-free swine herd (Midwest Research Swine, Gibbon, Minn.) and placed in negative-pressure isolation units at the University of Minnesota animal facilities. An enzyme-linked immunosorbent assay (ELISA) was used to confirm that the pigs were negative for EMCV antibodies before infection. Animals were intraperitoneally inoculated with 2.9 x 108 PFU of EMCV 30/87 in a 1-ml volume. Handling of animals, including feeding and euthanasia, was in conformity with National Institutes of Health and University of Minnesota institutional animal care guidelines. Four to six pigs were euthanized on days 7, 21, 45, and 90 p.i. by using pentobarbital sodium. Tissues from the brain, heart, kidney, liver, spleen, skeletal muscle, and pancreas were collected for isolation of infectious virus and viral RNA, histopathology, and immunohistochemistry. Infection of cynomolgus macaques. | Fourteen 2- to 4-year-old specific-pathogen-free cynomolgus macaques (Macaca fascicularis) were purchased from Charles River BRF (Houston, Tex.) and housed at Gwathmey Incorporated (Cambridge, Mass.) animal facilities, where experiments were performed. Protocols for handling of the macaques, including feeding and euthanasia, were approved by both Gwathmey Inc. and the University of Minnesota Institutional Animal Care and Use Committee and were in conformity with the National Institutes of Health guidelines. Twelve macaques were intraperitoneally inoculated with 2.9 x 108 PFU of EMCV 30/87 in a 1-ml volume, and two macaques were inoculated with sterile buffer as controls. Macaques were monitored daily for clinical signs of EMCV disease, and three animals were sacrificed at days 14, 21, and 45 p.i. for collection of the brain, heart, kidney, liver, spleen, skeletal muscle, and pancreas. Tissues were analyzed for histopathologic and immunohistochemical changes and for the presence of infectious virus and viral RNA. Sample processing and histopathologic analysis. | Sections of heart, brain, liver, kidney, spleen, skeletal muscle, and pancreas from pigs, macaques, and mice infected with EMCV for 2, 7, 14, 21, or 45 to 50 days were collected for histopathologic analysis, virus isolation, and RNA isolation. Tissues for RNA isolation were snap-frozen in liquid nitrogen, whereas tissues for histopathology and in situ hybridization were fixed in 10% neutral buffered formalin and embedded in paraffin. Paraffin-embedded sections were cut at a thickness of 4 mum and stained with hematoxylin and eosin for histopathologic analysis. Serum samples collected from the animals at sacrifice were used to determine levels of virus-specific immunoglobulin G (IgG) by ELISA. In situ hybridization. | To detect the presence of viral genomes in pig tissues, in situ hybridization was performed by using a 414-bp VP1-specific probe as described previously for Theiler's murine encephalomyelitis virus . Briefly, a VP1 cDNA was subcloned from EMCV 30/87 into plasmid pUC18 by using BamHI and HindIII restriction sites, and the cDNA probe was prepared by digesting the VP1 plasmid with the NcoI and BamHI restriction enzymes . The probe was labeled with 35S-dATP by using the Random Primers DNA labeling system (Invitrogen) and was purified by using G-50 Sephadex Quick Spin columns (Roche, Indianapolis, Ind.). To prepare tissue samples, paraffin-embedded sections were deparaffinized by using xylene. Sections were first digested with 10 mug of proteinase K/ml in phosphate-buffered saline for 30 min at 37C and then treated with 0.1 M triethanolamine containing acetic anhydride. The sections were prehybridized in a buffer containing deionized formamide, Denhardt's solution, sodium chloride, salmon sperm DNA, yeast total RNA, and yeast tRNA for 4 h at room temperature before hybridization with a 35S-labeled 309-bp VP1 probe. Hybridization was performed overnight at 37C and was followed by extensive washes in reducing buffer at 55C. Air-dried slides were immersed in an NTB2 film emulsion (Eastman Kodak Co., Rochester, N.Y.) and exposed at 4C for 5 days. RT-PCR. | Pig, macaque, and mouse tissues collected at days 2, 7, 14, 21, 45, and 90 p.i. were analyzed for EMCV RNA by nested reverse transcription-PCR (RT-PCR) as described previously . Samples of brain, heart, liver, kidney, spleen, or skeletal muscle tissues were homogenized in TRIzol (Invitrogen), followed by chloroform extraction of total RNA. Five micrograms of total RNA was reverse transcribed by using an oligo(dT) primer and the Superscript II reverse transcription kit (Invitrogen) before the outer and nesting PCRs were performed using primer pairs specific for the VP1 or VP2 gene of EMCV 30/87 sequenced in our laboratory. The VP1 primers used for the outer PCR were GCCTCAGTTTGACCCTGCTTATG (5' primer) and CGGCTCTCGGAGTCATGTCAATC (3' primer) (product size, 592 bp), and those for the nesting reaction were CGTCTCACAGAAATTTGGGGCAAC (5' primer) and CCAGGCTTCCTGTGTTGTCAAATC (3' primer) (product size, 340 bp). The VP2 primers for the outer PCR were CAGTAGGCCGTCTTGTTGGTTATG for the 5' end and CACTTCAAGATCCACGGTGGTGTTG for the 3' end (product size, 499 bp), and the nesting primers were GGCACTGTTCATGATGGAGAACAC for the 5' end and GGTGATCCATAGCAAAGGGACCTTTC for the 3' end (product size, 350 bp). All primer sequences are given in 5'-to-3' orientation. The outer PCR was performed on 1 mul of the template cDNA by adding 0.2 mM deoxynucleoside triphosphates, 2 mM magnesium chloride, 10 pmol of each primer, and 1 U of Taq polymerase (Invitrogen). For the nested PCR, 2.5 mul of the first-round PCR mixture was used. PCR products were analyzed by agarose gel electrophoresis. Nucleotide sequence accession number. | Nucleotide sequences for EMCV 30/87 have been submitted to GenBank (accession number ). RESULTS : Molecular analysis of EMCV 30/87 and comparison with other EMCV strains. | The entire EMCV 30/87 genome was sequenced for comparison with other EMCV strains and correlated with the pathogenesis of virus infection in different animal species. The sizes of the LP, VP1, VP2, VP3, VP4, 2A, 2B, 2C, 3A, 3B, 3C, and 3D genes of EMCV 30/87 (GenBank accession number ) were similar to those of strains 2887A/91 (accession number ), EMCV R/45 (accession number ), PV21/58 (accession number ), EMC-B/58 (accession number ), EMC-D/58 (accession number ), and PV2/58 (accession number ). However, Mengo virus M/48 (accession number ) had an extra amino acid (proline) in the 2B protein at position 10 from the N terminus of the protein. Comparison of the nucleotide and predicted amino acid sequences for the coding regions encompassing the 12 genes demonstrated that EMCV 30/87 shared the highest overall nucleotide sequence identity (84.2%) and predicted amino acid identity (96.4%) with strain 2887A/91, isolated from an aborted porcine fetus in Belgium, and with strain PV21/58, a laboratory isolate from a virus isolated from a pig myocardium in Panama (Tables and ). EMCV 30/87 also had high nucleotide (84.0%) and amino acid (96.6%) sequence identities with EMCV R/45, isolated from the myocardium of a chimpanzee. Of the 12 EMCV genes, 2B was the most conserved among the eight EMCV strains, demonstrating 90.4 to 94.2% nucleotide sequence identity and 94.7 to 97.3% predicted amino acid identity (Tables and ). Phylogenetic comparison of the seven EMCV strains and Mengo virus revealed the closeness of EMC-B/58, EMC-D/58, and PV2/58 , all of which are variant mutants derived from EMCV-M. EMCV R/45 and 2887A/91 segregated together, and EMCV 30/87 appeared to segregate between the two clusters. Mengo virus M/48 segregated far from the seven EMCV strains. Perhaps the most intriguing finding was that EMCV 30/87 had the shortest poly(C) tract , which was observed both in a mixed population of low-passage-number EMCV 30 used to inoculate mice and in two plaque-purified strains of the virus (EMCV 30/PA and EMCV 30/PB). EMCV 30/87 had a poly(C) tract consisting of 14 C's (CUC5UC8), whereas the older EMCV isolates (EMCV R/45, PV2/58, PV21/58, EMC-B/58, and EMC-D/58) had 118 to 142 C's and Mengo virus M/48 had 61 C's. Strain 2887A/91, another recent isolate from an aborted pig fetus, also had a short poly(C) tract (C10UCUC3UC10) and was highly pathogenic. The 5' NTR of EMCV 30/87 consisted of approximately 724 nucleotides, a size comparable to that of strain 2887A/91 (728 nucleotides) but 135 nucleotides shorter than that of PV21/58, the virus with the longest poly(C) tract . The 10 to 21 nucleotides at the immediate 5' side of the poly(C) tract consisted of 60 to 75% A residues, leading to a highly conserved region in all EMCV strains and Mengo virus . Another conserved region was observed 4 to 6 nucleotides downstream of the poly(C) tract . Analysis of the 3' NTR of EMCV 30/87 revealed no major differences among eight EMCV strains. The 3' NTR consisted of 125 nucleotides in addition to a short poly(A) tail, which was comparable to the 3' NTRs of all other EMCV strains (107 to 126 nucleotides) and Mengo virus M/48 (125 nucleotides). EMCV-induced clinical disease in mice, pigs, and macaques. | All mice (n = 25) inoculated with >=4 PFU of EMCV 30/87 died between days 2 and 7 p.i. from acute encephalitis . The clinical signs in mice included hunched posture, ruffled fur, lethargy, anorexia, and hind limb paralysis. The clinical signs lasted for 24 to 72 h before mice became moribund with signs of severe respiratory failure. No mice died of sudden cardiac failure, which was observed in 16% of EMCV 30/87-infected pigs. Mice inoculated with 104 to 105 PFU of EMCV 30/87 developed disease 2 to 4 days p.i., whereas mice inoculated with 4 to 100 PFU of virus developed disease between days 5 and 7 p.i. Terminally, mice developed posterior limb paresis or complete paralysis before rapidly becoming moribund. Six of the 37 pigs inoculated with 2.9 x 108 PFU of EMCV 30/87 (16.2%) died of acute myocardial failure 2 to 3 days later. The rest of the pigs did not develop any overt clinical signs. None of the 12 cynomolgus macaques inoculated with 2.9 x 108 PFU of EMCV 30/87 (the same virus dose as that given to pigs) developed detectable clinical signs. Electrocardiographic profiles were not determined for mice, pigs, or macaques. A summary of the survival curves of mice, pigs, and macaques following EMCV infection is presented in Fig. . Sham-inoculated pigs (n = 2), mice (n = 4), and macaques (n = 2) remained asymptomatic throughout the experimental period and had no fatalities. Pathogenesis of EMCV 30/87 in mice. | All EMCV 30/87-infected mice died within 2 to 7 days. Following intraperitoneal inoculation, there was widespread virus distribution, as demonstrated by detection of EMCV RNA in the blood, spleen, kidneys, liver, heart, brain, and skeletal muscle by RT-PCR (Fig. ; Table ). Viral RNA could also be localized by in situ hybridization, particularly in tissue sections from the heart and brain . Histopathologic changes were observed in the heart and brain but not in the liver, spleen, kidney, pancreas, or skeletal muscle. In the heart, there were large foci of lymphocytic infiltration in the myocardium , accompanied by degeneration and necrosis of cardiac myocytes and sarcoplasm. In the brain, similar foci of lymphocytic infiltration were observed in the cerebral cortices , hypothalami, and hippocampi. To investigate whether EMCV persistence and chronic disease can occur in mice, we inoculated mice (n = 8) with lower titers (2 to 10 PFU), but they did not develop either clinical disease or anti-EMCV antibodies, indicating that the virus titers were too low to establish an infection. In addition, no EMCV RNA was detected in the brain, heart, spleen, pancreas, blood, kidney, liver, or skeletal muscle in these mice 21 days p.i. , further indicating that no infection was established. Pathogenesis of EMCV 30/87 in pigs. | The detailed clinical and histopathologic profile of EMCV 30/87 in swine has been described previously . Limited experiments performed to provide data for direct comparison with viral pathogenesis in mice and cynomolgus macaques reaffirmed that the virus replicated primarily in the brain, heart, and spleen, rarely in the liver, and not in the kidneys and skeletal muscles during the first 7 days. However, in the chronic phase of the disease, persistent viral RNA was more widespread and was detected in the brain, liver, heart, spleen, kidney, pancreas, and skeletal muscles. Histopathologic changes in both the acute and chronic phases were confined to the heart and brain. Pigs sacrificed in the acute phase of the disease (days 7 to 21 p.i.) exhibited severe myocardial lesions, including multiple foci of lymphocytic infiltration, degeneration, and necrosis . Brain changes included infiltration of lymphocytes in the cerebral cortices and meninges, as well as perivascular cuffing in the cerebral cortex and hippocampi. In the chronic phase of the disease (days 45 to 90 p.i.), multiple discrete foci of myocardial mineralization and lymphocytic infiltration were observed , which were larger and more numerous than the lesions observed in macaques. Brain lesions in the chronic stages of disease were characterized by perivascular cuffing in the cerebral cortex , medulla, and cerebellum. EMCV persistence in the heart, brain, liver, kidney, spleen, skeletal muscle, pancreas, and mesenteric lymph nodes during the acute phase, but primarily in the heart and brain in the chronic phase, was demonstrated by using in situ hybridization and RT-PCR. There was mild lymphocytic infiltration in the spleen in the acute phase, but no pathological changes were detected in the liver, pancreas, kidney, skeletal muscles, or mesenteric lymph nodes at any stage of the disease. EMCV pathogenesis in cynomolgus macaques. | As shown in Fig. , none of the 12 infected macaques developed overt clinical signs, and there were no fatalities among them, even though the same dose of virus (2.9 x 108 PFU) was used to infect both pigs and macaques. Following EMCV 30/87 inoculation, viral RNA was detected in the blood, spleen, liver, heart, kidney, brain, and skeletal muscles (Fig. ; Table ). The viral RNA localized by in situ hybridization in the heart and brain were associated with several pathological changes. However, although viral RNA was detected in the kidney, liver, spleen, pancreas, or skeletal muscles at day 7 p.i., no histopathologic abnormalities were demonstrated in these organs, as illustrated by the micrograph showing viral RNA in the kidney despite the normal histology of the organ . Histopathologic changes in the brain and heart were observed in three of the four macaques at day 7 p.i. In the myocardium, there were numerous large foci of inflammation associated with necrosis and degeneration of cardiac myocytes. The lesions were similar to those observed in acutely infected pigs but were less frequent, suggesting that primate myocytes were not as susceptible to the porcine virus. Histopathologic changes in the brain included numerous foci of perivascular cuffing , which were most prominent in the cerebrum and the cerebral cortex. There were histopathologic changes in 7 of 8 spleens (87.5%), 3 of 8 hearts (37.5%), and 2 of 8 brains (25%) for macaques infected for 21 and 45 days. Splenic lesions were characterized by highly prominent germinal centers in the chronically infected macaques compared to those in uninfected or 7-day-infected macaques . Myocardial lesions were characterized by small discrete foci of lymphocytic infiltration , with minimal disruption of cardiac myofiber arrangement. Brain lesions included small foci of infiltrating lymphocytes in blood vessel walls. Analysis of viral RNA distribution by in situ hybridization and RT-PCR demonstrated EMCV RNA most consistently in the spleen (75% of the cases at day 45 p.i.), but also in the heart (50%), brain (50%), pancreas (25%), and kidney (25%), as demonstrated in Table . Serum samples from macaques infected for 7 days had low levels of virus-specific IgG, whereas high levels of EMCV-specific IgG were detected at both days 21 and 45 p.i. . DISCUSSION : Studies have suggested that the poly(C) tract in EMCV strains and Mengo virus may be a determinant in pathogenesis, virulence, and host range . Recombinant Mengo virus M/48 variants with shortened poly(C) tracts (C0 or C24) were highly attenuated, producing either no histopathologic changes (C0) or mild meningitis (C24) in 4-week-old mice . However, the Mengo virus variants with short poly(C) tracts had growth kinetics and plaque sizes similar to those of wild-type Mengo virus M/48 (C63) in vitro, and they were immunogenic in mice, pigs, and macaques, indicating that the viruses were infectious in these animals . After sequencing the genome of EMCV 30/87 and determining that it had the shortest poly(C) tract (CUC5UC8) of all wild-type EMCV strains, we determined the pathogenesis and virulence of the virus in various animal species. Our data demonstrate that EMCV 30/87, a low-passage-number (4 or 5 passages) natural isolate with a short poly(C) tract (CUC5UC8), is highly virulent, resulting in substantial fatalities in mice and pigs and severe pathological changes in mice, pigs, and macaques. Similar findings were observed with the wild-type strain 2887A/91 (C10UCUC3UC10), which induced acute fatal encephalitis characterized by hind limb paralysis as early as day 1 p.i. in mice and had a 50% lethal dose (LD50) of 1 PFU . The direct correlation between the length of the poly(C) tract and virulence may be specific to Mengo virus M/48, which is phylogenetically different from most of the EMCVs, as shown in Fig. . This hypothesis is supported by studies using poly(C) tract variants from both EMCV R/45 and PV2/58, which demonstrated no correlation between the length of the poly(C) tract and virulence . Another possible explanation for the correlation between virulence and the length of the poly(C) tract in Mengo M variants is that continuous passage and genetic engineering of EMCV strains and Mengo virus result in significant attenuation, as demonstrated with strain 2887A/91 . Comparison of the parent strain 2887A/91 and an infectious cDNA clone generated from the parent virus demonstrated that the recombinant 2887A/91 clone had smaller plaque sizes. Mice inoculated with the infectious 2887A/91 cDNA clone developed delayed clinical signs (from day 5 p.i.), with an LD50 of 100 PFU, whereas mice infected with the 2887A/91 parent virus developed clinical signs from day 1 p.i., and the LD50 was 1 PFU . A more intriguing fact is that EMCV 30/87 and strain 2887A/91, both recent isolates from pigs, have 4- to 10-fold-shorter poly(C) tracts than the older EMCVs , suggesting that evolutionary pressure may favor the emergence of EMCV strains with shorter poly(C) tracts, at least in pigs. A number of more recent EMCV strains have been isolated from pigs and rodents, but their genomic structures, in particular the lengths of their poly(C) tracts, have not been determined . It was demonstrated previously that experimental infection of pigs with EMCV 30/87 resulted in acute fatal cardiac failure for 16% of the pigs and persistent infection in myocardial and CNS cells for the remaining 84% . In addition, EMCV 30/87 productively infected primary human cardiomyocytes, suggesting that undetected EMCV in porcine neural or myocardial tissues transplanted into humans may result in severe infection in recipients. The studies described in this report were also undertaken in order to determine the ability of EMCVs to cause interspecies disease outbreaks by comparing the pathogenesis and disease phenotypes induced by EMCV 30/87 in C57BL/6 mice, pigs, and cynomolgus macaques. Infection of mice, pigs, and cynomolgus macaques resulted in similar EMCV 30/87 pathogenesis, characterized by generalized virus distribution followed by localization of virus in the myocardium and CNS, which were the primary sites of virus replication and histopathologic changes in the three animal species. Viral RNA was detected in the blood, spleen, kidneys, liver, heart, and brain in 75 to100% of infected macaques, and in more than 50% of infected mice, within 7 days p.i. Interestingly, the virus spread was slower in pigs, requiring as long as 21 days for detection of viral RNA in all the organs. The most severe lesions (based on the extent of lymphocytic infiltration and the number of lesions) were observed in pigs, followed by mice, with macaques having the least severe disease. The differences in disease phenotype among the three animal species were intriguing. All infected mice developed acute encephalitis characterized by paresis and flaccid paralysis of the posterior limbs and resulting in death from acute respiratory failure between days 2 and 7 p.i. In contrast, a small proportion of pigs (16%), the original host of the virus, died from sudden cardiac failure 2 days after infection, whereas the rest of the animals remained overtly asymptomatic for as long as 90 days p.i., despite extensive myocardial and CNS histopathologic changes. All EMCV 30/87-infected macaques remained asymptomatic for the entire experimental period (45 days), despite extensive myocardial pathological changes and virus persistence in more than 50% of the animals. Although limited to mice only, previous studies suggested that EMCV strains isolated from pigs or primates could induce severe disease in other animals. EMCV R/45, isolated from a 5-year-old chimpanzee suffering from acute fatal myocarditis in 1945, induced paralysis of the hind legs and interstitial myocarditis in mice . Mengo virus M/48, isolated from the cerebrospinal fluid of a paralyzed rhesus macaque, induced paralysis, characterized by lesions in brain and spinal cords, and myocarditis in mice . EMC-B/58 and EMC-D/58, variants of EMC-M isolated from a naturally infected domestic pig suffering from severe myocarditis, have been widely used to study the pathogenesis of viral diabetes in a mouse model . Strain 2887A/91 induced an acute fatal encephalitis characterized by paralysis of posterior limbs as early as 1 day after infection in mice . Since we have demonstrated that porcine EMCV can productively infect human cardiomyocytes, cynomolgus macaques were infected with EMCV 30/87 as a model of pig-to-human xenozoonosis, designed to provide an indication of porcine EMCV virulence and pathogenesis in humans. The absence of severe clinical disease or death in macaques infected with EMCV 30/87 may suggest that porcine EMCV isolates transmitted during xenotransplantation may not produce severe clinical disease in humans. However, the absence of overt disease may have been due to the fact that the macaques were 2 to 4 years old at the time of EMCV infection, whereas the pigs and mice were 5 weeks old at infection. The finding of widespread virus distribution among internal organs and severe histopathologic changes in the macaques is more significant, particularly because recent studies in our laboratory have demonstrated the transmission of EMCV 30/87 and induction of disease in mice following transplantation (intra-abdominal) of porcine myocardial or pancreatic tissues . Using a more clinically relevant model, we have demonstrated that EMCV 30/87-infected porcine islet cells can transmit the virus and induce acute fatal disease in mice (L. A. Brewer, R. S. LaRue, B. Hering, and M. K. Njenga, submitted for publication). Given that long-term immunosuppression will probably accompany xenotransplantation, the findings that a porcine strain of EMCV can establish severe infection in primates and that it can be transmitted during transplantation indicate the need to develop a rapid test to screen pig tissues harvested for transplantation for EMCV RNA. FIG. 1. : Phylogenetic relationship between EMCV 30/87 and other EMCV strains. Phylogenetic relationship between EMCV 30/87 and other EMCV strains. Following alignment of the contiguous predicted amino acid sequences for the translated regions of the viruses (LP, VP4, VP2, VP3, VP1, 2A, 2B, 2C, 3A, 3B, 3C, and 3D), a rooted phylogram was generated by maximum parsimony analysis using Mengo virus M/48 as the outgroup. Absolute distances are listed above each branch, with bootstrap confidence levels given below in parentheses. GenBank accession numbers for the viruses are given in Materials and Methods. FIG. 2. : Survival curves of EMCV-infected mice, pigs, and macaques. Survival curves of EMCV-infected mice, pigs, and macaques. Six-week-old C57BL/6 mice (n = 30), 5-week-old pigs (n = 35), and 2- to 4-year-old cynomolgus macaques (n = 12) were intraperitoneally inoculated with EMCV 30/87 and monitored daily for clinical signs and fatality for 45 days. Sham-inoculated pigs (n = 2), mice (n = 4), and macaques (n = 2) remained asymptomatic and had no fatalities. FIG. 3. : Pathological changes and detection of viral RNA in mice infected with EMCV 30/87. Pathological changes and detection of viral RNA in mice infected with EMCV 30/87. Five-week-old C57BL/6 mice were intraperitoneally inoculated with 102 to 103 PFU of EMCV 30/87, and histopathologic changes and the presence of EMCV RNA in the blood (Bl), heart (Ht), brain (Br), spleen (S), pancreas (P), liver (L), kidney (K), and skeletal muscle (SM) were analyzed at 2 to 8 days p.i. All EMCV 30/87-infected mice inoculated with more than 102 PFU of virus died between days 2 and 8 p.i. Histopathologic changes were confined to the brain and heart. In the heart, there were large foci of lymphocytic infiltration in the myocardium (A), accompanied by degeneration and necrosis of cardiac myocytes, whereas in the brain, similar lymphocytic infiltrations were observed in the cerebral cortices (B), hypothalami, and hippocampi. EMCV RNA was detected in both the heart (C) and brain (D) by in situ hybridization, but also in the blood, kidney, spleen, and skeletal muscle by RT-PCR (E). In the brain and heart, which had high levels of EMCV RNA, nested RT-PCR using VP2 primers followed by electrophoresis resulted in two visible bands, the larger (499 bp) representing the primary reaction product and the smaller (350 bp) representing the nested reaction product. For histopathology, heart and brain sections were embedded in paraffin and 4-mum-thick sections were stained with hematoxylin and eosin. In situ hybridization was performed using a 35S-labeled VP1 cDNA probe. Results for mice inoculated with 2 to 10 PFU are shown in the right portion of panel E. FIG. 4. : EMCV-induced pathological changes and detection of viral RNA in pigs. EMCV-induced pathological changes and detection of viral RNA in pigs. Five-week-old pigs were intraperitoneally inoculated with 2.9 x 108 PFU of EMCV 30, and histopathologic changes and the presence of EMCV RNA in the heart, brain, spleen, pancreas, liver, kidney, and skeletal muscle were analyzed 7, 21, 45, and 90 days p.i. In the acute phase (7 days p.i.), inflammation and degenerative changes were observed in the heart (A), and brain (B), but no changes were observed in the other organs. In the chronic phase (21 to 90 days p.i), infiltration of lymphocytes in the heart (C) and perivascular cuffing in the brain (D) were evident throughout the infection period up to day 90 p.i. and were accompanied by persistence, as demonstrated by localization of EMCV RNA by in situ hybridization (E and F). For histopathology, heart and brain sections were embedded in paraffin, and 4-mum-thick sections were stained with hematoxylin and eosin. In situ hybridization was performed by using a 35S-labeled VP1 cDNA probe. FIG.5. : EMCV-induced pathological changes and detection of viral RNA in cynomolgus macaques. EMCV-induced pathological changes and detection of viral RNA in cynomolgus macaques. Adult macaques were intraperitoneally inoculated with 2.9 x 108 PFU of EMCV 30/87, and histopathologic changes were analyzed 7, 21, and 45 days later. At day 7 p.i., there were acute inflammatory and mild degenerative changes in the myocardium (A) and brain (B, arrows), whereas there were no detectable changes in kidneys (C), livers, and spleens from the same animals. In contrast, EMCV RNA was widely distributed in tissues at day 7 p.i., as demonstrated by in situ hybridization (black grains) in the myocardium (D), brain (E), and kidneys (F) and by RT-PCR in the heart (Ht), brain (Br), liver (L), spleen (S), kidney (K), pancreas (P), and skeletal muscle (SM) (J). At days 21 and 45 p.i., mild lymphocytic infiltration was observed in the heart (I, arrow), and the formation of prominent germinal centers was observed in the spleen (H, arrows), in contrast to spleens from 7-day EMCV-infected macaques (G, arrow). EMCV RNA was most consistently observed in the spleen and pancreas at days 21 and 45 p.i. (J). In tissues with high levels of EMCV RNA at days 7 (all tissues), 21 (spleen), and 45 (spleen) p.i., nested RT-PCR using the VP2 primers followed by agarose gel electrophoresis resulted in two visible bands, the larger (499 bp) representing the primary reaction product and the smaller (350 bp) representing the nesting reaction product. For histopathology, macaque tissue sections were embedded in paraffin and 4-mum-thick sections were stained with hematoxylin and eosin. In situ hybridization was performed using a 35S-labeled VP1 cDNA probe. FIG. 6. : Detection by ELISA of EMCV-specific IgG in serum samples from experimentally infected cynomolgus macaques. Detection by ELISA of EMCV-specific IgG in serum samples from experimentally infected cynomolgus macaques. (A) Virus-specific IgG levels in serum samples from macaques infected for 7, 21, or 45 days. Data are presented as mean A630 readings (+- standard errors) from four serum samples at each time point performed at serum dilutions between 1:500 and 1:2,000. (B) Profile of virus-specific IgG levels through the 45-day experimental period as determined at a 1:500 serum dilution. TABLE 1 : Percent nucleotide sequence identity between the genes of EMCV 30/87 and those of other EMCV strains and Mengo virus TABLE 2 : Percent predicted amino acid sequence identity between proteins of EMCV 30/87 and those of other EMCV strains and Mengo virus TABLE 3 : Comparison of the length of the poly(C) tract and the nucleotide sequences of the adjoining regions in EMCV 30/87 with those in other EMCV strains and Mengo virus M/48 TABLE 4 : Percent EMCV RNA-positive animals following EMCV 30/87 infection Backmatter: PMID- 12915543 TI - Recombinant Sindbis/Venezuelan Equine Encephalitis Virus Is Highly Attenuated and Immunogenic AB - Venezuelan equine encephalitis virus (VEEV) is an important, naturally emerging zoonotic virus. VEEV was a significant human and equine pathogen for much of the past century, and recent outbreaks in Venezuela and Colombia (1995), with about 100,000 human cases, indicate that this virus still poses a serious public health threat. The live attenuated TC-83 vaccine strain of VEEV was developed in the 1960s using a traditional approach of serial passaging in tissue culture of the virulent Trinidad donkey (TrD) strain. This vaccine presents several problems, including adverse, sometimes severe reactions in many human vaccinees. The TC-83 strain also retains residual murine virulence and is lethal for suckling mice after intracerebral (i.c.) or subcutaneous (s.c.) inoculation. To overcome these negative effects, we developed a recombinant, chimeric Sindbis/VEE virus (SIN-83) that is more highly attenuated. The genome of this virus encoded the replicative enzymes and the cis-acting RNA elements derived from Sindbis virus (SINV), one of the least human-pathogenic alphaviruses. The structural proteins were derived from VEEV TC-83. The SIN-83 virus, which contained an additional adaptive mutation in the nsP2 gene, replicated efficiently in common cell lines and did not cause detectable disease in adult or suckling mice after either i.c. or s.c. inoculation. However, SIN-83-vaccinated mice were efficiently protected against challenge with pathogenic strains of VEEV. Our findings suggest that the use of the SINV genome as a vector for expression of structural proteins derived from more pathogenic, encephalitic alphaviruses is a promising strategy for alphavirus vaccine development. Keywords: Introduction : Venezuelan equine encephalitis virus (VEEV) is a member of the Alphavirus genus in the Togaviridae family. VEEV is an enveloped virus with a nonsegmented, positive-sense RNA genome of approximately 11.5 kb. The 5' two-thirds of the genome encodes four nonstructural proteins (nsP1 to nsP4) that form an enzyme complex required for viral replication . After viral RNA entry into the cytoplasm, a nonstructural polyprotein is translated directly from the viral genome and utilized in the production of a full-length, negative-sense replicative RNA intermediate. This RNA is then used as a template for synthesis of positive-sense genomic RNA and for transcription of a subgenomic 26S RNA. The ca. 4-kb subgenomic RNA corresponds to the 3' one-third of the viral genome and is translated into a structural polyprotein that is proteolytically cleaved into the capsid and envelope glycoproteins E2 and E1 . Two hundred forty copies of the capsid protein combine with the genomic viral RNA to form an icosahedral nucleocapsid. Finally, the nucleocapsid buds from the plasma membrane to acquire a lipid envelope with embedded protein spikes containing E1-E2 heterodimers . VEEV was a significant human and equine pathogen for much of the past century, and recent epidemics indicate that VEEV still represents a serious public health threat. Furthermore, VEEV is a potential biological warfare and terrorism threat. Alphaviruses in the Venezuelan equine encephalitis (VEE) complex are serologically classified into six distinct antigenic subtypes . Historically, only members of subtypes IAB and IC were associated with major epidemics and equine epizootics. During the most recent major outbreak in Venezuela and Colombia (1995) involving subtype IC VEEV, about 100,000 human cases occurred, with more than 300 fatal encephalitis cases estimated . VEEV infection of mice leads to a biphasic disease with initial replication in lymphoid tissues followed by viremia and penetration into the central nervous system (CNS), where the virus replicates until the death of the infected animal (-, ). The result of the CNS infection is acute meningoencephalitis that leads to massive death of neuronal cells . During VEEV epizootics, equine mortality due to encephalitis can reach 83%; in humans, while the overall mortality rate is low (<1%), neurological disease including disorientation, ataxia, mental depression, and convulsions can be detected in up to 14% of infected individuals, especially children . Sequelae of VEE-related clinical encephalitis in humans and rats are also described . The predominant pathological findings in fatal human VEE cases reveal the following: (i) in the CNS, edema, congestion, hemorrhages, vasculitis, meningitis and encephalitis; (ii) in the lungs, interstitial pneumonia, alveolar hemorrhage, congestion, and edema; (iii) in lymphoid tissue, follicular necrosis and lymphocyte depletion; and (iv) in the liver, diffuse hepatocellular degeneration . A small animal (mouse) model for VEE-induced encephalitis and lymphotropism is well established ; however, mice do not develop the pulmonary and hepatic symptoms that were described at high rates (91 and 61%, respectively) in fatal human VEE cases . The live-attenuated TC-83 vaccine strain (TC-83) was developed 4 decades ago by serial passaging of the virulent, subtype IAB Trinidad donkey (TrD) VEEV strain in guinea pig heart cell cultures . Currently, TC-83 is still the only VEEV strain available for vaccination of laboratory workers and military personnel. More than 8,000 humans have been vaccinated during the past 4 decades . The cumulative data unambiguously demonstrated that nearly 40% of vaccinated people develop a disease with some symptoms typical of natural VEEV infection, including febrile, systemic illness and other adverse effects . The TC-83 vaccine has also been used for immunization of equines and can lead to febrile clinical illness with viremia levels that are potentially sufficient to infect mosquitoes and initiate a transmission cycle . These findings indicate significant residual virulence of TC-83 strain in both humans and horses . Residual virulence has also been detected in mice, where TC-83 was uniformly lethal for the C3H/HeN strain after intracerebral (i.c.) inoculation and produced clinical illness in BALB/c and C3H/HeN mice for almost 14 days after subcutaneous (s.c.) inoculation . A formalin-inactivated version of the TC-83 vaccine, C-84, is used for individuals who fail to seroconvert to TC-83 . However, repeated boosters with C-84 are required to generate and maintain an effective immune response, and the induced neutralizing antibodies do not persist for a long time. A more promising candidate vaccine was described recently, when attenuation of the VEEV TrD strain was achieved by the introduction of lethal mutations into the PE2 furin cleavage site of an infectious cDNA clone, followed by selection of a second-site suppressor mutation in the E1 glycoprotein . This virus is highly attenuated for laboratory rodents. This strategy minimized the potential for direct reversion to virulence. However, the potential for reversion to wild-type virulence via compensatory mutations remains. To further overcome the problem of possible reversion to wild-type virulence of live-attenuated VEEV vaccine strains, we have applied an alternative approach using a recombinant virus capable of inducing a protective immune response. Utilizing the genome of Sindbis virus (SINV), a relatively human-nonpathogenic alphavirus, as a vector, we developed a chimeric SIN/VEE virus expressing all of the structural proteins of the TC-83 strain. This recombinant virus, SIN-83, replicated to high titers in cell culture and protected mice against lethal challenge with high doses of recently isolated, naturally circulating subtype IC and ID strains of VEEV. A distinguishing feature of this SIN-83 chimeric virus is its inability to cause any detectable clinical disease in either adult or weanling mice after either s.c. or i.c. inoculation. MATERIALS AND METHODS : Cell cultures. | BHK-21 cells were obtained from the American Type Culture Collection (Rockville, Md.). Vero cells were kindly provided by Charles M. Rice (Rockefeller University, New York, N.Y.). Both cell lines were propagated in Alpha minimal essential medium (MEM) supplemented with 10% fetal bovine serum (FBS) and vitamins. Mice. | Weanling (male and female) NIH Swiss mice purchased from Harlan (Indianapolis, Ind.) were maintained for 6 days after birth in a BSL-3 facility prior to experimental infection. Plasmid constructs. | All plasmids were constructed by standard recombinant DNA techniques. The parental pToto1101 plasmid, encoding the SINV genome , and the pTC-83 plasmid, encoding the genome of VEEV TC-83 , were obtained from Charles M. Rice (Rockefeller University, New York, N.Y.) and Richard Kinney (Centers for Disease Control, Fort Collins, Colo.), respectively. Fragments containing the SINV subgenomic promoter and the 5' untranslated region (UTR) of the VEEV subgenomic RNA were generated by PCR amplification, cloned into the pRS2 plasmid for sequencing, and then used for generating the cDNA clone of the chimeric SIN-83S virus genome. The sequence of the junction region between the SINV and VEEV sequences is shown in Fig. . The construct, pSIN-83S, contained the promoter for SP6 RNA polymerase, followed by nucleotides (nt) 1 to 7601 of the SINV genome, nt 7536 to 11382 of VEEV TC-83 (with an additional C->T mutation of nt 7555), an AGGCCTTGGG sequence, and a 355-nt sequence containing the SINV 3'UTR (starting from nt 11394), poly(A) followed by an XhoI restriction site. A map and se-quences of the plasmid are available from the authors upon request. pSIN-83 was different from pSIN-83S by a single point mutation at nt 4065, changing Ser795 in nsP2 to Thr. Plasmids pZPC and pSH, containing infectious cDNAs of VEEV strains ZPC738 (subtype ID) and SH3 (subtype IC), respectively, will be described elsewhere (a publication describing these two clones is in preparation). The ZPC738 strain was isolated from a sentinel hamster exposed in a tropical lowland forest of Zulia State, Venezuela, in 1997, and strain SH3 was isolated from a febrile human during the 1992-3 Venezuelan epidemic/epizootic . RNA transcription and transfection. | Plasmids were purified by centrifugation in CsCl gradients. Prior to transcription, they were linearized by the restriction enzymes immediately downstream from the poly(A) sequences in the genomes. RNAs were synthesized by SP6 RNA polymerase (Invitrogen) or T7 RNA polymerase (Epicentre) (in case of the TC-83-coding plasmid) in the presence of cap analog. The yield and integrity of transcripts were monitored by gel electrophoresis in nondenaturing conditions. For electroporation, aliquots of transcription reactions were used without additional purification, and RNAs were transfected into BHK-21 cells using previously described conditions . Viruses were harvested after cytopathic effect (CPE) development, usually 24 h postelectroporation. Viral replication analysis. | Cells were seeded at a concentration of 5 x 105 cells/35-mm-diameter dish. After 4 h of incubation at 37C, monolayers were infected at an appropriate multiplicity of infection (MOI) (as indicated in figure legends) for 1 h at 37C, and then the inoculum was replaced with 1 ml of complete medium. At selected times postinfection, media were replaced by fresh media, and titers of virus in the harvested samples were determined by plaque assay on BHK-21 cells . RNA analysis. | The protocol used for labeling SINV-specific RNAs with [3H]uridine in the presence of dactinomycin is described in the legend to Fig. . RNAs were isolated from the cells by using TRIzol reagent, as recommended by the manufacturer (Gibco-BRL, Bethesda, Md.), denatured with glyoxal in dimethyl sulfoxide, and analyzed by agarose gel electrophoresis using the previously described conditions . Analysis of protein synthesis. | BHK-21 cells were seeded into 6-well dishes at a concentration of 5 x 105 cells/well. Four hours later, the cells were infected with different viruses at an MOI of 10 PFU/cell in 200 mul of Alpha MEM supplemented with 1% FBS at room temperature for 1 h. Media were then replaced with Alpha MEM containing 10% FBS, and incubation was continued at 37C. At 12 h postinfection, the cells were washed three times with phosphate-buffered saline (PBS) and then incubated for 30 min at 37C in 1 ml of RPMI medium lacking methionine, supplemented with 0.1% FBS and 20 muCi of [35S]methionine. After the incubation, cells were scraped from the dish into PBS, pelleted by centrifugation, and dissolved in 200 mul of standard loading buffer. Aliquots of the samples were analyzed on sodium dodecyl sulfate-10% polyacrylamide gels. After electrophoresis, gels were dried and analyzed by autoradiography. Immunization and challenge with virulent VEEV. | Weanling female 6-day-old mice were inoculated s.c. with VEE TC-83 or SIN-83 viruses at doses of 103, 104, 105 or 106 PFU in a total volume of 100 mul of PBS. After vaccination, each cohort of 10 animals was maintained for 21 days without any manipulation. Immunized animals were then challenged with VEEV subtype ID strain ZPC738 (half of the animals) or subtype IC strain SH3 (half of the animals) by subcutaneous injection into the medial thigh with 106 PFU (ca. 106 50% lethal doses [LD50]) per animal in 0.1 ml of PBS. Five nonvaccinated, 4-week-old mice were challenged in the same manner at the same time, with each strain. Mice were observed twice daily for clinical illness (ruffled, depressed, anorexia, and/or paralysis) and/or death. Other groups of mice were vaccinated with the described above doses of VEE TC-83 or SIN-83 viruses and kept for 28 days postvaccination for final bleeding that was performed, after isoflurane narcosis, using cardiac puncture. Collected serum was heat inactivated at 56C for 30 min and kept frozen at -70C until further serological tests. Antibody assays. | Plaque reduction neutralization tests were performed on BHK-21 monolayers. A stock of VEEV TC-83 was incubated for 1 h at 37C with different dilutions of sera taken individually from mice infected with VEE TC-83 or SIN-83 viruses. Then cell monolayers were incubated with a virus-serum mixture for 1 h at 37C, overlaid with 0.5% agarose, maintained for 36 h at 37C, and stained with crystal violet. Comparative virulence of VEE TC-83 and SIN-83. | Six-day-old mice (male and female) were inoculated i.c. or s.c. with 2 x 104, 2 x 105, or 2 x 106 PFU of VEE TC-83 and SIN-83 viruses. Final volumes per animal were 20 mul (in PBS) for i.c. and 100 mul for the s.c. inoculation route. Animals were monitored twice daily as described previously for at least 2 months. Long-term pathogenicity study. | Six-day-old NIH Swiss mice (male and female) were inoculated s.c. into the medial thigh with 2 x 106 PFU of VEEV TC-83 or with SIN-83 virus (in 100 mul of PBS); 10 animals were used per virus, and 6 noninfected animals of the same age were included as controls. The body weight of each animal was determined weekly for 6 weeks, starting on day 7 of age (day 1 postinfection). Weights of animals that died were excluded from the study. RESULTS : Recombinant SIN-83 virus. | Using previously generated infectious cDNA copies of the SINV and VEEV genomes, we designed a recombinant virus named SIN-83 and tested its pathogenicity and ability to induce a protective immune response against VEEV challenge. The cis-acting RNA elements of the recombinant genome, which are required for replication and transcription of the subgenomic RNA (5' UTR, 3' UTR, and the subgenomic promoter), were derived from SINV (see Materials and Methods). All of the nonstructural protein genes were SINV specific as well. Structural genes and the 5' UTR of the subgenomic RNA were acquired from the VEEV TC-83 genome. This strategy of virus design was developed to maintain the optimal combinations of (i) factors essential for RNA replication, which include replicative enzymes and recognized RNA sequences, and (ii) factors required for efficient translation of the subgenomic RNA (the sequence and secondary structure of the 26S 5' UTR). The promoter element, located upstream of the subgenomic RNA transcription start, and the four 5'terminal nucleotides of the subgenomic RNA were designed to be SINV specific, because they represent the end of nsP4 and the termination codon of the nsP-coding open reading frame. The additional C->T mutation was introduced into position 24 of the 26S 5' UTR to compensate for the G->A mutation in position 4 and to maintain the computer-predicted 5' terminal secondary structure of the SIN-83S subgenomic RNA close to that of the VEEV subgenomic RNA. The in vitro-synthesized RNAs of VEE TC-83, SIN-83S (the initial construct), and SIN Toto1101 viruses were electroporated into BHK-21 cells. All of the RNAs had similar infectivity (data not shown), and rescued viruses were harvested 24 h later and used for all subsequent experiments. By that time, all cells exhibited profound cytopathic effects (CPE). We noticed that the original SIN-83S virus, recovered after RNA transfection, initially had a small-plaque phenotype that was unstable. The large plaques developed after one passage and rapidly became dominant. The entire viral genome from one large plaque-forming isolate (SIN-83 lp) was sequenced, and the only difference with respect to the original sequence was identified at position 4065, leading to a S795->T mutation in nsP2. This mutation was introduced into the pSIN-83S plasmid, which was then named pSIN-83. The SIN-83 virus, rescued from this cDNA, had a large-plaque phenotype and was indistinguishable from the variant SIN-83 lp, based on plaque morphology and replication kinetics . It also replicated at a significantly higher rate and to higher titers than the original SIN-83S virus . This SIN-83 variant was used for the next experiments. Replication of SIN-83 recombinant virus. | In the initial experiments, we compared the basic replication characteristics of the rescued recombinant SIN-83 virus with those of VEEV TC-83 and SINV Toto1101. We analyzed the ability of all three viruses to replicate in commonly used cell lines and the synthesis of virus-specific RNAs and proteins during viral replication. All three viruses were able to replicate in both BHK-21 and Vero cells (Fig. and B) with similar efficiencies, but they caused very different morphological changes in the cells during CPE development. Besides, the SINV Toto1101-infected Vero and BHK-21 cells died noticeably faster than those infected with VEEV TC-83. However, the release of SINV Toto1101 was slower, particularly in Vero cells, and final titers in the media were always substantially lower compared to those of VEEV TC-83. SIN-83 and SINV Toto1101 developed CPE at similar rates and with comparable changes in cell morphology. The chimeric virus replicated better than SINV, but not to the same titers as VEEV TC-83 (Fig. and B). Analysis of viral protein synthesis demonstrated that the SIN-83 virus efficiently expressed the structural proteins of the same molecular weight as those of VEEV TC-83 . In pulse-labeling experiments, the presence of the additional protein bands corresponding to SINV nsP2 and P123 was detected in lysate of SIN-83 virus-infected cells. This could be an indication that packaging of the chimeric genome into nucleocapsids was probably less efficient than packaging of SINV and VEEV genomes by homologous structural proteins. Examination of RNA replication and transcription showed that all of the cis-acting elements in the genome of the recombinant SIN-83 virus were fully functional , and the subgenomic RNAs demonstrated expected differences in sizes: 4,106, 3,915, and 4,171 nt [without poly(A) tail] for SIN Toto1101, VEE TC-83, and SIN-83 viruses, respectively. In multiple experiments, the rates of VEEV TC-83 genome replication and synthesis of the subgenomic RNA were always lower than those of SINV Toto1101 and SIN-83. This less efficient RNA replication level was in agreement with slower CPE development but contradicted the rates of virus release. Immunization with SIN-83 virus protects mice against VEEV challenge. | To compare the immunogenicity of the recombinant SIN-83 and VEE TC-83 viruses, 6-day-old NIH Swiss mice were infected s.c. with both viruses at doses ranging from 103 to 106 PFU (see Materials and Methods for details). After 28 days, we detected VEEV-specific neutralizing antibodies in the sera of both groups , although the titers in VEEV TC-83-immunized animals were higher. This was not surprising, considering the higher replication levels of this virus in cell culture. Other immunized animals were challenged with 106 PFU (similar106 LD50) of VEEV strain ZPC738 (subtype ID) or SH3 (subtype IC). All of the mice survived, regardless of the doses of SIN-83 and VEE TC-83 viruses used for immunization (103, 104, 105, and 106 PFU). After challenge, none of the animals developed any detectable clinical symptoms of VEE for more than 2 months, the entire period of observation. All of the mice from the nonimmunized control group were hunched and less active on day 3 or 4 postinfection with either the ZPC738 strain or the SH3 strain of VEEV. Between days 4 and 7, the nonimmunized mice developed anorexia, lethargy, and hind limb paralysis, and the CNS symptoms always progressed to stupor and finally to coma. Except for one animal infected with the SH3 strain (death on day 8 postinfection), all animals died by day 7. The cause of death was confirmed by isolating VEEV from the brains. The results of these experiments indicated that the chimeric SIN-83 virus could replicate in vivo and generate an effective immune response against VEEV challenge, with an efficiency similar to that of the VEEV vaccine strain TC-83. Comparative virulence of TC-83 versus SIN-83. | To compare the virulence of the VEE TC-83 and SIN-83 viruses, 6-day-old mice were inoculated i.c. or s.c. with different doses of each virus ranging from 2 x 104 to 2 x 106 PFU (see Materials and Methods). As shown in Fig. , VEEV TC-83 was virulent for weanling mice regardless of the inoculation route. The i.c. injection of 2 x 106 PFU of VEEV TC-83 caused 100% mortality. A 10-fold-lower dose was also uniformly lethal, but survival time was noticeably longer. A lower mortality level was observed in the TC-83 group that received 2 x 104 PFU i.c., and only one animal died. However, it should be noted that CNS sequelae, including ataxia and/or paralysis, developed in seven out of nine mice infected i.c. with 2 x 104 PFU of VEEV TC-83. Animals surviving the i.c. inoculation were also much smaller when compared to the animals of the same age from the SIN-83 virus-inoculated group (data not shown). None of the animals inoculated with SIN-83 virus died or developed any CNS signs or sequelae for at least 21 days postinfection. VEEV TC-83 was less pathogenic for weanling mice after s.c. inoculation (mortality rate, 10 to 20%), but many of the surviving animals developed clinical disease and/or CNS sequelae (three of eight for 106 PFU inoculum, three of nine for 105, and four of nine for 104 PFU). None of the SIN-83-inoculated animals had detectable clinical illness. We also monitored the body weights of mice for 6 weeks after s.c. inoculation with 2 x 106 PFU of either VEE TC-83 or SIN-83 virus . Animals infected with VEEV TC-83 at the age of 6 days were highly inhibited in their growth compared to those infected with SIN-83 or compared to the noninfected control group of the same age. The most pronounced difference in body weight was between 2 and 4 weeks postinfection. At the age of 6 weeks (5 weeks postinfection), the difference became insignificant, indicating that mice were able to compensate for the initial body weight loss. SIN-83-infected mice did not exhibit any weight loss compared to the uninfected control group. DISCUSSION : Recent VEE epidemics and equine epizootics, as well as the novel threat from bioterrorism, have underscored the need for a safe and effective vaccine. Previously, the attenuated VEEV TC-83 (United States) strain and VEEV strain 230 (former Union of Soviet Socialist Republics) were developed by serial passaging of the VEEV TrD in guinea pig heart cells (83 passages) and 230 passages in embryonated eggs, respectively. The main disadvantage of both viruses is the significant residual pathogenicity. For example, vaccination with VEEV TC-83 is associated with adverse effects in up to 40% of people . However, both strains were used for decades for protection of laboratory personnel working with VEEV. Viral pathogenesis appears to be determined by multiple factors that include the structure of the glycoprotein spikes on the surface of virions and the ability to replicate in specific tissues and interfere with the innate antiviral response at the cellular and organismal levels. Thus, mutations in the nonstructural proteins and cis-acting elements of viral genomes can play a significant role in virus attenuation. Attenuating mutations in genome regions other than those encoding the structural proteins have been identified for poliovirus , yellow fever virus , VEEV , Rift Valley fever virus , Semliki Forest virus, , and SINV , among others. However, despite a great deal of accumulated data, our knowledge about the roles of the nonstructural proteins in viral replication and interaction with host cells is far from complete. This is certainly true for VEEV as well, and the strategy of further attenuation of VEEV with point mutations in the genome does not look completely predictable. In addition, alphaviruses are known for their very high mutation rates, and the stability of the introduced attenuating point mutations during virus propagation will always be a great concern . In the present study, we attempted to develop a new, safer, attenuated alphavirus that can elicit a protective immune response against VEEV infection. The rationale was to create a chimeric viral genome that encodes the following: (i) the replicative enzymes and other RNA elements essential for replication derived from SINV, one of the least human-pathogenic alphaviruses, and (ii) structural proteins derived from the attenuated TC-83 strain of VEEV. The idea of creating attenuated chimeric viruses is not new; it was successfully applied in the development of vaccines against Japanese encephalitis and dengue viruses, using the 17D attenuated yellow fever virus genome as a vector . Chimeric tick-borne encephalitis/dengue type 4 viruses were designed for vaccine development as well . Recombinants between different alphaviruses were also described previously. (i) Western equine encephalitis virus (WEEV) is believed to be a natural recombinant between ancestral EEE-like and SIN-like viruses . (ii) Chimeras between distant alphaviruses, SINV and Ross River virus (RRV), and the more closely related eastern equine encephalitis virus and WEEV were found to be viable as well . SIN/RRV chimeric viruses could grow to high titers and were mainly applied to the study of interactions between structural and nonstructural regions in alphavirus genomes. However, they were attenuated compared to the original SINV and RRV strains, and based on this, we expected a significant level of attenuation from SIN/VEE recombinants. The initially constructed SIN/VEE chimera, SIN-83S, was designed based on SINV Toto1101 and VEEV TC-83 sequences and demonstrated heterogeneity in plaque morphology, which indicated that the virus was capable of additional adaptation, resulting in higher replication rates. In the plaque-purified variant that we used for sequencing, the large-plaque phenotype and more efficient replication were due to a single point mutation in nsP2, S795->T. However, we cannot rule out the possibility that other variants forming large plaques could have different mutations. Because it was not the goal of the present study, we did not explore the possibility of different mutations and their potential role in recognizing the changed subgenomic promoter or in other functions, but this is now under investigation. The redesigned chimera named SIN-83, containing the S795->T mutation in nsP2, was stable and could replicate to titers similar to those of the parental viruses, SINV Toto1101 and VEEV TC-83, in commonly used cell lines . We did not detect alterations in the replication of the genomic RNA and transcription of the subgenomic 26S RNA . This was an indication that all of the cis-acting RNA elements functioned properly. VEEV structural proteins were expressed in the cells infected with SIN-83 as efficiently as in the cells infected with VEEV TC-83, suggesting that the subgenomic RNA of VEEV can have structural elements that promote its translation under conditions of translational shutoff, which occurs in both VEEV- and SINV-infected cells . This was a point of concern, because the 5'-terminal sequence of the VEE 26S RNA is different from the same fragment of other alphaviruses; it is 20 nt shorter than average, and there is no computer-predicted, stable stem-loop structure in the sequence coding the amino-terminal portion of the capsid protein. This structural element in SINV and Semliki Forest virus was previously shown to function as a translational enhancer that promotes translation of viral subgenomic RNA in the conditions of virus-specific translation inhibition . However, in the VEEV genome, this fragment remains very G-C rich and appears to function in the same fashion, probably together with the short 5' UTR, but this assumption needs further investigation. SIN-83 virus induced an immune response in weanling mice that was sufficient to protect them against high doses (106 LD50) of subtype ID and IC VEEV strains. The same protective effect was confirmed in adult (6 to 8 weeks old) mice (data not shown). The use of weanling mice allowed us not only to test the efficiency of the immune response caused by SIN-83 but also to compare the residual pathogenicity of this virus with that of VEEV TC-83. SIN-83 was significantly more attenuated; i.c. infection with 2 x 106 PFU did not produce any clinical symptoms. In contrast, the same doses of TC-83 were lethal, and virus was easily detected in mouse brains. However, we did not study if TC-83-induced mouse death was a result of additional mutations, making virus more pathogenic, or whether death resulted from residual pathogenicity. The s.c. administration of VEEV TC-83 was less lethal than i.c. inoculation. However, the high rate of CNS sequelae in animals surviving TC-83 infection, after either s.c. or i.c. inoculation, indicates that the virus is able to penetrate into the CNS and replicate efficiently after inoculation by either route. Virus clearance from the CNS in animals surviving challenge is probably associated with irreparable destruction of neuronal tissue, which is reflected in the high rate of CNS sequelae, such as ataxia and paralysis. The cause of these findings was not investigated further but could be a result of more-efficient virus replication in the brain, rather then of a higher neurotoxicity at the single cell level. Moreover, we believe that high titers of neutralizing antibodies in surviving animals infected with TC-83 probably reflect high levels of replication, leading to disease in weanling mice. The ability of TC-83 to induce clinical illness was previously described in mice and in equines and is sometimes comparable to that caused by enzootic VEEV strains . In contrast to VEEV TC-83 infection, animals vaccinated with SIN-83 virus did not develop CNS symptoms or sequelae, even upon i.c. inoculation, indicating very low pathogenicity of this chimeric virus. In the course of the experiments, we also noticed that prolonged alteration of the immune system or other adverse VEEV TC-83-induced effects during the most intensive phase of animal development led to a significant reduction in growth. This effect was detectable for several weeks postvaccination . We can only speculate about the mechanisms leading to slower growth of the animals; some of the possibilities include metabolic factors, such as the reduced energy recovery from food caused by bacterial overgrowth in the gut (e.g., because of the reduced level of immunoglobulin A), and endocrine/exocrine factors due to the direct viral or immune-mediated effects on the pancreas and/or CNS, or a combination of several different factors. Further pathogenesis studies with young mice are needed to understand the underlying mechanisms for these alterations in animal development. Importantly, SIN-83 did not cause any effect on mouse growth after s.c. inoculation of 2 x 106 PFU, at least. In conclusion, we have shown that a recombinant alphavirus with a genome encoding the entire replicative enzyme complex of SINV and structural proteins of VEEV can replicate efficiently in vitro and is highly attenuated in vivo. Our data strongly suggest that the recombinant SIN-83 virus is more attenuated than the TC-83 vaccine strain of VEEV. However, further studies are needed to determine its safety and the efficacy of the immune response. The use of the SINV genome as a vector for expression of structural proteins derived from more-pathogenic, encephalitis alphaviruses, including VEEV, eastern equine encephalitis virus, and WEEV, appears promising for further alphavirus vaccine development. FIG. 1. : Schematic representation of the recombinant SIN-83S virus genome and the junction between SINV and VEEV sequences in the region of the subgenomic promoter. Schematic representation of the recombinant SIN-83S virus genome and the junction between SINV and VEEV sequences in the region of the subgenomic promoter. Bold letters represent SINV-specific sequence, and underlined letters represent VEEV-specific sequence. Lowercase letters indicate mutations introduced into the VEEV sequence to make the junction more Sindbis-like and to preserve the putative secondary structure of the 5' UTR in the VEEV subgenomic RNA. FIG. 2. : Analysis of plaque morphology and virus replication in BHK-21 cells. Analysis of plaque morphology and virus replication in BHK-21 cells. (A) The originally designed SIN-83S virus, stock prepared from plaque-purified virus, which formed large plaques, designated SIN-83 lp, and SIN-83 virus containing the S795->T mutation in nsP2 found in SIN-83 lp were titrated on monolayers of BHK-21 cells. Plaques were allowed to develop for 36 h prior to fixation and staining. (B) BHK-21 cells were infected with SIN-83S, SIN-83 lp, and SIN-83 viruses at an MOI of 5 PFU/cell. At the indicated times, media were replaced, and virus titers were determined as described in the Materials and Methods. The experiment was performed twice with very similar results. FIG. 3. : Analysis of virus replication in BHK-21 and Vero cells. Analysis of virus replication in BHK-21 and Vero cells. BHK-21 (A) or Vero (B) cells were infected with VEE TC-83, SIN-83, and SIN Toto1101 viruses at an MOI of 1 PFU/cell. At the indicated times, media were replaced, and virus titers were determined as described in the Materials and Methods. The experiments were also performed at MOI of 0.1 and 10 PFU/cell with very similar differences in virus growth (data not shown). FIG. 4. : Analysis of protein synthesis and replication of virus-specific RNAs in infected cells. Analysis of protein synthesis and replication of virus-specific RNAs in infected cells. BHK-21 cells were infected with VEE TC-83, SIN-83, and SIN Toto 1101 viruses at an MOI of 10 PFU/cell. (A) At 12 h postinfection, the cells were labeled with [35S]methionine and analyzed on a sodium dodecyl sulfate-10% polyacrylamide gel as described in Materials and Methods. The additional 35S-labeled protein bands in the lysate of SIN-83-infected cells are marked by stars. (B) At 2 h postinfection, medium in the wells was replaced by 1 ml of Alpha MEM supplemented with 10% FBS, dactinomycin (1 mug/ml), and [3H]uridine (20 muCi/ml). After 3 h of incubation at 37C, RNAs were isolated from the cells and analyzed by agarose gel electrophoresis as described in Materials and Methods. FIG. 5. : Survival of mice infected with VEE TC-83 and SIN-83 viruses. Survival of mice infected with VEE TC-83 and SIN-83 viruses. Six-day-old NIH Swiss mice were inoculated i.c. (A) or s.c. (B) with indicated doses of viruses. Animals were monitored for 2 months. No deaths occurred after day 7 postinfection in any of these experiments. FIG. 6. : Mouse body weights after infection with VEE TC-83 or SIN-83 viruses. Mouse body weights after infection with VEE TC-83 or SIN-83 viruses. Six-day-old NIH Swiss mice were inoculated s.c. with 2 x 106 PFU of VEE TC-83 or SIN-83 viruses. Ten animals were used per virus, and six noninfected animals of the same age were included as a control group. The body weight of each animal was determined weekly. Weights of the VEEV TC-83-infected mice that died were excluded from the study. Error bars indicate standard deviations. TABLE 1 : Neutralizing antibody titers in mice infected with different doses of VEE TC-83 and SIN-83 viruses Backmatter: PMID- 12915548 TI - Transforming Growth Factor beta1 Receptor II Is Downregulated by E1A in Adenovirus-Infected Cells AB - Transforming growth factor beta1 (TGF-beta1) signaling is compromised in many tumors, thereby allowing the tumor to escape the growth-inhibitory and proapoptotic activities of the cytokine. Human adenoviruses interfere with a number of cellular pathways involved in cell cycle regulation and apoptosis, initially placing the cell in a "tumor-like" state by forcing quiescent cells into the cell cycle and also inhibiting apoptosis. We report that adenovirus-infected cells resemble tumor cells in that TGF-beta1 signaling is inhibited. The levels of TGF-beta1 receptor II (TbetaRII) in adenovirus-infected cells were decreased, and this decrease was mapped, by using virus mutants, to the E1A gene and to amino acids 2 to 36 and the C-terminal binding protein binding site in the E1A protein. The decrease in the TbetaRII protein was accompanied by a decrease in TbetaRII mRNA. The decrease in TbetaRII protein levels in adenovirus-infected cells was greater than the decrease in TbetaRII mRNA, suggesting that downregulation of the TbetaRII protein may occur through more than one mechanism. Surprisingly in this context, the half-lives of the TbetaRII protein in infected and uninfected cells were similar. TGF-beta1 signaling was compromised in cells infected with wild-type adenovirus, as measured with 3TP-lux, a TGF-beta-sensitive reporter plasmid expressing luciferase. Adenovirus mutants deficient in TbetaRII downregulation did not inhibit TGF-beta1 signaling. TGF-beta1 pretreatment reduced the relative abundance of adenovirus structural proteins in infected cells, an effect that was potentiated when cells were infected with mutants incapable of modulating the TGF-beta signaling pathway. These results raise the possibility that inhibition of TGF-beta signaling by E1A is a means by which adenovirus counters the antiviral defenses of the host. Keywords: Introduction : Transforming growth factor beta1 (TGF-beta1) is a prototype member of a family of multifunctional cytokines . Originally discovered as a fibroblast growth factor, TGF-beta1 was soon found to play an important role in a variety of physiological processes including immunoregulation, the cell cycle, apoptosis, and formation of the extracellular matrix . In epithelial cells, TGF-beta1 negatively affects the cell cycle primarily through transcriptional upregulation of cyclin-dependent kinase inhibitors . In the immune system, TGF-beta1 along with interleukin-10 functions to control and limit the extent of the adaptive immune response . The signaling pathways for all members of the TGF-beta1 family are similar . Intracellular signaling is initiated upon the binding of the active cytokine to the TGF-beta receptor II (TbetaRII) homodimer and the assembly of a heterotetrameric complex consisting of receptors I and II. TbetaRII is a ubiquitously expressed constitutively active serine/threonine kinase . Once the heterotetrameric receptor complex is formed, TbetaRII phosphorylates TbetaRI and thereby greatly enhances TbetaRI serine/threonine kinase activity. The Smad family of proteins includes secondary mediators of TGF-beta signaling . Receptor-specific Smads that are phosphorylated by activated TbetaRI associate with Smad 4 and other factors to form a transcriptionally competent complex that enters the nucleus and modulates gene expression. The TGF-beta1 signaling pathway is inactivated in many tumors, presumably allowing the tumors to escape TGF-beta1-mediated growth inhibition and apoptosis . Frequently, inhibition of TGF-beta1 signaling occurs by either abolition of the function of a common mediator, Smad 4, or by interference with TbetaRII function. Some of the reported mechanisms of TbetaRII downregulation include inhibition of promoter activity , decrease in mRNA stability , and intracellular retention . Human adenovirus causes a number of benign diseases and may establish persistency in lymphoid cells . Quiescent epithelial cells are believed to be the main target of acutely replicating adenovirus in vivo. Infection is divided into two stages, early and late. Early genes begin to be expressed prior to viral DNA replication and encode proteins that usurp the cell . Progression into the late stage of infection and successful completion of the viral life cycle require replication of the viral DNA genome. The cellular DNA synthesis machinery may facilitate viral genome replication; upon infection, cell cycle perturbations in quiescent epithelial cells, primarily due to the adenovirus E1A proteins, are observed . In addition, the adenovirus genome encodes a number of proteins that counteract host cell apoptosis, whether it is induced by unscheduled cell cycle progression or mediated by the immune system . Because of unscheduled entry into the cell cycle and protection against apoptosis, adenovirus-infected cells are forced into a tumor-like state. Increased levels of active TGF-beta1 are created at the site of inflammation through the release of active cytokines by a subpopulation of macrophages and regulatory T cells and by local activation of extracellular matrix-associated latent complexes . It seems possible that the growth-inhibitory and proapoptotic functions of TGF-beta1 would be both inhibitory to viral DNA replication and detrimental to the survival of infected epithelial cells. In addition or alternatively, the immunoregulatory function of TGF-beta1 could diminish the antiadenovirus immune response and accordingly may be beneficial for in vivo adenovirus replication and possibly persistence. Therefore, it is reasonable to consider whether disruption of the TGF-beta1 signaling pathway might occur in adenovirus-infected cells. Here we report that adenovirus mediates a decrease in TbetaRII protein levels and that the E1A proteins are responsible for the effect. This decrease is accompanied by a reduction in steady-state TbetaRII mRNA levels. TGF-beta1-mediated signaling in infected cells is inhibited; adenovirus mutants that fail to downregulate TbetaRII do not inhibit TGF-beta1-mediated signaling. Finally, activation of the TGF-beta1 pathway decreases the abundance of adenovirus structural proteins in infected cells. MATERIALS AND METHODS : Cell lines. | Human A549 lung adenocarcinoma and human HepG2 hepatocellular carcinoma cell lines were purchased from the American Type Culture Collection. A549 cells were grown in Dulbecco's modified Eagle medium (DMEM; JRH Biosciences, Lenexa, Kans.) supplemented with 10% fetal calf serum (FCS; HyClone, Logan, Utah). HepG2 cells were grown in DMEM-F12 medium supplemented with 10% FCS. Adenovirus mutants. | Ad2, Ad5, and rec700 were used as wild-type controls. rec700 is a recombinant virus derived from Ad2 and Ad5 with an E1A region from Ad5 . E1A.2-36 and E1A.81-120 are adenovirus mutants that lack amino acids 2 to 36 and 81 to 120, respectively, in the E1A proteins (kindly provided by Elizabeth Moran, Temple University). pm975 and 12Swt express only the 13S and 12S E1A isoforms, respectively . dl808 lacks the E4 region (deletion encompasses map units 92.0 to 97.1) . dl7001 and dl327 lack the entire E3 region, with the exception of the gene for E3-12.5K present in dl327 . dl764 and dl753, mutants, derived from rec700, lack the RIDbeta and RIDalpha genes, respectively . 12S.2-36 and 12S.928 are 12Swt-based mutants with an E1A N terminus deletion and a point mutation, respectively. dl313 expresses E1A proteins lacking amino acids 220 to 289. dl312 lacks the E1A region . Ad/E3 is an E1-negative replication-defective adenovirus vector expressing E3 proteins under the control of the cytomegalovirus (CMV) promoter . 176-9 (dC-term) is a 12S mutant with a deletion of amino acids 224 to 284 in the E1A protein . dl118 is an E1B deletion mutant . Mutants E1A.2-36, E1A.81-120, pm975, 12Swt, 12S.2-36, 12S.928, dl313, dl312, and dl118 are all in a dl309 genetic background. dl309 is an Ad5 mutant that lacks the genes for the E3 RIDalpha, RIDbeta, and 14.7K proteins. Antibodies and reagents. | Rabbit anti-TbetaRII antibodies used for Western analysis were purchased from Santa Cruz Biotechnology (catalog numbers sc-400 and sc-220; Santa Cruz, Calif.). Affinity-purified goat anti-TbetaRII raised against an extracellular receptor domain was purchased from R&D Systems (catalog number AF-241-NA; Minneapolis, Minn.) and used for immunoprecipitations. A rabbit anti-Ad5 antibody was purchased from the American Type Culture Collection. The anti-phospho-Smad 2 antibody was from Upstate Biotechnology (Charlottesville, Va.). Recombinant human TGF-beta1 was purchased from R&D Systems; it was reconstituted and stored in accordance with the manufacturer's instructions. Western analysis. | HepG2 or A549 cells were washed three times with phosphate-buffered saline and lysed on ice in radioimmunoprecipitation assay buffer (50 mM Tris [pH 8.0], 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS], 1 mM EDTA, 1 mM NaF) for 30 min. Aspirated buffer was centrifuged at 12,000 rpm in an Eppendorf centrifuge (model 5415C) for 10 min, and the protein concentration of the supernatant was measured by the Bio-RadDC protein assay (Bio-Rad Laboratories, Hercules, Calif.). Equal protein amounts of 50 (TbetaRII) or 30 mug (adenovirus late proteins) were loaded into each lane of the SDS-8% polyacrylamide gel electrophoresis (PAGE) gel. Proteins were transferred to a polyvinylidene difluoride membrane (Immobilon-P; Millipore, Bedford, Mass.) and incubated with a combination of anti-TbetaRII antibodies (1:300 dilution each) or the anti-Ad5 antibody (1:1,600 dilution). Following application of the secondary horseradish peroxidase-conjugated antibody and subsequent washes, a signal was generated with a commercial chemiluminescence substrate (LumiGLO; KPL, Gaithersburg, Md.). The signal was detected by autoradiography and quantified by densitometry using FluorChem software (Alpha Innotech Corporation, San Leandro, Calif.). Analysis of late protein synthesis. | A549 cells were maintained under 10 or 0.2% FCS or 0.2% FCS plus 5 ng of TGF-beta1/ml for 3 days prior to infection and throughout the infection. During 3 days of pretreatment, medium was refreshed once to sustain the activity of the cytokine. Fresh medium was also supplied at the beginning of the infection. RPA. | HepG2 cells were mock infected or infected with adenovirus mutants and maintained in the presence of AraC. Total RNA was isolated from infected and mock-infected HepG2 cells with TRIzol reagent (Invitrogen, Carlsbad, Calif.) according to the manufacturer's instructions. Radioactively labeled RNA antisense probes were generated by the Riboprobe Combination System Sp6/T7 (Promega, Madison, Wis.). The TbetaRII probe was synthesized with T7 polymerase from a pBluescript TbetaRII probe plasmid linearized with HindIII. To generate the pBluescript TbetaRII probe, a HindIII-PstI TbetaRII fragment recovered from the mycTbetaRII plasmid (kindly provided by Yoav Henis, Tel Aviv University) was cloned into the pBluescript (-) multiple cloning site. The GAPDH (glyceraldehyde-3-phosphate dehydrogenase) probe was synthesized with Sp6 polymerase from an XbaI-linearized template (kindly provided by Rob Fleming, Saint Louis University). Thirty (TbetaRII) or 6 (GAPDH) mug of total RNA was hybridized to the freshly made radioactively labeled probes and processed with a commercially available RNase protection assay (RPA) kit (RPAIII; Ambion, Dallas, Tex.) according to the manufacturer's instructions. Immunoprecipitations. | A549 cells were mock infected or infected with rec700 at a multiplicity of infection of 50 PFU/cell and maintained in the presence of 20 mug of AraC/ml. At 18.5 h postinfection (p.i.), cells were washed and incubated with DMEM lacking cysteine and methionine for 30 min. At 19 h p.i., the preincubation medium was changed to fresh cysteine- and methionine-negative DMEM supplemented with 100 muCi of a [35S]methionine-cysteine mixture (EasyTag Express protein labeling mixture; Perkin-Elmer, Boston, Mass.)/ml. After 15 min of incubation in radioactive medium, cells were either immediately placed on ice (0-h chase) or washed three times with prewarmed cold DMEM and incubated in DMEM containing 10% FCS for the time period shown in Fig. . At the end of a chase period, the cells were placed on ice, washed four times with ice-cold phosphate-buffered saline, and scraped into 100 mul of 0.5% NP-40 lysis buffer (20 mM sodium phosphate, 250 mM NaCl, 30 mM sodium pyrophosphate, 5 mM EDTA, 10 mM NaF). Cell lysates were normalized to the amount of protein in the lysate and combined with 0.25 mug of the antibody and 20 mul of 50% protein G-agarose (Roche, Indianapolis, Ind.). Following overnight incubation, the agarose beads were washed with 0.5% SDS lysis buffer; immunoprecipitated proteins were dissociated from the agarose in 2x Laemmli buffer and separated on SDS-8% PAGE gels. Gels were dried, and the signal was visualized by autoradiography and phosphorimaging with a STORM phosphorimager, with subsequent quantitation by ImageQuant, version 4.0, software. Luciferase assay. | HepG2 cells seeded in 24-well dishes were transfected with 0.5 mug of 3TP-lux (kindly provided by J. Massague, Sloan-Kettering Institute) and 0.3 mug of pCMV-beta Gal plasmid. At 6 h posttransfection, HepG2 cells were mock infected or infected with rec700 at 50 PFU/cell. Infections were maintained in the presence of freshly supplied AraC at 20 mug/ml. At 12 h p.i., medium was replaced with serum-free medium containing 5 ng of human recombinant TGF-beta1 (R&D Systems)/ml. At 26 h p.i., cells were lysed, and luciferase activity was measured with a luminometer by using the luciferase assay system (Promega). To determine beta-Gal activity, the same volume of cell lysate was incubated in assay buffer (200 mM sodium phosphate [pH 7.3], 2 mM MgCl2, 100 mM beta-mercaptoethanol, 1.33 mg of o-nitrophenyl-beta-d-galactopyranoside [ONPG]/ml) and measured on a spectrophotometer at 405 nm. In place of infection, in some experiments cells were cotransfected with 0.5 mug of either 13S E1A or 12S E1A plasmids (kindly provided by G. Chinnadurai, Saint Louis University). RESULTS : TbetaRII is downregulated in the course of adenovirus infection. | Figure illustrates TbetaRII protein levels observed in human A549 cells infected with rec700, an Ad5-Ad2-Ad5 recombinant that functions as a wild-type adenovirus . Two forms of TbetaRII are shown: the mature form of the receptor and the short-lived TbetaRII precursor. (This precursor has been shown to be sensitive to endoglycosidase H digestion in vitro in extracts obtained from Mv1Lu cells ). TbetaRII levels in adenovirus-infected cells were significantly decreased ; in multiple experiments, only 5 to 20% of TbetaRII proteins remained at 19 h p.i. Blocking DNA synthesis with AraC maintains adenovirus infection in the early stage . As shown in Fig. (lane 3), TbetaRII levels in AraC-treated infected cells were decreased, indicating that the decrease is mediated by an adenovirus early protein. Similar results were obtained with HepG2 cells (data not shown). A time course of TbetaRII downregulation in adenovirus-infected cells is presented in Fig. . Only the mature form of TbetaRII is shown. The decrease in TbetaRII levels first became evident between 13 and 15 h p.i. E1A is responsible for the decrease in TbetaRII protein levels in adenovirus-infected cells. | A set of adenovirus mutants was used to map TbetaRII downregulation to an early protein function. Wild-type rec700, Ad2, and Ad5 were highly effective in decreasing TbetaRII levels (Fig. , lanes b, f, and g). Virus mutants lacking most of the E4 (dl808) or E1B (dl118) region were similar to wild-type virus in clearing TbetaRII (Fig. , lane e, and C, lane c). Thus, the E1B and E4 regions are not required to downregulate TbetaRII. Several mutants with lesions in the E3 region were examined inasmuch as E3 proteins downregulate a number of cell surface receptors, including epidermal growth factor receptor, Fas, and TRAIL receptors 1 and 2 . Somewhat surprisingly, deletion of all or nearly all of the E3 region (dl7001 and dl327; Fig. , lanes c and d) or deletion of the gene for RIDalpha (dl753; Fig. , lane b) or RIDbeta (dl764; Fig. , lane i) did not affect TbetaRII downregulation. In addition, expression of all E3 genes from the deleted E1 region of replication-defective adenovirus (Ad/E3) failed to remove TbetaRII at 2 days p.i. (Fig. , lane g). These results indicate that E3 proteins are not required to decrease TbetaRII levels. Given that downregulation of TbetaRII is an early function and that the early E1B, E3, and E4 regions are not required to mediate the decrease in TbetaRII, the immediate-early E1A region was examined. Mutant dl312 lacks most of the E1A region ; TbetaRII levels in cells infected with dl312 were not decreased (Fig. , lane e), establishing that E1A expression is required for TbetaRII downregulation. There are two interpretations of this result: first, the E1A protein itself could be responsible for TbetaRII downregulation; second, considering that E1A is required for efficient induction of adenovirus genes, an E1A-inducible viral protein(s) could be responsible. We favor the first interpretation because under the conditions of the experiment, namely, high multiplicity of infection (50 PFU/cell) and a long period p.i. (48 h), it is known that adenovirus early genes are expressed in an E1A-independent manner . The E1A gene is expressed as two major alternatively spliced 13S and 12S mRNAs, which encode proteins of 289 (289R) and 243 (243R) amino acids, respectively. A mutant expressing either the 13S or 12S isoform of E1A (pm975 or 12Swt, respectively) was effective in downregulating TbetaRII (Fig. , lane h, and C, lane e) (pm975 was not quite as effective as the wild type). The finding that 12S E1A alone is functional excludes a rigorous requirement for the E1A CR3 domain, which is present in the 289R protein but absent from the 243R protein, in suppressing TbetaRII levels. Interestingly, when adenoviruses with mutations in E1A were examined, a defect in TbetaRII downregulation was observed. Mutant dl313 was mostly defective in clearing TbetaRII (Fig. , lane d). Mutant dl313 expresses E1A proteins that are truncated at their C termini (deletion of amino acids 220 to 289), suggesting that the C-terminal portion of the E1A proteins is required for downregulation of TbetaRII. In accord with this suggestion, a phenotype similar to that of dl313 was observed in a 12Swt mutant (dC-term) with an E1A whose C terminus lacked amino acids 224 through 284 (including the C-terminal binding protein [CtBP] binding site) (Fig. , lane f). (Mutant dl313 also lacks the E1B region, but given that the E1B mutant dl118 was not defective in downregulating TbetaRII [Fig. , lane c], we presume that the defect is due to the deletion in the C-terminal region of E1A.) In addition to the C-terminal region of E1A, the N-terminal region is also important because a 12Swt mutant (12S.2-36) with an E1A lacking amino acids 2 to 36 was completely incapable of TbetaRII downregulation (Fig. , lane c). On the other hand, the absence of the pRB binding site in another 12Swt mutant (12S.928) had no effect on the efficiency of TbetaRII downregulation (Fig. , lane d). The importance of the N-terminal E1A domain for effective TbetaRII downregulation was further confirmed by examining TbetaRII levels in cells infected with E1A mutants that express both the 13S and 12S forms of E1A . Mutants E1A.2-36 and E1A.81-120 (derived from dl309) express mutant E1A proteins lacking amino acids 2 to 36 and 81 to 120, respectively. As expected, the E1A.2-36 mutant was unable to downregulate TbetaRII, whereas E1A.81-120 was as efficient as the parent virus. All infections were monitored in parallel for expression of E1A and the E2 DNA binding protein (DBP) by immunofluorescence; typically, 90 to 95% of cells stained positive for E1A when applicable, and 60 to 70% were DBP positive, with a homogenous pattern of staining in cells incubated with AraC (data not shown). We conclude that the 13S (289R) and 12S (243R) forms of E1A force downregulation of TbetaRII protein, that the N-terminal (amino acids 2 to 36) and C-terminal (amino acids 224 to 284) regions of the E1A proteins are required for TbetaRII downregulation, and that the pRb-binding region of the E1A protein is not required. Adenovirus's ability to decrease TbetaRII protein levels correlates with its ability to downregulate TbetaRII mRNA. | As a means to examine whether the decrease in TbetaRII protein levels was due to a decrease in TbetaRII mRNA levels, an RNase protection assay was employed to determine TbetaRII mRNA levels in mock-infected cells and cells infected with a limited set of E1A mutants. The densities of the RNase-protected fragments were analyzed with FluorChem software (Alpha Innotech Corporation) and are presented in the graphs under the corresponding autoradiography data (Fig. and B). The density of each TbetaRII fragment was normalized against the density of the corresponding GAPDH signal. Figure shows that the decrease in TbetaRII mRNA (seen as two RNase-protected bands) was most apparent in rec700-infected cells (compare lane e with lane d) (2.7- and 2.4-fold decreases in the TbetaRII signal compared to signals for mock infection and infection with E1A-negative mutant dl312, respectively). TbetaRII mRNA levels were also decreased in cells infected with a mutant expressing only the 13S or 12S isoform of E1A (pm975 or 12Swt, respectively; lanes f and g), but the decrease in TbetaRII mRNA due to these two mutants was consistently less than that due to the wild-type adenovirus (Fig. and B). TbetaRII mRNA levels in cells infected with dl312 (Fig. , lane h) were similar to those in mock-infected cells (lane d). Under the conditions of infection used (50 PFU/cell, 43 h p.i.), early and late virus genes are expressed by dl312 (; our unpublished observations). Therefore, it is likely that TbetaRII mRNAs are not downregulated by dl312 because it does not make the E1A protein although it makes other adenovirus proteins. Figure illustrates the TbetaRII mRNA levels with two 12Swt-based mutants. Mutant 12S.928 (Fig. , lane g), which is defective in binding pRB, decreased TbetaRII mRNA levels similarly to the parent virus (12Swt; Fig. , lane e), whereas 12S.2-36 had no effect on TbetaRII mRNA levels (Fig. , lane f). GAPDH mRNA, which was used as a control, was not affected by these viruses (Fig. and B, lanes i to m). Interestingly, a DeltaCtBP mutant that is partially defective in decreasing TbetaRII protein levels (Fig. , lane f) was as efficient as the parent 12Swt virus in decreasing the corresponding mRNA levels (Fig. , compare lanes c and d), suggesting that E1A second-exon functions, including encoding the CtBP binding domain, are dispensable for the TbetaRII mRNA decrease in infected cells. In conclusion, in accord with the results of the previous section, E1A proteins, and particularly the N-terminal amino acids of E1A, decrease TbetaRII mRNA in adenovirus-infected cells. Dynamics of the TbetaRII protein as determined by pulse-chase analysis are affected in adenovirus-infected cells. | There is a quantitative difference in steady-state TbetaRII mRNA and protein levels in infected cells: whereas TbetaRII protein levels were decreased 10-fold by 19 h p.i. , there was only a 2.4- to 2.7-fold decrease in steady-state TbetaRII mRNA levels at this time compared to levels in mock-infected cells . This difference was not due to the host cell shutoff that is observed during the late stages of infection inasmuch as experiments were done under conditions where infection was limited to the early phase. A TbetaRII protein pulse-chase experiment was performed to examine potential differences in the rates of synthesis and the half-lives of TbetaRII in adenovirus- and mock-infected cells. A549 cells were infected at 50 PFU/cell and maintained in freshly supplied AraC throughout the experiment. At 19 h p.i., cells were metabolically labeled with a [35S]cysteine-methionine mixture for 15 min and chased in cold medium (Fig. and D). TbetaRII was immunoprecipitated, resolved by SDS-PAGE, and detected by a STORM phosphorimager as well as autoradiography. The phosphorimager signal was quantified with ImageQuant software (Fig. and E). Representative autoradiograms and phosphorimager quantitations of the TbetaRII-specific signal from two independent experiments are shown in Fig. . Since a 2.4- to 2.7-fold decrease in TbetaRII mRNA was observed in infected cells, the decrease in the levels of TbetaRII protein in infected cells detected at 0 min of chase, compared to the level in mock-infected cells, should not be more than this, unless the translation step is affected. TbetaRII mRNA possesses extended 5' and 3' untranslated regions; therefore, translational regulation of TbetaRII abundance is conceivable. In multiple experiments (more than five), there was approximately a twofold difference in the TbetaRII between the adenovirus- and mock-infected cells at 0 min of chase (Fig. and D; quantitation in Fig. and E); this suggests that TbetaRII translation was not markedly affected by adenovirus infection. Curiously, for both the mock-infected and infected samples, there was an increase in detectable TbetaRII following the chase, peaking at about 30 min of chase. This increase was reproducible and was seen in A549 and HepG2 cells. Further, it occurred even when the chase was conducted in the presence of cycloheximide to inhibit the elongation of translation (data not shown). One explanation is that the antibody used does not efficiently detect the initial form of TbetaRII; it detects only a form that arises from posttranslational modifications. In any event, this increase in detectable TbetaRII was reproducibly lower in infected cells than in mock-infected cells; this is apparent in Fig. , in which the curves for the infected and mock-infected samples are superimposed. To calculate the TbetaRII half-life, data points obtained for both mock-infected and infected cells were fitted to an exponential equation with Excel. The fitness of the exponential trend was confirmed by the corresponding R2 values. Because of the delay in peak accumulation of TbetaRII, only data points taken from the 60- to 150-min chase interval were used to determine the TbetaRII half-life. A TbetaRII half-life of 40 min in mock-infected A549 cells was calculated . This calculated half-life is shorter than the reported TbetaRII half-life of 1.5 to 2 h in mink lung epithelial cells , which may be a species or cell line phenomenon. Importantly, no significant difference in TbetaRII half-life between the mock- and adenovirus-infected cells was found . At this point we cannot explain why, if the half-lives of the protein in mock- and adenovirus-infected cells are the same and the rate of precursor synthesis is decreased 2-fold, a 3-fold decrease in TbetaRII mRNA produces a 10-fold decrease in TbetaRII protein in infected cells, as determined by Western blotting. We have repeated pulse-chase and immunoprecipitation of TbetaRII using two additional antibodies raised against the N-terminal and C-terminal sequences of the protein; results were similar to those presented in the Fig. . In addition, we have tried two alternative lysis buffers: radioimmunoprecipitation assay buffer (used in a study by Wells et al. to immunoprecipitate TbetaRII from mink cells in a pulse-chase experiment) and Laemmli buffer (containing 1% SDS). Both methods of lysing cells yielded results similar to the one shown in Fig. . The events occurring during the initial TbetaRII processing in the infected cells that lead to decreased peak detection of TbetaRII (as shown in Fig. and D) could potentially contribute to TbetaRII downregulation. TGF-beta1 signaling in adenovirus-infected cells is inhibited. | Given that TbetaRII levels in infected cells are decreased and that the protein required for the decrease is E1A, then E1A should inhibit TGF-beta1-induced signaling. To examine this prediction, we used a transient-transfection system in which HepG2 cells were transfected with p3TP-lux, a TGF-beta1-responsive reporter plasmid , and pCMV-betagal as a transfection control. HepG2 cells were used because they had a higher efficiency of transfection than A549 cells. Subsequently, cells were either mock-infected or infected with the wild-type adenovirus or mutants. As shown in Fig. , upon treatment of mock-infected cells with 1 or 5 ng of active human recombinant TGF-beta1/ml, a robust activation of the signaling pathway occurred, as is evident from the increased luciferase activity. In contrast, very little activation of the TGF-beta1 signaling pathway occurred in rec700-infected cells, as illustrated by a minimal increase in luciferase production . Adenovirus mutants expressing only the 12S E1A wild-type or mutant isoform were analyzed for TGF-beta1 signaling. Signaling was inhibited in cells infected with 12Swt and the 12S mutant deficient in pRB binding (12S.928), but not in cells infected with the mutant harboring a deletion in the N terminus (12S.2-36) . These results are consistent with results in Fig. , which showed that TbetaRII protein levels were downregulated by 12Swt and by the 12S.928 mutant but not by the 12S.2-36 mutant. As a further check for the ability of E1A to block TGF-beta1 signaling, we cotransfected the p3TP-lux reporter with plasmids expressing the 13S and 12S forms of E1A. As shown in Fig. , both E1A isoforms reduced TGF-beta-induced luciferase synthesis, confirming a role for E1A in inhibiting TGF-beta1 signal transduction and indicating that E1A can suppress TGF-beta1 signaling in the absence of other adenovirus proteins. Finally, we examined the levels of phosphorylated Smad 2 in cells infected with rec700. Levels of phospho-Smad 2 in mock-infected cells increased upon TGF-beta1 treatment, as determined by Western analysis using the anti-phospho-Smad 2 antibody (Fig. , lanes a and b). Little Smad 2 phosphorylation was induced in rec700-infected cells at 20 h p.i. (Fig. , lane d), and no phosphorylated Smad 2 was detected in infected cells at 24 h p.i. (Fig. , lane e). The lack of TGF-beta1-inducible Smad 2 phosphorylation in infected cells suggests that TbetaRII downregulation by adenovirus contributes to the loss of TGF-beta1 signaling observed in infected cells. TGF-beta1 reduces the accumulation of adenovirus late proteins and virus yields. | Large amounts of structural proteins are produced during the late stage of infection, and expression is dependent on adenovirus genome replication. Synthesis of late proteins is followed by virion assembly and release. Therefore, the levels of structural proteins in infected cells can be used to evaluate progression through the viral life cycle. To determine whether TGF-beta1 signaling has an effect on adenovirus life cycle progression, the accumulation of adenovirus late proteins as well as E1A proteins in TGF-beta1 treated cells was examined. A549 cells undergo growth arrest, but not apoptosis, upon treatment with active TGF-beta1 (data not shown). In our experiments, prior to infection, A549 cells were maintained in 10 or 0.2% FCS or 0.2% FCS containing 5 ng of active TGF-beta1/ml for 3 days. Cells were subsequently infected at 35 PFU/cell with wild-type dl309 or mutant E1A.2-36 or E1A.81-120 adenoviruses. Infected cells were kept under conditions identical to the pretreatment environment. At 24 h p.i., levels of adenovirus structural and E1A proteins were examined by Western analysis using anti-Ad5 and anti-E1A antibodies, respectively. The relative abundances of structural and E1A proteins were quantified by densitometric analysis. In addition, the ratio of signal in serum-starved cells to that in TGF-beta1-treated cells was calculated for structural and E1A proteins (Fig. and B, ratios of lanes b to c, e to f, and h to i; summarized in Fig. ). Serum starvation had a negligible effect on the levels of structural proteins for all three viruses tested (Fig. , compare lanes a and b, d and e, and g and h). However, the levels of structural proteins in serum-starved cells maintained in the presence of TGF-beta1 were decreased (Fig. , lanes c, f, and i). This negative effect was more pronounced in cells infected with E1A.2-36, a mutant (similar to 12S.2-36) incapable of downregulating TbetaRII and inhibiting TGF-beta1 signaling . While the ratios of the amounts of structural proteins in serum-starved cells to the amounts in TGF-beta1-treated cells for dl309 and E1A.81-120 infections were very similar (1.9 and 1.74, respectively; Fig. ), the ratio was higher in cells infected with E1A.2-36 (3.43; Fig. ). These results indicate that, although TGF-beta1 pretreatment decreases adenovirus structural protein levels in infected cells, a greater effect is observed when the mutant cannot inhibit TGF-beta1 signaling. Although the levels of E1A proteins were somewhat lower in TGF-beta1-treated cells, the decreases in E1A for all adenovirus mutants were very similar. To assess the effect of TGF-beta1 treatment on virus yields, A549 cells were pretreated as described above and subsequently infected with 10 PFU of E1A.2-36 or E1A.81-120/cell. Following infection, pretreatment conditions were restored. Infected cells together with supernatants were collected at 0, 1, 2, and 3 days p.i. and freeze-thawed three times, and total virus yield was determined by plaque assay for each sample. E1A.2-36 and E1A.81-120 had similar growth kinetics in control cells but not in TGF-beta-treated cells . Although E1A.81-120 was as efficient as its wild-type parent virus in decreasing TbetaRII levels , it produced a decrease in late protein synthesis in TGF-beta-treated cells at 24 h p.i. that was similar to the decrease observed in cells infected with wild-type virus (Fig. and C). Accordingly, growth of E1A.81-120 was delayed in TGF-beta-treated cells compared to that in control cells (about a 1.5-log-unit difference at 24 h p.i. and about a 1-log-unit difference at 48 h p.i.; Fig. ). The E1A.2-36 mutant showed greater delays in growth at 24 and 48 h p.i. (2.5 and 2 log units, respectively), correlating with the mutant's inability to downregulate TbetaRII and to control TGF-beta signaling. DISCUSSION : TGF-beta1 is a critical factor in the homeostasis of the immune system. TGF-beta1 is produced by regulatory T cells and macrophages to control immune system activation . Viruses have evolved to prevent the immune system from premature destruction of infected cells and to limit the antiviral immune response upon infection. Additionally, a TGF-beta1-imposed G1/S block in the cell cycle and the induction of apoptosis would probably threaten completion of the viral life cycle. Therefore, it seems reasonable that modulation of TGF-beta1 production and activity would be executed by many infectious agents. Indeed, TGF-beta1 was reported to partially induce reactivation of latent Epstein-Barr virus . TGF-beta1 promoter activity was induced by the human papillomavirus E6 oncoprotein . TGF-beta1 secretion induced by HIV antigens was shown to account for inhibition of gamma interferon induction in response to human immunodeficiency virus . Finally, the human T-cell lymphotropic virus type 1 Tax protein was found to inhibit Smad-mediated signaling at the transcriptional level . Our data also support the idea of TGF-beta1 pathway modulation by an infectious agent. Our results show that TbetaRII is downregulated in adenovirus-infected cells. The decrease in TbetaRII becomes apparent at 13 to 15 h p.i. and is a function of E1A. A limited number of cellular genes undergo negative regulation by E1A, including genes encoding two surface receptors, neu and HER2 , major histocompatibility complex class I (MHC-I) genes , and some differentiation genes . MHC-I downregulation is mediated only by the Ad12 E1A (a member of adenovirus subgroup A) and not Ad5 E1A (subgroup C). Conversely, a number of cellular genes are positively regulated by E1A, including Golgi-associated GP73 as well as genes whose expression is controlled by E2F . According to our mapping studies, amino acids 2 to 36 in the E1A N terminus are particularly important for TbetaRII downregulation in infected cells. The CtBP binding site in the sequence encoded by the second exon also seems to be involved. On the other hand, the pRB binding site and the CR3 sequences are dispensable. The ability of adenovirus mutants with E1A lesions to downregulate TbetaRII protein levels correlates with the ability to decrease TbetaRII mRNA; the decrease in TbetaRII mRNA is likely due to transcriptional repression by E1A. The contribution of the E1A N-terminal-binding cellular factors to the TbetaRII decrease in infected cells is of considerable interest, as CBP/p300, TATA binding protein, and AP-1 all bind to E1A N-terminal sequences . The AP-1 complex, in particular, may play a role in positive regulation of the TbetaRII promoter . The ability of E1A to bind CtBP also plays a role in efficient TbetaRII downregulation in adenovirus-infected cells. CtBP, a transcriptional corepressor, belongs to a family of proteins with both nuclear and cytoplasmic functions . As a transcriptional repressor, CtBP has been shown to modulate TGF-beta1-mediated transcription. In addition, members of the CtBP family may play a role in Golgi regulation . Modulation of TGF-beta1 signaling by E1A has been studied by several groups in the context of cellular transformation. E1A was able to prevent TGF-beta1-induced upregulation of p21CIP, p15INK , and junB . Other groups showed that E1A could relieve TGF-beta1-induced growth inhibition, consistent with the role of E1A as a transforming agent . The ability to recruit p300 plays an important role in Smad-mediated transactivation , and E1A has been shown to displace p300 from Smad-containing complexes . The p300 binding site in E1A was shown to be required for E1A's ability to modulate TGF-beta1 signaling . In a transient-transfection system, E1A was able to interact with Smads 1, 2, and 3, suggesting that E1A may directly modulate the activity of Smad transcription complexes . Last, in a mouse cell line stably transfected with E1A, decreased TbetaRII protein and mRNA levels were observed along with diminished TbetaRII promoter activity . A 10-fold decrease in TbetaRII levels, as determined by Western blotting, was observed in adenovirus-infected cells at 19 h p.i. However, at the same time only an similar2.5-fold decrease in steady-state TbetaRII mRNA levels was seen, suggesting that some factor(s) in addition to the decrease in TbetaRII mRNA is responsible for the decrease in TbetaRII protein. This factor does not appear to be a decrease in the half-life of TbetaRII, because the half-lives in infected and uninfected cells were similar, as determined in a pulse-chase experiment in which radiolabeled TbetaRII was analyzed by immunoprecipitation and SDS-PAGE. Also, the factor is probably not a decrease in translation of TbetaRII because there was at most a twofold reduction due to infection in 35S-labeled TbetaRII detected by immunoprecipitation at the onset of the chase period and at least a twofold reduction would be expected because the TbetaRII mRNAs are less abundant in infected cells. The analysis is complicated by the fact that different antibodies were used for immunoprecipitation and Western blotting: the antibody used for immunoprecipitation (from R&D Systems) did not react with TbetaRII by Western blotting, and the antibody used for Western blotting (from Santa Cruz Biotechnology) produced a large background by immunoprecipitation. Nevertheless, we present the following working hypothesis to explain the quantitative differences. First, the initial rates of translation of TbetaRII in uninfected and infected cells are the same; the twofold decrease in TbetaRII seen in infected cells is due to the corresponding decrease in TbetaRII mRNA levels. Second, in uninfected cells, the TbetaRII precursor (form X) is processed to the higher-molecular-weight form (form Y), which is more readily detected by the antibody than form X, thereby accounting for the apparent increase in TbetaRII seen in the first similar30 min of chase. In infected cells, the development of form Y is inhibited by E1A or its degradation is accelerated. This would account for the decrease in the peak of TbetaRII observed in infected cells following 30 min of chase . Form Y is the species detected in uninfected cells by Western blotting, a method that measures the steady-state levels of TbetaRII. Since adenovirus inhibits the accumulation of TbetaRII, a 10-fold difference between infected and uninfected cells is observed by Western blotting. The antibody used in the immunoprecipitations was selected by the manufacturer for its ability to react with and neutralize cell surface TbetaRII (hypothetical form Y); therefore, the antibody may not have as high of an affinity for the misfolded or incompletely modified TbetaRII as it does for the mature cell surface receptor. TGF-beta1 signaling in adenovirus-infected cells was decreased, as determined by use of a TGF-beta-responsive reporter. Either the 13S or 12S E1A isoform was sufficient for inhibition of TGF-beta signaling in a transient-transfection system, as shown in this report and reports from other groups . Based on our findings and reports of others, there are at least three potential mechanisms of TGF-beta1 signaling inhibition in adenovirus-infected cells: (i) displacement of p300 from Smad-containing transcriptional complexes by E1A, (ii) direct modulation of Smad-mediated transcription by E1A, and (iii) E1A-mediated downregulation of TbetaRII levels. The ability of E1A to downregulate TbetaRII levels is dependent on the N-terminal sequences that are also critical for binding transcriptional modulators such as p300 and YY1, as well as S4 and S8 regulatory components of the proteasome . The individual contributions of TbetaRII downregulation and p300 sequestration to the overall inhibition of TGF-beta1 signaling by E1A in infected cells should be determined in future studies. A key question is whether the adenovirus-mediated downregulation of TbetaRII and TGF-beta1 signaling that we had observed in cultured cells is physiologically relevant. Unfortunately, there is not an established animal model to study adenovirus pathogenesis. Accordingly, we explored the effect of TGF-beta1 on the adenovirus life cycle in vitro using A549 cells, a cell line commonly used to study adenoviruses. It is known that treatment with TGF-beta1 results in growth inhibition but not apoptosis of A549 cells . According to our data , TGF-beta1 treatment prior to and throughout the infection decreased the accumulation of adenovirus structural proteins, which correlated with decreased virus yields in cytokine-treated cells. This occurred in cells that were maintained in low FCS. Interestingly, there was more of a decrease with a mutant that does not downregulate TbetaRII. The decrease is not due to TGF-beta1-induced apoptosis inasmuch as A549 cells are resistant. The decrease was not caused by low serum because levels of E1A and structural protein accumulation in cells in 10 and 0.2% serum were equal. These results raise the possibility that TGF-beta1 inhibits the adenovirus life cycle and that E1A counteracts this inhibition. One possibility is that there is a delayed transition into the S phase of the cell cycle in cells infected with mutants incapable of inhibiting the TGF-beta1 pathway and that this could delay late protein synthesis. Alternatively, adenovirus gene expression may be directly regulated by the Smad-containing transcriptionally active complexes. In conclusion, we have shown that E1A downregulates TbetaRII in adenovirus-infected cells and that the activation of TGF-beta1 pathway is severely compromised in the course of infection with wild-type adenovirus. The negative effect exerted by TGF-beta1 on the accumulation of adenovirus structural proteins and virus growth may have played a role in adenovirus's acquisition of the capability to specifically target TGF-beta1 signaling in the host cells. FIG. 1. : Adenovirus infection downregulates the TbetaRII protein. Adenovirus infection downregulates the TbetaRII protein. (A) A549 human lung adenocarcinoma cells were infected with rec700 at 50 PFU/cell. Some of the infected cells were maintained in 20 mug of AraC/ml. Cell lysates collected at 19 h p.i. were subjected to Western blotting using anti-TbetaRII antibodies (Santa Cruz Biotechnology). Upper and lower arrows, mature and immature forms of TbetaRII, respectively. (B) A549 cells were infected with rec700 at 10 PFU/cell. Cell lysates were collected at the indicated hours p.i. and analyzed by Western blotting. Equal protein amounts were loaded per lane. FIG. 2. : The E1A 13S or 12S protein, including amino acids 2 to 36 and the CtBP binding site, is required to downregulate the TbetaRII protein in adenovirus-infected cells. The E1A 13S or 12S protein, including amino acids 2 to 36 and the CtBP binding site, is required to downregulate the TbetaRII protein in adenovirus-infected cells. A549 (A and B) or HepG2 (C and D) cells were infected with wild-type (wt) or mutant adenoviruses at 50 PFU/cell. Cell lysates were harvested at 24 h p.i. for the wild type and most mutants and at 48 h p.i. for dl313, 12Swt, dC-term, Ad/E3, 12S.2-36, 12S.928, and dl312. Cell lysates were subjected to Western blotting with anti-TbetaRII antibodies (Santa Cruz Biotechnology). (E) Schematic of the adenovirus (Ad) genome. The E1A proteins activate the transcription of adenovirus genes and deregulate the cell cycle by suppressing or activating cellular proteins and genes. E1B proteins suppress cellular apoptosis. E3 proteins confer a stealth function to the virus by inhibiting immune cell-mediated apoptosis. E4 proteins function in gene regulation, in part by facilitating degradation of p53; they are also required for viral mRNA transport from the nucleus. Virus DNA replication is necessary for late protein synthesis derived from the major late transcription unit. At about 24 h p.i., virions begin to assemble in the cell nucleus, and after 2 to 3 days cell lysis begins to occur, with the release of virions. FIG. 3. : E1A is required to reduce TbetaRII mRNA levels in adenovirus-infected cells. E1A is required to reduce TbetaRII mRNA levels in adenovirus-infected cells. HepG2 cells were infected with wild-type (wt) or mutant adenoviruses at 50 PFU/cell and maintained in the presence of AraC throughout the infection. Total RNA was isolated at 19 (mock, rec700, and pm975) and 43 h p.i. (12Swt, 12S.2-36, 12S.928, dl312, and dCtBP). An RNase protection assay was performed using 30 and 6 mug of total RNA for detection of TbetaRII and GAPDH mRNA, respectively. Lanes a and b (A and B), full-length GAPDH and TbetaRII probes, respectively, hybridized with yeast (Torula sp.) total RNA without subsequent RNase treatment; lanes c (A and B), both probes hybridized with yeast total RNA and treated with RNase. Radioactively labeled DNA markers (A, lane n) are 300, 200, and 100 bp. RNase-protected TbetaRII and GAPDH mRNA fragments were quantified by densitometry (bar graphs). The densities of TbetaRII fragments normalized against the density of GAPDH fragments are presented under the corresponding data. FIG. 4. : Stability of the TbetaRII protein as determined by pulse-chase analysis in adenovirus-infected cells. Stability of the TbetaRII protein as determined by pulse-chase analysis in adenovirus-infected cells. (A and D) In two independent experiments, A549 cells were mock infected or infected with rec700 at 50 PFU/cell and maintained in the presence of AraC throughout the infection. At 19 h p.i., cells were pulsed with [35S]methionine-cysteine for 15 min and chased in 10% FCS-supplemented nonradioactive medium for the indicated time periods. Immunoprecipitated, radioactively labeled TbetaRII was quantified by phosphorimager at each time point of chase (B and E, respectively). (C) Calculations of the TbetaRII half-life (t1/2). FIG. 5. : E1A inhibits TGF-beta-induced signal transduction in adenovirus-infected cells as determined by using the 3TP-lux reporter plasmid. E1A inhibits TGF-beta-induced signal transduction in adenovirus-infected cells as determined by using the 3TP-lux reporter plasmid. (A) HepG2 cells were transfected with 3TP-lux and pCMV-betaGal plasmids. After 6 h of transfection, cells were infected with rec700 at 50 PFU/cell or mock infected. Infections were maintained in the presence of AraC. Cells were treated with TGF-beta1 from 12 to 26 h p.i.; subsequently, cells were lysed and luciferase and beta-Gal activities were measured. Luciferase values were normalized against beta-Gal activity for each sample. Each experimental condition was done in triplicate, and the average values are shown. (B) HepG2 cells were transfected with 3TP-lux and CMV beta-Gal plasmids and an empty vector or a plasmid expressing either the 13S or 12S isoform of E1A. At 18 h posttransfection, cells were treated with recombinant TGF-beta1 (3 ng/ml) for 8 h. (C) HepG2 cells were treated as described for panel A. Cells were maintained from 12 to 48 h p.i. in the presence of 5 ng of TGF-beta1/ml. (D) A549 cells were infected with 50 PFU/cell of rec700 or mock infected and maintained in the presence of AraC. At 20 (lane d) or 24 (lane e) h p.i. cells were mock treated or treated with 5 ng of TGF-beta1/ml for 20 min. Levels of phospho-Smad 2 (arrow) were determined by Western analysis. Molecular weight marker positions are shown. FIG. 6. : TGF-beta1 suppresses adenovirus late protein synthesis and virus yields in infected cells. TGF-beta1 suppresses adenovirus late protein synthesis and virus yields in infected cells. A549 cells were maintained in 10 or 0.2% FCS or 0.2% FCS plus 5 ng of TGF-beta1/ml as indicated throughout the experiment. Following 3 days of pretreatment, cells were infected with adenovirus mutants at 10 PFU/cell. At 24 h p.i., cell lysates were collected and analyzed by Western blotting using anti-Ad5 (A) and anti-E1A (B) antibodies. Equal protein concentrations were loaded in all lanes. The signal from every lane was measured and quantified with FluorChem software (Alpha Innotech Corporation). (C) Ratio of the signal in serum-starved cells to that in cytokine-treated cells for each adenovirus mutant. (D) A549 cells were treated and infected as for panels A and B. Cell lysates and supernatants were collected, and total virus yields at the indicated times postinfection were determined. Backmatter: PMID- 12915580 TI - A Bromodomain-Containing Protein from Tomato Specifically Binds Potato Spindle Tuber Viroid RNA In Vitro and In Vivo AB - For the identification of RNA-binding proteins that specifically interact with potato spindle tuber viroid (PSTVd), we subjected a tomato cDNA expression library prepared from viroid-infected leaves to an RNA ligand screening procedure. We repeatedly identified cDNA clones that expressed a protein of 602 amino acids. The protein contains a bromodomain and was termed viroid RNA-binding protein 1 (VIRP1). The specificity of interaction of VIRP1 with viroid RNA was studied by different methodologies, which included Northwestern blotting, plaque lift, and electrophoretic mobility shift assays. VIRP1 interacted strongly and specifically with monomeric and oligomeric PSTVd positive-strand RNA transcripts. Other RNAs, for example, U1 RNA, did not bind to VIRP1. Further, we could immunoprecipitate complexes from infected tomato leaves that contained VIRP1 and viroid RNA in vivo. Analysis of the protein sequence revealed that VIRP1 is a member of a newly identified family of transcriptional regulators associated with chromatin remodeling. VIRP1 is the first member of this family of proteins, for which a specific RNA-binding activity is shown. A possible role of VIRP1 in viroid replication and in RNA mediated chromatin remodeling is discussed. Keywords: Introduction : Viroids are the smallest known pathogens of higher plants. They consist of an infectious, single-stranded, covalently closed circular RNA molecule of ca. 250 to 400 nucleotides (nt). Viroids replicate autonomously, causing symptoms in specific host-viroid combinations of developmental disorders . Viroids are divided into two structural and functional distinct families. Potato spindle tuber viroid (PSTVd) is the type member of the Pospiviroidae and replicates in the nucleus. Avocado sunblotch viroid, the type member of the Avsunviroidae, replicates in the chloroplast of plant cells . Replication of both groups of viroids proceeds via RNA intermediates of complementary polarity. There is no evidence for viroid-encoded polypeptides. Thus, all genetic functions of the viroid replicon, such as cell entry and movement, replication, host specificity, and pathogenicity must be encoded in the RNA sequence and the resulting secondary structure, which is believed to provide specific signals for protein factors of the host. The predicted secondary structure of most viroids is a rod-like structure with five structural domains and an unusually high content of G and C bases . Despite the simple genome of viroids and the extensive knowledge of their structure and domains, little is known about host factors that interact with these structural elements. PSTVd replicates in the nucleus and accumulates eventually in the nucleolus as a circular single-stranded molecule, which is defined as positive polarity . Inhibition of PSTVd replication by alpha-amanitine indicates that viroid RNA strands of positive and negative polarity are most probably synthesized by the DNA-dependent RNA polymerase II (Pol II) . This enzyme can also accept PSTVd RNA as a template and "transcribe" it in vitro, initiating the synthesis at specific start sites of the circular PSTVd "plus" RNA template . Oligomeric replication intermediates of negative and positive polarity suggests that replication occurs by a rolling-circle-type mechanism . Some viroid-host protein complexes have been described so far; however, the significance or function of most of them for the viroid life cycle and/or pathogenicity is still unclear. Using in vitro-reconstituted nuclear complexes, PSTVd RNA was found to be associated with histones and two proteins of 41 and 31 kDa, suggesting that a similar complex occurs in nuclei of infected cells . Another protein that potentially interacts with viroids was described by Hiddinga et al. . These authors identified a host-encoded 68-kDa protein whose phosphorylation depended on the presence of PSTVd. The protein was immunologically related to a similar double-stranded RNA-dependent protein kinase in mammalian cells that has been implicated in the downregulation of protein synthesis after virus infection. Hammond and Zhao have isolated the gene of another serine/threonine protein kinase, which is induced after viroid infection. The kinase, called PKV, has a molecular mass of 55 kDa, a putative nuclear localization signal and autophosphorylating activity. However, there was no evidence for binding or direct activation of the kinase activity after binding to viroid RNA, and the role of these kinases in viroid life cycle or pathogenesis signal transduction remains elusive. Two chloroplast proteins from avocado were recently shown to interact with avocado sun blotch viroid, and one of them was involved in maturation of the viroid RNA . Recently, it was shown by in situ hybridization that most PSTVd RNA is localized in the nuclei of infected cells, but replicative forms are also present in the phloem . In the same study, it was also shown that PSTVd invades only sink leaves but not source leaves. PSTVd is confined to the vascular tissue of sink-source transition leaves. Further, PSTVd could not be localized in some flower organs, such as petals, anthers, and ovaries. Two groups could independently demonstrate an interaction between hop stunt viroid and the phloem protein PP2, an abundant lectin present in phloem exudates . It was suggested that this protein, although not exhibiting sequence-specific RNA binding, might be involved in transport of viroids and possibly also other RNAs. In the present study, we applied an RNA-ligand screening procedure for the identification of a novel tomato RNA-binding protein VIRP1. We show that the protein interacts specifically with PSTVd RNA. Furthermore, we identified an in vivo complex of the VIRP1 with the PSTVd RNA. We examined the expression of VIRP1 in different healthy and PSTVd-infected tissues and discuss the possible function of this tomato protein within the plant cell and during the viroid life cycle. MATERIALS AND METHODS : Plasmids and in vitro RNA transcription. | The plasmids used in the course of the present studies for in vitro synthesis of the labeled RNA transcripts were as follows. EcoRI-linearized pHa106 served as a template DNA for longer-than-unit-length positive-strand PSTVd. EcoRI-linearized pSP-Av5.8(-) was used for the production of the pentamer form of PSTVd(+). PvuII-linearized pGEM-3Zf(-) (Promega) and XbaI-linearized pBluecript II KS(+) (Stratagene, La Jolla, Calif.) served as templates for the production of control RNAs. For the actin transcript, a DraI/EcoRI fragment from plasmid pSac3 was inserted into the EcoRI/SmaI sites of pT3T7 vector and, after HindIII linearization, it was used as a template DNA for the negative-strand RNA synthesis. An EcoRI-linearized plasmid was used as a template for the synthesis of the 46-mer U1-RNA stem-loop II containing nt 50 to 92 of the human U1-RNA . The template used for transcription of potato U1 snRNA was a pU1EH HindIII-linearized plasmid . Transcription reactions were carried out at 37C in a final volume of 20 mul under conditions recommended by the manufacturers. After incubation for 1.5 h, the template was digested with RNase-free DNase, followed by phenolization after a further 5 min. The synthesized RNA was separated from the unincorporated nucleoside triphosphates by chromatography through a 2-ml column of Biogel A (0.5 M; Bio-Rad) in 10 mM Tris-HCl-1 mM EDTA (pH 8.0). In vitro transcripts were checked prior to use by 5% polyacrylamide gel electrophoresis (PAGE) containing 8 M urea. Screening of a cDNA expression library by RNA-ligand. | A lambda-ZAPII cDNA expression library was constructed from poly(A)+ mRNA isolated from PSTVd viroid-infected Lycopersicum esculentum (cultivar Rentita) leaves. The screening assay was carried out essentially as previously described by Sagesser et al. . cDNA cloning and DNA sequencing. | For further analysis of the lambda-ZAPII phage clones, an excision procedure using the R408 helper phage to generate subclones in pBluescript SK(-) (Stratagene) phagemid vector was performed . Templates for DNA sequence analysis were purified by using Qiagen plasmid kit according to manufacturer's protocols. Dideoxynucleotide chain termination reactions were conducted with deoxyadenosine 5'-[gamma-[35S]thio]triphosphate (Amersham) for DNA sequence analysis on both strands or by automatic sequencing. The sequencing reactions were done with the U.S. Biochemicals T7 Sequenase version 2.0 kit as previously described by Rouer . Introduction of sequencing priming sites throughout the cDNA cloned in pBluescript SK(-) was generated by using the transposon gammadelta . Sequence analysis and database searches were done by using FASTA, ALIGNMENT, MAP, and BESTFIT routines of the University of Wisconsin Genetic Computer Group Package, version 8.0 , and tools available through National Center for Biotechnology Information (NCBI), The Arabidopsis Information Resource, and European Bioinformatics Institute websites. The nucleic acid sequences reported here were submitted to the EMBL database under accession numbers , , , and . Northern and Southern blot analysis. | RNA was purified from PSTVd-infected and noninfected tomato leaves as described previously . Poly(A)+ mRNA purification was carried out with a Dynabead mRNA purification kit (Dynal, Oslo, Norway) according to the manufacturer's instructions. For RNA gel blot analysis, 15 mug of total RNA and poly(A)+ mRNA extracted from 100 mug of total RNA were electrophoresed on 1% agarose gel containing formaldehyde or guanidine thiocyanate and then blotted onto Nytran membrane (Schleicher & Schuell). Blots were prehybridized at 65C in 50% formamide as described earlier with an alpha-32P-labeled 0.7-kb transcript specifying an internal portion of the VIRP1 cDNA (nt 695 to 1435) or a labeled DNA probe containing the RsaI fragment of the cDNA (nt 136 to 2339). In cases where a DNA probe was used, formamide was omitted from the hybridization solution, and in all cases an additional washing step with 0.1x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate) was performed. Equal loading of RNA was verified by ethidium bromide staining of the gel before transfer to the membrane. Alternatively, the RNA blot was hybridized with a radioactively labeled RNA probe for the constitutively expressed actin gene. Genomic DNA from L. esculentum (cultivar Rentita) leaves was extracted as described by Saghai-Maroof et al. . For DNA gel blot analysis, 10 mug of genomic DNA was digested with restriction enzymes, fractionated on 0.7% agarose gels, and transferred onto Nytran membranes (Schleicher & Schuell). Hybridization was performed at 50C with an internal VIRP1 BamHI fragment (encompassing nt 695 to 1435) RNA probe as described above for Northern blot analysis. Membranes were washed at a final stringency of 1% saline sodium citrate buffer and 1% sodium dodecyl sulfate (SDS) at 50C for 1 h before autoradiography. Protein expression and purification. | Two oligonucleotides, His1 (5'-TAT GCA TCA CCA TCA CCA TCA CGT GTT AAC CCG G-3') and His2 (5'-GAT CCC GGG TTA ACA CGT TAT GGT GAT GGT GAT GCA-3') were synthesized, annealed, and ligated into the NdeI/BamHI site of pET-3a , producing pHis1. This vector includes six histidine residues following the initiator methionine. The pVIRP1 (nt 1 to 2770) and pVIRP1Delta (nt 1091 to 2403) cDNAs were excised from the pBluescript SK(-) vector by SmaI/XhoI digestion and then inserted into the HpaI site of the pHis1 expression vector (the XhoI site had been blunted with the Klenow fragment), yielding pHis-VIRP1 and pHis-VIRP1Delta, respectively. The original library clone lambdaVIRP1Delta contained a stretch of nine GA nucleotides derived from the cloning procedure, in addition to several nucleotides from the vector; library clone lambdaVIRP1 contained several nucleotides from the vector. Expression of recombinant protein was performed in Escherichia coli BL21(DE3) cells by IPTG (isopropyl-beta-d-thiogalactopyranoside) induction in Luria-Bertani (LB) medium supplemented with 100 mug of ampicillin/ml. Expression and purification of the proteins was essentially as described previously . E. coli-expressed fusion protein pHis-VIRP1Delta was further purified under denaturing conditions by using Ni-nitrilotriacetic acid (NTA) columns according to the manufacturer's instructions (Qiagen, Hilden, Germany). The protein was renatured by stepwise dialysis against native buffer (10 mM Tris-HCl [pH 8.0], 100 mM NaCl, 10% [vol/vol] glycerol, and 1 mM dithiothreitol). Northwestern blot analysis. | Recombinant E. coli BL21(DE3) harboring plasmids pHIS-VIRP1 and pHis-VIRP1Delta was grown at 30C in LB medium containing 100 mug of ampicillin/ml. RNA-protein interactions analyzed by Northwestern assay were conducted essentially as reported previously . Electrophoretic mobility shift assay (EMSA). | Protein-RNA interaction analyses were performed by mixing protein and nucleic acid in binding buffer (10 mM HEPES-NaOH [pH 8.0], 50 mM KCl, 100 mM EDTA, and 5% glycerol) in a final volume of 10 mul. All samples contained 1 mug of yeast tRNA and similar6 x 103 cpm of the radioactively labeled RNA. Unless otherwise indicated, the RNA-protein mixtures were kept for 60 min at room temperature (22C). The complexes were separated by electrophoresis through a 6% nondenaturing polyacrylamide gel (29:1 [acrylamide/bisacrylamide ratio]) by using 0.5x TBE buffer (50 mM Tris base, 50 mM boric acid, and 1 mM EDTA [pH 8.3]) at a constant voltage of 10 V cm-1 and room temperature. Gels were prerun for 1 h before samples were loaded. Gels were dried and exposed to X-ray film. Purification of extracts and IP. | Tissue irradiation and homogenization was performed essentially as described previously . Briefly, leaf tissue from PSTVd-infected tomato was irradiated with 20 J of UV light/cm2. After irradiation, crude extract was obtained by homogenizing the tissue in immunoprecipitation (IP) buffer (0.1 M Tris-HCl [pH 9], 0.1 M NaCl, 0.1 M 2-mercaptoehanol, 10 mM EDTA) at a ratio of 5 ml/g (fresh weight) of tissue. The extract was clarified by low-speed centrifugation. IP was performed with an extract volume corresponding to 0.2 g of fresh tissue. Then, 5 mg of protein A-Sepharose (Amersham) was equilibrated in IP buffer by rotating the mixture at 4C for 1 h. The beads were treated with 10 mul of VIRP1 rabbit polyclonal antibody (prepared from purified HIS-VIRP1Delta protein at the Department of Applied Biochemistry and Diagnostics, Institute of Molecular Biology and Biotechnology) in a volume of 200 mul, incubated for an additional hour, and subsequently incubated with the PSTVd-infected tomato extracts. After incubation overnight at 4C, protein A-Sepharose was pelleted and washed four times with IP buffer. For proteinase K digestion after IP, immunoprecipitates were mixed with 60 mug of proteinase K in a volume of 50 mul of 10 mM Tris-HCl (pH 7.8)-5 mM EDTA-1 mM 2-mercaptoethanol-0.5% SDS, followed by incubation at 37C for 45 min. After digestion, the preparations were treated with phenol, and RNAs were recovered by ethanol precipitation. RNAs were separated by denaturing 5% PAGE, electroblotted onto nylon membranes (Hybond-N; Amersham), and fixed by exposure to UV irradiation. Membranes were hybridized at 70C in the presence of 50% formamide with a alpha-32P-labeled PSTVd RNA probe. RESULTS : Isolation of a PSTVd RNA-binding protein. | We made use of a cDNA expression library cloned in lambda-ZAPII and originating from PSTVd-infected tomato leaf tissues. This library was subjected to the RNA-ligand screening procedure that we developed previously . As RNA probe we used a linear alpha-32P-labeled PSTVd positive-strand RNA derived from plasmid pHa106 . The RNA transcript Ha106 represents a full monomeric unit of PSTVd, flanked by terminal repeats of the central conserved region that is highly conserved among viroids of the PSTVd group. The RNA transcript Ha106 is highly infectious and assumes a secondary structure that allows in vitro processing to a circular molecule , thus representing a functional viroid RNA. A total of ca. 600,000 plaques were screened for filter-absorbed recombinant proteins that are able to bind to the synthetic PSTVd positive-strand RNA ligand Ha106. This primary screening delivered 18 plaques exhibiting RNA binding . Since plaques were confluent and contaminated with phages from different clones, bacteriophages were collected from the signaling plaques and subjected to a secondary screening , during which they were purified to homogeneity. This was confirmed in a tertiary screening, which delivered only signaling plaques. In this way we obtained 14 independent cDNA clones that maintained binding affinity after a third round of purification. It is noteworthy that we had only a very low number of false-positive clones, which demonstrates the reliability of the screening procedure. The size of the cDNA inserts varied between 1.3 and 2.7 kb. However, partial nucleotide sequencing of the inserts revealed an identical overlapping nucleotide sequence. The cDNA with the longest insert, now designated lambdaVIRP1, was used for further analysis. The lambda clone was converted to plasmid DNA by phage rescue excision and then sequenced on both strands. The cDNA insert consisted of 2,770 bp. Sequence analysis revealed that the insert begins with a 5'-untranslated region (5'-UTR) of 224 bases, followed by an AUG codon at position 225 and an open reading frame (ORF) of 1,809 bases. An in-frame terminator UAA codon 6 bp upstream of the AUG triplet suggests that the AUG codon at position 225 serves as the initiation codon. The predicted ORF, which encodes a polypeptide with 602 amino acids, is followed by a 3'-untranslated region (3'-UTR) of 716 bp. Sequences of the 14 clones started at various sites of the 5'-UTR or ORF but had the same sequence. However, their 3'-UTRs showed differences. Four classes of 3'-UTRs designated 3'-UTR-a, 3'-UTR-b, 3'-UTR-c, and 3'-UTR-d were detected, with lengths of 341, 370, 659, and 716 nt, respectively. All four 3'-UTRs ended with an 18- to 25-nt poly(A) tail. At position 2136, a sequence motif UUGUAA was located in all four 3'-UTRs; however, an A-rich noncanonical positioning signal is localized only at position 2596 of 3'-UTR-c and 3'-UTR-d (accession no. , , , and ). None of the four 3'-UTRs contained a full mRNA polyadenylation control signal, as determined by analysis of more than 5,000 Arabidopsis thaliana genes . Analysis of VIRP1 sequence. | The lambdaVIRP1-encoded protein (VIRP1) of 602 amino acids has a calculated molecular mass of 65,586 kDa and a predicted basic pI of 8.24. VIRP1 contains some functional domains or regions of biased composition. This includes a proline-rich domain (amino acids 322 to 404) and a serine-rich domain (amino acids 566 to 595). Residues 302 to 305 provide possible sites for amidation, and there are several potential sites for phosphorylation, N myristoylation, and N glycosylation. An ATP/GTP-binding site motif A (residues 21 to 28) can be found and ---most importantly ---a bipartite nuclear localization signal (residues 328 to 345) . VIRP1 shows limited sequence similarity with a number of proteins, including the gene products of brm and fsh , RING3 and SNF2 . However, in each case the similarity is limited to the same single domain. This is a conserved domain of 110 amino acids rich in aromatic amino acids, which is called bromodomain and is found in several eukaryotic transcription factors . Recently, it was shown that the bromodomain binds acetylated lysines . Also, the Arabidopsis protein At1g05910 has homology to VIRP1 in the bromodomain region. At1g05910 itself is homologous to yeast YTA7, a protein, which belongs to the AAA family. The bromodomain has been proposed to be a "chromatin tag," which targets proteins to the dynamic chromatin in the nucleus . Except for the partial similarity to proteins from other organisms, there is similarity to two Arabidopsis proteins of unknown function (At5g10550 and At5g65330). In this case, the homology is extended to amino- and carboxy-terminal parts of the proteins, including regions from positions 98 to 145 and from 185 to 290 and also the regions from positions 390 to 470 and from 520 to 602 of VIRP1 (data not shown). Structure and transcript levels of the VIRP1 gene. | VIRP1 expression was readily detected in flower, fruit, seeds, and in the leaves . In viroid-infected plants expression of VIRP1 mRNA was repressed at least twofold compared to healthy controls. Separate examination of flower revealed the presence VIRP1 mRNA in sepals as well as in petals. To identify the copy number of the gene, genomic DNA isolated from tomato was hybridized with a VIRP1-specific RNA probe spanning nt 695 to 1435 at stringent hybridization conditions. After digestion of genomic tomato DNA with RsaI, a single signal of similar2.4 kb was detected. Mixed digestion with RsaI and EcoRI delivered a unique hybridization signal at similar2.2 kb. Digestion of genomic DNA with BamHI was expected to give only one fragment but delivered an additional weaker fragment (arrow). This suggests the existence of a second related gene with homology to VIRP1 . Specificity of binding of protein VIRP1 to PSTVd RNA. | To verify the RNA-protein interaction and to determine the degree of specificity we used three different methods. The first method used was the plaque lift assay, which represents a variation of the screening procedure. In this assay, the phage lambdaVIRP1 expressing the recombinant VIRP1 protein was mixed with a nonrecombinant phage at a known ratio, and the resulting plaques transferred to filters (plaque lift). As in the screening procedure, the filters were allowed to react with an RNA ligand. However, we used here various kinds of radioactively labeled RNA transcripts . Only when PSTVd RNA (monomeric or multimeric form) was used as ligand could plaques originating from recombinant phages be discriminated from those generated by control phages. The two other methods required a different mode of expression of the recombinant VIRP1. Therefore, the cDNAs of VIRP1 and VIRP1Delta (encompassing the full coding region or the C-terminal part from nt 1091 to 2403, the smallest cDNA clone isolated but still capable of binding) were subcloned in a modified pET3a expression vector, pHis1, which allows expression of the recombinant protein at high levels under the control of T7 RNA polymerase promoter . The modification of the pHis1 vector allows the fusion of six histidine residues to the amino terminus. Northwestern assays were performed with crude protein extracts from E. coli cells in order to confirm the binding capacity of the protein to PSTVd RNA. Protein extracts from E. coli cells containing plasmids pHis-VIRP1 or pHis-VIRP1Delta were separated on SDS-10% PAGE gels and transferred to a nitrocellulose membrane. A signal could be detected after the membrane was probed with radioactively labeled PSTVd positive-strand RNAs; however, no signal was visible when a transcript of U1 RNA or other unrelated RNAs were used (not shown). Both the full-length and the C-terminal part of VIRP1 were binding in this assay, thus confirming the specific interaction of the expressed protein, once immobilized on nitrocellulose membranes, with PSTVd RNA. To further test the specificity of interaction, the VIRP1Delta protein was purified after overexpression in E. coli BL21(DE3) and subsequent chromatography on a Ni-NTA-agarose column. The purified protein was used to test the interaction of VIRP1Delta and PSTVd in solution by EMSAs. The protein was incubated with radioactively labeled monomeric PSTVd positive-strand RNA in the presence of tRNA as competitor, and the complexes were analyzed on native polyacrylamide gels . Retardation of PSTVd RNA was observed in the presence of VIRP1Delta protein. Using unlabeled PSTVd RNA as a competitor, the retardation of the radioactively labeled RNA could be reversed. No retardation of VIRP1Delta was observed when U1 RNA or plasmid-derived RNA transcripts were used . The collective data of Fig. to demonstrated that interaction of VIRP1 and PSTVd RNA occurs in a sequence-specific manner. VIRP1 protein and PSTVd RNA form a complex in vivo. | The search for cellular factors that assist steps in the biological cycle of viroids in the present work has been done by applying the in vitro RNA-ligand screening. Although the interaction between VIRP1 and PSTVd RNA is specific, as demonstrated by three independent approaches, its possible functional significance depends on whether the observed interaction occurs also in vivo. To identify the in vivo complex, IP assays were performed with a VIRP1-specific and a nonspecific polyclonal antibody. PSTVd-infected tissue was UV irradiated in order to "freeze" the interaction between the VIRP1 protein and the PSTVd RNA . Irradiation of the plant tissue prevents possible disruption of the complex during the homogenization procedure, thus resulting in enrichment in PSTVd RNA-protein complexes. The immunoprecipitates were treated with proteinase K, and the digestion products were analyzed by denaturing polyacrylamide gel and by Northern blot hybridization with a probe for detecting positive-strand PSTVd forms. A signal corresponding to the circular positive-strand RNA (and traces of linear positive-strand RNA) could be detected in the sample immunoprecipitated with VIRP1-specific polyclonal antibody, but no signal was observed when an unrelated antibody was used . DISCUSSION : Expression and posttranscriptional modification of VIRP1. | In this work, we present the sequence of a cDNA, which encodes a protein binding specifically to PSTVd viroid RNA. It is the first host factor identified for this nuclearly replicating family of viroids. Three proteins have been isolated in our laboratory by this method: VIRP1 (described here); a second, as-yet-uncharacterized viroid-binding protein (VIRP2 [unpublished results]); and a novel RNA-binding protein from Triturus carnifex . A fourth protein was isolated at another laboratory by this approach , thus also verifying the specificity of this method for RNA-protein interactions that may be transient in vivo and not easily detected biochemically. An interesting finding regarding the expression of VIRP1 is that it is repressed at least two- to threefold in leaves of viroid-infected plants. This observation was verified in several experiments. At the moment, it is unclear whether this is an indirect interaction (VIRP1 is among a subset of genes that are indirectly downregulated upon viroid infection) or whether this is a direct interaction involving the binding of PSTVd to VIRP1. Ectopic expression of VIRP1 will clarify this question. Another interesting finding is that VIRP1 is expressed in petals in infected and noninfected (data not shown) plants, an organ that has been found to be free of PSTVd RNA in in situ hybridization experiments in tomato and Nicotiana benthamiana plants . Zhu et al. have shown that PSTVd RNA can in principle replicate independently in petals, if the initial RNA is produced from a transgene expressed in this organ, but most probably it cannot be transported to petals after a normal mechanical infection of leaves. The presence of VIRP1 in petals indicates that it might have a role in replication and that is not always in the same cells and tissues with viroid RNA in a complex. In the course of the RNA-ligand screening, we have identified four identical VIRP1 ORFs that display an unusual organization of their 3'-UTRs. All four UTRs share the first 343-bp sequence; the next 26 nt are shared only by UTR-b, UTR-c, and UTR-d; the next 289 nt are shared by UTR-c and UTR-d and, finally, UTR-d has a unique terminal region before the poly(A) sequence. Therefore, it seems that the termination of transcription and/or polyadenylation of VIRP1 are subject to some mechanism of control. Indeed, in some cases we could detect multiple transcripts around the expected length; however, we do not know yet the mechanism underlying this complexity (unpublished results). It is possible that different signals on these UTRs regulate differential transcript stability, export, or translatability. Recently, a new mechanism of translational regulation has been discovered that involves small nearly perfect or perfect complementary RNAs, the so-called micro-RNAs, which are also present in plants . Interestingly, one of these micro-RNAs matches perfectly with transcripts of transcription factors involved in signaling cascades, indicating that this flexible regulation system is applied for regulatory proteins. It will be interesting to investigate whether VIRP1 is also subject to such regulation. Alternatively, VIRP1 mRNA might be subject to stability regulation through other RNA-binding proteins. Possible function of VIRP1. | The most striking feature of the identified VIRP1 is the presence of a bromodomain. This is a conserved domain of ca. 110 amino acids that is characterized by four alpha-helices and is present in several eukaryotic nuclear proteins . The conserved domain can be arranged as a single, double, or multiple units and frequently occurs in combination with a wide spectrum of various other functional protein domains. At least 40 different motifs can be found in bromodomain-containing proteins, e.g., AAA motifs and MYB, SAND, SET, or WD40 domains (data retrieved from the CDART-NCBI programme; ). The complexity of the bromodomain-containing proteins is indicative of their multiple roles. However, a unifying role is emerging: the bromodomain is thought to be a "tag" for localization of the corresponding protein in the nucleus and, in particular, in dynamic chromatin . Further, it is clear that a number of bromodomain-containing proteins have a role in development in different organisms. For example, Drosophila brahma and its murine homologue Brg1 are components of chromatin remodeling complexes of the so-called SWI/SNF class (; reviewed in reference ). Bromodomain-containing proteins such as the Drosophila trx-G group of genes (ash1, brm, and fsh) and Caenorhabditis elegans genes lin-49 and lin-59 are important for somatic or embryonic development, indicating how chromatin dynamics may influence cell fate. Some of these genes (trx-G) act as positive regulators; for example, lin-49 and lin-59 are thought to regulate Hom-C genes, the counterparts of Hox genes in C. elegans . GCN5 interacts with acetylated histone H3 and H4 peptides, and this interaction could potentially induce repression of Gcn5p acetyltransferase activity . It has also been reported that the bromodomain of human Gcn5p interacts with Ku-DNA-dependent protein kinase, a kinase activated by DNA and important in DNA repair and replication . It is possible that bromodomain-containing proteins act to initiate and maintain an epigenetic mark, resulting in a specific transcriptional state during development, independent of the specific function of the regulated genes, as more or less "structural" chromatin state maintainers. This would also explain the relatively high expression levels of these proteins. In accordance with this view, VIRP1 also has a relatively high expression level. Since viroid symptoms include developmental disorders (such as leaf malformation and dwarfism), it is possible that the VIRP1-viroid interaction is involved in symptom formation. It is likely that VIRP1 has another cellular (RNA) target so that viroid RNA could compete for VIRP1, thus removing the protein at least in part from its natural substrate and/or location. This interaction might cause some or all of the pleiotropic symptoms, which characterize viroid infection. In this context it is interesting that PSTVd RNA has been found to be in a complex with nuclear histones . VIRP1 is the first protein that has the bromodomain associated with an RNA-binding domain. In view of the importance of RNA-mediated methylation in plants as an additional regulatory and defense mechanism, this type of protein gains a special interest. Since PSTVd is capable of initiating methylation of homologous DNA sequences in the nucleus , it is conceivable that this bromodomain-containing protein may play a role in transmitting the RNA-mediated sequence specificity to the methylation machinery of the nucleus. In this case, host genome sequences that share the necessary extent of homology with parts of viroid RNA will be methylated. This mechanism has also been discussed as a possible pathogenicity mechanism . Based on examinations by protein motif analysis programs, VIRP1 does not appear to contain a typical RNA-binding recognition motif. It will be interesting to characterize the RNA-binding domain, which is localized at the carboxy-terminal part of the protein, as well as to determine the specific primary and secondary structure of the RNAs that can interact with VIRP1. The homology of VIRP1 with a bromodomain-containing protein from Arabidopsis, which also contains an AAA domain and is the yeast and human homologue of Tat-binding protein 7 (YTA-7 or TBP-7), is interesting. TBP-7 has been shown to interact in human cells with TBP-1 , a protein with possibly dual roles in transcription and 26S protease complexes and a regulator of retroviral transcription . Interestingly, the tomato homologue to TBP-1, LeMa-1 , has been isolated earlier in our laboratory by using a different screening method. Although no RNA-binding capacity of bacterially expressed LeMa-1 could be verified thus far in vitro, it is conceivable that the three proteins ---VIRP1, TBP-7 and LeMA-1 ---might exist as a complex in vivo since two (VIRP1 and TBP-7) could interact through their bromodomains and the third, LeMa-1, possibly interacts with TBP-7. The complex of TAT-binding and associated proteins is involved in elongation of retroviral transcripts in mammalian cells . In some plants, retroviroid-like elements have been described . It is therefore possible that viroids are derived from retroviral elements. Further, PSTVd has been shown to be replicated by Pol II . VIRP1 is a good candidate to be connected with PSTVd transcription by Pol II. Possible role of VIRP1 in viroid life cycle. | It has been proposed that multimeric PSTVd, which is assumed to be replicated via DNA-dependent RNA Pol II in the nucleoplasm, is transported into the nucleolus to be processed into monomeric molecules, thereby accounting for the high concentration of PSTVd detected in the nucleolus . However, it has been recently reported that direct transport of fluorescent PSTVd monomer transcripts into the nucleolus is also possible . VIRP1 is the ideal candidate for several possible steps in viroid life cycle: (i) transferring viroid RNA to the nucleus, since it contains nuclear localization signals; (ii) exporting it from the nucleus, e.g., through phosphorylation-dephosphorylation of the serine-rich domain at the carboxy terminus; (iii) tethering viroid RNA to the Pol II transcription complex through interactions of its bromodomain; and (iv) transferring the RNA-mediated methylation signal, thereby mediating a possible mechanism of viroid-induced repression of genes. The presence of circular viroid in the immunoprecipitated complex indicates a possible nucleolar localization of the protein or a role in transporting the viroid monomer there in order to be stored until it is transferred to other cells. Also, a possible role in cell-to-cell internuclear or long distance transport cannot be excluded. Mutational analysis of VIRP1, as well as specification of the viroid domain(s) which bind to it, will elucidate this interesting viroid-host interaction. FIG. 1. : Detection of PSTVd RNA-binding protein VIRP1. Detection of PSTVd RNA-binding protein VIRP1. A total of about 600,000 plaques from a cDNA expression library from tomato leaves were subjected to a primary screening, by using a longer-than-unit-length PSTVd RNA transcript as a radioactively labeled probe. (A) The tomato cDNA expression library was plated at ca. 5,000 PFU/plate. The black arrow shows the signal corresponding to the viroid binding protein 1 clone after a primary screening. (B) The clone picked from the primary screening was plated at a density of 25 PFU/plate, and the PSTVd RNA-binding properties of positive clone 1 were confirmed by secondary screening. Plaque lift and binding assays with alpha-32P-labeled RNA transcripts were performed as described in Materials and Methods. FIG. 2. : Bromodomain of VIRP1. Bromodomain of VIRP1. Consensus protein motif pfam00439 (protein motif data collection Pfam 8.0) is aligned with a region of VIRP1. The data were retrieved from . Identical amino acids are in red, and the conservative changes are in blue. FIG. 3. : Expression and tissue distribution of VIRP1 mRNA in tomato. Expression and tissue distribution of VIRP1 mRNA in tomato. mRNA extracted from 100 mug of total RNA was size separated in 1% agarose-formaldehyde (A and C) or agarose-guanidine thiocyanate (B) gels. RNA was blotted and hybridized with an alpha-32P-labeled 0.7-kb VIRP1 antisense RNA probe. (A) Northern blot of RNAs from healthy (lane H) and PSTVd-infected (lane I) tomato leaves. The same blot was also probed with an actin RNA transcript for loading controls as shown in the lower panel (lanes 1 and 2, respectively). (B) Northern blots of RNAs from various tissues. Poly(A)+ mRNA from leaf (L, lanes I [infected] and H [healthy]), seed (lane Sd), flower (lane Fl), and fruit (Fr) were hybridized with antisense VIRP1 RNA. M, DNA molecular weight marker. (C) Northern blot of total RNA from petals (lane 1) and sepals (lane 2) of infected plants hybridized with VIRP1 DNA. Equal loading was verified by visualization of 28S and 18S rRNAs on the blot. FIG. 4. : Southern analysis of tomato genomic DNA. Southern analysis of tomato genomic DNA. (A and B) Southern analysis of tomato total DNA. Lane 1, digestion with RsaI; lane 2, digestion with RsaI/EcoRI; lane 3, digestion with HindIII; lane 4, digestion with BamHI; lane M, lambda DNA molecular weight marker digested with BstEII. (C) Restriction map of VIRP1 and the probe used for hybridization. The coding region is shown by a heavy black box. FIG. 5. : Binding of VIRP1 by plaque lift assay. Binding of VIRP1 by plaque lift assay. Binding specificity of VIRP1 and PSTVd RNA was tested by plaque lift assay. A mixture (1:1) of lambdaVIRP1 and lambda-ZAPII phage was plated out and tested for binding with different RNAs. (A and B) Only when monomeric (A) or multimeric (B) PSTVd positive-strand RNA forms (for details see Materials and Methods) were used as RNA probes could positive signals be discriminated from those due to background. (C and D) When the same mixture was plated at a lower density of 25 PFU/plate and allowed to interact with either potato U1 RNA (C) or human U1 RNA (D), only background signals were visible. In panels C and D the numbers of signals visible on the autoradiograph were the same as the numbers of plaques per plate. The black arrows show the signals corresponding to VIRP1 plaques, which interact with PSTVd (positive signals), while the open arrows indicate the background signals. FIG. 6. : Northwestern analysis of VIRP1. Northwestern analysis of VIRP1. Specificity of binding of VIRP1 protein with PSTVd positive-strand RNA was confirmed by Northwestern assay. Crude extracts of E. coli BL21 cells harboring either pHIS, pHis-VIRP1 or pHis1-VIRP1Delta or without plasmid (see the text for details) were separated by SDS-10% PAGE, transferred to nitrocellulose membranes and probed with radioactively labeled transcripts. In panels A and C, hybridization was done with PSTVd positive-strand RNA; in panel B, hybridization was done with U1 RNA. (A and B) BL21(pHIS) or BL21(pHIS-VIRP1) crude cell extracts. Lanes 1 and 3, noninduced cells; lanes 2 and 4, IPTG-induced cells. (C) BL21 cell extracts. Lane I1, BL21(pHis-VIRP1) cell extracts; lane I2, BL21 cell extracts; lane II1, BL21(pHis1-VIRP1Delta) cell extracts; lane II2, BL21 cell extracts. The arrows indicate the VIRP1 and VIRP1Delta proteins, respectively. In panel C, parts I and II present results from two different experiments. FIG. 7. : Specificity of binding of PSTVd RNA by EMSA. Specificity of binding of PSTVd RNA by EMSA. Radioactively labeled RNAs were incubated with or without purified VIRP1Delta protein for 60 min at room temperature. Mixtures were electrophoresed on 6% native polyacrylamide gels and subjected to autoradiography. When PSTVd RNA (lanes 1 to 3) was incubated with VIRP1Delta protein (lane 2), retardation due to RNA-protein complex formation could be observed, which could be competed for in the presence of 100-fold excess nonlabeled PSTVd RNA (lane 3). In contrast, when other RNA transcripts, e.g., pGEM-3Zf(-) (lanes 4 and 5), potato U1 snRNA (lanes 6 and 7), or pBluescript II KS(+) (lanes 8 and 9), were incubated together with VIRP1Delta (lanes 5, 7, and 9), no such retardation was detected. FIG. 8. : VIRP1-PSTVd RNA in vivo complexes. VIRP1-PSTVd RNA in vivo complexes. PSTVd-infected extracts from irradiated tissue were used for IP assays. After IP, the samples were treated with proteinase K, and RNAs were recovered by ethanol precipitation after phenolization. Lane 1, RNAs extracted from IP reaction with a VIRP1Delta-specific polyclonal antibody; lane 2, nonspecific polyclonal antibody used for the IP; lane 3, no antibody was used. Extracted RNAs were separated by PAGE, electroblotted, and hybridized with a alpha-32P-labeled negative-strand PSTVd RNA probe. Backmatter: PMID- 12915552 TI - Passive Immunization with Neutralizing Antibodies Interrupts the Mouse Mammary Tumor Virus Life Cycle AB - Mouse mammary tumor virus (MMTV) infects the host via mucosal surfaces and exploits the host immune system for systemic spread and chronic infection. We have tested a neutralizing rat monoclonal antibody specific for the retroviral envelope glycoprotein gp52 for its efficiency in preventing acute and chronic mucosal and systemic infection. The antibody completely inhibits the superantigen response and chronic viral infection following systemic or nasal infection. Surprisingly however, the antibody only partially inhibits the early infection of antigen-presenting cells in the draining lymph node. Despite this initially inefficient protection from infection, superantigen-specific B- and T-cell responses and systemic viral spread are abolished, leading to complete clearance of the retroviral infection and hence interruption of the viral life cycle. In conclusion, systemic neutralizing monoclonal antibodies can provide an efficient protection against chronic retroviral amplification and persistence. Keywords: Introduction : The retrovirus mouse mammary tumor virus (MMTV) is transmitted from infected mothers to offspring via milk during the first 2 weeks after birth. MMTV targets three major cell types for infection: dendritic cells (DC) and B lymphocytes early in infection and epithelial cells of exocrine secretory organs later on (for a review see reference ). Once MMTV has crossed the mucosa it infects DC and B cells in the Peyer's patches (PP) before spreading to peripheral organs . Chronic infection of the mammary gland leads to viral transmission via milk and to mammary tumor development after retroviral insertion next to proto-oncogenes . Adult mice can be chronically infected by subcutaneous (s.c.) injection or by nasal administration of the virus . After entry into the cell, viral RNA is reverse transcribed and the resulting viral DNA is integrated into the genome in target DC and B cells. This leads to expression of a viral superantigen (SAg) at the cell surface . Oral, s.c., and nasal routes of infection proceed with similar kinetics, which are dominated by efficient priming of a SAg T-cell response via an interaction between infected B cells and DC-primed SAg-reactive cognate T cells. SAg-activated T cells are slowly deleted from the peripheral T-cell repertoire. This deletion represents one of the most sensitive readouts for chronic infection and systemic spread of MMTV. While the virus persists in all lymphoid and nonlymphoid organs, MMTV establishes a chronic immune response with germinal centers, persisting antigen (Ag), and virus-specific B and T cells only in the draining lymph node (LN). This leads to a weak life-long neutralizing antibody (Ab) response against the viral envelope (Env) protein gp52 . Although in susceptible mice this chronic Ab response is unable to clear the virus infection, resistant mouse strains such as I/LnJ have evolved a strong neutralizing Ab response that controls chronic infection and prevents infection of the progeny . This is in agreement with the role of a strong neutralizing antibody response in interruption of the viral life cycle . Since the early steps of MMTV infection at mucosal sites are very similar to those of human immunodeficiency virus type 1 infection and because sensitive assays allow monitoring of the early host response, MMTV is considered an ideal animal model with which vaccine strategies to block entry and spread of a retrovirus can be designed. Although prior immunization studies have shown evidence for the role of a humoral immune response in reduction of MMTV infection , questions regarding the in vivo role of passive transfer of neutralizing Abs in the protection of peripheral and mucosal entry sites of viral infection are unanswered. For example, both mucosal secretory immunoglobulin A (IgA) and systemic IgG have been shown to counteract viral challenge at mucosal surfaces . However it is unclear whether passively transferred systemic IgG is effective in preventing mucosal virus infection. We have recently demonstrated that induction of strong neutralizing Ab responses after virus infection did not influence infection of peripheral lymphoid organs but strongly inhibited mammary gland infection, tumor development, and virus transmission to the next generation . These observations raised the question of whether neutralizing Abs were able to prevent virus production and amplification in nonlymphoid tissue rather than initial infection of B and T lymphocytes and DC. We therefore decided to study the fate of MMTV in the presence of preexisting neutralizing Abs against Env proteins. To do this, we succeeded in generating a neutralizing rat monoclonal Ab (MAb; 2B3) against the viral major external Env glycoprotein (gp52). Our data show that a single parenteral dose of MAb 2B3 can efficiently block MMTV SAg-induced T-cell responses as well as the appearance of viral particles in the milk after s.c. or mucosal challenge with MMTV. Although the MAb was unable to prevent virus entry into draining LN cells and reverse transcription in the first days after s.c. injection, MMTV infection was not amplified by the SAg response and was completely cleared within a few days. These data demonstrate a virus neutralization pathway that completely inhibits viral amplification and virus spread to the mammary gland epithelium but that allows viral uptake. MATERIALS AND METHODS : Animals and immunization. | BALB/c mice and Lewis rats were purchased from Harlan Olac (London, United Kingdom). MMTV strain SW-infected mice were obtained from IFFA Credo (L'Arlabesque, France) and bred in the Institut Suisse de Recherche sur le Cancer animal facilities. Adult female mice (7 to 10 weeks old) or 7-day-old mice were used in all experiments. BALB/c mice were passively immunized by intraperitoneal (i.p.) injection of various doses of purified MAbs or phosphate-buffered saline (PBS) prior to or after MMTV SW infection. The concentration of rat IgG MAbs in the serum of passively immunized mice was assessed by enzyme-linked immunosorbent assay (ELISA). Abs and flow cytometry analysis. | The Abs used in this study included goat anti-rat IgG-Fcgamma-horseradish peroxidase (HRP; Jackson Laboratory, Bar Harbor, Maine), rabbit anti-rat IgG (Jackson), rat anti-mouse CD29 MAb (9EG7, IgG2a; Pharmingen), polyclonal rabbit anti-gp52 IgG, and polyclonal sheep anti-gp52 IgG (kindly provided by P. Hainaut, University of Liege, Liege, Belgium) and an HRP-conjugated donkey anti-rabbit Ab (NA-934; Amersham Life Science, Little Chalfont, United Kingdom). Flow cytometry was performed by using anti-Vbeta8.2-fluorescein isothiocyanate (FITC; F23.2) , anti-Vbeta6-FITC (44-22-1) , anti-B220-FITC (RA3-6B2; BD Pharmingen, San Diego, Calif.) or Cy-Chrome, anti-CD11c-FITC (HL3; BD Pharmingen), anti-CD11b-FITC (M1/79; BD Pharmingen), F4/80-FITC (BD Pharmingen), anti-CD4-phycoerythrin (PE) or Cy-Chrome (GK1.5; BD Pharmingen); anti-CD8-PE (53-6.7; BD Pharmingen), or anti-major histocompatibility complex class II (MHC-II; I-A/I-E)-FITC or -PE (M5/114; Boehringer GmbH, Mannheim, Germany). The antibodies recognizing murine CD138 (syndecan-1 and 281-2; BD Pharmingen) and F4/80 (BD Pharmingen) were biotinylated and visualized with streptavidin conjugated to Cy-Chrome (BD Pharmingen). Peripheral blood lymphocytes (PBLs) were isolated from heparinized blood samples by centrifugation through a Ficoll (Pharmacia, Uppsala, Sweden) cushion, and the layer containing lymphocytes was collected. Mice were sacrificed, LNs and spleens were isolated and homogenized, and single-cell suspensions were used. Lymphocytes were preincubated with anti-FcgammaRII MAb (2.4G2) whole supernatant and stained in one step with FITC, PE, and biotinylated or Cy-Chrome-labeled MAbs as described previously . Dead cells were excluded from acquisition by forward scatter and side scatter gating. Data were analyzed with CELLQuest software (Becton Dickinson Immunocytometry Systems, San Jose, Calif.). For separation of B cells, T cells, DC, and macrophages from draining LNs, tissues were homogenized and cells were stained with either anti-CD11b-FITC, anti-CD11c-FITC, or anti-F4/80-FITC, together with anti-CD4-PE, anti-CD8-PE, and B220-Cy-Chrome MAbs. T cells were identified by excluding cells positive for CD11b, CD11c, F4/80, and B220 and gating on CD4+ and CD8+ cells. B cells were identified as CD11b- CD11c- F4/80- CD4- CD8- B220+ cells. For separation of macrophages and DC, cells were stained with a combination of CD11c-FITC, syndecan-PE, CD19-PE, CD3-PE, and F4/80-biotin, followed by streptavidin-Cy-Chrome. Macrophages were syndecan- CD11c- CD19- CD3- F4/80+, and DC were syndecan- CD19- CD3- F4/80- CD11c+. Virus isolation and infection. | Milk containing MMTV SW was obtained from lactating virus-infected mothers. The pool of infected milk was diluted 1:3 in PBS and centrifuged at 600 x g for 10 min to remove cells and to skim as previously described . Aliquots were stored at -70C. For generation of MAbs in rats, viral particles of MMTV SW were purified from infected milk . Briefly, MMTV-infected milk was placed on a linear 20 to 60% sucrose gradient and ultracentrifuged for 2 h at 95,000 x g and 4C in a T11170 rotor (Centrikon). The fraction between the 40 and 20% sucrose solutions was isolated, diluted in PBS, and centrifuged for 2 h at 95,000 x g and 4C. The virus pellet was dissolved in PBS, and the protein concentration was determined by using the Bio-Rad AG (Reinach, Switzerland) protein assay. Purified virus (1 mg/ml) was aliquoted and stored at -70C. For all s.c. infections 20 mul of diluted milk containing approximately 4 x 109 MMTV SW particles was injected into the footpads of mice. After 4 days mice were sacrificed and the draining popliteal LNs were removed. For nasal infection, adult mice were anesthetized by injecting a mixture of 1.5 mg of ketamine (Ketasol; E. Graub, Bern, Switzerland) and 0.35 mg of xylazine (Rompun; Bayer, Zurich, Switzerland). Subsequently, 1010 MMTV SW particles were administered to each nostril. Six days after infection, the nasally associated lymphoid tissue (NALT) lymphocytes were isolated as described previously . For infection of suckling mice 7-day-old BALB/c mice were foster nursed by MMTV SW-infected mothers. Production of virus-specific MAbs. | Eight-week-old female Lewis rats received one s.c. injection and one intramuscular injection of 30 mug of purified MMTV SW diluted on a volume basis in complete Freund's adjuvant into the base of the tail followed by three injections of 15 mug of purified MMTV SW mixed on a volume basis in incomplete Freund's adjuvant every 3 days. Inguinal and para-aortic LNs were removed at day 12, and lymphocytes were fused with the myeloma cell lines (NS1) by following standard protocols. Cells were cultured in the presence of hypoxanthine-aminopterin-thymidine-selective medium. Supernatants were tested for the presence of virus Env-specific Abs by ELISA. The Ab isotype was determined by using an isotyping kit (RMT RC1; Serotec) according to the manufacturer's recommendations. Abs were affinity purified on protein G-Sepharose 4 Fast Flow columns (Pharmacia), and purity was monitored by sodium dodecyl sulfate-8% polyacrylamide gel electrophoresis (SDS-8% PAGE). The protein concentration was determined by measuring UV absorbance at 280 nm. Immunoprecipitation and Western blotting. | One hundred microliters of hybridoma supernatant was incubated with 20 mul of MMTV-containing milk for 1 h at 4C, followed by incubation for 1 h at 4C with a mouse anti-rat kappa chain MAb (MAR 18.5; ATCC TIB 216) linked to protein A-Sepharose (Pharmacia). Samples were centrifuged at 10,000 x g for 15 s at 4C and washed three times with 50 mM Tris (pH 7.4)-150 mM NaCl-5 mM EDTA-1% Triton X-100. For analysis, aliquots were denatured by being boiled three times in SDS-PAGE buffer for 5 min, fractionated by SDS-8% PAGE, and electrophoretically transferred to nitrocellulose filters (Schleicher and Schuell; 0.2-mum pore size). Nitrocellulose filters were immunostained by using a polyclonal rabbit anti-gp52 IgG (Hainaut), followed by an HRP-conjugated donkey anti-rabbit Ab. Ab binding was visualized by incubation with HRP substrate (Amersham). Polyclonal goat anti-gp52 serum was used as a positive control. ELISA. | Ninety-six-well plates (F96 Nunc Maxisorp; Life Technologies, Basel, Switzerland) were coated overnight with 200 ng of purified MMTV SW/ml. Alternatively, coating was performed with rabbit anti-rat IgG and plates were blocked with 5% of nonfat dry milk (Rapilait, Sulgen, Switzerland) in PBS containing Tween 20 (0.1% [vol/vol]). After incubation of the hybridoma supernatant on the coated plate for 2 h at 37C, specific Abs were detected with phosphatase-conjugated goat anti-rat Fcgamma-IgG and left for 1 h at 37C. The reaction was developed by adding 50 mul of o-phenylenediamine (P-1526; Sigma, St. Louis, Mo.) and stopped by addition of 50 mul of 2 M sulfuric acid. The absorbance was read at 492 nm with Spectra Count (Meriden). In vivo neutralization assay. | Sera were collected by tail bleeding of immunized mice and depleted of complement by heating for 30 min at 56C. Sera were mixed with MMTV and PBS in equal amounts (final serum concentration: 1:3) and incubated on ice for 1 h. The mixture was s.c. injected into the hind footpads of BALB/c mice, and after 4 days the popliteal LNs were removed. Cells were stained with anti-Vbeta6-FITC or -Vbeta8.2-FITC and anti-CD4-PE and analyzed by flow cytometry. PCR. | DNA extracted from cells of different origins was analyzed by PCR as described previously . In brief, 50 ng of DNA isolated from LNs, spleens, and mammary glands was analyzed by PCR with specific primers for exogenous MMTV SW. Primers for Mtv-6, -8, and -9 were used as positive controls. To semiquantify the number of MMTV SW-infected cells in 105 total cells, DNA was extracted from lymphocytes derived from BALB.D2 mice containing two copies of the endogenous provirus Mtv-7, which is highly homologous to exogenous MMTV SW, per cell. The SAg sequence of Mtv-7 was amplified from serial dilutions of BALB.D2 DNA into BALB/c DNA, keeping the total DNA concentration constant. The PCR product was detected by liquid hybridization with a radioactive oligonucleotide probe as described previously and the specific primer 5'-CAA GGA GGT CTA GCT CTG GCG-3'. PCR products were separated on a 6% denaturing polyacrylamide gel, which was dried and then exposed to X-Omat film (Eastman Kodak Company, Rochester, N.Y.). Quantitative real-time PCR. | Quantitative PCR was performed with a Light Cycler machine and a FastStart DNA Master SYBR Green I kit (Roche, Rotkreuz Switzerland). Genomic DNA of samples and controls was prepared by standard proteinase K lysis and phenol-chloroform extractions. The positive control consisted of genomic DNA of spleen cells isolated from BALB/c mice 6 months after infection with MMTV SW. Naive BALB/c mouse DNA was used as a negative control. Quantification of integrated MMTV SW was based on specific amplification of the SAg sequence. Amplification of the chromosome sequence D12Mit199 was used to quantify DNA. The primers were as follows: SW forward, 5'-CTGCCATCCAAATCTTTGCT-3'; SW reverse, 5'-ATGCATGCCACATATGTACACA-3'; D12Mit199 forward, 5'-CTGCCATCCAAATCTTTGCT-3', and reverse, 5'-ATGCATGCCACATATGTACACA-3'. The PCR conditions for cycling were 5 s at 95C, 20 s at 65C, and 6 s at 72C for 45 cycles in a 1x master mixture containing 4 mM MgCl2, 20 ng of DNA, and primers for MMTV SW at 0.25 muM each or primers for D12Mit199 at 0.5 muM each. The sensitivity and detection limit were assessed through serial dilutions of DNA. Amplification plots were analyzed by the second-derivative method with the Light Cycler data analysis, version 3.5, software (Roche). Corrections for amplification efficiency were included. Statistical analysis. | Significant differences in protection were assessed by the Mann-Whitney U test; probability (P) values <0.01 were considered significant. RESULTS : Systemic neutralizing Abs protect peripheral and mucosal lymphoid tissue from SAg responses. | Env-reactive MAbs were obtained after immunization of Lewis rats with purified infectious MMTV SW. We selected the IgG2a and kappa chain gp52-specific MAb 2B3 for further studies. Two antibodies (4B8 and 10B9) from the same fusion were used as isotype-matched controls . To test whether the 2B3 MAb could inhibit MMTV SW infection and SAg-mediated Vbeta6+ CD4+ T-cell proliferation in vivo, BALB/c mice were i.p. injected with various doses of MAb 2B3 (100 to 0.1 mug). Five hours later mice were challenged with MMTV s.c. by injection into the hind footpad. After 4 days the mice were sacrificed, the draining popliteal LNs were removed, and the percentage of Vbeta6+ CD4+ T cells was determined by flow cytometry. Control mice which had been preinjected with PBS or MAb 4B8 had a normal MMTV-induced stimulation of SAg-reactive CD4+ Vbeta6+ T cells of from 11 to 24% . Three micrograms of MAb 2B3 was able to fully block the SAg-reactive T-cell response. The percentage of Vbeta8.2+ CD4+ T cells (14%) nonreactive to MMTV SW was not affected by MMTV SW infection (data not shown). Thus, i.p. administration of MAb 2B3 efficiently prevented SAg-specific T-cell responses in the draining popliteal LNs. These data suggest that productive MMTV infection and expression of viral SAg by the host cell were aborted. We studied whether mucosa-associated lymphatic tissue such as NALT was also protected by the 2B3 MAb. Therefore 3 or 10 mug of MAb 2B3 was i.p. injected into adult BALB/c mice and 12 h later mice were nasally infected with MMTV SW. Six days after nasal challenge mice were sacrificed and NALT was isolated. As shown in Fig. , the 2B3 MAb significantly inhibited the SAg-specific Vbeta6+ CD4+ T-cell response. While injection of 10 mug completely blocked the Vbeta6+ CD4+ T-cell increase (10.9% Vbeta6+ T cells in 2B3-treated infected mice and 10.5% in naive animals), protection was incomplete with 3 mug (11.8%). Full protection of newborn mice fed by MMTV SW-infected mothers was achieved only when 100 mug of MAb 2B3 was injected . This indicates that a 10-fold-higher dose of the MAb was required for complete prevention of the SAg response following intestinal mucosal infection. Taken together, our results indicate that an Env-specific MAb can efficiently neutralize viral infection at both s.c. and mucosal sites. Passive immunization induces long-lasting protection. | To evaluate the efficiency and duration of protection conferred by 10 mug of MAb 2B3 (around 1013 IgG molecules), mice were challenged s.c. with MMTV SW 5 h and 10 days after passive i.p. antibody injection. As shown in Fig. , the SAg stimulation of CD4+ Vbeta6+ T cells was completely inhibited even when infection occurred 10 days after injection of MAbs (12% Vbeta6+ T cells among CD4+ T cells). The rat MAb titers, as measured by ELISA, in the serum of passively immunized mice 5 h after i.p. administration of 10 mug of MAbs ranged from 1.5 to 3 mug/ml, and titers varied between 0.9 to 1.5 mug/ml at day 10 (data not shown). When MAb 2B3 was injected 48 h after virus injection, the SAg response was normal . Injection of antibody 24 h after virus injection showed a partial inhibition of the SAg response. These results show that once infection has occurred no effect on the outcome of the SAg response is observed, clearly demonstrating that the neutralizing effect occurred during the first phase of retroviral infection. Passive immunization interrupts the viral life cycle. | One of the features of MMTV infection is the peripheral clonal deletion of SAg-reactive CD4+ T cells. Deletion of SAg-reactive T cells represents the most sensitive assay available for detection of chronic MMTV infection and systemic spread. To assess whether this deletion is inhibited by MAb 2B3, female adult mice were immunized i.p. with 100 mug of 2B3, with the nonneutralizing MAb 4B8, or with PBS (control mice). After 12 h groups of three to five mice were challenged either by s.c. injection into the hind footpad or by nasal administration of MMTV SW. At days 30, 90, and 120 the percentage of Vbeta6+ CD4+ T lymphocytes in the blood was determined. At day 90 after s.c. and nasal infection, with the exception of 1 nasally infected animal, all control mice (injected with MAb 4B8 or PBS) had deleted the Vbeta6+ CD4+ T cells (Fig. and ). In contrast, none of the mice injected with the 2B3 MAb had deleted SAg-reactive T cells even after 120 days. This implies that chronic viral infection is inhibited in mice passively immunized with a single dose of systemic MAbs. To test whether 2B3-treated mice could transmit the virus to the next generation, adult female mice were mated 12 to 16 weeks after infection with uninfected adult male mice. The peripheral deletion of Vbeta6+ CD4+ T cells in the PBLs of newborn mice was assessed 4 to 5 weeks after birth by flow cytometry. In the parents no deletion was detected 210 to 240 days after passive immunization with 2B3 . Three consecutive litters (L1 to L3) were analyzed since viremia can be increased after consecutive lactations. None of the offspring from MMTV-challenged mothers treated with MAb 2B3 showed deletion of SAg-reactive T cells, in contrast to control mice (mice injected with MAb 4B8 or PBS). This lack of transmission indicates that the 2B3 MAb completely blocks chronic MMTV SW infection after s.c. or nasal application of the virus. MMTV infection is not productive after passive immunization. | The capacity of MAb 2B3 to inhibit viral entry and reverse transcription in target cells was tested by injecting 100 mug of MAb 2B3 i.p. into mice and infecting them 12 h later with MMTV SW injected into the hind footpad. Two or 4 days later the draining popliteal LNs were removed and infection was determined by semiquantitative PCR . In control mice viral DNA was detected in the draining LNs on days 2 and 4 after challenge. An increase in numbers of viral copies per cell was observed on day 4, when infected B cells are preferentially amplified with SAg-mediated help . In 2B3-treated mice, infection of the draining LN cells was clearly detectable. Similar intensities were observed on days 2 and 4 after injection in 2B3-treated mice, confirming the absence of a preferential amplification due to SAg-mediated help . Based on amplification of the highly homologous endogenous Mtv-7 sequence in serially diluted BALB.D2 DNA, we estimated that the early infection was reduced about 10-fold by the MAb 2B3 treatment. We further assessed the reduction of retroviral infection by treatment with MAb 2B3 using quantitative real-time PCR specific for the MMTV SW SAg sequence . Mice were injected with the MAb and infected as previously described. MMTV SW infection of B cells, T cells, DC, macrophages, and unsorted cells in the draining LN, which were purified by fluorescence-activated cell sorting, was quantified on day 2 after virus injection. On day 2, compared to chronically infected splenocytes, B cells and DC from MMTV-injected control mice showed 50- to 100-times-weaker infection . Viral DNA was not detected in T cells and macrophages. All these data are in complete agreement with our previously published results, indicating that early infection occurs in DC and B cells at very low levels before SAg-driven viral amplification . In the MAb-treated cells, no MMTV SW infection was detected. The sensitivity of this quantitative PCR is limited for the following reasons: only 20 ng of genomic DNA (4,000 cells) per real-time PCR assay was used, and the observed detection limit corresponded to 1 to 3 copies (data not shown). These results show that the 2B3 pretreatment reduced the early infection in B cells and DC. Using the more sensitive but less quantitative semiquantitative PCR we found about 10-fold reductions in infection in both classical MMTV targets, DC and B cells (data not shown). Therefore, the tropism of MMTV remained the same in the presence of the MAb but was reduced about 10-fold. Next, the impact of the 2B3 pretreatment on chronic MMTV SW infection was determined by semiquantitative PCR. Two months after infection, 2B3-pretreated mice had no detectable infection in the draining LN, nondraining LNs, spleen, or mammary gland . Taken together, our results demonstrate that MAb 2B3 does not prevent initial reverse transcription but significantly decreases the early infection in B cells and DC. Therefore both SAg-driven amplification of infected B cells and chronic infection, but not acute infection, are abolished in the presence of neutralizing antibodies. B-cell differentiation is prevented by Ab treatment. | During the SAg response, infected B cells differentiate to plasmablasts with SAg-mediated cognate-T-cell help. The induced plasmablasts modulate the B-cell marker CD45R (B220R) and induce expression of the plasma cell marker CD138 (syndecan-1) . As shown in Fig. the induction of B-cell differentiation into MHC-II+ B220low plasmablast B cells is completely blocked in 2B3-treated mice. Moreover, in these mice no activation and CD138 expression of MHC-II+ B220low B cells were observed and amplification of SAg-reactive Vbeta6+ CD4+ T cells was lacking . These results show that the MAb 2B3 efficiently blocks SAg-mediated B- and T-cell activation. Passive immunization renders mice susceptible to superinfection with MMTV SW. | In the course of a normal MMTV infection, neutralizing Abs that completely block superinfection with a second MMTV isolate are obtained . To address the question of whether 2B3-treated mice generate a natural neutralizing Ab response, mice were rechallenged with MMTV SW 2 months after the initial MMTV SW infection. As shown in Table , control mice (first injection: isotype-matched Ab IgG2a) could easily be infected with MMTV SW (second injection), as measured by the deletion of Vbeta6 T cells. MMTV SW-infected mice deleted Vbeta6 T cells after the first injection and did not respond to a second challenge due to the presence of neutralizing Abs. Mice treated with MAb 2B3 did not delete Vbeta6 T cells after the first MMTV injection but did delete them after the second MMTV challenge. By challenging mice with another MMTV strain (C4) that is cross-neutralized by the SW-specific neutralizing-Ab response, we obtained similar results (data not shown). These results demonstrate that the 2B3 MAb blocks not only SAg presentation and T-cell deletion but also the generation of long-lasting neutralizing Abs. In addition, the 2B3 MAb was not capable of protecting mice from MMTV infection 2 months after passive immunization. DISCUSSION : In this study we show that a neutralizing MAb specific for MMTV Env gp52 is able to completely inhibit retroviral spread and persistence in lymphoid and nonlymphoid organs. In the presence of the Ab, however, complete protection from infection is not observed in the natural targets for MMTV, B cells and DC. Despite the approximately 10-fold-reduced, albeit clearly detectable, infection, no SAg-mediated amplification of infected B cells was observed. Our data point out the potency of systemic neutralizing Abs against both free virus and virus-infected cells in LNs and mucosa and demonstrate the protective effect of passive immunization in the early phases of retroviral infection. In viral infections, neutralizing Abs that inhibit attachment to the host cell membrane are naturally generated after infection against viral surface proteins such as Env (for a review see reference ). Such neutralizing Abs not only protect the host from reinfection by released virus but also, as has been conclusively shown, act in infected cells by blocking virus uncoating , virus replication , release of infective viral particles , chronic infection, and cell-to-cell transmission of the virus . We demonstrate in this study that antibody treatment completely prevents chronic infection in lymphoid and epithelial tissues and that even tough initial infection of the natural MMTV targets, DC and B cells, is not prevented. In previous studies immunization of mice with inactivated MMTV, purified gp52, or synthetic Env peptides failed to prevent SAg-induced B- and T-cell responses after MMTV infection, whereas passive immunization with anti-SAg MAbs reduced the SAg response . It has been suggested that polyclonal anti-MMTV Env Abs also act in the postinfection phase by blocking the SAg response . Although these results would partly explain the mechanism of MAb 2B3, our observations that administration of MAb 48 h after MMTV infection did not prevent the SAg-induced T-cell stimulation argue against this hypothesis. These data are in agreement with previous studies demonstrating that zidovudine (AZT) treatment 48 h after MMTV infection was incapable in blocking virus amplification and chronic infection in the differentiated infected B cells . Similar to results of our present study, AZT treatment 24 h after virus injection induced only partial inhibition. Most likely the first round of infection is not completed after 24 h. In other virus systems passively transferred Abs have been shown to eradicate early virus dissemination but are inefficient late in infection . It is generally accepted that higher neutralization titers are required to block direct cell-to-cell spread than to neutralize infection by free virus. This explains why passive immunization after infection is less efficient. It has previously been shown that infected lymphocytes can produce MMTV particles , but the role of these released virus particles in chronic infection is debatable . Due to the absence of any neutralizing effect when the MAb was administered 48 h after virus injection, it is unlikely that free virus plays a role during the late phases of infection. Since natural MMTV infection occurs by gradual delivery of MMTV through the physiological oral route to PP , we tested the ability of parenterally administered anti-Env Abs to prevent mucosal MMTV infection. The present data show the striking capacity of a single low dose of 2B3 IgG given i.p. to inhibit SAg-induced T-cell stimulation in the NALT . As for the draining LN, the data suggest that systemic IgG gains access to mucosal tissue to counter MMTV. This is in agreement with previous reports indicating that systemic virally induced IgG can protect against viral infection at the mucosal surface . In contrast to results for nasal infection of adult mice, only a partial protection was observed in PP of newborn mice when 10 mug of Abs was administered. The discrepancy between the effects of MAb 2B3 on NALT and PP infection is unlikely to be a result of higher virus load in the pups (up to 1012 viral particles/ml) since infection levels in the neonatal PP are not higher than those in adult draining LNs . Most likely, the accessibility of PP for systemic Abs is highly restricted. The accessibilities of Abs to NALT and PP are different, since most subepithelial capillaries in the mouse PP dome, unlike those in villi, lack endothelial fenestrae and the dome capillary network is less permeable to macromolecules . The late step of MMTV infection is characterized by the systemic spread of the virus followed by peripheral deletion of the CD4+ T-cell population expressing the appropriate T-cell receptor, Vbeta. In this study no deletion of Vbeta6+ T cells in MAb 2B3-immunized mice 4 months after infection or in pups from MAb 2B3-treated infected mice, even after three litters, was observed. It has been recently demonstrated that IgG immune complexes can activate B cells by binding to both IgM and MyD88-dependent Toll-like receptors (TLR) on B cells . TLR4 has recently been shown to act as a receptor for MMTV . Absence of TLR4 binding, for example, could lead to the absence of B-cell activation. It is likely that in passively immunized infected mice the majority of viral particles enter B cells as immune complexes coated with neutralizing Abs. This explains why the virus amplification loop in B cells could be abolished. In agreement with this interpretation preexisting neutralizing Abs inhibited B-cell proliferation and differentiation in the draining popliteal LN 6 days after MMTV infection. Intriguingly, differentiation into large CD138+ B220low MHC-II+ B cells was completely blocked in mice that were treated with MAb 2B3 prior to MMTV infection. Moreover, MAb 2B3 was able to prevent the generation of neutralizing B-cell memory responses by the host since the mice were fully susceptible to challenge with MMTV SW 2 months after 2B3 MAb immunization and MMTV SW infection. We have tested whether MMTV SW entered cell subsets other than B cells and DC, which are the classical target cells for early MMTV infection in MAb 2B3-treated mice. By quantifying the number of viral copies in sorted B cells, T cells, DC, and macrophages, we found no evidence for such a target cell shift. We have recently demonstrated that stronger mammary gland infection and virus transmission to the next generation occur in mice generating only low neutralizing-Ab titers upon MMTV challenge . These natural Abs were incapable of clearing the virus from infected lymphocytes but could prevent virus spread from lymphocytes to mammary epithelial cells and among epithelial cells in most mouse strains. Mouse strains with genetic resistance to MMTV infection have been described. Several of these strains generate higher neutralizing titers after MMTV infection . Consistent with those studies we show here that passive immunoprophylaxis providing preexisting Abs before viral challenge can completely change the fate of a chronic infection. Although neutralizing Ab treatment does not prevent virus entry into a few target cells, this infection is insufficient to induce SAg-driven virus amplification and a detectable host immune response. As a consequence persistent retroviral infection of B cells is not established. FIG. 1. : Immunoprecipitation of MMTV SW with 4B8 (IgG2a[kappa]) (lane 1 [counting from the left]), 10B9 (IgG2a[lambda]) (lane 2), and 2B3 (IgG2a[kappa]) (lane 3). Immunoprecipitation of MMTV SW with 4B8 (IgG2a[kappa]) (lane 1 [counting from the left]), 10B9 (IgG2a[lambda]) (lane 2), and 2B3 (IgG2a[kappa]) (lane 3). Controls included Ab 2B3 without virus (lane 4), MMTV incubated with goat anti-gp52 (lane 5), and native MMTV (lane 6). FIG. 2. : (A) Female adult BALB/c mice were i. (A) Female adult BALB/c mice were i.p injected with different doses (100 to 0.1 mug) of MAb 2B3, the isotype-matched 4B8 MAb, or PBS and challenged by s.c. injection of MMTV SW 5 h later. Four days after infection lymphocytes isolated from draining LNs were analyzed by flow cytometry using anti-Vbeta6-FITC and anti-CD4-PE MAbs. Data are mean values of Vbeta6 expression on gated viable CD4+ T cells from two to four LNs. This assay was done twice with similar results. The hatched bar represents Vbeta6 percentages in control BALB/c mice +-2 standard deviations. (B) Adult female mice were i.p. immunized with 3 or 10 mug of MAb 2B3. Control mice were injected with PBS. Twelve hours later mice were challenged nasally with MMTV SW. Six days later the percentage of Vbeta- CD4+ T cells in NALT was determined by flow cytometry. *, P < 0.01; **, P < 0.005. (C) Seven-day-old mice were i.p. injected with 10 or 100 mug of MAb 2B3 or PBS. Twelve hours later mice were infected by foster nursing on MMTV SW-infected lactating females. Seven days later, the percentage of Vbeta+ CD4+ cells in PP from these mice was determined by flow cytometry. **, P < 0.005. (D) Female BALB/c mice were i.p. injected with 10 mug of MAb 2B3, control MAb 4B8, or PBS. After 5 h or 10 days mice were challenged by s.c. injection with MMTV SW into the hind footpad. Four days later the draining popliteal LNs were isolated and cells were analyzed by flow cytometry. The mean values and values for each individual experiment measuring Vbeta6+ cells among CD4+ T cells in single draining LNs are shown. Four to six animals were analyzed separately in each group. The assay was done two to four times with similar results. *, P < 0.001. (E) Mice were s.c. infected with MMTV SW and 24 or 48 h later 10 mug of MAb 2B3 was injected i.p. The mean values and values for each experiment measuring Vbeta6+ CD4+ T cells from six individual draining LNs are shown. FIG. 3. : Female adult mice were i. Female adult mice were i.p. immunized with 100 mug of MAb 2B3. As controls, mice were injected with the same dose of isotype-matched MAb 4B8 or PBS. After 12 h mice were challenged s.c. with MMTV SW in the left hind footpad (A) or via the nasal route (B). At the indicated time points the percentage of CD4+ Vbeta6+ T cells in PBLs was determined by flow cytometry. Results for individual mice are shown. Boxes, PBS; circles, 4B8; triangles, 2B3. FIG. 4. : Female adult mice were i. Female adult mice were i.p. immunized with 100 mug of MAb 2B3. As controls, mice were injected with the isotype-matched MAb 4B8 or PBS. After 12 h mice were s.c. challenged with MMTV SW in the hind footpad. Two or 4 days later the draining popliteal LNs were removed. (A, top) Integrated copies of MMTV SW viruses on days 2 and 4 were detected by PCR using primers specific for Mtv-7 and the exogenous MMTV SW SAg. Analysis with primers specific for endogenous Mtv-6, -8, and -9 sequences served as internal controls to confirm equal amounts of DNA in each sample. (Bottom) To semiquantify infection levels, PCR with the identical MMTV SW primers was performed with serial dilutions of BALB.D2 mouse DNA in BALB/c mouse DNA. (B, top) Levels of MMTV SW infection on day 2 were quantified by real-time PCR in total LN cells and sorted B cells (B), T cells (T), DC, and macrophages (M). Values for MMTV SW infection were normalized with respect to D12Mit199values and are presented as percentages of the chronic infection level. Duplicate analysis showed that measurement errors for MMTV SW and D12Mit199 are 35 and 3%, respectively. *, samples negative for the MMTV SW sequence. (Bottom) The specificity of MMTV SW amplification was evaluated by comparing amplification plots of nontemplate control (ntc) and genomic DNA extracted from a chronically MMTV SW-infected BALB/c mouse (chronic SW) and a noninfected BALB/c mouse. In addition specificity was shown by the melting curve. (C) Female adult mice were passively immunized (i.p.) with 100 mug of 2B3 followed by challenge with MMTV SW s.c. 2 (lanes 1 to 6) or 11 months (lane 7) later. Draining popliteal LNs, nondraining LNs, spleens, and mammary glands were removed. DNA was analyzed by PCR with primers specific for the exogenous MMTV SW SAg and Mtv-7 as well as primers specific for the endogenous Mtvs. FIG. 5. : Flow cytometry analysis of lymphocytes isolated from the draining popliteal LN 6 days after MMTV SW infection. Flow cytometry analysis of lymphocytes isolated from the draining popliteal LN 6 days after MMTV SW infection. (A) Among naive mice, 5.4% of all MHC-II+ cells express low levels of B220 (B220low B cells). In MMTV SW-infected mice, the percentage of MHC-II+ B220low B cells increases to 17.3%. In 2B3-treated infected mice, B220 expression (6.4%) is comparable to that in uninfected control mice. (B) The percentages of CD138-expressing cells gated on MHC-II+ B220low B cells are shown. (C) The percentages of Vbeta6+ cells were determined by flow cytometry gated on CD4+ T cells. The data are representative of two independent experiments with similar results. TABLE 1 : Effect of 2B3 on virus transmission from mothers to pups TABLE 2 : Susceptibility of mice to superinfection Backmatter: PMID- 12915575 TI - Maedi-Visna Virus and Caprine Arthritis Encephalitis Virus Genomes Encode a Vpr-Like but No Tat Protein AB - A small open reading frame (ORF) in maedi-visna virus (MVV) and caprine arthritis encephalitis virus (CAEV) was initially named "tat" by analogy with a similarly placed ORF in the primate lentiviruses. The encoded "Tat" protein was ascribed the function of up regulation of the viral transcription from the long terminal repeat (LTR) promoter, but we have recently reported that MVV and CAEV Tat proteins lack trans-activation function activity under physiological conditions (S. Villet, C. Faure, B. Bouzar, G. Verdien, Y. Chebloune, and C. Legras, Virology 307:317-327, 2003). In the present work, we show that MVV Tat localizes to the nucleus of transfected cells, probably through the action of a nuclear localization signal in its C-terminal portion. We also show that, unlike the human immunodeficiency virus (HIV) Tat protein, MVV Tat was not secreted into the medium by transfected human or caprine cells in the absence of cell lysis but that, like the primate accessory protein Vpr, MVV and CAEV Tat proteins were incorporated into viral particles. In addition, analysis of the primary protein structures showed that small-ruminant lentivirus (SRLV) Tat proteins are more similar to the HIV type 1 (HIV-1) Vpr protein than to HIV-1 Tat. We also demonstrate a functional similarity between the SRLV Tat proteins and the HIV-1 Vpr product in the induction of a specific G2 arrest of the cell cycle in MVV Tat-transfected cells, which increases the G2/G1 ratio 2.8-fold. Together, these data strongly suggest that the tat ORF in the SRLV genomes does not code for a regulatory transactivator of the LTR but, rather, for a Vpr-like accessory protein. Keywords: Introduction : Lentiviruses comprise a genus of complex, nononcogenic retroviruses that infect diverse mammalian hosts, including primates (human immunodeficiency virus [HIV] and simian immunodeficiency virus), carnivores (feline immunodeficiency virus), and ungulates (equine infectious anemia virus, bovine immunodeficiency virus, and the small-ruminant lentiviruses [SRLV]). In addition to the three essential retroviral genes, gag, pol, and env, lentiviruses carry a variable number of regulatory and accessory genes, most of which are situated between the end of pol and the beginning of env. The most complex of these viruses, HIV type 1 (HIV-1), encodes three regulatory genes, tat, rev, and vif, and three accessory genes, vpu, vpr, and nef. The three conserved open reading frames found between pol and env in the SRLVs maedi-visna virus (MVV) and caprine arthritis-encephalitis virus (CAEV) have classically been assigned to the regulatory genes tat, rev, and vif. Some MVV-infected sheep progress to a stage of chronic debilitating disease affecting mainly the lungs (progressive interstitial pneumonia) and the central nervous system (progressive demyelinating encephalomyelitis) many years after infection . Goats infected by the related CAEV can develop leukoencephalomyelitis (young kids) or chronic arthritis and mastitis (adult animals) . In vivo, MVV and CAEV chiefly replicate in cells of the monocyte/macrophage lineage, and viral replication has been found to be dependent on the differentiation of monocytes into macrophages . Unlike the primate and feline lentiviruses, MVV and CAEV do not cause immune deficiency in their infected hosts. This property correlates with their inability to cause productive infection of CD4+ T lymphocytes. Some studies have reported an increase in the expression of genes under the control of the MVV or CAEV long terminal repeat (LTR) or a stabilization of the corresponding mRNA in the presence of the corresponding Tat peptide, but the increases in the levels of activity were inconstant and often low, ranging from 1.6-fold to 46-fold . We have recently compared the transactivation activities of HIV-1 Tat with Tat proteins from MVV and CAEV on their homologous LTRs in different cell types, including macrophages, from natural host species and found little or no transactivation by SRLV Tat proteins under conditions in which the HIV-1 Tat produced a 60-fold increase in expression of the reporter gene . We concluded that transactivation of the homologous LTR was unlikely to be the principal function of the SRLV Tat proteins. In the present study, we attempted to identify the possible natural activities of the SRLV Tat proteins through an analysis of their cellular localization and potential function. In transfected cells, the SRLV Tat protein is strongly concentrated in the nucleus, and this localization depends on the presence of signals on the C-terminal portion of the protein. In contrast to HIV-1 Tat, the SRLV Tat protein is never found in the culture supernatant of intact transfected human or ovine cells. It is, however, incorporated into both MVV and CAEV virus particles released from productively infected cells. Since all these characteristics are shared by the Vpr protein of HIV-1, we tested the SRLV Tat proteins for their ability to mediate an important function of HIV-1 Vpr, namely, cell cycle arrest in the G2 phase . We found that the ratio of cells in G2 to those in G1 increased 2.8-fold in unsynchronized cultures transfected with tat from MVV compared to that in mock-transfected controls. Taken together, the results presented in this paper clearly show that the lentiviruses of sheep and goats do not encode a potent transactivator of their own LTRs but, rather, a protein with functions closely related to those of the Vpr proteins of the primate lentiviruses. MATERIALS AND METHODS : Plasmid constructs. | Plasmids directing the expression of green fluorescent protein (GFP) linked to the MVV Tat peptide or to the 5' or 3' portions of the molecule were constructed by inserting the relevant genetic elements between the XhoI and HindIII sites of pEGFP (Clontech, Hampshire, United Kingdom). The complete tat sequence was amplified from the p8Xsp5-RK1 (KV-1772) recombinant genomic plasmid with primers UC1-tot and LC1-tot, and the truncated sequences were generated by using LC1-5' in place of LC1-tot or UC1-3' in place of UC1-tot. A FLAG epitope-labeled Tat protein from HIV-1 and equivalent fusion proteins from MVV and CAEV were expressed from plasmids constructed by first inserting a 633-bp BglII-PstI fragment containing the cytomegalovirus promoter from pUT535 (Cayla, Montpelier, France) into plasmid BSK+ (Stratagene, Amsterdam, The Netherlands) and then adding a 158-bp SalI-ApaI fragment containing the simian virus 40 (SV40) polyadenylation signal from pUT614 (Cayla) to create the BSK-CMV-pA vehicle. The HIV-1 Tat sequence and the SRLV equivalents were tagged with the FLAG epitope by amplification from pSG5 Tat (kindly supplied by L. Gazzolo) for HIV-1, from p8Xsp5-RK1 for MVV, or from pK9 Kb for CAEV with 5' primers containing the FLAG coding sequence (H-Flag5', V-Flag5', and C-Flag5', respectively) and the relevant 3' primers (Tat HIV3, LC1tot, and Tat 04, respectively) as follows: UC1-tot, 5'CCCTCGAGGCGAAGAAGTACCAAGAAG3'; LC1-tot, 5'CCAAGCTTATCACAGTTCAACATT TC3'; LC1-5', 5'CCAAGCTTTCATTCGCCAGGTACTCTT 3'; UC1-3', 5'CCCTCGAGGCAGACTACAACGATGGCT 3'; C-Flag5', 5'GGAATTCATGGACTACAAGGACGACGATGACAGTGAAGAACTGCCTCAAAGAAGGGAGACACATCCAGAAGAACTTGTAAGGAACGTACGGGAAAGAGAAAG3'; Tat 04, 5'CCAAGCTTGATTATGTTCCCCACCC3'; V-Flag5', 5'GGAATTCATGGACTACAAGGACGACGATGACAAGGAAGAAGTACAAGAAGAAGCCAGGAGGCTTAGTAGAAGTAGAGGGAGTATTTCAATTTTATGAAGA3'; H-Flag5', 5'TGAATTCATGGACTACAAGGACGACGATGACAAGGAGCCAGTAGATCCTAGCTAGAGCCCTGGAAGCATCCAGGAAGT CAGCCTAAAACTGCTTGTACCACTT3'; and Tat HIV3, 5'CCAAGCTTCGAGCTATTCCTTC3'. Cells. | Human epithelioid carcinoma HeLa cells were maintained at 37C in 5% CO2 in Dulbecco's minimal essential medium (Life Technologies, Cergy Pontoise, France) supplemented with 10% fetal calf serum (Life Technologies). Large T immortalized goat embryo fibroblasts (TIGEF) and goat synovial membrane (GSM) cells were maintained in Eagle's minimal essential medium (Life Technologies) supplemented with 10% fetal calf serum. Transfection of plasmid DNAs and virus production. | HeLa cells or TIGEF were seeded into 60-mm-diameter (2 105 cells/plate) or 100-mm-diameter (5 105 cells/plate) tissue culture petri dishes (Elvetech, Lyon, France) and maintained in culture for 24 h before transfection with Effectene (Qiagen, Courtaboeuf, France), according to the manufacturer's recommendations. Briefly, for a 60-mm dish, 8 mul of enhancer was added to 1 mug of the relevant plasmid DNA in DNA condensation buffer and incubated for 3 min at room temperature prior to the addition of 10 mul of Effectene reagent. After a further 10 min at room temperature, the mixture was added to the culture dish and incubated for 5 h before the medium was changed. For 100-mm dishes, double quantities of DNA and reagents were used. CAEV-pBSCA virus stock was generated by Lipofectamine (Life Technologies) transfection of 5 mug of plasmid pBSCA DNA into 5 105 GSM cells which had been seeded into a 25-cm2 tissue culture flask 24 h previously . After 18 h and three washes with phosphate-buffered saline (PBS), the medium was replaced and the virus-containing supernatant medium was collected at 3, 5, and 7 days posttransfection. Supernatants were stored at -80C after they were clarified through 45-mum-pore-size membranes (Millipore, St. Quentin, France). MVV strain K1514 was produced on GSM cells under standard conditions . Immunoprecipitation assay. | Supernatant (1 ml) was collected from 60-mm culture dishes 48 h after transfection and immediately mixed with 100 mul of 10x radioimmunoprecipitation assay (RIPA) buffer (0.5 M Tris [pH 7.5], 0.5 M NaCl, 5% deoxycholic acid, 2% sodium dodecyl sulfate [SDS], 0.1 M EDTA). The cell layer was then scraped into 1 ml of 1x RIPA buffer, transferred into a 1.5-ml Eppendorf tube, and clarified by centrifugation for 10 min at 1,300 x g. Specific antigens were sought by adding 1 mul of anti-FLAG antibody (Bio-Rad, Marnes-la-Coquette, France) or 5 mul of anti-SV40 large T antibody (Santa Cruz Biotechnology, Santa Cruz, Calif.) in 20 mul of 1x RIPA buffer to 50 mul of a 6% suspension of protein A-Sepharose (Sigma, La Verpillere, France); this mixture was then added to the cell supernatant or lysate preparations in a fresh Eppendorf tube, which was stored overnight at 4C on a rocker. The Sepharose beads were then recovered by centrifugation at 500 x g for 2 min and washed five times in 1x RIPA buffer; the captured antigens were released for immunoblot analysis by resuspension in 50 mul of 2x reducing buffer (80 mM Tris [pH 6.8], 12% glycerol, 2% SDS, 2% beta-mercaptoethanol, bromophenol blue). Preparation of virions and cell lysates. | To test whether tagged SRLV Tat proteins could be incorporated into virions, TIGEF in 100-mm dishes were first infected with either CAEV or MVV (multiplicity of infection [MOI], 1) for 2 h and then cultured in fresh medium for 24 h. The TIGEF were then transfected with the relevant plasmids encoding FLAG-Tat-tagged proteins as described above, and the culture supernatants were collected 72 h later. After clarification at 1,300 x g for 10 min, virions were pelleted from the supernatants by centrifugation at 100,000 x g for 45 min at 4C and then recovered in 45 mul of lysis buffer (10 mM Tris [pH 7.5], 0.15 M NaCl, 2 mM EDTA, 0.5% Nonidet P-40). The cell layers were washed twice with PBS and disrupted with trypsin. The cells were recovered by centrifugation at 500 x g for 5 min, resuspended in 100 mul of cold lysis buffer, and maintained on ice for 4 min before clarification in a microcentrifuge for 5 min. The cell lysates and virion preparations were stored at -80C until used. For density gradient analysis, six culture dishes (diameter, 100 mm) of TIGEF were infected and then transfected as described above. At 72 h posttransfection, the supernatants were pooled and the virions were pelleted by centrifugation at 100,000 x g for 45 min. The pellets were resuspended in 500 mul of 10 mM Tris (pH 7.5)-100 mM NaCl-1 mM EDTA and layered onto an 11-ml linear sucrose gradient (20 to 60% [wt/vol] sucrose in 10 mM Tris [pH 7.5]-100 mM NaCl-1 mM EDTA). The gradient was centrifuged for 16 h at 100,000 x g at 10C, and 12 fractions (0.9 ml) were collected from the top of the tube. Materials contained in each fraction were pelleted and dissolved in 30 mul of lysis buffer and then subjected to immunoblot analysis. Immunoblot analysis. | Proteins from lysates (10 mul) or supernatants (20 mul) in sample reducing buffer were separated on 16.5% polyacrylamide gels in the presence of SDS. The proteins were transferred from the gel to nitrocellulose transfer membranes (Schleicher & Schuell, Ecquevilly, France), which were probed with monoclonal antibodies directed against the FLAG epitope (diluted 10,000 times; Bio-Rad), SV40 large T antigen (diluted 1,000 times; Santa Cruz Biotechnology), or CAEV capsid protein (CAEP8B1 monoclonal antibody, diluted 1,000 times; VMRD Inc., Pullman, Wash.) Bound antibodies were detected by incubation with horseradish peroxidase-conjugated goat anti-mouse immunoglobulin G (diluted 3,000 times; Bio-Rad) and visualized by enhanced chemiluminescence (ECL kit; Amersham, Saclay, France) according to the manufacturer's instructions. Flow cytometry analysis of G2 arrest of cell cycle. | The cell cycle was studied quantitatively by flow cytometric determination of the nuclear DNA content after cells were stained with propidium iodide (Sigma) as previously described . Briefly, 5 105 transfected cells were harvested following dissociation with trypsin; they were then rinsed and resuspended in 1 ml of PBS. Cell suspensions were then fixed by the addition of 500 mul of 1% paraformaldehyde and a 30-min incubation in ice. The fixed cells were rinsed once with PBS, incubated for 30 min at room temperature in a solution containing RNase A (10 mg/ml), and then stained by the addition of 2 mul of 10-mg/ml propidium iodide solution. The DNA content was finally analyzed for more than 10,000 events by FACScan flow cytometry (Becton Dickinson, Le Pont de Claix, France). The cell cycle distribution of fluorescent transfected cells was analyzed with Cell Quest software. Background levels of fluorescence were determined by using mock-transfected cells. RESULTS : Subcellular localization of MVV Tat protein. | We evaluated the intracellular localization of an SRLV Tat protein by comparing the distribution of fluorescence in TIGEF transfected with a plasmid (pGFP-Tat) expressing a GFP-MVV Tat fusion protein with that in cells transfected by the pGFP plasmid, which expresses only the fluorescent marker. GFP was distributed throughout the cytoplasm and nucleus of the transfected cells, as observed by fluorescence microscopy at 48 h posttransfection , whereas the GFP-Tat fusion protein was essentially present in the nucleus . To establish which portion of the MVV Tat protein was responsible for this localization, we constructed two further plasmids expressing either the N-terminal (activator) portion or the leucine-rich C-terminal portion of the protein as fusions with GFP. TIGEF transfected with the construct expressing the N-terminal portion of MVV Tat and examined 48 h later showed a distribution of fluorescence throughout the whole cell , like the distribution for GFP alone, whereas the cells transfected with the construct expressing the C-terminal portion of MVV Tat showed a nuclear localization of fluorescence , like the distribution for the intact Tat protein. MVV Tat protein is not released from transfected cells. | It has been clearly demonstrated that the HIV-1 Tat protein is secreted into the culture medium by infected, or transfected, cells , so we investigated whether an SRLV Tat protein was also present in the supernatant medium of our transfected cells. To facilitate detection of the Tat proteins, they were expressed as fusion proteins with the FLAG epitope by the plasmids pTat-Flag-MVV and pTat-Flag-HIV. At 48 h after transfection of HeLa cells, the supernatant media were collected and analyzed by immunoblot analysis following specific immunoprecipitation with the antibody to the FLAG epitope. The cell monolayer was also lysed and subjected to the same analysis. As expected, the HIV-1 Tat protein was easily detectable in both the cell lysate and the supernatant medium, but MVV Tat was absent from the supernatant medium, although clearly present in the cell lysate . Because the secretion of the Tat protein into the supernatant might be a property that is characteristic of natural host cells for the virus, we repeated the experiment using caprine TIGEF as hosts for the two plasmids. Again, the HIV-1 Tat protein was clearly present in the supernatant, although in a smaller quantity than when it was expressed in HeLa cells, and the MVV Tat protein was seen only in the cell lysate . TIGEF constitutively express the SV40 large T antigen, which was used for their immortalization, so we probed the samples with the antibody to SV40 large T as an internal control. The antigen was clearly present in the cell lysates and absent from either supernatant, suggesting that the secretion of HIV-1 Tat into the medium was a specific event and not simply due to cellular leakage (data not shown). In summary, MVV Tat, unlike HIV-1 Tat, is not actively secreted into the culture medium from transfected human or caprine cell cultures. Incorporation of tagged Tat into virions. | We next tested whether SRLV Tat proteins, although not released free into the supernatant of transfected cells, were incorporated in the virions on productive infection. TIGEF were infected at an MOI of 1 with either MVV or CAEV and then transfected 24 h later with pTat-Flag-MVV or the equivalent pTat-Flag-CO plasmids, respectively. Supernatants and cell lysates were collected 72 h later and analyzed by immunoblotting with the anti-FLAG monoclonal antibody. Control samples were mock infected and then transfected similarly. As shown in Fig. and B, cell lysates from cells transfected with MVV or CAEV Tat-Flag plasmids contained an abundant 11-kDa band recognized by the anti-FLAG antibody, whether they were infected with a replicative virus or not. In contrast, an 11-kDa band was present only in the supernatants of infected cells and never in those of the noninfected controls. Since the result described above suggested that SRLV Tat proteins might be incorporated into budding virions, we analyzed the supernatant from CAEV-infected and transfected cells by density gradient ultracentrifugation. Cells were infected and transfected as described above, the supernatant medium was subjected to high-speed centrifugation, and the resuspended pellet was placed on the top of a linear sucrose gradient as described in Materials and Methods. After ultracentrifugation, the gradient was divided into 12 fractions, and these were analyzed for the CAEV Tat fusion protein with the anti-FLAG antibody and for the major CAEV capsid antigen by using an anti-CAEV p25 Gag antibody. Proteins were separated by SDS-polyacrylamide gel electrophoresis and analyzed by immunoblotting with the antibodies mentioned above, and the Tat-FLAG fusions always colocalized with the CAEV capsid proteins . This strongly suggests that the SRLV Tat proteins are incorporated into the viral particle. Expression of Tat increases the proportion of cells arrested in the G2 phase of the cell cycle. | The possible effect of the SRLV Tat proteins on the cell cycle of the infected cells was investigated by using TIGEF transiently transfected with an MVV Tat-GFP fusion protein and then fixed and stained with propodium iodide. This procedure allowed simultaneous evaluation of the expression of the Tat fusion and of the quantity of cellular DNA. Control cells were transfected with GFP alone. DNA profiles were determined at 72 and 120 h posttransfection. At 72 h, for cells transfected with GFP alone, 46% of the transfected cells were in the G1/S phase and 35% were in the G2/M phase of their cycles; for Tat-transfected cells, only 24% were in the G1/S phase compared to 51% in G2/M at the same time . Assuming that the time spent in M and S is unchanged, this represents an increase of 2.8-fold in the ratio of cells in G2 to G1 in Tat-transfected cells. At 120 h posttransfection, the effect was less pronounced, but the Tat-transfected cells still showed a 1.5-fold increase in the G2/G1 ratio. DISCUSSION : It has recently been shown that the proteins translated from the open reading frames identified as tat in the SRLV do not function primarily as up regulators of transcription from the viral LTR, as do the Tat peptides of the primate lentiviruses . In the present study we pursued a comparison of the SRLV Tat proteins with the different primate lentivirus accessory and regulatory proteins. We used vectors expressing SRLV Tat proteins fused to identifiable markers to examine the intracellular distribution of the protein, its secretion into the supernatant medium, and its possible incorporation into released virions. Finally, we evaluated the capacity of an SRLV Tat peptide to mediate an important function of an HIV-1 accessory protein which is the specific arrest in the G2 phase of the cell cycle. Cells transfected with a plasmid encoding native MVV Tat fused to GFP showed a clear nuclear fluorescence, with occasional, possibly perinuclear, patches . We did not observe the nucleolar concentration typical of HIV-1 Tat . A similar pattern of expression is obtained when GFP is fused to the C-terminal, leucine- and cysteine-enriched portion of the MVV Tat protein, but the N-terminal portion (the so-called activator domain) does not direct a nuclear concentration. The C-terminal portion of MVV Tat begins within the leucine-rich conserved region and extends to the natural terminator following another well-conserved cysteine-rich domain. The intermediate region, although globally not well conserved between different viral isolates, always contains many basic amino acids, often in individually conserved positions, which may constitute a nuclear localization signal. Studies involving site-directed mutagenesis of these amino acids in the present construct may clarify their role. Using the FLAG epitope fused to the SRLV Tat proteins, we showed that these proteins are not secreted into the medium from transfected cells, unlike tagged HIV-1 Tat, which is secreted from the same cells under the same conditions. In natural infections, HIV-1 Tat excreted from an infected cell is believed to enter neighboring cells and to contribute to the development of pathology. The amino acid sequence of HIV-1 Tat contains no identifiable secretion signal , and such leaderless secreted proteins are believed to be externalized through the endoplasmic reticulum and Golgi apparatus by a nonclassical pathway which is still poorly understood . We show in this study that FLAG-tagged HIV-1 Tat is secreted from transfected caprine cells (TIGEF), although less efficiently than from human (HeLa) cells. This suggests that the mechanism for export of leaderless secreted proteins is conserved in mammals but that HIV-1 Tat is specifically better adapted to the human mechanism. The absence of MVV Tat from the supernatants of either human or caprine cells suggests that it does not employ this cellular mechanism. The findings described above indicate important functional differences between HIV-1 Tat and the SRLV Tat proteins, which are emphasized by the finding that the latter are exported from actively infected cells in association with the viral particles . A comparison with other regulatory and accessory proteins of the primate lentiviruses showed a considerable similarity to the Vpr accessory protein. This is a small 96-amino-acid (14-kDa) protein which is expressed in the nucleus and is exported in the viral particles budding from infected cells . FLAG labeling of HIV-1 Vpr did not alter its distribution in infected cells or viral particles . Characteristic functions of HIV-1 Vpr include arrest of infected cells in the G2/M phase of their cycle by inhibition of p34cdc2 activity , participation in nuclear import of the viral preintegration complex in nondividing cells , and a weak (two- to sevenfold) trans-activation of the viral LTR . Alignment of the primary amino acid sequences of the SRLV Tat proteins and HIV-1 Vpr protein showed that both sequences contain an acidic N-terminal domain, a central leucine-rich domain, and an arginine-rich C-terminal portion . The C-terminal portion of the SRLV Tat is usually described as cysteine rich; however, it can be seen that this sequence is also relatively rich in arginine residues. Thus, the primary structures of SRLV Tat and HIV-1 Vpr proteins are very similar, and the same main domains are conserved. In contrast, a corresponding examination of the primary structure of HIV-1 Tat protein revealed no noteworthy similarity with the structures of either SRLV Tat or HIV-1 Vpr. This resemblance of the SRLV Tat proteins to HIV-1 Vpr is reinforced by our finding that cells transfected with a plasmid expressing an MVV Tat-GFP fusion protein accumulate in the G2/M phase of the cell cycle, with a 2.8-fold increase in cells in G2/M compared to those in G1/S at 72 h posttransfection. The effect persisted at lower levels until at least 120 h posttransfection. This result indicates that MVV Tat, like HIV-1 Vpr, induces specific arrest in the G2 phase of the cell cycle. In conclusion, the different properties of the SRLV Tat proteins, including primary amino acid sequence, cellular and virion localization, absence of strong transactivation of the viral LTR, and cellular arrest in the G2 phase of the cycle, lead to the conclusion that this protein is better considered an accessory protein similar to the Vpr protein of the primate lentiviruses and that the SRLV do not encode a veritable Tat protein. FIG. 1. : Localization of wild-type and mutant MVV Tat-GFP fusion proteins. Localization of wild-type and mutant MVV Tat-GFP fusion proteins. The TIGEF cell line was transfected with pGFP (A), pGFP-Tat (B), pGFP-5' (C), and pGFP-3' (D) plasmid constructs; 48 h posttransfection, the cells were observed by fluorescence microscopy. Magnification, x400. FIG. 2. : Release of the Tat protein in the supernatant of Tat-expressing cells. Release of the Tat protein in the supernatant of Tat-expressing cells. The human HeLa (A) or goat TIGEF (B) cell lines were transfected with the pTat-Flag-MVV and pTat-Flag-HIV plasmid constructs. At 48 h posttransfection, proteins from supernatants (lanes 2, 4, and 6) and cell lysates (lanes 1, 3, and 5) were immunoprecipitated using an anti-FLAG monoclonal antibody. Immunoblots of immunoprecipitated fractions from mock-transfected (lanes 1 and 2), pTat-Flag-MVV-transfected (lanes 3 and 4), and pTat-Flag-HIV-transfected (lanes 5 and 6) cells were performed with a specific monoclonal antibody directed against the FLAG epitope. FIG. 3. : MVV and CAEV Tat proteins are associated with viral particles. MVV and CAEV Tat proteins are associated with viral particles. TIGEF were inoculated (MOI, 1) with MVV (A) or CAEV (B and C) or were mock infected. At 24 h postinoculation, the cells were transfected with pTat-Flag-MVV (A) and pTat-Flag-CO (B and C) plasmids. At 48 h posttransfection, the supernatants were collected and virions were harvested from pellets by ultracentrifugation; cell lysates were then obtained following treatment with the lysing solution. (A and B) Proteins from the supernatants (lanes 1 and 3) and cell lysates (lanes 2 and 4) were analyzed by immunoblot assay by using an antibody directed against the FLAG epitope (diluted 10,000 times). Lanes 1 and 2, cells infected with MVV (A) or CAEV (B); lanes 3 and 4, mock-infected cells. (C) Pellets of virions were resuspended and loaded on the top of a linear sucrose gradient. After 16 h of ultracentrifugation, 12 fractions were collected and proteins were extracted and then analyzed by immunoblot assay with an anti-p25 Gag antibody to reveal the major CAEV capsid protein (diluted 1,000 times) or an anti-FLAG antibody (diluted 10,000 times) to reveal the CAEV Tat protein. FIG. 4. : G2 arrest of cell cycle in Tat-expressing cells. G2 arrest of cell cycle in Tat-expressing cells. TIGEF were transfected with plasmid pGFP as a negative control or with plasmid pGFP-Tat. Cells were harvested 72 and 120 h posttransfection, and the DNA was stained with propidium iodide. Cells were analyzed by flow cytometry of more than 10,000 events with Cell Quest analysis software. Histograms of the flow cytometry analysis of the DNA content in transfected TIGEF harvested 72 h posttransfection are shown in panel A. The ordinate indicates the number of cells, and the abscissa indicates the DNA content. (B) The ratios of the percentages of cells in G2/G1 phases of the cell cycle are shown. Values shown at the top of each bar represent the mean value of results of at least three independent transfection experiments. FIG. 5. : Comparison of the primary structures of the SRLV and HIV-1 Vpr and Tat proteins. Comparison of the primary structures of the SRLV and HIV-1 Vpr and Tat proteins. The putative domains of MVV Tat, HIV-1 Tat, and HIV-1 Vpr are shown as stained boxes. The HIV-1 Tat protein contains an N-terminal proline-rich domain (Pro), a cysteine-rich domain (Cys), a core, a basic domain (basic), and a glutamine-rich domain (Gln). The HIV-1 Vpr protein contains an N-terminal acidic domain (acidic), two leucine-rich domains (Leu), and a basic domain (basic). MVV Tat protein contains an N-terminal acidic domain (acidic), a leucine-rich domain, and a cysteine-rich domain. Backmatter: PMID- 12915550 TI - Activation of Mitogen-Activated Protein Kinase and NF-kappaB Pathways by a Kaposi's Sarcoma-Associated Herpesvirus K15 Membrane Protein AB - The K15 gene of Kaposi's sarcoma-associated herpesvirus (also known as human herpesvirus 8) consists of eight alternatively spliced exons and has been predicted to encode membrane proteins with a variable number of transmembrane regions and a common C-terminal cytoplasmic domain with putative binding sites for SH2 and SH3 domains, as well as for tumor necrosis factor receptor-associated factors. These features are reminiscent of the latent membrane proteins LMP-1 and LMP2A of Epstein-Barr virus and, more distantly, of the STP, Tip, and Tio proteins of the related gamma2-herpesviruses herpesvirus saimiri and herpesvirus ateles. These viral membrane proteins can activate a number of intracellular signaling pathways. We have therefore examined the abilities of different K15-encoded proteins to initiate intracellular signaling. We found that a 45-kDa K15 protein derived from all eight K15 exons and containing 12 predicted transmembrane domains in addition to the cytoplasmic domain activated the Ras/mitogen-activated protein kinase (MAPK) and NF-kappaB pathways, as well as (more weakly) the c-Jun N-terminal kinase/SAPK pathway. Activation of the MAPK and NF-kappaB pathways required phosphorylation of tyrosine residue 481 within a putative SH2-binding site (YEEVL). This motif was phosphorylated by the tyrosine kinases Src, Lck, Yes, Hck, and Fyn. The region containing the YEEVL motif interacted with tumor necrosis factor receptor-associated factor 2 (TRAF-2), and a dominant negative TRAF-2 mutant inhibited the K15-mediated activation of the Ras/MAPK pathway, suggesting the involvement of TRAF-2 in the initiation of these signaling routes. In contrast, several smaller K15 protein isoforms activated these pathways only weakly. All of the K15 isoforms tested were, however, localized in lipid rafts, suggesting that incorporation into lipid rafts is not sufficient to initiate signaling. Additional regions of K15, located presumably in exons 2 to 5, may therefore contribute to the activation of these pathways. These findings illustrate that the 45-kDa K15 protein engages pathways similar to LMP1, LMP2A, STP, Tip, and Tio but combines functional features that are separated between LMP1 and LMP2A or STP and Tip. Keywords: Introduction : Kaposi's sarcoma (KS)-associated herpesvirus (KSHV) or human herpesvirus 8 is a human type 2 gammaherpesvirus found in all forms of KS , in primary effusion lymphoma (PEL) , and in the plasma cell variant of multicentric Castleman's disease . Strong epidemiological evidence suggests that KSHV plays an indispensable role in the pathogenesis of KS but that additional factors, such as immune suppression or coinfection with HIV, are required for the manifestation of this tumor . KSHV is present in the endothelial and spindle (tumor) cells of KS lesions, in PEL cells, and in perifollicular B cells of multicentric Castleman's disease, where it persists in a latent form with limited viral gene expression . In these tumor cells, lytic viral replication occurs in a subpopulation of KSHV-infected cells . The K15 gene of KSHV is located adjacent to the terminal repeat region at the right end of the KSHV long unique coding region and consists of eight differentially spliced exons . The sequences of all of the K15 cDNA clones isolated so far (, ; M. M. Brinkmann et al., unpublished data) are predicted to contain a common C-terminal cytoplasmic region linked to a variable number of transmembrane domains . The cytoplasmic region (amino acids [aa] 355 to 489) contains one putative SH2-binding site motif, Y481EEVL, a second tyrosine-containing motif (Y431ASIL) of the general Yxxtheta consensus found in SH2-binding sites and cytoplasmic internalization motifs, a putative proline-rich SH3-binding site, and a putative TRAF-binding site (A473TQPTDD) . These sequence motifs are conserved between the two highly divergent M and P genotypes of KSHV that have been found in this region of the KSHV genome, suggesting the conservation of associated functional properties . Phosphorylation of Y481 in the YEEVL motif and binding of TRAF-1, -2, and -3 to the cytoplasmic domain of K15 have been observed. The combination of multiple transmembrane regions with a cytoplasmic domain that can be phosphorylated on tyrosines and/or interact with TRAFs is reminiscent of the latent membrane proteins LMP1 and LMP2A of Epstein-Barr virus (EBV). LMP1 and LMP2A are both located at the ends of the coding region of the EBV genome, with LMP2A located in the position corresponding to that of K15, while LMP1 occupies the position corresponding to that of another KSHV membrane protein, K1. K1 has transforming properties and triggers a number of intracellular signaling pathways . LMP1 is considered to be essential for the tumorigenic properties of EBV by virtue of its ability to activate the pathways for NF-kappaB, c-Jun N-terminal kinase (JNK)/SAPK, p38 MAPK, and JAK/STAT (reviewed in reference ). Activation of these signaling pathways requires the multimerization of LMP-1 molecules in the membrane through its six membrane-spanning domains and the interaction of cellular signaling components with different regions in its cytoplasmic domain . Among these, so-called C-terminal activator region 1 (CTAR-1) interacts with TRAF-1, -2, -3, and -5 and activates the NF-kappaB pathway, while the more distal region CTAR-2 associates directly with TRADD and indirectly with TRAF-2 and initiates the cascades leading to the activation of JNK/SAPK and NF-kappaB . The region between CTAR-1 and -2 has been reported to be involved in triggering of the JAK/STAT cascade . LMP2A consists of an N-terminal cytoplasmic domain of 119 aa and a membrane anchor region of 12 transmembrane segments. The cytoplasmic domain contains several SH2-binding sites surrounding tyrosine residues, of which two, Y74 and Y85, are grouped as an ITAM (immune receptor tyrosine-based activation motif) (reviewed in reference ). Tyrosine 112 is part of a YEEA motif that resembles the YEEI/L consensus sequence for Src SH2-binding domains. LMP2A is expressed in latently EBV-infected circulating B cells and in B-lymphoma cells of large-cell lymphoma in immunocompromised individuals or those with Hodgkin's disease (reviewed in reference ). In B cells, binding of the Src kinase Lyn to the YEEA motif leads to phosphorylation of the ITAM, which in turn leads to recruitment of the tyrosine kinase Syk . This sequence of events is similar to the early signaling steps that occur after triggering of the B- or T-cell receptor: here, the initial phosphorylation of the ITAM in cytoplasmic components of the receptor complex leads to the recruitment of Syk (in B cells) or Zap70 (in T cells) and binding of the tandem SH2 domains in the recruited kinases to the two phosphorylated tyrosine residues in the ITAM. The interactions of LMP2A with the components of these signaling cascades lead to inhibition of B-cell receptor signaling . LMP2A interacts with the WW domain containing ubiquitin ligases Nedd4, WWP2/AIP2, and AIP4 via a PPPPY motif and thereby induces the ubiquitination and rapid proteasome-mediated turnover of the LMP2A-associated protein tyrosine kinases (PTKs) . It is thought that this contributes to the survival of EBV in infected B cells, since the effect of LMP2A would be to inhibit the activation of infected B cells by antigen and with it the activation of the viral lytic cycle . However, transgenic expression of LMP2A in murine B cells increases B-cell survival and therefore LMP2A may also make a more direct contribution to the survival of EBV-infected B cells. Unlike its counterpart LMP1, however, LMP2A is not required for the immortalization of EBV-infected B cells. LMP2A is also expressed in the tumor-derived epithelial cells of nasopharyngeal carcinoma (reviewed in reference ). Experimental expression of LMP2A in epithelial cells leads to the acquisition of phenotypic characteristics of transformation such as growth in soft agar and dedifferentiation and the activation by LMP2A of the phosphatidylinositol 3 kinase/Akt pathway is important for this LMP2A-induced phenotype . In epithelial cells, LMP2A phosphorylation can be triggered by adherence of these cells to extracellular matrix proteins and involves the regulator of Src, the C-terminal Src kinase Csk . LMP2A also binds to and is phosphorylated by the mitogen-activated protein kinase (MAPK) ERK1 . A recent report found that LMP2A activates the MAPKs Erk and JNK in B cells , but others have not observed Erk activation by LMP2A in epithelial cells . Similar to the EBV LMPs, at least two New World primate gamma2-herpesviruses, herpesvirus saimiri (HVS) and herpesvirus ateles (HVA), also encode membrane-associated proteins, STP, Tip (HVS), and Tio (HVA), that can either bind to TRAFs (STP) or are phosphorylated by members of the Src kinase family (Tip, Tio) (; reviewed in reference ). STP of HVS strains A and C is required for the transforming and/or lymphomagenic potential of these viruses (reviewed in reference ). HVS and HVA are T lymphotropic, and HVS strain C Tip is phosphorylated by the Src kinase Lck and can inhibit T-cell receptor-induced signaling . Tip may also have a modulatory role in STP-mediated transformation . Given the presumed roles of LMP1, LMP2A, STP, Tip, and Tio in immortalization, dedifferentiation, maintenance of latency, and regulation of the viral lytic replication cycle, we wanted to investigate whether KSHV K15-derived proteins have similar functions. As a first step, we sought to identify cellular signaling pathways triggered by K15 proteins and show here that the longest K15 isoform, but not the smaller isoforms, activate the Ras/MAPK and NF-kappaB pathways following phosphorylation of a tyrosine residue in a YEEVL motif by several members of the Src kinase family. MATERIALS AND METHODS : Cell culture methods, transfections, and MEK1/2 inhibitors. | The cell lines HEK (human embryonic kidney) 293-T, HEK 293, and Cos7 were cultured in Dulbecco's modified Eagle's medium (Gibco) supplemented with 10% heat-inactivated fetal calf serum (FCS), 50 IU of penicillin per ml, 50 mug of streptomycin per ml, and 300 mug of l-glutamine per ml at 37C in humidified air with 5% CO2. For transfections, cells were grown to subconfluence in six-well plates (Greiner). HEK 293-T and 293 cells were transfected with FuGENE transfection reagent (Roche; FuGENE-to-DNA ratio of 3 mul to 1 mug), Cos7 cells were transfected with the Effectene transfection reagent (Qiagen) in accordance with the manufacturer's instructions. Total amounts of transfected DNA were adjusted with salmon testis DNA (Sigma). The specific MEK1/2 inhibitors PD98059 and UO126 (Calbiochem) were used at a concentration of 50 muM from a 50 mM stock solution in dimethyl sulfoxide (DMSO). The inhibitors were added to the medium 14 h after transfection, and cells were incubated for a further 8 h before being extracted as described below. DNA constructs. | The full-length K15 cDNA clone (K15 aa 1 to 489) from the BCP-1 KSHV isolate (primary effusion cell line infected with KSHV; ) was generated by reverse transcription (RT)-PCR amplification of the K15 region with the forward primer LRH6for, which binds 192 bp upstream of the putative start codon of K15 (5'-ATCCGGAATATTTATGAGCC-3'), and primer ex8arev, which binds to the start region of open reading frame 75 (ORF75) 245 nucleotides downstream of the K15 stop codon (5'-CTCCAACCACAGCCCAGTGACG-3'). The resulting PCR product was cloned into pGEM-T (Promega) and sequenced. The PCR product was then cloned into the eucaryotic expression vector pFJEA by digestion with EcoRI (clone MBK15). Clone 35, which contains a full-length cDNA (exons 1 to 8) of K15 from the BCBL-1 PEL cell line cloned into the EcoRI/XbaI sites of pFJEA was kindly provided by J.-K. Choi . Clone 35 contains an artificial Kozak sequence, begins at the putative start codon of K15 in exon 1, and contains a Flag tag at its C-terminal end. Plasmid K15 ex1-8 (aa Delta473 to 489) lacking the C-terminal 17 aa and plasmid K15 ex1-8 Y481F carrying the Y481->F point mutation in the Y481EEVL motif were generated by PCR amplification of clone 35 with the upstream primer LAMPB5' (5'-ACAGGATCCTAAAATGAAGACACTCATATTCTTCTGG-3') and the reverse primer ex8ad1 (5'-ACGCCCGGGTTAGGCTTGGGCGGTGTCTATACGGAAGGC-3') or ex8aY:F (5'-TATGAATTCCTAGTTCCTGGGAAATAAAACCTCCTCAAACAGGTC-3'), respectively. PCR products were cloned into pGEM-T and subsequently cloned into the NotI site of pFJEA and sequenced. The cDNAs of the K15 splice variants K15 ex1/6-8, K15 ex1 as/6-8, and K15 ex1 as/4-8 were generated by RT-PCR amplification of the K15 region (BCP-1 KSHV isolate) with primers LR65 (5'-ACGATCTAGATCCGGAATATTTATGAGC-3') and ex8arev, followed by a nested PCR with forward primer FJLampfor (5'-ATCGAATTCCATTTACAACAACTCTATTG-3'), binding 45 nucleotides upstream of the K15 start codon, and reverse primer FJLamprev (5'-ATCGGTACCTAGTTCCTGGGAAATAAAAC-3'), binding at the K15 stop codon. PCR products were cloned into pGEM-T, sequenced, and subsequently cloned into the EcoRI and KpnI sites of pFJEA. A plasmid containing the entire predicted cytoplasmic domain (aa 355 to 489) of K15 fused to glutathione S-transferase (GST) (pGEX8A, GST-K15355-489) has been previously described . A further GST fusion plasmid in which the carboxy-terminal 17 aa of K15 were deleted (GST-K15355-472) was generated by PCR with the 5' primer GSTex8afor (5'-CAGGGATCCTAAATAGTTACCGACAGAGAAGGGG-3') and the 3' primer ex8arevd1. Further deletion constructs of the K15 C-terminal region of the GST-K15 fusion protein were generated with plasmid pGEX8A (GST-K15355-489) as the template together with primer GSTex8afor and either reverse primer ex8arevd2 (5'-ACTCCCGGGACCGGACACTAAAATACTGGC-3') to generate GST-K15355-438 or reverse primer ex8arevd4 (5'-ACTCCCGGGTGTGGTGTAAGTTCTGGTCCC-3') to generate GST-K15355-373. PCR products were cloned into pGEX-3X (Amersham Pharmacia) and sequenced. The LMP1-K15355-489 chimera, consisting of the six amino-terminal transmembrane domains of LMP1 and the cytoplasmic region of K15, was constructed by amplification of the transmembrane domain of LMP1 from plasmid pSG5-LMP1 with primers LMPLAMPa5' (5'-GACGGTACCATGGACTACAAGGACGATGACGACAAGGAACAC GACCTTGAGAGGGGCCCACCG-3') and LMPLAMPa3' (5'-CCTTCTCTGTCGGTAACTATTTACTTCATCACTGTGTCGTTGTCCATGG-3'). Primer LMPLAMPa5' introduces the Flag epitope tag DYKDDDDK (sequence underlined), while the first 25 nucleotides of LMPLAMPa3' are homologous to a region within the carboxy terminus of K15 (sequence underlined). The resulting amplicon was purified and used as a primer together with primer LAMPa3' (5'-TATGAATTCCTAGTTCCTGGGAAATAAAAC-3') in a PCR with genomic BCP-1 DNA as the template. The resulting product was cloned into pcDNA3.1 (Invitrogen) and sequenced. cDNA plasmids expressing Flag-tagged TRAF-1, -2, and -3 were kindly provided by M. Rothe. Myc-tagged PTK expression plasmids for Src, Fyn, Hck, Lck, and Yes were kindly provided by S. Muller, Erlangen, Germany. pSRalpha-HA-JNK1 and pSRalpha-HA-Erk2 ; pCMV-HA-p38 ; the AP-1 reporter constructs pRTU14 (luciferase reporter gene under the control of a minimal promoter and four TREs [TPA response elements]) and pRTU1 (lacking the TREs) ; GST-c-Jun ; dominant negative mutant forms of TRAF-2 (pRK-TRAF-2 [aa 87 to 501]) , TRAF-6 (aa 300 to 524) , and Ras (N-17); expression vectors pCis2 and pRKH5; and the construct pSV-LMP1 have been described previously. Dominant negative Erk2 C3 (Y185->F) , RSV Raf C4B , and eucaryotic expression vector KRSPA (Rous sarcoma virus promoter) were kind gifts from S. Ludwig. The NF-kappaB reporter plasmid p3EnhkappaBconA-Luc containing three NF-kappaB sites upstream of the firefly luciferase gene and pSG5-LMP1 were provided by A. Eliopoulos. Expression and purification of recombinant proteins and GST fusion protein binding assays. | GST-c-Jun, GST-K15355-489, and GST-K15355-472 for radioactive immunocomplex kinase assays were purified as described in reference . The purified GST fusion proteins eluted with glutathione were analyzed by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE). For GST pulldown experiments, Escherichia coli M15 cultures transformed with the GST-K15 expression plasmids or GST alone were grown at 37C in 2YT medium plus ampicillin, induced at an optical density at 600 nm of 0.2 to 0.3 with 1 mM isopropyl-beta-d-thiogalactopyranoside (IPTG), and harvested by centrifugation 6 h after induction. The pellet was resuspended in 1,000 mul of phosphate-buffered saline (PBS) plus protease inhibitors per 10 ml of culture to an optical density at 600 nm of 2.5. After sonication on ice, Triton X-100 was added to a 1% final concentration, and the cells were kept on ice for 10 min and then centrifuged for 15 min at 14,000 x g at 4C. The supernatant was adsorbed onto 20 to 50 mul of glutathione Sepharose beads for 1 h at 4C, after which the beads were washed twice in PBS and then once in 1% NP-40 lysis buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1 mM EDTA, 1% NP-40). HEK 293 or 293-T cells were transfected as mentioned in the section on coimmunoprecipitation but with 1 mug of cDNA expression constructs per well of a six-well plate of either Flag-tagged TRAF-1, -2, or -3 (HEK 293) or the myc-tagged Src kinase Src, Lck, Yes, Fyn, or Hck (HEK 293-T). At 48 h after transfection, cells were washed once on ice with PBS and lysed in 1% NP-40 lysis buffer containing protease inhibitors for 10 min on ice. Cell lysates were centrifuged at 14,000 x g for 10 min at 4C and precleared by incubation with 50 mul of glutathione beads for 1 h at 4C. Cleared lysates were then incubated for 1 h or overnight at 4C with either GST control protein or GST-K15 fusion protein immobilized to glutathione beads. The beads were washed three times in 1% NP-40 lysis buffer, and proteins adsorbed to beads were boiled in 50 mul of SDS electrophoresis sample buffer and analyzed by SDS-PAGE and Western blotting. Generation of rabbit polyclonal and mouse monoclonal antibodies to the C-terminal K15 domain. | The carboxy-terminal region of K15 was transferred from plasmid pGEX8A to plasmid pGEX6A. The GST-K15355-489 fusion protein was prepared from 1-liter E. coli (M15) cultures, and the K15 portion of the protein was cleaved by incubation in the presence of PreScission protease while still bound to GST beads in accordance with the manufacturer's (Amersham Pharmacia) instructions. The cleaved C-terminal cytoplasmic domain of K15 (aa 355 to 489) was eluted with PBS and used to immunize two rabbits in accordance with standard protocols. Noncleaved GST-K15355-489 fusion protein was used to immunize mice in accordance with standard protocols. Immunocomplex kinase assays and Western blotting (immunoblotting). | As indicated in the figure legends, HEK 293-T cells were cotransfected with 1 mug of hemagglutinin (HA)-tagged MAPK and 1 mug of K15 cDNA expression constructs with FuGENE. Where indicated, 700 ng of dominant negative mutants or their respective empty expression vectors were also cotransfected. After transfection, cells were maintained in medium containing 1% FCS to downregulate serum-activated signaling pathways. At 24 h after transfection, cells were lysed in TBS-T buffer (20 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1 mM EDTA, 1% Triton X-100; the protease inhibitors phenylmethylsulfonyl fluoride [PMSF; 1 mM], leupeptin [50 muM], aprotinin [100 U/ml], benzamidine [200 muM], and pepstatin A [1 muM]; and the phosphatase inhibitors sodium orthovanadate, beta-glycerophosphate, sodium molybdate, sodium fluoride, sodium pyrophosphate at 0.5 mM). Cleared lysates were incubated overnight with anti-HA monoclonal antibody 12CA5 (Boehringer Mannheim) immobilized to protein G Sepharose beads (Amersham Pharmacia) to immunoprecipitate HA-tagged kinases. Afterwards, samples were centrifuged and the supernatants were analyzed for expression of K15 constructs by Western blotting. Beads with bound MAPK were washed twice with TBS-T and twice with kinase reaction buffer (20 mM Tris-HCl [pH 7.4], 20 mM NaCl, 1 mM dithiothreitol, 10 mM MgCl2, 2 muM ATP, 0.5 mM beta-glycerophosphate, 0.5 mM sodium orthovanadate). In vitro kinase reactions to assay the activity of immunoprecipitated HA-MAPK were performed in kinase reaction buffer in the presence of 10 muCi of [gamma-32P]ATP (10 mCi/ml; Hartmann Analytics) per reaction sample with 200 ng of bovine myelin basic protein (MBP; Sigma), ATF2 fusion protein (34 kDa; Cell Signaling), or purified GST-c-Jun as a substrate. Kinase reactions were stopped after 30 min at 25C with electrophoresis sample buffer, and the reaction mixtures were heated for 5 min at 100C, separated by SDS-PAGE, blotted onto Hybond-C membranes (Amersham Pharmacia), and analyzed by autoradiography. For the Src kinase assay, HEK 293-T cells were transfected with either 1 mug of expression plasmids for myc-tagged Src kinase Src, Hck, Lck, Yes, or Fyn or the empty expression vector. After transfection, cells were kept in complete medium with 10% FCS and cells were lysed 24 h posttransfection in TBS-T as mentioned above. Immunoprecipitation of the myc-tagged Src kinase was carried out overnight with anti-c-myc antibody-coated beads (Clontech). The in vitro kinase reaction was done as described above but with 200 ng of purified GST-K15355-489 or GST-K15355-472 fusion protein as the substrate. For detection of proteins by Western blotting, cleared cell lysates or fractions of the sucrose gradient containing K15 proteins were not boiled prior to SDS-PAGE. As indicated, the following primary antibodies were used for immunostaining of immunoblots: rabbit anti-JNK1 C-17, mouse anti-Erk2 D-2, rabbit anti-p38 MAPK H-147, anti-caveolin 1 (all Santa Cruz Biotech), mouse anti-c-myc 9E10 (BIOMOL), mouse antiphosphotyrosine P-Thyr-100 (Cell Signaling), and mouse anti-Flag M2 (Sigma). Immunoblots were analyzed with horseradish peroxidase-coupled secondary antibodies (DAKO) and a standard enhanced-chemiluminescence reaction (Perkin-Elmer). Coimmunoprecipitation. | One day prior to transfection, 3 x 105 HEK 293 cells were plated into each well of a six-well plate. The cells were transiently transfected with 500 ng of either plasmid LMP1-K15355-489 or the deletion mutant form LMP1-K15355-472 and 1 mug of the cDNA expression construct for Flag-tagged TRAF-1, -2, or -3 with the FuGENE transfection reagent. At 48 h posttransfection, cells were washed once in PBS and lysed in 1% NP-40 lysis buffer containing protease inhibitors (see above). Lysates were cleared by centrifugation, and supernatants were precleared with 50 mul of protein G Sepharose (Amersham Pharmacia). To 100 mul of precleared lysate, 0.5 mul of rabbit anti-K15 serum was added. After 1 h of incubation on ice, the protein-antibody complexes were incubated with 20 mul of protein G Sepharose beads at 4C with gentle shaking for 2 h. Beads were washed three times with 1% NP-40 lysis buffer, and bound proteins were recovered by boiling in SDS-PAGE sample buffer. Eluted proteins were analyzed by Western blotting as described above. Luciferase-based reporter assays. | To assay NF-kappaB activity, HEK 293-T cells were transiently cotransfected with 50 ng of the NF-kappaB luciferase reporter plasmid p3EnhkappaBconA-Luc containing three tandem repeats of the NF-kappaB sites from the immunoglobulin Gkappa promoter and the indicated amounts of K15 expression constructs or LMP1, as mentioned in the figure legends, per well of a six-well plate. At 24 h posttransfection, cells were washed with PBS and lysed in reporter lysis buffer (Promega). Luciferase activities were measured in cleared lysates with a luciferase assay system in accordance with the manufacturer's (Promega) instructions. NF-kappaB activity was calculated as fold induction compared to that of mock (empty expression vector)-transfected controls. For analysis of AP-1 activity, 50 ng of the reporter plasmid pRTU14 or the negative control reporter pRTU1 was cotransfected with different amounts of K15 or LMP1 expression constructs as indicated in the figure legends. Dominant negative mutant forms of TRAF-2, TRAF-6, Ras, Raf, and Erk2 or the respective empty expression vectors were cotransfected as indicated in the figures. After transfection, cells were maintained in medium containing 1% FCS to downregulate serum-activated signaling pathways. Cells were lysed 24 h posttransfection in reporter lysis buffer, and luciferase activity was assayed as mentioned above. Flotation gradients. | One day prior to transfection, Cos7 cells were plated in 75-cm2 tissue culture flasks (Costar) and transfected the following day at similar50% confluency with Effectene (3.1 mug of DNA per 75-cm2 flask) in accordance with the manufacturer's (Qiagen) instructions. At 48 h posttransfection, cells were washed once with PBS, trypsinized, resuspended in PBS, centrifuged, washed with 10 ml of PBS, counted, and pelleted. The cell pellets were resuspended in the respective volumes of ice-cold TNE buffer (10 mM Tris-HCl [pH 7.4], 150 mM NaCl, 5 mM EDTA plus protease inhibitors) containing 1% Triton X-100 to adjust the cell concentration to 107 cells/ml for all samples. Cells were lysed for 30 min on ice, homogenized with a 200-mul pipette (10 strokes), and centrifuged for 10 min at 900 x g at 4C. One milliliter of the supernatant was mixed with 1 ml of 85% sucrose in TNE and pipetted onto the bottom of an ultracentrifuge tube (14 by 95 mm; Beckman). Six milliliters of a 35% sucrose solution in TNE was layered on top, followed by 3.5 ml of a 5% sucrose solution in TNE. The samples were subjected to ultracentrifugation at 200,000 x g in an SW40 rotor for 22 h at 4C, and 1-ml fractions were collected starting at the top of the gradient. The pellet was resuspended in 200 mul of TNE containing 1% Triton X-100. Equal amounts of fractions 1 to 12 (20 mul) were analyzed by SDS-PAGE and Western blotting (of the pellet fraction, only 4 mul was loaded). Fractions 4 and 5 contain lipid rafts and the proteins associated with them. Fractions 9 to 12 contain soluble and solubilized proteins of the nonraft membrane. RESULTS : Proteins encoded by the K15 gene and localization in lipid rafts. | The KSHV K15 gene encodes several membrane proteins with a varying number of transmembrane regions and a common cytoplasmic domain . The longest of the K15-encoded proteins, containing all eight exons, is predicted to give rise to a membrane protein of 489 aa with up to 12 transmembrane domains. When the K15 gene was transiently expressed in Cos7, HEK 293, or HEK 293-T cells, a protein of similar45 kDa was detected by immunoblotting with a polyclonal antiserum raised against the cytoplasmic domain of K15 (Fig. and data not shown). Similar results were obtained with two independently constructed expression vectors for the eight-exon K15 cDNA (see Materials and Methods) which were used throughout this study. In addition, we and others have recently described several splice variants of K15. The K15 splice variants examined in this study are shown in Fig. . Whereas K15 ex1/6-8 uses exons 1, 6, 7, and 8, the splice variants K15 ex1 as/6-8 and K15 ex1 as/4-8 use an alternative splice (as) in exon 1 together with exons 6, 7, and 8 or exons 4 to 8, respectively. This implies the use of a different reading frame in exon 1. These three splice variants are predicted to encode proteins with two to four transmembrane domains. When transiently expressed in Cos7 cells, proteins of approximately 33 to 35 kDa (K15 ex1/6-8), 34 and 21 kDa (K15 ex1 as/4-8), and 26 and 21 kDa (K15 ex1 as/6-8) were detected by immunoblotting with K15 polyclonal antiserum (Fig. , C, and D). The 33- to 35-kDa protein seen with K15 ex1/6-8 corresponds to the size expected for this cDNA , as do the 34- and 26-kDa proteins seen with K15 ex1 as/4-8 and K15 ex1 as/6-8, respectively (Fig. and D). The additional 21-kDa protein seen with K15 ex1 as/4-8 and K15 ex1 as/6-8 (Fig. and D) may be the result of alternative splicing, proteolytic processing, or internal initiation of translation. As predicted, the different K15 protein isoforms examined appeared to be membrane proteins, since subcellular fractionation experiments (data not shown) indicated that all K15 protein isoforms can be pelleted with cellular membranes after high-speed centrifugation (100,000 x g) of cellular extracts from which other organelles (e.g., nuclei, mitochondria) have been depleted by centrifugation at 1,000 x g. In addition, all K15 isoforms are located in lipid rafts, as shown by flotation experiments . Lipid rafts are membrane microdomains rich in cholesterol and sphingolipids that are resistant to solubilization with certain nonionic detergents like Triton X-100 at 4C. Hence, lipid rafts and proteins associated with them float up to a position of lower density (5 to 35%) in a sucrose gradient in the presence of 1% Triton X-100, whereas soluble proteins and solubilized membrane proteins remain at the bottom of the gradient at the higher sucrose concentration (35 to 42.5%). As shown in Fig. (top), the 45-kDa protein derived from the eight-exon K15 cDNA expression construct (K15 ex1-8; aa 1 to 489) could be detected in fractions 4 and 5 of the sucrose gradient, which represent the interface of 5 and 35% sucrose. These fractions correspond to the position of lipid rafts, as indicated by the presence of caveolin 1 (Fig. , bottom). Caveolin 1 is a component of caveolae, which represent a specialized subset of lipid rafts . K15 proteins of 33 to 35 kDa (splice variant K15 ex1/6-8 in Fig. ) and 21 kDa (splice variants K15 ex1 as/4-8 in Fig. and K15 ex1 as/6-8 in Fig. ) also floated to the same position in the sucrose gradient. Only a portion of the K15 proteins was incorporated into lipid rafts . On the basis of densitometry scanning of the blots shown in Fig. , we estimate that 21% of the 45-kDa immunoreactive K15 protein (K15 ex1-8) (excluding the material in the pellet fraction) was present in lipid rafts. For the smaller 33- to 35-kDa K15 ex1/6-8 splice variant, a comparable amount (26%) was found in lipid rafts, in contrast to only 5% in the case of K15 ex1 as/4-8 and a similar small proportion in the case of K15 ex1 as/6-8. It thus appears that the K15 ex1 as/4-8 and K15 ex1 as/6-8 proteins were incorporated less efficiently than the 45-kDa (K15 ex1-8) and 33- to 35-kDa (K15 ex1/6-8) isoforms. The eight-exon 45-kDa K15 protein (K15 ex1-8; aa 1 to 489) was also detected in fractions 10, 11, and 12 (sucrose concentrations of 35 to 42.5%) of the sucrose gradient, where solubilized membrane proteins would be expected (Fig. , top). This result is reminiscent of LMP1 and LMP2A, which are both localized to lipid rafts and also to nonraft cell membranes . The 45-kDa K15 protein could also be detected in the pellet fraction containing cytoskeleton-associated proteins (Fig. , top), which was also observed for the LMP1 protein but not for LMP2A by Higuchi et al. . A chimeric LMP1/K15 protein, containing the six LMP1 transmembrane domains fused to the cytoplasmic region of K15, was also found in lipid rafts and nonraft cell membranes . K15 protein binds to and is phosphorylated by members of the Src kinase family. | Since K15 resembles the LMP2A protein of EBV with respect to its genomic localization at the right end of the genome, its protein structure with 12 putative transmembrane domains, the presence of putative binding sites for SH2 and SH3 motifs in the cytoplasmic domain , and its localization in lipid rafts, we examined whether K15 has functional characteristics in common with LMP2A. It has previously been shown that K15 is constitutively phosphorylated on the tyrosine within the distal putative SH2-binding motif Y481EEVL . Since LMP2A is phosphorylated at its ITAM by members of the nonreceptor protein tyrosine kinase (PTK) family, we investigated whether K15 is a substrate for these kinases. The complete cytoplasmic C-terminal domain of K15 fused to GST (GST-K15355-489) and the deletion mutant form GST-K15355-472 lacking the distal SH2-binding motif (Y481EEVL) were expressed in E. coli and purified as described in Materials and Methods. The purified fusion proteins were analyzed by SDS-PAGE (data not shown), and equal amounts of protein were used as the substrate in an in vitro kinase assay with immunoprecipitated myc-tagged PTKs Src, Fyn, Lck, Hck, and Yes from transiently transfected HEK 293-T cells. Figure shows that all of the kinases examined phosphorylated the C-terminal domain of K15 (GST-K15355-489, Fig. , top, upper part), whereas the GST-K15355-472 fusion protein lacking the Y481EEVL motif was not phosphorylated (Fig. , bottom, upper part), nor was unfused GST protein . Expression of the myc-tagged PTKs was examined with an anti-c-myc antibody . Since the sequence deleted in the mutant form GST-K15355-472 contains two threonine residues (T474 and T478) that could be targets for phosphorylation by other cellular kinases, we probed the Western blots of the kinase assay reaction mixtures with an anti-phosphotyrosine antibody . The result shows that GST-K15355-489, but not GST-K15355-472, was phosphorylated on tyrosine, indicating that Y481 in the YEEVL SH2-binding motif was phosphorylated by the PTKs examined. To investigate the binding of the C-terminal K15 domain to PTKs, we used a GST pulldown approach. GST-K15355-489 protein expressed in E. coli was used to bind the myc-tagged PTKs Src, Hck, Fyn, Lck, and Yes from transiently transfected HEK 293-T cells. GST alone served as a negative control. As shown in a Western blot probed with anti-c-myc antibody, all of the PTKs examined could be precipitated with the GST-K15355-489 protein . In order to identify the binding site of the PTKs within the K15 cytoplasmic domain, serial deletion mutant forms of GST-K15355-489 were constructed and used in a GST pulldown experiment. As shown in Fig. , all of the PTKs analyzed still bound strongly to all deletion mutants lacking, respectively, the distal SH2-binding motif Y481EEVL (GST-K15355-472), the region of aa 439 to 489 (GST-K15355-438), and the PPLP motif (aa 387 to 390) together with the YASIL motif (aa 431 to 435) (GST-K15355-373). As described previously , Y481EEVL, Y431ASIL, and P387PLP are conserved between the K15 P and M variants and could be candidates for SH3- and SH2-binding motifs. Our results suggest that the most proximal region of the cytoplasmic domain of K15 contributes to the binding of Src kinases but do not exclude a role for the PPLP or YASIL motif in this interaction. Amino acids 473 to 489 of the C terminus of K15 are required for TRAF-2 but not for TRAF-1 and -3 binding. | We have previously shown that the carboxy-terminal domain of K15 (aa 355 to 489) is able to interact with TRAF-1, -2, and -3 in a GST pulldown assay . In order to map the binding site for TRAF-1, -2, and -3 in K15, we used GST-K15355-489 and GST-K15355-472 in a pulldown experiment with detergent-solubilized extracts from HEK 293 cells that had been transiently transfected with Flag-tagged cDNA constructs of TRAF-1, -2, or -3. The results confirmed our previous observation that the carboxy-terminal region of K15 interacts with TRAF-1, -2, and -3, with TRAF-3 binding more strongly than TRAF-1 and -2 (Fig. , top). Deletion of aa 473 to 489 abolished the binding of TRAF-2 to K15 but not the binding of TRAF-1 and -3 (Fig. , bottom). These results were further confirmed by coimmunoprecipitation experiments with the chimeric construct LMP1-K15355-489 , which was cotransfected with Flag-tagged cDNA constructs expressing TRAF-1, -2, or -3 into 293 cells . At 48 h after transfection, proteins were immunoprecipitated with the anti-K15 polyclonal serum raised against the C-terminal domain of K15 and immunoblotted with an anti-Flag monoclonal antibody. The results shown in Fig. mirror those of the GST pulldown experiment, with TRAF-3 being most efficiently immunoprecipitated with the LMP1-K15355-489 chimera, followed by TRAF-1 and -2 (Fig. , lanes 1 to 3). When LMP1-K15355-489 was replaced with the deletion mutant form LMP1-K15355-472, binding of TRAF-2 was lost, while TRAF-1 and -3 were still efficiently precipitated (Fig. , lanes 4 to 6). These results suggest a strong interaction of TRAF-1 and -3 with aa 355 to 472 and a weaker, possibly indirect, interaction of TRAF-2 with aa 472 to 489 of K15. Activation of the NF-kappaB pathway by the 45-kDa K15 protein involves Y481. | The ability of TRAF-1, -2, and -3 to interact with K15 resembles their binding to CTAR-1 and -2 of LMP1, where they provide a link to downstream events that result in activation of the NF-kappaB pathway . We investigated whether the different K15 isoforms can activate NF-kappaB by cotransfecting corresponding expression vectors (Fig. and Materials and Methods) together with a luciferase reporter plasmid containing three NF-kappaB sites into HEK 293-T cells. As shown in Fig. , K15 ex1-8, expressing the longest K15 isoform (aa 1 to 489), showed five- to sixfold activation, whereas the smaller K15 isoforms (K15 ex1/6-8, K15 ex1 as/6-8, and K15 ex1 as/4-8) did not significantly increase basal NF-kappaB activity. LMP-1 induced approximately 30-fold activation of this reporter in our experiments (data not shown). In comparison, we observed only moderate (threefold) NF-kappaB activation with the LMP1-K15355-489 chimeric protein . Protein expression of the different constructs in HEK 293-T cells was comparable on Western blots (not shown). These results indicate a contribution of the transmembrane regions of K15 to NF-kappaB activation, which can also partly be provided by the LMP1 transmembrane regions. A deletion construct lacking aa 473 to 489, including the Y481EEVL motif, and the K15 point mutant form of Y481->F481 did not activate NF-kappaB , indicating that phosphorylation of Y481, most likely by members of the PTK family, is required for initiation of NF-kappaB signaling. The activity of the AP-1 transcription factor is induced by K15. | Mitogenic transcription factor AP-1 is a dimer of Jun-Jun or Jun-Fos family proto-oncoproteins, and it has been shown that LMP1 is an inducer of AP-1 activity in HEK 293 cells . To test whether K15 can activate the AP-1 transcription factor, HEK 293-T cells were transiently cotransfected with 50 ng of the reporter plasmid pRTU14, which consists of a luciferase reporter gene under the control of a minimal promoter and four AP-1-binding sites (TREs), and increasing amounts of different K15 expression constructs , or plasmid pSV-LMP1 coding for the LMP1 protein as a positive control. The results depicted in Fig. show that full-length K15 ex1-8 (aa 1 to 489) was a more potent inducer of AP-1 activity than was LMP1, with K15 activating AP-1 up to seven- to eightfold and LMP1 activating AP-1 up to three- to fourfold in a dose-dependent manner compared to cells transfected with control plasmids. No increase in luciferase activity was measured with the reporter control plasmid pRTU1 lacking the TREs (data not shown). AP-1 activation of the K15 deletion mutant form lacking the Y481EEVL motif (K15 ex1-8 Delta473 to 489) and the point mutant form K15 ex1-8 Y481F was significantly reduced. This indicates that phosphorylation of tyrosine Y481 of the Y481EEVL motif is important for K15-induced AP-1 activation. Interestingly, the LMP1-K15355-489 chimera did not induce AP-1 activity. The three K15 splice variants (K15 ex1/6-8, K15 ex1 as/6-8, and K15 ex1 as/4-8; Fig. and ) showed no clear-cut activation although they contain the entire K15 cytoplasmic domain. This may suggest that the regions encoded by exons 2 to 5 are also important for induction of AP-1 activity. A K15 protein induces the kinase activity of the MAPKs Erk2 and JNK1 but not that of p38. | After we had established that K15 is able to induce the AP-1 transcription factor, the next step was to elucidate the signaling pathways that mediate the signaling effects of the K15 protein. Three major MAPK pathways have been described to date that regulate AP-1 activity in the cell: the MEK1/2-Erk1/2 pathway, the MKK4/7(SEK1)-JNK1 cascade, and the MKK3/6-p38 pathway . LMP1 induces the activity of the transcription factor AP-1 specifically via the JNK1 pathway but does not induce the activity of the Erk2 kinase . Furthermore, LMP1 induces the activity of the MAPK p38 . In order to investigate the ability of the longest K15 isoform to induce any of these pathways, we examined the activation of the MAPKs Erk2, JNK1, and p38 in in vitro immunocomplex kinase assays. LMP1 was included in the JNK1 and p38 kinase assays as a positive control. As depicted in Fig. , the 45-kDa K15 protein (K15 ex 1-8; aa 1 to 489) was able to induce the kinase activity of cotransfected HA-tagged Erk2 kinase in transiently transfected HEK 293-T cells as measured by phosphorylation of the myelin basic protein (MBP). In contrast to the strong activation of Erk2 kinase by K151-489 , we only observed weak activation of JNK1 in cotransfected HEK 293-T cells . As expected, the positive control LMP1 strongly induced the kinase activity of cotransfected HA-JNK1 in HEK 293-T cells . Equal amounts of HA-tagged Erk2/JNK1 being immunoprecipitated and used for the kinase assay reactions were confirmed by Western blotting with antibodies to Erk2/JNK1 protein (Fig. and B, middle parts). The activation of the Erk2 and JNK1 kinase activities in HEK 293-T cells induced by the K15 protein depended on its intact cytoplasmic domain (Fig. and B), since no Erk2 or JNK1 activities were observed upon cotransfection with either the deletion construct K15 ex1-8 Delta473 to 489 or the point mutant form K15 ex1-8 Y481F. This suggests that phosphorylation of Y481, presumably by members of the PTK family, is crucial for K15 signaling activity in these two MAPK pathways. In similar cotransfection experiments with K15 expression vectors and HA-p38 kinase in HEK 293-T cells followed by an in vitro kinase assay with recombinant ATF-2 as the substrate, we did not observe an effect of K15 proteins on p38 kinase activity, in contrast to LMP1, which activated this pathway (data not shown; , ). These findings demonstrate that the 45-kDa K15 protein specifically induced the classical MEK1/2-Erk2 MAPK pathway and, but to a lesser extent, the SEK1-JNK1 pathway but not the MKK-p38 cascade. Dominant negative mutant forms of TRAF-2, Ras, Raf, and Erk2 and the MEK1/2-specific inhibitors PD98059 and UO126 reduce K15-induced activities of Erk2 and AP-1. | Since we have demonstrated that K15 bound TRAF-2 via aa 472 to 489 of its C-terminal domain and that this region is crucial for induction of NF-kappaB and AP-1 activity by K15, we wanted to elucidate whether a dominant negative mutant form of TRAF-2 could block K15-induced Erk2 and AP-1 activation. In order to further examine the downstream signaling routes linking K15 to Erk2, we also tested if coexpression of dominant negative mutant forms of Ras (Ras N17), Raf, or Erk2 could inhibit K15-induced Erk2 and AP-1 activation in immunocomplex kinase assays and luciferase-based AP-1 reporter experiments . In addition, we tested the effects of two chemical MEK1/2-specific inhibitors, PD98059 and UO126 , on the induction of Erk2 and AP-1 by K15 . MEK1/2 is the upstream kinase that phosphorylates and thereby activates Erk1/2. Transient coexpression of dominant negative TRAF-2 with HA-Erk2 and full-length K15 reduced K15-induced Erk2 activity, as shown by decreased phosphorylation of the Erk2 substrate MBP compared to coexpression with the empty TRAF-2 vector pRKH5 in HEK 293-T cells . An even more marked inhibition by dominant negative TRAF-2 was seen in the AP-1 luciferase-based reporter assay . In order to show the specificity for TRAF-2 and since TRAF-6 is involved in LMP1-induced activation of the p38 MAPK pathway , we also analyzed the effect of a dominant negative TRAF-6 mutant form on K15-induced AP-1 activation. Figure shows that dominant negative TRAF-6 had no effect on AP-1 induction by K15 compared to cotransfected pcDNA3.1 vector. Dominant negative Ras N17 inhibited K15-induced Erk2 and AP-1 activation, as shown in the immunocomplex kinase assay and the reporter assay , in contrast to the empty expression vector pCis2. The dominant negative Raf mutant form reduced MBP phosphorylation by Erk2 in the immunocomplex kinase assay but showed only moderate inhibition in the AP-1 reporter assay compared to the empty expression vector KRSPA (data not shown). In the AP-1 reporter assay, the dominant negative Erk2 mutant form (Erk C3, Y185->F185) was able to inhibit AP-1 activation by K15, in contrast to KRSPA . In addition to dominant negative mutant forms of the components of the Ras/Erk pathway, we also found that two chemical inhibitors of MEK1/2, PD98059 and UO126, inhibited K15-induced Erk2 activity in an immunocomplex kinase assay . Likewise, in the AP-1 reporter assay, both PD98059 and UO126 at 50 muM inhibited K15-induced AP-1 activity by approximately 50 to 60% (data not shown). These results suggest that K15 signals via the Ras/Erk pathway involving the components Ras, Raf, MEK, and Erk2 and that TRAF-2 may be involved in linking this pathway to K15 phosphorylated on Y481. DISCUSSION : We show here that the major isoform of K15-derived proteins, containing all eight exons and presumably 489 aa, activated the Ras/MAPK pathway, as well as the NF-kappaB pathway and, more weakly, the MAPK JNK. Activation of these pathways depended on Y481 in a carboxy-terminal YEEVL motif that is reminiscent of the YEEA motif surrounding Y112 of EBV LMP2A. In B cells, Y112 of LMP2A is phosphorylated by, and important for the recruitment of, the Src kinases Lyn and Fyn . We could show that several Src kinases, including Src, Fyn, Yes, Hck, and Lck, could phosphorylate Y481 of K15 in vitro . However, sequences located between aa 355 and 373, at the beginning of the cytoplasmic domain of K15, appear to contribute to the interaction of Src kinases with the K15 cytoplasmic domain, since a GST-K15 fusion protein containing this region still bound to all of the Src kinases tested . This observation does not exclude an additional role for the PPLP or YASIL motif in Src kinase binding. A Y481F mutant form abolished the ability of the major K15 protein to activate the Ras/MAPK pathway, as well as JNK and NF-kappaB (Fig. , , and ), indicating that phosphorylation of Y481, presumably by members of the Src kinase family, represents a first step in the activation of these pathways. Constitutive phosphorylation of K15 Y481 in B cells has been reported by others , although the kinases involved were not identified. This group also found that a CD8-K15 chimeric protein could inhibit B-cell receptor-initiated signaling in the B-cell line BJAB but was unable to initiate Ca2+ mobilization after cross-linking with an antibody to CD8 . In our study, the nature of the transmembrane anchor appeared to be important for the ability of K15-derived proteins to initiate different signaling pathways, since only the longest isoform, containing 12 putative transmembrane domains , was capable of initiating efficient signaling (Fig. and ). Even the six transmembrane domains of LMP1, which can mediate efficient clustering of LMP1 in the cell membrane as a prerequisite for constitutive intracellular signaling , were unable to induce activation of an AP-1 reporter when fused to the cytoplasmic region of K15 (Fig. and ) and only showed moderate activation of an NF-kappaB reporter compared to the longest K15 isoform . Although we did not examine Ca2+ mobilization as a consequence of K15-initiated signaling, we think that it is therefore possible that the CD8-K15 chimeric protein used by Choi et al. may have been also unable to trigger these signaling pathways. Lipid rafts serve as platforms in cellular membranes that can incorporate and recruit membrane-associated molecules required for the assembly of signaling cascades (reviewed in reference ). We therefore examined whether the longest K15 isoform (which is capable of inducing intracellular signaling) differs from the smaller isoforms (which are not) with respect to their incorporation into lipid rafts. We found that at least the isoform K15 ex1/6-8, which contains the complete first exon, appeared to be incorporated into lipid rafts to roughly the same extent as the longest K15 isoform (Fig. and B). In contrast, the two isoforms K15 ex1 as/6-8 and K15 ex1 as/4-8, which use an alternative splice acceptor in exon 1 and therefore contain only 6 aa derived from the first exon in a different reading frame , may be incorporated less efficiently into lipid rafts (Fig. and D). Since the LMP1-K15 chimeric protein was also incorporated into lipid rafts , it appears as if differential incorporation into lipid rafts is not responsible for the fact that these membrane proteins differ in the ability to initiate intracellular signaling. Sequences encoded by K15 exons 2 to 5 may therefore contribute to intracellular signaling by providing sites for posttranslational modification or by interacting with other intracellular components. Is there a functional role for the smaller K15 isoforms? The study by Choi et al. suggests that a signaling-inactive K15 variant could inhibit intracellular signaling pathways. It is therefore conceivable that the smaller K15 isoforms may play a role in modulating the effect of the eight-exon isoform on the activation of the pathways reported here. This would represent a parallel to the presumed role of the LMP2 splice variant LMP2B, which is devoid of the amino-terminal intracellular domain of LMP2A, but could modify LMP2A-induced signaling by affecting the clustering of LMP2A molecules in the cell membrane. RT-PCR studies of PEL cell lines suggest that the transcript encoding the longest K15 isoform is more strongly expressed than those encoding the smaller isoforms. A recent report attributed the presence of an approximately 23-kDa K15-derived protein in PEL cell lines to internal proteolytic processing of the approximately 45-kDa (exons 1 to 8) K15 precursor protein. We have seen proteins of a similar size in PEL cells that are, however, not associated with cellular membranes, in contrast to the approximately 21-kDa proteins translated from K15 splice variants and shown in Fig. to be associated with lipid rafts (Brinkmann et al., unpublished data). Whether the balance in virus-infected cells is therefore in favor of the 45-kDa activating K15 protein or in favor of the smaller nonactivating forms is unclear and may be difficult to predict from the relative abundance of individual transcripts. In activating the Ras/MAPK pathway, the K15 eight-exon protein resembles LMP2A, which has recently been reported to activate Erk in HEK 293 cells, thereby increasing the stability of c-jun , although others have not seen activation of Ras/MAPK by LMP2A in epithelial cells . We did not investigate whether one of the downstream effects of Ras/MAPK activation by K15 is the increased stability of c-jun but found moderately increased JNK activity following transfection of K15. It remains a possibility that this weak activation of the JNK pathway occurs indirectly, perhaps as a consequence of the activation of other pathways, such as the Ras/MAPK pathway. Our preliminary studies of cDNA arrays also indicate that the expression of c-jun may be increased at the transcriptional level following K15 transfection (data not shown). There may therefore be multiple routes for LMP2A and K15 to increase the activity of the JNK/c-jun pathway which would contribute to the increased activity of the AP-1 transcription factor following K15 expression, as reported here. Our results also suggest an involvement of TRAF-2 in the K15-induced activation of the Ras/MAPK pathway, since a dominant negative mutant form of TRAF-2 inhibited the activity of Erk2 and AP-1 in K15-transfected cells . We also observed weak binding of TRAF-2 to the cytoplasmic domain of K15, which was abolished by deletion of the last 17 aa, including the Y481EEVL motif, as well as the sequence ATQPTDD, which appears to be distantly related to the TRAF-binding consensus site PxQxS/T (residues identical to those of CTAR-1 of LMP1 are underlined) . In contrast, the stronger interaction with TRAF-1 and -3 seems to occur through a more membrane-proximal region of K15. It is therefore conceivable that the interaction of the K15 cytoplasmic domain with TRAF-2 occurs in an indirect manner. TRAF-2 mediates the activation of the p38/MAPK pathway, as well as the activation of NF-kappaB induced by LMP-1, CTAR-1 and -2, and possibly JNK . The effect of LMP1 on p38/MAPK also involves TRAF-6 downstream of TRAF-2 , and TRAF-6 has recently been shown to be involved in the interleukin-1-induced activation of p38/MAPK via Ras and in the CD40-induced activation of Erk via a Ras-independent pathway . However, K15-induced AP-1 activity was not affected by cotransfection of a dominant-negative mutant form of TRAF-6 , suggesting that other intracellular factors may link the cytoplasmic domain of K15, phosphorylated on Y481, to the activation of the Ras/MAPK pathway. This issue, and the identity of cellular and viral genes affected by K15-induced Ras/MAPK activation, is currently under investigation. FIG. 1. : ORF K15 expression constructs used in this study and their putative protein products. ORF K15 expression constructs used in this study and their putative protein products. The K15 ORF is multiply and alternatively spliced (as). The major transcript identified in PEL cells by RT-PCR is fully spliced and contains all eight exons (K15 ex1-8; aa 1 to 489). It encodes a membrane protein with up to 12 transmembrane domains and a cytoplasmic C-terminal domain (aa 355 to 489). The C-terminal domain contains motifs reminiscent of SH2, SH3, and TRAF-like binding sites. The distal TRAF-like binding site and the distal SH2-binding motif Y481EEVL are deleted in construct K15 ex1-8 Delta473 to 489. K15 ex1-8 Y481F carries a point mutation in the distal SH2-binding motif (Y481->F481EEVL). The LMP1-K15355-489 chimera was constructed by fusing the six transmembrane domains of LMP1 to the cytoplasmic C-terminal end (aa 355 to 489) of K15. The splice variants K15 ex1/6-8, K15 ex1 as/6-8, and K15 ex1 as/4-8 differ in the number of transmembrane domains they contain, but all contain the C-terminal domain. FIG. 2. : The 45-kDa K15 protein (aa 1 to 489) is associated with lipid rafts. The 45-kDa K15 protein (aa 1 to 489) is associated with lipid rafts. Cos7 cells were transiently transfected with K15 expression construct K15 ex1-8 (A); the natural splice variant K15 ex1/6-8 (B), K15 ex1 as/4-8 (C), or K15 ex1 as/6-8 (D); or the LMP1-K15355-489 chimera (E). At 48 h after transfection, Cos7 cells were lysed in TNE buffer and extracts were analyzed on a flotation sucrose gradient as described in Materials and Methods. After ultracentrifugation, 1-ml fractions were collected starting at the top of the gradient and analyzed by Western blotting with a rabbit antibody to K15. Lipid raft-associated proteins are localized at the interface of 5 and 35% sucrose (fractions 4 and 5). Soluble proteins and solubilized membrane proteins of the nonraft plasma membrane are localized to the higher-density sucrose fractions (fractions 10 to 12, 35 to 42.5% sucrose). P, pellet. (A, top) The 45-kDa form of K15 ex1-8 (aa 1 to 489) localizes to lipid rafts (fractions 4 and 5, interface of 5 and 35% sucrose) and to the fractions containing solubilized membrane proteins (fractions 10 to 12, 35 to 42.5% sucrose). (A, bottom) Western blotting of the samples from the upper part of this panel probed with anti-caveolin 1. Endogenous caveolin 1 served as a positive control for lipid raft localization. (B) The 33- to 35-kDa protein derived from splice variant K15 ex1/6-8 is localized in the lipid raft fraction and high-density sucrose fractions (fractions 8 to 11). *, possible dimeric form. (C) The similar21-kDa protein derived from splice variant K15 ex1 as/4-8 (see text) is found in lipid rafts and high-density sucrose fractions. (D) The similar21-kDa protein derived from K15 ex1 as/6-8 is localized in lipid rafts and also in fractions containing solubilized membrane proteins (see text). (E) The LMP1-K15355-489 chimera is found in lipid rafts and high-density sucrose fractions. FIG. 3. : The tyrosine residue of the distal SH2-binding motif (Y481EEVL) of the K15 C-terminal domain is phosphorylated by members of the Src kinase family of PTKs. The tyrosine residue of the distal SH2-binding motif (Y481EEVL) of the K15 C-terminal domain is phosphorylated by members of the Src kinase family of PTKs. Purified GST-K15355-489 and GST-K15355-472 proteins were used as substrates in in vitro kinase assays with immunoprecipitated (IP) myc-tagged PTKs and [gamma-32P]ATP. (A) The PTK family members Src, Fyn, Yes, Lck, and Hck phosphorylate the C-terminal domain of K15 in an in vitro immunocomplex kinase assay (top, upper part). When the last 17 aa encompassing the SH2-binding motif (Y481EEVL) of the C terminus were deleted (GST-K15355-472), no phosphorylation was observed (bottom, upper part). The expression level of the myc-tagged immunoprecipitated Src kinases was detected with an anti-c-myc antibody (top, middle part, and bottom, lower part). The amounts of GST-K15355-489 fusion protein and nonfused GST protein used in the in vitro kinase reactions were also analyzed with a Coomassie-stained SDS gel (top, lower part). (B) The kinase reaction mixtures shown in panel A were analyzed by Western blotting with an anti-phosphotyrosine antibody. Expression of the GST-K15 fusion proteins was detected with an anti-K15 monoclonal antibody. FIG. 4. : The nonreceptor PTKs Src, Fyn, Yes, Lck, and Hck bind to the C-terminal domain of K15. The nonreceptor PTKs Src, Fyn, Yes, Lck, and Hck bind to the C-terminal domain of K15. In a GST pulldown experiment, the GST-K15355-489, GST-K15355-472, GST-K15355-438, and GST-K15355-373 fusion proteins immobilized on glutathione Sepharose beads were incubated with lysates of HEK 293-T cells transfected with expression vectors for myc-tagged PTKs. Bound PTKs were detected by Western blotting with an antibody to c-myc. FIG. 5. : Interaction of TRAF-1, -2, and -3 with the carboxy-terminal domain of K15. Interaction of TRAF-1, -2, and -3 with the carboxy-terminal domain of K15. (A) Flag-tagged cDNA expression constructs of TRAF-1, -2, and -3 were transiently transfected into HEK 293 cells, and the interaction of the TRAF proteins with the carboxy-terminal domain of K15 (aa 355-489) was examined by GST pulldown assays as described in Materials and Methods. (Top) Interaction of TRAF with the entire carboxy-terminal domain of K15. Lanes: 1 to 3, TRAF-1-transfected cells; 4 to 6, TRAF-2-transfected cells; 7 to 9, TRAF-3-transfected cells; 1, 4, and 7, TRAF proteins bound to GST-K15355-489; 2, 5, and 8, proteins bound to GST; 3, 6, and 9, input cell lysates. (Bottom) Interaction with the carboxy-terminal region of K15 from which the terminal 17 aa, containing the distal TRAF-like binding site, have been deleted (GST-K15355-472). Samples were loaded as in the upper part of this panel. (B) HEK 293 cells were transiently cotransfected with Flag-tagged TRAF-1 (lanes 1 and 4), TRAF-2 (lanes 2 and 5), and TRAF-3 (lanes 3 and 6) expression constructs together with the LMP1-K15355-489 (lanes 1 to 3) or LMP1-K15355-472 (lanes 4 to 6) chimera. At 48 h after transfection, detergent-extracted cell lysates were immunoprecipitated (IP) with rabbit anti-K15 serum and analyzed by Western blotting, followed by staining with an anti-Flag monoclonal antibody. FIG. 6. : The 45-kDa K15 protein activates the transcription factor NF-kappaB in a luciferase-based reporter assay. The 45-kDa K15 protein activates the transcription factor NF-kappaB in a luciferase-based reporter assay. HEK 293-T cells were transiently cotransfected with 50 ng of the luciferase reporter plasmid p3EnhkappaBconA-Luc containing three NF-kappaB binding sites upstream of the luciferase gene and different amounts of K15 expression constructs (200 and 500 ng of DNA). At 24 h after transfection, cells were lysed and analyzed for luciferase activity. Shown is the relative fold activation compared to that of empty-vector (mock)-transfected (200 and 500 ng of DNA) cells based on triplicate samples. Equal expression levels of K15 proteins were analyzed by Western blotting with rabbit K15 antiserum (data not shown). FIG. 7. : Transcription factor AP-1 is activated by K15. Transcription factor AP-1 is activated by K15. HEK 293-T cells were transiently cotransfected with the AP-1 luciferase reporter plasmid pRTU14 containing four AP-1-binding sites upstream of the luciferase gene and different amounts of K15 expression constructs (100, 200, 500, and 1,000 ng). After transfection, cells were grown in medium containing 1% FCS, lysed after 24 h, and analyzed for luciferase activity. Shown is the relative fold activation compared to that of mock-transfected (100, 200, 500, and 1,000 ng) cells based on triplicate samples. Equal expression levels of K15 proteins were analyzed by Western blotting with rabbit K15 antiserum (data not shown). FIG. 8. : Immunocomplex kinase assays with MAPKs Erk2 and JNK1. Immunocomplex kinase assays with MAPKs Erk2 and JNK1. HEK 293-T cells were transiently cotransfected with 1 mug of the HA-tagged MAPK Erk2 or JNK1 and 1 mug of the K15 or LMP1 expression construct. After transfection, cells were grown in medium containing 1% FCS and lysed after 24 h in TBS-T buffer containing phosphatase inhibitors. The cell lysates were then subjected to immunoprecipitation with HA-antibody 12C5 coupled to protein G beads overnight. MBP or purified GST-c-Jun fusion protein served as the substrate in the in vitro kinase reaction mixture with Erk2 (A) or JNK1 (B), respectively. The immunoprecipitated (IP) kinase was incubated with its substrate and [gamma-32P]ATP for 30 min at 25C in kinase reaction buffer and subsequently analyzed by autoradiography (top of panels A and B). Western blots of the kinase reactions were probed with specific Erk2 or JNK1 antibodies to ensure equal expression levels of the MAPKs (middle of panels A and B). Cell lysates were analyzed for equal expression of K15 constructs by Western blotting and probing with a K15 antibody (bottom of panels A and B). FIG.9. : Dominant negative-negative (d. Dominant negative-negative (d.n.) mutant forms of TRAF-2, Erk2, Raf, and Ras and the MEK1/2 inhibitors UO126 and PD98059 reduce the Erk2 and AP1 activation induced by K15. (A) Seven hundred nanograms of a dominant negative mutant form of TRAF-2, Ras, Raf, or Erk2 or the respective empty expression vector, pRKH5, pCis2, or pKRSPA, was cotransfected with 1 mug of HA-Erk2 and 1 mug of K15 ex1-8 (aa 1 to 489) or empty K15 expression vector (mock) where indicated. Erk2 kinase activity was monitored by phosphorylation of MBP (top). The Western blot probed with anti-Erk2 antibody shows equal expression levels of the Erk2 kinase for all samples. (B) HEK 293-T cells were cotransfected with 1 mug of HA-Erk2 and 1 mug of K15 ex1-8 (aa 1 to 489) expression construct or empty K15 expression vector. At 14 h after transfection, DMSO alone or the MEK1/2 inhibitor PD98059 or UO126 was added to the medium at 50 muM from a 50 mM stock solution in DMSO. Cells were incubated for a further 8 h before protein extraction. In vitro kinase assays were performed as described in the legend to Fig. and Materials and Methods. The Western blot probed with anti-Erk2 antibody shows equal expression levels of the Erk2 kinase for all samples. (C) AP-1 luciferase-based reporter assay. HEK 293-T cells were transiently cotransfected with 50 ng of the AP-1 reporter plasmid (pRTU14), 500 ng of the K15 ex1-8 (aa 1 to 489) expression construct, and 350, 700, or 1,000 ng of a d.n. mutant form of Ras, TRAF-2, TRAF-6, or Erk2 or the respective empty expression vector, pCis2, pRKH5, pcDNA3.1, or pKRSPA. Shown are the luciferase values (in relative light units) of the different dominant negative mutant forms with respect to those of the corresponding empty expression vectors, which were set at 1. Backmatter: PMID- 12915573 TI - Point Mutations in Exon I of the Herpes Simplex Virus Putative Terminase Subunit, UL15, Indicate that the Most Conserved Residues Are Essential for Cleavage and Packaging AB - The herpes simplex virus UL15 and UL28 genes are believed to encode two subunits of the terminase involved in cleavage and packaging of viral genomes. Analysis of the UL15 protein sequence and its herpesvirus homologues revealed the presence of 20 conserved regions. Twelve of the twenty regions conserved among herpesviruses are also conserved in terminases from DNA bacteriophage. Point mutations in UL15 were designed in four conserved regions: L120N (CR1), Q205E (CR2), Q251E (CR3), G263A (CR3), and Y285S (CR4). Transfection experiments indicated that each mutant gene could produce stable UL15 protein at wild-type levels; however, only one mutant (Q251E) was able to complement the UL15-null virus. Each mutation was introduced into the viral genome by marker transfer, and all mutants except Q251E were unable to form plaques on Vero cells. Furthermore, failure to form plaques on Vero cells correlated with a defect in cleavage and packaging. Immunofluorescence experiments indicated that in cells infected with all mutant viruses the UL15 protein could be detected and was found to localize to replication compartments. Although wild-type and mutant Q251E were able to produce A, B, and C capsids, the rest of the mutants were only able to produce B capsids, a finding consistent with their defects in cleavage and packaging. In addition, all mutant UL15 proteins retained their ability to interact with B capsids. Therefore, amino acid residues 120, 205, 263, and 285 are essential for the cleavage and packaging process rather than for association with capsids or localization to replication compartments. Keywords: Introduction : Herpes simplex virus type 1 (HSV-1) DNA cleavage and packaging involves at least seven essential gene products: UL6, UL15, UL17, UL25, UL28, UL32, and UL33 (reviewed in references and ). This complex process has been compared to the better-studied bacteriophage, T4 and lambda, in which monomeric genomes are released from the concatemeric DNA by cleavage and subsequently packaged into preassembled capsids (reviewed in references and ). In the lambda phage system, gpA and gpNu comprise a two-subunit terminase that binds to and cleaves concatemeric DNA at specific sequences. The terminase, bound to the DNA, docks onto the prohead at a unique portal vertex. Translocation of the DNA into the capsid requires ATP hydrolysis. The DNA is cleaved at specific sites, and the terminase, with the remaining DNA still attached, detaches from the prohead . Other phage proteins such as gpD and gpW act to stabilize DNA in the capsids . Recent experiments with HSV indicate that the herpesviruses may encode similar cleavage and packaging machinery. Newcomb et al. recently reported that the HSV-1 UL6 protein localizes to a unique vertex on procapsids and forms a 12-member ring in vitro ; these results suggest that UL6 forms a unique portal through which the viral DNA enters the capsid. After packaging, UL25 may play a role in stabilizing DNA inside the capsid . UL17 and UL32 may act to transport capsids to replication compartments (RCs) that are believed to be the sites of cleavage and packaging . UL15 and UL28 are believed to comprise two subunits of the terminase . The function of UL33 is unknown , but it has recently been reported to associate with UL28 and UL15 . By analogy with phage, the terminase complex would be expected to bind viral DNA, transiently associate with the capsids at a unique portal vertex, carry out site specific cleavage of viral DNA, and translocate the DNA inside the capsids in an ATP-dependent fashion. Several lines of evidence support the notion that UL15 and UL28 make up two subunits of the terminase. Protein sequence analysis reveals homology between UL15 and gp17 (a subunit of the T4 terminase), including the Walker A and B boxes characteristic of ATP-binding domains . A point mutation in the Walker A box in the HSV-1 UL15 gene prevents cleavage and packaging, suggesting that ATP hydrolysis is required for the process . In vitro experiments indicate that UL28 and its homolog in human cytomegalovirus (HCMV) bind to a packaging (pac) site, which is believed to comprise part of the cleavage recognition sequence . Furthermore, UL15 and UL28 transiently associate with capsid intermediates and associate with each other as determined by coimmunoprecipitation and copurification . Under certain experimental conditions, UL28 requires the presence of UL15 to be transported to the nucleus . Mutations in UL15 and UL28 homologues from HCMV have been shown to promote resistance to the same antiviral compounds , indicating a possible interaction between the two proteins. Taken together, these lines of evidence suggest UL15 and UL28 may comprise two subunits of the putative terminase. We used sequence analysis and site-directed mutagenesis here to provide additional evidence that UL15 is part of the terminase complex and to elucidate regions of UL15 that are essential in the cleavage and packaging process. The homology alignments reveal regions that are highly conserved among UL15 homologues in the herpesvirus family and bacteriophage terminases. It seems likely that such evolutionarily conserved regions will be important for biological function. Mutations in exon I of the UL15 gene were engineered in highly conserved regions and then analyzed for biological activities exhibited by the wild-type UL15 protein, including (i) the ability to localize to RCs, (ii) participation in cleavage and packaging, and (iii) transient association with capsid intermediates. MATERIALS AND METHODS : Cells and viruses. | African green monkey kidney cells (Vero; American Type Culture Collection, Rockville, Md.) were propagated and maintained as described previously . MV cells are a Vero cell line expressing a full-length UL9 containing mutation G354A in motif V. MV cells potentiate HSV-1 infection and have been used in the present study to enhance viral infection (B. Martincheva et al., unpublished data). C2 cells, which express the UL15 gene products and are permissive for UL15 mutants, were propagated as described previously . The KOS strain of HSV-1 was used as the wild-type virus, and UL15-null viruses, hr81-1 and hr81-2, which contain the ICP6::lacZ mutagen cassette inserted in UL15 exon I and exon II, respectively, were described previously . Four point mutations in UL15 were introduced into the viral genome as described below. The G263A mutant with a mutation in the Walker A Box of UL15 was previously described . An Sf21 cell extract containing the UL15 protein expressed from a recombinant baculovirus was a kind gift from Daniel Tenney (Bristol-Myers Squibb). Alignments. | Using the HSV-1 UL15 protein sequence, a PSI-BLAST search of the NCBI database was performed to identify possible homologues . The eight human herpes UL15 homologues were aligned using CLUSTALW (Blosum algorithm) within MacVector . Using the lambda family bacteriophage (HK97) gp2 (terminase subunit) protein sequence, a PSI-BLAST search of the NCBI database was performed to identify homologues from other bacteriophage and the herpesvirus family . Homologues gp2 (HK97), UL15 (HSV-1), gp17 (kvp40), and gp17 (T4) and the consensus sequence from the above alignment were aligned using CLUSTALW (identity algorithm) within MacVector . Sequential PCR mutagenesis. | The G263A mutation in plasmid pcDNAUL15C (containing a CMV promoter and the genomic UL15 sequences, including exon I, UL16, UL17, and exon II) has been previously reported . In the present study, four new point mutations were introduced into the UL15 gene. For mutagenesis, pUC119-UL15GE1 plasmid, containing exon I of UL15, was used as the template for two-step PCR with the oligonucleotides listed in Table . In each case, the primers were designed to introduce a silent mutation resulting in the introduction of a new restriction enzyme site. The PCR products were subcloned into pUC119-UL15GE1, replacing the existing wild-type UL15 sequences. These mutant clones were confirmed by sequence analysis and detection of the new restriction site within the plasmid. Subsequently, exon I from the pUC119-UL15GE1 mutant plasmids was subcloned into pcDNA-UL15C, replacing the wild-type exon I. The pcDNA-UL15C mutant plasmids were used for transient-transfection and transient-complementation assays, and the pUC119-UL15GE1 plasmids were used for marker transfer. Transient transfection and complementation. | Vero cells were transfected with 3 mug of pcDNA-UL15C plasmids containing wild-type or mutant versions of the UL15 gene by use of Lipofectamine Plus according to the manufacturer's protocol (Gibco-BRL). Cells were harvested at 20 h posttransfection and subjected to Western blot analysis. Transient-complementation assays were performed as previously described . Briefly, Vero cells were transfected with 3 mug of pcDNA-UL15C plasmids containing wild-type or mutant UL15 genes driven by a CMV promoter and superinfected at a multiplicity of infection (MOI) of 3 PFU/cell with UL15-null virus (hr81-1 or hr81-2). Progeny virus were harvested and assayed on the complementing cell line (C2). The transient-complementation index was determined by dividing the viral titer resulting from the mutant plasmid by the viral titer resulting from the empty plasmid. Construction of UL15 mutants by marker transfer. | Marker transfer experiments were carried out as described previously . Vero cells were transfected with pUC119-UL15GE1 plasmids and infectious hr81-1 DNA (containing a lacZ insertion in exon I of the UL15 gene). Plaques were stained with neutral red in the presence of the chromogenic substrate X-Gal (5-bromo-4-chloro-3-indolyl-beta-d-galactopyranoside); white plaques were purified three times on C2 cells before viral stocks were prepared. Recovery of the desired mutant was verified by Southern blot analysis to detect the introduction of new restriction sites within the viral genome. Preparation of cell extracts. | Vero cells, infected with KOS or UL15 mutant viruses at an MOI of 10 PFU/cell, were harvested and prepared for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis as previously described , except that samples were sonicated prior to boiling. Capsid preparation. | Capsid preparation was essentially as previously described except that six large (225-cm2) flasks of MV cells were infected at an MOI of 5 PFU/cell for each virus. Sucrose gradients (20 to 50%) were prepared with a Gradient Master (Biocomp). Sucrose gradient fractions were collected with a piston gradient fractionator (Biocomp). After fraction collection, the samples were precipitated with trichloroacetic acid, and the pellets were washed with ethanol and then resuspended in loading buffer (200 mM Tris [pH 8.8], 1% [vol/vol] 2-mercaptoethanol, 2% [wt/vol] SDS, 10% [vol/vol] glycerol, 0.001% [wt/vol] bromphenol blue). The samples were boiled for 5 min and stored at -20C. Western blot analysis. | Samples (10 to 30 mul) were subjected to SDS-PAGE on 10 or 12% polyacrylamide gels and transferred to enhanced chemiluminescence (ECL) membranes. To detect UL15 gene products, alphaAS9 rabbit polyclonal antibody (generously provided by Daniel Tenney at Bristol-Myers Squibb), was used at a 1:2,000 dilution in TBST (10 mM Tris [pH 8.0], 150 mM NaCl, 0.05% Tween) overnight at 4C . To detect VP5 gene products, alpha-HSV MCP ICP5 monoclonal antibody (ABI, Columbia, Md.) was used at a dilution of 1:2,000 in TBST overnight at 4C. Membranes were developed either by ECL (Amersham) or by using alkaline phosphatase (AP; Promega or Bio-Rad) as instructed by the manufacturer. DNA isolation and Southern blot hybridization. | Vero cells (1.5 x 106) were infected at an MOI of 10 PFU/cell for 18 h and harvested, and the total DNA was isolated, as described previously . DNA was then digested with BamHI and subjected to electrophoresis and Southern blot analysis . Membranes were probed with a 32P-labeled BamHI SQ junction fragment . Immunofluorescence and imaging. | Immunofluorescence was carried out as described by Burkham et al. , except as noted. Cells were fixed in 1.5% formaldehyde in phosphate-buffered saline (PBS) for 15 min and permeabilized in cold acetone at -20C for 2 min. The primary antibodies were applied for 3 h; 39S, a monoclonal antibody (ICP8) , was used in 3% normal goat serum (NGS) in PBS (1:200) to detect RCs, and alphaAS9, a polyclonal antibody in 3% NGS in PBS (1:150), was used to detect UL15. Secondary antibodies, i.e., fluorescein isothiocyanate-conjugated goat anti-mouse and Texas red-conjugated goat anti-rabbit in 3% NGS in PBS (1:200), were applied for 1 h. Images were arranged and labeled by using Adobe Photoshop 5.0 and Illustrator 7.0. RESULTS : UL15 is highly conserved among herpesviruses. | As previously reported, the HSV-1 UL15 protein sequence is highly conserved within the Herpesviridae family . To initiate a structure-function analysis of UL15, we identified potentially important residues by extending the sequence comparison between HSV-1 UL15 and other Herpesviridae family members. More than 50 UL15 homologues from human and animal herpesviruses were identified. Figure shows an alignment of UL15 homologues from all eight human herpes viruses. Twenty highly conserved regions (CRs) were designated within the HSV-1 UL15 protein coding sequence and its homologues. These regions were chosen on the basis of the following criteria: a region must either exhibit 70% homology (similarity and identity) among all eight human herpesviruses or it must exhibit 70% homology to a CR of a bacteriophage terminase. Nineteen CRs were identified that met the first criterion (CR2 to CR20). CR1 exhibited 64% homology among the herpesviruses, but a subregion of CR1 revealed 70% similarity compared to bacteriophage terminases (see below). Although it would have been possible to combine some of the CRs into larger regions, this was not done because we wanted to distinguish between CRs that demonstrated homology to bacteriophage terminases and those that did not. The high degree of sequence conservation probably indicates important regions for biological function. Of the 20 CRs, the most highly conserved are CR3 and CR6 containing the Walker A and B boxes, respectively, which are found in ATP-binding proteins. CRs of UL15 are also conserved in terminases of bacteriophage. | Previous protein sequence analysis suggested a relationship between the UL15 protein of channel catfish herpes virus and the gp17 protein encoded by bacteriophage T4 . More recent work has suggested homology between HSV-1 UL15 and terminases in the T4 family . This homology was confined primarily to the Walker A and B boxes that make up the ATP-binding domain. Although the previous analysis of amino acid sequence conservation between bacteriophage terminases has not revealed extensive homology outside the ATP-binding domain, we have now extended this analysis by using the identity algorithm within the CLUSTALW program. The large subunit of the terminase from several members of the lambda and T phage families, including gp2 (HK97), gp17 (T4), and gp17 (KVP40), were aligned with HSV-1 UL15 . In confirmation of previous results, the proteins were homologous in the Walker A and B box regions but, in addition, 10 other regions of homology were identified. These include CRs 1, 3, 4, 6, 8, 9, 13, 14, 16, 17, 19, and 20 . This high degree of homology suggests an evolutionary relationship between phage and herpesvirus terminases and may help to identify key residues within the HSV-1 UL15 that are essential for terminase function. Point mutations in exon I. | In order to determine whether the CRs in UL15 play a role in terminase function, point mutations were designed in the first four CRs, three of which are also conserved in bacteriophage. In the present study we report five mutations: one in CR1 (conserved in bacteriophage and herpesviruses), one in CR2 (conserved in herpesviruses only), two in CR3 (one conserved in herpesviruses and one conserved in bacteriophage), and one in CR4 (conserved in bacteriophges and herpesviruses) . Our intention was to identify residues that are either intrinsically important for terminase activity in general or are specifically important for herpesviruses. Mutations were engineered into the pcDNA plasmid containing the genomic version of the UL15 gene driven from the CMV promoter. To determine whether the altered genes could produce stable full-length UL15 protein, Vero cells were transfected with mutant and wild-type versions of the UL15 gene in the pcDNA-UL15 plasmid. At 20 h posttransfection, cells were harvested and subjected to Western blot analysis. Figure shows that wild-type and mutant proteins were expressed at similar levels. This indicates that all mutant UL15 proteins were relatively stable when expressed in the absence of other viral proteins. Four of the five exon I mutants exhibited severe growth defects. | To determine whether the mutant UL15 genes could complement the UL15-null viruses (hr81-1 and hr81-2), transient-complementation assays were performed as described in Materials and Methods. Vero cells were transfected with the pcDNA-UL15C plasmids containing wild-type or mutant versions of the UL15 gene and superinfected with either hr81-1 or hr81-2; the titers of the progeny were determined for the complementing cell line . Only the wild-type and mutant Q251E proteins were able to efficiently complement both exon I and exon II UL15-null viruses. As previously reported, the G263A mutation was unable to complement UL15 null mutants (36; data not shown). In addition, the L120N, Q205E, and Y285S mutations failed to complement UL15 null mutants in this assay . Three mutants ---Q205E, G263A, and Y285S ---exhibited complementation indices significantly less than that of the empty plasmid (36). This may indicate that these mutant proteins exert a transdominant effect. Further experiments will be necessary to explore this possibility. To determine the effect of UL15 mutations in the context of viral infections, each UL15 mutant gene was transferred into the viral genome by marker transfer as described in Materials and Methods. Progeny viruses were plaque purified three times and verified by gain of a restriction site within the UL15 gene, and viral stocks were made. Only mutant Q251E virus was able to form plaques on noncomplementing Vero cells . All of the mutants (L120N, Q205E, Q251E, G263A, and Y285S) were able to form plaques on the complementing cell line, C2, indicating that the growth defects were due to a UL15 mutation and not to a second mutation in another gene. Mutant UL15 proteins were detected in infected cell lysates. | To determine whether stable protein could be produced in the context of viral infection, Western blot analysis of cells infected with each mutant virus was performed as described in Materials and Methods. Antibody to VP5 was used as an internal control to demonstrate that efficient infection had occurred. Figure shows that all mutant viruses (lanes 2 to 7) produced wild-type (lane 1) levels of capsid protein VP5 (marked with an arrowhead). Figure shows that all mutants except Y285S (lane 7) produced wild-type levels of UL15. Interestingly, Y285S could produce stable protein after overexpression by transfection (see Fig. , lane 6). The Y285S mutant protein could, however, be detected either in infected cells by immunofluorescence or in capsids by sucrose gradient centrifugation and Western blot analysis (described below). Thus, the Y285S protein is made in transfected and infected cells; however, in the context of infection may be somewhat unstable. On the other hand, the synthesis of wild-type levels of mutant proteins by viruses L120N, Q205E, Q251E, and G263A indicates that no gross alterations in protein formation leading to global instability have occurred (Fig. , lanes 2, 3, 4, and 5, respectively). Cleavage but not DNA synthesis was inhibited by UL15 mutants. | Previously, it has been shown that UL15-null viruses are able to produce viral DNA but not able to carry out cleavage and packaging . To determine whether viruses bearing point mutations were also impaired in DNA cleavage, Vero cells (1.5 x 106) were infected with wild-type and mutant viruses at an MOI of 10 PFU/cell. Total DNA was isolated and subjected to Southern blot analysis, as described in Materials and Methods. The cleavage of the viral DNA was detected by the presence of S and Q fragments . All viruses were able to produce viral DNA as assessed by the presence of the joint SQ fragments. On the other hand, however, S and Q fragments were only observed in DNA harvested from cells infected with wild-type and mutant Q251E viruses (Fig. , lanes 1 and 4). This indicates that genomic cleavage does not occur in cells infected with the remaining mutant viruses (L120N, Q205E, G263A, Y285S, and hr81-1) (Fig. , lanes 2, 3, 5, 6, and 7, respectively). All UL15 mutant proteins were detected in RCs. | We and others have shown that viral gene expression, DNA replication, and encapsidation occur in large globular domains in the nucleus called RCs . RCs are detected with antibody to the major single-stranded DNA-binding protein (ICP8). It was previously reported that wild-type UL15 localized to RCs within the nucleus of infected cells . Immunofluorescence was performed to determine whether mutant forms of UL15 behaved like the wild type with respect to nuclear localization. In cells infected with wild-type virus, UL15 was detected in RCs, confirming previous results (Fig. , row 1). In cells infected with UL15-null virus (hr81-1), RCs were present, as detected by anti-ICP8 antibody, but UL15 was not detected (Fig. , row 2). In cells infected with all mutant viruses, UL15 was detected in RCs. Cells infected with mutant Q205E, are shown in Fig. , row 3, as a representative example. Thus, although UL15 protein from cells infected with mutant Y285S was undetectable by Western blot, UL15 was clearly detectable by immunofluorescence, although it exhibits a somewhat fainter staining pattern than that observed in cells infected with wild-type virus (data not shown). This finding suggests that all mutant proteins are synthesized, folded properly, and correctly transported to the RCs, but in the case of mutant Y285S, the UL15 proteins maybe somewhat unstable. Transient capsid association of UL15 was maintained in all of the UL15 mutant viruses. | Multiple bands that react with the UL15 antibody, alphaAS9, were previously reported to associate with capsids. The 81-kDa UL15 protein was found to be associated more strongly with B capsids than with C capsids . We wanted to examine whether the UL15 point mutations would alter the ability of the 81-kDa UL15 protein to bind to capsids. Western blot analysis with alphaAS9 (UL15) was performed on capsids isolated from MV cells infected with mutant or wild-type viruses. Capsid bands were purified by sucrose gradient centrifugation and collected with a piston gradient fractionator (Biocomp) as described in Materials and Methods. Figure shows Western blots of sucrose gradient fractions for wild-type (KOS) and mutant viruses. In the top panel for each virus, VP5 was detected, indicating which fractions contained capsids. This was confirmed visually as fraction samples were collected (data not shown). Mutant Q251E was able to produce A, B, and C capsids similar to wild type . All other mutants produced only B capsids similar to the UL15-null virus (hr81-1). In the bottom panel for each virus, the 81-kDa UL15 protein was detected by use of antibody alphaAS9 . The wild-type and Q251E 81-kDa UL15 proteins were found to associate more strongly with B capsids than C capsids. The UL15-null virus, hr81-1, did not produce UL15 protein. UL15 protein produced in cells infected with mutant viruses L120N, Q205E, G263A, and Y285S was also found to associate with B capsids . This suggests that CRs 1 to 4 are not essential for the association of UL15 protein with B capsids. In summary, it appears that the failure of mutants L120N, Q205E, G263A, and Y285S to cleave and package viral DNA is not due to the inability of UL15 to associate with B capsids. DISCUSSION : HSV-1 DNA cleavage and packaging is a complex process that bears many similarities to the cleavage and packaging process in the better-studied bacteriophage. By analogy with bacteriophage, HSV-1 appears to encode a terminase that is essential for encapsidation and the production of progeny virus. HSV-1 UL15 has been identified as a putative component of the terminase. By using sequence analysis, we were able to identify regions in the HSV-1 UL15 protein sequence conserved among homologues in the herpesvirus family. Even more striking was the observation that 12 of the 20 CRs were also found in HK97, T4, and Kvp40 bacteriophage terminases. This analysis extended the original work of Davison et al. and Mitchell et al. showing that UL15 was similar to T4 gp17 primarily in the region of the Walker A and B boxes of the ATP-binding domain . We propose that these evolutionarily conserved regions may be important for the biological functions of the terminase. In the present study, we introduced point mutations into the HSV-1 UL15 coding gene to determine which regions are important for function. By analogy with bacteriophage systems, if UL15 is a component of the terminase, we might expect that it takes part in several functions during encapsidation, including binding and transporting concatemeric DNA to the capsid, docking at the unique portal vertex of the capsid, cleavage and packaging of the viral DNA, and disassociation from the capsid. Direct biochemical analysis of the putative terminase subunits, UL15 and UL28, has been hampered by the insolubility of these proteins expressed in heterologous expression systems. Expression systems tested to date include insect cells infected with recombinant baculoviruses and bacteria transformed with expression vectors (D. Tenney, unpublished data). In addition to the alignment data described above, the strongest evidences to date that UL15 and UL28 are components of the terminase complex are their ability to transiently associate with B capsids during encapsidation ; the requirement of UL6, the portal protein , for the transient association; and the ability of UL28 to bind to pac sequences . It was anticipated that the introduction of subtle mutations in CRs of UL15 may result in mutants with a partial phenotype that would be able to carry out some but not all of the functions of the putative terminase. For instance, it might be possible to isolate a mutant that could cleave at the UL terminus but not the US terminus. Such data would support the assignment of UL15 as a component of the terminase. Several key residues and regions of UL15 were identified that are essential for cleavage and packaging. Mutations in amino acids 120 (CR1), 205 (CR2), 263 (CR3), and 285 (CR4) resulted in the inability to cleave viral DNA, indicating that these residues (and possibly the CRs they are a part of) may be important for this aspect of terminase activity. On the other hand, all mutant UL15 proteins were able to localize to RCs and to associate with B capsids. Thus, the localization to RCs and association with capsids are functions separable from the ability to participate in cleavage and packaging. These point mutations may affect downstream activity such as initiation of packaging, cleavage of the first termini, translocation of the DNA into the capsid, and the ATP hydrolysis necessary for the translocation of the DNA. The mutant G263A (CR3) results in a change in the invariant glycine in the Walker A box of the ATP-binding domain of the UL15 protein. This mutant protein may be defective in the binding and/or hydrolysis of ATP. Another mutation in CR3, Q251E, behaves like wild-type virus, indicating that a conservative change in this CR is tolerated. It is possible that a less conservative change at residue 251 would not be tolerated. Although highly conserved among herpesvirus homologues, amino acid 251 was not conserved among bacteriophage. This may indicate that amino acid 251 is not essential for terminase function per se. Mutant Y285S (CR4) is defective in cleavage and packaging. Interestingly, mutant Y285S protein was detected by immunofluorescence and was found to be associated with capsids but was not readily visible on Western blots of cells infected with mutant virus. The mutation at amino acid 285 may cause the mutant protein to be somewhat unstable in infected cell lysate; however, it may be stabilized by its association with capsids. Alternatively, mutant UL15 may be enriched in the capsid preparation, perhaps by binding better than wild-type UL15 to capsids or by binding irreversibly. It will be of interest to determine whether this mutant exerts a transdominant effect. Because of the apparent instability of the mutant protein in cell extracts, we cannot say whether the defect in cleavage and packaging is a result of the specific mutation in CR4 or due to the relative instability of the protein. The presence of the mutant protein in capsids, however, may indicate the former. Mutants L120N, Q205E, G263A, and Y285S fail to cleave and package viral DNA despite the fact that they can localize and associate with capsids in a fashion similar to wild type. These results suggest that residues in CRs 1 to 4 are involved in the DNA cleavage reaction. In summary, the mutational analysis described here has resulted in the identification of three separable functions of UL15: localization to RCs, association with capsids, and cleavage of viral DNA. The similarity of the cleavage phenotype in mutants in CRs 1 to 4 indicates that this reaction may be carried out in a synchronized fashion involving multiple regions of the protein. FIG. 1. : UL15 homology alignments: UL15 is highly conserved among the herpesvirus family. UL15 homology alignments: UL15 is highly conserved among the herpesvirus family. A PSI-BLAST search was performed on the HSV-1 UL15 protein sequence to obtain all its homologues. The eight human herpesvirus homologues were aligned using CLUSTALW (Blosum algorithm) within MacVector. We assigned 20 CRs between UL15 and its homologues based on the fact that many but not all of these regions are also conserved in bacteriophage terminase proteins (see Fig. ). The Walker A and B boxes of the putative ATP-binding domain are indicated. FIG. 2. : UL15 and phage alignments. UL15 and phage alignments. A PSI-BLAST search was performed on gp2 from HK97. UL15 (HSV-1), the UL15 consensus sequence from Fig. [UL15(con)], gp2 (HK97), gp17v (KVP40), and gp17 (T4) were aligned by using CLUSTALW (identity algorithm) within MacVector. Of the 20 CRs shown in Fig. , 12 are also conserved between HSV and the phage terminases. The numbers above each CR correspond to the CRs shown in Fig. . FIG. 3. : UL15 detected by Western blot analysis. UL15 detected by Western blot analysis. (A) Western blot analysis of transiently transfected cells. Vero cells (2.1 x 106) transfected with a pcDNA plasmid containing a wild-type or mutant version of the UL15 gene were harvested 20 h posttransfection, subjected to SDS-PAGE on a 12% gel, and transferred to an ECL membrane. Western blot analysis was performed by using alphaUL15 (1:2,000) and AP Immunestar development (Bio-Rad). The arrowhead indicates the band corresponding to the 81-kDa UL15 protein (lane 1). (B and C) Western blot analysis of infected cell lysates. Vero cells (2.1 x 106) were infected with wild-type or mutant virus at an MOI of 10 PFU/cell and harvested 20 h postinfection. The samples were subjected to SDS-PAGE on a 12% gel and then transferred to an ECL membrane. Western blot analysis was performed by using alphaVP5 (1:2,000) and ECL development (Amersham Pharmacia) (B) and alphaUL15 (1:2,000) and AP development (Promega) (C). In panel B, the arrowhead indicates the position of VP5. In panel C, the arrowhead indicates the position of the 81-kDa form of UL15. An extract from insect cells infected with a UL15 expressing baculovirus was used as a control for the presence of the 81-kDa form of UL15 (lane 9). The positions of molecular mass markers (in kilodaltons) are shown in lane 1. FIG. 4. : Southern blot analysis of viral DNA from cells infected with wild-type and mutant viruses. Southern blot analysis of viral DNA from cells infected with wild-type and mutant viruses. (A) The HSV-1 genome consists of two unique regions, UL and US, which are flanked by repeated sequences a, b, and c. The subscript n in "an" indicates multiple copies of the a sequence. The BamHI fragments corresponding to S, Q, and SQ junction fragments are indicated. (B) Vero cells (107) were infected with the indicated virus at an MOI of 10 PFU/cell for 18 h. Total DNA was digested with BamHI, subjected to electrophoresis, and transferred to a GeneScreen Plus membrane. The membrane was probed with a 32P-labeled BamHI SQ junction fragment. Arrows indicate the junction fragments SQ, Q, and S (from top to bottom). DNA from cells infected with wild-type and null mutant virus (hr81-1) are shown in lanes 1 and 7, respectively. Lanes 2 to 6 contained DNA from cells infected with mutants L120N, Q205E, Q251E, G263A, and Y285S, respectively. Although mutants L120N, Q205E, and hr81-1 appear to contain a reduced amount of SQ DNA compared to wild type and the other mutants, this difference was not reproducible, and we do not feel that it reflects a true reduction in levels of viral DNA synthesis. FIG. 5. : Immunofluorescence analysis. Immunofluorescence analysis. Vero cells were infected with wild-type virus (KOS) (top row), UL15-null virus (hr81-1) (middle row), or the UL15 mutant virus Q205E (bottom row) at an MOI of 10 PFU/cell and fixed for immunofluorescence at 8 h postinfection. Infected cells were stained with anti-ICP8 monoclonal antibody (39S) (left panels, green) and anti-UL15 polyclonal antibody (alphaAS9) (middle panels, red). The right-hand column labeled Merge shows the merged images. The color was adjusted on the hr81-1 panels to demonstrate the lack of UL15 in these cells. FIG. 6. : Western blot analysis of wild-type and mutant capsids. Western blot analysis of wild-type and mutant capsids. MV cells were infected with the indicated virus at an MOI of 5 PFU/cell for 20 h. Sucrose gradient fractions were collected as described in Materials and Methods. Gradient fractions for each virus were subjected to SDS-PAGE on a 10% gel, and Western blot analysis was performed by using alphaAS9 (alphaUL15) and AP development (Promega) as described in Materials and Methods. The alphaAS9 antibody cross-reacts with VP5 proteins, and this cross-reaction aids in the detection of fractions that contain capsids (top panel for each virus). We previously reported an 87-kDa protein which associated with C capsids that was detected by alphaAS9 . In the experimental conditions used in the present study, the 87-kDa protein was not observed. The bottom panel for each virus shows the position of the 81-kDa form of UL15. Fractions containing A, B, and C capsids are marked with brackets. TABLE 1 : Primers used in two-step PCR TABLE 2 : Designed point mutations in the UL15 protein TABLE 3 : Transient-complementation assay TABLE 4 : Virus growth assay Backmatter: PMID- 12915531 TI - Competition between the Sendai Virus N mRNA Start Site and the Genome 3'-End Promoter for Viral RNA Polymerase AB - The genomic and antigenomic 3'-end replication promoters of Sendai virus are bipartite in nature and symmetrical, composed of le or tr sequences; a gene start or gene end site, respectively; and a simple hexameric repeat. The relative strengths of these 3'-end promoters determines the ratios of genomes and antigenomes formed during infection and whether model mini-genomes can be rescued from DNA by nondefective helper viruses. Using these tests of promoter strength, we have confirmed that tr is stronger than le in this respect. We have also found that the presence of a gene start site within either 3'-end promoter strongly reduces 3'-end promoter strength. The negative effects of the gene start site on the 3'-end promoter suggest that these closely spaced RNA start sites compete with each other for a common pool of viral RNA polymerase. The manner in which this competition could occur for polymerase off the template (in trans) and polymerase on the template (in cis) adds insight into how the viral RNA polymerase switches between its dual functions as transcriptase and replicase. Keywords: Introduction : Sendai virus (SeV), a Respirovirus of the Paramyxovirinae subfamily, is a model nonsegmented negative-strand RNA virus. The first step in the SeV replication cycle is the production of mRNA from a helical subviral structure, the nucleocapsid (NC), in which the genome RNA is tightly and stoichiometrically associated with the viral NC protein (N protein). This N-subunit assembly, together with multiple attached viral polymerases (a complex of the P and L proteins) is the minimum subviral unit that is thought to retain infectivity . The synthesis of RNA genomes (or [+] antigenomes) and their assembly with N protein is coupled, and these viral RNAs are only found as nucleocapsids . Electron micrographs of SeV nucleocapsids show a flexible helical assembly with 13 N subunits per turn and variable pitch, in which each N subunit binds 6 nucleotides (nt) . For viruses of the Paramyxovirinae, efficient replication of model mini-genomes in transfected cells requires that their total length be a multiple of six, and viruses of this group whose genome has been entirely sequenced mostly have genome lengths that are multiples of six . Inefficient replication of non-hexamer-length mini-genomes was not due to the lack of encapsidation of the mini-genome but apparently was due to the inability of viral RNA-dependent RNA polymerase (vRdRP) to initiate at the mini-nucleocapsid 3' end . It has been suggested that nucleocapsid assembly begins with the first 6 nt at the 5' end of the nascent chain and continues by assembling 6 nt at a time until the 3' end is reached. The efficiency of the 3'-end promoter then presumably depends on the position of these cis-acting sequences relative to the N subunits, and this hexamer or N-subunit "phase" is determined by the precise length of the genome chain . The requirement for hexameric genome length of the Paramyxovirinae has recently been underscored for human parainfluenza virus type 2 (hPIV2), a rubulavirus, and measles virus, a morbillivirus. hPIV2 recovered from non-hexameric-length cDNAs were found to contain a biased distribution of mutations that all restored hexameric genome length . Similarly, an obligatorily diploid measles virus was recovered from DNA that encodes a defective H protein, again via mutations that restore hexameric genome length . The genomic and antigenomic promoters (G/Pr and AG/Pr) of the Paramyxovirinae are found within the terminal 96 nt or 16 N subunits of each RNA and are composed of leader (le) or trailer (tr) sequences; a gene start or gene end site, respectively; and a simple hexameric repeat (Fig. ; see also Fig. ). Although G/Pr and AG/Pr are similar in overall structure, they carry out separate functions. G/Pr is associated with le RNA, antigenome RNA, and mRNA synthesis, while AG/Pr is associated with tr RNA and genome RNA synthesis. The relative abundance of the various viral RNAs throughout the infection is regulated in large part via G/Pr and AG/Pr, and this ensures as well that a preponderance of genomes are produced during infection. These 3'-end promoters are (at least) bipartite in nature . There is both an end element comprising ca. the first 30 nt at the 3' ends (in which the first 12 nt are conserved between G/Pr and AG/Pr and across each genus), and a downstream element within the 5' untranslated region of the N gene ( nucleocapsids) or the 3' untranslated region of the L gene ([+] nucleocapsids). For SeV and hPIV3 and apparently all respiro- and morbilliviruses, the downstream element is a simple but phased hexameric repeat (3' [C1N2N3N4N5N6]3, bound to the 14th, 15th, and 16th N protein subunits [, ]). For rubulaviruses like SV5, [N1N2N3N4G5C6]3 is repeated in subunits 13, 14, and 15, such that all these conserved nucleotides are adjacent to the first 12 nt in the helical nucleocapsid with 13 subunits per turn . This common surface, comprising the conserved sequences of both elements of the bipartite promoter on two turns of the helix, is thought to mediate the initial interaction of vRdRP with resting nucleocapsids to initiate RNA synthesis at the genome 3' end (see below). The remarkable sequence simplicity of the downstream promoter element is presumably possible because its hexamer phase also contributes to its activity . This phase effect is thought to stem from the different chemical environments of the 6 nt bases associated with each N subunit, as revealed by chemical attack studies of resting SeV nucleocapsids . Adenosines in any hexamer phase are largely protected from dimethyl sulfate, whereas cytosine reactivity is highly variable and strongly depends on hexamer phase. Cytosines are highly reactive only in hexamer positions 1 and 6, precisely the positions of the conserved cytosines in the downstream element of G/Pr and AG/Pr . If this correlation is not coincidental, this suggests that the conservation of hexamer phase of this cis-acting sequence serves to make these cytosines accessible to vRdRP in resting NCs. Another feature strictly conserved among respiro-, morbilli-, and rubulaviruses is that the first (N) mRNA starts precisely opposite U56 (underlined) within the decameric gene start signal 3' 55A/UCCCA NUUUC/N66 (for SeV, hexamer phase indicated). We refer to this mRNA initiation signal as gene start 1 (gs-1). Curiously, none of the conserved cytosines of gs-1 are in hexamer positions 1 and 6. However, once vRdRP has initiated RNA synthesis at the genome 3' end and begun to elongate, we assume that conformational changes occur within N:RNA that alter vRdRP's interaction with the nucleotide bases, including gs-1. Since SeV RdRP is minimally a coiled-coil tetramer of P and a single L protein (with an aggregate molecular mass equivalent to 8 N-subunits) and this polymerase initiates RNA chains at two such closely spaced sites at the genome 3' end (only 56 nt apart), one would expect these two events to interfere with each other, at least under some conditions. However, detailed studies involving modified DI mini-genomes did not uncover any indication that gs-1 affected G/Pr strength; only the first 30 nt of le/tr regions were found to influence this property. We now report studies that show that in some cases exactly the same situation applies, namely, that only the le/tr sequences appear to influence promoter strength. More importantly, we find other cases in which gs-1 clearly decreases 3'-end promoter strength. The manner in which gs-1 negatively influences G/Pr strength is discussed. MATERIALS AND METHODS : Generation of recombinant SeV (rSeV). | Briefly, one 9-cm-diameter dish of BSR-T7 cells that endogenously express T7 RNA polymerase were transfected with 1.5 mug of pGEM-L, 5 mug of pGEM-N, 5 mug of pGEM-PHA (which does not express C proteins), and 15 mug of the various pSeV infectious constructs. After 24 h the cells were scraped into their medium and injected directly into the allantoic cavity of 9-day-old embryonated chicken eggs. Three days later, the allantoic fluids were harvested and reinjected undiluted into eggs. For further passages, the viruses were diluted 1/500 before injection. The presence of viruses was determined by pelleting allantoic fluids through a TNE (10 mM Tris-HCl [pH 8.0], 100 mM NaCl, 1 mM EDTA)-25% glycerol cushion for 20 min at 14,000 rpm in an Eppendorf 5417C centrifuge. Virus pellets were resuspended in sample buffer, and the proteins were separated by sodium dodecyl sulfate-10% polyacrylamide gel electrophoresis and stained with Coomassie brilliant blue. Virus infection and passage. | A549 or BSR T7 cells were infected at a multiplicity of infection (MOI) of 10 in Dulbecco's modified Eagle medium. After 1 to 2 h of absorption, the inoculum was removed and replaced with fresh medium containing 10% fetal bovine serum. Virus supernatant from 48-h-infected cells was cleared by filtration, treated for 1 h with trypsin (1.2 mug/ml), and used to reinfect fresh A549 or BSR T7 cells. Real-time PCR. | A 9-cm-diameter dish of A549 cells was infected with SeV at an MOI of 10. After 24 h, medium was removed, the cells were scraped, and total RNA was extracted using Trizol. One-tenth of this RNA was mixed with 0.5 mug of specific primer and converted to cDNA with Moloney murine leukemia virus reverse transcriptase for 1 h at 37C. One-tenth of the cDNA was used for real-time PCR, using TaqMan Universal Master mix and the ABI Prism 7700 sequence detector. Several dilutions of cDNA were tested to construct a standard curve that was used to obtain relative numbers for the samples. DI particle rescue with helper virus. | Dishes of BSR-T7 cells (diameter, 9 cm) were infected with either wild-type SeV (SeV-wt) or SeV-AGP55 at an MOI of 10. One hour later, the medium was removed and the cells were transfected with 5 mug of the DI-encoding plasmid. Twelve hours later the medium was replaced with fresh Dulbecco's modified Eagle medium containing 10% fetal bovine serum. The ND and DI particles present in the culture medium were then passaged on fresh cells as described above. DI particle rescue with plasmids. | Dishes of BSR-T7 cells (diameter, 9 cm) were transfected with 1.5 mug of pGEM-L, 5 mug of pGEM-N, 5 mug of pGEM-PHA (which does not express C proteins), and 5 mug of the various DI-encoding plasmids. RESULTS : The ratio of genomes to antigenomes was previously determined by (32)P-primer extension analysis. This ratio can also be determined by real-time reverse transcriptase PCR (RT-PCR) analysis, as follows (and as described in Materials and Methods). Four pairs of primers that span gene junctions are used; one primer of each pair specifically extends on genomes and lies upstream of the junction, and the other specifically extends on antigenomes and lies downstream of the junction. The four genome-specific and four antigenome-specific primers are used in individual reactions to synthesize cDNA with RT. Samples of the RT reaction mixture are then used in real-time PCR amplification along with the reverse primer located across the junction. This approach eliminates the contribution of mRNAs to the amount of antigenomes determined (only discistronic mRNAs that are very rare in SeV infections will be scored). At least three sets of primers (and fluorescent probes) were used for each determination, and the average genome/antigenome ratio from at least two infections of each SeV is shown. The ratios determined by RT-PCR agree well with those determined by primer extension analysis (in parentheses in Fig. ), and require considerably less viral RNA as starting material. Relative to SeV-wt infections, where ca. 10 times as many genomes as antigenomes are found intracellularly, SeV-AGP55 infections (in which the entire 55 nt of tr is replaced with the le sequence) contain equal levels of the two . (AGP refers to a virus or DI particle. When "AGP" is followed by a single number, as in AGP55, this indicates that positions 1 to 55 of AG/Pr have been exchanged with le sequences.) SeV-AGP65 infections, which contain gs-1 in addition to the le sequence at AG/Pr, yielded similar results . The same genome-to-antigenome ratios were found in virions (data not shown), since, similar to rabies virus infections , SeV genomes are not being selectively exported during virion assembly . Genome/antigenome ratios intracellularly should then simply reflect the frequencies with which vRdRP productively initiates at the template 3' ends and begins to encapsidate nascent le/tr RNAs, to produce genome or antigenome nucleocapsids. The efficiency of this process is one measure of promoter strength. By this criterion, the replacement of the tr region of AG/Pr with the le sequence alone (SeV-AGP55) strongly decreases AG/Pr strength. Since SeV-AGP55 replicates to levels similar to those of SeV-wt, it appears that G/Pr is inherently weak only in the presence of the stronger AG/Pr. Thus, as described for rabies virus infections , competition between G/Pr and AG/Pr for vRdRP appears to be the primary determinant of the relative amounts of SeV genomes and antigenomes formed The genome/antigenome ratio of 1 in AGP55/65 infections suggests that promoter strength is primarily determined by tr versus le sequences themselves, in agreement with previous studies of DI genome replication . However, it is also possible to construct SeV in which the first 42 or 48 nt of the le sequence of G/Pr is replaced with the equivalent tr sequences (SeV-GP42/48) without seriously affecting viral mRNA synthesis . SeV-GP42 and SeV-GP48 are notable in that their infections fail to induce apoptosis. Nevertheless, SeV-GP42/48 infections still accumulate 10 times as many genomes as antigenomes intracellularly, like SeV-wt (only SeV-GP42 is shown in Fig. ). This result was unexpected because, if promoter strength is determined primarily by the nature of first 30 nt of le/tr sequence, the modified G/Pr of SeV-GP42/48 should have been equivalent in strength to AG/Pr. Possible explanations for why G/Pr of SeV-GP42/48 remains weaker than AG/Pr include (i) the remainder of the le region (positions 49 to 55) is specifically important for G/Pr strength in the context of nondefective infections and (ii) in some situations, the presence of gs-1 within G/Pr does indeed reduce G/Pr strength, despite the presence of 48 nt of tr sequence. To directly examine this question, a matched series of rSeV isolates were prepared that bridged the transition from SeV-wt to SeV-AGP65 . Except for the ambisense SeV (i.e., SeV that expresses ambisense mRNA) that grow less well than SeV-wt, all the other SeV grow similarly to SeV-wt; i.e., they all accumulate similar levels of viral macromolecules intracellularly. SeV-wt+MCS contains a cassette at the L gene end/tr junction with three elements, a poly(A)-stop site that terminates mRNAs transcribed from the antigenome 3' end, a multicloning site (MCS), and a duplicated L gene end site. SeV-wt+MCS infections produce genome/antigenome ratios of 10 like SeV-wt; thus, the inclusion of this cassette appears not to have affected relative promoter strengths. SeV-AGP56-65 is identical to SeV-wt+MCS except that the 10 nt adjacent to tr (i.e., the L gene end site) are converted to the complement of gs-1. Remarkably. the simple inclusion of these 10 nt (gs-1) in AG/Pr of AGP56-65 leads to transcription of a short (80 nt) ambisense mRNA from this antigenome that terminates just before the opposing L gene. SeV-AGP56-65 infections, however, now accumulate almost as many antigenomes as genomes. Thus, gs-1 placed within AG/Pr not only functions well but also apparently acts to equalize the two promoters, presumably by weakening AG/Pr. SeV-AGP52-65 and AGP48-65 differ from AGP56-65 by the further conversion of the 4 and 8 nt adjacent to the ambisense gene start site. The exchange of these nucleotides further increases the relative amounts of antigenomes, but only slightly. Thus, AG/Pr can be weakened relative to G/Pr in one of two ways: by exchanging the resident tr sequences for le sequences or by introducing an ectopic gs-1 within this 3'-end promoter. The introduction of gs-1 within AG/Pr automatically destroys the L gene end site. It thus remains possible that the destruction of this sequence is responsible (at least in part) for weakening AG/Pr. We therefore attempted to prepare derivatives of wt+MCS in which the L gene end site is mutated to something other than the complement of gs-1. The complement of the L gene end site (5' UAAGAAAAA) was mutated to 5' UGCCGCAUG and UUUCUUUUU. However, we were unable to prepare these rSeV in three separate attempts, and we sought another manner to examine this question. The mutated L gene end sites were built into mini-genomes expressing green fluorescent protein (GFP) , and these were recovered from DNA by coinfection with SeV-AGP55 (which contains two weak replication promoters [see below]). We were again unable to recover the 5' UGCCGCAUG mini-genome (DI-Ge-1; replicable genomes may not tolerate certain sequences here). However, DI-Ge-3 (5' UUUCUUUUU) could be recovered almost as well as the wt. When RNA from these mixed infections were examined for their ratio of DI genomes to antigenomes, DI-Ge-3 and DI-wt were found to contain the same excess of genomes over antigenomes (approximately threefold in this case; the ratio of DI genomes and antigenomes are not necessarily the same as that of ND genomes and antigenomes). Thus, the loss of the complement of the L gene end site per se does not alter AG/Pr strength. We conclude that it is the presence gs-1 within AG/Pr that is responsible for weakening this 3'-end promoter. Competition between DI genomes and ND helper genomes for vRdRP. | The relative amounts of genomes and antigenomes formed during infection is one measure of the relative strengths of G/Pr and AG/Pr of nondefective SeV. This manner of characterizing promoters stems from early studies of copy back DI genomes, which have deleted most of the genome and simultaneously exchanged >96 nt of G/Pr with the equivalent sequences of AG/Pr . Copy back DI genomes do not transcribe mRNAs and have a clear replicative advantage over ND virus. Moreover, in contrast to their ND helper virus, these coinfections accumulate equal amounts of copy back DI genomes and antigenomes. Further studies suggested that their replicative advantage did not stem from their shorter length, and that relative promoter strength is determined by the terminal 96 nt of G/Pr and AG/Pr. Internal-deletion (int-Delta) DIs that maintain the wt promoter constellation can also be generated on high-MOI passage of SeV stocks. However, these DIs are considerably more difficult to generate, and the reasons for their replicative advantage over their ND helper virus are unknown. In further contrast to copy back DIs, int-Delta DI genomes cannot be recovered from DNA using SeV-wt as helper, even though they are efficiently amplified when the replication substrates (N, P, and L) are provided from plasmids. More importantly, the same int-Delta DIs are efficiently recovered from DNA using SeV-AGP55 or AGP65 as helper virus (Fig. and ). Thus, although int-Delta DIs presumably have no advantage in competing with wt helper virus for limiting RdRP during recovery from DNA, int-Delta DIs can clearly compete with ND helpers whose AG/Pr is weakened (as defined by ND genome/antigenome ratios). We have adapted this system to study promoter strength by including a GFP transcription unit within int-Delta DIs. BSR T7 cells infected with various helper viruses are transfected with plasmids expressing various mini-antigenomes (see time line in Fig. ). The particles released 36 to 48 h postinfection are then used to infect fresh cultures that are examined by fluorescent microscopy for GFP expression. This simple test reveals whether the DI particles have been generated from plasmid DNA during the initial infection/transfection, and amplified by the helper virus during subsequent coinfection (P1) . Using this DI reporter system, we have reexamined whether gs-1 within G/Pr affects the ability of G/Pr to compete for vRdRP during coinfection. A series of mini-genome constructs that varied only in their G/Pr was used to determine what changes were required so that the int-Delta DI could be recovered and amplified by SeV-wt helper (5 top constructs, Fig. ). We first exchanged the first 42 or 55 nt of le of G/Pr with the equivalent tr sequences (GP42/wt and GP55/wt), but only background fluorescence was found at passage 1. We then added a further 12 nt of the trailer sequence but maintained the remainder of DI GP55/wt, so that the GFP expression was retained and now started at nt 68 rather than nt 56 (DI GP68+start/wt). However, this DI as well could not be recovered (in six attempts) with SeV-wt. Only when the mRNA start site was eliminated from G/Pr of DI GP55/wt, by exchanging it for the equivalent AG/Pr sequence (DI GP96/wt), was this DI recovered with SeV-wt. (The elimination of gs-1 of course also eliminates GFP expression, and DI RNA levels were determined by real-time RT-PCR [Fig. , bottom].) Thus, int-Delta DI genomes successfully compete for vRdRP with wt helper genomes only when the mRNA start site within G/Pr has been eliminated; simply moving gs-1 downstream by 12 nt (DI 68+start/wt) has no effect. Similar experiments were carried out with DI wt/AGP55, which contains le in place of tr sequences at AG/Pr (bottom three constructs in Fig. ). Just as DI-wt/wt cannot be rescued by coinfection with SeV-wt, DI wt/AGP55 cannot be rescued by coinfection with SeV-AGP55 (presumably because both G/Pr and AG/Pr of the DI and ND-helper virus are again equivalent). When the first 42 nt of G/Pr are exchanged for the equivalent tr sequences, this DI GP42/AGP55 cannot be rescued by SeV-AGP55. However, DI GP96/AGP55, in which the first 96 nt are trailer sequence and gs-1 is eliminated, can be rescued with SeV-AGP55. Taken together, these results indicate that gs-1 within G/Pr clearly decreases the ability of this 3'-end promoter to compete for vRdRP during recovery from DNA with helper SeV and subsequent coinfection. DISCUSSION : The 96-nt G/Pr contains 55 nt of le sequence (light color, weak), gs-1 (oval with arrow) and a simple hexamer repeat (BB box), whereas the symmetrical AG/Pr contains tr sequence (dark color, strong), the L gene end site (empty oval), and a BB box . AG/Pr is a stronger replication promoter than G/Pr, as defined by the preponderance of genomes formed during virus replication. In these and other experiments, the two BB boxes appear to be interchangeable without effect (data not shown). The major finding of this work is that gs-1 within 3'-end promoters negatively influences promoter strength. Our results are summarized schematically in Fig. . G/Pr contains two elements that weaken this 3'-end promoter relative to AG/Pr; the presence of the le as opposed to the tr sequences, and gs-1. Simply replacing the tr sequence of AG/Pr with le sequence reduces AG/Pr strength such that SeV-AGP55 infections now accumulate equal amounts of genomes and antigenomes . Similarly, introducing gs-1 into AG/Pr of SeV-wt+MCS (SeV-AGP56-65) reduces the genome/antigenome ratio from 12 to 2.5 . Thus, the stronger AG/Pr apparently can be weakened in two different ways. G/Pr has been examined via DI mini-genomes as well as ND-SeV, because gs-1 is not essential for DI-genome replication. Exchanging the first 42 or 48 nt of the le region of G/Pr of ND viruses with the equivalent tr sequences (SeV-GP42/48) does not diminish the genome/antigenome ratio of 10, nor does the analogous exchange of the first 55 nt of G/Pr improve the ability of int-Delta DI genomes to be rescued by SeV-wt . Thus, simply exchanging le for tr sequences apparently cannot overcome the negative effects of gs-1 within G/Pr. The simultaneous elimination of the mRNA start site within G/Pr of int-Delta DI genomes (DI GP98/wt and DI GP98/AGP55) was found to be essential for their rescue by SeV-wt and SeV-AGP55, respectively . Thus, the ability of DI genomes to compete with helper virus genomes for vRdRP is dominated by the presence of gs-1 within the DI genome. We note that there are also examples at AG/Pr where the converse situations applies, namely, where the presence of le as opposed to the tr sequences predominates, and effects due to the presence of gs-1 are not detected (e.g., SeV-AGP55 and SeV-AGP65 rescue int-Delta DI genomes with equal efficiency). Almost all our DI constructs were efficiently amplified when their replication substrates were provided via plasmids (in the absence of other competing SeV templates) . The inability of some constructs to be rescued with SeV-wt, coupled with their efficient rescue with SeV-AGP55/65, suggests that DI rescue depends on the ability of the DI 3'-end promoters to compete with the helper genomes for vRdRP. It is from these experiments that the conclusion that gs-1 negatively affects 3'-end promoter strength comes through most clearly. On what basis, though, does gs-1 within G/Pr diminish the ability of DI genomes to compete with ND genomes for replication substrates? Simple competition of resting ND and DI N:RNAs for a common pool of vRdRP (off the template) may be part of the explanation. In this case, the competition implies that the eventual transcriptases and replicases cannot be distinguished off the template. However, it is also possible that this competition occurs in cis, once vRdRP has engaged G/Pr, as this might offer an explanation for why the presence of gs-1 within G/Pr is so deleterious for 3'-end promoter strength. A model for this in-cis competition is shown in Fig. . vRdRP in this cartoon (elongated oval) has already initiated at the genome 3' end and cleared the promoter. In those cases where RNA synthesis and assembly with N become coupled, vRdRP is committed to replication and this P4-L does not reenter the pool until it has completed N:RNA synthesis. G/Pr strength, or its relative ability to initiate RNAs that are subsequently coassembled with N, is presumably conditioned by the presence of the le versus tr sequences within G/Pr. These sequences as genome would promote vRdRP initiation, and as nascent [+] le RNA would promote N assembly. When RNA synthesis and its assembly with N do not become coupled, vRdRP releases the nascent le chain at or near gs-1. If vRdRP is simultaneously released from the template along with the le RNA, this P4-L rejoins the vRdRP pool off the template and is free to interact with all the available 3'-end promoters in the coinfected cell . However, if vRdRP is not released from the template along with le RNA, it may then be free to scan the template (in both directions) for the nearby gs-1 as well as the genome 3' end, similar to vRdRP that has released its mRNA at gene junctions . Scanning confers an advantage to this 3'-end promoter, precisely because vRdRP is restricted to reinitiating on this template. The presence of gs-1 within G/Pr weakens this replication promoter by diverting these scanning vRdRP from reinitiating at the genome 3' end, thus decreasing its relative replication promoter strength. A corollary of this competition-in-cis model is that relative initiation from gs-1 will similarly depend on competition with the 3'-end promoter; i.e., gs-1 will be initiated more frequently when present within or near G/Pr (near weaker le sequence) than within or near AG/Pr (near stronger tr sequence). This aspect has recently been tested with DI mini-genomes containing tandem 96-nt promoters in which gs-1 of the external promoter was corrupted and the replicative ability of the internal promoter was disabled, such that gs-1 of the internal promoter was now found at position 146, i.e., 50 nt downstream of the external 96-nt replication promoter. In this location, gs-1 was used 15 times more frequently when downstream of a weak G/Pr than downstream of a strong AG/Pr (D. Vulliemoz, P. Le Mercier, and L. Roux, personal communication). Again, the obvious conclusion is that gs-1 and nearby replication promoters compete with each other, presumably in cis, for a common pool of vRdRP in coinfected cells. FIG. 1. : Genome 3'-end promoter (G/Pr) in resting nucleocapsids. Genome 3'-end promoter (G/Pr) in resting nucleocapsids. A model of the helical N-subunit assembly of the SeV resting genome nucleocapsid is shown, based on the electron micrographic reconstruction of Egelman et al. , with a pitch of 5.3 nm and 13 subunits per turn. The position of the genome RNA within this structure is unknown; the RNA line is placed on the outside surface of the N assembly for clarity and because it is just long enough when extended to occupy this position. The N subunits (and the 6 nt positions within each subunit) are numbered from the genome 3' end (lower right). An enlargement of the bipartite 3'-end promoter (dark subunits on the left) is shown on the right, where the 3'-terminal 96 nt are placed within each subunit in groups of six. The 3'-terminal 12 nt and the [C1N2N3N4N5N6]3 repeat conserved in all genomes and antigenomes of respiro-and morbilliviruses are highlighted. The position of the N gene start site (gs-1) is also indicated. FIG. 2. : 3'-end promoter mutations of nondefective SeV and the ratio of genomes to antigenomes formed during infection. 3'-end promoter mutations of nondefective SeV and the ratio of genomes to antigenomes formed during infection. (A) The SeV genomes are shown as a series of boxes (not drawn to scale). (B) The 96-nt genome 3'-end promoter (on the left) and antigenome 3'-end promoter (right) are shown as narrow boxes. The remainder of the genome (from the N gene AUG codon to the UAA that terminates the L open reading frame) is shown as wider boxes. The genome and antigenome 3'-end promoters, G/Pr and AG/Pr, are composed, respectively, of 55 nt of le (white box) or tr sequence (black box), gs-1 (white circle with forward arrow) or the complement of the L gene end signal (black circle with stopped arrow), a short spacer sequence, and the [C1N2N3N4N5N6]3 repeat or its complement (BB box). The cassettes added at the L gene end/tr junction to control ambisense mRNA expression are marked with a dotted line. White and grey indicate the various G/Pr and AG/Pr sequences present at each 3'-end promoter. In naming the viruses, GP and its associated numbers refer to those nucleotides of G/Pr that have been replaced with the equivalent sequences of AG/Pr, or vice versa for AGP viruses. Total RNA from cytoplasmic extracts of A549 cells infected for 24 h with 20 PFU/cell of the various ND-SeV was used to estimate the relative amounts of genomes and antigenomes by real-time RT-PCR. Four sets of primers and probes that spanned different gene junctions (A) were used (to minimize the contribution of mRNA to the antigenome determination [see Materials and Methods]). At least three sets of primers and probes were used for each determination, and the average G/AG ratio from at least two infections of each SeV is shown. The numbers in parentheses show the ratios determined by 32P-labeled primer extension. FIG. 3. : Defective-interfering mini-genomes expressing GFP. Defective-interfering mini-genomes expressing GFP. (A) DI-wt, in which the entire viral coding sequences (from the N gene AUG to the UAA of the L open reading frame) were replaced with the GFP open reading frame, is shown schematically as for Fig. . The sequences present at L gene end site of DI-wt, DI-Ge-1 and DI-Ge-3 (see text) are indicated. (B) A time line of DI particle recovery from DNA is shown (see text for details). Green fluorescent photos of the first passage cultures (P1) that contained equal densities of cells are shown. To measure viral RNA levels, two additional virus passages were carried out to minimize spill-over from biased genome/antigenome ratios during DI recovery from DNA. Genome/antigenome levels were then determined by real-time RT-PCR with the two primers pairs and probe as indicated in panel A. Transfection of DI-wt without helper virus (first photo in panel A), and DI-Ge-1 transfection with helper virus (third photo in panel A) serve as negative controls that show the amplification specificity for the DI RNAs. Error bars, standard deviations. FIG. 4. : DI particle recovery from DNA with various nondefective helper SeV. DI particle recovery from DNA with various nondefective helper SeV. The various DI constructs are shown schematically and are named as for Fig. and . For those constructs with a functional gs-1 driving GFP expression, recovery was scored (+ or -) by fluorescent microscopy at P1 . For those constructs without functional gs-1 (marked with asterisks), recovery was scored by determining DI genome levels by real-time RT-PCR and is plotted in the bar graph below. Error bars, standard deviations. FIG. 5. : A model for promoter competition in cis for vRdRP. A model for promoter competition in cis for vRdRP. The various 96-nt G/Pr and AG/Pr are shown schematically as for the other figures, with the weak le sequences shown by light shading and the strong tr sequences shown by dark shading. The first 12 nt of each promoter are conserved and are shown as a separate shading. The larger oval near the end of the le/tr regions represents vRdRP that has initiated at the 3' end, has cleared this promoter, and has just released the le/tr RNAs. If vRdRP remains attached to the N:RNA without a nascent chain, it is free to scan the template (in either direction) for another RNA start site, which can be either the 3'-end promoter or gs-1 (horizontal arrows below oval). The relative strength of the 3'-end promoter, i.e., its ability to compete for limiting vRdRP during infection relative to other 3'-end promoters, is indicated by the size of bent arrow at the 3' ends. Relative 3'-end promoter strength is conditioned both by the presence of the le/tr sequences and that of gs-1 (circle with bent arrow). Backmatter: PMID- 12915561 TI - Identification of Novel Immunodominant CD4+ Th1-Type T-Cell Peptide Epitopes from Herpes Simplex Virus Glycoprotein D That Confer Protective Immunity AB - The molecular characterization of the epitope repertoire on herpes simplex virus (HSV) antigens would greatly expand our knowledge of HSV immunity and improve immune interventions against herpesvirus infections. HSV glycoprotein D (gD) is an immunodominant viral coat protein and is considered an excellent vaccine candidate antigen. By using the TEPITOPE prediction algorithm, we have identified and characterized a total of 12 regions within the HSV type 1 (HSV-1) gD bearing potential CD4+ T-cell epitopes, each 27 to 34 amino acids in length. Immunogenicity studies of the corresponding medium-sized peptides confirmed all previously known gD epitopes and additionally revealed four new immunodominant regions (gD49-82, gD146-179, gD228-257, and gD332-358), each containing naturally processed epitopes. These epitopes elicited potent T-cell responses in mice of diverse major histocompatibility complex backgrounds. Each of the four new immunodominant peptide epitopes generated strong CD4+ Th1 T cells that were biologically active against HSV-1-infected bone marrow-derived dendritic cells. Importantly, immunization of H-2d mice with the four newly identified CD4+ Th1 peptide epitopes but not with four CD4+ Th2 peptide epitopes induced a robust protective immunity against lethal ocular HSV-1 challenge. These peptide epitopes may prove to be important components of an effective immunoprophylactic strategy against herpes. Keywords: Introduction : Genital,dermal, and ocular herpes simplex virus (HSV) infections cause prevalent, lifelong recurrent infections, with a spectrum of clinical manifestations, including cold sores, genital lesions, corneal blindness, and encephalitis . Despite the availability of many interventional strategies, there has been a constant increase of HSV prevalence during the last 3 decades . Several challenges face the development of an effective herpes vaccine that could help control this epidemic, including the uncertainty about the exact immune correlates of protection, the identification of immunogenic epitopes, and the development of an effective and safe immunization strategy . Despite previous emphasis on antibody (Ab) and CD8+ T-cell responses , there is growing evidence to support a pivotal role for the T-helper type 1 (Th1) subset of CD4+ T cells in antiherpesvirus immunity . CD4+ T cells are required for the protection of mice from HSV challenge . In humans, CD4+ T cells are stimulated in vivo following an HSV infection and the integrated CD4 memory response to HSV type 1 (HSV-1) appears to occur in up to 0.2% of circulating CD4+ T cells . Severe herpetic infections are often seen in immunocompromised individuals with impaired T-cell immunity, such as AIDS and transplant patients, where the immune defect is predominantly displayed in CD4+ T cells . While it is believed that CD4+ T-cell responses are important for protection in general, the importance of Th1- versus Th2-type immune responses for protection against HSV-1 infection is still under investigation. These findings, along with the important role of CD4+ cells in supporting both B-cell and CD8+ T-cell functions , suggest that a successful immunotherapeutic or immunoprophylactic strategy against HSV-1 should include immunodominant CD4+ Th-cell epitopes . Several lines of evidence point to glycoprotein D (gD) being a major target of HSV-1 immune clearance mechanisms and, therefore, an excellent candidate antigen (Ag): (i) it is highly conserved and antigenically cross-reactive with HSV-1 and HSV-2 , (ii) it is among the main target Ags for CD4+ T cells in HSV-seropositive individuals , (iii) it appears to play a vital role in initiating the infectious process , and (iv) importantly, a recent large human genital HSV clinical trial shows that immunization with gD recombinant protein mixed with alum and 3-O-deacylated-monophosphoryl lipid A adjuvants induced significant protection against clinically apparent genital herpes in women who were seronegative for both HSV-1 and HSV-2 . The gD epitopes and the immune mechanism that provide protective immunity are not yet fully understood. Knowledge of immune responses to gD and of their correlation with protection is limited primarily to the identification of B-cell epitopes , with only a limited number of T-cell epitopes having been reported to date . We hypothesized that characterization of the CD4+ T-cell epitope repertoire of gD could be vital in the case of HSV infection. After infection, CD4+ T cells directed to the immunodominant epitopes of gD might have been inactivated and T cells specific for subdominant epitopes might have escaped T-cell tolerance . The present study was aimed at identifying the CD4+ T-cell epitopes of HSV-1 gD and determining the role of CD4+ Th1 versus CD4+ Th2 T-cell epitopes in protection against an ocular HSV-1 challenge. We analyzed HSV-1 gD for CD4+ T-cell epitopes by a combined approach of computational identification of candidate T-cell epitopes followed by in vitro and in vivo biologic validation. Many of the predicted gD sequences were established as forming natural epitopes recognized by HSV-1-primed CD4+ T cells, and their in vivo function was confirmed as inducing a CD4-dependent protective immunity against HSV challenge. Among 12 predicted gD sequences, we found four new immunodominant peptide epitopes (gD49-82, gD146-179, gD228-257, and gD332-358) containing naturally processed epitopes that selectively induced a strong Th1 subset of CD4+ T cells in mice. Four other epitope peptides (gD22-52, gD77-104, gD200-234, and gD287-317) were found to preferentially induce the Th2 subset of CD4+ T cells. Interestingly, immunization with the newly identified immunodominant CD4+ Th1 peptide epitopes, but not CD4+ Th2 peptide epitopes, induced CD4-dependent protective immunity against lethal HSV-1 challenge. The peptide epitopes described in this report could form part of an immunotherapeutic or an immunoprophylactic strategy for treating herpes. MATERIALS AND METHODS : T-cell epitope prediction. | The gD sequence (HSV-1 strain 17) was loaded into the new prediction software (TEPITOPE) to predict promiscuous epitopes . The TEPITOPE algorithm is a Windows application that is based on 25 quantitative matrix-based motifs that cover a significant part of human HLA class II peptide binding specificity . The algorithm permits the prediction and parallel display of ligands for each of the 25 HLA-DR alleles. The TEPITOPE prediction threshold was set at 5%, and 12 regions , predicted to bind at least 50% of the major histocompatibility complex (MHC) class II molecules were picked. The peptide gD253-278 (LPPELSETPNATQPELAPEDPEDSAL) from the region gD257-287 and the peptide gB728-761 (NAAMFAGLGAFFEGMGDLGRAVGKVVMGIVGGVV), which were not picked by the TEPITOPE program as containing potential T-cell epitopes, were synthesized and were used as negative controls. Synthesis of peptides. | A total of 14 peptides (12 gD TEPITOPE-selected peptides and 2 control peptides not selected by TEPITOPE, i.e., gD253-278 and gB728-761), each consisting of 27 to 34 amino acids, were synthesized by BioSource International (Hopkinton, Mass.) on a 9050 Pep Synthesizer Instrument by using solid-phase peptide synthesis and standard 9-fluorenylmethoxy carbonyl technology (PE Applied Biosystems, Foster City, Calif.). Peptides were cleaved from the resin by using trifluoroacetic acid-anisole-thioanisole-anisole-ethanedithiol (EDT)-water (87.5:2.5:2.5:2.5:5%) followed by ether (methyl-t-butyl ether) extraction and lyophilization, as previously described . The purity of peptides was between 75 and 96%, as determined by reversed-phase high-performance liquid chromatography (Vydac C18) and mass spectroscopy (Voyager MALDI-TOF System). Stock solutions were made at 1 mg/ml in water, except for peptide gD146-179, which was solubilized in 5% dimethyl sulfoxide. All peptides were aliquoted and were stored at -20C until assayed. All studies were conducted with the immunogen emulsified in Montanide-ISA-720 adjuvant (M-ISA-720) (Seppic, Fairfield, N.J.) at a 30:70 ratio, which was immediately injected in mice. HSV-1. | The McKrae strain of HSV-1 was used in this study. The virus was triple plaque purified and was prepared as previously described . UV-inactivated HSV-1 was made by exposing the live virus to a Philips 30-W UV bulb for 10 min at a distance of 5 cm. Heat-killed virus was made by heating virus solution at 100C for 5 min. HSV inactivation was confirmed by the inability to produce plaques when tested on Vero cells. Mice and immunization. | Six- to 8-week-old C57BL/6 (H-2b), BALB/c (H-2d), and C3H/HeJ (H-2k) mice, purchased from the Jackson Laboratory (Bar Harbor, Maine), were used in all experiments. Groups of five mice per strain were immunized subcutaneously with peptides in M-ISA-720 adjuvant on days 0 and 21. Peptide-specific T-cell assay. | Twelve days after the second immunization, spleen and inguinal lymph nodes were removed and placed in ice-cold serum-free HL-1 medium supplemented with 15 mM HEPES, 5 x 10-5 M beta-mercaptoethanol, 2 mM glutamine, 50 U of penicillin, and 50 mug of streptomycin (GIBCO-BRL, Grand Island, N.Y.) (CM) . The cells were cultured in 96-well plates at 5 x 105 cells/well in CM, with recall or control peptide at a 30-, 10-, 3-, 1-, or 0.3-mug/ml concentration, as previously described . The cell suspensions were incubated for 72 h at 37C in 5% CO2. One microcurie of [3H]thymidine (Dupont NEN, Boston, Mass.) was added to each well during the last 16 h of culture. The incorporated radioactivity was determined by harvesting cells onto glass fiber filters and counting on a Matrix 96 direct ionization counter (Packard Instruments, Meriden, Conn.) . Results were expressed as the mean counts of cell-associated [3H]thymidine per minute recovered from wells containing Ag minus the mean counts of cell-associated [3H]thymidine per minute recovered from wells without Ag (Deltacpm) (average of triplicate). The stimulation index (SI) was calculated as the mean counts of cell-associated [3H]thymidine per minute recovered from wells containing Ag divided by the mean counts of cell-associated [3H]thymidine per minute recovered from wells without Ag (average of triplicate). For all experiments the irrelevant control peptide gB141-165 and the T-cell mitogen concanavalin A (Sigma, St. Louis, Mo.) were used as negative and positive controls, respectively. Proliferation results were confirmed by repeating each experiment twice. A T-cell proliferative response was considered positive when Deltacpm was > 1,000 and when the SI was > 2, as previously reported . CD4 and CD8 blocking of T-cell responses. | Mice were immunized with a gD peptide, and splenocytes were isolated and stimulated as described above. Stimulated splenocytes were then incubated with the immunizing or control peptide with or without the anti-CD4 monoclonal Ab (MAb) GK1.5 (PharMingen, San Diego, Calif.). Identically prepared cultures were incubated with anti-CD8 MAb 53-6.7 (PharMingen) or with an isotype-matched control Ab, all at 10 mug/ml. It has been previously shown that these antibodies show the expected specificities and function as expected . Cytokine analysis. | T cells were stimulated with either immunizing peptides (10 mug/ml), the irrelevant control peptide (10 mug/ml), UV-inactivated HSV-1 (multiplicity of infection [MOI] = 3), or with concanavalin A (0.5 mug/ml) as a positive control. Culture media were harvested at 48 h (for interleukin 2 [IL-2]) or 96 h (for IL-4 and gamma interferon [IFN-gamma]) and were analyzed by specific sandwich enzyme-linked immunosorbent assay following the manufacturer's instructions (PharMingen). Flow cytometry analysis. | gD peptide-stimulated T cells were phenotyped by double staining with anti-CD4 and anti-CD8 MAbs and fluorescence-activated cell sorter analysis. After 4 days of stimulation with a 10 muM concentration of each peptide, 106 cells were washed in cold phosphate-buffered saline-2% bovine serum albumin buffer and were incubated with phycoerythrin anti-CD4 (PharMingen) or with fluorescein isothiocyanate anti-CD8 (PharMingen) MAbs for 20 to 30 min on ice. Propidium iodide was used to exclude dead cells. For each sample, 20,000 events were acquired on a FACSCalibur and were analyzed with CellQuest software on an integrated-power Macintosh G4 (Becton Dickinson, San Jose, Calif.). Bone marrow-derived DC. | Murine bone marrow-derived dendritic cells (DC) were generated by using a modified version of the protocol that was described previously . Briefly, bone marrow cells were flushed from tibias and femurs with RPMI 1640, and a single-cell suspension was made. A total of 2 x 106 cells were cultured in 100-mm tissue dishes containing 10 ml of RPMI 1640 supplemented with 2 mM glutamine, 1% nonessential amino acids (GIBCO-BRL), 10% fetal calf serum, 50 ng of granulocyte-macrophage colony-stimulating factor per ml, and 50 ng of IL-4 (PeproTech Inc.) per ml. Cells were fed with fresh media supplemented with 25 ng of granulocyte-macrophage colony-stimulating factor per ml and 25 ng of IL-4 per ml every 72 h. After 7 days of incubation this protocol yielded 50 x 106 to 60 x 106 cells, with 70 to 90% of the nonadherent cells having typical DC morphology. This was routinely confirmed by fluorescence-activated cell sorter analysis of CD11c, class II, and DEC-205 surface markers of DC. CD4+-T-cell responses to HSV-infected DC. | CD4+ T-cell lines specific to each of the gD peptides were derived from spleens of immunized mice in vitro, as previously described . To test the biological relevance of these peptide epitopes, CD4+ T-cell effector cells were incubated with irradiated DC (T-cell-to-DC ratio = 50:1) infected with UV-inactivated HSV-1 (MOI of 3, 1, 0.3, or 0.1) and lipopolysaccharide (1 mg/ml) matured for 24 h. As a control CD4+ T cells were also incubated with mock-infected DC. DC and CD4+ T cells were incubated for 5 days at 37C. [3H]thymidine was added to the cultures 18 h before harvesting. Proliferative responses were tested in quadruplicate wells, and the results were expressed as mean counts per minute +- standard deviation. In some experiments, splenocytes from immunized or control mice were restimulated in vitro by incubation with heat-inactivated or UV-inactivated HSV-1. In vivo depletion of CD4+ and CD8+ T cells and HSV-1 challenge. | Twenty days after the second dose of peptide immunogens, mice were infected with 2 x 105 PFU per eye of HSV-1 in tissue culture media administered as eyedrops in a volume of 10 mul. In some experiments, immunized mice were intraperitoneally injected with six doses of 0.1 ml of clarified ascetic fluid in 0.5 ml of phosphate-buffered saline containing MAb GK1.5 (anti-CD4) or MAb 2.43 (anti-CD8) on days -7, -1, 0, 2, and 5 postinfection. Flow cytometry analysis of spleen cells consistently revealed a decrease in CD4+ and CD8+ T cells in treated mice to levels of <3% of those of normal mice. Statistical analysis. | Figures represent data from two or three independent experiments. The data are expressed as the mean +- standard error of the mean and were compared by using Student's t test on a STATVIEW II statistical program (Abacus Concepts, Berkeley, Calif.). RESULTS : Prediction of potential HSV-1 gD T-cell epitopes by the TEPITOPE algorithm. | The deduced primary amino acid sequence of the HSV-1 glycoprotein gD was analyzed by the newly developed T-cell epitope prediction TEPITOPE algorithm . Twelve regions bearing putative antigenic and immunogenic determinants were detected by using a stringent threshold . These regions are probably less constrained than other parts of the molecule and therefore more readily accessible to proteolysis, an event that precedes T-cell epitope presentation in association with MHC molecules . Medium-sized peptides corresponding to the 12 identified regions were synthesized. Ten of them belong to the external N-terminal portion of gD (gD1-29, gD22-52, gD49-82, gD77-104, gD96-123, gD121-152, gD146-179, gD176-206, gD200-234, and gD228-257). One lies adjacent to the hydrophobic membrane anchor domain of gD (gD287-317), and one is part of the proposed hydrophilic C-terminal cytoplasmic portion of gD (gD332-358). It should be noted that, of the 12 predicted regions, six mapped to nonglycosylated regions of gD (gD1-29, gD49-82, gD146-179, gD228-257, gD287-317, and gD332-358). Peptides containing predicted gD epitopes elicited potent CD4+ T cells in mice with diverse MHC backgrounds. | T-cell immunogenicity studies of the selected gD peptides emulsified in the M-ISA-720 adjuvant were carried out in H-2b, H-2d, and H-2k mice. In an initial experiment, we investigated the optimal dose response to peptide gD1-29 and found no differences between doses of 50, 100, and 200 mug. Subsequent experiments used 100 mug (at day 0) and 50 mug (at day 21) of each peptide. Peptide-specific T-cell responses were determined from both the spleen and the lymph node cells. Depending on the peptides and strain of mice used, significant T-cell proliferative responses were generated by every gD peptide. Thus, each of the 12 chosen regions contained at least one T-cell epitope . The strongest T-cell responses were directed primarily, although not exclusively, against five peptides (gD1-29, gD49-82, gD146-179, gD228-257, and gD332-358). The T-cell responses of H-2b, H-2d, and H-2k mice were focused on the same three peptides (gD49-82, gD146-179, and gD332-358), suggesting that they contain dominant T-cell epitopes . In contrast, gD200-234 and gD228-257 appeared to be genetically restricted to H-2d mice. The levels of T-cell response were relatively high with a Deltacpm of >=10,000 for most peptides and up to 50,000 cpm for gD332-358 . Although relatively moderate compared to those for the remaining gD peptides, the responses to gD22-52, gD77-104, and gD96-123 were also significant . A peptide derived from the region gD257-287 (peptide gD253-278), which was not picked by the TEPITOPE program as containing a potential T-cell epitope, did not induce any T-cell response in any of the strains of mice used. The specificity of the proliferative responses was ascertained by the lack of response after restimulation of T cells with an irrelevant peptide (gB141-165) and the lack of response in adjuvant-injected control mice (data not shown). CD4 T cells and not CD8 T cells were important. | These responses were abolished by a MAb against CD4 molecules but not by a MAb against CD8 . Collectively, these results indicate that gD49-82, gD146-179, gD228-257, and gD332-358 peptides are highly immunogenic and contain at least one immunodominant CD4+ T-cell epitope. Priming of multiple and elevated T-cell responses by pools of gD-derived peptides. | The ability of pools of gD peptides to simultaneously induce multiple T cells specific to each peptide within the pool was explored. In these experiments, the immunogenicity in H-2d mice of pooled versus individual peptides was compared side by side to investigate if there was any agonistic or synergistic interaction between the peptide epitopes composing the pool. As a control, H-2d mice were injected with M-ISA-720 alone. Immunization with a pool of gD1-29, gD49-82, and gD332-358 peptides generated multiepitopic and significantly more elevated T-cell responses specific to each peptide (P < 0.001) . Thus, when evaluated individually, each peptide induced a relatively lower response (P < 0.001) . In a similar experiment, the responses induced by a pool of gD96-123, gD146-179, and gD287-317 peptides were also at higher than the responses induced when individual peptides were employed (data not shown). Several selected gD epitope peptides were identified as targets of IFN-gamma- and IL-2-secreting CD4+ T cells. | To determine the subset of CD4+ T cells induced by each gD peptide, we studied the pattern of peptide-specific IL-2, IL-4, and IFN-gamma cytokines. As shown in Fig. , five profiles of cytokines were found: (i) peptides gD1-29, gD49-82, gD96-123, gD146-179, gD228-257, and gD332-358 induced strong Th1 cytokine (IFN-gamma and IL-2) secretion more efficiently than did the remaining peptides; (ii) the gD287-317 peptide induced low levels of Th1 cytokine secretion; (iii) the gD22-52 and gD77-104 peptides preferentially induced Th-2 cytokine (IL-4); (iv) the gD200-234 peptide induced a mixed IFN-gamma-IL-4 response, since both cytokines were induced to a comparable extent; and (v) finally, the gD121-152 and gD176-206 peptides did not induce production of IL-2, IL-4, or IFN-gamma cytokines . Overall, most peptides induced IFN-gamma and IL-2 production, indicating that the selected HSV-1 gD peptides emulsified in the M-ISA-720 adjuvant preferentially elicited a polarized Th1 immune response . Antibody blocking of T-cell activity revealed that these cytokines were mainly produced by CD4+ T cells and only slightly by CD8+ T cells . T cells induced by gD-peptides react to the native viral protein. | Experiments were performed to determine if CD4+ T cells induced by gD synthetic peptides were reactive to naturally processed epitopes, as presented by HSV-1-infected cells. CD4+ T-cell lines specific to gD1-29, gD49-82, gD146-179, gD228-257, and gD332-358 immunodominant Th1 epitopes and to the gD77-104 subdominant epitope were established from H-2d mice. Recognition of autologous bone marrow-derived DC infected with UV-inactivated HSV-1 by each of these T-cell lines was then monitored in a proliferation assay . T-cell lines specific to gD1-29, gD49-82, gD146-179, gD228-257, and gD332-358 recognized UV-inactivated HSV-infected DC. No response was observed when autologous mock-infected DC were employed as Ag-presenting cells. The CD4+ T-cell line induced by subdominant gD77-104 peptide also failed to recognize HSV-infected DC . Overall, these results indicated that processing and presentation of the epitopes contained in the gD1-29, gD49-82, gD146-179, gD228-257, and gD332-358 peptide sequences occurred in HSV-infected cells. In subsequent experiments, T-cell proliferation and IFN-gamma production by immune T cells from gD peptide-immunized mice were monitored against whole HSV as Ag by using immune spleen cells as the source of responding T cells and Ag-presenting cells. T cells from H-2b, H-2d, and H-2k mice immunized with gD49-82, gD146-179, gD228-257, and gD332-358 exhibited significant proliferation and IFN-gamma production upon in vitro stimulation with heat-killed HSV-1. Under the same conditions, T cells from the adjuvant-injected control mice did not respond to heat-inactivated HSV-1-stimulation . Thus, these responses were Ag specific and were not due to a mitogenic effect of viral particles. These HSV-1-specific T-cell responses were strongly reduced by anti-CD4 MAb treatment but not by anti-CD8 MAbs . Lack of immunodominance in HSV-primed T-cell responses to selected gD-peptides. | To determine the fine specificity of broadly reactive T cells associated with viral immunity and to explore immunodominance in the context of HSV infection, proliferation of primed T cells obtained from 20 HSV-1-infected H-2d mice were evaluated against gD by using each of the 12 peptides as Ag . Although the selected peptides stimulated moderate gD-specific T-cell responses, surprisingly, the in vivo HSV-primed T cells were reactive in vitro to up to 10 of the 12 gD peptides (depending on the individual mouse). Despite a difference between individual mice, a broad range of gD-specific T cells were primed in all of the 20 infected mice and a unique array of T-cell responses was identified from each of the 20 infected mice analyzed . Seven peptides (gD1-29, gD49-82, gD96-123, gD146-179, gD228-257, gD287-317, and gD332-358) induced a response in more than 85% of the HSV-infected mice . The responses were found to the immunodominant epitopes, gD1-29, gD49-82, gD146-179, gD287-317, and gD332-358 and also to gD22-52, gD77-104, gD96-123, and gD121-152, which represent subdominant epitopes. Immunization with a pool of the newly identified CD4+ Th1 immunodominant gD peptide epitopes resulted in prolonged survival upon a lethal HSV-1 challenge. | To investigate whether the Th1- or Th2-type immune responses are more important for protection from HSV-1 infection, mice were immunized with either CD4+ Th1 peptide epitopes (gD49-82, gD146-179, gD228-257, and gD332-358) or CD4+ Th2 peptide epitopes (gD22-52, gD77-104, gD200-234, and gD287-317) . The previously described protective epitope gD1-29 was excluded from these experiments. Groups of 10 H-2d mice were immunized twice with the following: a pool of gD49-82, gD146-179, gD228-257, and gD332-358 emulsified in M-ISA-720 adjuvant or a pool of gD22-52, gD77-104, gD121-152, and gD200-234 emulsified in M-ISA-720 adjuvant or M-ISA-720 alone (adjuvant-injected control), or they were left untreated (nonimmunized control). Twenty days after the second dose of peptide vaccine, mice were challenged with 2 x 105 PFU of the McKrae strain of HSV-1 per eye as described in Materials and Methods. Mice were monitored for 4 weeks after ocular HSV-1 challenge for their ability to withstand a lethal infection with HSV-1. All of the mice that died following challenge did so between days 8 and 12 postinfection. All of the H-2d mice immunized with the pool of Th1 peptide epitopes survived the lethal HSV-1 challenge. In contrast, less than 50% of the H-2d mice immunized with the pool of Th2 peptide epitopes survived. Only 10% of adjuvant-injected and 10% of nonimmunized control H-2d mice survived the HSV-1 challenge . These results demonstrate that the Th1 peptide epitopes provided more protection against lethal ocular HSV-1 challenge than did the Th2 peptide epitopes (P = 0.0003). Depletion of CD4+ T cells abolishes the protective response. | To determine the involvement of CD4+ and CD8+ T cells in induced protection, mice were immunized with gD49-82, gD146-179, gD228-257, and gD332-358 peptides and were then divided into 4 groups of 10. The groups were then either depleted of CD4+ T cells, depleted of CD8+ T cells, left untreated (none), or treated with irrelevant antibodies (rat immunoglobulin G [IgG] or IgG control). All four groups were then challenged with HSV-1 as described above. Depletion of CD4+ T cells resulted in the death of all infected mice (Table ; P = 0.0001). However, depletion of CD8+ T cells or injection of control rat IgG antibodies did not significantly impair the induced protective immunity (P = 0.47 or 1, respectively) . These results demonstrated that, in this system, CD4+ T cells are required and CD8+ T cells are not sufficient for protective immunity against lethal ocular HSV-1 challenge. DISCUSSION : HSV infection is a major cause of a spectrum of morbidity and clinical diseases in humans . During the last decade, HSV vaccine development has primarily focused on various forms of recombinantly expressed glycoproteins . Of the 11 known HSV glycoproteins, gD is the most highly conserved and the most antigenically cross-reactive between HSV-1 and HSV-2 . gD has emerged as an excellent candidate Ag for inducing a protective immunity in animal models against ocular and genital infections with both types of HSV . It has recently been reported in a large human genital HSV trial that immunization with recombinant gD induced significant protection in women who were seronegative for HSV but was ineffective in seropositive women and in men . The antibody response to this vaccine was similar to natural HSV infection, but demonstration of its ability to induce T-cell responses was restricted by the limited knowledge of gD T-cell epitopes (, , -). Identification and molecular characterization of the CD4+ T-cell epitope repertoire of gD would contribute to a better understanding of the immune correlates of protection and would help in developing effective immunotherapeutic and immunoprophylactic strategies. There are three principal findings in the present report. First, the combined approach of epitope prediction performed by using the TEPITOPE algorithm and the rapid determination of immunogenicity in mice, with a wide range of MHC restriction, proved to be a useful strategy for identifying T-cell epitope-bearing regions from the sequence of a key immunogenic HSV protein. Second, appropriate medium-sized peptides carefully selected from the putative gD sequence and emulsified in a human-compatible adjuvant M-ISA-720 mainly induced potent IFN-gamma- and IL-2-secreting CD4+ T cells that recognized naturally processed epitopes. Third, and of particular interest, immunization with a pool of newly identified CD4+ Th1 gD peptide epitopes elicited a CD4-dependent protective immunity against a lethal ocular HSV-1 challenge. Among the advantages of an epitope-based vaccine approach over the conventional whole-protein approach is the possibility of including multiple immunodominant and subdominant epitopes. This strategy may offer an opportunity to elicit responses superior to that induced following HSV infection or even following immunization by recombinantly expressed glycoproteins. Several gD peptide epitopes selected in the present study induced potent HSV-specific CD4+ Th1-like T-cell responses, with production of high levels of IFN-gamma and IL-2, cytokines that play a significant role in host response to HSV . In addition, we demonstrated an enhancement in the immunogenicity of gD peptides injected as a pool, supporting the concept that combining potent immunogenic epitopes may be a practical way to create a broader and a more potent T-cell response. Numerous potential T-cell epitopes could theoretically be generated from a protein Ag. However, in practice, T cells tend to focus upon just a few immunodominant epitopes, whereas discrete cryptic epitopes remain hidden to the immune system . Identification of such immunogenic epitopes from among the often hundreds or thousands of amino acids comprising HSV proteins could be a cumbersome and laborious process . Traditional approaches for identifying such epitopes from among the often hundreds or thousands of amino acids that cover the entire sequence of a protein Ag have generally involved synthesizing scores of overlapping peptides ("overlapping-peptide method") and extensive screening of T-cell clones isolated from whole Ag-stimulated cells . The overlapping-peptide method might not automatically map the same epitope peptides predicted in our study from using the TEPITOPE algorithm. Indeed, using the overlapping-peptide method, one can imagine that the regions where these peptides overlap might by themselves contain "junctional epitopes" . Shortening or lengthening a peptide by one or a few amino acids might sometime result in missing junctional epitopes that might be present in the overlapping regions. In addition, a major drawback of the overlapping-peptide method is the number of peptide sequences that need to be synthesized and tested, thus making it an expensive, labor-intensive, and time-consuming process . In addition, progress on the mapping of T-cell epitopes has been slow due to reliance on studies of clones, an approach that generally involves extensive screening of T-cell precursors isolated from whole Ag-stimulated cells . Candidate T-cell peptide epitopes can be also screened based on their affinity to interact in vitro with HLA class II molecules . However, the high polymorphism of HLA class II molecules is a major drawback in the identification of CD4+ T-cell epitopes with large population coverage . Another alternative to cloning T cells has employed the elegant technology of tetramer-guided epitope mapping, which offers the opportunity for a straightforward cloning of the Ag-specific T cells through single-cell sorting . However, besides needing to form pools of overlapping peptides with this technology, there are concerns that relevant peptides present in the pool will be competed out by irrelevant peptides. In addition, the relative instability of MHC class II tetramers (compared to MHC class I tetramers) points out that the tetramer approach still needs improvement . Relatively laborious strategies identified small subsets of candidate epitopes by sequencing peptides eluted from purified MHC molecules from pathogen-infected cells and then testing their MHC binding affinity . High-affinity peptides are then tested for their ability to induce pathogen-specific T cells . Murine class II-restricted Th lymphocyte epitopes have been described as cross-reacting with HLA-DR molecules, highlighting the feasibility of murine models for evaluating epitope-based vaccines destined for human use . In the study reported here, to allow reduction of the number of peptides and T cells to be assayed, we employed the TEPITOPE algorithm to quickly identify regions bearing potential T-cell epitopes from the sequence of HSV-1 gD. The T-cell immunogenicity of selected epitopes was verified in mice of diverse MHC haplotypes. The predictive power of the present strategy was evaluated by comparing the TEPITOPE-identified regions to the published T-cell epitopes of HSV-1 and HSV-2 gD proteins. The latter epitopes were identified with other methods (e.g., overlapping peptides and truncated forms of gD protein) and other algorithms (e.g., prediction of amphipathic and hydrophilic parameters) . The present strategy accurately identified regions bearing all of the previously reported gD epitopes: gD1-23 ; gD241-260 , and gD290-314 . More importantly, the present approach identified several novel, not previously described, Th1 immunodominant CD4+ T-cell epitopes that elicited protective immunity. HSV-primed T cells, induced in mice following HSV-1 infection, were found to recognize nearly all of the selected gD peptides. This supports the concept of extensive diversity in the array of gD epitopes being recruited during HSV infection . The selected peptides were immunogenic in mice of diverse MHC backgrounds, and most showed high immunogenicity in rabbits, as well (L. BenMohamed, unpublished data). These T-cell epitopes were thus promiscuously recognized by a broad variety of T-cell receptor molecules in diverse MHC haplotypes. A careful comparative analysis of the location of the broadly reactive peptides within the HSV-1 gD sequence revealed two relevant characteristics: (i) most of the immunogenic T epitopes were segregated to two regions within either the N-terminal or the C-terminal domain of gD, while multiepitopes corresponding to different MHC restrictions have been observed in relatively small regions of other protein Ags , and (ii) most of the identified peptide epitopes were confined to nonglycosylated regions of gD . This is consistent with a possible structural influence of glycosylation on suppressing the development of herpes-specific T-cell responses, by some as-yet-undefined process, as has been recently reported for simian immunodeficiency virus glycoproteins . Accumulating evidence in both animal models and humans indicates that CD4+ T-cell immunity is directly or indirectly related to the control of HSV infection . Mice with depressed CD4+ T cells are prone to develop severe herpetic diseases or mortality upon HSV challenge . Often severe ocular and genital herpetic infections are seen in immunocompromised individuals with impaired T-cell immunity. This is true in AIDS and transplant patients where the immune defect is predominantly dependent upon CD4+ T cells . While it is believed that CD4+ T-cell responses are important for protection in general, the Th1- or Th2-type immune response correlates of protection from HSV-1 infection are poorly understood. In the present study, to investigate the role of Th1- or Th2-type immune responses, mice were immunized either with CD4+ Th1 or CD4+ Th2 peptide epitopes, in an effort to polarize the immune response. The new CD4+ Th1 peptide epitopes in gD, but not Th2 peptide epitopes, induced a CD4-dependent protective immunity against a lethal ocular HSV-1 challenge. This result supports an important role for Th1 CD4+ T cells in antiherpes immunity . However, because CD4+ cells play an important role in supporting both B and CD8+ cell function , we cannot rule out that Abs and cytotoxic T lymphocytes contributed to the observed protection, as both might have been indirectly affected by in vivo depletion of CD4+ T cells . A possible role of Ab response is supported by the finding that many of the selected gD peptides in this study were also found to induce strong antibody responses (BenMohamed, unpublished). In addition, some of the immunizing peptides (e.g., gD228-257 and D332-358) contain the previously described linear B-cell epitopes . However, it is unlikely that preexisting antibody specific to these immunizing peptides will have a major effect, since CD4 depletion was done after antibody response to the peptides occurred but before the boost of these antibodies would have happened following the HSV-1 challenge. FIG. 1. : Ag-specific T-cell proliferation induced in BALB/c (H-2d), C57BL/6 (H-2b), and C3H/HeJ (H-2k) mice following immunization with TEPITOPE-selected gD peptides. Ag-specific T-cell proliferation induced in BALB/c (H-2d), C57BL/6 (H-2b), and C3H/HeJ (H-2k) mice following immunization with TEPITOPE-selected gD peptides. Mice were immunized with the peptide indicated at the top of each panel, and peptide-specific T-cell responses were determined following in vitro stimulation of spleen T cells with the recall peptide (solid symbols) or the irrelevant control peptide gB141-165 (open symbols). The spleen T cells were obtained from H-2d (circles), H-2b (triangles), and H-2k (squares) mice 14 days after the second (final) immunization. The means +- standard deviations from two separate experiments (five mice/group/experiment) are shown. FIG. 2. : Simultaneous induction of multiple Ag-specific T cells to pools of gD-derived peptides. Simultaneous induction of multiple Ag-specific T cells to pools of gD-derived peptides. H-2d mice were immunized with a pool of gD1-29, gD49-82, and gD332-358 or with the individual peptides. The peptides indicated above each panel were used for in vitro stimulation of spleen T cells from mice immunized with the pooled peptides (solid squares), mice immunized with individual peptide (open squares), or control mice injected with adjuvant alone (open circles). The results shown are representative of three separate experiments. The P values shown are for pooled versus single peptide. FIG. 3. : Induction of IL-2, IL-4, and IFN-gamma by gD peptides emulsified in M-ISA720 adjuvant. Induction of IL-2, IL-4, and IFN-gamma by gD peptides emulsified in M-ISA720 adjuvant. H-2d mice were immunized with individual peptides as described. Splenic T cells were stimulated with the immunizing peptide (solid bars) or the control irrelevant peptide gB141-165 (open bars). IL-2, IL-4, and IFN-gamma secreted into the culture media were quantified by a specific sandwich enzyme-linked immunosorbent assay, as described in Materials and Methods. FIG. 4. : The gD peptide-induced T-cell responses were relevant to the native viral protein. The gD peptide-induced T-cell responses were relevant to the native viral protein. (A) Splenocyte-derived CD41-T-cell lines from gD1-29-, gD49-82-, gD77-104-, gD146-179-, gD228-257-, or gD332-358-immunized H-2d mice were incubated with autologous bone marrowderived DC (at an effector-to-target ratio of T cells to DC of 50:1). Target DC had been infected with UV-inactivated HSV-1 at an MOI of 3, 1, 0.3, or 0.1 (HSV Infected Dendritic Cells). Control cultures contained CD41 T cells and mock-infected DC. The [3H]thymidine uptake as observed in three independently performed experiments is shown. (B) Proliferative responses of CD41 T cells from H-2d (closed circles), H-2b (closed triangles), and H-2k (closed squares) mice immunized with gD49-82, gD146-179, gD228-257, and gD332-358 gD peptides were determined by using heat-killed HSV-1 at an MOI of 3, 1, 0.3, 0.1, 0.03, or 0.01 for in vitro stimulation. [3H]thymidine incorporation was determined after 3 days of stimulation. Spleen cells from adjuvantinjected H-2b (open circles), H-2d (open triangles), and H-2k (open squares) mice were used as control. FIG. 5. : The predicted peptides contain epitopes that are naturally processed and presented to the host immune system during the course of HSV infection. The predicted peptides contain epitopes that are naturally processed and presented to the host immune system during the course of HSV infection. T cells from spleens of H-2d mice 21 days after HSV-1 infection recognized the selected gD peptides as determined by in vitro lymphoproliferative responses to individual peptide (HSV-1 infected). Spleen cells from mock-infected mice were used as controls. TABLE 1 : Peptides bearing potential T-cell epitopes identified within the HSV-1 glycoprotein D (gD) by using the TEPITOPE algorithm TABLE 2 : CD4+ dependence of T-cell proliferation and cytokine secretion induced by gD peptides TABLE 3 : Immunization with newly identified Th1 gD peptide epitopes in the M-ISA-720 adjuvant confers protective immunity to a lethal HSV-1 challenge TABLE 4 : Immunization with the newly identified gD peptide epitopes in the Montanide adjuvant induced a CD4+-T-cell-dependent protective immunity against a lethal HSV-1 challenge Backmatter: PMID- 12915553 TI - Silent Point Mutation in an Avian Retrovirus RNA Processing Element Promotes c-myb-Associated Short-Latency Lymphomas AB - The avian leukosis virus DeltaLR-9 causes a high frequency of B-cell lymphomas within weeks after injection into 10-day-old chicken embryos. These lymphomas result from proviral integrations into the oncogene c-myb. In contrast, LR-9, which lacks the 42-nucleotide gag gene deletion of DeltaLR-9, does not cause a high frequency of c-myb-associated short-latency lymphomas. Although viral replication rates and spliced env mRNA levels were found to be similar for both viruses, DeltaLR-9 exhibited an increase in readthrough transcription compared to LR-9. The DeltaLR-9 deletion is located in the region of the gag gene corresponding to the matrix (MA) protein as well as in the negative regulator of splicing (NRS) element. To test whether disruption of the NRS or of the MA protein was responsible for inducing short-latency lymphomas, we generated viruses with NRS point mutations that maintained the wild-type Gag amino acid sequence. One of the mutant viruses induced an even higher incidence than DeltaLR-9 of short-latency lymphomas with viral integrations into c-myb. Thus, we propose that disruption of the NRS sequence promotes readthrough transcription and splicing to the downstream myb gene, causing overexpression of a slightly truncated Myb protein, which induces short-latency tumors. Keywords: Introduction : A defining characteristic of retroviruses is that they randomly integrate their reverse-transcribed genome into the genome of the host cell. Selection for a specific integration site can occur when a retrovirus integrates into a proto-oncogene, resulting in a clonal tumor. When an avian leukosis virus (ALV) is injected into 1-day-old chicks, it typically induces clonal B-cell lymphomas exhibiting proviral DNA integrated into the oncogene c-myc . These B-cell lymphomas result from the deregulated expression of c-myc and take several months to develop. However, if an ALV is injected into 10- to 12-day-old chicken embryos, approximately 14% of the hatched chickens develop short-latency B-cell lymphomas that cause death within weeks . Moreover, injection of chicken embryos with the recombinant ALV EU-8 (which has a 42-nucleotide [nt] deletion in the gag gene) leads to short-latency B-cell lymphomas in 40 to 80% of the chickens . Most of these short-latency lymphomas contain proviral DNA integrated into the first intron of c-myb . Transcription initiates at the ALV promoter in the 5' long terminal repeat (LTR) and reads through the polyadenylation sequence in the 3' LTR to transcribe the downstream c-myb gene (see Fig. ). Splicing then occurs from the 5' splice site in the 5' end of the viral gag gene to the 3' splice site of the c-myb second exon . Translation of the ALV-myb hybrid RNA yields a Myb protein that is truncated at its N terminus by 20 amino acids . Truncated Myb, in contrast to full-length Myb, is highly oncogenic when expressed from retroviral vectors introduced into 12-day-old chicken embryos . Further, the transcriptional profile of the short-latency, c-myb-associated tumors is distinct from that of c-myc-associated lymphomas . The higher incidence of short-latency tumors obtained with EU-8 than with other ALVs has been attributed to a 42-nt deletion in the gag gene of EU-8, which was derived from ring-necked pheasant virus . This same region (nt 735 to 776) was deleted from the recombinant ALV LR-9 to produce DeltaLR-9. When injected into 10-day-old chicken embryos, DeltaLR-9 also causes mortality from short-latency B-cell lymphomas at a higher frequency than LR-9 . The DeltaLR-9 genomic deletion is located within the gag gene, and it coincides with the 5' portion of a viral RNA element termed the negative regulator of splicing (NRS) . When the NRS of DeltaLR-9 was inserted into the intron of a heterologous splicing reporter construct, splicing was increased relative to that in a construct containing the wild-type (WT) LR-9 NRS . The NRS has previously been characterized in Rous sarcoma virus (RSV), and its sequence is 95% identical to that of the NRS of the ALV LR-9 . The RSV NRS is a 230-nt-long cis-acting element, which limits both RNA splicing and readthrough transcription . Both of these processes are critical to c-myb-associated short-latency tumorigenesis. In RSV, the NRS suppresses splicing of the src mRNA but not of env mRNA . The NRS has two regions necessary for function. A purine-rich region at its 5' end binds arginine/serine-rich (SR) proteins and hnRNP H . The 3' end of the NRS contains a sequence that is similar to a 5' splice site and binds U1 snRNP . U1 snRNP binding is thought to be important for NRS activity; mutations in its binding site inhibit NRS activity, and a complementary U1 snRNA mutation partially rescues activity . The deletion in the DeltaLR-9 genome is also within the region of gag corresponding to the matrix (MA) protein, thus generating a truncated MA protein (see Fig. ) . To determine whether the mutation in MA or the NRS was important for short-latency tumorigenesis, we tested viruses with point mutations that code for a WT Gag protein but alter the NRS. In addition, the binding of various splicing factors to the NRS RNA was studied with the NRS mutants. Finally, to further study how the DeltaLR-9 deletion promotes tumorigenesis, its effects on viral replication, env mRNA splicing, and readthrough transcription were examined. MATERIALS AND METHODS : Cells and viruses. | Chicken embryo fibroblasts (CEFs) were cultured in medium 199 with the addition of 2% tryptose phosphate, 1% calf serum, 1% heat-inactivated chick serum, and 1% antibiotic-antimycotic (Life Technologies). The permuted LR-9 and DeltaLR-9 ALV constructs are both inserted into the vector pUC13 and were a gift from Bill Hayward. U916A, G919A, and WT NRS ALVs were constructed by inserting RSV fragments from the BspEI site (nt 700) to the KpnI site (nt 4,995) into the corresponding sites in LR-9. U916A and G919A are both silent mutations, which were originally described in RSV by O'Sullivan et al. . Ten micrograms of SacI-linearized viral constructs was transfected into CEFs as previously described . Reverse transcriptase assay. | After harvesting, the virion-containing medium was assayed for reverse transcriptase activity as described by Telesnitsky et al. . Quantitation was performed on an Instant Imager (Packard). RNase protection assay. | The readthrough probe contains the entire LR-9 LTR and flanking sequences. The permuted, single-LTR LR-9 genomic plasmid was digested with SacI (nt 255) and HindIII (nt 7,040), and the LTR fragment was inserted into the pGEM-4Z cloning vector (Stratagene). The construct was linearized with BanI prior to in vitro transcription with Sp6 RNA polymerase. Total cellular RNA was harvested with RNA Bee (Tel-Test). Readthrough RNase protections were carried out as described previously by using a hybridization temperature of 42C and an RNase digestion temperature of 37C. The bands were quantified on an Instant Imager (Packard), and the values were adjusted for size differences. The fraction of readthrough RNA was determined relative to the sum of normally polyadenylated and readthrough RNA bands. mRNA selection and Northern blotting. | Polyadenylated RNA was selected from total cellular RNA by binding to oligo(dT) cellulose, and the selected RNA was used for Northern blotting as described by Sambrook and Russell . Viral RNA was hybridized to the readthrough (LTR) probe described above, which is complementary to both spliced and unspliced viral RNAs. The bands were quantified on an Instant Imager (Packard). Chicken injection and tumor detection. | Virus was collected at day 11 after transfection of CEFs, and 105 to 106 infectious U was injected into the chorioallantoic veins of 48 10-day-old chicken embryos from the inbred SC White Leghorn line (Hyline International, Dallas, Iowa). Thirty-seven of the injected embryos hatched, and 34 were monitored for tumor formation. Chickens infected with different viruses were kept in separate cages, and they were sacrificed when ill or at 10 weeks after hatching. Each chicken was examined post mortem for tumors by gross examination and by tissue section staining as described by Neiman et al. . Tissue samples were collected for DNA extraction. All chicken manipulations were performed without knowledge of the experimental hypothesis. DNA extraction and nested PCR. | DNA from chicken bursa and liver tissue was extracted as described by Sambrook and Russell . Nested PCR was carried out to detect integrations into c-myb by using primers specific for the proviral LTR and for myb exon 2, as described by Neiman et al. . The upstream c-myb integration was detected with a 3' primer (5'-CGGCACGGAGCGCGCTTTGCGTGCGTG-3') upstream of exon 1 of c-myb in conjunction with the same LTR primers . Affinity selections. | The viral constructs were used as templates for PCRs with a T7 promoter (underlined)-containing 5' primer (5'-TAATACGACTCACTATAGGGAGGTCCGGAGTGCATCGAGAAACC-3') hybridizing to NRS nt 703 to 727 and a 3' primer (5'-AGGGAAGGATACAAACCACTCCCCACA-3' for LR-9, DeltaLR-9, and the WT; 5'-AGGGAAGGATACAATCCACTCCCCACA-3' for U916A; and 5'-AGGGAAGGATATAAACCACTCCCCACA-3' for G919A) hybridizing to nt 930 to 907. RNA synthesis, linkage to adipic acid dihydrazide-agarose beads, and affinity selections were performed as described previously . Both affinity-selected proteins and RNAs were examined. For RNA analysis, the washed pellet was digested with 50 mug of proteinase K (Roche) for 30 min at 37C in 300 mul of proteinase K buffer. The samples were then boiled for 30 s and extracted with phenol and with phenol-chloroform-isoamyl alcohol (25:24:1). After ethanol precipitation, the pellet was resuspended in 20 mul of 8 M urea, electrophoresed on a 6% polyacrylamide-8 M urea gel, and transferred to a Zeta-Probe GT blotting membrane (Bio-Rad) with a transblot semidry transfer apparatus (Bio-Rad). The blot was probed with a U1 snRNA-specific probe , which was linearized with EcoRI prior to transcription with T3 polymerase. Affinity-selected proteins were separated on a sodium dodecyl sulfate-13% polyacrylamide gel, transferred, and probed as described by Sambrook and Russell . Antibodies against ASF/SF2 were obtained from A. Krainer, and antibodies against hnRNP H were obtained from I. Mattaj. A Coomassie-stained gel (GelCode Blue staining reagent; Perbio) was used to normalize the signals. RESULTS : LR-9 and DeltaLR-9 exhibit similar replication rates and env mRNA levels. | The 42-nt deletion in the gag gene of DeltaLR-9 ALV disrupts the NRS sequence, and it also deletes residues 119 to 132 of the MA protein . In an effort to determine how the deletion in DeltaLR-9 promotes tumorigenesis, we first assayed its effect on viral replication. CEFs were infected with equivalent reverse transcriptase units of either LR-9 or DeltaLR-9 virus. To monitor growth rates, medium was collected every other day starting with day 3 and assayed for viral reverse transcriptase activity. Based on reverse transcriptase levels, the replication rates of LR-9 and DeltaLR-9 were found to be virtually identical . Thus, we conclude that although DeltaLR-9 has an extensive deletion in MA and in the NRS sequence, this does not affect the replication rate of the virus in CEFs. Next, we wanted to determine whether the deletion affected levels of spliced env mRNA in ALV. In RSV, NRS mutations increase src but not env mRNA levels . LR-9 and DeltaLR-9 constructs were transfected into CEFs, RNA was harvested after 6 days, and the relative amount of spliced env mRNA was determined by Northern blotting with an LTR-specific probe. In this assay, approximately 35% of LR-9 RNA was found to be spliced, and DeltaLR-9 RNA was spliced to a similar level . Therefore, the deletion in the NRS did not result in an increase in splicing to the viral env gene, which is consistent with the finding that replication was not affected by the deletion. ALV with an NRS deletion has increased readthrough transcription. | The short-latency tumors induced by DeltaLR-9 are likely due to overexpression of truncated Myb, which is translated from a readthrough transcript as shown previously for the related EU-8 ALV . Since the NRS has been implicated in promoting polyadenylation of RSV transcripts , we next wanted to determine whether the rate of readthrough transcription is different for DeltaLR-9 than for the LR-9 virus. To this end, plasmids containing LR-9 and DeltaLR-9 DNAs were transfected into CEFs, and total cellular RNA was harvested after 11 days. The levels of readthrough transcription were compared to those of correctly terminated transcripts by RNase protection with a readthrough probe that was complementary to the LR-9 LTR and flanking sequences . In these experiments, it was not important where the virus was integrated, because only readthrough into viral sequences was assayed. LR-9 was determined to have an average of 10% readthrough transcripts (relative to the total amount of readthrough and normally terminated polyadenylated RNA), and DeltaLR-9 had a 60% higher amount of readthrough (P < 0.005) . We conclude that one of the factors that causes DeltaLR-9 to induce short-latency lymphomas is that it exhibits a higher level of readthrough transcription into downstream cellular genes than does LR-9. In addition, it was previously shown that the NRS of DeltaLR-9 promotes splicing to a downstream reporter gene . Is the deletion in the MA protein or the NRS sequence of DeltaLR-9 responsible for the increase in short-latency lymphomas? | The short-latency tumors induced by DeltaLR-9 may be due to either an altered MA protein or the disruption of the overlapping NRS sequence (or both). To distinguish between these possibilities, we generated point mutations that changed the NRS sequence but did not affect the amino acid sequence of the MA protein. We constructed two viruses, U916A and G919A, each having a single NRS point mutation but lacking the 42-nt deletion of DeltaLR-9. U1 snRNP binding has been shown previously to be important for NRS function , and both of these mutations were predicted to disrupt the U1 snRNP binding consensus sequence . Nevertheless, only one of these point mutations, G919A, led to an increase in readthrough transcription and splicing to src in RSV . Since both readthrough and splicing are thought to be important for the generation of truncated Myb, we predicted that the U916A mutant virus would behave like WT virus. The viruses with point mutations, U916A and G919A, and the WT NRS control virus were constructed by placing the corresponding RSV fragments (nt 700 to 4,995), described previously , into the LR-9 background . The replication abilities of the constructs were confirmed in CEFs through reverse transcriptase assays (data not shown). To determine the tumorigenicity of the viral constructs, we injected 12 10-day-old chicken embryos with one of the following four test viruses: the viruses with point mutations, U916A and G919A, the WT NRS (negative control) virus, and DeltaLR-9 (positive control). Out of the total of 48 injected chicken embryos, 37 chicks hatched and 34 were monitored for deaths and the development of tumors for 10 weeks after hatching. Surviving chickens were sacrificed at week 10, and all chickens were examined for tumors at the time of their death. We asked whether short-latency lymphomas (causing death within 10 weeks after hatching) would arise at the same frequency with the viruses carrying NRS point mutations as with the positive control virus, DeltaLR-9. The percentages of chickens that died, or were sacrificed for humane reasons, before 10 weeks after hatching (defined as the short-latency period) and that exhibited lymphomas are plotted in Fig. . Lymphomas were detected both by gross examination and by histological analysis. The group infected with the positive control, DeltaLR-9, showed a rate of mortality from lymphomas of 44% (four of nine birds), which is similar to the 54% that Smith et al. observed by week 10 with the same viral construct . Remarkably, the G919A viral construct, which has only a single point mutation compared to the WT, induced a 75% (six of eight birds) rate of mortality from short-latency lymphomas in this same time period . Moreover, one of these G919A-induced lymphomas was extraordinarily large; the metastatic tumor filled the full length of the chicken's body cavity. The group receiving the negative control WT virus, which has a WT NRS sequence, exhibited a 10% (1 of 10 birds) rate of mortality from lymphomas, similar to the rate that was found by Smith et al. for LR-9 . Those infected with the other silent mutant, U916A, showed only a 14% (one of seven birds) mortality rate that was close to that for the negative control . U916A did not induce any short-latency tumors observable by gross examination, but one bird, which died 7 weeks after hatching, exhibited a large number of neoplastic follicles (20% of the total) in the bursa (for a detailed discussion of neoplastic follicles, see reference ). Additional symptoms exhibited by most chickens that died from short-latency lymphomas included anemia, ataxia, cachexia, and hepatomegaly. Short-latency lymphomas are associated with integrations into c-myb. | We next wanted to determine the integration sites of proviruses in chickens that died from short-latency lymphomas . Proviral integrations into the c-myb gene were seen previously in tumors induced by the EU-8 ALV, which has an NRS deletion identical to that of DeltaLR-9 . DNA was extracted from both bursa and liver tissues of the chickens. Proviral integration into c-myb was assayed by nested PCR by using primers in the viral LTR and either in c-myb exon 2 or upstream of c-myb exon 1 (primer location not shown). A PCR product indicated that there was an integration into c-myb in the same transcriptional orientation as c-myb. All of the chickens that died with lymphomas during the 10-week period after hatching (defined here as short-latency lymphomas) did indeed show proviral integrations into c-myb . In addition, a number of birds sacrificed at week 10 were also found to have lymphomas. We asked whether the chickens that developed lymphomas but did not die by week 10 also had proviral integrations into the c-myb gene. The percentage of chickens that developed these lymphomas is shown in Fig. . DeltaLR-9 induced lethal lymphomas in 44% (four of nine) and nonlethal lymphomas in 33% (three of nine) of the infected chickens; all of these birds exhibited proviral integrations into c-myb . Nonlethal lymphomas were induced by U916A in 28% (two of seven) of the chickens, but only one of these chickens had a c-myb integration . Of the chickens infected with the G919A ALV mutant, 75% (six of eight) developed lymphomas and died before week 10; all of these short-latency lymphomas were associated with proviral integrations into c-myb. No nonlethal lymphomas were observed in the remaining two birds infected with G919A. Finally, the negative control (WT) induced nonlethal lymphomas in 50% (5 of 10) of the chickens. Interestingly, only one of these chickens had a proviral integration into c-myb. The non-myb-associated, nonlethal (at week 10) lymphomas induced by either the WT NRS virus or the U916A mutant are likely to have proviral integrations into other proto-oncogenes. There were no morphological differences between the short-latency lymphomas and the nonlethal lymphomas (by 10 weeks) observable by either gross examination or histological staining (data not shown). Finally, by this PCR assay, c-myb integrations into DNA from both the bursa and liver were observed in the absence of obvious tumor formation in six chickens (three infected with U916A, one with G919A, and two with WT) but not in a control uninfected chicken. Even though observable tumors did not develop in these chickens, there appears to have been some selection for cells with proviral integration into c-myb. Integration sites cluster within 900 nt upstream of c-myb exon 2. | To determine whether specific integration sites correlated with the induction of short-latency lymphomas, we mapped the approximate sites of proviral insertion into c-myb based on the sizes of the PCR products. Strikingly, most of the integration sites clustered within 900 nt upstream of the start of c-myb exon 2 . In addition, one integration site was found upstream of c-myb exon 1 in the bursa and the liver of a G919A-infected chicken (by using an alternate upstream myb primer). There was no apparent difference between the integration sites found in lymphomas induced by the different viral constructs. In addition, short-latency lymphomas and lymphomas that did not lead to death in the time course of the experiment exhibited similar proviral integration sites. Most chickens that had an integration into c-myb intron 1 had only one such integration; however, multiple c-myb integrations (two or three) were seen in six chickens. Effects of mutations on the binding of factors to the NRS. | The NRS was shown to be important in preventing tumorigenesis since the virus with the NRS point mutation G919A caused a high frequency of short-latency lymphomas, even though it encoded a WT MA protein. Since the U916A mutant did not cause a high frequency of short-latency lymphomas, we wanted to determine whether the mutant RNAs differed in the binding of U1 snRNP or other factors. Affinity selection experiments were carried out with NRS RNAs from LR-9, DeltaLR-9, U916A, G919A, and WT (RSV) viral constructs . The NRS RNAs (RSV nt 703 to 930) were transcribed in vitro and covalently linked to agarose beads. After incubation with HeLa nuclear extract, the levels of binding of hnRNP H and ASF/SF2 proteins and U1 snRNA were determined by Western and Northern blot analysis, respectively. These particular factors were selected because they have been shown previously to have a functional interaction with the RSV NRS . The levels of binding of both hnRNP H and ASF/SF2 were found to be substantially reduced for DeltaLR-9 RNA relative to those for LR-9 RNA and RNAs of all the other constructs . This suggests that the primary binding sites for ASF/SF2 and hnRNP H in the NRS are in the region spanning nt 735 to 776. However, DeltaLR-9 RNA did show a low level of binding of both ASF/SF2 and hnRNP H, suggesting that there may be additional binding sites outside of the deleted region. McNally and McNally reported that ASF/SF2 binding was greatly decreased by the deletion of nt 715 to 748 of the RSV NRS , suggesting that there may be additional binding sites upstream of the deletion in DeltaLR-9, as shown in Fig. . Surprisingly, levels of U1 snRNA binding were found to be identical for LR-9 and DeltaLR-9 RNAs in this in vitro assay . Thus, ASF/SF2 binding to nt 735 to 776 of the NRS does not seem to be necessary for U1 snRNP recruitment to the 3' end of the NRS. Furthermore, the two viruses with NRS point mutations, U916A and G919A, both showed sharply diminished U1 snRNA binding . Thus, the difference in short-latency tumor induction by the U916A and G919A viruses could not be explained by differential U1 snRNP binding in this assay. In summary, the 5' region of the NRS (specifically nt 735 to 776) binds hnRNP H and ASF/SF2, and the 3' region of the NRS binds U1 snRNP . We were unable to differentiate between U916A and G919A on the basis of the binding of these factors; however, there may be other proteins that bind differentially to these two mutants. DISCUSSION : In this study, LR-9 and DeltaLR-9 ALVs, which differ only by a 42-nt deletion in the gag gene, were found to have the same replication rates and the same amounts of spliced env mRNA. However, readthrough transcription was found to be higher with DeltaLR-9 than with LR-9. In addition, the DeltaLR-9 NRS was shown previously to increase splicing to a downstream reporter gene . When injected into 10-day-old chicken embryos, the G919A, but not the U916A, NRS point-mutation virus showed an even higher rate of tumor formation than DeltaLR-9. Thus, the deletion in the viral MA protein was not found to be necessary for short-latency lymphoma formation. We conclude that the high frequency of short-latency tumors with proviral integrations into c-myb is associated with inactivation of the NRS, implicating RNA processing as the prime determinant of short-latency tumorigenesis in this system. The NRS is not necessary for viral replication in CEFs. | The replication rates of LR-9 and DeltaLR-9 were found to be identical in CEFs. Thus, a 42-nt deletion in the 5' portion of the NRS did not impair replication rates. Similarly, O'Sullivan et al. did not find substantial replication impairment in RSV mutants bearing single point mutations in the 3' portion of the NRS. However, some replication impairment was observed at early time points with RSV containing three scattered point mutations in the 3' region of the NRS . The DeltaLR-9 deletion is also in the MA protein-encoding region (corresponding to residues 119 to 132) of the virus. MA is the adapter protein that anchors the viral core to the cell's Env-studded lipid bilayer so that new virions can bud out of the cell . Consistent with our findings, Nelle and Wills found that the entire C-terminal half of RSV MA (residues 87 to 155) is dispensable for budding and infectivity in QT6 cells. Thus, this region of the virus does not seem to be necessary for replication, even though removing it deletes portions of the MA protein and of the NRS. Furthermore, the level of splicing to the viral gene env was not affected by the 42-nt deletion in DeltaLR-9 or by NRS point mutations in RSV . The NRS promotes polyadenylation. | Expression of truncated Myb to induce short-latency lymphomas necessitates readthrough transcription followed by splicing to c-myb . Readthrough transcription was found to be increased in the NRS mutant DeltaLR-9 in this study. The level of readthrough transcription was determined to be similar10% for LR-9. The deletion in the NRS increased the level of readthrough by 60%. The NRS was previously implicated by Miller and Stoltzfus in the regulation of polyadenylation of RSV RNA . In that study, large deletions, which overlapped in a 606-nt region including the NRS, led to readthrough of 50% of the viral RNA. More recently, Fogel et al. and O'Sullivan et al. found that more-limited mutations in the NRS also increased the level of readthrough transcription in RSV. Thus, one of the functions of the WT NRS appears to be the promotion of efficient polyadenylation at a site 6.5 kb downstream. Since polyadenylation can be promoted by an upstream 3' splice site and the ALV env 3' splice site is weak , the NRS may promote polyadenylation either directly or indirectly through interaction with the 3' splice site. The NRS regulation of polyadenylation may be modulated through hnRNP H or SR protein binding. This study reports that DeltaLR-9, which exhibited reduced hnRNP H and SR protein binding , showed an increase in readthrough transcription and a concomitant increase in short-latency tumor formation . There are two GGGA sequences in the deleted region (nt 735 to 776); this sequence has been shown to be a binding site for hnRNP H . Further, a member of the hnRNP H family, DSEF-1, has been reported to bind to a G-rich region downstream of the simian virus 40 poly(A) sequence, where it helps to recruit cleavage-stimulating factor to aid both in cleavage and in polyadenylation . hnRNP H may perform a similar function for the NRS. Binding of hnRNP H to an RSV NRS fragment containing nt 740 to 770 was previously reported by Fogel and McNally . However, an RSV mutant bearing extensive point mutations in this region failed to affect the polyadenylation frequency in transient transfection assays with a replication-defective virus . SR proteins may also modulate polyadenylation efficiency. For instance, SR proteins SAP75 and U2AF65 can interact with the poly(A) polymerase; however, this interaction has been seen to exert a negative effect on polyadenylation efficiency . Nevertheless, different SR proteins may have a positive effect on polyadenylation. Besides hnRNP H and SR proteins, components of the U1 snRNP, which also binds to the NRS, have been shown to recruit polyadenylation factors. The U1 snRNP-associated protein U1A has been found to interact with the cleavage-polyadenylation specificity factor and to aid in its recruitment . In addition, the U1 snRNP 70,000-molecular-weight protein has been found to bind to the polyadenylation polymerase and to inhibit polyadenylation of bovine papillomavirus RNA . In that case, however, the U1 snRNP binding site is immediately upstream of the AAUAAA polyadenylation signal. Thus, the NRS may promote polyadenylation not only through interactions with hnRNP H and SR proteins but also through binding of U1 snRNP or other factors. The NRS is a bipartite element: the function of the 5' end is not limited to recruiting U1 snRNP to the 3' end. | Our experiments showed that a single point mutation in the NRS can cause short-latency lymphomas that are associated with integrations into the oncogene c-myb, similar to the previously characterized lymphomas induced by DeltaLR-9. Thus, we have observed that mutations at either end of the NRS (the 5' end in DeltaLR-9 and the 3' end in G919A) can increase the frequency of short-latency lymphomas. This confirms the original observation that both 5' and 3' NRS regions are important for the function of the NRS . However, it has been proposed that the main function of the 5' region of the NRS is to bind SR proteins in order to recruit U1 snRNP to the 3' region of the NRS. McNally and McNally found that U1 snRNP binding to the NRS was greatly reduced in an affinity selection assay with a mutant that was missing the 5' region of the NRS containing the SR protein binding sites. In contrast, affinity selection experiments carried out in our lab demonstrated that the 3' region of the NRS (nt 903 to 939) was sufficient to bind U1 snRNP (data not shown). In addition, in the present study the DeltaLR-9 mutation (nt 735 to 776) did not inhibit U1 snRNP recruitment to the full-length NRS construct . Our data indicate that the function of the 5' region of the NRS is not limited to recruiting U1 snRNP to the 3' end since the mutation in DeltaLR-9 partially inhibits NRS function (suppression of readthrough and splicing) but the DeltaLR-9 NRS is nevertheless able to recruit U1 snRNP efficiently in vitro . The G919A NRS mutant causes a high incidence of short-latency lymphomas. | In the short-latency lymphomas studied here, the locations of proviral integration sites in c-myb intron 1 were similar to those in EU-8-induced tumors previously observed . Since most of the integrations that produced lymphomas clustered within 900 nt upstream of the start of exon 2, it is possible that the processivity of the transcribing polymerase may be a limiting factor for tumor development (reviewed in reference ). In addition, all of the selected integrations are downstream of a transcriptional pause site found in intron 1 of c-myb, which has been studied in human and murine c-myb genes . Also, integration into c-myb alone may be sufficient to cause short-latency lymphomas, because two of four tumors examined by inverse PCR had no other integrations (data not shown). We observed that a silent NRS point mutation (G919A) caused short-latency lymphomas effectively, showing that a defective NRS and not a truncated MA protein is the cause of the high frequency of short-latency tumor induction by DeltaLR-9. In the case of DeltaLR-9, increased readthrough may contribute to increased expression of truncated Myb, thus explaining, at least partially, why DeltaLR-9 causes a higher level of short-latency lymphomas than LR-9. Perhaps the NRS functions in ALVs to limit readthrough transcription into a wide variety of downstream cellular genes. Interestingly, the U916A mutant in RSV shows RNA readthrough levels that are identical to those of WT RSV, while readthrough is increased by 50% with the RSV G919A mutant virus . Thus, short-latency tumor induction with ALV mutants G919A and DeltaLR-9 correlates with increased readthrough of the poly(A) site, as determined in RSV for the mutants with point mutations . In addition, once readthrough has occurred, splicing to c-myb is most likely also increased in NRS mutants, analogous to the increase in splicing to src in RSV . RSV has two alternate 3' splice sites (for env and the cell-derived src) . O'Sullivan et al. found that the WT NRS in RSV limits splicing to the src gene but does not affect env splicing. Similarly, we did not observe an increase in env mRNA levels with an ALV bearing a deletion in the NRS. We propose that in an ALV, which lacks src, the NRS regulates splicing to a downstream cellular 3' splice site, like that of c-myb, after readthrough transcription . Furthermore, the RSV U916A mutant virus showed only a minimal increase in splicing to src, while RSV G919A showed a threefold increase in splicing to src compared to WT RSV . Considering the increase in both readthrough transcription (1.5-fold) and splicing (threefold) relative to those in the WT NRS virus, we predict that integration of the G919A virus into c-myb should promote approximately five times more expression of truncated Myb protein. U916A, on the other hand, does not exhibit a substantial increase in readthrough transcription or in splicing to src in RSV , and it also did not cause a high frequency of short-latency lymphomas in this study. Thus, the U916A mutant's phenotype is similar to that of the WT NRS virus in vivo; however, the U916A NRS differs from the WT NRS in that it does not bind U1 snRNP in vitro . Additional work will be necessary to resolve the apparent inconsistency between the effects of this NRS mutation in the viral context, assayed in vivo, and as an isolated RNA element assayed in vitro. In conclusion, the ALV NRS seems to inhibit both readthrough and splicing to downstream cellular genes, such as c-myb. In this manner, the NRS suppresses short-latency tumorigenesis, and it may also limit the occurrence of transducing viruses like RSV. FIG. 1. : Integration of the DeltaLR-9 ALV provirus into c-myb results in the expression of a truncated Myb protein. Integration of the DeltaLR-9 ALV provirus into c-myb results in the expression of a truncated Myb protein. The NRS element spans nt 703 to 930 within the ALV gag gene. DeltaLR-9 has a deletion of 42 nt (nt 735 to 776), which lies within both the NRS and the gag region corresponding to the MA protein (nt 381 to 904). When DeltaLR-9 integrates into the first intron of c-myb, transcription initiates normally in the 5' LTR but occasionally reads through the polyadenylation sequence in the 3' LTR to transcribe c-myb. Splicing then occurs from the viral 5' splice site (5'ss) to the 3' splice site (3'ss) of c-myb exon 2. Translation produces a truncated and highly oncogenic Myb protein that is missing its first 20 N-terminal amino acids . Viral integrations into c-myb were detected by nested PCR with 5' primers complementary to the viral LTR and 3' primers complementary to c-myb exon 2. FIG. 2. : Replication rates and env mRNA levels are identical for LR-9 and DeltaLR-9, but DeltaLR-9 shows an increase in readthrough transcription. Replication rates and env mRNA levels are identical for LR-9 and DeltaLR-9, but DeltaLR-9 shows an increase in readthrough transcription. (A) To study replication rates, reverse transcriptase (RT) activity was measured every 2 days after infection of CEFs with LR-9 and DeltaLR-9. Results representative of those from four experiments are shown. (B) Polyadenylated viral RNA was detected by Northern blotting with an LTR probe. Both LR-9 and DeltaLR-9 generated similar35% spliced env mRNA. A representative gel and the averages of results from three normalized experiments are shown. (C) Readthrough transcription was measured by RNase protections with a readthrough probe, which includes the LTR and flanking sequences and is similar to the one described by O'Sullivan et al. . Briefly, the 653-nt-long probe protects 397 nt of correctly terminated 3' RNA and 498 nt of readthrough RNA. The 255-nt fragment that represents the 5' RNA initiated in the 5' LTR is not shown. The percentage of readthrough transcription relative to the total amount of 3' RNA [readthrough and normally terminated 3' poly(A)] was found to be similar10% for LR-9. The figure shows a representative gel and the averages of results from four normalized experiments. FIG. 3. : Infection of 10-day-old chicken embryos with DeltaLR-9 and G919A ALVs causes high mortality from short-latency lymphomas. Infection of 10-day-old chicken embryos with DeltaLR-9 and G919A ALVs causes high mortality from short-latency lymphomas. (A) The mutants with silent point mutations, U916A and G919A, and the WT control were generated by placing a gag-pol fragment (BspEI-KpnI) from the respective RSV constructs into the LR-9 background. The NRS sequences from nt 703 to 930 are shown for LR-9 and DeltaLR-9 (black lines) and U916A, G919A, and WT (gray lines; derived from RSV). Point mutations are indicated by X's. The point mutations are in the 3' region of the NRS, do not overlap with the DeltaLR-9 deletion in the 5' region, and do not affect the Gag amino acid sequence. The line over the WT NRS sequence denotes the NRS region that is similar to the 5' splice site consensus sequence. (B) Ten-day-old chicken embryos were infected with the viruses shown above, and their mortality from lymphomas was monitored for 10 weeks after hatching (short-latency period). FIG. 4. : All short-latency tumors were associated with viral integrations into c-myb. All short-latency tumors were associated with viral integrations into c-myb. All of the chickens that died with short-latency lymphomas during the 10 weeks after hatching (bars on left; also plotted in Fig. ) had viral integrations into c-myb as observed by nested PCR. The bursa and liver of each chicken that exhibited nonlethal lymphomas 10 weeks after hatching were also examined for viral integration into c-myb (bars on right). These nonlethal lymphomas were associated with integrations into c-myb in all of the DeltaLR-9-induced tumors but to a lesser extent in the tumors induced by U916A and WT viruses. G919A caused only short-latency lymphomas. FIG. 5. : Proviral integration sites cluster within 900 nt upstream of c-myb exon 2. Proviral integration sites cluster within 900 nt upstream of c-myb exon 2. Each bar represents an independent integration site that was determined by nested PCR. Nearly all integration sites were within the 3' third of the 2.8-kb first intron of c-myb; in addition, one integration site was found about 1.4 kb upstream of myb exon 1. Larger boxes indicate integrations associated with short-latency lymphomas, and smaller boxes indicate integrations that resulted in lymphomas that were not lethal within 10 weeks after hatching. FIG. 6. : Reduced binding of hnRNP H and ASF/SF2 proteins, but not U1 snRNA, to the DeltaLR-9 NRS RNA. Reduced binding of hnRNP H and ASF/SF2 proteins, but not U1 snRNA, to the DeltaLR-9 NRS RNA. (A) Affinity selections were performed with in vitro transcribed NRS RNA (nt 703 to 930), which was covalently bound to agarose beads and incubated in HeLa nuclear extract. Bound hnRNP H and ASF/SF2 proteins were visualized by Western blotting, and bound U1 snRNA was visualized by Northern blotting. Equal loadings of total selected proteins in each lane were confirmed by a Coomassie-stained gel (data not shown). NE, nuclear extract. (B) The NRS has a bipartite structure, binding ASF/SF2 and hnRNP H at its 5' end and U1 snRNP at its 3' end. FIG. 7. : Comparison of the phenotypes of NRS mutants in RSV and ALV. Comparison of the phenotypes of NRS mutants in RSV and ALV. The RSV NRS mutant G919A, but not U916A, was previously found to cause increased readthrough of the viral poly(A) site and increased splicing to v-src compared to WT RSV . Accordingly, G919A transformed CEFs more rapidly than WT RSV . The RSV U916A mutant virus was indistinguishable from WT virus in these assays. Similarly, in the LR-9 ALV background, the NRS G919A mutant virus and viruses with the 42-nt deletion (Delta42), but not the U916A mutant virus, caused a high frequency of short-latency lymphomas. We propose that the RSV NRS regulation of splicing to v-src is similar to the ALV NRS regulation of splicing to downstream cellular genes like c-myb. 5'ss, 5' splice site; 3'ss, 3' splice site; ND, not determined. Backmatter: PMID- 12915545 TI - Heterologous Human Immunodeficiency Virus Type 1 Lentiviral Vectors Packaging a Simian Immunodeficiency Virus-Derived Genome Display a Specific Postentry Transduction Defect in Dendritic Cells AB - Heterologous lentiviral vectors (LVs) represent a way to address safety concerns in the field of gene therapy by decreasing the possibility of genetic recombination between vector and packaging constructs and the generation of replication-competent viruses. Using described LVs based on human immunodeficiency virus type 1 (HIV-1) and simian immunodeficiency virus MAC251 (SIVMAC251), we asked whether heterologous virion particles in which trans-acting factors belonged to HIV-1 and cis elements belonged to SIVMAC251 (HIV-siv) would behave as parental homologous vectors in all cell types. To our surprise, we found that although the heterologous HIV-siv vector was as infectious as its homologous counterpart in most human cells, it was defective in the transduction of dendritic cells (DCs) and, to a lesser extent, macrophages. In DCs, the main postentry defect was observed in the formation of two-long-terminal-repeat circles, despite the fact that full-length proviral DNA was being synthesized and was associated with the nucleus. Taken together, our data suggest that heterologous HIV-siv vectors display a cell-dependent infectivity defect, most probably at a post-nuclear entry migration step. As homologous HIV and SIV vectors do transduce DCs, we believe that these results underscore the importance of a conserved interaction between cis elements and trans-acting viral factors that is lost or suboptimal in heterologous vectors and essential only in the transduction of certain cell types. For gene therapy purposes, these findings indicate that the cellular tropism of LVs can be modulated not only through the use of distinct envelope proteins or tissue-specific promoters but also through the specific combinatorial use of packaging and transfer vector constructs. Keywords: Introduction : Lentiviral vectors (LVs) are useful tools for the delivery of genes with potential therapeutic applications. The major advantage of LVs over other gene delivery systems lies in their ability to pass through an intact nuclear membrane barrier and to integrate into the host genome. For gene therapy purposes, this translates into the possibility of integrating the transgene of interest, ensuring its durable expression, and the ability to target highly differentiated cells . LVs have now been developed from primate lentiviruses, such as the human and simian immunodeficiency viruses (HIV-1, HIV-2, SIVMAC251, and SIV1A11) , as well as from nonprimate lentiviruses . Common to all such vectors is the physical separation of functions normally encoded by the retroviral genome into separate constructs: one or more packaging constructs that encode the trans-acting factors required for particle formation and infectivity and a gene transfer vector. The latter represents essentially a miniviral genome, devoid of viral open reading frames but bearing the cis elements required for viral infectivity. As a consequence, the life cycle of LVs mirrors that of lentiviruses in a single-round infection but LVs are themselves nonreplicative. A major concern in the use of primate LVs is the pathogenicity of their parental viruses in humans. Although LVs are nonreplicative, the possibility exists that replication-competent viruses will be generated following recombination between vector and packaging constructs during the phases of vector production . It is clear that the possibility of such an event, although small, may temper enthusiasm over the use of LVs in gene therapy. Theoretically, a possible strategy to avoid formation of replication-competent viruses would be to lower the homology of regions shared between the packaging construct and the transfer vector, thus decreasing the likelihood of recombination. This may be achieved, for example, by engineering heterologous LVs in which packaging and vector constructs are derived from different but related viruses. It is well documented that particles of certain viruses can cross package and mobilize the genomic RNA of a related virus, giving rise to heterologous virion particles, or RNA pseudotypes, as in the case of the spleen necrosis virus and more recently HIV-1, SIV, and the feline immunodeficiency virus . Cross packaging is the result of a promiscuous interaction between elements present in the different viral species and involved in viral RNA packaging in cis (the packaging sequence, PSI) and in trans (the nucleocapsid domain of the Gag polyprotein, NC). This interaction is in most cases nonreciprocal. Indeed, spleen necrosis virus virion particles incorporate and mobilize murine leukemia virus viral genomic RNA, while the opposite does not occur. Similarly, HIV-1 has been shown to cross package HIV-2 and SIV RNAs , although the reciprocal situation does not occur (in the case of HIV-2) or occurs inefficiently (in the cases of SIV and feline immunodeficiency virus) . Since HIV-1's cross packaging of SIV RNA occurs efficiently , we used previously derived LVs to develop heterologous LVs in which trans-acting factors belonged to HIV-1 and cis elements in the transfer vector belonged to SIVMAC251 (HIV-siv). Heterologous HIV-siv LVs have been previously engineered and shown to be able to transduce aphidicolin-arrested cells . However, a close inspection of such vectors in comparison with homologous LVs and in a larger spectrum of cells had not been carried out . This analysis is required if such vectors are to be used in gene therapy applications. In this study, we analyzed the behavior of heterologous HIV-siv LVs in close comparison with that of homologous LVs in different human cell types. We show that despite the fact that the infectivity of heterologous HIV-siv LVs was similar to that of homologous LVs in most cells, they were, unlike the latter, incapable of transducing differentiated dendritic cells (DCs) and, to a lesser extent, macrophages. We characterized the infectivity defect of HIV-siv vectors in DCs and were able to show that the major impairment occurs in the formation of two-long-terminal-repeat (2LTR) circles, despite the fact that full-length (FL) proviral DNA was synthesized and associated with the nucleus. MATERIALS AND METHODS : Cells. | The human monocyte-like cell line U937 and the human myeloid leukemia cell line HL60 were obtained through the centralized facility for AIDS reagents supported by European Union program EVA/MRC and the United Kingdom medical research council. Primary lymphocyte and monocyte fractions were obtained from peripheral blood mononuclear cells of healthy donors at the Etablissement Francais du Sang de Lyon as described previously . Briefly, after isolation of peripheral blood mononuclear cells by centrifugation in lymphocyte separation medium (Eurobio, Les Ulis, France), cells were layered onto a four-step discontinuous density gradient (75 to 50.5 to 40 to 30% Percoll at a density of 1.130 g/ml; Pharmacia, Uppsala, Sweden) to separate monocytes from peripheral blood lymphocytes (PBLs). After centrifugation at 1,000 x g for 25 min, monocytes and lymphocytes were recovered as low- and high-density fractions, respectively. Prior to freezing, monocytes were further purified by negative selection by using a cocktail of hapten CD3, CD7, CD19, CD45RA, and CD56 anti-immunoglobulin E antibodies coupled to MACS microbeads by following the manufacturer's instructions (Miltenyi Biotec, Paris, France). U937 and HL60 cells, cells from the human T-cell line Jurkat , and PBLs were cultured in complete RPMI 1640 medium supplemented with 10% fetal calf serum (FCS). Human 293T and HeLa P4P5 fibroblasts (HeLa cells expressing the CD4 and CCR5 receptors and bearing an HIV-LTR-beta-galactosidase reporter cassette, obtained from Pierre Charneau, Pasteur Institute, Paris, France) were maintained in complete Dulbecco modified Eagle medium supplemented with 10% FCS. For stimulation, PBLs were treated for 24 h with 1 mug of phytohemagglutinin (PHA; catalog no. L8902; Sigma)/ml supplemented with 150 U of human recombinant interleukin-2 (IL-2; catalog no. 136; obtained through the National Institutes of Health [NIH] AIDS reagent and reference program)/ml or with 25 ng of IL-7 (R&D Systems)/ml. DCs were differentiated from monocytes upon culture for 4 to 6 days in granulocyte-macrophage colony-stimulating factor (GM-CSF; 100 ng/ml; Schering-Plough) and IL-4 (100 ng/ml; R&D Systems) in complete RPMI 1640 medium plus 5% FCS or human AB serum, as described previously . At the end of the differentiation period, cells were CD14- (CD1a+ in FCS), had increased surface expression of HLA-DR, CD80, and CD86, and upon maturation with lipopolysaccharide, CD40L, or tumor necrosis factor alpha, had upregulated CD83. DCs differentiated in this manner were fully competent in phagocytosis of latex beads and in the mixed lymphocyte reaction. Macrophages were obtained from purified monocytes. Briefly, monocytes were washed and plated at 106/ml/well of a 6-well plate in complete RPMI 1640 medium in the absence of serum for 3 h at 37C. By this time the majority of monocytes had adhered to the plate, and FCS and GM-CSF were added at final concentrations of 10% and 100 ng/ml, respectively. Macrophages were cultured for 6 or 17 days under these conditions. LV DNA constructions. | The HIV-1- and SIVMAC251-based LV systems have been described elsewhere (in references and and in reference , respectively). The packaging constructs, SIV3+ and 8.2, are hereafter referred to as SIV and HIV, respectively. Both constructs express the complete sets of structural and accessory viral proteins with the exception of Env and Vpu (in HIV-1) and are devoid of the packaging sequence, PSI, to impair incorporation of FL viral genomic RNA into virion particles. The lentivirus-derived vector genomes gae-pgk and pRRLsin.PPT.hPGK.GFPpre are hereafter referred to as siv and hiv, respectively. The vesicular stomatitis virus G (VSVg) envelope protein that pseudotypes lentiviral particles efficiently and allows for broad cell tropism was expressed from the plasmid MD.G . Virion particle production and purification. | LVs were produced by calcium phosphate DNA transfection of 293T cells and were pseudotyped with the VSVg envelope. Forty-eight hours after transfection, supernatant of transfected cells was harvested and purified essentially, as described before . Briefly, supernatant was filtered (pore size, 0.45 mum) and purified through a double-step sucrose cushion (45 to 25% [wt/vol]). After ultracentrifugation for 2 h at 25,000 rpm in a Beckman centrifuge, the interface between the 45 and 25% sucrose was harvested, diluted three times in phosphate-buffered saline, and recentrifuged under the same conditions onto a single 25% sucrose cushion. At the end of the ultracentrifugation, virions were resuspended in RPMI 1640 medium devoid of serum and supplemented with 10 mM MgCl2 and 200 muM deoxynucleoside triphosphates, aliquoted, and frozen. Normalization of virions contained in the different viral preparations was carried out either by normalization for protein content by using the exogenous reverse transcriptase assay as described previously or by determining the infectious titers after transduction of HeLa or 293T cells, which are used as standard cells in our laboratory. Generally, quantifications determined by protein content and infectious titers agreed. Cell transduction with LVs. | A total of 105 cells were used for transduction with LVs. Adherent cells were plated onto 12-well plates (24-well plates for macrophages) the day before transduction. Nonadherent cells were instead plated onto 96-well plates the same day as transduction. Transductions were carried out in the presence of 6 mug of polybrene (Sigma)/ml for 2 h at 37C. Cells were then spun directly in the plate; the medium was removed and replaced with fresh medium. Cell transduction was examined 3 to 4 days posttransduction by flow cytometry by determining the percentage of cells expressing the enhanced green fluorescent protein (eGFP) reporter gene carried by the LVs. When growth arrested, cells were gamma-irradiated at 3,000 rad. To distinguish and discard highly contaminated viral preparations, control infections were routinely performed in the presence of nucleoside analogs zidovudine (at 10 mug/ml; catalog no. 3485) and 2',3'-dideoxycytidine (at 20 mug/ml; catalog no. 220), both obtained through the AIDS reagent and reference program, NIH. These infections allow for the distinction between eGFP transgene expression following lentiviral transduction and eGFP protein carryover and internalization (transduction as opposed to pseudotransduction). Western blot analysis and antibodies. | Cell and virion lysates were prepared and analyzed by using standard procedures. A monoclonal anti-CA antibody that recognized both SIV and HIV CA (catalog no. 3537; obtained from the AIDS reagent and reference program, NIH) was used; the monoclonal anti-VSVg antibody was clone P5D4 (Sigma). Slot blot analysis of viral genomic RNA. | Slot blot analysis of viral genomic RNA was carried out essentially as described before . Briefly, virions produced by calcium phosphate transfection of 293T cells were purified by ultracentrifugation through sucrose as described above, normalized by using exogenous reverse transcriptase activity or infectious titers, and transferred onto a nylon membrane by using a slot blot apparatus. The membrane was incubated overnight at 42C in a solution containing 10% polyethylene glycol, 1.5x SSPE (1x SSPE is 0.18 M NaCl, 10 mM NaH2PO4, and 1 mM EDTA [pH 7.7]), 7% sodium dodecyl sulfate (SDS), and 100 mug of salmon sperm DNA/ml with a 32P-end-labeled DNA probe that hybridizes to the eGFP gene sequence 5'-GGCCGTTTACGTCGCCGTCCAGC-3' present on both hiv- and siv-derived vector genomes. The membrane was washed in 0.1% SDS-0.2x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate) and exposed for phosphorimager analysis. Analysis of reverse transcriptase intermediates in transduced cells. | For PCR analysis, DCs and HeLa cells were analyzed in parallel. Transduction was carried out at multiplicities of infection (MOIs) ranging from 0.5 to 2, as described above. Cell aliquots were harvested at 2 and 24 h posttransduction and lysed in a solution containing 50 mM KCl, 10 mM Tris-HCl (pH 8.3), 2.5 mM MgCl2, 0.5% NP-40, 0.5% Tween-20, and 120 mug of proteinase K/ml for 1 h at 60C. Proteinase K was inactivated for 30 min at 95C, and lysates were used directly for PCR analysis (in fivefold dilutions). PCRs were carried out with 2.5 U of Taq DNA polymerase (Promega) in the presence of a solution containing 1.5 mM MgCl2, 5 mM Tris-HCl (pH 8.0), 10 mM NaCl, 0.01 mM EDTA, 0.1 mM dithiothreitol (DTT), 5% glycerol, 0.1% Triton X-100, 200 muM deoxynucleoside triphosphates, and 10 pmol of each primer in a final volume of 50 mul. Amplifications were carried out by using 30 to 40 cycles as follows: 92C for 30 s, 55C for 30 s, and 72C for 45 s. Primer sequences were as follows (from 5' to 3'; nucleotide numbers within brackets refer to the complete SIVMAC251 sequence [accession number ]): minus-strand strong stop (MSSS) primers PE103, AGTCGCTCTGCGGAGAGGCTG (nucleotides [nt] 507 to 527), and PE83, TGCTAGGGATTTTCCTGC (nt 789 to 807); FL product primers 39, CCGTCGTGGTTGGTTCCTGCCG (nt 878 to 899), and 40, GCTAGATACCCAGAAGAGTTGGAAG (nt 294 to 309); 2LTR primers PE107, AGCTGCCATTTTAGAAGTAAGCC (nt 664 to 686), and PE151, TCTGACAGGCCTGACTTGC (nt 318 to 336); mitochondrial DNA (as in reference ) primers 98, GAATGTCTGCACAGCCACTTTCCAC, and 99, GATCGTGGTGATTTAGAGGGTGAAC; and actinup, CGAGAAGATGACCCAGATC, and actindown, TGCCGCCAGACAGCACTGTG. Probe sequences were, respectively, primer PE107; primer PE107; primer 40; probe 100, TGGGGTTTGGCAGAGATGT; and actinprobe, GGAGAAGAGCTACGAGCTGC. The products obtained after PCR amplification were transferred onto a nylon membrane and hybridized with the specific 32P-end-labeled probe. The intensity of the signals obtained after hybridization was determined by phosphorimager analysis. Subcellular fractionation. | Cells were transduced and analyzed 24 h posttransduction by following a procedure described previously, with minor modifications . Briefly, cells were washed in cold phosphate-buffered saline and then resuspended in a solution containing 10 mM Tris-HCl (pH 8.0), 1.5 mM MgCl2, 10 mM KCl, and 2 mM DTT. Cells were vortexed and incubated for 1 min on ice. Cells were then spun at 5,500 rpm for 5 min, and the supernatant was transferred to a new tube (fraction 1, cytoplasmic fraction). The fraction was clarified by centrifugation at 14,000 rpm in a Hedaeus centrifuge for 10 min. The nuclear fraction was washed twice in the buffer mentioned above, and the supernatant was pooled and clarified as described above (fraction 2, control for integrity of the nuclear fraction). The pellet containing the nuclei was lysed and vortexed in a solution containing 10 mM Tris-HCl (pH 7.4), 5 mM MgCl2, 160 mM KCl, 1 mM DTT, 0.6% SDS, and glass beads (fraction 3, nuclear fraction). All fractions were adjusted for identical salt concentrations and incubated at 60C for 1 h after addition of 120 mug of proteinase K/ml. Lysates were treated for PCR analysis, as described above. RESULTS : Cell-specific infectivity defect of an HIV-1 LV packaging an siv-derived genome. | To develop heterologous LVs with HIV-1 packaging an siv-derived genome, we used comparable, although not identical, versions of HIV-1 and SIVMAC251 LVs . The HIV-1 and the SIVMAC251 packaging constructs, referred to as HIV and SIV, respectively, are devoid of the packaging sequences PSI and env, code for Gag-Pro-Pol, and express the complete set of accessory viral genes. Both lentivirus-derived vector genomes, referred to as hiv and siv, bear an eGFP expression cassette. Besides this, they carry the cis sequences required for transcription, splicing (the major viral splice donor and acceptor sites), nuclear export (the Rev-responsive element [RRE]), packaging (PSI), and reverse transcription (primer binding site [PBS] and 3' polypurine tract [PPT]) of the vector genome. They also bear a central polypurine tract-central termination sequence (cPPT-CTS). The cPPT increases the efficacy of cell transduction by LVs , although the function of this sequence is actually debated . The woodchuck hepatitis virus posttranscriptional regulatory element (Wpre) that increases transgene expression by stabilizing mRNAs in cis is present on hiv but not on siv. The presence of this element increases transgene expression in HIV-1 vectors , but for unknown reasons it is detrimental during the transduction of certain cell types by SIV LVs . As such, Wpre was not present in our siv-derived vector genome. Homologous HIV-hiv and SIV-siv LVs were produced along with the heterologous HIV-siv vector by calcium phosphate transfection of 293T cells with the DNAs shown in Fig. . All LVs were pseudotyped with the VSVg envelope to allow for broad cell tropism of viral particles. As a first step towards the characterization of heterologous LVs, virions were normalized by exogenous reverse transcriptase activity and their infectious titers were determined upon transduction of HeLa cells. While all vectors had similar levels of infectivity, SIV-hiv LV titers were drastically reduced . Given the data described in the literature, this defect is not surprising and it is most probably caused by the inability of SIV virions to cross package an hiv genome . For this reason, SIV-hiv LVs were not used any further in subsequent experiments. To normalize the different viral preparations, we routinely determined their infectious titers in 293T or HeLa cells, highly permissive to both SIV and HIV transduction . This method very closely, albeit not identically, mirrored normalization by protein content, as determined by exogenous reverse transcriptase activity and Western blotting (data not shown). To exclude the possibility of pseudotransduction, i.e., the presence of intracellular eGFP signal in target cells due to protein carryover or phagocytosis rather than viral transduction, control test infections were routinely performed in the presence of reverse transcriptase inhibitors. Normalized amounts of virions were used for parallel transduction of HeLa cells, as a control, and DCs, the latter being an important cell target in gene therapy and in HIV pathogenesis. HeLa cells and DCs were transduced with the same MOIs in logarithmic dilutions . As expected, HeLa cells were equally well transduced by all vectors, irrespective of their cycling status (data not shown). DCs were more refractory to transduction with LVs than were HeLa cells, as indicated by the lower percentages of transduced cells for the same MOIs. However, both homologous SIV and HIV vectors transduced DCs well and routinely; transduction efficiencies of around 60 to 90% were attained with an MOI of 10. MOIs higher than 10 were avoided, since we noticed a negative impact on DC viability and morphology under these conditions (data not shown). The homologous SIV vector was between five- and 10-fold less infectious than its HIV-1 counterpart in human DCs, a defect present at low but not at high MOIs. This observation is reminiscent of the Fv1-like cell restriction recently described for primate lentiviruses and may explain why HIV-1 vectors transduce human cells better than their SIV counterparts (see below). In contrast to homologous SIV and HIV LVs, transduction of DCs with the HIV-siv vector was severely impaired (at least 100-fold). This impairment cannot be explained by loss of infectivity of heterologous HIV-siv LVs, as these do transduce HeLa cells efficiently. As certain postentry blocks may be determined by the entry pathway used by the virus , LVs were pseudotyped with an HIV-1 R5 envelope protein (JR-FL) and used for DC transduction . Similar to what was observed with VSVg-pseudotyped LVs, heterologous HIV-siv LVs were unable to transduce DCs. These results show a drastic infectivity defect of HIV-siv LVs in DCs that is independent from the viral entry pathway used. The infectivity defect of the HIV-siv vector is specific for DCs and, to a lesser extent, macrophages. | To determine whether the infectivity defect of HIV-siv virions was specific for DCs, different human cell lines and primary blood cells were analyzed for their susceptibility to transduction with LVs. Virions were purified as described above and used in the transduction of the human monocytic cell lines U937 and HL60 and primary macrophages at MOIs of 5, 0.5, and 0.05 . Similarly, identical amounts of virions were used to transduce the lymphocytoid Jurkat cell line and primary PBLs that had been activated with PHA-IL-2 or IL-7 for 24 h . As transduction of unstimulated PBLs did not yield appreciable transduction rates, it was not examined further (data not shown). All the cell lines tested in this study were transduced more efficiently than primary cells, regardless of the cycling statuses of the cells (data not shown). Generally, the homologous HIV-hiv vector had an advantage compared with SIV in the transduction of all the human cells tested in this study. The magnitude of this advantage varied among the cell lines (from five- to eightfold) and was more marked at low rather than high MOIs, at which SIV and HIV transduction rates were close to equal. With primary lymphocytes instead, HIV maintained its advantage over SIV at all MOIs. Contrary to what was observed with DCs, heterologous HIV-siv virion particles were able to efficiently transduce all the other cells tested in this study (with the exception of primary macrophages; see below). The infectivity of HIV-siv virions neared that of homologous HIV-hiv particles and was consistently higher than that of SIV. The only difference among homologous and heterologous HIV vectors was a slightly lower infectivity of the latter (one- to threefold), probably due to intrinsic differences between the hiv and siv genomic constructs. With primary macrophages however, the difference between homologous and heterologous HIV vectors reached 10-fold. This defect was not as marked as it was in DCs but was nevertheless readily distinguishable. Overall, these results show that the infectivity defect of HIV-siv virion particles is restricted to DCs and, to a lesser extent, macrophages, suggesting a cell-specific infectivity defect. Characterization of HIV-siv virion particles. | The fact that HIV-siv virion particles efficiently transduced different cell types argues against an intrinsic defect of such particles. However, to exclude completely this possibility, homologous HIV and SIV virions were compared with HIV-siv virion particles with respect to their protein profiles and genomic RNA contents. Normalized amounts of virion particles were analyzed by Western blotting by using an anti-VSVg and an anti-CA antibody that recognizes both SIV and HIV-1 CAs . As expected, virions contained similar amounts of Gag and VSVg proteins regardless of the vector genome incorporated (Fig. , lanes 2 and 3 and 5 and 6). Gag processing by the viral protease was also unmodified in HIV particles, regardless of the vector genome packaged. Next, the amount of viral genomic RNA incorporated into the particles was determined by slot blot analysis by using procedures described previously . RNA derived from MOI-normalized amounts of virion particles was transferred onto a nylon membrane by using a slot blot apparatus and hybridized with a 32P-end-labeled antisense probe specific for the common eGFP gene sequence . As controls for plasmid DNA carryover, viral RNA preparations were treated with RNase A. Densitometry analysis revealed no major differences among the levels of hiv- and siv-derived vector genomes incorporated into HIV virion particles. Taken together, our results indicate that there is no major structural difference among homologous and heterologous HIV-1 virion particles. Characterization of the early steps of transduction with HIV-siv LVs by PCR. | As the transduction defect of HIV-siv LVs was more dramatic in DCs, we decided to focus our efforts on these cells. To precisely characterize this defect, accumulation of reverse transcriptase intermediates was determined by semiquantitative PCR after cell transduction. To this end, the homologous SIV-siv vector and the HIV-siv vector were compared side by side. As the two vectors contain the same genome, PCR analysis is performed by using the same primers and the results obtained for the two vectors can be directly compared. HeLa cells and DCs were transduced in parallel at MOIs between 0.5 and 2 in different experiments, i.e., at MOIs at which differences between HIV and SIV vectors were minor in DCs. Cell aliquots were harvested at 2 and 24 h posttransduction and lysed, and intermediates representing the MSSS, the FL, and 2LTR circles were analyzed by PCR by using primers designed to distinguish each of these forms. Cell lysates were amplified in fivefold dilutions, and PCR was performed in linear conditions for semiquantitative analysis. The products obtained after PCR were transferred onto a nylon membrane and hybridized to a 32P-labeled internal probe specific for each amplified product . Consistent with the similar infectivities of SIV-siv and HIV-siv virion particles, accumulation of MSSS, FL, and 2LTR circle products was equivalent following transduction of HeLa cells. In DCs, the amounts of MSSS were also equivalent following transduction with the SIV-siv and HIV-siv vectors. FL products accumulated normally at 2 h, but levels of these products were six- to eightfold lower at 24 h posttransduction with HIV-siv than they were with SIV-siv. FL products were, however, readily obtained after transduction of DCs with HIV-siv. In contrast, accumulation of 2LTR circles was drastically reduced in DCs transduced with the heterologous HIV-siv vector (at least 100-fold compared to that in DCs transduced with SIV). Our results indicate that HIV-siv virion particles display a slight defect in FL proviral DNA accumulation and a major defect in the formation of 2LTR circles specifically in DCs. A defect in 2LTR circle formation may be explained by several possibilities: instability of proviral DNA, impairment of nuclear entry, or block at a post-nuclear migration step. Proviral DNA of HIV-siv LVs is associated with the nucleus following subcellular fractionation of transduced DCs. | To begin to distinguish among the hypotheses mentioned above, PCR analysis was performed with cytoplasmic and nuclear cell fractions obtained 24 h posttransduction of DCs . Nuclear and cytoplasmic extracts were obtained according to procedures previously established , and the presence of FL proviral DNA was determined by limiting-dilution PCR, as described above. Amplification of actin DNA was used as a control for the fractionation procedure and nuclei integrity. As expected, actin DNA was amplified essentially from the sole nuclear fraction, and only trace amounts could be found in the cytoplasmic and nuclear wash fractions, probably as a result of mechanical breakage of a small proportion of nuclei. When FL proviral DNA was examined, the amount was six- to eightfold lower following transduction with HIV-siv than that following transduction with the SIV-siv vector (as also shown in Fig. ), but in both cases, nearly all of the FL proviral DNA was associated with the nucleus. These results suggest that the preintegration complex (PIC) of the heterologous HIV-siv vector is able to access the nucleus, although it remains formally possible that proviral DNA might be tightly tethered outside the nuclear membrane rather than inside it. DISCUSSION : In addition to its purpose with regard to gene therapy, one objective of this study was to determine whether functions specified by cis elements and trans-acting factors of closely related lentiviruses were conserved during the transduction of different human cell types. The data collected here confirm earlier findings, obtained by using heterologous LVs on chemically arrested cells and primary human macrophages, that this is the case but add the notion that the extent of function conservation may be influenced on a cell-specific basis. Comparison of HIV-1 and SIVMAC251 LVs in the transduction of human cells. | Although the data presented here are not meant to constitute an exhaustive comparative study of HIV-1 and SIVMAC251 vectors, they provide some points of interest towards this end. Generally, HIV-1 LVs (both heterologous and homologous) have an advantage over SIV LVs in the transduction of all the human cells tested in this study, and reciprocally, SIV LVs seem more infectious in simian cells (Vero cells) (data not shown). However, with the exception of lymphocytes, the difference in transduction efficiencies between HIV and SIV lowers with increasing MOIs. This observation may reflect general adaptation of these lentiviruses for replication in their natural host and their ability to overcome restrictions such as the one specified by the resistance factor 1 (Ref1) and the lentiviral resistance gene (Lv-1) in primate cells that target incoming retroviral capsid proteins at different points postentry. Although this may partly explain the relative transduction efficiencies of SIV and HIV LVs in human cells, we ignore at present the question of whether the phenomenon observed in this study is specified through similar mechanisms. At any rate, an Lv-1-like restriction cannot explain the difference between homologous and heterologous HIV-1 LVs in DCs, as they share identical capsids and should thus be subjected to the same restrictions. Cell-dependent transduction defect of HIV-siv LVs. | We have determined that the transduction defect of HIV-siv LVs observed in DCs is independent from several parameters: the viral entry pathway, as similar results were obtained whether LVs were pseudotyped with VSVg or HIV-1 R5 (JR-FL) envelope proteins; the presence or absence of viral accessory proteins of HIV-1; the specific transgene expression cassette; the Wpre element; the degree of differentiation of DCs; and the blood donor (data not shown). In addition, the HIV-siv transduction defect was not rescued by treatment of DCs with As2O3, a drug that enhances retroviral transduction efficiency by stimulating reverse transcription (, ; data not shown). One possible explanation for the transduction defect of HIV-siv LVs may be their inability to transduce nondividing cells. We believe this explanation unlikely for several reasons. HIV-siv vectors are fully capable of transducing growth-arrested gamma-irradiated or aphidicolin-treated cells (data not shown) or PBLs treated with IL-7, a cytokine that promotes G0-to-G1b transition but not cell proliferation ; their proviral DNA is associated with the nuclei of transduced DCs, suggesting that PICs cross the nuclear membrane; and the degrees of transduction inhibition are different between DCs and macrophages, although both are nondividing cells, suggesting involvement of other factors besides cell cycle arrest, as recently suggested for lymphocytes . Interestingly, both macrophages and DCs are derived from monocytes, a cell population completely refractory to lentiviral transduction (; data not shown). In these cells, LV transduction results in normal levels of FL proviral DNA in the absence of 2LTR circles . Thus, it is possible to speculate that monocytes bear a restrictive cell environment that eases with differentiation into DCs or macrophages and becomes accessible to a certain extent to transduction with LVs. However, these cells may still express a repressive activity (or lack a facilitating activity) that homologous but not heterologous HIV-siv LVs are able to surmount, probably because one or more viral functions required for transduction are suboptimal in the latter. Viral determinants of the cell-specific transduction defect of HIV-siv LVs. | We believe that in addition to a cellular determinant, the defect of heterologous HIV-siv LVs in DCs is due to the swapping of cis and trans viral functions. Indeed, when these functions belong to the same virus, as in homologous LVs, transduction of DCs proceeds normally. Thus, we hypothesize the existence of an interaction between a cis element(s) and a trans-acting viral factor(s) required during transduction with LVs and normally influenced by cellular factors, hence its dependence on the cell environment. In its heterologous configuration, the interaction is suboptimal, resulting in a transduction defect only in restrictive cells. The identity of such interacting elements is currently unknown. As viral accessory proteins of HIV-1 are not necessary for transduction of DCs, we believe that a product of the structural Gag-Pro-Pol polyprotein may be involved in the interaction with a cis element (, ; data not shown). The siv vector contains six major viral elements: PBS, PPT, cPPT-CTS, PSI, RRE, and LTRs. PBS and PPT are nearly identical between HIV-1 and SIVMAC251, while PSI and RRE (which interact with NC and Rev, respectively) are required for assembly of infectious virion particles apparently normal for heterologous LVs. On the other hand, both the cPPT and the LTRs play a role during the early phases of transduction. The cPPT induces a structure on proviral DNA termed DNA flap and augments transduction efficiency of LVs , although the underlying mechanism remains controversial . Whether the cPPT acts by recruiting viral trans-acting factors and/or cellular proteins is currently unknown. The LTRs interact with the viral integrase (IN) through short attachment sequences (att) of about 10 nt positioned at their ends . The IN-att interaction displays sequence specificity, although elements of closely related viruses functionally recognize one another, as indicated by the infectivity of heterologous retroviral particles (, , , , ; this study). However, it is possible that the affinity of the IN-att interaction and thus the extent of its functional conservation is influenced by competition with factors differentially expressed in human cells, only a few of which are actually known . Postentry block after transduction of DCs with HIV-siv LVs. | Our data suggest that the defect of HIV-siv LVs in DCs occurs at a post-nuclear migration step prior to integration. Indeed, postintegration restrictions in several cells have been described before for wild-type lentiviruses, but these do not apply to LVs whose expression, once integrated, proceeds independently of all viral elements. The first defect observed upon transduction of DCs with HIV-siv LVs is a lower accumulation of FL proviral DNA that may result from lower efficiency of reverse transcription or from increased degradation of proviral DNA ends, as described for certain murine leukemia virus and HIV-1 NC mutants . We believe the latter unlikely. Indeed, similar kinetics of FL proviral DNA accumulation were obtained by using PCR primers at different positions in the 5' LTR (data not shown), indicating that if end degradation occurs, it is similar for both homologous and heterologous LVs. Once formed, FL proviral DNA is associated with the nucleus, where formation of 2LTR circles is severely reduced. Impairment of 2LTR circles has been generally taken as an indirect measure of a nuclear import defect. However, if the presence of 2LTRs is a measure of proviral DNA entry into the nucleus, as 2LTRs form only after ligation by nuclear ligases, their absence may in principle be due to impairment at a post-nuclear migration step. The fractionation data presented here argue in favor of this hypothesis and are supported by findings of similar defects in p12 mutants of the Moloney murine leukemia virus and by results of infecting human macrophages with certain strains of SIV (with the due caveat that the current fractionation techniques do not allow us to unequivocally distinguish between DNA inside the nucleus and DNA tethered outside it). Our understanding of the events that, after nuclear migration of PICs, lead to integration is still incomplete, as is our understanding of the identity of cellular proteins and the role that these proteins may play . Thus, it is possible that once in the nucleus the path that results in integration is a complex one and blocks at these stages may result in sequestration of proviral DNA in the nucleus in a form that does not allow its integration. In conclusion, the finding described here raises the possibility that an interaction between cis elements and trans-acting viral factors exists that is required for early steps of viral infection and that is modulated to different extents by cellular factors. For a more applicative approach, the results presented here show that restriction of transgene expression by LVs may be attained not only through the use of Env proteins or cell-specific promoters but also with the choice of specific combinations of packaging and transfer vectors. As HIV-siv LVs are impaired in the transduction of professional antigen-presenting cells but not other cell types, these vectors could be used to deliver genes of therapeutic interest, possibly lowering the risks of an immune response against the transgene itself. FIG. 1. : Schematic representation of the SIVMAC251- and HIV-1-based LVs used. Schematic representation of the SIVMAC251- and HIV-1-based LVs used. Structures of the SIVMAC251 and HIV-1 packaging constructs (A) and of the respective vector genomes (B) carrying an eGFP reporter gene under the control of the phosphoglycerate kinase 1 promoter (PGK). Original names of the constructs are reported within parentheses (capital letters for the packaging constructs, italic small letters for the vector genomes). SD and SA, major splice donor and acceptor sites, respectively. The U3s of both siv and hiv are self-inactivating and result in an impaired U3 viral promoter (U3*). cPPT actually denotes a composite cPPT-CTS element . Accessory genes carried by the packaging constructs are not detailed here. (C) Viral preparations were normalized by exogenase reverse transcriptase activity and used to transduce HeLa cells. The estimated infectious titers per milliliter are reported. FIG. 2. : Cell-specific infectivity defect of HIV-siv vectors in DCs. Cell-specific infectivity defect of HIV-siv vectors in DCs. VSVg (A and B)- and JR-FL (C)-pseudotyped LVs were produced by calcium phosphate DNA transfection into 293Tcells and purified by ultracentrifugation through sucrose, and their infectious titers were determined upon transduction of target HeLa or 293T cells. HeLa cells (105) and DCs were then transduced in parallel with logarithmic dilutions of homologous SIV-siv and HIV-hiv vectors along with the heterologous HIV-siv vector. Successful transduction events were analyzed by flow cytometry by determining the percentage of eGFP-positive cells 72 to 96 h posttransduction. Results obtained from a representative experiment are shown here, with the percentages of transduced eGFP-positive cells as the ordinates and the MOIs as the abscissas. FIG. 3. : Transduction of primary cells and different cell lines with homologous HIV, SIV, and heterologous HIV-siv LVs. Transduction of primary cells and different cell lines with homologous HIV, SIV, and heterologous HIV-siv LVs. Virions, produced as described in the legend to Fig. , were used to transduce monocyte/macrophage-like cell lines and primary macrophages (A) and lymphoid Jurkat cells along with primary PBLs stimulated with PHA-IL-2 or IL-7 (B). Results from representative experiments are depicted. The percentages of transduced eGFP-positive cells (ordinates) are shown as a function of MOIs (abscissas). FIG. 4. : Characterization of HIV-siv virion particles. Characterization of HIV-siv virion particles. (A) Western blot analysis of cell lysates (cell-assoc.; lanes 1 to 3) and virion particles (virion-assoc.; lanes 4 to 6) produced after calcium phosphate transfection of DNA into 293T cells. Virion particles were purified onto a double-step sucrose gradient prior to analysis. Western blots were probed by using an anti-CA antibody that recognizes a common epitope present on both HIV and SIV CA and with an antibody that recognizes the VSVg envelope, as indicated. The positions of migration of molecular mass markers (in kilodaltons) are indicated on the right. The products generated upon viral protease processing of SIV (p57-p45-p27) and HIV-1 (p55-p41-p24) Gag polyproteins and recognized by the anti-CA antibody are shown on the left. (B) Genomic-vector RNA incorporation as determined by slot blot analysis on normalized amounts of virion particles by using a probe that hybridizes to the common eGFP gene sequence present on both siv and hiv vector genomes. RNase A treatment of viral RNA preparations was included as a control for DNA contaminations. Twofold dilutions of viral RNA were used as a control for the linearity of the assay. Std. dilutions, standard dilutions. FIG. 5. : Semiquantitative PCR analysis of reverse transcription intermediates upon cell transduction. Semiquantitative PCR analysis of reverse transcription intermediates upon cell transduction. Normalized amounts of SIV-siv and HIV-siv vectors were used to transduce 105 HeLa cells and DCs, as indicated. Cell aliquots were harvested at 2 and 24 h posttransduction, lysed, and analyzed by PCR. PCR analysis was conducted with fivefold serial dilutions of starting material by using primers that recognized specifically the MSSS, the FL, and the 2LTR forms produced during the reverse transcription process, as indicated. Amplification of mitochondrial DNA (mtDNA) was used for DNA input and quality control. PCR products were transferred onto a nylon membrane and hybridized with P32-labeled specific internal probes prior to phosphorimager analysis. FIG. 6. : Subcellular fractionation and localization of proviral DNA in DCs. Subcellular fractionation and localization of proviral DNA in DCs. DCs were transduced with SIV-siv and HIV-siv vectors and fractioned 24 h later into cytoplasmic (fraction 1), wash (fraction 2), and nuclear (fraction 3) fractions, as indicated. PCR was then conducted with serial dilutions of samples by using a primer specific for FL proviral DNA and one specific for actin as a control for nuclear integrity. The amplification products obtained after PCR analysis were transferred onto a nylon membrane and hybridized with P32-labeled specific internal probes prior to phosphorimager analysis. Backmatter: PMID- 12915582 TI - Murid Herpesvirus 4 Strain 68 M2 Protein Is a B-Cell-Associated Antigen Important for Latency but Not Lymphocytosis AB - This work describes analyses of the function of the murid herpesvirus 4 strain 68 (MHV-68) M2 gene. A frameshift mutation was made in the M2 open reading frame that caused premature termination of translation of M2 after amino acid residue 90. The M2 mutant showed no defect in productive replication in vitro or in lungs after infection of mice. Likewise, the characteristic transient increase in spleen cell number, Vbeta4 T-cell-receptor-positive CD8+ T-cell mononucleosis, and establishment of latency were unaffected. However, the M2 mutant virus was defective in its ability to cause the transient sharp rise in latently infected cells normally seen in the spleen after infection of mice. We also demonstrate that expression of M2 is restricted to B cells in the spleen and that M2 encodes a 30-kDa protein localizing predominantly in the cytoplasm and plasma membrane of B cells. Keywords: Introduction : Murid herpesvirus 4 is an endogenous pathogen of free-living rodents of the Apodemus genus, e.g., wood mice . Infection of laboratory mice by murid herpesvirus 4 strain 68 (MHV-68; also called gammaHV-68) is an amenable model system for the study of gammaherpesvirus pathogenesis and for the development of therapeutic strategies against these viruses . Following intranasal inoculation of mice with MHV-68, a productive infection occurs in the lung . This is cleared around day 10 postinfection (p.i.) by CD8+ T cells , though the virus persists in a latent form in epithelial cells at this site . MHV-68 spreads to the spleen, where it also becomes latent, predominantly within B lymphocytes, but also in macrophages and dendritic cells . Establishment of latency in the spleen is associated with a marked splenomegaly, increase in lymphocytes in the spleen (splenic lymphocytosis) , and a subsequent peripheral mononucleosis that resembles that caused by primary infection of humans by Epstein-Barr virus (EBV) . Splenic lymphocytosis, which peaks at day 14 p.i., is driven by CD4+ T cells and is dependent on MHV-68-infected B cells in the spleen . Concomitant with the lymphocytosis is a sharp rise in the number of latently infected B cells, the resolution of which to a relatively constant baseline level is achieved by CD8+ T cells . CD8+ T cells along with antibody are important in the long-term control of persistent infection . Mononucleosis peaks at around day 35 p.i. and is characterized by the massive expansion of a diverse CD8+ T-cell population expressing the Vbeta4 T-cell receptor (TCR) chain in conjunction with a variety of TCR alpha chains . This differs from both EBV-associated mononucleosis and classical superantigen reactions in that the majority of T cells are not largely virus specific, nor is their expansion major histocompatibility complex (MHC) dependent . However, the Vbeta4+ T-cell expansion is dependent upon the presence of MHV-68-infected B cells and CD40 ligand-dependent CD4+ T-cell help . The left end of the unique region of the MHV-68 genome has attracted considerable interest due to the presence of four open reading frames (ORFs) (M1, M2, M3, and M4) and eight viral tRNA-like genes (vtRNAs), none of which have a homologue in other gammaherpesvirus genomes . M2 has attracted particular interest because its expression is restricted to latent infection in vitro and in vivo , where the peak of expression occurs around day 14 p.i. in the spleen (i.e., during splenomegaly), consistent with a role in the establishment of latency . The 193-amino-acid M2 protein, furthermore, contains an actively recognized CD8+ T-cell epitope, suggesting that immune regulation of M2-expressing cells is critical for the resolution of splenomegaly and maintenance of the host-virus equilibrium that underlies viral persistence . Construction of murine gammaherpesvirus with a mutation in M2. : All molecular, cellular, biological, and virological techniques were performed exactly as described in detail by Macrae et al. unless otherwise specified. To assess the function of M2, we generated a recombinant virus with a 4-bp deletion in the M2 ORF. To generate the recombinant, we used a modification of a technique that we had developed for rescue of wild-type (MHV-68) virus from the defective MHV-76 strain . Here, the 9.5-kbp deletion at the left end of the unique region of the MHV-76 genome (relative to MHV-68) was repaired by homologous recombination with a cosmid (cA8) that contains the relevant region from the wild-type MHV-68 genome (nucleotide sequence 115165 to 26842). To construct such a recombinant MHV-68 (rescuant) with a lesion in the M2 gene, we utilized a SalI restriction site at genome coordinate 4337 that is unique within cA8 to generate a 4-bp deletion within the M2 ORF . This deletion caused a frameshift that would result in premature termination of translation after amino acid residue 90 and mutated the known cytotoxic T-cell epitope in M2 . Specifically, following restriction with SalI (Invitrogen), cA8 was digested with mung bean nuclease (Invitrogen) to remove the 5' overhang and then religated to generate cA8DeltaM2. To ensure that the predicted frameshift mutation had been introduced, cosmids that lacked a HincII restriction site within the targeted SalI site were identified and subjected to DNA sequence analysis. To generate recombinant virus, BHK-21 cells were cotransfected with cA8DeltaM2 targeting vector and MHV-76 DNA as described previously , and the resulting plaques containing MHV with the reinserted 9.5-kbp DNA fragment containing the M2 mutation [76(RDeltaM2)] were identified by PCR with primers that amplified M2 . A wild-type rescuant [76(WTR)] containing the intact M2 gene was similarly generated with cA8 instead of cA8DeltaM2. After five rounds of plaque purification by limiting dilution, recombinant virus was characterized by Southern blot hybridization and PCR analysis to ensure that it was free of contaminating MHV-76. The presence and integrity of the M2 and mutated M2 ORF were first tested by restriction digestion of infected-cell DNA with HincII followed by Southern blot hybridization with the M2 ORF as a probe. The results showed the presence of the expected 2.2-kbp and 354-bp HincII fragments in the wild-type rescuant 76(WTR) (the expected 58-bp fragment was too small to be visualized on the blot). In 76(RDeltaM2) DNA the expected 2.2-kbp fragment was present, and the 354-bp fragment had been replaced by the predicted fragment of 412 bp due to the deletion of the HincII site at position 4337 and amalgamation of the 354- and 58-bp fragments. The integrity of the rest of the leftmost unique portion of the genome was confirmed by Southern analysis of HindIII-digested DNA with cA8 as a probe and by PCR analysis for M1 to M4 ORFs as described by Macrae et al. . Further, the gross integrity of the genomes of the two rescuant viruses was confirmed by restriction analysis of viral DNA with BamHI, HindIII, and EcoRI (data not shown). Finally, the integrity of the 4-bp mutation in 76(RDeltaM2) was confirmed by sequencing the DNA obtained by PCR amplification of the M2 ORF. M2 is not required for virus replication in vitro. : To assess potential effects of the mutation in M2 on MHV-68 replication, we performed single-step virus growth curves (multiplicity of infection [MOI] of 5) and multistep growth curves (MOI of 0.01) for MHV-76, 76(WTR), and 76(RDeltaM2) in BHK-21 cells. As shown in Fig. , all three viruses replicated with similar efficiency and in a similar fashion as that reported previously for MHV-68 . This result was not totally unexpected, since M2 is not expressed to any great extent during productive MHV-68 infection , and MHV-76 replicates normally in tissue culture . Recent experiments by Jacoby et al. , furthermore, have yielded similar results with respect to lack of a contribution by M2 to virus replication. Effect of M2 on productive replication in the lung. : The influence of M2 on productive infection in vivo was assessed by infecting groups of BALB/c mice with MHV-76, 76(WTR), and 76(RDeltaM2) and monitoring infectious virus in the lungs at various times p.i. by plaque assay. The data in Fig. demonstrated that there was no significant quantitative difference between the maximal titers achieved by infectious wild-type rescuant 76(WTR) (the equivalent of MHV-68) and those achieved by MHV-76; however, MHV-76 was cleared slightly more quickly. By contrast, the yield of 76(RDeltaM2) was minimally but significantly higher than that of either MHV-76 or 76(WTR) (day 5, P < 0.001), and infectious virus remained detectable through day 10 p.i. The results obtained for MHV-76 and 76(WTR) are similar to those reported previously . Unexpectedly, the mutation in M2 appeared to slightly enhance virus production in vivo. The reason for this is not clear, and, although the difference was statistically significant, it is unlikely that this relative increase in titer by only 1 log over 76(WTR) in the context of a 4.5-log titer is of biological significance, particularly given that the rate of clearance of 76(RDeltaM2) from the lungs was similar to that of 76(WTR). Recent experiments by Jacoby et al. also showed the lack of a contribution of M2 to virus replication. Thus, M2 does not negatively influence virus replication upon primary infection. Splenic lymphocytosis is not affected by mutation of M2. : The total number of splenic lymphocytes was determined in mice at various time points after infection with MHV-76, 76(WTR), and 76(RDeltaM2). As illustrated in Fig. , 76(WTR) induced a characteristic transient increase in splenic lymphocyte number that peaked at day 14 p.i. MHV-76 infection also caused an increase in cellularity, but cell numbers were significantly lower than those detected after 76(WTR) infection (day 14, P < 0.001; day 21, P < 0.01). In contrast, 76(RDeltaM2) induced a transient increase in spleen cell number of similar intensity and duration as that induced by 76(WTR). Thus, the mutation in M2, unlike the deletion present in MHV-76, did not affect the ability of MHV-68 to cause splenic lymphocytosis. It was possible, however, that even though the mutation in M2 did not influence the overall number of cells in the spleen, specific cell subpopulations could have been affected. Therefore, we determined the relative numbers of CD19+ B cells and CD8+ and CD4+ T cells within spleens at day 14 p.i. by fluorescence-activated cell sorting (FACS) analysis. The results indicated that similar relative levels of CD19+ and CD8+ cell populations were observed in mice infected with 76(WTR), 76(RDeltaM2), and MHV-76. However, while there was a similar level of CD4+ cells in mice infected with 76(WTR) and 76(RDeltaM2), the number of this subset was significantly elevated by comparison in MHV-76-infected animals (P = 0.007 by t test). In a similar fashion, the activation status of CD8+ and CD4+ T cells was determined by coanalysis of the levels of CD62 ligand (CD62L) expression, as cells expressing low levels of CD62L (CD62Llow) are more activated. As illustrated in Fig. , there was no difference in the numbers of activated CD8+ T cells in mice infected with any of the three viruses. However, there was a significantly increased number of activated CD4+ cells in MHV-76-infected mice compared to those infected with 76(WTR) and 76(RDeltaM2) (P = 0.007 by t test). The relative increase in CD4+ CD62Llow cells in MHV-76-infected mice was similar to the relative increase in CD4+ cells, indicating that there was no difference in the proportion of activated CD4+ cells in these mice. Thus, mutation in M2 did not affect the relative proportions or activation of the major spleen cell subsets that contribute to MHV-68-induced lymphocytosis. Disruption of M2 affects the normal pattern of latency. : Latent virus in the spleen was first measured at various times p.i. by an infective center (IC) assay . In mice infected with 76(WTR), measurable latent (i.e., reactivated) virus could be detected at day 7 p.i. and peaked at day 14 p.i.; by day 31 p.i. a level of approximately 400 ICs per spleen had been established. By contrast, in mice infected with MHV-76, the day 14 spike in the number of latently infected cells characteristic of wild-type MHV-68/76(WTR) was not observed, and overall the amount of latently infected cells detected was significantly less (day 14, P < 0.001). MHV-76 ICs were still detectable at day 31 p.i., at levels of approximately 50 per spleen, i.e., one-eighth of that obtained with wild-type MHV-68 infection. ICs in 76(RDeltaM2)-infected mice were detectable at day 7 p.i., and by day 14 latency had been established at a relatively constant level of approximately 5,000 ICs per spleen through day 31 p.i. At day 14 p.i. this was a significantly lower level than that observed in 76(WTR)-infected mice (P < 0.001) and significantly higher than that in MHV-76 infection (P < 0.001). The number of latently infected cells did not decrease from day 14 through 21 p.i. and was 1 log higher than that in 76(WTR) infection at day 31 p.i. (P < 0.01). These results indicated that mutation of M2 severely affected the characteristic acute rise in ICs (latently infected cells) within the spleen during primary infection and ultimately resulted in an increased level of latently infected cells. It should be noted that we did not detect preformed infectious virus in the spleen at any of the time points analyzed, which would have indicated a significant presence of productively infected cells. Thus, we conclude that the results obtained with the IC assay represented reactivation from latent infection. Because the mutation in M2 could be negatively affecting the ability of the virus to reactivate from latency, also resulting in lower estimates of latently infected cells, we determined the viral genome copy number in the spleens of infected mice by quantitative fluorescent PCR (QF-PCR). We have previously shown that genome copy number correlates directly with the number of latently infected cells in MHV-68-infected mice as determined by a limiting dilution reactivation assay . DNA was therefore isolated from the spleens of mice at days 14 (peak of lymphocytosis) and 46 p.i. and subjected to QF-PCR. The results indicated that, at day 14 p.i. with all three viruses, viral DNA copy number was in good agreement with the IC assay, i.e., DNA copy number and ICs obtained from mice infected with MHV-76 and 76(RDeltaM2) were approximately 10-fold lower than in 76(WTR)-infected mice (compare Fig. ). A positive correlation between copy number and ICs was also observed at day 46 p.i. However, the number of viral genomes was 50-fold higher in 76(RDeltaM2)-infected mice than in 76(WTR)-infected mice, a difference that was statistically significant (P < 0.0001) as determined by the Student t test. By contrast, in our IC assay, the number of latently infected cells in the spleens of 76(RDeltaM2)-infected mice was only 10-fold higher than that in the 76(WTR)-infected mice. This suggests that the actual level of latency established by the M2 mutant virus is higher than reactivable latency, indicating a possible defect in the ability of the mutant virus to reactivate from latency. To confirm that the effects on latency were due to the lesion in M2, we generated a second recombinant virus with a 256-bp deletion within the ORF (coordinates 4239 to 4494). This virus had a similar phenotype in vivo as 76(RDeltaM2) (data not shown). Thus, the effects seen were due to the mutation of M2 and not other secondary mutations in the viral genome. Finally, to confirm that the observed properties of the 76(RDeltaM2) virus were attributable to the frameshift mutation within the M2 gene, we determined the sequence of the M2 ORF amplified from DNA extracted from three separate viral plaques resulting from reactivation from three different mice at day 14 p.i. with 76(RDeltaM2). The results indicated that, in all cases, the only mutation in the M2 ORF was the 4-bp deletion that we had originally introduced (data not shown). Thus, we found no evidence for the selection of M2 revertant or pseudorevertant viruses in vivo and concluded that the striking reduction in the number of latently infected cells observed at day 14 p.i. in mice infected with 76(RDeltaM2) was indeed a consequence of the mutation within M2. A decrease in the peak of cells latently infected with M2 mutant viruses was also observed by Jacoby et al. . This fits well with previous observations that M2 is expressed during latency and that the peak of M2 expression in the spleen appears to occur around day 14 p.i . There was, however, a significantly greater number of latently infected cells present in the spleens of mice infected with the M2 mutant virus than in spleens of mice infected with MHV-76 at day 14 p.i. Because MHV-76 lacks M1 to M4 and the vtRNAs, one or more of these genes in addition to M2 likely contributes to the transient spike or rise in latently infected cells characteristically observed at this point following primary infection. At least part of this is due to the M3 protein, a soluble chemokine binding protein , since viruses with M3 deleted have a defect in their ability to cause this transient rise in latently infected splenic lymphocytes . A transient rise in latently infected cells shortly after infection is seen with other gammaherpesviruses such as EBV. The biological function of this is not known, but it is generally believed to promote viral persistence by ensuring that a critical pool of latently infected cells is established prior to maturation of the host antiviral immunosurveillance. Our results suggest that this may not be the case, at least for MHV-68, since both 76(RDeltaM2) and MHV-76 are capable of establishing long-term latency. However, it is possible that the transient rise in latently infected cells may play a more critical role, such as that proposed in EBV infection, in the context of a naturally acquired MHV-68 infection as opposed to one that is laboratory acquired. Surprisingly, despite the dramatic reduction in the number of latently infected spleen cells as a consequence of M2 mutation, the splenic lymphocytosis that is concomitant with the peak splenomegaly at day 14 p.i. was not significantly affected. Thus, for the first time we have been able to separate the spike in latently infected cells from the concurrent lymphocytosis, in which all spleen lymphocyte subsets are expanded . By contrast, MHV-76 was severely disabled in its ability to induce lymphocytosis, strongly suggesting that one or more of the genes M1, M3, and M4 or the vtRNAs are primarily responsible for this phenomenon. The precise causes of the transient lymphocytosis and rise in latently infected cells are unclear, but it has been assumed that they are interdependent. Specifically, in addition to their temporal overlap during primary infection, both are dependent upon latently infected B cells and CD4+ T cells . This led to two alternative hypotheses . The first is that the virus drives the expansion of latently infected B cells and that CD4+ and CD8+ T-cell expansion is in reaction to this. The second is that virus infection causes activation and expansion of CD4+ T cells, which in turn drive the expansion of latently infected B cells and CD8+ T cells. In the light of our present results, however, it seems likely that an alternative mechanism is operating, namely, that first the transient expansion of latently infected cells and splenic lymphocytosis are not dependent upon one another; second, transient expansion of the number of latently infected cells is virus driven, with M2 playing a major role; and third, the expansion of noninfected B cells, and CD4+ and CD8+ T cells, appears to be in response to latently infected cells, but this is dependent upon neither high levels of these cells nor M2 expression. Finally, as indicated above, other viral genes not present in MHV-76 are key to the expansion of spleen lymphocyte number. Given that M2 was required for the expansion of latently infected cells within the spleen during acute infection, it was somewhat surprising that mutation of M2 did not negatively affect the establishment of long-term latency. This is supported by similar observations obtained with other M2 mutants and with MHV-76 . Interestingly, the level of both reactivable latency and latent genomes maintained long-term by our M2 mutant in the spleen was significantly higher than that for either wild-type virus or MHV-76 . An elevation of the level of genome-positive cells in the spleen after peritoneal infection was also observed by Jacoby et al. , using a similar mutation in M2 (introduction of a stop codon). It is possible, therefore, in a way analogous to EBV LMP-2A, that M2 functions in part by inhibiting or preventing the overexpansion of virus-infected cells within the host. Since this expansion is not observed in mice infected with MHV-76, this effect is most likely dependent upon other gene products present at the left end of the MHV-68 genome. M2 does not affect Vbeta4-positive mononucleosis. : To assess whether the MHV-68-induced mononucleosis is in response to the M2 gene product, C57BL/6 mice were infected with MHV-76, 76(WTR), and 76(RDeltaM2), and their peripheral blood was assayed by FACS at day 35 p.i. as described previously . Two parameters of mononucleosis were measured: the presence of activated CD8+ lymphocytes (CD62Llow) and the levels of Vbeta4+ CD8+ T cells. As shown by the representative data presented in Fig. , the pattern of CD62L staining was similar regardless of the strain of infecting virus. Specifically, in each case the majority (99%) of CD8+ T cells expressed low levels of CD62L. By contrast, mice infected with either 76(WTR) or 76(RDeltaM2) had elevated levels of Vbeta4+ CD8+ cells (24 and 22%, respectively) relative to mice infected with MHV-76, which as expected had relatively normal levels (4%) of Vbeta4+ CD8+ T cells. Thus, the mutation in M2 affected neither the presence of elevated levels of activated CD8+ T cells in the periphery nor the high level of Vbeta4+ CD8+ cells in this population. Thus, MHV-68-induced mononucleosis is dependent on a gene other than M2 that is present within the leftmost 9.5 kbp of the unique portion of the MHV-68 genome, i.e., that deleted in MHV-76. Like lymphocytosis in the spleen, the viral factors responsible for the generation of either the peripheral CD8+ T-cell mononucleosis or the abnormal expansion of Vbeta4 TCR+ cells within this population have not been identified. However, like the transient expansion of spleen cells, Vbeta4 TCR+ CD8+ expansion is dependent upon the presence of MHV-68-infected B cells and CD40 ligand-dependent CD4+ T-cell help . The mutation in M2 affected neither mononucleosis nor expansion of Vbeta4+ T cells . However, while MHV-76 induced peripheral mononucleosis, it did not promote expansion of Vbeta4+ T cells. This indicates that specific viral products are required for Vbeta4+ T-cell expansion and that they reside within the deletion present in MHV-76, excluding M2. M2 expression occurs within B cells in the spleen. : Our previous analyses have shown that M2 mRNA is actively expressed at day 14 p.i. in the spleens of mice infected with MHV-68 and is detectable sporadically in splenic B cells at later time points p.i. . In addition to B cells, MHV-68 also infects splenic macrophages and dendritic cells . To ascertain in which spleen cell type(s) M2 is expressed, C57BL/6 mice were infected intranasally with MHV-68. B cells (total, activated, and resting), macrophages, and dendritic cells were then isolated by FACS from the spleens at day 14 p.i., and total RNA extracted from these cells was subjected to nested reverse transcription-PCR (RT-PCR) as described previously . In addition to M2, the expression of the MHV-68 M3 and cellular beta-actin mRNAs were analyzed using primers shown in Table . M3 was chosen because it is also characteristically expressed in the spleen at this time point ; detection of the beta-actin mRNA served as an indicator of mRNA integrity and a positive control for amplification. In serial dilution assays, our RT-PCR assay was able to detect M2 mRNA in as few as one S11 cell in 105 MHV-68-negative cells (BJAB B-lymphoma cells; data not shown). S11 is a latently infected murine B-cell tumor line that expresses M2 mRNA in virtually all cells , and thus the level of M2 expression in these cells is likely to be an accurate representation of its expression in latently infected cells in vivo. Additionally, upon spiking cDNA synthesized from the total RNA of MHV-68-negative cells with a full-length M2 cDNA, we determined that the sensitivity of our PCR assay was sufficient to detect between one and five copies of the M2 target (data not shown). To detect gene expression in spleen cell subsets, RNA from a total of at least three independent sorts for each cell type was analyzed. As shown in Table , whereas M3 transcripts were detected in all cell subsets analyzed, M2 mRNA was detected only in B cells, both activated and resting. beta-Actin mRNA was equivalently amplified in all RNA samples analyzed (data not shown). Our other studies have shown that the frequency of infection within each cell subset (determined by limiting dilution analysis-PCR), though highest in activated B cells, is well within the level of detection of this assay (10-5), as described above. Thus, these data strongly suggest that M2 expression in the spleen is B cell specific. Previous analyses of M2 have shown that it is expressed within the latently infected S11 B-cell line in vitro and during latency in vivo, most abundantly (as determined by RT-PCR) during the peak of lymphocytosis in the spleen at day 14 p.i . Here, we extend these analyses by showing that M2 is expressed predominantly, if not solely, within B cells and not in splenic macrophages and dendritic cells that also harbor latent virus. A previous report of M2 expression in spleen and peritoneal exudate cells following intraperitoneal inoculation of muMT knockout mice , which fail to develop mature B cells due to an inability to express the immunoglobulin M heavy chain (mum), would seem to be inconsistent with our finding here of B-cell-restricted expression of M2. However, recent findings indicate that muMT mice do indeed generate B cells, through an mum-independent pathway, that specifically produce immunoglobulin A . It is possible, therefore, that such B cells were the source of M2 expression in MHV-68-infected muMT mice. The M2 protein is cytoplasmic and plasma membrane associated in B cells. : The predicted amino acid sequence of the M2 protein lacks sufficient homology to any known protein or functional motif that would suggest its role during MHV-68 infection. Therefore, as a first step in defining M2 function, we transiently expressed it within mouse A20 B cells . Because of the lack of an adequate M2-specific antibody, M2 was expressed as either an N- or C-terminal fusion with enhanced green fluorescent protein (EGFP) by using the vectors pEGFP-C1 and pEGFP-N1 (Clontech) to aid in its detection. Immunoblot analysis of transfected A20 cells revealed the presence of a novel protein of similar60 kDa in the lysates of those cells transfected with an expression vector encoding either the N- or C-terminal EGFP-M2 fusion protein (data not shown). Accounting for the mass of the EGFPs, the molecular mass of M2 is similar30 kDa. This was in good agreement with the size of a hemagglutinin-tagged version of M2 stably expressed in A20 cells (data not shown) and previously reported unpublished observations by others . Of particular note was the observation that the molecular mass of M2 in the fusion was considerably greater than its predicted molecular mass of 22 kDa. The molecular mass of the protein by in vitro transcription and translation analysis was 26 kDa (data not shown). Thus, the molecular mass of 30 kDa for native M2 is most likely due to a combination of an aberrant migration of the unmodified polypeptide backbone in sodium dodecyl sulfate-polyacrylamide gel electrophoresis and a contribution of approximately 4 kDa by posttranslational modification(s). To determine the subcellular localization of the M2 protein, transiently expressed EGFP-M2 was examined in A20 cells at 24 h posttransfection by confocal microscopy with a Leica TCSNT confocal system. In some experiments, cells were fixed in 4% (wt/vol) paraformaldehyde and stained with either propidium iodide (PI; in phosphate-buffered saline containing 10 mug of RNase A [Qiagen] and 1 mug of PI [Molecular Probes] per ml for 30 min at 37C in the dark) to label DNA or tetramethyl rhodamine isocyanate (TRITC)-labeled reagents to detect specific cellular markers. Cells were observed using a combination of laser light wavelength and band filter specific for EGFP (488 and 525/50 nm) or PI-TRITC (568 and 600/30 nm). Image analysis was then used to form a picture with the resulting two images overlaid. Representative results are shown in Fig. . Cells transfected with empty vector (pEGFP-C1) displayed a diffuse pattern of EGFP-specific fluorescence that was present throughout the cytoplasm and nucleus as evidenced by the covisualization of EGFP and PI . In cells transfected with pEGFP-C1/M2 the EGFP-M2-specific fluorescence was primarily localized in patches near the cell margins and in the nucleus, as well as diffusely at a lower intensity throughout the nucleus and cytoplasm. However, nuclear staining was not universal and there appeared to be little crossover between the EGFP-M2 and PI fluorescence. A similar pattern of staining was seen with the C-terminal fusion protein encoded by pEGFP-N1/M2 and when a hemagglutinin epitope-tagged molecule was expressed in the same cells (data not shown). These results were consistent with M2 being principally localized in patches at or near the plasma membrane and within areas in the nucleus that did not contain chromatin. Because the cytoplasm in B cells is small and it is thus often hard to distinguish membrane and cytoplasmic fluorescence, we immunostained pEGFP-C1/M2-transfected cells with TRITC-labeled anti-MHC class II (monoclonal antibody SW 73.2 ) . MHC class II glycoprotein molecules are principally localized to the plasma membrane of B cells. The EGFP-M2-specific fluorescence was as observed previously, and a proportion of the patches of EGFP-M2 clearly colocalized with MHC class II, as evidenced by yellow staining when the two images were overlaid. Similar results were obtained when anti-CD19 (another B-cell surface marker) was used instead of anti-MHC class II (data not shown). Additionally, no colocalization was observed between EGFP-M2 and BODIPY ceramide (a marker for the Golgi apparatus) or LysoTracker Red DND-99 (a marker for lysosomes) (data not shown). Thus, the above results are consistent with a proportion of the M2 protein being cytoplasmic and a proportion being localized in patches either at or extremely close to the plasma membrane. Putative function of M2. : Our M2 expression data suggest that the effects of mutation of M2 as discussed above are B cell specific. Thus, in some respects, M2 function may be analogous to the perceived collective function of the EBV growth program of latency gene expression (which includes several B-cell-restricted genes): principally, to drive the rapid expansion of infected cells to establish a critical pool of latently infected cells. The function of M2, however, remains enigmatic. Preliminary biochemical analyses failed to demonstrate any posttranslational modification of the protein that might have shed light on this problem. While the apparent effects of M2 expression in the spleen and its localization to the cytoplasm and plasma membrane may argue for a role in cell signaling, the precise function of M2 and its specific role in MHV-68 latency remain to be determined. FIG. 1. : Targeted mutation of the M2 gene in MHV. Targeted mutation of the M2 gene in MHV. (A) Schematic representation of the left end of the MHV-68 genome. ORFs are shown as shaded arrows, and the eight vtRNAs are shown as small arrowheads. TRs, terminal repeat elements. Note that the M2 ORF is within the second exon of the M2 mRNA. The positions of the HincII restriction sites (genomic coordinates given) used to evaluate recombinant virus DNA are shown above. The position of the M2 probe used in Southern analysis of viral DNA is shown below. (B) Sequence of the M2 gene and protein surrounding the site of the introduced 4-bp deletion (shaded box) at the unique SalI restriction site. The amino acid sequence of the known cytotoxic T-lymphocyte epitope within M2 is shown in underlined italic boldface. (C) Southern blot analysis of DNA isolated from BHK-21 cells infected with MHV-76, 76(WTR), or 76(RDeltaM2) and digested with HincII. The blot was hybridized to a 32P-labeled M2 DNA probe (nucleotides 4000 to 4606). The positions of molecular size markers (in base pairs) within the gel are shown to the left of the autoradiograph. FIG. 2. : Biological characterization of viruses. Biological characterization of viruses. In each panel data obtained from infection with MHV-76 , 76(RDeltaM2) , and 76(WTR) are shown. All infections of mice (BALB/c) were intranasal with 2 x 105 PFU of virus. (A) Single-step growth curve comparing virus replication in BHK-21 cells infected at an MOI of 5. Data are representative of two independent experiments, each done in duplicate. (B) Multistep growth curve comparing replication of viruses in BHK-21 cells at an MOI of 0.01. Data are representative of two independent experiments, each done in duplicate. (C) Virus replication in the lung; data shown are the mean virus titers obtained from four mice per group at each time point. (D) Total spleen cell number within infected mice; shown are the mean cell numbers obtained from four infected mice per group at each time point. (E) Latently infected cells in the spleens of infected mice as determined by IC assay; data from four mice per group are shown at each time point. No preformed infectious virus was detected in this assay (data not shown). (F) Quantification of viral DNA in the spleen. Shown are the viral genome copy numbers per 300 ng of DNA as determined by QF-PCR at the indicated days p.i. The results are the mean values obtained from four mice per time point. In all panels, where indicated, bars represent standard errors of the means and asterisks over points represent a statistically significant difference from the wild-type value. FIG. 3. : FACS analysis of spleen cell subsets. FACS analysis of spleen cell subsets. Spleen cells from infected mice (as indicated) were stained with monoclonal antibodies to the indicated cell surface markers and analyzed by FACS. In each case the percentage of positive cells was calculated, and the means with standard errors of four individual animals are shown. Asterisks represent a statistically significant difference from the wild-type value. FIG. 4. : FACS analysis of peripheral blood lymphocytes. FACS analysis of peripheral blood lymphocytes. Peripheral blood cells of mice infected for 35 days with the indicated viruses were stained with labeled with monoclonal antibodies either to CD8 and Vbeta4 or to CD8 and CD62L and subjected to FACS. Data are representative of at least three separate experiments. FIG. 5. : Confocal microscopic analysis of M2 expression. Confocal microscopic analysis of M2 expression. A20 cells were transfected with either pEGFP-C1 (A to C) or pEGFP-C1/M2 (D to I) and after 48 h were either fixed, permeabilized, and then counterstained with PI or stained with TRITC-conjugated antibody to MHC class II antigens and then fixed. Cells were then visualized using a Leica TCNTS confocal microscope with combinations of laser light wavelength and band filter specific for EGFP (488 and 525/50 nm [A, D, and G]) or for PI and TRITC (568 and 600/30 nm [B, E, and H]). Image analysis with Leica software enabled the overlaying of these two images (C, F, and I). All images are shown at the same magnification. TABLE 1 : RT-PCR primers used in this study TABLE 2 : MHV-68 M2 and M3 expression in spleen at day 14 p.i. Backmatter: PMID- 12915556 TI - Edmonston Measles Virus Prevents Increased Cell Surface Expression of Peptide-Loaded Major Histocompatibility Complex Class II Proteins in Human Peripheral Monocytes AB - Gamma interferon (IFN-gamma) induces expression of the gene products of the major histocompatibility complex (MHC), whereas IFN-alpha/beta can interfere with or suppress class II protein expression. In separate studies, measles virus (MV) was reported to induce IFN-alpha/beta and to up-regulate MHC class II proteins. In an attempt to resolve this paradox, we examined the surface expression of MHC class I and class II proteins in MV-infected peripheral monocytes in the presence and absence of IFN-alpha/beta. Infection of purified monocytes with Edmonston B MV resulted in an apparent increase in cell surface expression of HLA-A, -B, and -C class I proteins, but it had no effect on the expression of HLA-DR class II proteins. MV-infected purified monocytes expressed IFN-alpha/beta, but no measurable IFN-gamma expression was detected in supernatant fluids. Class II protein expression could be enhanced by coculture of purified monocytes with uninfected peripheral blood mononuclear cell (PBMC) supernatant. MV infection of PBMCs also did not affect expression of class II proteins, but the expression of HLA-A, -B, and -C class I proteins was increased two- to threefold in most donor cells. A direct role for IFN-alpha/beta suppression of MHC class II protein expression was not evident in monocytes since MV suppressed class II protein expression in the absence of IFN-alpha/beta. Taken together, these data suggest that MV interferes with the expression of peptide-loaded class II complexes, an effect that may potentially alter CD4+-T-cell proliferation and the cell-mediated immune responses that they help to regulate. Keywords: Introduction : Measles virus (MV) is an immunosuppressive virus that can mediate profound transient suppression of a wide array of immune responses. These include decreased in vitro proliferative responses to mitogens and MV antigens, lymphopenia, decreased natural killer (NK) cell activity, increased plasma immunoglobulin E (IgE), and alterations in cytokine production . The virus is localized to monocytes/macrophages in infected peripheral blood mononuclear cells (PBMC) . Monocyte/macrophage activation represents a key step in the clearance of many virus infections . Once activated, monocytes/macrophages rapidly produce cytokines (e.g., alpha/beta interferon [IFN-alpha/beta] and tumor necrosis factor alpha) and nitric oxide, which have direct antiviral activity . Monocytes have a significant role in the immune response to virus infection in part because they express major histocompatibility complex (MHC) class I and class II proteins on their cell surfaces. Class I and class II proteins are critical to the presentation of virus-derived antigens to CD8+ and CD4+ T cells, respectively. Viral-antigen presentation to CD8+ T cells is mediated by MHC class I complexes of heavy chain and beta2-microglobulin, together with antigenic peptides derived from processed virus proteins . Class I complexes alert cytotoxic T cells to the infected state of the cell. In MV infections, clearance by T lymphocytes occurs during the rash phase of infection, i.e., 4 to 10 days after the onset of disease . Viral-antigen presentation by class II proteins is critical to the activation of CD4+ T cells, which mediate events such as B-cell help in antibody production and cytokine release. Antigenic proteins processed by antigen-presenting cells (APCs) (i.e., monocytes/macrophages, B cells, and dendritic cells) are deposited in late endocytic compartments, subjected to proteolytic cleavage, and presented by MHC class II proteins to CD4+ T cells . During MV infections, class II protein-dependent, CD4+-T-cell-mediated events, such as delayed-type hypersensitivity skin test responses to recall antigens, are suppressed prior to and long after the MV-induced rash . Some antiviral cytokines (e.g., IFN-alpha/beta and IFN-gamma) can activate monocytes/macrophages to up-regulate MHC class I protein expression and enhance processing of viral peptides so that they can be efficiently displayed by MHC proteins at the surfaces of infected cells . Transport of class II proteins is under tight control since their constitutive expression is limited to classical APCs and regulated by IFN-gamma . IFN-alpha/beta and IFN-gamma also serve to inhibit viral growth and limit the spread of infection . Although IFN-alpha/beta is synthesized by most infected cell types, synthesis of IFN-gamma is restricted to NK cells, NK T cells, and T cells . IFN-gamma is reportedly induced following MV infection . An examination of IFN-gamma levels in the plasma of measles patients found elevated levels of the cytokine, primarily during the rash phase of the illness . Recent studies of MV-infected Zambian children revealed that IFN-gamma production was primarily by CD8+ T cells and NK cells up to 7 days following the rash . In this study, we examine how MV affects IFN-induced expression of surface MHC proteins in human monocytes. Results from primary human monocyte cultures and from monocyte cell lines suggest that MV suppresses IFN-gamma-induced up-regulation of class II protein expression independently of IFN-alpha/beta. We suggest that these observations have important implications for understanding host resistance to MV infection. MATERIALS AND METHODS : Cell culture. | Blood from healthy adult donors was collected into heparinized tubes, and PBMC were isolated from lymphocyte separation medium (ICN/Cappel, Aurora, Ohio)-generated buffy coats. Cells were collected and processed as described previously . Monocytes were isolated using CD14 microbeads (Mitenyi Biotec, Auburn, Calif.) according to the manufacturer's protocol. Purity was assessed by flow cytometry (CD14, CD11b, and CD3) and was >99%. Viability as determined by trypan blue exclusion was >95%. In all experiments, cells were cultured in the presence of heat-inactivated human AB serum lacking MV antibodies. Duplicate experiments with heat-inactivated fetal bovine serum (FBS) were done routinely in parallel, with similar results observed. Some human donor cells were activated by FBS and showed increased surface MHC expression in the absence of virus. The limited numbers of donor cells prevented use of the same donors in every experiment. THP-1 cells (a monocytic cell line) were maintained in RPMI medium supplemented with 0.05 mM beta-mercaptoethanol and 10% FBS (HyClone, Logan, Utah). MelJuSo cells (a human melanoma cell line) were maintained in RPMI medium supplemented with 10% FBS. Virus preparation. | All experiments were done with the MV Edmonston (Edm) strain propagated in Vero cells (American Type Culture Collection, Manassas, Va.). Preparation of viral stocks has been described previously . To UV inactivate the virus, 1 ml of virus was placed in a 35-mm2 petri dish on ice and exposed to 400 muW of shortwave UV light/cm2 for 8 h. UV inactivation of the virus was confirmed by a plaque assay of the killed virus on a Vero monolayer. In vitro assays. | PBMC (107) or purified monocytes (PM) (106) were plated in 24-well plates (Corning Incorporated, Corning, N.Y.) and infected with MV at a multiplicity of infection (MOI) of 1 for 1 h at 37C in the absence of serum. Serum subsequently was added to a final concentration of 5%, and the cells were maintained at 37C for various incubation times prior to analysis. Control cells were mock infected with uninfected Vero cell lysate. For some experiments, purified monocytes were infected with MV to create APCs. After 48 h, the medium was removed from the cells and replaced with CD14-depleted PBMC (peripheral blood leukocyte) supernatant previously cultured for 48 h. This supernatant was prepared by depleting PBMC of CD14+ monocytes. CD14-depleted PBMC were kept in culture and then transferred to the infected monocytes after 48 h. The APCs subsequently were examined for surface MHC proteins by flow cytometric analysis. Cells were routinely stained for cytosolic MV nucleoprotein (NP) as an indicator of infection. Cytokine assays. | PM (106) were treated with 1,000 U of IFN-gamma (Biosource, Camarillo, Calif.; R & D Systems, Minneapolis, Minn.) per ml or 200 U each of IFN-alpha and IFN-beta (Biosource) per ml for 48 h at 37C. In antibody neutralization experiments, >1,000 U of antibody per ml of PM was used. Anti-IFN-alpha and -IFN-beta antibodies were purchased from Chemicon International (Temecula, Calif.). A direct enzyme-linked immunosorbent assay (ELISA) was used to quantitate the levels of IFN-gamma or IFN-alpha in experimental samples. Human IFN-gamma (HuIFN-gamma) (Biosource) or HuIFN-alpha (R & D Systems) was used as the assay standard. Cytokine controls and experiment samples were added to anti-IFN antibody-coated microtiter wells (Biosource or R & D Systems) and assayed according to the manufacturer's protocol. In brief, biotin-labeled anti-IFN antibody was used for detection and streptavidin-horseradish peroxidase was used for enumeration of wells with an ELISA plate reader at a wavelength of 450 nm. All standards and unknowns were run in duplicate for each experiment. Flow cytometry. | In brief, cells were washed in phosphate-buffered saline containing 0.5% bovine serum albumin and 0.05% sodium azide (FACS buffer) and resuspended in FACS buffer containing appropriately diluted antibody. Cells were stained for surface MHC class I and II protein expression with anti-HLA-A, -B, and -C locus products (clone G46-2.6; BD Bioscience Pharmingen, San Diego, Calif.), anti-HLA-DR (clone Tu36; BD Bioscience Pharmingen), anti-CD14 (clone RMO52; Beckman Coulter, Fullerton, Calif.), anti-CD3 (clone UCHT1; Beckman Coulter), fluorescein isothiocyanate-labeled KK2 (anti-MV NP; Chemicon International), and CD11b (clone Bear 1; Beckman Coulter). After being stained for 20 min at 4C, the cells were washed in FACS buffer and then fixed with 4% paraformaldehyde for 20 min at 4C. Cells were permeabilized in FACS buffer containing 0.1% saponin (permeabilization buffer [PB]) for 10 min at room temperature. Cells were washed once in PB and then resuspended in PB containing a specific antibody at the appropriate dilution. Internal staining for MV NP was carried out for 30 min at room temperature. Stained cells were analyzed by use of four-color flow cytometry (Beckman Coulter EPICS-XL) and EXPO32 ADC software (Beckman Coulter). Side-scatter versus forward-scatter parameters were used to gate cells by size. Side-scatter versus cell marker plots were used to analyze each specific mononuclear cell population. Quadrant analysis using FL1 (fluorescein isothiocyanate) versus FL2 (phycoerythrin) was used to differentiate infected cells from uninfected cells. Mean fluorescence intensity measurements from the generated plots are displayed as graphs in the figures. Slight variations in antibody staining from experiment to experiment were common, but trends for any given donor were always consistent. RESULTS : MV Edm infection of human PBMC does not significantly alter MHC class II protein expression. | MHC class II protein expression on the surfaces of PBMC from 10 donors with no documented recent exposure to MV or MV vaccine was examined following infection with Edm or mock infection with Vero cell lysate. MV infection, confirmed by testing for MV NP expression in the PBMC specimens, showed levels of NP expression among the PBMC populations examined ranging from 2 to 8%. Detection of cytosolic NP was used to demonstrate successful infection with MV. Levels of cytosolic NP do not correlate with the quantity of MV-derived peptides that would be available for loading onto MHC complexes. The range of CD14+ expression of the PBMC specimens examined before and after infection at an MOI of 1 was 13 to 18%. As a surrogate for evaluating the antigen-presenting capability in MV-infected cells, cell surface expression of peptide-loaded MHC proteins on CD14+ PBMC was evaluated by use of monoclonal antibody Tu36, which binds to peptide-loaded class II alpha/beta complexes at the cell surface . Class I A, B, and C complexes were measured with antibody G46-2.6 at the cell surface. In the presence or absence of virus, similar90% of cells scored positive for MHC class I and similar20% scored positive for MHC class II proteins. At MOIs higher than 1, there was considerable loss of the CD14 protein on cell surfaces (data not shown). No statistically significant differences in the levels of MHC class II protein expression were observed between infected and uninfected monocytes , suggesting that MV infection does not measurably alter the expression of MHC class II complexes on CD14+ MV-infected monocytes within the PBMC population studied. A small and consistent increase in class II protein expression was observed in 50% of donors. However, this increase was not statistically significant when it was analyzed by a Student t test. In contrast, infection of PBMC with Edm resulted in a two- to threefold increase of peptide-loaded MHC class I complexes at the cell surfaces of more than half of the CD14+ PBMC specimens examined . These results showed that MV infection of PBMC does not significantly increase MHC class II protein surface expression but appears to augment increased peptide-loaded MHC class I expression. Significant increased class II protein surface expression does not occur despite the elevated levels of IFN-gamma expression that result from infection . To further evaluate this effect, we examined MHC expression on PM from two donors, C and F, following Edm or mock infection . As predicted from the PBMC studies, no significant increase of surface MHC class II-peptide-loaded complexes was detected 24 to 48 h after infection. Kinetic studies have shown that increased class I surface expression mediated by IFN-gamma in cell lines precedes the induction of class II protein expression and peaks earlier ; therefore, we also examined cells at 72 h after infection. At the later time point, surface class II protein levels began to decrease, not increase (data not shown). As in PBMC, an approximately twofold elevation in MHC class I HLA-A, -B, and -C proteins was detected in the PM at 48 h after infection . UV-inactivated MV also elevated class I proteins in this assay (data not shown). MHC class II protein expression on MV-infected CD14+ cells is not significantly affected by increasing MOI. | MHC class II protein expression on PM or PBMC was examined following infection with 0, 0.1, 1, 5, and 10 MOIs of Edm . With increasing amounts of inoculating virus, no significant differences in peptide-loaded MHC class II protein expression were observed for Edm-infected PM or PBMC at 48 h after infection. In addition, similar levels of MHC class II protein expression were detected when anti-DP and -DQ class II antibodies were used, suggesting that levels of class II DP and DQ proteins, like levels of DR proteins, were not affected by Edm infection (data not shown). Furthermore, identical profiles were seen in tests for surface class II DR proteins with the DR-specific antibody L243, indicating that class II molecules in MV-infected monocytes had not lost the Tu36 epitope (data not shown). To confirm that MHC class II protein expression could be elevated on PM and PBMC, mock-infected and Edm-infected cells were treated with 1,000 U of recombinant HuIFN-gamma per ml of sample . The units of cytokines used to treat the cells were determined on the basis of published literature and on dose-response experiments on PM (data not shown). Interestingly, although levels of MHC class II proteins were increased on IFN-gamma-treated PBMC, MV appeared to suppress IFN-gamma-mediated up-regulation of surface MHC class II protein expression on PM cells , and it did so even after 72 h (data not shown). Similar levels of MV NP were detected for treated and untreated PBMC and PM (data not shown), though IFN-gamma has been shown to inhibit MV replication . Taken together, these results further suggest that MV infection results in a failure to up-regulate class II protein expression. It is also possible that MV infection results in the elevation of additional factors in PM that further interfere with class II protein surface expression. MV infection is associated with IFN-alpha production. | MV induces infected human monocytes to produce IFN-alpha, and in turn, IFN-alpha suppresses MV replication in those cells . To determine the levels of IFN expression associated with Edm infection of PBMC or PM, supernatants from Edm-infected PBMC and PM were examined for IFN-alpha and IFN-gamma expression . Equivalent numbers of CD14+ cells were examined in the PM and PBMC. Measurements of IFN-alpha and IFN-gamma levels in the supernatants of Edm-infected PBMC or PM showed considerable induction of IFN-alpha (similar100 U/ml) in all virus-infected supernatants and induction of IFN-gamma (similar17 to 20 U/ml) in MV-infected PBMC . These findings are consistent with reports of IFN-alpha being produced primarily by monocytes and consistent with the localization of MV NP to CD14+ cells (data not shown). As a control, the levels of IFN-alpha expression by Edm-infected and uninfected Vero cells were determined, and as expected, no substantial levels of expression were observed. It thus appears that inhibition of increased MHC class II protein expression on PM occurs in the presence of IFN-alpha. Up-regulation of MHC proteins by IFN-gamma is suppressed in MV-infected CD14+ cells. | MHC expression on PM or PBMC was examined following infection of cells with live virus or treatment with UV-inactivated virus . IFN-gamma-induced expression of both class I and class II proteins in Edm-containing PM was inhibited compared with expression in the control . Thus, only the IFN-gamma-treated control PM up-regulated class I and class II surface complexes. In contrast, IFN-gamma-treated PBMC elevated class I and class II protein expression in both control and Edm-infected cells , further suggesting that the PM may lack factors necessary to activate increased class II protein expression. PM are depleted of numerous cytokine-producing cells. Suppression of class II protein expression was also detected following treatment of PM with killed MV . Treatment with UV-inactivated virus did not up-regulate class II protein surface expression, and the inactivated virus may suppress IFN-gamma-induced up-regulation of class II protein expression. These same observations were made when supernatants from Edm-infected monocytes were inactivated and then added to a fresh culture of monocytes (data not shown). These results suggested that viral replication was not required to inhibit IFN-gamma-induced up-regulation of MHC proteins in PM . Since IFN-alpha was detected in the monocyte-derived supernatant, we tested whether neutralizing IFN-alpha/beta would allow IFN-gamma to up-regulate class II protein expression. However, antibody neutralization of IFN-alpha/beta also failed to prevent the inhibition of IFN-gamma-induced up-regulation of class II protein expression, thus implying that IFN-alpha/beta did not directly inhibit IFN-gamma. Taken together, the data suggest that IFN-alpha/beta may not play a role in the inhibition of IFN-gamma-induced up-regulation of surface class II peptide-loaded complexes on peripheral monocytes. IFN-alpha/beta does not prevent IFN-gamma-induced elevation of class II proteins in PM. | Class I and class II proteins were analyzed on PM 48 h following treatment with IFNs . As expected, IFN-gamma treatment elevated both class I and class II proteins, whereas IFN-alpha/beta treatment elevated only class I protein expression. Contrary to findings for cell lines , IFN-alpha/beta treatment did not inhibit IFN-gamma-induced expression of surface class II proteins in any donor tested. The effects of type I IFNs on IFN-gamma-induced class II protein expression have previously been reported to be tissue specific and not to occur in human PM . Only class I proteins are elevated in MV-infected class II protein-positive cell lines. | Class II protein-positive cell lines MelJuSo and THP-1 were infected with Vero cell-adapted Edm for 48 h, and the cell surfaces were scored for class II protein expression . Levels of class I proteins were recorded for comparison. Both cell lines were readily infected with Edm (similar40% of both cells lines scored positive for MV NP). Greater than 90% of both cell lines scored positive for class I and class II protein expression; only 4% of THP-1 cells scored positive for class II protein expression. MV infection elevated class I proteins two- to fourfold , and this augmentation was dose dependent (data not shown). Class II protein expression was not elevated at 48 h or at 24 or 72 h (data not shown) after infection. In addition, no IFN-alpha was detected in the supernatant by ELISA (data not shown). We therefore conclude that MV-infected cell lines also fail to up-regulate class II proteins following infection. Under these conditions, no IFN-alpha was released into the supernatant, suggesting that at a minimum, IFN-alpha did not play a role in suppressing class II protein enhancement following MV infection. Thus, the data do not support a role for IFN-alpha/beta in suppressing class II protein levels in Edm-infected monocytic cells. PBMC-derived factors promote an increase in surface class II protein expression in Edm-infected CD14+ cells. | It might be inferred from the data that MV-infected PM respond to exogenous IFN-gamma if they are sufficiently activated. To test this, surface class II protein levels on Edm-infected PM were examined following treatment with supernatants derived from CD14-depleted PBMC. At 96 h, similar16% of CD14 cells scored positive for class II protein without the PBMC-derived supernatant, whereas similar78% of CD14 cells scored positive for class II protein with the PBMC-derived supernatant. Similarly, similar17% of CD14 cells scored positive for class I protein without the PBMC supernatant, whereas for both mock- and Edm-infected cells, similar65% of CD14 cells scored positive for class I protein with the PBMC supernatant. The supernatant derived from peripheral blood leukocytes contained similar20 U of IFN-gamma per ml as determined by ELISA. Treatment with the PBMC-derived supernatant stimulates PM and restores class II protein levels to those of the control . We confirmed a partial role for IFN-gamma in this restoration through antibody blocking experiments (data not shown). It is highly likely that additional cytokines, which contribute to activation of the monocytes, also play a role. These factors, in concert with IFN-gamma, appear responsible for elevating class II proteins in stimulated Edm-infected monocytes. DISCUSSION : It has been reported that MV up-regulates the surface expression of class II proteins in human PM and, furthermore, that IFN-gamma, the activator of class II transcription, is not involved in this process . The type I IFNs (i.e., IFN-alpha and -beta) are induced in MV-infected monocytes and appear to inhibit viral pathogenesis by halting viral replication . Interestingly, IFN-alpha and -beta have recently been shown to interfere with or suppress class II protein expression in many viral systems . Thus, it appears from the literature that MV infection results in an up-regulation of MHC class II molecules in an IFN-alpha/beta-rich milieu. In an effort to resolve this paradox, the present study was done to examine the influence of type I and type II IFNs on class II protein surface expression in MV-infected monocytes. We were surprised to find that MV infection of purified monocytes did not significantly alter the expression of surface-displayed MHC class II proteins, a finding in contrast with that from the earlier report . Moreover, MV-infected purified monocytes failed to up-regulate class II proteins following stimulation with IFN-gamma, an event that readily occurs in the absence of virus when cells are treated with exogenous IFN-gamma. We found that IFN-alpha/beta, at least in the specific case of MV-infected human monocytes, does not appear to play a role in this inhibition. Finally, we showed that the addition of MV-infected PBMC culture supernatant to purified monocytes results in greater surface expression of class II molecules than does the addition of exogenous IFN-gamma alone, suggesting that additional factors play a role in MHC class II protein regulation in this system. Expression of MHC class II proteins on monocytes is central to the development of CD4-mediated T-cell responses. Activated APCs expressing MHC class II molecules along with appropriate costimulatory molecules present processed antigens to CD4+ T cells. CD4+ T cells in turn regulate both cellular and humoral immune responses. Cytokines produced by activated CD4+ T cells, CD8+ T cells, and innate immune cells orchestrate the developing immune response. Failure to up-regulate class II protein expression following MV infection may provide a mechanism to delay the onset of CD4-mediated T-cell responses, such as macrophage activation and IFN-gamma production, thereby facilitating MV replication and spread. CD8+ T cells are known to clear MV-infected cells in the context of substantial suppression of CD4+-T-cell-mediated events . In addition, an investigation of cytokine levels in children with measles found that it was predominantly the CD8+-T-cell and NK cell lymphocyte population, not the CD4+-T-cell population, that produced IFN-gamma . Findings from our study of MV-infected monocytes are consistent with those observations since they demonstrate a circumstance in which class I-restricted activation may occur in the context of reduced class II protein-restricted activation. A number of microorganisms evade immune surveillance by preventing signaling via class II proteins . For example, Toxoplasma gondii causes down-regulation of MHC class II molecules and an inability to up-regulate class I molecules , and the Epstein-Barr virus gene BZLF2 encodes a 42-kDa protein that interacts with the HLA-DR beta chain and inhibits class II protein-dependent antigen presentation in B cells . Furthermore, human cytomegalovirus encodes a protein, US2, that targets class II DR-alpha and DM-alpha molecules for degradation by the proteosome . Human immunodeficiency virus infection has also been shown to impair class II protein expression in infected primary monocytes . Two fairly well characterized mechanisms that explain how some viruses interfere with class II protein signaling have been described. In these mechanisms, IFN-alpha/beta and/or viral gene products down-regulate class II gene transcription by down-regulating levels of the class II transcriptional activator (CIITA) protein . Paramyxoviruses are reported to interfere with the cytokine signaling pathways of the IFNs . We examined the influence of type I and type II IFNs on class II protein expression in MV-infected monocytes. Our results show that MV inhibits IFN-gamma-induced expression of class II proteins, and they strongly suggest that this inhibition does not occur via IFN-alpha/beta. IFN-alpha/beta was of interest because of the known role of IFN-alpha/beta in suppressing IFN-gamma-induced expression of class II proteins in CIITA-transfected human cells , mouse cytomegalovirus-infected mouse macrophages , human cytomegalovirus-infected human endothelial cells , and human parainfluenza virus 3-infected human cell lines . However, this did not appear to be the mechanism by which increased class II protein surface expression was prevented in MV-infected monocytes. The above-mentioned studies also report a role for viral gene products in suppressing class II gene transcription. On the basis of recent studies of Sendai virus, it has been suggested that MV likely uses its nonstructural V and C proteins to obstruct IFN responses . We plan to examine correlations between V and C protein expression and MHC expression in MV-infected monocytes. An examination of specific viral proteins which interfere with the IFN response is the subject of ongoing work in our laboratory. Wild-type and vaccine strains of MV appear to differ both in their response to type I and type II IFNs and in their ability to induce IFN-alpha/beta in infected cells . Vaccine strains, or strains that have been laboratory adapted in Vero cells, which use CD46 for entry, were shown to induce significant levels of IFN-alpha/beta, whereas wild-type strains and B95a-passaged strains induced 10- to 80-fold less IFN-alpha/beta . In our model system, vaccine strains use CD46 for entry, readily infect the monocyte population, and secrete IFN-alpha. Though wild-type MV strains use the receptor CDw150 or the human signaling lymphocyte activation molecule for entry, recent data suggest that they also use CD46 for entry, albeit with lower affinity . Pilot experiments carried out with wild-type strains of MV in this system revealed low levels of infection and no up-regulation of class II proteins and did not detect IFN-alpha in supernatant fluids (unpublished observations). Since wild-type viruses reportedly inhibit the secretion of IFN-alpha/beta , we plan to examine the impact of wild-type entry via the human signaling lymphocyte activation molecule on class II protein surface expression. It is possible that wild-type viruses not yet adapted to Vero cells up-regulate class II molecules. Jacobson et al. reported that much of the cytotoxic-T-lymphocyte (CTL) response to measles in humans was MHC class II protein restricted. Others have reported that although CD4+- and CD8+-T-cell killing of MV-infected cells was intact in the system examined, these CTLs were slow to be generated . It was proposed that this ability to delay the generation of CTLs would eventually allow killing but only after additional viral spread . We are keen to determine whether such delays are promoted by reduced class II protein signaling. Though MV may not directly target antigen presentation by class II proteins, its probable effects on cytokines that regulate class II protein transcription and membrane trafficking may ultimately affect surface deposition of class II complexes on infected human monocytes. We are equally eager to determine whether obstructing the IFN pathway helps MV induce and sustain viremia. FIG. 1. : PBMC from 10 human donors were infected with Edm at an MOI of 1, and 48 h later, cell surface class II (A) or class I (B) complexes were measured by flow cytometric analysis. PBMC from 10 human donors were infected with Edm at an MOI of 1, and 48 h later, cell surface class II (A) or class I (B) complexes were measured by flow cytometric analysis. Mean fluorescence intensity (MFI) values from three independent experiments are shown. I-shaped bars indicate standard deviations of values for mock- and Edm-infected cells (P < 0.01). Class II protein levels in mock-infected and Edm-infected cells were statistically indistinguishable. After infection, significant differences in class I protein levels were measured for donors B, C, D, F, G, and J. Ten thousand total events were collected. FIG. 2. : Purified monocytes from donors C and F were infected with Edm at an MOI of 1, and 48 h later, surface class I and class II complexes were measured by flow cytometric analysis. Purified monocytes from donors C and F were infected with Edm at an MOI of 1, and 48 h later, surface class I and class II complexes were measured by flow cytometric analysis. Mean fluorescence intensity (MFI) values are shown for representative experiments done in duplicate on the same day. I bars indicate standard deviations of values for mock- and Edm-infected cells (P < 0.01). Ten thousand total events were collected. FIG. 3. : PM (A) or PBMC (B) from donors E and F were either mock infected or infected with Edm at an MOI of 1 for 48 h. PM (A) or PBMC (B) from donors E and F were either mock infected or infected with Edm at an MOI of 1 for 48 h. At the time of inoculation, cells were treated or not treated with 1,000 U of IFN-gamma/ml. Cells were analyzed for surface expression of class I or class II proteins by flow cytometry, and mean fluorescence intensity (MFI) values are reported. Ten thousand total events were collected. FIG. 4. : PM from two donors were either treated or not treated with UV-inactivated Edm (UV-Edm) at an MOI of 1 for 48 h, and some cells were simultaneously treated with IFN-gamma. PM from two donors were either treated or not treated with UV-inactivated Edm (UV-Edm) at an MOI of 1 for 48 h, and some cells were simultaneously treated with IFN-gamma. For cells from each donor, the bars represent values for cells that were untreated (first bar) or treated with 1,000 U of IFN-gamma/ml (second bar), UV-Edm (third bar), UV-Edm with 1,000 U of IFN-gamma/ml (fourth bar), UV-Edm with 1,000 U of IFN-gamma/ml and 1,000 U of anti-IFN-alpha/beta/ml (fifth bar), or UV-Edm with 1,000 U of IFN-gamma/ml and 1,000 U of neutralizing anti-IgM/ml (sixth bar). Surface class II proteins were stained with anti-HLA-DR, and mean fluorescence intensity (MFI) values are shown, with I bars indicating standard deviations. Ten thousand total events were collected. FIG. 5. : PM from two donors were either untreated (first bar) or treated for 48 h with 1,000 U of IFN-gamma/ml (second bar), 200 U of IFN-alpha/beta/ ml (third bar), 1,000 U of IFN-gamma/ml and 200 U of IFN-alpha/beta/ml (fourth bar), or 1,000 U of IFN-gamma/ml, 200 U of IFN-alpha/beta/ml, and 1,000 U of neutralizing anti-IFN-alpha/beta per ml (fifth bar). PM from two donors were either untreated (first bar) or treated for 48 h with 1,000 U of IFN-gamma/ml (second bar), 200 U of IFN-alpha/beta/ ml (third bar), 1,000 U of IFN-gamma/ml and 200 U of IFN-alpha/beta/ml (fourth bar), or 1,000 U of IFN-gamma/ml, 200 U of IFN-alpha/beta/ml, and 1,000 U of neutralizing anti-IFN-alpha/beta per ml (fifth bar). Surface class II proteins were stained with anti-HLA-DR, and class I proteins were stained with an antibody that recognizes HLA-A, -B, and -C proteins. FIG. 6. : The human melanoma cell line MelJuSo and the monocytic cell line THP-1 were either mock infected with Vero cell lysate or infected with Edm at an MOI of 1, and 48 h later, cell surface expression of class I (anti-HLA-A, -B, and -C) or class II (anti-HLA-DR) protein was measured by flow cytometric analysis. The human melanoma cell line MelJuSo and the monocytic cell line THP-1 were either mock infected with Vero cell lysate or infected with Edm at an MOI of 1, and 48 h later, cell surface expression of class I (anti-HLA-A, -B, and -C) or class II (anti-HLA-DR) protein was measured by flow cytometric analysis. Mean fluorescence intensity (MFI) values from three independent experiments are shown. MV-induced elevation of class I protein on MelJuSo and THP-1 cells was statistically significant (P < 0.01). Ten thousand total events were collected. FIG. 7. : PM from donor F were either mock infected or infected with Edm at an MOI of 1 for 48 h. PM from donor F were either mock infected or infected with Edm at an MOI of 1 for 48 h. The media were removed and replaced with supernatants from CD14-depleted PBMC which had been in culture for 48 h. After 48 h, the cells were scored for class II or class I proteins by flow cytometry, and mean fluorescence intensity (MFI) values are shown, with I bars indicating standard deviations. Ten thousand total events were collected. TABLE 1 : Levels of IFN-alpha and IFN-gamma in supernatants after infection with MV TABLE 2 : Surface MHC class II molecules analyzed by a dose-response assay Backmatter: PMID- 12915554 TI - Link between Genome Packaging and Rate of Budding for Rous Sarcoma Virus AB - The subcellular location at which genomic RNA is packaged by Gag proteins during retrovirus assembly remains unknown. Since the membrane-binding (M) domain is most critical for targeting Gag to the plasma membrane, changes to this determinant might alter the path taken through the cell and reduce the efficiency of genome packaging. In this report, a Rous sarcoma virus (RSV) mutant having two acidic-to-basic substitutions in the M domain is described. This mutant, designated Super M, produced particles much faster than the wild type, but the mutant virions were noninfectious and contained only 1/10 the amount of genomic RNA found in wild-type particles. To identify the cause(s) of these defects, we considered data that suggest that RSV Gag traffics through the nucleus to package the viral genome. Although inhibition of the CRM-1 pathway of nuclear export caused the accumulation of wild-type Gag in the nucleus, nuclear accumulation did not occur with Super M. The importance of the nucleocapsid (NC) domain in membrane targeting was also determined, and, importantly, deletion of the NC sequence prevented plasma membrane localization by wild-type Gag but not by Super M Gag. Based on these results, we reasoned that the enhanced membrane-targeting properties of Super M inhibit genome packaging. Consistent with this interpretation, substitutions that reestablished the wild-type number of basic and acidic residues in the Super M Gag M domain reduced the budding efficiency and restored genome packaging and infectivity. Therefore, these data suggest that Gag targeting and genome packaging are normally linked to ensure that RSV particles contain viral RNA. Keywords: Introduction : Gag polyproteins form retrovirus-like particles that bud from the plasma membrane. Gag is synthesized by soluble ribosomes and is targeted to the site of budding by the membrane-binding (M) domain located in the N-terminal region . In most retroviruses (e.g., human immunodeficiency virus type 1 [HIV-1]), the M domain contains both myristate and a cluster of basic residues, which form hydrophobic and electrostatic interactions with membrane phospholipids, respectively . In Rous sarcoma virus (RSV), an avian retrovirus, the M domain maps to the first 86 residues in Gag . Although it is not myristylated, the RSV M domain contains 11 basic residues, and according to the nuclear magnetic resonance structure of the M domain, these basic residues are present on the surface of the molecule . Because substitutions of two or more basic residues in the M domain prevent targeting and budding, RSV Gag may bind to the plasma membrane by forming electrostatic interactions with acidic phospholipids . In addition to budding, Gag is responsible for packaging two copies of the retroviral genome. Efficient packaging requires the presence of basic residues and one or two copies of a zinc finger motif (Cys-X2-Cys-X2-His-X4-Cys) in the nucleocapsid (NC) domain of Gag . These elements enable specific interactions between Gag and the packaging signal (Psi), which is located in the 5' end of the viral RNA (vRNA) . Aside from the interaction between NC and Psi, many aspects of genome packaging remain unknown. Of particular interest is the subcellular location(s) at which Gag first binds vRNA. Recently, RNA targeting motifs known as A2 response elements have been identified in the genomes of several retroviruses . In addition, the RNA-targeting protein Staufen has been implicated in genome packaging by HIV-1 Gag . Although the functional significance of these findings is unclear, these data suggest that RNA-targeting proteins might enhance the efficiency of genome packaging by transporting vRNA to the sites of Gag assembly. Given the importance of the M domain in Gag targeting, it is possible that changes in this domain would alter the pathway by which Gag trafficks through the cell, and such changes might also affect the efficiency of genome packaging. There is evidence for this possibility from RSV, where a twofold reduction in the efficiency of genome packaging is caused by adding the myristylated, nine-amino-acid Src M domain to the N terminus of Gag . The mutant, known as Myr1E, is also noninfectious, and the packaged vRNA is monomeric. Stimulated by the potential role of the M domain in genome packaging, we characterized an RSV mutant with two substitutions, E25K and E70K, in the M domain. Although the acidic-to-basic substitutions enhanced budding by this mutant, designated Super M, the extracellular particles were noninfectious and contained only 1/10 the amount of vRNA found in wild-type (WT) particles. To identify the cause(s) of these defects, we considered recent data that suggest that RSV Gag must transit the nucleus to package vRNA . In contrast to WT, Super M Gag did not accumulate in the nucleus when cells were treated with leptomycin B (LMB), an inhibitor of nuclear export. In addition, the importance of Gag-Gag interactions in RSV targeting was tested by deleting the NC domain and examining the subcellular distribution of mutant and WT proteins. Interestingly, deletion of the NC sequences prevented WT Gag from accumulating on the plasma membrane, but Super M Gag targeting was unaffected. Together, these results suggest that the Super M substitutions alter the targeting properties of RSV Gag. Because the altered targeting phenotypes of Super M are likely caused by an enhanced affinity for the plasma membrane, it was possible that packaging and infectivity would also be disrupted. To test this possibility, additional substitutions that reestablished the WT number of basic and acidic residues were introduced to the Super M Gag domain, and these changes reduced the budding rate and restored genome packaging and infectivity. Hence, these observations suggest a relationship between genome packaging and membrane targeting by RSV Gag. MATERIALS AND METHODS : Proviral expression vectors and mutagenesis. | The WT gag gene used in these studies was obtained from the Prague C strain of RSV . To create pRS.V8-EGFP, gag was cloned via the unique SstI and HpaI sites into pRCAS-EGFP, a proviral plasmid that expresses green fluorescent protein (GFP) from the nonessential v-src region of the genome . The substitutions present in Super M Gag and the charge-balanced mutant proviral plasmids were generated by oligonucleotide-directed mutagenesis using previously described methods . The sequences of the mutagenic oligonucleotides and the resulting substitutions (underlined) are as follows: for Super M, the E25K oligonucleotide sequence is 5'-CTAAGAAGAAAATAGGGGCCATG and the E70K oligonucleotide sequence is 5'-GAAATCGGGAAAGTTAAAAACC; for K6,18E-Super M, the K6E oligonucleotide sequence is 5'-GCCGTCATAGAGGTGATTTCGTCC and the K18E oligonucleotide sequence is 5'-CTATTGCGGGGAAACTAGTCCTTC; for K13,18E-Super M, the K13E oligonucleotide sequence is 5'-GTCCGCGTGTGAAACCTATTG and the K18E oligonucleotide has the sequence shown above. Analysis of particle release. | Duplicate cultures of QT6 (quail) cells were transfected with mutant or WT proviral plasmids by the calcium phosphate method. Both cultures were labeled with l-[35S]methionine (0.1 muCi/mul, >1,000 Ci/mmol) in methionine-free Dulbecco's modified Eagle medium (DMEM) at 16 h posttransfection. One set was labeled for only 5 min, while the other was labeled for 2.5 h. After the 5-min labeling, the media were discarded, and the cells were lysed in radioimmunoprecipitation assay buffer. For the 2.5-h labeling, media and cell lysate fractions were prepared. After proteins were immunoprecipitated with antiserum raised against purified RSV, the samples were denatured and separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The amounts of labeled Gag in the lysates of the cells labeled for 5 min and the amounts of capsid (CA) protein released into the media after the 2.5-h labeling were quantified by phosphorimagery. The budding efficiency was calculated by dividing the amount of CA released after 2.5 h by the amount of Gag in the 5-min lysate. The data are expressed as percentages of the WT efficiency of release, which was set to 100%. Pulse-chase analyses were performed by transfecting identical cultures of QT6 cells with WT or mutant proviral (pRS.V8-EGFP) plasmids. After 16 h, the cells were washed and labeled for 10 min with l-[35S]methionine (0.1 muCi/mul, >1,000 Ci/mmol) in methionine-free DMEM. At the end of the pulse (time zero), one culture from each set was placed on ice, and media and cell lysates were prepared with radioimmunoprecipitation assay buffer. For the remaining cultures, the labeling media were removed and serum-free DMEM containing an excess of unlabeled methionine was added. At various times, one culture from each set was placed on ice, and media and cell lysates were prepared. After full-length Gag and the products of Gag proteolysis were immunoprecipitated, the labeled proteins were detected and quantified by phosphorimagery. To determine the rate of Gag disappearance from the cell, the amount present at each time point was divided by the amount detected immediately after the pulse. To determine the rate of Gag accumulation in the medium, the amount of CA present at each time point was quantified, and this number was multiplied by the ratio of the number of methionines in full-length Gag to the number in CA (correction factor = 3). The extrapolated value was then normalized by the amount of Gag detected in the cell lysates immediately after the pulse. Infectivity assays. | QT6 cells were transfected with mutant or WT proviral DNA. After transfection for 16 h, the cells were washed and fresh media were added. Virus-containing media were collected 48 h later, and the cellular debris was pelleted by centrifugation for 10 min at 2,000 x g. The cell-free supernatants were transferred to new tubes, and reverse transcriptase (RT) assays were performed to determine the concentration of virus. Turkey embryo fibroblasts (TEFs) were then inoculated with equal numbers of particles. After 24 h, the media were changed, and the cells were passaged 1:3 every 3 to 4 days so that infectivity could be determined at days 3, 7, and 14 postinfection. To measure infectivity, the cultures were trypsinized and stored on ice. Approximately 10,000 cells were then examined for the presence of GFP fluorescence with a FACScan (BD Biosciences). As mentioned above, infected cells expressed GFP from the nonessential v-src region of the viral genome. Hence, the percentage of infected cells was determined by dividing the number of green fluorescent cells by the total number examined. To determine whether infectivity by K6,18E-Super M or K13,18E-Super M resulted from unintended suppressor mutations, virus-containing media were collected from the TEFs at 2 weeks postinfection, the cellular debris was removed by centrifugation (10 min at 2,000 x g), and aliquots of the media were used to determine the virus concentrations by RT assay. Equal numbers of WT and mutant virus were added to fresh, uninfected TEFs, and the percentage of infected (GFP-fluorescent) cells was determined at 1 week postinfection. At this time point, the number of cells infected with the charge-balanced viruses would be similar to the number infected by the WT if the mutant viruses had made additional, compensatory changes (e.g., reversion of the Super M substitutions) during the primary infection. However, cultures inoculated with the charge-balanced mutants contained one-fourth to one-third the number of fluorescent cells found in WT-infected cultures (data not shown). Analysis of physical properties and protein composition of extracellular virus. | For electron microscopy, QT6 cells were seeded on 60-mm-diameter Permanox dishes and transfected with WT or Super M proviral DNA. After 16 h, the cultures were washed with 0.1 M sodium cacodylate (pH 7.4) and fixed in 4% paraformaldehyde-0.5% glutaraldehyde for 1 h at 4C. Fixed cells were then incubated for 30 min with 1% osmium tetroxide, dehydrated, and embedded in Epon 812. Thin sections were stained with uranyl acetate and lead citrate, and extracellular particles were detected by electron microscopy. To determine the density of Super M virus, particle-containing media were collected after labeling transfected QT6 cells for 2.5 h with [35S]methionine. Media were then centrifuged (10 min at 2,000 x g) to remove cellular debris and layered onto a 10 to 50% gradient of sucrose suspended in phosphate-buffered saline (PBS). As an internal control, labeled particles produced by pGag-GFP were added to the same gradient. The samples were then centrifuged at 83,500 x g for 16 h at 4C. Sixteen fractions of 0.7 ml each were collected from the bottom of the gradient, and the labeled Gag proteins were immunoprecipitated, resolved by SDS-PAGE, and quantified by phosphorimagery. The amount of WT or mutant protein in each fraction is expressed as a percentage of the total detected in the entire gradient. The density of each fraction was determined by examining a 50-mul aliquot by refractometry. The levels of Env and Pol incorporated into Super M particles were examined by Western blotting. Particles were collected from transfected QT6 cells, and RT assays were performed using exogenous templates (poly[A]) and primers (oligo[dT]). Virus samples corresponding to equal amounts of RT activity were denatured, and the proteins were resolved by SDS-PAGE and blotted to nitrocellulose membranes. To detect the amount of CA in each sample, the membranes were incubated with rabbit polyclonal antisera specific for RSV Gag and the products of Gag proteolysis (anti-Gag). To detect the levels of Env incorporated, membranes were incubated with rabbit antisera raised against the TM subunit of RSV Env (anti-TM). Primary antibody binding was visualized by enhanced chemiluminescence after incubation of the membranes with goat anti-rabbit antibodies conjugated to horseradish peroxidase. Analysis of genome packaging. | QT6 cells were transfected with mutant or WT proviral DNA. After 16 h, the transfection media were removed, the cells were washed, and fresh media were added. Forty-eight hours later, virus-containing media were collected and the cellular debris was removed as described above. To concentrate the extracellular virus, the cell-free media were centrifuged at 126,000 x g for 40 min at 4C over a solution of 20% sucrose in PBS. The virus pellet was then resuspended in PBS, and a portion was used to determine the concentration of virus by RT assay. Particle-associated RNA was then prepared with the TRI reagent (Sigma) according to the supplier's protocol and linear acrylamide (Ambion; 10 mug/ml) as a coprecipitant. After resuspension in diethyl pyrocarbonate (DEPC)-treated water, RNA sample volumes corresponding to equal numbers of particles were denatured and blotted to nylon membranes with a slot blot apparatus. The amount of genomic RNA packaged by each virus was detected by hybridization with 32P-labeled antisense riboprobes specific for the gag gene, as previously described . The amounts of labeled RNA bound to each sample were quantified by phosphorimagery. In a similar manner, the steady-state levels of genomic RNA present in both WT and Super M-producing cells were examined. The levels of genome packaging by the WT and Super M were also examined by RNase protection assay (RPA) and Northern blotting (data not shown). For the Northern blot analyses, particle-associated RNA was purified as described above. RNA sample volumes corresponding to equal numbers of virus were denatured, resolved on 2.2 M formaldehyde-1% agarose gels in MOPS (morpholinepropanesulfonic acid) electrophoresis buffer, and blotted to nylon membranes by standard techniques. Hybridizations with a 32P-labeled gag-specific riboprobe were performed as described above. For analysis by RPA, RNA from equivalent numbers of Super M and WT particles was denatured and hybridized with a 32P-labeled antisense gag riboprobe (532 nucleotides [nt]) and the RPA III kit (Ambion). Protected fragments (502 nt) were separated by electrophoresis on 5% acrylamide-8 M urea gels and quantified by phosphorimagery. To determine whether Super M contained vRNA dimers, QT6 cells were transfected with WT or mutant proviral DNA and the extracellular particles were collected from the media by ultracentrifugation. RNA from equivalent numbers of particles was purified and resolved on agarose gels under native conditions as described by Fu and Rein . After electrophoresis, the RNA was denatured by incubating the gel in 6% formaldehyde for 30 min at 65C. After several rinses in DEPC-treated water, the gel was soaked in 20x SSC (1x SSC is 0.15 M NaCl plus 0.15 M sodium citrate), and the RNA was blotted to nylon membranes. Blotted RNA was hybridized with the 32P-labeled gag antisense riboprobe and visualized by autoradiography (data not shown). Packaging of heterologous, Psi-containing RNA. | The 160-nt RSV minimal packaging sequence, MPsi, was defined by Banks et al. . To create pCMV.MPsi.gfp, we used PCR to amplify the DNA sequence between nt 156 and 315, which encodes MPsi in the RSV Prague C proviral plasmid pATV-8. The sequences of the primers used for PCR were as follows: sense, 5'-AATAGATCTGATCCTGCCCTCATCC; antisense, 5'-TTAAGATCTGCGGCCGCCGTCTTCCAACG. These primers created BglII sites (underlined) that flank the MPsi sequence, so the PCR product was digested and inserted into the BglII site (nt 610) of pEGFP-N2 (Clontech), which expresses GFP from the cytomegalovirus immediate-early promoter. Positive clones containing a single insert in the proper orientation were identified by restriction digestion and confirmed by DNA sequencing. To create the WT gag-only expression vector, the Prague C gag sequences from the proviral plasmid pRC.V8 were isolated by digestion with SstI and HpaI. This fragment was inserted into pEGFP-N2 after the gfp sequences were removed by digestion with SstI and NotI. In this case, the NotI end of the vector DNA was treated with the Klenow fragment of E. coli DNA polymerase I for ligation to the blunt HpaI end of the gag fragment. The resulting plasmid, designated pCMV.GagPR, was transfected into QT6 cells and found to synthesize full-length Gag proteins that were released into the medium and processed by protease (PR) (data not shown). To create pCMV.Super M.GagPR, the SstI/BspEI fragment (453 bp) encoding the Super M changes was exchanged with the corresponding WT sequences in pCMV.GagPR and the resulting clones were screened by DNA sequencing. The abilities of WT and Super M Gag to package heterologous mRNA were determined as follows. First, both pCMV.MPsi.gfp and pEGFP-N2, designated MPsi(-).gfp, were cotransfected into QT6 cells with either the WT or Super M Gag expression vector. All plasmids were transfected at equimolar amounts, and the total amount of DNA per transfection was 15 mug/35-mm-diameter plate. After 12 h, the transfection media were removed, and the monolayers were washed. The cultures were then trace labeled with [35S]methionine (0.1 muCi/mul, >1,000 Ci/mmol) in complete DMEM containing 1% fetal bovine serum. After approximately 20 h of labeling, the media were collected, the cellular debris was removed by low-speed centrifugation, and a portion of each supernatant was used to determine the concentrations of radiolabeled particles by immunoprecipitation, SDS-PAGE, and phosphorimager analyses. Particles were then pelleted from the remaining media, and the particle-associated RNA was purified with the TRI reagent (Sigma) and linear acrylamide as a coprecipitant. The cell-associated RNA was also purified from the monolayers in a similar manner. Upon resuspension of the RNA in DEPC-treated water, equal volumes were mixed with a 32P-labeled antisense riboprobe (306 nt) specific for the 5' ends of MPsi.gfp and MPsi(-).gfp mRNA. After hybridization for 16 h at 42C, RNase digestions were performed with the RPA III kit (Ambion). Protected fragments were resolved on 4% acrylamide-8 M urea gels and quantified by phosphorimagery. Because the labeled riboprobe contained both MPsi and gfp sequences, the presence of the MPsi.gfp mRNA yielded a protected band of 269 nt, whereas MPsi(-).gfp mRNA yielded a protected band of 204 nt. To calculate the packaging efficiency for each RNA species, the amount detected in the medium was divided by the steady-state level detected in the cell. This value was then normalized by the number of radiolabeled particles present in the medium. In each experiment, the efficiency of MPsi.gfp packaging by WT Gag was set to 100%. To determine the abilities of WT and Super M Gag to select for the MPsi-containing mRNA in this assay, the efficiency of MPsi.gfp packaging was divided by the efficiency of MPsi(-).gfp packaging. Hence, this number represents the factor of enrichment for Psi-containing mRNA in WT and Super M Gag particles. To compare the abilities of wild-type and Super M Gag to package nonspecific RNA, QT6 cultures were cotransfected with pCMV.MPsi(-).gfp and either pCMV.GagPR or pCMV.Super M.GagPR vectors. After 12 h, the cells were trace labeled with [35S]methionine and particle-associated RNA was purified as described above. Nonspecific packaging of MPsi(-).gfp mRNA was quantified by RPA and normalized to the amount of radiolabeled Gag present in the medium. For comparison, the level of MPsi(-).gfp packaging by the WT was set to 100%. Subcellular targeting properties of Super M. | Construction of the pGag-GFP expression plasmid was described previously . To create pDeltaNC-GFP, the region between nt 1044 and 1865 in gag was amplified by PCR. The primer sequences were as follows: sense, 5'-GGGCAAGGGTCAGGG; antisense, 5'-GTTTGGGGCCCCCTCTCTCTATTGAC. To prepare the PCR product for cloning, we made use of the endogenous EspI site at nt 1419 in gag and the ApaI site engineered by the antisense PCR primer (underlined). The digested fragment was then exchanged with the EspI/ApaI fragment of pGag-GFP. The resulting sequences encode a chimera in which GFP is fused to the C terminus of the seventh residue of the NC domain in Gag. Hence, the remaining C-terminal residues of Gag, including the I domains and the zinc fingers, are absent in DeltaNC-GFP. To create pSuperM-GFP and pSuperM-DeltaNC-GFP, the SstI/BspEI fragment encoding the Super M Gag substitutions was exchanged with the corresponding WT sequences in pGag-GFP and pDeltaNC-GFP, respectively. The subcellular distributions of the GFP-tagged chimeras in transfected QT6 cells were examined by confocal microscopy as previously described . To determine the effects of LMB treatment on WT and mutant Gag-GFP localization, transfected QT6 cells were treated with the drug at a final concentration of 10 ng/ml for 3 h before examination by confocal microscopy . RESULTS : In a previous report, the 11 basic residues in the RSV Gag M domain were shown to be essential for Gag targeting to the plasma membrane . To further investigate the possibility that Gag binds the plasma membrane by forming electrostatic interactions with acidic phospholipids, additional basic residues were introduced into the M domain, and these changes were shown to enhance budding from mammalian (COS-1) cells. The most dramatic increase in budding was observed with a mutant, designated Super M, whose Gag contains two substitutions, E25K and E70K . These substitutions simultaneously increased the number of basic residues in the M domain from 11 to 13 and decreased the number of acidic residues from 6 to 4. Hence, the Super M substitutions cause a net shift of +4 in the M domain. Analysis of Super M virus budding and infectivity. | To further our analysis of Super M, the mutant sequence was subcloned into pRS.V8-EGFP, a proviral plasmid derived from pRCAS-EGFP . Metabolic labeling of transfected QT6 cells subsequently showed that Super M produced three times more virus than the WT control, whereas the level of the previously described Src chimera, Myr1E, was enhanced by about twofold . The enhanced rate of release of these two mutants was confirmed by pulse-chase analyses. As shown in Fig. , the pulse-labeled mutants disappeared from the cells more quickly than the WT, with the levels of Super M and Myr1E being reduced to less than one-half of WT levels after chasing for 30 min. Concurrent with their increased rate of disappearance from the cell, Super M and Myr1E proteins accumulated more rapidly in the media . Therefore, Super M and Myr1E bud much faster than WT virus. Interestingly, 50 to 60% of the amount of wild-type Gag synthesized during the pulse was recovered in the cell and media fractions at the end of the chase (120 min), whereas approximately 80% of Super M Gag was accounted for. Given the correlation between increased recovery and enhanced budding, it was possible that rapid escape from the cytoplasm reduced Gag loss due to premature processing by the viral PR or degradation by cellular machinery. To determine whether PR affected recovery, PR- versions of the WT and Super M were expressed from gag-only vectors and analyzed as shown in Fig. . In this case, approximately 90% of the pulse-labeled WT and Super M proteins were recovered at the final chase point (data not shown). Thus, the difference in recovery between PR+ versions of the WT and Super M is likely due to differences in the relative budding rates rather than a gross change in protein stability. As mentioned above, Myr1E was previously shown to be noninfectious . Since Super M and Myr1E exhibit similar, enhanced budding rates, we considered the possibility that Super M would be noninfectious as well. To test this, WT and mutant viruses were collected from transfected QT6 cells. Media containing equal numbers of particles were then transferred to primary cultures of TEFs. The kinetics of virus spreading were determined by examining the cultures at days 3, 7, and 14 postinfection. Infected cells were detected by fluorescence-activated cell sorting (FACS) because the proviruses expressed gfp from the nonessential, v-src region of the genome . To determine the sensitivity of this method, serial twofold dilutions of infected (GFP-positive) cells were made with uninfected (GFP-negative) cells. Subsequent analyses revealed that FACS reliably detected GFP-positive cells in a linear range from 1 to 100% of the total population examined (10,000 cells) (data not shown). Despite the apparent sensitivity of this assay, fluorescent cells were not detected in the TEF cultures incubated with Super M virus . Moreover, no infectivity was detected when 104-fold more Super M virus (relative to the WT control) was incubated with the TEF cultures (data not shown). Therefore, Super M is noninfectious. Physical properties and composition of the Super M virus. | To uncover the cause(s) of the Super M infectivity defect, several features of the mutant virus were examined. Thin sections of Super M-producing cells examined by electron microscopy revealed particles of normal size with electron-dense cores like those of mature, WT virus . Analysis of particles by isopycnic sucrose gradient centrifugation showed that the density of Super M virus was like that of WT particles, with the peak of each population banding at a density of 1.16 g/ml in sucrose . To determine the levels of Env and Pol incorporation, particles were collected from transfected QT6 cells and analyzed by RT assays using exogenous templates and primers. Virus sample volumes corresponding to equal amounts of RT activity were denatured, separated by SDS-PAGE, and probed by Western blotting. With a primary antibody specific for RSV Gag and Gag-derived proteins (anti-Gag), equal amounts of CA (27 kDa) were detected in the WT and Super M lanes . Since these samples were first normalized by RT activity, this result indicates that the ratios of Gag to Pol for WT and Super M viruses are similar. Likewise, an antibody specific for the TM subunit of Env (anti-TM) yielded bands of equal intensities when the WT and Super M samples were normalized by RT activity. Here, however, the TM bands migrated faster than the published mobility of 37 kDa, a difference probably due to fewer glycosylation sites in the TM subunit of this particular strain of RSV . In contrast to the levels of Env and Pol, the amounts of genomic RNA packaged by Super M were significantly reduced. This defect was first identified by slot blot analyses using RNA purified from equal numbers of WT and mutant particles . Specifically, Super M contained only 10% of the amount of vRNA found in WT virus. Northern blotting and RPAs confirmed the severity of this defect (data not shown). To determine whether the Super M packaging deficiency was due to a paucity of vRNA inside the cell, perhaps as a result of vRNA instability due to the Super M mutations, RNA was also purified from the transfected monolayers. As shown in Fig. , however, the Super M-expressing cells contained normal amounts of genomic RNA. Therefore, the Super M packaging defect is not due to a gross reduction in the level of intracellular vRNA. Although the mutations that cause the two acidic-to-basic substitutions in Super M lie downstream of the packaging signal, it was possible that these mutations somehow altered folding in the 5' end of the vRNA and disrupted stem-loop structures that are critical for Psi function. Alternatively, the Super M changes might have altered folding in the NC domain and prevented Gag from recognizing the Psi sequence. To test these possibilities, a plasmid expressing MPsi.gfp, an RNA consisting of the 160-nt RSV minimal packaging sequence, MPsi, at the 5' end of a nonviral (gfp) mRNA, was generated . With this vector, designated pCMV.MPsi.gfp, the abilities of WT and mutant Gag proteins (expressed from pCMV.GagPR, a gag-only expression vector) to package identical, MPsi-containing mRNA were compared. To determine the specificity of packaging in this system, the pCMV.MPsi.gfp plasmid was cotransfected with pCMV.MPsi(-).gfp, a vector that expresses a gfp mRNA lacking the MPsi sequence. Under these conditions, Super M Gag packaged about one-fourth the amount of MPsi.gfp found in WT virus-like particles (VLPs) . Thus, the genome packaging defect seen in provirus-transfected cells is likely due to the amino acid substitutions in the Super M protein rather than to the mutations made in the genomic RNA. With MPsi.gfp, however, the Super M defect was less pronounced (i.e., genome packaging was 1/10 WT levels, whereas MPsi.gfp packaging was 1/4 WT levels). A possible explanation for this difference is that the number of MPsi.gfp mRNAs (produced by the cytomegalovirus immediate-early promoter) may have exceeded the number of viral genomes that Gag normally encounters in an infected cell. Nevertheless, packaging of MPsi.gfp mRNA by WT and Super M Gag was apparently specific because incorporation of the MPsi- control was only 1/5 to 1/10 as efficient. This result reveals that, although the absolute amounts differed, the ratios of MPsi to MPsi(-).gfp mRNA in WT and Super M VLPs were similar . Thus, the ability to select for MPsi-containing RNA suggests that the Super M substitutions do not weaken NC's specificity for vRNA. Instead, the Super M domain may prevent packaging by an indirect mechanism, perhaps by enabling Gag to leave the site of genome packaging more quickly. Given recent data suggesting that RNA is required to serve a structural role during retrovirus assembly , the ability of Super M to incorporate nonspecific, cellular RNA was investigated. In this case, either WT or Super M gag-only (pCMV.GagPR) expression vectors were cotransfected into QT6 cells with pCMV.MPsi(-).gfp, which makes mRNA that can only be packaged nonspecifically. The amounts of MPsi(-).gfp mRNA in the cell and medium fractions were determined by RPA, and the packaging efficiency was calculated by normalizing to the amount of Gag released. Under these conditions, Super M Gag packaged about one-half the amount of MPsi(-).gfp mRNA found in WT VLPs . Hence, this result suggests that Super M is able to "compensate" for the vRNA packaging defect by incorporating nonspecific, cellular RNA. However, because Super M Gag exhibited a twofold reduction in the relative level of MPsi(-).gfp incorporation, it seems likely that the rapid rate of budding inhibits nonspecific RNA packaging, albeit to a much lesser degree than vRNA packaging. Despite the reduced amount of RNA incorporation, Super M virus appeared normal by electron microscopy and exhibited WT density in sucrose . Therefore, the level of total RNA incorporation is apparently sufficient to enable Gag-Gag interactions and efficient release. Previous studies of Myr1E provide a precedent for the finding that M domain changes can reduce genome packaging in RSV. However, packaging in Myr1E is reduced by only twofold. More strikingly, the vRNA packaged by this mutant is monomeric. To determine whether Super M exhibits a dimerization defect, vRNA was purified and resolved on agarose gels under native conditions. Although the genomes of WT virus were clearly dimeric and those from Myr1E migrated as monomers, we were unable to resolve a clear band of vRNA from the Super M samples (data not shown). Efforts to increase the signal by loading purified RNA from 10-fold-more Super M particles resulted only in higher backgrounds. One possible explanation for these results is that the amount of vRNA packaged by Super M is below the level of detection for this particular assay. Targeting properties of Super M Gag. | Previous studies of HIV-1 Gag have identified M domain changes that enhance membrane binding but do not stimulate the release of extracellular particles . Instead, these mutants bind membranes indiscriminately, and, in some instances, particles have been detected in the endoplasmic reticulum or in intracellular vesicles. To examine the membrane specificity of Super M, GFP was fused in place of a C-terminal region of Gag that is nonessential for budding . Confocal microscopy revealed that Super M Gag-GFP, like an analogous construct bearing the WT M domain, was found at the plasma membrane under steady-state conditions . In addition, Super M Gag-GFP did not accumulate at intracellular membranes, which is consistent with the enhanced rate of extracellular virus production by this mutant. Recent data suggest that at least some RSV Gag passes through the nucleus . Since nuclear trafficking is inhibited by the Src sequence present on the Myr1E M domain, we asked whether Super M is similarly deficient by treating pGag-GFP- and pSuperM-GFP-transfected cells with LMB, an inhibitor of the CRM-1 pathway of nuclear export . As expected, LMB treatment trapped WT Gag-GFP in the nucleus . However, Super M Gag-GFP was found at the plasma membrane in LMB-treated cells and did not appear in the nucleus. Although the functional significance of nuclear trafficking by WT Gag is unclear, it is possible that the acidic-to-basic substitutions in the Super M domain prevent nuclear targeting by enhancing Gag's affinity for the plasma membrane. This interpretation is supported by the role of NC in Super M targeting. Although NC is best known for packaging vRNA during assembly, studies of HIV-1 Gag have revealed that NC is also required for efficient plasma membrane targeting . This is because NC contains the I domains, which enable Gag to bind RNA and use it as a scaffold for assembly . Since RNA binding promotes interactions between Gag proteins, the requirement for the I domains in membrane targeting suggests that Gag cannot bind to the plasma membrane as a monomer. Upon oligomerization, however, Gag proteins form a multivalent membrane-binding complex that can stably bind membranes with high avidity. Prompted by the importance of NC and the I domains in membrane targeting by HIV-1 Gag, we investigated the role of NC in the subcellular distribution of RSV Gag by deleting the majority of the NC sequence from our GFP-tagged chimera . As shown in Fig. , this mutant adopted a diffuse pattern in the cytoplasm of transfected cells. Thus, the NC domain is required for efficient plasma membrane localization by RSV Gag. In contrast, removal of the NC domain did not prevent plasma membrane targeting by Super M Gag-GFP . This result is significant because it suggests that Super M is not dependent on RNA interactions for targeting to the plasma membrane. Because Gag normally binds the viral genome in infected cells, Super M's ability to target in the absence of RNA interactions suggests that genome packaging is inhibited by Super M's enhanced membrane-binding potential. This interpretation implies that, by reducing Super M's membrane-binding potential, its dependence on RNA binding for plasma membrane targeting would be increased and genome packaging would be restored. Second-site suppressors of the Super M phenotype. | To determine whether Super M's genome packaging and infectivity defects are due to enhanced membrane targeting, pairs of lysine-to-glutamate substitutions were made at residues 6 and 18 or at residues 13 and 18, while maintaining the original E25K and E70K substitutions present in Super M Gag . Hence, although the positions of the charged residues were changed, these substitutions restored the balance of 11 basic and 6 acidic residues found in the WT M domain. Consequently, reestablishing the charge balance reduced the budding efficiency of proviral constructs to within 50% of WT levels . Moreover, these substitutions restored both the ability to trap mutant Gag-GFP proteins in the nuclei of LMB-treated cells and the dependence on NC for plasma membrane targeting . These results suggest that Super M's membrane-binding potential can be weakened by reducing the net positive charge in the M domain. As predicted, the charge-balancing substitutions also restored genome packaging and the ability to infect cultured cells . Although the slower replication kinetics exhibited by the charge-balanced mutants could be due to some defect in virus entry, K6,18E-SuperM and K13,18E-SuperM bud only half as well as the WT, so the relative rates of spreading by these mutants are slowed, at least in part, by their reduced budding efficiencies. The observed infectivity was not due to additional, unintended mutations (e.g., reversion) because mutant particles isolated at the end of these experiments (2 weeks postinfection) replicated with the same kinetics as those initially collected from the transfected QT6 cells (data not shown). Thus, the ability to affect genome packaging and infectivity by modulating the net charge in the M domain suggests that Gag targeting and genome packaging are linked in RSV. DISCUSSION : The data presented here reveal that small changes (i.e., two substitutions) in the M domain can have a profound impact on genome packaging and infectivity of RSV. Whether the infectivity block exhibited by Super M is due entirely to the genome packaging defect is unclear. Given the severity of the packaging defect, it is likely that Super M incorporates no more than one vRNA copy per particle. If this is the case, then an inability to form vRNA dimers would also block infectivity. Alternatively, if a few Super M particles contained vRNA dimers, then additional defects (e.g., tRNA primer incorporation) would remain to be identified. However, the correlation between the vRNA content and infectivity of the mutants described here suggests that the genome packaging deficiency is the major reason why Super M is noninfectious. Hence, these data raise the question of how the M domain affects genome packaging by RSV Gag. Although the severity of the Super M packaging defect is similar in magnitude to that caused by deleting both zinc fingers in NC (both mutants package only 1/10 the amount of vRNA found in WT particles), numerous studies of RSV Gag have shown that the MA region, including the M domain, is not involved in direct vRNA interactions . For instance, MA can be replaced by the Src M domain without affecting the levels of vRNA incorporation . Moreover, an RSV Gag chimera containing NC sequences from murine leukemia virus (MLV) packages the MLV genome. Hence, RNA specificity is not determined by the MA region in chimeric Gag proteins . This result is consistent with the observation that RSV MA does not bind vRNA with any specificity in vitro . So, if the M domain is not involved in vRNA binding, then how does it affect genome packaging? Given the well-established role of the M domain in Gag targeting, one possibility is that the M domain directs Gag to the site of genome packaging. Recent data suggest that the RSV M domain contains a nuclear localization signal (NLS), and because the absence of nuclear import by Myr1E Gag correlates with a genome packaging defect, it was proposed that genome binding occurs in the nucleus . However, the dispensability of the RSV M domain in vRNA packaging contradicts this hypothesis. Indeed, vRNA packaging is not affected when the MA region is replaced by the Src M domain, a change that would block nuclear trafficking by removing the putative NLS in the M domain . Likewise, deletions or single substitutions in the NLS in HIV-1 MA prevent nuclear import, but these changes do not affect vRNA incorporation . Moreover, retroviral vector systems in which Gag specifically packages mRNA that exists exclusively in the cytoplasm have been designed . In those studies, recombinant alphaviruses (e.g., Semliki Forest virus) or poxviruses (e.g., vaccinia virus) were used to synthesize transcripts in the cytoplasm of Gag-expressing cells, and the high titers achieved by this approach reveal that Gag can readily package cytoplasmic mRNA. Therefore, these data argue against a role for nuclear trafficking in genome packaging. Consequently, the Super M packaging defect may not be due to a lack of nuclear import per se, but the absence of a nuclear phase suggests that the Super M substitutions alter Gag targeting in some manner. This possibility may be significant because, if genome packaging occurs at a specific location in the cytoplasm, then perhaps the Super M substitutions prevent targeting to this site as well. Alternatively, genome packaging may not require Gag transport to a particular location in the cell but may depend on the accumulation of Gag proteins at the site of synthesis. In this case, packaging would be inhibited if Gag begins targeting from the site of synthesis before binding to the viral genome. Support for this possibility is provided by the role of the NC domain in Gag targeting . The importance of NC, and in particular the I domains, in both genome packaging and Gag targeting suggests that these two functions are normally linked. The basis of this connection seems to be the need for Gag to use RNA as a scaffold for assembly. Indeed, the I domains, in conjunction with the zinc fingers in NC, facilitate specific interactions with the Psi sequence in the viral genome. As additional Gag proteins attach nonspecifically to vRNA via the I domains, they assemble into a multivalent complex that can bind membranes with high avidity. Since the formation of this multivalent complex relies on RNA interactions, genome packaging may normally precede Gag targeting. Consequently, the requirement for RNA interactions in Gag targeting may ensure that the retroviral genome is packaged before nascent particles are released from the cell. Based on the model described above, the vRNA packaging defects exhibited by Super M and Myr1E may be due to the ability of these mutants to target to the plasma membrane before genome binding and Gag-Gag interactions occur. Support for this interpretation is provided by the enhanced budding properties of these mutants and the maintenance of Super M targeting in the absence of the NC domain. Although Gag-Gag interactions may be unnecessary for Super M targeting, this mutant nevertheless assembles efficiently because the extracellular particles resemble the WT in appearance and density . Importantly, reestablishing the WT balance of basic and acidic residues in the Super M Gag domain improved genome packaging and restored infectivity . Hence, these data provide further evidence that the genome packaging defect is caused by the enhanced membrane-binding potential of the Super M Gag domain. The idea that genome packaging precedes Gag targeting is not new. A well-known example of this mechanism is provided by the Mason-Pfizer monkey virus (MPMV). Unlike those of RSV and HIV-1, MPMV Gag proteins first form complete capsids in the cytoplasm, which then target to the plasma membrane for budding . A precedent for M domain changes alleviating the assembly prerequisite for Gag targeting is provided by the effects of substitution R55W in the MA domain of MPMV Gag . Specifically, this change prevents the accumulation of intracellular capsids and enables particle assembly at the plasma membrane. Interestingly, the mutant was also shown to bud faster than WT MPMV, so the R55W change may enhance membrane affinity. Like Super M, the R55W mutant is noninfectious, and further studies suggested that the infectivity defect is due to the incorporation of fewer Env glycoproteins. However, the efficiency of genome packaging was not examined in these studies and therefore may be defective as well. This prediction is based on the assumption that MPMV genomes are present at the intracellular sites of capsid assembly, though surprisingly the presence of vRNA in intracellular particles has not been reported. Despite the likelihood that the extra basic residues in the Super M domain enhance Gag's affinity for the plasma membrane, other factors could explain why this mutant buds so quickly. For instance, we have shown that Super M does not localize to the nucleus in LMB-treated cells. Hence, if WT Gag normally travels through the nucleus before targeting to the plasma membrane, then Super M may bud faster simply by avoiding the nuclear "detour." In addition, it is possible that the substitutions in Super M promote Gag-Gag interactions. That the MA region could be a site of direct, albeit weak, contact between Gag proteins is suggested by cross-linking studies, yeast two-hybrid data, and the observation that purified HIV-1 MA forms trimers upon crystallization . Thus, the Super M substitutions may enhance interactions between Gag proteins and enable the formation of a multivalent membrane-binding complex in the absence of the I domains and RNA binding. Certainly, additional studies will be needed to test these hypotheses, and the insights gained by these efforts should shed more light on the relationship between membrane targeting and genome packaging by Gag. FIG. 1. : Location of the Super M substitutions in RSV Gag. Location of the Super M substitutions in RSV Gag. (A) Diagram of the RSV proviral DNA used in these experiments (top) and the RSV Gag polyprotein (bottom). The 5' long terminal repeat (LTR) promotes transcription of genome-length mRNA, which can be encapsidated or used for the synthesis of Gag and Gag-Pol proteins. In addition, full-length vRNA can be spliced for the synthesis of Env glycoproteins and, in the recombinant virus used here, for the synthesis of GFP. Intact Gag polyproteins drive budding and are subsequently cleaved by the viral PR into the mature products: MA, CA, NC, etc. The locations of the domains required for budding are indicated below Gag. The M domain is essential for plasma membrane targeting. The I domains promote Gag-Gag interactions and the assembly of dense particles. The L domain is required for a late step in budding. (B) The secondary structure of the RSV M domain consists of five helices (rectangles) connected by flexible loops. The locations of the 11 basic and 6 acidic residues in the M domain are depicted according to their charge. The two acidic-to-basic substitutions in Super M Gag, E25K and E70K, are indicated. FIG. 2. : Super M budding and infectivity. Super M budding and infectivity. (A) QT6 (quail) cells were transfected with proviral (pRCAS-derived) plasmids and labeled for 2.5 h with [35S]methionine. Viral proteins were immunoprecipitated from the cell and media fractions, separated by SDS-PAGE, and quantified by phosphorimager analysis. The amounts of viral proteins in the media were normalized to the levels of intracellular gag expression. The budding efficiency of WT virus was set to 100% for comparison to Super M and the previously described RSV mutant Myr1E, which encodes the Src membrane-binding domain as an extension from the N terminus of Gag. The data are the averages of 10 (Super M) or 4 (Myr1E) experiments, and the error bars measure 1 standard deviation (SD) from the mean. Pulse-chase analyses of Gag levels in the cell (B) and medium (C) fractions were performed with QT6 cells transfected with proviral plasmids. The percentage of Gag in the cells at each time point was determined by dividing the amount remaining by the amount detected at the beginning of the chase. Similarly, the amount of Gag present after the pulse was used to normalize the amount of viral antigen detected in the medium at each time point. The data are the averages of four experiments, and the error bars measure 1 SD from the mean. (D) WT and mutant viruses were collected from transfected QT6 cells, normalized by RT assay, and transferred to cultures of TEFs. Infected cells were detected by FACS analysis because the proviruses expressed gfp from the long terminal repeat. The numbers of infected (fluorescent) cells are expressed as percentages of the total population examined at days 3, 7, and 14 postinfection. The results show the averages of six experiments, and the error bars measure 1 SD from the mean. FIG. 3. : Characterization of Super M particles. Characterization of Super M particles. (A) Electron micrographs of Super M (left) and WT (right) virus released from QT6 cells transfected with proviral plasmids. (Scale bars = 100 nm). (B) Viral particles were produced by metabolically labeled cells and separated by sucrose density gradient centrifugation. Gag proteins were immunoprecipitated from each gradient fraction, resolved by SDS-PAGE, and quantified by phosphorimager analysis. The amount of Gag in each fraction is expressed as a percentage of the total recovered from all of the fractions collected. WT virus was used as a control for normal density. (C) WT and Super M virions were collected from transfected cells, pelleted through 20% sucrose, and resuspended in PBS. Virus samples corresponding to equivalent amounts of RT activity were denatured, separated by SDS-PAGE, and transferred to nitrocellulose membranes for Western blot analyses. CA bands were detected with polyclonal Gag-reactive antiserum (alpha-Gag). Env incorporation was determined by using an antibody against the TM subunit (alpha-TM). FIG.4. : Genome packaging properties of Super M. Genome packaging properties of Super M. (A) QT6 cells were transfected with proviral plasmids, and particles were prepared as for Fig. . RNA was purified from equivalent numbers of cells and particles, bound to nylon membranes using a slot blot apparatus, and then probed with radiolabeled, antisense RNA specific for the viral genome. Labeled bands were quantified by phosphorimager analysis, and the amounts of vRNA detected in WT cell and virus samples were defined as 100%. To ensure that the RNAs were analyzed within the linear range of this assay, the samples in the right column of each blot contained 1/10 of the total RNA blotted in the left column. (B) Vectors expressing gfp mRNA with or without the RSV minimal packaging sequence (MPsi) were created. These were cotransfected into QT6 cells with WT or Super M gag-only expression vectors (pCMV.GagPR). The cultures were trace labeled with [35S]methionine in 1% serum-containing media. After the labeling, the media were collected and aliquots of each were used to determine the numbers of radiolabeled VLPs present. The remaining media were used for the purification of particle-associated RNA, and the cell-associated RNA was purified from the transfected monolayers. The amounts of MPsi.gfp and MPsi(-).gfp mRNA in both the cell and media fractions were determined by RPA and quantified by phosphorimager analysis. IP, immunoprecipitation. (C) A representative autoradiograph shows the amounts of MPsi.gfp and MPsi(-).gfp mRNA detected in the cell and media fractions by RPA. The specific gag-only expression vector, WT or Super M (SM), cotransfected is indicated above each lane. (D) The packaging efficiencies of MPsi and MPsi(-).gfp mRNA were calculated by dividing the amount detected in the medium by the amount present in the cell. This ratio was then normalized by the number of particles present in the medium. For comparison, the efficiency of MPsi.gfp packaging by WT Gag was set to 100%. The results are the averages of four experiments, and the error bars measure 1 standard deviation (SD) from the mean. In addition, the abilities of WT and Super M Gag proteins to select for Psi-containing mRNA were calculated by determining the ratio of MPsi to MPsi(-).gfp packaged into extracellular particles. (E) MPsi(-).gfp RNA was coexpressed with either WT or Super M Gag in QT6 cells. RNA was purified from cell and medium fractions, and the amounts of MPsi(-).gfp (arrowhead) present were determined by RPA (left). The relative abilities of WT and Super M Gag to package nonspecific RNA were calculated by normalizing the amount of MPsi(-).gfp RNA in the medium to the amount of Gag released (right). For comparison, the amount of MPsi(-).gfp packaging by WT Gag was set to 100%. The data are the averages of six experiments, and the error bar measures 1 SD from the mean. FIG. 5. : Subcellular localization of Super M Gag. Subcellular localization of Super M Gag. (A) Gag-GFP was created by replacing nonessential, C-terminal sequences of RSV Gag with GFP. The DeltaNC-GFP construct contains a large deletion encompassing the I domains in NC. Super M versions of Gag-GFP and DeltaNC-GFP were generated by replacing the WT M domain sequences with those with the E25K and E70K substitutions. (B) QT6 cells transfected with Super M or WT Gag-GFP vectors were left untreated or were treated with an inhibitor of nuclear export (LMB) prior to being examined by confocal microscopy. The untreated cells revealed the steady-state localization patterns of WT and Super M Gag-GFP and also served as controls for cells treated with LMB. The DeltaNC panels show the effects of deleting NC on the localization of Gag-GFP proteins with WT or Super M domains. FIG. 6. : Creation and analysis of suppressors of Super M. Creation and analysis of suppressors of Super M. (A) The WT numbers of positively and negatively charged residues were restored in the Super M domain by basic-to-acidic substitutions at positions 6 and 18 or at positions 13 and 18. (B) Analysis of the budding efficiencies of the charge-balanced mutants was performed with proviral vectors as for Fig. . (C) The localization patterns of DeltaNC-GFP and Gag-GFP proteins containing charge-balanced M domains in untreated or LMB-treated cells were examined as for Fig. . (D) The genomic RNA content of extracellular virus was determined by using proviral vectors as for Fig. . (E) Infectivity assays were performed at 1 and 2 weeks postinfection as for Fig. . For the results shown in panels B, D, and E, the data are the averages of at least three experiments and the error bars represent 1 standard deviation from the mean. Backmatter: PMID- 12915579 TI - Surface Downregulation of Major Histocompatibility Complex Class I, PE-CAM, and ICAM-1 following De Novo Infection of Endothelial Cells with Kaposi's Sarcoma-Associated Herpesvirus AB - Under selective pressure from host cytotoxic T lymphocytes, many viruses have evolved to downregulate major histocompatibility complex (MHC) class I and/or T-cell costimulatory molecules from the surface of infected cells. Kaposi's sarcoma-associated herpesvirus (KSHV) encodes two proteins, MIR-1 and MIR-2, that serve this function during lytic replication. In vivo, however, KSHV exists in a predominantly latent state, with less than 5% of infected cells expressing discernible lytic gene products. Thus, mechanisms of immune evasion that depend on genes expressed only during lytic replication are unlikely to be active in most KSHV-infected cells. As a result, we searched for evidence of similar defensive strategies extant during latency, employing culture systems that strongly favor latent KSHV infection. We measured cell surface levels of immunomodulatory proteins on both primary dermal microvascular endothelial cells (pDMVEC) infected through coculture with induced primary effusion lymphoma cells and telomerase-immortalized DMVEC infected directly with cell-free virus. Employing a panel of antibodies against several endothelial cell surface proteins, we show that de novo infection with KSHV leads to the downregulation of MHC class I, CD31 (PE-CAM), and CD54 (ICAM-I) but not CD58 (LFA-3) or CD95 (Fas). Furthermore, flow cytometry with a fluorescently labeled monoclonal antibody to the latency-associated nuclear antigen (LANA) revealed that downregulation occurred predominantly on KSHV-infected (LANA-positive) cells. Although the vast majority of infected cells displayed this downregulation, less than 1% expressed either immediate-early or late lytic proteins detectable by immunofluorescence. Together, these results suggest that downregulation of immunomodulatory proteins on the surface of target cells may represent a constitutive mode of immune evasion employed by KSHV following de novo infection. Keywords: Introduction : Herpesviruses constitute a family of human pathogens that persist indefinitely in the host organism and often contribute to a variety of disease states. Viral persistence depends on the ability of herpesviruses to establish a latent infection within cells. This process requires the successful evasion of host immune defenses, which usually includes the restricted expression of viral genes during latency. Although such restriction lowers their overall antigenic profile, herpesviruses must employ additional mechanisms to ensure prolonged protection from host immune responses. One such mechanism involves the downregulation of immunoregulatory proteins, such as major histocompatibility complex (MHC) class I, from the surface of infected cells. This serves to limit their recognition by cytotoxic T lymphocytes (CTLs) and dampen the inflammatory response to viral infection. Examples of human herpesvirus proteins that perform this function include the ICP47 protein of herpes simplex virus 1 , the interleukin-10 protein of Epstein-Barr virus , and the US2, US3, US6, and US11 proteins of human cytomegalovirus . Kaposi's sarcoma-associated herpesvirus (KSHV) is a recently discovered member of the gamma (lymphotropic) subfamily of herpesviruses and is responsible for several distinct diseases in humans, including Kaposi's sarcoma (KS), primary effusion lymphoma (PEL), and multicentric Castleman's disease . While pathologically diverse, these disorders are all associated with an immunocompromised state and pose a significant threat to human immunodeficiency virus-infected individuals and solid-organ transplant recipients worldwide. In contrast, KSHV infection in healthy patients is usually asymptomatic. In all infected individuals, KSHV elicits a humoral and cellular host immune response directed against both lytic and latent proteins of the virus , yet this response, even in healthy persons, is unable to eradicate KSHV from the body. This suggests that KSHV, like other human herpesviruses, possesses the ability to evade immune responses during infection. Recent studies have identified several KSHV lytic proteins that exert potential immunoregulatory roles during lytic replication. These include inhibition of apoptosis by viral Bcl-2 and open reading frame (ORF) K7 , complement deregulation by ORF 4 , Th2-type polarization by viral MIP-II , and inhibition of the interferon antiviral response by viral interferon response factor 1 (vIRF-1) , vIRF-3 , and viral interleukin-6 . Furthermore, KSHV also encodes two immediate-early proteins, MIR-1 (encoded by ORF K3) and MIR-2 (encoded by ORF K5), that downregulate immunoregulatory proteins such as MHC class I and ICAM-1 from the surface of cells to limit their recognition by immune cells (14, 15, 26-28, 61). KSHV, however, most often follows the general paradigm of gammaherpesvirus infection, namely a primarily latent phase of infection marked by a highly restricted pattern of viral gene expression. The prevalence of KSHV-infected cells undergoing lytic (productive) infection both in vitro and in vivo is typically low (1 to 5%), with the remaining infected cells harboring the virus in its latent form . Thus, mechanisms of immune evasion contingent upon genes expressed solely during the lytic cycle would protect only a small fraction of KSHV-infected cells. It follows, therefore, that KSHV may additionally require mechanisms of evasion that are active during latency. Several KSHV latent proteins have likewise been implicated in viral immune evasion. These include disruption of p53 by latency-associated nuclear antigen (LANA) and LANA2 , inhibition of apoptosis by viral FLIP , downregulation of the B-cell receptor by ORF K1 , and deregulation of the interferon response by vIRF-II . Additionally, work by Brander et al. has shown that KSHV-infected PEL cells express lower levels of MHC class I and possess increased resistance to CTLs compared with Epstein-Barr virus-infected or primary B cells . However, the relevance of these findings to de novo infection of endothelial cells with intact KSHV remains unknown. Therefore, we investigated the potential viral immune evasion strategies of KSHV in two endothelial cell culture systems that strongly favor latent infection with KSHV. Specifically, we tested telomerase-immortalized dermal microvascular endothelial cells (DMVEC) infected with concentrated virus and primary DMVEC (pDMVEC) infected via coculture with induced BCBL-1 cells, a PEL line harboring KSHV. With a panel of antibodies against numerous endothelial cell surface proteins, we examined these immortalized and primary cells for signs of immune modulation following de novo infection with KSHV. MATERIALS AND METHODS : Cell culture. | BCBL-1 and BJAB cells were grown in RPMI 1640 medium (Gibco, Rockville, Md.) supplemented with 10% fetal bovine serum, 10 mM HEPES (pH 7.5), 100 U of penicillin per ml, 100 mug of streptomycin per ml, 2 mM l-glutamine, 0.05 mM beta-mercaptoethanol, and 0.02% (wt/vol) sodium bicarbonate. Cells were maintained at 2.5 x 105 to 8.0 x 105/ml and were used for up to 2 months after the thaw date. pDMVEC were obtained from Clonetics Corporation (Walkersville, Md.) at low passage and grown in EGM-2 MV medium (Clonetics) with 50 U of penicillin per ml and 50 mug of streptomycin per ml substituted for amphotericin B and gentamicin. Creation of T4 TIME cells. | Telomerase-immortalized microvascular endothelial (T4 TIME) cells were generated from pDMVEC with the stable incorporation of the human telomerase gene (hTERT) by retroviral transduction. Briefly, the pBABE retroviral plasmid containing the hTERT gene (a gift from R. Weinberg) was packaged into a Moloney murine leukemia virus pseudo typed with the vesicular stomatitis virus glycoprotein G. This virus was then transduced into pDMVEC by spinfection for 2 h at 1,000 x g, and stable clones were selected in the presence of 0.5 mug of puromycin per ml for 1 month. hTERT expression was then confirmed with the Trapeze telomerase detection kit (Intergen, Manhattan N.Y.). One of the resultant clones, T4, expressed high levels of telomerase, as determined by telomerase repeat amplification analysis and has survived, to date, over 80 passages without detectable changes in morphology, growth rate, or susceptibility to infection with KSHV compared with early-passage T4 TIME cells (data not shown). Coculture infection of pDMVEC cells with KSHV. | pDMVEC at early passage (<15 passages) were infected by coculture with the primary effusion lymphoma cell line BCBL-1, based on methods described by Sakurada et al. . Briefly, BCBL-1 cells were induced with a 12-h exposure to 20 ng of O-tetradecanoyl phorbol 13-acetate per ml and 300 muM sodium butyrate and, after 3 days, cocultured directly with pDMVEC at various ratios for 12 h in RPMI 1640 medium supplemented with 10% fetal bovine serum and 100 mug of endothelial cell growth supplement per ml (BD Pharmingen, San Jose, Calif.). The pDMVEC were then washed vigorously to remove the BCBL-1 cells and cultured for an additional day in EGM-2 endothelial cell medium lacking hydrocortisone. We found that low levels of hydrocortisone, normally present in EGM-2 endothelial growth medium, led to a modest two- to threefold induction of immediate-early lytic protein expression in the immunofluorescence assay (data not shown). As a result, we eliminated hydrocortisone from the growth medium in all short-term experiments to minimize the levels of spontaneous lytic reactivation. For controls, pDMVEC were cocultured with uninduced BCBL-1 cells or BCBL-1 cells induced in the presence of 0.5 mM phosphonoformic acid (Sigma Aldrich, St. Louis, Mo.). Cell-free virus infection of T4 TIME cells. | BCBL-1 cells were induced as described above, and virus was collected from the supernatant on the sixth day after induction by centrifugation at 13,000 x g for 3 h. The viral pellet was resuspended at 100th the original volume in EGM-2 MV culture medium lacking hydrocortisone, and aliquots were frozen at -80C. A small aliquot of concentrated virus was used to determine the number of viral genome equivalents by Southern blot analysis with a fluorescently labeled single-stranded KSHV DNA probe complementary to ORF 73 (Random Primer DNA chemiluminescence kit; New England Nuclear, Boston, Mass.). Approximately 5,000 viral genome equivalents per endothelial cell were then used to infect T4 TIME cells in the presence of 8 mug of Polybrene per ml (Sigma Aldrich) for 2 h. Hydrocortisone was omitted from the culture medium 24 h before addition of the virus. For controls, a sister plate was infected with an equivalent amount of virus that was first heat inactivated for 30 min at 65C or UV-inactivated at 40 W for 60 min with 254-nm UV light. Cells were then washed to remove the Polybrene and incubated for 2 days in fresh EGM-2 medium lacking hydrocortisone. Flow cytometry. | pDMVEC and T4 TIME cells were infected with KSHV as described above and harvested for flow cytometry at the indicated times. Cells were detached from the plate with a 0.25% trypsin-EDTA solution (Clonetics), and 0.5 x 106 cells per sample were then transferred to a 96-well plate after assessing cell count and viability by trypan blue exclusion. For surface staining, samples were washed twice with 1x PBSA (1x phosphate-buffered saline [PBS] with 0.09% sodium azide) and blocked for 30 min with antibody staining buffer (3% fetal bovine serum in PBSA). Cell surface staining was then carried out for 1 h at 4C in the dark. All cell surface antibodies and isotype controls were obtained from BD Pharmingen and used at the recommended dilution of 0.25 mug of antibody/106 cells in antibody staining solution. Apoptotic cells were identified with the annexin V fluorescein isothiocyanate apoptosis detection kit as described by the manufacturer (BD Pharmingen). Cells were then washed twice with cold 1x PBSA and permeabilized for nuclear staining with the paraformaldehyde-saponin-based bromodeoxyuridine Fix/Perm kit from BD Pharmingen as described by the manufacturer. Nuclear staining to identify KSHV-infected cells was performed with a rat monoclonal antibody to the KSHV LANA protein (Applied Biosciences Inc., Columbia, Md.) conjugated to the fluorochrome Alexa 488 per the manufacturer's protocol (Molecular Probes, Eugene, Oreg.) and used at a 1:1,000 dilution in BD Pharmingen Perm/Wash buffer supplemented with 3% fetal bovine serum. Flow cytometry was performed with a FACSCalibur flow cytometer and analyzed with Cell Quest Pro software (Becton Dickinson, Bedford, Mass.). Prior to analysis, all samples were gated by forward and side scatter to eliminate dead cells. Immunofluorescence. | T4 TIME cells and pDMVEC (with and without KSHV) were grown overnight on Becton Dickinson eight-well culture slides coated with fibronectin. The slides were then fixed in 95% ethanol-5% acetic acid for 10 min at room temperature and permeabilized in PBS with 0.5% Triton X-100 for 20 min at room temperature as described by Moses et al. . Nonspecific binding was blocked by incubating the cells with 10% goat serum supplemented with 1% glycine and 3% bovine serum albumin and then with the appropriate primary antibody diluted in blocking buffer for 60 min. To determine the extent of KSHV infection, cells were assayed for LANA expression with Alexa 488-conjugated monoclonal antibody to LANA at a 1:1,000 dilution. Rabbit polyclonal antibodies to RTA (ORF 50), (a gift from D. Lukac), ORF 45 (a gift from Y. Yuan), MIR-1 (ORF K3) and MIR-2 (ORF K5) (both gifts from G. Hayward), MCP (ORF 25) and SCAF (ORF 17.5) were all used at 1:500 except MIR-2, which was used at 1:1,000. Following staining with primary antibodies, cells were washed twice with 1x PBS and then incubated with goat anti-rabbit immunoglobulin conjugated to Texas Red (Jackson Labs, West Grove. Pa.) diluted 1:150 for 60 min in the dark. Cells were then washed twice with 1x PBS and subsequently stained with 0.5 mug of 4',6'-diamidino-2-phenylindole (DAPI) (Sigma) per ml in 180 mM Tris (pH 7.5) for 30 min at 4C in the dark. The slide was then coated with Antifade mounting solution (Biomeda, Foster City, Calif.) and sealed with a coverslip. Paracrine experiments. | Medium was harvested from T4 TIME cells 48 h after infection with cell-free virus (as described above). The medium was cleared of cell debris and viral particles by ultracentrifugation at 75,000 x g for 30 min in an SW-55Ti rotor (Sorvall, Newtown, Conn.). Naive T4 TIME cells were then incubated with the virus-free supernatant for 48 h and prepared for flow cytometry as described above. Naive T4 TIME cells incubated with medium from mock-infected T4 TIME cells prepared similarly served as controls. RESULTS : KSHV infection of telomerase-immortalized endothelial cells. | To investigate potential KSHV-specific immune evasion strategies, we assessed changes in the expression of specific cell surface proteins following de novo infection of endothelial cells with KSHV. We focused our initial efforts on DMVEC because they represent a probable target cell of KSHV within Kaposi's sarcoma lesions in vivo . Previous attempts at infecting similar cells with KSHV, however, have resulted in initial infection rates of less than 5% . Therefore, we took advantage of recent findings indicating that human pDMVEC first immortalized with the human telomerase gene (hTERT) support higher rates of infection with concentrated KSHV . We created hTERT-immortalized microvascular endothelial cells (TIME cells) through the retroviral insertion of hTERT into pDMVEC and the subsequent selection of antibiotic-resistant colonies (see Materials and Methods). Infection of one of the resultant clones, T4, with the concentrated viral supernatant from the KSHV-infected PEL line BCBL-1 resulted in a dramatic morphological change in many of the target endothelial cells to a spindle-like phenotype within 24 h . This change is characteristic of KSHV infection of endothelial cells and coincided closely with the onset of latent gene expression, as evidenced by the detection of LANA by the immunofluorescence assay . In contrast, incubation of T4 cells with either UV-treated (not shown) or heat-inactivated KSHV caused neither LANA staining nor morphological changes associated with infection (Fig. and D). The proportion of cells infected with KSHV correlated with the amount of input virus, determined by viral genome copy number (see Materials and Methods). However, with viral concentrations sufficient to generate 100% infection of all of the target cells, we often observed excessive cell death. To avoid this issue, we optimized the initial infection level in most experiments to approximately 50% by maintaining a constant ratio of KSHV genome copy number to target cell of approximately 5,000:1. This high ratio may reflect the small proportion of DNA-containing particles that are infectious virions as well as the relatively low susceptibility of endothelial cells to KSHV infection (C. Tomescu and D. H. Kedes, unpublished observations). With initial infection rates of approximately 50%, the number of infected (LANA-positive) cells remained stable for several days. However, the proportion of KSHV-infected cells began to decrease with continued culture. This reduction began approximately 4 days postinfection, so that by 3 weeks fewer than 15% of the cells remained positive (data not shown). This reduction may reflect either loss of the virus from within infected endothelial cells over time or a growth advantage of uninfected over infected cells within the culture. As a result, we restricted most of our analyses to the first 72 h postinfection. We next characterized the frequency of viral lytic gene expression among the newly infected T4 TIME cells. Lagunoff and colleagues reported that incubation of a similarly derived TIME cell line with KSHV resulted in predominantly latent infection, with lytic protein expression confined to less than 1% of infected cells . To explore whether our T4 TIME cells displayed a similar latent bias after infection, we assayed for KSHV lytic protein expression by immunofluorescence with antibodies specific for four immediate-early lytic gene products, ORF 50 (RTA), ORF 45, K3 (MIR-1), and K5 (MIR-2), as well as two late lytic gene products, ORF 17.5 (SCAF) and ORF 25 (MCP). We also costained samples for latent protein expression with a monoclonal antibody to LANA to identify the KSHV-infected cells in the culture. With this approach, we observed a relative paucity of both immediate-early and late lytic protein expression in the great majority of KSHV-infected (LANA-positive) cells (Fig. , third and fourth columns). Specifically, immunofluorescence assays detected the immediate-early lytic proteins RTA, MIR-1, and MIR-2 in only 1% of KSHV-infected T4 TIME cells (Fig. , fourth column). ORF 45 antibodies gave similar results (data not shown), while even fewer cells demonstrated late lytic proteins such as SCAF (not shown) and MCP (Fig. , fourth column). This last finding may reflect premature termination of the lytic cascade prior to expression of late lytic proteins in the few cells undergoing spontaneous lytic reactivation. Under higher magnification, reactivity with antibodies to lytic proteins in most KSHV-infected cells was indistinguishable from background levels with secondary antibody alone (Fig. , third column) despite characteristic LANA staining (Fig. , first column). In parallel assays, the same set of lytic antibodies reacted with 10 to 15% of similarly induced BCBL-1 cells (not shown), reflecting the higher levels of lytic reactivation that characterize this PEL line . Together, these results argue that KSHV infection of T4 TIME cells, in the absence of inducing agents, is predominantly latent. Nevertheless, we cannot rule out the distinct possibility that low levels of lytic proteins, expressed in quantities below the sensitivity of our immunofluorescence assay, are present in some cells following KSHV infection. Downregulation of MHC class I, PE-CAM, and ICAM-1 after de novo infection of T4 TIME cells with KSHV. | To investigate potential KSHV-specific immune evasion strategies that might arise following initial infection, we examined newly infected T4 TIME cells with a panel of antibodies to endothelial cell surface proteins involved in immune recognition . This analysis included monoclonal antibodies to MHC class I, MHC class II, CD31 (PE-CAM), CD40, CD54 (ICAM-1), CD58 (LFA-3), CD80 (B7.1), CD86 (B7.2), and CD95 (Fas). While four of these nine proteins, MHC class II, B7.1, B7.2, and CD40, were only poorly expressed on T4 TIME cells, the five remaining proteins were well represented prior to KSHV infection. We monitored the cell surface expression of this set and observed a significant downregulation of three of the proteins shortly after infection. Included among these was MHC class I, which was distinctly downregulated within 48 h of KSHV infection . In contrast, incubation of T4 TIME cells with heat- or UV-inactivated virus led to no change in MHC class I levels compared with uninfected, naive cells (data not shown). The number of cells with downregulated MHC class I varied proportionally with the amount of input virus and, in turn, with the level of infection as reflected by LANA reactivity in immunofluorescence assays (not shown). Furthermore, flow cytometry of cultures that were approximately 50% infected (LANA positive) most often demonstrated a widened distribution of MHC class I signal rather than a biphasic curve with two distinct populations, indicating that the extent of downregulation per cell varied considerably . Nevertheless, the mean KSHV-associated reduction in MHC class I ranged between two and fivefold compared to mock-infected control cultures in four separate experiments. These measurements, however, represent a minimal estimate, since approximately 50% of the cells in the cultures were uninfected. In addition to MHC class I, KSHV infection of T4 TIME cells also led to surface downregulation of PE-CAM (CD31) and ICAM-1 (CD54) (Fig. and C, respectively). These two proteins, normally expressed on endothelial cells, provide adhesion and costimulatory signaling to CTLs during T-cell-endothelial cell interactions . Furthermore, inhibition of these signaling pathways greatly reduces the ability of CTLs to kill target cells . In contrast, KSHV infection of T4 TIME cells led to no appreciable change in the level of the cell surface protein LFA-3 (CD58) while slightly increasing the levels of another marker, Fas (CD95) (data not shown). These findings argue that the downregulation of MHC class I, ICAM-1, and PE-CAM during infection is specific and not merely a reflection of global or indiscriminant downregulation. To investigate the kinetics of immunoregulatory protein downregulation, we measured cell surface expression of MHC class I, PE-CAM, ICAM-1, and LFA-3 on T4 TIME cells at 12-h intervals following de novo infection with KSHV. For MHC class I, downregulation was not evident at 0 (not shown) or 12 h postinfection . However, downregulation was apparent by 24 h and reached its peak by 48 h . In contrast, LFA-3 levels remained unchanged or slightly elevated throughout this period . PE-CAM and ICAM-1 downregulation kinetics (not shown) were similar to those of MHC class I. For all three proteins, cell surface levels at 72 h (not shown) postinfection were similar to those at 48 h, suggesting that a new, lower expression set point may have been established. By 96 h postinfection, however, the percentage of LANA-positive cells within the culture began to decrease (see above). This led to a concomitant reduction in the number of cells that exhibited downregulation of immunoregulatory proteins, although the profile of proteins affected on these individual cells was unchanged from that at the earlier times. Apoptosis was not directly responsible for a decrease in the proportion of cells downregulating immunoregulatory proteins, since greater than 97% percent of all cells exhibiting downregulation were viable, as evidenced by a lack of annexin V reactivity (data not shown). At all times, the proportion of cells exhibiting surface protein downregulation correlated directly with the proportion of KSHV-infected (LANA-positive) cells within the culture (data not shown). These findings suggest that direct infection with KSHV may be required to induce downregulation on cells within the mixed culture. However, they do not exclude the possibility that downregulation may be variable on infected cells and that soluble paracrine factors or direct cell-to-cell contact may contribute to similar changes on uninfected neighboring cells. To investigate the possibility of a paracrine effect, we incubated naive T4 TIME cells for 48 h with virus-cleared medium removed from a culture of T4 TIME cells infected with KSHV 48 h earlier (see Materials and Methods). Flow cytometry of cells treated with this conditioned medium, however, revealed no downregulation of any of the cell surface proteins that we tested . Instead, treatment of naive cells with conditioned medium from KSHV-infected cells led to a modest upregulation of ICAM-1 and PE-CAM . These findings may reflect a response to proinflammatory cytokines potentially released by cells within the KSHV-infected culture. In any case, these results indicate that soluble factors are not sufficient to elicit downregulation in uninfected cells. However, cell-to-cell contact, as a mechanism of downregulation, remained a formal possibility. Addressing this question required a method to distinguish infected from uninfected cells within a mixed population while simultaneously measuring levels of cell surface proteins. Development of flow cytometric approach to identify KSHV-infected cells. | Since the downregulation of immunomodulatory proteins by KSHV did not appear to involve paracrine effects, we next tested if the effects were dependent on direct infection with KSHV. To investigate this possibility, we developed a flow cytometric approach, similar to that described by Kellam et al., to identify KSHV-positive cells following de novo infection . As a marker of infection, we chose to monitor the expression of LANA because it is an abundant viral protein expressed in all KSHV-infected cells . We first conjugated the highly reactive fluorochrome Alexa 488 to an anti-LANA rat monoclonal antibody to identify LANA-expressing cells in culture and then devised a fixative process to detect LANA expression in the nucleus while preserving membrane integrity for cell surface analysis (see Materials and Methods). To validate this approach, we tested the ability of this conjugated antibody to detect LANA expression within the KSHV-infected BCBL-1 cell line by flow cytometry. Greater than 95% of BCBL-1 cells were positive for LANA with this technique, while the reactivity of the same antibody with the KSHV-uninfected B-cell line BJAB was minimal . Furthermore, fluorescent microscopy revealed that BCBL-1 cells prepared for flow cytometric analysis with the fluorescently labeled LANA antibody exhibited the speckled nuclear staining pattern that is characteristic of the LANA immunofluorescence assay (Fig. , inset) (-, ). We next used this approach to identify newly infected T4 TIME cells following KSHV infection. LANA-specific flow cytometry detected evidence of KSHV infection of T4 TIME cells within 24 h of their incubation with the virus . Dot plot analysis demonstrated a marked shift in fluorescence in a subset of the cells that stained brightly for LANA (Fig. , gate R3). The mean fluorescence intensity of infected cells was also well above the background of mock-infected cells . In subsequent experiments, we compared cells with the highest (top 25%) LANA signal (Fig. , gate R3) to those with the lowest (bottom 25%) (Fig. , gate R2) to unambiguously delineate any potential differences in the surface protein expression profile of infected versus uninfected cells. The latter group stained with the same low intensity as naive cells and therefore likely represented only background staining. This strict gating approach allowed us to focus on KSHV-infected (LANA-positive) cells in the culture while minimizing potentially confounding effects from background fluorescence. Of note, in determining absolute infection rates, this flow cytometric technique, although specific for KSHV-infected cells, was less sensitive than parallel immunofluorescence assays with the same anti-LANA monoclonal antibody. This difference likely reflects the advantage of the immunofluorescence assay over flow cytometry in distinguishing between the distinct intranuclear, punctate LANA staining pattern and nonspecific background fluorescence. As a result, we used an anti-LANA immunofluorescence assay in each experiment to more accurately assess overall infection rates. Downregulation of MHC class I, PE-CAM, and ICAM-1 on KSHV-infected T4 TIME cells. | We next investigated whether the cells that exhibited immunoregulatory protein downregulation tracked with the population that gated as LANA positive by flow cytometry. KSHV infection of T4 TIME cells followed by staining for both intranuclear LANA and surface proteins demonstrated that increased expression of LANA generally correlated with lower expression of MHC class I, shifting the population from the upper left quadrant (low LANA/high MHC class I) to the lower right quadrant (high LANA/low MHC class I) in flow cytometry analyses . A similar association was also evident between LANA expression and the downregulation of PE-CAM and ICAM-1 , although ICAM-1 downregulation was also present in cells with intermediate to low LANA expression. In contrast, LFA-3 surface expression remained essentially unchanged in the same population, showing no correlation with LANA expression . Although these data suggested a trend between level of LANA expression and downregulation of MHC class I, PE-CAM, and, to a lesser extent, ICAM-1, the lack of signal intensity from intranuclear LANA fluorescence in our flow cytometry assay prevented a more definitive assessment of this relationship. Specifically, the LANA flow cytometry assay was unable to separate sufficiently infected cells with intermediate levels of LANA signal from uninfected cells with similar but nonspecific background fluorescence (note the overlap in LANA staining between KSHV-infected and mock-infected cells in Fig. .) Therefore, in subsequent experiments, we used the strict gating approach shown in Fig. to compare downregulation between the cells exhibiting the greatest (top 25%) and the least (bottom 25%) LANA fluorescence intensity and excluded the remainder of the cells for which the LANA signal overlapped with background staining exhibited on mock-infected cells. This gating focused the comparison between the populations that most likely represented infected and uninfected cells. Using four-color flow cytometric staining, we identified a close association between the downregulation of MHC class I and the downregulation of both PE-CAM and ICAM-1 (but not LFA-3) following KSHV infection, suggesting that these processes arise on a similar population of cells (Fig. , second column, lower left quadrants). To determine whether this population comprised infected or uninfected cells in the mixed culture, we stained T4 TIME cells for both cell surface protein expression and intranuclear LANA expression and gated on cells with the highest (top 25%) and lowest (bottom 25%) LANA signal as described above. We found that the bulk of the KSHV-infected population (highest LANA expressers) exhibited downregulation of MHC class I, PE-CAM, and ICAM-1, as evidenced by a shift in the cell population from the upper right quadrant to the lower left quadrant (Fig. , compare the third and fourth columns). Approximately 70% of KSHV-infected (LANA-positive) cells exhibited downregulation of MHC class I, while less than 3% of the same population showed downregulation of LFA-3 (Fig. , fourth column). In addition, codownregulation of MHC class I with PE-CAM and ICAM-1 was pronounced on the high LANA expressers but only minimal on the low LANA expressers (Fig. , compare the third and fourth columns, lower left quadrants). Together, these results indicate that the downregulation of immunoregulatory proteins occurs preferentially on cells directly infected with KSHV (highest LANA expressers). Primary DMVEC infected by coculture display similar patterns of MHC class I, PE-CAM, and ICAM-1 downregulation. | We next asked if KSHV infection of pDMVEC would recapitulate the pattern of immunomodulatory protein downregulation we observed following direct infection of T4 TIME cells. To circumvent the inefficiency of pDMVEC infection with cell-free virus, we modified a coculture technique first reported by Sakurada and coworkers, demonstrating that direct incubation of BCBL-1 cells with primary human umbilical vein endothelial cells (HUVEC) resulted in high-efficiency KSHV infection . In brief, we incubated BCBL-1 cells with O-tetradecanoyl phorbol 13-acetate and sodium butyrate for 12 h to initiate lytic replication and, after 3 days (the earliest time at which we detected released infectious virions [; C. M. O'Connor and D. H. Kedes, unpublished results]), cocultured these cells with adherent pDMVEC for 12 h in medium modified to support the growth of both cell types (see Materials and Methods). Infection of pDMVEC with the coculture method resulted in a dramatic morphological change and the onset of LANA expression within 24 h that was reminiscent of KSHV-infected T4 TIME cells (not shown). These changes were specific to infection with KSHV rather than general effects of the coculture system, since incubation with either uninduced BCBL-1 cells or BCBL-1 cells induced in the presence of the herpesvirus DNA replication inhibitor phosphonoformic acid induced no discernible change in the pDMVEC target cells (not shown). We were able to vary the proportion of infected cells by adjusting the ratio of induced BCBL-1 cells to pDMVEC within each coculture. For most experiments, we optimized infection rates of the pDMVEC to approximately 50% with a 10-fold excess of induced BCBL-1 cells, mimicking the infection rates we achieved in T4 TIME cells infected with approximately 5,000 genome copies per cell (see above). Results from flow cytometric analyses of newly infected pDMVEC cells paralleled the patterns of MHC class I downregulation in KSHV-infected T4 TIME cells, with again little to no change in surface expression of LFA-3 . As with the morphological changes we described above, downregulation was specific to infection by KSHV, since incubation with either uninduced BCBL-1 cells (not shown) or BCBL-1 cells induced in the presence of phosphonoformic acid led to no such reduction on the pDMVEC target cells . PE-CAM and ICAM-1 were likewise downregulated by KSHV following coculture infection, and their downregulation tracked with MHC class I downregulation (Fig. and D, lower left quadrants). In addition, LANA costaining by flow cytometry revealed an association between high levels of LANA expression and the downregulation of MHC class I , PE-CAM, and ICAM-1 (not shown) but not LFA-3 on newly infected pDMVEC. We next characterized the relative levels of viral lytic gene expression within the newly infected pDMVEC as we had for the T4 TIME cells. Sakurada and colleagues reported that coculture infection of HUVEC results in a predominantly latent system of infection and that detection of lytic expression by immunofluorescence assay was minimal . To confirm whether KSHV infection within pDMVEC displayed a similar latent predisposition after coculture, we assayed for KSHV lytic protein expression by the immunofluorescence assay with the same set of antibodies to immediate-early and late lytic proteins we had assayed in the T4 TIME cell system (see above). We found that coculture infection of pDMVEC cells led to a predominantly latent infection, with immunofluorescent evidence of immediate-early expression of the lytic proteins RTA and MIR-2 confined to less than 1% of infected cells (Fig. , third column). The remaining 99% of the infected cells demonstrated staining for those two proteins at levels no greater than with secondary antibody alone (Fig. , third column), despite characteristic LANA staining (Fig. , first column). Immunofluorescence assays with antibodies to the ORF 45, SCAF, and MCP proteins showed a similarly low prevalence in the pDMVEC cultures (not shown). These findings paralleled the results we obtained after direct cell-free KSHV infection of T4 TIME cells and indicate that de novo infection of either primary or telomerase-immortalized endothelial cells leads primarily to viral latency, with only rare immunofluorescence assay evidence of lytic gene expression within the population. Finally, to assess the duration of KSHV-induced downregulation in primary cells, we examined changes in cell surface protein expression in pDMVEC cells 3 weeks postinfection. To facilitate this, we infected pDMVEC with a lower initial ratio of induced BCBL-1 to pDMVEC than we used for short-term cultures (1:1 versus 10:1, respectively) and let the virus spread slowly through the culture. A similar approach has led to long-term KSHV infection with small initial amounts of purified virus to infect immortalized endothelial cells . By 3 weeks postinfection, approximately 95% of cells became LANA positive in the immunofluorescence assay (Fig. , first column) and exhibited morphological changes consistent with infection by KSHV (not shown). Long-term-infected pDMVEC also continued to display a primarily latent repertoire of KSHV gene expression, with immunofluorescence assay evidence of immediate-early lytic proteins such as RTA and MIR-2 confined to an extremely small (<1%) proportion of the infected cells (Fig. , third column). Along with the nearly 100% LANA reactivity in these long-term cultures, virtually all of the cells continued to display marked downregulation of MHC class I compared with a smaller fraction at day 2 postinfection, when LANA reactivity was confined to less than 10% of the cells . PE-CAM and ICAM-1 were also downregulated by KSHV following long-term coculture infection, and their downregulation paralleled the reduction in MHC class I (Fig. , lower left quadrants). These findings suggest that downregulation of immunoregulatory molecules by KSHV is not a transient process but rather a continuous phenomenon associated with infection. DISCUSSION : Here we describe the surface downregulation of immunoregulatory proteins by KSHV in newly infected endothelial cells in culture. Using flow cytometry, we observed clear downregulation of MHC class I, PE-CAM, and ICAM-1 on both immortalized and primary endothelial cells following de novo infection with KSHV. This downregulation appears to be a specific phenomenon rather than a result of a global downregulatory mechanism, since levels of other surface markers, including Fas (not shown) and LFA-3 (Fig. and ), remain unchanged or slightly elevated after infection. Furthermore, the downregulation that we observed is not a transient mechanism associated only with the early events of viral infection. Instead, downregulation of immunoregulatory proteins persists in KSHV-infected cells even after 3 weeks in culture . However, we cannot exclude the possibility that long-term infection with KSHV requires continual reinfection of naive cells rather than the long-lived growth and expansion of KSHV-infected cells from earlier time points postinfection. Importantly, work by others has demonstrated that KSHV infection of similarly derived TIME cells results in a predominantly latent profile of viral gene expression . The present work is the first demonstration of efficient KSHV infection of pDMVEC by coculture, although infection of HUVEC cells by this method also results in a primarily latent viral gene expression profile . Similarly, our analysis of both immediate-early and late lytic protein expression by immunofluorescence also indicates a mainly latent infection not only for the T4 TIME cells but also for pDMVEC. In both systems, less than 1% of KSHV-infected (LANA-positive) cells expressed detectable lytic proteins after KSHV infection (Fig. and ). The number of cells exhibiting surface molecule downregulation varied proportionally with the amount of input virus and correlated roughly with the extent of KSHV infection within each infected culture. This suggests that direct infection with KSHV may lead to immunoregulatory protein downregulation and led us to investigate downregulation on an individual-cell basis. To this end, we created a flow cytometric assay targeting the LANA protein of KSHV to identify newly infected cells in culture. This assay exhibits both high sensitivity and specificity when tested on mixtures of KSHV-infected PEL cells and uninfected B cells and also distinguishes newly infected T4 TIME cells from uninfected cells in the same culture . T4 TIME cells, however, are less well separated because they have lower overall levels of LANA expression than BCBL-1 cells (Fig. and B). Nevertheless, it is apparent that, when used together with staining of immunoregulatory molecules on the surface of cells, this approach confirms that downregulation of MHC class I, PE-CAM, and ICAM-1 but not LFA-3 occurs primarily on KSHV-infected (LANA-positive) cells (Fig. , fourth column). These results are supported by paracrine experiments showing that virus-free medium from KSHV-infected cells fails to induce downregulation of the same markers on naive cells . However, we also found that a small fraction (approximately 15%) of LANA-negative cells exhibited immunoregulatory protein downregulation. While it is possible that the poor LANA expressers (the R2 gate in Fig. ) contain a subset of infected cells with LANA levels that fall below the detection limit of our assay, they may also represent an essentially uninfected group of cells. Therefore, we cannot exclude the possibility that infected cells may exert a downregulatory effect on some of their uninfected neighbors. Similarly, a small but consistent minority (approximately 25%) of KSHV-infected (LANA-positive cells) failed to undergo immunoregulatory protein downregulation. This may be explained by a threshold effect of downregulation where measurable immunoregulatory protein downregulation requires a critical level of viral gene expression. In agreement with this hypothesis, our recent findings indicate that the degree of downregulation best parallels the percentage of infected cells in the culture exhibiting 25 or more discrete punctate LANA dots per nucleus in the immunofluorescence assay (Tomescu and Kedes, unpublished observations). In contrast, infection of T4 TIME cells or pDMVEC with smaller amounts of virus leads to fewer dots per nucleus and a correspondingly lower degree of immunoregulatory protein downregulation. These analyses indicate that downregulation of immunomodulatory proteins following de novo infection occurs preferentially on KSHV-infected cells and that manifestation of the effect, and perhaps its magnitude, may depend on initially delivering a sufficient number of viral genomes into each cell or, in longer-term infections, generating sufficient copies of the viral episome. The incubation of naive T4 TIME cells with conditioned medium from KSHV-infected T4 TIME cells leads to a modest upregulation of PE-CAM and ICAM-1, likely reflecting the inflammatory nature of KSHV infection . Likewise, incubation of naive T4 TIME cells with either UV- or heat-inactivated KSHV leads to similar upregulation during mock infection (data not shown). Finally, flow cytometry indicates that even within KSHV-infected cultures, a fraction of cells demonstrates an increased level of PE-CAM and ICAM-1 cell surface expression compared to cells in parallel cultures not exposed to virus (data not shown). The size of this population of cells is proportional to the percentage of cells judged to be uninfected (LANA negative) in each experiment by parallel immunofluorescence assay (not shown). The remainder of the cells within these KSHV-infected culture has levels of PE-CAM and ICAM-1 that are at or below the basal levels of these proteins on naive T4 TIME cells. One explanation for these results is that that the presence of KSHV induces a strong inflammatory reaction in T4 TIME cells but that infection of these cells with KSHV is sufficient to prevent this upregulation. This may have important implications for immune evasion since PE-CAM and ICAM-1 signaling by endothelial cells is required to induce the full activation and killing capacity of CTLs . Furthermore, these findings emphasize the strength of this immunomodulatory effect and highlight the ability of KSHV to induce downregulation in the presence of an inflammatory environment normally aimed at upregulating these molecules. In vivo, immunohistochemical studies indicate that while PE-CAM and ICAM-1 expression is detectable in Kaposi's sarcoma lesions, their expression within tumor spindle cells themselves is often patchy or absent . Therefore, downregulation of PE-CAM and ICAM-1 expression by KSHV may constitute a continual protective strategy used by the virus within the inflammatory environment of the lesion . The downregulation of MHC class I and/or costimulatory molecules is a common immune evasion tactic used by many viruses, including several members of the herpesvirus family . For herpesviruses, these strategies help to establish and maintain infection in the host despite the continuous presence of a virus-specific cellular immune response. The importance of such defensive mechanisms in herpesviruses is particularly well exemplified by human cytomegalovirus, which encodes four separate proteins to block MHC class I presentation at three fundamentally different steps . Recently, several groups have shown that KSHV encodes two immediate-early lytic proteins that serve an analogous function. These proteins, MIR-1 and MIR-2, are capable of downregulating MHC class I proteins as well as T-cell costimulatory molecules when overexpressed in heterologous cells in vitro . At least in these single-gene experiments, the resultant downregulation is sufficient to provide cells with resistance to immune cells in culture and therefore may protect KSHV-infected cells undergoing lytic replication in vivo. The contribution of MIR-1 and MIR-2 to immune evasion by KSHV during latency, however, is less clear. Analysis of viral mRNA expression both in vivo within Kaposi's sarcoma lesions and in vitro in infected PEL cell lines indicates that, in most cells, KSHV gene expression is restricted to a small subset of latent genes . In addition, our immunofluorescence assay of lytic protein expression in the culture systems described in this report also argues that only a small proportion (approximately 1%) of all infected cells undergoes spontaneous lytic reactivation, with the remainder of cells harboring KSHV in the latent form. A priori, an effective overall immune evasion strategy employed by KSHV would protect not only this small number of lytically infected cells during their relatively short life span, but also the more predominant population of latently infected cells that may persist within the host for prolonged periods. Since the majority of KSHV-infected cells adopt a latent phenotype, it follows that downregulation of immunoregulatory proteins both in vivo and in the two models of infection that we describe here may depend on a latent gene or combination of latent genes. Nevertheless, we cannot rule out the distinct possibility that extremely low levels of lytic gene products, including MIR-1 and/or MIR-2, remain below the detection levels of our assays yet are responsible, at least in part, for the immunoregulatory protein downregulation that we observed following de novo infection with KSHV. Recent studies have shown that MIR-1 and MIR-2 function as type 3 ubiquitin ligases, targeting surface molecules for internalization and subsequent destruction through the ubiquitin-proteosome pathway . Consequently, their enzymatic nature allows the theoretical possibility that low levels of these proteins (potentially undetectable by the immunofluorescence assay) may be sufficient to induce surface molecule downregulation. In addition, a recent study has shown that MIR-2 may represent a unique KSHV immediate-early lytic protein whose expression may be independent of the main lytic transactivator of KSHV, RTA . Therefore, determining the importance of latent and/or lytic genes in biologically meaningful modes of immune evasion during the different stages of KSHV infection and tumorigenesis will depend on not only determining the genes that are capable of inducing these effects but also assessing the nature of the balance between the latent and lytic state within individually infected cells. FIG. 1. : Morphological changes associated with infection of T4 TIME cells by KSHV. Morphological changes associated with infection of T4 TIME cells by KSHV. KSHV-infected (A) and mock-infected (C) T4 TIME cells were identified by phase microscopy, with superimposed DAPI staining of the nuclei 48 h after infection. (B and D) Cells from panels A and C, respectively, stained for LANA expression and visualized by fluorescence microscopy. FIG. 2. : KSHV-infected T4 TIME cells stained for latent and lytic protein expression and analyzed by immunofluorescence with the indicated antibodies 48 h after infection. KSHV-infected T4 TIME cells stained for latent and lytic protein expression and analyzed by immunofluorescence with the indicated antibodies 48 h after infection. The first, second, and third columns represent immunofluorescence assays at high magnification (40x) of a single field for LANA, DAPI, and the indicated lytic protein, respectively. Note that the background reactivity of KSHV-infected T4 TIME cells to the indicated lytic antibodies was indistinguishable from that with secondary antibody alone. The fourth column represents a parallel culture of infected cells stained with the same antibodies at low magnification (10x), with insets showing magnified views of the indicated (boxed) reactive cell. Values represent the percentage of KSHV-infected (LANA-positive) cells that exhibited reactivity to the corresponding lytic antibody. FIG. 3. : Downregulation of immunoregulatory proteins on T4 TIME cells following infection with KSHV. Downregulation of immunoregulatory proteins on T4 TIME cells following infection with KSHV. Flow cytometric analysis of KSHV-infected (green line) and mock-infected (pink line) T4 TIME cells stained for surface protein expression with the indicated antibodies 36 h after infection. The shaded purple histogram indicates the isotype control. Note that ICAM-1 downregulation was unusually pronounced in this experiment. Mock-infected cells were incubated with heat-inactivated KSHV; UV-inactivated KSHV gave similar results. FIG. 4. : Kinetics of MHC class I downregulation following KSHV infection of T4 TIME cells. Kinetics of MHC class I downregulation following KSHV infection of T4 TIME cells. Flow cytometric analysis of KSHV-infected (green line) and mock-infected (pink line) T4 TIME cells stained with antibodies to MHC class I (A) and LFA-3 (B) at the indicated times postinfection. The results at 0 and 72 h (not shown) were identical to those at 12 and 48 h, respectively. Note that the levels of KSHV infection per cell and, as a result, the number of cells downregulating MHC class I were slightly lower in this experiment than in most others (see, for example, Fig. ). FIG. 5. : Absence of soluble paracrine effects in downregulation of immunoregulatory proteins by KSHV. Absence of soluble paracrine effects in downregulation of immunoregulatory proteins by KSHV. (A to D) Flow cytometric analysis of naive T4 TIME cells incubated with virus-free medium from KSHV-infected (green line) or mock-infected (pink line) T4 TIME cells and stained with the indicated antibodies. FIG. 6. : Flow cytometry of LANA staining in KSHV-infected B cells and T4 TIME cells. Flow cytometry of LANA staining in KSHV-infected B cells and T4 TIME cells. (A) KSHV-infected BCBL-1 cells (green line) and uninfected control BJAB cells (pink line) were stained with a fluorescently labeled monoclonal antibody targeting LANA. Inset: A typical KSHV-infected BCBL-1 cell after LANA staining for flow cytometry but visualized by fluorescence microscopy. (B) KSHV-infected T4 TIME cells (green line) and mock-infected T4 TIME cells (pink line) stained for LANA as in panel A. Inset: A typical KSHV-infected T4 TIME cell after LANA staining for flow cytometry but visualized by fluorescence microscopy. (C) Dot plot analysis of LANA expression in mock-infected T4 TIME cells. (D) Dot plot analysis of LANA expression in KSHV-infected T4 TIME cells. Gates R2 and R3 represent cells after virus exposure with the least (bottom 25%) and the greatest (top 25%) reactivity with LANA monoclonal antibody, respectively. Gate R1 was used to isolate live cells by forward and side scatter (not shown). Note that essentially no mock-infected T4 TIME cells fell within the R3 gate. FIG. 7. : Association between LANA expression (x axis) and downregulation of immunoregulatory proteins (y axis) in T4 TIME cells. Association between LANA expression (x axis) and downregulation of immunoregulatory proteins (y axis) in T4 TIME cells. (A to H) Flow cytometry of mock-infected and KSHV-infected cells following membrane and nuclear staining with the indicated antibodies 48 h after infection. Note the shift in the cell population in KSHV-infected samples from the upper left quadrant to the lower right quadrant for MHC class I, PE-CAM, and ICAM-1 but not LFA-3. Quadrants were set based on mock-infected control samples for the indicated surface antibodies and isotype controls for LANA staining. FIG. 8. : Downregulation of immunoregulatory markers on KSHV-infected but not mock-infected T4 TIME cells. Downregulation of immunoregulatory markers on KSHV-infected but not mock-infected T4 TIME cells. Mock-infected (first column) and KSHV-infected (second column) T4 TIME cells were stained for surface protein expression with the indicated antibodies 48 h postinfection. The third and fourth columns show downregulation of immunoregulatory proteins on cells from the second column but first gated on the lowest (bottom 25%, blue cells) or the highest (top 25%, green cells) LANA-expressing cells (see Fig. ). Quadrants were set based on mock-infected control samples for the indicated surface antibodies. Values represent the percentage of the gated cell population present within each quadrant. FIG. 9. : Downregulation of immunoregulatory proteins on pDMVEC cells following coculture infection with KSHV. Downregulation of immunoregulatory proteins on pDMVEC cells following coculture infection with KSHV. (A and B) Flow cytometric analysis of KSHV-infected (green line) and mock-infected (pink line) pDMVEC cells stained for surface protein expression with the indicated antibodies 48 h after infection. Mock-infected pDMVEC cells were cocultured with BCBL-1 cells induced in the presence of phosphonoformic acid. (C and D) Simultaneous downregulation of PE-CAM and ICAM-1, respectively, with MHC class I on the surface of pDMVEC cells after coculture infection. Quadrants were set as in Fig. . (E and F) Expression of MHC class I but not of LFA-3 decreased on cells displaying the highest levels of LANA staining. Quadrants were set as in Fig. . Note that the levels of KSHV infection per cell and, as a result, the number of cells exhibiting downregulation was slightly lower in this experiment than in most others. FIG. 10. : KSHV-infected pDMVEC cells stained for latent and lytic protein expression with the indicated antibodies either 2 (A) or 23 (B) days after coculture infection. KSHV-infected pDMVEC cells stained for latent and lytic protein expression with the indicated antibodies either 2 (A) or 23 (B) days after coculture infection. The first, second, and third columns represent immunofluorescence assays of a single field for LANA, DAPI, and the indicated lytic protein, respectively. Note that the background reactivity of KSHV-infected pDMVEC cells to the indicated lytic antibodies was indistinguishable from that with secondary antibody alone (not shown for day 23). FIG. 11. : Continued downregulation of MHC class I, PE-CAM, and ICAM-1 on pDMVEC cells after long-term KSHV infection. Continued downregulation of MHC class I, PE-CAM, and ICAM-1 on pDMVEC cells after long-term KSHV infection. (A) pDMVEC cells stained for MHC class I expression by flow cytometry at days 2 and 23 postcoculture infection. (B) Simultaneous downregulation of PE-CAM and ICAM-1 with MHC class I on the surface of KSHV-infected pDMVEC cells 23 days after coculture infection. Quadrants were set as in Fig. . Backmatter: PMID- 12915568 TI - Specific Association of Glycoprotein B with Lipid Rafts during Herpes Simplex Virus Entry AB - Herpes simplex virus (HSV) entry requires the interaction of glycoprotein D (gD) with a cellular receptor such as herpesvirus entry mediator (HVEM or HveA) or nectin-1 (HveC). However, the fusion mechanism is still not understood. Since cholesterol-enriched cell membrane lipid rafts are involved in the entry of other enveloped viruses such as human immunodeficiency virus and Ebola virus, we tested whether HSV entry proceeds similarly. Vero cells and cells expressing either HVEM or nectin-1 were treated with cholesterol-sequestering drugs such as methyl-beta-cyclodextrin or nystatin and then exposed to virus. In all cases, virus entry was inhibited in a dose-dependent manner, and the inhibitory effect was fully reversible by replenishment of cholesterol. To examine the association of HVEM and nectin-1 with lipid rafts, we analyzed whether they partitioned into nonionic detergent-insoluble glycolipid-enriched membranes (DIG). There was no constitutive association of either receptor with DIG. Binding of soluble gD or virus to cells did not result in association of nectin-1 with the raft-containing fractions. However, during infection, a fraction of gB but not gC, gD, or gH associated with DIG. Similarly, when cells were incubated with truncated soluble glycoproteins, soluble gB but not gC was found associated with DIG. Together, these data favor a model in which HSV uses gB to rapidly mobilize lipid rafts that may serve as a platform for entry and cell signaling. It also suggests that gB may interact with a cellular molecule associated with lipid rafts. Keywords: Introduction : Herpes simplex virus (HSV) is typically responsible for mucosal lesions of the mouth and genital organs in humans, from where it spreads and establishes lifelong latent infections in sensory neurons. Periodically, the virus reactivates, multiplies, and is transported through the axon back to a portal of entry . Binding to host cell surfaces and entry is a complex process involving the essential viral glycoproteins B (gB), gD, gH, and gL and multiple cellular molecules, each with various levels of affinity and avidity (reviewed in references and ). In current models, gC and/or gB binds cell surface heparan sulfate proteoglycans, bringing the viral envelope and plasma membrane close enough for fusion to occur . As part of this process, gD must bind to a specific receptor, which can be either herpesvirus entry mediator A (HVEM or HveA), a member of the tumor necrosis factor receptor family, nectin-1 (HveC) or nectin-2 (HveB), two members of the Immunoglobulin superfamily, or a particular type of modified heparan sulfate proteoglycans (HSPG) 3-OST-3 . These interactions may then recruit the other essential viral glycoproteins into a functional fusion unit. In addition, entry may involve plasma membrane rearrangement, signaling events, and/or recruitment of additional cellular molecules. Accumulated evidence indicates that plasma membrane microdomains, or lipid rafts, that are highly enriched in cholesterol and sphingolipids play a crucial role in the lateral organization of the plasma membrane . It has been proposed that constitutive or transient enrichment of a variety of signaling molecules in these defined microdomains plays a major role in the organization of signal transduction. These domains retain substantial lateral mobility and are viewed as highly ordered moving platforms that carry specific proteins. Several viruses have taken advantage of lipid rafts for one or more aspects of their replication cycle (reviewed in references , , , and ). Such mechanisms include viral particle assembly , budding from the plasma membrane , signaling , fusion , and virus entry . It was proposed that human immunodeficiency virus (HIV) entry is inhibited by the presence of drugs that remove cholesterol . The inhibitory effect was reversed by addition of exogenous cholesterol, indicating that cholesterol-enriched lipid rafts play an important role in HIV entry. Since HIV and HSV enter cells by direct fusion, an intriguing possibility is that rafts also play a role in HSV entry (reviewed in reference ). Receptors for HIV, including CD4 and CCR5 , and for murine leukemia virus are found in lipid rafts. Binding of gp120 to cells further recruits a larger number of HIV receptors into lipid rafts . One evident question is whether the receptors for HSV associate with lipid rafts. It has recently been shown that some members of the tumor necrosis factor receptor superfamily, including CD120a, CD40, and the p75 neurotrophin receptor, are localized in rafts . Cross-linking of CD40 with antibodies results in stable association with lipid rafts, leading to activation of tyrosine kinases and mobilization of tumor necrosis factor receptor-associated factors (TRAFs) . Such events are essential for downstream events, such as NF-kappaB activation and interleukin expression. Epstein-Barr virus has exploited this signaling route by virtue of the localization of the latent membrane protein-1 (LMP1) in rafts. Association of LMP1 with rafts is responsible for signaling events that mimic those of a constitutively active CD40 receptor . An essential role for rafts has been described for the initiation of Fas-mediated cell death signaling . As a tumor necrosis factor receptor-like protein, HVEM contains TRAF-binding motifs in its cytosolic tail , and several reports demonstrated activation of NF-kappaB during early events of HSV infection . In addition, several proteins are rapidly phosphorylated during HSV entry . Hence, it is of interest to know whether HVEM is constitutively distributed in raft-like structures on cell surfaces or redistributes there as a consequence of virus binding. If so, is such an association important for virus entry? Similar questions may be asked about a second HSV receptor, nectin-1 (HveC). Strong evidence for membrane raft-dependent scaffolding of signaling complexes has come from studies on the T- and B-cell immunoreceptors and the Fcepsilon receptor, all members of the same protein family as nectin-1 . Here we present data investigating the importance of intact lipid rafts at the cell membrane for efficient virus replication. Second, we analyzed whether two HSV receptors, HVEM and nectin-1, as well as HSV glycoproteins showed an association with lipid rafts during virus entry. We found that HVEM and nectin-1 were not associated with rafts in uninfected cells and that this distribution did not change during infection. Similarly, gD, gC, and gH were not found associated with these microdomains during infection. By contrast, a fraction of gB was associated with rafts after virus attachment and during entry. This last observation points to a unique function for gB during HSV entry that involves lipid rafts and suggests the existence of a gB receptor(s) that is enriched in the cholesterol-rich microdomains. MATERIALS AND METHODS : Cells and viruses. | African green monkey kidney (Vero) and human embryonic kidney (HEK) 293T cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 5% and 10% fetal calf serum (FCS), respectively. B78-H1 mouse melanoma cells expressing gD receptor HVEM (A10) or nectin-1 (C10) were grown in DMEM supplemented with 5% FCS and 500 mug of G418 per ml. B78-H1 cells (3E5) expressing a green fluorescent protein (GFP)-tagged form of HVEM (enhanced green fluorescent protein [EGFP] at the C terminus of HVEM) were cultured similarly. These cells were shown to be permissive for infection by HSV (C. Whitbeck, unpublished results). HSV-1 (KOS) and an HSV-1 strain (KOS/tk12) that carries the lacZ gene under the control of the ICP4 promoter were both purified on sucrose gradients as described elsewhere . HSV-1 (KOS/tk12) and vesicular stomatitis virus were kindly provided by P. G. Spear and R. N. Harty, respectively. Antibodies and reagents. | Rabbit R47 and R137 sera were raised against HSV gC-1 and gH-1/gL-1, respectively . Monoclonal antibody (MAb) DL6 recognizes a linear epitope on gD ; MAb SS-10 was generated by immunizing mice with gB purified from an extract of HSV-1-infected BHK cells (unpublished data); MAb CW10 was raised against a recombinant form of HVEM (200t) expressed in a baculovirus expression system (C. Whitbeck, unpublished data); MAbs CK6 and CK41 were raised against a recombinant form of nectin-1 . CK41 was also labeled with phycoerythrin at Molecular Probes. MAb to flotillin-2/ESA was obtained from BD Transduction Laboratories. Anti-mouse and anti-rabbit immunoglobulin secondary antibodies coupled to horseradish peroxidase (HRP) were purchased from Kirkegaard and Perry Laboratories. Cholera toxin B subunit peroxidase conjugate (CTB-HRP) was from Sigma and used at 40 ng/ml. 7-Amino-actinomycin D (BD Pharmingen) was used for the exclusion of nonviable cells in flow cytometric assays following the manufacturer's instructions. Production and purification of HSV-1 glycoproteins. | gC(457t) contains the full ectodomain of gC1 truncated at residue 457 . gD(285t), a form of the ectodomain of gD truncated at residue 285 , binds the receptor molecules HVEM and nectin-1 with higher affinity than gD(306t) . The ectodomain comprising the first 730 amino acids of gB was expressed by recombinant baculovirus-infected insect cells. Based on the nucleotide sequence of the HSV-1 gB open reading frame, we synthesized two PCR primers in order to amplify and modify the gB ectodomain coding region for cloning and expression in a recombinant baculovirus. The first primer, 5'-CGGGATCCGGCGGCTCCGACTTC-3', hybridized to the noncoding strand of the gB ORF immediately beyond the region coding for the predicted signal sequence and incorporated a BamHI restriction enzyme cleavage site (bold letters). The second primer, 5'-GCGTGATCAGGCGGCGTTGGCGTCGGCGTGGATGAC-3', hybridized to the coding strand of the cloned gB open reading frame immediately prior to the transmembrane region coding sequence (residue 730) and incorporated a BclI restriction enzyme cleavage site (bold letters). After cloning into the BamHI site of the pVT-Bac transfer vector, the gB coding region was positioned downstream of and in frame with the mellitin signal sequence coding region. The resulting plasmid construct was cotransfected with baculovirus DNA (Baculogold; Pharmingen) into Sf9 cells grown in monolayer culture. After 4 days, the culture supernatant (containing recombinant progeny virus) was plated onto Sf9 cell monolayers under Grace's insect cell medium containing 1% agarose. Recombinant virus plaques were picked and amplified, and infected cell cultures were screened for the expression of gB by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblot analysis with R69 antiserum. The truncated protein called gB(730t) was purified from cells with a DL16 immunosorbent column. This MAb recognizes a dimer-specific epitope on gB, and hence all of the purified protein is dimeric. Conditions for elution and concentration were essentially the same as described for gD1(306t) . Cholesterol sequestration and virus entry assay. | Vero, A10, or C10 cells were seeded in 96-well plates and grown overnight to reach about 4 x 104 cells per well. Cells were then incubated for 30 min at room temperature with serial dilutions of methyl-beta-cyclodextrin (MbetaCD) or nystatin (both reagents from Sigma) in cell culture medium containing 30 mM HEPES. After three washes with medium, cells were infected with HSV-1 (KOS/tk12) at a multiplicity of infection of 10 and allowed to attach to the cells at 4C for 1 h. The temperature was then shifted to 37C to allow infection to proceed synchronously. Five hours later, cells were lysed by adding an equal volume of DMEM containing 1% NP-40. beta-Galactosidase activity was determined by adding substrate (chlorophenol red-beta-d-galactopyranoside; Roche) and measuring the absorbance at 570 nm in a microtiter plate reader. In replenishment experiments, Vero cells were first treated with 7.5 mM MbetaCD for 30 min. MbetaCD was washed out as before, and various amounts of water-soluble cholesterol balanced with MbetaCD (Sigma) in DMEM without serum were added. After 30 min, the cholesterol-containing medium was washed out, cells were infected with HSV-1 (KOS/tk12), and the process of entry was assayed as before. To deplete cholesterol from purified virions, 2 x 106 PFU was incubated for 30 min at room temperature with serial dilutions of MbetaCD and then diluted 60-fold with cell culture medium containing HEPES. Vero cells seeded into 96-well plates as above were infected with the treated virus at a multiplicity of infection of 10, and entry was assayed as before. Plaque formation assay. | Vero cell monolayers in 48-well plates were incubated for 30 min at room temperature with serial dilutions of MbetaCD in cell culture medium containing 30 mM HEPES. After three washes with medium, cells were infected with 100 PFU of HSV-1 (KOS) or vesicular stomatitis virus at 37C. One hour postinfection, the medium was removed, and the cells were overlaid with DMEM containing 1% carboxymethylcellulose and 5% FCS and incubated for an additional 24 h. The cells were then fixed with methanol-acetone (2:1 ratio) for 20 min at -20C and air dried. Virus titers were determined by an immunoperoxidase assay with a mixture of anti-gB, -gC, -gD, and -gH/gL polyclonal antisera (HSV) or stained with crystal violet (vesicular stomatitis virus). Flow cytometry. | C10 or A10 seeded in six-well plates was incubated for 30 min at room temperature with serial dilutions of MbetaCD in cell culture medium containing 30 mM HEPES. After three washes with medium, cells were detached with 0.02% EDTA in phosphate-buffered saline (PBS) (Versene; Gibco-BRL), pelleted, and resuspended in 100 mul of PBS containing 3% FCS. Cells were stained on ice for 1 h by the addition of 5 mug of CK41 per ml directly labeled with phycoerythrin to detect nectin-1. After a PBS wash, cells were fixed with 3% paraformaldehyde in PBS and analyzed by fluorescence-activated cell sorting (FACS). Transient transfection of 293T cells. | HEK-293T cells (80% confluent) grown in a T75 flask were transfected with 5 mug of endotoxin-free (Qiagen) purified plasmid pBEC14, encoding HVEM , or pBG38, encoding nectin-1 , with GenePorter, as recommended by the manufacturer (Gene Therapy Systems). Twenty-four hours later, cells were fractionated on sucrose gradients as described below. Plasmids pBEC14 and pBG38 were kindly provided by P. G. Spear. Isolation of low-density detergent-insoluble membrane fractions on sucrose gradients. | Low-density detergent-insoluble membrane microdomains were isolated essentially as described by others with some modifications . Briefly, confluent cell monolayers in a T75 flask were washed twice with PBS and then scraped into 1 ml of ice-cold MNE buffer (25 mM MES [2-{N-morpholino}ethanesulfonic acid, pH 6.5], 150 mM NaCl, 2 mM EDTA) containing 1% Triton X-100 (Fluka) and a cocktail of protease inhibitors (Roche). Cells were further homogenized with 10 strokes in a Dounce homogenizer, and then 800 mul of the homogenate was adjusted to 45% sucrose (prepared in MNE) and placed at the bottom of an ultracentrifuge tube. A discontinuous gradient was formed by overlaying the homogenate sequentially with 1.6 ml of 35% and of 5% sucrose (both prepared in MNE). These mixtures were centrifuged at 200,000 x g and 4C for 16 h in an SW50 swinging-bucket rotor. The gradient was fractionated from the top, and 12 fractions of 400 mul each were collected. A 13th fraction was obtained by extracting the pellet at room temperature with 400 mul of 4% SDS in 125 mM Tris, pH 6.8. Isolation of detergent-resistant membranes by sequential centrifugation. | Nonionic detergent-insoluble glycolipid-enriched membranes (DIG), which are soluble in octylglucoside, were prepared as described by others with minor modifications. Briefly, confluent cells (T25 flask) were scraped into 1 ml of ice-cold MNE buffer containing 1% Brij-96 (Fluka) and a cocktail of protease inhibitors (Roche). After 10 strokes in a Dounce homogenizer, the nuclei were pelleted by centrifugation at 500 x g for 5 min. Supernatants were sequentially centrifuged at 10,000 x g for 10 min and then at 200,000 x g for 1 h with an SW50 rotor. Supernatants containing detergent-soluble membranes as well as cytosolic proteins are referred to as DSM. The insoluble pellet was extracted with 200 mul of octylglucoside solution (50 mM beta-octylglucopyranoside in 20 mM Tris [pH 8], 200 mM NaCl, and protease inhibitors). Insoluble membranes still attached to the nuclei and the pellet from the 10,000 x g centrifugation run were also both reextracted with 200 mul of octylglucoside solution. The octylglucoside-soluble fractions were pooled and referred to as DIG. Measuring binding of virus or soluble glycoprotein to the cell surface. | Confluent C10 cells (T75 flask) were washed once with cold culture medium containing 30 mM HEPES and then incubated for 1 h at 4C with HSV-1 (KOS) (multiplicity of infection, 10 to 20). Cells were then either left on ice or shifted to 37C for various times before being washed with cold PBS. Cells were then scraped into 1 ml of ice-cold MNE buffer containing 1% Triton X-100, homogenized, and fractionated as described earlier in the section on isolation of low-density detergent-insoluble membrane fractions on sucrose gradients. Alternatively, cells were grown to confluency (T25 flask), washed, and incubated as before but now in the presence of 0.1 muM purified soluble gB(730t), gC(457t), or gD(285t). Cells were then either left on ice or shifted to 37C for various times before being washed with cold PBS. The DIG were then isolated as described in the section on isolation of detergent-resistant membranes by sequential centrifugation. Western and dot blot analyses. | To analyze the distribution of proteins in the fractions obtained from centrifugation experiments, an equal volume of each fraction was mixed with SDS-PAGE sample buffer (Pierce), and the proteins were resolved by SDS-PAGE (Novex) under denaturing conditions, followed by Western blotting. In some experiments, the low-density fractions (3 to 6) from a sucrose gradient were pooled, precipitated with methanol-chloroform, and solubilized in gel sample buffer. After transfer, the membranes were reacted with specific antibodies, washed, and then incubated with secondary antibodies coupled to HRP. Bound antibodies were revealed by enhanced chemiluminescence (ECL; Amersham) and exposure to film. For detection of the ganglioside GM1, 50 mul per fraction was blotted onto a nitrocellulose membrane, blocked with PBS containing 5% nonfat dry milk and 0.2% Tween, and incubated with CTB-HRP diluted in the same buffer. Bound proteins were revealed by ECL as before. RESULTS : Effect of cholesterol depletion or chelation on HSV entry. | One way to tell if lipid rafts are important for virus entry is to deplete cells of cholesterol with agents that sequester cholesterol, such as cyclodextrins, or agents that chelate cholesterol, such as statins, and then examine the effect of such treatment on virus entry. We treated Vero and B78-H1 cells with methyl-beta-cyclodextrin (MbetaCD) or nystatin for 30 min, washed out the drug, and then infected them with HSV-1 carrying the lacZ (KOS/tk12) gene under the control of the ICP4 promoter . At 6 h postinfection, the levels of beta-galactosidase activity in cell extracts were determined and used as a measure of virus entry . Virus entry into Vero cells was inhibited in a dose-dependent fashion, with 50% inhibition occurring at 7.5 mM MbetaCD and 80 mug of nystatin per ml . To determine whether this effect was dependent on a particular receptor, we repeated the experiment with cell lines bearing a single HSV receptor. A10 cells, which express HVEM, and C10 cells, which express nectin-1, were both derived by transfection of the mouse melanoma cell line B78-H1 . MbetaCD inhibited HSV entry into A10 and C10 cells in a dose-dependent fashion . This indicated that cholesterol was important for HSV entry with either the HVEM or nectin-1 receptor. Based on trypan blue dye exclusion or with actinomycin D in a flow cytometric assay, cells remained viable at drug concentrations as high as 30 mM MbetaCD or 200 mug of nystatin per ml (data not shown). One possible explanation for the inhibition of HSV entry by cholesterol-sequestering drugs is that residual drug present in or on the cells was toxic for the virus. To address this question, HSV-1 (KOS/tk12) virions were incubated with increasing concentrations of MbetaCD, then diluted 60-fold, and tested for infectivity on Vero cells . Treatment of virus with 7.5 mM MbetaCD inhibited its ability to infect cells by 90%. However, virus treated with 1 mM MbetaCD entered Vero cells almost as efficiently as control untreated virus. Since cells were washed extensively after MbetaCD treatment and before infection , it is unlikely that residual drug left on the cells was as high as 1 mM. So, although a high concentration of MbetaCD was toxic for HSV, this toxicity was not sufficient to explain the inhibition of entry into cells treated with MbetaCD. Another possibility is that MbetaCD inhibited virus attachment to cells. We treated C10 cells with 30 mM MbetaCD, washed them extensively, and added HSV-1 (KOS) for 1 h at 4C. Extracts of untreated and MbetaCD-treated cells contained the same amount of glycoprotein D (gD), as measured by SDS-PAGE and Western blotting . In addition, drug treatment did not alter the amount of nectin-1 detected in the same extracts . Surface expression of nectin-1 and HVEM (not shown) was not significantly affected by MbetaCD, as demonstrated by FACS analysis. Together, these observations rule out the possibility that MbetaCD inhibits HSV entry by either preventing virus binding to cells or changing the receptor levels present in the cells. We also evaluated the effect of cholesterol sequestration on HSV replication in a plaque formation assay. Vero cells were treated with MbetaCD for 30 min, the drug was washed out, and then cells were infected with HSV-1 (KOS). As a control, MbetaCD-treated cells were also infected with vesicular stomatitis virus, which is known to be raft independent . Plaques were stained 24 h later and counted . The number of HSV plaques was reduced in a dose-dependent manner, with 50% inhibition occurring at 10 mM MbetaCD. In contrast up to 30 muM MbetaCD did not reduce the number of vesicular stomatitis virus plaques made. These results underscore the importance of plasma membrane cholesterol content for efficient viral replication and support the observations made in the entry assays. In addition, we observed that the few plaques that formed in cells treated with MbetaCD were markedly smaller (not shown). However, plaque sizes were normal after longer periods of infection, and the reduction in plaque number in drug-treated cells was less severe (not shown). These observations suggest that cholesterol depletion imposes a delay in the onset of infection by HSV. Cholesterol replenishment reverses effect of MbetaCD. | To show that the inhibitory drug effects were specific, we replaced the cholesterol after sequestration and tested whether HSV entry was restored. We treated Vero cells with 7.5 mM MbetaCD for 30 min, a concentration that inhibited entry by 50% (Fig. , arrow). The drug was washed out, and increasing concentrations of cholesterol were then added to reconstitute plasma membrane cholesterol levels. Excess cholesterol was washed out, and the cells were infected with HSV-1 (KOS/tk12) . The addition of cholesterol completely reversed the inhibitory effect of MbetaCD in a dose-dependent fashion. We conclude that plasma membrane cholesterol is very important during HSV entry because sequestration or chelation of cholesterol prior to HSV infection inhibited entry, an effect reversed by the addition of exogenous cholesterol. Since rafts are maintained via cholesterol (reviewed in reference ), we wondered whether inhibition of HSV entry mediated by cholesterol-sequestering drugs was due to raft dispersion. Effect of cholesterol depletion after HSV entry. | To test whether cholesterol sequestration had an effect after entry but during replication, Vero cells were first infected with HSV-1 (KOS/tk12) and at various times postinfection treated for 30 min with 7.5 mM MbetaCD . MbetaCD markedly inhibited HSV replication when it was added at 30 min postinfection, but the level of inhibition decreased when the drug was added at later times. Normal entry was observed when MbetaCD was applied to cells between 2 to 3 h after infection. Interestingly, at any time following infection, MbetaCD-mediated inhibition of entry was not reversible by replenishment of cholesterol. These data suggest that rafts may be important during the early events of HSV replication, like fusion, but play no role later, once virus entry has occurred. They also provide further evidence that MbetaCD is not toxic for cells. Distribution of HSV receptors on the plasma membrane: are the viral receptors HVEM and nectin-1 found in rafts? | All these experiments suggested that cholesterol-rich rafts might play a role in HSV entry. One possibility is that HSV receptors are associated with rafts or redistribute into rafts as a consequence of binding to virus glycoproteins. To assess these possibilities, we used sucrose gradients to isolate low-density detergent-insoluble fractions from cells. B78 cells stably expressing HVEM-GFP or nectin-1 were extracted in Triton X-100 and analyzed for receptor distribution into low- and high-density fractions . HVEM-GFP was used because its larger molecular size allowed us to visualize it more easily than native HVEM by Western blotting. The ganglioside GM1 and the protein flotillin-2 are both associated with rafts and were used as markers. GM1 was mostly enriched in the low-density fractions (numbers 3 to 7), and a substantial portion of flotillin-2 was recovered in fractions 4 and 5. This is in agreement with other reports . Neither HVEM-GFP nor nectin-1 was found in the low-density sucrose fractions (numbers 1 to 6). These proteins were only recovered in the high-density fractions (numbers 7 to 12), where nonraft proteins are found. Similarly, when 293T cells were transiently transfected with plasmids expressing wild-type HVEM or nectin-1 , neither receptor was found in the GM1/flotillin-2-rich fractions. As a second approach, we used a more general procedure for two reasons. First, some proteins, such as prion protein, are enriched in semiordered lipid margins, located at the interface between the highly ordered lipid raft domains and the fluid glycerolipid membrane domain . These margin domains are solubilized in Triton X-100 but not in the milder detergent Brij-96. Second, due to strong interactions with the cytoskeleton, some raft-associated molecules, such as TRAF-3, only minimally associate with low-density fractions in sucrose gradients but are mainly recovered in the pellet of the gradient together with the cytoskeleton . Nectin-1 associates with afadin, which in turn associates with the cytoskeleton via actin . Thus, it was possible that one or both receptors might be minimally associated with rafts. A less stringent way to examine such interactions is to solubilize the membranes with a milder detergent, such as Brij-96, and then to use centrifugation to separate glycolipid domains (DIG) and the cytoskeleton. Then DIG which also contain rafts are solubilized and separated from the cytoskeleton with octylglycoside . Therefore, we lysed HVEM-GFP cells and C10 cells with Triton X-100 or Brij-96 and isolated the DIG fractions . In no case was HVEM found in fractions containing DIG , although a substantial amount of flotillin-2 and all of GM1 were detected in these fractions. These results are consistent with those obtained by sucrose gradient centrifugation. Similarly, nectin-1 was not detected in the DIG recovered from cells lysed with Triton X-100 (Fig. , top panel). However, a small but reproducible fraction of nectin-1 was found associated with DIG recovered from cells solubilized with Brij-96, suggesting that a small amount of the protein may associate with rafts (Fig. , bottom). However, the amount of nectin-1 found in the DIG fraction never exceeded the amount of the transferrin receptor found in this fraction (not shown). The transferrin receptor is a classical marker of a nonraft membrane protein . This suggests that nectin-1 detected in DIG did not come from a pool of molecules associated with rafts but rather was a contaminant of the fractionation procedure. We therefore conclude that HVEM and nectin-1 are not associated with lipid raft fractions in HVEM cells or nectin-1 cells regardless of the fractionation technique or the detergent used. Distribution of HSV glycoproteins on plasma membrane during entry. | We asked whether HSV receptors redistribute into lipid rafts after initial virus association or during entry. Manes et al. proposed this process for HIV gp120 and its receptors . Accordingly, C10 cells were infected with HSV and lysed with Triton X-100 at 0, 15, or 30 min postinfection. Lysates were subjected to sucrose gradient centrifugation . GM1 was detected with the low-density fractions 3 to 6 at all three times postinfection . To facilitate analysis, we chose a representative nonraft fraction (fraction 11) and compared this with pooled fractions 3 to 6 (Fig. and ). The raft marker flotillin-2 was found primarily in pooled fractions 3 to 6 . However, nectin-1 was excluded from these fractions and instead was located entirely in the nonraft fractions at 0, 15, and 30 min postinfection . During this time period, gD also was restricted to nonraft fractions , suggesting that any association of gD with nectin-1 at the time of or shortly after infection occurred outside lipid rafts. Similarly, gC and gH, two other glycoproteins involved in entry, were restricted to nonraft fractions . Interestingly, a substantial fraction of gB associated with rafts at 4C, and this proportion remained constant through the early steps of entry . However, when purified virions were fractionated on a sucrose gradient in the absence of cells, gB could not be detected in raft fractions (not shown). Similarly, gB was not selectively associated with DIG compared to gD and gH/gL when transfected in 293T cells (not shown). Thus, it is unlikely that gB interacts with rafts during the extraction procedure or during sedimentation. The association of gB with rafts requires the initial binding of the virus to a cell surface. Distribution of soluble HSV glycoproteins after association with plasma membrane. | These studies suggested that gB interacts specifically with lipid rafts during virus entry. Because this association was observed at time zero, it was possible that the ectodomain of virion gB might associate with a raft-associated cell molecule. To test this, we incubated cells with purified ectodomains of gB, gC, or gD and evaluated the association of the different glycoproteins with the DIG fraction. Also, to mimic the situation of virus entry, the proteins were added to the cells at 4C and the temperature was shifted to 37C for various times . At 4C and time zero, soluble gB was already associated with the DIG fraction (Fig. , 0 min). Interestingly, the apparent amount and proportion of gB in the DIG fraction increased during incubation at 37C for up to 1 h (Fig. , DIG). During the same period, the proportion of the full-size gB ectodomain found in the DSM was constant, although there was an accumulation of degradation products over time (Fig. , DSM). By contrast, gC and gD (Fig. , lower panel) were almost excluded from DIG at all time points. The low levels of gD found in DIG are consistent with the low levels of nectin-1 found in this fraction (Fig. , top panel). The fact that gC was also found exclusively in the DSM is important because it, like gB, associates with heparan sulfate on cells . Thus, it is unlikely that the distribution of gB into DIG is related to heparan sulfate binding. If so, gB but not gC would associate with a specific type of HSPG enriched in DIG, or gB and gC could have differential effects when bound to HSPG. The association of soluble gB with DIG corroborates our observations made with virion gB. Together, the data suggest that the interaction of gB with rafts does not depend on its transmembrane domain or carboxy terminus but is mediated by the ectodomain. DISCUSSION : Requirement of cholesterol for HSV entry. | The entry of HSV into cells is a complex process that is not completely understood. It involves four viral glycoproteins, gB, gD, and gH/gL, and at least one of the gD receptors (reviewed in references and ). Lipid rafts represent a privileged microdomain that may facilitate virus entry (reviewed in references , , , , , and ). We found that sequestration of cholesterol, an essential constituent of rafts, inhibits entry into cells expressing either HVEM or nectin-1. Entry was restored by replenishing the cells with cholesterol, confirming the requirement for cholesterol. Since cholesterol is an essential structural molecule of rafts , we hypothesize that rafts may be directly involved during entry. HSV entry may also require the presence of cholesterol in the plasma membrane independently of rafts. In particular, fusion may require cholesterol regardless of the maintenance of rafts. Fusion of alphaviruses requires the presence of both cholesterol and sphingolipids in target membrane, but the fusion activity is not dependent on the presence of DIG . Similarly, depleting plasma membrane cholesterol inhibits HIV-1 entry, but the presence of HIV-1 receptors in rafts is not required for infection . These observations suggest that cholesterol modulates the HIV-1 entry process independently of its ability to promote raft formation. Cholesterol is, however, essential for the conformation and function of CXCR4 and CCR5 . Alternatively, it was proposed that binding of HIV may be favored by the presence of CD4 in rafts, but rafts may then disperse prior to the membrane fusion reaction . Since gB was found in lipid rafts during attachment of HSV, this suggests a function for these structures during entry, in particular for fusion. HVEM and nectin-1 are not associated with lipid rafts. | HIV infection induces lateral membrane diffusion following interaction of the viral envelope with cell surface receptors, which are necessary for infection . We found that HVEM was not associated with rafts in uninfected cells regardless of which detergent was used for extraction. This was a surprise, because other tumor necrosis factor receptors such as CD40, CD120a, and the p75 neurotrophin receptor are associated with rafts . The cytosolic tail of CD120a containing the death domain is necessary and sufficient for both localization of the receptor to lipid rafts and signaling apoptosis. This implies that receptor localization to lipid rafts is important for the function of this receptor . The absence of a death domain in HVEM may explain its particular membrane distribution . On the other hand, CD40 engagement by MAbs leads to a membrane raft-restricted recruitment of TRAF-3 and TRAF-2 to CD40's cytoplasmic tail. This indicates that the membrane raft structure plays an integral role in CD40 signaling . A similar domain is likely important for HVEM function. However, binding of gD did not result in redistribution of HVEM into rafts. Moreover, deletion of part of the cytosolic tail of HVEM comprising the TRAF-2 and TRAF-5 binding sites required for NF-kappaB activation had no effect on its function as an HSV receptor . Based on the model proposed for the T- and B-cell receptors and the Fcepsilon receptor , we hypothesized that the immunoglobulin family member nectin-1 would also associate with rafts during entry. Nectin-1 is located at tight junctions , and tight junctions are found in membrane microdomains . However, we found that nectin-1 was not associated with lipid rafts in uninfected cells and that this distribution did not change during infection. If rafts are involved in HSV entry, it is not because of the localization of the gD receptors in these structures. Localization of nectin-1 in tight junctions and rafts in vivo should be confirmed. gB associates with lipid rafts during entry. | The interesting finding that gB but not gD, gC, or gH associated with rafts after virus attachment and during entry may point to a unique function for gB during HSV entry that involves association with lipid rafts. We hypothesize that a cellular gB receptor(s) is enriched in rafts, a situation that would explain the unique association of gB with microdomains during entry . Localization of this unknown molecule in rafts would also account for the inhibition of virus entry by MbetaCD and nystatin. Rafts are likely to be less than 70 nm in diameter . Thus, the envelope of HSV (which is 150 to 200 nm in diameter) could potentially interact simultaneously with both raft and nonraft domains during attachment. This would explain how only gB is found associated with rafts . Among the herpesviruses, gB is the most highly conserved glycoprotein, suggesting that this molecule possesses a central function which is common for the entry of different herpes viruses . It is of interest that cytomegalovirus gB activates the interferon response pathway via an interaction with an unidentified cellular receptor different from HSPG . Activation of these genes involves the mitogen-activated protein kinase pathway and protein kinase C. Liquid-ordered domains represent the preferential location for the initiation of the mitogen-activated protein kinase as well as other signaling cascades (reviewed in reference ). Thus, the potential cytomegalovirus gB receptor may also localize in rafts, where it elicits signaling upon cytomegalovirus attachment. The association of HSV gB with lipid rafts during entry may be necessary to elicit rapid signaling events. Transient tyrosine phosphorylation of several proteins occurs within minutes after infection, followed by activation of NF-kappaB. Both of these events support the idea that signaling cascades parallel HSV entry (data not shown) . The direct involvement of HSV gB in signaling remains to be investigated. The role of rafts during HSV infection has been investigated by others. Sucrose gradient fractionation suggested that the UL56 gene product is a membrane protein associated with rafts . A minor fraction of the tegument protein vhs as well as VP16 and gH localizes to DIG in HSV-infected cells . We found that gH as well as gB, gC, and gD was excluded from DIG in the virion envelope (data not shown). Similarly, these glycoproteins were not in DIG from either cells infected with HSV for 8 h or cells transfected with the individual glycoproteins (data not shown). We suspect that the traces detected in rafts are likely a consequence of contamination. Another interesting observation was that treatment of herpes simplex virions with a cholesterol-sequestering drug inhibited entry. Replenishment with cholesterol partially restored entry. Thus, cholesterol may play an important role in determining the structure of the viral envelope. Whether cholesterol is important for the organization of rafts on the virion is not known. However, rafts are found in the Golgi complex , a cellular compartment important for the final envelopment of virus (reviewed in reference ). Thus, it is possible that some rafts may be incorporated in the viral envelope. Similar experiments with HIV showed that MbetaCD-treated virus was less infectious . The association of gB with rafts during entry could represent a step in our understanding of fusion. Semliki Forest virus fusion protein E1 inserts preferentially in membrane domains enriched in cholesterol and sphingolipid . By analogy, HSV gB may preferentially bind to a raft-associated molecule and play a direct role in fusion through these microdomains. The other glycoproteins would contribute through strong and specific interactions with cognate receptors, bringing the viral envelope and plasma membrane into close apposition. Alternatively, conformational changes in the entry glycoproteins induced by binding to receptors could result in the formation of a fusion complex which includes the association of gB with rafts. Identification of the gB binding molecules as well as characterization of the individual involvement of each glycoprotein during this process are essential to answer these questions. FIG. 1. : Effect of cholesterol depletion or chelation on HSV entry. Effect of cholesterol depletion or chelation on HSV entry. Vero cells (A and B) and B78 A10 and C10 cells (C) were treated with increasing concentrations of MbetaCD (A and C) or nystatin (B). The drugs were washed out, and the cells were infected with HSV-1 (KOS/tk12) and assayed for beta-galactosidase activity at 6 h postinfection. Results presented in A and B are the mean of three independent experiments done in duplicate, with standard deviations. The results shown in panel C are representative of three experiments. A value of 100% entry represents the beta-galactosidase activity at 6 h in the absence of MbetaCD. FIG. 2. : Effect of cholesterol depletion from virions on HSV entry. Effect of cholesterol depletion from virions on HSV entry. HSV-1 (KOS/tk12) particles (2 x 106 PFU) were treated with various amounts of MbetaCD for 30 min at room temperature. MbetaCD was then diluted to a final concentration of 0.1 mM or less, a concentration that has no effect on cells, and virus was allowed to infect Vero cells (multiplicity of infection of 10). beta-Galactosidase activity at 6 h postinfection was determined as described for Fig. . The results presented are representative of two independent experiments done in triplicate, with standard deviations. FIG. 3. : Effect of cholesterol depletion from cells on expression of nectin-1 and on attachment and replication of HSV. Effect of cholesterol depletion from cells on expression of nectin-1 and on attachment and replication of HSV. (A) C10 cells were untreated (-) or treated (+) with 30 mM MbetaCD for 30 min. The drug was washed out, and HSV-1 (KOS) was allowed to bind to the cells for 1 h at 4C. Total cell proteins were extracted, resolved by SDS-PAGE, and probed with anti-nectin-1 MAb CK6 (upper panel) or anti-gD MAb DL6 (lower panel). (B) C10 cells were treated with increasing concentrations of MbetaCD (indicated at the right), then stained with anti-nectin-1 MAb CK41 directly labeled with phycoerythrin, and analyzed by FACS. The control (open area) represents the fluorescence of unstained and drug untreated cells. (C) Vero cells were treated with increasing concentrations of MbetaCD. The drug was washed out, and the cells were infected with HSV1 (KOS) or vesicular stomatitis virus (VSV) for 24 h at 37C. Then the cells were fixed, and plaques werevisualized by immunoperoxidase staining (HSV) or crystal violet staining (vesicular stomatitis virus). FIG. 4. : Effect of cholesterol replenishment on HSV entry. Effect of cholesterol replenishment on HSV entry. Vero cells were untreated (left, rectangles) or treated (right, rectangles) with 7.5 mM MbetaCD for 30 min. MbetaCD was washed out, and various amounts of cholesterol in MbetaCD were added. After 30 min, the cholesterol was removed, and the cells were infected with HSV-1 (KOS/tk12) and assayed for beta-galactosidase activity at 6 h postinfection. The results presented are representative of two independent experiments done in triplicate. FIG. 5. : Effect of cholesterol depletion from cells on the course of HSV entry. Effect of cholesterol depletion from cells on the course of HSV entry. Vero cells were infected with HSV-1 (KOS/tk12). At various times following infection (indicated on the abscissa), cells were treated for 30 min with MbetaCD. The drug was washed out, and medium or cholesterol (chol) was added. After 30 min, the cholesterol was removed, and infection was allowed to proceed. beta-Galactosidase activity at 6 h postinfection was determined. The results shown are representative of three independent experiments, and 100% entry represents the beta-galactosidase activity at 6 h in the absence of MbetaCD. FIG. 6. : Association of HSV receptors with detergent-insoluble low-density sucrose fractions containing rafts. Association of HSV receptors with detergent-insoluble low-density sucrose fractions containing rafts. 3E5 cells stably expressing HVEM-GFP (A), C10 cells stably expressing nectin-1 (B), and 293T cells transiently transfected with a plasmid expressing the HVEM (C) or nectin-1 (D) gene were homogenized in Triton X-100 and fractionated on sucrose gradients. Proteins (10 mul of each fraction) were resolved on SDS-PAGE and transferred to nitrocellulose, and the distribution of HSV receptors was analyzed by Western blotting with MAb CW10 to HVEM or MAb CK6 to nectin-1. Fraction numbers are indicated at the top of the figure (P for pellet). T and B refer to the top and bottom of the gradient, respectively. The distribution of the known raft-associated protein flotillin-2 (flot-2) was analyzed with a commercial MAb and served as a positive control for rafts (A, B, C, and D, middle panels). Also shown is the distribution of the raft-specific ganglioside GM1 analyzed by dot blot with CTB-HRP (A, B, C, and D, lower panels). Sizes are shown on the left (in kilodaltons). FIG. 7. : Association of HSV receptors with detergent-insoluble glycolipid complexes. Association of HSV receptors with detergent-insoluble glycolipid complexes. 3E5 cells (A), C10 cells (B), and control parental B78-H1 cells were fractionated into detergent-soluble (DSM) and insoluble (DIG) membrane fractions with either Triton X-100 (TX-100) or Brij-96. The distribution of the two HSV receptors was analyzed by Western blotting as described for Fig. . At the top are indicated the receptor expressed (control, HVEM, or nectin-1), the detergent used (Triton X-100 or Brij-96), and the fraction, DSM (M) or DIG (D). Sizes are shown on the left (in kilodaltons). FIG. 8. : Association of HSV receptors and glycoproteins with detergent-insoluble low-density sucrose fractions during entry. Association of HSV receptors and glycoproteins with detergent-insoluble low-density sucrose fractions during entry. C10 cells were infected with HSV (KOS), and at various times postinfection, cells were extracted in Triton X-100 and fractionated on a sucrose gradient as in Fig. . (A) The distribution of GM1 in the different sucrose gradient fractions was analyzed by dot blot in noninfected cells (NV) and after infection for various times, indicated at the right of the panel. (B) Low-density fractions 3 to 6 containing the raft marker GM1 (raft) were pooled, precipitated with methanol-chloroform, resuspended in gel sample buffer, and resolved by SDS-PAGE. The distribution of flotillin-2 (flot-2) and nectin-1 at various times during infection (indicated at the top of the panel) was compared to that found in 10 mul of the GM1-negative fraction 11 (non-raft) by Western blotting on separate membranes. The distribution of the glycoproteins (indicated at the right) in raft and nonraft fractions was analyzed with MAb DL6 to gD, polyclonal antibody R47 to gC, polyclonal antibody R137 to gH, and MAb SS-10 to gB. Sizes are shown on the left (in kilodaltons). FIG. 9. : Association of HSV soluble glycoproteins with detergent-insoluble glycolipid complexes. Association of HSV soluble glycoproteins with detergent-insoluble glycolipid complexes. (A) Soluble gB, (B) gC, or (C) gD (0.1 muM each) was bound to C10 cells at 4C, and the temperature was shifted at 37C for various times (in minutes), as indicated at the top of the panel. Cells were extracted in 1% Brij-96, and the distribution of each glycoprotein into DSM and DIG was analyzed as described for Fig. . For each experiment, the control (c) without glycoproteins (A and B) or on parental nectin-1-negative B78-H1 cells (C) is shown. Panel C also shows the distribution of nectin-1. Sizes are shown on the left (in kilodaltons). FIG. 10. : Model of HSV attachment involving lipid rafts. Model of HSV attachment involving lipid rafts. Initial attachment of HSV is mediated through the interaction of gB and/or gC with HSPG. These interactions are likely to occur in nonraft domains, since gC and a fraction of gB are found in DSM. Then a specific interaction of gD with a cognate receptor (here, nectin-1) is essential for fusion to occur. Both gD and nectin-1 are excluded from rafts during attachment and throughout the entry process. However, a fraction of gB is associated with rafts from the moment of attachment and during entry. Such a distribution can be explained by the interaction of gB with an unknown receptor enriched in rafts. The size of the HSV virion (150 to 200 nm) would allow its envelope to interact with a plasma membrane region that can encompass at least one raft microdomain (70 nm). Backmatter: PMID- 12915559 TI - The Moloney Murine Leukemia Virus Repressor Binding Site Represses Expression in Murine and Human Hematopoietic Stem Cells AB - The Moloney murine leukemia virus (MLV) repressor binding site (RBS) is a major determinant of restricted expression of MLV in undifferentiated mouse embryonic stem (ES) cells and mouse embryonal carcinoma (EC) lines. We show here that the RBS repressed expression when placed outside of its normal MLV genome context in a self-inactivating (SIN) lentiviral vector. In the lentiviral vector genome context, the RBS repressed expression of a modified MLV long terminal repeat (MNDU3) promoter, a simian virus 40 promoter, and three cellular promoters: ubiquitin C, mPGK, and hEF-1a. In addition to repressing expression in undifferentiated ES and EC cell lines, we show that the RBS substantially repressed expression in primary mouse embryonic fibroblasts, primary mouse bone marrow stromal cells, whole mouse bone marrow and its differentiated progeny after bone marrow transplant, and several mouse hematopoietic cell lines. Using an electrophoretic mobility shift assay, we show that binding factor A, the trans-acting factor proposed to convey repression by its interaction with the RBS, is present in the nuclear extracts of all mouse cells we analyzed where expression was repressed by the RBS. In addition, we show that the RBS partially repressed expression in the human hematopoietic cell line DU.528 and primary human CD34+ CD38- hematopoietic cells isolated from umbilical cord blood. These findings suggest that retroviral vectors carrying the RBS are subjected to high rates of repression in murine and human cells and that MLV vectors with primer binding site substitutions that remove the RBS may yield more-effective gene expression. Keywords: Introduction : Retroviral vectors based upon the Moloney murine leukemia virus (MLV) have been used widely due to their high efficiency of stable gene transfer. Transcription from the MLV long terminal repeat (LTR) is severely repressed in mouse embryonal carcinoma (EC) cells and mouse embryonal stem (ES) cells . Characteristics of MLV that have been shown to mediate poor expression in undifferentiated ES and EC cell lines include the inadequate function of the enhancer in the 5'LTR due to the lack of transcriptional activator binding sites , the negative control region (NCR) located in the U3 region of the LTR , and the repressor binding site (RBS) . The RBS is an 18-bp DNA element located downstream of the 5'LTR which directly overlaps the MLV primer binding site (PBS) by 17 of its 18 bp . The MLV PBS functions by binding a proline tRNA molecule that primes reverse transcription. The existence of the overlapping RBS was first determined due to a spontaneously occurring single-base G-to-A mutation that occurred within the PBS sequence that allowed expression by MLV in F9 EC cells . This single base mutation was called the B2 mutation . Like the MLV PBS, the B2 PBS binds to a proline tRNA molecule, but with a single base-pair mismatch . Several other mutations that have been made within the MLV PBS have also been shown to relieve the repression conveyed by the RBS element . Due to the requirement that vectors derived from MLV must contain a functional PBS to initiate reverse transcription, the overlapping RBS sequence cannot be removed simply by removing the MLV PBS. In order to eliminate this repressive element and still have a functional vector, several groups have replaced the MLV PBS with the PBS from dl587rev, a recombinant between MLV and an endogenous mouse retrovirus . The dl587rev PBS sequence contains the B2 mutation plus five additional base pairs different from the MLV PBS and is a perfect match for the glutamine tRNA . The RBS has been shown to repress transcription from the MLV LTR within the context of an MLV genome when located away from its normal position, in either orientation, upstream of the MLV transcription start site or when positioned downstream within an intron . In addition to repressing transcription from the MLV LTR, the RBS was shown to substantially repress transcription from two internal heterologous viral promoters, simian virus 40 (SV40) and adenovirus major late promoter, which were inserted downstream of the RBS into an MLV vector with an enhancerless LTR . The mechanism by which the RBS represses transcription is not known. The RBS is thought to function by interacting with an unknown trans-acting factor. In support of this hypothesis, the repressive activity of the RBS was demonstrated to be saturable by transfection of increasing amounts of DNA containing the RBS sequence, suggesting that the repressive activity is mediated by a trans-acting factor or factors . With an exonuclease III protection assay, a factor from PC13 EC cell extracts was shown to bind a MLV PBS probe, but not a B2 PBS probe . Using an electrophoretic mobility shift assay (EMSA), a predominantly nuclear protein in F9 EC cell nuclear extracts was identified that bound to an MLV PBS probe, but not a B2 PBS probe . Several studies have described the repressive activity of the RBS to be stem cell specific , because it has been shown to have repressive activity in several murine EC cell lines (F9, PC13, and PCC4) and D3 ES cells, but not in 3T3 embryonic fibroblasts. Retroviral vectors with multiple modifications, including the dl587rev PBS, have been demonstrated to be superior to vectors containing only MLV components , but the specific role of the PBS change has not been analyzed. We evaluated whether the RBS alone was sufficient to cause repression by incorporating this element into a lentiviral vector derived from human immunodeficiency virus type 1 (HIV-1). We also wanted to determine if the repressive activity mediated by the RBS present in ES and EC cell lines was also present in murine hematopoietic stem cells and more differentiated cells and if the repressive activity mediated by the RBS was present in human cells. Because some of these cell types are inefficiently transduced by MLV-based retroviral vectors, we used a series of HIV-1-based lentiviral vectors into which the MLV, B2, or the dl587 PBS sequences were inserted downstream from various internal promoters. We observed that the PBS sequences acted in the context of the lentiviral vector genome as they do within the MLV vector genome context in that the MLV PBS sequence, containing the overlapping RBS, repressed expression from several internal promoters, while the B2 and dl587 PBS sequences did not. Repression specific to the RBS was also documented in a variety of murine and human cells, including hematopoietic stem and progenitor cells. MATERIALS AND METHODS : Cell culture. | The following cell lines were obtained from the American Type Culture Collection (ATCC) and cultured according to their recommendations: K562 (ATCC CCL-243), U937 (ATCC CRL-1593.2), PA-1 (ATCC CRL-1572), Tera-2 (ATCC HTB-106), NCCIT (ATCC CRL-2073), 70Z/3 (ATCC TIB-158), STO-SNL/2 (ATCC CRL-2225), ES-D3 (ATCC CRL-1934), Jurkat (ATCC TIB-152), KG1a (ATCC CCL-246.1), AMJ2-C11 (ATCC CRL-2456), F9 EC (ATCC CRL-1720), NIH/3T3 (ATCC CRL-1658), and CCRF-CEM (ATCC CCL-119). DU.528 were obtained from Joanne Kurtzberg (Duke University Medical Center, Durham, N.C.) and cultured in Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 10% fetal bovine serum (FBS), 10% horse serum (HS), 2 mM l-glutamine, 0.1 mM sodium pyruvate, 100 U of penicillin/ml, and 100 mug of streptomycin/ml. BM185 cells have been previously described and were cultured in RPMI medium supplemented with 5% FBS, 0.01 mM 2-mercaptoethanol (Sigma, St. Louis, Mo.), 100 U of penicillin/ml, and 100 mug of streptomycin/ml. WTc.F cells (ES cells derived from C57Bl/6 mice) , a gift of Andrew Kung (Harvard Medical School, Boston, Mass.), were cultured on irradiated STO-SNL/2 feeder layers in Dulbecco's modified Eagle's medium (DMEM) supplemented with 15% FBS, 4 mM l-glutamine, 0.1 mM nonessential amino acids, 0.1 mM 2-mercaptoethanol, 100 U of penicillin/ml, and 100 mug of streptomycin/ml. FDCP-Mix cl.A4 was obtained from Lez Fairbairn (Paterson Institute for Cancer Research, Manchester, United Kingdom) and cultured in Iscove's modification of Dulbecco's medium (IMDM) supplemented with 20% HS and 10 ng of murine interleukin-3 (IL-3; Biosource International, Camarillo, Calif.)/ml. Mouse embryonic fibroblasts (MEFs) were isolated by trypsinization of embryos dissected at 13.5 days of gestation from outbred CF-1 mice (Charles River Laboratories, Wilmington, Mass.). Each embryo was harvested separately, the brain and internal organs were removed, and the carcasses were minced and incubated in 0.05% trypsin for 30 to 45 min at 37C. Single-cell suspensions obtained after trypsinization were plated in 10-cm dishes in DMEM supplemented with 10% FBS, 100 U of penicillin/ml, 100 mug of streptomycin/ml, and 2 mM l-glutamine. Experiments with MEFs were performed on early passages (less than five). Mouse bone marrow (BM) cells were isolated by flushing bone marrow from adult mouse femurs with a 27[1/2]-gauge needle. Bone marrow stromal cell populations were isolated by breaking the femurs into small pieces and culturing them with isolated marrow cells. Cells were allowed to adhere to tissue culture plates for 4 to 6 days in IMDM containing 20% HS, 20% FBS, 2 mM l-glutamine, 100 U of penicillin/ml, 100 mug of streptomycin/ml, 0.1 mM 2-mercaptoethanol, and 6 muM hydrocortisone. Adherent cells were rinsed 5 to 7 times with phosphate-buffered saline (Irvine Scientific, Santa Ana, Calif.) every day for 10 to 15 days to remove nonadherent cells. Adherent cells were allowed to expand to confluency for an additional 2 to 3 weeks. Remaining CD45+ hematopoietic cells were removed with a magnetic column after staining with rat anti-mouse CD45 antibody and magnetic bead-conjugated goat anti-rat immunoglobulin G. CD45- cells were placed back into culture and later verified to be CD45- by flow cytometry. CD34+ cells were isolated from human umbilical cord blood obtained from normal deliveries, using Miltenyi MiniMACS magnetic separation columns (Miltenyi Biotech, Sunnyvale, Calif.) after Ficoll-Hypaque (Amersham Pharmacia Biotech, Piscataway, N.J.) density gradient centrifugation. Use of these cord blood samples was approved by the Committee on Clinical Investigations at Childrens Hospital, Los Angeles, Calif. To isolate CD34+ CD38- cells, CD34+ cells were washed in phosphate-buffered saline and incubated for 30 min at 4C in fluorescein isothiocyanate-CD34 (HPCA2; Becton Dickinson Immunocytometry Systems, San Jose, Calif.) and phycoerythrin-CD38 (leu 17; Becton Dickinson Immunocytometry Systems). CD34+ CD38- cells were then isolated by fluorescence-activated cell sorting (FACS), using the gating previously described , on a FACSVantage flow cytometer (Becton Dickinson Immunocytometry Systems) and using LysysII software (Becton Dickinson Immunocytometry Systems). Vector construction. | The pL-eGFP-SN vector was constructed by inserting a BglII-NotI fragment containing the enhanced green fluorescent protein (eGFP) gene (Clontech Laboratories, Palo Alto, Calif.) into the HpaI site of pL-X-SN as previously described . Modifications made to pL-eGFP-SN to generate pLD-eGFP-SN, pM-eGFP-SN, pMD-eGFP-SN, and pMND-eGFP-SN were described previously . Self-inactivating (SIN) lentiviral vectors containing the U3 region of the MND LTR (MNDU3; the U3 region from the myeloproliferative sarcoma virus [MPSV] LTR with the NCR removed) as an internal promoter with each of the PBS sequences were constructed as follows. pCCL-hCMV-eGFP (kindly provided by Luigi Naldini, Cell Genesys, Foster City, Calif.) was digested with ClaI and SalI to remove a fragment containing hCMV-eGFP to generate pCCL-X. pMND-Neo was digested with ClaI and Asc1 to isolate a fragment containing the MNDU3 promoter. The MNDU3 enhancer-promoter fragment was blunted and ligated into the EcoR5 site of pIC-20H to generate pMNDU3-20H. pMNDU3-20H was digested with BamHI and BglII to isolate a fragment containing the MNDU3 enhancer-promoter. The MNDU3 enhancer-promoter fragment was inserted into the BamHI site of pCCL-X to generate pCCL-MNDU3-X. pMND-eGFP-SN was digested with XhoI to release a fragment containing eGFP. The eGFP fragment was ligated into the XhoI site of pCCL-MNDU3-X to generate pCCL-MNDU3-eGFP. The MLV PBS was inserted into the SalI site downstream of the MNDU3 promoter using 5'-phosphorylated oligonucleotides which anneal to create the PBS (underlined) flanked by SalI sites and a unique MluI site used to verify insertion: 5'-TCGACACGCGTGGGGGCTCGTCCGGGATCGGGAGACCCCG-3' and 5'-TCGACGGGGGCTCCCGATCCCGGACGAGCCCCCACGCGTG-3'. The B2 PBS was inserted into the SalI site downstream of the MNDU3 promoter using 5'-phosphorylated oligonucleotides which anneal to create the PBS flanked by SalI sites and a unique MluI site used to verify insertion: 5'-TCGACACGCGTGGGGGCTCGTCCGAGATCGGGAGACCCCG-3' and 5'-TCGACGGGGGCTCCCGATCTCGGACGAGCCCCCACGCGTG-3'. The dl587 PBS was inserted into the SalI site downstream of the MNDU3 promoter using the following 5'-phosphorylated oligonucleotides which anneal to create the PBS flanked by SalI sites and a unique MluI site used to verify insertion: 5'-TCGACACGCGTGGAGGTTCCACCGAGATTTGGAGACCCCG-3' and 5'-TCGACGGGGTCTCCAAATCTCGGTGGAACCTCCACGCGTG-3'. SIN lentiviral vectors containing different internal promoters were constructed using pCCL-hCMV-eGFP. The human cytomegalovirus (hCMV) internal promoter in this vector was removed with ClaI and BamHI and replaced with the human elongation factor 1a (hEF-1a) promoter, which was acquired from pV4.1e-hF.IX (kindly provided by Hiroyuki Nakai, Stanford University, Stanford, Calif.) by PCR with Pfu Turbo polymerase (Stratagene, La Jolla, Calif.) using the primers 5'-GAAGATCGATCGTGAGGCTCCGGTG-3' and 5'-GGTAGGATCCACGACACCTGAAATG-3', followed by digestion with ClaI and BamHI. To replace the hCMV promoter in pCCL-hCMV-eGFP with the human ubiquitin C (hUbiqC) promoter, the plasmid was first digested with ClaI, blunt ended with Pfu Turbo polymerase, and then digested with BamHI. The hUbiqC promoter fragment was prepared from pFUGW (kindly provided by Carlos Lois, California Institute of Technology, Pasadena, Calif.) by digesting with PacI, then blunt ending with T4 polymerase (Invitrogen, Carlsbad, Calif.), followed by digestion with BamHI. To replace the hCMV promoter in pCCL-hCMV-eGFP with the mouse phosphoglycerate kinase (mPGK) promoter, the hCMV promoter was removed by digestion with ClaI and AgeI, and an oligonucleotide containing restriction sites ClaI-EcoRV-BamHI-AgeI was inserted. The mPGK promoter was excised from pic20H-mPGK with HincII and BglII and inserted into EcoR5 and BamHI. To replace the hCMV promoter in pCCL-hCMV-eGFP with the SV40 promoter, the hCMV-eGFP cassette was removed by digestion with ClaI and SalI and an oligonucleotide containing restriction sites ClaI-XhoI-EcoRV-BamHI-SmaI-SalI was inserted. An eGFP fragment was excised from pMND-eGFP-SN with BamHI and inserted into this multicloning site. The SV40 promoter was excised from pL-X-SN with XhoI and StuI and inserted into this multicloning site after digestion with XhoI and EcoRV. The MLV PBS was inserted into a BamHI site downstream of the hEF-1a and hUbiqC promoters and upstream of the mPGK promoter using 5'-phosphorylated oligonucleotides that anneal to create the PBS flanked by BamHI sites and a unique NheI site used to verify insertion: 5'-GATCCGCTAGCGGGGGCTCGTCCGGGATCGGGAGACG-3' and 5'-GATCCGTCTCCCGATCCCGGACGAGCCCCCGCTAGCG-3'. The MLV PBS was inserted into an XhoI site upstream of the SV40 promoter using 5'-phosphorylated oligonucleotides that anneal to create the PBS flanked by XhoI sites and a unique NheI site used to verify insertion: 5'-TCGAGGCTAGCGGGGGCTCGTCCGGGATCGGGAGAC-3' and 5'-TCGAGTCTCCCGATCCCGGACGAGCCCCCGCTAGCC-3'. The MLV PBS was inserted into an AgeI site downstream of the SV40 promoter using 5'-phosphorylated oligonucleotides that anneal to create the PBS flanked by AgeI sites and a unique NheI site used to verify insertion: 5'-CCGGTGCTAGCGGGGGCTCGTCCGGGATCGGGAGAA-3' and 5'-CCGGTTCTCCCGATCCCGGACGAGCCCCCGCTAGCA-3'. Vector supernatant production. | MLV vector supernatants were generated by a stably transduced GP+E-86 packaging cell line . Cells were plated close to confluency in T75 flasks, grown at 37C for 48 h in 10 ml of DMEM supplemented with 10% FBS, 100 U of penicillin/ml, 100 mug of streptomycin/ml, 2 mM l-glutamine. Vector supernatants were filtered through a 0.45-mum filter and frozen at -80C until used. Vesicular stomatitis virus glycoprotein-pseudotyped lentiviral vector supernatants were generated by transient transfection of 293T cells (ATCC CRL-1268) as previously described using 10 mug of vector plasmid, 10 mug of pRDelta8.9 packaging plasmid , and 2 mug of pMD.G(VSV) envelope plasmid . Twelve hours after transfection, cells were treated with 10 mM sodium butyrate (Sigma Scientific, Inc., Brighton, Mich.) for 12 h as previously described . After 12 h of exposure to sodium butyrate, the cells were washed twice with phosphate-buffered saline and refed with fresh medium. Thereafter, supernatants were collected every 12 h for 3 to 5 days, filtered through a 0.2-mum filter flask (Nalgene, Rochester, N.Y.), and concentrated by ultracentrifugation at 50,000 x g for 140 min as previously described . Pellets were resuspended in serum-free DMEM and stored at -80C until used. Vector supernatant titer determination. | Vector supernatant titers were determined by endpoint dilution. 293 cells (ATCC CRL-1573) were seeded at 105 cells/well in six-well cell culture plates (Corning Inc., Miami, Fla.) in DMEM supplemented with 10% FBS, 100 U of penicillin/ml, 100 mug of streptomycin/ml, 2 mM l-glutamine and placed in a 37C incubator for 12 h. Cells were then transduced with 1-ml serial dilutions (i.e., 10-1, 10-2, 10-3) of vector supernatant and analyzed by flow cytometry for eGFP expression 48 h later. Titers were calculated by multiplying the number of cells at the time of vector supernatant addition by the percentage of eGFP-positive cells determined by flow cytometry divided by 100, multiplied by the dilution factor to yield the number of infectious units (IU) per milliliter. Titers ranged between 0.5 x 106 and 10 x 106 IU/ml before ultracentrifugation and 0.5 x 108 and 5 x 108 IU/ml after ultracentrifugation. Lentiviral vector transductions. | Target cells that were transduced to determine whether they possessed RBS-mediated repressive activity were transduced in parallel and under the same conditions as F9 EC and 293 cells in order to control for small differences in vector titer that could contribute to differences in transduction efficiency. Most cell types were transduced for 12 h using a final vector concentration of 1 x 106 to 2 x 106 IU/ml and a multiplicity of infection of 5 to 10 in their normal growth medium. CD34+, CD34+ CD38-, and primary mouse BM cells required higher concentrations of vector for efficient transduction and were transduced with a final vector concentration of 5 x 107 to 10 x 107 IU/ml plus Polybrene at 8 mug/ml. After transduction, cells were passaged in culture for 6 to 10 days and then analyzed by flow cytometry for eGFP expression. PCR for relative copy number. | Genomic DNA was isolated using a DNeasy tissue kit (Qiagen, Valencia, Calif.) for use as template in a semiquantitative PCR to determine the relative vector copy number in transduced cells. To generate the standard curve, we used a GP+E-86 cell clone containing five copies of the MND-eGFP-SN vector, as previously described . The standard curve was generated using dilutions representing 4, 2, 1, 0.5, 0.25, 0.13, 0.06, and 0.03 copies/cell. DNA of the five-copy clone was mixed with DNA of nontransduced GP+E-86 cells, so that the total input template DNA was maintained constant. Template DNA was diluted to 100 ng in 44 mul of double-distilled (ddH2O) in a PCR tube. Three microliters of this solution, containing 7 ng of template DNA, was removed and placed into a second tube, leaving 41 mul containing 93 ng of template DNA in the first tube. Master mix solutions were then added to make a final volume of 50 mul containing 2.5 U of Pfu Turbo polymerase, 1x Pfu PCR buffer (Stratagene), each primer at a 0.2 muM concentration, and each deoxynucleoside triphosphate at a 0.2 mM concentration. The tube containing 93 ng of template DNA was used for eGFP PCR, and the tube containing 7 ng of template DNA was used for beta-actin PCR to control for varying input template content. eGFP primer sequences used were sense, 5'-ATGGTGAGCAAGGGCGAGGAGCTG-3', and antisense, 5'-GCCGTCGTCCTTGAAGAAGATGGTG-3', yielding a product of 314 bp. beta-Actin primer sequences used were sense, 5'-GTACCACAGGCATTGTGATG-3', and antisense, 5'-GCAACATAGCACAGCTTCTC-3', yielding a product of 219 bp, as previously described . Reactions were performed on a Perkin-Elmer GeneAmp 9600 PCR system. Both the eGFP and beta-actin PCRs were conducted for 28 cycles with denaturation at 94C for 30 s, annealing at 65C for 1 min, and extension at 72C for 1 min. Reaction products were separated by electrophoresis on 0.8% agarose-EtBr gels and visualized using an Eagle Eye charge-coupled device camera (Stratagene). Nuclear extracts and EMSAs. | Nuclear extracts were prepared essentially as described elsewhere . Nuclear extracts were prepared without dialysis against modified Dignam buffer D and were left in Dignam buffer C (20 mM HEPES [pH 7.9], 25% [vol/vol] glycerol, 0.3 M NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, 0.5 mM dithiothreitol) as previously described . EMSAs were performed using previously described binding conditions . Probes were prepared by annealing complementary single-stranded oligonucleotides and then end labeling with [gamma-32P]ATP (Amersham Biosciences) using T4 polynucleotide kinase (Invitrogen). Extended 28-bp PBS probes were used due to reported increased binding in EMSA compared to the 18-bp core PBS sequence (underlined below) . Oligonucleotide probes were as follows: MLV PBS, 5'-GGGGGCTCGTCCGGGATCGGGAGACCCC-3'; B2 PBS, 5'-GGGGGCTCGTCCGAGATCGGGAGACCCC-3'; dl587 PBS, 5'-GGAGGTTCCACCGAGATTTGGAGACCCC-3' (only one strand for each is shown). Radioactively labeled double-stranded DNA oligonucleotides were purified using G-25 columns (Amersham Biosciences) according to the manufacturer's instructions. Radioactive labeling ranged from 200,000 to 400,000 cpm/ng. Each binding reaction was performed with 5 to 15 mul of nuclear extract in a total volume of 30 mul of Thornell binding buffer (25 mM HEPES [pH 7.9], 1 mM EDTA, 10% [vol/vol] glycerol, 5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride) containing 25 ng of poly(dI-dC)/ml, 5 mM NaCl, 5 mM KCl, 3 mM MgCl2, and 0.1 mM ZnCl2, as previously described . Prior to the addition of probe, binding reaction mixtures were preincubated at 30C for 20 min, and then 0.5 ng of double-stranded radioactively labeled probe was added in a 1-mul volume and reactions were incubated for an additional 20 min at 30C. Binding reactions were run on a 6% native acrylamide gel. The gel was loaded at room temperature, run for 20 min at 130 V, and then transferred to 4C and run at 175 V for approximately 4 h in Tris-glycine buffer (5 mM Tris [pH 8.5], 38 mM glycine, 0.2 mM EDTA). After electrophoresis, gels were dried and opposed to film for 1 to 4 days at -80C with intensifying screens. Mouse bone marrow harvest, transduction, and transplantation. | Mice were purchased from Jackson Laboratories (Bar Harbor, Maine) and maintained at the Childrens Hospital Los Angeles Central Animal Facility. All studies were approved by the Institutional Animal Care and Use Committee at Childrens Hospital Los Angeles. Donor marrow was harvested from 8- to 12-week-old male B6/SJL mice 2 days after 5-fluorouracil intravenous injection (150 mg/kg of body weight; American Pharmaceutical Partners, Los Angeles, Calif.). The marrow was cultured at a density of 2 x 106 cells/ml in IMDM supplemented with 20% FBS, 10 ng of murine IL-3 (Biosource International)/ml, 2.5 ng of murine stem cell factor (Biosource International)/ml, 50 ng of human IL-6 (Biosource International)/ml, 2 mM l-glutamine, 100 U of penicillin/ml, 100 mug of streptomycin/ml, and 0.01 mM 2-mercaptoethanol. Harvested cells were prestimulated in cytokine-containing medium for 12 h prior to transduction. Lentiviral supernatant was added to a final concentration of 2 x 107 IU/ml. Polybrene was added to a final concentration of 6 mug/ml. Recipient 8- to 12-week-old female C57Bl/6 mice were exposed to two doses of 600 cGy of gamma irradiation (128 cGy/min from a cesium-137 source) on two consecutive days (total of 1,200 cGy). One to two hours following the second dose of radiation, 2 x 106 to 4 x 106 transduced donor BM cells were injected into the tail vein of each recipient in 200 mul of phosphate-buffered saline containing 50 U of heparin/ml. Antibiotics were added to the drinking water for 3 weeks posttransplantation (200 mug of Maxim-200/ml; Phoenix Pharmaceuticals, St. Joseph, Mo.). The donor-recipient pairs were congenic at the CD45 locus, with donors expressing CD45.1 and recipients expressing CD45.2 isoforms of CD45, which are easily distinguishable by flow cytometry using commercially available antibodies (BD Biosciences, Palo Alto, Calif.). RESULTS : Modifications made to the MLV vector increase its frequency of expression in F9 EC cells. | We made a series of retroviral vectors containing from one to three modifications to the MLV vector to remove cis-acting elements reported to restrict expression of MLV in F9 EC cells. Figure shows the arrangement of elements contained in the full-length MLV vector constructs (not drawn to scale). The vector expresses eGFP from the 5'LTR and neomycin resistance (Neor) from an internal SV40 promoter. Figure shows the contribution each of the three modifications, both alone and in combination, made to the frequency of expression in F9 EC cells. To control for differences in titer between vector supernatants, 3T3 cells, which are permissive for MLV expression, were transduced in parallel and under the same conditions as F9 EC cells. Values are expressed as the percentage of eGFP-positive F9 EC cells relative to the percentage of eGFP-positive 3T3 cells 7 days after transduction. Figure shows that the unmodified MLV vector, L, expresses rarely in F9 EC cells. Replacing the MLV PBS with the dl587 PBS in vector LD, thereby removing the RBS, alleviates repression in a substantial fraction of F9 EC cells. Replacing the MLV U3 with the MPSV U3 alone in vector M, while leaving the MLV PBS and therefore the RBS in place, alleviates very little repression in F9 EC cells. MPSV is a variant of the Moloney sarcoma virus that had greater expression in myeloid cells , and the MPSV LTR was isolated and shown to express better than the MLV LTR in EC cells . Among the seven single-base differences between the MLV and the MPSV enhancer repeats, one of these single-base differences has been shown to introduce a functional Sp1 transcription factor binding site in the MPSV enhancer repeat that is not present in the MLV enhancer repeat. This Sp1 site accounts for much of the increase in activity of this enhancer in F9 EC cells . Replacing the MLV U3 with the MPSV U3 and replacing the MLV PBS with the dl587 PBS in combination in vector MD, thereby removing the RBS, alleviates repression in a substantial fraction of F9 EC. Removing the NCR from the U3 of MD to make the triply modified vector MND further alleviates repression in F9 EC cells. In each case where the MLV PBS, and therefore the RBS, is present expression is nearly completely repressed. Figure shows the contribution that each of these three modifications made to expression in F9 EC cells after selection in G418 for expression of Neor from a downstream internal SV40 promoter. Each of these F9 EC pools was selected with 0.75 mug of G418/ml for 10 days from a pool of cells that was transduced with a dilution of vector supernatant that transduced 3T3 cells in parallel to <10% eGFP positivity, so that most cells within these pools should contain only a single copy of the vector. In 0.75 mug of G418/ml, all nontransduced F9 EC cells in control wells were dead within 7 days. Following G418 selection, the cells were passaged in culture for an additional 2 weeks before analysis of eGFP expression by flow cytometry. Figure demonstrates the same pattern of increased expression of eGFP from the modified vectors as in unselected F9 EC cells. Again, in each case, removing the RBS by replacing the MLV PBS with the dl587 PBS made the largest contribution to increased expression, while changing the enhancer or deleting the NCR made only more modest improvements. Vectors containing the MLV PBS, and therefore the RBS (L and M), expressed in less than 5% of the G418-selected F9 cells, whereas the vectors lacking the RBS (LD, MD, and MND) expressed in 80 to 90% of selected cells. The RBS repressed expression from an internal MNDU3 promoter in a SIN lentiviral vector. | To determine if the RBS was capable of repressing expression outside of its normal retroviral genome context, a series of SIN lentiviral vectors were constructed having an internal MNDU3 promoter driving eGFP expression with one of the three PBS sequences under investigation inserted between the promoter and the eGFP transgene . SIN lentiviral vectors do not efficiently express transcript from their own LTR after reverse transcription, and thus eGFP expression reflects the activity of the internal promoter . Each vector supernatant in this series was produced and the titer was determined simultaneously on 293 cells. Based on these titer values, F9 EC and 293 cells were transduced in parallel using identical conditions. After 6 to 10 days the cells were harvested and analyzed for eGFP expression by flow cytometry. Flow cytometric analyses from a typical experiment are shown in Fig. . The averages of 15 experiments are plotted in Fig. . In 293 cells, all four vectors were expressed at the same frequency. In contrast, the lentiviral vector containing the MLV PBS shows essentially no expression in the F9 EC cells, whereas the other vectors, with no PBS or with the B2 or dl587 PBS, all expressed at similar frequencies. An alternative explanation for the results shown in Fig. is that the vector containing the MLV PBS was not transferred or integrated at the same frequency as the other vectors in the series. To rule out this possibility, we performed semiquantitative PCR on genomic DNA from transduced 293 and F9 EC cells. Figure shows that the vector containing the MLV PBS was transferred at a similar relative frequency as the vectors containing the B2 and dl587 PBS and the vector containing no PBS. Thus, the difference in expression by the vector carrying the MLV PBS is not due to poor gene transfer and therefore reflects repression of expression. The RBS repressed expression from internal SV40, hUbiqC, hEF-1alpha, and mPGK promoters. | Previous studies have demonstrated that within the context of the retroviral genome, the RBS was able to repress transcription from two downstream internal heterologous viral promoters, SV40 and AdMLP . To determine whether the RBS also repressed transcription from cellular promoters, we constructed SIN lentiviral vectors analogous to the vector series introduced in Fig. , but having either the SV40, hUbiqC, mPGK, or hEF-1a promoter driving eGFP expression, with and without the MLV PBS sequence inserted either immediately upstream or immediately downstream of the promoter. F9 EC cells and 293 cells were transduced in parallel with the various vector constructs using identical conditions. Cells were passaged in culture for 6 to 10 days and then analyzed for eGFP expression by flow cytometry. Figure shows the average results from two separate experiments. The data are presented as the percentage of eGFP-positive F9 EC cells normalized to the percentage of eGFP-positive 293 cells in the same vector arm, relative to the F9 EC/293 transduction ratio achieved by vector with no MLV PBS multiplied by 100. Figure shows that the SV40 promoter and the three cellular promoters examined were substantially repressed by the RBS, although to varying degrees. The SV40 promoter was repressed greater than 90% whether the MLV PBS was placed upstream or downstream of the promoter. The hUbiqC promoter was repressed greater than 60%, the hEF-1a promoter was repressed greater than 75%, and the mPGK promoter was repressed greater than 80%. In comparison, the MNDU3 promoter was repressed greater than 95%. These results indicate that although the RBS is capable of substantially repressing heterologous cellular promoters in this context, the repression is not complete when compared to the repression of the MNDU3 promoter. Repression by the RBS is more pronounced in undifferentiated stem cells but is not stem cell specific. | Previous studies have described the repressive activity of the RBS as stem cell specific on the basis that restriction is nearly complete in undifferentiated ES and EC cell lines but not in differentiated 3T3 fibroblasts . One study demonstrated that substantial RBS-mediated repressive activity was present in the mouse hematopoietic progenitor cell line FDCP . Previous studies have not examined RBS activity in primary mouse cells. To determine the extent of RBS activity in cell types other than EC and ES cell lines, the series of SIN lentiviral vectors containing the different PBS sequences and the MNDU3 promoter driving eGFP expression was used to screen for RBS repressive activity in primary mouse cells and mouse cell lines. The data presented in Fig. demonstrate that RBS-mediated repressive activity is present to varying degrees in mouse cells other than stem cells. Near-complete repression of the vector containing the MLV PBS was seen in F9 EC and ES-D3 cells, as previously reported, and also in WTc.F cells, an ES cell line generated from C57BL/6 mice . In addition, p300 and CBP knockout versions of WTc.F were tested, and similar repression was observed (data not shown). The vector with the MLV PBS was heavily repressed in the three heterogeneous primary mouse cell populations we analyzed. Expression was repressed in greater than 90% of mouse bone marrow stromal cells (CD45-, adherent cells isolated from adult mouse bone marrow), greater than 90% of whole mouse BM cells isolated from adult mice, and greater than 80% of MEFs isolated from 13.5-day embryos of outbred CF-1 mice. In addition, we observed that expression was repressed to varying degrees in four hematopoietic cell lines at different stages of differentiation: FDCP-Mix cl.A4 (hematopoietic progenitor), AMJ2-cll (macrophage), 70Z/3 (pre-B cell), and BM185 (pre-B cell). Both the 3T3 and STO cell lines showed no substantial repression when compared to the human 293 cell line. Semiquantitative PCR was performed on selected cell types to verify that differences in expression were not attributable to differences in the level of gene transfer . Although the panel of cells examined is limited, these data demonstrate that RBS-mediated repressive activity is not a stem-cell-specific, cell-line-specific, or mouse-strain-specific activity. The RBS repressed expression in whole mouse bone marrow and its differentiated progeny after bone marrow transplant. | As described above , expression was repressed in greater than 90% of whole mouse BM cells transduced with the SIN lentiviral vector containing the MLV PBS. To determine the fate of the progeny of the bone marrow progenitors in this cell population carrying repressed vector, we transplanted transduced murine BM cells into radiation-ablated mice. Figure shows the percentage of cells expressing eGFP in donor mouse bone marrow transduced for transplantation but maintained in culture for 1 week. Approximately 10 to 12% of the cells expressed eGFP from each vector, except <1% expression was seen in cells transduced by the lentiviral vector containing the MLV PBS. After 10 weeks, the transplanted mice were sacrificed and hematopoietic cells were harvested from the bone marrow, peripheral blood, thymus, and spleen. Figure shows the percentage of donor-derived CD45.1+ cells that expressed eGFP from each mouse in each hematopoietic tissue examined. Expression was nearly completely absent from the vector containing the MLV PBS in CD45+ cells from all four hematopoietic compartments examined. This stands in contrast to the B2 and dl587 PBS vector arms that yielded frequencies of eGFP expression comparable to the vector having no PBS. The absence of eGFP expression in these differentiated hematopoietic cell populations may not reflect active RBS-mediated repression but instead could be due to an inherited epigenetic mechanism of silencing, such as DNA methylation or chromatin condensation, that occurred secondary to RBS-mediated repression in a repopulating hematopoietic progenitor. To rule out the possibility that the lack of eGFP expression observed in the cells transduced with the vector containing the MLV PBS was due to a difference in engraftment efficiency in this arm, harvested cells from all arms were stained with anti-CD45.1 antibody to differentiate donor-derived CD45.1 cells from recipient-derived CD45.2 cells. Figure shows the percentage of harvested cells that were derived from donor cells according to CD45.1 antibody staining. The frequency of CD45.1+ cells recovered from blood, thymus, spleen, and bone marrow of mice transplanted with cells transduced with the MLV PBS-containing vector was in the same range as in the other three vector arms and in the mock-transduced arm. Thus, it can be concluded that the lack of expression seen in the MLV PBS vector arm was not due to inefficient engraftment of donor cells. To rule out the possibility that the observed lack of eGFP expression seen in the cells transduced by the vector containing the MLV PBS was due to inefficient gene transfer into the donor bone marrow, semiquantitative PCR was performed on genomic DNA isolated from an aliquot of the transduced donor BM cells and bone marrow isolated from each recipient mouse. Donor bone marrow was transduced to the same relative efficiency in each vector arm. All three bone marrow samples from mice transplanted with bone marrow transduced with the vector containing the MLV PBS were positive for eGFP sequence, and the signal intensity was not significantly different from that seen in the 11 other mice in the three other vector arms. Therefore, the lower expression seen from the vector with the MLV PBS was not due to inefficient gene transfer but rather reflects repression of expression. The RBS repressed expression in the human hematopoietic cell line DU.528 and primary human CD34+ CD38- cells isolated from umbilical cord blood. | Previous studies have not examined the effects of the RBS in human cells. The series of SIN lentiviral vectors containing the different PBS sequences and the MNDU3 promoter driving eGFP expression was used to screen for RBS repressive activity in a panel of primary human cells and human cell lines. Of all the cell types we examined, only DU.528 and primary CD34+ CD38- cells demonstrated any repressive activity . DU.528 is a hematopoietic progenitor cell line capable of generating progeny with characteristics of at least three hematopoietic lineages, in vitro: T-lymphoid, granulocytic/monocytic, and erythroid . CD34+ CD38- cells constitute about 3.5% of CD34+ cells present in umbilical cord blood and are enriched for primitive hematopoietic progenitor cells . About 50% of the cells in each of these cell populations were repressed for expression of the MLV PBS-containing vector. In contrast, little repression was seen in transduced CD34+ cells, which are a heterogeneous population of cells that are more mature than the CD34+ CD38- cells. In contrast to mouse EC cells , substantial repression was not observed in the three human EC cells we tested: Tera-2, PA-1, and NCCIT. In addition, we did not observe substantial repression in any of the five human hematopoietic cell lines we tested: U937 (myeloid), KG1a (myeloid), CEM (T cell), Jurkat (T cell), and K562 (myeloid). These findings suggest that expression of a factor(s) that binds to the RBS and represses expression occurs mainly in the more primitive and pluripotent human hematopoietic stem and progenitor cells. Binding factor A was present in mouse cells that have RBS-mediated repressive activity. | The differential binding of factor A, from EC cell nuclear extracts, to a MLV PBS probe, but not a B2 PBS probe, has been previously demonstrated by an EMSA . The EMSA gel pictures shown in Fig. demonstrate that the factor A bandshift was present in nuclear extracts from primary mouse cells and mouse cell lines other than EC cells, in each cell type where the MLV PBS was repressive, including cell types where repression by this element was only partial (i.e., BM185 cells). We observed RBS-mediated repressive activity in the human cell line DU.528 and primary CD34+ CD38- cells in our expression assay. To determine if an orthologue of factor A protein was detectable in human cells that showed repression by the MLV PBS, we generated nuclear extracts from DU.528 cells for use in the EMSA. In two experiments, differential binding of any factor to the MLV PBS and B2 PBS probes was not observed. Due to the number of cells required to generate nuclear extracts, we have not attempted this assay with primary human CD34+ CD38- cells. DISCUSSION : Repression by the RBS outside of the context of MLV genome. | Previous studies have demonstrated that the RBS is active in either orientation when positioned upstream of the MLV transcription start site or downstream of its normal position within an intron. Although the minimal length of the RBS sequence required for repressive activity was shown to be 18 bp, all previous studies have examined the RBS within the context of the MLV genome, leaving open the possibility that other sequences within the MLV genome might be required to cooperate with the RBS for full repressive activity. The only elements from MLV present in the lentiviral vectors we studied were the U3 region from MPSV (with the NCR deleted) and the MLV PBS. In this context, the internal MNDU3 promoter was fully repressed by the RBS, suggesting that additional elements outside of the PBS sequence and the MPSV U3 are not required for full PBS activity, although our study does not exclude the possibility that the RBS may cooperate with elements present in the lentiviral vector backbone. No effect was seen from the B2 or dl587 PBS. Repression of the SV40, hUbiqC, mPGK, and hEF-1a promoters by the RBS. | A previous study demonstrated that within the context of the MLV genome, the RBS was able to repress transcription from a position upstream of two internal heterologous viral promoters, SV40 and AdMLP . In this study, we showed that in addition to the SV40 viral promoter, three cellular promoters were each substantially repressed in F9 EC cells when the RBS was present in the vector, although each of the heterologous promoters was repressed to differing degrees and none were as fully repressed as the MNDU3 promoter. Nevertheless, these experiments demonstrate that the mechanism of repression of the RBS is able to dominantly repress transcription directed by many different transcription factors. RBS-mediated repressive activity in mouse cells is not stem cell specific. | Previous studies have referred to RBS-mediated repressive activity as an undifferentiated stem cell-specific activity. Most studies of the RBS have focused on repression of MLV in EC and ES cell lines compared to the lack of repression observed in 3T3 fibroblasts. Our study of RBS-mediated repression in mouse cells demonstrates that the repressive activity of the RBS, although most pronounced in ES and EC cell lines, is not restricted to undifferentiated stem cells. Our findings demonstrate that there is substantial RBS-mediated repressive activity in several mouse progenitor and differentiated hematopoietic cell lines as well as three heterogeneous primary cell populations isolated from embryonic and adult mice. The only mouse cells we tested that did not demonstrate repression of vector expression by the RBS were 3T3 and STO fibroblasts. Both 3T3 and STO cells were originally isolated from primary MEF cultures , which in this study showed nearly the same degree of repression as ES and EC cell lines. The lack of RBS-mediated repressive activity in the 3T3 and STO cell lines might be due to either aberrant gene expression following transformation, a founder effect, or the loss of repressor expression during differentiation. Vectors derived from MLV have been used for a variety of purposes where stable gene transfer and expression is necessary, including preclinical studies leading to gene therapy clinical trials. Often times, these initial expression studies are performed in mouse cell culture or animal models. This study demonstrated that several more-differentiated hematopoietic cell lines possess an incomplete but substantial degree of RBS-mediated repressive activity, compared to ES cells. Primary isolates of MEFs and mouse bone marrow stromal cells isolated from adult mice also demonstrated RBS-mediated repressive activity comparable to ES and EC cells. Although we have not examined fully differentiated cells outside the hematopoietic compartment, it is possible that many differentiated cell types in the mouse may possess substantial RBS-mediated repressive activity that may affect the outcome of gene expression studies that employ MLV-based vectors containing the MLV PBS. Several commercially available retroviral expression systems contain the MLV PBS and so may not be ideal tools for gene expression studies in mice. RBS-mediated repressive activity in human cells. | Our data using the DU.528 human hematopoietic cell line and primary human CD34+ CD38- hematopoietic stem cells provide the first evidence that RBS-mediated repressive activity is present in human cells. The RBS is hypothesized to effect transcriptional repression by an undefined mechanism through an as-yet-unidentified trans-acting factor or factors identified by EMSA as binding factor A. If mouse binding factor A is the trans-acting factor that conveys repression through its interaction with the RBS, then human cells that repress expression from vectors containing the RBS may express an orthologue. We have been unable to demonstrate using an EMSA that nuclear extracts from the human cell line DU.528 contain an obvious differentially binding factor analogous to the differentially binding factor in the nuclear extracts of mouse cells that repress expression of vectors containing the RBS. This may be because the binding conditions used in the EMSA were optimized for use with nuclear extracts from mouse cells that contain high RBS-mediated activity, and the conditions have not been optimized for human nuclear extracts. The observation that human cells possess RBS-mediated repressive activity has important implications for human gene therapy studies. Past and ongoing clinical trials have been performed using MLV-based vectors containing the MLV or B2 PBS. Because both the tRNA sequence and the PBS sequence are copied during reverse transcription, the double-stranded DNA product of reverse transcription of a vector having the B2 PBS contains a single base mismatch within the PBS region. The cellular machinery that repairs this single base mismatch in the double-stranded DNA provirus corrects one or the other base so that approximately 50% of the progeny provirus contain a B2 PBS and 50% revert to the MLV PBS sequence . Although our studies with human cell lines representing mature hematopoietic lineages did not detect any repressive activity of the RBS, it has not yet been determined whether RBS-mediated repressive activity is present in primary differentiated human hematopoietic lineages or in differentiated nonhematopoietic cell types. We were able to detect RBS-mediated repressive activity in human CD34+ CD38- cells, but it is not yet known whether this repression is maintained or reversed when these cells differentiate in vivo. If the RBS-mediated repressive activity we observed in human CD34+ CD38- cells is sustained through differentiation, as we observed in mouse bone marrow transplant experiments, then integrated proviruses containing the RBS in patients might be repressed. At present, there are MLV-based vectors readily available with PBS replacements which do not contain the RBS (MND , MSCV , and HSC1 ). Our results suggest that it would be advisable to use the modified vectors instead of those containing the RBS for gene transfer and expression studies in mice and that there is the potential for the RBS present in some clinically approved retroviral vectors to have deleterious effects on vector expression in human gene therapy clinical trial patients. FIG. 1. : Modifications made to the MLV vector increase its frequency of expression in F9 EC cells. Modifications made to the MLV vector increase its frequency of expression in F9 EC cells. (A) MLV vector provirus diagram showing the arrangement of elements contained in the full-length MLV vector construct (not drawn to scale). The vector expresses eGFP from the 5'LTR and neomycin resistance from an internal SV40 promoter. Modifications made to the vector are depicted on the left side of the figure. (B) Frequency (+- standard deviation) of eGFP expression in unselected F9 EC cells relative to 3T3 cells transduced in parallel. (C) Frequency of eGFP expression in F9 EC cell pools after selection in G418 for expression of Neor from a downstream SV40 promoter. FIG. 2. : The RBS repressed expression in F9 EC cells when positioned downstream of an internal MNDU3 promoter expressing eGFP in a SIN lentiviral vector. The RBS repressed expression in F9 EC cells when positioned downstream of an internal MNDU3 promoter expressing eGFP in a SIN lentiviral vector. (A) Lentiviral vector provirus diagram showing the arrangement of elements in the SIN lentiviral vector series having an internal MNDU3 promoter driving eGFP expression and one of the three PBS sequences inserted between the promoter and the eGFP transgene. (B) Representative flow cytometric analyses of F9 EC and 293 cells transduced in parallel with vectors containing the indicated PBS sequences. (C) Averages (+- standard deviation) of flow cytometry data from 15 experiments performed as described for panel B. The percentage of eGFP-positive F9 EC cells was normalized to the percentage of eGFP-positive 293 cells in the same vector arm to control for small differences in vector titer, relative to the F9 EC/293 transduction ratio achieved with vector with no PBS, multiplied by 100. (D) Semiquantitative PCR demonstrates that gene transfer by the vector containing the MLV PBS occurred at relatively the same frequency as for vectors containing the B2, dl587, or no PBS. Mock-transduced cells served as negative controls for eGFP detection by PCR. FIG. 3. : The RBS repressed expression in F9 EC cells by a lentiviral vector with an internal hUbiqC, hEF-1alpha, mPGK, SV40, or MNDU3 promoter. The RBS repressed expression in F9 EC cells by a lentiviral vector with an internal hUbiqC, hEF-1alpha, mPGK, SV40, or MNDU3 promoter. (A) Lentiviral vector provirus diagram showing arrangement of elements in the vector series having one of five promoters and either no PBS or an MLV PBS inserted either upstream (<-) or downstream (->) of the promoter expressing eGFP. (B) Percentage of eGFP-positive F9 EC cells normalized to the percentage of eGFP-positive 293 cells in the MLV PBS vector arm, relative to the F9 EC/293 transduction ratio achieved with vector with no PBS, multiplied by 100 (+- standard deviation). FIG. 4. : Repression by the RBS is not stem cell specific. Repression by the RBS is not stem cell specific. (A) eGFP expression in mouse cells from a SIN lentiviral vector with an internal MNDU3 promoter driving eGFP expression with one of the PBS sequences placed between the promoter and the eGFP transgene. Data are presented as the percentage of eGFP-positive target cells (indicated on the x axis) normalized to the percentage of eGFP-positive 293 cells in the same vector arm, relative to the target cell/293 transduction ratio achieved with vector with no PBS, multiplied by 100 (+- standard deviation). (B) Semiquantitative PCR demonstrates that gene transfer by the vector containing the MLV PBS occurred at relatively the same frequency as that with vectors containing the B2, dl587, or no PBS. FIG. 5. : The RBS repressed expression in whole mouse bone marrow and its differentiated progeny after bone marrow transplant. The RBS repressed expression in whole mouse bone marrow and its differentiated progeny after bone marrow transplant. (A) eGFP expression of donor bone marrow kept in culture for 7 days following transduction. (B) Mean eGFP expression (+ standard deviation) in CD45.1+ donor cells harvested from recipient mice 10 weeks after transplant with transduced bone marrow shown in panel A. Circles represent values of individual mice. (C) Percentage of donor CD45.1+ cells recovered from transplanted mice. (D) Semiquantitative PCR demonstrates that gene transfer by the vector containing the MLV PBS occurred at relatively the same frequency as the B2, dl587, and no PBS vectors in donor bone marrow and bone marrow harvested from each recipient mouse. FIG. 6. : The RBS repressed expression in the human hematopoietic cell line DU.528 and primary human CD34+ CD38- cells isolated from umbilical cord blood. The RBS repressed expression in the human hematopoietic cell line DU.528 and primary human CD34+ CD38- cells isolated from umbilical cord blood. (A) eGFP expression in human cells from a SIN lentiviral vector with an internal MNDU3 promoter driving eGFP expression with one of the PBS sequences placed between the promoter and the eGFP transgene. (B) Semiquantitative PCR demonstrates that gene transfer by the vector containing the MLV PBS occurred at relatively the same frequency as the B2, dl587, and no PBS vectors. FIG. 7. : EMSA for differentially binding factor A. EMSA for differentially binding factor A. Differential binding of the factor A bandshift, to an MLV PBS probe but not a B2 PBS probe, was observed in nuclear extracts of several mouse cell lines and primary mouse cells, but not from human DU.528 cells. Backmatter: PMID- 12915538 TI - Mutations in the N Termini of Herpes Simplex Virus Type 1 and 2 gDs Alter Functional Interactions with the Entry/Fusion Receptors HVEM, Nectin-2, and 3-O-Sulfated Heparan Sulfate but Not with Nectin-1 AB - Multiple cell surface molecules (herpesvirus entry mediator [HVEM], nectin-1, nectin-2, and 3-O-sulfated heparan sulfate) can serve as entry receptors for herpes simplex virus type 1 (HSV-1) or HSV-2 and also as receptors for virus-induced cell fusion. Viral glycoprotein D (gD) is the ligand for these receptors. A previous study showed that HVEM makes contact with HSV-1 gD at regions within amino acids 7 to 15 and 24 to 32 at the N terminus of gD. In the present study, amino acid substitutions and deletions were introduced into the N termini of HSV-1 and HSV-2 gDs to determine the effects on interactions with all of the known human and mouse entry/fusion receptors, including mouse HVEM, for which data on HSV entry or cell fusion were not previously reported. A cell fusion assay was used to assess functional activity of the gD mutants with each entry/fusion receptor. Soluble gD:Fc hybrids carrying each mutation were tested for the ability to bind to cells expressing the entry/fusion receptors. We found that deletions overlapping either or both of the HVEM contact regions, in either HSV-1 or HSV-2 gD, severely reduced cell fusion and binding activity with all of the human and mouse receptors except nectin-1. Amino acid substitutions described previously for HSV-1 (L25P, Q27P, and Q27R) were individually introduced into HSV-2 gD and, for both serotypes, were found to be without effect on cell fusion and the binding activity for nectin-1. Each of these three substitutions in HSV-1 gD enhanced fusion with cells expressing human nectin-2 (ordinarily low for wild-type HSV-1 gD), but the same substitutions in HSV-2 gD were without effect on the already high level of cell fusion observed with the wild-type protein. The Q27P or Q27R substitution in either HSV-1 and HSV-2 gD, but not the L25P substitution, significantly reduced cell fusion and binding activity for both human and mouse HVEM. Each of the three substitutions in HSV-1 gD, as well as the deletions mentioned above, reduced fusion with cells bearing 3-O-sulfated heparan sulfate. Thus, the N terminus of HSV-1 or HSV-2 gD is not necessary for functional interactions with nectin-1 but is necessary for all of the other receptors tested here. The sequence of the N terminus determines whether nectin-2 or 3-O-sulfated heparan sulfate, as well as HVEM, can serve as entry/fusion receptors. Keywords: Introduction : Human herpes simplex virus type 1 (HSV-1) and HSV-2, porcine pseudorabies virus (PRV), and bovine herpesvirus type 1 (BHV-1) are members of the alphaherpesvirus subfamily and have similar viral and cellular requirements for entry into cells . Initial attachment is usually mediated by interaction of virion envelope glycoprotein C (gC) and/or gB with cell surface heparan sulfate. The subsequent interaction of virion gD with one of its receptors triggers the penetration of virus, which occurs by fusion of the viral envelope with a cell membrane, and requires four viral glycoproteins, gB, gD, gH, and gL. HSV-induced cell fusion requires the same four viral glycoproteins ---gB, gD, gH, and gL ---as well as a gD receptor . Transfection of cells with plasmids expressing gB, gD, gH, and gL is sufficient, in the absence of virus infection, to induce fusion with target cells expressing HSV entry receptors . Three classes of cell surface molecules can serve as entry/fusion receptors for HSV-1 or HSV-2 . These include herpesvirus entry mediator (HVEM), a member of the tumor necrosis factor receptor family; nectin-1 and nectin-2, two members of the immunoglobulin superfamily; and specific sites in heparan sulfate generated by the action of certain isoforms of 3-O-sulfotransferase (3-OST). Human and, as shown here, mouse forms of HVEM can mediate entry of both HSV-1 and HSV-2, but not PRV or BHV-1 . Human and mouse forms of nectin-1 can serve as entry receptors for HSV-1, HSV-2, PRV, and BHV-1 , whereas human and mouse forms of nectin-2 have limited entry activity. Human nectin-2 can mediate the entry of HSV-2 and PRV but has considerably less activity with HSV-1, unless there is an amino acid substitution at position 25 or 27 of gD . Mouse nectin-2 can mediate the entry of PRV but not BHV-1 or HSV strains . Specific sites in heparan sulfate generated by human or mouse 3-O-sulfotranferases, 3-OST-3A or 3-OST-3B, can serve as entry receptors for HSV-1 . X-ray structures of a truncated form of HSV-1 gD, alone and in complex with truncated HVEM, were recently determined . A portion of the gD ectodomain forms an immunoglobulin fold in both structures, with N-terminal and C-terminal extensions. The N-terminal extension forms a hairpin in the complex with HVEM but is disordered in the crystal of gD alone. The interface between gD and HVEM involves only two short segments of gD (amino acids 7 to 15 and 24 to 32) in the N-terminal hairpin. These observations suggest that the N terminus of gD is conformationally flexible and that gD may have different conformations upon binding different receptors. A single amino acid substitution at position 27 (Q27P or Q27R) in HSV-1 gD prevents binding to HVEM and entry via HVEM , a finding consistent with the contact sites identified in the X-ray structure, but has no negative effect on binding to nectin-1 or entry via nectin-1 . Intriguingly, these substitutions confer the ability to use nectin-2 as an entry/fusion receptor . The inability of PRV and BHV-1 to use HVEM as an entry/fusion receptor can be explained by significant divergence of PRV and BHV-1 gD sequences from HSV-1 and HSV-2 gD in the N-terminal domain that has the contact sites for HVEM. However, there must be structural conservation elsewhere in the proteins to explain the fact that all four viruses can use nectin-1 as an entry/fusion receptor. These observations imply that the interfaces of gD with different receptors are different. In the present study, we investigated whether the N-terminal domains of HSV-1 and HSV-2 gD were critical for functional interactions with each of the known human and mouse entry/fusion receptors. We identified amino acid substitutions or deletions that significantly altered or reduced physical and functional interactions with HVEM, nectin-2 and 3-O-sulfated heparan sulfate but were without effect on interactions with nectin-1. MATERIALS AND METHODS : Cells. | Chinese hamster ovary (CHO-K1) cells were provided by J. Esko (University of California, San Diego). CHO-IEbeta8 cells, obtained by the stable transfection of CHO-K1 cells with pMLP01, express the Escherichia coli lacZ gene under control of the HSV-1 ICP4 promoter . CHO-K1 cells stably expressing human HVEM, nectin-1, or nectin-2 were originally designated CHO-HveA35 , CHO-Prr1 , or CHO-Prr2 , respectively. Here they are called CHO-HVEM, CHO-nectin-1, and CHO-nectin-2, respectively. All cells were grown in Ham F12 medium with 10% fetal bovine serum. Plasmids. | Plasmids encoding HSV entry receptors included pcDNA3-based constructs as follows: pBEC10 expressing human HVEM , pBG38 expressing human nectin-1 , human 3-OST-3B whose expression generates HSV-1 entry receptors , and pDS106 expressing mouse HVEM and pCR13 expressing mouse nectin-1. HSV-1 and HSV-2 gD mutants were generated by using the QuickChange site-directed mutagenesis kit (Strategene). HSV-1 gD mutants were generated by using pAZT3 , which contains the open reading frame (ORF) of HSV-1(KOS) gD in pUC19 (nucleotides 138419 to 139603 cloned between the EcoRI and SphI restriction sites). HSV-2 gD mutants were generated by using pMY1, constructed by subcloning the EcoRI-KpnI fragment, containing the HSV-2(333) ORF, from pAZD2 into pUC19. pMY1 contains nucleotides 141016 to 142197 cloned between the EcoRI and KpnI sites in pUC19. Nucleotide numbers are from the published sequences of the HSV-1 genome (NC_001806) and the HSV-2(HG52) genome (NC_001798). Plasmids expressing the various forms of gD and mutants were generated by subcloning the EcoRI-SphI fragment (for HSV-1 gD) or the EcoRI-KpnI fragment (for HSV-2 gD) from the pUC19-based plasmids into pCAGGS/MCS . Plasmids expressing soluble gD:Fc hybrids were constructed to contain the first 320 codons of HSV-1(KOS) gD (excluding the signal sequence) or the first 319 codons of HSV-2(333) gD (wild-type or mutant forms) fused to the C terminus of the rabbit immunoglobulin G (IgG) heavy chain. The appropriate region of each gD gene was amplified by PCR with primers that added an upstream HindIII site and a downstream KpnI site and the pUC19-based plasmids as templates. The PCR products were cloned into pDM19, which is a pcDNA3-based plasmid and contains the cytomegalovirus promoter upstream of a HindIII site and 231 codons of the rabbit IgG heavy-chain gene downstream of a KpnI site. For HSV-1 gD:Fc plasmids, the primer pairs consisted of 5'-GGACCAAAGCTTGAATTCCTTTTGTGTGGT-3' and 5'-TCGAGCTCGGTACCCGGCGATCAGGCCCAT-3'. For HSV-2 gD:Fc plasmids, the primer pairs consisted of 5'-GGACCAAAGCTTGAATTCGTGTGCATCGCG-3' and 5'-TCGAGCTCGGTACCCGATGATCAGGCCCGG-3'. The introduced restriction endonuclease recognition sequences are shown in boldface. All plasmids used in the present study were verified by DNA sequencing through the gD genes and are listed in Table . Viruses and viral entry assay. | HSV-1(F), HSV-1(U10), HSV-1(KOS), HSV-1(KOS)Rid1, HSV-1(KOS)tk12, and HSV-1(KOS)Rid1/tk12 were described previously . Entry assays were based on quantitation of beta-galactosidase expressed from the viral genome or expressed in CHO-IEbeta8 cells as a result of viral entry. CHO-IEbeta8 cells are induced to express beta-galactosidase upon viral entry and delivery of the HSV-1 transactivator VP16 into cells . At low-input multiplicities of infection, the beta-galactosidase activity detected from the reporter gene in virus or cell is proportional to the number of cells infected; when all cells are infected, the enzyme activity reaches a plateau (L. Kwon and P. G. Spear, unpublished data). Parental CHO-K1, CHO-HVEM, or CHO-nectin-1 cells were plated in 96-well plates (ca. 5 x 104 cells per well). The next day, the cells were incubated with serial dilutions of the beta-galactosidase reporter viruses, HSV-1(KOS)tk12 or HSV-1(KOS)Rid1/tk12, at 37C for 6 h. Alternatively, CHO-IEbeta8 cells were transfected with plasmids expressing human HVEM (pBEC10), human nectin-1 (pBG38), mouse HVEM (pDS106), or mouse nectin-1 (pCR13) for 6 h and then plated in 96-well plates overnight. Cells were incubated with serial dilutions of HSV-1(KOS), HSV-1(KOS)Rid1, HSV-1(F), or HSV-1(U10) at 37C for 6 h. The cells were then washed, permeabilized, and incubated with the beta-galactosidase substrate, O-nitrophenyl-beta-d-galactopyranoside (ONPG; Sigma). The reaction was monitored at 410 nm in a Spectra Max 250 enzyme-linked immunosorbent assay (ELISA) reader. Preparation of soluble gD:Fc fusion proteins. | Soluble gD:Fc fusion proteins were produced as described previously with modification. Briefly subconfluent CHO-K1 cells in six-well plates were transfected with gD:Fc-expressing plasmids for 6 h by using Lipofectamine (Invitrogen) in serum-free medium and changed into medium containing 10% fetal bovine serum overnight. Cells were then washed and incubated with serum-free medium, and supernatants were collected 48 h later. The culture supernatants were clarified by low-speed centrifugation, and the concentrations of gD:Fc proteins were determined by ELISA by using an anti-rabbit Fc detection system and rabbit IgG for the standard curve. Receptor-binding assay. | The receptor-binding assay was described previously . CHO cells expressing HSV-1 or HSV-2 entry/fusion receptors were grown on coverslips and incubated with gD:Fc at 10 ng/ml and 37C for 30 min. Cells were washed, fixed with methanol at -20C for 5 min, and incubated with Alexa 488-conjugated goat anti-rabbit IgG (Molecular Probes). Binding was visualized by confocal microscopy. Alternatively, cell ELISA (CELISA) was performed. Cells grown on 96-well plates were incubated with serial dilutions of gD:Fc for 30 to 60 min at 37C. Cells were washed and fixed with 2% formaldehyde and 0.2% glutaraldehyde for 10 min. The cells were then sequentially incubated with biotinylated anti-rabbit IgG (Sigma), Amdex streptavidin-conjugated horseradish peroxidase (HRP; Amersham), and HRP substrates (BioFx Lab). Binding was monitored at 370 nm. Assay for cell surface expression of gD. | CELISA was carried out to detect cell surface expression of gD, as described previously . Briefly, subconfluent CHO-K1 cells in six-well plates were transfected with pCAGGS-based plasmids expressing various forms of gD, along with gB, gH, and gL for 6 h and replated on 96-well plates overnight. The cells were then incubated with polyclonal anti-gD rabbit serum (R7) at 37C for 30 min, fixed, and sequentially incubated with biotinylated anti-rabbit IgG, streptavidin-conjugated HRP, and HRP substrates. Binding was monitored at 370 nm. Cell fusion assay. | Details of the cell fusion assay were described previously . Effector (CHO-K1) cells were transfected with a mixture of plasmids expressing HSV glycoproteins (gB, gD, gH, and gL) and T7 RNA polymerase (pCAGT7). Plasmids expressing the HSV-1(KOS) glycoproteins were described elsewhere : pPEP98 (gB), pPEP100 (gH), and pPEP101 (gL). Plasmids expressing the HSV-2(333) glycoproteins have also been described : pAZB2 (gB), pAZD2 (gD), pAZH2 (gH), and pAZL2 (gL). The target cells were CHO-HVEM, CHO-nectin-1, CHO-nectin-2, or CHO-K1 cells transiently transfected with human 3-OST-3B, mouse HVEM or mouse nectin-1. All target cells were transfected or cotransfected with a plasmid expressing luciferase under control of the T7 promoter (pT7EMLuc). Effector and target cells were subsequently detached and mixed in a 1:1 ratio and replated for 18 h. Luciferase activity was quantitated by a luciferase reporter assay system (Promega) by using a TD-20/20 luminometer (Turner Designs). Immunofluorescence and microscopy. | Indirect immunofluorescence was carried out as described previously . The primary antibodies used were mouse monoclonal anti-nectin-1 antibody R1.302 and rabbit polyclonal anti-HVEM antibody R11874 . The secondary antibodies used were Alexa 488-conjuated goat anti-mouse IgG and Alexa 488-labeled anti-rabbit IgG. Immunofluroscence observations were made with a Zeiss LSM 510 confocal microscope equipped with a x100, 1.4 numerical aperture oil immersion objective lens. Orthogonal sections were made in a z stack of 30 images (0.6-mum interval) according to the LSM 510 operating manual. RESULTS : Effect of substitutions at position 25 in HSV-1 gD on HVEM-mediated entry. | Viral mutants with amino acid substitutions in the N-terminal region of HSV-1 gD have been described previously. As mentioned above, Rid mutants of HSV-1(KOS) have substitutions at position 27 (Q27P or Q27R) that significantly reduce their ability to use human HVEM as an entry receptor and enhance their ability to use human nectin-2 . The U10 mutant of HSV-1(F) has a substitution at position 25 (L25P) and has enhanced ability to use human nectin-2 as an entry receptor . Results have not previously been reported for effects of the U10 mutation on viral entry via human HVEM or for effects of any of these mutations on entry via mouse HVEM. Therefore, we inoculated CHO cells expressing human or mouse HVEM with the Rid and U10 mutants and HSV-1 parental strains and assessed the efficiency of viral entry by quantitation of beta-galactosidase, which was expressed from a reporter gene under control of the immediate-early HSV-1 ICP4 promoter, either in the viral genome or in the cell. Because the inoculated cells were permeabilized for beta-galactosidase quantitation at 6 h after the addition of virus, viruses that failed to enter the cells or entered at a very slow rate would score as impaired for entry. As positive controls, CHO cells expressing human or mouse nectin-1 were also inoculated. As negative controls, CHO-K1 cells were inoculated. The results presented in Fig. show, first, that the human and mouse forms of nectin-1 or HVEM were similar in their ability, or failure, to mediate entry of the mutant and wild-type forms of HSV-1. Second, whereas the Rid mutation (Q27P) eliminated the ability of the mutant to infect cells expressing human or mouse HVEM, the U10 mutation (L25P) had a lesser effect but reduced the efficiency of infection for cells expressing either form of HVEM. At least 10 times more of the U10 mutant than the wild-type parental strain F was required to achieve similar levels of entry. Third, the Rid mutant appeared to infect nectin-1-expressing cells less efficiently than did HSV-1(KOS), whereas there was little difference between the U10 mutant and HSV-1(F) in this regard. Finally, some differences between the parental strains are evident. Lower input multiplicities of HSV-1(F) than HSV-1(KOS) achieved similar levels of viral entry on the nectin-1-expressing and HVEM-expressing cells. Also, infection of the control CHO cells could be detected at high input concentrations of HSV-1(F) but not at all for HSV-1(KOS). This latter difference with respect to CHO cells was reported previously and indicates that there is a weak endogenous strain-dependent CHO entry receptor for HSV-1. It should be noted that the amino acid sequences of the gD genes of these two strains are identical in the ectodomain and differ by only two amino acids in the membrane-proximal region of the cytoplasmic tail . The explanation for the observed differences between HSV-1(KOS) and HSV-1(F) in the efficiency of viral entry or early events leading to gene expression could lie in these differences or in differences in other genes. We conclude that mouse and human HVEM are indistinguishable in their entry activities, at least for the viral strains analyzed in Fig. , despite our previous mentions of unpublished and preliminary findings that HSV-1(KOS)Rid mutants could infect cells via mouse HVEM . We also conclude that the Q27P substitution has a more profound effect on entry via HVEM than does the L25P substitution, although the fact that the mutant viral strains are not isogenic could have influenced the results. In the present study, we have assessed the effects of both of these mutations, as well as others, in the HSV-1(KOS) and HSV-2(333) genetic backgrounds. Mutations in the N-terminal region of HSV-2 gD have not yet been described. Amino acid substitutions and deletions were introduced into the HSV-1 and HSV-2 gD genes. The altered forms of gD were expressed as soluble hybrid proteins in which the ectodomain of gD was fused to the Fc region of rabbit IgG. The hybrids were used to assess the effects of the mutations on ability of gD to bind to cell surface nectin-1 and HVEM. Binding to the other receptors studied here, nectin-2 and 3-O-sulfated heparan sulfate, could not be assessed in this way, probably because of lower affinities of binding . The full-length mutant forms of gD were also expressed in CHO cells, along with gB, gH, and gL, to quantitate effects of the mutations on ability of the glycoprotein-expressing cells (effector cells) to induce fusion with cells expressing the various receptors (target cells). This cell fusion assay provided a convenient way to test the functional activity of multiple gD mutants and yielded results likely to be predictive of the effects of the mutations on viral entry. Effects of mutations in the N termini of HSV-1 and HSV-2 gDs on binding of gD:Fcs to nectin-1 and HVEM. | Binding of the mutant forms of gD:Fcs to entry receptors was assessed qualitatively by immunofluorescence and quantitatively by measuring the amounts bound to intact cells in monolayer (CELISA). Figure illustrates differences in the cellular localization of human HVEM and nectin-1 in stably expressing CHO cells, as determined by antibody binding and by the binding of wild-type forms of HSV-1 or HSV-2 gD:Fcs. As previously shown , antibodies specific for nectin-1 localized to regions where two cells come in contact , reflecting the ability of this cell adhesion molecule to engage in homophilic trans-interactions between cells. Nectin-2 exhibits the same cellular localization . Antibodies to HVEM, on the other hand, were distributed all over the cell surface , at least all over the surface not in contact with the substrate. The other panels of Fig. show that HSV-1 gD:Fc (panels G and H) and HSV-2 gD:Fc (panels I and J) bound to the cells in patterns similar to those of the anti-nectin-1 and anti-HVEM antibodies. CHO cells expressing mouse nectin-1 or HVEM bound both HSV-1 and HSV-2 gD:Fcs in patterns indistinguishable from those shown in Fig. for the human receptors (data not shown). Control CHO cells devoid of exogenous receptors failed to bind the anti-nectin-1 antibodies, anti-HVEM antibodies or gD:Fcs (data not shown). We have noted previously the colocalization of anti-nectin-1 antibodies and gD:Fcs at cell contact sites and the enhancement of gD:Fc binding when cell adhesions are disrupted, indicating that homophilic trans-interactions of nectin-1 can reduce, but not prevent, gD:Fc binding . In the first 37 amino acids of HSV-1 gD, which includes the N-terminal hairpin present in the gD-HVEM complex, there are only two amino acid differences from HSV-2 gD (A7P and D21N), as shown in Fig. . Thus, the amino acid substitutions previously characterized in HSV-1 could also be investigated in HSV-2. The mutations introduced into HSV-1(KOS) and HSV-2(333) gDs and gD:Fcs are listed in Table , along with the names of the relevant plasmids. These mutations included the U10 (L25P) and Rid (Q27P and Q27R) substitutions and combinations of both. They also included a triple substitution (L25P/Q27R/T230I) found in HSV-1(ANG), which results in a phenotype similar to that of the Rid mutants . In addition, various deletions encompassing the regions in gD that make contact with HVEM were made. The wild-type and mutant forms of the gD:Fcs were tested by immunofluorescence for their ability to bind to cells expressing nectin-1 or HVEM. The results (not shown) can be summarized briefly. All of the HSV-1 and HSV-2 mutant gD:Fcs bound to cells expressing either the human or mouse forms of nectin-1, in a pattern similar to that observed for the wild-type gD:Fcs . On the other hand, all of the mutant gD:Fcs, except for L25P, failed to bind to cells expressing either the human or mouse forms of HVEM. Both the HSV-1 and HSV-2 L25P mutant forms of gD:Fc bound to cells expressing either HVEM or nectin-1 in patterns indistinguishable from wild-type. To quantify the binding to nectin-1 and HVEM, serial dilutions of the wild-type and mutant gD:Fcs were incubated with cells expressing human or mouse nectin-1 or HVEM in the CELISA assay. Figure shows the results obtained with the HSV-1 and HSV-2 substitution mutants, and Fig. shows the results obtained with the deletion mutants. All of the mutants bound to human or mouse nectin-1 with certain quantitative differences observed. Specifically, all of the HSV-1 substitution mutants bound more efficiently than wild-type gD:Fc to nectin-1, human or mouse, and all of the HSV-2 substitution mutants, except for L25P, also bound with greater efficiency (based on equivalent binding at input concentrations 1/3 to 1/10 that of wild-type gD:Fc). The HSV-2 L25P mutant was indistinguishable from wild-type HSV-2 gD:Fc in its binding to nectin-1. All of the deletion mutants bound to nectin-1 as efficiently as did wild-type forms of the gD:Fcs, except for HSV-2 deletion 7-21, which required up to 10 times higher concentrations for binding equivalent to that of wild-type gD:Fc. On the other hand, none of the substitution or deletion mutants bound to either form of HVEM, except for HSV-1 and HSV-2 L25P, a finding consistent with the immunofluorescence results (not shown). Effects of mutations in the N termini of HSV-1 and HSV-2 gDs on cell fusion. | It was previously shown that cells expressing the Q27P mutant of HSV-1 gD, along with wild-type forms of HSV-1 gB, gH, and gL, had significantly reduced ability to induce fusion with target cells expressing HVEM, enhanced ability to induce fusion with cells expressing nectin-2, and unchanged ability to induce fusion with cells expressing nectin-1 , a finding consistent with the entry phenotypes of Rid mutant viruses. None of the other gD mutations described here have been tested previously for their effects on cell fusion. CHO cells were transfected with each wild-type or mutant form of HSV-1 or HSV-2 gD, along with wild-type forms of HSV-1 or HSV-2 gB, gH, and gL in homologous combinations. These effector cell populations were then divided, and portions were replated in 96-well dishes for CELISA assays to quantitate the cell surface expression of gD, while the remainder of each population was mixed with target cells expressing each of the known HSV-1 or HSV-2 fusion/entry receptors for cell fusion assays. Cell fusion was quantitated by measuring the activity of luciferase, which can be expressed only after effector and target cells have fused. The results are presented in Fig. for the substitution mutants and in Fig. for the deletion mutants. Panels A and C in both figures show that the mutant and wild-type forms of gD were expressed at equivalent levels on the surfaces of the effector cells. The cell fusion results (panels B and D in Fig. and ) demonstrate the following. (i) All of the amino acid substitutions in HSV-1 gD, except L25P, reduced cell fusion with human HVEM to ca. 5% of the wild-type level and with mouse HVEM to 10 to 20% of the wild-type level. Similar results were obtained with the HSV-2 substitution mutants except that the reduction in levels of cell fusion was not quite so pronounced. (ii) All amino acid substitutions at positions 25 and/or 27 in HSV-1 gD significantly increased cell fusion activity with target cells expressing human nectin-2 and reduced fusion with cells expressing 3-OST-3B. These mutations in HSV-2 gD were without effect on the already significant level of cell fusion observed with wild-type HSV-2 gD and human nectin-2. (iii) All of the amino acid substitutions in HSV-2 gD reduced cell fusion with parental CHO-K1 cells to 20 to 35% of the wild-type level. (iv) All of the deletions in HSV-1 or HSV-2 gD significantly reduced cell fusion with cells expressing any of the receptors except nectin-1. Values close to the background values found in the absence of gD were noted for some of the deletions, particularly for those missing at least amino acids 24 to 32 and for all target cells except those expressing nectin-1 and those expressing human nectin-2 or 3-OST-3B when tested with the HSV-1 mutants. (v) None of the amino acid substitutions or deletions had any appreciable effect on cell fusion when the target cells expressed either human or mouse nectin-1. Figure demonstrated that deletion 7-21 in HSV-2 gD, but not in HSV-1 gD, reduced the efficiency of binding of the soluble gD:Fc to nectin-1; however, neither mutation had any discernible effect on fusion with nectin-1-expressing cells . It is possible that (i) the fusion assay cannot discriminate among different levels of activity above a certain threshold, (ii) the different forms of gD used in the binding and fusion assays were affected differently by the deletion in the case of HSV-2 (both forms were checked for presence of the appropriate deletion and absence of unwanted mutations), or (iii) a wide range of binding affinities can be associated with the same level of fusion activity. As we noted previously, mutations in nectin-1 that reduced gD binding did not necessarily affect HSV entry activity . A deletion of 7 to 21 amino acids in HSV-1(F) gD was previously characterized . Virus expressing this mutant gD appeared indistinguishable from wild-type HSV-1(F) for plaque formation on Vero cells and replication in a human cell line, suggesting that the receptor used by this mutant might have been nectin-1. We can conclude that the N-terminal region of either HSV-1 or HSV-2 gD is dispensable for the binding of gD to nectin-1 and for fusion with cells expressing nectin-1, whereas this region is required for full activity with all of the other receptors tested. Moreover, the amino acid sequence within the N terminus governs the level of activity with each of these receptors. DISCUSSION : We were surprised to find that all but one of the mutations described here had effects on functional interactions of both HSV-1 and HSV-2 gDs with all of the known HSV fusion/entry receptors except nectin-1. The results with HVEM were largely predictable, but the other results were not, for the reasons given below. The importance of amino acid Q27 in the HSV-1 gD-HVEM interaction was evident from the crystal structure of the HVEM-gD complex . This showed that Q27 is part of a short extended strand that forms hydrogen bonds with a beta strand in HVEM. The side chain of Q27 points toward an alpha-helix in gD that is sandwiched between the N-terminal hairpin and the immunoglobulin fold. This side chain inserts into a pocket too small to accommodate an arginine residue (as in Q27R). The Q27P substitution would be expected to distort the beta-strand interaction. New information presented here is that mutations in the same position in HSV-2 gD also disrupted physical and functional interactions with HVEM. It is not yet known how similar are the structures of HSV-1 and HSV-2 gD. An earlier study performed with soluble truncated forms of gD (306 amino acids instead of 285 in the form that was crystallized) used circular dichroism to estimate the secondary structure content of both HSV-1 and HSV-2 gDs . HSV-2 gD exhibited properties consistent with little or no alpha-helical content compared to 10% for HSV-1 gD. Also, the beta-sheet content was estimated to be higher for HSV-2 gD than for HSV-1 gD. However, similar or different the structures prove to be, substitutions Q27P or Q27R in either HSV-1 or HSV-2 gD disrupt interactions with HVEM, both human and mouse. The gD amino acid L25 also points toward the alpha-helix mentioned above and, although not part of the extended strand that forms hydrogen bonds with the beta-strand in HVEM, the L25P substitution might be expected to affect this interaction. The carbonyl group of L25 forms a hydrogen bond via a water molecule with the hydroxyl group of Y23 in HVEM. Despite these structural findings, our results show that the L25P substitution in HSV-1 or HSV-2 gD had little or no effect on the binding of gD:Fc to HVEM or on fusion with cells expressing HVEM. This mutation in HSV-1(F)U10 did, however, reduce the efficiency of viral entry into cells expressing HVEM. Perhaps reduced efficiency of binding and cell fusion would also have been observed with the L25P mutation in HSV-1(F) gD, if tested as the HSV-1(KOS) gD mutants were tested. Alternatively, other factors may influence how the mutant form of gD functions in the context of the virion. The key point, however, is that not all of the identified contacts between HVEM and gD are equally important for a stable and functional interaction. This point was also recently documented in a study investigating the effects of amino acid substitutions in HVEM on interactions with HSV-1 gD . Interestingly, substitutions in two of the HVEM residues (T35 and V36) that make contacts with Q27 reduced binding to HSV-1 gD but had little effect on entry activity of the altered HVEM. Substitutions at Y23 completely abolished binding to gD and viral entry activity. Both findings seem contradictory to the ones presented here, but the latter finding, at least, can be explained by the fact that Y23 makes extensive contacts with other residues of gD and not just L25. The most surprising findings were that the gD mutations had such different effects on functional interactions with human nectin-1 and nectin-2. These two proteins are related in sequence and structure, and homologous regions of the two receptors have been found to be critical for HSV entry . Amino acid substitutions in loops between the putative F and G beta strands and the C' and C" beta strands of the N-terminal immunoglobulin folds of both nectin-1 and nectin-2 have been shown to reduce HSV entry activity. Also, other amino acid substitutions in the loop between the C' and C" beta strands of human and mouse nectin-2 can enhance or confer entry activity for both HSV-1 and HSV-2. It seems reasonable to predict that contact sites on HSV-2 gD for these two receptors are similar, and yet the N terminus of gD is required for fusion activity with nectin-2 but not with nectin-1. Moreover, the amino acid sequence in the N terminus influences the level of fusion activity with nectin-2. This is evident not only from the effects of mutations in the N terminus of HSV-1 gD but also from results obtained in a study designed to identify the regions of HSV-2 gD that permit fusion activity with nectin-2. The present study showed that as many as 7 amino acid differences between HSV-1 and HSV-2 in the first 53 amino acids, i.e., upstream of the immunoglobulin fold, are largely responsible for the differences in ability of HSV-1 and HSV-2 gDs to induce fusion with cells expressing human nectin-2 . The results presented here show that contact regions for nectin-1 must be present downstream of amino acid 32 in gD. It seems likely, based on the considerations outlined above, that these regions would also make contact with nectin-2 but, in the case of nectin-2, would not be sufficient for fusion activity. Possibly there are additional contact regions for both nectin-1 and nectin-2 within the N-terminal 32 amino acids, and these are necessary for functional activity with nectin-2 but not with nectin-1. Alternatively, portions of gD within the first 32 amino acids may influence the conformation of the rest of gD in ways that are critical for nectin-2 binding and function for not for nectin-1. Dissociation constants for the interaction of HSV gD with nectin-1 are in the micromolar range , whereas it has not been possible to determine dissociation constants for gD-nectin-2 interactions. It seems likely that the affinity of HSV-2 or HSV-1 Rid gD for nectin-2 is lower than for nectin-1 or that the trans-homophilic interactions of nectin-2 dimers are of higher affinity than those of nectin-1 dimers and less easily disrupted to allow for gD binding or both. Physical interactions of HSV-1 Rid gD or HSV-2 gD with human nectin-2 have not been demonstrable by the CELISA used here, by ELISA, or by plasmon resonance, whereas all three methods have been used to demonstrate interactions of gD with nectin-1 . Only by immunofluorescence have we been able to demonstrate a specific colocalization of HSV-1 Rid1 gD with human nectin-2, but not with a mutant of nectin-2 devoid of viral entry activity . Negative effects of the N-terminal mutations in HSV-1 gD on fusion activity with cells expressing 3-OST-3B are consistent with the tentative identification of putative heparan sulfate-binding domains in gD . The electron density maps of the gD crystals included two large spherical features tentatively identified as sulfate ions. These features were found in basic domains that are candidates for heparan sulfate-binding sites. One such region was a positively charged pocket formed by beta strands of the N-terminal hairpin and including amino acids K1, R35, and R36. The new information presented here includes the effect of the L25P substitution in HSV-1 gD on functional interactions with HVEM, the effects of amino acid substitutions in the N terminus of HSV-2 gD on interactions with all of the HSV-2 fusion/entry receptors, the effects of deletions in the N termini of both HSV-1 and HSV-2 gDs on interactions with all of the HSV-1 and HSV-2 fusion/entry receptors, and the testing of wild-type and mutant forms of HSV-1 and HSV-2 gDs for functional interactions with all of the mouse fusion/entry receptors. The latter results extend our findings that mouse nectin-1 is a functional entry receptor for both HSV-1 and HSV-2, as well as for PRV and BHV-1 . Viral entry results for mouse HVEM have not previously been reported. We show here that mouse HVEM is indistinguishable from human HVEM in its ability to mediate entry of HSV-1(KOS), HSV-1(F), and HSV-1(F)U10 and in its inability to mediate entry of HSV-1(KOS)Rid1. In a previous discussion and review , we had mentioned unpublished preliminary results suggesting that mouse HVEM could mediate the entry of HSV-1 Rid mutants. It is possible that mouse HVEM is not quite as inactive for HSV-1 Rid entry as human HVEM, a finding consistent with the cell fusion results shown in Fig. . The next stage of this work is to introduce some of the mutant forms of gD into the viral genome to test whether the receptor specificities observed in the cell fusion assay also apply to viral entry. We predict that some of the HSV-1 and HSV-2 deletion mutants will be able to infect cells expressing only nectin-1, among the known HSV entry receptors. Such mutants will be invaluable for various studies such as identifying the key receptors for entry of virus into various cell types in culture and in vivo, and for investigating viral signaling via various receptors. FIG. 1. : (A) Amino acid sequence alignments for the N termini of HSV-1(KOS) and HSV-2(333) gDs. (A) Amino acid sequence alignments for the N termini of HSV-1(KOS) and HSV-2(333) gDs. Two segments (amino acids 7 to 15 and 24 to 32, underlined) in HSV-1 gD were shown to be engaged in the interface with HVEM . Amino acid substitutions in HSV-1 gD at positions 25 (L25P) and 27 (Q27P or Q27R) have previously been described , as discussed in the text. Wild-type forms of HSV-1 and HSV-2 gDs differ at two positions within this region, as marked by the colons. (B) Effects of the Q27P mutation in HSV-1(KOS) and the L25P mutation in HSV-1(F) on entry of virus into CHO cells expressing human or mouse HVEM. In the left panels, CHO-nectin-1 cells, CHO-HVEM cells, CHO-K1 cells transfected with plasmids expressing mouse nectin-1 or mouse HVEM, and control CHO-K1 cells were inoculated with serial dilutions of the beta-galactosidase-expressing viruses, HSV-1(KOS)tk12 or HSV-1(KOS)Rid1/tk12. In the right panels, CHO-IEbeta8 cells (which can be induced to express beta-galactosidase by infection with viruses carrying HSV VP16) were transfected with plasmids expressing the receptors as indicated or empty vector and were then inoculated with serial dilutions of HSV-1(F) or HSV-1(F)U10. After 6 h, the cells were lysed for the quantitation of beta-galactosidase activity as a measure of viral entry. Values were optical densities at 410 nm (OD410) of the ONPG reaction product; the means of triplicate determinations with standard deviations for one representative experiment are shown. Similar results were obtained in three additional experiments. FIG. 2. : Localization of human nectin-1 or HVEM in CHO-nectin-1 or CHO-HVEM cells by immunofluorescence. Localization of human nectin-1 or HVEM in CHO-nectin-1 or CHO-HVEM cells by immunofluorescence. The cells on coverslips were fixed and stained by incubation with an anti-nectin-1 monoclonal antibody (B and E) or an anti-HVEM rabbit serum (D and F), followed the addition of by Alexa 488-labeled secondary antibodies. Phase-contrast images paired with panels B and D are shown in panels A and C, respectively. Replicate cells were incubated with HSV-1 gD:Fc (G and H) or HSV-2 gD:Fc (I and J), followed by fixation and incubation with Alexa 488-labeled anti-Fc antibodies. Vertical lines across the larger panels in E to J indicate the plane of the z sections shown to the left. Horizontal lines indicate the plane of the z sections shown at the top for the antibody staining only. Bar, 10 mum. FIG. 3. : Quantitation of the binding of wild-type and mutant (substitution mutant) gD:Fcs to CHO cells expressing human and mouse forms of nectin-1 and HVEM. Quantitation of the binding of wild-type and mutant (substitution mutant) gD:Fcs to CHO cells expressing human and mouse forms of nectin-1 and HVEM. CHO-nectin-1 cells, CHO-HVEM cells, or CHO-K1 cells transfected with mouse nectin-1 (mNectin-1) or mouse HVEM (mHVEM) were incubated with serial dilutions of the wild-type or substitution mutant forms of gD:Fc indicated (HSV-1 [A] and HSV-2 [B]). The cells were then washed, fixed, and incubated with an anti-Fc detection system. The horse radish peroxidase reaction product was quantitated by determining the OD370. The values presented are means and standard deviations of triplicate determinations. The results presented for this experiment are representative of two other experiments. Symbols: *, L25P; , Q27P; , Q27R; , L25P/Q27P; , L25P/Q27R; , L25P/Q27R/T230I; , wild type. FIG. 4. : Quantitation of the binding of wild-type and mutant (deletion mutant) gD:Fcs to CHO cells expressing human and mouse forms of nectin-1 and HVEM. Quantitation of the binding of wild-type and mutant (deletion mutant) gD:Fcs to CHO cells expressing human and mouse forms of nectin-1 and HVEM. CHO-nectin-1 cells, CHO-HVEM cells or CHO-K1 cells transfected with mouse nectin-1 (mNectin-1) or mouse HVEM (mHVEM) were incubated with serial dilutions of the wild-type or deletion mutant forms of gD:Fc indicated (HSV-1 [A] and HSV-2 [B]). The cells were then washed, fixed, and incubated with an anti-Fc detection system. The HRP reaction product was quantitated by determining the OD370. The values presented are means and standard deviations of triplicate determinations. The results presented for this experiment are representative of two other experiments. Symbols: *, Del 7-21; , Del 7-15; , Del 24-32; , Del 7-15/24-32; , Del 7-32; , wild type. FIG. 5. : Expression of wild-type and mutant (substitution mutant) forms of gD on the surfaces of effector cells and fusion of these cells with target cells expressing various human and mouse fusion receptors. Expression of wild-type and mutant (substitution mutant) forms of gD on the surfaces of effector cells and fusion of these cells with target cells expressing various human and mouse fusion receptors. CHO-K1 cells were transfected with mixtures of plasmids expressing HSV-1 (A and B) or HSV-2 (C and D) gB, gD (wild-type [WT] or substitution mutant forms as indicated), gH, and gL. In control mixtures, the gD plasmid was left out and replaced with empty vector (NO gD). The cell populations were then divided, and a portion was plated in 96-well plates for CELISA (A and C). The remaining cells were mixed with the various target cells indicated for cell fusion assays (B and D). In panels A and C, anti-gD bound is expressed as HRP product measured at OD370. In panels B and D, cell fusion is expressed as units of luciferase, normalized to the values obtained with wild-type gD on HVEM-expressing cells. The results are expressed as the means and standard deviations of three independent experiments, each done in triplicate. FIG. 6. : Expression of wild-type and mutant (deletion mutant) forms of gD on the surfaces of effector cells and fusion of these cells with target cells expressing various human and mouse fusion receptors. Expression of wild-type and mutant (deletion mutant) forms of gD on the surfaces of effector cells and fusion of these cells with target cells expressing various human and mouse fusion receptors. CHO-K1 cells were transfected with mixtures of plasmids expressing HSV-1 (A and B) or HSV-2 (C and D) gB, gD (wild-type [WT] or deletion mutant forms as indicated), gH, and gL. In control mixtures, the gD plasmid was left out and replaced with empty vector (NO gD). The cell populations were then divided, and a portion was plated in 96-well plates for CELISA (A and C). The remaining cells were mixed with the various target cells indicated for cell fusion assays (B and D). In panels A and C, anti-gD bound is expressed as HRP product measured at OD370. In panels B and D, cell fusion is expressed as units of luciferase, normalized to the values obtained with wild-type gD on HVEM-expressing cells. The results are expressed as the means and standard deviations of three independent experiments each done in triplicate. TABLE 1 : Summary of gD mutations and plasmids encoding the wild-type and mutant genes Backmatter: PMID- 12915581 TI - Differences in the N Termini of Herpes Simplex Virus Type 1 and 2 gDs That Influence Functional Interactions with the Human Entry Receptor Nectin-2 and an Entry Receptor Expressed in Chinese Hamster Ovary Cells AB - Amino acid differences at seven positions in the N termini of the glycoproteins D (gDs) specified by herpes simplex virus type 1 (HSV-1) and HSV-2 are largely responsible for the significantly higher cell fusion activity of HSV-2 gD with Chinese hamster ovary cells expressing human nectin-2 or only an endogenous hamster receptor. Keywords: Introduction : Herpes simplex virus type 1 (HSV-1) and HSV-2 cause mucocutaneous lesions that are indistinguishable clinically, but the two serotypes exhibit differences in gene sequences, biology, pathogenesis, and epidemiology. In addition, the two serotypes exhibit some differences in the efficiency of cell entry, dependent on the entry receptors expressed, as outlined below. Heparan sulfate chains on cell surface proteoglycans provide initial binding sites for both HSV-1 and HSV-2, but each serotype recognizes somewhat different structural features of heparan sulfate . After the binding of HSV-1 or HSV-2 to heparan sulfate, another viral glycoprotein, glycoprotein D (gD), engages any one of several cell surface receptors to initiate the process of viral entry . Viral entry occurs by fusion of the viral envelope with a cell membrane and requires the envelope glycoproteins gB, gH, and gL, as well as gD, and a gD receptor on the cell. The human HSV gD receptors include herpesvirus entry mediator (HVEM) , a member of the tumor necrosis factor receptor family; nectin-1 and nectin-2 , members of the immunoglobulin superfamily that are cell adhesion molecules present in cadherin-based adherens junctions ; and sites in heparan sulfate generated by the action of specific 3-O-sulfotransferases . HSV-1 and HSV-2 differ in their ability to use the various gD receptors for viral entry. Human HVEM or nectin-1 can serve as entry receptors for both serotypes, whereas nectin-2 has a higher entry activity for HSV-2 strains than for HSV-1 strains, and 3-O-sulfated heparan sulfate has a higher entry activity for HSV-1 strains than for HSV-2 strains. Also, Chinese hamster ovary (CHO) cells express an endogenous receptor for HSV-2 entry that has only minimal activity for HSV-1 entry . The activity of the HSV entry receptors can be assessed by quantitative viral entry assays , as well as by quantitative cell fusion assays in which cells expressing HSV gB, gD, gH, and gL are tested for their ability to fuse with cells bearing entry/fusion receptors . The activities of the viral glycoproteins can be assessed similarly by using mutant viruses in viral entry assays or mutant forms of the viral glycoproteins in cell fusion assays. The purpose of the present study was to define the viral determinants responsible for the greater ability of HSV-2, compared to wild-type HSV-1, to interact functionally with human nectin-2 and the endogenous CHO receptor. HSV-2 gD confers enhanced fusion activity with the CHO receptor and nectin-2. : Since gD is the ligand for all known HSV entry receptors, it seemed likely that HSV-2 gD determines the enhanced fusion activity of HSV-2 glycoproteins with the CHO receptor and nectin-2. To test this hypothesis, quantitative cell fusion assays were done with the homologous set of HSV-1 or HSV-2 glycoproteins or with mixtures in which each of the glycoproteins in the HSV-1 or HSV-2 set was substituted with its heterologous counterpart. CHO cells were transfected with plasmids expressing the appropriate mixtures of HSV glycoproteins and T7 polymerase (termed effector cells). The plasmids used for expression of the HSV-1(KOS) glycoproteins were previously described . A number of plasmids were generated for the present study, as described in Table , and the inserts were sequenced to confirm the desired coding of amino acid sequence. CHO cells stably expressing HVEM, nectin-1, or nectin-2 or control CHO cells were transfected with a plasmid carrying the firefly luciferase gene linked to a T7 promoter (termed target cells). The effector and target cells were then mixed in 1:1 ratios, plated, and incubated for 18 h prior to lysis for quantitation of luciferase as a measure of cell fusion . Figure shows that the homologous mixtures of HSV-1 and HSV-2 glycoproteins induced levels of cell fusion with each receptor consistent with viral entry assays done with the same receptors . The cell fusion assay may be more sensitive in that HSV-1 glycoproteins exhibited low but detectable activities with the endogenous CHO receptor and nectin-2, whereas HSV-1(KOS) entry with these receptors is more difficult to detect. Figure also shows that substitution of HSV-1 gD (gD1) with gD2 in the otherwise HSV-1 set significantly enhanced cell fusion with target cells expressing the CHO receptor only or also nectin-2 but not with target cells expressing nectin-1 or HVEM. In the converse combination, substitution of gD2 with gD1 in the otherwise HSV-2 set significantly reduced cell fusion with the CHO receptor and nectin-2 but not with nectin-1 or HVEM. Thus, gD is the major determinant of the differential ability of HSV-1 and HSV-2 glycoproteins to induce cell fusion with the CHO receptor and nectin-2. Other findings emerged from the results presented in Fig. . Substitution of gB2 with gB1 in the HSV-2 set significantly enhanced fusion, but only with target cells expressing nectin-1 and HVEM, whereas substitution of gB1 with gB2 in the HSV-1 set reduced cell fusion with target cells expressing any of the receptors. It has been reported that deletions from the cytoplasmic C terminus of gB2 can enhance cell fusion induced by HSV-2 glycoproteins with cells in which the fusion receptors were not identified and that the effect was not necessarily correlated with altered cell surface expression of this glycoprotein . We have confirmed these findings and showed further that substitution of wild-type gB2 with truncated gB2 had the same effect on cell fusion with nectin-1 and HVEM (A. Zago and P. G. Spear, unpublished results), as did substitution with gB1 . It seems likely that the cytoplasmic tail of wild-type gB2 has a domain that is inhibitory for cell fusion, at least with certain fusion receptors, whereas such an inhibitory activity in wild-type gB1 is not so evident. The greater ability of HSV-1 glycoproteins, compared to HSV-2 glycoproteins, to induce cell fusion with nectin-1 and HVEM is probably accounted for by this inhibitory domain in gB2. We chose to use wild-type gB2 in the present study. Substitutions of gH and gL were not done separately because these glycoproteins are known to function as a heterodimer . With nectin-1 and HVEM, substitutions of gH/gL in either combination with the other glycoproteins had little effect on cell fusion . The same was true for the CHO receptor and nectin-2 when gH1/gL1 was substituted with gH2/gL2. However, substitution of gH2/gL2 with gH1/gL1 in the otherwise HSV-2 mixture reduced cell fusion activity with the CHO receptor and nectin-2. This phenomenon requires further exploration beyond the scope of the present study. Mapping of domains in gD2 that confer cell fusion activity with the CHO receptor and nectin-2. : Plasmids expressing hybrid forms of gD were constructed . These hybrids had the first 66 or 98 amino acids from gD1 or gD2 and the remainder from gD2 or gD1, respectively. Plasmids expressing the hybrid proteins were used in cell fusion assays, along with plasmids expressing the wild-type forms of gD1 or gD2, in combination with the other HSV-1 or HSV-2 glycoproteins. The results presented in Fig. show that the gD2/1 hybrid having the first 66 amino acids from HSV-2 resembled gD2 in its enhancement of fusion with target cells expressing the CHO receptor only or also nectin-2 when combined with the HSV-1 glycoproteins. The activity of the gD2/1 hybrid was ca. 60 to 80% that observed with gD2 regardless of whether the other glycoproteins were from HSV-1 or HSV-2 and regardless of receptor (except that the gD2/1 hybrid had activity comparable to that of gD2 when the receptors were nectin-1 or HVEM and the other glycoproteins were from HSV-2). Also, the converse gD1/2 hybrid resembled gD1 in its poor cell fusion activity with target cells expressing the CHO receptor only or also nectin-2 when combined with either the HSV-1 or HSV-2 glycoproteins. The gD1/2 hybrid was slightly more active than the gD2/1 hybrid with the other HSV-1 glycoproteins and slightly less active with the other HSV-2 glycoproteins when the receptors were nectin-1 and HVEM. Results similar to those shown in Fig. for CHO cells were also obtained with hybrids in which the first 98 amino acids were switched. We conclude from these results (i) that the hybrids had no gross conformational abnormalities and had fusion activities comparable to those of gD1 and gD2, at least when tested with gB2 and gH2/gL2 and the fusion receptors, nectin-1 and HVEM, and (ii) that the superior activity of gD2 in fusion activity with the CHO receptor and nectin-2 is largely due to amino acid differences between gD2 and gD1 within the first 66 amino acids. Figure shows an alignment of the amino acid sequences of gD1 and gD2 in the first 66 amino acids. There are seven differences in sequence, five of which are also noted on a backbone trace of the structure of gD1 . For the five positions most likely to affect secondary and tertiary structure (i.e., positions 7, 21, 42, 43 and 45), gD1 was mutated by substituting the amino acid present in gD1 for the one present in gD2. These substitutions were made individually and in pairs (A7P, A7P/P45E, A7P/D21N, A42P/G43S, and P45E). When these mutant forms of gD1 were tested for ability to induce the fusion of target cells expressing the CHO receptor only or also nectin-2, none exhibited as high as 50% of the activity observed with gD2, using either the HSV-1 or HSV-2 forms of gB and gH/gL, except in the case of the HSV-1 set of glycoproteins tested with target cells expressing only the CHO receptor. Only the gD1 mutants carrying the A7P substitution, either alone or in combination with P45E or D21N, exhibited enhanced cell fusion activity (50 to 60% that of gD2) under these conditions (data not shown). None of the gD1 mutants tested with either HSV-1 or HSV-2 gB and gH/gL differed significantly from wild-type gD1 in cell fusion activity when the receptors were nectin-1 or HVEM (data not shown). We conclude that the A7P substitution in gD2 contributes to the higher level of fusion activity of gD2 but that multiple substitutions within the first 53 amino acids of gD2 (the domain containing all of the differences between gD1 and gD2 in the first 66 amino acids), including A7P, are required for the gD2-like levels of cell fusion activity with the CHO receptor and nectin-2. The levels of gD expressed on the surfaces of glycoprotein-expressing cells used in the cell fusion assays were assessed by an immunoassay, in which the live cells in the monolayer were incubated with a type-common anti-gD rabbit serum, R#7, and then fixed and incubated with a detection system as previously described . Comparable levels of cell surface expression were consistently noted for gD1, gD2, all of the gD1/gD2 hybrids, and all HSV-1 mutants (data not shown), indicating that the differences in cell fusion observed could not be explained by expression levels of gD. The results presented here and elsewhere show that the amino acid sequence in the N-terminal region of gD, outside of the Ig-fold, influences whether human nectin-2 and an endogenous CHO receptor can serve as entry/fusion receptors. We showed here (i) that gD2 is more active than gD1 in inducing the fusion of cells expressing human nectin-2 or only an endogenous CHO receptor, regardless of whether the viral glycoproteins gB, gH, and gL were from HSV-1 or HSV-2 and consistent with the greater activity of these receptors for HSV-2 entry , and (ii) that as many as 7 amino acid differences between gD1 and gD2 in the first 53 amino acids of the N terminus are largely responsible for the greater activity of gD2. It was previously shown that specific amino acid substitutions in the N terminus of gD1 enable functional interactions with human nectin-2. Substitutions Q27P or Q27R in gD1 confer the ability of HSV-1 to use nectin-2 as an entry or fusion receptor . Also, substitution L25P in gD1 enables HSV-1 to use nectin-2 as an entry receptor . Interestingly, the mutations at position 27 were also shown to enhance, by about 10-fold, the limited ability of HSV-1 strain KOS to infect CHO cells . Other effects of the mutations at position 27 in gD1 include enhanced affinity of isolated soluble mutant gD1 for nectin-1 and significantly reduced ability of HSV-1 to use HVEM as an entry or fusion receptor , a finding consistent with the location of amino acid 27 in a contact region identified in the X-ray structures . It should be noted that L25 and Q27, and in fact all of the positions in the contact region from positions 24 to 32, are identical in gD1 and gD2. Thus, substitutions at positions 25 or 27 enable gD1 to interact functionally with nectin-2 but are not necessary for gD2 to interact with nectin-2. Attempts were made to isolate the endogenous CHO receptor, without success. Oligonucleotides matching conserved sequences in nectin-1 and nectin-2 were tested for the ability to prime amplification of homologous sequences from reverse transcripts of RNA extracted from CHO cells or from plasmids extracted from a library of CHO expression clones. Antibodies raised against human or mouse nectin-2 were tested for their ability to bind to CHO cells by immunofluorescence. Identification of the endogenous CHO receptor would probably require expression cloning, which would in turn require a cell line with a high level of resistance to HSV-2 entry. Some pilot experiments done with the B78H1 cell line were not satisfactory. We abandoned the effort to identify the receptor because the results obtained in the present study were so similar for human nectin-2 and the endogenous CHO receptor, suggesting that the endogenous CHO receptor could be the Chinese hamster homolog of nectin-2 despite our inability to amplify nectin-2 gene sequences by PCR. Human nectin-2 and the endogenous CHO receptor are distinct because certain mutations in gD2 can inhibit cell fusion to a much greater extent with control CHO cells than with CHO cells expressing nectin-2 . Although the results summarized here indicate that the N-terminal regions of gD1 and gD2 influence entry and cell fusion with nectin-2 or the endogenous CHO receptor, further studies are required to determine whether the N terminus makes direct contact with these receptors, influences the conformation of other regions in gD that are involved in direct contacts, or influences interactions with the other viral glycoproteins that are required for cell fusion. FIG. 1. : Cell fusion induced by plasmids expressing homologous or heterologous mixtures of the HSV-1 or HSV-2 glycoproteins. Cell fusion induced by plasmids expressing homologous or heterologous mixtures of the HSV-1 or HSV-2 glycoproteins. The cell fusion assay was performed by transfecting CHO effector cells with plasmids expressing the indicated glycoproteins or vector alone (neg). The effector cells were mixed with CHO target cells expressing only the CHO endogenous receptor or also nectin-2, nectin-1, or HVEM. Effector cells transfected with an homologous set of HSV-1 or HSV-2 glycoproteins are the controls. The glycoprotein substituted is indicated with a minus (-), and the substituting protein is indicated with a plus (+). Cell fusion was quantified by measurements of luciferase activity (arbitrary units divided by 10 to give the numbers shown on the y axis). The values shown are means and standard deviations from one experiment performed in triplicate. The results presented are representative of at least two other additional experiments. FIG. 2. : Cell fusion induced by gD2/1 or gD1/2 hybrid molecules in combination with the other HSV-1 or HSV-2 glycoproteins. Cell fusion induced by gD2/1 or gD1/2 hybrid molecules in combination with the other HSV-1 or HSV-2 glycoproteins. These hybrids have the first 66 amino acids from gD1 or gD2 and the remainder from gD2 or gD1, respectively. The CHO effector cells were transfected with plasmids expressing the indicated glycoproteins or vector alone (neg). Effector cells were mixed with CHO target cells expressing only the CHO endogenous receptor or also nectin-2, nectin-1, or HVEM. Cell fusion was quantified by measurements of luciferase activity (arbitrary units divided by 100 to give the numbers shown on the y axis). The values shown are means and standard deviations from one experiment performed in triplicate. The results presented are representative of at least two other additional experiments. FIG. 3. : (A) Alignment of the 66 N-terminal amino acids of HSV-1(KOS) gD and HSV-2(333) gD. (A) Alignment of the 66 N-terminal amino acids of HSV-1(KOS) gD and HSV-2(333) gD. Dots indicate identity. The positions where the amino acids differ are indicated. (B) Backbone trace of the crystal structure of gD1 bound to HVEM . HVEM is shown in black, and gD1 is shown in gray. Residues that have been substituted in the gD1 mutants are displayed in different colors. TABLE 1 : Expression plasmids generated for this study Backmatter: PMID- 12915557 TI - Variability at Human Immunodeficiency Virus Type 1 Subtype C Protease Cleavage Sites: an Indication of Viral Fitness? AB - Naturally occurring polymorphisms in the protease of human immunodeficiency virus type 1 (HIV-1) subtype C would be expected to lead to adaptive (compensatory) changes in protease cleavage sites. To test this hypothesis, we examined the prevalences and patterns of cleavage site polymorphisms in the Gag, Gag-Pol, and Nef cleavage sites of C compared to those in non-C subtypes. Codon-based maximum-likelihood methods were used to assess the natural selection and evolutionary history of individual cleavage sites. Seven cleavage sites (p17/p24, p24/p2, NC/p1, NC/TFP, PR/RT, RT/p66, and p66/IN) were well conserved over time and in all HIV-1 subtypes. One site (p1/p6gag) exhibited moderate variation, and four sites (p2/NC, TFP/p6pol, p6pol/PR, and Nef) were highly variable, both within and between subtypes. Three of the variable sites are known to be major determinants of polyprotein processing and virion production. P2/NC controls the rate and order of cleavage, p6gag is an important phosphoprotein required for virion release, and TFP/p6pol, a novel cleavage site in the transframe domain, influences the specificity of Gag-Pol processing and the activation of protease. Overall, 58.3% of the 12 HIV-1 cleavage sites were significantly more diverse in C than in B viruses. When analyzed as a single concatenated fragment of 360 bp, 96.0% of group M cleavage site sequences fell into subtype-specific phylogenetic clusters, suggesting that they coevolved with the virus. Natural variation at C cleavage sites may play an important role, not only in regulation of the viral cycle but also in disease progression and response to therapy. Keywords: Introduction : One of the most dramatic changes in the human immunodeficiency virus type 1 (HIV-1)-AIDS epidemic has been the rapid emergence and devastating spread of subtype C viruses . HIV-1 C now accounts for >56% of all circulating viruses and is the most commonly transmitted subtype worldwide . Subtype C predominates in southern Africa and India and is increasing in frequency in China and Brazil . The disproportionate increase in C viruses relative to other HIV-1 strains suggests that subtype C may be more easily transmitted or that it has a higher level of "fitness" at the population level. One possible explanation is that founder effects relating to the ongoing introduction of subtype C into new population groups with different host factors, or different social and sexual practices, may be responsible for the rapid spread. However, founder and host effects cannot account for the fact that C viruses are overtaking preexisting virus subtypes in several different geographical regions, including Yunnan Province in China and the southern region of Brazil . It is increasingly evident that additional (nonhost) viral factors are also contributing to the rapid spread of HIV-1 C. Viral studies indicate that subtype C has distinct genetic and phenotypic properties that differentiate it from other HIV-1 subtypes. Various studies have postulated that an extra NF-kappaB binding site in the long terminal repeat , a prematurely truncated Rev protein , or a 5-amino-acid insertion in Vpu may influence viral gene expression, altering the transmissibility and pathogenesis of C viruses . Factors related to viral entry and pathogenesis, such as the CCR5 and non-syncytium-inducing properties of C isolates , may also contribute to the increased spread of C viruses. One area of research that is receiving consideration is the possibility that C viruses have a more active, catalytically efficient protease . The C protease is highly conserved at the amino acid level and has a distinct signature sequence that differentiates it from those of subtypes A, B, and D . A subset of these signature residues, present in the hinge (M36I/R41K/H69K) and alpha-helix (I89M) of C (and A) proteases, has been linked to increased catalytic activity . Another signature pattern, identified in >80% of C viruses from South Africa, is T12S/T15V/L19I/I93L . The 12S, 15V, and 19I residues of this motif are located near the N terminus of the protease in an extended beta-chain. The 93L polymorphism is located within a hydrogen-bonded turn immediately upstream from the protease-reverse transcriptase (RT) cleavage site, in close proximity to 12S/15V/19I and the dimerization domain . All of these polymorphisms lie outside the catalytic site of the protease, in regions that would be expected to alter the enzyme's activity toward its natural cleavage sites, leading to adaptive (compensatory) changes in the cleavage site itself. Since protease inhibitors (PIs) are currently the most active antiretroviral drugs used for the treatment of HIV-1 , it is important to collect information, not only on the C protease but also on its drug responsiveness, substrate-inhibitor specificities, and cleavage site characteristics. This information is needed to design PIs that are maximally effective against C viruses and to obtain new insights into the mechanism of drug resistance. Studies have shown that resistance mutations in the B protease are associated with impaired proteolytic processing, decreased enzymatic activity, and a failure to produce mature infectious virions . Compensatory cleavage site mutations can partially compensate for these defects . In this report, we describe the natural variability of subtype C cleavage sites in viruses from Africa, India, and Brazil and compare the results to cleavage site patterns in representative B and group M viruses. MATERIALS AND METHODS : Cleavage site characteristics. | The HIV-1 protease is a small, 99-amino-acid aspartic enzyme that mediates the cleavage of Gag, Gag-Pol, and Nef precursor polyproteins. These reactions occur late in the viral life cycle, during virion assembly and maturation at the cell surface. The process is highly specific, temporally regulated, and essential for the production of infectious viral particles (-, ). As shown in Fig. , the main structural proteins are formed by cleavage of the Pr55gag polyprotein into matrix (MA; p17), capsid (CA; p24), nucleocapsid (NC; p7), p6gag, and two spacer peptides, p2 and p1. The viral enzymes are formed by cleavage of Pr160gag-pol, a fusion protein derived by ribosomal frame shifting . Although Pr160gag-pol also contains p17, p24, and p2, its C-terminal cleavage products are NC, a transframe protein (TFP), p6pol, protease (PR), reverse transcriptase (RTp51), RNase H (RTp66), and integrase (IN) . In total, 12 proteolytic reactions are required to generate a mature infectious virion. Each reaction occurs at a unique cleavage site that differs in amino acid composition . Some cleavage sites contain phosphorylated Ser/Thr or Tyr residues that alter the sites' susceptibilities to cleavage . P6gag, the major phosphoprotein of HIV-1, plays an essential role in the release of virus from the membranes of infected cells . Sequence data and construction of cleavage site fragments. | A total of 84 full-length nucleotide sequences were selected for analysis. These sequences included two C isolates from South Africa, TV001 and TV002 , in addition to another 25 subtype C, 30 subtype B, and 27 representative group M reference sequences (including A [n = 3], B [n = 4], C [n = 5], D [n = 3], F1 and F2 [n = 4], G [n = 2], H [n = 2], J [n = 2], and K [n = 2] subtypes) extracted from the Los Alamos database . Sequences were selected based on the patient being treatment naive. Since the prevalence of drug resistance in untreated patients has been reported to range from 1 to 11% , sequences were also screened and excluded from the study if they were found to contain primary resistance mutations. The majority of the sequences were obtained by direct DNA PCR amplification and cloning of peripheral blood mononuclear cells. Nucleotide sequences were aligned by CLUSTAL W and manually edited with the codon alignment of the Genetic Data Environment (GDE version 2.2) program . Calculation of the pairwise distance matrix, phylogenetic inference, and tree construction were performed on a dual-processor Linux computer by using the PAUP version 4.0b2a program (Sinauer Associates, Sunderland, Mass.) and a GDE for Linux HIV-1 interface . Thirty-base-pair segments, consisting of 15 nucleotides (5 amino acids) on each side of the 12 cleavage sites, were extracted and concatenated into a 360-bp nucleotide sequence. Reconstruction of ancestral cleavage site sequences. | To examine the evolutionary histories of individual cleavage sites, Phylogenetic Analysis under Maximum Likelihood (PAML) software was used to identify amino acid and nucleotide substitutions along each branch of the tree. Branch lengths were estimated using a nucleotide substitution model; amino acid sequences were deduced from the reconstructed nucleotide triplets. The analyses involved the use of maximum-likelihood methods and a time-reversible model which assume different substitution rates, base frequencies, and transition/transversion rate ratios (kappa) . Using this approach, we were able to reconstruct the ancestral sequences and internal nodes for each of the 12 protease cleavage sites in the B, C, and M group data sets. The number of proximal ancestors for each data set was n - 1, which translated into 29 ancestral sequences for subtype B, 26 sequences for subtype C, and 26 sequences for the group M viruses. The most recent common ancestor (MRCA) nucleotide sequence for each virus in the three data sets was saved and translated into its corresponding amino acids. Diversity and cleavage site polymorphisms. | Nucleotide diversity at cleavage sites was measured using a Kimura 2-alpha parameter model with a distance matrix implemented in the MEGA program version 2.0 (Arizona State University, Tempe). Amino acid diversity was measured using a Poisson distribution method implemented in the same MEGA package. P values for diversity measurements were calculated by applying the t test to the distance matrix of each data set. To determine whether the sequences had evolved over time, amino acid profiles for individual Gag, Gag-Pol, and Nef cleavage sites were compared to the inferred MRCA for that site. Assessment of positive selection pressure. | Nucleotide sequences were also analyzed with Codeml, a program from the PAML software package . The likelihood ratio test and recently developed codon-based models were used to assess natural selection and adaptive evolution at the amino acid level. These selection models use maximum-likelihood scores to account for variation in the dn/ds (nonsynonymous/synonymous) ratio (omega) at individual codons along the length of the sequence. High rates of synonymous mutation are indicative of conservation and a strict requirement for biological function, while high rates of nonsynonymous substitution are indicative of adaptive change in response to host selection pressure. An individual amino acid was considered to be positively selected if the dn/ds ratio was significantly greater than 1.0. RESULTS : Viral characteristics. | No primary RT- or PI-resistant mutations were detected among the 84 full-length sequences selected for study. Although attempts were made to include only sequences amplified directly from HIV-1 proviral DNA and to match these sequences based on duration of infection, plasma viral load, and CD4 count, this proved difficult. A surprisingly small amount of full-length sequence data was available from treatment-naive patients infected with subtype B, and when available, it was often poorly annotated. Despite these limitations, the frequency and pattern of naturally occurring polymorphisms observed in this study were remarkably similar to those reported for a control group of subtype B infections treated with nucleoside reverse transcriptase inhibitors but not with PIs or nonnucleoside reverse transcriptase inhibitors . Most of the non-C sequences came from regions of the world where treatment is not yet readily available. Genetic diversity and patterns of amino acid variability at individual cleavage sites. | Variation at the 12 cleavage sites of subtypes B and C and group M is shown in Table . Seven (58.3%) sites (p17/p24, p24/p2, NC/p1, NC/TFP, PR/RT, RT/p66, and p66/IN) were found to be relatively well conserved, both over time and between subtypes, with a mean intrasubtype distance ranging from 0.40 +- 0.20% to 7.49 +- 5.62%. The remaining five (41.7%) sites exhibited moderate (p1/p6gag) to extensive (p2/NC, TFP/p6pol, p6pol/PR, and Nef) variation, with mean intrasubtype diversities reaching levels as high as 42.42 +- 15.16%. For the purposes of this study, we have referred to these three patterns as conserved, moderately variable, and variable. Polymorphisms were more common among C than B viruses (P < 0.0001). Overall, six cleavage sites (p17/p24, p2/NC, NC/p1, PR/RT, RT/p66, and Nef) had signficantly higher levels of diversity among C viruses (P < 0.0001); five sites (p24/p2, p1/p6gag, NC/TFP, p6pol/PR, and p66/IN) had similar levels of diversity in both subtypes, and one site, TFP/p6pol, was highly variable among B and group M viruses (mean distances, 16.7 and 24.7%, respectively) but relatively conserved in C viruses (mean distance, 7.6%) (P < 0.0001). Compared to the M data set, the level of polymorphism at C cleavage sites was as wide ranging as that observed for the entire M group, a data set containing nine different HIV-1 subtypes. One cleavage site, p17/p24, was significantly more diverse in subtype C (mean divergence, 5.15%) than in group M (mean divergence, 2.23%) and subtype B (mean divergence, 2.31%) viruses (<0.0001 for both comparisons). Seven sites (p24/p2, p2/NC, NC/p1, p6pol/PR, RT/p66, p66/IN, and Nef) exhibited similar levels of diversity in both data sets (P = 0.004 to 0.961). Only four group M cleavage sites (p1/p6gag, NC/TFP, TFP/p6pol, and PR/RT) had mean diversities that were significantly greater than that observed for subtype C (P < 0.0001). For >50% of the sites, the variability of B viruses was significantly lower than that observed for subtype C or group M (P < 0.0001), despite the fact that the B viruses covered a broader time frame. The observed polymorhpisms were not randomly distributed across the variable cleavage sites but were confined to specific amino acids, most of which were positively selected in C but not in B viruses . The least variable residues were the P1 positions of p1/p6gag, TFP/p6pol, and Nef and the P1' positions of p6pol/PR and Nef. As shown, these positions flank the scissile bond, with P1 located immediately upstream and P1' located immediately downstream of the cleavage junction. The most variable positions were the P1, P3, P4, P5, P3', and P5' residues of p2/NC; the p3' residue of p1/p6gag; the P1' and P4' residues of TFP/p6pol; the P1 to P4 residues of p6pol/PR; and the P3 to P5 and P2' residues of Nef. Subtyping and phylogenetic-tree analysis. | When subjected to phylogenetic analysis , the concatenated 360-bp fragments of the group M data set fell into eight subtype-specific clusters representing subtypes A, B and D, C, F, G, H, J, and K, with the cleavage sequences for subtypes B and D segregating together in the same subcluster. This pattern was supported by high bootstrap values, by high-score maximum-likelihood trees, and by phylogenetic analysis of the deduced amino acids. These findings reveal the subtype-specific nature of protease cleavage sites and suggest that the evolution of cleavage sites parallels that of the full-length genome. With the exception of C.98IN022, B.AR00.ARMS008, and B.US.P896, all of the cleavage site fragments in the B and C data sets segregated into two distinct monophyletic groups representing either subtype B or subtype C viruses (data not shown). The longer branch lengths in the C subcluster were reflective of the increased diversity of C viruses relative to subtype B. Identification and dating of common ancestors. | Maximum-likelihood methods were next used to reconstruct the internal nodes of the phylogenetic tree and to estimate the times of divergence of individual sequences from their MRCA. These estimates were determined by measuring the number of substitutions along each branch of the tree. MRCAs for the B, C, and group M data sets are shown in Fig. . Two different patterns were observed based on the relationship between a given sequence and its MRCA. Conserved (n = 7) and moderately variable (n = 1) cleavage sites shared the same (identical) MRCA among all three data sets. The proximal location of the MRCA relative to the root of the tree suggests that, for these sequences, cleavage site diversification occurred after subtype divergence. In contrast, variable cleavage sites (n = 4) showed a high degree of divergence both from their subtype-specific MRCA and from the group M MRCA. Ancestral nodes for the variable sites were located closer to the tips of the tree (data available upon request). Variability of cleavage sites relative to other regions of the HIV-1 genome. | For C viruses, the average intersequence divergence among the concatenated cleavage site fragments was higher (10.1%) than those observed for the Gag (9.8%) and Pol (5.8%) proteins but lower (16.5%) than that observed for Nef. In contrast, B cleavage sites were significantly less diverse (4.8%; P < 0.0001) than those of the Gag (7.0%), Pol (6.0%), and Nef (14.7%) proteins of subtype B. Among group M viruses, cleavage site diversity was significantly higher (12.1%) than that calculated for Pol (5.9%) but lower than that determined for Gag (15.7%) and Nef (18.5%). These results are presented in more detail in Table . Physical-chemical properties of amino acids at P1-P1' cleavage junctions. | Overall, excluding the highly conserved asparagine (N) residue at the P1 position of NC/p1 and NC/TFP, >97.0% of P1-P1' amino acids in group M were nonpolar. Of these, 77.8% were hydrophobic, 21.5% were small amino acids (78.9% proline, 20.8% alanine, and 0.3% glycine), 0.3% were polar uncharged (one serine and three glutamine), 0.2% were polar charged (one arginine and one aspartic acid), and 0.3% were ungrouped (cysteine) residues. These amino acids were localized at specific positions within the cleavage sequence. The small amino acids were localized primarily to the P1' position of p17/p24, p24/p2, TFP/p6pol, p6pol/PR, and PR/RT. Hydrophobic amino acids were concentrated at the P1-P1' junction. As previously reported for HIV-1 B , the P1-P1' amino acids of subtype C fell into two different patterns defined by the size of the P1' amino acid: type I, represented by p2/NC and p1/p6gag, and type II, represented by p17/p24 and p24/p2. Both types carried large nonpolar, hydrophobic amino acids (leucine, tyrosine, phenylalanine, and methionine) at position P1 and either a large (type I) or small (type II) hydrophobic amino acid (proline, alanine, or glycine) at P1'. MRCAs and subtype-specific signature patterns. | A summary of amino acid signature patterns relative to the subtype B and C and group M MRCAs is shown in Tables and . Mutations at cleavage sites defining the enzymatic (PR/RT, RT/p66, and p66/IN) and structural (p17/24, p24/p2, and NC/p1) components of HIV-1 were relatively uncommon and, when detected, were found at greater frequencies among C versus B viruses. In total, only 1.2, 0.8, and 1.1% of the 840 amino acids at each of the PR/RT, p66/IN, and p24/p2 cleavage sites carried substitutions. As a result, the majority of sequences at conserved sites (81.5 to 96.7%) were identical to both the subtype-specific MRCA and the common ancestor of the group M viruses. Several cleavage sites involved in the regulation of polyprotein processing and protease activation, p2/NC, TFP/p6pol, and p6pol/PR, were highly variable and differed significantly from both the subtype-specific and group M MRCAs. With the exception of the TFP/p6pol site, which was more variable in subtype B, divergence from the MRCA was greatest for C and M viruses. None of the p2/NC and p6pol/PR sequences in the group M data set was identical to the M MRCA signatures for p2/NC and p6pol/PR, STAIM/MQKGN and TSFSF/PQITC, respectively. Positive selection of amino acids at protease cleavage sites. | The concatenated 360-bp cleavage site fragments were next compared internally to assess the mutational behavior of P1-P5 and P1'-P5' sites in the absence of drug therapy. As described in Materials and Methods, the analyses were performed using codon-based maximum-likelihood methods that test for the variation in selection pressure (dn/ds) at individual amino acids along the length of the sequence. Application of the likelihood ratio test indicated that the best-fit model for subtype C and group M sequences was the positive-selection (discrete) model (chi2 = 54.47; P < 0.0001 and 62.34 and P < 0.0001, respectively), while for B viruses the neutral model performed as well as the positive model (chi2 = 1.2; P > 0.05). Overall, when analyzed as a single concatenated fragment, all three datasets were found to be under negative (purifying) selection, with dn/ds (omega) values ranging from 0.26 to 0.31 for all sites. Only 20 (16.6%) of the 120 amino acids within the 12 cleavage sites of subtype C were found to be under positive selection, with a omega3 value of 1.7. Group M and subtype B cleavage sites had fewer positively selected amino acids: 11.1 and 6.7%, respectively . DISCUSSION : The presence of polymorphisms in the protease of subtype C would be expected to result in adaptive (compensatory) changes in the natural cleavage sites that are recognized and cleaved by the C enzyme. To test this hypothesis, we examined the prevalences and patterns of cleavage site mutations in the Gag, Gag-Pol, and Nef proteins of subtype C compared to those of non-C viruses. Using phylogenetic and ancestral reconstruction methods, we identified two groups of natural cleavage site sequences. The first group consisted of eight sequences, seven of which were highly conserved in all HIV-1 subtypes, and an eighth sequence which was moderately variable. Mutations at these sites were uncommon and, when present, were observed at relatively equivalent frequencies among different subtypes. These cleavage sites, which defined the main structural (MA, CA, and NC) and enzymatic (RT, RNase H, and integrase) proteins of HIV-1, were under strong negative (purifying) selection pressure, had a common ancestor, and showed little genetic evolution over time. The second group consisted of five cleavage sites that were under positive (diversifying) selection pressure, exhibited extensive inter- and intrasubtype variability, and showed little (or no) resemblance to the common ancestor of group M or to the subtype-specific MRCA. Our data showing that the majority (58.3%) of cleavage sites are highly conserved in all subtypes were not unexpected, given the narrow specificity of the HIV-1 protease relative to cellular proteases, such as pepsin . The strong purifying selection pressure exerted on these sites is presumably a reflection of the need to maintain the spatial configuration of the enzyme-substrate complex, conserve the hydrophobic nature of the scissile bond, and retain the biological activity of functionally important sites, such as the P1' proline of p17/p24 and the P1 and P1'-P5' residues of NC/p1. Cleavage of the P17/p24 site is known to play an important role in virion maturation, while processing of NC/p1 is required for ribosomal frame shifting and Gag-Pol expression (, -, ). In B viruses, cleavage of p17/24, p24/p2, and NC/p1 has been shown to be suboptimal, with the NC/p1 site being rate limiting . It has been suggested that the slow, regulated cleavage of these structural proteins may represent a common strategy to ensure that the assembled virions have the full complement of proteins needed to bud from the cell surface, bind to a new cell, and initiate a new round of viral replication . The carboxyl terminus of NC is particularly interesting. Unlike other cleavage sites, which carry an aromatic amino acid at P1 and either a leucine or proline residue at P1', the C termini of NC/p1 and NC/TFP carry an asparagine (N) residue at P1 opposite a phenylalanine (F) residue at P1' (, -). In this study of 84 untreated patients, no mutations were detected at the P2 or P1'-P3' positions of NC/p1 or NC/TFP and only a single N->C mutation was detected at P1. Taken together, these findings underscore the unique nature and limited mutability of the NC/p1 and NC/TFP cleavage junctions. Although these sites were strongly conserved in natural infection, recent studies have shown that an A->V substitution at the P2 positions of NC/p1 and NC/TFP is a common adaptive change, occurring in 29% of PI-resistant patients taking indinavir, saquinavir, and/or ritonavir for the treatment of subtype B . This valine substitution is frequently associated with an M46 I or L mutation (and possibly a V82 mutation) in the protease and leads to altered polyprocessing and improved growth of protease-mutated viruses. Whether similar "second-locus" mutations will be observed during the treatment of non-B subtypes remains to be established. The identification of common patterns may facilitate the development of broad-based inhibitors with increased specificity and improved binding to the mutated protease. These secondary inhibitors might preempt (or delay) the emergence of resistance. Our analyses also revealed important differences among HIV-1 subtypes. Particularly intriguing was the identification of five cleavage sites that exhibited extensive variability across all subtypes, with C viruses being significantly more variable than subtype B. Variation was restricted to a few specific amino acids, most of which were positively selected in C but not in B viruses. In contrast to conserved sites, variable cleavage sites tended to be those with regulatory rather than structural or enzymatic functions. At least four of the variable sites (p2/NC, p1/p6gag, TFP/p6pol, and p6pol/PR) are known to play major roles in the regulation of polyprotein processing and, in the case of TFP/p6pol, in the activation of the protease enzyme (-, , ). Studies of subtype B have shown that p2/NC is the initial and most rapidly processed cleavage site, controlling both the rate and the order of Gag and Gag-Pol polyprocessing . Our results indicate that p2/NC is by far the most variable cleavage site, with intrasubtype diversity ranging from 18.7% in subtype B to levels of 42.4% in subtype C. The p1/p6gag cleavage product, p6gag, is a major phosphoprotein that is critical to the release of mature, infectious virions . Although not well studied, phosphorylation of Gag and Gag-Pol sequences has been shown to alter susceptibility to cleavage, attenuating or even preventing the proteolytic process . The TFP/p6pol cleavage site, defining the N terminus of p6pol, was the only site to have a significantly higher level of diversity among B than among C viruses. TFP/p6pol is a novel cleavage site located 8 amino acids downstream from NC in the TFP domain of Gag-Pol . Although TFP/p6pol lies outside (and upstream) of the protease, the EDL tripeptide of this cleavage site (ENL in the case of C viruses) has been postulated to have a major influence on protease activation and on the timing and specificity of Gag-Pol cleavage, delaying the release of the protease until after the viral particle has budded from the cell membrane. Such a mechanism may protect the cell from the cytotoxic effects of proteolysis . The observed subtype variation in the cleavage sites controlling the initiation and rate of Gag and Gag-Pol processing (p2/NC) and the activation of protease (TFP/p6gag) suggests that there may be important differences in the way that B and C viruses regulate polyprocessing and virion assembly. Subtle cleavage site differences could, over time, have a major differential impact on the pathogenesis of HIV-1 subtypes and on response to therapy. Early treatment studies suggest that C viruses give an excellent initial response to highly active antiretroviral therapy but that the duration of the response may be less than that reported for B viruses. In summary, our results point to important inter- and intrasubtype differences in protease cleavage sites, especially in the p2/NC, TFP/p6pol, and p6pol/PR sites. The main limitations of our study relate to the cross-sectional nature of the data sets and the limited availability of well-matched pretreatment controls for use in the B data set. Despite these limitations, the potential impact of our findings on HIV-1 disease progression and response to therapy warrants further investigation, both at the patient level and in vitro using site-directed mutagenesis. The separate monophyletic clustering of B and C cleavage sites suggests that cleavage sites have evolved in a subtype-specific manner. The divergence between ancestral and contemporary sequences in the C data set and the location of an ancestral node distal to the group M MRCA suggest that variation in C cleavage sites began early, prior to the diversification of HIV-1 subtypes. A more detailed investigation of C cleavage sites, both over time and in response to therapy, is in progress. The present study forms the baseline for these ongoing studies. FIG. 1. : Schematic of the Gag and Gag-Pol processing sites showing the 12 individual protease cleavage sites: 5 cleavage sites in Gag (p17/p24, p24/p2, p2/NC, p7/p1, and p1/p6gag), 6 cleavage sites in Gag-Pol (NC/TFP, TFP/p6pol, p6pol/PR, PR/RT, RT/p66, and p66/IN), and a single site in Nef. Schematic of the Gag and Gag-Pol processing sites showing the 12 individual protease cleavage sites: 5 cleavage sites in Gag (p17/p24, p24/p2, p2/NC, p7/p1, and p1/p6gag), 6 cleavage sites in Gag-Pol (NC/TFP, TFP/p6pol, p6pol/PR, PR/RT, RT/p66, and p66/IN), and a single site in Nef. The frequency of amino acid substitution at each of these cleavage sites is shown in Fig. and Tables and . FIG.2. : Amino acid polymorphisms at Gag, Gag-Pol, and Nef cleavage sites. Amino acid polymorphisms at Gag, Gag-Pol, and Nef cleavage sites. The letters refer to the amino acid substitutions; the numbers in parentheses refer to the number of times the substitution was observed. Each cleavage site sequence consists of the 5 amino acids upstream and the 5 amino acids downstream of the scissile bond, indicated by a shill. The labeling of amino acids is according to the convention of P1 to P5 going from the scissile bond toward the amino terminus and P1' to P5' going toward the carboxy terminus. Positively selected amino acids are marked with asterisks. Dots represent amino acids that are identical to those in the M MRCA. FIG. 3. : Phylogenetic relationships of the South African Tygerberg virology (TV) cleavage site sequences relative to other subtypes in the group M data set. Phylogenetic relationships of the South African Tygerberg virology (TV) cleavage site sequences relative to other subtypes in the group M data set. This representative maximum-likelihood tree is based on concatenation and analysis of the 12 protease site nucleotide sequences as a single segment of 360 bp. An indication of the degree of sequence dissimilarity is given by the distance from the central node. The percentage of bootstrap trees out of 1,000 replications supporting a particular phylogenetic group is shown alongside the node considered. TABLE 1 : Sequences used for the analysis of protease cleavage sitesa TABLE 2 : Inter- and intrasubtype diversity at 12 cleavage sites, expressed as amino acid distances between sequences TABLE 3 : Relationship between cleavage site signature patterns and common ancestors TABLE 4 : Relationship between cleavage site signature patterns and common ancestors Backmatter: PMID- 12915546 TI - Rotavirus Infection Stimulates the Cl- Reabsorption Process across the Intestinal Brush-Border Membrane of Young Rabbits AB - Rotavirus is a major cause of infantile gastroenteritis worldwide. However, the mechanisms underlying fluid and electrolyte secretion associated with diarrhea remain largely unknown. We investigated the hypothesis that loss of Cl- into the luminal contents during rotavirus infection may be caused by a dysfunction in the chloride absorptive capacity across the intestinal brush-border membrane (BBM). The luminal Cl- concentrations in the entire small intestine of young rabbits infected with lapine rotavirus decreased at 1 and 2 days postinfection (dpi), indicating net Cl- absorption. At 7 dpi, luminal Cl- concentrations were slightly increased, indicating a moderate net Cl- secretion. By using a rapid filtration technique, 36Cl uptake across BBM was quantified by modulating the alkali-metal ion, electrical, chloride, and/or proton gradients. Rotavirus infection caused an identical, 127% +- 24% increase in all Cl- uptake activities (Cl-/H+ symport, Cl- conductance, and Cl-/anion exchange) observed across the intestinal BBM. The rotavirus activating effects on the symporter started at 1 dpi and persisted up to 7 dpi. Kinetic analyses revealed that rotavirus selectively affected the capacity parameter characterizing the symporter. We report the novel observation that rotavirus infection stimulated the Cl- reabsorption process across the intestinal BBM. We propose that the massive Cl- reabsorption in villi could partly overwhelm chloride secretion in crypt cells, which possibly increases during rotavirus diarrhea, the resulting imbalance leading to a moderate net chloride secretion. Keywords: Introduction : Rotavirus is the leading cause of infantile gastroenteritis and is responsible for high morbidity in developed countries and high mortality in developing areas of the world . The virus infects the mature enterocytes in the upper two-thirds of villus epithelia , and the idea is now gaining ground that diarrhea is not necessarily a consequence of any physical lesion but can precede it, as if cell dysfunction were the cause rather than the consequence of the histological damage . Recently, it was shown that rotavirus impairs both Na+-d-glucose (SGLT1) and Na+-l-leucine symport activities across intestinal brush-border membrane (BBM) isolated from young rabbits . Because SGLT1 is known to be involved in intestinal water reabsorption under physiological conditions , we propose that the mechanism of rotavirus diarrhea might involve generalized inhibition of Na+-solute symport systems and hence of water reabsorption . However, rotavirus disease is known to be multifactorial , and additional mechanisms might be needed. Rotavirus has been reported to lower intestinal disaccharidase activities both in vivo in young mice and in vitro in human Caco-2 cells . It has been proposed that the capacity of rotavirus to increase fluid and electrolyte secretion might be attributed in part to the action of the enterotoxin NSP4 or its cleavage product, NSP4-(112-175), after it is released from virus-infected cells . This peptide, which induces diarrhea in neonatal mice as did full-length NSP4 and NSP4-(114-135) peptide , would bind to a yet-unidentified apical membrane receptor in villus enterocytes or crypt cells or both and would directly or indirectly trigger signal transduction pathways to enhance net chloride secretion . However, the question arises as to the physical accessibility and binding capacity of luminal enterotoxin to the cells of the crypt region . Thus, the cellular and molecular bases by which rotavirus infection and NSP4 induce alterations in net chloride secretion remain largely unknown . It is widely accepted that increased luminal chloride concentrations can be due to decreased absorption in the villus cells and/or increased secretion in the crypt cells . To date, however, there has been little information concerning a possible dysfunction in the chloride absorptive process in rotavirus disease. It was earlier proposed that massive loss of chloride in the stools associated with the genetic disease chloridorrhea could be due to severe malfunctioning of the intestinal BBM Cl-/H+ symporter . The nonobligatory Cl-/H+ symporter has previously been well characterized for both BBM and basolateral membrane (BLM) vesicles isolated from guinea pig intestine and can account for Cl-/Cl- exchange and Cl- conductance activities in addition to symport of both H+ and Cl- . In the present study we investigated the hypothesis that loss of Cl- into the luminal contents during rotavirus infection may be caused by a dysfunction in chloride transport across the intestinal BBM vesicles purified from young rabbits, which have proved to be a good animal model for studying the pathogenesis of rotavirus diarrhea . We report the novel observation that rotavirus infection caused an identical increase in all Cl- uptake activities (Cl-/H+ symport, voltage-activated Cl- conductance, and Cl-/anion exchange) observed across the intestinal BBM. We propose that the massive Cl- reabsorption in villi could partly overwhelm chloride secretion in crypt cells, which possibly increases during rotavirus diarrhea, the resulting imbalance leading to a moderate net chloride secretion. A preliminary account of this work has already been published . MATERIALS AND METHODS : Rabbit inoculations and sample collection. | Specific-pathogen-free (SPF) 4-week-old New Zealand albino hybrid rabbits were obtained from Charles River (France) and were maintained at the animal house of the University of Paris-Sud, ChAtenay-Malabry, as previously described . Rabbits were orally inoculated with 2 ml of the lapine rotavirus stock suspension, strain La/RR510 (105.7 infectious particles/rabbit) . At appropriate times after infection, three to seven animals were killed by cervical dislocation after stunning. For each animal, the entire contents of the small intestine were collected and then filtered to remove debris before performing measurements of ionic concentrations (ion-selective electrodes for Cl- [ISA Biologie] and Na+ and K+ [Kone Optima]). Rotavirus present in the intestinal fluid of infected rabbits was quantified by enzyme immunoassay (IDEIA rotavirus test; DAKO Diagnostics, Cambridgeshire, England). For the transport study, the entire small intestine was removed, rinsed with saline at room temperature, everted, and distributed into plastic bags for storage at -80C, as described previously . BBM vesicle preparation. | Rabbit intestinal BBM vesicles were prepared by using the magnesium precipitation method as described previously . They were suspended at about 20 mg of membrane protein per ml in the appropriate membrane buffer (20 mM HEPES, 40 mM citric acid, 100 mM concentration of either Tris gluconate, K+ gluconate, or KCl-0.02% LiN3, supplemented to a total osmolarity of 560 mOsM with sorbitol and adjusted to pH 7.5 with Tris base) and stored in liquid nitrogen until the day of transport assay, as described previously . All the vesicles were normal according to a series of previously defined criteria . Membrane protein concentrations were measured with a Bio-Rad protein assay kit, with bovine serum albumin as the standard. Transport assay and expression of results. | Chloride transport was assayed by using 36Cl and a rapid filtration technique as described previously . Before use, H36Cl (0.4 mCi mmol-1; Amersham) was neutralized with Tris base . Briefly, 5 mul of a BBM vesicle preparation was used to carry out uptake measurements by mixing with 45 mul of transport buffer formed by the membrane buffer supplemented with either constant or variable concentrations of unlabeled substrate (5 to 100 mM), 5 mM 36Cl, and the amount of sorbitol necessary to secure a total osmolarity of 660 mOsM and was adjusted to either pH 7.5 or pH 5.0 with Tris base (final concentrations and pH in the incubation mixtures). When present, valinomycin (Sigma), a potassium specific ionophore, was dissolved in ethanol and was allowed to evaporate to dryness before being mixed with the membrane preparation . Initial uptake rate measurements were carried out for 4 s at 22C . Results are expressed as absolute velocities, v, in nanomoles seconds-1 milligram of protein-1 +- standard deviation (SD), of the pool of several experiments performed with 3 to 7 different BBM vesicle preparations. Uptake data were statistically compared by using a global F test . Kinetic analyses. | Uncorrected, initial absolute entry rates as a function of the substrate (cis Cl-) concentration were fitted by nonlinear least-squares regression analysis to an equation containing one saturable, Michaelian transport system plus a diffusional component: where Vmax and Kt are the capacity and affinity parameters of classic Michaelis-Menten kinetics, respectively, S is the substrate, and Kd is an apparent diffusion constant . To perform each fit, the procedure of Fletcher and Powell as modified by van Melle and Robinson was used. To test the fit of data to equation 1, we used the commercial program Stata (Integral Software, Paris, France). For statistical evaluation, fits were compared either within each given condition (F test) or between pairs of conditions (F' test), as described previously . To determine which kinetic parameters were affected, fits were performed under the restriction of one parameter common to the two sets of data (F" test) . All calculations were performed on Apple Macintosh microcomputers. RESULTS : Ionic concentrations in the intestinal contents of noninfected and infected rabbits. | Luminal K+ concentrations were not significantly modified by rotavirus infection (Fig. , lower curve). Similarly, rotavirus had no effect on luminal Na+ concentrations (data not shown). On the contrary, the Cl- concentrations in the intestinal lumen appeared to undergo a biphasic change during the time course of infection (Fig. , upper curve). At 1 and 2 days postinfection (dpi), luminal Cl- concentrations decreased significantly below control values, indicating a net Cl- absorption. At 3 dpi, at the time of maximal virus shedding and net increase in intestinal fluid volume of infected rabbits , a slight increase in Cl- concentrations was apparent. At 7 dpi, coinciding with the time of mild diarrhea , luminal Cl- concentrations were significantly increased (by about 24%) compared to control values, indicating a moderate net Cl- secretion. Thereafter, luminal Cl- contents returned to basal level. Evidence for pH gradient-dependent uphill transport of Cl- across intestinal BBM from young rabbits. | Considering our previous reports that an alkaline-inside pH gradient can furnish the energy necessary for the uphill transport of Cl- across intestinal BBM and BLM purified from guinea pigs , we specifically looked for the presence of overshoot phenomena in BBM from young rabbits. In the presence of a pHo/pHi (o [out] and i [in] indicate the extra- and intravesicular spaces, respectively) gradient of 5.0/7.5, Cl- uptake took place uphill (Fig. , upper curve), yielding a fast overshoot with the characteristics of the intestinal BBM Cl-/H+ symporter . When imposing an outward-directed Cl- gradient ([Cl-]o/[Cl-]i = 15/100 mM), cold trans Cl- induced overshoots both in the absence of a pH gradient (Fig. , curve Delta) and in its presence (pHo/pHi = 5.0/7.5; data not shown), a result indicating the existence of Cl-/Cl- exchange activity in rabbit BBM . In contrast, no Cl- overshoot was ever observed under equilibrated pH conditions (pHo/pHi = 7.5/7.5) in the absence or presence of alkali-metal ions (either K+, Na+, or K+ plus Na+) and electrical gradients (Fig. , the two lower curves), indicating the absence of any Cl-/Na+, Cl-/K+, and/or 2 Cl-/Na+/K+ symporter. Total initial Cl- uptake involves Cl-/H+ symport, rheogenic Cl- conductance, and Cl- anion exchange activities across noninfected rabbit intestinal BBM. | To further investigate the mechanism(s) underlying the pH gradient-dependent uphill transport of Cl- evidenced in rabbit intestinal BBM, the initial Cl- entry rates were quantified in a diversity of experimental conditions aimed at modifying the alkali-metal ion, electrical, and/or Cl- gradients, both in the absence and presence of an alkaline-inside pH gradient . Evidence for a Cl-/H+ symporter. | In the complete absence of alkali-metal ions, a 2.5-unit pH gradient caused a significant increase in the initial Cl- entry rate, confirming the presence of a Cl-/H+ symporter (Table , condition F versus A). To test whether the activation of Cl- uptake by an alkaline-inside pH gradient was either electrical (rheogenic) or electroneutral, we tested the effect of short circuiting the membrane potential with 100 mM equilibrated K+ concentrations and valinomycin. In these conditions, voltage clamping had a small but insignificant effect on the initial influx rate (Table , condition I), indicating that the uphill Cl- uptake driven by a pH gradient involved mainly an electroneutral Cl-/H+ symporter. In the absence of a pH gradient, the initial Cl- entry rate (Table , condition D), which was quantitatively identical to that obtained in the complete absence of alkali-metal ions (Table , condition A), could be taken as the reference of basal, initial Cl- entry rate in rabbit BBM. Evidence for rheogenic Cl- conductance. | To further assess the contribution to total Cl- uptake of a possible potential-sensitive conductance pathway, we studied the effect of valinomycin in the presence of an inward-directed K+ gradient ([K+]o/[K+]i = 100/0 mM), both with and without an alkaline-inside, 2.5-unit pH gradient. Both in the absence and in the presence of such a pH gradient, cis K+ caused small increases in the initial Cl- entry rate, further increased when valinomycin was added (Table , conditions B, C versus D and G, and H versus I). All of these increases were abolished when the membrane potential was short circuited to zero (Table , conditions D and I). These results can be interpreted as being due to the stimulation of a Cl- conductance pathway by the positive-inside membrane potential created by the inward-directed K+ gradient plus valinomycin . When alone, K+ had by itself just a slight electrical effect, indicating that rabbit BBM had a low intrinsic permeability for this ion. The results in Table further confirm the absence of any electroneutral Cl-/Na+, Cl-/K+, or 2 Cl-/Na+/K+ symporter in our rabbit BBM preparations. The observation that the activations caused by 100 mM cis of either K+, Na+, or an equimolar mixture of Na+ plus K+ on Cl- uptake were identical and fully abolished under voltage clamping conditions (Table , condition B versus D) indicated that the ion effects were purely electrical. Evidence for electroneutral Cl-/Cl- exchange activity. | Cl-/Cl- and Cl-/HCO3- exchange activities have both been reported to be present in enterocyte membranes of several animal species, including rat, guinea pig, and rabbit . Accordingly, we investigated the effects on 36Cl- uptake of imposing gradients of either chloride, pH, or both together in the presence of short-circuiting conditions to assess an electroneutral mechanism. Table gives the results obtained by using a combination of four different, meaningful situations (conditions D, E and I, J). Both in the absence of a pH gradient and in its presence, cold trans Cl- significantly increased the initial Cl- influx. These results provide evidence for the coexistence in the same vesicles of Cl-/Cl- exchange activity which is activated by cis H+ and a Cl-/H+ symporter which is stimulated by trans chloride. Similar activation of proton-gradient-driven Cl- uptake by trans bicarbonate has also been reported in both villus and crypt cell BBM from rabbit intestine . From the whole set of results, we conclude that total initial Cl- uptake in rabbit intestinal BBM vesicles occurred with three distinct activities: proton-coupled Cl- transport, voltage-activated Cl- conductance, and Cl-/anion exchange. Activating effect of rotavirus infection on Cl- transport across intestinal BBM. | To establish whether or not Cl- uptake activities were affected in the course of rotavirus infection, Cl- transport was studied by using BBM from infected rabbits. In support of earlier observations , virus shedding into the intestinal fluid of infected rabbits peaked at 3 dpi (data not shown). We previously reported that, at this time, significant inhibition of both SGLT1 and Na+-l-leucine symport activities was observed in the absence of any apparent histological damage (see Table in reference ). We therefore undertook to determine intestinal Cl- transport at 3 dpi. Chloride uptake experiments were repeated in the absence or presence of K+, electrical, and/or Cl- gradients, with or without a pH gradient. In all the study conditions, rotavirus infection had a general activating effect on Cl- uptake activities observed across the rabbit intestinal BBM . The increases in the initial Cl- entry rates were about the same (overall mean of 127% +- 24%) irrespective of the imposed gradient, suggesting that the quantitative participation of any of the activities to total initial Cl- uptake could be expected to be unaffected by rotavirus infection. All Cl- uptake values at 1 min were slightly activated, but no Cl- overshoot was ever observed under equilibrated pH conditions in the absence or presence of a K+ gradient with or without valinomycin (data not shown). The Cl- uptake values at equilibrium (2 h) remained constant (7.52 +- 3.07 nmol mg of protein-1; n = 18) and were identical to those found for noninfected rabbits (7.84 +- 1.55 nmol mg of protein-1; n = 18; see Fig. ), a result indicating that the apparent vesicular volume, which can be interpreted as a measure of the functional vesicle yield , did not change during rotavirus infection. Similar conclusions were reached when studying the effect of rotavirus infection on intestinal BBM Na+-solute symport activities in young rabbits . Quantification of Cl-/H+ symport, rheogenic Cl- conductance, and Cl-/Cl- exchange activities by intestinal BBM from noninfected and infected rabbits. Having established that initial Cl- influx into BBM occurred with three different properties, Cl-/H+ symport, rheogenic Cl- conductance, and Cl-/anion exchange, we next evaluated the quantitative participation of each activity. Among the many experimental conditions used, we selected those that would enable these calculations to be made. The relevant results are listed in Table , where the effects of electrical, trans chloride, and proton gradients were calculated in terms of the uptake increments. Taken as a whole, our results indicate that rotavirus infection did not markedly modify the proportion of either activity. About 10 and 90% of total initial Cl- uptake involved electrical and electroneutral mechanism(s), respectively. Although infection appeared to abolish the voltage-activated Cl- conductance activity (Table , condition C), the presence of this weakly operative component does not affect the general conclusion that Cl-/anion exchange and Cl-/H+ symport activities participated to a great extent (on average, 49 and 40%, respectively) to total Cl- influx, both in infected and control BBM. Evolution with time of proton-coupled Cl- uptake by intestinal BBM from infected rabbits. | To further confirm that rotavirus infection activated Cl- transport, the proton-gradient-driven Cl- uptake was studied as a function of time after infection. As illustrated in Fig. , all proton-coupled Cl- uptakes across BBM from infected animals were statistically different from that of the control group. Strong activation of Cl-/H+ symport activity occurred at times as short as 1 dpi and persisted up to 7 dpi. The increases remained practically identical during the first 2 days after infection, about 132 and 109%, respectively. Thereafter, at 3 and 7 dpi there seemed to be a progressive fall in Cl-/H+ symport activity, although the increases were still considerable, 91 and 56%, respectively. From the whole set of observations, we conclude that rotavirus infection strongly enhanced chloride influx across rabbit BBM. Effect of rotavirus infection on the kinetics of pH gradient-dependent Cl- uptake by BBM. | To understand the mechanisms involved in activation of intestinal Cl- transport during rotavirus infection, kinetic analyses were performed as a function of cis chloride concentration in the presence of a pH gradient (pHo/pHi = 5.0/7.5). As previously established with intestinal BBM and BLM from guinea pigs , the kinetics of Cl- uptake in the presence of an alkaline-inside pH gradient can be described by an equation involving one Michaelian, concave hyperbola plus one linear, nonsaturable, diffusion-like component (equation 1). The values for different groups of infected animals were all statistically different from those of the control group . The effects were highly significant at 3 dpi according to the F' test at P < 0.01. At 7 dpi a significant difference was apparent when the limits of probability were increased to P < 0.05, agreeing with the data on increased Cl- uptake during rotavirus infection . Also in agreement with the data shown in Fig. , no significant difference according to an F' test was revealed between infected rabbits at 3 and 7 dpi. Therefore, the relevant data were pooled and are shown as an overall fit (3 plus 7) in Table . To determine which kinetic parameters were indeed affected, the two sets of data at zero and 3 plus 7 dpi were fitted again under the restriction of one parameter common to the two sets. The F" test revealed a significant loss of information with a common Vmax, meaning that the parameter in question differed from one set to another. We conclude, therefore, that Vmax was the only parameter systematically increased by rotavirus infection. In contrast, Kt remained practically unchanged, exhibiting a relatively high value, about 35 mM, compared to 10 mM for guinea pig BBM . DISCUSSION : We report the novel observation that infection of young rabbits by lapine rotavirus, La/RR510 strain, caused a strong, identical increase in all Cl- transport systems observed across intestinal BBM. Both in infected and control BBM, the rheogenic Cl- conductance activity was weakly operative, whereas the electroneutral systems predominated: Cl-/anion exchange and Cl-/H+ symport activities represented, on average, 49 and 40%, respectively, of the total initial Cl- influx. Incidentally, such parallel increases in all three Cl- uptake activities, caused by rotavirus infection, would appear to strengthen the concept of the same transport entity, namely a single, mobile carrier acting as a nonobligatory Cl-/H+ symporter . As previously established with intestinal BBM and BLM from guinea pigs and similar to the present BBM results, external protons activated Cl-/Cl- exchange (Table , conditions E and J), a result providing evidence for the functioning of a symporter . In contrast, proton inhibition would be expected for an antiporter (exchanger) , as was the case for the erythrocyte Cl-/HCO3- exchanger mediated by anion exchanger AE1 . Taken as a whole, whether three distinct transport mechanisms or a single one was involved does not affect the general conclusion that rotavirus infection enhanced Cl- transport across rabbit villus cell BBM. The molecular identities of the proteins that mediate the anion exchange functions remain unknown . Whereas AE1 appears to code for a Cl-/HCO3- exchanger, it has yet to be established whether or not the intestinal AE2 codes for a Cl-/OH- antiporter or a Cl-/H+ symporter . AE2 was first localized in rabbit ileum BBM and then in BBM of both jejunum and ileum of the guinea pig and rabbit . More recently, AE2 was detected in both BBM and BLM of villus enterocytes and crypt cells from mice when using antibody to AE2 amino acids 109 to 122 . However, when using antibody to amino acids 1224 to 1237, AE2 was present in ileal BLM and not in BBM of rabbits and mice , further confusing AE2 localization. Nevertheless, it has been speculated that AE2a is the main isoform present in the intestinal BBM . To date, practically nothing is known about the localization and physiological significance of the AE2 isoforms throughout the gut. To explain a selective rotavirus effect on the Vmax parameter characterizing the Cl-/H+ symporter, two mechanisms can be envisaged, namely an increase in transporter numbers or a rise in the transporter turnover rate . Such mechanisms could not be demonstrated in the present paper, mainly because of practical problems of availability and specificity of antibodies to AE2 . Nevertheless, the fact that the Cl- uptake values at equilibrium that provide a measure of the effective transport capacity of a given vesicle population were the same in control and infected rabbits strongly suggests that turnover rate could be the parameter affected by rotavirus infection. A similar mechanism has been proposed to explain glucose malabsorption resulting from rotavirus infection, although here infection reduced the Vmax parameter characterizing SGLT1 without affecting the density of phlorizin-binding sites and the SGLT1 protein antigen present in the BBM preparations . The present results indicate a biphasic change in the luminal Cl- concentrations : net Cl- absorption at the earliest times after rotavirus infection and net Cl- secretion at a late stage of infection, coinciding with the appearance of mild diarrhea . However, the increase in luminal Cl- concentrations was found to be moderate, which would seem to be in line with observations that the ionic concentrations in the stools of rotavirus-positive children are much lower than those occurring in the secretory diarrheas caused by secretagogues, such as the enterotoxins of Vibrio cholerae and Escherichia coli . Davidson et al. reported that net Na+ and Cl- fluxes in jejunal epithelium from piglets infected with human rotavirus did not differ from those in noninfected animals. Conversely, Starkey et al. , by using an in vitro perfusion technique and intestinal segments, showed that Cl- transport exhibited secretion at 3 dpi, a time at which Cl- concentrations in the luminal contents from neonatal mice were maximally elevated. They predicted that increased net Cl- secretion could not be due to loss of chloride absorption but rather to the presence of a secretory component. However, the authors did not provide any experimental evidence for this suggestion. We show for the first time that rotavirus infection caused no chloride malabsorption but rather activation of the absorptive capacity of Cl-. In principle, activation of Cl- uptake that persisted up to 7 dpi would result in net absorption, fully agreeing with our data in the first two days after infection . However, this result appears to be incompatible with the slight increase in net Cl- secretion at a late stage of infection. We propose that the massive Cl- reabsorption in villi could overwhelm chloride secretion in crypt cells, which possibly increases during rotavirus diarrhea, the resulting imbalance leading to a moderate net chloride secretion. It should be noted that our BBM preparations were derived mostly from enterocytes , which leaves open the question of the rotavirus effect on the Cl- transport systems present in crypt cell BBM . The extent to which rotavirus can modify Cl- transport in small-intestine crypt cells remains unknown and will be dealt with elsewhere. Taken as a whole, the substantial chloride reabsorption reported in the present paper and the massive loss of water due to inhibition of Na+-solute symport systems may be determining mechanisms (but not necessarily the only ones) in the watery diarrheas caused by rotavirus. Although oral rehydration solution appears to be safe and effective in all forms of acute diarrhea in childhood, progress in the understanding of intestinal pathophysiology may lead to the development of new drugs to treat the clinically important disease caused by rotavirus worldwide . FIG. 1. : Ionic concentrations in the intestinal fluid of noninfected and experimentally infected rabbits. Ionic concentrations in the intestinal fluid of noninfected and experimentally infected rabbits. Potassium and chloride concentrations in the luminal contents from the entire small intestine of each animal were determined as a function of time after infection. Results are expressed in millimolars +- SD, with n = 8 to 43 determinations per point obtained from 4 to 12 rabbits. Because the results for four to nine noninfected rabbits at 0, 8, and 57 dpi were statistically indistinguishable, data have been pooled and are shown as a horizontal line. Identical letters indicate results found to be statistically indistinguishable according to a global F test (P < 0.01). FIG. 2. : Time course of chloride uptake by intestinal BBM purified from noninfected rabbits: effect of H+, electrical, and/or Cl- gradients. Time course of chloride uptake by intestinal BBM purified from noninfected rabbits: effect of H+, electrical, and/or Cl- gradients. Cl- uptake was determined with 15 mM cis36Cl as substrate. The extra- and intravesicular spaces contained a 20 mM HEPES-40 mM citric acid buffer adjusted with Tris base to give at zero time a pHo/pHi gradient of either 7.5/7.5 (, , Delta) or 5.0/7.5 and was supplemented with appropriate mixtures of either Tris gluconate, K+ gluconate, or KCl to obtain the following ionic concentration gradients: [K+]o/[K+]i = 0/0 mM and [Cl-]o/[Cl-]i = 15/0 mM ; [K+]o/[K+]i = 100/0 mM and [Cl-]o/[Cl-]i = 15/0 mM ; and [K+]o/[K+]i = 100/100 mM and [Cl-]o/[Cl-]i = 15/100 mM (Delta). When K+ was present, valinomycin was added at 10 mug mg protein-1. Results are expressed as absolute uptake values in nanomoles milligrams of protein-1 +- SD, with n = 16 to 26 determinations per point derived from four different membrane preparations. Because the uptake values at equilibrium (2 h) were identical, data have been pooled (n = 18). So as not to overload the figure, the uptake values at 4 s, 60 s, and equilibrium were used. FIG. 3. : Rotavirus activating effects on BBM Cl-/H+ symporter as a function of time after infection. Rotavirus activating effects on BBM Cl-/H+ symporter as a function of time after infection. The initial Cl- entry rates were determined with 15 mM cis36Cl. The extra- and intravesicular spaces contained a 20 mM HEPES-40 mM citric acid-100 mM Tris gluconate buffer adjusted with Tris base to give an initial pHo/pHi gradient of 5.0/7.5. Results are expressed in nanomoles milligrams of protein-1 second-1 +- SD, with n = 18 to 72 determinations per point obtained from three to seven different membrane preparations. Identical letters indicate results found to be statistically indistinguishable according to a global F test (P < 0.01). TABLE 1 : Chloride uptake by intestinal BBM purified from noninfected controls and infected rabbits at 3 dpi: effect of potassium, electrical, and/or chloride gradients on the initial CI- entry rates in the absence and presence of a pH gradient TABLE 2 : Quantification of rheogenic CI- conductance, CI-/CI- exchange, and CI-/H+ symport activities by intestinal BBM vesicles purified from noninfected controls and experimentally infected rabbits TABLE 3 : Kinetics of pH gradient-dependent CI- uptake by intestinal BBM purified from noninfected controls and infected animals at 3 and 7 dpi Backmatter: PMID- 12915567 TI - Latent Herpes Simplex Virus Infection of Sensory Neurons Alters Neuronal Gene Expression AB - The persistence of herpes simplex virus (HSV) and the diseases that it causes in the human population can be attributed to the maintenance of a latent infection within neurons in sensory ganglia. Little is known about the effects of latent infection on the host neuron. We have addressed the question of whether latent HSV infection affects neuronal gene expression by using microarray transcript profiling of host gene expression in ganglia from latently infected versus mock-infected mouse trigeminal ganglia. 33P-labeled cDNA probes from pooled ganglia harvested at 30 days postinfection or post-mock infection were hybridized to nylon arrays printed with 2,556 mouse genes. Signal intensities were acquired by phosphorimager. Mean intensities (n = 4 replicates in each of three independent experiments) of signals from mock-infected versus latently infected ganglia were compared by using a variant of Student's t test. We identified significant changes in the expression of mouse neuronal genes, including several with roles in gene expression, such as the Clk2 gene, and neurotransmission, such as genes encoding potassium voltage-gated channels and a muscarinic acetylcholine receptor. We confirmed the neuronal localization of some of these transcripts by using in situ hybridization. To validate the microarray results, we performed real-time reverse transcriptase PCR analyses for a selection of the genes. These studies demonstrate that latent HSV infection can alter neuronal gene expression and might provide a new mechanism for how persistent viral infection can cause chronic disease. Keywords: Introduction : A fascinating attribute of herpes simplex virus (HSV) is its ability to enter a quiescent state and establish a lifelong latent infection in sensory neurons that innervate the site of primary, productive infection. Following productive infection by HSV at the site of inoculation, the virus spreads to and enters sensory neurons, where it establishes a latent infection. Latent infection forms a reservoir of virus for recurrent infection, disease, and transmission to other individuals . HSV type 1 (HSV-1) is usually associated with primary infections of the orofacial area and latent infection of the trigeminal ganglion, while HSV-2 is usually associated with genital infections and latent infection in sacral ganglia. Although both primary and recurrent infections are usually self-limited, HSV can cause serious diseases such as neonatal disseminated herpes, viral encephalitis, and blinding keratitis . Also, genital herpes infection has been associated with an increased risk for human immunodeficiency virus infection . Additionally, long-term neurological symptoms have occasionally been associated with HSV infection . Latent HSV infection entails repression of the productive cycle of gene expression . Although the viral locus encoding the latency-associated transcripts (LATs) contributes to this repression , host functions are also likely to play a role in the repression of viral gene expression. In mouse models of latent HSV infection, the presence of infiltrating immune cells and cytokines in latently infected ganglia suggests that a local immune response may contribute to the maintenance of latent infection. Low-level expression of productive-cycle genes during latency may provide an antigenic stimulus for immune effectors that could repress HSV gene expression . Some evidence suggests that neuronal functions also help maintain latency . Thus, latent HSV infection appears to involve functions encoded by the pathogen and host functions including nonspecific and specific immune responses and neuronal functions. Control of transcription is likely to be a central process in the molecular interactions between HSV and its mammalian host. Transcript profiling with DNA microarrays provides an opportunity to identify host genes whose expression is affected by the presence of a latent HSV infection. Subsequent analyses of these genes could potentially deepen our understanding of specific mechanisms of virus-host interactions. Microarrays have been used to study productive HSV infection in cell culture, both to monitor viral gene expression and to detect HSV-induced changes in cellular gene expression . Microarray approaches have also been used to identify changes in cellular gene expression under conditions known to reactivate HSV from latent infection in experimental animal models . However, changes in host gene expression during the maintenance of latency were not reported in these studies. Analysis of gene expression in infected tissues by using microarrays can be complicated greatly by cellular heterogeneity . Latent HSV genomes reside in the nuclei of neuron cell bodies in trigeminal ganglia. However, trigeminal ganglia comprise not only clusters of sensory neurons and their nerve fibers but also nerve fibers derived from cells located outside the trigeminal ganglion that pass through or terminate within the ganglion and small glial cells called satellite cells that completely envelop the neuronal cell bodies . Also contained in trigeminal ganglia are Schwann cells, endothelial (capsular) cells of the microvasculature, blood cells, and motor neurons. During the first several days following corneal inoculation of mice with HSV, the virus replicates in corneal epithelial cells and is then cleared by the host immune response. During this period, HSV enters nerve terminals and is transported to neuronal cell bodies, where it undergoes acute replication. Nonspecific and specific immune effector cells infiltrate along the trigeminal nerve and form foci around individual infected neurons . Although viral replication is not detectable and most infiltrating cells are cleared by 30 days after corneal inoculation, low levels of immune effector cells remain in ganglia harboring latent HSV . We set out to use microarrays to test the hypothesis that HSV perturbs host gene expression in latently infected ganglia. To address the challenge of identifying gene expression changes in neurons that harbor latent HSV within the context of a complex tissue, we have performed microarray analyses in replicate experiments and applied statistical tests to identify those host genes whose expression changes significantly with latent HSV infection. We found significant changes in the expression of the immune response and, of particular interest, in the expression of neuronal genes, including several with roles in the regulation of gene expression and neural transmission. MATERIALS AND METHODS : Cells and virus. | The KOS strain of HSV was propagated and assayed on Vero cell monolayers as previously described . Infection protocol. | Seven-week-old male HSD/ICR mice (Harlan Sprague-Dawley) were anesthetized and either infected with 2 x 106 PFU of HSV per eye or mock infected with a medium by inoculation onto scarified corneas as previously described . At the indicated times postinoculation, the animals were sacrificed and trigeminal ganglia were removed, rapidly frozen in dry ice, and stored at -80C. Mice were housed in accordance with institutional and National Institutes of Health (NIH) guidelines on the care and use of animals in research, and all procedures were approved by the Institutional Animal Use Committee of Harvard Medical School. Microarray hybridization. | RNA was isolated from trigeminal ganglia by using RNA STAT-60 (Tel-Test, Friendswood, Tex.) and was pooled for each treatment group. Synthesis of [33P]-labeled cDNA, hybridization conditions, and data collection were exactly as described previously , except for the use of mouse total RNA (15 mug) and mouse C0t 10 DNA (10 mug) in hybridizations to mouse cDNA arrays. Array hybridizations performed in duplicate were highly reproducible, as reflected by the fact that the coefficient of variation (calculated as the standard deviation divided by the mean) averaged less than 0.2 for genes whose intensities were above a detection threshold, as reported by Chiang et al. . The dynamic range of detection spanned 3 orders of magnitude. Spot intensities were normalized to the median array spot intensity for each individual filter (global normalization). Data preprocessing was performed, and scatterplots were generated with Expression Explorer (Millennium proprietary software). Statistical analysis. | To obtain estimates of measurement error, we assumed an error model with a fixed error component (e.g., background variation) and a proportional error component (e.g., variation in the amount of spotted DNA). Measurement error (E) was modeled separately for each group of arrays corresponding to an individual tissue sample. We modeled the fixed error (A) as being the median standard deviation of the expression level for genes in the 5th percentile of expression (an average over replicates) and the coefficient for proportional error (C) as the coefficient of variation (standard deviation/mean) for genes in the 95th percentile of mean expression. The measurement error estimate for each gene within each group of arrays from a given sample was then calculated as , where A and C are defined above and x is the replicate-averaged expression level of the gene. The observed standard deviation was then regularized with the estimated measurement error by using the method of Baldi and Long with their default value of 10 pseudocounts. For each gene in each of the three independent experiments, we obtained a P value from the two-sided Student t test comparing latently infected and mock-infected samples. The P values from independent experiments were then combined by using Fisher's method . A Bonferroni correction based on the number of genes tested was then applied so that the single-hypothesis-test P value corresponding to a significant difference (multiple-hypothesis-corrected alpha = 0.05) was <= 9.78 x 10-6 and that for a highly significant difference (multiple-hypothesis-corrected alpha = 0.001) was <= 1.96 x 10-7. Although Bonferroni's correction can be overly conservative for certain analyses , its use here is consistent with the primary goal of determining with high confidence whether or not latent HSV impacts gene expression in the host ganglion, as opposed to a goal of generating a list of hypothetically differentially expressed genes with high false-positive rates. In situ hybridization | Methods for ganglion tissue preparation and in situ hybridization have been described previously . Plasmids used as probes for cellular gene expression included cDNAs encoding Chrm1, Ptprs, and Kcnc1. These genes were selected on the basis of their known neuronal specificity and the availability of specific reagents at the time the tests were conducted. Plasmid pIPH, used as a probe for HSV LATs, was included with each analysis to confirm latent or mock infection in ganglia and to verify the neuronal localization of HSV. DNA probes were generated by radiolabeling plasmids with alpha-35S-dATP and alpha-35S-dCTP (Amersham, Arlington Heights, Ill.) via nick translation. Two to six mock-infected or latently HSV infected ganglia were serially cryosectioned (thickness, 6 mum). Sequential sections from at least four different locations in each ganglion were hybridized with the various probes described above. Real-time RT-PCR. | Real-time reverse transcriptase PCR (RT-PCR) was performed on the ABI Prism 7700 Sequence Detection system (Applied Biosystems, Foster City, Calif.). Primers were designed according to the manufacturer's guidelines. Aliquots of RNA from each sample of pooled ganglia were treated extensively with RQ1 DNase, purified by using a MasterPure RNA Purification kit (Epicentre Technologies, Madison, Wis.), and reverse transcribed with a mixture of gene-specific downstream PCR primers. PCR primers were based on clone sequences and homology data retrieved from the Mouse Genome Database . We used SYBR Green fluorescence detection for all assays (2x PCR Master Mix; Applied Biosystems). The specificity of each assay was verified by electrophoretic separation of exponential-phase PCR products on polyacrylamide gels stained with a SYBR Gold nucleic acid gel stain (Molecular Probes, Eugene, Oreg.). Each set of samples analyzed included cDNAs derived from mock- and HSV-infected samples from a given experiment, RT-negative controls, and a panel of primer pairs that always included those for mouse glyceraldehyde-3-phosphate dehydrogenase (Gapdh) and LAT. Gapdh was used as the normalization standard after it was determined that the amounts of mouse Gapdh RNA per total RNA measured spectrophotometrically were comparable in the latently infected and mock-infected samples. LAT was included to confirm the identities of latently versus mock-infected samples. Threshold cycle numbers (CT) for each specific and reference (Gapdh) cDNA within a given sample were measured at least twice in duplicate or triplicate from each of two cDNA preparations of a given RNA sample. Replicate CTs were within 1 CT of each other. CTs were averaged (AvCT). CT is inversely proportional to the amount of starting material and represents a power of 2; the amount is calculated as 1/(2f)-CT, where f is amplification efficiency. Based on the assumption that each sequence was amplified with comparable efficiency in all mock- and HSV-infected pairs of samples, the difference between the AvCT of the reference gene and that of the specific gene (DeltaCT) [AvCT(Gapdh) - AvCT(specific gene) = DeltaCT(specific gene)] inversely correlates with the relative amount of the specific gene in a given sample. To compare genes in mock- and HSV-infected samples within each of the three experimental sets, the difference between their DeltaCTs was determined [DeltaCT(gene in mock-infected sample) - DeltaCT(gene in HSV-infected sample) = DeltaCT(infection)]. The ratio of gene expression in HSV-infected samples to that in mock-infected samples is calculated as 2DeltaCT(infection) for each experimental set. RESULTS : Microarray analysis of host gene expression in latently infected ganglia. | We hypothesized that latent HSV infection alters neuronal gene expression in ways that might affect neuronal physiology and viral gene expression. To investigate whether latent HSV infection alters neuronal gene expression, we conducted transcript profiling using microarray technology. We used a well-characterized mouse model in which HSV-1 strain KOS is inoculated onto a scarified mouse cornea, undergoes productive infection in the corneal epithelium and trigeminal ganglion for several days, and then establishes a latent infection in trigeminal ganglia by day 30 postinoculation . RNA was prepared from pooled ganglia (10 to 16 for each sample) harvested as described previously at day 30 postinfection (p.i.). RNA samples from latently and mock-infected mice were used to generate radiolabeled cDNA probes. Probes were hybridized to duplicate nylon filters, with each printed in duplicate with 2,556 PCR-amplified, mostly full length mouse cDNA clones plus quality and normalization control clones. Scatterplots of duplicate hybridizations demonstrated the filter-to-filter consistency of the arrays , while scatterplots of latently infected versus mock-infected ganglia revealed numerous differences consistent with increases and decreases in host transcript levels . Thus, differences in gene expression between the experimental and control groups were detected by microarray analysis. Several factors could potentially limit the confidence with which one can interpret microarray results. First, there is experimental variability of microarray measurements, as has been observed by others . Second, only a small fraction of cells in trigeminal ganglia harbor latent HSV, so that even large differences in gene expression in individual infected cells might give rise to very small differences in expression on a whole-ganglion basis. Thus, we wished to evaluate relatively small differences in gene expression. For these reasons, we performed three independent experiments that entailed three separate infections of mice and preparation of cDNA probes to increase the statistical power of the analysis. Microarray sets used for each experiment came from separate printings. Each probe was hybridized on replicate arrays such that each clone was represented at least four times. Differences between latently HSV infected and mock-infected samples were evaluated statistically in each experiment by using a variant of the Student t test in which standard deviations were regularized by an independent estimate of measurement errors. The P values from each experiment were combined, and a correction was applied to account for the total number of genes tested. Only genes whose expression differed in the same direction in all three experiments were included. Differences between latently and mock-infected samples were statistically significant for 51 genes, and 8 of these differences were highly significant (Tables and ). (The entire annotated list of genes represented in the data set, along with averaged, normalized values and ratios of gene expression in HSV-infected samples to that in mock-infected samples, is available for viewing at .) Altered expression of immune response genes. | Infiltration of inflammatory cells in latently infected ganglionic tissue has been observed . Thus, it was not surprising that about half of the 51 genes in Tables and are involved in the immune response and that all of these immune response genes exhibited increased expression in latently infected versus mock-infected ganglia. This group included genes expressed specifically by immune cells, such as the CD3 subunit gene Cd3g , and/or in target cells involved in antigen presentation or in responding to immune stimuli, such as the major histocompatibility complex class I genes H2-D1 and beta2 microglobulin, which can be expressed in the nervous system . These results confirm and extend previous observations of a persistent immune response in latently infected ganglia , thus validating the array results. Alteration of neuronal gene expression. | Strikingly, a surprising number of the genes in Tables and (mostly in Table ) are known to be expressed in nervous system tissue and/or neurons. Twenty-one genes of this class exhibited increased expression in latently infected versus mock-infected ganglia, and eight exhibited decreased expression. Many of the genes whose expression was altered play definite roles in neuronal physiology. These include genes encoding neurotransmitter receptors, such as Gprc1g , Chrm1 , and Gabbr1 ; voltage-gated ion channels, such as Kcnc1 and Kcnab ; proteins involved in signaling, including Trc8 , Ptprs , Adrbk1 , Gpr56 , Ppp3Ca , Rgs4 , Rab27b , and Gh ; proteins involved in neurite extension, axonal elongation, and cell-matrix interactions ("axonal remodeling"), such as Prss12 , Ulk1 , Col15a1 , Adam23 , and n-chimaerin ; and proteins involved in catabolic functions, such as Man2b1 , Ctsd , Usp20 , Ube1x , and Acox2 . Several genes, such as Stat1 , Hipk2 , Apobec1 , and Clk2 , are involved in gene expression. One gene, Aebp1, encodes a protein described variously as being involved in either gene expression or axonal remodeling or both . In situ hybridization analysis was performed on a few genes that appeared significantly altered in all statistical analyses in an attempt to determine the cell types that expressed the cellular transcripts. Neuronal expression of Chrm1 was observed in sensory neurons in mock-infected (data not shown) and latently infected ganglia. Neuronal expression of Ptprs was also observed in mock-infected (data not shown) and latently infected ganglia. Finally, neuronal expression of Kcnc 1 was observed in mock-infected and latently infected ganglia (data not shown). We confirmed that these ganglia were latently infected by in situ hybridization with a LAT probe (data not shown). Specific neuronal expression of another gene, Stat1, in mouse trigeminal ganglia has recently been shown by immunohistochemistry . Although expression of Gh in neurons has not been documented to our knowledge, the expression of Gh receptors and target genes has been demonstrated . Confirmation by real-time RT-PCR. | To verify these changes in gene expression by an alternate method, we quantified the expression of 11 genes (Tables and ) representing various neuronal functions by RT-PCR using real-time technology. Mouse Gapdh was assayed as a normalization control. As shown in Fig. , Stat1 transcripts, which were more abundant in latently infected than in mock-infected ganglia in the three array analyses, were also reproducibly more abundant in latently infected than in mock-infected ganglia relative to the abundance of Gapdh transcripts in the RT-PCR assay. The mean ratios of expression in latently infected ganglia to expression in mock-infected ganglia were of similar magnitude by the two assays. Thus, we verified changes in expression of specific neuronal genes in latently infected ganglia by RT-PCR. Time course of host gene expression. | The altered levels of specific mouse RNAs in latently infected ganglia on day 30 p.i. could reflect changes in cellular gene expression specific to the latent state or changes induced initially during acute viral replication. To distinguish between these possibilities, in one of the three experiments we harvested trigeminal ganglia from HSV- and mock-infected mice on days 3 and 10 p.i., and we assayed the RNA by using the same arrays used for the 30-day samples. Limiting the analysis just to the 52 genes in Table , those whose expression was greater in latently infected ganglia formed two expression profile clusters: one, of special interest, comprising 15 genes showing maximal expression on day 30 p.i. (Fig. and Table ) and the other showing maximal expression on either day 3 or day 10 p.i. (Fig. and Table ). As expected, most of the host immune response genes exhibited peak expression on day 3 or 10 p.i., with peak levels of lymphocyte markers such as Cd3d, Cd3g, Cxcr6, Lyn, Axl, Il2rg, Lck, and Ptprc at day 10 p.i. (Many other genes known to be involved in classical host immune responses exhibited increased expression at days 3 and 10, as expected. Only those whose expression was significantly increased at day 30 are examined here.) The peak expression of Stat1 on day 3 p.i. was consistent with high expression of cytokines at this time . In contrast, most of the genes that are known to be expressed relatively abundantly in nervous tissue and/or neurons, including those shown by in situ hybridization to be specifically expressed in neurons of latently infected trigeminal ganglia, were expressed in greatest abundance on day 30 p.i., well after cessation of acute viral replication in ganglia. Additionally, the eight genes whose expression was lower in latently infected ganglia all exhibited the greatest decrease relative to levels in mock-infected ganglia at day 30 p.i. . Thus, taken together, these data show that latent infection rather than acute infection was associated with altered expression of these neuronal genes. DISCUSSION : An important problem in virology is to understand the interaction of latent viral infections with the host cell and to identify host responses to virus latency. In these studies, we observed up-regulation or down-regulation during latent infection of the levels of a number of host genes, including several thought to be expressed specifically in neurons. Our confidence that levels of expression of the genes listed in Tables and are truly altered is based on stringent statistical criteria applied to the microarray hybridization data coupled with validation by quantitative real-time RT-PCR. Evidence that expression of the genes listed in Table is altered only during latent infection and that these changes are not remnants of acute infection is provided by the results in Fig. . In contrast to our results, Hill and colleagues did not identify changes in gene expression between uninfected and latently infected ganglia. One reason for the differing results could be that they looked at a total of only 149 mouse genes in a "stress/toxicology" set, whereas our data set included 2,500 genes in a diverse set. Potential effects of latent HSV infection on host gene expression. | Our microarray data reflect changes in gene expression in the entire trigeminal ganglion, which is a heterogeneous tissue composed of many cell types. The alterations in expression of neuron-specific genes observed in these studies could represent a direct effect of HSV on cellular gene expression in latently infected cells. Alternatively, the virus may act indirectly on cellular gene expression, for example, by inducing neuronal injury or by inducing a persistent immune response to latent infection that includes expression of cytokines known to alter host gene expression. The increased expression of Stat1 is consistent with the latter mechanism. Although it is easier to explain the decreased expression of certain genes by such an indirect mechanism, to date, little or no precedent exists for immune modulation of neuronal gene expression. Direct and indirect effects of latent HSV infection on cellular gene expression are not necessarily mutually exclusive. These two mechanisms make specific predictions about the levels of expression of particular cellular genes in latently infected versus uninfected neurons. Preliminary studies using in situ hybridization to examine this issue have been inconclusive due to the qualitative nature of this technique and the apparent variability of expression from cell to cell. Thus, PCR-based methods to examine individual cells may be necessary to test these models. Potential effects of changes in host gene expression on viral gene expression. | Alternatively, changes in cellular gene expression may represent a host response that contributes to the repression of productive-cycle HSV gene expression characteristic of latent infection. Of particular interest are cellular genes encoding proteins that regulate gene expression. Aebp1 has been reported to be a transcriptional repressor, and Hipk2 has been reported to be a corepressor . A relative of Clk2, Clk1, has been reported to induce neuronal differentiation of PC12 cells in a manner akin to nerve growth factor , which helps to maintain HSV latency in both in vitro and in vivo models . The increased expression of Apobec1, an RNA-editing enzyme, is especially intriguing, given the abundant expression of LAT, an RNA of uncertain coding capacity, in latently infected neurons. Altered expression of genes may provide clues to changes in signaling pathways that regulate gene expression. The increased expression of Chrm1 is interesting in that it is ordinarily down-regulated by ciliary neurotrophic factor (CNTF), which has been reported to promote HSV reactivation in humans . Certain cellular genes may maintain the virus in a state of readiness for reactivation. Stat1 may also be in this category, as it is activated by CNTF. Other neuronal genes whose expression is altered could also influence viral gene expression less directly. For example, Chrm1 couples to G proteins that can activate mitogen-activated protein kinase cascades that lead to changes in gene expression . These various hypotheses make specific, testable predictions regarding the effects of knocking out or overexpressing these genes in latently infected ganglia. Cellular proteins that help to maintain latency or promote reactivation could potentially serve as targets for therapeutic intervention. Potential implications for HSV infections in humans. | Given that the expression of several genes with known roles in neurotransmission and signaling is altered in latently infected ganglia, our results, using a mouse model and a virus strain with which no spontaneous reactivation has ever been observed , raise the possibility that latent HSV can affect sensation. Moreover, several of the genes whose expression is altered, including Gprc1g, Gabbr1, Kcnab2, and Kcnc1, affect neuronal excitability in an interconnected manner , which suggests a mechanism by which changes in neuronal physiology are induced by latent virus and could alter sensation. This predicts that neurons in latently infected ganglia would exhibit alterations in ion channel function such as have been observed in neurons in vitro . Consistent with this idea, postherpetic allodynia and hyperalgesia have been observed in mice latently infected with HSV for at least 40 days p.i. . Interestingly, there are reports of patients with recurrent HSV with symptoms of abnormal sensation, including pain, both before and even long after clinically evident disease . Although these symptoms are often ascribed to reactivating HSV , or could be due to sensory nerve or central nervous system injury, our results provoke the speculation that these clinical phenomena and perhaps more subtle changes in sensation may be due to effects of latent HSV infection on sensory neurons. These studies demonstrate that latent herpes simplex virus infection can alter neuronal gene expression and might provide a new mechanism for how persistent viral infection can cause chronic disease. FIG. 1. : Scatterplots comparing 33P signal intensities in microarrays after background subtraction and normalization. Scatterplots comparing 33P signal intensities in microarrays after background subtraction and normalization. Each point represents the log10 phosphorimager intensity value normalized to the median intensity of the microarray for each filter. (A and B) Results from duplicate array hybridizations. The x and y axes each represent one of the two duplicate arrays. HSV, one representative probe generated from pooled ganglia 30 days after infection. Mock, one representative probe generated from pooled ganglia 30 days after mock infection, and from the same experiment as the HSV probe shown. (C) Results for HSV-infected versus mock-infected ganglia. y axis, one of the two duplicate HSV arrays; x axis, one of the duplicate Mock arrays. FIG. 2. : Neuronal expression of host genes confirmed by in situ hybridization. Neuronal expression of host genes confirmed by in situ hybridization. Mouse trigeminal ganglia latently infected with HSV at 30 days p.i. hybridized with radiolabeled DNA probes for cellular genes. Neurons were identified morphologically by their large size and staining pattern. (A) Muscarinic acetylcholine receptor gene (Chrm1). Grains indicate nuclear and cytoplasmic distribution of the mRNA in neurons (magnification, x200). (B) Protein tyrosine phosphatase sigma gene (Ptprs). Grains cluster over neurons (magnification, x400). FIG. 3. : Quantitative real-time RNA PCR. Quantitative real-time RNA PCR. This example illustrates that the level of Stat1 RNA normalized to Gapdh RNA is higher in latently infected ganglia (HSV) than in mock-infected ganglia (Mock). DeltaRn, fluorescence difference from background determined by using Sequence Detection System software (Applied Biosystems). Each assay was performed in triplicate. The relative level of Stat1 was determined by the difference in average cycle number at a constant threshold DeltaRn. Gapdh, mouse Gapdh mRNA; Stat1, mouse Stat1 mRNA. View is expanded around exponential phase. FIG. 4. : Kinetic expression profiles of selected genes whose up-regulation at day 30 p. Kinetic expression profiles of selected genes whose up-regulation at day 30 p.i. correlated significantly with HSV latency. HSV/Mock, ratios of mean normalized signal intensities in HSV-infected samples to those in mock-infected samples.. Each bar for day 3 or 10 p.i. represents an average ratio from one experiment, and each bar for day 30 p.i. represents average ratios from all three experiments. TABLE 1 : Changes in mRNA levels during latent HSV infection following peak expression during acute infection TABLE 2 : Changes in mRNA levels associated with latent HSV infection Backmatter: PMID- 12915558 TI - Involvement of the Matrix and Nucleocapsid Domains of the Bovine Leukemia Virus Gag Polyprotein Precursor in Viral RNA Packaging AB - The RNA packaging process for retroviruses involves a recognition event of the genome-length viral RNA by the viral Gag polyprotein precursor (PrGag), an important step in particle morphogenesis. The mechanism underlying this genome recognition event for most retroviruses is thought to involve an interaction between the nucleocapsid (NC) domain of PrGag and stable RNA secondary structures that form the RNA packaging signal. Presently, there is limited information regarding PrGag-RNA interactions involved in RNA packaging for the deltaretroviruses, which include bovine leukemia virus (BLV) and human T-cell leukemia virus types 1 and 2 (HTLV-1 and -2, respectively). To address this, alanine-scanning mutagenesis of BLV PrGag was done with a virus-like particle (VLP) system. As predicted, mutagenesis of conserved basic residues as well as residues of the zinc finger domains in the BLV NC domain of PrGag revealed residues that led to a reduction in viral RNA packaging. Interestingly, when conserved basic residues in the BLV MA domain of PrGag were mutated to alanine or glycine, but not when mutated to another basic residue, reductions in viral RNA packaging were also observed. The ability of PrGag to be targeted to the cell membrane was not affected by these mutations in MA, indicating that PrGag membrane targeting was not associated with the reduction in RNA packaging. These observations indicate that these basic residues in the MA domain of PrGag influence RNA packaging, without influencing Gag membrane localization. It was further observed that (i) a MA/NC double mutant had a more severe RNA packaging defect than either mutant alone, and (ii) RNA packaging was not found to be associated with transient localization of Gag in the nucleus. In summary, this report provides the first direct evidence for the involvement of both the BLV MA and NC domains of PrGag in viral RNA packaging. Keywords: Introduction : The RNA packaging process for retroviruses involves a recognition event of the genome-length viral RNA by the viral Gag polyprotein precursor (PrGag), which acts to initiate the morphogenesis of virus particles (see reference for review). The mechanism underlying this genome recognition event is poorly understood, but many biochemical and genetic analyses have revealed that this event involves the interaction between stable RNA secondary structures at the 5' end of the viral genome and, in many cases, amino acids in the nucleocapsid (NC) domain of PrGag . The genome recognition event is an important step in the morphogenesis of infectious retrovirus particles. This recognition event leads to the predominant packaging (encapsidation) of the genome-length viral RNA into assembling particles . This discrimination process, which is primarily a viral RNA-protein interaction, is known to strongly favor the full-length viral RNA over spliced viral RNAs and cellular mRNAs . In general, the RNA sequences necessary and sufficient for the RNA packaging process are located in a region that includes the 5'-noncoding region along with the 5' half of the gag gene . These sequences are referred to as the "packaging" or "encapsidation" signal. The packaging signal can enhance the packaging of RNAs containing the signal over those not containing the packaging signal. However, in the absence of RNAs containing a packaging signal, cellular RNAs can be packaged into particles, which supports the conclusion that RNA is a structural element in retrovirus particles . The Gag polyprotein precursor alone is sufficient for the formation of particles, and the presence of RNA aids in efficient particle production . Since activation of the viral protease does not typically occur until after virus particle release, the genome recognition event involves full-length viral RNA and unprocessed Gag polyprotein . Gag includes three main domains (matrix [MA], capsid [CA], and NC). Attempts to address the RNA packaging process have focused on identifying RNA-binding domains within unprocessed Gag (and the processed Gag products) and demonstrating the specificity of binding of viral RNA to Gag either in cell-free reactions or in the uptake of viral RNA into virus particles. The NC proteins of all retroviruses share the characteristics of a high percentage of basic residues and zinc binding domains involved in RNA packaging and in many instances alter the specificity of RNA binding . It is not clear whether NC actually confers all of the selective recognition of the viral genomic RNA . The p2 spacer peptide between CA and NC of human immunodeficiency virus type 1 (HIV-1) PrGag has been suggested to play a role in selective RNA packaging . However, previous studies with HIV-1 have indicated that MA does not play a role in RNA packaging . Bovine leukemia virus (BLV), as well as other deltaretroviruses, replicates to low titers in animals and is poorly infectious in cell culture. Cocultivation is typically used to infect permissive host cells. Because of these difficulties, information regarding the molecular details of the virus life cycles, including virus assembly, is limited. To more easily study the RNA packaging step in the virus assembly pathway of BLV, we have used a virus-like particle (VLP) model system to identify the protein determinants of BLV RNA packaging . As expected, many of the basic amino acid residues and all zinc finger residues mutated in the NC domain of PrGag led to significant reductions in RNA packaging. Interestingly, many of the basic residues in MA were also found to significantly influence RNA packaging. Together, these observations provide the first evidence that both the MA and NC domains of BLV PrGag are involved in viral RNA packaging. MATERIALS AND METHODS : VLP vector and mutagenesis. | The vector used to produce VLPs, pGag-HA, has a hemagglutinin (HA) epitope tag fused to the C terminus of the BLV PrGag. The construction of this vector and the ability of the vector to produce VLPs have been described previously . This vector transcribes a single RNA transcript and no spliced RNA. The RNA is translated to express PrGag in the absence of the viral protease. Therefore, the VLPs do not have mature cores. The primary BLV RNA packaging signal, consisting of two stable RNA stem-loop structures (i.e., SL1 and SL2), are located just downstream of the gag start codon (-). Due to the presence of the packaging signal, this RNA transcript can be packaged into VLPs through the interaction with PrGag. In this system, the VLP expression vector transcribes only one RNA species due to the absence of the BLV splice donor and splice acceptor sites. Site-directed mutations were introduced into pGag-HA by using the QuickChange XL kit (Stratagene, La Jolla, Calif.) according to the manufacturer's instructions. All derivatives made of pGag-HA were sequenced to verify the correct introduction of the desired mutation and the absence of undesired mutations. Cells and transfections. | To express BLV PrGag in mammalian cells, COS-1 cells were transfected with pGag-HA or a derivative and placed under G418 selection to obtain stable cells expressing PrGag. Cells were grown in 100-, 60-, or 35-mm-diameter dishes in Dulbecco's modified Eagle's medium (GIBCO BRL, Gaithersburg, Md.) supplemented with 10% Fetal Clone III serum (HyClone, Logan, Utah). Superfect (Qiagen, Valencia, Calif.) was used for transfection of cells. Two days posttransfection, cells were placed under G418 selection until resistant colonies formed (similar3 weeks). Approximately 100 G418-resistant colonies were pooled and used for VLP analysis. Quantitation of VLP production. | Supernatant from pooled stable cell clones producing VLPs was harvested and clarified by low-speed centrifugation (5 min at 700 x g) and then subjected to ultracentrifugation for 1 h at 40,000 x g at 4C. Pelleted VLPs were then resuspended in radioimmunoprecipitation assay (RIPA) buffer (1% IGEPAL CA-630, 50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.5% deoxycholate [DOC], 5 mM EDTA, 0.1% sodium dodecyl sulfate [SDS]). Lysates prepared from VLP-producing cells and VLP samples were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose, and BLV Gag was detected with a primary antibody directed against the HA epitope tag (Covance, Princeton, N.J.) and an antimouse immunoglobulin (Ig) horseradish peroxidase-linked whole antibody (from sheep) as a secondary antibody (Amersham, Arlington Heights, Ill.). Quantitation of band intensities was done with the Quantity One software package with the Chemi Doc 2000 Documentation System (Bio-Rad, Richmond, Calif.). Quantitation of viral RNA in cells and VLPs and determination of RNA packaging efficiencies. | Total RNA was extracted from transfected cells by using a cellular RNA extraction kit (Qiagen) according to the manufacturer's instructions. Viral RNA was extracted with a QIAamp viral RNA kit (Qiagen) according to the manufacturers' instructions. To control for DNA contamination, the RNAs isolated were treated with DNase I (Invitrogen) and used in reverse transcription-PCR (RT-PCR) without the addition of reverse transcriptase to confirm the absence of DNA. Detection of the RNA transcript that expresses PrGag in cells and its incorporation into VLPs due to the RNA packaging signal was done by quantitative real-time RT-PCR analysis. Quantitative real-time RT-PCR was carried out with the Qiagen RT-PCR Master SYBR Green kit in a 20-mul reaction volume. RNA templates were reverse transcribed at 50C for 25 min followed by a denaturing step at 95C for 15 min. PCR was then performed with the primers +BLV979 (5'-AACCGCCATCGTGCTTGGGCA-3') and -BLV1183 (5'-CGCTTCAGCGGCGGCTATTGC-3'). These primers amplify an approximately 200-bp-long fragment within the gag gene. The PCR protocol consisted of 50 cycles of denaturation (95C for 15 s), annealing (60C for 20 s), and extension (72C for10 s). For each step, the temperature transition rate was 20C/s. The PCR was monitored with a Light Cycler (Roche Diagnostics) after each elongation step by SYBR Green I dye binding to amplified products. Quantitation was carried out with an external standard curve. Standard RNA was synthesized by in vitro transcription of a linearized plasmid containing a BLV proviral DNA representing the 5' end of the gag gene. All in vitro transcripts were DNase I treated and tested for DNA contamination. Standard curves were constructed from 10-fold serial dilutions of synthetic RNA transcripts. After real-time RT-PCR was completed, logarithmic values of fluorescence for each dilution were plotted against cycle number. A baseline was set just above the fluorescence background, and a crossing point was determined with amplification curves obtained during the initial exponential phase of amplification. The specificity of the amplified product was then determined by melting curve analysis. Melting curve acquisitions were done immediately after PCR was completed, by heating at 95C for 0 s, cooling to 75C for 15 s, and heating slowly at 20C/s until 95C with continuous fluorescence recording. RNA packaging efficiencies were determined by the ratio of viral RNA detected from VLPs to viral RNA detected from total RNA recovered from VLP-producing cells divided by the ratio of Gag protein detected from VLPs to Gag protein detected from VLP-producing cells. RNA packaging efficiencies of mutants were then compared relative to that of the parental pGag-HA vector. Confocal microscopy. | Transfected cells were grown on coverslips, fixed with 4% paraformaldehyde, and permeabilized with Triton X-100, both diluted in phosphate-buffered saline. The cells were then incubated with anti-HA antibody followed by incubation with Alexa Fluor 488-conjugated goat anti-mouse IgG (Molecular Probes). Images were collected with a Bio-Rad MRC 600 confocal microscope. RESULTS : Effects of mutations in the NC domain of PrGag on RNA packaging. | In order to determine the influence of NC on RNA packaging, a series of NC mutants were made in pGag-HA and tested for their ability to influence RNA packaging. Alanine residues were introduced in place of basic amino acid residues and in place of the conserved residues in each of the two zinc finger binding domains . The panel of NC mutants was analyzed in parallel with the parental vector. Figure shows the RNA packaging efficiencies of the mutants. There were three groups of mutant phenotypes. The first group of mutants had no RNA packaging defects compared to the parental vector. These mutants included K17A and R18A. A second group of mutants had mild RNA packaging defects (40 to 70% reduction in VLP production) (i.e., R12A, K17A/R18A, K59A, K66A, and K68A). The third group of mutants had severe RNA packaging defects compared to the parental vector (85 to 95% reduction) (i.e., C24A, C27A, R35A, C49A, H57A, and R60A). For many of the mutants, VLP production was reduced two- to threefold that of the parental vector (i.e., R12A, K17A, R18A, K17A/R18A, R35A, and C49A). This reduction is within the linear detection range for the immunoblot assay used (described below). As predicted, many of the basic amino acid residues that were mutated led to significant reductions in RNA packaging efficiencies. The mutated basic residues that resulted in the greatest reductions were R35A and R60A. Interestingly, each of these residues lies between the conserved histidine and cysteine residues in the first (i.e., R35A) and second (i.e., R60A) zinc finger domains. Effects of mutations in the MA domain of PrGag on RNA packaging. | Previous studies with other retroviruses have provided little support for a role of the MA domain in retroviral RNA packaging . To test whether basic residues in the BLV MA domain influence RNA packaging, a panel of mutants that changed basic residues in MA to alanine was created . The RNA packaging efficiencies of the mutants were then analyzed in parallel with the parental vector. Figure shows a representative protein analysis of Gag from cells and VLPs. The K52A and K53A MA mutants led to the lowest levels of VLPs (8 and 17%, respectively). These reductions are within 13-fold of the level of the parental vector and are within the linear range of detection for the assay, which was at least 16-fold . Figure shows the resulting RNA packaging efficiencies for the panel of MA mutants tested. As with the NC mutants, there were three general groups of mutant phenotypes. Several basic residues that are not conserved among BLV, human T-cell leukemia virus type 1 (HTLV-1), and HTLV-2 led to no or small changes in RNA packaging efficiencies (i.e., R31A, H48A, K53A, R69A, and R72A). A second group led to modest reductions in RNA packaging efficiencies (i.e., R27A and K52A). Finally, a third group had significant reductions in RNA packaging (66 to 92%) compared to that of the parental vector (K41A and H45A). The H45A MA mutant had a reduction in RNA packaging efficiency (92%) that was as great as any of those in the individual NC mutants, indicating that BLV MA plays an important role in RNA packaging. The K49 residue was not analyzed in this study. In other retroviruses, such as HIV-1, basic amino acid residues in MA can influence PrGag membrane targeting and are proposed to form a positively charged patch that promotes membrane binding . To determine whether MA mutants that significantly reduced RNA packaging also influenced PrGag membrane localization, the cellular distribution of Gag was analyzed by confocal microscopy. Figure shows the effects of several basic amino acid mutants on Gag distribution in cells. These mutants (i.e., R27A, K41A, H45A, K52A, and K53A) had a cellular distribution of Gag that was comparable to that observed with the wild type. Specifically, Gag expression yielded a stippled appearance along the plasma membrane for the parental Gag and for these mutants . Importance of charged amino acid residues in MA on RNA packaging. | The MA residues R27 and H45, when mutated, revealed significant reductions in RNA packaging efficiencies. To further analyze the importance of these positively charged amino acid residues on RNA packaging, mutations were made at either R27 or H45 to create either a conservative (i.e., R27H or H45R) or nonconservative (i.e., R27G or H45G) change at these residues. Figure shows the effects of these amino acid substitutions on RNA packaging. R27G had a phenotype comparable to that of R27A, while R27H had a phenotype that was comparable to that of the parental vector. Furthermore, H45G had an RNA packaging defect similar to that of H45A, while H45R had a phenotype equivalent to that of the parental vector. These data indicate that there is a correlation between the absence of a positively charged amino acid at positions 27 or 45 and a defect in RNA packaging. The cellular distribution of Gag was not influenced by these mutations, indicating there was no correlation between the RNA packaging defects and Gag membrane localization . Combined effect of MA and NC mutations on RNA packaging. | To test for the potential interplay between MA and NC on RNA packaging, a mutant construct was created in which both MA and NC mutations were introduced into the same vector and then assayed for the ability to package RNA. In particular, the MA/NC double mutant MA H45A/NC C24A was tested. The combined mutant had a 10-fold reduction in VLP production and a severe defect in RNA packaging (98% reduction compared to that of the parental strain) that was significantly lower than the RNA packaging efficiency of either mutant alone . This indicates that the combination of mutations together further reduced the efficiency of RNA packaging. Lack of transient nuclear localization of Gag. | A recent study with Rous sarcoma virus (RSV) revealed that RSV Gag protein enters the nucleus by a nuclear-targeting sequence in the MA domain and is subsequently transported to the cytoplasm by using a CRM1-mediated nuclear export pathway . Transient expression of a dominant-negative CRM1 or by treating cells with leptomycin B (LMB; a drug that attaches to the central domain of CRM1 to disrupt its interaction with nuclear export signals) resulted in the redistribution of Gag from the cytoplasm to the nucleus. The MA mutant, Myr1E, was insensitive to the effects of LMB treatment, apparently because it bypassed the nuclear compartment during virus assembly . Myr1E has a defect in RNA packaging, which implies that RSV nuclear localization of Gag might be involved in viral RNA-Gag interactions. Since we discovered BLV MA mutants with RNA packaging defects like Myr1E, we hypothesized that the viral RNA-Gag interactions involved in BLV RNA packaging might occur in the nucleus after transient nuclear localization of Gag. To test whether BLV Gag is localized in the nucleus, 293T cells transiently transfected with the BLV VLP construct PR+ were treated with either LMB or transiently cotransfected with a dominant-negative CRM1 (DeltaCAN). In addition, fetal lamb kidney cells chronically infected with BLV (FLK-BLV) were either treated with LMB or transiently transfected with DeltaCAN. When 293T cells were either cotransfected with an RSV Gag-green fluorescent protein (GFP) construct and DeltaCAN or transfected with Gag-GFP and then treated with LMB, there was a distinct nuclear localization of Gag-GFP, comparable to that originally reported by Scheifele et al. . When 293T cells were either cotransfected with the BLV vector and DeltaCAN or treated with LMB after transfection with the BLV vector, no BLV Gag was observed in the nucleus of cells . Comparable results were obtained with LMB treatment or transfection of DeltaCAN into FLK-BLV cells. One interpretation of these data is that BLV Gag does not enter the nucleus, implying that the genome recognition event occurs in the cytoplasm. However, it is possible that BLV Gag does enter the nucleus and is exported by a CRM1-independent pathway. DISCUSSION : This is the first report to provide direct genetic evidence that both the MA and NC domains of BLV PrGag are protein determinants of RNA packaging. The unique observation in this study is the identification of several basic amino acid residues in MA that when mutated, led to RNA packaging defects as great as, if not greater than, those observed for single-amino-acid changes in NC. These observations imply that BLV, and perhaps other deltaretroviruses, use distinct viral RNA-protein interactions in genome recognition and RNA packaging. An interesting observation made with the MA mutants analyzed in this study is that the basic residues in MA that influenced RNA packaging had no effect on Gag membrane localization. This is in contrast to studies with HIV-1 that indicate a role for the basic amino residues of MA in membrane localization (by the formation of a charged patch that is thought to promote membrane binding) . However, our data support those from a previous study of the basic residues in the HTLV-1 MA, which found that HTLV-1 MA basic residue mutants led to normal transport of Gag to the membrane as well as cleavage of PrGag . Many of these HTLV-1 mutants had reduced infectivity. Our study found that several of the BLV MA basic residue mutants affected RNA packaging and did not influence Gag membrane targeting (although they did lead to reductions in virus particle production), suggesting that the reduced infectivity observed with the HTLV-1 mutants may be (in some cases) related to an RNA packaging defect. The locations of the basic residues that led to defects in RNA packaging help to provide some mechanistic insights into potential viral RNA-protein interactions involved in genome recognition. Among the NC mutants, the residues that lie within the first and second zinc finger domains, including R35 and R60, had a significant impact on RNA packaging. Similar residues in the HIV-1 NC zinc finger have been implicated in interactions with SL3 RNA and have been shown to be important for RNA packaging by mutagenesis studies . The basic residues in the BLV MA that caused the greatest defects in RNA packaging (i.e., K41 and H45) are located in alpha-helix B . Presently, no structural data are available regarding the potential viral RNA-MA protein interactions. However, it has been previously reported that BLV MA specifically interacts with the viral RNA in a region that spans the locations of SL1 and SL2 . This suggests that the contacts made by BLV MA with the viral RNA may also involve an interaction with SL1 and/or SL2. Interestingly, the HIV-1 MA protein has been found to bind to the viral RNA in the pol gene, via the basic amino acids in MA, and virus mutants that disrupted this RNA-protein interaction had delayed replication kinetics, although no role in RNA packaging was reported . The BLV MA mutants in this study that influenced RNA packaging do not overlap with the known RNA sequence of gag known to be involved in RNA packaging . This indicates that the mutations introduced into the VLP vector influence RNA packaging by the amino acid change in MA and not the nucleotide changes in the RNA. This conclusion is supported by the data in Fig. , which show the importance of charged amino acid residues (and not just nucleotide changes) in RNA packaging. The magnitude of the defects in RNA packaging described here in this study for the most extreme MA and NC mutants (i.e., a MA-NC double mutant) are in the range of those observed previously when SL1 or SL2 was mutated . Taken together, these observations suggest that residues in both MA and NC make direct contacts with the packaging signal or perhaps adjacent sequences that would influence genome recognition. The interaction of NC with RNA sequences outside of those RNA secondary structures of the RNA packaging signal has been shown directly or implied with murine leukemia virus and spleen necrosis virus and is believed to play an important role in RNA packaging . The data in this report provide evidence for a role of both the MA and NC domains of Gag in BLV RNA packaging. This observation implies that distinct viral RNA-protein interactions occur in BLV genome recognition and RNA packaging to that of other retroviruses. Further studies will determine the precise RNA-protein interactions involved in BLV RNA packaging, the location in the cell at which genome recognition occurs, and whether both MA and NC play a role in RNA packaging of other deltaretroviruses. FIG. 1. : Mutagenesis of basic residues and the zinc fingers in the BLV NC domain of PrGag results in RNA packaging defects. Mutagenesis of basic residues and the zinc fingers in the BLV NC domain of PrGag results in RNA packaging defects. (A) The BLV NC domain of PrGag. Amino acid residues in boldface indicate basic residues, and boldface residues that are underlined indicate basic and/or charged residues conserved among BLV, HTLV-1, and HTLV-2. The locations of the two zinc finger (Cys-His box) domains within NC are indicated. (B) Relative RNA packaging efficiencies of BLV NC mutants. RNA packaging efficiencies were determined by the ratio of viral RNA to the amount of Gag from VLPs (see Materials and Methods for details). The RNA packaging efficiency with wild-type BLV NC was set at 100, and the values for the NC mutants (+- standard deviation) are relative to that of the wild type (WT). Mutants were tested in parallel with the parental construct, and each experiment was done in triplicate. FIG. 2. : Mutagenesis of basic residues in the BLV MA domain of PrGag results in RNA packaging defects. Mutagenesis of basic residues in the BLV MA domain of PrGag results in RNA packaging defects. (A) The BLV MA domain of PrGag. Amino acid residues in boldface indicate basic residues, and boldface residues that are underlined indicate basic and/or charged residues conserved among BLV, HTLV-1, and HTLV-2. The alpha-helices in MA are indicated, based upon nuclear magnetic resonance solution structure analysis of BLV MA . (B) Quantitation of VLP production. Preparation of lysates from VLP-producing cells and VLPs and immunoblot analysis are described in Materials and Methods. Representative results from the immunoblot analysis are shown. WT, wild type. (C) Linear range of detection for immunoblot analysis. Protein was diluted (1:1, 1:2, 1:4, 1:5, 1:8, 1:10, and 1:16) and was subjected to immunoblot analysis. The band intensity (arbitrary units) for each dilution was determined with the Quantity One software package of the Chemi Doc 2000 Documentation System (Bio-Rad) and plotted against the amount of protein used for immunoblot analysis (16, 8, 4, 3.2, 2, 1.6, or 1 mul) to determine if protein detection within the dilution range was linear. (D) Relative RNA packaging efficiencies of BLV MA mutants. RNA packaging efficiencies were determined by the ratio of viral RNA to the amount of Gag from VLPs (see Materials and Methods for details). The RNA packaging efficiency with wild-type BLV MA was set at 100, and the values for the MA mutants (+- standard deviation) are relative to that for the wild type. Mutants were tested in parallel with the parental construct, and each experiment was done in triplicate. (E) Cellular distribution of PrGag for selected MA mutants. Cells stably transfected with VLP constructs containing mutations in MA were grown on coverslips, fixed, and incubated with an anti-HA Ig followed by incubation with Alexa Fluor 488-conjugated antimouse Ig. Images were obtained with a confocal microscope. FIG. 3. : Requirement of charged amino acid residues in BLV MA for RNA packaging. Requirement of charged amino acid residues in BLV MA for RNA packaging. RNA packaging efficiencies were determined by the ratio of viral RNA to the amount of Gag from VLPs (see Materials and Methods for details). The RNA packaging efficiency with wild-type (WT) BLV MA was set at 100, and the values for the MA mutants (+- standard deviation) are relative to that for the wild type. (A) Mutants were tested in parallel with the parental construct, and each experiment was done in triplicate. (B) Cellular distribution of PrGag. Cells stably transfected with VLP constructs were grown on coverslips, fixed, and incubated with antibodies prior to confocal microscopy. FIG. 4. : Combined influence of NC and MA mutations on RNA packaging. Combined influence of NC and MA mutations on RNA packaging. RNA packaging efficiencies were determined by the ratio of viral RNA to the amount of Gag from VLPs (see Materials and Methods for details). The RNA packaging efficiency with wild-type BLV was set at 100, and the values for the mutants (+- standard deviation) are relative to that for the wild type (WT). Mutants were tested in parallel with the parental construct, and these experiments were done in triplicate. FIG. 5. : Lack of transient nuclear localization of BLV Gag. Lack of transient nuclear localization of BLV Gag. 293T cells were transiently transfected alone with an RSV Gag-GFP construct or a BLV vector (PR+) and treated with LMB (10 ng/mul) or were transiently cotransfected with the RSV or BLV construct and a dominant-negative CRM1 (DeltaCAN) (10 mug per approximately 5 x 105 cells). In parallel, FLK cells chronically infected with BLV were treated with LMB (10 ng/mul) or were transiently transfected with DeltaCAN (10 mug per approximately 5 x 105 cells). Forty-eight hours posttransfection, cells were fixed and then analyzed by confocal microscopy. Backmatter: PMID- 12915532 TI - Respiratory Syncytial Virus Infection Sensitizes Cells to Apoptosis Mediated by Tumor Necrosis Factor-Related Apoptosis-Inducing Ligand AB - Respiratory syncytial virus (RSV) is an important cause of respiratory tract disease worldwide, especially in the pediatric population. For viruses in general, apoptotic death of infected cells is a mechanism for reducing virus replication. Apoptosis can also be an important factor in augmenting antigen presentation and the host immune response. We examined apoptosis in response to RSV infection of primary small airway cells, primary tracheal-bronchial cells, and A549 and HEp-2 cell lines. The primary cells and the A549 cell line gave generally similar responses, indicating their appropriateness as models in contrast to HEp-2 cells. With the use of RNase protection assays with probes representing 33 common apoptosis factors, we found strong transcriptional activation of both pro- and antiapoptotic factors in response to RSV infection, which were further studied at the protein level and by functional assays. In particular, RSV infection strongly up-regulated the expression of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) and its functional receptors death receptor 4 (DR4) and DR5. Furthermore, RSV-infected cells became highly sensitive to apoptosis induced by exogenous TRAIL. These findings suggest that RSV-infected cells in vivo are susceptible to killing through the TRAIL pathway by immune cells such as natural killer and CD4+ cells that bear membrane-bound TRAIL. RSV infection also induced several proapoptotic factors of the Bcl-2 family and caspases 3, 6, 7, 8, 9, and 10, representing both the death receptor- and mitochondrion-dependent apoptotic pathways. RSV also mediated the strong induction of antiapoptotic factors of the Bcl-2 family, especially Mcl-1, which might account for the delayed induction of apoptosis in RSV-infected cells in the absence of exogenous induction of the TRAIL pathway. Keywords: Introduction : Human respiratory syncytial virus (RSV) is the most common cause of mortality associated with serious viral bronchiolitis and pneumonia in infants throughout the world (reviewed in reference ). RSV also is increasingly recognized as an important cause of respiratory tract disease in adolescents and adults and as a leading cause of mortality associated with viral pulmonary infections in the elderly . Currently, an RSV vaccine is not yet available, although several live attenuated vaccine candidates are under clinical trials . Apoptosis, or programmed cell death, is a common response of cells to infection with viruses . Apoptosis can be induced when specific transmembrane death receptors are engaged by exogenous Fas ligand, tumor necrosis factor (TNF), or TNF-related apoptosis-inducing ligand (TRAIL) borne on the membrane of effector immune cells or, in the case of TRAIL, also expressed as a secreted trimer. Apoptosis also can occur when intracellular changes trigger a mitochondrion-mediated pathway. Each pathway includes formation of intracellular complexes for the activation of initiator caspases (e.g., caspases 8, 9, and 10), which in turn activate effector caspases (e.g., caspases 3, 6, and 7), which in turn cleave cellular protein substrates and trigger the classic degenerative changes specific for apoptosis. In some situations apoptosis can contribute to pathogenesis, but more typically it is an important factor in host defense that hastens the death of infected cells and thereby limits the replication and spread of the virus (see reference for a review). As a consequence, many viruses have evolved various mechanisms to inhibit or evade apoptosis. In addition, death by apoptosis instead of necrosis can significantly affect the efficiency of capture of viral antigens by antigen-presenting cells and presentation to T cells. The most-potent antigen-presenting cells, dendritic cells, were shown to internalize apoptotic but not necrotic cells and process them for presentation by both major histocompatibility complex class I and class II molecules; some studies also demonstrated subsequent efficient induction of T-cell responses . For example, loading of dendritic cells with human immunodeficiency virus-infected apoptotic, but not necrotic, cells induced proliferation of human immunodeficiency virus-specific CD4+ and CD8+ cells . In the context of virus infection in vivo, a recombinant rabies virus that was engineered to express cytochrome c, an effective apoptosis-inducing protein, was associated with attenuated pathogenicity and an increased induction of virus-neutralizing antibodies . The concept that apoptosis can play a significant role in pathogenesis and the host immune response is of direct relevance for vaccine development. Information concerning apoptosis in RSV-infected cells has been inconsistent and conflicting. Two early studies identified potential mediators of apoptosis induced in response to RSV infection, namely, interleukin-1 converting enzyme and Fas , but only in the latter study was apoptosis detected, and then only at a low level. A third study also documented elevated apoptosis in response to RSV infection but provided evidence that the Fas pathway was not an important mediator. Instead, the authors proposed that RSV-induced apoptosis is initiated as part of an endoplasmic reticulum-associated stress response involving caspase 12 . However, this hypothesis was placed in doubt by the subsequent finding that the human caspase 12 gene contains a frameshift and additional loss-of-function mutations that should preclude the expression of functional caspase 12 . Other recent studies implicated RSV in inducing antiapoptotic pathways. In one case, RSV infection of HEp-2 cells induced the antiapoptotic factor IEX-1L . In a second case, treatment of RSV-infected cells with an inhibitor of phosphatidylinositol 3-kinase (PI-3K) resulted in more-rapid apoptosis, implying that under normal conditions signaling through the PI-3K pathway mediates an inhibition of RSV-induced apoptosis . The idea that apoptosis is inhibited or delayed in RSV-infected cells is consistent with the minimal cytopathic effect observed in response to RSV infection of primary human airway cells in a reconstituted differentiated pseudostratified mucociliary epithelium in vitro . In the present study, we investigated apoptosis during RSV infection, in particular in primary cell cultures as well as in transformed cell lines. RNase protection assays (RPAs) using a broad panel of probes identified up-regulation of both pro- and antiapoptotic factors, representing both the death receptor and mitochondrial pathways. In particular, RSV strongly up-regulated expression of both TRAIL and its functional receptors and strongly sensitized cells to apoptosis induced by exogenous TRAIL. In the absence of exogenous TRAIL, apoptosis occurred very late in RSV infection, a delay that might reflect the strong up-regulation of antiapoptotic factors, including Mcl-1. MATERIALS AND METHODS : Cells and viruses. | Primary normal human bronchial-tracheal epithelial cells (NHBE) (with retinoic acid to inhibit cell differentiation) and small airway epithelial cells (SAEC) from a single donor were obtained from Biowhittaker Inc. (Walkersville, Md.). The cells were passaged up to three times according to the manufacturer's recommendations using the supplied trypsin and medium. The A549 type II alveolar adenocarcinoma cell line was obtained from the American Type Culture Collection (Manassas, Va.) and used up to passage 100. The HEp-2 line (CCL-23) was obtained from the American Type Culture Collection and used up to passage 400. Recombinant RSV strain A2 and influenza A virus purified strain H1N1 (Advanced Biotechnologies Inc., Columbia, Md.) were used at a multiplicity of infection (MOI) of 2 PFU per cell. Quantitation of RSV was performed by plaque assay using monoclonal antibody (MAb) staining of plaques . For quantitation of intracellular RSV, cells were collected with a cell scraper, washed, resuspended in a small volume of cell culture medium (1 ml of medium per 106 cells), subjected to freezing and thawing, and centrifuged to generate a clarified supernatant that was quantitated by plaque assay. For some experiments, RSV was inactivated by UV light and confirmed to lack residual infectivity by plaque assay. Analysis of mRNA. | RPA was done using a RiboQuant Multi-Probe RPA system with three different mixtures of probe templates, namely, hAPO-1c, hAPO-2b, and hAPO-3d (BD PharMingen, San Diego, Calif.), according to the manufacturer's recommendations. The data were quantitated using a PhosphorImager 445-SI (Molecular Dynamics, Sunnyvale, Calif.). For RT-PCR of TRAIL mRNA, total RNA was subjected to reverse transcription using a random primer mix and SuperScriptII RNase H- reverse transcriptase (Invitrogen, Carlsbad, Calif.) followed by PCR with human-mouse TRAIL PCR Primer Pair (R&D Systems, Minneapolis, Minn.) and Vent DNA polymerase (New England Biolabs, Beverly, Mass.) using the following parameters of PCR: melting at 94C for 35 s; annealing at 55C for 35 s; synthesis of DNA at 72C for 45 s (30 cycles were performed). Flow cytometry analyses. | Annexin V staining was performed using fluorescein isothiocyanate (FITC)-labeled annexin V (BD Biosciences, San Diego, Calif.) according to the manufacturer's recommendations. Terminal deoxynucleotidyltransferase dUTP nick end labeling (TUNEL) assay was performed using FlowTACS In Situ TUNEL-based apoptosis detection kit (R&D Systems) according to the manufacturer's recommendations. For flow cytometry analysis of cell surface TRAIL, cells were stained with primary anti-human TRAIL MAb 687 or mouse MAb 002 (isotype control) (both from R&D Systems), followed by secondary antibody, R-phycoerythrin (R-PE)-labeled goat anti-mouse immunoglobulin G (IgG) (Caltag Laboratories, Burlingame, Calif.). Death receptor 4 (DR4) expression was analyzed using goat polyclonal IgG specific to extracellular domain of human DR4 or normal goat IgG (isotype control) (both from R&D Systems), followed by R-PE-labeled swine anti-goat IgG (Caltag Laboratories). DR5 was analyzed using goat polyclonal IgG specific to extracellular domain of human DR5 (R&D Systems). For quantitation of intracellular Mcl-1 by flow cytometry, cells were fixed and permeabilized using a DAKO IntraStain fixation and permeabilization kit (DAKO Corporation, Carpinteria, Calif.) and stained with human Mcl-1-specific purified rabbit polyclonal IgG or, as a control, normal rabbit IgG (both from Santa Cruz Biotechnology, Inc., Santa Cruz, Calif.), and this was followed by staining with FITC-labeled goat anti-rabbit IgG (Caltag Laboratories). Analysis of enzymatic activity of caspases. | Analysis of caspase activity was performed using the pan-caspase inhibitor valyl-alanyl-aspartyl fluoromethyl ketone (VAD-FMK) conjugated with FITC (Promega, Madison, Wis.). A total of 500,000 cells were labeled with 10 mul of a 1:50 dilution of FITC-VAD-FMK for 20 min at 37C, followed by flow cytometry. For analysis of enzymatic activities of individual caspases, cell lysates were prepared and normalized according to protein concentration determined using the BCA Protein Assay reagent kit (Pierce, Rockford, Ill.). Analysis was performed using caspases 8, 3, 6, and 9 (Colorimetric activity assays; R&D Systems) according to the manufacturer's recommendations. Briefly, cell lysate samples were diluted with incubation buffer in 96-well plates, p-nitroaniline-labeled peptide substrates specific for the individual caspases were added, and the plates were incubated at 37C for 2 h and read at 405 nm. Western blot analysis. | Lysates of RSV-infected or mock-infected cells were normalized for protein concentration, subjected to electrophoresis on a sodium dodecyl sulfate-4 to 20% polyacrylamide gel, transferred to a polyvinyl difluoride membrane (Invitrogen), and analyzed by incubation with rabbit anti-Mcl-1 antibodies (Santa Cruz Biotechnology) at a 1:200 dilution followed by peroxidase-labeled goat anti-rabbit IgG (WesternBreeze; Invitrogen) according to the manufacturer's recommendations. Immunohistochemical studies. | A549 cells were seeded on glass coverslips in six-well plates and infected with RSV. At different times postinfection, cells were rinsed with phosphate-buffered saline and fixed with 2% formaldehyde at room temperature for 10 min, which was followed by two washes with phosphate-buffered saline. To detect Mcl-1 protein, cells were incubated with a 1:50 dilution of human Mcl-1-specific rabbit antibody (Santa Cruz Biotechnology), and this was followed by staining with a 1:400 dilution of FITC-conjugated donkey anti-rabbit antibody (Jackson Immunoresearch Laboratories, West Grove, Pa.). To visualize nuclei, cells were stained with the fluorescent DNA-intercalating dye 4,6-diamidino-2-phenylindole (DAPI) (1 mug/ml) for 30 min at room temperature. Analysis of soluble TRAIL in cell media. | The level of TRAIL in medium was determined using TRAIL ActivELISA (Imgenex, San Diego, Calif.) according to the manufacturer's recommendations. Briefly, a 96-well plate was covered with anti-TRAIL polyclonal antibodies, blocked with bovine serum albumin, and incubated with cell media or standard dilutions of recombinant soluble TRAIL. Soluble TRAIL captured on the plate was then detected by TRAIL-specific detecting antibody followed by incubation with alkaline phosphatase-conjugated secondary antibody and color reaction with p-nitrophenyl phosphate. Optical density was measured at 405 nm and compared with those for the standard dilutions. Sensitization of RSV-infected cells to apoptosis induced by TRAIL. | A549 cells were infected with RSV. After incubation at 37C for the various times indicated in Results, synthetic TRAIL was added to a final concentration of 500 or 250 ng/ml with or without a TRAIL-specific antibody (2 mug/ml) that increased the ability of TRAIL to induce apoptosis (both reagents from Alexis Corporation, San Diego, Calif.). The level of apoptosis was quantitated by annexin V staining followed by flow cytometry analysis. RESULTS : RSV replication in primary lung epithelial cells. | One possible factor in the inconsistency in results from previous studies of RSV-induced apoptosis was the use of established cell lines for most of the published experiments. We therefore evaluated RSV infection of primary NHBE and primary human SAEC in addition to two transformed human cell lines, A549 and HEp-2. The A549 line originated from type II alveolar adenocarcinoma cells, and the HEp-2 line originally consisted of laryngeal epidermal carcinoma cells but now is available commercially as a line that is mixed with HeLa cervical carcinoma cells due to cross-contamination . As will be seen later, results obtained with the A549 cells generally resembled those of the primary cells, whereas this often was not the case with HEp-2 cells. The NHBE and SAEC cells were compared to A549 cells for the ability to support the replication of RSV . With regard to release of the virus into the medium, the A549 cells appeared to lag behind the primary cells at the 12- and 24-h points, but by 48 h the titer of virus released from A549 cells was similar to that for the primary cells. With regard to cell-associated virus, the A549 cells lagged behind the primary cells at 12 h, and at 24 and 48 h there was a consistent range of titers, with that of the NHBE cells being the highest, followed by those of the SAEC and A549 cells. Thus, each of the cultures supported efficient RSV replication, although there were cell-specific differences in kinetics and efficiency of virus release. Following RSV infection of NHBE cells, cell rounding was observed in a small fraction of cells at 12 h postinfection, followed by more extensive cell rounding and some syncytium formation at the later time points . In SAEC, extensive cell rounding and some syncytium formation started at 12 h. In A549 cells, significant changes were not observed at 12 and 24 h postinfection, but extensive syncytium formation and cells rounding started at 48 h. Generally, the appearance of morphological changes in monolayers of the three cell types correlated well with accumulation of the infectious virus in cells and media , with SAEC being the most sensitive and A549 the least sensitive to RSV infection. Also, A549 cells exhibited greater syncytium formation and less cell rounding than the primary cells. These cytopathic effects likely reflect a combination of direct virus-mediated damage and cell-mediated apoptosis. Cell rounding is associated with, but is not specific for, apoptosis, while syncytium formation is mediated by the viral F protein but can amplify apoptotic signals . RSV induces apoptosis in A549 and primary lung epithelial cells. | To monitor apoptosis more directly, RSV-infected NHBE, SAEC and A549 cells were analyzed by staining with annexin V . Annexin V has a high affinity for membrane phospholipid phosphatidylserine (reviewed in reference ), which is translocated from the inner to the outer leaflet of the plasma membrane at an early stage of apoptosis . In A549 cells, there was little staining above that of the control cells at 24 h, but by 48 h there was extensive staining . In NHBE cells, there was a moderate increase in the percentage of apoptotic cells at 24 h, and there was a significant increase (to 55%) at 48 h. In SAEC cells, a more rapid increase in the percentage of apoptotic cells was observed, with approximately 70% of the cells staining at 24 h. The percentages of annexin V-positive mock-infected NHBE and SAEC were higher than that of A549 cells, likely due to a lower viability of the primary cells in vitro. Thus, apoptosis measured by annexin V staining generally paralleled or followed the production of infectious virus and the appearance of gross morphological changes noted in Fig. . Annexin V staining of cells infected with UV-inactivated RSV was essentially equivalent to that of the uninfected control, indicating that RSV replication is necessary for the induction of apoptosis . As a comparison, cells infected in parallel with influenza A virus exhibited a somewhat lower level of apoptosis than those infected with RSV . Apoptosis in A549 cells also was monitored by the TUNEL assay, which detects DNA fragmentation and thus measures a downstream step in apoptosis . Positive TUNEL staining was observed at 48 and 72 h postinfection, but not at 24 h, which correlates with the results of annexin V staining. In situ staining of RSV-infected A549 cells with the fluorescent DNA-intercalating dye DAPI demonstrated strong condensation of chromatin, a characteristic apoptotic change, in a small fraction (<10%) of RSV-infected cells at 48 h postinfection but not at the earlier time points (not shown). RSV induces activation of initiator caspases 8, 9, and 10 and effector caspases 3, 6, and 7. | Total caspase activation in RSV-infected cells was studied with FITC-conjugated VAD-FMK, which binds irreversibly to activated caspases. A549 cells were mock infected or infected with RSV for 12, 24, or 48 h; placed in suspension by treatment with trypsin; washed; stained with FITC-VAD-FMK; and assayed by flow cytometry . At 12 h postinfection, the number of FITC-VAD-FMK-positive cells was the same in RSV- and mock-infected cells. At 24 h a small increase in the number of the positive cells was observed in RSV-infected but not mock-infected cells; at 48 h approximately 50% of RSV-infected cells were positive, whereas uninfected control cells did not show an increase in staining. RPAs were used to monitor the levels of caspase mRNAs in RSV-infected NHBE, SAEC, and A549 and HEp-2 cells . Marked increases in the levels of mRNAs encoding initiator caspases 8 and 10 and effector caspases 3, 6, and 7 were detected in the primary NHBE and SAEC at the earliest time point analyzed (12 h). Marked increases also were observed for each of these caspase mRNAs in A549 cells by 24 h and, especially, 48 h except for caspase 6 mRNA, which was significantly increased only at 72 h. In contrast, in HEp-2 cells there was only a modest, twofold increase in the levels of mRNA for caspases 10, 3, and 6, and there was no change for the mRNAs for caspase 8 and 7 . The increases in caspases 3, 8, and 6 in A549 cells were accompanied by increase in the corresponding enzymatic activities ; caspases 7 and 10 were not assayed. Caspase 9 mRNA was not detectable in NHBE and SAEC cells, while RSV infection of A549 and HEp-2 cells induced a small (up to twofold) increase in caspase 9 mRNA (data not shown). The lack of caspase 9 mRNA detectable by RPA in the primary cells might be due to a low level of transcription of this gene. However, caspase 9 enzymatic activity was detected in A549 cells beginning 36 h postinfection . No increase in caspase 2 and 5 mRNA was detected with RSV infection (not shown). Thus, RSV infection resulted in up-regulation of mRNA for caspases 8 and 10, which are initiator caspases associated with the death receptor pathway; weak up-regulation of caspase 9, an initiator caspase associated with the mitochondrial pathway; and up-regulation of caspases 3, 6, and 7, which are effector caspases utilized by either pathway. RSV induces expression of pro- and antiapoptotic proteins of the Bcl-2 family. | The Bcl-2 family contains a number of pro- and antiapoptotic factors whose balance can play a role in the decision to initiate apoptosis. RNase protection demonstrated up-regulation of a number of pro- and antiapoptotic Bcl-2 family members in response to RSV infection. mRNA encoding three proapoptotic members, Bid, Bax and Bak, was up-regulated two- to sevenfold in RSV-infected NHBE, SAEC, and A549 cells, whereas only Bak was up-regulated in HEp-2 cells . Up-regulation also was observed for antiapoptotic members of the Bcl-2 family. mRNA for Mcl-1 was strongly up-regulated in RSV-infected NHBE, SAEC, and A549 and HEp-2 cells . The up-regulation of Mcl-1 mRNA was detected at 12 h for NHBE and SAEC cells and 24 h for A549 cells, which were the earliest times tested, and by 16 to 24 h for HEp-2 cells. In RSV-infected primary cells, Mcl-1 mRNA reached levels that were 200 to 400% of the amount of the abundant GAPDH (glyceraldehyde-3-phosphate dehydrogenase) mRNA, representing an increase of up to 10-fold. An increase in mRNA for two other antiapoptotic Bcl-2 family members, Bcl-W and Bcl-xL, also was observed in all four cell cultures in response to RSV infection, although the increase was less marked in HEp-2 cells . Also, the level of these mRNAs relative to GADPH mRNA was less than for Mcl-1 . Up-regulation of mRNA for Bfl-1, another antiapoptotic member of Bcl-2 family was detected in RSV-infected A549 cells but not in the other cell lines (not shown). The accumulation of Mcl-1 protein in RSV-infected A549 cells was analyzed at 12 and 24 h postinfection by intracellular staining and flow cytometry and Western blotting . Flow cytometry analysis demonstrated a slight increase in the median expression of Mcl-1 protein at 12 h and significant (P < 0.001) increase in Mcl-1 at 24 h. Western blotting showed elevated levels of the Mcl-1 protein at 12 and 24 h post-RSV infection. The increase in Mcl-1 protein was also detected by immunohistochemistry at 12 and 24 h postinfection . RSV induces production of TRAIL in infected cells. | RPAs with NHBE and SAEC primary cells and the A549 cell line showed a strong up-regulation of TRAIL mRNA , whereas no increase was observed in HEp-2 cells (not shown). TRAIL is a member of the TNF family and, together with its functional receptors DR4 and DR5 , comprises one of the death receptor pathways for apoptosis. The up-regulation of TRAIL mRNA was particularly strong in primary NHBE and SAEC, where it increased from the low level seen in mock- or UV-RSV-infected cell controls to a level that was within 55 to 75% of GAPDH mRNA . TRAIL mRNA also was induced in A549 cells, but its abundance relative to GADPH mRNA was nearly 10-fold lower than in primary cells. In each of the three cell types, up-regulation of TRAIL mRNA was detected at the earliest time point tested (12 h for primary cells and 24 h for A549 cells). The expression of transmembrane TRAIL protein on the cell surface was quantitated in mock-infected and RSV-infected A549 cells by flow cytometry . An increase in the number of TRAIL-positive cells was observed at 48 and 72 h after RSV infection but not at 24 h postinfection . TRAIL also can be expressed in a soluble, truncated form that results from proteolytic cleavage of the ectodomain from the transmembrane anchor . However, soluble TRAIL in the medium of RSV-infected or mock-infected cells was not detected by an antigen-capture enzyme-linked immunosorbent assay that had a lower limit of detection of 40 pg/ml (data not shown). RSV induces expression of TRAIL receptors DR4 and DR5. | mRNA encoding the two principal functional receptors for TRAIL, DR4 and DR5, was quantified by RPA. In NHBE, both DR4 and DR5 mRNAs were strongly up-regulated by 12 h postinfection and remained elevated at 24 and 48 h, with peak levels of 35% (for DR4) and 95% (for DR5) of that of GAPDH mRNA . In SAEC, DR4 and DR5 mRNAs were markedly elevated at 12 h but diminished thereafter, with the peak levels approximately 50 and 130%, respectively, that of GAPDH mRNA. In A549 cells, the levels of DR4 and DR5 mRNA increased steadily at 24, 48, and 72 h, but the levels relative to GAPDH mRNA were substantially lower than for primary cells. In HEp-2 cells, the up-regulation of DR4 or DR5 mRNA in response to RSV infection was twofold or less, and the level of expression was much lower than for primary or A549 cells. We also quantified the levels of TRAIL decoy receptor 1 (DcR1) and DcR2) mRNAs, which can bind TRAIL and block its effects; no detectable increase in these mRNAs was detected after RSV infection in any of the cells (data not shown). Expression of DR4 and DR5 proteins on the surface of RSV-infected A549 cells was assayed by flow cytometry. Up-regulation of DR4 was detected at 48 h and increased at 72 h postinfection . Expression of DR5 in RSV-infected cells was elevated at 24, 48, and 72 h. Thus, RSV infection up-regulates the expression of DR4 and DR5 receptors for TRAIL on the surface of cells. We also observed an increase in the expression of Fas mRNA after RSV infection (data not shown), consistent with the results of O'Donnell et al. . Up-regulation of Fas was observed in NHBE, SAEC, and A549 cells, but not in HEp-2 cells. Infection of primary cells with RSV resulted in a 2- to 6-fold increase in Fas mRNA, while infection of A549 cells resulted in a 2- to 12-fold increase; these values were about 2 to 20% that of GAPDH mRNA (data not shown). RSV sensitizes cells to apoptosis induced by TRAIL. | Increased surface expression of the DR4 and DR5 TRAIL receptors can sensitize a cell to apoptotic killing by exogenous TRAIL, which is present on the surface of a variety of immune cells, including natural killer cells, CD4+ T lymphocytes, macrophages, and dendritic cells (, -, , ), and which also can be released as a secreted soluble protein. To determine whether RSV infection resulted in an increased sensitivity to TRAIL-mediated apoptosis, A549 cells were infected with RSV, and 24 h later, exogenous TRAIL was added to the medium to a final concentration of 500 ng/ml together with a TRAIL-specific antibody that enhances TRAIL activity by cross-linking and increasing its local concentration. Before TRAIL and anti-TRAIL antibody were added, there was no difference in cell morphology between RSV-infected and noninfected cells; however, within 3 h after the addition of TRAIL plus anti-TRAIL antibody, the RSV-infected cells exhibited dramatic cytopathic effects. RSV-infected cells treated with TRAIL plus anti-TRAIL antibody rounded up, and within 5 h after the addition, most of the cells were floating in the medium (Fig. , left panels). No significant changes were observed in uninfected cells treated in parallel with TRAIL plus anti-TRAIL antibody, nor in untreated RSV-infected cells . Also, addition of anti-TRAIL antibody alone to uninfected or RSV-infected cells had no effect (not shown). Addition of TRAIL to RSV-infected cells in the absence of the TRAIL-specific antibody induced very little cytopathic effect after 5 h of incubation; following a further 3 h of incubation, the amount of cytopathic effect was increased but remained considerably less than that observed after 5 h of treatment with TRAIL plus TRAIL-specific antibody (not shown). To quantify the percentage of apoptotic cells, the cell monolayers were harvested together with floating cells, stained with annexin V-FITC, and analyzed by flow cytometry (Fig. , right panels). The percentage of annexin V-positive cells in the three negative controls, namely, mock-infected cells, mock-infected cells treated with TRAIL plus anti-TRAIL antibody, and RSV-infected cells without TRAIL or anti-TRAIL antibody, was 24 to 26%. In contrast, 42% of RSV-infected cells treated with TRAIL plus anti-TRAIL antibody were annexin V positive. Thus, RSV strongly sensitizes cells to rapid TRAIL-induced apoptosis. The time course of acquisition of sensitivity to TRAIL-induced apoptosis during the first 25 h of RSV infection was studied in A549 cells . Cells were infected with RSV and at various times postinfection were exposed to TRAIL (250 ng/ml) plus TRAIL-specific antibody and incubated for an additional 5 h before being collected and analyzed for annexin V staining. There was a steady increase in sensitivity to exogenous TRAIL with increasing time after infection, with a marked increase in sensitivity evident by 11 h postinfection , long before the production of infectious progeny virus . DISCUSSION : We examined the induction of apoptosis following RSV infection of two types of human primary pulmonary epithelial cells, namely, NHBE and SAEC, and two cell lines commonly used for RSV studies, namely, A549 cells and HEp-2 cells. RPAs were used to quantify the mRNAs for 33 common apoptosis factors, which identified a number of pro- and antiapoptotic factors that were up-regulated in response to RSV infection. In general, results obtained with the two primary cells were in close agreement. Results with A549 cells also were generally in good agreement with those of the primary cells, although sometimes the magnitude of the response was reduced; in contrast, the transcriptional profile of pro- and antiapoptotic response in HEp-2 cells often was distinct, suggesting that HEp-2 cells are not a reliable substrate for such studies. The results with the primary and A549 cells showed that RSV induces apoptosis in a large fraction of infected cells, although the programmed degenerative changes usually occurred late in infection compared to virus production. RSV infection was shown to strongly up-regulate the components of the TRAIL death receptor pathway and strongly sensitized the cells to apoptosis mediated by exogenous TRAIL. Evidence was found of activation of components of both the death receptor and mitochondrial apoptotic pathways. In addition, there was strong up-regulation of antiapoptotic factors, notably Mcl-1, which might account for the apparent slow onset in the execution of apoptosis. TRAIL is an apoptosis-inducing ligand that belongs to the TNF superfamily, together with TNF alpha and FAS ligand (FasL or CD95) . TRAIL is a type II transmembrane protein that can undergo cleavage to release a C-terminal soluble domain that can form trimers. Both the soluble and membrane-associated forms can induce apoptosis , although in the present study we were unable to detect soluble TRAIL in medium from RSV-infected cells. Expression of TRAIL has been detected in a number of human tissues, most predominantly in spleen, lungs, and prostate ; inducers include alpha/beta and gamma interferons . TRAIL, like TNF alpha and FasL, is an inducer of the death receptor (external) pathway of apoptosis and is involved in apoptosis induced by reovirus , adenovirus , and cytomegalovirus ; in addition, measles virus induces TRAIL production by dendritic cells . TRAIL can bind to five different receptors: DR4 and DR5 are functional receptors that mediate apoptotic responses and were strongly up-regulated by RSV infection, and DcR1 , DcR2 , and osteoprotegrin are decoy receptors that bind TRAIL without inducing apoptosis and were not detected in the present study. Binding of TRAIL to DR1 or DR2 leads to formation of death-inducing signaling complex consisting of (i) DR4 or DR5 trimers, (ii) an adaptor molecule such as FAS-associating protein with death domain (FADD), and (iii) pro-caspase 8, which is activated by self-cleavage upon binding the complex. Caspase 8 activates effector caspases 3 and 7, which in turn cleave cellular substrates, resulting in irreversible apoptosis . Alternatively, formation of death-inducing signaling complex can involve caspase 10 instead of caspase 8 . TRAIL also can induce apoptosis through mitochondrion-dependent pathway involving the cleavage of Bid protein by activated caspase 8, the loss of mitochondrial potential, and the release of cytochrome c from mitochondria, resulting in activation of caspase 3 . Transcripts for Bid, Bax, and Bak, three proapoptotic members of the Bcl-2 family associated with the mitochondrion-dependent pathway, were found to be up-regulated in the present study. Infection with RSV also induced production of DR4 and DR5, with elevation of DR5 on the cell surface occurring earlier than that of DR4. Importantly, RSV infection efficiently sensitized cells to rapid apoptotic death induced by exogenous TRAIL added at 24 h postinfection, a time when no apoptotic changes could be detected by observation. This presumably represents a massive induction of apoptosis mediated by the RSV-induced TRAIL surface receptors, consistent with previous studies demonstrating a correlation between the surface expression of TRAIL receptors and sensitivity to TRAIL-induced apoptosis . It seems likely that RSV-mediated sensitization of infected cell monolayers to TRAIL-induced apoptosis in vitro reproduces a mechanism of host defense in vivo whereby infected cells render themselves susceptible to killing mediated by (i) soluble TRAIL or cell surface TRAIL expressed by neighboring RSV-infected cells or (ii) immune cells bearing cell surface TRAIL. TRAIL has been shown to mediate the cytotoxic effect of a variety of immune cells, including NK cells, CD4+ T lymphocytes, macrophages, and dendritic cells and in some studies was up-regulated on the surface of CD8+ T lymphocytes (, -, , ). Interestingly, CD4+ cell-mediated cytotoxicity was involved in protection of the upper respiratory tract from RSV infection in a study of an experimental RSV vaccine in mice . It is noteworthy that the infected cell monolayers exhibited increased sensitivity to TRAIL-mediated apoptosis as early as 6 h postinfection, the earliest time examined, and were extremely sensitive by 16 to 25 h. Thus, the up-regulation of the TRAIL receptors in response to RSV infection provides a death ligand-dependent pathway for the rapid destruction of infected cells early in the viral replicative cycle. In the absence of deliberate activation of this pathway by added TRAIL, apoptosis eventually was induced in a large fraction of RSV-infected cells, but this happened late in infection, after extensive release of progeny virus. The pathway for this late apoptosis is unclear and might involve soluble TRAIL that was below the level of detection or might involve internal activation of the mitochondrion-dependent pathway. Increased activity was observed for factors associated with each pathway. RSV also induced the early production of both pro- and antiapoptotic proteins of the Bcl-2 family: the proapoptotic members were Bid, Bax, and Bak, and the antiapoptotic members were Mcl-1, Bcl-W, and Bcl-xL. In particular, mRNA for Mcl-1 was strongly up-regulated to a level that was two- to fourfold higher than that of the abundant GAPDH mRNA marker. The level of Mcl-1 protein was shown by Western blot, flow cytometry, and immunohistochemistry to be increased. All members of the Bcl-2 family have at least one of the four Bcl-2 homology (BH) functional domains, BH1 to BH4 (reviewed in reference ); for example, Mcl-1 and Bcl-2 each have BH1, BH2, and BH3 domains . The ratio between the levels of the proapoptotic and antiapoptotic members of the family plays a role in the decision between apoptosis versus cell survival . Bcl-2, the prototype member of the family, is localized in the outer mitochondrial membrane, the nuclear envelope, and the endoplasmic reticulum . Apoptosis induced by TRAIL can be delayed or inhibited by overexpression of the Bcl-2 or Bcl-xL antiapoptotic protein, an effect that has been demonstrated in various cells , including lung cancer cells . It also has been suggested that Mcl-1 might be involved in the inhibition of the mitochondrial pathway of TRAIL-induced apoptosis . In differentiating human myeloblastic leukemia U937 cells, antisense depletion of Mcl-1 resulted in rapid induction of apoptosis, although in that study the involvement of TRAIL was not studied . Up-regulation of Mcl-1 and possibly other antiapoptotic factors of the Bcl-2 family could account for the delayed onset of apoptosis in RSV-infected cells. In an in vitro model of well-differentiated pseudostratified mucociliary epithelium, RSV infection resulted in minimal cytopathic changes, suggesting that apoptosis was inhibited . It might be that antiapoptotic proteins also are induced in that experimental system. The up-regulation by RSV of both pro- and antiapoptotic members of the Bcl-2 family might be mediated by NF-kappaB. Transcriptional activation by NF-kappaB has been demonstrated for at least two members of the Bcl-2 family, namely, Bfl-1 and Bcl-xL , both of which are up-regulated in RSV-infected cells in the present study. Activation of NF-kappaB by RSV infection has been reported by several groups . This is mediated by interaction of RSV with toll-like receptor 4 and possibly by the production of double-stranded RNA during the infection. In addition, engagement of receptors DR4 and DR5 by TRAIL activates the NF-kappaB pathway , which might represent another mechanism of NF-kappaB activation during RSV infection. In contrast, the up-regulation of Mcl-2 is not mediated by NF-kappaB, but instead by the PI-3K pathway, which involves activation of the serine/threonine kinase Akt-1 . Infection of cells with RSV activates the PI-3K pathway , which might explain the stronger up-regulation of Mcl-1 than of the other members of the Bcl-2 family here . It will be important to directly demonstrate whether one or more antiapoptotic factors indeed inhibit apoptosis in RSV-infected cells. Since apoptosis limits virus replication and spread and enhances antigen presentation, inhibition of apoptosis could be a significant factor in viral pathogenesis and the host immune response. Suppression of Mcl-1 synthesis in vivo, for example, might result in the rapid induction of apoptosis in RSV-infected cells, thereby blocking spread of the virus to adjacent cells. Antisense oligonucleotides against Bcl-2, another antiapoptotic member of the family, have been demonstrated to be effective against various cancers by promoting apoptotic death of tumor cells and are currently in clinical trials for using alone or in combination with chemotherapeutic drugs (, ; reviewed in reference ). It also will be interesting to see if individual RSV proteins have any role in pro- or antiapoptotic responses. For some viral genes, this can be directly investigated using available gene-deletion viruses . Modification of an RSV vaccine virus to induce rather than inhibit apoptosis might increase its attenuation and immunogenicity. FIG. 1. : RSV replication in normal NHBE, SAEC, and the A549 cell line following infection with an input MOI of 2. RSV replication in normal NHBE, SAEC, and the A549 cell line following infection with an input MOI of 2. (A) Kinetics of accumulation of cell-associated RSV and free virus released into the medium. The data represent the average titers from three wells of six-well plates per time point for each type of cell. (B) Photomicrographs illustrating cytopathological changes in NHBE, SAEC, and A549 cells at various time points following RSV infection at an MOI of 2. The 0-h time point represents the cells immediately before the infection. Mock-infected cells incubated and photographed in parallel at all time points looked similar to that at 0 h and are not shown. FIG. 2. : Kinetics of apoptosis induced by RSV. Kinetics of apoptosis induced by RSV. (A) Quantitation of annexin V-positive cells by flow cytometry in the primary NHBE and SAEC cells and A549 cells. The percentages of annexin V-positive cells infected with RSV, UV-RSV, or influenza virus or mock infected are shown as the mean of two samples per group for the primary cells and four samples per group for A549 cells. (B) TUNEL assay for apoptosis of RSV-infected or mock-infected A549 cells. The results are expressed as the ratio of the median fluorescence for RSV-infected cells to that for mock-infected cells (fluorescent infected cells/fluorescent uninfected cells). The experiment was performed three times, and representative data are shown. A positive control for DNA fragmentation obtained by treatment of uninfected cells with DNase I (not shown) gave results similar to the RSV-infected culture at 72 h; a negative control using unlabeled RSV-infected cells (not shown) gave results similar to the mock-infected control at 24 h. (C) Fluorescence microscopy showing chromatin condensation (arrow) in A549 cells at 48 h postinfection. Only a minority (<10%) of the RSV-infected culture showed chromatin condensation; similar cells were not observed in the uninfected control. FIG. 3. : Induction and activation of intracellular caspases in response to RSV infection. Induction and activation of intracellular caspases in response to RSV infection. (A) Flow cytometry analysis of activated caspases in RSV-infected A549 cells harvested at the indicated times postinfection. The cells were labeled with FITC-VAD-FMK, and the percentage of positive cells is indicated in each panel. The experiment was performed three times, and representative data are shown. (B) Kinetics of accumulation of mRNA for caspases 8, 10, 3, 6, and 7 in NHBE, SAEC, and A549 and HEp-2 cells (caspases 8 and 10 are initiator caspases; caspases 3, 6, and 7 are effector caspases). RPA was performed, and the level of each caspase mRNA is expressed as a percentage of the GAPDH housekeeping mRNA. Note the differences in the scale of the y axis. (C) Kinetics of enzymatic activity for caspases 3, 6, 8, and 9 in RSV-infected A549 cells. Each experiment was performed twice, and representative data are shown. FIG.4. : RSV induces expression of pro- and antiapoptotic members of the Bcl-2-family. RSV induces expression of pro- and antiapoptotic members of the Bcl-2-family. (A and B) Kinetics of accumulation of mRNA for (A) proapoptotic proteins Bid, Bax, and Bak and (B) antiapoptotic proteins Mcl-1, Bcl-W, and Bcl-xL in NHBE, SAEC, A549, and HEp-2 cells that were mock-infected or infected with RSV or UV-inactivated RSV. RPAs were performed and the level of each Bcl-2 family member mRNA is shown relative to that of GADPH mRNA measured in the same gel lane. (C) Expression of Mcl-1 protein in RSV-infected A549 cells is shown at various times postinfection by intracellular immunostaining and flow cytometry. The median fluorescence +- standard error was derived from three samples in each group. (D) Western blot analysis of Mcl-1 protein in RSV-infected or mock-infected A549 cells harvested 12 or 24 h postinfection; Mcl-1 is 37 kDa . (E) Immunohistochemical staining of Mcl-1 in RSV-infected A549 cells. A549 cells were infected with RSV or mock infected, fixed 24 h postinfection, and stained with anti-Mcl-1 antibody followed by an FITC-conjugated secondary antibody. FIG. 5. : Induction of TRAIL mRNA and protein in response to RSV infection. Induction of TRAIL mRNA and protein in response to RSV infection. (A) Production of TRAIL mRNA in RSV-, UV-RSV-, or mock-infected A549 cells, NHBE, and SAEC. Total intracellular RNA was harvested at the indicated times postinfection, and the amount of TRAIL mRNA measured by an RPA is shown as a percentage of the GAPDH mRNA measured in the same gel lane. (B) Expression of transmembrane TRAIL protein at the surface of RSV- or mock-infected A549 cells measured by flow cytometry at the indicated times postinfection. The percentage of TRAIL-positive cells is expressed as the mean +- standard error based on three samples per group. Results obtained with an isotype control antibody were similar to those obtained with mock-infected control cells (not shown). FIG. 6. : Induction of the DR4 and DR5 TRAIL receptors in RSV-infected NHBE, SAEC, and A549 and HEp-2 cells. Induction of the DR4 and DR5 TRAIL receptors in RSV-infected NHBE, SAEC, and A549 and HEp-2 cells. (A) Kinetics of DR4 and DR5 mRNA accumulation in the four different cell cultures. Intracellular RNA was isolated at the indicated times and the amount of DR4 and DR5 mRNA was determined by an RPA and is represented here as a percentage of the GAPDH mRNA. (B) Expression of DR4 and DR5 protein at the surface of RSV-infected A549 cells. The numbers in each box are the ratio of the median fluorescence for RSV-infected cells to that for mock-infected cells (fluorescence infected cells/fluorescence uninfected cells), based on three samples per group. The isotype control consisted of cells processed with a normal goat IgG and was similar to mock-infected cells (not shown). FIG.7. : Infection with RSV sensitizes cells to TRAIL-induced apoptosis. Infection with RSV sensitizes cells to TRAIL-induced apoptosis. RSV-infected or mock-infected A549 cells were treated for 5 h, beginning 24 h postinfection, with added TRAIL (500 ng/ml) in the presence of TRAIL-specific antibody or were mock treated. The cells were photographed to visualize TRAIL-induced apoptosis (panels at left) and collected and analyzed for apoptosis by annexin V staining (panels at right); the percentage of annexin V-positive-cells (bar) is indicated for each histogram. The experiment was performed four times, and representative data are shown. FIG. 8. : Kinetics of acquisition of sensitivity to TRAIL-induced apoptosis during RSV infection. Kinetics of acquisition of sensitivity to TRAIL-induced apoptosis during RSV infection. RSV-infected A549 cells were infected, and replicate cultures were treated with TRAIL (250 ng/ml) in the presence of TRAIL-specific antibody for 5 h beginning at 6, 11, 16, 19, 22, or 25 h postinfection. A control culture (M) was mock infected and processed in parallel with the 6-h sample. The cells were collected and stained with annexin V, and the percentage of apoptotic cells was determined. Backmatter: PMID- 12915564 TI - Complementation of a Deletion in the Rubella Virus P150 Nonstructural Protein by the Viral Capsid Protein AB - Rubella virus (RUB) replicons with an in-frame deletion of 507 nucleotides between two NotI sites in the P150 nonstructural protein (DeltaNotI) do not replicate (as detected by expression of a reporter gene encoded by the replicon) but can be amplified by wild-type helper virus (Tzeng et al., Virology 289:63-73, 2001). Surprisingly, virus with DeltaNotI was viable, and it was hypothesized that this was due to complementation of the NotI deletion by one of the virion structural protein genes. Introduction of the capsid (C) protein gene into DeltaNotI-containing replicons as an in-frame fusion with a reporter gene or cotransfection with both DeltaNotI replicons and RUB replicon or plasmid constructs containing the C gene resulted in replication of the DeltaNotI replicon, confirming the hypothesis that the C gene was the structural protein gene responsible for complementation and demonstrating that complementation could occur either in cis or in trans. Approximately the 5' one-third of the C gene was necessary for complementation. Mutations that prevented translation of the C protein while minimally disturbing the C gene sequence abrogated complementation, while synonymous codon mutations that changed the C gene sequence without affecting the amino acid sequence at the 5' end of the C gene had no effect on complementation, indicating that the C protein, not the C gene RNA, was the moiety responsible for complementation. Complementation occurred at a basic step in the virus replication cycle, because DeltaNotI replicons failed to accumulate detectable virus-specific RNA. Keywords: Introduction : Rubella virus (RUB) is the sole member of the genus Rubivirus in the family Togaviridae (for a review, see reference ). The RUB virion consists of a genomic, single-stranded RNA enclosed in a quasispherical capsid composed of multiple copies of the viral capsid protein, C, which is in turn surrounded by a lipid bilayer envelope in which are embedded two virus glycoproteins, E1 and E2. The RUB genome is 9,762 uncleotides (nt) in length, of positive-polarity, and contains two long open reading frames (ORFs). The 5'-proximal ORF, or nonstructural protein ORF (NS-ORF), is translated from the genome RNA into a 240-kDa precursor that is proteolytically cleaved at a single site by a virus-encoded protease into two products: an N-terminal product of 150 kDa (P150) and a C-terminal product of 90 kDa (P90). The 3'-proximal ORF, or structural protein ORF (SP-ORF), which is translated from a subgenomic (SG) RNA, encodes the virion proteins in the order 5'-C-E2-E1-3'; processing of these proteins is mediated by the cellular enzyme signal endopeptidase. The RUB NS proteins function in viral RNA replication. From predictions based on computer alignment with sequences from other viruses, P150 contains (from N terminus to C terminus) a methyltransferase domain, a Y domain, a proline hinge domain, an X domain, and a protease domain that catalyzes the cleavage of the NS precursor; P90 contains helicase and RNA-dependent RNA polymerase (RDRP) domains . Of these, the activities of the protease and helicase domains have been confirmed experimentally . While it has been hypothesized that the X domain functions in trans cleavage mediated by the protease , putative functions for the Y and proline hinge domains have not been proposed. Recently, we developed a RUB replicon (RUBrep) in which the SP-ORF was replaced with a reporter gene, such as that coding for chloramphenical acetyltransferase (CAT) or green fluorescent protein (GFP) . While the replicon RNA itself is replicated and the reporter gene is expressed only in the initially transfected cell, in the presence of wild-type helper virus, the replicon is amplified, packaged, and spread to other cells. In this context, the replicon resembles naturally occurring RUB defective-interfering RNAs that maintain the NS-ORF but contain large deletions within the SP-ORF. None of the RUBrep/GFP constructs with deletions in the NS-ORF was capable of replication (as detected by GFP expression) as expected. However, of these deletion mutants, only a construct with an in-frame deletion between two NotI sites at nt 1685 and 2192 of the genome (thus designated DeltaNotI) could be amplified by wild-type helper virus. This deletion encompassed amino acid residues 551 to 720 of P150, a region between the Y and proline hinge domains. This finding indicated that the NS-ORF is required in cis for amplification by wild-type helper virus as well as for self-replication, probably due to a cis preference by the viral replicase proteins for the RNA template from which they were translated . We hypothesized that only the NotI region of the NS proteins, which has no known or suggested function based on computer alignments, could be complemented by wild-type helper virus, but then were surprised to find that when DeltaNotI was introduced into an infectious cDNA clone (Robo402), a viable virus was produced that replicated to within threefold of wild-type virus . This indicated that another viral gene could complement deletions of the NotI region. In this report, we show that the C protein can complement deletions of the NotI region of the P150 NS protein. MATERIALS AND METHODS : Recombinant DNA methods. | Recombinant DNA manipulations were performed basically as described by Sambrook et al. with minor modifications. Escherichia coli JM109 was used as the bacterial host. Restriction enzymes and T4 DNA ligase were obtained from New England BioLabs (Beverly, Mass.) or Roche Molecular Biochemicals (Indianapolis, Ind.) and used essentially as recommended by the manufacturers. Standard PCR mixtures contained 400 ng of each oligonucleotide primer, 20 ng of linearized plasmid template, 200 muM each deoxynucleotide triphosphate, and 5 U of Ex Taq DNA polymerase (PanVera/TaKaRa, Madison, Wis.) in 1x buffer provided by the manufacturer in a total volume of 50 mul. The cycling protocol was 20 s at 98C, 20 s at 50C, and 1 to 3 min at 70C for 35 cycles followed by one cycle of 10 min at 72C. Generation of constructs and site-directed mutagenesis. | The infectious genomic cDNA clones Robo402, Robo402-DeltaNotI, and Robo402/IRES (internal ribosome entry site construct), all of which use pBR322 as a backbone plasmid, were described previously . The SP6 RNA polymerase promoter-cDNA insert from each of these clones was transferred to pUC18 to create Robo502, Robo502-DeltaNotI, and Robo502/IRES, respectively. An additional genomic cDNA construct, Robo503, was generated, which could be linearized prior to in vitro transcription with either EcoRI or SpeI (this was necessary for constructs containing the CAT gene, which contains an internal EcoRI site). The replicon constructs, RUBrep/GFP and RUBrep/GFP-DeltaNotI, were described previously . For use in this study, the SP6 RNA polymerase promoter-cDNA insert in each of these constructs was transferred to pUC18. To create the mutated constructs used in this study, both PCR amplification with mutagenic oligonucleotide primer or primers followed by restriction fragment replacement with the amplification product into the parent vector and a three-round asymmetric PCR-three-fragment ligation strategy were employed. Constructs with fusions of the C or E2 genes with the GFP gene or C, E2, or E1 with 3'-terminal deletions fused with GFP were initiated with mutagenic oligonucleotides that placed an XbaI site following the 3' end of the gene or the deletion within the gene that allowed in-frame fusion with an XbaI site at the 5' end of the GFP cassette. Constructs that contained progressive deletions of the C gene from the 5' end were initiated with mutagenic oligonucleotides that placed an XbaI or SpeI site and an ATG initiation codon upstream from the nucleotides following the deletion. RUBrep constructs expressing a fusion protein between the complete C gene of Sindbis virus (SIN) or the SIN C gene with 3' deletions and GFP were initiated with an upstream oligonucleotide containing EcoRI and SpeI sites and the 5' 21 nt of the SIN C gene and downstream oligonucleotides that contained an XbaI site followed by 15 to 16 nt complementary to the sequence of the SIN C gene upstream from the deletion site, allowing for introduction into SpeI-XbaI-digested RUBrep/GFP; XhoI-linearized pTE5'2J was used as the template. A series of mutations in the 5' nine codons of the C gene were initiated with mutagenic oligonucleotides containing the desired mutation or mutations with 8 similar12 nt on either side of the mutation. Mutations created in RUBrep/GFP were transferred to RUBrep/CAT . To generate replicons containing the NotI deletion, the HindIII-BglII fragment from RUBrep/GFP-DeltaNotI (containing the SP6 RNA polymerase promoter and the 5' end of the RUB genome through nt 5355, a region containing the NotI deletion) was used to replace the corresponding fragment in the RUBrep/GFP or RUBrep/CAT construct. Plasmids expressing cassettes containing various regions of the RUB NS-ORF and SP-ORF as well as the SIN SP-ORF under control of the human cytomegalovirus (CMV) immediate-early promoter were created with VR1012 vector (Vical, Inc., San Diego, Calif.). Each cassette was amplified by PCR with an upstream oligonucleotide that contained a restriction site appropriate for introduction into the multiple cloning site of VR1012, an ATG for initiation of translation if necessary, and the 5' 15 nt from the region to be expressed. The downstream oligonucleotide contained an appropriate restriction site, the complement of a termination codon if necessary, and nucleotides complementary to the 3' 15 nt from the region to be expressed. In vitro transcription and transfection. | Robo, RUBrep, and VR1012 constructs were purified on CsCl isopycnic density gradients prior to use. Robo502 and its derivatives were linearized with EcoRI, while Robo503 and its derivatives were linearized with SpeI prior to in vitro transcription, which was carried out as previously described . The transcription reaction mixtures were used directly for cell transfection without DNase treatment or phenol-chloroform extraction. Vero and BHK-21 cells were transfected with Lipofectamine 2000 as previously described . GFP expression was monitored by direct examination of the transfected monolayer with a Zeiss Axioplan microscope with epifluorescence capability. For CAT activity assay, Vero cells were transfected with transcripts from replicon-CAT-DeltaNotI constructs. At 4 days posttransfection, cells were lysed and used for CAT activity assay basically as described by Seed and Sheen . Replicon-specific RNAs species synthesized in transfected cells were analyzed by Northern assay with a NorthernMax-Gly kit (Ambion, Houston, Tex.) as previously described with 32P-labeled, nick-translated pGEM/GFP as a probe . RESULTS : C gene complements NotI deletions in RUB replicons in cis. | We showed previously that transcripts from a replicon construct expressing the GFP gene, RUBrep/GFP, produced GFP following transfection of Vero cells, while transcripts from a construct with the NotI region of the NS-ORF deleted (RUBrep/GFP-DeltaNotI) did not . However, in the presence of coinfecting wild-type helper RUB or cotransfection with transcripts from the infectious cDNA clone, Robo402, the GFP signal was amplified for both constructs. Unexpectedly, transcripts from a Robo402 construct containing the NotI deletion, Robo402-DeltaNotI, yielded viable virus that replicated to within threefold of Robo402 virus. The only difference between Robo402-DeltaNotI and RUBrep/GFP-DeltaNotI is the SP-ORF. Therefore, we hypothesized that a structural gene or protein could complement the NotI region of the NS-ORF, allowing both replication of Robo402-DeltaNotI virus and amplification of RUBrep/GFP-DeltaNotI in the presence of wild-type helper virus. To this end, a series of replicons were constructed that contained in-frame fusions of the individual structural proteins with GFP. As shown in Fig. , while RUBrep/GFP-DeltaNotI did not express GFP in transfected cells, RUBrep/C-E2-E1(1-9)-GFP-DeltaNotI (which contains the SP-ORF through the first nine amino acids of the E1 gene fused to the GFP gene) and RUBrep/C-E2-GFP-DeltaNotI (which contains the C and E2 genes fused with GFP gene) produced GFP, indicating that the structural proteins can complement the NotI deletion in the NS-ORF. Further analysis showed that RUBrep/C-E2(1-7)-GFP-DeltaNotI, which contains the C gene through the first seven amino acids of the E2 gene fused to the GFP gene, and RUBrep/C-GFP-DeltaNotI, which contains the C gene fused with GFP gene, did express GFP in transfected cells, while RUBrep/E2-GFP-DeltaNotI, which contains the E2 gene fused with GFP gene, did not. When introduced into the wild-type replicon without the NotI deletion, RUBrep, all of these constructs expressed GFP. Thus, the moiety responsible for complementation of the NotI deletion was the C gene. Next, to examine whether the complete C gene or only part would suffice to complement the NotI deletion, a series of C-GFP fusions that contained progressive deletions of the C gene from either the 5' or 3' end, or both, were generated. As shown in Table , when present in RUBrep-DeltaNotI, none of the 5' deletions, including one that only deleted 8 codons, expressed GFP, and only the smallest 3' deletion, which deleted 23 codons (i.e., the E2 signal sequence at the C terminus of C) expressed GFP. In RUBrep, all of these constructs expressed GFP. Thus, the majority of the C gene was required for complementation. We also made a series of constructs expressing a fusion protein between the C gene of the alphavirus SIN, or 3'-truncated fragments of the SIN C gene, and GFP in both RUBrep and RUBrep-DeltaNotI. As shown in Table , all of these RUBrep constructs expressed GFP, while none of these RUBrep-DeltaNotI constructs expressed GFP, and thus the SIN C protein cannot substitute for the RUB C protein in complementing the NotI deletion. To test expression with a different reporter gene, the CAT gene was employed. As shown in Table , similar to the GFP constructs, RUBrep/CAT-DeltaNotI did not express CAT; however, RUBrep/C-CAT-DeltaNotI did. In contrast to the fusions of progressive 3' deletions of C to GFP, CAT fusions containing the N-terminal 88 codons of the C gene expressed CAT. This finding was likely related to the different virus production efficiencies following transfection of the versions of the infectious clone used to generate the CAT and GFP constructs. RUBrep/GFP and RUBrep/CAT were built by using Robo502 and Robo503, respectively. Robo503 has a SpeI 3' linearization in place of the EcoRI linearization site in Robo502; this substitution was necessary because the CAT gene contains an internal EcoRI site. Unexpectedly, while the transfection efficiencies of the Robo502 and Robo503 transcripts were similar, virus production was more rapid in Robo503-transfected cells than in Robo502-transfected cells such that by 4 days posttransfection, the difference in virus produced was 100-fold greater (data not shown). In the replicon context, this would lead to greater production of the C protein-reporter gene fusion proteins. When the RUBrep/C-CAT-DeltaNotI deletion series was tested in BHK cells, similar results were obtained as in Vero cells: namely that RUBrep/CAT-DeltaNotI or RUBrep/C-CAT-DeltaNotI fusions containing the 5' 58 codons did not express CAT, while RUBrep/C-CAT fusions containing 88 or more codons of the C gene expressed GFP (data not shown). Thus, the complementation phenomenon was not cell specific. Thus far in this study, replication of RUBrep-DeltaNotI constructs was assayed by reporter gene expression. To determine if complementation by the C gene was actually operating on SG RNA synthesis and/or translation of the reporter gene from the SG RNA, viral RNA synthesis in cells transfected with transcripts from the series of RUBrep/GFP-DeltaNotI constructs was assayed by Northern blotting of intracellular RNA extracted from transfected cells. As shown in Fig. , in cells transfected with constructs that did not express GFP (i.e., RUBrep/GFP-DeltaNotI, RUBrep/E2-GFP-DeltaNotI, and RUBrep/C-GFP-DeltaNotI containing less than the 277 codons of the C gene), no viral RNA was detected, while in cells transfected with constructs that expressed GFP, RUBrep/C(1-277)-GFP-DeltaNotI, RUBrep/C-GFP-DeltaNotI, RUBrep/C-E2(1-9)-GFP-DeltaNotI, RUBrep/C-E2-GFP-DeltaNotI, and RUBrep/C-E2-E1(1-7)-GFP-DeltaNotI, both genomic and SG RNAs were readily detectable. RUBrep containing all of these constructs synthesized both genomic and SG RNAs (data not shown). This result revealed that the defect in DeltaNotI mutants and complementation by the C gene was at a basic point in the virus replication cycle, prior to accumulation of detectable virus RNA, and was not at the level of SG RNA synthesis and/or reporter gene translation. In subsequent experiments, complementation was assayed both by reporter gene expression and synthesis of viral RNA. The C gene complements the NotI deletions in trans. | Since amplification of replicons by wild-type helper virus occurs in trans, we next conducted experiments to determine if the C gene expressed in trans could complement RUBrep/GFP-DeltaNotI. These experiments were performed by cotransfecting cells with RUBrep/GFP-DeltaNotI transcripts and transcripts from RUBrep/CAT constructs; complementation was detected by expression of GFP from RUBrep/GFP-DeltaNotI. While RUBrep/CAT was unable to complement RUBrep/GFP-DeltaNotI, demonstrating that the NS proteins cannot complement the NotI deletion, RUBrep/C-CAT fusion constructs containing minimally the 5' 31 codons as well as RUBrep/C-E2-CAT exhibited complementation (Fig. and Table ). This result was consistent with the ability of these C-CAT constructs to complement in cis , with the exception that only 31 codons of the C gene was necessary for trans complementation, while 88 codons was necessary for cis complementation. None of the C-CAT fusion constructs that contained 5' deletions of the C gene, including the shortest, which deleted only eight codons, was able to complement RUBrep/GFP-DeltaNotI. Additionally, none of the RUBrep/SIN-C-CAT constructs complemented in trans . As an independent system with which to analyze trans complementation, cells were cotransfected with RUBrep/GFP-DeltaNotI, and plasmid vectors expressing regions of the RUB SP-ORF under control of the CMV immediate-early promoter; as described above, complementation was detected by expression of GFP by RUBrep/GFP-DeltaNotI. CMV vectors expressing the complete SP-ORF, the C gene, or minimally the 5' 88 codons of the C gene exhibited complementation, while vectors expressing E2 and E1, the 5' 58 codons or less of the C gene, or the SIN SP-ORF did not complement. Using the CMV vector that expressed the shortest 5' region of the C gene able to complement, CMV RUB-C(1-88), a series of progressive 5' deletions of the C gene were made; however, none of these CMV constructs, including the shortest 5' deletion of eight codons, was able to complement RUBrep/GFP-DeltaNotI. These results were consistent with trans complementation using RUBrep/CAT constructs, with the exception that only the 5' 31 codons of the C gene was necessary for complementation by RUBrep/CAT, while the 5' 58 codons were necessary for complementation by the CMV vector. The CMV vector was also used to analyze whether the NS-ORF region of the RUB could complement the NotI deletion. As shown in Table , CMV vectors expressing the entire RUB NS-ORF, the P150 gene, the P90 gene, or the NotI region were used to cotransfect cells with RUBrep/GFP-DeltaNotI transcript; no GFP expression was observed. Determination of whether the C protein or the C gene RNA is the complementing factor. | Since the defect in DeltaNotI replicons is early in the replication cycle, before the accumulation of detectable virus RNA, we initially hypothesized that the C gene RNA was the moiety responsible for complementation. The C gene RNA could be provided by either the genomic or SG RNA, and therefore complementation by a RUB construct in which the "junction" untranslated region (UTR) between the NS- and SP-ORFs was replaced with the encephalomyocarditis virus IRES (Robo502/IRES) was tested; Robo502/IRES does not synthesize SG RNA . When the NotI deletion was made in Robo502/IRES, the resulting construct, Robo502/IRES-DeltaNotI, yielded viable virus, indicating that the SG RNA was not necessary for complementation . Robo502/IRES-DeltaNotI virus replicated about threefold less well than did Robo502/IRES virus, the same as the differential between Robo502 and Robo502-DeltaNotI virus. When cells were cotransfected with RUBrep/GFP-DeltaNotI transcripts and transcripts from either Robo502/IRES or Robo502/IRES-DeltaNotI, complementation occurred, indicating that SG RNA was not necessary for complementation in trans. In addition to the lack of an SG RNA, the IRES constructs also lack the 5' UTR of the SG RNA, and thus this RNA sequence is not required for complementation. We next focused on the 5' nine codons of the C gene that are necessary for complementation by using the C-GFP construct that contains the complete C gene. As shown in Fig. , the sequence at the 5' end of the C gene contains two in-frame AUGs (designated AUG1 and AUG2) separated by seven codons. A series of mutations of these sequences, shown in Fig. , were made in the C-GFP fusion protein cassette, in which the complete C gene is fused with GFP, and the mutated cassettes were introduced into (i) RUBrep/GFP to ascertain replication of the wild-type replicon containing the cassette (Northern blotting) and to determine if translation of the fusion protein occurred (GFP expression), (ii) RUBrep/GFP-DeltaNotI to assay for cis complementation (both Northern gels and GFP expression), and (iii) RUBrep/CAT to assay for trans complementation (GFP expression in cells cotransfected with RUBrep/GFP-DeltaNotI). It was found that RUBrep/GFP containing all of these mutations replicated, and thus none adversely affected viability. When AUG1 or both AUG1 and AUG2 were mutated to AUA, neither translation of the fusion protein nor cis or trans complementation was detected; however, when AUG2 was mutated to AUA, translation and complementation occurred. These results indicate that translation of the C gene cannot initiate at AUG2. To use a different mutation to abrogate initiation of translation at the first AUG of the C gene, AUG1 was replaced with UAAUAA with the same result that translation of the fusion protein and complementation were not detected. However, when the UAAUAA was placed immediately upstream of AUG1, translation and complementation occurred. Next, 1- and 2-nt deletions were made in codon 6 of the C gene to maintain AUG1 in its wild-type context, including initiation of translation, but prevent normal translation of the C-GFP fusion protein. As expected, GFP expression was not detected in RUBrep/GFP, and neither mutant was capable of complementation. These results indicate that production of the C protein is necessary for complementation. Finally, to dissect the presence of C protein production from the effect of the RNA sequence at the 5' end of the C gene on complementation, silent mutations were made in each of the seven codons between AUG1 and AUG2; these mutations would preserve the amino acid sequence of the N terminus of the C protein but perturb the RNA sequence. None of the individual mutations had an effect on complementation, and when all of these mutations were combined in one mutant construct, complementation was still detected, providing further evidence that the C protein, rather than its encoding RNA, was the moiety responsible for complementation of the NotI deletion in P150. DISCUSSION : The goal of this study was to investigate whether a structural gene or protein of RUB could complement deletions of the NotI region of the NS-ORF that rendered replicons nonviable. The first indication of this phenomenon was a result in a previous publication , in which we found that while the NotI deletion was lethal in replicons, virus with the same deletion was viable. Since the only difference between virus and replicons was the SP-ORF, we initiated this study with the hypothesis that one of the structural genes or proteins complemented the NotI gene or protein. By constructing DeltaNotI replicons that expressed via the SG RNA in-frame fusions of the individual structural proteins with GFP, we were able to determine that the moiety responsible for complementation of the NotI deletion was the C gene. A series of RUBrep/C-GFP fusions that contained progressive deletions of the C gene from either the 5' or 3' end, or both, showed that codons 1 to 277 of the C gene (in other words, the C gene lacking the E2 signal sequence) were required for replication. Similar results were obtained in both Vero and BHK cells, and thus complementation was not cell specific. We also found that the C gene of the related alphavirus SIN could not complement the NotI deletion, and thus complementation was specific to the RUB C gene. While initial results were based on GFP expression, Northern analysis revealed that in cells transfected with DeltaNotI replicon constructs that failed to express GFP, replicon-specific RNA was not detectable, indicating that complementation occurred at a basic step in the replication cycle, prior to accumulation of detectable replicon RNA, and not at the level of SG RNA synthesis and/or reporter gene translation. As expected, since wild-type RUB was able to amplify replicons bearing NotI deletions , we also found that the C gene could complement RUBrep/GFP-DeltaNotI in trans when expressed either from another replicon or from a plasmid, resulting in replication of RUBrep/GFP-DeltaNotI. The complete C gene was not required for complementation in trans, since the 5' 31 codons of the C gene in RUBrep replicons or the 5' 58 codons of the C gene in a CMV-driven plasmid were sufficient for complementation. To some extent, if not completely, the difference between the amount of the C gene required for complementation in cis versus that required in trans was due to differences in efficiencies of virus production following transfection of the parent infectious clones. The RUBrep/C-GFP-DeltaNotI constructs used to assay cis complementation were based on Robo502, while the RUBrep/C-CAT constructs used to assay trans complementation were based on Robo503. Robo502 and Robo503 vary in the sequence of the additional nucleotides added to the transcript following the oligo(A) (A20) tract due to residual sequences in the restriction site used for linearization (EcoRI and SpeI, respectively): namely GAAUU in Robo502 transcripts and CUAG in Robo503 transcripts. While Robo502 and Robo503 transcripts have the same transfection efficiency, Robo503 transcripts replicate more rapidly than do Robo502 transcripts, which would hypothetically result in a more rapid and higher-level production of C protein. As predicted if this hypothesis were correct, when C-GFP fusions with 3' deletions were introduced into RUBrep/CAT-DeltaNotI, the construct with only the 5' 88 codons of the C gene replicated (in contrast to the 277 codons required for replication of RUBrep/C-GFP-DeltaNotI constructs). Significantly, none of the constructs with 5' deletions of the C gene of any size were able to complement either in cis or in trans from any of the vectors employed. While complementation is traditionally associated with proteins, RNA sequences or structures that function at long range in cis or function both in cis and in trans have been described for diverse RNA viruses . Because the C genes in both Robo502/IRES and CMV-driven plasmids (neither of which synthesizes the RUB SG RNA nor contains the 5' UTR of the SG RNA) were found to be capable of complementing the NotI deletion, the SG RNA is not involved in complementation, and the complementation moiety resides within the coding sequences for the C protein. Concentrating on the 5' nine codons of the C gene, which deletion analysis indicated were essential for complementation, we found that mutations that abrogated translation of the C gene but only changed 1 nt within these codons also eliminated complementation, while mutations that changed the RNA sequence of these codons without changing the encoded amino acid sequence had no effect on complementation. Thus, the C protein and not the C gene RNA is the moiety responsible for complementation. Interestingly, in an alignment between the C proteins of RUB and SIN , the N terminus of the SIN protein is aligned with the second Met residue of the RUB protein, and the N-terminal eight residues of the RUB protein, which are essential for complementation, are not included in the alignment. (Overall, in this alignment, the two C proteins share 17% identity and 26% similarity.) The mechanism by which the C protein complements the NotI region of the P150 replicase protein was not investigated in this study; however, complementation occurred at an early step in the replication cycle, since RUBrep/GFP-DeltaNotI failed to accumulate detectable replicon-specific RNA. Since complementation occurred in cis, initial replication of input replicon transcripts would theoretically have to occur to allow for synthesis of the C protein (none was introduced by infecting virions, since this study was done entirely with transcripts), although translation of C protein from partially degraded transcripts cannot be completely ruled out. If initial replication is required to produce C protein, then complementation operates in the process of amplification, rather than initiation, of virus RNA synthesis. Interestingly, it has been reported that the C protein colocalized with P150 on tubular structures in RUB-infected cells late in infection . The C protein contains a motif between residues 28 and 56 that binds the RUB genomic RNA , and RNA binding by the C protein has been shown to be regulated by a phosphorylation site at Ser-46 . Whether these activities assist in RNA replication as well as playing a role in their obvious function of encapsidation is not known; it is conceivable that C protein binding of the genome RNA is a factor in efficient release from the replication complex. However, since only the N-terminal 31 residues of the C gene were needed for complementation in trans by RUBrep/C-CAT, these activities may not be involved in complementation of the NotI deletion. Besides a specific step in the viral replication cycle, C protein complementation of the NotI region of P150 may involve binding to cellular factors. The C protein has been shown to interact with two mitochondrial proteins, p32 and Par-4, and the p32-binding domain within the C protein has been mapped to the N-terminal region of the protein . While the function and utility of these interactions in RUB replication have not been elucidated, p32 belongs to the family of cellular defense collagens . The C protein has also been associated with induction of apoptosis in RK13 cells, a cell line exquisitely sensitive to RUB-induced cytopathic effect . Finally, it is possible that C protein complementation of the NotI domain of the P150 protein is mediated through some general activity rather than direct compensation for NotI domain function: for example, binding to transcripts and stabilizing them or efficiently targeting them. This would be a function that the C protein in virions possibly plays during virus infection, and under this scenario, replicons with NotI deletion would simply replicate too inefficiently to establish themselves before degradation occurred or some cellular defense mechanism was fully induced. In this regard, the function of the NotI domain is not known (since it has no detectable homology with replicase proteins of other viruses), and it is notable that when aligned with the NotI domain, the N terminus of the C protein shares little homology (22% similarity and 16% identity) with this domain , and we found that neither P150 nor the NotI domain could complement DeltaNotI replicons when expressed in trans. FIG. 1. : Mapping the moiety responsible for complementation. Mapping the moiety responsible for complementation. In panel A and Table are shown the results of experiments with cassettes expressed by RUBrep or RUBrep-DeltaNotI. The cassettes consisted of either a reporter gene (GFP or CAT) or a portion of the RUB or SIN SP-ORF fused in frame with GFP or CAT. The GFP version of each cassette was used to assess replication of the wild-type replicon, RUBrep (replication column), and of the replicon with the NotI deletion, RUBrep-DeltaNotI (cis complementation column), containing the cassette. Replication or complementation was detected by both GFP expression using fluorescence microscopy of transfected cells (GFP) and virus-specific RNA production using Northern blot analysis (B [RNA]). trans complementation was assessed by cotransfecting cells with transcripts of the RUBrep construct containing the CAT version of the cassette and RUBrep/GFP-DeltaNotI transcripts; replication of RUBrep/GFP-DeltaNotI was detected by GFP expression. Panel A summarizes experiments with fusion proteins containing portions of the RUB SP-ORF. A schematic diagram of the ORF with the coding sequences for the C, E2, and E1 proteins as well as the E2 and E1 signal sequences (which remain attached to C and E2, respectively, following processing) is shown at the top of the panel and the portion of the ORF contained in each cassette, along with the reporter gene, is shown under this diagram. The C-E2-E1(1-9)-GFP/CAT cassette contains the complete C and E2 genes and the first nine amino acids of the E1 gene fused to the reporter gene, and the C-E2(1-7)-GFP/CAT cassette contains the complete C gene and the first seven amino acids of the E2 gene fused to the reporter gene. For the Northern blot shown in panel B, Vero cells were transfected with transcripts from RUBrep/GFP-DeltaNotI (lane 1), RUBrep-DeltaNotI constructs expressing a series of C-GFP-DeltaNotI fusions that contained progressive 3'-terminal deletions of the C gene (lanes 2 to 11), RUBrep/C-GFP-DeltaNotI (lane 12), RUBrep/C-E2-GFP-DeltaNotI (lane 13), RUBrep/E2-GFP-DeltaNotI (lane 14), RUBrep/C-E2-E1(1-9)-GFP-DeltaNotI (lane 15), or RUBrep/C-E2(1-7)-GFP-DeltaNotI (lane 16). Four days posttransfection, total RNA was extracted and analyzed by gel electrophoresis, blotting, and probing with 32P-labeled pGEM-GFP DNA. The positions of migration of the replicon genomic (G) RNAs and SG RNAs (which vary in size) are denoted. FIG. 2. : Viability and trans complementation by infectious cDNA clone constructs. Viability and trans complementation by infectious cDNA clone constructs. Shown are genomic diagrams of Robo502 and Robo502/IRES, in which the junction-UTR between the NS- and SP-ORFs was replaced by the IRES of encephalomycarditis virus, without and with the NotI deletion in the P150 gene. As shown, all of these constructs give rise to viable virus, as shown by the titers and plaque morphologies of virus in transfection culture fluid harvested 4 days posttransfection. To test for the ability to complement the NotI deletion in trans, Vero cells were cotransfected with transcripts from one of these constructs and RUBrep/GFP-DeltaNotI transcripts, and replication of RUBrep/GFP-DeltaNotI was detected by GFP expression. FIG. 3. : Effect on complementation of mutations at the 5' end of the C gene. Effect on complementation of mutations at the 5' end of the C gene. Using the C-GFP/CAT cassette, a series of mutations at the 5' end of the C gene were made. (The 5' end of the C gene has two in-frame AUGs separated by seven codons). Under the wild-type sequence, each mutated sequence is given with the mutation in boldface and designated with an asterisk. Each mutated cassette was assessed for replication in wild-type replicon RUBrep/GFP, cis complementation in RUBrep/GFP-DeltaNotI, and trans complementation in RUBrep/CAT (in cells cotransfected with RUBrep/GFP-DeltaNotI transcripts) as described in the legend to Fig. and shown in Table . FIG. 4. : Alignment of the sequences of RUB C protein with those of the SIN C protein and NotI domain. Alignment of the sequences of RUB C protein with those of the SIN C protein and NotI domain. The entire RUB and SIN C protein sequences are included in the alignment; amino acids are numbered from the N-terminal Met residue. In the RUB C X NotI alignment, only the N-terminal 100 residues of the C protein are used. The numbering of the NotI domain represents the residue number within the P150 protein. The alignment was made with the MacVector version 6.5.3 program. TABLE 1 : Experiments with fusion proteins containing terminally truncated fragments of the RUB and SIN C genes TABLE 2 : GFP and CAT activity in Vero cells TABLE 3 : Detection of trans complementation with a plasmid-CMV vector Backmatter: PMID- 12915539 TI - Effects of Modification of the Transcription Initiation Site Context on Citrus Tristeza Virus Subgenomic RNA Synthesis AB - Citrus tristeza virus (CTV), a member of the Closteroviridae, has a positive-sense RNA genome of about 20 kb organized into 12 open reading frames (ORFs). The last 10 ORFs are expressed through 3'-coterminal subgenomic RNAs (sgRNAs) regulated in both amounts and timing. Additionally, relatively large amounts of complementary sgRNAs are produced. We have been unable to determine whether these sgRNAs are produced by internal promotion from the full-length template minus strand or by transcription from the minus-stranded sgRNAs. Understanding the regulation of 10 sgRNAs is a conceptual challenge. In analyzing commonalities of a replicase complex in producing so many sgRNAs, we examined initiating nucleotides of the sgRNAs. We mapped the 5' termini of intermediate- (CP and p13) and low- (p18) produced sgRNAs that, like the two highly abundant sgRNAs (p20 and p23) previously mapped, all initiate with an adenylate. We then examined modifications of the initiation site, which has been shown to be useful in defining mechanisms of sgRNA synthesis. Surprisingly, mutation of the initiating nucleotide of the CTV sgRNAs did not prevent sgRNA accumulation. Based on our results, the CTV replication complex appears to initiate sgRNA synthesis with purines, preferably with adenylates, and is able to initiate synthesis using a nucleotide a few positions 5' or 3' of the native initiation nucleotide. Furthermore, the context of the initiation site appears to be a regulatory mechanism for levels of sgRNA production. These data do not support either of the established mechanisms for synthesis of sgRNAs, suggesting that CTV sgRNA production utilizes a different mechanism. Keywords: Introduction : Positive-stranded RNA viruses are near the ultimate in genetic compactness, often with individual nucleotides serving multiple functions in their life cycles. The virion RNA itself must serve as the mRNA for its early expressed genes and as the template for production of progeny, in addition to having numerous cis-acting elements for replication, regulation, and assembly. The expression of multiple genes usually occurs by a combination of strategies: production of subgenomic RNAs (sgRNAs); processing of polyproteins; use of multipartite genomes, frameshifts, and readthroughs; or alternate initiation of open reading frames (ORFs). sgRNAs may represent a more advanced level of regulation, allowing control of amounts and timing of individual gene products. Most RNA viruses are relatively small and produce few sgRNAs, but as viruses increase in size the complexity of gene expression increases. Some of the larger RNA viruses have as many as 11 sgRNAs. Citrus tristeza virus (CTV), a member of the Closteroviridae, is the largest plant RNA virus; it has a positive-sense single-strand RNA genome of ~20 kb organized into 12 ORFs , with the last 10 3' ORFs expressed through a nested set of 3'-coterminal sgRNAs . The Closteroviridae appears to be intermediate between the small alpha-like viruses and those of the large Nidovirales. The large sizes of the replicase-associated genes and interdomain areas, in addition to the large number of sgRNAs, suggest similarity to the Nidovirales. However, the lack of discontinuous mRNAs with a common leader , characteristic of the Arteriviruses and Coronaviruses of the Nidovirales , suggest greater similarity to the alpha-like viruses, although most of the sgRNAs of the toroviruses and okaviruses of the Nidovirales are contiguous with the 3' end of the genome like CTV . RNA-dependent RNA polymerase (RdRp) phylogeny, however, clearly places the closteroviruses into the alphavirus-like supergroup . Synthesis of the different CTV sgRNAs is regulated in both amounts and timing . Levels of sgRNAs are based on promoter strength and position within the genome . An unusual characteristic of CTV is that relatively large amounts of 3'-terminal negative-stranded sgRNAs accumulate along with 5'-terminal positive-stranded sgRNAs. The 5'-terminal sgRNAs apparently are produced by termination near the 3' controller elements during genomic RNA synthesis . The amount of production of the minus-stranded sgRNAs is controlled by the p23 gene product . We have characterized two different sgRNA controller elements to a region approximately 50 nucleotides upstream of the controlled ORF, but we have been unable to definitively determine the mechanism used by CTV to synthesize sgRNAs. Two general strategies of production of sgRNAs have been proposed for positive-stranded RNA viruses (reviewed in reference ). Most members of the alphavirus-like supergroup, and some other plant viruses, produce sgRNAs by promotion. Specific promoter elements on the genomic negative strand allow precise replicase complex recognition for initiation of sgRNA synthesis at internal sites, followed by continued synthesis to the terminus of the template . A second model proposed for the Nidovirales (Arteriviridae and Coronaviridae), the nodaviruses, and some plant viruses postulates synthesis of sgRNA minus strands by termination, followed by transcription of sgRNA plus strands from the negative templates (, , , , ; reviewed in reference ). Grouping viruses by these different transcription strategies results in groups similar to those derived from RdRp phylogeny. With either strategy, the replicase complex must precisely interact with the negative-stranded template, either internally at the full-length minus strand for promotion or at the 3' end of the sgRNA minus strand for transcription of sgRNA plus strands. By convention, the +1 nucleotide is defined as the template nucleotide used to initiate sgRNA synthesis. In general, alpha-like viruses are expected to follow the +1 pyrimidine and +2 adenylate rule (in reference to the negative strand of the genome) as initiation nucleotides for the genomic positive strand and sgRNA , and usually within each virus the same first nucleotide is found for both types of RNAs . Most of the small RNA viruses of plants have guanylates as 5' termini of the genomic and sgRNAs . Adenylate, however, is a common 5' terminus of RNAs for animal alpha-like viruses and members of the Nidovirales but infrequent for plant viruses . Uridylate and cytidylate as 5' termini are uncommon but have been also reported for rubella virus and Oat chlorotic stunt virus sgRNAs, respectively . Mutagenesis of the initiation nucleotide is a useful tool to examine sgRNA synthesis. A characteristic observed of viruses that produce their sgRNAs by promotion is that mutagenesis of the +1 nucleotide prevents or greatly decreases sgRNA synthesis in vivo and in vitro . This observation has allowed confirmation of mapping of the 5' termini of sgRNAs . However, in viruses that produce their sgRNAs via a termination mechanism, modification of this nucleotide can inhibit accumulation of the sgRNA plus strand without affecting minus-strand synthesis . This observation can be useful in defining whether the positive- or negative-stranded sgRNA is produced first, and by which mechanism. Members of the Closteroviridae produce up to 11 sgRNAs. When a viral replicase complex must interact with numerous internal cis-acting elements on the genomic RNA minus strand (promotion mechanism) or with numerous promoters on the 3' end of the minus-stranded sgRNA (termination mechanism) to initiate sgRNA synthesis, what common traits should be expected? Would the sgRNAs be expected to have the same 5'-terminal nucleotide? Would the 5' termini of genomic and sgRNAs tend to be the same? How much flexibility is there in initiation of synthesis of the sgRNA plus strands? The 5' termini of sgRNAs of other members of the Closteroviridae have been examined. An adenylate is the 5' terminus of several sgRNAs of the closteroviruses, Beet yellows virus (BYV) and Beet yellow stunt virus (BYSV) . The exception is the BYV p6 sgRNA that contains a guanylate similar to the 5' terminus of the genomic RNA . In contrast, the 5' termini of the sgRNAs of the crinivirus Sweet potato chlorotic stunt virus were reported to be adenylate, guanylate, or even uridylate, whereas the 5' ends of both genomic RNAs (RNA1 and RNA2) were guanylates. The 5' termini of the CTV genomic RNA and the most highly produced sgRNAs, p20 and p23, are adenylates . In contrast to that of other related viruses , the CTV CP core controller element mapped to positions -47 to -5 and did not contain the +1 nucleotide corresponding to the 5' terminus of the sgRNA . Several questions arise. If the +1 nucleotide is essential for the synthesis of the positive strand of the sgRNAs in viruses that produce sgRNAs either by internal promotion on the full-length minus strand or by initiation at the 3' end of the sgRNA minus strand produced by termination, how does initiation occur without a conserved +1 nucleotide as a specific component of the controller element? How much flexibility is there for the nucleotides surrounding the +1 site? What is the effect of modifications of these nucleotides on sgRNA production in CTV? In this work, we examined the 5' termini of the CTV sgRNAs of intermediate (CP and p13) and low (p18) abundance and found that they were all adenylates. We then mutated the nucleotides corresponding to the 5' termini of these sgRNAs and found that, instead of preventing sgRNA synthesis and confirming our mapping, accumulation of sgRNAs continued, generally with little reduction. Mapping the initiation sites of some of these +1 nucleotide mutants showed that the CTV replicase complex can use an alternate purine (pyrimidine in the negative strand), preferably an adenylate, as the initiating nucleotide for sgRNA synthesis. Further mutagenesis of the nucleotides surrounding the initiation sites of the sgRNAs suggested that initiation of CTV sgRNA synthesis can occur within a range of several nucleotides around the original +1 site and that the context of the initiation site can modulate the efficiency of sgRNA synthesis. MATERIALS AND METHODS : Determination of the 5' termini of CTV sgRNAs. | The nucleotide numbering and sequences in this study are according to Satyanarayana et al. (T. Satyanarayana, S. Gowda, M. A. Ayllon, and W. O. Dawson, submitted for publication) (GenBank accession no. ). All sequences are presented as a positive-stranded RNA. Total RNA was isolated from bark tissue of young sweet orange [Citrus sinensis (L.) Osb.] infected with the CTV T36 isolate, using the guanidine-HCl procedure with modifications previously described , and was used as a template to map the 5' termini for CP, p18, and p13 wild-type sgRNAs. Similarly, total RNA extracted from Nicotiana benthamiana mesophyll protoplasts infected with RNA transcripts of the CP +1 A/C mutant was used as a template to determine the sgRNA 5' terminus of this mutant. Four negative-sense primers, PM37 (5'-GTTTACGGAACTGAAAGAAG-3'), C502 (5'-GGAACTGAAAGAAGACTCGGCAGCA-3'), C244 (5'-CGATTATATCACCCACGTTCA-3'), and PM36 (5'-CTTAAAGTCATCATTCATGAG-3'), complementary to positions 16247 to 16228, 16241 to 16217, 16894 to 16874, and 17439 to 17419 of the T36 genome, were used for primer extension analysis of CP, CP +1 A/C mutant, p18, and p13 sgRNAs, respectively, as described by Karasev et al. . Briefly, 0.3 pmol of primer (~106 dpm), 5' labeled with 150 muCi of [gamma-32P]ATP (New England Nuclear; 6,000 Ci/mmol) using polynucleotide kinase (Promega), was annealed to total RNA extracted from CTV T36-infected plants or mesophyll protoplasts inoculated with RNA transcripts of the CP +1 A/C mutant. Primer extension reactions were performed with avian myeloblastosis virus reverse transcriptase (U.S. Biochemicals) at 42C for 1 h. Extension products were analyzed in a 6% denaturing polyacrylamide gel side by side with the sequencing ladders obtained on the respective fragments of the CTV genome, cloned into the replicon CTV-DeltaCla333 or into pBluescript SK (Stratagene) vector, with the same primers used for the primer extension reactions. Total RNA extracted from N. benthamiana mesophyll protoplasts infected with RNA transcripts of the p20 +1 A/G mutant was used as a template to determine the alternate sgRNA 5' terminus of this mutant. Total RNA was denatured at 90C, 3' polyadenylated by yeast poly(A) polymerase (U.S. Biochemicals), and reverse transcribed using avian myeloblastosis virus (U.S. Biochemicals) and oligo(dT) (M111; 5' GGTCTCGAG(T)18 3' ). Amplification of the 5' end of the p20 +1 A/G sgRNA was carried out with the negative-sense primer C582 (5' GTACCTGCAGATCCTGATGGTCTCCGTTCA 3', corresponding to positions 17943 to 17924, with the PstI restriction site underlined and nontemplate nucleotides in italics) and M111. The amplified products were cloned into pGEM-T Easy vector according to the manufacturer's instructions (Promega). Nucleotide sequencing was performed with an automatic sequencer (Applied Biosystems model 373) at the Interdisciplinary Center for Biotechnology Research DNA sequencing core facility of the University of Florida (Gainesville, Fla.). Plasmid constructions. | All the mutants used in this study were constructed in a CTV replicon, CTV-DeltaCla333, which does not produce any sgRNA . Plasmid pCTV-DeltaCla333 contains a deletion of nucleotide 10852 (ORF 1b) to nucleotide 18526 (leaving the last 493 nucleotides of the p23 gene, followed by the 3' noncoding region of CTV), with unique restriction sites (XhoI, PstI, StuI, and ClaI) added at the end of ORF 1b to facilitate insertion of DNA fragments . To generate CP, p18, p13, p20, and p23 constructs, DNA fragments of the infectious clone pCTV9 , corresponding to positions 15895 to 16631, 16574 to 16894, 17120 to 17568, 17510 to 17943, and 18163 to 18526, respectively, were amplified by PCR using a positive-sense primer containing an XhoI restriction site and a negative-sense primer providing a blunt end. The amplified products were digested with XhoI and ligated into pCTV-DeltaCla333, which was previously digested with XhoI and StuI. The DNA fragments used to create the sgRNA mutants were generated by overlap extension PCR using primers containing specific single nucleotide substitutions that were introduced in the position corresponding to the 5' terminus of the sgRNA and also in the nucleotides around this position in the genomic RNA. All the nucleotide mutations introduced were confirmed by sequencing as described above. Protoplast transfections and Northern blot hybridization. | Isolation and polyethylene glycol-mediated transfection of mesophyll protoplasts from N. benthamiana were performed following the procedures described by Navas-Castillo et al. and Satyanarayana et al. . In vitro-capped transcripts were generated from NotI-linearized DNA constructs using SP6 RNA polymerase (Epicentre Technologies) and were used directly for protoplast inoculation (~1 x 106). Protoplasts were harvested at 4 days postinoculation, and total RNA was extracted . The 5'-terminal 600 nucleotides and the 3'-terminal 900 nucleotides of the CTV T36 isolate were cloned in pGEM-7Zf vector (Promega) and used to generate positive- and negative-stranded RNA-specific digoxigenin-labeled riboprobes. The riboprobes were examined by their specificity and equalized using the double-stranded RNA as described previously . Total RNAs were analyzed by Northern blot hybridization using 5' and 3' positive- and negative-stranded RNA-specific riboprobes. The relative amounts of positive-stranded sgRNA were quantified using Northern blot films by scanning and densitometry with the OS-SCAN program (Oberlin Scientific, Oberlin, Ohio). The values of the mean and standard deviation of positive-stranded sgRNA produced for each construct were obtained from at least two to six independent protoplast transfections. RESULTS : Determination of the 5' termini of CTV sgRNAs corresponding to the CP, p18, and p13 genes. | Previously, the 5' termini of p20 and p23 sgRNAs, the two most abundantly produced sgRNAs, were mapped to an adenylate , which also was the nucleotide at the 5' terminus of the genomic RNA . To examine whether the presence of an adenylate at the 5' termini of the CTV sgRNAs is a general phenomenon, we determined the 5' termini of other sgRNAs produced at lower amounts than the p20 and p23 sgRNAs: those of the intermediately abundant CP and p13 genes, and the low-produced p18 gene . Total RNA extracted from plants infected with CTV and primers located approximately 100 nucleotides downstream of the start codon of the corresponding gene were used for a primer extension reaction to synthesize complementary DNAs (cDNAs). The same primers were used to generate the sequence ladder of the corresponding genomic sequence that was loaded in lanes adjacent to the primer extension products. In our control reactions of the known p20 sgRNA , the primer extension product tended to migrate slightly slower than sequencing products, as observed by others . The cDNA product of the sgRNA for CP migrated as a single band coinciding with nucleotide 16115 of the adjacent sequence ladder . The primer extension products of the sgRNAs corresponding to the p18 and p13 genes migrated as single bands of lower intensity mapping at positions 16751 and 17316, respectively . The results of the primer extension analysis suggested that each of the CP, p18, and p13 sgRNAs started with an adenylate and had 5' leaders of 37, 38, and 10 nucleotides, respectively. Thus, 5 of 10 CTV sgRNAs mapped so far have adenylates as their 5' termini, and the genomic sequence surrounding this nucleotide is an AU-rich region. We attempted to map the 5' termini of the sgRNAs corresponding to the remaining five genes of CTV but did not obtain unequivocal results, probably due to the lower abundance of these sgRNAs. Effect of modification of the initiation site context of the CP sgRNA. | We do not yet know whether CTV produces its sgRNAs by internal promotion on the full-length minus strand or by termination during minus-strand synthesis followed by transcription from the sgRNA minus strand to amplify the sgRNA plus strand. Thus, the +1 nucleotide would be the template nucleotide to initiate sgRNA synthesis on the genomic negative strand, if the sgRNAs were produced by promotion, or the nucleotide near the minus-stranded sgRNA 3' terminus, if the sgRNAs were produced by termination. By convention, we will refer to the +1 nucleotide in the positive strand sequence corresponding to the 5' terminus of the sgRNA (and also its complement in the minus strand). Thus, although we refer to numbering in the positive strand sequence, all mutations in the plus strands create complementary changes in the minus strand sequences. It previously has been shown that mutation of the +1 nucleotide can prevent positive-stranded sgRNA synthesis in vivo and that this process can be used to corroborate the primer extension results and/or uncouple positive- and negative-stranded sgRNA synthesis. We thus examined the effects of alteration of the +1 nucleotide of different sgRNAs of CTV. Since full-length CTV produces 10 3'-terminal sgRNAs , it is difficult to visualize alterations to one specific sgRNA. To examine effects of mutations on the production of one sgRNA, we cloned a selected gene, including its controller element, into the CTV replicon CTV-DeltaCla333, which will produce only the sgRNA of the inserted controller element . In vitro transcripts of these mutants were used to inoculate N. benthamiana mesophyll protoplasts, and total RNA was extracted and analyzed by Northern blot hybridization using four different riboprobes: positive- and negative-stranded RNA-specific probes corresponding to the genomic 3' and 5' ends. We mutated the +1 nucleotide of the CP sgRNA from A to either C, U, or G to create the mutants CP +1 A/C, CP +1 A/U, and CP +1 A/G, respectively . Analysis of the positive- and negative-stranded RNAs from these mutants showed that none of the mutations of the +1 nucleotide prevented production of the CP sgRNA. Yet, there were some reductions in the accumulation of this sgRNA. CP +1 A/C, CP +1 A/U, and CP +1 A/G produced about 80, 50, and 25% as much positive-stranded sgRNA as the wild-type CP construct . There was no indication that any of the mutations affected the ratio of positive to negative strands of the sgRNA. Accumulation of the sgRNA of each mutant appeared to be reduced approximately the same proportion in each strand. With the continued synthesis of the sgRNAs after alteration of the +1 nucleotide, it was not known whether the replicase complex initiated with the substituting nucleotide at the +1 position, or whether it chose to initiate at an alternative position. Thus, we needed to identify the 5' termini of mutant sgRNAs. We were able to determine the 5' termini of the sgRNAs from wild-type infections of whole plants. However, these mutants were capable of replication only in protoplasts in which only about 0.1% or less of the protoplasts become infected with CTV transcripts , resulting in much lower levels of sgRNAs. These lower levels of sgRNAs in protoplasts were thus near our detection limit when using primer extension. For this reason, we chose to examine the 5' terminus of the more-abundant sgRNA of the CP +1 A/C mutant. The primer extension reaction was performed using total RNA and a primer located approximately 130 nucleotides downstream of the native +1 nucleotide of the CP sgRNA. The cDNAs obtained comigrated as doublet bands . The upper band coincided with an adenylate at nucleotide 16116 of the adjacent sequence ladder obtained with the same primer , while the lower band mapped between the positions 16116 (A) and 16117 (U), which might be an artifact of the reaction. We commonly found double bands from analysis of primer extension reactions with RNA extracted from protoplasts, but not with RNA extracted from plants . It was clear that the 5' terminus did not correspond to the substituting cytidylate at nucleotide 16115 that was positioned in the native +1 site and that initiation began at a different site. Our best interpretation was that the adenylate in position 16116 represented the 5' terminus of the CP +1 A/C sgRNA , suggesting the preference of the CTV replicase complex to use an adenylate (uridylate in the negative strand) as initiating nucleotide for CP sgRNA synthesis. The apparent initiation at an alternate adenylate (+2 A of the wild-type sgRNA leader) was examined by mutating it to a cytidylate, generating CP +1 +2 A/C (or CPCC) . The amount of sgRNA produced in this construct decreased substantially to less than 5% of that of the wild-type CP construct . The amount of the minus-stranded sgRNA decreased proportionally . Attempts to map the 5' terminus of this mutant were not successful. It was possible that the new 5' terminus of this construct (CPCC) was also an adenylate, and the closest one was at position 16120 (+6 position of the wild-type sgRNA leader). We next created a mutant containing a substitution of this +6 A to C (CPCC +6 A/C), and this construct still produced similar amounts of sgRNA (approximately 5%) , suggesting that the CTV replicase complex could have selected a different nucleotide to initiate sgRNA synthesis. We examined whether creating an alternate adenylate nearer the native +1 site of the CPCC mutant would increase the efficiency of sgRNA synthesis. We substituted the nucleotides at positions +3, -1, and -2 from U to A to generate the mutants CPCC +3 U/A, CPCC -1 U/A, and CPCC -2 U/A, respectively . Constructs CPCC -1 U/A and CPCC -2 U/A did not produce increased amounts of sgRNA, but CPCC +3 U/A produced a fourfold increase of CP sgRNA compared to CPCC, suggesting that the presence of an adenylate at the +3 position was more favorable than at the -1 or -2 positions. Effect of modification of the initiation site contexts of other sgRNAs. (i) p13 and p23 sgRNAs. | The p13 and p23 sgRNAs are produced in intermediate and high amounts, respectively. The 5' termini of p13 and p23 sgRNAs mapped to adenylates at positions 17316 and 18353, respectively, in a similar context: UXAAUU (+1 nucleotide underlined). Therefore, it is possible that mutations near the +1 nucleotide might have similar effects on production of these sgRNAs. We created mutants p13 +1 A/C and p23 +1 A/C by substitution of the +1 adenylate with cytidylate and analyzed positive- and negative-stranded RNA accumulation by Northern blot hybridization using specific riboprobes corresponding to the 3' end of the genomic RNA . The mutation of the +1 adenylate decreased positive-stranded sgRNA accumulation of both genes to approximately 40% of that of the wild-type constructs, and the decrease in sgRNA minus-strand accumulation was proportional . Next, we substituted the adenylates at positions +1 and +2 with cytidylates to generate constructs p13 +1 +2 A/C and p23 +1 +2 A/C . Both mutants replicated in protoplasts and produced less than half as many positive- and negative-stranded sgRNAs as the mutants with only the +1 adenylate substitution . These data suggested that the CTV replicase complex likely follows similar rules for the production of CP, p13, and p23 sgRNAs. (ii) p20 sgRNA. | The p20 sgRNA is often produced at the highest amount. The 5' terminus of this sgRNA was mapped to the adenylate at position 17713 , in the context AGUAUAA (+1 nucleotide underlined) . The sequence surrounding the +1 nucleotide of the p20 sgRNA was slightly different from that surrounding the +1 nucleotide of the other three CTV sgRNAs examined (CP, p13, and p23). To examine the effect of modifications of the nucleotides near the +1 nucleotide on p20 sgRNA synthesis, the mutant p20 +1 A/C was created by substitution of the +1 adenylate by cytidylate . This mutant produced reduced amounts of both positive- and negative-stranded sgRNAs . The accumulation of the positive strand was approximately 25% of that of the wild-type construct . Other adenylates were located at positions +3 and +4. The adenylate at the +3 position was replaced by a cytidylate to create the mutant p20 +1 +3 A/C . Yet, this mutant produced approximately the same amount of positive-stranded sgRNA as the mutant with only the +1 adenylate changed (p20 + 1A/C) . The p20 sgRNA leader contained a uridylate at the +2 position. To determine if a change of this uridylate affected the accumulation of the p20 sgRNA, this nucleotide was replaced with a cytidylate, resulting in the mutant p20 +1 A +2 U/C . This mutant replicated and produced both plus- and minus-stranded sgRNAs in increased amounts compared to the mutant with only the +1 adenylate mutated (p20 +1 A/C) . Subsequently, we analyzed the effect of the substitution of the +1 A with other nucleotides, U and G, resulting in mutants p20 +1 A/U and p20 +1 A/G, respectively . The p20 +1 A/U mutant produced 50% as much plus sgRNA as wild type, whereas accumulation of sgRNA from the p20 +1 A/G mutant was higher than that of the wild-type construct. We determined the 5' terminus of the p20 +1 A/G mutant by polyadenylating the 3' terminus of the negative-stranded sgRNA followed by cloning and sequencing. The sequences of six independent clones showed that the 5' terminus of the p20 +1 A/G sgRNA was the guanylate at the +1 position (nucleotide 17713). These results suggested that the context near the +1 site of the p20 sgRNA greatly modulates its production and that the CTV replicase complex can efficiently use a guanylate (cytidylate in the negative strand) as an alternate initiation nucleotide for p20 sgRNA synthesis. (iii) p18 sgRNA. | CTV sgRNAs are produced in higher amounts when positioned nearer the 3' terminus of the genome . Yet, the p18 sgRNA is produced in low amounts even though it is located near the 3' end. The 5' terminus of the p18 sgRNA was mapped to an adenylate at nucleotide 16751 in the positive strand context UUAUU (+1 nucleotide is underlined) (Fig. and ). To examine the effects of the modifications of the nucleotides in and around the +1 position, the +1 adenylate of p18 sgRNA was mutated to cytidylate to generate the construct p18 + 1A/C . This mutant produced the positive-stranded sgRNA at about the same level as the wild type , suggesting that mutation of the +1 nucleotide did not greatly affect accumulation of p18 sgRNA. We then substituted the +1 A with U or G, generating p18 +1 A/U and p18 +1 A/G . The amounts of positive-stranded sgRNA accumulated by both of these mutants were considerably more than that of the wild-type construct . Again, sgRNA plus- and minus-strand accumulations were proportional . These results suggested that the native context for the +1 nucleotide was unfavorable for optimal synthesis of the p18 sgRNA, with most alterations having a positive effect. Next, we analyzed the effect of additional substitutions of the uridylate in the +2 position (p18 +1 A +2 U/C) or the adenylate in the -3 position (p18 +1 -3 A/C) . Mutant p18 +1 A +2 U/C accumulated approximately 50% of sgRNA plus strand compared with the p18 wild-type construct. Modification of both adenylates in positions +1 and -3 prevented measurable synthesis of the sgRNA, suggesting that the adenylate at the -3 position could function as the alternate 5' terminus of the p18 +1 A/C mutant. Production of positive-stranded 5'-terminal sgRNAs. | We previously found that, in addition to the positive-stranded 3'-terminal sgRNA and its complementary negative strand, each controller element of the 10 3' genes also produced a positive-stranded 5'-terminal sgRNA, apparently by termination during genomic plus-strand synthesis . To analyze whether the mutations near the +1 site, which reduced or altered sgRNA production, altered or uncoupled production of the 5'-terminal sgRNA, we examined the accumulation of this sgRNA from total RNA isolated from N. benthamiana protoplasts infected with all of these mutants by Northern blot hybridizations using probes specific to the first 600 nucleotides of the 5' end of the genomic RNA. All of these mutants accumulated positive- but not negative-stranded 5'-terminal sgRNAs (data not shown) in proportion to the accumulation of the 3'-terminal sgRNAs, indicating that all three sgRNAs were produced in concert. DISCUSSION : CTV expresses its 3'-terminal genes through a nested set of 3'-terminal sgRNAs produced at different levels . We were able to map the 5' termini of three more sgRNAs (CP, p13, and p18) in addition to the two previously mapped (p20 and p23 ). All were found to possess a +1 adenylate, the same as the 5' end of the genome . Additionally, the 5' termini of the p13, p18, p20, and p23 sgRNAs of other CTV strains also map to an adenylate . Thus, all CTV sgRNAs mapped so far have an adenylate as the 5' terminus, suggesting that initiation of RNA synthesis from a uridylate on the minus strand is preferred by the CTV replicase complex. Sequence similarity at the 5' ends of the genomic RNA and the 5' leaders of sgRNAs is a characteristic of numerous viruses , including other members of the Closteroviridae such as BYV and Sweet potato chlorotic stunt virus . This sequence similarity has been proposed to be involved in assisting the viral replicase complex to recognize and interact with specific minus-strand signals for the initiation of synthesis of plus-strand genomic RNA and sgRNAs. The first nucleotides of the 5' end of the genomic RNA (AAUUUCA) and the positive genomic sequence around the 5' termini of sgRNAs (underlined, CP [UUAAUUGA], p18 [UUAUUGAA], p13 [UCAAUUGCU], p20 [GUAUAACGU], p23 [UUAAUUCGA]) were similar but not identical AU-rich regions. In general, modification of the +1 nucleotide has been a useful tool for characterizing viral sgRNA production. Alteration of this nucleotide in viruses that produce sgRNAs by internal initiation (promotion), including alpha-like viruses, tends to abolish or greatly decrease sgRNA synthesis and can be used to confirm the mapping of sgRNAs 5' termini . Modification of the +1 nucleotide of the sgRNAs of two viruses thought to produce their sgRNAs by termination, Tomato bushy stunt virus and Flock house virus, greatly inhibited accumulation of positive-stranded sgRNAs with little effect on negative-strand levels , suggesting that the negative strands are produced first. The RdRp domain of CTV falls within the evolutionary lineage of the alphavirus supergroup. Therefore, it would be expected that the viral replicase complex follows a mechanism of promotion for sgRNA synthesis. Yet, modification of the initiation nucleotides of CTV sgRNAs did not stop their production as expected and, consequentially, was not useful to confirm the primer extension mapping. Additionally, alpha-like viruses that produce sgRNAs by promotion are expected to follow the +1 pyrimidine and +2 adenylate rule (in reference to the minus strand) as initiation nucleotides for positive-stranded RNAs . CTV did not strictly follow this rule for initiation of sgRNAs, often initiating with a +1 uridylate, +2 uridylate, and +3 adenylate. CTV also did not comply with our expectations of the alternative mechanism, production of sgRNA minus strands by termination, because modification of the +1 nucleotides of the different CTV sgRNAs did not uncouple plus- and minus-strand sgRNA accumulation or significantly affect their ratios. Another abnormality of CTV is that the initiation nucleotide of its sgRNAs is not positioned within the controller (promoter-terminator) element. The +1 nucleotide of the CP is positioned 5 nucleotides outside of the core element . Thus, the results of the modification of the +1 nucleotides of CTV sgRNAs suggest that CTV does not follow either of the characterized patterns of sgRNA synthesis: production of sgRNAs via internal promotion or production of negative-stranded sgRNAs by termination and sgRNA plus strands by transcription. Instead, it appears that CTV sgRNA production follows an alternative or modified mechanism. Studies performed in vitro have shown the flexibility of Brome mosaic virus, Cowpea chlorotic mottle virus, and Cucumber mosaic virus RdRps to select +1 sites for sgRNA synthesis by using alternate positions and/or nucleotides when the +1 site is modified (-, , , , ). All these examples showed that in vitro these RdRps always initiate with a guanylate or adenylate as authentic or alternate 5' termini for sgRNA synthesis, respectively, but most important is the flexibility to use an alternative initiation position. After substitution of the native +1 adenylate of the CTV CP sgRNA at position 16115 with a cytidylate, the replicase complex initiated with an alternate adenylate at position 16116 instead of using the cytidylate at position 16115. Modification of the alternate adenylate (16116) to cytidylate further reduced sgRNA production. We technically were not able to map the alternate 5' nucleotides of the other sgRNAs produced at lower levels, but additional mutations were consistent with the continued preference of the replicase complex to initiate sgRNA synthesis with adenylates. In general, if the sgRNAs had a second adenylate adjacent or near the +1 nucleotide, as in the CP, p13, and p23 sgRNA leaders, substitution of the +1 adenylate with another nucleotide resulted in relatively small reductions of sgRNA accumulation, perhaps by half, suggesting that the replicase complex had flexibility to initiate with a nearby alternate adenylate, since modification of such an alternate adenylate further reduced sgRNA accumulation. Providing adenylates in alternate positions near the original +1 site sometimes increased sgRNA accumulation, but not always. For example, providing an alternate adenylate at positions -1 or -2 in the CP constructs with the native +1 and +2 adenylates removed did not restore wild-type levels of sgRNA synthesis. However, the insertion of an adenylate at +3 increased the CP sgRNA accumulation levels. These results suggest that the position of the initiation adenylate and its context are important for CTV sgRNA accumulation. It appears that the CTV replicase complex prefers to initiate with an adenylate (uridylate in the negative strand). The fact that other members of the Closteroviridae can initiate genomic and sgRNAs with guanylate suggests that the CTV replicase complex might also initiate with other nucleotides. Substitution of the +1 adenylate of both the p20 and p18 genes with guanylate resulted in increased production of sgRNA. Interestingly, the CTV replicase complex initiated synthesis of p20 +1 A/G sgRNA with the guanylate (cytidylate in the negative strand) in the +1 site more efficiently than with the +1 adenylate. The 5' terminus of the p18 sgRNA from a full-length virus of a different CTV isolate with a natural substitution of the adenylate at position 16751 with guanylate (similar to the p18 +1 A/G mutant) has also been mapped . In this case, sgRNA synthesis did not initiate with the guanylate at position 16751, but instead with the adenylate at position 16748. The CTV replicase complex appears to be flexible in using an initiation adenylate in other positions downstream (isolate with a natural mutation of the p18 +1 A to G) or upstream (CP +1 A/C mutant) of the native +1 nucleotide, or using a different nucleotide in the native +1 position (p20 +1 A/G). Thus, if the CTV replicase complex initiated sgRNAs with a purine, preferably with adenylates, as the 5' terminus in all of the mutants examined herein, it would suggest that the replicase complex could initiate at positions ranging from -5 to +6 (p23 +1 +2 A/C mutant) relative to the native +1 position and still produce at least minimal amounts of sgRNA. Although the CTV replication complex appears to prefer an adenylate, or at least a purine, as the 5' terminus of its sgRNAs, we still cannot exclude that some of the alterations observed in levels of sgRNA synthesis were due to context rather than a specific initiation nucleotide. Previously, we have shown that both position within the genome and promoter strength control the level of sgRNA production . The sequence context of the initiation site of CTV sgRNAs, which is outside the core promoter, appears to be an additional regulatory mechanism to control sgRNA synthesis. Mutation of the +1 A to C or U of the highly abundant p20 sgRNA led to decreased sgRNA production, whereas mutation of the +1 A to G resulted in increased levels of sgRNA accumulation. Thus, even the most abundant sgRNA appears to be capable of increasing synthesis even further. The p18 sgRNA is produced minimally even though it is positioned within the genome to allow high levels of production. Substitution of the +1 adenylate with any other nucleotide increased p18 sgRNA accumulation, suggesting the native context was designed to be down-regulated. These data not only emphasize the importance of the +1 nucleotide context but also demonstrate a tool for examining and manipulating gene regulation in vivo. FIG. 1. : Determination of the 5' termini of CTV sgRNAs. Determination of the 5' termini of CTV sgRNAs. (A) Schematic diagram of the CTV genome. Boxes represent the ORFs with the respective numbers and encoded products. (B) Northern blot analysis of positive-stranded RNAs isolated from N. benthamiana mesophyll protoplasts inoculated with CTV9 full-length infectious clone virions. The genomic RNA (gRNA) and the 3'-terminal sgRNAs are indicated by an arrow and a right brace, respectively. (C) Analysis of the primer extension products in a 6% polyacrylamide gel side by side with the sequencing ladders obtained from the cloned fragments. The primer extension experiment was performed using total RNA isolated from plants infected by the CTV T36 isolate of Florida as template and the primers PM37 (nucleotides 16247 to 16228), C244 (nucleotides 16894 to 16874), and PM36 (nucleotides 17439 to 17419) specific to CP, p18, and p13 sgRNAs, respectively. Sequencing reactions were performed with the same primers. The first 10 nucleotides of the sequence of the 5' leader of the sgRNA are presented to the right. The 5'-to-3' direction is from top to bottom. FIG. 2. : Modification of the initiation site context of CP sgRNA. Modification of the initiation site context of CP sgRNA. (A) Schematic diagram of the CTV-DeltaCla333, the parental plasmid used for cloning the CP controller element and mutants. The numbers indicate the genomic region deleted to create this construct. The restriction enzymes sites introduced at the end of ORF 1b in CTV-DeltaCla333 are indicated at the top. The sgRNA sequence around the +1 nucleotide of the CP gene of CTV and related mutants cloned into CTV-DeltaCla333 between XhoI and StuI restriction sites are presented. The underlined nucleotide in the sequence represents the +1 nucleotide. The mutated nucleotides are indicated in bold in each mutant. The numbers indicate the termini of the cloned fragments. The mean and standard deviation (SD) of the amount of positive-stranded sgRNA produced by each construct is indicated onthe right. (B) Northern blot analysis of total RNA isolated from N. benthamiana mesophyll protoplasts inoculated with RNA transcripts from wild-type and mutants of CP sgRNA. The blots were hybridized with 3' 900-nucleotide positive- and negative-stranded digoxigenin-labeled RNA probes. (C) Analysis of the primer extension product of the CP +1 A/C mutant in a 6% polyacrylamide gel side by side with the sequencing ladder obtained on the cloned fragment. The primer extension experiment was performed using the primer C502 (nucleotides 16241 to 16217) and total RNA isolated from N. benthamiana mesophyll protoplasts inoculated with RNA transcripts from the CP +1 A/C mutant. The sequencing ladder presented on the right was generated using the same primer. The sequence of the first 10 nucleotides of the sgRNA 5' leader is presented to the left in the 5'-to-3' direction from top to bottom. FIG. 3. : Mutational analysis of the 5' termini of p13 and p23 sgRNAs. Mutational analysis of the 5' termini of p13 and p23 sgRNAs. (A) Schematic diagram of the CTV-DeltaCla333 genome as outlined the legend for Fig. . The sgRNA sequences around the position +1 of the p13 and p23 genes of CTV and related mutants were cloned into CTV-DeltaCla333. The +1 nucleotide is underlined, and the mutated nucleotides in each mutant are indicated in bold. The numbers indicate termini of the cloned fragments. The mean and the standard deviation (SD) of the sgRNA positive strand produced by each construct are presented on the right. (B) Northern blot analysis of the total RNA isolated from N. benthamiana mesophyll protoplasts inoculated with RNA transcripts of constructs of wild-type p13 and p23 sgRNAs and their respective mutants using 3' 900-nucleotide positive- and negative-stranded digoxigenin-labeled riboprobes. FIG. 4. : Modification of the initiation site context of p20 sgRNA. Modification of the initiation site context of p20 sgRNA. (A) Schematic diagram of the CTV-DeltaCla333 genome used to clone wild-type and mutants of p20 sgRNA. The nucleotides in bold indicate the mutated nucleotide in each construct, and the underlined nucleotides represent the +1 nucleotide. The numbers indicate termini of the cloned fragments. The mean and the standard deviation (SD) of the amount of sgRNA positive strand produced by each construct are presented on the right. (B) Northern blot analysis of the total RNA isolated from N. benthamiana mesophyll protoplasts inoculated with RNA transcripts of wild-type and p20 sgRNA mutants, hybridized with 3'-end digoxigenin-labeled RNA probes. FIG. 5. : Modification of the initiation site context of p18 sgRNA. Modification of the initiation site context of p18 sgRNA. (A) Schematic diagram of the CTV-DeltaCla333 genome (details as outlined in the legend for Fig. ). The sgRNA sequence around the +1 nucleotide of p18 gene of CTV and related mutants were cloned into CTV-DeltaCla333. Nucleotide +1 is underlined in the wild-type sgRNA sequence. The nucleotides in bold represent the mutated nucleotides in each construct. The numbers indicate termini of the fragments cloned into CTV-DeltaCla333. On the right is presented the mean and the standard deviation of positive-stranded sgRNA produced by each construct. (B) Northern blot analysis of RNA isolated from N. benthamiana mesophyll protoplasts inoculated with RNA transcripts from the wild-type p18 sgRNA construct and its mutants. The blots were hybridized with 3'-end positive- and negative-stranded digoxigenin-labeled RNA probes. Backmatter: PMID- 12915541 TI - Identification of a Mutation in Editing of Defective Newcastle Disease Virus Recombinants That Modulates P-Gene mRNA Editing and Restores Virus Replication and Pathogenicity in Chicken Embryos AB - Editing of P-gene mRNA of Newcastle disease virus (NDV) enables the formation of two additional proteins (V and W) by inserting one or two nontemplated G residues at a conserved editing site (5'-AAAAAGGG). The V protein of NDV plays an important role in virus replication and is also a virulence factor presumably due to its ability to counteract the antiviral effects of interferon. A recombinant virus possessing a nucleotide substitution within the A-stretch (5'-AAgAAGGG) produced 20-fold-less V protein and, in consequence, was impaired in replication capacity and completely attenuated in pathogenicity for chicken embryos. However, in a total of seven serial passages, restoration of replication and pathogenic capacity in 9- to 11-day-old chicken embryos was noticed. Determining the sequence around the editing site of the virus at passage 7 revealed a C-to-U mutation at the second nucleotide immediately upstream of the 5'-A5 stretch (5'-GuUAAgAAGGG). The V mRNA increased from an undetectable level at passage 5 to ca. 1 and 5% at passages 6 and 7, respectively. In addition, similar defects in another mutant possessing a different substitution mutation (5'-AAAcAGGG) were restored in an identical manner within a total of seven serial passages. Introduction of the above C-to-U mutation into the parent virus (5'-GuUAAAAAGGG) altered the frequency of P, V, and W mRNAs from 68, 28, and 4% to 15, 44, and 41%, respectively, demonstrating that the U at this position is a key determinant in modulating P-gene mRNA editing. The results indicate that this second-site mutation is required to compensate for the drop in edited mRNAs and consequently to restore the replication capacity, as well as the pathogenic potential, of editing-defective NDV recombinants. Keywords: Introduction : Newcastle disease virus (NDV) belongs to the new genus Avulavirus within the family Paramyxoviridae. NDV is the causative agent of one of the most devastating diseases of poultry worldwide. The negative-strand RNA virus genome of NDV encodes six genes (in the order 3'-NP-P-M-F-HN-L-5'). Except for the P gene mRNA, all express one major viral structural protein from a single open reading frame (ORF). The P gene mRNAs of the subfamily Paramyxovirinae express several proteins by the use of alternative reading frames and by a process of mRNA editing. The mRNA editing, which involves insertion of pseudotemplated G residue(s) occurs by a polymerase stuttering mechanism . Like in other members of paramyxovirinae, the P gene mRNA of NDV is edited by inserting one or two G residues into a G run within a conserved editing locus (5'-AAAAAGGG) . As a result, three P gene derived mRNA species are produced which encode for the P, V, and W ORFs (unedited, with +1 and +2 frameshifts, respectively). The V and W proteins are colinear to the N-terminal half of the P protein but have different carboxy-terminal parts. The majority of the accessory proteins, including V and/or C proteins encoded by the P genes of different members of the subfamily Paramyxovirinae have been shown to be dispensable for virus replication in cell cultures . Interestingly, however, mutants lacking V or C accessory proteins were strongly attenuated in pathogenicity in vivo . The mechanism of the in vivo attenuation was demonstrated to involve the interferon system in which accessory proteins, particularly V or C proteins, inhibit interferon signaling or suppress interferon production (reviewed in references and ). Additional roles of paramyxovirus accessory proteins in viral RNA transcription and synthesis, as well as in virus assembly and propagation, were also recently uncovered, demonstrating functional versatility of these proteins . The dual function of the V protein of NDV in virus replication, as well as in virus virulence, was shown by generating mutants lacking the entire V and W proteins or only the unique C-terminal part of the V protein or by constructing a mutant expressing low levels of V protein . The C terminus of the V protein, which is highly rich in cysteines, is of particular interest since it is conserved in all three genera of paramyxovirinae, with the exception of human parainfluenza virus types 1 and 3 . NDV mutants lacking the entire V and W proteins or only the C-terminal portion were severely impaired in virus pathogenicity and virus propagation both in vitro and in vivo, indicating that the C terminus of V protein is responsible for this incompetence. Very recently, the NDV V protein was demonstrated to have an interferon-antagonist activity, and this activity was shown to be located in the C terminus of the V protein . Therefore, the mechanism by which V protein plays a role in virulence is presumably due to its ability to antagonize the antiviral effects of interferon and its exact role in virus replication remains to be established. As opposed to the mutations that completely abolished the function of NDV V protein, a single A-to-G change within the A-run only reduced the editing frequency and thus lowered the level of additional proteins generated by RNA editing . The mutant NDV-P1 expressing low-level V protein grew in 9- to 11-day-old embryonated eggs to a titer of 6.7 log10 50% embryo infectious dose (EID50)/ml at passage level 5. In passages 6 and 7 the titers of the mutant in embryonated eggs have increased to 7.1 and 9.3, respectively. The latter titer is identical to the titer of the parent virus. This dramatic increase in replication efficiency was also accompanied with increase in pathogenicity for 9- to 11-day-old embryonated specific-pathogen-free (SPF) chicken eggs. Compared to the parent virus, the mutant at passage level 5 was attenuated by as much as 106-fold for young embryos. The degree of attenuation was, however, dramatically decreased in subsequent passages and the mutant was almost as pathogenic as the parent virus at passage level 7 for 9- to 11-day-old embryonated SPF chicken eggs. Here we describe a compensatory mutation, which is responsible for the restoration of the replication efficiency and pathogenicity in 9- to 11-day-old embryonated SPF chicken eggs. Moreover, introducing the mutation into the backbone of the parent virus provided direct evidence for the role of the second-site mutation in modulating the editing frequency of NDV. A mechanism for how NDV restores RNA editing to ensure the expression of P-gene derived proteins for efficient NDV replication and virulence is discussed. MATERIALS AND METHODS : Construction of editing-site mutants. | The plasmid pflNDV, expressing the full-length antigenome RNA of the lentogenic ND vaccine strain, Clone-30 , was used to introduce mutations. Construction and generation of the mutant NDV-P1, possessing an A-to-G replacement at position 5'-A3 (5'-AAgAAGGG) was described previously . Another mutant containing a replacement at position 5'-A4 in the A-stretch of the editing site was constructed by performing PCR with the template pflNDV with forward primer 4 (5'-GCTCCTCGCGGCTCAGACTCG-3', nucleotides 151 to 171) and reverse primers A4 (5'-CCATGGGCCCTGTTTAGCATTGGACG-3', nucleotides 2269 to 2294). In order to introduce the identified second-site mutation into the backbone of the parent Clone-30, PCR was performed with the same forward primer 4 and a reverse primer T2 (5'-CCATGGGCCCTTTTTAACATTGGACG-3'). PCR products were then digested with AatII/ApaI and cloned into the same sites of pflNDV. The region newly introduced into each clone was sequenced to rule out PCR-introduced errors. The resultant full-length clone with a single nucleotide substitution at position 5'-A4 and the clone possessing the identified second-site mutation were named NDV-PC4 and NDV-T2, respectively . Recovery of recombinant viruses. | Transfections and virus recoveries were done as described preiviously . Approximately 1.5 x 106 BSR-T7/5 cells stably expressing phage T7 RNA polymerase were grown overnight in 3.2-cm-diameter dishes. Cells were transfected with plasmid mixtures containing 5 mug of pCite-NP, 2.5 mug of pCite-P, 2.5 mug of pCite-L, and 10 mug of one of the full-length clones by using a mammalian transfection kit (CaPO4 transfection protocol; Stratagene). Three days after transfection, supernatant was harvested and inoculated into the allantoic cavity of 9- to 11-day-old embryonated SPF chicken eggs. After 3 to 4 days of incubation, the presence of virus in the allantoic fluid was determined by rapid plate hemagglutination (HA) test with chicken erythrocytes. The infectious titer, expressed as the EID50 was calculated by using the method of Reed and Muench . Determining sequences around the P gene mRNA editing site. | Total RNA was isolated from BSR-T7/5 cells infected with serially passaged recombinant viruses by using the RNeasy kit (Qiagen). Reverse transcription by avian myeloblastosis virus reverse transcriptase on 1 mug of total RNA was primed with NDV P gene-specific oligonucleotide P#13 (5'-CCACCCAGGCCACAGACGAAG-3', nucleotides 2176 to 2196) or oligo(dT) primer to amplify only mRNAs. DNA amplification was then performed with NDV P gene-specific oligonucleotides P#13 and P#17 (5'-ATGAATTCAGCTGTTGGA-3', nucleotides 2671 to 2688). The PCR products were analyzed on a 1% agarose gel and used directly for sequencing or cloned by using Invitrogen's TOPO TA cloning kit. Sequences of ca. 300 bp flanking the editing site were determined from independent colonies and examined for the presence of any mutation around the conserved editing site. Serial passaging of recombinant viruses in embryonated eggs. | Recombinant viruses were independently passaged serially in 9- to 11-day-old embryonated SPF chicken eggs. The titer of the progeny obtained after each passage was determined by inoculating 10-fold serial dilutions into the allantoic cavity of embryonated eggs. The eggs were observed daily and NDV-specific mortality was scored during the incubation period of 7 days. The HA test was carried out, and the infectious titer, expressed as the EID50, and the mortality (expressed as 50% embryo-lethal dose [ELD50]) was calculated by using the method of Reed and Muench . RESULTS : Serial passage of NDV-P1 yielded progeny viruses that exhibit enhanced replication and restored pathogenicity. | Recently, we have shown that NDV mutants lacking the complete V and W proteins or only the cysteine-rich carboxy-terminal part of the V protein were severely impaired in cell culture growth and completely unable to propagate in 9- to 11-day-old embryonated eggs. In contrast, a mutant NDV-P1 possessing a single nucleotide substitution in the 5'-A5G3 sequence (5'-AAGAAGGG, Fig. ) grew in eggs, albeit to very low infectious titers. The modification in NDV-P1 led to expression of V protein at a 20-fold-lower level and extreme attenuation for the chicken embryos. The findings indicated that the V protein of NDV plays an important role in virus replication, as well as in virus virulence . The infectious titers of NDV-P1 after the fifth and sixth egg passages were 6.7 and 7.1 log10 EID50/ml, respectively. One further passage yielded a titer of 9.3 log10 EID50/ml, a value that is identical to the titer obtained for the parent virus . To determine whether there is also a concomitant increase in virus virulence, 10-fold serial dilutions of passages 5, 6, and 7 of NDV-P1 were inoculated into 11-day-old embryonated SPF chicken eggs and incubated for 1 week. Consistent with the previous observations, NDV-P1 at passage level 5 was not lethal for embryos even with a dose as high as 6 log10 EID50. Passage 6 of NDV-P1 showed some embryo mortality and passage 7 was nearly as virulent as the parent virus . The ELD50 subsequently dropped from >6.7 log10 EID50 at passage 5 to 4.7 and 1.1 log10 EID50 at passages 6 and 7, respectively. For comparison, the ELD50 of the parent virus is 0.3 log10 EID50. Chicken embryos inoculated with the parent virus, rNDV, already started to die at 3 days postinoculation, at doses higher than 4 log10 EID50/ml . The phenotypic changes of NDV-P1 between passages 5 and 7 initially suggested reversion of the introduced mutation at the editing site. We therefore sequenced several independent clones from passage 7 to control the presence or absence of the introduced mutation at the 5'-A5 stretch. In contrast to our expectation, all of the clones maintained the introduced A-G substitution. The stability of the introduced mutation suggested the occurrence of a second-site compensatory mutation. Localization of a second-site compensatory mutation. | In order to localize the mutation responsible for enhancing the replication capacity and restoring pathogenicity of NDV-P1, we sequenced several independent clones derived from passages 5, 6, and 7 . Approximately 300 bp flanking the editing-site sequence were examined for the occurrence of any deviation from the NDV-P1 virus sequence. All clones derived from passage 5 contained no alterations in the sequenced region. For passage 6, 4 of 102 clones contained a C-to-U subtitution (5'-GuUAAgAAGGG), which is the second nucleotide immediately upstream of the 5'-A5G3 editing-site sequence. Surprisingly, the proportions of independent clones possessing this mutation at passage 7 dramatically increased to 88% , demonstrating the growth advantage of NDV-P1 progeny viruses possessing this compensatory mutation. Sequencing of PCR products derived from genomic RNA showed that the second-site mutation is incorporated during genome replication at position 2278 of NDV genome. This mutation is referred as C-U2278 here. These results indicate that a total of seven serial passages in NDV-P1 enabled the virus to introduce a compensatory mutation that overcame the dramatically decreased replication capacity and pathogenicity in chicken embryos. The second-site compensatory mutation C-U2278 partially restores P-gene mRNA editing frequency. | In earlier studies we have demonstrated the presence of a link between the level of V-protein expression and the efficiency of virus replication, as well as virus pathogenicity in chicken embryos . To determine whether the exponential increase in the proportion of clones possessing the second-site compensatory mutation during serial passages of NDV-P1 is accompanied by increased editing frequency, the proportions of P, V, and W ORFs were determined from the above-sequenced plasmid clones. As shown in Fig. and Table , the amount of edited mRNAs increased from an undetectable level at passage 5 to 1 and 6% at passages 6 and 7, respectively. The total amount of edited mRNAs of 6% at passage 7 is more than fivefold-lower than that of the parent virus. However, both V and W mRNAs were detected in passage 7 at a proportion comparable to that of the parent virus, demonstrating restoration of P-gene mRNA editing, albeit at a low level. Similar defects in another editing-site mutant were also restored by the C-U2278 compensatory mutation. | To determine the involvement of the second-site C-U2278 mutation in restoring defects caused by a different substitution mutation, we constructed another mutant with a single base substitution at position 5'-A4 (5'-AAAcAGGG, Fig. ). The mutant NDV-PC4 was recovered as previously described and serially passaged in 9- to 11-day-old embryonated SPF chicken eggs for a total of seven passages. The infectious titers and embryo mortalities at passages 3 and 7 were then determined. As shown in Fig. , the mutation in NDV-PC4 suppressed the efficiency of replication by a 100-fold at passage 3, which was restored to parent virus level within four further passages. Consistent with the observation made for NDV-P1, NDV-PC4 at passage 3 was completely safe for 9- to 11-day-old embryonated eggs even at a dose as high as 6 log10 EID50 . This phenotype was dramatically altered within the subsequent four passages, and the mutant is nearly as pathogenic as the parent virus at passage 7. The ELD50 dropped from >6 log10 EID50 at passage 3 to 0.2 log10 EID50 at passage 7. To determine whether the restoration in replication capacity and pathogenic potential of NDV-PC4 mutant was also accompanied by the second-site C-U2278 mutation and increased editing frequency, the region around the editing site was amplified and cloned. The sequence around the editing site was then determined from independent clones derived from passages 3 and 7. The results show that the introduced point mutation in the 5'-A5 stretch has been retained. As expected, all sequenced clones derived from passage 3 do not contain the C-U2278 mutation. In contrast, 89% of the clones derived from passage 7 possess the second-site compensatory mutation . The mutation in NDV-PC4 suppressed the editing frequency to undetectable level at passage 3, and the patterns of editing at passage 7 were 96, 4, and 0% for P, V, and W mRNAs, respectively . The absence of W mRNA in passage 7 of NDV-PC4 may be explained by the relatively small number of total colonies analyzed compared to that of NDV-P1 . Taken together, the results demonstrate that seven serial passages in embryonated chicken eggs enabled editing-defective NDV mutants to introduce the compensatory mutation that partially overcame the decrease in editing frequency and restored virus replication and virulence. The second-site compensatory mutation C-U2278 modulates P-gene mRNA editing. | Although the C-U2278 mutation clearly plays an important role in partially restoring the editing frequency of an editing-defective mutant, it was of interest to determine the influence of this mutation in the backbone of the parent virus. We therefore introduced this mutation upstream of the 5'-A5G3 sequence of the parent virus (5'-GuUAAAAAGGG). The modified full-length cDNA clone, NDV-T2, together with three support plasmids expressing NDV NP, P, and L proteins, was transfected into BSR-T7/5 cells, and virus recovery was done as described in Materials and Methods. HA was detected in eggs already after the first passage. After one further passage, a titer of 9.4 log10 EID50/ml was obtained, which is very similar to that of the parent virus (9.3 log10 EID50/ml). In addition, the ELD50 of the recombinant NDV-T2 was 0.2 log10 EID50, a value again comparable to that of the parent virus (0.3 log10 EID50). We then determined the editing frequency by sequencing independent clones derived from the second passage. Surprisingly, this alteration gave rise to a virus that produces only 15% unedited P mRNA, which is 4.5-fold less than that produced by the parent virus. The rest 85% of the mRNAs were edited, constituting 44% V and 41% W mRNAs (Fig. and Table ). In the parent virus backbone, the C-U2278 mutation predominantly affected W mRNA by increasing the insertion of two G's by 10-fold, whereas V mRNA increased by a moderate 1.5-fold. Therefore, this mutation is a key determinant in modulating NDV P-gene mRNA editing by influencing the overall frequency and pattern of mRNA editing. DISCUSSION : The V protein of NDV plays a crucial role both in virus replication and pathogenicity. Earlier studies demonstrated that NDV mutants lacking both V and W proteins or lacking only the unique cysteine-rich C-terminal region of the V protein produced 5,000-fold-less infectious progeny in cell culture than that produced by the parent virus. Surprisingly, the mutants were completely unable to propagate in 9- to 11-day-old embryonated chicken eggs . These findings demonstrated that the V protein of NDV is indispensable for virus replication, although P-gene-derived accessory proteins of Paramyxoviridae were categorized as nonessential gene products . Another mutant, NDV-P1, possessing a single-nucleotide substitution in the 5'-A5 run was, in contrast, viable in embryonated eggs. This mutation resulted in downregulation of V protein expression instead of complete abrogation . NDV-P1 was dramatically attenuated for chicken embryos with an ELD50 of >6.7 log10 EID50 compared to 0.3 log10 EID50 for the parent virus. The involvement of V protein in virulence is presumably associated with its ability to counteract the antiviral effects of interferon, which is encoded by the carboxy terminus of V . The studies described in this report show how editing-defective NDVs restore P-gene mRNA editing and hence secure expression of V and presumably W protein to overcome the selective pressure placed on them for better replication in embryonated eggs. The mutants NDV-P1 and NDV-PC4 possess single-nucleotide substitutions at the 5'-A3 and -A4 positions, respectively, in the 5'-A5G3 editing-site sequence. The effects of the mutations in NDV-P1 and NDV-PC4 at low passage levels were very dramatic resulting in an at least 100-fold titer decrease compared to that of the parent virus. In addition, both mutants were completely nonpathogenic for 9- to 11-day-old chicken embryos even at high doses. All of the sequenced independent clones derived from low passages of both mutants represented unedited P mRNA, demonstrating that the edited V and W mRNAs are below the detection limit. Within a total of seven passages, the mutants showed restored efficiency of replication and increased pathogenicity for chicken embryos. The proportion of edited mRNAs augmented from an undetectable level (0%) at low passages to 6 and 4% for NDV-P1 and NDV-PC4, respectively, at passage 7. In NDV-P1 passage 7, although not in NDV-PC4, both V and W mRNAs were detected, demonstrating restoration of P-gene mRNA editing, albeit at a low level. The possibility that W protein may play a role in restoring virus replication and virulence could not be ruled out from these experiments. However, in previous experiments, recombinants either lacking V and W proteins or lacking only the C terminus of V protein exhibited similar defect in replication and virulence . This finding strongly suggested that, V rather than W is responsible for restoring virus virulence and presumably virus replication. The striking result of the studies described in this report is the identification of a second-site, compensatory mutation capable of modulating mRNA editing. The proportion of the progeny virions of NDV-P1 and NDV-PC4 that incorporated the C-U2278 mutation increased from 0% at low passages to 86 and 89%, respectively, at passage 7. The occurrence of an identical second-site mutation in both recombinants strongly indicates that the C-U2278 substitution is essential to compensate for the loss of function caused by disruption of the A-stretch. Introduction of the C-U2278 mutation into the backbone of the parent virus provided a direct evidence for the key role of this second-site mutation in modulating the editing process. The mutant NDV-T2 edited 85% of its P-gene derived mRNAs as opposed to only 32% for the parent virus. This mutation not only raised the total amounts of edited mRNAs but also considerably altered the editing pattern of P, V, and W mRNAs from 68, 28, and 4% to 15, 44, and 41%, respectively. Although P proteins of negative-strand RNA viruses are generally known to be required for efficient virus replication, it is surprising that NDV-T2 is able to replicate to the level of the parent virus despite a 4.5-fold-lower level of P mRNA. The fact that P, V, and W proteins differ only at their C terminus, as well as the finding showing the involvement of V protein in virus replication, suggest that V and/or W proteins might at least partially compensate for the functions of P protein. To provide further insight for the requirement of P protein in NDV replication, we designed a full-length cDNA that should produce only edited mRNAs but not P mRNA. Despite repeated recovery attempts, we failed to rescue an infectious virus (data not shown), suggesting that P protein is not entirely dispensable for NDV replication or completely replaceable by V and W proteins. Similar to NDV, ca. 30% of Sendai virus (SeV) mRNAs are edited by mostly inserting a single G residue. The effects of introducing alterations into the sequence around the SeV P-gene mRNA editing site (5'-AACA6G3) have been extensively studied . In a recombinant SeV in which the 5'-A-run was shortened from six to five nucleotides, no mRNA editing could be detected. However, substitutions of the upstream nucleotides of the 5'-AACA5G3 sequence from AAC to AAU or AUU restored editing activity to 24 or 44%, respectively. In the latter case, significant fractions of the mRNAs contained two to six G insertions. Hausmann et al. also provided experimental evidence that the insertions of the nontemplated G residues occur by a stuttering process. Furthermore, approximately six nucleotides upstream of the editing site, particularly the positions more proximal to the 5'-A6G3 sequence, were shown to determine the number of times the polymerase stutters before normal transcription resumes . Our data are thus consistent with these findings, showing that the nucleotides immediately upstream of the editing site are key determinants in modulating the frequency and pattern of insertion of pseudotemplated G residues. Moreover, based on sequence comparison around the editing site of different paramyxoviruses, the NDV and SeV are the only known paramyxoviruses that do not have a U residue at the second nucleotide upstream of the 5'-A5G3 or 5'-A6G3 sequences, respectively. As shown for NDV-T2, the presence of U at position 2278 increased the level of edited mRNAs and altered the pattern of editing. It would be of interest to determine whether other members of the subfamily Paramyxovirinae produce high percentage of edited transcripts in a pattern similar to that of NDV-T2. The only difference between NDV-T2 and the parent virus is the formation of unique base pairings in the upstream sequence. As shown in Fig. , the unique upstream C:U/A:U pairing in NDV-T2 is more efficient than the corresponding C:C/G:U pairing in rNDV in promoting one G insertion. For inserting two G's, the upstream 3'-ACA template strand of NDV-T2 would require to base pair with 5'-UUA, whereas the corresponding pairing in the parent rNDV would be between 3'-ACG and 5'-CUA. The result is generation of 41% W mRNA (with two G insertion) in NDV-T2 as opposed to only 4% W mRNA in the parent virus . It appears that the nature and most likely the stability of the mRNA-template hybrid in the most upstream 5'-A5G3 sequence of NDV is a key determinant in controlling how many times the polymerase stutters and when to resume normal transcription after an editing event. However, the editing process is less efficient in the absence of an intact A-stretch at the editing site, suggesting that mutations within the uridine-rich region of the negative-strand genome may induce secondary structure formation that prevent or weaken specific interaction with the transcription-editing complex. Therefore, nucleotide insertion and editing-site recognition might be two distinct processes that cooperatively regulate P-gene mRNA editing. Further studies should reveal the exact mechanism involved in the editing of paramyxoviruses. FIG. 1. : Recombinant NDV constructs. Recombinant NDV constructs. A schematic representation of the NDV gene order is shown in the negative-strand genomic RNA. The nucleotide sequences around the editing site (positions 2274 to 2291) are presented in a positive sense (5' to 3'). The introduced nucleotide modifications (lowercase letters) and amino acid substitutions (one-letter code) in the P proteins of mutant viruses are shown in boldface. FIG. 2. : Infectious titers of recombinant NDVs. Infectious titers of recombinant NDVs. Embryonated chicken eggs were inoculated at a dose of 4 log10 EID50 per egg and incubated for 4 days. Tenfold serial dilutions of the harvested allantoic fluid samples were prepared and inoculated into 9-to 11-day-old embryonated eggs (10 eggs/dilution). After 3 to 4 days of incubation, an HA test was carried out, and the calculated infectious titers in the log10 EID50/ml are shown on the top of the bars. FIG. 3. : Pathogenicity of serially passaged NDV-P1 in SPF chicken embryos. Pathogenicity of serially passaged NDV-P1 in SPF chicken embryos. Embryonated eggs were inoculated with the parent virus at passage 3 (rNDV) or the mutant NDV-P1 at passage 5 (P1-5), passage 6 (P1-6), or passage 7 (P1-7) and then incubated for 7 days or until the embryos had died. NDV-P1, which was completely safe for 11-day-old chicken embryos at passage 5, became nearly as virulent as the parent virus at passage 7. FIG. 4. : Pattern of P-gene mRNA editing in cells infected with serially passaged NDV-P1. Pattern of P-gene mRNA editing in cells infected with serially passaged NDV-P1. The sequences around the editing site from independent clones derived from rNDV or NDV-P1 at passage 5 (P1-5), passage 6 (P1-6), or passage 7 (P1-7) were determined. The percentages of unedited mRNAs representing P ORF (P) or edited mRNAs with insertion of one G residue coding for V ORF (V) or with insertions of two G residues coding for W ORF (W) are shown on the top of the bars. FIG. 5. : (A) Infectious titers of recombinant NDV-PC4 at passages 3 and 7. (A) Infectious titers of recombinant NDV-PC4 at passages 3 and 7. Titers were determined as described under the legend for Fig. . (B) Pathogenicity of NDV-PC4 at passage 3 (PC4-3) and passage 7 (PC4-7) in SPF chicken embryos. Mortality was determined as described in the legend for Fig. . (C) Editing frequency of NDV-PC4 at passages 3 (PC4-3) and passage 7 (PC4-7). mRNA editing was determined from independent clones as described under the legend for Fig. . FIG. 6. : Comparison of putative mRNA-template hybrids of rNDV and NDV-T2 during the editing process. Comparison of putative mRNA-template hybrids of rNDV and NDV-T2 during the editing process. The template negative-strand genomes (top strands) were written 3' to 5', and the mRNA chains (bottom strands) were written 5'to 3'. The editing frequency of both viruses for P, V, and W mRNAs are shown in percentage. The only difference between the two viruses at position 2278 (C-U2278) was shown in boldface lowercase. The unique putative base pairings are underlined. As much as 85% of P-gene derived mRNAs in NDV-T2 are edited, whereas only 32% of the mRNAs in the parent virus represent V and W mRNAs. TABLE 1 : Occurrence of a second-site compensatory C-U2278 mutation in editing-site mutants at various passages Backmatter: PMID- 12915549 TI - Herpes Simplex Virus Type 2 Virion Host Shutoff Protein Regulates Alpha/Beta Interferon but Not Adaptive Immune Responses during Primary Infection In Vivo AB - The herpes simplex virus (HSV) virion host shutoff (vhs) protein, the product of the UL41 (vhs) gene, is an important determinant of HSV virulence. vhs has been implicated in HSV interference with host antiviral immune responses, down-regulating expression of major histocompatibility complex molecules to help HSV evade host adaptive immunity. The severe attenuation of vhs-deficient viruses in vivo could reflect their inability to escape immune detection. To test this hypothesis, BALB/c or congenic SCID mice were infected intravaginally (i.vag.) with the HSV type 2 (HSV-2) vhs null mutant 333d41 or the vhs rescue virus 333d41R. vhs-deficient virus remained severely attenuated in SCID mice compared with rescue virus, indicating that vhs regulation of adaptive immune responses does not influence HSV pathogenesis during acute infection. Innate antiviral effectors remain intact in SCID mice; prominent among these is alpha/beta interferon (IFN-alpha/beta). The attenuation of HSV-2 vhs mutants could reflect their failure to suppress IFN-alpha/beta-mediated antiviral activity. To test this hypothesis, 129 and congenic IFN-alpha/beta receptor-deficient (IFN-alpha/betaR-/-) mice were infected i.vag. with wild-type virus, vhs null mutants 333-vhsB or 333d41, or the vhs rescue virus 333d41R. Whereas vhs-deficient viruses showed greatly reduced replication in the genital mucosa of 129 mice compared with wild-type or vhs rescue viruses, they were restored to nearly wild-type levels of replication in IFN-alpha/betaR-/- mice over the first 2 days postinfection. Only wild-type and vhs rescue viruses caused severe genital disease and hind limb paralysis in 129 mice, but infection of IFN-alpha/betaR-/- mice restored the virulence of vhs-deficient viruses. vhs-deficient viruses replicated as vigorously as wild-type and rescue viruses in the nervous systems of IFN-alpha/betaR-/- mice. Restoration was specific for the vhs mutation, because thymidine kinase-deficient HSV-2 did not regain virulence or the capacity to replicate in the nervous systems of IFN-alpha/betaR-/- mice. Furthermore, the defect in the IFN-alpha/beta response was required for restoration of vhs-deficient virus replication and virulence, but the IFN-alpha/beta-stimulated protein kinase R pathway was not involved. Finally, vhs of HSV-2 has a unique capacity to interfere with the IFN-alpha/beta response in vivo, because an HSV-1 vhs null mutant did not recover replication and virulence after i.vag. inoculation into IFN-alpha/betaR-/- mice. These results indicate that vhs plays an important role early in HSV-2 pathogenesis in vivo by interfering with the IFN-alpha/beta-mediated antiviral response. Keywords: Introduction : Herpes simplex virus types 1 and 2 (HSV-1 and HSV-2) are ubiquitous human pathogens that cause a variety of serious diseases. Primary HSV infections typically initiate at mucosal surfaces, where lytic replication in epithelial cells generates mucosal lesions. Peripheral replication amplifies the virus load and increases uptake of virus into sensory nerve termini. Establishment of latent infections in sensory neurons secures shelter for the virus throughout the life of the host. Periodic reactivations result in disease recurrence and provide an opportunity for transmission to new hosts. The virion host shutoff (vhs) protein of HSV plays a significant role in promoting pathogenesis at the cell and organismal levels. In the infected cell, vhs possesses both endo- and exonuclease activity and mediates the rapid shutoff of protein synthesis via degradation of both cellular and viral mRNAs . As a component of the virion tegument, vhs is released directly into the cytosol and can immediately exert its effects on the newly infected cell. Cellular mRNAs are almost completely degraded within 6 h of infection with HSV-1, and within just 2 h by HSV-2 . The activity of HSV-2 vhs is also approximately 40-fold stronger than that of HSV-1 . In vivo, vhs null mutants of HSV-1 and HSV-2 are profoundly attenuated, implicating vhs as a virulence factor that helps HSV establish robust infection. HSV-1 lacking vhs activity replicates to 1,000-fold-lower titers in the cornea, trigeminal ganglia, and brain of mice than wild-type virus and has impaired capacity to enter the central nervous system, establish latency, and undergo reactivation (-). vhs-deficient HSV-2 strains (333-vhsB or 333d41) also replicate much less efficiently than wild-type virus in the genital mucosa and nervous tissue of mice and cause less disease . vhs has been implicated in down-regulating major histocompatibility complex (MHC) class I and class II molecules. This activity has functional implications because it is associated with reduced cytotoxic T-lymphocyte recognition of HSV-infected cells . Thus, vhs-deficient viruses may be unable to evade critical adaptive immune responses during acute infection, leading to their attenuation. Alternatively, the attenuation of HSV-1 and HSV-2 vhs null mutants within 24 to 48 h postinfection (-) suggests a pivotal role for mediators of innate immunity in controlling acute infection. A key mediator of innate antiviral immunity to virus infection is the alpha/beta interferon (IFN-alpha/beta) response. The many subtypes of IFN-alpha and IFN-beta are released from infected cells and bind to a single IFN-alpha/beta receptor (IFN-alpha/betaR). Receptor-mediated signal transduction stimulates expression of numerous IFN-stimulated genes whose products can interfere with viral replication by disrupting macromolecular synthesis. HSV-1 in culture is relatively insensitive to the effects of IFN-alpha/beta, being reduced 2- to 5-fold for replication, compared to a more than 10,000-fold reduction for vesicular stomatitis virus . In addition, HSV-1 virulence is not substantially increased in mice lacking the IFN-alpha/betaR (IFN-alpha/betaR-/-) , suggesting that HSV has a means to effectively counter the IFN-alpha/beta response. Determinants of HSV-1 IFN resistance have been mapped to ICP0, Us11, and gamma34.5 . Each of these proteins has been confirmed as a contributor to HSV-1 virulence (, , ; S. Ward and D. Leib, personal communication). Evidence for HSV-1 vhs as an additional inhibitor of the IFN-alpha/beta response is controversial and is not borne out in vivo . Replication of an HSV-1 vhs null mutant is slightly elevated and prolonged in the corneas of IFN-alpha/betaR-/- mice compared to wild-type mice. However, the titer of vhs-deficient HSV-1 is 1,000-fold lower than that of wild-type virus at all times in IFN-alpha/betaR-/- mice , indicating that HSV-1 vhs does not play a significant role in counteracting IFN-alpha/betaR signaling in vivo. Mediators of IFN resistance in HSV-2 have not been well characterized, but several observations suggest that vhs may contribute in this case. Genetic studies have mapped IFN resistance to a 7.4-kb region of the HSV-2 genome that contains the vhs gene and portions of two flanking genes . Another genetic locus influencing HSV-2 resistance to IFN-alpha/beta has been mapped 60 kb distant from the vhs locus . This region contains the UL13 viral kinase, which influences either vhs synthesis or activity . In addition to genetic evidence, HSV-2 is even more resistant to IFN-alpha/beta in vitro than HSV-1 , and HSV-2 vhs activity is more potent than that of HSV-1 vhs (-, ). These observations led us to investigate whether HSV-2 vhs interferes with the host IFN-alpha/beta response. Here we use a mouse model of genital infection with HSV-2 to examine the role of vhs in countering host adaptive immune and innate responses to primary infection. MATERIALS AND METHODS : Cells and viruses. | Vero cells were maintained in Dulbecco's modified Eagle's medium supplemented with 3% newborn and 3% fetal calf serum. SB5, a plaque-purified stock of HSV-2 strain 333, was obtained from the American Type Culture Collection (VR-2546). HSV-1 strain UL41NHB and HSV-2 strain 333d41 were obtained from David Leib. The vhs open reading frame (ORF) in UL41NHB is disrupted by insertion of the Escherichia coli lacZ gene under the control of the human cytomegalovirus immediate-early 1 (IE1) promoter at a ScaI site. 333d41 has a targeted deletion within the UL41 gene between two XcmI sites, resulting in a 939-bp excision and complete lack of vhs activity. 333d41R was constructed by cotransfecting full-length 333d41 DNA with p41SB5-B, a plasmid containing a BglII genomic fragment of HSV-2 333 that includes the UL41 gene . Briefly, 1 mug of each DNA mixed with Lipofectamine (Invitrogen) was used to transfect Vero cells. Viral progeny were screened by PCR for the deletion within vhs by using the primers 5'-GCG TCC AAC CGA TAA ATC AAG C-3'(forward) and 5'-TTA CCC AAA AGT CCC TGT TCC C-3' (reverse), with 30 cycles of 95C for 30 s, 66.1C for 30 s, and 72C for 60 s. PCR products were analyzed on agarose gels stained with ethidium bromide. Isolates were plaque purified three times, and identity of the rescue virus was confirmed by Southern blot analysis using a BamHI digest (data not shown). The RNase activity of 333d41R was confirmed by Northern blot analysis in an mRNA degradation assay described elsewhere and was identical to that of SB5 (data not shown). HSV-2 strains 333-vhsB and DeltaTK- were the generous gifts of Jim Smiley. 333-vhsB contains a cassette consisting of the ICP6 promoter driving the lacZ gene . The cassette was inserted at codon 30 of the vhs ORF, disrupting the UL41 gene. DeltaTK- contains a 180-bp deletion in the thymidine kinase (tk) ORF of strain 333 that inhibits tk activity. Cell lysate stocks of each strain were prepared as previously described . Titers of virus stocks were determined by simultaneous titration on Vero cell monolayers to ensure equivalent doses of virus in the inocula. Animals and infections. | BALB/c mice were purchased from the National Cancer Institute (Frederick, Md.). CB-17.SCID mice were purchased from Taconic (Germantown, N.Y.). IFN-alpha/betaR-/- and congenic 129 [129Sv(ev)] mice were generously provided by Michel Aguet and were obtained from Skip Virgin. Protein kinase R-deficient (PKR-/-) mice on the 129 background were generously provided by Bob Silverman and were obtained from Skip Virgin. 129, IFN-alpha/betaR-/-, and PKR-/- mice were bred and housed in the Department of Comparative Medicine, Saint Louis University School of Medicine. All mice were maintained in accordance with institutional and Public Health Service guidelines and were used at 6 weeks of age. Vaginal swab and neural tissue titers. | Vaginal vaults were swabbed twice at 10 h and days 1 through 4 postinfection using type 1 calcium alginate swabs (Puritan). The swabs were placed together into vials containing 1 ml of phosphate-buffered saline and stored at -80C. At 6 days postinfection, mice were sacrificed and the spinal cord, brainstem, and brain were dissected, placed in microcentrifuge tubes containing phosphate-buffered saline and 1-mm glass beads, and stored at -80C. Thawed tissues were disrupted using a Mini BeadBeater 8 (Biospec Products). Viral titers in swab samples and tissues were determined by standard plaque assay . Disease scores. | Severity of genital and neurologic disease was assessed on days 2 through 6 postinfection using the following scale: 0, no signs; 1, mild erythema and edema of the external genitalia; 2, moderate erythema and edema of the external genitalia; 3, severe erythema and edema of the external genitalia with presence of lesions; 4, bilateral hind limb paralysis; 5, death. Statistics. | Significance of difference in viral titer between groups on individual days was determined by the t test. The nonparametric Kruskall-Wallis test was used to determine significance of difference in disease scores between groups on individual days. RESULTS : vhs-deficient mutants of HSV-1 and HSV-2 replicate poorly compared to wild-type virus in the mouse cornea and genital mucosa, respectively. They also cause minimal disease and have reduced capacity to replicate in the nervous system . We reasoned that if adaptive immune responses are responsible for the attenuation of vhs-deficient viruses, then replication and virulence should be restored in mice lacking adaptive immune functions. To test this hypothesis, BALB/c and congenic CB-17.SCID mice were infected i.vag. with 2 x 106 PFU of the vhs-deficient mutant 333d41 or the vhs rescue virus, 333d41R. Titers of vhs rescue virus collected from daily vaginal swabs of BALB/c and SCID mice were significantly higher than those of 333d41 on days 2 to 4 postinfection . Correspondingly, BALB/c mice infected with vhs rescue virus rapidly developed severe genital inflammation, and lesions were present by 6 days postinfection . Inflammation provoked by 333d41 infection was comparatively moderate. The same dichotomy was observed in SCID mice, where rescue virus caused pronounced inflammation but vhs-deficient virus did not . Interestingly, genital inflammation was more severe in BALB/c mice than SCID mice, suggesting an immunopathologic component to the inflammatory response. Virus titer in the nervous system was determined on day 6 postinfection. vhs rescue virus replicated to high titers in the spinal cord of BALB/c mice and, as anticipated, levels of 333d41 were much lower . Brainstem and brain tissue of BALB/c mice contained less rescue virus, but 333d41 could not be detected. In SCID mice, rescue virus replicated to high titer in the central nervous system, but titers of 333d41 remained low . A second vhs-deficient virus, 333-vhsB, also remained attenuated in SCID mice (data not shown). These results indicate that adaptive immune responses were not responsible for the reduced capacity of vhs-deficient virus to replicate or cause disease and, thus, that vhs does not play a significant role in counteracting the adaptive immune response during acute infection. SCID mice lack T and B lymphocytes, but they possess normal innate immune mechanisms. A key mediator of innate antiviral responses is IFN-alpha/beta. Although vhs of HSV-1 does not appear to influence IFN-mediated antiviral resistance in vivo , three observations suggested that HSV-2 may utilize vhs to counteract the IFN-alpha/beta response. IFN sensitivity of a wild-type HSV-2 strain has been mapped to a region that includes the vhs locus , a UL13- HSV-2 strain with reduced vhs activity permits more IFN production in vivo , and HSV-2 vhs activity is much stronger and faster than that of HSV-1 . We therefore examined the pathogenesis of vhs-deficient HSV-2 strains in mice genetically deficient in the receptor for all IFN-alpha/beta (IFN-alpha/betaR-/-). IFN-alpha/betaR-/- and wild-type 129 mice were infected i.vag. with 2 x 106 PFU of 333 (clone SB5), 333d41, or vhs rescue virus 333d41R. In 129 mice, 333d41 replicated approximately 100-fold less efficiently in the genital mucosa at all time points compared to wild-type and vhs rescue viruses . In IFN-alpha/betaR-/- mice, replication profiles of SB5 and 333d41R were similar to those seen in wild-type 129 mice over the first 2 days postinfection but were maintained at higher levels thereafter . Strikingly, titers of vhs-deficient 333d41 were increased 20- to 50-fold at all time points in IFN-alpha/betaR-/- mice compared with 129 mice (P < 0.0001). Levels of 333d41 replication in IFN-alpha/betaR-/- mice were not statistically different from levels of wild-type and rescue viruses over the first 2 days postinfection . These data indicate a significant recovery of replication by vhs-deficient virus in the vaginal mucosa of mice lacking the capacity to mount an IFN-alpha/beta-mediated antiviral response. 333-vhsB was slightly more attenuated than 333d41 in 129 mice but also showed significantly enhanced replication in the genital mucosa of IFN-alpha/betaR-/- mice (data not shown). Signs of genital disease in 129 mice infected with wild-type and vhs rescue viruses became readily apparent by 4 days postinfection and were severe 5 to 6 days postinfection . Most mice developed hind limb paralysis by day 6 postinfection. The disease course in IFN-alpha/betaR-/- mice infected with SB5 or 333d41R was similar to that in 129 mice, although accelerated by approximately 1 day . Whereas infection of 129 mice with vhs-deficient virus had resulted in mild genital inflammation , IFN-alpha/betaR-/- mice showed severe genital disease (P = 0.013 and 0.001 on days 5 and 6, respectively) that developed with approximately the same kinetics as disease caused by wild-type and vhs rescue viruses . Furthermore, signs of neurologic disease were equally apparent in IFN-alpha/betaR-/- mice infected with vhs-deficient, wild-type, or vhs rescue viruses. These data indicate that the virulence of vhs-deficient virus is markedly increased in IFN-alpha/betaR-/- mice relative to that in wild-type virus. At 6 days postinfection, the titers of 333d41 were significantly lower than those of wild-type and vhs rescue viruses in spinal cord and brainstem and could not be detected in the brain of 129 mice . In contrast, 333d41 replicated to levels equivalent to wild-type and rescue viruses in the spinal cord and brainstem of IFN-alpha/betaR-/- mice (Fig. ; P < 0.0001 compared to 129 mice for both tissues). Only the brain contained lower titers of 333d41 than the wild-type and rescue viruses. Thus, replication of vhs-deficient virus is largely restored in the nervous systems of IFN-alpha/betaR-/- mice. We observed in these experiments that the 333d41R virus had a slightly less virulent phenotype than the wild-type 333 derivative SB5. Specifically, development of genital disease was moderately delayed in IFN-alpha/betaR-/- mice , and titers of 333d41R were lower than those of SB5 in the brains of 129 mice (P = 0.005) . 333d41R possesses wild-type vhs activity, as determined by degradation of cellular glyceraldehyde-3-phosphate dehydrogenase mRNA (data not shown). 333-vhsB, 333d41, and 333d41R viruses were constructed from wild-type HSV-2 strain 333, whereas SB5 is a clonal isolate of 333. It is possible that a subtle difference in the virulence of 333 and SB5 exists, perhaps due to clonal selection or different passage histories. Alternatively, a secondary mutation could have arisen in the construction of the 333d41R virus. In preliminary studies, 333-vhsBR, a different rescue virus which replicates like SB5 in the cornea , was equivalent to 333d41R in the genital model (data not shown). This result suggests that 333d41R does not contain an additional attenuating mutation relative to 333-vhsB and, instead, it implies that a difference in the pathogenicity of SB5 and 333 is discriminated in the vaginal model of HSV infection. We have taken the conservative approach of comparing 333d41 to the 333d41R rescue virus rather than to SB5 to evaluate statistical significance of differences resulting from the vhs mutation. To rigorously demonstrate vhs-specific phenotypic restoration of HSV-2 vhs mutants in IFN-alpha/betaR-/- mice, groups of 129 and IFN-alpha/betaR-/- mice were infected with a 333 strain containing a deletion in the tk gene, HSV-2 DeltaTK- . tk is dispensable for HSV replication in dividing cells but not in neurons , and tk-deficient viruses are less virulent than wild-type virus . As expected, replication of DeltaTK- in the genital mucosa of 129 mice and IFN-alpha/betaR-/- mice was equivalent to that of wild-type and rescue viruses , and genital inflammation in 129 mice was mild . In distinct contrast to 333d41, the virulence of DeltaTK- in IFN-alpha/betaR-/- mice was not restored to wild-type levels , and DeltaTK- could be detected only at low levels in the nervous systems of IFN-alpha/betaR-/- mice . These observations indicate that the enhanced capacity of 333d41 to replicate and cause disease in IFN-alpha/betaR-/- mice was specific to the vhs lesion. Another criterion of specificity for viral inhibitors of host proteins involved in key antiviral pathways is the demonstration that recovery of virulence occurs only in the absence of the particular host protein. This criterion may be difficult to meet for a viral protein such as vhs that likely employs the relatively nonspecific mechanism of mRNA degradation to block host antiviral responses. To begin to dissect IFN-stimulated antiviral pathways inhibited by vhs, as well as to evaluate specificity, we focused on PKR. PKR, one of the best-characterized IFN-alpha/beta-stimulated genes, directs the shutoff of host cell protein synthesis. We investigated the role of HSV-2 vhs in inhibition of the PKR pathway by using mice that lack PKR (PKR-/-). Interestingly, we found that replication in the genital mucosa and virulence of 333d41 were not restored in PKR-/- mice relative to wild-type virus. Titers of vhs-deficient virus were higher in PKR-/- mice than in wild-type 129 mice only in the spinal cord (Fig. ; P = 0.0016). The PKR-deficient phenotype of PKR-/- mice was confirmed by the capacity to restore replication of the gamma34.5-deficient HSV-1 strain, 17termA (data not shown). These results indicate that vhs does not counteract the antiviral activities of PKR and that restoration of replication and virulence of vhs-deficient virus is not a universal phenomenon in mice with genetic defects (even defects in a specific IFN-alpha/beta-induced antiviral pathway). vhs instead must interfere with a separate antiviral mechanism stimulated by IFN-alpha/betaR engagement. vhs-deficient HSV-1 does not regain the capacity to replicate and cause disease in the corneas of IFN-alpha/betaR-/- mice , indicating that HSV-1 vhs is not an inhibitor of the IFN-alpha/beta response in the eye. We therefore investigated whether a unique property of genital epithelial cells or of HSV-2 vhs resulted in HSV-2 interference with the IFN-alpha/beta response in the genital tract. As shown in Fig. , the HSV-1 vhs null mutant UL41NHB did not replicate efficiently in the genital mucosa of 129 mice after i.vag. inoculation. Titers of UL41NHB were only slightly elevated in IFN-alpha/betaR-/- mice and were not increased relative to those of wild-type HSV-1 . Virulence of wild-type virus was enhanced by infection of IFN-alpha/betaR-/- mice, but vhs-deficient HSV-1 remained attenuated . Neither virus was detectable in the nervous system (data not shown). Thus, recovery of a vhs mutant's capacity to replicate and cause disease similar to wild-type virus in the absence of an IFN-alpha/beta response is unique to HSV-2 vhs and is not a general property of the i.vag. route of infection. DISCUSSION : The vhs protein clearly contributes to HSV virulence in the host, because vhs-deficient mutants of HSV are highly attenuated in various mouse models of HSV infection . One hundred- to 1,000-fold less replication occurs in the genital mucosa after infection with vhs-deficient HSV-2, and little virus reaches the nervous system. We have made several observations regarding the role of HSV-2 vhs in pathogenesis that largely explain the attenuation of vhs-deficient viruses. First, we determined that adaptive immunity is not responsible for attenuation of vhs-deficient HSV-2 or HSV-1 during acute infection by demonstrating that the mutant viruses remain attenuated during primary infection of SCID mice. Second, we have shown that vhs interferes with an aspect(s) of the IFN-alpha/beta response, because replication and especially virulence of vhs-deficient HSV-2 are significantly increased relative to wild-type and rescue viruses in IFN-alpha/betaR-/- mice. Third, inhibition of the IFN-alpha/beta response is a unique property of the HSV-2 vhs protein, because vhs-deficient HSV-1 replication and virulence do not improve relative to wild-type HSV-1 after i.vag. (Fig. to ) or corneal infection of IFN-alpha/betaR-/- mice. Fourth, vhs does not assist HSV-2 evasion of the IFN-alpha/beta response by interfering with the PKR pathway. The IFN-alpha/beta response rapidly constructs a formidable barrier to virus infection; the dramatic effect of IFN-alpha/beta on the pathogenesis of a virus without vhs as a countermeasure attests to the significance of HSV-2 vhs as a modulator of this innate host defense mechanism. In order to prove specificity in the restoration of the wild-type phenotype to vhs-deficient viruses in IFN-alpha/betaR-/- mice, it was important to demonstrate that (i) an attenuated viral strain with a mutation in a gene irrelevant to IFN-alpha/beta remains attenuated in IFN-alpha/betaR-/- mice, and (ii) the vhs mutant virus is not restored in a host with a different genetic lesion . We have satisfied the first of these specificity criteria with the tk- mutant virus, DeltaTK-. We observed that DeltaTK- replicated poorly in the nervous systems of 129 mice and, unlike 333d41, did not replicate well in the nervous systems of IFN-alpha/betaR-/- mice, even with an inoculum of 2 x 106 PFU. Therefore, restoration was specific to the vhs mutation. Leib et al. observed that a tk- HSV-1 mutant inoculated at the same dose replicated more efficiently in the corneal epithelium of IFN-alpha/beta/gammaR-/- mice than wild-type mice but was barely detectable in the nervous system through 5 days postinfection. Taken together, these data suggest that loss of alpha/beta or alpha/beta/gamma IFN receptors does not globally restore the capacity of HSV mutants to replicate in the nervous system. Infection of PKR-/- mice satisfied the second criterion for specificity in that 333d41 was not restored for replication or virulence. While the PKR mutation is not independent of the IFN-alpha/beta receptor signaling pathway, our data indicate that this separate host genetic mutation does not restore vhs-deficient viruses to virulence and, therefore, restoration is specific to the IFN-alpha/betaR mutation. The fact that attenuation of 333d41 was not reversed by deletion of the host PKR gene is intriguing. This result suggests that although PKR mediates an important antiviral pathway that is induced by IFN-alpha/betaR engagement, PKR is not the primary target of vhs-mediated host inhibition. HSV already encodes two proteins, US11 and ICP34.5, demonstrated to block PKR-mediated protein synthesis inhibition at different stages . With two specific inhibitors of the PKR pathway, it may not be necessary for HSV to encode yet another inhibitor. Lack of PKR pathway disruption also suggests that vhs may discriminately inhibit host antiviral functions. However, if vhs utilizes its RNase activity to inhibit inducible mediators of the IFN-alpha/beta response, it may have a limited capacity to block the PKR pathway due to preformed PKR in cells that can be directly activated by virus infection . Determining which IFN-related antiviral pathways are opposed by vhs and the mechanisms of vhs action will be the focus of future experiments. If the mechanism of IFN-alpha/beta pathway inhibition by vhs does involve RNA degradation, as seems likely, vhs could be unique as a general, relatively nonspecific inhibitor of the inducible IFN-alpha/beta response. In this manner, vhs could be broadly useful to the virus across cell types and tissues that may vary in their dependence on individual IFN-induced antiviral pathways. Nevertheless, some evidence for tissue specificity in HSV-2 vhs interference with the IFN-alpha/beta response may have arisen in our present experiments. Restoration of vhs-deficient HSV-2 replication in the genital mucosa of IFN-alpha/betaR-/- mice was not comprehensive because titers were lower than those for wild-type virus at 3 and 4 days postinfection. Despite reduced mucosal replication, vhs-deficient virus replicated as vigorously as wild-type virus in the spinal cord and brainstem. Thus, differences may exist in the particular antiviral pathways stimulated by IFN-alpha/betaR engagement in specific tissues, and some of these pathways, for example, PKR, may be less inhibited by vhs than others. Such cell-type-specific differences in IFN induction have been noted in reovirus infections of cardiac myocytes versus skeletal muscle cells . In the genital mucosa, NK cells, neutrophils, or even early antiviral T-cell responses may also help reduce vhs-deficient virus replication after 48 h. We are currently investigating the reason for the later decline in vhs-deficient virus titer in the mucosa of IFN-alpha/betaR-/- mice relative to that with wild-type virus. Production of IFN-alpha/beta and IFN-stimulated gene induction in the genital mucosa will be interesting areas of future investigation. Our observation that HSV-2 vhs counteracts the IFN-alpha/beta response fits logically into a picture of HSV-2 pathogenesis for three reasons. First, vhs enters the cytoplasm with the infecting virion and, thus, can immediately counter rapid IFN synthesis or receptor-mediated induction of the antiviral state. Indeed, IFN-alpha can suppress immediate-early gene expression , so the virus must have an immediate counterresponse. Second, because vhs degrades mRNA, it could counteract a wide variety of host responses essential to induction of the antiviral state, including IFN-beta and -alpha4 produced by the initially infected cells. Finally, HSV-2 is even more resistant to IFN-alpha/beta than HSV-1 , and HSV-2 vhs is faster and approximately 40-fold more active than HSV-1 vhs . An effect of HSV-2 vhs on the host IFN-alpha/beta response therefore should be more readily discernible than an effect of HSV-1 vhs. Our results highlight an important difference in the pathogenic mechanisms of HSV-2 and HSV-1. We observed that the presence or absence of functional lymphocytes emphatically does not alter the pathogenesis of either HSV-1 (data not shown) or HSV-2 vhs-deficient viruses in a naive host. This observation indicates that the adaptive immune response is not responsible for attenuating the pathogenesis of viruses that lack vhs activity, at least during a rapidly progressing primary infection in BALB/c mice. Because the IFN-alpha/betaR-/- mice used had a 129 genetic background, it is formally possible that vhs could interfere with primary adaptive immune responses in 129 mice. However, the generally greater resistance of 129 mice to HSV infection should make recovery of virulence by vhs-deficient virus more challenging. Despite observations that vhs interferes with the expression of MHC class I and class II molecules , our result was not entirely surprising given that replication of vhs-deficient viruses in the genital mucosa of wild-type mice is curtailed within 48 h. Nevertheless, vhs could play a role in adaptive immune responses during reactivation in an immune host, when evasion of memory cytotoxic T-lymphocyte responses would confer a significant advantage to the virus. Our results also do not preclude the possibility that vhs retards the induction of adaptive immune responses through its interference with MHC expression and potentially through its inhibition of the IFN-alpha/beta response. IFNs induce the activation and maturation of dendritic cells , the most potent antigen-presenting cell type for naive T cells. IFN-alpha/beta thus establishes an important link between early innate responses and developing antiviral adaptive immunity. If less IFN-alpha/beta is produced by wild-type virus-infected cells than by cells infected with vhs-deficient virus, the adaptive immune response to wild-type virus may be delayed or reduced in magnitude. We are currently investigating this possibility. In any event, our observation that vhs counteracts the innate antiviral IFN-alpha/beta response during acute infection, coupled with its capacity to down-regulate MHC molecules crucial in adaptive immune responses, mark vhs as a multifunctional immune evasion mechanism of HSV-2. FIG.1. : Pathogenesis of vhs-deficient HSV-2 is not restored in SCID mice. Pathogenesis of vhs-deficient HSV-2 is not restored in SCID mice. Groups of four BALB/c and congenic CB.17-SCID mice were infected i.vag. with 2 x 106 PFU of vhs-deficient HSV-2, strain 333d41, or the vhs rescue virus 333d41R. (A) Replication in the genital mucosa of BALB/c and SCID mice. The titer of virus in vaginal swab samples was determined by standard plaque assay. Data points represent the geometric mean +- standard error of the mean (SEM). (B) Severity of genital and neurological disease in BALB/c and SCID mice. Mice were scored for signs of disease as described in Materials and Methods. Values represent the arithmetic mean +- SEM. (C) Replication in the nervous systems of BALB/c and SCID mice. Mice were sacrificed 6 d postinfection. Brain, brainstem, and spinal cord tissues were disrupted by bead beating. Viral titers were determined by standard plaque assay. The dashed line indicates the limit of detection. Bars represent the geometric mean +- SEM. FIG. 2. : Replication in the genital mucosa of 129 and IFN-alpha/betaR-/- mice. Replication in the genital mucosa of 129 and IFN-alpha/betaR-/- mice. Groups of mice were infected i.vag. with 2 x 106 PFU of the indicated HSV-2 strains. The titer of virus in vaginal swab samples of 129 mice (n = 10 to 12) (A) and IFN-alpha/betaR-/- mice (n = 6 to 8) (B) was determined by standard plaque assay; for DeltaTK-, n = 5. Data points represent the geometric mean +- standard error of the mean. The dashed line indicates the limit of detection. *, P = 0.0178; **, P < 0.0001 (for 333d41 compared with 333d41R). FIG. 3. : Severity of genital and neurological disease in 129 and IFN-alpha/betaR-/- mice. Severity of genital and neurological disease in 129 and IFN-alpha/betaR-/- mice. Mice as described in the legend for Fig. were scored for signs of disease. Values for 129 (A) and IFN-alpha/betaR-/- (B) mice represent the arithmetic mean +- standard error of the mean. **, P < 0.001, as determined by Kruskall-Wallis nonparametric test for 333d41 compared with 333d41R. FIG. 4. : Replication in the nervous systems of 129 and IFN-alpha/betaR-/- mice. Replication in the nervous systems of 129 and IFN-alpha/betaR-/- mice. Mice as described in the legend for Fig. were sacrificed on day 6 postinfection. Brain, brainstem, and spinal cord tissues of 129 (A) and IFN-alpha/betaR-/- (B) mice were disrupted by bead beating. Viral titers were determined by standard plaque assay. The dashed line indicates the limit of detection. *, P = 0.032 to 0.012; **, P < 0.0001 for 333d41 compared with 333d41R as determined by t test. n = 9 to 11 for 129 samples; n = 5 for IFN-alpha/betaR-/- samples due to deaths of mice. For DeltaTK-, n = 5. FIG. 5. : Pathogenesis of 333d41 is not restored in PKR-/- mice. Pathogenesis of 333d41 is not restored in PKR-/- mice. Groups of seven to nine wild-type 129 and PKR-/- mice were infected i.vag. with 2 x 106 PFU of SB5 or 333d41. n = 4 for 129/SB5. (A) The titer of virus in vaginal swab samples was determined by standard plaque assay. Data represent the geometric mean +- standard error of the mean (SEM). (B) Mice were scored for signs of disease as described in the legend for Fig. . Values represent the arithmetic mean +- SEM. (C) Viral titers in brain, brainstem, and spinal cord tissues were determined as described in the legend for Fig. . Dashed lines indicate limits of detection. FIG. 6. : Pathogenesis of vhs-deficient HSV-1 is not restored in IFN-alpha/betaR-/- mice. Pathogenesis of vhs-deficient HSV-1 is not restored in IFN-alpha/betaR-/- mice. Groups of four to five 129 and IFN-alpha/betaR-/- mice were infected i.vag. with 2 x 106 PFU of HSV-1 KOS or vhs-deficient UL41NHB. (A) Replication in the genital mucosa. The titer of virus in vaginal swab samples was determined by standard plaque assay. Values represent the geometric mean +- the standard error of the mean (SEM). The dashed line indicates the limit of detection. (B) Severity of genital and neurological disease. Mice were scored for signs of disease as described in Materials and Methods. Values represent the arithmetic mean +- SEM. Backmatter: PMID- 12915570 TI - Phylogenetic and Evolutionary Relationships among Torovirus Field Variants: Evidence for Multiple Intertypic Recombination Events AB - Toroviruses (family Coronaviridae, order Nidovirales) are enveloped, positive-stranded RNA viruses that have been implicated in enteric disease in cattle and possibly in humans. Despite their potential veterinary and clinical relevance, little is known about torovirus epidemiology and molecular genetics. Here, we present the first study into the diversity among toroviruses currently present in European swine and cattle herds. Comparative sequence analysis was performed focusing on the genes for the structural proteins S, M, HE, and N, with fecal specimens serving as sources of viral RNA. Sequence data published for animal and human torovirus variants were included. Four genotypes, displaying 30 to 40% divergence, were readily distinguished, exemplified by bovine torovirus (BToV) Breda, porcine torovirus (PToV) Markelo, equine torovirus Berne, and the putative human torovirus. The ungulate toroviruses apparently display host species preference. In phylogenetic analyses, all PToV variants clustered, while the recent European BToVs mostly resembled the New World BToV variant Breda, identified 19 years ago. However, we found ample evidence for recurring intertypic recombination. All newly characterized BToV variants seem to have arisen from a genetic exchange, during which the 3' end of the HE gene, the N gene, and the 3' nontranslated region of a Breda virus-like parent had been swapped for those of PToV. Moreover, some PToV and BToV variants carried chimeric HE genes, which apparently resulted from recombination events involving hitherto unknown toroviruses. From these observations, the existence of two additional torovirus genotypes can be inferred. Toroviruses may be even more promiscuous than their closest relatives, the coronaviruses and arteriviruses. Keywords: Introduction : Toroviruses, coronaviruses, arteriviruses, and roniviruses (order Nidovirales) are evolutionary related enveloped positive-stranded RNA viruses of vertebrates and invertebrates. Their common ancestry is evident from sequence identity in their replicase proteins and from similarities in genome organization, gene order, and replication strategy . There are, however, marked differences in genome size, host range, and virion architecture . Toroviruses are morphologically unique and typically occur as a collection of discoidal, kidney-, and rod-shaped particles, with a tubular nucleocapsid enveloped by a membrane carrying large spikes . Whereas coronaviruses and arteriviruses have been studied in detail, little is known about toroviruses. One important reason for this gap is that toroviruses have not been propagated in cell culture, with the sole exception so far of the Swiss equine isolate Berne virus (BEV). Early seroepidemiological surveys based upon BEV cross-neutralization assays and enzyme-linked immunosorbent assay (ELISA) indicated that toroviruses occur in a wide variety of ungulate hosts (cattle, sheep, goats, and swine) (, , , , ; for a review, see reference ). Toroviruses (ToVs; acronyms used are according to the recommendations of the ICTV Coronavirus Study Group) seem to be associated with asymptomatic enteric infections in swine but can cause serious, at times fatal, diarrheal disease in cattle (, -). Berne virus-related bovine toroviruses (BToVs), designated Breda virus (BRV), were first seen in Breda, Iowa, during an outbreak of neonatal calf diarrhea, in which 15% of the affected animals died . Both gnotobiotic and conventionally reared calves developed mild to severe diarrhea upon experimental BRV infection . Toroviruses have also been implicated in human gastroenteritis. In fecal samples from children and adults with diarrhea, torovirus-like particles cross-reactive with BEV- and BRV-specific antisera were detected by immunoelectron microscopy , and torovirus antigens were detected by ELISA . In addition, torovirus RNA was detected in stool specimens of pediatric patients by reverse transcription (RT)-PCR with primers designed after the 3' nontranslated region (3'-NTR) of the Berne virus genome . Recently, we completed the nucleotide sequence of the Berne virus genomic RNA (S. L. Smits and R. J. de Groot, unpublished data). With a genome length of 28 kb, toroviruses are among the largest RNA viruses, rivaled in size only by the coronaviruses. The 5'-most two-thirds of the genome are taken up by two huge overlapping open reading frames (ORFs), 1a and 1b, encoding polyproteins from which the various subunits of the viral replicase/transcriptase are derived (; S. L. Smits and R. J. de Groot, unpublished data). Downstream of ORF1b, there are four cistrons of 5 kb, 0.7 kb, 1.2 kb, and 0.5 kb (as ordered from 5' to 3'). These encode the structural proteins, the spike (S), membrane (M), hemagglutinin-esterase (HE), and nucleocapsid (N) proteins, respectively . The structural proteins are translated from a 3'-coterminal nested set of subgenomic mRNAs , produced by discontinuous and nondiscontinuous RNA synthesis . Sequence data for toroviruses other than Berne virus are limited to the S, M, HE, and N genes of the bovine torovirus BRV , the N gene of porcine torovirus (PToV) field variant Markelo , and the HE gene from a putative human torovirus variant . Clearly, many questions regarding the genetic diversity of toroviruses need to be addressed. For instance, to what extent do the toroviruses currently present in the field resemble the early BEV and BRV? These viruses were isolated more than two decades ago from different hosts in different geographic locations . Are toroviruses host specific, or does transmission occur between vertebrate species? What is the geographic distribution of torovirus genotypes? To increase our insight into torovirus phylogeny and evolution, we performed comparative sequence analysis of selected European porcine and bovine variants, focusing on the genes for the structural proteins. The toroviruses of cattle and swine display host species preference and belong to distinct genotypic lineages. However, we also provide evidence for multiple intertypic recombination events which involved BToV Breda- and PToV Markelo-like ancestors as well as hitherto unknown toroviruses. MATERIALS AND METHODS : Specimens. | Pooled fecal samples (20 to 60 animals per specimen) from 75 Dutch veal calf farms with 38 to 930 calves, aged 1 to 52 weeks , were kindly provided by W. van der Poel (National Institute of Public Health and the Environment, Bilthoven, The Netherlands). Bovine and porcine fecal samples, taken from diarrheic animals in Italian herds, were screened for the presence of toroviruses by (immuno) electron microscopy . Specimens containing torovirus RNA were identified by diagnostic RT-PCR, targeted to the 3'-NTR of the torovirus genome. A fecal sample from a naturally infected pig containing the Dutch PToV variant Markelo (P-MAR) was described previously , as were two torovirus-positive specimens obtained during an RT-PCR survey of Hungarian swine and cattle herds . All samples selected for further genetic analysis are listed in Table . RNA purification, cDNA synthesis, and PCR amplification. | RNA was extracted from fecal samples and concentrated according to the guanidinium isothiocyanate silica protocol of Boom et al. with minor modifications . RT-PCR amplification of torovirus sequences was performed with RNase H-free Moloney murine leukemia virus reverse transcriptase (Superscript II) and Taq polymerase (Gibco-BRL, Life Technologies) or with the Expand long-template PCR system (Roche Diagnostics GmbH), according to the instructions of the manufacturers. Precautions to avoid carryover of amplification products included physical separation of the pre- and post-PCR procedures. In each experiment, multiple water controls were taken along. Amplicons were gel purified and cloned into vector pGEM-T Easy (Promega Corp., Madison, Wis.). Sequence reactions were performed commercially (BaseClear Labservices) with the ABI Prism BigDye Terminators version 3.0 cycle sequencing kit (Applied Biosystems) and an ABI Prism 3100 genetic analyzer. The nucleotide sequences determined for each RT-PCR product on at least two independent clones and, with few exceptions, in both orientations were analyzed and deposited in the EMBL database. Primers used for RT-PCR and sequence analysis are listed in Table . Phylogenetic analysis. | Torovirus sequences were aligned with the DNAStar program MegAlign (ClustalV method; DNAStar, Inc.) to determine pairwise differences. Multiple sequence alignments were generated with the program Multalin with the Dayhoff symbol comparison table. The sequences of the genes coding for structural proteins of BToV Breda 1, EToV Berne, and HToV were extracted from the NCBI database (accession numbers , , , , and ). Phylogenetic and molecular evolutionary analyses were conducted with Mega version 2.1 . To detect sites of recombination, the program PhylPro was used . The distribution of synonymous substitutions (Dss) was determined by pairwise comparison of protein-coding nucleotide sequences via sliding-window analysis. Gaps were excluded from the alignments. The number of synonymous substitutions per synonymous site (Ks) was estimated for overlapping 240-nucleotide gene segments with a 60-nucleotide step size with the program K-estimator 4.5 . Dss profiles were generated by plotting the calculated Ks values against nucleotide position. Nucleotide sequence accession numbers. | The newly determined sequences have been deposited in the EMBL database under accession numbers through . RESULTS | Identification of new bovine and porcine torovirus field variants. | During routine electron microscopic analysis of clinical samples, torovirus-like particles were occasionally observed in the feces of diarrheic Italian cattle and swine . A diagnostic RT-PCR, targeted to the 3'-NTR of the viral genome , detected torovirus RNA in one out of four bovine (B6) and in three out of four porcine (P4, P9, and P10) samples. Four additional bovine specimens containing torovirus RNA (B145, B150, B155, and B156) were identified upon RT-PCR screening of a library of pooled fecal samples from 75 Dutch cattle farms, collected during the winter of 1998 . Also included in our studies were fecal samples containing the Dutch PToV variant Markelo (P-MAR) and two specimens (B1314 and P78) identified as torovirus positive in a previous RT-PCR survey of Hungarian swine and cattle herds . All samples, with date, location, and host of origin, are listed in Table . Viral RNA extracted from feces served as a template for RT-PCR amplification of the genes for the structural proteins (Fig. ; Table ). To avoid primer-directed selection of sequence variants, nonbiased primers were designed after elements conserved among BRV, BEV, and PToV, including the 3'-NTR and the intergenic regions . For the amplification of S genes, the positive-stranded primer was designed after ORF1b sequences, conserved between BEV and BRV . In all cases, RT-PCR amplicons of the anticipated size were obtained (not shown), which were analyzed either directly or after cloning. Genetic diversity of torovirus field variants. | Pairwise comparison of toroviral nucleic acid sequences revealed considerable diversity, with identities ranging from 55.9 to 100% (Tables , , and ). Phylogenetic trees were generated for each gene by neighbor-joining , maximum parsimony, and maximum-likelihood methods (not shown); these methods yielded essentially identical topologies. Three distinct torovirus genotypes with apparent preferences for horse, cattle, and swine could be discerned, represented by BEV, BRV, and P-MAR, respectively; HToV apparently constitutes a fourth genotype . While all new PToV strains resembled P-MAR, the BToV strains seemed mostly related to BRV. The S gene of the Dutch BToV variant B145 shared 93% identity with that of the prototypic New World BToV variant Breda virus . The B150 S gene was partially sequenced. Within its 3'-most 2,436 nucleotides, the B150 S gene was 98% and 95% identical to those of B145 and BRV, respectively. Sequence identities between the BToV S genes and those of Berne virus and P-MAR were considerably lower (77% and 73%, respectively; Fig. ). In trees constructed for the M genes, the new BToV variants clustered with BRV (93 to 94% identity); among these new variants, sequence identity in the M gene was more than 99% (Table and Fig. ). Remarkably, however, in the HE gene, only a subset of the BToV variants (B6, B145, B156, and B1314) grouped with BRV . The B150 and B155 HE sequences were 99% identical to each other but quite distinct from those of the other BToVs (67 to 73% identity) and equally far removed from those of the PToV variants (65 to 68% identity; Table and Fig. ). Moreover, for all new BToV variants, the N genes did not group with that of BRV but, surprisingly, with those of the PToV strains . Note, however, that the new variant BToV N genes still formed a separate cluster. The PToV strains P-MAR, P9, P10, and P78 displayed 5 to 8% divergence and clustered in trees constructed for M, HE, and N (Fig. , C, and D). However, there was an anomaly in the phylogeny of variant P4: in the N and M genes, P4 grouped with the other PToVs, but in trees based upon HE sequences, it was clearly distinct . In the HE gene, P4 is only 77% identical to the other PToVs, versus 92 to 94% sequence identity in the M and N genes (Tables to ). Evidence for recombination of torovirus sequences. | Conflicting phylogenetic data, such as obtained for the different regions of the BToV and PToV genomes, are indicative of recombination. For example, the B6, B145, B156, and B1314 sequences seem to have resulted from a genetic exchange between BToV and PToV. In order to identify recombination breakpoints, we employed phylogenetic profile analysis (PhylPro) , a computer graphic method based on the premise that phylogenetic relationships derived from different regions of a multiple sequence alignment will be similar when no recombination has occurred. Recombination junction sites create areas of low phylogenetic correlation and in PhylPro graphs are depicted as sharp downward peaks. Within the HE genes of B6, B145, B156, and B1314, single recombination breakpoints were detected, located approximately 1,100 nucleotides from the ATG initiation codon . Visual inspection of the alignment suggested a template switch between residues 1070 through 1110: sequences coding for the ectodomain of a BRV-type HE were effectively fused to those encoding a PToV-type transmembrane region and cytoplasmic domain . PhylPro analysis of the B150 and B155 sequences revealed three potential recombination sites , one located near the 5' end of the HE gene, approximately 80 nucleotides downstream of the ATG codon. Visual inspection of the alignment suggested a template switch within the region comprising residues 70 through 110 . The other two breakpoints were located at the 3' end of the HE gene between residues 1070 and 1110, one at the exact same position as was found for B6, B145, B156, and B1314 and one immediately upstream of it (Fig. to C). PhylPro comparison of PToV variant P4 with the other PToV variants yielded inconclusive results, with multiple, weak recombination signals scattered throughout the HE region (not shown). This prompted the question of whether the sequence divergence in the PToV HE genes resulted from recombination or merely from antigenic drift. As an alternative approach to test for recombination, we determined the distribution of synonymous substitutions (Dss). "Silent mutations" are fixed at random locations within viral genomes and at an approximately constant rate, providing a molecular clock . The number of synonymous substitutions per synonymous site, Ks , thus reflects the evolutionary distance between two related coding sequences. If viruses have diverged from a common ancestor without the occurrence of recombination, pairwise sequence comparison should yield essentially identical Ks values throughout the genomes. Differences in Ks value between genomic regions, however, would be a tell-tale sign of recombination. The coding sequences of the various torovirus variants were compared pairwise by sliding-window analysis, during which Ks values were estimated for overlapping 240-nucleotide gene segments with the program K-estimator 4.5 . Dss profiles were generated by plotting the calculated Ks values against nucleotide position . Comparison of P-MAR with the PToV variants P10 and P9 yielded roughly constant Ks values across the M, HE, and N genes . Apparently, no intertypic genetic exchanges have occurred in this part of the genome after P9, P10, and P-MAR branched off from their most recent common ancestor. In contrast, pairwise comparison of P4 with the other PToV variants (Fig. and not shown) revealed that within the HE region, comprising nucleotides 20 to 950, the number of synonymous substitutions per synonymous site was four- to fivefold higher than in the flanking HE sequences or in the adjacent M and N genes. The Dss profiles are suggestive of a double recombination event, during which a segment of the PToV HE gene was exchanged for the corresponding HE sequences from an as yet unidentified torovirus. However, the data do not allow a definitive conclusion as to whether the P4 HE sequence is recombinant or, rather, that of the other PToVs. Dss profiles obtained for the BToV variants were fully consistent with the results of phylogenetic analysis and PhylPro scanning. Comparison of B6 and B145 gave a Dss profile with constant, low Ks values throughout the M, HE, and N coding regions. In contrast, the B145-BRV comparison revealed an abrupt increase in Ks value at the deduced recombination breakpoint within the B145 HE gene, coinciding with a sudden decrease in Ks value in the B145-P-MAR plot . Thus, the Dss profiles also indicate a single recombination event in the HE gene, during which BRV- and PToV-type sequences were fused. The Dss profiles of the B150-B6 and B150-B145 pairs showed that, within the region comprising the codons for residues 50 to 1100 of the HE gene, the number of synonymous substitutions per synonymous site was approximately eightfold higher than in the surrounding sequences . In profiles comparing B150 to BRV, Ks values were low in the M gene and in the 5' region of the HE gene, with a sharp eightfold increase in the number of synonymous substitutions per synonymous site around position 40. Conversely, the B150 versus P-MAR profile showed a sixfold drop in Ks value near the 3' end of the HE gene, around nucleotide 1100 . The combined data suggest that the HE sequences amplified for B150 and B155 are a composite derived from a BRV-like parent, a PToV-like parent, and an unknown torovirus. DISCUSSION : The identification and characterization of new torovirus field variants is arduous. Although seroprevalence data indicate that torovirus infections are common in ungulates , naturally infected animals do not display distinctive clinical signs . Moreover, fecal or respiratory shedding of virus, detectable by ELISA or RT-PCR, appears to be short-lived . For the present study, we used torovirus-positive fecal specimens from cattle and swine in Italy, the Netherlands, and Hungary which had been either preselected by electron microscopy or prescreened by diagnostic RT-PCR. This allowed the first study into the genetic diversity among toroviruses currently present in European herds. Four genotypes, displaying 30 to 40% sequence divergence, were readily distinguished. These were encountered in horses, cattle, swine, and humans and are represented by BEV, BRV, PToV, and HToV, respectively. Although the ungulate toroviruses may not be species specific (see below), they do display host species preference. In phylogenetic analyses, all PToV variants clustered, whereas the recent European BToVs resembled the New World BToV variant BRV, which was isolated 19 years ago . For example, in the S, M, and HE genes (6.7 kb), the European variant B145 shared 93% overall sequence identity with BRV. Intriguingly, chimeric torovirus sequences were amplified from a number of fecal specimens, suggesting the existence of two additional torovirus genotypes. The detection of chimeric genes would imply intertypic genetic exchange via RNA recombination in natural torovirus populations. However, the fact that pooled fecal samples served as a source of viral RNA poses an obvious caveat. Herds and even individual animals may be subject to coinfection with nonrecombinant toroviruses of different genotypes. Consequently, the chimeric HE sequences could have been generated during RT-PCR, in particular during the reverse transcription step. RNase H-deficient Moloney murine leukemia virus reverse transcriptase was used throughout our experiments to minimize the frequency of template switching, but even under these conditions, recombination in vitro may still have occurred . Indeed, in simulation experiments in which we mixed synthetic T7 RNA polymerase transcripts of P-MAR and BRV HE at a 1:1 ratio and performed RT-PCR with nonbiased primers, chimeric amplicons were obtained. However, the in vitro recombination frequency was only 5% (4 out of 76 clones tested), which is in good agreement with the values (3%) reported by Luo and Taylor . Most amplicons (95%) were nonrecombinant and either PToV or BRV specific. Moreover, template switching occurred at seemingly random locations (not shown). In contrast, RT-PCR amplification of HE sequences from the various bovine and porcine specimens gave strictly reproducible results. In independent RT-PCR assays, RNA extracted from each individual sample yielded consistently and exclusively one particular chimeric sequence; parental nonrecombinant HE genes were never detected. Moreover, sequences of identical chimeric structure were obtained with different primer sets and, most notably, were detected in different fecal specimens. The B6, B145, B156, and B1314 HE sequences differed at several nucleotide positions, as would be expected for genes of viruses subject to genetic and/or antigenic drift. Yet they all had identical recombination breakpoints and parents; the same was true for the HE genes of specimens B150 and B155. Based upon the collective evidence, we postulate that the chimeric sequences detected by RT-PCR are not in vitro artifacts but faithfully reflect the genomic organization of torovirus field variants. During PToV divergence, a recombination event occurred, apparently involving a donor virus of a hitherto unknown genotype. Thus, the coding sequences for the PToV HE ectodomain were partly replaced, presumably through a double template switch . Our data do not allow a conclusion as to whether P4 or the P-MAR-like viruses are the recombinant progeny. As the P4 variant is the oldest field variant, it may well represent the parental PToV type; a definitive answer will have to await identification of the other recombination partner. The phylogenetic data for the BToV variants are best explained by two sequential recombination events from which the B150/B155 lineage arose. An initial genetic exchange between BRV- and PToV-like parents gave rise to the B6/B145/B156/B1314 lineage. Apparently, a single template switch occurred at the 3' end of the HE gene, as a result of which part of the BToV HE gene and the complete BToV N gene were replaced by the corresponding PToV sequences. A B6-type descendant then partook in a second recombination event, this time entailing a double template switch during which the coding sequences for the ectodomain of a BRV-type HE protein were replaced by those of yet another unknown torovirus . The occurrence of intertypic recombination implies cross-species transmission. At present, it cannot be established which hosts served as "mixing vessels" to produce the recombinant PToV and BToV variants. Also, the natural hosts of the unknown parental viruses of the PToV and BToV recombinants remain to be identified. How long ago did the recombination events take place? Our data set is too limited to allow more than an educated guess. From the sequence alignments and the sampling dates of the field variants (1990, 1998, and 2000 for B6, B145, B156, and B1314, respectively), we estimated the rate of synonymous substitutions in BToV to be ~3 x 10-3 per synonymous site per year. This is somewhat lower than the rates reported for influenzavirus, picornavirus, and human immunodeficiency virus (~13 x 10-3) , but in the same order as the rate estimated for the coronavirus transmissible gastroenteritis virus. Sanchez et al. determined the total mutation fixation rate at ~0.7 x 10-3 substitutions/site/year. From their data set, we recalculated a rate of ~1.4 x 10-3 for synonymous substitutions (S. L. Smits and R. J. de Groot, unpublished data). Under the assumptions of (i) a monophyletic origin of the new BToV strains and (ii) the existence of an approximate molecular clock , the B6-type and B150-type lineages would have branched during the mid-1990s. Although this rough estimate should be considered with caution, the data suggest that the extant recombinant BToV variants emerged only recently. The genetic exchanges possibly resulted in antigenic shifts which allowed the recombinant viruses to evade existing herd immunity. It is of note that in two instances the ectodomain of the HE protein was altered as a result of the recombination. HE is a prominent toroviral envelope glycoprotein and hence a potential prime target for the humoral immune response. Most likely, the recombination mechanism involved similarity-assisted template switching (homologous recombination) . In each case, the donor sequence neatly replaced the homologous region of the acceptor sequence, and the crossovers occurred within stretches of 20 to 40 nucleotides that are highly conserved (98 to 100% identity) among currently known torovirus genotypes. Recombination has been well documented for RNA viruses , including other nidoviruses (for reviews, see references , , and ). Homologous recombination between related strains was observed for coronaviruses and arteriviruses, both under experimental conditions and in nature . The genus Coronavirus consists of three antigenic clusters, which in turn can be subdivided into at least 13 different species on the basis of host specificity and genotype . While intraspecies recombination among coronaviruses may be common in the field , there are only a few recognized cases of interspecies recombination. The best-known exceptions are the type II feline coronaviruses, which originated from genetic exchanges between canine and type I feline coronavirus strains . The type II feline coronaviruses do not form a monophyletic lineage but arose from multiple independent recombination events . The frequency of successful intertypic genetic exchange (those instances in which the recombinant offspring can establish itself in the host population) is determined among others by (i) properties inherent to the viral replicase, (ii) the frequency of cross-species transmission and the odds of double infections (not only of the host, but also of host cells), (iii) the viability of the recombinant progeny, and (iv) the "evolutionary gain," i.e., the increase in fitness of the recombinant viruses relative to their nonrecombinant parents . Coronaviruses and toroviruses are closely related by evolution, occupy similar niches, and possess error-prone replicases that promote RNA recombination. The apparent difference between coronaviruses and toroviruses, with respect to the frequency of natural intertypic recombinant viruses in the field, may therefore reflect differences in viral "life style" and survival strategy. The evidence that toroviruses cause gastroenteritis in animals and humans ranges between firm and tenuous (, , , , , , , -), and their veterinary importance and etiological role in human disease remain to be corroborated. Epidemiological studies are hampered by our limited knowledge of toroviral genetic and antigenic diversity. Thus far, most diagnostic assays rely on BEV- or BRV-derived antigens and sequences and are hence of narrow specificity. The data presented in this report not only offer an exciting glimpse into torovirus genetics but also furnish new diagnostic tools, including a set of antigenically distinct HE proteins. FIG. 1. : Electron microscopic detection of toroviruses in fecal specimens from diarrheic swine (P4 and P9) and cattle (B6). Electron microscopic detection of toroviruses in fecal specimens from diarrheic swine (P4 and P9) and cattle (B6). Virions were analyzed directly (B6) or after immunoaggregation with convalescent-phase serum from PToV P4-infected animals (P4 and P9). Negative staining was performed with 2% sodium phosphotungstic acid. Bars, 100 nm. FIG. 2. : Torovirus genome organization and schematic outline of the strategies used for RT-PCR amplification of the genes for the structural proteins. Torovirus genome organization and schematic outline of the strategies used for RT-PCR amplification of the genes for the structural proteins. The upper panel shows a schematic representation of the genome. Boxes represent the genes for the polymerase (ORF1a and ORF1b), the spike protein (S), the membrane protein (M), the hemagglutinin-esterase (HE), and the nucleocapsid protein (N). Open arrows indicate intergenic regions and transcription-regulating sequences. Also indicated are the cap structure (black dot) and the poly(A) tail (An). The lower panel shows a schematic outline of the RT-PCR assays employed to amplify S, M, HE, and N sequences. The orientations and positions of the oligonucleotides on the torovirus genome are shown. FIG. 3. : Unrooted neighbor-joining trees depicting the phylogenetic relationships among torovirus field variants. Unrooted neighbor-joining trees depicting the phylogenetic relationships among torovirus field variants. Trees were constructed for (A) the S gene, (B) the M gene, (C) the HE gene, and (D) the N gene with the Kimura-2 parameter method. Confidence values calculated by bootstrap analysis (1,000 replicates) are indicated at the major branching points. Branch lengths are drawn to scale; the scale bar represents 0.05 nucleotide substitution per site. The torovirus reference strains BRV, BEV, P-MAR, and HToV are italicized and underlined. The tree shown for the HE gene was based upon an alignment corresponding to residues 1 to 1049 of B145. Note that BEV was not included in this tree; as a result of a large deletion, Berne virus has lost most of its HE gene . FIG. 4. : Identification of recombination sites in the M, HE, and N genes of torovirus field variants. Identification of recombination sites in the M, HE, and N genes of torovirus field variants. (A) Recombination sites were identified with the program PhylPro . This phylogenetic profile method introduces the phylogenetic correlation measure, i.e., the principle that phylogenetic relationships in different regions of an aligned sequence will be similar when no recombination has occurred. An alignment of the coding regions of the M, HE, and N genes of P-MAR, P4, P9, P10, BRV, B6, B145, B150, and B155 was generated. For each individual sequence in the alignment, the phylogenetic correlations were computed at every position with a sliding-window technique, with window limits fixed at 15 differences. Shown are the phylogenetic profiles of B6 and B145, with x and y axes indicating the nucleotide positions and the phylogenetic correlation, respectively; as a reference, the genes for M, HE, and N, drawn to scale, are depicted as boxes (top). For clarity, profiles of P4, B150, and B155 were hidden in the graph. (B) Phylogenetic profiles of B150 and B155; for clarity, profiles of P4, B6, and B145 were hidden. Window limits were fixed at 10 differences. (C) Alignments of HE sequences surrounding the recombination sites; arrows and numbers correspond to those in panels A and B. Nucleotide differences with respect to the B145 sequences are shown. Nucleotide positions given are numbered from A1 of the B145 HE gene. FIG. 5. : Distribution of synonymous substitutions (Dss) in the M, HE, and N genes of torovirus field variants. Distribution of synonymous substitutions (Dss) in the M, HE, and N genes of torovirus field variants. Multiple alignments were generated for the protein-coding nucleotide sequences; all gaps were excluded. The sequences were compared pairwise by sliding-window analysis, during which the number of synonymous substitutions per synonymous site (Ks) was estimated for overlapping 240-nucleotide gene segments with a 60-nucleotide step size. Dss profiles were generated by plotting the calculated Ks values against nucleotide positions. As a reference, the genes for M, HE, and N, drawn to scale, are depicted as boxes (top). The Dss profiles shown were produced by pairwise comparison (A) of P4, P9, and P-MAR, (B) of B145 to BRV, P-MAR, or B6, (C) of B145 to B150 or B6 and of B150 to B6, and (D) of B150 to BRV or P-MAR, with the B145-B6 graph providing a baseline. FIG. 6. : Hypothetical model for genetic exchanges among torovirus field variants. Hypothetical model for genetic exchanges among torovirus field variants. Torovirus genomes are depicted schematically, with the various genes represented by boxes. (A) Presumptive recombination event during PToV divergence, involving an ancestral PToV strain (P-Anc; genes indicated by white boxes) and an unknown toroviral parent (?; genes indicated by hatched boxes). Exchange of HE sequences resulted in recombinant progeny (P-Rec). (B) Presumptive recombination events during BToV divergence. During an initial single recombination event, the 3' end of the HE gene and the N gene and the 3'-NTR of a BRV-like BToV variant (indicated in white) were replaced by the corresponding PToV sequences (dotted), giving rise to the B6/B145/B156/B1314 lineage. A subsequent double recombination event, during which BToV HE sequences were replaced by those of an as yet unidentified torovirus (?, indicated in black), resulted in the B150/B155 lineage. TABLE 1 : Virus variants and characteristics TABLE 2 : Oligonucleotide primers used for RT-PCR TABLE 3 : Sequence relationships between toroviral M genes TABLE 4 : Sequence relationships between toroviral HE genes TABLE 5 : Sequence relationships between toroviral N genes Backmatter: PMID- 12915585 TI - Early- and Intermediate-Stage Variants of Simian Immunodeficiency Virus Replicate Efficiently in Cells Lacking CCR5 AB - Primate lentiviruses are thought to use the chemokine receptor CCR5 as the major coreceptor for entry into cells. Here we show that some variants of simian immunodeficiency virus (SIV) replicate efficiently in peripheral blood mononuclear cells (PBMCs) lacking a functional CCR5. There were differences in the replication patterns of sequential variants that evolved during SIVMne infection; the late-stage pathogenic variants were unable to replicate in PBMCs lacking CCR5, whereas the early- and intermediate-stage viruses replicated as well in PBMCs lacking CCR5 as they did in cells with wild-type CCR5. The coreceptor specificities of these sequential variants were compared using indicator cell lines expressing known SIV coreceptors. Among the known SIV coreceptors, there were none that were functional for the early and intermediate variants but not the late-stage variants, suggesting that the coreceptor used for replication in PBMCs may be a coreceptor that has not yet been described. Because some variants replicate with high efficiency in peripheral blood cells using this as yet uncharacterized cellular receptor, this coreceptor may be important for viral entry of some target cell populations in the host. Keywords: Introduction : The entry of primate lentiviruses into cells involves a series of interactions with host cell receptors, starting with binding of the viral envelope to the CD4 protein; this leads to a conformational change that allows binding of the envelope protein to a second receptor. This coreceptor can be one of a variety of seven transmembrane-spanning G-protein-coupled chemokine receptors. The chemokine receptor CCR5 is used by most isolates of human immunodeficiency virus type 1 (HIV-1) and is thought to be the coreceptor that is expressed on cells that are the earliest targets for virus replication after transmission. Some later-stage isolates of HIV-1 use CXCR4 as a coreceptor, and many of these variants also retain the ability to use CCR5 as a coreceptor, perhaps allowing some flexibility for later-stage viruses to infect a wider range of target cells. A minority of HIV-1 variants can use other G-protein-coupled receptors for entry, although the efficiency of infection is reduced relative to that for CCR5-mediated entry, and the importance of these alternate coreceptors in infection is unclear (reviewed in references , , and ). Simian immunodeficiency virus (SIV) also uses CCR5 as a coreceptor . In addition, several other cellular receptors can function as coreceptors for SIV, including GPR1, GPR15, CCR8, APJ, STRL33, and CHEMR23 (, , , , ; G. Alkhatib, F. Liao, E. A. Berger, J. M. Farber, and K. W. C. Peden, Letter, Nature 388:238, 1997). Many pathogenic SIVs can efficiently infect cells by using both CCR5 and GPR15 as a coreceptor with CD4 . However, studies of SIV infection of macaques suggest that GPR15 may not be important for virus replication in vivo because mutations that abolish GPR15-mediated entry of SIVmac239 did not affect the levels of virus in the infected animal . In general, it remains unclear if coreceptors other than CCR5 play a major role in SIV or HIV replication at any stage of infection and disease in the host. Previous studies of SIV infection of Macaca nemestrina demonstrated that the late-stage variants are more pathogenic than the cloned infecting virus, SIVMneCL8, from which they arose in the host . Viruses that evolved at intermediate (35-week-postinfection) and late (170-week-postinfection) stages in animals infected with SIVMneCL8 replicated 100- and 3,000-fold better, respectively, than the parental viral clone when transmitted to naive monkeys . Some of the increased replication was attributable to the inability of these intermediate- and late-stage viruses to elicit neutralizing antibody responses. However, because neither the intermediate nor the late viruses could elicit detectable neutralizing antibodies, the similar30-fold-higher replication of the two viruses may reflect an additional selective advantage of the late-stage variant . In our initial studies of the coreceptor specificities of these different SIVMne variants, we found no differences in the coreceptor preferences of these viruses; all of the SIVMne variants tested were able to infect cells in culture by using either CCR5 or GPR15, but not CXCR4 . To examine if coreceptors other than CCR5 were important for replication of these variants in relevant cell types such as T lymphocytes and macrophages, we examined the ability of the viruses to infect human peripheral blood mononuclear cells (PBMCs) from donors who were homozygous for a 32-bp deletion of CCR5 (Delta32 PBMCs). The viruses examined were chimeric viruses encoding envelope surface unit proteins that were obtained from intermediate [35-week-postinfection; SIVMneCL8(35wkSU)] and late [170-week-postinfection; SIVMneCL8(170wkSU)] stages of infection in animals inoculated with SIVMneCL8 . Thus, the only difference between viruses was in the envelope surface protein. Blood samples from four HIV-1-seronegative donors, two with a wild-type CCR5 allele (35-9 and 53-5) and two with the homozygous Delta32 allele (07-5 and 43-1), were provided without indication of the CCR5 status. PBMCs were isolated from whole blood by Ficoll-Hypaque centrifugation and cultured by methods similar to those described previously . SIV particles were made by transfecting 293T cells with the proviral clones and then passaging cell-free virus harvested 2 days later for 1 to 2 weeks in CEMx174 cells. The infectious titer was determined using sMAGI cells . SIVmac239 and the SIV/HIV-1 chimera SHIV162 were also generated in CEMx174 cells; in the case of SHIV162, the virus titer was determined using MAGI-CCR5 cells . Phytohemagglutinin (PHA)-stimulated PBMCs were infected with each virus at a multiplicity of infection of 0.01. Every 2 to 3 days, medium was replaced, and the level of virus replication was assessed by assaying the filtered supernatant for SIV p27gag. The results of the first infection study are shown in Fig. , and they represent the averages of results from duplicate infections. SHIV162 was used as a control for these experiments because it is dependent on CCR5 for entry and thus would not be expected to replicate in PBMCs lacking a functional CCR5 protein. As expected, SHIV162 replicated to high levels in the PBMCs from two donors and not in the PBMCs from two others. Upon unblinding of the experiment, the donors were shown to have the expected CCR5 genotype. SHIV162 replicated to high levels in PBMCs from each of the wild-type CCR5 donors, with peak levels of p27gag production of similar19 to 58 ng/ml (Fig. ; Table ). Somewhat lower peak levels of replication (similar6 to 15 ng/ml) were observed for each of the SIVs tested in PBMCs from the two donors with a wild-type CCR5 allele. In PBMCs from donors with a deletion in CCR5, the infection profile of SIVMne170SU followed a similar pattern as that observed with the CCR5-dependent SHIV162 control virus; there was a low level of virus replication (peak of <=0.6 ng of p27gag/ml) of the chimera carrying the envelope gene from late-stage infection. Surprisingly, the parental, early-stage SIVMne clone, SIVMneCL8, replicated to high levels in the cells lacking CCR5. Similarly, the virus carrying the prototype envelope gene from intermediate stages of infection also replicated in PBMCs lacking CCR5. In both cases, the replication kinetics of these early- and/or intermediate-stage viruses were very similar in PBMCs from wild-type and Delta32 CCR5 donors . The infection experiments were repeated on a second occasion, and similar results were obtained. Table shows the peak viral levels observed in both infection experiments in which we used PBMCs from the same four donors, two with wild-type CCR5 and two with Delta32 CCR5. In summary, the control virus SHIV162 replicated to similar100-fold-higher peak levels in each of the four experiments with PBMCs carrying a wild-type CCR5 gene than in experiments with PBMCs carrying Delta32 CCR5. A similar difference (10- to 100-fold) between peak levels of p27gag in Delta32 PBMCs and those in wild-type PBMCs was observed for both the full-length proviral clone from 170 weeks postinfection, SIVMne170, and the chimera encoding the envelope from this provirus, SIVMneCL8(170wkSU). In contrast, both SIVMneCL8 and SIVMneCL8(35wkSU) replicated to similar levels in PBMCs from wild-type and Delta32 CCR5 donors (approximately one- to threefold differences in peak viral levels). In the second experiment, SIVmac239 was also examined because previous studies have shown that this virus can replicate in PBMCs lacking CCR5 from some donors but not others . Replication of SIVmac239 was only three- to fourfold lower in Delta32 CCR5 PBMCs; the replication pattern was very similar to that of the intermediate-stage SIVMne variant in the second experiment, where it was tested in parallel . Thus, the major novel finding is that the early- and intermediate-stage SIVMne variants can replicate in PBMCs in the absence of CCR5, whereas late-stage SIVMne variants are much more CCR5 dependent. To determine whether the coreceptor that is used by early- and intermediate-stage SIVMne variants is among those that have been described, we examined the ability of the viral variants to infect a set of HOS-CD4+ (GHOST) cell lines that express various coreceptors known to support infection by different strains of SIV. It was previously shown that all SIVMne variants use CCR5 and GPR15 for entry into target cells . None of the SIVMne variants was able to use STRL33 (now called CXCR6) or CXCR4 as a coreceptor for entry, at least at the levels expressed on GHOST cells that permit HIV-1 infection . To examine other potential coreceptors, we infected a panel of GHOST4 cells expressing the following coreceptors: CCR8, GPR1, APJ, and CHEMR23. We compared infection of these cells with infection of GHOST cells expressing CCR5 (GHOST cells expressing CCR8, GPR1, and CCR5 were kindly provided by Dan Littman). In the case of APJ and CHEMR23, the cells were constructed by transducing the parental indicator cells (GHOST4), which are HOS cells modified to stably express the human CD4 gene and humanized green fluorescent protein (GFP) under the control of the HIV-2 long terminal repeat, with retroviral particles carrying a pBabe-puromycin vector containing the coreceptor gene of interest (APJ or CHEMR23). The results of infection of GHOST cells expressing various coreceptors with the different SIVMne variants and relevant control viruses are shown in Table . As expected, all viruses tested were able to use CCR5. However, neither the early-stage variant SIVMneCL8 nor the intermediate-stage variant SIVMneCL8(35wkSU) was able to use any of these additional coreceptors. The late-stage variant SIVMneCL8(170wkSU) was able to use CCR8 and GPR1 in addition to CCR5 and GPR15, although the efficiency of alternative coreceptor usage was reduced compared to CCR5 usage. Summary and conclusions. : The results of these studies suggest that some SIV variants can replicate to high levels in PBMCs lacking CCR5. To our surprise, this was true for early- and intermediate-stage variants but not for later-stage variants that evolved from them. This result was unexpected because the late-stage virus SIVMne170 replicated to much higher levels in the animal, as judged by both plasma and lymph node viral RNA levels, and the late-stage virus was more pathogenic . This suggests that the alternative coreceptor used by early-stage viruses may not be important for high levels of virus replication per se. Thus, entry with the use of this coreceptor, if relevant in vivo, may be for a select subset of targets, perhaps cells important early in infection and/or with viruses that cause a more prolonged disease course. Additional studies with sequential viruses from monkeys infected with other SIV or simian/human immunodeficiency virus (SHIV) strains may provide insights as to whether there is a relationship between CCR5-independent replication and stage of infection for primate lentiviruses. One might assume that the opposite is the case for HIV-1 infection in humans, because in some individuals, the late-stage HIV-1 variants acquire the ability to use the CXCR4 coreceptor to infect PBMCs. However, there is no evidence that well-characterized sequential variants from the same source have been tested for replication in PBMCs that lack CCR5. Both GPR15 and STRL33 are expressed on activated PBMCs (-, ). However, there was not a correlation between the use of these coreceptors and the ability of the SIV variant tested to replicate in PBMCs lacking CCR5. The coreceptors tested were of human origin, so species-specific differences in the GHOST cell and the PBMC experiments cannot account for the findings. None of the SIVMne variants efficiently infect GHOST cells expressing STRL33, whereas GPR15 is used by all of the SIVs tested, including the late-stage variants ; thus, infection via one of these coreceptors would not appear to explain the observed differences in replication of the SIV variants in Delta32 PBMCs. However, we cannot completely rule out that the differences we observed between viruses reflect a difference between their abilities to use a given coreceptor when expressed on PBMCs and those when expressed on GHOST cells. This could occur if PBMCs express much higher levels of the coreceptor in question than do the GHOST cell lines, because it has been noted that, for some coreceptors such as STRL33, the levels of expression have a major impact on viral replication . Thus, we cannot rule out that early- and late-stage variants differ in their abilities to infect PBMCs via STRL33. Coreceptor-independent entry is not a likely explanation for these findings because these SIVs are dependent on a coreceptor for entry; SIVMneCL8 did not infect HeLa cells expressing CD4 alone, but it could infect HeLa cells expressing CD4 and CCR5 . In GHOST cells, we did observe low-level infection of the late-stage variants with various alternative coreceptors such as CCR8 and GPR1, suggesting that the late-stage viruses may have more flexibility in the coreceptors that they use for entry. However, we found no evidence that a known SIV coreceptor can be used for entry by the early- and intermediate-stage SIVMne variants, but not the late-stage variant. There were no coreceptors that distinguished SIVmac239 from the early- and/or intermediate-stage SIVMne variants, and previous results , which were confirmed here, indicate that SIVmac239 can infect PBMCs from some donors lacking CCR5. The earlier studies by Zhang et al. , which included studies with inhibitor molecules that block CCR5 entry, suggested that a coreceptor other than CCR5 may be used by SIVmac239 for replication in PBMCs, although the specific coreceptor could not be identified . Our studies indicate that the same is true for some variants of SIVMne. Collectively, these studies suggest that the coreceptor that facilitates high-level replication of some SIVs in PBMCs lacking CCR5 may be a protein that has not yet been identified as a coreceptor for primate lentiviruses. Because this coreceptor permits entry in peripheral blood cells, this suggests that the as yet unidentified coreceptor may be important for replication in relevant cell types. FIG. 1. : Levels of virus production, as measured by p27gag antigen capture assay, in PBMCs infected with SIV and SHIV. Levels of virus production, as measured by p27gag antigen capture assay, in PBMCs infected with SIV and SHIV. For these experiments, PBMCs were isolated from the whole blood of four HIV-1-seronegative donors, two with wild-type CCR5 (, 53-3; , 35-9) and two with CCR5 with a 32-bp-deletion (*, 43-1; , 07-5). After Ficoll-Hypaque centrifugation, the cells were cultured with 1 mug of PHA/ml in RPMI 1640 supplemented with 10% fetal bovine serum (RPMI complete) and 10 U of human recombinant interleukin-2/ml for 3 days. The cells were pelleted by centrifugation at 300 x g, washed with RPMI complete to remove the PHA, and cultured for an additional 24 h in RPMI complete with interleukin-2. Duplicate cultures of 2 x 106 cells/ml were infected with each virus at a multiplicity of infection of 0.01 in 1 ml of RPMI complete plus 10 U of interleukin-2/ml. The infectious titers of the viruses were determined by sMAGI assay for SIV or MAGI-CCR5 assay for SHIV162 . Fourteen hours later, the cells were pelleted at 300 x g, washed three times with RPMI complete to remove any residual cell-free virions, and resuspended in 5 ml of RPMI complete plus interleukin-2 (10 U/ml). Every 2 to 3 days, 2 ml of cell-free supernatant was removed and replaced with fresh RPMI complete supplemented with 10 U of interleukin-2/ml. Cell-free supernatant that was collected was stored at -80C and used to monitor viral replication by assaying for p27gag antigen with an enzyme-linked immunosorbent assay kit (Coulter). TABLE 1 : Peak p27gag levels in PBMCs infected with different SIVs and SHIV TABLE 2 : Infection of GHOST4 indicator cells with variants of SIVMne Backmatter: PMID- 12915528 TI - Chimeric Influenza A Viruses with a Functional Influenza B Virus Neuraminidase or Hemagglutinin AB - Reassortment of influenza A and B viruses has never been observed in vivo or in vitro. Using reverse genetics techniques, we generated recombinant influenza A/WSN/33 (WSN) viruses carrying the neuraminidase (NA) of influenza B virus. Chimeric viruses expressing the full-length influenza B/Yamagata/16/88 virus NA grew to titers similar to that of wild-type influenza WSN virus. Recombinant viruses in which the cytoplasmic tail or the cytoplasmic tail and the transmembrane domain of the type B NA were replaced with those of the type A NA were impaired in tissue culture. This finding correlates with reduced NA content in virions. We also generated a recombinant influenza A virus expressing a chimeric hemagglutinin (HA) protein in which the ectodomain is derived from type B/Yamagata/16/88 virus HA, whereas both the cytoplasmic and the transmembrane domains are derived from type A/WSN virus HA. This A/B chimeric HA virus did not grow efficiently in MDCK cells. However, after serial passage we obtained a virus population that grew to titers as high as wild-type influenza A virus in MDCK cells. One amino acid change in position 545 (H545Y) was found to be responsible for the enhanced growth characteristics of the passaged virus. Taken together, we show here that the absence of reassortment between influenza viruses belonging to different A and B types is not due to spike glycoprotein incompatibility at the level of the full-length NA or of the HA ectodomain. Keywords: Introduction : There are three types of influenza viruses: A, B, and C. Originally, the members of the influenza A virus type were defined by their serologic properties by using polyclonal antisera made against internal proteins of the viruses. All members of type A influenza viruses cross-react with polyclonal antibodies made from an influenza A virus but not with those made from an influenza B or C virus . This classification was later confirmed when the entire sequences of the genomes of influenza A, B, and C viruses were obtained. The genes coding for the surface glycoproteins can vary dramatically among different viruses belonging to the same type, but the genes coding for the internal proteins are always more related among strains of one type than those of strains belonging to two different types. Thus, type A, type B, and type C influenza viruses can be reliably classified. Sequencing has also proven that influenza type A, B, and C viruses have evolved from a common ancestor . One of the hallmarks of RNA viruses with segmented genomes is the ability to undergo reassortment. Thus, the segmented negative-strand RNA viruses readily reshuffle RNA segments in progeny viruses derived from two parent viruses infecting the same cell. For example, human influenza A viruses have been shown to undergo reassortment of genes by capturing RNA segments from avian influenza A viruses, resulting in novel human viruses with altered pathogenicity and/or potential to cause pandemics . In contrast, reassortment of genes between a type A and a type B virus has never been observed, suggesting that these viruses have become different species, if speciation is defined as having lost the ability to mate with another influenza virus, resulting in an exchange of genetic information. The absence of reassortment between influenza viruses belonging to different types has been puzzling because it has been shown that the RNA-dependent RNA polymerase of an influenza A virus can recognize the promoter sequence of an influenza B virus. Specifically, we have shown that an influenza A virus whose neuraminidase (NA) gene is flanked by the 5' and 3' noncoding sequences of the influenza B virus NS gene is viable . This virus is infectious ---albeit attenuated ---and this is proof that the absence of reassortment between influenza viruses of different types is not due to the divergence of promoters, since the heterologous influenza A virus RNA-dependent RNA polymerase appears to recognize the promoter of the influenza B virus NS gene . In vitro experiments with minigenomes of type A and type B influenza viruses have confirmed this finding . In the present study, we show that recombinant influenza A viruses whose full-length NA or HA ectodomain is derived from an influenza B virus are viable. This finding suggests that the "mixing" of influenza A and B virus proteins is compatible with the rescue of infectious virus and that the NA or HA of an influenza A virus can be functionally replaced with the corresponding protein from an influenza B virus. MATERIALS AND METHODS : Cells and viruses. | Madin-Darby canine kidney (MDCK) and 293T cells were maintained in Dulbecco modified Eagle medium (DMEM) with 10% fetal calf serum and antibiotics . Influenza A/WSN/33 (H1N1) (WSN), B/Yamagata/16/88 (Yamagata), and recombinant viruses WSN-BNA, WSN-BNA/A65, WSN-BNA/ACT, WSN-BNA/ACT-ATM, and WSN-BHA/ACT-ATM were propagated in MDCK cells in DMEM containing 0.3% bovine serum albumin. Virus titers were measured by plaque assay on MDCK cells. Parental B/Yamagata virus, as well as all recombinant viruses, containing the NA of the influenza B virus were grown and plaqued in the presence of 1 mug of TPCK (l-1-tosylamide-2-phenylmethyl chloromethyl ketone)-treated trypsin (Sigma Co.)/ml. For preparation of virus stocks used to analyze proteins by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), viruses were propagated in 10-day-old embryonated chicken eggs for 3 days at 37C. Viruses were purified from allantoic fluid by spinning the supernatant at 10,000 rpm for 30 min and by ultracentrifugation through a 30% sucrose cushion. For enzyme-linked immunosorbent assays (ELISAs), viruses were amplified in MDCK cells for 48 h (WSN, WSN-BNA, and WSN-BNA/A65) or 72 h (WSN-BNA/ACT and WSN-BNA/ACT-ATM) at 37C or for 72 h at 35C (Yamagata) and purified from cell culture supernatant by using the same protocol. Construction of plasmids. | All of the plasmid constructs described here were cloned by using the same strategy, i.e., by insertion of a PCR product into the SapI sites of the plasmid pPolI-SapI-Rb . This strategy positions the PCR product between a truncated human RNA polymerase I promoter and a hepatitis delta virus ribozyme sequence in such a way that a negative (vRNA sense) transcript is produced from the polymerase I promoter. The sequence of all PCR inserts was confirmed, and nucleotide changes that had been introduced by PCR were corrected by using QuikChange XL site-directed mutagenesis (Stratagene, La Jolla, Calif.) when appropriate. The cloning of influenza A/WSN virus NA and hemagglutinin (HA) genes into the pPolI transcription plasmid has been described previously . The HA and NA genes of B/Yamagata/16/88 were amplified by reverse transcription-PCR (RT-PCR) and cloned into pPolI-SapI-Rb to create the plasmids pPolI-HA-Yamagata and pPolI-NA-Yamagata, respectively. All NA genes and the HA genes described here were generated by PCR amplification of the appropriate gene fragments from pPolI-NA-WSN, pPolI-HA-WSN, pPolI-NA-Yamagata, and pPolI-HA-Yamagata, followed by ligation into pPolI-SapI-Rb. All primers covering the 3' and 5' noncoding regions (NCRs) of WSN NA or WSN HA contain the SapI sites that are used for inserting the NA or the HA constructs into pPolI-SapI-Rb. pPolI-BNA contains the Yamagata NA open reading frame (ORF; 466 amino acids [aa]) flanked by the 3' (19-nucleotide [nt]) and 5' (31-nt) vRNA NCRs of the WSN NA vRNA. pPolI-BNA/A65 contains the Yamagata NA ORF flanked by the 3' and 5' ends of the WSN NA vRNA. In addition, the 3' NCR is extended by insertion of additional 65 nt between the 3' NCR and the start codon of the NA. This 65-nt sequence corresponds to the amino-terminal codons of the WSN NA ORF (nt 20 to 84 of the WSN NA segment). The AUG codons present in these 65 nt were mutated to UUG. In addition to the 3' and 5' NCRs, the pPolI-BNA/ACT also contains the cytoplasmic domain (six amino-terminal amino acids) of the influenza A/WSN virus NA, whereas the transmembrane domain (29 aa) and the ectodomain (425 aa) are derived from the B virus NA. In pPolI-BNA/ACT-ATM, the cytoplasmic tail (12 aa) and the transmembrane domain (29 aa) of the influenza B virus NA were replaced by the corresponding WSN NA domains (aa 1 to 35). In addition, the 5' NCR is extended by insertion of 36 nt between the stop codon and the 5' NCR. This 36-nt sequence corresponds to nt 1346 to 1381 of WSN NA (coding for the 11 carboxy-terminal amino acids of WSN NA followed by the stop codon). pPolI-BHA/ACT-ATM contains the 3' (32 nt) and 5' (48 nt) NCRs, cytoplasmic tail (10 aa), and transmembrane domain (28 aa) of the A/WSN HA and the ectodomain (545 aa) from the B/Yamagata HA. In pPolI-BHA (583 aa) the entire ORF of the B/Yamagata HA is flanked by the A/WSN HA NCRs. Generation of recombinant viruses. | For the plasmid-only rescue of recombinant WSN viruses containing the NA or the HA of influenza B/Yamagata virus, we used a published protocol . Briefly, a coculture of MDCK and 293T cells was transfected with four expression plasmids coding for the PB1, PB2, and PA proteins, and nucleoprotein (NP) of WSN and seven pPolI transcription plasmids, each coding for one of the WSN virus vRNA segments, omitting either the NA or the HA plasmid. The eighth segment was provided in the form of one of the constructed pPolI plasmids encoding all or part of the influenza B virus NA protein (BNA, BNA/A65, BNA/ACT, or BNA/ACT-ATM) or the pPolI plasmid encoding BHA/ACT-ATM. A total of 0.5 to 1 mug of each plasmid was transfected by using Lipofectamine 2000 (Invitrogen, Carlsbad, Calif.). At 12 h posttransfection the medium was replaced by DMEM containing 0.3% bovine serum albumin, 10 mM HEPES, and 1 mug of TPCK-rreated trypsin/ml. At 3 to 5 days posttransfection, virus-containing supernatant was inoculated into 7-day-old embryonated chicken eggs. Eggs of this young age were chosen because, due to their immunological immaturity, they induce low levels of interferon after infection . Allantoic fluid was harvested after 3 days of incubation at 37C and assayed for the presence of virus by hemagglutination of chicken red blood cells or by plaque formation in MDCK cells. The NA or the HA segment of the recovered mutant viruses were analyzed by RT-PCR and sequencing. Briefly, RNA was extracted from virus-containing allantoic fluid and from cell culture supernatant by using the RNeasy Mini kit (Qiagen, Inc.). The RNA was reverse transcribed with Superscript II RNase H- (Invitrogen) according to the manufacturer's instructions. The PCR was performed by using a universal 3' NCR primer and an NA- or HA-specific 5' NCR primer. In order to confirm the presence of the recombinant HA or NAs in the recombinant viruses, RT-PCR products were subcloned into pGEM-T (Promega Co.), identified by digestion with appropriate restriction enzymes, and sequenced. Indirect immunofluorescence analysis. | Confluent monolayers of MDCK cells in 35-mm dishes were infected with recombinant viruses at a multiplicity of infection (MOI) of 2. At 12 h postinfection, cells were fixed and permeabilized by treatment with methanol (-20C) for 5 min, followed by ethanol (-20C) for 30 s. After blocking with 3% bovine serum albumin in phosphate-buffered saline (PBS), cells were incubated for 40 min with a monoclonal antibody directed against NA (9A6) or HA (15B6) (kindly provided by the Mount Sinai Hybridoma shared research facility) of influenza B/Panama/45/90 virus and a polyclonal serum raised against WSN virus (8236). After three washes with PBS containing 0.5% Triton X-100, cells were incubated for 1 h with Texas red-conjugated anti-mouse immunoglobulin G (IgG; Rockland) and with fluorescein isothiocyanate-labeled rabbit immunoglobulins (Dako A/S, Copenhagen, Denmark). After three additional washes, infected cells were analyzed by fluorescence microscopy. Growth curve of recombinant viruses. | MDCK cells (106) in 35-mm dishes were infected with recombinant viruses (WSN-BNA, WSN-BNA/A65, WSN-BNA/ACT, WSN-BNA/ACT-ATM, WSN-BHA/ACT-ATM, and rescued WSN [rWSN]) at an MOI of 0.001. Samples of supernatants were collected at different time points postinfection and plaqued on MDCK cells. PNGase F digestion and SDS-PAGE. | Carbohydrate residues were removed from glycoproteins (HA and NA) of wild-type and recombinant viruses by treating ca. 20 mug of sucrose-purified virions with PNGase F (New England Biolabs, Inc.) according to the manufacturer's instructions. Proteins were analyzed by SDS-12% PAGE and staining with Coomassie brilliant blue. ELISA. | Dynex Immulon 4HBX flat-bottom microtiter plates (Thermo Labsystems USA, Beverly, Mass.) were coated with serial twofold dilutions in PBS of wild-type influenza A/WSN and B/Yamagata viruses; with recombinant WSN-BNA, WSN-BNA/A65, WSN-BNA/ACT, and WSN-BNA/ACT-ATM viruses; or with PBS only. All viruses were amplified in MDCK cells and purified from cell culture supernatant as described above. Coating was performed in duplicates at room temperature for 15 h, starting with 0.25 mug (50 mul of a 5-mug/ml solution) of virus. Plates were then blocked with PBS buffer containing 1% bovine serum albumin at room temperature for 90 min. Coated and blocked ELISA plates were incubated with 5 mug of a monoclonal antibody recognizing influenza B/Yamagata NA (9A6)/ml at room temperature for 1 h. The plates were washed four times with PBS and then incubated with a 1:2,000 dilution of the secondary antibody (anti-mouse IgG-horseradish peroxidase [Roche Diagnostics Co., Indianapolis, Ind.]) at room temperature for 1 h. Plates were washed again four times with PBS, and the color was developed with ABTS [2,2'azinobis(3-ethylbenzthiazolinesulfonic acid)] (Roche). The optical density was read at 405 nm, and the average of duplicate samples was calculated after subtraction of the value (averaged) obtained from PBS-coated wells. RESULTS : Generation of influenza A viruses with chimeric NA or chimeric HA segments derived from an influenza B virus. | Four viruses possessing chimeric influenza B virus NA segments were generated in a background of seven WSN virus segments. The structures of the parental influenza A (WSN) and B (Yamagata) virus segments, as well as of the chimeric NA constructs, are depicted in Fig. . For all constructs, the noncoding sequences were derived from the A/WSN strain to promote efficient replication by the WSN polymerase. For the BNA construct, the entire coding region was derived from the B virus NA. The ORF of the NB that starts upstream of that of the B virus NA was eliminated. The BNA/A65 construct possesses coding regions of the entire influenza B virus NA ORF. However, this construct contains additional 65 nt derived from the 5' end of the A/WSN NA ORF (with the AUG codons present in this sequence changed to UUG). This sequence was included because sequences from within the coding regions of influenza virus RNA segments have been found to influence levels of viral RNA synthesis and packaging . Two additional constructs were generated to assess whether replacement of the transmembrane and/or cytoplasmic domain in the influenza B virus NA with those of the influenza A virus NA would alter the phenotype of the virus. In one chimera, BNA/ACT, only the cytoplasmic tail was derived from the A/WSN NA. In the other chimeric NA construct, BNA/ACT-ATM, the transmembrane and cytoplasmic domains of the influenza B virus NA segment were replaced by those of A/WSN NA. Similar to the insertion in BNA/A65, a 36-nt sequence derived from the carboxy-terminal portion of the WSN NA ORF was included in the 5' NCR of BNA/ACT-ATM . We also constructed a virus expressing a chimeric HA segment that contains the carboxy-terminal cytoplasmic and transmembrane domains of A/WSN HA (with the entire ectodomain deriving from the influenza B virus HA) . However, WSN viruses containing the entire influenza B/Yamagata virus HA segment flanked by the NCRs of the WSN virus HA (Fig. B) could not be rescued. All recombinant viruses were generated by using previously published protocols . The resulting viruses were plaque purified, amplified on MDCK cells and the chimeric segments were analyzed by RT-PCR and sequenced. By these methods, each of the viruses was found to have the expected recombinant gene (data not shown). The identity of the chimeric viruses was further demonstrated by indirect immunofluorescence . MDCK cells were infected with wild-type A/WSN, wild-type B/Yamagata, WSN-BNA, WSN-BNA/A65, WSN-BNA/ACT, and WSN-BNA/ACT-ATM viruses and WSN-BHA/ACT-ATM virus . Virus-infected cells were stained with both an antiserum raised against WSN virus and either an influenza B virus NA-specific antibody or an influenza B virus HA-specific antibody . All infected cells showed presence of the heterologous glycoprotein as well as the WSN components of recombinant viruses. Growth characteristics of recombinant viruses. | The multicycle growth of the chimeric viruses at 37C was compared in MDCK cells . rWSN was included for comparison of the growth kinetics of the recombinant viruses. rWSN virus had been rescued by using the same WSN "background" plasmids as those employed for the recombinant viruses analyzed here. Thus, these viruses are isogenic differing only in one gene. The viruses possessing the full-length influenza B virus NA proteins, BNA and BNA/A65, grew to titers similar to that of the rWSN virus. However, WSN-BNA/ACT and WSN-BNA/ACT-ATM viruses which express chimeric proteins possessing portions of the WSN NA protein displayed impaired growth compared to the rWSN virus . This experiment was repeated with similar results. In contrast to recombinant WSN viruses expressing influenza B virus NAs, the WSN virus expressing the B virus HA (WSN-BHA/ACT-ATM) produced an HA titer of <=8 compared to an HA titer of 256 for wild-type A/WSN virus and did not form plaques in MDCK cells. In order to increase the growth properties of this recombinant virus, it was serially passaged in MDCK cells. Virus obtained from the sixth passage showed a titer of 108 PFU/ml and produced visible plaques in MDCK cells. This WSN-BHA/ACT-ATM (P6) virus was used for further analysis. The growth dynamics were determined in MDCK cells and compared to those of rWSN . Remarkably, this virus grows to titers comparable to that of wild-type A/WSN virus, if not to higher titers. In order to clarify whether adaptive amino acid changes had occurred in the genes coding for BHA/ACT-ATM or for WSN NA that could account for the markedly increased growth characteristics of P6, we cloned both genes by RT-PCR. Four independent PCR products of each full-length HA, as well as of NA, were sequenced. Two of the four RT-PCR products of the A/WSN NA had one mutation each (L95R and E366G), but there is no evidence that any of these mutations has biological significance. Interestingly, in all four RT-PCR products of BHA/ACT-ATM the same amino acid change was found. This mutation replaces histidine at position 545 with tyrosine (H545Y). aa 545 is the carboxy-terminal amino acid of the influenza B HA ectodomain, immediately followed by the transmembrane region . Analysis of HA sequences available in the database (Influenza Sequence Database, Los Alamos National Laboratory ) revealed that all influenza B virus HAs contain a histidine at this position, whereas in all influenza A virus H1 HAs the last amino acid of the ectodomain is a tyrosine. In order to investigate whether the improved growth characteristics of WSN-BHA/ACT-ATM virus of the sixth passage compared to the originally rescued virus was a result of the single amino acid change H545Y in BHA/ACT-ATM, we introduced this mutation into pPolI-BHA/ACT-ATM by site-directed mutagenesis, thus creating pPolI-BHA/ACT-ATM (H545Y). This plasmid was used for the rescue of WSN-BHA/ACT-ATM (H545Y) virus. Recombinant WSN-BHA/ACT-ATM (H545Y) virus was plaque purified and amplified on MDCK cells, and the presence of the mutated nucleotide sequence in the chimeric HA segment of this recombinant virus was confirmed by RT-PCR and sequencing of the RT-PCR product (data not shown). This recombinant WSN virus containing the mutated HA chimera (WSN-BHA/ACT-ATM [H545Y]) was found to grow to the same titers as the passaged WSN-BHA/ACT-ATM (P6) virus , suggesting that the tyrosine at the boundary of ectodomain and transmembrane domain of the HA protein may be responsible for the improved growth characteristics. Incorporation of the BHA/ACT-ATM protein into virions. | We wanted to investigate whether the extent of incorporation of the chimeric HA protein into virions differed from that of wild-type influenza A virus HA. To this end, we amplified WSN and WSN-BHA/ACT-ATM (P6) viruses in 10-day-old embryonated chicken eggs and partially purified viruses from the allantoic fluid as described in Materials and Methods. To obtain a better resolution of the bands corresponding to NP, HA1, and NA, virions were treated with PNGase F to remove N-linked carbohydrate residues from the polypeptides. The proteins of both treated and untreated virions were separated by SDS-12% PAGE and then visualized by staining with Coomassie brilliant blue. As expected, the HA1 subunit of the chimeric protein, which is 19 aa longer than the WSN HA1, runs slower in the gel . The incorporation of BHA/ACT-ATM into recombinant virions is comparable to that of wild-type influenza A virus HA based on the amount of NP and M1 present in the preparations. Levels of influenza B NA proteins in chimeric viruses. | In order to determine the relative amounts of the NA proteins in the rescued viruses, we performed ELISAs. Influenza B/Yamagata and A/WSN viruses, as well as recombinant WSN-BNA, WSN-BNA/A65, WSN-BNA/ACT, and WSN-BNA/ACT-ATM viruses, were amplified in MDCK cells and purified from cell culture supernatants as described in Materials and Methods. The purity of the virus preparations was confirmed by SDS-PAGE and silver staining (not shown). Equivalent amounts of viruses were then used in the ELISAs. ELISA plates were coated with serial twofold dilutions of recombinant and parental viruses, starting with 0.25 mug as described in Materials and Methods. The relative amounts of the influenza B virus NA proteins incorporated into virions was measured as intensity of alphaB NA antibody binding to virions . Both recombinant WSN viruses containing the entire influenza B virus NA protein (WSN-BNA and WSN-BNA/A65 viruses) showed an approximately twofold (2.2- and 2.4-fold, respectively)-higher NA protein content compared to wild-type influenza B/Yamagata virus. Results obtained from the measurement of the NA activity of equivalent amounts of the viruses by using 2'-(4-methyl-umbelliferyl)-alpha-d-N-acetylneuraminic acid (Mu-NANA) as a substrate support this finding: the NA activity of WSN-BNA virus and WSN-BNA/A65 virus is 1.8- and 2-fold, respectively, greater than that of B/Yamagata virus. However, in chimeric WSN viruses expressing BNA/ACT or BNA/ACT-ATM, the levels of NA protein incorporation into virions were 35 and 14%, respectively, of that of wild-type influenza B virus. Again, the content of the NA as measured by ELISA corresponds roughly to the NA activity as determined by NA assay with Mu-NANA as a substrate, i.e., the NA activities of WSN-BNA/ACT virus and WSN-BNA/ACT-ATM virus were 27 and 29% of that found for wild-type influenza B/Yamagata virus. Values obtained from control WSN-coated wells did not differ from those of PBS-coated wells (data not shown). DISCUSSION : Although type A and type B influenza viruses are known to cocirculate in humans, the formation of reassortant viruses has never been observed in vivo or, for that matter, in vitro . Given that the HA and the NA of influenza A and B viruses have similar functions, we explored the possibility that the type B glycoproteins could functionally replace those of type A influenza viruses. Here we report the generation of infectious influenza A viruses expressing either the HA or the NA of an influenza B virus. The NA protein of influenza B/Lee/40 virus had been shown earlier to complement the influenza A virus NA in an NA-deficient type A virus (NWS-Mvi) and to be packaged into virions when overexpressed in cells infected with NWS-Mvi . However, the formation of an infectious (nondefective) chimeric virus expressing an influenza B virus glycoprotein has not been demonstrated before. We employed a plasmid-only transfection system for the generation of transfectant influenza A viruses expressing either a full-length influenza B virus NA, chimeras between type A and type B NAs, or a chimeric HA protein. Both recombinant viruses expressing the full-length influenza B virus NA (WSN-BNA and WSN-BNA/A65 viruses) grow to titers comparable to that of the isogenic influenza A virus . Thus, influenza B virus NAs carrying the NCRs of the corresponding influenza A virus segment can fully replace the function of the type A NA in viruses growing in tissue culture. The phenotype of these viruses in animal models has not been investigated. The amount of NA per total viral protein appears to be twice that found in purified influenza B virus preparations, as determined by ELISAs . However, it is not clear whether the increased amount of NA content per total micrograms of protein reflects a higher incorporation rate, since we have not determined the amount of NA expressed on the surface of infected cells. The addition of 65 nt derived from the amino-terminal portion of the influenza A virus NA ORF to the 3' NCR of the recombinant NA in BNA/A65 does not appear to change the phenotype of this virus. In a recently published report it has been postulated that the amino-terminal 21 nt of the NA coding region are critical for the packaging of the NA vRNA segment . It is likely that a similar packaging signal is present in the ORF of the influenza B virus NA, since the WSN-BNA virus does not appear to have a lower growth rate. In contrast to the recombinant A/WSN viruses expressing full-length type B NAs, the rescued viruses containing chimeric NA proteins (WSN-BNA/ACT and WSN-BNA/ACT-ATM viruses) are compromised in growth in MDCK cells. They grow to an at least 1-log-lower titer than wild-type influenza A virus . The ELISA shows that these purified viral preparations possess only 35 and 14% of the NA protein compared to wild-type virus . The reason for the lower NA content in WSN-BNA/ACT and WSN-BNA/ACT-ATM viruses remains unknown. One possible explanation could be that the protein is less stable due to replacement of the cytoplasmic tail or of the cytoplasmic tail plus the transmembrane region of the influenza B virus NA (with those of the type A NA). Alternatively, packaging of the chimeric NA proteins could be decreased, as has been shown with transfectant influenza A viruses carrying point mutations in the cytoplasmic tail of the NA . However, studies performed in our laboratory with NA/TAIL negative-strand viruses showed that deletion of the entire cytoplasmic tail did not result in reduced packaging of the NA proteins into virions . In addition to the four recombinant viruses expressing all or part of the influenza B virus NA protein, we rescued an influenza A virus containing a chimeric HA protein (WSN-BHA/ACT-ATM virus). This virus did not grow to high titers and failed to form plaques in MDCK cells, but after it was passaged in MDCK cells we obtained a WSN-BHA/ACT-ATM (P6) virus that forms plaques and grows to titers comparable to that of wild-type influenza A virus . We suggest that the occurrence of an adaptive amino acid change confers the increased growth to the recombinant WSN-BHA/ACT-ATM (P6) virus. Both the HA and the NA of different influenza A viruses have been shown to acquire adaptive mutations as a result of altered HA-NA combinations in a virus . These mutations led to the restoration of a functional match of the two influenza viral glycoproteins. It was assumed that mutations would result in a functional balance between the two opposing activities of the HA and NA: one being involved in binding to sialic acid containing receptors and the other one being involved in the removal of sialic acids from virion and cell surfaces, respectively . Our analysis of the passaged WSN-BHA/ACT-ATM (P6) virus showed that there was no consistent change in the NA, but all HA clones shared a mutation (H545Y). According to the structural organization of influenza A virus HA proteins proposed by Nobusawa et al. , the amino acid at position 545 corresponds to the carboxy-terminal amino acid of the influenza A virus HA ectodomain, and the transmembrane domain starts at position 546. Introduction of this mutation into BHA/ACT-ATM and rescue of recombinant WSN-BHA/ACT-ATM (H545Y) showed that this amino acid change could fully restore the growth characteristics of the passaged virus (P6). The tyrosine at this position of the HA protein appears to play an important role for productive infection of recombinant influenza A virus expressing this chimeric HA protein in tissue culture. One possible explanation may be that the tyrosine close to the transmembrane region may be needed for maintaining the stability of the BHA/ACT-ATM protein. Alternatively, this tyrosine might be an important functional component of the transmembrane domain of the wild-type influenza A virus HA. An additional chimeric construct coding for the full-length type B HA flanked by the influenza A virus HA NCRs has been generated. However, this HA segment could not be rescued into the A/WSN virus background. Since the lack of reassortment between influenza A and B viruses is not due to incompatibilities at the level of their RNA promoters or of their NA proteins (the present study), why have A/B NA gene reassortants never been obtained? One possibility is that the expression of NB from the wild-type B virus NA gene interferes with influenza A virus replication. Another explanation for the lack of reassortment may be due to incompatibilities at the level of the ribonucleoprotein complexes. For instance, NP-RNA complexes of influenza A virus are not efficiently replicated by the RNA-dependent RNA polymerase of influenza B virus and vice versa . Thus, during mixed influenza A and B virus infections it is likely that RNAs are only replicated by the homologous viral RNA polymerase. A heterologous ribonucleoprotein complex containing P (polymerase) proteins and NP derived from influenza A and B viruses would be inactive and would not be packaged into viruses. In any case, our results clearly show that lack of influenza A/B virus reassortment is not due to an incompatibility at the level of their NA proteins or at that of the HA ectodomain. FIG. 1. : NA and HA genes used to generate recombinant influenza A/WSN viruses. NA and HA genes used to generate recombinant influenza A/WSN viruses. (A) Schematic diagram of the influenza B/Yamagata/16/88 and A/WSN/33 parental and chimeric NA genes showing the ORFs of the NAs. NCRs are indicated by lines. CT, cytoplasmic tail; EC, ectodomain; TM, transmembrane domain. In BNA/A65 and BNA/ACT-ATM the NCRs have been extended by insertion of 65 nt into the 3' NCR (nt 20 to 84) and insertion of 36 nt (nt 1298 to 1333) into the 5' NCR, respectively. Flu B wt NA, influenza B/Yamagata wild-type virus NA; Flu A wt NA, influenza A/WSN wild-type virus NA. The asterisk in Flu B wild-type NA indicates the position of the start codon of the NB protein which is upstream of that of the NA protein. This upstream sequence coding for the amino terminus of the NB protein is not included in the chimeric constructs. (B) Schematic representation of the influenza B/Yamagata/16/88 and A/WSN/33 virus HA and the chimeric BHA-ACT/ATM segments. Flu B wt HA, influenza B/Yamagata wild-type virus HA; Flu A wt HA, influenza A/WSN wild-type virus HA. The ORFs are shown as boxes. The NCRs are represented as lines. CT, cytoplasmic tail; EC, ectodomain; TM, transmembrane domain. For details, see Materials and Methods. FIG. 2. : Immunofluorescence analysis of MDCK cells infected with recombinant viruses. Immunofluorescence analysis of MDCK cells infected with recombinant viruses. (A) Recombinant influenza A viruses expressing recombinant influenza B virus NAs. Subpanels: A and D, wild-type influenza B/Yamagata virus, B and C, wild-type influenza A/WSN virus; E and F, WSN-BNA virus; G and H, WSN-BNA/A65 virus; I and J, WSN-BNA/ACT virus; K and L, WSN-BNA/ACT-ATM virus. (B) Recombinant influenza A virus expressing the HA ectodomain of influenza B virus. Subpanels: A and B, mock infection; C and D WSN-BHA/ACT-ATM virus. alpha Flu B NA and alpha Flu B HA, monoclonal antibodies recognizing influenza B/Yamagata virus NA (9A6) or HA (15B6), respectively; alpha Flu A, polyclonal rabbit antiserum (i.e., 8236) raised against A/WSN virus. FIG. 3. : Growth characteristics of recombinant influenza viruses. Growth characteristics of recombinant influenza viruses. MDCK cells were infected at an MOI of 0.001 with rWSN , WSN-BNA , WSN-BNA/A65 , WSN-BNA/ACT , and WSN-BNA/ACT-ATM viruses (A) or with wild-type WSN and WSN-BHA/ACT-ATM (P6 and H545Y ) viruses (B). Plaque assays were performed on MDCK cells. h p.i., hours postinfection. FIG. 4. : Protein gel analysis of influenza A/WSN and recombinant WSN-BHA/ACT-ATM (P6) viruses. Protein gel analysis of influenza A/WSN and recombinant WSN-BHA/ACT-ATM (P6) viruses. Proteins of purified viruses, either untreated (-) or treated with PNGase F (+), were separated by SDS-PAGE on a 12% gel and stained with Coomassie brilliant blue. Positions of untreated (-) and treated (+) proteins are indicated. The band of added PNGase F is shown to the right by an arrow. The positions of the deglycosylated influenza A/WSN HA1 (Flu A HA1) and WSN-BHA/ACT-ATM (Flu B HA1) viruses are indicated by arrows on the right. HA0, HA precursor protein; HA2, HA2 HA subunit. FIG. 5. : Content of influenza B virus NA in purified chimeric viruses as determined by ELISA. Content of influenza B virus NA in purified chimeric viruses as determined by ELISA. ELISA plates were coated with 50 mul of serial twofold dilutions of recombinant purified WSN-BNA , WSN-BNA/A65 , WSN-BNA/ACT , WSN-BNA/ACT-ATM , and wild-type B/Yamagata viruses . The starting concentration was 5 mug/ml. Plates were incubated with an influenza B virus NA-specific monoclonal antibody (9A6) and developed with a secondary horseradish peroxidase-conjugated anti-mouse IgG. Backmatter: PMID- 12915544 TI - Human Cytomegalovirus US2 Causes Similar Effects on Both Major Histocompatibility Complex Class I and II Proteins in Epithelial and Glial Cells AB - The human cytomegalovirus (HCMV) glycoprotein US2 specifically binds to major histocompatibility complex (MHC) class I heavy chain (HC) and class II proteins DRalpha and DMalpha, triggering their degradation by proteasomes. Effects of US2 on class II proteins were originally characterized in HCMV- or adenovirus vector-infected U373 astroglioma cells. Here, we have extended characterization of US2-mediated degradation of class II DRalpha to two other cell lines, including biologically relevant epithelial cells. Comparison of the effects of US2 in cells expressing both class I and II proteins demonstrated only a slight preference for class I HC. Moreover, US2 caused degradation of DRalpha and DMalpha when these proteins were expressed by transfection without DRbeta, invariant chain (Ii), or DMbeta. Therefore, US2 binds to alpha chains of DR and DM and triggers endoplasmic reticulum degradation without formation of class II DR alphabeta/Ii or DM alphabeta complexes. Similar levels of degradation of class II alpha were observed in cells expressing vastly different amounts of class II, suggesting that cellular factors, other than class II, were limiting. We concluded that US2 has broad effects in a variety of cells that express both class I and II proteins and is relevant to HCMV infection in vivo. Keywords: Introduction : Human cytomegalovirus (HCMV) is a ubiquitous beta-herpesvirus that causes lifelong infections characterized by low-level persistence, latency, and bouts of reactivation . HCMV expresses a plethora of proteins that allow escape from host immunity, and these likely play a major role in persistence. The HCMV genome contains a cassette of genes in the US2 to US11 region that encode eight homologous glycoproteins. Five of these, US2, US3, US6, US10, and US11, inhibit the major histocompatibility complex (MHC) class I pathway and antigen presentation to CD8+ T lymphocytes . We have proposed that these and other HCMV immune evasion strategies function during a specific time of viral life cycle, or in specific cells, creating a window of opportunity for transient virus replication and spread (reviewed in references , , and ). HCMV not only blocks MHC class I-mediated presentation of viral antigens to CD8+ T cells but also inhibits the class II antigen presentation pathway, allowing the virus to avoid detection by CD4+ T cells (reviewed in references and ). MHC class II proteins are normally expressed by professional antigen-presenting cells (APCs) like B cells, macrophages, and dendritic cells (DCs). HCMV may infect certain APCs, especially monocyte-macrophages, but only at a specific stage of activation and often abortively or inefficiently (reviewed in reference ). HCMV also replicates in epithelial cells of alimentary and upper respiratory tracts, vascular endothelial cells, and glial cells (reviewed in reference ), cells that express MHC class II naturally, especially after induction by gamma interferon (IFN-gamma). HCMV inhibits IFN-gamma signaling and induction of class II promoters, reducing the inducible expression of class II genes . However, even before this inhibition of class II transcription, HCMV early glycoproteins US2 and US3 inhibit class II antigen presentation by destroying or abolishing the functions of class II proteins . US2 binds to class II DR and causes rapid and efficient proteosome-mediated degradation of only the alpha chain of the class II DR alphabeta complex . US2 also causes degradation of the alpha chain of DM , an MHC class II complex required for loading of antigenic peptides onto class II DR complexes (reviewed in reference ). HCMV US3 binds to class II DR alphabeta heterodimers, inhibiting binding of the invariant chain (Ii), leading to intracellular mislocalization and reduced peptide loading of DR complexes . Our laboratory's previous characterization of the effects of US2 on class II proteins involved U373 astroglioma cells expressing the human class II transactivator (CIITA) and infected with either HCMV or with replication-defective adenovirus (Ad) vectors . Recent studies involving bacterially produced, in vitro-translated, or Ad-expressed US2 lead to suggestions that US2 might not cause degradation of class II proteins in vitro or in DCs . These authors also suggested that US2 might not generally affect class II in other cells and that effects might be limited to U373 cells, and they suggested that glial cells do not normally express class II. In fact, glial cells and other cell types that HCMV infects frequently in vivo, including epithelial and endothelial cells, express class II proteins, although generally lower levels are produced compared with DCs and expression is often upregulated by IFN-gamma (reviewed in reference ). DCs and other so-called professional APCs such as monocyte-macrophages are probably much less relevant in terms of the effects of HCMV US2 and US3. These cells are poorly infected by HCMV, expression of HCMV gene products is often relatively low, and class II proteins are expressed at high levels . Rehm et al. found insufficient levels of US2 were expressed by using Ad vectors, so that neither class I or class II proteins were substantially affected. It is not clear whether US2 might have caused degradation of class II in DCs had enough US2 been delivered. To characterize whether US2 has general effects on MHC class II proteins in a variety of cells, we characterized US2-mediated degradation of class I and II proteins in several different cell lines, including two human epithelial cell lines. US2 showed only a slight preference for class I compared with class II proteins. US2 triggered efficient degradation of the class II DRalpha and DMalpha proteins when these were expressed without the DRbeta or DMbeta chains. Cells with substantial differences in levels of class II proteins displayed similar degradation at any given level of US2, supporting the hypothesis that cellular factors apart from MHC proteins are limiting in US2-induced endoplasmic reticulum (ER) degradation. MATERIALS AND METHODS : Cells. | All cell lines were propagated in media supplemented with 10% fetal bovine serum (HyClone, Logan, Utah), 2 mM l-glutamine, 100 U of penicillin/ml, and 100 mug of streptomycin (Biowhittaker, Walkersville, Md.)/ml, unless otherwise stated. HeLa, U373-MG (American Type Culture Collection, Manassas, Va.), and 293 (Microbix, Toronto, Ontario, Canada) cells were propagated in Dulbecco's modified Eagle's medium (DMEM; Mediatec, Herndon, Va.). His16 and Neo6 cells were derived by stable transfection of the human astroglioma cells U373-MG with the human class II transactivator CIITA and pSV2His or pSV2Neo, respectively . His16 cells were propagated in DMEM lacking histidine (University of California San Francisco Cell Culture Facility, San Francisco, Calif.) and supplemented with 0.5 mM histidinol (Sigma, St. Louis, Mo.). Neo6 cells were grown in DMEM supplemented with 200 mug of G418 sulfate (GIBCO)/ml. HeLa-CIITA cells were a gift from Philip Benaroch, Institut Curie, Paris, France, and were grown in DMEM supplemented with 300 mug of hygromycin (Sigma)/ml. 2E12 cells , which are AGS human gastric carcinoma cells stably transfected with CIITA, were a gift from Lindsey Hutt-Fletcher, University of Missouri, Kansas City, and were propagated in Ham's F12 medium supplemented with 0.4 mug of puromycin/ml. Recombinant Ad's. | Replication-defective (E1-) Ad vectors expressing HCMV US2, US3, US7, and US11 have been described previously . Briefly, the relevant genes were amplified by PCR from HCMV strain AD169 and inserted into plasmid pDeltaE1sp1Btet . 293 cells were cotransfected with pJM17, a plasmid containing the Ad type 5 genome , and each of the above plasmids. Recombinant Ad vectors were screened by PCR. The resulting viruses, AdtetUS2, AdtetUS3, AdtetUS7, and AdtetUS11, were plaque purified twice to eliminate wild-type virus. For expression of the glycoproteins, cells were coinfected with these viruses and a second virus, Adtet-trans, using 20% of the amount of AdtetUS virus. Infections were carried out in the absence of tetracycline, which represses the tet-trans promoter. AdBHG10T7Rbetagal (Adbetagal [49]) was obtained from Frank Graham, McMaster University, Hamilton, Ontario, Canada. Adbetagal was constructed by recombination between a shuttle plasmid containing the T7 promoter, beta-galactosidase gene, and simian virus 40 poly(A) sequences and the circular Ad5 genomic plasmid pBHG10, as described elsewhere . Other recombinant Ad's (E1-, E3-) encoding US11 and US2 were obtained from Armin Rehm, Max Delbruck Center for Molecular Medicine, Berlin, Germany. Briefly, the genes were coupled to the HCMV immediate-early promoter and bovine growth hormone poly(A) signal in a shuttle vector. Recombination with the viral vector was performed in Escherichia coli for US11 and in 293 cells for US2 . We refer to these viruses as AdUS2 and AdUS11, as opposed to our tetracycline transactivator-regulated AdtetUS viruses. All Ad vectors were propagated and titers were determined on 293 cells as described previously . Plasmids and antibodies. | Eukaryotic expression constructs encoding human leukocyte antigen (HLA)-DRalpha or HLA-DRbeta were gifts from Elizabeth Mellins, Stanford University, Palo Alto, Calif. Rabbit antiserum to the N-terminal peptide of US2 has been described previously . The monoclonal antibody (MAb) HC10, specific for class I heavy chain (HC) , was obtained from Hidde Ploegh, Harvard Medical School, Boston, Mass. MAb's to HLA-DRalpha (DA6.147) and HLA-DRbeta (HB10A) were gifts from Peter Cresswell, Yale University, New Haven, Conn. The MAb to HLA-DMalpha was a gift from John Trowsdale, Cambridge University, Cambridge, United Kingdom. The MAb (B3/25) to human transferrin receptor was obtained from Roche Molecular Corp., Indianapolis, Ind. Transfections. | For transient expression of class II proteins, HeLa cells were transfected with plasmid DNA containing relevant genes by using the Geneporter 2 transfection reagent, according to the manufacturer's recommendations (Gene Therapy Systems, San Diego, Calif.). Briefly, 106 cells were seeded into 60-mm Falcon tissue culture dishes (Becton Dickinson, Franklin Lakes, N.J.) and transfected the next day with 4 to 6 mug of DNA in serum-free Opti-MEM (GIBCO). For expression of HCMV glycoproteins, the transfected cells were infected with Ad vectors 48 to 60 h posttransfection after the cells had reached 4 x 106 cells/dish. Metabolic labeling of cells and immunoprecipitation of proteins. | Cells were radiolabeled after 18 to 20 h of infection with recombinant Ad vectors. In general, cells were plated in 150-mm Nunclon tissue culture dishes (Nalge Nunc International, Rochester, N.Y.) at a density of 6 x 106 (His16, Neo6, or 2E12) or 15 x 106 (HeLa-CIITA) cells per dish and infected the next day. Infected or control cells were washed in DMEM lacking methionine and cysteine (GIBCO) and then incubated for 1 h in the same medium containing 1% dialyzed fetal bovine serum (starvation medium). Cells were labeled for various time periods in starvation medium supplemented with [35S]methionine-cysteine (Promix; Amersham Pharmacia Biotech, Piscataway, N.J.). For longer periods, e.g., 3 h, adherent cells were labeled in tissue culture dishes with 50 to 100 muCi of [35S]methionine-cysteine/ml. For short pulse labels, cells were trypsinized, washed twice in DMEM lacking methionine and cysteine, and labeled in suspension with 500 muCi of [35S]methionine-cysteine/ml. To chase the label, cells were incubated in medium containing 10-fold excess methionine and cysteine. The cells were then washed with cold phosphate-buffered saline, and cell extracts were made using Nonidet P-40-deoxycholate (NP-40-DOC) lysis buffer (50 mM Tris-HCl [pH 7.5], 100 mM NaCl, 1% NP-40, 0.5% DOC, 1 mg of bovine serum albumin [BSA]/ml, and a cocktail of protease inhibitors). For denaturation of protein complexes, cell extracts were made in small volumes with sodium dodecyl sulfate (SDS) lysis buffer comprised of 50 mM Tris (pH 7.4), 100 mM NaCl (Tris-saline), 1% SDS, 1 mg of BSA/ml, and a cocktail of protease inhibitors and boiled for 10 min. SDS was then diluted 10-fold by addition of Tris-saline containing 1% Triton X-100, 1 mg of BSA/ml, and a cocktail of protease inhibitors. Proteins of interest were immunoprecipitated by incubating cell extracts with the appropriate antibody for 2 h and subsequently with protein-A agarose beads for 2 h, as described previously . Electrophoresis and autoradiography. | Immunoprecipitated proteins were eluted by boiling in 2% SDS and then subjected to electrophoresis using 10 to 12% polyacrylamide gels, as described elsewhere . The gels were fixed in 30% acetic acid and 10% methanol and treated with Enlightning (New England Nuclear, Beverly, Mass.), dried, and exposed to X-ray film (Eastman Kodak Company, Rochester, N.Y.) or PhosphorImager screens (Molecular Dynamics, Sunnyvale, Calif.). Protein bands were quantified by using the PhosphorImager BAS2500 system (Molecular Dynamics). RESULTS : Class II complexes are lost in a dose-dependent manner in US2-expressing epithelial and glial cells. | His16 cells are U373 cells transfected with CIITA and express substantial quantities of class II alpha, beta, and Ii, and they efficiently present antigens to CD4+ T cells . To extend our observations beyond His16 cells, we obtained HeLa cervical carcinoma and AGS gastric carcinoma cells that had both been stably transfected with CIITA . Initially, the levels of expression of class II proteins were examined in a pulse-chase labeling format. His16, HeLa-CIITA, and 2E12 (AGS-CIITA) cells were labeled in suspension for 2, 4, or 8 min, and the label was chased for 0 or 20 min. All three cell types expressed class II proteins, although at lower levels in the HeLa-CIITA cells . In longer labeling periods and in chase samples, the beta chain was coprecipitated with alpha chains by MAb DA6.147, an antibody that recognizes the cytoplasmic domain of the class II alpha chain but that can also precipitate alphabeta complexes. With all the cell lines, the detection of class II proteins increased with longer periods of labeling or chase , due in part to acquisition of antibody epitopes . Normalization of class II expression to cell number indicated that HeLa-CIITA and 2E12 cells express ~10% and ~60% of class II proteins compared to His16 cells. The respective parental cell lines expressed much lower levels of class II proteins (data not shown), although it should be noted that U373 cells (without CIITA transfection) less efficiently present antigens to CD4+ T cells . Replication-defective Ad vectors have been used to study the biochemical effects of HCMV proteins in U373 cells because infection of U373 cells with HCMV is frequently inefficient. Here, we have used an Ad vector to deliver US2 into HeLa-CIITA and 2E12 epithelial cells, as well as His16 glial cells, and in every case class II was lost in a US2 dose-dependent manner . The control Ad vector, Adtet-trans, slightly increased the amount of class II in a dose-dependent fashion (data not shown), similar to previous observations from our laboratory . Since it has been suggested that the effects of US2 were related to overexpression of a glycoprotein that accumulated in the ER or to consequences of CIITA expression , we compared the effects of US2 to HCMV US3 and US7, homologous glycoproteins that do not cause degradation of MHC proteins . There was no obvious effect of expressing US3 or US7 in either His16 or HeLa-CIITA cells . Similar results were also obtained with Neo6 and 2E12 cells (data not shown). Therefore, US2 causes degradation of class II alpha with similar doses in several cell types. Degradation of class II alpha chains in cells expressing only DRalpha or DMalpha. | US2 triggers degradation of the class II alpha chain shortly after its synthesis while not affecting beta and Ii chains . However, it is not clear whether US2 binds preferentially to alphabeta or alphabeta/Ii complexes or whether beta or Ii are required for degradation. To analyze this further, we transiently transfected HeLa cells with plasmids encoding just DRalpha, DRbeta, or DMalpha chains and then delivered various doses of US2. When expressed alone in HeLa cells, DRalpha was degraded in a dose-dependent manner but DRbeta was not affected by US2 expression . When both alpha and beta were cotransfected into HeLa cells (HeLa/DRalpha/DRbeta), both the alpha and beta chains disappeared because an antibody specific to the alpha chain was used and the alpha chain was lost. US2 also caused degradation of DMalpha when DMalpha was expressed without DMbeta in HeLa cells (Fig. , lower panel). Therefore, US2 can bind to and promote degradation of DRalpha and DMalpha without other components of these class II complexes. Evidence for a limiting cellular factor. | Previous studies from our laboratory, involving mutant forms of US2, demonstrated that binding of US2 to MHC class I and II proteins was not sufficient for their degradation . This and other lines of research have suggested that cellular factors or events beyond binding of US2 to MHC proteins are required for the degradation process. This is an important area of investigation because US2- and US11-mediated degradation of MHC proteins is a model for ER degradation in mammalian cells . His16 and 2E12 cells expressed substantially higher levels of class II proteins than HeLa cells transfected with the DRalpha gene. Therefore, we reasoned that US2 might cause much more rapid degradation in transfected HeLa cells, unless some cellular factor was limiting. First, to quantify the levels of expression of DRalpha in His16 cells and transiently transfected HeLa cells, cells were labeled for 3 h and DRalpha was immunoprecipitated using the alpha-specific antibody DA6.147. In the left panel of Fig. , class II alpha/beta/Ii complexes were precipitated using MAb DA6.147 from His16 cell extracts made using NP-40-DOC lysis buffer. Under these conditions, the alpha and Ii chain bands form a tight doublet. In the middle lane, class II complexes were first denatured with SDS to separate Ii and alpha chains and then immunoprecipitated with MAb DA6.147. Clearly, DA6.147 much less efficiently precipitated denatured DRalpha under these conditions, but the band indicates the position of the alpha chain in this gel. In the right lane, DRalpha was precipitated from transfected HeLa extracts made using NP-40-DOC buffer (there is no beta or Ii in these cells). To compare the amount of alpha chain in the left and right lanes, we used a PhosphorImager to quantify the entire alpha/Ii band from His16 cells and then calculated the contribution of the alpha chain to the band intensity, based on the relative quantities of methionine and cysteine. This difference, combined with the fact that fourfold more HeLa-DRalpha cells were used, established that His16 cells expressed 20- to 40-fold more class II DRalpha than was the case in transfected HeLa cells. In Fig. we established that 2E12 cells express approximately 60% the class II DRalpha that is expressed in His16 cells. Different doses of US2 were delivered into 2E12, His16, or HeLa cells transfected with DRalpha or DRbeta. The loss of DRalpha was similar in the three different cells , even though there were 10- to 40-fold more class II proteins expressed in 2E12 and His16 cells. No loss of DRbeta was observed. Expression of US2 was very similar in His16 and transfected HeLa cells at any dose of AdtetUS2 , and similar levels of US2 were observed in 2E12 cells (data not shown). Further, in His16 cells and HeLa cells transfected with the DMalpha gene, degradation of class II DMalpha was similar or identical at any given dose of US2 . Thus, degradation of both DRalpha and DMalpha occurred with a similar US2 dose response, even though substantially different amounts of the MHC class II proteins were present in cells. This supports the existence of cellular factors other than MHC proteins and US2 that are involved and limiting in US2-induced ER degradation of MHC proteins. Relative effects of US2 on class I HC versus class II alpha chain. | Based on in vitro experiments involving recombinant forms of US2 , it was suggested that US2 only affects MHC class I HC and has little or no effect on class II alpha. In mammalian cells, however, our group has found that US2 affects both proteins . The relative effects of US2 on MHC class I and II proteins have not been compared in cells that express both and where different doses of US2 could be delivered. His16 cells and Neo6 cells, a second CIITA-transfected U373 cell line that expresses approximately half the class II expressed in His16 cells , were analyzed. When these cells were infected with low to intermediate doses of AdtetUS2 (10 to 40 PFU/cell), more class I HC was degraded than class II DRalpha in both His16 and Neo6 cells . However, the preference for HC was only 1.5- to 2-fold greater than that for class II alpha. At higher doses of US2 (80 PFU/cell), both proteins were efficiently degraded. Comparison of Ad vectors expressing US2. | The report of Rehm et al. suggested that an Ad vector expressing US2 did not affect the stability of class II DR proteins in DCs. In preliminary experiments, we were unable to substantially express US2 or US3 by using Ad vectors in monocyte-macrophages or DCs cultured from blood (R. Tomazin and N. Hegde, unpublished data). Low levels of US2 were expressed in these cells, but there was no obvious effect on MHC proteins or antigen presentation (Tomazin and Hegde, unpublished). Our Ad vectors, e.g., AdtetUS2 and AdtetUS11, rely on the use of the HCMV immediate-early (IE) promoter that is upregulated by coinfecting cells with a second Ad vector expressing a tetracycline transactivator protein . Rehm et al. constructed AdUS2 and AdUS11 that rely on constitutive activity of the HCMV IE promoter . We found that Ad vectors utilizing constitutive promoters coupled to HCMV US2 to US11 genes produced poor yields, apparently because these ER-retained US2 to US11 proteins were toxic to 293 cells used to produce the Ad viruses (reference and data not shown). To compare our Ad vectors with those of Rehm et al., His16 cells were infected with AdtetUS2 or AdUS2 of Rehm et al. . Both US2-expressing Ad vectors caused degradation of class II DRalpha and class I HC in these cells . However, AdUS11 also reduced DRalpha in these cells, whereas our AdtetUS11 did not. Since AdUS11 grew to significantly lower titers, larger amounts of virus stock were required in order to obtain similar levels of US11 expression (data not shown). Moreover, AdUS11 inhibited expression of an unrelated cellular protein, the transferrin receptor, but this was not the case with AdtetUS11 (Fig. ; compare to Ad-betagal lanes). US2 delivered by using AdUS2 also markedly reduced the stability of class II DRalpha in HeLa-CIITA and 2E12 epithelial cells (data not shown). We concluded that US2 delivered by both Ad vectors caused degradation of both class I and II proteins in a variety of cells, and this cannot be the reason for any discrepancies. DISCUSSION : The results presented here demonstrate that US2 causes degradation of MHC class II alpha in several different cells: gastric epithelial cells, cervical carcinoma cells, and two different clones of astroglioma cells. The extent of degradation of class II alpha was similar or identical with each cell type at any given dose of US2. In cells expressing both class I and II proteins, there was a 1.5- to 2-fold preference for class I HC over class II alpha. The half-life of class II alpha was reduced from ~45 to 60 min down to 2 to 5 min by expressing intermediate levels of US2 (data not shown). Other homologous HCMV glycoproteins, US3 and US7, had no effect on the stability of MHC proteins. Therefore, US2 causes degradation of class II proteins in several relevant cells and does not show a high degree of preference for MHC class I versus class II. The second important set of observations involved US2-mediated degradation of class II DRalpha and DMalpha in cells expressing these alpha chains alone, i.e., without beta and Ii chains. There was no detectable effect on the DRbeta chain in cells expressing DRbeta alone, providing an important negative control for nonspecific effects of US2. These results have important implications for our understanding of how US2 functions. Normally, DRalpha and DRbeta chains associate rapidly in the ER and fold extensively to form alphabeta complexes (reviewed in reference ). Obviously, US2 binding does not depend upon protein surfaces that evolve after alphabeta complex formation. Instead, US2 can bind directly to free alpha chains. This highlights the rapid nature of US2-DRalpha interactions in the ER. These results also ruled out the suggestion that U373 cells transfected with CIITA express disproportionate amounts of class II proteins, leading to induction of unfolded protein response and ER degradation , because the transiently transfected HeLa cells do not express CIITA. Cells that contained substantially different amounts of class II alpha exhibited similar levels of degradation of alpha chain at any given concentration of US2. These observations support the hypothesis that US2 triggers processes leading to ER-associated degradation by recruiting cellular proteins onto MHC proteins. The cellular factors limiting in US2-mediated degradation are unlikely to include components of the Sec61 protein translocon because US2 expression at high levels does not apparently affect Sec61-mediated forward translocation of proteins into the ER. The cytoplasmic tail of US2 is required for degradation of MHC proteins , and our group recently showed that swapping this domain in place of the US3 cytoplasmic tail produced a chimera that could cause degradation of MHC proteins . Unlike wild-type US3, US2/US3 chimeras capable of degrading MHC proteins bound p97 adenosine triphosphatase (ATPase) , a protein involved in the extraction of ER proteins into the cytoplasm . Therefore, p97 ATPase may be a candidate cellular protein that is limiting in the US2-provoked retrotranslocation and degradation pathway, but others are also likely to be involved. One major question that prompted these studies was whether or not US2 affects class II generally in any cell, in a subset of cells, or not at all. Based on extrapolation of structural analyses of the binding of US2 to class I HC, Gewurz et al. predicted that US2 should bind to the class II beta chain, rather than the alpha chain, and this group reported that US2 failed to bind to or cause degradation of class II complexes in vitro . However, in these in vitro studies, US2 also failed to destabilize certain class I proteins that are known to be degraded in US2-expressing mammalian cells . Thus, these in vitro and structural studies failed to reproduce several effects of US2 previously observed in mammalian cells. In the studies of Rehm et al. , relatively large input doses (500 PFU/cell) of an Ad vector expressing US2 were used in attempts to deliver US2 into DCs. There was little or no effect on the stability of MHC class I HC, compared with control cells, although band intensities were not quantified (see Fig. in reference ). Given the well-established effects of US2 on class I HC in other cells, it appears that inadequate amounts of US2 were delivered, and any conclusions about whether or not US2 can inhibit class II in DCs or generally in other cell types are not warranted. Alternatively, DCs may lack factors required by US2 to cause class I and II degradation. We did not pursue efforts to test US2's effects in DCs here for several reasons. First, DCs and monocyte-macrophages are inefficiently infected by HCMV, especially when these cells are cultured from blood and infected by laboratory strains of HCMV passaged on fibroblasts . Moreover, Ad vectors often do not infect DCs well . For example, the human immunodeficiency virus nef protein can reduce cell surface expression of MHC class I molecules in a variety of cells (reviewed in reference ) but did not affect class I when expressed by Ad vectors in DCs . Our preliminary efforts to transduce monocyte-macrophages and DCs cultured with Ad vectors lead to the conclusion that HCMV proteins could not be delivered effectively (R. A. Tomazin and N. R. Hegde, unpublished results). Therefore, we do not know whether US2 functions in DCs or macrophages to inhibit class II antigen presentation. The question of whether US2 functions in professional APCs is likely irrelevant for at least two reasons. First, the high levels of MHC proteins expressed in APCs, coupled with low levels of viral gene expression, make it difficult to imagine that US2, US3, and other HCMV immune evasion proteins could be effective in these cells in vivo. Second, US2 and US3 are intracellular membrane proteins that function only in HCMV-infected cells. In vivo, only a small fraction of APCs are infected (reviewed in reference ), and other, uninfected APCs are free to take up HCMV antigens and effectively prime or initiate immune responses. Consistent with this, there are robust and sustained anti-HCMV CD4+ T-cell responses in humans (reference and references therein). Therefore, it is highly unlikely that HCMV benefits from inhibiting the class II pathway in APCs. Rather, we believe that US2 and US3 function to prevent class II-mediated presentation of endogenous viral antigens in cells such as endothelial, glial, or epithelial cells or other class II-expressing cells that are important hosts for HCMV replication in vivo. Herpesviruses extensively target their structural proteins to endosomal compartments (reviewed in reference ), where peptides destined to be loaded onto class II are generated (reviewed in references and ). We have evidence for highly efficient loading of peptides derived from endogenous HCMV antigens, e.g., gB, onto class II proteins (C. Dunn, D. M. Lewinsohn, and D. C. Johnson, unpublished results). Such presentation will cause endothelial or epithelial cells to be recognized by anti-HCMV CD4+ T cells that can lyse the cells or produce antiviral cytokines. Therefore, we believe that the real function of US2 and US3 is to prevent presentation of endogenous HCMV antigens via the class II pathway and recognition of virus-producing cells, rather than presentation of exogenous antigens by professional APCs (reviewed in references and ). The observations that US2 targets several distinct MHC proteins is not novel. US3 affects classical and nonclassical class I proteins, as well as class II proteins . HCMV UL16 affects MIC-B, UL16-binding protein-1 (ULBP-1), and ULBP-2 . Human immunodeficiency virus nef targets CD4 and class I molecules for demise in lysosomes and alters sorting and peptide loading of class II proteins . The Kaposi's sarcoma herpesvirus K3 and K5 proteins target MHC class I proteins HLA-A, -B, -C, and -E for internalization from the cell surface followed by degradation , and K5 also downregulates B7-2 and ICAM-1 . Therefore, many viral immunomodulatory proteins have apparently evolved to affect multiple molecules that exhibit only limited homology but with shared structural features. FIG. 1. : Expression of class II proteins by various cell lines. Expression of class II proteins by various cell lines. His16, HeLa-CIITA, and 2E12 cells were labeled in suspension for 2, 4, or 8 min with [35S]methionine-cysteine, and the label was chased for 0 or 20 min. NP-40-DOC cell extracts derived from 6 x 106 (His16, 2E12) or 15 x 106 (HeLa-CIITA) cells were mixed with anti-DRalpha MAb DA6.147, and MHC class II complexes immunoprecipitated. FIG. 2. : US2 causes loss of class I and II proteins. US2 causes loss of class I and II proteins. (A) His16, HeLa-CIITA, or 2E12 cells were left uninfected (0 PFU/cell) or were infected with AdtetUS2 and Adtet-trans using 3 and 0.6, 10 and 2, 30 and 6, 100 and 20, or 300 and 60 PFU/cell, respectively, for 18 h. Infected cells were labeled with [35S]methionine-cysteine for 3 h, and cell extracts were made using NP-40-DOC lysis buffer. Class II complexes were immunoprecipitated from 6 x 106 (His16, 2E12) or 15 x 106 (HeLa-CIITA) cells using MAb DA6.147. (B) His16 (6 x 106) and HeLa-CIITA (15 x 106) cells were left uninfected (0 PFU/cell) or were infected with AdtetUS2, AdtetUS3, or AdtetUS7, in each case with Adtet-trans using 30 and 6 PFU/cell or 150 and 30 PFU/cell, respectively, for 18 h. Infected cells were labeled with [35S]methionine-cysteine for 1 min, and the label was chased for 20 min. Cell extracts were made with 1% SDS lysis buffer and denatured at 100C, and the SDS was diluted 10-fold with 1% Triton X-100 buffer. Lysates were divided equally into two, and class I HC or class II alpha chain (DRalpha) was immunoprecipitated with the MAb's HC10 or DA6.147, respectively. FIG. 3. : US2 causes degradation of class II alpha when expressed without the beta chain. US2 causes degradation of class II alpha when expressed without the beta chain. HeLa cells were transiently transfected with plasmids containing either the class II alpha (HeLa/DRalpha) or beta genes (HeLa/DRbeta), or both plasmids (HeLa-DRalpha/DRbeta), or one containing the DMalpha gene (HeLa-DMalpha). After 48 h, these transfected HeLa cells (4 x 106) and His16 cells (1 x 106) were left uninfected (0 PFU/cell) or infected with 3 and 0.6, 10 and 2, 30 and 6, 100 and 20, or 300 and 60 PFU/cell of AdtetUS2 and Adtet-trans, respectively, for 18 h. The cells were then labeled with [35S]methionine-cysteine for 3 h. Class II alpha was immunoprecipitated using MAb DA6.147, and class II beta was immunoprecipitated with MAb HB10A. alphabeta complexes were precipitated from HeLa cells transfected with both DRalpha and DRbeta plasmids with MAb DA6.147, and DMalpha was precipitated with MAb 5C1. Note that fewer His16 cells were used, and extracts of these cells were made using SDS buffer, as in Fig. , allowing selective precipitation of DRalpha but reducing binding of MAb DA6.147 compared with extracts of transfected HeLa cells made using NP-40-DOC buffer (see Fig. ). FIG. 4. : Similar degradation in cells with different amounts of class II proteins. Similar degradation in cells with different amounts of class II proteins. (A) His16 cells (1 x 106) or DRalpha-transfected HeLa cells (4 x 106) were labeled with [35S]methionine-cysteine for 3 h. His 16 cells were lysed in NP-40-DOC buffer (left lane denoted alpha/beta/Ii)) or in buffer containing 1% SDS, and the extract was denatured at 100C before the SDS was diluted with 1% Triton X-100 buffer (middle lane, denoted alpha). Transfected HeLa cells (HeLa-DRalpha) were lysed in NP-40-DOC buffer. All samples were immunoprecipitated with anti-DRalpha MAb DA6.147. (B) His16 or 2E12 cells (1 x 106) or transfected HeLa cells (4 x 106) were left uninfected (0 PFU/cell) or infected with 3 and 0.6, 10 and 2, 30 and 6, 100 and 20, or 300 and 60 PFU of AdtetUS2 and Adtet-trans/cell, respectively, for 18 h. Infected cells were labeled with [35S]methionine-cysteine for 3 h, extracts were made in every case with 1% SDS lysis buffer, proteins were denatured, and the SDS was diluted with 1% Triton X-100 buffer. Class II alpha or beta chains were immunoprecipitated from the cell extracts with anti-alpha MAb DA6.147 or anti-beta MAb HB10A, respectively. The intensities of protein bands were quantified by phosphorimager analyses. The data are the averages of three independent experiments, and the density of protein bands obtained from cells left uninfected (no AdtetUS2) was considered 100%. Data points without error bars are due to standard deviations that are too small for depiction. (C) Cells were labeled as in panel B, and US2 was immunoprecipitated with rabbit polyclonal serum to an N-terminal peptide of US2. The amount of US2 was quantified by phosphorimager analyses from the three experiments depicted in panel A. (D) The effects of US2 on class II DMalpha were analyzed as for panel B, except that HeLa cells were transfected with a plasmid containing the DMalpha gene and compared to His16 cells that also express DM. Immunoprecipitation was performed with anti-DMalpha MAb 5C1. FIG. 5. : US2 shows preference for class I HC over class II alpha chain. US2 shows preference for class I HC over class II alpha chain. His16 or Neo6 cells (6 x 106) were left uninfected (0 PFU/cell) or infected with 5 and 1, 10 and 2, 20 and 4, 40 and 8, or 80 and 16 PFU/cell of AdtetUS2 and Adtet-trans, respectively, for 18 h. Cells in suspension were labeled for 1 min, the label was chased for 20 min, cell extracts were prepared using SDS lysis buffer, proteins were denatured, and the SDS was diluted as described in the legend for Fig. . Extracts were then divided equally into two, and class II alpha (DRalpha) or class I HC was immunoprecipitated with MAbs DA6.147 or HC10, respectively. The amount of alpha or beta that remained after expression of each dose of US2 expression was quantified by phosphorimager analyses and compared to that in uninfected cells (100%). The data represent the averages of three independent experiments. Data points without error bars are due to standard deviations that are too small for depiction. FIG. 6. : Different Ad vectors expressing US2 cause degradation of MHC proteins. Different Ad vectors expressing US2 cause degradation of MHC proteins. His16 cells (6 x 106) were infected with 50 PFU/cell of Ad viruses AdUS2 and AdUS11 (described by Rehm et al. ) or Adbeta-gal (described by Tomanin et al. ), or AdtetUS2, AdtetUS3, or AdtetUS11 (described elsewhere by our group [, ]). The cells were labeled in suspension with [35S]methionine-cysteine for 1 min, and the label was chased for 20 min. Cell extracts were made using SDS lysis buffer, proteins were denatured, and the SDS was diluted as described in the legend for Fig. . Extracts were divided into three, and class II alpha chain (DRalpha), class I HC (panel A), or transferrin receptor (TfR; panel B) were immunoprecipitated with MAbs DA6.147, HC10, or B3/25, respectively. Backmatter: PMID- 12915584 TI - High Circulating Frequencies of Tumor Necrosis Factor Alpha- and Interleukin-2-Secreting Human T-Lymphotropic Virus Type 1 (HTLV-1)-Specific CD4+ T Cells in Patients with HTLV-1-Associated Neurological Disease AB - Significantly higher frequencies of tumor necrosis factor alpha- and interleukin-2-secreting human T-lymphotropic virus type 1 (HTLV-1)-specific CD4+ T cells were present in the peripheral blood mononuclear cells of HTLV-1-associated myelopathy/tropical spastic paraparesis (HAM/TSP) patients than in those of asymptomatic carriers with similar provirus loads. The data suggest that HTLV-1-specific CD4+ T cells play a role in the pathogenesis of HAM/TSP. Keywords: Introduction : HTLV-1 (human T-lymphotropic virus type 1) is the causative agent of adult T-cell leukemia/lymphoma and HTLV-1-associated myelopathy/tropical spastic paraparesis (HAM/TSP) and is distantly related to human immunodeficiency virus type 1 (HIV-1). However, whereas infection with HIV-1 results in disease in almost all of those infected, infection with HTLV-1 usually results in persistent asymptomatic carriage; less than 7% will develop either adult T-cell leukemia/lymphoma or HAM/TSP. The cause of these differing outcomes is unknown. Despite the fact that HTLV-1 causes disabling and fatal diseases, there is no vaccine and no satisfactory treatment for either the malignancy or the inflammatory diseases. HAM/TSP is characterized by a chronic, progressive, inflammatory, demyelinating myelopathy that resembles progressive spinal forms of multiple sclerosis. The pathogenesis of this disease is unclear but is thought to involve the host's immune response to the virus . The lesions in the central nervous system (CNS) contain infiltrates of inflammatory cells. Until now, most work on the cell-mediated immune response to HTLV-1 has focused on the cytotoxic T-lymphocyte (CTL) response. CTL responses are critical to clearing virus infections by the specific lysis of infected cells and secretion of antiviral cytokines such as gamma interferon (IFN-gamma) and tumor necrosis factor alpha (TNF-alpha) . In addition to CTLs, acquired immunity is also critically dependent on the generation of virus-specific CD4+ T-helper responses. By secreting cytokines upon recognition of cognate antigen, the CD4+ T-helper cell is involved in the coordination of the immune response to the invading pathogen. Activated T-helper cells can produce interleukin-2 (IL-2) and IFN-gamma, which augment CTL effector functions (Th1), and IL-4, IL-5, IL-6, and IL-10, which promote B-cell and antibody responses (Th2). T-helper cells can also activate professional antigen-presenting cells by secretion of cytokines or upregulation of CD40L on the cell surface . Several observations on HTLV-1 demonstrate the potential importance of CD4+ T cells in the cellular immune response to the virus. (i) CD4+ T cells are the main subset of cells infected with HTLV-1 in vivo . (ii) Infected CD4+ T cells spontaneously secrete proinflammatory, neurotoxic cytokines such as TNF-alpha and IFN-gamma , and high levels of these cytokines have been demonstrated in the serum, cerebrospinal fluid, and spinal cord lesions of HAM/TSP patients . (iii) HTLV-1 infection in CD4+ T cells can impair T-helper function . (iv) CD4+ CD45RO+ T cells are the predominant infiltrating subset (65%) of total CD3+ T cells in newer CNS lesions of HAM/TSP patients . Others have also detected more CD4+ T cells in the CNS lesions of HAM/TSP patients with a short history of disease than in patients with a long history of disease (when CD8+ T cells predominate) . We recently reported that IFN-gamma-secreting, HTLV-1-specific CD4+ T cells were significantly more frequent in HAM/TSP patients than in asymptomatic carriers (ACs) . This observation suggested that these HTLV-1-specific CD4+ T cells could be involved in the initiation and pathogenesis of HAM/TSP. According to the bystander damage hypothesis , the inflammatory and neurotoxic cytokines TNF-alpha and IFN-gamma could cause the nonspecific cell death and inflammation seen in HAM/TSP patients. In this study, we tested the hypotheses that (i) HAM/TSP patients and ACs with similar proviral loads would differ in the frequencies of TNF-alpha- and IL-2-secreting HTLV-1-specific CD4+ T cells and (ii) there is a correlation between the proviral load and the frequency of Th1-type CD4+ T cells in either HAM/TSP patients or ACs. We therefore used a flow cytometric assay with intracellular cytokine staining to detect IFN-gamma-, TNF-alpha-, and IL-2-positive, HTLV-1-specific CD4+ T cells. The cells were stimulated with overlapping peptides (20 amino acids, overlapping by 14 or 15) corresponding to all of the major proteins of HTLV-1. Peptides were grouped into one pool and added to cell culture medium to achieve a final concentration of 1 mug of each peptide per ml prior to incubation at 37C. This strategy should detect the total T-helper type 1 response to all potential epitopes of HTLV-1. We studied HTLV-1-infected ACs and patients with HAM/TSP attending the HTLV-1 clinic at St. Mary's Hospital, London, England. The diagnosis of HAM/TSP was made in accordance with World Health Organization criteria . Control samples were obtained from healthy volunteers in our laboratory. Peripheral blood mononuclear cells (PBMCs) were isolated by standard density gradient centrifugation and cryopreserved until use. Nonspecific cytokine production by PBMCs was induced by a combination of 0.1 ng of phorbol myristate acetate (PMA) per ml (Sigma) and 0.5 mug of A23187 (Sigma) per ml, or Staphylococcus enterotoxin B (SEB) at 5 mug/ml was added to the culture medium. Twenty nanomolar concanamycin A (a potent inhibitor of perforin activity) was added to the culture medium to prevent CTL-mediated lysis via the perforin-dependent cytotoxic pathway . Ten micrograms of brefeldin A (Sigma) per ml was added for the last 5 h of culture to inhibit protein secretion. HTLV-1 DNA was quantified by real-time PCR with a LightCycler (Roche, Mannheim, Germany) . HTLV-1 primers SK43 and SK44 were used. The beta-globin copy number of each sample was similarly quantified with primers PC03 and PC04 . Standard curves were generated for both PCRs with genomic DNA from C10-PBL cells, which contain one Tax copy per cell. We have previously published data on the flow cytometric detection of IFN-gamma-positive, CD4+ T cells in HTLV-1-infected patients . We established that our 6-h flow cytometric assay could also be used to detect the cytokines TNF-alpha, TNF-beta (lymphotoxin), and IL-2. As positive controls, we used the nonspecific T-cell stimulators PMA-A23187 and SEB. Figure illustrates the cytokine characterization of CD4+ T cells after 6 and 18 h of in vitro culture with brefeldin A inhibition of protein secretion. The results showed that IFN-gamma, TNF-alpha, and IL-2 protein expression could be detected after 6 h. TNF-beta was only detectable after 18 h of culture and only from a small percentage of CD4+ T cells. Isotype controls showed that the staining was specific (data not shown). Similar data were obtained from three uninfected controls, two HAM/TSP patients, and two ACs. We applied the 6-h assay to our cohort of HAM/TSP patients and ACs with similar proviral loads . We initially also stained for TNF-beta-positive, HTLV-1-specific CD4+ T cells in two HAM/TSP patients (TAQ and TAL) and two ACs (HT and HBI) but did not detect any positive cells. We therefore focused only on TNF-alpha and IL-2 staining for the other seven subjects in each group. Figure shows the flow cytometric data from a representative HAM/TSP patient (TBO). We characterized the cytokine expression of HTLV-1-specific CD4+ T cells in nine HAM/TSP patients and nine ACs with similar proviral loads. The data show that there were large populations of hitherto undetected TNF-alpha- and IL-2-positive CD4+ T cells that are HTLV-1 specific in HAM/TSP patients but not in ACs. The differences were statistically significant for all three cytokines (IFN-gamma, P = 0.0003; TNF-alpha, P = 0.00056; IL-2, P = 0.0003 [Mann-Whitney, two-tailed P values]). There were also small numbers of cells that coexpressed two cytokines, but consistent with previous reports , these were much less frequent than single-cytokine expression in the one cell. Triple-cytokine secretors were not tested for, but these would be expected to be even less common. We tested for correlations between individual cytokines and proviral loads in both HAM/TSP subjects and ACs by nonparametric statistical analysis (Spearman rank correlation). The data are shown in Fig. . There was a significant correlation between the frequency of total IL-2-positive HTLV-1-specific CD4+ T cells in ACs and the proviral load (two-tailed P = 0.0138, r = 0.8000). There was a tendency to a positive correlation between proviral load and both IFN-gamma and TNF-alpha in the ACs, but this did not reach statistical significance. The data on HAM/TSP patients showed no significant correlation between the proviral load and the specific-cytokine frequency for the three cytokines tested . The analysis of T-cell responses with a single cytokine such as IFN-gamma may not be representative of the entire Th1-type response. This may account for the conflicting reports of correlation and noncorrelation between antigen loads and cytokine responses, particularly for HIV . We therefore estimated the total frequency of Th1-type HTLV-1-specific CD4+ T cells in these subjects and tested for a correlation with proviral load; the data are shown in Fig. . We show that the estimated total frequency did not correlate with the proviral load in HAM/TSP patients but there was a significant positive correlation in ACs (HAM/TSP patients, two-tailed P = 0.3853, r = 0.3264; ACs, two-tailed P = 0.0214, r = 0.7615). The proviral load in HTLV-1 infection has been shown to correlate with the risk of inflammatory diseases such as HAM/TSP . There is typically a strong CTL response to the virus in both HAM/TSP patients and ACs . The CTL response has been shown to exert significant selection pressure on the Tax protein in ACs but not in HAM/TSP patients. Recent work has suggested that CTLs exert a beneficial effect by lightening the proviral load and protecting against HAM/TSP . However, there is also evidence that CTLs contribute to the inflammation seen in HAM/TSP. HTLV-1-specific CTLs are found at higher frequency in cerebrospinal fluid than in peripheral blood, implying either that they preferentially migrate into the CNS or that they selectively expand in the CNS (or both) . Also, some of these clones secrete inflammatory cytokines, chemokines, and matrix metalloproteinases . It is possible to reconcile these apparently contradictory data with the concept of a threshold proviral load . In this model, in an individual with a proviral load below the threshold, sufficient T-cell receptors on HTLV-1-specific CTLs are engaged to ensure efficient cell killing (protective) but insufficient to elicit production of inflammatory cytokines such as TNF-alpha and IFN-gamma. As the proviral load increases, specific CTLs commence cytokine production, in addition to their cytotoxic function (which itself may be pathogenic as well as protective). Indeed, it has been shown that above a threshold proviral load of similar1% of PBMCs, the risk of HAM/TSP rises exponentially . However, in acquired immunity, the CD4+ T-helper cell also plays a critical role in the immune response. In HTLV-1, little is known about the role of the CD4+ T-helper cell. In this study, we used overlapping peptide libraries of all major HTLV-1 proteins to quantify the Th1 CD4+ T-cell response to the virus in both HAM/TSP patients and infected ACs. Production of IFN-gamma is frequently used as a marker of Th1 CD4+ T cells in ELISpot or intracellular cytokine staining assays. However, certain Th1 cells produce IL-2 or TNF-alpha but not IFN-gamma. Previously, we showed that HAM/TSP patients have a predominance of Th1-type HTLV-1-specific CD4+ T cells . In this study, we wished to determine whether there are similarly large populations of other Th1-type cytokine (TNF-alpha, TNF-beta, and IL-2)-secreting, HTLV-1-specific CD4+ T cells in HAM/TSP patients compared to ACs with similar proviral loads. These cytokines may play an important role both in the immune response to HTLV-1 and in the pathogenesis of inflammatory neurological disease. A heavy proviral load has been shown to be associated with increased risk and severity of HAM/TSP . Therefore, to elucidate a potential mechanism of pathogenesis of HAM/TSP, we selected a group of patients and HTLV-1-infected ACs with similar loads so that any observed differences would not reflect proviral load differences . We show that total TNF-alpha- and IL-2-positive, HTLV-1-specific CD4+ T cells were significantly more frequent in HAM/TSP patients than in ACs. TNF-alpha is a powerful proinflammatory cytokine and has been implicated in many inflammatory disorders . It also plays a central role in coordination of the inflammatory response . The discovery of numerous TNF-alpha+, HTLV-1-specific CD4+ T cells in patients with inflammatory CNS disease, but not in ACs, supports the hypothesis that bystander damage is inflicted by invading CD4+ T cells in the early, active lesions of HAM/TSP. IL-2 has an important role in lymphocyte proliferation and in arming effector T cells, including both T-helper cells and CTLs. IL-2 has also recently been shown to rescue secondary expansion and function in CTLs depleted of CD4+ T-cell help . The high frequency of IFN-gamma+, TNF-alpha+, and IL-2+, HTLV-1-specific CD4+ T cells raises the possibility that such cells contribute to the pathogenesis of inflammatory CNS disease by forming self-perpetuating inflammatory lesions. The observation that these cells are found in abundance in patients with HAM/TSP but not in ACs with similarly high proviral loads suggests that the high Th1-cell frequency is specifically associated with the disease HAM/TSP. Analyses of total IFN-gamma-positive, virus-specific T cells have been reported in several articles . Our results show that other Th1 cytokines also need to be examined if a more complete picture of CD4+ T-helper responses in other chronic infections with viruses such as HIV, cytomegalovirus, Epstein-Barr virus, and hepatitis B or C virus is to be obtained. Similarly, detection of IFN-gamma alone may give an incomplete measure of virus-specific CD8+ T-cell frequency. We observed positive correlations between the estimated total Th1 (and total IL-2-positive) CD4+ T-cell frequencies and the proviral load in ACs but not in HAM/TSP patients. There are several interesting implications. The frequency of the Th1 response rises as the proviral load increases in ACs but not in HAM/TSP patients. Yet the proviral loads are not dissimilar between the two groups. This suggests that there are fundamental qualitative (functional), as well as quantitative, differences in the CD4+ T-cell response to HTLV-1. What might such qualitative differences be? It is not possible to reach a robust conclusion about the efficiency or effectiveness of the antiviral T-cell response from the frequency of such cells in the circulation . However, the greater frequency of HTLV-1-specific CD4+ T cells in patients with HAM/TSP and a given proviral load implies either that their CD4+ T cells proliferate faster (or die more slowly) than those of ACs or that more HTLV-1 antigen is expressed in HAM/TSP patients than in ACs with a similar proviral load (or both). In conclusion, we show here that there are significantly higher frequencies of TNF-alpha- and IL-2-positive, HTLV-1-specific CD4+ T cells in HAM/TSP patients than in HTLV-1-infected ACs with similar proviral loads. These results confirm and extend our previous work on IFN-gamma-positive, HTLV-1-specific CD4+ T cells with another assay. The relative importance of these three cytokines in the pathogenesis of HAM/TSP is unknown. There were positive correlations between the estimated total Th1 (and total IL-2 response) and proviral load in ACs but not in patients with HAM/TSP. The data suggest that there are quantitative and qualitative differences in the T-helper cell populations between HAM/TSP patients and ACs. We have preliminary evidence that the HTLV-1 antigens recognized by CD4+ T cells in HAM/TSP patients and ACs do not differ (unpublished data); however, it remains possible that the epitopes within these antigens differ systematically. The data presented in this report have important implications for the understanding of the pathogenesis of HAM/TSP disease, the cellular immune response to HTLV-1, and possible therapeutic approaches. FIG. 1. : Th1-type cytokine functional characterization of CD4+ T cells. Th1-type cytokine functional characterization of CD4+ T cells. Staining for intracellular IFN-gamma, TNF-alpha, TNF-beta, and IL-2 was conducted following stimulation with SEB or with PMA and A23187 (a calcium ionophore) for 6 or 18 h of in vitro culture of PBMCs. Brefeldin A was added for the last 5 h to inhibit protein secretion. TNF-beta was only detected after 18 h of culture and only at a low frequency. Cells were gated on the CD4hi population for SEB stimulation and controls or on all lymphocytes for PMA stimulation. FIG. 2. : High frequencies of TNF-alpha- and IL-2-positive CD4+ T cells in HAM/TSP patients after 6 h of in vitro culture with HTLV-1 peptides. High frequencies of TNF-alpha- and IL-2-positive CD4+ T cells in HAM/TSP patients after 6 h of in vitro culture with HTLV-1 peptides. Data on TBO, a patient with HAM/TSP, are shown. Cells were gated on the CD4hi population. The total number of events collected was 200,000 for the peptide-stimulated culture and 100,000 for the control (no peptides). FIG. 3. : Correlation between frequency of single-cytokine-positive, HTLV-1-specific CD4+ T cells and proviral load (PL). Correlation between frequency of single-cytokine-positive, HTLV-1-specific CD4+ T cells and proviral load (PL). There is a significant positive correlation between IL-2-positive HTLV-1-specific CD4+ T-cell frequency and proviral load in ACs. There is a trend toward a positive correlation for all three cytokines in ACs, but this does not reach significance for IFN-gamma and TNF-alpha. Data were analyzed by Spearman rank correlation. The solid line represents a least-squares regression line. The asterisk indicates a significant correlation. FIG. 4. : Correlation between total Th1-type, HTLV-1-specific CD4+ T-cell frequency and proviral load. Correlation between total Th1-type, HTLV-1-specific CD4+ T-cell frequency and proviral load. The total Th1 frequency correlates positively with the proviral load in ACs but not in HAM/TSP patients. Analysis was done by Spearman rank correlation. The solid line represents a least-squares regression line. The asterisk indicates a significant correlation. TABLE 1 : HTLV-specific CD4+ T-cell frequencies in HTLV-1-infected subjects Backmatter: PMID- 12915529 TI - Targeting of the Turnip Yellow Mosaic Virus 66K Replication Protein to the Chloroplast Envelope Is Mediated by the 140K Protein AB - Turnip yellow mosaic virus (TYMV), a positive-strand RNA virus in the alphavirus-like superfamily, encodes two replication proteins, 140K and 66K, both being required for its RNA genome replication. The 140K protein contains domains indicative of methyltransferase, proteinase, and NTPase/helicase, and the 66K protein encompasses the RNA-dependent RNA polymerase domain. During viral infection, the 66K protein localizes to virus-induced chloroplastic membrane vesicles, which are closely associated with TYMV RNA replication. To investigate the determinants of its subcellular localization, the 66K protein was expressed in plant protoplasts from separate plasmids. Green fluorescent protein (GFP) fusion and immunofluorescence experiments demonstrated that the 66K protein displayed a cytoplasmic distribution when expressed individually but that it was relocated to the chloroplast periphery under conditions in which viral replication occurred. The 66K protein produced from an expression vector was functional in viral replication since it could transcomplement a defective replication template. Targeting of the 66K protein to the chloroplast envelope in the course of the viral infection appeared to be solely dependent on the expression of the 140K protein. Analysis of the subcellular localization of the 140K protein fused to GFP demonstrated that it is targeted to the chloroplast envelope in the absence of other viral factors and that it induces the clumping of the chloroplasts, one of the typical cytological effects of TYMV infection. These results suggests that the 140K protein is a key organizer of the assembly of the TYMV replication complexes and a major determinant for their chloroplastic localization and retention. Keywords: Introduction : A universal feature of eukaryotic positive-strand RNA viruses is that replication of their genomes is closely associated with intracellular membranes (reviewed in reference ). Most purified viral RNA replication complexes copurify with membrane extracts from infected cells (reviewed in reference ) and, although in some cases RNA synthesis activity can be solubilized , in vivo and in vitro studies suggest that the presence of membranes and/or phospholipids is essential for at least some steps of RNA replication . It was proposed that these membranes can play both a structural and a functional role in the replication complex. Electron microscopy observations of infected cells revealed that many positive-stranded RNA viruses induce proliferation and/or reorganization of the intracellular membranes of their host to create a membrane compartment in which RNA replication takes place. Depending on the virus, a variety of membrane systems can be concerned, including the early and late endomembrane systems , the nuclear envelope , the vacuole , the endosomes and lysosomes , the peroxisomes , chloroplasts , and mitochondria . The fact that distinct types of membranes are involved in the replication of different viruses suggests the establishment of specific interactions between such host membranes and virus-encoded proteins. A number of viral proteins that target replication complexes to intracellular membranes have been identified . Membrane interaction of host-encoded factors that are part of the viral replication complex has also been reported . Despite this universal association of positive-strand RNA virus replication complexes with intracellular membranes, little is known about the mechanisms by which the viral replication complexes are targeted to and assembled on specific membrane sites. Characterizing these structures and the mechanisms of their localization may help to identify general principles in positive-strand RNA virus replication. We address here this question by studying the assembly of the replication complex of Turnip yellow mosaic virus (TYMV), the type member of the tymovirus group. TYMV shares viral replication features with positive-strand RNA viruses from other members of the alphavirus-like supergroup of viruses and has proven useful in investigating fundamental aspects of viral multiplication . TYMV is a small spherical plant virus that infects members of the Cruciferae (reviewed in reference ). Upon infection, TYMV triggers the development of typical cytological abnormalities that appear to be confined to the chloroplasts . These include the swelling and clumping of the chloroplasts and the appearance of peripheral structures consisting of membrane vesicles, 50 to 100 nm in diameter, that are likely to result from the invagination of the chloroplast envelope into the organelle . These small vesicles are closely associated with TYMV RNA replication, as revealed by previous in vivo RNA labeling observations and immunocytochemical experiments with antibodies raised against the viral replication complex . Consistent with these results, membrane-associated extracts that selectively synthesize TYMV negative-strand RNAs have been isolated from TYMV-infected plant cells . The TYMV genome is composed of a monopartite, positive-sense RNA genome of 6,718 nucleotides (nt) . It is capped and directs the expression of two extensively overlapping nonstructural proteins: 69-kDa (69K) protein and 206K protein . A third open reading frame (ORF) encodes the 20-kDa coat protein (CP), which is expressed from a subgenomic RNA. The 206K protein is the only viral protein required for TYMV RNA replication . It shows considerable amino acid sequence similarities with nonstructural putative replication proteins of several positive-strand RNA viruses and domains indicative of methyltransferase, NTPase/helicase, and RNA-dependent RNA polymerase (RdRp) activities have been highlighted in its sequence . This large nonstructural protein also contains a papain-like cysteine proteinase domain located between the methyltransferase and the NTPase/helicase domains that is responsible for the cotranslational proteolytic cleavage of the 206K protein in vitro . Mapping of the cleavage site (between A1259 and T1260) revealed that the resulting N-terminal protein product of 140 kDa (140K) contains the methyltransferase, proteinase, and NTPase/helicase motifs, whereas the C-terminal 66K protein encompasses the RdRp domain . This cleavage was also demonstrated to be functional in vivo , and both the 140K and 66K viral proteins appear to be essential for the replication of TYMV RNA genome . We showed previously that, during viral infection, the TYMV 66K protein localizes to the virus-induced membrane vesicles present at the chloroplast envelope . It is not known, however, whether the 66K protein carries the determinants for its own localization. In the present study, we used the ability to express the TYMV-encoded replication proteins in plant protoplasts from separate plasmids to investigate the determinants of their localization. Green fluorescent protein (GFP) fusion and immunofluorescence experiments provided the opportunity to trace the 66K protein within transfected or infected protoplasts. We demonstrate that the 66K protein displays a cytoplasmic distribution when expressed individually and that its targeting to the chloroplast envelope in the course of the viral infection is solely dependent on the expression of the 140K protein. Our studies also demonstrate that the 140K protein is targeted to the chloroplast envelope in the absence of other viral factors and that it induces the clumping of the chloroplasts, one of the typical cytological effects of TYMV infection. MATERIALS AND METHODS : Plasmid constructions. | All DNA manipulations were performed by using standard techniques . The sequences of PCR-generated DNA fragments were confirmed by DNA sequencing, and the overall structures of all plasmids were confirmed by restriction analysis. Laboratory designations for plasmids are given in parentheses. Intermediate plasmids. | A DNA fragment corresponding to the 140K ORF was amplified by PCR with Pfu DNA polymerase (Promega) and the primers 5'-CCGGAATTCGAACTAGTCATGGCCTTCCAATTAGCATTGGACGCCC-3'and 5'-GGAAGCTTGAATTCAGGCCCCGTTGAGTTTGGGGCCGCG-3' with E17 as a template. After digestion with SpeI and EcoRI, the fragment was cloned in the similarly restricted pGadGH vector (Clontech) to create pGad-140K. A DNA fragment corresponding to the 66K ORF was amplified by PCR with E17 as a template, Pfu DNA polymerase, and primers 5'-CCGGAATTCGAACTAGTCACCCCCAGCGCATCCCCCACCCACCGTTCG-3' and 5'-GGCCGCTCGAGGATCCTATTGGACGTAGTGAAGCAATTCAGACTC-3'. The resulting PCR product was digested with SpeI and BamHI and cloned into the SpeI-BglII-restricted pGadGH vector to create pGad-66K. Cloning of the AvrII-SacII DNA fragment of E17 into the SpeI-SacII-restricted pGad-66K created pGad-JN115. The NcoI-PstI fragment of p17AE corresponding to the N terminus of the 66K ORF was cloned into the similarly restricted vector pBS-Nco+, a derivative of pBlueScript II KS(+) in which a unique NcoI restriction site replaced the SmaI site, to create pBS-66-Nterm. The C terminus of the 66K ORF was then excised from pGad-66K by digestion with PstI and SmaI and recloned into the PstI-EcoRV-restricted pBS-66-Nterm to create the plasmid pBS-66K. A similar strategy was used to create pBS-His66K, except that plasmid p8AF was used to provide the N-terminal His6-tagged version of the 66K ORF. In order to obtain the clone pBS-66KHis in which the His6 tag is fused to the C-terminal of the 66K ORF, a DNA fragment was amplified by PCR with E17 as a template, Pfu DNA polymerase, and the primers 5'-ACTCGCCACGACTGGCCATCCG-3' and 5'-GCGAATTCAAGCTTAATGGTGATGGTGATGGTGTTGGACGTAGTGAAGCAATTCAGACTCAGCG-3'. The resulting PCR product was digested with PstI and HindIII and cloned into the similarly restricted pBS-66-Nterm to create pBS-66KHis. Full-length cDNA constructs and derivatives. | Plasmid E17, from which genomic length infectious transcripts corresponding to the entire genome of TYMV can be obtained, was described previously . Plasmid E17-stopDelta (dpA3) was constructed in two steps. Frist, E17 was digested with SalI and SacII, followed by insertion of the adaptor 5'-TCGACTCTAGAACCGC-3'. The resulting plasmid E17Delta contains an in-frame deletion within the 66K ORF (i.e., amino acids 74 to 165 missing). Then, the BamHI-SalI fragment of pGad-140K was cloned in the similarly restricted E17Delta to create E17-stopDelta, in which a stop codon truncates the 206K ORF at the cleavage site (amino acid 1259) and in which nt 3872 to 4366 of the TYMV genome are deleted (amino acids 1 to 165 of the 66K ORF). To obtain the plasmid E17-stop69K (gdA2), a DNA fragment was amplified by PCR with Pfu DNA polymerase and primers 5'-GGAAACAGCTATGACCATG-3' and 5'-CCTTCGGAATGGACCATGGGTAGGTCTGTATCGACGATCGAATTGAATCAACTGTGGATTCGAGAATC-3' with E17 as a template. After digestion with HindIII and NcoI, the fragment was cloned in the similarly restricted E17 to create E17-stop69K, in which a stop codon truncates the 69K ORF at amino acid 30 without modification of the 206K ORF. Plasmid E17-stopCP (C4+mcs) was constructed in two steps. First, a DNA fragment was amplified by PCR with Pfu DNA polymerase and primers 5'-GACCCTTCTACCGGCCTCCATC-3' and 5'-CACGTGCCCGGGACGCGTCTGGTAGGACGGTGGCGACGGTGAC-3' with E17 as a template. After digestion with SstII and XmaI, the fragment was cloned in the similarly restricted E17. The resulting plasmid C4 contains a deletion within the CP ORF (i.e., nt 5708 to 6058 of the TYMV genome missing). Then, the adaptor oligonucleotide 5'-GTGAATAGGCCTCGAGGGCCTTCGAACTAGTACGTATCTAGAGCGGCCGCCC-3' was inserted in the SmaI restriction site to create a stop codon interrupting the CP ORF at amino acid 26. Expression vectors. | All plant expression vectors derive from pDH51, a pUC-based plant expression cassette that includes the cauliflower mosaic virus (CaMV) 35S promoter and terminator . To facilitate future cloning of the genes of interest and enhance their translation, the original polylinker was replaced by the Omega adaptor 5'-TATTTTTACAACAATTACCAACAACAACAAACAACAAACAACATTACAATTACTATTTACAATTACCATGGTACTAGTCCCGGGATCCT-3'. It contains the 5'-untranslated leader of tobacco mosaic virus (TMV) (the omega element; underlined), a sequence that functions as a translational enhancer and several restriction sites. The corresponding plasmid was designated pOmega and was further used for cloning the TYMV genes. To construct the expression vector encoding the 66K protein and its histidine-tagged derivatives, the 66K ORFs were excised from pBS-66K, pBS-His66K, and pBS-66KHis by digestion with NcoI and XhoI and recloned into the NcoI-SalI-restricted pOmega vector to create pOmega-66K, pOmega-His66K, and pOmega-66KHis, respectively. To obtain the expression plasmid encoding the 140K protein, the 140K ORF was excised from pGad-140K by digestion with SpeI and SalI and recloned into the similarly restricted pOmega vector to create pOmega-140K. To obtain the expression plasmid encoding the 206K protein, a SmaI site was first inserted in pOmega-140K in place of the SalI site with the use of the adaptor oligonucleotide 5'-TCGAGCCCGGGC-3'. The SacI-SmaI DNA fragment of pGad-JN115 was then excised and recloned into the similarly restricted pOmega-140K to create pOmega-206K (dpB16). To construct the expression vector pOmega-CP encoding the TYMV CP, a DNA fragment was amplified by PCR with Pfu DNA polymerase and primers 5'-GGACTAGTCCATGGAAATCGACAAAGAACTCGC-3' and 5'-GCTCTAGAGGATCCTTAGGTGGAAGTGTCCGT-3' with E17 as a template. The resulting PCR product was digested with SpeI and XbaI and cloned into the similarly restricted pOmega expression vector. A similar strategy was used to construct POmega-69K (gdG3) encoding the TYMV 69K protein with a His6 tag at the C terminus, except that the plasmid E17-206K-stop was used as a template for the PCR. This clone contains three point mutations aiming at preventing the expression of the overlapping 206K protein without introducing any modification of the 69K ORF. The primers used for PCR were 5'-TAATACGACTCACTATAGGGCTCGAGACTAGTCATGAGTAACGGCCTTCCAATAAGCATTGGACGC-3' and 5'-GTCGACTCTAGAGCATGCTCAGTGATGGTGATGGTGATGATCGGTGTCGGGGGCGCTGCCGTAGTC-3'. The resulting DNA fragment was digested with SpeI and SphI and cloned into the similarly restricted pOmega expression vector. To construct GFP gene fusions, the polylinker of pOmega was first replaced by the adaptor oligonucleotide 5'-CTAGTCGCGAGCTCTGCAGATCTCTAGATCCATGGGATCCGTACGCGGCCGCCCTAGGTACGTAGTCGACTCGAGCATG-3' to create pOmega-adaptGFP (dpC4). The DNA fragment corresponding to the enhanced green fluorescent protein (EGFP) gene (Clontech) was then excised from pCK-EGFP by digestion with NcoI and BsrGI and cloned into the NcoI-BsiWI-restricted pOmega-adaptGFP to create the expression vector pOmega-EGFP (dpD1) encoding free EGFP. Cloning of the SpeI-XhoI fragment of pGad-66K into the AvrII-XhoI-restricted pOmega-EGFP created pOmega-EGFP-66K (dpE7). Similarly, cloning of the SpeI-SalI fragment of pGad-140K into the AvrII-SalI restricted pOmega-EGFP created pOmega-EGFP-140K (ajB2). In order to obtain the plasmid pOmega-66K-EGFP (dpG40) in which the EGFP gene is fused at the C terminus of the 66K protein coding sequence, the SpeI-PstI fragment of pBS-66-Nterm was first cloned into the similarly restricted pOmega-EGFP to create pOmega-66-Nterm-EGFP (dpF30). Then, a DNA fragment was amplified by PCR with Pfu DNA polymerase and primers 5'-ACTCGCCACGACTGGCCATCCG-3' and 5'-GCTCTAGATTGGACGTAGTGAAGCAATTCAGAC-3' with E17 as a template. The resulting PCR product was digested with PstI and XbaI and cloned into the similarly restricted pOmega-66-Nterm-EGFP vector to create pOmega-66K-EGFP. In vitro transcription and translation. | Capped in vitro transcripts were obtained from AgeI-linearized DNA templates by using the Message Machine Transcription system (Ambion) according to the supplier's instructions. In vitro translation reactions were carried out in micrococcal nuclease-treated reticulocyte lysate as described previously . Preparation and infection of protoplasts. | Protoplasts of Arabidopsis thaliana ecotype Columbia were prepared from a cell suspension culture as described previously . A total of 106 protoplasts were transfected with 5 mug of (i) viral RNA, (ii) in vitro transcript, (iii) plasmid expression vector, or (iv) a mixture of both in cotransfection experiments. Transfected protoplasts were incubated at 24C in the dark for the indicated period of time as described previously . Under these conditions, routine transfection levels were ca. 10% of transfected protoplasts, and the efficiency of cotransfection was estimated to be at least 70% of the transfected protoplasts. Plant inoculations. | For plant inoculations, aliquots of 4 x 105 protoplasts were collected at 48 h posttransfection, lysed in 50 mul of inoculation buffer (20 mM sodium phosphate [pH 7.5]), and rubbed on two young leaves of 5- to 6-week-old Chinese cabbage (Brassica pekinensis cv. Granaat) plants, using celite as an abrasive. The plants were grown at 20 to 25C with a 16-h daylength. At 4 weeks postinoculation, both the inoculated and uninoculated young developing leaves were collected, and detection of the viral CP in the samples was performed as described previously . SDS-PAGE and immunoblot analysis. | Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblot analysis with polyclonal anti-66K or anti-CP antibodies were performed as previously described , except that the anti-CP antibody was diluted 10,000-fold. Immunofluorescent labeling, GFP observation, and epifluorescence microscopy. | TYMV-infected or plasmid-transfected protoplasts (similar106 cells) were harvested for immunofluorescence staining at 18 to 48 h posttransfection and were processed essentially as described previously , except that goat anti-rabbit immunoglobulin G conjugated to Alexafluor-488 (Molecular Probes) diluted 2,000-fold in phosphate-buffered saline-1% bovine serum albumin was used as a secondary antibody. To observe GFP fluorescence in living cells, similar105 cells were harvested, immobilized between microscope slides and coverslips, and examined immediately. Fluorescence microscopy and image acquisitions were performed as previously described by using a Leica DMR epifluorescence microscope. Standard filters for fluorescein isothiocyanate were used to monitor and record Alexafluor-488 and EGFP fluorescence. Chloroplasts were identified in fixed cells by DAPI (4',6'-diamidino-2-phenylindole) staining of the chloroplastic DNA as described previously or in living cells by the chlorophyll autofluorescence that was monitored and recorded by using the rhodamine channel. RNA extraction and Northern blot hybridization. | The protoplasts were collected by centrifugation at 80 x g for 5 min at room temperature, and nucleic acids were extracted as described previously . Total RNA (2.5 mug) was loaded onto 1% agarose-formaldehyde gels and, after they were blotted onto charged nylon membranes (Roche), the membrane was hybridized with a digoxigenin (DIG)-labeled RNA probe according to the supplier's instructions (Roche). After stringent washes (in 0.2x SSC [1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate]-SDS 0.1% at 68C), the blot was incubated with anti-DIG antibody conjugated to alkaline phosphatase and then visualized with CDP-Star as a substrate according to the supplier's instructions (Roche). For the production of the DIG-labeled probe, plasmid 1AH was constructed by subcloning the 3,044-bp XhoI-EcoRI fragment of E17 in the similarly restricted pBluescript II KS(+). The probe was prepared by in vitro transcription with T7 RNA polymerase of BglII-restricted 1AH, in the presence of DIG-UTP (Roche) according to the supplier's instructions. The 338-nt riboprobe hybridized to nt 5984 to 6318 of the viral genomic RNA and was suitable for detection of the viral genomic plus-strand RNA and the subgenomic RNA. RESULTS : Construction of plasmids expressing the 66K protein and its derivatives in plant cells. | In order to study the localization of the viral proteins independently of the viral infection, we first constructed the expression plasmid vector pOmega, a derivative of pDH51 that allows the expression of cloned genes in plant cells under the transcriptional control of the CaMV 35S promoter. The TMV omega element, which functions as a translational enhancer , was inserted downstream of the promoter sequence. Cloning of the TYMV 66K protein coding sequence into pOmega resulted in pOmega-66K, whereas pOmega-His66K and pOmega-66KHis encode derivatives of the 66K protein with in frame His6 tags at the N or C terminus, respectively. The use of GFP, which allows direct fluorescence microscopy of living cells, was also explored. EGFP, a fluorescence-enhanced version of GFP (Clontech), was fused in frame to the N or C terminus of 66K protein to create two fusions, EGFP-66K and 66K-EGFP, that were encoded by the plasmids pOmega-EGFP-66K and pOmega-66K-EGFP, respectively. The expression in Arabidopsis cells of the 66K protein and its derivatives was analyzed by transfection of each of the plasmid constructs in protoplasts, followed by immunoblotting with anti-66K specific antibodies . As shown in lane 2, in cells transfected with pOmega-66K the anti-66K antibodies recognized a major protein that comigrated with the 66K protein produced in the course of the viral infection (lane 1). In cells expressing the His-tagged derivatives (lanes 3 and 4) or the GFP-fused versions of the 66K protein (lanes 5 and 6), the anti-66K antibody-reactive band shifted to a higher position that was consistent with the expected molecular mass of the fusion proteins. The 66K protein and its derivatives have a cytoplasmic distribution when expressed individually. | In order to determine the intracellular sites of accumulation of the TYMV 66K protein in the absence of other virus-encoded proteins, Arabidopsis protoplasts were transfected with the expression vectors pOmega-66K, pOmega-His66K, or pOmega-66KHis. At 46 h posttransfection, the transfected cells were fixed, and indirect immunofluorescence microscopy was carried out by using anti-66K antibodies and secondary antibodies conjugated to Alexafluor-488 fluorochrome. Chloroplasts were identified in fixed cells by DAPI staining of the chloroplastic DNA. Figure to D' show representative epifluorescence micrographs of such cells. Transfected cells displayed a bright fluorescent staining of the 66K protein (green) throughout the cytoplasm. The protein was excluded from subcellular compartments such as the nucleus, the vacuole, or chloroplasts, as evidenced by confocal microscopy (data not shown). Mock-infected cells processed and imaged in parallel showed negligible fluorescence . The cytoplasmic distribution of the 66K protein was confirmed by protoplast transfection experiments with the expression vectors pOmega-EGFP-66K or pOmega-66K-EGFP. In these cases, the subcellular distribution of the fusion proteins can be directly observed by fluorescence microscopy in living cells due to the green fluorescence of the GFP moiety. When expressed from pOmega-EGFP-66K or pOmega-66K-EGFP, the GFP-fused 66K proteins appeared distributed diffusely in the cytoplasm (Fig. and G). In some cases, the fluorescent proteins were also observed as punctate structures of variable size and intensity distributed throughout the cytoplasm. These foci do not correspond to chloroplasts, as evidenced by the simultaneous observation of chlorophyll red autofluorescence in living cells, or staining of the chloroplastic DNA with DAPI in fixed cells. When EGFP was produced in a free form from pOmega-EGFP, diffuse fluorescence characteristic of cytosoluble GFP was observed throughout the cell . These studies demonstrate that the 66K protein, when expressed individually, is not targeted to the chloroplast but that it accumulates in the cytoplasm. These data therefore suggest that targeting of the 66K protein to the chloroplast envelope in the course of the viral infection, as can be seen in Fig. and H' and as evidenced previously , is not an intrinsic property of the 66K protein but rather a process that is dependent upon virus infection. The 66K protein and some of its derivatives can support RNA replication in trans. | To test whether the 66K protein, and its fused derivatives, expressed in trans from expression plasmids can support TYMV RNA replication and subgenomic mRNA synthesis, the expression vectors pOmega-66K, pOmega-His66K and pOmega-66KHis, pOmega-EGFP-66K, and pOmega-66K-EGFP expressing wild-type 66K, His-tagged 66K derivatives, and GFP-fused 66K proteins, respectively, were introduced into plant cells, together with a defective TYMV-derived replication template. The variant RNA to be used as a template for replication, named E17-stopDelta, contains two alterations aimed at preventing the synthesis of a functional 66K protein . First, a stop codon was introduced at amino acid 1259 of the 206K ORF, i.e., at the C terminus of the 140K protein . This stop codon was followed by a 494-nt deletion (nt 3872 to 4366 of the TYMV genome), the region that formerly encoded the first 165 amino acids of the 66K protein. As expected, upon in vitro translation in rabbit reticulocyte lysate of the corresponding transcript, the 66K protein was not produced, but wild-type amounts of the 69K and the 140K proteins were synthesized (data not shown). Upon transfection of Arabidopsis protoplasts with E17-stopDelta transcript, neither accumulation of viral progeny was detectable by Northern blot hybridization (Fig. , lane 2), nor accumulation of the CP was detectable by Western blot (Fig. , lane 2). These observations are consistent with the essential role played by the 66K protein within the replication process and its debilitation in the E17-stopDelta transcript. In contrast, in cells cotransfected by this defective replication template and the plasmid construct pOmega-66K expressing wild-type 66K protein, both genomic and subgenomic RNA (Fig. , lane 3) and a large amount of CP (Fig. , lane 3) could be detected. This demonstrates that the 66K protein produced in trans from an expression vector is able to complement the E17-stopDelta defective RNA. Similar data were observed when the expression vectors pOmega-His66K (Fig. and B, lanes 4) and pOmega-66KHis (Fig. and B, lanes 5) were used, demonstrating that the presence of the His6 tag at one or the other terminus of the protein did not interfere with its functionality. On the other hand, it appeared that the GFP fusion proteins did not support genomic RNA replication and CP synthesis, as shown by the absence of viral RNA progeny and CP in cells that were cotransfected with E17-stopDelta transcript and the plasmid constructs pOmega-EGFP-66K or pOmega-66K-EGFP (Fig. and B, lanes 9 and 11). Thus, the GFP moiety is apparently deleterious for the proper functioning of the 66K protein in the replication process. Western blotting experiments with the anti-66K antibodies were used as controls to verify that all of the constructs used were properly expressed . Given the alterations introduced in E17-stopDelta, it appears extremely unlikely that the replication observed in the cotransfection experiments is due to reversion of the defective replication template or to recombination with the expression vector. Nevertheless, experiments were carried out to verify that wild-type RNA was absent from the progeny of the complementation experiments. For that purpose, a fraction of the protoplasts used in the experiment was collected 46 h posttransfection, lysed and used to inoculate Chinese cabbage plants. Those were inspected 4 weeks postinoculation for disease symptoms, and the presence of CP in the inoculated and young expanding leaves was assayed by Western blotting. This potent assay failed to detect any wild-type RNA molecule in the progeny (data not shown), thereby demonstrating that the 66K protein produced from an expression vector can transcomplement a transcript in which the copy of the 66K gene has been disabled. The 66K protein and its derivatives relocalize to the chloroplast envelope in infected cells. | Because the above data suggested that viral infection is required for proper targeting of the 66K protein to the chloroplast envelope, it was of interest to investigate the subcellular distribution of the 66K protein and its derivatives within cells during the infection process. For this purpose, Arabidopsis protoplasts were cotransfected with the expression vector pOmega-66K, together with E17-stopDelta transcript. As shown above, under these conditions transcomplementation allows the viral replication to take place. Indirect immunofluorescence microscopy was then carried out on the transfected cells by using anti-66K antibodies. Figure and J' show a representative epifluorescence micrograph of such cells. In cells where replication occurs, the 66K protein encoded by pOmega-66K appeared targeted to the periphery of chloroplasts, whereas the cytoplasm was almost completely depleted of any fluorescence. Cells transfected with E17-stopDelta transcript alone showed negligible fluorescence . Similar experiments were also performed with the expression vector pOmega-EGFP-66K or pOmega-66K-EGFP that were transfected together with wild-type TYMV RNA to provide a functional 66K protein. Localization of the GFP-fused 66K proteins was observed by fluorescence microscopy of the living cells. As shown by representative epifluorescence micrographs of such cells, both EGFP-66K and 66K-EGFP fusion proteins appeared targeted to the chloroplast envelope in cells where replication occurs. Similar observations were made on fixed cells, although GFP fluorescence was reduced relative to unfixed cells (data not shown). In control experiments, cotransfection of TYMV viral RNA with pOmega-EGFP demonstrated that viral infection had no influence on the distribution of free EGFP, which presented a typical cytosoluble protein distribution pattern (data not shown). Within cotransfected cells, the subcellular distribution of the plasmid-encoded 66K protein and its derivatives, in the shape of rings around the chloroplasts, therefore matched perfectly that of wild-type 66K produced during TYMV viral infection (Fig. and H'). We conclude that during virus multiplication the 66K protein and its derivatives are relocated within the cell and that targeting of the 66K protein to the chloroplast envelope where the replication takes place is dependent upon virus infection. This leads to the assumption that proteins produced during the viral infection (e.g., 69K, CP, 206K, and/or 140K) are necessary to target and/or anchor the 66K protein to the chloroplast envelope. Targeting of the 66K protein to the chloroplast envelope is not dependent on the 69K or CP protein. | In order to evaluate further the role played by each of the TYMV-encoded proteins on the targeting of the 66K protein to the chloroplast envelope, we examined the effect of disrupting either the 69K movement protein or the 20K CP in the viral genome. For that purpose, two mutant plasmids were obtained : E17-stop69K, in which a stop codon truncates the 69K ORF at amino acid 30 without modification of the overlapping 140K ORF, and E17-stopCP, in which the CP gene is interrupted at amino acid 26. In vitro translation experiments in rabbit reticulocyte lysate of the corresponding transcripts revealed the expected profile of expression (data not shown). Arabidopsis protoplasts were transfected with each of the mutant transcripts, and indirect immunofluorescence microscopy was then carried out on the transfected cells by using anti-66K antibodies. As shown by representative epifluorescence micrographs of such cells in Fig. and C, the 66K protein encoded by both transcripts was properly targeted to the chloroplast envelope. No major differences in terms of protein distribution were observed compared to wild-type E17 transcripts transfected in parallel . Similar results were obtained in experiments involving the expression vector pOmega-EGFP-66K, which was cotransfected together with each of the mutant transcripts. In both cases, the GFP-66K protein, observed by fluorescence microscopy of the living cells was efficiently targeted to the chloroplast envelope (Fig. and E). As a complementary approach, we also investigated whether the localization of the 66K protein was affected by the simultaneous expression of the 69K protein or CP. For this purpose, expression vectors encoding the 69K protein or the CP (pOmega-69K and pOmega-CP, respectively) were constructed and Arabidopsis protoplasts were transfected with these plasmids together with the construct pOmega-66K. Indirect immunofluorescence microscopy was then carried out on the cotransfected cells by using anti-66K antibodies. Examination of hundreds of cells revealed that, in both cases, the 66K protein remained in the cytoplasm and was not targeted to the chloroplast envelope. Representative epifluorescence micrographs of such cells are shown in Fig. and G. These experiments clearly demonstrate that the 69K movement protein and CP are not required for the proper targeting of the 66K protein to the chloroplast envelope. Targeting of the 66K protein to the chloroplast envelope is dependent on the 140K protein. | To further assess the role played by the other virus-encoded proteins, the expression vector pOmega-206K encoding the 206K protein was constructed. Upon its transfection into Arabidopsis protoplasts, analysis of protein extracts revealed that the 66K protein was easily detected by immunoblotting with anti-66K antiserum (Fig. , lane 8), therefore demonstrating that the 206K protein produced in this system is properly processed. The inability to detect the uncleaved 206K product is consistent with protoplast infection experiments with viral RNA (Fig. , lanes 1 and 7) and is most likely due to the fast and efficient processing of the 206K protein into its 140K and 66K protein products in plant cells. Arabidopsis protoplasts were transfected with the expression vector pOmega-206K, and indirect immunofluorescence microscopy was then carried out on the transfected cells by using anti-66K antibodies. Figure and A' show representative epifluorescence micrographs of such cells in which the periphery of chloroplasts displayed a bright fluorescent staining. These results demonstrate that expression of the 206K protein allows targeting of the 66K protein to the chloroplast envelope. Similar experiments were also performed by using the expression vectors pOmega-EGFP-66K or pOmega-66K-EGFP that were transfected together with pOmega-206K. Localization of the GFP-fused 66K proteins was observed by fluorescence microscopy of the living cells. As shown by representative epifluorescence micrographs of such cells, both EGFP-66K and 66K-EGFP (Fig. and C') fusion proteins were targeted to the chloroplast envelope. Despite the proper targeting of the 66K protein and its GFP fusion derivatives to the chloroplast envelope, it should be noted that the pattern of localization observed was somehow different from the one observed during the viral infection process. In particular, the fluorescent signal corresponding to the 66K protein often formed partial rings or arcs, as opposed to the images of circles typically seen during viral infection. These arcs appeared to localize preferentially at the junctions between clumped chloroplasts (see arrows in Fig. '). In addition, the clumping of chloroplasts within cells transfected with pOmega-206K was much more severe than the one that typically appears during TYMV infection , leading to the formation of crowded masses of chloroplasts. Precise data concerning the timing of processing of the 206K protein are currently lacking. However, the fact that the unprocessed 206K protein was not detected in protoplasts transfected with pOmega-206K (Fig. , lane 8) and the observation that targeting of the 66K protein to the chloroplasts could be achieved during transcomplementation experiments in which only the 140K and 66K processed products are provided , led to the hypothesis that the 140K cleavage product rather than the unprocessed 206K polyprotein may be involved in mediating chloroplast targeting of the 66K protein. To investigate more precisely the role played by the 140K protein in this process, pOmega-140K, an expression vector encoding the 140K protein was constructed and Arabidopsis protoplasts were cotransfected with the expression vector pOmega-66K together with the pOmega-140K construct. Indirect immunofluorescence microscopy was then carried out on the transfected cells by using anti-66K antibodies. As seen in Fig. D and E (representative epifluorescence micrographs of such cells), when the 66K protein was expressed in the presence of the 140K protein, it appeared distributed at the periphery of chloroplasts. Similar experiments were also performed by cotransfection of the expression vectors pOmega-EGFP-66K or pOmega-66K-EGFP together with the expression vector pOmega-140K. Observation by fluorescence microscopy of the living cells confirmed that, in both cases, the GFP-fused 66K proteins were efficiently targeted to the chloroplast envelope by the simultaneous expression of the 140K protein (Fig. to H). In some cases, the distribution of the 66K protein was in the shape of rings (Fig. and F), but in others the labeling formed arcs that seem to be present preferentially at the junctions between clumped chloroplasts (arrows in Fig. ). Altogether, these results demonstrate that relocalization of the 66K protein to the chloroplast envelope during viral replication is solely dependent on the expression of the 140K protein. The 140K protein localizes to the chloroplast envelope in the absence of other viral components. | To explore the independent localization properties of the 140K protein, the expression vector pOmega-EGFP-140K was constructed, encoding the 140K protein fused to EGFP. Arabidopsis protoplasts were transfected with this plasmid, and the intracellular distribution of the EGFP-140K protein was observed by fluorescence microscopy of the living cells. Representative epifluorescence micrographs of such cells are shown in Fig. to K'. Expression of EGFP-140K induced a clumped chloroplast distribution, demonstrating that the 140K viral protein is involved in that process. The EGFP-140K fusion protein displayed a perfect localization at the chloroplast envelope, in the shape of partial or complete rings, with a stronger labeling at the junctions between adjacent chloroplasts (Fig. ' and K'). These data therefore demonstrate that the 140K protein is targeted to the chloroplast envelope independently of any other TYMV-encoded proteins or RNA replication and is able to induce the clumping of the chloroplasts, one of the typical cytological effects of TYMV infection. DISCUSSION : A universal feature of eukaryotic positive-strand RNA viruses is that replication of their genomes is closely associated with intracellular membranes (reviewed in reference ). In the case of TYMV, replication complexes colocalize with virus-induced membrane vesicles that are thought to result from the invagination of the chloroplast envelope . We showed previously by immunodetection experiments and electron microscopy observations performed on infected plants and protoplasts that the TYMV 66K protein localizes to these membrane vesicles during viral infection . We investigated here the intracellular localization and organellar targeting of TYMV 66K protein in Arabidopsis cells with the aim of defining the determinants for this subcellular localization. We were interested in locating the 66K protein when it is expressed individually in transfected cells, and the effectiveness of noninvasive techniques, such as GFP tagging, had already been documented for this purpose . However, since GFP fusions may debilitate 66K protein function or affect its location in cells, the 66K protein was also expressed with His6 tags or as an unfused version, which could be detected by immunofluorescence. The 66K protein and its derivatives were properly expressed and were readily detectable in transfected cells either by Western blotting or by fluorescence microscopy . In all cases, the 66K protein displayed a cytoplasmic distribution and the presence of the His tag or the GFP moiety had no apparent effect on its subcellular localization. In some cases, discrete spots were observed that did not colocalize with chloroplasts but whose nature was not investigated in details. These foci may result from the aggregation of the 66K protein, since it has been previously reported that the 66K protein formed aggregates when produced in large amounts in insect cells and that similar tendancies to aggregation or oligomerization have previously been reported for other viral RdRps, including Brome mosaic virus (BMV) 2a and poliovirus 3D proteins . Remarkably, the 66K protein produced from the expression vector pOmega-66K was functional since it had the potential to transcomplement a replication-defective RNA . The His-tagged versions behaved comparably, demonstrating that the presence of the His6 tag at one or the other terminus of the protein did not interfere with its functionality. It is clear that the rescued mutant produced less positive-strand RNA than did the wild-type viral RNA, but these amounts were not quantified. The low complementation efficiency obtained in the rescue experiment may be due to the cis-preferential replication of TYMV that has been previously reported . In contrast, fusion of EGFP to the N or C terminus of the 66K protein abolished its ability to support viral replication, possibly because the GFP moiety interfered with the proper folding of the protein or its binding to other components of the replication complex. Under conditions where viral replication occurs, i.e., in the transcomplementation experiments, the 66K protein encoded by pOmega-66K was relocated to the chloroplast envelope . This suggests that the proper targeting of the 66K protein to the replication complexes depends on the expression of other virus-encoded proteins. Interestingly, although the GFP-fused versions of the 66K protein were not functional in replication , their distribution was indistinguishable from that of the functional proteins in coinfection experiments involving wild-type viral RNA . This suggests that the ability of the GFP fusions to establish the proper interactions needed for chloroplast localization is not impaired and that they might constitute suitable markers for studying the determinants of 66K protein localization, as long as a functional 66K protein is provided simultaneously to allow the viral replication to proceed. These data also indicate that the GFP fusion proteins did not interfere in a dominant manner with the proper targeting of the replication complexes to the chloroplast envelope. The results obtained with the various TYMV-derived mutants, as well as in the coexpression experiments, clearly demonstrated that neither CP nor the 69K movement protein is involved in the chloroplastic localization of the 66K protein . These data are consistent with previous results demonstrating that these two viral proteins are not required for viral replication . In contrast, expression of the 140K protein, either in the form of the 206K polyprotein precursor or as an independent protein, allowed the relocation of the 66K protein to the chloroplast envelope . Our studies also demonstrate that the 140K protein is targeted to the chloroplast envelope in the absence of other viral factors and that it induces the clumping of the chloroplasts, a typical cytopathic effect of TYMV infection . Altogether, these findings reveal a novel role of the 140K protein as a key organizer of the assembly of the TYMV replication complexes and a major determinant for their chloroplastic localization and retention. The results presented here address several important aspects of TYMV replication complex formation and reveal multiple parallels between membrane localization of RNA replication complexes by TYMV and other alphavirus-like viruses, such as BMV or Semliki forest virus (SFV). Like BMV 1a and SFV nsP1 , the TYMV 140K protein is able to interact with cellular membranes in the absence of other viral factors and to relocate the viral protein encompassing the RdRp domain (BMV 2a, SFV nsP4, and TYMV 66K, respectively) that otherwise is cytoplasmic . This observation is consistent with the amino acid sequence similarities existing among these proteins and the fact that the 140K protein encompasses the methyltransferase domain, which has been identified as the main membrane association determinant of BMV 1a and SFV nsP1 . In the case of BMV, the recruitment of 2a to intracellular membranes mediated by 1a correlates with their ability to interact, as evidenced in vitro or in two-hybrid experiments . It is thus expected that the TYMV 140K and 66K proteins would also interact with each other, and experiments are currently under way to test this hypothesis. Interestingly, although the assembly of multiprotein replication complexes on intracellular membranes appears to be a general aspect of genome replication by all positive-strand RNA viruses, unknown variations in virus-host interaction lead different viruses to direct replication complex assembly to different membrane sites. For instance, the replication complexes of BMV, SFV, and TYMV target the endoplasmic reticulum, the endosomal or lysosomal membranes, and the chloroplast envelope, respectively . Although a detailed basis for this specificity remains to be determined, it has been proposed that tethering of the replication complexes to particular membranes can be controlled by a direct interaction between a helical peptide region and phospholipids and/or by specific interactions with membrane-bound host proteins . Further experiments are needed to identify the membrane association determinants of the TYMV 140K protein, as well as its eventual chloroplast partner(s). This might be facilitated by the fact that preparative procedures to isolate chloroplast envelopes as well as methods for the detailed analysis of their lipid and protein composition have already been described . The viral replication process often results in proliferation, vesiculation, and severe rearrangement of the cellular membrane compartments. The importance of these rearrangements in virus replication is not yet clear, but it has been suggested that this process may provide a means of increasing the key lipid constituents and/or the membrane surface area used as a scaffold to assemble the replicative machinery. The compartmentalization resulting from the membrane vesiculation may also be critical for viral replication by diverting the viral RNA from the translation machinery , by increasing the local concentration of replication complexes and their associated components , or by avoiding host defense responses directed against double-stranded RNA replicative intermediates . The cytological abnormalities induced by TYMV infection appear to be confined to the chloroplasts. In infected cells, the chloroplasts are swollen, rounded, and clumped together . The data presented here demonstrate that the expression of the sole 140K protein is sufficient to promote the clumping of chloroplasts. This process is therefore unrelated to the production of CP or to the accumulation of TYMV particles between adjacent chloroplasts as previously suggested . Although the molecular mechanisms enabling the 140K protein to alter the intracellular distribution of chloroplasts still remain to be defined, it is noteworthy that its protein counterparts in TMV and BMV (proteins 126K and 1a, respectively) can form multimers that result in a higher-order protein structure . Therefore, we propose a model in which the 140K protein would interact peripherically with the chloroplast envelope membrane and in which intermolecular self-interactions between multiple 140K proteins would result in the formation and stabilization of chloroplast aggregates. It will be interesting to investigate this model in the future. Another conspicuous cytopathic effect of TYMV infection is the occurrence of characteristic double membrane vesicles at the chloroplast periphery that are thought to result from invaginations of the chloroplast envelope into the organelle and that serve as sites for viral RNA replication . Information concerning the origin and biogenesis of these TYMV-induced vesicles is still missing, but it is tempting to draw an analogy with membrane vesiculation in BMV-infected cells, which was reported to be induced by the expression of the sole viral protein 1a . Further ultrastructural studies on cells expressing the TYMV 140K protein will be necessary to clarify whether this individual viral protein also triggers membrane vesiculation processes similar to those occurring during viral infection. This may help to identify the intracellular structures induced by TYMV replication in plant cells and give new insights about the specific virus-host interactions that ultimately determine a successful viral infection. FIG. 1. : Schematic representation of the genomic organization of wild-type TYMV RNA and mutant genomes used in the present study. Schematic representation of the genomic organization of wild-type TYMV RNA and mutant genomes used in the present study. Open bars denote viral ORFs. The encoded 206K protein is proteolytically processed at a peptide bond signified by a filled square. Deletions are indicated by broken bars, and introduced stop codons are indicated by asterisks. FIG. 2. : Expression of the TYMV 66K protein and its derivatives in plant cells. Expression of the TYMV 66K protein and its derivatives in plant cells. Arabidopsis protoplasts were transfected with TYMV RNA (lanes 1 and 7) or plasmids pOmega-66K (lane 2), pOmega-His66K (lane 3), pOmega-66KHis (lane 4), pOmega-EGFP-66K (lane 5), pOmega-66K-EGFP (lane 6), pOmega-206K (lane 8), and water (lane 9). The cells were harvested at 27 h posttransfection, and total proteins were extracted and subjected to SDS-8% PAGE and immunoblot analysis with anti-66K polyclonal antibodies. The position of molecular weight markers (Biolabs) is indicated. FIG. 3. : The 66K protein and its derivatives have a cytoplasmic distribution when expressed individually but relocalize to the chloroplast envelope in infected cells. The 66K protein and its derivatives have a cytoplasmic distribution when expressed individually but relocalize to the chloroplast envelope in infected cells. Arabidopsis protoplasts were transfected with water (A); with the expression plasmids pOmega-66K (B), pOmega-His66K (C), pOmega-66KHis (D and D'), pOmega-EGFP (E), pOmega-EGFP-66K (F), or pOmega-66K-EGFP (G); and with the viral RNA (H and H'). Transfections were also performed with transcript E17-stopDelta alone (I), or together with the plasmid pOmega-66K (J and J'), or with plasmid pOmega-EGFP-66K (K) or pOmega-66K-EGFP (L), together with viral RNA. The protoplasts were collected at 22 h posttransfection (H) or 46 h posttransfection (A to G and I to L), and they were either observed directly for GFP fluorescence (E to G, K, and L) or processed for indirect immunofluorescence labeling by using anti-66K antiserum followed by secondary antibodies coupled to Alexafluor-488 (A to D' and H to J'). The cells were observed by epifluorescence microscopy. Scale bars, 10 mum. Panels D', H' and J' are enlargements of panels D, H, and J, respectively, in which the DAPI staining (blue) of the chloroplastic DNA was acquired and superimposed to the fluorescence signal of the viral protein (green) to visualize the location of chloroplasts. FIG. 4. : The 66K protein and some of its derivatives can support TYMV RNA replication in trans. The 66K protein and some of its derivatives can support TYMV RNA replication in trans. Arabidopsis protoplasts were transfected with transcript E17 (lane 1), transcript E17-stopDelta (lane 2), or combinations of transcript E17-stopDelta with the plasmids pOmega-66K (lane 3), pOmega-His66K (lane 4), pOmega-66KHis (lane 5), pOmega-EGFP-66K (lane 9), and pOmega-66K-EGFP (lane 11). Each of the expression plasmids was also transfected individually (lanes 6 to 8 and lanes 10 and 12, respectively). The transfected protoplasts were collected at 46 h posttransfection and used both for RNA and protein analyses. (A) Equivalent amounts of total RNA prepared from the transfected protoplasts were analyzed by Northern blotting with a single-stranded, DIG-labeled RNA probe complementary to positive-strand TYMV RNA. The migration positions of genomic (g) and subgenomic (sg) RNAs are indicated. (B and C) Equivalent amounts of proteins were analyzed by SDS-PAGE on 15% (B) or 8% (C) polyacrylamide gels. The gels were electroblotted onto a nitrocellulose filter, and proteins were revealed by Western blotting with anti-CP (B) or anti-66K (C) antiserum. FIG. 5. : Targeting of the 66K protein to the chloroplast envelope is not dependent on the 69K movement protein or CP. Targeting of the 66K protein to the chloroplast envelope is not dependent on the 69K movement protein or CP. Arabidopsis protoplasts were transfected with transcript E17 (A), with transcript E17-stop69K alone (B) or together with plasmid pOmega-EGFP-66K (D), with transcript E17-stopCP alone (C) or together with plasmid pOmega-EGFP-66K (E), or with plasmid pOmega-66K together with plasmid pOmega-69K (F) or pOmega-CP (G). The protoplasts were collected at 25 h (A to C) or 46 h (D to G) posttransfection, and they were either observed directly for GFP fluorescence (D to E) or they were processed for indirect immunofluorescence labeling by using anti-66K antiserum, followed by secondary antibodies coupled to Alexafluor-488 (A to C and F to G). The cells were observed by epifluorescence microscopy. Scale bars, 10 mum. FIG. 6. : Targeting of the 66K protein to the chloroplast envelope is dependent on the 140K protein, which localizes to the chloroplast envelope in the absence of other viral components. Targeting of the 66K protein to the chloroplast envelope is dependent on the 140K protein, which localizes to the chloroplast envelope in the absence of other viral components. Arabidopsis protoplasts were transfected with the plasmid pOmega-206K alone (A and A') or pOmega-206K together with the plasmids pOmega-EGFP-66K (B) or pOmega-66K-EGFP (C and C') or with the plasmid pOmega-140K together with the plasmids pOmega-66K (D and E), pOmega-EGFP-66K (F), or pOmega-66K-EGFP (G and H), or with the plasmid pOmega-EGFP-140K (I to K'). The protoplasts were collected at 22 h (A and A') or 46 h (B to K') posttransfection, and they were either observed directly for GFP fluorescence (B to C' and F to K') or they were processed for indirect immunofluorescence labeling by using anti-66K antiserum, followed by secondary antibodies coupled to Alexafluor-488 (A, A', D, and E). The cells were observed by epifluorescence microscopy. Scale bars, 10 mum. Panels A', C', J', and K' are enlargements of panels A, C, J, and K, respectively. In order to visualize the location of chloroplasts, the DAPI staining (blue) of the chloroplastic DNA was acquired and superimposed onto the fluorescence signal of the viral protein (green) in panels A' and E, whereas in panels C', H, and K' the chlorophyll autofluorescence (red) was acquired and superimposed onto the GFP fluorescence (green). Backmatter: PMID- 12915547 TI - Novel Recombinant Parapoxvirus Vectors Induce Protective Humoral and Cellular Immunity against Lethal Herpesvirus Challenge Infection in Mice AB - Orf virus (ORFV; Parapoxvirus ovis) was used to develop a novel vector system for the generation of effective and safe live vaccines. Based on the attenuated ORFV strain D1701-V, recombinants were produced that express the glycoproteins gC (D1701-VrVgC) or gD (D1701-VrVgD) of the alphaherpesvirus of swine, pseudorabies virus (PRV). Expression of gC and gD was also demonstrated on the surface of recombinant virus-infected murine cells that do not produce infectious ORFV. Single or combined immunization with the ORFV recombinants protected different mouse strains of a host species nonpermissive for ORFV against a fulminant, lethal PRV challenge infection equal to immunization with PRV live vaccine. Most notably, even a single immunization with D1701-VrVgC was protective, whereas two applications of D1701-VrVgD were required for immune protection. The higher protective capacity of D1701-VrVgC correlated with the induction of a strong specific humoral immune response. This suggestion was supported by transfer experiments using sera from recombinant-immunized mice, which resulted in partial gC but not gD antibody-mediated protection of the naive recipients. Remarkably, immunization of different immune-deficient mice demonstrated that the application of the PRV gC-expressing recombinant controlled the challenge infection in the absence of either CD4+ or CD8+ T cells, B cells, or an intact perforin pathway. In contrast, D1701-VrVgD-immunized mice lacking CD4+ T cells exhibited reduced protection, whereas animals lacking CD8+ T cells, B cells, or perforin resisted the challenge infection. The present study demonstrates the potential of these new vector vaccines to efficiently prime both protective humoral and cell-mediated immune mechanisms in a host species nonpermissive for the vector virus. Keywords: Introduction : Vaccines based on live virus are excellent inducers of long-term immunity by eliciting protective humoral and cell-mediated immune responses against the inserted antigen. To this end, poxviruses are one of the most versatile expression systems for foreign antigens and have been considered as vectors for human and veterinary live vaccines . Long-term immunity induced by vaccinia virus (VACV) or other poxviruses, however, might result in unsuccessful revaccination or reduced protection against VACV-encoded foreign antigens . For safety reasons, different strategies are used to develop attenuated, host-restricted, or replication-deficient poxviruses, which retain their ability to activate the host's immune response (for review, see references and ). The host range-restricted attenuated VACV strain MVA (modified VACV Ankara) was found to induce lower levels of VACV-neutralizing antibodies than wild-type VACV and is currently widely used as a vector vaccine. It cannot grow in human cells and is propagated on primary chicken embryo fibroblasts (CEFs). Avipoxvirus vectors, which show an abortive replication in mammalian cells, are also produced in CEFs . However, products from CEFs do not represent optimal safety profiles due to different adventitious contaminants, in contrast to production in permanent cell lines. Recently, the genus Parapoxvirus (PPV) of the family Poxviridae, and in particular the type species Orf virus (ORFV), has been proposed as candidate for novel vector vaccines. Arguments in favor of an ORFV vector include the very restricted host range (sheep and goats), its tropism restricted to the skin, the lack of systemic infection, a short-term vector-specific protective immunity, and the exceptionally strong stimulation of fast innate cellular immune mechanisms at the site of infection (for review, see references and ). In addition to cytokines, chemokines, and alpha/beta interferon (IFN-alpha/beta) as part of the host's inflammatory response against the infection, major histocompatibility complex class II-positive dendritic cells accumulate in the infected skin, which represent professional antigen-presenting cells for the subsequent induction of a specific immune response . CD4+ T cells dominate the local accumulation of B and T cells and were found to be of importance for the development of ORFV-specific antibodies . It is worthwhile to stress the short-lived duration of ORFV-specific immunity, which allows frequent reinfections . A most important feature of ORFV in the context with its use as a vaccine is the absence of systemic virus spread, even in immunocompromised individuals or after intravenous injection of high virus doses . Occasional transmission of wild-type ORFV to humans often remains unrecognized . A prime candidate for use as a recombinant vector is the highly attenuated, cell culture-adapted ORFV strain D1701, which is almost apathogenic in sheep . This attenuated virus strain possesses various immunostimulatory properties (for review, see reference ). After adaptation of D1701 on the nonruminant Vero cell line, a new variant (D1701-V) was obtained without altered immunogenic properties and also lacking pathogenicity, even in immunosuppressed sheep . To investigate the immunogenicity of recombinant ORFV against a clinically relevant pathogen, the Alphaherpesvirus of swine, Pseudorabies virus (PRV; Herpesvirus suis type 1) was chosen. The neurotropic PRV has a broad host range with a high mortality, including rodents, which are used as models to investigate the role of viral proteins in neurotropism and neurovirulence of PRV (for review, see reference ). Moreover, mice are commonly used to investigate immunorelevant virus components in the PRV-specific immune response as well as to evaluate the immunogenicity and protective capacity of vaccines against lethal PRV infection. Among the 10 different PRV glycoproteins, particularly the glycoproteins gB, gC, and gD are important for the antiviral humoral and cellular immune responses . Several reports demonstrated some protective effect after passive immunization with anti-gC and anti-gD antibodies and the immunogenic relevance of gC and gD by using recombinant VACVs or glycoprotein-encoding plasmid DNA for immunization . The present study describes the generation of ORFV recombinants expressing the PRV glycoproteins gC and gD, which are correctly processed in ORFV permissive and nonpermissive cell cultures. The attenuated ORFV strain D1701-VrV was used as parental virus for recombinant construction, in which the Escherichia coli lacZ gene replaces the virus-encoded VEGF-E gene, which is a functional homologue of the mammalian vascular endothelial growth factor and represents an important virulence factor of ORFV . The results presented demonstrate the powerful potential of this vector system to protect against a fulminant, lethal herpesvirus infection. Even single immunization with a mixture of both recombinants or with the gC-expressing D1701-VrV recombinant alone protected mice against a lethal PRV challenge infection. Experiments using different immune-deficient mice revealed that the induced PRV glycoprotein gC-specific humoral response is necessary, although not sufficient to control the challenge infection. Moreover, it was found that the ORFV recombinant-induced immune mechanisms are able to compensate for the lack of either B cells, CD4+ or CD8+ T cells, or perforin. MATERIALS AND METHODS : Cells and virus. | The attenuated ORFV strain D1701 , originally propagated in the bovine kidney cell line BKKL-3A, was adapted to the simian cell line Vero (D1701-V) and propagated as described recently . After appearance of cytopathogenic effect (CPE), cells were harvested after trypsin treatment (0.125 mg/ml; Difco, Augsburg, Germany) and centrifuged at 30,000 x g, and the resulting pellet was sonified. Titration and plaque purification of virus were performed with Vero cells in six-well plates with 0.9% (wt/vol) agarose (SeaPlaque agarose; BMA, Rockland, Maine) in minimal essential medium. The recombinant ORFV D1701-VrV contains the functional E. coli lacZ gene replacing the viral VEGF-E gene, which is present in two copies due to its location in the inverted terminal repeats of the D1701 genome (Fig. and B) After removal of the VEGF-E gene, a lacZ gene cassette was inserted into the EcoRV site of plasmid pdV550 . Construction of recombinant transfer plasmids. | For insertion into the VEGF-E gene locus of ORFV D1701-V, the PRV glycoprotein genes were cloned into plasmid pdV-Rec1, which was obtained by linker insertion of a multiple cloning site into pdV550 . The complete coding sequence of the PRV gC gene was excised from plasmid pALM-20 (generously provided by L. W. Enquist, Princeton University, Princeton, N.J.) as a 1.47-kbp NcoI-HinfI fragment . After fill-in reaction with Klenow polymerase, this fragment was blunt end ligated into the single EcoRV site of pdVRec1, resulting in plasmid pdVgC. The complete PRV gD gene was obtained from plasmid pgDBSII (kindly provided by A. Jestin, AFSSA, Ploufragan, France) by HindIII-BamHI digestion and subsequently ligated into the HindIII and BamHI sites of pdVRec1, resulting in plasmid pdVgD. The correct construction of all plasmids was verified with restriction enzyme digests and DNA sequencing. Generation and selection of recombinant viruses. | For the generation of ORFV recombinants expressing PRV gC (D1701-VrVgC) or gD (D1701-VrVgD), Vero cells were infected with D1701-VrV at a multiplicity of infection (MOI) of 0.1 50% tissue culture infectious dose (TCID50) and were transfected 2 h later with 2 mug of the recombinant transfer plasmid (pdVgC or pdVgD) and 4 mul of SuperFect transfection reagent according to the recommendations of the manufacturer (Qiagen, Hilden, Germany). Four days later, the cells were lysed by multiple freeze-thawing cycles and plaque titrated on Vero cells. Using a 1% SeaPlaque agarose overlay containing 300 mug of Bluo-Gal per ml (Invitrogen Life Technologies, Karlsruhe, Germany), white virus plaques could be identified after successful exchange of the lacZ gene cassette in D1701-VrV. Immunostaining of virus plaques with a polyclonal goat anti-PRV serum (diluted 1:500), peroxidase-conjugated antigoat immunoglobulin G (IgG) antibodies (diluted 1:500; Dianova, Hamburg, Germany), and the Vector-VIP substrate kit for peroxidase (Vector Laboratories, Burlingame, Calif.) confirmed expression of the foreign genes. After three to four consecutive rounds of plaque purification, virus stocks were prepared in Vero cells. For immunization, virus stocks were briefly sonicated, and cell debris was removed by centrifugation before virus titration. DNA analyses. | Viral DNA of the recombinant viruses was prepared according to the alkaline lysis procedure followed by phenol extraction as described in reference . Restriction endonuclease analysis of the purified viral DNA and Southern blot hybridization were performed as described previously with PRV glycoprotein gene-specific or lacZ gene-specific, radioactively labeled probes. RNA isolation and Northern blot hybridization. | For enrichment of viral early RNA, cells were pretreated with 100 mug of cycloheximide (CH) per ml for 1 h before infection (MOI of 10). After an additional 8 h in the presence of CH, total cellular RNA was prepared with TriZOL reagent (Invitrogen Life Technologies, Leek, The Netherlands). RNA (10 mug) was separated in formaldehyde-containing agarose gels and transferred onto nylon membranes (Hybond N+; Amersham Biosciences, Freiburg, Germany). Hybridization using PRV glycoprotein gene-specific or ORFV ANK3 gene-specific probes was performed overnight at 45C in a mixture containing 3x SSPE (1x SSPE is 0.18 M NaCl, 10 mM NaH2PO4, and 1 mM EDTA [pH 7.7]), 1.0% nonfat dried milk, 2.0% sodium dodecyl sulfate (SDS), 0.5% (vol/vol) diethyl-pyrocarbonate (DEPC; Serva, Heidelberg, Germany), 7.0% (wt/vol) dextran sulfate, and 60% deionized formamide (Invitrogen Life Technologies). Western blot analysis. | Cells infected at an MOI of 10 were harvested, collected by centrifugation, and lysed in a mixture of 12.5 mM Tris-HCl (pH 6.8), 2.5% SDS (wt/vol), 10% sucrose (wt/vol), 0.02% bromphenol blue (wt/vol), and 5% 2-mercaptoethanol (vol/vol). Total cell proteins were resolved by electrophoresis in an SDS-10% ProSieve 50 (BMA) gel and electroblotted onto a polyvinylidene difluoride Western blotting membrane (Roche Applied Sciences, Mannheim, Germany) for 30 min at 100 V in a mixture of 25 mM Tris, 192 mM glycine, and 20% (vol/vol) methanol (pH 8.6). The membranes were blocked in TBST buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 0.05% Tween 20), containing 5% nonfat dried milk for 1 h and then incubated with primary antibody for 2 h at room temperature. To this end, undiluted PRV gC-specific mouse monoclonal hybridoma supernatant A18b containing 1% nonfat dried milk, the polyclonal PRV gD-specific rabbit antiserum 016/00 (diluted 1:2,500 in TBST containing 1% nonfat dried milk), or the monoclonal antibody (MAb) 4D9 (diluted 1:100) directed against the 39K major envelope protein of ORFV was used. After being washed with TBST, the membranes were incubated for 1 h with a 1:100 dilution of species-specific secondary anti-IgG (H+L) antibodies conjugated to peroxidase (Dianova) followed by staining with phosphate-buffered saline (PBS) containing 20% methanol, 0.02% diaminobenzidine, 0.06% chloronaphthol, and 0.006% H2O2. Fluorescent analysis of protein surface expression. | For flow cytometric analyses using the FACSCalibur fluorescence-activated cell sorter (FACS) (Becton Dickinson, Heidelberg, Germany), cells were harvested 24 h after infection (MOI of 10) and immunostained with the PRV gC-specific MAb A18b or with the 1:100-diluted PRV gD-specific antiserum 016/00 and 1:100-diluted species-specific secondary anti-IgG (H+L) antibodies conjugated to fluorescein isothiocyanate (Dianova). All incubations and washing steps were performed at 4C in PBS containing 3% fetal bovine serum. Mice. | Wild-type BALB/c, C57BL/6, and 129/Sv/Ev mice and mutant CD4-/- 129/Sv/Ev, B-cell deficient C57BL/6 (muMT), CD8-/- C57BL/6, or perforin-/- C57BL/6 (PKOB) mice were used. All of these mice were kindly provided by R. M. Zinkernagel (Institute for Experimental Immunology, University of Zurich, Zurich, Switzerland) and bred at the Federal Research Center for Virus Diseases of Animals, Tuebingen, Germany. Mice of both sexes were challenge infected at 8 to 10 weeks of age. Immunization and challenge infection of mice. | Mice were injected intramuscularly (i.m.) with 107 TCID50s of the ORFV recombinants expressing the PRV glycoproteins in a total volume of 0.2 ml (0.1 ml for each hind leg). Immunization was repeated at 2-week intervals, and 2 weeks after the last immunization, mice were bled from the retroorbital plexus for serum collection. Immediately thereafter, the BALB/c mice were challenge infected intraperitoneally (i.p.) with 102 PFU (corresponding to 30 50% lethal doses [LD50s]) of the highly virulent PRV strain NIA-3, and the C57BL/6 and 129/Sv/Ev mice were infected with 103 PFU (corresponding to 25 LD50s). The LD50 was determined for each wild-type mouse strain (four animals per group) by i.p. infection using serial dilutions of PRV NIA-3 ranging from 105 to 3.3 PFU, and was calculated as described previously . As negative controls, mice were immunized with 107 TCID50s of the parental ORFV D1701-VrV, and as positive controls, mice were immunized with 107 PFU of the PRV live vaccine Begonia (Intervet International BV, Boxmeer, The Netherlands). Antibody analyses. | Sera from immunized and control mice were analyzed for PRV-specific IgG1 and IgG2a antibodies by enzyme-linked immunosorbent assay (ELISA) using PRV-coated microtiter plates and substrate of the commercially available Checkit Aujeszkytest II (Intervet, Unterschleissheim, Germany). The plates were incubated with serial twofold dilutions (starting with a 1:80 dilution) of each serum for 90 min at 37C, washed as recommended, and incubated with 1:5,000-diluted biotinylated goat anti-mouse IgG1 or biotinylated goat anti-mouse IgG2a antibody (Southern Biotechnology, Birmingham, Ala.) for 90 min at 37C. The plates were washed and incubated with a 1:2,000 dilution of peroxidase-conjugated avidin (BD PharMingen) for 1 h at 37C. Antibody titers were expressed as the reciprocal of a serum dilution exhibiting an at least twofold increase in optical density over that of the negative control serum that was obtained from D1701-VrV-immunized mice. The standard deviation was calculated from the mean titer of sera obtained from the individual mice (eight animals per group immunized separately with the individual ORFV recombinants and five animals per group vaccinated with PRV live vaccine or with a combination of both recombinants). Serum neutralizing antibodies directed against PRV were determined by a complement-independent neutralization assay. Sera were inactivated at 56C for 30 min, and twofold dilutions of the sera (starting with a dilution of 1:20) were incubated in triplicates together with 50 PFU of PRV strain NIA-3 for 90 min at 37C in flat-bottom 96-well plates. Thereafter, Vero cells were added, and 3 days later, the serum neutralizing antibody titer was determined as the highest serum dilution resulting in 100% reduction of CPE. RESULTS : In vitro characterization of D1701 recombinants in ORFV permissive cells. | Virus recombinants were selected and plaque purified as described in Materials and Methods. The correct substitution of both copies of lacZ by the PRV gC and gD genes, respectively, was verified by Southern blot hybridization with probes specific for lacZ, gC, or gD (data not shown). Specific transcription of the PRV glycoprotein genes, which are controlled by the strong early VEGF-E promoter, was examined by Northern blot analysis of total RNA isolated 8 h postinfection (p.i.) from CH-treated cells. The results demonstrated specific transcription of early mRNA of the expected size for gC (Fig. lane 1) as well as for gD (Fig. , lane 2). To control for comparable levels of viral gene expression in the cells infected with the individual recombinants, the transcription rate of the early ORFV gene ANK-3 was tested. As shown in Fig. , cells infected with D1701-VrVgC (lane 1), D1701-VrVgD (lane 2), or the parental virus, D1701-VrV (lane 3), synthesized comparable amounts of viral mRNA. In addition, each PRV glycoprotein gene-specific probe detected a transcript approximately 3.8 kb in size, which was not found in uninfected cells (lane 4). This can be explained by partial read-through of the early transcription stop motif (T5NT) of the original VEGF-E by using another stop motif 1.65 kb downstream, as also found for the ANK-3 gene. In addition, Northern blot analysis also demonstrated substantial foreign gene expression at later times p.i. in the absence of CH (data not shown), which is also found for the VEGF-E gene of D1701-V (M. Henkel and H.-J. Rziha, unpublished data). Translation of the PRV glycoproteins was tested in cells infected with each D1701-VrV recombinant by Western blot analysis. Using the gC-specific MAb A18b, all known processing products of gC, the 58-kDa precursor, the 74-kDa pre-Golgi form, and the 92-kDa mature form could be demonstrated in D1701-VrVgC-infected cells (Fig. , lane 1) as in PRV-infected cells (Fig. , lane 5). Similarly, cells infected with D1701-VrVgD exhibited authentic expression of gD compared to PRV-infected cells. The gD-specific polyclonal antiserum detected the precursor (45 kDa) and the mature glycosylated form (60 kDa) of gD . Comparable synthesis of both forms of the ORFV major envelope protein (39 and 31 kDa) was found in Vero cells infected with the recombinant viruses and the parental D1701-VrV , which indicated similar viral protein synthesis. Finally, flow cytometry demonstrated surface expression of gC or gD on Vero cells infected with the individual D1701-VrV recombinants, but an approximately 10-fold-higher synthesis of gC than of gD (Fig. and D). This is in agreement with the Western blot results, although a comparable amount of the specific transcripts seemed to be synthesized in the productively infected Vero cells , and both recombinants grew to equal virus titers (data not shown). From the experiments done, the reasons for a lower level of gD synthesis in permissive Vero cells compared to the level of gC synthesis remain obscure and need further investigation. Expression of gC and gD in D1701-VrV recombinant-infected cells nonpermissive for ORFV. | Since one objective of this study was to evaluate the efficiency of the D1701-VrV recombinants in mice, a nonpermissive host for ORFV, the foreign gene expression and production of infectious progeny were tested in mouse cell lines. Northern and Western blot analyses of 3T3 and L929 cells infected with D1701-VrVgC3 or D1701-VrVgD showed the correct expression of both PRV glycoproteins (data not shown). In contrast to the permissive Vero cells, surface expression of comparable amounts of gC and gD was found by flow cytometry in both mouse cell lines infected with the individual recombinant virus (Fig. , C, E, and F). ORFV production was tested by titration experiments (single-step virus growth curve) in 3T3 and L929 cells infected with a D1701-VrVgC3 MOI of 1.0 or 10.0. Production of infectious progeny was detectable in neither cell lysates (Fig. and C) nor supernatants of both infected mouse cell lines (Fig. and D), whereas high titers of the recombinant viruses were produced in Vero cells (Fig. , solid triangles) indistinguishable from those with the parental D1701-VrV (data not shown). Collectively, the results demonstrate the in vitro expression of the inserted PRV glycoproteins also on the surface of infected cells nonpermissive for ORFV. PRV-specific serum antibody response in immunized mice. | To evaluate the immunogenicity of the PRV gC- and gD-expressing ORFV recombinants, three different mouse strains, BALB/c, 129/Sv/Ev, and C57BL/6, were immunized i.m. up to three times at 2-week intervals. Groups of animals were vaccinated with D1701-VrVgC and D1701-VrVgD alone or in combination at 107 TCID50s per dose. As controls, animals were immunized with the same dose of a PRV live vaccine (Begonia), with the parental D1701-VrV, or with PBS. PRV-specific serum antibodies of IgG1 and IgG2a subclasses were determined by ELISA. As shown in Fig. immunization of BALB/c mice with the PRV live vaccine induced higher specific IgG2a than IgG1 serum antibody titers, very similar to the results found after a single combined immunization with D1701-VrVgC plus D1701-VrVgD. A second combined immunization increased the specific IgG1 and IgG2a antibody titers 16-fold and 4-fold, respectively, whereas a third immunization did not change that picture . Essentially the same results were found in the various immunization regimens in 129/Sv/Ev mice . In contrast, significantly lower titers of IgG2a serum antibodies to PRV were found after all immunizations of C57BL/6 mice, resulting in a slight prevalence of the specific IgG1 subclass . Individual immunizations with both D1701-VrV recombinants were performed to determine their contribution to the antigen-specific immune response. The ELISA results demonstrated that the specific IgG1 and IgG2a serum antibody titers induced by D1701-VrVgC in the different mouse strains were nearly identical to those found after the first and second immunizations with the combination of both recombinants . In contrast, no specific serum antibodies were detectable after a single application of D1701-VrVgD, and after the second immunization, approximately 30- to 85-fold-lower titers of PRV-specific IgG1 and IgG2a antibodies were induced in all mouse strains (Fig. to C). Due to the generally low induction of specific antibodies by D1701-VrVgD, no antigen-specific IgG2a serum antibodies were detectable after the second immunization of C57BL/6 mice . These results suggest a prevalence of a PRV gC-specific antibody response after simultaneous immunization with D1701-VrVgC and D1701-VrVgD. Finally, sera were tested for the presence of complement-independent PRV-neutralizing antibodies . All sera obtained after the first immunization with either the PRV live vaccine or with each D1701-VrV recombinant alone or in combination did not contain PRV-neutralizing activity (cutoff, dilution of 1:20). After booster immunization, the sera of all three mouse strains exhibited low titers of PRV-neutralizing antibodies. A third combined immunization with both recombinants slightly increased the titers of PRV-neutralizing serum antibodies . In conclusion, these results indicate that D1701-VrVgC and D1701-VrVgD were both able to induce comparable amounts of antigen-specific neutralizing antibodies in serum despite their clearly different capabilities of inducing antigen-specific IgG antibodies. Protective capacity of D1701-VrVgC3 and D1701-VrVgD. | The protection experiments were performed with a 25- to 30-fold LD50 of PRV for challenge infection of the different mouse strains (eight animals per group) and are summarized in Table . All nonimmunized (PBS) or D1701-VrV-immunized animals succumbed to challenge infection within 72 to 96 h, whereas a single application of the PRV live vaccine (Begonia) protected all wild-type mouse strains. Even a single i.m. application of the combination of D1701-VrVgC and D1701-VrVgD mediated 100% protection to C57BL/6 or 129/Sv/Ev mice, and seven of eight BALB/c mice were protected . The same protection rates were found after a single immunization with D1701-VrVgC alone. Using D1701-VrVgD, seven of eight C57BL/6 or 129/Sv/Ev mice, but only one of eight immunized BALB/c mice, survived the challenge infection . Thus, priming of BALB/c mice with D1701-VrVgD was much less protective than with D1701-VrVgC. Two separate immunizations with each recombinant protected all animals of each wild-type mouse strain, except for one BALB/c mouse immunized with D1701-VrVgD . All protected animals were observed over a period of up to 9 weeks and did not show clinical symptoms at any time after challenge infection. Interestingly, BALB/c mice were also completely protected against a 10-fold-higher dose of challenge virus after two immunizations with D1701-VrVgC alone or with a combination of both recombinants, but only 50% of BALB/c mice immunized twice with D1701-VrVgD survived this higher infectious dose (data not shown). The data show that already a single application of the new ORFV recombinants could mediate protection of the different mouse strains against a lethal PRV infection, whereby the recombinant D1701-VrVgC appeared to have a higher protective capacity. Serum transfer experiments. | Since a strong antibody response was induced by the gC recombinant, the role of antigen-specific serum antibodies in protection against lethal challenge infection was further scrutinized. Therefore, sera from BALB/c mice immunized twice with either D1701-VrVgC or D1701-VrVgD were used for passive immunization of naive recipients. Twenty-four hours after intravenous transfer of 0.1, 0.3, or 0.5 ml of serum, the antigen-specific antibody titers of the recipients were determined immediately before challenge infection (data not shown). Control animals received 0.5 ml of serum from mice immunized twice with D1701-VrV, and all died between 72 and 96 h after challenge infection . Transfer of 0.1 ml of serum from D1701-VrVgC-immunized mice (DgC serum) did not protect against challenge infection, although the specific serum antibody titers (1:530 titer of IgG1, 1:1,200 titer of IgG2a) were comparable to those of BALB/c mice after one active immunization with D1701-VrVgC mediating protection (Fig. and ). However, two BALB/c mice died 30 to 50 h later compared to the controls (Fig. , squares). After passive immunization with 0.3 ml of DgC serum, the recipients exhibited higher serum antibody titers (1:1,920 titer of IgG1, 1:2,970 titer of IgG2a) than mice after a single immunization with live D1701-VrVgC; nevertheless, only one of six animals was protected. However, again two animals survived 30 to 50 h longer than the controls (Fig. , circles). Finally, compared to actively single-immunized BALB/c mice, approximately threefold-higher antigen-specific antibody titers were found in the sera of animals passively immunized with 0.5 ml of DgC serum, but again only a low level of protection against the PRV challenge infection (two of six mice) was obtained. Notably, the death of four mice of this group was significantly delayed (Fig. , triangles). These results indicate that PRV gC-specific serum antibodies alone are not sufficient for protection. Passive immunization with sera from D1701-VrVgD-immunized mice did not result in detectable antibody titers: none of the animals was protected against challenge infection, and the time to death of only two animals, each receiving 0.3 or 0.5 ml of serum, was prolonged by 50 h . Immunization of immune-deficient mice. | To further elucidate the role of the humoral response in controlling the challenge virus, experiments with B-cell-deficient muMT mice were performed. Due to the finding that the protection rate after combined immunizations equaled that after immunization with D1701-VrVgC alone, in the following experiments, mice were only immunized with the individual recombinants. Protection of five of eight muMT mice was found after single administration of D1701-VrVgC or D1701-VrVgD, respectively (data not shown), whereas after two immunizations with each recombinant alone, seven of eight mice were protected . With the PRV live vaccine, four of five muMT mice were protected . These results demonstrate that the Ig-deficient mice are still capable of controlling challenge infection after the different immunizations. It has been reported that successful protection of mice against lethal PRV infection might depend on IFN-gamma-producing CD4+ T cells . To investigate whether similar immune effector mechanisms might be responsible for the protective effect of the new PRV glycoprotein-expressing ORFV recombinants, 129/Sv/Ev CD4-/- knockout mice were investigated. The specific antibody response was generally lower in CD4-/- mice after immunization with the ORFV recombinants than in wild-type 129/Sv/Ev mice . Single immunization with D1701-VrVgC induced two- to threefold-lower PRV-specific IgG1 and IgG2a serum antibody levels than in wild-type mice, and antigen-specific serum antibodies were not detectable even after two immunizations with D1701-VrVgD. PRV live vaccination (Begonia) of the CD4-/- mice induced no detectable IgG1 and 16-fold-lower titers of specific IgG2a serum antibodies compared to those in wild-type mice. Furthermore, none of the sera obtained from the different immunizations exhibited PRV-neutralizing activity (data not shown). After different single immunizations, CD4-/- mice were clearly less protected against challenge infection than were the congenic 129/Sv/Ev mice. Single administration of D1701-VrVgC protected 62.5% of the CD4-deficient animals, but single administration of D1701-VrVgD protected only 25% (data not shown), and only two of four PRV live-vaccinated mice survived the challenge infection. However, a booster immunization with D1701-VrVgC protected all CD4-/- mice against challenge infection, but only 50% of mice survived the challenge infection after two administrations of the D1701-VrVgD recombinant . Altogether, protection of CD4-/- mice by a single application of D1701-VrVgC was reduced, which appeared to correlate with the reduced induction of specific antibodies in serum, but could be compensated for by a second immunization. This was not the case with D1701-VrVgD, although partial protection of the CD4-/- mice could be achieved. The reduced protection of CD4-deficient mice mediated by the ORFV recombinants might be a consequence of a reduced capacity of cytolytic T-cell effector mechanisms in those animals. Therefore, challenge experiments with perforin knockout (PKOB) and CD8-deficient mice (which had been immunized twice with each recombinant alone) were performed. The results, however, demonstrated 100% protection of PKOB or CD8-/- mice after administration of D1701-VrVgC or D1701-VrVgD, as well as of the PRV live vaccine . Therefore, these results indicate that after immunization with the new ORFV recombinants, the control of the PRV challenge infection can be maintained in the absence of CD4+ or CD8+ T cells, and the perforin pathway seems to be not required for immune protection. DISCUSSION : Like other poxviruses, ORFV possesses properties suitable for the development of virus vector vaccines (i.e., the stable incorporation of foreign DNA, easy propagation, and inexpensive production). In addition, the restricted virus host range in vivo, as well as its immunomodulatory properties , makes ORFV an interesting candidate for novel virus vector vaccines. So far, however, the potential of recombinant ORFV to induce protective immunity against microbial infections has not been investigated. In this report, for the first time, we describe the development of novel ORFV recombinants for vaccination against a fulminant herpesvirus infection and demonstrate their protective capacity in mice as a nonpermissive host species for ORFV replication. A single application of the individual PRV glycoprotein expressing ORFV recombinants alone or in combination conferred solid protection against PRV infection to the majority of animals of the different mouse strains tested. Also using a 10-fold-higher infectious dose of 300 LD50s, 75 and 25% of the BALB/c mice immunized with the recombinants expressing glycoprotein gC or gD, respectively, survived the challenge infection. Most importantly, even a single vaccination with D1701-VrVgC gave protection from lethal challenge infection equal to that provided by vaccination with the PRV live vaccine. In contrast, twofold vaccination with D1701-VrVgD was required for solid protection. In addition, we found high titers of gC-specific serum IgG antibodies in D1701-VrVgC-immunized animals, but significantly lower titers of gD-specific antibodies after administration of D1701-VrVgD. This indicated an important role for the vector-induced humoral immunity in protection against PRV infection and might explain the lower protective capacity of the recombinant expressing glycoprotein gD. PRV glycoprotein gC represents one of the major viral proteins that trigger the humoral immune response in mice and swine, and gC-specific MAbs can be sufficient to protect against PRV infection . This has been attributed to the neutralizing activity of antibodies specific for the heparin binding domain of this glycoprotein, which is necessary for host cell attachment . However, despite nondetectable PRV-neutralizing activity in sera of mice after a single vaccination with the ORFV gC recombinant, the majority of animals were protected against PRV infection. Even after a second application, the gC-specific neutralizing antibody titers were low, coinciding with observations made after vaccination of mice with a PRV live vaccine , a gC-specific DNA vaccine , or purified glycoprotein gC . In the present study, intravenous transfer of sera obtained from mice immunized twice with D1701-VrVgC resulted in delayed mortality, but could not solidly protect the recipient mice, even when serum antibody titers before PRV challenge infection were higher than those found in mice protected after a single active immunization with D1701-VrVgC. This is in contrast to the protection of mice after transfer of sera obtained from mice immunized with inactivated PRV . In that study, however, serum administration at the same site immediately before i.p. challenge infection might result in rapid inactivation of the virus, but this does not entirely reflect the in vivo situation of an actively immunized host. Furthermore, neutralizing antibodies directed against other components of the virus might contribute to the protective capacity of the sera from mice immunized with whole virus particles. In conclusion, our results indicate that gC-specific antibodies induced by the ORFV recombinant play an important role early after PRV infection, but are not solely responsible for solid long-term protection. This is corroborated by the observation that even B-cell-deficient mice were protected after vaccination with the gC recombinant, which indicates that cellular mechanisms of immunity can counterbalance the lack of specific antibodies. The importance of cell-mediated immunity in protection against virus infections is generally acknowledged. In herpesvirus infections, including PRV, the development of antigen-specific CD4+ effector T cells has been frequently reported to be important for protective immunity in mice . Using CD4-deficient mice, we also found that the lack of antigen-primed CD4+ T cells in PRV Begonia-immunized mice results in impaired resistance against PRV infection. The PRV-specific immunity induced by a single administration of the ORFV recombinants expressing PRV glycoprotein gC or gD was less protective in CD4-deficient mice than that in congenic wild-type animals. Thus, both glycoprotein-specific vector vaccines obviously enable priming of antigen-specific CD4+ T cells, participating in protection against PRV challenge infection. Most notably, however, a second immunization with D1701-VrVgC, but not with D1701-VrVgD, induced substantial immune serum antibodies independent of T-helper cells (only twofold-lower titers than in congenic mice), and subsequently all CD4-/- mice were protected against challenge infection. The generation of antiviral serum IgG antibodies with the help of CD4- CD8- T cells has been reported in CD4-/- mice . However, as discussed above, transfer of the induced gC antibodies alone was not sufficient for solid protection, and, therefore, the presence of immune T cells seems to be required for full protective capacity of immune serum antibodies, as also reported for herpes simplex virus type 2 (HSV-2) infection . The D1701-VrVgD recombinant exhibits a rather limited capability to induce gD-specific humoral immunity. The approximately 10-fold-lower translation (Fig. and E) and surface expression (Fig. and D) of gD compared to gC in Vero cells productively infected with the recombinants might explain the reduced specific humoral response. However, comparable levels of synthesis of both PRV glycoproteins were demonstrated in the recombinant-infected murine cells that do not support production of infectious ORFV. Although not proven in vivo, it seems unlikely that substantially smaller amounts of gD than of gC are synthesized in mice immunized with each recombinant. Thus, the protective capacity of D1701-VrVgD seems to be primarily dependent on the induction of cell-mediated immune mechanisms. So far, the effector functions of the vector vaccine-induced CD4+ T cells remain to be clarified. Besides providing help to antigen-specific B cells, they might also be important for the generation of PRV-specific CD8+ cytotoxic T lymphocytes (CTLs). Recently it has been reported that CD4+ T cells are dispensable for primary differentiation of virus antigen-specific CD8+ CTLs, but secondary CTL expansion upon reencounter with antigen is wholly dependent on the presence of T-helper cells during the priming process . However, although CTLs are generated in PRV-vaccinated mice and swine, CD8+ T cells have been shown to be dispensable for protection against subsequent lethal PRV infection, at least in the murine model . Similarly, our results using CD8 knockout mice indicate that effector functions of antigen-primed CD8+ T cells are of minor importance as well for the ORFV vector vaccine-mediated protection against PRV infection. In addition, the complete protection of perforin knockout mice or of NK cell-depleted mice (data not shown) after twofold vaccination with the ORFV recombinants argues against the necessity of NK cell-mediated cytolysis of infected target cells for protection. Collectively, these observations are in agreement with the general concept that lymphocyte-mediated cytotoxicity particularly is important for the elimination of noncytolytic viruses prior to their release from infected cells, whereas cytopathic viruses, like herpesviruses, are controlled most efficiently by antiviral interleukins or IFNs . Previous studies have shown that a functional IFN-gamma system is important for protective PRV-specific immunity induced by PRV vaccines and that IFN-gamma secretion by antigen-specific CD4+ T-helper cells reduces morbidity and mortality as well of HSV-infected mice . In the presence of the Th1 cytokine IFN-gamma, an enhanced expression of inducible nitric oxide synthetase (iNOS) by macrophages has been reported to improve destruction of phagocytically ingested virus particles . Characteristic for a Th1-type immune response is the IFN-gamma-induced expression of Igs of the IgG2a isotype . In immunized C57BL/6 mice, we observed an inverse relationship between antigen-specific IgG1 and IgG2a isotypes as compared to that in the immunized BALB/c and 129/Sv/Ev mice. However, that should not be misinterpreted as a shift towards a Th2-type immune response. In C57BL/6 mice, the gene for IgG2a is deleted and the isotype IgG2c is expressed , and commercial anti-IgG2a antibodies do not cross-react substantially with IgG2c . Interestingly, we have found that in BALB/c mice, the ORFV recombinants are less protective than in the other mouse strains tested. BALB/c mice have been shown to exhibit higher susceptibility to microbial infections, including herpesvirus infections, due to being less competent to elicit Th1-controlled cellular immunity than, for example, C57BL/6 mice . Altogether, the cytokine secretion pattern of the vector vaccine-induced PRV-specific T cells, particularly in the absence of CD4+ T cells, and the importance of IFN-gamma now need to be investigated to elucidate additional immune effectors stimulated by the protective ORFV recombinants. Importantly, preliminary experiments indicate that the ORFV recombinants elicit their protective immunogenicity even with preexisting immunity to the viral vector. Prime immunization of BALB/c mice with the parental ORFV recombinant D1701-VrV did not impair the development of a gC-specific humoral immune response and protection against PRV infection after subsequent application of the recombinant D1701-VrVgC (unpublished observations). This might be due to the short-lived ORFV-specific adaptive immunity even in sheep, where productive replication of the virus occurs . Thus, it seems likely that in a nonpermissive host species, the induced ORFV-specific immunity does not prevent vector-encoded foreign gene expression. Hence, induction of immunity to different pathogens by sequential immunization with diverse ORFV vector vaccines should be possible. In this regard ORFV vector vaccines might be superior to recombinants based on VACV, where with preexisting VACV-specific immunity, a limited foreign antigen-specific primary immune response was found in murine and primate model systems . In addition, propagation of recombinants based on host range-restricted and replication-deficient VACV and avipoxvirus strains requires cumbersome preparation of CEFs or complementing cell lines . VACV strain MVA was found to regain growth in Vero cells , but the safety of that Vero cell-derived virus variant remains to be shown. In contrast, ORFV recombinants merge the advantage of a natural restricted host range with the ability to be maintained in a permanent cell line, which meet the criteria of state-of-the-art production of biologicals. Current studies shall substantiate the capability of the ORFV recombinants to induce an antigen-specific memory immune response. In addition to protection against lethal infection with a cytolytic herpesvirus, the ORFV vector system was successfully used to protect rats, also nonpermissive for ORFV, against challenge infection with the neurotropic, noncytolytic Borna disease virus even 8 months after the last immunization (unpublished data). This indicates a more common applicability of ORFV recombinants to control acute or persistent virus infections requiring different immune defense mechanisms . Therefore, studies are needed to dissect the immunostimulatory properties of ORFV in more detail. FIG. 1. : Construction of the ORFV recombinants (A) The map locations of HindIII fragments and of the inverted terminal repeats (ITR) of the genome of ORFV strain D1701-V are depicted. Construction of the ORFV recombinants (A) The map locations of HindIII fragments and of the inverted terminal repeats (ITR) of the genome of ORFV strain D1701-V are depicted. (B) The PstI-HindIII fragment containing the VEGF-E and adjacent genes was cloned as plasmid pORF-PA. (C) The singular StyI restriction site in pORF-PA was used to delete the VEGF-E gene by a bidirectional Bal31 digest that resulted in plasmid pdV-550. (D) A synthetic linker covering the indicated restriction sites was inserted into the EcoRV site. The obtained plasmid, pdV-Rec1, contains the early promoter of VEGF-E (Pvegf-e) and the original early transcription stop motif T5NT. (E) The NcoI-HinfI fragment of plasmid pALM-20 containing the complete PRV gC gene was blunt-end ligated into the EcoRV site of pDV-Rec1, resulting in plasmid pdV-gC. (F) The use of the HindIII-BamHI fragment of plasmid pgDBSII allowed cloning of the complete gD gene of PRV in pdV-Rec1 to obtain plasmid pdVgD. FIG. 2. : Expression of PRV gC and gD in ORFV recombinant-infected cells. Expression of PRV gC and gD in ORFV recombinant-infected cells. Vero cells were infected (MOI of 10) with D1701-VrVgC (lanes 1), D1701-VrVgD (lanes 2), D1701-VrV (lanes 3), and PRV Begonia (lanes 5) or mock infected (lanes 4). For Northern blot hybridization using radioactively labeled probes specific for PRV gC (A), PRV gD (B) and the ORFV early gene ANK3 (C), total RNA was isolated from CH-treated Vero cells. The sizes of the respective specific transcripts are indicated in kilobase pairs to the left. Cell lysates were obtained 24 h p.i., and Western blot analysis was performed with the gC-specific MAb A18b (D), the gD-specific rabbit serum 016/00 (E), or the ORFV 39K-specific MAb 4D9 (F). The apparent molecular mass of the detected proteins is indicated in kilodaltons. FIG. 3. : Surface expression of PRV gC and gD on ORFV recombinant-infected cells. Surface expression of PRV gC and gD on ORFV recombinant-infected cells. Flow cytometry of the indicated cells 24 h after infection with D1701-VrVgC (A to C) or D1701-VrVgD (D to F). Nonfixed infected cells (dark lines) or noninfected cells (bright lines) were stained 24 h after infection with the antigen-specific antibodies. The diagrams show the number of counted cells exhibiting specific fluorescence intensities. FIG. 4. : Single-step growth curve of D1701-VrVgC. Single-step growth curve of D1701-VrVgC. Vero cells (solid triangles), 3T3 cells (open squares), or L929 cells (solid squares) were infected with a MOI of 10 (A and B) or 1.0 (C and D) and harvested at the indicated times after infection. Cell lysates (A and C) and supernatants (B and D) were separately titrated on Vero cells in triplicate to determine infectious virus progeny. Bars indicate standard deviations. FIG. 5. : PRV-specific serum antibody response in immunized mice. PRV-specific serum antibody response in immunized mice. Sera from immunized BALB/c (A), 129/Sv/Ev (B), and C57BL/6 (C) mice were taken 2 weeks after the final vaccination. PRV-specific IgG1 (shaded columns) and IgG2a (white columns) antibodies were determined by subclass-specific ELISA. The mice were immunized as indicated with D1701-VrVgC (D-gC) and D1701-VrVgD (D-gD) alone or in combination as well as with the PRV live vaccine Begonia. The ratios of IgG2a to IgG1 subclasses are given above the columns. Bars indicate standard deviations. (D) PRV-neutralizing antibody titers of the different sera. Note the different titer scale from those of panels A to C. FIG. 6. : PRV challenge infection of mice after intravenous transfer of recombinant ORFV-immune sera. PRV challenge infection of mice after intravenous transfer of recombinant ORFV-immune sera. Sera were obtained from BALB/c mice after twofold immunization with each ORFV recombinant alone and passively transferred to naive recipients. The survival times of individual animals are depicted after transfer of the indicated volumes of D1701-VrVgC immune sera (A) or D1701-VrVgD immune sera (B). Mice receiving serum from twofold D1701-VrV-immunized animals served as a control. Challenge infection with 300 LD50s of PRV was performed 24 h after serum transfer. TABLE 1 : Protection of different immunized mouse strains against lethal PRV challenge infection TABLE 2 : ORFV recombinant-mediated protection of immune-deficient mice Backmatter: PMID- 12915542 TI - Local Character of Readthrough Activation in Adenovirus Type 5 Early Region 1 Transcription Control AB - Wild-type early activity of the adenovirus 5 E1b gene promoter requires readthrough transcription originating from the adjacent upstream E1a gene. This unusual mode of viral transcription activation was identified by genetic manipulation of the mouse betamaj-globin gene transcription termination sequence (GGT) inserted into the E1a gene. To facilitate further study of the mechanism of readthrough activation, the activities of GGT and a composite termination sequence CT were tested in recombinant adenoviruses containing luciferase reporters driven by the E1b promoter. There was a strict correlation between readthrough and substantial downstream gene expression, indicating that interference with downstream transcription was not a unique property of GGT. Blockage of readthrough transcription of E1a had no apparent effect on early expression of the major late promoter, the next active promoter downstream of E1b. A test for epistatic interaction between termination sequence insertions and E1a enhancer mutations suggested that readthrough activation and E1a enhancer activation of the E1b promoter are mechanistically distinct. In addition, substitution of the human cytomegalovirus major immediate-early promoter for the E1b promoter suppressed the requirement for readthrough. These results suggest that readthrough activation is a "local" effect of a direct interaction between the invading transcription elongation complex and the E1b promoter. DNase I hypersensitivity footprinting provided evidence that this interaction altered an extensive E1b promoter DNA-protein complex that was assembled in the absence of readthrough transcription. Keywords: Introduction : In cells infected with adenovirus type 5, transcription from the E1a promoter does not terminate upstream of the adjacent E1b gene promoter. Instead, the E1a transcripts usually are elongated through the E1b promoter and invade the E1b gene coding region . Artificial termination of this readthrough transcription by insertion of the mouse beta-major globin terminator (GGT) dramatically reduces early, but not late, E1b gene expression . The simplest interpretation of this result is that readthrough transcription is required for wild-type (wt) activity of the E1b promoter early after infection. Strong evidence for this unusual mode of gene regulation was provided by the observation that base substitution mutations in GGT that specifically relieve transcription termination restore early E1b promoter activity in cis . This result indicated that the inhibition caused by GGT arises solely from the transcription termination properties of the sequence insertion. This stimulation of early E1b transcription was designated readthrough activation. Initial attempts to map the target of activation to particular E1b promoter elements by assaying for epistasis between insertion of GGT and promoter mutations provided no evidence for a particular target sequence in the promoter . These findings raised the possibility that readthrough transcription activates the E1b promoter by a mechanism other than by direct recruitment of sequence-specific DNA binding proteins to the promoter. One line of evidence suggested that a global property of the early template could be involved. Despite the fact that the relief of inhibition by GGT occurs after the onset of viral DNA replication, the requirement for readthrough is not complemented in trans when late-infected cells are superinfected with templates whose replication is prevented . This result suggests that the readthrough requirement is a cis-dominant property of the early template. Among the possibilities are transcription-driven template remodeling, an activity of the E1a enhancer that mediates cis-acting global stimulation of early viral gene transcription, or a direct interaction between the invading transcription elongation complex and a prereplicative E1b promoter complex. We report here the results of experiments that were designed to investigate further the nature of readthrough activation. To validate a stringent connection between readthrough transcription and activation of the early E1b promoter, we tested whether a different transcription termination sequence also inhibited early E1b transcription. A second set of experiments examined the relationship between readthrough and certain global aspects of early viral transcription control. Several lines of evidence implicated a "local" rather than a global mechanism for readthrough activation of early E1b gene expression. Consistent with this notion, preventing readthrough produced a slightly altered pattern of DNase I hypersensitivity of the E1b promoter in early-infected cells, suggesting that the interaction affected the structure or composition of the promoter-protein complexes. MATERIALS AND METHODS : Cells and viruses. | Monolayers of HeLa cells or 293 cells were maintained as described previously . Adenovirus stocks were prepared from infected monolayers of 293 cells as cell lysates or purified by two rounds of density gradient centrifugation . Virus stocks were quantified by plaque titration on monolayer cultures of 293 cells . The origins of wt adenovirus type 5, dl309 , dl2004 , and recombinant strains in2011, sub2040, sub2041, sub2046, and sub2047 have been described previously. The sub2040 series strains contain the dl343 E1a chain-termination mutation , a chloramphenicol acetyltransferase (CAT) reporter gene in place of the E1b coding region, and, except for sub2046, a termination sequence insertion. Recombinant DNA. | A double-stranded oligonucleotide containing the AT-rich transcription termination sequence from the human gastrin gene (GaT) and four unpaired bases at each end for ligation to an NcoI site was produced by annealing the synthetic oligonucleotides 5'-CATGGTTTTTTTTTAATTTTTATTTTATTTTATTTTT-3' and 5'-CATGAAAAATAAAATAAAATAAAAATTAAAAAAAAAC-3' (Hershey Medical Center Core Facility). The double-stranded oligonucleotide was inserted into the SmaI site of pBLSK+ (Stratagene) in either orientation to produce pBLGT. A 313-bp BstYI DNA fragment containing the wt (pA) or mutated (pA") form of the 5' end of the mouse betamaj-globin gene transcription termination sequence was inserted into the BamHI site of pBLGT, and the clones were screened for orientation. Plasmids with the BstYI DNA fragment upstream of the AT-rich sequence were designated pCT (composite terminator) and pCTdpm (double point mutations [see reference ]), and the inserts were designated pAGaT and pA"GaT, respectively. Plasmids for strains containing a luciferase reporter gene were constructed from cloning vector pACCMVpLpA (obtained from Alex F. Y. Chen, Mayo Clinic ). This plasmid contains an expression cassette with adenovirus nucleotides (nt) 1 to 454, the major immediate-early promoter (MIEP) of human cytomegalovirus, a polylinker cloning site, the simian virus 40 poly(A) site, and adenovirus nt 3334 to 6231. The NotI site at the junction of the simian virus 40 poly(A) sequence and adenovirus nt 3334 was inactivated by partial digestion with NotI, repair of the recessed 3' ends with DNA polymerase I (Klenow), and religation. The second NotI site is located at the junction of adenovirus nt 454 and the 5' end of the MIEP. An NcoI-XbaI DNA fragment containing the luciferase (Luc) gene from pGL3 (Promega) was inserted into the HindIII site in the polylinker of the modified version of pACCMVpLpA to produce pACMIEPLuc. NcoI-XbaI DNA fragments containing pAGaT or pA"GaT were excised from pCT and pCTdpm, respectively, and inserted into the unique NotI site of pACMIEPLuc. After screening for the proper orientation of the inserted DNA, plasmids were identified that contained adenovirus nt 1 to 454, a copy of the wt or mutated composite termination sequence, and an MIEP-Luc expression unit. Although this configuration removes the E1a TATA box and major transcription start site, upstream initiation is expected to supply about 20% of the normal rate of E1a promoter activity in infected cells . Accordingly, these arrangements of viral DNA and transcription termination sequences were designated exon 1 insertions. These two plasmids, pACCTMIEPLuc and pACCTdpmMIEPLuc, were the source of virus strains sub2020 and sub2021. The MIEP in pACCTMIEPLuc and pACCTdpmMIEPLuc was excised by digestion with ClaI and SalI and replaced by inserting a 377-bp XbaI DNA fragment from p362CAT that contains the E1b promoter, including about 360 bp 5' of the transcription initiation site. The resulting plasmids, pACCTE1bLuc and pACCTdpmE1bLuc, which contain an E1b-Luc expression unit, were the source of virus strains sub2022 and sub2023, respectively. The plasmid pE1a-dl343, obtained from P. Hearing, has been described previously. It contains the dl343 E1a allele, which results in premature termination of translation . p343/112 was constructed by substituting adenovirus nt 1 to 1009 (SmaI site) from plasmid pE1a-dl343 for the corresponding region in plasmid pXC101 , which includes the dl112 E1b allele . The region containing the termination sequence copy, MIEP-Luc expression unit, and 3' adenovirus DNA sequences was excised with BstXI from pACCTMIEPLuc or pACCTdpmMIEPLuc and joined to a XbaI-ClaI DNA fragment from p343/112 that contains adenovirus nt 1 to 1338 and vector sequences. The resulting plasmids contained the termination sequence inserted in E1a exon 2 (position 1338). These two plasmids, pAC343CTMIEPLuc and pAC343CTdpmMIEPLuc, were the source of virus strains sub2026 and sub2027, respectively. A similar strategy was used to isolate plasmids pAC343CTE1bLuc and pAC343CTdpmE1bLuc, the sources of sub2024 and sub2025, respectively. Plasmid pAC343E1bLuc, the source of sub2100, was constructed by exchanging an EcoRI/BstXI DNA fragment from pACCTE1bLuc, which contains the E1b-Luc gene and the downstream adenovirus sequences for the corresponding E1b promoter and gene sequences as an XbaI/SphI DNA fragment of p343/101. The resulting plasmid contained the dl343 E1a allele and the E1b-Luc expression unit. To construct plasmids with exon 2 insertions of GGT, the region of pXC1/t-CAT or pXC1/t(dpm)-CAT containing the GGT copy and some flanking adenovirus sequences was excised with ClaI and PshAI and substituted for the corresponding segment of pAC343CTE1bLuc. The resulting plasmids, pAC343GGTE1bLuc and pAC343GGTdpmE1bLuc, were the sources of sub2028 and sub2029, respectively. A similar strategy was used to construct pAC343GGTMIEPLuc and pAC343GGTdpmMIEPLuc, the sources of sub2030 and sub2031, respectively, except that the replacement in pAC343CTMIEPLuc was accomplished by a ClaI/ClaI fragment exchange. The plasmid pterm112 has been described elsewhere . The E1a enhancer region was removed by excising a BsrGI-SacII fragment containing adenovirus type 5 nt 196 to 357 from pterm112 to produce pterm112-2, the source of in2011-2. A similar strategy was used to produce plasmid p343/112-2, the source of dl340/112-2. Recombinant viruses. | Recombinant adenoviruses were constructed by overlap recombination of plasmids into the dl309 or dl340 genome as described previously . Structures of recombinant viral genomes were determined by restriction enzyme analysis of DNA samples enriched for viral DNA by a modified Hirt procedure . Viral isolates were plaque purified two additional times prior to stock preparation. Recombinant adenoviruses containing luciferase reporter genes were constructed in a dl309 genetic background (E3 variant) from the plasmids described above. Recombinant strains used for the analysis of genetic interaction between transcription termination and the E1a enhancer were constructed in a dl340 background . The strain in2011, originally isolated in a dl309 background, was reconstructed in a dl340 background for the purposes of this work. RNA analysis. | HeLa cells were coinfected with 20 PFU each of wt adenovirus type 5 and a test strain per cell. Cytoplasmic RNA or nuclear RNA was harvested 4.25 h postinfection (hpi) for measurement of early gene expression. Cytoplasmic RNA or nuclear RNA was prepared as described previously , assayed by hybridization and protection from digestion by nuclease S1, and quantified as described previously . All hybridization probes were prepared by digestion with an appropriate restriction enzyme, labeling at the 5' end with T4 polynucleotide kinase and [gamma-32P]ATP, and digestion with a second restriction enzyme prior to isolating DNA fragments labeled at one 5' end. To detect MIEP-Luc RNA, a 1,003-bp probe was prepared as a ScaI (labeled end)-HindIII fragment from pACCTMIEPLuc. To detect E1b-Luc RNA, a 580-bp probe was prepared as a ScaI (labeled end)-HindIII fragment from pACCTE1bLuc. To detect wt E1b RNA, we used a 713-bp probe prepared as a KpnI (labeled end)-XbaI fragment from plasmid pXC7 . The probe for detecting E3 transcription was prepared as described previously . To detect E1a transcription downstream of the transcription termination sequences in MIEP-Luc strains (readthrough transcription), a 1,376-bp probe was isolated as an NdeI (labeled end)-ScaI fragment from pACCMVpLpA. To detect readthrough transcription from E1b-Luc strains, a 1,215-bp probe was prepared as a HpaI (labeled end)-SacII fragment of pXC1 . The preparation of probes for simultaneous detection of wt and dl112 E1b RNA, or wt and dl309 E3 RNA, was described previously . A probe for detection of unspliced major late promoter (MLP) transcripts was prepared as a HindIII (labeled end)-XhoI fragment from plasmid pHinC . Band intensities were quantified using a Molecular Dynamics PhosphorImager. After subtraction of background lane intensity, E1b- or MIEP-dependent transcription was normalized for template copy number and gel loading as described previously . Luciferase assays. | Monolayer cultures of HeLa cells were coinfected with 20 PFU of wt adenovirus type 5 or a test strain/cell. At 4 hpi, cell lysates were prepared and the luciferase activity was assayed by using the luciferase assay system with reporter lysis buffer (Promega Corp.) according to the manufacturer's protocol. Light emission was quantified on an FB12 luminometer (Zylux Corp.). All values were normalized to extract protein concentrations (DC protein assay; Bio-Rad Corp.). DNase I footprinting in isolated nuclei. | HeLa cells growing in 150-mm dishes were infected at about 60% confluence with 50 PFU of the indicated virus/cell. Only CsCl-purified virus was used for in vivo footprinting. At 4 to 5 hpi, monolayers were washed once with ice cold Tris-buffered saline (TBS) and then removed from the surface in 5 ml of cold TBS/dish. Nuclei for DNase I digestion were prepared as described by Cordingley et al. . The cells were collected by centrifugation at 800 x g for 10 min at 4C and suspended in 5 ml of homogenization buffer (0.01 M Tris-HCl [pH 7.4], 0.015 M NaCl, 0.06 M KCl, 0.15 mM spermine, 0.5 mM spermidine, 0.01 M EDTA, 0.1 mM EGTA, 0.2% NP-40, 5% sucrose, 0.001 M phenylmethylsulfonyl fluoride) per dish and kept on ice for 3 min. The cells were lysed with a few strokes of a Dounce homogenizer (15 ml) with a B pestle until few intact cells remained. The suspension was carefully layered over 3.5 ml of homogenization buffer containing 10% sucrose and no NP-40. Nuclei were collected at the bottom of the tube by centrifugation for 10 min at 1,200 x g at 4C and suspended in an equal volume of homogenization buffer, but with 0.2 mM EDTA, 0.2 mM EGTA, and no NP-40 or sucrose (wash buffer). The nuclei were collected again by centrifugation as before for 3 min and suspended in wash buffer at a concentration of about 3.5 x 107 nuclei/ml. Up to 32 U of DNase I (1,000 U/ml; Promega)/ml was added to 0.5-ml aliquots of nuclei, and digestion was initiated by adding 5 mul of 0.2 M CaCl2, 0.4 M MgCl2. Digestion was for 15 min at 4C and was terminated by adding 0.05 ml of 1 M Tris-HCl (pH 8.5), 0.1 M EDTA. Digested nuclei were stored at -80C. For DNA extraction , the nuclei were thawed and diluted with 5.25 ml of 6 M guanidine-HCl, 0.1 M sodium acetate, pH 5.5. The solution was rocked gently at room temperature for 1 h, and the solution was carefully layered under 12.6 ml of absolute ethanol in a 30-ml Corex tube. The solution was mixed gently by inversion, and the precipitate was collected by centrifugation. The pellet was washed two times with 5 ml of absolute ethanol, dried briefly, and suspended in 0.9 ml of 0.1x TE (1x TE is 0.01 M Tris-HCl [pH 8.0], 0.001 M EDTA). After at least 16 h at 4C, the suspension was transferred to a clean centrifuge tube and digested for 1 h at 37C with 100 U of HindIII/ml to reduce the viscosity. Nucleic acid was extracted three times with an equal volume of TE-saturated phenol and once with chloroform-isoamyl alcohol (24:1), precipitated with ethanol, suspended in water, and stored at -20C. Naked DNA controls were prepared by digesting purified DNA from infected cells with 0.2 U of DNase I/ml for 15 min at 4C. The DNA was extracted with phenol and chloroform, precipitated, and suspended in water for cycled primer extension. A 50-pmol aliquot of primer was 5'-end labeled by adding 50 pmol of [gamma- 32P]ATP (6,000 Ci/mmol) and 25 U of T4 polynucleotide kinase in 0.05 ml. The labeled primer was isolated by electrophoresis in a denaturing 20% polyacrylamide gel, or a 3% Biogel (Qbiogene, Inc.) and then purified by using MerMaid spin column (Qbiogene, Inc.) chromatography according to the manufacturer's protocols. Twenty cycles of primer extension were performed using 25 U of Taq DNA polymerase (Fisher Scientific)/ml in buffer B with 2 mM MgCl2, 0.2 mM deoxynucleoside triphosphates, approximately 8 x 106 dpm (about 0.5 pmol) of labeled primer, and DNA from about 7 x 106 nuclei. Extension products were extracted once with chloroform-isoamyl alcohol, precipitated, and suspended in loading buffer containing 80% formamide for 6-to-8% denaturing polyacrylamide gel electrophoresis. Dried gels were exposed to Kodak XAR-5 film for autoradiography. Plots of the scans of autoradiograms were generated using NIH Image, version 1.62. Primers. | The r-strand primers were as follows: for detection of wt E1b genomes, adenovirus type 5 nt 1817 to 1794, 5'-GATGAGCCCCACAGAAACTCCAA-3'; and for detection of genomes containing E1b-CAT fusion genes, 5'-GGCCGTAATATCCAGCTGAACG-3', from a position 77 nt further 3' with respect to the E1b promoter. The l-strand primers were as follows: Ad 5 nt 1325 to 1350, 5'-CGACATCACCTGTGTCTAGAGAATGC-3'; and for detection of genomes containing GGT, 5'-GCAGTAGGTAGAACCCTTG-3' in the mouse betamaj-globin sequence (nt 5403 to 5421), from a position 48 nt further 5' with respect to the E1b promoter. RESULTS : Activity of a composite termination sequence. The mouse betamaj-globin transcription termination sequence (GGT) inhibits readthrough transcription and dramatically reduces early E1b RNA synthesis when it is inserted into exon 2 of the E1a coding region . For GGT, inactivation of the poly(A) addition signals, which are required for termination of transcription in exon 2 , relieves the inhibition of early E1b gene transcription. Therefore, the ability of GGT to terminate upstream transcription is necessary for inhibition of downstream gene transcription. However, the GGT element is 1,560 bp in length, and some other intrinsic property of the sequence might also be necessary, but not sufficient, for inhibition of downstream promoter function by the exon 2 insertions. In fact, E1a exon 1 insertions of either wt or mutated GGT prevent readthrough and inhibit downstream gene transcription (; L. F. Maxfield and D. J. Spector, unpublished data). Since transcription termination of GGT is expected to be poly(A) site dependent , another activity of GGT aside from termination of transcription also might affect downstream promoter activity. To address these issues, we designed another candidate termination sequence. pAGaT was constructed by deleting all but the 5' 350 bp of GGT, essentially retaining only the 3' part of the globin gene exon 3 that includes the poly(A) sites (pA) , and joining it to a 35-bp transcription termination sequence from the human gastrin gene (GaT) . The gastrin element by itself was incapable of terminating transcription in the context of the virus (D. J. Spector, unpublished data). From here on, the hybrid element will be referred to as CT, or composite terminator. To test both the termination properties of CT and its possible effect on downstream gene transcription, a series of four adenovirus strains was constructed. Two strains contained a copy of CT in the first (sub2022) or second (sub2024) exon of E1a and a luciferase reporter gene in place of the E1b coding sequences (E1b-Luc) . Two matching strains (sub2023 and sub2025) were constructed with base substitution mutations in the poly(A) signal of the pA segment of CT (mCT). These mutations relieve termination by GGT . To compare the activity of CT with that of GGT, two more strains (sub2028 and sub2029) were constructed with GGT or inactivated GGT inserted in the second exon of E1a and a luciferase gene in place of the E1b coding region. Finally, control strain sub2100 contained the E1b-Luc gene but no termination sequence insertion . A nonsense mutation in the E1a coding region of each strain eliminated the possibility that functional E1a proteins might be made from some strains but not others. The dose of E1a was controlled in each infection by complementation (see below). E1a activation of the E1b promoter was obtained in the presence or absence of readthrough (data not shown). Downstream gene expression was assayed by two methods. First, luciferase activity was determined after infection of HeLa cells or 293 cells with the different recombinant viruses. In HeLa cells, E1a and E1b proteins were provided in trans by coinfection with wt virus. E1a and E1b proteins are constitutively expressed in 293 cells . Accordingly, any defect in E1b-Luc expression conferred by the CT or GGT insertion resulted from a cis-acting mechanism. Second, steady-state levels of cytoplasmic E1b-Luc RNA in infected HeLa cells were measured by hybridization and protection from S1 nuclease digestion. Again, E1a and E1b proteins were provided in trans by coinfection with wt virus. Early RNA from the adenovirus E3 gene was used as an internal control because early E3 transcription is not affected by termination sequence insertions in E1 . By either assay, insertion of CT or GGT in exon 2 of the E1a gene reduced E1b expression in a poly(A) site-dependent manner . The inhibition of E1b expression by CT was not as efficient as that by GGT. When CT was inserted into the first exon of the E1a gene, E1b gene expression was reduced more dramatically than in the strains with exon 2 insertions. The luciferase expression was near background, and mRNA levels were indistinguishable from the background. In this location, inhibition of E1b gene expression by the exon 1 insertion was poly(A) site independent, a result also obtained with GGT (data not shown). To determine whether inhibition of downstream gene expression was associated with a block in readthrough transcription in these strains, steady-state levels of readthrough E1a RNA were measured by hybridization-nuclease protection, using a probe from the E1b promoter region. This probe also detected abundant wt E1a mRNAs from the complementing strain. Readthrough transcripts were detected from the control strain sub2100 and from strains sub2025 and sub2029 with exon 2 insertions of termination sequences with inactivated poly(A) sites . However, readthrough transcripts did not accumulate to detectable levels from the exon 1 insertion strains, sub2022 and sub2023, or from strains sub2024 and sub2028 with wt termination sequence insertions in exon 2. Lack of readthrough transcripts always was accompanied by reduced expression of the E1b-Luc gene. As well, detectable readthrough transcription always was associated with higher levels of E1b-Luc expression. These data show that CT had transcription termination and downstream gene-inhibition activities similar to those of GGT, although CT was probably a less potent terminator in the virus. Therefore, the ability to inhibit downstream promoter function was not unique to GGT. Rather, the data strongly support the hypothesis that the ability of sequence elements to prevent readthrough transcription from E1a is both necessary and sufficient for inhibition of downstream E1b gene expression. Effect of readthrough transcription on early activity of the MLP. | If readthrough activation originates from a global property of the early viral template, such as transcription-driven remodeling, one might expect that the early function of other viral promoters would be affected. Insertion of GGT in E1a, which is in the left half of the genome, does not affect the transcription of the E3 promoter in the right half of the genome . However, leftward readthrough transcription from the strong E4 promoter at the right end of the genome could conceivably produce any global template changes required for activation of the E2 and E3 transcription units. Any distal effects of transcriptional readthrough from E1a might only be manifest in the left half of the genome. The MLP, whose early activity is comparable to that of the other early promoters , is the next promoter downstream of E1b that is active early. Accordingly, an experiment was performed to determine whether the activity of the MLP was affected by GGT insertions in E1a. Strain in2011 has an insertion of GGT in exon 1 of E1a, and sub2040 has an exon 2 insertion. Both of these strains fail to accumulate readthrough transcripts and produce dramatically reduced amounts of early E1b RNA . sub2047 is identical to sub2040, except that the former has an inactivated GGT sequence . dl2004 and sub2046 are the parental strains of in2011 and sub2040, respectively, without the GGT insertions . All of these strains have defects in the E1a and E1b coding regions, and so infections were performed in 293 cells to provide the early E1 proteins in trans. The results indicate that the ratio of MLP to E3 RNA was the same for each strain . Therefore, the effect of transcription termination in E1a did not extend to the MLP, the next promoter downstream of E1b. The relationship between transcription termination and the activity of the E1a enhancer on E1b early gene expression. | The E1a enhancer mediates two kinds of activities. One regulates E1a gene expression, whereas the second activates the transcription of the other early viral genes in cis . One possible role for readthrough activation is to facilitate a functional interaction between the E1a enhancer and the E1b promoter. If so, then mutations that interfere with both readthrough and enhancer function might be expected to exhibit epistasis, that is, the reduction in expression caused by one mutation should mask any additional phenotype from the second in double mutant strains. To evaluate this possibility, early gene expression was assayed from three recombinant viruses. Strain dl340/112-2 (Enh- RT+) has a deletion in the enhancer-encapsidation region of the genome (nt positions 195 to 357) that removes both the E1a-specific enhancer components (element I) and the enhancer that governs all early viral transcription in cis (element II) . Strain in2011 (Enh+ RT-) has a GGT insertion that prevents readthrough. Strain in2011-2 (Enh- RT-) has both mutations. The control strain dl343-112 (RT+ Enh+) has the same E1b and E3 genetic backgrounds as the other strains. HeLa cells were coinfected with wt adenovirus type 5 and a recombinant virus, and early E1b and E3 RNA were assayed at 4.25 hpi. Single infections were performed with dl2004 and wt adenovirus type 5 to provide standards for the RNAs made from the test strain and complementing strain, respectively. After subtraction of the background (E1b wt), each mutation alone resulted in about a 10-fold decrease in E1b RNA in cis, whereas the double mutation produced a 50-fold decrease . In a second experiment, each single mutant was down 15- to 20-fold and the double mutant was down about 300-fold. As expected , enhancer mutants also made less E3 RNA. These data provided no evidence for an epistatic interaction between the two kinds of mutations. Therefore, transcription termination probably did not affect substantially any functional interaction between the E1a enhancer and the E1b promoter. Furthermore, as observed in a similar analysis of double mutants with defects in readthrough and the E1b promoter , the data show that readthrough effectively activated a highly weakened E1b promoter (compare Enh- RT- with Enh- RT+). Suppression of the requirement for readthrough transcription by a strong downstream enhancer-promoter. | The results of the experiments above are consistent with the notion that activation requires only an upstream source of readthrough transcription and the downstream E1b promoter. That is, the mechanism of action depends only on an interaction between the invading RNA polymerase elongation complex and the E1b promoter. One might expect such a localized interaction to stimulate the activity of weak promoters but to be unnecessary for strong downstream promoters. If so, then the requirement should be relieved, or suppressed, by substituting a strong promoter for E1b. Accordingly, the E1b promoter was replaced with the human cytomegalovirus MIEP in recombinant adenoviruses with various termination sequence insertions. Strains sub2020 and sub2026 have CT insertions in the first and second exons of E1a, respectively. sub2021 and sub2027 are versions of these two strains with inactivated CT (mCT). sub2030 and sub2031 have the wt and inactivated forms of GGT, respectively, inserted into the second exon of E1a. MIEP-Luc gene expression was assayed either by measuring luciferase activity or by determining steady-state levels of cytoplasmic MIEP-Luc RNA in HeLa cells coinfected with wt virus as described above. The data indicate that MIEP mediated a high level of luciferase expression or RNA synthesis regardless of whether the termination element was wt or mutant. Assays for readthrough RNA yielded the expected result: readthrough was observed only when an exon 2 termination element was inactivated . Because sub2030 and sub2031 grew poorly and produced low-titer stocks, a relatively large volume of lysate was required to achieve the desired multiplicity of infection. Under these conditions, a general inhibition of viral gene expression was observed, as reflected in the low levels of viral RNA in the corresponding samples. These results show that readthrough and downstream gene expression were completely uncoupled when the E1b promoter was replaced by MIEP. That is, MIEP functioned at a high level regardless of whether readthrough was present or absent. Therefore, substitution of the relatively weak E1b promoter with a strong downstream promoter relieved or suppressed the requirement for readthrough activation in this region of the viral genome. This result further supported the notion that readthrough acts through a local interaction at the site of the E1b promoter. Effect of readthrough on E1b promoter structure. | To examine the effect of readthrough on DNA-protein interactions at the E1b promoter in infected cells, DNase I footprinting was performed . Proteins bound to DNA can protect the DNA from DNase I digestion or distort the double helix to produce hypersensitive sites of digestion. At high multiplicities of infection and a corresponding high copy number of early templates per nucleus, only a small fraction of templates may be engaged in promoter complex formation at a given gene at any particular time . Under these conditions, regions protected from DNase I digestion may be difficult to observe against a high background of viral DNA molecules with a different structure. However, hypersensitive sites will appear as enhanced signals. Viral infections at low multiplicities failed to produce detectable signals after DNase I digestion of isolated nuclei, so high-multiplicity conditions were used for the analysis presented here. Hypersensitivity of positions in viral DNA digested in infected cell nuclei was scored as the reproducible appearance of bands that were not observed either with undigested nuclei or with DNA purified from infected cell nuclei before digestion. A very conservative approach was adopted. A band was considered indicative of hypersensitivity only if the relative difference from the surrounding background in the same lane was reproducibly greater than the relative difference from the surrounding background in the digested naked DNA lane. DNase I digestion was performed with nuclei infected either with wt virus (dl309) or strain sub2047, which has a mutated GGT and thus allows readthrough , or nuclei were coinfected with both strains and analyzed under conditions that permitted distinct assay of the different templates. Identical digestion profiles of the E1b promoter region were obtained under all of these conditions. An example from the sub2047 genomes in coinfected cells is shown in Fig. . For the r-strand , the strongest hypersensitive sites mapped approximately at positions 1525 (band B), 1550 (band C), and 1605 (band D). Surprisingly, the region between bands B and C showed evidence of protection from digestion, which suggests a particularly high level of occupancy of this region. Band A at about position 1460 and band E at position 1695 were modestly hypersensitive. For the l-strand , three stronger hypersensitive sites were obtained near positions 1505 (band I), 1545 (band G), and 1675 (band F), and three more modestly hypersensitive sites mapped near positions 1440 (band K), 1465 (band J), and 1520 (band H). With one exception (band I), these sites mapped at or near the boundaries of regions protected by KB cell nuclear extract proteins from DNase I digestion in vitro . Note that the footprinting reported here was performed in HeLa cells. In vitro site III, which contains a consensus sequence for binding of transcription factor Ap2, was most obviously delineated by both protection and hypersensitive boundaries in isolated nuclei. To determine the effect of preventing readthrough transcription on the DNase I hypersensitivity profile, footprinting was performed on nuclei from HeLa cells infected with sub2040 , alone or in combination with dl309. The same results were obtained in either case, and the assay of sub2040 templates in coinfected cells is shown in Fig. . Most of the hypersensitive sites, as well as the protected region, obtained from readthrough-containing strains were observed. However, hypersensitive site D did not appear in the absence of readthrough. These data suggest that an extensive E1b promoter DNA-protein complex was assembled in the absence of readthrough transcription but that readthrough altered the structure of the complex. DISCUSSION : The analysis of recombinant adenovirus strains reported here reveals several important properties of readthrough activation in the E1 region of adenovirus type 5. First, the capacity to inhibit downstream gene expression by preventing readthrough transcription was not restricted to the transcription termination sequence, GGT. Rather, a different composite termination sequence, CT, was competent to inhibit downstream E1b promoter activity. Accordingly, it is likely that any means that prevent readthrough into the E1b promoter region are sufficient for inhibition of E1b promoter activity. Second, the mechanism of readthrough activation appears to be local. There was no evidence for a direct relationship between the readthrough requirement and the distal E1a enhancer, which has a global effect on expression of early genes, nor was there any effect on the early expression of the next promoter downstream, the MLP. Furthermore, simply replacing the downstream target promoter with a strong promoter relieved the requirement for readthrough in early-infected cells. Finally, readthrough was associated with a change in the structure of the E1b promoter complex. Activity of the composite termination sequence CT. | The evidence that CT terminated transcription is indirect and relies on three observations. The first was the lack of accumulation of steady-state transcripts downstream of the element. Steady-state measurements in themselves do not rule out rapid degradation of a short-lived readthrough precursor. The second observation was the restoration of readthrough transcripts by inactivation of the poly(A) signals. The dependence of transcription termination by RNA polymerase II on poly(A) signals is well documented . Third, there was a complete correlation between inhibition of E1b promoter activity and both the insertion of wt CT and the lack of the accumulation of readthrough transcripts. This finding was consistent with the relationship between these events established by the use of GGT, a known transcription terminator . The composite termination sequence was assembled from two components, the region of GGT that contains the poly(A) addition sites and an AT-rich sequence from the human gastrin gene . Current evidence suggests that poly(A) sites can combine with downstream regions that promote both pausing of RNA polymerase and destabilization of the RNA-DNA hybrid to terminate transcription . The termination region of GGT is about 500 bp downstream of the poly(A) signals and is very AT rich . For termination by GGT, the distance between the two functional elements is essential for efficient termination . In the case of CT, the AT-rich sequence from the gastrin gene may be particularly potent, so as to reduce the requirement for additional distance. On the other hand, the fact that CT did not inhibit E1b transcription as strongly as GGT did suggests that the former element was not as efficient a termination sequence. Additional spacing between its two functional components might increase the activity of CT. Since CT is only about 370 bp, an insertion that increases its activity might result in an element that is as potent as GGT but still much shorter and, therefore, more versatile as a gene control element. Another possibility is that the difference in activity resides in the difference between the AT-rich sequences of the two elements. When either GGT or CT was inserted into exon 1 of E1a, the loss of steady-state readthrough transcripts and the inhibition of downstream gene expression became poly(A) site independent. A large, unspliced exon initiated in E1a and extending through the terminator to the poly(A) sites in the E1b promoter could be very unstable and fail to accumulate. However, the lack of readthrough activation suggests that the readthrough transcripts were not made at all. A more likely explanation may be related to the processivity of elongating RNA polymerase. In recombinant strains with exon 1 insertions, a copy of the wt or mutated termination sequence was placed at adenovirus nt 454, followed by a MIEP-Luc or E1b-Luc expression unit. This configuration removed the E1a TATA box, major transcription start site (position 499), and the promoter proximal splice site, most likely resulting in E1a transcription initiation from upstream initiation sites . It has been reported that removal of promoter proximal splice signals from mammalian genes results in a marked reduction in levels of transcription . The elongation complex recruits processivity factors as it transcribes the DNA, and a major component of elongation factors are mRNA processing factors . If the recruitment of any of these factors by the elongation complex requires assembly of splice junctions , then encountering an AT-rich region prematurely, even in the absence of a poly(A) site, could result in the interaction between a polymerase with reduced processivity and the AT-rich element, resulting in disengagement of the polymerase from the template. The local character of readthrough activation: the effectors and targets of readthrough activation. | The fact that some cis-acting property of early transcription templates is required for readthrough activation suggested that the mechanism might involve a global process that affects the establishment of early transcription. Transcription-driven template remodeling or enhancer activity at a distance was among the possibilities. The results here argue against these kinds of general effects. Readthrough had no effect on the MLP, the nearest active early promoter downstream of E1b and whose activity is comparable to that of E1b during the early phase of infection. One possibility is that readthrough activation decreases with distance, a notion supported by the reduced activity of the E1b promoter in strains with GGT insertions compared to strains with CT insertions . Alternatively, MLP function could be completely unresponsive to the readthrough mechanism. In fact, a reported candidate termination element in the MLP could reduce the level of early readthrough transcription independently. Evidence that viral enhancer function and readthrough activation were genetically separable activities argues against the involvement of enhancer activity at a distance in readthrough activation. Residual enhancer activity and/or readthrough could have contributed to the apparent lack of an epistatic interaction between the mutations. If so, there still might be a direct mechanistic relationship between readthrough and enhancer function. Alternatively, the enhancer could play an indirect role by increasing the amount of readthrough. However, it seems more likely that any interaction between the enhancer and the E1b promoter is direct and independent of readthrough. The ability of a strong downstream promoter to suppress the readthrough requirement suggests that the strength and/or structure of the target promoter is important in the interaction. Readthrough activation might deliver an initiation or elongation factor that compensates for the inability of the native E1b promoter to recruit the relevant factors efficiently. At the same time, readthrough could counterbalance potential interference from a nearby upstream promoter, in this case, E1a. The strength of each transcription unit and their separation could be critical factors, and the distance between them could influence the degree to which interference or activation predominate at the downstream promoter. A recent observation in our laboratory suggesting that readthrough activation occurred with plasmid transcription units in a transient expression assay (L. Shen, A. M. Rowzee, and D. J. Spector, unpublished data) would imply that the interaction is unrelated to virus life cycle processes. This finding is consistent with the results of the experiments described here, which focus attention directly on the two transcription units and point to a local mechanism for readthrough activation that involves a direct interaction between the invading transcription elongation complex and the early E1b promoter. The footprinting results support the notion that readthrough activation is local rather than global. An E1b promoter-protein complex was assembled in the absence of readthrough, and readthrough was accompanied by a change in the structure of the complex. Previous attempts to map genetically a target of activation to particular E1b promoter elements by assaying for epistasis between GGT and promoter mutations provided no evidence for a particular sequence-specific DNA binding protein target . Accordingly, the structural change may not involve the recruitment of a DNA binding protein. Rather, readthrough could confer new or different protein-protein interactions, perhaps involving one or more general transcription factors as suggested above, at the promoter. The application of chromatin immunoprecipitation assays should help address these issues as well as help determine which step in the transcription cycle of this promoter requires readthrough. FIG. 1. : Early expression of the E1b-Luc gene and readthrough transcription from recombinant adenovirus strains with transcription termination sequence insertions 5' of the E1b promoter. Early expression of the E1b-Luc gene and readthrough transcription from recombinant adenovirus strains with transcription termination sequence insertions 5' of the E1b promoter. (A) Cytoplasmic RNA was prepared at 4.25 h after infection of HeLa cells coinfected with the indicated virus and wt adenovirus type 5. Viral RNA was assayed by hybridization and protection from nuclease S1 digestion. For the E1b-Luc probe with the labeled site in the luciferase gene, protection of RNA originating in E1b should produce a 207-nt band (diagram above the autoradiogram). Nuclease-sensitive AT-rich regions in the RNA-DNA hybrids probably produced the bands migrating faster than the predicted sizes. For the sample infected with wt virus only, E1b RNA was assayed with a wt probe (E1b wt). E1b-Luc RNA levels were quantified, normalized to the sub2100 value (1.00), and plotted to generate the graph at the bottom of the panel. (B) Protection of the E3 probe by wt E3 RNA or E3 RNA from recombinant virus (dl309 background) yielded a 445-nt product for wt E3 RNA or a 190-nt product for dl309 E3 RNA as described previously . The vertical arrow in the diagram above indicates a position of sequence divergence between the probe and the RNA, and the caratindicates a splice. (C) Luciferase activity in samples prepared from cells infected with the indicated viruses was determined as described in Materials and Methods. The diagrams on the right show the genome structures in the left end region of the E1b-Luc viruses. The arrangements of the two transcription starts, termination sequence insertions, Luc gene, and the dl343 mutation (*) are indicated. mCT, mutated CT; mGGT, mutated GGT. (D) Readthrough transcription from E1b-Luc strains. Samples were assayed for readthrough transcription as described in Materials and Methods. The probe, labeled at adenovirus type 5 nt 1573, contained part of the E1b promoter and upstream adenovirus sequences. A 234-kb labeled product (RT) should be obtained from readthrough transcription in the CT insertion strains, since they did not contain adenovirus sequences 5' of 1339. The vertical arrow indicates the position of sequence divergence between the wt adenovirus probe and the RNA encoded by GGT or CT. The major band is from protection of wt E1a exon 2 transcripts (splice junction at position 1229) from the complementing strain, and the faster-migrating E1a band probably represents an S1-sensitive site in the E1a mRNA-DNA hybrid. FIG. 2. : Early activity of the adenovirus MLP from strains with GGT insertions. Early activity of the adenovirus MLP from strains with GGT insertions. Nuclear RNA was prepared from infected 293 cells at 4.5 hpi and assayed by hybridization and protection from S1 nuclease digestion using probes for the major late transcription unit (MLP) or E3. Strains in2011 (E1a exon 1 insertion) and sub2040 (E1a exon 2 insertion) contained GGT. All other strains are controls that allowed readthrough. sub2047 contained inactivated GGT. FIG. 3. : Test for epistatic interaction of readthrough and enhancer mutations that reduce early E1b transcription. Test for epistatic interaction of readthrough and enhancer mutations that reduce early E1b transcription. HeLa cells were coinfected with wt adenovirus type 5 and dl343/112 (RT+ Enh+), in2011 (RT- Enh+), dl340/112-2 (RT+ Enh-), or in2011-2 (RT- Enh-). Separate cultures were infected with either dl2004 (test), which has the same E1b and E3 genetic background as the test strains, or wt virus. Nuclear RNA prepared from samples harvested at 4.25 hpi was analyzed for E1b or E3 RNA by hybridization and nuclease protection. The arrows indicate the positions of the relevant protection products. The marker was labeled HaeIII-digested DNA fragments of pBR322. FIG. 4. : Early expression of the MIEP-Luc gene and readthrough transcription from recombinant adenovirus strains with transcription termination sequence insertions 5' of the MIEP. Early expression of the MIEP-Luc gene and readthrough transcription from recombinant adenovirus strains with transcription termination sequence insertions 5' of the MIEP. Cytoplasmic RNA was prepared at 4.25 h after infection of HeLa cells coinfected with the indicated virus and wt adenovirus type 5. Viral RNA was assayed by hybridization and protection from nuclease S1 digestion, using probes for the MIEP-Luc transcription unit (A) or E3 (B). For the MIEP-Luc probe with the labeled site in the luciferase gene, protection of RNA originating in E1b should produce a 227-nt band. Protection of the E3 probe by wt E3 RNA or E3 RNA from recombinant virus (dl309 background) yielded the same products as in Fig. . (C) Luciferase activity in samples prepared from cells infected with the indicated viruses was determined as described in Materials and Methods. The diagrams on the right show the genome structures of the MIEP-Luc viruses in the left end region. Thearrangements of the two transcription starts, termination sequence insertions, Luc gene, and the dl343 mutation (*) are indicated. (D) Readthrough transcription from MIEP-Luc strains. Samples were assayed for readthrough transcription as described in Materials and Methods. The probe, labeled in the MIEP sequence, protected a 407-nt readthrough product. The vertical arrow indicates a position of sequence divergence between the MIEP probe and the RNA encoded by GGT or CT. FIG.5. : Effect of readthrough on DNase I hypersensitivity patterns of the E1b promoter. Effect of readthrough on DNase I hypersensitivity patterns of the E1b promoter. Nuclei were prepared 4.25 h after coinfection of HeLa cells with dl309 and either sub2040 (no readthrough) or sub2047 (readthrough) and digested with 32 U of DNase I per ml (32) or mock digested (0). The DNA was extracted, and a sample of mock-digested DNA from sub2047-infected cells was digested with 1.25 U of DNase I/ml. Primer extensions were performed with an r-strand primer (A) or l-strand primer (B) specific for the sub strain. The positions of the cleavage sites in viral DNA were calculated by comparison to the migration of marker (M) DNA fragments prepared from a HaeIII digest of pBR322. Scans of the bracketed areas are shown to the left. The scan of undigested nuclei is of the 2047 (0) sample. The letters and arrows indicate positions of hypersensitivity. The approximate positions of the arrows are provided in the text. The absence of band D in the sub2040 profile is designated D*. (C) Summary of footprinting results. The E1b promoter region is denoted by nucleotide position. The transcription start site is at position 1702. The rectangles represent regions (I to V, GC and CAP) that are protected from DNase I digestion in vitro in KB cells . TA indicates the location of the TATA box. The arrows represent the DNase I hypersensitive sites mapped in isolated nuclei (top, r-strand; bottom, l-strand). Backmatter: PMID- 12915577 TI - Cell Cycle Regulation by Kaposi's Sarcoma-Associated Herpesvirus K-bZIP: Direct Interaction with Cyclin-CDK2 and Induction of G1 Growth Arrest AB - In order to cope with hostile host environments, many viruses have developed strategies to perturb the cellular machinery to suit their replication needs. Some herpesvirus genes protect cells from undergoing apoptosis to prolong the lives of infected cells, while others, such as Epstein-Barr virus Zta, slow down the G1/S transition phase to allow ample opportunity for transcription and translation of viral genes before the onset of cellular genomic replication. In this study, we investigated whether Kaposi's sarcoma-associated herpesvirus (KSHV) K-bZIP, a homologue of the Epstein-Barr virus transcription factor BZLF1 (Zta), plays a role in cell cycle regulation. Here we show that K-bZIP physically associates with cyclin-CDK2 and downmodulates its kinase activity. The association can be detected in the natural environment of KSHV-infected cells without artificial overexpression of either component. With purified protein, it can be shown that the interaction between K-bZIP and cyclin-CDK2 is direct and that K-bZIP alone is sufficient to inhibit CDK2 activity. The interacting domain of K-bZIP has been mapped to the basic region. The result of these associations is a prolonged G1 phase, accompanied by the induction of p21 and p27 in a naturally infected B-cell line. Thus, in addition to the previously described transcription and genome replication functions, a new role of K-bZIP in KSHV replication is identified in this report. Keywords: Introduction : Herpesvirus replication is often broken down into early, delayed-early, late, and latent phases. The cellular environment required for each phase is unique, and a number of viral genes have evolved to interact with host factors to create a more favorable environment for viral replication. Regulation or deregulation of the cell cycle is one strategy often used by viruses to alter the cellular environment. In the literature, there is a preponderance of evidence showing that viral gene products interact with cell cycle proteins such as cyclins and CDKs or their regulators such as p53 and retinoblastoma protein . Some of these interactions lead to accelerated cell cycle progression and cellular proliferation, in some cases contributing to the transformation of target cells. Other herpesvirus gene products delay cell cycle progression and transiently arrest the infected cells at the G1 stage. This usually occurs among the first stages of the viral life cycle and involves immediate-early and/or early gene products , such as ICP0 and ICP27 of herpes simplex virus, IE2 and UL69 of cytomegalovirus, and Zta of Epstein-Barr virus (EBV). It was postulated that lengthening the G1 phase would give additional time for the virus to complete transcription and translation of early genes and for viral DNA replication to occur before the onset of competing cellular genomic replication (see reference for a review). The molecular mechanisms by which viral gene products interfere with G1 to S progression differ from virus to virus. Both EBV Zta and cytomegalovirus IE2 are known to bind p53 and stabilize it . On the other hand, cytomegalovirus IE2 also binds retinoblastoma protein and releases E2F, which transcriptionally activates cyclin E gene expression . By these means, cytomegalovirus blocks cellular DNA replication while at the same time it activates certain cell cycle pathways to provide a more suitable environment for viral DNA replication. The overall consequence of these interactions is to slow down the G1 to S transition . Kaposi's sarcoma-associated herpesvirus (KSHV), also known as human herpesvirus 8, is a member of the gammaherpesvirus family, which includes EBV and herpesvirus saimiri. KSHV infection is associated with all types of Kaposi's sarcoma, including AIDS-associated, endemic, and renal transplant-related Kaposi's sarcoma . It has also been implicated in B-cell lymphoproliferative diseases such as primary effusion lymphoma and multicentric Castleman's disease . Like other herpesviruses, KSHV encodes both latent and lytic genes . The latent genes are thought to be primarily responsible for the maintenance of latency and directly involved in cell transformation . The lytic genes are involved either directly in viral replication (genome replication, transcription, etc.) or in providing a cellular environment conducive for viral replication (e.g., B-cell activation, immune modulation of host response, target cell recruitment) or both . Products from both latent and lytic genes have been shown to interact with cell cycle proteins or their regulators. Perhaps the best known is v-cyclin, which directly interacts with and activates cellular CDK6, thereby accelerating cell cycle progression . Other viral gene products such as latency-associated nuclear antigen (LANA) interact with the cell cycle regulator p53 and thus can potentially perturb both the cell cycle and apoptosis programs . Previously, we reported the identification of an early KSHV gene product, K-bZIP, which is the positional and structural homolog of EBV Zta . We show here that K-bZIP interacts directly with cyclin-CDK2 and downmodulates the kinase activity of CDK2, resulting in G1 growth arrest of the host cells. The association of K-bZIP and CDK2 can be detected in the naturally infected BCBL-1 line, and the expression of K-bZIP but not of a mutant lacking the interaction domain results in slower growth of cells. This growth arrest is accompanied by the increased expression of p21 and p27. Our results suggest that K-bZIP, in addition to being a transcriptional regulator, is also directly involved in cell cycle regulation. MATERIALS AND METHODS : Plasmids. | Plasmids encoding the full-length K-bZIP (K-bZIP/wt, residues 1 to 273) were constructed in pcDNA3.1 (Invitrogen). Cloning introduced a CpoI site and either a hemagglutinin (HA) tag, a Flag tag, or a T7 tag to the N terminus as described previously . The resulting plasmids were denoted pHA-K-bZIP, pFlag-K-bZIP, and pT7-K-bZIP, respectively. A natural spliced variant of K-bZIP, K-bZIPDeltaLZ, which carries a deletion of the leucine zipper region, was inserted in-frame 3' of the Flag tag sequence of modified pcDNA3.1 as described previously . The resulting plasmid was denoted pFlag-K-bZIPDeltaLZ. An expression plasmid with a deletion of the K-bZIP basic region (BR) (residues 123 to 189) (pFlag-K-bZIPDeltaBR) was constructed as described previously . Deletion fragments of K-bZIP were amplified by PCR with Pfu Turbo (Stratagene) with primers as described previously and then cloned into a modified pGEX4T-2 (Amersham-Pharmacia), which introduced a CpoI site to produce glutathione S-transferase (GST) fusion proteins. To prepare additional K-bZIP deletion mutants in this study, K-bZIP deletion fragments were amplified with the following primer pairs (italics designate CpoI sites and capital letters denote the viral sequence): K-bZIP 107F (5'-aac ggt ccg CTC TCT CAC ACA CCA CCA AGA-3') and K-bZIP 157R (5'-aac gga ccg CTT AAC TAC AGA CGC AGG CAC-3'), and K-bZIP 158F (5'-aac ggt ccg GCC GAA GTA TGT GAT CAG TC-3') and K-bZIP 208R (5'-aac gga ccg CTT ATG TGC CTC CAA TCT CGC-3'). Amplified fragments were then digested with CpoI (Fermentas) and inserted in frame downstream of the GST coding sequence. Expression plasmids for CDK2, cyclin A, and cyclin E were a generous gift from D. J. Templeton (Case Western Reserve University). Cell culture, transient transfection, establishment of K-bZIP stable transfectants, and growth rate analysis. | Epithelial human embryonic kidney cell lines 293 and 293T, as well as a cervical carcinoma cell line, HeLa cells, were grown in monolayer cultures in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum in the presence of 5% CO2. To establish stable expression of Flag-control, Flag-K-bZIP, Flag-K-bZIPDeltaBR, and Flag-K-bZIPDeltaLZ, 293 cells were transfected with 10 mug of each expression plasmid, which contained the G418 resistance gene. After 2 to 3 weeks of selection with medium containing 400 mug of G418 per ml (Gibco-BRL), individual G418-resistant colonies were cloned and tested for the expression of K-bZIP and its mutants by Western blot analysis. These stable cell lines were grown at 37C in DMEM supplemented with 10% fetal bovine serum and 200 mug of G418 per ml in the presence of 5% CO2. To establish BCBL-1 cells stably expressing HA-K-bZIP, cells were transfected with 20 mug of pHA-K-bZIP, which contains a G418 resistance gene. Stable cells were collected as described previously . These stable cell pools were maintained in RPMI 1640 supplemented with 15% fetal bovine serum and 200 mug of G418 per ml in the presence of 5% CO2. Both viral lytic replication and HA-K-bZIP expression were induced by treatment of log-phase HA-K-bZIP stable BCBL-1 cells with tetradecanoyl phorbol acetate (TPA) (20 ng/ml; Sigma). For determination of growth rate, 5 x 104 cells from each of the cell clones were seeded in 60-mm plates. Cell numbers were counted at the indicated time postseeding. Triplicate plates were seeded for each cell clone. Cell cycle analysis. | For cell cycle analysis; BCBL-1, GA10, and BCP-1 cells were collected, washed once with phosphate-buffered saline (PBS), suspended in 0.5 ml of cold (4C) PBS-0.1% glucose, fixed with 5 ml of ice-cold 70% ethanol for at least 45 min at 4C, washed with PBS, and treated with RNase A (0.1 mg/ml) in a propidium iodide (69 mM; Sigma)-sodium citrate (38 mM) solution. Cell cycle analysis was performed by fluorescence-activated cell sorting (FACScan; Becton Dickinson) with the ModFit LT version 2.0 software (Verity Software House Inc.). Immunoprecipitation and immunoblot analyses. | BCBL-1 cells were rinsed in ice-cold PBS, and 107 cells were lysed in EBC lysis buffer (50 mM Tris-HCl [pH 7.5], 120 mM NaCl, 0.5% NP-40, 50 mM NaF, 200 muM Na2VO4, 1 mM phenylmethylsulfonyl fluoride) supplemented with a protease inhibitor cocktail (Roche). After centrifugation (15,000 x g for 10 min at 4C), 20 mul of protein A- and protein G-Sepharose beads (Upstate) were added to the supernatants and incubated overnight at 4C. Then 500 mug of each of the cleared supernatants was reacted with 3 mug of anti-CDK2 (SC-163; Santa Cruz), anti-cyclin A (SC-751; Santa Cruz), anti-cyclin E (SC-481; Santa Cruz), anti-cyclin D (Upstate), anti-CDK4 (SC-260-G; Santa Cruz), anti-p53 (SC-126; Santa Cruz), anti-p107 (SC-318; Santa Cruz), anti-cdc2 (SC-54; Santa Cruz), anti-p21 (SC-397; Santa Cruz), anti-p27 (SC-1641; Santa Cruz), antitubulin (SC-8035; Santa Cruz), or anti-HA (Babco) for 3 h to overnight at 4C with gentle rotation. The immune complex was captured by the addition 20 mul of a protein A- and protein G-Sepharose bead mixture and rocked for an additional 2 h at 4C. Beads were washed four times with EBC buffer and then boiled for 5 min in 20 mul of 2x sodium dodecyl sulfate (SDS) sample buffer (125 mM Tris-HCl [pH 6.8], 4% SDS, 10% 2-mercaptoethanol, 20% glycerol, 0.6% bromophenol blue). 293T cells were cotransfected with 2 mug of pT7-K-bZIP and 3 mug of pFlag-CDK2 or pFlag-empty expression plasmids with FuGene 6 (Roche) according to the supplier's recommendations. The cells were harvested 48 h after transfection and lysed in EBC buffer. Then 500 mug of cell lysates was immunoprecipitated with the addition of 25 mul of anti-Flag antibody-conjugated agarose (Sigma). The beads were washed four times with EBC buffer and then boiled for 5 min in 20 mul of 2x SDS sample buffer. Protein samples from total cell lysates (50 mug/lane) or immunoprecipitates were subjected to SDS-10% polyacrylamide gel electrophoresis (PAGE) and then transferred to a polyvinylidene difluoride membrane (Biotechnology Systems) with a semidry transfer apparatus (Amersham Pharmacia). After blocking for 1 h at room temperature with 5% skim milk in TBST (20 mM Tris-HCl [pH 7.5], 137 mM NaCl, 0.05% Tween 20), the membranes were incubated with primary antibodies for 2 h at room temperature. The membranes were subsequently washed with TBST three times for 10 min each at room temperature and then incubated with horseradish peroxidase-conjugated antibodies for 1 h at room temperature. The membranes were then washed three times with TBST and visualized with enhanced chemiluminescence reagents (Amersham-Pharmacia). Final dilutions of the first antibodies for immunoblotting were 1:4,000 for the rabbit anti-K-bZIP antibody , 1:3,000 for the anti-T7 antibody (Novagen), 1 mug of antiactin (SC-1615; Santa Cruz), 1 mug of anti-CDK2, 1 mug of anti-p21, 1 mug of anti-p27 and 1 mug of anti-p53 per ml in TBST containing 5% skim milk. Immunofluorescence assay. | Forty-eight hours post-TPA treatment, BCBL-1 cells were fixed with 3.7% paraformaldehyde in PBS for 5 min at room temperature and subsequently treated with 1.0% Triton X-100 followed by 1.0% NP-40 in PBS for 10 min each at room temperature. After being washed twice with 0.2% Tween 20 in PBS, cells were smeared on a coverslip (Fisher). After being blocked with PBS-2% bovine serum albumin (Fisher), cells were incubated with anti-K-bZIP rabbit serum (1:2,000) and anti-CDK2 mouse monoclonal antibody (5 mug/ml; B.D. Transduction Labs) in PBS-2% bovine serum albumin for 1 h at room temperature. After being washed four times with PBS, Alexa Fluor 555-conjugated goat anti-rabbit immunoglobulin G F(ab')2 (1:3,000) (Molecular Probes) and Alexa Fluor 488-conjugated goat anti-mouse immunoglobulin G F(ab')2 (1:2,000) (Molecular Probes) in PBS-2% bovine serum albumin were applied as secondary antibodies and allowed to react for 1 h at room temperature. DNA was stained with 1 muM To-Pro3 (Molecular Probes) in PBS for 1 min at room temperature. Imaging was performed with a confocal microscope equipped with an argon-krypton laser (LSM510-MicroSystem; Carl Zeiss Co., Ltd). Preparation and purification of GST fusion proteins. | GST fusion proteins were expressed in Escherichia coli strain BL21 transformed with the following plasmids encoding distinct domains of K-bZIP: pGEX-K-bZIP F.L. (residues 1 to 237), pGEX-K-bZIP I (residues 1 to 121), pGEX-K-bZIP II (residues 107 to 157), pGEX-K-bZIP III (residues 122 to 189), pGEX K-bZIP IV (residues 158 to 208), pGEX K-bZIP V (residues 190 to 237), or pGEX4T-2. The GST fusion proteins were then purified with glutathione-Sepharose beads (Amersham-Pharmacia) by a standard procedure. Bacterial cells (500 ml) were cultured in Luria broth for each construct. Protein expression was induced with 1 mM (final concentration) isopropyl-beta-d-thiogalactopyranoside (IPTG). Bacterial cells were washed once in PBS and then lysed by sonication in PBS containing 1% Triton X-100 and 1% Sarkosyl. After clearing by centrifugation at 7,000 x g for 10 min at 4C, glutathione-Sepharose beads (500 mul of a 1:1 slurry in PBS) were added to the lysates for affinity purification. After incubation overnight at 4C with rotation, the beads were washed four times in PBS containing 1% Triton X-100 and 1% Sarkosyl. The proteins immobilized on the glutathione-agarose beads were quantified by Coomassie blue staining, with bovine serum albumin as a protein standard. In vitro interaction assay. | GST-protein beads containing approximately 2.0 mug of proteins were resuspended in binding buffer (20 mM HEPES [pH 7.9], 150 mM NaCl, 1 mM EDTA, 4 mM MgCl2, 1 mM dithiothreitol, 0.1% NP-40, and 10% glycerol, supplemented before use with 1 mg of bovine serum albumin per ml, 0.5 mM phenylmethylsulfonyl fluoride, and 1 x protease inhibitor cocktail), and then incubated for 30 min at 4C with 10 mul of in vitro translated proteins, which were labeled with [35S]methionine with the TNT coupled transcription and translation system (Promega). The beads were washed four times with binding buffer and then resuspended and boiled in 2x SDS sample buffer. After proteins were separated by SDS-PAGE, radiolabeled polypeptides retained on the beads were visualized by autoradiography. In vitro kinase assay. | cyclin-CDK2 complexes were immunoprecipitated from 250 mug of total cell lysate from 293 stable cells, BCBL-1 cells, or HA-BCBL-1 cells in EBC buffer with anti-CDK2 antibody (SC-163; Santa Cruz). Immunocomplexes were washed three times with kinase buffer (20 mM HEPES [pH 7.5], 10 mM MgCl2, 10 mM MnCl2, 1 mM dithiothreitol, 1 mM NaF, 0.1 mM Na2VO4). Kinase reactions were performed for 15 min at 30C with histone 1 (Upstate) as a substrate. In vitro CDK2 kinase inhibition assays were performed with the full-length GST fusion with K-bZIP. The full-length GST-K-bZIP protein was recovered from bacterial lysates with glutathione-Sepharose beads, separated by SDS-PAGE, excised, equilibrated in SDS electrophoresis buffer, and electroeluted into a cellulose tube. The purified protein in SDS electrophoresis buffer was dialyzed against PBS overnight. Then 2 mul of CDK2-cyclin A (Upstate) was incubated with purified GST-K-bZIP for 1 h at 4C in kinase buffer (10 mul of kinase buffer and 10 mul of PBS containing 0.5 mug of purified GST-K-bZIP as indicated in the figure legends) without ATP. Total protein amounts were adjusted with bovine serum albumin. Kinase reactions were performed for 5 min at 25C in a volume of 30 mul (20 mul of kinase buffer and 10 mul of PBS containing GST-K-bZIP, 10 muCi of [gamma-32P]ATP [Amersham-Pharmacia], 50 muM cold ATP, and 0.5 mug of histone 1 as a substrate). Reactions were stopped by addition of 30 mul of SDS sample buffer, and the mixtures were boiled, fractionated by SDS-PAGE, and visualized with a phosphoimager. Kinase activity was measured with Quantity One software (Bio-Rad). RESULTS : KSHV lytic replication prolongs the G1 phase of naturally infected B cells. | In our investigation of TPA-induced reactivation of KSHV, we observed a general increase in G1 phase of treated cells. These results are illustrated in Table . KSHV latent cell lines BCP-1 and BCBL-1 were used in this study, with GA10, a KSHV-free B-cell line, as a control. We measured the fraction of cells in G1 phase by flow cytometry at different times after TPA treatment. TPA treatment of BCP-1 resulted in a marked increase in the G1 fraction, from 47% before treatment to 71% at 48 h after treatment. TPA treatment of BCBL-1 cells had a less robust effect on the G1 fraction, from 37% before treatment to 46% at 72 h, perhaps reflecting the fact that only 10 to 20% of BCBL-1 cells respond to TPA treatment. We also monitored the molecular events of BCP-1 cells after TPA treatment by Western blot analysis. As shown in Fig. , there was a rapid rise of the level of K-bZIP within 12 h, consistent with its classification as an early-lytic gene product. This was followed by a gradual rise of p21 and p27 protein levels. The amount of p53 did not seem to change significantly in TPA-treated BCP-1, likely due to the high basal level of p53 in these cells. Thus, at least for this cell line, the increase in p21 may not all be attributed to p53 activation. We wished to elucidate the mechanism underlying the cell cycle delay and hypothesized that, in a manner similar to EBV Zta, KSHV K-bZIP may play a role in this process. K-bZIP physically interacts with cyclin-CDK2 in naturally infected B cells. | To explore the possible role of K-bZIP in cell cycle regulation, we asked whether K-bZIP interacts directly with cell cycle regulators in BCBL-1 cells without artificial overexpression of the K-bZIP gene product. To this end, BCBL-1 cells were treated with TPA to induce the expression of K-bZIP, and cell extracts isolated at 48 h after treatment were subjected to immunoprecipitation-Western blot analyses with antibodies against various CDKs, cyclins, and additional unrelated molecules as controls. Among the CDKs and cyclins tested, K-bZIP only coprecipitated with CDK2 and its associated cyclins, cyclins E and A, as shown in Fig. . This is interesting, as Polson et al. demonstrated that K-bZIP is phosphorylated by cyclin-CDK2. p53, which is also known to interact with K-bZIP , was used as a positive control, and antibody against the irrelevant epitope HA was used as a negative control. To further establish the interaction between K-bZIP and cyclin-CDK2 in vivo, we studied their subcellular colocalization in BCBL-1 cells. For this experiment, we used highly specific mouse monoclonal antibodies against CDK2, allowing them to be distinguished from the rabbit immunoglobulin G antibody against K-bZIP. Alexa Fluor 555-conjugated anti-rabbit immunoglobulin G and Alexa Fluor 488-conjugated anti-mouse immunoglobulin G were used to detect K-bZIP (red) and CDK2 (green), respectively. As shown in Fig. , K-bZIP was localized in the nucleus in a diffuse pattern, and CDK2 was primarily present in Cajal bodies, as we showed previously . The merged image indicates that a fraction of K-bZIP and CDK2 were colocalized in the Cajal bodies in this environment. K-bZIP interacts with CDK2 in the absence of other viral factors. | The studies described above were conducted in physiologically relevant KSHV-infected BCBL-1 cells. However, the situation was complicated by the presence of other KSHV gene products in these cells. Additionally, to rule out the possibility that the observed association in BCBL-1 cells was due to nonspecific effects of the polyclonal antibodies used, we constructed CDK2 and K-bZIP expression vectors tagged with Flag and T7 sequences, respectively. These constructs were linked to a cytomegalovirus promoter and cotransfected into 293T cells. Forty-eight hours after transfection, the cells were harvested and lysed. CDK2 was immunoprecipitated with Flag antibody-coated beads, and the coprecipitates were probed with anti-T7 antibody to detect K-bZIP. As shown in Fig. , T7-K-bZIP coprecipitated with Flag-CDK2 but not with the Flag-tagged vector control. Reciprocal experiments with the T7-CDK2 and Flag-K-bZIP expression vectors confirmed this result (data not shown). These data showed that K-bZIP without other viral products is able to interact with CDK2. K-bZIP interacts directly with cyclin-CDK2 via the basic region and downmodulates CKD2 kinase activity. | To determine whether the interaction between K-bZIP and cyclin-CDK2 is direct, we incubated in vitro-translated [35S]CDK2, cyclin A, or cyclin E with purified GST-K-bZIP. As shown in Fig. , full-length GST-K-bZIP pulled down CDK2, cyclin A, and cyclin E. There was some background association of GST with in vitro-translated cyclins A and E, which however could be distinguished from the increased intensities of the GST-K-bZIP pulldowns. To map the interaction domain of K-bZIP, we constructed a series of K-bZIP deletion mutants (Fig. , upper panel) and subjected them to the GST pulldown assay. The results in Fig. , lower panel, showed that the N-terminal half of K-bZIP was ineffective in pulling down CDK2, cyclin A, or cyclin E. By contrast, the basic region appeared to bind CDK2, cyclin A, and cyclin E with an intensity even higher than that of the full-length K-bZIP. The leucine zipper region showed limited affinity toward cyclin A and CDK2. These data suggest that the interaction between K-bZIP and the cyclin-CDK2 complex is direct and that the basic region of K-bZIP is the primary domain interacting with this complex, consistent with the presence of a cyclin-binding motif in this region (see Discussion). This direct interaction has a functional impact on CDK2. In an in vitro kinase assay of the CDK2 complex with histone 1 as a substrate, the addition of GST-K-bZIP significantly reduced the kinase activity (Fig. , upper panel). The lower panel provides the proper reaction control, which indicates that the amounts of histone 1 were similar and the only difference in these two reactions was the addition of purified GST-K-bZIP. These results suggest that direct association of K-bZIP with the cyclin-CDK2 complex leads to inhibition of the kinase activity. K-bZIP expression affects cell growth kinetics. | To study the biological consequence of K-bZIP interaction with cyclin-CDK2, we established stable clones of 293 cells bearing Flag-K-bZIP. Cells transfected with Flag-K-bZIPDeltaBR, a mutant devoid of the basic region, were used as controls. As shown in Fig. , the growth rate of cells bearing wild-type K-bZIP (K-bZIP #1) was significantly lower than the parental line. An independently isolated K-bZIP-293 line (K-bZIP #2) behaved similarly, suggesting that the results are not due to clonal variations of selectants. 293 cells bearing a mutation that deletes the basic region, expected to be incapable of interacting with CDK2, grew with kinetics nearly identical to that of the parental 293 cells. In cells expressing the full-length construct, CDK2 activity was barely detectable, as shown in Fig. . In agreement with the kinetics data, the CDK2 activity in K-bZIPDeltaBR-transfected cells remained similar to that of vector-transfected cells. The downmodulation of CDK2 was also accompanied by an increase in the levels of p21 . It has been reported that phosphorylation by CDK2 targets p21 for degradation . Inhibition of CDK2 activity by K-bZIP would thus be expected to increase the p21 protein level, as we observed here. To offer an independent confirmation of this notion, HeLa cells were transiently transfected with K-bZIP, and the levels of p21 protein were measured. As shown in Fig. , K-bZIP expression led to an increased level of p21. These data suggest that the significant growth-inhibitory effect of K-bZIP is caused by a combined effect of binding of K-bZIP to cyclin-CDK2 and the increased expression of CDK2 inhibitor p21. Recently, Wu et al. showed that K-bZIP enhanced p21 expression through interaction with C/EBPalpha, a transcriptional factor for the p21 locus. These authors mapped the interacting domain to be primarily in the leucine zipper domain of K-bZIP. To distinguish the relative contributions of p21 increase by CDK2 and C/EBPalpha, we constructed a K-bZIP mutant, K-bZIPDeltaLZ, devoid of the leucine zipper region (amino acids 189 to 237), and developed a corresponding, stable 293 clone. K-bZIPDeltaLZ bound cyclin-CDK2 but failed to interact with C/EBPalpha (data not shown). As shown in Fig. , increased p21 expression was observed in a K-bZIPDeltaLZ clone, albeit at a lower level, compared to the wild-type K-bZIP clones. These data suggest that K-bZIP mediated the increase in p21 expression through both CDK2 and other pathways, such as C/EBPalpha. K-bZIP overexpression inhibits growth of BCBL-1 and extends G1 phase of reactivated cells. | Having demonstrated that K-bZIP expression leads to growth inhibition of 293 cells, we wished to extend this analysis to the more appropriate context of BCBL-1 cells. BCBL-1 cells were transfected with HA-K-bZIP, which carries the G418 resistance marker gene. Stable clones selected in medium supplemented with G418 were pooled, and their growth rate was compared to that of vector-transfected BCBL-1 cells which had been similarly selected. As shown in Fig. , the K-bZIP-BCBL-1 pool displayed a significantly slower growth rate than the vector-transfected BCBL-1 cells. This slower growth was accompanied by much reduced CDK2 kinase activity, as shown by the in vitro kinase assay . We also studied the G1 fraction of K-bZIP-BCBL-1 cells after TPA treatment and compared it with that of the parental BCBL-1 cells. As shown in Fig. , the constitutive expression of K-bZIP increased the fraction of cells in G1 before TPA treatment (0 h) to 57% (from the original 37% in parental BCBL-1 cells, Table ). The HA epitope tag allowed us to distinguish between the ectopic and endogenous forms of K-bZIP by size, and their levels through the course of TPA treatment are shown in Fig. . Because TPA treatment induces the expression of the endogenous viral K-bZIP, we would expect it to work in concert with the ectopically expressed HA-K-bZIP to further augment the G1 fraction increase. Indeed, the G0/G1 fraction increased from 54% at 4 h to 63% at 24 h and to 80% at 72 h after TPA treatment. At the same time, the S fraction was reduced from 27% at 4 h posttreatment to only 7% at 72 h. In parallel, the levels of both p21 and p27 increased significantly. This showed that a combined action of exogenous and endogenous K-bZIP expression has a devastating effect on the cell cycle progression of the infected cell. DISCUSSION : Previously, we and others independently reported the identification of a basic leucine zipper protein of KSHV, and we designated this gene K-bZIP. K-bZIP is the structural and positional analogue of the EBV Zta gene. EBV Zta functions as a potent transcriptional factor which triggers lytic replication when overexpressed in latent EBV cell lines, and it, together with Rta, represents the main latent to lytic replication switch for the virus. There is considerable evidence indicating that KSHV Rta, the ortholog of EBV Rta, is a potent activator of KSHV replication. Overexpression of KSHV Rta in the BCBL-1 cell line leads to reactivation of the latent KSHV genome . In contrast, overexpression of K-bZIP has not yielded lytic replication and, as shown in this report, it results in a reduction of the growth rate in the latent cells. We recently reported that K-bZIP is a transcriptional coregulator of KSHV Rta . Through association, K-bZIP suppresses the transactivation ability of KSHV Rta in a promoter-dependent manner. Others have shown that K-bZIP is a coregulator of CBP, p53, and C/EBPalpha . Thus, K-bZIP's transcriptional role seems at present to be limited to its activity as a coactivator or a corepressor. By analogy to EBV Zta, K-bZIP may have other nontranscriptional functions in the KSHV life cycle. EBV Zta is known to bind the viral DNA replication origin and is involved in EBV genomic replication . K-bZIP was similarly shown to assist in the formation of a potential KSHV DNA prereplication complex , though its specific role in genome replication remains elusive. EBV Zta has also been shown to mediate cellular growth arrest by stabilizing p53, inducing the expression of p21 and p27 , thereby contributing to the inhibition of CDK2 activity. In this study, we explored the role of K-bZIP in modulating cell cycle progression. CDK2 plays a pivotal role in cell cycle progression. Phosphorylation by CDK2 of retinoblastoma protein or its related molecules, the p130 and p107 pocket proteins, sets the stage for entry into the S phase by releasing and activating E2F, a transactivator of cyclin E and DNA polymerase . This is viewed as a major pathway by which CDK2 advances G1 to S transition . Recently, it was shown that CDK2 has additional substrates which facilitate this transition in a retinoblastoma protein-independent manner. These substrates include E2F1, E2F5, histone H1, HIRA, and NPAT . HIRA and NPAT are both involved in the histone biosynthesis which accompanies DNA replication . CDK2 also phosphorylates its own inhibitors, p21 and p27, targeting them for degradation . Thus, CDK2 activity controls multiple pathways leading to S-phase progression. It is therefore no surprise that multiple signal transduction pathways such as growth factors, oncogenes, and DNA repair machinery all target CDK2 activity to control cell cycle. Viruses have also evolved ways to target cyclin-CDK2 to modulate the cell cycle to suit their particular needs during replication. Some of these viral gene products, primarily those involved in latency or transformation, accelerate G1 to S transition, while others do the opposite. A number of factors interact directly with cell cycle regulators such as CDKs and cyclins or interact with checkpoint proteins such as the CDK inhibitors p53 and retinoblastoma protein; others transcriptionally modulate CDK inhibitors . There are yet other viral gene products, such as v-cyclin of KSHV, which represent hyperactive homologs of the cell cycle proteins themselves . The consequences of all these interactions are either activation or inactivation of the cyclin-CDK complexes, which leads to acceleration or deceleration of progression through the cell cycle. In this report, we have shown that KSHV K-bZIP targets CDK2 directly and induces growth delay at the G0/G1 phase. K-bZIP achieves this end by physically associating via its basic region with cyclin-CDK2. This interaction is direct, as demonstrated by in vitro studies with purified protein preparations, and can be detected in naturally infected B cells without overexpression of either component. In addition, some of the K-bZIP and CDK2 are found to be colocalized in the nucleus. The consequence of this interaction is the inhibition of CDK2 kinase activity. The inhibitory mechanism is likely to be similar to those of p21, p27, and p57 . This class of factors, which also include retinoblastoma protein, HIRA, and E2F1, bind cyclin-CDK2 tightly and are characterized by cyclin-binding RXL motifs, surrounded by S/TPXK/R motifs, which serve as sites of phosphorylation by CDK2 . We observed that the K-bZIP basic region contains two RXL motifs (KRRLH and TRQLQ), similar to p21, which are both deleted in our K-bZIPDeltaBR construct. These may account for the observed association with cyclins E and A. KRRLH is present in the N-terminal half of the basic region, where as TRQLQ is in the C-terminal half. As shown in Fig. , both halves interact with the cyclin-CDK2 complex, in agreement with the above assertion. Interestingly, precisely 14 amino acids preceding each of the RXL motif is an S/TPXK/R motif (TPPR for RRL, and SPTR for RQL). Both sites were shown by Polson et al. to be phosphorylated by CDK2. Our results also showed that K-bZIP binds in vitro-translated CDK2. We are not sure that this interaction is direct, due to the likely presence of cyclins in the reticulocyte lysates. In either case, we have presented strong evidence that K-bZIP is able to bind the cyclin-CDK2 complex. This tight binding mediates both a potent inhibition of cyclin-CDK2 kinase activity, presumably by blocking the interaction of substrates with cyclin-CDK2 complexes, and a profound reduction in cell growth, as seen in 293 cells overexpressing K-bZIP. Because this cell line is devoid of any other KSHV proteins, the data suggest that K-bZIP alone is sufficient to inhibit cell cycle progression. In BCBL-1 cells, the effect of overexpression of K-bZIP is less dramatic, likely due to the presence of KSHV latent proteins such as v-cyclin and LANA, both of which have been shown to have the potential to accelerate cell cycle progression . The deceleratory effect of K-bZIP may be blunted by these latent viral proteins, which normally occupy a separate and distinct phase of the viral life cycle. After TPA treatment, which represses the expression of latent genes and activates viral lytic genes, including K-bZIP, a significant fraction (nearly 80%) of the cells were in the G0/G1 phase, while only 7% were in S phase. This rather dramatic effect was caused by a combination of the ectopically expressed HA-K-bZIP and the endogenous K-bZIP . The above experiments demonstrate the ability of K-bZIP to delay the G1 to S transition, presumably allowing ample time for virus transcription, translation, and initiation of viral DNA replication to take place. Based on the present work and published reports, there are at least three nonmutually exclusive pathways by which K-bZIP can inhibit CDK2 and induce growth arrest. The first, as shown in this study, is direct binding to cyclin-CDK2. There are other indirect effects of K-bZIP on CDK2 activity. For instance, we observed increased protein levels of both p21 and p27 in cells that overexpressed K-bZIP. As described above, CDK2 phosphorylation of p21 and p27 targets these molecules for degradation. K-bZIP binding would thus have the potential to increase the protein levels of p21 and p27 by shielding them from this degradation signal. A second mechanism, recently reported by Wu et al. , is increased expression of p21 mediated by K-bZIP through binding C/EBPalpha, a factor known to drive the transcription of p21. These authors demonstrated that K-bZIP overexpression in mouse fibroblasts lead to a prolonged G0/G1 phase, accompanied by an increase in p21 transcripts. C/EBPalpha was a necessary component in this process. While their data were primarily based on studies in mouse fibroblasts, the cell cycle inhibition and p21 increase are consistent with what we observed in K-bZIP-overexpressing BCBL-1 cells. It is thus possible that this mechanism also operates in BCBL-1 cells. Finally, it was reported that K-bZIP interacts with and presumably stabilizes p53 . We were able to reproduce these data in KSHV-infected cells . As p53 is a known transcriptional activator of p21, it is conceivable that this interaction may stabilize p53 and thereby lead to increased transcription of p21 in cells overexpressing K-bZIP. The stabilization of p53 by cytomegalovirus IE2 and EBV Zta is a suggested mechanism by which these two viral gene products inhibit cell growth . All three mechanisms described above converge on the inactivation of CDK2, which results in the delayed cell cycle progression from G1 to S. There are other implications of the interaction between CDK2 and K-bZIP. Polson et al. showed that CDK2 phosphorylates K-bZIP at Thr111 and Ser167 in vitro. It is conceivable that K-bZIP inhibits CDK2 activity by serving as a high-affinity substrate. In return, CDK2 phosphorylates K-bZIP, potentially modifying its function. Phosphorylation of bZIP transcription factors may alter their transactivation activity, DNA binding ability, stability, and/or subcellular localization. We have previously shown that Meq, a bZIP protein encoded by Marek's disease herpesvirus, is colocalized with CDK2 in Cajal bodies and phosphorylated by cyclin-CDK2 . CDK2 phosphorylation at Ser42 of Meq weakens its DNA binding affinity and translocates Meq into the cytoplasm. Likewise, herpes simplex virus UL42 DNA synthesis processivity factor is known to bind and be phosphorylated by CDK1 . It is worth noting that CDK2 plays an important role in the replication of other DNA viruses such as human papillomavirus , cytomegalovirus , and polyomavirus . Whether CDK2 plays a direct role in KSHV replication and how CDK2 phosphorylation of K-bZIP may alter its function will require further investigation. The availability of CDK2 phosphoacceptor site mutants of K-bZIP should facilitate such work. In summary, we have shown that K-bZIP interacts directly with and inhibits cyclin-CDK2, thereby prolonging the G1 phase of KSHV-infected cells. Our study and that of Wu et al. clearly establish a nontranscriptional role of K-bZIP in KSHV replication. FIG. 1. : Effect of KSHV reactivation on p53, p21, and p27 levels. Effect of KSHV reactivation on p53, p21, and p27 levels. After TPA treatment, BCP1 cells were lysed with EBC buffer. Protein extract (100 mug/lane) from each time point (in hours) was loaded. K-bZIP was detected with anti-K-bZIP rabbit serum. p53, p21, and p27 antibodies were purchased commercially. Nonspecific reaction of antibody was used as a loading control. Protein expression levels of both p21 and p27 were increased after viral reactivation. FIG. 2. : Coimmunoprecipitation assay. Coimmunoprecipitation assay. (A) Coimmunoprecipitation assay with the KSHV-positive BCBL-1 cell line. BCBL-1 cells induced to viral lytic replication with TPA (48 h) were harvested with EBC buffer, and the same amounts of lysates (500 mug) were precipitated with antibodies against different molecules and then immunoblotted with anti-K-bZIP rabbit serum. (B) Colocalization of K-bZIP with CDK2. Confocal analysis was performed with anti-K-bZIP rabbit serum and anti-CDK2 mouse monoclonal antibody 48 h after TPA induction. K-bZIP (red) and CDK2 (green) were detected with Alexa Fluor 555-conjugated goat F(ab')2 anti-rabbit immunoglobulin G and Alexa Fluor 488-conjugated goat F(ab')2 anti-mouse immunoglobulin G. The nucleus was counterstained with To-Pro-3 (blue). This panel is representative of 10 different fields. (C) Association between K-bZIP and CDK2-cyclins in 293T cells. 293T cells were cotransfected with the indicated plasmids. Cell lysates were precipitated with Flag antibody-conjugated agarose, and coimmunoprecipitation of K-bZIP was detected with anti-T7 antibody. The expression of T7-tagged K-bZIP in total lysates is shown in the same blots as a control. IP, immunoprecipitation; W.B, Western blotting; alpha, anti. FIG. 3. : In vitro interaction and interacting domain of K-bZIP. In vitro interaction and interacting domain of K-bZIP. (A) Domains of K-bZIP and GST-K-bZIP mutants I to V are indicated in the panel. The results of GST pulldown assays for in vitro-translated (IVT), 35S-labeled full-length CDK2, cyclin A, and cyclin E are shown. FL, full length; TA, transactivation domain; BR, basic region; LZ, leucine zipper domain. (B) Inhibition of CDK2-cyclin kinase activity by K-bZIP. CDK2-cyclin kinase activity was measured by in vitro kinase assay with histone 1 as a substrate. The kinase activity of half of the reaction was measured by Quantity One. Relative kinase activity is shown. Protein amounts used in this assay are indicated at the bottom of the panel. The Coomassie-stained gel of the other half of the reaction is shown as a loading control. BSA, bovine serum albumin. FIG. 4. : (A) K-bZIP influence on cell cycle division. (A) K-bZIP influence on cell cycle division. For determination of cell growth rate, 5 x 104 cells of Flag-293-K-bZIPwt 1, Flag-293-K-bZIPwt 2, Flag-K-bZIPDeltaBR, and Flag-293-vector were seeded in 60-mm plates and cultured in 10% fetal bovine serum-DMEM. Cell numbers were counted at the indicated time points postseeding. Triplicate plates were prepared for each cell clone. (B) Inhibition of CDK2 kinase activity in Flag-293-K-bZIPwt cell lines. CDK2 kinase activity was determined by in vitro kinase assay after immunoprecipitation (IP) of CDK2 from stable cell lines. Histone 1 was used as a substrate for the assay. The amounts of the CDK2 were measured by immunoblotting. W.B., Western blotting. (C) Upregulation of p21 in Flag-293-K-bZIPwt and Flag-K-bZIPDeltaLZ cell lines. Protein expression levels of p21 in stably K-bZIP-expressing cell lines were measured. Total cell lysates were prepared from the stable cell lines. Actin served as an internal control for the amounts of protein on the membrane. (D) Upregulation of p21 in transiently transfected HeLa cells. HeLa cells were transfected pFlag-empty vector or pFlag-K-bZIPwt. At 72 h posttransfection, total cell lysates were obtained, and 50 mug of protein was loaded in each lane and probed with anti-p21 antibody. Actin served as an internal control for the amount of protein on the membrane. FIG. 5. : (A) K-bZIP slows down cell cycle division. (A) K-bZIP slows down cell cycle division. A total of 5 x 104 K-bZIP-BCBL-1 cells or parental BCBL-1 cells were seeded in 60-mm plates and cultured in 15% fetal bovine serum-RPMI 1640. Cell numbers were counted at the indicated time points. Triplicate plates were seeded for each cell clone. (B) Inhibition of CDK2 kinase activity in K-bZIP-BCBL-1 cells. CDK2 kinase activity was determined by the in vitro kinase assay after induction of K-bZIP expression (72 h). CDK2 was immunoprecipitated from K-bZIP-BCBL-1 or parental BCBL-1 cells. Histone 1 was used as a substrate for the assay. The amounts of CDK2 were measured by immunoblotting. W.B, Western blotting. FIG. 6. : KSHV reactivation and K-bZIP expression prolong G0/G1 phase. KSHV reactivation and K-bZIP expression prolong G0/G1 phase. (A) Cell cycle analysis of TPA-treated K-bZIP-BCBL-1 cells. Cell cycle analysis after induction by TPA was carried out with flow cytometry. Cell cycle analysis was performed with ModFit LT. The percentage of cells in G0/G1 phase is indicated. (B) Upregulation of p21 and p27 in K-bZIP-BCBL-1 cells after induction. Protein extract (100 mug/lane) from each time point (in hours) was loaded. Protein expression levels of p21, p27, and total K-bZIP in HA-K-bZIP-BCBL-1 cells were measured by immunoblotting. Actin served as an internal control for the amount of protein on the membrane. alpha, anti. TABLE 1 : G0/G1 phase after TPA treatment Backmatter: PMID- 12915536 TI - Replication of Hepatitis C Virus Subgenomes in Nonhepatic Epithelial and Mouse Hepatoma Cells AB - The hepatitis C virus (HCV) pandemic affects the health of more than 170 million people and is the major indication for orthotopic liver transplantations. Although the human liver is the primary site for HCV replication, it is not known whether extrahepatic tissues are also infected by the virus and whether nonprimate cells are permissive for RNA replication. Because HCV exists as a quasispecies, it is conceivable that a viral population may include variants that can replicate in different cell types and in other species. We have tested this hypothesis and found that subgenomic HCV RNAs can replicate in mouse hepatoma and nonhepatic human epithelial cells. Replicons isolated from these cell lines carry new mutations that could be involved in the control of tropism of the virus. Our results demonstrated that translation and RNA-directed RNA replication of HCV do not depend on hepatocyte or primate-specific factors. Moreover, our results could open the path for the development of animal models for HCV infection. Keywords: Introduction : Hepatitis C virus (HCV) is an enveloped, positive-stranded RNA virus that belongs to the Flaviviridae, a family that includes other human pathogens such as Yellow fever virus, Dengue virus, and West Nile virus . Although broad tissue and species tropisms are hallmarks of these viruses, HCV replication has so far been detected only in human and chimpanzee livers. Moreover, for reasons that are not yet understood, the amount of HCV RNA even in infected liver tissue is generally below one copy of RNA per hepatocyte on average. Hence, this viral RNA can be detected only with PCR, making it difficult to determine whether secondary sites for viral replication exist in the infected host . HCV encodes a single polyprotein that is processed proteolytically into 10 polypeptides . Three of these products are structural proteins required for capsid formation (core) and assembly into enveloped viral particles (E1 and E2). Four of the products are enzymes including cysteine and serine proteases (NS2 and NS3), an ATP-dependent helicase (NS3), and an RNA-directed RNA polymerase (NS5B). The functions of the remaining three polypeptides, p7, NS4B, and NS5A, are not yet known . For study of HCV replication in tissue culture cells, the structural proteins can be replaced with a selectable marker, such as the neomycin phosphotransferase . Replication of such subgenomic HCV replicons has so far been demonstrated only in the human hepatoma cell line Huh7. Although the apparent restriction of HCV replication to Huh7 cells would be consistent with the narrow host and tissue tropism of HCV infections, direct evidence for a role of hepatocyte-specific factors in HCV replication has so far been lacking. In fact, other members of the Flaviviridae whose RNA genomes are replicated in a manner very similar to that of HCV generally exhibit broad tissue and host tropism. Moreover, the efficient replication of HCV in Huh7 cells depends on adaptive mutations located in the nonstructural (NS) genes . Such considerations raise the possibility that additional mutations might allow HCV replication to occur in nonhepatic or nonhuman cells. This notion is supported by results described in this report. We show that amplification of HCV RNA can occur in HeLa and mouse hepatoma cells, indicating that host factors required for RNA replication are not hepatocyte specific and not restricted to cells of human origin. MATERIALS AND METHODS : Cell culture. | Cells were purchased from the American Type Culture Collection . The Huh7-derived cell lines GS4.1 and GS4.5 are subclones derived from cell lines FCA1 and FCA4, respectively . Cell line Bsp8 is a Huh7-derived cell line expressing HCV-N subgenomic replicon 1bneoDeltaS . All cultures were grown in Dulbecco's modified Eagle's medium (Gibco-Invitrogen) supplemented with 10% fetal bovine serum, l-glutamine, nonessential amino acids, penicillin, and streptomycin. The conditions used for the transfection of cells with total RNA were identical to those used for the transfection with in vitro-transcribed RNA . Colonies were selected with G418 at a concentration of 1 mg/ml. RNA transfection. | All the plasmids were linearized with ScaI, and RNA was synthesized with the MEGAscript kit (Ambion). In vitro-transcribed RNA was purified as previously described . Total cellular RNA was extracted with Trizol reagent (Invitrogen). The conditions used for the transfection of cells with total RNA were identical to those used for the transfection with in vitro-transcribed RNA . Colonies were selected with G418 at a concentration of 1 mg/ml. RNA analysis. | Total cellular RNA was extracted with Trizol reagent. Five micrograms of total RNA was fractionated on 1% agarose gels containing 2.2 M formaldehyde and transferred onto a nylon membrane. Membranes were hybridized with riboprobes specific for plus-stranded HCV replicon RNA, human papillomavirus (HPV) E6, and mouse albumin mRNA as described previously . The HPV and mouse albumin probes spanned nucleotides 811 to 1491 (GenBank accession number ) and nucleotides 1501 to 1988 (GenBank accession number ), respectively. Reverse transcription-PCR and DNA sequencing. | Nucleotide and amino acid numbers correspond to the HCV type 1b genome Con-1 (AJ238799). HCV replicons were isolated and cloned from established cell lines by PCR amplification of three fragments spanning the entire NS region from position 3420 to 9410. The untranslated regions at the 5' and 3' ends of HCV RNA were cloned separately for nucleotide sequence analysis. DNA synthesis was carried out with Superscript II reverse transcriptase provided in a cDNA synthesis kit (Gibco-Invitrogen). The DNA oligomers used as primers for the reverse transcription reaction mapped to positions 485 to 465, 5492 to 5473, 7256 to 7234, 9410 to 9388, and 9616 to 9597. The reaction mixtures were incubated for 1 h at 45C. PCR was performed with an Advantage PCR kit (Clontech). One microliter of the cDNA reaction mixture was used for PCRs with 19- to 23-nucleotide-long primers that yielded fragments spanning positions 1 to 464, 1387E to 5082, 5016 to 7226, 7154 to 9387, and 9239 to 9616. Position 1387E refers to an oligomer specific for the encephalomyocarditis virus (EMCV) internal ribosome entry site (IRES) element located upstream of NS3. The PCR products were cloned into plasmid pGEM-T Easy (Promega). Four clones of each fragment were sequenced with an ABI automatic DNA sequencer, and a consensus sequence was established with the help of a sequence assembly program (Genetics Computer Group). Long reverse transcription-PCR was performed with an Advantage-GC kit (Clontech) with a pair of primers beginning at positions 1415E, upstream of NS3, and 7989 within NS5B. The PCR conditions were modified as follows: step 1, 95C for 3 min; step 2, 5 cycles, 30 s at 95C and 6 min at 72C; step 3, 27 cycles, 30 s at 95C and 6 min at 68C; step 4, 68C for 6 min. PCR products were gel purified and digested with HindIII and MfeI and replaced with the corresponding fragment in plasmid I377/NS3-3'. Plasmid construction. | All plasmids were derived from the parental HCV Con-1 replicon I377/NS3-3' (AJ242652). Subgenomes containing consensus mutations were constructed by replacing DNA restriction fragments with the corresponding fragments from the pGEM-T Easy cDNA libraries (see above). The resulting plasmids with the amino acid changes in the NS region are listed in Table . Sequence files for each plasmid are available upon request. Immunofluorescence. | Cells were plated on coverslips in six-well plates at least 16 h before treatment, washed with phosphate-buffered saline, and fixed with cold methanol-acetone (1:1) for 15 to 20 min. Next, the cells were blocked in phosphate-buffered saline containing 10% fetal bovine serum for 30 min at room temperature and then incubated with anti-NS5A antibodies (a gift from Chen Liu) and fluorescein isothiocyanate-conjugated goat anti-mouse immunoglobulin antibodies (Jackson Laboratories). In addition, cells were stained with the DNA-binding fluorochrome DAPI (4',6'-diamidino-2-phenylindole). Coverslips were mounted with antifade agent (Molecular Probes), examined with a Nikon immunofluorescence microscope, and photographed with a charge-coupled device camera. RESULTS : HCV replication in cells of nonhepatic origin. | As HCV exhibits a very narrow host range and infects only humans and chimpanzees, we asked whether this limitation was due to determinants of RNA replication. Because efficient replication of subgenomes depends on genetic adaptations of the replicon , presumably to compensate for subtle variations in the cellular environments among cells from different tissues, we hypothesized that replication in cells of nonhepatic origin would require additional, cell-type-specific adaptive mutations. Transfection of several primate- and rodent-derived cell lines with subgenomic RNA transcribed from plasmid DNA carrying previously identified adaptive mutations in Huh7 cells did not yield cell lines expressing replicons (Table and results not shown). To increase the chance for the selection of RNA subgenomes capable of replicating in cells of nonhepatic origin, we used subgenomic RNA isolated from Huh7 cell lines that replicate HCV RNA. Because of the high rate of nucleotide incorporation errors that occur during RNA-directed RNA synthesis, this population of viral subgenomes exhibited a much larger genetic heterogeneity than did RNA transcribed from a DNA template in vitro previously used for the transfection of Huh7 cells. Upon transfection of BHK, Vero, CV-1, HT1080, and HeLa cells with total RNA obtained from Huh7 cell lines GS4.1, GS4.5, and Bsp8, we obtained G418-resistant cell clones only with HeLa cells. The number of clones ranged from approximately 2 (Bsp8) to 50 (GS4.1) per 10 mug of total RNA depending on the origin of the RNA used for the transfections. Replicons in these three Huh7-derived cell lines contained different adaptive mutations and replicated two different HCV 1b genomes . Several HeLa-derived colonies obtained with total RNA from GS4.1 cells were subsequently expanded into seven stable cell lines (SL1 to SL7; Fig. , lanes 4 and 8 to 12). The amounts of viral RNA present in early passages of these cell lines examined ranged from 0.05 to 7.5 ng/10 mug of total RNA, which corresponded to 20 to 3,000 copies of RNA per cell. In general, the amounts of RNA increased upon passage of cells and reached levels that were comparable to those obtained with the most productive Huh7-derived cell lines such as GS4.1 (lanes 2 and 4 to 6). As expected, expression of viral gene products could be confirmed by immunofluorescence with antibodies directed against NS5A . As with GS4.1 cells, more than 90% of SL1 cells expressed viral proteins. However, in contrast to Huh7 cell lines where the accumulation of HCV RNA declines approximately 100-fold when cells become confluent, viral replication in HeLa cells was not affected by the growth conditions of the cells, i.e., SL1 cells continued to produce high amounts of viral RNA even when they became confluent (results not shown) . Adaptation of HCV replicons. | To determine whether HCV replication in HeLa cells led to the selection of subgenomes with cell-type-specific adaptive mutations, we compared the efficiency by which G418-resistant colonies formed in Huh7 and HeLa cells transfected with total RNA isolated from GS4.1 and SL1 cells. Total RNA from GS4.1 cells led to the selection of approximately 166 G418-resistant colonies per ng of viral RNA in Huh7 cells compared with only 4 colonies in HeLa cells . In contrast, total RNA from SL1 cells yielded 160 colonies in HeLa cells compared with about 20 in Huh7 cells. These results indicated that replication in HeLa cells led to the selection of variants with cell-type-specific adaptive mutations that were responsible for the 40-fold increase in colony formation efficiency between amplified RNA in GS4.1 and SL1 cells. Nucleotide sequence analysis of HCV cDNA clones obtained from the SL1 and SL2 cell lines confirmed this view. These data showed that replicons in the two HeLa cell lines maintained the previously identified adaptive mutations in GS4.1 cells and acquired several additional mutations that resulted in amino acid changes in the NS region (Fig. and Table ). Notably, some of the new mutations formed clusters in the NS4B and NS5A regions. In the case of SL1 cells, we observed a deletion of 43 amino acids near the C terminus of NS5A. Of particular interest were mutations in the amino-terminal region of NS4B, because they have so far not been found in cDNAs from replicons in Huh7 cells and hence could have been responsible for the observed adaptation of replicating RNA . Moreover, one mutation at position 1749 was present in both SL1 and SL2 cells. In contrast to the results obtained with the NS regions, we could not detect any mutations in the 5' and 3' untranslated regions of replicons expressed in SL1 and SL2 cells. Mouse hepatoma cells can support HCV RNA replication. | The discovery of several additional mutations in cDNA clones obtained from SL1 and SL2 cells prompted us to examine whether total RNA from these cell lines could yield colonies in cells that did not appear to be permissive for HCV replication after transfection with subgenomic RNA or total RNA from Huh7-derived cell lines. In addition to the five cell lines that we examined initially with Huh7-derived RNA, we examined five additional hepatoma and hepatocyte-derived cell lines . G418-resistant colonies were obtained only with the mouse hepatoma cell line Hepa1-6 after transfection with total RNA from SL1 cells (Fig. , lanes 4 to 6, 9, and 12). As with HeLa cells, the amounts of RNA ranged from 300 to 1,000 copies of RNA per cell and a large fraction of the cells expressed viral proteins . In contrast to Huh7 and HeLa cells, the amount of HCV RNA in the mouse cell lines appeared to vary between cell passages (Fig. , lanes 6 to 14). Interestingly, total RNA isolated from one of the mouse cell lines, MH1, did not produce significantly more colonies in Hepa1-6 cells than did total RNA from SL1 cells, suggesting that the subgenomes present in SL1 cells were already adapted for replication in the mouse cells . In support of this interpretation, nucleotide sequence analysis of viral cDNAs cloned from three mouse cell lines showed that the majority of the mutations identified in SL1 cells were maintained . Surprisingly, the deletion in NS5A identified in four of four clones sequenced from SL1 cells was not present in replicons isolated from mouse cells, indicating that a subpopulation of replicons without the deletion was still present in these (SL1) cells. Cell-derived HCV RNA is more efficient than in vitro-transcribed RNA in initiating replication in HeLa and mouse hepatoma cells. | So far, our results showed that replication of HCV subgenomes in HeLa and mouse cells led to the selection of replicons with several novel mutations. The majority of these mutations were located in the NS3, NS4B, and NS5A regions. Moreover, the results showed that cell-derived RNA carrying some or all of these mutations was much more efficient in establishing G418-resistant colonies in HeLa cells than was RNA derived from Huh7 cells . Based on these observations, we surmised that introduction of these mutations into available subgenomic replicons should alter or expand their tissue and host tropism. To test this hypothesis, we constructed 13 subgenomic replicons that carried mutations in NS3, NS4B, and NS5A alone or in combination with each other as described in Table . Of the 13 constructs examined, only two, pZS2 and pZS25, yielded a small number of G418-resistant colonies in HeLa cells . Viral RNA replication was confirmed by Northern blot analysis of total RNA isolated from six cell lines derived from those colonies. None of the variants yielded colonies in Hepa1-6 cells. Moreover, negative-control experiments with in vitro-transcribed RNA derived from a variant containing a frameshift mutation in NS5B did not yield any colonies that could be expanded into cell lines. Notably, save for one, all replicons were permissive for replication in Huh7 cells, albeit with significantly different efficiencies . Interestingly, both pZS2 and pZS25 carried mutations in NS4B that were conserved in replicons from two independent HeLa cell lines, SL1 and SL2. In addition, these replicons had the S2204I mutation in NS5A that was previously found to be one of the most potent adaptive mutations for HCV replication in Huh7 cells. Because both replicons replicated very efficiently in Huh7 cells, the results suggested that the NS4B mutations could have contributed to the observed expansion of the tissue tropism of HCV replicons. In support of this hypothesis, the subgenome with the highest efficiency in Huh7 cells, pZS11 lacking mutations in NS4B , did not yield any colonies in HeLa cells. However, the number of colonies obtained with in vitro transcripts was too low to draw firm conclusions . To further explore the basis for the observed low colony formation efficiency of in vitro-transcribed RNA in HeLa cells, we examined whether replication in HeLa cells led to the selection of adaptive mutations that were not discovered previously when we sequenced cDNA clones from SL1 and SL2 cells. For this purpose, cDNA clones were isolated from total RNA obtained with pZS2- and pZS25-derived cell lines, respectively. Nucleotide sequence analysis of both cDNA clones did not reveal any additional consensus mutations, suggesting that the two subgenomes were sufficiently adapted for replication in HeLa cells (results not shown). However, as mentioned above, we could not exclude the possibility that a minor population of subgenomic replicons with additional mutations were present in these cell lines. To overcome this problem, we developed a method for the isolation and cloning of cDNAs spanning the NS3 to NS5B region (see Materials and Methods). We produced replicon cDNA libraries from GS4.1, SL1, and MH4 cells. Approximately 2,000 cDNA clones were pooled and subsequently used for in vitro transcription of subgenomic RNA. With Huh7 cells, the colony formation efficiency of the pooled clones was comparable to that of the most efficient subgenomes, such as pZS2 or pZS25, and did not vary significantly with the origin of the total RNA used for cDNA cloning . Consistent with previous results, colony formation in HeLa and mouse cells was origin dependent, i.e., save for one case, colonies were observed only with clones derived from SL1 and MH4 cell lines. Notably, with this strategy we were able for the first time to obtain G418-resistant colonies with Hepa1-6 cells by using in vitro-transcribed RNA. To confirm the presence of viral RNA, we expanded 11 colonies and performed Northern blot analysis with total RNA. All 11 RNA samples analyzed contained viral RNA ranging from approximately 0.1 to 1 ng/5 mug of total RNA (results not shown). Taken together, the results supported the hypothesis that mutations identified in subgenomic replicons expressed in HeLa and mouse cells play a role in adaptation of the replicons to certain cell-type-specific conditions. Importantly, the results invoked the possibility that differences exist between amplified RNA in cells and in vitro-transcribed RNA that influence the efficiency by which HCV subgenomes initiate replication in HeLa and mouse hepatoma cells. DISCUSSION : HCV is known as a species- and tissue-specific virus. This report now shows that replication of HCV can occur in cells derived from tissues other than liver, indicating that cellular factors required for RNA replication are expressed in cell types other than hepatocytes. One interpretation of this result is that the apparent tropism of HCV for hepatocytes is determined primarily at the level of virus entry or assembly or, alternatively, that HCV can infect many other tissues but has escaped detection due to very low amounts of RNA replication or accumulation. Extrahepatic tissues could serve as reservoirs for HCV that, as with human immunodeficiency virus, could provide a source of viruses that are refractory to antiviral therapy and, importantly, can be responsible for infection of liver grafts following orthotopic liver transplantation . Such a scenario would have profound implications for antiviral therapy. For example, the targeting of drugs to secondary sites of viral replication and the analysis of drug metabolism in cells other than hepatocytes would become important factors for the development of successful antiviral therapies. Although proof for HCV replication in cells of nonhepatic origin is still lacking, there is ample evidence for the presence of viral RNA in lymphocytes and other tissues . The results of this study will encourage further investigations that might provide convincing evidence for HCV replication in extrahepatic tissues. It is conceivable that HCV quasispecies in hepatocytes and other tissues exhibit differences in their composition due to the selection of variants with cell-type-specific adaptations. As shown in this report, replication of subgenomes in HeLa cells led to the accumulation of clusters of mutations in the NS3, NS4B, and NS5A regions including a deletion in NS5A . Mutations and deletions in NS5A have been found previously in genomes that replicated in Huh7 cells, which could suggest that expression of the natural form of this protein in cell culture somehow interferes with RNA replication . However, mutations in the amino terminus of NS4B have previously not been observed. Notably, in both SL1 and SL2 cells, the mutations changed two or one glutamine residues, respectively, to one of the two basic amino acids arginine and histidine. Moreover, the mutation V1749A was present in all five cell lines examined (Table and Fig. ). Thus far, our results showed that these mutations appeared to be required for replication in HeLa cells, because only replicons pZS2 and pZS25 carrying these mutations yielded colonies after transfection with in vitro-transcribed RNA (Tables and ). However, due to the low efficiency in colony formation obtained with in vitro-transcribed RNA, our results did not yet provide definitive proof for such a conclusion (see below). The amino terminus of NS4B is predicted to reside on the cytoplasmic side of endoplasmic reticulum membranes and may interact with other host or viral proteins required for RNA replication . As an integral endoplasmic reticulum membrane protein, NS4B might provide a scaffold for the assembly of replication complexes and act as a regulator for RNA replication. More importantly, a recent study revealed that NS4B can induce particular membrane structures, called membranous webs, proposed to be the site for HCV replication . Interestingly, genetic analyses with an HCV-related pestivirus identified the amino-terminal region of NS4B as a determinant for cytotoxicity caused by high levels of virus replication . Although the exact mechanism by which NS4B exerts this activity is still unknown, it might interact with cell-type-specific factors and cause the selection of variants with adaptive mutations as shown in this study. For reasons that we do not yet understand, we could not yet obtain subgenomes that replicated with high efficiency, transiently or permanently, in HeLa or mouse cells (Tables and ). Although it is conceivable that we missed a critical mutation, because it was for some reason underrepresented in our cDNA clones, it is puzzling that it did not arise following the transfection of HeLa cells with RNA. Based on the experience with Huh7 cells, we would have expected that such an event would have occurred and eventually led to the identification of the critical adaptive mutation(s). Nevertheless, our results indicated that the mutations in NS4B and NS5A in replicons pZS2 and pZS25 were sufficient to establish replication in a small number of HeLa cells, because, based on our sequence analysis of cDNA clones, cell lines obtained with these subgenomes did not contain any additional mutations (results not shown). Moreover, transfection of HeLa and mouse cells with heterogeneous populations of subgenomes that should have represented the populations of amplified RNA in cells, with one exception, did not yield more colonies than in vitro-transcribed RNA . Hence, based on our results we were speculating that in vitro-transcribed RNA exhibited some toxicity in HeLa or mouse cells. However, this is an unlikely scenario, because we found that in vitro-transcribed RNA did not alter the colony formation efficiency in HeLa cells when added to total RNA isolated from HCV-containing cell lines, such as GS4.1 and SL1 (results not shown). Similarly, addition of small amounts of in vitro-transcribed RNA to total RNA from normal Huh7 or HeLa cells did not yield any G418-resistant colonies. These results also indicated that cellular mRNAs did not influence the colony formation efficiency. In addition, we did not observe an increase in colony formation efficiency when we used cured HeLa cells that were obtained through the treatment of subgenome-expressing cells with an HCV polymerase inhibitor. It is conceivable that establishment of HCV replicons in HeLa cells requires certain adaptive mutations that are not required for the maintenance of replicons during the expansion of G418-resistant colonies. Because the amount of viral RNA in total cellular RNA is too low to permit detection of transient replication, we have not been able to directly test this hypothesis. Finally, our results could indicate that amplified viral RNA isolated from cells exhibits physical differences from RNA that is transcribed in vitro, such as methylation of certain residues or other, so far unrecognized modifications at the termini of viral RNA that are required for initiation of replication in HeLa cells. In summary, we have shown in this report that HCV RNA replication is not restricted to the human hepatoma cell line Huh7 but instead can occur in HeLa cells and hepatoma cells derived from mice. These findings suggest that it may be possible to develop a mouse model for HCV infection. Establishment of such a model will depend on the isolation of HCV variants that can infect mouse hepatocytes or transgenic mice that express the still elusive HCV receptor. FIG. 1. : Physical map of HCV. Physical map of HCV. The open reading frame is flanked by the IRES and the 3' untranslated region containing a polypyrimidine tract (poly U-UC), respectively. Subgenomes contain the neomycin phosphotransferase gene (neo) in lieu of the structural genes (C, core; gpE1 and gpE2, envelope proteins; p7) and the NS protein 2 (NS2 protease). Translation of the remaining NS proteins (NS3, protease; NS4A, cofactor for protease; NS4B; NS5A; NS5B, polymerase) is regulated by the IRES derived from EMCV. Proteins of unknown function for viral replication are marked with white rectangles. The cleavage sites for the three proteases required for processing of the polyprotein are indicated with circles. FIG. 2. : Replication of HCV subgenomic replicons in HeLa cells. Replication of HCV subgenomic replicons in HeLa cells. (A) Detection of HCV viral RNA. Total RNA (5 mug) was isolated from HeLa cell lines that were established from G418-resistant cell colonies and analyzed by Northern blot analysis. Blots were hybridized with radiolabeled RNA probes corresponding to the HCV NS5 region to detect viral RNA (vRNA) and the DeltaE1 region of HPV present in HeLa cells. In vitro-transcribed HCV RNA (1 ng plus 5 mug of total RNA from Huh7 cells, lane 7) served as a marker (M) and control for the hybridization reaction, and 28S rRNA served as a control for the amount of RNA present in each sample analyzed. GS4.1 is a Huh7-derived cell line expressing HCV subgenomic replicons. RNA in SL1 cells was analyzed from cells harvested at the indicated passage (p). SL3 to SL7 were analyzed at passage 3. (B) Immunohistochemical analysis of HCV replication in HeLa cells. Expression of NS5A in GS4.1 and SL1 cells (passage 26) was detected with a monoclonal antibody bound to fluorescein isothiocyanate-conjugated antibody. Parental Huh7 and HeLa cells served as controls. FIG. 3. : Sequence analysis of HCV subgenomes in HeLa and mouse hepatoma Hepa1-6 cells. Sequence analysis of HCV subgenomes in HeLa and mouse hepatoma Hepa1-6 cells. (Left) Physical map of HCV subgenomic RNA including the positions of the first amino acid NS3 and the last residue of the polyprotein. The IRES for the translation of the NS genes is indicated (EMCV-IRES). (Right) Mutations causing amino acid changes identified in cDNAs isolated from subgenomic replicons present in GS4.1 (passage 21) and the indicated HeLa (SL1, passage 26; SL2, passage 5) and Hepa1-6 (MH1 [passage 12], MH2 [passage 4], and MH4 [passage 4]) cell lines are depicted with horizontal bars. Four independent clones were sequenced from each PCR fragment that was amplified from cDNAs obtained from total RNA purified from the indicated cell lines. Mutations present in more than one cell line are indicated with their amino acid positions. Mutations that occurred in 50% of the clones analyzed are indicated with an asterisk. Mutations that occurred in only one of four clones analyzed are not included in the figure. A deletion identified in cDNA clones obtained from SL1 cells spanning amino acids 2371 to 2413 is indicated (Delta). FIG. 4. : Replication of HCV subgenomes in mouse hepatoma cells. Replication of HCV subgenomes in mouse hepatoma cells. (A) Detection of HCV viral RNA. RNA in Hepa1-6 cell lines (MH1 to MH5) that were established from G418-resistant cell colonies was analyzed as described in the legend to Fig. except that a radiolabeled probe specific to the mouse albumin cDNA was used in lieu of the probe against HPV. MH4 and MH5 were analyzed at passage 4. (B) Immunohistochemical analysis of HCV replication in MH1 cells (passage 3). Expression of NS5A was detected as described in the legend to Fig. . Hepa1-6 cells served as a negative control. TABLE 1 : Cell lines tested in this study TABLE 2 : Colony formation efficiency of total cellular RNA TABLE 3 : Consensus mutations in replicons isolated from HeLa and mouse hepatoma cell clones TABLE 4 : Colony formation efficiency of mutant replicons in Huh7 cells TABLE 5 : Colony formation efficiency of in vitro-transcribed RNA Backmatter: PMID- 12915566 TI - Newcastle Disease Virus V Protein Is a Determinant of Host Range Restriction AB - It has been demonstrated that the V protein of Newcastle disease virus (NDV) functions as an alpha/beta interferon (IFN-alpha/beta) antagonist (M. S. Park, M. L. Shaw, J. Munoz-Jordan, J. F. Cros, T. Nakaya, N. Bouvier, P. Palese, A. Garcia-Sastre, and C. F. Basler, J. Virol. 77:1501-1511, 2003). We now show that the NDV V protein plays an important role in host range restriction. In order to study V functions in vivo, recombinant NDV (rNDV) mutants, defective in the expression of the V protein, were generated. These rNDV mutants grow poorly in both embryonated chicken eggs and chicken embryo fibroblasts (CEFs) compared to the wild-type (wt) rNDV. However, insertion of the NS1 gene of influenza virus A/PR8/34 into the NDV V(-) genome [rNDV V(-)/NS1] restores impaired growth to wt levels in embryonated chicken eggs and CEFs. These data indicate that for viruses infecting avian cells, the NDV V protein and the influenza NS1 protein are functionally interchangeable, even though there are no sequence similarities between the two proteins. Interestingly, in human cells, the titer of wt rNDV is 10 times lower than that of rNDV V(-)/NS1. Correspondingly, the level of IFN secreted by human cells infected with wt rNDV is much higher than that secreted by cells infected with the NS1-expressing rNDV. This suggests that the IFN antagonist activity of the NDV V protein is species specific. Finally, the NDV V protein plays an important role in preventing apoptosis in a species-specific manner. The rNDV defective in V induces apoptotic cell death more rapidly in CEFs than does wt rNDV. Taken together, these data suggest that the host range of NDV is limited by the ability of its V protein to efficiently prevent innate host defenses, such as the IFN response and apoptosis. Keywords: Introduction : Newcastle disease virus (NDV), an avian paramyxovirus, is classified as the only member of the genus Avulavirus belonging to the family Paramyxoviridae within the order Mononegavirales . NDV is an economically important pathogen, since periodic outbreaks affect the poultry industry. NDV is also considered a potential oncolytic agent in the treatment of cancer because it can selectively kill tumor cells . NDV isolates are categorized as velogenic (highly virulent), mesogenic (intermediate), or lentogenic (nonvirulent), depending on the severity of the disease they cause . A critical molecular determinant for the pathogenicity of NDV appears to be the cleavage site of the fusion (F) protein . The NDV genome is 15,186 nucleotides long, and it consists of six transcriptional units that encode the nucleocapsid protein (NP), phosphoprotein (P), matrix protein (M), F protein, hemagglutinin protein (HN), and the polymerase protein (L). Two additional proteins, V and W, are expressed by mRNAs, which are derived from the P gene via RNA editing . These V and W proteins share their amino (N)-terminal domains with the P protein and vary at their carboxy (C) termini. The NDV V protein, similar to other paramyxovirus V proteins, has a cysteine-rich C-terminal domain which binds two atoms of Zn2+ . It has been demonstrated that plasmid-mediated expression of the NDV V protein or of its C-terminal domain inhibits the alpha/beta interferon (IFN-alpha/beta) response . We now show, using reverse genetics, that this IFN antagonist activity is important for virus replication in vivo. In addition, we found that V activity is restricted to avian hosts. Moreover, we show that the NDV V protein and the influenza A virus NS1 protein are functionally interchangeable and that the host restriction of NDV can be partially overcome by the expression of the influenza A virus NS1 protein. Several viruses have evolved strategies to regulate IFN-related responses through the synthesis of IFN-alpha/beta antagonists . Specifically, influenza viruses and paramyxoviruses use distinct virus-specific proteins to counteract the IFN response . Some influenza viruses and paramyxoviruses, including simian virus 5 (SV5) and respiratory syncytial virus, have been shown to inhibit the IFN response in a species-specific manner , suggesting that the IFN-alpha/beta antagonist activity may affect the host range restriction of viruses. It was previously shown that IFN-alpha/beta cytokines are mediators of apoptotic death in virus-infected cells . Successful viral replication requires evasion of proapoptotic mechanisms in order to achieve efficient virus production and spread of progeny . Recently, it was shown that a mutant of SV5, a virus closely related to NDV, lacking the C-terminal cysteine-rich domain of its V protein induced increased cytopathic effects (CPE) in infected cells . In addition, SV5 requires expression of the small hydrophobic (SH) gene to efficiently prevent apoptosis induced by viral infection . In this context, it should be noted that NDV does not code for an SH protein. Here, we show through the use of recombinant NDV (rNDV) mutants defective in V protein expression that the NDV V protein plays an important role in preventing IFN responses and apoptosis in chicken cells, but not in human cells. These species-specific effects may provide a partial explanation for the limited host range of NDV. MATERIALS AND METHODS : Cells and viruses. | Chicken embryo fibroblasts (CEFs) were prepared from 10-day-old specific-pathogen-free chicken embryos (Charles River SPAFAS, North Franklin, Conn.). The CEFs were maintained in minimum essential medium containing 10% fetal bovine serum. Vero, Hep-2, and A549 cells were maintained in Dulbecco's modified Eagle medium containing 10% fetal bovine serum. rNDV and rNDV-GFP viruses were previously generated by reverse genetics from the corresponding full-length cDNA copies derived from the NDV Hitchner B1 strain . Construction of plasmids. | pNDV/B1 containing the full-length cDNA of the lentogenic NDV Hitchner B1 strain has been described previously . To construct rNDV V(-)2ch, we substituted 2 nucleotides in the RNA-editing locus of the P gene . First, the SacII-XbaI NDV cDNA fragment was subcloned into plasmid pSL1180 (Amersham Pharmacia Biotech), and then site-directed mutagenesis PCR was performed using the primers Vko(+) (5'-ATC GTC CAA TGC TAA gAA aGG CCC ATG GTC G) and Vko(-) (5'-CGA CCA TGG GCC tTT cTT AGC ATT GGA CGA T) (the lowercase letters represent the mutated nucleotides). The mutated fragment was then reintroduced into the full-length NDV clone. pNDV V(-)stop has a full-length cDNA copy of a mutated NDV genome encoding a truncated V protein . In order to selectively block the expression of the C terminus of the NDV V protein, a stop codon was introduced into the open reading frame (ORF) of V using the following primers: Vstop(+) (5'-AAG GGC CCA TGG TCt AGC CCC CA), containing an ApaI site (underlined), and Vstop(-) (5'-TGC CGC TTC TAG AGT TGG ACC TTG), containing an XbaI site (underlined). The lowercase letter represents the mutation introducing the stop codon in V without affecting the P frame. To insert the NS1 genes of the influenza A/PR/8/34 virus or the NDV V gene as an extra transcriptional unit into the NDV mutant genomes, an NruI site was introduced between the HN and L genes of rNDV V(-)2ch and rNDV V(-)stop. A pNDV/B1 fragment digested with SpeI and BsiWI was subcloned into pSL1180, and then site-directed mutagenesis was performed using the primers SDM Nru(+) (5'-AAC AGC TCA TGG TTC GCG ATA CGG GTA GGA CA) and SDM Nru(-) (5'-TGT CCT ACC CGT ATC GCG AAC CAT GAG CTG TT), containing an NruI site (underlined). A new transcriptional unit encoding the NS1 or V protein was then inserted between the HN and L genes using the NruI restriction site. Three nucleotides (lowercase) in the RNA-editing locus of the new V gene were changed to block RNA editing (UUUUUCCCC to UUcUUuCCgC). Generation of rNDV mutants. | rNDV mutants were generated using a reverse genetics system established previously . Briefly, Hep-2 or A549 cells in six-well plates were infected with MVA-T7 and then transfected with rNDV V(-)2ch, rNDV V(-)stop, rNDV V(-)/V, or rNDV V(-)/NS1 together with pTM1-NP, pTM1-P, and pTM1-L. After overnight incubation, the transfected cells were cocultured with CEF cells, and 3 or 4 days posttransfection, the supernatants were injected into the allantoic cavities of 7- or 8-day-old embryonated chicken eggs. Allantoic fluids were harvested 3 or 4 days postinjection, and viral growth was confirmed by hemagglutination assay or by infection immunofluorescence assay (IFA) in freshly infected Vero cells. The insertion of the new transcriptional units and the introduction of the mutations were confirmed by reverse transcription-PCR followed by sequencing analysis. Viral growth kinetics. | Embryonated chicken eggs were inoculated with nonrecombinant wild-type (wt) NDV, wt rNDV, or rNDV mutants (100 TCID50/egg). Monolayers of CEFs and Vero, Hep-2, and A549 cells were infected with each virus at a multiplicity of infection (MOI) of 0.01. Viral growth was analyzed at different time points after infection. The 50% tissue culture infective dose (TCID50) of each virus present in the allantoic fluid or in cell supernatants was determined by IFA. For this purpose, 96-well plates containing Vero cells were infected with serial 10-fold dilutions of the samples, and the presence of NDV was determined by IFA using an anti-NDV rabbit polyclonal serum. IFAs. | For the analysis of viral growth and viral protein expression, confluent CEFs were infected with the recombinant virus (four wells per dilution). The cells were incubated for 2 days and fixed with 2.5% formaldehyde containing 0.1% Triton X-100. The cells were incubated with anti-NDV rabbit polyclonal serum or anti-influenza virus NS1 rabbit serum and then were washed and stained with fluorescein isothiocyanate-conjugated anti-rabbit immunoglobulins (DAKO). Viral protein expression was examined by fluorescence microscopy. Quantification of IFN production from infected cells. | To measure IFN production, cells were infected with rNDV or rNDV mutants at an MOI of 1. Supernatants from infected cells were harvested 24 h postinfection (p.i.) and then treated with UV light for 7 min to inactivate the infectivity of the virus. Serial 10-fold dilutions of UV-treated supernatants were added to fresh Vero cells or CEFs. The UV-treated supernatants were removed 20 h posttreatment, and then the cells were infected with rNDV-GFP at an MOI of 1. Green fluorescent protein (GFP) expression was analyzed by fluorescence microscopy 24 h p.i. The ability of the UV-treated supernatants to inhibit rNDV-GFP replication is indicative of the antiviral action of the IFN present in the sample. DNA fragmentation and Hoechst staining. | The extent of apoptosis was estimated by measuring the degree of DNA fragmentation. Briefly, cells were harvested and lysed with lysis buffer (10 mM EDTA, 50 mM Tris-HCl [pH 8.0], 0.5% sodium lauryl sarcosine) containing 100 mug of proteinase K/ml at 55C for 2 h. The DNA was extracted with phenol-chloroform and precipitated with ethanol. After RNA was removed by RNase treatment, the DNA was analyzed by 2% agarose gel electrophoresis to visualize the fragmented DNA. For observation of the morphologies of infected cell nuclei, the infected cells were fixed with methanol containing 10% formaldehyde and stained with the DNA-binding dye Hoechst 33258. Terminal deoxynucleotidyltransferase-mediated dUTP-fluorescein isothiocyanate nick end labeling (TUNEL) staining. | CEFs or Hep-2 cells were infected with wt rNDV or rNDV mutants at an MOI of 1. The cells were fixed at different time points with 4% paraformaldehyde for 1 h at room temperature and then permeabilized by adding 0.1% Triton X-100 in 0.1% sodium citrate on ice for 2 min. Terminal transferase was used to label free 3'OH ends in genomic DNA with fluorescein-dUTP as described in the manufacturer's manual (in situ cell death detection kit, fluorescein; Roche). The nicked fluorescence-labeled DNA was detected by fluorescence microscopy. RESULTS : Generation of an NDV editing-defective mutant. | It was previously found that expression of the V protein of NDV prevents the activation of the IFN-alpha/beta response in plasmid-transfected cells . In order to determine whether the NDV V protein also mediates the evasion of the IFN response during NDV infection, we generated an rNDV defective in V protein expression by inactivating the RNA-editing signal responsible for V mRNA synthesis [rNDV V(-)2ch] . It has been shown that three P gene-derived proteins with identical N-terminal sequences, P, V, and W, are produced in wt NDV-infected cells at frequencies of similar68, 29, and 2%, respectively . However, the level of V protein expression from an NDV editing-defective mutant containing a single-nucleotide change in the RNA-editing locus was <2% . In spite of repeated egg passages, NDV V(-)2ch did not grow to detectable hemagglutinin titers, but virus could be detected by IFA when allantoic fluids were used to infect cultured cells (not shown). These data indicated that this mutant NDV was compromised for replication in eggs. In order to further analyze the replication properties of rNDV V(-)2ch, we conducted multicycle growth kinetics studies in CEFs and Vero, Hep-2, and A549 cells, as well as in 6- and 10-day-old embryonated chicken eggs. Supernatants from infected cell cultures and the allantoic fluid from embryonated eggs were harvested every 24 h for 3 days. NDV titers were determined by measuring the TCID50 in Vero cells. The titers of rNDV V(-)2ch grown in CEFs and in 10-day old embryonated chicken eggs were 1,000- and 10,000-fold lower, respectively, than those of wt rNDV. However, in Vero cells, which do not produce IFN , the titer of rNDV V(-)2ch was only 20-fold lower than of that of wt rNDV . Replication of rNDV V(-)2ch was also more efficient in 6-day-old embryonated chicken eggs, which are known to be IFN deficient , than in 10-day-old embryonated chicken eggs . Interestingly, both wt rNDV and rNDV V(-)2ch replicated poorly in human Hep-2 and A549 cells, reaching similar500-fold-lower titers than wt rNDV in CEFs . To demonstrate that the replication-defective phenotype of rNDV V(-)2ch was due to impaired V protein expression, we reinserted the V gene as an additional transcriptional unit between the HN and L genes and generated rNDV V(-)/V . This insertion improved the replication of the virus to levels comparable to those of wt rNDV in both embryonated chicken eggs and cell cultures, except for a delayed growth pattern . Generation of an NDV mutant expressing a truncated V protein lacking its C-terminal domain. | It has been demonstrated that plasmid-mediated expression of the C-terminal domain of NDV V inhibits the IFN-alpha/beta response . To selectively block expression of the unique C terminus of the V protein in an infectious NDV, an rNDV mutant expressing a truncated V protein lacking its C-terminal domain [rNDV V(-)stop] was generated. One nucleotide was changed in the P gene to create a premature stop codon in V without affecting the P ORF . A similar mutant had been shown to grow to titers even lower than those of an editing-defective mutant NDV . In our studies, maximal titers of rNDV V(-)stop in 6- or 10-day-old embryonated chicken eggs were also similar104-fold lower than those of rNDV V(-)2ch . rNDV V(-)stop did not grow to measurable titers in CEFs (data not shown). These results indicate that the C-terminal domain of the NDV V protein plays a critical role in viral replication. Expression of the influenza A virus NS1 protein compensates for the lack of the V C-terminal domain in infectious NDV. | In order to study the contribution of the IFN antagonist activity of the C-terminal domain of V in NDV replication, we generated an rNDV expressing the IFN antagonist NS1 protein of influenza A virus in the backbone of rNDV V(-)stop. The NS1 ORF was inserted into pNDV V(-)stop between the HN and L genes as an extra transcriptional unit . Expression of the influenza virus NS1 protein from rNDV V(-)/NS1-infected CEFs was detected using anti-influenza virus NS1 rabbit polyclonal antibody . Interestingly, the insertion of the influenza A virus NS1 gene into the rNDV V(-)stop genome restored viral replication in both embryonated chicken eggs and CEFs to wt rNDV levels . These data indicate that the influenza A virus NS1 protein complements the loss of function of the C-terminal domain of the NDV V protein in the context of infectious viruses, suggesting that the NDV V and the influenza NS1 proteins are functionally interchangeable IFN-alpha/beta antagonists in chicken cells. Expression of the influenza A virus NS1 protein enhances replication of NDV in human cells. | The growth properties of rNDV V(-)/NS1 were also analyzed in human cell lines. Hep-2 or A549 cells were infected at an MOI of 0.01, and viral titers in the supernatant were determined at different times p.i. Interestingly, titers of rNDV V(-)/NS1 were similar10-fold higher than those of rNDV, rNDV V(-)2ch, or rNDV V(-)/V in both human cell lines . These results suggest that the influenza A virus NS1 protein prevents the cellular antiviral response more efficiently than the NDV V protein in human cells, resulting in increased viral replication. Therefore, the host range of NDV for avian species might be partly explained by the IFN-alpha/beta antagonist activity of its V protein. NDV V prevents IFN production in CEFs but not in human Hep-2 cells during viral infection. | In order to investigate whether the impaired replication of rNDV mutants correlated with higher levels of IFN induced in infected cell monolayers, we measured the IFN present in the supernatants of infected cells using a bioassay, as described in Materials and Methods. In these experiments, we did not use rNDV V(-)stop because this virus could not be grown to the titers required to achieve an MOI of 1. Hep-2 cells and CEFs were infected with wt rNDV or rNDV mutants at an MOI of 1. Supernatants were collected 24 h p.i., UV treated, and used to pretreat Vero cells or CEFs for 20 h before challenging them with NDV-GFP. Bioactive IFN present in the samples was monitored based on its ability to prevent NDV-GFP replication, as measured by GFP expression. As expected, no IFN was detected in IFN-deficient Vero cells infected with any of the rNDVs. Thus, when supernatants of infected Vero cells were added to new Vero cells, viral replication of NDV-GFP was not inhibited. In contrast, when Vero cells were treated with human IFN-beta, NDV-GFP replication was completely abolished (data not shown). Similarly, pretreatment of Vero cells with supernatant from wt rNDV, rNDV(-)2ch, or rNDV V(-)/V-infected Hep-2 cells induced an antiviral state that prevented NDV-GFP replication, indicating that all three viruses induced similar levels of IFN secretion in human Hep-2 cells . Interestingly, the levels of IFN present in supernatants from rNDV V(-)/NS1-infected Hep-2 cells was decreased similar10-fold. To monitor levels of IFN secreted by infected CEFs, we replaced Vero cells with the homologous CEFs as the indicator cell line, since IFN from one species (chicken) usually does not recognize the receptor of a distantly related species (monkey). When supernatants from rNDV V(-)2ch-infected CEFs were used to pretreat new CEFs, NDV-GFP replication was substantially inhibited. However, preteatment of CEFs with supernatants from CEFs infected with wt rNDV, rNDV V(-)/V, or rNDV V(-)/NS1 did not inhibit NDV-GFP replication . These data correlate with the replication properties of rNDV mutants in CEFs. They also correlate with the observed levels of activation of an IFN-responsive promoter following infection with the different viruses (data not shown). Taken together, our results indicate that the NS1 protein of influenza A virus inhibits IFN production in NDV-infected CEFs and Hep-2 cells, while the NDV V protein prevents IFN production only in NDV-infected CEFs but not in human Hep-2 cells. Delayed apoptosis in NDV-infected chicken cells but not in human cells. | When CEFs were infected with the different rNDVs at an MOI of 1, we observed that infection with rNDV V(-)2ch showed increased CPE compared to CEFs infected with wt rNDV, rNDV V(-)/V, or rNDV V(-)/NS1 at 36 h p.i. (Fig. , top). In contrast, in Hep-2 cells, the degrees of CPE were similarly high upon infection with wt rNDV-, rNDV V(-)2ch-, or rNDV V(-)/V-infected Hep-2 cells at 36 h p.i. rNDV V(-)/NS1 was exceptional in Hep-2 cells and induced low levels of CPE (Fig. , top). wt rNDV-, rNDV V(-)2ch-, or rNDV V(-)/V-infected Hep-2 cells were detached from the monolayers and had the appearance of apoptotic cells, while rNDV V(-)/NS1-infected and mock-infected Hep-2 cells had similar appearance with only a few detached cells. To determine whether the increased CPE seen in NDV-infected cells was indicative of the induction of apoptosis, Hoechst 33258 staining was performed to detect chromatin condensation, which is one of the hallmarks of apoptosis. The majority of CEFs infected with rNDV V(-)2ch showed chromatin condensation at 36 h p.i. in this assay. In contrast, low levels of chromatin condensation, similar to those in mock-infected CEFs, were observed in wt rNDV-, rNDV V(-)/V-, or rNDV V(-)/NS1-infected CEFs (Fig. , bottom). On the other hand, human Hep-2 cells infected with the same rNDVs showed high levels of chromatin condensation indicative of apoptosis, with the exception of rNDV V(-)/NS1-infected cells (Fig. , bottom). The presence of fragmented chromosomal DNAs was also analyzed in infected CEFs or Hep-2 cells. Cells were harvested 36 h p.i., and DNA was purified and analyzed for fragmentation as described in Materials and Methods. Fragmented DNA was not detected in rNDV-infected CEFs at that time. However, infection of CEFs by rNDV V(-)2ch resulted in clear DNA fragmentation . DNA fragmentation was also apparent in Hep-2 cells infected by wt rNDV, rNDV V(-)2ch, or rNDV V(-)/V, but not in rNDV V(-)/NS1-infected Hep-2 cells . To confirm the chromatin condensation and DNA fragmentation results, TUNEL assays were performed. Virus-infected CEFs and Hep-2 cells were TUNEL stained at 24 and 48 h p.i. to detect the presence of nicked chromosomal DNA, as described in Materials and Methods. As shown in Fig. , TUNEL-positive CEFs were apparent only after rNDV V(-)2ch infection, with levels increasing from 24 to 48 h p.i. By contrast, TUNEL-positive human Hep-2 cells were detected following infection with wt rNDV, rNDV V(-)2ch, or rNDV V(-)/V, while only very low levels of nicked-DNA staining were observed in rNDV V(-)/NS1-infected Hep-2 cells and only at 48 h p.i. Taken together, these data indicate that V expression delays apoptosis in NDV-infected CEFs but not in human Hep-2 cells. Interestingly, expression of the influenza A virus NS1 protein by NDV results in delayed apoptosis even in Hep-2 cells. DISCUSSION : It has been demonstrated that the V protein of NDV has IFN antagonist activity and that this activity is associated with the C terminus of the protein . This finding was based on experiments in which plasmids expressing the entire ORF of V or the C-terminal domain of V were transfected into cells, and the replication of an IFN-sensitive virus was used as a readout to measure anti-IFN activity. Here, we have used reverse genetics to generate NDV mutants which are defective in the expression of V in order to study the activity of the V protein in the context of infectious viruses. The editing-defective mutant NDV V(-)2ch grows in IFN-competent cells to titers which are 3 to 4 log units lower than those achieved by the wt rNDV . In Vero cells, which are defective in IFN synthesis, the growth difference between rNDV V(-)2ch and rNDV is less pronounced. This finding and the fact that an isogenic recombinant NDV expressing the full-length V as a separate transcriptional unit has wt-like growth characteristics confirm the idea that the V protein provides NDV with the ability to grow in IFN-competent CEFs. More importantly, it was found that the IFN antagonist of influenza A viruses (NS1) could functionally replace the V protein. In an NDV backbone which was unable to express the C terminus of the V protein, expression of the NS1 protein resulted in a virus [rNDV V(-)/NS1] indistinguishable from wt rNDV with respect to growth in CEFs. Furthermore, this virus showed an increased (1.5-log-unit) replication titer in human cells over that of wt rNDV . This finding suggests that replacing the V protein with the influenza virus NS1 protein enhances the ability of NDV to grow in human cells. Although rNDV V(-)/NS1 grows to higher titers in Hep-2 and A549 cells than wt NDV, the absolute titers of the virus are lower than those obtained in CEFs (avian cells represent the natural host of NDV). We showed earlier that the IFN antagonist NS1 may influence the replication properties of a virus. Exchanging the NS1 gene in a mouse-adapted influenza virus with the NS1 gene of the 1918 human pandemic influenza virus resulted in a dramatic decrease in replication in the mouse, suggesting a species-specific property of the IFN antagonist NS1 . It has also been reported that the SV5 V protein blocks IFN signaling by causing the proteasomal degradation of the transcription factor STAT1 in human cells but not in mouse cells . Similarly, respiratory syncytial virus was shown to exhibit host range restrictions mediated by viral IFN antagonist proteins . In addition, intracellular STAT2 protein can act as a host range determinant of paramyxoviruses . Interestingly, Maeno et al. reported that superinfection of HeLa cells with Sendai virus and NDV resulted in the complementation of the growth of NDV. In retrospect, we suggest that the enhanced yield of NDV in the superinfected cultures is the result of the expression in human cells of functional IFN antagonists by Sendai virus. By this mechanism, the host restriction of NDV in human cells was alleviated. The experiments reported in the present paper show that direct expression of the influenza virus NS1 protein in a recombinant NDV leads to a similar phenotype. Additional evidence for the mechanism involving IFN comes from direct measurements of IFN secreted from Hep-2 cells infected with these recombinant viruses. Only the NS1-expressing virus [rNDV V(-)/NS1] in Hep-2 cells appears to suppress the induction of human IFN. Hep-2 cells infected by other recombinant viruses, including the wt rNDV, induce an IFN response which can be measured by blocking virus replication in Vero cells . As expected, in chicken cells, both the V protein- and the NS1 protein-expressing viruses suppress the synthesis of IFN , again demonstrating the species specificity of these viral IFN antagonists. The ability of NDV V to suppress IFN production in CEFs might be due to an inhibition of IFN regulatory factors (IRFs), as reported for the V proteins of SV5 and Sendai virus , or to an inhibition of STAT factors involved in a positive feedback of IFN synthesis . In this respect, the V protein of the paramyxoviruses SV5, hPIV2, and SV41 and mumps and Nipah viruses are known to inhibit STAT function . Compared to rNDV, rNDV V(-)2ch exhibited an increased CPE in chicken cells. We suggest that this effect is related to an antiapoptotic activity of the viral V protein. Three assays were used to further characterize this finding. Hoechst 33258 staining showed condensation of nuclear chromatin in rNDV V(-)2ch-infected chicken cells. DNA fragmentation analysis and TUNEL assays confirmed the differences found in cells infected by the NDV mutant defective in V compared to those infected by viruses expressing V or NS1 . In human cells, however, infection by rNDV V(-)2ch is very similar to that by viruses expressing V protein with respect to chromatin condensation, DNA fragmentation, and TUNEL assays. Only the NDV recombinant expressing the NS1 protein appears to cause CPE without apoptotic changes . In our experiments using the NDV Hitchner B1 strain, we did not observe a significant level of apoptotic markers in infected CEFs. This appears to be inconsistent with an earlier report by Lam, who found apoptosis in chicken embryo cells 12 h p.i. by the GB strain of NDV . This strain of NDV is highly virulent, and thus, cells may undergo a stress-related response to infection. In accord with our results, NDV has been reported to be an excellent inducer of IFN-alpha and of tumor necrosis factor alpha in both mouse and human macrophages and lymphocytes . Also, the expression of TRAIL (tumor necrosis factor-related apoptosis-inducing ligand) (and thus of apoptosis) was induced by NDV in human monocytes . Several reports have demonstrated that IFN-alpha/beta and IFN-gamma induce apoptosis in human tumor cells . NDV infection of human tumor cells can lead to the upregulation of HLA, cell adhesion molecules, and chemokines and, finally, to apoptosis . It is not clear whether the antiapoptotic activity of the NDV V protein is mediated by its IFN antagonist activity. Further experiments are required to clarify this issue. Interestingly, mutations in the V gene of SV5 can also result in viruses that induce both IFN and cell death . In terms of the antiapoptotic activity of the NS1 protein expressed by infectious NDV, previous experiments involving infectious influenza virus expressing a full-length NS1 protein support this notion . However, there is one report which suggests that apoptosis is stimulated by the influenza virus NS1 protein when it is expressed by plasmids in the absence of infectious virus . Adaptation of a virus to its host is a complex phenomenon and is the result of a balancing act involving direct interactions of virus and host components. With respect to NDV, several features have been shown to affect the growth and virulence of the virus in its host. Specifically, the cleavage site of the NDV F protein has been found to be an important factor contributing to the pathogenesis of the virus . We suggest here that the V protein is an additional virulence factor of NDV that affects the IFN and apoptosis responses of the infected host. Furthermore, we suggest that the V protein affects the host range of the virus via its species-specific IFN antagonist activity. FIG. 1. : Schematic representation of rNDV mutants. Schematic representation of rNDV mutants. The P and V mRNAs expressed from wt NDV are indicated at the top. The P and V mRNAs are identical except for one nontemplated G nucleotide in the V mRNA. The amino acids encoded by the indicated nucleotide sequences are also shown. Amino acids in italics show coding differences in the P and V proteins due to mRNA editing. Mutations introduced into rNDV V(-)2ch are indicated by a solid triangle. The mutation introduced into rNDV V(-)stop is also indicated by a circled "stop." rNDV V(-)2ch is deficient in expression of V and W proteins due to an impairment in RNA editing. To selectively block the expression of the carboxy terminus of the V protein, a stop codon was created in the V ORF without affecting the P frame in rNDV V(-)stop. The NDV V or the influenza virus (Flu) PR8 NS1 gene was inserted between the HN and L genes in the context of rNDV V(-)2ch or rNDV V(-)stop, respectively, to generate rNDV V(-)/V and rNDV V(-)/NS1. For details, see Materials and Methods. FIG. 2. : Growth kinetics of rNDV mutants. Growth kinetics of rNDV mutants. CEFs and Vero, Hep-2, and A549 cells were infected with each indicated virus at an MOI of 0.01 in the presence of 10% allantoic fluid. In addition, 10-day-old embryonated chicken eggs were inoculated with 100 TCID50/egg. Each supernatant from infected cells and embryonated eggs was harvested every 24 h for 3 days. Titers were determined by measuring the TCID50 in Vero cells. Note that the mutations responsible for the loss of the V function in rNDV V(-)/V and rNDV V(-)/NS1 are due to the insertion of an editing mutation and a stop codon, respectively. FIG. 3. : Expression of the influenza A virus NS1 protein by rNDV V(-)/NS1. Expression of the influenza A virus NS1 protein by rNDV V(-)/NS1. Expression of the influenza virus NS1 protein or of NDV viral proteins was detected in infected CEFs by immunofluorescence using anti-rabbit influenza virus NS1 polyclonal antibody or anti-rabbit NDV polyclonal antibody, respectively. Magnification, x10. FIG. 4. : IFN production by rNDV mutant-infected cells. IFN production by rNDV mutant-infected cells. To measure IFN production after viral infection, cells were infected with rNDV or rNDV mutants at an MOI of 1. Supernatants (Sup) from infected cells were harvested 24 h p.i. and then treated with UV light for 7 min to inactivate the infectivity of the virus. UV-treated supernatants were added to new Vero cells or CEFs. The UV-treated supernatants were removed 20 h posttreatment, and then the cells were infected with rNDV-GFP at an MOI of 1. NDV-GFP expression was analyzed by fluorescence microscopy 24 h p.i. Note that the mutations responsible for the loss of the V function in rNDV V(-)/V and rNDV V(-)/NS1 are due to the insertion of an editing mutation and a stop codon, respectively. FIG.5 : Induction of apoptosis by rNDV mutants. Induction of apoptosis by rNDV mutants. CEFs and Hep-2 cells were infected with wt rNDV or rNDV mutants at an MOI of 1 or mock infected. (A and B) The morphologies of infected cells at 36 h p.i. were analyzed by phase-contrast microscopy (CPE; magnification, 310). To observe chromatin condensation in cell nuclei, the infected cells at 36 h p.i. were fixed with methanol containing 10% formaldehyde and stained with the DNA-binding dye Hoechst 33258 (magnification, 340). (C) DNA fragmentation was analyzed 36 h p.i. Lane 1, uninfected cells; lane 2, wt rNDV-infected cells (MOI 5 1); lane 3, rNDV V(2)2ch-infected cells (MOI 5 1); lane 4, rNDV V(2)/V-infected cells (MOI 5 1); lane 5, rNDV V(2)/NS1-infected cells (MOI 5 1); L, 1-kb DNA ladder. (D) TUNEL staining of rNDV-infected CEFs and Hep-2 cells at 24 and 48 h p.i. Note that the mutations responsible for the loss of the V function in rNDV V(2)/V and rNDV V(2)/NS1 are due to the insertion of an editing mutation and a stop codon, respectively. ---Continued. TABLE 1 : Titers of rNDV V(-) mutants in embryonated chicken eggs Backmatter: PMID- 12915551 TI - Interferon Regulatory Factor 7 Regulates Expression of Epstein-Barr Virus Latent Membrane Protein 1: a Regulatory Circuit AB - We have shown previously that interferon regulatory factor 7 (IRF7), a multifunctional protein intimately involved in latent Epstein-Barr virus (EBV) infection, is induced as well as activated by EBV latent membrane protein 1 (LMP1), the principal EBV oncoprotein. Since the LMP1 promoter (LMP1p) contains an interferon-stimulated response element (ISRE), we hypothesized that IRF7 might be able to regulate LMP1 expression and thus participate in a regulatory circuit between these two genes. In this study, IRF7 was shown first to activate LMP1p in transient transfection assays. Compared with EBV nuclear antigen 2 (EBNA2), the most potent viral transactivator of LMP1p, IRF7 has a lesser effect (approximately 10% that of EBNA2) on induction of LMP1p. Study with IRF7 deletion mutants showed that IRF7 functional domains have similar effects on both the beta interferon (IFN-beta) and LMP1 promoters in BJAB and 293 cells, and study with IRF7 phosphomimetic mutants showed that IRF7 phosphorylation may be involved in the activation of these two promoters. Further, the ISRE in LMP1p responds to IRF7 induction and IRF7 binds to this element. In the EBV-positive cell line P3HR1, which lacks the complete EBNA2 and EBV-encoded leader protein genes and hence expresses low-level LMP1, IRF7 alone can notably increase the endogenous LMP1 mRNA and protein levels. These results indicate that LMP1 is regulated by this host cell gene in addition to the viral factor, EBNA2, and may help to explain how LMP1 is expressed in type II latency in the absence of EBNA2. Moreover, IRF7 can regulate a viral gene in addition to a host cellular gene such as the IFN-beta gene. Together with the previous data that LMP1 can induce IRF7 expression and facilitate IRF7 phosphorylation and nuclear translocation, these results suggest a positive regulatory circuit between IRF7 and LMP1. Keywords: Introduction : Interferons (IFNs) are a large family of multifunctional secreted proteins involved in antiviral defense, cell growth regulation, and immune activation . The IFN regulatory factors (IRFs) are a growing family of virus- and IFN-inducible cellular proteins that now consists of 10 members, including the recently discovered IRF-10 as well as four Kaposi's sarcoma-associated herpesvirus (KSHV)-encoded viral IRFs . A hallmark of IRF proteins (but not including viral IRFs) is an N-terminal DNA-binding domain (DBD), which contains a five-tryptophan-residue repeat. This repeat forms a helix-turn-helix motif determining the characteristic DNA-binding site, 5'-(A/G)NGAAANNGAAACT-3'. Distinct properties and functions of IRF family members depend on their different C-terminal sequences and structures . IRFs are critical in the regulation of expression of type I IFNs (IFN-alpha and IFN-beta), IFN-stimulated genes, and other cytokines and chemokines; three of them, IRF3, IRF5, and IRF7, are direct inducers of virus-mediated signaling . IRF7 was cloned and identified within the biologic context of Epstein-Barr virus (EBV) latency and is expressed predominately in spleen, thymus, and peripheral blood leukocytes. IRF7 has been implicated in IFN gene activation in response to virus infection . In addition to IFNs, IRF7 expression is also stimulated by other factors such as sodium butyrate , lipopolysaccharide, topoisomerase II inhibitors, 12-O-tetra-decanoylphorbol-13-acetate (TPA), and tumor necrosis factor alpha (TNF-alpha) . Cellular factors, including NF-kappaB, are involved in the induction of IRF7 by TNF-alpha . Interestingly, IRF7 also can be induced by the principal EBV oncoprotein, latent membrane protein 1 (LMP1) . IRF7 appears to be a multifunctional protein involved in multiple physiological processes during viral infection, including regulation of the host immune system, IFN gene expression, and EBV latency, as well as acting as a putative oncogene (L. Zhang, C. Der, and J. S. Pagano, unpublished results). IRF7 has also been reported to participate in monocyte differentiation . EBV LMP1 is an integral membrane protein with six transmembrane N-terminal domains and a long C-terminal cytoplasmic tail. LMP1, as a key EBV oncogenic factor, is a constitutively active receptor-like molecule that does not need the binding of a ligand. Expression of LMP1 is essential for growth transformation of B lymphocytes . Several studies have demonstrated that efficient LMP1 promoter activation is dependent upon virus-specific factors; some host cellular factors are also involved in these processes (, , , -, ). EBV nuclear antigen 2 (EBNA2) is the best-characterized viral factor that transactivates the LMP1 promoter (LMP1p) . Also, EBV-encoded leader protein (EBNA-LP) can activate the LMP1p, especially in cooperation with EBNA2 . In addition to LMP1p, it is well established that EBNA2 can transactivate a number of other genes, including the cellular CD21, CD23, c-Bcl-2, and c-fgr and EBV LMP2A and LMP2B genes and the EBNA-1 C promoter . EBNA2 does not bind directly to DNA but exerts its function by interacting with the cellular protein, recombination signal-binding protein (RBP-Jkappa), and PU.1 protein; other cellular transcriptional factors mediate the activation of LMP1 by EBNA2 . The latency-associated nuclear antigen encoded by KSHV can also activate LMP1p . LMP1 stimulates the expression of IRF7 , ICAM-1, LFA-3, CD40, EBI-3, the transporter associated with antigen processing 2 (Tap-2), Fas, and TNF receptor-associated factor 1 in EBV-negative Burkitt's lymphoma (BL) cells and of the epidermal growth receptor in epithelial cells (, -, , ). LMP1 also induces or enhances expression of matrix metalloproteinase 1 , matrix metalloproteinase 9 , Bcl-2 , cyclooxygenase-2, and vascular endothelial growth factor as well as fibroblast growth factor 2 and hypoxia-inducible factor 1alpha (N. S. Wakisaka and J. S. Pagano, unpublished results), all of which are associated with tumor invasion and metastasis. Therefore, LMP1 is a key mediator for host viral defense as well as oncogenesis. Because of a conserved DBD in the N-terminal region of all members of the IRF family, IRFs can bind to consensus or similar sequences: IFN-stimulated response element (ISRE), positive regulatory element, and IFN consensus sequence, all of which have conserved GAAA repeats. LMP1 has an ISRE within its promoter . In this paper, we show that IRF7 can bind to and regulate the LMP1 promoter and that IRF7 can stimulate expression of LMP1 mRNA and protein independently of as well as cooperatively with EBNA2. MATERIALS AND METHODS : Plasmids. | The LMP1 promoter construct pGL2(-512/+72)-luciferase plasmid, which has two directly repeated copies of the LMP1 -512/+72 promoter element, was provided by Jeffery Lin and Elliott Kieff . The LMP1 promoter construct pgLRS(-259)-CAT and its mutations, pgLRS(-259)-ISRE--CAT, pgLRS(-259)-Jkappa--CAT, and pgLRS(-259)-ISRE-/Jkappa--CAT, were gifts from Anna Sjoblom and Lars Rymo . The wild-type IRF7A expression plasmid pcDNA3/IRF7A was constructed in this laboratory . The IFN-beta promoter (IFNbetap) construct (pGL3/IFNbetap-Luc) and a series of pCMV2-Flag plasmids containing wild-type IRF7A and its deletion or substitution mutations were described previously and were gifts from Rongtuan Lin and John Hiscott. The pBS-LMP1 plasmid for RNase protection assays (RPA) was provided by Paul Farrell . The EBNA2-expressing plasmid construct was made by inserting a 2.1-kb MluI-AhaIII restriction fragment containing the EBNA2b coding region from the AG876 virus into the pGEM derivative pHD 101-3 . IRF7-DN, which lacks the DBD of the human IRF7A gene, was provided by Tom Maniatis. Cell lines. | BJAB is an EBV-negative human BL cell line. P3HR-1, an EBV-positive cell line from which portions of the EBNA-LP and EBNA2 open reading frames have been deleted, expresses low levels of LMP1. Both BJAB and P3HR1 cells were grown in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS) and penicillin-streptomycin (5 U/ml and 5 mug/ml, respectively). The 293 cell line (derived from human kidney epithelial cells) was grown in Dulbecco's modified Eagle's medium with 10% FBS and penicillin-streptomycin (5 U/ml and 5 mug/ml, respectively). Transfections. | BJAB and P3HR-1 cells were transfected with an electroporator at 210 V and 975 muF with 5-mug quantities of the various promoter constructs or vector and increasing concentrations of the IRF7 clones in 0.5 ml of RPMI 1640 medium containing 10% FBS. Vector DNA was added to equalize the total amount of DNA used in all transfections. After electroporation, cells were resuspended in 10 ml of complete medium and incubated for 48 h. Cells were harvested, washed twice with phosphate-buffered saline, and lysed in appropriate buffer. Reporter assays. | For luciferase assays, the transfected cells were collected at 48 h posttransfection and resuspended in 200 mul of 1x reporter lysis buffer (Promega). The cells were placed in a dry ice-isopropanol bath for 2 min, thawed at 37C, vortexed, and then centrifuged at 16,000 x g for 20 s. Cell lysates (20 mul each) were combined with luciferase assay reagent (Promega), and the relative light units (RLU) were measured in an Lmax luminometer (Molecular Devices Corp.). The pcDNA3 empty vector alone was used in these assays to determine the ability of IRF7 to activate the vector alone, and the transfection efficiencies were normalized by beta-galactosidase (beta-Gal) values. Chloramphenicol acetyltransferase (CAT) assays were performed as described previously ; results were analyzed by use of ImageQuant software (version 5; Molecular Dynamics). All reporter assay results presented are from single experiments representative of multiple independent trials or are averages derived from multiple independent repetitions. Western blotting with enhanced chemiluminescence detection. | Cells were collected and washed with 5 ml of phosphate-buffered saline. The pellets were resuspended in 200 mul of cell lysis buffer and mixed with 200 mul of 2x sodium dodecyl sulfate loading buffer. Lysates were separated on sodium dodecyl sulfate-10% polyacrylamide electrophoresis gels for detection of IRF7 and LMP1 and 8% gels for EBNA2. The proteins were then transferred to nitrocellulose membranes. The membranes were blocked in 5% milk for 1 h at room temperature (RT), rinsed in Tris-buffered saline with Tween (TBST) three times for 5 min each time, and then incubated with the specific primary antibody at RT for 1 h or at 4C overnight. After three rinses for 5 min each rinse in TBST, the membranes were incubated with appropriate horseradish peroxidase-conjugated secondary antibodies (Amersham Biosciences) (1:3,000) for 1 h at RT and subsequently washed with TBST three times for 5 min each time. Specific signals were detected by enhanced chemiluminescence (ECL) following the manufacturer's instruction (Amersham Pharmacia Biotech). IRF7 was detected with rabbit polyclonal antibody (Santa Cruz) (1:300), and EBNA2, LMP1, and Flag were detected with mouse monoclonal antibodies PE2 (DAKO) (1:500), CS1-4 (DAKO) (1:100), and M2 (Sigma) (1:3,000), respectively. EMSA. | BJAB and 293 cells in 100-mm-diameter dishes were transfected with 5 mug of Flag-tagged IRF7A or its mutants. Cells were collected and lysed in binding buffer (10 mM Tris-HCl [pH 7.5], 1 mM EDTA, 50 mM NaCl, 2 mM dithiothreitol [DTT], 5% glycerol, 0.5% NP-40, 10 mug of bovine serum albumin [BSA]/mul) with protease inhibitors at 48 h posttransfection. The whole-cell lysates were used for electrophoretic mobility shift assays (EMSA), as described previously . Poly(dI-dC) (Amersham) at a final concentration of 62.5 mug/ml was added to reduce nonspecific binding. The double-stranded sequence containing LMP1 ISRE is GATCCAACAGGAAATGGAAAGGCAGTG. Another sequence containing mutated LMP1 ISRE is GATCCAACAGGAggTGGAggGGCAGTG (the mutated nucleotides are indicated by lowercase letters). These double-stranded sequences with 5'-GATC adhesive ends were labeled (using Klenow fragments) with [alpha-32P]dCTP. For competitor assays, 100-fold excess cold probe or AP-1 sequence (Promega) was added to the binding mixture. For supershift assays, whole-cell lysates were incubated with 0.2 mug of Flag antibody M2 (Sigma) before probe was added. Protein-DNA complexes were separated on 5% 60:1 acrylamide gels. RNA isolation and RPA. | P3HR1 cells were transfected with different amounts of IRF7 and EBNA2 expression plasmids together with CD4 expression plasmid. Using a QIAgen RNeasy mini kit, total RNAs were isolated from CD4-conjugated magnetic bead (Dynal)-selected cells. RPA were performed with total RNA using an RNase protection kit II (Ambion). The hybridization temperature was 37C. The human GAPDH probe was supplied by US Biochemicals Inc. The LMP1 probe for RPA, which was from the pBS-LMP1 construct and labeled with [alpha-32P]UTP by in vitro transcription, corresponds to nucleotides 169033 to 169423 in the B95-8 EBV genome. The protected fragments were 220 and 90 bp . RESULTS : IRF7 upregulates the LMP1 promoter. | The positive regulators of the LMP1 promoter (LMP1p) identified to date center mostly on the EBNA2-responsive enhancer region to which RBP-Jkappa binds . Here, we examined whether another region of LMP1p that contains an ISRE would be responsive to induction by IRF7. BJAB, an EBV-negative human BL cell line, was used for this study because it has little endogenous IRF7. The assayed LMP1p construct, pLMP1(-512/+72), has two directly repeated copies of the LMP1 -512/+72 promoter element . Using luciferase assays, we examined the effect of the presence of IRF7 on the regulation of LMP1p after transient expression of IRF7. In cells transfected with IRF7A, LMP1p activity was increased about 10-fold compared with that produced by the vector pcDNA3 alone . The regulatory effects were clearly dose dependent, with increasing amounts of IRF7A resulting in elevated levels of luciferase activity . In this assay, EBNA2, which is known to transactivate LMP1p, was used as a positive control. Compared with IRF7, EBNA2 had a greater effect on transactivation of LMP1p (up to 100-fold with 5 mug of EBNA2 expression plasmid). However, the cotransfection of IRF7A and EBNA2 resulted in up to 200-fold activation of LMP1p; the combined effect was generally greater than the effect of IRF7 alone plus that of EBNA2 alone . Thus, LMP1p can respond to induction by IRF7 independently of the EBNA2 enhancer region, and the two can function together to activate the promoter. Western blot analyses show that exogenous IRF7 was expressed in the transfected BJAB cells and that a low level of endogenous IRF7 is detected in these cells (Fig. , first lane). To elaborate on the significance of the results obtained with BJAB cells, we tested the effects of IRF7 on LMP1p in another human cell line, 293. These cells contain no detectable endogenous IRF7 and EBNA2, and LMP1p constructs cannot be activated by endogenous factors in 293 cells (data not shown). Luciferase assays show that IRF7 can activate LMP1p in 293 cells. However, the effect is less striking than in BJAB cells. IRF7 consistently activates LMP1p in 293 cells at levels two- to threefold higher than those seen with the vector-only control . The effect is approximately threefold less than that obtained with BJAB cells. IRF7 functional domains and phosphorylation in LMP1p activation. | Functional domains of IRF7 protein have been defined based on effects of mutated IRF7 on the IFNA4, IFN-beta, and RANTES promoters . This study showed that the C terminus of IRF7 protein contains a constitutive activation domain (CAD), a signal response domain, a virus-inducible activity domain (VAD), and an inhibition domain (ID), as shown in Fig. . Also, a nuclear export sequence (NES) was identified between amino acids 416 and 467 which overlaps with the ID . To check whether these functional domains have similar functions for LMP1p, we transfected IRF7 domain deletion mutants IRF7(Delta247-467) (designated as IRF-DeltaID), IRF7(Delta151-278) (IRF7-DeltaCAD), IRF7(Delta247-372) (IRF7-DeltaVAD), IRF7(Delta151-467) (IRF7-DeltaCAD/ID), and IRF7(Delta416-467) (IRF7-DeltaNES), together with the LMP1p construct pLMP1 (-512/+72) or the IFNbetap construct pGL3/IFNbetap-Luc, in 293 cells. Interestingly, these IRF7 functional domains have similar functional effects on two different promoters, IFNbetap and LMP1p, in the same cell line, 293 . IRF-DeltaID has a higher level of activation ability than wild-type IRF7 for both IFNbetap and LMP1p, although the level of activation is much higher for IFNbetap than for LMP1p. Although deletion of the VAD (IRF7-DeltaVAD) impaired transactivation greatly, the VAD is not necessary for retention of IRF7 activity in the absence of virus infection, since IRF7-DeltaID (Delta247-467) without the VAD also has very high constitutive activity. In contrast, the CAD is necessary for constitutive IRF7 activity. All the other deletion mutants have lower activation capabilities than wild-type IRF7, and IRF7-DeltaNES almost abolished activation of LMP1p . The dominant-negative mutant, IRF7-DN, which lacks the DBD, cannot activate LMP1p , indicating that direct or indirect binding of IRF7 to LMP1p is necessary for LMP1p activation. Similar results were obtained with BJAB cells (data not shown). The sequence and phenotype of each of the IRF7A point mutants with serines replaced by phosphomimetic aspartic acids, IRF7(D483/487), IRF7(D475-479), and IRF7(D477/479), are shown in Fig. . These substitution mutants show that serines at different sites have different functional capacities in LMP1p induction. IRF7(D483/487) produced much less induction than wild-type IRF7A; however, IRF7(D477/479) has much higher inducing ability. Substitution of two more serines at serine-475 and serine-476 (D475-479) did not result in a higher level of activity. The results suggest that only phosphorylation of serine-477 and serine-479 is necessary for LMP1p activation in the absence of virus infection . Western blot analysis shows that IRF7 and its mutants are all expressed in the transfected cells . The bulk of each mutated protein was located in the nucleus, except for IRF7-DeltaID and IRF7-DeltaNES, both of which were located only in the nucleus with none of either detected in the cytoplasmic extracts (data not shown). The ISRE of LMP1p responds to IRF7. | The LMP1 regulatory sequence (LRS) is defined as nucleotides 169477 to 170151 of the B95-8 genome, which corresponds to -634 to +40 relative to the LMP1 transcription start site. Because IRF7 can bind to ISRE-like elements, we tested whether the ISRE element in LMP1p is responsive to IRF7. BJAB cells were transfected with pcDNA3/IRF7A expression plasmid together with pgLRS(-259)-CAT LMP1p constructs and its mutants. As expected, both pgLRS(-259)-CAT (wild type) and the pgLRS(-259)-RBP-Jkappa--CAT mutant have similar activities after IRF7 induction; however, both the pgLRS(-259)-ISRE--CAT mutant and the pgLRS(-259)-ISRE-/RBP-Jkappa--CAT mutant could not be activated by IRF7 . From these data, we can conclude that the ISRE, not the RBP-Jkappa binding site in LMP1p, is the target for IRF7. IRF7 binds to the ISRE of LMP1p. | Further, whether IRF7 binds to the ISRE of LMP1p was investigated by EMSA. BJAB and 293 cells were transfected with IRF7 or its mutants, and whole-cell lysates were prepared for EMSA after 48 h. The ISRE probe that contains LMP1p ISRE and its flanking sequences was synthesized and labeled by 32P. As indicated in Fig. , the Flag-tagged wild-type IRF7A and IRF7(D477/479) can bind to the ISRE plus flanking sequences of LMP1p (see details in Materials and Methods) specifically. However, binding of IRF7-DeltaID (Delta247-467), which can activate both LMP1p and IFNbetap in both BJAB and 293 cells, was not detected (Fig. , lane 7) but did produce a weak signal in the supershift assay (Fig. , lane 12). So it seems that this IRF7 mutant has very weak DNA-binding activity. Consistent with this result, this mutant could activate the IFNA4 promoter but it could not be detected by EMSA as binding to the PRDI-III sequences in this promoter . Recently, Yang et al. showed that this IRF7 mutant could activate transcription in insect cells but that it had very weak DNA-binding activity. Actually, some IRF7 protein-DNA complexes are not readily detected by EMSA . IRF7-DN, which lacks the DBD, cannot bind to the ISRE in LMP1p (Fig. , lane 9). When the ISRE sequence in the probe was mutated (see details in Materials and Methods), no specific band was detected with cell lysates from cells transfected with wild-type IRF7 or any IRF7 mutant (Fig. , right panel). Supershift assays with monoclonal Flag antibody confirmed that the protein-DNA complex contains IRF7 (Fig. , lanes 10 to 14). Exogenously expressed IRF7 induces expression of LMP1 in P3HR1 cells. | To determine whether the levels of endogenous LMP1 actually change when exogenous IRF7 is expressed, we transfected an EBV-positive cell line exhibiting low levels of LMP1 to determine whether increased expression of LMP1 would occur when the endogenous promoter was activated by IRF7. The P3HR1 genome is deleted from the EBNA2 gene, which transactivates LMP1p, and the cells express very low levels of LMP1. P3HR1 cells were transfected with increasing amounts of IRF7 and cotransfected with EBNA2 and IRF7A. The Western blotting results demonstrated that LMP1 protein levels increased with increasing amounts of transfected IRF7 . LMP1 could be barely detected in the vector-alone control, although endogenous IRF7 protein was detected in these P3HR1 cells (Fig. , lane 1). The endogenous IRF7 may be inactive and therefore does not detectably induce LMP1. Transiently transfected IRF7A or EBNA2 or cotransfection with IRF7 and EBNA2 each resulted in elevated LMP1 protein levels. These results suggest that IRF7A can upregulate LMP1 protein levels in the context of the EBV genomic, native promoter elements. To confirm this interesting result with RPA, we also assayed LMP1 RNA levels in P3HR1 cells after transfection. The RPA results show a pattern of increase in LMP1 RNA levels corresponding to that of LMP1 protein (shown in Fig. ). In the vector-alone control, endogenous LMP1 mRNA could not be detected in P3HR1 cells. With increasing amounts of IRF7 or EBNA2 or the combination of IRF7 and EBNA2, the LMP1 mRNA level is elevated gradually. By titrating in IRF7 with EBNA2, we could essentially reconstitute normal LMP1 levels as expressed in the parental JiJoye cell line . Thus, IRF7 can activate the endogenous LMP1 promoter independently or together with EBNA2 in cell culture and the results confirm those obtained with the promoter-reporter assays. DISCUSSION : Our results clearly show that IRF7 can activate the LMP1 promoter by binding to its ISRE and that IRF7 can directly induce LMP1 expression. Furthermore, the study with IRF7 deletion mutants showed that IRF7 functional domains have similar functions for two different promoters, IFNbetap and LMP1p, in both 293 and BJAB cells. Our results with IFNbetap confirm those of Lin et al. . From our results together with those from the study of Lin et al. , it can be concluded that IRF7 functional domains are constitutive and independent of its targets and cell lines. It has been reported that NF-kappaB plays an important role in the activation of IFNbetap independently of the IRF family , but no NF-kappaB binding site has been identified in LMP1p. It is unknown whether IRF7 can activate transcription without involvement of another transcription factor. IRF3 has been shown to associate with IRF7 in IFN gene activation . In this study, we used a constitutively active form of IRF7, IRF7(Delta247-467), which cannot dimerize with either IRF3 or IRF7. However, this mutant has very high activity on both the promoters, indicating that dimerization is not necessary for IRF7 transactivation ability, even though it has been reported that artificially induced IRF7 dimers can bind DNA and induce target genes independently of phosphorylation or other downstream effects of viral infection . IRF7 is modified by phosphorylation after exposure to various stimuli, including viral infection, double-stranded RNA , chemotherapeutic DNA-damaging drugs , and expression of specific viral gene products such as LMP1 . This phosphorylation is thought to regulate IRF7 activity through elements in the C terminus which dictate nuclear localization, dimer formation, and promoter targeting . Two groups reported different results in investigations of the identity of phosphorylated sites needed for IRF7 transactivation activity. Marie et al. showed that murine IRF7 activity is dependent on phosphorylation of serine-425/426, which correspond to human IRF7A serine-471/472 . Hiscott and colleagues showed that human IRF7A activity depends on the phosphorylation of serine-477/479. Our results at least support the latter, but we did not test for the effect of serine-471/472. Several studies have demonstrated that LMP1 is regulated by virus-specific factors, especially EBNA2 and EBNA-LP, mediated by some host cellular factors that tether to the LMP1 promoter (, , , -, , , ). Actually, EBNA2 is not the only viral activator for LMP1. In KSHV-infected cells as well as in cells coinfected with both KSHV and EBV, LMP1 is activated by the KSHV antigen latency-associated nuclear antigen in the absence of EBNA2 . To our knowledge, no report has shown that LMP1 is regulated directly by a host cellular factor. In this study, we demonstrated that LMP1 can be activated by cellular IRF7 without the involvement of any viral factor. This interesting finding may account for a puzzling phenomenon in EBV biology: LMP1 is expressed in type II latency but EBNA2 is absent in type II cells . Type II latency is exemplified by nasopharyngeal carcinoma (NPC), and in fact, Zhang and Pagano have detected high levels of expression of IRF7 in most NPC tissues tested as well as in LMP1-positive NPC lines passaged in nude mice (L. Zhang and J. S. Pagano, unpublished results). So the LMP1 levels detected in type II cells may be the consequence of IRF7 expressed in type II cells. Therefore, like EBNA2, which is the key viral antigen for maintaining type III latency, IRF7 may play an equally important role in maintaining type II latency through regulation of LMP1 expression. IRF7 functions as a multifunctional protein. First, IRF7 was cloned on the basis of its ability to bind to and repress EBNA-1 Qp in EBV type III latency . Second, IRF7, like other IRFs, plays a primary role in regulating the expression of type I IFN genes in virus-infected cells . Zhang and Pagano showed that IRF7 was involved in the activation of Tap-2 by LMP1 in B lymphoma cells ; Tap-2 mutations are involved in diseases associated with the immune system. Moreover, IRF7 has properties consistent with a putative oncogene (Zhang et al., unpublished). Here, we found that IRF7 can regulate expression of a viral gene in addition to host cellular genes such as IFNA genes. Expression of IRF7 mRNA and protein is clearly inducible by LMP1 in EBV-infected cells . Since LMP1 can activate NF-kappaB and since there is an NF-kappaB binding site in the IRF7 promoter, it is likely that IRF7 induction by LMP1 is mediated by NF-kappaB, at least in part. In fact, Zhang et al. have found that a suppression of IkappaB can block the induction of IRF7 by LMP1. Furthermore, overexpression of NF-kappaB can induce IRF7 expression in DG75, the EBV-negative BL cell line . However, the results of a promoter-reporter assay indicated that LMP1 could not activate IRF7 promoter constructs in HeLa cells, suggesting that IRF7 induction by LMP1 is cell type specific and that NF-kappaB activation needs cooperation with other transcription factors . In addition to induction of IRF7 expression, LMP1 also regulates IRF7 protein activity: LMP1 augments the phosphorylation status and facilitates the nuclear localization of IRF7 . Based on our results, a regulatory circuit between LMP1 and IRF7 is outlined : IRF7 activates the LMP1 promoter by binding to its ISRE, which results in LMP1 protein expression. In turn, the LMP1 expression can induce IRF7 expression and regulate its protein activity. In addition, IRF7 may be autoregulated through binding of IRF7 protein to the ISRE in the first intron or the IRF-E in the IRF7 promoter since IRF7 is capable of binding to these elements; preliminary results indicate that IRF7 can induce the IRF7 promoter in transient transfection assays (S. Ning and J. S. Pagano, unpublished results). P3HR1 cells are defective for EBNA2 expression ; therefore, little if any LMP1 is detected. The absence of LMP1 leads to low levels of IRF7. We show here that IRF7 can induce a low level of LMP1 in these cells. However, since EBNA2 is totally absent and since the IRF7 and EBNA2 inducing effects are independent and EBNA2 is the stronger inducer and normally necessary for LMP1 induction, the level of LMP1 remains subnormal in P3HR1 cells. Our results suggest that IRF7 potentiates the oncogenic effect of LMP1. Since LMP1 can induce IRF7, the net effect should be augmented expression of this EBV oncoprotein. In other work, Zhang et al. have found that IRF7 by itself has oncogenic properties in 3T3 cells, as determined by focus formation and colony formation as well as production of tumors, and that the combination of LMP1 and IRF7 produces at least additive effects (Zhang et al., unpublished). Whether IRF7 is essential for effects of LMP1 including oncogenesis remains to be determined. FIG. 1. : IRF7 upregulates the LMP1 promoter in BJAB cells. IRF7 upregulates the LMP1 promoter in BJAB cells. (A) Schematic illustration of the LMP1 promoter. Numbers indicate the nucleotide sites relative to the transcriptional start site. The ISRE sequence is shown. (B) Promoter activity was monitored by investigating luciferase activity. BJAB cells were transfected with pGL2(-512/+72)-luciferase plasmid (LMP1p) and IRF7 or with IRF7 plus EBNA2. Luciferase activity was analyzed at 48 h after transfection. RLU levels and severalfold activation relative to the basal level of reporter gene in the presence of the vector after normalization with beta-Gal activity were measured. Equal amounts of cell lysates were assayed for luciferase activity. Each data point represents the average of seven to nine repeats in five independent experiments. Error bars represent means +- standard errors (SE). (C) Western blot analysis showing that IRF7 was expressed in transfected BJAB cells. Monoclonal IRF7 antibody was used at a dilution of 1:300. A total of 100 mug of protein was loaded in each lane. Tubulin was used as loading control. FIG. 2. : Effect of IRF7 mutants on LMP1 and IFN-b promoters. Effect of IRF7 mutants on LMP1 and IFN-b promoters. (A) Schematic representation of the IRF7 mutants used in this study. The amino acid sequence from 468 to 491 is shown, serines replaced with phosphomimetic aspartic acids are in bold, and the sites are indicated as subscripts. SRD, signal response domain (20). (B) Comparison of the effects of IRF7 mutants on the IFN-b and LMP1 promoters in 293 cells. 293 cells were transfected with pGL3/IFNb-Luc or pGL2(2512/172)-Luc and a series of expression plasmids encoding IRF7 or IRF7 mutants as indicated. RLU levels and severalfold activation relative to the basal level of reporter gene in the presence of the vector after normalization with b-Gal activity were measured. Equal amounts of cell lysates were assayed for luciferase activity. Each data point represents the average of seven to nine repeats in five independent experiments. Error bars represent means 6 SE. (C) Western blot analysis of transfected IRF7 mutants. Monoclonal Flag antibody M2 was used at a dilution of 1:3,000. Lanes: 1, pCMV2-Flag; 2, IRF7A; 3, IRF7-DID; 4, IRF7-DCAD; 5, IRF7-DVAD; 6, IRF7-DCAD/ID; 7, IRF7-DNES; 8, IRF7(D475-479); 9, IRF7(D483/487); 10, IRF7(D477/479). FIG. 3. : The ISRE in LMP1p responds to IRF7. The ISRE in LMP1p responds to IRF7. IRF7 and pLRS(-259)-CAT or its mutants as indicated were transfected into BJAB cells. Transfected cells were collected after 48 h, and CAT assays were performed . Equal amounts of cell lysates were used. The CAT signals were quantitated by Molecular Dynamics PhosphorImager and normalized with beta-Gal activities. Each data point represents the average of eight repeats in two independent experiments. Error bars represent means +- SE. FIG. 4. : IRF7 binds to the ISRE in LMP1p. IRF7 binds to the ISRE in LMP1p. BJAB cells were transfected with 5 mug of Flag-tagged IRF7A or its mutants as indicated. Cells were collected and lysed in binding buffer (10 mM Tris-HCl [pH 7.5], 1 mM EDTA, 50 mM NaCl, 2 mM DTT, 5% glycerol, 0.5% NP-40, 10 mug of BSA/mul) with protease inhibitors at 48 h posttransfection. The binding reaction mixture contained 20 mug of total proteins, 62.5 mug of poly(dI-dC)/ml, 2 mM DTT, 10 mug of BSA/mul, and 40,000 cpm of 32P-labeled probe. (Left panel) EMSA with the double-stranded sequence containing LMP1p ISRE (GATCCAACAGGAAATGGAAAGGCAGTG). Lanes 10 to 14 show supershift with Flag antibody. (Right panel) EMSA with the mutated LMP1p ISRE (GATCCAACAGGAggTGGAggGGCAGTG [the mutated nucleotides are indicated by lowercase letters]). For competitor assays, 100-fold excess cold probe or AP-1 sequences were added to the binding mixture. For supershift assays, 0.2 mug of Flag antibody was added to each sample before the probe was added. Protein-DNA complexes were separated on 5% 60:1 acrylamide gels. FIG. 5. : Expression of LMP1 protein is upregulated in P3HR1 cells transiently transfected with IRF7 alone or together with EBNA2. Expression of LMP1 protein is upregulated in P3HR1 cells transiently transfected with IRF7 alone or together with EBNA2. P3HR1 cells contain an EBV genome from which EBNA2 and the C terminus of EBNA-LP have been deleted. Lanes 4 to 6 contain lysates from P3HR1 cells transfected with increasing amounts of IRF7, and lanes 7 to 10 contain lysates from P3HR1 cells transfected with different combinations of IRF7 and EBNA2. LMP1, EBNA2, and IRF7 were probed with anti-LMP1 (CS1-4), anti-EBNA2 (PE2), and anti-IRF7 (H-246) at dilutions of 1:100, 1:500, and 1:300, respectively. Equal amounts of proteins were loaded in each lane, as verified by tubulin blotting. FIG. 6. : IRF7 upregulates LMP1 mRNA in P3HR1 cells. IRF7 upregulates LMP1 mRNA in P3HR1 cells. P3HR1 cells were transfected with IRF7 or EBNA2 or combinations of different amounts of IRF7 and EBNA2. Transfected cells were selected using anti-CD4 magnetic beads, and total RNAs were isolated. LMP1 and GAPDH probes were labeled with [alpha-32P]UTP and used for RPA. Saccharomyces cerevisiae RNA was used as the negative control, and RNA from JiJoye cells was used as the positive control. Equal amounts of RNAs were loaded in each lane, as verified by GAPDH signals. A histogram quantifying LMP1 mRNA signals by normalization with GAPDH mRNA signals is shown at the bottom. FIG. 7. : A scheme for the proposed regulatory circuit between IRF7 and LMP1. A scheme for the proposed regulatory circuit between IRF7 and LMP1. IRF7 protein can activate the LMP1 promoter, as demonstrated in this study. The activated LMP1 promoter results in elevated LMP1 protein levels, which in turn induce IRF7 mRNA and protein and regulate IRF7 protein activity through phosphorylation as indicated . Thick, solid arrows indicate identified regulatory pathways, and broken arrows indicate pathways which are proposed but still under study. Backmatter: PMID- 12915586 TI - Susceptibility of Human Hepatitis Delta Virus RNAs to Small Interfering RNA Action AB - In animal cells, small interfering RNAs (siRNA), when exogenously provided, have been reported to be capable of inhibiting replication of several different viruses. In preliminary studies, siRNA species were designed and tested for their ability to act on the protein expressed in Huh7 cells transfected with DNA-directed mRNA constructs containing hepatitis delta virus (HDV) target sequences. The aim was to achieve siRNA specific for each of the three RNAs of HDV replication: (i) the 1,679-nucleotide circular RNA genome, (ii) its exact complement, the antigenome, and (iii) the less abundant polyadenylated mRNA for the small delta protein. Many of the 16 siRNA tested gave >80% inhibition in this assay. Next, these three classes of siRNA were tested for their ability to act during HDV genome replication. It was found that only siRNA targeted against HDV mRNA sequences could interfere with HDV genome replication. In contrast, siRNA targeted against genomic and antigenomic RNA sequences had no detectable effect on the accumulation of these RNAs. Reconstruction experiments with nonreplicating HDV RNA sequences support the interpretation that neither the potential for intramolecular rod-like RNA folding nor the presence of the delta protein conferred resistance to siRNA. In terms of replicating HDV RNAs, it is considered more likely that the genomic and antigenomic RNAs are resistant because their location within the nucleus makes them inaccessible to siRNA-mediated degradation. Keywords: Introduction : Human hepatitis delta virus (HDV) has a 1,679-nucleotide (nt) single-stranded circular RNA genome that is replicated by RNA-directed RNA synthesis, most probably involving host RNA polymerase II . During this replication, three RNA species accumulate , as represented in Fig. . The genome and its exact complement, the antigenome, are considered unit length. They exist primarily in a circular conformation but also in a linear conformation; these two conformations can be resolved using appropriate conditions of gel electrophoresis . The third RNA species consists of relatively lower amounts of an 800-nt polyadenylated RNA (of the same polarity as the antigenome), which is translated to produce a 195-amino-acid protein, known as the delta antigen (deltaAg-S), and is essential for HDV genome replication . An increasing number of reports have shown that small interfering RNAs (siRNA) can be exogenously provided to cells undergoing animal virus replication and achieve inhibition . For the following reasons, we were specifically interested in the possible susceptibility to siRNA of HDV RNAs. (i) The HDV genomic and antigenomic RNAs can fold into an unbranched rod-like structure with 74% of the bases paired , and this folding might interfere with siRNA action. (ii) The delta protein has the ability to bind double-stranded RNA and thus might also interfere. (iii) While several reports indicate that HDV genomic and antigenomic RNAs are predominantly located in the cell nucleus , two recent studies using cell fractionation indicated that much of the genomic RNA (but not the antigenomic RNA) might be cytoplasmic and thus possibly accessible to siRNA attack . With these questions in mind, our first objective was to design and test siRNA species specific for sequences on HDV mRNA, genomic RNA, and antigenomic RNA. As represented in Fig. , the initial strategy was to use expression vectors to produce within Huh7 cells DNA-directed mRNA species that contain these three sequences. As indicated, a total of 16 siRNA, each containing a 21-nt region based on the nucleotide sequence of Kuo et al. , were designed and constructed using a Silencer siRNA construction kit (Ambion). The locations and sequences of the target sites are listed in Table . The assay in each case was for the translation and accumulation of the delta protein, as detected by immunoblotting. The results and their quantitation are presented in Fig. . Also shown is the immunoblot for the internal control, green fluorescent protein (GFP). An expression vector for GFP was cotransfected, and as expected, the siRNA directed against HDV sequences did not inhibit the accumulation of GFP protein. Consistent with the siRNA experience of others , we found that some of the designed siRNA were able to give 80 to 95% inhibition of the HDV mRNA sequences (Fig. , lanes 4, 6, and 7). Furthermore, as planned, it was possible to obtain siRNA that attacked the genomic and antigenomic RNA sequences (Fig. , respectively). It should be noted that in Fig. , the insert of 657 nt of antigenomic sequence (from position 660 to 4) should thus have been able in large part to fold with the 585 nt spanning the delta antigen open reading frame (at position 1596 to 1011) into an extensive amount of unbranched rod-like structure. Even if this potential folding did occur in vivo, it did not confer resistance to functioning siRNA (lanes 14, 15, and 16). We next examined whether those siRNA with proven activity could interfere with HDV genome replication, as assayed by the accumulation of unit-length HDV RNA species. Some results are shown in Fig. , along with quantitation. We observed that inhibition of RNA accumulation occurred only with siRNA 4 and 6, which targeted the mRNA sequences and caused a significant reduction of delta protein accumulation . (In panel A, the data shown are for antigenomic RNA; however, similar results were obtained for genomic RNA [data not shown]. Also, siRNA 14, 15, and 16 failed to reduce HDV RNA accumulation.) In the above-described experiment, the gel electrophoresis conditions used do not separate the circular from the linear conformations of unit-length HDV RNAs. However, if the siRNA treatment led to single endonucleolytic cuts on unit-length circles to produce linears, we expect that this would have inhibited further RNA accumulation . Therefore, to further test this preliminary interpretation that unit-length genomic and antigenomic RNAs were resistant to siRNA, we carried out the following additional experiment. We used transfection to express HDV RNA multimers by DNA-directed RNA transcription. From previous studies, we knew that these would be posttranscriptionally processed to form unit-length RNA circles and yet, because of a 2-nt deletion in the open reading frame for the small delta protein, would be unable to make the essential delta protein and undergo RNA-directed transcription and replication . We then used gel analysis conditions capable of separating both linear and circular conformations of unit-length HDV RNA. Some data are shown in Fig. . As can be seen from the quantitation, cotransfection with siRNA 10, 11, 12, and 13, specific for genomic RNA sequences, did not reduce the accumulation of unit-length genomic RNA. Furthermore, the presence of this siRNA did not cause a reduction in the fraction of RNAs with a circular conformation (lanes 10 to 13 relative to lane C). In similar experiments, we expressed antigenomic nonreplicating RNAs and found that they were not sensitive to siRNA 4, 6, and 9 (data not shown). We interpret these data as evidence against siRNA action, even for inducing a single nick on the nonreplicating HDV RNAs. In addition, the delta protein was not present and thus could not be the basis for the observed resistance. Our studies show that HDV circular RNAs, whether transcribed from an RNA template or from a DNA template , were resistant to siRNA attack. In the case of the DNA-directed transcript, the expression vector was such that the primary (nonreplicating) multimeric transcript was via host RNA polymerase II and that RNA (prior to ribozyme processing and ligation to form unit-length circles) should have undergone both 5'-capping and 3' polyadenylation. Others have shown that for a host mRNA precursor, the intronic regions are resistant to siRNA while the exons are sensitive . Our studies show that the unit-length circular RNA processed out of the multimeric transcript was not only stable but also resistant to siRNA attack. In summary, these studies support the interpretation that during genome replication, the only HDV RNA directly susceptible to inhibition mediated by siRNA is the mRNA. At least under the conditions of these experiments, the resistance of the genomic and antigenomic RNAs was not dependent on RNA structure, RNA conformation, or the presence of the small delta protein. Further experiments will be needed to determine if this "resistance" was in fact due to inaccessibility based on nuclear localization of genomic and antigenomic unit-length RNAs. It remains possible that some of the genomic RNA is cytoplasmic but is somehow inaccessible to attack by siRNA. For example, it could be protected by a host RNA-binding protein. For this or maybe other reasons, these HDV RNAs are probably resistant to siRNA because they are simply not accessible to the RISC, a protein-RNA effector nuclease complex that recognizes and destroys target RNAs . In this respect, our findings with HDV are analogous to those reported for siRNA action on the replication of respiratory syncytial virus and influenza virus . That is, siRNA cannot target the replicating viral RNA transcripts directly but only indirectly via action on the viral mRNA species. FIG. 1. : Representation of three main species of HDV RNA. Representation of three main species of HDV RNA. Also indicated are the genomic and antigenomic ribozymes (cleavage site is shown as a circle) and the open reading frame (ORF) for the deltaAg . At the right is the number of molecules of each RNA per average liver cell for an infected woodchuck and chimpanzee, as previously reported . Indicated on the 1,679-nt genomic RNA is the origin for the nucleotide numbering, according to the sequence of Kuo et al. . FIG. 2. : Transfected siRNAs could target HDV mRNA species. Transfected siRNAs could target HDV mRNA species. Huh7 cells were transfected with one of three plasmids that express an HDV mRNA species. (A) pDL444 expressed an mRNA equivalent to normal HDV mRNA. (B and C) pDL444 was modified to contain 657 nt of extra sequences in the 3' untranslated region, which led to the transcription of either partial genomic HDV RNA sequences (position 4 to 660) (B) or antigenomic RNA sequences (position 660 to 4) (C). The constructs were cotransfected with a plasmid expressing GFP. After 2 days the total protein was extracted and examined by immunoblotting to detect both delta protein and GFP. Detection and quantitation were with a bioimager (Fuji). The 16 HDV-specific siRNA indicated in the figure were designed and delivered as a cotransfection using Lipofectamine 2000 (Invitrogen). As a negative control (lanes C) we used siRNA against glyceraldehyde-3-phosphate dehydrogenase. FIG. 3. : Effect of transfected siRNA on the accumulation of replicating HDV RNAs. Effect of transfected siRNA on the accumulation of replicating HDV RNAs. Cells were transfected with pDL553 to initiate HDV genome replication. siRNA species (at 30 nM) were cotransfected as indicated and as previously described in Fig. . At day 2, total RNA was extracted, glyoxalated, and analyzed by electrophoresis in a 1% agarose gel. HDV antigenomic RNA was then detected by Northern assay (A). Similarly, total protein was examined by immunoblotting to detect deltaAg-S (B). For both panels, bioimager data were subjected to quantitation, expressing the amount of signal detected relative to that obtained for the control transfection. As a negative control (lane C), we used siRNA against endogenous glyceraldehyde-3-phosphate dehydrogenase (GAPDH); treatment with this siRNA reduced GAPDH mRNA but had no effect on HDV RNA levels (data not shown). Also shown in panel B is the immunoblot assay for expression of GFP, which was cotransfected as a control. FIG. 4. : Transfected siRNA did not target nonreplicating unit-length HDV genomic RNA. Transfected siRNA did not target nonreplicating unit-length HDV genomic RNA. Cells were transfected with pDL542 to achieve the transcription and accumulation of nonreplicating unit-length genomic HDV RNA circles. Cotransfected with this plasmid was either a glyceraldehyde-3-phosphate dehydrogenase control siRNA (lane C) or genomic HDV-specific siRNA (lanes 10 to 13), as indicated at the top of the figure and as previously described for Fig. . At day 2, total RNA was extracted, glyoxalated, and analyzed by electrophoresis in a 3% agarose gel. Linear and circular forms of HDV genomic RNA were then detected by Northern assay. After hybridization and quantitation using the bioimager, we deduced the amounts of linear and circular RNAs, as summarized in the histogram. TABLE 1 : Locations and sequences of siRNA targets Backmatter: PMID- 12915583 TI - Protective Efficacy of an AIDS Vaccine, a Single DNA Priming Followed by a Single Booster with a Recombinant Replication-Defective Sendai Virus Vector, in a Macaque AIDS Model AB - We previously demonstrated the excellent protective efficacy of DNA priming followed by Gag-expressing Sendai virus (SeV) boosting (DNA prime/SeV-Gag boost vaccine) against a pathogenic simian-human immunodeficiency virus (SHIV89.6PD) infection in macaques. Here we show that we established a practical, safer AIDS vaccine protocol, a single DNA priming followed by a single booster with a recently developed replication-defective F deletion SeV-expressing Gag, and show its protective efficacy against SHIV89.6PD infections. Keywords: Introduction : Virus-specific cellular immune responses play an important role in the control of immunodeficiency virus infections . DNA vaccines, recombinant-viral-vector-based vaccines, and their combinations are promising AIDS vaccine methods because of their potential for inducing cellular immune responses. Recently, some of these AIDS vaccines inducing virus-specific cellular immune responses have been reported to prevent AIDS progression in macaque models using pathogenic simian-human immunodeficiency viruses (SHIV) . Members of our laboratory previously reported the potential of a recombinant Sendai virus (SeV) vector for inducing virus-specific cellular immune responses and the excellent protective efficacy of a vaccine system consisting of DNA priming and a Gag-expressing SeV (SeV-Gag) booster in a macaque AIDS model . In a preclinical trial, DNA priming-SeV-Gag booster vaccination induced high levels of virus-specific T cells, the viremia in all the vaccinated macaques was controlled, and the animals were protected from AIDS progression after a pathogenic SHIV (SHIV89.6PD) challenge. SeV is an enveloped virus with a negative-sense RNA genome. It causes fatal pneumonia in mice, its natural host, but is thought to be nonpathogenic in primates, including humans . The recombinant SeV vector system has been shown to induce efficient gene transfer in vitro . A recent analysis confirmed that SeV infection is nonpathogenic in macaques and that its transmission between them is inefficient. Intranasal SeV inoculation of macaques induces antigen expression localized in the nasal mucosa and its primary lymph nodes (LN), the submandibular LN and the retropharyngeal LN. Furthermore, a safer replication-defective SeV vector lacking the F gene, F(-)SeV, has recently become available . The DNA prime/SeV-Gag boost regimen reported previously consists of a series of four vaccinations with defective proviral DNA and a single intranasal booster with a replication-competent SeV-Gag . In this study, we established a practical, safer DNA prime/SeV boost system, a single DNA priming followed by a single booster with a replication-defective F(-)SeV-expressing simian immunodeficiency virus (SIV) Gag protein, F(-)SeV-Gag, for clinical use as an AIDS vaccine. The DNA priming-boosting regimen, called DNA/F(-)SeV-Gag version 2, used in this study consists of a single DNA vaccination and an intranasal F(-)SeV-Gag booster at week 6 after the priming. The DNA used for the priming (referred to as CMV-SHIVdEN) was constructed from an env and nef deletion SHIV DNA (SIVGP1 DNA) by replacing the 5' long terminal repeat region with a cytomegalovirus promoter with an immediate-early enhancer and the 3' long terminal repeat region with simian virus 40 poly(A). At DNA vaccination, the animals received 5 mg of CMV-SHIVdEN DNA intramuscularly. At the booster vaccination, the animals received 6 x 109 cell-infectious units of F(-)SeV-Gag intranasally. All the animal experiments in this study were performed in accordance with the guidelines for laboratory animals of the National Institute of Infectious Diseases. We measured virus-specific T-cell levels by flow-cytometric analysis of gamma interferon (IFN-gamma) induction after specific stimulation as described previously . In brief, lymphocytes were cocultured with autologous herpesvirus papio-immortalized B lymphoblastoid cell lines infected with a recombinant vaccinia virus vector expressing SIV Gag for Gag-specific stimulation. Alternatively, lymphocytes were cocultured with B lymphoblastoid cell lines infected with vesicular stomatitis virus G-pseudotyped SIVGP1 for SHIV-specific stimulation. Gag-specific T-cell levels and SHIV-specific T-cell levels were calculated by subtracting IFN-gamma-positive T-cell frequencies after nonspecific stimulation from those after Gag-specific stimulation and SHIV-specific stimulation, respectively. Values exceeding 0.05% were considered positive. In the first experiment, we examined the distribution of the F(-)SeV-Gag vector after the booster. Two cynomolgus macaques (Macaca fascicularis), C97-018 and C94-030, received the DNA prime/F(-)SeV-Gag boost vaccine. Neither of them showed apparent clinical symptoms. Both of them were euthanized 2 weeks after the booster. We extracted RNA from the cells prepared from each tissue taken at autopsy and examined F(-)SeV-Gag distribution by detection of SIV gag after nested reverse transcription-PCR amplification as previously described . The vector was detected in the nasal mucosa, the tonsil, the submandibular LN, and the retropharyngeal LN but was undetectable in other lymphoid tissues, including peripheral LN . Reverse transcription-PCR with SeV N-specific primers showed similar results (data not shown). These results indicate that F(-)SeV-Gag distribution was restricted to the nasal mucosa and its primary lymphoid tissues. Flow-cytometric analysis of Gag-specific IFN-gamma induction in peripheral blood mononuclear cells (PBMC) at the autopsy showed efficient Gag-specific T-cell induction in both of the vaccinated macaques. Then we examined the distribution of Gag-specific CD8+ T cells after the booster. While Gag-specific CD8+ T cells were detected in all the analyzed tissues (the nasal mucosa, the tonsil, the retropharyngeal LN, and the axillary LN), the levels in the nasal mucosa and the tonsil harboring the vector were higher than those in the axillary LN and PBMC . In the second experiment, we evaluated the protective efficacy of the DNA/F(-)SeV-Gag version 2 system in a SHIV model. Three rhesus macaques (Macaca mulatta), R00-020, R00-023, and R00-024, were vaccinated for the challenge experiment. None of them showed apparent clinical symptoms after the vaccination. At week 6, just before the F(-)SeV-Gag booster, SHIV-specific CD8+ T cells were undetectable in all three animals and significant levels of SHIV-specific CD4+ T cells were detected in one animal (R00-023) but not in the others. At week 7, 1 week after the booster, all the macaques showed high levels of SHIV-specific CD4+ and CD8+ T cells, indicating efficient expansion of the number of SHIV-specific T cells as a result of the F(-)SeV-Gag booster . The levels peaked at around 1 week after the booster. Macaques R00-020 and R00-023 maintained detectable levels of SHIV-specific CD8+ T cells as well as CD4+ T cells until challenge at week 19, about 3 months after the booster. Macaque R00-024 also maintained SHIV-specific CD4+ T cells until challenge, but its SHIV-specific CD8+-T-cell level declined to a marginal level before challenge. These three macaques were intravenously challenged with 10 50% tissue culture infective doses (TCID50) of SHIV89.6PD 13 weeks after the booster. In the control experiment performed previously , all four naive control macaques showed acute depletion of peripheral CD4+ T lymphocytes 2 weeks after the challenge. After the acute phase, the viremia in three of the macaques was not controlled and they developed AIDS and were euthanized within a year. In contrast, all three macaques vaccinated with the DNA/F(-)SeV-Gag version 2 system were protected from acute AIDS progression . Among them, two macaques, R00-020 and R00-023, were protected from acute CD4+-T-cell depletion. Their peak viral loads in plasma were greatly reduced (geometric mean, 5.9 x 105 copies/ml) compared to those in the controls (geometric mean, 1.7 x 107 copies/ml) and were undetectable at the set point. Acute viremia was not controlled in one macaque (R00-024), and peripheral CD4+ T cells were lost during the acute phase, but viremia was controlled and CD4+ T cells had recovered at the set point. We then examined SHIV-specific T-cell levels after challenge . In macaques R00-020 and R00-023, which showed higher levels of protection, SHIV-specific CD8+ T cells were detectable in PBMC even at week 1 after challenge and more vigorous secondary responses were observed after that. In contrast, SHIV-specific CD8+ T cells were undetectable at week 1 in macaque R00-024, which showed a lower level of protection, although vigorous secondary responses were observed after that. These results are compatible with those of a previous study, indicating that rapid secondary responses of virus-specific CD8+ T cells are important for controlling acute viremia . Levels of SHIV-specific CD4+ T cells were augmented in all the vaccinated macaques at week 1.5 after challenge . Both of the macaques in which acute viremia was controlled maintained these levels after that. In macaque R00-024, SHIV-specific CD4+ T cells became undetectable at week 2, but the levels recovered after that. Thus, control of SHIV infections by the DNA/F(-)SeV-Gag vaccine led to maintenance of virus-specific CD4+ T cells at the set point. Recently, several kinds of DNA prime/viral vector boost vaccines have been shown to induce protective efficacy against pathogenic SHIV infections . In those studies, DNA vaccinations were performed more than once. In the present study, we evaluated the efficacy of a DNA prime/SeV boost system with minimum numbers of vaccinations in an SHIV model. Furthermore, this study is the first preclinical trial of a replication-defective F(-)SeV vector as a vaccine tool. In macaques, the vector distribution was localized in the nasal mucosa, its primary LN, and the tonsil after an intranasal F(-)SeV-Gag booster. The distribution was similar to that with replication-competent SeV-Gag. Gag-specific CD8+ T cells were detected predominantly in the nasal mucosa and the tonsil, suggesting that the vector-derived antigen presentation was predominant in these tissues. In addition to the local induction, the intranasal F(-)SeV-Gag injection induced Gag-specific CD8+ T cells to be produced systemically. These results indicate that the F(-)SeV vector is a promising vaccine tool for inducing virus-specific cellular immune responses. In a previous study , members of our laboratory showed that the production of SHIV-specific T cells was efficiently induced in all four macaques vaccinated with the DNA/SeV-Gag version 1 system consisting of a series of four DNA vaccinations and a replication-competent SeV-Gag booster. All of them maintained SHIV-specific CD8+ T cells for more than 3 months until challenge, and acute viremia was controlled without acute CD4+-T-cell depletion after SHIV89.6PD challenge. Two of three macaques vaccinated with the DNA/F(-)SeV-Gag version 2 system in this study showed similar levels of protection; acute viremia was controlled without acute CD4+-T-cell depletion, and they were protected from AIDS progression. On the other hand, in the remaining macaque (R00-024) with a marginal level of SHIV-specific CD8+ T cells at challenge, acute viremia was uncontrolled and there was acute CD4+-T-cell depletion. However, even in macaque R00-024, SHIV89.6PD infections were controlled at the set point, CD4+ T cells recovered, and the animal was protected from AIDS progression. While two levels of protection were observed, this study indicates that even the DNA/F(-)SeV-Gag version 2 AIDS vaccine system with a minimum number of vaccinations can induce protective immune responses in a macaque AIDS model. FIG. 1. : Gag-specific CD8+-T-cell levels in tissues after an intranasal F(-)SeV-Gag booster. Gag-specific CD8+-T-cell levels in tissues after an intranasal F(-)SeV-Gag booster. Samples were obtained from two cynomolgus macaques, C97-018 and C94-030, which were euthanized 2 weeks after an intranasal F(-)SeV-Gag booster. Numbers of Gag-specific CD8+ T cells are shown as percentages of the total number of CD8+ T cells are shown as percentages of the total number of CD81 T cells. The lower panels show representative dot plots of IFN-g induction in CD81 T cells after Gag-specific stimulation. Dot plots gated on CD31 CD81 lymphocytes of the tonsil (left) and the retropharyngeal LN (right) in macaque C97-018 are shown. Ax LN, axillary LN; RP LN, retropharyngeal LN. FIG. 2. : SHIV-specific-T-cell levels in PBMC before (A) and after (B) challenge. SHIV-specific-T-cell levels in PBMC before (A) and after (B) challenge. For three vaccinated rhesus macaques, numbers of SHIV-specific CD4+ T cells (left panels) and numbers of SHIV-specific CD8+ T cells (right panels) are shown as percentages of the total numbers of CD4+ and CD8+ T cells, respectively. (A) The levels at week 6 [just before the F(-)SeV-Gag booster], week 7 (1 week after the booster), week 8 (2 weeks after the booster), and week 19 (just before SHIV challenge) after the initial DNA priming are shown. (B) The levels at weeks 1, 1.5, 2, and 20 after challenge are shown. FIG. 3. : Changes in peripheral CD4+-T-cell counts and plasma viral loads after SHIV89. Changes in peripheral CD4+-T-cell counts and plasma viral loads after SHIV89.6PD challenge. The left panels show results for previously described naive control macaques (17), and the right panels show results for macaques vaccinated with the DNA prime/F(-)SeV-Gag boost vaccine. (A) Changes in relative peripheral CD4+-T-cell counts (per microliter). For each animal, the CD4 count at the challenge is considered to be 100, and the values for the CD4 counts relative to that at the challenge are shown. (B) Changes in plasma SHIV RNA copy numbers (numbers of copies per milliliter) quantified as described previously . TABLE 1 : Distribution of SIV gag RNA 2 weeks after a F(-)SeV-Gag booster Backmatter: PMID- 12915533 TI - Tsg101 Control of Human Immunodeficiency Virus Type 1 Gag Trafficking and Release AB - The structural precursor polyprotein of human immunodeficiency virus type 1, Pr55gag, contains a proline-rich motif (PTAP) called the "late domain" in its C-terminal p6 region that directs release of mature virus-like particles (VLPs) from the plasma membranes of gag-transfected COS-1 cells. The motif binds Tsg101 (vacuolar protein-sorting protein 23, or Vps23), which functions in endocytic trafficking. Here, we show that accumulation of the wild-type (wt) Gag precursor in a fraction of COS-1 cytoplasm enriched in multivesicular bodies and small particulate components of the plasma membrane (P100) is p6 dependent. Cleavage intermediates and mature CA mainly partitioned with more rapidly sedimenting larger material enriched in components of lysosomes and early endosomes (P27), and this also was p6 dependent. Expression of truncated or full-length Tsg101 proteins interfered with VLP assembly and Gag accumulation in the P100 fraction. This correlated with reduced accumulation of Gag tagged with green fluorescent protein (Gag-GFP) at the plasma membrane and colocalization with the tagged Tsg101 in perinuclear early endosomes, as visualized by confocal microscopy. Fractionation analysis and confocal examination both indicated that the N-terminal region of Tsg101, which contains binding sites for PTAP and ubiquitin (Ub), was required for Gag trafficking to the plasma membrane. Expression of FLAG-tagged Tsg101 with a deletion in the Ub-binding pocket inhibited VLP release almost completely and to a significantly greater extent than expression of the wt tagged Tsg101 protein or Tsg101-FLAG containing a deletion in the PTAP-binding region. The results demonstrate that Gag associates with endosomal trafficking compartments and indicate that efficient release of virus particles from the plasma membrane requires both the PTAP- and Ub-binding functions of Tsg101 to recruit the cellular machinery required for budding. Keywords: Introduction : All retroviruses have in common three genes, gag, pol, and env, which specify the structural and enzymatic functions of the virus (reviewed in reference ). The gag gene alone is sufficient for assembly and release of immature virus-like particles (VLPs) from infected cells. Maturation to form the infectious particle requires a protease (PR) encoded in pol. The gag-encoded protein (Gag) contains distinct domains involved in assembly and release. The region required for release by budding from the plasma membrane is called the late (L) domain . The L domain of Gag is a Pro-rich motif that is highly conserved in retroviruses (, , , , , , ; reviewed in reference ); other enveloped viruses, including rhabdoviruses, filoviruses, and Epstein-Barr virus, and cellular proteins also have Pro-rich motifs . The L domains in retroviruses differ in amino acid sequence and location within the respective viral structural proteins but are functionally exchangeable , suggesting commonality of function. Recently, the PTAPP motif, or L domain, of human immunodeficiency virus type 1 (HIV-1), which is located in the C-terminal p6 region of the Gag precursor polyprotein, was found to bind to the protein product of the tsg101 gene . Tsg101 was originally identified by the reversible neoplasia associated with its functional inactivation in murine fibroblasts . A cell line deficient in Tsg101 (SL6) shows a variety of nuclear, microtubule, and mitotic spindle abnormalities , and tsg101 null mutant mice show defective cell proliferation and early embryonic death . Sequence analysis has suggested , and experimental evidence has shown, that Tsg101 can function in both the modulation of transcription and the inhibition of ubiquitination and protein decay . The latter effects are mediated by an N-terminal region that contains a ubiquitin (Ub) conjugase (E2)-like domain. The domain lacks the active-site Cys residue crucial to Ub conjugation but binds Ub elsewhere in this region . The N-terminal domain of Tsg101 also is the minimal binding region required for its interaction with HIV-1 Gag . Tsg101 appears to be ubiquitous in the cell and exhibits cell cycle-dependent localization in the Golgi complex , suggesting that it is highly dynamic. Tsg101-deficient SL6 cells show defective endosomal trafficking , and an orthologue of Tsg101, Stp22p/Vps23, is a class E vacuolar protein sorting (Vps) protein in the endosomal sorting machinery of Saccharomyces cerevisiae . Both the mammalian and the yeast proteins have been shown to recognize Ub and act in the removal of endosomal protein-Ub conjugates through a multivesicular body (MVB) . The C-terminal region of Tsg101 contains a binding site for Vps28, and an upstream coiled-coil domain may facilitate Vps37 binding . Together, the three proteins form a complex, ESCRT-1 (endocytic sorting complex required for trafficking). The ability of Tsg101 to bind Ub is critical to the function of the complex. ESCRT complexes 2 and 3 have also been described recently . Recent studies have demonstrated that Tsg101 is required for HIV-1 production , implicating the endocytic pathway in this process. During investigation of the effects of perturbation of the Tsg101 protein level on Gag assembly, we found that Tsg101 overexpression diminished Gag release, an effect similar to that of p6 deletion. This observation suggested a means of identifying the basis for the Tsg101 binding requirement in Gag release and prompted us to examine cells for alterations associated with p6 deletion and Tsg101 overexpression. We found that wild-type (wt) Gag was stably associated with a particulate fraction enriched in MVBs and plasma membrane components where Gag precursors lacking the p6 region (Deltap6) did not stably accumulate. Expression of full-length or truncated Tsg101 proteins inhibited association of wt Gag with this fraction, promoted Tsg101-Gag colocalization with perinuclear endosomes, and diminished VLP release. Disruption of the PTAP-binding function of Tsg101 resulted in perinuclear sequestration of Tsg101-Gag complexes. Disruption of the Ub-binding function of Tsg101, which is required for correct sorting of cargo in endocytic trafficking, prevented Tsg101-Gag colocalization in both the perinuclear and plasma membrane regions and was very inhibitory to VLP release. These results identify an endosome-enriched subcellular fraction in which Gag accumulation is dependent on both p6 and Tsg101, and they also indicate that Gag trafficking and VLP release require the PTAP- and Ub-binding functions of Tsg101. MATERIALS AND METHODS : Plasmids. | pgp-RRE-r carries wt gag, pol, and vif. The pgp-RRE-r gagDeltap6 construct contains three point mutations in the first codon of p6, converting it to a stop (ochre) codon, as described previously . The mutation did not affect PR activity . pIND-hTsg101-FLAG (encoding wt human Tsg101) and mutants derived from it have been described by Feng et al. . Constructs encoding Tsg101 C-terminally tagged with Myc (pLLEXP1-hTsg101-myc) and Tsg101-Myc deletion mutants have been described by Li et al. . Tsg101 fused to enhanced yellow fluorescent protein (Tsg101-EYFP) was constructed by amplifying Tsg101 with engineered BamHI and EcoRI cleavages and cloning the resulting DNA into the BglII and EcoRI sites of pEYFP-C1 (Clontech, Palo Alto, Calif.). Hrs was amplified with primers containing restriction sites for AvrII and HindIII. The DNA, with a C-terminal FLAG tag generated by PCR, was cloned into the pLLEXP1 vector (described in reference ) by using NheI and HindIII sites. Cell culture, transfection, and preparation of cytoplasmic extracts. | COS-1 cells were cultured in Dulbecco's modified Eagle medium supplemented with fetal bovine serum and antibiotics to 60% confluency at 37C. The cells were transfected with pCMV-rev, wt or mutated pgp-RRE-r, and either wt or mutated pIND-h-Tsg101-FLAG with pVgRXR or wt or mutated pLLEXP1-hTsg101-myc by use of the FuGene 6 reagent (Roche). Rev is an HIV-1-encoded protein required for expression of gag and pol . pVgRXR encodes both monomers of a heterodimeric ecdysone-inducible receptor that, together with the hormone, forms a transcriptional activator complex that binds the promoter driving tsg101 expression. Treatment of the transfected cells with the ecdysone analogue ponasterone (pa) (5 muM) 24 h after transfection induced high-level expression of pIND-h-Tsg101-FLAG. A low level of Tsg101 expression was obtained in cells transfected with pIND-h-Tsg101-FLAG and pVgRXR in the absence of pa. At 48 h posttransfection, the culture medium was separated, and cells were harvested by scraping with a rubber policeman and collected by centrifugation. The pelleted cells were washed three times with cold phosphate-buffered saline (PBS), swollen in 1 ml of cold hypotonic buffer (10 mM Tris [pH 7.4]-1 mM MgCl2) containing protease inhibitors (for 15 min at 4C), and disrupted with 35 strokes of a Dounce homogenizer (type B pestle). Fractionation of the cell lysate was performed as described previously . The lysate was spun for 10 min at 1,000 x g and 4C to separate nuclei and any unbroken cells from the cytoplasmic fraction (S1). Where indicated, particulate material that sedimented at 27,000 x g (P27) was isolated from the S1 fraction, the supernatant was spun at 100,000 x g, and the resulting pellet (P100) was collected. Virus particles in culture supernatants passed through a 0.45-mum-pore-size filter were isolated by ultracentrifugation through a cushion of 20% sucrose at 30,000 rpm for 80 min at 4C (in a Beckman SW41 rotor). The pelleted viral particles were resuspended at 4C in similar100 mul of PBS containing 2% sodium dodecyl sulfate. Western blot analysis. | Proteins were separated by electrophoresis through sodium dodecyl sulfate-12.5% polyacrylamide gels. Following electrophoresis, the gels were transferred to nitrocellulose filters and analyzed by Western blotting. An anti-Tsg101 mouse monoclonal antibody (Santa Cruz, Inc.), an anti-Myc mouse monoclonal antibody (Sigma), an anti-CA rabbit polyclonal antibody raised against a native form of the CA protein , and a mouse anti-FLAG antibody (Sigma) were used as specified below. Proteins were visualized by chemiluminescence using Lumi-Light reagents (Roche). Measurements of relative protein levels were determined by densitometry using NIH Image (version 1.62) software. Confocal microscopy. | COS-1 cells were transfected with Gag C-terminally tagged with green fluorescent protein (GFP) by using pCMV-Gag-EGFP and, where indicated, with DNA encoding Tsg101 or Hrs. At 48 h posttransfection, the cells were washed once in PBS and fixed in 4% formaldehyde (Fisher) in Ca2+-free, Mg2+-free PBS for 20 min. Samples were then washed three times for a total of 5 min with PBS, permeabilized with 0.1% Triton X-100 for 5 min, and washed three more times with PBS. After being blocked for 10 min in PBS containing 1% bovine serum albumin, the cells were incubated with the primary antibody for 1 h at 37C, rinsed with PBS, and then incubated with the fluorescently tagged secondary antibody for 30 min at 37C. The nuclear stain 4',6'-diamidino-2-phenylindole (DAPI) (Molecular Probes) was added in the last 10 min. After a rinse, the cells were mounted by using p-phenylenediamine (PADA) and Immunomount. Images were captured on an inverted fluorescent/differential interference contrast Zeiss Axiovert 200M microscope equipped with an AxioCam HRm camera (Zeiss) and a mercury arc lamp light source, using a 63x Plan-Apochromat (NA 1.40) oil objective, and operated by AxioVision (version 3.1; Zeiss) software. Twenty to thirty optimal sections along the z axis were acquired in increments of 0.4 mum. Figures show the central image or, where indicated, z sections from the adherent surface through the nucleus. The fluorescent data sets were deconvolved by using the constrained iterative method (AxioVision, version 3.1). The following excitation and emission wavelengths (lambdaex and lambdaem, respectively) were used for imaging: with DAPI or Alexa Fluor (for the goat anti-mouse secondary antibody), lambdaex = 360 +- 20 and lambdaem = 460 +- 25; with fluorescein isothiocyanate (for GFP), lambdaex = 480 +- 20 and lambdaem = 535 +- 25; with Texas Red (for tetramethyl rhodamine isothiocyanate or YFP), lambdaex = 560 +- 25 and lambdaem = 645 +- 35. RESULTS : p6-dependent Gag localization. | Previous studies have demonstrated that the structural precursor polyprotein Gag is sufficient for assembly of VLPs and for their release into the culture medium. The C-terminal p6 region of the protein plays a critical role in this process . COS-1 cells transfected with pCMV-rev and wt pgp-RRE-r express HIV-1 Gag and Gag-Pol precursors and Vif and assemble, release, and process the precursors, as indicated by detection of the mature CA protein in VLPs following Western blot analysis using a polyclonal antibody against the CA protein (Fig. , lanes 1 and 2). Deletion of p6 from Gag (Deltap6) blocked mature VLP formation (Fig. , lanes 3 and 4). Because p6-dependent Gag release has not been observed in all cell lines , two preparations are shown in Fig. to demonstrate that VLP release was p6 dependent under the conditions used in this study. To examine the basis of the p6 requirement for Gag release, we determined if p6 was required for the localization of Gag to a particular subcellular fraction. A previously described procedure was used to separate membrane-enriched particulate material based on sequential differential centrifugation at 27,000 x g and 100,000 x g . As shown in Fig. , both the full-length wt Gag precursor and the truncated Deltap6 protein, as well as the proteolytically processed p41, CA-p2, and mature CA products derived from them , were detected in the P27 fraction (lanes 1 and 3). Typically, the p41 and CA proteins derived from Gag were more abundant than those derived from Deltap6. The wt Gag protein and its cleavage products were also detected in the P100 fraction (Fig. , lane 2). The amount of wt Gag precursor in the P100 fraction was similar2- to 5-fold greater than that detected in the P27 fraction (n > 8). This estimate and those described below were based on comparison with the densitometry signals obtained by diluting a control (typically S1) lysate. In contrast, the amount of the Deltap6 precursor in the P27 fraction reproducibly exceeded the amount in the P100 fraction (Fig. ; compare lanes 3 and 4). The results indicated that accumulation of full-length Gag in the P100 fraction was p6 dependent and that accumulation of Gag-derived products in the P27 fraction was also p6 dependent. Truncated Tsg101 proteins, previously shown to interfere with Gag release, and the full-length Tsg101 protein interfere with Gag accumulation in the P100 fraction. | We next determined if the p6 dependence of Gag localization to the P100 fraction reflected the involvement of Tsg101 . Cells were cotransfected with DNA encoding Gag and either full-length Tsg101-Myc or truncated Tsg101 proteins , and the effects on VLP release and P100 accumulation were examined. Coexpression of Gag and full-length Tsg101-Myc or truncated Tsg101-Myc proteins containing the Gag binding site (Fig. , lanes 2 to 5) resulted in detection of similar2-fold less VLP in the medium than was released by cells transfected with Gag alone (lane 1). This finding is consistent with those of previous studies . Truncated Tsg101 proteins lacking the Gag binding site (Fig. , lanes 6 and 7) affected VLP formation to a lesser extent. Examination of the P27 and P100 fractions indicated that, with the exception of the P100 sample derived from cells expressing fragment F (amino acids [aa] 256 to 391) (lane 12), all of the P100 fractions contained less of the full-length Gag precursor than the respective P27 fraction (compare lanes 11 and 12 to lanes 1 and 2, 3 and 4, 5 and 6, 7 and 8, and 9 and 10). In addition, less p41, CA-p2, and CA accumulated in the P27 and P100 samples derived from cells expressing truncated Tsg101 proteins with the Gag binding site than in those without the Gag binding site (Fig. , lanes 3 to 8). These observations indicated that accumulation in the P100 fraction and efficient VLP release were linked and related to the Tsg101-Gag interaction. However, the fact that the nonbinding Tsg101 fragments (fragments E and F), which interfered minimally with VLP formation (Fig. , lanes 6 and 7), had different effects on P100 accumulation (Fig. , lanes 9 to 12) indicated that the relationship was not strictly correlated. This may reflect the heterogeneity of the P27 and P100 fractions. P27- and P100-associated endosomal markers. | As noted above, the P27 and P100 fractionation procedure separates membrane-enriched particulate components, in contrast to methods described previously (e.g., in reference ). To test for constituents of various subcellular particulate compartments, the P27 and P100 fractions were examined for enrichment of specific protein markers in an immunoblot analysis . Cathepsin D, a marker for hydrolytically active lysosomes , and early endosome antigen 1 (EEA1) and hepatocyte growth factor-regulated tyrosine kinase substrate (Hrs), both markers for early endosomes , were all enriched in the P27 fraction (Fig. , lane 1). In contrast, a marker for late endosomes and MVBs, the mannose-6-phosphate receptor (M6PR) protein , and Na+ K+ ATPase, an integral plasma membrane protein , were both enriched in the P100 fraction (Fig. , lane 2). Thus, although undoubtedly heterogeneous, the P27 and the P100 fractions nevertheless contained constituents that were specific for distinct endosomal compartments. Confocal microscopy of cells overexpressing Tsg101. | Taken together, the results shown in Fig. , , and suggested that Gag might be associated with endosomal compartments during assembly and that overexpression of wt Tsg101 or Tsg101 truncation mutants interfered with this process. To determine if the change in Gag distribution detected by biochemical fractionation reflected altered intracellular Gag localization, COS-1 cells expressing either Gag-GFP alone, Gag-GFP and full-length Tsg101-Myc, or Gag-GFP and N-terminally truncated Tsg101 (fragment E, aa 140 to 391) were examined by deconvoluting confocal microscopy. Figure shows the immunofluorescence detected in raw images of these cells and in z sections through the indicated area of the cells. Cells expressing Gag-GFP alone exhibited the expected fluorescence staining pattern ; i.e., staining was punctate, and most of the signal was membrane proximal. Some Gag was also detected near the DAPI (blue)-stained nucleus. In contrast, although the raw image indicated punctate staining in cells coexpressing full-length Tsg101-Myc , z sectioning revealed that most of the Gag-GFP colocalized with Tsg101 in the perinuclear region. Perinuclear colocalization of Tsg101 and Gag was detected whether the area of the cell examined was centrally (view 1) or peripherally (view 2) located. In contrast, in cells coexpressing the Tsg101 mutant lacking the N-terminal p6-binding domain (fragment E), much less of the Gag-GFP protein accumulated with Tsg101 near the nucleus . These observations were highly reproducible in four independent experiments, in each of which 40 to 50 cells were examined. The results indicate that expression of Tsg101-Myc interfered with accumulation of Tsg101-Gag complexes at the plasma membrane, and they support the results of the biochemical fractionation. Taken together, the results in Fig. and above suggested that Tsg101 overexpression might have promoted the accumulation of full-length Gag in the early-endosome-enriched P27 fraction. To examine this possibility, the localization of Gag-GFP to this compartment in the absence or presence of tagged Tsg101 was tested by using FLAG-tagged Hrs protein to identify early-endosome vesicles. As shown in Fig. , in the absence of Tsg101 overexpression, most of the Gag-GFP protein was detected at the plasma membrane and a subpopulation of Gag-GFP was detected in the perinuclear region, in agreement with the results described above and in previous reports . Moreover, much of the perinuclear subpopulation colocalized in the same z-section plane with the Hrs-FLAG-marked early endosomes. Coexpression of Gag-GFP, Hrs-FLAG, and Tsg101 tagged with YFP (Tsg101-YFP) (red fluorescence) indicated that virtually all of the perinuclear Gag-Tsg101 complexes were associated with the Hrs-FLAG-marked endosomes . These results suggest that Gag was sequestered in early endosomes following an increase in the level of Tsg101 above physiological levels. Effects of Tsg101 N-terminal mutations. | Previous studies have demonstrated the following. (i) The N-terminal region of Tsg101 contains binding sites for Ub and for the PTAP motif in p6 . Mutation of Val43, Asn45, and Asp46 to Ala reduces Ub binding 3- to 8-fold without affecting p6 binding; mutation of Met95 to Ala reduces binding of p6 or Ub-modified p6 similar50- to 300-fold, respectively, without affecting Ub binding . (ii) Some of the Gag protein associated with Tsg101 is ubiquitinated . (iii) The Ub-binding function of Tsg101 is required for the recognition and proper sorting of endocytic cargo . Thus, either or both Tsg101 binding functions could be involved in Gag trafficking. To determine the contributions of Tsg101 Ub binding and PTAP binding to Gag localization and release, the effects of FLAG-tagged Tsg101 proteins containing mutations in the Ub- and PTAP-binding regions were examined. wt and mutant Tsg101 expression was induced by transfection with pIND-h-Tsg101-FLAG, a construct that encodes the protein under the control of a modified promoter containing Drosophila ecdysone-responsive DNA elements. A high level of Tsg101-FLAG protein expression was obtained by treatment with the ecdysone analogue pa at 5 muM for 24 h, starting at 24 h after transfection. As described further below, a low level of Tsg101-FLAG expression also occurred in the absence of pa, presumably due to nonspecific or leaky promoter induction. This level was sufficient to inhibit VLP release . Figure shows the effects of the Tsg101 mutations on VLP release. As expected based on the results shown above (Fig. and ), cells cotransfected with DNA encoding Gag and the wt Tsg101-FLAG protein (Fig. , lane 3) produced less VLP than control cells (lane 1). Addition of the Tsg101 inducer pa 24 h after transfection increased this inhibitory effect (Fig. , lane 4). Expression of Tsg101-FLAG with a deletion in the Ub-binding pocket (Delta41-43) inhibited VLP release almost completely and to a significantly greater extent than expression of the wt FLAG-tagged Tsg101 protein in the absence (Fig. , lane 7) or presence (lane 8) of the inducer. The effect of Tsg101-FLAG containing a deletion in the PTAP-binding region (Fig. , lanes 11 and 12) was similar to the effect of wt Tsg101-FLAG expression. As expected, VLP release was blocked by p6 deletion (Fig. , lanes 2, 5, 6, 9, 10, 13, and 14). Western blot analyses of cytoplasmic extracts prepared from the transfected cells indicated that comparable levels of wt Gag (55-kDa), the Deltap6 precursor protein (similar50-kDa), or the p41 cleavage intermediate accumulated in the absence (Fig. , lanes 1 and 2) and presence (lanes 3 to 14) of Tsg101-FLAG expression. Levels of the CA-p2 protein in cells transfected with the Delta93-95 mutant were two- to fivefold higher than those in cells transfected with wt Tsg101-FLAG or the Delta41-43 mutant in three of three experiments (data not shown). This effect was not p6 dependent, and the significance is not known. Reprobing the blot with an antibody against Tsg101 showed that pa had induced comparable levels of the tagged proteins (Fig. , lanes 3, 5, 7, 9, 11, and 13). In each case, the Tsg101-FLAG protein was expressed at a higher level than endogenous Tsg101. Despite the inhibitory effect on VLP release observed in the absence of inducer , no tagged protein was detected by Western blot analysis under these conditions (Fig. , lanes 4, 6, 8, 10, 12, and 14). However, treatment of the cells with a fluorescently labeled secondary antibody targeted to the anti-FLAG primary antibody that recognized the tagged Tsg101 protein and examination by confocal microscopy revealed that the wt, Delta41-43, and Delta93-95 Tsg101-FLAG proteins were all expressed in the absence of pa . In agreement with the results obtained following coexpression of Gag-GFP and Tsg101-Myc , Tsg101-Gag complexes were detected in both peripheral and perinuclear regions of cells expressing Gag-GFP and the wt FLAG-tagged Tsg101 protein (Fig. , yellow fluorescence in merged image). In contrast, expression of the Delta41-43 mutant prevented colocalization in both peripheral and perinuclear regions (Fig. , lack of yellow fluorescence in merged image). Expression of the Delta93-95 mutant did not block perinuclear colocalization . Areas enclosed by dashed white rectangles in Fig. were deconvolved to enhance the resolution. Taken together with the inhibitory effects of these Tsg101 mutations on VLP release, the results suggest that the Ub- and PTAP-binding functions of Tsg101 both contribute to the trafficking of Gag to release sites at the plasma membrane and that they play distinct roles. DISCUSSION : Previous studies have shown that the p6 domain at the C terminus of Gag is the determinant of viral release by budding from the plasma membrane. Here, we provide evidence that the p6 region is responsible for accumulation of the structural precursor in a cellular fraction (P100) that was enriched in late-endosome, MVB, and plasma membrane components. Gag lacking p6 (Deltap6) was largely excluded from accumulating in this fraction, although the presence in the P100 fraction of cleavage intermediates derived from the Deltap6 precursor indicated that association occurs, consistent with the detection of the Deltap6 precursor at the plasma membrane in electron microscopic studies . The p6 region was also responsible for accumulation of cleavage intermediates and mature CA in a fraction that was enriched in early endosomes. We showed that the requirement for p6 for P100 accumulation was related to the interaction of p6 with Tsg101 by demonstrating that expression of full-length or truncated Myc-tagged Tsg101 proteins reduced Gag accumulation in the P100 fraction and inhibited Gag release, an effect similar to that of p6 deletion. The ability of one of the Tsg101 fragments (Tsg101 A) to inhibit Gag release has been established previously . All of the Tsg101 proteins containing the N-terminal binding region (Tsg101 A through D) interfered with P100 accumulation, suggesting a requirement for a region in Tsg101 between aa 313 and 391. This is the region of Vps28 binding . Nonbinding fragment E, but not nonbinding fragment F, also interfered with P100 accumulation, suggesting that the region in Tsg101 between aa 140 and 256 is also important. Analysis by confocal microscopy supported the notion that Tsg101-Gag interaction and release were linked, as Gag was sequestered in a perinuclear location when tagged Tsg101 was expressed. Much of the sequestered Gag protein appeared to be trapped in early endosomes, since tagged Gag and Tsg101 colocalized with perinuclear vesicles bearing tagged Hrs, a marker of that endocytic compartment. Because these observations were made under conditions of Tsg101 overexpression, our findings do not establish the early endosome as a site of Gag localization. However, the observation that the Ub-binding site in Tsg101 was critical for VLP release provides strong support for the hypothesis that Gag transits through early endosomes, because this Tsg101 function is required for the entry of cargo into endocytic compartments in both yeast and mammalian cells . Deletion of the entire N-terminal domain bearing the PTAP- and Ub-binding pockets interfered least with VLP release, indicating that the observed changes in subcellular location were due to events mediated through interaction with the PTAP- and Ub-binding regions. Expression of Tsg101-FLAG with a mutation in the PTAP-binding pocket produced a sequestering effect similar to that of wt Tsg101-Myc or Tsg101-FLAG. Expression of Tsg101-FLAG with a mutation in the Ub-binding pocket prevented Tsg101-Gag colocalization in perinuclear and membrane-proximal locations. This mutant blocked Gag release almost completely and functioned in a dominant-negative manner. Thus, it interfered extensively with the cellular machinery required for budding. The PTAP-binding site in Gag has been shown to be critical for recruitment of Tsg101 to the cell periphery . Indeed, in our study, disruption of the PTAP-binding pocket in Tsg101 by deletion of aa 93 to 95 blocked this recruitment . Interestingly, disruption of the Ub-binding pocket by deletion of aa 41 to 43 did not interfere with Gag or Tsg101 accumulation at the cell periphery but interfered with their colocalization . This suggests that the Ub-binding function contributes to productive recruitment. Furthermore, our finding that the Delta41-43 mutant interfered with VLP release to a significantly greater extent than the Delta93-95 mutant suggests that productive recruitment requires both events in perinuclear early endosomes and at the budding site on the plasma membrane. The Ub-binding function of Tsg101 has been shown to be critical for ESCRT-1 recognition and sorting of cargo in the endocytic trafficking pathway . Depletion of Tsg101's ESCRT-1 binding partner, Vps28, which might have occurred following Tsg101-FLAG or Tsg101-Myc expression in our studies, has been shown to result in endosomal accumulation of Ub-protein conjugates and to retard epidermal growth factor receptor trafficking to a compartment where it undergoes sorting . Perhaps the binding of ubiquitinated Gag through the PTAP-binding pocket of one Tsg101 molecule promotes interaction at the plasma membrane with a second Tsg101 protein through its Ub-binding pocket. Presumably the latter Tsg101 molecule is complexed to the membrane fusion machinery required for virus budding. Precisely how Tsg101 is involved in the membrane pinching-off and resealing events that result in the release of assembled Gag particles from the membrane is unclear. The possibility that Tsg101 controls the trafficking or function of a membrane-associated cellular protein that regulates the final exocytosis event is attractive. Tsg101 function is required for entry into the sorting MVB , and it is noteworthy that the invagination event associated with fusing with the MVB is the only known example where vesicle formation is directed away from the cytoplasm, as occurs during virus budding. Most likely, Gag, through its interaction with Tsg101, is linked to other Vps proteins that determine the formation of the sorting MVB and cause the machinery to direct invagination out of the cell of vesicles containing the maturing virus particles. If this strategy is indeed unique to the virus, the requirement for the interaction of Gag with Tsg101 may provide a potential target for interference with HIV propagation. FIG. 1. : p6 dependence of VLP release and intracellular Gag localization. p6 dependence of VLP release and intracellular Gag localization. VLPs (A) and cytoplasmic extracts (B) were prepared from cells transfected with wt gag or gag-pol (lanes 1 and 2) or Deltap6 gag or gag-pol (lanes 3 and 4). Duplicate VLP samples are shown. Cytoplasmic extracts derived from cells expressing wt (lanes 1 and 2) or Deltap6 (lanes 3 and 4) Gag were fractionated sequentially to obtain the P27 (lanes 1 and 3) and P100 (lanes 2 and 4) fractions. Proteins were identified by Western blotting with an anti-CA antibody. Molecular weight markers (in thousands) are given on the left. FIG. 2. : Effects of Tsg101 overexpression on VLP release and Gag accumulation in particulate cytoplasmic fractions. Effects of Tsg101 overexpression on VLP release and Gag accumulation in particulate cytoplasmic fractions. (A) (Left) Schematic diagram showing full-length and truncated Tsg101 proteins used in this study, as follows: fragment A, aa 1 to 160; fragment B, aa 1 to 256; fragment C, aa 1 to 312; Tsg101 protein D, aa 1 to 391; fragment E, aa 140 to 391; fragment F, aa 256 to 391. Filled square, Gag binding region in the UEV domain. (Center) Western blot analysis of cytoplasmic extracts for adventitious Tsg101 expression by use of an anti-Myc antibody. (Right) The blot was reprobed with an anti-Tsg101 monoclonal antibody. The antibody recognizes full-length Tsg101 and fragments B, C, and E. Molecular weight markers are shown to the left of the gels. (B) VLPs were isolated from the culture media of cells transfected with DNA encoding Gag or Gag-Pol and the empty vector (lane 1), full-length Tsg101-Myc (lane 5), or fragments of Tsg101-Myc (lanes 2 to 4, 6, and 7). (C) P27 (lanes 1, 3, 5, 7, 9, and 11) and P100 (lanes 2, 4, 6, 8, 10, and 12) fractions were isolated from cytoplasmic extracts prepared from the cells. FIG. 3. : Examination of P27 and P100 fractions for endogenous markers of endosomal compartments. Examination of P27 and P100 fractions for endogenous markers of endosomal compartments. P27 (lane 1) and P100 (lane 2) fractions prepared from cells expressing Gag were examined for endosomal protein markers by Western blotting using marker-specific antibodies. FIG. 4. : Confocal microscopy of cells expressing Gag-GFP and wt or N-terminally truncated Tsg101-Myc. Confocal microscopy of cells expressing Gag-GFP and wt or N-terminally truncated Tsg101-Myc. Shown are confocal images of fixed COS-1 cells transfected with Gag-GFP alone (A), Gag-GFP and wt Tsg101-Myc (B), or Gag-GFP and N-terminally truncated Tsg101-Myc (fragment E) (C). Serial z sections from the adherent cell surface through the nucleus are shown for the region indicated by the dashed white square. In panel B, view 1 corresponds to the upper white square and view 2 corresponds to the lower white square. Green, Gag; red, Tsg101-Myc; blue, nuclei stained with DAPI. FIG. 5. : Confocal microscopy of cells expressing Gag-GFP, Hrs-FLAG, and Tsg101-YFP. Confocal microscopy of cells expressing Gag-GFP, Hrs-FLAG, and Tsg101-YFP. Shown are confocal images of fixed COS-1 cells transfected with Gag-GFP and Hrs-FLAG (A) or Gag-GFP, Hrs-FLAG, and Tsg101-YFP (B). Cells expressing Gag-GFP and Hrs-FLAG contained fluorescence in peripheral and perinuclear regions, and both regions are shown. Fluorescence was exclusively perinuclear in cells expressing Gag-GFP, Hrs-FLAG, and Tsg101-YFP. The area within the dashed white square in panel A was deconvolved and enlarged to enhance visualization of Hrs-Gag colocalization. Green, Gag; red, Tsg101-YFP; blue, Hrs-FLAG. There is some nonspecific staining of nuclei (n) with the anti-FLAG antibody. FIG. 6. : Effects of N-terminal Tsg101 mutations on VLP release. Effects of N-terminal Tsg101 mutations on VLP release. VLPs were isolated from the media of cells expressing wt Gag (lanes 1, 3, 4, 7, 8, 11, and 12) or the Deltap6 precursor protein (lanes 2, 5, 6, 9, 10, 13, and 14). FIG. 7. : Effects of N-terminal Tsg101 mutations on Gag and Tsg101 accumulation in the cytoplasm. Effects of N-terminal Tsg101 mutations on Gag and Tsg101 accumulation in the cytoplasm. Cytoplasmic extracts were prepared from cells expressing Gag (lanes 1, 3, 4, 7, 8, 11, and 12) or the Deltap6 protein precursor (lanes 2, 5, 6, 9, 10, 13, and 14) and either wt Tsg101-FLAG or the indicated Tsg101-FLAG mutant and were analyzed by Western blotting using an anti-CA (A) or anti-Tsg101 (B) antibody. FIG. 8. : Effects of N-terminal Tsg101 mutations on Gag and Tsg101 distribution. Effects of N-terminal Tsg101 mutations on Gag and Tsg101 distribution. Shown are confocal images of COS-1 cells transfected with Gag-GFP and either wt Tsg101-FLAG (A), Delta41-43 Tsg101-FLAG (B), or Delta93-95 Tsg101-FLAG (C) in the absence of pa. Red fluorescence indicates leaky (uninduced) Tsg101-FLAG expression from the transfected tsg101 DNA. Areas within dashed white rectangles were deconvolved to enhance the resolution. Green, Gag; red, Tsg101-FLAG; yellow, colocalization. Backmatter: PMID- 12915578 TI - Chromosomal Distribution of Endogenous Jaagsiekte Sheep Retrovirus Proviral Sequences in the Sheep Genome AB - A family of endogenous retroviruses (enJSRV) closely related to Jaagsiekte sheep retrovirus (JSRV) is ubiquitous in domestic and wild sheep and goats. Southern blot hybridization studies indicate that there is little active replication or movement of the enJSRV proviruses in these species. Two approaches were used to investigate the distribution of proviral loci in the sheep genome. Fluorescence in situ hybridization (FISH) to metaphase chromosome spreads using viral DNA probes was used to detect loci on chromosomes. Hybridization signals were reproducibly detected on seven sheep chromosomes and eight goat chromosomes in seven cell lines. In addition, a panel of 30 sheep-hamster hybrid cell lines, each of which carries one or more sheep chromosomes and which collectively contain the whole sheep genome, was examined for enJSRV sequences. DNA from each of the lines was used as a template for PCR with JSRV gag-specific primers. A PCR product was amplified from 27 of the hybrid lines, indicating that JSRV gag sequences are found on at least 15 of the 28 sheep chromosomes, including those identified by FISH. Thus, enJSRV proviruses are essentially randomly distributed among the chromosomes of sheep and goats. FISH and/or Southern blot hybridization on DNA from several of the sheep-hamster hybrid cell lines suggests that loci containing multiple copies of enJSRV are present on chromosomes 6 and 9. The origin and functional significance of these arrays is not known. Keywords: Introduction : Jaagsiekte sheep retrovirus (JSRV) is an exogenous betaretrovirus, responsible for ovine pulmonary adenocarcinoma in sheep. The sheep genome contains a family of endogenous retroviral sequences (enJSRV) closely related to JSRV . Interestingly, the integration pattern, as assayed by Southern blot hybridization, does not vary much among a wide range of domestic sheep breeds and even among wild sheep, including bighorn sheep and mouflon sheep, the likely ancestor of domestic breeds. Thus, it appears that these viruses became fixed in the genome early in the development of the genus Ovis and have not shown much recent movement. Closely related endogenous viruses are also found in the genus Capra, which includes wild and domestic goats. Indeed related viruses were found in several other genera of the subfamily Caprinae of the family Bovidae . Much of the interest in this family of viruses has been directed at developing ways to distinguish exogenous JSRV from enJSRV, and to this end restriction maps and partial sequences were derived for several of the enJSRV loci . This led to the identification of diagnostic restriction endonuclease cleavage sites and hybridization probes that allowed the isolation of two infectious clones of JSRV . Recently, three complete enJSRV proviral loci have been sequenced . These were defective in producing virus particles and differed from infectious JSRV by numerous point mutations and deletions. Sequence differences between the long terminal repeats (LTRs) of JSRV and enJSRV seem to be particularly important in determining the differences in expression in different tissues. While JSRV LTRs appear to be most active in lung cells, the enJSRV LTRs seem to be most active in the uterus . Further identification and characterization of the enJSRV loci will be necessary in determining their functional significance. We have sought insight into the distribution of endogenous proviral loci in sheep chromosomes by using two approaches to mapping. In the first approach, fluorescence in situ hybridization (FISH) was used to localize the endogenous viral loci to sheep and goat chromosomes. The second approach made use of a recently developed panel of sheep-hamster hybrid cell lines. Each of the cell lines in this panel contains one or a few sheep chromosomes, and collectively the panel contains the entire sheep genome. This panel allows one to study individual sheep chromosomes in the absence of the rest of the genome, and it was used to detect enJSRV loci by PCR and Southern blotting. Using these approaches we report that the distribution of enJSRV proviruses on the genome is unusual because two chromosomes carry loci with several copies of the provirus. MATERIALS AND METHODS : Ovine and caprine cell cultures. | Metaphase spreads were prepared from caprine and ovine cell lines: CAT2 (caprine testis-derived fibroblasts with a 6/15 translocation ), SGH2 (skin fibroblasts derived from a fertile ewe-goat hybrid ), MFP2 (fetal lung-derived fibroblasts) obtained from H. Hayes, TIGEF (T-immortalized goat embryonic fibroblast) developed in our laboratory , and IDO5 (sheep dermal fibroblasts), generously given by Chappuis, Institut Merieux, Lyon, France. H921 fibroblasts were obtained from explants of cardiac muscle of a sheep with lung tumor and maintained in culture for six passages. OF686 fibroblasts emerged from epithelial culture of a nasal tumor from sheep. CAT2, SGH2, and MPF2 were cultured in complete minimal essential medium (MEM) (MEM [GIBCO BRL] supplemented with 2 mM l-glutamine [GIBCO BRL], 1% antibiotic-antimycotic solution with 100 U of penicillin/ml, 100 mug of streptomycin/ml, and 250 ng of amphotericin B [GIBCO BRL[/ml, and 10% decomplemented fetal calf serum [GIBCO BRL]) at 37C with 5% CO2. TIGEF, IDO5, OF686, and H921 were cultured in complete Dulbecco's modified Eagle medium (DMEM) medium (DMEM medium [GIBCO BRL], 2 mM l-glutamine, 10% decomplemented fetal calf serum, 10 U of penicillin/ml and 10 mug of streptomycine/ml). Chromosomal localization of enJSRV. (i) Preparation of metaphase spreads. | Metaphase spreads were obtained by a double thymidine synchronization as previously described . Briefly, on day 1, cells were seeded in complete medium and incubated at 37C, with 5% CO2. At 30 to 40% confluence, cells were incubated in complete medium supplemented with 0.8 mg of thymidine (Sigma)/ml for 14 to 15 h. On day 2, cells were extensively washed three times for 20 min with MEM and incubated for 8 h with complete medium. A second thymidine synchronization was performed using the same conditions. On day 3, cells were washed with MEM medium and incubated in complete medium supplemented with 10 mug of bromodeoxy-uridine (Sigma)/ml to induce R-banding. After an hour of incubation, 0.5 mug of fluorodeoxyuridine (Sigma)/ml was added to the culture. Mitotic cells appeared as round cells at the surface of the culture 6 to 8 h after bromodeoxy-uridine treatment and were mechanically detached from the plastic. Supernatants were harvested and mitotic cells were collected by a 10-min centrifugation at 430 x g. The pellets were washed once in phosphate-buffered saline (PBS) without calcium or magnesium. Approximately 1.5 x 106 cells were incubated for 13 min at 37C in 10 ml of a 0.22 mum-pore-size-filtered hypotonic solution containing 16% decomplemented fetal calf serum and 2 mug of EDTA/ml in pure sterile water. The mitotic nuclei were prefixed for 5 min at room temperature by adding 1 ml of 3:1 methanol-acetic acid solution and were centrifuged for 10 min at 430 x g. The pellets were washed three times after a 10-min incubation at room temperature in 10 ml of 3:1 methanol-acetic acid solution and then kept overnight at 4C. Finally, nuclei were resuspended in 200 to 500 mul of 3:1 methanol-acetic acid solution, spread on cold slides, and stored at -20C. Culture flasks were vigorously shaken, and the cells not attached to the plastic were recovered by centrifugation. (ii) FISH. | The JSRV plasmid , containing the full-length genome of the South-African strain of JSRV, generously provided by Gilles Querat (Marseille, France), was used to generate the probe. The genomic biotinylated 7.5-kb JSRV probe was obtained by random priming as previously described ; briefly, 200 ng of denatured JSRV DNA was incubated for 2 h at 37C in 50 mul of OLBF buffer (50 mM tris-HCl [pH 8], 5 mM MgCl2, 200 mM HEPES [pH 6.6], 5.4 U of Pd(N6) hexanucleotides (Boehringer, Meylan, France)/ml, 10 mM 2-B-mercaptoethenol), containing 40 muM (each) dATP, dCTP, and dGTP (Pharmacia), 20 muM biotin 11-dUTP (Sigma), 652 mug of bovine serum albumin (Sigma)/ml, and 2.5 U of Klenow fragment (Eurogentech). After a 2-h incubation at 37C, the reaction was stopped by addition of 42 mM EDTA. The JSRV probe was coprecipitated with 4 mug of sonicated salmon sperm DNA and 20 mug of sonicated sheep competitor DNA, dissolved in 20 mul of hybridization buffer (50% formamide, 10% dextran sulfate in SSCP buffer [pH 7] [40 mM NaH2PO4-Na2HPO4, 240 mM NaCl, 30 mM Na3C6H5O7 2H2O]), and heat denatured. Hybridization was performed as previously described with minor modifications. Slides were washed an hour at room temperature in 2x SSC (300 mM NaCl, 30 mM Na3C6H5O7 [pH 7]) and treated with 100 mug of RNase A (Sigma)/ml in 2x SSC (Sigma) for 1 h at 37C. After a 15-min wash in 2x SSC at room temperature, the slides were dehydrated in an ethanol series (70, 80, and 90%). The metaphase spreads were denatured by a 2-min incubation at 72C in 70% formamide in 2x SSC and dehydrated in an ethanol series at -20C. After a 2-h incubation at 37C in hybridization buffer (50% formamide, 10% dextran sulfate in SSCP buffer), the denatured JSRV probe was added to the slides and they were covered with "steam-cooking" paper. Slides were hybridized overnight at 37C in a moist chamber and then successively washed at room temperature as follows: 15 min in 2x SSC; three times for 3 min in 50% formamide-2x SSC; four times for 2 min in 2x SSC; and four times for 2 min in 1x BN buffer (0.5 M NaHCO3, 25 ml of IGEPAL (Sigma)/liter, and 0.1 mg of sodium azide [pH 8]/ml). Slides were then incubated for 10 min at room temperature in 1x BN supplemented with 5% nonfat dry milk, followed by 30 min at 37C in a dark, moist chamber with 50 mul of a 1/400 dilution of avidin-FITC (Vector) in 1x BN supplemented with 5% nonfat dry milk, followed by three washes for 2 min in 1x BN at 42C. After a 10-min incubation, signal amplification was performed with 50 mul per slide of a 1/20 dilution of goat serum (Vector) in 1x BN. Each slide was incubated for 30 min in the dark at 37C with 50 mul of a 1/20 dilution of biotinylated antiavidin antibodies (Vector) in 1x BN and washed three times for 2 min at 42C in 1x BN. After 10 min at room temperature in 1x BN supplemented with 5% nonfat dry milk, a second treatment with avidin-FITC was done under the same conditions. Slides were washed in PBS and stained in the dark with 60 mul of 0.5% propidium iodide in PBS for 10 min. R-banding was revealed by addition of 10 mul of antifade mixture (90% glycerol, 0.1% p-phenylene diamine dihydrochloride sigma [pH 11] in PBS) . The slides were examined with a Zeiss MCD80DX microscope coupled with a charge-coupled device camera, and metaphases were analyzed with the CytoVision System (Applied Imaging). Chromosomes were identified according to the new International System for Chromosome Nomenclature of Domestic Bovids . Sheep-hamster somatic cell hybrid lines. | A panel of 30 sheep-hamster cell lines was obtained from the Eleanor Roosevelt Institute for Cancer Research, Denver, Colo. A full description of the construction and characterization of these lines is found in the work of Burkin et al. . Chinese hamster ovary cells and the hybrid cell lines were grown at 37C in Ham's F-12 media supplemented with 5% fetal bovine serum to generate enough cells for cryopreservation and DNA isolation. DNA isolation. | Each cell line was grown to confluency in a roller bottle, trypsinized, and then incubated at 50C for 3 h in 10 ml of extraction buffer containing 10 mM Tris-Cl (pH 8.0), 0.1 mM EDTA (pH 8.0), 20 mug of pancreatic RNase/ml, 0.5% sodium dodecyl sulfate (SDS), and 100 mug of proteinase K/ml. DNA from lysed cells was extracted twice in equilibrated phenol (pH 8.0), precipitated by addition of 0.7 vol of isopropanol, spooled with a glass rod, and resuspended in Tris-EDTA. DNA was also isolated from a normal sheep lung (98AO1) for use as a control. gag PCR. | Each 50-mul reaction consisted of 35 ng of genomic DNA, standard 1x PCR buffer, 200 muM deoxynucleoside triphosphates, 20 pmol (each) of gag primer (sense [P1], 5'-GCTGCTTTRAGACCTTATCGAAA-3', and antisense [P2], 5'-ATACTGCAGCYCGATGGCCAG-3') and 1.5 U of Taq polymerase. The thermocycler was programmed for an initial cycle of 94C for 3 min and then 30 cycles of 94C for 30 s, 58C for 30s, and 72C for 30 s, with a final extension of 72C for 5 min. PCR products were analyzed by electrophoresis in 1.5% agarose in 1x Tris-borate EDTA. Southern blot hybridization. | Ten micrograms of each genomic DNA was cleaved with EcoRI, XbaI, or BamHI, and fragments were separated on a 0.7% agarose gel and transferred by a PosiBlot 30-30 pressure blotter (Stratagene) to a nylon membrane (MSI, Westborough, Mass.) as described previously . The DNA probe was generated by PCR with P1 and P2 primers for JSRV capsid from gel-purified fragments with the addition of [alpha32-P]dCTP to the reaction. Each membrane was hybridized with 32P-labeled capsid probe with a specific activity of 106 cpm of probe/ml of UltraHyb (Ambion) at 42C. Following hybridization, the membrane was washed at 42C twice for 5 min (2x SSC, 0.1% SDS) and then twice for 15 min (0.1% SSC, 0.1% SDS) followed by exposure to a PhosphorImager screen for either 5 h or 2 weeks by autoradiography. RESULTS : Southern blot hybridization on sheep genomic DNA cut with any of several restriction enzymes generally shows multiple bands of widely varying intensity when probed with capsid or envelope probes The dark bands presumably represent dropout fragments conserved in several endogenous proviruses. Southern blots such as these with a wide variety of restriction enzymes led to the conclusion that there are 15 to 20 enJSRV proviruses in the sheep and goat genomes , but they do not provide any information on their distribution in the sheep or goat genome. One way of examining the distribution of proviruses in the genome is by FISH. The enJSRV chromosomal distribution in goat (2n = 60) and sheep (2n = 54) chromosomes was analyzed by FISH on four ovine cell lines, two caprine cell lines, and one goat-sheep hybrid cell line. Chromosome identification was carried out using the R-banding karyotype. The positions were confirmed on multiple metaphase spreads for each cell line. For a given cell line, only the sites reproducibly found are reported. Using a biotinylated full-length JSRV genome as a probe, enJSRV copies were consistently localized on the two pairs of metacentric chromosomes 1 and 2 and on the acrocentric chromosomes 6, 8, 10, and 14 in sheep and on the acrocentric chromosomes 1, 2, 3, 6, 7, and 18 in goats (Fig. and B; Table ) Except for positions 1q45 and 2q41, enJSRV copies were integrated in different sites in goat and sheep genomes. All the species-specific copies were confirmed by the use of a goat-sheep hybrid cell line derived from a fertile goat-sheep hybrid (Fig. ; Table ). In the ovine cell lines, the signal was clearly stronger in position q12 of chromosome 6, suggesting a multicopy enJSRV locus. This was confirmed in ovine cell lines MPF2, OF686, H921, and IDO5 . Depending on the quality of the metaphase spreads and of the hybridization, three to four individual hybridization sites were observed. enJSRV was absent from the equivalent position in caprine chromosome 2. Additional sites were present in position 5q15 in the OF686 cell line and position 2q12.2 in the H921 cell line . Despite the fact that these two cell lines were derived from nasal tumor and heart of a sheep with pulmonary adenocarcinoma, no exogenous betaretrovirus was detected by PCR and nested PCR (data not shown), suggesting that the additional copies were of endogenous origin. Another way of examining the distribution of proviruses in the sheep genome is to make use of the panel of sheep-hamster cell lines constructed by fusing sheep cells with Chinese hamster ovary cells . Total genomic DNA was extracted from each of the cell lines and used as a template for PCR with primers from the gag region of JSRV. As indicated in Table products of the expected size (229 bp) were amplified from all of the lines except R928-6 A3 (chromosomes 16 and 21), R891-16R4B (chromosome 16), and R891-16R5A (chromosome 17). Since many of the lines contain more than one sheep chromosome, it is not possible to determine which of the chromosomes in these lines carry proviruses. However, from these data it can be concluded that at least chromosomes 1, 2, 3, 4, 6, 7, 8, 9, 10, 13, 14, 25, 26, and X carry gag sequences of proviral loci. Three endogenous proviruses have been cloned and sequenced . These have been designated enJS56A1, enJS59A1, and enJS5F16. Sheep genomic sequences adjacent to the proviruses were determined for each of these clones (GenBank accession numbers , , and , respectively). Primers were designed for these sequences and were used in PCR with DNA from each of the hybrid cell lines as a template. By determining which lines gave amplification products, it was possible to assign enJS56A1 to chromosome 6 and enJS59A1 to the X chromosome. The enJS5F16 flanking primers amplified a product from line R894-25C, which carries chromosomes 12, 13, and 15. It did not amplify a product from line R891-29R8A, which carries chromosomes 13 and 19, or line R891-27R1B, which carries chromosome 13. Thus, either chromosome 12 or chromosome 15 must carry enJS5F16 (data not shown). The assignment of enJS56A1 to chromosome 6 was further confirmed by FISH on metaphase chromosomes with a probe derived from the DNA flanking the provirus. This probe colocalized with virus probes that hybridize with the multicopy locus (data not shown). In order to further examine the loci on individual chromosomes, DNA preparations from sheep-hamster cell lines that contain a single sheep chromosome were digested with EcoRI and subjected to Southern blot hybridization with a capsid probe All of the available JSRV and enJSRV sequences have an EcoRI site just downstream of the beginning of the pol open reading frame. The capsid probe should therefore detect junction fragments between the 5' LTR and sheep DNA. Hybridization results indicate that cell lines had bands of various sizes, suggesting one or more sites of random enJSRV integration in the sheep chromosomes. No background hybridization was evident with the Chinese hamster ovary cell line (negative control) or with the lines containing chromosomes 16 and 17, which were negative for enJSRV by PCR . The cell lines containing chromosomes 6 and 9 had bands, which migrated at approximately 6 kb, with intensities 5- to 10-fold higher than those of the other bands. This suggests that chromosomes 6 and 9 contain multiple copies of enJSRV that have the same size restriction fragment hybridizing to the capsid probe. A possible explanation for this is that the copies form part of a tandemly repeated array of enJSRV genomes. In order to test this, DNA samples from the cell lines containing chromosomes 6 and 9 were then cleaved with EcoRI, BamHI, or XbaI and subjected to Southern blot hybridization. None of the available provirus sequences contains an XbaI site; therefore, the number of XbaI bands that hybridize with the probe should indicate the number of integrations. For chromosome 6 and 9, there are single XbaI bands that migrate in the high-molecular-weight region of the gel where the size resolution is poor . Therefore, all of the proviruses appear to reside on large XbaI fragments. All of the available provirus sequences contain a single BamHI site near the end of pol. Thus, left-end junction fragments and full-length copies of the putative tandem repeat sequence should hybridize with the gag probe. For chromosome 6, there is a strong band of hybridization around 6 kb and two weaker bands of hybridization at higher molecular weights . This would be consistent with a tandem array of 6-kb units and another locus. For chromosome 9, all three digests gave a single strong band of hybridization. This is again consistent with a single tandem array of 6-kb repeats. The junction fragments either are not resolved from the 6-kb band in the EcoRI and BamHI digests or they are two dilute to be visualized in these blots. These results suggest that chromosomes 6 and 9 each have several enJSRVs integrated in a tandem fashion. The 6-kb bands seen for chromosomes 6 and 9 correspond well with prominent bands seen in EcoRI and BamHI digests of sheep genomic DNA . DISCUSSION : The presence of an endogenous retrovirus family related to JSRV has been known since the first JSRV nucleic acid probes were used on Southern blots of sheep genomic DNA. The distribution of the proviruses in the sheep genome appears to be quite stable, since the pattern of hybridization shows little variation among domestic sheep breeds and even among wild sheep . This suggests that there has not been much movement of these viruses for many thousands of years. This is quite different from the type C mouse endogenous viruses, which vary considerably from strain to strain . The Southern blot pattern, however, does not provide information on the location of proviruses on the sheep genome or allow the identification of individual loci, which may be of considerable interest in assessing their possible effects on the host physiology or their potential for interaction with exogenous JSRV. For example, there is a high level of expression from one or more enJSRV loci in the reproductive tract of the ewe . However, little is known about which loci are expressed, and the significance of this expression can only be speculated at this time. The chromosomal sites of enJSRV were physically mapped by FISH in sheep and goat chromosomes using a biotinylated full-length JSRV probe. Using seven ovine and caprine cell lines, we consistently mapped enJSRV to seven chromosomal positions in sheep and eight in goat. Only two chromosomal positions (1q45 and 2q41) are common in the sheep and goat genomes, suggesting that some of the enJSRV copies entered into the small ruminants genomes after the speciation. The 7.5-kb probe we used reliably detected single proviral copies. A multicopy locus was identified on ovine chromosome 6, in position q12-13, with clear multiple sites of hybridization probably corresponding to at least three or four enJSRV sequences. Additional insight into the distribution of viruses among the chromosomes was gained by determining whether a gag PCR product could be amplified from a panel of sheep-hamster cell lines, each of which carries one or a few sheep chromosomes. The results indicate that at least half of the sheep chromosomes carry at least one copy of the gag portion of the proviral genome. This is consistent with what is predicted by the Poisson distribution for random integration of 15 to 20 endogenous proviruses on the 28 chromosomes of the sheep genome. All of the chromosomes identified as having proviral loci by FISH were also identified by PCR on the sheep-hamster hybrids. It is not surprising that more proviral loci were identified by PCR, since it should be a more sensitive technique due to the exponential amplification of the signal. In addition, the PCR primers were designed to detect the relatively conserved gag sequences of the virus. The FISH hybridization probe was a full-length clone of the exogenous JSRV genome and may not be sensitive enough to reliably detect truncated and/or relatively divergent enJSRV genomes. An unexpected aspect of the enJSRV distribution in the sheep genome is the presence of two loci, which apparently contain several copies of a proviral genome. Two lines of evidence suggest that they may exist in a tandem repeat configuration. First, prior to the use of high-resolution R-banding for karyotyping the sheep chromosomes as described above, two to four loci were consistently detected by FISH on sheep metaphase chromosomes with probes of 2 and 4 kb, representing the 5' and 3' portions of the genome, respectively (data not shown). These probes did not reliably detect single proviral copies but readily detected these loci, suggesting multiple copies of provirus. Second, Southern blot analysis of DNA from sheep-hamster cell lines that carry either chromosome 6 or chromosome 9 suggests that multiple copies of enJSRV are present and that they may be organized in a head-to-tail tandem array. This type of organization in retroviral proviruses is unusual but not unprecedented. For example, a tandem HERV-K proviral locus has recently been described in humans , and a tandem duplication of avian sarcoma virus has been found in transformed quail cells . The significance of the arrays must await their further characterization and the determination of whether or not they are transcribed and expressed at some time during the life of the sheep. FIG. 1. : enJSRV sequences in the sheep genome. enJSRV sequences in the sheep genome. Sheep genomic DNA was digested with BamHI (B), EcoRI (E), HindIII (H), PstI (P), SacI (S), or XbaI (X) and subjected to agarose gel electrophoresis and Southern blot hybridization using a JSRV capsid region probe. FIG. 2. : FISH analysis with a full-length JSRV probe on ovine and caprine metaphase spreads. FISH analysis with a full-length JSRV probe on ovine and caprine metaphase spreads. Chromosomes were identified by R-banding, and integration sites are indicated by arrows. (A) Ovine MPF2 cells (fetal lung-derived fibroblasts). (B) Ovine OF686 cells. (C) Sheep-goat hybrid SGH2 cells. (D) Multicopy locus on position q12 of ovine chromosome 6 from four different lines. FIG. 3. : Analysis by Southern blot hybridization on sheep-hamster hybrid cell lines. Analysis by Southern blot hybridization on sheep-hamster hybrid cell lines. (A) DNA from cell lines containing single sheep chromosomes was digested with EcoRI, and Southern blots were probed with a JSRV capsid region probe. The sheep chromosome contained by the cell line is shown above the lane. (B) Sheep genomic DNA (gDNA) and DNA from hybrid cell lines containing single sheep chromosomes 6 or 9 were digested with XbaI (X), EcoRI (E), or BamHI (B), and Southern blots were probed with a JSRV capsid region probe. TABLE 1 : enJSRV distribution in ovine and caprine chromosomes determined by fluorescent in situ hybridization TABLE 2 : Detection of enJSRV loci in sheep-hamster hybrid cell lines by PCR Backmatter: PMID- 12915527 TI - Reovirus Receptors and Pathogenesis AB - Keywords: null: null REOVIRUS ATTACHMENT PROTEIN sigma1 BINDS TO CELL SURFACE CARBOHYDRATE AND JAM1 : The sigma1 protein is a fibrous trimer consisting of an elongated tail domain that inserts into the virion and a globular head domain that projects away from the virion surface . The recently determined crystal structure of the C-terminal half of T3D sigma1 reveals that the tail is formed in part by a triple-beta spiral and that the head is formed by a compact eight-stranded beta-barrel . T3 sigma1 contains receptor-binding domains in both the tail and head regions. A domain in the tail binds alpha-linked sialic acid , whereas a domain in the head binds junctional adhesion molecule 1 (JAM1) . In T3D sigma1, these domains are dissociable by treatment of sigma1 with intestinal proteases, such as trypsin or chymotrypsin , which likely accounts for the attenuated virulence of this strain after oral inoculation . The T1 sigma1 tail also binds cell surface carbohydrate , but this molecule has not been identified. The capacity of T3 reovirus to bind sialic acid influences infection of cultured cells. Both T1 and T3 reoviruses can infect L929 cells, a murine fibroblast cell line commonly used to propagate reovirus. However, only T3 strains can infect murine erythroleukemia (MEL) cells . This growth restriction is sialic acid dependent , and serial passage of non-sialic-acid-binding T3 strains in these cells results in selection of viruses that have acquired the capacity to bind sialic acid . The MEL-adapted phenotype is conferred by single point mutations in a region of the sigma1 tail implicated in sialic acid binding . Although the sialic acid-binding region of T3D sigma1 was not included in the crystal structure , molecular modeling suggests that this region is contained within the triple beta-spiral . Substantial evidence has accumulated to suggest that the sigma1 head binds to proteinaceous receptors on the cell surface . A flow cytometry-based expression-cloning approach was used to identify such molecules by use of a non-sialic-acid-binding strain as an affinity ligand . A neural precursor cell (NT2) cDNA library was selectively enriched by fluorescence-activated cell sorting for cDNAs that confer binding of fluoresced virions to transfected cells. Four clones were identified that conferred virus binding to all transfected cells. Each encoded JAM1, a member of the immunoglobulin superfamily postulated to regulate formation of intercellular tight junctions . JAM1-specific monoclonal antibodies inhibit reovirus binding and infection, and expression of JAM1 in nonpermissive cells rescues reovirus growth . Most importantly, the sigma1 protein binds directly to JAM1 with an apparent Kd of similar6 x 10-8 M , providing confidence that JAM1 is a reovirus receptor. Surprisingly, JAM1 serves as a receptor for both prototype and field-isolate strains of all three reovirus serotypes (; J. A. Campbell and T. S. Dermody, unpublished observations). Therefore, JAM1 does not appear to explain the serotype-dependent differences in reovirus tropism in the murine CNS. These observations suggest that proteinaceous receptors other than JAM1, or unique carbohydrate-based coreceptors, influence reovirus pathogenesis. SIALIC ACID AS A DETERMINANT OF REOVIRUS ATTACHMENT AND DISEASE : To dissect the contribution of sialic acid to T3 reovirus attachment, infection, and disease, the S1 gene segments of a MEL-adapted strain, T3C44-MA, and its non-sialic-acid-binding parental strain, T3C44, were introduced into the genetic background of T1L by reassortment . The resultant viruses, termed T3SA+ and T3SA-, respectively, differ by a proline-to-leucine substitution at amino acid 204 in sigma1 which confers the capacity to bind sialic acid . HeLa cells are permissive for infection by both T3SA+ and T3SA-; however, viral yields from a single replication cycle are significantly higher for T3SA+ when HeLa cells are infected with equivalent multiplicities of infection of these strains . This enhanced growth is sialic acid dependent, as removal of cell surface sialic acid by neuraminidase or incubation of virions with the soluble sialic acid analog sialyllactose (SLL) decreases yields of T3SA+ to the same levels as those of T3SA- . Radioligand binding studies using T3SA+ and T3SA- indicate that the capacity to bind sialic acid enhances the association rate of virus for cells and increases the avidity of binding . Kinetic analyses using inhibitors of sialic acid and JAM1 binding demonstrate that sialic acid is engaged first in the adsorption process, as the inhibitory effect of SLL on infection by T3SA+ occurs at early but not late time points. However, a sigma1-specific monoclonal antibody that blocks virus binding to JAM1 inhibits viral infectivity at both early and late times during adsorption. These data suggest that reovirus attaches to cells via an adhesion-strengthening mechanism by which initial low-affinity binding to sialic acid facilitates secondary higher-affinity binding to JAM1 . In addition to attaching virus to the cell surface, the capacity of T3 reovirus to bind sialic acid influences tissue tropism and disease phenotypes. Both T3SA+ and T3SA- grow well in the intestine following peroral inoculation of newborn mice and disseminate to distant sites . However, T3SA+ produces higher titers in the brain and liver at early times after inoculation than those produced by T3SA-. Importantly, by day 12 following inoculation, though, titers of both viruses are equivalent in all organs tested . These findings correlate well with observations made using cultured cells and suggest that the enhancement of viral attachment due to sialic acid binding also occurs in vivo. Despite the differences in the kinetics of spread exhibited by T3SA+ and T3SA- following infection of newborn mice, the disease phenotypes associated with these strains differ dramatically. Animals infected with T3SA+ but not T3SA- develop jaundice, steatorrhea, and oily fur . These findings are consistent with previous studies of reovirus infections of mice and correlate with injury to bile duct epithelium (-, ). Histological analysis indicates that T3SA+ and T3SA- vary strikingly in the pattern of liver injury in infected mice . Liver sections from animals infected with T3SA+ demonstrate a robust inflammatory response concentrated in the portal areas, whereas liver sections from animals infected with T3SA- contain only mild inflammatory infiltrates . Concordantly, immunohistochemical staining reveals viral antigen in bile duct epithelial cells of animals infected with T3SA+ . Viral antigen is also present in liver tissue of animals infected with T3SA-, but the antigen is primarily localized to hepatocytes . These findings suggest that utilization of sialic acid as a coreceptor targets reovirus to bile duct epithelial cells. Interestingly, the disease produced by T3SA+ in mice is similar in some respects to biliary atresia in human infants, which has been linked to reovirus in approximately 50% of cases in one study . As such, infection of newborn mice by sialic acid-binding reoviruses may have utility as a model to dissect the pathogenesis of bile duct injury in humans. REOVIRUS ATTACHMENT AND DISASSEMBLY ARE REQUIRED FOR INDUCTION OF APOPTOSIS : In addition to conferring viral attachment, engagement of reovirus receptors also induces postbinding signaling events that may influence disease pathogenesis. Reovirus induces apoptosis in cultured cells and in vivo . Insight into mechanisms by which reovirus elicits apoptosis was first developed from studies of viral prototype strains that vary in the capacity to induce this cellular response. T3D induces apoptosis to a greater extent than T1L in L cells , Madin-Darby canine kidney (MDCK) cells , and HeLa cells . Differences in the capacity of these strains to induce apoptosis are determined primarily by the viral S1 gene ; however, the M2 gene makes a secondary contribution to the magnitude of the apoptotic response . Linkage of the S1 gene to the efficiency of reovirus-induced apoptosis suggests that the capacity of attachment protein sigma1 to bind different receptors regulates proapoptotic signaling. Sialic acid-binding strain T3SA+ induces apoptosis to a much greater extent than non-sialic-acid-binding strain T3SA- in both HeLa cells and L cells . Removal of cell surface sialic acid with neuraminidase or blockade of virus binding to sialic acid by SLL abolishes the capacity of T3SA+ to induce apoptosis . These findings indicate that the capacity of T3 reovirus to bind sialic acid significantly enhances its capacity to induce an apoptotic response. However, JAM1 also plays a critical role in apoptosis. Although T3SA+ can bind and enter cells via a JAM1-independent pathway mediated by binding sialic acid, T3SA+ is incapable of inducing apoptosis in the absence of JAM1 binding . Thus, reovirus binding to both sialic acid and JAM1 is required to induce maximal levels of apoptosis. Since T1 and T3 reoviruses bind different carbohydrate coreceptors, it is possible that serotype-specific differences in disease are attributable to differences in the induction of apoptosis in vivo. Although ligation of sialic acid and JAM1 is necessary for apoptosis induced by reovirus, viral attachment to the cell surface alone is not sufficient. Inhibitors of acid-dependent viral disassembly block apoptosis by reovirus , indicating a requirement for postattachment entry steps. However, viral transcription is dispensable, as inhibitors of viral RNA synthesis do not diminish the capacity of reovirus to induce apoptosis . Linkage of the M2 gene with the efficiency of apoptosis induction by reovirus provides additional support for the idea that viral entry steps are required for proapoptotic signaling . The M2 gene encodes the mu1 protein, which mediates penetration of the virus into the cytosol following viral disassembly in endosomes . These findings suggest that receptor binding and disassembly must occur within the same cellular compartment to elicit an apoptotic response. It is also possible that virus-receptor interactions couple with mu1-mediated membrane penetration to elicit apoptosis . A critical component of the signaling cascade that leads to apoptosis of reovirus-infected cells is the transcription factor NF-kappaB, which is known to play important roles in regulating cellular stress responses, including apoptosis . Reovirus activates NF-kappaB in several cell lines, including L cells, MDCK cells, and HeLa cells . NF-kappaB complexes activated by reovirus are comprised of NF-kappaB subunits p50 and p65 (RelA) . Apoptosis induced by reovirus is significantly reduced in cells treated with a proteasome inhibitor and in cells expressing a transdominant inhibitor of NF-kappaB . In addition, reovirus-induced apoptosis is blocked in cells deficient in the expression of the p50 or p65 NF-kappaB subunits. These results demonstrate that NF-kappaB plays a proapoptotic role during reovirus infection . Reovirus also activates c-Jun N-terminal kinase and extracellular signal-related kinase , but the involvement of these signaling molecules in NF-kappaB activation and apoptosis induction is not understood. In addition to NF-kappaB, several other cellular molecules have been implicated in reovirus-induced apoptosis. The calcium-dependent protease calpain is activated during reovirus infection, and calpain inhibitors block apoptosis induced by reovirus . It is not known how the calpain and NF-kappaB pathways couple to cause apoptosis during reovirus infection. Cellular gene expression is required for apoptosis induced by reovirus . In some cell types, reovirus infection leads to expression of death receptors DR4 and DR5 and their proapoptotic ligand TRAIL , suggesting that apoptosis is induced by autocrine or paracrine mechanisms. However, mitochondrial injury during reovirus infection has been documented , providing evidence that intrinsic pathways are also involved in apoptosis induced by reovirus infection. CONCLUSIONS : A precise understanding of the serotype-dependent differences in tropism exhibited by reovirus in the murine CNS remains elusive. T1 and T3 reoviruses vary in the types of cell surface carbohydrate used as coreceptors , but both serotypes bind JAM1 . These observations make it unlikely that JAM1 is the sole determinant of reovirus tropism in the murine CNS. It is possible that JAM1 serves as a serotype-independent reovirus receptor at some sites within the host and that other as yet unidentified receptors confer serotype-dependent tropism in the CNS. Definitive assessment of the role of JAM1 in reovirus pathogenesis awaits the results of studies using JAM1-null mice. The nature of the carbohydrate bound by reovirus plays an important role in viral attachment and apoptosis induction , and new evidence indicates that reovirus strains that vary in sialic acid utilization also vary in the capacity to produce biliary tract disease . However, both T3SA+ and T3SA- infect neurons, and these viruses display equivalent 50% lethal dose values following peroral inoculation of mice . These data suggest that sialic acid binding is not the primary determinant of neural tropism exhibited by T3 reovirus. It is possible that the unidentified carbohydrate bound by T1 reovirus directs infection to ependymal cells in the CNS. If so, engagement of JAM1 on neurons by non-sialic-acid-binding T3 reovirus would be expected, with a lethal outcome as the result. Studies using viruses containing chimeric sigma1 proteins in which the T1 and T3 carbohydrate-binding domains are reciprocally exchanged will help to clarify the role of carbohydrate coreceptors in reovirus disease. Differences in receptor utilization might also influence pathogenesis by virtue of activating different types of signaling pathways. Since T1 and T3 differ in the capacity to induce apoptosis , a property linked to receptor binding , it is possible that postattachment signaling plays a role in the production of disease. Support for this idea comes from studies of reovirus-induced myocarditis in which treatment of mice with inhibitors of calpain to inhibit proapoptotic signaling ameliorates tissue injury . Additional studies are required to confirm an association of apoptosis with reovirus-induced disease. However, it appears that the role of reovirus receptors in disease pathogenesis is more complex than simply mediating the virus docking event. A final remarkable observation is the localization of JAM1 to tight junctions. In addition to reovirus, several other viruses bind receptors expressed at regions of cell-cell contact . Like JAM1, the coxsackievirus and adenovirus receptor CAR is expressed at tight junctions . Nectins, which serve as receptors for herpes simplex virus , are expressed at adherens junctions . Interestingly, each of these viruses is capable of infecting both epithelial surfaces and neurons in some types of host organisms. Junctional regions are sites of enhanced membrane recycling, endocytic uptake, and intracellular signaling . Therefore, it is possible that viruses have selected junction-associated proteins as receptors to usurp the physiologic functions of these molecules. Such viruses would be expected to display common themes in attachment and internalization, which may extend to conserved modes of pathogenesis and disease production. FIG. 1. : Crystal structure of reovirus attachment protein sigma1. Crystal structure of reovirus attachment protein sigma1. The crystal structure of T3D sigma1 includes residues 245 to 455 . The three monomers of the sigma1 trimer are shown in red, orange, and blue. Each monomer consists of a C-terminal head domain formed by a compact beta-barrel and an N-terminal fibrous tail that contains three beta-spiral repeats. Based on analysis of patterns in aligned sigma1 sequences, the beta-spiral likely begins at residue 167 of T3D sigma1 and comprises eight repeats. The N-terminal five repeats, which are not included in the crystal structure, are shown in gray. The spiral has been extended using translated and rotated sigma1 repeats to generate a model that depicts the approximate dimensions of the molecule. Amino acids Asn198, Arg202, and Pro204 have been implicated in the interaction of T3D sigma1 with sialic acid . The approximate location of these residues in the model (shown in ball-and-stick representation on the right) suggests that they form a binding site for sialic acid. Residues 1 to 167 are not shown; these residues are predicted to form a triple alpha-helical coiled coil structure . This figure was prepared by Thilo Stehle (Harvard University) (published with permission) with the program RIBBONS . FIG. 2. : (A to D) Liver histopathology in mice following infection with T3SA- and T3SA+. (A to D) Liver histopathology in mice following infection with T3SA- and T3SA+. ND4 Swiss Webster mice (2 to 3 days old) were inoculated perorally with phosphate-buffered saline (A) or 2.5 x 103 PFU of either T3SA- (B) or T3SA+ (C and D). At 8 days postinoculation, liver tissue was harvested, embedded in paraffin, thin sectioned, and stained with hematoxylin and eosin. Magnification, x100 (C) or x400 (A, B, and D). (E to H) Immunohistochemical localization of reovirus antigen in bile duct epithelial cells. ND4 Swiss Webster mice (2 to 3 days old) were inoculated perorally with 2.5 x 103 PFU of either T3SA- (E and F) or T3SA+ (G and H). At 6 days postinoculation, liver tissue was harvested, embedded in paraffin, thin sectioned, and stained for reovirus antigen using rabbit anti-reovirus serum and horseradish peroxidase. Dark-brown staining indicates the presence of reovirus antigen. Magnification, x400. Representative sections from two separate animals are shown. Modified from Barton et al. with permission from the publisher. FIG. 3. : Model of reovirus-induced apoptosis. Model of reovirus-induced apoptosis. Reovirus infection is initiated by attachment of the virus to carbohydrate coreceptors and JAM1. For T3 reoviruses, the carbohydrate bound is sialic acid (SA). After attachment to cellular receptors, reovirus enters cells by receptor-mediated endocytosis. Within an endocytic compartment, the viral outer capsid is removed to generate infectious subvirion particles (ISVPs). During virion-to-ISVP conversion, sigma3 is degraded and lost from virions, sigma1 undergoes a conformational change, and mu1 is cleaved to form particle-associated fragments delta and phi. Removal of sigma3 exposes hydrophobic domains in mu1 that facilitate interactions of ISVPs with endosomal membranes, leading to delivery of core particles into the cytoplasm and concomitant activation of the viral transcriptase. Viral attachment and disassembly must occur within the same cellular compartment to activate NF-kappaB. Activation of NF-kappaB also might be achieved by mu1-mediated membrane penetration acting in synergy with viral receptor engagement. Once activated, NF-kappaB translocates to the nucleus, where it induces the expression of proapoptotic genes. TABLE 1 : Properties of reovirus pathogenesis attributable to sigma1 protein Backmatter: PMID- 12915555 TI - Kaposi's Sarcoma-Associated Herpesvirus/Human Herpesvirus 8 Transcriptional Activator Rta Is an Oligomeric DNA-Binding Protein That Interacts with Tandem Arrays of Phased A/T-Trinucleotide Motifs AB - Kaposi's sarcoma associated herpesvirus (KSHV)/human herpesvirus 8 (HHV-8) encodes an immediate early transcriptional activator, Rta, which mediates viral reactivation from latency and lytic viral replication. Here we report the purification and characterizations of HHV-8 Rta and its interaction with Rta-responsive DNA elements. The Rta response element (RtaRE) in the promoter of the KSHV/HHV-8 K8 open reading frame was mapped to a 47-bp sequence (RtaRE1) and a 60-bp sequence (RtaRE2) upstream of the TATA motif. A comparison of the K8 RtaREs with other viral RtaREs revealed a pattern of multiple A/T triplets spaced with a periodicity of 10 or 20 bp. Substitutions of the in-phase A/T trinucleotides of the RtaRE1 with G/C bases greatly diminished Rta responsiveness and Rta binding. By contrast, base substitutions in an out-of-phase A/T-trinucleotide sequence had no effect. Importantly, multimers of (A/T)3N7 and N5(A/T)5N6(A/T)4 motifs supported a strong Rta response in a copy number-dependent manner. No specific sequence motifs in the spacer regions could be discerned. Potent Rta response, however, was obtained with phased A/T trinucleotides with 7-bp spacers of arbitrary sequences with high G/C content. Lengthening of the phased A/T motifs or lowering of the G/C content of the spacers resulted in a reduction in Rta response. Finally, Escherichia coli-derived Rta is an oligomer of 440 kDa in molecular size and binds RtaRE as an oligomer. These results support a model of Rta transactivation wherein the subunits of the Rta oligomer make multiple contacts with a tandem array of phased A/T triplets in the configuration of (A/T)3(G/C)7 repeats. Keywords: Introduction : Kaposi's sarcoma-associated herpesvirus (KSHV), also known as human herpesvirus 8 (HHV-8), was first discovered in tumor tissues of Kaposi's sarcoma patients and subsequently linked to several other human cancers, including an AIDS-associated malignancy called body cavity-based lymphoma or primary effusion lymphoma and an atypical lymphoproliferative disorder known as multicentric Castleman's disease . The nucleotide sequence of KSHV/HHV-8 shows it to be highly related to oncogenic, lymphotropic, gammaherpesviruses, such as Epstein-Barr virus (EBV) and the rhadinovirus subfamily of gammaherpesviruses, including simian herpesvirus saimiri and avian Marek's disease virus. The regulatory circuit for KSHV/HHV-8 gene expression during the lytic cycle and reactivation from latency resembles that of EBV but with interesting differences. Based on protein sequence similarities and genome location with respect to their EBV counterparts, two putative KSHV/HHV-8 viral transcription activators, Rta and Zta (K-bZIP), encoded by open reading frame (ORF) 50 and ORF K8, respectively, have been identified . Like their EBV equivalents, the polypeptides for KSHV/HHV-8 Rta and Zta (K-bZIP) are derived from the splicing of a complex set of RNA transcripts . While EBV Rta and Zta act synergistically and are both important for EBV reactivation in latently infected B cells, KSHV/HHV-8 Rta alone appears to be responsible for viral reactivation . The KSHV/HHV-8 Rta is a polypeptide of 691 amino acid residues in size. It shares protein sequence homology with the immediate early protein 2 (IE2) of the bovine herpesvirus 4 ; with the putative transactivator, EDRF1(ORF 50), of the herpesvirus saimiri ; and with the EBV Rta (otherwise called BRLF1) . The highest degree of sequence homology among these proteins lies in regions corresponding to amino acid residues 1 to 240 and 611 to 652 of KSHV/HHV-8 Rta. The sequence and functional similarities suggest structural and functional conservation for this family of proteins. Rta is a sequence-specific DNA-binding protein that recognizes Rta-responsive DNA motifs located in the transcriptional regulatory regions of immediate early and early genes . This notwithstanding, the interaction between Rta and its responsive elements is incompletely understood. Unlike conventional sequence-specific transcription factors such as c-Jun, CREB/ATF-1, NF-kappaB, and the steroid hormone receptor, which bind well-defined cis-regulatory elements that are often no more than 8 to 10 bp in sequences, KSHV/HHV-8 Rta and the related bovine herpesvirus IE2 appear to interact with DNA elements that consist of longer and less-defined sequences . Furthermore, how Rta interacts with cellular transcription factors, transcriptional coactivators, and components of the basal transcriptional machinery to augment mRNA transcription also remains unresolved. Here, we describe the identification of a 47-bp sequence in the promoter of K8 ORF as a minimal Rta response element (RtaRE), which we termed RtaRE1. The 47-bp K8 RtaRE supports potent Rta response. Sequence comparison of the 47-bp K8 RtaRE with similar elements found in polyadenylated nuclear RNA (PAN), K12 ORF, and ORF 57 promoters revealed a consensus of triplet A/T motifs (underlined in Table ) spaced every 7 or 17 nucleotides apart. Based on these data, another RtaRE (RtaRE2) with a similar sequence pattern was also identified in the region downstream of the 47-bp motif in the K8 promoter. Mutational analyses indicate that these phased A/T trinucleotides and, to a lesser extent, the G/C contents of the intervening spacer sequences are critical for potent Rta-mediated transactivation. We showed further that multimers of (A/T)3N7 and N5(A/T)5N6(A/T)4 motifs support strong Rta response in a copy number-dependent manner. Importantly, optimal Rta response was obtained with phased A/T trinucleotides with arbitrary 7-nucleotide spacer sequences of high G/C content. Biochemical analyses of Rta showed it to be a multisubunit, possibly hexameric DNA-binding protein. These results suggest that Rta oligomers make contacts with multiple A/T triplets repeated in phase every 10 or 20 bp in the context of G/C-rich sequences to bring about transcriptional activation. MATERIALS AND METHODS : Plasmid construction. | The coding sequence of Rta was PCR amplified from a BCBL1 cDNA library using primers 5'-AGATCTCCATGGCGCAAGATGACAAG and 5'-GAATTCTCAGTCTCGGAAGTAATTACG. The PCR product was inserted into PCR2.1 TA cloning vector (Invitrogen, Inc., San Diego, Calif.). A BglII-EcoRI fragment containing the cDNA of Rta was then purified and cloned into pcDNA3.1 vector via BamHI and EcoRI sites to obtain CMV-Rta. A DNA fragment containing the 220-bp sequence upstream of the K8 coding region was generated by PCR by using genomic DNA of TPA-induced BCBL1 cells as a source of the DNA template together with the primer pair 5'-CTCGAGAGTGTTCGCAAGGGCGTCTG and 5'-AGATCTTTGGCAGGGTTACACGTTTA. The PCR product was first cloned into a TA cloning vector, PCR2.1, released by XhoI and BglII digestion, and inserted into a promoterless reporter plasmid, pA3Pluc, to produce RtaRE220-Luc. The RtaRE deletions were also derived by PCR by using the RtaRE220-Luc as a template. The nucleotide sequences for upstream primers are as follows: primer I, CTCGAGGTTTGGTGCAAAGTG; primer II, CTCGAGGCTCATTTTCTCACG; primer III, CTCGAGTTTGTGAAACAATAATGA; primer 1, CTCGAGGAGTTAACCTACAGAT; and primer 2, CTCGAGCTACTTAAAATAGCTC. The same downstream primer, namely, AAGCTTGGCAGGGTTACACGT, was used in all deletion constructs. Again, PCR products were inserted into pCR2.1 TA cloning vector, isolated as XhoI and HindIII fragments, and then cloned into SalI- and HindIII-digested PA3Luc vector. Plus- and minus-strand oligonucleotides corresponding to regions a, b, c, and d of RtaRE220 were chemically synthesized. The plus strand of each region contains a 5' BamHI protruding end, and the minus strand contains a 3' BglII protruding end. After the plus- and minus-strand oligonucleotides had been annealed, the double-stranded DNAs were phosphorylated at the 5' end with T4 polynucleotide kinase (New England Biolabs). They were then ligated in tandem, digested with BamHI and BglII restriction endonucleases exhaustively, and inserted into the BamHI site of the pUC19 vector. Plasmid clones that harbor oligonucleotide inserts were identified by PCR with the universal pUC primer pair. The copy number of each oligonucleotide insert was determined by polyacrylamide gel electrophoresis (PAGE) of the PCR products against appropriate DNA markers. SalI/XmaI fragments containing multimers of RtaRE regions a to d were inserted into an enhancerless reporter plasmid, pDeltaE-RSV-luc, via the same sites. To alter the non-A/T bases in the K8 ORF promoter region randomly, a 90-mer oligonucleotide, 5'-CTAGGGTACCATCCCGGGTTTATTTTTAANNNTTTNNNNAAANNGNNNTTAANNNNNNNATTNNNCNNAAAATAGTCGACAGATCTGTCA-3' (N represents a mixture of all four bases equally represented in the reaction), flanked by KpnI and SmaI restriction endonuclease sequences in the 5' end and SalI and BglII sequences in the 3' end, was generated by chemical synthesis. Double-stranded DNA was obtained by primer extension by using the 90-mer as a template and a primer of the sequence 5'-TGACAGATCTGTCGACTA-3'. The double-stranded DNA was then digested by XmaI/SalI and inserted into similarly digested pDeltaE-RSV-luc plasmid. Thirty clones were picked up and analyzed for response to Rta-mediated transactivation. Results from five representative clones are shown. Mutagenesis of the A/T motifs in the 47-bp RtaRE was preformed similarly. The sequences of the wild-type and mutant (M1 to M5) oligonucleotides used are 5'-CTAGGGTACCATCCCGGGTTTATTTTTAACAGTTT(CCC for M1)GGTGCAAA(GGG for M2)GTGGAGTTAA(CGG for M3)CCTACAGATT(GCC for M4) CTACTTAAAA(GGG for M5) TAGTCGACAGATCTGTCA-3', with the mutated bases underlined and the respective substitutions (C for T and G for A in M1 to M5) in parentheses. Complementary oligonucleotide pairs for Con1, Con2, and RtaRE2 sequences were made to contain 5'-phosphorylated SmaI/SalI, BglII/BamHI, and SmaI/SalI protruding ends, respectively. Con1 sequences 5'-CCGGGAAAGCGACCTTAAGCGACCGATTGCGACCGAAAGCGACCG-3' (plus strand) and5'-TCGACGGTCGCTTTCGGTCGCAATCGGTCGTTAACGGTCGCTTTC-3' (minus strand) and RtaRE2 sequences 5'-CCGGGCATTTTCTCACGAATCTGGTTGATTGTGACTATTTGTGAAACAATAATGATTAAAGGGG-3'(plus strand) and 5'-TCGACCCCTTTAATCATTATTGTTTCACAAATAGTCACAATCAACCAGATTCGTGAGAAAATGC-3' (minus strand) were annealed by incubation in a water bath at 70C and allowed to cool to 40C over a period of 45 min. The double-stranded Con1 DNA was then inserted directly into pDeltaE-RSV-luc via SmaI/SalI sites. Reporter plasmids containing tandem repeats of the Con2 sequence, namely, 5'-GATCTAAAAGTCGACAAAAG-3' (plus strand) and 5'-GATCCTTTTGTCGACTTTTA-3'(minus strand), were generated similarly, as described earlier for the constructs of regions a to d. The sequences of all constructs were confirmed by DNA sequencing. Cell lines. | Human embryonic kidney 293 (HEK293) cells were grown in Dulbecco's modified Eagle medium supplemented with 100 IU of penicillin/ml, 0.1 mg of streptomycin (STR)/ml, and 10% heat-inactivated fetal bovine serum and cultured at 37C in a humidified incubator with 5% CO2. Transfection and luciferase assays. | Lipofectamine (GIBCO BRL) was used in DNA transfection of HEK293 cells according to the manufacturer's protocol. Briefly, cells were plated at a density of 1.0 x 105 cells/well in 12-well plates 1 day before transfection. For each well, 0.5 mug each of CMV-Rta or Rta-null pcDNA3.1 and 0.5 mug of the reporter plasmid were used. To control for variability in transfection, 0.1 mug of pRL-TK containing an HSV thymidine kinase (TK) promoter-driven Renilla luciferase reporter cassette (Promega) was included as an internal control. Cells were harvested 48 h after DNA transfection, and the firefly luciferase activity was assayed and normalized against Renilla luciferase activity. For luciferase assays, cell lysates were prepared by dissolving the DNA-transfected cells from each well in 200 mul of the reporter lysis buffer (Promega). Twenty microliters of the lysate was placed in each well of a 96-well plate. After injection of 100 mul of a luciferase substrate buffer, the luciferase activity was measured immediately in an MLX microtiter plate luminometer (DYNEX Technologies). Immunofluorescence analysis. | Forty-eight hours after DNA transfection, cells on coverslips were fixed with 4% paraformaldehyde for 10 min on ice, followed by incubation with 0.1% Triton X-100 solution for 10 min on ice. The coverslips were washed with phosphate-buffered saline (PBS) three times, immersed in 3% bovine serum albumin (BSA) in PBS for 1 h at room temperature, and incubated overnight at 4C with a rabbit polyclonal antibody against Rta (1:500) in the same buffer. After five washes with PBS, fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit immunoglobulin G (Sigma; 1:10,000) in PBS containing 3% BSA was applied for 1 h at room temperature. Cells were washed again with PBS five times, then mounted with Fluoromount-G (Southern Biotechnology Associates, Inc., Birmingham, Ala.) solution containing 0.5 mug of 4'-6-diamidine-2-phenylindole (DAPI)/ml for fluorescence microscopy. Photographs were taken with appropriate filters and images overlaid in a computer to create two-color images. Expression and purification of KSHV/HHV-8 Rta. | The complete coding sequence of Rta was inserted into an expression vector, pTrc2His2-TOPO (Invitrogen), whose product is tagged with a c-Myc epitope and a hexahistidine extension at the COOH terminus, to produce pTrc2His2-TOPO-Rta. Competent Escherichia coli BL21 cells harboring the pTrc2His2-TOPO-Rta plasmid were grown in 10 ml of Luria-Bertani (LB) broth containing 100 mug of ampicillin/ml overnight at 37C. The overnight culture was used to inoculate 1 liter of LB broth containing 100 mug of ampicillin/ml, and the culture was incubated for 3 to 4 h until it reached an A600 of 0.8 to 1.0. At this point, Rta expression was induced for 6 to 7 h with 1 mM isopropylthio-beta-d-galactoside (IPTG) at 37C with shaking. The Rta-expressing cells were then harvested by centrifugation and resuspended in 10 ml of lysis buffer (1 M NaCl, 10 mM imidazole, 25 muM phenylmethylsulfonyl fluoride [PMSF] in PBS buffer [pH 7.4]). After sonication with a microtip at 70% duty cycle for four bursts of 60 s each, the cell suspension was centrifuged in a Sorvall SS-34 rotor at 12,000 rpm for 30 min at 4C to pellet unbroken cells and debris. The supernatant was then incubated with 1 ml of Ni2+-nitrilotriacetic acid (NTA) resin (Qiagen) overnight at 4C. The protein-bound gel matrix was packed into a column (1.5 by 10 cm) and washed with 40 volumes of the same buffer containing 20 mM imidazole. Rta protein was then eluted with a 10-ml gradient of 40 to 160 mM imidazole. Proteins in each fraction were analyzed by sodium dodecyl sulfate (SDS)-12% PAGE followed by immunoblotting with an anti-c-Myc antibody (Invitrogen). Fractions containing Rta were dialyzed against buffer D (20 mM HEPES [pH 7.9], 150 mM KCl, 0.2 mM EDTA, 0.5 mM PMSF, 0.5 mM dithiothreitol [DTT], 20% glycerol) and stored frozen at -70C. Gel electrophoretic mobility shift assay (EMSA). | A 116-bp EcoRI/HindIII fragment containing the 47-bp RtaRE from pUC19-RtaRE (see Fig. and ) and a 69-bp XmaI/SalI fragment containing the mutant or wild-type RtaRE derived from the corresponding DeltaE-RSV-luc constructs (see Fig. ) were labeled with [alpha-32P]dATP and [alpha-32P]dCTP, respectively, by Klenow enzyme, resolved in a 6% polyacrylamide gel, and isolated. Protein-DNA binding reactions were carried out as described previously with minor modifications . Typically, each binding reaction was in 20 mul, in a binding buffer containing 10 mM HEPES [pH 7.9], 40 mM KCl, 5 mM MgCl2, 10% glycerol, 10 mM beta-mercaptoethanol, 0.5 mg of BSA/ml, and 0.1 mM EDTA. Each reaction mixture contained 2 mul of alpha-32P-labeled DNA (similar5 ng) and 5 mul (similar0.1 mug) of Rta protein in the presence of other DNA competitors in indicated amounts and was incubated at room temperature for 1 h. A rabbit antibody generated against a peptide containing amino acid residues 527 to 539 (KKRKALTVPEADT) of Rta (a generous gift of Gary Hayward) was used in the supershift experiment. Reaction mixtures were electrophoresed at 35 mA in a 5% nondenaturing polyacrylamide gel (30:1 acrylamide to bisacrylamide) in Tris-glycine-EDTA buffer at 4C for 1.5 h. The gel was then dried on a piece of Whatman filter paper and autoradiographed. Sucrose gradient centrifugation. | Rta protein (100 mul, similar2 mug), together with 30 mul of a protein standard containing ferritin (440 kDa), adolase (158 kDa), and BSA (67 kDa) at a concentration of 4 mg/ml each, was gently layered on top of a 10-ml sucrose gradient generated in a 12-ml ultracentrifuge tube by the freezing and thawing of a 15% sucrose solution containing 300 mM ammonium acetate, 1 mM DTT, and 1x protease inhibitor cocktail (Roche Molecular Biochemicals). Ultracentrifugation was carried out in a Beckman SW40Ti rotor at 32,000 rpm at 4C for 20 h. Five hundred-microliter fractions were collected from the bottom of the centrifuge tubes after puncturing with a needle. Protein fractions were lyophilized, resuspended in 50 mul of SDS-PAGE loading buffer, and resolved in a 4 to 20% SDS-polyacrylamide gradient gel. Coomassie blue staining and immunoblotting were used to identify protein markers and Rta, respectively. Chemical cross-linking. | Bis-sulfosuccinimidyl suberate (BS3), a water-soluble, homobifunctional chemical cross-linker, was used in protein cross-linking studies. Purified wild-type Rta (approximately 0.2 mug in 10 mul) was incubated at room temperature with 1 mul of freshly prepared BS3 at different concentrations (0.25, 0.5, 2.5, 5, 25, and 50 mM) for 30 min or at 0.5 mM for different times (5 to 30 min). The reaction was quenched by adding 1 mul of 1 M Tris buffer (pH 8.0) followed by incubation at room temperature for 15 min. The samples were then mixed with SDS-PAGE loading buffer containing 100 mM beta-mercaptoethanol, heated at 100C for 5 min, resolved in 4 to 20% SDS-PAGE gels, and analyzed by immunoblotting with the Myc epitope antibody. RESULTS : Characterizations of a KSHV/HHV-8 RtaRE. | KSHV/HHV-8 Rta (referred to as Rta herein) potently activates the transcription of KSHV/HHV-8 K8 ORF and several other KSHV/HHV-8 early genes . To study the interaction between Rta and its response elements, we generated a 220-bp DNA fragment containing the sequence that lies immediately upstream of the K8 coding region by PCR and fused it to a promoterless firefly luciferase reporter plasmid, pA3Pluc-FL, to produce RtaRE220-Luc. To determine if the 220-bp sequence can mediate transactivation by Rta, HEK293 cells were cotransfected with RtaRE220-Luc and an Rta expression plasmid, CMV-Rta, in which the Rta cDNA was placed under the control of the human cytomegalovirus immediate early enhancer/promoter. As expected, under these conditions, CMV-Rta induced a greater than 100-fold increase in gene expression compared to the Rta-null, pcDNA3.1 vector control (Fig. and B), indicating that the 220-bp region contains DNA elements that mediate a potent Rta response. To delineate further the cis element responsible for Rta transactivation, four 5' deletion mutants (deletion mutants II, III, 1, and 2) in the 220-bp sequence were constructed, and the ability of each mutant to mediate transactivation by Rta was determined after cotransfection with CMV-Rta. Deletion mutant II, lacking 57 bp in the 5' region of the 220-bp fragment, showed approximately 70 to 80% of the activity of the full-length fragment. A deletion of 104 bp (deletion mutant III) greatly reduced the absolute level of Rta-mediated transcription and the relative level of transactivation (factor of activation) that compares the promoter activity in the presence or absence of Rta (Fig. , compare II and III). Smaller deletions (constructs 1 and 2) in the sequence spanning nucleotides -163 and -117 also diminished the overall promoter activity but diminished the activation levels to a lesser extent. Approximately 6- to 10-fold transactivation could still be observed with deletion construct III, although on the whole, the promoter was largely inactive. These results suggest that multiple cooperating Rta-responsive regions reside in the 220-bp sequence. In an effort to locate the minimal RtaRE, we derived synthetic oligonucleotides that cover different parts (Fig. , a to d) of the -181 to -117 regions. These oligonucleotides were ligated as tandem multimers and inserted into a reporter plasmid, pDeltaE-RSV-luc, which contains an enhancer-less promoter, derived from the long terminal repeat of the Rous sarcoma virus, upstream of the firefly luciferase reporter gene. This promoter extends 142 nucleotides upstream of the Rous sarcoma virus mRNA start site, retains elements that are potential NF-Y and C/EBP binding sites, and has a relatively low basal level activity. As show in Fig. , only the 47-bp sequence in region d supported a potent Rta response. Further, the levels of Rta transactivation increased significantly with an increase in the copy number of the 47-bp sequence (Fig. , lanes 5 to 7). Tandem copies of sequences that contain the TTTATTTTTAA (region a) or the TTAAAATA (region c) motifs showed reduced, albeit detectable, response to Rta (Fig. , lanes 1 and 4), while multimers of the oligonucleotide containing the GTTAA sequence (region b) showed lower Rta response (Fig. , lanes 2 and 3). These results suggest that the 47-bp sequence corresponding to region d contains a minimal RtaRE, and subregions of this element support only partial or reduced Rta response. Finally, multimers of the 47-bp sequence promote potent Rta-mediated transactivation. Partial purification and characterizations of Rta derived from E. coli. | Several gammaherpesviruses encode Rta-like proteins. Earlier studies have indicated that Rta is a nuclear protein that functions as a sequence-specific DNA-binding transcription factor . The molecular details of how Rta interacts with its target DNA element remain incompletely understood. Consistent with a role of Rta in transcriptional activation, an immunofluorescence assay revealed that Rta localized to the nucleus in a diffuse staining pattern . To derive and purify Rta protein in sufficient quantity for biochemical studies, we expressed the full-length Rta with a c-Myc epitope tag and a hexahistidine extension in E. coli. The recombinant Rta protein was then partially purified by Ni2+-NTA Sepharose affinity chromatography and detected by Coomassie blue staining and immunoblotting (Fig. , lanes 1 and 2, respectively). The E. coli-derived Rta is approximately 74 kDa in size, in agreement with its predicted size based on amino acid sequence (Fig. , lane 1). The Rta expressed in BCBL-1 cells, however, has been reported to have an apparent size that is much greater . To compare the Rta derived from different expression systems, we expressed it in 293 cells after DNA transfection. As shown in Fig. , the Rta expressed in 293 cells after DNA transfection has an apparent molecular size of 120 kDa, significantly larger than its E. coli-derived counterpart (Fig. , lane 2), suggesting that it is posttranslationally modified as previously described . Rta/RtaRE interaction. | To determine if Rta is a DNA-binding protein, we incubated the partially purified Rta with double-stranded DNA cellulose and eluted the bound protein with buffer solutions containing increasing concentrations of NaCl . Consistent with the notion that Rta is a DNA-binding protein, the E. coli-derived Rta bound tightly to double-stranded DNA cellulose and eluted preferentially with buffer solutions that contained NaCl at 0.6 M or greater concentrations (Fig. , lanes 4 to 6). The interaction between Rta and RtaRE was examined by EMSA using the partially purified Rta protein and a p32-labeled DNA fragment (1xRtaRE) containing a single copy of the 47-bp minimal RtaRE. As shown in Fig. , Rta bound to RtaRE to produce a specific protein/DNA complex, C1 (Fig. , lane 1). An unlabeled RtaRE-containing DNA fragment competed specifically for Rta binding (Fig. , lanes 1 to 5). By contrast, nonspecific competitors such as poly(dG-dC) did not compete with RtaRE for Rta binding even when added in 200-fold molar excess (1,000 ng) of the probe (lanes 9 to 11). Interestingly, poly(dI-dC) competed with RtaRE when used in excess (lane 8). Likewise, poly(dA-dT) also competed for Rta binding (data not shown). Inclusion of a specific antibody generated against amino acid residues 527 to 539 (KKRKALTVPEADT) of Rta in the binding reaction resulted in a supershift (lane 13), while addition of a nonspecific control antibody had no effect (lane 14), indicating that Rta is present in the complex. Unexpectedly, a slower-migrating protein/DNA complex, C2, was often detected in Rta EMSA when excess unlabeled RtaRE was used as a competitor (lanes 4 and 5). Rta is an oligomeric DNA-binding protein. | The appearance of the C2 complex in EMSA prompted us to consider the possibility that native Rta may be oligomeric and comprise multiple RtaRE-binding sites such that extra copies of RtaRE DNA may be incorporated into holo-Rta to produce the C2 complex. To determine if Rta is oligomeric, both chemical cross-linking and sucrose gradient centrifugation were carried out. Purified Rta protein was incubated at room temperature with increasing concentrations of a chemical cross-linker, BS3, for 30 min or with 0.5 muM BS3 for increasing duration. The cross-linked Rta was then resolved in 4 to 20% polyacrylamide gradient gels (Invitrogen) and analyzed by immunoblotting with the c-Myc antibody. As shown, protein species with molecular sizes that correspond to Rta dimer, trimer, and tetramer appeared in increasing abundance that correlated with the duration of incubation and the concentration of BS3 used . Minor bands suggestive of larger Rta species were also seen, but they could not be adequately resolved in the gel, and their abundances were low, possibly because of the inefficiency of the cross-linker to covalently join all subunits and the difficulty in transferring high-molecular-size protein species to a nitrocellulose membrane. Consistent with the notion that Rta is an oligomer, sucrose gradient centrifugation further revealed the presence of a major Rta species with a Stokes radius slightly smaller than that of ferritin (440 kDa) and a minor species slightly smaller than adolase (158 kDa) . Judging from the predicted molecular size of 74 kDa for Rta, it appears that the major Rta species may be a hexamer and that the minor species may be a dimer. To demonstrate that the oligomeric form of Rta is the protein species responsible for RtaRE binding, the gradient fractions containing the 440- and 150-kDa species were separately pooled and concentrated by filtration through a Microcon centrifugal filter unit (molecular weight cutoff, 10,000; Microcon YM-10, Millipore Corp.). The sucrose-containing buffer in which these protein fractions were kept was also exchanged with an appropriate storage buffer (buffer D) following filtration and concentration. The protein preparations were then assayed for RtaRE-binding activity by EMSA. As shown in Fig. , the 440-kDa species, but not the 150-kDa species, bound the RtaRE probe to produce a protein/DNA complex that is indistinguishable from the unfractionated Rta preparation . Rta antibody supershift further indicates that this complex contains Rta. These results show that each native Rta molecule most likely exists as a 440-kDa hexamer and possibly contains multiple DNA binding sites as a result. A/T nucleotides are critical for mediating Rta response. | The oligomeric structure of Rta suggests that one possible way by which it interacts with DNA is via binding to multiple short nucleotide sequence motifs spaced over arbitrary intervening sequences. This would explain why a consensus Rta-binding motif is not immediately apparent. Assuming that Rta is a hexamer or a trimer of dimers, a complete RtaRE may consist of three or six Rta-binding motifs. In view of these results, we compared the DNA sequences of the 47-bp K8 RtaRE with RtaREs of PAN, K12, and ORF 57 promoters previously reported in the literature . As shown in Table , multiple stretches of A/T motifs that range from 3 to 11 residues long can be identified in the RtaREs of all four promoters. Furthermore, a sequence periodicity of (A/T)3N7(A/T)3N7(A/T)3N7 ---i.e., triplet A/T motifs (underlined in Table ) spaced 7 or 17 nucleotides apart ---can be discerned, with the RtaREs of K8 and PAN showing the greatest conformity to this pattern and the RtaRE of K8 containing the highest number of repeats. To test if the A/T motifs in the K8 ORF promoter mediate RtaRE, we generated a 90-mer oligonucleotide that contained the sequence of the 47-bp RtaRE and 12 additional upstream nucleotides flanked by KpnI and SmaI restriction endonuclease sequences in the 5' end and SalI and BglII sequences in the 3' end. During chemical synthesis of the 90-mer, the A/T bases of the K8 ORF promoter sequence were kept unchanged while most of the G/C bases except for two (Fig. , positions 28 and 49) were replaced with 25% each of A, T, G, and C bases. The 90-mer was then made double stranded by using a primer that annealed to the 3' sequence corresponding to the SalI and BglII sites. The duplex DNA was then inserted into pDeltaE-RSV-luc via SmaI and SalI sites. Multiple plasmid clones were picked, sequenced, and cotransfected with CMV-Rta to determine their abilities to support Rta-mediated transactivation. Three highly Rta-responsive (HR) clones and two clones of lower Rta responsiveness (LR) (Fig. and B) ---with levels of Rta transactivation varying from 20- to 100-fold ---were sequenced and compared . The DNA sequences of all five clones conformed to the original 90-mer design, with the A/T-rich sequences largely unchanged expect for one base at position 19, which was fortuitously omitted during the synthesis. This single-base deletion occurred within an in-phase A trinucleotide and an out-of-phase T trinucleotide (see below) and did not affect the activity of the element, possibly because it did not alter the phasing of the critical A/T motifs (see below). The DNA sequences outside of the A/T motifs in the HR clones deviated significantly from that of the original K8 RtaRE, indicating that they are not absolutely critical for Rta response. The most obvious difference between the HR and LR clones resides in the presence of exceedingly long runs of A/T in the LR clones (Fig. , underlined), suggesting that while A/T bases are important, they need to be kept at a specific length for optimal Rta response. The RtaRE consists of phased A/T triplets, quadruplets, or quintets spaced with a periodicity of 10 or 20 bp. | To demonstrate further the importance of the A/T motifs in the (A/T)3N7 repeats, we replaced both in-phase (M2 to M5) and out-of-phase A/T motifs (M1) found in the K8 RtaRE with G/C residues and at the same time kept the remainder of the sequence unchanged. Each mutant element was then inserted into the pDeltaE-RSV-luc reporter and tested for its response to Rta. In keeping with the notion that phased A/T trinucleotides are critical for a strong Rta transactivation, G/C substitutions of the in-phase A/T motifs greatly reduced Rta responsiveness (Fig. , M2 to M5), while the out-of-phase TTT trinucleotide could be altered without significant loss of Rta response (Fig. , M1). As expected, the reporter activity of each mutant RtaRE sequence in response to Rta also correlated with its strength of Rta binding . The importance of the phased A/T trinucleotides is shown further by using a consensus sequence, Con1 (AAAGCGACCTTAAGCGACCGATTGCGACCGAAAGCGACCG). Con1 contains four phased A/T nucleotides (underlined) whose sequences are identical to those in the 47-bp K8 RtaRE. However, unlike the 47-bp RtaRE, the A/T motifs in Con1 are placed adjacent to arbitrary heptanucleotide sequences of high G/C content. As shown in Fig. , when inserted into pDeltaE-RSV-luc, Con1 supported potent Rta-mediated transactivation, whose level significantly exceeded that of a solo 47-bp K8 RtaRE (compare lanes 1 and 7). With the idea that RtaRE consists of in-phase A/T motifs in mind, we constructed luciferase reporters that contain tandem copies of the N5T(A)4N6(A)4 motif sequence GGATCTAAAAGTCGACAAAA (termed Con2) and tested their abilities to mediate Rta response. As indicated in Fig. , weak but significant Rta responses (especially in transactivation) were seen with single and double Con2 constructs (1xCon2 and 2xCon2, lanes 9 and 10 and lanes 11 and 12), while a potent Rta response, with activity comparable to that of Con1 (lane 7) and exceeding that of the 47-bp K8 RtaRE (lane 1), was seen with triple Con2 (3xCon2, lanes 13 and 14). A close inspection of the DNA sequence in deletion construct III also revealed a stretch of sequence (-116 to -56) that conforms to the [(A/T)3(G/C)7]3-6 pattern. This element (termed RtaRE2), when alone in the context of the K8 promoter, had a relatively modest response to Rta (Fig. , deletion III). However, when inserted into the pDeltaE-RSV-luc vector, it supported strong Rta transactivation . In aggregate, these data indicate that the K8 promoter contains two RtaREs. These two elements act synergistically to promote Rta-mediated mRNA transcription. When alone, each of these two elements responds moderately to Rta. However, when inserted upstream of a relatively robust promoter, such as that derived from the Rous sarcoma virus long terminal repeat, potent Rta response can be achieved. Finally, these data support the notion that Rta preferentially interacts with multiple, phased A/T triplets and, to a lesser extent, A/T quadruplets and quintets with G/C-rich intervening spacer sequences to promote strong transactivation. DISCUSSION : Immediate early proteins that exhibit sequence and functional similarity to KSHV/HHV-8 Rta are common among herpesviruses; the R transcriptional activator encoded by the EBV BRLF1 gene and similar proteins encoded by bovine herpesvirus (BHV-4 IE2) and simian herpesvirus saimiri are examples . To date, the exact mechanism of action of this family of viral transcriptional activator remains incompletely understood. Rta appears to be able to activate transcription through two distinct mechanisms, namely, by direct DNA binding and via association with a DNA-binding transcription factor RBP-Jkappa protein, which interacts with RBP-Jkappa sites in several Rta-targeted promoters . In this study, we have analyzed the promoter of the K8 ORF and identified a region of 47 bp that responds to Rta-mediated transactivation. This 47-bp sequence, by itself or in the form of multimers, promotes potent transcriptional activation by Rta in a copy number-dependent manner. In agreement with published results , EMSAs using the 47-bp element as a probe indicate that Rta is a sequence-specific DNA-binding protein. The E. coli-derived Rta is oligomeric, possibly consisting of six subunits, and binds DNA as an oligomer. A DNA sequence comparison between the 47-bp RtaRE and similar elements found in PAN, K12 ORF, and ORF 57 promoters revealed the presence of multiple A/T-trinucleotide motifs spaced 7 or 17 nucleotides apart. Mutational analyses indicate that the phased A/T motifs are critical for Rta response, while the spacer regions and the out-of-phase A/T motifs can be base substituted without adverse effects on the activity of the element. Most importantly, tandem copies of (A/T)3N7 motif (A/T)3GCGACCG sequences and N5T(A)4N6(A)4 motif GGATCTAAAAGTCGACAAAA sequences promote strong Rta-mediated transactivation. Based on this sequence pattern, another RtaRE in the K8 promoter (termed RtaRE2) downstream of the 47-bp element was also identified. RtaRE2 also supported a strong Rta response when inserted upstream of the RSV promoter . Taken together, these results suggest that Rta oligomers (hexamers?) bind multiple, phased A/T triplets ---i.e., [(A/T)3(G/C)7]3-6 ---in RtaRE to bring about potent transcriptional activation. This KSHV/HHV-8 Rta recognition sequence differs significantly from the previously reported consensus recognition site (GNCCN9GGNG) of the R transactivator of EBV . We think that RtaRE1 most likely extends upstream of the 47-bp repeat and includes the long stretch of A/T sequence as indicated in Table , judging from the reduction in promoter activity when the latter region was excluded (Fig. , construct II). RtaREs act synergistically when multimerized, as indicated by the combined action of RtaRE1 and RtaRE2 in the K8 promoter and that of multimers of the 47-bp motif . Finally, RtaREs support a strong Rta response when incorporated in the context of a strong promoter, such as that of the Rous sarcoma virus. The phasing of the A/T nucleotides in RtaRE provides an explanation for the moderate Rta response of A/T sequences linked together arbitrarily . It further explains a recent methylation interference analysis of Rta and the RtaRE of PAN promoter, which revealed little contact of G/C nucleotides with Rta . Consistent with our conclusion, extensive analyses of the PAN RtaRE have indicated the center TAA motif (see Table ) to be critical for both Rta binding and transactivation . The longer A/T motifs at the 5' and 3' borders of PAN promoter have also been shown to contribute to optimal Rta activation, although they are more tolerant of base substitutions . This is similar to what we describe here for the M5 mutation in K8 promoter . It should be pointed out that the mutations we introduced into the K8 RtaRE altered three A/T bases simultaneously and may have a greater and more unambiguous effect than the single and double base substitutions in the PAN promoter reported by Song et al. . The nucleotide sequences of PAN and K12 RtaREs share extensive homology (also see Table ). The sequence of the K8 RtaRE, however, deviates significantly from them, especially in the spacers in between the phased A/T motifs . Our results indicate that optimal Rta response is obtained with phased A/T trinucleotides with heptanucleotide spacers of high G/C content. Increase in the length of the A/T motifs (Fig. , Con1 versus Con2) or in the A/T content in the spacer regions diminishes (Fig. , HR versus LR), while increase in the G/C content of the spacers augments, Rta responsiveness . The need for three copies of Con2 (six A/T motifs) to support a potent Rta response may reflect a lower affinity of the A/T motifs in Con2 for Rta due to the increased length of A/T nucleotides; thus, additional binding motifs are needed for optimal interaction with Rta. Consistent with this notion, phased A/T hexanucleotides and heptanucleotides showed little Rta response (data not shown). Support for this conclusion can also be found in the G/C-rich nature of the spacers adjacent to the center TAA trinucleotide of the PAN RtaRE (see Table ). A/T substitutions in these spacer regions have been shown to attenuate both Rta binding and transactivation . As might be expected, multiple phased A/T triplets can be readily discerned in the upstream regions of three immediate early genes of KSHV/HHV-8, namely, the vIRF, vOX-2, and K7 genes . Whether these elements indeed mediate Rta transactivation remains to be determined. The exact subunit stoichiometry in each Rta oligomer is not clear at present. Chemical cross-linking and fluid dynamic analyses of the holo-Rta suggest that it may be a hexamer. The oligomeric nature of Rta suggests that it has the potential to contact multiple DNA motifs and is consistent with its requirement for multiple (three or more) phased A/T trinucleotides for potent transactivation. Finally, the critical importance of the A/T trinucleotides for Rta response and the sensitivity of Rta/RtaRE interaction to competition by poly(dI-dC) and poly(dA-dT) are reminiscent of the interaction between TATA box and TATA-binding protein (TBP). Because the minor groove structure of the dI-dC base pair is identical to that of the dA-dT base pair and TBP binds TATA sequence in the minor groove, TBP binding to the TATA motif is strongly competed by poly(dI-dC) . Whether each subunit of the Rta oligomer contacts the minor groove of each of the phased A/T motifs remains to be determined. FIG. 1. : Identification of RtaREs in K8 promoter. Identification of RtaREs in K8 promoter. (A) DNA sequence of 220-bp upstream of the K8 ORF. A DNA fragment containing a 220-bp sequence immediately upstream of the K8 coding region and four nested 5' deletions (II, III, 1, and 2) were generated by PCR with appropriate primers (see Materials and Methods) and then fused to a promoterless firefly luciferase reporter plasmid, pA3Pluc-FL, to produce RtaRE220-Luc and four other luciferase reporter plasmids. Arrows mark the 5' end of each construct. (B) Localization of RtaREs. HEK293 cells were cotransfected with luciferase reporter plasmid and an Rta expression plasmid, CMV-Rta, as described in Materials and Methods. At least three independent sets of luciferase assays, each with triplicate DNA transfection, were performed (a total of nine transfections) for each plasmid construct. All firefly luciferase reporter assays were normalized against HSV TK promoter-driven Renilla luciferase reporter activities. Standard deviations of the reporter activities and the extent of transactivation by Rta are also shown. FIG. 2. : Mapping of minimal RtaRE. Mapping of minimal RtaRE. (A) The sequences of the oligonucleotides (a to d) covering different parts of the -181 to -116 region are as indicated, with the A/T-rich regions underlined. (B) These oligonucleotides were synthesized with 5' BamHI and 3' BglII protruding ends. They were ligated as tandem multimers and inserted upstream of the TATA box of an enhancerless promoter in a reporter plasmid, pDeltaE-RSV-luc (see Materials and Methods). DNA transfections were carried out with or without CMV-Rta, and luciferase assays were as described in the legend for Fig. . The copy numbers, ranging from 1 (1x) to 6 (6x), of the individual sequences tested (a to d) are indicated below each column. FIG. 3. : Characterizations of Rta. Characterizations of Rta. (A) Rta is a nuclear protein. Immunofluorescence was carried out as detailed in Materials and Methods by using rabbit anti-Rta antibody (Rta). Cells were also stained with DAPI (Dapi). (B) Coomassie blue staining (lane 1) and immunoblot (lane 2) of partially purified E. coli-derived Rta. (C) Immunoblot analysis of E. coli-derived Rta (lane 1) and Rta expressed in 293 cells (lane 2). FIG. 4. : Rta binds RtaRE specifically. Rta binds RtaRE specifically. (A) Rta is a DNA-binding protein. Fifty microliters of partially purified Myc- and hexahistidine-tagged Rta was incubated with 0.5 ml of double-stranded DNA cellulose and washed three times with a buffer containing 100 mM NaCl, 20% (vol/vol) glycerol, 5 mM MgCl2, 0.1 mM EDTA, and 20 mM HEPES. The DNA cellulose-bound proteins (lane 7) were eluted with 0.5 ml of the same buffer solution containing increasing concentrations of NaCl, as indicated (lanes 1 to 6). The eluted fractions were immunoblotted with an antibody against c-Myc. (B) Rta forms a protein/DNA complex with RtaRE. EMSAs were carried out as described in Materials and Methods by using partially purified Rta and a 32P-labeled DNA fragment (1xRtaRE) containing a single copy of the 47-bp minimal RtaRE. DNA competitors that included unlabeled RtaRE DNA fragment (lanes 2 to 5), poly(dI-dC) (lanes 6 to 8), and poly(dG-dC) (lanes 9 to 11) were added in the amounts indicated to binding reactions. C1 denotes the Rta/RtaRE complex. C2 denotes a slower-migrating protein/DNA complex in reactions where excess unlabeled RtaRE was used as a competitor (lanes 4 and 5). Lanes 13 and 14 contained an antibody specific for Rta (Rta Ab) and a control antibody against the E. coli maltose binding protein (NS Ab), respectively. FIG. 5. : Rta is an oligomeric DNA-binding protein. Rta is an oligomeric DNA-binding protein. (A and B) Chemical cross-linking of Rta. Purified E. coli-derived Rta was chemically cross-linked with BS3 (see Materials and Methods for details) at room temperature with increasing time (A) or amount of BS3 (B), resolved in 4 to 20% polyacrylamide gradient gels, and analyzed by immunoblotting with the c-Myc antibody. Cross-linked Rta species with molecular sizes corresponding to different oligomers are as indicated. (C) Sucrose density gradient sedimentation analysis of Rta. Partially purified Rta protein, together with a protein standard containing ferritin (440 kDa), adolase (158 kDa), and BSA (67 kDa), was applied over a sucrose gradient and centrifuged in a Beckman SW40Ti rotor at 32,000 rpm at 4C for 20 h. The peak positions (lanes 8, 15, and 18) of the protein markers in the gradient are indicated. (D) The oligomeric form of Rta is responsible for RtaRE binding. The gradient fractions corresponding to the 440- and 150-kDa species were pooled separately, concentrated by Microcon YM-30 filtration, and assayed for RtaRE binding (lanes 1 to 6 and 8 to 14, respectively) as described in the legend for Fig. . Increasing amounts of 440- and 150-kDa Rta were used in lanes 2 to 5 and 9 to 12, respectively. Lanes 1 and 8 contain free probe in the binding buffer. Lane 7 contains the free probe only. Rta antibody was included in reactions in lanes 5 and 13. An antibody against the E. coli maltose binding protein (MBP-Ab) was included in lanes 6 and 14 as a nonspecific antibody control. The Rta/RtaRE complex is as indicated. A fast-migrating nonspecific band (NS) is as noted. FIG. 6. : A/T nucleotides are critical for Rta response. A/T nucleotides are critical for Rta response. (A and B) DNA sequences of chemically synthesized RtaREs. A 90-mer oligonucleotide that contains the sequence of the 47-bp RtaRE and 12 additional upstream nucleotides flanked by KpnI and SmaI sites in the 5' end and SalI and BglII sites in the 3' end was generated by chemical synthesis with the A/T bases kept unchanged and most of the G/C bases except for two (positions 28 and 49) replaced with 25% each of A, T, G, and C bases. After primer extension, double-stranded nucleotides were inserted into pDeltaE-RSV-luc. Multiple clones were sequenced and screened for Rta responsiveness after DNA transfection and reporter assays. The sequences of five representative clones, of which three (HR1 to HR3) (A) were highly responsive to Rta and two (LR1 and LR2) (B) showed lower response to Rta, are compared with the original sequence in the K8 ORF promoter (WT). The consensus sequence contains the unaltered bases of the original oligonucleotide. Long runs of A/T bases in LR clones are underlined. (C) Rta responses of HR and LR sequences. HEK293 cells were cotransfected with an Rta expression plasmid, CMV-Rta, and individual luciferase reporter plasmids containing the HR and LR sequences, as described in the legend for Fig. . Reporter activities of each construct with or without Rta were measured, and activation was computed. Standard deviations of activation are also shown. FIG. 7. : Phased A/T triplets flanked by G/C-rich spacers are critical for potent Rta transactivation. Phased A/T triplets flanked by G/C-rich spacers are critical for potent Rta transactivation. (A) Multiple phased A/T triplets are critical for Rta response. Six 90-mer oligonucleotides (wild-type and M1 to M5) that contain the sequence of the 47-bp K8 RtaRE (bracketed) and 12 additional upstream nucleotides were generated by chemical synthesis, as described in the legend for Fig. . Each of the oligonucleotides M1 to M5 contains three specific base substitutions, with the A and T bases replaced with G and C residues, respectively (indicated below the sequence). After primer extension, the double-stranded DNA was then inserted into pDeltaE-RSV-luc and confirmed by DNA sequence analyses. The constructs were then cotransfected with CMV-Rta into 293 cells, as described in the legend for Fig. . Activation of each construct was calculated and is shown. A pDeltaE-RSV-luc-derived reporter containing two copies of the 47-bp RtaRE (2x47-bp) (same as 2xd in Fig. ) and pDeltaE-RSV-luc (vector) were included as controls. Phased A/T motifs critical for Rta response are in boxes, and the nonessential T trinucleotide is marked with dots beneath it. (B) Rta-binding activities of mutant RtaREs. EMSAs were as described in the legend for Fig. , except that the various wild-type and mutant RtaREs were 69 bp in length. Wild-type and mutant RtaREs were labeled to similar specific activities, as in described in Materials and Methods. Approximately the same amount of DNA probe was used in each reaction. Possibly because of the decreased size of the DNA probe, the Rta/DNA complexes migrated faster than the complex in Fig. . (C) Phased A/T triplets with G/C-rich heptanucleotide spacers support potent Rta transactivation. Complementary oligonucleotide pairs for Con1 and Con2 sequences were made to contain 5'-phosphorylated SmaI/SalI and BglII/BamHI protruding ends, respectively. A/T motifs are boxed. Con1 sequence was inserted directly into pDeltaE-RSV-luc via SmaI/SalI sites. Luciferase reporter constructs containing the indicated numbers of Con2 sequence were produced as described in Materials and Methods and characterized. The Rta response of each construct as determined by firefly luciferase activity was measured as described above and marked as appropriate (lanes 7 to 14). All firefly luciferase reporter assays were normalized against TK-Renilla luciferase reporter activities, as described in the legend for Fig. . The 1x47-bp, 2x47-bp, and pDeltaE-RSV-luc empty vector controls (lanes 1 and 2, 3 and 4, and 5 and 6, respectively) are as described in the legend for panel A. (D) RtaRE2 supports strong Rta transactivation in DeltaE-RSV-luc. A complementary oligonucleotide pair of RtaRE2 with 5'-phosphorylated SmaI/SalI sites was inserted directly into pDeltaE-RSV-luc and analyzed for Rta response. A reporter containing the 47-bp RtaRE1 sequence was included for comparison. Luciferase assays were as described in the legend for panel C. FIG. 8. : DNA sequences upstream of the translational start sites of KSHV/HHV-8 vIRF, vOX-2, and K7 immediate early genes contain multiple phased A/T triplets. DNA sequences upstream of the translational start sites of KSHV/HHV-8 vIRF, vOX-2, and K7 immediate early genes contain multiple phased A/T triplets. Continuous strings of A/T bases are denoted in bold. A/T triplets that are in the same phase are shaded, underlined with solid or dotted lines, or indicated in white over a black background. The potential TATA elements are double underlined. TABLE 1 : Sequence comparison of RtaREs Backmatter: PMID- 12915571 TI - Epidemiology, Genetic Diversity, and Evolution of Endemic Feline Immunodeficiency Virus in a Population of Wild Cougars AB - Within the large body of research on retroviruses, the distribution and evolution of endemic retroviruses in natural host populations have so far received little attention. In this study, the epidemiology, genetic diversity, and molecular evolution of feline immunodeficiency virus specific to cougars (FIVpco) was examined using blood samples collected over several years from a free-ranging cougar population in the western United States. The virus prevalence was 58% in this population (n = 52) and increased significantly with host age. Based on phylogenetic analysis of fragments of envelope (env) and polymerase (pol) genes, two genetically distinct lineages of FIVpco were found to cooccur in the population but not in the same individuals. Within each of the virus lineages, geographically nearby isolates formed monophyletic clusters of closely related viruses. Sequence diversity for env within a host rarely exceeded 1%, and the evolution of this gene was dominated by purifying selection. For both pol and env, our data indicate mean rates of molecular evolution of 1 to 3% per 10 years. These results support the premise that FIVpco is well adapted to its cougar host and provide a basis for comparing lentivirus evolution in endemic and epidemic infections in natural hosts. Keywords: Introduction : Feline immunodeficiency virus (FIV) is a lentivirus in the family Retroviridae that infects members of the family Felidae exclusively. First described in domestic cats , FIV-related lentiviruses have also been documented from a number of wild felid species, including lions (Panthera leo) and leopards (Panthera pardus) in Africa, Pallas' cats (Otocolobus manul) in central Asia, and cougars (Puma concolor) in North and South America . While all feline lentiviruses analyzed to date represent a monophyletic group with similar organizations and structures , sequence divergence among viruses isolated from different cat species is often high. For example, sequence similarity between the env genes of FIV in the domestic cat and that in the cougar (FIVpco) is only 39% . The clinical manifestation of FIV infection in domestic cats typically involves chronic immune dysfunction, opportunistic infections, and behavioral disorders. In contrast, the limited data available do not indicate that FIV in naturally infected wild cats is associated with overt disease . The apparently low virulence coupled with a high genetic divergence among lentiviruses both within and among wild cat species have been interpreted as the outcome of a long history of coevolution between the virus and natural host in the Felidae . The dynamics and natural history of retroviral infections in natural populations remain poorly understood. At the same time, the recognition of human immunodeficiency virus (HIV) infection as a zoonotic disease, which originated from simian immunodeficiency virus (SIV) in wild primates, has highlighted the importance of wildlife species as sources of viral epidemics . More generally, wildlife populations are likely to hold many important clues to the maintenance, emergence, and evolution of viral pathogens . Not surprisingly, an increasing number of studies in recent years have examined natural lentivirus infections in primate hosts at both the intrahost and the population levels. Studies addressing the evolution and epidemiology of FIV have been limited to domestic- and feral-cat populations , where the virus is reported to have been introduced only within the last few decades . No studies have examined the epidemiology and evolution of endemic lentiviruses within a specific, free-ranging population of wild feline hosts. Such a study provides the opportunity to investigate how the virus is transmitted and how it evolves in a natural system where it has putatively persisted for extended periods of time. To this end, phylogenetic analysis of sequence data obtained from naturally infected individuals can be used first to identify routes and rates of virus transmission within a host population. Secondly, obtaining serial samples from the same individuals and from a large number of infected animals from the same population makes it possible to assess the rate and course of molecular evolution in the virus over time. Determining the evolutionary characteristics of endemic viruses is significant, given that strong selective regimes and high rates of nucleotide substitution are factors frequently considered a hallmark of epidemic retroviruses, such as HIV. The number of analytic tools available to estimate rates of molecular evolution from sequence data has increased rapidly in recent years. Specifically, methods that incorporate time data in a phylogenetic framework and which are particularly suited for the analysis of quickly evolving organisms, such as RNA viruses, have been introduced . In these applications, the time of isolation for each virus sequence is taken into account in the reconstruction of phylogenetic relationships, so as to provide an estimate of the expected amount of molecular change occurring per time interval. Therefore, these methods offer a way of comparing the courses of virus evolution both within and among host individuals, and among different viral genes, from sequence data. Moreover, new methods for rate estimation, such as Bayesian inference approaches , have recently become available and represent a significant improvement over earlier approaches by not being reliant on a known genealogy for temporally spaced sequences. Here, we describe the prevalence and molecular evolution of endemic FIVpco in a population of free-ranging cougars. Prevalence in this population was high (58% overall), particularly among older individuals. Using data from two viral genes, pol and env, we document the occurrence of two divergent types of FIVpco that were specific to the study area. In addition, we found that virus sequences were characterized by low intrahost genetic diversity and purifying selection and were evolving at a rate that was considerably lower than estimates for HIV type 1 (HIV-1) over the short term. These results stand in contrast to those observed in symptomatic lentiviral infections and give insights into possible courses and mechanisms of host-virus adaptation in natural host systems. MATERIALS AND METHODS : Sample collection. | Blood samples were collected from 52 individuals in a population of free-ranging cougars in the Snowy Mountain Range (SR) in southeast Wyoming between 1997 and 2001. This mountain range is separated from other nearby ranges by marginal cougar habitat and thus was likely to represent a demographic unit. Sex and age were recorded for each animal captured. Individuals were classified into the following four age categories: kittens (0 to 12 months), yearlings (13 to 24 months), young adults (25 months to 4.5 years), and old adults (>4.5 years). Several animals were caught and sampled repeatedly, including nine infected cougars that were sampled at intervals ranging from 11 to 49 months. Because the captured animals had been fitted with radio collars and monitored in the field, extensive information on movement and space use was available for most individuals, and maternity was known for all kittens sampled. DNA isolation. | Genomic DNA was extracted from lymphocytes of captured animals using a Super Quik-Gene isolation kit (AGTC, Denver, Colo.) at Wyoming State Veterinary Laboratory, Laramie. DNA and serum were then sent to the University of Montana, where all remaining work was conducted. For one infected individual (SR627) that died during the course of the study, a second DNA sample was obtained from brain rather than blood. Serology. | Evidence of FIV infection was determined by using a flow cytometry assay that is based on serological recognition of cells infected with a cougar FIV isolate (M. Poss, unpublished data). Cougar serum was incubated with either a feline thymic lymphoma cell line (3201) or 3201 cells infected with a cougar lentivirus. The cells were then washed twice in phosphate-buffered saline containing 2% fetal calf serum and incubated with cat anti-immunoglobulin G-fluorescein isothiocyanate. The cells were washed and analyzed on a FACSCalibur (Becton Dickinson, San Jose, Calif.), and data were statistically evaluated using Cell Quest software. Serum was considered positive for FIV antibody if there was a significant difference in fluorescence intensity between infected and uninfected 3201 cells that were incubated with the serum. All test results were confirmed by immunoblotting (data not shown). PCR amplification of pol and env. | Fragments of the proviral pol and env genes were amplified from serial dilutions of DNAs of infected individuals. First, a 462- to 527-bp fragment of pol was amplified by nested PCR. The primers used for the first round were 2479F (5' TAG AAG CAT TAA CAG AAA TAG TAG AGA 3') and 3028R (5' TTG TAA TTT ATC TTC AGG AGT TT 3') or 1259F (5' GAA GGA AAG GTA AAA AGA GCA GAT 3') and 1261R (5' ATC TTC AGG AGT TTC AAA TCC CCA 3') . If the first round was amplified with 2479F-3028R, the 1259F-1261R pair was used for the second round. In a few cases, instead of 1259F, a third forward primer, 2581F (5' AAA TCA GGA AAA TGG AGA A 3'), was used with primer 1261R. If the first round had been amplified using 1259F-1261R, the primers for the second round were 2506F (5' GGT AAA AAG AGC AGA TCC TA 3') and 3012R (5' AGT TTC AAA TCC CCA CCA TAG 3'). The conditions for the first round of PCR were as follows: 3 min at 94C, followed by 30 cycles of 94C for 30 s, 44C for 40 s, and 71C for 50 s, and a final extension for 5 min at 72C. For the second round of amplification, the conditions were 3 min at 94C, followed by 30 cycles of 94C for 30 s, 48C for 30 s (52C for 2506F-3012R), and 71C for 50 s, and a final extension for 5 min at 71C. In addition, a 719-bp fragment of env was amplified by nested PCR. The first-round primers were 7175F (5' ATT GCA TAT TGG GAT TTT A 3') and 8267R (5' TAT CTT AGA CAC TCG TTG G 3'), while the second round employed primers 7406dF (5' CCR TGG GGW GGR AGT AGR T 3') and 8115R (5'GTG CCA GTG GTT GCT CCT ATC A 3'). The conditions for the first round were 3 min at 94C, followed by 10 cycles of 94C for 30 s, 40C for 30 s, and 72C for 90 s, followed by 25 cycles of 94C for 30 s, 52C for 30 s, and 72C for 70 s. The conditions for the second round were 3 min at 94C, followed by 30 cycles of 94C for 30 s, 50C for 40 s, and 71C for 50 s, and a final extension for 5 min at 72C. The PCR products from pol and env amplifications were cloned into a PCR4 vector (Invitrogen, Inc., San Diego, Calif.). One to 12 env clones were sequenced for each time point, with a maximum of four clones originating from the same PCR to reduce the chance of resampling. For pol, multiple clones were obtained only from the nine individuals for whom serial samples were available. All primer sites were removed from the sequences prior to analysis. A number of precautionary measures were taken to minimize the probability of contamination of samples: DNA extractions took place in Wyoming at a facility where no FIV work is conducted. At the University of Montana, PCRs for FIV gene amplification were set up in a room that was kept free of plasmids and PCR products. Samples from all animals were first assessed by PCR for pol, regardless of the animals' serological status. None of the samples from seronegative animals yielded a PCR product. Samples that were FIV positive were then evaluated by PCR for env. At most, samples from two infected individuals, determined by pol PCR to be infected by unrelated viruses, were amplified at the same time. Sequential samples from the same individuals were analyzed as the samples were received and hence were never subjected to PCR at the same time. Finally, PCR were randomly checked by heteroduplex mobility assay for diversity within the amplification reaction and against the clone generated for an animal to ensure (i) that clones represented the original sample, (ii) that we were not missing diversity within the population of amplified sequences, and (iii) that no contamination from our reference strains or other cloned products had occurred. To determine Taq polymerase error, a 719-bp cloned cougar lentivirus env sequence was amplified and cloned, and 10 clones were sequenced. A 681-bp section (after primer sequences were excluded) was compared to the original sequence, and the polymerase error rate per nested PCR procedure was defined as the number of nucleotide changes observed per site. Phylogenetic analysis. | Alignments were conducted in Lasergene99 (version 4.06) from DNASTAR, Inc. (Madison, Wis.), using the CLUSTAL W algorithm, and were adjusted manually. For each of the two gene segments, the most appropriate model of nucleotide substitution was found with the program MODELTEST , using the full set of available sequences. In both cases, the best model, based on Akaike's information criterion, was a General Time Reversible + gamma (GTR + Gamma) model . Unless otherwise indicated, the estimated parameters were used in all further analyses. To determine the affiliation of FIVpco from the SR population with sequences from other North American cougar populations, a maximum-likelihood (ML) tree was constructed from the pol data. For this analysis, only one randomly chosen sequence was used per individual. For animals from which sequential samples were available, a sequence from the most recent sampling date was selected. Additional FIVpco pol sequences from North American cougars were obtained from GenBank (accession numbers , , , , , , , , and ). As an outgroup, we used a virus sequence from a cougar on Vancouver Island (British Columbia, Canada; accession number ) that represents a distinct clade of FIVpco relative to which all western North American types (except those from California) form a monophyletic group . To examine the ancestral relationships of FIVpco from the SR population, an ML tree was constructed from the entire set of env sequences. As for pol, the sequence from a Vancouver Island cougar was used as an outgroup (accession number ). The large number of sequences and a high number of polytomies found in the tree prevented a full heuristic search, and the search was therefore terminated after 100,000 iterations or 168 h of computing time. While additional trees with higher likelihood probably exist, they are unlikely to differ significantly from the tree obtained. Support for individual clusters was assessed through a bootstrap analysis with 500 replicates, using a neighbor-joining (NJ) tree algorithm and the same likelihood settings used in obtaining the ML tree. Figures for both pol and env trees were generated with the help of the program TreeView . Analysis of env sequences. | The complete alignment of all env sequences obtained from SR cougars (n = 292) was analyzed in several ways for evidence of positive or negative selection. First, the proportions of possible nucleotide substitutions for all sites, as well as for synonymous and nonsynonymous sites (Pamilo-Bianchi-Li method) were calculated in the program MEGA2 based on pairwise comparisons. Second, sequencewide evidence of positive or negative selection was sought with a Z test using MEGA2 based on the number of synonymous substitutions per synonymous site (KS) and the number of nonsynonymous substitutions per nonsynonymous site (KA). This test evaluates whether the null hypothesis KS = KA can be rejected in favor of one of the alternative hypotheses (KS > KA or KS < KA). Third, a codon-based test for selection was also performed using ML to test for selection acting on only a fraction of the codon sites. Two models of codon selection were fitted to the sequence data . Model M1 assumes that nonsynonymous substitutions can be either fatally deleterious (dN /dS ratio; omega0 = 0) or neutral (omega1 = 1) to selection. Model M2 extends model M1 by including a third category of otherwise-selected codons, with the nature of the selection indicated by the value of omega2 as positive (omega2 > 1) or negative (omega2 < 1). The parameters of models M1 and M2 were estimated using PAML software from an NJ tree constructed under a GTR model, and selection models were compared using likelihood ratio testing. Different initial values of omega2 were used, as advised by the user documentation, because the software is known to suffer from convergence problems . All codon sites were also classified as experiencing purifying, neutral, or some other type of selection using an empirical Bayesian approach also provided in PAML. The analyses were performed twice: (i) after removing those codon sites at which there were gaps or stop codons or sites having ambiguous nucleotides in some sequences (16 of 226 codons) and (ii) after removing those sequences containing gaps, stop codons, or ambiguous nucleotides (2, 35, and 2 of 292 sequences, respectively). To determine whether nucleotide changes were evenly distributed across the region of env and pol amplified or whether they occurred at particular hot spots, the number of observed nucleotide states per site was calculated using the program MacClade version 3.0 and plotted as the average number of states per nucleotide found within a sliding window of 10 nucleotides. Rate estimates of molecular evolution. | Rate estimates for the evolution of env and pol were based on all sequences available for an individual or on a subset of sequences for the population as a whole. Two estimation techniques were employed. First, we used an ML approach implemented in the program TipDate . Because this technique requires a known tree topology, NJ trees were first constructed for the respective sets of sequences using PAUP* . To facilitate adequate rooting of the within-individual trees, two or three intermediately distant sequences from other individuals were included as outgroups during tree construction. These taxa (and the branches leading to them) were subsequently removed prior to rate estimation. For the population as a whole, the data set was limited to one randomly selected sequence per individual and time point, to keep the estimation computationally feasible, and the root was found by the midpoint method. Once NJ trees had been obtained, they were entered into the program TipDate version 1.2 together with the respective sequence data and the sampling date for each sequence (to the nearest decimal fraction of a year) to find the ML rate of substitution, as well as an estimated absolute age of the corresponding tree (i.e., time to the most recent common ancestor), including 95% CIs. For the second estimation approach, we employed a recently developed Monte Carlo Markov chain (MCMC) framework for the Bayesian estimation of evolutionary rates and population parameters . This technique also uses sets of dated sequences but offers several specific advantages compared to the ML estimation. First, it does not assume that tree topology is known and it is not limited to estimation of evolutionary rates. Instead, it takes account of the uncertainty in any phylogenetic estimate and the error inherent in any sampling scheme by drawing representative samples from the space of plausible population sizes, evolutionary rates, and phylogenies . The result is an estimate of the posterior probability distribution of evolutionary rates (marginalized over phylogenies and population size). The method applied uses a coalescent prior on divergence times . Under this prior, viral population size within and between hosts was assumed to remain constant over the time spanned by the genealogy. Previous applications have shown that rate estimation using this method is robust to different assumptions of population history . An additional value of the Bayesian framework is that it allows specific prior information to be incorporated into the analysis. In our case, since the approximate ages of the cougar hosts were known, we made the assumption that the most recent common viral ancestor within an individual was unlikely to be much older than the infected animal itself or the time since infection. Thus, we used an exponential distribution with expectation equal to the age of the animal (as estimated in the field) or the known time since infection (based on seroconversion) as a prior for the age of the root of the tree (i.e., the age of the most recent common ancestor of all viral sequences in an infected individual). This had the effect of influencing the depth of the within-host viral phylogeny, since the depth of the phylogeny is necessarily constrained to be bounded by the time to infection because it is assumed that either a single virus or a homogenous group of viruses establishes infection in the host. No prior on the age of the root was used for the estimation of evolutionary rates for the SR population as a whole. Nucleotide sequence accession numbers. | One env sequence per SR individual has been submitted to GenBank under accession numbers to and to , as well as one pol sequence per individual ( to and to [see Fig. ]). Nexus files of the complete set of pol and env sequences used in the analysis can be obtained from or from the authors. RESULTS : FIVpco seroprevalence. | We tested 52 different cougars, of which 30 (58%) showed serological evidence of FIVpco infection. The prevalence increased significantly with age class (Spearman's rank correlation; r = 1.0; P < 0.001) and reached 100% among old adults . Overall, similar proportions of males and females were infected, but the ratios were more variable among younger age classes . Of all kittens sampled, 25% were positive. All five young (kittens or yearlings) that had been born to uninfected females tested negative. In contrast, 50% of the young born to infected females (10 of 20) were positive at the time of sampling (when the individuals were between 4 and 16 months old). Two adult females seroconverted over the course of the study. In both cases, the cats had given birth to young near the time of seroconversion, suggesting that they may have become infected during mating. Phylogenetic analysis of pol sequences. | Virus sequences were obtained from all seropositive individuals for which peripheral blood mononuclear cells were available (28 of 30), while none of the seronegative samples yielded a PCR product. An ML tree constructed from pol sequences identified two distinct groups of FIVpco in the SR population that were >20% divergent (corrected distances), with high bootstrap support for the respective nodes . Previously published pol sequences from other cougar populations all affiliated with these two lineages . One lineage (lineage I), which represented 75% of infected cats in the SR population, included sequences from northern Wyoming and British Columbia. The second FIVpco lineage identified (lineage II), which was found in seven SR cougars (25%), contained isolates from Arizona, Colorado, northern Wyoming, and Arizona-Nevada. Phylogenetic analysis showed that genetic divergence of FIVpco based on pol sequences was high in Rocky Mountain cougars, as documented previously . Within lineage I, the corrected divergence among sequences from different areas was 8 to 10%, while differences within lineage II ranged from 2 to 15%. The high diversity found in Rocky Mountain FIVpco contrasted with low degrees of sequence differentiation among pol sequences from within the SR population . All samples originating from the SR population formed well-supported groups within their respective lineages, with only 0 to 3% sequence divergence within the groups. Together, these data indicate that SR cougars were infected with highly related FIV types that were clearly distinguishable from viruses found in other parts of the Rocky Mountains. Phylogenetic analysis of env sequences. | The relationship of virus sequences from animals within the SR population was evaluated based on env because env sequences were longer than pol sequences and because we made an initial assumption that the evolutionary rate in env would be higher than that for pol. The general topology of the ML tree of env sequences from the SR population was congruent with the clustering of sequences derived from SR animals in the pol tree . Specifically, pol and env sequences from an individual had the same phylogenetic affiliation relative to those from other individuals. At the same time, the genealogical structure of the env tree was more defined and there was stronger support for substructure in each of the two lineages, possibly due to the larger amount of phylogenetic information contained in the longer env fragment. As in the pol analysis, clusters of virus genotypes were often not unique to specific individuals but rather to entire groups . This resembles the situation described for a domestic-cat population following what appeared to be a singular introduction of FIV and suggests a close epidemiological link among individuals within groups. In most cases, an obvious explanation for such links was apparent. For example, shared or closely related virus genotypes were found in all six known mother-offspring groups, indicating recent vertical-transmission events. Close clustering was also observed for sequences derived from animals for which we lacked familial information but for which field observations demonstrated adjacent or overlapping home ranges . Observational data on home ranges following capture were unavailable or yielded no evidence for spatial overlap only for the three remaining individuals (SR626, SR631, and SR643) that also had virus genotypes closely related to those found in other SR cougars. Nucleotide divergence in env. | Based on all 292 available clones, uncorrected proportions of possible nucleotide substitutions (p distances) among env sequences derived from the same infected animal varied between 0 and 2% . This range applied to comparisons both within and among different sampling time points. Different individuals carrying virus from the same phylogenetic lineage exhibited distances ranging from 0.5 to 5.5% but with a bimodal distribution centered around 1 and 4%. The uncorrected average difference among viruses from the two lineages was 14%. It is noteworthy that there were virtually no insertions or deletions among the env sequences. The error rate of the Taq polymerase was experimentally determined to be 0.015% per site (one substitution observed in 6,810 nucleotides). Because this value is at least an order of magnitude smaller than the average within-host genetic difference among FIV fragments , it can be assumed that Taq error did not significantly affect our results. Proportion of synonymous and nonsynonymous changes in env. | Within individuals, the mean proportion of synonymous changes per synonymous site (KS) in env sequences was very similar to the proportion of nonsynonymous changes (KA) . Among individuals infected with virus from the same viral lineage, KS was twice times as high as KA, and this increased to a 10-fold difference for comparisons between the two lineages. Concordant with the observation that synonymous changes were more common, a test for possible deviations from neutrality indicated that molecular evolution in env was dominated by purifying selection (Z = 10.41; P < 0.001). For the codon-specific analysis, ML estimates consistently indicated that 60% of sites experienced strong purifying selection regardless of the model and parameters considered. Meaningful estimates of the proportion of sites under neutral or positive selection were precluded by convergence problems (data not shown). PAML routinely returned an estimate of omega2 that matched the seed value (range, 0.7 to 10), suggesting failure to converge. The classification of sites using empirical Bayes' estimates identified 37 to 47% of sites as experiencing purifying selection but, as with the ML estimation, it was inconclusive in respect to the proportion of neutral sites and sites under positive selection (data not shown). Distribution of molecular changes across env andpol sequences. | We plotted the average number of different character states observed per nucleotide in a sliding window of 10 nucleotides for both pol and env in order to assess whether the amplified fragments contained both variable and conserved regions . Both gene fragments were found to contain segments in which nucleotide diversity was elevated, but overall, the distributions of substitutions in both viral genes were uniform. Evolutionary rates of FIVpco within the SR population. | The two approaches used to estimate FIV evolution within the SR population yielded evolutionary-rate estimates close to 1% per decade for both env and pol . Estimates obtained with the MCMC approach were slightly higher (1.54% for both genes) and were associated with smaller variances than the ML estimates. The latter showed a higher rate for env (1.14%) than for pol (0.77%). However, the CIs of all estimates obtained with both ML and MCMC overlapped substantially, indicating no significant difference in evolutionary rates among the different estimation approaches or between the two genes examined. Concordant with the lower mean rate for pol using ML, the estimated time to the most recent common ancestor of all SR sequences (i.e., the ancestor of both viral lineages) was higher for pol (168.6 years) than for env (79.6 years). All estimates were clearly distinguishable from zero, demonstrating sufficient evolutionary signal in the two data sets to identify substitution rates. In comparison, the uncertainty associated with the MCMC estimates was much smaller than for ML estimates and suggested that the two viral lineages found in Rocky Mountain cougars had likely separated 36 to 153 years ago. Evolutionary rates of FIVpco within infected individuals. | Substitution rates were estimated from sequential samples from nine infected individuals . Mean rates using ML ranged from 0.39 to 2.35% per 10 years (overall mean, 1.27%) for pol and 0.06 to 3.88% per 10 years for env (overall mean, 1.42%). Thus, evolutionary rates derived for individual animals were similar to the previous estimates of similar1% per decade for the population as a whole. As before, there was no evidence that the evolutionary rates for the two genes were different. Some problems with uncertainty in the estimation were encountered in that most ML estimates were not significantly different from zero. This suggests that overall, the number of changes in the virus that had occurred within an infected individual was small relative to the time intervals between samples (8 to 49 months). Also, the fact that precision was generally higher for estimates for env, which was based on a fragment that was 220 bp longer than pol, points to the fact that longer sequences were more likely to contain evolutionary signal and thus enhanced the ability to estimate rates. While most estimated times to the most recent common viral ancestor fell within or close to the assumed maximum time since infection, estimated tree ages were much higher (>30 years) in a few cases. Such long time periods, which would span several cougar generations and thus transmission events, appear unrealistic given the small number of viruses presumed to be transmitted during natural retrovirus infections . We suspect that unrealistically high tree ages and very low rates were related to difficulties in accurately estimating the tree topology when the signal in the data was weak. However, in light of these results, it remains possible that substantial viral sequence diversity occasionally survives transmission bottlenecks. The described difficulties in generating reasonable and precise rate estimates based on sequence data from individuals could be largely overcome by employing a Bayesian MCMC approach. This method takes the uncertainty of the tree topology into account. In addition, we used prior knowledge about maximum time since infection for each individual to improve estimation. The mean rates obtained ranged from 1.03 to 6.21% per decade for pol (average, 2.91%) and from 1.08 to 3.89% for env (average, 2.34%) . These results therefore confirmed the earlier observation that the evolutionary rates of both genes were similar. There was no difference among rate estimates for viruses in lineages I and II. In all cases, the assumed maximum time since infection fell within the 95% CIs of the estimate for the time since the most recent common ancestor. Whereas evolutionary rates rarely exceeded 2% per decade in most individuals, data for three animals (SR620, SR623, and SR630) were suggestive of higher rates in pol and env in at least three of the four estimates obtained per cat. Notably, these three cougars must have become infected less than a year before the first serial virus sample, because they either were newborn kittens (SR620 and SR630) or had seroconverted since the last sample (SR623). Thus, higher substitution rates appeared to be associated with more recent infections. DISCUSSION : The overall 58% FIV seroprevalence determined for the SR population is higher than most previous reports for free-ranging cougars in North America . The most comprehensive serological survey of North American cougars (n = 225) reported 31% overall FIV prevalence . FIV prevalence levels of >30% are also unusual for feral domestic-cat populations . Our data suggest that high FIV prevalence in cougars in general may be related to effective transmission of the virus both from mothers to infants or yearlings and among adult individuals. More than half of the young born to infected dams carried the virus at the time of our sampling. Young cougars remain with their mothers for 1 to 2 years. Because our samples were often taken several months before independence, the actual rate of transmission from mother to offspring may be even higher. This high frequency of vertical transmission is in accordance with reports for FIV in domestic cats but markedly higher than that found for lentiviruses in natural primate host populations . Vertical transmission is unlikely to be responsible for all infections in cougars, however, given that the prevalence of infection continued to increase with age after reaching maturity. Territorial fights, mating, or other forms of social contact most likely facilitate horizontal transmission of FIVpco in cougars, based on observations of domestic cats . The cooccurrence of several distinct FIV lineages within the same population of hosts has been described previously for free-ranging cougars and African lions . While the emergence of regional FIV subtypes is certainly not surprising for an animal with a wide geographic distribution, the intermixing of these distinct types suggests extensive cougar movements, current and/or historic, from or to the SR population. At the same time, SR pol sequences formed monophyletic clusters within their respective lineages, indicating that the two virus strains found within the population had been able to rapidly and effectively spread on a local scale. The presence of two viral lineages in the SR population also raises the possibility that individuals could become coinfected with both types. Indeed, coinfection with divergent FIVpco genotypes was previously documented in a cougar . However, we had no evidence for such coinfection in this population, suggesting that it occurs at a low frequency. Furthermore, coinfection with divergent virus types would have been a prerequisite for the generation of possible recombinant types carrying pol and env from different viral lineages. Such genetic reassortment of viral genotypes has been documented, for example, for HIV both in vivo and in vitro . However, the same clusters of infected individuals were identified for both pol and env in our data set, thus providing no evidence for ancestral recombination among lineages. It should be noted, however, that coinfection and recombination involving highly related viruses remain possible. Because of the small sequence differences involved, such events would be very difficult to detect but at the same time would be expected to be inconsequential in respect to the general genealogical patterns. Phylogenetic analysis of env identified a number of well-supported clusters containing closely related sequences . Adult female cougars in particular shared highly related viruses with their offspring, supporting the premise of vertical transmission . Sequence divergence between viruses in the mother and offspring was usually very small but ranged up to >1%. This was likely due to the fact that samples from mother and kitten were typically obtained at different times, and the interval between sampling and infection was unknown for both. In addition to the close affiliation of FIVpco sequences obtained from females and their offspring, closely related virus sequences were also found in individuals with overlapping or neighboring home ranges . For example, spatial movement data collected using radio telemetry showed that cougars that had contiguous or overlapping home ranges were infected with viruses that clustered together in the phylogenetic tree (600, 603, 607, and 608 in lineage I and 604, 605, 606, and 611 in lineage II). Adult females frequently share home ranges with offspring from a previous mating, and hence, some of these virus clusters may also be derived from related animals. The relationship of all animals in the population is currently under investigation and will illuminate mechanisms of viral transmission among cougars. However, the bimodal distribution of genetic distances among env sequences from different individuals within the two FIVpco lineages suggests that genetic distances of up to 2% among individuals reflect recent infection events. In contrast, distances of 4%, observed among viruses from different individuals, likely represent genotypes separated by more virus generations and potentially more than one transmission event. In contrast to the common local spread of a particular genotype, the introduction of a new virus into the population appears to be relatively rare. The env sequence most distinguishable from all other SR genotypes was isolated from a yearling male, SR622, which was probably not born in the study area. Unlike female offspring, whose presence close to the mother's territory is often tolerated even after maturity, male yearlings are usually expelled from the natal area and tend to disperse further away from it . One plausible scenario is that SR622 had immigrated into the SR population from a neighboring or distant area, where it had become infected with a related, yet distinguishable, variant of FIVpco. Because retroviral diversity increases with time since infection, we expected to observe larger divergence in env in some individuals. Instead, intrahost env diversity in cougars averaged <1%, which is similar to values for pol in cougars and domestic cats but lower than the 1.1 to 2.5% reported for env in domestic cats . This low sequence diversity within hosts is remarkable given that cougars can reach ages of >10 years in the wild and that some of the individuals we examined were known to have harbored infections for several years. Also, we found that within individuals, sequences taken at time points that were 8 to 49 months apart were not more different from each other than sequences derived from the same time point. This may be due, in part, to the fact that we derived our sequence data from proviral DNA, and FIVpco may reside in long-lived cell populations. However, these data suggest that evolution of FIV env within cougar hosts may occur at a lower rate than in epidemically occurring retroviruses. It was also noteworthy that there was a virtual absence of insertions or deletions in env fragments from FIVpco, while such changes are commonly observed in domestic-cat FIV and primate lentiviruses. Concordant with a low genetic diversity among intrahost isolates and the close relation of sequences from putative family groups, the substitution rates we estimated for cougar FIVpco were low compared to estimates for HIV or SIV. At the same time, it should be noted that evolutionary rates have been obtained using a wide range of techniques and that some of these are not equivalent to our approach, making a comparison with published estimates difficult. The only study we are aware of that estimated evolutionary rates in FIV was done with domestic cats and found no changes in the consensus sequence of pol or gag in a naturally infected cat over a 3-year period . For a part of the env gene that included a variable region, a rate of 3.4% per 10 years was estimated in the same study, similar to estimates obtained here. Although results were variable, our estimates suggest an overall rate of 1 to 3% per 10 years for both pol and env. In comparison, studies of HIV evolution within hosts or small groups of hosts reported higher rates of 0.3 to 1.0% per year , while rates up to 3% per year have been calculated for SIV . Interestingly, HIV-1 evolves considerably more slowly at the population level, where, using sequence data collected over a decade or more, rates of 1 to 2 and 4 to 6% per decade have been estimated for pol (based on all substitutions) and env (V3; based on synonymous changes alone), respectively . Whereas FIVpco evolution within some infected cougars in our study appeared to be faster than for the population as a whole, our data do not support rates that would exceed half a percent per year. Comparisons of our estimates to those for primate lentiviruses should be done cautiously, however, because most of the studies focused on particularly variable regions of env and, in the case of SIV, were almost all conducted with experimentally rather than naturally infected individuals . More studies of naturally infected primate and feline hosts are needed to determine whether a lower rate of molecular evolution, relative to HIV, could indeed be a characteristic of natural host-virus systems or of feline lentiviruses. Given the estimated substitution rate of <0.5% per year, the time intervals between our samplings (8 to 49 months) were at the low end of what would yield measurable changes in the virus. Clearly, extending the time frame of virus sampling should improve the precision of rate estimates, as would longer sequence data. For the data collected here, the use of a Bayesian MCMC approach that integrated over all possible tree topologies markedly improved our estimates for evolutionary rates from virus sequences compared to the ML approach. Specifically, incorporating prior information on the probable time since infection made it possible to obtain estimates for virus evolution within individuals where ML rates had not been distinct from zero . In almost all cases, the probable time since infection fell within the obtained 95% CIs of both ML and MCMC estimates, indicating that this prior information was in agreement with the data. No prior for tree age was used for the analysis of virus sequences from all SR individuals combined. The obtained estimates overall suggest that the split between the two FIVpco lineages in Rocky Mountain cougars occurred within the last 200 years . This relatively recent time point is consistent with the fact that FIVpco from this region is distinct from virus found in cougars elsewhere in North America and that cougar populations in the western United States are currently recovering from low population levels due to persecution during the 19th century and the first half of the 20th century . Several factors may be involved in a lower rate of molecular evolution in FIVpco than in other lentiviruses, including the fidelity of replication and the length of virus generation time. Reported error rates for the replication enzymes of different retroviruses vary by a factor of 20 , raising the possibility that lower mutation rates are due to a higher fidelity of the viral reverse transcriptase. In addition, low evolutionary rates may be related to longer virus generation times. Although in HIV, the generation time is as short as 1 to 3 days, it is thought to be substantially longer for a small percentage of viruses that establish latent infections in long-lived cell types, such as macrophages . Therefore, it may be that cell types facilitating slower virus replication comprise the majority of FIVpco-infected cells in cougars, resulting in slow virus population turnover. Although the question of cell tissue tropism in FIVpco remains speculative, the observation that HIV variants targeting long-lived cells outlive fast-replicating variants suggests a potential selective regime that could favor slowly replicating viruses in a coevolutionary process. Evasion of the host immune system is frequently considered a partial explanation for the high rate of molecular evolution observed in many lentiviruses, especially for the envelope glycoprotein. At the same time, evidence for positive selection has been more or less limited to studies of HIV-1 in humans and is controversial even there . Higher proportions of nonsynonymous changes also characterize SIV infection of macaques, which are not a natural host of primate lentiviruses. In African green monkeys, a natural host for SIV, the proportion of synonymous substitutions was higher than that of nonsynonymous changes . In the domestic cat, a strong role of positive selection in FIV lacks support based on reported KS/KA ratios that rarely exceed 1 . Consistent with the latter results, we found evidence for selection largely against rather than in favor of nonsynonymous changes in cougar FIVpco env, supporting the idea that molecular evolution of this gene occurs under some functional constraint. Results for the identification of sites under positive selection were ambiguous but suggest that the number of such sites is likely to be low. In addition, the low rate of nonsynonymous substitutions relative to synonymous substitutions between hosts argues against any process of host-specific adaptation as the virus moves from one host to the next during virus transmission. This lends some weight to the argument that forms of FIV maintained in wild cats are well adapted to their host species. Still, compared to nonpathogenic lentiviruses in other natural hosts, which are often characterized by rapid accumulation of evolutionary changes resulting from high rates of replication , intrahost genetic diversity and substitution rates were low in the cougar FIVpco system. This shows that despite overall similarities, life history strategies and evolutionary characteristics may vary considerably among different lentiviruses in their natural hosts. FIG. 1. : Prevalence of FIVpco infection in a population of free-ranging cougars in southeast Wyoming by sex and age class. Prevalence of FIVpco infection in a population of free-ranging cougars in southeast Wyoming by sex and age class. Prevalence was determined by serology using a flow cytometric assay and confirmed by PCR. The sample sizes are shown above the bars (n = 52 [total]). For individuals that were sampled repeatedly, only infection status for the most recent sample was considered. FIG. 2. : ML tree of partial FIVpco pol sequences from free-ranging cougars in the Rocky Mountains (United States and Canada). ML tree of partial FIVpco pol sequences from free-ranging cougars in the Rocky Mountains (United States and Canada). Taxa designated SR represent animals from the study area. For each SR individual, one sequence was randomly selected from a larger data set. Taxa designated Pco represent previously described FIV pol sequences from other cougar populations in western North America. The state or province where the animal was captured is indicated in boldface letters (YNP, Yellowstone National Park). Branch lengths of >0.02 are shown below the branches. The FIVpco sequence from a Vancouver Island cougar (PLV1695) was used as an outgroup. The values next to tree nodes represent bootstrap values of >70 based on 500 replicates using an NJ tree algorithm and the same model of substitution used for the ML tree. FIG. 3. : ML tree of 292 partial FIV env sequences from 28 infected cougars in a population in southeast Wyoming. ML tree of 292 partial FIV env sequences from 28 infected cougars in a population in southeast Wyoming. Mother-kitten groups are indicated by open boxes, and animals with related viruses and confirmed overlap in home ranges are indicated by shaded boxes. Where identical sequences were identified, their number is indicated in parentheses next to the taxon symbol. Branch lengths of .0.01 are shown below the branches. The ML search was terminated after 100,000 iterations (see the text). See Fig. for further descriptions. FIG. 4. : Partitioning of genetic diversity in FIV env sequences from 28 naturally infected cougars based on p distances observed in pairwise comparisons. Partitioning of genetic diversity in FIV env sequences from 28 naturally infected cougars based on p distances observed in pairwise comparisons. For each category of comparison (e.g., within individuals, same time point), frequencies sum up to 1. FIG. 5. : Plots of nucleotide diversity for FIV env (A) and pol (B) gene fragments. Plots of nucleotide diversity for FIV env (A) and pol (B) gene fragments. Diversity is expressed as the average number of character states found per sliding window of 10 nucleotides. TABLE 1 : Parameters of molecular-evolution models TABLE 2 : Average proportions of synonymous and nonsynonymous sites from pairwise comparisons of FIVpco env sequences TABLE 3 : ML and MCMC estimates of evolutionary rates for FIVpco TABLE 4 : ML and MCMC estimates of evolutionary rates for two FIVpco genes within naturally infected cougars Backmatter: PMID- 11818032 TI - Kinetic comparison of tissue non-specific and placental human alkaline phosphatases expressed in baculovirus infected cells: application to screening for Down's syndrome AB - Abstract | Background | In humans, there are four alkaline phosphatases, and each form exibits a characteristic pattern of tissue distribution. The availability of an easy method to reveal their activity has resulted in large amount of data reporting correlations between variations in activity and illnesses. For example, alkaline phosphatase from neutrophils of mothers pregnent with a trisomy 21 fetus (Down's syndrome) displays significant differences both in its biochemical and immunological properties, and in its affinity for some specific inhibitors. Results | To analyse these differences, the biochemical characteristics of two isozymes (non specific and placental alkaline phosphatases) were expressed in baculovirus infected cells. Comparative analysis of the two proteins allowed us to estimate the kinetic constants of denaturation and sensitivity to two inhibitors (L-p-bromotetramisole and thiophosphate), allowing better discrimination between the two enzymes. These parameters were then used to estimate the ratio of the two isoenzymes in neutrophils of pregnant mothers with or without a trisomy 21 fetus. It appeared that the placental isozyme represented 13% of the total activity of neutrophils of non pregnant women. This proportion did not significantly increase with normal pregnancy. By contrast, in pregnancies with trisomy 21 fetus, the proportion reached 60 --80% of activity. Conclusion | Over-expression of the placental isozyme compared with the tissue-nonspecific form in neutrophils of mother with a trisomy 21 fetus may explain why the characteristics of the alkaline phosphatase in these cells is different from normal. Application of this knowledge could improve the potential of using alkaline phosphatase measurements to screen for Down's syndrome. Keywords: Background : Alkaline phosphatase (AP, orthophosphoric monoester phosphohydrolase, alkaline optimum, EC 3.1.3.1) is a group of ubiquitous enzymes found in nearly every organ. So far, four different human isoenzymes have been identified: the tissue non-specific isozyme (NSAP) is expressed in numerous tissues , while the three specific genes have more restrictive expression: in intestin (intestinal AP; ), placenta (PLAP; ) or thymus and testis (germ-cell AP; ). AP are zinc-containing dimeric proteins which catalyze the hydrolysis of phosphomonoester with release of inorganic phosphate and alcohol at alkaline pH. The catalytic mechanism was first deduced from the structure of the bacterial enzyme and was recently confirmed from the structure of a human isozyme . It involves the activation of a serine by a zinc atom, the formation of a phosphorylenzyme, the hydrolysis of the phosphoseryl by a water molecule activated by a second zinc atom and the release of the phosphate or its transfer to an acceptor. Four main catalytic functions have been attributed to these enzymes, hydrolase activity on low molecular weight phosphomonoesters , phosphotransferase activity , protein phosphatase activity and pyrophosphatase activity . The physiological role of AP is poorly known, except for the involvement of the NSAP isoenzyme in the mineralisation of bone tissue . In blood, only NSAP is found in the serum and in neutrophils . These cells contain a wide variety of enzymes functionally active in host defense. Among them AP, probably serving in membrane upregulation, has been identified in specific inclusions, the secretory granules, representing a highly mobilizable storage compartment. AP is detectable in differentiated granulocytes, including myelocytes, meta myelocytes, band forms and segmented neutrophils. With pregnancy, AP increases in the serum. This phenomenon originates from the PLAP synthesized in the placenta from the 7th week of pregnancy which passes into the mothers serum [,-]. Besides this activity in the serum, the AP activity also increases in the neutrophils of pregnant women, but it is the NSAP isozyme which is responsible . As yet, little is known of the mechanisms regulating AP activity during the course of normal gestation. Three distinct mechanisms presumably act in combination to elicit AP activity: i) the physiological hyperleucocytosis occurring with a steady increase in leucocyte count during pregnancy ; ii) the rise in placental hormonal secretions, estrogen and mainly progesterone, results in an increase in AP activity correlated with an elevation of steady state mRNA levels as a consequence of enhanced gene transcription ; iii) the induction by granulocyte colony stimulating factor (G-CSF), one of the most important modulators responsible for NAP activity . The characteristics of AP from blood neutrophils of women whose fetuses had trisomy 21 differ from those with normal pregnancies. An elevated AP activity has been reported in affected individuals . An ectopic expression of PLAP seems to appear since i) AP is more stable to heat and urea denaturation , ii) AP is more sensitive to inhibitors L-homo-arginine, EDTA, L-phenylalanine, L-p-bromotetramisole and sodium thiophosphate and iii) AP is less recognized by anti-NSAP antibodies and shows a reaction with anti-PLAP antibodies . These variations in the respective levels of expression of NSAP and PLAP in serum and neutrophils can lead to these enzymes being used as markers to detect trisomy 21 fetuses. However, this method has been reported to be controversial while others found it to be reliable . This discrepancy may originate from the difficulty of separating the two isozymes, NSAP and PLAP. Thus, we studied some of their properties after in vitro production of recombinant enzymes in baculovirus infected cells in order to provide data useful to differentiate the two isozymes. Results : Comparisons of enzyme stability | Enzymes were denatured by urea and by heating. In the presence of urea, we observed a decrease of active enzyme concentration until a plateau suggesting that the denaturation was reversible . Thus, the kinetic of denaturation was analyzed with scheme 1, Figure 2 | Inactivation of alkaline phosphatase by urea: effect of incubation of PLAP and NSAP with 7.4 M urea on activity. Inactivation of alkaline phosphatase by urea: effect of incubation of PLAP and NSAP with 7.4 M urea on activity. where E represents the native enzyme, Ed the inactive denatured enzyme, kd the denaturation rate constant and kr the renaturation rate constant. The variation of remaining activity (E/E0) with time t follows equation 1: Analysis of data by non-linear regression gave an estimation of the two rate constants kd and kr for the two enzymes . The renaturation rate constant (kr) was significantly different from zero confirming the reversibility of the urea denaturation. Placental AP appeared to be more resistant to urea denaturation than NSAP while the renaturation rate constants were not significantly different. The heat denaturation rate differed between the two enzymes at the all temperature assayed. Denaturation was irreversible and the simplest model which fitted to the data is illustrated by scheme 2, where Ed represents the irreversible form of the denatured enzyme. The variation of remaining activity (E/E0) with time t follows: Analysis of data led to the rate constant kd for the two enzymes . Placental AP appears to be 15 fold more resistant than NSAP. Figure 3 | Inactivation of alkaline phosphatase by temperature: effect of incubation of PLAP and NSAP at 56C on activity. Inactivation of alkaline phosphatase by temperature: effect of incubation of PLAP and NSAP at 56C on activity. Comparison of kinetic constants of NSAP and PLAP | Hydrolysis of p-nitrophenylphosphate by the two recombinant enzymes did not reveal any significant difference in their affinity for this substrate. Km were estimated to 171 +/- 12 muM and 180 +/- 15 muM respectively. For each enzyme, we observed a decrease of Km with pH (from 9.5 to 8) but without any meaningful difference between the two enzymes. Sodium thiophosphate is a full competitive inhibitor of AP , and the inhibition was thus studied according to scheme 3, where E represents the enzyme, S the substrate p-nitrophenylphosphate, and I the reversible inhibitor. For clarity, free substrate and inhibitor are omitted from the presentation as are the products. Variation of the hydrolysis rate of the substrate (v) with substrate and inhibitor concentrations follows equation 3: Placental AP was more than four times more sensitive to this inhibitor than NSAP . Figure 4 | Inhibition of PLAP and NSAP by thiophosphate. Inhibition of PLAP and NSAP by thiophosphate. L-p-bromotetramisole is an uncompetitive inhibitor specific for alkaline phosphatase, it binds to the phosphorylenzyme intermediate preventing dephosphorylation as phenylalanine . Binding of inhibitor is shown on scheme 4, where E represents the free enzyme; ES, the Michaelian complex; EX the phophorylenzyme and EXI, the inhibitor bound on the phosphoryl enzyme. The variation of the hydrolysis rate of the substrate (v) with substrate and inhibitor concentrations follows equation 4: L-p-bromotetramisole appeared to be a potent inhibitor of the two enzymes. Placental AP was more sensitive than NSAP . Figure 5 | Inhibition of PLAP and NSAP by L-p-bromotetramisole. Inhibition of PLAP and NSAP by L-p-bromotetramisole. Quantification of NSAP and PLAP in neutrophils | Phosphatase activity from neutrophils of pregnant women with a normal or with a trisomy 21 fetus were slightly different (1.8 +/- 0.4 and 1.2 +/- 0.2 n I.U. per mg protein, p = 0.0013). Kinetic constants were used to estimate the relative amounts of NSAP and PLAP. As the greater difference between the two enzymes was the resistance to temperature denaturation, the stability of neutrophil AP was recorded and analyzed considering that there was a mixture of the two enzymes. Thus, remaining activity follows the sum of equation 2 weighted by the proportion of the two enzymes : where kda represents the denaturation rate constant of the PLAP component, kdb, the denaturation rate constant of the NSAP component, and a the relative proportion of PLAP. As the rate constants were already determined, the fit allowed the proportion of the two isozymes to be unambiguously determined. In non pregnant women, PLAP represented 13% of the total AP activity of neutrophils. This proportion did not significantly increase with normal pregnancy. By contrast, in pregnancies with trisomy 21 fetus, we found a mean of PLAP of 67%. Analysis of nine individuals revealed that PLAP proportion in neutrophils varied from 60 to 80% for pregnancies with trisomy 21 . The same analysis was performed using the differential sensitivity to urea, to thiophosphate and to tetramisole and the same result was obtained: the proportion of PLAP in neutrophils reached 60 --80% of AP activity in pregnancies with a trisomy 21 fetus. Table 1 | AP activity (nkat/mg. protein) and Percentage of PLAP in neutrophils of pregnant women pregnant or not and bearing either trisomy 21 fetus or normal fetus. Discussion : Comparison of alkaline phosphatase isoenzymes necessitates their purification. As the preparation of reasonable amounts of purified alkaline phosphatase from human tissues is a rather complex undertaking , we chose to express two isozymes in vitro. The PLAP gene had been already expressed in vitro using transfected simian cells, baculovirus and Pichia pastoris. Production did not significantly differ from the data reported by Davies et al., i.e. 10 U/ml. In the present study, NSAP was produced in baculovirus infected cells but expression was weak, about 5-fold lower than the expression obtained with PLAP. The lower stability of NSAP compared to PLAP may contribute to this under-expression of the NSAP isozyme. The kinetics of urea denaturation distinguish the two isoenzymes. Denaturation appeared to be monophasic in this study for the two enzymes. This is not in contradiction with the report of Hung and Chang who evidenced a biphasic denaturation of the enzyme because in the first denaturation phase, the enzyme remains fully active, thus this step was not analysed in the present study. The relative resistance of PLAP to denaturation has been known for a long time. Here we confirm this result and we show that the stability originates from a decrease of the rate of the denaturation step leading to reversible non active form. The renaturation rate constant of the two enzymes were not significantly different. PLAP also appeared to be more stable than NSAP to temperature denaturation as first described by McMaster et al. and since repeatedly confirmed . As for urea denaturation, the stability seems to originate from a lower rate of the reversible denaturation while the denaturation rate constants leading to an irreversible denatured form were not significantly different. As denaturation of AP depends on the incubation buffer , we may hypothesize that it would be possible to find conditions for which differences in denaturation between the two enzymes are still more pronounced. By comparison with the GCAP isozyme, which differs from PLAP by only 7 aminoacids, Watanabe et al. identified glutamate 429 as the main amino-acid responsible for the relatively high stability of PLAP. The Km of PLAP for p-nitrophenylphosphate was slightly lower than the Km found by Chang et al.. Determination of Km with the human enzyme from the same source (Sigma) confirmed this difference suggesting that it rather originates from experimental conditions than from the in vitro expression of the enzyme. L-p-bromotetramisole appeared to be a potent inhibitor to AP with a slight specificity for PLAP compared to NSAP. This result is in contrast with the data available for levamisole which, although related to L-p-bromotetramisole, is known to be specific of NSAP . NSAP activity of neutrophils have been reported to increase during pregnancy , and presence of a heat stable AP in neutrophils can be a useful marker for the screening of trisomy 21 fetuses . We sequenced the neutrophil NSAP phosphatase from trisomy 21 pregnant women and trisomy 21 children but no mutation was detected indicating that the differences in alkaline phosphatase characteristics do not originate from a mutant allele of the non specific alkaline phosphatase gene (unpublished data). This suggests that the heat stability of neutrophil AP associated with Down's syndrome may originate from the relative expression of PLAP and NSAP. Our results might indicate that PLAP is expressed in neutrophils and its proportion increases in Down's syndrome pregnancy. However, this conclusion is in contradiction with the report of Peleg et al. which reported no difference in AP stability in neutrophils of pregnant women bearing a trisomy 21 fetus. Among the hypotheses which can be proposed, loss of the PLAP component during enzyme preparation is possible since the authors discarded soluble proteins. The presence of PLAP can be used in conjunction with other markers in the serum of mothers which are currently used to detect Down's syndrome. The screening procedure using the serum concentration of alpha-fetoprotein, human chorionic gonadotropin, pregnancy-associated plasma protein-A and unconjugated oestriol combined with nuchal translucency has a detection rate of 85 --90 per cent with 5 per cent false positives . So, using the presence of PLAP in neutrophils might improve the detection rate. Is the presence of PLAP in neutrophils due to the actual existence of a trisomy 21 fetus ? or is the presence of PLAP in neutrophils responsible for a predisposition of trisomy 21 pregnancy ? As neutrophils of trisomy 21 patients contain only NSAP , we can reject the second hypothesis. Then the presence of PLAP in neutrophils is not genetically determined and seems to be a consequence of the presence of a trisomy 21 fetus. Materials and Methods : Subjects and sample isolation | Nine blood samples from unrelated women bearing a fetus with trisomy 21 were examined. Patients were 36 +/- 6 years of age. Blood was collected during weeks 19 and 20 of gestation following amniocentesis and karyotyping. Two control groups, with same number of samples from women of the same age and at the same gestational age were done: pregnant women with normal pregnancy and not-pregnant women. The permission of all patients was obtained before blood was collected. Neutrophils were immediately isolated by the procedure of Gainer and Stinson . Extraction of the enzyme was immediately performed after blood taking and carried out in 25 mM phosphate buffer pH7 in the presence of 2% Triton X-100. The cells were sonicated and the homogenate spun at 10 krpm, 1 hour at 2C. The clear supernatant was collected and frozen at -20C before biochemical determination for less than three days. Preliminary experiments showed that congelation in these conditions did not affect the phosphatase activity. In vitro gene expression and protein purification | The NSAP encoding gene was cloned by RT-PCR and inserted into the baculovirus transfer vector pBacPaK9 (Clontech). Recombination in the virus BacPAK6 was performed by standard protocols . The placental AP was produced form the recombinant baculovirus constructed by Davis et al.. Enzymes were partially purified on DEAE columns and precipitated by acetone according to Masuhara et al.. AP assays and kinetic measurements | Kinetic measurements were performed in triplicate at 37C, at pH 9.5 in 0.1 mol/L diethanolamine-HCl buffer containing 45 mmol/L MgCl2, with an ionic strength maintained constant (0.2) with NaCl, using sodium p-nitrophenylphosphate as substrate. The amount of reaction product (paranitrophenate) was quantified using absorption at 400 nm (apparent epsilon = 18500 L.mol-1.cm-1 at pH 9.5). All the kinetics were carried out for at least 5 minutes and the initial velocities determined by the slopes using the software included in the Perkin Elmer UV-Visible spectrophotometer Lamda 15. The reversible inhibitors sodium thiophosphate and L-p-bromotetramisole (2,3,5,6-tetrahydro-6-phenylimidazo-[2,1-b]-thiazol; Fig. ) were incubated for five minutes with the enzyme before the addition of substrate. Data were analysed by multiple nonlinear regression using a non-linear global optimization method based on simulated annealing (J. Czaplicki, V. Marcel and D. Fournier, manuscript in preparation) with equations solved according to the rapid equilibrium hypothesis. Figure 1 | Inhibitors and substrate used. Inhibitors and substrate used. Stability of enzymes and sensitivity to inhibitors | For inactivation studies, the concentration of native enzyme was calculated from the residual activity after preincubation of the protein either with urea or at high temperature, without substrate. Urea denaturation was carried out by incubating each protein with freshly 7.4 M urea solution in 0.1 mol/L diethanolamine-HCl buffer pH 9.5, 45 mmol/L MgCl2, 0.2 ionic strength at 20C for 30 minutes. The variation of the remaining proportion of non-denatured enzyme with time was estimated by sampling aliquots every five minutes and recording the remaining activity. As a control, the remaining activity was determined by incubation of proteins without urea. Heat inactivation used the same protocol but incubation was at 56C, 65C or 70C. As the results obtained at the three temperature are in accordance, only the results obtained at 56C are presented. Backmatter: PMID- 11825341 TI - Decreased insulin binding to mononuclear leucocytes and erythrocytes from dogs after S-Nitroso-N-Acetypenicillamine administration AB - Abstract | Background | Nitric oxide (NO) and oxygen free-radicals play an important part in the destruction of beta-cells in auto- immune diabetes although the precise mechanism of interaction is still not known. This study was designed to examine any possible diabetogenic effect of NO by investigating any differences in cellular binding of insulin to its receptor on the cell membranes of erythrocytes and mononuclear leucocytes of dogs treated with the NO donor, S-nitroso-N-acetylpenicillamine (SNAP) and controls treated with captopril. Results | The result obtained showed decreased binding of insulin to its receptor on the cell membranes of erythrocytes and mononuclear leucocytes. Mononuclear leucocytes from SNAP-treated dogs had decreased ability to bind insulin (16.30 +- 1.24 %) when compared to mononuclear leucocytes from captopril-treated controls (20.30 +- 1.93 %). Similar results were obtained for erythrocytes from dogs treated with SNAP (27.20 +- 1.33 %) compared with dogs treated with captopril (34.70 +- 3.58 %). Scatchard analysis demonstrated that this decrease in insulin binding was accounted for by a decrease in insulin receptor sites per cell, with mononuclear leucocytes of SNAP-treated dogs having 55 % less insulin receptor sites per cell compared with those of captopril-treated controls (P < 0.05). Average affinity and kinetic analysis revealed a 35 % decrease in the average receptor affinity, with mononuclear leucocytes from captopril-treated controls having an empty site affinity of 12.36 +- 1.12 x 10-8 M-1 compared with 9.64 +- 0.11 x 10-8 M-1 in SNAP-treated dogs (P < 0.05). Conclusion | These results suggest that acute alteration of the insulin receptor on the membranes of mononuclear leucocytes and erythrocytes occurred in dogs treated with S-nitroso-N-acetylpenicillamine. These findings suggest the first evidence of the novel role of NO as a modulator of insulin binding and the involvement of NO in the aetiology of diabetes mellitus. Keywords: Background : Nitric oxide (NO) is a physiologically important signal molecule regulating a variety of biological functions such as smooth muscle relaxation, neurotransmission, and immune processes. Excess induction of NO, however, has also been implicated as a cause of diverse pathological conditions such as inflammation, neuro-degenerative diseases, cardiovascular disorders, and possibly cancer. These detrimental effects of NO have been attributed to reactive nitrogen species such as oxides of nitrogen (NOx) and peroxynitrite (ONOO-), which are formed by the reaction of NO with oxygen and superoxide respectively . Reactive nitrogen species can oxidize, nitrate, and nitrosate bio-molecules such as proteins, DNA, and lipids, thus altering their functions. S-nitroso-N-acetylpenicillamine has been shown to be potent smooth muscle relaxants in a variety of vascular smooth muscle preparations . It has been shown to produce significant inhibition of spontaneous contractions in myometrial strips from oestrogen-primed, and progesterone-treated rats. The relaxant effect of SNAP was found to be exerted by a mechanism independent of cGMP elevation but may be due to the activation of a calcium-activated potassium channel . Experiments were also done with SNAP, to validate the significance of NO in the development of diabetes. The results showed cleavage of the DNA into nucleosomal fragments . This NO donor cause the activation of poly(ADP ribose) synthase in cell nuclei, and this activation depletes intracellular NAD+ by consuming it as its substrate . It is proposed that internucleosomal DNA cleavage induced by NO activates poly(ADP ribose) synthase, which in turn depletes intracellular NAD+ and causes poly-ADP-ribosylation of nuclear proteins, resulting in the deterioration of beta-cell function . Insulin binding to receptors is the first event in insulin action, and this first step represents a major control point for insulin's effects in vivo. Insulin binding to receptors is not a fixed biologic process, but is subjected to modulation by alterations in either receptor number or affinity . In the present study, we investigated the acute effects of a dose of SNAP on the binding of insulin to its receptor on the cell membranes of mononuclear leucocytes and erythrocytes and whether an insulin-receptor defect is of importance in the hyperglycemic condition. Results : Figure and Figure summarizes the ability of non-radioactive insulin to competitively inhibit the binding of 125-I-insulin to the insulin receptor on the cell membranes of mononuclear leucocytes and erythrocytes respectively, in dogs treated with 20 mg/kg of SNAP, and controls treated with 20 mg/kg of captopril or 2 ml of DMSO. Comparison of the plots showed that insulin receptor on the cell membranes of erythrocytes and mononuclear leucocytes from dogs treated with SNAP binds significantly less 125-I-insulin than cells from the captopril-treated controls and those administered with 2 ml of DMSO at unlabelled insulin concentrations. Figure 1 | Competitive binding curves showing the effect of 20 mg/kg of SNAP , 20 mg/kg of captopril and 2 ml of DMSO on the binding of insulin to its receptor on mononuclear leucocytes. Competitive binding curves showing the effect of 20 mg/kg of SNAP , 20 mg/kg of captopril and 2 ml of DMSO on the binding of insulin to its receptor on mononuclear leucocytes. Percentage of 125-I-insulin. bound is plotted as a function of the unlabelled insulin concentration. Statistical significant differences are indicated by *P < 0.05 and *P < 0.01. Figure 2 | Competitive binding curves showing the effect of 20 mg/kg of SNAP , 20 mg/kg of captopril and 2 ml of DMSO on the binding of insulin to its receptor on erythrocytes. Competitive binding curves showing the effect of 20 mg/kg of SNAP , 20 mg/kg of captopril and 2 ml of DMSO on the binding of insulin to its receptor on erythrocytes. Percentage of 125-I-insulin. bound is plotted as a function of the unlabelled insulin concentration. Statistical significant differences are indicated by *P < 0.05 and *P < 0.01. The percentage 125-I-insulin bound to the insulin receptor on the cell membrane of mononuclear leucocytes of dogs treated with SNAP (16.3 0 +- 1.24 & 10.50 +- 1.14 %) was significantly lower than the percentage 125-I-insulin bound at 1.5 h to those of captopril-treated controls (20.30 +- 1.93 & 16.00 +- 1.12 %) or controls administered with DMSO (19.21 +- 2.21 & 14.15 +- 2.63 %) at very low unlabelled insulin concentrations (0 and 1 ng/ml). Statistical analysis of the percentage insulin bound at these low insulin concentrations revealed that the differences for the SNAP-treated and captopril-treated dogs were statistically significant [P = 0.020 and P = 0.031; Figure . The remainder of the curve showed that percentage 125-I-insulin bound in SNAP-treated dogs being appreciably less than those of the captopril-treated controls, but not statistically significant (P > 0.05), with the exception of the percentage 125-I-insulin bound at 2.5 ng/ml (P = 0.006). A similar trend was seen on examination of the erythrocytes binding profile . The maximum percentage of 125-I-insulin bound to the insulin receptor using captopril was 34.70 +- 3.58 % and dogs administered with DMSO, 32.60 +- 4.67 % compared with 27.20 +- 1.33 % for erythrocytes of dogs with SNAP (P > 0.05). Further analysis showed that the percentage of 125-I-insulin bound to the insulin receptor on the cell membranes of erythrocytes of SNAP-treated dogs at unlabelled insulin concentrations greater than 0.5 ng/ml, were appreciably less and statistically significant than those of the captopril-treated controls (P < 0.05). Comparison of the competition curves of percentage 125-I-insulin bound to the insulin receptor on erythrocytes of the SNAP-treated dogs and captopril-treated controls showed slopes that decreased steadily to values of 15.45 +- 1.02 % and 25.45 +- 1.76 % respectively (P = 0.007) at an unlabelled insulin concentration of 10 ng/ml. The integrated area under the curve for mononuclear leucocytes from dogs treated with 20 mg/kg of SNAP was 88.65 +- 5.28 % x 10 ng/ml compared with 131.76 +- 5.57 % x 10 ng/ml in captopril-treated controls (P < 0.05) and 126.65 +- 4.3.28 % in dogs administered with DMSO. In this study, bound/free (B/F) ratio of the labelled hormone is expressed as a function of the bound hormone giving a Scatchard Plot for mononuclear leucocytes and erythrocytes . Curvilinear plots were obtained for the controls treated with 20 mg/kg of captopril and 2 ml of DMSO, and dogs treated with 20 mg/kg of SNAP. A greater B/F implies that there is more bound hormone than free. Comparison of the plots showed that insulin receptor on the cell membranes of mononuclear leucocytes for captopril-treated controls and dogs treated with DMSO had maximum B/F values of 0.221 +- 0.03 and 0.219 +- 0.13 respectively, compared with 0.182 +- 0.01 for SNAP-treated dogs at the 1.5 h time interval (P < 0.05). Further statistical analysis revealed that although the B/F 125-I-insulin ratios for the SNAP-treated dogs were appreciably less than the captopril-treated controls, the differences were only statistically significant at an unlabelled insulin concentration of 10 ng/ml (P = 0.017). Figure 3 | Scatchard plot showing the effect of 20 mg/kg of SNAP , 20 mg/kg of captopril and 2 ml of DMSO on the binding of insulin to its receptor on mononuclear leucocytes. Scatchard plot showing the effect of 20 mg/kg of SNAP , 20 mg/kg of captopril and 2 ml of DMSO on the binding of insulin to its receptor on mononuclear leucocytes. Bound/Free ratio is plotted as a function of the insulin bound (B). Statistical significant differences are indicated by *P < 0.05 and **P < 0.01. Figure 4 | Scatchard plot showing the effect of 20 mg/kg of SNAP , 20 mg/kg of captopril and 2 ml of DMSO on the binding of insulin to its receptor on erythrocytes. Scatchard plot showing the effect of 20 mg/kg of SNAP , 20 mg/kg of captopril and 2 ml of DMSO on the binding of insulin to its receptor on erythrocytes. Bound/Free ratio is plotted as a function of the insulin bound (B). Statistical significant differences are indicated by *P < 0.05 and **P < 0.01. The slopes of both plots of the bound/free 125-I-insulin ratio of the insulin receptor for erythrocytes from SNAP-treated and captopril-treated dogs, are quite different, as demonstrated in Figure . The bound/free 125-I-insulin ratio of the insulin receptor of dogs treated with SNAP at the 1.5 h time point was 0.374 +- 0.03 compared with 0.531 +- 0.06 for dogs treated with captopril (P = 0.009) and 0.543 +- 0.06 in DMSO-treated dogs. Further statistical analysis of the Scatchard plots indicated that there were statistically significant differences in the bound/free 125-I-insulin ratio of the insulin receptor for dogs treated with SNAP, compared with the captopril-treated controls. This was evident at unlabelled insulin concentrations of 0.1 and 5 ng/ml (P = 0.008 and P = 0.013 respectively). To analyze these changes in affinity more precisely, the data have been plotted on an average receptor affinity graph as described by DeMeyts and Roth . Graphical representation for the average receptor affinity profiles is depicted in Figures and at the 1.5-h time point. Examination of the slopes of the plots for mononuclear leucocytes from SNAP-treated dogs and captopril-treated controls are not different, especially at unlabelled insulin concentrations of 5 ng/ml and 10 ng/ml. However, the average receptor affinity values of the SNAP-treated dogs were slightly less than those of their captopril-treated counterparts at the other unlabelled insulin concentrations (P > 0.05). Figure 5 | Average affinity profile showing the effect of 20 mg/kg of SNAP , 20 mg/kg of captopril and 2 ml of DMSO on the binding of insulin to its receptor on mononuclear leucocytes. Average affinity profile showing the effect of 20 mg/kg of SNAP , 20 mg/kg of captopril and 2 ml of DMSO on the binding of insulin to its receptor on mononuclear leucocytes. Average receptor affinity is plotted as a function of the receptor occupancy. Figure 6 | Average affinity profile showing the effect of 20 mg/kg of SNAP , 20 mg/kg of captopril and 2 ml of DMSO on the binding of insulin to its receptor on erythrocytes. Average affinity profile showing the effect of 20 mg/kg of SNAP , 20 mg/kg of captopril and 2 ml of DMSO on the binding of insulin to its receptor on erythrocytes. Average receptor affinity is plotted as a function of the receptor occupancy. Analysis of the data shows that mononuclear leucocytes of dogs treated with captopril have an 'empty site' affinity (Ke) of 12.36 +- 1.12 x 10-8 M-1, which begin to decrease when approximately 1.48 % of total receptor sites was occupied. With increasing occupancy of the receptor sites by 125-I-insulin, the average receptor affinity progressively decreased to the 'filled site' affinity (Kf) of 5.23 +- 0.46 x 10-8 M-1, when 71.02 % of available receptor sites were occupied . The comparable values of Ke and Kf for mononuclear leucocytes of dogs administered with DMSO were 11.25 +- 0.8 x 10-8 M-1 (2.00 % receptor occupancy) and 5.03 +- 0.45 x 10-8 M-1 (70.01 % receptor occupancy) respectively. The comparable values of Ke and Kf for mononuclear leucocytes of dogs treated with 20 mg/kg of SNAP were 9.64 +- 0.11 x 10-8 M-1 (2.16 % receptor occupancy) and 4.80 +- 0.85 x 10-8 M-1 (66.41 % receptor occupancy) respectively (P < 0.05). Further statistical analysis of the data showed no significant differences in the average receptor affinity values between the groups at low unlabelled insulin concentrations (0.1 -- 2.5 ng/ml). A similar pattern was seen for the average receptor affinity of the insulin receptor on the cell membrane of erythrocytes of the SNAP-treated dogs. The Ke value of the insulin receptor for erythrocytes from dogs treated with SNAP was 8.56 +- 0.79 x 10-8 M-1 when 2.61 % of available receptor sites were occupied. This was lower than that of the captopril-treated controls, 8.96 +- 0.91 x 10-8 M-1 when 1.60 % of available receptor sites were occupied (P > 0.05). The Ke value of the insulin receptor for erythrocytes from dogs administered with DMSO was 8.70 +- 0.68 x 10-8 M-1 when 3.11 % of available receptor sites were occupied. The calculated mean value for Kf for the captopril-treated controls was 4.92 +- 0.43 x 10-8 M-1 at an occupancy of 70.72 % compared with 4.53 +- 0.57 x 10-8 M-1 at an occupancy of 78.30 % for dogs treated with SNAP. The calculated mean value for Kf for dogs treated with DMSO was 4.84 +- 0.20 x 10-8 M-1 at an occupancy level of 71.5 %. The mean differences between empty and filled site affinities of the insulin receptor on the erythrocyte cell membrane of captopril-treated controls and SNAP-treated dogs or between the values of DMSO-treated and SNAP-treated dogs were not statistically significant (P > 0.05). On administration of SNAP, a decrease in the number of insulin receptor sites on the cell membrane on mononuclear leucocytes was observed in the dogs. There was a 55 % decrease in the number of insulin receptor sites per cell of dogs treated with SNAP (11.90 +- 1.57 x 104) compared with that of mononuclear leucocytes for captopril-treated controls (22.10 +- 1.81 x 104) at the 1.5-h time point. The number of insulin receptor sites per cell in dogs treated administered with DMSO is 21.20 +- 1.25 x 104. The decrement in the insulin receptor sites per cell between SNAP-treated and captopril-treated dogs was statistically significant [P = 0.003, Figure ]. Figure 7 | Bar graph showing the effect of 20 mg/kg of SNAP, 20 mg/kg of captopril and 2 ml of DMSO on the number of insulin receptor sites on the cell membranes of mononuclear leucocytes. Bar graph showing the effect of 20 mg/kg of SNAP, 20 mg/kg of captopril and 2 ml of DMSO on the number of insulin receptor sites on the cell membranes of mononuclear leucocytes. Statistical significant differences are indicated by *P < 0.05 and **P < 0.01. Figure showed that the number of insulin receptor sites on the cell membrane of erythrocytes for SNAP-treated dogs was appreciably less than those of captopril-treated dogs at the 1.5-h time point. The calculated insulin receptor sites per cell for dogs treated with captopril and DMSO was 92 +- 6 and 89 +- 6 respectively, compared with 60 +- 4 insulin receptor sites per cell in dogs treated with SNAP at the 1.5-h time point (P = 0.285). Figure 8 | Bar graph showing the effect of 20 mg/kg of SNAP, 20 mg/kg of captopril and 2 ml of DMSO on the number of insulin receptor sites on the cell membranes of erythrocytes. Bar graph showing the effect of 20 mg/kg of SNAP, 20 mg/kg of captopril and 2 ml of DMSO on the number of insulin receptor sites on the cell membranes of erythrocytes. Discussion : This study is a sequel to a previous study by McGrowder et al. which found that SNAP-treated dogs displayed postprandial hyperglycaemia. Captopril was used as the control drug based on results from a study conducted by Winocour et al. which found that low-dose therapy had no significant effect on blood glucose control in hypertensive insulin-treated diabetic individuals and that it lowers blood pressure by a mechanism which is different from SNAP. The study by McGrowder et al. also reported that SNAP at 20 mg/kg caused significant reduction of the haemodynamic parameter, mean arterial pressure and a significant increase in heart rate via the release of nitric oxide. Captopril at a dose of 20 mg/kg had less of an effect on both mean arterial and heart rate. The mean arterial blood pressure-lowering effect of captopril (an angiotensin converting enzyme inhibitor) is related to a reduction in the peripheral arterial vascular resistance. The hypotensive response to captopril is accompanied by a decrease in plasma aldosterone and angiotensin II levels and an increase in plasma renin levels. In response to the decrease in heart rate, there is an increase in heart rate through the baroreflex. Insulin resistance is a common pathological finding in patients with impaired glucose tolerance as well as in subjects with type II diabetes. Our results have shown that the impaired glucose tolerance observed in dogs treated with SNAP is due to decreased insulin binding to receptors. This receptor defect is due to a modulation of the receptor affinity and receptor binding sites which appears to effect physiologic changes in insulin binding. Analysis of the data clearly demonstrated that erythrocytes and mononuclear leucocytes isolated from dogs acutely treated with 20 mg/kg of SNAP have decreased ability to bind insulin. Decreased binding was highly significant at low unlabelled insulin concentrations, i.e. low occupancy levels. Kinetic analysis of the data demonstrated that the decreased binding of insulin to its receptor on erythrocytes and mononuclear leucocytes was attributed primarily to a significant decrease in the number of insulin receptors sites per cell, and secondarily to a reduction in the average affinity of the receptor for insulin. Using the Scatchard method of analysis, graphs of B/F vs [B] for both mononuclear leucocytes and erythrocytes [Figure and Figure ] yielded curvilinear plots with upward concavity. These results suggest that the mononuclear leucocytes and erythrocytes from the dogs used in this study have similar binding kinetics as those expressed in the literature for lymphocytes, mononuclear leucocytes, erythrocytes and the insulin target tissues . The data in Figures and clearly demonstrate that insulin binding is decreased at lower non-radioactive insulin concentrations, but the curves converge at the higher non-radioactive insulin levels. This pattern suggests a change in receptor affinity. The decrease in insulin binding attributed by a decrease in receptor affinity was confirmed by the decrease in the average affinity (K) for both mononuclear leucocytes and erythrocytes. To investigate whether the decrease in insulin binding was attributed to a decrease in the number of receptor sites per cell, the Scatchard plot for the data was analyzed. Using this method of analysis, the x-intercept represents the number of insulin receptor sites per cell. Calculations revealed that there was a 35 % decrease in the number of receptor sites per erythrocyte in SNAP-treated dogs compared with captopril-treated controls. There was a significant 55 % decrease in the number of receptor sites per mononuclear leucocyte from SNAP-treated dogs. Therefore, the decrease in insulin binding by the acute effects of SNAP is primarily as a result of a decrease in the number of receptor sites per cell. The cooperative interactions among receptor sites can be explained in terms of the 'negative cooperativity' model. Negative cooperativity is a frequent occurrence in hormone-receptor systems in which there are site-site interactions, resulting in a decrease in the apparent affinity of receptor for insulin when fractional saturation of the receptor increases. According to this model and calculation of the number of receptor sites using Scatchard analysis, the decreased insulin binding observed is primarily due to a decrease in the number of receptor sites. It can be suggested that that the lower number of receptor sites per cell in dogs treated with SNAP, could be the result of primary alteration in the receptor or may be secondary to some other alteration in the integrity of the membrane . The NO released from SNAP may damage mononuclear leucocytes and erythrocytes in the short term with direct effects on membrane structure, membrane fluidity, cross-linking and function. The molecular nature and site-site interactions of the receptor sites may include mechanisms such as intra-molecular changes in tertiary or quaternary structure of the receptor, association and dissociation of the receptor molecules, clustering of receptors in the membrane or phase transitions in the membrane itself. These changes in membrane integrity could be responsible for the decreased number of receptor sites per cell. The resulting membrane dysfunction can impair transport of glucose across the cell membrane resulting in the observed hyperglycaemia. The contributing effect of any alteration in receptor affinity was evaluated by plotting the data on an average affinity graph as described by De Meyts and Roth . There was a significant 22 percent decrease in the empty site affinity for mononuclear leucocytes, supporting the concept that there can be alterations in both receptor affinity and receptor sites, both contributing to the decreased insulin binding. Some of the insulin receptor on the cell membranes of erythrocytes and mononuclear leucocytes from dogs treated with SNAP may be affected by the cytotoxic effect of NO and may become desensitized. Desensitization is associated with a total lack of insulin effect despite remaining insulin receptors. Several possibilities exist to explain the mechanism of changing receptor affinity and desensitization. Firstly, fluidity may be an important factor in modulating insulin binding and action. Secondly, the insulin receptor may be covalently associated with another protein that modulates receptor affinity. It is therefore a possibility that NO released from SNAP may alter the interaction of insulin with its receptors, thus affecting the ability of insulin to differentially regulates its receptor and this regulator protein . Nitric oxide may also affect the formation of the insulin receptor complex. A third possibility is that the receptor undergoes some form of post-translational modification that alters binding and signal transmission properties . The post-translational modification could involve a change in the redox state of the receptor. The insulin receptor is composed of major sub-units linked by disulfide bonds to various oligomeric forms. Reduction of the oxidized forms of the receptor could modify the affinity of insulin . Both circulating erythrocytes and mononuclear leucocytes exhibit insulin receptors. The most remarkable difference between these two cell types is the age dependence of the insulin receptor on erythrocytes. Clearly major perturbations in the mean cell age of an erythrocyte preparation will have some effect on insulin binding. Therefore there is more variability in insulin binding to erythrocytes than mononuclear leucocytes at the limiting, low bound/free ratios of 125-I-insulin. This variability affects the affinity profiles derived from this data. The advantage of using these circulating blood cells for investigating the receptor status in human and animals is that they are more easily accessible than cells of primary insulin target organs, such as adipocytes and muscles. It has been found that the characteristic of insulin binding to human adipocytes (including affinity constants for the binding reaction) were similar to the characteristics of insulin binding to mononuclear leucocytes in obese subjects . Other studies have also confirmed that changes in insulin receptors of mononuclear leucocytes seem to mirror the events of more traditional insulin target tissues . Therefore mononuclear leucocytes represents a more uniform population of cells capable of same receptor-mediated function as adipocytes, providing a clearer reflection of the insulin receptor status in target tissues. The main objective of the study was to evaluate any binding abnormalities in dogs with impaired glucose tolerance as a result of treatment with the nitric oxide donor, S-nitroso-N-acetylpenicillamine. This result is useful in elucidating the possible diabetogenicity of nitric oxide. Whether or not NO plays a role in the inhibition of insulin action is presently under investigation in our laboratory. We will examine some physiological assessment of the cells and measurement of post-receptor signaling such as tyrosine kinase activity, insulin receptor substrate-1 (IRS-1) phosphorylation, and phosphotidylinositol 3-kinase activity. Conclusion : In summary, the present report demonstrated that impaired glucose tolerance observed in SNAP-treated is associated with a decrease in insulin receptor binding. The mechanism of decreased insulin binding was due primarily to a significant reduction in the number of receptor sites per cell and to a lesser extent, a reduction in the average receptor affinity. These findings for the first time suggest a novel role of NO as a modulator of insulin binding. Materials and Methods : Animals -- Experimental design | The protocol was conducted in accordance with the guidelines of the University of the West Indies Animal Committee. Eighteen normal Mongrel dogs (9 males and 9 females) of 2 --3 years, with average weight of 11 +- 0.4 kg were obtained from the Pre-Clinical Animal House of the Department of Basic Medical Sciences, University of the West Indies. The animals were maintained in the animal house under the supervision of attendants and a veterinary consultant. Dogs were fed on a diet of Purina Laboratory Chow (Purina, St. Louis MO. U.S.A.) and water administered ad libitum. The dogs were divided into three groups, test and controls. An oral glucose tolerance test was performed on each dog. Briefly, after an 18-h fast, dogs were anaesthetized with sodium pentabarbital (30 mg/kg i.v.). Subsequently, the fasting blood sample was taken and in control experiments 20 mg/kg body weight of captopril (Sigma Chemicals Co. Ltd. St. Louis MO. USA) dissolved in water was administered intravenously. In test experiments, S-nitroso-N-Acetylpenicillamine (SNAP; Sigma Chemicals Co. Ltd. St. Louis MO. USA) was dissolved in 2 ml of dimethyl sulphoxide (DMSO at a concentration of 0.0014 mol/L) and administered at 20 mg/kg body weight to the dogs. Additional blood samples were collected 1.5-h after administration of a glucose load of 1.75 g/kg body weight. In a previous study by McGrowder et al. pilot experiments were carried out where a low dose of 5 mg/kg body weight of SNAP or captopril was first administered and its effect on the blood glucose observed during oral glucose tolerance test. Subsequent experiments involved increasing the doses of SNAP and captopril, based on the effect of the previous dose on the blood glucose levels. The result showed that SNAP at 20 mg/kg had a significant effect on blood glucose levels especially at the 1.5-h time point, while a similar dose of captopril had little or no effect. In this study blood samples were taken at the 1.5-h time point. Each sample was collected in an EDTA tube and immediately placed on ice for subsequent biochemical analysis. Preparation of purified erythrocytes | The mononuclear leucocytes and erythrocytes were separated using a Percoll density gradient. Mononuclear leucocytes were separated from the erythrocytes by the use of pasteur pipettes. The erythrocytes receptor assay was performed according to a modification of the method by Ghambir et al . The erythrocytes were washed three times by centrifugation (4C, 4500 rpm) in 10 ml of buffer G containing 1% human serum albumin, (pH 7.8) for 10 minutes. On each occasion, the supernatant was removed and the cells re-suspended in buffer G, and re-spun. After the final wash of the cells, the supernatant was removed and the cells were left in 4 ml of buffer G containing 1 % human serum albumin. This suspension contained 4 --6 x 109 cells/ml. Preparation of purified mononuclear leucocytes | Mononuclear leucocytes were re-suspended in 4 ml of cold Buffer M and centrifuged at 1100 rpm for 10 minutes at 4C. This was repeated three times, and on each occasion the supernatant was removed and the cells re-suspended in buffer M. After the final washing, the cells were left in 1 % human serum albumin (approximately 2 ml) in buffer M. The mononuclear leucocytes had a concentration of 1.70 -- 2.00 x 107 cells/ml. Binding of 125-I to mononuclear cells and erythrocyte cells | Receptor binding were evaluated by the use of the Scatchard analysis of competitive-inhibition curves, which is a modification of the original technique described by Gambir et al . Lyophilized unlabelled non-radioactive insulin (10 mg; Sigma porcine, 23.5 U/mg, Sigma Chemical Co., St. Louis, MO USA) was reconstituted in 0.1 M HCl (10 ml) and 100 mul quantities were dispensed and stored at -70C until required. Monoiodinated A14 [125-I] insulin (Amershan, Arlington Heights, IL USA; specific radioactivity 50 muCi) was dissolved in deionised water (500 mul) and aliquots of 5 mul stored at -70C until needed. One hundred and fifty (150) mul aliquots of the washed erythrocyte suspension were added to a series of Eppendorf tubes in triplicate. To each sample tube was added, standard (50 mul) and 125-I-insulin (50 mul) and the incubated for 3 hours. Similar experiments were carried out using mononuclear leucocytes. The mononuclear leucocyte reaction was stopped with cold buffer M and the tubes were centrifuged at 1100 rpm for 2 minutes. The cells were washed twice, and the radioactivity of the pellet was determined using a Gamma Counter (Abbott Auto Logic Gamma Counter). The erythrocyte reaction was stopped by adding 1 ml of cold saline and centrifuged at 1100 rpm for 1 minute. The supernatant was discarded and 100 mul of 40 % formalin was added to harden the red cell pellet. The radioactivity of the pellet was also determined using a Gamma Counter. The data was analyzed by Scatchard analysis . The receptor affinity and receptor numbers were derived for the physiological range of insulin i.e. between 0.1 and 100 ng/ml. Specific insulin binding (SB) was calculated as the percentage of radioactive insulin bound by 4 x 109 and 2 x 106 cells/ml for erythrocytes and mononuclear leucocytes, respectively. Non-specific binding was assessed by the amount of radioactive insulin bind in the presence of 100 ng/ml unlabelled insulin. Competitive binding curves were obtained for each erythrocyte and mononuclear leucocyte suspension. From these curves, the insulin receptor affinity and number of the receptor sites were determined by the Scatchard analysis . Cell binding analysis | The results of the binding studies are presented in three ways: (1) the percentage binding of 125-I-insulin as a function of the total insulin concentration (competitive curve), (2) the bound-free insulin ratio plotted as a function of the bound insulin (Scatchard plot) and, (3) the average affinity profile calculated according to the method of De Meyts and Roth . The total binding capacity or concentration of the binding sites was derived from the point where the linear extrapolation of the curve intercepts the horizontal axis and this was used to calculate the number of receptor sites per cell . Experimental data suggest that the insulin receptor consist of homologous binding sites that undergo negatively cooperative site-site interactions such that the affinity of the receptors for insulin is inversely related to the receptor occupancy. The average affinity profile expresses the relationship between the average affinity for insulin (K) and the receptor occupancy (Y). The average affinity falls as a function of receptor occupancy (negative co-operativity) until the lowest observable affinity (Kf) is reached. The fractional occupancy necessary to produce Kf is designated yf, and Ke represents the highest observable affinity of the receptors and is exhibited in the native or 'empty site' state. Calculations and statistical analysis | All results shown in the figures are expressed as means +- S.E.M. Integrated area under the curve (iUAC) was calculated by subtracting the rectangle corresponding to the basal value from the total area under the curve . Analysis of the data was done using the Sigma Plot and Sigma Statistics software packages (Jandel Scientific). To evaluate the effects of SNAP and captopril the biochemical parameters, values for each group were compared by either a paired student's test or two-way analysis of variance (ANOVA) followed by the Bonferroni multiple comparison test . P values less than 0.05 was considered to indicate significance in all cases. Abbreviations : SNAP, S-nitroso-N-acetylpenicillamine; DMSO; dimethyl sulfoxide; NO, nitric oxide, DNA, deoxyribonucleic acid; cGMP, cyclic guanosine monophosphate; SB, specific insulin binding. Backmatter: PMID- 11825346 TI - Quality and methods of developing practice guidelines AB - Abstract | Background | It is not known whether there are differences in the quality and recommendations between evidence-based (EB) and consensus-based (CB) guidelines. We used breast cancer guidelines as a case study to assess for these differences. Methods | Five different instruments to evaluate the quality of guidelines were identified by a literature search. We also searched MEDLINE and the Internet to locate 8 breast cancer guidelines. These guidelines were classified in three categories: evidence based, consensus based and consensus based with no explicit consideration of evidence (CB-EB). Each guideline was evaluated by three of the authors using each of the instruments. For each guideline we assessed the agreement among 14 decision points which were selected from the NCCN (National Cancer Comprehensive Network) guidelines algorithm. For each decision point we recorded the level of the quality of the information used to support it. A regression analysis was performed to assess if the percentage of high quality evidence used in the guidelines development was related to the overall quality of the guidelines. Results | Three guidelines were classified as EB, three as CB-EB and two as CB. The EB guidelines scored better than CB, with the CB-EB scoring in the middle among all instruments for guidelines quality assessment. No major disagreement in recommendations was detected among the guidelines regardless of the method used for development, but the EB guidelines had a better agreement with the benchmark guideline for any decision point. When the source of evidence used to support decision were of high quality, we found a higher level of full agreement among the guidelines' recommendations. Up to 94% of variation in the quality score among guidelines could be explained by the quality of evidence used for guidelines development. Conclusion | EB guidelines have a better quality than CB guidelines and CB-EB guidelines. Explicit use of high quality evidence can lead to a better agreement among recommendations. However, no major disagreement among guidelines was noted regardless of the method for their development. Keywords: Background : The objective of guidelines development is to assist physicians and patients in making optimal health care decisions, which in turn should improve the quality of clinical practice . Different methods are used to develop guidelines. Some are developed by a consensus of experts while others also use a formal way to appraise the literature and create evidence-based (EB) guidelines. In general, evidence-based guidelines are considered to provide better recommendations for practice than consensus-based guidelines but are time consuming and expensive to create . This belief that EB guidelines are superior to other types of guideline is based on our normative views of methods for guidelines development and not on empirical comparison of practice recommendations using different methods for development of guidelines. To date no formal evaluation has been performed to detect if there are differences in the quality and recommendations between evidence-based and consensus-based (CB) guidelines. If guidelines developed by using consensus or evidence-based methods have the same quality and agree in the recommendations, then obviously resources spent on the laborious and time-consuming process of locating and appraising evidence can be used elsewhere. Otherwise, if evidence based guidelines have a better quality and their recommendations differ from those guidelines produced by consensus, then creation of evidence based guidelines may become the only acceptable method of guideline development. In this paper, we explore if there are differences in the quality and recommendations between EB and CB guidelines. Methods : To enable meaningful comparison, multiple recommendations produced by a given guideline method should be available. This objective is best met by focusing on the guidelines that comprehensively attempt to guide clinicians in the management of one disorder. Since breast cancer is an important disease and various organizations have produced guidelines using different methods , we conducted a comparison study of comprehensive breast cancer guidelines. We assessed both the differences in the quality as measured by using different quality instruments assessment and the level of agreement among guidelines according to the method of development. 1. Identification and assessment of instruments for measurement of the quality of guidelines Since there is no uniformly accepted instrument for evaluation of the quality of guidelines, we first performed a comprehensive literature search to identify published tools for assessment of clinical practice guideline quality. We searched MEDLINE (1996 --2000) using the keywords: guidelines, practice guidelines, quality, "weights and measures", "scale", psychometrics, reproducibility. Any article considered relevant to evaluate quality of guidelines was retrieved. The list of references of each article was also scanned. After an assessment of 14 papers by four of us, four instruments to assess the quality of guidelines were identified . An additional instrument (SIGN) was identified through Evidence-based Health Discussion Group . For additional details on the instruments for evaluation of guidelines readers are referred to the Appendix (see ). To assess their reliability and reproducibility, we applied all identified instruments to each guideline (see below). We calculated the coefficient of agreement (kappa) among evaluators for each guideline . A good interobserver agreement was considered if kappa value exceeded 0.4 . In our evaluation, two instruments had a kappa interobserver agreement K > 0.4 among all investigators in 6 of 8 guidelines . When it comes to evaluation of the quality of breast cancer guidelines these instruments performed better than others and probably can be recommended for future use. Table 1 | Interobserver agreement of instruments for assessment of the guidelines quality 2. Identification and classification of breast cancer guidelines A literature search was conducted for published breast cancer guidelines using MEDLINE for the years 1996 -- April 2000. The following keywords were used in combination: Guidelines, Practice Guidelines, recommendations, breast neoplasms. An Internet search was also performed, using the method described by Sanders et al . 131 articles were retrieved, and reviewed for their content. We considered any article that fit the definition of the National Library of Medicine for practice guidelines: directions or principles presenting current or future rules of policy for the health care practitioner to assist him in patient care decisions regarding diagnosis, therapy, or related clinical circumstances . Eight papers referred to breast cancer guidelines , and were selected for the analysis. Each guideline was classified as CB, when there was no consideration about the quality of evidence used to make practice recommendations; as EB, when there was an explicit consideration of the quality of evidence in the development of guidelines; or as consensus based with no explicit consideration of evidence (CB-EB) when there were considerations about the evidence, but not in explicit manner. From these eight guidelines, three were classified as EB three as CB-EB and two as CB . Table 2 | Classification of Breast Cancer Guidelines according to the method of development. 3. Evaluation of guidelines Each guideline was evaluated independently by three of us using each of the instruments. All discordances were resolved by a consensus meeting. Each guideline was scored according to the instructions of each instrument. The quality and rank was determined by the quotient of items scored positively by the total items scored for each instrument. 4. Evaluation of agreement among guidelines Using instruments to evaluate practice guidelines yields conclusions regarding normative aspects of the guidelines development , but does not necessarily mean that recommendations provided by guidelines using different methods will produce different management advice to our patients. To assess if recommendations among various guidelines differ, we need to determine the level of agreement among guidelines for each specific decision point. Since NCCN (National Comprehensive Cancer Network) guidelines were presented in explicit, algorithmic format, we used this one to identify the decision points for matched comparison with other guidelines. These guidelines have been developed by the leading 18 cancer institutions in the US and have been constantly updated and re-evaluated. They have also been developed to closely mimic clinical practice. Therefore, we feel that selection of decision points based on the NCCN guidelines were appropriate. We identified fourteen decision points in the management of stage I and II breast cancer that were linked to specific recommendations in the other guidelines for our comparison. Comparison of recommendations for advanced stages of breast cancer has not been performed since there was only one guideline that included it . Subsequently, four of us evaluated each of these decision points in each guideline examining level of agreement among various guidelines. Since matching between recommendations in the guidelines that were presented in non-algorithmic format was poor, we decided to use NCCN guidelines as a benchmark. We classified agreement of each guideline with the NCCN guidelines as having full agreement, partial agreement and disagreement. It was considered that guidelines agree with the NCCN if the management recommendation was the same; the guidelines were considered to disagree if they provided different recommendations. A partial agreement was judged to exist if the guideline recommended the same management but in a broadly defined sense and not in explicit, clear manner. Each of these decision points was also classified as supported by high quality evidence or not. High quality evidence was considered to be based on randomized trials (RCT) or systematic reviews (SR)/meta-analysis (MA). If the quality evidence was not based on RCT or SR/MA or was not stated, it was classified as low quality evidence. Subsequently, we performed a regression analysis to assess the contribution of the quality of evidence to the total score obtained by each instrument for the evaluation of the guidelines quality. Independent variable was the proportion of decisions supported by high quality evidence while dependent variable was score obtained by each instrument. A regression analysis was performed after it has assessed that the distribution of the variables was normal by Wilks-Shapiro test. Results : Evaluation of the quality of guidelines | The results of the quality of each guideline according to each instrument are shown in Table . Overall, EB guidelines had higher scores than CB, and the CB-EB category ranked in the middle . As expected, the instruments for the evaluation of quality are based on the number of desired built-in normative features of good guidelines development, as initially recommended by Institute of Medicine. This is further confirmed by the evaluation of the contribution of the quality of evidence to the final quality score: the regression analysis performed showed that the quality of guidelines, as measured by these instruments, is a function of the percentage of high quality evidence that each guideline contains. This suggests that evidence plays a major role in the composition of the quality scales. If the quality of evidence is poor, paying attention to other quality domains in the development of guidelines will not result in higher quality scores. Fig illustrates a relationship between the quality of evidence and the total quality score using the two instruments that achieved best agreement among evaluators . It is quite remarkable to note that up to >94% variation in the score could be explained by the quality of evidence alone. Figure 1 | Average score of each guideline according to the method of development Average score of each guideline according to the method of development Acronyms and abbreviations: ACCC -- Association of Community Cancer Centers; CMA-Canadian Medical Association; ICSI -- Institute for Clinical Systems Improvement; MPS -- Multi Professional Societies; NCCN -- National Comprehensive Cancer Network; NHMRC -- The National Health and Medical Research Council; SIGN -- Scottish Intercollegiate Guidelines Network; SSO -- Society of Surgical Oncology. EB: evidence-based guidelines; CB: consensus-based guidelines EB-CB: consensus-based guidelines with no explicit considerations of evidence Figure 2 | A relationship between quality of evidence and total guideline quality score. A relationship between quality of evidence and total guideline quality score. Note that up to 94% of variation in the quality score can be explained by the quality of evidence. Table 3 | Quality of breast cancer guidelines Evaluation of agreement among guidelines | The agreement among each guideline for the 14 decision points is shown in Table . We obtained no major disagreements among guidelines, but the EB guidelines had a better agreement with the decision points in any situation than CB-guidelines and CB-EB guidelines. The fact that no major disagreements were seen regardless the method of development can probably be explained by the vagueness of recommendations by CB guidelines. As shown in Table , the number of decision points supported by high quality evidence is highest in the EB guidelines and zero in CB guidelines. The use of high quality evidence was significantly associated with a higher level of concordance among the decision points. When the source of evidence was of good quality (RCT or SR), we had 18 full agreements and 23 partial agreements (Chi square = 0.610, degrees of freedom = 1, p = 0.435). When the source of evidence was not stated or was of lower quality, we had 17 full agreements and 40 partial agreements (Chi-Square 9.281, Degrees of freedom 2, p= 0.002). This means that recommendations based on high quality evidence may lead to less disagreement and potentially less practice variation. Table 4 | Level of agreement between NCCN guideline and other breast cancer guidelines. Discussion : Guidelines have been increasingly used in medical decision-making. Different methods have been used in guideline development. Does it matter how guidelines were produced? Most authors believe that it matters very much and that guidelines produced using evidence-based methods are superior to other methodologies of development . However, empirical investigations to assess if guidelines produced by different methods have different quality and result in different recommendations have not been performed. Here, we report such a study. Using formal instruments for evaluation of the quality of guidelines we found that EB-guidelines had substantially higher score than CB-guidelines or guideline that considered evidence in a less formal way (CB-EB). As discussed above (see Results), this is not a surprising result, since the instruments for the guidelines evaluation measure the quality based on the number of desired normative characteristics in a particular guideline. Since appraisal of evidence is considered inherently important for the development of a good guideline, one would then expect that the guidelines that pay more attention to its evidence basis (i.e., those that are evidence-based) would receive higher quality score than other types of the guidelines (i.e. guidelines developed solely by a consensus process) (see Fig ). This is also evident in our finding that variation in the total quality score can be up to 94% explained by the quality of evidence (see Fig ). Not all instruments for evaluation of guidelines performed equally well. Only two of the instruments available to address the quality of guidelines had a good level of agreement among evaluators (k > 0.4) in most of guidelines. This result raises concern about the reproducibility of results using the other instruments reported in the literature. In general, a few studies have been done to evaluate reproducibility of the instruments for assessment of the guidelines quality. Any future study attempting to address the quality of guidelines should take this finding into account. A more interesting question is to assess if the recommendations among guidelines produced by different methods actually differ. We found no instance of total disagreement among guidelines regardless of the method of development. We also found that EB and CB-EB guidelines had more points of agreement with our benchmark guidelines (NCCN) than guidelines developed using exclusively consensus method. We also found that when high-quality evidence existed in the literature (see Results) less disagreement was found among various guidelines. This is not completely surprising because formulation of guidelines does not happen in a vacuum. Most guideline developers are experts in the field who have knowledge of the literature. When evidence is unequivocal, less disagreement may be expected. Consequently, less practice variation may be found when high-quality evidence exists. Conclusions : In conclusion, EB guidelines have a better quality than CB guidelines as measured by the quality assessment instruments used in this study. The explicit use of high quality evidence is desirable and can lead to a better agreement among recommendations. However, no major disagreement among guidelines was noted regardless of the method for their development. Competing Interest : none declared Acknowledgements : We thank Dr.Stephen Edge for reviewing our paper and his helpful comments and constructive critique. Backmatter: PMID- 11835696 TI - Characterization of a protease produced by a Trichoderma harzianum isolate which controls cocoa plant witches' broom disease AB - Abstract | Background | Several Trichoderma strains have been reported to be effective in controlling plant diseases, and the action of fungal hydrolytic enzymes has been considered as the main mechanism involved in the antagonistic process. However, although Trichoderma strains were found to impair development of Crinipellis perniciosa, the causal agent of cocoa plant witches' broom disease, no fungal strain is available for effective control of this disease. We have then undertaken a program of construction of hydrolytic enzyme-overproducing Trichoderma strains aiming improvement of the fungal antagonistic capacity. The protease of an indian Trichoderma isolate showing antagonistic activity against C. perniciosa was purified to homogeneity and characterized for its kinetic properties and action on the phytopathogen cell wall. Results | A protease produced by the Trichoderma harzianum isolate 1051 was purified to homogeneity by precipitation with ammonium sulfate followed by hydrophobic chromatography. The molecular mass of this protease as determined by SDS-polyacrylamide gel electrophoresis was about 18.8 kDa. Its N-terminal amino acid sequence shares no homology with any other protease. The purified enzyme substantially affected the cell wall of the phytopathogen C. perniciosa. Western-blotting analysis showed that the enzyme was present in the culture supernatant 24 h after the Trichoderma started to grow in casein-containing liquid medium. Conclusions | The capacity of the Trichoderma harzianum protease to hydrolyze the cell wall of C. perniciosa indicates that this enzyme may be actually involved in the antagonistic process between the two fungi. This fact strongly suggest that hydrolytic enzyme over-producing transgenic fungi may show superior biocontrol capacity. Keywords: Background : Since the pioneering work of Weindling and Fawcett on the use of Trichoderma strains to control damping-off caused by Rhizoctonia solani Kuhn in citrus, considerable attention has been focused over the past 20 years on the isolation of fungal antagonist that could be as effective as pesticides in the repression of fungal pathogens . However, the molecular basis of biological control is not clearly understood. Proposed mechanisms resulting in biocontrol are competition for the substrate , the ability to colonize the niche favored by the pathogen, antagonism by antibiotics and the action of cell wall-degrading enzymes . It has been claimed that a battery of antibiotic substances produced by Trichoderma spp. are responsible for their antagonistic properties . On the other hand, chitinolytic enzymes produced by Trichoderma harzianum Rifai have been suggested as being responsible for the observed antifungal activity . Indirect evidence suggested that T. harzianum antagonize first and foremost by antibiosis leading to cell death, followed by degradation of the cell wall by chitinolytic enzymes . In addition, T. harzianum growing in liquid medium containing Botrytis cinerea cell wall produced considerable proteolytic activity , which was suggested to be involved in the antagonistic process. This possibility is further supported by the finding that an alkaline protease (prb1) is also induced in the presence of Rhizoctonia solani cell wall . Cocoa plant witches' broom is a serious disease caused by the hemibiotrophic fungus Crinipellis perniciosa. This fungus is currently causing severe damage to many cocoa plantations in South America. Phytosanitation and, to a lesser extent, chemical control are recommended in certain situations but are not effective and/or practical in all circumstances. Nevertheless, some Trichoderma spp isolates were previously found to control the development of C. perniciosa in field conditions and, consequently, to control the disease (Jose L. Bezerra, personal comm.). Production of hydrolytic enzymes by two of these fungal isolates (1051 and TVC) in liquid medium was then studied . Both fungi were found to produce substantial amounts of chitinases, N-acetylglucosaminidase, beta-glucanases, cellulases, amylases and proteases. One chitinase and one amylase produced by the isolate 1051 were purified and characterized . The chitinase drastically affected the C. perniciosa cell wall in vitro, indicating that this enzyme may play an important role during the antagonistic process between T. harzianum and the phytopathogen C. perniciosa. Nevertheless, substantial information provides support for the concept that the extraordinary capacity of T. harzianum to attack the structures of phytopathogens is related to the synergistic action of hydrolytic enzymes including chitinases, glucanases, lipases and proteases . The properties of several hydrolases produced by fungi showing antagonistic capacity against the causal agent of cocoa plant witches' broom disease are now being investigated in this laboratory . Here, we report the purification and some properties of one protease produced by T. harzianum 1051 grown in liquid medium containing casein as carbon source. Materials and methods : Microorganisms | Trichoderma harzianum, isolate 1051, was obtained from the collection of the Centro Nacional de Pesquisa de Monitoramento e Avaliacao de Impacto Ambiental (CNPMA/EMBRAPA, Jaguariuna, SP, Brasil). The phytopathogen Crinipellis perniciosa was kindly provided by Dr. Jose Luiz Bezerra from CEPLAC/Ilheus, BA, Brasil. Both fungi were maintained by serial passages in Bacto-dextrose-agar medium. Enzyme production and assay | For enzyme production, T. harzianum was grown in TLE liquid medium as previously described , for 72 h at 28C. Protease activity was assayed using a reaction system containing 500 muL of a 1% solution of hammarstein casein dissolved in 0.1 M HEPES buffer, pH 8.0, 250 muL of the HEPES buffer, and 0 --250 muL of enzyme solution. Reaction was conducted for 20 min at 37C and stopped by addition of 1.25 mL of 10% TCA, incubation (40 min) on ice and centrifugation (10 min, 5,000 rpm). The absorbance of the supernatant was determined at 280 nm. One unit of proteolytic activity corresponds to the amount of enzyme required to cause an increase of 1.0 A280 nm unit in the absorbance of the supernatant within 20 min. All enzyme assays were carried out in triplicate. In all cases, the standard deviation values were smaller than 3% of the mean values. Protease purification | Four 1 liter-erlenmeyer flasks, each containing 500 mL of TLE liquid culture medium [0.1 % bactopeptone; 0.03 % urea; 0.2 % KH2PO4; 0.14 % (NH4)2SO4; 0.03 % MgSO4.7H2O; 0.03 % CaCl2.6H2O; 1 mL of 0.01 % trace elements solution (Fe2+, Mn2+, Zn2+, and Co2+), 0.02 % glucose, pH 5.5) containing 0.5% casein] were inoculated with 2 x 107 spores of T. harzianum 1051 and the culture incubated for 72 h with agitation (120 rpm) at 28C. The culture supernatant were then collected by filtration on filter paper and the proteins precipitated with 20% ammonium sulfate. This was followed by centrifugation at 28.800 g for 30 min, and reprecipitation of the supernatant with 60% ammonium sulfate. The resulting pellet (hereafter called crude enzyme extract) was ressuspended in 50 mM sodium phosphate buffer, pH 5.0, and loaded on a Phenyl-Sepharose column (10 x 1.8 cm) equilibrated with the phosphate buffer. Proteins were then eluted with a negative linear gradient formed with 50 mL of 50 mM sodium phosphate buffer, pH 5.0, containing 1.0 M ammonium sulfate, and 50 ml of the same buffer. Samples of 3.0 ml were collected at a flow rate of 54 mL/h. SDS-PAGE and immunoblotting analysis | Protease samples were submitted to electrophoresis in polyacrylamide gels under denaturing conditions, according to Laemmli . Proteins present in the gels were stained with silver reagent according to the method of Blum et al.. For immunoblotting analysis, proteins present in the SDS-polyacrylamide gels were transferred to a nitrocellulose membrane using a LKB Multiphor II (Pharmacia). The membranes were then incubated overnight with 2 % skimmed milk solution at 4C, and the proteins immunodetected using the polyclonal mouse antibody specific for the T. harzianum 1051 protease, alkaline phosphatase-linked mouse anti-IgG antibody (Sigma Chem.Co) and color development reagent (132 muL of nitro blue tetrazolium + 66 muL of 5-bromo-4-chloro-3-indolyl dissolved in 20 mL of phosphate buffered saline). For production of the mouse polyclonal antibody against the T. harzianum 1051 protease, three mices received intraperitoneally at 20-day intervals three doses (30 mug each) of the purified protease. The first injection was given with complete Freund's adjuvant, the second with incomplete Freund's adjuvant and the third without any adjuvant. The animals were bled 45 days after the first injection and serum was collected by centrifugation after blood clotting. Protein sequencing | The N-terminal amino acid sequence of the Trichoderma 1051 protease was determined by the Edman degradation method using an automatic sequencer (Applied Biosystems 477A Automatic Protein Sequencer) fitted on line to a 120 A PTH-analyzer. Comparison of the protease amino acid sequence with other proteins was done using the Swiss-Prot databank. Enzymatic digestion of Crinipellis perniciosa mycelium and scanning electron microscopy (SEM) analysis | Disks (5 mm) of solid culture medium containing C. perniciosa mycelium were placed in microcentrifuge tubes containing 1 mL of either T. harzianum 1051 culture supernatant showing proteolytic activity (220 U/mL) or the corresponding purified protease (75 U/mL). After incubation for 0, 24 and 48 h at 37C, the supernatants were used for determination of reducing sugars by the dinitrosalicylic acid method (DNS assay) according to Miller , glucose by the glucose oxidase method and protein by a simplification of the method of Lowry et al.. The C. perniciosa mycelium disks submitted to the action of the protease samples were then fixed for 2 h at 4C in 2% (v/v) glutaraldehyde and 2% (w/v) paraformaldehyde buffered with 0.05% sodium cacodylate buffer, pH 7.2, and postfixed in 1% (w/v) osmium tetroxide in the same buffer. The specimens were washed with buffer and dehydrated in 30 --100% (v/v) acetone, critical-point dried in CO2 and sputter-coated with gold. Materials were examined microscopically at an accelerating voltage of 10.0 kV. Results and discussion : T. harzianum 1051 growing in casein-containing liquid medium produces substantial proteolytic activity as previously reported . The supernatant of T. harzianum 1051 cultures were then used to purify the protein showing proteolytic activity against casein. Upon freeze-drying of the culture supernatant, there was no significant loss of enzyme activity. The proteases present in the freeze-dried culture supernatant reconstituted with water were only slightly precipitated with ammonium sulfate at the concentration of 20%, but was fully precipitated with this salt at the final concentration of 60 % at pH 5.0. We then used the enzyme sample obtained by precipitation with 20 % and 60 % ammonium sulfate for further purification of the enzyme using a Phenyl-Sepharose column . The proteins were eluted as several peaks both before and after the ammonium sulfate gradient was started. Nevertheless, the bulk proteolytic activity was eluted as a single peak, i.e., only the proteins eluted with an ammonium sulfate concentration of 0.23 --0.16 M (fractions 28 --37) were active against casein. The enzyme sample resulting from pooling of fractions 30 --37 contained a single protein as demonstrated by SDS-PAGE analysis . The protein fractions eluted from the column after the ammonium sulfate gradient showed very low proteolytic activity and several protein species, as analyzed by SDS-PAGE (not shown). The summary of the purification protocol indicated that the enzyme was purified 4.06-fold (Tab. ). The higher amount of total protein in the pellet resulting from precipitation of the freeze-dried culture supernatant with ammonium sulfate must be resulting from artifact, since this salt may react with Lowry reagents. The molecular mass of the enzyme was estimated to be 18.8 kDa . Figure 1 | Chromatography of the crude enzyme preparation of Trichoderma harzianum 1051 on a Phenyl-Sepharose column. Chromatography of the crude enzyme preparation of Trichoderma harzianum 1051 on a Phenyl-Sepharose column. protease activity; Absorbance at 280 nm; --- (NH4) 2SO4 negative gradient. Figure 2 | SDS-PAGE of the protease from Trichoderma harzianum 1051 purified on a Phenyl-Sepharose column. SDS-PAGE of the protease from Trichoderma harzianum 1051 purified on a Phenyl-Sepharose column. Mw, molecular weight standards; A, purified protease. Proteins were silver stained. Table 1 | Summary of the purification protocol of the protease produced by Trichoderma harzianum 1051. The effect of the pH on the protease activity was determined at the pH range of 7.0 --10.0. At this range, maximal activity was determined at the pH values of 7.0 and 8.0, but substantial proteolytic activities were encountered at pH 9.0 and 10.0 . The effect of acidic pH (<7.0) could not be determined, since the casein precipitates at acid pH conditions. This fact has made it difficult to determine the optimal pH for the purified protease, and the presented data do not exclude the possibility that the enzyme is mostly active at pH values lower than 7.0. The optimal pH value determined for the proteolytic activity produced by T. viride in mixed culture with Sclerotium rolfsii in soil, was 6.0 , this value being lower than the optimal pH determined for the proteolytic activities from Trichoderma virens, and for the extracellular proteases from Streptomyces spp. G157 and S. pactum. The T. harzianum 1051 enzyme was mostly active at 37C . This optimal temperature value is very close to the optimal temperatures determined for the proteases from Streptomyces spp. G157 , and S. pactum. The purified enzyme was only fairly stable at 37C and pH 8.0 (not shown). It lost about 50% of its original activity after pre-incubation for 15 min at this temperature and pH, followed by the incubation with the substrate for 20 min at 37C, as stated in materials and methods. Figure 3 | The effect of pH (A) and temperature (B) on the activity of the purified protease from Trichoderma harzianum 1051. The effect of pH (A) and temperature (B) on the activity of the purified protease from Trichoderma harzianum 1051. SDS-PAGE analysis revealed that several proteins were present in the culture supernatant of Trichoderma harzianum cultures grown on casein-containing medium . As casein has a molecular mass of about 23 kDa, the possibility that most of those proteins resulted from hydrolysis of the substrate present in the culture can be excluded, since most proteins have molecular mass values higher than 23 kDa. Western-blotting analysis using anti-protease antibodies raised in mice against the purified T. harzianum 1051 protease revealed that the antibodies recognized three proteins, indicating that the T. harzianum 1051 produces and secretes at least three proteases, one with a molecular mass of 18 kDa, and the other two with molecular masses of about 16 --17 kDa . The 18.8 kDa protein recognized by the antibody corresponds to the protein purified and analyzed as above corresponds to the purified protease as demonstrated in Fig. . Nevertheless, the possibility that the 17 --18 kDa proteases result from proteolysis of the 18 kDa protein can not be excluded, since both proteins were present in the culture medium already 24 h after the beginning of culture. However, there is still the possibility that these two smaller proteins were recognized unspecifically by the anti-protease antibody. Figure 4 | SDS-PAGE (A) and Western blot (B) analysis of the proteins present in the culture supernatant of Trichoderma harzianum SDS-PAGE (A) and Western blot (B) analysis of the proteins present in the culture supernatant of Trichoderma harzianum 1051 grown in liquid medium containing 0.5% casein, for different growing times: 6 h, 24 h, 48 h, and, 72 h. Mw, molecular mass standards. Samples of culture medium containing about 100 mug of proteins were diluted in sample buffer and applied in the gel. Proteins in the SDS-PAGE were silver stained. The nitrocellulose membrane was developed with the polyclonal antibodies raised in mice against the Trichoderma harzianum 1051 purified protease, and alkaline phosphatase-conjugated anti-mouse IgG and the color developing reagents nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl. The arrows indicate position of the 16, 17, and, 18.8 kDa proteins. The N-terminal amino acid sequence of the T. harzianum 1051 18.8 kDa protease (VPSTVWTVCP10) showed no homology with any other protease. It did, however, show homology of about 54% --75% with the amino acid sequences of proteins produced by Ectatomma tuberculatum, Ascaris lumbricoides, Escherichia coli, and Rhizobium meliloti. The effect of the proteolytic activity-exhibiting T. harzianum 1051 culture supernatant and of the corresponding purified protease (Fig. and Fig. ) on the C. perniciosa cell wall was tested by measuring the amounts of reducing sugar and protein produced during incubation of the enzyme samples with the phytopathogen mycelium (Tab. ). As much as 0.202 mug of protein/muL of reaction mixture, as measured by the method of Lowry et al. were present in the reaction mixture 48 h after the beginning of the period of incubation of the purified native enzyme with the Crinipellis mycelium (Tab. ). At the same time, no protein was detected in the reaction mixture containing heat-denatured enzyme. The purified native and the denatured enzyme produced nor reducing sugar neither glucose. However, these products were present in the reaction mixture containing crude enzyme extract and the phytopathogen mycelium. This is probably due to the presence of other enzymes besides the protease. The action of the protease on the C. perniciosa mycelium was further confirmed by SEM analyses which revealed that the purified enzyme could hydrolyze the cell wall of the phytopathogenic fungus C. perniciosa as soon as 48 h after contact . As expected, no hydrolytic effect was observed for the denatured enzyme on the phytopathogen mycelium . This finding supports previous observation that the T. harzianum isolate 1051 controls witches' broom disease in the field (J. L. Bezerra, personal comm.) and is able to impair the growth of C. perniciosa when both fungi are allowed to grow on the same plate containing solid medium (dual culture). The most common event involved in the antagonistic process is perforation of the phytopathogen cell wall by lytic enzymes produced by the antagonist . The effects of purified Trichoderma chitinase on the cell wall of Rhizoctonia solani and Sclerotium rolfsii and on C. perniciosa mycelium were reported before supporting the involvement of hydrolytic enzymes in the antagonistic process. As a consequence, it has been claimed that hydrolytic enzymes and the genes encoding them may be useful for production of transgenic microorganisms with superior biocontrol capability and the development of transgenic plants with high resistance to pathogens . Further studies on molecular characterization of the protease produced by witches' broom disease-controlling T. harzianum, and studies about its possible participation in the antagonistic process are being carried out. Figure 5 | Scanning electron microscopy of the Crinipellis perniciosa mycelium. Scanning electron microscopy of the Crinipellis perniciosa mycelium treated for 48 h with native (A), and heat-denatured (B) Trichoderma harzianum 1051 purified protease. Table 2 | Amounts of reducing sugar, glucose and protein in the reaction medium containing Crinipellis perniciosa mycelium and crude enzyme extract, native and heat-denatured Trichoderma harzianum 1051 purified protease (75 U), or Trichoderma harzianum 1051 culture medium containing proteolytic activity (220 U). Hydrolysis was conducted for 48 hours. For assay of reducing sugar, glucose and protein see Materials and Methods. Conclusions : The Trichoderma harzianum isolate 1051 which was previously shown to control cocoa plant witches' broom disease in the field, impairs growth of the disease agent, Crinipellis perniciosa, in solid medium. One protein with molecular mass of 18.8 kDa may account for the bulk of the proteolytic activity secreted by the Trichoderma during its growth in culture medium containing casein. Mouse polyclonal antibody produced against the 18.8 kDa protease recognizes two other proteins with molecular masses of 16 --17 kDa. The 18.8 kDa purified protease was able to hydrolyze the phytophatogen C. perniciosa cell wall in vitro, indicating that this enzyme may play a role in the antagonistic interaction between the Trichoderma and the phytopathogen. Backmatter: PMID- 11825347 TI - Resource use data by patient report or hospital records: Do they agree? AB - Abstract | Background | Economic evaluations alongside clinical trials are becoming increasingly common. Cost data are often collected through the use of postal questionnaires; however, the accuracy of this method is uncertain. We compared postal questionnaires with hospital records for collecting data on physiotherapy service use. Methods | As part of a randomised trial of orthopaedic medicine compared with orthopaedic surgery we collected physiotherapy use data on a group of patients from retrospective postal questionnaires and from hospital records. Results | 315 patients were referred for physiotherapy. Hospital data on attendances was available for 30% (n = 96), compared with 48% (n = 150) of patients completing questionnaire data (95% Cl for difference = 10% to 24%); 19% (n = 59) had data available from both sources. The two methods produced an intraclass correlation coefficient of 0.54 (95% Cl 0.31 to 0.70). However, the two methods produced significantly different estimates of resource use with patient self report recalling a mean of 1.3 extra visits (95% Cl 0.4 to 2.2) compared with hospital records. Conclusions | Using questionnaires in this study produced data on a greater number of patients compared with examination of hospital records. However, the two data sources did differ in the quantity of physiotherapy used and this should be taken into account in any analysis. Keywords: Background : Concurrent economic evaluation alongside clinical trials is an increasingly used method for undertaking health related cost-effectiveness studies. The collection of resource use data is a prerequisite for these evaluations. There are essentially three ways to collect resource use data. Data can be collected from clinicians, patients and routine medical records. A drawback of collecting data from clinicians, through clinical proforma or case report, is that they cannot be blinded to patient participation in the study, increasing the chance of introducing a biased assessment of outcome. Disadvantages of collecting data from medical records include: access to the relevant medical records may be difficult particularly if treatment is spread over a number of different providers; the high cost in terms of research time to access such records; the accuracy of some data collection systems may be questionable; finally, some resource use data such as time off work, or over the counter medicines, may not be available from medical records. The alternative of asking the patient for their resource use is attractive, as relevant questions can be readily included within patient self completed questionnaires. Consequently such a method is easier and less costly form of data collection compared with medical records. The disadvantages to using patient completed questionnaires, both prospective and retrospective, concern their accuracy. The accuracy of self completed questionnaires is affected by recall error, item completion rates and questionnaire response rates. Hence, patients may report inaccurate levels of resource use, they may not complete the relevant resource use questions consistently if at all and finally, some patients will not return their questionnaires. There has been some previously published work looking at the accuracy of patient reported resource use data, but the results have been somewhat contradictory. Good agreement has been demonstrated in reports of specialised diagnostic procedures , hospital admission and specialist consultation . These same studies showed over reporting of clinic attendance and blood testing . Similarly, another study showed that agreement between patient report and medical records was higher for procedures that generated a test report than for those that were documented using a physician note . However, studies have also reported major discrepancies between medical records and patient recall , with others reporting that patient self-report tended to systematically underestimate resource use compared with medical records . A previous study of physiotherapy attendance showed fair agreement between patient interviews and insurance registers . As part of an economic evaluation conducted alongside a randomised controlled trial in the field of orthopaedics, we identified a priori, physiotherapy as an area of resource use likely to differ between the study arms. This trial, conducted with local research ethics committee approval between December 1993 and December 1994 and reported elsewhere, compared care from an orthopaedic medicine specialist (OM) with care from conventional orthopaedic surgeon led services (OS) for the management of non-surgical orthopaedic patients. The largest proportion of physiotherapy referrals was expected to be carried out at the orthopaedic unit in which the trial was taking place. We therefore planned to use the hospital physiotherapy department's records to provide estimates of resource use with patient reports supplementing hospital records (for example, if a patient had received physiotherapy from a distant facility). However, once the study was underway we found that the hospital physiotherapy department's system of record keeping using physiotherapist diaries, meant that individual patient data were difficult and time consuming to access. We also found that more patients than expected were being referred to physiotherapists in the community, at health centres or to other hospitals. In total patients were referred to 31 different centres for their physiotherapy which meant much of this data was impossible to access due to the logistical and financial constraints of the study. We did, however, succeed in obtaining hospital data for a sample of patients, and this gave us the opportunity to compare the agreement of patient and hospital reported data for the estimation of total resource use. Therefore, the aim of this paper is to compare the results of retrospective patient self-reported physiotherapy use with routinely collected hospital record data. Methods : Patients included in the study were all over 18 years and had all been referred to their local orthopaedic hospital with a musculoskeletal condition, which was classified by existing hospital procedures as unlikely to require surgery. The sample was approximately 50% male, had a mean age of 45 years and 38% (n = 315) were referred for physiotherapy. Patient self-report (PSR) | Physiotherapy resource use questions were included in a postal follow-up questionnaire mailed to patients 3 and 12 months after their initial outpatient appointment. Patients were asked firstly if they had had any NHS physiotherapy for their condition over the past three months; if the answer was 'yes', they were then asked to give the date and location of hospital or clinic, for each treatment session they attended. Outpatient department records (ODR) | Referral for NHS physiotherapy was recorded for all patients in their outpatient department notes, together with the location of the referral. This gave definitive data on physiotherapy referrals, but no details of actual attendances. Physiotherapy department records (PDR) | A systematic search of the physiotherapy department records of the hospital, in which the study was taking place, enabled us to retrieve a sample of referral histories. In addition, a small number of external physiotherapy clinics routinely reported back to the hospital with the number and dates of patients' treatment sessions. The sample of patients for the reliability exercise was selected if their physiotherapy data fulfilled the following four criteria: * PSR and ODR data indicated that physiotherapy had been prescribed; * The two data sources agreed about the location of the clinic supplying treatment; * Numbers of attendances were available from both PSR and PDR; * The time frame of the PDR data matched that of the PSR data (ie the first and fourth 3 month periods post-consultation). The rate of agreement was measured by an intraclass correlation coefficient (ICC) which assesses the conformity between two quantitative measurements, reporting the proportion of the variability due to variation among subjects. This was calculated from an analysis of variance (ANOVA) two-way random effects model . This was conducted on log transformed data, to correct for non-normality. In addition the calculation of limits of agreement and a paired t-test were also used . Results : Figure summarises the responses of the two methods of data collection. From the 315 patients referred for physiotherapy the main source of loss for the PDR data was the inability to trace patients referred to outlying clinics. For the PSR data the actual questionnaire response rate compares favourably with similar studies , but 26% of the patients referred for physiotherapy did not report attendance. However, 17 (21%) of these subjects did have attendance data available from PDR. PSR data on physiotherapy attendance was available for 17% more patients than PDR data (95% Cl 10% to 24%). Figure 1 | Physiotherapy attendance data collection Physiotherapy attendance data collection There were 59 (19%) patients with data available from both sources; four of these cases only had PDR data for the 6 month period not covered by the follow-up questionnaire; 3 differed in the reported location of the physiotherapy clinic; leaving 17% (n = 52) of patients fulfilling the four criteria for inclusion in the reliability exercise. The ICC of 0.54 (95% Cl 0.31 to 0.70) indicates reasonable agreement between PDR and PSR, as would be expected with two methods of measuring the same quantity. However in this situation the ICC is a relative measure and should be assessed with respect to the other methods of assessing agreement. Forty eight percent (n = 25) of the study group report fewer visits with PDR compared with PSR; with a mean of 1.3 (95% Cl 0.4 to 2.2) fewer visits reported in physiotherapy department notes (Table). Figure also illustrates this point with the majority of the points falling below zero, indicating a lower number of visits for PDR. In addition the limits of agreement, denoting the range in which 95% of the differences should lie, stretches from eight fewer visits for PDR to five more visits. This would indicate that the level of agreement is inadequate. Although the distribution of the differences is approximately Normally distributed, there may be a tendency for the difference to increase with the mean of the two scores. A log transformation was carried out, but this had no effect on the relationship. Figure 2 | Scatterplot of the difference in reported visits by mean reported visits for the two methods Scatterplot of the difference in reported visits by mean reported visits for the two methods Table | Number of physiotherapy visits reported The table also shows that the difference in reporting observed between the two methods may not be the same for both trial groups, but the small numbers in the OS group make this difficult to say for certain. Discussion : This study has shown that there are differences in reported use of physiotherapy between patient completed questionnaire and hospital records. In contrast, with studies reporting underestimation of resource use , our study suggests that self reported data gives a higher estimate of resource use compared with hospital records. These findings agree with other studies looking at resource use for routine health service contacts . Assuming that most of those referred for physiotherapy receive it, the study suggests patient self-report is more accurate than physiotherapy department notes in estimating the number of patients attending for physiotherapy treatment, although even this gives a lower estimate than we would expect considering the actual number of referrals made. In addition, we have also shown that differences in estimates of the number of physiotherapy sessions attended can occur when patient self-reported data and hospital records are compared. In this instance, even for patients for whom hospital and self-report agreed that physiotherapy treatment had taken place we cannot be sure that PDR data represents the more reliable measure as there is some doubt as to whether all the physiotherapy data were either recorded at all or recorded in such a form that it was easily accessible to the researcher. This might explain the discrepancy between the two methods. It is also possible that recall problems led patients to report additional visits outside the reference period of three months. However, even if the hospital estimates of the number of physiotherapy visits per patient were more accurate, the use of patient reported data is not likely to bias the results of our evaluation unless there were systematic differences between the trial groups in terms of accuracy. In the table there is some evidence that this could be happening, with patients allocated to OM reporting on average one extra visit per patient when compared with the OS group. It would be prudent, if necessary, to undertake a sensitivity analysis using both estimates to ensure that the two data sources do not lead to different estimates of physiotherapy costs. Whilst the differences between the two methods of data collection are statistically significant, they may not be of economic significance. For example, if in a sensitivity analysis it were shown that the results of the cost effectiveness analysis were not affected by substantial changes in physiotherapy use as was the case in this study , then a difference of approximately 1 visit per patient would be of no significance. However, if physiotherapy were to prove to be one of the dominant costs in the evaluation then the addition or subtraction of one extra visit per patient could easily be of economic significance. The use of sensitivity analyses in these situations would also improve the generalisability of the findings. Where levels of nonresponse are high these may be combined with multiple imputation methods to assess the results with respect to both uncertainty round resource use and their costs. Conclusions : This study has shown that there can be a disparity between hospital records and patient self-reported data with neither source necessarily being the more accurate. We would recommend that in similar situations, the decision on which data source to use be based upon a pilot study comparing self-report and hospital data against a rigorously collected 'gold standard'. We would also recommend that where data from both sources are available and there are material differences in resource use, their impact be assessed using sensitivity analysis. Competing interests : None declared. Backmatter: PMID- 11876827 TI - Probing substrate binding to Metallo-beta-Lactamase L1 from Stenotrophomonas maltophilia by using site-directed mutagenesis AB - Abstract | Background | The metallo-beta-lactamases are Zn(II)-containing enzymes that hydrolyze the beta-lactam bond in penicillins, cephalosporins, and carbapenems and are involved in bacterial antibiotic resistance. There are at least 20 distinct organisms that produce a metallo-beta-lactamase, and these enzymes have been extensively studied using X-ray crystallographic, computational, kinetic, and inhibition studies; however, much is still unknown about how substrates bind and the catalytic mechanism. In an effort to probe substrate binding to metallo-beta-lactamase L1 from Stenotrophomonas maltophilia, nine site-directed mutants of L1 were prepared and characterized using metal analyses, CD spectroscopy, and pre-steady state and steady state kinetics. Results | Site-directed mutations were generated of amino acids previously predicted to be important in substrate binding. Steady-state kinetic studies using the mutant enzymes and 9 different substrates demonstrated varying Km and kcat values for the different enzymes and substrates and that no direct correlation between Km and the effect of the mutation on substrate binding could be drawn. Stopped-flow fluorescence studies using nitrocefin as the substrate showed that only the S224D and Y228A mutants exhibited weaker nitrocefin binding. Conclusions | The data presented herein indicate that Ser224, Ile164, Phe158, Tyr228, and Asn233 are not essential for tight binding of substrate to metallo-beta-lactamase L1. The results in this work also show that Km values are not reliable for showing substrate binding, and there is no correlation between substrate binding and the amount of reaction intermediate formed during the reaction. This work represents the first experimental testing of one of the computational models of the metallo-beta-lactamases. Keywords: Background : The overuse of antibiotics in the clinic and for agricultural uses has resulted in a tremendous selective pressure for antibiotic resistant bacteria. These bacteria become resistant by a number of mechanisms, such as producing enzymes that hydrolyze or inactivate the antibiotics, producing efflux pumps that transport the antibiotic out of the cell, or modifying their cell wall components so they no longer bind effectively to the antibiotics . The most common, least expensive, and effective antibiotics are the beta-lactam containing antibiotics, such as the penicillins, cephalosporins, and carbapenems . These antibiotics are mechanism-based inhibitors of transpeptidase, a bacterial enzyme required for the production of a strong viable cell wall . In response to the widespread use of beta-lactam containing antibiotics, bacteria have acquired the ability to produce beta-lactamases, which are enzymes that hydrolyze and inactivate beta-lactam containing antibiotics. There are over 300 distinct beta-lactamases known, and these enzymes have been grouped by a number of classification schemes . For example, Bush has developed a scheme, based on the enzymes' molecular properties, that has four distinct beta-lactamase groups . One of the more alarming groups are the Bush group 3 enzymes, which are Zn(II) dependent enzymes that hydrolyze nearly all known beta-lactam containing antibiotics and for which there are no or very few known clinical inhibitors [,,-]. The metallo-beta-lactamases have been further divided by Bush into subgroups based on amino acid sequence identity: the Ba enzymes share a >23% sequence identity, require 2 Zn(II) ions for full activity, prefer penicillins and cephalosporins as substrates, and are represented by metallo-beta-lactamase CcrA from Bacteroides fragilis, the Bb enzymes share a 11% sequence identity with the Ba enzymes, require only 1 Zn(II) ion for full activity, prefer carbapenems as substrates, and are represented by the metallo-beta-lactamase imiS from Aeromonas sobria, and the Bc enzymes have only 9 conserved residues with the other metallo-beta-lactamases, require 2 Zn(II) ions for activity, contain a different metal binding motif than the other metallo-beta-lactamases, prefer penicillins as substrates, and are represented by the metallo-beta-lactamase L1 from Stenotrophomonas maltophilia. A similar grouping scheme (B1, B2, and B3) based on structural properties of the metallo-beta-lactamases has recently been offered . The diversity of the group 3 beta-lactamases is best exemplified by the enzymes' vastly differing efficacies towards non-clinical inhibitors; these differences predict that one inhibitor may not inhibit all metallo-beta-lactamases [,-]. To combat this problem, we are characterizing a metallo-beta-lactamase from each of the subgroups in an effort to identify a common structural or mechanistic aspect of the enzymes that can be targeted for the generation of an inhibitor. It is hoped that this inhibitor, when given in combination with an existing antibiotic, will prove to be an effective therapy against bacteria that produce a metallo-beta-lactamase. This work describes our efforts on metallo-beta-lactamase L1 from S. maltophilia. S. maltophilia is an important pathogen in nosocomial infections of immunocompromised patients suffering from cancer, cystic fibrosis, drug addition, AIDS and in patients with organ transplants and on dialysis . This organism is inherently resistant to most antibiotics due to its low outer membrane permeability and to beta-lactam containing antibiotics due to the production of a chromosomally expressed group 2e beta-lactamase (L2) and a group 3c beta-lactamase (L1) . L1 has been cloned, over-expressed, and partially characterized by kinetic and crystallographic studies . The enzyme exists as a homotetramer of ca. 118 kDa in solution and in the crystalline state. The enzyme tightly binds two Zn(II) ions per subunit and requires both Zn(II) ions for full catalytic activity. The Zn1 site has 3 histidine residues and 1 bridging hydroxide as ligands, and the Zn2 site has 2 histidines, 1 aspartic acid, 1 terminally-bound water, and the bridging hydroxide as ligands. Spencer and coworkers used the crystal structure and modeling studies to propose a substrate binding model and identified several active site residues that were involved in substrate binding . However, this model has not been tested experimentally. In order to prepare tight binding inhibitors of the metallo-beta-lactamases, knowledge about how substrate binds to the enzymes is needed so that all substrate-enzyme binding contacts can be maintained in any proposed inhibitor. This work describes our efforts at understanding how substrates bind to metallo-beta-lactamase L1. Several site-directed mutants of L1 were generated and characterized, and the results from these studies reveal that none of the active site residues predicted from earlier computational studies are essential for tight substrate binding. Figure 1 | Active site residues that were mutated in this study. Active site residues that were mutated in this study. Figure was rendered using Rasmol v. 2.6. The coordinates were obtained from the Protein Data Bank using the accession number 1 sml. Results : Wild type L1 | Wild-type L1 was over-expressed in Escherichia coli and purified as previously described . This procedure produced an average of 50 --60 mg of >90% pure, active protein per 4 L of growth culture. Circular dichroism spectra were collected on wild-type samples to ensure L1 expressed using the pET26b expression system had the correct secondary structure. The CD spectrum of wild type L1 showed an intense, broad feature at 190 nm and a smaller feature at 215 nm (see Additional file 1: CD spectra). These features are consistent with a sample with significant alpha/beta content. The Compton and Johnson algorithm was used to estimate secondary structure in the samples; wild-type L1 was estimated to have 38.3% alpha-helix, 26.7% beta-structure (9.3% antiparallel beta-sheet, 2.1% parallel beta-sheet, and 15.3% beta-turn), and 34.9% other structure. These estimates are in excellent agreement with the crystallographically determined secondary structure of ~40% alpha-helix and 30% beta-structure . Metal analyses on multiple preparations of wild-type L1 demonstrated that the enzyme binds 1.9 +- 0.2 Zn(II) ions per monomer , in agreement with previous results . Table 2 | Metal Content of Wild-type L1 and L1 mutants Steady state kinetic studies were performed on multiple preparations of wild type L1, and the resulting kinetic data are shown in Tables ,,. When using nitrocefin as substrate and 50 mM cacodylate, pH 7.0, as buffer, wild-type L1 exhibited a kcat value of 38 +- 1 s-1 and a Km value of 12 +- 1 muM. The inclusion of 100 muM ZnCl2 in the assay buffer resulted in slightly lower values of Km and higher values for kcat. The inclusion of higher concentrations of Zn(II) did not further affect the steady-state kinetic constants. Apparently, the purified, recombinant enzyme does not bind its full complement of Zn(II); therefore, 100 muM Zn(II) was included in all subsequent kinetic studies. Table 3 | Steady-state kinetic constants for hydrolysis of cephalosporins by wild-type L1 and L1 mutants. Table 4 | Steady-state kinetic constants for hydrolysis of penicillins by wild-type L1 and L1 mutants. Table 5 | Steady-state kinetic constants for hydrolysis of carbapenems by wild-type L1 and L1 mutants Wild-type L1 exhibited kcat values of 41 +- 1 s-1, 1.9 +- 0.1 s-1, 42 +- 1 s-1, and 82 +- 5 s-1 for the cephalosporins, nitrocefin, cefoxitin, cefaclor, and cephalothin. For these same substrates, the Km values were 4 +- 1 muM, 1.1 +- 0.1 muM, 13 +- 1 muM, and 8.9 +- 1.5 muM, respectively. Two penicillins were tested as substrates, and penicillin G and ampicillin exhibited Km values of 38 +- 12 muM and 55 +- 5 muM and kcat values of 600 +- 100 s-1 and 520 +- 10 s-1, respectively . Three carbapenems were also used as substrates for L1, and biapenem, imipenem, and meropenem exhibited Km values of 32 +- 1 muM, 57 +- 7 muM, and 15 +- 4 muM and kcat values of 134 +- 4 s-1, 370 +- 5 s-1, and 157 +- 9 s-1, respectively . L1's preference for penicillins and carbapenems over cephalosporins, as exemplified by the kcat values, is in agreement with previous studies and supports L1's placement in the beta-lactamase 3c family . Rapid-scanning visible spectra of 25 muM wild-type L1 with 5 muM nitrocefin demonstrated a decrease in absorbance at 390 nm, an increase at 485 nm, and a rapid increase and slower decrease in absorbance at 665 nm. These spectra are similar to those previously reported for wild-type L1 and nitrocefin , and the features can be attributed to substrate decay, product formation, and intermediate formation and decay, respectively. Under these conditions, 2.2 muM intermediate was formed during the first 10 milliseconds of the reaction , and the rate of decay of this intermediate corresponds to the steady-state kcat . To probe further the binding of nitrocefin to wild-type L1, stopped-flow fluorescence studies were conducted as previously described . The reaction of wild-type L1 with nitrocefin under steady-state conditions at 10C resulted in a rapid decrease in fluorescence followed by a rate-limiting return of fluorescence . Fitting of the data, as described by Spencer et al., resulted in a KS value for wild-type L1 of 38 +- 5 muM . Figure 2 | Intermediate formation by wild-type L1 and L1 mutants. Intermediate formation by wild-type L1 and L1 mutants. The spectra were collected using rapid scanning Vis studies, and the absorbance values at 668 nm were converted to concentration values as described in Materials and Methods. Typical reactions were conducted with 25 muM L1 (or mutant) and 5 muM nitrocefin in 50 mM cacodylate, pH 7.0, containing 100 muM ZnCl2 at 25C. Figure 3 | (A) Stopped-flow fluorescence of wild-type L1 with nitrocefin. (A) Stopped-flow fluorescence of wild-type L1 with nitrocefin. In a typical reaction, 5 --10 muM wild-type L1 was mixed with various concentrations of nitrocefin, and the reaction was monitored for up to 1 second at 10C, using the conditions described in Materials and Methods. The data were fitted to a double exponential using SigmaPlot v. 6.10.(B) Plot of observed rate constant versus concentration of nitrocefin. Solid lines were fitted to the data as described in Materials and Methods. Ser224 mutants | (the BBL numbering scheme proposed in reference 41 was used throughout this manuscript). All sequenced subclass Ba and Bb metallo-beta-lactamases (except VIM-1) have a lysine residue at position 224 , and all computational models for substrate binding to the metallo-beta-lactamases assume that the invariant carboxylate on substrates forms an electrostatic interaction with this lysine. In L1, the residue at position 224 is a serine , and the substrate-binding model for L1 predicts that this serine residue interacts with the carboxylate on substrate via a water molecule . To test the proposed role of Ser224 in L1, serine was changed to an alanine (S224A), aspartic acid (S224D), and lysine (S224K), and these mutants were characterized using metal analyses, CD spectroscopy, steady-state kinetics, and pre-steady state kinetic studies. Small-scale growth cultures showed that all three mutants were over-expressed at levels comparable to those of wild-type L1. Large-scale over-expression and purification of the mutants showed that all three mutants were isolatable at levels comparable to those of wild-type L1. Metal analyses of the S224A and S224D mutants showed that both mutants bind nearly two Zn(II) ions , like wild-type L1 ; however, the S224K mutant binds only 1.0 Zn(II) per protein. CD spectra of the mutants were similar to those of wild-type L1 (see Figure in Additional materials). Steady-state kinetic studies were conducted with all three mutants in buffer containing 100 muM ZnCl2 to ensure that both Zn(II) binding sites were saturated in these studies. Addition of higher concentrations of Zn(II) did not result in different values for the steady-state kinetic constants in Tables ,,. When the cephalosporins were used as substrates, the S224A and S224K mutants exhibited 2- to 4-fold changes in Km values . In studies with cefoxitin, cefaclor, and cephalothin as substrate, the observed kcat values for the S224A and S224K mutants were 2- to 7-fold lower; however, the kcat values when using nitrocefin as substrate were slightly higher (< 2-fold). On the other hand, the S224D mutant exhibited 3- to 50-fold higher Km values and 2- to 20-fold lower kcat values for the cephalosporins tested. A similar trend was observed in kinetic studies when using penicillins as substrates . Generally, the S224A and S224K mutants exhibited small changes in Km and kcat, while the S224D mutant yielded 20- to 40-fold increased values for Km and >10-fold decreases in kcat when using the penicillins as substrates. When the carbapenems were used as substrates however, the changes in Km values were relatively smaller than with the other substrates, and 2- to 37-fold changes in kcat were observed . Rapid-scanning Vis studies of the S224X mutants were conducted to probe whether the mutations caused changes in the amount of intermediate that accumulates during catalysis. When 50 muM S224A was reacted with 5 muM nitrocefin, 1.7 muM intermediate formed during the first 10 milliseconds of the reaction , and rate of decay of this intermediate was equal to the steady-state kcat . In spite of utilizing a number of reaction conditions, the S224K and S224D mutants yielded rapid-scan spectra with no detectable absorbances at 665 nm , indicating that the intermediate is not stabilized as well in these mutants as in wild-type L1. Stopped-flow fluorescence studies at 10C with the S224A, S224D, and S224K mutants and nitrocefin as the substrate resulted in KS values of 39 +- 10, 213 +- 63, and 33 +- 5 muM, respectively. Asn233 mutants | Two-thirds of all sequenced metallo-beta-lactamases have an Asn at position 233 , and this residue was predicted and shown to be involved with substrate binding and activation by interacting electrostatically with the substrate beta-lactam carbonyl. However, in L1, Asn233 is 14 A away from the modeled position of the substrate beta-lactam carbonyl . To test the role of Asn233 in substrate binding, the Asn was changed to a leucine (N233L) and to an aspartic acid (N233D), and these mutants were characterized by using metal analyses, CD spectroscopy, steady-state kinetics, and pre-steady state kinetic studies. Small-scale growth cultures showed that both mutants were over-expressed at levels comparable to that of wild-type L1. Large-scale over-expression and purification of the mutants showed that both mutants were isolatable at levels comparable to that of wild-type L1. Metal analyses of the N233L and N233D mutants showed that both bind nearly two Zn(II) ions , like wild-type L1 . CD spectra of the mutants were similar to those of wild-type L1. Steady-state kinetic studies were conducted with both mutants in buffer containing 100 muM ZnCl2 to ensure that both Zn(II) binding sites were saturated in these studies. Addition of higher concentrations of Zn(II) did not result in different values for the steady-state kinetic constants in Tables ,,. With all substrates tested, the N233L and N233D mutants exhibited Km values that differed less than a factor of 4 than that observed for wild-type L1 (Tables ,,). The kcat values exhibited by these mutants for all substrates also differed by less than a factor of 4, except when biapenem and meropenem were used as substrates for the N233D mutant. With these two substrates, there was a 19-fold and 45-fold decrease in the kcat values when using biapenem and meropenem, respectively . The steady-state kinetic data generally support the prediction that Asn233 does not play a large role in binding or catalysis. However, rapid-scanning Vis studies of N233L and N233D with nitrocefin demonstrate that no detectable amounts of intermediate are formed during the reaction, even when using a wide number of reaction conditions . Stopped-flow fluorescence studies at 10C with the N233L and N233D mutants and nitrocefin as substrate resulted in KS values of 26 +- 9 and 25 +- 8 muM, respectively. Tyr228 mutants | The substrate-binding model showed that Tyr228 in L1 was position-conserved with Asn233 in the other crystallographically characterized metallo-beta-lactamases [,,-]. Spencer and coworkers postulated that Tyr228 is part of an oxyanion hole that interacts with the beta-lactam carbonyl on substrate and helps to stabilize the putative tetrahedral intermediate formed during substrate turnover . To test this hypothesis, Tyr228 was changed to an alanine and to a phenylalanine to afford the Y228A and Y228F mutants, respectively. Small-scale growth cultures showed that both mutants were over-expressed at levels comparable to those of wild-type L1. Large-scale over-expression and purification of the Y228A and Y228F mutants showed that both mutants were isolatable at levels comparable to those of wild-type L1. Metal analyses of the mutants showed that both bind nearly two Zn(II) ions , like wild-type L1 , and CD spectra of the mutants were similar to those of wild-type L1. Steady-state kinetic studies were conducted with both mutants in buffer containing 100 muM ZnCl2 to ensure that both Zn(II) binding sites were saturated in these studies. Addition of higher concentrations of Zn(II) did not result in different values for the steady-state kinetic constants in Tables ,,. When cephalosporins were used as substrates, the Y228A and Y228F mutants exhibited Km values that were 6- to 45-fold higher than those observed for wild-type L1 . The largest change in Km was observed when cefaclor was used as substrate, and the smallest change was observed when nitrocefin was used as substrate. The Tyr228 mutants exhibited < 4-fold change in kcat values for the cephalosporins tested , suggesting that Tyr228 is not playing a large role in catalysis. When penicillins were used as substrates, the Tyr228 mutants exhibited 3- to 13-fold increased Km values and < 2-fold changes in kcat, as compared to the values ascertained using wild-type L1 . On the other hand when carbapenems were used as substrates, the Tyr228 mutants exhibited < 6-fold increases in Km values as compared to those values for wild-type L1 . Interestingly, there was a 2- to 8-fold drop in kcat values for the Tyr228 mutants, as compared to values observed for wild-type L1, when using the carbapenems as substrates. Rapid-scanning Vis spectra of the reaction of the Y228A and Y228F mutants with nitrocefin demonstrated a marked decrease in the amount of intermediate formed with these mutants . In reactions with 50 muM mutant and 5 muM nitrocefin, only 0.75 and < 0.30 muM intermediate formed for the Y228F and Y228A mutants, respectively. The concentration of mutants were varied between 25 to 150 muM to ensure that all of the substrate was bound; however, none of the reactions resulted in the detection of intermediate at levels observed for wild-type L1 (data not shown). Stopped-flow fluorescence studies at 10C of nitrocefin hydrolysis by Y191A and Y191F resulted in KS values of 85 +- 9 and 22 +- 6 muM, respectively. Ile164 and Phe158 mutants | All crystallographically characterized metallo-beta-lactamases have a flexible amino acid chain that extends over the active site [,,-]. Previous NMR studies on CcrA have shown that this loop "clamps down" on substrate or inhibitor upon binding, and there is speculation that the distortion of substrate upon clamping down of the loop may drive catalysis . The crystal structure of L1 showed that there is a large loop that extends over the active site, and modeling studies have predicted that two residues, Ile164 and Phe158, make significant contacts with large, hydrophobic substituents at the 2' or 6' positions on penicillins, cephalosporins, or carbapenems . To test this prediction, Ile 164 and Phe158 were changed from large, hydrophobic residues to alanines to afford the I164A and F158A mutants. Small-scale growth cultures demonstrated the I164A and F158A mutants were over-expressed at levels comparable to that of wild-type L1 (data not shown). Large-scale over-expression and purification of the mutants resulted in comparable quantities of isolatable enzymes, which had identical CD spectra as wild-type L1 and bound slightly less Zn(II) than wild-type L1 . All steady-state kinetic studies were conducted in buffers containing 100 muM ZnCl2 to ensure that both metal binding sites were saturated during the studies. When using the cephalosporins, nitrocefin, cefoxitin, and cephalothin, as substrates and the I164A mutant, there were 2- to 10-fold (only for cefoxitin) increases in Km and 2- to 4-fold increases in kcat observed . However when cefaclor was used as substrate, the I164A mutant exhibited a 3-fold decrease in Km and a 1.5-fold decrease in kcat . On the other hand, the Km and kcat values for the I164A mutant when the penicillins or carbapenems were used as substrates were very similar to those numbers exhibited by wild-type L1 (Tables and ). When the cephalosporins were used as substrates for the F158A mutant, the Km values observed were 7- to 31-fold higher than those determined for wild-type L1, and surprisingly, the kcat values were 2- to 31-fold higher than those exhibited by wild-type L1 . As with the I164A mutant, the changes in Km and kcat for the penicillins and carbapenems were relatively small, as compared with the values obtained with the cephalosporins . Rapid-scanning Vis studies on nitrocefin hydrolysis by I164A and F158A showed a marked decrease in intermediate accumulation, with the I164A mutant generating < 0.30 muM intermediate and the F158A producing no detectable intermediate . Stopped-flow fluorescence studies at 10C resulted in a KS value of 31 +- 11 muM for the I164A mutant. The reaction of F158A with nitrocefin was so rapid, we could not determine a KS value for this mutant. Discussion : beta-Lactam containing antibiotics constitute the largest class of antibiotics, and these compounds are relatively inexpensive to produce, cause minor side effects, and are effective towards a number of bacterial strains. Nonetheless, bacterial resistance to these antibiotics is extensive, most commonly due to the bacterial production of beta-lactamases . In fact, there have been over 300 distinct beta-lactamases reported, and most of these enzymes utilize an active site serine group to nucleophilically attack the beta-lactam carbonyl, resulting in a hydrolyzed product that is covalently attached to the active site. To combat these enzymes, beta-lactamase inhibitors such as clavulanic acid, sulbactam, and tazobactam have been given in combination with a beta-lactam containing antibiotic to treat bacterial infections . One class of beta-lactamases that are particularly unaffected by the known beta-lactamase inhibitors and have been shown to hydrolyze almost all known beta-lactam containing antibiotics including late generation carbapenems at high rates are the metallo-beta-lactamases . Although there are no reports of metallo-beta-lactamases isolated from major pathogens , these enzymes are produced by pathogens such as B. fragilis, S. maltophilia, and P. aeruginosa. It is inevitable that the continued and extensive use of beta-lactam antibiotics will result in a major pathogen that produces a metallo-beta-lactamase. Efforts to solve the crystal structure of one of the metallo-beta-lactamases with a bound substrate molecule have failed, most likely due to the high activity of the enzymes towards all beta-lactam containing antibiotics . Therefore, computational studies have been used extensively to study substrate binding, the role of the Zn(II) ions in catalysis, the protonation state of the active site, and inhibitor binding [,,-]. All of the substrate binding models have made assumptions before the substrate was docked into the active site , and some of these assumptions have been shown to be invalid for certain substrates . With L1, two key assumptions were made: (1) the bridging hydroxide functions as the nucleophile during catalysis and (2) Zn1 coordinates the beta-lactam carbonyl . With these assumptions and after energy minimizations, Ser224 was predicted to hydrogen bond to the substrate carboxylate , reminiscent of the role predicted for Lys224 in CcrA . Ullah et al. predicted that Phe158 and Ile 164 form hydrophobic interactions with bulky substituents on the substrate, suggesting that the loss of these residues would only affect binding of substrates with large aromatic substituents . In the modeling studies on CcrA , Asn233 was predicted to interact with the beta-lactam carbonyl on substrate, and mutagenesis studies have supported this prediction . Although Asn233 is sequence conserved in L1, it is located 14A away from the modeled position of the beta-lactam carbonyl and was predicted not to play a role in substrate binding to L1. On the other hand, the substrate-binding model predicted that Tyr228 was in position to offer a hydrogen bond to the beta-lactam carbonyl and participate in an oxyanion hole that was proposed to form as the substrate was hydrolyzed . By using the crystal structure and modeling studies on L1, Ullah et al. proposed a reaction mechanism for the enzyme . To test this proposed mechanism and the proposed roles of the amino acids discussed above, site-directed mutagenesis studies were conducted on metallo-beta-lactamase L1 and reported herein. The overlap extension method was used to prepare the site-directed mutants, and a variety of studies were used to probe whether the single point mutations resulted in large structural changes in the mutant enzymes. (1) The over-expression levels of mutants were analyzed with SDS-PAGE to ensure that the mutations did not result in changes in the over-expression levels of the enzymes. With a few L1 mutants and with other enzyme systems in the lab, single point mutations often result in depressed levels of over-expression . In the case of the mutants described here, all of the mutants over-expressed at levels comparable to wild-type L1 (data not shown). (2) The total amounts of the mutants isolatable after chromatography were compared with wild-type L1 levels. We have found, in particular with metal binding mutants of L1 (G. Periyannan, R.B. Yates, and M.W. Crowder, unpublished results) and glyoxalase II , that single point mutations can result in over-expressed mutants being processed into inclusion bodies and unisolatable as soluble proteins. In the case of the mutants described here, all of the mutants were isolated at levels comparable to wild-type L1. (3) CD spectra were collected for all mutants and compared to the spectrum of wild-type L1. Although we did not expect a large change in the secondary structure of L1 upon single point mutations, CD spectroscopy is the most common structural technique to characterize site-directed mutants. All of the mutants described here exhibited CD spectra that were very similar, or identical, to that of wild-type L1 (see Additional file 1: CD spectra). (4) Metal analyses on the mutants were used to probe whether point mutations caused a significant change to the metal binding site as to preclude metal binding. The crystal structures of the metallo-beta-lactamases reveal a complex and far-reaching hydrogen-bonding network around the metal binding sites, and disruption of this network is predicted to affect metal binding . With all of the mutants described here except the S224K mutant, each mutant binds wild-type or near-wild-type levels of Zn(II) after purification. The S224K mutant exhibited a 50% reduction in metal binding , and we postulate this is due to electrostatic repulsions between the newly introduced Lys with Zn2. In spite of the mutants binding significant amounts of Zn(II), we included 100 muM ZnCl2 in all of the kinetic buffers to ensure saturation of the metal binding sites and to facilitate direct comparison of the kinetic data. (5) All mutants were stable to multiple freeze/thaw cycles and to prolonged storage (> 3 weeks) at 4C, retaining > 95% of their activity. With these five lines of evidence, we were confident that none of the point mutations resulted in large structural changes in L1 and that any kinetic differences could be attributed to the changed amino acid. As a first approximation of substrate binding, we examined the steady-state kinetics of 4 cephalosporins, 2 penicillins, and 3 carbapenems (Tables ,,) and compared the Km values of the mutants with those of wild-type L1. The substrates tested were chosen because they exhibited low Km values in previous kinetic studies , and we believed that we could saturate the enzymes with substrate even if there was large change in binding with the point mutations. The Tyr228 mutants exhibited increased Km values for 8 of the 9 substrates tested, with the smallest changes in Km observed when the carbapenems were used as the substrate. This result supports the proposed role of Tyr228 in substrate binding. In contrast, the results on the Ser224 mutants suggest that this residue is not important in substrate binding, since the S224A and S224K mutants did not exhibit any significant increases (by a factor of >= 10) in Km for any of the substrates tested. Only when Ser224 was replaced with an Asp residue was there significant increases in the observed Km value for 6 of the 9 substrates tested, and the largest changes were exhibited when the penicillins were used as substrates. This result supports the observation of differential binding modes of substrates to the beta-lactamases, depending on the structure of the substrate . The only remaining mutants that exhibited significant changes in the Km values were the I164A and F158A mutants. The I164A mutant exhibited increased Km values only when using cefoxitin as the substrate, suggesting an interaction of the isoleucine group with the methoxy group on cefoxitin. The F158A mutant exhibited higher Km values when using the cephalosporins as substrates, suggesting an interaction of the cephalosporins' substituents with the phenylalanine on the loop that extends over the active site. None of the other mutants exhibited vastly different values for Km with any of the substrates tested. An examination of the kcat values of the mutants revealed some surprising results. The S224D mutants displayed decreased kcat values for 7 of the 9 substrates tested. Since similar results were not observed with the S224K and S224A mutants, we do not propose a catalytic role for Ser224. Instead, we predict that the insertion of an aspartic acid into the active site at position 224 results in a change in the hydrogen bonding network in L1; this hydrogen bonding network is extensive in all metallo-beta-lactamases that have been characterized crystallographically . The N233D mutant also exhibited greatly reduced kcat values for biapenem and meropenem but not for imipenem or any of the other substrates tested. This mutation is also predicted to affect the hydrogen bonding network around the active site, and apparently, interactions of the enzyme with the 4' substituent of the carbapenems has an effect on catalysis. More surprisingly are the increases in kcat of the F158A mutants. We are uncertain why the mutation of residues on the loop that extends over the active site would affect kcat, since substrate and product binding have been predicted to be very fast in the reaction of nitrocefin with L1. However, we do note that the kcat/Km values of wild-type L1 and F158A differ by a factor less than 2. The inability to propose a consistent binding model also supports the recent proposal that different substrates of L1 are hydrolyzed by different mechanisms and further suggests that using steady-state kinetic constants may not be a valid way to probe substrate binding to L1. In addition, the minimal kinetic mechanism of nitrocefin hydrolysis by L1 has been reported, and this mechanism predicts that Km does not accurately reflect substrate binding. By using this mechanism , Km is equal to {(k-1 + k2)k3k4} / {k1(k3k4 + k2k4 + k2k3}. To probe more directly the reaction, stopped-flow absorbance studies were conducted, and the substrate decay rates (390 nm) were studied as a function of nitrocefin concentration. While nitrocefin is a nontypical substrate, as a result of the dinitro-substituted styryl substituent , it is the substrate about which the most is known about its hydrolysis mechanism. Therefore, kinetic studies with nitrocefin as substrate allowed for us to evaluate the effect of point mutations on the reaction mechanism of L1. There was no clear dependence on substrate decay rates with nitrocefin concentration (data not shown). We did note though that the amount of intermediate formed during the reactions varied considerably depending upon which mutant of L1 was used in the study. All of the mutants exhibited decreases in intermediate formation, and the S224D, S224K, F158A, N233D, and N233L mutants yielded rapid-scanning data consistent with no detectable intermediate. These same mutants exhibited vastly differing Km values. Clearly there is no correlation of Km with the presence of the reaction intermediate. Apparently, the ability to observe the intermediate is not governed entirely by the choice of substrate , and it also depends on precise arrangement of active site residues. It is also possible that the site-directed mutants could be utilizing a different mechanism to hydrolyze nitrocefin . Recently, Spencer and co-workers reported that stopped-flow fluorescence studies can be used to monitor the reaction of L1 with nitrocefin and that an initial binding step can be directly monitored . By increasing the concentration of nitrocefin, the rate of the initial binding step increased to a maximum, and fitting of these data yielded a binding constant (called KS herein) for nitrocefin. Each of the L1 mutants were studied using the stopped-flow fluorescence studies, and the resulting data were fitted as reported by Spencer et al. . All of the mutants exhibited KS values identical, within error, to wild-type L1, except the S224D and the Y228A mutants. The placement of a negative charge at position 224 drastically affects nitrocefin binding and results in a 6-fold decrease in binding affinity . To a lesser degree, the aromatic portion of Tyr228 must have an effect on the binding site as the KS value for nitrocefin binding to this mutant is decreased by a factor of 2; however, the hydroxyl group probably does not form a hydrogen bond to the substrate as proposed. By using nitrocefin as substrate and Km values alone, a completely different conclusion is reached regarding important substrate binding residues. The results presented here suggests that none of the residues in this study are essential for tight nitrocefin binding, possibly because other parts of the active site accommodate the loss of certain binding contacts. Table 6 | Km and KS values for Wild-type L1 and mutants with nitrocefin as substrate Spencer et al. also reported stopped-flow fluorescence studies when using cefaclor and meropenem as substrates, and KS values for these substrates were reported to be 710 +- 180 and 272 +- 112 muM, respectively . However in our hands, the rates of substrate hydrolysis were so fast when using wild-type L1 that we could not use substrate concentrations high enough to saturate the enzyme. Similarly, we could not determine KS values for penicillin G or ampicillin because the observed rates of hydrolysis at low substrate concentrations were too fast to observe data, even at 10C. Conclusions : The results presented herein indicate that none of the active site residues identified with computational studies are essential for tight substrate binding. These data also indicate that the use of Km values to describe substrate binding to L1 is unreliable and that there is no correlation between intermediate accumulation and substrate binding affinity. These results demonstrate that new computational studies are now needed to probe substrate binding to L1, and these studies are currently underway. The results presented herein can be used to guide these new computational studies, which will lead to the design of potential inhibitors and hopefully a way to combat penicillin resistance in bacteria. Materials and Methods : Materials | E. coli strains DH5alpha and BL21(DE3) pLysS were obtained from Gibco BRL and Novagen, respectively. Plasmids pET26b and pUC19 were purchased from Novagen. Primers for sequencing and mutagenesis studies were purchased from Integrated DNA Technologies. Deoxynucleotide triphosphates (dNTP's), MgSO4, thermopol buffer, Deep Vent DNA polymerase, and restrictions enzymes were purchased from Promega or New England Biolabs. Polymerase chain reaction was conducted using a Thermolyne Amplitron II unit. DNA was purified using the Qiagen QIAQuick gel extraction kit or Plasmid Purification kit with QIAGEN-tip 100 (Midi) columns. Wizard Plus Minipreps were acquired from Promega. Luria-Bertani (LB) media was made following published procedures . Isopropyl-beta-thiogalactoside (IPTG), Biotech grade, was procured from Anatrace. Phenylmethylsulfonylfluoride (PMSF) was purchased from Sigma. Protein solutions were concentrated with an Amicon ultrafiltration cell equipped with YM-10 DIAFLO membranes from Amicon, Inc. Dialysis tubing was prepared using Spectra/Por regenerated cellulose molecular porous membranes with a molecular weight cut-off of 6 --8,000 g/mol. Q-Sepharose Fast Flow was purchased from Amersham Pharmacia Biotech. Nitrocefin was purchased from Becton Dickinson, and solutions of nitrocefin were filtered through a Fisherbrand 0.45 micron syringe filter. Cefaclor, cefoxitin, and cephalothin were purchased from Sigma; penicillin G and ampicillin were purchased from Fisher. Imipenem, meropenem, and biapenem were generously supplied by Merck, Zeneca Pharmaceuticals, and Lederle (Japan), respectively. All buffers and media were prepared using Barnstead NANOpure ultrapure water. Generation of site-directed mutants of L1 | The over-expression plasmid for L1, pUB5832, was digested with NdeI and HindIII, and the resulting ca. 900 bp piece was gel purified and ligated using T4 ligase into pUC19, which was also digested with NdeI and HindIII, to yield the cloning plasmid pL1PUC19. Mutations were introduced into the L1 gene by using the overlap extension method of Ho et al., as described previously . The oligonucleotides used for the preparation of the mutants are shown in Table . The ca. 900 bp PCR products were digested with NdeI and HindIII and ligated into pUC19. The DNA sequences were analyzed by the Biosynthesis and Sequencing Facility in the Department of Biological Chemistry at Johns Hopkins University. After confirmation of the sequence, the mutated pL1PUC19 plasmid was digested with NdeI and HindIII, and the 900 bp, mutated L1 gene was gel purified and ligated into pET26b to create the mutant overexpression plasmids. To test for overexpression of the mutant enzymes, E. coli BL21(DE3)pLysS cells were transformed with the mutated over-expression plasmids, and small scale growth cultures were used . Large-scale (4 L) preparations of the L1 mutants were performed as described previously . Protein purity was ascertained by SDS-PAGE. Table 1 | Oligonucleotides used in preparation of L1 mutants Metal content | The concentrations of L1 and the mutants were determined by measuring the proteins' absorbance at 280 nm and using the published extinction coefficient of epsilon280 nm = 54,804 M-1*cm-1 or by using the method of Pace . Before metal analyses, the protein samples were dialyzed versus 3 x 1 L of metal-free, 50 mM HEPES, pH 7.5 over 96 hours at 4C. A Varian Inductively Coupled Plasma Spectrometer with atomic emission spectroscopy detection (ICP-AES) was used to determine metal content of multiple preparations of wild type L1 and L1 mutants. Calibration curves were based on three standards and had correlation coefficient limits of at least 0.9950. The final dialysis buffer was used as a blank, and the Zn(II) content in the final dialysis buffers was shown to be < 0.5 muM (detection limit of ICP) in separate ICP measurements. The emission line of 213.856 nm is the most intense for zinc and was used to determine the Zn content in the samples. The errors in metal content data reflect the standard deviation (sigman-1) of multiple enzyme preparations. Steady-state kinetic studies | Steady-state kinetic assays were conducted at 25C in 50 mM cacodylate buffer, pH 7.0, containing 100 muM ZnCl2 on a HP 5480A diode array UV-Vis spectrophotometer at 25C. The changes in molar absorptivities (Deltaepsilon) used to quantitate products were (in M-1cm-1): nitrocefin, Deltaepsilon485 = 17,420; cephalothin, Deltaepsilon265= -8,790; cefoxitin, Deltaepsilon265= -7,000; cefaclor, Deltaepsilon280= -6,410; imipenem, Deltaepsilon300= -9,000; meropenem, Deltaepsilon293= -7,600; biapenem, Deltaepsilon293= -8,630; ampicillin, Deltaepsilon235= -809; and penicillin G, Deltaepsilon235 = -936. When possible, substrate concentrations were varied between 0.1 to 10 times the Km value. In kinetic studies using substrates with low Km values (cefoxitin, nitrocefin, and cephalothin) or with small Deltaepsilon values (penicillin and ampicillin), we typically used substrate concentrations varied between ~ Km and 10 x Km and used as much of the DeltaA versus time data (that was linear) as possible to determine the velocity. Steady-state kinetics constants, Km and kcat, were determined by fitting initial velocity versus substrate concentration data directly to the Michaelis equation using CurveFit . The reported errors reflect fitting uncertainties. All steady-state kinetic studies were performed in triplicate with recombinant L1 from at least three different enzyme preparations. Circular dichroism | Circular dichroism samples were prepared by dialyzing the purified enzyme samples versus 3 x 2 L of 5 mM phosphate buffer, pH 7.0 over six hours. The samples were diluted with final dialysis buffer to ~75 mug/mL. A JASCO J-810 CD spectropolarimeter operating at 25C was used to collect CD spectra. Stopped -- flow/Rapid-Scanning UV-Visible Spectrophotometry | Rapid-scanning Vis spectra of nitrocefin hydrolysis by L1 and the L1 mutants were collected on a Applied Photophysics SX.18MV stopped-flow spectrophotometer equipped with an Applied Photophysics PD.l photodiode array detector and a 1 cm pathlength optical cell. A typical experiment consisted of 25 muM enzyme and 5 muM nitrocefin in 50 mM cacodylate buffer, pH 7.0 containing 100 muM ZnCl2, the reaction temperature was thermostated at 25C, and the spectra were collected between 300 and 725 nm. Data from at least three experiments were collected and averaged. Absorbance data were converted to concentration data as described previously by McMannus and Crowder . Stopped-flow fluorescence studies of nitrocefin hydrolysis by L1 were performed on an Applied Photophysics SX.18MV spectrophotometer, using an excitation wavelength of 295 nm and a WG320 nm cut-off filter on the photomultiplier. These experiments were conducted at 10C using the same buffer in the rapid-scanning Vis studies. Fluorescence data were fitted to kobs = {(kf [S]) / KS + [S])} + kr as described previously or to kobs = kf [S] + kr by using CurveFit v. 1.0. Abbreviations : AES, atomic emission spectroscopy; bp, base pairs; CD, circular dichroism; epsilon, extinction coefficient; ICP, inductively coupled plasma; kcat, turnover number; kDa, kilodaltons; Km, Michaelis constant; KS, substrate binding constant; LB, Luria-Bertani media. Backmatter: PMID- 10984461 TI - Opportunities at the Intersection of Bioinformatics and Health Informatics --A Case Study AB - Abstract This paper provides a "viewpoint discussion" based on a presentation made to the 2000 Symposium of the American College of Medical Informatics. It discusses potential opportunities for researchers in health informatics to become involved in the rapidly growing field of bioinformatics, using the activities of the Yale Center for Medical Informatics as a case study. One set of opportunities occurs where bioinformatics research itself intersects with the clinical world. Examples include the correlations between individual genetic variation with clinical risk factors, disease presentation, and differential response to treatment; and the implications of including genetic test results in the patient record, which raises clinical decision support issues as well as legal and ethical issues. A second set of opportunities occurs where bioinformatics research can benefit from the technologic expertise and approaches that informaticians have used extensively in the clinical arena. Examples include database organization and knowledge representation, data mining, and modeling and simulation. Microarray technology is discussed as a specific potential area for collaboration. Related questions concern how best to establish collaborations with bioscientists so that the interests and needs of both sets of researchers can be met in a synergistic fashion, and the most appropriate home for bioinformatics in an academic medical center. Keywords: Introduction : This paper provides a "viewpoint discussion" based on a presentation made to the American College of Medical Informatics (ACMI) Symposium, which was held in February 2000 at San Marco Island, Florida. The activities of one day of that three-day symposium centered on the intersection of bioinformatics and health informatics. The implicit theme was to help identify opportunities for researchers in health informatics to become involved in the rapidly growing field of bioinformatics. In general, there are two types of such opportunities. One set of opportunities occurs where bioinformatics research itself intersects with the clinical world. The second set of opportunities occurs where bioinformatics research can benefit from the technologic expertise, techniques, and approaches that health informaticians have used extensively in the clinical arena. Several related questions were also addressed. One question concerned how best to establish collaborations with bioscientists so that the interests and needs of both sets of researchers can be met in a synergistic fashion. A second question concerned the appropriate home for bioinformatics in an academic medical center. Since these questions are very broad and complex, they can be difficult to discuss in abstract terms. this paper uses the experience of the Yale Center for Medical Informatics (YCMI) as a case study to structure a discussion of these issues. What is Bioinformatics? : The term "bioinformatics" has been used with different scopes and meanings by different groups of researchers. The term could refer to a range of activities: In this paper, "bioinformatics" is loosely defined to include the first two areas discussed above ---informatics involving the biosciences, including genomics. The Spectrum of Biomedical Informatics at the Yale Center : outlines the spectrum of biomedical informatics activities at the YCMI. The three areas of research are increasingly likely to intersect in the near future. Genomic Informatics | Over the past decade, the YCMI has been involved in a number of projects in support of genomics and genetics. An early project explored the use of parallel computation in biological sequence analysis and genetic linkage analysis, in collaboration with faculty in the Department of Computer Science. Another project provided Internet-based informatics support for the collaborative Genome Center, involving the Albert Einstein College of Medicine and Yale, to map human chromosome 12. Current YCMI activities in genomic informatics focus on three areas. One long-standing collaboration, with the laboratory of Prof. Kenneth Kidd (Genetics), centers on human genome diversity. A second major collaboration is with Prof. Michael Snyder, director of Yale's Yeast Genome Analysis Center. A recent and rapidly growing set of collaborations involve the support of microarray technology, as discussed further below. All these collaborations have involved the development of databases and informatics tools for internal use at Yale and also for providing public access to the data via the Web. With regard to integrating genomic informatics with clinical practice, now that many genes and gene fragments have been identified, there are tremendous opportunities to work up from the genetic blueprint to explore gene expression, functional genomics, and structural genomics on a massive scale, including their implications for human disease. Clinical Informatics | Clinical informatics and genomic informatics are at far ends of the spectrum shown in . Here the field of medical and health informatics has long been confronting the development of informatics techniques that deal with the fuzziness of clinical data and disease. At the YCMI, we have been working on many different projects: Thus, we anticipate increasing interplay between the genomic and clinical levels, which will have a major impact on our informatics activities. Neuroinformatics | In , neuroinformatics sits between genomic informatics and clinical informatics and is really a place-holder for informatics activities involving many different tissues and organ systems. Once the genetic blueprint is known, the next step is to determine what it means in a range of different tissues and organ systems, including the kidney, the liver, the gastrointestinal tract, the heart, and the endocrine system, among others. Neuroinformatics focuses on the central nervous system, which is clearly the most complex organ system. The goal of the field of neuroinformatics is to provide enabling informatics technology that supports neuroscience research at many different levels, . These include the genetic level, the cellular level, the physiologic and pharmacologic levels, and, eventually, the level of behavioral research. The research includes experimental microanatomic studies (e.g., imaging cells) and macroanatomic studies (e.g., imaging brains). Each type of experiment uses different experimental techniques and generates different types of experimental data. Historically, different laboratories have tended to focus on one or two of these levels. At each of these levels, large amounts of experimental data are being generated in a form that can be stored and analyzed online. To fully understand a neuroscience phenomenon, however, it is ultimately important to gather data at many different levels and analyze all those data in an integrated fashion. In summary, neuroinformatics in a sense "connects" genomic informatics and clinical informatics in the areas of neuroscience research. As such, it is representative of many other bioscience disciplines that are addressing similar issues in other tissues and organ systems. The three levels shown in represent a spectrum of informatics activities that will become increasingly integrated in many different ways. Table 1 | The Spectrum of Biomedical Informatics at the Yale Center for Medical Informatics Table 2 | Examples of Experimental Neuroscience Data at Different Levels of Brain Function Bioinformatics and Health Informatics: Selected Areas of Intersection : Health informaticians have many potential opportunities to become involved in collaborations involving bioinformatics. One set of opportunities occurs where bioinformatics research intersects the clinical world: A second set of opportunities occurs where bioinformatics research can benefit from techniques and approaches that informaticians have used extensively in the clinical arena: Storing and Analyzing Microarray Data: A Case Study : The microarray, a recently developed technology, offers a wide range of informatics opportunities.,,, Yale is currently installing two microarray analyzers, one in the School of Medicine and one on Yale's main campus. These use a technology whereby "DNA chips" measure whether and to what degree different genes are expressed in experimental tissue samples. Each DNA chip can analyze the presence or absence and the approximate level of expression of tens of thousands of genes. For example, one group of Yale researchers will use this technology to study hematologic disease involving white blood cells (WBCs). They will take WBCs at different stages of cell differentiation and in a single experiment see which of roughly 10,000 genes are expressed in two samples that have been combined (e.g., a normal sample and a cancerous sample at the same stage of differentiation). The test for each gene (really a small DNA fragment that is part of a gene) is seen on the microarray as a single spot in a massive array of spots. The two samples (from normal and abnormal WBCs) will be tagged with fluorescent markers of different colors (red and green), so that each gene can be tested in both samples in a single microarray experiment. Each experiment will generate 10,000 data points. Each point will have several associated values, reflecting the actual intensity and relative intensity of both fluorescent markers at each of the 10,000 spots. The researchers estimate they will ultimately perform up to five such experiments a day. In addition, other laboratories will be using the same machine to generate similar numbers of data for many different experiments. It takes roughly 10 minutes to perform each microarray analysis. Slides can be loaded to allow the machine to perform analyses automatically 24 hours a day. As a result, we see microarray experiments as an exciting opportunity to expand our activities in genomic informatics: As a result, a broad, stable infrastructure of informatics staff and faculty will be required to support the high volume of microarray experiments that will soon be performed. Establishing Collaboration with Bioscientists: Informatics Support vs. Informatics Research : If health informaticians are to become centrally involved in bioinformatics, they need to establish robust collaborations with bioscientists. In the forging of such collaborations, it is important to understand that informaticians can play two general types of informatics roles ---specifically, providing informatics support for bioscience research and performing informatics research that uses the bioscience domain as a context for addressing basic informatics research issues. In this regard, it is important to point out that bioscientists typically have motivations that are very different from those of the clinician collaborators with whom many health informaticians have worked in the past. Clinician collaborators are, typically, primarily involved in clinical practice and are looking for additional interesting research projects in which they can become involved. For such collaborators, embarking on a clinical informatics research project allows them to provide their clinical expertise and participate in sophisticated research that relates to their field. If this research results in additional visibility and publications, they have reason to be happy. Bioscientist collaborators may have a very different motivation. They are already doing research. They are looking to informaticians to provide tools to help them perform their research more effectively. They are, typically, not looking for additional areas for research peripheral to their field, nor do they want to devote their time to such projects. They have more than enough research in their own field to keep them busy, and their time is precious. As a result, they will not be satisfied with the clinical informatics model of serving as domain experts in informatics research projects, even if the projects are in their field. They want help solving their immediate research problems. As a result, bioscientists typically want informaticians to provide them with informatics support. Conversely, academic informaticians want at some level to be performing informatics research, although they are certainly willing to provide informatics support if this leads to interesting research opportunities. The YCMI's neuroinformatics experience in the national Human Brain Project (HBP) provides, as a case study, a chance to discuss these issues more concretely. The YCMI's HBP work involves the integration of multidisciplinary sensory data, using the olfactory system as a model system. This HBP work involves both neuroinformatics support and neuroinformatics research. Neuroinformatics Support | Our neuroinformatics support activities involve building a variety of databases and tools. In general, we have attempted to build databases that can serve the needs of our collaborating laboratories and also serve as pilot resources for the field as a whole. These databases include ORDB, containing information about olfactory receptors; OdorDB, containing information about odor molecules; NeuronDB, containing information about neuronal cell properties; and ModelDB, containing neuronal models that can be searched, examined, downloaded via the Web, and run locally. The development of these databases involved a great deal of practical work with our collaborators to finetune their design, functionality, and interface so that they can be readily useful to, and usable by, neuroscientists. It is important to emphasize, however, that "just" performing good neuroinformatics support requires that informatics faculty work closely with the neuroscientists to understand the biological problems, to appreciate the needs of the neuroscience researchers, and to develop well-structured solutions to enable neuroscience research. Neuroinformatics Research | In developing these tools, we have been able to define interesting neuroinformatics research projects. As discussed in more detail below, however, this did not happen immediately. Our current neuroinformatics research includes: The Problem and the Solution | In a neuroscience collaboration, one would like to strike a balance between neuroinformatics support and neuroinformatics research, so that both can be pursued in a synergistic fashion. The problem that we encountered in our HBP work was essentially that of "the chicken and the egg." At the start of our HBP activities, in particular, there was no critical mass of neuroinformatics support activities to provide a context for neuroinformatics research. In addition, our neuroinformatics support activities were applied and pragmatic, reflecting the real-world needs of our neuroscience collaborators. It was therefore difficult (inretrospect, impossible) to perform neuroinformatics research that was directly tied to our collaborators' immediate research needs. The problem was that we had not achieved a robust level (a critical mass) of neuroinformatics support activities to provide a context for neuroinformatics research. Once we had achieved a sufficiently robust level of neuroinformatics support (which took about five years), we could then embark on neuroinformatics research that was built on our support and therefore directly tied to our collaborators' research needs. This, in turn, meant that the results of our neuroinformatics research could be folded back to enhance our neuroinformatics support, in a fully synergistic fashion. For example, as described above, we were able to integrate the operational versions of all four of our HBP databases into our EAV/CR data model. We believe that this provides a strong pilot proof of concept for the EAV/CR model and also provides a robust, flexible database environment for the further development of these and future HBP databases at Yale. Finding an Academic Home for Bioinformatics : An important question concerns the most appropriate home for bioinformatics in an academic medical center. One possible academic home is in a bioscience department. To the extent that a particular computational technique is unique to a department, then that department may well be a logical home for researchers who focus on that technique. This would be particularly true if such faculty members need to be fully trained in that department's discipline. To the extent that bioinformatics faculty members require broad training in informatics issues and have skills that are applicable across many bioscience fields, however, there is logic to basing those faculty members in a broader academic unit containing colleagues who share this informatics background. Two general types of such a unit are: A unit of the later type would promote ---among all faculty, staff, and students ---work at the intersection of clinical and bioscience informatics as well as a broader appreciation of biomedical informatics as a whole. As the current trends in bioinformatics continue, the latter model is likely to become an increasingly logical solution. It is clear, however, that many historical, political, and practical considerations will influence how any individual academic medical center approaches this issue. Backmatter: PMID- 10984462 TI - The Interactions Between Clinical Informatics and Bioinformatics --A Case Study AB - Abstract For the past decade, Stanford Medical Informatics has combined clinical informatics and bioinformatics research and training in an explicit way. The interest in applying informatics techniques to both clinical problems and problems in basic science can be traced to the Dendral project in the 1960s. Having bioinformatics and clinical informatics in the same academic unit is still somewhat unusual and can lead to clashes of clinical and basic science cultures. Nevertheless, the benefits of this organization have recently become clear, as the landscape of academic medicine in the next decades has begun to emerge. The author provides examples of technology transfer between clinical informatics and bioinformatics that illustrate how they complement each other. Keywords: Introduction : The Stanford Medical Informatics (SMI) laboratory was created in the mid-1980s as part of a reorganized Stanford Knowledge Systems Laboratory (KSL). The KSL was a federation of investigators interested in applying artificial intelligence methods to real-world problems in engineering, science, and medicine. The core competency of the SMI centered on the research interests of Dr. Edward Shortliffe, who had well-established research programs in the medical applications of expert systems (such as the MYCIN system) and the creation of an electronic infrastructure to support such systems. The initial focus of the laboratory was clearly on the clinical applications of artificial intelligence. The KSL itself was an outgrowth of the Stanford Heuristic Programming Project, which was formed around the DENDRAL program for interpreting mass spectroscopy data using a production rule formalism. The excitement about production rule systems led to the application and refinement of these ideas into rule-based expert systems and to MYCIN, a system for diagnosing infectious diseases and recommending treatment. At the same time, other projects arose at the Heuristic Programming Project that focused more on the support of basic science. The MOLGEN I project studied planning techniques for molecular biological experiments,, the MOLGEN II project modeled the process of scientific discovery in molecular biology,, and the PROTEAN project looked at interpreting NMR experimental information about three-dimensional structures using constraint satisfaction techniques. These projects did not reside explicitly in the SMI laboratory at the time of its formation, but they were staffed by close collaborators and students in the KSL, so the opportunities in basic science were always clear. Most important, there was an early recognition that the basic methodological approaches taken in the two disciplines could often be transferred to produce novel contributions. By the late 1980s, it became clear that there was an impending explosion of data that would become available in molecular biology and that the SMI was in a good position to apply, to basic biology, the research strategies it had been using in clinical medicine. The National Library of Medicine encouraged Stanford (and other institutions at which it funded training programs) to seriously consider expanding the training mission to include the application of computational technologies to basic biology. Starting in 1992, therefore, faculty in what eventually came to be called "bioinformatics" were hired at SMI, and students with interests in this area were recruited. The philosophy behind the addition of bioinformatics to the SMI research and training mix was not that a separate track should be introduced but that the existing training and research experience in the SMI and KSL could be adapted in an evolutionary manner. In the case of training, the five elements of the medical informatics training program (core informatics, domain biology, computer science, probability/statistics/decision science, and ethical/legal/social issues) were very well suited to training in both bioinformatics and clinical informatics, so the training program expanded quite naturally. Some projects focused on the medical domain, others on the biological domain. The problems in the two domains are clearly different, but the methodologies that are used to approach them are shared and form the basis of a cohesive program. Thus, the SMI houses both bioinformatics and clinical informatics efforts at Stanford. The new laboratory logo combines the traditional medical caduceus with a DNA double helix to stress the interactions between the two fields. There are also active programs in bioinformatics research in the Departments of Biochemistry, Structural Biology, Pathology, Genetics, Mathematics, and Computer Science, and work in clinical informatics is being done in the Departments of Medicine, Computer Science, Anesthesiology, and Pathology. The focus on methodology, and the management of the informatics training program (which sends students to many of the other departments to work on projects), are what make the SMI a focal point for efforts in these areas. The pursuit of bioinformatics and clinical informatics together is not without some difficulties. Practitioners in clinical medicine and basic science do not instantly understand the distinction between the scientific goals of their domains and the transferability of methodologies across the two domains. They sometimes question whether informatics investigators are really devoted to the solution of scientific problems or are simply enamored of computational methodologies of unclear significance. It is therefore imperative that informatics investigators (and their students) be able to work collaboratively with physicians and scientists in a manner that makes it clear that the creation of excellent, well-validated methods for solving problems in these domains is the paramount goal. Usually, the particular skills of informatics students and the quality of their contributions to the research efforts they join significantly allay these concerns. Students in informatics have a much richer understanding of computer science, statistics, and information technology than do students in collaborating disciplines, and they think more like engineers in creating solutions. Figure 1 | The Stanford Medical Informatics logo stresses the combination of clinical medicine, basic biology, and computer science. The Stanford Medical Informatics logo stresses the combination of clinical medicine, basic biology, and computer science. Variations are found at . Technology Transfer Between Subdisciplines : One of the most compelling reasons for housing clinical informatics and bioinformatics under the same roof is the opportunities this provides for rapid technology transfer between the two subdisciplines. Because there are still barriers between the two fields (bioinformatics professionals may not even know the clinical informaticians at their institutions), there are also opportunities for accelerating progress in both fields by knowing the problems and literature in each. Even in the days of the Heuristic Programming Project, ideas from the DENDRAL project on mass spectroscopy were applied to the MYCIN project on diagnosing disease. Success with MYCIN inspired some to pursue planning and discovery efforts in the MOLGEN framework. Two particularly illustrative areas of technology transfer that have had significant impact are the development of Bayesian methods for reasoning and the development of frameworks for knowledge representation and acquisition. Probabilistic Methods in Clinical Informatics Applied to Biological Structure | In the mid 1980s, the SMI experienced a "probabilistic revolution," during which many projects in medical diagnosis began to focus on the use of probability theory as an alternative to other uncertainty calculi that had been proposed. (As an aside, the interest in these methods stemmed, in large part, from student exposure to courses in probabilistic decision making, which created a cadre of converts to this mode of thinking.) In particular, a great interest developed in the use of Bayesian belief networks as an organizing paradigm for understanding diagnosis (compute the most likely diagnosis), treatment (compute the highest utility), knowledge acquisition (acquire the best numbers), and other elements of medical reasoning. The work that resulted made an impact on medical informatics and launched a number of successful careers in the application of these technologies in medicine and beyond.,,, The furor over probabilistic methods in clinical informatics spilled over in many ways to the work that was going on in bioinformatics. First, the protean project (which focused on computing three-dimensional molecular structures from sparse and noisy data) moved to adopt a probabilistic representation of structure in which an ill-defined "prior" structure was updated with uncertain data to compute the most likely a posteriori structural estimate. The impetus for this change in representation (from an initial formulation as a discrete constraint satisfaction or combinatorial optimization problem) can be traced directly to the success of probabilistic, Bayesian methods in the clinical side of the laboratory., Second, in a more direct application of the clinical informatics experience, Bayesian methods were applied to problems in sequence analysis. In particular, dependencies between positions in biological sequences were modeled as dependencies in a Bayesian net, and good performance was demonstrated in the ability of such models to characterize and recognize patterns in biological sequence. Again, abandoning the traditional assumption of the independence of columns in a multiple alignment could be traced directly to the work in clinical diagnosis. Essentially, this work was diagnosing an alpha helix based on probabilistic evidence. Knowledge Representation and Structured Knowledge Acquisition | A second area in which methodologies from clinical informatics have inspired efforts and contributions in bioinformatics is that of knowledge representation. The experiments in supporting structured representations of clinical medicine ---in particular, the representation of clinical treatment protocols (in the Oncocin project) ---led the clinical informatics researchers at the SMI to focus attention on the creation of general-purpose frame representation systems that allowed strict ontological modeling of domains, and used the resulting ontologies to automatically create templates for the structured acquisition of knowledge. The PROTEGE project continues to be a source of software for structured data modeling and knowledge acquisition. PROTEGE has recently been adopted by the World Wide Web Consortium as a recommended platform for authoring RDF documents, and meetings of PROTEGE user groups (including health care organizations) interested in robust methods for data modeling are held annually. Around 1995, an SMI bioinformatics effort to create three-dimensional structural models from combinations of experimental data (as reported in the literature) was focusing on the structure of the bacterial ribosome, a critical cellular molecular ensemble., The algorithms used to create the three-dimensional models worked well, but it was a struggle to collect all the relevant data from the literature. Aware of the work on ontological modeling on the clinical side, we proposed the creation of an ontology of biological objects as well as the experiments that supply the data used in modeling. As a test bed, we proposed to gather the published literature about the ribosome (having identified approximately 200 papers that contain reports of primary data) and create a set of structured representations for the data. The framebased system that was chosen for RiboWEB and the organization of the system were closely related to those of the PROTEGE system. In addition, a long history of creation and maintenance of controlled terminologies in clinical medicine helped inform the creation of a controlled vocabulary for describing structural biological experiments. The RiboWEB system now serves as a test bed for the use of structured representations of biological data. RiboWEB makes it possible to compare a new piece of data with all the relevant data reported previously, it allows inconsistencies in the literature to be identified, and it allows three-dimensional models to be built on the basis of a selection and interpretation of a subset of the available information. Once again, there was a technology transfer of informatics techniques from the clinical to the biological domain. Blurring the Boundary Between Clinical Informatics and Bioinformatics | The success of the RiboWEB project was responsible, in part, for the newest efforts in the SMI to create a pharmacogenetics knowledge base (PharmGKB, available at ) that combines a diverse array of data, from genomic data to cellular/molecular phenotype and clinical phenotype. The data modeling and knowledge base infrastructure developed at SMI form the basis for this project, which aims to provide scientists nationwide with integrated access to data over the Web. Data will be acquired in structured form and will be distributed using the same kinds of structured representations that were developed for RiboWEB. The PharmGKB project is the first project at SMI that blurs the distinction between bioinformatics and clinical informatics. The knowledge base will hold genomic sequences, multiple alignments, and structural information, which is the stuff of bioinformatics. At the same time, it will store clinical patient records of diseases, medications, side effects, and laboratory results, which are the traditional material of clinical informatics. Similarly, the analytic capabilities will be based on algorithms introduced in both fields. Clearly, for projects such as this, there may not be a useful distinction between clinical informatics and bioinformatics, and in many ways the strategy of combining the training and research environments for these two subjects is vindicated. As the biological world enters the post-genome age, the interplay between basic biological data (sequences, structures, pathways, and genetic networks) and clinical information systems is, clearly, critical. Genes and structures are useful only in the context of the functions and phenotypes that they produce, and so we look toward continuing interaction and, perhaps, unification of the two fields. One of the ultimate goals of both bioinformatics and clinical informatics is to have robust computational models of physiology that will enable us to model, store, retrieve, and analyze the effects, on patients, of disease, medications, and the environment. Bioinformatics approaches these models from the bottom up, while clinical informatics approaches them from the top down. The other critical technologies of clinical informatics, including knowledge representation, data mining, automated diagnosis, and information retrieval, can be viewed as technologies supporting this goal. The training program curriculum at SMI has recently been updated and generalized to be appropriate for both clinical informatics and bioinformatics. This process was surprisingly simple. We now require 1) instead of medical physiology, any biology or physiology of interest to the student; 2) as before, significant coursework in probability, statistics, or decision analysis, or a combination of these; 3) a substantial amount of core computer science; and 4) nontechnical courses in the ethical, social, legal, or business aspects of the field. We have also generalized our introductory core informatics courses to introduce the principles of data representation and algorithms in biomedicine that make it a challenging field. The resulting generalized curriculum has been tested for both bioinformatics and clinical informatics students and appears to strike an appropriate balance between domain knowledge and methodologic knowledge. Recognizing the success of this generalized curriculum, we have changed the name of the training program from Medical Information Sciences (the name of the degree program as it was defined in 1982) to Biomedical Informatics. The Need for Informatics to Co-exist with Other Areas of Biocomputation : As we look to the future of informatics, it becomes clear that informatics is one aspect of a larger area of endeavor, which can be loosely called "biomedical computation." At a recent Stanford faculty retreat, we found that most biomedical computation efforts fall into one of six affinity groups. (The summary abstract book is available in the archives, at .) These groups are divided roughly into those whose investigators think mostly (but not exclusively) about computing directly with and about physical systems and those whose investigators think mostly about information acquisition, storage, retrieval, and management. Both use significant computation skills and require a strong understanding of computer science. The affinity groups are: There was a broad consensus at this retreat that an umbrella organization for these affinity groups makes sense, to support joint teaching, research, and shared infrastructure and to provide a focus for the application of advanced computational techniques to problems in biomedicine. The creation of such an organization is a work in progress that promises to solidify and extend Stanford's commitment to these areas. Backmatter: PMID- 10984463 TI - Electronic Health Record Meets Digital Library --A New Environment for Achieving an Old Goal AB - Abstract Linking the electronic health record to the digital library is a Web-era reformulation of the long-standing informatics goal of seamless integration of automated clinical data and relevant knowledge-based information to support informed decisions. The spread of the Internet, the development of the World Wide Web, and converging format standards for electronic health data and digital publications make effective linking increasingly feasible. Some existing systems link electronic health data and knowledge-based information in limited settings or limited ways. Yet many challenging informatics research problems remain to be solved before flexible and seamless linking becomes a reality and before systems become capable of delivering the specific piece of information needed at the time and place a decision must be made. Connecting the electronic health record to the digital library also requires positive resolution of important policy issues, including health data privacy, government envouragement of high-speed communications, electronic intellectual property rights, and standards for health data and for digital libraries. Both the research problems and the policy issues should be important priorities for the field of medical informatics. Keywords: Introduction : The idea of linking automated clinical data and knowledge-based information to support health care, research, and education gained prominence in the early 1980s when the vision of Integrated Advanced Information Management Systems (IAIMS) was first articulated. The National Library of Medicine's Unified Medical Language System (UMLS) project was initiated in 1986 to facilitate the development of IAIMS systems capable of linking and integrating different types of machine-readable biomedical information, including patient records, the biomedical literature, factual databanks, and expert systems. As expected by those involved with IAIMS, UMLS, or both, the amount of useful patient data, clinical information, and biomedical knowledge in electronic form has increased dramatically since the 1980s. Also as anticipated, computers have continued to decrease in cost and size while increasing in speed, functionality, and ease of use. Not so generally predicted, but even more critical for the integration of the electronic health record and the digital library, the spread of the Internet and the development of the World Wide Web have simultaneously: These developments make integrated access to computer-based health records and related knowledge-based information increasingly feasible. The rise of Web technology has also contributed to expansive definitions of both the electronic health record and the digital library. In fact, these two concepts now meet at the edges. Electronic Health Record Defined : In a 1998 concept paper, the National Committee on Vital and Health Statistics (NCVHS) built on previous work by the American Medical informatics Association to describe three types of computer-based health records. In the NCVHS view, patient, personal, and population health records are needed to facilitate co-ordination, research, and assessment for clinical care and public health and to permit individuals to participate more actively in their own health care. Patient records "record clinical care and are used by delivery systems in which doctors, nurses, and other health professionals provide an array of hospital, primary care, and other ambulatory and institutional health services." Electronic patient records are increasingly using multimedia formats, and they are being aggregated into clinical data repositories and warehouses by many medical centers.,, Personal or consumer-oriented health records are "for individual use, including assessment of health status and linkage with physicians' records." They may also include knowledge-based information, such as health education and disease management advice, that is specifically relevant to the person, or links to such information, as are currently available in several Web-based systems. Population health records are derived "from the health care system and have been made as non-identifiable as possible for public health and research applications. They may also incorporate survey data." Population health records are used for monitoring public health and the outcomes of care and also for health services research, including quality management. In other words, the population health record envisioned by the NCVHS is a database derived from both health care and health survey data. The population health data sets now accessible via the Web generally do not combine the two types of data, and their interactive search capabilities are limited, in large part by the need to preserve confidentiality., Digital Library Defined : The term "digital library" may have originated with the National Science Foundation in 1994, when the first opportunity for federal digital library grants was announced. The concept is not sharply defined, and librarians and computer scientists tend to emphasize different aspects of it. A good working definition was proposed by Clifford Lynch in 1995, that is, an electronic information access system that offers the user a coherent view of an organized, selected, and managed body of information. The digital library focuses on information accessible via the Internet and encompasses the following materials: A digital library is not a single entity. It requires the technology to link different resources and assumes that the linkages between individual digital collections and information services will be transparent to the end user. Most existing systems have yet to achieve this transparency, but there are robust examples of usefully integrated digital collections, including some in biomedicine. The linked information resources and services of NLM's National Center for Biotechnology Information represent one of the best current instantiations of the digital library concept at a national level. Given an expansive definition of the digital library, collections of data aggregated from individual health records, like the clinical data warehouse or the population health data set, can be viewed as part of the larger digital library needed to support biomedical research, education, and informed health care decisions. This is a useful view for institutions that develop and provide access to digital health data and to knowledge-based information. Regarding aggregated health data as part of the digital library not only opens up new funding opportunities but may also encourage fruitful multidisciplinary cooperation on problems common to knowledge-based information and aggregated health data, including permanent retention of electronic information or the need to implement variable user access privileges. Requirements for Connecting the Electronic Health Record and the Digital Library : Although aggregations of health data can be profitably viewed as components of a digital health library, connecting individual electronic health records to other electronic information in the digital library remains a highly desirable goal. Such connections are useful both to find the universe of "patients like this one" and to facilitate access to scientific evidence, accepted treatment practices, and other decision support tools relevant to particular patients. Early in the UMLS project, three requirements for effective linking between disparate sources of machine-readable health information were identified ---technical connections, organizational connections, and conceptual connections. Today, it is clear that a fourth requirement for linking the electronic health record and the digital library is public policy that permits the technical, organizational, and conceptual connections to be made. One policy issue that casts a long shadow over the development of electronic health records and their effective connection to the digital library is health data privacy. In the United States, a lack of consensus about the legitimate uses of health data, and about the appropriate balance between these uses and a patient's right to confidentiality, has thus far stymied efforts to pass federal health data privacy legislation. The absence of federal legislation increases public unease about the electronic storage and transmission of identifiable health information. The failure to enact U.S. health privacy legislation by Aug 21, 1999, triggered the Health Insurance Portability and Accountability Act's requirement that the Department of Health and Human Services issue privacy regulations that will have the force of law. Proposed regulations were issued in November 1999 and have elicited about 40,000 comments from interested parties, reflecting the degree of controversy surrounding public policy in this area. These comments must be analyzed and addressed before final federal privacy regulations can be issued. In addition to privacy, policy issues such as government support for telecommunications, intellectual property rights, and standards also affect our ability to forge effective technical, organizational, and conceptual connections between the electronic health record and the digital library. These policy issues are discussed in the context of the particular type of connections to which they apply. Technical Connections | The technical connections needed to link the electronic health record and the digital library include the computing equipment, telecommunications, platform-independent software, logon procedures, and access controls needed to ensure access to different information sources. These technical matters already have been greatly simplified by the spread of the Internet and the Web. It is safe to assume that general technical advances, not driven specifically by health or digital library applications, will continue to make it easier to connect the electronic health record and the digital library. Technical aspects of format standards for all types of information (including health records and electronic publications) will continue to converge and to increase in flexibility and functionality, as the current move to XML illustrates., Technology will drive solutions to now-difficult problems, including efficient user access control (maybe via systems that combine biometric scanning and small portable devices with embedded computers) and clinical data entry (maybe via voice recognition systems that make use of intelligent thesauri). Cheaper mass storage is likely to be a significant part of the solution to the problem of making digital information permanently available. Access to the Internet is increasing rapidly among the U.S. population, but ensuring that everyone who needs health information has appropriate information technology remains an important challenge. A range of federal programs has been designed to promote connectivity, including the NLM's Internet connections grants for health institutions. After a slow start, the implementation of the universal service provisions of the Telecommunications Act of 1996 is now making access more affordable for some rural health care providers and libraries. Nonetheless, reliable high-speed access to the Internet remains an unsolved technical or economic problem in some geographic areas. As yet, the public Internet cannot deliver the quality of service needed for many routine health applications. There is reason to hope that engineering developments will make the Next Generation Internet better able to meet the requirements of health care, public health, and biomedical research. The National Research Council's Computer Science and Telecommunications Board recently released a report of an NLM-commissioned study of the technical capabilities required to make the Next Generation Internet suitable for many health applications. According to this report, health applications do not have unique technical requirements, but they do demand complex and variable combinations of technical capabilities and a higher degree of security than other types of applications. To ensure that health requirements are addressed in network engineering research and development, the biomedical community must increase its interaction with the Internet engineering community. The NLM has recently funded the implementation phase of a series of innovative medical applications of Next Generation Internet capabilities that may help achieve this goal. Organizational Connections | Organizational connections are the mutual agreements both in and between institutions to provide or obtain access to information on different systems. Even in single organizations, linking systems from different departments or facilities can be difficult, although institutions that have embraced IAIMS (e.g., Vanderbilt University and the Oregon IAIMS) have shown that it can be done. Lack of incentives for interorganizational cooperation is certainly one of the primary obstacles to forming aggregate health databases that cross institutional boundaries, as the case histories of many attempts to develop community health information networks illustrate. Perhaps the biggest interorganizational problem currently affecting connections between individual health records and digital libraries is the difficulty of arranging access to electronic publications produced by many different organizations, including commercial companies and scientific societies. Fearing loss of revenue, many publishers wish to use restrictive licence agreements that in essence eliminate the "fair use" rights available under U.S. copyright law and guidelines. In addition, publishers' customary license agreements may not support remote access by faculty, students, and practitioners ---let alone patients ---in clinics, offices, and homes. Working directly with publishers or through third-party consolidators, individual libraries and consortia can be successful in negotiating arrangements that support the uses that are most valuable to their users, but the negotiations may be time-consuming and expensive. To date, most biomedical journals available in electronic form continue to be available in print. Often, access to the electronic version means that a library must obtain a print subscription as well. One way or another, providing access to electronic journals often costs libraries more than purchasing print journals, many of which are already staggeringly expensive. Further complicating the licensing picture is the Uniform Computer Information Transactions Act (UCITA), recently proposed by the National Conference of Commissioners on Uniform State Laws. This document, which is intended to provide a uniform commercial framework for the licensing of information in all 50 states, presents numerous potential problems related to shrink-wrap and click-on licenses, user rights under the Copyright Act, consumer protection issues, and transaction costs. As a proposed model state law, UCITA will be debated, adopted, modified, or rejected by individual state legislatures. In the meantime, a major national policy debate about electronic intellectual property rights continues as society attempts to grapple with the impact of new technologies on the ability to reproduce, distribute, control, and publish information. The 1998 Digital Millennium Copyright Act ostensibly preserves fair use of electronic intellectual property, but under fairly murky conditions, which will probably be clarified in subsequent court cases. There is also a strong push to legislate new sui generis rights for databases, to protect the investment of those who have expended resources in aggregating facts or data that are themselves not covered by current copyright law. Current law protects the entire compilation, but not individual facts in it. The range of data affected could be enormous, including stock quotes, phone numbers, baseball scores, gene sequences, names of procedures in coding systems, and clinical values in aggregated health databases. Rising concerns about costs and restrictive access to electronic journals have led to serious efforts to develop alternative publication avenues. The Association of Research Libraries' Scholarly Publishing and Academic Resources Coalition (SPARC) seeks to promote the creation of lower-cost, peer-reviewed electronic journals to compete with specific commercial journals that are thought to be significantly overpriced. The National Institutes of Health have announced plans to host a free permanent repository of full-text scientific articles in the life sciences, called PubMed Central. (Earlier versions of this proposal were referred to as e-Biomed and e-Bioscience.) PubMed Central will be built by NLM's National Center for Biotechnology Information, with advice from an advisory board of nongovernment experts. The full-text articles it contains will be linked to their corresponding bibliographic citations in the PubMed database, which will be expanded to include any non-MEDLINE journals or articles deemed to be within the scope of PubMed Central. PubMed Central may become a primary dissemination mechanism for some existing and some new electronic journals. Since PubMed Central will permanently archive all the material it accepts, including high-resolution graphics, some publishers are expected to deposit articles on a delayed basis after their original electronic publication. As recent counter-moves by some publishers indicate, PubMed Central and other alternative electronic publishing and archiving mechanisms are bound to have an effect on scientific publishing. Conceptual Connections | As any current Web user knows, technical connections and permission to use electronic resources mean little unless an appropriate conceptual connection is made between the user's information need and pertient content in different digital resources. The NLM's UMLS project is focused on enabling these conceptual connections. It builds intellectual "middleware" ---electronic knowledge sources and related lexical programs ---to help system developers build applications that can interpret user inquiries and find and integrate relevant information from different information sources. The NLM continues to expand the UMLS products to improve their utility and to make them easier for system developers to use. Some success has been achieved, and the UMLS data and programs are now in production and research use in a wide variety of applications at NLM and elsewhere., Many challenging informatics research questions remain to be answered, however, before flexible and seamless linking becomes a reality and before systems become capable of delivering the specific piece of information needed when and where it is needed. Linking electronic health records and digital library resources will become easier when each resource of interest adheres to content conventions or standards, such as a defined set of data elements and the use of controlled vocabularies or classifications in some of these data elements. This is not to say that different types of resources will ever have the same set of content standards, just that it is easier if each follows some set of standards that can be factored into system development. Reliance on standards has certainly contributed to the success of the National Center for Biotechnology Information's efforts to integrate different genome resources. A 1990 comparison found considerable similarity in the types of content standards that existed for printed literature (which constituted the vast bulk of what was available in 1990) and those needed in the clinical arena. These types include: Although the types of required standards are similar, the extent to which standardization had been achieved and the impetus for standardization in the health care and library arenas were very different in 1990, and they remain very different today. Health Data Standards | In 1990, there was almost no standardization of clinical information and transactions, although de facto quasi-standard formats for some administrative and billing transactions were in wide use, primarily because of Medicare requirements. Today, there is substantially more standardization of the format and content of clinical transactions, particularly for laboratory tests and drugs, and a fair likelihood that standardization of other types of clinical information will increase. Factors that have contributed to the move toward standardization include the general trend toward electronic commerce, increasing automation in health care institutions, the spread of managed care, pressures to measure health care quality and outcomes, and heightened awareness of the potential importance of content standards, including controlled vocabularies. Some of these factors played into the enactment of the administrative simplification provisions of the Health Insurance Portability and Accountability Act of 1996 (HIPAA), which has had a significant effect on the development of content and messaging standards for both administrative and clinical data. In addition to its privacy requirements, HIPAA's administrative simplification provisions instruct the Secretary of Health and Human Services to establish national standards for a variety of electronic administrative health transactions, for coding systems and unique identifiers to be used in these transactions, and for ensuring the security of electronic health data. Although HIPAA's focus is administrative transactions, the line between administrative and clinical data is not a sharp one, particularly for claims attachments, which can contain detailed clinical information. HIPAA also expanded the role of the National Committee on Vital and Health Statistics (NCVHS) to provide wide-ranging advice to the Secretary and the Congress on health data policy, and specifically charged the NCVHS to develop advice on any necessary federal action to promote complete electronic medical records by August 2000. HIPAA's direct impact on clinical standards will come in the establishment of standards for claims attachments. The proposed rule for claims attachments will designate the HL7 (Health Level 7) clinical transactions standard and LOINC (Logical Observation Identifiers, Names, and Codes) as HIPAA standards. The expectation that LOINC will become a HIPAA standard, plus its recent selection as a target system for the exchange of laboratory test data between the Department of Defense and the Department of Veterans Affairs in the Government Computer-based Patient Record (GCPR) project, has prompted an interagency federal contract to support LOINC's ongoing maintenance and free electronic distribution. Federal agencies are also exploring government-wide licensing as a method of facilitating use of clinical vocabularies. HIPAA has undoubtedly increased the pace of development of health data standards. The Act requires the government to consult with existing consensus standards groups and to select currently available standards whenever possible. The HIPAA standards selection process has encouraged existing standards groups to cooperate to complete or modify existing standards or to develop new ones to meet HIPAA requirements. Speculation that the recommendations made by the NCVHS in 2000 may lead to additional clinical data standards has also speeded development and fostered collaboration between previously competing groups. Although there is probably still a long, hard road to travel before a complete set of standards for electronic health data is achieved, its likelihood is much greater today than it was a decade ago. Digital Library Standards | Meanwhile, the digital library has yet to achieve the level of standardization that applies to the organization, management, and sharing of printed materials. To state the obvious, there is substantial overlap in the printed materials acquired by different libraries, and considerable cost savings result if a cataloging or indexing record created by one library can be used by another. A relatively small number of large players, including the Library of Congress and the NLM, create and disseminate bibliographic records for significant segments of the printed literature. If local libraries use the same standards employed by these large players, locally produced records can be easily integrated with records obtained from outside sources. As a result, adoption of a standard by key producers or processors of bibliographic records for printed literature in essence guarantees the adoption of that standard by many other institutions. As of 1999, the factors that promoted standardization in records for the printed literature do not yet apply to most components of the digital library. Although there is great interest, discussion, and exploration of the issues involved in describing, organizing, obtaining access to, and archiving electronic information, consensus on a reasonable strategy for accomplishing these functions has yet to emerge. Some promising collaborative tests are underway, but libraries are not yet generating or sharing large numbers of descriptive records for electronic resources. Among the many reasons for this are the rapidly evolving nature of electronic information sources; uncertainty about whether separate descriptive records (distinct from standard descriptive data embedded in the electronic resources) will be necessary in the long term; slow progress in developing efficient mechanisms for integrated searching of Web documents and records in large existing catalog databases; and unwillingness to expend resources describing information sources to which access cannot be guaranteed, either because licenses will become prohibitively expensive or the sources will just disappear from the Web. Under these circumstances, it is not surprising that the development of standards for electronic information sources and the necessary harmonization of these standards with those for printed literature are both in their infancy. Given the NLM's mission to improve access to the results of research, it is fair to ask what the Library is doing to promote the standardization and cooperation that will improve current access to digital health information and ensure future access. The NLM's strategy is to use its own products and services as test-beds for technical and organizational approaches to organizing and managing digital information. The Library's goal is to gain experience from concrete experiments so that it can contribute to the development of workable national standards and strategies and also provide useful advice to other publishers of electronic information. Current NLM activities are focused on identifying the metadata and mark-up tagging needed for the NLM's own electronic publications and services, and on developing workable models for collaborative arrangements for organizing electronic resources, with MEDLINEplus, the Library's new consumer health Web resource, as a test case for such collaboration. The proposed Dublin Core set of metadata for electronic resources is the focus of considerable serious activity and publicity in the Web and library communities. As yet, it falls short of a reasonable standard because its extreme flexibility works against the creation of consistent metadata by different content producers. Like many other organizations, the NLM found it necessary to create additional metadata tags and to impose additional restrictions to promote consistency of application when adapting the Dublin Core for use in Profiles in Science. Metadata may also be a good way for Web publishers to announce their intentions regarding the permanence of the documents they are publishing. The NLM has initiated work to define and convey appropriate levels of permanence for the different types of electronic publications it produces and to develop the policy and procedures needed to guarantee those levels of permanence. An important theme in discussions about organizing and archiving Web-based information is the need for persistent, unique identifiers for electronic resources that provide shorter, more permanent links than current URLs. Of the several efforts to establish an infrastructure to support assignment and look-up of permanent unique identifiers for digital entities, the Digital Objective Identifier (DOI) initiative may be the most prominent. Neither the DOI nor any other system has yet achieved broad acceptance, and today it appears likely that there will be several different systems for different types of digital objects, including some built on existing national bibliographic systems for print publications. The success of PubMed and its approach to linking MEDLINE citations to the full text of articles available on publishers' Web sites has alerted scientific publishers to something that medical librarians have known for years. The NLM's article citation numbers are permanent unique identifiers for journal articles, and MEDLINE itself is a publicly accessible directory that includes the unique identifier, a standard description of the article to which the identifier refers and, in an increasing number of cases, the URL for the full text of the article. Many biomedical publishers now use NLM identifiers to link journal article references in their electronic publications to the corresponding MEDLINE records and, if available, to the electronic full text of the articles. The Library's formats for SGML/XML submissions of citations, abstracts, and now full text for PubMed Central probably also have a standardizing effect on electronic journal publishing. A New Environment for Achieving an Old Goal : "Electronic health record meets digital library" is the Web-era reformulation of the long-standing informatics goal of seamless integration of automated clinical data and relevant knowledge-based information to allow informed decisions. When this goal first gained currency, the assumption was that health care professionals were the decision makers. Clinicians remain a primary target for integrated "just-in-time" information services, but these services are also needed by public health professionals and, in an era in which individuals are assuming more responsibility for their health, by patients and the well public. Technical developments during the last decade have set the stage for achieving rapid delivery of specific knowledge relevant to an individual person or community at the time and place of need. Systems that can link electronic health data and knowledge-based information in limited settings or limited ways are here today. Their existence partly explains the increased public interest in resolving relevant policy issues, including health data privacy, the "digital divide," electronic intellectual property rights, and standards for health data and for electronic publishing. While the environment is highly favorable for progress, many important informatics research problems remain to be solved before electronic health records and digital library resources are linked "seamlessly" to support decisions and actions that improve individual and community health. Solutions are likely to come from the work of multidisciplinary teams of health professionals, medical informaticians, librarians, computer scientists, linguists, and behavioral scientists and from continuing serious effort to achieve positive resolution of policy issues. These are worthy priorities for the beginning of this millennium. Backmatter: PMID- 10984464 TI - Improving Clinical Communication --A View from Psychology AB - Abstract Recent research has studied the communication behaviors of clinical hospital workers and observed a tendency for these workers to use communication behaviors that were often inefficient. Workers were observed to favor synchronous forms of communication, such as telephone calls and chance face-to-face meetings with colleagues, even when these channels were not effective. Synchronous communication also contributes to a highly interruptive working environment, increasing the potential for clinical errors to be made. This paper reviews these findings from a cognitive psychological perspective, focusing on current understandings of how human memory functions and on the potential consequences of interruptions on the ability to work effectively. It concludes by discussing possible communication technology interventions that could be introduced to improve the clinical communication environment and suggests directions for future research. Keywords: Introduction : Communication between health care workers accounts for the major part of the information flow in health care, and growing evidence indicates that errors in communication give rise to substantial clinical morbidity and mortality. Covell has reported that about 50 percent of information requests by clinicians in clinic were met by colleagues rather than by documented sources. Safran et al. reviewed the information transactions in a hospital with a mature computer-based record system and still found that about 50 percent of information transactions occurred face-to-face between colleagues, with e-mail and voicemail accounting for about another quarter of the total. Given the importance of interpersonal communication as a means of information exchange, it is not surprising that communication failures are a large contributor to adverse clinical events and outcomes. In a retrospective review of 16,000 in-hospital deaths, communication errors were found to be the leading cause, twice as frequent as errors due to inadequate clinical skill. Furthermore, in a study of primary care physicians, about 50 percent of all detected adverse events were associated with communication difficulties. The causes and remedies of poor communication in the health care system are consequently of critical interest to the study of informatics. Much work has examined the dynamics of communication between individual health care providers and patients, and this body of work can help optimize that interaction.,,, Similarly, there are bodies of work on nurse and physician perceptions of, and satisfaction with, their communication,,,, and on physician teaching behaviors.,,, However, there is a paucity of detailed data about the effect of communication behaviors on overall organizational efficiency and effectiveness in health care. Some studies have investigated the use of communication technologies in health care settings. One study in a nursing home found that the number of telephone calls nurses received constituted a significant communication burden and that most calls were routine or informative only. Introduction of a voicemail system allowed the vast majority of calls to be transferred to that medium, saving both time and unnecessary extra communication. Spurck et al. found that nurses and physicians felt that their hospital's telecommunication system was more effective when nurses were given portable phones to carry, and noted several efficiency gains. Another study looked at the effects of voicemail on internal and external customer satisfaction. None of these studies, however, explored the basis for individual communication choices or the cumulative effect of those choices on clinical teams or on the wider organization. An exception is a study by Coiera and Tombs, who observed that the communication behaviors of individuals in hospital teams are often individually inefficient or unsuccessful and, when taken as a whole, result in an interrupt-driven environment in the organization. A large number of factors might influence communication behavior in organizations, including the nature of the available communication infrastructure, the nature of the work undertaken, and the practices that are routinely applied in the organization by individuals. A limiting factor in any communication analysis is the cognitive capacity of individuals to undertake their work, and in studies of high cognitive workload, it has often been shown that error or inefficiency results when cognitive limits are exceeded. In this paper, we focus on the cognitive limits of individuals and how these might explain the inefficient and interruptive clinical environment described in the Coiera and Tombs study. We do so by drawing together the available clinical findings with empirical research from cognitive psychology. In particular, we are interested in human memory, and we speculate that the burden on memory that results from a highly pressured working environment is a significant contributor to the interruptive behaviors observed. By adopting this approach, we hope to achieve several goals. Our first goal is to provide a principled framework in which existing findings about communication patterns in health care organizations can be understood and can direct further research. Our second goal is to promote an understanding, based on empirical psychological research, of the sorts of problems in real working environments that can, potentially, be addressed by the appropriate application of new communication technologies. In addition, in discussing the social environment into which new communication technologies are introduced, we aim to caution against an approach that does not make the social environment the object of study. Enthusiastic applications of new technologies do not always have the consequences expected of them. This is not least because they are always introduced into a social environment, and this often acts strongly to modify the ways in which their capabilities are harnessed. Finally, we hope to stimulate psychologically and sociologically informed research into the effective application of new communication technologies to hospitals. In the next sections, we introduce relevant concepts from cognitive psychology and then relate these to the existing clinical communication findings. In particular, we discuss specific types of memory error that are likely to be seen in interrupt-driven environments. We then use this model to discuss the likely consequences of introducing different communication technologies into this environment. Clinical Communication Patterns : Coiera and Tombs observed the communication patterns of eight physicians and two nurses in an English district general hospital. The available channels of communication for these highly mobile professionals consisted of face-to-face meetings, both impromptu and planned; desktop telephones; paging; written notes for colleagues in patient notes; notes at ward desks; notice boards; and pigeon holes for personal memos. Voicemail and e-mail were not supported, and mobile telephones were not used. The subjects in this study, like those in others, made little or no use of more formal sources of information, with the exception of data from the medical record. One of the communication behaviors observed was a bias toward interruptive communication. Interruptive, or synchronous, communication methods require the simultaneous interaction of the two parties to the communication: the telephone and face-to-face discussions are two such methods. In contrast, an asynchronous method, such as writing a note or leaving a voicemail or answerphone message, allows the recipient to deal with the communication at a time of his or her choosing. The authors reported that staff showed strong preferences for making telephone calls and for taking advantage of chance face-to-face meetings with colleagues. There was little evidence that staff's own experience of interruptions encouraged them to adopt more "considerate" communication methods when contacting their colleagues. The authors also observed that the reliance on synchronous methods can be a source of inefficiency for the person attempting to communicate. Recipients may be unavailable or occupied, or the communication channel may be busy, and tasks remain undone until these conditions change. Coiera and Tombs' observations are of interest from a psychological perspective, for two reasons. First, cognitive psychology may offer explanations for reliance on synchronous communication. Second, the interrupt-driven nature of hospital work may foster conditions that are likely to result in impairments to memory during the working day, which potentially contribute to clinical errors. In both cases, the application of psychological theory will allow future observational studies to be designed that specifically investigate the bias to synchronous communications and whether and under what circumstances the postulated memory impairments occur. We therefore introduce in brief some of the relevant concepts from cognitive psychology and discuss how these concepts might apply in this environment. Human Memory : A knowledge of the way in which human memory is believed to function is key to understanding the probable effects of working in an interrupt-driven environment. It is also key to understanding the requirements of technologies that might be introduced to support those in such environments. The functioning of human memory has been the subject of empirical study in cognitive psychology for many decades. A basic division of memory into short- and long-term functional components provides the cornerstone on which the working of memory is understood. Our knowledge is believed to be stored in notional "repositories," known as long-term memory. Remembering medical facts, significant dates, events from childhood, and how to drive a car all draw on long-term memory. Much of the time, most of our fund of knowledge and meaning in long-term memory is inactive; that is, it is not the current focus of attention. Working memory is believed to be the activated state of information held in memory. It may be equated with the component of memory we associate with attention. Working memory actively processes information, whether the information is sensory input (for example, sounds, sensations, or sights currently being experienced) or items from long-term memory. When carrying out a mental calculation, making a plan to do something, recalling a phone number, or writing a note, it is working memory that allows the various "pieces" of information to be attended to, integrated, and manipulated. Working memory has some interesting characteristics. In particular, it is extremely limited in its capabilities. The number of items ---such as thoughts, sensory impressions, and plans ---that can be held in working memory is very small., Furthermore, items in working memory are easily disturbed by each other. This is particularly the case when someone is distracted from thinking about one task by a new one that supervenes. An intention to carry out an act can be forgotten by the intrusion of another plan, even when only ten seconds separates the intention from the intrusion. Working memory is also severely limited in duration. Without conscious attention to plans or other items in working memory, the accurate memory persists no longer than about 20 seconds. This decay can be overcome by acts of conscious self-reminder: these serve to refresh and re-prioritize the items in working memory. If there are competing demands on working memory, however, such as executing another task or communicating with a co-worker, then such rehearsal of intention becomes impossible, with the same effect ---a plan may be forgotten. Considerable empirical evidence shows the powerful negative effects of both interference and diversion of attention, on working memory. The characteristics of the process of forgetting are not random; in particular, two serial position effects, known as the primacy and the recency effects, are known to affect retention in systematic ways. The primacy effect describes the tendency toward superior recall of items that have resided longest in working memory; the recency effect describes a similar superiority of items most recently added to working memory. The combined result of these two effects is poor retention of items in the middle of the "mental list." In addition, a distracting task before recall can obliterate the recency effect but not affect the primacy effect. A further distinction may be drawn in long-term memory between retrospective memory and prospective memory. Retrospective memory refers to the factual, autobiographical, and instructional ("how-to") knowledge we possess. Prospective memory, in contrast, is the memory for a future act, or the memory to remember to do something. It necessarily draws on retrospective memory and entails complex planning and coordination. Like retrospective memory, prospective memory relies on working memory for its processing work. For example, in remembering that you need to contact someone later in the day, you draw on your retrospective memory in deciding how to make the contact. Failures of Working Memory in an Interrupt-driven Environment | Those who work in an interrupt-driven environment are likely to suffer failures of working memory. As interruptions occur, interfering with the active cognitive rehearsal of what is to be done and generating new tasks for their recipients, prospective plans may be partly or fully forgotten. They may or may not be recalled subsequently, depending on appropriate cues for recall. The greater the number of such plans, the more the effect will be exacerbated. Since planning for prospective activities is an activity of working memory, which is limited in the number of distinct items it can retain, then forcing more items into working memory will cause some plans to be displaced from it, perhaps to be forgotten. The tasks that should suffer most from interruptions are all but the oldest tasks in a nurse's or doctor's mental "to do" list. Coiera and Tombs proposed that immediate acknowledgement of a message seemed to be needed in such an interruptive environment to permit workers to complete a task. Good psychological reasons may be advanced for this and for the "selfish" behavior observed. When a worker's working memory is operating to capacity, the highest priority is likely to become the reduction of this mental burden by completing the tasks that are consuming memory resources. Reinforcing this, it is probable that when the consequences of errors can be so serious, it is difficult for a doctor or nurse to feel that he or she has truly "handed over" responsibility for a task without an explicit acknowledgement from the recipient. Existing asynchronous methods of communication, such as hand-written notes, voicemail, and e-mail do not easily or routinely offer this feature. The consequence may be that the task cannot be removed from working memory. Errors of Reality Monitoring and Temporal Association | Two further types of memory error may be made more likely by interruptive working environments. "Reality monitoring" is the ability to discriminate between "true" and "false" memories. True memories are memories of events, objects, and actions that really occurred or were experienced. False memories derive from the imagination that something occurred or was experienced. Confusion between the two is a quite ordinary occurrence. It may be seen as a consequence of the way in which memory functions as creatively constructed representations rather than as simple records of sensory information. Childhood memories, for example, are often a composite of the original experience interwoven with what other people said about it, as well as the embroideries that are overlaid on the original fabric as the experience is mentally revisited over time. Reality monitoring errors may be either omissive or repetitive. If the memory of an intention to act is confused as a memory of having acted, an error of omission will occur. If the memory of the performed act is mistaken as the memory of a plan to act, the error will be one of repetition. For example, if an intention to take a dose of medicine is confused as the action of having done so, a dose will be missed. If the action is mistaken as the intention, an extra dose will be taken. The second type of errors are those of "temporal association." These errors are thought to be strongly associated with routine and frequently performed actions. The more repetitive and routine an action is, the more difficult it can be to decide whether a memory of the action is today's or yesterday's memory. For example, did I really clean my teeth this morning or am I remembering cleaning my teeth yesterday morning? Both reality monitoring and temporal association errors can, therefore, result in the omission or repetition of tasks. Given that they are fundamental tendencies of the memory system, we may expect them to be additional sources of memory failure in an interruptive working environment. These errors will be especially likely under particular circumstances. Outstanding tasks that are simple, routine, and repetitive are particularly vulnerable. Under conditions of high work pressure, when there is insufficient time to perform a reality check, errors may be more readily accepted. Furthermore, since junior doctors, like nurses, are required to undertake many more routine and repetitive tasks than are senior doctors (for example, ordering laboratory tests and securing ward beds for new admissions), it is probable that they are more vulnerable to failures of reality monitoring and temporal association than are senior doctors. The Effects of Expertise on Memory | There is a large body of research on the effects of skill acquisition on problem solving and memory,,,, which indicates that with experience, some components of tasks can be performed automatically. They are sufficiently well learned that once set in train they do not rely on working memory to be enacted, thereby freeing components of working memory for alternative use. This means that the probability of a memory error is greater for less experienced members of staff. Junior medical officers, for example, are novices in the practice of any given speciality of hospital medicine. As the tasks associated with each level of seniority differ, this will also be true of newly appointed specialist medical staff. Since experts need to rely less on general attentional resources than do novices, it is probable that more experienced doctors at every level of the hierarchy, as well as nurses and other health care professionals, will suffer less from the effects of interruptions in the performance of specific tasks than will their less experienced colleagues. In the Coiera and Tombs study, the greatest communication burden actually fell on the most junior staff, whom one would expect to be the group most likely to make errors in such circumstances. Discussion : New Communication Technologies | Coiera and Tombs observed two main contributors to the interrupt-driven nature of the hospital environment ---the behaviors of hospital workers and the characteristics of the work itself. The work was highly mobile, conducted in multidisciplinary teams, and involved many simultaneous tasks and responsibilities. The existing communication environment in the hospital in which the study took place relied largely on synchronous communications and did not support mobility. Individuals responded to these factors by favoring synchronous communications even when they were not necessary or even productive, thus increasing the interruptiveness of the working environment. The previous exploration of memory functioning demonstrates that working in a busy and interrupt-driven environment can over-extend the capabilities of the human cognitive system. In such an environment, there is a premium on immediate task completion and reliance on synchronous communication. Such behaviors permit ambiguities and uncertainties to be dealt with on the spot and can thus be construed as reasonable adaptations to working in such an environment. Coiera and Tombs suggested technologies that could reduce the interrupt-driven nature of hospital work. Portable telephones could support mobility, and asynchronous communications technologies such as voicemail and e-mail, with acknowledgements, could fulfill the initiator's need for immediate task completion without generating an interrupt for the recipient. The introduction of new technologies seldom, however, permits such straightforward predictions to be made. Social influence approaches,,, to studying information technology have shown that the use of new technologies is not predicted solely by the characteristics of the technologies themselves. Instead, the human environment into which they are introduced is critical in shaping how their capabilities are actually used. Attitudes of key individuals and organizational norms are among the important factors that shape adoption and use. There are thus many uncertainties about the actual use of technologies. Correspondingly, this is a fertile area for research, no less in the medical field than in the world of the office. In the following sections, some of the difficulties with these simple predictions about introducing new technologies are explored, to underline the difficulties that unexpectedly arise when apparently simple solutions are introduced into complex human work environments. Increasing Interruption and the Bias to Synchronous Communication | What might be achieved by providing staff with mobile phones? The general effect of introducing mobile phones is to make individuals more available. Since failure to reach individuals results in further attempts to make contact, mobile phones would be expected to reduce the call failure rate and thus the overall call traffic for the organization. For the individual caller, it could mean a reduction in the number of outstanding tasks, as call recipients are easier to reach. This means that the corresponding number of items in working memory that are associated with a task no longer compete for attention. For the recipient, however, the picture is less clear. If providing mobile phones only reduces the number of call re-tries, we would expect interruption levels to be unchanged. However, if the ease of contacting individuals has the effect of creating additional calls for conversations that would not have occurred previously, then the overall interruption level would increase. We would predict in these circumstances that each new call generates an interruption, and an addition to working memory. The synchronous bias hypothesis predicts that individuals preferentially use synchronous communication channels. Mobile telephones, by making synchronous communication easier, would thus be predicted to result in new calls being made and consequently would result in an increase in the overall interruption rate for individuals. In such circumstances, new synchronous technologies therefore would not, on their own, resolve the practical or the cognitive difficulties faced by those in interruptive working environments. Asynchronous Messaging | What might the provision of asynchronous technologies mean? The technology would permit the message sender to achieve task completion independently of the recipient's location and current activity. The recipient may choose a convenient time to consult and act on his or her messages. The cognitive benefits of voicemail and e-mail for both message senders and recipients could be substantial. For callers, independent completion of communication tasks reduces the number of pending tasks in working memory. For call recipients, the receipt of fewer interruptive calls would be likely to contribute to greater chunks of uninterrupted time and greater ability to rehearse and recall existing outstanding tasks. It would allow completion of more tasks, fewer errors in task completion, and fewer forgotten tasks. These probable benefits from moving some communication tasks from synchronous to asynchronous channels would result from a decreased incidence in the factors contributing to memory errors, such as distraction, interference, and new involuntary additions to prospective memory. There is much to be learned about the degree to which and circumstances under which callers might choose to employ asynchronous channels. While call recipients would probably choose to deal with their calls at one time, the same is not likely of callers. Since each undone communication task remains an item in prospective memory, it is instead desirable to carry out communication tasks as the need for them occurs. Tasks that do not require immediate acknowledgement or completion lend themselves to asynchronous methods of communication. Under what circumstances might the employment of an asynchronous method form the sender's first preference, rather than the last resort, when attempts at synchronous communication have failed? Might callers choose their communication methods on the basis of the demands of the task, and could careful design of the technologies encourage a shift to task-based use of communications? The Effects of Social Influence on Technology Use | There are a cluster of very interesting questions relating to the use of communication technologies and status. How might the choice and use of synchronous and asynchronous communication be affected (if at all) by an individual's status in the hospital hierarchy? Might the greater certainty of connection with junior staff encourage more "selfish" behavior by those higher in the hierarchy? That is, for routine communication, might more senior members of a team feel more free to interrupt junior members with synchronous communication and, conversely, would more junior members, more reluctant to interrupt, tend to use asynchronous voice messaging or e-mail to communicate with senior staff? As research with office,,, and health care workers indicates, it is probable that staff would influence one another's adoption and patterns of use and that local norms might evolve. Since communication technologies are essentially shared tools, the degree to which they may be fully exploited is contingent on all parties to a communication being prepared to use, and feeling satisfied with their ability to use, the capabilities provided. If, for example, one member of a team were reluctant to access text messages, the behavior of other members of the team would either have to evolve to accommodate that antipathy or to exert influence to convince that team member to behave differently. Thus, it is probable that differences in use of available technologies might be seen from team to team, shaped by key individuals. Effects on the Nature of Conversations | Coiera and Tombs indicated that chance face-to-face meetings were an important medium of communication. One of the reasons that such meetings were so eagerly seized on was precisely the difficulties the study participants had in either locating colleagues or setting up synchronous conversations. These meetings often provided the first opportunity one colleague had to confer with another on a particular matter and sometimes substituted for the failed attempts at earlier conversations. We might expect that the use of voicemail and e-mail would cause some changes to the content of opportunistic exchanges between colleagues. Instead of communication of the nature "I need to speak to you about X," we could expect that opportunistic exchanges might instead be more of the nature "Did you read/receive/act on my voice/text message to you about Y?" or "Thanks for your voice message about Z; I'll see to it this afternoon." In other words, instead of representing the first opportunity one colleague has to communicate with another, opportunistic meetings might become opportunities to confirm earlier communications and, perhaps, elaborate on them. Message Acknowledgement | Coiera and Tombs suggested that the need for acknowledgement of receipt of a message was one of the drivers behind the preference for synchronous communication. The cognitive reasons for this preference were discussed above: without confidence that the receiver has taken over the task, it remains an unfinished task in the caller's working memory. Acknowledgments could, perhaps, be required for different purposes. Has the message arrived safely in the recipient's "in" tray? Has the recipient listened to or seen the message yet? The message sender might require different types of acknowledgment, to be able to feel that the communicated task has truly been delegated and that he or she is thus able to "remove" that task from working memory. Additional interesting questions are raised by research in the office on acknowledgments mediated by e-mail. In some circumstances, agreements mediated by e-mail were not viewed by recipients as equivalent in strength or reliability to face-to-face agreements on a course of action. Co-workers wished to look each other "in the eye" when negotiating and agreeing on commitments. Might this also be true in the hospital environment? Future Research | We have hypothesized that in some cases, over-dependence on synchronous channels of communication may come about because these conditions result in excessive burdens on memory. At face value, synchronous communication may seem to provide the best way to complete some tasks and reduce unfinished tasks held in working memory; in fact, difficulties with this channel, ranging from unanswered calls to not knowing the location of a colleague, often mean that many inefficiencies are introduced into the process. These are hypotheses that require further examination in the field. Studying cognitive phenomena in a naturalistic setting is a method of enquiry that is increasing in popularity and is increasingly seen to complement the older laboratory-based method of enquiry, to the benefit of both. Laboratory studies, with their strict experimental controls, enable cause-and-effect relationships to be postulated and examined and theory to be developed. However, laboratory experiments are often criticized as unrepresentative of real life. Field studies, on the other hand, lack this control and therefore cannot demonstrate cause-and-effect relationships. This does not mean, however, that laboratory-based theory cannot be applied to the field and used to shape and guide enquiry. Given the complexity of this field of study, we propose a program of research to address its many facets. At least three types of approach are required. First, future studies should be designed specifically to examine memory errors in the hospital setting. Structured observation techniques could be used to focus on the occurrence of memory errors, and interviews with study participants could draw on their own accounts of observed data. While self-accounts do not offer proofs of theory, they enrich and inform the observer's understanding; combined observation and interviews are the basis of ethnographic approaches. The same approach is required to further understand communication behaviors. Furthermore, semi-experimental studies may bridge the need to understand behaviors in a working environment while executing controlled experiments. For example, artificially structured memory tasks could be given to staff to carry out during their routine work. Second, and informed by the ethnographic approach, quasi-experimental studies, are needed to assess the consequences of introducing new communication technologies. For example, asynchronous technologies may be introduced to a mobile and distributed team. Studies of relevant behaviors using before and after comparisons ---as well as comparisons over the same period with a different team that has not used the new technologies ---should also be carried out. Because of the complexity of the phenomena under study, separate studies investigating communication behaviors and memory errors will be required. Third, the introduction of new technologies also requires investigation of the social environment into which they are introduced. The ethnographic approach,, a method of observation focused on learning the meaning for participants of particular behaviors, will help characterize the social aspects of the communication environment, and drive hypothesis construction and further experiments. Conclusions : This report has taken a cognitive psychological approach to set the scene for future investigations of clinical communication behavior. It has outlined some of the negative consequences that interruptive communication patterns could have on the ability of hospital staff to manage successfully their current and impending task loads. It has detailed several characteristics of human memory that might contribute to greater cognitive burdens. It has indicated some of the characteristics of tasks that contribute to the probability that they will be forgotten in interrupt-driven circumstances. It has suggested that the degree of an individual's expertise is one characteristic that could be used to discriminate between the differing vulnerabilities of individuals to interruptions. Finally, it has combined this cognitive approach with a social information approach to consider the effects of new communication technologies in the interrupt-driven environment of the hospital. This approach reveals great scope for studying existing communication patterns and work practices in hospitals to understand their effects on memory and task performance. This is a necessary precursor to developing an understanding of the consequences of introducing new communication technologies. Backmatter: PMID- 10984465 TI - Comparative Evaluation of Three Continuous Speech Recognition Software Packages in the Generation of Medical Reports AB - AbstractObjective: To compare out-of-box performance of three commercially available continuous speech recognition software packages: IBM ViaVoice 98 with General Medicine Vocabulary; Dragon Systems NaturallySpeaking Medical Suite, version 3.0; and L&H Voice Xpress for Medicine, General Medicine Edition, version 1.2. Design: Twelve physicians completed minimal training with each software package and then dictated a medical progress note and discharge summary drawn from actual records. Measurements: Errors in recognition of medical vocabulary, medical abbreviations, and general English vocabulary were compared across packages using a rigorous, standardized approach to scoring. Results: The IBM software was found to have the lowest mean error rate for vocabulary recognition (7.0 to 9.1 percent) followed by the L&H software (13.4 to 15.1 percent) and then Dragon software (14.1 to 15.2 percent). The IBM software was found to perform better than both the Dragon and the L&H software in the recognition of general English vocabulary and medical abbreviations. Conclusion: This study is one of a few attempts at a robust evaluation of the performance of continuous speech recognition software. Results of this study suggest that with minimal training, the IBM software outperforms the other products in the domain of general medicine; however, results may vary with domain. Additional training is likely to improve the out-of-box performance of all three products. Although the IBM software was found to have the lowest overall error rate, successive generations of speech recognition software are likely to surpass the accuracy rates found in this investigation. Keywords: Introduction : Changes in health care are increasing the demand for electronic records in large organizations. Medical professionals who do not have access to transcription services must type their own chart entries, which requires typing skill and significant amounts of time. Because of increased sharing of patient care across multiple facilities, VA New England is interested in evolving technology-based approaches to enhancing documentation of patient care in an electronic form. This study was undertaken in part to assess the potential use of speech recognition software in busy clinical settings without transcription support, prior to a decision on significant capital investment. By the close of the 1990s, speech recognition software had become a potentially viable and affordable substitute for transcription, costing approximately $2,000 per workstation with software. Software that converts the spoken word to text has been used in many specialized health care settings (e.g., radiology and cardiology). A search of medical and psychological journal listings (using MEDLINE and PsychLit) revealed few published articles evaluating speech recognition software in health care settings. It is noteworthy that the majority of these studies evaluated discrete speech recognition software. A number of software reviews have been published in the popular press and computer trade magazines (e.g., PC Magazine, Nov 1999), but none of these publications has provided a systematic comparison of continuous speech recognition software package performance. Zafar et al. conducted the only published study evaluating continuous speech recognition software. Their article includes a comprehensive overview of the process of continuous speech recognition, a comparative evaluation of the specifications of three different products (IBM ViaVoice Gold, Dragon Systems NaturallySpeaking, and Philips SpeechMagic), and findings from tests conducted with these three products. The authors reported accuracy rates as high as 98 percent with significant training of the software, but their methods for categorizing and counting recognition errors were not described. Although the authors tested three different software packages, they did not present comparative data that would allow evaluation of the relative accuracy of each software package. In addition, it was unclear from the results whether the recognition rate reported was for a single dictator or was the average of the three authors' recognition rates. Furthermore, the high rate of recognition may be related to the skill of the small number of dictators who evaluated the software. The primary objective of the present study was to compare the relative accuracy of three continuous speech recognition software packages. The study was designed to evaluate out-of-box performance that may be expected, with minimal training of the software, by both experienced and inexperienced dictators. In addition, the study was designed to allow for comparative evaluation of error rates within specific categories of errors (e.g., medical abbreviations and punctuation). Materials and Methods : Materials | Software | The following continuous speech recognition packages were evaluated in this study: IBM ViaVoice 98 with IBM General Medicine vocabulary (IBM, Armonk, New York); Dragon NaturallySpeaking Medical Suite, version 3.0 (Dragon Systems, Inc., Newton, Massachusetts); and L&H Voice Xpress for Medicine, General Medicine Edition, version 1.2 (Lernout & Hauspie, Burlington, Massachusetts). Hardware | Pentium II computers of identical specifications were used for comparison of the three software packages. Each computer was a 333-MHz system equipped with 128 MB of RAM and a Creative SoundBlaster-compatible sound card. Subjects | Twelve physicians working with the VA New England Healthcare System participated in the study. Participants were all male (an accident of availability) and ranged in age from 29 to 59 years, with a mean age of 46 years. The majority (11 of 12) spoke English as a native language. All participants reported using a computer at least two hours a day, and almost half (5 of 12) reported previous experience with voice dictation software. Only two participants reported ongoing use of speech recognition software at the time of the study. Medical Record Samples | Four sample medical record entries (two progress notes, one assessment summary, and one discharge summary) were chosen for dictation based on vocabulary, formatting complexity, and length. All material was drawn from actual records. Identifying information contained in the reports was altered to preserve patient confidentiality, and the reports were edited for typographic errors. Procedure | All testing of speech recognition software was conducted in a single day for each clinician. Each clinician completed enrollment and dictation trials for the three software packages. Because of the potential threats to validity posed by a within-subjects design (e.g., fatigue, learning effects), the order of use for each software package was counterbalanced using a Latin square design. Enrollment | Each clinician completed the standard voice enrollment for each software package. Although additional training of the speech model is available in each package, and extended training of voice recognition software will improve performance and ease of use, an abbreviated training period (less than 60 min) was chosen in this study to evaluate this software for use in a setting in which extensive training may not be practically feasible. Following voice enrollment, participants completed two medical chart dictations (a medical progress note, and one psychological report) for practice and then began the dictation trials. Dictation Trials | For the scored trials, each clinician completed a 707-word medical discharge summary and a 257-word medical progress note. The progress note contained 227 words of general English vocabulary, 17 words of medical vocabulary, 13 medical abbreviations, 10 numbers, and 48 punctuation marks. The discharge summary contained 568 words of general English vocabulary, 98 words of medical vocabulary, 41 medical abbreviations, 90 numbers, and 216 punctuation marks. Medical vocabulary was defined as words that were unlikely to appear outside a medical context (e.g., erythematous); English vocabulary was defined as words that can be found in nonmedical prose (e.g., trauma). Dictators were instructed not to correct errors in dictation either by voice or by typing. Time to complete each dictation was recorded, and a copy of the generated file was saved for evaluation of speech recognition errors. Participants provided a subjective rating of each software package following completion of the dictation trial with each package and then again following completion of trials with all three packages. Dictation Sample Scoring | Errors in dictation were assessed by word-for-word comparison of a printed copy of the dictation samples and the captured dictations. Errors in recognition were categorized and recorded in 43 distinct groups, which cluster into five broad areas: general English vocabulary misrecognition, medical vocabulary misrecognition, extra word insertion, number misrecognition, and punctuation misrecognition. ( shows a breakdown of each error type within the broad categories.) Several scoring procedures were implemented to improve the consistency of scoring: 1) all dictations were scored in a group (of three investigators and two research assistants) so that all scoring issues could be decided by group consensus; 2) each dictation (across packages) for a single participant was scored by the same investigator; and 3) after all dictation samples for the medical progress note had been scored, investigators performing all subsequent scoring were blinded to the software used. Rules were also implemented to ensure consistency of scoring. For example, the maximum number of possible errors in a dictation was based on the number of items in the dictation sample, and parts of speech that improved grammar but were not present in the dictation sample were not counted as errors. In addition, many minor rules (not described in the present paper) specific to each error category were used to guide scoring (e.g., "low back pain" is an acceptable substitute for "LBP"). Table 1 | Voice Scoring Categories Results : The study was designed to evaluate the accuracy of three continuous speech recognition products in the generation of medical chart entries. Examination of combined errors across categories (general English vocabulary, medical vocabulary, medical abbreviations, numbers, and punctuation) revealed that the IBM software had the lowest mean error rate (6.6 to 8.4 percent) followed by the Dragon software (12.0 to 13.9 percent) and then the L&H software (13.8 to 14.6 percent). Examination of the overall error rate for recognition of vocabulary alone (General English vocabulary, medical vocabulary, medical abbreviations) yielded similar results. IBM had the lowest mean error rate (7.0 to 9.1 percent) followed by L&H (13.4 to 15.1 percent) and then Dragon (14.1 to 15.2 percent). Order Effects | Repeated measures analyses of variance were conducted to test for any order effects that may have resulted from the within-subjects design. Order of use for the three software packages was entered as a between-subjects variable, and total error scores for each package were entered as the within-subjects variables. Analysis of the progress note data revealed no significant effect for order of use (F[2,8] = 0.013, P < 0.987). Similar results were found with the discharge summary data (F[2,9] = 0.248, P < 0.785). Time to Complete Dictation Trials | The length of time needed to dictate the 257-word progress note was consistent across products, with no significant differences (F[2,22] = 1.05, P < 0.336) in the mean dictation time for IBM (M = 6.0 min), Dragon (M = 5.2 min), and L&H (M = 6.4 min) software. The length of time needed to complete a 938-word discharge report, however, was significantly different between packages (F[1,10] = 10.9, P < 0.01], with Dragon taking the shortest amount of time (M = 12.2 min), followed by IBM (M = 14.7 min), and L&H (M = 16.1 min). Comparative Error Rates Across Packages | Analyses of variance of the discharge summary data revealed significant differences in error rates among software packages in the recognition of general English vocabulary, medical vocabulary, medical abbreviations, numbers, and punctuation. ( shows ANOVA results.) Slightly different findings emerged in the analyses of the progress note data. Significant differences in error rates were found in the recognition of general English vocabulary, medical abbreviations, and numbers but not in recognition of medical vocabulary and punctuation. Overall Speech Recognition Error Rates | To examine the overall error rates for the three packages, the ratio of errors (observed errors/possible errors) was examined for each category in which a significant difference was found among packages. shows that for the progress note, IBM had the lowest rate of errors for general English vocabulary (M = 7.65 percent), medical abbreviations (M = 23.78 percent), and numbers (M = 13.64 percent). The overall error rate for all items combined (words plus numbers and punctuation) was lowest for IBM (8.40 percent) followed by L&H (13.85 percent) and Dragon (13.88 percent). A series of pairwise comparisons showed that the IBM error rate was significantly different from both the Dragon error rate and the L&H error rate for both English vocabulary and medical abbreviations. The error rate for recognition of numbers was significantly different between IBM and L&H and also between Dragon and L&H, but not between IBM and Dragon. Similar results were found for the discharge summary. IBM had the lowest rate of errors for general English vocabulary (M = 6.22 percent), medical vocabulary (M = 9.10 percent), medical abbreviations (M = 13.01 percent), numbers (M = 10.56 percent), and punctuation (M = 3.70 percent). shows that the overall error rate (all items combined) was lowest for IBM (6.62 percent) followed by Dragon (12.03 percent) and L&H (14.62 percent). A series of pairwise comparisons showed that the IBM error rate was significantly different from the L&H error rate for all error categories. Significant differences between IBM and Dragon emerged for English vocabulary, medical vocabulary, and medical abbreviations, but not for punctuation and numbers. Significant differences between L&H and Dragon were found for both medical abbreviations and numbers. Overall Recognition Rates by Previous Dictation Experience | A series of independent sample t-tests were conducted to determine whether participants who had significant experience with dictation (at least two years of experience with either a transcription service or voice recognition software) achieved better rates of correct recognition than participants who had no experience with dictation. Analysis of the progress note data showed that past dictation experience was not related to the performance of the IBM package (t = 0.952, P < 0.366), the Dragon package (t = 1.80, P < 0.105), or the L&H package (t = 1.53, P < 0.366). Analysis of the discharge data revealed similar results. Comparative Evaluation of Findings | A series of paired sample t-tests were conducted to determine whether there was a significant difference between mean error rates for each package between dictation trials. Mean percent error rates for the discharge summary and progress note data were not found to be statistically different for IBM (t = -0.677, P < 0.512), Dragon (t = -0.354, P < 0.730), and L&H (t = 0.964, P < 0.356) software packages. Subjective Ratings | In response to the question "Would you use this software again?" 100 percent of participants replied "Yes" for both the IBM and Dragon products, whereas only 66 percent of participants replied "Yes" for the L&H product. Following use of all three products, participants were asked to rank order each system on the basis of their perception of the product 's performance. The IBM product received the most favorable responses, with 92 percent (11 of 12 participants) ranking it number one, and 8 percent (1 of 12 participants) ranking it number two. The Dragon product was rated number one by 17 percent (2 of 12 participants) and number two by 83 percent (10 of 12 participants), whereas the L&H product was ranked number three by 100 percent (12 of 12 participants). Table 2 | ANOVA Results for the Progress Note and Discharge Summary Table 3 | Mean Error Rates and Percentage of Misrecognized Words in Each Category and Across Products for the Progress Note Dictation Table 4 | Mean Error Rates and Percentage of Misrecognized Words in Each Category and Across Products for the Discharge Summary Dictation Discussion : Results of this study suggest that, in generating medical record entries, the out-of-box performance of IBM ViaVoice 98 is better than that of software developed by Dragon and L&H; however, recognition rates may vary, depending on speech domain. Specifically, across both dictation samples, the IBM product was superior to the other two products in the recognition of general English vocabulary, medical abbreviations, and numbers. Findings for medical vocabulary and punctuation were not consistent across test dictation trials, so conclusions about the relative performance of the three packages for these two error categories are limited. Surprisingly, previous experience with dictation services did not have an effect on the overall error rates. This result may be due partly to two factors: 1) participants had some time to practice dictation of the two practice notes prior to the scored dictation trials, and 2) participants were reading from a script and thus did not experience some of the slowing and stopping that is typical of inexperienced dictators as they learn to compose notes as they dictate. This latter point may have also had some effect on recognition rates across the three packages. Specifically, recognition rates may have been enhanced by the absence of pauses and hesitations that would have been present if the dictators were required to compose medical entries instead of reading from a script. Also notable was the high rate of recognition errors found in the use of technology that is more advanced than discrete speech recognition. Error rates reported in studies examining discrete speech recognition software have ranged from 1 to 3 percent, which is significantly lower than the 6 to 8 percent reported in the present study for the most accurate of the three packages evaluated. Based on the results of Zafar et al., however, it is reasonable to conclude that extended training of the speech model, extended training of the dictator in package-specific dictation conventions, and the addition of vocabulary not contained in products as shipped would improve the recognition rates substantially. The first methodologic challenge of this study was to develop a scoring protocol that would ensure the consistency of scoring across packages within a subject and across subjects within the sample. The primary strategy used in developing this protocol was to limit the degree of interpretation that was needed in scoring an error. This strategy was adopted because interpretive scoring protocols are difficult to standardize and are threatened by any potential interpretive biases. In pursuit of this strategy, we developed some simple rules for scoring, which did not allow the evaluator to score dictations on the basis of an interpretation of what the evaluator thought went wrong in the dictation. Although a rigid application of these rules removed the guesswork and potential errors in scoring, these rules did not allow us to develop a better understanding of the types of errors that may be inherent in the speech models of the software evaluated, and they may have contributed to higher absolute error rates. Some methodologic challenges were difficult to address because of time and cost constraints. In this study design, for example, we were unable to determine whether words missing from the dictation sample were missing because they were misrecognized by the software or because the dictator omitted them. In fact, in several dictation samples, it was clear that the dictator had omitted a word or phase in preference for some other style of presenting the information. Scoring of these instances, however, was not altered to take account of the dictator 's missed words, as our scoring methodology was designed to limit the degree of interpretive bias that might be present. With greater resources, it might have been possible to audiotape the dictators' speech, have that audiotape transcribed to a typed report, and evaluate the recognition errors for the dictation sample on the basis of the transcribed report. Having 64 dictation samples transcribed, however, was beyond the scope of our budget. Although this may seem to be a serious source of error in our findings, the study was designed so that the occurrence of this type of error should be distributed evenly across software packages. With advances in technology since the time of this study, several improvements in the present methodology have become possible. Future studies should make use of digital voice recorders that are designed to interface with speech recognition software. Use of this technology would allow each participant in the study to dictate each report only once and then use the same speech file with each software package. This would eliminate many potential confounders of a within-subjects design (e.g., practice effects, fatigue) and it would also be more economically feasible to have the digital speech file transcribed for direct comparison with the dictation sample. Conclusions : With increasing power, decreasing cost of computing hardware, and recent increases in the sophistication of speech recognition software, the use of speech recognition to replace transcription in real-world settings is finally becoming feasible. Accuracy of recognition, which has traditionally been a major problem, has increased dramatically because of improvements in speech recognition technology and availability of medical vocabularies. As a result, it is becoming possible to think of productively employing off-the-shelf speech recognition products in clinical settings. However, a thorough understanding of the business problems that need to be solved and the technical and functional attributes of available products is essential to any successful undertaking. The study discussed here is important because it represents one of the veryfew attempts at a robust evaluation of performance attributes of commercial speech recognition products. Although technology will continue to evolve, and products will come and go, knowledge of the approach taken in this study should be of value both to health care organizations considering speech recognition implementations and to researchers contemplating further investigation of speech recognition technology. Backmatter: PMID- 10984466 TI - Variables That May Enhance Medical Students' Perceived Preparedness for Computer-based Testing AB - AbstractObjective: To identify variables that may enhance medical student's preparedness for computer-based administration of the United States Medical Licensing Examination (USMLE). Design: A cross-sectional survey of 301 medical students who completed a self-administered questionnaire. Measurements: The questionnaire was designed to obtain information about students' computer resources, personal experience with computers, computer expertise, opinions about computers, experience with computer-based testing, perceived preparedness for the computer-based USMLE, and demographic variables. Variables related to students' perceived preparedness for the computer-based USMLE were identified by ordinal logistic regression. Results: A significant regression model yielded four significant predictors: perceived preparedness for USMLE content (P < 0.0001), opinions about computers (P < 0.0012), gender (P < 0.0001), and a gender by computer-based testing experience interaction (P < 0.0004). Computer resources, personal experience with computers, computer expertise, age, race, and year of medical school were not significant predictors. Conclusion: Students' perceived preparedness for computer-based administration of high-stakes examinations may be facilitated by preparing them for examination content, by enhancing their opinions about computers, and by increasing their computer-based testing experiences. Keywords: Introduction : Computer-based testing (CBT), defined as the use of computers to administer tests, is now used in highstakes examinations such as the Graduate Record Examination and nursing licensure examinations. Instant scoring, flexibility in scheduling, and enhanced security are some of the advantages of CBT. Computer-based versions of Steps 1, 2, and 3 of the United States Medical Licensing Examination (USMLE) have just begun to replace paper-and-pencil forms. Recent research suggests that most North American medical students have access to, and have used, computers in some capacity., From 1994 to 1999, there have been increases in the number of graduating medical students in the United States who reported that adequate time was allocated toward instruction in the use of computers during medical school. In addition, more than half the respondents indicated that computer-based programs had been a part of course instruction. The transition to computer-based administration of the step examinations, however, has prompted questions about the differential impact of paper-and-pencil versus computer-based testing on test performance. These differential effects are referred to as mode effects. Furthermore, there is some concern that mode effects may be related to examinee characteristics such as computer experience, opinions about computers, and demographics. While speeded tests have stringent time limits to measure processing speed, power tests are designed to measure examinee knowledge of a specific content area. Providing unlimited time for examinees to take the USMLE is impractical; therefore, time limits are imposed for this power test. Research suggests that computer versions of timed power tests can be constructed to measure the same trait as corresponding paper forms. A meta-analysis of 21 studies that examined timed power tests completed by young adults and adults revealed a correlation of 0.97 between these modes of administration. The mean and standard deviation of the scaled differences was -0.03, which indicated that the computer forms tended to be slightly harder than the paper forms, with relatively small variation from study to study. The American Psychological Association has guidelines for the development, use, and interpretation of computerized tests. These guidelines indicate that paper and computer versions of the same test may be considered equivalent if the rank order of scores of individuals tested in both modes are similar and if the means, dispersions, and shapes of the score distributions are approximately the same. To date, field tests of computer versions of the USMLE indicate that CBT does not have a differential effect on test performance. Research into the relationship between mode effects and examinee characteristics is less conclusive. It appears that neither level of computer experience nor anxiety about using computers affects performance on computer tests compared with paper tests., On the other hand, findings with regard to mode effects due to gender, age, and race have been mixed. Results obtained from 1,114 examinees who completed computer and paper versions of the Graduate Record Examination (GRE) indicated that on the analytic scale, Asian examinees performed better than their paper scores predicted on the computer version, compared with African-Americans. With regard to the verbal scale of the GRE, nonwhite male examinees performed better on the computer version than their paper scores predicted, compared with nonwhite female examinees. On the same scale, younger African-American examinees performed better on the computer version than older African-American examinees. With regard to the quantitative scale, the age pattern held, and the overall performance of male examinees on the computer version was better than would be expected from their paper scores, whereas female examinees performed better on the paper version. As worldwide interest in CBT increases, it would be useful to know how to help medical students prepare for this transition. To date, variables that may influence student preparation for CBT have not been reported in the medical education literature. The specific purpose of this study was to examine the relationship between perceived preparation for computer-based administration of the USMLE step examinations and the following predictor variables: computer resources, computer expertise, personal experiences with computers, experience with CBT, opinions about computers, perceived preparedness for USMLE content, and demographics. Method : Design and Instrument | A cross-sectional survey of 301 students who attended a public medical school in the southeastern United States was conducted. First-, second-, third-, and fourth-year students completed a self-administered questionnaire designed to obtain information about several aspects of computer use. Information about personal computer resources and computer expertise at the time of entering medical school was assessed by summing four and five differentially weighted experiences, to yield composite "resource" (e.g., own a computer, have an e-mail address) and "expertise" (e.g., self-taught, degree in computer science) variables, respectively . Specific computer experiences were assessed by the sum of 17 differentially weighted experiences to yield a composite "experience" variable. Experience with CBT was the sum of four equally weighted experiences, and opinions about computers was the sum of scores obtained from the Opinions about Computers scale. The Opinions about Computers scale consisted of eight items derived from two reliable and valid instruments, the Computer Opinion Survey and the Computer Anxiety Scale. It used a five-point Likert scale ranging from "strongly disagree" to "strongly agree." An internal consistency reliability coefficient (Cronbach alpha) of 0.79 was observed in this study, compared with previously observed alphas of 0.94 and 0.95 for the Computer Opinion Survey and the Computer Anxiety Scale, respectively. Information about participant preparedness for completing the USMLE on computer and preparedness for USMLE content was obtained from the questionnaire, as were demographic data including student gender, age, ethnicity, and year in medical school. Degree of preparedness was assessed on a five-point scale ranging from "very unprepared" to "very prepared." Data Analysis | Based on a review of the literature, the following predictor variables were identified: computer resources; expertise; experiences; CBT experience; opinions about computers; perceived preparedness for USMLE content; gender; age, race, year of medical school; and interactions between gender, age, race and personal computer experiences, CBT experience, and opinions about computers (nine interactions). Degree of perceived preparedness for completing the USMLE step examinations on computer was the criterion variable. A stepwise ordinal logistic regression model (SAS System for Windows, version 6.14) was used to analyze the data. An alpha level of 0.05 was the criterion for statistical significance of the relationship between the criterion and predictor variables and individual predictor variables. Table 1 | Computer Resources, Expertise, and Experiences of Medical Students Results : Eighty-seven percent (69 of 79), 87 percent (62 of 71), 99 percent (74 of 75), and 82 percent (62 of 76) of first-, second-, third-, and fourth-year students, respectively, completed the questionnaire. Male students made up 52 percent of the group, and the mean ages of the first-, second-, third-, and fourth-year students were 26, 27, 27, and 29 years, respectively. Eighty percent were white, and 20 percent were African-American, American Indian, Asian, or Hispanic. With the exception of race (there are proportionally more under-represented minority students at this school), these demographics are comparable with those of students in other U.S. medical schools. Most respondents owned a computer (70.4 percent), had access to a computer (98.5 percent), had an e-mail address (94.8 percent), and were connected to the Internet (48.3 percent) . With regard to specific computer experiences, the data were consistent with national findings. The majority of students had used e-mail (97 percent), educational software (75 percent), and conducted online literature searches (88 percent). Twenty-seven percent had used computers to do statistics. Similarly, 96 percent, 79 percent, 85 percent, and 29 percent of students nationwide have experience using e-mail, using educational software, conducting online literature searches, and using computers to do statistics, respectively. The proportional odds assumption, required for the ordinal logistic regression model, was tested for goodness of fit and met (psi2 = 4.0087, P = 0.2605). The relationship between the criterion and predictor variables (P < 0.0001) was significant. Significant predictors of perceived preparedness for completing the USMLE on computer were perceived preparedness for USMLE content (P < 0.0001), positive opinions about computers (P < 0.001), gender (P < 0.0001), and a gender by CBT experience interaction (P < 0.0001) . Computer resources, expertise, experiences, age, race, year of medical school, and the interactions of perceived preparation for USMLE content, CBT experience, and opinions about computers with age and race were not significant predictors. Bivariate correlation analyses among the predictor variables indicated very few substantial associations. While year of medical school and computer experience were significantly related to four other variables, exclusion of these two variables from the analysis did not change the statistical significance of the remaining variables or the final regression model. Table 2 | Results of Stepwise Ordinal Logistic Regression Analysis to Predict Preparedness for Computerized Administration of the U.S. Medical Licensing Examination (USMLE) Discussion : This study extends previous research by examining variables related to medical student preparedness for the computer-based USMLE. Such information is very relevant to test-taking skill development and to designing medical informatics curricula, now considered a requisite component of medical education. The results indicate that students who felt more prepared for the computer-based version of the USMLE step examinations were also more likely to feel more prepared for USMLE content, be male, be female with CBT experience, and have more positive opinions about computers. The findings of this study are consistent with research that indicates a weak relationship between CBT performance and computer expertise. At first glance, the results seem incongruent with studies that suggest that mode of test administration (i.e., paper versus computer) is not influenced by opinions about, or anxiety toward, computers., It has been suggested, however, that these variables might become influential as computer tasks become more complex,, a phenomenon labeled "situation-specific anxiety." Because it is a licensing examination for physicians, the USMLE is both a high-stakes and complex computer task. The specific anxiety associated with preparing for it could be characterized, therefore, as "USMLE anxiety." Thus, students who do not feel prepared for USMLE content, and who have negative opinions about computers, may experience additional anxiety that makes them feel relatively more unprepared for the computer-based versions of the USMLE. Previous research suggests that mode effects due to demographic variables are dependent also on the nature of computer tasks. Although gender did not interact with mode of test administration for relatively low-stakes tests such as psychological assessments, gender differences in perceived self-efficacy regarding complex computer tasks were evident in other studies., It may be possible, however, to decrease gender differences by preparing female students for complex computer tasks. Consequently, female students with more experience in relevant computer tasks might feel more prepared for high-stakes, complex computer tasks, which was observed in the present study. Specifically, female students with CBT experience were more likely to feel prepared for the computer-based USMLE. The results suggest variables that could be examined to identify students who may be apprehensive about the transition of high-stakes examinations to CBT and, consequently, allow schools to tailor interventions for these students. The findings indicate that medical schools can prepare students for computerized administration of the USMLE by readying them for USMLE content and by increasing student exposure to CBT. The latter may be achieved by increasing the integration of CBT into basic science and clinical courses. Current data indicate that 31 percent of medical students nationwide experience CBT as part of the medical education curriculum. There is, therefore, ample room for improvement in this area. Although the external validity of this study is limited by the use of a single medical school, this concern is mitigated by similarities in demographics and computer experiences between the sample of students in this study and the national population of medical students. At the same time, we recognize that each school has its own culture, and additional studies would be useful to confirm the generalizability of the findings. The present study was concerned with assessing a subjective variable that could be measured only by respondent self-report. The psychology literature suggests that self-report data can be reliable and valid under the following conditions: when an emotional state is being assessed (e.g., how do you feel?) rather than a request for causal attribution (why do you feel this way?); when the questions are not threatening, which may not be the case, for instance, when one is being asked about sexual behavior; and when questions are simply stated. We believe that the item "To what extent do you feel prepared to take the USMLE on computer?" meets these criteria. The psychology literature also suggests a strong positive link between perception of ability to perform specific tasks and subsequent performance. This appears to be the case for computer use. In addition, the medical education literature suggests positive relationships between medical students' perceptions of skills and evaluations and those rendered by more objective benchmarks. With regard to students' reports of readiness, for instance, Fincher et al. found a correlation of 0.58 between reported readiness for internship training, as assessed by a single item, and subsequent evaluation by internship directors. Furthermore, positive relationships between medical students' perceptions of their clinical skills and ensuing clerkship performances have been reported.,, Finally, we believe that the self-report data obtained from the participants in the present study are conceptually sound. To ascertain the latter, data obtained from the fourth-year students in our study were compared with responses by the same students to items on the Association of American Medical Colleges (AAMC) Graduating Questionnaire. In terms of the relationship between actual computer experiences and confidence in having specific skills and knowledge, there were several consistencies between the information obtained via the questionnaire used in our study and that used by the AAMC. Ninety-seven percent, 90 percent, and 26 percent of students in our study stated that they had used e-mail, done an online literature search, and done statistics on computer, respectively. On the AAMC Graduating Questionnaire, 93 percent, 96 percent, and 24 percent of the same cohort of students indicated that they were "confident they had the appropriate knowledge and skills to, a) use electronic mail, b) carry out reasonably sophisticated searches of the medical education databases, and c) use a statistical software package," respectively. Evidence of consistency between students' reported confidence in having specific computer skills and knowledge in conjunction with their reported computer experiences suggests that students would be consistent, too, in reporting their perceived preparedness for computer-based testing of the USMLE and their actual preparedness. A multi-item scale would probably be a more desirable approach to assessing the outcome variable, but we found no precedent in the literature for a scale with enough questions to improve reliability and validity given the specific nature of the outcome variable (i.e., perceived preparedness for taking the USMLE by computer). In conclusion, this study indicates that some students do not feel prepared for the computer-based version of the USMLE. However, implementation of activities designed to increase preparation for examination content, improve opinions about computers, and increase CBT experiences may help to mitigate this perceived lack of preparedness. Backmatter: PMID- 10984467 TI - Exploring Performance Issues for a Clinical Database Organized Using an Entity-Attribute-Value Representation AB - AbstractBackground: The entity-attribute-value representation with classes and relationships (EAV/CR) provides a flexible and simple database schema to store heterogeneous biomedical data. In certain circumstances, however, the EAV/CR model is known to retrieve data less efficiently than conventionally based database schemas. Objective: To perform a pilot study that systematically quantifies performance differences for database queries directed at real-world microbiology data modeled with EAV/CR and conventional representations, and to explore the relative merits of different EAV/CR query implementation strategies. Methods: Clinical microbiology data obtained over a ten-year period were stored using both database models. Query execution times were compared for four clinically oriented attribute-centered and entity-centered queries operating under varying conditions of database size and system memory. The performance characteristics of three different EAV/CR query strategies were also examined. Results: Performance was similar for entity-centered queries in the two database models. Performance in the EAV/CR model was approximately three to five times less efficient than its conventional counterpart for attribute-centered queries. The differences in query efficiency became slightly greater as database size increased, although they were reduced with the addition of system memory. The authors found that EAV/CR queries formulated using multiple, simple SQL statements executed in batch were more efficient than single, large SQL statements. Conclusion: This paper describes a pilot project to explore issues in and compare query performance for EAV/CR and conventional database representations. Although attribute-centered queries were less efficient in the EAV/CR model, these inefficiencies may be addressable, at least in part, by the use of more powerful hardware or more memory, or both. Keywords: Introduction : A problem that data modelers commonly encounter in the biomedical domain is organizing and storing highly diverse and heterogeneous data. For example, a single patient may have thousands of applicable descriptive parameters, all of which need to be easily accessible in an electronic patient record system. These requirements pose significant modeling and implementation challenges. One increasingly popular solution that addresses these issues is to model data using the entity-attribute-value (EAV) approach. This model, which was historically seen first in LISP association lists, has been used in multiple applications, including the HELP system, and the Columbia-Presbyterian Clinical Data Repository., The EAV approach offers many advantages, including its flexibility and ability to store heterogeneous data in a simple, easily maintained format. Our group has recently enhanced this representation to permit data modeling with classes and relationships (EAV/CR). Despite its significant benefits, EAV design has the potential to be less efficient than "conventional" database schemas when accessing data. In particular, attribute-centered queries, where the query criterion is based on the value of a particular attribute, are most likely to show impaired performance. This is especially true when query criteria combine one or more simple conditions in Boolean fashion. An example of such a query is "find all patients with blood cultures positive for either Streptococcus pneumoniae OR Candida albicans." The reason for the potential performance degradation is that the relatively fast AND, OR, and NOT operations that would be required if we were operating on conventional schema tables must be converted to considerably slower set-based equivalents (set intersection, union, and difference, respectively) for EAV tables. Attribute-centered queries are important for research questions; their performance is not critical for the care of individual patients. The contents of the production (patient-care) database serve, however, as the basis for any research databases that an institution must support, and the question is how such research databases must be implemented. There are two ways to do so: The first solution is easy to set up, but queries may not perform well. The second solution promises better performance but is more elaborate because of the data transformations involved. The required transformations can be extensive if one has to support multiple research efforts. To determine which solution is applicable in given circumstances, it would be useful to have performance metrics that compare the relative performance of attribute-centric queries on EAV schemas and equivalent queries on conventional schemas. We have not found any published reports that quantify these performance differences systematically. Of related interest are the influences on query performance exerted by factors including database size and hardware configuration (e.g., amount of available physical memory). The questions addressed in this paper are: Answers to these questions would provide valuable insights into the viability of the EAV approach for supporting large-scale databases. In this paper, we explore issues in database performance for real-world microbiology data stored using an EAV/CR and conventional approach. In particular, we examine: 1) the differences in execution time for a set of clinically relevant database queries operating on an EAV/CR and conventional schema containing the same set of data, 2) the relative efficiency of different EAV/CR query strategies, and 3) the effect of database size and physical memory on database performance. Finally, we examine monitors of system performance and identify potential performance bottlenecks for the competing database schemas. Background : The EAV model, which has been described in great detail elsewhere,,, can be visualized conceptually as a database table with three columns: 1) "Entity ID" or "Object ID"; 2) "Attribute," or a pointer to a separate "Attributes" table; and 3) "Value," containing the value corresponding to the particular entity and attribute. In an electronic patient record system, the entity ID usually refers to a specific patient-associated event stored as a patient ID and time stamp. The attribute and value columns correspond to a specific parameter (e.g., potassium) and its value, respectively. Each entity-attribute-value triplet occupies one row in the table. Since attributes are not represented as columns, this single table can store all values in the database. EAV/CR enhances the basic EAV approach by using the object-oriented concepts of "classes" and "relationships" to facilitate data modeling. Classes in EAV/CR are data structures that store the attributes (possibly including other classes) associated with a particular type of entity. For example, a bacterial isolate from a blood culture could be modeled with the class "Bacterial Organism" that contains attributes "Organism_ID" (which references an organism name), "Quantity_cultured", and "Antimicrobial_tested." (The last attribute is multivalued, because multiple antimicrobials are typically tested against a single culture.) "Antimicrobial_tested," in turn, is itself a class with the attributes "Antimicrobial_ID" (which references a drug name) and "Sensitivity_result." As in object-oriented modeling, objects represent instances of a class. Thus, in our previous example, every cultured bacterial isolate possesses an object ID as an instantiation of class "Bacterial Organism." Relationships are used to describe inter-class interactions. Metadata, or information about the data contained in the database, is then applied to describe, maintain, update, and access the data. This overall approach allows the system to incorporate new attributes without altering the underlying physical database schema. Conventional schemas, in contrast, model parameters as distinct columns. For example, the parameter "Organism_name" could be represented as a column in a table "Microbial_Organism." As the database evolves to accommodate diverse data, the number of parameters grows and the number of columns and tables that are needed increases. Thus, with a rapidly evolving and expanding domain such as medicine, where new tests and concepts are commonplace, the underlying schema requires frequent modification. This is particularly true of clinical medicine, where there can be dozens of specialty-specific tables, each with a constantly growing list of fields. The ability of the EAV approach to store diverse data in a single (or few) tables greatly simplifies the physical schema of the database. It provides an efficient way to access all data pertinent to a specific object (i.e., entity-centered queries). In an EMR, this often takes the form of "Select all data (clinical findings, lab tests, etc.) for patient X." In a conventional schema with numerous tables, the user would need to search every table containing patient data to extract the relevant data for the patient, making query development laborious. Furthermore, not all these tables may contain data for a given patient; as a result, many tables may be searched unnecessarily. As stated previously, it is the attribute-centered queries that are of concern to the developer who is focused on efficiency. To systematically quantify the performance of an EAV/CR schema compared with its conventional counterpart, we chose the domain of microbiology. This domain, which has been the subject of many previous database models,, provides the potential to work with complex data structures where class attributes exist as both primitive types (e.g., strings) as well as other classes. Methods : In this section, we describe our test database and schema and various aspects of the evaluation strategy that we devised. In each case we provide the rationale of our design decisions. Data Description | Each specimen collected from a patient in the Veterans Administration (VA) system is assigned a unique ID. The ID of the ordering physician is also recorded against the specimen. One or more tests (e.g., Gram stain and culture) may then be run on each specimen. For each culture, a panel of antimicrobial sensitivity tests is performed. The data are stored in DHCP, an M language database that serves as the clinical patient repository used nationwide by VA Medical Center (VAMC) hospitals. The entity-relationship diagram for the original data is shown in (details have been omitted for clarity). All classes/entities can be traced to the class "Microbiology Specimen," which sits at the "top" of the class diagram. Classes contain either primitive types or other classes as attributes, which enable classes to relate to one another. It should be noted that the entities "bacterial culture," "fungal culture," etc. shown in are represented in separate tables in DHCP. During the process of data transformation, we merged these into a single table, since the separate tables have similar structures. The original, raw, antimicrobial sensitivity data had the names of antimicrobials (e.g., "streptomycin," "chloramphenicol") hard-coded as columns in a table. This represents poor table design. With each new antimicrobial introduced into the institution, the table structure needs updating, and all forms dependent on the table require redesign with each change. It may be possible to use this design in the M language, where all disk-based data structures are sparse by design. In relational database management systems (RDBMSs), however, such a design wastes much space, because NULL values of empty columns are still recorded. Further, RDBMSs have a limit on the maximum number of columns per table (e.g., 255), which might be exceeded in some circumstances. Even if space and column constraints are unimportant, necessitating alterations in the table structure and application code each time the number of objects changes is not a good system design approach. Other parts of the DHCP schema are correctly designed. For example, in the Pharmacy subschema, names of medications are not hard-coded as columns but treated as data: the ID of the drug in the Orders table references a list of drugs in a Drugs table. Therefore, to make comparisons between EAV and conventional schemas more realistic, we first transformed the sensitivity data into the (correct) row-modeled form, where the antimicrobial ID was an attribute in a column and referenced a table of antimicrobial drug names. Data Extraction | We extracted a dataset from VA Connecticut DHCP that contains all the available online data, which range from 5/1/87 to 9/26/98 and include results from more than 135,000 microbiology test specimens and more than 400,000 antimicrobial sensitivity tests for more than 28,000 patients. The extracted data were transformed for storage into a SQL Server 7.0 RDBMS. The extraction was done in two steps. We first created a conventional schema from the data (transforming the data where appropriate, as described above). We then created an EAV/CR schema from the conventional one by assigning a unique object ID to each instance of a particular class. In all, there were 965,529 instantiated objects in our database. The Conventional Schema | The conventional schema is shown in . All primary and foreign keys, as well as fields used for search (e.g., antimicrobial and organism names), are indexed. It should be noted that the culture and antimicrobial sensitivity tables can be considered special-purpose or homogeneous EAV tables (because each stores only a single class of data in row-modeled form). In contrast, "EAV" databases contain general-purpose tables that store heterogeneous data (across many classes). The EAV/CR Schema | The EAV/CR physical schema is shown in . As in all EAV/CR schemas, there are two parts to the schema: metadata and data. (The EAV/CR metadata schema are not illustrated. They have been described in detail, and a summary is available on the Web at ). Attributes are segregated by data type (e.g., string values are stored in "EAV_TEX,T" whereas dates are held in "EAV_DATE"). The table EAV_OBJECTS is of particular interest. It accommodates situations where one object can contain another object and permits objects to be associated with one another. In this table, the "Value" field is the ID of a "child" object. Thus, in our logical schema, tests are children of a specimen, cultures are children of a test, and antimicrobial sensitivity results are children of a culture. In other words, when data are hierarchic (as in consecutive one-to-many relationships), each record in EAV_OBJECTS represents one node of a tree. To traverse down an object hierarchy starting with a particular object, we locate that object's ID in the Object_ID column of EAV_OBJECTS. We then gather all Values (child Object IDs) for this entity. We then recurse, searching for these IDs in the Object_ID column, and so on. One well-known method of minimizing the number of recursive queries when traversing an object hierarchy is to redundantly store the ID of the "ancestor" object against every record in a tree, so that all descendants of the ancestor can be retrieved in a single SQL statement. If a significant percentage of queries also use the "ancestor" class, the number of joins required to retrieve the data is also minimized. To allow us to explore this means of schema optimization and its impact on queries, we created an "ancestor" field containing the microbiology specimen ID associated with each test value in EAV_TEXT. We indexed all object ID columns and created a compound index on attribute-value pairs for tables EAV_TEXT and EAV_DATE, since queries typically select data using both an "attribute ID" and "value" together as search criteria. In addition, a separate index was created for the Value column of EAV_OBJECTS, since the nature of a query might make it desirable to traverse an object hierarchy from descendant to ancestor. Expanding the Size of the Data | To measure the effects of database size on performance, we replicated the contents of our initial two databases (EAV/CR-Small and Conventional-Small) and reassigned newly created, unique IDs for each new object. The replicated data were then appended onto the corresponding, original database to create "new" databases with twice the number of objects (EAV/CR-Medium and Conventional-Medium). This process was then repeated to yield databases four times larger than our originals (EAV/CR-Big and Conventional-Big). Since the underlying schemas did not change, queries did not require modification within a given representation. The "Big" data set represents the equivalent of 30 to 40 years' worth of microbiology data for all patients in VAMC Connecticut. The benchmark timings for EAV/CR-Medium and Conventional-Medium were intermediate between the smallest and largest representations, with the data showing an approximately linear trend between the "small" and "big" databases. The purpose of studying three database sizes was to allow us to assess better how relative performance varied with size. To minimize information overload in the tables presented in the remainder of the paper, we omit further discussion of the two "medium" databases. The EAV/CR representation consumed approximately four times the storage of our conventional schema. EAV/CR-Big consumed 1,177 Mb, with indexes accounting for more than 62 percent. Our largest conventional representation (Conventional-Big) was 301 MB, with approximately 32 percent accounted for by indexes. It is true that EAV/CR is more space-efficient for sparse data. In this case, however, prior to importing the raw data, we had transformed the bulk of the original sparse data (e.g., the antimicrobial sensitivity results and the cultures) into dense, row-modeled facts in columns that are mostly IDs (long integers). It is well known that a row-modeled conventional table will always take significantly less space than the equivalent facts represented in EAV form. This is because a single set of facts is represented as a single row in a row-modeled table but as multiple rows in an EAV table, one per fact. For example, to fully describe a single antimicrobial sensitivity result, we store several related facts linked to the specimen ID: the micro-organism that was isolated, the antimicrobial tested, and the sensitivity of the former to the latter. In the EAV representation, each row has the extra overhead of Object ID and Attribute ID, plus accompanying indexes. Query Benchmarking | To compare performance between the two competing physical representations, we developed several clinically oriented queries, as follows. The two attribute-centered queries were: System Description: Memory Allocation | Tests were conducted on a Dell Dimension XPS D300 running Windows NT Server 4 with a 300-MHz Pentium II processor and 66-MHz system bus. Execution times were recorded using the SQL Server 7.0 Profiler tool. We ran our benchmarks with 192 Mb and 256 Mb of physical RAM and allocated 64 Mb and 128 Mb, respectively, to SQL Server 7. We did this to ensure that the additional RAM was specifically allocated to the database. This left 128 Mb of RAM for the operating system and its associated tasks. The system employed a Western Digital UDMA, 9.5mS, 5400 RPM hard drive. We want to emphasize, however, that in our benchmarks it is the ratio of timings between the EAV/CR and conventional schemas that are important, rather than the absolute timings. Alternative Formulation of EAV/CR Queries | Formulation of a complex query against an EAV/CR database is significantly more difficult and error-prone than is formulation of a functionally identical SQL query against an equivalent conventional schema. This is because the physical schema of an EAV/CR database ---the actual tables holding the data ---differs markedly from its logical schema, which specifies how the classes modeled in the database are related to each other. For example, when considering Query 1, the "culture," "organism," and "patients" tables do not physically exist in the EAV/CR schema. In a conventional schema, in contrast, logical and physical schemas are identical. Queries against an EAV/CR system must, therefore, be formulated in terms of the semantics of the logical schema and must then be translated into operations on the tables in the physical schema. We present four ways to do this, the third and fourth being slight modifications of the second, as follows: Method One | A single (large) SQL statement can be created, which extracts the desired data by joining the necessary tables. In an EAV/CR schema, however, a small set of physical tables is used to model many more logical tables. This results in the same tables being used repeatedly under different aliases, with "self-joins" being performed. In , which illustrates this approach for Query 1, the table EAV_TEXT is used twice. As the logical schema becomes more complex and more classes have to be traversed, such SQL gets progressively harder to write and debug, because it often incorporates a large number of nested constructs. To complicate matters, when this SQL is sent to an RDBMS, the latter parses the code and tries to formulate an optimal strategy before executing it. Query optimization is a problem known to be factorial with respect to the number of tables or aliases participating in the join. (Factorial 8, or 8!, equals 8 x 7 x 6... x 2 x 1 = 40,320. An excellent discussion of the query optimization problem is posted at .) Difficulties manifest, therefore, as the number of aliases increase. While DBMS optimizers are claimed by their vendors to use clever heuristics in such a situation, simply evaluating each of the numerous alternatives takes time. (In an experiment reported by Nadkarni and Brandt, an attempted 20-alias join that would have yielded only 50 rows in the result set caused an RDBMS to freeze.) The only justifiable circumstances for using a single-statement approach for EAV/CR query are for queries with a modest number of aliases (as in ), or for "stored" queries, a facility available in many DBMSs. (In execution of stored queries, the parsing phase is bypassed, because the query has already been "compiled," so to speak, with a plan of execution already worked out.) Method Two | The second method uses a "divide and conquer" strategy. A series of simple queries are specified in terms of individual classes in the conceptual schema. Each query accesses one or two tables at a time to create a temporary table. The temporary tables are then combined with joins, using appropriate set operations (intersection, union, difference), if necessary. When the final result is obtained, temporary tables created along the way are deleted. In this particular case, the DBMS optimizer has no problem with individual queries. In fact, the execution plan devised by optimizers often involves creating temporary tables, and we are explicitly telling the DBMS what to do at each step rather than relying on the DBMS to determine its own strategy. illustrates this method for Query 1. It is, of course, possible that, if a DBMS spends enough time analyzing a complex query, it will arrive at a solution superior to what we have manually specified. However, for highly complex ad hoc queries (unlike stored procedures), the time required to devise an execution plan may be much greater than the time required to actually execute it. Method Three | A slight modification of the second method is to repopulate and empty existing indexed temporary tables instead of creating and deleting them. The potential rationale is that if intermediate result sets created by a simple query are large, then, when the result sets are to be combined, the query optimizer will often create temporary indexes to speed up the join process. If we maintain several indexed tables for the sole purpose of holding intermediate results, then, when the tables are populated through INSERT queries, the indexes are also populated, and subsequent joins will be speeded up. illustrates this approach. Notice the very great similarity to the code in . The only differences are that we use INSERT INTO... SELECT instead of SELECT... INTO, and TRUNCATE TABLE instead of DROP TABLE. This strategy, however, is double-edged. If intermediate results are small, then the DBMS may perform joins in memory rather than on disk. If so, we have created disk-based indexes for nothing. Index maintenance uses machine resources and time, because whenever rows are removed or added to a table, the indexes on the table must also be maintained. Method Four | A minor modification of the third method is to use initially un-indexed temporary tables, which are indexed only after all records are inserted. Record insertion is expected to be faster because the indexes do not have to be maintained each time a new record is added to the corresponding table. The query is illustrated in . We applied all four methods to Query 1 operating on EAV/CR-Big with 128Mb RAM allocated to the database. We found that our second method (developing simple queries that created tables and later removed them) gave the best results. Measuring Caching Effects | When the exact same query is sent to a DBMS multiple times, the execution times for the second and subsequent runs are often significantly shorter than for the first. This is because the DBMS uses caching, either in memory or on disk. The SQL string of a query is stored (so that the text of a new query may be compared with it), along with the execution plan. If the result set is small enough, it may also be cached. Therefore, in an extreme case, a repeated query will not cause any disk activity at all: the DBMS, having decided that it is identical to a previous one, will instantaneously return a stored result. It is important to determine caching effects for conventional versus EAV/CR schemas, especially if queries on the latter consist of a batch rather than a single statement. (All the individual queries in a batch might not be cached.) To measure caching effects, each query was executed five times in succession immediately following system start-up for each combination of system memory and database size. We used five runs to control for "cold" database, or start-up effects. We report the timings for every query result in two ways: as initial query execution time and as cached execution time (the average of the four succeeding runs). The former probably simulates the real-world situation more closely than the latter, since it is less likely that two different users of the system will perform successive identical ad hoc queries. Utilizing the Ancestor Field | We also compared execution times for queries that made use of the ancestor field construct with execution times for those that did not. As stated previously, the use of the ancestor field allows the composer of a query to "cheat" and shorten the traversal path in suitable circumstances. It was necessary to quantify the benefit obtained while recognizing that queries may not always be able to take advantage of this field. Monitoring System Performance | Operating system performance was monitored with Windows NT performance monitor in conjunction with SQL Server Objects and Counters. We recorded the following parameters: Page read frequency is a surrogate measure of disk I/O (input/output) efficiency. A high percentage of hard disk activity and a large number of disk reads per second suggest a potential disk I/O bottleneck. Finally, a high degree of processor activity suggests a potential processor-related bottleneck. We performed these evaluations for Query 1 during initial execution (when there was no data caching) and the fifth run of the sequence of five (where caching effects were presumed to be most significant). SQL Server 7 and Windows NT provide system measurements every second; we averaged these values over the duration of the query's execution to obtain summary measurements. Figure 1 | Entity-relationship diagram for the microbiology data. Entity-relationship diagram for the microbiology data. For reasons of space, some tables, which are not relevant to the queries in the text, are not shown. Figure 2 | The conventional physical database schema. The conventional physical database schema. Fields in some tables are not shown, for reasons of space; for the most part, only fields linking to other fields are illustrated. Figure 3 | EAV physical database schema. EAV physical database schema. Since all EAV tables share the same structure, the details of two tables have been omitted. Figure 4 | Database query developed using a single SQL statement. Database query developed using a single SQL statement. In this and the next three figures, attribute ID 20 refers to organism name and ID 9 refers to patient ID. Figure 5 | Database query developed using temporary tables created and deleted dynamically to store interim data. Database query developed using temporary tables created and deleted dynamically to store interim data. Figure 6 | Database query developed using previously created and indexed tables to store interim data. Database query developed using previously created and indexed tables to store interim data. Figure 7 | Database query developed using un-indexed tables to store interim data, which are indexed only after data are inserted into the tables. Database query developed using un-indexed tables to store interim data, which are indexed only after data are inserted into the tables. Results and Interpretation : Comparing Different EAV/CR Query Strategies for Attribute-centered Queries | As stated earlier, there are at least four different strategies to use when performing attribute-centered queries of EAV/CR data. Prior to contrasting the EAV/CR and conventional schemas for different memory configurations and database sizes, we had to decide which of these strategies was most viable. We report the comparative benchmarks with each approach for Query 1 (finding all patients with cultures positive for P. aeruginosa). Similar results were observed when these strategies were applied to Query 2 (find all specimens with S. pneumoniae and C. albicans). The time difference of 10 percent between the multiple-statement and single-statement approaches is not particularly dramatic. We hypothesized, however, that as the number of classes involved in a query increased, these differences would become increasingly pronounced. We tried out a new query, "Find all patients with cultures positive for P. aeruginosa that showed resistance to ceftazidime." (This query also involves the "Antimicrobial_Tested" class and "Antimicrobial Name" attribute.) The initial execution times for the multiple-statement and single-statement queries were 194 sec compared with 338 sec, respectively. In other words, using multiple statements almost halved execution time. (Our experience is in line with that reported in a well-known database article, which stated that giant, "elegant" SQL statements are often too complex for the limited intelligence of a DBMS optimizer.) Because of these results, we used the multiple statement approach, with creation of temporary tables, to benchmark EAV/CR query performance for the remaining analyses. Comparing Performance of the EAV/CR and Conventional Representations for Entity-centered Queries | Queries on the conventional and EAV/CR schemas gave comparable execution times: To summarize, increasing the volume of the data by a factor of four (in the "Small" versus "Big" databases) had little effect on entity-centered queries. This is expected, because for such queries, which use indexes well and return small amounts of data, search time tends to increase logarithmically rather than linearly with data size. The results indicate that, compared with the EAV/CR schema, the conventional schema may be marginally more efficient (if at all, since results for Query 4 are inconclusive), but the differences are not large enough to be of major practical importance. Comparing Performance of the EAV/CR and Conventional Representations for Attribute-centered Queries | Query 1: Find All Patients with Cultures Positive for P. aeruginosa . The conventional queries were three to five times faster than their EAV/CR counterparts in the initial run, and 2 to 12 times faster in the cached run. We analyze the results below. Increasing database size and system memory had little influence on the ratio for the initial run. With the smallest database, greater available RAM improved caching of EAV/CR queries, but this effect was lost for the largest database. Using the ancestor field in the EAV/CR query provided little or no benefit for the smallest database but reduced the time for the large database. Here, the benefits were most pronounced with 64 Mb of RAM (with time reduced by a third), but less so (only one tenth) as more RAM was added. The number of page reads per second was consistently larger in the EAV/CR schema, particularly as the database increased in size. This suggests that physical disk I/O is a likely bottleneck in this query operating on the EAV/CR model. Query 2: Find All Specimens Positive for S. pneumoniae and C. albicans . Here, the EAV/CR queries that used the "Ancestor" field were much faster than queries on the conventional schema: twice as fast with the initial query, six to eight times as fast when cached. Changing database size or increasing system memory had little effect on these ratios. When the ancestor field was not used, the EAV/CR query took longer, with the difference in execution times widening for initial runs as database size increased. With 64 MB of RAM allocated, the ratios were 3 for the small databases and 5.6 for the large databases. As with Query 1, however, the ratios decreased (to 3.2 and 3.3, respectively) when 128 Mb of RAM was allocated. Our results indicate that while EAV/CR is generally less efficient than its conventional equivalent, the inefficiency can be mitigated, at least in part, by additional memory. The dramatic benefits of the ancestor field (specimen ID) in Query 2, which uses this field specifically, should be put in perspective. Specifically, incorporation of this extra field in the schema amounts to deliberate denormalization, a standard space-for-time tradeoff. (We could also have done this in the conventional schema, by adding a specimen ID column redundantly to the cultures and antimicrobial sensitivity tables. In this case, the altered conventional schema would almost certainly have outperformed the EAV/CR schema considerably.) Our results do suggest that, if performance for certain types of commonly executed queries is critical in a particular database, judicious use of ancestor fields can exert more of an impact than the schema's underlying design (i.e., EAV/CR or conventional). In many electronic patient record systems, for example, almost every item of data is tagged with the ID of the patient to whom it directly or indirectly pertains, because a very large portion of queries relates to individual patients. Table 1 | Execution Times (in Seconds) for Query 1: "Find All Patients with Cultures Positive for Pseudomonas aeruginosa" Table 2 | Execution Times (in Seconds) for Query 2: "Find All Specimens Positive for Both Pseudomonas aeruginosa and Candida albicans" Discussion : We performed a pilot exploration of EAV/CR database efficiency issues that can help guide designers and programmers of EAV/CR databases, such as electronic patient record systems, as to the performance penalty they could expect if they tried to execute complex queries on the EAV/CR data directly, rather than exporting subsets of the data to a conventional database for querying purposes. However, the following limitations of our study must be noted. We examined performance monitors for a single database engine running on a single operating system platform. We chose this configuration largely because of the available built-in system monitoring tools. The results generated here may not completely map to other database engines and operating systems. (Identical queries conducted on the two schemas using a Microsoft Access 97 database engine running on Windows 98, however, yielded qualitatively similar results.) We cannot predict how performance would be affected by increasing the database to sizes larger than those tested here. Our results suggest, however, that the difference in performance appears to widen as database size increases for certain queries and that this can be offset, at least in part, by adding system memory. We did not make use of the parallel processing capabilities of the SQL Server database engine, since we employed a single processor machine for testing. Investigations are under way to study additional methods to optimize query generation and execution. Our databases were populated with data from one specific domain, microbiology. We cannot predict how performance times would differ in a database populated from increasingly heterogeneous sources of data. We would expect, however, that the effort needed to develop queries would increase disproportionately in the conventional schema compared with the EAV/CR schema. The logical schema of our test database may not be the best candidate for demonstrating the full potential value for an EAV/CR schema, since the data do not exhibit sparseness. The Windows NT operating system performs certain background tasks for system maintenance that could conceivably affect specific performance runs. We tried to minimize these effects by generously dedicating 128 Mb of physical RAM to the operating system at all times. These are the lessons we learned from this exercise: While attribute-centered EAV/CR queries are slower than queries on conventional schemas, this does not disqualify EAV/CR schemas from consideration for warehousing biomedical data. For a complex schema, the increased execution times are significantly offset by ease of database and query maintainability in the EAV/CR model. The conventional query's speed advantage (being three to five times faster) may appear to be discouraging for EAV/CR, but the fact is that the longest EAV/CR query in our fairly large data set took less than 3.5 min to complete. Continuing enhancements of CPU speed, increasing RAM capacity, and the availability of affordable multiprocessor machines will lower absolute times further, even though they would benefit conventional and EAV schemas equally. In any case, long query execution times tend to be far less critical for non ---real-time, attribute-centered queries, which are often submitted in batch mode. (Patient-centered queries typically submitted in an electronic patient record system, in contrast, demand quick response time.) Perhaps the most compelling justification for moving EAV/CR data subsets to specially designed, conventionally structured data marts is when users who demand response times of a few seconds repeatedly query these subsets. Conclusions : This paper describes a pilot project to explore issues in query performance for EAV/CR and conventional database representations. Although we found that attribute-centered queries performed less efficiently in the EAV/CR model, we feel that many of the benefits inherent in the EAV/CR representation help offset this decrease in performance. Purchasing more memory, additional processors, or faster hardware, or a combination of these, may prove a very cost-effective approach to handling these potential inefficiencies, particularly if the alternative is to maintain two parallel versions of a database with different structures. We plan to continue building and optimizing query strategies to support this representation as well as testing its ability to handle increasingly complex databases for alternative biomedical domains. Backmatter: PMID- 10984468 TI - GEM --A Proposal for a More Comprehensive Guideline Document Model Using XML AB - AbstractObjective: To develop a guideline document model that includes a sufficiently broad set of concepts to be useful throughout the guideline life cycle. Design: Current guideline document models are limited in that they reflect the specific orientation of the stakeholder who created them; thus, developers and disseminators often provide few constructs for conceptualizing recommendations, while implementers de-emphasize concepts related to establishing guideline validity. The authors developed the Guideline Elements Model (GEM) using XML to better represent the heterogeneous knowledge contained in practice guidelines. Core constructs were derived from the Institute of Medicine's Guideline Appraisal Instrument, the National Guideline Clearinghouse, and the augmented decision table guideline representation. These were supplemented by additional concepts from a literature review. Results: The GEM hierarchy includes more than 100 elements. Major concepts relate to a guideline's identity, developer, purpose, intended audience, method of development, target population, knowledge components, testing, and review plan. Knowledge components in guideline documents include recommendations (which in turn comprise conditionals and imperatives), definitions, and algorithms. Conclusion: GEM is more comprehensive than existing models and is expressively adequate to represent the heterogeneous information contained in guidelines. Use of XML contributes to a flexible, comprehensible, shareable, and reusable knowledge representation that is both readable by human beings and processible by computers. Keywords: Introduction : Over the last decade, clinical practice guidelines have become increasingly important repositories of knowledge about ideal practice. Built on a careful analysis and understanding of research evidence combined with expert consensus, a flood of guidelines are being created in an effort to diminish inappropriate practice, to improve health outcomes, and to control costs of care. Sponsoring organizations most often publish practice guidelines as paper-based, prose documents, which sometimes include algorithmic flowcharts. These publications are typically unavailable during clinical consultations. Zielstorff noted that while electronic dissemination can "solve the problem of accessibility to the guideline itself... access to knowledge embedded in the guideline can still be problematic." She called for defining guideline knowledge in "standard interchange formats that permit installation in a wide variety of technical infrastructures." Guideline querying, electronic distribution, and decision support systems that implement the guideline's recommendations can all be facilitated when the knowledge contained in the guideline is represented in digital form. The proposed Guideline Elements Model (GEM) is intended to serve as a document model, i.e., an idealized abstraction of a guideline document that masks certain details to bring forth others (after Degoulet and Fieschi). By describing concepts pertinent to guideline representation, attributes of those concepts, and relationships among the concepts, GEM promotes translation of natural language guideline documents into a format that can be processed by computers. The framework is intended to be useful to developers, disseminators, implementers, maintainers, and end users of guidelines. XML (the Extensible Markup Language) offers a powerful technology for representing electronic documents. It allows both computers and human beings to access the information in a document and extract it for reuse or modification. In XML, guideline documents are conceptualized as a hierarchy of elements ---basic units of information that store data and define structure by virtue of their position in a tree. Element tags demarcate text and provide labels that characterize the semantic content of the element. Tagging a document does not require programming skill, yet it can create a computer-processible representation of the knowledge contained in a guideline. An XML document type definition (DTD) models the names of allowable elements and attributes in the document, the content of each element, and the structure of the document (i.e., the order and cardinality of each element). A validating parser can ensure that any tagged document conforms to the DTD. A parser (such as Internet Explorer 5) can read an XML document file and populate a tree in memory, thereby exposing all the elements and attributes to manipulation by an application. That application might, for example, select certain components for presentation to the user (e.g., costs) or interact with facts in a clinical database to provide guideline-based decision support. Several stakeholders combine their efforts to develop, disseminate, implement, and maintain the knowledge contained in guidelines throughout the guideline life cycle. Not surprisingly, therefore, the models that have been created to represent this knowledge vary, depending on the orientation of the stakeholder. Models devised by those who build and evaluate practice guidelines tend to emphasize descriptions of the methods applied in guideline development, issues of guideline testing and maintenance, and details about the sponsors, objectives, and intended audiences; however, they often provide few details for conceptualizing recommendations. For example, the National Guideline Clearinghouse (NGC) model contains most of the concepts found in the health services literature, yet it provides only a single slot for "Major Recommendations." On the other hand, informatics researchers tend to model guideline recommendation components in great detail but often fail to address the concerns of the health services community, such as evidence strength and quality, potential biases of sponsors and developers, and validity checks. For example, although our own augmented decision table model for representing guidelines knowledge recognized the importance of "collateral information," it fell short in defining a comprehensive set of considerations. One of the most ambitious efforts to build an electronic guideline representation is the Guidelines Interchange Format (GLIF) of the Intermed Collaboratory. GLIF includes constructs for the name of a guideline and its authors, purpose, and eligibility criteria, but it specifies only a vaguely defined component called didactics to "provide background or supporting information." The emphasis in GLIF is on detailed specification of guideline recommendations. Since different stakeholders who work with a single guideline have different information needs and require different computer applications to support those needs, an ideal electronic model of a guideline should be capable of representing all pertinent aspects of the document. In this work we attempt to bridge the deficiencies in existing designs by proposing a guideline model that includes a more comprehensive set of components. It would be naive to believe that any model could completely meet the needs of all current and future users. We therefore set a goal of creating a "more comprehensive" model that would represent the most important guideline components and be sufficiently flexible to allow future extensions if needed. In this work, we describe our approach to developing GEM, followed by a presentation of the model itself. We next illustrate how a variety of guideline models from both the health services and informatics literature map to GEM, and we briefly describe our experience modeling a variety of published guidelines with GEM. We conclude with a discussion of attributes of an ideal guideline document model. Approach to Model Development : We relied on three primary sources to form the core set of guideline elements. First, for concepts related to guideline development and evaluation, the Institute of Medicine's Provisional Instrument for Assessing Clinical Guidelines provides a detailed set of evaluation criteria for practice guidelines. The purpose of this instrument was to define the attributes of ideal guidelines, to encourage systematic guideline development, and to provide a standardized approach and structure for the assessment of guideline documents. It consists of 46 questions related to seven guideline attributes: validity, clarity, multidisciplinary process, clinical flexibility, reliability and reproducibility, clinical adaptability, and scheduled review. Pertinent constructs for GEM were extracted from these questions. Second, for concepts related to guideline dissemination, the NGC provides an online resource for evidence-based guidelines. Sponsored by the U.S. Agency for Healthcare Research and Quality (formerly the Agency for Health Care Policy and Research), the American Medical Association, and the American Association of Health Plans, it provides a schema for classification of the components of guideline documents. This model includes a set of "key attributes" for summarizing each guideline to facilitate search and retrieval of information from the NGC Web site and comparison between guidelines. The NGC also offers several sets of controlled vocabulary constructs to describe concepts in the model. Third, for concepts related to implementation of guideline recommendations, we apply a set of constructs derived originally from the augmented decision table model. This approach has been used to represent knowledge derived from: The original model has been substantially enhanced and extended. GEM represents a key infrastructure component for a proposed object-oriented framework for development of computer-based guideline implementations. Constructs extracted from these sources were supplemented by additional concepts derived from published models. We searched the MEDLINE database (1990 to 1999) using the OVID search engine. The search strategy looked for "practice guidelines" as a subject heading and "guideline" as a text word. The results were combined with "knowledge representation" or "model" or "evaluation." In another search we looked for papers that had been published in the Journal of the American Medical Informatics Association, Methods of Information in Medicine, or the proceedings of AMIA annual meetings that addressed guidelines. The bibliographies of selected articles were also searched for relevant publications, as were the authors' reference files. Articles were selected that modeled and categorized the content of clinical practice guideline documents. Papers that failed to specify detailed modeling constructs and designs that described models of guideline implementations without describing document models were excluded. Markup using GEM tags was the natural outgrowth of a system that has been in use in our laboratory since 1995. Members of the Guidelines Review Group at Yale have collaborated with the Committee on Quality Improvement of the American Academy of Pediatrics to review and critique proposed evidence-based guidelines prior to publication and at the time of scheduled review. "Logical analysis" is our name for the cognitive task by which recommendation components are extracted from the natural language text of clinical practice guidelines and specified in a computable format. The first step in logical analysis has been to mark up paper-based documents using colored highlighters to identify and categorize guideline components, such as recommendations, evidence, costs, patient preferences, and clinical options. The GEM hierarchy permits much more detailed categorization than is possible with the physical highlighting system. Proposed Model : As shown in , GEM can be depicted as a directed graph with Guideline Document as the root. The major concepts in the first tier of the GEM hierarchy below the root level are identity, developer, purpose, intended audience, method of development, target population, knowledge components, testing, and review plan. Each of these elements, in turn, comprises one or more additional levels of guideline constructs. Components of GEM are defined as XML elements. Elements have distinct names and are delimited with start and end tags, e.g., Diabetic Nephropathy <Title> Elements may contain other elements, they may store text, or they may be empty. Elements may appear as often as required. Most elements store information that is presented literally in the guideline text itself, e.g., release date, name of sponsoring organization, and recommendation text. A small number of metalevel tags provide information about the guideline, which has been interpreted, e.g., developer.type. To indicate whether an element's content is explicitly stated in the guideline document or was inferred by the person who performed the markup, each element has an attribute called "source." The source attribute can take values of "explicit," "inferred," or in some cases "NGC" (to indicate that the NGC structured vocabulary is used). GEM has been proposed as an ASTM E31.25 standard representation for guideline documents. Following ASTM and HL7 conventions, element names in GEM are formatted in lower case and words are separated by periods. In this report, italics indicate specific elements (e.g., title, decision.variable). The complete GEM hierarchy, definitions for all elements, a GEM template, the document type definition, and the schema can be viewed at . In the next sections, we describe the major elements of the GEM hierarchy. Identity | Information that identifies a particular guideline document and describes it in general terms is clustered in the identity construct. The identity element includes the guideline's complete title, a citation that references its publication, its release date, its availability (in electronic and print formats), and a person or organization that can be contacted for further information. The status element indicates whether the guideline has been updated or revised. Since many current guidelines are released as packages that may include patient education materials, foreign language versions, quick reference guides, and technical reports, a construct for companion.documents is included. An entry stored in the adaptation element indicates whether the guideline has been adapted from another publication. Developer | The organization responsible for development of the guideline is identified and described. A developer.type element (e.g., medical specialty society, federal government agency, managed care organization) provides a structured description of the guideline's sponsor. The formal name of the committee within the developing organization as well as its members' names and individual or committee expertise are represented. In addition, sources of financial support for the guideline's development, the names of organizations that have endorsed the guideline, and reference to other organizations' guidelines on the same topic are included. Purpose | Purpose elements describe the main health practices, services, or technologies addressed by the guideline and reasons for the guideline's development. Guideline category classifies the major focus of the guideline, e.g., diagnosis, treatment, or prevention. The rationale for guideline development (e.g., evidence of inappropriate practice, wide practice variation) is subtly different from the objective of the guideline (e.g., to increase use of a particular test, to diminish inappropriate use of a therapy), and either (or both) may be described. The health.outcome element stores the specific health outcomes or performance measures that the guideline is intended to affect. The available.option element describes the principal alternative preventive, diagnostic, or therapeutic interventions that are available. Exception refers to factors that may permit an exception to be made in applying the guidelines, including home and family situation and constraints on the health care delivery system. Strategies, performance measures, and plans for implementing the recommendations may be stored in the implementation.strategy element. Intended Audience | The intended.audience element refers to the health care providers whose behavior the guideline is intended to influence. It includes both professional.group and care.setting constructs, indicating where a guideline recommendation may be applicable, e.g., office, intensive care unit, or a particular health maintenance organization. The clinical.speciality element applies the NGC structured vocabulary to categorize the intended users. Method of Development | The validity of a guideline's recommendations is closely tied to concepts incorporated in development.method. Evidence-based guideline development processes relate recommendations directly to the scientific evidence that supports them. Such constructs are clearly important to developers and implementers and to end users of guideline recommendations as well, as they decide whether the recommendations should influence their behavior. The description.evidence.collection element refers to approaches taken by the guideline developers to identify and retrieve scientific evidence. The method.evidence.collection element stores an NGC structured construct; number.source.documents refers to the number of documents identified during evidence collection. Evidence.time.period refers to the publication dates of the evidence. Method.evidence.grading stores criteria used to gauge the quality of information from different sources and may include a formal rating.scheme. Method.evidence.combination refers to formal methods of synthesis used to develop summary measures that reflect the strength of scientific evidence, e.g., meta-analysis, decision analysis, or formal group judgment techniques. Specification.harm.benefit describes qualitatively the anticipated benefits, potential risks, or adverse consequences associated with implementing the guideline recommendations, while quantification.harm.benefit stores mathematical models and numeric estimates. The role.value.judgment element stores information related to whose values were applied in determining the relative desirability of a health practice. For example, guidelines that optimize health care from the point of view of the individual patient, the payor, and society may well differ. Likewise, the role.patient.preference element stores information about how preferences were applied to determine guideline policies. Target Population | Target.population refers to the group of persons who are the subject of the guideline recommendations. The eligibility element may include criteria ---the inclusion.criterion and exclusion.criterion ---that determine the specific portion of the target population for which recommendations are applicable. For example, a guideline on managing otitis media in young children defines the inclusion criteria as "age 1 through 3 years" and "otherwise healthy except for otitis media with effusion." Exclusion criteria are specified as "craniofacial or neurologic abnormalities" and "sensory deficits." The NGC specifies sex and age ranges for categorization of the target population. Testing | The external.review element stores information about the findings of persons and groups outside the sponsoring organization, who have reviewed recommendations. The pilot.testing element stores text that refers to testing of the guideline's recommendations in clinical settings. Revision Plan | The scheduled.review and expiration elements store review and expiration dates that help determine the validity of the recommendations in light of new evidence. Knowledge Components | Knowledge components store and categorize the expert knowledge that is the salient feature of clinical practice guidelines. We have classified knowledge.components into three high-level classes ---recommendation, definition, and algorithm ---because the sub-elements of each of these call for different approaches to processing . Each knowledge component and its sub-tree in the GEM hierarchy are discussed below. Recommendation | Recommendations are the unique components that distinguish guidelines from other clinical publications; recommendations are intended to influence practitioners' behavior. When recommendations are analyzed into atomic concepts (and perhaps encoded in a structured vocabulary), they can be executed by a computer's logic. Recommendations can be categorized as conditional or imperative statements. While conditional statements clearly delineate the situations in which they apply, imperatives are broadly applicable to the target population and do not impose constraints on their pertinence. Conditional recommendations can be described in rules that take the form: If CONDITION then ACTION(s) {because REASON(s)} A condition, in turn, is specified by one or more combinations of a decision.variable and its value linked by comparison operators, e.g., <decision.variable>platelet count</decision.variable><value>less than 50,000</value>. In many cases, the value of a decision variable is not explicitly stated in guideline text but is implied to be true or present. Fulfillment of the condition triggers at least one guideline-specified action. Reason elements explain why the action has been triggered. The evidence.quality that led the guideline developers to call for a particular recommendation and the strength they attach to a particular recommendation are also tagged. The flexibility element describes optional conditions or actions that relate to a particular rule and are often recognizable by the presence of "or" statements in the guideline text. Defining a condition and executing an action often entail an economic burden that can be described in cost elements associated with an individual decision.variable or action or with the higher-level conditional. Information about the relationships between recommendations is stored in the link element. Such links might define a temporal sequence or a part-whole relationship or relate one part of the hierarchy to another. A reference slot can be used to store citations to specific evidence that supports a particular recommendation. The logic element summarily stores the Boolean connectives that link component decision variables and actions; for example: IF decision.variable 1 AND decision.variable 2 THEN action 1 OR action 2. At deeper levels of the conditional tree, elements store information that describes in detail individual decision variables and actions. Specific elements define quantitative test parameters for individual decision variables (sensitivity, specificity, predictive.value) and benefits and risks or harms associated with individual actions. In contrast to conditional recommendations, imperative recommendations present broadly applicable directives (which parallel the actions in a conditional recommendation), e.g., A major aspect of initial treatment should consist of lifestyle modifications, such as weight loss, reduction of salt and alcohol intake, and exercise.... The laboratory must use a screening procedure that will detect sickle hemoglobin in the newborn.... Test results must be reported in understandable language that includes the identified phenotype, diagnostic possibilities, and sources where additional information may be obtained. The laboratory also should inform the infant's mother of the screening result, unless prohibited by law. Imperatives often include terms such as "require," "must," and "should" but do not contain conditional text (e.g., "if," "when," "whenever") that would limit their applicability to specified circumstances. With the exception of decision.variable elements (which exist only in the conditional sub-tree), most of the deeper-level elements of the knowledge components hierarchy are similarly applicable to both imperative and conditional statements. Definition | A definition element stores important guideline terminology as well as the meaning of the terms. For example, a guideline that advises on appropriate diagnostic testing for children with febrile seizures includes a careful definition of "simple febrile seizure." A Centers for Disease Control guideline on hepatitis B immunization recommends more intense immunoprophylaxis for infants of "high risk mothers," a high-level concept defined to include intravenous drug abusers and women with sexually transmitted diseases during pregnancy or pre-existing liver disease. Indeed, the American Academy of Neurology has issued a guideline that is expressed as a set of case definitions ---rather than recommendations ---for HIV-associated neurologic disease. Algorithm | Many (but not all) guidelines include an algorithm that is graphically represented in flowcharts. This describes a temporal sequence of activities and the branching decision logic that implement the guideline's recommendations. In GEM, a flowchart can be included en bloc as an algorithm element, or it can be broken down into its component parts. The GLIF specification consists of a collection of "guideline steps," which are linked in a directed graph. The GEM algorithm hierarchy includes elements derived from the GLIF steps model: action.step, which specifies a clinical action that is to be performed in the patient-care process; conditional.step, which directs flow from one guideline step to another on the basis of evaluation of a criterion; branch.step, which directs flow in alternate directions; and synchronization.step, which represents a convergence of other steps. Figure 1 | High-level concepts in the Guideline Elements Model. High-level concepts in the Guideline Elements Model. Figure 2 | Detailed model of the knowledge components hierarchy. Detailed model of the knowledge components hierarchy. Mapping : Although considerable research has been focused on representing guideline knowledge, no single model has gained wide acceptance. We selected, from our literature review, a sample of existing models that we think are representative of a range of guideline document models, to explore our hypothesis that current models are limited in their comprehensiveness. We attempted to map GEM elements to constructs described in these published models. Each publication was reviewed by at least two authors to establish whether a particular concept was described in the publication. Conflicts were resolved by discussion among the authors. In general, we tended to be liberal in these mappings, because we recognized that the published specifications of models might be incomplete. To represent "health services" models, we selected the Institute of Medicine's Appraisal Instrument, the NGC model, a proposal for structured abstracts of scientific papers that describe clinical practice guidelines, and a recent evaluation of guideline quality. Informatics models of guideline documents include the Arden Syntax, an international standard for encoding logic in decision support systems; GLIF, a knowledge representation intended to facilitate guideline sharing; a relational model that captures both structured guideline content and procedural logic; PRESTIGE, a generic approach to representation of guideline knowledge in the European Community; a Web-based guideline dissemination system available nationally on the Kaiser Permanente intranet,; and a model based on augmented decision tables. In , we show the high-level concepts from GEM that are represented in these selected models. In general, health services models were more explicit with regard to developer, purpose, intended audience, and method of development concepts than were informatics models, as indicated by the density of the gray bars in those segments. On the other hand, although every model had a construct for "recommendation," the informatics models tended to atomize knowledge components into more detailed constructs than did the health services models. The display is somewhat deceptive, because each element is given equal visual weighting. Nonetheless, it is clear that many informatics models lack constructs for encoding knowledge about guideline development methodology and validity assessment. Likewise, the health services models under-specify features that facilitate implementation of recommendations. Figure 3 | GEM constructs represented in a variety of guideline models. GEM constructs represented in a variety of guideline models. A high-level concept was considered to be present if it or any of its subordinate concepts was described. See text for sources. Guideline Markup with GEM : The usability and expressive adequacy of GEM were tested by applying it to a selection of guidelines. The authors modeled published guidelines as GEM documents from a variety of disciplines that represented areas of their expertise. Only practice guidelines sponsored by national organizations were modeled, although we believe that the process should be applicable to local guidelines. No effort was made in this study to model critical pathways or clinical trial protocols, which may require additional elements. Electronic versions of the guidelines were marked up using Microsoft XML Notepad 1.5, by copying pertinent text and pasting it into an empty GEM Schema template. The guidelines can be viewed using Internet Explorer 5.0 (or later versions) at . An example of a portion of a GEM document is shown in . As might be expected, there was substantial variation in the use of GEM elements from guideline to guideline. No two guidelines ---whether produced by the same organization on different topics or produced by different organizations and covering the same health condition ---are constructed identically. We noted that a number of tags were not used to model any of the documents. Moreover, as has been noted in the evaluation of GLIF endocing,, there was considerable variation in the way modelers analyzed the guidelines. GEM offers flexibility with respect to the granularity at which individual elements are encoded, and this flexibility was exploited. This effect was not formally studied and will be the subject of future research. Figure 4 | Identity elements from a guideline depicted hierarchically as a GEM document. Identity elements from a guideline depicted hierarchically as a GEM document. Opening and closing tags are shown in boldface. Contact, status, and patient.resource are empty. The complete markup of this guideline may be viewed at . Discussion : We propose a document model for practice guidelines that can store and organize the heterogeneous information they contain. Although the elements identified in this work could be added to most existing health services and informatics models, GEM describes concepts and knowledge more comprehensively than do other current models. An ideal guideline knowledge model should be: GEM markup can be used as a first step in translating paper-based, narrative guidelines into formats that can be processed electronically. Developers can use a GEM-encoded document as a set of empty slots to be filled to create a high-quality guideline; e.g., a fully tagged document could facilitate decision table verification of guideline logic, or an XSL-formatted template can help automate the extraction of components that indicate methodological quality. Disseminators can use XML's Web capabilities to publish guidelines. Implementers can use the tagged data for assistance in encoding recommendations, understanding terminology, and even direct execution. For example, a conditional recommendation's decision variables could be automatically extracted from a GEM document and used to label a data collection control, while the potential values are used to name radio buttons. End users can select various aspects of interest from GEM-encoded documents (e.g., the quality of evidence that supports individual recommendations, the costs of interventions) or compare guidelines on the same topic from different sources. Use of XML for representation of GEM offers a number of advantages. The self-descriptive capability of XML improves searching for, indexing, and locating information. Moreover, the open XML standard facilitates development of tools for document processing. XML is an intrinsic part of the Web, with presentation and parsing capabilities built into Web browsers. Software to process XML documents is expected to become ubiquitous and inexpensive. Over their lifetime, documents represented in XML can be used and reused in a multitude of ways, including (most likely) some that have not yet been invented. GEM has several limitations. The model is simply an abstraction of the guideline document and, as such, must rely on extrinsic systems to apply it in ways that are useful. GEM does little to resolve the ambiguities that are present in many guidelines. It can, however, faithfully present them to a user for resolution. Use of a system that forces developers to define recommendations as if-then-else statements might help avoid introduction of ambiguous statements. Although GEM extends the work of multiple researchers, this model is probably not comprehensive. Additional elements, attributes, and relationships may be necessary to adequately encode guidelines, depending on the needs of stakeholders. The XML representation can be updated easily to accommodate these needs. Since the model currently incorporates more than 100 elements, effective markup with GEM will require training and practice to achieve optimal results. Next steps in our work with GEM will involve refining the model and building and evaluating tools that facilitate activities throughout the guideline life cycle. We are working to create parsing and editing tools, specifically designed for guideline markup, that will promote consistent encoding. We envision Web-based tools for guideline developers that will allow them to collaborate effectively without face-to-face meetings ---a major source of guideline development expense. Another goal of this project is to create clinical decision support tools automatedly from guideline documents stored in GEM format. Boxwala et al. have described an architecture for a guideline execution engine using ActiveX, which operates from guidelines encoded in GLIF. We plan to apply ASP (active server pages) technology to dynamically configure intranet Web pages from XML documents. GEM is intended to meet the needs of a wide variety of stakeholders in the guidelines initiative. More comprehensive models are necessary to describe fully the heterogeneous knowledge contained in clinical practice guidelines. The extensibility and computability of XML make it ideal for guideline document representation. We offer this model as an open, extensible framework and welcome contributions from others working in this area. Backmatter: PMID- 10984469 TI - Corpus-based Statistical Screening for Phrase Identification AB - AbstractPurpose: The authors study the extraction of useful phrases from a natural language database by statistical methods. The aim is to leverage human effort by providing preprocessed phrase lists with a high percentage of useful material. Method: The approach is to develop six different scoring methods that are based on different aspects of phrase occurrence. The emphasis here is not on lexical information or syntactic structure but rather on the statistical properties of word pairs and triples that can be obtained from a large database. Measurements: The Unified Medical Language System (UMLS) incorporates a large list of humanly acceptable phrases in the medical field as a part of its structure. The authors use this list of phrases as a gold standard for validating their methods. A good method is one that ranks the UMLS phrases high among all phrases studied. Measurements are 11-point average precision values and precision-recall curves based on the rankings. Result: The authors find of six different scoring methods that each proves effective in identifying UMLS quality phrases in a large subset of MEDLINE. These methods are applicable both to word pairs and word triples. All six methods are optimally combined to produce composite scoring methods that are more effective than any single method. The quality of the composite methods appears sufficient to support the automatic placement of hyperlinks in text at the site of highly ranked phrases. Conclusion: Statistical scoring methods provide a promising approach to the extraction of useful phrases from a natural language database for the purpose of indexing or providing hyperlinks in text. Keywords: Introduction : Modern computer-based retrieval systems have the potential to retrieve from a large database those documents that satisfy a Boolean query composed of virtually any words and phrases the operator may desire. Given this power, it is reasonable to ask what purpose indexing could serve. There is actually the potential for a large benefit. As shown by a number of studies,,,,,, there is great inconsistency in the terms people use to describe the same subject. In the words of Bates, "In study after study, across a wide range of environments, it has been found that for any target topic, people will use a very wide range of different terms, and no one of those terms will occur very frequently." Indexing can alleviate this problem by expanding the list of terms by which a document may be accessed. Thus, one path to improved indexing is to obtain a list of terms (words and phrases) sufficient to include a high percentage of the terms that people will actually use in querying a database, and add sufficient synonymy information to allow a query expressing a particular concept to access those documents that are indexed with an expression synonymous with the query. The Unified Medical Language System (UMLS) includes not only a large list of important terms but also a synonymy capability relating these terms in the Metathesaurus. It is intended, among other things, to provide a solution to the indexing problem just out-lined., The system as it stands is, of course, incomplete and will, for the forseeable future, stand to benefit from increased coverage of the latest terminology in the various fields covered by MEDLINE. Our hypothesis is that statistical information about the occurrence of phrases in MEDLINE can provide a useful screen for candidate phrases that are of similar quality to the material already in the UMLS. A person generally does not possess the kind of information that is available in this way. This information can, however, be readily obtained by automatic processing and can serve as a guide to terms that would make useful additions to UMLS. Such guidance may be important, given the limited human resources that are available to integrate terminology into the Metathesaurus. The development of a controlled vocabulary of indexing phrases is not the only use to which our methods can contribute. The ranking of phrases by quality can also be used as an aid to automatically place hyperlinks in text. We are currently involved in a project to link phrases in MEDLINE documents to appropriate sections and subsections of books in the field of bio-medicine that may provide the reader with additional information about the subjects of the phrases. Here the most useful phrases in a MEDLINE record are marked as hot links that are "clickable" to reach a book or books of potential interest. The first book, Molecular Biology of the Cell, is available at . To see the book links, the user must select a single document and, when it is displayed, click on the "book" link to the right of the document. In the applications sections of this paper, we show how this type of linkage can be produced automatically on the basis of phrase ranking. The hyperlinks viewable at differ mainly in incorporating some human review of the phrase lists used. Several methods have been used historically in attempts to extract useful words and phrases from document collections for purposes of indexing. Since Luhn's pioneering work on indexing,, the importance of term frequency information has been recognized. The frequencies of both phrases and the words that compose them are important in the phrase extraction method of Jones et al. A second kind of information that can be helpful in indexing is the distribution of frequencies of a term within documents. It has been proposed that non-content-bearing terms are well modeled by a single Poisson distribution, whereas content-bearing terms require a two-Poisson or some more complicated model.,, In fact, one of our scoring methods is based on the degree that a term's distribution deviates from a Poisson distribution. The greater this deviation, the more likely that the term is a useful one. In addition to this method we employ other relatively simple scoring methods based on term frequencies, co-occurrence, and word suffixes. The objective is to locate phrases that are grammatically acceptable and specific in their meaning, yet occur with sufficient frequency in the database to make them useful additions to UMLS. There are many methods of noun phrase extraction based on natural language processing that we have not examined. Proprietary methods such as CLARIT and NPtool were not of interest, since we seek to understand the methods in as much detail as possible. The transformation-based parsing developed by Brill,, hidden Markov part-of-speech tagging as in the Xerox Tagger, and parsing based on a probabilistic grammer as in CHOPPER are potentially of greater interest. However, these are complex tools designed for a different task than ours. They seek to assign part-of-speech tags as a basis for natural language parsing, whereas we seek to identify those phrases that are not only syntactically correct but also readily recognizable by human beings as useful and descriptive of a subject area. Even if natural language parsing methods can contribute to the accomplishment of our task, we must still ask whether their complexity is necessary to its accomplishment. We seek to show in what follows that simpler methods suffice. We will examine the Xerox Tagger to show that it adds little to what can be accomplished by our scoring methods. Another phrase extraction task that has been studied is phrase extraction with the purpose of improving retrieval by expanded automatic indexing on test collections. The methods of phrase identification are based on part-of-speech tagging as well as some statistical methods. This area is exemplified by the work of Fagan and Lewis and Croft. Interestingly, while there has been some success with this approach in improving retrieval, the results are not consistently good. This led Lewis and Jones to comment that "... automatically combining single indexing terms into multiword indexing phrases or more complex structures has yielded only small and inconsistent improvements over the simple use of multiple terms in a query." We mention this area mainly to distinguish it from our own work. Different goals and different methods of evaluation characterize the two approaches. Instead of seeking to improve retrieval in some automatic system, we seek to identify those phrases that are the most user friendly, and we evaluate our success by how well we are able to identify a set of phrases (UMLS) that are maintained by human beings because they are found descriptively useful. We begin with a description of the different data sets we study and how they are constructed. We then present the scoring methods that are designed to distinguish useful phrases from simple co-locations of terms. We describe our approach to evaluation of the scoring methods and present results on the effectiveness of the scoring methods when applied to a large database of MEDLINE records. Besides the six scoring methods that we find useful, we evaluate two other methods and find that they do not add significantly to overall effectiveness. We discuss application of our methods to extraction of candidates for UMLS and also as a procedure for marking text with hyperlinks. The paper concludes with a discussion and description of future directions. Data Sources and Preparation : We consider word pairs and word triples from two different sources. The first source is the UMLS, developed by the National Library of Medicine. The UMLS (9th edition, 1998) was obtained from the National Library of Medicine on CD. (Information regarding its availability for research purposes may be found at .) Our second data source is the set of 304,057 MEDLINE records with abstracts and entry dates in the year 1996. We shall refer to this document set as MED96. These two data sets are processed somewhat differently, because they differ considerably in content. However, we will use a procedure to normalize text strings that is the same for both. We normalize text in three steps: All alphabetic characters are lowercased, all non-alphanumeric characters are replaced by blanks, and multiple blank spaces between words are converted to single blank spaces. The UMLS is processes as follows: First, all text strings are obtained from the UMLS "mrcon" (concept name) file. From the resulting set of strings, any containing punctuation marks or stop words are deleted. For this purpose a list of 310 common stop words is used. Finally, the remaining strings undergo normalization and removal of any duplicates. The result is a set we denote by Uall. This is the set of all phrases that we obtain from UMLS. From Uall we extract the subset of strings consisting of two words each. The result is 156,086 word pairs, denoted by U2. In the same way we extract all three-word phrases from Uall. The result is 103,367 word triples, denoted by U3. MED96 is processed somewhat differently. We first process the titles and abstracts of the MED96 records, breaking at punctuation marks and stop words. The resulting set of strings is normalized and made unique, to produce the set Mseg. This is the set of longest phrases that we obtain from MED96. By M2 we denote the set of all contiguous word pairs that can be obtained from the members of Mseg. For example, the four-word string "escherichia coli cell growth" from Mseg yields the three overlapping two-word phrases "escherichia coli", "coli cell", and "cell growth" in M2. By M3 we denote the set of all contiguous word triples that can be obtained from the same source. The difference in the processing of the UMLS and MED96 is perhaps worth emphasizing. The strings in Uall are essentially a subset of the strings that occur in the UMLS "mrcon" file, except for lower casing, and as such by and large represent syntactically reasonable and semantically meaningful phrases. The subset U2 is just those strings in Uall that are composed of two words. There are longer phrases in Uall that could be broken up into contiguous two-word phrases and added to U2, but we do not do this because we do not know whether these would be of high quality. The same applies to the derivation of U3. The U2 and U3 sets represent our gold standard for good phrases, and we seek to keep their quality as high as possible. On the other hand, Mseg is a large set of strings that are obtained from all the text in MED96. Many of these are not, as phrases, of high quality. The M2 set is derived from Mseg by taking all those strings in Mseg that consist of two words as well as all those contiguous word pairs that may be obtained from longer phrases in Mseg. The longer phrases in Mseg are broken up in this way and added to M2 because, even if the longer phrases are of poor quality, some two-word substrings may be of good quality and such potential should not be ignored. The same basic method applies to the derivation of M3. It will then be the task of the scoring procedures that we introduce to separate the good from the bad. Scoring Methods : In this section, we define the various scoring methods we want to apply to the word pairs in the set M2 and the word triples in the set M3 extracted from the MED96 database. Our goal is to define scoring methods that will allow us to find the most useful phrases occurring in a database. We only define the methods and give some justification for their choice here. Their systematic evaluation is the subject of the next sections. We begin by describing scoring methods for the word pairs in the set M2. When these have been described we indicate the modifications necessary for application of the same methods to M3. Method I | Given a word pair in the set M2, we perform a simple count of the number of MED96 documents that contain that word pair (phrase frequency). Dividing this count by the normalization factor N (the size of MED96), the corresponding score s1 is The normalization factor is a constant and is optional here, but it might allow one to compare results across databases more readily. Rationale: Phrases as well as single words follow a Zipf-like distribution, with a plethora of very low frequency phrases and progressively fewer examples in the higher-frequency categories. Rare phrases are of only limited value as discriminators. Naturally, the UMLS tends to avoid very low frequency terms, which explains the utility of frequency as a scoring method. Method II | Given a word pair in the set M2, we count the number of documents in MED96 that contain both words, even if not as a contiguous pair. The result is called the co-occurrence, and the score s2 is It is evident that this score always lies between 0 and 1. Rationale: As an example, consider two word pairs, "diabetes mellitus" and "wide tumor." For "diabetes mellitus," the phrase frequency (the number of the MED96 documents that contain "diabetes mellitus") is 2,465, the co-occurrence (the number of the MED96 documents that contain both "diabetes" and "mellitus" but not necessarily as a contiguous pair) is 2,468, and thus the score s2 is 0.99. For "wide tumor," the phrase frequency is 3, the co-occurrence is 352, and the score s2 is 0.008. Two words that tend to co-occur only in the form of a phrase often form a high-quality phrase. Method III | Given a phrase in the set M2, we examine all occurrences of the phrase throughout the text of MED96. As described in the previous section, the text of MED96 is broken at stop words and punctuation marks, and the resulting phrases compose the elements of Mseg. Each occurrence of a phrase that immediately precedes one of these break points (at a stop word or a punctuation mark) is counted in phraseend for that phrase. For example, the word pair "lipoprotein cholesterol" occurs in the sentence fragment"... serum total and high-density lipoprotein cholesterol, C-reactive protein, and plasma fibrinogen." Here it occurs just before a comma, and hence this occurrence will contribute 1 to the score phraseend for the phrase "lipoprotein cholesterol." The meaning of "end" in this context is that "lipoprotein cholesterol" is at the right-hand end of the longer phrase "high-density lipoprotein cholesterol" that this sentence fragment contributes to Mseg. In the same sentence fragment is also the phrase "density lipoprotein," but since this occurrence of "density lipoprotein" does not immediately precede a stop word or punctuation mark, it does not add to the score phraseend for "density lipoprotein." Again, normalizing by the total number of MED96 documents, the score s3 is Rationale: The scoring method s3 is a quasi-syntactic categorization. The head of a phrase tends to occur at the right-hand end. The number of times that a phrase ends at a stop word or a punctuation mark is a measure of the likelihood that its last word is a head and, therefore, of whether the phrase can stand alone. For example, "central nervous" will be followed immediately by a stop word or punctuation mark much less frequently than will the phrase "nervous system." Method IV | The score s4 is obtained as an odds ratio based on the last three characters of the last word in the phrase. The definition is where the number p(good phrase|l1l2l3) is the probability of being a good phrase given the last three letters l1l2l3. From a simple rearrangement of the Bayes theorem, we can infer s4, i.e., where p(l1l2l3|good phrase) is obtained as the distribution of the last three letters of the last word over all phrases of Uall, and p(l1l2l3) is obtained as the distribution of the last three letters of the last word over all the phrases in M2. Rationale: The scoring method s4 is based on the characteristic suffixes that tend to be applicable to different word classes and different parts of speech. For example, if the last three characters of the last word in a word pair are "-ely," as in "bind cooperatively," the phrase may not be of very high quality (s4 = 0.044). However, if the last three characters of the last word in a word pair are "-ine" (often the suffix of a chemical or medicine), such as "basophil histamine," "biogenic amine," and "catalytic histadine," the phrase may be of high quality (s4 = 2.45). Method V | Our next scoring method is based on the hypergeometric distribution. For a given word pair in the set M2, let nf equal the number of MED96 documents that contain the first word in the pair and ns equal the number of MED96 documents that contain the second word in the pair. Again, let N denote the total number of MED96 documents. If x denotes the co-occurrence of the two words and if we assume the words are randomly distributed, then x obeys the hypergeometric probability distribution, defined by Using this distribution we may obtain the P value, i.e., the probability that the actual co-occurrence is as great as or greater than the observed co-occurrence if the words are assumed to be randomly distributed: where min(nf, ns) is the smaller of the numbers nf and ns. Then s5 is given by Rationale: If the observed co-occurrence of a word pair is quite above the expected value for a random incident, the phrase may be a useful one. For example, for the word pair "surgically curable" we have nf = 1,583, ns = 148, N = 304,057, and co-occurrence = 3. The estimated co-occurrence (Eco) from the hypergeometric distribution is The words "surgically" and "curable" appear together at a near random level in the database. However, for the word pair "immunodeficiency virus," we have nf = 3,505, ns = 11,143, N = 304,057, and co-occurrence = 2,845. Also, the expected co-occurrence (Eco) from the hypergeometric distribution is The observed co-occurrence (= 2,845) for the words "immunodeficiency" and "virus" in the MED96 database is far above the random level (Eco = 128). The score s5, which is the negative logarithm of the P value, is the measure of the discrepancy from a random incident (s5 = 1.36859 for the word pair "surgically curable" and s5 = 3,527.45 for the word pair "immunodeficiency virus"). Method VI | Our final scoring method is based on the distribution of the within-document term frequencies. We define a randomly distributed phrase as one whose distribution among documents is described by a Poisson distribution. For such a phrase, the probability P(k) that fjd, the number of occurrences of phrase j in document d, is equal to k is given by where the parameter lambdaj is the average number of occurrences of j per document over the whole database. Therefore, we can find the probability p that the given phrase j occurs one or more times in d: We denote by q (= 1 - p) its complement, i.e., the probability that j does not occur in d. Let us consider an experiment that consists of N repeated independent Bernoulli trials with parameter p. Let E(= N p) refer to the expected number of documents containing the phrase considered. If a phrase occurs multiple times in few documents, we say it has a tendency to clump. We measure the tendency to clump by how much the observed number of documents containing the phrase (i.e., phrase frequency) falls below the expectation E. For a given word pair we calculate the P value, i.e., the probability that phrase frequency would be less than or equal to that observed if it were generated by the Poisson distribution of equation (9). Then the score s6 is given by Rationale: Intuitively, the occurrences of a term sensitive to content will have a greater tendency to clump than will those of a non-content-bearing term. This is common with names of things. Therefore, if the phrase considered carries content, we expect that the observed phrase frequency will be much less than E. The scoring method s6 is a measure of this clumping compared with a Poisson distributed phrase. For example, s6 is 168.08 for the name "ulcerative colitis," which is highly specific, but s6 is 0.65 for "common cancer," which is a general concept. The same scoring methods discussed for the word pairs in the set M2 can be applied to the word triples in the set M3 with a slightly altered definition of co-occurrence. Given a word triple in the set M3, phrase frequency is unchanged as the number of documents in MED96 that contain the word triple. However, for s2, co-occurrence is the number of documents that contain both the first word and the second and third words contiguously as a word pair. The same definition of co-occurrence applies when computing s5 and in equation (6), ns is the number of documents that contain the second and third words as a word pair. The value of phraseend in the score s3 in equation (3) is the number of phrases extracted from MED96 documents in which the given word triple in the set M3 occurs at the right-hand end. The score s4 likewise has the obvious interpretation where only p(l1l2l3) is changed to the distribution of letters (l1l2l3) appearing at the end of word triples in the set M3. Evaluation Method : Here we assume a given set of phrases M and a scoring method S that computes a real number for each phrase in the set M. The scoring method S allows us to rank the set M so that the phrases are in order of decreasing score. We also assume that we have available a golden set of good-quality phases G (this generally requires human judgment). The evaluation methods we use are measures of how well the scoring method S moves phrases in the set M I G to the top of the listing of M by rank (the lowest ranks). In other words, we consider the phrases in M I G to be the relevant phrases we are attempting to find in the set M. This allows us to view the problem as a retrieval problem and to apply some of the standard measures used in information retrieval science. In particular, we will apply recall and precision, which are the most commonly used measures in information retrieval. Because recall and precision are generally defined for a given rank and the results are different for each rank considered, we will also use the 11-point average precision as a single summary measure for the complete ranking. We will further use interpolated recall-precision curves as a graphic way of viewing performance. The 11-point average precision and interpolated recall-precision curves are widely used in presenting the results of retrieval experiments.,, Other measures are used in the information retrieval setting, such as the E-measure, expected search length, and relevance information. While these measures have some advantages in specialized settings, they are less intuitive and less well known, and we feel they offer no advantage in our setting. Let us assume that the number of phrases in M is N and that the phrases are represented by the list indexed in rank order, where the order is that of decreasing score S. Further, let Then the precision (Pr) and the recall (Rr) of S for the retrieval down to rank r is defined by and respectively. (Here parallelXparallel denotes the number of elements in the set X.) In words, Pr is the fraction of phrases retrieved down to rank r that are in M intersection G, and Rr is the fraction of phrases in M intersection G that are found in the retrieval down to rank r. Since the precision is usually high early in the ranks and becomes progressively lower at higher ranks, and since the recall is low at the early ranks but increases with increasing rank, it is possible to gain a useful picture of performance by graphing precision as a function of recall (a so-called recall-precision curve). However, precision does not always strictly decrease as one moves down the ranks. Because of this, it has become common to perform an interpolation on the precision value associated with a given recall level, in which that precision is replaced by any higher precision that may occur at a higher recall level. For example, if the precision 0.38 is calculated from equation (14) and the corresponding recall is 0.10, but a precision of 0.43 is found at a recall level of 0.20, then the value 0.38 is replaced by 0.43 as the accepted precision at recall level 0.10. In this way noise in the data may be reduced and the curve smoothed. We apply interpolation to obtain precision values at the 11 recall values of 0, 10, 20, 30, 40, 50, 60, 70, 80, 90, and 100 percent. (Notice that the precision at 0 recall is the highest precision found at one or more ranks of retrieval, since at zero ranks Pr in equation (14) is undefined). These 11 recall-precision pairs are used to produce interpolated recall-precision curves. We also average the 11 precision values together to produce the 11-point average precision as an overall summary performance measure. Results : Here we give the results of applying the scoring methods (defined under Scoring Methods) to the word pairs M2 and word triples M3 (described under Data Sources and Preparation). For the purpose of evaluation, the set of good phrases for M2 is G2 = U2 intersection M2 and the set of good phrases for M3 is G3 = U3 intersection M3. (The values U2 and U3 are defined under Data Sources and Preparation.) There are 26,131 phrases in G2 and 9,234 phrases in G3. The 11-point interpolated recall-precision curves for our scoring methods applied to the sets M2 and M3 are shown in and . It can be clearly seen that each scoring method moves the relevant phrases toward the top of the lists. The 11-point average precision for each scoring method has been computed. is the list of scoring methods and the 11-point average precision values on the set of word pairs M2, and provides corresponding data for the word triples in the set M3. For either M2 or M3 it is possible to combine the different scoring methods to produce a composite score. For example, we may take the linear combination of the logarithm of each score si with a coefficient xi and denote the resulting score as s: Because the logarithm and the exponential functions are monotonically increasing functions of their arguments, s as defined by equation (15) is equivalent to its exponential for ranking purposes, and we would have obtained the same results if we have defined s as a product of the factors .. Ranking the phrases in order of a decreasing combined score s with the coefficients xi where i = 1... 6, we can obtain the 11-point average precision. Through numeric study, we may seek the coefficients for which the combined score s gives the maximum 11-point average precision. We have iteratively maximized the 11-point average precision on one coefficient at a time. For example, x1 is varied with the remaining five coefficients x2, x3, x4, x5, and x6 fixed. This procedure is repeated until none of the coefficients can be altered to increase the 11-point average precision. We find that the combined score s with the coefficients x1 = -0.8, x2 = 3.9, x3 = 0.9, x4 = 3.1, x5 = 1.2, and x6 = 1.7 ---i.e., ---gives the maximum 11-point average precision we are able to achieve for the word pairs in the set M2. The result is listed in the final row of . Likewise, the combined score s with the coefficients x1 = -1.0, x2 = 1.5, x3 = 1.7, x4 = 2.6, x5 = 1.0, and x6 = 2.2 ---i.e., ---gives the maximum 11-point average precision, listed in the final row of , that is applicable to the word triples in the set M3. To get an idea how much each score contributes to the maximum 11-point average precision, we subtract its contribution from the combined score s. The resultant score (which is the optimal combined score s minus the contribution of an individual scoring method) and the 11-point average precision value after applying it are listed in the third and the fourth columns of for the word pairs in the set M2. The third and the fourth columns of provide corresponding data for the word triples in the set M3. We can see that the precision of each scoring method does not directly add to the total precision found for the combined score s. For example, although the scoring method s6 alone gives the largest 11-point average precision for both word pairs and triples, it does not make the largest contribution to the combined score s. This implies in particular that our scoring methods are not independent of each other. In computing the combined scores given in equations (16) and (17), we have employed an optimization procedure to choose the coefficients. In such a computation there is always the possibility of overtraining, so that the results are applicable only to the particular data set on which we have done the training. We suspected that this would not be a problem in the current situation, because we are training only six parameters and are employing a very large number of data, namely, 304,057 MEDLINE documents. Furthermore, in actuality there are only five independent parameters, because ranking is not changed when one multiplies all scores by a constant. To test for overtraining, we randomly split the set of documents into disjoint data sets, MED1 and MED2, where MED96 = MED1 MED2, parallelMED1parallel = 152,028, and parallelMED2parallel = 152,029. We then re-estimated the parameters for equation (16) on each subset independent of the other to produce the optimal 11-point average precision on that subset. The results are given in along with the coefficients for the whole of MED96 for comparison. Two things stand out here. First, we see that the coefficients obtained on MED1 and MED2 are almost identical. Second, we see that there is a significant difference between the coefficients obtained on the subsets and the whole database for scores s1 and s3. This suggests that database size might be a factor in these scores. To complete the comparison, we tested the effectiveness of each set of coefficients listed in on the two subsets MED1 and MED2. The results are given in . These results suggest that all three sets of coefficients have very close to the same effectiveness on MED1 and MED2. They are all within 1 percent of each other. This effectively rules out any significant overtraining. At the same time it suggests that the effectiveness of the composite scores is not very sensitive to the coefficients assigned to log(s1) and log(s3). While this is true, it is also true that we can degrade the composite quite drastically if we make these coefficients too large. A large coefficient for the contribution of a single score will cause this score to dominate the composite and the effectiveness of the composite to approach the effectiveness of the single score, as recorded in Tables and . Figure 1 | Recall-precision curves for the scoring methods applied to the set M2. Recall-precision curves for the scoring methods applied to the set M2. The 11 precision values are interpolated so that precision is a nonincreasing function of recall. Figure 2 | Recall-precision curves for the scoring methods applied to the set M3. Recall-precision curves for the scoring methods applied to the set M3. The 11 precision values are interpolated so that precision is a nonincreasing function of recall. Table 1 | Contribution of Individual Word-pair Scoring Methods to Combined Precision Table 2 | Contribution of Individual Word-triple Scoring Methods to Combined Precision Table 3 | Optimal Coefficients Table 4 | Eleven-point Average Precision for Word Pairs Other Methods Tested : From our survey of the literature, the method of phrase identification that seems the closest in spirit to the method we have developed is that of Jones et al. We tested their method on our task. The score of a phrase is given by a product W F N2, where W is the sum of the frequencies of the individual words that make up the phrase, F is the phrase frequency, and N is the number of distinct non-stop words in the phrase. Since we consider only phrases without stop words and since we apply the method to word doubles and word triples separately, the factor N2 may be ignored for our purposes. With this scoring method, the 11-point average precision on M2 is 0.06, and the 11-point average precision on M3 is 0.107. Tables and show that in each case the result is not as good as s1, which is equivalent to using F alone. This suggests that the factor W may be extraneous. As further support for this conclusion, we attempted to use the W F N2 score to improve the combined scores for both word pairs and word triples; however, we were unable to improve our results in either case. Part-of-speech tagging has been a popular method of extracting phrases from text for a variety of purposes.,,,,,, We were naturally interested in including part-of-speech tagging as part of our system. To examine the possible benefits, we obtained access to the Xerox Part-of-Speech Tagger. The tagger employed the SPECIALIST lexicon and was trained on MEDLINE text. To apply tagging to our problem, we required a method of scoring word pairs and triples based on tagging. We give here the details for word pairs only. The trained tagger was used to tag the MED96 corpus. By this means each occurrence of a member of M2 in MED96 has a tag pair associated from the tagging. We then constructed two lists of tag pairs. First, a set TM2 of tuples (t1t2,m) was constructed, where t1t2 is any tag pair that occurs as a tag pair of an instance of some member of M2 in MED96, and m is the number of different members of M2 that occur at least once in MED96 with the tag pair t1t2. The set TG2 is constructed on the basis of the same set of tag pairs as TM2, the difference being that for (t1t2,g) TG2, g is the number of different members of G2 that occur at least once in MED96 with the tag pair t1t2. On the basis of these tag pair lists, we can estimate the important probabilities. For any tag pair t1t2 with (t1t2,m) TM2, we set where parallelM2parallel denotes the size of the set M2. Likewise, for (t1t2,g) TG2, we set Then, for any tag pair t1t2 seen in conjunction with a member of M2, we associate an odds score Finally, if for any word pair w1w2M2 we let the set of tag pairs that correspond to all occurrences of w1w2 in MED96 be denoted by T(w1w2), we define the score In an exactly analogous manner, a scoring function may be constructed for word triples in M3. We might question the use of a straight average in equation (21). We tried a weighted average, in which the weight for score(t1t2) was the number of occurrences of w1w2 in MED96 with tag pair t1t2. This actually gave worse results. We tested the scoring functions for word pairs and word triples based on tagging in the same way the six scoring methods were evaluated before (see Evaluation Method). We found that the tagging score for word pairs produced an 11-point average precision of 0.166, while that for word triples produced an 11-point average precision of 0.167. These results are competitive with the results given for the scoring methods listed in Tables and . An important question, however, is whether these methods add significantly to the methods already presented. We attempted to improve the composite scoring method for word pairs (described under Evaluation Method) by adding some fraction of log score(w1w2) to it. We were able to improve the score only from 0.37 (bottom row ) to 0.38. Likewise, we attempted to improve the composite score for word triples. Here we were marginally more successful, improving the composite score from 0.32 (bottom row ) to 0.34. Applications : We have applied the combined scoring methods derived as described under Evaluation Method to the MED96 data set described under Data Sources and Preparation. Of the 584,315 word pairs obtained from MED96 and ranked, the top one third were selected as a set B2 of sufficient quality to warrant consideration for inclusion in UMLS. Of the 206,522 word triples obtained and ranked, the top one half were selected as a set B3 of sufficient quality to warrant consideration for inclusion in UMLS. The result is 157,867 two-word phrases in B2 and 95,006 three-word phrases in B3 that are not in UMLS. While the methods described here can serve as a screen for the extraction of useful phrases, they can also form part of a system for marking useful phrases in text. Each marked phrase may then serve as a hyperlink to other texts that contain the same phrase. To illustrate the results of such processing, we have taken the sets B2 and B3, just defined, as the candidate phrases. The text of a document is broken into segments at punctuation marks and stop words. Each segment consisting of at least two words is examined, and the highest ranked member of B2 and the highest ranked member of B3 are selected for marking. If the two selected phrases overlap, only the word triple is marked, but if they do not overlap, both are marked. In some cases there is no word triple, and then the selected word pair is marked. In other cases no phrase scores high enough to be selected. Following is the result of such marking on a sample document from MED96. Title: Impaired glucose tolerance at five-year follow-up of Young men with borderline hypertension. Abstract: Recent studies suggest that patients with essential hypertension have impaired glucose tolerance and are hyperinsulinemic compared with normotensive subjects. The aims of the study were (1) to follow blood pressures of 56 young men with borderline hypertension for 5 years, (2) to investigate glucose tolerance in these subjects, and (3) to determine the relation of insulin/ glucose metabolism to structural vascular changes and hemodynamic patterns in borderline hypertension. METHODS: Thirty-nine young (age 22-34 years) male subjects with borderline hypertension (SBP 140-160 and or DBP 85-95 mmHg initially and 17 normotensive control subjects (SBP 110-130 and DBP 60-80 mmHg) participated in the study. Blood pressure was measured, a standard oral glucose tolerance test (OGTT) was performed, and glucose, insulin and C-peptide were determined before and 30, 60, 90 and 120 minutes after a standard 75-g glucose load. Post-ischemic forearm vasodilatory responses were examined by plethysmography. RESULTS: At follow-up, the borderline hypertensives had maintained significantly higher blood pressures than control subjects. Borderline hypertensives also had significantly impaired glucose tolerance compared to control subjects. The insulin response had a somewhat more sluggish descent, but did not differ significantly from the response of normotensives. The C-peptide response pattern resembled that of insulin, but C-peptide was significantly elevated after 120 min. On the whole group level, there were only weak relations of insulin to blood pressure. By contrast, fasting insulin and post-load insulin levels were strongly correlated with body mass index, the waist-hip circumference ratio, triglyceride, and both total and LDL cholesterol. Across the whole group, there were significant correlations between forearm minimal vascular resistance and fasting insulin (r = +0.37 p = 0.007) and insulin area-under-the-curve (r = +0.28 p = 0.044). However, Rmin was even more strongly correlated with body mass index, suggesting that this relationship was related to degree of obesity. CONCLUSION: Borderline hypertension in young men is a persistent condition which is associated with impaired glucose tolerance without hyperinsulinemia. This finding suggests that impaired glucose tolerance might be a more primary phenomenon in early hypertension devoid of lipid metabolic aberrations. While the processing shown here is not perfect, it does mark most of the interesting and useful phrases that one might wish to follow as links to other documents. The main improvement that appears necessary is the elimination of throw-away phrases as "five-year follow," "Recent studies suggest," and "Thirty-nine." Here, "five-year follow" should be "five-year follow-up," but "up" is on our stop list. As a rule, any phrase that contains the word "suggest" or "suggests" should be dropped. Likewise, phrases that are numbers are not useful as links. Thus, simple rules can be added to the system to improve the processing. of particular note, syntactic parsing would appear capable of adding little to the analysis, since almost all the marked phrases appear syntactically reasonable. Discussion : There are a number of limitations of the work reported here. One of these is our definition of a phrase. We exclude stop words from phrases. Because of this limitation, we cannot detect such phrases as "vitamin A," "hepatitis A," or "cancer of the lung." There are two things that can be done to help alleviate this problem. First, we can leave the letter "A" off the stop list while processing. This will generate significantly more phrases because the letter "A" is so prevalent in the language. If this is a burden on machine memory or disk space, we can even limit the processing to just those phrases that contain the letter "A" as one of the words in the phrase. As a second step, we can take good phrases that have been identified and use them to find additional useful phrases that contain stop words. For example, if we have identified the phrase "lung cancer" as useful, we may then find "cancer of the lung" as a modified form. This could be accomplished in a general way by noting that the one is a rearrangement of the other with stop words added. It could also be accomplished, and with less error, by using a template that matches "UV" to "V of the U." We have also excluded phrases with more than three words. Relatively fewer useful phrases have four or more words in them. However, it would be useful to be able to identify phrases such as "high pressure liquid chromatography" and "left main coronary artery." We are currently examining methods by which we can extend our processing to obtain these longer phrases. A second limitation of our approach is our dependence on the UMLS. First, we would like to point out that of the individual scoring methods presented under Scoring Methods, only s4 depends for its derivation on the UMLS. The main use of the UMLS is to demonstrate that these scoring methods are effective in ranking useful phrases above non-useful phrases in a large list of phrases extracted from MED96. This is based on the assumption that the majority of phrases found in M2 and M3 are not useful and hence, if we can rank the phrases from the UMLS that occur in these lists near the top, we are succeeding in differentiating good phrases from bad. There still remains the question of the dependence of s4 and the composite scores on the UMLS. Clearly there is a dependence, for we could not define these scores without the UMLS or some large set of good phrases. The question is whether the dependence is reasonable or not. We believe this question is answered in the affirmative by the results of the cross-validation testing presented under Results. We found the composite score (which includes a contribution from s4) derived from one subset of MEDLINE, say MED1, does not lose effectiveness when applied to a disjoint subset of MEDLINE, MED2. Such results justify the application of the composite scores to new material in MEDLINE. We would, however, warn that the optimal choice of coefficients for MEDLINE might not be optimal for some other area of application. Another aspect of our treatment that deserves comment is the fact that we derive the composite scores as a log linear sum of the individual scores. This is inspired by the inherent simplicity of the approach and by the wide success of log linear models in statistics. Although there is a strong similarity, our approach is somewhat different in that we seek to optimize the resultant ranking for retrieval rather than maximize the likelihood of the data. In other words, we seek to solve a slightly different problem. There is, at least theoretically, the possibility to do better modeling if we have detailed knowledge of the dependencies between the different individual scores. However, we are not in possession of this detailed knowledge, and this leads us to follow what is, for the present, a more feasible approach. Finally, there is the question of whether some other methods of scoring may not prove useful for our task. To begin, we may ask why part-of-speech tagging does not prove more more successful in the task of identifying useful phrases. One issue is the performance of the Xerox tagger. In the initial description of the tagger, accuracy of more than 96 percent is claimed. The implementation we used was trained on medical text, has been tested and used extensively in-house, and has performed well, and we see no reason to question the figure of 96 percent. If this figure is reasonably accurate, then we could not expect to see much improvement even if we were to use a tagger with 98 percent accuracy. We believe the lack of benefit we see from tagging stems from two sources. First, two of our scoring methods are based on properties of phrases that are syntactically important. The score s3 is based on how often the phrase appears to have a potential head word as its last word, while score s4 is based on the expectation that the last word in the phrase has a three-letter suffix that would be seen in a high-quality phrase. Since the Xerox tagger uses the suffix of an unknown word in predicting its ambiguity class, there is clearly overlap in the information used by s4 and the tagger. While one could, in principle, test the level of dependence between s4 and the tagger scoring, we have not attempted to do this. The second point is that the task of identifying high-quality phrases is as much a problem of semantics as it is a problem of syntax. Tagging cannot help with the semantic problem, and thus the performance of tagging alone on the task at hand would seem to be limited. What we have said notwithstanding, there are other methods of part-of-speech tagging, that might yield different and more favorable results for our task. Other approaches that may prove useful are methods that require one to work from known good phrases to obtain related phrases that may also be of good quality. Examples include the work of Hersh et al., in which known good head words for phrases were used to locate numerous phrases built from them, and the work of Cooper and Miller in locating good phrases that are lexical variations of MeSH terms (the lexical indexing system PostDoc) or that co-occur at a high level with MeSH terms (the statistical indexing system Pindex). While we do not question the effectiveness of these approaches, we have avoided them because they imply a strong correlation between what one has already given as good phrases and what one can find with the methods. Our aim has been to accomplish a more general type of processing that would not bind us so strongly to prior knowledge. In our current and ongoing work we are examining two ways of improving the system. First, as mentioned above, we would like to allow phrases longer than three words. We are seeking to do this by examining more closely the statistical dependency between words that occur in text, the idea being that a word that occurs at the left end of a phrase may belong there if the dependency is sufficiently strong. Second, we would like to find a way to score phrases more accurately as to how laden with content or subject matter they are. Bookstein et al., have developed methods for this purpose that make use of the distribution of terms within a document. Those that are content bearing tend to be uneven in their distribution. Unfortunately, we have access only to titles and abstracts of documents and will have to take a different approach, more related to how the terms are distributed relative to other terms within the whole database. Backmatter: PMID- 10984470 TI - Bioinformatics and Clinical Informatics --The Imperative to Collaborate AB - Keywords: null: null Standard Data Models : In less than a decade, the Human Genome project (HGP) has generated a large amount of biological data that is likely eventually to lead to a qualitative change in the way in which clinical medicine (diagnostics, prognostics, and therapeutics) is practiced. A central intellectual and technologic asset to this effort has been GenBank and related genomic and protein databases (e.g., the SWISS-PROT, Exon-Intron, and IMGT databases). Their standardized data models have allowed research laboratories throughout the world to rapidly populate them with the very latest information. In turn, these databases are freely available throughout the world via the Internet and have seeded, accelerated, and inspired thousands of research projects. In contrast, there are few, if any, consequential shared national clinical databases. Specifically, patient data in one information system can only rarely be transferred to another to expedite patient care. This, despite decades of research and development of clinical record systems. This marked contrast is deceptive. The HGP has benefited from the elegant simplicity of the genetic code. In essence, at the level of primary structure, the genetic information coded by any organism is simply a sequence of characters drawn from a very limited alphabet. Consequently, there are only a very few items that GenBank requires be submitted for an entry to be a valid (and useful) component of its database. The clinical care of human beings is far more complex, requiring at the minimum a detailed record of the history of multiple clinical interventions and outcomes, relevant life history, and clinical measurements that span several modalities, from serum chemistry to brain imaging. It is not surprising that the data model required to capture all this information is extremely complex, as is evidenced by the Health Level 7 Reference Information Model. It is a remarkable tribute to the persistence of the individuals involved in these standardization efforts, that they have been able to arrive at a reasonably adequate standardized representation of not only the many descriptors but much of the process and business of clinical care. As the HGP moves from the acquisition of raw genomics data to the biological function of the discovered genes and their clinical importance, the bioinformatics community will have to address very similar complexities. That is, the clinical annotation of genomic data sets, particularly for human beings, will essentially provide the equivalent, if not identical, challenge of the creation of a comprehensive medical record. Even prior to encompassing the entirety of clinical annotation, the genomics community has faltered in developing shared and standardized data models where the simplicity of the genome no longer dominates. For example, there are several competing technologies for the massively parallel measurement of gene expression using microarrays. Some of these arrays use two probes per gene and are constructed using robotic spotting techniques. Others are constructed with oligonucleotides using photolithographic techniques. Although all these techniques measure gene expression, a widely adopted standard to represent the results across all microarray technologies has yet to emerge. The GATC proposal, for example, is a possible candidate for such a data model, but its usage is currently spotty and controversial. To clinical informaticians, this will be all too reminiscent of the challenge of creating a shared data model for laboratory results. Standardized Vocabularies : The lack of widely accepted standardized vocabularies for clinical care has greatly hampered the development of automated decision support tools and clinical research databases. The impossibility of guaranteeing that a serum sodium or systolic blood pressure has the same code or term throughout our hospital system is troublesome. Fortunately, several efforts in the private and public realm (e.g., LOINC) are addressing this issue. The National Library of Medicine has invested large resources to enable these different vocabularies to be interoperable, at least at a basic level. The same problems are not unknown in bioinformatics. Even at this early stage of the HGP, DNA sequences that were previously not known to be part of the same gene have different names and are joined in only some databases (with varying levels of confidence). As the HGP ventures into diverse areas of bioscience (as well as into the clinical area), vocabulary issues are also important. Indeed, the lack of a standardized vocabulary already arises in genomics as well, in annotations. For example, despite the fact that the basic data element of GenBank is the sequence (which has an easily standardized representation), there are diverse annotations that are very nonstandardized right now. Errors : A recent report by the Institute of Medicine highlights the immense mortality and morbidity due to medical errors. Clinical informaticians (e.g., Bates et al. and Kuperman et al.,) have been instrumental in demonstrating how automated systems can be used to reduce this error rate. These industrial processing and quality improvement techniques are not without relevance to the HGP. It is well known that the mouse genome database has been contaminated with entries of the rat genome and that the specification of 5' to 3' polarity of a gene sequence has been found to be inverted.,, And these are only some of the known errors in a very large effort. These sources of error can be reduced, as they have in many industries, by the application of increased process automation and automated interception of human error before it becomes consequential. The architectures of clinical order entry systems, designed for complex clinical enterprises to prevent erroneous and dangerous clinical behavior, can inform the design of genome sequencing and expression profiling processes to prevent the kind of errors we are already finding in genomic databases. Noise : It is well known that the noise in clinical measurements leads to erroneous decision making. The archetypal example is in the intensive care unit, where multiple physiologic monitors each has its own alarm module. Because of the noisy nature of the biological signals that are monitored, the alarms are ignored or switched off because of their high false-positive rate. The noisy nature of the monitored signals thus has a significant impact on the provision of care and the decision-making ability of care providers who are working under conditions of uncertainty and data overload. Similar noise considerations arise in genomics. For example, with gene microarrays, we can measure the expression of tens of thousands of genes at a time. There are several sources of noise in these measurements: within a microarray, across microarrays, and from the intrinsic variability of the biological systems being measured. Yet in 1999, several reports, which appeared in scientific journals of the first rank, included changes in expression so small as to be indistinguishable from noise. Such changes are, in essence, a false-positive result. These false positives are potentially extremely costly. A biological researcher might decide to invest several months investigating a gene's regulation because a microarray experiment showed it to be increased or decreased under a particular set of conditions. In clinical informatics there is a rich literature of the techniques that can be used to identify false positives and reduce noise (e.g., filtering, signal fusion,,,,). Many of these techniques are transferable to the genomic domain. Privacy : In 1997, the Institute of Medicine reported significant lacunae in both technology and policy in protecting confidential patient data. Among the problems of greatest concern that were emphasized by this report were the relatively unrestricted access by third parties to these data for secondary uses and the inadequacy of the anonymization process (in both practice and theory,). Subsequently, the clinical informatics community has developed several model confidentiality policies and cryptographic identification systems. As the fruits of the HGP are translated first into clinical research protocols and then into clinical practice, personally identifiable genomic data will find their way into some form of information system. The challenges posed to the security and privacy of such data will dwarf any encountered to date with conventional clinical data. The reasons are twofold: First, genomic information is likely to be much more predictive of current and future health status than most clinical measurements. Second, with very few exceptions, an individual's genome is uniquely identifying. This identifiability is much more reliable, persistent, and specific than typically cited identifiers, including a person's name, social security number, date of birth, and address. At the very least, the architects of information systems storing genetic data should learn from all the mistakes of and designs developed for the security architectures and privacy policies of conventional clinical information systems. Conversely, the extreme concerns posed by the storage of personal genetic data is likely to generate new policies and security architectures that will enhance the confidentiality of clinical information systems. Moreover, when personal genetic data becomes incorporated into routine medical practice, the safeguards for the confidentiality of the medical record will be crucial to the confidentiality of the genetic data referenced there. Costs of Acquiring Data : "Getting the data in" has often been cited by authorities in clinical informatics as being among the most difficult challenges in successfully deploying clinical information systems. In particular, the costs of acquiring detailed and structured data from the clinical care process have been daunting. Voice and handwriting recognition information systems have not been broadly adopted, because of a variety of performance and usability issues. The cost and practicability issues will continue to present obstacles to clinical information system utility and deployment until better solutions are arrived at. In contrast, the HGP has managed to achieve significant economies of scale in sequencing technology. Gene microarrays alone have dropped in cost by a factor of two in just the last year. Here again, once genomic investigators attempt to bridge the gulf from purely genomic data sets to phenotypically (i.e., clinically) annotated data sets, they will be confronted with the same challenges of clinically oriented, codified data acquisition. The questions of which user interfaces are the most cost efficient, reliable, and generalizable to multiple clinical domains are among the implementation and design challenges that they will face. Although they have yet to arrive at definitively successful answers, clinical informaticians have already completed several decades worth of engineering and ethnographic studies addressing the very same questions. Extracting Knowledge from Data : The first rough draft of the human genome was reported to have been completed in May of this year. It is likely that a complete, high-quality human DNA reference sequence will be available by 2003. Yet the function of the vast majority of the genes in the human genome will be unknown. The minority of genes with documented function are likely to have many more functions and interactions that are unknown. Consequently, one of the primary applications of information technologies in genomics is the application of machine learning techniques to determine how genes are functionally interdependent and how these interdependencies are reflected in the biological and clinical behavior of the system in which they operate., Many of these machine learning techniques were previously applied to the task of extracting knowledge from clinical databases, and some were even developed first in the clinical domain.,,,,,,,, To be sure, the genomic era has challenged these machine learning techniques to the extreme, because of the high dimensionality of data sets (e.g., tens of thousands of measurements per experiment) and the relatively few cases and experiments from which investigators are attempting to glean knowledge. Summary : Without being exhaustive, this brief review suggests the multiple points of commonality between the genomic and clinical strands of the biomedical informatics research agenda. It also suggests that the training of investigators in informatics should include a set of core competencies that at least cover these common points. In this fashion, the joint research agenda might be well served to the mutual benefit of biomedical science and clinical care. The Stanford Medical Informatics educational program, described in this issue, illustrates this benefit. Backmatter: PMID- 11882257 TI - What kind of evidence is it that Evidence-Based Medicine advocates want health care providers and consumers to pay attention to? AB - Abstract | Background | In 1992, Evidence-Based Medicine advocates proclaimed a "new paradigm", in which evidence from health care research is the best basis for decisions for individual patients and health systems. Hailed in New York Times Magazine in 2001 as one of the most influential ideas of the year, this approach was initially and provocatively pitted against the traditional teaching of medicine, in which the key elements of knowing for clinical purposes are understanding of basic pathophysiologic mechanisms of disease coupled with clinical experience. This paper reviews the origins, aspirations, philosophical limitations, and practical challenges of evidence-based medicine. Discussion | EBM has long since evolved beyond its initial (mis)conception, that EBM might replace traditional medicine. EBM is now attempting to augment rather than replace individual clinical experience and understanding of basic disease mechanisms. EBM must continue to evolve, however, to address a number of issues including scientific underpinnings, moral stance and consequences, and practical matters of dissemination and application. For example, accelerating the transfer of research findings into clinical practice is often based on incomplete evidence from selected groups of people, who experience a marginal benefit from an expensive technology, raising issues of the generalizability of the findings, and increasing problems with how many and who can afford the new innovations in care. Summary | Advocates of evidence-based medicine want clinicians and consumers to pay attention to the best findings from health care research that are both valid and ready for clinical application. Much remains to be done to reach this goal. Keywords: Background : Evidence-Based Medicine (EBM) is based on the notion that clinicians, if they are to provide, and continue to provide, optimal care for their patients, need to know enough about applied research principles to detect studies published in the medical literature that are both scientifically strong and ready for clinical application. This opportunity for continuing to improve the quality of medical care stems from the huge ongoing public and private investment in biomedical and health research. The challenges in applying new knowledge, however, are considerable, and EBM does not address them all. Two that EBM tries to address are as follows. First, the advance of knowledge is incremental, with many false steps, and with breakthroughs few and far between, so that only a very tiny fraction of the reports in the medical literature signal new knowledge that is both adequately tested and important enough for practitioners to depend upon and apply. Second, practitioners have limited time and little understanding of research methods. To help practitioners meet these challenges, EBM advocates have created procedures to objectively identify and summarize evidence as it accumulates on clinical topics, and resources that allow users to find the current best evidence when and where it is needed for decisions concerning health and health care . This paper reviews the origins, aspirations, philosophical limitations, and practical challenges of evidence-based medicine. Discussion : The history and precepts of EBM | Evidence-Based Medicine (EBM), the term and current concepts, originated from clinical epidemiologists at McMaster University . Although the term has been adopted by many disciplines and adapted to their use (eg, as Evidence-Based Nursing, Evidence-Based Clinical Practice, Evidence-Based Pharmacy, and so on), the objectives of these congeners are the same and I will use the generic term in this essay. EBM advocates want patients, practitioners, health care managers and policy makers to pay attention to the best findings from health care research that meet the dual requirements of being both scientifically valid and ready for clinical application. In doing so, EBM advocates proclaimed a new paradigm and seemingly pitted EBM against the traditional knowledge foundation of medicine, in which the key elements are understanding of basic mechanisms of disease coupled with clinical experience. The latter is epitomized by the individual authority ("expert"), or, better still, collective medical authority, such as a panel of experts convened by a professional society to provide practice guidelines based on collective expert opinion. EBM claims that experts are more fallible in their recommendations (of what works and what doesn't work in caring for patients) than evidence derived from sound systematic observation (that is, health care research). This is especially so during recent decades when applied research methods have been developed for observation and experimentation in increasingly naturalistic and complex clinical settings. Furthermore, because applied research methods are based on assessing probabilities for relationships and the effects of interventions, rather than underlying mechanistic explanations, EBM posits that practitioners must be ready to accept and deal with uncertainty (rather than seeking the reductionist allure of basic science), and to acknowledge that management decisions are often made in the face of relative ignorance of their underlying nature or true impact for individual patients. A fundamental assumption of EBM is that practitioners whose practice is based on an understanding of evidence from applied health care research will provide superior patient care compared with practitioners who rely on understanding of basic mechanisms and their own clinical experience. So far, no convincing direct evidence exists that shows that this assumption is correct. Nevertheless, the New York Times Magazine "Year in Review" included EBM as one of the most influential ideas of 2001 . Basic versus applied health research | Scientific approaches to studying health care problems developed at a leisurely pace until the end of World War II when some of the public funding that had been dedicated to killing was reallocated to saving lives through health research. Initial investments were directed firstly to basic research, to better understand the determinants and pathophysiology of disease, and medical schools reflected this stage of development in their teaching of the basic sciences of biology, pathology, physiology and biochemistry, as the foundation of medical knowledge. Increasing shares of investment were then allocated to the development and applied testing of innovations in clinical settings. Although these applied research methods were initially rooted in the observational techniques of epidemiology, clinical epidemiologists such as Archie Cochrane in the UK, Alvan Feinstein in the US, and David Sackett in Canada, pioneered and legitimized the use of experimentation in clinical settings, leading to the randomized controlled trial becoming the hallmark of testing. It is important to recognize that experimental designs were added to observational designs, not substituted for them. Different methods, observational or experimental, are needed for exploring different questions. The first trial in which randomization was formally described and applied was published in the British Medical Journal in 1948 and heralded a new era of antibiotic treatment, streptomycin for tuberculosis. Today, methodologies from other scientific disciplines have been added. For example, nonexperimental and qualitative research methods have been adopted from the social sciences. Thus, the research methods of medical science are pluralistic and expanding, driven by attempts to address a broader range of questions, and undoubtedly by the priority that people place on personal health, the obvious benefits that biomedical research has already brought, and the prospect that these benefits are just the beginning. EBM does not clearly address the role of basic science in medical discovery, except to indicate that, in most circumstances of relevance to individual patient care, basic science alone does not provide valid and practical guidance. There are some exceptions: certain deficiency disorders, such as type 1 diabetes mellitus, for example. However, even though basic science provides definitive evidence that insulin deficiency is the underlying problem in this disorder, determining which of many possible ways of delivering exogenous insulin therapy results in the best care for patients has required myriads of applied research studies, with clear evidence concerning the benefit of multiple dose insulin regimens coming less than a decade ago . In many such situations, empirical solutions, tested by applied research methods, are "holding the fort" until basic understanding -- of mechanisms and interventions -- is forthcoming. This will continue to be the case for the foreseeable future, the marvelous advances in genomics notwithstanding. This schism between basic and applied research is, however, more rhetoric than reality. Rather, basic and applied research are different ends of a spectrum of health research, progressing from "bench" to "bedside". The best applied research studies are often founded on excellent basic science findings, even if basic research is neither necessary nor sufficient for the management of most medical problems. This is because applied research is a complementary way of knowing, not a participant in a scientific turf war competing to be the best way of knowing. Nevertheless, from a pragmatic, clinical focus, applied research provides evidence to practitioners and patients that is often better suited for the specific problems they must deal with. Confusion between the objectives of science and those of the practice of medicine has perhaps led to much of the misunderstanding and criticism leveled at EBM. An example of the interplay between basic and applied clinical research | An example illustrates the complex relationship between basic and applied research, and in turn, between both of these and clinical practice. Narrowing of the arteries to the front part of the brain (the internal carotid artery and its tributaries, the anterior and middle cerebral arteries) is associated with stroke, in which a part of the brain dies when it loses its blood supply. Narrowings of the internal carotid artery above the level of the neck can be bypassed through connecting the superficial temporal artery (STA), on the outside of the head, with a branch of the middle cerebral artery (MCA), just inside the skull. STA-MCA bypass (also known as extracranial-intracranial (EC/IC) bypass) is an elegant (and expensive) surgical procedure that is both technically feasible in a high proportion of cases and leads to increased blood supply to the part of the brain beyond the narrowing of the ICA. This increased blood supply was thought, on physiological grounds, to be exactly what the brain needed to prevent future strokes in people who had had minor strokes in this vascular distribution. About 200 reports of case series of patients undergoing STA-MCA bypass were reported in the medical literature up to 1985, almost all of them interpreted by their surgeon authors as indicating benefits for patients. In these case series studies, patients are described before and after undergoing the procedure, and sometimes compared with findings in previous reports ("historical controls"), with and without the procedure. In 1985, a large randomized controlled trial was reported . This showed no reduction in the subsequent rate of stroke with the procedure when compared with the rate for patients who did not have the procedure. On further analysis, it was found that patients with STA-MCA bypass who had higher rates of blood flow were actually worse off, and that surgery blunted the natural rate of recovery from the initial stroke that led to selection of patients for surgery . Dissemination of these findings was rapid and led to the elimination of this procedure for attempting to prevent stroke recurrence. Subsequently, randomized controlled trials were conducted for patients with narrowing of the internal carotid artery in the neck. Removing this narrowing surgically, a procedure called carotid endarterectomy, had been practiced for longer than STA-MCA bypass at the time but had not been adequately tested in controlled trials. Its use was brought into question because of the negative findings of the STA-MCA bypass trial, which appeared to undermine the physiological rationale for the procedure, namely, that opening a partially blocked internal carotid artery would reduce the risk for subsequent stroke. As it happens, several randomized controlled trials of carotid endarterectomy that ensued showed that it has substantial benefit for symptomatic patients with severe narrowing of the carotid artery, but not for those who had mild narrowing or who had no symptoms associated with the narrowing . These trials of STA-MCA bypass and carotid endarterectomy have led to better understanding of the basic mechanisms of stroke, elimination of a harmful surgical procedure, promotion of another procedure, and provision of evidence for tailoring the findings to individual patients . These advances in knowledge have benefited many patients. Unfortunately, surveys of patient care also show that some patients receive endarterectomy when they are unlikely to benefit from it, while others who might benefit are not offered it . In fact, there are numerous examples of underapplied evidence of both the benefits and harms of treatments . Eliminating this mismatch between who could benefit and who is offered health care interventions is the prime objective of EBM. The nuts and bolts of EBM | A current definition of EBM is "the explicit, judicious, and conscientious use of current best evidence from health care research in decisions about the care of individuals and populations" . A more pragmatic definition is a set of tools and resources for finding and applying current best evidence from research for the care of individual patients. This practical definition reflects the fact that there are now many information resources in which evidence from health care research has been pre-graded for validity by people with expertise in research methods, and, better still, also assessed by experienced practitioners for clinical relevance. Thus, the user's task is changing from the largely hopeless one of reading the original medical literature to find out about current best care, to one of finding the right pre-assessed research evidence, judging whether it applies to the health problem at hand, and then working the evidence into the decision that must be made. Grades of the quality of evidence are derived from scientific principles of epidemiology and its offspring, clinical epidemiology. The grades are based on several notions, the most elementary of which are as follows. First, studies that take more precautions to minimize the risk of bias (for example, through using reliable and valid measures of health care outcomes) are more likely to reveal useful truths than those that take fewer precautions. Second, studies based in patient populations that more closely resemble those that exist in usual clinical practice are more likely to provide valid and useful information for clinical practice than studies based on organisms in test tubes, creatures in cages, very select human populations, or unachievable clinical circumstances (such as extra staff to provide intensive follow up, far beyond the resources in usual clinical settings). Third, studies that measure clinical outcomes that are more important to patients (eg, mortality, morbidity and quality of life, rather than liver enzymes and serum electrolytes) are more likely to provide evidence that is important to both practitioners and patients. Simple guidelines for critical appraisal of health care research evidence are widely available in print [for example, ] and on the internet (for example, ). Optimal study designs differ for determining the cause, course, diagnosis, prognosis, prevention, therapy and rehabilitation of disease, so the rules for assessing validity differ for these different questions. For example, randomized allocation of participants to an intervention and control group is held to be better than non-random allocation for controlling bias in intervention studies. This is not merely a matter of logic, common sense or faith: non-random allocation usually results in more optimistic differences between intervention and control groups than does random allocation . Similarly, in observational study designs for assessing the accuracy of diagnostic tests, independent interpretation of the tests that are being compared is known to result in less optimistic reports of test performance . Although most guidelines for critical appraisal are not comprehensive or fully rigorous, they provide an effective filter for the reliability and validity of health care research that screens out about 98% or more of the medical research literature as not being ready for clinical use . Of those studies that make it through the filter, systematic reviews provide the firmest base for the application of evidence in practice ; the past decade has seen the Cochrane Collaboration forging a world wide effort to summarize evidence concerning the effects of health care interventions . Once individual studies have been assembled and graded for quality, the collected evidence then can be used to make recommendations for practice, preferably with each recommendation being labeled according to the level of evidence that supports it. Various systems for indicating the level of evidence for collected evidence are available, for example, from the Centre for Evidence-Based Medicine in Oxford and in EBM books . Some difficulties with the term Evidence-Based Medicine | Many objections to EBM are based on the notion that it advocates cook-book medicine, that is, treating patients strictly according to a formula or algorithm derived from a research study. In fact, this was never intended by the advocates of EBM, but it was perhaps not initially clearly emphasized that evidence from research can be no more than one component of any clinical decision. Other key components are the circumstances of the patient (as assessed through the expertise of the clinician), and the preferences of the patient . Just how research evidence, clinical circumstances, and patients' wishes are to be combined to derive an optimal decision has not been clearly stated, except that "clinical judgment and expertise" are viewed as essential to success . Figure 1 | Basic elements of clinical decision making Basic elements of clinical decision making Even more problematic, the term evidence is commonly used for many types of evidence of relevance to clinical practice, not just health care research evidence. For example, clinicians collect evidence of patients' circumstances and wishes. Thus, it is hardly surprising that the term evidence-based medicine is confusing to many, who do not appreciate that its evidence is narrowly defined as having to do with systematic observations from certain types of research. The very name has been an impediment to getting across its main objective, namely, that health care research is nowadays producing important results that, if applied, can benefit patients more than the treatments that clinicians are experienced in recommending. Using the technical definition of EBM, evidence from heath care research is a modern, never-before-available complement to traditional medicine. Perhaps a better name would be "certain-types-of-high-quality-and-clinically-relevant-evidence-from-health-care-research-in-support-of-health-care-decision-making"...an accurate but mind-numbing descriptor. Philosophical issues -- from a lay perspective | The main original paper on EBM proposed EBM as a paradigm shift, based on Thomas Kuhn's definition of paradigms: ways of looking at the world that define both the problems that can be legitimately addressed and the range of admissible evidence that may bear on their solution. Nevertheless, it is fair to say that not very much attention was paid by the originators of EBM to the philosophy of science. It is also easy to agree with Alan Chalmers that most scientists and EBM advocates are ignorant of the philosophy of science and give little or no thought to constructing a philosophical basis for their activities. According to Guba and Lincoln , in the basic science that underpins traditional medicine, the workings of the human body and basic mechanisms of disease can be discovered by observations of an individual human or organism using instruments that are objective and bias free. These mechanisms then can be discerned by inductive logic and known to a certainty. By contrast, applied research deals with more complex phenomena than disease mechanisms; often relies on experimentation rather than (just) observation; recognizes that observations of complex phenomena can be biased and takes measures to reduce bias; has groups of patients as the basis of observation; uses probabilities to judge truth, rather than expecting certainty; and uses deductive and Bayesian logic to progress. Certainly, there are differences between the approaches of basic and applied research, but are they mutually exclusive, as in a paradigm shift, or complementary ways of knowing, as in a pluralistic version of epistemology? The latter view seems to be more tenable. The expectation of EBM that doctors should keep abreast of evidence from (certain-types-of-health-care-) research raises many issues. First, what is "valid" health care research? Second, what are the "best" findings from this research? Third, when is health care research "ready" for application? Fourth and fifth, to whom and how does one apply valid and ready evidence from health care research? EBM provides a set of increasingly sophisticated tools for addressing these questions, but, at present, the result is only partly as good as EBM advocates hope it will become. Meanwhile there is much to criticise about EBM from both philosophical (AV Kulkarni, personal communication, 2000) and practical perspectives. For example, it is difficult to be smug about the superiority of the research methods advocated by EBM when the results of studies that are similar methodologically not infrequently disagree with one another. Meanwhile, it has been shown that the findings of observational studies agree more often than not with the findings of allegedly more potent RCTs . While holes can be picked in these arguments against the ascendancy of RCTs , there is no way to win the argument without a universal standard of truth. Furthermore, the issue of when a research finding is ready for clinical application remains mired in the lack of a satisfactory resolution of how findings from groups can be applied to individuals. For one thing, our understanding of how to determine what patients want is primitive. Also problematic, the circumstances in which patients are treated can vary widely from location to location (including locations that are right across the street from one another): the resources, expertise and patients are often quite different and the same research evidence cannot be applied in the same way, or not at all. Finally, we do not have convincing studies showing that patients whose clinicians practise EBM are better off than those whose clinicians do not practise EBM: no one has done a randomized controlled trial of EBM with patient outcomes as the measure of success. Such a trial would be impossible to do as the control group could not be effectively isolated from the research that EBM is attempting to transfer, and it would be regarded as unethical to attempt to do so. This situation is unfortunate in the sense that, even if it is accepted that current research is generating valuable findings for health care, there are many questions about whether the EBM movement is doing anything useful to accelerate the transfer of these findings into practice. Nevertheless, we do have limited evidence that the concepts of EBM are teachable . David Hume and his followers took pains to point out the differences between is and ought. The is of EBM is that science is producing new and better ways of predicting, detecting and treating disease than were imaginable at the middle of the past century. The ought of the EBM movement, which annoys many practitioners, and would perturb Hume and his followers, is that EBM advocates believe that clinicians ought to be responsible for keeping up to date with these advances and ought to be prepared to offer them to patients. Thus, EBM has taken on the tones of a moral imperative. But it is premature to get very preachy about the ought of EBM, not that this has stopped EBM's more ardent advocates. Worse still, the interventions that advocates of EBM insist ought to be provided in all appropriate individual circumstances would undoubtedly have some important adverse effects. For one, full implementation would cost much more than the resources currently available for health care, even accounting for some cost effective innovations and deletion of existing but ineffective practices. The increased costs of care would lead to unaddressed (let alone resolved) dilemmas in distributive justice. Second, interventions that save lives and reduce suffering in the short term may end up prolonging life beyond the point of senescence and misery. EBM advocates try to ameliorate the latter problem by declaring that patients' values ought to be incorporated into clinical decisions, but without assuring that we know how to do this. Indeed, there is a continuing tension here between the consequentialist, population-based origins of epidemiology (doing the greatest good for the greatest number), which generates most of the best evidence that EBM advocates hope to convince practitioners and patients to pay attention to, and the deontological or individualistic approach of medicine, doing the greatest good for the individual patient, which practitioners are sworn to do. Although some components of EBM have been derided as representing ultrapragmatic utilitarianism, EBM doesn't offer a credible solution to this tension, nor even take a clear stance on it, perhaps reflecting the dual origins of many EBM advocates: most of the leaders are trained in both epidemiology and a clinical discipline, and do both research and clinical practice. In weighing the philosophical issues raised by EBM, many epistemological issues certainly merit intense discussion. However, it is the ethical issues that I believe to be of highest concern. Will the proceeds of the new science of medicine be fairly distributed in society? Given the already stupendous and wildly escalating costs of health care, driven particularly by newer diagnostic and therapeutic interventions, how can resources be optimally and fairly allocated within the health care sector and across all sectors of public expenditure? Can the long-term consequences (for example, unproductive and miserable longevity) of the short-term gains that are regularly documented by health care research continue to be ignored? How can patient's wishes be informed, determined and taken into account in health care decision making? Should some of the funds for health research be diverted into some other sector (continuing education?) so that the health care system can catch up to the current state of knowledge? Is EBM a waste of time if we lack adequate understanding of practical methods of changing practitioner and patient actions ? One hopes that the attention of philosophers will be drawn to these questions, as well as to the continuing debate about whether EBM is a new paradigm and whether applied health care research findings are more valid for reaching practical decisions about health care than basic pathophysiological mechanisms and the unsystematic observations of practitioners. Summary : Evidence-Based Medicine has evolved substantially from its origins a decade ago, becoming less pretentious and more practical. Nonetheless, it must continue to evolve and address several important issues that will otherwise limit its value as an adjunct to health care decisions. Pressing matters include agreement on what constitutes "best" evidence; appropriate generalization beyond research projects; accurate and efficient communication with practitioners, patients and policy makers; and moral issues including distributive justice and individual autonomy. Given the substantial investment of society and commerce in fundamental and applied health research, and the high expectations of society for reducing the burden of illness, attention to these matters should have high priority. Competing interests : The author is one of the originators of the concepts of Evidence-Based Medicine, and is the developer and editor of a number of evidence-based publications, including ACP Journal Club, acpjc.org and Best Evidence; co-editor for Evidence-Based Medicine; coordinating editor for Evidence-Based Mental Health, Evidence-Based Nursing; associate editor for WebMD's Scientific American Medicine; and advisory board member for Clinical Evidence. Backmatter: